(Recent Advances in Phytochemistry 33) Gordon M. Cragg - Michael R. Boyd (Auth.) - John T. Romeo (Eds.) - Phytochemicals in Human Health Protection - Nutrition - and Plant Defense-Springer US (1999)

recent advances in phytochemistry
volume 33
Phytochemicals in
Human Health
Protection, Nutrition,
and Plant Defense
RECENT ADVANCES IN PHYTOCHEMISTRY
Proceedings of the Phytochemical Society of North America
General Editor: John T. Romeo, University of South Florida, Tampa, Florida
Recent Volumes in the Series:
Volume 24 Biochemistry of the Mevalonic Acid Pathway to Terpenoids

Proceedings of the Twenty-ninth Annual Meeting of the Phytochemical
Society of North America, Vancouver, British Columbia, Canada, June,
1989
Volume 25 Modern Phytochemical Methods

Proceedings of the Thirtieth Annual Meeting of the Phytochemical Society
of North America, Quebec City, Quebec, Canada, August, 1990
Volume 26 Phenolic Metabolism in Plants

Proceedings of the Thirty-first Annual Meeting of the Phytochemical
Society of North America, Fort Collins, Colorado, June, 1991
Volume 27 Phytochemical Potential of Tropical Plants

Proceedings of the Thirty-second Annual Meeting of the Phytochemical
Society of North America, Miami Beach, Florida, June, 1992
Volume 28 Genetic Engineering of Plant Secondary Metabolism

Proceedings of the Thirty-third Annual Meeting of the Phytochemical
Society of North America, Pacific Grove, California, June-July, 1993
Volume 29 Phytochemistry of Medicinal Plants

Proceedings of the Thirty-fourth Annual Meeting of the Phytochemical
Society of North America, Mexico City, Mexico, August, 1994
Volume 30 Phytochemical Diversity and Redundancy in Ecological

Interactions
Proceedings of the Thirty-fifth Annual Meeting of the Phytochemical
Society of North America, Sault Ste. Marie, Ontario, Canada, August,
1995
Volume 31 Functionality of Food Phytochemicals

Proceedings of the Thirty-sixth Annual Meeting of the Phytochemical
Society of North America, New Orleans, Louisiana, August, 1996
Volume 32 Phytochemical Signals and Plant-Microbe Interactions

Proceedings of a jOint meeting of the Phytochemical Society of North
America and the Phytochemical Society of Europe, Noordwijkerhout, The
Netherlands, April, 1997
Volume 33 Phytochemicals in Human Health Protection, Nutrition, and

Plant Defense
Proceedings of the Thirty-eighth Annual Meeting of the Phytochemical
Society of North America, Pullman, Washington, July, 1998
A Continuation Order Plan is available for this series. A continuation order will bring delivery of each new
volume immediately upon publication. Volumes are billed only upon actual shipment. Forfurther information
please contact the publisher.
recent advances in phytochemistry
volume 33
Phytochemicals in
Human Health
Protection, Nutrition,
and Plant Defense
Edited by
John T. Romeo
University of South Florida
Tampa, Florida
Springer Science+Business Media, LLC

Library of Congress Cataloging-in-Publication Data
Phytochemicals in human health protection, nutrition and plant delense

/edited by John T. Romeo.
p. cm. -(Recent advances in phytochemistry; v. 33)
Includes bibliographical relerences and index.
ISBN 978-1-4613-7123-6 ISBN 978-1-4615-4689-4 (eBook)
DOI 10.1007/978-1-4615-4689-4
1. Materia medica, Vegetable. 2. Pharmacognosy. 3. Plan!
bioactive compounds. 4. Phytochemistry. 1. Romeo, John T.
II. Series.
QK861.R38 voI. 33
[RS164]
572'.2 s-dc21
[615' .32] 99-37365
CIP
Cover:
Flax seed. The richest natural source 01 the plant lignan, secoisolariciresinol diglycoside
(SOG), translormed to cancer-protecting mammalian lignans by intestinal bacteria.
(Courtesy 01 Lilian U. Thompson)
Intsia bijuga. Heartwood. Cross section 01 vesicles, one 01 which (lighter color) is lull 01
pentahydroxy Ilavonoid, robinetin. Oarker adjacent vessel contains a mixture 01 other
Ilavonoids and tri- and tetra-hydroxy stilbenes. Compounds are assumed to lunction in
decay resistance (Courtesy 01 W. E. Hillis).
Proceedings 01 the 38th Annual Meeting 01 the Phytochemical Society 01 North America
on Phytochemicals in Human Health Protection, Nutrition, and Plant Oelense, held July
26-31,1998, in Pullman, Washington
ISBN 978-1-4613-7123-6
© 1999Springer Science+Business Media New York
Originally published by Kluwer Academic / Plenum Publishers in 1999
Softcover reprint of the hardcaver Ist editian 1999
AII rights reserved

No part 01 this book may be reproduced, s!ored in a retrieval system, or transmitted in
any lorm or by any means, electronic, mechanical, photocopying, microlilming, recording,
or otherwise, wi!hout written permission lrom the Publisher
To G. H. N. Towers, phytochemical pioneer.
PREFACE
Since 1994, the Phytochemical Society of North America has devoted its
annual symposia to topics with biological perspectives. Our last four volumes
have dealt with medicinal plants (1994), plant/insect interactions (1995), food
phytochemicals (1996), and plant/microbe interactions (1997), respectively.
The Symposium held in Pullman, Washington, July 26-31, 1998 brought
many aspects of these previous symposia once again to the forefront. This
time, however, there was greater emphasis on the potential applications of
phytochemistry to the diverse topics of human health and nutrition and plant
defense. As we learned about innovative uses of molecular biology as it is being
applied to these topics, we were reminded once again of the biochemical
foundation on which these advances rest. On the occasion of the 75 th birthday of
G.H. Neal Towers, which we were privileged to celebrate, a perspective of
where we began and how far we have advanced was made patently real for those
in attendance.
The papers assembled in this volume were presented during the Sympo-
sium. Roughly grouped under three broad topics, they include: I. Drug Discov-
ery and Pathway Engineering toward New MedicinallNutriceutical Targets
(papers by Cragg, Croteau, Thompson, Costa, McLaughlin, Dixon, and Matern),
2. Roles for Polyphenols-Biosynthesis and Applications (Gross, Hillis, Haslam,
and Ferreira), 3. New Chemical Prospects and Plant Defense (Asakawa, Selmar,
Houghton, and Mizutani).
It is estimated that 80% of the world's population relies on traditional med-
icines for primary health care. For the rest of the world, about 114 of all pre-
scriptions contain plant extracts or active principles derived from higher plants.
The role of the U.S. National Cancer Institute in the development of new drugs
is reviewed by Cragg. A road map through this agency is provided along with
current protocols and collaborative procedures. Plant natural products continue
to be major players in cancer treatment, accounting for up to 62% of commer-
cially available drugs. Successes and prospects are discussed. Currently, approx-
imately 1,000 Pacific yew trees must be harvested to produced a kilogram oftaxol,
used for treating breast, ovarian, and other cancers. Treatment for ovarian cancer
Vll
viii PREFACE
alone would consume 90,000 trees annually. Walker and Croteau have focused
their work on elucidating the biosynthesis of taxol and its related taxoids. They
use a sequential approach, which examines the enzymes catalyzing each trans-
formation, one which involves both in vivo and in vitro studies and molecular
genetics. Ultimately, this information will be utilized to engineer biological
systems to improve yields of taxoids.
The anticancer effects of lignans, particularly effective against mammary,
colon, skin, and prostate cancers, are discussed by Thompson. Diets high in flax,
the richest source, contain plant lignans that are acted upon by bacterial flora in
the colon of humans to produce enterodiol and enterolactone, the putative pro-
tective agents. Epidemiological and animal studies indicate that both hormone-
related (weak or antiestrogenic effects) and non-hormone-related (antioxidant,
antiangiogenic) mechanisms may be responsible. Costa and colleagues demon-
strate that the major enzymatic steps for synthesizing these protective lignans
are now understood. There is potential for using gene transfer and biotechnolog-
ical manipulation of regulatory enzymes to enhance levels of the beneficial
compounds.
McLaughlin and Chang update their studies of Annonaceous acetogenins.
Since 1993, by using relatively simple bioassays, over 200 bioactive compounds
have been described and evaluated. Structural/activity studies have identified
. compounds having both antitumor and pesticidal effects. Acetogenins inhibit
mitochondrial electron transport systems, and, as a consequence, they are espe-
cially toxic to multiple drug resistant tumor cells and pesticide resistant insects
that possess ATP-dependent xenobiotic efflux systems. In another paper which
links plant defense and human health protection, Dixon and colleagues show how
molecular biology can answer definitively such questions as the efficacy of phy-
toalexins. By cloning and manipulating genes, not only can plant resistance to
disease be improved, but added human health benefits are likely. Antiestrogens,
which are chemopreventive for breast cancer, and soy, which is correlated with
decreased prostrate cancer, may also have hormonelike activity in plants. Matern
also emphasizes structure/activity relationships in his discussion of the medici-
nal potential and biosynthesis of coumarins, compounds isolated from many plant
and microbial sources. Coumarin roles as anti-HIV inhibitors and as treatment
for skin disorders are among those singled out. He details recent developments
in biosynthesis and localization, and points out that cladistic analysis of DNA
sequences should lead to classification of enzymes, and mutational studies to the
identification of domains responsible for catalytic specificities. Biotechnological
generation of plants with modified coumarin biosynthesis and future medical
applications are probable.
The importance of phenolic compounds and their potential for human bet-
terment was apparent in several papers. The widely accepted view that tannins
function as insect feeding deterrents was both reinforced and challenged by
Gross. The diversity of tannins, and our recognition of new emerging roles for
PREFACE ix
them, such as viral inhibitors, which enable some insects to thrive by feeding on
phenolic containing plants, forces us to rethink long-held ideas concerning their
functions as feeding deterrents. Similarly, we realize their vast potential for
medical applications. Extensive enzyme studies have both identified metabolic
intermediates and have provided tools for the elucidation of biochemical reac-
tion mechanisms. There is a need for cellular localization studies, information
on transport vehicles, knowledge of final deposition sites, and application of
molecular biology for insight into the unique structures of tannin-synthesizing
enzymes. Hillis examines an often neglected source of natural products, the heart-
wood, a major feature of which is increased formation of secondary metabolites,
especially phenolics. He examines the anatomy of heart wood and its associated
tissues, and also the mechanisms by which compounds are selectively con-
centrated there. Potential fruit of such studies could lead toward more efficient
syntheses of complex natural substances.
The health-promoting effects of tea, fruit juices, and red wine are at least
partially attributable to polymeric proanthocyanidins. The chemical work of
Ferreira and colleagues demonstrates how conformation analysis is leading to an
understanding of the complexation of these compounds with other biomolecules,
and thus, a better understanding of their salutary effects. Haslam and colleagues
discuss astringency and polyphenollprotein interactions. Structure/activity rela-
tionships, the role of water, and hydrophobic effects and interactions are
described. While admitting that structural problems still remain to be solved, they
call for increasing emphasis on studies ranging from metabolism through ecology,
to practical and applied problems.
A group of papers focus more directly, but not exclusively, on plant defense.
Asakawa writes on the phytochemistry ofbryophytes. An imposing array of active
terpenoids and aromatic compounds are present in Liverworts with about 80% of
the sesqui- and diterpenoids being enantiomers of those found in higher plants.
Only about 5% of bryophytes have been studied chemically. Many are distin-
guished by characteristic odors and taste, pungency and bitterness, antimicrobial
and antifungal activity, insect antifeedant activity, nematocidal and piscicidal
activity, and plant growth inhibitory activity. Increasing medical applications are
being discovered, ranging from cytotoxic and anti-HIV activity to possible
chemoprevention of osteoporosis and allergy. Similarly, interest in the genus Bud-
dleja is relatively recent. A number of uses of extract of this plant, which include
wound healing, treatment for liver and bronchial complaints, and antifungal,
antibacterial, analgesic, and sedative effects are known. Houghton and Mensah
show that an understanding ofthe role ofthe plant's contained iridoids, f1avonoids,
and phenylethanoids is emerging. Selmar summarizes the biology of cyanogenic
glucosides and related nutritional problems. The idea that high concentrations
in certain plants function as repellents is generally accepted, but the role of low
concentrations, widespread in most food plants, is less understood. A putative
function as a metabolic mediator involved in signaling and influencing plant
x PREFACE
metabolism and development has been proposed. Finally, Mizutani summarizes

years of study of isolations and characterizations of ecochemicals. Sequiterpenes,
oligostilbenes, and isoflavonoids are emphasized, and evidence for a variety of
traditional ecological roles is provided.
The Symposium was organized by Norman G. Lewis and G.H. Neal Towers.
Their success in bringing together such a distinguished group of speakers was
exceptional. The array of friendly people at Washington State University who
made our stay pleasant are especially thanked. As in past years, my editing of this
volume was greatly facilitated by the assistance, patience, and friendship of Dawn
McGowan.
John T. Romeo
University of South Florida
CONTENTS
I. Natural Product Drug Discovery and Development: The United

States National Cancer Institute Role .................... .
Gordon M. Cragg, Michael R. Boyd, Rita Khanna, David 1. Newman,
and Edward A. Sausville
2. Taxol Biosynthesis: A Review of Some Determinant Steps . . . . . . . . .. 31

Kevin Walker and Rodney Croteau
3. Role of Lignans in Carcinogenesis ............................. 51

Lilian U. Thompson
4. Toward Engineering the Metabolic Pathways of Cancer-Preventing

Lignans in Cereal Grains and Other Crops ................ 67
Michael A. Costa, Zhi-Qiang, Laurence B. Davin,
and Norman G. Lewis
5. Simple (Bench-Top) Bioassays and the Isolation of New

Chemically Diverse Antitumor and Pesticidal Agents
from Higher Plants ................................... 89
Jerry L. McLaughlin and Ching-Jer Chang
6. Molecular Controls for Isoflavonoid Biosynthesis in Relation to

Plant and Human Health ............................... 133
Richard A. Dixon, Pedro Canovas, Ze-Jian Guo, Xian-Zhi He,
Chris Lamb, and Fiona McAlister
7. Medical Potential and Biosynthesis of Plant Coumarins ............. 161

Ulrich Matern
8. Biosynthesis, Biodegradation, and Cellular Localization of

Hydrolyzable Tannins ................................. 185
Georg G. Gross
xi
xii CONTENTS
9. The Formation of Heartwood and Its Extractives: An Overview ...... 215

WE. Hillis
10. Recent Advances in the Chemistry of Proanthocyanidins ............ 255

Daneel Ferreira, Hendrik van Rensburg, Elfranco Malan, Johan
Coetzee, and Reinier J. J. Nel
11. Astringency and Polyphenol Protein Interactions .................. 289

Edwin Haslam, Michael P. Williamson, Nicola J. Baxter,
and Adrian J. Charlton
12. Phytochemistry of Bryophytes: Biologically Active Terpenoids and

Aromatic Compounds from Liverworts ................... 319
Yoshinori Asakawa
13. Biologically Active Compounds from Buddleja Species ............. 343

Peter J. Houghton and Abraham Y. Mensah
14. Cyanide in Foods: Biology of Cyanogenic Glucosides

and Related Nutritional Problems ........................ 369
Dirk Selmar
15. Plant Ecochemicals from the Viewpoint of Plant Defense ........... 393
Junya Mizutani
Index ....................................................... .421

Chapter One
NATURAL PRODUCT DRUG DISCOVERY AND

DEVELOPMENT
The United States National Cancer Institute Role
Gordon M. Cragg,l* Michael R. Boyd, I Rita Khanna,2

David 1. Newman, I and Edward A. Sausville '
IDevelopmental Therapeutics Program

Division of Cancer Treatment and Diagnosis
National Cancer Institute
Bethesda, Maryland 20892
2 Technology Development and Commercialization Branch
Office of the Director

National Cancer Institute
Bethesda, Maryland 20892
Introduction 2
Anticancer Agents Derived from Natural Resources: The NCI Role ...... . 3
Current Status of the NCI Natural Products Drug Discovery Program ... . 7
Drug Discovery ............................................. . 7
Preclinical Development ...................................... . 9
Clinical Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 10
Natural Products Drug Development: The Supply Issue ................ 10
Paclitaxel (Taxol®) ........................................... 13
Potential Anti-HIY Agent, Michellamine B ......................... 15
Collaboration in Drug Discovery and Development: The NCI Role . . . . . .. 17
Screening Agreement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 18
Collaboration in Preclinical Development: Rapid Access to Intervention
Development (RAID) ....................................... 18
* Author to whom enquiries should be addressed.

Phytochemicals in Human Health Protection. Nutrition. and Plant Defense, edited by Romeo.
Kluwer Academic / Plenum Publishers, New York, 1999.
2 G. M. CRAGG et al.
National Cooperative Drug Discovery Group Program (NCDDG) 18

Source Country Collaboration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 19
Distribution of Extracts from the NCI Natural Products Repository. . . .. 21
DTP WWW Homepage ....................................... 22
International Cooperative Biodiversity Group Program (lCBG), a
Multiagency Program ....................................... 22
Small Business Innovation Research (SBIR) and Small Business
Technology Transfer (STTR) Programs ......................... 23
New Directions in Natural Product Drug Discovery .................... 24
Exploration of New Environments ............................... 24
Unexplored Potential of Microbial Diversity ........................ 24
Targeting Natural Products ...................................... 25
Combinatorial Biosynthesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 25
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 26
INTRODUCTION
Over the ages, humans have relied on nature for their basic needs for the pro-
duction offoodstutTs, shelters, clothing, means of transportation, fertilizers, flavors
and fragrances, and not least, medicines. Plants have formed the basis of sophisti-
cated traditional medicine systems that have been in existence for thousands of
years in countries such as China l and India. 2 These plant-basedsystems continue
to play an essential role in health care, and it has been estimated by the World Health
Organization that approximately 80% of the world's inhabitants rely mainly on tra-
ditional medicines for their primary health care. 3 Plant products also play an impor-
tant role in the health care systems of the remaining 20% of the population mainly
residing in developed countries. Analysis of data on prescriptions dispensed from
community pharmacies in the United States from 1959 to 1980 indicates that about
25% contained plant extracts or active principles derived from higher plants, and
at least 119 chemical substances, derived from 90 plant species, can be considered
as important drugs currently in use in one or more countries. 3 Of these 119 drugs,
74% were discovered as a result of chemical studies directed at the isolation ofthe
active substances from plants used in traditional medicine. Well-known examples
of plant-derived medicinal agents include: the antimalarial drug quinine, obtained
from the bark of Cinchona officinalis; the analgesics, codeine and morphine from
Papaver somniferum; the antihypertensive reserpine from Rauwolfia serpentina;
and the cardiac glycoside, digoxin, from Digitalis purpurea. 4
While marine organisms do not have a history of use in traditional medi-
cine, the ancient Phoenicians employed a chemical secretion from marine mol-
luscs to produce purple dyes for woolen cloth, and seaweeds have long been used
to fertilize the soil. The world's oceans, covering more than 70% of the earth's
NATURAL PRODUCT DRUG DISCOVERY AND DEVELOPMENT 3
surface, represent an enormous resource for the discovery of potential chemother-

apeutic agents. All but two ofthe 28 major animal phyla are represented in aquatic
environments, with eight being exclusively aquatic, mainly marine. 5
The discovery of penicillin from the filamentous fungus, Penicillium
notatum, and the broad therapeutic use of this agent in the 1940s, ushered in a
new era in medicine and the "Golden Age" of antibiotics, and promoted the inten-
sive investigation of nature as a source of novel bioactive agents. Microorganisms
are a prolific source of structurally-diverse bioactive metabolites and have yielded
some of the most important products of the drug industry, including the
penicillins, aminoglycosides, tetracyclines, cephalosporins, and other classes of
antibiotics that have revolutionized modern medicine.
This interest in nature as a source of potential chemotherapeutic agents
continues, and an analysis of the number and sources of anticancer and anti-
infective agents, reported mainly in the Annual Reports of Medicinal Chemistry
from 1984 to 1995 covering the years 1983 to 1994, indicates that over 60% of
the approved drugs developed in these disease areas are of natural origin. 6
ANTICANCER AGENTS DERIVED FROM NATURAL

SOURCES: THE NCI ROLE
The United States National Cancer Institute (NCI) was established in 1937,
its mission being "to provide for, foster, and aid in coordinating research related
to cancer." In 1955, NCI set up the Cancer Chemotherapy National Service Center
(CCNSC) to coordinate a national voluntary cooperative cancer chemotherapy
program, involving the procurement of drugs, screening, preclinical studies, and
clinical evaluation of new agents. By 1958, the initial service nature of the orga-
nization had evolved into a drug research and development program with input
from academic sources and substantial participation of the pharmaceutical indus-
try. The responsibility for drug discovery and preclinical development at NCI now
rests with the Developmental Therapeutics Program (DTP), a major component
of the Division of Cancer Treatment and Diagnosis (DCTD). Thus, NCI has, for
the past forty years, provided a resource for the preclinical screening of com-
pounds and materials submitted by grantees, contractors, pharmaceutical and
chemical companies, and other scientists and institutions, public and private,
worldwide, and has played a major role in the discovery and development of many
of the available commercial and investigational anticancer agents. During this
period, more than 400,000 chemicals, both synthetic and natural, have been
screened for antitumor activity.
Initially, most of the materials screened were pure compounds of synthetic
origin, but the program also recognized that natural products were an excellent
source of chemicals with a wide variety of biological activities. During the early
4 G. M. CRAGG et at.
years of the CCNSC, the screening of natural products was concerned mainly with
the testing of microbial fermentation products, and, prior to 1960, only about 1,500
plant extracts were screened for antitumor activity. Plants have a long history of
use in the treatment of cancer, 7 although many of the claims for the efficacy of such
treatment should be viewed with some skepticism because cancer, as a specific
disease entity, is likely to be poorly defined in terms of folklore and traditional
medicine. 8 Earlier work on the isolation of active antitumor agents from Podophyl-
lum peltatum L., the Mayapple, found throughout the eastern U.S. and used by early
American cultures for the treatment of skin lesions and warts, and the discovery
and development of vinblastine and vincristine, from the rosy periwinkle, Catha-
ranthus roseus (L.) G. Don, used in the treatment of childhood leukemia and other
cancers, however, provided convincing evidence that plants could be sources of a
variety of novel potential cancer chemotherapeutic agents (Fig. 1).8 Epipodophyl-
lotoxin, isolated as the active antitumor agent from various species of Podophyl-
lum, was semisynthetic ally converted into the clinically active agents, etoposide
and teniposide. Thus, the decision was made to explore plants more extensively as
sources of agents with antitumor activity, and, in 1960, an interagency agreement
was established with the United States Department of Agriculture (USDA)
for the collection of plants for screening in the CCNSC program. A small
number of animal extracts, mainly of marine origin, also were tested beginning in
1960, but by the end of 1968, only 1,000 animal extracts had been screened.
The pace of investigation of marine invertebrates accelerated in the 1970s, and by
1982, over 16,000 extracts had been screened. In contrast, however, from 1960 to
1982, over 180,000 microbial fermentation products and over 114,000 plant-
derived extracts were tested for in vivo antitumor activity, mainly using the LI210
and P388 mouse leukemia models. Extracts showing significant activity were
subjected to bioassay-guided fractionation, and the isolated active agents were
submitted for secondary testing against panels comprising four to eight animal
tumor models and human tumor xenografts. 9 Those agents showing significant
activity in the secondary panel were assigned priorities for preclinical and
clinical development.
Of the 92 anticancer drugs commercially available prior to 1983 in the
United States and approved worldwide between 1983 and 1994, approximately
62% can be related to natural origin. 6 While the majority of these drugs were
discovered outside the NCI program, the NCI did playa significant role in the
development of many of them. Two plant-derived agents, paclitaxel (taxol®)
and camptothecin, were discovered through the NCI program, and, while camp-
tothecin failed as a clinical candidate in the 1970s, its derivatives, topotecan,
9-amino camptothecin, and irenotecan (CPT-II), are currently showing clinical
efficacy against a variety of cancer disease types (Fig. I ).10 Other plant-derived
drugs in clinical trials are homoharringtonine, isolated from the small Chinese
evergreen tree, Cephalotaxus harringtonia var. Drupacea (Sieb. & Zucc.)
Koidzumi, and 4-ipomeanol, a pneumotoxic furan derivative produced by
A'
R-+~ o
HO
0
0
HO
Camptothecin : R' = R2 = R3 = H
Etoposide: R = CH3;
Topotecan: R'
R2 = H ; R3 = CH 3CH2
= OH; R2 = CH2N(CH3b; R3 = H
Teniposide : R = (A
OH
N
N
H
""""/
o
II
~
CH'CO OOH
o 0
II II
Ph-C-NH-CH-CH-C-O
I
Ph
I
OH
.H
HO . ~, 0
: H :
PhCO OCCH 3
II II
o 0
Taxal (Ph = Phenyl)

Vincristine: R = CHO Paclitaxel
Vinblastine: R = CH 3
Figure 1. Commercial plant-derived anticancer drugs,

<
~Oll_
elL"
o
Ipomeanol
Harringtonine: R = I - ~
C02 Ma
OH HO O
Homoharringtonine: R = ~_
C02 Me
Figure 2. Plant-derived anticancer drugs in clinical development.
sweet potatoes (Ipomoea hatatas) infected with the fungus, Fusarium solani
(Fig. 2). Homoharringtonine has shown activity against various leukemias and
is in Phase III clinical trials, while ipomeanol is in early clinical trials for
treatment of patients with lung cancer. I I A number of plant-derived agents were
entered into clinical trials by the NCI, but the trials were terminated due to lack
of efficacy or unacceptable toxicity. I I Among these agents were acronycine,
bruceantin, maytansine, and thalicarpine, all of which could serve as cytotoxins
for linking to monoclonal antibodies or other "carrier" molecules targeted to
specific tumors.
Many of the commercial drugs of microbial origin, such as actinomycin D,
doxorubicin (adriamycin), and mitomycin C, were discovered by research groups
associated with the pharmaceutical industry, and this trend continues, generally
in close collaboration with the NCI in the developmental phases. Much of the
drug discovery effort in the marine area, however, has been supported by the NCI
through contract or grant mechanisms. While no marine-organism-derived agent
has yet been approved for commercial development, several agents, including
bryostatin I and dolastatin lO (Fig. 4), are in clinical trials; 12 bryostatin I is
showing some promising activity in trials against melanoma. 13
Most drugs currently available for cancer therapy are effective predominantly
against rapidly proliferating tumors, such as leukemias and lymphomas, but, with
some notable exceptions, such as paclitaxel, show little useful activity against the
slow-growing adult solid tumors, such as lung, colon, prostatic, pancreatic, and
brain tumors. In the early 1980s, the NCI program was discontinued because it was
perceived that few novel active leads were being isolated from natural sources. Of
particular concern was the failure to yield agents possessing activity against the
solid tumor disease-types. This apparent failure might, however, be attributed more
to the nature of the primary screens being used at the time, rather than to a
deficiency of nature. Continued use of the primary P388 mouse leukemia screen
appeared to be detecting only previously identified active compounds or chemical
structure types having little or no activity against solid tumors. In retrospect, these
results might be attributed to the use of a single disease-specific model as the
primary screen that filtered out those agents with potential specificity against
tumors other than mouse leukemia or closely related human diseases.
In an attempt to overcome this deficiency, NCI developed an alternative,
disease-oriented, preclinical anticancer drug discovery strategy aimed at the dis-
covery of new agents for disease-specific clinical trials in relevant cancer patient
populations. 14
CURRENT STATUS OF THE NCI NATURAL PRODUCTS

DRUG DISCOVERY PROGRAM
Drug Discovery
During 1985-1990, the NCI developed a new in vitro primary screen based
upon a diverse panel of human tumor cell lines. 14 The screen currently comprises
sixty cell lines derived from nine cancer types, and organized into subpanels
representing leukemia, lung, colon, central nervous system, melanoma, ovarian,
renal, prostate and breast. In late 1998, a preliminary prescreen comprising three
cell lines will be introduced, and all materials will be tested in the prescreen.
Those materials showing significant activity in one or more of the three lines will
be advanced to the sixty cell line screen for further evaluation.
With the development of the new in vitro screening strategy, the NCI once
again turned to nature as a potential source of novel anticancer agents, and a new
natural products acquisition program was implemented in 1986. Contracts for the
cultivation and extraction of fungi and cyanobacteria and for the collection of
marine invertebrates and terrestrial plants were initiated in 1986, and with the
exception of fungi and cyanobacteria, these programs continue to operate. Marine
organism collections originally focused in the Caribbean and Australasia, but have
now expanded to the Central and Southern Pacific and to the Indian Ocean (off
East and Southern Africa) through a contract with the Coral Reef Research Foun-
dation, which is based in Palau in Micronesia. Terrestrial plant collections have
been carried out in over 25 countries in tropical and subtropical regions world-
wide through contracts with the Missouri Botanical Garden (Africa and Mada-
gascar), the New York Botanical Garden (Central and South America), and the
University of Illinois at Chicago (Southeast Asia), and have been expanded to the
continental United States through a contract with the Morton Arboretum.
In carrying out these collections, the NCI contractors work closely with
8 G. M. CRAGG et a/.
qualified organizations in each of the source countries. Botanists and marine

biologists from source country organizations collaborate in field collection
activities and taxonomic identifications, and their knowledge of local species and
conditions is indispensable to the success of the NCI collection operations. Source
country organizations provide facilities for the preparation, packaging, and ship-
ment of the samples to the NCI's Natural Products Repository (NPR) in Freder-
ick, Maryland. The collaboration between the source country organizations and
the NCI collection contractors, in turn, provides support for expanded research
activities by source country biologists, and the deposition of a voucher specimen
of each species collected in the national herbarium or repository is expanding
source country holdings of their biota. When requested, NCI contractors also
provide training opportunities for local personnel through conducting workshops
and presentation of lectures. In addition, through its Letter of Collection (LOC)
and agreements based upon it, the NCI invites scientists nominated by Source
Country Organizations to visit its facilities, or equivalent facilities in other
approved U.S. organizations, for 3-12 months to participate in collaborative
natural products research. Representatives of most of the source countries have
visited the NCI and contractor facilities for shorter periods to discuss colla-
boration. 15 Contract collections of plants are now being de-emphasized in favor
of establishing direct collaborations with qualified organizations in the source
countries (discussed below).
Dried plant samples (0.3-1 kg dry weight) and frozen marine organism
samples (-I kg wet weight) are shipped to the NPR in Frederick where they
are stored at -20°C prior to extraction with a I: I mixture of methanol:
dichloromethane and water to give organic solvent and aqueous extracts. All
extracts are assigned discrete NCI numbers and returned to the NPR for storage
at -20°C until requested for screening or further investigation. After testing in
the in vitro human cancer cell line screen, active extracts are subjected to bioas-
say-guided fractionation to isolate and characterize the pure, active constituents.
Agents showing significant activity in the primary in vitro screens are selected
for secondary testing in several in vivo systems. Those agents exhibiting
significant in vivo activity are advanced into preclinical and clinical development.
Of the 51 anticancer agents currently in active preclinical or Phase I development
by NCI (excluding biologics), 29 are either natural products or derived from
natural products, with the source organisms being 12 microbial, 2 marine, 6 plant,
and 4 animal in origin, together with 5 synthetic compounds based on natural
products models. 6
As part of the response of the National Institutes of Health (NIH) to the
AIDS epidemic, DTP developed a screening program for the large-scale testing
of synthetic and natural materials for anti-HIV activity.16 The screen measured
the effect of materials on the growth of human Iymphoblastoid cells in the pres-
ence or absence of the human immunodeficiency virus (HIV-I), 17 and from 1988
to 1996, over 90,000 extracts were tested in this screen. In late 1996, the screen-
DECISION NETWORK PROCESS
STAGES:
-·IIA------.· liB -III--+IND

Acquisition (CTEP)
Proof of Pharmacology
Screening Phase I
Activity Toxicology
Protocol
(INO-directed)
INOA Filing
Optimize Formulation
Schedule
Bulk
Synthesis
Preliminary
Tox& Ph arm
OVERALL MANAGEMENT:
CANCER OR AIDS OPERATING COMMITIEE
Figure 3. The NCI decision network process.
ing of extracts was discontinued, and alternative assays involving the use of target
enzymes are now being used.
Preclinical Development
Those agents showing significant in vivo activity are presented to the NCI
Division of Cancer Treatment and Diagnosis (DCTD) Decision Network Com-
mittee (DNC), and, if approved by the DNC, the agent is entered into preclinical
and clinical development. The Decision Network Process divides the preclinical
drug development process into stages designated as DNIIA, DNIIB, and DNIII
as described below (Fig. 3).
• An adequate supply of natural product is procured to permit preclini-
cal and clinical development (discussed in detail below).
• Formulation studies are performed to develop a suitable vehicle to sol-
ubilize the drug for administration to patients, generally by intravenous
injection or infusion in the case of cancer. The low solubility of many
natural products in water poses considerable problems, but these can
be overcome by use of co-solvents or emulsifying agents (surfactants)
such as Cremophore EL (polyoxyethylated castor oil).
• Pharmacological evaluation determines the best route and schedule of
administration to achieve optimal activity of the drug in animal models,
the half-lives and bioavailability of the drug in blood and plasma, the
rates of clearance and the routes of excretion, and the identity and rates
of formation of possible metabolites.
• In the final preclinical step, toxicological studies are performed to
determine the type and degree of major toxicities. in rodent and dog
models. These studies help to establish the safe starting doses for
administration to human patients in clinical trials.
Clinical Development
Phase I studies are conducted to determine the maximum tolerated dose

(MTD) of a drug in humans and to observe the sites and reversibilities of any
toxic effects. Once the MTD has been determined and the clinicians are satisfied
that no insurmountable problems exist with toxicities, the drug advances to Phase
II clinical trials. These trials generally are conducted to test the efficacy of the
drug against a range of different cancer disease types. In those cancers where
significant responses are observed, Phase III trials are conducted to compare the
activity of the drug with that of the best chemotherapeutic agents currently avail-
able for the treatment of those cancers. In addition, the new drug may be tried in
combination with other effective agents to determine if the efficacy of the com-
bined regimen exceeds that of the individual drugs used alone.
Some promising anticancer and anti-HIV agents currently in development
are shown in Figures 4 and 5.
NATURAL PRODUCT DRUG DEVELOPMENT: THE SUPPLY

ISSUE
The critical first step in the development of any natural product drug is the
procurement of an adequate supply to meet the requirements for preclinical and
clinical investigation. While total synthesis may be considered as a potential
route for bulk production of the active agent, it is worth noting that the structures
of most bioactive natural products are complex, and bench-scale syntheses
often are not readily adapted to large-scale economic production. Isolation from
the natural source, therefore, often provides the most economically viable method
of production. It also should be noted that, of the established plant-derived
commercial anticancer drugs, vinblastine and vincristine are still produced by
isolation from Catharanthus rose us grown in various regions worldwide, while
etoposide and teniposide are semisynthetically produced from natural precursors
isolated from Podophyllum emodii harvested in India and Pakistan (Fig. I).
The problems associated with the large-scale production of paclitaxel also
have been resolved through semisynthesis from natural precursors, such as
baccatin III and IO-desacetylbaccatin Ill, isolated from the needles of various
Taxus species. IS
NATURAL PRODUCT DRUG DISCOVERY AND DEVELOPMENT II
Bryostatin-1
NSC 339555
HO
Flavopiridol UCN-01
NSC 649890 NSC 638850
KRN5500
NSC 650426
Dolastatin 10
NSC 376128
Figure 4. Some natural product-derived anticancer agents in development by the NCI.
H,C
H,C
CH,
CH,
Calanolide A
Michellamine B
CH,
Prostratin
Conocurvone
Figure 5. Some natural product-derived anti-AIDS agents discovered by the NCr.
The initial raw material collection sample (0.3-1.0 kg) generally will yield
enough extract (10-40 g) to permit isolation of the pure, active constituent in
sufficient milligram quantity for complete structural elucidation. Subsequent
secondary testing and preclinical development, however, might require gram or
even kilogram quantities, depending on the degree of activity and toxicity of the
active agent.
In order to obtain sufficient quantities of an active agent for early pre-
clinical development, recollections of 5-50kg of the raw material, preferably
from the original collection location, might be necessary. Should the preclinical
studies justify development ofthe agent towards clinical trials, considerably larger
amounts of material would be required. The performance of large recollections
necessitates surveys of the distribution and abundance of source organisms as
well as determination of the variation of drug content in the various parts in the
case of plants, and the fluctuation of content with time and season of harvesting.
In addition, the potential for mass cultivation or aquaculture of the source organ-
ism needs to be assessed. If problems are encountered due to scarcity of the wild
source or inability to adapt it to cultivation, a search for alternative sources is
necessary. Other species of the same genus, or closely related genera, can be ana-
lyzed for drug content, and techniques, such as tissue culture, can be investigated.
Paclitaxel (Taxol®)
Probably the most significant drug discovered and developed through the
NCI natural products program is the plant-derived agent, paclitaxel (Fig. I). Pacli-
taxel was isolated through an NCI contract with Drs. Monroe Wall and Mansukh
Wani of Research Triangle Institute from the bark of the Pacific Yew tree (Taxus
brevifalia) in 1969 from samples collected by the USDA as part of the early
exploratory plant screening program. Like many other potential anticancer agents
at that time, paclitaxel only showed moderate activity against the then current
mouse leukemia models and was not considered of particular interest. It was only
the observation of its activity in new test systems (the 816 melanoma and several
human tumor xenografts) developed in the mid- to late 1970s that revived inter-
est. This interest was further heightened by the discovery of its unique mecha-
nism of action by Dr. Susan Horwitz of Albert Einstein School of Medicine;
paclitaxel polymerizes tubulin and stabilizes microtubules, thereby inhibiting
mitosis and cell division. '9
These observations promoted the development ofpaclitaxel which advanced
through preclinical studies (e.g. animal toxicology) to initiation of Phase I clinical
trials in 1983. The early trials were fraught with serious problems of toxicity, par-
ticularly allergic reactions including anaphylaxis, which brought it close to being
dropped from clinical studies. The toxicity was traced back to the poor solubility
of taxol in aqueous systems which required the use of high concentrations of
the emulsifying agent, Cremophore EL (a castor oil derivative), in the preparation
of a suitable vehicle for parenteral administration; Cremophore EL is known
to elicit hypersensitivity reactions. These problems were alleviated by the use
ofionger infusion times (e.g. 24 hours every 14-21 days) and premedication with
anti-allergy drugs.
Due to the slow progress of paclitaxel through Phase I clinical trials and
doubts about its clinical efficacy, only sufficient drug for a moderate number of
trials was isolated in bulk (several kilograms) from the Pacific Yew bark. This
later became a problem when important activity was found in Phase II trials in
ovarian cancer in 1987 and interest in the drug greatly intensified. The observa-
tion of approximately 30% response rates in trials with patients having refractory
ovarian cancer resulted in a tremendous demand for the drug. 20 The yield ofpacli-
taxel isolated was about 1 gram per 301bs of bark, and the average Pacific Yew
tree (about 100 years in age) yielded about 20lbs of bark (equivalent to 1.5 trees
per gram). Given that about 12,000 women were dying annually in the U.S. from
advanced ovarian cancer, and that the usual treatment required about 2 grams of
paclitaxel per patient, 24,000 grams were needed, amounting to the destruction
of 36,000 trees. Meanwhile, significant activity also was observed in the treat-
ment of patients with metastatic breast cancer (40,000 deaths per year), and
responses were being observed also in patients with other forms of advanced
malignancy, including lung cancer, malignant melanoma, and lymphomas.
A detailed analysis of the paclitaxel supply crisis and its eventual solution
has been published, 18 so only a brief review is presented here. The initial source
was the bark of the Pacific Yew, Taxus brevi/alia, an understory tree growing in
the forests of the Pacific northwest from northern California into British Colum-
bia. The taxol supply needs for preclinical and early clinical studies were met
easily by bark collections in Oregon between 1976 and 1985, ranging in size from
2,000 pounds to 15,000 pounds. Later observations of responses in the treatment
of patients with a variety of solid tumors, including malignant melanoma and
ovarian cancer, led to an escalation in demand for drug, resulting in several 60,000
pound-collections between 1987 and 1989. These collections raised concerns
about their impact on the continued existence of the tree, but inventories con-
ducted by the USDA Forest Service and Bureau of Land Management and funded
by Bristol-Myers Squibb (BMS) determined that the tree was abundant (estimates
of>100 million trees) on government land. Over 1.6 million pounds of bark were
harvested under strictly controlled conditions in each of the years 1991 and 1992
by Hauser Northwest, a subsidiary of Hauser Chemical Research (HCR) under
contract to BMS. These collections resulted in the production of hundreds of
kilograms of paclitaxel by HCR.
Both NCI and BMS realized that alternative sources of paclitaxel would
need to be developed to permit its eventual marketing as a clinical drug, and NCI
organized workshops in 1990 and 1992 to promote research into various aspects
of paclitaxel. Analytical surveys of the needles of a number of Taxus species
collected from several countries, including Canada, Georgia, Mexico, Russia,
Ukraine, and the United States were performed, and the content of paclitaxel and
key baccatin precursors in various Taxus cultivars was determined. Though the
paclitaxel content ofthe needles generally was lower than that of the bark, needles
of several species and cultivars were found to be relatively good sources of bac-
catin precursors. The pioneering studies, by the French research team of Greene,
Poitier and coworkers, of the semisynthetic conversion of 10-desacetylbaccatin
III, isolated from the needles of T baccata, to paclitaxel,21 and the subsequent
development of more efficient conversion processes,22 led to the large-scale pro-
duction of paclitaxel and related compounds, such as taxotere,23 from renewable
needle resources, and the solution of the supply problem. Significant advances
also have also been made in the production of paclitaxel through plant tissue
culture using technology developed by the company Phyton Catalytic, working
with BMS,24 while there also have been promising developments in the isolation
from microbial sources. 25
Potential Anti-HIV Agent, Michellamine B
Michellamine B (Fig. 5) was isolated as the main in vitro active anti-HIV

agent from the leaves of the 1iana, Ancistrocladus korupensis, collected in the
Korup region of southwest Cameroon. 26 Initially, the plant was tentatively
identified as A. abbreviatus, but collections of this and several other known
Ancistrocladus species failed to yield any michellamines or show any anti-HIV
activity. Subsequent detailed taxonomic investigation of the source plant com-
pared to authentic specimens of A. abbreviatus revealed subtle but distinctive
morphological differences, and the species was determined to be new to science,
and officially named Ancistrocladus korupensis. 27 Michellamine B shows in vitro
activity against a broad range of strains of both HIV-I and HIV-2, including
several resistant strains of HIV-I. 26 The species appears to be distributed mainly
within the Korup National Park, and vine densities are of the order of one large
vine per hectare. Fallen leaves collected from the forest floor contain michel-
lamine B, and collections of these leaves provided sufficient biomass for the iso-
lation of enough drug for completion of preclinical development. It was clear,
however, that extensive collections of fresh leaves could pose a possible threat to
the wild source. Thus far, no other Ancistrocladus species has been found to
contain michellamine B, and investigation of the feasibility of cultivation of the
plant as a reliable biomass source was initiated in 1993 through a contract with
the Center for New Crops and Plant Products of Purdue University, working in
close collaboration with the University of Yaounde 1, the World Wide Fund for
Nature Korup Project, Missouri Botanical Garden, Oregon State University, and
the NCI contractor, Program Resources, Inc. An extensive botanical survey was
undertaken, and the range and distribution of the species were mapped out. Dried
leaf samples from representative vines were shipped to NCI for analysis of
michellamine B content. Plants indicating high concentrations were re-sampled
for confirmatory analysis, and those repeatedly showing high concentrations were
targeted for cloning via vegetative propagation. A medicinal plant nursery was
established to hold and maintain the A. korupensis collection at the Korup Park
Headquarters in Mundemba. In keeping with the NCI policies of collaboration
with source countries, all cultivation studies were performed in Cameroon and
involved local popUlations, particularly those in the regions adjacent to the Korup
National Park.
Based on the observed activity, the NCI committed michellamine B to
INDA-directed preclinical development. Unlike many natural products, formula-
tion presented no problem since the drug is readily water-soluble as its diacetate
salt. Continuous infusion studies in dogs indicated that in vivo effective anti-HIV
concentrations could only be achieved close to toxic dose levels. Thus, despite
these observations and the in vitro activity against an impressive range of HIV-l
and HIV-2 strains, there were serious disadvantages which precluded advance-
ment of michellamine B to clinical trials. The difference between the toxic dose
level and the anticipated level required for effective antiviral activity was small,
indicative of a narrow therapeutic index. Further toxicology studies in primates
confirmed the very narrow therapeutic index, and indicated potential neurologi-
cal toxicity. Based on these observations, NeI discontinued further studies aimed
at clinical development.
Despite this decision, it is possible that the pharmacological and toxico-
logical profiles can be improved through analogue synthesis. Such studies could
require substantial quantities of the natural product, or the successful synthetic
studies of Bringmann and his group could provide a satisfactory solution. 28 The
isolation of the novel antimalarial compounds, the korupensamines (Fig. 6), from
10'
OH OCH3 OH OCH3
g'
CH3
HO CH3
OH CH3
9
korupensamine A korupensamine B
II' 10' 10'

OCH 3 0CH3 OH OCH3
g' g'
CH3 CH3
HO CH3 HO
11
OH
korupensamine C korupensamine D
Figure 6, Korupensamines A-D: plant-derived potential antimalarial compounds.

A. korupensis, provides another class of potential medicinal agents from this

plant. 29 The korupensamines, which are equivalent to the "monomeric" units of
the michellamines, are essentially inactive against HIY, whereas the michel-
lamines exhibit only weak antimalarial activity.
COLLABORATION IN DRUG DISCOVERY AND

DEVELOPMENT: THE NCI ROLE
As noted above, much of the NCI drug discovery and development effort
has been, and continues to be, carried out through collaborations with research
organizations and the pharmaceutical industry worldwide. Many of the naturally
derived anticancer agents were developed through such efforts. Thus, the discovery
and preclinical development of etoposide and teniposide, semisynthetic derivatives
of the natural product epipodophyllotoxin, were performed by Sandoz investiga-
tors, and the NCI played a substantial role in the clinical development.
Though paclitaxel (taxol®) was discovered by Wall and Wani with NCI IS
contract support, the key to solving the supply problem was the semisynthetic con-
version ofbaccatin III derivatives to paclitaxel (and taxane analogs) pioneered by
the French group led by Poitier,21 followed by the development of alternative con-
version methods by the Holton group,22 supported by the NCI and Bristol-Myers
Squibb. IS The semisynthetic analog, taxotere (docetaxel), produced through a col-
laborative agreement between the Centre National de la Recherche Scientifique
(CNRS) and Rhone-Poulenc Rorer, after undergoing extensive clinical evaluation
in Europe and North America under auspices of organizations, such as the Euro-
pean Organization for Research and Treatment of Cancer (EORTC), and the Cana-
dian and U.S. National Cancer Institutes, 23 is now in clinical use in Europe and
North America. Indeed, there is close collaboration between the EORTC, the
United Kingdom Cancer Research Campaign (CRC), and the NCI in the preclini-
cal and clinical development of many anticancer agents, such as bryostatin I, dolas-
tatin 10, aphidicolin glycinate, rhizoxin, pancratistatin, and phyllanthoside.
Drugs, such as bleomycin, aclacinomycin, and deoxyspergualin, were dis-
covered by the Umezawa group at the Institute of Microbial Chemistry in Japan
and developed in collaboration with the NCI; a number of the agents currently
in early clinical development at the NCI, such as UCN-OI, and quinocarmycin
and spicamycin analogs, are the result of collaborations between Japanese com-
panies, such as Kyowa Hakko Kogyo, Fujisawa Pharmaceutical Co. Ltd, and Kirin
Brewery Ltd, and the NCL I2
The DTP of the NCI thus complements the efforts of the pharmaceutical
industry and other research organizations through taking positive leads, which
industry might consider too uncertain to sponsor, and conducting the "high risk"
research necessary to determine their potential utility as anticancer drugs. In pro-
moting drug discovery and development, the DTP/NCI has formulated various
mechanisms for establishing collaborations with research groups worldwide.
18 G. M. CRAGG et ai.
Screening Agreement
In the case of organizations wishing to have pure compounds tested in the

NCI drug screening program, such as pharmaceutical and chemical companies or
academic research groups, the DTPINCI has formulated a screening agreement
which includes terms stipulating confidentiality, patent rights, routine and non-
proprietary screening and testing versus non-routine, and levels of collaboration
in the drug development process. Individual scientists and research organizations
wishing to submit pure compounds for testing generally consider entering into
this agreement with the NCI DCTD. Should a compound show promising anti-
cancer activity in the routine screening operations, the NCI will propose the estab-
lishment of a more formal collaboration, such as a Cooperative Research and
Development Agreement (CRADA) or a Clinical Trial Agreement (CTA).
Collaboration in Preclinical Development: Rapid Access to

Intervention Development (RAID)
RAID is a new program designed to facilitate translation to the clinic of

novel, scientifically meritorious therapeutic interventions originating in the aca-
demic community. The RAID process makes available to the academic research
community, on a competitive basis, NCI resources for preclinical development of
drugs (see above). The process functions as a collaboration between the NCI and
the originating laboratory, and tasks may be apportioned to either the NCI or the
originating laboratory, depending on the facilities and expertise available in the
latter. While the RAID process is similar to the Decision Network Process dis-
cussed above, the products of the RAID program are returned directly to the orig-
inating laboratory for proof-of-principle clinical trials. It is assumed that most of
the products in the RAID program will be studied clinically under investigator-
held INDs (Investigational New Drug approvals granted by the FDA) within the
originating (or a collaborating) institution. NCI may consider assuming respon-
sibility for clinical trials sponsorship if unanticipated circumstances develop pre-
cluding clinical development by the originating institution. The RAID process
cannot be used by private industry (which can interact with NCI through the DN
process), nor can it be used to develop a product already licensed to a company;
however, the existence of research collaborations between the academic investi-
gators and companies does not affect the eligibility for support from RAID for
an individual product, provided the product is not licensed to a company.
Full details may be obtained from the DTP Website (see below).
National Cooperative Drug Discovery Group (NCDDG) Program
In the late 1970s and early 1980s, many significant discoveries were made
in such fields as biochemistry, molecular biology, embryology, and carcinogene-
sis that had the potential for the development of new strategies and agents for
cancer treatment; most investigators, however, were working only in their own
areas of expertise without the benefit of close liaison with experts in the many
disciplines required to discover and develop new therapies and strategies. In
response to the need to coordinate these research efforts, the NCI initiated the
NCDDG Program in the early 1980s with the goal of bringing together scientists
from academia, industry, and government, in the form of consortia, in a focused
effort aimed at the discovery of new drugs. 30 The inclusion of an industrial com-
ponent in almost all consortia has had positive effects in helping to orient the aca-
demic component( s) towards drug development, and maintaining a focus on the
final outcomes of drug discovery in terms of clinical trials and marketable prod-
ucts, as well as contributing high quality scientists and resources to the Program.
Involvement of NCI Staff has enabled the NCI to contribute its considerable
resources and expertise in cancer drug development, including extensive com-
puterized databases and repositories of compounds tested over more than 40
years, primary and secondary screening systems, and all the resources necessary
for preclinical development of agents meeting selection criteria of the NCI Deci-
sion Network Committee or RAID program. The consortia, headed by a Princi-
pal Investigator, submit proposals based on independent ideas, rather than in
response to specific topics proposed by the NCI, thereby permitting the widest
scope and the greatest degree of innovative science, and encouraging diversity in
the discovery of new drugs and therapeutic approaches.
The National Cooperative Natural Product Drug Discovery Group
(NCNPDDG) Program is one of four such programs, the other three being
directed at studies of Mechanisms of Action, Specific Diseases (e.g. lung and
colon cancer), and Preclinical Model Development. Since 1989, twelve NCN-
PDDGs have been awarded encompassing the study of all natural sources, includ-
ing plants, marine bacteria and invertebrates, microalgae, cyanophytes, and
dinoflagellates, and using a variety of assays, such as molecular targets, mecha-
nism-based assays, cell lines and in vivo systems.
Source Country Collaboration
Drug Discovery: Memorandum of Understanding. As discussed, the

collections of plants and marine organisms have been carried out in over 25
countries through contracts with qualified botanical and marine biological orga-
nizations working in close collaboration with qualified source country organiza-
tions. The recognition of the value of the natural resources (plant, marine and
microbial) being investigated by the NCI and the significant contributions being
made by source country scientists in aiding the performance of the NCI collec-
tion programs have led the NCI to formulate its LOC, specifying policies aimed
at facilitating collaboration with, and compensation of, countries participating in
the drug discovery program. IS
With the increased awareness of genetically-rich source countries to the

value of their natural resources and the confirmation of source country sovereign
rights over these resources by the U.N. Convention of Biological Diversity, orga-
nizations involved in drug discovery and development are increasingly adopting
policies of equitable collaboration and compensation in interacting with these
countries. 3 ! Particularly in the area of plant-related studies, source country sci-
entists and governments are committed to performing more of the operations
in-country, as opposed to the export of raw materials. The NCI has recognized
this fact for several years, and has negotiated Memoranda of Understanding
(MOU) with a number of source country organizations suitably qualified to
perform in-country processing. In considering the continuation of its plant-
derived drug discovery program, the NCI has de-emphasized its contract
collection projects in favor of expanding closer collaboration with qualified
source country scientists and organizations. In establishing these collaborations,
the NCI undertakes to abide by the same policies of collaboration and compen-
sation as specified in the Lac. A number of other organizations and companies
have implemented similar policies. 3 ! Through this mechanism, collaborations
have been established with organizations in Bangladesh, Brazil, China, Costa
Rica, Iceland, Korea, Mexico, New Zealand, Pakistan, Panama, Russia, South
Africa, and Zimbabwe.
Drug Development: The Calanolides. In 1988, an organic extract of the

leaves and twigs of the tree, Calophyllum lanigerum, collected in Sarawak,
Malaysia in 1987, through the NCI contract with the University of Illinois at
Chicago (UIC) in collaboration with the Sarawak Forestry Department, showed
significant anti-HIV activity. Bioassay-guided fractionation of the extract yielded
(+)-calanolide A (Fig. 5) as the main in vitro active agent. 32 Attempted recollec-
tions in 1991 failed to locate the original tree, and collections of other specimens
of the same species gave only trace amounts of calanolide A. In 1992, a detailed
survey of C. lanigerum and related species was undertaken by UIC and botanists
ofthe Sarawak Forestry Department. As part of the survey, latex samples of Calo-
phyllum teysmanii were collected and yielded extracts showing significant anti-
HIV activity. The active constituent was found to be (-)-calanolide B which was
isolated in yields of 20 to 30%. While (-)-calanolide B is slightly less active
than (+)-calanolide A, it has the advantage of being readily available from the
latex which is tapped in a sustainable manner by making small slash wounds in
the bark of mature trees without causing any harm to the trees. A decision was
made by the NCI DNC to proceed with the preclinical development of both the
calanolides, and, in June of 1994, an agreement based on the NCI Letter of Col-
lection was signed between the Sarawak State Government and the NCI. Under
the agreement a scientist from the University of Malaysia Sarawak was invited to
visit the NCI laboratories in Frederick to participate in the further study of the
compounds.
The NCI obtained patents on both calanolides, and, in 1995, an exclusive

license for their development was awarded to Medichem Research, Inc., a small
pharmaceutical company based near Chicago. Medichem Research had devel-
oped a synthesis of (+)-calanolide A 33 under a Small Business Innovative
Research (SBIR; see below) grant from the NCI. The licensing agreement
specified that Medichem Research negotiate an agreement with the Sarawak State
Government. Medichem Research, in collaboration with the NCI, has advanced
(+)-calanolide A through preclinical development and has been granted an INDA
for clinical studies by the U.S. Food and Drug Administration (FDA). The
Sarawak State Government and Medichem Research formed a joint venture
company, Sarawak Medichem Pharmaceuticals Incorporated (SMP), in late 1996,
and SMP has sponsored Phase I clinical studies with healthy volunteers. It has
been shown that doses exceeding the expected levels required for efficacy against
the virus are well tolerated. Trials using patients infected with HIV-l are expected
to begin in late 1998.
Meanwhile, by late 1995, the Sarawak State Forestry Department, VIC, and
the NCI had collaborated in the collection of over 50 kg of latex of C. teysmanii,
and kilogram quantities of (-)-calanolide B have been isolated for further devel-
opment towards clinical trials. The development of the calanolides is being
facilitated through the signing of a Cooperative Research and Development
Agreement (CRADA) between Medichem Research and the NCI in which NCI
is contributing research knowledge and expertise.
The development of the calanolides is an excellent example of collabora-
tion between a source country (Sarawak, Malaysia), a company (Medichem
Research, Inc.), and the NCI in the development of promising drug candidates,
and illustrates the effectiveness and strong commitment of the NCI to policies
promoting the rights of source countries to fair and equitable collaboration and
compensation in the drug discovery and development process. The development
of the calanolides has been reviewed as a "Benefit-Sharing Case Study" for the
Executive Secretary of the Convention on Biological Diversity by staff of the
Royal Botanic Gardens, Kew. 34
Distribution of Extracts from the NCI Natural

Products Repository
In carrying out the collection and extraction of thousands of plant and

marine organism samples worldwide, the NCI has established a Natural Products
Repository (NPR) which is a unique and valuable resource for the discovery of
potential new drugs and other bioactive agents. The rapid progress made in the
elucidation of mechanisms underlying human diseases has resulted in a prolifer-
ation of molecular targets available for potential drug treatment. The adaptation
of these targets to high throughput screening processes has greatly expanded the
potential for drug discovery. In recognition of this potential, the NCI has devel-
oped policies for the distribution of extracts from the NPR to qualified organiza-
tions, subject to the signing of a legally-binding Material Transfer Agreement
(MTA) (see DTP WWW Homepage below).
To be considered for access to the NPR, organizations are required to
submit short proposals outlining the nature of their screening systems and demon-
strating the capability to process active extracts and develop any isolated active
agents towards clinical trials and commercial production. Approved organizations
have to enter into an MTA with DCTD, with one of the key terms being the
requirement that the recipient organization negotiate suitable terms of collabora-
tion and compensation with the source country(ies) of any extract(s) shown to
exhibit significant activity in the organization's screens. While the current poli-
cies only apply to the testing of extracts in screens pertaining to activity against
cancer, AIDS, and related opportunistic infections, as well as diseases of concern
to developing countries (e.g. malaria, parasitic diseases), the extension to testing
against all human diseases currently is being considered.
DTP WWW Homepage
The NCI DTP offers access to a considerable body of data and background
information through its WWW Homepage:http://dtp.nci.nih.gov/
Publicly available data include results from the human tumor cell line
screen and AIDS antiviral drug screen, the expression of molecular targets in
cell lines, and 2D and 3D structural information. Background information is
available on the drug screen and the behavior of "standard agents," NCI investi-
gational drugs, analysis of screening data by COMPARE,14 the AIDS antiviral
drug screen, and the 3D database. Data and information are only available on
so-called "open compounds" which are not subject to the terms of confidential
submission.
In providing screening data on extracts, they are identified by code numbers
only; details of the origin of the extracts, such as source organism taxonomy and
location of collection, may only be obtained by individuals or organizations pre-
pared to sign agreements binding them to terms of confidentiality and require-
ments regarding collaboration with, and compensation of, source countries. Such
requirements are in line with the NCI commitments to the source countries
through its LOC and the MTA.
International Cooperative Biodiversity Group (ICBG) Program,

a Multiagency Program
The ICBG Program resembles NCDDG Programs in structure, in that con-

sortia are formed comprising academic, industrial, and U.S. government organi-
zations. Organizations from developing countries, however, are also required
components. This Program is jointly sponsored by the National Science Founda-
tion (NSF) and components of the National Institutes of Health (NIH), including
the NCI, the National Institute of Allergy and Infectious Diseases (NIAID), the
National Heart, Lung, and Blood Institute (NHLBI), and the National Institute of
Mental Health (NIMH). The goals of the Program are research into drug discov-
ery from natural sources, linked to the identification, inventory, and conservation
of biodiversity, a primary concern of the NSF, and economic development in
developing countries. 30 All these goals are linked to the provision of suitable
training and infrastructure building.
Five awards, four involving countries in Central and South America and one
involving the West African countries of Cameroon and Nigeria, were awarded in
1993 and 1994, and are administered through the NIH Fogarty International
Center. A significant challenge in the development of the ICBGs was the estab-
lishment of principles related to intellectual property rights and the protection of
the rights of the participating source (developing) countries, including commu-
nities and indigenous peoples. While it was possible to develop guidelines for use
in negotiating contracts and agreements, no single set of contractual terms could
apply to all participants, and awardees have developed unique mechanisms and
agreements to suit the particular circumstances of the organizations and countries
involved. 35 As integrated conservation and development projects, the long term
evaluation of this Program will depend on how successful the projects are in
demonstrating the economic value of biodiversity in providing new pharmaceu-
ticals and sustainable natural products-based industries for the participating
developing countries. The Program was recently recompeted. Five new awards,
involving countries in Central and South America (Argentina, Chile, Mexico,
Panama, and Suriname), Africa (Cameroon, Madagascar, and Nigeria), and
Southeast Asia (Vietnam) were awarded in September, 1998.
Small Business Innovation Research (SBIR) and Small Business

Technology Transfer (STTR) Programs
The SBIR program is a set-aside program designed to support innovative

research by small U.S. business concerns (500 or less employees) that have
the potential for commercialization of the subject of the research. The program
is divided into two phases. Phase I covers a six-month period for feasibility
studies of a proposed project, and currently can be funded to the extent of
$100,000. Phase II covers a two-year period for development of any project
considered of sufficient promise to clinical application and commercialization,
and currently can be funded to the extent of $750,000. The STTR program
supports cooperative research and development with potential for commer-
cialization between small business concerns and U.S. non-profit research
organizations.
The research topics should be in areas of emerging and high priority, and
natural product topics of interest include:
• New biological methods for production of bioactive natural products.

• New systems for the large-scale production of active agents for
preclinical and clinical development
• Newer methods for the isolation and purification of active agents.
• Methods for the isolation, purification, identification, cultivation, and
extraction of microbes from unusual habitats.
Further information on the Small Business programs and funding opportu-
nities throughout the National Institutes of Health may be obtained from the NIH
homepage:http://www.nih.gov/grants/funding/sbir.htm
NEW DIRECTIONS IN NATURAL PRODUCT

DRUG DISCOVERY
Exploration of New Environments
As discussed, the potential of the marine environment as a source of

novel drugs remains largely unexplored. Despite the more intensive investigation
of terrestrial flora, it is estimated that only 5-15% of the approximately 250,000
species of higher plants have been systematically investigated chemically
and pharmacologically/6 and the potential of large areas of tropical rainforests
remains virtually untapped. The continuing threat to biodiversity through
the destruction of terrestrial and marine ecosystems lends an urgency to the
need to expand exploration of these resources as a source of novel bioactive
agents.
Unexplored Potential of Microbial Diversity
Until recently, microbiologists were greatly limited in their study of natural

microbial ecosystems due to an inability to cultivate most naturally occurring
microorganisms. In a report recently released by the American Academy of
Microbiology entitled "The Microbial World: Foundation of the Biosphere," it is
estimated that "less than 1% of bacterial species and less than 5% of fungal
species are currently known," and recent evidence indicates that millions of
microbial species remain undiscovered. 37
New development of procedures for cultivating and identifying microor-
ganisms will aid microbiologists in their assessment of the earth's full range of
microbial diversity. In addition, procedures based on the extraction of nucleic
acids from environmental samples will permit the identification of microorgan-
isms through the isolation and sequencing of ribosomal RNA or rONA (genes
encoding for rRNA); samples may be obtained from soils and marine habitats, as
well as extreme habitats, such as hot springs, deep-sea vents, sea ice, and polar
lakes. Valuable products and information are certain to result from the cloning
and understanding of the novel genes which will be discovered through these
processes.
The report concludes that "these new microorganisms provide a vast
untapped reservoir of genetic and metabolic diversity, the harvesting and study
of which will have far-reaching, positive effects for society in areas such as
enhanced food production, medicine (e.g. antibiotic discovery), bioremediation
of waste materials, and agriculture. 37"
Targeting Natural Products
A recurring liability of natural products, at least in the area of cancer

chemotherapy, is that, although many generally are potent, they have limited sol-
ubility in aqueous solvents and exhibit narrow therapeutic indices. These factors
have resulted in the demise of a number of promising leads, such as bruceantin
and maytansine.
An alternative approach to utilizing such agents is to investigate their poten-
tial as "warheads" attached to monoclonal antibodies specifically targeted to epi-
topes on tumors of interest. 38 While this is not a new area of research to the NCr,
the DTP is well established to refine and expand this approach to cancer therapy.
The DTP has a wide range of potent, natural product chemotypes to explore as
potential "warheads," and also has the capability to produce clinical grade mon-
oclonal antibodies (Mabs) through its Biological Resources Branch.
Another strategy of interest is the use of antibodies as vectors for enzymes
capable of activating a nontoxic drug precursor (prodrug) to a potent cytotoxic
moiety.39 After injection and localization of an antibody-enzyme conjugate at the
tumor, a nontoxic prodrug is administered, and while remaining innocuous to the
normal tissues, it is converted to the cytotoxin by the enzyme localized at the
tumor. This approach, called "antibody-directed enzyme prodrug therapy"
(ADEPT) provides further potential for the application of potent natural products
to cancer treatment.
Combinatorial Biosynthesis
Advances in the understanding of bacterial aromatic polyketide biosynthe-

sis have lead to the identification of multifunctional polyketide synthase enzymes
(PKSs) responsible for the construction of polyketide backbones of defined
chain lengths, the degree and regio-specificity of ketoreduction, and the
regiospecificity of cyclizations and aromatizations, together with the genes encod-
ing for the enzymes. 40 A set of rules for manipulating the early steps of aromatic
polyketide biosynthesis through genetic engineering has been developed, permit-
ting the biosynthesis of polyketides not generated naturally ("unnatural natural
products"). Since polyketides constitute a large number of structurally-diverse
natural products exhibiting a broad range of biological activities (e.g. tetr-

acyclines, doxorubicin, and avermectin), the potential for generating novel mol-
ecules with enhanced known bioactivities, or even novel bioactivities, appears to
be high.
The NCI is promoting this area of research through the award of grants to
consortia composed of multidisciplinary groups devoted to the application of
combinatorial biosynthetic and/or combinatorial chemical techniques to the
generation of molecular diversity for testing in high throughput screens related
to cancer.
CONCLUSION
As illustrated in the foregoing discussion, nature is an abundant source of

novel chemotypes and pharmacophores. However, it has been estimated that only
5 to 15% of the approximately 250,000 species of higher plants have been sys-
tematically investigated for the presence of bioactive compounds/ 6 while the
potential of the marine environment has barely been tapped. 5 The Actinomycetales
have been extensively investigated and have been, and remain, a major source of
novel microbial metabolites;41 however, less than 1% of bacterial and less than
5% of fungal species are currently known, and the potential of novel microbial
sources, particularly those found in extreme environments/ 7 seems unbounded.
To these natural sources can be added the potential to investigate the rational
design of novel structure type,; within certain classes of microbial metabolites
through genetic engineering, as has been elegantly demonstrated with bacterial
polyketides. 40 The proven natural product drug discovery track record, coupled
with the continuing threat to biodiversity through the destruction of terrestrial
and marine ecosystems, provides a compelling argument in favor of expanded
exploration of nature as a source of novel anticancer agents.
REFERENCES
I. CHANG, H-M., BUT, PP-H. 1986. Pharmacology and Applications of Chinese Materia
Medica, Vols I and 2, Singapore, World Scientific Publishing.
2. KAPOOR, L.D. 1990. CRC Handbook of Ayurvedic Medicinal Plants. Boca Raton,
Florida, CRC Press.
3. FARNSWORTH, N.R., AKERELE, 0., BINGEL, A.S., SOEJARTO, D.O., GUO, Z. 1985.
Medicinal plants in therapy. Bull. WHO. 63: 965-98/.
4. KINGHORN A.D. 1994. The discovery of drugs from higher plants. In: The Discovery
of Natural Products with Therapeutic Potential (VP Gullo, ed.), Butterworth-Heinemann,
Boston, pp. 81-108.
5. MCCONNEL, 0., LONGLEY, R.E., KOEHN, F.E. 1994. The discovery of marine natural
products with therapeutic potential. In: The Discovery of Natural Products with Thera-
peutic Potential (VP Gullo, ed.), Butterworth-Heinemann, Boston, pp. /09-174.
6. CRAGG, G.M., NEWMAN, OJ., SNADER, K.M. 1997. Natural products in drug dis-
covery and development. 1. Nat. Prod. 60: 52-60.
7. HARTWELL 1.L. 1982. Plants Used Against Cancer. Quarterman, Lawrence, MA.
8. CRAGG, G.M., BOYD, M.R., CARDELLINA II, 1.H., NEWMAN, OJ., SNADER, K.M.,
MCCLOUD, T.G. 1994. Ethnobotany and drug discovery: The experience of the US
National Cancer Institute. In: Ethnobotany and the Search for New Drugs. Ciba Foun-
dation Symposium 185, (OJ Chadwick, J Marsh, eds.), Wiley & Sons, Chichester, U.K.,
pp. 178-196.
9. DRISCOLL, 1.S. 1984. The preclinical new drug research program of the National Cancer
Institute. Cancer Treat. Rep. 68: 63-76.
10. POTMEISEL, M., PINEDO, H. 1995. Camptothecins: New Anticancer Agents. Boca
Raton, Florida, CRC Press.
II. CRAGG, G.M., BOYD, M.R., CARDELLINA 1I, 1.H., GREVER, M.R., SCHEPARTZ,
S.A., SNADER, K.M., SUFFNESS, M. 1993. Role of plants in the National Cancer Insti-
tute drug discovery and development program. In: Human Medicinal Agents from Plants.
Am Chern Soc Symposium Series 534, (AD Kinghorn, MF Balandrin, eds.), Amer. Chern.
Soc., Washington, DC, pp. 80-95.
12. CHRISTIAN, M.e., PLUDA, 1.M., HO, T.e., ARBUCK, S.G., MURGO, AJ.,
SAUSVILLE, E.A. 1997. Promising new agents under development by the Division of
Cancer Treatment, Diagnosis, and Centers of the National Cancer Institute. Scm. Oncol.
24: 219-240.
13. PHILIP, PA., REA, D., THAVASU, P., CARMICHEL, 1., STUART, N.S.A., ROCKETT,
H., TALBOT, D.e., GANESAN, T., PETTIT, G.R., BALK WILL, E, HARRIS, A.L. 1993.
Phase I study ofbryostatin I: Assessment of interleukin 6 and tumor necrosis factor alpha
induction in vivo. 1. Natl. Cancer Inst. 85: 1812-1818.
14. BOYD, M.R., PAULL, K.D. 1995. Some practical considerations and applications of the
National Cancer Institute in vitro anticancer drug discovery screen. Drug Dev. Res. 34:
91-109.
IS. MAYS, T.D., MAZAN, K.D., CRAGG, G.M., BOYD, M.R. 1997. "Triangular privity"-a
working paradigm for the equitable sharing of benefits from biodiversity research
and development. In: Global Genetic Resources: Access, Ownership, and Intellectual
Property Rights, (KE Hoagland and AY Rossman, eds.), Association of Systematics
Collections, Washington DC, pp. 279-298.
16. BOYD, M.R. 1988. Strategies for the identification of new agents for the treatment of
AIDS: A national program to facilitate the discovery and preclinical development of new
candidates for clinical evaluation. In: "AIDS: Etiology, Diagnosis, Treatment and Pre-
vention" (VT. DeVita, S. Hellman and S.A. Rosenberg, cds.), 1.B. Lippincott, Philadel-
phia, pp. 305-319.
17. WEISLOW, O.S., KISER, R., FINE, D.L, BADER, 1., SHOEMAKER, R.H., BOYD, M.R.
1989. New soluble-formazan assay for HIV-I cytopathic effects: Application to high-flux
screening of synthetic and natural products for AIDS-antiviral activity. 1. Natl. Cancer
Inst. 81: 577-586.
18. CRAGG, G.M., SCHEPARTZ, S.A., SUFFNESS, M., GREVER, M.R. 1993. The taxol
supply crisis. New NCI policies for handling the large-scale production of novel natural
product anticancer and anti-HIV agents. 1. Nat. Prod. 56: 1657-1668.
19. HORWITZ, S.B. 1992. Mechanism of action of taxol. Trends Pharmacol. Sci. 13:
134-136.
20. MCQUIRE, WP, ROWINSKY, EX, ROSENSHEIM, N.B., GRUMBINO, Ee.,
CETTINGER, D.S., ARMSTRONG, D.K., DONEHOWER, R.e. 1989. Taxol: A unique
antineoplastic agent with significant activity in advanced ovarian epithelial neoplasms.
Ann. Intern. Med. III: 273-279.
21. DENIS, IN., GREENE, A.E., GUENARD, D., GUERITTE-VOEGELEIN, E, MAN-

GATAL, L., POTIER, P. 1988. Highly efficient, practical approach to natural taxo!' 1 Am.
Chern. Soc. 1\0: 5917-5919.
22. HOLTON, R.A., LIU, IH., GENTILE, L.N., BEIDIGER, RJ. 1992. Semi-synthesis of
taxo!. Second NCI Workshop on Taxol and Taxus, Alexandria, VA, National Cancer Insti-
tute, September, 1992; GEORG, G.!., CHERUVALLATH, Z.S., HIMES, R.H., MEJIL-
LANO, M.R. 1992. Novel biologically active taxol analogues: Baccatin III
13-[N-p-chlorobenzoyl-(2'R,3'S)-3'-phenylisoserinateland baccatin III \3 [N-benzoyl-
(2'R,3'S)-3'-(p-chlorophenyl)isoserinate]. Bioorg Med Chern Lett. 2: 295-298.
23. CORTES, lE., PAZDUR, R. 1995. Docetaxe!' 1 Clin. Onco!. 13: 2643-2655.
24. Paclitaxel production, marketing heats up. Chern. & Eng. News, 1998, June 15: II
25. STROBEL, G.A., HESS, WM., FORD, E., SIDHU, R.S., YANG, X. 1996. Taxol from
fungal endophytes and the issue of biodiversity. 1 Industrial Microbiology 17: 417-423.
26. BOYD, M.R., HALLOCK, Y.E, CARDELLINA, IH., II, MANFREDI, K.P., BLUNT,
lW, MCMAHON, IB., BUCKHEIT, R.W, JR., BRINGMANN, G., SCHAFFER, M.,
CRAGG, G.M., THOMAS, D.W, JATO, J.G. 1994. Anti-HlV michellamines from
Ancistrocaldus korupensis. 1 Med. Chern. 37: 1740-1745.
27. THOMAS, D.W, GEREAU, R.E. 1993. Ancistrocladus korupensis (Ancistroclada ceae):
A New Species of Liana from Cameroon. Novon. 3: 494-498.
28. BRINGMANN, G., HARMSEN, S., HOLENZ, 1, GEUDER, 1, GOTZ, R., KELLER,
P.A., WALTER, R., HALLOCK, Y.E, CARDELLINA, IH., II, BOYD, M.R. 1994. "Bio-
mimetic" oxidative dimerization of korupensamine A: Completion of the first total syn-
thesis of michellamines A, B, and C. Tetrahedron. 50: 9643-9648.
29. HALLOCK, Y.E, MANFREDI, K.P., BLUNT, lW, CARDELLINA, J.H., II, SCHAF-
FER, M., GLUDEN, K-P., BRINGMANN, G, LEE, A.Y., CLARDY, 1, FRANCOIS, G.,
BOYD, M.R. 1994. Korupensamines A-D, novel antimalarial alkaloids from Ancistro-
cladus korupensis. 1 Org. Chern. 59: 6349-6355.
30. SUFFINESS, M, CRAGG, G.M., GREVER, M.R., GRIFO, FJ., JOHNSON, G., MEAD,
J.A.R., SCHEPARTZ, S.A., VENDITTI, 1M., WOLPERT, M. 1995. The National Co-
operative Natural Products Drug Discovery Group (NCNPDDG) and International
Cooperative Biodiversity Group (ICBG) Programs. Internat. J. Pharmacognosy. 33 Sup-
plement: 5-16.
31. BAKER, n., BORRIS, R.P., CARTE, B, CRAGG. G.M., GUPTA, M.P., IWU, M.M.,
MADULlD, D.R., TYLER, YE. 1995. Natural product drug discovery and development:
New perspectives on international collaboration. 1 Nat. Prod. 58: 1325-1357.
32. KASHMAN, Y., GUSTAFSON, K.R., FULLER, R.W, CARDELLlNA, IH., II,
MCMAHON, J.B., CURRENS, MJ., BUCKHEIT, R.W, HUGHES, S.H., CRAGG,
G.M., BOYD, M.R. 1992. The Calanolides, a novel HIV-inhibitory class of coumarin
derivatives from the tropical rainforest tree, Calophyllum langerum. 1 Med. Chern. 35:
2735-2743.
33. FLAVIN, M.T., RIZZO, ID., KHILEVICH, A., KUCHERENKO, A., SHEINKMAN,
A.K., VILAYCHACK, Y, LIN, L., CHEN, W, GREENWOOD, E.M., PENGSUPARP, T.,
PEZZUTO, J, HUGHES, S.H., FLAVIN, T.M., CIBULSKI, M., BOULANGER, WA.,
SHONE, R.L., XU, Z.-Q. 1996. Synthesis, chromatographic resolution, and anti-human
immunodeficiency virus activity of (±)-calanolide A and its enantiomers. 1 Med. Chern.
39: 1303-1313.
34. TEN KATE, K, WELLS, A. 1998. Benefit-Sharing Case Study. The access and benefit-
sharing policies of the United States National Cancer Institute: a comparative account of
the discovery and development of the drugs Calanolide and Topotecan. Submission to the
Executive Secretary of the Convention on Biological Diversity by the Royal Botanic
Gardens, Kew.
35. ROSENTHAL,1. 1997. OECD Proceedings: Investing in Biological Diversity. The Cairns
Conference, Australia, 25-28 March, 1996. OECD Publications, Paris, pp. 253-273.
36. BALANDRIN, M.E, KINGHORN, A.D., FARNSWORTH, N.R. 1993. Plant-derived
natural products in drug discovery and development. An overview. In: Human Medicinal
Agents from Plants (AD Kinghorn and MF Balandrin, eds.) Am. Chern. Soc. Symposium
Series 534, Amer. Chern. Soc., Washington, DC, pp. 2-12.
37. YOUNG, P. 1997. Major microbial diversity initiative recommended. ASM News. 63:
417-421.
38. SAUSVILLE, E.A. 1997. Targeted toxins. In: Encyclopedia of Cancer, Vo!' III. Acade-
mic Press, Inc., pp. 1703-1714.
39. MELTON, R.G., SHERWOOD, R.E 1996. Antibody-enzyme conjugates for cancer
therapy. 1. Nat!. Cancer Inst. 88: 153-165.
40. MCDANIEL, R., EBERT-KHOSLA, S., HOPWOOD, D.A., KHOSLA, C. 1995. Ratio-
nal design of aromatic polyketide natural products by recombinant assembly of enzymatic
units. Nature. 375: 549-554.
41. HORAN, A.C. 1994. Actinomycetes: A continuos source of novel natural products. In:
The Discovery of Natural Products with Therapeutic Potential (VP Gullo, ed.), Butter-
worth-Heinemann, Boston, pp. 3-30.
Chapter Two
TAXOL BIOSYNTHESIS
A Review of Some Determinant Steps
Kevin Walker and Rodney Croteau
Institute of Biological Chemistry

Washington State University
Pullman, Washington 99164-6340
Introduction .................................................. 32
Alternative Sources of Taxol Supply ............................. 32
Semi-synthesis ............................................ 32
Total Synthesis ......... '. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 34
Biological Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 34
Elucidating the Biosynthetic Pathway ......................... , . . .. 35
Formation of the Universal Diterpene Precursor:
Geranylgeranyl Diphosphate ................................. 35
Taxadiene Synthase: The First Committed Enzyme of
Taxol Biosynthesis ......................................... 36
Taxoids Are Biosynthesized by a Mevalonate-Independent
Pathway .................................................. 37
Hydroxylation of the Taxadiene Nucleus ......................... 37
Subsequent Hydroxylation Steps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 38
Acylation Reactions ........... , ................................ 38
Taxadienyl Acetate as the Third Specific Intermediate in
Taxol Biosynthesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 38
Conversion of IO-Deacetylbaccatin III to Baccatin III . . . . . . . . . . . . .. 40
C4/C20 Epoxidation and Oxetane Ring Formation. . . . . . . . . . . . . . . . . .. 40
Oxidation at C9 ................. , ....... , ................... 41
C 13 Side-Chain Formation and Assembly ,........................ 42
Origin of Side-Chain N-Benzoate Function ........................ 44
Conclusion ........................... , . . . . . . . . . . . . . . . . . . . . . . .. 45
Phytochemicals in Human Health Protection, Nutrition, and Plant Defense, edited by Romeo.
Kluwer Academic I Plenum Publishers, New York, 1999.
31
32 K. WALKER AND R. CROTEAU
INTRODUCTION
Taxol l (Fig. I) is one of the structurally more complex members ofthe taxoid
family characterized by the tricyclic diterpene taxane ring system. Taxol was first
purified from Pacific Yew (Taxus brevi/alia) bark in 1970 2 after extracts from this
material showed cytostatic activity against cancer cells. 3 Classical antineoplastic
drugs (e.g., vinblastine, colchicine, podophyllotoxin, maytansine, and others) act
by inhibiting polymerization of micro tubules. However, in 1979, Taxol was found
to exhibit a novel cytostatic mechanism that involves the promotion oftubulin poly-
merization and stablization of assembled microtubules with consequent blocking
of normal mitotic spindle development and cell division. 4 Discodermolides 5 and
epothilones6 are other natural products with this unusual mode of action. Taxol, as
a new, alternative cell spindle poison, was ultimately promoted to clinical testing.
Taxol and its semi-synthetic congener, TaxotEre (Fig. I ),7 have attracted con-
siderable interest during the last ten years for their use in chemotherapy. 8 Currently,
investigations on dose-dense scheduling have provided both shorter treatment time
and palliative therapy for advanced disease in higher risk patients undergoing
taxane chemotherapy.9 These two taxanes, either alone or in combination with cis-
platin, or anthracyclines, doxorubicin and epirubicin, 10. I I have proven to be
effective in the treatment of ovarian, breast, head, neck, and lung cancers. 10.12.13
The increased utilization ofTaxol, and structurally-related taxoids, in cancer
chemotherapy has precluded obtaining this compound exclusively from the origi-
nallimited source, the bark of the Pacific yew (Taxus brevi/alia). Only 100~ 170
mg Taxol can be isolated from -I kg bark from a mature tree (-0.02% yield).14.15
Roughly 1,000 trees are needed to produce I kg Taxol, and a full ten course treat-
ment requires about 3 g Taxo!. Taxol for the treatment of ovarian cancer alone would
consume -90,000 trees annually,15 placing an exorbitant long term demand on this
limited resource; therefore, alternative sources of supply are being sought.
Alternative Sources of Taxol Supply
Semi-Synthesis. Semisynthesis currently provides the major com-

mercial supply ofTaxo!' lO-Deacetylbaccatin III and baccatin III (Fig. I) can be
isolated from the needles of yew species as a renewable resource and then
Figure 1. Coupling offarnesyl diphosphate to isopentenyl diphosphate to form geranylgeranyl

diphosphate, followed by cyclization to taxa-4(5),11(12)-diene involving a transient 1S-
verticillene intermediate and an intramolecular proton transfer. A single monoprotic base may
be responsible for removal of the I la-hydrogen and reprotonation at C7. Also illustrated is the
cytochrome P450-mediated hydroxylation of taxa-4(5),11(12)-diene, with migration of the
double bond, to form taxa-4(5), II (12)-dien-5a-ol, and the subsequent elaboration of this 5a-
01 to Taxol and related taxoids. A, geranylgeranyl diphosphate synthase; B, taxa-4(5), II (12)-
diene synthase; C, taxa-4(5),II(l2)-diene P450 hydroxylase; OPP, diphosphate moiety; Bz,
benzoyl; Ac, acetyl; * denotes a semi-synthetic compound.
~
o><
l'
to
~{~- o
(/J.
:i...,
::c
tTl
farnesyl and isopentenyl geranylgeranyl (/J.
diphosphates diphosphate Vi
MM/'
G. / ' T "'-B:8
~.B BH~
----+~''''.. 5~ ,
~
04
1 H
H 20 H 20 H H 20
~'-M'--
verticillene taxa-4(5), 11 (l2)-diene
Rl0 OOH
~ ----+
~ I"'" ~
R20"'" 1 ~ H~ 0
~"'~OH ----+
OH OBz OAe
----+
H 20
taxa-4(5), 11 (l2)-dien-5a-ol R, = Ac, R2 = N-benzoylphenylisoserine: Taxol

RJ = Ac, R2 = N-tigloylphenylisoserine: Cephalomannine
R, = H, R2 = H: 10-deactylbaccatin III
R, = Ac, R2 = H: Baccatin III w
w
R, = H, R2 = N-(t-butoxycarbonyi)phenylisoserine: Taxotere*
chemically acylated with various synthetic precursors of the phenylisoserine

side-chain to form Taxol or Taxotere. I6-20
Total Synthesis. Taxol has been prepared by a number of elegant

total syntheses,21-23 but these multi-step routes, requiring precise regio- and
stereospecific control, are costly and low yielding, thus making this approach
unrealistic for commercial production?I-23 Recently, Wender et al. have developed
a synthetic approach that employs pinene as a starting material and that yields
greater quantities of Taxol than previous syntheses. 24 ,25 Even with increased
yields, considerable further optimization will be required for this approach to
become a commercial reality.
Biological Sources, In higher plants, the production of taxanes appears to

be restricted to species of Taxus and Austrotaxus, However, additional producers
ofTaxol have been found among diverse genera of endophytic fungi,z6 including
Taxomyces andreanae, Pestalotiopsis, Fusarium, and Alternaria, isolated
from various Taxus species. 26 Most of these fungal symbionts are able
to produce Taxol de novo, but not in commercially viable amounts, Recently,
fermentation of Alternaria alternata was reported to yield about 116 mg
TaxollL. 14 Molecular genetic approaches aimed at elucidating and exploiting
the Taxol biosynthetic pathway may lead to improved microbial methods for
producing this drug.27
Another biological source of Taxol is Taxus cell cultures which respond
to methyl jasmonate elicitation with enhanced production of the drug. 28- 3o
Ketchum et al. have achieved the most rapid accumulation ofTaxol in cell culture
yet reported/o amounting to 23.4mg/L per day in elicited cultures. The most
abundant taxoids (detected by UV-HPLC) produced in these cultures are Taxol
(20%), l3-acetyl-9-dihydrobaccatin III (a promising new taxane derivative 31 ,32),
and baccatin VI which together account for 39-62% of the total UV-active
taxoids,30
Plant cell cultures also provide an excellent tool for the in vivo and in vitro
elucidation of the complex Taxol biosynthetic pathway by offering consistent
product titers and selection of high-yielding lines. Furthermore, the isolation and
purification of Taxol from cell cultures requires fewer steps than isolation from
needles or bark because the quantities of interfering phenolics, lipids, and other
contaminating metabolites are less per gram of tissue. Most importantly, novel
taxane derivatives potentially can be generated by genetic manipulation of cells33
and/or by optimizing culture conditions,30,34
Given the projected increase in Taxol usage as a chemotherapeutic
agent, and the necessary reliance on biological methods of production, it is impor-
tant to understand the enzymatic reactions involved in Taxol biosynthesis, espe-
cially the rate-limiting steps of the pathway, since manipulation of this pathway
could provide the basis for the improved production oftaxoids. The up-regulation
TAXOL BIOSYNTHESIS 35
of rate-limiting steps by overexpression of the corresponding genes in intact plants

or cell cultures could be expected to raise the titers of these medicinally desirable
compounds to commercially significant levels. s This review summarizes recent
developments in the study ofTaxol biosynthesis.
ELUCIDATING THE BIOSYNTHETIC PATHWAY
Formation of the Universal Diterpenoid Precursor:

Geranylgeranyl Diphosphate
Geranylgeranyl diphosphate (GGPP) synthase catalyzes the coupling offar-

nesyl diphosphate (FPP) and isopentenyl diphosphate (IPP) to form geranylger-
anyl diphosphate (Fig. I). Genes encoding this farnesyl transferase are of interest
because this enzyme produces the branch-point progenitor of a variety of diter-
penes 35 38 and tetraterpenes (carotenoids )/9.40 including the gibberrelin plant hor-
mones,41 the phytol side-chain of chlorophyll and diterpenoid natural products
such as casbene,42 oryxalexins,43 and Taxo1. 44
The universal diterpenoid precursor GGPP is converted by taxadiene syn-
thase to taxa-4( 5), II (l2)-diene (Fig. I) in the first committed step of the Taxol
biosynthetic pathway in Taxus spp. Since Taxol biosynthesis occurs in non-
photosynthetic, terminally differentiated stem tissue, or in non-photosynthetic,
undifferentiated cell cultures, the molecular regulation of GGPP synthase is of
interest, particularly in the context of flux control in the formation of this branch-
point precursor in an instance where high level production to drive essential,
primary metabolic processes is absent. 45 To examine regulation, isolation of the
gene coding for this synthase is necessary.
A hybridization probe showing significant homology to angiosperm GGPP
synthases was discovered fortuitously while screening size-selected, cloned peR
products initially amplified for other purposes from an induced Taxus cell cDNA
library.45 This probe ultimately yielded a cDNA encoding a GGPP synthase from
r canadensis, and the clone was confirmed by functional expression in yeast. The
cDNA specifies an open reading frame of 1,179 nucleotides and a deduced protein
of393 residues (42.6 kDa) containing an N-terminal plastidial transit peptide. 45
To assess the role ofGGPP synthase in the control of pathway flux, the time
course ofTaxol production, mRNA levels, and GGPP synthase and taxadiene syn-
thase activities were evaluated in non-induced and methyl jasmonate-induced T.
canadensis suspension cells. Regulation of gene expression was assessed by RNA-
blot analysis, which showed that steady state mRNA levels for both GGPP syn-
thase and taxadiene synthase of induced cells were discernibly higher than those
of the non-induced controls. 45 The induced activity levels of these early enzymes
of the Taxol pathway were also found to increase with time, with a one day lag
behind peak mRNA levels. 45 The significantly higher activity of GGPP synthase
over taxadiene synthase activity, as well as the coincidence of the time course
curves for both enzymes, suggests that the prenyltransferase is not rate-limiting in
induced Taxol production.
Taxadiene Synthase: The First Committed Enzyme of

Taxol Biosynthesis
Taxadiene synthase catalyzes the cyclization ofGGPP to taxa-4(5),II(l2)-

diene 44 and, in constructing the taxane skeleton, it constitutes the committed step
in the biosynthesis ofTaxol and related taxoids (Fig. I). Taxadiene synthase activ-
ity was first detected in cell-free extracts of T brevifalia stems, and the enzyme
was purified by a series of traditional chromatographic steps.46 The enzyme is
monomeric (-79 kDa), exhibits a pH optimum at 8.5, a requirement for divalent
metal ion (Mg2+) and a low Km value of 3/lM for the substrate. 46 Mechanistic
evaluation of this enzymatic reaction has shown that the cyclization involves an
intramolecular hydrogen transfer and proceeds without detectable free interme-
diates, or the preliminary formation of the endocyclic isomer, taxa-4(20), II (12)-
diene,47 which was earlier proposed as the likely progenitor ofTaxol based on the
abundance of taxane metabolites with the 4(20)-double bond. 48 .49 These results
suggest a mechanism for the taxadiene synthase reaction involving cyclization of
geranylgeranyl diphosphate to a transient verticillyl cation, followed by intramol-
ecular transfer of the CII a-proton to C7 to initiate the transannular B/C-ring
closure to the taxenyl cation, and terminating deprotonation at C5 to yield
taxa-4(5),II(12)-diene directly.47.5o
An homology-based cloning strategy, based upon sequences of terpenoid
synthases of angiosperms, was applied to access a hybridization probe for this
gymnosperm synthase. The resulting screen of a cDNA library from T brevi(a-
lia stems yielded a full-length clone that was functionally expressed in E. cali. 51
The cDNA sequence specifies an open reading frame of 2,586 nucleotides, and
the deduced full-length protein (862 residues, -98.3 kDa) includes a long pre-
sumptive plastidial transit peptide and the typical DDXXD substrate binding
motif.51 Comparison of the translated sequence to the databases showed high
homology to abietadiene synthase (46% identity, 67% similarity), a diterpene syn-
thase from grand fir. 5I To increase the solubility of the recombinant taxadiene syn-
thase the protein has been heterologously overexpressed as a thioredoxin fusion
which resembles the native enzyme in general properties. 52
To assess the role of taxadiene synthase in the control of pathway flux,
Taxol-producing suspension cell cultures of T canadensis were employed; the
enzyme from this source53 appears to be identical to that from T brevifalia stems
which was previously characterized in detail. 51 The time-course of enzyme
activity during Taxol accumulation in developing cell cultures indicated a rise in
taxadiene synthase activity prior to the onset ofTaxol accumulation, and that per-
sisted at high level into stationary phase. 53 As importantly, the synthase activity
measured in vitro exceeded the maximum rate ofTaxol production in viva. 53 These
results suggest that transformations downstream of the taxadiene synthase are

slower than this first committed pathway step.
Taxoids Are Biosynthesized by a Mevalonate-Independent

Pathway
The two enzymes described above (OOPP synthase and taxadiene synthase)
are both translated as preproteins bearing N-terminal targeting sequences that
direct them to the plastids for processing to mature forms. It is now generally
accepted that the monoterpenes, diterpenes, and tetraterpenes of higher plants
are biosynthesized exclusively in plastids via isopentenyl diphosphate 54 .55 derived
from the mevalonate-independent pathway.56,57 Thus, it was entirely consistent
when Zenk and co-workers58 demonstrated that the taxoid, taxuyunnanine C,
derived from [U-\3C]- or [1-\3C]glucose in T chinesis cell cultures, yielded a label-
ing pattern (evaluated by NMR spectroscopy) that was inconsistent with the meval-
onate pathway but that bore salient features reminiscent ofthe alternative pathway
for isoprenoid biosynthesis,56 Particularly compelling was the observation oflong
range 13C_13C coupling in taxuyannine C, which proved that a contiguously-labeled
C 3 precursor (precluded by the classical mevalonate pathway) undergoes intramol-
ecular rearrangement during the formation ofthe precursor isoprene unit (Fig. 2).58
In T chinesis cell cultures, [1 ,2- 13 C]-acetate was incorporated only into the acetate
side-chains, not the tricyclic terpenoid core.
Hydroxylation of the Taxadiene Nucleus
Taxol biosynthesis continues with extensive oxidative modifications of the

progenitor olefin, taxa-4( S), 11 (12)-diene, followed by elaboration ofthe acyl side-
chains. The first of these oxidations constitutes the second specific step of Taxol
biosynthesis and leads to the formation of taxa-4(20), 11 (l2)-dien-Sa-ol. An in
vivo experiment using T brevifalia stem tissue demonstrated conversion of
labeled taxa-4(20), 11 (12)-dien-Sa-ol into advanced taxoids (lO-lS% incorpora-
tion), including 10-deacetylbaccatin III, cephalomanine, and Taxol, thereby
confirming this metabolite as a pathway intermediate. 59 Microsomal preparations
from Taxus stems and cell cultures were shown to catalyze the NADPH and O 2
dependent conversion of taxa-4( S), 11 (12)-diene exclusively to the Sa-ol congener
Figure 2. Labeling pattern of the taxane skeleton of tax-

ayunnanine derived from [U-"C 6 ]glucose. Arrows denote
the long-range "C_"C coupling observed within each iso-
prenoid unit that could only result if the mevalonate-
independent pathway to isopentenyl diphosphate applies.
(Fig. 1).59 This microsomal hydroxylase fulfilled all of the expected requirements
of a cytochrome P450 mixed function monooxygenase (heme thiolate protein),
including blue light (450 nm) reversal of CO inhibition. 59 Whether the mechanism
for this hydroxylation involves an epoxide intermediate or a regiospecific radical
rebound-oxygen insertion at C5 directly is not yet known. The rearranged double
bond and the stereochemistry of this first oxygenated intermediate constrain the
mechanistic possibilities for subsequent elaboration at this site of the oxetane
moiety ofTaxol. 50
Subsequent Hydroxylation Steps
Although experimental evidence has confirmed that the first oxygenation

oftaxa-4(5),II(l2)-diene occurs at C5, no compelling data are available to define
the order in which subsequent oxygenations and acylations take place. However,
consideration of the relative abundances of naturally-occurring taxoids bearing
an oxygen function at each carbon of the taxane ring20.60.61 has suggested the order
of oxygenation after C5 to be CIO followed by C2 or C9. 48 Metabolites bearing
an oxygen at C 13 are next in abundance, while the oxygenations at C7 and C I
are considered to occur late in the Taxol biosynthetic pathway, possibly after
oxetane ring formation. 4R •5o As a further complication to pathway prediction, acy-
lations of existing hydroxyl functions may precede the introduction of new
hydroxyl groupS.48.62
Microsomal preparations from r canadensis cells, when optimized to
sustain P450-type reactions 63 and incubated with taxa-4(5),II(l2)-diene or taxa-
4(20), II (12)-dien-5a-ol in the presence of NADPH and O2, yield polar products
with chromatographic properties and mass spectra corresponding to taxadiene
diol, triol, and tetraol (Fig. 3 ),50 the precise structures of which are under inves-
tigation (Lovy, Hefner, Fuller, and Croteau, unpublished results). Consistent with
these in vitro observations that indicate the central role of cytochrome P450 oxy-
genases in taxoid metabolism,46.59 Eisenreich et al. have recently demonstrated
that the oxygen atoms of several taxoids are, in fact, derived from molecular
oxygen as expected. 64
Acylation Reactions
Taxadieny! Acetate as the Third Specific Intermediate in Taxa! Biosynthe-

sis. Although optimized microsomal preparations, when incubated with taxadi-
ene or taxadienol, yielded products that were identified as a diol, triol, and tetraol,
taxadienyl acetate appeared to be a more efficient substrate for subsequent
cytochrome P450 microsomal hydroxylations 50 in yielding a single major product
that, based on LC-MS evaluation, appeared to be a taxadiene pentaol monoac-
etate (Fig. 3).50 Taxadienyl benzoate was not detectably oygenated by this system,
indicating that the efficiency of the acetate ester as the precursor was not due
~
or><
OJ
HO o
[/J
:i....,
~ ~
----=--=.
---+
A~---+ ::r:
11"'" ; '" II""': "', ....'9 t'I'I
:: ~ :: ?/ [/J
1 H 4 OH 1 H 4 OAe OAe
~ H 20 H 20 20
en
taxadien-5a-ol taxadien-5a-yl acetate presumed taxadiene pentaol monoacetate
HO HO
B
""'OH
---+ "''''OH
---+
20 20
presumptive taxadiene diol presumed taxadiene tetraol

Figure 3. Acetylation of taxa-4(20), II( 12)-dien-5a-ol at C5 to form taxa-4(20), II (12)-dien-5a-yl acetate. The putative pentaol monoacetate formed
by sequential cytochrome P450-dependent hydroxyl at ions is also shown, as are the sequential hydroxylations of taxa-4(20),11 (l2)-dien-5a-ol to the
level of the tetraol. The hydroxylation patterns proposed are based on the relative abundances of naturally-occurring taxoids. A, taxa-4(20),II(l2)-
dien-5a-ol acetyl transferase; B, P450-mediated hydroxylations.
~
'-0
simply to increased lipophilicity but rather to improved recognition of this

substrate by this set of hydroxy lases that appears to operate sequentially. 50 The
efficient utilization of taxadienyl acetate as a substrate for subsequent oxidative
modification prompted a search for the taxadienol transacetylase that catalyzes
this presumptive third step of Taxol biosynthesis.
Croteau and coworkers found that soluble enzyme preparations from
induced T. canadensis and T. cuspidata suspension cell cultures catalyzed the
acetyl CoA-dependent acetylation oftaxadienol; transacylase activity in the mem-
branous fractions was negligible. 65 The product of this acetyl CoA: taxadienol-
O-acetyl transferase reaction was confirmed by radiochromatographic and
GC-MS analysis to be the expected Sa-yl acetate. Following determination of the
time-course of acetyl transferase activity in methyl jasmonate-induced cell cul-
tures, the operationally soluble enzyme was partially purified by a combination
of anion exchange, hydrophobic interaction and affinity chromatographies. 65 The
enzyme was shown to be a SOkDa monomeric protein with pI -4.7, pH optimum
at 9.0, and high affinity for the co-substrates (Km values of 4.2 J..lM and S.S J..lM
for taxadienol and acetyl CoA, respectively).65 This transferase is regiospecific
since it does not acetylate the ClO or CI3 hydroxyl groups of the more advanced
taxoids lO-deacetylbaccatin III and baccatin III. It is insensitive to monovalent
and divalent metal ions, is only weakly inhibited by p-hydroxymercuribenzoate,
N-ethylmaleimide and coenzyme A, and resembles, in general properties, the few
other O-acetyl transferases of higher plant origin that have been examined. 65
A time-course assessment of transacetylase activity and Taxol accumula-
tion in induced T. canadensis cell cultures showed that the induction of the
transacetylase was not in itself sufficient to promote Taxol production, thereby
suggesting that downstream step(s) are of greater significance in the control of
pathway flUX.65 This observation is not surprising when considering the number
of subsequent modifications necessary to convert the relatively unfunctionalized
taxadienyl acetate to the highly functionalized Taxo!.
Conversion of 10-Deacetylbaccatin 111 to Baccatin 111. Another

transacylase, that operates later in the Taxol biosynthetic pathway, catalyzes the
acetylation of lO-deacetylbaccatin-III (lO-DAB) to baccatin-III (cf. Fig. I );66 the
latter is a direct precursor of Taxol. 48 Crude cell-free extracts from roots of T.
baccata saplings, when incubated with lO-DAB and 14C_ or 3H-Iabeled acetyl-
coenzyme A as the acetyl donor, yielded radiolabeled baccatin 111. 66 Product
formation was confirmed by co-chromatography with authentic baccatin III.
The reaction is strictly dependent on the addition of 10-DAB and is specific for
the ClO-hydroxyl group of the taxane ring.66
C4/C20 Epoxidation and Oxetane Ring Formation
Various taxoids derived from taxadiene differ in their oxygenation patterns,

the nature ofthe side-chains (e.g. acetyl, benzoyl, cinnamoyl, 3-(dimethylamino)-
R~00"o7~"o~o
20 20 20 AcO 20
4(5)-ene 4(20)-ene-5a-ol 4(20)-epoxy-5a-ol oxetane ring
Figure 4. Proposed progression ofbiosynthetic transformations at C4, C5, and C20 oftaxoids
leading to the oxetane moiety.
3-phenylpropionyl, and modified phenylisoserinyl), and the functional grouping

at the C4/C20 position. 48 Many of the naturally occurring taxoids have been cat-
egorized into three groups based primarily on structural differences at C4/C20: 49 •67
Group I taxoids contain compounds in which C4/C20 comprises an exocyclic
methylene; in Group II, C4/C20 bears an oxirane (epoxide) function; Group III
metabolites bear an oxetane ring about C4/C20. Gueritte-Voegelein et al.
have proposed the following sequence of biosynthetic transformations in the
construction of Taxol: 49 4(5)-ene ~ 4(20)-ene-5a-ol ~ 4(20)-epoxy-5a-ol ~
oxetane, thus describing a progression from Group I taxoids to Group II and
Group III (Fig. 4). It is likely that the aforementioned C5-acetylation precedes
oxirane formation and ring expansion to the oxetane.
Based on assessment of the abundances of naturally-occurring taxanes,
epoxidation of the taxane skeleton at C4/C20 likely occurs late in the Taxol
pathway, perhaps after all other ring oxygenations have occurred, but before the
conversion of the C9 hydroxyl to a ketone function. 48 Three groups have sug-
gested mechanisms for the elaboration of the 4(20)-epoxide to the unique oxetane
ring (ring D) ofthe taxoids. 48 One mechanism (Fig. 5a) proposes initial hydrolytic
opening of the epoxide, or nucleophilic attack by acetate to form the acetoxy-
diol;68 this transformation has been modeled in a non-enzymatic system using a
3,4-cyclohexenone epoxide derivative. 69 A second mechanism (Fig. 5b) proposes
direct rearrangement of an a-activated epoxide, but attempts to model this reac-
tion have failed. 70 The third (Fig. 5c), and most appealing, mechanism posits an
intramolecular rearrangement of the a-acetoxyepoxide in which the acetate group
migrates from C5 to C4.49 To date, none of these proposed mechanisms has been
substantiated by direct biochemical evidence.
Oxidation at C9
The C9 position of the taxane skeleton is likely hydroxylated prior to the

formation of the oxetane ring (ring D). However, based on the relative abundance
of naturally occurring taxoid metabolites that possess both the C9-oxo group and
the oxetane ring, the C9-hydroxyl is probably oxidized to a ketone after D-ring
formation. 50 The presumptive dehydrogenase responsible for this transformation
is currently under investigation.
a)
x = H or acetvl
b)
x = activated leaving group
c)
Figure 5. Proposed mechanism for oxetane ring formation (see text for details).
C13 Side-Chain Formation and Assembly
An assessment of structure-activity relationships of the taxoids has indi-

cated that the N-benzoyl phenylisoserine side-chain moiety at C13 is essential for
cytostatic activity.71 The biosynthesis of this Taxol side-chain has been investi-
gated by Floss and co-workers,7274 and the phenylisoserine core was found to
originate from phenylalanine. Since cinnamate esters of the taxoids are common,
and phenylalanine is known to be converted to cinnamic acid by phenylalanine
ammonia lyase, the origin ofphenylisoserine via cinnamate was considered. Thus,
E-cinnamic acid was postulated to be converted to the Z-isomer, then epoxidized
from the re-face, followed by epoxide opening with concurrent amination.72 This
pathway was ultimately ruled out since neither E-[ring-2H5]cinnamic acid nor the
Z-[ring-2H5]cinnamic acid epoxide was incorporated into Taxol (Fig. 6).
Conversely, the successful incorporation of both labeled phenylalanine and
~-phenylalanine into the C 13 side-chain ofTaxol (Fig. 6) prompted the search for
an aminomutase that isomerized phenylalanine directly to the ~-isomer (which
could subsequently undergo hydroxylation to phenylisoserine).74 Phenylalanine
aminomutase activity was first observed in cell-free extracts of young T brevifa-
lia sapling stems that were incubated with (S)-[U- 14C]phenylalanine. Following
~""
I """ C0 2 H_ _ _....
~
Oi""" ----1~. 04o,H
""
I
C0 2 H
E-cinnamic acid Z-cinnamic acid Z-cinnamic acid epoxide
~H2 + AcO
~C02H
t::JH 2 0 .?'
~-phenylalanine ~O""""" o
. C~U 6H OH
~H2 ~ N-debenzoyltaxol
l~oO
2
~C02H
~I
phenylisoserine
o NH 0
crY~'
o
OH
Taxol
Figure 6. Summary of C 13 side-chain formation and assembly. The phenylpropanoid moiety
of the side-chain is derived from phenylalanine by initial rearrangement to ~-phenylalanine. The
cinnamate derivatives are not incorporated into taxol in vivo. A, phenylalanine ammonia lyase;
B, phenylalanine aminomutase; C, baccatin 111-13-0-phenylisoserinyl transferase; D, N-
debenzoyltaxol-N-benzoyl transferase.
the reaction, the mixture was diluted with carrier (S)-phenylalanine and (R,s)-I3-
phenylalanine, and these metabolites were derivatized to the corresponding N-
benzoyl methyl esters which were extensively purified to indicate a minimum
of 0.17% conversion to l3-phenylalanine. The configuration of the enzymatically
formed product was established as (3R)-I3-phenylalanine by similar means,
but involved conversion to the (I S)-camphanoate methyl ester and co-
chromatography with the authentic diastereomic derivative. Hence, this mutase
H~
.f
tl; :,yNH2 phenylalanine aminomutase
~
~ S ... , Ho
// C02H
phenylalanine p-phenylalanine
Figure 7. Summary of the stereochemical course of the phenylalanine aminomutase reaction

(see text for details).
produces the enantiomer of f3-phenylalanine that corresponds in configuration to

the phenylisoserine moiety ofTaxol. 74
Insight into the mechanism of the aminomutase reaction was gained by an
experiment in which a I : I mixture of unlabeled phenylalanine and (S)-[2- 1S N, ring-
2Hs]phenylalanine was deployed as substrate. The resulting f3-phenylalanine was
converted to the N-benzoyl methyl ester as before, and GC-MS analysis revealed
that the migration of the amino group from the a- to the f3-carbon is strictly
intramolecular. 74 The steric course of the aminomutase reaction was also deter-
mined, by conversion of (2S,3R)-[ring, 3-2H6]- and (2S,3S)-[ring, 2,VH 7]pheny-
lalanine to the product, and shown to proceed with retention of the C3 pra-R
hydrogen and partial migration of the C3 pra-S hydrogen of phenylalanine to C2
off3-phenylalanine (results summarized in Fig. 7).74 Although little is known about
the properties and cofactor requirements of this enzyme, the aminomutase of Taxus
represents the first example of such an isomerase from higher plants, and it is the
first example of a phenylalanine aminomutase from any source. 74
No information is available as yet on the conversion of f3-phenylalanine
to phenylisoserine. However, Floss and coworkers have provided tentative
evidence, from in viva experiments with T. brevifalia cambial tissue, that the
penultimate step of Taxol assembly involves the addition of phenylisoserine to
C 13 of baccatin III to provide N-debenzoyltaxol, which is then N-benzoylated
(Fig. 6).73
In plant cell culture, the accumulation of baccatin III does not strictly
parallel the production of Taxol; instead, baccatin 1lI accumulates well after cell
necrosis. 30 Although this observation might suggest the role of baccatin III as a
Taxol degradation product and thus seemingly contrasts with the report that bac-
catin III is converted to Taxol in T. brevifalia cambial tissue'73 too little is known
about late-stage pathway steps or potential differences in flux control between
intact tissue and cultured cells to draw firm conclusions at present.
Origin of the Side-Chain N-Benzoate Function
Incorporation experiments with [ring-2Hs]-labe1ed a- and f3-phenylalanines

and [ring-2Hs]phenylisoserine indicate that the side-chain N-benzoate residue
(and presumably the C2-benzoate ofTaxol) originates, in part, from each phenyl-
propanoid. 72 •75 In higher plants, benzoate has been shown to derive from cinnamic
acid, most likely via 3-hydroxy-3-phenyl propionate; subsequent reactions yield
benzoyl CoA and acetyl COA. 76 Since in T. brevifalia, cinnamic acid is not incor-
porated into Taxol, the conversion ofphenylisoserine and (X- and ~-phenylalanine
to the benzoate moiety must occur by an alternate route.
CONCLUSION
Engineering biological systems to improve the yields ofTaxol and related

taxoids used in drug semi-synthesis should be based on a complete understand-
ing of the target pathway, i.e., the enzymes catalyzing each transformation, par-
ticularly those responsible for rate-limiting steps, and the genes encoding these
proteins. The sequence of oxygenations and acylations leading from taxa-
4( 5), II (12)-diene to Taxol can be revealed by a systematic, stepwise approach
which combines information from both in vivo and in vitro studies, and molecu-
lar genetics. This sequential approach can be extended to assess the timing of the
oxidation of the C9 hydroxy and the formation of the oxetane ring, and to define
the coordination of N-benzoylphenylisoserine side-chain production and assem-
bly. The contribution of each enzyme to pathway flux then can be assessed by
in vivo studies, the slow steps identified, and a suitable cloning strategy devised
to acquire the corresponding genes. Subsequent overexpression of genes con-
trolling slow pathway steps in engineered Taxus and/or derived cell cultures would
be expected to increase overall pathway flux and improve the production yields
of Taxol.
ACKNOWLEDGMENTS
We thank Joyce Tamura for preparation of the manuscript. The work by

the authors described herein was supported by Cytoclonal Pharmaceutics Inc.,
McIntire-Stennis Project 0967, and NIH Grant CA 55254.
REFERENCES
I. Paclitaxel is the generic name for Taxol which is a registered trademark of Bristol-
Myers Squibb. Because of the greater familiarity with the word "Taxol", wc use it
throughout.
2. WANI, M.e., TAYLOR, H.L., WALL, M.E., COGGON, P., MCPHAIL, A.T. 1971. Plant
antitumor agents. VI. Isolation and structure of Taxol, a novel antileukemic and
antitumor agent from Taxus brevi{olia. 1. Am. Chern. Soc. 93:2325-2327.
3. WALL, M.E., WAN!, M.e. 1995. Paclitaxcl: From discovery to clinic. In: Taxane
Anticancer Agents: Basic Science and Current Status. ACS Symposium Series 583 (G. I.
Georg, T.T. Chen, I. Ojima, D.M. Vyas., eds.) Washington, DC, pp. 18-30.
4. SCHIFF, P.B., FANT, 1., HORWITZ, S.B. 1979. Promotion of microtubule assembly in
vitro by Taxol. Nature 277:665-667.
5. KOWALSKI, R.1., GIANNAKAKOU, P., GUNASEKERA, S.P., LONGLEY, R.E., DAY,
B. w., HAMEL, E. 1997. The microtubule-stabilizing agent discodermolide competitively
inhibits the binding of pac litaxel (Taxol) to tubulin polymers, enhances tubulin nucleation
reactions more potently than paclitaxel, and inhibits the growth of paclitaxel-resistant
cells. Mol. Pharmacol. 52:613-622.
6. NICOLAOU, K.C., ROSCHANGAR, E, VOURLOUMIS, D. 1998. Chemical biology of
epothilones. Angew. Chern. Int. Ed. Eng. 37:2015-2045.
7. Taxotere is a registered trademark of Rhone-Poulenc Rorer; the generic name for this
compound is docetaxel. Due to its greater familiarity, the word "Taxotere" is used
throughout.
8. MENHARD, B., EISENREICH, w., HYLANDS, P.1., BACHER, A., ZENK, M.H. 1998.
Taxoids from cell cultures of Taxus chinensis. Phytochemistry 49: 113-125.
9. LOFFLER, T.M. 1998. Is there a place for "dose-dense" weekly schedules of the taxoids?
Semin. Oncol. 25:32-34 (suppl. (2).
10. GOLDSPIEL, B.R. 1997. Clinical overview of the taxanes. Pharmacotherapy
17:IIOS-125S.
II. PAGANI, O. 1998. Taxoids in combination with epirubicin: The search for improved
outcomes in breast cancer. Semin. Oncol. 25:23-26 (suppl. 12).
12. DIERAS, V 1998. Taxanes in combination with doxorubicin in the treatment of metasta-
tic breast cancer. Semin. Oncol. 25: 18-22 (suppl. 12).
13. NABHOLTZ, 1.M., MACKEY, J., SMYLIE, M., TONKIN, K. 1998. Taxane-based
three-drug combination in metastatic and adjuvant treatment of breast cancer. Semin.
Oncol. 25:27-31 (suppl. 12).
14. ROHR,1. 1997. Biosynthesis ofTaxol. Angew. Chern. Int. Ed. Engl. 36:2190-2195.
15. STIERLE, A., STROBEL, G., STIERLE, D. 1995. The search for a Taxol-producing
microorgansim among the endophytic fungi of the Pacific yew, Taxus brevifolia. J. Nat.
Prod. 58:1315-1324.
16. GEORG, G.I., ALI, S., ZYGMUT, J., JAYASINGHE, L.R. 1994. Taxol: A novel antitu-
mor agent. Exp. Opin. Ther. Patents 4: 109-120.
17. HOLTON, R.A., BIEDIGER, R.1., BOATMAN, P.O. 1995. Semisynthesis of Taxol and
Taxotere. In: Taxol: Science and Applications. (M. Suffness, ed.) CRC Press, Boca Raton,
FL, pp. 97-121.
18. COMMER<,:ON, A., BOURZAT, J.D., DIDIER, E., LAVELLE, E 1995. Practical semi-
synthesis and antimitotic activity of docetaxel and side-chain analogs. In: Taxane Anti-
cancer Agents: Basic Science and Current Status. (G.1. Georg, T.T. Chen, I. Ojima, D.M.
Vyas, cds.) American Chemical Society, Washington, DC, pp. 233-246.
19. GUENARD, D., GUERITTE-VOEGELEIN, E, POTIER, P. 1993. Taxol and Taxotere:
Discovery, chemistry, and structure-activity relationships. Ace. Chern. Res. 26: 160-167.
20. KINGSTON, D.G.I., MOLINERO, A.A., RIMOLDI, J.M. 1993. The taxane diterpenoids.
In: Progress in the Chemistry of Organic Natural Products. (w. Hcrz, G.w. Kirby, R.E.
Moore, W. Steglich, e. Tamm, eds.) Springer-Verlag, New York, 206 p.
21. HOLTON, R.A., KIM, H.-B., SOMOZA, e., LIANG, E, BIEDIGER, R.1., BOATMAN,
P.O., SHINDO, M., SMITH, e.e., KIM, S., NADIZADEH, H., SUZUKI, Y, TAO, e.,
VU, P., TANG, S., ZHANG, P., MURTHI, K.K., GENTILE, L.N., L1U, J.H. 1994. First
total synthesis of Taxol. 2. Completion of the C and 0 rings. 1. Am. Chern. Soc.
116:1599-1600.
22. NICOLAOU, K.e., YANG, Z., LIU, J.1., UENO, H., NANTERMET, P.G., GUY, R.K.,
CLAIBORNE, C.F., RENAUD, I, COULADOUROS, E.A., PAULVANNAN, K.,
SORENSEN, E.I 1994. Total synthesis ofTaxol. Nature 367:630-634.
23. MASTERS, J.1., LINK, 1.T, SNYDER, L.B., YOUNG, WB., DANISHEFSKY, S.1. 1995.
A total synthesis ofTaxol. Angew. Chern. Int. Ed. Eng. 34:1723-1726.
24. WENDER, P.A., BAD HAM, N.F., CONWAY, S.P., FLOREANCIG, P.E., GLASS, TE.,
HOUZE, J.B., KRAUSS, N.E., LEE, D.S., MARQUESS, D.G., MCGRANE, P.L., MENG,
W, NATCHUS, M.G., SHUKER, A.1., SUTTON, Ie., TAYLOR, R.E. 1997. The pinene
path to taxanes. 6. A concise stereocontrolled synthesis of Taxol. I Am. Chern. Soc.
119:2757-2758.
25. WENDER, P.A., BADHAM, N.F., CONWAY, S.P., FLOREANCIG, P.E., GLASS, TE.,
GRANICHER, e., HOUZE, IB., JANICHEN, 1., LEE, D.S., MARQUESS, D.G.,
MCGRANE, P.L., MENG, W, MUCCIARO, TP., MUHLEBACH, M., NATCHUS, M.G.,
PAULSEN, H., RAWLINS, D.B., SATKOFSKY, J., SHUKER, A.1., SUTTON, J.e.,
TAYLOR, R.E., TOMOOKA, K. 1997. The pinene path to taxanes. 5. Stereocontrolled
synthesis of a versatile taxane precursor. J. Am. Chern. Soc. 119:2755-2756.
26. STROBEL, G.A., HESS, WM., FORD, E., SIDHU, R.S., YANG, X. 1996. Taxol from
fungal endophytes and the issue of Taxol has biodiversity. 1. Ind. Microbiol. Biotech.
17:417-423.
27. LONG, D.M., SMIDANSKY, E.D., ARCHER, A.1., STROBEL, G.A. 1998. in vivo
addition of tclomeric repeats to foreign DNA generates extrachromosomal DNAs in the
Taxol-producing fungus Pestalotiopsis microspora. Fung. Genet. BioI. 24:335-344.
28. MIRJALILI, N., LINDEN, 1.e. 1996. Methyl jasmonate induced production of Taxol in
suspension cultures of Taxas cU5pidata: Ethylene interaction and induction models.
Biotechnol. Prog. 12:110-118.
29. YUKIMUNE, Y., TABATA, H., HIGASHI, Y., HARA, Y. 1996. Methyl jasmonate-
induced overproduction of paclitaxel and baccatin III in Taxas cell suspension cultures.
Nature Biotech. 14: 1129-1132.
30. KETCHUM, R.E.B., GIBSON, D.M., CROTEAU, R.B., SCHULER, M.L. 1999. The
kinetics of taxoid accumulation in cell suspension cultures of Taxus following elicitation
with methyl jasmonate. Biotechnol. Bioeng. 62:97-105.
31. KLEIN, L.L., LI, L., MARING, C.1., YEUNG, e.M., THOMAS, S.A., GRAMPOVNIK,
D.1., PLATTNER, 1.1. 1995. Antitumor activity of 9(R)-dihydrotaxane analogs. 1. Med.
Chern. 38: 1482-1492.
32. ALDER, 1.D., JARVIS, K.P., MARSH, K.e., KLEIN, L.L., CLEMENT, 1.1. 1996. Pre-
clinical in vivo efficacy of two 9-dihydrotaxane analogues against human and murine
tumours. Brit. 1. Cancer 73:560-564.
33. HAN, K.-H., FLEMING, P., WALKER, K., LOPER, M., CHILTON, WS., MOCEK, u.,
GORDON, M.P., FLOSS, H.G. 1994. Genetic transformation of mature Taxus: An
approach to genetically control the in vitro production of the anticancer drug Taxol. Plant
Sci. 95:187-196.
34. MA, W, PARK, G.L., GOMEZ, G.A., NIEDER, M.H., ADAMS, TL., AYNSLEY, 1.S.,
SAHAI, O.P., SMITH, R.J., STAHLHUT, R.W, HYLANDS, P.J. 1994. New bioactive
taxoids from cell cultures of Taxas baccata. 1. Nat. Prod. 57:116-122.
35. CLARKE, S. 1992. Protein isoprenylation and methylation at carboxyl-terminal cysteine
residues. Annu. Rev. Biochem 6\:355-386.
36. RILLING, H.C., BREUNGER, E., EPSTEIN, WW, CRAIN, PF. 1989. Prenylated pro-
teins: Demonstration of a thioether linkage to cysteine of proteins. Science 247 :318-320.
37. KLEINIG, H. 1989. The role of plastids in isoprenoid biosynthesis. Annu. Rev. Plant
Physiol. Plant Mol. BioI. 40:39-59.
38. SCHULTZ, G., SOLL, 1., FIEDLER, E., SCHULZE-SIEBERT, D. 1985. Synthesis of
prenylquinones in chloroplasts. Plant Physio!. 64:123-129.
39. BARTLEY, G.E., SCOLNIK, P.A., GIULIANO, G. 1994. Molecular biology of
carotenoid biosynthesis in plants. Annu. Rev. Plant Physio!. Plant Mo!. Bio!. 45:287-301.
40. BARTLEY, G.E., SCOLNIK, P.A. 1995. Plant carotenoids: Pigments for photoprotection,
visual attraction, and human health. Plant Cell 7:1027-1038.
41. SUN, T.-P., KAMIYA, Y. 1994. The Arabidopsis GAl locus encodes the cyclase
ent-kaurene synthetase A of gibberellin biosynthesis. Plant Cell 6: 1509-1518.
42. DUDLEY, M.W., GREEN, T.R., WEST, c.A. 1986. Biosynthesis of the macrocyclic diter-
pene casbene in castor bean (Ricinus communis L.) seedlings. The purification and prop-
erties of farnesyl transferase from elicited seedlings. Plant Physio!. 81 :343-348.
43. WEST, C.A., LOIS, A.E, WICKHAM, K.A., REN, Y.-Y. 1990. Diterpenoid phytoalex-
ins: Biosynthesis and regulation. Recent Adv. Phytochem. 24:219-248.
44. KOEPP, A.E., HEZARI, M., ZAJICEK, 1., VOGEL, B.S., LAFEVER, R.E., LEWIS, N.G.,
CROTEAU, R. 1995. Cyclization of geranylgeranyl diphosphate to taxa-4(5), II( 12)-
diene is the committed step of Taxol biosynthesis in Pacific yew. 1. Bio!. Chern.
270:8686-8690.
45. HEFNER, 1., KETCHUM, R.E.B., CROTEAU, R. 1998. Cloning and functional expres-
sion of a cDNA encoding geranylgeranyl diphosphate synthase from Taxus canadensis
and assessment ofthe role of this prenyltransferase in cells induced for Taxol production.
Arch. Biochem. Biophy. 360:62-74.
46. HEZARI, M., LEWIS, N.G., CROTEAU, R. 1995. Purification and characterization of
taxa-4(5),II(l2)-diene synthase from Pacific yew (Taxus brevifolia) that catalyzes the first
committed step of Taxol biosynthesis. Arch. Bioch. Biophys 322:437--444.
47. LIN, X., HEZARI, M., KOEPp, A.E., FLOSS, H.G., CROTEAU, R. 1996. Mechanism of
taxadiene synthase, a diterpene cyclase that catalyzes the first step of Taxol biosynthesis
in Pacific yew. Biochemistry 35:2968-2977.
48. FLOSS, H.G., MOCEK, U. 1995. Biosynthesis of Taxa!. In: Taxal: Science and Appli-
cations. (M. Suffness, ed.) CRC Press, Boca Raton, FL, pp. 191-208.
49. GUERITTE-VOEGELEIN, E, GUENARD, D., POTIER, P. 1987. Taxol and derivatives:
A biogenetic hypothesis. 1. Nat. Prod 50:9-18.
50. HEZARI, M., CROTEAU, R. 1997. Taxol biosynthesis: An update. Planta Medica
63:291-295.
51. WILDUNG, M.R., CROTEAU, R. 1996. A cDNA clone for taxadicne synthase, the
diterpene cyclase that catalyzes the committed step of Taxol biosynthesis. 1. Bio!.
Chern. 271:9201-9204.
52. HUANG, K.-X., HUANG, Q.-L., WILDUNG, M.R., CROTEAU, R., SCOTT, A.1. 1998.
Overproduction, in Escherichia coli, of soluble taxadiene synthase, a key enzyme in the
Taxol biosynthetic pathway. Protcin Express. Purif. 13:90-96.
53. HEZARI, M., KETCHUM, R.E.B., GIBSON, D.M., CROTEAU, R. 1997. Taxol produc-
tion and taxadiene synthase activity in Taxus canadensis cell suspension cultures. Arch.
Biochem. Biophys. 337:185-190.
54. MC CASKILL, D., CROTEAU, R. 1999. Strategies for bioengineering the development
and mctabolism of glandular tissues in plants. Nature Biotechno!. 17:31-36.
55. MC CASKILL, D., CROTEAU, R. 1999. Isopentenyl diphosphate is the terminal product
of the deoxyxylulose-5-phosphate pathway for terpenoid biosynthesis in plants. Tetrahe-
dron Lett. 40:653-656.
56. ROHMER, M., KNANI, M., SIMONIN, P., SUTTER, B., SAHM, H. 1993. Isoprenoid
biosynthesis in bacteria: A novel pathway for the early steps leading to isopentenyl diphos-
phate. Biochem. 1. 295:517-524.
57. EISENREICH, w., SCHWARZ, M., CARTAYRADE, A., ARIGONI, D., ZENK, M.H.,
BACHER, A. 1998. The deoxyxylulose phosphate pathway of terpenoid biosynthesis in

plants and microorganisms. Chern. BioI. 5:R221~R223.
58. EISENREICH, W, MENHARD, B., HYLANDS, PJ., ZENK, M.H., BACHER, A. 1996.
Studies on the biosynthesis of Taxol: The taxane carbon skeleton is not of mevalonoid
origin. Proc. Nat. Acad. Sci. USA 93:6431~6436.
59. HEFNER, 1., RUBENSTEIN, S.M., KETCHUM, R.E.B., GIBSON, D.M., WILLIAMS,
R.M., CROTEAU, R. 1996. Cytochrome P450-catalyzed hydroxylation of taxa-
4(5), II (12)-diene to taxa-4(20), II (12)-dien-5a.-ol: The first oxygenation step in
Taxol biosynthesis. Chern. BioI. 3:479-489.
60. APPENDINO, G. 1995. The phytochemistry of the yew tree. Nat. Prod. Rep. 12:349~360.
61. SHI, Q.-W, ORITANI, T., SUGIYAMA, T., KIYOTA, H. 1998. Three novel bicyclic 3,8-
secotaxane diterpenoids from the needles of the Chinese yew, Taxus chinensis var. mairei.
1. Nat. Prod. 61:1437~1440.
62. CROTEAU, R., HEFNER, 1., HEZARI, M., LEWIS, N.G. 1996. Taxol biosynthesis:
Cyclization and early hydroxylation steps of the pathway. Curr. Top. Plant Physiol.
15:94~\o4.
63. HEFNER, 1., CROTEAU, R. 1996. Microsome preparation from woody plant tissues.
Methods Enzymol. 272:243~250.
64. EISENREICH, W, MENHARD, B., LEE, M.S., ZENK, M.H., BACHER, A. 1998.
Multiple oxygenase reactions in the biosynthesis of taxoids. 1. Am. Chern. Soc.
120:9694~9695.
65. WALKER, K., KETCHUM, R.E.B., HEZARI, M., GATFIELD, D., GOLENIOWSKI, M.,
BARTHOL, A., CROTEAU, R. 1999. Partial purification and characterization of taxa-
4(20),II(l2)-dien-5a.-ol: O-acetyltransferase: The third enzyme of Taxol biosynthesis.
Arch. Biochem. Biophys., in press.
66. ZOCHER, R., WECKWERTH, W, HACKER, c., KAMMER, 8., HORNBOGEN, T.,
EWALD, D. 1996. Biosynthesis ofTaxol: Enzymatic acetylation of 10-deacetylbaccatin-
III to baccatin-III in crude extracts from roots of Taxus baccala. Biochem. Biophys. Res.
Comm.229:16-20.
67. MILLER, R.W 1980. A brief survey of Taxus alkaloids and other taxane derivatives. 1.
Nat. Prod 43:425-437.
68. DELLA CASA DE MARCANO, D.P., HALSALL, T.G., CASTELLANO, E., HODDER,
OJ.R. 1970. Crystallographic structure determination of the diterpenoid baccatin V, a
naturally occurring oxetane with a taxane skeleton. 1. Chern. Soc., Chern. Commun.
1382~1383.
69. BERKOWITZ, WF., AMARASEKARA, A.S. 1985. An approach to the D-ring of
baccatin III (Taxol, cephalomannine). Tetrahedron Lett. 26:3663~3664.
70. SWINDELL, C.S., BRITCHER, S.F. 1986. Construction of the taxane C-ring epoxy
alcohol moiety and examination of its possible involvement in the biogenesis of the taxane
3-oxetanol structure. 1. Org. Chern 51:793~797.
71. KINGSTON, D.G. 1994. Taxol: The chemistry and structure-activity relationships of a
novel anticancer agent. Trends Biotechnol. 12:222~227.
72. FLEMING, PE., MaCEK, u., FLOSS, H.G. 1993. Biosynthesis of taxoids. Mode of
formation of the Taxol side-chain. 1. Am. Chern. Soc. 115:805~807.
73. FLEMING, PE., KNAGGS, A.R., HE, X.-G., MOCEK, u., FLOSS, H.G. 1994. Biosyn-
thesis of taxoids. Mode of attachment of the Taxol side-chain. 1. Am. Chern. Soc.
116:4137-4138.
74. WALKER, K.D., FLOSS, H.G. 1998. Detection of a phenylalanine aminomutase in
cell-free extracts of Taxus brevi/alia and preliminary characterization of its reaction.
1. Am. Chern. Soc. 120:5333~5334.
75. WALKER, K. 1997. The use of stereospecifically-labeled phenylpropanoid compounds
to determine the steric course of key reaction steps in the biosynthesis of Taxol,
hyoscyamine, and cycloheptyl fatty acids. Ph.D. Thesis, University of Washington,
Seattle, 142 p.
76. BJORKLUND, lA., LEETE, E. 1992. Biosynthesis of the benzoyl moiety of cocaine
from cinnamic acid via R-(+)-3-hydroxy-3-phenylpropanoic acid. Phytochemistry
31 :3883-3887 (and references therein).
Chapter Three
ROLE OF LIGNANS IN CARCINOGENESIS
Lilian U. Thompson
Department of Nutritional Sciences

Faculty of Medicine
University of Toronto
Toronto, Ontario
Canada M5S 3E2
Introduction .. , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 51
Sources of Mammalian Lignan Precursors . . . . . . . . . . . . . . . . . . . . . . . . . .. 53
Anticancer Effects of Lignans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 54
Mammary Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 54
Colon Cancer ............................................... 56
Skin and Prostate Cancers ..................................... 57
Mechanism of Lignan Effect. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 58
Application in Cancer Prevention and Treatment. . . . . . . . . . . . . . . . . . . . .. 60
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 61
INTRODUCTION
Epidemiological studies suggest that diets rich in whole grains, fruits, and
vegetables are associated with a lower risk of cancer. I Although the low fat and
high fiber content of such diets appear to contribute to their health benefits, many
studies suggest that their non-nutritive components or phytochemicals also
contribute to anticancer effects. The lignans are thought to be one of these
phytochemicals.
Lignans are phenolic compounds formed by the union of two cinnamic acid
residues and, thus, have a dibenzylbutane skeleton structure. 2 What has been sug-
gested to be cancer protective, however, are the mammalian Iignans enterodiol
Phytochemicals in Human Health Protection, Nutrition. and Plant Defense, edited by Romeo.
51
52 L. U. THOMPSON
(ED) and enterolactone (EL) which are produced when plant lignans are acted
upon by the bacterial flora in the colon of humans or animalsY ED is produced
from the plant lignan secoisolariciresinol diglycoside (SDO) and EL from
matairesinol (Fig. I), although there may be other plant lignan precursors that
have not yet been identified. ED may be converted to EL, but not the reverse,
under the colonic conditions. The mammalian lignans differ from plant lignans
in that mammalian lignans have the hydroxyl groups in the meta position while
plant lignans have the oxygenated substituents primarily in the para position
(Fig. I). Once mammalian lignans are produced in the colon, they undergo
enterohepatic circulation i.e. they are absorbed, transported to the liver, and
secreted in bile. 5.6 A portion reaches the kidney and eventually is excreted in the
urine. Thus, urinary excretion of ED and EL has been used as an index of plant
lignan intake.
Secoisolariciresinol Diglycoside (SJX;) Matairesinol

OH
HO~OH o
H,CO CHi) ----L---0~CHi)H
OH
HO~OH
HO CH20 ----L----°~CH20H
OH
colonic hydrolysis colonic dehydroxylation

bacteria dehydroxylation bacteria demethylation
demethylation
o
H
H CHi)H
CHi)H oxidation
~
colonic bacteria
OH
OH
Enterodiol (ED) Enterolactone (EL)
Figure 1. Metabolism of plant lignans to mammalian lignans.
ROLE OF LIGNANS IN CARCINOGENESIS 53
That mammalian lignans have a role in carcinogenesis was indicated by the

following early observations. The chemical structure of the lignans is similar to
estradiol and tamoxifen, a breast cancer drug, suggesting that they may have weak
estrogenic/antiestrogenic properties (Fig. 2). Non-human primates, on their usual
diet, have high urinary lignan excretion and high resistance to the carcinogenic
effects of estrogens, even when combined with carcinogens. 7 The Finnish and
Americans, on traditional diets and with a high risk of cancer, have lower urinary
lignan excretion than the Japanese with a low risk of cancer. 8- 1O The Japanese diet
is rich in isoflavones, phytoestrogens with similar chemical structures as the
Iignans. Similarly, the urinary lignan excretion in breast cancer patients and omni-
vores (with high cancer risk) is significantly lower than in vegetarians (with low
cancer risk).8,lO This was supported by a recent case-control study on breast cancer
patients which showed a three-fold reduction in breast cancer risk with increased
urinary excretion oflignans. 11 Also, the incidence of prostate cancer in Hongkong
and Portugal is lower than in the UK and, interestingly, the plasma and prostatic
fluid lignan and/or isoflavone concentration in men in these countries are higher
than those in the UK. 12
While the above observations suggest a relationship between Iignans and
cancer, they do not establish causality. Therefore, about a decade ago, we
embarked on a research program to test the hypothesis that lignans have anti-
cancer effects. Since the mammalian lignans or their major precursors were not
commercially available at that time, our plan was to identify a food source that
is rich in the plant lignan precursor, to use this as a food model to determine the
anticancer effect of lignans, and, if the results were encouraging, to then isolate
the precursor and test its anticancer effects and potential mechanism(s) of action.
This paper will describe the results of these studies and those of others.
SOURCES OF MAMMALIAN LIGNAN PRECURSORS
The mammalian lignans are produced in the human colon; thus, an in vitro
fermentation method using human fecal inoculum to simulate human colonic
HO~
OH
1713-Estradiol
Figure 2. Chemical structure of estradiol and tamoxifen.
54 L. U. THOMPSON
Table 1. Mammalian lignan production from various foods (ug/100g)

As is basis Dry Basis
Food group Number of foods Range Mean Range Mean
Defatted flaxseed 67,541 67,541 78,045 78,045
Full fat flaxseed 52,679 52,679 57,857 57,857
Other oilseeds 4 161-1,130 638 168-1,288 709
Legumes 7 201-1,787 562 226-1,956 624
Legume hulls 5 222-535 371 245-596 413
Cereal brans 5 181-651 486 204-743 542
Whole grain cereals 9 115-924 359 129-1,052 407
Vegetables 27 21-407 144 200-6,344 1,533
Fruits 7 35-181 84 229-1,080 548
Seaweeds (dried) 2 653-1,147 900 729-1,267 998
'Summarized from reference."
fermentation was used to screen 66 common plant foods for their precursors. 13
This showed that precursors are widely distributed in oil seeds, whole grains,
legumes, fruits, vegetables, and seaweeds, but flaxseed, also known as linseed,
contains the highest concentration with values 75-800 times higher than the other
foods (Table 1). The result was in agreement with an early observation that urinary
lignan excretion in rats fed flaxseed was over 100 times higher than the other ten
grains that were tested. 3 This finding was further confirmed in a recent direct
analysis of the precursors SDO and matairesinol in many plant foods. 14 .15
Although the lignan concentration differs with variety, planting location and year
of harvest,16 flaxseed remains the richest source of the mammalian lignan pre-
cursors, particularly the SDO. Human feeding studies showed a 5 to 195 fold
increase in urinary lignan excretion upon diet supplementation with 10 to 50 g of
flaxseed per day.13.17.20 All these findings suggest that flaxseed is a good' food
model to test the anticancer effects of lignans.
ANTICANCER [~FFECTS OF LIGNANS
Mammary Cancer
Because the effect of flaxseed may differ depending on whether it is given

before or after carcinogen exposure, flaxseed was tested in rats at different stages
of carcinogenesis. An initial study involved the feeding of diets supplemented
with 5% or 10% flaxseed or defatted flaxseed for 4 weeks to determine whether
flaxseed can reduce cell proliferation prior to injection with the mammary car-
cinogen dimethylbenzanthracene (DMBA) and reduce the number of nuclear
aberrations after carcinogen injection. 21 Cell proliferation and nuclear aberrations
are considered early markers of cancer risk since highly proliferating cells are
more susceptible to carcinogen damage,22 while nuclear aberrations are lethal
events that have been associated with carcinogen exposure and susceptibility of
the cells to carcinogen.23 Both cell proliferation and nuclear aberration were sig-
nificantly reduced by the flaxseed diets. 21 The urinary lignan excretion increased
with flaxseed dose and was negatively related to the number of nuclear aberra-
tions in the mammary epithelial cells. This suggests the potential of lignans in
flaxseed to reduce the number of tumors if provided prior to carcinogen expo-
sure. Indeed, when the experiment was repeated with 5% flaxseed fed only at the
pre-initiation stage and with tumor as the endpoint (i.e. up to 20 weeks after
DMBA treatment), a reduction in the number of tumors, but not in tumor size,
was observed compared with the control. 24 In contrast, when provided only after
carcinogen exposure (early promotion stage), 5% flaxseed significantly reduced
the tumor volume by over 50%, but not the tumor number. 24
To establish that the anticancer effect of flaxseed was due to the mammalian
lignans produced from SDG, the SDG was isolated from flaxseed 25 and tested for
its effect at the early promotion stage. 26 SDG, fed at a level equivalent to that con-
sumed in the 5% flaxseed diet (i.e. 1.5mg/day), resulted in 47% fewer tumors
than the control. It did not reduce significantly the tumor incidence and volume.
The results were in contrast to that seen with flaxseed,24 and are perhaps related
to the fact that in the flaxseed study, the basal diet (control) was fed for 4 weeks
prior to carcinogen injection, while in the SDG study, the basal diet was fed for
only one week. Nonetheless, this provided the first direct evidence that a mam-
malian lignan precursor can influence carcinogenesis although the nature of the
effect depended on the experimental design.
Flaxseed (supplemented at a level of 2.5% and 5%) and the SDG and
flaxseed oil equivalent to the amount in 5% flaxseed diet were also tested for their
effect on tumor development when fed to rats starting 13 weeks after carcinogen
treatment i.e. when the tumor size was about I cm in diameter. 27 After seven weeks
of treatment, all treatment groups had established tumor volumes that were more
than 50% smaller than the initial volume, while the tumor volumes in the control
group did not significantly change (Table 2). The fact that both the SDG and oil
groups had tumor sizes which did not differ significantly from that of the 5%
flaxseed group indicate that both components contribute to the effect seen with
flaxseed. However, the number and volume of new tumors that appeared during
the seven week treatment were the lowest in the SDG group, suggesting a stronger
cancer protective effect of the SDG than the oil when provided at a late promo-
tion stage of carcinogenesis. Considering all tumors (i.e. established and new),
the volumes in animals fed the 2.5% flaxseed, 5% flaxseed, and SDG equivalent
to that in 5% flaxseed level, were significantly lower than those fed the control
or flaxseed oil diet. A significant negative relationship was observed between the
urinary mammalian lignan excretion and the established tumor size at the end of
the study, providing further evidence that mammalian lignans are effective in
56 L. U. THOMPSON
Table 2. Tumor characteristics of rats fed the high fat basal diet (BD) alone or
supplemented with secoisolariciresinol diglycoside (SDG, 2,OOOnmollday), flaxseed
oil (OIL, 1.82% w/w) or flaxseed (2.5% or 5% , w/w)*
Established
tumor
volume, em3
New tumor Total tumor New tumor
Dietary group Initial Final volume, em 3 volume, em 3 number
BD 0.741 0.635 0.213 0.321 2.4
SDG 0.668 0.304 0.052 0.115 1.2
OIL 0.644 0.299 0.187 0.223 2.7
2.5% F 0.700 0.187 0.094 0.126 1.8
5.0% 0.569 0.149 0.128 0.135 2.4
*All values are geometric means, except the new tumor number.
Summarized from reference,"
reducing established tumor growth. The flaxseed oil does not have lignans; thus,
its inhibitory effect on established tumor growth may be related to the metabolic
effect of its alpha linolenic acid which is present in high concentration (57% of
total fatty acids).
Colon Cancer
The mammalian lignans are produced from plant precursors in the colon;34
therefore, this is a body site where they may exert their primary effects. To deter-
mine the role of lignans on colon carcinogenesis, rats were treated with the colon
carcinogen azoxymethane and then fed diets supplemented with either 5% or 10%
flaxseed or defatted flaxseed for 4 weeks.2x Upon sacrifice, the number of aber-
rant crypts (AC) or AC per focus (ACF) in their colons was determined. ACF are
thought to be valid putative preneoplastic markers of colon carcinogenesis. 29
Compared with the control, a significantly lower number of AC (41-53%) and
ACF (48-57%) were observed with the flaxseed diets, indicating a protective
effect on colon cancer. However, the urinary lignan excretion or the flaxseed levels
did not have a linear relationship with the number of AC or ACF suggesting that
there is a dose limit to the protective effect of the lignans. The effect of flaxseed
also did not differ significantly from that of defatted flaxseed, indicating that the
effect was not due to the oil content.
To further establish the role of the SDG, in a second study, groups of
azoxymethane-treated rats were fed either the basal diet (control) or the basal diet
supplemented with a lower level of flaxseed (2.5% or 5%) or defatted flaxseed,
or SDO equivalent to that in the 5% flaxseed (i.e. 1.5 mg/day).30 The defatted
flaxseed diets had the same lignan content as the flaxseed diets but lacked the
flaxseed oil, so the role of the flaxseed oil could be further differentiated from
that of the SDG. After a longer feeding time of 100 days, a lower number of ACF
and AC per ACF was again observed in all treatment groups compared with the
control. The control group already had four microadenomas and two polyps, while
the treatment groups did not have any, indicating that the control group was pro-
gressing faster towards tumor formation than those fed either the flaxseed or the
SDG. The observed effect was thought to be unrelated to the fiber content of
flaxseed since no significant differences were observed among diets in pH and
cecal short chain fatty acids, the products of fiber fermentation. On the other hand,
a significant negative relationship was observed between the urinary lignan excre-
tion and the number of AC per ACE Together with the fact that SDG caused an
effect which did not differ significantly from that of the 5% flaxseed, this rela-
tionship further suggests that the mammalian lignans derived from the SDG in
flaxseed have colon cancer protective effect.
Although the above studies did not have tumor formation as the endpoint,
a subsequent in vitro study showed that the cell proliferation of four human colon
tumor cell lines (HCT-15, LS 174T, Caco-2, and T-84) can be inhibited by the
mammalian lignans, with EL being more effective than ED. 31 Cell proliferation
was inhibited by 55-88% at 100 uM ED or EL concentration, a level that is high
but is achievable in the colon with the intake of up to 25-50g flaxseed. 31 The
effect was not estrogen-related since estradiol has no effect on the cell prolifera-
tion of these tumor cells.
Skin and Prostate Cancers
The effect of various levels of flaxseed (2.5%, 5% or 10%) on the lung

metastases of melanoma cell line B 16BL6 injected directly to the blood stream
was determined in mice. 32 Feeding the diets two weeks before and two weeks after
the melanoma cell injection resulted in a 32%, 54%, and 63% lower number of
lung tumors in mice fed the above levels, respectively, compared with the control.
The flaxseed diets also lowered the tumor cross-sectional area and volume in a
dose-dependent manner. When SDG was provided to the mice at levels equiva-
lent to that found in the above flaxseed diets, a similar dose-dependent reduction
in number, area, and volume of lung tumors was observed (unpublished data).
This suggests that flaxseed has the ability to reduce experimental metastasis and
that the mammalian lignans from the SDG may in part be responsible for this
effect. Whether the lignans can inhibit other types of metastatic cancer, such as
breast cancer, remains to be determined.
The effect of flaxseed or SDG on other types of cancer has not been tested
in vivo. However, rye bran, which contains a significant amount of lignans, was
tested for its ability to influence prostate tumor development in rats transplanted
with Dunning R3327PAP prostate tumor. 33 At 33% dietary level, rye (both non-
heat and heat-treated) did not stop but did delay the prostate tumor development.
In contrast, the 33% rye endosperm with no lignans had no significant effect.
58 L. U. THOMPSON
While this result may be suggestive of a Iignan effect on prostate cancer, it is not
conclusive.
MECHANISM OF LIGNAN EFFECT
Many mechanisms have been suggested for the effect of lignans on car-
cinogenesis although they are based primarily on results derived from in vitro
studies. 34- 47 Lignans have weak estrogenic/antiestrogenic as well as non-hormone
related properties which are summarized in Table 3. 35-47 At high concentrations
and in the presence of estradiol, EL and ED appear to have anti estrogenic and
antiproliferative effects on estrogen receptor positive breast cancer cells, while at
low concentrations, they are more estrogenic. 35 -39 They bind to human and rat
alpha-fetoprotein, a plasma carcino-embryonic protein that regulates the growth
of proliferative and tumor cells. 40 This is important since competition with estro-
gen for binding to alpha fetoprotein may help modulate the proliferative-
enhancing activity of the estrogen on the target tissue. They have been shown to
bind to nuclear type II estrogen binding site,41 a component of the genome that
regulates estrogen-stimulated growth. They inhibit the activities of enzymes
involved in sex hormone synthesis, such as aromatase, 5-alpha reductase, and 17-
beta hydroxy steroid dehydrogenase. 42 -44 Aromatase is a cytochrome P-450
enzyme that catalyzes the formation of estrogen from androstenedione; thus, it
can lower the amount of available estrogen that may increase breast tumor growth.
The 5-alpha reductase is needed in the conversion of testosterone to the main
prostatic androgen, 5 alpha dihydrotestosterone; thus, its inhibition may have
implications in prostate cancer development. The l7-beta hydroxy steroid dehy-
Table 3. Biological properties of Iignans

Stimulate growth of MCF-7 cells; Inhibit growth of MCF-7 cells in presence of estradiol l4
In low concentrations, stimulate DNA synthesis in MCF-7 cells; in high concentrations,
inhibit DNA synthesis in MCF-7 cells"
Stimulate growth of MCF-7 and T47D cells; stimulate MCF- 7 cell progesterone receptor3"
Inhibit proliferation of ZR-71-1 breast cancer cellline l7
Induce pS2 expression in MCF-7 cells lR
Bind to alpha- feto protein; competes with estradiol and estrone l9
Bind to rat nuclear type II binding site (bioflavonoid reccptor)4U
Stimulate SHBG synthesis by HepG2 livcr cells4()
Inhibit aromatase enzyme in placental microsome, lEG-3 human choriocarcinoma cells,
human preadipose tissue 4 '-42
Inhibit 5-alpha reductase and 17 -beta hydroxy steroid dehydrogenase in genital skin
fibroblast 43
Inhibit steroid binding to sex steroid binding protein44
Inhibit proliferation of bovine and human vascular endothelial cells 4s
Hydroxy radical scavenging activity'6
drogenase catalyzes the reversible reactions between 17-keto and 17-hydroxy

steroids and, therefore, is involved in both androgen and estrogen biosynthesis
and likely in the pathophysiology of estrogen-responsive breast cancer. Lignans
can stimulate the synthesis of sex hormone binding globulin (SHBG) in HepG2
liver cells. 41 Since SHBG has been shown to exert a negative control on estradiol
action in MCF -7 human breast cancer cells, their increased production is thought
to be protective against hormone-related cancers.
Other mechanisms unrelated to sex hormone changes also have been
suggested. Enterolactone has been shown to reduce endothelial cell proliferation
and thus angiogenesis, the formation of blood vessels that provide nutrients to
the growing tumors. 46 Purified SDG 47 and flaxseed extracts48 also have been
shown to have antioxidant effects.
Despite strong evidence of potential mechanisms from in vitro studies, sup-
porting in vivo data are limited. Because the main mechanism suggested for the
anticancer effect of lignans is its antiestrogenic effect, in vivo studies have con-
centrated on demonstrating that the flaxseed or the ED and EL formed from SDG
have hormonal effects. In studies with men, no sex hormone changes were
observed upon feeding 13.5 g/day flaxseed in bread for 6 weeks.17 However, in
premenopausal women fed 109/day raw flaxseed over three menstrual cycles, a
lengthening of the luteal phase length but not the menstrual cycle length was
observed. 49 No significant changes in plasma sex hormones and SHBG were seen
although there was a decreased level of plasma estradiol to progesterone ratio. In
contrast to premenopausal women, an estrogenic effect was observed in post-
menopausal women fed flaxseed as indicated by an increase in their vaginal cell
maturation index and reduction in hot flushes. 50 .51
To determine whether any hormonal effects of flaxseed are due to its
lignans, rats were fed for four weeks either the basal diet (control) or the basal
diet supplemented with 2.5%, 5%, or 10% flaxseed or the SDG equivalent to that
consumed in flaxseed (i.e. 0.75, 1.5 or 3.0mg/day). This resulted in a dose- related
cessation or lengthening of estrous cycles in both the flaxseed- and SDG-fed rats,
similar to the effect in rats given the anti estrogen tamoxifen (1 mg/kg body
weight).52 Estrus cycling changes also were observed in the offspring of rats fed
flaxseed or SDG during pregnancy and lactation. 53
Overall, the above in vivo data suggest that flaxseed has some hormonal
effects which can partly be attributed to its SDG. It has an anti estrogenic effect
when blood estrogen levels are high (as in premenopausal women and in animal
studies), and an estrogenic effect when the blood estrogen levels are low (as in
the case of postmenopausal women). Although these observations appear to agree
with the in vitro studies, the lignan concentrations found to be cancer protective
in vitro 35 - 47 are often much higher (1-100 umollL) than the blood lignan levels
«1 umol/L) reported for vegetarians or human subjects fed flaxseed.12.54-5R
Perhaps multiple mechanisms (i.e. both hormone- and non-hormone-related) take
effect at the same time and result in cancer protection even at low blood lignan
60 L. U. THOMPSON
concentrations. It is also likely that phytochemicals, including the lignans, are

effective even at low blood levels because they act synergistically or additively.
For example, a mixture of 0.5 umol/L each of seven phytoestrogens, including
EL, has been shown to inhibit the 5 alpha reductase and 17 beta hydroxysteroid
dehydrogenase at the same level as 100umollL of EL.44
APPLICATION IN CANCER PREVENTION

AND TREATMENT
Lignan bioavailability studies are important as they will help in relating in

vitro and in vivo data, in understanding the mechanisms, in determining the organs
most affected by lignans, and in estimating the optimum level and frequency of
lignan intake for safety and anticancer effects. Little work has yet been done in
this area. However, it has been shown that mammalian lignan production, based
on urinary lignan excretion, is linear at up to 5% flaxseed level of intake or up
to 4.4 umol SDG/day intake in rats,25 and at up to 25 g/day flaxseed intake in
humans. 54 From rat studies with radiolabelled SDG, some lignans still remain in
the gastrointestinal tract 48 hours after intake,58 perhaps due to their slow rate of
metabolism and enterohepatic circulation. 54.58 The tissues with the highest lignans
are those involved in SDG metabolism (i.e. intestinal, hepatic, and renal) and
certain estrogen sensitive tissues (i.e. uterus).58 Interestingly, the mammary gland
had only low levels of the SDG metabolites despite the fact that cancer protec-
tive effects have been observed in this organ, suggesting that the effect of lignan
is not entirely due to direct binding with the organ. Daily versus occasional intake
of lignan rich foods such as flaxseed can at least double as well as stabilize the
plasma lignan level. 54 Because of slow lignan metabolism, a once a day vs. several
times a day intake of the lignan-rich food is sufficient to maintain high blood
lignan levels. The SDG appears to be stable during processing since the intake of
raw flaxseed or flaxseed in a muffin or bread formulation resulted in similar
urinary lignan excretion in human subjects. 54
Recognizing that flaxseed and its lignans may have potential health bene-
fits including anticancer effects, food companies are now marketing flaxseed con-
taining products. Whole or milled flaxseed has been incorporated into baked
products such as breads, muffins, bagels, rolls, cookies, or breakfast cereals. In
a quick survey of Toronto supermarkets, we found at least 12 different brands of
breads and two breakfast cereals containing flaxseed. 20 Flaxseed containing
muffins and bagels are also sold in bakery shops and restaurants. These products
were not available about a decade ago.
The mammalian Iignans produced from commercial and homemade prod-
ucts have a significant dose dependent relationship with the amount of flaxseed
added. 20 If the base grains and flaxseed source are the same, the amount of
flaxseed is a good predictor of mammalian lignans produced from these foods.
Considering the mean lignan value for all commercial breads (12.4 umolll 00 g),
two 30-g slices of flaxseed-containing bread will provide about 7.4 umol of
lignans. 2o This amount is higher than the estimated lignan intake of nonvegetari-
ans (4.6umol/day) with a plant food intake on the upper end of current diet rec-
ommendations (i.e. 8 servings of whole grains and 8 servings of fruit and/or
vegetables) and is the same as the lignan intake of vegetarians (7.4umol/day) who
may have a higher intake of plant foods (i.e. 9 servings of whole grains, 10 serv-
ings of fruits and/or vegetables, 2 servings of legumes, and one serving of
oilseeds). Thus, an individual may be able to double or triple their intake of
tignans with the addition of some flaxseed containing breads in their diet.
CONCLUSION
Epidemiological studies suggest that lignans may have some cancer pro-
tective effects. Animal studies have shown that flaxseed, the richest source of
mammalian lignan precursor, can (a) reduce the mammary tumor incidence,
number, and size depending on the time of exposure, (b) reduce the development
and growth of colon tumors, and (c) reduce the tumor size and incidence oflung
metastasis of melanoma cells. The SDG, ED, and EL produced effects similar to
flaxseed, confirming that mammalian lignans derived from SDG have anticancer
effects. In vitro studies and limited in vivo studies suggest that both hormone
related (i.e. weak estrogenic/antiestrogenic) and non-hormone related (i.e. antiox-
idant, anti angiogenic ) mechanisms may be responsible for this effect. Although
bioavailability studies suggest some similarities in the metabolism of lignans in
vitro, and in rats and humans, the anticancer effects of lignans have yet to be
demonstrated in prospective studies in humans. If confirmed to be cancer pro-
tective without adverse effects, increased tignan consumption not only from
whole grains, legumes, fruits, and vegetables but also from a small amount of
flaxseed or flaxseed-supplemented products may be recommended.
ACKNOWLEDGMENT
The author thanks her graduate students, postdoctoral fellows, technicians

and collaborators for their research assistance and the Natural Sciences and Engi-
neering Research Council, Flax Council, Health Canada and Cancer Research
Society for financial support of her research program on lignans.
REFERENCES
1. WORLD CANCER RESEARCH FUND, AMER1CAN INSTlTUTE FOR CANCER

RESEARCH, 1997. Food, Nutrition, and the Prevention of Cancer: A Global Perspective.
AICR, Washington, D.C., 670 p.
62 L. U. THOMPSON
2. SETCHELL, K.D.R. 1995. Discovery and potential clinical importance of mammalian

lignans. pp. 82-98 in Flaxseed in Human Nutrition (S.c. Cunnane, L.u. Thompson, eds.),
AOCS, Champaign, IL.
3. AXELSON, M., SJOVALL, J., GUSTAFSSON, B.E., SETCHELL, K.D. 1982. Origin of
lignans in mammals and identification of a precursor from plants. Nature 298:
659-660.
4. BORRIELLO, S.P', SETCHELL, K.D., AXELSON, M., LAWSON, A.M. 1985.
Production and metabolism of lignans by the human faecal flora. 1. Appl. Bacteriol. 58:
37-43.
5. AXELSON, M., SETCHELL, K.D. 1980. Conjugation of lignans in human urine. FEBS
Lett. 122: 49-53.
6. SETCHELL, K.D.R., ADLERCREUTZ, H. 1988. Mammalian lignans and phytoestro-
gens: Recent studies on their formation, metabolism and biological role in health and
disease. pp. 315-345 in Role of the Gut Flora, Toxicity and Cancer (1.R. Rowland, eds.),
Academic Press, London.
7. ADLERCREUTZ, H., FOTSIS, I., BANNWART, c., WAHALA, K., MAKELA, I.,
BRUNOW, G., HASE, I. 1986. Determination of urinary lignans and phytoestrogen
metabolites, potential antiestrogens and anticarcinogens, in urine of women on yarious
habitual diets. J. Steroid Biochem. 25: 791-797.
8. ADLERCREUTZ, H., FOTSIS, I., HEIKKINEN, R., DWYER, J.T., WOODS, M.,
GOLDIN, B.R., GORBACH, S.L. 1982. Excretion of the lignans enterolactone and
enterodiol and of equol in omnivorous and vegetarian postmenopausal women and in
women with breast cancer. Lancet 2: 1295-1299.
9. ADLERCREUTZ, H., HONJO, H., HIGASHI, A., FOTSIS, I., HAM ALAIN EN, E.,
HASEGAWA, I., OKADA, H. 1991. Urinary excretion oflignans and isoflavonoid phy-
toestrogens in Japanese men and women consuming a traditional Japanese diet. Am. J.
Clin. Nutr. 54: 1093-1100.
10. ADLERCREUTZ, H.T, GOLDIN, B.R., GORBACH, S.L., HOCKERSTEDT, K.A.V,
WATANABE, S., HAMALAINEN, E.K., MARKKANEN, M.H., MAKELA, TH.,
WAHALA, K.T, HASE, TA., FOTSIS, I. 1995. Soybean phytoestrogen intake and
cancer risk. J. Nutr. 125: 757S-770S.
II. INGRAM, D., SANDERS, K., KOLYBABA, M., LOPEZ, D. 1997. Case-control study
of phyto-oestrogens and breast cancer. Lancet 350: 990--994.
12. MORTON, M.S., CHAN, P'S.E, CHENG, c., BLACKLOCK, N., MATOS-FERREIRA,
A., ABRANCHES-MONTEIRO, L., CORREIA, R., LLOYD, S., GRIFFITHS, K. 1997.
Lignans and isoflavonoids in plasma and prostatic fluid in men: samples from Portugal,
Hong Kong, and the United Kingdom. Prostate 32: 122-128.
13. THOMPSON, L.u., ROBB, P., SERRAINO, M., CHEUNG, E 1991. Mammalian lignan
production from various foods. Nutr. Cancer 16: 43-52.
14. MAZUR, W, FOTSIS, I., WAHALA, K., OJALA, S., SALAKKA, A., ADLERCREUTZ,
H. 1996. Isotope dilution gas chromatographic-mass spectrometric method for the deter-
mination ofisoflavonoids, coumestrol, and lignans in food samples. Anal. Biochem. 233:
169-180.
15. MAZUR, WM., DUKE, J.A., WAHALA, K., RASKU, S., ADLERCREUTZ, H. 1998.
Isoflavonoids and lignans in legumes: Nutritional and health aspects in humans. J. Nutr.
Biochem. 9: 193-200.
16. THOMPSON, L.u., RICKARD, S.E., CHEUNG, E, KENASCHUK, E.O.,
OBERMEYER, WR. 1997. Variability in the anticancer lignan levels in flaxseed. Nulr.
Cancer 27: 26-30.
17. SHULTZ, TO., BONORDEN, W.R., SEAMAN, W.R. 1991. Effect of short-term flaxseed
consumption on lignan and sex hormone metabolism. N utr. Res. II: 1089-1100.
18. LAMPE, 1.W, MARTINI, M.C., KURZER, M.S., ADLERCREUTZ, H., SLAVIN, 1.1..
1994. Urinary lignan and isoflavonoid excretion in premenopausal women consuming
flaxseed powder. Am. 1. Clin. Nutr. 60: 122~128.
19. CUNNANE, S.c., HAMADEH, MJ., LIEDE, A.C., THOMPSON, loU, WOLEVER,
T.M., JENKINS, DJ. 1995. Nutritional attributes of traditional flaxseed in healthy young
adults. Am. 1. Clin. Nutr. 61: 62--68.
20. NESBITT, P.D., THOMPSON, loU 1997. Lignans in homemade and commercial
products containing flaxseed. Nutr. Cancer 29: 222~227.
21. SERRAINO, M., THOMPSON, loU 1991. The effect of flaxseed supplementation on
early risk markers for mammary carcinogenesis. Cancer Lett. 60: 135~142.
22. RUSSO, 1., RUSSO, I.H. 1978. DNA labelling index and structure of the rat mammary
gland as determinants of its susceptibility to carcinogenesis. 1. Natl. Cancer Inst. 61:
1451~1459.
23. SHARKEY, M., BRUCE, R. 1986. Quantitation of nuclear aberrations as a screen for
agents damaging to mammary epithelium. Carcinogenesis 7: 1991 ~ 1995.
24. SERRAINO, M., THOMPSON, loU 1992. The effect of flaxseed supplementation on the
initiation and promotional stages of mammary tumorigenesis. Nutr. Cancer 17: 153~ 159.
25. RICKARD, S.E., ORCHESON, 1..1., SEIDL, M.M., LUYENGI, 1.., FONG, H.H.,
THOMPSON, loU 1996. Dose-dependent production of mammalian lignans in rats and
in vitro from the purified precursor secoisolariciresinol diglycoside in flaxseed. 1. Nutr.
126: 2012~2019.
26. THOMPSON, loU, SEIDL, M., RICKARD, S., ORCHESON, 1.., FONG, H. 1996.
Antitumorigenic effect of a mammalian lignan precursor from flaxseed. Nutr. Cancer 26:
159~165.
27. THOMPSON, loU, RICKARD, S.E., ORCHESON, 1..1., SEIDL, M.M. 1996. Flaxseed
and its lignan and oil components reduce mammary tumor growth at a late stage of car-
cinogenesis. Carcinogenesis 17: 1373~1376.
28. SERRAINO, M., THOMPSON, loU 1992. Flaxseed supplementation and early markers
of colon carcinogenesis. Cancer Lett. 63: 159~165.
29. BIRD, R.P. 1995. Role of aberrant crypt foci in understanding the pathogenesis of colon
cancer. Cancer Lett. 93: 55-71.
30. JENAB, M., THOMPSON, L.u. 1996. The influence of flaxseed and lignans on colon
carcinogenesis and beta-glucuronidase activity. Carcinogenesis 17: 1343~ 1348.
31. SUNG, M.-K., LAUTENS, M., THOMPSON, L.U 1998. Mammalian lignans inhibit the
growth of estrogen-independent human colon tumor cells. Anticancer Res. 18:
1405-1408.
32. YAN, 1.., YEE, 1.A., LI, D., MCGUIRE, M.H., THOMPSON, L.U 1998. Dietary flaxseed
supplementation and experimental metastasis of melanoma cells in mice. Cancer Lett.
124: 181~186.
33. LANDSTROM, M., ZHANG, 1.X., HALLMANS, G., AMAN, P., BERGH, A.,
DAMBER, 1.E., MAZUR, W, WAHALA, K., ADLERCREUTZ, H. 1998. Inhibitory
effects of soy and rye diets on the development of Dunning R3327 prostate adenocarci-
noma in rats. Prostate 36: 15 I ~ 161.
34. ADLERCREUTZ, H., MAZUR, W 1997. Phytoestrogens and western diseases. Ann.
Med. 29: 95-120.
35. MOUSAVI, Y, ADLERCREUTZ, H. 1992. Enterolactone and estradiol inhibit each
other's proliferative effect on MCF-7 breast cancer cells in culture. 1. Steroid Biochem.
Mol. BioI. 41: 615~619.
36. WANG, c., KURZER, M.S. 1997. Phytoestrogen concentration determines effects on
DNA synthesis in human breast cancer cells. Nutr. Cancer 28: 236-247.
37. WELSHONS, WV, MURPHY, C.S., KOCH, R., CALAF, G., JORDAN, Vc. 1987.
64 L. U. THOMPSON
Stimulation of breast cancer cells in vitro by the environmental estrogen enterolactone

and the phytoestrogen equol. Breast Cancer Res. Treat. 10: 169-175.
38. HIRANO, 1., FUKUOKA, K., OKA, K., NAITO, 1., HOSAKA, K., MITSUHASHI, H.,
MATSUMOTO, Y 1990. Antiproliferative activity of mammalian lignan derivatives
against the human breast carcinoma cell line, ZR-75-1. Cancer Invest. 8: 595-602.
39. SATHYAMOORTHY, N., WANG, T.T.Y, PHANG, TM. 1994. Stimulation of pS2
expression by diet-derived compounds. Cancer Res. 54: 957-961.
40. GARREAU, B., VALLETTE, G., ADLERCREUTZ, H., WAHALA, K., MAKELA, T.,
BENASSAYAG, C., NUNEZ, E.A. 1991. Phytoestrogens: new ligands for rat and human
alpha-fetoprotein [published erratum appears in Biochim Biophys Acta 1991 Dec
3;1133(1): 113]. Biochim. Biophys. Acta. 1094: 339-345.
41. ADLERCREUTZ, H., MOUSAVI, Y, CLARK, 1., HOCKERSTEDT, K.,
HAM ALAIN EN, E., WAHALA, K., MAKELA, 1., HASE, 1. 1992. Dietary phytoestro-
gens and cancer: In vitro and in vivo studies. J. Steroid Biochem. Mol. BioI. 41: 331-337.
42. ADLERCREUTZ, H., BANNWART, C, WAHALA, K., MAKELA, 1., BRUNOW, G.,
HASE, 1., AROSEMENA, P.J., KELLIS, J.T., Jr., VICKERY, L.E. 1993. Inhibition of
human aromatase by mammalian lignans and isoflavonoid phytoestrogens. 1. Steroid
Biochem. Mol. BioI. 44: 147-153.
43. WANG, C, MAKELA, 1., HASE, 1., ADLERCREUTZ, H., KURZER, M.S. 1994.
Lignans and flavonoids inhibit aromatase enzyme in human preadipocytes. 1. Steroid
Biochem. Mol. BioI. 50: 205-212.
44. EVANS, B.A.J., GRIFFITHS, K., MORTON, M.S. 1995. Inhibition of 5a-reductase in
genital skin fibroblasts and prostate tissue by dietary lignans and isoflavonoids. 1.
Endocrinol. 147: 295-302.
45. MARTIN, M.E., HAOURIGUI, M., PELISSERO, C., BENASSAYAG, C., NUNEZ, E.A.
1996. Interactions between phytoestrogens and human sex steroid binding protein. Life
Sci. 58: 429-436.
46. FOTSIS, 1., PEPPER, M., ADLERCREUTZ, H., FLEISCHMANN, G., HASE, 1.,
MONTESANO, R., SCHWEIGERER, L. 1993. Genistein, a dietary-derived inhibitor of
in vitro angiogenesis. Proc. Natl. Acad. Sci. U.S.A. 90: 2690-2694.
47. PRASAD, K. 1997. Hydroxyl radical-scavenging property ofsecoisolariciresinol diglu-
cos ide (SDG) isolated from flax-seed. Mol. Cell. Biochem. 168: 117-123.
48. AMAROWICZ, R., WANASUNDARA, W.M.V.N., WANASUNDARA, P.K.J.P.D.,
SHAHIDI, F. 1993. Antioxidant of the ethanolic extract of flax seed in ~
carotene-linoleate model system. J Food Lipids I: 111-117.
49. PHIPPS, W.R., MARTINI, M.C., LAMPE, 1.w., SLAVIN, J.L., KURZER, M.S. 1993.
Effect of flax seed ingestion on the menstrual cycle. 1. Clin. Endocrinol. Metab. 77:
1215-1219.
50. WILCOX, G., WAHLQVIST, M.L., BURGER, H.G., MEDLEY, G. 1990. Oestrogenic
effects of plan foods in postmenopausal women. Sr. Med. J. 301: 905-906.
51. MURKIES, A.L., WILCOX, G., DAVIS, S.R. 1998. Clinical review 92: Phytoestrogens.
1. Clin. Endocrinol. Metab. 83: 297-303.
52. ORCHESON, L.1., RICKARD, S.E., SEIDL, M.M., THOMPSON, L.v. 1998. Flaxseed
and its mammalian lignan precursors cause a lengthening or cessation of estrous cycling
in rats. Cancer Lett. 125: 69-76.
53. TOU, J., CHEN, 1., THOMPSON, L. 1998. Flaxseed and its lignan precursor, secoiso-
lariciresinol diglycoside, affect pregnancy outcome and reproductive development in rats.
1. Nutr. 128: 1861-1868
54. NESBITT, P.D., LAM, Y, THOMPSON, L.v. 1999. Human metabolism of mammalian
lignan precursors in raw and processed flaxseed. Am. 1. Clin. Nutr. (in press).
55. ADLERCREUTZ, H., FOTSIS, 1., LAMPE, 1., WAHALA, K., MAKELA, 1., BRUNOW,
G., HASE, T 1993. Quantitative detennination of Iignans and isoflavonoids in plasma of

omnivorous and vegetarian women by isotope dilution gas chromatography-mass spec-
trometry. Scand. 1. Clin. Lab. Invest. Supp!. 215: 5-18.
56. ADLERCREUTZ, H., FOTSIS, T, WATANABE, S., LAMPE, J., WAH ALA, K.,
MAKELA, T, HASE, T 1994. Determination of lignans and isoflavonoids in plasma by
isotope dilution gas chromatography-mass spectrometry. Cancer Detect. Prevo 18:
259-271.
57. MORTON, M.S., WILCOX, G., WAHLQVIST, M.L., GRIFFITHS, K. 1994. Determi-
nation of Iignans and isoflavonoids in human female plasma following dietary supple-
mentation. 1. Endocrino!. 142: 251-259.
58. RICKARD, S.E., THOMPSON, L.v. 1998. Chronic exposure to secoisolariciresinol
diglycoside alters lignan disposition in rats. 1. Nutr. 128: 615-623.
Chapter Four
TOWARD ENGINEERING THE METABOLIC

PATHWAYS OF CANCER-PREVENTING
LIGNANS IN CEREAL GRAINS AND
OTHER CROPS
Michael A. Costa, Zhi-Qiang Xia, Laurence B. Davin, and

Norman G. Lewis
Institute of Biological Chemistry

Washington State University
Pullman, Washington 99164-6340
Introduction ................................................... 67
Dietary Lignans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 68
Cancer-Preventing Lignans in Plants ............................... 70
Lignan Precursors and Biosynthetic Pathways. . . . . . . . . . . . . . . . . . . . . . .. 76
Laccases .................................................... 76
Dirigent Proteins ....................... '. . . . . . . . . . . . . . . . . . . . .. 76
PinoresinollLariciresinol Reductases . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 80
Secoisolariciresinol Dehydrogenase . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 81
Secoisolariciresinol Glucosyltransferase .......................... 82
Enhanced Levels of Cancer-Preventing Lignans in Plants: Towards
Enginnering Their Metabolic Pathways. . . . . . . . . . . . . . . . . . . . .. 82
INTRODUCTION
Lignans, ubiquitous constituents of vascular plants, have a number of prop-

erties that are of use to humans: some can protect against the onset of various
cancers,1 whereas others have antimitotic, antiviral, antibacterial, and antifungal
Phytochemicals in Human Health Protection. Nutrition, and Plant Detense, edited by Romeo.
Kluwer Academic / Plenum Publishers, New York. 1999.
67
68 M. A. COSTA et al.
properties. 2 Certain lignans can also function, for example, as antioxidants,

platelet activating factor receptor antagonists, and anti-tubercular agents, and
others are disinfectants, moth repellants, and insecticides. 3- 5 Because of their
important applications from a health and economic perspective, intensified
efforts are being expended to both understand and manipulate their biosynthetic
pathways. Accordingly, this chapter highlights progress being made towards
deciphering the 8-8' linked lignan metabolic pathway and its utility to the
bioengineering of human foodstuffs.
The lignans are found in essentially all terrestrial land plants,6 although
their amounts can vary markedly with both tissue type and species. They are pre-
dominantly dirneric phenylpropanoids, with the most commonly reported
interunit linkage being the 8-8', as illustrated by pinoresinol (1), secoisolari-
ciresinol (2), and matairesinol (3) (Fig. I); this particular contribution focuses on
the importance of such lignans in helping prevent the onset of certain
malignancies.
DIETARY LIGNANS
It is sometimes assumed that Western populations conscientiously consume

nutritionally well balanced diets that are conducive to the maintenance of good
health. Many aspects, however, are overlooked, neglected, or are unknown as
regards the underlying causes for certain populations being healthier, and how we
can alter our own diets in order to attain improved health benefits. For example,
although the well-defined Western diet may be balanced in terms of daily nutri-
ent intake, the importance of consuming a diet containing increased levels of sub-
stances having roles in disease prevention is often neglected. Moreover, a truly
balanced dietary intake might also be expected to have positive economical
ramifications with respect to the reduction of health care and medical costs.
Recent information obtained from various studies has added to our knowl-
edge as to which types of compounds provide seemingly beneficial health effects;
this resulted primarily from studies of populations consuming higher levels of
plant substances compared to those on omnivorous diets, and provided needed
insight as to why certain dietary plant components are salutary to our health.
For example, the typical Western diet, which consists of a low level of
dietary fiber and a high content of animal fat/protein, adversely contributes to
the high incidence of breast, prostate, and colon cancers, and other disorders
such as coronary heart disease. On the other hand, a diet rich in grain products
results in a decreased risk of these hormone-dependent cancers, 7 even though
the mechanisms by which diet influences hormone concentrations are not
clearly understood. It has also been observed that breast cancer patients have a
lower fiber intake and lower urinary excretion of the "mammalian" lignans,
enterodiol (4) and enterolactone (5), obtained via metabolism in the gastroin-
(X0H HOD (')
»
z
1 0M, (')
g"'>. M,O I 8'
m
8 ~
"'' ' 8 '"0
:;>;:I
m
,O"v <:
HO .# ° ~ OH ~
-l
J?' OMe OMe Z
(+)-Pinoresinol la (-)-Pinoresinol Ib (;)
r
MeO_ ~ /'-. 8 /'-. ~ .OMe

OH - /'.-
Meo:crM [J)
HO X ' ( X o M e I I0
.' ~
~
HO ~ Z
HO' r 8''''-,./ 0 H g' OH HO .# 8 IT 8""'" """ 'OH (')
m
:;>;:I
m
»
r
OMe
(;)
OH OH OH OH S;
(-)-SecoisolariciresinoI2a (+ )-Seeoisolariciresinol 2b (-)-MatairesinoI3a (+)-MatairesinoI3b Z
[J)
»
z
HO MeO o
OH OGle
HO~8'0 ~
OH 1.# 8 OGle ::t:
HO m
:;>;:I
o (')
P"I e5'"0
~ OH [J)
OH
Enterodiol 4 Enterolaetone 5 Secoisolariciresinol diglucoside 6
0-
'-0
Figure 1. Various 8-8' linked lignans and the "mammalian" lignans, enterodiol 4 and enterolactone 5.
Table 1. Lignan excretion by women consuming

vegetarian vs. non-vegetarian diets 8
Dietary group Average total urinary lignan (I1molJday)
Premenopausal
Vegetarian 4.16
Omnivore 2.89
Breast cancer 2.34
Postmenopausal
Vegetarian 8.09
Omnivore 1.99
Breast cancer 2.08
testinal tract of the plant lignans, secoisolariciresinol (2) and matairesinol (3). In
contrast, those population groups that have a positive correlation between
high grain intake and urinary enterolactone (5) excretion appear to be more pro-
tected, e.g. young vegetarians. In this respect, Table 18 shows various dietary
groups and the presumed health risks associated with low consumption of daily
levels of grain, as revealed by differences in enterolactone excretion. Note also
that a high lignan excretion has been observed in subjects with a low risk of colon
cancer.9
Such differences initially were thought to be due to either the amount or
type of dietary fiber ingested, but comparisons of populations consuming Western
vs. vegetarian diets established similar levels of total dietary fiber intake. Other
studies showed that increased fiber levels in the diet did not have distinct health
benefits,1O suggesting that additional factors, associated with plant food items in
the diet, had a beneficial effect. 11.12
Eventually, two classes of plant-derived phenolics, the "mammalian"
lignans and the isoflavonoids, were attributed roles as cancer-preventive sub-
stances. Both classes of these metabolites are diphenolic compounds having
certain structural similarities to the natural and synthetic estrogens and antie-
strogens,13 and the term "phytoestrogen" was subsequently coined. In addition to
cancer prevention, dietary phytoestrogens also are believed to play a role in
limiting the effects of menopause, osteoporosis, and heart disease. Indeed, the
importance of lignans in medicine, human health, and nutrition has since become
well recognized. lo
CANCER-PREVENTING LIGNANS IN PLANTS
The role of the "mammalian" lignans, enterodiol (4) and enterolactone (5),
in helping prevent carcinogenesis has been summarized in detail in numerous
contributions,14.15 and the reader is referred to Chapter 3 by Thompson in this
CANCER-PREVENTING LIGNANS IN CEREAL GRAINS AND OTHER CROPS 71
volume. They are formed from secoisolariciresinol (2) and matairesinol (3),
through metabolism by facultative gastrointestinal microflora,16 and differ from
plant lignans in that they have phenolic hydroxyl groups only in the meta posi-
tion of the aromatic rings (Fig. 2). Possible pathways for the conversion of 2 and
3 into 4 and 5, respectively, have been described elsewhere; 17 mammalian lignans
were first detected in the urine of female rats and humans. 18,19
The presence of mammalian lignans in humans can influence sex hormone
metabolism and various biological activities, including intracellular enzyme
levels, protein biosynthesis, growth factor action, and malignant cell proliferation
and angiogenesis. These properties have been interpreted as providing the basis
MeO MeO
HO HO
OH OH
Secoisolariciresinol 2 Matairesinol 3
MeO MeO
OH
"'",,/OH
OMe
HO HO
OH
OH
._---------------..,.
OH
Enterodiol 4 Enterolactone 5
Figure 2. Formation of cnterodiol 4 and enterolactone 5 by action of gastrointestinal flora on

sccoisolariciresinol 2 and matairesinol 3. respectively.
for cancer growth inhibition and prevention of cancer initiation. 14 Proposed mech-
anisms on how mammalian lignans (so-called phytoestrogens) may exert their
effects include that of decreasing free estrogen concentrations resulting from
stimulation of sex hormone binding globulin synthesis in the liver. \3 However,
mammalian lignans are also effective in preventing the initiation and growth of
animal tumors,15 and it has been shown that synthetic enterolactone (5) is cyto-
toxic to human lymphoid cells. IS Although the precise physiological and phar-
macological effects of the dietary lignans are not yet established, a positive
correlation has been identified between diet and reduction in onset of "Western
diseases", including breast, prostate, and colon cancers.
The dietary lignans, secoisolariciresinol, matairesinol, and secoisolari-
ciresinol diglucoside (SOG, 6) are present in grains, various seeds, berries,
fruits, and vegetables (Tables 2-5) as shown. The main known source is
flaxseed, which contains SOG in amounts 75 to 800 times higher than in other
foods (Table 5).15,10,2\ Additionally, legumes such as peanuts, alfalfa sprouts,
kudzu leaf, and soybeans, also contain secoisolariciresinol (see Table 3).22
Table 2. Lignans obtained from various plant parts with

dietary and/or medicinal value
Plant
Plant species organ Lignan Proposed health benefit Ref.
Legumes (e.g .• Seed Secoisolariciresinol 2 Cancer chemoprevention 22
soybean, Matairesinol 3
Glycine max)
Tea (Camellia Leaf Secoisolariciresinol 2 Antioxidant 28
sinensis) Matairesinol 3 Cancer chemoprevention
Heart disease prevention
Flax (Lin urn Seed Secoisolariciresinol Decreases cholesterol levels 15

usitatissimum) diglucoside 6 Decreases rate of onset of
Secoisolariciresinol 2 lupus (in mice)
Cancer chemoprevention
Sesame Seed Sesamin 7 Reduced cholesterol levels 26

(Sesamum Sesamolinol 8 Antioxidant
indicum) Prevent heart disease
Mayapple Root Podophyllotoxin 9 Cancer treatment 6

(Podophyllum rhizomes (etoposide 1O,
sp.) teniposide ll)
Schizandra Fruit Gomisin A 12 Prevents hepatic disease 30--32

chinensis
Creosote bush Leaf Nordihydroguaiaretic Antioxidant 33
(Larrea acid 13
tridendata)
Table 3. Lignan contents reported present in various legumes 22

Plants Secoisolariciresinol 2* Matairesinol 3*
Glycine max (soybean) 13-273 O-trace
Phaseolus vulgaris (kidney bean) 56-153 O-trace
Phaseolus lunatus (lima bean) 158-184 trace
Apios americana (groundnut) 20.8-58.1 2.2-4.6
Cajanus cajan (pigeon pea) 18.7-50.3 o
Cicer arietinum (garbanzo bean) 6.7-8.1 o
Viciafaba (broad bean) 26-31.8 131-trace
Vigna mungo (black gram) 45.7-240 70.8-262
Arachis hypogaea (peanut) 333 trace
Puer aria lobata (kudzu leaf) 476 trace
Sophora japonica (Japanese pagoda tree) 1,590 3.8
*~gl\ 00 g dry weight.
Table 4. Enterodiol 4 and enterolactone 5 production

from various foods 23
Food Enterodiol 4* Entero1actone 5*
Flaxseed meal 68,204 9,841
Wheat bran 327 296
Oats 99 279
Brown rice 143 190
Lentil 1,092 864
Strawberry 376 400
Pear 410 670
Apple 6 223
Mckuba seaweed 1,083 184
*~gl\ 00 g dry weight.
Table 5. Lignan content of various foods 2l

Food Secoisolariciresinol2* Matairesinol 3*
Flaxseed 369,900 1,087
Sunflower seed 610 o
Wheat (whole grain) 32.9 2.6
Barley (whole grain) 58.0 o
Ryc (whole grain) 47.1 65
Pumpkin seeds 21,370 o
Carrots 192 2.9
Broccoli 414 23.1
Cranberry 1510 o
*~gl1 00 g dry weight.
Other data have also been presented ls .23 indicating the in vitro formation of
"mammalian" lignans following digestion of human dietary food items (Table 4).
Indeed, the comparison of foods in Table 4 serves as a possible indicator of the
large differences in amounts of "mammalian" lignans produced from different
foods in the diet, and presumably reflects the levels of lignans present in the plant
materials consumed. Of special interest is the high level produced when flaxseed
is used as substrate. 23 On the other hand, the values present in mekuba seaweed
are unusual, given that the phenylpropanoid pathway is presumed to be absent (to
any great extent) in algae. 24 Whether the mammalian lignans are exclusively due
to secoisolariciresinol and matairesinol has also been questionable, given that
Adlercreutz l4 noted that, in rats fed rye (Secale cereale) bran, more mammalian
lignans were excreted than the lignans deemed present. Another important aspect
is the effect that different levels and types of fiber in the diet may have on lignan
metabolism, since this may alter the populations of bacterial species present in
the intestine. Indeed, Reddy et at. 25 summarized the effects of diverse dietary
habits on activities and levels of the gut microflora, revealing differences in con-
centrations of anaerobes and aerobes and in associated metabolism of fecal bile
acids.
In addition to 2 and 3, other lignans have important pharmacological
roles (Table 2, Fig. 3).2 These include pinoresinol (1) from Forsythia in term edia ,
a blood platelet activating factor, and sesamin (7) and sesamolinol (8), antioxi-
dants from sesame (Sesamum indicum) seeds, whose intake reduces the choles-
terol levels in humans. 2 Indeed, the mechanism by which sesamin exerts
its antioxidant effects has been studied using rats fed experimental diets: admin-
istration of sesamin resulted in an increase in tocopherol levels/ 6 where the
role of sesamin on tocopherol levels appears to involve competition with y-
tocopherol oxidation. Akimoto et al.27 also discuss its multiple biological
functions, which include interference with linoleic acid metabolism, hypocholes-
terolemic action, enhancement of hepatic detoxication of chemicals and
alcohol, protective effect on hypertension, cardiovascular hypertrophy, and chem-
ically induced mammary tumors. Moreover, sesamin is a hydroxyl radical
scavenger and a potent inhibitor of NADPH-dependent microsomal lipid
peroxidation.
Interestingly, a high tea (both black and green) intake is also negatively
correlated with the occurrence of coronary heart disease. Mazur et al. 28 have
reported relatively high levels of secoisolariciresinol and matairesinol in tea as
being (partly) responsible for its antimutagenic, anticarcinogenic, and antioxida-
tive properties. Other lignans of biological importance include podophyllotoxin
(9) from the mayapple (Podophyllum peltatum) rhizomes which, as its semi-
synthetic etoposide (10) and teniposide (11) derivatives, is among only a few
plant compounds effcctive for cancer treatment. 6•29 Another is gomisin A (12),
isolated from Schizandra chinensis fruits, which has the ability to protect the
(+)-Sesamin 7 (+)-SesamolinoI8
OH
t~.8,0
O~
~ 0
MoOVOM, OMe
(~)-Podophyllotoxin 9 Teniposide II
HO
HO
MeO OH
(+)-Gomisin A 12 Nordihydroguaiaretic acid 13
Figure 3. Selected lignans having dietary and pharmacological significance.
liver from a variety of liver-damaging agents, e.g. the hepatotoxic compounds,

galactosamine and CCI 4 . 3O- 32 Finally, nordihydroguaiaretic acid (13), from the
creosote bush (Larrea tridendata) has powerful antioxidant properties/ 3 and
a derivative is also effective against the human immunodeficiency virus
(HIV).34.35
LIGNAN PRECURSORS AND BIOSYNTHETIC PATHWAYS
Lignans are found in vascular plants and accumulate in various tissues

including roots, stems, leaves, flowers, fruits, and seeds. 6 They typically occur as
optically pure dimeric substances, formed by the coupling of two monomeric
C 6 C3 moieties derived from the phenylpropanoid pathway.2
The biochemical pathway to the lignans of interest inthis chapter was first
delineated in Forsythia intermedia (Fig. 4).2.17,36-46 Entry in the pathway involves
the first known example of regio- and stereoselective intermolecular phenoxy
radical coupling in plants, producing (+)-pinoresinol (Ia) through the coupling
of two E-coniferyl alcohol (14) molecules. 38 Further conversion into
(-)-secoisolariciresinol (2a) and (-)-matairesinol (3a) occurs via sequential
reductive and dehydrogenative steps, respectively.36.39.45.4 6
Laccases
While plants are able to control or confer specificity to phenolic coupling

reactions in a precise, predictable way, all known one-electron oxidants (such as
02/laccase; H202/peroxidase) in vitro catalyze phenol coupling reactions that are
non-specific,z4,38 If more than one coupling site is present, however, such as is the
case with coniferyl alcohol, a range of racemic compounds are produced via non-
regio and non-stereospecific coupling.
In this regard, we isolated a highly glycosylated laccase from F intermedia
stems of M, - 100 kDa. When coniferyl alcohol is incubated in its presence,
it generates free radicals which undergo non-specific coupling reactions to give
predominantly (±)-pinoresinols (Ia/I b), (±)-dehydrodiconiferyl alcohols
(I5a/I5b), and (±)-erythrolthreo guaiacylglycerol 8-0~' coniferyl alcohol
ethers (16a/I6b), where neither regiochemistry nor stereochemistry is controlled
(Fig. 5A).38 Since then, the cDNA clone for this laccase-like protein was obtained
from a F intermedia stem cDNA library by using a polymerase chain reaction
guided strategy, and the 2-kilobase pair sequence was found to have the expected
four copper-binding regions, as well as potential glycosylation sites. 24
Dirigent Proteins
Determination of how vascular plants actually specify the outcome of phe-

nolic coupling reactions in vitro has been pioneered by this research group.38 In
this regard, we have isolated a protein which, when incubated with E-coniferyl
alcohol (14) and a one electron oxidant in vitro, is able to engender only forma-
tion of (+)-pinoresinol (la) (Fig. 5B), rather than the racemic products (±) la/lb,
15a/15b, and 16a/16b. This protein has a M, - 78kDa as determined by gel per-
meation chromatography and analytical ultracentrifugation; on the other hand,
SDS-polyacrylamide gel electrophoresis analysis gives a subunit M, - 2f6 kDa,
("J
OH ~OH ~OH ~
tTl
Pinoresinoll :;0
"t:I
le- oxidant 0"""~ OMe lariciresinol
Dirigent protein
[:g
reductase tr'~
)3 r> OM'
NADPH
3 ~""" 0 ~"""HO
. OMe
t"'"
OH HOY HOY
I
OMe OMe
Coniferyl alcohol 14 (+)-Pinoresinol la (+)-Lariciresinol 19a r:/l
~
Z
("J
Pinoresinoll tTl
lariciresinol I NADPH [:g
reductase >-
t"'"
MeO MeO ~
OH Z
o r:/l
(-)-Secoisolariciresinol OH
HO HO
dehydrogenase ~
NAD(Pt ~
::r:
tTl
::tI
("J
OH OH
(-)-Matairesinol 3a (-)-Secoisolariciresinol 2a ~
r:/l
Figure 4. Biosynthetic pathway to the lignan (-)-matairesinoI3 in Forsythia interrnedia.

-..J
-..J
A OH
HO
j
OMe (±)-Dehydrodiconiferyl
alcohols ISallSb
OH
OH
M~~O"
OMe
I e- oxidation 8.8'
HO
OH o
(and other OMe (±)-Pinoresinols tall b
Coniferyl alcohol 14 resonance hybrids)
8-0-4'
H&HOg ~OH
O~
""" OMe
I
# OMe
OH erythrolthreo (±)-Guaiacylglycerol
8-0-4' coniferyl alcohol ethers 16a/16b
B
OH
le- oxidation
Dirigent protein
OH
Coniferyl alcohol 14 Putative enzyme-bound intermediate (+)-Pinoresinolla
Figure 5. A. Non-specific phenoxy radical coupling catalyzed by laccase. B. When coniferyl

alcohol 14 is incubated in presence of laccase and the dirigent protein, only (+}-pinoresinol la
is formed.
suggesting that the native protein may be a trimer.38 This 78 kDa protein, which
does not have any measurable (oxidative) catalytic capacity on its own, can, in
the presence ofa one-electron oxidant, align the coniferyl alcohol radicals formed
in such a way that stereoselective coupling occurs. The term dirigent protein
(Latin, dirigere: to guide or align) was introduced to describe the proposed unique
function of this 78 kDa protein. 38 Note, however, that stereoselective coupling is
only observed when E-coniferyl alcohol is used as a substrate, but not with either
E-p-coumaryl (17) or E-sinapyl (18) alcohols.
OH OH
MeO OMe
OH OH
p- Coumaryl alcohol 17 Sinapyl alcohol 18
To further study the specificity of the coupling reaction in vitro, two dis-
tinct but very similar cDNAs, psdFi 1 and psdFi2, encoding the dirigent protein,
were obtained40 by screening a cDNA library made from F. intermedia young
stems. 39 This was done by using a -370-base pair peR product generated from
primers designed from the amino terminus and several internal amino-acid
microsequences of the dirigent protein. The gene sequence encodes a protein of
18 kDa and the protein is preceded by a signal peptide. It contains four potential
N-glycosylation sites and four potential phosphorylation sites. It has no introns
and shows no homology to any other proteins of known function. This cDNA was
cloned into a baculovirus expression vector system and was expressed in
Spodoptera Jrugiperda Sf9 insect cells. SDS-PAGE revealed that the insect
culture produced recombinant dirigent protein consisting of - four bands ranging
in size from 22 to 27kDa (Fig. 6, lane 4), vs. the single band of 27kDa for the
native protein (Fig. 6, lane 2). When subjected to an enzymatic deglycosylation
procedure, only a single band of -18 kDa was present (Fig. 6, lane 5), which cor-
responds with the calculated molecular weight based on the amino acid sequence.
In vitro assays confirmed that only the (+)-enantiomer of pinoresinol (la) was
being formed when the recombinant dirigent protein was incubated with
E-coniferyl alcohol (14) and laccase as shown (Fig. 7).40 In addition, the dirigent
protein was also cloned into an inducible promoter vector and was expressed
in Schneider 2 (S2) Drosophila melanogaster cells. This convenient system
produces the dirigent protein' in stable transformed insect cells.
A 2 3 4 5
kDa
94.0-
67.0 -
43.0 -
30.0 -
20.1 -
Figure 6. A. SDS-PAGE and B. western blot of native
Forsythia intermedia (lanes I and 2) and recombinant
(Spodopterafbaculovirus expressed, lanes 3 to 5) diri-
B
2 3 4 5 gent protein. Lanes I (lOOllg) and 3 (25Ilg): crude total
30.0 - cellular protein extracts. Lanes 2 and 4 (OAllg, each):
purified proteins. Lane 5 contains OAllg of the degly-
20.1 - cosylated form of the recombinant protein: u
In flaxseed, it was shown that the opposite enantiomeric form of pin ore sino I
is formed: when flaxseed cell-free extracts are incubated in presence of E-[3H]-
coniferyl alcohol, only (-)-pinoresinol (1 b) is generated. 17 A flaxseed cDNA
library has been constructed and it is being screened to obtain the corresponding
dirigent protein gene. How both the dirigent protein genes and the correspond-
ing proteins from Forsythia and Linum differ in engendering the formation of (+)-
and (-)-pinoresinol (la and Ib), respectively, is of great interest.
PinoresinollLariciresinol Reductases
The bifunctional, NADPH-dependent, (+ )-pinoresinol/(+ )-Iariciresinol

reductase catalyzes the enantiospecific reduction of (+ )-pinoresinol (1 a) into (+)-
200 (+ )-pinoresinol ] a
a0-
:s
o
:~ 100
g
o
:g Figure 7. Chiral HPLC analy-
0::
sis of [9,9'-3HJ-pinoresinol 1
obtained following incubation
with lac case and recombinant
dirigent protein; as can be
o 10 20 30 seen essentially only the (+)-
Elution volume (ml) antipode la is formed.
lariciresinol (19a) and (+)-lariciresinol (19a) into (-)-secoisolariciresinol (2a),

respectively; (-)-pinoresinol (Ib) and (-)-lariciresinol (19b) do not serve as sub-
strates. 36A1 ,42 This enzyme was purified to apparent homogeneity from Forsythia
intermedia (>3,000 fold) and has a Mr - 35 kDa as shown by SDS-PAGE. 39 It is
a type A reductase since only the pro-R hydrogen on the nicotinamide ring of
NADPH was abstracted and transferred to the lignan product. 36
The cDNA clone for this gene, plrFi I, obtained by using techniques similar
to those used for the previous dirigent protein clones, was overexpressed in
Escherichia coli cells by using a pSBETa vector system,47 since it is an cytoso-
lic (nonglycosylated) protein. The heterologously expressed protein showed the
same enantiospecificity as the native protein. 39
With the pinoresinolliariciresinol reductase in hand from Forsythia inter-
media, attention was next directed to obtaining comparable reductases from
western red cedar (Thuja plicata), since it also forms significant amounts of
pinoresinol (I) derived lignans. By using a polymerase chain reaction guided
strategy, two distinct cDNAs, plrTp I and plrTp2, were isolated. 48 Interestingly,
when heterologously expressed in E. coli, the corresponding proteins catalyzed
distinct pinoresinolliariciresinol properties. The first, plrTp I, catalyzed the
NADPH-dependent conversion of (-)-pinoresinol (1 b) into (-)-lariciresinol
(19b), then to (+)-secoisolariciresinol (2b), whereas the second, plr-Tp2,
isolated from the same plant source, converted (+)-pinoresinol (Ia) into (+)-
lariciresinol (19a), and subsequently into (-)-secoisolariciresinol (2a).48 This
indicates that in western cedar two types of reductase exist with different
enantiospecifities. 48
Interestingly, in flax, which accumulates large amount of secoisolari-
ciresinol diglucoside (6), we have demonstrated that only (-)-pinoresinol is effec-
tively converted into (+ )-secoisolariciresinoi. A cDNA has recently been obtained
from the flaxseed cDNA library; it is currently being transferred to an E. coli
expression system (1.D. Ford, unpublished results, Washington State University)
to assess its enantiospecificity.
Secoisolariciresinol Dehydrogenase
The enzyme catalyzing the dehydrogenation of secoisolariciresinol (2) to

form matairesinol (3), secoisolariciresinol dehydrogenase, has been purified to
apparent homogeneity (>6,000 fold) from Forsythia intermedia stems. 46 It is
a NAD(Pt dependent reductase specific for (-)-secoisolariciresinol (2a) as
substrate, with only (-)-matairesinol (4a) formed as product; the corresponding
(+)-enantiomer (2b) is not used as substrate. 45 ,46
The cDNA encoding this enzyme has been isolated by using screening pro-
cedures similar to those described above. This 32 kDa cytosolic protein is encoded
by an 831 base pair cDNA.4h When expressed in E. coli, the recombinant protein
converted (-)-secoisolariciresinol (2a) into (-)-matairesinol (3a) as did the native
protein from F intermedia. Interestingly, the formation of an intermediary lactol
MeO MeO MeO

OH
OH
HO HO HO
OH OH OH
(-)-Secoisolariciresinol 2a Lactol20 (-)-Matairesinol 3a
Figure 8. Recombinant dehydrogenase in vitro conversion of (-)-secoisolariciresinol 2 into

(-)-matairesinoI3.
(20) compound was also detected (Fig. 8) with the recombinant protein, but not
with the corresponding native enzyme.
Secoisolariciresinol Glucosyltransferase
This protein catalyses the final step in SDO formation in flax (L. usitatis-
simum) seed, with maximum secoisolariciresinol glucosyltransferase activity
being observed at week two of flax seed development; it appears to be localized
in the seed coat (1.D. Ford, Washington State University, unpublished results). A
partially purified putative secoisolariciresinol diglucosyl transferase, when incu-
bated in the presence of UDP[ 1-3H]glucose, converts (±)-secoisolariciresinols
(2a/2b) into (+)-[3H]secoisolariciresinol diglucoside (6) as shown by HPLC
analysis and liquid scintillation counting. Further confirmation of the product was
obtained by IH NMR analysis of the acetylated derivative of the enzymatic
product. This glucosyltransferase, capable of converting secoisolariciresinol (2)
into SDO is being purified from flax seeds.
ENHANCED LEVELS OF CANCER-PREVENTING

LIGNANS IN PLANTS: TOWARDS ENGINEERING THEIR
METABOLIC PATHWAYS
With essentially all of the genes encoding the biochemical pathway from
coniferyl alcohol to secoisolariciresinol and matairesinol in hand, the opportunity
now exists to biotechnologically enhance existing levels in lignan-producing
plants (e.g. flax) or to introduce (regulatory genes or perhaps even the entire
pathway) into those which do not normally biosynthesize the lignans to any con-
siderable extent (e.g. rice, rye and wheat). As shown in Tables 3 and 4, lignans
differentially accumulate in foods, such as wheat, flax, sesame, fruits, and veg-
etables. Accordingly, manipulation of regulatory enzymatic steps would appear
to be a useful strategy to enhance/increase levels of these beneficial compounds.
Any biotechnological approach must, however, be competitive with other

approaches that might also be suitable for achieving an increased dietary intake
of these plant lignans. These include: increasing intake of the food item(s)
containing the particular phytochemical; increasing the concentrations of the
product through conventional plant breeding and/or using biotechnology and gene
transfer; selective processing such as milling or extraction to enhance the con-
centration of the chemical in the food; and direct addition of the phytochemical
to the food. 49 Of these options, the use of biotechnology and gene transfer is
perhaps the most effective approach, on the basis of achieving both relatively
short timelines, as well as minimal intervention and alteration of normal dietary
habits.
Moreover, the prognosis for bioengineering metabolic pathways of cancer-
preventing lignans is feasible, given the numerous examples of engineered plants,
e.g. tomato, cotton, potato, and pea (virus), for purposes such as insect control
and disease resistance. Indeed, examples of food crop plants bioengineered for
improved production traits include Bacillus thuringiensis toxin genes introduced
into transgenic tomato, maize, rice, potato, and soybean,50 and other crops have
been successfully bioengineered to improve shelf storage life or nutritional
value. 51.52 Currently few, if any, food plants have been bioengineered specifically
for the purpose of enhancing health benefits to the consumer. Shintani and Del-
laPenna 53 successfully elevated vitamin E content in Arabidopsis seeds, by over-
expressing y-tocopherol methyltransferase, which limits the step in the conversion
of y-tocopherol, to a-tocopherol. It is hoped that this can be applied to soybean,
corn, and canola. Thus, the potential exists for developing transgenic plants to
contain increased levels of mammalian lignan precursors.
In addition to obtaining the genes producing the specific enzymes in the
lignan biosynthetic pathways, most oftheir promoter elements have also been iso-
lated (J.H. Jeon, Washington State University, unpublished results). Accordingly,
the goal of increasing levels of these phytoestrogen precursors and directing their
location of synthesis and accumulation in normally consumed foods, such as
cereal grains, seems to be testable at this point. Given that the lignan biosynthetic
pathways are present in many important food crops, such as wheat, flax, sesame,
and numerous fruits and vegetables, this allows for potential manipulation of the
enzyme components and locations of product accumulation through genetic engi-
neering. Indeed, our first attempts are now being made to manipulate lignan con-
tents in foods in order to promote health, by increasing the levels of dirigent
protein, pinoresinolllariciresinol reductase, and secoisolariciresinol dehydroge-
nase in plants such as flax. The "nutraceutically" enhanced foods, which are
defined as food substances that provide medical or health benefits,49.s4 produced
by these manipulations should provide a novel approach to the prevention and
treatment of diseases such as cancer by allowing for increased intake of biolog-
ically active compounds found in natural plant material that is consumed in the
normal daily diet.
ACKNOWLEDGMENTS
The authors thank the United States Department of Agriculture (9603622)

and Mcintire-Stennis, for generous support of these studies.
REFERENCES
I. SETCHELL, K.D.R., ADLERCREUTZ, H. 1988. Mammalian lignans and phyto-oestro-

gens. Recent studies on their formation, metabolism and biological role in health and
disease. In: Role of the Gut Flora in Toxicity and Cancer. (Rowland, LR., ed.) Academic
Press, London, pp 315-345.
2. LEWIS, N.G., DAVIN, L.B. 1999. Lignans: Biosynthesis and function. In: Comprehen-
sive Natural Products Chemistry. Vol. \. (Barton, Sir D.H.R., Nakanishi, K., Meth-Cohn,
0., eds), Elsevier, London, pp 639-712.
3. FAURE, M., LISSI, E., TORRES, R., VIDALA, L.A. 1990. Antioxidant activities of
lignans and flavonoids. Phytochemistry 29: 3773-3775.
4. HARMATHA, J., NAWROT, 1. 1984. Comparison of the feeding deterrent activity of
some sesquiterpene lactones and a lignan lactone towards selected insect storage pests.
Biochem. Syst. Ecol. 12: 95-98.
5. MACRAE, D.W., TOWERS, G.H.N. 1984. Biological activities of lignans. Phytochem-
istry 23: 1207-1220.
6. AYRES, D.C., LOIKE, 1.D. 1990. Chemistry and Pharmacology of Natural Products.
Lignans. Chemical, Biological and Clinical Properties. Cambridge University Press,
Cambridge, England, pp. 402.
7. ADLERCREUTZ, H., HOECKERSTEDT, K., BANNWART, c., BLOIGU, S.,
HAMALAINEN, E., FOTSIS, T., OLLUS, A. 1987. Effect of dietary components, includ-
ing lignans and phytoestrogens, on enterohepatic circulation and liver metabolism of
estrogens and on sex hormone binding globulin (shbg). 1. Steroid Biochem. 27:
1135-1144.
8. HERMAN, c., ADLERCREUTZ, H., GOLDIN, B.R., GORBACH, S.L., HOCKERST-
EDT, K.A.Y., WATANABE, S., HAMALAINEN, E.K., MARKKANEN, M.H.,
MAKELA, T.H., WAHALA, K.T., HASE, T.A., FOTSIS, T. 1995. Phytoestrogen intake
and cancer risk. 1. Nutr. 125: 757-770.
9. KORPELA, 1.T., KORPELA, R., ADLERCREUTZ, H. 1992. Fecal bile acid metabolic
pattern after administration of different types of bread. Gastroenterology 103: 1246-1253.
10. SLAVIN, 1., JACOBS, C., MARQUART, L. 1997. Whole-grain consumption and chronic
disease: protective mechanisms. Nutr. Cancer 27: 14-21.
II. ADLERCREUTZ, H. 1984. Does fiber-rich food containing animal lignan precursors
protect against both colon and breast cancer? An extension of the "fiber hypothesis". Gas-
troenterology 86: 761-764.
12. HORWITZ, c., WALKER, A.R.P. 1984. Lignans. Additional benefits from fiber. Nutr.
Cancer 6: 73-76.
13. KURZER, M.S., Xu, X. 1997. Dietary phytoestrogens. Annu. Rev. Nutr. 17: 353-38\.
14. ADLERCREUTZ, H. 1996. Lignans and isoflavonoids: Epidemiology and possible role
in prevention of cancer. In: Natural Antioxidants and Food Quality in Atherosclerosis and
Cancer Prevention. (Kumpulainen, J.T., Salonen, J.K., eds), The Royal Society of
Chemistry, Cambridge, England, pp 349-355.
15. THOMPSON, L., ORCHESON, L., RICKARD, S., JENAB, M., SERRAINO, M.,
SEIDL, M., CHEUNG, F. 1996. Anticancer effects of flaxseed Iignans. In: Natural
Antioxidants and Food Quality in Atherosclerosis and Cancer Prevention. (Kumpulainen,
J.T., Salonen, J.K., eds), The Royal Society of Chemistry, Cambridge, England, pp
356-364.
16. BORRIELLO, S.P, SETCHELL, K.D.R., AXELSON, M., LAWSON, A.M. 1985.
Production and metabolism of lignans by the human faecal flora. l Appl. Bacteriol.
58: 37-43.
17. FORD, J.D., DAVIN, L.B., LEWIS, N.G. 1999. Lignans and health: Cancer chemopre-
venti on and biotechnological opportunities with plant lignans. In: Plant Polyphenols 2:
Chemistry and Biology. (Gross, G.G., Hemingway, R.W, Yoshida, T., eds), Plenum Press,
New York, pp ( •• ).
18. SETCHELL, K.D.R., LAWSON, A.M., MITCHELL, F.L., ADLERCREUTZ, H., KIRK,
D.N., AXELSON, M. 1980. Lignans in man and in animal species. Nature 287: 740-742.
19. STITCH, S.R., TOUMBA, J.K., GROEN, M.B., FUNKE, C.w., LEEMHUIS, 1., VINK,
1., WOODS, G.F. 1980. Excretion, isolation and structure of a new phenolic constituent
of female urine. Nature 287: 738-740.
20. NESBITT, P.D., THOMPSON, L.u. 1997. Lignans in homemade and commercial
products containing flaxseed. Nutr. Cancer 29: 222-227.
21. ADLERCREUTZ, H., MAZUR, W 1997. Phyto-oestrogens and Western diseases. Anal.
Med. 29: 95-120.
22. MAZUR, WM., DUKE, lA., W AHALA, K., RASKU, S., ADLERCREUTZ, H. 1998.
Isoflavonoid and lignan in legumes: Nutritional and health aspects in humans. Nutr.
Biochem. 9: 193-200.
23. THOMPSON, L.u., ROBB, P, SERRAINO, M., CHEUNG, F. 1991. Mammalian lignan
production from various foods. Nutr. Cancer 16: 43-52.
24. lEWIS, N.G., DAVIN, L.B., SARKANEN, S. 1999. The Nature and function of lignins.
In: Comprehensive Natural Products Chemistry. Vol. 3. (Barton, Sir D.H.R., Nakanishi,
K., Meth-Cohn, 0., eds), Elsevier, London, pp 618-739.
25. REDDY, B.S., COHEN, L.A., McCOY, G.D., HILL, P., WEISBURGER, 1.H., WYNDER,
E.L. 1989. Nutrition and its relation to cancer. Adv. Cancer Res. 32: 327-345.
26. KAMAL-ELDIN, A., PETTERSSON, D., APPELQVIST, L.-A. 1996. The in vivo antiox-
idant properties of sesame lignans. In: Natural Antioxidants and Food Quality in Ather-
osclerosis and Cancer Prevention. (Kumpulainen, J.T., Salonen, J.K., eds), The Royal
Society of Chemistry, Cambridge, England, pp 230-235.
27. AKIMOTO, K., ASAMI, S., TANAKA, T., SHIMIZU, S., SUGANO, M., YAMADA, H.
1996. Antioxidant activity of sesamin on NADP-dependent lipid peroxidation in liver
microsomcs. In: Natural Antioxidants and Food Quality in Atherosclerosis and Cancer
Prevention. (Kumpulainen, J.T., Salonen, 1.K., cds), The Royal Society of Chemistry,
Cambridge, England, pp 241-246.
28. MAZUR, WM., W AHALA, K., RASKU, S., SALALlA, A., HASE, T., ADLER-
CREUTZ, H. 1998. Lignan and isoflavonoid concentrations in tea and coffee. Br. 1. Nutr.
79: 37-45.
29. GOEl, H.C., PRASAD, H.C., SHARMA, A. 1998. Antitumor and radioprotective action
of Podophyllum hexandrum. Ind. 1. Exp. BioI. 36: 583-587.
30. KO, K.M., IP, S.P., POON, M.K.T., WU, S.S., CHE, c.T., NG, K.H., KONG, Yc. 1995.
Effect of a lignan-enriched fructus schisandrae extract on hepatic glutathione status in
rats: protection against carbon tetrachloride toxicity. Planta Medica 61: 134-137.
31. IP, S.P, MAK, D.H.F., LI, pc., POON, M.K.T., KO, K.M. 1996. Effect of a lignan-
enriched extract of Schisandra chinensis on aflatoxin BI and cadmium chloride-induced
hepatotoxicity in rats. Pharmacology and Toxicology 78: 413-416.
32. YAMADA, S., MURAWAKI, Y, KAWASAKI, H. 1993. Preventive effect of gomisin A,
a lignan component of Schizandra fruits, on acetaminophen-induced hepatotoxicity in

rats. Biochemical Pharmacology 46: 1081-1085.
33. OLIVETO, E.P. 1972. Nordihydroguaiaretic acid: A naturally occuring antioxidant.
Chern. Ind.: 677-679.
34. GNABRE, IN., BRADY, IN., CLANTON, OJ., ITO, Y, DITTMER, I, BATES, R.B.,
HUANG, R.C.C. 1995. Inhibition of human immunodeficiency virus type I transcription
and replication by DNA sequence-selective plant lignans. Proc. Natl. Acad. Sci. USA 92:
11239-11243.
35. GNABRE, IN., ITO, Y, MA, Y, HUANG, R.C. 1996. Isolation of anti-HIV-I lignans
from Larrea tridentata by counter-current chromatography. I Chromatogr. A 719:
353-364.
36. CHU, A., DINKOVA, A., DAVIN, L.B., BEDGAR, D.L., LEWIS, N.G. 1993. Stere-
ospecificity of (+)-pinoresinol and (+ )-Iariciresinol reductases from Forsythia intermedia.
1. BioI. Chern. 268: 27026-27033.
37. DAVIN, L.B., BEDGAR, D., KATAYAMA, T, LEWIS, N.G. 1992. On the stereoselec-
tive synthesis of (+ )-pinoresinol in Fonythia saspensa from its achiral precursor,
coniferyl alcohol. Phytochemistry 31: 3869-3874.
38. DAVIN, L.B., WANG, H.-B., CROWELL, A.L., BEDGAR, D.L., MARTIN, D.M.,
SARKANEN, S., LEWIS, N.G. 1997. Stereoselective bimolecular phenoxy radical cou-
pling by an auxiliary (dirigent) protein without an active center. Science 275: 362-366.
39. DINKOVA-KOSTOVA, A.T., GANG, D.R., DAVIN, L.B., BEDGAR, D.L., CHU, A.,
LEWIS, N.G. 1996. (+)-Pinoresinoll(+)-Iariciresinol reductase from Fonythia interme-
dia: Protein purification, eDNA cloning, heterologous expression and comparison to
isoflavone reductase. 1. BioI. Chern. 271: 29473-29482.
40. GANG, D.R., COSTA, M.A., FUJITA, M., DINKOVA-KOSTOVA, A.T., BURLAT, V,
MARTIN, W, SARKANEN, S., DAVIN, L.B., LEWIS, N.G. 1999. Regiochemical
control of monolignol radical coupling: A new paradigm for lignin and lignan biosyn-
thesis. Chemistry & Biology 6: 143-151.
41. KATAYAMA, T, DAVIN, L.B., LEWIS, N.G. 1992. An extraordinary accumulation of
(-)-pinoresinol in ccll-free extracts of Forsythia intermedia: Evidence for enantiospecific
reduction of (+ )-pinoresinol. Phytochemistry 31: 3875-3881.
42. KATAYAMA, y, DAVIN, L.B., CHU, A., LEWIS, N.G. 1993. Novel benzylic ether
reductions in lignan biogenesis in For.lythia intermedia. Phytochemistry 33(3): 581-591.
43. UMEZAWA, T, DAVIN, L.B., KINGSTON, D.G.!., YAMAMOTO, E., LEWIS, N.G.
1990. Lignan biosynthesis in Forsythia sp. 1. Chern. Soc., Chern. Commun.: 1405-1408.
44. UMEZAWA, T, DAVIN, L.B., LEWIS, N.G. 1990. Formation of the lignan, (-)-
secoisolariciresinol by cell-free extracts of Forsythia intermedia. Biochem. Biophys. Res.
Commun. 171(3): 1008-1014.
45. UMEZAWA, T, DAVIN, L.B., LEWIS, N.G. 1991. Formation of lignans, (-)-
secoisolariciresinol and (-)-matairesinol with For.lythia intermedia cell-free extracts.
1. BioI. Chern. 266: 10210-10217.
46. XIA, Z.-Q., COSTA, M.A., DAVIN, L.B., LEWIS, N.G. 1999. Purification and charac-
terization of (-)-secoisolariciresinol dehydrogenase: Cloning and recombinant expres-
sion.: (submitted).
47. SCHENK, P.M., BAUMANN, S., MATTES, R., STEINBIB, H.-H. 1995. Improved high-
level expression system for eukaryotic genes in Escherichia coli using T7 RNA poly-
merase and rare AcgtRNAs. BioTechniques 19: 196-200.
48. FUJITA, M., GANG, D.R., DAVIN, L.B., LEWIS, N.G. 1999. Recombinant
pinoresinolllariciresinol reductases from western red cedar (Thaja plicata) catalysc oppo-
site enantiospecific conversions. 1. BioI. Chern. 274: 618-627.
49. HAUMANN, B.F. 1993. Designer foods. Inform 4: 354-366.
50. HILDER, VA., GATEHOUSE, A.M.R. 1990. Transforming plants as a means of crop
protection against insects. Outlook on Agriculture 19: 179-183.
51. YOUNG, R. 1995. Improved tomato products: Utilising modern biotechnology. Food
Australia: Official journal of CAFTA and AIFST 47: 203-204.
52. KRAMER, M.G., REDENBAUGH, K. 1994. Commercialization of a tomato with an
antisense polygalacturonase gene: the FLAVR SAVR® tomato story. Euphytica 79:
293-297.
53. SHINTANI, D., DELLAPENNA, D. 1998. Elevating the vitamin E content of plants
through metabolic engineering. Science 282: 2098-2100.
54. BROWER, V 1998. Nutraceuticals: Poised for a healthy slice of the healthcare market?
Nature Biotech. 16: 728-731.
Chapter Five
SIMPLE (BENCH-TOP) BIOASSAYS AND THE

ISOLATION OF NEW CHEMICALLY DIVERSE
ANTITUMOR AND PESTICIDAL AGENTS
FROM HIGHER PLANTS
Jerry L. McLaughlin and Ching-Jer Chang
Department of Medicinal Chemistry and Molecular Pharmacology

School of Pharmacy and Pharmacal Sciences
Purdue University
West Lafayette, Indiana 47907
Introduction ............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 89
Methodology .................................................. 91
Recent Progress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 93
Non-Acetogenin Compounds ................................... 93
New Annonaceous Acetogenin Compounds ........................ 100
Biological Studies with the Annonaceous Acetogenins ............... 120
Summary ..................................................... 124
INTRODUCTION
Phytochemists cannot easily survive as basic scientists. Applied science has

become essentially the only source for significant amounts of research funds. In
our market-driven economy, government and industry have forced us to face this
fact. A major purpose of science today is to solve the problems of society. So
which of society's problems are solvable by phytochemistry? The answers to this
question are numerous, but our laboratory for the past 20 years has focused on
quests for new antitumor and pesticidal agents from higher plants. Cancer death
claims 1,500 people a day (one per minute) in the United States. l The market for
89
90 1. L. McLAUGHLIN AND C.-J. CHANG
natural, environmentally friendly pesticides will gross nearly one billion dollars
in 1998. Thus, the discovery of phytochemicals as new antitumor and pesticidal
agents is justifiable and should be fundable.
Advances in separation technology (chromatography) and in structural elu-
cidation methods (spectral methods and x-ray crystallography) have simplified
the isolation and characterization of new phytochemicals. Bioassays remain
the most insurmountable roadblock in our quest for new bioactive substances.
Examples of such are: the complex panel of 60 human tumor cell lines at the
National Cancer Institute (NCI, Frederick, Maryland); the robotic, mechanism-
based, screening methods employed by the U.S. pharmaceutical industry; the use
of human tumor xenografts in athymic mice; and the costly panels of indicator
pests maintained by industrial pesticide researchers.
For the independent phytochemical investigator, establishment of cell
culture laboratories requires a full-time commitment of manpower and special-
ized equipment that consumes one's entire research budget. Here at Purdue Uni-
versity we have been aided by a Cell Culture Laboratory funded through the
Purdue Cancer Research Center with NCI support; yet the fees for these services
(up to $120.00/sample in six cell lines) are exorbitant, and we can afford only
confirmatory cell culture tests for our isolated compounds. Similarily, a battery
of mechanism-based assays is too expensive to create and maintain, and we, at
best, can only operate a few of these (protein kinase C, protein-tyrosine kinase)
in our new anticancer drug discovery efforts.2.3 Similarily, the cost of sterile
quarters for maintenance and testing of athymic mice is prohibitive to most
academicians. In pesticide research, the growth and propagation of indicator
pests is perhaps less troublesome and burdensome, but it still consumes funds for
specialized facilities, pest containment, and maintenance manpower. We have
concluded that our best investment of research funds is not on complicated
biological methodology, but on the chemical side of this work.
Thus, our laboratory has sought, developed, and employed general, uncom-
plicated, and affordable bioassay alternatives. Four simple, yet effective, bench-
top bioassays have given us the advantage of detecting both novel and known
phytochemicals with a broad range of structural as well as biological diversity.
These bioassays are: (l) lethality to the larvae of brine shrimp (Artemia salina)
(a rapid, general, bioassay for antitumor and pesticidal compounds), (2) the inhi-
bition of crown gall tumors, induced by plasmid transfer and expression from
Agrobacterium tumefaciens, on discs of potato (Solanum tuberosum) tubers (an
antitumor bioassay), (3) the inhibition or stimulation of frond proliferation of
duckweed (Lemna minor) (an assay for plant growth stimulants and inhibitors),
and (4) lethality to the larvae of yellow fever mosquitoes (Aedes aegyptii) (a test
for pesticides).
Since 1984, over 320 diverse bioactive plant components have been
isolated and characterized in our laboratory by using these methods both for
screening the extracts and for directing the fractionations. Three previous
BIOASSAYS AND ISOLATION OF NEW AGENTS 91
reviews summarize our progress until 1993. 4- 6 This paper updates our progress to
1997.
METHODOLOGY
The materials and methods required for the brine shrimp and potato disc
bioassays, respectively, are described in our original papers/'s and the methods
are updated in later reports including a validation studyY In a recent article, all
of the currently used methods for all four bioassays are described with the back-
grounds, complete references, and statistical considerations. 'o These bioassays
have now become popular. Analyses of literature citations show that the original
brine shrimp test (BST) paper has been quoted hundreds of times and is the most
useful of these methods. 7 These methodologies have been taught by us in over 20
international workshops, and an effort has been made to bring the bioassays
directly to the people where most of the uninvestigated plant species of the world
are located (SE Asia, Africa, Central America, and South America). To conserve
space, only the materials and methods of the BST are presented herein (Table 1,
Fig. 1). Recent papers describe these procedures in Spanish and Chinese and
extend the reading audience. 11,12
In our laboratory, each researcher conducts his/her own brine shrimp bioas-
says on his/her own bench. This simple methodology promotes self-reliance and
rapid turn around (24 hours), and these advantages are important for this type of
work. All known antitumor agents (except prodrugs) are detected by the brine
shrimp and potato disc tests, and all known pesticides, tested so far, are detected
Compound or Extract : O. 5 ml
(20 mg) + 2 ml s o l v c n t - - - - - - - - - - - - - - - + - - -......~ 0.5 ml
1------;- 0.5 ml
for 100 ppm take

0,5ml
0.2 ml + 1.8 ml solvent -----------+---I-~ : 0.5 ml
~----I- 0.5ml
for 10 ppm take
0.2 ml + 1.8 ml solvent

,-----1----
- - - - - - - - - - + - -......
: ..
0.5 ml
0.5 ml
'------;-- 0.5 ml
for 1 ppm take
0.5ml
0.2 ml + 1.8 ml solvent :
------+----II~ 0.5 ml
~--- O.Sml
Figure 1. Flow chart for alternate dilution procedure for brine shrimp bioassay.
92 J. L. McLAUGHLIN AND C.-J. CHANG
Table 1. Materials and procedures for brine shrimp lethality bioassay

Materials
Artemia salina cysts (brine shrimp eggs from pet store). Most pet stores carry these as a type
of fish food.
Sea salt (also from pet store).
Small tank (hatching chamber) to grow shrimp with dividing dam, cover, and lamp to attract
shrimp (The hatching chamber can be conveniently made from a covered plastic soap
dish).
Syringes: 5ml, 0.5ml, 100JlI, and IOJlI (smaller syringes are not needed if dilutions are
prepared in the alternative dilution procedures as described in Figure I).
Two dram vials (9 per sample + I control)
Procedures
1. Prepare sea water according to directions on box (38 g sea salt per liter of water), filter.
2. Put sea water in small tank, add shrimp eggs to one side of the divided tank, and cover
this side. The lamp above the other side will attract the hatched shrimp.
3. Allow two days for the shrimp to hatch and mature as nauplii (in warmer climates,
hatching may take place sooner, but the authors use 48-72hr nauplii to be consistent).
4. Prepare vials for testing; for each fraction, test initially at 1,000, 100, 10 Jlglml; prepare
three vials at each concentration for a total of nine vials; weigh 20 mg of sample and add
2 ml of solvent (20 mgl2 ml); from this solution transfer 500, 50, or 5 JlI to vials
corresponding to 1,000, 100, or 10 Jlg/ml, respectively. Evaporate solvent under nitrogen
and then put under high vacuum for about 30 min; volatile solvents will evaporate
overnight. Alternatively, materials may be dissolved in DMSO (dimethylsulfoxide), and up
to 50 JlI may be added per 5 ml of brine before DMSO toxicity will affect the results. An
alternative dilution procedure is illustrated in Fig. I; this procedure avoids the need for
100 JlI and 10 JlI syringes.
5. After two days (when the shrimp larvae are ready), add about 4 ml of sea water to each
vial, count 10 shrimp per vial (30 shrimp per dilution), and adjust the volume with sea
water to 5 mllvial. Place the vials, uncovered, under the lamp. Be sure that vials are not
overheated by the lamp.
6. Twenty-four hours later count and record the number of survivors.
7. Analyze the data with the Finney computer program for probit analysis to determine LC 50
values and 95% confidence intervals. A copy of this program for IBM PCs is available
from J.L. McLaughlin; a $10.00 donation is requested to cover costs, postage, and
handling.
8. Additional dilutions at less than 10 Jlg/ml may be needed to determine the LC so values for
potent materials; also, intermediate concentrations, e.g., at 750, 500, and 250 Jlg/ml can be
prepared and tested to narrow the confidence intervals. By starting with 2mglml (step 4
above) dilutions at 100, 10, and I Jlg/ml are easily prepared for more potent materials.
by the brine shrimp and mosquito larvae tests. Thus, false negatives are not a
major problem. In screening randomly collected plants (over 3,500 species, so
far), extracts of about 5% have shown activity at LC 50 values <500ppm in
the BST. Those with LC 50 values <200 ppm are worthy of confirmatory testing in
the potato disc test and then in cell culture in our search for new antitumor
compounds.
In the past four year period (1993-1997), 98 papers from our laboratory
describing over 200 compounds and their biological effects have been published
dealing with this research. Thus, a good level of productivity is possible with a
small group of dedicated scientists using bioassays that really work. Combinato-
rial chemists, molecular biologists, and marine chemists now all compete with
phytochemists for research funds. Consequently, work on higher plants in the
U.S., in the future, will have to be conducted with some unusual flair to give it a
competitive edge. Plant research funds are increasingly difficult to obtain.
Researchers in the countries where unique uninvestigated plant species are abun-
dant are especially urged to collect, screen, and fractionate their available plant
species. New compounds are out there waiting to be found and applied to the
needs of society, and action must be taken soon before the dwindling plant species
of these regions are made extinct.
RECENT PROGRESS
Using the above methods, since 1993, our laboratory has found 207 bioac-
tive compounds which are briefly described and biologically evaluated in this
review. Confirmatory cytotoxicity tests have been made in the panel of human
tumor cell lines at the Cell Culture Laboratory, Purdue Cancer Center. These cell
lines were carefully selected to cover the major types of problematic human solid
tumors. They are A-549 (lung), MCF-7 and MCF-71 Adr (breast), HT-29 (colon),
A-498 (renal), PC-3 (prostate), and PACA-2 (pancreatic). Resistance ratios
(MCF-7 vs. MCF-7/adr, which is resistant to adriamycin) help to select com-
pounds and extracts which are effective against multiple drug resistant (MDR)
tumors. ED50 values are presented in Ilg/ml (ppm). Our seven day MTT assay is
based on mitochondrial reduction of the tetrazoliurn to a blue formazan dye. 13 At
Eli Lilly and Company, bullatacin (61) gave an ED50 value of 1O-14 1lg/m l against
human leukemia (CRF-LEM) cells, and, at Abbott Laboratories, the adjacent bis-
tetrahydrofuran (THF) Annonaceous acetogenins (51, 61, 70) all gave cytotoxic-
ity ED50 values showing potencies at least three orders of magnitude greater than
adriamycin; thus, we have confidence in, and independent confirmation of, the
high levels of cytotoxic potencies presented herein. Cytotoxic ED50 values <41lg1
ml are considered significant. 14 Adriamycin is always run as a positive control.
Most of our recent work has been focused on the potent antitumor and pes-
ticidal acetogenins from the Annonaceae. However, several non-acetogenin com-
pounds are quite promising and will be discussed first followed by the
presentation of our new actogenins and a discussion of various biological studies.
Non-Acetogenin Compounds
I. Euphorbia lagascae Spreng. (Euphorbiceae): We isolated piceatannol

(3,4,3',5'-tetrahydroxy-trans -stilbene) from the seeds of this species from Spain
in 1984. 15 Its in vivo (P388, murine leukemia, determined at NCI) antitumor activ-
ity is due to inhibition of protein tyrosine kinases (PTKs).16 We have determined
94 1. L. McLAUGHLIN AND c.-1. CHANG
that it selectively inhibits, at the substrate binding site, the Syk PTK that medi-
ates the degranulation of antigen-stimulated RBL-2H3 rat tumor mast cells.l7
Thus, this compound and/or its analogs may find future application in the
prevention or treatment of allergies as well as tumors.
2. Knema glomerata Merrill (Myristicaceae): Work on the stem bark ofthis
Philippine species led to the isolation of one new phenylalkylphenol, kneglom-
eratanol (1), and two new acetophenones, kneglomeratanones A (2) and B (3),
together with seven known phenolic compounds (4-10) and three known fla-
vanoids. All of these compounds showed moderate to significant toxicities to three
human tumor cell lines, inhibited the growth of the potato disc crown gall tumors
(PO), and were lethal in the brine shrimp test (BST).18 Since these compounds
do not show selective cytotoxicities, they are not being studied further.
3. Melia volkensii Gurke (Meliaceae): The root bark of this folkloric
medicine from Kenya has yielded ten cytotoxic triterpenes: meliavolin (11),
meliavolkin (12), and melkianin A (13);19 meliavolen (14), melianione (15), 3-
episapelin A (16), and nimbolin B (17)/° meliavolkenin (18);21 and meli-
avolkensins A and B (19, 20).22 Seven of these (11, 12, 14, 15, 18-20) are new
compounds. X-ray crystallography of 13 and the acetate of 11, Mosher esters,
derivative preparations, and spectral analyses (especially NOESY) were used to
elucidate the structures. All of these compounds (especially 17) showed interest-
ing selectivities for the colon cell line (HT-29). Work is continuing with this
species.
OH
o
l' : 2" R,
HO~' (CH2)6~-
2' 9' \0' 0 1"
oj'
HO~'
2 6 2' 0
(CH ,)n ~ CH 3
3 5
4
OH
3 Rl=COCH3, n=8 6 R2=H, n=10
R,
, '"""~: R'O¥(CH')~
~ R2
2 Rj=COCH3. R2=OH, n=4

4 Rj=R2=H, n=8
9 R j =COCH3. R2=OH, n=6
5 Rl =R2=H, n=6
10 Rj=COOH, R2=H, n=8
7 Rl=H, R2=OH, n=6
8 Rl=H, R2=OH, n=8

Cmpd BST LCSO PO EOSO Human Cancer Cell Lines ( g/mI)
(~g/mI) % TIC A-549 MCF-7 HT-29

I I
0.21 29 5.99 2.32 9.33
2 0.37 38 3.89 3.44 18.84
3 1.92 38 2.30 1.12 6.S6
4 1.37 39 1.85 2.09 2.62
5 0.45 54 35.01 7.76 28.59
6 7.65 41 2.37 2.41 3.25
7 0.47 62 3.49 2.69 3.24
8 0.33 45 3.26 1.84 2.87
9 0.18 30 3.22 2.87 4.47
10 1.24 61 IS.00 3.06 2.98
27
o Ul o
12
if'.
; - 21
,01 A
·0
5' 7
I o'~9 AcO'"
3
6'
11 (11 a, 25a-OH, 11 b, 25b-OH) 12
8t
~'7 OH
18
.
'1 '
cr
~ E 24 26
(t
17 20 "OH
0 3
"O A 4'?'" 0\\"
~'r' 2' l' 0" S' 17' 29 28
5' ~ I 7' '"
13
2~h
20
",,\.~
E
24
23
26
14, R ~ H
96 1. L. McLAUGHLIN AND C.-1. CHANG
16
H 9H
HO HO~c:>VV27
;~~23 l'oH
-".:. 26
1
17
H
r=':
14:"" 13"""'" ...... ,,~o
' 0 3 A
Acd" "OCOC= CHCH 3 4(
'7 (
'1' 0\"
I 5';:'-" I 7' 29
CH3 6'
17 18
0 26
4'r
5'
(f
:;:,-..
1'1'3
0'"
2. ~H
HI"
21
OCH 3
2 'I 27
~OCH3
o ,).('H
6' 19, R= 20 23 20,R=
Cmpd BST A-549 MCF-7 HT-29

.,
II
12
0.34
1.83
11.25 ., 5.34., 9.5x10.,
5,7xlO 2,6xlO 1.2xlO
13 1.83 6,65 3,61 2,3]
.,
14
15
54
88,5
4,33
2.51
5,66
>]0
8.49x10 .,
8.48x]0
16 154.4 5,98 9,77
., 1.46
.,
·3
17 20,1 3.12 5,90xlO <]0
18 0,55 10.33 4.30 6,7xlO .,
19
20
106.6
89.3
13.26
10.63
3.06
2.02
4.9xlO .,
2.5xlO
4. Persea americana MilLvar. americana (Lauraceae): The bark of this

tree, commonly known as avocado, a native of Mexico, yielded a new C-20 alkyl-
alkene methyl ester designated as persealide (22); 22 was cytotoxic with moder-
ate EDso values (I-Lg/ml) in A-549 (3.07), MCF-7 (2.87), and HT-29 (3.17).23
Continued work with the green fruits, which are considerably more bioactive, has
led to three linear trihydroxylated hydrocarbons, 1,2,4-trihydroxynonadecane
(23), I ,2,4-trihydroxyheptadec-16-ene (24), and 1,2,4-trihydroxyheptadec-16-yne

(25); all three show selective cytotoxicities toward the prostate cell line (PC-3),
with activities in the range of adriamycin. The acetylenic compound (25) is ten
times more potent than rotenone in the yellow fever mosquito larvae (YFM)
test. 24 We hope to obtain sufficient amounts of 25 for in vivo testing against
PC-3.
OH OH
OH
16
22 23, R= -CH,CH,CH, 24, R= -CH=CH, 25, R= -C",CH
MCF·7 A-549 HT-29 A-498 PC-3 PACA-2 YFM
Cmpd
23 2.98 3.21
I 3.01
I 2.69
I 1.17 4.65 1.79
24 3.41 4.43 2.62 3.62 4.62xlO·' 5.28 2.13xI0·'
25 4.84 6.52 8.94 4.06 6.07xI0·' >10 7.76xlO·'
adriamycin 2.2IxW' 8.05xI0·' 5.34xlO·) 4.85xI0·) I.IlxIO·) 2.42xlO·)
rotenone 7.46xI0·'
5. Euphorbia poisonii Pax (Euphorbiaceae): The latex of this Nigerian tree

is used as a pesticide as well as for homicides; it is extremely irritating and causes
blindness when in contact with the eyes. Certain phorbol esters had been previ-
ously isolated. BST directed fractionation led us to nine compounds: bioactive
19-acetoxyingol-3, 12-diacetate-8-nicotinate-7-phenylacetate (26) and its less
active congeners, 19-acetoxyingol-3, 12-diacetate-8-hydroxy-7 -phenyl acetate
(27) and 19-acetoxyingol-3-acetate-8, l2-dihydroxy-7 -phenylacetate (28),25 and a
new tigliane diterpene, 12-deoxyphorbol 13-(9, I O-methylene) undecanote (29),
together with five known diterpenes (30-34). In our panel of six human solid
tumor cell lines, all but 28 were selectively cytotoxic for the kidney carcinoma
cell line (A -498) with the potency of 31 exceeding that of adriamycin by ten thou-
sand times;26 potent skin irritancy was noted when handling 33 (resininferatoxin);
it is an irritant by interacting with vanilloid receptors, expressed by polymodal
primary sensory neurons.27 Hitherto unknown compound 31 demonstrated the
strongest bioactivities and showed excellent selectivity of over one million times
for the A-498 (renal) cell line; the presence of the cycIopropyl ring in the unusual
l3-(9,IO-methylene) undecanoate ester would seem to enhance this effect. We
98 J. L. McLAUGHLIN AND c.-J. CHANG
o o
OAe OAe
16
1711 .. ,1~8
7 9 "H q,
lC-~_O:r 6' ~c-o OR,
"3' >- I
N 26
27 28
H,C
RI R2 R
Ae 0
0,
;e-cH, -Q-
V_ ~
-~
OH
29 33 OCHl
30 Ac 34 H
ctC_CH,-Q
/ -
31 H 0
/c~
32 H 0
-~
I
Cmpd BST A·549 MCF·7 HT·29 A·498 PC·3
I I2.16 I3.71 I3.04 I2.05
PACA·2
26 2.00
I2.31
27 >30 >30 >30 24.99 >30 >30
28 >30 >30 >30 >30 15.41 >30

-4
29 191 18.9 41.8 30.77 2.18x10 27.7 30.34
30 12.0 26.1 41.3 29.47 1.59xl0

"
35.1 26.9
31
,2
2.85xlO
'\
2.30 3.83xlO
-\
<10
-, -\
1.92
1.I0xl0 2.44xl0
32 114 11.6 70.0 26.5 46.0 17.9 37.3

_2
33 281 1.12 4.29 3.00 1.08x1O 2.90 2.44
-2 -2
34 5.41xlO 30.9 7.41 1.97x1O 4.56 3.55
adriamycin -3 -\ -, -, , ,
4.16xlO 3.76xlO 3.90xlO 1.87x1O 6.37xlO 1.63xl0
hope to reisolate more of31 and test it in vivo against A-498 xenografts in athymic
mice.
6. Serenoa repens (Bart.) Small (Palmae): This is the saw palmetto of
Florida and the Carolinas the berries of which are extracted and sold popularly
as a nutritional supplement to shrink benign prostatic hypertrophy. BST directed
fractionation led to the isolation of l-monoacylglycerides, l-monolauricin (45)
and l-monomyristin (46).28 These compounds are unusual because mammalian
systems produce 2-monoacylglycerides, and, if phosphorylated at C-3 by kinases,
45 and 46 would form powerful surfactants, as are formed from phospholipids
by snake venom lipases; 45 and 46 were only moderately cytotoxic with selec-
tivity for PACA-2 (pancreas) and A-498 (lung). A use patent has been filed with
commercial interest coming from the herb trade industry. Additional bioactive
lipoid compounds are present in trace amounts.
R=
45,
o
II
R= -c~
46,
Cmpds. BST A498 PC-3 PACA-2
45 79.2 3.77 23.3 2.33
46 53.3 3.58 8.84 1.87
adriamycin
7. Monarda fistulosa L. (Lamiaceae): Wild bergamot is popular as an

herbal tea and is used in France to treat cancer; this Indiana species was active
in our bioassays, and activity-directed fractionation yielded the known com-
pounds, thymol (47) and thymoquinone (48); hydrothymoquinone (49) was pre-
pared by reduction of 48. Surprisingly, 49 was generally cytotoxic, but 48 was
selective for PC-3 (prostate) in the potency range of adriamycin. In the NCI cell
panel, 48 was active against PC~3 but was especially selective for SF-539 (CNS
carcinoma).29 None of the compounds inhibits oxygen uptake in mitochondria,
although 47, surprisingly, is twice as potent as rotenone in the YFM assay.
8. Miscellaneous: Space limitations have necessitated that we delete
descriptions of several additional, diverse, significantly cytotoxic and/or interest-
ing compounds found in Aralia spinosa (Araliaceae) from Indiana;30 Cunning-
47 48 49 50
Cmpds PACA·2
47 19.2
48 53% 15 3.48 4.30 2.92 1.82 3.86xI0·' 2.12
49 55% 2.84 4.4IxlO·' 3.30x 10" 6.18xI0·' 1.18x 10" 2.79xlO·'
50 >100 1.98 2.73 3.09 3.10 1.63 1.99
adriamycin .- 4.99xI0·' 2.53xI0·' 3.5IxlO·2 4.28xlO·' 2.76xlO·' 1.33x10·'
hamia konishi Hayata (Pinaceae) from Taiwan;3l Taiwania cryptomerioides

Hayata (Taxodiaceae) from Taiwan;32 Annona sengalensis Pers. (Annonaceae)
from Nigeria;33 Ocimum gratissimum L. (Lamiaceae) from Nigeria;34 and Popil-
/iajaponica Newman (Scarabaeidae) (Japanese beetles).35 Dihydroguaiaretic acid
(50) showed low activity in the BST assay but was generally cytotoxic and active
against Caenorhabditis elegans (LC 50 10.1 ppm), indicating that it is responsible
for the anthelmintic activity attributed to Pycnanthus angolensis (Welw.) Warb.
(Myristicaceae) from Nigeria. 36
New Annonaceous Acetogenin Compounds
I. Asimina triloba DunaJ. (Annonaceae): Asimicin (51), the first 4-hydrox-

ylated acetogenin, was reported by us as a major antitumor (in vivo active in P388
leukemia) and pesticidal component of the seeds and bark of this Indiana tree,
the common paw paw. 37 U.S. patents were awarded for the pesticidal use of the
acetogenins 38 and the composition of matter of asimicin. 39 Several firms (Dow
Elanco, AgriDyne, American Cyanamid, DuPont, S.c. Johnson) have confirmed
the pesticidal effects, and licensees of the patents are being sought. By coppic-
ing to collect the small branches and twigs, paw paw biomass can be dried and
processed to produce a potent acetogenin mixture. 40 ,41 The results of BST analy-
ses of monthly biomass collections from a single tree were astonishing and
demonstrated a dynamic flux in acetogenin composition depending on the season;
May, June, and July are optimum, and the winter months are poorest for biomass
collection by a factor of more than 25 times. 42 HPLC/MS/MS analyses identify
these months as the peak for acetogenin content. 43 Currently, we are assaying twig
BIOASSAYS AND ISOLATION OF NEW AGENTS IOl
samples collected on the same day from 670 trees growing in the same planta-
tion in Maryland; bioactivities vary as much as 900 times from one tree to the
next, showing unexpected germplasm variability.
We have now isolated 43 additional bioactive compounds from the seeds
and bark of A. triloba. These include 18 adjacent bis-THF acetogenins which can
be divided into three types: the asimicin type [asimicin (51), asimin (52), asimi-
cacin (53), asiminecin (54), asiminocin (55), asimilobin (56), parviflorin (57), cis-
and trans-asimicinones (58, 59), and lO-hydroxyasimicin (60)]; the bullatacin
type [bullatacin (61), bullatin (62), squamocin (63), motrilin (64), bullanin (65),
cis-and trans-bullatacinones (66, 67), and bullatetrocin (68)]; and the trilobacin
type [trilobacin (69), trilobin (70), asitribin (71), cis-and trans-trilobacinones (72,
73), and 10-hydroxytrilobacin (74)].44-54 Also included are 16 mono-THF com-
pounds [annonacin, annonacin A, cis-and trans-annonacin A-ones, gigantetrocin
A, cis-and trans-gigantetrocin-A-ones, cis-and trans-isoannonacins, murisolin
(75), 16, 19-cis-murisolin (76), murisolin A (77), cis-and trans-murisolinones (78,
79), and asiminenins A and B (80, 81 ).5253.55 Only one nonadjacent bis-THF ace-
togenin, trilobalicin (82), has been found. 54 The complicated absolute stereo-
chemistries of most of these acetogenins have been determined by IH NMR
analyses of Mosher esters/" sometimes aided by the formation of formaldehyde
acetal derivatives. 57 Some are extremely potent, e.g. 55, 65, and 70 are nonselec-
tively cytotoxic at ED50 values <1 0-12 Ilg/m l. The addition of the fourth hydroxyl
decreases potencies as in 60, 68, and 74.53 HPLC has permitted the difficult
separation of the 2,4-cis and trans ketolactones whose activities are usually
comparable. 4R Compound 56 has its adjacent bis-THF ring system located at C-
10 to C-17 and has only one flanking hydroxyl group at C-18. Compound 57 is
a C-35 acetogenin, and its hydroxylated bis-THF ring system starts at C-13
instead of at C-15. Some of the absolute stereochemistries are not shown here,
but they are available in the published papers and in our latest review paper. 58
~
~&a37
3 35 37
2 R, R,
0
'0
~o
0
0
Rs R3 X Y
A B C D E RI R2 R3 R4 RS R6
I I
51 threo lrans Ihreo trans {hreo X OH H H H H
52 lhrco truns {hyeo Irans [hreo X OH OH H H H
53 {hyco trum [hreo trans (hreo X H H OH H H
54 (hreo trans threo trans {hreo X H H H OH H
55 [hreo trans threo trans rhreo X H H H H OH

58,59 [hreo trans threo trans (hreo Y H H H H
60 thyeo trans [hreo trans threo X OH OH H H H
61 [hreo lrans (hreo trans erythro X OH H H H H
62 (hreo trans threo trans erythro X H OH H H H
63 threo trans (hreo trans erythro X H H OH H H
64 Ihyeo trans threo trans erythro X H H H OH H
65 threo trans threo (rans erythro X H H H H OH
66,67 [hreo trans (hreo trans erythro Y H H H H
69 Ihreo [runs erythro CIS {hreo X OH H H H H
70 lhreo lruns erYlhro CIS threo X H OH H H H
71 Ihrco trans erythro CIS (hreo X H H H OH H
72,73 thrco tran5 erythro CiS threo Y H H H H
74 thrco trans erythro cis threo X OH OH H H H
".:!"~' ".",
32~[CH2),
0"
56 57
~H
31 =
32
34
OH o
68
~In\li
~~~..~
, ~
~
3
o4-o 0 L
35 37
OH OH y
A B C D R
75 Lhreo trans threo satd. X
76 thrco cis threo satd. X
77 erylhro or threo Iran... erythro or threo saId. X
78,79 threo Irans [hreu satd. y
80 threo CIS {hreo unsatd. X
81 threo trans threo unsatd. X

..ns
erythro ./
\
OH
0
22
32
OH 0
OH OH
82
Cmpd SST A-549 MCF-7 HT-29 A-498 PC-3 PACA-2
51 2.56x10
·2
8.43x10 -. 8.52x 10 <10
·12
52 ·3 _9 -, ·12
4.6x10 7.99x10 9.57xlO <10
.J ·9 ·12 ·11
53 5.7x10 3.58x10 <10 <10
.J ·9 ·12
54 4.9x10 3.29x10 2.74x10 <10
.J.:! ·12 ·12
55 4.9x 10'3 3.lx10 2.9x10 <10
56 1.06
-, 2.14 1.47
.,
3.01x10 6.30xlO 2.26xlO I.04x10
57 -2 -12
1.72
.,
8.8x 10 <10 5.49x10
58
., ·7 ·7
2.2
I.08x10 I.lxlO 2.0xlO
59 ·7 ·7 .,
2.9xlO <10 <10 1.5 x 10
60 4.3xI0" 6.73xI0" 3.27xlO" 7.58xI0') >1 5.27xI0-' >1

.) ·11 ·12
61 1.59x 10 1.25xl0 >10 I.0xlO
-J ·6 ·6 ·6
62 4.0x10 9.39x 10 8.33x10 3.78x10
63
·9 -, ·It
9.7x10 2.94x10 1.52xl0 3.79x10
·12 ·14 -6
64 1.I0xl0 8.52x10 I.96x10 4.46xI0
.) ·14 ·12
65 6.Ox 10 3.llxl0 3.22x10 4.77xI0
66,67 3.0xlO I.0x10 5.0x10
68 3.1xlO· 1 3.52xlO- 1 4.97x10" 3.32x 10. 5 >1 >1 >1
69
.J .J .,
8.7x 10 8.02x I 0 3.39x1O <10
70 9.7x10
.J
5.71x 10 2.95x10 2.20x 10
..
71 ·2 -4 ·5
1.69 1.30
-,
2.35x10 2.25x10 1.24x 10 7.04x10 L25x10
72,73 2.0xIO· 2 4.64xI0'6 I.25xI0" 1.57 4.9IxI0-' 2.75 6.86xI0-2
74 2.6xlO· 1 I.DOxIO· s 1.88xlO· g

1.39 I.00xl0·2 3.78xI0" 1.96xlO· 1
75
., ., 3.15 -, ·9
2.36
.J
2.07x 10 5.90x10 6.58x I 0 I.09xlO 1.50xl0
76 3.46xI0-' 3.4lx10 1.58xlO' 2 1.27 4.16 1.42 I.53x10

104 I. L. McLAUGHLIN AND c.-I. CHANG
77 2.07xI0" 5.90xlO
. 3.15 6.58xlO
.,
1.09xlO
., 2.36 1.50xlO
.,
78 12.3 1.48xlO
.,
7.93xlO
., 7.54xlO
., 3.44 1.48 1.07xlO
.,
79 18.2
., ., 1.16 1.23
., .,
2.76xlO 2.90xlO 2.14xlO 5.90xlO
80 4.73x10
.,
2.85xlO
.,
2.39xlO
.,
8.12xlO
.,
6.26xlO
., 1.66 8.58xlO
.
81 5.82x10
., 3.22x10
. 3.61xlO
. 6.94xW-' 5.72xlO
., 3.66xlO
.,
6.34xlO
.7
82 5.79 5.78XIO·' 1.59XIO·7 2.28 5.97XIO·' 9.80XIO·' 2.75XlO·'
2. Asimina longifolia Kral (Annonaceae): This species is the long-leafed

dwarf paw paw tree from Georgia. Activity-directed fractionation has yielded 14
new acetogenins. These include four adjacent bis-THF acetogenins (longimicins
A-D, 83-86) which represent the asimicin type; however, the locations of the
hydroxyl flanked ring systems are shifted along the aliphatic chain toward the
lactone ring; SAR's show that the position of the bis-THF ring moiety is essen-
tial for maximization of bioactivities with the positioning at C-15 to C-24 being
optimum. 59 Nine of the other new compounds are mono-THF acetogenins:
longicin (87) and cis-and trans-goniothalamicinone (88, 89);60 longifolicin (90),
longicoricin (91), and cis-and trans-gigantetroneninone (92, 93);61 4-acetyl
annonacin (94) and 4-acetyl xylomaticin (95).62 Longanin (96) and the known
compound, giganin (97), both lack THF rings and are less active (but still active
in the range of adriamycin); the absolute stereochemistry of 97 was proven. 63
Some promising selectivities were exhibited, e.g. 90 was quite active (EDso
<10- 7 Ilg/ml) against the prostate (PC-3) cell line, and 72 was excellent (EDso
1.25 x 10-9 Ilg/ml) against the pancreatic (PACA-2) cells. We hope to test these
in vivo in athymic mice. Chlorination did not improve cytotoxicities and seems
not to be a worthwhile pursuit. 64
threo trans
trans
Cmpd y
83 OH H o 6
84 OH H o
85 OH H o 6
86 H OH 2
S
R "
~ R
~~O
x
~ 0
OH OH 0 y
R Rl R2 b A S C
87 X OH OH threo trans erythro
88,89 y OH 6 threo trans threo
90 X H OH threo trans threo
91 X H OH 6 threo trans threo
94 X OAc OH threo Irans threo
9S X OAc OH 6 threo trans threo
''''''', OH -~.~
_M, "
22 21 ~: .. 2 16
~ ~~' . .
34 (CH,l8 Ill: 11 '4 0 9 b 1 a
OH OH a
92,93 96
32 ............... (CH2)'0~(CH2h~(CH2)5
:
OH :OH m 4
I
0
..l35
0
1
97
Cmpds SST I A-549 I MCF-7 I HT-29 I A-498 I PC-3 I PACA-2

83 18.7 2.95xlO· 1 8.89xl0· 1 5.25xlO· 1 5.44xl0· 1 7.01xl0·2 1.73xl0"
84 7.34 1.43xl0" 1.54xlO" 3.32x 10.3 6.40xlO" 2.20 7.92x10· 2
8S 9.42 4.55xl0" 8.80xlO" 1.00 1.27xlO" 2.9 1.09
86 4.58 4.93x10" 2.15x10" 1.16x 10" 3.53x10" 2.42x10" 1.69xlO"
87 0.11 l.77x10·6 >1 2.4x 10.5 1.99x10" 4.26xlO·3 1.25x10"
88,89 0.14 2.06x10" 9.67x10" 4.05x10" 2.91x10" 1.37xl0" 1.33xl0· 3
90 3.52 1.13xlO·6 1.23x1O· 5 1.23 4.55x10" <10-7 4.22x1O"
91 1.56 1.04 2.31 1.36x10·3 1.71 3.04xlO" 1.36
92,93 0.11 1.0xlO" 1.98 4.81xlO" 1.12xlO" 8.35xlO" 1.99xlO"
94 21.86 3.38xlO·5 2.65xlO" 1.86xlO·5 3.59x10" 3.56xl0" 1.40xlO· 3
9S 34.12 1.25x 10.6 3.04x10" 1.12x10" 2.66x10" 3.51xlO" 6.22xlO"
96 890 1.88x10· 1 1.03 6.88x 10.3
97 12.3 4.91x10" 3.39 5.99xlO" 3.86x10" 3.96xlO" 1.11x10"
98 2.01xlO" 2.83xl0· 11 <10- 12

2.15
3. Asimina parviflora (Michx.) DunaJ. (Annonaceae): This species is the

dwarf paw paw from Georgia. The twigs have yielded two new (98, 57) aceto-
genins. 65 Parvifloracin (98) is the first nonadjacent bis-THF acetogenin with 35
carbons. Both 98 and 57 are potent and selectively cytotoxic; subsequently, we
found 57 in Annona bullata and Asimina triloba; its absolute stereochemistry was
determined. 66 Hoye and Ye,67a have synthesized 57 in a 14-step reaction sequence,
and it is the most potent (of six acetogenins and five standard pesticides tested)
against pesticide susceptible and resistant cockroaches. 67b
trans
98
4. Xylopia aromatica (Mart.) Lam. (Annonaceae): The bark of this species

from Venezuela has yielded eight new acetogenins with bizarre structural diver-
sities. Xylopianin (99), xylopiacin (100), and xylomaticin (101) are new mono-
THF compounds; 99 and 100 are unusual in having hydroxyls at C-8; 101 is like
annonacin but with 37 instead of 35 carbon. 6x Xylopien (102) and xylomatenin
(103) are similar, with hydroxyls at C-8 and C-lO, respectively, but with a double
bond at C-23/24. 69 Venezenin (104) represents an unusual C-37 acetogenin
lacking any THF rings and possessing a carbonyl and a double bond located two
methylenes away from a vicinal diol; the acetonide (105) and the dihydro (106)
derivatives both greatly enhanced the bioactivities; oxidation of the double bond
followed by cyclization yielded annomontacin-IO-one (107) and l8/21-cis-
annomontacin-IO-one (108), both of which showed enhanced bioactivities. 70
Aromin (109) and aromicin (110) represent a new type of nonadjacent bis-THF
ring acetogenin (one THF ring at C-4 to C-7 and the other at C-16 to C-19);
reducing the carbonyl of 109 to the dihydro derivative (111) enhanced the
cytotoxicities somewhat. 71 The cis-trans keto lactone mixture (112, 113) of 104
was also isolated. 72
~ 3 35 37
~
04'_ 0
o
y
BIOASSAYS AND ISOLATION OF NEW AGENTS ]07
Cmpd A B m n R
99 saId. trans X
100 saId. trans 9 X
101 saId. trans X
102 unsatd. trans 6 9 X
103 unsatd. Irans 4 9 X
107 satd. trans 6 X
108 saId. cis 6 X
111,112 saId. trans Y
trans
OH
0 0
OR, 0 threo
Cmpd RI R2 Cmpd R m
104 a H 95 0 II
105 a acetonide 96 a 13
106 H,OH H 97 H,OH H
Cmpd BST A-549 MCF-7 HT-29 A-498 PC-3 PACA-2
99 3.3xlO- 1 4.26xI0-' 18.47 6.63

-6 -7
100 0.49 1.36xl0 1.39xl0 I.Ilxlo- 1
-4 -6
101 0.11 1.54x I 0 1.47xl0 3.04xlO· 1
_3 -3
102 0.87 3.64x10 7.llxl0 7.26xI0- 3
-3 -3
103 0.77 <10 6.31x10 <10
-2 -2
104 9.33 1.08xl0 <10 1.58
-3
105 <10 3.47x10 9.32x I 0
-3 -2
106 <10 5.72xI0 2.6
107 0.84
_3 ., -,
7.86x10 7.47x10 3.46x10
108 0.98 -3 -, -2
4.x10 5.6x 10 4.7x10
·1 ·1
109 31.86 2.89x10 1.22 1.18 7.50x10 5.98x10
·1 ·1
8.99x10
110 10.28 2.17xI0

.,
4.24x10
-, 1.09
., ·1 -,
2.61x10 3.01x10 2.25x10
111 >50 4.67x10

.J
1.84x10
-2
2.30xlO
·2
2.09x10
-2
5.68xlO
-2
1.45xl0
-,
112,113 1.04xl0
·1
1.24 1.55 3.48 -, 2.22
2.15x10
adriamycin 5.20xlO
·2
5.75x10
-1
6.41x10
·2
3.42x10
-2
1.21x10
-, -2
2.55x10
5. Rollinia mucosa (Jacq.) Baill. (Annonaceae): We obtained the leaves of

this Central American fruit tree from the conservatory of the Missouri Botanical
Garden. So far, eleven novel acetogenins have been isolated and reported.
However, HPLC/(+)ESI-MS/MS detected over 40 acetogenins including a host of
new compounds in the bioactive extracts (F005).73 Sylvatacin 114 and 12,15-cis-
sylvatacin (115) are nonadjacent bis-THF acetogenins with potent selectivities
for the lung (A-549) and pancreatic (PACA-2) cell lines; their absolute stereo-
chemistries were solved by a procedure using the pattern recognition of the
NOESY spectra of their 1,4-diol formaldehyde acetal derivatives (116, 117)
which still retained the selective cytotoxicities. 74 Compounds 114 and 115 are
abundant and are prime candidates for in vivo testing; we suggest initial work
against A-549 (lung) xenografts.
Mucocin (118) from these bioactive extracts was detected by its unusual 'H
NMR shifts; it is the first acetogenin reported that possesses an hydroxylated
tetrahydropyran (THP) ring; 118 and its formaldehyde acetal (119) also showed
potent selectivities for the lung and pancreatic cell lines; 118, like all acetogenins
tested so far, inhibited oxygen uptake by rat liver mitochondria, discounting spec-
ulation that a new mode of action was introduced by the THP ring. 7s Muconin
(120), the first adjacent THF/THP acetogenin, and mucoxin (121), the first THF-
ring-hydroxyated acetogenin, expand the novelty of the compounds available in
the Rollinia genus and show some selectivities for the PACA-2 (pancreas) and
MCF-7 (breast) celilines. 76 Rollinecins A (122) and B (123) are two mono-THF
acetogenins closely related to longicin (87); compounds 122 and 123, however,
are C-37 compounds and have epimeric 14-0H's with the THF ring system at C-
17 to C-22 instead of having a 10-OH with the ring system at C-13 to C-18; they
were separated by HPLC, and their absolute stereochemistries were proven by a
modified Mosher ester method. 77 •78 Muricatetrocin C (124) and rollidecins A-D
(125-128) all bear vicinal diols whose absolute stereochemistries were proven by
analyzing their per-Mosher esters;78.79 compounds 127 and 128 were initially
detected with HPLC/MS/MS. 80 Compounds 125-128 are among the first adjacent
bis-THF acetogenins lacking a flanking hydroxyl on the lactone side, are selec-
tively ca. 1,000 times the cytotoxic potency of adriamycin, and show unusually
good activity against A-498 (renal); we hope to test 128 in athymic mice against
A-498 tumor xenografts.
34 34
34
114 P=trans 115 P=as 116 P=trans 117 P=as
35 ····37 35
CH 3(CHz)a I D CH 3(CH;z),
3. 2 1 3.
D
118 119
"
" OH 6H
37
37
120 121
trans
34
OH
122 14R 123 14S 124
R,
A n R, R2
125 threo OH OH
126 erythro 5 OH OH
127 H H
128 7 H H
1 <101 1.58xlO·1 <IO J

114 8.0xlO 3.79xlO·\ 4.52x1OSi
115 1.1 <10-8 1.15xlO·[ 5.23xl0·[ 1.41 1.80 <10"8
116 9.7xlO <10.8 1.50xlO·[ 1.62x20· 2 1.55 1.20 <10-8
117 3.4xlO
., 6. 17xlO·5 5. 18xlO·[ 1.20 2.76 2.06 <10-8
118 1.3 1.0xlO~ 1.8 9.4xlO·1 2.6 1.6xlO·1 47x10
119 1.5 4.8xlO~ 3.3 1.3 >10 8.9xlO·1 3.3x104
120 4.5xI0·3 3.6xlO·' 2.4x104 1.8,10.1 5.8xlO· 1 5.4,104
121 3.6x10·2 3.7xlO·3 6.lxlO·[ 8.4xlO·[ 3.1xlO·1 3.3xlO·[

.[
122 3.1xlO 1.I4xlO-I 1.44 1.00 7.25xlO-I 2.62x10-l 3.47,10·'
123
.[ -I 2.72 1.44 2.29xlO· 4 3.62xlO-I 2.53xlO·4
1.3 x 10 4.23xlO
.[ .;,
124 7.6xlO 5.55xlO 3.19 1.98 3.39xlO·2 1.35xlO·' 5.69xlO·'
·1
125 4.2xlO 1.04xlO 1.78 1.42 5.40x10· 1 1.65xlO·' 1.41xlO·6
126 ·1 ., 1.32 1.69 2.28xlO·' 1.73xlO·' 3.44xlO.6

2.8xlO 3.73xlO
127 1.32 1.07 6.26xlO·' 1.44 2.86xlO·1 1.08xlO·[
128 5.87 5.05 5.39 3.99 1.90 1.03
adriamycin 3.6xlO·' 1.0xW' 1. 8x 10·' 2.6xlO·' 4.2xlO·3 3.4xW'
6. Goniothalamus giganteus Hook. f. Thomas (Annonaceae): In bioactiv-

ity-directed fractionations of the bark extracts of this tree from Thailand, two
major chemical classes ofbioactive components have been discovered: the potent
cytotoxic acetogenins and the less cytotoxic styryllactones. Two more of the latter
have now been found, gonioheptolides A (129) and B (130), which bear novel
eight-membered lactone rings.Rl
129 R=Me Rl=H

OR
OR,
0
130 R=Et Rl=H
Cmpd BST A-549 MCF-7 HT-29
129 >400 4.25 72.6 13.68
130 >400 24.31 10.93 6.91

Two highly unusual acetogenins isolated from this species are giganin (96),
the first acetogenin to have neither THF nor epoxide rings,82 and goniocin (131),
the first tris-THF ring acetogenin; the cytotoxic potency of 131 approached that
of adriamycin.83 Gonionenin (132) was isolated, and the C-21122 double bonds
in 132 and in gigantetronenin were oxidized with m-chloroperbenzoic acid to give
epoxides which were then cyclized using perchloric acid to give pairs of adjacent
bis-THF [cyclogonionenin (133) and cyclogonionen C (134)] and nonadjacent
bis-THF [gigantecin (135) and C-18/21-cis-gigantecin (136)] acetogenins. The
resulting bis-THF compounds (133-136) showed enhanced bioactivities, with
135 and 118 being over a billion times more potent than adriamycin against
certain cell lines. 84 Similarly, goniodenin (137) was found with a cis-C-2l/22
double bond and was converted to a pair of tris-THF acetogenins; cyclogonio-
den ins T (138) and C (139), which showed generally enhanced cytotoxicities,
especially for the pancreatic (PACA-2) cellline. 85 Gigantransenins A-C (140-142)
are C-37 mono-THF acetogenins each with an unprecedented trans double
bond; their cytotoxic potencies were close to those of adriamycin.80 Cis-
gigantrionenin (143) and 4-acetyl gigantetrocin A (144) both have unusual
features with 144 showing good selectivity.87 The absolute configurations of
gigantetrocin A, goniothalamicin, and several other acetogenins from this and
other species were determined through formaldehyde acetal derivatives and
Mosher ester methodology.57
Recently, 4-deoxyannomontacin (145), with a mixture of cis-and trans-
annonmontacinones (146, 147),88 and 4-deoxygigantecin (148), with a mixture of
cis- and trans-gigantecinones (149, 150),89 have been isolated; compounds 145
and 148 are less cytotoxic, respectively, than annomontacin (151) and gigantecin
(152), but 145 shows good selective potencies for A-549 (lung) and MCF-7
(breast). These new compounds were equivalent or superior to rotenone in the
YFM test. Compound 145 is abundant and is proposed for in vivo testing against
A-549 (lung) and MCF-7 (breast) tumor xenografts.
c<s
37
22 ~ 21
OH OH 35
2
1 36
1°
10
o
131 132
threo
A I trans
37
threo~
I
threo
/(CH 2) , , \ OH
34 22 21 0 1817 0 14 13 10
OH OH
133 A=trans 134 A=cis 135 A=trans 136 A=cis

threo
Ihreo lrans ~ Irans 37
H 3 '36
22 21 \, . I 0
CH,(CH 2h, 18 17'01413 '0 10 4
1
OH o
137 138 A=trans 139 A=cis
trans trans
threo threo 37 threo
\ I I
35 OH
35
OH '36
23 21 I 0
23 I
CH 3 (CH 2)& r 18- 17 0-14 13 10 2 CH 3 (CH 2)g 10
22 1 1
OH OH 0 0
140 R= -OH 141 R= 11110H 142
15.-
143 144
OH
34
145 R=A; 1436, 147 R=B; 151 R=C A B C
148 R=A; 149,150 R=B; 152 R=C A B C
-Cmpd YFM
131
132 21.7
.J ., -4
1.34xl0 4.54xlO 1.12xl0
133 0.19
.J .J .,
1.41xl0 1.19xlO 1.54xl0
134 2.71
.J ., .,
1.07xl0 1.79xl0 4.08xlO
135
., ·12 -4 ·12
3.44xlO <10 1.0xlO <10
·1 ·12 ·12 ·12
136 2.49x10 <10 <10 <10
137 0.85 1.86xl0

., 8.40 4.45x 10
.J ., 1.21
.,
8.98x10 1.88xl0
138 35.3 -4 ., ., ., -4 ·8
6.37x10 7.63x 10 4.67x10 1.67x 10 9.38x10 <10
139 40.1
., 2.57 1.07
., .,
4.27x I 0 3.64x10 5.21x10 3.83x 10
140 3.6
., ·2
1.5 1.5
., .,
1.6xlO 1.0xlO 1.8xlO I.7x 10
141 5.8
., -2
1.4 1.6 -, 1.5
2.lxlO 2.1xlO 7.lx10
142 4.2
-, 2.2x 10-2 1.3 1.5 1.5 1.1
1.8xlO
143 2.52 -1 -, -6 -1 -, .,
5.99x 10 2.68x 10 6.94x10 1.39x 10 I.IIxlO 1.15xl0
144 6.78 <10

-2
8.5x10
-, <10
-2
1.55xl0
-, 1.02 <10
-2
145 I.3x 10-' 6.45x10-' 5.77xI0-' 1.41x10-' 1.50x10-' 1.73x10-' 1.00xI0-' 1.30
146,147 3.lxI0-' 2.60 3.21 2.55x 10-' 1.44 1.01 6.78xI0-' 1.00
148 4.0xlO- 1 -, 1.0 1.43x10-' 3.28x I 0-' 1.50x 10-' 3.93xlO-' 0.68
I.3lxlO
149,150 3.27 2.14xI0-' >1 >1 2.l2xlO-' 1.08xlO-) >1 0.27
151 13 I.3xxlO-' 2.3x 10-' 3.4xlO-2
152 2.22 2.19xI0-' 4.1 Ix 10-' 2.68xI0"
rotenone IT
adriamycin 4.40x I 0-) 5.94xI0-2 2.9IxI0- 2 7.20xI0) 3.70xlO- 2 2.33x I 0- 2
7. Annona muricata L. (Annonaceae): This is a popular tropical fruit tree

called guanabana or sour sop. Thousands of tons of the seeds are available as
a by-product of its juice industry. The partitioned ethanol extract (F005) of
the seeds, obtained from the Dominican Republic, was bioactive (BST LC 50
1.6 ~g/ml)90 and yielded muricatetrocins A (153) and B (154) and gigantetrocin
B (155), along with gigantetrocin A; these were all selectively cytotoxic for the
colon (HT-29) cells, and their absolute configurations were solved using partial
Mosher esters. 91 Five new diastereomeric mono-THF acetogenins, cis-annonacin
(156) and cis-annonacin-IO-one (157) (which are related to the trans-ring
annonacin series) and cis-goniothalamicin (158), arianacin (159), and javoricin
(160) (which are structural isomers of goniothalamicin) were found; 156 was
selectively cytotoxic to the colon (HT-29) cells (10,000 times the potency of
adriamycin) and is proposed for xenograft testing. 92
The leaves of guanabana, collected in Java, also showed considerable bioac-
tivity (F005, BST LC 50 0.17 ~g/ml). New acetogenins from the leaves include
annomuricins A (161) and B (162),93 muricatocins A (163) and B (164),94 anno-
muricin C (165) and muricatocin C (166) (each containing five hydroxyl
groupS),95 annomutacin (167), (2,4-trans)-lOR-annonacin A-one (168), and (2,4-
cis)-I OR-annonacin A-one (169),96 murihexocins A (170) and B (171)97 and anno-
hexocin (172)98 (each containing six hydroxyl groups and the last including an
unusual 1,3,5-triol), annopentocins A (173), B (174), and C (175) (which are
mono-THF compounds with the flanking hydroxyl on the right one carbon away
from the ring), and (2,4-cis-and trans)-annomuricin D-ones (176, 177) (which
possess an erythro vicinal diol between the mono-THF and the keto lactone
rings. 99 No bis-THF acetogenins, so far, have been found in guanabana, but some
of these mono-THF compounds show interesting cytotoxic potencies and selec-
tivities, e.g. 153 against HT-29 (colon) cells.
ri QH
A
C
-
R--
m Cmpd
~
n A B C 0
~~
Cmpd RI R2 I~HI Y
153 -OH -OH 4 axial 156 S S cis R OH
154 -OH -OH equatorial 157 S S CIS 4 ~O
155 1I110H IIiIOH 6 equatorial 158 S S cis R OH
1
R R trans S 4 OH
159
160 R R trans OH
-----
35 Y 35
33 33
OH OH~ OH OH
I 0
10 I 0
2 1 2 1
0 0
Cmpd W X Cmpd W Y
161 threo erythro 164 threo pseudo-erylhro
162 erythro eyrthro 165 erythro pseudo-erythro
163 threo threo 166 erythro pseudo-erythro

_. - - - --------~----
etylhro {run\, Ihref). 9H ~Y,.~35 etylhru truns threv
/'
./(CH2h~ 2 21
0'"
18
/"(CH2)5~1~O
17 (CH )
4 2
1
0
32
/(CH2h1~~'
?~o
20 190'16
/.,
15 11 R
~A
(CH2)S',
2 33
34 35
0
34 20H OH 11 25 0 OH OH 0
167 168 A = trans 169 A = cis
~
32 (CHZ)s
Cmpd
~
~- R RI
170 H H H OH OH OH OH
171 H OH OH H H OH OH
-------------------
172
173 H
R2
OH
J R3
H
-r--- 14-
OH
174 OH H OH H
175 H(OH) OH(H) OH(H) H(OH)
176,177
------
Cmpd PACA-2
153
154 1.8 53 4.9x10

~ 1
1.86 2.8x10
, ~
155 2.6 60 2.5x 10

~ 1
6.3x 10
·1
4.lx10
, ~
~ 1
156 2.3 28
~8
2.3xlO 118 1.0xl0

1 ·1
157 32
~ ~4
1.8 3.5x10 2.9x10 9.0xlO
1
158 47
~ ~J
5.2 I.3x 10 1.05 5.3x 10
159 7.1 26 4.7xlO 4.0xlO 4.4
160 4.9 47
., ·1
1.8
I.7x 10 2.3x 10
·1
161 0.63 3.3x10 >1.0 >1.0
1 ·1
162 0.69
~
1.59xlO >1.0 4.35xlO
163 0.70
~ , ~ 1
1.56
7.55x10 1.23xl0
1 1
164 0.56 3.34x 10
~
1.03xl0
~
1.66
1 1
165 0.61 3.08x10
~
2.28x 10
~
1.54
166 0.60
., 1.48
9.09x10 6.45x I0
167 0.039 J.57x10 >1.0 >1.0
1 1
168,169 0.11 1.74xl0
~
5.70xlO
~
>1.0
170 29 1.32 12.54 3.00 2.51 1.71xl0

~ , 9.73xl0
.\ .\
171 34 1.08 6.95 2.30 4.92 1.26x1O 4.13xlO
172 34
.\
2.26
.\
2.36
., .\
3.4xlO 7.8xlO 1.95xl0 7.7xlO
.\ .\ ·2
173 8.9 I.7lxlO 17.93 1.63 6.07x10 1.14 3.58xlO
174 11.2
., 3.56 1.64 -\ -\
2.74x10 3.79x10 2.12xlO 162.x10
175 13.8
., 2.97 1.24
.\
2.06x10 2.58x10 2.28x10 4.28x10
176,177 4.8
-, -\ ., -\
1.32 <10
-2
<10 6.llxlO <10 1.22xl0
-------- -----~----
8. Annona bullata Rich. (Annonaceae): In screening over 80 species of

Annonaceae, this Cuban species is one of the most potent (F005, BST LC 50
0.003~g/ml, PD 78% tumor inhibition). Bullatacin (61) and the (2,4-cis and
trans )-bullatacinones were the first acetogenins isolated from this species,lOo and
a patent has now been issued protecting these potent compounds. lol A total of 45
acetogenins have now been found in the bark extracts. Among three pairs of
new (2,4-cis and trans)-ketolactones, [lO-hydroxybuliatacinones (178, 179), 12-
hydroxybullatacinones (180, 181), and 29-hydroxybullatacinones (182,183)], the
mixture 180, 181 was somewhat selective for the breast cell line (MCF-7).I02 We
also reported the related series of (2,4-cis and trans )-30-,31-, and 32-hydroxy-
bullatacinones (184-189), which exhibited potent selectivities for the colon (HT-
29) cell line. lo3 Bullacin (190) is a new C-35 acetogenin with an unprecedented
C-6 hydroxyl group.66 A mixture of (2,4-cis and trans)-bulladecinones (191,192)
represent the first acetogenins having adjacent bis-THF rings at C-12 and C-16,
respectively, and having only one hydroxyl group adjacent to one side of the bis-
THF ring system. I04 C-12, 15-cis-Bullatanocin (193) and C-12, 15-cis-bullatalicin
(194) were separated by HPLC from the more abundant bullatanocin (195) and
bullatalicin (196); all of these nonadjacent bis-THF compounds are extremely
potent and selective for the lung (A-549) and pancreatic (PACA-2) cell lines; 105,106
compound 170 was active (75% TGI vs. 78% for cisplatin) against A-2780 human
ovarian xenografts in athymic mice,107 so these compounds should be given high
priority in in vivo studies against the A-549 and PACA-2 tumors. 32-Hydroxy-
bullatacin (197), 31-hydroxybullatacin (198), 30-hydroxybul\atacin (199), and
(2,4-cis and trans)-28-hydroxybullatacinones (200, 201) were also isolated; 200
and 201 were separated by HPLC; and the absolute configurations of 197-201,
as well as those of 184-189 and 191, 192, were defined by derivative
preparations and advanced Mosher ester methodology; all of these compounds
were potently cytotoxic, with 200 being surprisingly more active than 201, its
diastereomer. 108
9. Annona squamosa L. (Annonaceae): A series of 16 "annonins", reported
by a research group from Bayer AG and isolated from the seeds of this species
orythro
\ A
~
./ (CH2l.~ (CH2l '" (CH2l. ~_ 2 37
:w I 2 2
R3 OH o , 0 0
Cmpd [--R-Il~
----Ri R3 A
178 OH H H cis
179 OH H H trans
180 H OH H cis
181 H OH H trans
182 H H OH cis
183 H H OH trans
7°
/hreo
erythro
Ifans I trans ~6 37
• , 16/ ~(CH) 4R. 35
", 29 . 0
ISR 0
OR
Cmpd-- Rl R2
184,185 H H OH H
186,187 H OH H H
188,189 OH H H H
200,201 H H H OH
A
4R~37
(CH 2)5--...-""-U ~y g
o
37
34
A B A B
193 cis (12S) threo (24R) 195 trans (12R) threo (24R)
194 cis (l2S) erythro (24S) 196 trans (l2R) erythro (24S)
Cmpd
197
R
H
--r R1
OH
R2
H
R3
198 H H OH H
199 H H H OH
----~---- ----_._------
Cmpd- I BST- A-549 MCF-7 HT-29 A-498 PC-3 PACA-2
.2 -l !4 .2
178,179 7.00xlO 8.52xlO 1.86xl0 6.93xlO
180,181
., ., ., 1.51
4.43x10 1.30xlO 2.15x10
182,183 7.53xlO
.,
2.40x10
.,
1.35xl0
. 1.57xl0
.,
184,185
., ., ., ·12
4.05x10 1.66xl0 1.05xlO <10
186,187 1.58xl0
.,
3.29x10
.11
7.63xlO
. 1.09xlO
·12
188,189
., ., ., <10
·12
4.67x10 1.25xl0 1.61xl0
190 7.0xlO
., 1.79xl0
.,
1.00xlO
.,
5.23xlO
.,
191,192
., ., ., .,
1.37xlO 3.37xlO 1.07xl0 2.29x10
193 5.60x10
.,
1.51xlO
.,
8.53x10
.,
5.81xl0
.,
8.23xlO
.,
6.81xlO
. 1.25xl0
·IS
194
., ·11
>1.0 >1.0
., ., ·14
8.19x10 1.88xl0 3.48x10 2.15xl0 2.81x10
195
., ·11 ., ., ., 4.39xlO
., 8.90xl0
·IS
4.33x10 2.15xlO 2.lOxl0 3.67xlO 6.39xlO
196 1.24 ·13 ., >1.0

., ., ·IS
1.16x10 2.79x10 2.59xlO 1.99xlO 6.09xlO
197
., ., ., 1.48
., 1.62xlO-2 <10
.,
8.00x10 <10 <10 <10
198
., -8 ., 1.21
., ., <10.,
5.72xlO <10 <10 <10 3.16xlO
199
., ., ., 1.17
., ., <10.,
6.55x10 <10 <10 <10 4.33xlO
200 7.98x10
.,
4.38x10
. 4.46xl0
.,
6.90xl0
.. 4.35 2.33 1.02xl0
.,
201
., ., 1.71 1.45
., 9.04
.,
7.98xlO 3.18x10 1.00xlO 1.19xlO
-.- ... ~~-------------- -- ------~~---
from India, contained some errors in the reported structures; I09 we reassigned the
relative stereochemistry of bullatalicin (196) and the bullatalicinones and identi-
fied the erroneous structures as bullatalicin, bullatanocin, and squamostatin A. 110
The bark has been an excellent source of bullatacin (61) for biological evalua-
tions; while purifying some bullatacin by HPLC, a shoulder peak was isolated
and identified as squamotacin (202), a bullatacin analogue in which the hydrox-
ylated bis-THF ring system starts at C-13 instead of C-15; molvizarin (203), a
C-35 analog of bullatacin (61), was also isolated; 202 and 203 showed remark-
able cytotoxic selectivity for the prostate (PC-3) cell line, with 202 showing a
potency more than 100 million times that of adriamycin; data for bullatacin (61),
which is more generally cytotoxic, are given below for comparative purposes. III
A patent has been filed, and 202 should be a major target compound for
synthesis.
Four new mono-THF acetogenins, cis- and trans-mosinones A (203, 204),
mosin B (205), and mosin C (206), along with annoreticuin (207), all possess a
carbonyl group at C-9; all of these compounds show selectivity against the pan-
creatic cell line (PACA-2), with potencies 10 to 100 times that of adriamycin.1I2
A patent has been filed, and 206 is proposed for in vivo tests against PACA-2
xenografts.
CH,
OH OH
m n
202
203
51
Cmpd ~--A-5~ T MCF-7 HT-29 A-498 PC-3 PACA-2
202 6.80x I O· 2. 77x 10·' >I

.J
>1
., 4
1.00xl0 l.72xlO 1.33xl0
203
., ., >1
.; ., ·8 .;
5.26x10 6.30x10 7.32x10 7.09xlO 4.47x10 7.66x10
51
.; ·6 ., >1
., ·9 ·9
1.59xl0 2.44x10 5.9xlO 4.85xlO <10 <10
adriamycin 2.57x1O 2.84xI0·2 3.47xlO

.,
4.156xlO
.,
2.49x 10
., 3.42xlO
.,
3.17xlO
.,
-----------_. --
0,·35
120 I. L. McLAUGHLIN AND c.-I. CHANG
R3=~ B
~
R,-
~(E)
\ 20 0
I 15 2
9
2
0
OH OH 0 0
R.=~ d~37
R2 =
"0 0
Cmpd. A B C D E N C-1S/C-20
203,204 threo trans threo R, R. RIR
205 lhreo/erythro trans threolerythro R1 R, RlS or SIR
206 threo cis threo R1 R, RlS
207 threo trans threo R1 R, RlR
203,204 4.39xI0·' >1 >1 >1 >1 3.19xlO·2 2.18xlO·'
205 2.93xlO· 1 9.44xI0· 1 >1 >1 >1 >3.50xlO- 1 2.5IxI0"
206 1.54xlO- 1 5.96xlO- 1 >1 >1 >1 >1 1.1 7x 10"
207 4.09xlO-' 2.73xlO- 1 >1 >1 >1 9.64xI0·' 2.39xlO"
adriamycin 2.57xlO- 1 5.27x 10- 3 1.99x 10- 1 2.00xI0-' 1.02xlO-' 3.2IxI0-' 1.79xlO-'
Biological Studies with the Annonaceous Acetogenins

Londershausen et at., 113 at Bayer AG, initially observed that acetogenin-
treated insects had substantially lower total levels of ATP, similar to the effect of
antimycin A, a known inhibitor of the mitochondrial electron transport system
(ETS). Mitochondrial enzymes were tested, and squamocin (63) (annonin I) was
2.5-5 times as potent as rotenone in inhibiting NADH:ubiquinone oxidoreduc-
tase. Concurrently, our collaborators at the University of Ottawa 114 observed a
lower level of oxygen consumption in acetogenin-fed insects and experimentally
located the site of action of asimicin (51) (and F005 from paw paw) within mito-
chondrial complex 1. Our collaborators at Michigan State University lO7 were
working, concurrently, with bullatacin (61) (the 24-epimer of asimicin) in SF9
insect cells, as well as in insect and mammalian mitochondria, and arrived at the
same conclusion. Tests with our bullatacin (61) in a whole host of mechanism-
based antitumor assays at several firms (Merck, Glaxo, Bristol-Meyers Squibb,
and Abbott) failed to identify any other mechanism of antitumor action.
Subsequently, Friedrich et at. 115 and Espositi et al. 116 confirmed that the
acetogenins bind competitively with respect to the ubiquinone site at complex I,
whereas, rotenone binds non-competitively suggesting an alternative site.
Hollingworth et at. 117 concluded that bullatacin (61) is the most potent among the
several chemically diverse types of complex I ETS inhibitors known. In our lab-
oratory, Landolt et al. ll8 showed several structure-activity relationships (SARs)
among 20 structurally diverse acetogenins in the inhibition of oxygen uptake by
rat liver mitochondria, and some are more potent than rotenone in this subcellu-
lar system. Mitochondrial testing of 14 additional acetogenins found trilobacin
(69) and asiminacin (54) to be more potent than bullatacin (61).119 Similar SAR
conclusions were obvious when we tested 44 acetogenins in the yellow fever mos-
quito larvae (YFM) assay, with bullatacin (61) and trilobacin (69) giving the best
potencies (LC so values 0.10 and 0.67ppm, respectively; rotenone gave 1.2ppm);
compounds with YFM LC so values of <1.0ppm are considered as new lead
candidates for pesticide development. 12o Summaries of this mechanism and
these SARs have been published in our recent reviews on the Annonaceous
acetogenins. 41.58, 121
A second mode of action helps to explain the selectivity of the acetogenins
in inhibiting tumor vs. normal cells; an ubiquinone-linked NADH oxidase is acti-
vated in the plasma membranes of tumor cells, and this enzymatic activity is
potently inhibited by bullatacin (61) and other acetogenins. 122 The net effect of
this action and the action on the ETS results in intracellular ATP depletion, and
apoptosis (programmed cell death) is a consequence of the build up of radical
oxygen.123 Indeed, we have recently demonstrated that bullatacin (61) induces
apoptosis in WEHI 231 B cells, a murine B cell line that is at the immature stage
of B cell maturation (Geahlen and McLaughlin, unpublished).
The cell inhibition activities of several Annonaceous acetogenins, covering
the three major structural classes ofbis-adjacent, bis-nonadjacent, and single THF
ring compounds and their respective ketolactone rearrangement products, were
tested blind at Wayne State University in an in vitro disc diffusion assay against
three murine (P388, P03, and M17/Adr) and two human (H8 and HI2S) cancer-
ous cell lines, as well as a non-cancerous rat GI epithelial cell line (l18). The
results demonstrated a dose-dependent inhibition of cancerous cell growth, while
non-cancerous cell growth was not inhibited by the same dosages, All of the ace-
togenins, irrespective of their various structural types, inhibited the growth of
adriamycin resistant tumor cells and non-resistant tumor cells at the same levels
of potency. These results showed that the Annonaceous acetogenins are an
extremely potent class of compounds and their inhibition of cell growth can be
effective for drug resistant cancer cells, while exhibiting only minimal toxicity to
normal non-cancerous cells,124
Resistance mechanisms which require an ATP-dependent transporter would
be obvious targets for the useful application of these compounds as antitumor
agents, antimalarials, pesticides, or in any of the several systems in which such
ATP-dependent efflux is a significant factor, An important earlier observation
from the Ncr cell panel was that adriamycin resistant cell lines are almost always
more susceptible to the acetogenins than the respective wild type cells, e,g. MCF-
710gIO GIso -5.86 molar and MCF-7/Adr 10gIO GI so -6.49 molar for bullatacin (61);
the acetogenins also routinely give more impressive selectivity and potency data
in the panel than almost all of the clinically used antitumor drugs. In the
COMPARE program,125 against the NCI bank of 175 "standard agents" for bul-
latacin (61) at all three levels (GI so , TGI so, and LC so ), there was nothing with a
correlation coefficient above 0.7, indicating no close mechanistic relatives. Con-
cerning bullatacin (61) vs. all of the compounds tested to that time, most of the
close matches were with related acetogenins (which we had provided earlier). The
NCI concluded that, " ... the mechanism of action has definite uniqueness."
We decided to confirm, in our own laboratory, the effects of the acetogenins
against multi-drug resistant (MDR) human tumor cells in vitro. Bullatacin (61)
was effectively cytotoxic to MCF-7/Adr (MDR) breast cells, while it was more
cytostatic to the parental MCF-7/wt (wild type) cel1. 126 This assay system was
extended to include 14 of our structurally diverse acetogenins; bullatacin-indexes
illustrated the MDR phenomenon (bullatacin 1.0; adriamycin 258.5; vincristine
141.3; vinblastine 18.8) and suggested several SARs: I) the bis-adjacent THF
acetogenins with three hydroxyls and the bullatacin-type stereochemistry (threo-
trans-threo-trans-erythro) from C-15 to C-24 are generally potent; 2) a spacing
of 13 carbon atoms between the I5-hydroxyl and the y-Iactone seems to be
optimum; and 3) some of the mono-THF compounds, e.g. gigantetrocin A are
potent. 127 Thus, we believe that this is good rational evidence to suggest in vivo
studies with athymic mice bearing xenografts of MDR human tumors; we have
applied for a patent protecting the use of acetogenins in treating MDR tumors.
A logical extension of this work was into the pesticide field. At Purdue's
Entomology Department, we compared six Annonaceous acetogenins (two
each from the mono-THF, nonadjacent bis-THF, and adjacent bis-THF classes)
with five commercial insecticides representing the amidohydrazone, carbamate,
organophosphate, and pyrethroid classes. Dietary toxicities were determined to
insecticide-resistant and insecticide-susceptible strains of Blattella germaniae
(the German cockroach), 2nd and 5th instars. The speed of kill (LTso) values per-
mitted ranking of all 11 compounds, and parviftorin (57) was ranked second with
cypermethrin (a synthetic pyrethroid) having a slight edge. Resistance ratios were
calculated and were, as predicted, lower for the acetogenins than for the synthetic
compounds. Only hydramethylnon (an inhibitor of ATP production at ETS
complex III) gave comparable resistance ratios. Some of the acetogenins were
even more effective against the resistant vs. the nonresistant roaches. Thus, the
acetogenins show great promise in treating insecticide-resistant insects in a new
type of baited "Roach Motel".67b
A glance at the cytotoxicity data tabulated above for our new acetogenins
reveals some remarkable potencies and selectivities for certain tumor cell types.
For example, squamotacin (202), the ring shifted bis-THF analog of bullatacin
(61), is selective for the prostate (PC-3) cell line, with a potency of over 100
million times that of adriamycin; the nonadjacent bis-THF bullatanocinlbullatal-
icin series (193-196) and the sylvaticin series (114-117) are extremely potent
(some LCso's at <IO-8Ilg/ml) for the pancreatic (PACA-2) cell line; the unique
tris-THF, cyclogoniodenin T (138), and the 9-carbonyl mono-THF compounds
(203-207) also showed selectivity for PACA-2; the mono-THF, longicin (87), was
surprisingly potent to PACA-2 and the lung (A-549) cell lines; and the asimicin
series (51-55) is extremely potent (LCso's < IO-12Ilg/ml) against the colon (HT-
29) cell line. We predict that specific acetogenins may well become useful in the
therapy of specific tumor types as well as their MDR descendants, and patents
for some of these specific applications have been submitted.
Limited in vivo antitumor tests of the Annonaceous acetogenins have given
additional encouraging results. Uvaricin originally showed in vivo activity against
3PS (P388) (murine lymphocytic leukemia) (157% TIC at I Amg/kg), and rolli-
nones (147% T/C at I Amg/kg) and asimicin (51) (124% T/C at 25.0 Ilg/kg) were
similarly active, with 51 showing fifty times the potency of the other twO. 128 We
reported the activity ofbullatacin (61) and (2,4-cis and trans)-bullatacinones (66,
67) against U2l0 (murine leukemia) in normal mice and in inhibiting tumor
xenografts of A2780 (human ovarian carcinoma) in athymic mice. 107 Bullatacin
(61), effective at only 50.0Ilg/kg, was over 300 times more potent than taxol
against U2l0, and 61, again at 0.05mg/kg, and bullatalicin (196), at I mg/kg
were nearly as effective as cis-platinum, while causing much less weight loss.
Previously unpublished results, obtained by the late G. Grindey at the Eli Lilly
Co., showed similar effectiveness of61 (67% tumor growth inhibition at 50 Ilg/kg)
against X-5563 plasma cell myeloma grafts in normal mice. A favorable ratio of
cytotoxic EDso values vs. LDso values in mice suggested that bullatacin (61)
should be less toxic in vivo than currently used drugs such as cis-platinum. 129 The
difficulties involved in securing additional in vivo studies, at either the NCI
or within the pharmaceutical industry, have been the major road block in the
development of these compounds as new, clinically useful, antitumor agents.
Ames test results (Sitek Research Laboratories, unpublished results) on
F005 from paw paw were negative in 9 out of 10 tests and only slightly positive
(2.5% above background reversions) on one histidine mutant of Salmonella
typhimurium after enzyme activation of the extract. This negative result demon-
strates that the acetogenins are not mutagenic, and these results were to be
expected because the acetogenins, unlike most other antitumor agents, do not
exert their primary effects by interfering with DNA; they inhibit ATP production.
In other unpublished results (Asta Laboratories), bullatacin (61) was 3PS
active (131 % TIC at 50llg/kg) in mice and emetic at 1851lg/kg in pigs; this latter
result demonstrated that the acetogenins likely explain the former use of Eli Lilly's
fluid extract of paw paw seeds as an emetic preparation. 130 Thus, emesis is a def-
inite safety factor should someone ingest excessive amounts of these materials
either intentionally or unintentionally. Antiemetic drugs, as are administered with
cis-platinum, may likely be needed for antitumor treatment with the acetogenins,
and/or emesis can be used as a convenient signal that overdosage has occurred.
Emesis is actually an advantage concerning the potential pesticidal applications.

We have now published four comprehesive compilations reviewing the chemistry
and biology of the Annonaceous acetogenins,58.12I.128.131 and our fifth review is
being prepared. I32
SUMMARY
Four simple (bench-top) bioassays are serving us well for the detection and
fractionation monitoring of new plant antitumor and pesticidal agents. These are:
(1) lethality to the larvae of brine shrimp (Artemia salina); (2) the inhibition of
crown gall tumors, induced by plasmid transfer and expression from Agrobac-
terium tumefaciens, on discs of potato (Solanum tuberosum) tubers; (3) the inhi-
bition or stimulation of frond proliferation of duckweed (Lemna minor); and (4)
lethality to the larvae of yellow fever mosquitoes (Aedes aegyptii). Since 1984,
over 320 chemically diverse bioactive plant components have been isolated and
characterized in our laboratory by using these methods. Recently, bioactive com-
pounds from the Meliaceae, Lauraceae, Euphorbiaceae, Laminaceae, and other
plant families have been isolated, but our most exciting leads have been with
the potent acetogenins from the Annonaceae; these compounds are powerful
inhibitors of mitochondrial electron transport systems and of the NADH oxidase
that is prevalent in the plasma membranes of tumorous cells. The consequence is
ATP depletion, and this is especially toxic to multiple drug resistant tumor cells
and pesticide resistant insects that possess ATP-dependent xenobiotic efflux
systems. Structural activity relationship studies (in mitochondrial preparations
and against mosquito larvae) help to define the optimum structural features. This
paper has presented the chemistry and biological testing results of 207 plant
components recently isolated using the simple bioassays described folIowed by
cytotoxicity testing in a panel of six human tumor cell lines.
REFERENCES
I. American Cancer Society, "Cancer Facts and Figures-1997", American Cancer Society,
Atlanta, 1997, pp. 1-7.
2. CHANG, c.-1., ASHEN DEL, c.L., CHAN, T., GEAHLEN, R.L., MCLAUGHLIN, 1.L.
1991. Mechanism-based discovery of natural anticancer agents. In: Advances in New
Drug Development. B.-K. Kim et al. (eds.). Pharmaceutical Society of Korea, Seoul,
pp.448-457.
3. CHANG, c.-1., ASHENDEL, c.L., GEAHLEN, R.L., HUNG, M.-C., MCLAUGHLIN,
1.L., WATERS, 0.L. 1997. Oncogene-moldu1ated signal transduction inhibitors from
plants, In: Phytochemical Diversity: A Source of New Industrial Products, S. Wrigley,
M. Hayes, R. Thomas, and E. Chrystal (eds.), The Royal Society of Chemistry, Infor-
mation Series, London, pp. \70-\78.
4. MCLAUGHLIN, 1.L. 1991. Crown gall tumours in potato discs and brine shrimp lethal-
ity: Two simple bioassays for higher plant screening and fractionation. In: Methods in
Plant Biochemistry, Vol. 6, K. Hostettmann (ed.), Academic Press, London, pp. 1-31.
5. MCLAUGHLIN, lL., CHANG, C.1., SMITH, D.L. 1991. Bench-top bioassays for the
discovery of bioactive natural products: An update. In: Studies in Natural Products
Chemistry, Vol. 9, Atta-Ur-Rahman (ed.), Elsevier, Amsterdam, pp. 383--409.
6. MCLAUGHLIN, lL., CHANG, C.1., SMITH, D.L. Simple bench-top bioassays (brine
shrimp and potato discs) for the discovery of plant antitumor compounds: Review of
recent progress. In: Human Medicinal Agents from Plants, A.D. Kinghorn and M.F.
Balandrin (eds.), ACS Symposium Series 534, American Chemical Society, Washington,
D.C., 1993, pp. 112-137.
7. MEYER, B.N., FERRIGNI, N.R., PUTNAM, lE., JACOBSEN, L.B., NICHOLS, D.E.,
McLAUGHLIN, 1.L. 1982. Brine shrimp: A convenient general bioassay for active plant
constituents. Planta Medica 45:31-34.
8. FERRIGNI, N.R., PUTNAM, 1.E., ANDERSON, B., JACOBSEN, L.B., NICHOLS, D.E.,
MOORE, D.S., McLAUGHLIN, 1.L. 1982. Modification and evaluation of the potato disc
assay and antitumor screening of Euphorbiaceae seeds. 1. Nat. Prod. 45:679-686.
9. ANDERSON, 1.E., GOETZ, C.M., MCLAUGHLIN, 1.L., SUFFNESS, M. 1991. A blind
comparison of simple bench-top bioassays and human tumor cell cytotoxicities as anti-
tumor prescreens. Phytochem. Anal. 2: 107-111.
10. MCLAUGHLIN, 1.L., ROGERS, L.L., ANDERSON,lE. 1998. The use of biological
assays to evaluate botanicals. Drug Information Journal 32:513-524.
II. MCLAUGHLIN, 1.L., COLMAN-SAIZARBITORIA, T., ANDERSON, lE. 1995.
Tres bioensayos simples para quimicos de productos naturales. Revista de la Sociedad
Venezolana de Quimica, 18(4), 13-18.
12. MCLAUGHLIN, 1.L. GU, Z.M. 1997. Two simple bioassay methods to detect antitumor
compounds. China Materia Medica. 22:617-619.
13. MONKS, A., SCUDIERO, D., SKELHAN, P., SHOEMAKER, R., PAULL, K., VISTICA,
D., HOSE, c., LANGLEY, 1., CRONICE, P., VAIGO-WOLFF, A., GRAY-GOODRICH,
M., CAMBELL, H., MAYO, 1., BOYD, M. 1991. Feasibility of a high-flux anticancer
drug screen using a diverse panel of cultured human tumor cell lines. 1 Nat. Can. Inst.
83:757-766.
14. GERAN, R.I., GREENBURG, N.H., MACDONALD, N.N., SCHUMACHER, A.A.,
ABBOTT, B.1. 1972. Protocols for screening chemical agents and natural products against
animal tumors and other biological systems. Cancer Chemotherapy Reports. 3: part 3,
1-103.
15. FERRIGNI, N.R., MCLAUGHLIN, J.L. POWELL, R.G., SMITH, JR., C.R. 1984. Use
of potato disc and brine shrimp bioassays to detect activity and isolate piceatannol as the
antileukemic principle from the seeds of Euphorbia lagascae. 1. Nat. Prod. 47:347-352.
16. GEAHLEN, R.L. AND MCLAUGHLIN, lL. 1989. Piccatannol (3,4,3i5i-tetrahydroxy-
trans-stilbene) as a natural occurring protein-yyrosine kinase inhibitor. Biochcm.
Biophys. Res. Com. 165:241-245.
17. OLIVER, J.M., BURG, D.L., WILSON, B.S., McLAUGHLIN, 1.L., GEAHLEN, R.L.
1994. Inhibition of mast cell FceR I-mediated signaling and effector function by the
Syk-selective inhibitor, piceatannol. 1. BioI. Chern. 269:29697-29703.
18. ZENG, L., GU, Z.-M .• FANG, X.-P., McLAUGHLIN, 1.L. 1994. Kneglomeratanol, kne-
glomeratanones A and B, and related bioactive compounds from Knema glomerata. 1.
Nat. Prod. 57:376-381.
19. ZENG, L., GU, Z.-M., FANWICK, P.E., CHANG, C.1., SMITH, D.L., McLAUGHLIN,
lL. 1995. Additional bioactive triterpenoids from Melia volkensii (Meliaciae). Hetero-
cycles. 41:741-752.
20. ZENG, L., GU, Z.-M., FANWICK, P.E., CHANG, C.J., SMITH, D.L., McLAUGHLIN,
J.L. 1995. Two new bioactive triterpenoids from Melia volkensii (Meliaceae). Tetrahe-
dron. 51 :2477-2488.
21. ZENG, L., GU, Z.-M., CHANG, C.J., WOOD, K.Y, McLAUGHLIN, J.L. 1995.
Meliavolkenin, A new bioactive triterpenoid from Melia volkensii (Meliaceae). Bioorg.
Med. Chern. 3:383-390.
22. ZENG, L., GU, Z.-M., CHANG, C.J., SMITH, D.L., MCLAUGHLIN, J.L. 1995. A pair
of new apotirucallane triterpenes, meliavolkensins A and B, from Melia volkensii
(Meliaceiae). Bioorg. Med. Chern. Lett. 5:181-184.
23. YE, Q., GU, Z.-M., ZENG, L., CHANG, C.J., MCLAUGHLIN, J.L., SASTRODI-
HARDJO, S. 1996. Persealide: A novel, biologically active component from the bark of
Persea americana (Lauraceae). IntI. J. Pharmacog. 34:70-72.
24. OBERLIES, N.H., ROGERS, L.L., MARTIN, J.M., McLAUGHLIN, J.L. 1998. Cyto-
toxic and insecticidal constituents of the unripe fruit of Persea americana. J. Nat. Prod.
61:781-785.
25. FATOPE, M.O., ZENG, L., OHAYAGA, J.E., MCLAUGHLIN, J.L. 1996. Stereostruc-
tures and bioactivities of new 19-acetoxyingol diterpenes from the latex of Euphorbia
poisonii (Euphorbiaceae). Bioorg. Med. Chern. 4:1679-1683.
26. FATOPE, M.O., ZENG, L., OHAYAGA, J.E., SHI, G., McLAUGHLIN, J.L. 1996. Selec-
tively cytotoxic diterpenes from Euphorbia poisonii. J. Med. Chern. 39: 1005-1008.
27. SZALLASI, A. 1993. The vanilloid (capsaicin) receptor: Receptor types and species dif-
ferences. Gen. Pharm. 25:223-243.
28. SHIMADA, H., TYLER, YE., McLAUGHLIN, J.L. 1997. Biologically active acylglyc-
erides from the berries of saw-palmetto (Serenona repens, Palmae). J. Nat. Prod.
60:417-418.
29. JOHNSON, H.A., ROGERS, L.L., ALKIRE, M.L., MCLAUGHLIN, J.L. MCCLOUD,
T.G. 1998. Bioactive monoterpenes from Monardafistulosa (Lamiaceae). Nat. Prod. Lett.
11:241-250.
30. RATNAYAKE, S., MCLAUGHLIN, J.L. 1993. Petroselinic acid from the ripe berries of
Aralia spinosa. IntI. J. Pharmacog. 31 :35-37.
31. HE, K., SHI, G., ZENG, L., YE, Q., McLAUGHLIN, J.L. 1997. Konishiol, a novel
sesquiterpene, and bioactive components from Cunninghamia konishii. Planta Medica.
63: 158-160.
32. HE, K., ZENG, L., SHI, G., ZHAO, G.-X., McLAUGHLIN, J.L. 1997. Bioactive com-
pounds from Taiwania cryptomerioides. J. Nat. Prod. 60:38-40.
33. FATOPE, M.O., ANDU, O.T., TAKEDA, Y, ZENG, L., SHI, G., SHIMADA, H.,
McLAUGHLIN, J.L. 1996. Bioactive ent-kaurene diterpenoids from Annona senegalen-
sis. J. Nat. Prod. 59:301-303.
34. NJOKU, C.J., ZENG, L., ASUZU, LU., OBERLIES, N.H., McLAUGHLIN, J.L. 1997.
Oleanolic acid, a bioactive component of the leaves of Ocimum gratissimum (Lamiaceae).
IntI. J. Pharmacog. 35:134-137.
35. ROGERS, L.L., McLAUGHLIN, J.L. 1998. Cholesterol from Japanese beetles, Popillia
japonica. Nat. Prod. Lett. 11:237-240.
36. NJOKU, C.1., HOPp, D.C., ALALI, E, ASUZU, u., McLAUGHLIN, J.L. 1997. Dihy-
droguaiaretic acid: A bioactive component of the stem bark of Pycnanthus angolensis
(Myristicaceae). Planta Medica. 63:580--581.
37. RUPPRECHT, J.K., CHANG, C.1., CASSADY, J.M., MCLAUGHLIN, J.L., MIKOLA-
JCZAK, K.L., WEISLEDER, D. 1986. Asimicin, a new cytotoxic and pesticidal
acetogenin from the paw paw, Asimina triloba (Annonaceae). Heterocycles, 24:
1197-1201.
38. MIKOLAJCZAK, K.L., MCLAUGHLIN, J.L., RUPPRECHT, J.K. "Control of Pests with
Annonaceous Acetogenins," (pesticidal use patent on acetogenins) U.S. Patent No.

4,721,727, issued January 26,1988.
39. MIKOLAJCZAK, K.L., MCLAUGHLIN, 1.L., RUPPRECHT, 1.K. "Control of Pests with
Annonaceous Acetogenins," (divisional patent on asimicin) u.s. Patent No. 4,855,319
issued August 8, 1989.
40. RATNAYAKE, S., RUPPRECHT, J.K., POTTER, W.M., MCLAUGHLIN, 1.L. 1993.
Evaluation of various parts of the pawpaw tree, Asimina trilaba (Annonaceae), as
commercial sources of the pesticidal Annonaceous acetogenins. 1. Econ. Entom.
85:2353-2356.
41. MCLAUGHLIN, 1.L., ZENG, L., OBERLIES, N.H., ALFONSO, D., JOHNSON, H.A.,
CUMMINGS, B.A. 1997. Annonaceous acetogenins as new natural pesticides: Recent
progress. In: Phytochemical Pest Control Agents, P. Hedin, R. Hollingworth, 1. Mujamoto,
E. Mesler, and D. Thompson (eds.), ACS Symposium book, Washington, D.C., pp.
117-130.
42. JOHNSON, H.A., MCLAUGHLIN, 1.L., GORDON, 1. 1996. Monthly variations in bio-
logical activity of Asimina trilaba (The North American paw paw tree). In: Progress in
New Crops: The Third National Symposium, 1. Janick (ed.), ASHS, Alexandria, pp.
609-614.
43. GU, Z.-M., JOHNSON, H.A., ZHOU, D., WU, 1., GORDON, 1., MCLAUGHLIN, 1.L.
1998. Quantitative evaluation of Annonaceous acetogenins in monthly samples of
paw paw (Asimina trilaba) twigs by liquid chromatography/tandem mass spectrometry.
Phytochem. Anal. 10:72-38.
44. ZHAO, G.X., HUI, Y-H., RUPPRECHT, 1.K., MCLAUGHLIN, 1.L., WOOD, K.V 1992.
Additional bioactive compounds and trilobacin, a novel highly cytotoxic acetogenin, from
the bark of Asimina trilaba. 1. Nat. Prod. 55:347-356.
45. ZHAO, G.X., NG, 1.H., KOZLOWSKI, 1.E, SMITH, D.L., McLAUGHLIN, 1.L. 1994.
Bullatin and bullanin: Two novel, highly cytotoxic acetogenins from Asimina tri/aba.
Heterocycles. 38: 1897-1908.
46. ZHAO, G.X., MIESBAUER, L.R., SMITH, D.L., McLAUGHLIN, 1.L. 1994. Asimin,
asiminacin, and asiminecin: Novel highly cytotoxic asimicin isomers from Asimina
tri/aba. 1. Med. Chern. 37:1971-1976.
47. ZHAO, G.X., GU, Z.-M., ZENG, L., CHAO, J.E, WOOD, K.V, KOZLOWSKI, 1.E,
McLAUGHLIN, 1.L. 1995. 1995. The absolute configuration of trilobacin and trilobin, a
novel highly potent acetogenin from the stem bark of Asimina tri/aba (Annonaceae).
Tetrahedron. 51:7149-7160.
48. ZHAO, G.X., CHAO, 1.E, ZENG, L., McLAUGHLIN, 1.L. 1996. (2,4-cis)-Asimicinone
and (2,4-trans)-asimicinone: Two novel ketolactone acetogenins from Asimina tri/aba
(Annonaceae). Nat. Toxins. 4:128-134.
49. ZHAO, G.X., CHAO, 1.E, ZENG, L., RIESER, M.J., McLAUGHLIN, 1.L. 1996. The
absolute configuration of adjacent bis-THF acetogenins and asiminocin, a novel highly
potent asimicin isomer from Asimina tri/aba. Bioorg. Med. Chern. 4:25-32.
50. WOO, M.H., CHO, K.Y, ZHANG, Y, ZENG, L., GU, Z.-M., McLAUGHLIN, 1.L. 1995.
Asimilobin and cis- and trans-murisolinones, novel bioactive Annonaceous acetogcnins
from the seeds of Asimina tri/aba. 1. Nat. Prod. 58:1533-1542.
51. WOO, M.H., ZENG, L., MCLAUGHLIN, 1.L. 1995. Asitribin and asimenins A and B,
novel bioactive Annonaceous acetogenins from the seeds of Asimina tri/aba. Heterocy-
cles.41:1731-1742.
52. WOO, M.H., ZENG, L., YE, Q., GU, Z.-M., ZHAO, G.-X., McLAUGHLIN, 1.L. 1995.
16,19-cis-Murisolin and murisolin A, two novel bioactive mono-tetrahydrofuran
Annonaceous acetogenins from Asimina tri/aba seeds. Bioorg. Med. Chern. Lett.
5:1135-1140.
128 1. L. McLAUGHLIN AND C.-1. CHANG
53. HE, K., SHI, G., ZHAO, G.-X., ZENG, L., YE, Q., SCHWEDLER, J.T., WOOD, K.Y.,
McLAUGHLIN, 1.L. 1996. Three new adjacent bis-tetrahydrofuran acetogenins with four
hydroxyl groups from Asimina triloba. 1. Nat. Prod. 59:1029-1034.
54. HE, K., ZHAO, G.-X., SHI, G., ZENG, L., CHAO, 1.-E, McLAUGHLIN, 1.L. 1997. Addi-
tional bioactive Annonaceous acetogenins from Asimina triloba (Annonaceae). Bioorg.
Med. Chern. 5:501-506.
55. ZHAO, G.-X., RIESER, M.1., HUI, Y-H., MIESBAUER, L.R., SMITH, D.L.,
McLAUGHLIN,1.L. 1993. Biologically active acetogenins from the stem bark of Asimina
triloba. Phytochemistry. 33: 1065-1 073.
56. RIESER, M.1., HUI, Y-H., RUPPRECHT, lK., KOZLOWSKI, lE, WOOD, K.Y.,
McLAUGHLIN, lL., HOYE, T.R., HANSON, P.R., ZHUANG, Z.-P. 1992. Determina-
tion of absolute configurations of stereogenic carbinol centers in Annonaceous aceto-
genins by 1H- and 19F-NMR analysis of Mosher ester derivatives. 1. Am. Chern. Soc.
114: 10203-10213.
57. GU, Z.-M., ZENG, L., FANG, X.-P., COLMAN-SAIZARBITORIA, T., HUO, M.,
MCLAUGHLIN, 1.L. 1994. Determining absolute configurations of stereocenters in
Annonaceous acetogenins through formaldehyde acetal derivatives and Mosher ester
methodology. 1. Org. Chern. 59:5162-5172.
58. ZENG, L., YE, Q., OBERLIES, N.H., SHI, G., GU, Z.-M., HE, K., McLAUGHLIN, J.L.
1996. Recent advances in Annonaceous acetogenins. Nat. Prod. Repts. 13 :275-306.
59. YE, Q., HE, K., OBERLIES, N.H., ZENG, L., SHI, G., EVERT, D., McLAUGHLIN, 1.L.
1996. Longimicins A-D: Novel bioactive acetogenins from Asimina longifolia
(Annonaceae), and structure-activity relationships of asimicin type of Annonaceous
acetogenins. 1 Med. Chern. 39: 1790-1796.
60. YE, Q., ZENG, L., ZHANG, Y, ZHAO, G.-x., MCLAUGHLIN, lL., WOO, M.H.,
EVERT, D.R. 1995. Longicin and goniothalamicinone: Novel bioactive monotetrahydro-
furan acetogenins from Asimina longifolia. J. Nat. Prod. 58: 1398-1406.
61. YE, Q., ALFONSO, D., EVERT, D., MCLAUGHLIN, 1.L. 1996. Longifolicin, longi-
coricin, and gigantetroneninone, three novel bioactive monotetrahydrofuran Annonaceous
acetogenins from Asimina longifolia. Bioorg. Med. Chern. 4:537-545.
62. YE, Q., ZENG, L., SHI, G., EVERT, D., McLAUGHLIN, 1.L. 1996. 4-Acetyl annonacin
and 4-acetyl xylomaticin: Two novel bioactive mono-tetrahydrofuran Annonaceous
acetogenins from Asimina longifolia (Annonaceae). Nat. Prod. Lett. 8:291-298.
63. YE, Q., MCLAUGHLIN, 1.L., EVERT, D. 1996. Absolute stereochemistries of giganin
and loganin, bioactive non-tetrahydrofuran ring Annonaceous acetogenins from Asimina
longlfDlia. Heterocycles 43: 1607-1612.
64. YE, Q., SHI, G., HE, K., McLAUGHLIN, 1.L. 1996. Chlorinated Annonaceous aceto-
genins and their bioactivities. 1. Nat. Prod. 59:994-996.
65. RATNAYAKE, S., GU, Z.-M., MIESBAUER, L.R., SMITH, D.L., WOOD, K.Y., EVERT,
D.R., MCLAUGHLIN, J.L. 1994. Parviftoracin and parviftorin: Cytotoxic bis-
tetrahydrofuran acetogenins with 35 carbons from Asimina parvifiora (Annonaceae). Can.
J. Chern. 72:287-293.
66. GU, Z.-M., FANG, X.-P', ZENG, L., WOOD, K.Y., MCLAUGHLIN, lL. 1993. Bullacin:
A New Cytotoxic Annonaceous acetogenin from Annona bullata. Heterocycles.
36:2221-2228.
67a. HOYE, T.R., YE,Q. 1996. Highly efficient synthesis of the potent antitumor Annonaceous
acetogenin (+)-parviftorin. J. Am. Chern. Soc. 118:1801-1802.
67b. ALALI, EQ., KAAKEH, W, BENNETT, G.W, MCLAUGHLIN, J.L. 1998. Annona-
ceous acetogenins as natural pesticides: Potent toxicity against insecticide susceptible
and resistant German cockroaches (Dictyoplera: Blattellidae). 1 Econ. Entom. 91:
641-649.
68. COLMAN-SAIZARBITORIA, T., ZAMBRANO, 1., FERRIGNI, N.R., GU, Z.-M., NG,
1.H., SMITH, D.L., MCLAUGHLIN, 1.L. 1994. Bioactive Annonaceous acetogenins from
the bark of Xylopia aromatica. 1. Nat. Prod. 57:486--493.
69. COLMAN-SAIZARBITORIA, T., GU, Z.-M., MCLAUGHLIN, 1.L. 1994. Two new
bioactive monotetrahydrofuran Annonaceous acetogenins from the bark of Xylopia aro-
matica. 1. Nat. Prod. 57:1661-1669.
70. COLMAN-SAIZARBITORIA, T., GU, Z.-M., ZHAO, G.X., ZENG, L., KOZLOWSKI,
J.F., MCLAUGHLIN, 1.L. 1995. Venezenin: A new bioactive Annonaceous acetogenin
from the bark of Xylopia aromatica. J. Nat. Prod. 58:532-539.
71. ALFONSO, D., COLMAN-SAIZARBITORIA, T., ZHAO, G.-X., SHI, G., YE, Q.,
SCHWEDLER, J.T., MCLAUGHLIN, 1.L. 1996. Aromin and aromicin, two new bioac-
tive Annonaceous acetogenins possessing an unusual bis-THF ring structure, from
Xylopia aromatica .. Tetrahedron. 52:4215-4224.
72. COLMAN-SAIZARBITORIA, T., ALFONSO, D., MCLAUGHLIN, J.L. 1996. 2,4-cis-
and trans-Venezinones: New ketolactone Annonaceous acetogenins, lacking THF-rings,
from Xylopia aromatica. Phytochem. Anal. 7:313-317.
73. GU, Z.-M., ZHOU, D., WU, 1., SHI, G., ZENG, L., McLAUGHLIN, 1.L. 1997. Screen-
ing for Annonaceous acetogenins in bioactive plant extracts by liquid chromatogra-
phy/mass spectrometry. J. Nat. Prod. 60:242-248.
74. SHI, G., ZENG, L., GU, Z.-M., MACDOUGAL, 1.M., McLAUGHLIN, J.L. 1995.
Absolute stereochemistries of sylvaticin and 12,15-cis-sylvaticin, bioactive C-20,23-cis
non-adjacent bis-tetrahydrofuran Annonaceous acetogenins, from Rollinia mucosa.
Heterocycles. 41: 1785-1796.
75. SHI, G., ALFONSO, D., FATOPE, M.O., ZENG, L., GU, Z.-M., ZHAO, G.-X., HE, K.,
MACDOUGAL, J.M., McLAUGHLIN, J.L. 1995. Mucocin: A new Annonaceous aceto-
genin bearing a tetrahydropyran ring. J. Am. Chern. Soc. 117:10409-10410.
76. SHI, G., KOZLOWSKI, 1.F., SCHWEDLER, J.T., WOOD, K.V, MACDOUGAL, J.M.,
McLAUGHLIN, J.L. 1996. Muconin and mucoxin: Additional nonclassical bioactive
acetogenins from Rollinia mucosa. 1. Org. Chern. 61:7988-7989.
77. SHI, G., ZENG, L., HE, K., YE, Q., MCLAUGHLIN, J.L. 1997. A new aspect of Moshers
method. Absolute configuration of epimeric carbinols in Annonaceous acetogenins. Nat.
Prod. Lett. 10:125-132.
78. SHI, G., ZENG, L., HE, K., YE, Q., GU, Z.-M., McLAUGHLIN, J.L., MACDOUGAL,
J.M. 1997. Applying Moshers method to acetogenins bearing vicinol diols. The absolute
configurations of muricatetrocin C and rollidecins A and B new bioactive Annonaceous
acetogenins from Rollinia mucosa. Bioorg. Med. Chern. 59:548-551.
79. SHI, G., YE, Q., HE, K., McLAUGHLIN, 1.L., MACDOUGAL, J.M. 1997. Rollinecins
A and B: two new bioactive Annonaceous acetogenins from Rollinia mucosa. 1. Nat. Prod.
60:417-418.
80. GU, Z.-M., ZHOU, D., WU, 1., SHI, G., MCLAUGHLIN, 1.L. 1997. Isolation of new
bioactive Annonaceous acetogenins from Rollinia mucosa guided by liquid chromatog-
raphy/mass spectrometery. Bioorg. Med. Chern. 5:1911-1916.
81. FANG, X.-P., ANDERSON, lE., QIU, X.-X., KOZLOWSKI, IF., CHANG, C.l,
MCLAUGHLIN, 1.L. 1993. Gonioheptolides A and B: Novel eight-membered-ring lac-
tones from Goniothalamus giganteus (Annonaceae). Tetrahedron. 49: 1563-1570.
82. FANG, X.-P., SONG, R., GU, Z.-M., RIESER, M.J., MIESBAUER, L.R., SMITH, D.L.,
MCLAUGHLIN, 1.L. 1993. A new type of cytotoxic Annonaceous acetogenin: Giganin
from Goniothalamus giganteus. Bioorg. Med. Chern. Lett. 3: 1153-1156.
83. GU, Z.-M., FANG, X.-P., ZENG, L., MCLAUGHLIN, lL. 1994. Goniocin from Gonio-
thalamus giganteus: The first tri-THF Annonaceous acetogenin. Tetrahedron Lett.
35:5367-5368.
84. GU, Z.-M., FANG, X.-P., ZENG, L., SONG, R., NG, 1.H., WOOD, K.V, SMITH, D.L.,
McLAUGHLIN, 1.L. 1994. Gonionenin: A new cytotoxic Annonaceous acetogenin from
Goniothalalmus giganteus and the conversion of mono-THF acetogenins to bis-THF
acetogenins. 1. Org. Chern. 59:3472-3479.
85. ZHANG, Y, ZENG, L., WOO, M.H., GU, Z.-M., YE, Q., WU, EE., McLAUGHLIN, 1.L.
1995. Goniodenin, a novel bioactive Annonaceous acetogenin from Goniothalamus
giganteus and its conversion to tri-THF acetogenins. Heterocycles. 41: 1743-1755.
86. ZENG, L., ZHANG, Y, McLAUGHLIN, 1.L. 1996. Gigantransenins A, B, and C, novel
mono-THF acetogenins bearing trans double bonds, from Goniothalamus giganteus
(Annonaceae). Tetrahedron Lett. 31 :5449-5452.
87. ZENG, L., ZHANG, Y, YE, Q., SHI, G., HE, K., McLAUGHLIN, 1.L. 1996. Cis-
Gigantrionecin and 4-acetyl gigantetrocin A, two new bioactive Annonaceous acetogenins
from Goniothalamus giganteus, and the stereochemistries of acetogenin 1,2,5-triols.
Bioorg. Med. Chern. 4:1271-1279.
88. ALALI, E, ZENG, L., ZHANG, Y, YE, Q., HOPp, D.c., SCHWEDLER, 1.T.,
MCLAUGHLIN, 1.L. 1997. 4-Deoxyannomontacin and (2,4-cis and trans)-
annomontacinone, new bioactive mono-tetrahydrofuran Annonaceous acetogenins from
Goniothalamus giganteus. Bioorg. Med. Chern. 5:549-555.
89. ALALI, EQ., ZHANG, Y, ROGERS, L.L., MCLAUGHLIN, 1.L. 1997. (2,4-cis and
trans )-Gigantecinone, and 4-deoxygigantecin, bioactive non-adjacent bis-tetrahydrofuran
Annonaceous acetogenins, from Goniothalamus giganteus. 1. Nat. Prod. 60:929-933.
90. RIESER, MJ., FANG, X.-P., RUPPRECHT, 1.K., HUI, Y-H., SMITH, D.L., MCLAUGH-
LlN,1.L. 1992. Bioactive single-ring acetogenins from seed extracts of Annona muricata.
Planta Medica. 59:91-92.
91. RIESER, M.1., FANG, X.-P., ANDERSON, 1.E., MIESBAUER, L.R., SMITH, D.L.,
McLAUGHLIN, 1.L. 1993. 1994. Muricatetrocins A and Band gigantetrocin B: Three
new cytotoxic monotetrahydrofuran-ring acetogenins from Annona muricata. Helvetica
Chim. Acta, 76:2433-2444 (1993) and erratum 77: 882.
92. RIESER, MJ., GU, Z.-M., FANG, X.-P., ZENG, L., WOOD, K.V, McLAUGHLIN, 1.L.
1996. Five novel mono-tetrahydrofuran ring acetogenins from the seeds of Annona
muricata. 1. Nat. Prod. 59: 100-108.
93. WU, EE., GU, Z.-M., ZENG, L., ZHAO, G.-X., Zhang, Y, McLAUGHLIN, 1.L. Sas-
trodihardjo, S. 1995. Two new cytotoxic monotetrahydrofuran Annonaceous acetogenins,
annomuricins A and B, from the leaves of Annona muricata. 1. Nat. Prod. 58:830-836.
94. WU, F.E., ZENG, L., GU, Z.-M., ZHAO, G.-X., ZHANG, Y, SCHWEDLER, 1.T.,
McLAUGHLIN, 1.L., Sastrodihardjo, S. 1995. Muricatocins A and B: Two new bioactive
monotetrahydrofuran Annonaceous acetogenins from the leaves of Annona muricata.
1. Nat. Prod. 58:902-908.
95. WU, EE., ZENG, L., GU, Z.-M., ZHAO, G.-X., ZHANG, Y, SCHWEDLER, 1.T.,
MCLAUGHLIN, J.L., SASTRODIHARDJO, S. 1995. New bioactive monotetrahydrofu-
ran Annonaceous acetogenins, annomuricin C and muricatocin C, from the leaves of
Annona muricata. 1. Nat. Prod. 58:909-915.
96. WU, EE., ZHAO, G.-X., ZENG, L., ZHANG, Y, SCHWEDLER, 1.T., McLAUGHLIN,
1.L., SASTRODIHARDJO, S. 1995. Additional bioactive acetogenins, annomutacin and
(2,4-cis and trans)-IOS-annonacin-A-ones, from the leaves of Annona muricata. 1. Nat.
Prod. 58:1430-1437.
97. ZENG, L., WU, E-E., GU, Z.-M., McLAUGHLIN, 1.L. 1995. Murihexocins A and B,
two novel-THF acetogenins with six hydroxyls, from Annona muricata (Annonaceae).
Tetrahedron Lett. 36:5291-5294.
98. ZENG, L., WU, E-E., McLAUGHLIN, 1.L. 1995. Annohexocin, a mono-THF acetogenin
with six hydroxyls, from Annona muricata (Annonaceae). Bioorg. Med. Chem. Lett.
5: I 865-1868.
99. ZENG, L., WU, F.-E., MCLAUGHLIN, 1.L., SASTRODIHARDJO, S. 1996. Five new
mono-tetrahydrofuran ring acetogenins from the leaves of Annona muricata. J. Nat. Prod.
59: 1035-1042.
100. HUI, y'-H., RUPPRECHT, 1.K., LlU, Y-M., ANDERSON, J.E., SMITH, D.L., CHANG,
CJ., McLAUGHLIN, 1.L. 1989. Bullatacin and bullatacinone: Two highly potent bioac-
tive acetogenins from Annona bullata. 1. Nat. Prod. 52:463-477.
101. MCLAUGHLIN, J.L., HUI, Y.-H. 1993. Chemotherapeutically Active Acetogenins.
(bullatacin and bullatacinone) U.S. Patent No. 5,229,419, issued July 20, 1993.
102. GU, Z.-M., FANG, X.-P., HUI, Y.-H., MCLAUGHLIN, 1.L. 1994. 10-,12-, and 29-
Hydroxybullatacinones: New cytotoxic Annonaceous acetogenins from Annona bullata
Rich. (Annonaceae). Natural Toxins. 2:49-55.
103. GU, Z.-M., FANG, X.-P., MIESBAUER, L.R., SMITH, D.L., MCLAUGHLIN, J.L. 1994.
30-,31-, and 32-Hydroxybullatacinones: Bioactive terminally hydroxylated Annonaceous
acetogenins from Annona bullata. 1. Nat. Prod. 56:870-876.
104. GU, Z.-M., FANG, X.-P., ZENG, L., KOZLOWSKI, 1. F., McLAUGHLIN, J.L. 1994.
Novel cytotoxic Annonaceous acetogenins: (2,4-cis and trans)-bulladecinones from
Annona bullata (Annonaceae). Bioorg. Med. Chern. Lett. 4:473-478.
105. GU, Z.-M., FANG, X.-P., RIESER, M.J., HUI, X.H., MIESBAUER, L.R., SMITH, D.L.,
WOOD, K.Y., MCLAUGHLIN, J.L. 1993. New cytotoxic Annonaceous acetogenins:
Bullatanocin and bullatanocinone, from Annona bullata (Annonaceae). Tetrahedron.
49:747-754.
106. GU, Z.-M., ZENG, L., MCLAUGHLIN, 1.L. 1995. Isolation and structural elucidation of
bioactive C-12,15-cis non-adjacent bis-THF Annonaceous acetogenins. Heterocycles.
41 :229-236.
107. AHAMMADSAHIB, K.I., HOLLINGWORTH, R.M., MCGOVREN, J.P., HUI, Y.-H.,
MCLAUGHLIN, J.L., 1993. Mode of action of bullatacin: A potent antitumor and
pesticidal Annonaceous acetogenin. Life Sciences. 53: 1113-1120.
108. GU, Z.-M., ZENG, L., SCHWEDLER, J.T., WOOD, K.Y., McLAUGHLIN, J.L. 1995.
New bioactive adjacent bis-THF Annonaceous acetogenins from Annona bullata. Phyto-
chemistry. 40:467-477.
109. NONFON, M., LlEB, F., MOESCHLER, H., WENDISCH,D. 1990. Four annonins from
Annona squamosa. Phytochemistry. 29: 1951-1954.
110. FANG, X.-P., GU, I.-M., RIESER, MJ., HUI, Y.-H., MCLAUGHLIN, 1.L., NONFON,
M., LlEB, F., MOESCHLER, H.-F., WENDISCH, D. 1993. Structural revisions of some
non-adjacent bis-tetrahydrofuran Annonaceous acetogenins. 1. Nat. Prod. 56: I 095-· I 100.
I II. HOPp, D.C., ZENG, L., GU, Z.-M., MCLAUGHLIN, 1.L. 1996. Squamotacin: An
Annonaceous acetogenin with cytotoxic selectivity for the human prostate tumor cell line
(PC-3). J. Nat. Prod. 59:97-99.
112. HOPp, D.C., ZENG, L., Gu, Z.-M., KOZLOWSKI, 1.F., McLAUGHLIN, J.L. 1997. Novel
mono-tetrahydrofuran ring acetogenins, from the bark of Annona squamosa, showing
cytotoxic selectivities for the human pancreatic carcinoma cell line, PACA-2. 1. Nat. Prod.
60:581-586.
113. LONDERSHAUSEN, M., LEICHT, w., LIEB, F., MOESCHLER, M. 1991. Molecular
mode of action of annonins. Pest. Sci. 33: 427-438.
114. LEWIS, M.A., ARNASON, J.L., PHILOGENE, BJ.R., RUPPRECHT, J.K., MCLAUGH-
LIN, J.L. 1993. Inhibition of respiration at site I by asimicin, an insecticidal aetogenin of
the paw paw, Asimina triloba (Annonaceae). Pesticide Biochem. Physiol. 45: 15-23.
115. FRIEDRICH, T., OHNISHI, T., FORCHE, E., KUNZE, 8., JANSEN, R., TROWITZSCH,
w., HOFLE, G., REICHENBACH, H., WEISS, II. 1994. Two binding sites for naturally
occurring inhibitors in mitochondrial and bacterial NADH: Ubiquinone oxidoreductase
(Complex I). Biochem. Soc. Trans. 22:226-230.
116. ESPOSITI, M.D., GHELLI, A., BATTA, M., CORTES, D., ESTORMELL, L.F. 1994.
Natural substances (acetogenins) from the family Annonaceae are powerful inhibitors of
mitochondrial NADH dehydrogenase (complex (I). Biochem. 1. 301:161-167.
117. HOLLINGWORTH, R.M., AHAMMADSAHIB, K.I., GADELHAK, G., MCLAUGH-
LIN, 1.L. 1994. New inhibitors of complex I of the mitochondrial electron transport chain
with activity as pesticides. Biochem. Soc. Trans. 22:230-233.
118. LANDOLT, 1.L., AHAMMADSAHIB, K.I., HOLLINGWORTH, R.M., BARR, R.,
CRANE, F.L., BUERCK, N.L., MCCABE, G.P., MCLAUGHLIN, 1.L. 1995. Determi-
nation of structure-activity relationships of Annonaceous acetogenins by inhibition of
oxygen uptake in rat liver mitochondria. Chemico-Biological Interactions 98:1-13.
119. ALFONSO, D., JOHNSON, H.A., COLMAN-SAIZARBITORIA, T., PRESLEY, C.P.,
MCCABE, G.P', MCLAUGHLIN, 1.L. 1996. SARs of Annonaceous acetogenins in rat
liver mitochondria. Natural Toxins. 4: 181-188 and erratum 295.
120. HE, K., ZENG, L., YE, Q., SRI, G., OBERLIES, N.H., ZHAO, G.-X., McLAUGHLIN,
J.L. 1997. Comparative SAR evaluations of Annonaceous acetogenins for pesticidal
activity. Pest. Sci. 49:372-378.
121. GU, Z.-M., ZHAO, G.-X.,OBERLIES, N.H., ZENG, L., McLAUGHLIN, 1.L. 1995.
Annonaceous acetogenins: Potent mitochondrial inhibitors with diverse applications. In:
Phytochemistry of Medicinal Plants: Recent Advances in Phytochemistry, vol. 28, 1.T.
Amason, R. Mata, 1.T. Romeo (eds.), Plenum Press, New York, 1995, pp. 249-310.
122. MORRE, D.1., DE CABO, R., FARLEY, c., OBERLIES, N.H., MCLAUGHLIN, J.L.
1995. Mode of action of bullatacin, a potent antitumor acetogenin: Inhibition of NADH
oxidase activity of HELA and HL-60, but not liver, plasma membranes. Life Sciences,
56:343-348.
123. WOLVETANG, E.1., JOHNSON, K.L., KRAMER, K., RALPH, S.1., LINNANE,
A. W. 1994. Mitochondrial respiratory chain inhibitors induce apoptosis. FEBS Lett.
339:40-44.
124. OBERLIES, N.H., JONES, 1.L., CORBETT, T.H., FOTOPOULOS, S.S., MCLAUGH-
LIN, 1.L. 1995. Tumor cell growth inhibition by several Annonaceous acetogenins in an
in vitro disk diffusion assay. Cancer Lett. 96:55-62.
125. PAULL, K.D., SHOEMAKER, R.H., HODES, L., MONKS, A., SCUDIERO, D.A.,
RABIN STEIN, L., PLOWMAN, 1., BOYD, M.R. 1989. Display and analysis ofpattems
of differential activity of drugs against human tumor cell lines: Development of mean
graph and COMPARE algorithm. 1. Nat. Can. Inst. 8 I: 10898-1092.
126. OBERLIES, N.H., CROY, Y.L., HARRISON, M.H., MCLAUGHLIN, 1.L. 1997. The
Annonaceous acetogenin bullatacin is cytotoxic against multidrug resistant human
mammary adenocarcinoma cells. Cancer Lett. 115:173-179.
127. OBERLIES, N.H., CHANG, C.J., McLAUGHLIN, 1.L. 1997. Structure-activity rela-
tionships of diverse Annonaceous acetogenins against multi drug resistant human mnam-
mary adenocarcinoma (MCF-7/adr) cells. 1. Med. Chern. 40:2102-2106.
128. RUPPRECHT, 1.K., HUI, Y-H., MCLAUGHLIN, 1.L. 1990. Annonaceous acetogenins:
A review. 1. Nat. Prod. 53:237-278.
129. HOLSCHNEIDER, C.H., JOHNSON, M.T., KNOX, R.M., REZAL, A., RYAN, W.J.,
MONTZ, E.1. 1994. Bullatacin in vivo and in vitro experience in an ovarian cancer model.
Cancer Chern. Pharm. 34: 166- I 70.
130. Anonymous. 1898. Lilly's Handbook of Pharmacy and Therapeutics, 13th edition, Eli
Lilly and Company, Indianapolis, p. 89.
131. FANG, X.-P., RIESER, M.1., GU, Z.-M., LZENG, MCLAUGHLIN, 1.L. 1993. Annona-
ceous acetogenins: An updated review. Phytochem. Anal. 4:27-48 and 49-67.
132. ALALI, F.Q., LIU, x.-X., MCLAUGHLIN, 1.L. 1999. Annonaceous acetogenins: Recent
progress. J. Nat. Prod. 61: (in press).
Chapter Six
MOLECULAR CONTROLS FOR

ISOFLAVONOID BIOSYNTHESIS IN RELATION
TO PLANT AND HUMAN HEALTH
Richard A. Dixon,1 Pedro Canovas,I,2 Ze-Jian Guo, 1,2

Xian-Zhi He,1 Chris Lamb,2 and Fiona McAlister l
1Samuel Roberts Noble Foundation

Ardmore, Oklahoma 73402
2 Salk Institute for Biological Studies
La Jolla, California 92037
Introduction, .. , , ....................................... , ... , .. 134

Structure and Biological Activities of Isoftavonoids ............ , ..... , 134
Structure and Distribution ........ , ............................ 134
Isoftavonoids and Plant Health .. , ........ , .. , , ........ , ......... 135
Activity as Phytoalexins ....... , ... , , .... , , .. , ... , .. , . , ...... 135
Activity in Initiating the Rhizobium-Legume Symbiosis ........... 136
Isoftavonoids and Human Health ........ , .... , .................. 137
Isoftavonoids as Phytoestrogens ., ........ , ................... 137
Isoftavonoids as Cancer Chemopreventive Compounds ............ 138
Enzymology and Molecular Biology of Key Steps in
Isoftavonoid Biosynthesis .. , ............................ 139
ChalconeSynthase, Chalcone Reductase, and the Formation of
5-Deoxyisoftavonoids ... , .. , ........ , .... , ........ " .... 139
Isoftavone Synthase ............... , .... , ... , ........ " ....... 143
The 4'-O-Methylation ofIsoftavones ............................. 144
Transcriptional Control of Phytoalexin Biosynthesis ................... 148
Pharmacological Approaches to the Dissection of Signal Pathways for
Isoftavonoid Biosynthesis .... , .. , ........... , .. , ..... , ... 151
Conclusions ...................... , , ................. , ......... 153
* To whom correspondence should be addressed.

133
134 R. A. DIXON et al.
INTRODUCTION
Isoflavonoids function in plant-microbe interactions as antimicrobial phy-

toalexins and regulators of bacterial nodulation genes,I-3 and dietary isoflavones
have been ascribed cancer chemopreventive activity in humans. 4 Many of the
genes encoding enzymes for the elaboration of isoflavonoids from phenyl-
propanoid precursors have been isolated. The reactions catalyzed by isoflavone
synthase and isoflavone O-methyltransferase are critical for the formation of both
isoflavonoid defense compounds and cancer chemopreventive compounds in
alfalfa. Cloning and manipulating the genes encoding these enzymes provides a
means to develop novel transgenic plants with improved disease resistance and,
perhaps, added health benefits for humans.
This chapter provides an introduction to the above referenced biological
activities of isoflavonoids and then describes in more detail the enzymology and
molecular biology of several key biosynthetic reactions. The isoflavonoid pathway
is transcriptionally induced in response to elicitation or infection. We have shown
that two distinct transcription factors, a basic leucine zipper protein and a
homolog of the human Ku autoantigen, are involved in activation of expression
of the chalcone synthase gene. Phosphorylation of these factors may control
their activity, and represents the end stage of the signal transduction pathway(s)
linking elicitor reception to the transcriptional activation of (iso )flavonoid
biosynthetic genes.
The exact nature and complexity of the signal transduction pathways for
induction of bioactive isoflavonoids remain to be determined. We here review
results of pharmacological studies aimed at addressing the role of active oxygen
species in induction of isoflavonoid biosynthesis during plant defense. Hydrogen
peroxide generated by the oxidative burst, a key early component of the hyper-
sensitive response to microbial pathogens, appears necessary but not sufficient
for the accumulation of isoflavonoid phytoalexins in soybean cells. As yet
unidentified diffusible signal molecules of high potency have also been impli-
cated for the induction of the phytoalexin response in cells proximal to those
responding directly to pathogen or elicitor.
STRUCTURE AND BIOLOGICAL ACTIVITIES

OF ISOFLAVONOIDS
Structure and Distribution
The flavonoids are a major class of phenylpropanoid-derived plant natural

products. Their fifteen carbon (C 6-C 3-C 6) backbone can be arranged as a 1,3-
diphenylpropane skeleton (flavonoid nucleus) or as a 1,2-diphenylpropane
skeleton (isoflavonoid nucleus). Although I,3-diphenylpropane flavonoid
MOLECULAR CONTROLS FOR ISOFLAVONOID BIOSYNTHESIS 135
derivatives are almost ubiquitous among terrestrial plants, the isoflavonoids are
restricted primarily to the Leguminosae, although they occur rarely in other
families such as the Apocynaceae, Pinaceae, Compositae, and Moraceae. 5•6
The nearly 900 naturally occuring isoflavonoid aglycones can be divided into
9 major classes based on differences in their basic carbon skeletons. 5 Members of
these various classes, some of which will be described in more detail later in this
article, are shown in Figure I. The isoflavones and pterocarpans are the most abun-
dant isoflavonoids, with 334 and 152 different structures, respectively, having been
described as of September 1994. This enormous diversity is caused by the various
substituents, for example methoxyl, prenyl, methylenedioxy, that can occur on
many different positions of the A and Brings.
The limited taxonomic distribution of the isoflavonoids is directly related to
the occurrence of the enzyme isoflavone synthase, which catalyzes the aryl migra-
tion reaction leading to the formation of an isoflavone from a flavanone (see below).
It is probable that this enzyme has evolved independently in the Leguminosae and
the other diverse taxa in which isoflavonoids are occassionally found.
Isoflavonoids and Plant Health
Activity as Phytoalexins. Many isofiavonoids have significant, broad-

spectrum antimicrobial activity and are therefore believed to help the plant fight
microbial disease. Antimicrobial isoflavonoids can be classified as pre-formed
"phytoanticipins", such as the prenylated isofiavones of lupin, which are synthe-
sized in various organs ofthe plant during seedling development, 7 and the inducible
"phytoalexins"~ such as the pterocarpans of bean, alfalfa, pea, and soybean.9
In spite of the large literature on isofiavonoid phytoalexins, there are still
HO%HO~I
I I ....- ~ H HO
~ ~ ~I
o ~ I CH 30 ~ OCH 3
OCH 3
OH
Daidzein (isoflavone) (-)Sativan (isoflavan) (-)Medicarpin (pterocarpan)
HO%I
~ ~ o
° ~I OH o >
Coumestrol (coumestan) (+ )Pisatin (pterocarpan)
Figure 1. Representative examples from four classes of isoftavonoids.

no clear rules defining structure/activity relationships. Many plant pathogens can

metabolize, and thereby detoxify, isoflavonoids,lo and structure/activity relations,
therefore, often depend on the fungus used in the bioassay.
The exact function of phytoalexins in disease resistance has, in most cases,
still to be confirmed. The phytoalexin "hypothesis" is based essentially on cor-
relative data; isoflavonoids accumulate in infected plant cells to levels shown to
be antimicrobial in vitro. II However, few studies have directly tested the hypoth-
esis that phytoalexins actually function in disease resistance. Isolates of Nectria
hematococca with reduced ability to degrade the pterocarpan pisatin have reduced
virulence on pea, suggesting that pisatin is a factor in the disease resistance
response. 12 To evaluate critically the hypothesis that isoflavonoids function as phy-
toalexins, it will be necessary to produce mutant or genetically engineered plants
lacking only the isoflavonoid phytoalexins. Mutants of Arabidopsis lacking the
indole phytoalexin camalexin showed increased disease symptoms following
infection with a compatible race, suggesting that this phytoalexin may playa role
in disease symptom limitation, but these plants were not less resistant to incom-
patible bacteria. 13 This would suggest, not surprisingly, that phytoalexins do not
in themselves constitute a major determinant of resistance in incompatible
interactions, since these result in not only phytoalexin accumulation but also the
hypersensitive response and associated metabolic changes. Now that genes encod-
ing late enzymes of the isoflavonoid phytoalexin pathway have been cloned, it
should be possible to test the functions of isoflavonoids as phytoalexins by direct
transgenic approaches.
Isoflavonoids can act as stimulatory, as well as inhibitory, factors in inter-
actions of legumes with fungi. For example, the isoflavones daidzein and genis-
tein, released in soybean root exudates, act at nanomolar concentrations as
chemoattractants for zoospores of Phytophthora sojae, and also induce their
encystment and germination. 14 Likewise, biochanin A and several pterocarpan
phytoalexins, including medic arpin and pisatin, stimulate germination of
Fusarium solani spores at micromolar concentrations. 15
Activity in Initiating the Rhizobium-Legume Symbiosis. Another manner in

which isoflavonoids may impact plant health is through their role in the establish-
ment of symbiotic nitrogen fixation. Flavonoid and isoflavonoid compounds
play critical roles as activators of Rhizobium nod genes, leading to the formation
by the bacteria of substituted Iipochitooligosaccharide signal molecules (Nod
factors) that in turn induce root hair curling and the cortical cell divisions that
characterize the early development of the nitrogen fixing legume root nodule. 16
F1avones in alfalfa and red clover root exudates potently activate nod gene
expression,3.17 whereas the major inducers of the nod genes of the soybean sym-
biont Bradyrhizobium are the isoflavones daidzein and genistein. 18 Reduced syn-
thesis of daidzein in soybean roots at suboptimal temperatures limits Rhizobial
colonization. 19
Isoflavonoids and Human Health
lsojiavonoids as Phytoestrogens. Dietary isoftavonoids can exert estro-

genic effects in animals. Reports from Australia from as early as 1946 documented
the occurrence of infertility in sheep that had grazed on clover rich in the isoflavone
formononetin,20 and breeding programs to select for low isoflavone lines of sub-
terranean clover were subsequently implemented? I It has even been suggested that
the natural population sizes of California quails during periods oflow food supply
are controlled by their feeding on isoflavone-rich legumes. 22
Daidzein and genistein share key structural features with the potent estro-
gen estradiol-17P, particularly the phenolic ring and the distance (11.5 A)
between the two hydroxyl groups (Fig. 2). These features determine their ability
to bind estrogen receptors, and isoflavonoids can thus exert both estrogenic and
anti-estrogenic activity, the latter by competing for receptor binding by estradiol.
Structural similarities have also been noted between isoflavones and tamoxifen
(Fig. 2), an antiestrogen which has been clinically tested as a chemopreventive
agent in women with high risk of breast cancer.4 Equol (a major metabolite of
dietary isoflavonoids formed by the gastrointestinal flora) (Fig. 2) and genistein
can displace bound estrogen and testosterone from human sex steroid binding
protein.23 Thus, phytoestrogens might also affect clearance rates of androgens and
estrogens and thus the availability of the hormones to target cells.
Similar to estradiol, formononetin stimulates mammary gland proliferation
HO HO
HO%II
~ ;:?
OH ° ~ I OCH 3
Genistein Formononetin
Biochanin A
oo~ H°'COu
~ 1 ;:?
~
I
OH
OH Equol
Estradiol 17-13
Tamoxifen
Figure 2. Structural similarities between isoflavonoid phytoestrogens and estradiol and

tamoxifen.
l38 R. A. DIXON et af.
and increases estrogen receptor and plasma prolactin levels in mice,24 although it
is 15,000 times less potent than estradiol for binding to murine mammary estro-
gen receptors. 24 Thus, its major biological activity is probably as an estrogen
agonist, although its concentration in the "Western" diet is probably too low
to cause physiological effects. However, in humans eating a soy-rich diet,
isoflavonoids may be present in the urine at very high levels. For example, human
adults given a daily diet containing 40 g of soy protein secreted 5.3 mg of equol
per day, compared with only 2-27 Jlg of estrone glucuronide, the principal urinary
estrogen that is released during the follicular phase of the menstrual cycle. 25 This
is a 100-fold increase in urinary equol above that observed in adults who consume
little soy products in their diet.
The observation that phytoestrogens can be assayed by activation of an
estrogen-regulated transcription system in yeast cells 26 provides a method for
rapid screening of these compounds. This approach is now being applied to
extracts from plants, such as Artemesia,27 used in herbal medicine for treatment
of conditions such as menstrual abnormalities. It is likely that several species will
contain bioactive flavonoids and isoflavonoids.
An interesting, but yet unresolved, question is whether plants themselves
contain estrogen receptor-like proteins, and, if so, whether phytoestrogens may
function hormonally in plants. Such a view would have held little weight ten
years ago, but the recent demonstration of the functions of the brassinosteroids in
plant growth and development give credence to the possibility that new
classes of plant regulators may await discovery. In this regard, it is interesting that
dioecious species, such as the Osage orange (Madura pomifera), contain
estrogen-receptor binding compounds, the levels and seasonal variations of which
are consistent with a possible role in sex determination. 26 Although the exact nature
of these compounds has yet to be determined, it is interesting that M. pomifera, a
member of the Moraceae, contains at least three prenylated isoflavonoids. 6
Isojlavonoids as Cancer Chemopreventive Compounds. There has been an

explosion recently of interest in dietary isoflavonoid phytoestrogens because of
the significant correlations between an isoflavone-rich soy-based diet and reduced
incidence of breast cancer or mortality from prostate cancer.28 An epidemiologi-
cal study of Singapore Chinese women (420 healthy controls and 200 with
histologically confirmed breast cancer) indicated that soy consumption was
directly correlated with reduced risk of cancer,29 and these effects appear to be
dietary rather than genetic.
One gram of powdered soybean chips contains nearly 800 Jlg of daidzein and
over 500 Jlg of genistein, whereas one gram of soy protein has approximately 150
Jlg of daidzein and 250 Jlg of genistein. 30 Because urinary excretion of daidzein,
genistein, and equol was shown to be at least lO-fold higher in farm workers from
Japan compared to Americans or Europeans, isoflavonoids found in soy products
might be the agents responsible for reduced cancer risk. 28
When administered neonatally, genistein effectively protects against

chemically-induced mammary tumors in rats. 31 The effects include increased
latency, reduced tumor incidence and multiplicity, and more rapid maturation
of undifferentiated end buds to differentiated lobules. Furthermore, genistein and
biochanin A inhibit the growth of human stomach cancer cell lines in vitro, appar-
ently by stimulating a signal transduction pathway leading to apoptosis. 32 When
these cancer cells were transplanted into mice, biochanin A, but not genistein,
significantly inhibited tumor growth. Although no clinical trials have been reported
documenting effects of controlled dietary supplementation with genistein on breast
cancer incidence in humans, it is known that a high soy diet containing up to 45
mg of isoflavones per day causes changes in the menstrual cycle that may help
reduce cancer risk. 33 It has been suggested recently that the high levels of
isoflavones in breast milk of humans consuming a high soy diet may provide the
infant with protection against cancer later in life. 34
Not all the effects of isoflavones on human health are necessarily associ-
ated with their estrogenic activity. Genistein also inhibits DNA topoisomerase and
tyrosine protein kinase,35 as well as possessing antioxidant activity and being a
cell cycle inhibitor. 36 Kinase inhibition is generally regarded as being specific for
tyrosine kinases, such as epidermal growth factor (EGF) receptor, although at
higher concentrations genistein also inhibits protein histidine kinase. 37 Genistein
blocks EGF-mediated tyrosine phosphorylation in vivo in human epidermal car-
cinoma cells. 35 When specifically targeted to the B-cell-specific receptor CD-19
by conjugation to a monoclonal antibody, genistein selectively inhibited CD-19-
associated tyrosine kinase activities, resulting in death of human B-cell precur-
sor leukemia cells. 38 Other isoflavones such as daidzein do not inhibit tyrosine
kinase activity, and are therefore used as controls in pharmacological experiments
utilizing genistein.
Unlike other isoflavonoids, genistein only exerts toxicity at concentrations
greatly in excess of those at which it first exerts its biological effects, making it
an important subject for future studies on cancer chemoprevention. Further infor-
mation on the clinical effects of isoflavonoid phytoestrogens can be found in
several reviews. 3942
ENZYMOLOGY AND MOLECULAR BIOLOGY OF KEY

STEPS IN ISOFLAVONOID BIOSYNTHESIS
Chalcone Synthase, Chalcone Reductase, and the Formation of

5-Deoxyisoflavonoids
The first C 15 precursor of the isoflavonoids is the chalcone derived from

the head-to-tail condensation of 4-coumaroyl CoA and three molecules of
malonyl CoA catalyzed by the enzyme chalcone synthase (CHS). CHS is a
dimeric polyketide synthase, subunit Mr approximately 42,000, which catalyzes

the addition, condensation, and cyclization reactions leading to the formation
of 2',4,4',6'-tetrahydroxychalcone (naringenin chalcone) (Fig. 3). CHS has been
purified and characterized, and its genes cloned, from many plant species. 43 -47
The genetic model plant Arabidopsis thaliana contains a single CHS
gene,48 which is clearly sufficient for the basic functions of plant growth and
development. However, in most legume species, CHS is encoded by multi gene
families, consisting of 6-8 members in green bean (Phaseolus vulgaris),49 at
least 7 in soybean,50 at least 7 in pea,45 at least 4 in subterranean clover,46 6 or 7
in Kudzu vine (Pueraria lobata),5! and more than 7 in alfalfa. 47 Gene family
members are often tightly clustered in the genome,44,46,49 suggesting that they have
arisen from fairly recent gene duplication events. The multiple forms of CHS in
legumes may have evolved to serve particular specializations of the flavonoid
pathway, for example in production of isoflavonoid phytoalexins and flavonoid-
lisoflavonoidlchalcone nodulation gene inducers. However, there is no direct evi-
dence currently to support this hypothesis. In alfalfa, at least 5 different members
3X ~-SCOA
CH
I 2
COO-
+
CoAS
/3'<:::
I
g
OH
3CoASH
c~~
t
CoAS
t
Malonyl CoA 0 ~ 0 o o o
4-Co"m",o," Co. 'C~
NADPH 1 CHS • ''''"'.~'''
YY
HOyyOH , , ( ) - o : H S
OH 0
-= CoAS
0 o OH 0
I
(2' ,4,4' ,6'-Tetrahydroxychalcone

}-. H20
(naringenin chalcone)
+CHS
HO~OH
fj ~ OH
:::,..
I I -
o
2' ,4,4'-Trihydroxychalcone
(isoliquiritigenin)
Figure 3. Reactions catalyzed by chalcone synthase in the presence and absence of chalcone
reductase.
of the CHS gene family are constitutively expressed in roots and root nodules,
but not in the aerial parts of the plant. However, these family members are
expressed in leaves, at the onset of the isoflavonoid phytoalexin defense response,
following exposure to pathogens or elicitors. 47 The CHS proteins encoded by the
different gene family members are generally very similar in primary sequence,47
and it is not known if they possess different kinetic properties or are differentially
localized in the cell.
Many isoflavonoids lack the 5-hydroxyl group (6'-hydroxyl, chalcone num-
bering), and are thus derived from 2',4,4'-trihydroxychalcone rather than from
naringenin chalcone. The 5-deoxy isoflavonoids are particularly prevalent in
legume roots, and the pterocarpan phytoalexins are invariably of this class. 13C_
Labelling studies indicated that the 5-hydroxyl group is lost prior to the cycliza-
tion of the A-ring of the chalcone,52 presumably at the polyketide stage. Crude
extracts from elicited cell cultures of Glycyrrhiza echinata were shown to produce
2',4,4' -trihydroxychalcone and its corresponding flavanone liquiritigenin, in addi-
tion to naringenin chalcone, from malonyl CoA and 4-coumaroyl CoA in the pres-
ence of high concentrations of NADPH. 53 The chalcone was produced first and
then converted to the flavanone by chalcone isomerase present in the preparation.
The activity was described as 6'-deoxychalcone synthase, and was demonstrated
also in G. echinata protoplasts. 53
Purified soybean CHS was shown to require the presence of a separate
protein, given the trivial name of "chalcone reductase" (CHR), for NADPH-
dependent formation of (iso )liquiritigenin. 54 The reductase was purified to appar-
ent homogeneity and was shown to be a monomer of Mr 34,000, that catalyzed
the transfer of the pro-R-hydrogen of [4-3H]NADPH to the polyketide bound to
CHS, with resultant loss of the hydroxyl function as water (Fig. 3). The enzyme
exhibited approximately 90% maximal activity at a molar ratio (CHS : reductase)
of 2: I, and could co-act with CHS from parsley, a species that does not synthe-
size 6'-deoxychalcone derivatives. 54 This latter point suggests that the multiple
forms of CHS found in legumes are unlikely to be involved differentially in the
formation of 6'-deoxy- and 6'-hydroxy-chalcones, although this remains to be
proven experimentally.
CHR is encoded by a small gene family in soybean 55 and alfalfa,56-59 and
has also recently been cloned from Pueraria lobata and Glycyrrhiza echinata;59.6o
the gene does not appear to be present in species such as carrot and parsley that
do not accumulate 5-deoxy isoftavonoids. 55 The enzyme possesses a leucine
zipper domain, but it is not yet known if this is involved in interactions with CHS.
Although the enzyme is a polyketide reductase, it does not share significant
sequence identity to the reductases of fatty acid synthesis; rather, it is related to
a mammalian aldose reductase and to 2,5-diketo-D-gluconic acid reductase from
Corynebacterium. 55
It is still not clear why co-action of CHR with CHS results in no more than
100 A
50
~ 25
.:;
U
a:I
Q)
>
~
<i.i 100
a:
75
Figure 4. Relative transcrip-
50 tion rate kinetics of (A)
CHS (00) and (8) CHR (00)
in relation to PAL (0) in
25 elicitor-treated alfalfa cell sus-
pension cultures. Values from
nuclear transcript run-on
analyses are normalized to
3.0 100% maximum level. Data
Hours post elicitation from ref 59.
50% formation of the 6'-deoxycha1cone. Interaction of recombinant CHR with a

CHS heterodimer containing a single active site produced no significant differ-
ence in 6'-deoxy to 6'-hydroxy product ratio from that observed with wild-type
CHS, indicating that the production of both cha1cones can not result from the
presence of two functionally distinct active sites (i.e. one coupled to CHR and
one not).61 Recently, the three dimensional structure of CHS has been resolved
by X-ray crystallography, and a mechanism was proposed for the chlacone reduc-
tion involving the polyketide exiting the active site of CHS for reduction, and re-
entering for final cyclization (J.-L. Ferrer, 1.M. Jez, M.E. Bowman, R.A. Dixon
and 1. Noel, unpublished results).
Nuclear transcript run-on experiments have demonstrated closely coo-
rdinated induction of CHR with CHS at the transcriptional level in elicited
alfalfa cell suspension cultures (Fig. 4).62 We have recently isolated genomic clones
corresponding to two distinct alfalfa CHR genes (1. Clouse, R.A. Gonzales and
R.A. Dixon, unpublished results). Sequence analysis of their promoter regions
reveals the presence of both G-box and H-box elements proximal to the trancrip-
tion start site, a feature shared with legume CHS genes (Fig. 5, see below). It is
MOLECULAR CONTROLS FOR ISOFLAVONOlD BIOSYNTHESIS 143
BeanchslS
-72
H-Box G' G' [fATAl ATG

Alfalfa chr7 ----------1I...-[H}-()------......L..I--
Figure 5. Diagramatic representations of the bean chs15 and alfalfa chr7 promoters. SB =
silencer box. Boxes marked G' contain single base mismatches from the chslS G-box core
sequence (CACGTG). See text for details.
likely, but not yet proven, that these elements bind transcription factors responsi-
ble for the co-ordinated transcriptional activation of CHS and CHR.
Isoftavone Synthase
In 1984, it was shown that an activity present in microsomes from elicitor-

treated soybean cell suspension cultures could catalyze the conversion of narin-
genin to genistein, or of 2(S)-liquiritigenin to daidzein, in the presence of
NADPH. 63 The crude microsomal preparation was labile, with a half-life of only
10 min at room temperature, and the reaction was dependent on NADPH and mol-
ecular oxygen. The reaction proceeds in two steps. Naringenin is first converted
in a cytochrome P450-catalyzed reaction requiring NADPH and O2 to the corre-
sponding 2-hydroxyisoflavanone. This relatively unstable compound then under-
goes dehydration to yield genistein (Fig. 6). The origin of the 2-hydroxyl was
determined from studies on the isoflavone synthase activity in microsomes from
elicited cell cultures of Pueraria lobata. 64 1XO from 180 2 was incorporated into
the 2-hydroxyl group, resulting in a 2-hydroxyisoflavanone with molecular ion
HOWl ' b
H
V
I OH
HO
HO
~ NADPH
• OH
OH Genistein
(-)(25) Naringenin 2,5,7,4'- Tetrahydroxy-
isoflavanone
Figure 6. The reactions catalyzed by the "isoftavone synthase" complex.

shifted by two mass units, whereas there was no corresponding shift in the mol-
ecular ion of the isoflavone, consistent with the subsequent dehydration reaction.
The dehydration reaction appears to be catalyzed by an activity present predom-
inantly in the cytoplasmic supernatant. The 2-hydroxyisoflavanone can also
spontaneously convert to the isoflavone. The enzyme is stereoselective, and (2R)-
naringenin is not a substrate.
The model for the reaction pathway of "isoflavone synthase" (Fig. 6)65 there-
fore involves P450-catalyzed hydroxylation coupled to aryl migration, a reaction
with mechanistic similarities to the well-described proton migration mechanism
of some P450 reactions. Similarities between the mechanism ofIFS and other reac-
tions, such as ring condensation of ent-7-hydroxy-kaurenoic acid to GA I2 aldehyde
in gibberellin biosynthesis, and sterol demethylation, have been discussed. 65
To date, there have been no reports on the purification to homogeneity, or
the molecular cloning, of either of the two enzymes of the IFS complex. The
flavanone 2-hydroxylase cytochrome P450 from Pueraria has been solubilized
with Triton X-IOO, and partially purified by ion exchange chromatography; the
enzymatic reaction could be reconstituted by addition of NADPH cytochrome
P450 reductase. 66 The 2-hydroxyisoflavanone dehydratase has been purified from
elicitor-treated P. lobata cells, and is a soluble monomeric enzyme of subunit Mr
38,000. 60 •67 It is not yet clear whether this enzyme physically associates with the
P450 hydroxylase catalyzing the aryl migration.
The low abundance and extreme lability of the isoflavone synthase
cytochrome P450 make purification by classical procedures a daunting task. The
approach taken in our lab to cloning IFS has been to isolate cDNAs encoding
cytochrome P450s by a PeR-based strategy similar to that reported recently for
Arabidopsis P450s. 68 This is made possible by the existence of small motifs with
significant sequence homology in all known P450s. By using this strategy, we
have isolated 6 novel P450s from an alfalfa cell suspension library, and have
shown that 5 of these are elicitor-inducible (C.L. Steele and R.A. Dixon, unpub-
lished results). After suitable modifications to optimize expression, full length
cDNAs have been expressed in E. coli and in insect cell cultures, and functional
analysis is underway.
The 4'-O-MethyJation of Isoflavones
The isoflavonoid phytoalexins of several species, including alfalfa and

chickpea, are methylated at the 4'-position of the B-ring (Fig. I). In these
species, 4'-O-methylation appears necessary prior to further modification of
the isoflavonoid nucleus by 2' -hydroxylation and subsequent reduction to
isoflavanone, the substrate for formation of pterocarpans (Fig. 7).69 Surprisingly,
particularly in view of the relative ease of purification and cloning of phenolic 0-
methyltransferases from plants, the exact nature of the enzymatic step resulting in
4'-O-methylation of isoflavonoids is still unclear.
COO-
+3 \:OSCOA
malonyl-CoA
phenylalanine p-coumaroyl-CoA
CHS+CHR ~
H0yYHIr<}-OH
4,2',4'-trihydroxychalcone ~ '=-'
CHI ~ 0
oH
7,4'-{jihydroxyflavanone H 0 W - O -
IFS ~°
daidzeinHO~) r-\. . H
~
I°
7-IOMT/
CH,0'O?-o- /' .'·IOMT
I °I '_ OH lonnononel'nH"ti'J I"\. CH,

isoformononetin ~
IFOH ~ 0 -
,.hydm><ylo~oo,,'o HO~o,,",
IFR ~ 0 HO
(3A,....."''' HO~OC"'
PTS~O HO
HO
(6a, 11 aR)-medicarpin
or (-)-medicarpin
Figure 7. The late stages of biosynthesis of isoflavonoid phytoalexins in alfalfa.
Paradoxically, radio labelled precursor feeding experiments with eli-

cited alfalfa seedlings indicated that, although 2',4,4'-trihydroxychaIcone and
formononetin were good precursors of medicarpin, daidzein (the presumed sub-
strate of the 4' -O-methyitransferase) was not incorporated. 70 These results were
originally interpreted as indicating a requirement for methylation of the B-ring

during the aryl migration reaction catalyzed by isoftavone synthase. 70 That this can
not be obligatory, at least in vitro, is clear from the demonstration that the aryl
migration catalyzed by the 2-hydroxyisoftavanone synthase described above occurs
in the absence of methylation in species in which the 4' -hydroxyl group is either
free (e.g. in soybean) 66 or methylated (e.g. in alfalfa).7! A mutant of subterranean
clover (Trifolium subterraneum), which produced greatly reduced levels of for-
mononetin and biochanin A, accumulated high levels of daidzein, suggesting that
daidzein is indeed the immediate precursor offormononetin. 72
The contradiction between the precursor feeding studies and the enzymo-
logical and genetic studies pointing to daidzein as a substrate for 4'-O-methylation
has been compounded by attempts to demonstrate the enzymatic basis for the origin
of the 4'-methoxyl group of isoftavones. It would be predicted that an isoftavone
4' -O-methyltransferase would catalyze the conversion of daidzein to formononetin,
or of genistein to biochanin A. However, in a study of isoftavone 4'-0-
methylation in chickpea cell cultures, the enzyme initially described as a 4'-0-
methyltransferase was later shown to be an isoftavone 7-0-methyltransferase
which methylated the A-ring of genistein to yield prunetin (5,4'-dihydroxy-7-
methoxyisoftavone). This enzyme was reported to be a dimer ofM r 110,000, with
a pH optimum of 9.0 and a Km for genistein of 80 )lM. 73
Treatment of alfalfa cell suspension cells with yeast elicitor results in a
massive induction of a similar isoftavone 7-0-methyltransferase activity/4 which
methylates the A-ring of daidzein to produce isoformononetin (4'-hydroxy-7-
methoxyisoftavone).74 Isoformononetin is a rare naturally occurring compound
which is unlikely to be involved in the formation of medicarpin, and which we have
not shown to be present in our alfalfa cell cultures. The enzyme is a monomer of
Mr 41 ,000 that can be photoaffinity labelled with [3H]-S-adenosyl-L-methionine. 74
Partially purified alfalfa isoftavone 7-OMT had a pH optimum of 8.5, a Km value
of 20)lM for daidzein, and exhibited a low level of 4'-O-methyltransferase
activity resulting in the formation of formononetin. 74 It has proven impossible to
purify this 4'-OMT activity further.75 The extremely low level of daidzein 4'-0-
methyltransferase activity in elicited alfalfa cultures contrasts with the strongly
increased extractable activity of the isoftavone 7-O-methyltransferase in parallel
with all the other known enzymes in the pathway leading to medicarpin. 74 .76
A major problem in the purification of isoftavone O-methytransferase
(IOMT) from alfalfa was contamination with high levels of co-purifying
caffeic acid 3-0-methyltransferase (COMT) activity. We therefore recently devel-
oped a substrate-based affinity chromatographic system to purify the 41 kDa iso-
ftavone 7-0MT to homogeneity.75 The purified isoftavone O-methyltransferase
methylated the 7-position (A ring) of daidzein or genistein, and could also methy-
late, less efficiently, the 5-hydroxyl group of genistein. It was inactive against
formononetin.
Four internal peptide sequences were obtained from the purified protein,
one of which had high (72%) sequence identity to a region of a catechol-O-
methyltransferase from barley. All 4 internal peptides respectively had about
55% amino acid sequence identity to 4 regions of 6a-hydroxymaackiain 3-0-
methyl transferase from Pisum sativum,77 but had no sequence identity to the
alfalfa COMT or chalcone 2'-0-methyltransferase (ChaIOMT) genes previously
cloned in our laboratory. Sequence information from three internal tryptic pep-
tides was used to design oligonucleotide primers to facilitate cloning of a full-
length IOMT cDNA. At least two different IOMT genes are present in the genome
of tetraploid alfalfa, whereas a single gene is present in chickpea and cowpea. 78
Sequences hybridizing to alfalfa IOMT were not present in the genomes of
tobacco and potato. The alfalfa IOMT8 cDNA was functionally expressed in E.
coli, and the properties of the enzyme shown to be essentially identical to those
of the enzyme purified from cell cultures. 78 Strong induction ofIOMT transcripts
preceded the accumulation of medicarpin in elicitor-treated cell cultures.
We believe that the enzyme with isoflavone 7-0MT activity in vitro methy-
lates the 4'-position in vivo. The unexpected precursor feeding results in alfalfa
can be explained if the OMT is in a "metabolic compartment" or "channel," and
its association with the enzymes producing its substrate or removing its product
could account for the different regiospecificities observed in vivo and in vitro.
The isoflavone synthase and 2' -hydroxylase are both microsomal cytochrome
P450s, with which the 4' -OMT could be physically associated. Thus, only
daidzein formed in situ by microsomal isoflavone synthase, but not exogenously
supplied daidzein, might act as substrate for the OMT. Such close associations
might be necessary if the methyltransferase could only act on the 4'-position
during the aryl migration of the B-ring, the pK of the A-ring 7-hydroxyl making
it the preferred site for methylation in a non-associated in vitro system. We have
expressed the alfalfa IOMT in the yeast two-hybrid system as a fusion to the gal4
DNA binding domain, and used this to screen an elicited alfalfa culture cDNA
library fused to the gal4 transactivation domain for expressed proteins interact-
ing with the methyltransferase. This approach failed to reveal any alfalfa proteins
capable of interacting with the methyltransfrease in yeast cells (IT. Reddy and
R.A. Dixon, unpublished results).
It is now possible to test the methyltransferase channel hypothesis by mol-
ecular genetic strategies. We have recently generated a large number of indepen-
dent alfalfa transformants harboring the IOMT cDNA sequence in both sense and
antisense orientations, under control of the constitutive cauliflower mosaic virus
35S promoter (X.Z. He and R.A. Dixon, unpublished results). Some plants con-
taining sense constructs express high levels of IOMT transcripts in the leaves, an
organ in which IOMT is not normally expressed. Cell suspension cultures have
been initiated from these plants to test for the impact of the transgene expression
on isoflavone metabolite levels. Reduction of medicarpin formation in antisense
and sense-suppressed lines would be evidence for an involvement of IOMT in the

4'-O-methylation of daidzein.
TRANSCRIPTIONAL CONTROL OF PHYTOALEXIN

BIOSYNTHESIS
In elicited plant cells, the accumulation of isoflavonoids is usually accom-

panied by co-ordinated increases in the extractable activities of all the biosynthetic
enzymes of the pathway. These increases are primarily the result of increased
transcription of the genes encoding the various biosynthetic enzymes. 62 . 79
Two major questions have now become the focus of much research atten-
tion in several laboratories worldwide. What are the signal transduction pathways
linking elicitor perception at the cell surface to increased transcription of phy-
toalexin biosynthetic genes, and how are these integrated within the total program
of induced defense responses? A related question concerns the number of differ-
ent transcriptional activators that may be necessary to orchestrate the complete
pathway response. To address these questions, it is necessary to isolate the
transcriptional regulators responsible for the elicitation response. Because of the
ease of selection for mutations affecting synthesis of colored flavonoid deriva-
tives, genetic approaches have been successfully used for the cloning of tran-
scriptional regulators of the anthocyanin pathway.80 Such a strategy is much less
easy for colorless isoflavonoids, and the genetic intractability of many of the
legume species used for isoflavonoid research is a further problem. Hopefully, the
development of model legume genetic systems such as Medicago truncatula 81
will soon make it possible to use functional genomics approaches, such as high
frequency T-DNA or transposon insertion or activation tagging, coupled with high
throughput metabolite profiling, to isolate genes involved in the regulation of
isoflavonoid synthesis.
An alternative approach to understanding the mechanisms underlying the
transcriptional control of isoflavonoid biosynthetic genes is to use a combination
of molecular and biochemical techniques to identify the cis-elements in
isoflavonoid pathway gene promoters that confer infection- or elicitor-inducibil-
ity.H2.83 This information can be used to design probes for the isolation of the
transcription factors that bind to these sequences, and cloned transcription factors
can then be used as "substrates" to search for molecules that might interact with
these factors to modulate their activity. We describe below such a series of exper-
iments using an elicitor-inducible CHS gene promoter as the starting point for
walking back up the signal transduction pathway for elicitor modulation of
isoflavonoid synthesis. 79 •84
The bean CHS 15 promoter contains several cis-elements that have been
implicated in transcriptional control of the gene following elicitation or infection,
and also in developmentally controlled tissue-specific expression. These elements
have been identified by functional analysis of CHS promoter/reporter gene fusions

and promoter mutations thereof in electroporated protoplasts or transgenic
plants,85-87 by measurement of in vitro DNA binding activity of plant extracts,84.88
by analysis of DNAase I hypersensitive sites in chromatin,89 and by assay oftran-
scription from promoter templates in vitro. 90 In essence, the qualitative express-
sion pattern of the CHS15 promoter is determined by a small region containing
a proximal H-box (CCTACC) element (H-box I) and a G-box (CACGTG) element
(Fig. 5). Other upstream H-boxes and "silencer boxes" may be involved in
determining quantitative levels of expression.
A standard procedure for cloning transcription factors is to use single or
multimerized labelled oligonucleotides corresponding to the target DNA
sequence to screen cDNA expression libraries. However, labelled multimers of
the functionally implicated downstream CHS15 H-box did not detect expression
of any sequence-specific DNA binding proteins in such a screen, although in vitro
binding studies indicated that the same DNA binding proteins present in bean cell
extracts could bind both synthetic H-box oligomers and CHS promoter segments
harboring H-box /,84 We therefore resorted to classical purification to isolate H-
box binding factors from whole cell extracts of suspension cultured bean cells. A
combination of ion-exchange and DNA affinity chromatography on immobilized
H-box oligomers resulted in the purification of two distinct fractions, KAP I and
KAP2, that both bound the CHS15 H-box. R4 The two fractions were resolved from
one another by a KCl gradient on the DNA affinity column, suggesting that they
have different binding affinity for the H-box sequence. SDS-PAGE analysis of
KAP I revealed a single polypeptide of Mr 97,000. KAP 2 preparations consisted
of two polypeptides, of Mr 76,000 and 56,000, although the higher Mr polypep-
tide was not present in all preparations. The 56,000 Mr component of KAP 2 was
purified from approximately 30 kg of bean cells, yielding sufficient amounts for
microsequencing of internal tryptic peptides. Oligonucleotide primers were
designed based on the peptide sequence information, and KAP2 cDNA clones
obtained by standard library screening approaches (W Lindsay, C. Lamb and R.A.
Dixon, unpublished results). KAP 2 cDNAs encode a polypeptide of Mr 77,000,
that shows significant sequence identity (approximately 30% overall, spread over
the whole sequence) to human and mouse Ku autoantigen (W Lindsay, F. McAl-
ister, C. Lamb and R.A. Dixon, unpublished results). Ku autoantigen, which is
found in the serum of humans with the autoimmune disease lupus erythemato-
sus, has been ascribed a role as a transcriptional regulator. 91
In most legume species we have tested, KAP2 is encoded by a single gene
(F. McAlister, W Lindsay, C. Lamb and R.A. Dixon, unpublished results). Its
protein and transcripts are both oflow abundance. Both KAPI and KAP2 appear
to be localized primarily in the cytoplasm in unelicited cells, but are found in the
nuclear fraction following elicitation,84 consistent with a role in elicitor-mediated
transcriptional activation of CHS. Preliminary results indicate that baculovirus-
expressed KAP2 can activate transcription from a minimal rice PAL transcription
start site fused to the CHS15 H-box region in an in vitro transcription system (Q.
Zhu, F. McAlister, R.A. Dixon and C. Lamb, unpublished results). Studies are in
progress to confirm the nuclear localization of KAP2 following elicitation, by
using immunocytochemical and fluorescent gene-fusion approaches. The phos-
phorylation status of KAP2 does not affect its DNA binding activity,84 but may
impact its ability to interact with KAPI and possibly other transcription factors.
It may also affect nuclear transport of KAP2.
Screening a cDNA expression library with the region of the CHS15 pro-
moter containing both the H-box I and the G-box led to the cloning of a DNA
binding protein, G/HBFl, that binds both the G-box and H-box (in the context
of the CHS15 promoter).92 G/HBF I is a member of the basic leucine zipper family
of transcription factors, and its closest homo logs in plants are a member of the
opaque family of factors that control zein gene expression in maize, and the
CPRF2 factor that binds the G-box in the parsley CHS promoter. 93
Western blot analysis revealed that elicitation or infection result in the rapid
appearance of a phosphorylated form of GIHBF -1. 92 Use of E. coli expressed
G/HBFl as a substrate in in vitro kinase assays showed that this is associated with
the rapid induction of the protein kinase activity that phosphorylates G/HBF-1.92
To isolate the protein kinase involved in signal transduction to G/HBF -I, we cloned
G/HBF-l as a fusion to the gal4 DNA binding domain as a bait in the yeast two-
hybrid system, and used this to screen a soybean cell culture cDNA library (from
elicited and unelicited cultures) fused to the gal4 transactivation domain for
expressed proteins interacting with G/HBF-1. This screen resulted in the
identification of a calmodulin domain protein kinase (CDPK) that specifically
interacts with G/HBFl (P. Canovas, J.T. Reddy, C. Lamb and R.A. Dixon, unpub-
lished results). CDPKs are calcium dependent, and elicitor-mediated CHS tran-
scription also requires calcium. The two-hybrid screen also identified a protein
phosphatase that might be involved in G/HBF -1 regulation. Baculovirus-expressed
G/HBF-l kinase phosphorylates G/HBF-l in vitro (P. Canovas, S. Hedrick, R.A.
Dixon and C. Lamb, unpublished results). Studies are now in progress to address
functionally the role of G/HBF-l kinase by transgenic approaches.
Transcription of the downstream isoflavone reductase gene is slightly
delayed compared to that of PAL or CHS genes in elicited alfalfa cell suspen-
sions, consistent with its responding to different transcriptional regulators. 48 This
idea is confirmed by the results of in vivo (functional) and in vitro analyses of
the alfalfa isoflavone reductase promoter (S. Yin, B. Miao and N.L Paiva, per-
sonal communication), which appears to be regulated by transcription factor(s)
recognizing sequences not involved in CHS regulation. Availability of alfalfa
CHR and isoflavone O-methyltransferase gene promoters, coupled with genetic
approaches in model legumes, will now enable us to address whether common
or disparate transcriptional regulators are involved in coordinating the transcrip-
tional activation of the isoflavonoid pathway.
PHARMACOLOGICAL APPROACHES TO THE

DISSECTION OF SIGNAL PATHWAYS FOR
ISOFLAVONOID BIOSYNTHESIS
The above studies have identified a protein kinase cascade tightly linked to
the transcriptional activation of genes involved in the elicitor/infection-induced
synthesis of isoftavonoid phytoalexins. However, the induced hypersensitive
defense response in plant cells is a multi component one, involving rapid pro-
duction of active oxygen species (the oxidative burst), induction of anti-oxidative
enzymes, metabolic changes including cell wall reinforcement and phytoalexin
accumulation, production of various pathogenesis-related proteins, and genera-
tion of long distance signals for activation of systemic defense responses. 1,83,94,95
Signal transduction pathways must operate intracellularly (from elicitor reception
to gene activation) and intercellularly, the latter at both short distances (to signal
metabolic changes to the periphery of the hypersensitive lesion), and long dis-
tances (for activation of systemic responses). Understanding the complexities of
these various signal pathways and at what levels they diverge or converge is a
major unresolved problem in molecular plant pathology. We have used a some-
what simplistic pharmacological approach to attempt to dissect signaling path-
ways in a model soybean cell culture system that exhibits a strong oxidative burst,
anti oxidative defenses, and isoftavonoid accumulation in an avirulence gene
dependent manner.
Exposure of suspension cultures of soybean cv Williams 82 to Pseudomonas
syringae pv glycinea (Psg) harboring the avrA avirulence gene, which is recog-
nized by the Rpt resistance gene in Williams 82, results in an oxidative burst that
generates maximum concentrations of hydrogen peroxide approximately 4 to 6
hours after exposure to the bacteria. 96 A burst with similar kinetics is observed in
response to yeast elicitor. However, the large oxidative burst is not observed if the
soybean cells are exposed to Psg harboring the avrC avirulence gene, the corre-
sponding resistance gene for which is absent from cv Williams 82. Likewise, accu-
mulation of the isoftavonoid phytoalexin glyceollin is observed in cultures exposed
to Psg: avrA or yeast elicitor, but not to Psg: avrc. 97
The oxidative burst in cells exposed to Psg: avrA is inhibited by the protein
kinase inhibitor K252a, and the burst can be induced, in the absence of bacteria,
by the protein phosphatase inhibitor cantharidin, indicating the operation of
a finely poised protein phosphorylation/dephosporylation cascade controlling the
activity of the membrane associated NADPH oxidase responsible for active
oxygen generation. 96 The oxidase, and subsequent production of hydrogen
peroxide, is inhibited by the suicide inhibitor diphenylene iodonium (DPI).
Hydrogen peroxide generated in the oxidative b,urst acts as a diffusible
signal for induction of genes encoding anti oxidative enzymes such as glutathione
S-transferase,98 but very high concentrations of hydrogen peroxide only weakly
I Elicitor I Serine protease

inhibitors
~ ~
1 Receptor HriR~e:::c::ep~to:;;:rrl-----+I------------
[X]
!<t>
~~[(:)l
[(:-)l~~do kinase cascade
~
Phosphorylation
of GIHBF-l and
other transcription
factors
Salicylic .
acid PhenYlalamne)
t PAL
L - Cinnamic 'lcid enzymes
L 1 Downstream phenylpropanoid/
isoflavonoid pathway enzymes
I Isoflavonoid phytoalexins I
Figure 8. Scheme for signal transduction pathways for induction of the oxidative burst and
isoflavonoid phytoalexin accumulation in soybean cells. See text for details.
induce genes that encode phytoalexin biosynthetic enzymes such as PAL and
CHS. Treatment with DPI completely abolishes glyceollin accumulation
following exposure of soybean cells to Psg: avrA or yeast elicitor, suggesting that
hydrogen peroxide may act as a signal for phytoalexin induction. 97
Several serine protease inhibitors, including phenylmethylsulfonyl fluoride
(PM SF) and diisopropylfluorophosphate (DFP), although themselves inactive
as inducers of the oxidative burst and phytoalexin production, greatly potentiate
these responses if added along with Psg: avrA or yeast elicitor, and potentiation of
the oxidative burst by PMSF or DFP is completely inhibited by K252a or
DPI, suggesting that the potentiated burst operates through the kinase-mediated
pathway leading to stimulation of NADPH oxidase. 97 Interestingly, PMSF or
DFP treatment also results in a strong oxidative burst in soybean cells exposed to
Psg: avrC, and this is again blocked by DP!. However, this burst does not result in
glyceollin accumulation. Thus, generation of hydrogen peroxide is necessary, but
not in itself sufficient, for induction of isoflavonoid phytoalexin accumulation.
Clearly, a signal for isoflavonoid induction is generated in the Psg: avrA interac-
tion that is not generated in response to Psg:avrC, even when a synergistic inter-
action between Psg: avrC and PMSF or DFP leads to a strong oxidative burst. A
scheme summarizing potential signal pathways for the oxidative burst and
isoflavonoid phytoalexin induction is shown in Figure 8.
Treatment of cell suspension cultures of soybean, alfalfa, or tobacco with
the macromolecular elicitor cryptogein (a 10.3 kDa peptide from Phytophthora
cryptogea) results in the release of low molecular mass (500-2,OOODa) endoge-
nous elicitors that can pass through a dialysis membrane into the surrounding
culture medium. 99 The active compounds induce PAL and IFR transcripts and
enzymatic activity, accumulation of glyceollin, and an oxidative burst in soybean
cells. Release of the factors from soybean cells is inhibited by the RNA synthe-
sis inhibitor actinomycin D, but not by DP!. The activity of the factors released
from tobacco cells in response to cryptogein, or from alfalfa cells in response
to yeast elicitor, is unaffected by pre-treatment with sodium periodate or pro-
teinase K. In contrast, treatment of the factor(s) from soybean cells with pro-
teinase K or trypsin destroys its ability to induce glyceollin, suggesting that it
may be a small peptide. 99 The diffusible elicitors have now been extensively
purified, but structural characterization has been hampered by their extremely low
abundance, in spite of their high biological activity. This potency suggests that
they may indeed be important signal molecules for the intercellular transmission
of the elicitation response.
CONCLUSIONS
Several important advances have been made during the last few years in
our understanding of the molecular biology of isoflavonoid biosynthesis and its
control, and of the important effects of dietary isoflavonoids on human health.
The challenge for the future will be to devise strategies for manipulating the levels
of isoflavonoid compounds in transgenic plants, not only in legumes in which
these compounds occur naturally, but also in other plant species including major
food crops. The aims of these studies will be to increase either natural plant
disease resistance or dietary intake of health-promoting phytochemicals.
ACKNOWLEDGMENTS
We acknowledge the work done by several past and current postdocs,

faculty and visitors in the Noble Foundation and Salk Institute laboratories in
areas related to those described in the above article. Special thanks go to Drs G.
Murray Balance, Robert Gonzales, Wolfgang Droge-Laser, Susie Hedrick,
William Lindsay, J.T. Reddy, and Christopher Steele. We thank Cuc Ly for
artwork. Parts of this article are based on a larger article from the corresponding
author, to appear in Vol. I of "Comprehensive Natural Products Chemistry", Else-

vier Publishing, Co. The work in the author's laboratories was supported by the
Samuel Roberts Noble Foundation.
REFERENCES
1. DIXON, R.A., HARRISON, M.J. 1990. Activation, structure and organization of genes
involved in microbial defense in plants. Adv. Genet. 28: 165-234.
2. DIXON, R.A., PAIVA, N.L. 1995. Stress-induced phenylpropanoid metabolism. Plant
Cell 7: 1085-1097.
3. REDMOND, 1.w., BATLEY, M., DJORDJEVIC, M.A., INNES, R.W, KUEMPEL, P.L.,
ROLFE, B.G. 1986. Flavones induce expression of nodulation genes in Rhizobium. Nature
323: 632-636.
4. COWARD, L., BARNES, N.e., SETCHELL, K.D.R., BARNES, S. 1993. Genistein,
daidzein, and their ~-glycoside conjugates: antitumor isoflavones in soybean foods from
American and Asian diets. J. Agric. Food Chern. 41: 1961-1967.
5. TAHARA, S., IBRAHIM, R.K. 1995. Prenylated isoflavonoids-an update. Phytochem-
istry 38: 1073-1094.
6. LEE, S.-J., WOOD, A.R., MAIER, e.G.-A., DICON, R.A., MABRY, T.J. 1998. Preny-
lated flavonoids from Mac/ura pomifera. Phytochemistry, in press:
7. INGHAM, lL., TAHARA, S., HARBORNE, lB. 1983. Fungitoxic isoflavones from
Lupinus albus and other Lupinus species. Z. Naturforsch. 38c: 194-200.
8. VANETTEN, H., MANSFIELD, 1.W, BAILEY, 1.A., FARMER, E.E. 1994. Two classes
of plant antibiotics: phytoalexins versus "phytoanticipins". Plant Cell 6: 1191-1192.
9. DICON, R.A. 1986. The phytoalexin response: elicitation, signalling and the control of
host gene expression. BioI. Rev. 61: 239-291.
10. VANETTEN, H.D., MATTHEWS, D.E., MATTHEWS, P.S. 1989. Phytoalexin
detoxification: importance for pathogenicity and practical implications. Annu. Rev.
Phytopathol. 27: 143-164.
11. BHATTACHARYYA, M.K., WARD, E.WB. 1987. Temperature-induced susceptibility of
soybeans to Phytophthora megasperma f.sp. glycinea: phenylalanine ammonia-lyase and
glyceollin in the host; growth and glyceollin I sensitivity of the pathogen. Physiol. Mol.
Plant Pathol. 31: 407--419.
12. KISTLER, H.C., VANETTEN, H.D. 1984. Regulation of pis at in demethylation in Nectria
haematococca and its influence on pisatin tolerance and virulence. 1. Gen. Microbiol. 130:
2605-2613.
13. GLAZEBROOK, 1., AUSUBEL, EM. 1994. Isolation ofphytoalexin-deficient mutants of
Arabidopsis thaliana and characterization of their interactions with bacterial pathogens.
Proc. Natl. Acad. Sci. USA 91: 8955-8959.
14. MORRIS, P.E, WARD, E.WB. 1992. Chemoattraction of zoospores of the soybean
pathogen, Phytophthora sojae by isoflavones. Physiol. Mol. Plant Pathol. 40: 17-22.
15. RUAN, Y, KOTRAIAH, V, STRANEY, D.e. 1995. Flavonoids stimulate spore germi-
nation in Fusarium solani pathogenic on legumes in a manner sensitive to inhibitors of
cAMP-dependent protein kinase. Mol. Plant-Microbe Interact. 8: 929-938.
16. DENARIE, 1., DEBELLE, E, PROME, J.-e. 1996. Rhizobium lipo-chitooligosaccharide
nodulation factors: Signaling molecules mediating recognition and morphogenesis. Annu.
Rev. Biochem. 65: 503-535.
17. PETERS, N.K., FROST, J.w., LONG, S.R. 1986. A plant flavone, luteolin, induces
expression of Rhizobium meliloti nodulation genes. Science 233: 977-980.
18. KOSSLAK, R.M., BOOKLAND, R., BARKEI, 1., PAAREN, H.E., APPLEBAUM, E.R.
1987. Induction of Bradyrhizobium japonicum common nod genes by isoflavones isolated
from Glycine max. Proc. Natl. Acad. Sci. USA 84: 7428-7432.
19. ZHANG, E, SMITH, D.L. 1996. Genistein accumulation in soybean (Glycine max [L.]
Merr.) root systems under suboptimal root zone temperatures. 1. Exp. Bot. 47: 785-792.
20. SHUTT, D.A. 1976. The effects of plant oestrogens on animal reproduction. Endeavour
75: 110-113.
21. GILDERSLEEVE, R.R., SMITH, G.R., PEMBERTON, 1.1., GUILBERT, e.L. 1991.
Detection of isoflavone in seedling subterranean clover. Crop Sci. 31: 889-892.
22. LEOPOLD, A.S., ERWIN, M., OH, 1., BROWNING, B. 1976. Phytoestrogens: adverse
effects on reproduction in California quail. Science 191: 98-100.
23. MARTIN, M.E., HAOURIGUI, M., PELISSERO, e., BENASSAYAG, e., NUNEZ, E.A.
1996. Interactions between phytoestrogens and human sex steroid binding protein. Life
Sci. 58: 429-436.
24. WANG, w., TANAKA, Y, HAN, Z., HIGUCHI, e.M. 1995. Proliferative response of
mammary glandular tissue to formononetin. Nutr. Cancer 23: 131-140.
25. SETCHELL, K.D.R., BORRIELLO, S.P, HULME, P, KIRK, D.N., AXELSON, M.
1984. Nonsteroidal estrogens of dietary origin: possible roles in hormone-dependent
disease. Am. 1. C1in. Nutr. 40: 569-578.
26. MAIER, e.G.A., CHAPMAN, K.D., SMITH, D.W. 1995. Differential estrogenic activi-
ties of male and female plant extracts from two dioecious species. Plant Sci. 109: 31-43.
27. LEE, S.-1., CHUNG, H.-Y, MAIER, e.G.-A., WOOD, A.R., DICON, R.A., MABRY,
TJ. 1998. Estrogenic flavonoids from Artemesia vulgaris L. 1. Agri. Food. Chem. 46:
3325-3329.
28. ADLERCREUTZ, H., HONJO, H., HIGASHI, A., FOTSIS, T, HAMALLAINEN, E.,
HASEGAWA, T, OKADA, H. 1991. Urinary excretion of lignans and isoflavonoid phy-
to estrogens in Japanese men and women consuming a traditional Japanese diet. Am. 1.
Clin. Nutr. 54: 1093-1100.
29. LEE, H.P, GOURLEY, L., DUFFY, S.w., ESTEVE, 1., LEE, 1., DAY, N.E. 1991. Dietary
effects on breast-cancer risk in Singapore. Lancet 337: 1197-1200.
30. BARNES, S., GRUBBS, e., SETCHELL, K.D.R., CARLSON, 1. 1990. Soybeans inhibit
mammary tumors in models of breast cancer. pp. 239-253 in Mutagens and Carcinogens
in the Diet (M.W. Pariza, ed.) Wiley-Liss, Inc, New York.
3 I. LAMARTINIERE, e.A., MOORE, 1., HOLLAND, M., BARNES, S. 1995. Neonatal
genistein chcmoprevents mammary cancer. Proc. Soc. Exp. BioI. Med. 208: 120-123.
32. YANAGIHARA, K., ITO, A., TOGE, T, NUMOTO, M. 1993. Antiproliferative effects
of isoflavones on human cancer cell lines established from the gastrointestinal tract.
Cancer Res. 53: 5815-582 I.
33. CASSIDY, A., BINGHAM, S., SETCHELL, K.D. 1994. Biological effects of a diet of
soy protein rich in isoflavones on the menstrual cycle of premenopausal women. Am. 1.
Clin. Nutr. 60: 333-340.
34. FRANKE, A.A., L.1., e., WANG, w., SHI, C.Y. 1998. HPLC analysis of isoflavonoids
and other phenolic agents from foods and from human fluids. Proc. Soc. Exp. BioI. Med.
217: 263-288.
35. AKIYAMA, T, ISHIDA, 1., NAKAGAWA, S., OGAWARA, H., WATANABE, S., ITOH,
N., SHIBUYA, M., FUKAMI, Y 1987. Genistein, a specific inhibitor oftyrosine-spccific
protein kinases. 1. BioI. Chern. 262: 5592-5595.
36. TAKANO, T, TAKADA, K., TADA, H., NISHIYAMA, S., AMINO, N. 1993.
Gcnistein, a tyrosine kinase inhibitor, blocks the cell cycle progression but not Ca'+ influx
induced by BAY K8644 in FRTL-5 cells. Biochem. Biophys. Res. Commun. 190:
801-807.
37. HUANG, 1., NASR, M., KIM, Y., MATTHEWS, H.R. 1992. Genistein inhibits protein
histidine kinase. 1. BioI. Chern. 267: 15511-15515.
38. UCHUN, EM., EVANS, W.E., FORSYTH, C.J., WADDICK, K.G., AHLGREN, L.T,
CHELSTROM, L.M., BURKHARDT, A., BOLEN, 1., MYERS, D.E. 1995. Biotherapy
of B-cell precursor leukemia by targeting genistein to CD 19-associated tyrosine kinases.
Science 267: 886-891.
39. ADLERCREUTZ, H., HOCKERSTEDT, K., BLOIGU, S., HAMALAINEN, E.,
FOTSIS, T, OLLUS, A. 1987. Effect of dietary components, including lignans and phy-
toestrogens, on enterohepatic circulation and liver metabolism of estrogens and sex
hormone binding globulin (SHBG). 1. Steroid Biochem. 27: 1135-1144.
40. MESSINA, M.J., PERSKY, V, SETCHELL, K.D.R., BARNES, S. 1994. Soy intake and
cancer risk: A review of the in vitro and in vivo data. Nutr. Cancer 21: 113-131.
41. KNIGHT, D.C., EDEN, lA. 1996. A review of the clinical effects of phytoestrogens.
Obst. Gynecol. 87: 897-904.
42. WISEMAN, H. 1996. Role of dietary phyto-estrogens in the protection against cancer
and heart disease. Biochem. Soc. Trans. 24: 795-800.
43. NIESBACH-KLOSGEN, u., BARZEN, E., BERHNARDT, 1., ROHDE, w., SCHWARZ-
SOMMER, Z., REIF, H.J., WIENAND, u., SAEDLER, H. 1987. Chalcone synthase
genes in plants: a tool to study evolutionary relationships. 1. Mol. Evol. 26: 213-225.
44. AKADA, S., DUBE, S.K. 1995. Organization of soybean chalcone synthase gene
clusters and characterization of a new member of the family. Plant Mol. BioI. 29:
189-199.
45. AN, c., ICHINOSE, Y., YAMADA, T, TANAKA, Y., SHIRAISHI, T, OKU, H. 1993.
Organization of the genes encoding chalcone synthase in Pisum sativum. Plant Mol. BioI.
21: 789-803.
46. ARIOLI, T, HOWLES, P.A., WEINMAN, J.J., ROLFE, B.G. 1994. In Trifolium subter-
raneum, chalcone synthase is encoded by a multigene family. Gene 138: 79-86.
47. JUNGHANS, H., DALKIN, K., DIXON, R.A. 1993. Stress responses in alfalfa (Med-
icago sativa 1.) XV Characterization and expression patterns of members of a subset of
the chalcone synthase multigene family. Plant Mol. BioI. 22: 239-253.
48. FEINBAUM, R.L., AUSUBEL, EM. 1992. Transcriptional regulation of the Arabidopsis
thaliana chalcone synthase gene. Mol. Cell. BioI. 8: 1985-1992.
49. RYDER, TB., HEDRICK, S.A., BELL, IN., LIANG, X., CLOUSE, S.D., LAMB,
C.J. 1987. Organization and differential activation of a gene family encoding the plant
defense enzyme chalcone synthase in Phaseolus vulgaris. Mol. Gen. Genet. 210:
219-233.
50. AYABE, S., UDAGAWA, A., FURUYA, T. 1988. Stimulation of chalcone synthase activ-
ity by yeast extract in cultured Glycyrrhiza echinata cells and 5-deoxyflavanone forma-
tion by isolated protoplasts. Plant Cell Rep. 7: 35-38.
51. NAKAJIMA, 0., SHIBUYA, M., HAKAMATSUKA, T, NOGUCHI, H., EBIZUKA, Y.,
SANKAWA, U. 1996. cDNA and genomic DNA clones of chalcone synthase from Puer-
aria lobata. BioI. Pharmacol. Bulletin 19: 71-76.
52. DEWICK, P.M., STEELE, M.I 1982. Biosynthesis of the phytoalexin phaseollin in
Phaseolus vulgaris. Phytochemistry 21: 1599-1603.
53. AYABE, S.I., UDAGAWA, A., FURUYA, T 1988. NAD(P)H-dependent 6'-deoxychal-
cone synthase activity in Glycyrrhiza echinata cells induced by yeast extract. Arch.
Biochem. Biophys. 261: 458-462.
54. WELLE, R., GRISEBACH, H. 1988. Isolation of a novel NADPH-dependent reductase
which coacts with chalcone synthase in the biosynthesis of6'-deoxychalcone. FEBS Lett.
236: 221-225.
55. WELLE, R., SCHRODER, G., SCHILTZ, E., GRISEBACH, H., SCHRODER, 1. 1991.
Induced plant responses to pathogen attack. Analysis and heterologous expression of the
key enzyme in the biosynthesis of phytoalexins in soybean (Glycine max L. Merr. cv.
Harosoy 63). Eur. J Biochem. 196: 423-430.
56. BALLANCE, G.M., DIXON, R.A. 1994. Medicago sativa cDNAs encoding chalcone
reductase. Plant Physiol. 107: 1027-1028.
57. SALLAUD, c., EL-TURK, 1., BIGARRE, L., SEVIN, H., WELLE, R., ESNAULT, R.
1995. Nucleotide sequences of three chalcone reductase genes from alfalfa. Plant Physiol.
108: 869-870.
58. SALLAUD, c., ELTURK, 1., BREDA, c., BUFFARD, D., DEKOZAK, I., ESNAULT,
R., KONDOROSI, A. 1995. Differential expression of cDNA coding for chalcone reduc-
tase, a key enzyme of the 5-deoxyflavonoid pathway, under various stress conditions in
Medicago sativa. Plant Sci. 109: 179-190.
59. AKASHI, T, FURUNO, T, FUTAMI, K., HONDA, M., TAKAHASHI, T, WELLE, R.,
AYABE, S. 1996. A cDNA for polyketide reductase (Accession No. D83718) that cat-
alyzes the formation of 6'-deoxychalconc from cultured Glycyrrhiza echinata L. cells.
Plant Physiol. III: 347-348.
60. HAKAMATSUKA, T, EBIZUKA, Y., SANKAWA, U. 1994. Pueraria lobata (kudzu
vine): in vitro culture and the production of isoflavonoids. pp. 336-400 in Biotechnology
in Agriculture and Forestry (Y. P. S Bajai, ed.), Springer-Verlag, Berlin.
61. TROPF, S., KARCHER, 8., SCHRODER, G., SCHRODER, 1. 1995. Reaction mecha-
nisms of homodimeric plant polyketide synthases (stilbene and chalcone synthase). A
single active site for the condensing reaction is sufficient for synthesis of stilbenes, chal-
cones, and 6'-deoxychalcones. 1. BioI. Chern. 270: 7922-7928.
62. NI, W, FAHRENDORF, T, BALLANCE, G.M., LAMB, C.J., DIXON, R.A. 1996. Stress
responses in alfalfa (Medicago sativa L.). XX. Transcriptional activation of phenyl-
propanoid pathway genes in elicitor-treated cell suspension cultures. Plant Mol. BioI. 30:
427-438.
63. HAGMANN, M., GRISEBACH, H. 1984. Enzymatic rearrangement of flavanone to
isoflavone. FEBS Lett. 175: 199-202.
64. HASHIM, M.F., HAKAMATSUKA, T, EBIZUKA, Y., SANKAWA, U 1990. Reaction
mechamism of oxidative rearrangement of flavanone in isoflavone biosynthesis. FEBS
Lett. 271: 219·-222.
65. HAKAMATSUKA, T, HASHIM, M.F., EBIZUKA, Y., SANKAWA, U 1991. P-450-
dependent oxidative rearrangement in isoflavone biosynthesis: reconstitution ofP-450 and
NADPH:P450 reductase. Tetrahedron 47: 5969-5978.
66. KOCHS, G., GRISEBACH, H. 1986. Enzymic synthesis ofisoflavones. FEBS Lett. 155:
311-318.
67. HAKAMATSUKA, T, MORI, K., ISHIDA, S., EBIZUKA, Y., SANKAWA, U 1998.
Purification of2-hydroxyisoflavone dehydratase from the cell cultures of Pueraria lobata.
Phytochemistry 49: 497-505.
68. MIZUTANI, M., WARD, E., OHTA, D. 1998. Cytochrome P450 superfamily in
Arabidopsis thaliana: isolation of cDNAs, differential expression, and RFLP mapping
of multiple cytochromcs P450. Plant Mol. BioI. 37: 39-52.
69. HINDERER, W, FLENTJE, U, BARZ, W 1987. Microsomal isoflavone X-and 3'-
hydroxylases from chickpea (Cicer arietinum L.) cell suspensions induced for ptero-
carpan phytoalexin formation. FEBS Lett. 214: 101-106.
70. DEWICK, P.M., MARTIN, M. 1979. Biosynthesis ofpterocarpan, isoflavan and coumes-
tan metabolites of Medicago sativa: chalcone, isoflavone and isoflavanonc precursors.
71. KESSMANN, H., CHOUDHARY, A.D., DIXON, R.A. 1990. Stress responses in
alfalfa (Medicago sativa L.) III. Induction of medicarpin and cytochrome P450 enzyme
activities in elicitor-treated cell suspension cultures and protoplasts. Plant Cell Rep. 9:
38-41.
72. WONG, E., FRANCIS, C.M. 1968. Flavonoids in genotypes of Trifolium subterraneum-
I. The normal flavonoid pattern of the Geraldton variety. Phytochemistry 7: 2123-2129.
73. WENGENMAYER, H., EBEL, 1., GRISEBACH, H. 1974. Purification and properties of
a S-adenosylmethionine: isoflavone 4'-0-methyltransferase from cell suspension cultures
of Cicerarietinum L. Eur. 1. Biochem. 50: 135-143.
74. EDWARDS, R., DIXON, R.A. 1991. Isoflavone O-methyltransferase activities in elicitor-
treated cell suspension cultures of Medicago sativa. Phytochemistry 30: 2597·-2606.
75. HE, X.-Z., DIXON, R.A. 1996. Affinity chromatography, substrate/product specificity and
amino acid sequence analysis of an isoflavone O-methyltransferase from alfalfa (Med-
icago sativa L.). Arch. Biochem. Biophy. 336: 121-129.
76. DIXON, R.A., CHOUDHARY, A.D., DALKIN, K., EDWARDS, R., FAHRENDORF, T,
GOWRI, G., HARRISON, MJ., LAMB, C.J., LOAKE, GJ, MAXWELL, C.A., ORR,
1., PAIVA, N.L. 1992. Molecular biology of stress-induced phenylpropanoid biosynthe-
sis in alfalfa. pp. 91-138 in Phenolic Metabolism in Plants (H. A. Stafford, R.K., Ibrahim,
eds.), Plcnum Press, New York.
77. WU, Q., PREISIG, c.L., VANETTEN, H.D. 1998. Isolation of the cDNAs encoding
(+)6a-hydroxymaackiain 3-0-methyltransferase, the terminal step for the synthesis of the
phytoalexin pisatin in Pisum sativum. Plant Mol. BioI. 35: 551-560.
78. HE, X.-Z., REDDY, 1.T, DIXON, R.A. 1998. Stress responses in alfalfa (Medicago sativa
L.) XXII. cDNA cloning and characterization of an elicitor-inducible isoflavone 7-0-
methyItransferase. Plant Mol. BioI. 36: 43-54.
79. DIXON, R.A., HARRISON, MJ., PAIVA, N.L. 1995. The isoflavonoid phytoalexin
pathway: from enzymes to genes to transcription factors. Physiol. Plant. 93: 385-392.
80. HOLTON, TA., CORNISH, E.C. 1995. Genetics and biochemistry of anthocyanin
biosynthesis. Plant Cell 7: 1071-1083.
81. BENABEN, V, DUC, G., LEFEBVRE, V, HUGUET, T 1995. TE7, an inefficient
symbiotic mutant of Medicago truncatula Gaertn. cv jemalong. Plant Physiol. 107:
53-62.
82. LINDSAY, w.P., LAMB, CJ., DIXON, R.A. 1993. Microbial recognition and activation
of plant defense mechanisms. Trends Microbiol. I: 181-186.
83. LAMB, c., DIXON, R.A. 1997. The oxidative burst in plant disease resistance. Annu.
Rev. Plant Physiol. Plant Mol. BioI. 48: 251-275.
84. YU, L.M., LAMB, CJ., DIXON, R.A. 1993. Purification and biochemical characteriza-
tion of two proteins which bind to the H-box cis-clement implicated in transcriptional
activation of plant defcnse genes. Plant 1. 3: 805-816.
85. DRON, M., CLOUSE, S.D., DIXON, R.A., LAWTON, M.A., LAMB, C.J. 1988. Glu-
tathione and fungal elicitor regulation of a plant-defense gene promoter in electroporatcd
protoplasts. Proc. Natl. Acad. Sci. USA 85: 6738-6742.
86. HARRISON, M.1., CHOUDHARY, A.D., DUBERY, I., LAMB, C.J., DIXON, R.A. 1991.
Cis-elements and trans-acting factors for the quantitive expression of a bean chalcone
synthase gene promoter in electroporated alfalfa protoplasts. Plant Mol. BioI. 16:
877-890.
87. FAKTOR, 0., LOAKE, G., DIXON, R.A., LAMB, CJ. 1997. The G-box and H-box in a
39 bp region of a French bean chalcone synthase promotr constitute a tissue-specific reg-
ulatory element. Plant 1. II: 1105-1113.
88. HARRISON, M.1., LAWTON, M.A., LAMB, C.J., DIXON, R.A. 1991. Characterization
of a nuclcar protein which binds to three elements within the silencer region of a bean
chalcone synthase gene promoter. Proc. Natl. Acad. Sci., USA 88: 2515-2519.
89. LAWTON, M.A., CLOUSE, S.D., LAMB, C.J. 1990. Glutathione-elicited changes in
chromatin structure within the promoter of the defense gene chalcone synthase. Plant Cell
Rep. 8: 561-564.
90. ARIAS, J.A., DIXON, R.A., LAMB, C.J. 1993. Dissection of the functional architecture
of a plant defense gene promoter using a homologous in vitro transcription system. Plant
Cell 5: 485-496.
91. GIFFIN, W, TORRANCE, H., RODDA, DJ., PREFONTAINE, G.G., POPE, L.,
HACHE, R.J.G. 1996. Sequence-specific DNA binding by Ku autoantigen and its effects
on transcription. Nature 380: 265-268.
92. DROGE-LASER, W, KAISER, A., LINDSAY, WP., HALKIER, 8., LOAKE, G.A.,
DOERNER, P.W, DIXON, R.A., LAMB, CJ. 1997. Rapid stimulation of a soybean
protein-serine kinase that phosphorylates a novel blIP transcription factor, G/HBF-I, in
thc induction of early transcription-dependent defenses. EMBO 1. 16: 726-738.
93. ARMSTRONG, G.A., WEISSHAAR, 8., HAHLBROCK, K. 1992. Homodimeric and
heterodimeric leucine zipper proteins and nuclear factors from parsley recognize diverse
promoter elements with ACGT cores. Plant Cell 4: 525-537.
94. SOMSSICH, I.E., HAHLBROCK, K. 1998. Pathogen defence in plants-a paradigm of
biological complexity. Trends Plant Sci. 3: 86-90.
95. RYALS, 1., LAWTON, K.A., DELANEY, T.P., FRIEDRICH, L., KESSMANN, H.,
NEUENSCHWANDER, u., UKNES, S., VERNOOIJ, 8., WEYMANN, K. 1995. Signal
transduction in systemic acquired resistance. Proc. Natl. Acad. Sci. USA 92: 4202-4205.
96. SHIRASU, K., NAKAJIMA, H., RAJASEKHAR, Y.K., DIXON, R.A., LAMB, CJ. 1997.
Salicylic acid potentiates an agonist-dependent gain control that amplifies pathogen
signals in the activation of defense mechanisms. Plant Cell 9: 261-270.
97. GUO, l.-J., LAMB, C., DIXON, R.A. 1998. Potentiation of the oxidative burst and
isoflavonoid phytoalexin accumulation by serine protease inhibitors. Plant Physiol., in
press:
98. LEVINE, A., TENHAKEN, R., DIXON, R.A., LAMB, CJ. 1994. H,O, from the oxida-
tive burst orchestrates the plant hypersensitive disease resistance response as a local
trigger of programmed cell death and a diffusible inducer of cellular protectant genes.
Cell 79: 583-593.
99. GUO, l.-J., LAMB, c., DIXON, R.A. 1997. Release and biological activity of diffusible
signal compounds from elicited plant cells. J. Plant Physiol. 151: 699-710.
Chapter Seven
MEDICINAL POTENTIAL AND BIOSYNTHESIS

OF PLANT COUMARINS
Ulrich Matern
Institute of Pharmaceutical Biology

Philipps-University
Deutschhausstrasse 17 A
D-35037 Marburg, Germany
Introduction ................................................... 161

Current Aspects of Bioactivity .................................... 163
Biosynthesis of Coumarins ....................................... 166
Simple Coumarins ........................................... 166
Polyoxygenated and Alkylated Simple Coumarins .................. 168
Psoralen ................................................... 170
Oxygenated Psoralens ........................................ 176
Conclusions ................................................... 176
INTRODUCTION
Coumarins comprise a large group of natural compounds isolated from

many plant or microbial sources, and novel derivatives are still being reported
each year. 1.2 Their classification is based on the 2H-benzopyrane-2-one core struc-
ture, which almost always bears an oxygen in position 7 as in umbelliferone (Fig.
1). Despite the wide taxonomic distribution of these metabolites, the mechanism
of coumarin biosynthesis has been resolved only partially, and few in vitro studies
had addressed this problem till the early eighties. 3-4 The situation changed
markedly, however, when the activation of coumarin biosynthesis in response to
fungal challenge or other stress conditions was reported from various plants and
plant cell cultures. 5- 21 Induced cell cultures of the Apiaceae, in particular, paved
161
162 U. MATERN
7 - Methylesculin
Umbelliferone R1 =OH; R2 =H
Esculetin R1 = R2 =OH
Robustic acid
OMe
O~
Phellopterin
Figure I. Coumarins of medicinal or ecological potential.

MEDICINAL POTENTIAL AND BIOSYNTHESIS OF PLANT COUMARINS 163
the road to detailed molecular and regulatory studies, and significant progress
has been accomplished recently. Furthermore, it has long been known that
some coumarins show remarkable bioactivities, e.g. the anticoagulant effect of
4-hydroxycoumarins or the anti proliferative and phototoxic action of linear
furanocoumarins as well as the inhibitory action on 5-lipoxygenases. 22 - 24 More
recently, new and surprisingly diverse activities have been ascribed to other
coumarins, some of which have received public attention and may serve as lead
structures in drug development. A few such examples will be considered before
briefly reviewing some recent developments in the research of biosynthesis.
CURRENT ASPECTS OF BIOACTIVITY
Bioactivity guided screenings of plant extracts for anti-HIV activity under

the auspices of the U.S. National Cancer Institute (NCI) identified several novel
coumarins from the tropical rainforest tree Calophyllum lanigerum, Guttiferae,
with strong inhibitory activity even to those HIV-I strains which were resistant
to nucleoside or non-nucleoside anti-HIV drugs."f2 5 These compounds were
identified as angular dihydropyranocoumarins and designated as calanolides
(Fig. 2).26-28 A number of derivatives was prepared and revealed that the stereo-
configuration of substituents at positions 10, II and 12 of the dihydropyrane-ring
was crucial for HIV-inhibition, whereas the saturation of the 7,8-double bond or
replacement of the 12-hydroxy by an amino-function (Fig. 2) did not significantly
affect the activity. Also the structure of the substituent at position 4 of the
coumarin-ring appears irrelevant, since the inophyllums or cordatolides isolated
from related species and bearing an aromatic and methyl group, respectively,
turned out as effective?9.3o Retroviruses like HIV-I depend essentially on three
enzyme activities for replication: reverse transcriptase copies the viral RNA to
DNA, a protease digests the polycistronic primary product of translation to func-
tional proteins, and an integrase inserts the viral DNA copy into the host genome.
The calanolides inhibit the reverse transcriptase (Fig. 2), and kinetic analysis of
calanolide A revealed a complex mode of inhibition involving two binding sites. 31
Binding at one site was competitive to either deoxynucleotide triphosphates or
template/primer, while binding at the other site was noncompetitive. Overall, the
data suggested that calanolide A binds near the active site of reverse transcrip-
tase, thus interfering with nucleotide binding, and additionally in the region of
the pyrophosphate binding site.31
Some synthetic coumarins appear to be capable also of inhibiting HIV pro-
tease, and a prominent example from the NCI catalogue, a tetrameric 4-
hydroxycoumarin (Fig. 3), additionally showed significant inhibition of HIV-I
integrase activity.32 Chemical derivatization of this compound as well as molec-
ular modeling studies have been conducted and revealed that at least two of the
4-hydroxycoumarin units are required for integrase inhibition, and the position
164 U. MATERN
( + ) - Calanolide A
)?:~
~~n~
~ 0
(:t
••'
110
,.............. ~
11
12
OH
12 - Oxocalanolide A ( - ) - Calanolide A Calanolide F
ECso 0.15 11M inactive EC so 2.8 11M

IC so 9.3 11M IC 50 12.7 11M
Figure 2. Structures and bioactivities of calanolides. The anti-HIV-I activity (EC so ) and the
cell toxicity (lC so ) are indicated.
as well as the spatial orientation of the two hydroxy groups in conjunction with
the two pyrone-carbonyls are essential components of the pharmacophor. 32 .33 The
integration of the viral DNA copy into the host genome proceeds in three steps,
beginning with the recession of the 3' end followed by cleavage of the host DNA,
strand insertion into the host, and finally sealing of the nicks by DNA ligase.
Retroviral integrases contain three conserved acidic residues (Asp, Asp, Glu),
which are critical for the nuclease activity; these residues are assumed to facili-
tate the coordination with divalent metal ions and the phosphodiester backbone
during the 3' recession of the viral DNA copy. Accordingly, it was proposed that
the oxygens of the 4-hydroxycoumarin rings become coordinated to the essential
metal ions and thereby disrupt their binding to the enzyme or enzyme-DNA
complex. 32 .33 Taken together, coumarins are now being employed to inhibit all
three pivotal enzyme activities of HIV-I replication, the reverse transcriptase, the
protease, as well as the integrase.
IG 50 = 1.5j..1M IG 50 = 43,4 ± 23,7 j..IM
Figure 3. Synthetic coumarin inhibitors of HI V-I integrase. Modification of the tetrameric (left)
to the dimeric 4-hydroxycoumarin (right) considerably reduced the inhibitory efficiency (lC,o).
Four other recent studies deserve mentioning, in which new applications of

coumarins or new concepts of coumarin action have been proclaimed. Natural
linear furanocoumarins (psoralens) possess photosensitizing and antiproliferative
activities, and compounds such as 8-methoxypsoralen are being used in combi-
nation with ultraviolet irradiation for the treatment of skin disorders (PUVA
therapy).23.34 Serious side effects from this kind of treatment are the induction of
skin erythema and the risk of mutation. In an attempt to modulate the two activ-
ities, furanocoumarin isosters were synthesized, yielding some linear pyrrolo-
coumarins and 8-azapsoralens with strong antiproliferative effects which showed
no or much lower photomutagenic activity than 8-methoxypsoralen under irradi-
ation. 35 These compounds are under clinical evaluation. Besides the inhibition of
lipoxygenase activity in rat platelets and leukocytes, esculetin (Fig. I) is known
to exert antiproliferative effects and is thereby recommended to improve the cap-
illary blood flow. Based on a new assay system, Huang et al. proposed recently
that the anti proliferative activity of esculetin is due to the alteration of transduc-
tion signals that lead to DNA synthesis rather than to the inhibition of Iipoxyge-
nase or cyc100xygenase activities. 36 The receptors for growth factors such as
PDGF and EGF are protein kinases that directly affect the rate of nuclear tran-
scription and proliferation. Esculetin inhibited protein tyrosine kinase but not
protein kinase C activitity, indicating that it may act partly through inhibition of
protein kinase to reduce smooth muscle cell proliferation. 36 It is noteworthy in
this context that a plant hormone-like activity was proposed for 7-methylesculin
166 U. MATERN
(7 -O-methylesculetin 6-0-glucoside) (Fig. I) which influenced the rate of cell

division in tobacco leafy galls induced by Rhodococcus or Agrobacterium Spp.37
Along the same lines, the prenylated coumarin robustic acid (Fig. I) was recently
proposed to be an effective inhibitor of eucaryotic signal-regulated protein
kinases. 38 This compound strongly inhibited rat liver cAMP-dependent protein
kinase (cAK) but neither rat brain calcium- and phospholipid-dependent (PKC)
nor wheat embryo calcium-dependent (CDPK) protein kinases nor avian gizzard
calmodulin-dependent myosin light chain kinase (MLCK). It was suggested that
the lack of cAK in plants could make this enzyme an appropriate target for non-
lethal plant defense compounds: elevated cAMP concentration is a starvation
signal leading to cAK activation, and, therefore, the inhibition of cAK would
block the cAK-mediated aggression of fungal pathogens, insects, or other herbi-
vores. 38 Finally, the linear furanocoumarin phellopterin (Fig. I) was recently
shown in vitro to bind specifically and with high affinity to benzodiazepine recep-
tors in rat cortical membrane. 39 If this effect and the previously mentioned
concepts can be confirmed, the structures of these coumarins may lead to the
development of compounds with medicinal or ecological value.
BIOSYNTHESIS OF COUMARINS
Simple Coumarins
Coumarins may be derived from the polyketide or the phenylpropanoid

pathway. Some plant coumarins, in particular those substituted at carbon 4 like
the Gerbera 4-hydroxy-5-methylcoumarin and the microbial coumarins, are gen-
erated from polyketides and will not be considered any further in this review.3AoAI
Most plant coumarins, however, are formed from L-phenylalanine and cinnamic
or 4-coumaric acid generated via the shikimate and general phenylpropanoid
pathways (Fig. 4) (S. A. Brown, personal communication).3.5 Enzymes of the
general phenylpropanoid pathway have been thoroughly studied from several
plant cells that had been induced for coumarin accumulation. 42 -44 Nevertheless,
their pattern of activation is difficult to reconcile with the time course of accu-
mulation of coumarins, since the regulation ofbergaptol O-methyltransferase, the
only coumarin-specific enzyme cloned so far, in elicited parsley cells differed
greatly from that of phenylalanine ammonia-lyase. 45 Similarly, maximal induc-
tion of xanthotoxol O-methyltransferase in Ruta graveolens cell cultures was
observed 36--48 hours post elicitation, while phenylalanine ammonia-lyase and
4-coumarate: CoA ligase reached their maxima at 8-12 hours under these con-
ditions. 46 This discrepancy must be ascribed to the fact that the synthesis of
phenylpropanoids fuels several secondary pathways in the induced plant cells,
including those leading to cell wall reinforcements which require the activation
of cinnamic acids to their CoA-esters. 4 7,48 The shikimate pathway has been
6
eOOH eOOH eOOH
ONH'
OH
L-Phenylalanine trans-Cinnamic acid trans-4-Coumaric acid
/ ~
8° 8°
lfi
-...:::
OH
OH
OH
\ /
ro # 0
Coumarin
0 -HO CO .& 0
Umbelliferone
0
Figure 4. Formation of coumarin and umbellifcrone from trans-cinnamic or trans-4-coumaric

acid. The cyclization might proceed either by ortho-hydroxylation followed by lactone-
formation or by a mechanism involving a spirodienone intermediate.
reviewed recently, and the enzymes have been allocated almost exclusively to the
plastids. 49 Unfortunately, the cyclization of cinnamic or 4-coumaric acid to
coumarin and umbelliferone, respectively (Fig. 4), has not been conclusively
elaborated in vitro, and it remains to be established whether the acids or the cor-
responding CoA-esters are the physiological substrates of cyclization.</5o.51 Fur-
thermore, it is unclear whether the cyclization proceeds in two steps, i.e.
artha-hydroxylation followed by lactonization, or in one step as was proposed,
for example, via a spiro-intermediate (Fig. 4 ).cf 50.51 Preliminary evidence in the
literature pointed to the plastids as the site of coumarin cyclization, and this is
supported by recent data on the prenylation of umbelliferone. 51}-52
168 U. MATERN
Polyoxygenated and Alkylated Simple Coumarins
Cumarin and umbelliferone may be converted to various polyoxygenated

or 0- and C-alkylated simple coumarins (Fig. 5).3 Brown concluded from pre-
cursor feeding studies in Cichorium intybus and Daphne mezereum that esculetin
and daphnetin (7,S-dihydroxy-coumarin) are synthesized by an additional hydrox-
ylation of umbelliferone at C-6 or C_S. 53 .54 Furthermore, esculetin (Fig. 1) was
proposed as an intermediate between umbelliferone and scopoletin (7-hydroxy-
6-methoxycoumarin) in the course of puberulin (6,S-dimethoxy-7-0-
prenylcoumarin) biosynthesis in Agathosma puberula. 55 Circumstantial evidence
in support of this conclusion came from a transferase activity in sunflower which
O-methylates esculetin to scopoletin in vitro. 56 However, the physiological
significance of such substrate specificities, measured in crude extracts, may be
put into question; for example, the relative O-methyltransferase activities in
Ailanthus altissima cell cultures did not correlate properly with the pattern of
coumarin metabolites. 57 In any case, the results from Cichorium and Daphne spp.
contradicted the previous proposal claiming the additional oxygenation at the cin-
namic acid stage, i.e. caffeic and ferulic acid as the precursors of scopoletin in
tobacco. 58 At least in case of caffeic acid, the cyclization reaction was confirmed
in vitro, probably as a result of phenolase activity;51.59 this might favor again the
plastids as the site of coumarin synthesis.
Many plants are capable of accumulating umbelliferone, scopoletin, and the
related ayapin (6,7-methylenedioxycoumarin) or scoparone (6,7-dimethoxy-
coumarin) as phytoalexins in response to biotic or abiotic elicitation, e.g. wild
rubber tree, sunflower, citrus, tobacco, carrot, platanus, sweet potato, elm tree, or
potato,6.7.10.12.13.15-18.20.21.60 and the importance of scopoletin accumulation as a
factor of disease resistance was demonstrated in genetically defined tobacco
lines. 12 Cytokinin appears to mediate the stress response in tobacco and to control
the accumulation of coumarins, as was demonstrated in tissues transformed with
the isopentenyl transferase gene. 61 Furthermore, a tissue-specific and develop-
mental regulation seems to govern the accumulation of scopoletin in sunflower,
and maximal levels were reached at a rather late stage (72 hr and beyond). IS
Alkyl-side chain substitution, in particular with the dimethylallyl or prenyl
group, is commonly observed among the simple coumarins. C-Prenylation of
umbelliferone at C-6 or C-S with dimethylallyl diphosphate as a cosubstrate yields
demethylsuberosin and osthenol, respectively, as the precursors of linear and
angular furano- and pyranocoumarins (Fig. 5).3-5 The prenyltransferase catalyzing
the synthesis of demethylsuberosin was extensively purified from Ruta graveolens
and shown to be specific for C-6 of umbelliferone as well as for the isoprenoid
chain length of the cosubstrate. 62 Furthermore, the enzyme was assigned to the
plastidic membrane fraction, thus further strengthening the role of plastids in
coumarin biosynthesis. The latter point deserves particular mentioning in relation
to recent findings on the biogenesis of the prenyl portion of furanocoumarins in
~
tTl
o
n
~
r
'"t:I
~ ~ ~
tTl
~O®®+ 71
CCl Z
....,
HO .# 0 0 :;
8 r
Umbelliferone
~
to
(5
f/J
--------- / ~ i
~ :iltTl
f/J
CC1
1.# 0 Vl
o HO o o 0 o'Tl
'"t:I
Demethylsuberosin y- O-Prenylumbelliferone r
;..
Z
....,
("'J
o
c
Osthenol ~
Figure 5. Dimethylallyl diphosphate:umbelliferone dimethylallyltransferase reactions in Ammi majus. ~

Z
f/J
0"-
'-D
170 U. MATERN
celery. Detached celery leaves can be induced withjasmonate to accumulate pso-

ralens, and these psoralens incorporate carbons I and 2 of the prenyl residue into
the furan-ring (see below). Precursor feeding studies with I -deoxy-[5,5- ZHz]-D-
xylulose (DOX) revealed high incorporation rates of DOX (>30%) into the prenyl-
derived portion of the psoralens,52 which is in sharp contrast to the data published
previously on the incorporation of label from mevalonate into psoralens (less than
1%).3 The synthesis of DOX and its derivatives isopentenyl- and dimethylallyl
diphosphate must thus be ascribed to the mevalonate-independent pathway which
is supposed to operate in the plastids. 63 The concomitance of isopentenyl diphos-
phate isomerase activity is likely, since two genes encoding plastidic isomerases
were recently cloned from Arabidopsis.64
Most of the in vitro biosynthetic studies have been conducted with Api-
aceae cell cultures, which were induced with fungal elicitors to accumulate
coumarins, and Ammi majus (Bishop's weed) turned out particularly suitable for
such studies. These cells produced predominantly butenyl ethers of umbellifer-
one besides linear furanocoumarins. Microsomes prepared from the elicited cells
catalyzed both the 6-C- and the 7-0-prenylation to yield demethylsuberosin or
O-prenylumbelliferone (Fig. 5) as the precursors of psoralens and umbelliferone
butenyl ethers, respectively.65 Both prenyltransferases were active in the presence
of magnesium ions, whereas manganese ions fully supported the O-prenylation
but only one third of the maximal C-prenylation reaction. 65 Although further frac-
tionation of these particulate enzymes was not attempted, it appears likely that
both the activities reside in the plastids.
Psoralen
Furanocoumarins have been isolated from many unrelated plant families,I-3

and the capacity for their synthesis must have evolved several times. 66 Two
major classes, the linear and the angular furanocoumarins, can be distinguished,
which both originate from C-prenylated umbelliferones by oxidative cyclization
of the prenyl side-chain (Fig. 6); prenylation of umbelliferone at C-6 (deme-
thylsuberosin) leads to the psoralens, whereas prenylation at C-8 yields the pre-
cursor (osthenol) for the angular series. Cell cultures of the Apiaceae accumulate
predominantly psoralens upon treatment with fungal elicitors. 5.8 ,9 Microsomes pre-
pared from el icited Ammi majus cells catalyzed the synthesis of (+ )-marmesin from
demethylsuberosin in the presence ofNADPH and oxygen, which was further con-
verted to psoralen and bergaptol (5-hydroxypsoralen).5,67 The microsomal activi-
ties turned out to be fairly labile, and the microsomes from cells of related plant
species (Petroselinum crispum and Arracacia xanthorhizza) were not nearly as
efficient, although these cell cultures accumulated comparable amounts of pso-
ralens upon elicitation. The overall sequence of reactions (Fig. 6) was catalyzed by
three consecutive cytochrome P450-dependent monooxygenases, and inhibition
studies with commercial P450 inhibitors as well as the variable degree of elicitor-
3:::
m
tJ
n
~r<
['1,0,1 ~""
~ ,~ m
~ HOU 0 .10 ~ H,O Z
....,
HO1"';'~OVOAO l(o~oAo ~ ;;
Demethylsuberosin (+)-Marmesin r<
Psora len Acetone
~
tJ
III
~co/ is
[/).
o 0 -<
Z
....,
Umbelliferone ~ ::c:
m
[/).
ti'i
~~ o..,..,
Angelicin
HO b o 0 o ""r<
~ ~....,
I n
o
c:
3:::
Osthenol (-)-Columbianetin >-
;:>;;I
Figure 6. Sequence of reactions converting umbelliferone to linear (top) or angular (bottom) furanocoumarins.
Z
[/).
....,
172 U. MATERN
induction of the specific enzyme activities revealed that each step was catalyzed
by a different enzyme entity.5.8 It is likely that angular furanocoumarins are formed
in a similar way from osthenol (Fig. 6), but these reactions have not yet been
observed in vitro despite the fact that Ammi majus is capable of accumulating the
angular type of coumarins in its fruits at least. 68
The conversion of demethylsuberosin to (+)-marmesin had been proposed
to involve the epoxidation of the side-chain double bond and formation of the
corresponding diol. 3 However, although in principle P450 monooxygenases are
capable of inserting an "oxen" into 0Iefins,69.7o no such intermediate was observed
in vitro during the marmesin synthase reaction. It is more likely, therefore, that
the 7-hydroxyl group of demethylsuberosin favors the one step cyclization to
(+)-marmesin by delocalizing the side-chain double bond electrons. Model
mechanisms proposed for the interaction of the catalytic P450 oxo-derivative
(Fig. 7) with aliphatic double bonds support such a direct cyclization. 71 It should
be noted in this context that linear dihydropyrano- or pyronocoumarins such as
graveolone had been considered to originate from demethylsuberosin by a dif-
ferent cyclization reaction;3.72 preliminary evidence obtained with microsomes
from elicited parsley cells suggested, however, that graveolone is possibly formed
from (+)-marmesin. 5.8
The formation of psoralen from (+)-marmesin is a unique reaction in plant
secondary metabolism, since stoichiometric amounts of acetone must be
released. 3.5 The reaction had been proposed to be driven by the generation of the
carbenium ion of (+)-marmesin at C-3' followed by a 1,3-elimination step.3 This
hypothetical mechanism did not require the hydroxylation at C-3' and implied
a 2',3'-anti-elimination reaction. The identification of psoralen synthase as a
cytochrome P450-dependent monooxygenase in microsomes from elicited Ammi
majus or parsley cells refuted the proposal and pointed to an oxidative mecha-
nism involving iron-oxygen radical species. 5.8.67 After binding of the substrate,
P450 monooxygenases are reduced by cytochrome P450 reductase in the pres-
ence of molecular oxygen to yield both the reactive peroxyiron ([Felll-O-OH] and
oxoiron ([Fe'V-On enzyme species (Fig. 7). Accordingly, a sequential mechanism
appeared possible analogous to the side-chain dealkylation of cholesterol via
pregnenolone to the il '6 _steroid;69.71 this mechanism might involve the oxidative
functionalization of the isopropyloxy substituent in (+)-marmesin and removal of
one carbon by the reactive [Fe'V -0]" porphyryl radical cation (Fig. 7), leaving an
acetyl-derivative, followed by deacetylation through peroxyhemiketal formation
with the strong nucleophile [Felll-O-O-] (Fig. 7).69.71
In cooperation with the group of Boland, we revisited the mechanism of
the psoralen synthase reaction with stereospecifically deuterated substrates. 5.73
These studies revealed the elimination of the hydrogen in syn-orientation to the
isopropyloxy-substituent (Fig. 8). Furthermore, labeled acetone released from the
incubation was trapped with pentafluorobenzylhydroxylamine, clearly indicating
that the isopropyloxy side-chain was removed in one step. The microsomal
S(Cys) ~(Cys)
N~N N :
,;=-! N
/., 'Felll~1 RH e /.. .:::Fe ll
----/
N~I"
-N
N~'"'" "
-N
OH
2
RH
[Fe" l- 0 - OH]
[Fe 1v_ or
Figure 7. Schematic activation of cytochrome P450 yielding the ferric-hydroperoxy species
and the iron-oxo radical. The low-spin iron in the resting enzyme is converting to high-spin iron
upon substrate (RH) binding.
174 U. MATERN
HO~'"
.' <:Co
D,•••
0
H
I ~
bOO
~ + [Fe IV - 0)-
Fe IV 00 +
00
Felli +
~ OH
+
Felli + HDO
+
y
o
Figure 8. Schematic conversion of (+)-[3'-'H,jmarmesin to psoralen and acetone by syn-
elimination. The iron-oxo radical [Felv-Or abstracts the deuterium from carbon 3'. The reactive
intermediate loses the side-chain isopropyl radical by disproportionation, and the isopropyl
radical recombines with the 'HO radical in an oxygen-rebound process.
psoralen synthase accepted also the synthetic pseudosubstrate in which the iso-
propyloxy side-chain was replaced by an acetyl residue (Fig. 9). However, the
diol corresponding to (+ )-marmesin (iso-I,2-propylene residue), considered as
an intermediate in the hypothetical two step reaction, was not accepted; 73 this
diol is known as a natural plant metabolite (prandiolV Taken together, the
[Fe IV -Or porphyryl radical cation at the active site of psoralen synthase attacks
the Hs at C-3'of (+)-marmesin (Fig. 8). Homolytic abstraction of this hydrogen
yields a benzylic radical which then undergoes p-cleavage to psoralen and an
isopropyloxy radical. The latter radical reacts in an oxygen rebound process with
the adjacent porphyryl-Fe1V-OH center to acetone hydrate (Fig. 8).73 Synthetic
iron or manganese tetraphenylporphyryl complexes activated with iodosobenzene
also showed some psoralen synthase activity, but lacked the stereospecificity of
the microsomal enzyme catayzing both syn and anti eliminations. This empha-
sizes the impact of the spatial enzyme structure on the selectivity of the reaction
as was also demonstrated recently for camphor hydroxylase. 74
[Feill-O-OHJ
o
I
II
H3C - C- OH
Felli + HOO
Psoralen
Figure 9. Schematic conversion of (+)-[3' -'H,J-2'-acetyl-2',3' -dihydropsoralen to psoralen and

acetate by the ferric-hydroperoxy enzyme species. This compound served as a pseudosubstrate
for Ammi majus psoralen synthase.
176 U. MATERN
Oxygenated Psora lens
Elicitor-treatment of Ammi majus cells induced an additional P450

monoxygenase which hydroxylated psoralen to bergaptol (5-hydroxybergapten)
(Fig. 10)5.67 Assays under various conditions revealed that (+)-marmesin or
(±)-marmesin was not hydroxylated, thus ruling out a pathway to bergaptol via
5-hydroxymarmesin as considered previously.3 Xanthotoxol (8-hydroxypsoralen)
and the unstable hydroquinone 5,8-dihydroxypsoralen (Fig. 11) are likely formed
analogously from psoralen by the action of 5- and 8-monooxygenases, but these
hydroxylations have not been accomplished in vitro. Nevertheless, based on the
narrow specificity of bergaptol synthase, one or two additional monooxygenases
appear to be necessary for the synthesis of 5,8-dihydroxypsoralen (Fig. 11).
Neither one of these hydroxylated psoralens accumulates in umbelliferous
plants or induced cell cultures, because the compounds are readily methylated
to the major final products bergapten and xanthotoxin or, to a lesser degree,
isopimpinellin (Fig. 11). The O-methyltransferases methylating bergaptol to
bergapten (BMT) and xanthotoxol to xanthotoxin (XMT) were extensively
purified from the common rue and parsley cells,3.8,45,75 and the purification of
the common rue enzymes impressively demonstrated the high efficiency of
affinity chromatography.76 The expectation that the substrate specificities of the
O-methyltransferases might reveal the sequence of reactions from psoralen to
isopimpinellin, however, was disappointing. Parsley BMT catalyzed the methy-
lation of 5-hydroxy groups of furanocoumarins and of both hydroxy groups
in 5,8-dihydroxypsoralen, whereas XMT specifically methylated xanthotoxol
(8-hydroxypsoralen) to xanthotoxin, even in the presence of 8-hydroxybergapten
(Fig. 11). Furthermore, BMT methylated 5-hydroxyxanthotoxin at a higher rate
than bergaptol, and another, yet unidentified, transferase for the methylation
of 8-hydroxybergapten to isopimpinellin was not ruled out. Overall, the general
pattern of isopimpinellin biosynthesis has been elucidated (Fig. 11), but the
precise order of individual reaction steps remains to be established.
CONCLUSIONS
The conversion of umbelliferone to bergaptol has been accomplished in

vitro and requires the consecutive action of dimethylallyl diphosphate:umbellif-
erone 6-C-dimethylallyltransferase (prenyltransferase), marmesin synthase, pso-
ralen synthase, and psoralen 5-monooxygenase. All four particulate enzyme
activities can be induced in umbelliferous cell cultures by the addition of fungal
elicitor. Most likely, the prenyltransferase is associated with the plastids, whereas
the latter three enzymes were identified as cytochrome P450 monooxygenases
which must be assigned to the endoplasmic reticulum. Within the P450 enzymes,
marmesin and psoralen synthases catalyze unusual reactions (dihydrofuran for-
s::m
U
Pi
Z
;I>
t-
>-0
~
~
....,
l : ~
000
ceo Psoralen
~
~
U
/ ~ OH ttl
5C/J
~ i....,
~. . cCO
Ho'l 0 0 0
lo~o---lo
~ lloVo---lo ::r:
m
Bergaptol Bergapten C/J
(+)-Marmesin (jj
o
>Tj
~ OH /
• >-0
t-
;I>
Z
....,
~.CJCXlo n
HO' I o
5-Hydroxymarmesin c:
s::
;I>
Figure 10. Sequence of reactions converting (+ )-marmesin to bergapten.
c:z
C/J
-..)
-..)
178 U. MATERN
OH
cCCL000
Psoralen
•
Bergaptol
0
J OH
.
0 0
OH OH
Xanthotoxol 5,8-Dihydroxypsoralen
J
J
0
CH 3 0
Xanthotoxin
J
J
eGo
OH CH 3 0
..
0 I 0 I # 0 0
CHaO CHaO
5-Hydroxyxanthotoxin Isopimpinellin
Figure II. Pattern of reactions that may lead from psoralen to isopimpinellin.
mation and carbon-carbon bond cleavage with release of acetone, repectively).

Based on the proposal that the capacity of furanocoumarin biosynthesis must have
evolved several times,66 both these synthases conceivably originated from minor
base changes of ancestral genes. The current knowledge of conserved sequence
domains in P450 monooxygenases and the inducibility of the synthases in cell
cultures enable the cloning of their genes by PCR techniques. Cladistic analysis
of these DNA sequences should lead to the classification of the enzymes, and
mutational studies might identify the domain(s) responsible for the catalytic
specificities. The characterization of these domains will likely provide the tools
for the biotechnological generation of plants with enhanced or modified coumarin
biosynthetic capabilities. Such plants might hold promise for further medical
applications.
ACKNOWLEDGMENTS
The work cited from our laboratory was supported by the Deutsche Forschungs-
gemeinschaft and the Fonds der Chemischen Industrie. Critical reading of the
manuscript by M. Petersen, Marburg, is gratefully acknowledged.
REFERENCES
I. MURRAY, R.D.H. 1997. Naturally occurring plant coumarins. Progr. Chern. Org. Natural
Prod. 72: 1-119.
2. ESTEVEZ-BRAUN, A., GONZALEZ, A.G. 1997. Coumarins. Nat. Prod. Rep. 14:
465-457.
3. MURRAY, R.D.H., MENDEZ, J., BROWN, S.A. 1982. The Natural Coumarins. Occur-
rence, Chemistry and Biochemistry. lohn Wiley, 702 p.
4. O'KENNEDY, R., THORNES, R.D. 1997. Coumarins. Biology, Applications and Mode
of Action. John Wiley, 348 p.
5. MATERN, u., LOER, P, KREUSCH, D. 1999. Biosynthesis of coumarins. In press in
Comprehensive Natural Products Chemistry, Volume I: Polyketides and Other Secondary
Metabolites Including Fatty Acids and Their Derivatives (u. Sankawa, cd), Pergamon.
Oxford.
6. GIESEMANN, A., BIEHL, B. LIEBEREI, R. 1986. Identification of scopoletin as a phy-
toalexin of the rubber tree Hevea brasiliensis. J. Phytopathol. 117: 373-376.
7. TAL, 8., ROBESON, 0.1., 1986. The induction by fungal inoculation of ayapin and scopo-
letin biosynthesis in Helianthus anmlus. Phytochemistry 25: 77-79.
8. MATERN, u., STRASSER, H., WENDORFF, H., HAMERSKI, H. 1988. Coumarins and
furanocoumarins. pp. 3-21 in Cell Culture and Somatic Cell Genetics of Plants, Vol. 5:
Phytochemicals in Plant Cell Cultures (F. Constabcl and l.K. Vasil, eds.), Academic Press,
New York.
9. HAMERSKI, D., BEIER, R.C., KNEUSEL, R.E., MATERN, u., HIMMELSPACH, K.
1990. Accumulation of coumarins in elicitor-treated Ammi majus L. cell suspension
cultures. Phytochemistry 29: 1137-1142.
180 U. MATERN
10. SULISTYOWATI, L., KEANE, P.1., ANDERSON, 1.w. 1990. Accumulation of the
phytoalexin, 6,7-dimethyoxycoumarin, in roots and stems of citrus seedlings following
inoculation with Phytophthora citrophthora. Physiol. Mol. Plant Pathol. 37: 451-461.
II. AFEK, u., CARMELI, S., AHARONI, N. 1995. Columbianetin, a phytoalexin associated
with celery resistance to pathogens during storage. Phytochemistry 39: 1347-1350.
12. AHL GOY, P., SIGNER, H., REIST, R., AICHHOLZ, R., BLUM, w., SCHMIDT, E.,
KESSMANN, H. 1993. Accumulation of scopoletin is associated with the high disease
resistance of the hybrid Nicotiana glutinosa x Nicotiana debneyi. Planta 191: 200-206.
13. COXON, D.T., CURTIS, R.E, PRICE, K.R., LEVETT, G. 1973. Abnormal metabolites
produced by Daucus carola roots stored under conditions of stress. Phytochemistry 12:
/881-1885.
14. ECKEY-KALTENBACH, H., KIEFER, E., GROSSKOPF, E., ERNST, D.,
SANDERMANN Jr., H. 1997. Differential transcript induction of parsley pathogenesis-
related proteins and of a small heat shock protein by ozone and heat shock. Plant Mol.
BioI. 33: 343-350.
15. EL MODAFAR, e., CLERIVET, A., FLEURIET, A., MACHEIX, J.1. 1993. Inoculation
of Platanus acerifolia with Ceratocystis jimbriata f. sp. platani induces scopoletin and
umbelliferone accumulation. Phytochemistry 34: 1271-1276.
16. GARCIA, D., SANIER, e., MACHEIX, J.1., D'AUZAC, 1. 1995. Accumulation ofscopo-
letin in Hevea brasiliensis infected by Microcyclus ulei (P. Henn.) V Arx and evaluation
of its fungitoxicity for three leaf pathogens of rubber tree. Physiol. Mol. Plant Pathol.
47: 213-223.
17. GUTIERREZ, M.e., PARRY, A., TENA, M., JORRIN, 1., EDWARDS, R. 1995. Abiotic
elicitation of coumarin phytoalexins in sunflower. Phytochemistry 38: 1185-1191.
18. GUTIERREZ-MELLADO, M.e., EDWARDS, R., TENA, M., CABELLO, E,
SERGHINI, K., JORRIN, 1. 1996. The production of coumarin phytoalexins in different
plant organs of sunflower (Helianthus annuus L.). 1. Plant Physiol. 149: 261-266.
19. HEATH-PAGLIUSO, S., MATLIN, S.A., FANG, N., THOMPSON, R.H., RAPPAPORT,
L. 1992. Stimulation of furanocoumarin accumulation in celery and celeriac tissues by
Fusarium oxysporum f.sp. apii. Phytochemistry 31: 2683-2688.
20. MINAMIKAWA, T., AKAZAWA, T., IRITANI, I. 1963. Analytical study of umbellifer-
one and scopoletin synthesis in sweet potato roots infected by Ceratocystis fimbriata.
Plant Physiol. 38: 493-497.
21. VALLE, T., LOPEZ, 1.L., HERNANDEZ, 1.M., CORCHETE, P. 1997. Antifungal activ-
ity of scopoletin and its differential accumulation in Ulmus pumila and Ulmus campestris
cell suspension cultures infected with Ophiostoma ulmi spores. Plant Sci. 125: 97-101.
22. VALENTE, E.1., PORTER, W.R., TRAGER, w.F. 1978. Conformations of selected 3-
substituted 4-hydroxycoumarins in solution by nuclear magnetic resonance. Warfarin and
phenprocoumon. 1. Med. Chern. 21: 231-234.
23. MORII, K., MURAl, 0., HASHIMOTO, S., NAKAMURA, Y 1996. Highly regio- and
stereoselective photocyc1oaddition between coumarin and thymine by molecular recog-
nition. Tetrahedron Lett. 37: 8523-8526.
24. KIMURA, Y, OKUDA, H., ARICHI, S., BABA, K., KOZAWA, M. 1985. Inhibition of
the formation of5-hydroxy-6,8, 11,14-eicosateraenoic acid from arachidonic acid in poly-
morphonuclear leukocytes by various coumarins. Biochim. Biophys. Acta 834: 224-229.
25. BEUTLER, 1.A., CARDELLINA II, 1.H., MCMAHON, 1. B., SHOEMAKER, R.H.,
BOYD, M.R. 1995. Antiviral and antitumor plant metabolites. pp. 47-64 in Recent
Advances in Phytochemistry, Vol. 29: Phytochemistry of Medicinal Plants (1.T. Amason,
R. Mata, 1.T. Romeo, eds.), Plenum Press, New York.
26. KASHMAN, Y, GUSTAFSON, K.R., FULLER, R.W., CARDELLINA II, 1.H.,
MCMAHON, lB., CURRENS, M.J., BUCKHEIT Jr., R.W., HUGHES, S.H., CRAGG,
G.M., BOYD, M.R. 1992. The calanolides, a novel HIV-inhibitory class of coumarin
derivatives from the tropical rainforest tree, Calophyllum lanigerum. 1. Med. Chern. 35:
2735-2743.
27. MCKEE, T.e., CARDELLINA II, 1.H., DREYER, G.B., BOYD, M.R. (1995) The pseudo-
calanolides: Structure revision of calanolides C and D. J. Nat. Prod. 58: 916--920.
28. GALINIS, D.L., FULLER, R.W, MCKEE, T.C., CARDELLINA II, 1.H.,
GULAKOWSKI, R.1., MCMAHON, J.B., BOYD, M.R. 1996. Structure-activity
modifications of the HIV-I inhibitors (+)-calanolide A and (-)-calanolide B. J. Med.
Chern. 39: 4507-4510.
29. PATIL, A.D., FREYER, A.J., EGGLESTON, D.S., HALTIWANGER, R.C., BEAN, M.E,
TAYLOR, P.B., CARANFA, M.J., BREEN, A.L., BARTUS, H.R., JOHNSON, R.K.,
HERTZBERG, R.P., WESTLEY, J.W 1993. The inophyllums, novel inhibitors of HI V-I
reverse transcriptase isolated from the Malaysian tree, Calophyllum inophyllum Linn.
1. Med. Chern. 36: 4131-4138.
30. DHARMARATNE, H.R.W, WANIGASEKERA, WM.A.P., MATA-GREENWOOD, E.,
PEZZUTO, J.M. 1998. Inhibition of human immunodeficiency virus type-I reverse tran-
scriptase activity by cordatolides isolated from Calophyllum cordato-oblongum. Planta
Med. 64: 460-461.
31. CURRENS, M.J., MARINER, J.M., MCMAHON, J.B., BOYD, M.R. 1996. Kinetic
analysis of inhibition of human immunodefivciency virus type-I reverse transcriptase
by calanolide A. J. Pharmacol. Exp. Ther. 279: 652-661.
32. MAZUMDER, A., WANG, S., NEAMATI, N., NICKLAUS, M., SUNDER, S., CHEN,
J., MILNE, G.WA., RICE, WG., BURKE Jr., T.R., POMMIER, Y. 1996. Antiretroviral
agents as inhibitors of both human immunodeficiency virus type-I integrase and protease.
J. Med. Chern. 39: 2472-2481.
33. ZHAO, H., NEAMATI, N., HONG, H., MAZUMDER, A., WANG, S., SUNDER, S.,
MILNE, G.WA., POMMIER, Y., BURKE Jr., T.R. 1997. Coumarin-based inhibitors of
HIV integrase. 1. Med. Chern. 40: 242-249.
34. SAPAROY, S.M., ZHURAVEL, N.N., MOLLAEY, R.E., SUKHORUKOY, Y.L.,
POTAPENKO, A.Y. 1991. Effect of calcium ions on psora len-sensitized photohaemoly-
sis. 1. Photochem. Photobiol. 10: 159-164.
35. GUIOTTO, A., CHILIN, A., MANZINI, P., DALI.:ACQUA, E, BORDIN, E,
RODIGHIERO, P. 1995. Synthesis and antiproliferative activitivity of furocoumarin
isosters. Farmaco 50: 479-488.
36. HUANG, H.C., LAI, M.W, WANG, H.R., CHUNG, Y.L., HSIEH, L.M., CHEN, e.e.
1993. Antiproliferative effect of esculetin on vascular smooth muscle cells: Possible roles
of signal tranduction pathways. Eur. 1. Pharmacol. 237: 39-44.
37. VEREECKE, D., MESSENS, E., KLARSKOY, K., DE BRUYN, A., VAN MONTAGU,
M., GOETHALS, K. 1997. Patterns of phenolic compounds in leafy galls of tobacco.
Planta 201: 342-348.
38. WANG, B.H., TERNAI, B., POLYA, G. 1997. Specific inhibition of cyclic AMP-
dependent protein kinases by warangalone and robustic acid. Phytochemistry 44:
787-796.
39. BERGENDORFF, 0., DEKERMENDJIAN, K., NIELSEN, M., SHAN, R., WITT, R.,
AI, J., STERNER, O. 1997. Furanocoumarins with affinity to brain benzodiazepine
receptors in vitro. Phytochemistry 44: 1121-1124.
40. INOUE, T., TOYONAGA, T., NAGUMO, S., NAGAI, M. 1989. Biosynthesis of
4-hydroxy-5-methylcoumarin in a Gerbera jamesonii hybrid. Phytochemistry 28:
2329-2330.
41. SINGH, R., HSIEH, D.P.H. 1977. Aflatoxin biosynthetic pathway---elucidation by using
blocked mutants of Aspergillus parasiticus. Arch. Biochem.Biophys. 178: 285-292.
182 0. MATERN
42. HAHLBROCK, K., SCHEEL, D. 1989. Physiology and molecular biology of phenyl-
propanoid metabolism. Ann. Rev. Plant Physiol. Plant Mol. BioI. 40: 347-369.
43. DOUGLAS, C.J. 1996. Phenylpropanoid metabolism and lignin biosynthesis: From
weeds to trees. Trends Plant Sci. 1: 171-176.
44. SEELENFREUND, D., CHIONG, M., LOBOS, S., PEREZ, L.M. 1996. A full length
eDNA coding for phenylalanine ammonia-lyase from Citrus limon. Plant Physiol. Ill:
348.
45. LOZOYA, E., BLOCK, A., LOIS, R., HAHLBROCK, K., SCHEEL, D. 1991. Transcrip-
tional repression of light-induced flavonoid synthesis by elicitor treatment of cultured
parsley cells. Plant 1. 1: 227-234.
46. BOHLMANN, 1., GIBRALTARSKAYA, E., EILERT, 0. 1995. Elicitor induction of
furanocoumarin biosynthetic pathway in cell cultures of Rula graveolens. Plant Cell,
Tissue Organ Cult. 43: 155-161.
47. GRIMMIG, 8., MATERN, 0. 1997. Structure of the parsley caffeoyl-CoA 0-
mcthyltransferase gene, harbouring a novel elicitor responsive cis-acting element. Plant
Mol. BioI. 33: 323-341.
48. PETERSEN, M., STRACK, D., MATERN, 0. 1999. Biosynthesis of phenylpropanoids
and related compounds. In press in Annual Plant Reviews, Vol. 2: Biochemistry of Plant
Secondary Metabolism (M. Wink, ed.), Sheffield Academic Press, Sheffield.
49. WEAVER, L.M., HERRMANN, K.M. 1997. Dynamics of the shikimate pathway in
plants. Trends Plant Sci. 2: 346-351.
50. CONN, E.E. 1984. Compartmentation of secondary compounds. Ann. Proc. Phytochem.
Soc. Europe 24: 1-28.
51. MATERN, 0. 1991. Coumarins and other phenylpropanoid compounds in the defense
response of plant cells. Planta Med. 57: S 15-S20.
52. STANJEK, Y, PIEL, 1., BOLAND, W. 1999. Mevalonate-independent biosynthesis of
linear furanocoumarins in Apium graveolen.l' (Apiaceae). Phytochemistry, in press.
53. BROWN, S.A., 1985. Biosynthesis of6,7-dihydroxycoumarin in Cichorium intybus. Can.
1. Biochem. Cell BioI. 63: 292-295.
54. BROWN, S.A. 1986. Biosynthesis of daphnetin in Daphne mezereum L. Z. Naturforsch.
4Ic: 247-252.
55. BROWN, S.A., MARCH, R.E., RIVETT, D.E.A., THOMPSON, H.J. 1988. Intermediates
in the formation ofpuberulin by Agalhosma puberula. Phytochemistry 27: 391-395.
56. GUTIERREZ, M.e., PARRY, A., TENA, M., JORRIN, 1., EDWARDS, R. 1995. Abiotic
elicitation of coumarin phytoalexins in sunflower. Phytochemistry 38: 1185-1191.
57. OSOBA. O.A., ROBERTS, M.E 1994. Methyltransferase activity in Ailanthus allissima
cell suspension cultures. Plant Cell Rep. 13: 277-281.
58. FRITIG, B., HIRTH, L., OURISSON, G. 1970. Biosynthesis of the coumarins: Scopo-
letin formation in tobacco tissue cultures. Phytochemistry 9: 1963-1975.
59. SATO, M. 1967. Metabolism of phenolic substances by the chloroplasts-III. Phenolase
as an enzyme concerning the formation of esculetin. Phytochemistry 6: 1363-1373.
60. ANDREAE, S.R., ANDREAE, W.A. 1949. The metabolism of scopoletin by healthy and
virus infected potato tubers. Can. 1. Res. 27: 15-22.
61. HAMDI, S., CRECHE, 1., GARNIER, E, MARS, M., DECENDIT, A., GASPAR, T.,
RIDEAU, M. 1995. Cytokinin involvement in the control of coumarin accumulation in
Nicotiana ,abacum. Investigations with normal and transformed tissues carrying the
isopentcnyl transferase gene. Plant Physiol. Biochem. 33: 283-288.
62. DHILLON, D.S., BROWN, S.A. 1976. Localization, purification and characterization of
dimcthylallyl pyrophosphate:umbelliferone dimethylallyltransferase from Rula grave-
olens. Arch. Biochem. Biophys. 177: 74-83.
63. SPRENGER, G.A., SCHORKEN, 0., WIEGERT, T., GROLLE, S., DE GRAAF, A.A.,
TAYLOR, S.V, BEGLEY, T.P., BRINGER-MEYER, S., SAHM, H. 1997. Identification

of a thiamine-dependent synthase in Escherichia coli required for the formation of the
I-deoxy-D-xylulose 5-phosphate precursor to isoprenoids, thiamine, and pyridoxol.
Proc. Natl. Acad. Sci. USA, 94: 12857-12862.
64. CAMPBELL, M., HAHN, F.M., POULTER, c.D., LEUSTEK, T. 1997. Analysis of the
isopentenyl diphosphate isomerase gene family from Arabidopsis thaliana. Plant Mol.
BioI. 36: 323-328.
65. HAMERSKI, D., SCHMITT, D., MATERN, U. 1990. Induction oftwo prenyltransferases
for the accumulation of coumarin phytoalexins in elicitor-treated Ammi majus ceIJ sus-
pension cultures. Phytochemistry 29: 1131-1135.
66. BERENBAUM, M.R., ZANGERL, A.R. 1996. Phytochemical diversity: Adaption or
random variation? pp. 1-24 in Recent Advances in Phytochemistry, Vol. 30: Phytochem-
ical Diversity and Redundancy in Ecological Interactions (1.T. Romeo, 1.A. Saunders, P.
Barbosa, eds.), Plenum Press, New York.
67. HAMERSK1, D., MATERN, u., 1988. Biosynthesis of psoralens. Psoralen 5-
monooxygenase activity from elicitor-treated Ammi maius ceIJs. FEBS Lett. 239:
263-265.
68. ELGAMAL, M.HA, SHALABY, N.M.M., DUDDECK, H., HIEGEMANN, M. 1993.
Coumarins and coumarin glucosides from the fruits of Ammi majus. Phytochemistry 34:
819-823.
69. AKHTAR, M., WRIGHT, 1.N. 1991. A unified mechanistic view of oxidative reactions
catalyzed by P-450 and related Fe-containing enzymes. Nat. Prod. Rep. 8: 527-551.
70. BOLWELL, G.P', BOZAK, K., ZIMMERLlN, A. 1994. Plant cytochrome P450. Phyto-
chemistry 37: 1491-1506.
71. HALKIER, B.A. 1996. Catalytic reactivities and structure/function relationships of
cytochrome P450 enzymes. Phytochemistry 43: 1-21.
72. BEIER, R.C., IVIE, G.w., OERTLl, E.H. 1994. Linear furanocoumarins and graveolone
from the common herb parsley. Phytochemistry 36: 869-872.
73. STANJEK, V, MIKSCH, M., LUEER, P., MATERN, u., BOLAND, W. 1999.
Biosynthesis of psoralen: Mechanism of a cytochrome P450 catalyzed oxidative
<C->C-bond cleavage. Angew. Chem. Int. Ed. 38: 400-402.
74. GROVES,1.T. 1997. The importance of being selective. Nature 389: 329-330.
75. SHARMA, S.K., GARRETT, 1.M., BROWN, SA 1979. Separation of the S-
adenosylmethionine: 5- and 8-hydroxyfuranocoumarin O-methy1transferases of Ruta
graveolens L. by general ligand affinity chromatography. Z. Naturforsch. 34c: 387-391.
76. SHARMA, S.K., BROWN, S.A. 1979. Affinity chromatography of Ruta graveolens L.
O-methy1transferases. Studies demonstrating the potential of the technique in the
mechanistic investigation of O-methyltransferases. Can. 1. Biochem. 57: 986-994.
Chapter Eight
BIOSYNTHESIS, BIODEGRADATION, AND

CELLULAR LOCALIZATION OF
HYDROLYZABLE TANNINS
Georg G. Gross
Universitat Ulm
Abteilung Allgemeine Botanik
D-89069 Ulm, Germany
Introduction ................................................... 185

Biosynthesis of Hydrolyzable Tannins .............................. 187
Origin of Gallic Acid ......................................... 188
Formation of ~-Glucogallin .................................... 189
The Pathway to Pentagalloylglucose ............................. 190
Biosynthesis of Complex Gallotannins ........................... 197
Formation of Ellagitannins ..................................... 203
Degradation of Galloylglucoses ................................... 203
Localization of Gallotannins ...................................... 206
Conclusions and Perspectives ...................................... 208
INTRODUCTION
Tannins occupy a prominent position among plant polyphenols, not only

because of their widespread occurrence in the plant kingdom but also on grounds
of their technological, ecological, and medical significance. Unfortunately, it is
impossible to describe simply this large group of natural products on the basis
of a common, general chemical structure. Like some other heterogeneous groups
of secondary plant constituents (essential oils, for instance), several different
criteria are required to define tannins, the decisive factor traditionally being their
tendency to interact with aqueous solutions of proteins and other biological
Phytochemicals in Human Health Protection, Nutrition, and Plant Defense. edited by Romeo.
185
186 G.G.GROSS
macromolecules to form insoluble precipitates. A prominent example of this

property is tannage, i. e. the conversion of raw animal hides to leather. Other tannin
effects regard haze formation in beer, the taste, palatability, and digestibility of
wines, fruit, vegetables, and leaves (which is thought to deter herbivores), or
the inactivation of enzymes, particularly microbial exoenzymes. These manifold
effects cannot be ascribed to a particular tannin structure, except for the general
requirement of a high molecular weight of 800-3,000 daltons and sufficient
phenolic hydroxy groups to allow complexation.
Traditionally, plant tannins are classified as condensed tannins (to date often
referred to as proanthocyanidins according to the characteristic liberation of
colored anthocyanidin degradation products upon treatment with hot alcoholic
mineral acid) and hydrolyzable tannins. 1 The latter are characterized by a central
polyol moiety (usually ~-D-glucose) whose hydroxy functions are esterified with
one to five gallic acid (3,4,5-trihydroxybenzoic acid, 1) molecules (Fig. 1), or with
more complex derivatives of this triphenol. The fully galloylated glucose deriva-
tive, I ,2,3,4,6-penta-O-galloyl-~-D-glucose (2), is regarded as the immediate
precursor of the two subclasses of hydrolyzable tannins, i.e. gallotannins and
ellagitannins. Ellagitannins are supposed to result from oxidative processes by
o
eOOH
HO¥OH
OH
Gallic acid (1)
1,2,3,4,6-Penta-O-galloyl-B-D-glucose (2)
OH OH
OH OH
HO
HO
OH
OH
meta-Digallic acid (5) Ellagic acid (4)
Hexahydroxydiphenic
acid (3)
Figure 1. Structures of characteristic components of hydrolyzable tannins.

BIOSYNTHESIS, BIODEGRADATION, AND HYDROLYZABLE TANNINS 187
which C-C linkages are formed between adjacent galloyl residues of pentagal-
loylglucose to yield (R) or (S)-3,4,5,3',4',5'-hexahydroxydiphenoyl (HHDP, 3)
groups, followed by the subsequent formation of dimeric and oligomeric deriva-
tives that are connected via C-C or C-O-C bonds between the galloyl residues.
After the eventual hydrolytical release of these HHDP residues, the free diphenic
acid spontaneously rearranges to the extremely insoluble dilactone, e\lagic acid
(4), which became name-giving for this group of natural products. Gallotannins,
in contrast, originate by a quite different mechanism, i.e. by esterification of
further galloyl units to the pentagalloylglucose (2) core to yield digalloyl residues
which are attached via so-called meta-depside bonds (5). Substitution degrees of
as much as 10-12 galloyl residues have been reported for such gallotannins from
various sources. 2- 5 It should be noted that evidence, based on NMR spectroscopy,
has been presented that gallotannins are mixtures of meta- and para-depsides in
nature. 3- S As this view still awaits supporting experiments, only the traditional
meta-bonds, known from the literature for decades, are used in this chapter for
the illustration of gallotannin structures. To differentiate unequivocally these
polygalloylglucoses (gallotannins sensu strictu) from their one to five-fold sub-
stituted precursors (often referred to as "simple galloylglucose esters"), the term
"complex gallotannins" is used in this chapter.
Innumerable structures of hydrolyzable tannins and related compounds
bave been reported from many laboratories for decades, and this solid chemical
basis has prompted studies on the biochemistry of these complex molecules. This
question has been investigated in the author's laboratory mainly by extensive
enzyme studies, i.e. by a technique having the advantage that it not only allows
the unequivocal identification of metabolic intermediates but also provides oth-
erwise inaccessible information about "activated" intermediates as an indispens-
able tool for the elucidation of biochemical reaction mechanisms. Consequently,
the insights into the pathways described in this chapter have been obtained almost
exclusively by this laborious but also highly evidential method.
BIOSYNTHESIS OF HYDROLYZABLE TANNINS
The above outlined structural characteristics of hydrolyzable tannins

suggest that their biosynthesis must involve many reaction steps that contribute
to an extended biochemical pathway. With regard to a clear overview, the entire
biogenetic route can be conveniently subdivided into several sections which com-
prise individual specific challenges, in particular (i) the origin of gallic acid as
the principal phenolic constituent; (ii) the formation of ~-glucogallin (cf. 11 in
Fig. 3) as the first specific precursor of this pathway; (iii) the transformation of
this monoester to pentagalloylglucose (2), including questions on the role of
activated acyl donors and the exact structure of intermediates; (iv) the galloyla-
tion reactions involved in the conversion of pentagalloylglucose to complex
188 G. G. GROSS
gallotannins; and finally (v) the oxidative processes leading from pentagalloyl-
glucose to monomeric and oligomeric ellagitannins. The results obtained from
enzyme studies are reported in this section.
Origin of Gallic Acid
It is now generally accepted that benzoic acids (C 6 C 1 acids) are produced

in higher plants by side-chain degradation of cinnamic acids through a cinnamoyl-
CoA dependent ~-oxidation sequence. 6 However, particular problems have always
been encountered with respect to gallic acid (1). Despite numerous investigations,
the biosynthesis of this widespread plant constituent is still not yet fully clarified.
Rather conventional but unproven pathways involving side-chain degradation of
3,4,5-trihydroxycinnamic (8)7 or caffeic acid (7)8 to yield gallic acid (1) were for-
mulated many years ago (cf. Fig. 2, routes a and b). As a result of tracer experi-
ments with the fungus Phycomyces and various higher plants, a quite different
pathway (Fig. 2, route c/c') was proposed by others which postulated the direct
aromatization of shikimic acid or a biogenetically closely related compound, most
likely 5-dehydroshikimic acid (9).9.10
Enzyme studies, generally regarded as providing more reliable statements,
were not very helpful. Short communications on work with cell-free systems from
Cinnamic acid (6)

Caffeic acid (7) Trihydroxycinnamic
acid (8)
(b)
oA, 6H
5-Dehydroshikimic
(c')
0"
YOH
OH
Protocatechuic
(b, c') ))H
HOYOH
OH
Gallic acid (1)
acid (9) acid (10)
I
(c)
Figure 2. Proposed biosynthetic pathways to gallic acid.

mung bean seedlings II and leaves from Pelargonium 12 suggested the sequence
5-dehydroshikimic (9) ~ protocatechuic (10) ~ gallic acid (1) (route c' in
Fig. 2), but these preliminary results were never corroborated by detailed
investigations.
Ambiguous results regarding the two alternatives were obtained after
feeding experiments with carboxyl-labeled shikimic acid (the biosynthetic route
to gallic acid via C 6C3 compounds involves loss of the radioactive carboxyl group,
while it is retained in the direct aromatization) which suggested the existence of
two routes to gallic acid, the preferential one depending on the plant species and
its developmental status. 13- 16 Experiments with the herbicides L-AOPP (L-2-
aminooxy-3-phenylpropionic acid, an inhibitor of the deamination of L-
phenylalanine to cinnamic acid 6) and glyphosate (N-(phosphonomethyl)glycine,
a phosphoenolpyruvate analog that blocks the activity of 5-enolpyruvylshikimate
dehydrogenase) supported the direct aromatization of shikimate or shikimate
precursors to gallic acid and related phenolics. '7 . '8
Summarizing the conflicting evidence, it appears presently most plausible
to regard the direct aromatization of 5-dehydroshikimic acid (9) at least as a
significant, if not the predominant, route to gallic acid. Strong evidence support-
ing this conclusion has recently been obtained by feeding [13C]glucose to cultures
of the fungus Phycomyces blakesleeanus and to leaves of the dicotyledonous tree
Rhus typhina, followed by determination of isotope distributions of isolated gallic
acid and aromatic amino acids. '9 Interpretation of the resulting isotopomer pat-
terns by a retrobiosynthetic approach showed that gallic acid was derived in both
species from an early intermediate of the shikimate pathway, most probably 5-
dehydroshikimate (9). Notably, the carboxyl group of gallic acid was found to
originate from a C6C I intermediate of the shikimate pathway and not from the
side-chain of a C6C3 metabolite, thus ruling out this latter route as a major
pathway. It was concluded that dehydrogenation of 5-dehydroshikimate (9)
in both the fungus and the plant was the predominant pathway to gallic acid
(Fig. 2, route c) but the alternative route c' (Fig. 2) via protocatechuic acid (10)
could not be excluded on the basis of the available data.
Formation of ~-GlucogalIin
~-Glucogallin (l-O-galloyl-~-D-glucose, 11) has been known as a natural

constituent of the gallotannin from Rheum oiJicinale since the beginning of our
century (ceo) and was long ago proposed as the primary specific metabolite in
the biosynthesis of hydrolyzable tannins.21 Regarding the biosynthesis of this
ester, it was initially thought that a carboxyl-activated derivative of gallic acid
would participate, galloyl-coenzyme A being the most likely candidate by analogy
to the well known caffeoyl-CoA-dependent formation of chlorogenic acid and
related depsides.22. 23 Galloyl-CoA was therefore synthesized via the N-
hydroxysuccinimidyl derivative of 4-0-~-D-glucosidogallic acid, 24 but it was
190 G.G.GROSS
HO* ~
eOOH
OH
I
OH
Gallic acid (1)

UDP-Glucose
• \
'\
UDP
b
H~O'e ~
HO
OH
0
OH II
o
~
I1-Glucogallin (11)
OH
I
OH
OH
Figure 3. Biosynthesis of p-glucogallin (l-O-galloyl-p-D-glucose). UDP, uridine-5'-

diphosphate.
soon recognized that this thioester was not involved in the biosynthesis of 13-
glucogallin.25 It was found instead that a glucosyltransferase from oak leaves
catalyzed the efficient esterification of free gallic acid with UDP-glucose as the
energy-rich component, yielding f3-glucogallin and related I-O-acyl-f3-D-
glucoses (Fig. 3).26-2R Numerous analogous enzymes catalyzing the formation of
phenolic I-O-acylglucoses were isolated from various plant sources in the mean-
time, thus providing evidence that UDP-glucose functions as the general activated
donor in the esterification of glucose with phenolic acids.
The Pathway to PentagalloyJgJucose
The main pathway. It was apparent from early investigations on the biosyn-
thesis of gallotannins that crude extracts from young oak leaves produced digal-
loylglucose and trigalloylglucose in assay mixtures which contained f3-glucogallin
(11) as sole substrate. Again, as in the preceding formation of f3-glucogallin, no
requirement for galloyl-CoA was observed in these reactions. This surprising
result could be interpreted only with the assumption that f3-glucogallin exerted
an unexpected dual role, evidently acting not only as acceptor substrate but also
as the acyl donor required for such a transformation. 25 Considering the then
known compararatively low group-transfer potential ~Go' of acylglucoses (e.g.
glucose-I-phosphate, ca 21 kJ mol-I; glucose-6-phosphate, ca 10.5 to 12.5 kJ
mol-I ),29 the existence of such an enzyme activity was surprising. This question
was solved meanwhile by the finding that the related cinnamoy I ester, 1-0-
sinapoyl-f3-D-glucose, has an unexpectedly high ~Go' of 35.7kJmol-I.30 i.e. a
value that is comparable to well-known data for acyl-CoA thioesters (ca. 36 kJ
mol-I). It is reasonable to assume that the ~Go' of f3-glucogallin (11) is on the
same order of magnitude.
Subsequent investigations led to the isolation of an acyl transferase from oak
leaves that catalyzed the "disproportionation" of f3-glucogallin (11) by which the
galloyl moiety of the donor was transferred to the glucose-6-0H of the acceptor,
yielding 1,6-di-0-galloyl-f3-D-glucose (12) and free glucose as a deacylated
by-product (Fig. 4V I Two structurally closely related esters, 1-0-
protocatechuoyl- and l-O-p-hydroxybenzoyl-p-D-glucose, were also accepted as

substrates by this enzyme. 32
Analogous "disproportionation" reactions have been reported for the
biosynthesis of 1,2-disinapoylglucose in radish seedlings,33.3 4 3,5-dicaffeoylquinic
acid in sweet potato,35.36 and 1,3-di-isobutyroylglucose (Fig. 7, 17) in wild
tomato. 37 These observations support results from many other laboratories on the
role of l-O-acylglucoses as donors for the acylation of a wide array of aliphatic
and aromatic acceptors, including the formation of chi orogenic acid (caffeoyl
quinate) via caffeoylglucose as an alternative to the long established acyl-CoA
dependent synthesis 22 of this depside (references in 38 ). It is apparent that 1-0-
acylglucose esters, often regarded in the past as metabolically inert compounds,
occupy a prominent position in the secondary metabolism of higher plants which
is comparable to that of the generally acknowledged role of acyl-CoA esters.
It had been proposed on the basis of earlier studies 25 that the biosynthesis of
trigalloylglucose(s) should follow the same reaction mechanism as described above
for the formation of 1,6-digalloylglucose, i.e. by transacylation with p-glucogallin
serving as the donor substrate. This assumption was proven with an enzyme iso-
lated from leaves of staghorn sumac (Rhus typhina) that specifically catalyzed the
galloylation of the 2-0H of 1,6-digalloylglucose (12) to yield 1,2,6-tri-0-galloyl-
P-D-glucose (13) (Fig. 4).39 Again, as in the preceding step, the p-glucogallin
analogs, I-O-p-hydroxybenzoylglucose and l-O-protocatechuoylglucose,
replaced p-glucogallin as effective acyl donors, and tri-O-protocatechuoylglucose
was formed by this enzyme upon incubation of l-O-protocatechuoylglucose with
1,6-di-0-protocatechuoylglucose as acceptor. 40
Considering the above results, it is almost trivial to report that p-glucogallin
(11) also served as the galloyl donor in the subsequent acylation of 1,2,6-tri-0-
galloyl-p-glucose (13) to 1,2,3,6-tetra-0-galloyl-p-D-glucose (14), followed by
acylation of this intermediate to 1,2,3,4,6-penta-0-galloyl-p-D-glucose (2) (Fig.
4). The enzyme catalyzing the first of these steps was discovered in sumac leaves
and partially purified from green acorns of pedunculate oak (Quercus robur, syn.
Q. pedunculata). In addition to the natural substrate, 1,2,6-trigalloylglucose (13),
its isomer 1,3,6-trigalloylglucose-which is not an intermediate in the biosyn-
thesis of hydrolyzable tannins in oak or sumac-was an extremely efficient accep-
tor. However, 1,2,3,6-tetragalloylglucose (14) was the sole reaction product with
both substrates. 41
The final enzyme ofthe pathway, catalyzing the formation of 1,2,3,4,6-penta-
O-galloyl-p-D-glucose (2), was isolated from young oak leaves 42 and was recently
purified more than 1,OOO-fold to apparent homogeneity (Grundhi:ifer and Gross,
unpublished results). This acyltransferase strictly depended on 1,2,3,6-
tetragalloylglucose (14) as the acceptor, whereas the I ,2,4,6-isomer was inactive.
Side reactions. A summary of the above described enzyme studies suggests

a straight forward-directed metabolic pathway that starts with p-glucogallin (11)
'D
tv
H
° OH
71
O,;:;-c ~ OH oH
Lr tc
I1-Glucogallin (11) OH I1-Glucogallin (11)
",~O, ~O" oI ~
\. . \. .
HO~O,~ '\ HO~O,C 1 h OH ~
ft OH OH Glucose HO OH II Glucose
o o
I!-Glucogallin (11) 1,6-Di-O-galloyl-I!-D-glucose (12)
H OH
° OH OH
71
Lr h71
0, ~ OH OH
I1-Glucogallin (11)
O~j ~ 0" Lo"
'?o
tc ~ oH
1 h OH
\. .
HO~O,C '\
Glucose
HO 0
/
II
0
"O~o"JCloo
/
o /
0 II
0
O~~O" °;1 O~hOH

~ OH R
HO OH
HO OH o
OH HO o
1,2,6-Tri-O-galloyl-I!-D-glucose (13)
*
1,2,3,6-Tetra-O-galloyl-I!-D-glucose (14) @
rJ)
rJ)
1:0
o
[JJ
~
:r:
tTl
[JJ
po
1:0
H §
° OH
tTl
a
0.;::.
o
ix
T
~ I
OH
~
Ci
r..-Glucogallin (11) ?:i
HO:y1 g,o~O\ OH
"-- :"\-
tr: ~
HO~ p~O I ~
Glucose
OH *O=:/ p 'ft h OH Ci
:r:
7 O=CQ- 0 -<
~ I ~
HO OH
r;_ OH ~
OH HO OH
~
N
>-
1:0
l'
1,2,3,4,6-Penta-O-galloyl-B-D-glucose (2) tTl
~
Figure 4. The biosynthetic pathway from l3-glucogallin to pentagalloylglucose. ~
z
[JJ
'D
Y>
194 G. G. GROSS
and ends with pentagalloylglucose (2) as final product. Unfortunately, rami-

fications and side-reactions were encountered that did not fit this clear and attrac-
tive picture. At the beginning of this investigation, an unusual exchange reaction
between ~-glucogallin (11) and free glucose was found to occur with cell-free
extracts of oak leaves (Fig. 5).43 This reaction was catalyzed by a galloyltrans-
ferase with a specific requirement for D-glucose as the acceptor, while a variety
of phenolic l-O-acyl-~-D-glucoses could serve as donor substrates. 44 The natural
significance of this enzyme is still obscure; however, it was successfully employed
for the convenient and economic preparation of labeled ~-glucogallin.44
Other inconsistencies were encountered in the purification of the galloyl-
transferase catalyzing the formation of 1,2,6-trigalloylglucose (13). Initial unex-
plained problems were clarified only after recognizing the unexpected existence
of an interferring enzyme that effected the efficient formation of the same
product, 1,2,6-trigalloylglucose, but evidently without any requirement for the
established acyl donor, ~-glucogallin (11 ).45 As depicted in Figure 6, this enzyme
represents another example of a "disproportionation" reaction, in this instance,
however, with the participation of two molecules of 1,6-digalloylglucose
(12) which are converted to 1,2,6-trigalloylglucose (13) and anomeric 6-
galloylglucose as a partially deacylated by-product.
As a consequence of this "~-glucogallin-independent" transacylation, it was
concluded that not only monogalloylglucoses but also higher substituted deriva-
tives could act as acyl donors, as long as they possessed the energetically indis-
pensable l-O-galloyl group. This view was supported by a series of substrate
specificity studies which, however, also revealed that the reactivity of the higher
galloylated analogs was drastically diminished, most likely because of steric
HO~OH 0 16=0H OH
HO~O,
OH
o ~ I +
HO
OH
'c
II
OH HO~
o
I1-Glucogallin (11) ['4CjGlucose
jr OH
OH
HO~O\
H~O'C ~ I
HO 0 :;/ OH
OH 16=
HO~OH + OH
OH II
o
Glucose ['4CjI1-Glucogallin
Figure 5. Enzymatic galloyl-exchange between ~-glucogallin and glucose. Bold lines symbol-
ize the presence of an appropriate label (e.g. 14e) as prerequisite for the detection and
quantification of this reaction.
ttl
5
[/J
i
-l
::c
tr1
[/J
:0
ttl
5
vtr1
OH OH OH
OH OH OH Cl
r r r
l l l s::
o"c ~ OH v
0" ~ OH OH
OH
0"
'?
n ~ OH OH
OH
I
OO
n ~
+ 1,6-Digalloylglucose
'?o "'" HO
nh h
o "'" + ~ ~
h OH HO~O\ o'c I h OH
HO OH HO~II HO OH OH
HO~o,cg I ~
o 0 v
o=~" ~
1,6-Di-O-galloyl-f1.D-glucose (12) 6-0-Galloylglucose
v
13
HO OH
t<
N
;J>
1,2,6-Tri-O-galloyl-f1-D-glucose (13) ttl
r'
tr1
Figure 6. ~-Glucogallin "independent" enzymatic "disproportionation" of I ,6-digalloylglucose to 1,2,6-trigalloylglucose. ~
~
z
[/J
U>
""
196 G. G. GROSS
hindrance as a consequence of increasing bulkiness of the molecule due to

increasing substitution degrees, thus leaving I-mono- and 1,6-di-esters as the
predominating galloyl donors. 45
Galloylation of 1,6-digalloylglucose (12) by enzyme extracts from Rhus
was found to yield trace amounts of 1,3,6-trigalloylglucose, besides the 1,2,6-
substituted main product (13). Although enzyme assays had shown that this by-
product was efficiently transformed to 1,2,3,6-tetragalloylglucose (14) in the
subsequent step,41 it was impossible to assign any importance to this alternative
in vivo, due to the negligible supply of precursor. In addition to the main product
of this step, 1,2,3,6-tetragalloylglucose (14), cell-free extracts from sumac leaves
also were found to produce a certain amount of the 1,2,4,6-isomer as a by-
product; this side-reaction represented only a dead end, however, because the
enzyme catalyzing the subsequent step to the pentagalloylglucose level displayed
no affinity towards this compound. 42
In summary, it is evident that the significance of all these side-reactions is
negligible in comparison to the importance of the reactions involved in the "main
pathway", suggesting that pentagalloylglucose is produced in nature only accord-
ing to the sequence depicted in Figure 4.
General Characteristics of the Pathway. The above reported enzyme

studies have provided evidence of a linear pathway from gallic acid to 1,2,3,4,6-
pentagalloylglucose (2), the immediate precursor of both gallotannins and ellag-
itannins. One of the most striking discoveries in this connection was certainly the
surprisingly pronounced position-specificity of the individual enzyme-catalyzed
galloylation steps in this sequence. Interestingly, an identical sequence has been
reported for the chemical esterification of the hydroxyl groups of glucose in
studies with I-benzyl- or I-methyl-~-D-glucopyranose. This is explained by the
fact that, after the preferred semiacetal-OH at C-I, the primary 6-0H is more
reactive than the residual secondary hydroxy Is, the 2-0H among them being the
most reactive, due to an activating effect of the neighboring anomeric center. The
theoretically equivalent hydroxyls at C-3 and C-4 are discriminated against by
steric hindrance of access to the 4-0H group adjacent to the already substituted
bulky 6-position, resulting in a higher relative activity of the 3_0H.46.47
Most of the above described transferases were isolated from oak (Quercus
robur, Q. rubra), while sumac (Rhus typhina) was used only sporadically as an
enzyme source. Supplementary experiments were carried out, however, which
determined that the pathway to pentagalloylglucose was identical in both plant
groups. Comparison of the available data on the galloyltransferases involved in
this metabolic route (cf. Fig. 4) revealed a pronounced uniformity of their basic
properties, e.g. their optimal reactivity and stability in slightly acidic media (pH
4-6), their low 010 values (1.8-2.0) and sensitivity to higher temperatures, their
unusual cold tolerance, as expressed by residudal reaction rates of 10-25% even
at O°C, and their pronounced trend to unusually high molecular weights of about
260,000-450,000 daltons. An exception was the "~-glucogallin-independent" 1,6-

digalloylglucose "disproportionating" galloyltransferase (cf. Fig. 6)34 with a mol-
ecular weight of 56,000 daltons and a QIO value of 3.0, suggesting that this
enzyme does not belong on the main route to pentagalloylglucose.
It was a major concern during the progress ofthe enzyme studies that higher
galloylated substrates and products might prevent activity determinations in in
vitro assays due to their increasing tanning potential. While mono and digalloyl-
glucoses were inactive, significantly increasing precipitation of proteins (e.g.
hemoglobin20 or bovine serum albumin 48 ), inhibition of enzymes (e.g. ~-glucosi
dase 20 ), and complexation with basic or phenol-rich compounds (e.g. caffein 20 or
the protein-binding dye Coomassie brilliant blue 49 ) have been reported for three-
fold and higher galloylated glucose derivatives. Fortunately, it became evident
that the gallotannin synthesizing enzymes are resistant to the negative influences
of their polyphenolic substrates and products.
The pathway to pentagalloylglucose, as summarized in Figure 4, has long
been assumed to be unique. However, a reaction pattern with striking similarity to
the biosynthesis of galloylglucoses recently has been reported for the biosynthe-
sis of various isobutyroyl-~-D-glucoses in wild tomato (Lycopersicon penellii)
leaves. 37.50,51 As depicted in Figure 7, a UDP-glucose dependent glucosyltransferase
catalyzes the transformation of free isobutyric acid to l-O-isobutyroyl-~-D-glu
copyranose (15), and this metabolite has been found to serve as a donor for
subsequent trans acylation reactions yielding 1,3-di-, 1,2,3-tri-, and 1,2,3,4-tetra-
O-isobutyroylglucose. Also, acyl exchange reactions analogous to the reaction
depicted in Figure 5 have been observed with cell-free extracts from tomato. 37
Considering these results, it must be concluded that not only do phenolic 1-0-
acylglucoses occupy a central role as activated intermediates in secondary metab-
olism, but also that this activity must be ascribed to aliphatic analogs, thus cor-
roborating the presumed general importance of these ester acetals.
Biosynthesis of Complex Gallotannins
The transition from "simple" galloylglucoses to complex gallotannins is

marked by the addition offurther galloyl residues to 1,2,3,4,6-pentagalloyl-glucose
(2) to yield their characteristic meta-depside groups (5). Formally, this process can
be regarded as a continuation of the esterification reactions ofthe preceding steps;
it must be emphasized, however, that gallic acid now combines with phenolic
hydroxyls whose chemical properties are significantly different from those of the
aliphatic OH-groups of glucose. It was thus interesting to discover that cell-free
extracts from sumac (R. typhina) leaves catalyzed the acylation of pentagalloyl-
glucose exactly according to the same reaction mechnism as above, i.e. by
utilizing ~-glucogallin (11) as a specific galloyl donor. In these studies, the sequen-
tial galloylation of pentagalloylglucose to hexa-, hepta-, and octagalloylg-
lucoses was observed (Fig. 8), together with minor to trace amounts of nona- to
198 G. G. GROSS
+ UDP-Glucose
-UDP
"~~ OH \I
o
Isobutyric acid 1-0-lsobutyroyl-B-D-glucose (15)
~ OH II
o
+ (15), - Glc
\I
HO~\
Ac~~'c
o
II
0
~
1-0-lsobutyroyl-B-D-glucose (17) 1,3-Di-O-isobutyroyl-B-D-glucose
1. (15), - G"
+ (15), - Glc
1,2,3,4-Tetra-O-isobutyroyl-B-D-glucose 1,2,3-Tri-O-isobutyroyl-B-D-glucose
Figure 7. Biosynthesis of I-O-isobutryroyl-B-D-glucose in wild tomato (Lycopersicon penel-

Iii) and its transformation to di-, tri- and tetra-O-isobutyroyl-B-D-glucose esters. UDp, uridine-
5'-diphosphate; GIc, glucose.
o 2 3
Incubation time (h)
Figure 8. Enzymatic synthesis of complex gallotannins by galloylation of 1,2,3,4,6-pentagalloy-

Iglucose with cell-free extracts from leaves of Rhus typhina. (.), Hexa-, (0), hepta-, (_),
octagalloylglucose.
undecagalloylglucoses. 52 The gallotannin nature of the most abundant enzyme

reaction products, hex a- to octagalloylglucose, was proven by hydrolysis with
fungal tannase yielding gallic acid as the sole phenolic component, and by
methanolysis (which exclusively cleaves depside bonds) affording 1,2,3,4,6-
pentagalloylglucose (2) and methyl gallate in molar ratios of 1 : I, 1 : 2, and 1 : 3,
respectively.
Unequivocal structural proof of the reaction products was obtained
when three hexagalloylglucoses and three heptagalloylglucoses, isolated from
scaled-up enzyme assays with cell-free extracts from Rhus typhina, were
analyzed by lH and l3C NMR. It is important to note that the identified
structures, depicted in Figures 9 (16-19) and 10 (20, 21), and within certain
limits also the relative amounts of these in vitro reaction products, were
identical to those of the in vivo formed gallotannins from the related species R.
semialata. 3
More detailed studies revealed the existence of several isoenzymes in
sumac (R. typhina) that catalyzed the in vitro acylation ofpentagalloylglucose (2)
to higher substituted derivatives. Among these, three galloyltransferases (trans-
ferases A, B, and C) were separated according to their different molecular weights
that ranged from ca 170,000 to 360,000. Galloyltransferases A and B, character-
ized by molecular weights of 360,000 and 260,000, respectively, were found to
preferentially acylate the 4-position of pentagalloylglucose (2) to hexagalloyl-
glucose (16), followed by substitution of the 2-position of (16) to heptagalloyl-
glucose (17) (Fig. 9). Minor reactivities of these isoenzymes led to the sequence
pentagalloylglucose (2) 4 hexagalloylglucose (18) 4 heptagalloylglucose (17),
plus a trace activity towards heptagalloylglucose (19). Significant differences
between transferases A and B, apart from their molecular weights, were observed
for their temperature and pH sensitivity, enzyme A being less susceptible to
these factors than enzyme B. In summary, both isoenzymes promoted galloy-
lation at positions 2 and 4 of the galloylglucose core (Niemetz and Gross,
unpublished results).
Galloyltransferase C (Mf 170,000) has been purified to apparent homo-
geneity and was found to consist of four identical subunits of. Mf 42,000. 53 Sub-
strate specificity studies showed that pentagalloylglucose (2) was preferentially
galloylated to the hexagalloylglucose, 3-0-digalloyl-1 ,2,4,6-tetra-O-galloyl-~-D
glucose (20), and the heptagalloylglucose, 3-0-trigalloyl-I,2,4,6-tetra-O-galloyl-
(3-D-glucose (21). Thus, galloyltransferase C is specific in acylating penta- and
hexagalloylglucoses in the 3-position of the glucose core (Fig. 10). In summary,
the specificities of these three enzymes support earlier findings that the C-I and
C-6 positions of "Chinese gallotannin" from Rhus generally remain free of
depsidic substituents,3 in contrast to the gallotannins from Quercus infectoria 4
("Turkish gallotannin") or Paeonia lactiflora 2 .5 where depside bonds occur also
at C-6 of the glucose core.
7 OH
N
O""C :::,... , o
o
h OH
~
OH
O=C ~
rt/
cU OH
:y*~.~ HO h OH
~'"
H HO OH
1 ,2,3,4,6-Penta-O-galloyl-B-D-glucose (2)
OH H
7, ~ OH
""c
0 ... :::,... OH
7,°
Lr
OH ~O"":::"" OH H
O
?i
h h/ ~ OH H 7 0II '-':: OH
HO C ' o 4...... c ' ~ OH ~ 7

' PC '0 , h
' 0 II ,=,
HO :::,... <> 0
Y4 OH
O=C~ _ 0
",:QtX(,. '&~
, "'" ~ Ii o. . . c~
h OH
h-OH
HO OH H
HO OH HO OH*., o
H HO OH
HO ~
OH
OH
4= o
a
4-0-Digalloyl-1,2,3,6-tetra-O- ~
[/J
[/J
2-0-Digalloyl-1,3,4,6-tetra-O- galloyl-B-D-glucose (16)
galloyl-B-D-glucose (18)
IJj
25
[/J
~
::r:
tTl
[/J
.2ii
IJj
,
, OH
~ 25
71 otTl
O""~::-... OH Cl
o
lrr: LOH
-:? oII 6 ~ OH
~
71
HO~~'o HO II
~ ~C/°=y1 c'o~o.....c I II h OH ~
U OH
HO *O~~::-"'H ~
h .~
~
HO
o ~H *o~~ =Q-fio~ ~o o
1-& ::r:
o 0 OH >-<:
HO OH 7 o
~~*O~O
HO OH
.
7 I OH
h OH HO I
* es
HO::-'" OH
~" HO :::-.. OH ~
N
HOYOH OH H >-
IJj
OH r
tTl
2,3-Di-O-digalloyl-1,4,6-tri-O- 2,4-Di-O-digalloyl-1,3,6-tri-O- ~
galloyl-B-D-glucose (19) galloyl-B-D-glucose (17) ~
Z
[/J
Figure 9. Acylation of pentagalloylglucose (2) to hexagalloylglucoses (16, 18) and heptagalloylglucoses (17, 19) catalyzed by galloyl-
transferases A and B (cf. text) from leaves of Rhus typhina. ~-Glucogallin (11) serves as general galloyl donor in these transformations.
Main reactions are symbolized by bold arrows, minor reactions by thin arrows; the dashed arrow symbolizes trace reactivity.
N
o
OH OH
OH OH
0" :,... 1
tv
0" :,... 1 o
HO
o
II
C
6
a OH
OH HO
o
II
C
6
a OH
OH
tv
1 0 0 '" OH 1 0 ",OH
HO:"" 0 ~c""'O 1
o. . . c,.",-::. HO:"" 1
Y C'~
OHA II
tc OH
YC'~ 0 ~c/o o. . . c
tc h-
OH o=c ~ OH
HO~OH o=h-o: + Il-Glucogallin (11) I'" -
OH HO)--\H - Glc
O"c/ O
A
* : OH Q-OH
HO OH
1,2,3,4,6-Penta-0-galloyl- HOVOH 3-0-Digalloyl-1,2,4,6-tetra-0-

B-D-glucose (2) OH galloyl-B-D-glucose (20)
OH
OH
O~ :,... 1
HO
o
II
'<:c
6
a OH
OH
1 0 ",OH
HO
c,~
:,...
YO~c"""O O'c 1
tc fi
OH* o=c 0II OH
+ Il-Glucogallin (11) I'" -

O"c/O h- OH h--OH
- Glc OH HoM
r OH
1
HO :,... 0, ..... 0
OH*C/
* HO
:,...1
OH
3-0-Trigalloyl-1 ,2,4,6-tetra-0-
OH galloyl-r..-D-glucose (21) o
o
Figure 10. Transformation of 1,2,3,4,6-pentagalloylglucose (2) to 3-0-digalloyl-I,2,4,6-tetra-0-galloyl-~-D-glucose (20) and 3-0- Cl
trigalloyl-l,2,4,6-tetra-O-galloyl-~-D-glucose (21) by galloyitransferase C from leaves of Rhus typhina (see text for details). 23
Vl
Vl
Formation of Ellagitannins
In contrast to the limited distribution of gallotannins, ellagitannins occur in

many plant families where they display enormous structural diversity owing to
several possible binding sites of the hexahydroxydiphenoyl (HHDP) (3) bridges
at the pentagalloylglucose core, and particularly by their strong tendency to form
a multitude of dimeric and oligomeric derivatives. The challenge of this struc-
tural diversity traditionally has attracted the interest of chemists for decades;
extreme attention, however, is being paid currently to these compounds because
of their manifest biological activities, particularly their recently recognized
antimutagenic, anticarcinogenic or virus inhibiting effects. 54-56 This role as
promising chemotherapeutic agents has stimulated successful efforts on the
chemical synthesis of ellagitannins (for a recent review, see57 ).
Unfortunately, no similar progress can be recorded for the biosynthesis of
ellagitannins. It is generally assumed that the 3,4,5,3',4',5' -hexahydroxydiphenoyl
(HHDP, 3) residues of ellagitannins originate from the dehydrogenation of
spatially adjacent galloyl groups of 1,2,3,4,6-pentagalloylglucose (2). For the
common 4C I configuration of pentagalloylglucose, for instance, a metabolic
sequence has been proposed that leads to tellimagrandin II (22) and casuarictin
(23) as primary ellagitannin metabolites (Fig. 11).20.58 Numerous attempts
to unravel the mechanism of these oxidations have been published in the past
(for reviews, cf. 38.59,00); only free ellagic acid (4), however, was formed from
a variety of galloylated precursors in these experiments, while the occurrence
of true ellagitannins, characterized by glucose-bound HHDP (3) residues, was
never observed.
To increase the sensitivity of enzyme assays, [U-1 4C)pentagal\oylglucose
was prepared by photoassimilation of 14C02 in leaves of Rhus typhina. 61 With this
labeled substrate, a soluble protein fraction was isolated from leaves of the weed
Tellima grandiflora (Saxifragaceae) which suggested the oxidation of pentagal-
loylglucose (2) to tellimagrandin II (22) in the presence of FMN as a hydrogen
acceptor. 62 Severe inconsistencies were encountered upon detailed analyses,
however, and a critical reinvestigation of the system finally led to the conclusion
that the reaction product had been artificially formed in vitro, most likely by non-
enzymatic oxidation reactions and contamination of the enzyme preparation by
tight association of protein with in vivo biosynthesized ellagitannins. It has, thus,
to be stated that the old and challenging question of the mechism of ellagitannin
biosynthesis is still unanswered.
DEGRADATION OF GALLOYLGLUCOSES
In the course of investigations on the biosynthesis of gallotannins, the exis-

tence of a contaminating esterase was encountered that degraded both galloyl-
N
AOH ~
OH
O~C~OH H O Ig 0II OH
o I HO ~ C.......... O_CH, '.;:: OH
H0r7Y~'-O~O\ AOH HO r ~_O~O\ O'-COOH
HOY 0" /O~O'-C~OH -2[H] HO ~ I 0" / O~II / 0
OH "c 0 II
• OH 'c O=~
~6" O=b"O A \-y0H

OH HO OH HOVOH H~ OH
OH
1,2,3,4,6-Penlagalloylglucose (2) Tellimagrandin II (22)
2[H]
j-
Ho~OHIR
HO ~ c .......... O-CH, OH
OOH
o ~o\ I
HO r ~_o o~o,-c /0. OH
0 0 II
HO~ I \ / 0
OH co OC
HOOOOH
HO HO HO OH
o
CI
Casuariclin (23)
CI
Figure 11. Proposed steps in the oxidation of 1,2,3,4,6-penta-O-galloyl-f3-D-glucose (4C\ configuration) to monomeric
(5
C/l
C/l
ellagitannins by 4-6 and 2-3 coupling.
glucose substrates and enzymatically formed reaction products. 32 In studies on

the nature of this disturbing protein, the enzyme was purified from leaves of Q.
robur (pedunculate oak) more than 1900-fold to apparent homogeneity. From gel-
filtration experiments with Sephadex 0-200, a molecular weight of 300,000 was
estimated, while HPLC gel-filtration on a OPC-diol superformance column
(Merck) indicated a Mr of only 150,000. Two protein bands also were detected
after native electrophoresis on polyacrylamide-gel, while denaturing elec-
trophoresis in the presence of sodium dodecylsulfate revealed the existence of a
single polypeptide band of Mr 75,000. It was concluded that the native enzyme
preferentially existed as a tetramer of four apparently identical subunits in slightly
acidic medium, while it dissociated partially or completely into still highly
reactive dimers under the more alkaline conditions employed for OPC-diol
chromatography or native PAOE. 63
Detailed studies were carried out on the substrate-specificity of the enzyme
with a wide array of pure compounds. This hydrolase was inactive with different
ring-substituted methyl cinnamates, I-O-sinapoylglucose, and substrates with
nitro-substitution of the aromatic acid moiety. Hydrolysis occurred with simple
galloyl esters (methyl, ethyl, propyl gallate), naphthyl acetate (but not naphthyl
propionate or butyrate), mono- to hexa-substituted galloyl-~-D-glucoses, vari-
ously ring-substituted I-O-benzoyl-~-D-glucoses, and depsides such as meta-
digallic acid (5) or chi orogenic acid. The enzyme exhibited both pronounced
esterase and depsidase activities, i.e. properties which are traditionally ascribed
to fungal tannases. Other characteristics of this plant enzyme, e.g. pH and
temperature optima and its high molecular weight, closely resembled those of
microbial tannases (CC 8•64 ). By analogy to these enzymes, the esterase from oak
leaves was classified as a new "tannase" of plant origin.
There remains the question of an ecological role of such an enzyme in green
leaves. It has been recognized that condensed and hydrolyzable tannins contribute
to the defense of plants against ruminants (e.g. 65.66), and the existence of a tannin-
degrading enzyme in leaves would counteract this effect. In contrast to herbivo-
rous mammals, however, the situation with insects could be quite different. The
feeding deterrent role traditionally ascribed to tannins due to their astringency,
causing reduced palatability of plant parts, has been questioned; e.g. evidence has
been presented that the ellagitannin geraniin preferentially acts as a protoxin that
releases insect growth inhibitors, particularly ellagic acid, upon hydrolytic cleav-
age. 67 According to this view, hydrolyzable tannins could playa dual protective role
in plant-herbivore interactions, not only by direct protection against herbivorous
animals, but also indirectly in the form of their degradation products. The occur-
rence of tannase in green leaves could, thus, contribute to the latter process by
release of harmful products after cellular decompartmentalization under the attack
of an insect predator which brings the enzyme into contact with its tannin
substrates. Such protoxin-based defense strategies are well-documented for
ortho-coumaroyl glucosides, cyanogenic glucosides, and glucosinolates,68 and it is
206 G. G. GROSS
conceivable to add the system tannin-"tannase" to that list of chemical weapons in

higher plants.
LOCALIZATION OF GALLOTANNINS
In the course of the above investigations, questions regarding the intracellu-

lar localization of hydrolyzable tannins in those plant tissues which were used as
enzyme sources became increasingly important, and this problem was later
extended to the cellular distribution ofthe enzymes involved in their biosynthesis.
In general, much attention has been directed by several laboratories to the subcel-
lular compartmentation of polyphenolic compounds, addressing plastids, dic-
tyosomes, smooth or rough endoplasmatic reticulum, or vacuoles as possible
deposition sites (references in 69 ). However, it is common to all of these reports that
they suffer from the lack of specific reagents for the visualization of hydrolyzable
tannins, not allowing their unequivocal discrimination from other phenolic plant
constituents. One possible solution to this problem is the application of immuno-
cytochemical techniques which are known for their specificity and which, besides
their traditional use in the detection of macromolecules, have recently gained
increasing importance in the localization oflow molecular weight compounds. 70
For this purpose, a conjugate of 1,2,3,4,6-pentagalloylglucose (2), the
common and immediate precursor of both subclasses of hydrolyzable tannins, i.e.
gallotannins and ellagitannins, with keyhole limpet hemocyanin as a macromol-
ecular carrier, was used as an antigen for the production of polyclonal antibod-
ies in rabbits. Crude antisera were purified by affinity chromatography on
pentagalloylglucose-substituted Sepharose; the resulting mono specific antibodies
displayed strong affinity towards the antigen, pentagalloylglucose (2), and also
towards more complex gallotannins (e.g. 16-21) and ellagitannins, such as pedun-
culagin or vescalagin, while they were inactive towards monomeric and dimeric
condensed tannins (proanthocyanidins), low-molecular phenolics (e.g. 1, 4, 11)
and aliphatic low and high molecular weight carbohydrate polyols.69 Transmis-
sion electron microscopy of ultra-thin sections from caffeine-pretreated tissues
(to prevent leaching of phenolics into the cytoplasm 71 ), followed by glutaralde-
hyde/Os04 fixation, revealed that antibody-specific immunogold labeling was
exclusively associated with the walls of Rhus or Quercus leaf cells, while the
cytoplasmatic regions were essentially devoid of label. No labeling was observed
in controls with pre-immune serum. The results were corroborated by analogous
experiments with ultra-thin sections of freeze-cut plant material. Also, light-
microscopical examination of leaf sections, in which endogenous peroxidase was
blocked with trypsin, followed by treatment with antibody (replaced by
pre-immune serum in controls), anti-rabbit IgG-peroxidase complex, and final
staining with diaminobenzidine/CoCl 2 . led to the conclusion that hydrolyzable
tannins specifically accumulated within the cell walls of leaf tissues.
As a prerequisite for a related set of experiments, the galloyltransferase cat-

alyzing the biosynthesis of 1,2,3,4,6-pentagalloylglucose (2) from 1,2,3,6-
tetragalloylglucose (14) was purified from oak leaves to homogeneity and used
to raise antibodies against this pivotal catalyst in the biosynthesis of gallotannins
and ellagitannins (GrundhOfer and Gross, unpublished data). Crude sera were
purified by ammonium sulfate precipitation, chromatography on protein A, and
gel-filtration on Sephacryl S-300 to yield antibodies that exclusively reacted with
the antigen in immuno-blot experiments. Electron microscopy of ultra-thin sec-
tions from oak leaves, treated with anti-transferase and anti-rabbit immunogold
IgG, revealed that the enzyme was diffusely distributed over the entire undiffer-
entiated cytosol compartment of the cells, showing no association with walls,
membranes, or organelles, except for a slight label in the stroma of chloroplasts.
No labeling was observed in analogous experiments with leaves of Rhus, indi-
cating that the oak galloyltransferase significantly differed from its counterpart
in sumac leaves.
Summarizing these data, it appears that the biosynthesis of hydrolyzable
tannins occurs in the undifferentiated cytosolic compartment of oak leaves (with
the eventual participation of the chloroplast stroma), and this result nicely fits the
fact that this enzyme (as well all other enzymes of this pathway) has been
classified as "soluble". The gallotannin and ellagitannin products of these reac-
tions, on the other hand, were found to be exclusively associated with the sur-
rounding cell walls by the above reported immunological studies. This result is
in clear contrast to earlier reports on "tannin vacuoles" (e.g. 72-74). Considering the
rather unspecific staining procedures employed in these studies, however, it is
unclear what (poly)phenolic plant metabolites have actually been detected in the
investigations. It should be mentioned in this connection that control experiments
with spinach leaves, known to be devoid of hydrolyzable tannins, displayed no
immunogold labeling after treatment with anti-pentagalloylglucose or anti-
galloyltransferase IgG. On the other hand, our data were consistent with previous
observations on the immunocytochemically determined accumulation of
flavonoids in cell walls of Chrysosplenium americanum,75J6 and they are further
supported by recent considerations on chemical interactions between Iignins and
hydrolyzable tannins in wood. 77 A profound gap is encountered between our
knowledge of the cytosolic origin of hydrolyzable tannins and their final deposi-
tion in the walls of oak or sumac leaf cells. To date, no evidence for a cytosolic
compartmentation or an association with secretory transport vehicles, like Golgi
vesicles or other membrane systems, has been detected in our immunohisto-
chemical studies with anti-pentagalloylglucose. The simpliest explanation for the
currently available data is that hydrolyzable tannins are produced in the cytosol
by randomly distributed soluble enzymes, and that these cytotoxic agents are
eliminated by immediate and efficient excretion into the apoplasmatic cell wall
compartment where they accumulate, while their stationary concentration in the
cytoplasm remains below the detection limit.
208 G. G. GROSS
CONCLUSIONS AND PERSPECTIVES
On the basis of the investigations reported in this chapter, it is now possi-

ble to present a fairly detailed picture of a lengthy biogenetic sequence, starting
with a comparatively simple phenolic plant constituent, galIic acid, and ending
with highly substituted galIotannins characterized by complex chemical struc-
tures. The merit of our extensive enzyme studies in this field is that we not only
have knowledge of the individual metabolites along this pathway, but also that
we have now precisisely described the mechanisms involved in these transitions.
Unfortunately, this pleasant situation does not apply to our understanding of the
processes responsible for the oxidation of pentagalloylglucose to ellagitannins,
i.e. that subclass of hydrolyzable tannins which is not only more widespread in
nature but also which certainly will gain enormous importance in potential
medical applications in the future. It can only be hoped that fresh ideas and
sophisticated techniques will help to unravel this old but still challenging enigma.
(A similar problem applies to the biosynthesis of condensed tannins; despite many
attempts, the in vitro oxidation of monomeric precursors to dimeric or oligomeric
derivatives has never been observed. 78 Other upcoming challenges concern the
questions of the cellular localization of the biosynthetic machinery, the transport
vehicles, and the final depostion sites of hydrolyzable tannins-aspects which are
just in their infancy. It appears overdue for applying the arsenal of molecular bio-
logical methods to this field, for instance to gain insights into the possibly unique
molecular structures of the tannin-synthesizing enzymes, considering their sur-
prising resistance to unfavorable substrates and reaction products, and for devel-
oping probes that will allow us to monitor the development and disappearance of
enzymes related to these processes.
Finally, it appears appropriate to consider briefly the general significance
of hydrolyzable tannins in nature. It is well-documented that ellagitannins possess
anticarcinogenic properties; 79 as a recent example, the circular dimeric elIagitan-
nin, oenothein B, was found to suppress the transciption of mouse mammary
tumor virus. 80 In contrast, the impact of gallotannins (which represent the major
part of the biochemical pathways described in this chapter) as chemicals relevant
to mammalian health was recognized comparatively late. PentagalIoylglucose,
e.g. was recently found to promote tumor inhibition in mouse skin,sl and to in-
hibit Na+/K+-ATPase from horse kidney.82 Sulfated pentagalloylglucose-
derivatives were recognized as active anti-HIV agents in miceY Higher substi-
tuted pentagallyolglucose-derivatives (tannic acids) were found to possess
vasodilatory effects 84 or to inhibit the growth of bacteria causing food-borne ill-
nesses. 85 That such antiviral effects must not necessarily protect plants has been
shown in a study which revealed that caterpillars of gypsy moth (Lymantria
dispar) browsing on red oak leaves have better chances of surviving on trees rich
in tannins, simply because these polyphenols inhibit the development of a fatal
species-specific virus, lymantria-dispar-polyhedrosis-virus (LdNPV).86
The recent trends in this field have been eloquently described by Okuda et
al. 87 "The remarkable change in the definition of tannins, due to recent advances
in the chemistry of natural polyphenolic compounds, has been accompanied
by the changes in our recognition of the significance of their existence and
utilization, from that for leathering material to that for health effects."
ACKNOWLEDGMENTS
I thank the many colleagues and recent or previous coworkers that con-
tributed to the results reported in this chapter. Generous financial support from
the Deutsche Forschungsgemeinschaft (Bonn), the Fonds der Chemischen Indus-
trie (Frankfurt/M.) and from research grants of the University of Ulm is
gratefully acknowledged.
REFERENCES
1. FREUDENBERG, K. 1920. Die Chemie der natiirlichen Gerbstoffe. Springer, Berlin,

161 p.
2. NISHIZAWA, M., YAMAGISHI, T, NONAKA, G., NISHIOKA, I. 1980. Structure of
gallotannins in Paeoniae radix. Chern. Bull. 28: 2850-2852.
3. NISHIZAWA, M., YAMAGISHI, T, NONAKA, G., NISHIOKA, I. 1982. Tannins and
related compounds. Part 5. Isolation and characterization of polygalloylglucoses from
Chinese gallotannin. 1. Chern. Soc. Perkin Trans. I: 2963-2968.
4. NISHIZAWA, M., YAMAGISHI, T, NONAKA, G., NISHIOKA, I. 1983. Tannins and
related compounds. Part 9. Characterization of polygalloylglucoses from Turkish galls.
1. Chern. Soc. Perkin Trans. I: 961-965.
5. NISHIZAWA, M., YAMAGISHI, T, NONAKA, G., NISHIOKA, I., NAGASAWA, T,
OURA, H. 1983. Tannins and related compounds. XII. Isolation and characterization of
galloylglucoses from Paeoniae radix and their effect on urea-nitrogen concentration in
rat serum. Chern. Pharm. Bull. 31: 2593-2600.
6. LOSCHER, R., HEIDE, L. 1994. Biosynthesis of p-hydroxybenzoate from p-coumarate
and p-coumaroyl-coenzyme A in cell-free extracts of Lithospermum erithrorhizon cell
cultures. Plant Physiol. 106: 271-279.
7. ZENK, M.H. 1964. Zur Frage der Biosynthese der Gallussaure. Z. Naturforsch. 19b:
485-492.
8. NEISH, A.C., TOWERS, G.H.N., CHEN, D., EL-BASYOUNI, S.Z., IBRAHIM, R.K.
1964. The biosynthesis of hydroxybenzoic acids in higher plants. Phytochemistry 3:
485-492.
9. CONN, E.E., SWAIN, T 1961. Biosynthesis of gallic acid in higher plants. Chern. Ind.
592-593.
10. CORNTHWAITE, D.C., HASLAM, E. 1965. Gallotannins. IX. The biosynthesis of gallic
acid in Rhus typhina. J. Chern. Soc. 3008-3011.
11. TATEOKA, TN. 1968. Formation of protocatechuic acid from 5-dehydroshikimic acid in
the extract of mung bean seedling. Bot. Mag. Tokyo 81: \03-104.
12. KATO, N., SHIROYA, M., YOSHIDA, S., HASEGAWA, M. 1968. Biosynthesis of gallic
acid by a homogenate of the leaves of Pe/argonium inquinans. Bot. Mag. Tokyo 81:
506-507.
210 G.G.GROSS
13. HEIDE, L., FLOSS, H.G., TABATA, M. 1989. Incorporation of shikimic acid into
p-hydroxybenzoic acid in Lithospermum erythrorhizon cell cultures. Phytochemistry
28: 2643-2645.
14. SAIJO, R. 1983. Pathway of gallic acid biosynthesis and its esterification with catechins
in young tea shoots. Agric. BioI. Chern. 47: 455-460.
15. ISHIKURA, N. 1975. Incorporation rate of shikimic acid- 14C and phenylalanineYC into
gallic acid in Rhus and Acer leaves. Experientia 31: 1407-1408.
16. ISHIKURA, N., HAYASHIDA, S., TAZAKI, K. 1984. Biosynthesis of gallic and ellagic
acids with 14C-Iabeled compounds in Rhus and Acer leaves. Bot. Mag. Tokyo 97: 355-367.
17. AMRHEIN, N.A., TOPp, H., lOOP, O. 1984. The pathway of gallic acid biosynthesis in
higher plants. Plant. Physiol. 75s: 18.
18. LYDON, 1., DUKE, S.O. 1988. Glyphosate induction of elevated levels of hydroxyben-
zoic acids in higher plants. Agric. Food. Chern. 36: 813-818.
19. WERNER, I., BACHER, A., EISENREICH, W 1997. Retrobiosynthetic NMR studies
with l3C-Iabeled glucose. Formation of gallic acid in plants and fungi. 1. BioI. Chern.
272: 25474-25482.
20. HASLAM, E. 1989. Plant Polyphenols. Vegetable Tannins Revisited. Cambridge Univer-
sity Press, Cambridge, 230 p.
21. HADDOCK, E.A., GUPTA, R.K., AL-SHAFI, M.K., LAYDEN, K., HASLAM, E.,
MAGNOLATO, D. 1982. The metabolism of gallic acid and hexahydroxydiphenic
acid in plants: Biogenetic and molecular taxonomic considerations. Phytochemistry 21:
1049-1062.
22. STOCKIGT, 1., ZENK, M.H. 1974. Enzymatic synthesis of chlorogenic acid from
caffeoyl coenzyme A and quinic acid. FEBS Lett. 42: 131-134.
23. ULBRICH, B., ZENK, M.H. 1980. Partial purification and properties of p-
hydroxycinnamoyl-CoA: shikimate p-hydroxycinnamoyl transferase from higher plants.
24. GROSS, G.G. 1982. Synthesis of galloyl-coenzyme A thioester. Z. Naturforsch. 37c:
778-783.
25. GROSS, G. 1983. Synthesis of mono-, di- and trigalloyl-~-D-glucose by ~
glucogallin-dependent galloyitransferase from oak leaves. Z. Naturforsch. 38c: 519-523.
26. GROSS, G.G. 1982. Synthesis of ~-glucogallin from UDP-glucose and gallic acid by an
enzyme preparation from oak leaves. FEBS Lett. 148: 67-70.
27. GROSS, G.G. 1983. Partial purification and properties of UDP-glucose: vanillate 1-0-
glucosyl transferase from oak leaves. Phytochemistry 22: 2179-2182.
28. WEISEMANN, S., DENZEL, K., SCHILLING, G., GROSS, G.G. 1988. Enzymatic syn-
thesis of I-O-phenylcarboxyl-~-D-glucose esters. Bioorg. Chern. 16: 29-37.
29. ATKINSON, M.L., MORTON, R.K. 1960. Free energy and the biosynthesis of
phosphates. pp. 1-95 in Comparative Biochemistry. Vol. II. Free Energy and Biological
Function (M. Florkin and H.S. Mason, eds.), Academic Press, New York-London.
30. MOCK, H.P., STRACK, D. 1993. Energetics of the uridine 5'-diphosphoglucose: hydrox-
ycinnamic acid acyl-glucosyltransferase reaction. Phytochemistry 32: 575-579.
31. SCHMIDT, S.W, DENZEL, K., SCHILLING, G., GROSS, G.G. 1987. Enzymatic syn-
thesis of 1,6-digalloylglucose from ~-glucogallin by ~-glucogallin 6-0-galloyltransferase
from oak leaves. Z. Naturforsch. 42c:87-92.
32. GROSS, G.G., DENZEL, K., SCHILLING, G. 1990. Enzymatic synthesis of di-O-
phenylcarboxyl-~-D-glucose esters by an acyltransferase from oak leaves. Z. Naturforsch.
45c: 37-41.
33. DAHL BENDER, B., STRACK, D. 1984. Enzymatic synthesis of 1,2-disinapoylglucose
from I-sinapoylglucose by a protein preparation from cotylesons of Raphanus sativus
grown in the dark. 1. Plant Physiol. 116: 375-379.
34. DAHLBENDER, B., STRACK, D. 1986. Purification and properties of 1-

(hydroxycinnamoyl)-glucose: hydroxycinnamoyltransferase from radish seedlings. Phy-
tochemistry 25: 1043-1046.
35. KOJIMA, M., KONDO, T. 1985. An enzyme in sweet potato root which catalyzes the
conversion of chlorogenic acid, 3-caffeoylquinic acid, to isochlorogenic acid, 3,5-di-
caffeoylquinic acid. Agric. BioI. Chern. 49: 2467-2469.
36. VILLEGAS, R.J.A., SHIMOKAWA, T., OKUYAMA, H. KOJIMA, M. 1987. Purification
and characterization of chlorogenic acid: chlorogenate caffeoyl transferase in sweet potato
roots. Phytochemistry 26: 1577-1581.
37. GHANGAS, G.S., STEFFENS, 1.e. 1995. I-O-Acyl-P-D-glucoses as fatty acid donors
in transacylation reactions. Arch. Biochem. Biophys. 316: 370-377.
38. GROSS, G.G. 1998. Biosynthesis of hydrolyzable tannins. in Comprehensive Natural
Products Chemistry. Vol. 3. Carbohydrates and Their Derivatives Including Tannins,
Celiulose, and the Related Lignins. (B.M. Pinto, ed.), Elsevier, Oxford. In press.
39. DENZEL, K., SCHILLING, G., GROSS, G.G. 1988. Biosynthesis of gallotannins. Enzy-
matic conversion of 1,6-digalioylglucose to 1,2,6-trigalloylglucose. Planta 176: 135-137.
40. GROSS, G.G., DENZEL, K. 1991. Biosynthesis of gallotannins. p-Glucogaliin-
dependent galioylation of 1,6-digalloylglucose to I ,2,6-trigalioylglucose. Z. Naturforsch.
46c: 389-394.
41. HAGENAH, S., GROSS, G.G. 1993. Biosynthesis of 1,2,3,6-tetra-O-galloyl-p-D-
glucose. Phytochemistry 32: 637-641.
42. CAMMANN, 1., DENZEL, K., SCHILLING, G., GROSS, G.G. 1989. Biosynthesis of
gallotannins. p-Glucogaliin-dependent formation of 1,2,3,4,6-pentagalioylglucosc by
enzymatic galioylation of 1,2,3,6-tetragalioylglucose. Arch. Biochem. Biophys. 273:
58-63.
43. GROSS, G.G., SCHMIDT, S.W, DENZEL, K. 1986. p-Glucogallin-dependent acyl-
transferase from oak leaves. I. Partial purification and characterization. 1. Plant Physiol.
126: 173-179.
44. DENZEL, K., WEISEMANN, S., GROSS, G.G. 1988. p-Glucogaliin-dependent acyl-
transferase from oak leaves. II. Application for the synthesis of I-O-phcnylcarboxyl-P-
D-C 4 C]glucose esters. 1. Plant Physiol. 133: 113-115.
45. DENZEL, K., GROSS, G.G. 1991. Biosynthesis of galiotannins. Enzymatic "dispropor-
tionation" of 1,6-digalloylglucose to 1,2,6-trigalioylglucose and 6-galloylglucose by an
acyltransferase from leaves of Rhus typhina L. Planta 184: 285-289.
46. WILLIAMS, 1.M., RICHARDSON, A.e. 1967. Selective acylation of pyranosides. I.
Benzoylation of methyl a-D-glycopyranosides of mannose, glucose and galactose.
Tetrahedron 23: 1369-1378.
47. REINEFELD, E., AHRENS, D. 1971. Der Einfluss der Konfiguration auf die partielie
Veresterung von D-Glucopyranosiden. Liebigs Ann. Chern. 747: 39-44.
48. KAWAMOTO, H., NAKATSUBO, F., MURAKAMI, K. 1996. Stoichiometric studies of
tannin-protein co-precipitation. Phytochemistry 41: 1427-1431.
49. KILKOWSKI, WJ., GROSS, G.G. 1999. Color reaction of hydrolyzable tannins with
bradford reagent, coomassie briliiant blue. Phytochemistry, in press.
50. GHANGAS, G.S., STEFFENS, J.e. 1993. UDPglucose: fatty acid transglucosylation and
transacylation in triacylglucose biosynthesis. Proc. Natl. Acad. Sci. USA 90: 9911-9915.
51. KUAI, J.-P., GHANGAS, G.S., STEFFENS, J.e. 1997. Regulation oftriacylglucosc fatty
acid composition. Uridine diphosphate glucose: fatty acid glucosyltransferase with over-
lapping chain-length specificity. Plant Physiol. 115: 158 J -1587.
52. HOFMANN, A.S., GROSS, G.G. 1990. Biosynthesis of gallotannins: Formation of
polygalioylglucoses by enzymatic acylation of 1,2,3,4,6-penta-O-galioylglucose. Arch.
Biochem. Biophys. 283: 530-532.
212 G. G. GROSS
53. NIEMETZ, R., GROSS, G.G. 1998. Gallotannin biosynthesis: Purification of ~

glucogallin: 1,2,3,4,6-pentagalloyl-~-D-glucose galloyitransferase from sumac leaves.
54. OKUDA, T. YOSHIDA, T., HATANO, T. 1989. Ellagitannins as active constituents of
medicinal plants. Planta Med. 55: 117-122.
55. FELDMAN, K.S., ENSEL, S.M. 1994. Ellagitannin chemistry. Preparative and mecha-
nistic studies of the biomimetic oxidative coupling of galloyl esters. 1. Amer. Chern. Soc.
116: 3357-3366.
56. YANG, C.S., LEE, M.-1., CHEN, L. YANG, G.- Y. 1997. Polyphenols as inhibitors of car-
cinogenesis. Environ. Health Perspect. 105, Supplement 4: 971-976.
57. FELDMAN, K., SAHASRABUDHE, K., QUIDEAU, S., LAWLOR, M.D. 1999. Syn-
thesis studies on complex ellagitannins. in Plant Polyphenols. Chemistry and Biology
(G.G. Gross, R.W Hemingway and T. Yoshida, eds.), Plenum, New York, in press.
58. HATANO, T., KIRA, R., YOSHIZAKI, M., OKUDA, T. 1986. Seasonal changes in the
tannins of Liquidarnbar forrnosana reflecting their biosynthesis. Phytochemistry 25:
2787-2789.
59. GROSS, G.G. 1992. Enzymes in the biosynthesis of hydrolyzable tannins. pp. 43-60 in
Plant Polyphenols. Synthesis, Properties, Significance (R.W Hemingway and P.E. Laks,
eds.), Plenum, New York.
60. GROSS, G.G. 1992. Enzymatic synthesis of gallotannins and related compounds.
pp. 297-324 in Phenolic Metabolism in Plants (H.A. Stafford and R.K. Ibrahim, eds.),
Plenum, New York.
61. RAUSCH, H., GROSS, G.G. 1996. Preparation of ['4C]-labelled 1,2,3,4,6-penta-0-
galloyl-~-D-glucose and related gallotannins. Z. Naturforsch. 51c: 473--476.
62. GROSS, G.G. 1994. In vitro studies on the biosynthesis of gallotannins and ellagitannins.
Acta Horticult. 381: 74-80.
63. NIEHAUS, 1.D., GROSS, G.G. 1997. A gallotannin degrading esterase from leaves of
pedunculate oak. Phytochemistry 45: 1555-1560.
64. SCALBERT, A. 1991. Antimicrobial properties of tannins. Phytochemistry 30:
3875-3883.
65. FURSTENBURG, D., VAN HOFEN, W 1994. Condensed tannin as anti-defoliate agent
against browsing by giraffe (Giraffa carnelopardalis) in the Kruger National Park. Compo
Biochem. Physiol. 109A: 425-431.
66. ROBBINS, e.T., HANLEY, T.A., HAGERMAN, A.E., HJELJORD, 0., BAKER, D.L.,
SCHWARTZ, e.C., MAUTZ, WW 1987. Role of tannins in defending plants against
ruminants: reduction in protein availability. Ecology 68: 98-107.
67. KLOCKE, l.A., VAN WAGENEN, B., BALANDRIN, M.E 1986. The ellagitannin
geraniin and its hydrolysis products isolated as insect growth inhibitors from semi-arid
land plants. Phytochemistry 25: 85-91.
68. MAIlLE, P. 1984. Das toxische Kompartiment der Pflanzenzelle. Naturwissenschaften
71: 18-24.
69. KASPAR, D., GRUNDHOFER, P., GROSS, G.G. 1998. Preparation of antibodies against
hydrolyzable plant tannins. Phytochem. Anal. 9: 107-111.
70. IBRAHIM, R. 1990. Immunocytochemical localization of plant secondary metabolites
and the enzymes involved in their biosynthesis. Phytochem. Anal. I: 49-59.
71. MUELLER, We., GREENWOOD, A.D. 1978. The ultrastructure of phenolic-storing
cells fixed with caffeine. 1. Exp. Bot. 29: 757-764.
72. CHAFE, S.C., DURZAN, D.1. 1973. Tannin inclusions in cell suspension cultures of white
spruce. Planta 113: 251-262.
73. DIERS, L., SCHOTZ, E, MEYER, B. 1973. Dber die Ausbildung von Gerbstoffvakuolen
bei Oenotheren. Cytobiologie 7: 10-19.
74. ZOBEL, A.M. 1986. Localization of phenolic compounds in tannin-secreting cells from
Sambucus racemosa L. shoots. Ann. Bot. 57: 801-810.
75. CHAREST, PM., BRISSON, L., IBRAHIM, R.K. 1986. Ultrastructural features of
flavonoid accumulation in leaf cells of Chrysosplenium americanum. Protoplasma 134:
95-1O\.
76. MARCHAND, L., CHAREST, PM., IBRAHIM, R.K. 1987. Localization of partially
methylated flavonol glucosides in Chrysosplenium americanum: Immunogold labeling.
J. Plant Physiol. 131: 339-348.
77. HELM, R.E, RANATUNGA, TD., CHANDRA, M. 1997. Lignin-hydrolyzable tannin
interactions in wood. J. Agric. Food Chern. 45: 3100-3106.
78. BOHM, B.A., SINGH, S., KOUPAI-ABYAZANI, M.R. 1994. Biosynthesis and turnover
of tannins in sainfoin (Onobrychis viciifolia). Polyphenols Actualites II: 18-19.
79. OKUDA, T, YOSHIDA, T, HATANO, T 1989. Ellagitannins as active constituents of
medicinal plants. Planta Med. 55: 117-122.
80. AOKI, K., MARUTA, H., UCHIUMI, E, HATANO, T, YOSHIDA, T, TANUMA, S.
1995. A macrocircular ellagitannin, oenothein B, suppresses mouse mammary tumor gene
expression via inhibition of poly(ADP-ribose) glycohydrolase. Biochem. Biophys. Res.
Commun. 210: 329-337.
81. YOSHIZAWA, S., HORIUCHI, T, SUGANUMA, M., NISHIWAKI, S., YATSUNAMI,
J., OKABE, S., OKUDA, T, MUTO, Y, FRENKEL, K., TROLL, w., FUJIKI, H. 1992.
Penta-O-galloyl-~-D-glucose and (-)epigallocatechin gallate, pp. 316-325. in Phenolic
Compounds in Food and Their Effects on Health II. Antioxidants and Cancer Prevention
(M. Huang, C. Ho, and C. Lee, eds.), ACS Symposium Ser. Vol. 507, Washington,
DC, 1992.
82. SATOH, K., NAGAI, E, USHIYAMA, K., YASUDA, I., SETO, T, KANO, I. 1997. Inhi-
bition of Na+, K+-ATP-ase by I ,2,3,4,6-penta-O-galloyl-~-D-glucose, a major constituent
of both moutan cortex and Paeoniae radix. Biochem. Pharmacol. 53: 611-614.
83. NAKASHIMA, H., ICHIYAMA, K., HIRAYAMA, E, UCHINO, K., ITO, M., SAITOH,
T., UEKI, M., YAMAMOTO, N., OGAWARA, H. 1996. Sulfated pentagalloyl glucose
(Y-ART-3) inhibits HIV replication and cytophatic effects in vitro, and reduces HIV
infection in hu-PBL-SCID mice. Antivir. Res. 30: 95-108.
84. GOTO, H., SHIMADA, Y, AKECHI, Y, KOHTA, K., HATTORI, M., TERASAWA, K.
1996. Endothelium-dependent vasodilator effect of extract prepared from the roots of
Paeonia lactifiora on isolated rat aorta. Planta Med. 62: 436-439.
85. CHUNG, K.-T, STEVENS, S.E. jr, LIN, W.-E, WEI, c.1. 1993. Growth inhibition of
selected food-borne bacteria by tannic acid, propyl gallate and related compounds. Lett.
Appl. Microbiol. 17: 29-32.
86. HUNTER, M.D., SCHULTZ, J.c. 1993. Induced plant defenses breached? Phytochemi-
cal induction protects an herbivore from disease. Oecologia 94: 195-203.
87. OKUDA, T., YOSHIDA, T, HATANO, T 1992. Polyphenols from asian plants. Structural
diversity and antitumor and antiviral activities, pp. 160-183. in Phenolic Compounds in
Food and Their Effects on Health II. Antioxidants and Cancer Prevention, pp. 316-325.
(M. Huang, C. Ho, and C. Lee, eds.), ACS Symposium Ser. Vol. 507, Washington, DC,
1992.
Chapter Nine
THE FORMATION OF HEARTWOOD AND ITS

EXTRACTIVES
An Overview
WE. Hillis
CSIRO Forestry and Forest Products

Private Bag lO, Clayton South MDC
Victoria 3169, Australia
Introduction ................................................... 215

Cambial Region ................................................ 216
Sapwood ...................................................... 220
Primary Metabolites and Their Conversion .......................... 222
Transition Zone ................................................ 223
Heartwood Formation ........................................... 228
Extractives .......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
The Role of Ethylene ........................................... 236
Containment of Discolored Wood or False Heartwood ................. 239
Formation of Specific Components of Extractives ..................... 243
Summary ..................................................... 244
INTRODUCTION
During evolution, trees developed several defense systems to maintain long

term existence. One of these was the development of durable heartwood, now
known to be due to the extractives present. The term "extractives" used in this
chapter covers a large number of organic compounds of different classes which
can be extracted from wood or bark with polar and non-polar solvents and which
are the end-products of metabolism. As well as decay resistance, the extractives
may confer color, insect resistance, lower fiber saturation point, and increased
crush resistance of wood.!
K1uwer Academic I Plenum Publishers, New York, 1999.
215
216 W. E. HILLIS
On the other hand, the presence of extractives can also affect pulping and
lower the yield, color, and quality of pulp for papermaking, as well as interfering
with the recovery of pulping chemicals. 2 Some extractives in heartwood can
degrade wood coatings, lower adhesion qualities in bonded products (as with
some eucalypts), and cause corrosion. Others have found important use as
perfumes, dyestuffs, and medicines. In the future, the increasing volume of wood
from short-rotation, fast-grown plantations will contain lesser amounts of
heartwood and its extractives.
Recently it has been found 3.4 that some cells and splits in heartwood contain
pure or high concentrations of complex compounds. These can be biosynthesised
in heartwood from carbohydrates (or lipids to a lesser extent) without formation
of appreciable amounts of waste products. When the complex biological mecha-
nisms of these conversions are ascertained, it should be possible to produce the
complex compounds with the exact properties required in high yield and with
little pollution from an inexhaustible resource.
The phenomenon of heartwood initiation and formation, which occurs in
some trees, is complex and involves various processes in the living tree, some of
which may vary from species to species. The living ray and axial parenchyma
playa basic role in the internally controlled heartwood formation. The efficiency
and age of the sapwood parenchyma before activity increases can be influenced
by environmental conditions, growth rate, and various properties of the tree. A
holistic understanding of the physiological, anatomical, cellular, and biochemi-
cal conditions involved in heartwood formation is required before the phenome-
non can be defined. This has been attempted in three earlier studies. (Bamber
and Fukazawa,s Hillis,6 and Higuchi.7) The present contribution summarizes and
updates the first two, the primary objective being to understand the formation of
extractives resulting from heartwood formation. The latter devotes consideration
to the biochemistry and the enzymological changes in wood.
CAMBIAL REGION
The structural elements of wood are formed in the cambial region, located
between the previously formed bark and the wood of stems and branches. Cambial
cells have a limited life of 4-33 days, and consist of very thin walls containing
mostly water, numerous Jiving organelles, nutrients, sugars, etc. inside a plas-
malemma with embedded enzymes. Most cells divide and expand in about 30
days to become wood, but injuries in the growing season result in exudates with
compositions markedly different from the sapwood or heartwood extractives in
the same tree. 6 Processes controlling growth vary as cells go through stages
of division and enlargement and maturation, until death occurs when maturation
is complete.
THE FORMATION OF HEARTWOOD AND ITS EXTRACTIVES 217
An amorphous intercellular layer forms between the primary wall of all

adjacent cells, and, after enlargement, cellulose molecules are laid down side by
side as elementary fibrils on the primary wall of fiber cells in particular. These
fibrils exist in a crystalline form. Other cells are produced at the same time as
the fibers. The elementary fibrils oriented in the same direction as the fibers form
microfibrils and cross-bridges by hydrogen bonding between them and provide
great strength and some rigidity. These linear layers of micro fibrils are surrounded
by amorphous hemicelluloses (highly branched polysaccharides) which fill the
spaces between the microfibrils and are intimately involved with cellulose. These
cells need to acquire rigidity for resistance to gravity, which is provided in land
plants by the aromatic-based lignins that surround and are bound to parts of the
hemicelluloses. In turn, other parts of the hemicelluloses are bound to the cellu-
lose regions. Polymeric lignin and condensed tannins frequently occur together,
and a reduction step is involved in their biosynthesis.
Lignin accumulates first at the cell corners, then in the lamellae between
the cells, and then surrounds the hemicelluloses around the cellulose microfibril.
The latter forms lamellae which constitute the primary and then the secondary
portion of the cell wall. When lignification is complete, all the fiber cells die with
a few exceptions. 8 The undried cell wall is porous and in Pinus resinosa sapwood
contains about 25% capillaries (about 160-600 nm diameter) which contain water
throughout the cell wall and collapse on drying. 9
In many timbers, the xylem contains concentric rings-the growth rings.
They are often distinct because of variation in the size of the cells between the
early part of growth and the last formed growth. Water availability and tempera-
ture frequently limit cell production and growth ring width. The differences in
volume and density of wood produced early in the growing season (earlywood)
as opposed to late (latewood) result in the structure of growth rings. The dis-
tinctness of growth rings varies between species and climatic conditions. They
help to verify and validate forest productivity and growth estimates.
Fibers (in hardwoods or angiosperms) and tracheids (in softwoods or gym-
nosperms) are the only thick-walled cells. The function of fibers is principally for
support and structural purposes, whereas tracheids function as support and as
water conducting systems surrounding living parenchyma cells. Earlywood fibers
or tracheids have thinner cell walls, larger lumens, and shorter lengths than those
in latewood. The cell wall of tracheids is interrupted by various types and sizes
of pits which, in the original condition, allow fluids to move between cells but
aspirate on loss of water in surrounding tissues. Extractives-like materials are
formed at, or transferred through, pit pairs of ray parenchyma to tracheids. 1O The
number of pits varies between species of angiosperms, and types of fibers and
can be aspirated or blocked in various ways.
Two types of parenchyma are formed in the cambium-the horizontal ray
parenchyma and the axial parenchyma cells which remain live in the sapwood.
218 W. E. HILLIS
The secondary layer of these small, short, thin-walled cells may be absent. Ray
tissues in angiosperm trees have a more complicated structure in comparison with
coniferous trees. Ray and axial parenchyma are usually of similar size, but in
Pseudosindora palustris the size of apotracheal parenchyma cells (located axially
but away from vessels) is much larger than those of the ray parenchyma. II There
may be storage parenchyma (both horizontal and axial) containing reserve mate-
rials such as starch or fats or specialized cells contiguous with vessels. The
efficiency of parenchyma in heartwood formation may differ. Polyphenols can be
present only in the ray parenchyma of some species, whereas in other species they
are present also in the axial parenchyma. The ray parenchyma cells play an impor-
tant role in heartwood formation, whereas axial parenchyma appears to have dif-
ferent fm:ctions. The specialized cells have numerous large pits between them
and the vessels. They contain living cytoplasm which seems to be involved espe-
cially in carbohydrate metabolism and short-distance transport. 12 Lignification of
the ray parenchyma in the genus Pinus occurs in the transition zone during
heartwood formation. 5.13
Morphological 14.15 and chemical studies6 have revealed differences in func-
tion and physiological condition between various ray parenchyma cells. Starch is
stored mainly in the axial parenchyma which can be arranged in characteristic
patterns. Administration of 14C-labelled phenylalanine to disks of Zelkova serrata
showed that radioactivity was more conspicuous in the ray parenchyma cells than
in the axial parenchyma cells, particularly in the transition zone where most of
the radioactivity was located. 16 This will be discussed later. The dimensions of
ray tissue vary in different plants from one cell wide (uniseriate) to several cells
wide (multiseriate) and from one to twenty cells high. The behavior of the nuclei
in these cells, within a segment of the ray, as well as across the sapwood (and
transition zone) can differ. The cells can occupy 8-40% of the volume of
angiosperms and about 16% of gymnosperms.
Vessels or pores are found only in the axial direction of hardwoods and with
a diameter larger than other cells. Their size differs between and within species,
e.g. in the latter, vessels in the earlywood have larger diameters than those in the
latewood. They are in long lengths of 1-3 meters, thin-walled, and interconnected
through various types of pits, and may be separate or arranged in groups (as
ring-or diffuse-porous) with characteristic patterns. They can be blocked with
tyloses or extractives.
The balloon-shaped tyloses extrude from the secondary wall of living
ray cells through pits into vessels if the pit size exceeds 10 /lm. If the pit size
is smaller, extractives tend to be extruded from the ray cells into vessels. 17
Tyloses are thin-walled and may be lignified. They are regularly produced in
connection with wounding and appear in the living wood bordering dead
wood, either near injuries involving loss of moisture or characteristically at the
heartwood periphery of several species. They develop before heartwood
extractives.
Resin canals in gymnosperms have a larger diameter than the other cells.
They develop in both vertical and horizontal directions to form an anastomozing
network. In some genera such as Pinus, the canals are lined with thin-walled
epithelial cells which function for several seasons and can produce resinous mate-
rials under high pressure (up to 10 atmospheres). The epithelial cells in resin
canals in the genus Pinus are lignified during the heartwood transition. 5
Due to the fragile nature of the cambial region and its small volume even
in its most active phase, it is difficult to assign the location of its extractives.
During the short periods of rapid metabolism, the cambial region in some euca-
lypts may be separated to allow its extractives to be examined. The main recog-
nizable compounds are (+) catechin, (-) epicatechin, possibly (-) epigallocatechin
gallate, shikimic acid, and 5 derivates of galloyl- and hexahydroxy-diphenic acid
linked to gluco-pyranose, synthesized by the shikimic acid and other pathways. 18
If the cambial region of eucalypts is injured during its most active period, a
fluid of red-brown kino is formed, sometimes in large amounts. Kinos contain
largely proanthocyanidins in various stages of polymerization with distinctive
monomeric polyphenols in some species, such as C-glucosides of flavonoids and
the lignan eudesmin (1) in Eucalyptus hemiphloia. I9 •2o Usually, kino is contained
in the wood in thin veins (1.5-5 mm) of up to 2 meters in length or in pockets
(when the parenchyma bridges are broken) which may contain large amounts.
Often, kino from eucalypts or resin from pines is exuded onto the surface of the
tree from damaged cambial regions and probably serves as defense mechanisms
for those regions.
1: Eudesmin 2: Matairesinol
CH07£H°0
3
HO
1
""
h
H
I""
h
OCH3
OH
3: Hydroxymatairesinol 4: a-Conidendrin
220 W. E. HILLIS
Increased amounts of ethylene are present during kino formation, and

administration of ethylene-releasing compounds to active eucalypt cambium leads
to kino formation. A relationship has been found between kino vein size and the
amount of ethylene applied. 21 Ethylene injection to the cambial region of some
Prunus or Acacia spp. results in the formation of carbohydrate gum and with
Pinus spp. of 0leoresin. 6
SAPWOOD
Sapwood is the outer annulus of tree stems, branches, and roots. In the
living tree, the fibrous elements die soon after their development has been com-
pleted. On the other hand, many of the ray and axial parenchyma cells remain
alive for a considerable time in the sapwood which has been defined as "the
portion of the wood that, in the living tree, contains living cells and reserve mate-
rials (e.g. starch)".22 The area of sapwood before it is transformed into heartwood
varies considerably among families, genera, and species. Heartwood may never
form, and living cells and starch have been found in the inner tissues of Acer
saccharum, Nothofagus cunninghamii, Alstonia scholaris, etc. (M.M. Chattaway,
pers. com.) over 100 years old. Ring-porous angiosperms usually have narrow
sapwoods and a small number of sapwood rings, whereas the width and ring
number in diffuse-porous woods can be greater and more variable.23 In some
species, all the ray parenchyma live throughout the sapwood. In other species,
they begin to die from the middle sapwood, and in another group, the rays begin
to die from the cambium. lo
The number of growth rings in the sapwood can be characteristic of some
species after the tree has grown out of the juvenile stage, and it remains more or
less constant throughout the life of the tree. The commencement of heartwood
formation varies considerably between genera and species: it is about 3-4 years
with Robinia species, about 5-7 years (1.5-2.5 cm width) with mature eucalypts,
3-12 years with Douglas-fir, about 14 years with P radiata, and 25-70 years with
P sylvestris. 23 There appears to be a strong genetic influence, with family differ-
ences, in the age of the tree when heartwood is formed.6.24.25 The number of heart-
wood rings appears to be essentially unrelated to tree growth in the post-juvenile
growth phase in P radiata. 26 The number of rings is often more consistent than
the width and cross-sectional area which can be influenced by location in the tree,
environmental conditions and site quality,26-29 growth rate/O,31 height in stem, age
of tree, and size of crown. 6 ,32-39 A highly significant relationship has been found
between sapwood area and foliar surface area of four eucalypts. 4o
The sapwood width of Cryptomeria japonica appears to be more dominant
than the number of growth rings in controlling heartwood formation. 41 P
canariensis, having a more narrow sapwood than predicted by age and growth,
tends to display a higher percentage of axial parenchyma in the inner xylem. 42
The intense resin soaking found in some heartwoods is explained by the high
proportion of living cells in the xylem and their ability to accumulate large
quantities of reserve starch. 42
Trees probably require a particular amount of sapwood, as shown, for
example, by the strong relationship between rate of growth and sapwood width. 43
Also, it has been concluded that heartwood is probably formed when the burden
of maintaining all the sapwood exceeds the capacity of the area or mass of foliage
and increasing respiration costs39 to supply the needs of sap stream conduction
and food storage. 5,44 In a study of heartwood formation, Gerrish 35 used the
"specific leaf burden" (SLB), defined as the ratio of the unit volume ofrespiring
tissue to the unit mass ofleaf. He considered that in the diffuse-porous, evergreen
Metrosideros polymorpha, heartwood formation is controlled by whole-tree ener-
getics rather than by the movement and behavior of the sapstream. Formation of
heartwood may be initiated when the SLB reaches a threshold value such that
all the sapwood in the tree can no longer be maintained. It is possible with
M. polymorpha stands that heartwood initiation is linked to canopy closure.
With the increasing use of plantations, fast-growing species, and shorter
times between harvests, the amount of sapwood in proportion to heartwood in
marketable timber and pulpwood is also increasing. This will affect the use of
wood in various applications. It has been proposed that the principle determin-
ing the amount of sapwood is the provision for the storage of food reserves such
as starch. 5 The patterns of the radial decrease of starch have a strong relationship
with the sapwood width, and trees with a narrow-sapwood and an abrupt decrease
in starch at the heartwood boundary have deep-colored heartwood with distinct
boundaries; other species show a gradual decrease of starch across the
sapwood. 10,23,45 Seasonal changes are observed only in the amount of free fatty
acids in the sapwood of Pinus sylvestris, with the levels being greater at the
beginning and end of the growing season. 46 Changes in vacuole content (where
polyphenols accumulate) also appear in two patterns; one shows no vacuoles in
the outer sapwood but the number gradually increases towards the sapwood/heart-
wood boundary. The other shows large amounts of vacuoles from the outer to the
inner sapwood. 47 Species with a large number of vacuoles in the outer sapwood
have narrow sapwoods. 49 The radial changes of vacuolar content vary according
to species but generally decrease in size with age and disintegrate in the transi-
tion zone and with the death of ray cells. 48 Relationships between width and age
of sapwood with other features have been summarized. 6
The parenchyma cells are the main living cells and they can respond to
injury. In some cases, the fibers in sapwood can retain their cytoplasm and starch. 5
Nobuchi et al. 48 grouped 20 coniferous species according to the survival curves
of sapwood ray cells. All cells in one type, such as that containing Pinus densiflora
and Thuja orientalis, survive from the cambium to the transition zone between
sapwood and heartwood. In another type, such as that containing Pseudotsuga
japonica and Cryptomeria japonica, parts of the ray parenchyma cells die in the
222 W. E. HILLIS
inner sapwood. In the third type, part of the ray parenchyma begins to die in the
outer sapwood, as with Picea abies and Larix leptolepis, and these species do not
show a sharp boundary between sapwood and heartwood. Several angiosperm
trees show an abrupt decrease of living ray parenchyma at the heartwood periph-
ery.48 There is only a weak direct relationship between radial changes in the
proportion of living ray cells and heartwood formation. 48
Sapwood contains low levels of extractives, as in the case of Quercus
Spp49-51 and Thuja plicata. In some species the composition of sapwood extrac-
tives can differ from those in heartwood. In gymnosperms, sapwood has a rela-
tively high free water content (in the lumen) of about 150% compared with 55%
in the heartwood. In Cryptomeria japonica, the radial distribution patterns of
methanol soluble extractives closely resemble those of moisture content. 41 On the
other hand, the sapwood of angiosperms usually contains about 83% free water
compared with an average of 81 % in the heartwood for a number of species. 6
However, the moisture content of the heartwood of some angiosperms may be
higher than that of the sapwood, as with some Eucalyptus, Quercus, Juglans, etc.
species. It is noteworthy there is no change in moisture content at the periphery
of teak heartwood. 52 On this basis, it is not always easy to support the claim that
loss of water is the first step in heartwood formation.
PRIMARY METABOLITES AND THEIR CONVERSION
One of the major functions of sapwood is the accumulation of storage mate-

rial in the form of carbohydrates (soluble sugars and particularly starch) or lipids
and nitrogenous materials. Carbohydrates are the main reserves in angiosperms,
and lipids in gymnosperms. A few genera can contain significant amounts of both
types of storage materials. Resin canals are also present in gymnosperms and are
under pressure in the sapwood, but their contents play little part in heartwood
conversion.
A variety of alterations in metabolism occur in the transition zone. In those
trees which contain both starch and lipids, the secondary substances formed
during heartwood formation originate primarily from starch and translocated car-
bohydrates. On the other hand, the lipid reserves, if present, playa less impor-
tant role. There is no evidence for the interconversion of fats and carbohydrates.
Angiosperms contain variable amounts of starch across the sapwood «8%),
and the amounts vary seasonally with the lowest being in the late summer. The
amount of starch in sapwood of many species is inadequate for the formation of
large amounts of extractives in the heartwood. Consequently significant amounts
of translocated carbohydrate would be required. There is little information,
however, about the dynamic translocation of sucrose or glucose from the cambial
region through the rays to the heartwood periphery, although, in one study, it
occurred within 19 days in early summer53 and 12 days during late autumn. 54
Saranpaa and H6ll 45 considered that the cell wall hemicelluloses of Pinus
sylvestris were partly hydrolyzed during heartwood formation and may be
involved. The polyphenols are formed first in the shikimate and phenylpropanoid
routes, with flavonoids from an additional route of the plant aromatic pathway. 55
Data indicate that the shikimate pathway in the plastids provides amino acids for
protein synthesis, and the cytoplasm provides, in addition, phenylalanine for the
phenylpropanoid and flavonoid segments. The chalcone synthase leading to
flavonoid synthesis is located on the endoplasmic reticulum membrane. 55
In pines, an abrupt change in the composition of extractives occurred in the
narrow transition zone where the triglyceride content decreased rapidly and fatty
acids increased. There was little change in the composition of the resin acids
across a cross-section, although the proportion of them increased more substan-
tially in the heartwood than in the sapwood. The extractives in the earlywood
tissues of Pieea abies had a somewhat higher content of fatty acids than the
corresponding latewood tissues and the degree of saturation of these fatty acids
appeared to be higher. 6
Triacylglycerols and fatty acids are the major reserve lipids of Robinia.
pseudoaeaeia, and they have an energy content that is much lower than that of
starch. 56 Only a small portion is used for heartwood formation. 57 The lipid classes
contain a variety of fatty acids with chain lengths of 14 to 24 carbon atoms and
up to 3 double bonds. At the sapwood/heartwood boundary, the amount of satu-
rated fatty acids begins to decline towards the center of the trunk with a shift
towards a higher degree of un saturation. Steryl esters of palmitic and stearic acids
are the most dominant, and these and other esters reach maximum value on the
heartwood side of the transition zone after which lower quantities are present. 57
The amount of free sterols increases markedly in the same location of the heart-
wood. The six sterols change in proportion across the stem with cholesterol reach-
ing its maximum at the outermost heartwood. In Teetona grandis, the lipid content
increases towards the transition zone with a simultaneous decrease in the amount
of starch. Datta and Kumar 8 suggested that lipids, together with sugars from the
hydrolyzed starch, give rise directly to extractives in the transition zone.
Changes in metabolism towards synthesis of extractives are accompanied
by activity of corresponding enzymes. It has been suggested that apart from
lipases, several other acylester hydrolases (e.g. phospholipases) catalyse lipolytic
reactions in the transition zone,56 and the enzyme turnover is most rapid at the
sapwood-heartwood boundary. Relatively high lipase activities have been found
at the heartwood periphery of R. pseudoaeaeia.
TRANSITION ZONE
In many species, particularly those with dark heartwoods, a distinct pale-

colored, narrow (usually about 2~5 growth rings (2 to 5mm» zone can be seen
224 W. E. HILLIS
between sapwood and heartwood of freshly cut cross-sections, and sometimes

around injured regions. It is also known as the "white zone", "dry zone", or "inter-
mediate wood" and contains living ray and axial parenchyma cells, which die at
the heartwood periphery. The zone is usually devoid of starch, often imperme-
able to liquids, with a moisture content abruptly lower than that of sapwood and
sometimes the heartwood. 6 The starch content falls abruptly in this zone before
the death of the ray parenchyma cells. 41 The decreased level of water in the tran-
sition zone has been associated with heartwood formation. 59 Whereas the mois-
ture content in the sapwood of conifers is considerably higher than that in the
heartwood, this is not always so in angiosperms. Moreover, the latter can show
seasonal differences. 5.6 The average moisture content of the transition zone in
several species can be classified into three types: the moisture content is between
that of the sapwood and the heartwood; it is similar to either; or it is lower than
either the sapwood or the heartwood. Moderate seasonal fluctuations can occur. 5
The water content of Cryptomeria japonica is low in the middle or inner parts of
the transition zone but increases again in the heartwood. 'o A marked and sharp
aspiration of most of the bordered pits between tracheids at the border of the
sapwood and transition zone, as welt as their encrustation, occurs in Pinus spp.
and C. japonica and precedes the decrease in moisture content so that passage of
fluid can occur only through the rays.48
Dry zones also precede the formation of wetwood (a form of pathological
heartwood) and surround it in healthy Abies grandis trees, whereas suppressed
trees form only dry zones. 60 The activity of gradually dying ray cells is appar-
ently responsible for the replacement of water by a gas in adjacent tracheids in
this and in other coniferous species. 61 The water is withdrawn from affected
tissues by hydrostatic tension,61.62 and the gas is prevented from spreading through
the system by aspirated bordered pits. This situation occurs with both heartwood
formation and in response to infection. A concentration of ethylene has been
found in this zone 63 .64 and it could be the gas responsible.
In Robinia pseudoacacia, at the transition zone only trace amounts of
non-structural carbohydrates were present. 65 In Picea mariana, the transition zone
contained only a few of the inorganic elements present in other morphological
tissues, and these were concentrated or localized in the tori and half-bordered pit
membranes of gymnosperms. 66
When discs of living sapwood of Cryptomeria japonica were allowed to
slowly dry (withering), the sugars present decreased. On the other hand, four
conioid (norlignan) heartwood constituents, agatharesinol (5) and sugiresinol (7)
derived from the shikimate biosynthetic pathway, and ferruginol (9) and dehy-
droferruginol derived from the mevalonate biosynthetic pathway, were synthe-
sized and accumulated in the inner sapwood in the latter half of the withering
process. 67 .68 It was considered that the withdrawal of water from sapwood was a
precondition for withering as well for heartwood formation. 69 Alternatively, the
J.
increased respiration and biological activity in this zone could expel free water
from the living cells preceding the formation of extractives.
OH
Ho-o--?;O-OH ,"'"
R
.ij
".
Ot-P H
5: Agathal'esinol R~H 7: Sugiresinol RI~R2~OH; R3~R4~R5~H 9: (+}-Ferruginol
6: Sequirin R9)H 8: Hydroxysugiresinol RI~R2~R49)H; R3~R5~H
The transition zone has a sharp boundary with both sapwood and heart-
wood or injured wood, and follows the same outline. The number of annual rings
and width of the transition zone have constant values at the different heights of
C. japonica and other species. 70 Different colored layers can be seen in the
transition zone adjacent to the heartwood of Taxus baccata and Pinus radiata
and some other species. The width of the transition zone can vary throughout the
year. 6 While it has been recorded in a number of angiosperms, it is less obvious
than in gymnosperms.
The percentage of living cells decreases gradually across the transition zone
to zero at the outer heartwood boundary of Cryptomeria japonica. IO ,41,48 Two
marked cytological changes have been observed in parenchyma cells in the tran-
sition zone of this species. The first is at the boundary between the sapwood and
the transition zone, where the cells show conspicuous vacuolarization (polyphe-
nols accumulate) and symptoms of senescence. The second change is seen at the
boundary between the transition zone and the heartwood and indicates involve-
ment with heartwood and extractives formation. 10 The disintegration of the nuclei
in the ray parenchyma cells and the disintegration of the vacuoles begins in the
transition zone after the decrease of the moisture content in the sapwood. This
disorganization of cell compartments can result in an altered71 or intensified72
metabolism to produce polyphenols, particularly in the inner transition zone and
outer heartwood when the protoplasmic membrane disintegrates to release extrac-
tives and cytosolic enzymes such as oxidases. Chattaway's concept72 has been sup-
ported by enzyme assay and oxygen uptake,73,74 nitrogen contents/5 and tracer
experiments. 16,65 The metabolic activity is enhanced at certain times of the vege-
tative period. 7o,76-78 The ray parenchyma cells are considered to be more impor-
tant than axial cells for heartwood formation in C. japonica. 41 Biological
factors as well as non-biological factors can change the nature of the secondary
metabolites after the death of all ray parenchyma cells.
It has been shown that the incorporation of radioactive compounds into
wood extractives in the transition zone in both gymnosperms and angiosperms is
quite high. 53 ,54,76.79,8o It was Harris 81 who first drew attention to the formation of
226 WE. HILLIS
polyphenols in the inner layers of the transition zone in Pinus radiata during the
flush of growth. They became incorporated in heartwood in early summer. In
Eucalyptus polyanthemos, the outer layers of the zone formed a methylellagic
acid glucoside, whereas the inner layers formed proanthocyanidin polymers as
well (WE. Hillis, unpublished). The degree of flavonoid formation in Prunus
yedoensis was much larger in the transition zone than in the inner sapwood. With
Cryptomeria japonica,82 one compound was present in greatest amounts in the
transition zone, whereas the other two major polyphenols were more abundant in
the heartwood. A marked change in the resin composition occurred in the transi-
tion zone of spruce and pine when the triglyceride content decreased and the fatty
acids increased. 6
Greatly increased respiratory activity has been observed at the
sapwood/heartwood boundary in Robinia pseudoacacia. 83 There are several indi-
cations of increased activity of enzymes between sapwood and heartwood. Lipase
is localized in the transition zone of Indian hardwoods,84,85 triacylglycerols are
hydrolyzed in the transition zone in Pinus sylvestris, and significant increases
occur in the amount and type of fatty acid composition at this time,86 Concen-
trations of phenol oxidizing enzymes have been observed at both the outer and
inner peripheries of the transition zone of some eucalypts, This increase could
be due to de novo synthesis and/or introduction of an activator. A number of
co-enzymes of key enzymes have been found to increase near the heartwood
boundary. Two phenol-oxidizing enzymes have been found in fresh P. radiata
heartwood,87 In several species, increased peroxidase and polyphenolic oxidase
activity have been found in the layers adjacent to the heartwood or the transition
zone where they were concentrated in the torus of the pits. Greatest activity occurs
during the winter and dormant seasons. 88 ,89 These oxidase enzymes would be
responsible for the oxidation and polymerization of major portions of polyphe-
nolic extractives. Other enzymes have been reported such as: succinate dehydro-
genase (in Acacia nilotica,84 indicating a zone of high respiration and possibly
associated with mitochondrial involvement in the formation of phenolics);58
adenosine triphosphatase (implicated in cellular transport, membrane permeabil-
ity and high-energy consuming processes in cellular metabolism and which dis-
appears in heartwood); acid phosphatase (concerned with degradation and
transport of carbohydrate); lipase; maltase; aldolase; and amylase, 58 The presence
of glucose-6-phosphatase indicates the operation of the hexose monophosphate
shunt. Its activity in Tectona grandis has been detected in the inner sapwood and
transition zone only when active conversion of metabolic substances is taking
place to form complex heartwood extractives, 58 Peroxidase activity in the inner
sapwood and transition zone peaks during the late growing season, to early
dormant season, and in a period before the maximum emanation of ethylene.
Peroxidase activity decreases in the transition zone of Shorea robusta,90 An abrupt
rise and decline in the activities of succinate dehydrogenase and acid phosphatase
occurs at the heartwood periphery of Azadirachta indica,91
Malic dehydrogenase (MD) and glucose-6-phosphate dehydrogenase

(G6PD) have enhanced activity in the transition zone surrounding the heartwood
of Pinus radiata, especially in the winter,74 and in the transition zone surround-
ing lesions caused by Sirex noctilio.92 G6PD activity was considerably greater in
the transition zone than in the adjacent inner sapwood, whereas less difference
was noticed with MD activity. Thus, it appears that pentose shunt activity
increases more dramatically in the transition zone than the citric acid cycle. Con-
sequently, it would be expected that these tissues would synthesize polyphenols
containing ftavonoids. During the dormant season, the activity of MD is
significantly higher in the transition zone adjacent to the heartwood than in the
middle and inner sapwood. There is also a substantial increase in the activity of
G6PD in this zone, particularly during the dormant season.
Shikimic acid, a key intermediate in the biosynthesis of polyphenols, has
been found in the axial and ray parenchyma of transition zones of different
species. It has increased sharply to over 2 percent in Quercus iberica, the heart-
wood of which contains ellagitannins. In Prunus yedoensis, quinic acid is present
in larger amounts. 6
Higuchi,59 as summarized by Bamber and Fukazawa,5 proposed that in the
inner sapwood, the relative contribution of the pentosephosphate pathway
increases, and this favors the formation of cinnamic acid as a precursor of heart-
wood extractives. The following processes concerned with metabolic changes
of ray parenchyma are likely: (I) The lignification of ray parenchyma cell
wall is initiated by the formation of phenylalanine ammonia lyase during the
aging process. In the development of lignification, the usual residual phenolic
compounds in the cells induce disorganization of mitochondria-containing
enzymes of the energy providing TCA cycle and related phosphorylation system.
Also, the pyruvic acid, which has been metabolized in the mitochondria, is
used as the building material for heartwood extractives such as terpenoids and
the A ring offtavonoids. (2) Concomitantly, sugar metabolism in ray parenchyma
cells alters gradually from glycolysis and the TCA cycle to the pentosephosphate
pathway favored for the synthesis of the B ring of ftavonoids and stilbenes.
Phenolic glucosides are changed to heartwood extractives by hydrolysis or
oxidative polymerization when the tonoplast is ruptured in the transition zone
and they come in contact with hydrolase, phenol oxidase, and peroxidase in the
cytoplasm. 5
The presence of more specific enzymes has been shown with phenylalanine
ammonia lyase and chalcone synthase in Robina pseudoacacia. There is a dra-
matic increase in the specific activity of these two enzymes in the innermost
tissues ofthe transition zone, indicating coloration of the parenchyma cells occurs
before cell death 65 .93 in the late autumn-winter period. 94 It is most noticeable in
the earlywood portion of the rings and coincides in this species with the forma-
tion of heartwood extractives. It coincides with enhanced ethylene production in
Pinus radiata. 64 Newly emerging polypeptides, preponderantly of a chalcone
228 W. E. HILLIS
synthase subunit, have been detected in the transition zone of R. pseudoacacia

shortly before cell death, pointing to a transient activation of metabolism. The
exclusive presence of these enzymes in the transition zone would control
flavonoid synthesis. Phenylalanine ammonia lyase was measurable only in
differentiating xylem and at the sapwood/heartwood transition zone. 93
HEARTWOOD FORMATION
The formation of heartwood involves a number of physiological, anatomi-

cal, cellular, and chemical changes. Heartwood has been defined as "the inner
layers of wood which, in the growing tree, have ceased to contain living cells and
in which the reserve materials (e.g. starch) have been removed or converted
into heartwood substances:>22 The color of these heartwood substances can be
intensified by chemical stains. Formation commences in the innermost portions
of sapwood, transition zone, or pith and expands outwards in a regular fashion as
the tree ages. Heartwood formation in roots is normally confined to the tapering
proximal portions of the rootS.44 The xylem with heartwood color does not always
coincide with the heartwood of the definition 22 of the International Association
of Wood Anatomists because living cells sometimes exist in the former,84 and the
definition requires modification.
Attempts have been made to classify types of heartwood and discolored wood
in different species, such as obligatory colored or regular heartwoods (e.g. Quercus
spp.), facultatively colored or irregular formation of heartwood (e.g. Fraxinus
excelsior), light colored heartwood or ripewood (e.g. Abies alba). 14 This will not
be possible until more biochemical and physiological characteristics of the meta-
bolic events involved can be defined. When formed, heartwood results from phys-
iological death due to internal factors, and it is the continuous, central portion of
stems, branches, and roots where it is often symmetrical. Its boundary can undu-
late vertically and horizontally (sometimes abruptly) in the tree. There are indica-
tions that heartwood formation begins in the earlywood. 94 The age of the tree when
heartwood is formed varies widely within and between genera. Many Eucalyptus
spp. produce heartwood at about five years growth. On the other hand, heartwood
forms in different Pinus spp. between 12-70 years. In P radiata, environmental
and genetic factors apparently are important for heartwood development but not
tree vigor in the post-juvenile growth phase. 26 Heartwood is formed abruptly, as
does the quantity of extractives, together with bordered pit aspiration in several
gymnosperms and the number of tyloses with development of cell wall in several
angiosperm species. The amount of extractives formed increases outwards as the
tree grows, in a regular pattern that is usually annual, and by accumulation during
in situ biosynthesis. However, with P. banksiana, for example, heartwood is pro-
duced at the average rate of 0.57 ring per year until the tree reaches about 70 years
age and thereafter at an average rate of one ring per year. 32
The annual rhythm of activities is similar in all species. Thus, annual ring
formation commences in the spring and ceases towards the end of summer. Heart-
wood formation is initiated in warm temperate zones during the decline of
cambial activity or dormant periods. Before the end of summer, the heartwood
region has expanded to about 70% of the annual ring in Robinia pseudoacacia,
with completion of the formation in winter. Similar results have been observed
with Larix leptolepis, Pinus spp., and Cryptomeria japonica. 95 •96
Translocated carbohydrates and reserve substances provide raw materials
and/or the energy for heartwood formation. Heartwood is typically formed after
changes in reserve substances, and, in the autumn and winter seasons, no con-
spicuous changes are observed. 94 The outer heartwood of Robinia pseudo acacia
appears to contain living parenchyma cells in late summer. 94 ,95 Living cells in the
outer heartwood gradually form heartwood extractives in the autumn and winter
seasons and then die. 95 .96
The extracellular stimulus for heartwood formation is not entirely a process
dependent on age or sapwood volume. Heartwood extractives formation may
happen at different times in neighboring parenchyma cells. Abrupt enlargements
of the heartwood periphery can appear over short radial distances and are part
of the major heartwood region. Also, there are concentric and symmetrical dark
rings in heartwood. The heartwood of a few tropical species Microberlinia
brazzavilliensis (trade name zebrano), Berlinia, Diospyros, and Dalbergia spp.
(see als0 97.98 ) contain regularly spaced, very dark colored, axially long stripes. In
zebrano, they are narrow (l--4mm) and alternate between equally narrow, paral-
lel, light colored stripes which occupy one growth ring. In Thuja plicata 99 and
other species,6 the heartwood contains concentric bands of light colored wood in
a target ring pattern. There is apparently some factor that eliminates the ability
of the sapwood available at the time to convert into heartwood without damag-
ing the cambium's ability to form new sapwood (lA.F. Gardner, pers. com.). Frost
could be a cause in temperate regions 99 but is not likely to be so for tropical and
semi-tropical species or in some stands subject to weather extremes. Perhaps, a
substance is translocated from the cambium to the inner sapwood where it accu-
mulates and reaches a threshold level at the transition zone to initiate heartwood
formation. 43 This compound could be of greater importance for initiating extrac-
tives formation than the maintenance of a particular sapwood volume. Com-
pounds such as water-soluble vitamins 75 or ethylene precursors could induce or
activate genes encoding new enzyme systems.
If an injury to a tree is of sufficient severity to destroy a part of the cambium
and interrupt the files of ray parenchyma with the cambial zone, then the sapwood
situated behind that tissue will not be transformed to heartwood. The affected
region is eventually overgrown by normal cells, which later are transformed to
heartwood, leaving an isolated area of "included sapwood." On the other hand,
injury to the tree can accelerate the formation of heartwood in the neighboring
region. IOO When Abies grandis trees were infested by the balsam woolly aphid
230 W. E. HILLIS
(Adelges piceae), the percentage of heartwood increased, both in terms of age and
area, with the amount formed varying with different intensities of attack. Neither
the nature nor the amount of phenolic extractives was affected. Aphids are known
to change the anatomy and function of the ray parenchyma cells. 101
Heartwood color is often much darker than the surrounding sapwood. Trees
with narrow sapwoods have darker colored heartwoods,47 apparently due to the
greater availability of primary metabolites at the boundary. In Robinia pseudoa-
cacia, the earlywood exhibits the dark color of the heartwood, whereas the late-
wood shows the light color of the' sapwood. 56 The heartwood of some angiosperms
can have irregular bands of different color intensity and width. 98 Since the
amounts of reserve metabolites depend on climatic and environmental factors,
the deposition of heartwood extractives varies from one year to another. 102 Heart-
wood is free from storage starch, most sugars except mannose, xylose, arabinose,
and fats. 45.103 ,104 The small amounts of galactose and arabinose in the heartwood
of Pinus sylvestris might have originated from the breakdown of hemicelluloses
during heartwood formation. 45 ,103
Occasionally, the individual cells of a file of ray parenchyma cells in heart-
wood have different appearances, indicating the presence of secondary metabo-
lites of different composition. Also, in cross-section, the volume of non-structural
material in the vessels can exceed that of the adjacent ray parenchyma cells, indi-
cating a continuous extrusion from the rays of extractives and precursors during
the period of intense metabolic activity.
The process of heartwood formation is accompanied by several physiolog-
ical and cytological changes. Following the examination of several species,14,7I,105
morphology and dimensions of the nuclei in ray parenchyma cells were used to
define the vitality or activity in the transformation of sapwood to heartwood.
Other workers have obtained similar results with other species. 6 Large and oval
cell nuclei (indicative of active tissue) become radially elongated, and the degree
of slenderness decreases with increasing distance from the cambium and then
shows a rapid increase of cell activity in the transition zone (in support of Chat-
taway's theory).72 There is a tendency to form bigger and longer cell nuclei and
to increase the ratio of nucleoli volume to nuclei volume in the transition zone,
with the volumes being largest in particular trees with dark heartwoods. In R.
pseudoacacia and P japonica in spring to mid-summer, the ray parenchyma cells
containing nuclei coincide with the boundary of the sapwood and transition zone,
where one-half of the rays are alive. After this time, ray cells containing nuclei
can sometimes be found in the outer heartwood during late summer. IO Heartwood
formation is considered to occur in a rather short period at the boundary between
the transition zone and heartwood. 10 Recently, Magel et al. 94.106 showed that both
the ray and axial parenchyma in the inner sapwood of R. pseudoacacia are rich
in cell organelles and appear to play an important role during physiological
changes in this region. Cell death in the transition zone starts with vacuolariza-
tion, followed by a decrease in the number of cell organelles, accompanied by
newly emerging osmiophilic droplets on the tracheid side of half-bordered

pit-pairs between ray parenchyma cells and tracheids. 47 ,107-109 These probably
appear later on the inner surface of cell walls and pits and represent in situ syn-
thesized heartwood extractives. The plastids in xylem parenchyma cells of Pinus
sylvestris show most changes in their ultrastructure and starch content prior to
heartwood formation. Although the role of plastids is not completely understood,
they may play a role in the synthesis of resin and polyphenols in heartwood
formation. 109
The concentration of several alkali and alkaline earth metals is often higher
in the sapwoods than in the heartwoods of angiosperms and gymosperms, and
this internal recycling must reduce the nutrient demand placed on the forest eco-
system. The greatest changes are at the sapwood-heartwood boundary where,
in some cases, an element is in peak concentration66,1 10-1 12 and is localized in the
tori and half-bordered pit membrane. The nature and extent of redistribution and
storage varies with the particular element involved, the species, and the site on
which the tree grows. 100.1 13-116
EXTRACTIVES
Different classes and amounts of extractives are characteristic of different

families and genera. The ability to synthesize a diverse array of secondary
metabolites, particularly in angiosperms, which may repel or attract other organ-
isms, has evolved as one facet of the strategy of tree species to survive in chang-
ing conditions. The formation of secondary products is controlled by the
availability of energy for different sequences, by regulation of the quantity and
activity of the enzymes concerned, and by their integration into the overall
programs of differentiation and development of the cell. Synthesis is based on
spatial control of the location of enzymes, precursors, intermediates, and prod-
ucts. Compartmentation makes possible the synthesis of secondary products in
an ordered manner from intermediates of primary metabolism. The nature and
amount of the components ofthe extractives that enabled survival ofa tree species
during evolution presumably were influenced by genetic characteristics of
the tree, environment, aggressive pathogens, or insects. The cambium of some
species can produce exudates on injury, and these could have been an early form
of defense.
Heartwood developed under a variety of growing conditions contains
extractives of similar composition in trees of the same species or genus. 6 ,117 On
the other hand, the composition of exudates resulting from injury can be sub-
stantially different from those normally found in the heartwood. 6 The composi-
tion can also differ according to the injury. The same fungus can result in different
extractives in different species. 6 Details of the formation of characteristic extrac-
tives under different circumstances will assist in the precise description of
232 WE. HILLIS
different parenchyma and other cells and the sequence of genes and enzymes
which produce these extractives.
As previously mentioned, the term "extractives" used here covers a large
number of organic compounds of different classes which can be extracted from
wood or bark with polar and non-polar solvents 116 and which are the end-
products of metabolism. Some carbohydrates and their derivatives can be removed
with water. Whereas terpenoids can be completely removed, polyphenolic and aro-
matic-based compounds (found in most species) may enter the cell wall and
combine with cell wall components. They may be polymerized or oxidized and
require a weak alkali extraction. In a few species, due to the nature of their initial
extractives, such as ebony, polymerization and other changes can result in materi-
als insoluble in solvents. The term "extraneous material" should be confined to
silica, calcium carbonate, and inorganic based compounds. Extractives are prod-
ucts resulting from enhanced metabolism at the transition zone or heartwood
periphery due to cytological changes in the ray and axial parenchyma. 16,53
Extractives convey a wide range of useful properties to wood, particularly
if or when they enter the cell wall. They have properties such as resistance to
fungi and bacteria as well as water repellency to improve durability, increased
insect resistance, increased crushing strength of wood, low fiber saturation point,
and chromogenic properties to improve appearance. On the other hand, their pres-
ence can affect pulping and lower the yield, color, and quality of pulp for paper-
making, as well as affect recovery processes. They can degrade wood coatings,
lower adhesion qualities in bonded products, cause corrosion, and result in
dermatitis in some people.
The amount of methanol-soluble compounds of the same type in heartwood
increases at an increasing rate from pith to outer heartwood in different species,
and the rate in the inner heartwood may be IOW. 6,118 It is frequently shown by color
intensity that the maximum values decrease from the lower parts to crown height,
probably due to the cell age of the ray parenchyma41 and the availability of
primary metabolites. Slower grown trees of the same species generally contain
larger amounts of extractives. 6 It has been proposed that the amount of polyphe-
nols in heartwood is related to the amount of carbohydrate reaching the periph-
ery.6 The effect of growth rate is exemplified by the arabinogalactan in the lumen
of heartwood tracheids of Larix occidentalis. In an old (300 years) slow grown
tree, the arabinogalactan content in some rings was 24.5%, whereas in a young
fast-grown tree it reached only 7.68%.119
Often, the amount of extractives at the heartwood periphery is in the vicin-
ity of 10%, although the amount may reach 30-35% in Eucalyptus marginata.
The amount of starch in the sapwood is less than this amount, and, consequently,
extra translocated carbohydrate is required for extractives synthesis. Detailed
studies have shown that, in some species, the amount and constituents of extrac-
tives reach their maximum 2-3 rings inside the heartwood periphery,41.82,12o
possibly due to aqueous solutions diffusing into the heartwood. 41 ,58 It may also be
due to the slow conversion of precursors to end-product.
In some cases, the lower amounts of extractives in inner heartwood is due
to progressive non-enzymatic insolubilization as the wood ages within the
trunk. 121 The extent of insolubilization is probably related to the ease with which
the compounds are oxidized in air when they copolymerize with different cell-
wall polymers. 122 Also, the inner heartwood of Quercus spp. can become darker
than the outer due to hydrolysis and polymerization of the ellagitannins present. 51
It has been claimed the total amount of extractives is the same throughout the
heartwood. 51 The heartwoods of a number of angiosperms become increasingly
acidic as the trees grow and become more mature, mainly due to acetic acid
released from the acetyl groups of hemicelluloses. Values of pH 2.6 or less have
been found in the interior of some eucalypts. Unstable proanthocyanidins
can form red compounds under these conditions over a period of time and then
oxidize to brick-red to brown insoluble polymers121 which have been observed
in red-brown colored heartwoods. Under acidic conditions, ellagitannins can
be slowly hydrolyzed, and polymerized as in Quercus petraea and Castanea
sativa. 49 ,5o After a sharp increase at the heartwood periphery, the amount of ellag-
itannins decreases but the amount of free ellagic and gallic acids increases as the
heartwood ages, 6
Extractives in the cell wall can enter from cell lumens during heartwood
formation. Another possible mechanism is that precursors of polymerized
polyphenols enter or diffuse into the compound middle lamellae, or intercellular
space between the cells, from the pits of the parenchyma cells. From that
location they infiltrate the walls of the fibers via the cell wall capillaries of
the S I and the less permeable, thicker S2 layers of the secondary waiL 123-125
About one-half of the total polyphenols in redwood and incense cedar heartwoods
is estimated to be in the cell wall, and these are not as easily extracted as the
remainder, 126
The amount of extractives in sapwood is small and relatively uniform from
cambium to outer heartwood. At the cambial region of Tsuga spp, (hemlock),
a mixture of phenolic depsides based on shikimic, quinic, ferulic, and caffeic
acids is formed along with glucosides of lignans specific for western or eastern
hemlock 127 Catechin, epicatechin, and proanthocyanidin dimers are generated in
the parenchyma of the sapwood. In the transition zone, the lignans matairesinol
(2), hydroxy-matairesinol (3), a-conidendrin (4), and traces of related phenols
are formed along with a small proportion of a lignin-like materiaL These are
deposited in the parenchyma and bordered pits during heartwood formation, 127
Conifers show a pronounced decrease in fatty acid esters and an increase in free
fatty acids of different composition when the sapwood changes to heartwood,6
While neutral lipids dominate in sapwood of Pinus spp" diterpene resin acids,
mono- and sesqui-terpene volatile oils, and fatty acids are the main extractives
234 W. E. HILLIS
in heartwood. 128 The sapwood extractives of Schinopsis spp. are mainly gallic acid
esters and CIS compounds, with the latter being mainly in the inner sapwood. At
the heartwood periphery, the gallic acid content decreases whereas polymers of
CIS compounds increase sharply.12s
When '4C-phenylalanine was administered to Zelkova serrata, it was found
converted into small amounts of polyphenols that were present in sapwood, mainly
keyakinol (10) and taxifolin 6-C-glucoside (11). In the parenchyma of the transi-
tion zone, the phenylalanine was converted into larger amounts ofkeyakinol (10),
aromadendrin 6-C-glucoside (12) and taxifolin 6-C-glucoside (11), and in the
heartwood into, kaempferol 6-C-glucoside (13), keyakinin (14), and quercetin 6-
C-glucoside (15). This suggests that C-glucosylation occurs at the stage offorma-
tion of the flavanonols in the transition zone, and later, at the heartwood boundary,
an enzyme catalyses their oxidative dehydrogenation to the flavonol derivatives. 16
The transformation of sapwood into heartwood in Douglas-fir alters the composi-
tion of phenolic extractives in three different ways: phenolic glycosides present in
sapwood are absent in heartwood; some polyphenols absent or in small amounts
in the sapwood are biosynthesized at the transition zone; some polyphenols are oxi-
dized to complex polymers which may become partially insolubilized. 121
OH
2
R 10: Keyakinol (Rl=OMe, R2=H)
11: Taxifolin 6-C-Glc (Rl=R2=OH)
12: Aromadendrin 6-C-Glc (R I=OH, R2=H)
OH
13: KaempferoI6-C-Glc (Rl=OH, R2=H)
2
R 14: Keyakinin (p.l=OMe, R2=H)
15:Quercetin 6-C-Glc (R I=R 2=OH)
Slight differences in amount and color intensity in a few annual rings

have been attributed to the availability, after cell wall formation, of carbohydrate
that can be translocated to the periphery when heartwood is formed. Occa-
sionally, some samples of a species with dark-colored heartwood have a
sharply defined outer band of 5 or more growth rings which are less brown than
the interior. In E. marginata, the total amount of weak alkali soluble extractives
was the same in both regions, but the outer band had a much higher proportion
of water-soluble extractives than the inner portions (WE. Hillis unpubl).
Oxidation, dimerization, and polymerization of extractives in heartwood in
different species by nonbiological factors, after the death of ray parenchyma,
undoubtedly take place, but this must occur after small-sized molecules
impregnate the cell wall.
In conifers, the resin canals are lined with epithelial cells which produce
an oleoresin that, in the sapwood of healthy trees, can achieve a pressure of several
atmospheres. The composition of the oleoresin found in pockets differs from the
resin in sapwood and heartwood.
Heartwood is a "conservative" taxonomic tissue, and its extractives are
characteristic of the family, genus, and even species to which the tree belongs.
Natural products can be used in chemotaxonomy to assist identification and
improve botanical classification. There are two categories. First, the phyletic char-
acters acquired in previous times from unknown causes have the greater taxo-
nomic value. Second, the physiological and biological characters acquired as
adaptions to ecological conditions have taxonomic value mostly limited to the
species leve!. For example, the heartwoods of the Cupressaceae are the sole source
of tropolones such as ~-thujaplicin (16), and the Pinaceae contain a number of
monocyclic terpenoids (17-20). Different classes of polyphenols are characteris-
tic of Prunus, Acacia, Nothofagus, and Eucalyptus genera. 6 However, a few
species have biochemical varieties in which the heartwood components are
significantly different from those normally found in the species. 129
q:H 0
2 2 2 2
16: (3-Thujaplicin 17: Terpinolene 18: (1-Terpinene 19:(3-Terpinene 20:y- Terpinene
One or more classes of aliphatic compounds, terpenoids, aromatic com-

pounds such as ftavonoids, lignans, stilbenes, tropolones, among others, are found
in the wood extractives from different tissues of different families, genera, and
species. 6 ,7 Two or more classes of extractives are usually formed at the heartwood
peripheries. The amount and type of specific metabolites varies extensively from
one species to another. Stained microscopic sections of eucalypts indicate the
ratio of these classes can vary in different ray parenchyma and this would be
expected, as shown by the different behavior of the inner and outer cells of a
segment of the ray parenchyma. 70 Individual heartwood polyphenols are thought
to be biosynthesized by a particular process and in a particular position. The
sequence of formation of these classes in normal circumstances may differ, and
this may be associated with the availability of energy at the time. Polyphenols
involving only the shikimic acid pathway (such as gallic acid) appear to be formed
before those involving the tricarboxylic acid cycle (W.E. Hillis unpub!.).
Most earlier studies have given attention to pinaceous species, which usually
lack axial parenchyma cells other than epithelial cells. More recently, attention has
been given to species containing significant proportions of axial parenchyma. The
236 W. E. HILLIS
two types of parenchyma cells contain different major components. The axial
parenchyma cells of Cryptomeria japonica form significantly more red-colored
materials than do ray parenchyma cells during heartwood formation. 130 The latter
contribute more to the dark color of heartwood color. The parenchyma materials
often are deposited around cross-field pittings and around bordered pit pairs on the
tracheid-lumen side, particularly in latewood. 129 The axial parenchyma cells are
also significantly larger than those ofthe ray parenchyma in Pseudosindora palus-
tris. The phenolic fractions of the ethyl acetate soluble extractives are larger than
the acidic and neutral fractions in the axial parenchyma. The same phenols and the
neutral fraction (mainly sterols) are more abundant in the axial than in the ray
parenchyma. 131
The outermost colored growth ring of the heartwood of Robinia pseudoa-
cacia contains the highest concentration of dihydrorobinetin (21) in winter,
whereas in late summer its concentration is highest in the next inward growth
ring. A hydroxycinnamic acid derivative shows a similar distribution. In contrast
to dihydrorobinetin, robinetin accumulates in significant amounts in the transi-
tion zone. 65 In another study,56 phospholipides in R. pseudoacacia decreased
significantly in the transition zone towards the heartwood boundary so that only
trace amounts were present in the heartwood. Steryl esters were highest at the
heartwood periphery, but free sterols reached their maximum value with increas-
ing depth within the trunk, while relative frequencies of individual free sterols
changed. Phospholipase enzymes in this species also showed a maximal activity
at the heartwood periphery. 56
The heartwood ofCryptomeriajaponica contains the conioids (norlignans)
agatharesinol (5), sequirin (6), sugiresinol (7), hydroxysugiresinol (8), and
ferruginol (9). The radial position of the maxima of each ofthese phenols changes
depending on the height in the tree. 41 In a later study, the abrupt formation of
heterogeneous droplets which appeared in the mainly living ray parenchyma of the
transition zone was considered to be related to the heartwood phenols and their
precursors. These substances migrated from the ray parenchyma to the dead tra-
cheids in the transition zone in the first phase of biosynthesis of heartwood phenols.
Then, in the second phase at the boundary between transition zone and heartwood,
the chemical composition of the oil-like droplets changed from lipid-like materi-
als to those with a phenolic nature. The latter are probably precursors ofthe heart-
wood phenols. In the third stage, changes to the mostly conioid components
increased at different rates inside the heartwood, possibly due to some ray
parenchyma still living in the heartwood or to other biological changes. 107
THE ROLE OF ETHYLENE
Ethylene has been associated with increased respiration and synthesis of

enzymes required for the synthesis of phenols. Consequently, it is noteworthy
that in some species (Prunus serotina,132 Eucalyptus tereticornis, Pinus radiata,
Juglans nigra,6 but not in Chamaecyparis obtusa I35 ), high amounts of ethylene
have been detected in the transition zone or adjacent tissues in the late dormant
and winter periods. The quantity of ethylene (and phenolic extractives) produced
per unit weight of wood by the living axial and horizontal parenchyma adjacent
to the heartwood periphery is influenced more by individually activated than
total parenchyma cells and, in some cases, by the level of oxygen. 133 When fresh
blocks of P radiata sapwood were stored in a container continuously ventilated
with humidified air containing ethylene, polyphenols were formed but in a
composition resembling that of injured wood. 64
The situation is shown more clearly by the transition zone surrounding a
zone of infection. The wound reaction appears to be most rapid during the
growing season, and more ethylene emanates from this zone than that adjacent
to the heartwood of Pinus radiata or from mechanical wounds. 63 After the ovipos-
itor of Sirex noctilio penetrates the bark of P radiata, it forms two tunnels, one
for its eggs and the other for the fungus Amylostereum oreolalum to feed the
hatched larvae. 63 The sapwood cells are ruptured in the process. Dominant trees
produce much greater amounts of ethylene in response to attack by Sirex noctilio
and lesser amounts by injury than do suppressed trees. 64 Three weeks later,
pinosylvins (22,23) are present in considerable quantities in the lesions. 63 .64
Further cytological and biochemical studies on extractives formation are needed
giving particular attention to each part of the transition zone.
OH
OH
H0Y-O~
::7 I -
HO :::,...
OH
OR
22: Pinosylvin (R~H)

21: Dibydrorobinetin 23: Pinosylvin monometbyl etber (R~CH3)
The injection of ethylene (via ethrel) into Rhus succedanea resulted in

heartwood polyphenols, increased enzyme activity and amounts of lipids and
extractives in Azadirachta indica and Samanea saman. 6 In Mesuaferrea, it led to
the formation of a type of heartwood with increased amounts and activity of pro-
teins and enzymes in the transition zone. Darker colored but living parenchyma
cells remained. 134 Ethylene also leads to the accumulation of dark staining organic
substances in the ray and axial parenchyma cells of C. obtusa. 135 Small amounts
have a curvilinear effect in the production of polyphenolic kinos from the
cambium of eucalypts (N.D. Nelson unpubl. data). Ethylene also induces the
formation of carbohydrate gum from the same region in Prunus spp. and rubber
in Hevea braziliensis. 6 Ethylene production correlates with peroxidase activity in
walnut and cherry, and both peak near the transition wood during early
dormancy. 132 This occurs also in Pinus species. 13
238 W. E. HILLIS
In addition to a marked increase in ethylene production induced in plants

by normal growth processes and regulated by IAA, ethylene can also be induced
by wounding and environmental stress, such as drought and infection. The amount
of ethylene produced by wounds differs in trees from different families. Also, the
high production rates in Abies firma, Pinus taeda, and P strobus are substantially
greater than the rates in other conifers. 136 However, no genetic control over
ethylene production within a species has been found. 137
Insect-or fungal-affected outer sapwood can react to the invasion by
microorganisms in lesions (for example in pines from Heterobasidium annosum
or Sirex noctiliolAmylostereum areolatum) by forming colored "protection wood"
or "reaction zones" defined as necrotic tissues that are enriched with inhibitory
extractives which are produced in a margin in advance of infection. 81 ,138 When
infected with H. annosum, the gas-filled, dry zone extends axially 0,1-1 m beyond
the infection in the inner pine sapwood, 61 and stilbenes (phytoalexins) and other
polyphenols are formed. 6,139 Ray cells are disorganized in advance of fungal
hyphae, and the polyphenols formed are concentrated at the edge of the reaction
zone, 6 The reaction zones in Picea abies affected by H. annosum accumulate
normal heartwood polyphenols but hydroxymatairesinol (2) increases 15 times
its normal concentration in heartwood in the narrow band encircling the affected
zone. 140 A 17-fold increase in ethylene, as compared with unaffected sapwood,
has been found in the transition zone oflesions caused by the Sirex-Amylostereum
complex,63,64 Ethylene is detectable in one day, extractives can be formed in 2
days and in considerable quantities in three weeks in dominant trees. The com-
position is different from that of heartwood, with antifungal stilbenes being the
major and toxic components,63,64
Tracheid cavitation and the formation of a dry zone by a pathogenic nema-
tode Bursaphelenchus xylophilus in Pinus spp. is the first symptom observed in
the two weeks following infection, 141 Kuroda proposed stimuli by the rapidly
moving nematode are significant for inducing the first physiological changes. 141
An accompanying excessive formation of volatile terpenes with low surface
tension by the ray parenchyma begins within three days of infection. These
terpenes are considered to be the cause of cavitation and dehydration of the
surrounding tracheids, In view of these findings, it appears that ethylene is
formed before the increased formation of terpenes, This nematode results in the
accumulation in the xylem of P strobus of two stilbenoid and two flavonoid
antifungal compounds. 142
The increase in ethylene production during the period of increased meta-
bolic activity and enhanced respiration can induce or activate enzymes, isoen-
zymes, and co-enzymes required for the formation of extractives. It could also
be responsible for the cavitation in the neighboring cells, allowing hydrostatic
tension to reduce the amount of water in the transition zone. However, the agent
initiating the formation of ethylene is unknown, In addition, ethylene is unlikely
to be the only inducer of increased activity, as injury of tissues can produce a
different ratio of components from that in heartwood, as exemplified by injury of

Pinus radiata sapwood. Furthermore, adjacent parenchyma can produce different
polyphenols found in the heartwood of some species.
CONTAINMENT OF DISCOLORED WOOD OR

FALSE HEARTWOOD
The first observable response to insect injury of sapwood containing resin

canals under pressure is the release of resin which limits the spread of sym-
biont fungi. Volatile and non-volatile components of the resin can have inhibitory
effects on the fungi. In situations where the resin pressure is insufficiently
high, the fungi can progress to activate the ray and axial parenchyma into the
production of compounds which differ in amount and type from heartwood
extractives.
The irregularly shaped discolored woods 143 in the interior of different types
of trees have received various classifications, such as facultative or false or patho-
logical heartwood, irregular heartwood, wound wood, red or blackheart. 6.144
Discolored woods are initiated by environmental factors and, in contrast to heart-
woods, their formation is not related to season. Wetwood, an area of water-soaked,
undecayed wood in heartwood (and in sapwood), is often darker than normal
heartwood and is often found in Abies and Tsuga spp.
Discolored woods may exist as large areas of irregularly shaped columns,
often with bands of different colored tissues not necessarily restricted to the center
of the trunk. These compartments, or localized containment systems, can be
formed in sapwood (occasionally in heartwood), and the extractives in them can
be different in type or amount from those in surrounding tissues. They may result
from external causes such as injury, fungal and microbial attack, or insect attack.
In contrast to heartwoods, their formation is not related to season. Compartmen-
talization is often a defense process by which boundaries are formed to isolate
the injured tissue and, thus, resist the spread of pathogens. 145 In other cases, non-
wood destructive organisms change the nature of localized compounds to permit
subsequent decay. As in heartwood, the cells are dead, and in some samples, the
discolored wood is enclosed by a transition, dry wood zone. The intensity of dark
color is often greatest in a thin margin at the periphery of the discolored wood
where, in the case of Nothofagus cunninghamii, fungal hyphae are detected in
greatest amount and the major portion of the extractives is polymerized mater-
ial. 71 The discolored woods of Robinia pseudoacacia and Maclura pomifera differ
from true heartwood in respect to color, frequency of deposits, moisture content,
and pH. 5 .65 Similar woods also are frequently observed in Fagus spp. and
Acer spp. The presence of rotten branches and injuries to the stem and roots of
F sylvatica are the main cause of discolored wood. 146 The major portion, if not
all, of the compounds in colored wood is polymerized material, and it appears
240 WE. HILLIS
as darker material adjacent to the sapwood boundary. Because of their large size
these polymerized materials are unable to penetrate the cell walls of the fibers in
contrast to the extractives of normal heartwood.
Considerable differences in the amounts of storage carbohydrates
and adenine nucleotides have been observed between different trunks of Fagus syl-
vatica. 104 However, with increasing depth in the trunks, these carbohydrates
decrease in amount up to the boundary of the discolored zone which does
not contain them. The pattern of radial distribution of adenine nucleotides, which
act as energy donors, was similar to the carbohydrates, particularly with the
decrease inside the discolored wood. This pattern indicates the synthesis of com-
pounds takes place at the sapwood-discolored wood boundary after a short phase
of enhanced metabolism, in a similar manner to heartwood extractives. 104
Discolored wood resulting from injury is part of a sequence which can lead
eventually-but not always-to decay. According to the vigor of the tree, the
severity of the wound, and the aggressiveness of the microorganisms, the changes
may stop at a particular stage. Dark sapwood discolorations may with subsequent
healing and diameter growth become incorporated into heartwood with darker
colors. 147
When invading microorganisms spread, they do so vertically through com-
partmentalized tissues. While the margins of the compartments remain intact, the
portion of the column distal to the wounds-the reaction zone-can still contain
pioneer microorganisms. In some cases, these compartments appear as dark bands
in sapwood and heartwood.
Dark irregular streaks characterize some species such as Jug/ans nigra,
Tectona gra n dis, and Acacia melanoxylon. Less common are the narrow axial
bands or streaks of varying color and intensity which are found in a regular or
characteristic arrangement in sapwood or usually heartwood. These streaks can
be localized in tangential directions, such as in Diospyros species,148 or be more
extensive as concentric groups. Most notable are the large number of narrow dark
brown to black streaks in a creamy yellow heartwood of Microberlinia brazzav-
illiensis and the red axial streaks in a brown heartwood of Berlinia Spp.98 Whether
the highly selective formation of specific compounds by the physiologically active
parenchyma is due to inherent genetic cellular differences or to different stimuli
of the parenchyma at the heartwood periphery is unknown.148.149
Stains may also form at the sapwoodlheartwood boundary, and their cause
is uncertain. Soil conditions or root injuries may provide the initial stimuli for
eventual stain formation. 147 The stains contain larger amounts of colored mater-
ial than heartwood, and in Quercus spp., they originate in the ray parenchyma and
eventually become incorporated in the heartwood. Stains can also represent the
early stages of attack by certain species of fungi on the non-living cells of nor-
mally colored, completely formed heartwood, and result in dark discolorations,
such as those caused by Stereum frustulatum, and Fistulina hepatica on oak,
Eucalyptus marginata, and other species. 6.'47 Such tissue has lower durability than
heartwood.
Target ring patterns of pale color have been observed in Thuja plicata,
Pseudotsuga menziesii, and other species6 and in moon rings of a few consecu-
tive growth rings in Quercus Spp.150 These paler colored rings contain less
extractives than the adjacent heartwood, although the ellagic acid content in
Quercus spp. is higher,'51 and have been attributed to an earlier, cold period when
heartwood formation was interrupted.
Wounds are prevented from becoming extensively colonized by pathogens
by two general types of responses: cytological and chemical. When living cells
gradually die under conditions suitable for cellular metabolism, starch, if present,
disappears, and antimicrobial compounds are formed in "reaction zones" to
inhibit decay organisms. 92 •l37 Fungal infection can result in the formation, in
wound or discolored wood, of uncharacteristic compounds, different from those
normally formed in the heartwood of that species. Normal heartwoods of Prunus
spp. contain flavonoids and proanthocyanidins, 15 I but P. domestica affected by
Stereum (now Chondrostereum) purpureum contains patches of the coumarin
scopoletin (24).153 P.jamasakura affected by Trametes (syn. Coriolus) versicolor,
contains large amounts of the lignan cyclo-olivil (25).154 P. yedoensis, affected by
Taphrina wiesneri,155 contains fluorescent gentisic acid (2,5 dihydroxy benzoic
acid) glycosides and other compounds. Stained wood, resulting from colonization
of C. purpureum in Malus pumila, contains at its boundary with sapwood the
biaryl compound aucuparin (26) and a pentacyclic triterpene (27).156 The same
pathogen formed A-pyrufuran (28) in Pyrus communi. 157 When the highly durable
heartwood in a living Thuja plicata is infected by a fungus, Sporothrix sp., a novel
lactone thujin (29) is produced from the thujaplicins e.g. 16. 158 Also, the discol-
ored heartwood in this species, possessing distinct, but irregular transverse
boundaries, can contain fungi including Cylindrocephalum sp. This fungus
reduces the amount of the toxic thujaplicins and lignans but does not affect the
structural integrity of the wood. '59
The amount and composition of terpenoids can also change markedly upon
fungal infestation or insect attack. The ratio of the components in the group of
sesquiterpenes resulting from infection of Ulmus glabra with Ceratocystis ulmi
was different from that when infected with Corio Ius versicolor, and also from
that infected with Chondrestereum purpureum. 160 When the sapwood of Pinus
contorta was attacked by Dendroctonus ponderosae and its associated microor-
ganisms, the amounts of total terpenes and particularly J3-phellandrene (30) were
much higher than those found in normal heartwood. 161
With Liriodendron tulip!era,162 the sapwood contained small amounts of a
single non-phenolic alkaloid-glaucine (31). This compound was the major com-
ponent of the heartwood extractives, which contained substantial amounts of five
other non-phenolic alkaloids, two phenolic alkaloids, two lignans, and one simple
242 W. E. HILLIS
CH30~OH OH
HO "- I I-oH
::::,.. OCH3
OH
24: Scopoletin 25: (+)-Cyclo-oliviI 26: Aucuparin
27: Triterpene 28: y-Pyrufuran
29: Thujin 30: p-Phellandrene 31: Glaucine
phenol. On the other hand, the discolored sapwood (probably originating from
fire) contained nine non-phenolic aporphine alkaloids, eight phenolic aporphine
alkaloids, and two lignans. The composition of these extractives differed from
those in the heartwood in the three discolored zones of the one cross-section. In
two of these zones, the amount was higher than in the heartwood, with glaucine
being the most abundant compound and varying five_fold. 163 In later studies, the
extraordinarily high concentration of glaucine in the discolored sapwood formed
in response to injury to the living tree and its high antifungal activity suggested
that injury induces the tree to stimulate the biosynthesis of glaucine thus
protecting itself against attack of microorganisms that invade the stem after
wounding. 163 This in turn causes deficiency in the O-methyltransferase system of
the tree, resulting in the formation of phenolic tetraoxygenated aporphine
alkaloids. 163 These antifungal alkaloids are probably metabolized in tum to
nonantifungal but antibiotic alkaloid pigments by the fungi.
FORMATION OF SPECIFIC COMPONENTS

OF EXTRACTIVES
Usually a mixture of components is found in heartwood extractives of a

species. Discrete and isolated formation of compounds can also occur. The lignan
conidendrin (4) was present in a high degree of purity in the white opaque parts of
floccosoids, or clusters oftracheids, in Tsuga heterophylla, whereas the major com-
ponent in the clear colorless parts was hydroxymatairesinol (3). The compounds
were not found in the rays but were in the pit cavities on either side of the mem-
brane and in the tracheid lumen. '64 Almost pure bayin (a flavone C-glucoside) (32)
sometimes fills some vessels of Castanospermum australe where the rays appear
empty.6 The complete extractives of Intsia bijuga heartwood contain mainly robi-
netin (a pentahydroxy flavonol) and 3 other flavonoids as well as tri- and tetra-
hydroxy stilbenes (33,34). These compounds require at least two biosynthetic
pathways. A few scattered vessels contain crystalline robinetin, which usually
fills the lumen. Adjacent vessels contain dark amorphous deposits presumably
containing stilbenes and other components. The parenchyma are empty. 165
HO
OH
HO yh
,;7
~I
I
R
h
oH
o OH
32: 8ayin 33: 3,S,4'-Trihydroxy stilbene (R~H)

34: 3,S,3',4'-Tetrahydroxy stilbene (R~OH)
The involvement of enzymatically active membranes of the half-bordered

pit pairs between ray parenchyma cells and vessels is clearly shown in Intsia
bijuga. The colorless precursors of the flavonol robinetin appear to pass through
enzymatically active membranes of the half-bordered pits on their route to the
vessel, with the distinctive yellow, water insoluble, birefringent robinetin emerg-
ing from the pits. '65 Apparently, the suite of enzymes attached to the pit mem-
branes of this particular vessel differs from those of some other vessels which
contain stilbenes derived from different precursors. Nobuchi et a1. 41 have observed
oil-like droplets on the tracheid side of the half-bordered pit-pairs between the
ray parenchyma cells and tracheids in the inner transition zone of Cryptomeria
japonica. No droplets were observed on the half-bordered pit-pairs between axial
parenchyma and tracheids.
Despite observations and proof that heartwood extractives, or their precur-
sors, are formed in living ray and axial parenchyma, there remains the unexplained
formation of compounds in a pure or concentrated condition in the shakes, splits
or cavities in the inner parts of large tree stems. Shakes are lengthwise and radial
244 WE. HILLIS
separations between cells and occur in short lengths from the pith. They have no
obvious connection with living tissues, and the deposits, which appear to be formed
in situ, continue to fill them when they enlarge with continued tree growth. 4
Almost pure or high concentrations of the lignans gmelinol (35) in shakes
of Gmelina leichhardtii and cyclo-olivil (25) in Olea cunninghamii, the aromatic
diterpene acid podocarpic acid (36) in shakes of several Dacrydium spp. and
Podocarpus spp, the methyl thujate (37) in Thuja plicata, and camphor (38) in
Dryobalanops aromatica and D. lanceolata have been reported. 6 Three different-
shaped deposits occurred close together in a heart check (split) in the pith region
of Tsuga heterophylla, and each deposit largely contained one of the lignans a-
conidendrin (4) or matairesinol (2) or hydroxymatairesinol (3).166 Other shakes in
T heterophylla contained a-conidendrin 166 or matairesinol. 166 Shake deposits in
T mertensiana contained matairesinol 167 as did those in Abies spp. and Pseudot-
suga menziesiii. 168
Deposits in the shakes of lntsia bijuga contained crystals of pure robinetin.
35: Gmelinol 36: Podocarpic acid
"'()' C""O
'OCH 3
37: Melhyllhujale 38: (+)-Camphor
A deposit in a shake of an Aftelia sp. showed adjacent irregular patches of yellow

and colorless crystals which were probably kaempferol-3-rhamnoside cf (8), and
(+)-dihydrokaempferol cf (7).4
SUMMARY
The width or volume of sapwood in a tree depends on genetic, biological,

and environmental processes. The storage materials present are mainly carbohy-
drates or lipids, which are used by living parenchyma cells for various purposes.
Phenolic and similar compounds are present in small amounts. The change from
sapwood to symmetrically-shaped heartwood is due to increased metabolism of
parenchyma cells at the transition zone or heartwood periphery after which all
cells are dead. The changes occur with the passive involvement of other cells.
Increased amounts of ethylene also are formed in some species in response to
insect injury of living sapwood.
A major feature of heartwood is the increased formation of extractives, or
secondary metabolites, or non-structural materials, of a complex composition
which is different from that of the sapwood. The amount and type of the com-
pounds is characteristic of the species, but the amount can be influenced by envi-
ronmental conditions. The formation of natural products probably arose during
evolution as a defense mechanism.
Heartwood extractives are formed at the periphery from starch stored in the
sapwood and from translocated carbohydrate from the cambium. After synthesis,
the parenchyma cells die, and the different cell compartments disintegrate, allow-
ing further changes to the compounds, such as non-specific oxidation and poly-
merization. Diffusion into the cell wall or entry into the lumens of neighboring
cells, notably the entry into vessels in angiosperms, occurs.
Various types of irregularly-shaped discolored wood (false heartwood) are
apparently a form of localized containment system resulting from injury. The
extractives are more highly polymerized or oxidized than those of heartwood
extractives and are less likely to diffuse into the cell wall. Heartwood, discolored
woods, and sapwood affected by microbial growth and other injuries can contain
compounds different in composition from those normally present in heartwood.
A particular fungus can result in different extractives in different species.
In a few species, components of extractives can be formed in a high degree
of purity in vessels of angiosperms, tracheids of gymnosperms, and even in shakes.
Why or how these vessels and tracheids accumulate these complex natural prod-
ucts, that are present in much smaller amounts, if at all, in neighboring tissues, is
not yet know. Understanding the mechanisms for these highly selective formations
could lead to processes to biosynthesize complex compounds without by-products
from simple substrates. These substrates could be provided by simple carbohy-
drates, such as those translocated from the leaves, and available in unlimited quan-
tities. Essential additions would be necessary to support specific enzyme systems.
Synthesis by enzyme systems, using simple inexhaustible resources, and requiring
minimum energy, would provide a more economical way to produce complex
natural substances than those now available.
ACKNOWLEDGMENTS
Dr. G. A Kile, ChiefCSIRO Forestry and Forest Products, provided facilities and
Ms. Heather Forster prepared several drafts of this chapter. I am grateful to Dr.
lA.F. Gardner, Dr. A.F.A. Wallis, and Mr. 1 Ilic for reviewing the manuscript.
246 WE. HILLIS
REFERENCES
1. HILLIS, W.E. 1989. Historical uses of extractives and exudates. In: Natural Products of
Wood Plants, J.W Rowe (ed.), Springer, Berlin, pp. 1-13.
2. HILLIS, WE. (ed.) 1962. Wood Extractives. Academic Press, New York. 513 p.
3. HILLIS, WE. 1996. Formation of robinetin crystals in vessels of Intsia species. lAWA 1.
14: 405-419.
4. HILLIS, WE. 1998. Deposits in heartshakes in wood. I and II. Wood Sci. Techno!' 32:
129-137; 139-143.
5. BAMBER, R.K., FUKAZAWA, K. 1985. Sapwood and Heartwood: A Review. Forest
Prod. Abstracts 8: 265-278.
6. HILLIS, WE. 1987. Heartwood and Tree Exudates. Springer Berlin, 268 p.
7. HIGUGHI, T. 1997. Biochemistry and Molecular Biology of Wood. Springer Berlin,
362 p.
8. WOLKINGER, E 1970. Morphologie and systematic verbreitung der lebenden holzfasern
bei striiucher und baumen. Holzforschung 24: 141-147.
9. HILLIS, WE. 1971. Distribution, properties and formation of some wood extractives.
Wood Sci. Techno!. 51: 272-289.
10. NOBUCHI, T. 1985. Cytological Studies on Heartwood Formation. Publication of Dept.
Wood Sci. Techno!. Kyoto Univ., Japan. 126 p.
11. TACHIBANA, S., SUMIMOTO, M. 1985. Studies on pitch problems 15. Mokuzai
Gakkaishi 31: 375-382.
12. FAHN, A. 1990. Plant Anatomy. Pergamon Press, London, 588 p.
13. YAMAMOTO, K. 1982. Yearly and seasonal process of maturation of ray parenchyma
cells in Pinus species. Res. Bul!. Exp. For. Hokkaido Univ. 39: 245-296.
14. BOSSHARD, H.H. 1968. On the formation of facultatively colored heartwood in
Beilschmiedia tawa. Wood Sci. Techno!. 2: 1-12.
15. YAMAMOTO, K., FUKAZAWA, K., ISHIDA, S. 1979. Study on the cell wall develop-
ment of ray parenchyma cells in genus Pinus. 3. Res. Bull. Exp. For. Hokkaido Univ.
38: 451-458.
16. YAMAMOTO, K., HAYASHI, Y. 1996. Radioautographic investigations on heartwood
formation in Zelkova serrata. In: Recent Advances in Wood Anatomy L.A. Donaldson,
A.P. Singh, B.G. Butterfield, L.1. Whitehouse (eds.), New Zealand For. Res. Inst.
pp. 198-203.
17. CHATTAWAY, M.M. 1949. The development of tyloses and secretion of gum in heart-
wood formation. Aust. 1. Sci. Res. B 2: 227-240.
18. HILLIS, WE., CARLE, A. 1958. The polyphenols and shikimic acid of eucalypt cambium
and wood. Holzforschung 12: 136-141.
19. HILLIS, WE., CARLE, A. 1963. The chemistry of eucalypt kino. IV Eucalyptus
hemiphloia kino. Aust. 1. Chern. 16: 147-159.
20. HILLIS, WE., HORN, D.H.S. 1965. Nuclear magnetic resonance spectra and structures
of some C-glycosyl f1avonoids. Aust. 1. Chern. 18: 531-542.
21. NELSON, N.D., HILLIS, WE. 1978. Genetic and biochemical aspects of kino vein for-
mation in Eucalyptus gomphcephala. II. Hormonal influence. Aust. For. Res. 8: 83-91.
22. ANONYMOUS 1957. International Glossary of Terms used in Wood Anatomy. Prepared
by the Int. Assoc. Wood Anatomists. Trop. Woods 107: 1-36.
23. HILLIS, WE. 1977. Secondary changes in wood. In: Recent Advances in
Phytochemistry. Vol. XI. EA. Loewus and vc. Runeckles (eds.). Plenum Press, pp.
247-309.
24. RINK, G. 1987. Heartwood color and quantity variation in a young black walnut progeny
test. Wood Fiber Sci. 19: 93-100.
25. RINK, G., PHELPS, IE. 1989. Variation in heartwood and sapwood properties among
10-year-old black walnut trees. Wood Fiber Sci. 21: 177-182.
26. WILKES, I 1991. Heartwood development and its relationship to growth in Pinus
radiata. Wood Sci. Techno!. 25: 85-90.
27. CLIMENT, I, GIL, L., PARDOS, I 1993. Heartwood and sapwood development and its
relationship to growth and environment in Pinus canariensis. For. Eco. Management.
59: 165-174.
28. de KORT, I. 1993. Relationships between sapwood amount, latewood percentage, mois-
ture content of Douglas-fir. lAWA I 14: 413--427.
29. YANG, K.c., HAZENBERG, G. 1992. Impact of spacings on sapwood and heartwood
thickness in Picea mariana and Picea glauca. Wood Fiber Sci. 24: 330-336.
30. ESPINOSA BANCALARI, M.A., PERRY, D.A., MARSHALL, ID. 1987. Leaf area-
sapwood area relationships in adjacent young Douglas-fir stands with different early
growth rates. Can. I For. Res. 17: 174-180.
31. HILLIS, W.E., DITCHBURNE, N. 1974. The prediction of heartwood diameter in radiata
pine trees. Can. I For. Res. 4: 524-529, 743.
32. LONG, IN., SMITH, F. W. 1988. Leaf area-sapwood area relationships of lodgepole pine
as influenced by stand density and site index. Can. 1. For. Res. 18: 247-250.
33. HAZENBERG, G., YANG, K.c. 1991. Sapwood/heartwood width relationships with tree
age in balsam fir. .lAWA I 12: 95-99.
34. RYAN, M.G. 1987. Sapwood volume for three subalpine conifers: Predictive equations.
Can. 1. For. Res 19: 1397-1401.
35. GERRISH, G. 1990. Relating carbon allocation patterns to tree senescence in
Metrosideros forests. Ecology 71: 1176--1184.
36. YANG, K.C., HAZENBERG, G. 1991. Relationship between tree age and sapwood/-
heartwood width in Populus tremuloides. Wood Fiber Sci. 23: 247-252.
37. YANG, G., HAZENBERG, G. 1991. Sapwood and heartwood width relationship to tree
age in Pinus banksiana. Can. 1. For. Res. 21: 521-525.
38. SELLIN, A. 1996. Sapwood amount in Picea abies determined by tree age and radial
growth rate. Holzforschung 50: 291-296.
39. SELLIN, A. 1994. Sapwood-heartwood proportion related to tree diameter, age and
growth rate in Picea abies. Can. 1. For. Res. 24: 1022-1028.
40. PAINE, T.D., MALINOSKI, M.K., SCRIVEN, G.T. 1990. Rating Eucalyptus vigor and
the risk of insect infestation: Leaf surface area and sapwood-heartwood ratio. Can. 1. For.
Res. 20: 1485-1489.
41. NOBUCHI, T., TAKAI, K., HARADA, H. 1987. Distribution of heartwood phenols in
the trunk of sugi (Cryptomeria japonica) and partial characterisation of heartwood
formation. Mokuzai Gakkaishi 33: 88-96.
42. CLIMENT, 1., GIL, L., PARDOS, 1.A. 1998. Xylem anatomical traits related to resinous
formation in Pinus canariensis. Trees: Structure and Function 12: 139-145.
43. BAMBER, R.K. 1976. Heartwood, its function and formation. Wood Sci. Techno!. 10:
1-8.
44. CREMER, K.W. 1993. Taper and heartwood in roots and tree stability. Aust. For. 56:
38--44.
45. SARANPAA, P., HOLL, w. 1989. Soluble carbohydrates of Pinus sylvestris sapwood and
heartwood. Trees: Structure and Function 3: 138-143.
46. SARANPAA, P., NYBERG, H. 1987. Seasonal variation of neutral lipids of Pinus
sylvestris sapwood and heartwood. Trees: Structure and Function I: 139-144.
47. NOBUCHI, T., TOKUCHI, N., HARADA, H. 1987. Variability of heartwood formation
and cytological features in broad leaved trees. Mokuzai Gakkaishi 33: 546--604.
48. NOBUCHI, T., TAKAHARA, S., HARADA, H. 1979. Studies on the survival rate of ray
248 WE. HILLIS
parenchyma cells with ageing process in coniferous secondary xylem. Bull. Kyoto Univ.
For. 51: 239-246.
49. KLUMPERS, K., SCALBERT, A., JANIN, G. 1994. Ellagitannins in European oak wood:
Polymerisation during wood ageing. Phytochemistry 36: 1249-1252.
50. VIRIOT, c., SCALBERT, A., HERVE du PENHOAT, C.L.M., MOUTOUNET, N. 1994.
Ellagitannins in woods of sessile oak and sweet chestnut, dimerisation and hydrolysis
during wood ageing. Phytochemistry 36: 1253-1260.
51. LAVISCI, P., SCALBERT, A., MASSON, D., JANIN, G. 1991. Quality of Turkey oak
(Quercus cerris) wood. Holzforschung 45: 291-295.
52. NOBUCHI, T., 1. SIRlRAT, 1., M. SAKAI. 1996. Seasonal changes of wood formation
and some characteristics of heartwood formation in teak (Tectona grandis) plantation.
Kasetsart 1. (Nat. Sci.) 30: 254-263.
53. HILLIS, WE., HASEGAWA, M. 1963. The formation of polyp heno Is in trees. 1. Admin-
istration of 14C glucose and subsequent distribution of radio activity. Phytochemistry 2:
195-199.
54. HASEGAWA, N., SHIROYA, M. 1968. Translocation and transformation of sucrose in
the wood of Prunus yedoensis. Bot. Mag. (Tokyo) 81: 141-144.
55. HRAZDINA, G. 1992. Biosynthesis of flavonoids, In: Plant Polyphenols. R.W
Hemingway, P.E. Laks (eds.). Plenum Press, pp. 61-72.
56. HILLINGER, G., HOLL, W, ZIEGLER, H. 1996. Lipids and lipolytic enzymes in the
trunkwood of Robinia pseudoacacia during heartwood formation. I,ll. Trees: Structure
and Function 10: 366-375; 376-381.
57. SARANPAA, P., PIISPANEN, R. 1994. Variation in the amount oftri-acylglycerols and
steryl esters in the outer sapwood of Pinus sylvestris. Trees: Structure and Function 8:
228-231.
58. DATTA, S.K., KUMAR, A. 1987. Histochemical studies of the transition from sapwood
to heartwood in Tectona grandis. IAWA Bull. 8: 363-368.
59. HIGUCHI, 1. 1976. Biochemical aspects of lignification and heartwood formation. Wood
Res. 59/60: 180-199.
60. COUTTS, M.P., RISHBETH, 1. 1977. The formation of wetwood in Grand fir. Eur. 1. For.
Path. 7: 13-22.
61. COUTTS, M.P. 1976. The formation of dry zones in the sapwood of conifers I. Eur. 1.
For. Path. 6: 372-381.
62. COUTTS, M.P. 1977. The formation of dry zones in the sapwood of conifers. II. Eur. 1.
For. Path. 7: 6-12.
63. SHAIN, 1., HILLIS, WE. 1972. Ethylene production in Pinus radiata in response to
Sirex-Amylostereum attack. Phytopath. 62: 1407-1409.
64. SHAIN, 1., HILLIS, WE. 1973. Ethylene production in xylem of Pinus radiata in rela-
tion to heartwood formation. Can. 1. Bot. 51: 1331-1333.
65. MAGEL, E., JAY-ALLEMAND, C, ZIEGLER, E. 1994. Formation of heartwood sub-
stances in the stemwood of Robinia pseudoacacia. II. Trees: Structure and Function. 8:
165-171.
66. SAKA, S., GORING, D.A.I. 1983. The distribution of inorganic constituents in black
spruce wood as determined by TEM-EDXA. Mokuzai Gakkaishi 29: 648-656.
67. OHASHI, H., IMAI, T., YOSHIDA, K., YASUE, M. 1990. Characterisation ofphysio-
logical functions of sapwood fluctuation of extractives in the withering process of
Japanese cedar sapwood. Holzforschung 44: 79-86.
68. OHASHI, H., IMAI, T. 1990. Characterisation of physiological functions of sapwood:
synthesis and accumulation of heartwood extractives in the withering process.
Holzforschung 44: 317-323.
69. OHASHI, H., KATO, N., IMAI, T., KAWAI, S. 1991. Characterisation of physiological
functions of sapwood. Fluctuation of heartwood extractives in the withering process of

Japanese cedar fed an inhibitor of phenylalanine ammonia lyase. Holzforschung 45:
245-252.
70. NOBUCHI, T., HARADA, H. 1983. Physiological features of the "white zone" ofsugi-
cytological structure and moisture content. Mokuzai Gakkaishi 29: 824-832.
71. FREY-WYSSLING, A., BOSSHARD, H.H. 1959. Cytology of the ray cells in sapwood
and heartwood. Holzforschung 13: 129-137.
72. CHATTAWAY, M.M. 1952. The sapwood-heartwood transformation. Aust. For. 16:
25-34.
73. MATSUKUMA, N., KAWANO, 1., SHIBATA, Y., KONDO, T. 1965. Studies on the inter-
mediate zone of sugi wood. Mokuzai Gakkaishi 11: 227-231.
74. SHAIN, L., MACKAY, 1.EG, 1973. Seasonal fluctuations in respiration of aging xylem
in relation to heartwood formation in Pinus radiata. Can. 1. Bot. 51: 737-741.
75. ZIEGLER, H. 1968. Biological aspects of heartwood formation. Holz Roh-Werkst. 26:
61-68.
76. HIGUCHI, T., ONDA, Y., FUKIMOTO, Y. 1969. Biochemical aspects of heartwood for-
mation with special reference to the site of biogenesis of heartwood components. Wood
Res. 48: 15-30.
77. MAGEL, E., DROUET, A., CLAUDOT, A.C., ZIEGLER, H. 1991. Formation of heart-
wood substances in the stem of Robinia pseudoacacia. Trees: Structure and Function
5: 203-207.
78. BAQUI, S.A., SHAH, 1.1 1985. Histoenzymatic studies in wood of Acacia auriculiformis
during heartwood formation. Holzforschung 39: 311-320.
79. CAIN, WE, SUESS, H.E. 1976. Carbon 14 in tree rings. 1. Geophysical Res. 81:
3688-3694.
80. WORBES, M., JUNK, Y.I 1989. Dating tropical trees by means of 14C from bomb tests.
Ecology 70: 503-507.
81. HARRIS, 1M. 1954. Heartwood formation in Pinus radiata. New Phytol. 53: 517-524.
82. NOBUCHI, T., MATSUNO, H., HARADA, H. 1985. Radial distribution of heartwood
phenols and the related cytological structure in the fresh wood of sugi (Cryptomeria
japonica). Mokuzai Gakkaishi 3: 711-718.
83. HOLL, W, LENDZIAN, K. 1973. Respiration in the sapwood and heartwood of Robinia
pseudoacacia. Phytochemistry 12: 975-977.
84. SHAH, J.J., BAQUI, S., PANDALAI, R.C., PATEL, K.P. 1981. Histochemical changes in
Acacia nilotica during transition from sapwood to heartwood. IAWA Bull. n.s. 2: 31-36.
85. NAIR, M.N., SHAH, J.1., PANDALAI, R.C. 1981. Wood anatomy and histochemical
changes of sapwood during heartwood formation in Bridelia retusa. Proc. Indian Acad.
Sci. (Plant Sci.) 90: 425-433.
86. SARANPAA, P., NYBERG, H. 1987. Lipids and sterols in Pinus sylvestris sapwood and
heartwood. Trees: Structure and Function I: 82-87.
87. SHAIN, L., MACKAY, lEG. 1973. Phenol-oxidizing enzymes in the heartwood of Pinus
radiata. For. Sci. 19: 153-155.
88. EBERMANN, R., STICH, K. 1982. Peroxidase and amylase isoenzymes in the sapwood
and heartwood of trees. Phytochemistry 21: 2401-2402.
89. EBERMANN, R., STICH, K. 1985. Distribution and seasonal variation of wood peroxi-
dase activity in oak. Wood Fiber Sci. 17: 391-396.
90. KUMAR, A., DATTA, S.K. 1989. Histochemical and histoenzymological changes during
heartwood formation in a timber tree Shorea robusta. Proc. Indian Academy Sci. (Plant
Sci.) 99: 21-27.
91. NAIR, M.N. 1988. Wood anatomy and heartwood formation of neem (Azadirachta
indica). Bot. I Linn. Soc. 97: 79-80.
250 W. E. HILLIS
92. SHAIN, L. 1979. Dynamic responses of differentiated sapwood to injury and infection.
Phytopath. 69: 1143-1147.
93. MAGEL, E., HUBNER, B. 1997. Distribution of phenylalanine ammonia-lyase and chal-
cone synthase within trunks of Robinia pseudoacacia. Botanica Acta 110: 314-322.
94. MAGEL, E.A., DROUET, A., CHALUDOT, A.C., ZIEGLER, H. 1991. Formation of
heartwood substances in the stem of Robinia pseudoacacia. Trees: Structure and
Function 5: 203-207.
95. NOBUCHI, T., SATO, T., IWATA, R., HARADA, H. 1984. Season of heartwood forma-
tion and the related cytological structure of ray parenchyma cells in Robinia pseudoaca-
cia. Mokuzai Gakkaishi 30: 628-636.
96. NOBUCHI, T., KAMIZONO, Y, HARADA, H. 1976. Cytological changes of
parenchyma cells associated with heartwood formation in three softwood species. Bull.
Kyoto Uni. For. 48: 178-186.
97. NAIRN, P.M. 1955. Wood Specimens: 100 Representations In Colour. Nema Press,
London, 206 p.
98. HILLIS, W.E. 1996. Coloured streaks in wood. In: Recent Advances in Wood Anatomy.
L.A. Donaldson, A.P. Singh, B.G. Butterfield, L.1. Whitehouse (eds.). New Zealand Forest
Res. Institute, pp. 295-300
99. MACLEAN, 1., GARDNER, 1.A.E 1958. Distribution of fungicidal extractives in target
ring pattern heartwood of western red cedar. For. Prod. 1. 8: 107-108.
100. DUJESIEFKEN, D., LIESE, w., BAUCH, 1. 1984. Discolouration in the heartwood of
oak trees. IAWA. Bull. n.s. 5: 128-132.
101. PURITCH, G.S. 1977. Distribution and phenolic composition of sapwood and heartwood
in Abies grandis and effects of the balsam woolly aphid. Can. 1. For. Res. 7: 54-62.
102. HILLIS, W.E., HUMPHREYS, ER., BAMBER, R.K., CARLE, A. 1962. Factors
influencing the formation of phloem and heartwood polyphenols. Holzforschung. 16:
114-121.
103. FISCHER, c., HOLL, W. 1992. Food reserves of Scots pine Pinus sylvestris. II. Seasonal
changes and radial distribution of carbohydrate and fat reserves in pine wood. Trees:
Structure and Function 6: 147-155.
104. MAGEL, E., HOLL, W. 1993. Storage carbohydrates and adenine nucleotides in trunks
of Fagus sylvatica in relation to discolored wood. Holzforschung 47: 19-24.
105. HUGENTOBLER, U.H. 1965. Zur cytologie der kern-holzbildung. Viereljahr
Naturforsch. Gcs. Zurich 110: 321-343.
106. MAGEL, E., HAUCH, S., HUBNER, B. 1996. Cytological and biochemical changes
during heartwood formation (abstract). IAWA. 1. 17: 257-258.
107. NOBUCHI, T., OHMURA, w., SAIKI, H. 1991. Histochemistry ofray parenchyma cells
and tracheids associated with biosynthesis of heartwood phenols in Japanese cedar.
Mem. Coil. Agric. Kyoto Univ. No. 138: 1-10.
108. SARANPAA, P. 1988. Plastids and glycolipids in the stemwood of Pinus sylvestris. Trees:
Structure and Function 2: 180-187.
109. SARANPAA, P. 1990. Heartwood formation in stems of Pinus sylvestris. Public. Dept.
Botany University of Helsinki No. 14: 1-87.
110. OKADA, N., KATAYAMA, Y, NOBUCHI, T., ISHIMARU, Y, YAMASHITA, 1., AOKI,
A. 1987. Trace elements in the stems of trees. I. Mokuzai Gakkaishi 33: 913-920.
Ill. OKADA, N., KATAYAMA, Y, NOBUCHI, T., ISHIMARU, Y, AOKI, A. 1993. Trace ele-
ments in trees. V. Comparisons of radial distributions among softwood stems. Mokuzai
Gakkaishi 39: 1111-1118.
112. OKADA, N., KATAYAMA, Y., NOBUCHI, T., ISHIMARU, Y., AOKI, A. 1993. Trace ele-
ments in the stems of trees. VI. Comparisons of radial distribution among hardwood
stems. Mokuzai Gakkaishi 39: 1119-1127.
113. BROWNRIDGE,1.D. 1984. The radial distribution ofCS ' ]7 and K 40 in tree stems. 1. Plant
Nutrition 7: 887-896.
114. OKADA, N., SATO, M., KATAYAMA., Y, NOBUCHI, T., ISHIMARU, Y., YAMASHITA,
H., AOKI, A. 1988. Trace elements in the stems of trees. II. Mokuzai Gakkaishi 34:
874-880.
115. TANEDA, K., OTA, M., NAGASHIMA, M. 1986. The radial distribution and concen-
tration of several chemical elements in woods of five Japanese species. Mokuzai
Gakkaishi 32: 833-841.
116. DE VISSER, P.H.B. 1992. The relations between chemical composition of oak tree rings,
leaf bark, and soil solution in a partly mixed stand. Can. 1. For. Res. 22: 1824-1831.
117. ROWE, 1. W (ed.) 1989. Natural Products of Woody Plants. Vo!. I and II Springer, Berlin
1243 p.
118. NAULT, 1. 1988. Radial distribution of thujaplicins in old growth and second growth
western red cedar, Thuja plicata. Wood Sci. Techno!. 22: 73-80.
119. COTE, WA., DAY, A.c., SIMSON, B.W, TIMELL, T.E. 1966. Studies on larch ara-
binogalactan I. Holzforschung 20: 178-192.
120. GARDNER, 1.A.F., BARTON, G.M. 1960. The distribution of dihydroquercetin in
Douglas-fir and western larch. For. Prod. 1. 10: 171-173.
121. DELLUS, Y, MILA, I., SCALBERT, A., MENARD, C., MICHON, Y, HERVE du
PENHOAT, C.L.M. 1997. Douglas-fir polyphenols and heartwood formation. Phyto-
chemistry 54: 1573-1578.
122. PENG, S., SCALBERT, A., MONTlES, B. 1991. Insoluble ellagitannins in Castanea
sativa and Quercus petraea woods. Phytochemistry 30: 775-778.
123. BLAND, D.E., HILLIS, WE. 1969. Microspectrophotometric investigation of lignin and
polyphenol distribution in wood sections. Appita 23: 204-210.
124. STREIT, W, FENGEL, D. 1994. Heartwood formation in quebracho colorado: Tannin
distribution and penetration of extractives in cell walls. Holzforschung 48: 361-367.
125. STREIT, W, FENGEL, D. 1994. On the changes of the extractives composition during
heartwood formation in quebracho colorado. Holzforschung 48: 15-20 (supp!.).
126. KUO, M.L., ARGANBRIGHT, D.G. 1980. Cellular distribution of extractives in redwood
and incense cedar. I, II. Holzforschung 34: 17-22; 41- 47.
127. HERGERT, H.L. 1992. The nature of non-proanthocyanidin units in condensed tannins
from conifer wood and bark. In: Plant Polyphenols R.W Hemingway, P.E., Laks, (eds.)
Plenum Press, New York, pp. 385-409
128. LANGE, W, KUBEL, H., WEISSMAN, G. 1989. Die verteilung der extraktstoffe in
stammholz von Pinus sylvestris. Holz als Roh-u. Werkstoff 47: 487-489.
129. HILLIS, WE. 1966. Variation in polyphenol composition within species of Euca~vptus.
130. KURODA, H., SHIMAJI, K. 1983. Distribution of coloring substances in sugi heartwood.
131. TACHIBANA, S., SUMIMOTO, M. 1985. Studies on pitch problems caused by pulping
and bleaching of tropical woods. XV Mokuzai Gakkaishi 31: 375-382.
132. NELSON, N.D. 1978. Xylem ethylene, phenol-oxidising enzymes and nitrogen and heart-
wood formation in walnut and cherry. Can. J. Bot. 56: 626--634.
133. NELSON, N.D. 1975. Extractives produced during heartwood formation in relation to
amounts of parenchyma in Juglans nigra and Quercus ruhra. Can. 1. For. Res. 5: 291-301.
134. JAYABALAN, M., SHAH, J.J., RAJARATHINAM, K., PAUL SING, G.A.D.,
VEERASAMY 1994. Experimental induction of heartwood formation by ethephon in
Mesuaferrea. J. Ecotox and Environ. Monitor. 4: 141-146.
135. YAMAMOTO, F., ANGELES, G., KOZLOWSKI, T.T. 1987. Effect of ethrel on stem
anatomy of Ulmus americana seedlings. IAWA Bull. n.s. 8: 3-9.
252 W E. HILLIS
136. YAMANAKA, K. 1986. Wound ethylene production by phloem and cambium tissues in
conifers. Mokuzai Gakkaishi 32: 136-139.
137. NELSON, N.D., RIETVELD, WI, ISEBRANDS, J.G. 1981. Xylem ethylene production
in five black walnut families in the early stages of heartwood formation. For. Sci. 27:
537-543.
138. SHAIN, L. 1995. Stem defense against pathogens. In: Plant Stems: Physiology and Func-
tional Morphology. B.L. Gartner (ed.). Academic Press, San Diego, pp. 383-406
139. HART, IH. 1981. Role of phytostilbenes in decay and disease resistance. Ann. Rev.
Phytopath. 19: 437-458.
140. SHAIN, L., HILLIS, WE. 1971. Phenolic extractives in Norway spruce and their effects
on Fomes annosus. Phytopath. 61: 841-845.
141. KURODA, K. 1991. Mechanism of cavitation development in the pine wilt disease. Eur.
I For. Path. 21: 82-89.
142. YAMADA, T., ITO, S. 1993. Chemical defense responses of wilt-resistant pine species,
Pinus strobus and P taeda, against Bursaphelenchus xylophilus infection. Ann. Phytopath.
Soc. Japan 59: 666-672.
143. BAUCH, I, BAAS, P. (eds.) 1984. Development and characteristics of discolored wood.
IAWA Bull. n.s. 5: 91-154.
144. SHIGO, A.L., HILLIS, WE. 1973. Heartwood, discolored wood and microorganisms in
living trees. Ann. Rev. Phytopath. 11: 197-222.
145. SHIGO, A.L. 1984. Compartmentalization: A conceptual frameworok for understanding
how trees grow and defend themselves. Ann. Rev. Phytopath. 22: 189-214.
146. RIEDER, A. 1997. Scope for influencing coloured-heartwood formation in beech.
Osterreichische Forstzeitung 108: 34-26.
147. SACHS, I.B., WARD, IC., BULGRIN, E. 1966. Heartwood stain in red oak. Holz
Roh-Werkst. 24: 489-497.
148. HILLIS, WE., SOENARDI, P. 1994. Formation of ebony and streaked woods. IAWA. I
15: 425-437.
149. HILLIS, WE. 1996. Colored streaks in wood, In: Recent Advances in Wood Anatomy.
Donaldson, L.A., Singh, A.P., Butterfield, B.G., Whitehouse, LJ. (eds.) New Zealand
Forest Res. Inst. pp. 295-300.
150. DUJESIEFKEN, D., BAUCH, I 1987. Biologische charakterisierung von eichenholz mit
mondring. Holz Roh-Werkst. 45: 365-370.
151. CHARRIER, B., JANIN, G., HALUK, IP., MOSEDALE, J.R. 1995. Color and chemical
characteristics of moon rings in oakwood. Holzforschung 49: 287-292.
152. HASEGAWA, M. 1958. Flavonoids contained in Prunus woods. I Japan For. Soc. 40:
111-121.
153. HILLIS, WE., SWAIN, T. 1959. Studies on the phenolic constituents of Victoria plum.
II. J. Sci. Food Agric. 10: 135-144.
154. HASEGAWA, M., SHIRATO, T. 1959. Abnormal constituents of Prunus woods. J. Japan
For. Sc. 40: 111-121.
155. FUJII, S., AOKI, J., KOMOTO, M., MUNAKATA, K. 1971. Fluorescent glucosides
accumulated in Taphrina wiesneri-infected cherry stems. Agric. BioI. Chern. 35:
1526-1541.
156. KEMP, M.S., HOLLOWAY, P.J., BURDEN, R.S. 1985. A pentacyclic triterpene induced
in the wood of Malus pumila infected with Chondostereum purpureum. 1. Chern. Soc.
M: 1848-1876.
157. KEMP, M.S., BURDEN, R.S. 1984. Isolation and structure determination ofy-pyrufuran
from the wood of Pyrus communis. 1. Chern. Soc. Perkin Trans. 1: 1441-1443.
158. JIN, L., WILSON, J.W, SWAN, E.P. 1988. Thujin, a novel lactone isolated from the dis-
colored heartwood of Thuja plicata. Can. 1. Chern. 66: 51-58.
159. van der KAMP, BJ. 1986. Effects of heartwood inhabiting fungi on thujaplicin content
and decay resistance of western red cedar. Wood Fiber Sci. 18: 421-427.
160. BURDEN, R.S., KEMP, M.S. 1984. Sesquiterpene phytoalexins from Ulmus glabra.
161. SHRIMPTON, D.M. 1973. Extractives associated with wound response of lodgepole
pine attacked by the mountain pine beetle and associated microorganisms. Can. l Bot.
51: 527-534.
162. CHEN, c.L., CHANG, H.M., COWLING, E.B., HUANG HSU, c.v., GATES, R. 1976.
Aporphine alkaloids and lignans formed in response to injury of sapwood in Lirioden-
dron tulipifera. Phytochemistry 15: 1161-1167.
163. HUANG HSU, c.v., CHEN, c.L. 1991. Antimicrobial activities of aporphine alkaloids
isolated from heartwood and discolored sapwood of Liriodendron tulipijera.
164. KRAHMER, R.L., HEMINGWAY, R.W, HILLIS, WE. 1970. The cellular distribution of
lignans in Tsuga heterophylla wood. Wood Sci. Techno!. 4: 122-129.
165. HILLIS, WE. 1996. Formation ofrobinetin crystals in vessels of lntsia species. lAWA 1.
17: 405-419.
166. MEYER, R.W, LENEY, L. 1968. Shake in coniferous wood, an anatomical study. For.
Prod. 1. 18: 51-56.
167. SCHROEDER, H.A., KOZLIK, C.J. 1972. The characterisation of wetwood in western
hemlock. Wood Sci. Techno!. 6: 85-94.
168. BARTON, G.M., GARDNER, lA.F. 1962. The occurrence of matairesinol in mountain
hemlock, western hemlock and balsam. l Org. Chern. 27: 322-323.
Chapter Ten
RECENT ADVANCES IN THE CHEMISTRY OF

PROANTHOCYANIDINS
Daneel Ferreira, Hendrik van Rensburg, Elfranco Malan,

lohan Coetzee, and Reinier 1. 1. Nel
FRD Research Unit for Polyphenol- and Synthetic Chemistry

Department of Chemistry
University of the Orange Free State
P.O. Box 339
Bloemfontein, 9300 South Africa
Introduction ................................................... 255

Synthesis of Oligomeric Proanthocyanidins .......................... 256
Synthesis of the First Profisetinidins with Epifisetinidol
Constituent Units .......................................... 256
Synthesis of Procyanidins under Neutral Conditions ................ 259
Synthesis of Ether-Linked Proteracacinidins ....................... 262
Miscellaneous ............................................... 267
Cleavage of the Interflavanyl Bond of Proanthocyanidins ............... 268
B-Type Proanthocyanidins ..................................... 268
A-Type Proanthocyanidins ..................................... 271
Enantioselective Synthesis of Flavonoids ............................ 275
Dihydroftavonols ............................................. 275
Flavan-3-ols ................................................ 277
Conformational Analysis of Dimeric Proanthocyanidins ................ 278
Conclusion .................................................... 282
INTRODUCTION
The oligo- and polymeric proanthocyanidins represent a major group of

phenolic constituents in woody and some herbaceous plants. 1•2 These compounds
usually originate by coupling at C(4) (C-ring) of an electrophilic ftavanyl unit,
255
256 D. FERREIRA et al.
generated from a fIavan-4-01 1 or a fIavan-3,4-dioI 3 (the chain extender units), to

a nucleophilic fIavanyl moiety, often a fIavan-3-011.3 (the chain terminating unit).
Recent developments were initiated by the numerous industrial applications and
by the growing realization that the oligo- and polymeric proanthocyanidins may
be credited for the profound health-promoting properties of tea, fruit juices, and
red wine. This is mainly due to their in vitro radical scavenging4 or antioxidant5
biological properties, while the polyflavanoids in red wine have recently been
implicated in protection against cardiovascular disorders/ e.g. the "French
paradox".7-9 The realization that many of these effects are closely associated with
the intramolecular interactions of the proanthocyanidins with other biomolecules
has triggered a substantial research effort towards understanding the chemical
dynamics of this complex group of natural products. The most relevant results
emanating from these studies constitute the subject of this review. The nomen-
clature delineated in reference I will be consistently used.
SYNTHESIS OF OLIGOMERIC PROANTHOCYANIDINS
The limited number of options that are available for the synthesis of proan-
thocyanidin oligomers has been comprehensively reviewed 10- 13 and need not be
discussed here. In principle, a chain extender fIavanyl unit with a good leaving
group (typically OH or SCH2Ph) at C(4) is converted into a strongly electrophilic
center (typically C(4) carbocation or A-ring quinone methide) which is then
trapped by the chain terminating nucleophilic fIavanyl unit, typically a fIavan-3-
01 moiety, to give the regiomeric (4~6)- and (4~8)-bifiavanoids. These then
serve as precursors to trifiavanoids via coupling with the electrophile at the A-
or D-rings, and eventually to tetraflavanoids and higher oligomers (Fig. I).
In the profisetinidin (RI = H), i.e. 5-deoxy (A-ring) series of compounds, the
bifiavanoids have a preference for D-ring coupling in the transformation to the
trimers, while the procyanidin-type bifiavanoids (RI = OH), i.e. 5-oxy (A-ring),
exhibit a remarkable propensity for coupling at C(8) (A-ring). For chain termi-
nating moieties possessing B-rings with nucleophilicity comparable to that of the
A-ring, the former rings often compete favorably as nucleophiles in the process
of interflavanyl bond formation. 13 ,14
A recent application of this protocol to the synthesis of a unique series of
di- and trimeric profisetinidins, and a novel utilization of 4-benzylsulfanylfIavan-
3-01s in the controlled synthesis of di- and trimeric procyanidins under neutral
conditions and some related transformations are now discussed.
Synthesis of the First Profisetinidins with Epifisetinidol

Constituent Units l5
The profisetinidins, with their 3',4',7-trihydroxyflavan-3-ol extender units,
are the most important polyflavanoids of commerce, forming the major con-
RECENT ADVANCES IN THE CHEMISTRY OF PROANTHOCYANIDINS 257
6.&OH
HO
('~OH
OH
LG = OH or SCH2Ph (electrophilic chain extender flavanyl unit)
I
H@forLG=OH
HO~ for LG = OH or SCH2Ph
OH ~OH
~ 0yyol'I"'~OH
HOW
0 l II'···~OH
~
@
OH ~OH
R' R'
C-4 carbocation A-ring quinone methide
~OH ~OH
HO
"""~OH """~OH
OH
nucleophilic flavan-3-ol chain
terminating unit
~OH
HO
""'~OH
~
IE ~OH
I""'~OH
Q OH
OH
(4 - 6)-biflavanoids OH
OH
(4 - 8)-biflavanoids
Tetraflavanoids, etc. _Repetition Triflavanoids via coupling at A- or D-ring with
C-4 carbocation or A-ring quinone methide
Figure 1. Flavan-3,4-diols or 4-benzylsulfanylflavan-3-0Is and flavan-3-0Is as precursors to

profisetinidin (R' = H) or procyanidins (R' = OH) oligomers with (2R,3S)-constituent flavanyl
units.
stituents of wattle and quebracho tannins (see ref. 15 for appropriate references).
Naturally occurring oligomers exhibit predominantly 2,3-trans relative stereo-
chemistry and possess either 2R,3S or 2S,3R absolute configurations. 10.' ,
Profisetinidins exhibiting 2,3-cis relative configuration of the chain extender
moieties are extremely rare and are hitherto restricted to two tentative (4~6)
bis-fisetinidols from Colophospermum mopane,'6 a promelacacinidin from
Acacia melanoxylon,17 and some proteracacinidins's from A. galpinii's and A.
caffra.'9
The methanol extract of the bark of Pithecellobium dulce (Roxb.) Benth
(Guamuchil, Madras thorn) afforded a series of mono-, di- and trimeric pro-
fisetinidins exhibiting both 2,3-trans and 2,3-cis relative configurations of the
constituent fisetinidol moieties,'5 thus offering the opportunity to rigorously cor-
roborate the structures of the 2,3-cis analogues via synthesis. Such an approach
was additionally motivated by the inability to unequivocally differentiate between
2,3-cis-3,4-trans- and 2,3-cis-3,4-cis-configurations of ABC-units in e.g. com-
pound 4 on the basis of 'H NMR coupling constants.
Thus, separate treatment of epifisetinidol-4~-01 (1) with catechin (2)
and epicatechin (3) under mild acidic conditions afforded the epifisetinidol-
(4~~6)- and (4~~8)-catechins (6) and (4), epifisetinidol-(4~~6)- and (4~~8)
epicatechins (7 and 5), respectively (Fig. 2), both couplings proceeding highly
stereoselectively.20 'H NMR data of the permethylaryl ethers in conjunction with
the circular dichroic properties,z'-23 based on the aromatic quadrant rule,z4 then
permitted unambiguous structure definition of compounds 4-7 and hence of the
natural products 4 and 5. Acid-catalyzed condensation of epifisetinidol-4~-01 (1),
with either catechin (2) or epicatechin (3) in a I : 6 molar ratio, stereoselectively
afforded both the dimeric profisetinidins 4 or 5 and the angular trimeric pro-
fisetinidins 8 or 9. The appropriate derivatizations gave the permethylaryl ether
triacetates with 'H NMR and CD data identical to those of the same derivatives
of the natural products.
The elegance of this simple biomimetic approach to the synthesis of proan-
thocyanidin oligomers was next demonstrated during synthesis of the "mixed"
profisetinidin trimers 10 and 11, i.e. analogues possessing different chain exten-
der units. Trimer 10 with its fisetinidol ABC- and epifisetinidol GHI-unit was
formed by acid-catalyzed reaction of the fisetinidol-(4a~8)-catechin biftavanoid
(12)25 and epifisetinidol-4~-01 (1). The remaining triftavanoid (11) with its epi-
catechin DEF-unit was similarly synthesized using the epifisetinidol-(4~~6)
epicatechin (7) (vide supra) in the acid-catalyzed condensation with fisetinidol-
4a-ol (13). Comparison of the 'H NMR and CD data of the pennethylaryl
ether triacetates of trimers 10 and 11 with those of the same derivatives of
the natural products, again provided unequivocal structural proof for the latter
compounds.
H0'(l(0i:: ex
'" OH I
OH
HO yyvi ",,·: : ,. .
~OH
0 ex
~ OH
OH
~'OH
OH
OH
2 - ,
3 -
~OH
HO
"""~OH
HO
OH
~OH
"""~OH
OH
, ~
OH IE
~ OH
,
4 -
OH
5 -
6 -
Figure 2. Synthesis of dime ric profisetinidins with epifisetinidol chain extender units.
Synthesis of Procyanidins under Neutral Conditions
Owing to the lability of the interflavanyl bond in procyanidins under either

acidic or basic conditions, the existing semi-synthetic methods'3 invariably lead
to an equilibrium between substrates and products. Such a labile bond and an
apparent preference of the electrophile for the di- and trimeric products (see Fig.
1) once condensation is initiated, furthermore, gives poor control regarding the
level of oligomerization. We, thus, assessed the effectiveness of the thiophilic
Lewis acids, dimethyl(methylthio )sulfonium tetrafluoroborate (DMTSF)27,28 and
silver tetrafluoroborate (AgBF4),29 to activate the C(4)-S bond in the 4-thioethers
~OH
o"" .. ~OH
II"
OH ~OH
HOJ: "" .. ~OH
~J.., "I
OH
OH
8 ~ =,
9 ~ = ~OH
HO "",.~OH
OH ~OH
HOJ:
""'~OH
0""" OH
OH
10 ~ =,
11 ~ =-=
HO
~I~'l'"
~
0
""".::::.... r8:IB
0H
OH
~OH
~. OH &OH HOYY°'iI""~
~
HO '8 0 E I OH
""'.::::....
~
DI
::::.... F OH = OH
OH QH
OH 12 13
(JC 0H (JC 0H
HO ",,,,1
HOycxo
,,"" ~ 1 OH
" W O ,I' ,.. OH 1
1
~ OH
~ ""'OH OH
OH SCH2Ph
2 ~-
14
3 ~-
DMTSF or AgBF 4
R
HO
"" •.
(JC
~1
0H
OH
15 ~ " "R= H
16 ~ " ! ,R = 413-epicatechin
17~,,=,R=H
Figure 3. Synthesis of procyanidin B-1 15 and B-2 16 using Lewis acid activation of 4~
benzy1su1fany1epicatechin 14.
of flavan-3-ols towards carbon nucleophiles, and hence to generate the inter-

flavanyl bond of procyanidins under neutral conditions. 30
Thus, treatment of a mixture of 4~-benzylsulfanylepicatechin (14), repre-
senting the (2R,3R)-2,3-cis-flavan-3-ol chain extender unit of the procyanidins,
and catechin (2), in THF with DMTSF (1: lO: 1 molar ratio) or AgBF4 (1: 10:2
: 5 molar ratio) at or below DoC, afforded procyanidin B-1 (15) (22 and 38% for
DMTSF and AgBF 4, resp.) and the analogous trimeric procyanidin (16) (10% for
DMTSF, not formed with AgBF4) (Fig. 3). Since aqueous work-up of the DMTSF
reaction afforded an acidic reaction mixture (PH ca. 2-5), all subsequent runs
were worked-up in a pH 6.85 buffer solution. The moisture sensitivity ofDMTSF
and the improved yield of 15 recorded for AgBF 4 prompted us to abandon the
former reagent. When a mixture of 413-benzylsulfanylepicatechin (14) and epi-
catechin (3) in THF was treated with AgBF4 at the above molar ratios, procyani-
din B-2 (17) was formed in 37% yield, without evidence of the formation of
regiomeric dimers or of higher oligomers. This protocol, thus, compares favor-
ably with the classical acid-catalyzed condensation of catechin-4a-ol and cate-
chin (2),20.26 which gave a mixture of procyanidins B-3 (20) and B-6, the trimeric
procyanidin C(2) (21) (ef Fig. 4) and its 4,6-regioisomer, and the presumed all-
trans-(4-t8)-linked tetraflavanoid analogue (10: 1 : 12: 1 : 3) (45% overall yield).
The scope of the thiophilic Lewis acid mediated interflavanyl bond forma-
tion was extended using 4-benzylsulfanylcatechin (18) (4: 1 mixture of 4~- and
4a_epimers),2o.26 representing the (2R,3S)-2,3-trans-flavan-3-01 extender unit of
the procyanidins, as source of the flavan-3-01 C(4) electrophilic moiety (Fig. 4).
Separate treatment of a mixture of the epimeric 4-benzylsulfanylcatechins (18)
and catechin (2) and epicatechin (3) in THF with AgBF4 afforded procyanidin
B-3 (20) (35%) and B-4 (19) (51%), respectively. The preferences for the for-
mation of 4~- and 4a-interflavanyl bonds by using the epicatechin- and catechin-
4-thiobenzyl ethers (14 and 18), respectively, and for the (4-t8)-interflavanyl
linkages were anticipated,2o.26 i.e. the thiobenzyl ethers are converted by Lewis
acids into relatively stable intermediates permitting both the regioselective attack
of the nucleophile via C(8), where the HOMO displays maximum amplitude, and
the stereoselectivity by approach from the sterically least hindered side.
Finally, procyanidin B-3 (20) and the 4-benzylsulfanylcatechin epimeric
mixture (18) were coupled by using AgBF 4 in THF to give the trimeric pro-
cyanidin (21) (26%) as the only isolable product (Fig. 4). The sequences towards
the procyanidins depicted in Figures 3 and 4 from using AgBF 4 as the thiophilic
Lewis acid, no doubt, offer advantages as far as control over the level of oligomer-
ization, reversibility, and "scattering" of the interflavanyl bond(s) are concerned
in comparison with the formation of these products under conditions previously
developed (see ref. 13 and references cited therein).
Synthesis of Ether-Linked Proteracacinidins

Proanthocyanidins possessing ether-type interfiavanyl linkages are ex-
tremely rare except for the A-type oligomers which contain the conventional
C(4)-tC(6)1(8) bond as well as an additional ether linkage connecting C(2) (C-
ring) and C(5)1C(7) (D-ring).!,2 Analogues possessing exclusive ether bonds are
hitherto restricted to the 1,4-dioxane-type profisetinidins,3!.32 the recently reported
(4-t7: 5-t5) doubly-linked proteracacinidin,!9 and two (C c O-C 4)-promela-
cacinidins. 33 Recently,34,35 we identified the first two [4-0-3]-linked bis-
teracacinidins (22 and 23), as well as two [4-0-4]-linked analogues (24 and 25)
of the same class. Since no attempt was previously made to synthesize the singly
ether-linked compounds or to establish their stereochemistry, we explored possi-
ble semi-synthetic routes towards compounds 22-25.
It was anticipated that the C(4) benzylic ether bonds in proteracacinidins
22-25 would be highly susceptible to solvolysis in aqueous acidic medium.
(X 0H (X
7"1
0H
Ho):CXo
"" .. ::::,... OH
HOWO"",.::::,... I OH
I :::::,...
1
OH
::::,... OH
H
OH SCH2Ph
2
18 ~ == i and ~ (41)
3
AgBF4
(X
W
0H
HO 0
I "" . : : :,. .
1
OH
:::::,... ~ OH (X0H
(X OH ';
yyvi""':::::"" 0 I
W"
0H HO
OH
HO 0 ·1
~OH
OH OH
~H ~
W
OH ( X IOH 19
HO ' 0
::::,...I "" .. ::::,... OH
~ OH (X0H
I
OH
HO
yyv i """:: :""
'
' 0
OH
~OH
OH
21
Figure 4. Synthesis of procyanidin B-3 20, B-4 19 and C-2 21 using Lewis acid activation of
the 4-benzylsulfanylcatechin epimers 18.
These conditions,20 which have been applied universally for the formation of
C(Sp3)~C(Sp2) interflavanyl linkages, would be less applicable to the formation
of the ether bonds in compounds 22-25. We, thus, selected to activate the elec-
trophilic attributes of one of the flavan-3,4-diol methyl ethers, e.g. 26, via for-
mation of the 4-chloroflavan-3-ol derivative 27, in order to permit the formation
of the crucial ether bond at a near neutral pH value.
°H (£rI 0H OH
HO~O2,,"" ~B HO
A I c 3 3
~ 4 ""'OH
QH ~ 'OX"'O~ OH
° ~ 0,1'1
'. 4
"I
.................
OH
HO.",3 _ 0
OH
Q
OH
HO
22 - !
23 - OH
24
H
H0'6?°H
I
° "",·aO
~ 'Cl"'O~
~ I 0 ""
OH
'. OH
HO"'" _ 0
25
The [4-0-3]-linked proteracacinidin derivative 29 was synthesized accord-

ing to the sequence outlined in Figure 5. Epioritin-4a-ol tri-O-methyl ether (26)
with 2R,3R,4R absolute configuration was treated with thionyl chloride in dry
THF to form the intermediate (2R,3S,4S)-4-chloro-flavan-3-ol (27) in quantita-
tive yield with the anticipated inversion of configuration at C( 4) of precursor 26.
Addition of a two molar excess of oritin-4a-ol tri-O-methyl ether (28) and
eventual acetylation gave the all-cis flavan-3,4-diacetate (30) and the epioritin-
(4~~3)-oritin-4a-ol permethylaryl ether diacetate (29) (15.4%) with 'H NMR
and CD data identical to those of the same derivative of the natural product 22.
Separate treatment of the intermediate 4~-chloroflavan-3-ol derivative (27) with
ent-oritin-4a-ol tri-O-methyl ether (31) and epioritin-4a-ol tri-O-methyl ether

(33) in anhydrous THF, and subsequent acetylation afforded ent-oritin-(4a~4)
epioritin-4a-ol permethylaryl ether diacetate (32) (14.9%) and the epioritin-
(4/3~4)-epioritin-4a-ol permethylaryl ether diacetate (34) (23.7%), respectively.
Compounds 32 and 34 were identical to the same derivatives of the natural prod-
ucts 24 and 25, respectively by comparison of their IH NMR and CD data.
The stereoselective coupling between the 4-chloroflavan-3-01 derivative
(27) and the flavan-3,4-diol derivatives (28 and 33) to give the (C 4 -O-C 3)- and
(C 4-O-C 4 )-proteracacinidins (29 and 34), respectively, with retention of the C(4)
configuration of the electrophilic precursor (27), may be explained in terms of a
neighboring group mechanism involving intramolecular displacement of the
quasi-axial C(4) chloro nucleoruge by the axial C(3) hydroxyl group. The tran-
sient protonated epoxide (35) then permits preferential attack of the nucleophilic
C(4) hydroxyl group of the flavan-3,4-diol derivatives (28 and 33) from the less
hindered /3-face, resulting in a highly stereoselective coupling step.
We anticipated that coupling of the 4/3-chloroflavan-3-01 (27) and the ent-
oritin-4a-ol derivative (31) would also proceed via the neighboring group mech-
anism. The rather unexpected formation of the 4a-ether bond (F-ring) in 32, i.e.
with inversion of configuration at C(4) of the e1ectrophile 27, presumably reflects
reaction conditions incapable of triggering the neighboring group mechanism,
hence resulting in an SN2-type mechanism which requires the approaching hy-
droxyl nucleophile to force out the nucleofugal chloride. The requisite alignment
for such a concerted process may be facilitated by mutual hydrogen bonding of
the C(3) hydroxyl groups of both 27 and 31, which is effectively permitted by
the axial C(4) hydroxyl group of flavan-3,4-diol (31) compared to the equatorial
orientation of the same functionality in nucleophiles 28 and 33. In addition,
unfavorable 1,3-diaxial interaction between H-2ax of a putative oxirane of type
35 and the approaching nucleophile 31 would not favor a neighboring group
mechanism.
The formation of the all-cis-flavan-3,4-diol derivative, isolated as diacetate
30, in the coupling reactions of the 4/3-chloroflavan-3-01 (27) and diol derivatives
(28 and 31) is explicable in terms of solvolysis of the electrophile 27 during work-
up and chromatography. Inversion of configuration is effected by intermolecular
hydrogen bonding between the axial C(3) hydroxyl group and water, hence per-
mitting SN2 displacement of chloride ion.
Characterization of the novel stereoisomeric biflavanoids 22-25 extends
both the rare series of proteracacinidin-type oligomers as well as the hetero-
geneity of the interflavanyl bond in natural oligomeric proanthocyanidins. The
co-occurrence of these unique ether-linked compounds and the carbon-carbon
coupled analogues in Acacia galpiniil8 presumably reflects the poor nucle-
ophilicity of the pyrogallol A-ring of the monomeric flavan-3,4-diol precursors,
thus permitting alternative centers to participate in interftavanyl bond forming
processes.
OMe OMe
I I
MoO'Oo:" '"
OMe (r OMe (r
MeOty~
I ° : :". ~
4 OAc = "OAc
QAc
OAc
0:OC"
~II"" ° I ""'"
30
N
b OMe O\1e
OMe ~
Me Meo~o~II,.0
MeO i) ~
:- OH
29 OH
"
ii) Ac 20/pyridine
OMe
OMe I
(r
Me0Yy0'lII," ~
VY""OH C1
27
MeO
OM< (J
~0'-lil' ~ I
:Y
aM,
i) l,JU,. :- OH
()~\lk
M,O
OH
JJ
i)
ii) Ac 20/pyridine OH
/
31
ii) Ac 20/pyridine
(rOMe OMe
M"'~::: I
MeO
O~", ".
A
ox: O
..........
lOMe
OMe
t "cf¥"O~
2> D I
OMe
11"",4 F~ OMe +30

,iI 0
AcO" = AcO"
"I
=
0
()
34
1:e ~ OMe 32
Figure 5. Synthesis the ether-linked proteracacinidin derivatives 29, 32 and 34.

Miscellaneous
The principle of condensing electrophilic and nucleophilic flavanyl units

under mild acidic conditions to form oligomers was also implemented to syn-
thesize a range of related proanthocyanidins. Notable among these are the syn-
theses of dimers, e.g. 36 exhibiting flavan chain extender units derived from
flavan-4-0Is e.g. 37, as electrophiles/ 6.37 conventional proguibourtinidin-type
biflavanoids/~ and procyanidins 2H-labeled at C(4) (C-ring), e.g. 38, using cate-
chin-4-0Is with C(4) deuterium labels which are available via reduction of (+)-
taxifolin with sodium borotrideuteride. 39 In the latter instance, formation of the
interflavanyl bond between the perbenzylaryl ethers was catalyzed by the Lewis
acid, titanium tetrachloride in dichloromethane. The free phenolic deuterium
labeled pro cyanidin B-3 (38) was then generated by hydrogenolysis of the benzyl
protecting groups.
35
(X 0H
'(j()", : : ,. .
HO 0 " I
(X
OH
: :,. (X
0H
I OH
1
HO 0"
HOyYO,l""'::::"" OH yyv'l""::::"" OH
~
1
~OH OH
OH
" ': :,.

37
36
(X
W
0H
HO 0, I
OH
::::,...1
= OH OH
HO~OI"'(XH
~OH
OH
38
CLEAVAGE OF THE INTERFLAVANYL BOND OF

PROANTHOCYANIDINS
B-Type Proanthocyanidins
The facile cleavage of the interflavanyl bond in proanthocyanidins exhibit-

ing C(5) oxygenation of the A-ring of their chain-extender units with sulfur40.41
and oxygen nucleophiles 42 under acid catalysis has played a key role in the struc-
ture elucidation of this complex group of natural products. In the 5-deoxy (A-
ring) series of compounds, e.g. the fisetinidol-( 4~8)- and -( 4~6)-catechin
profisetinidins (12, 41, and 43), and the analogous prorobinetinidins (39 and 44)
from the commercially important bark of Acacia mearnsii (black wattle),43.4 4 this
C(Sp3)_C(Sp2) bond is remarkably stable under a variety of conditions 25 .45 and
has hitherto resisted all efforts at cleavage in a controllable manner. Such a
stable interflavanyl bond hampered both the structure investigation of the
polyflavanoid tannins in black wattle bark and those from other commercial
sources, e.g. Schinopsis spp. (quebracho), as well as the establishment of the
absolute configuration of the chain-terminating flavan-3-01 moiety in the 5-
deoxyoligoflavanoids. We, therefore, assessed conditions to efficiently cleave the
interflavanyl bond in profisetinidins under conditions sufficiently mild to allow
the isolation and identification of the constituent flavanyl units. 46
Treatment of the fisetinidol-(4a~8)-catechin (12),15 representing a typical
tannin unit of commercial wattle extract with sodium cyanoborohydride
Rl
cix°
RI
~OR'
6'~
I B H
R20 2II"::::"",
" OR2
HO 0 B I
~.::::,... OH
OR2 A I C
::::,...
=4 OH
OH 11""&OR2
OH
~OH
OR2
HO =
12 I - , RI =R2 =H
~
39 I ' R' = OH, R2=H I E
,,
# OH
40 I - , RI = H, R2 = Me
OH
41 I - , Rl =R2=H
42 l - , Rl = H, R2 = Me 43 RI =H
44 RI=OH
[Na(CN)BH3]47 in trifluoroacetic acid (TFA) for 6h at O°C gave products com-

prising the starting material 12, catechin (2) (15%), and the (2R)-1-(2,4-
dihydroxyphenyl)-3-(3,4-dihydroxyphenyl)propan-2-01 (49) (Fig. 6). Similar
treatment of the fisetinidol-(4~~8)- and -(4a~6)-catechins (41 25 and 43 25 ), with
their respective more and less labile interflavanyl bonds compared with the C(4)-
C(8) bond in compound 12 under acidic conditions,48 also afforded a mixture
consisting of starting material 41 and 43, catechin (2) (17, 4% resp.), and the
(2R)-1,3-diarylpropan-2-01 (49) (18,4% resp.).
Similar conditions also affected cleavage of the interflavanyl bond in
the fisetinidol-(4a~8)-catechin permethylaryl ether (40) to afford tetra-O-
methylcatechin (47) (21%), the 1,3-diarylpropan-2-01 (50) (12%), and tri-O-
methylfisetinidol (53) (12%). Such a rupture of the interflavanyl bond in the
permethylaryl ether (40) introduced an important dimension to these cleavages
in relation to the chemistry of the 5-deoxy oligoflavanoids where the additional
chromatographic steps involved with derivatization are often prerequisites for
sample purity. The "liberation" of the chain-terminating flavan-3-01 unit 2 or 47,
irrespective of whether the phenol 12 or methyl ether 40 was employed, provided
a powerful probe towards addressing the hitherto unsolved problem of defining
the absolute configuration at the stereocenters of this moiety in naturally occur-
ring proanthocyanidins that are synthetically inaccessible.
The mild conditions affecting simple cleavage of the strong interflavanyl
bond in the profisetinidins (12, 41, and 43) prompted application of the same pro-
tocol to the procyanidins B-1 (15) and B-3 (20) and their respective permethy-
laryl ethers 55 and 56 with less rigid C(4)-C(8) linkages compared to those in
the profisetinidins 12 and 41. Treatment of procyanidin B-1 (15) with Na(CN)BH3
in TFA for 1 h at O°C gave a mixture comprising the starting material 15, cate-
chin (2) (20%), and epicatechin (3) (21 %). Under identical conditions, procyani-
din B-3 (20) afforded catechin (2) (35%) and a residue of starting material.
The permethylaryl ethers 55 and 56 gave, within 30min, respectively tetra-O-
methylcatechin (47) (31 %), tetra-O-methylepicatechin [permethylaryl ether of
(3)] (33%), and starting material 55 and tetra-O-methylcatechin (47) (56%) and
starting material 56.
Whereas the heterocyclic ring of the catechin DEF moiety invariably
remains intact during the reductive process, cleavage of both the (4~6)- and
(4~8)-interflavanyl bonds in the free phenolic profisetinidins 12, 41, and 43
is apparently associated with the simultaneous opening ofthe C-ring ofthe chain-
extender unit. Protonation of the electron-rich phloroglucinol D_ring49 ,50 in
profisetinidin (12) (Fig. 6), and concomitant delivery of the equivalent of a
hydride ion at C(2) (C-ring) of intermediate 45 effects the concurrent rupture of
the pyran C-ring and of the C(4)-C(8) bond to give catechin (2) and the a-quinone
methide intermediate (48), which is subsequently reduced to the 1,3-
diarylpropan-2-01 (49). The selective cleavage of the interflavanyl bonds in pro-
cyanidins B-1 (15) and B-3 (20), and their permethylaryl ethers 55 and 56 pre-
RIO 0 ..
(E(IE
0Rl
~
"":::".. ORI
D I F
:::".. OH
RI
12/40 ~ 2 RI =H
OH
47 RI =Me
45 RI =OH HO~~O.
-::r OH-::r OH
46 RI =OMe
: :". :J ~ I
:::".. OH
~:J
1 48
(S(
oMe
BI 1
-::r oow
Me ", ".:::".. OMe
47 +50
A I c
:::".. OH
RI
53 Rl =H
54 Rl =D
49 RI = R2 = H
50 RI = Me, R2=H
51 RI = H, R2 = D
52 RI = Me, R2=D
Figure 6. Proposed route to the cleavage of the interflavanyl bond and of the C-ring in
profisetinidins e.g. 12 and permethylaryl ether 40.
sumably results from the relative lability of this bond, imposing a high degree of
SN 1 character to the processes of protonation and delivery of hydride ion.
The mechanism for cleavage of the interfiavanyl bond in the profisetinidin
biflavanoids (Fig. 6) was corroborated using sodium cyanotrideuterioborohydride
[Na(CN)BD31 in TFA. Under these conditions the fisetinidol-(4a~8)-catechin
(12) was converted into catechin (2) (26%) and the (2R)-1,3-dideuterio-1,3-
diarylpropan-2-ol (51) (25%), while the permethylaryl ether (40) and the
fisetinidol-( 4~~8)-catechin hepta-O-methyl ether (42) both gave tetra-O-
methylcatechin (47) (12, 32% resp.), the dideuterio-l ,3-diarylpropan-2-ol tri-O-
methyl ether (52) (14, 16% resp.), and the 4~-deuteriofisetinidol derivative (54)
(12, 14% resp.). Formation of the deuteriated 1,3-diarylpropan-2-0Is (51 and 52)
(mixtures of diastereomers) confirmed the conjecture regarding the genesis ofthe

propan-2-01s via reduction of the o-quinone methide (48).
The protonated species 45/46 presumably also served as precursor to the
4~-deuteriotri-O-methylfisetinidol (54) via delivery of hydride ion from the ~
face in a predomimmt SN2 mode. Compound 54 persistently formed also when
fisetinidol-(4o.~8)- and (4~~8)-catechin hepta-O-methyl ethers (40 and 42)
were treated with Na(CN)BD3 in TFA. This observation prompted an investiga-
tion of the structural features of the substrates that direct the stereochemistry
of the delivery of hydride ion at C(4) in intermediates of type 45/46. Whereas
treatment of the epifisetinidol-(4~~8)-catechin hepta-O-methyl ether (57) with
Na(CN)BD3 afforded the 4~-deuteriotri-O-methylepifisetinidol (59) (18.5%),
tetra-O-methylcatechin (47) (32%), and the (2S)-1 ,3-dideuterio-l ,3-diarylpropan-
2-01 (6%, enantiomer of compound 52), the ent-fisetinidol-(4~~8)-catechin
hepta-O-methyl ether (58) gave 4o.-deuteriotri-O-methyl-ent-fisetinidol (13%, the
enantiomer of compound 54), tetra-O-methylcatechin (47) (24%), 1 and the (2S)-
1,3-dideuterio-l ,3-diarylpropan-2-01 (12%, enantiomer of 52).
Thus, the formation of the 4~-deuteriofisetinidol- and epifisetinidol deriv-
atives (54 and 59) from the reduction of the profisetinidin permethylaryl ethers
(40, 42, and 57) with Na(CN)BD3 in TFA, and of the enantiomer of compound
54 during reduction of the ent-fisetinidol-(4~~8)-catechin derivative 58, indi-
cated that the deuterium ion is consistently delivered at C(4) of a protonated
species of type 45/46 from the side opposite to the 2-aryl group of the C-ring.
This presumably indicates that delivery of hydride ion occurs from a complex
between the reducing agent and the C-ring heterocyclic oxygen lone pair trans to
the 2-aryl group, such transfer being most readily facilitated in an A-conformersl
of type 60.
The potential of this development towards the structural elucidation of
the proanthocyanidin condensed tannins, especially the 5-deoxy analogues, from
important commercial sources is clear. In addition, the method facilitates the
ready definition of the absolute configuration of the chain-terminating flavan-3-
01 moiety in 5-deoxyoligoflavanoids, especially in view of the demonstration that
these units may also comprise ent-catechin and ent-epicatechin. s2 .53
A-Type Proanthocyanidins
The double interflavanyl linkage in A-type proanthocyanidins introduces a
high degree of conformational stability which culminates in high-quality and
unequivocal NMR spectra, conspicuously free of the effects of dynamic rotational
isomerism at the dimeric level. Compounds of this class are readily recognizable
from the characteristic AB-doublet CJ3 .4 = 3-4Hz) of C-ring protons in the het-
erocyclic region of their lH NMR spectra, 54 and may possess either (20.,40.)- or
(2~,4~)-double interflavanyl bonds. Two fundamental structural problems, i.e.
establishment of the mode of linkage of the C- to the D-ring, and assignment of
RIO RIO
15 RI =H 20 RI =H
55 RI =Me 56 RI =Me
~OMe
MeO
"""~OMe MeO
~OMe
"""~OMe
OH
OMe OMe
57 58
OMe
((
Meo
v ' O Q 0 I' "" ~ I OMe
I
~ ""'OH
D
59
60
the absolute configuration at the stereocentres of the F-ring, have limited progress
in this field. These and related problems have hitherto been approached via exotic
spectroscopic methods 55 - 58 which prompted us to search for a simpler and general
chemical method that is based upon the reductive cleavage of the acetal func-
tionality of A-type proanthocyanidins. The potential to address these problems by
reduction of either of the c-o acetal bonds was demonstrated 59 for the known
procyanidins A-I (61) and A-2 (62), available from the skins of mature peanuts
(Arachus hypogea),60 by using Na(CN)BH3 in TFA. The readily accessible hepta-
O-methyl ethers (63 and 64) were selected as model compounds with a view to
using the O-substituents of the D-ring as probes for anticipated much simplified
1H NMR studies.
Separate treatment of the hepta-O-methylprocyanidins A-I (63) and A-2
(64) with Na(CN)BH3 in TFA for 1.5 h at O°C (Fig. 7) gave conversion to mix-
tures comprising the starting materials, and, as anticipated from cleavage "a", the
tetrahydropyrano[2,3-f]chromene derivatives (65) (5.2%) and (66) (7%). The
envisaged B-type procyanidin biflavanoids (67 and 68) from the "b" pathway were
not obtained, but instead, the respective monomeric units, i.e. tetra-O-methyl-ent-
catechin (69) (4%) and tri-O-methylcatechin (70) (3.4%) from the A-I derivative
(63), and tetra-O-methyl-ent-catechin (69) (3%) and tri-O-methylepicatechin (71)
(l.3%) from the A-2 derivative (64) were isolated.
Both the carbon-oxygen bonds of the acetal functionality in the procyani-
din A-I (63) and A-2 (64) derivatives are, thus, susceptible to reductive cleavage
under acidic conditions. This process is presumably triggered by the random pro-
tonation of the acetal oxygens and concomitant delivery of the equivalent of
hydride ion at the antibonding (cr*) orbitals of the carbon-oxygen bonds in a pre-
dominant SN2 manner. Such a transfer of hydride ion apparently occurs from
a complex between the reducing agent and the axial C(3) (C-ring) oxygen lone
pair, the proximity of the boron-hydrogen bonds to the backside of the acetal
carbon being a prerequisite for reduction of either one of the acetal bonds. Reduc-
tion, thus, leads to "inversion" of configuration at C(2)(C) of both B-type pro-
cyanidin intermediates (67 and 68), and of the tetrahydropyrano[2,3-f]chromene
derivatives (65 and 66). The chemistry and the unequivocal structure elucida-
tion, including assessment of absolute configuration at all the stereocentres of the
latter class of compounds, are well understood, 61-63 and facilitated confirmation
of the absolute stereochemistry ofring F in the natural product derivatives 63 and
64.
Biflavanoids (67 and 68) are prone to facile cleavage of their interflavanyl
bonds via protonation of the electron-rich phloroglucinol D_ring 49 .5o and attack of
hydride ion at C(4)(C)46 to give the ent-catechin derivative (69) from the ABC-
unit and respectively, the epicatechin and catechin derivatives (70 and 71) from
the DEF-moieties. The "liberation" of the latter two chain terminating flavan-3-
01 units unambiguously defines the D-ring oxygen that is involved in the acetal
RIO
ORI 63,641 ..
cleavage a
65 ~ =~
661 = ~
61
62 I:=
63 ~
,RI=H
, Rl=H
,Rl=Me
64
!= ,RI=Me
MeO
67 ~ = I
68 ~=I
MeO
OMe
70 ~ =~
69 71 ~=l
Figure 7. Cleavage of the acetal functionality of proanthocyanidin A-I and A-2 permethylaryl
ethers 63 and 64 with Na(CN)BH3 in TFA.
functionality of the parent compounds 63 and 64. It furthermore provides a pow-

erful probe towards addressing the hitherto unsolved problem of establishing the
absolute configuration at the stereocenters of this moiety in naturally occurring
A-type proanthocyanidins. The ftavan-3-01 unit (69), albeit with inversed C(2)
configuration, should facilitate the assignment of the absolute configuration at
C(3) (C-ring) of the parent compounds 63 and 64, especially in view of the inabil-
ity to differentiate between 3,4-cis- and 3,4-trans-configuration in these com-
pounds on the basis of 3JHH values. 55 The mode of the C-C linkage between the
constituent ftavan-3-01 units in the A-type procyanidin, e.g. (4~6) or (4 ~ 8)
is defined by the nature of tetrahydropyranochromene,61 i.e. [2,31], [3,2-g] or
[2,3-h], that is formed via reductive cleavage "a".
The protocol described here should thus contribute substantially towards
a straight forward chemically orientated structural definition of the A-class
proanthocyanidins.
ENANTIOSELECTIVE SYNTHESIS OF FLAVONOIDS
Dihydroflavonols
Owing to the ease of the reductive transformation, dihydroftavonol ~
ftavan-3,4-diol,20.64 the dihydroftavonols (four diastereomers for each hydroxyla-
tion pattern) are key compounds as precursors to the electrophilic ftavanyl chain
extender units in the semisynthetic approach to proanthocyanidin oligomers. Only
a limited number of dihydroftavonols with 2,3-trans configuration are available
commercially or from natural sources. Analogues possessing 2,3-cis configura-
tion are exceptionally rare and definitely not available for preparative applica-
tions, which clearly demonstrates the need for a synthetic protocol giving access
to the full range of dihydroftavonol diastereomers with phenolic oxygenation
patterns approximating those of the natural products.
Virtually all the synthetic efforts to synthesize enantiomerically enriched
dihydroftavonols had hitherto focussed on the Julia asymmetric epoxidation of
chalcones 65 -67 and the subsequent transformation of the chalcone epoxides into
dihydroftavonols. The literature covering these developments up to 1990 was
recently comprehensively reviewed. 68 Our own efforts in this regard focussed
mainly on chalcones exhibiting the hydroxylation patterns of naturally occurring
dihydroftavonols,69-71 in contrast to other approaches selecting chalcones with the
minimum number of, or which are devoid of phenolic oxygenation. 72- 76
Epoxidation of the chalcone methyl ethers (72-76) with hydrogen perox-
ide in the triphase system, aq. NaOH/poly-L or D-alanine/CCI4,65-67 gave the
(-)-trans-(77a-81a) (aR,/3S)70 and (+)-trans-epoxides (77b-81b) (as,/3R),70
respectively, in high yields (79-99%) and fair enantiomeric excess (50-85%)
(Fig. 8).77.78 Initial attempts towards cyclization of the epoxides to the corre-
- (i)
~~" 0 ",H ""<::::: R5
RIAJl:3 I
# R4
72 Rl = R3 = R5 = H, R' = MOM, R4 = OMe 77 Rl = R3 = R5 = H, R2 = MOM, R4 = OMe
73 Rl =R4 = OMe, R' = MOM, R3 =R5 = H 78 Rl = R4 = OMe, R2 = MOM, R3 = R5 = H
74 Rl =R4=R5 = OMe, R2 = MOM, R3 =H 79 Rl =R4 = R5 = OMe, R2 = MOM, Rl =H
75 Rl = R3 = R4 = OMe, R' = MOM, R5 = H 80 Rl =R3 =R4 = OMe, R2 = MOM, R5 = H
76 Rl = R' = R4 = R5 = OMe, R' = MOM 81 Rl = R3 =R4 = R5 = OMe, R2 = MOM
(XI
W
Rl 0 R4
I I""'~ R5
~ OH
R3 0
87 Rl = R3 =R5 =H, R4=OMe
-
88 Rl =R4= OMe,R3 =R5 =H
89 Rl = R4 = R5 = OMe, R3 = H (iii)
90 Rl = R3 = R4 = OMe, R5 = H +
91 Rl = R3 = R4 = R5 = OMe
,,(XI
W
Rl 0 R4
82 Rl =R3 =R5 =H, R4=OMe
I ""~ R5 83 Rl = R4 = OMe, R3 = R5 = H
84 Rl = R4 = R5 = OMe, R3 = H
~ OH 85 Rl = R3 = R4 = OMe, R5 = H
R3 0 86 Rl = RJ = R4 = R5 = OMe
92 Rl=RJ=R5=H,R4=OMe
93 Rl = R4 = OMe, R3 = R5 = H
94 Rl = R4 = R5 = OMe, R3 = H
95 Rl = R3 = R4 = OMe, R5 = H 77-96a = configuration shown
96 Rl = Rl = R4 = R5 = OMe b = enantiomer
MOM=CH20Me
Figure 8. Synthesis of dihydroftavonol derivatives. Reagents: i, 30% H,O,: 6M NaOH 1 : 0.32

(v/v), poly-L- or poly-D-alanine: chalcone 1: 1 (m/m), CCl.; ii, BnSH (4 equiv.), SnCl. (0.2
equiv.); iii, AgBF. (5 equiv.).
sponding (2R,3R)-2,3-trans-dihydroflavonols were hampered by low yields due

to the formation of isoflavones via aroyl migration, and by loss of optical purity
in the dihydroflavonols due to acid-catalyzed racemization. 70 In order to circum-
vent these problems, methods aimed at the initial opening of the oxirane func-
tionality followed by deprotection and cyclization were investigated. Treatment
of the series of epoxides (77a/b-81a/b) with BnSH/SnCI4 selectively cleaved
the C(~)-O bond of the oxirane functionality at -20°C and effectively depro-
tected the acetal function at O°C to give the corresponding a.,2'-dihydroxy-~
benzylsulfanyldihydrochalcones (82a/b-86a/b) as diastereomeric mixtures
(anti: syn ca. 2.3: 1) in 86-93% yield. These were then treated with the thiophilic
Lewis acid, silver tetrafluoroborate (AgBF4)' in CH 2 Cl 2 at O°C, to give the 2,3-
eJ
trans-dihydroflavonols (87a/b-91a/b) 2,3 ca. 12,0 Hz) in good yields (48-80%)
(47-84% ee) and, albeit in low proportions (5-15%), for the first time also, the
2,3-cis analogues (92a/b-96a/b) 02,3 ca. 6.0 Hz) (44-84% ee).
Although the Julia asymmetric epoxidation of chalcones has proven to be
a reliable reaction affording epoxides in good yields and moderate to high enan-
tiomeric excess, this protocol is hampered by long reaction times (up to 3 days),
reactions requiring continuous addition of oxidant and base, while degradation of
the polyamino acid often poses difficulties. We, thus, assessed the potential of
the adapted version recently developed by Roberts and his co-workers 76 which
involves a two-phase non-aqueous system consisting of oxidant, a non-
nucleophilic base, immobilized poly(amino acid)/9 and an organic solvent.
Thus, treatment of chalcones (72-74) with immobilized poly-(L)-
leucine (PLC),76 urea-hydrogen peroxide complex (UHP),80 and 1,8-
diazabicyclo[5.4.0]undec-7-ene (DBU) in dry THF, afforded the (-)-(a.R,~S)
trans-epoxychalcones (77a-79a) in high yields (64-80%) and improved optical
purity (85-95%)? The enantiomeric (+)-(a.S,~R)-trans-epoxides (77b-79b) were
similarly obtained by using immobilized poly-(D)-Ieucine (PDL) in the same two-
phase system (61-76% yield, 81-90% enantiomeric excess). We used these epox-
ides with superior enantiomeric excess to synthesize a series of both enantiomers
of polyoxygenated ~-hydroxydihydrochalcones.81 The significance of the higher
enantiomeric excess of the epoxides in relation to dihydroflavonol synthesis is
evident.
Flavan-3-ols
The flavan-3-0Is are the most common chain terminating flavanyl units in
naturally occurring proanthocyanidin oligomers.1.3 All four possible diastere-
omers of catechin (3,5,7,3',4'-pentahydroxylation) were encountered, hence
stressing the need for synthetic access, especially in view of the fact that only
(+)-catechin and (-)-epicatechin are readily available among this largest group of
naturally occurring C6 .C 3 .C 6 -metabolites. In order to address the issue of stereo-
control at C(2) and C(3) of the flavan-3-01 molecular framework, we designed a
concise protocol that is based upon the transformation of I ,3-diarylpropenes into

1,3-diarylpropane-l ,2-diols, via asymmetric dihydroxylation, which are then used
as chirons for essentially enantiopure flavan-3_0Is. 82 •83
Owing to the excellent results reported by Sharpless and his co-workers 84- 87
during asymmetric dihydroxylation of olefins with AD-mix-a and AD-mix-~,
these stereoselective catalysts were utilized for introduction of chirality at C(3)
in the flavan-3-01 framework. Treatment of the protected (£)-propenes (97-101)
at ODC with AD-mix-a in the two-phase system Bu'OH: H20 (1 : 1) afforded the
eJ
(+)-(lS,2S)-syn-diols (102a-l06a) 1•2 5.8-6.5 Hz) in high yields (80-86%) and
optical purity (99% ee). The (-)-(lR,2R)-syn-diols (I 02b-1 06b) were similarly
obtained by using AD-mix-~ (82-87% yield, 99% ee) (Fig. 9). The absolute
configuration of the diols was tentatively assigned according to the Sharpless
mode1 84•85 for AD-mix (Fig. 10).
The syn-diols were susceptible to simultaneous deprotection and cycliza-
tion by treatment with 3 M HCI in methanol. Application of these conditions to
diols (102a/b-I06a/b), followed by acetylation, afforded the 2,3-trans-(48-68%)
(107a/b-Illa/b) and, for the first time, 2,3-cis-flavan-3-01 methyl ether acetate
derivatives (17-22%) (112a/b-116a/b) in excellent enantiomeric excesses.
Confirmation of the absolute configuration of the flavan-3-01 derivatives was
obtained by comparison of CD data with those of authentic compounds.
The potential of this protocol in the chemistry of oligomeric proantho-
cyanidins is evident, especially in view of its aptitude to the synthesis of free phe-
nolic analogues. The latter compounds are as conveniently accessible by simply
replacing the stable O-methyl ethers by labile protecting groups.
CONFORMATIONAL ANALYSIS OF DIMERIC

PROANTHOCYANIDINS
Conformational analysis of proanthocyanidin oligomers is, in principle,

concerned with the conformation of the pyran heterocycles and with the
phenomenon of conformational isomerism due to restricted rotation about
the interflavanyl bond(s). Realization of the fact that the conformational itinerary
of the heterocyclic rings involves a dynamic equilibrium between E- and A-
conformers 51 had a huge impact in this field. 2 It is generally being accepted that
an understanding of the biological significance of the polymeric proanthocyani-
dins relies on insight into their complexation with other biopolymers. This has
led to substantial efforts to improve comprehension of the interaction of
polyflavanoids with proteins.
Much ofthis effort has been focussed on developing a more detailed under-
standing of the conformational preferences and flexibility of the proanthocyani-
din polymers and their interaction with polypeptides. The conformational
97 Rl =R2=R4=H,R3=OMe 102a,b Rl = R2 = R4 = H, R3 = OMe

98 R2=R4=H,Rl =R3=OMe 103a,b R2 = R4 = H, Rl = R3 = OMe
99 R2=H,Rl =R3 =R4=OMe 104a,b R2 = H, Rl = R3 = R4 = OMe
100 R4 = H, Rl = R2 = R3 = OMe 105a,b R4 = H, Rl = R2 = R3 = OMe
101 Rl = R2 = R3 = R4 = OMe 106a,b Rl = R2 = R3 = R4 = OMe
112a,b Rl = R2 = R4 = H, R3 = OMe 107a,b Rl = R2 = R4 = H, R3 = OMe

113a,b R2 = R4 = H, Rl = R3 = OMe 108a,b R2 = R4 = H, Rl = R3 = OMe
114a,b R2 = H, Rl = R3 = R4 = OMe 109a,b R2 = H, Rl = R3 = R4 = OMe
115a,b R4 = H, Rl = R2 = R3 = OMe 110a,b R4 = H, Rl = R2 = R3 = OMe
116a,b Rl=R2=R3=R4=OMe 11la,b Rl = R2 = R3 = R4 = OMe
102-116a = configuration shown

b = enantiorrer
Figure 9. Synthesis of flavan-3-ol derivatives. Reagents: i, AD-mix-a or AD-mix-~,

Bu'OH: H20 (l : l, v/v), MeS02NH2; ii, 3M HCl, MeOH-H,O (3 : 1, v/v), then AC20/Pyridine.
properties of the polyflavanoids have, thus, been studied by using a variety of

molecular mechanics and molecular orbital computations in combination with
crystal structures, time-resolved fluorescence, as well as IH and l3C NMR
methods. Representative references to these techniques may be found in the
papers listed in references 88-94, which in themselves are arguably the most
authoritative reports recently published on this important branch of the chemistry
of the proanthocyanidins. These results are summarized by using the significant
recent contributions of Hatano and Hemingway.92.94
l3-attack
OMOM
(IR,2R)-102b-I06b
i OMOM
a-attack
97-101
RI,R2,R3,R4=H,OMe
(1 S,2S)-102a-I06a
Figure 10. Sharpless model of asymmetric dihydroxylation.
NMR analysis of procyanidin B-1 (15) and B-3 (20) permitted full assign-
ment of the proton and carbon resonances for both the more extended 117 and
compact 118 conformers in the free phenolic form. In organic solvents, the more
extended rotamer 117 of procyanidin B-1 (15) is preferred over the more compact
rotamer (10: 7), but in water, the more compact rotamer dominates (10: 2). When
procyanidin B-3 (20) is dissolved in organic solvents, the more compact rotamer
is slightly preferred (8 : 10). With water as solvent, only trace proportions of the
more extender rotamer are detected. In this solvent, rotational conformation
exchange is detected despite the observation of two distinct and sharp sets of
signals for each rotamer. The heterocyclic ring of the ABC unit exists in an
approximate half-chair conformation in each rotamer for both procyanidin 8-1
(15) and B-3 (20). Coupling constants of the heterocyclic ring of the DEF moiety
in both 15 and 20 indicate substantial axial orientation of the E-ring (see 119 and
120 for E- and A-conformers of the DEF unit of 20). Lineshape analysis of 3-
H(F) indicated that the "abnormal" coupling constants of the F-ring were indica-
tive of a comparatively high-energy skewed-boat conformation for 15 and
between a half-chair and a skewed-boat conformation for 20 rather than to E- ~
A-conformational exchange, which has hitherto been used to explain the smaller
than anticipated coupling constants.
OH
~
BI
O 1"-
HOW ~ OH
A 1 c 3
~ - OH g
9'O H
OH ~ E 1
HO~'8
9 0 1"- ~ OH
DI F
~ OH
OH
OH
118 (compact rotamer of 20) 117 (extended rotamer of 20)

C3-C4-08-09 torsion angle, (-) CJ-C4-08-D9 torsion angle, (+)
OH
OH
Fl
~
+OH
OH
119 : E-conformer 120 : A-conformer
Hatano and Hemingway92 used NOE studies to assess the association of

(+)-catechin (2) and procyanidin B-3 (20) with oligopeptides. These efforts,
focussing on the complexation of (poly)flavanoids with peptides containing
proline residues in aqueous solutions, revealed site specific approach directed by
hydrophobic interaction of the aromatic rings of (+)-catechin (2) and procyani-
din B-3 (20), to conformationally accessible regions of peptides without strong
preference for interaction with proline residues. The observed intermolecular
NOE's indicating the preferred sites in the association of (+)-catechin (2) and pro-
cyanidin B-3 (20) with the tetra-peptide, Gly-Pro-Gly-Gly are shown in 121 and
122, respectively.
OH
121
e('7
NH3 ,N~
H l8: 'f-H
H+OC CO-HN H
OH
H.............
B I CO
I 0
CU2
HO
o ,1"" ~ OH N--\-H
H J H
OH
~OH
HO
"I"'~OH
OH
122
CONCLUSION
This review clearly demonstrates that considerable progress has been made
to gain insight into the complex factors that govern the chemistry of the proan-
thocyanidin oligomers. The section on the stereoselective synthesis of
dihydroftavonols and ftavan-3-0Is introduces a fresh breeze into this much
neglected area of proanthocyanidin chemistry, indicating that introduction of chi-
rality at C(2) and C(3) is indeed possible via utilization of the myriad of modern
synthetic methods that are available to researchers in this field. It may be antici-
pated that the rapid advances that have been made in conformational analysis of
these compounds will continue and contribute towards understanding of the intri-
cate principles governing the complexation of proanthocyanidins with other bio-
molecules. These advances may eventually also contribute towards a better under-
standing of the claimed health-promoting properties of this important group of
natural occurring polyphenols.
ACKNOWLEDGMENTS
We are grateful for the enthusiastic support of our co-workers, E.V Brandt,
B.C.B. Bezuidenhoudt, P.S. van Heerden, R.J.J. Nel, and P.l Steynberg. Financial
support by the Foundation for Research Development, Pretoria and the "Sentrale
Navorsingsfonds" of this University, is gratefully acknowledged.
REFERENCES
1. PORTER, L.1. 1994. Flavans and proanthocyanidins. pp. 23-55 in The Flavonoids.
Advances in Research Since 1986 (IB. Harborne, ed.), Chapman and Hall, London.
2. FERREIRA, D., BEKKER, R. 1996. Oligomeric proanthocyanidins: Naturally occurring
O-heterocycles. Nat. Prod. Rep. 13: 411-433.
3. HEMINGWAY, R.W 1989. Bifiavonoids and proanthocyanidins. pp. 571-651 in Natural
Products of Woody Plants I (lW Rowe ed.), Springer-Verlag, Berlin.
4. SALAH, N., MILLER, N.J., PARANGA, G., TI1BURG, L., BOLWELL, G.P.,
RICEEVANS, e. 1995. Polyphenolic fiavanols as scavengers of aqueous-phase radicals
and as chain-breaking antioxidants. Arch. Biochem. Biophys. 322: 339-346.
5. KANNER, 1, FRANKEL, E.N., GRANIT, R., GERMAN, B., KINSELLA, IE. 1994.
Natural antioxidants in grapes and wine. 1. Agric. Food Chern. 42: 64-69.
6. FUHRMAN, B., LAVY, A., AVIRAM, M. 1995. Consumption of red wine with
meals reduces the susceptibility of human plasma and low-density lipoprotein to lipid-
peroxidation. Am. I Clin. Nutr. 61: 549-554.
7. FRANKEL, E.N., WATERHOUSE, A.L., TEISSEDRE, P.L. 1995. Principal phenolic
phytochemicals in selected Californian wines and their antioxidant activity in inhibiting
oxidation of low-density lipoproteins. I Agric. Food Chern. 43: 890-894.
8. FRANKEL, E.N., KANNER, 1., GERMAN, lB., PARKS, E., KINSELLA, IE. 1993.
Inhibition of oxidation of human low-density lipoprotein by phenolic substances in red
wine. Lancet 341: 454-457.
9. RUF, Ie., BERGER, 1.L., RENAUD, S. 1995. Platelet rebound effect of alcohol-
withdrawal and wine drinking in rats-relation to tannins and Iipid-peroxidation. Arterio-
scler. Thromb. Vasco BioI. 15: 140-144.
10. ROUX, D.G., FERREIRA, D. 1982. Structure and function in the biomimetic synthesis
of linear, angular and branched condensed tannins. Pure and Appl. Chern. 54: 2465-
2478.
11. ROUX, D.G., FERREIRA, D. 1982. The direct biomimetic synthesis, structure and
absolute configuration of angular and linear condensed tannins. Fortschr. Chern. Org.
Naturst. 41: 47-76.
12. FERREIRA, 0., STEYNBERG, J.P., BURGER, IF.W, BEZUIDENHOUDT, B.e.B.
1992. Synthesis and base-catalyzed transformations of proanthocyanidins. pp. 255-295
in Recent Advances in Phytochemistry (H.A. Stafford and R.K. Ibrahim, eds.) Plenum
Press, New York.
13. FERREIRA, D., STEYNBERG, J.P., ROUX, D.G., BRANDT, E.Y. 1992. Diversity of
structure and function in oligomeric fiavanoids. Tetrahedron 48: 1743-1803.
14. COETZEE, I, STEYNBERG, IP., BRANDT, E.Y., FERREIRA, D. 1995. Oligomeric
fiavanoids. Part 18. Dimeric prorobinetinidins from Robinia pseudacacia. Tetrahedron
51: 2339-2352.
15. STEYNBERG, P.I, STEYNBERG, IP., BRANDT, E.Y., FERREIRA, D., HEMINGWAY,
R.W 1997. Oligomeric fiavanoids. Part 26. Structure and synthesis of the first profisetini-
dins with epifisetinidol constituent units. J. Chern. Soc., Perkin Trans. 1: 1943-1950.
16. MALAN, J.C.S., YOUNG, D.A., STEYNBERG, IP., FERREIRA, D. 1990. Oligomeric
fiavanoids. Part 10. Structure and synthesis of the first tetrahydropyrano[2,3-g]chromenes
related to (4,6)-bis-(-)-fisetinidol profisetinidins. 1. Chern. Soc., Perkin Trans. 1: 227-234.
17. FOO, L.Y. 1986. A novel pyrogallol A-ring proanthocyanidin dimer from Acacia
melanoxylon. I Chern. Soc., Chern. Commun.: 236-237.
18. MALAN, E., SIREEPARSAD, A. 1995. The structure and synthesis of the first dimeric
proteracacinidins from Acacia galpinii. Phytochemistry 38: 237-239.
19. MALAN, E., SIREEPARSAD, A., BURGER, lEW, FERREIRA, D. 1994. A novel
doubly-linked proteracacinidin analogue from Acacia cafJra. Tetrahedron Lett. 35:
7415-7416.
20. BOTHA, J.J., FERREIRA, D., ROUX, D.G. 1981. Synthesis of condensed tannins. Part
4. A direct biomimetic approach to [4,6]- and [4,8]-bifiavanoids. J. Chern. Soc., Perkin
Trans. 1: 1235-1245.
21. BOTHA, J.J., FERREIRA, D., ROUX, D.G. 1981. Synthesis of condensed tannins. Part
1. Stereoselective and stereospecific synthesis of optically pure 4-arylfiavan-3-ols, and
assessment of their absolute stereochemistry at C-4 by means of circular dichroism.
1. Chern. Soc., Perkin Trans. 1: 1213-1219.
22. VAN DER WESTHUIZEN, 1.H., FERREIRA, D., ROUX, D.G. 1981. Synthesis of
condensed tannins. Part 2. Synthesis by photolytic rearrangement, stereochemistry, and
circular dichroism of the first 2,3-cis-3,4-cis-4-arylfiavan-3-ols. 1. Chern. Soc., Perkin
Trans. I: 1220-1226.
23. BARRETT, M.W, KLYNE, W, SCOPES, P.M., FLETCHER, A.C., PORTER, L.1.,
HASLAM, E. 1979. Plant proanthocyanidins. Part 6. Chiroptical studies. Part 95.
Circular dichroism ofprocyanidins. J. Chern. Soc., Perkin Trans. I: 2375-2377.
24. DE ANGELIS, G.G., WILDMAN, We. 1979. Circular dichroism studies-I. A quad-
rant rule for the optically active aromatic chromophore in rigid polycyclic systems.
Tetrahedron 25: 5099-5112.
25. YOUNG, D.A., CRONJE, A., BOTES, A.L., FERREIRA, D., ROUX, D.G. 1985.
Synthesis of condensed tannins. Part 14. Bifiavanoid profisetinidins as synthons. The
acid-induced "phlobaphene" reaction. I Chern. Soc., Perkin Trans. I: 2521-2527.
26. DELCOUR, lA., FERREIRA, D., ROUX, D.G. 1983. Synthesis of condensed tannins.
Part 9. The condensation sequence ofleucocyanidin with (+)-catechin and with the resul-
tant procyanidins. 1 Chern. Soc., Perkin Trans. 1: 1711-1717.
27. TROST, B.M., MURAYAMA, E. 1981. Dimethyl(methylthio)sulfonium fluoroborate.
A chemoselective initiator for thionium induced cyclizations. 1 Am. Chern. Soc. 103:
6529-6530.
28. TROST, B.M., SATO, T. 1985. Dimethyl(methylthio)sulfonium tetrafiuoroborate initiated
organometallic additions to and macrocyclizations ofthioketals. 1. Am. Chern. Soc. 107:
719-721.
29. BARRETT, A.G.M., BEZUIDENHOUDT, B.C.B., HOWELL, A.R., LEE, A.C.,
RUSSEL, M.A. 1989. Redox glycosidation via thionoester intermediates. J. Org. Chern.
54: 2275-2277.
30. STEYNBERG, P.1., NEL, R.J.J., VAN RENSBURG, H., BEZUIDENHOUDT, B.C.B.,
FERREIRA, D. 1998. Oligomeric flavanoids. Part 27. Interflavanyl bond formation in

procyanidins under neutral conditions. Tetrahedron 54: 8153-8158.
31. DREWES, S.E., ILSLEY, AH. 1969. Dioxan-linked biflavanoid from the heartwood of
Acacia mearnsii. J. Chern. Soc. (C): 897-900.
32. YOUNG, DA, FERREIRA, D., ROUX, D.G. 1983. Synthesis of condensed tannins. Part
10. "Dioxan-linked profisetinidins". J. Chern. Soc., Perkin Trans. I: 2031-2035.
33. FOO, L.Y. 1989. Isolation of[4-0-4]linked biflavanoids from Acacia melanoxylon: First
examples of a new class of single ether-linked proanthocyanidin dimers. J. Chern. Soc.,
Chern. Commun.: 1505-1506.
34. COETZEE, J., MALAN, E., FERREIRA, D. 1998. Oligomeric flavanoids. Part 28. Struc-
ture and synthesis of ether linked [4-0-3]bis-teracacinidins, a novel class of naturally
occurring proanthocyanidins. J. Chern. Res., in press, paper 8/01913D.
35. COETZEE, J., MALAN, E., FERREIRA, D. 1998. Oligomeric flavanoids. Part 29. Struc-
ture and synthesis of novel ether-linked [4-0-4]bis-teracacinidins. Tetrahedron 54:
9153-9160.
36. MALAN, E., SIREEPARSAD, A., SWINNY, A, FERREIRA, D. 1997. The structure and
synthesis of a 7,8,4' -trihydroxyflavan-epioritin dimer from Acacia caffra. Phytochemistry
44: 529-531.
37. HATANO, T., YAMASHITA, A, HASHIMOTO, T., ITO, H., KUBO, N., YOSHIYAMA,
M., SHIMURA, S., ITOH, Y., OKUDA, T., YOSHIDA, T. 1997. Flavan dimers with lipase
inhibitory activity from Cassia nomane. Phytochemistry 46: 893-900.
38. MALAN, E., SWINNY, E., FERREIRA, D., STEYNBERG, P. 1996. The structure and
synthesis of proguibourtinidins from Cassia abbreviata. Phytochemistry 41: 1209-1213.
39. PIERRE, M-C, CheZE, C, VERCAUTEREN, J. 1997. Deuterium labeled procyanidin
synthesis. Tetrahedron Lett. 38: 5639-5642.
40. BETTS, M.J., BROWN, B.R., BROWN, P.E., PIKE, w.T. 1967. Degradation of condensed
tannins: Structure of the tannin from Common Heather. Chern. Commun.: 1110-1112.
41. THOMPSON, R.S., JACQUES, D., HASLAM, E., TANNER, R.J.N. 1972. Plant Proan-
thocyanidins. Part I. Introduction; the isolation, structure, and distribution in nature of
plant procyanidins. J. Chern. Soc., Perkin Trans. I: 1387-1399.
42. FOO, L. Y., PORTER, L.J. 1978. Prodelphinidin polymers: Definition of structural units.
J. Chern. Soc., Perkin Trans. I: 1186-1190.
43. DREWES, S.E., ROUX, D.G., SAAYMAN, H.M., FEENEY, J., EGGERS, S.H. 1966.
The stereochemistry of biflavanoids from black wattle bark: leucofisetinidin-( +)-catechin,
leucorobinetinidin-(+)-catechin and leucorobinetinidin-(+)-gallocatechin. Chern.
Commun.: 370-371.
44. DREWES, S.E., ROUX, D.G., SAAYMAN, H.M., EGGERS, S.H., FEENEY, J. 1967.
Some stereochemically identical biflavanols from the bark tannins of Acacia mearnsii.
J. Chern. Soc. (C): 1302-1308.
45. ROUX, D.G., PAULUS, E. 1962. Polymeric leuco-fisetinidin tannins from the heartwood
of Acacia mearnsii. Biochem. J. 38: 320-324.
46. STEYNBERG, P.J., STEYNBERG, J.P., BEZUIDENHOUDT, B.CB., FERREIRA, D.
1995. Oligomeric flavanoids. Part 19. Reductive cleavage of the interflavanyl bond in
proanthocyanidins. J. Chern. Soc., Perkin Trans. I: 3005-3012.
47. LANE, CF. 1975. Sodium cyanoborohydride, a highly selective reducing agent for
organic functional groups. Synthesis (3): 135-146.
48. McGRAW, G.w., HEMINGWAY, R.W. 1982. Electrophile aromatic substitution ofcate-
chin. Bromination and benzylation. J. Chern. Soc., Perkin Trans. I: 973-978.
49. BROWN, A.G., EYTON, W.B., HOLMES, A., OLLIS, W.O. 1969. Identification of the
thearubigenins as polymeric proanthocyandins. Nature 221: 742-744.
50. BEART, J.E., LILLEY, T.H., HASLAM, E. 1985. Polyphenol interactions. Part 2. Cova-
lent binding of procyanidins to proteins during acid-catalyzed decomposition; observa-

tions on some polym(;ric proanthocyanidins. 1 Chern. Soc., Perkin Trans. 2: 1439-1443.
51. PORTER, LJ., WONG, R.Y., BENSON, M., CHAN, B.G., VISHWANADHAN,
YN., GANDOUR, R.D., MATTICE, WL. 1986. Conformational analysis offlavans. 'H
NMR and molecular mechanical (MM2) studies of the benzopyran ring of 3',4',5,7-
tetrahydroxyflavan-3-0Is: the crystal and molecular structure of the procyanidin:
(2R,3S,4R)-3',4' ,5, 7-tetramethoxy-4-(2,4,6-trimethoxyphenyl)-flavan-3-01. 1 Chern. Res.
M: 830-850; S: 86-87.
52. STEYNBERG, P.l, BURGER, lEW, BEZUIDENHOUDT, B.C.B., STEYNBERG, lP.,
VAN DYK, M.S., FERREIRA, D. 1990. The first natural condensed tannins with (-)-
catechin "terminal" units. Tetrahedron Lett. 31: 2059-2062.
53. DELLE MONACHE, E, FERRARI, E, POCE-TUCCI, A., MARINI-BETTOLLO, G.B.
1972. Catechins with (+)-epi-configuration in nature. Phytochemistry II: 2333-2335.
54. JACQUES, D., HASLAM, E., BEDFORD, G.R., GREATBANKS, D. 1974. Plant proan-
thocyanidins. Part II. Proanthocyanidin-A2 and its derivatives. 1 Chern. Soc., Perkin
Trans. I: 2663-2671.
55. CRONJE, A., BURGER, lEW, BRANDT, E.Y, KOLODZIEJ, H., FERREIRA, D. 1990.
Assessment of 3,4-trans and 3,4-cis relative configuration in the A-series of (4,8)-linked
proanthocyanidins. Tetrahedron Lett. 31: 3789-3792.
56. KOLODZIEJ, H., SAKAR, M.l, BURGER, lEW, ENGELSHOWE, R., FERREIRA, D.
1991. A-Type proanthocyanidins from Prunus spinosa. Phytochemistry 30: 2042-2047.
57. GONZALEZ, A.G., IRIZAR, A.C., RAVELO, A.G., FERNANDEZ, M.E 1992. Type-A
proanthocyanidins from Prunus spinosa. Phytochemistry 31: 1432-1434.
58. BALDE, A.M., DE BRUYNE, T., PIETERS, 1., KOLODZIEJ, H., VAN DEN BERG HE,
D., CLAEYS, M., VLIETINCK, A. 1995. Tetrameric proanthocyanidins containing a
double interflavanoid (A-type) linkage from Pavetta owariensis. Phytochemistry 40:
933-938, and references cited.
59. STEYNBERG, PJ., CRONJE, A., STEYNBERG, lP., BEZUIDENHOUDT, B.C.B.,
BRANDT, E.Y, FERREIRA, D. 1997. Oligomeric flavanoids. Part 25. Cleavage of the
acetal functionality in A-type proanthocyanidins. Tetrahedron 53: 2591-2598.
60. KARCHESY, J.J., HEMINGWAY, R.W 1986. Condensed tannins: (4~--78; 2~--70--77)
linked procyanidins in Arachis hypogea L. 1 Agric. Food Chern. 34: 966--970.
61. STEYNBERG, lP., BURGER, lEW, YOUNG, D.A., BRANDT, E.Y, STEENKAMP,
lA., FERREIRA, D. 1988. Oligomeric flavanoids. Part 3. Structure and synthesis ofphlo-
batannins related to (-)-fisetinidol-( 4u,6)- and (4u,8)-( + )-catechin profisetinidins. 1
Chern. Soc., Perkin Trans. I: 3323-3329. Oligomeric flavanoids. Part 4. Base-catalyzed
conversions of (-)-fisetinidol-( +)-catechin profisetinidins with 2,3-trans-3,4-cis-flavan-3-
01 constituent units. 1 Chern. Soc., Perkin Trans. I: 3331-3338.
62. STEYNBERG, lP., BURGER, lEW, YOUNG, D.A., BRANDT, E.Y, FERREIRA, D.
1988. Oligomeric flavanoids. Part 6. Evidence supporting the inversion of absolute
configuration at 3-C associated with base-catalyzed A-/B-ring interchange of precursors
having 2,3-trans-3,4-cis-flavan-3-01 constituent units. Heterocycles 28: 923-935.
63. STEYNBERG, lP., BEZUIDENHOUDT, B.C.B., BURGER, lEW, YOUNG, D.A.,
FERREIRA, D. 1990. Oligomeric flavanoids. Part 7. Novel base-catalyzed pyran
rearrangements of procyanidins. 1 Chern. Soc., Perkin Trans. I: 203-208.
64. PORTER, LJ., FOO, L.y. 1982. Leucocyanidin: Synthesis and properties of (2R,3S,4R)-
(+)-3,4,5,7,3',4'-hexahydroxyflavan. Phytochemistry 21: 2947-2952.
65. JULIA, S., MASANA, 1, VEGA, lC. 1980. "Synthetic enzymes". Highly stereoselec-
tive epoxidation of chalcone in a triphasic toluene-water-poly-(S)-alanine system. Angew.
Chern. Int. Ed. Eng!. 19: 929-931.
66. JULIA, S., GUIXER, 1, MASANA, 1, ROCAS, 1, COLONNA, S., ANNUZIATA, R.,
MOLINARI, H. 1982. Synthetic enzymes. Part 2. Catalytic asymmetric epoxidation by

means of poly amino acids in a triphase system. 1. Chern. Soc., Perkin Trans. I: 1317-1324.
67. BANFI, S., COLONNA, S., MOLINARI, H., JULIA, S., GUIXER, 1. 1984. Asymmet-
ric epoxidation of electron-poor olefins. Part 5. Influence on stereoselectivity of the struc-
ture of polyaminoacids used as catalysts. Tetrahedron 40: 5207-5211, and references cited
therein.
68. BEZUIDENHOUDT, s.c.s., FERREIRA, D. 1992. Enantioselective synthesis of
flavonoids. pp. 143-165 in: Plant Polyphenols: Synthesis, Properties, Significance. R.W.
Hemingway, P.E. Laks, and S.H. Branham, (eds.), Plenum Press, New York.
69. BEZUIDENHOUDT, S.C.B., SWANEPOEL, A., AUGUSTYN, J.A.N., FERREIRA, D.
1987. The first enantioselective synthesis of polyoxygenated o:-hydroxydihydrochalcones
and circular dichroic assessment of their absolute configuration. Tetrahedron Lett., 28:
4857-4860.
70. AUGUSTYN, J.A.N., BEZUIDENHOUDT, s.c.s., FERREIRA, D. 1990. Enantioselec-
tive synthesis of flavonoids. Part I. Polyoxygenated chalcone epoxides. Tetrahedron 46:
2651-2660.
71. AUGUSTYN, J.A.N., BEZUIDENHOUDT, B.C.S., SWANEPOEL, A., FERREIRA, D.
1990. Enantioselective synthesis of flavonoids. Part 2. Polyoxygenated o:-hydroxydihy-
drochalcones and circular dichroic asseSSlPent of their absolute configuration. Tetrahe-
dron 46: 4429-4442.
72. ENDERS, D., ZHU, 1., RAABE, G. 1996. Asymmetric epoxidation of en ones with oxygen
in the presence of diethylzinc and (R,R)-N-methylpseudoephedrine. Angew. Chern. Int.
Ed. Eng!. 35: 1725-1728.
73. ELSTON, c.L., JACKSON, R.F.W, MACDONALD, S.lF., MURRAY, PJ. 1997. Asym-
metric epoxidation of chalcones with chirally modified lithium and magnesium tert-butyl
peroxides. Angew. Chern. Int. Ed. Eng!. 36: 410-412.
74. BOUGAUCHI, M., WATANABE, S., ARAI, T., SASAI, H., SHIBASAKI, M. 1997.
Catalytic asymmetric epoxidation of o:,~-unsaturated ketones promoted by lanthanoid
complexes. 1 Am. Chern. Soc. 119: 2329-2330.
75. LYGO, s., WAINWRIGHT, P.G. 1998. Asymmetric phase-transfer mediated epoxidation
of o:,~-unsaturated ketones using catalysts derived from Cinchona alkaloids. Tetrahedron
Lett. 39: 1599-1602.
76. LASTERRA-sANCHEZ, M.E., FELFER, U, MAYON, P., ROBERTS, S.M., THORN-
TON, S.R., TODD, C.J. 1996. Development ofthe Julia asymmetric epoxidation reaction.
Part 1. Application of the oxidation to enones other than chalcones. 1. Chern. Soc., Perkin
Trans. I: 343-348.
77. VAN RENSBURG, H., VAN HEERDEN, P.S., BEZUIDENHOUDT, B.C.S., FERREIRA,
D. 1996. The first enantioselective synthesis of trans- and cis-dihydroflavonols. 1. Chern.
Soc., Chern. Commun.: 2747-2748.
78. VAN RENSBURG, H., VAN HEERDEN, P.S., BEZUIDENHOUDT, s.c.s.,
FERREIRA, D. 1997. Stereoselective synthesis of flavonoids. Part 4. Trans- and cis-
dihydroflavonols. Tetrahedron 53: 14141-14152.
79. ITS UNO, S., SAKAKURA, M., ITO, K. 1990. Polymer-supported poly(amino acids) as
new asymmetric epoxidation catalyst of o:,~-unsaturated ketones. J. Org. Chern. 55:
6047-6049.
80. HEANEY, H. 1993. Novel organic peroxygen reagents for use in organic synthesis. Top
Curro Chern. 164 (Organic Peroxygen Chemistry): 1.
81. NEL, R.J.J., VAN HEERDEN, P.S., VAN RENSBURG, H., FERREIRA, D. 1998. Enan-
tioselective synthesis of flavonoids. Part 5. Poly-oxygenated ~-hydroxydihydro
chalcones. Tetrahedron Lett. 53: 14141-14152.
82. VAN RENSBURG, H., VAN HEERDEN, P.S., BEZUIDENHOUDT, B.C.S., FER-
REIRA, D. 1997. Enantioselective synthesis of the four catechin diastereomer derivatives.

Tetrahedron Lett. 38: 3089-3092.
83. VAN RENSBURG, H., VAN HEERDEN, P.S., FERREIRA, D. 1997. Enantioselective
synthesis of flavonoids. Part 3. trans- and cis-Flavan-3-o1 methyl ether acetates. 1.Chem.
Soc., Perkin Trans. I: 3415-3421.
84. SHARPLESS, K.B., AMBERG, W, BENNANI, Y.L., CRISPINO, G.A., HARTUNG, 1.,
JOENG, K., KWONG, H., MORIKAWA, K., WANG, Z., XU, D., ZHANG, X. 1992. The
osmium-catalyzed asymmetric dihydroxylation-A new ligand class and a process
improvement. 1. Org. Chern. 57: 2768-2771.
85. JOENG, K., SJO, P, SHARPLESS, K.B. 1992. Asymmetric dihydroxylation of enynes.
Tetrahedron Lett. 33: 3833-3836.
86. WANG, Z., ZHANG, X., SHARPLESS, K.B. 1993. Asymmetric dihydroxylation of aryl
allyl ethers. Tetrahedron Lett. 34: 2267-2270.
87. KOLB, H.C., VAN NIEUWENHZE, M.S., SHARPLESS, K.B. 1994. Catalytic asym-
metric dihydroxylation. Chern. Rev. 94: 2483-2547.
88. BALAS, L., VERCAUTEREN, 1. 1994. Extensive high-resolution reverse 2D NMR
analysis for the structural elucidation of procyanidin oligomers. Magn. Res. Chern. 32:
386-393.
89. BALAS, L., VERCAUTEREN, 1., LAGUERRE, M. 1995. 2D NMR Structure elucida-
tion of procyanidins: The special case of the catechin-(4a-8)-catechin-(4a-8)-catechin
trimer. Magn. Res. Chern. 33: 85-94.
90. STEYNBERG, J.P., BRANDT, E.Y., FERREIRA, D., HELFER, C.A., MATTICE, WL.,
GORNIK, D., HEMINGWAY, R.W 1995. Conformational analysis of oligomeric
flavanoids. 2. Methyl ether acetate derivatives of profisetinidins. Magn. Res. Chern. 33:
611-620.
91. HEMINGWAY, R.W, TOBIASON, EL., McGRAW, G.W, STEYNBERG, J.P 1996.
Conformation and complexation of tannins: NMR spectra and molecular search model-
ing of flavan-3-o1s. Magn. Res. Chern. 34: 424--433.
92. HATANO, T., HEMINGWAY, R.W 1996. Association of (+)-catechin and catechin-
(4a~8)-catechin with oligopeptides. Chern. Commun.: 2537-2538.
93. KHAN, M.L., HASLAM, E., WILLIAMSON, M.P. 1997. Structure and conformation of
the procyanidin B-2 dimer. Magn. Res. Chern. 35: 854-858.
94. HATANO, T., HEMINGWAY, R.W 1997. Conformational isomerism of phenolic pro-
cyanidins: preferred conformations in organic solvents and water. J. Chern. Soc., Perkin
Trans. 2: 1035-10443.
Chapter Eleven
ASTRINGENCY AND POLYPHENOL PROTEIN

INTERACTIONS
Edwin Haslam,l Michael P. Williamson,2 Nicola J. Baxter/ and

Adrian J. Charlton2
1 Department of Chemistry, Dainton Building
University of Sheffield
Sheffield, u.K., S3 7HF
2Department of Molecular Biology and Biotechnology
Krebs Institute, University of Sheffield
Sheffield, u.K., S 10 2TN
Introduction ................................................... 289

Protein: Polyphenol Complexation ................................. 290
Structure-Activity Relationships: Polyphenols ........................ 291
Structure-Activity Relationships: Proteins ........................... 293
Salivary Proline-Rich Proteins (PRPs) ............................ 296
Polyphenol: Protein Precipitation .................................. 299
Water ........................................................ 301
Hydrophobic Effects / Interactions ................................. 303
Astringency ................................................... 308
Postscript ..................................................... 315
INTRODUCTION
The earliest of Man's uses of plant materials rich in polyphenolic metabo-

lites was in the conversion of animal hides to leather, and archaeological records
relate to this operation in Mediterranean regions around 1500 Be. Whilst a
complete scientific understanding of the traditional tanner's art remains, at best,
incomplete, 1 different light has been thrown on facets of this question from other
sources. Thus, polyphenol interactions with proteins (and other biological mole-
cules and macromolecules) underlie a wide range of other apparently unrelated
properties of plant materials. These include: astringency; ecology and chemical
289
290 E. HASLAM et a/.
defense in plants; foodstuffs, nutrition and beverages; fruit and floral pigmentation;
natural glues, varnishes, and exoskeletons; and the influence of diet and the appli-
cation of herbal medicines in the treatment of certain pathological conditions. 2
PROTEIN: POLYPHENOL COMPLEXATION
The interaction between polyphenols and other molecules may be either

reversible or irreversible (Fig. 1). Irreversible complexation, which frequently
takes place autocatalytic ally under the agency of oxygen or in the presence of
enzymes (polyphenoloxidases), creates new covalent bonds between the polyphe-
nol and the other substrate(s) and may lead to the formation of clearly defined
new products (e.g. as in the manufacture of black tea). Conversely, reversible com-
plexation (Fig. 1) is mediated by the gamut of non-covalent forces (hydrophobic
effects, hydrogen bonding, solvation, desolvation) which fall under the umbrella
description of molecular recognition. Reversible polyphenol: protein interactions,
which are believed to underlie the property of astringency, may be studied in solu-
tion or by an investigation of the precipitation processes which ultimately ensue
following extensive complexation. It has been generally assumed, although rarely
proven, that there is a straightforward relationship between the results from these
two types of study.
Scientific studies of the reversible association of polyphenols with proteins
have a history going back almost 200 years to Sir Humphry Davy3 in 1803.
However, until such time as structurally defined plant polyphenols became avail-
able, they were poorly understood. Various quantitative investigations, completed
over the past 25 years, have given considerable insights into the area; "structure-
activity" relationships have been delineated and mechanisms proposed. Although
Protein ( H2 0)a + Polyphenol ( H~ )b
I 1
[ Protein ( H20)c : Polyphenol ( H20 )d]
[ soluble complexes I
/
New Products
\
[ Protein: Polyphenol] (H 2 0)m
[ insoluble complexes I
Figure 1. Reversible and irreversible protein: polyphenol complexation.

ASTRINGENCY AND POLYPHENOL PROTEIN INTERACTIONS 291
Table 1. Protein: polyphenol association. Some

quantitative methods of study
Solution (a) lH and LlC NMR spectroscopy
(b) microcalorimetry
(c) equilibrium dialysis
(d) enzyme kinetics and inhibition
Precipitation (a) radiolabelled protein
(b) haemanalysis
(c) turbidimetric titration
Model Studies (a) caffeine
(b) methylene blue
(c) peptides
(d) cyclodextrins, non-ionic detergents
as a sub-group the proanthocyanidins are believed to be most commonly respon-

sible for the range of properties normally attributed to "tannins" in plants, many
of these complexation studies (Table I) often have been pursued more conve-
niently with a series of biosynthetically interrelated esters of gallic and hexahy-
droxydiphenic acid 2 .4.5 which differ systematically in phenolic content, water
solubility, conformation, and molecular size.
STRUCTURE-ACTIVITY RELATIONSHIPS:
POLYPHENOLS
Precise measurements, obtained by a number of the techniques noted in

Table I and by using a range of different phenols and proteins, fully support and
quantify many of the earlier suggestions and proposals concerning the chemical
nature of the most effective "vegetable tannins." Thus, for example, equilibrium
dialysis studies with the protein BSA (MR 69,000) showed6 (Table 2) that mole-
cular size, the number and disposition of phenolic nuclei, conformationalflex-
ibility, and water solubility were the dominant features in the determination of
the strength of binding of a particular polyphenol to the protein. Good water solu-
bility depresses the effectiveness of the polyphenol in protein complexation,
whilst increasing molecular size [cf the "dimeric" ellagitannin, rugosin D (1)]
and conformational flexibility [cf 13-1,2,3,4,6-penta-O-galloyl-D-glucose (2) and
the ellagitannins castalagin (5) / vescalagin (6)] enhance association with the
protein. In the context of conformational mobility, it is important to note that the
biosynthetic introduction of a biphenyl linkage between two galloyl ester units in
a metabolite-i.e. the generation of a constrained hexahydroxydiphenoyl ester
group typical of the ellagitannins7-invariably leads to a decreased affinity for
protein, although the number of aromatic nuclei and phenolic groups remains
exactly the same.
292 E. HASLAM et al.
HO
HO
Rugosin D (1)
HO OH
HO~OHO~OH
~o~JS~
0t, O~O~
HO¥OH YOH
OH OH
!3-1,2,3,4,6-Penta-O-galloyl-D-glucose (2)
Table 2. Protein-polyphenol complexation. Values of the free energy of transfer

(_~G8.tr) of a protein, BSA, from an aqueous medium to an aqueous medium
containing a polyphenol ligand, (concentration-mf)
Polyphenol
13-1,3,6-Trigalloyl-D-glucose 636 0.9
13-1,2,3,6-Tetragalloyl-D-glucose 788 9.1
13-1,2,3,4,6-Pentagalloyl-D-glucose (2) 940 26.9
Castalagin (5 )\Vescalagin (6) 934 1.0
Rugosin D (1) 1,847 58.7
Determined20 in aqueous solution at 25°C, pH 2.2 and at a free ligand (polyphenol) concentration (mr) of
4.5 mole kg-I. Results are expressed in terms of the free energy of transfer (AG8 ,,,) of the protein BSA from
an aqueous medium to an aqueous medium containing the polyphenol ligand, (concentration-mr), High
values of the quantity -AG8 ," are indicative of strong complexation, conversely low values portray poor
complexation,
principal sites at which complexation with protein is mediated
,1..........
1 ",-' "'"
HO
~
Ol ~,O -"""
~
.& OH
""
"
(XI
OH
OH HO~O ..(.... OH
galloyl esters. etc. ~OH
OH
flavan-3-o1s
Figure 2. Principal aromatic sites in natural polyphenols for the complexation with proteins. '1
High resolution NMR studiesS- lO generally confirm what has always been
assumed, namely that the aromatic nuclei of polyphenols provide the principal
sites for association with proteins. The data obtained from equilibrium dialysis
studies of a range of simple phenols (resorcinol, catechol, pyrogallol) and BSA
gave Scatchard plots whose analysis showed that the affinity of both catechol and
pyrogallol for BSA was three orders of magnitude greater than that of resorci-
no1. 11 Interestingly, resorcinol, at 20 D e, has twice the solubility in water of cate-
chol and pyrogallo1. This observation also gives quantitative support to earlier
ideas that the o-dihydroxy- and o-trihydroxy- aromatic nuclei of natural polyphe-
nols represent the principal sites for complexation with proteins (Fig. 2).
Molar enthalpies of transfer (L1H 0 .tr ) may be analogously determined by
microcalorimetric methods. 6 Plots of (L1H0 .tr ) versus (T L1S 0 .tr ), for a range of
natural galloyl and hexahydroxydiphenoyl esters complexing with BSA, gave
linear plots (slope -1) indicative of enthalpy-entropy compensation. This pro-
portionality indicates that particular features of protein-polyphenol interactions
also may be examined, with suitable caution, by reference to molar enthalpies of
transfer (L1H0 ,tr). Under conditions in which the polyphenol: BSA complex was
precipitated, it was observed that there was little or no heat change on precipita-
tion. This favors the proposition (and often the assumption) that the structures
of the polyphenol: protein complexes are similar in the solid state to those in
solution.
STRUCTURE-ACTIVITY RELATIONSHIPS: PROTEINS
In terms of its findings in relation to polyphenol-protein interactions there

is little doubt that the paper by Butler and Hagerman 12 probably has been the most
influential, particularly in the way in which it has directed subsequent work in
this area. The Purdue workers used a competitive binding assay and showed that
the condensed proanthocyanidin polymer (tannin) from Sorghum bicolor, at

pH 4.9, had a relative affinity for various proteins and synthetic polyamides which
ranged over four orders of magnitude-indicating that this polyphenol interacts
quite selectively with proteins and protein-like polymers. Proteins were most
efficiently precipitated at or near their isoelectric points. Tightly coiled globular
proteins were not so readily precipitated as those which had a random coil con-
formation and, in particular, those which were rich in the amino acid proline
(proline-rich proteins-PRPs). They also showed that the relative affinities of
polypeptides and proteins for the polymeric proanthocyanidin were influenced by
their size. The low affinities of small peptides and low molecular weight oligomers
of proline for the tannin and the non-linearity of the increase in affinity which
was observed with higher oligomers of proline implied, they suggested, that
the complexation involved the deployment of multiple binding sites on both
substrates.
In their seminal work, Butler and Hagerman,12 besides using various
oligomers of polyproline, utilized calfskin gelatin and a proline-rich protein from
the rat parotid gland as examples of PRPs. Luck et al. 13 later demonstrated the
same high affinity of polyphenols for another group of PRPs-the milk proteins,
as, 13 and K-caseins. These amphipathic molecules are, under appropriate
conditions, able to form micelles which, it was suggested, have the ability to
solubilize the added polyphenols within the micellar structure.
This affinity of natural polyphenols for proteins also extends to simple pep-
tides, which are proline-rich and/or hydrophobic in character. Several groups have
used high resolution NMR spectroscopy to try to define more precisely the origins
of the strong affinity which PRPs have for polyphenols. 8- IO,13-15 In their studies,
Hatano and Hemingway9 used intermolecular nOes to look at the association of
(+)-catechin and the procyanidin catechin-(4a-8) -catechin (via their phenolic
nuclei) with oligopeptides ranging from Pro-Gly, Pro-Val, and Pro-Phe to Gly-
Pro-Gly-Gly and bradykinin (Arg-Pro-Gly-Phe~Ser-Pro-Phe-Arg). They con-
cluded that strong selectivity for prolyl residues was not apparent; rather
complexation was directed to the conformationally accessible hydrophobic
regions of the peptide. Although agreeing that polyphenols associate preferen-
tially with the available hydrophobic sites on a polypeptide, other workers have
come to somewhat different conclusions concerning the role of prolyl groups in
polyphenol complexation.
Haslam 15 made a detailed NMR study of the interaction of a series of galloy I
esters with the bioactive peptide bradykinin (3). In aqueous media, NMR and CD
studies suggest that bradykinin is in rapid equilibrium among many conformers
and does not show any persistent structural features such as 13-turns or internal
hydrogen bonds. Bradykinin is proline rich and with two phenyl residues is
hydrophobic in character; its strong association with and precipitation by natural
polyphenols is, therefore, not unexpected. Jeffrel 6 determined association con-
stants for the interaction of bradykinin with 13-1,2,3,4,6-penta-O-galloyl-D-
glucose (2), ~-l ,3,4,6-tetra-O-galloyl-D-glucose and ~-l ,3,6-tri-O-galloyl-D-

glucose. A rapid decrease in the affinity of the polyphenol for the peptide was
observed with decreasing galloyl content, and this mirrors exactly the data
obtained earlier with larger protein molecules, i.e. a strong dependence upon mol-
ecular size, the number and arrangement of phenolic groups, and water solubil-
ity of the polyphenol.
Bradykinin (3)
No evidence was obtained to show that the peptide underwent significant

changes from a random coil conformation upon complexation with the polyphe-
nols. The most significant proton chemical shift changes in the bradykinin were
associated with each of the three proline residues and the two phenylalanine
groups. Strong intermolecular nOes were observed between the 2-H galloyl
protons and the protons of these groups, suggesting that these amino acid side
chains were participating preferentially in the complexation with polyphenol. 1t-
1t Stacking was put forward as the mode of interaction with the phenylalanyl
residues in the peptide (Fig. 3). However, it was suggested that there is probably
not a specific mode of binding between the polyphenol and peptide substrates.
Rather the driving force was visualized as the relatively unselective association
of the aromatic nuclei of the polyphenol with prolyl and phenylalanyl groups on
the nonapeptide which together form a relatively hydrophobic area on the surface
of the peptide.
The role of the side chain of the amino acid arginine (arg) in these com-
Phenylalanine residues - 'It-It' stacking
H~
~--
>
HO
HO~ HO
Arginyl residues - hydrogen bonding
NH2
H'+A~
N N
1 H
HO f;i
HO~::>
HO
Figure 3. Possible non-covalent interactions of phenolic rings with amino-acid residues in

Bradykinin
plexations remains unclear. However, analogues of bradykinin, which lack either

of the terminal arginine groups at positions I and 9, were significantly inferior in
their capacity to bind the polyphenol ~-1,2,3,4,6-penta-O-galloyl-D-glucose, sug-
gesting that these residues may also have a specific role to play in the binding of
polyphenolic substrates. On the basis of observations in related fields, 16.1 7 it was
suggested that the 1t-electrons of the "electron rich" phenolic nucleii act as hydro-
gen bond acceptors forming quasi-hydrogen bonds to the hydrogen bond donor
(-NHC(NH)-NH2) of the arginyl side-chain (Fig. 3).
Salivary Proline-Rich Proteins (PRPs)
Salivary proline-rich proteins have a repetitive primary structure particu-

larly rich in the amino acids proline (P), asparagine (N) and glutamine (Q).
In a comprehensive study, two peptides (I and II, 19 and 22 amino acids, Fig. 4)
overlapping in sequence, corresponding to the mouse salivary proline-rich protein
MP 5 repeat sequence, were synthesized and their interactions studied CH NMR
spectroscopy) with a range of simple phenols and natural polyphenols, [e.g.
(--epigallocatechin-3-0-gallate (4) and ~-l ,2,3,4,6-penta-O-galloyl-D-glucose
(2)V,iO.i3.14 Analysis was conducted by measurement of proton chemical shift
changes induced upon complexation with the polyphenol (Fig. 4), by the deter-
mination of dissocation constants (~) for the soluble complexes (Table 3), and
uy the measurement of the intermolecular nOes observed in NOESY and ROESY
spectra of the complexes (Fig. 5).
peptide I
1 19
~ • • • • ~ • • • ~~® • • ~ • • ~~
.... ..•
QGP P P QGGP QQRPPQ PGNQ
peptide II 1
• • ~~® • • ~ • • ~~
~ ~ ~
22
Ali ppm
0.2
0.1
• • ~~® • • ~ • • ~~~ • • • • ~ • • • ~
1 n
Figure 4. Peptides I and II. Plot of the mean change in chemical shift (~8) in H20 / d 6 DMSO
(9; I v/v) on going from the free peptide II (22 amino acids) to a I; 5.6 molar ratio of peptide
and 13-1,2,3,4,6-pentagalloyl-D-glucose (2). Chemical shift changes are averaged over all
protons in each residue except for those (e.g. NH) which are exchangeable (e.g. prolyl groups
and arginyl group each have 7 protons).
Table 3. Self association (Ka) and complexation (Kd' dissociation constants of

complex) of phenols with the salivary peptide II (22 amino acids). 8 The
approximate number of binding sites is given by "n"
Phenollsolvent,·b K, (M) Kd (M- I ) n
propyl gallate' 5 x IO- J 8
(-)-epicatechin' 27 5 x 10-3 4
(- )-epicatechin-3-0-gallateh 48 2 x 10-3
(-)-epigallocatechin-3-0-gallate (4)b 50 2 X 10-3
procyanidin B2h 300** 5 X 10-5 1-2
theaflavin b 230
13-1,3,6-trigalloyl-D-glucose b 30 I x 10-4 2
13-1,2,3,4,6-pentagalloyl-D-glucose (2)b 241 5 x 10-' 2
Solvents: '90% waterllO% deuteriated water; b90% waterllO'l'o ['HoI DMSO. Measurements at 276 0 K using
Brucker AMK 500 and 600 spectrometers.
"estimated value for K,.
* *** ** *** * **
• • LlLl®• • Ll.·LlLlLl • • • • Ll • •
1
eLl 22
Figure S. Intermolecular nOes observed in NOESY and ROESY spectra of the complexes
between peptide II and aromatic protons of the 6-galloyl ester group of ~-l ,2,3,4,6-
pentagalloyl-D-glucose (2) such as to suggest that intermolecular contact (*) occured primar-
ily between the galloyl ester groups and specific residues in the peptide chain.
These experiments supported the data obtained earlier with larger protein
molecules, i.e. a clear dependence in the association upon molecular size, the
number and arrangement of phenolic groups, and water solubility of the polyphe-
no!. They also strongly suggested that the principal binding sites on these pep-
tides are the prolyl residues themselves together with the preceding (N-terminal)
amino acid side-chain. Furthermore, measurements in the complexation studies
with both (-)-epigallocatechin-3-0-gallate (4) and 13-1,2,3,4,6-penta-O-
galloyl-D-glucose (2) of significant entropy changes [+80 - 90j deg- I M- 1 and
255 j deg- I M- 1 respectively] strongly indicate that the processes are hydrophobi-
cally driven (water structure breaking) ones.
OH
~OH
H0I(Y0'l,"'VOH
~"'O
OH O~OH
~OH
OH
(-)-epigallocatechin-3-0-gallate ( 4 )
It was concluded that the principal binding sites on the two peptides are
the apolar methylene and methine groups on the prolyl residues themselves. The
interaction was visualized as a hydrophobically driven association between a
galloyl ring and the exposed hydrophobic surface of the pyrrolidine ring, prefer-
entially that face containing the a-proton. Hydrogen bonding of one or two phe-
nolic groups to the tertiary amide carbonyl group on the adjacent peptide linkage
was presumed to be a secondary interaction helping to stabilize the complex, (Fig.
6). Thus, proline-rich peptides and proteins which have an open, random-coil type
of conformation have a high affinity for polyphenols not only as a result of their
extended structures, but also by virtue of the prolyl groups themselves which, in
a figurative sense, provide 'sticky patches" on the protein for the phenolic nuclei
of the polyphenolic substrate.
Proline residues - hydrophobic stacking
Figure 6. Figurative visualization of the non-covalent hydrophobic interaction between the

pyrrolidine ring of a prolyl residue and a phenolic ring.
POLYPHENOL: PROTEIN PRECIPITATION
One of the first scientific papers on polyphenol (vegetable tannin): protein

precipitation was that of Sir Humphry Davy nearly two hundred years ago.) On
the basis of work over the past 4 or more decades,2.5.17 it may be concluded, almost
certainly, that the efficacy of polyphenol (vegetable tannin) binding to proteins
derives in large part from the fact that polyphenols are polydentate ligands able
to bind simultaneously / consecutively at more than one point to the protein struc-
ture. 6. 11 ,I) Protein precipitation is the result of a two stage process in which soluble
complexes are first formed (Fig. I). As the position of the equilibrium changes,
then, as a second stage, these complexes may aggregate and precipitate. In
general, when polyphenols bring about precipitation of proteins from solution,
two extreme situations may be envisaged in relation to the stoicheiometry of
the precipitate. At low protein concentrations the polyphenol associates at one
or more sites on the protein surface. Ultimately, this gives a monolayer which
is less hydrophilic than the protein itself (Fig. 7a). Aggregation and precipita-
tion then ensue. Conversely, when the protein concentration is high, the relatively
hydrophobic surface layer is formed by complexation of the polyphenol onto
the protein with concomitant cross-linking of different protein molecules by
the multi-dentate polyphenols (Fig. 7b). Precipitation ensues as above. This
proclivity to cross-link protein molecules at higher protein concentrations
explains the changing stoicheiometry of the aggregates with changing protein
concentrations-an observation first indicated by Davy! More polyphenol is, thus,
required to precipitate polyphenols from dilute solutions than from concentrated
ones. 18
An interesting corollary of this hypothesis is that simple phenols themselves
should also be capable of bringing about the precipitation of proteins from
aqueous solution, in concert with polyphenols, or, if a sufficient concentration
, ,
(a)
\
..
I
I
I
,
,,,
I
precipitation
I
hydrophobic suface layer
polyphenol
,
\
\
,,
\
\
\
, \
, I ;
\ \
(b) \
\
,48,0
\ 0, ..
, ·-0 ~
Figure 7. Polyphenol-protein precipitation processes: (a)-low initial protein concentrations;
(b }--high initial protein concentrations.
can be maintained in solution, in isolation to push the equilibrium in favor of the

phenol protein complex. In the first, and probably more typical instance in vivo
(Fig. 8b), they are envisaged as contributing to the formation of the hydrophobic
surface layer coating the protein. In the second instance (Fig. 8c), the hydropho-
bic layer would result from the presence of simple phenols alone. For many simple
phenols the limit is provided by their solubility in water. It can, for example, be
achieved with pyrogallol (1 M) in the presence of BSA (3 x 10-5 M). Such factors
are clearly of some import when assessing the total astringency of plant extracts
and beverages rich in simple phenolic metabolites.
( a) (b) (c)
polyphenol I protein
o simple phenol 0
Figure 8. Protein precipitation: the role of simple phenols and their contribution to a hydropho-
bic surface coating-(a)-polyphenols alone; (b}-polyphenols plus simple phenols; (c)-
simple phenols alone.
The precipitation of protein by polyphenols is, in the absence of chemical

changes, an equilibrium process (Fig. I) and often it may be reversed by the addi-
tion of further protein. During his observations3 of the astringent vegetable infu-
sions using gelatin, Davy remarked:
In ascertaining the proportions of tannin in astringent infusions, great care must be taken to
prevent the presence of any excess of gelatine for when this excess exists, I have found that a
small portion of the solid compound formed is re-dissolved.
However, it is interesting to note that typical proline rich proteins (PRPs)

and polypeptides, which have a strong affinity for polyphenols, are unable to res-
olubilize the precipitated complexes when added in excess. The phenomenon
appears to be a consequence of thermodynamic rather than kinetic factors.
However, protein-polyphenol complexes may be dissociated, without denatura-
tion of the protein, with solvents such as acetone,19 with caffeine,20.21 with urea,22
polyvinylpyrrolidone,23 polyethylene glycols, and detergents. 24.25
WATER
Nobel laureate Albert Szent-Gyorgy once cryptically remarked "Biology

has forgotten water, or never discovered it." The presence of water is intuitively
accepted and invariably tacitly ignored, but the role of water and water solubility
is one of the key factors in the phenomenon of polyphe no I complexation. During
the association processes, the reorganization of the various solvent shells provides
an important driving force for the interactions to take place (Fig. 1). Water is by
far the most common liquid in our environment. Its structure can be represented
by a somewhat distorted tetrahedron with the oxygen atom at the center; the
protons are directed towards two of the vertices, and the lobes of electron
density-the so-called electron lone pairs ("rabbits ears" or "water wings")-are
directed towards the other two. Because water was probably indispensable for the
genesis of life and because of its unique properties, in a pure form and as a
solvent, the structure of water (and ices) has been the subject of intense work
since the last century. The literature is voluminous. However, agreement on a
comprehensive theory of the molecular structure of bulk water has proved elusive.
From the point of view of intermolecular hydrogen bonding, the individual water
molecule is well placed in having both double donor (protons) and double accep-
tor (lone pairs) hydrogen bond functionality based about one central oxygen atom.
It is accepted that the structure and properties of bulk water derive from exten-
sive hydrogen bonding interactions between individual water molecules. It is
the nature and arrangement of these interactions which are still the subject of
debate.
The dual hydrogen bonding functionality of water molecules is seen most
clearly in the crystal structure of ordinary hexagonal ice (Fig. 9) in which each
water molecule has four nearest neighbors to which it is hydrogen bonded. These
four hydrogen bonds are spatially arranged with local tetrahedral symmetry; that
is the oxygen atoms of the neighboring water molecules occupy the vertices of a
regular tetrahedron surrounding the oxygen atom of the central molecule.
This propensity of water to enter into extensive hydrogen bonding networks
extends to the various clathrate hydrates which it forms. As in ice, the hydrogen
bonded water molecules are four coordinate in a distorted tetrahedral arrange-
ment. They are "ice-like" but differ in having large internal cages. 26 Owing to the
steric requirements imposed by the inclusion of other molecules as guests within
these voids, their hydrogen bond networks are organized differently from that of
hexagonal ice.
The melting of ice to produce ordinary liquid water clearly entails funda-
mental changes in the way in which water molecules are arranged relative to one
another. Essentially two models have been developed. Frank's "jlickering cluster"
theory 27.28 invokes the concept of clusters of three or four coordinate water mol-
ecules with a mean life-time of about a nano second. A second type of modeI/9
of which a number of variations have been proposed, is the uniform continuum
model, orignally suggested by Bernal and Fowler. Essential to this second model
is the assumption that all of the hydrogen bonds in liquid water remain intact.
However, they may be stretched, bent, and distorted to an extent that varies with
temperature and pressure; the existence of "broken" hydrogen bonds is not specif-
ically recognized. In the immediate vicinity of a given water molecule, the
arrangement of the quasi-equilibrium positions resembles that in ice, but further
• --....... 'lone pair ' of lo~

HII~iO. _----~ electron density
H
I ,
H"";;O"'H HI
H ,
_" / O"-H-.O. H
• 9 ---H-O.. H,' ··H..... I
, ·H ..... / O-H---·O.
H O---H-O HI' " .
I : ....... H...,'
H-'O... H ··o:...-H
H H, I '
/ O"-H-O H
····0 ---H-O. ," "'H I
Figure 9, The water molecule. I
H
"H /H
..... O.. --H-O
..... O-H- .. O
HI' ".
Part-structure of hexagonal ice.
i ........ H '
Hydrogen bonds are shown as , ........ ~H
dotted lines; covalent bonds as full
I
lines. H
away from this center the arrangement and connectivity may be significantly dif-
ferent. In particular, it is likely that the O-H .. ·.... O hydrogen bonds will be bent
and stretched to varying degrees.
HYDROPHOBIC EFFECTS / INTERACTIONS
In early studies of the mechanism of vegetable tannage, the interaction of

tannic acid with gliadin mono layers was shown to produce a less compressible
and considerably more rigid film. From later analogous work with collagen mono-
layers, Ellis and Pankhurseo concluded that "the tannin molecule shall have at
least more than one reactive group which shall be so spaced that they are able
to combine inter-molecularly with the protein. It is therefore not surprising to find
that such simple substances as mono-, di- and tri-hydric phenols are not capable
of tanning collagen monolayers. .......... To summarize, the reaction of collagen
monolayers with condensed tannins appears to be predominantly by hydrogen
bonding between the multifunctional tannin molecules and the keto-imide groups
of the protein." The idea that the most effective tannins (polyphenols) are multi-
functional is still widely accepted, but the view that the non-covalent interactions
take place primarily through the deployment of multiple hydrogen bonds between
the phenolic hydroxyl groups and the carbonyl groups of the peptide linkages has
undergone some modifications. The emphasis on hydrogen bonding had a number
of origins, e.g. that tannins were bound by modified collagen and synthetic nylon
polymers such as polycaprolactam, the latter containing the peptide linkage as
the "only reactive grouping. ,,22,31 However, it has long been recognized that the
strong charge solvating property of water and its own intrinsic hydrogen bonding
ability mean that the possibility of obtaining large binding energies from these
forces alone is improbable. 32
Evidence that might be interpreted as showing that hydrophobic interac-
tions are important in the formation of tannin-protein complexes had been noted
earlier, (for example the complexes could be dissociated by detergents and organic
solvents). Hoff and his colleagues/ 3 however, were the first to draw serious atten-
tion to the involvement of hydrophobic groups in the formation and stabilization
of tannin-protein complexes, The hydrophobic nature of tannins was, thus,
demonstrated by their adsorption on uncharged polystyrene resin, Complex for-
mation between condensed tannins and gelatin or poly-L-proline was enhanced
with increasing ionic strength and temperature as predicted for hydrophobic inter-
actions. According to Jencks, "Hydrophobic effects" are probably the single most
important factor providing the driving force for non-covalent intermolecular inter-
actions in aqueous media. 32 They may broadly be defined as an interaction of the
molecules with each other which is stronger than the interaction of the separate
molecules each with water. No mechanism is implied by this definition.
Water is a notoriously poor solvent for apolar compounds, such as hydro-
carbons and the noble gases, at moderate temperatures and pressures, This reluc-
tance to dissolve in water has been popularly attributed to the hydrophobicity of
these substances. For an apolar compound to dissolve in water, it must intrude
into a liquid that is characterized by an extended network of hydrogen bonds and
has a high cohesive energy. Many rationalizations have focused upon the large
losses of entropy which accompany the dissolution of non-polar solutes, such as
a noble gas or a hydrocarbon, in water. Frank and Evans 28 first sought to ratio-
nalize the unusual thermodynamic properties of non-polar solutes in water by
postulating a particular ordering of water molecules (structure making) around
the solute. They described the process as "iceberg" formation:
when a rare-gas or a non-polar molecule dissolves in water at room temperature, it modifies the
water structure in the direction of greater 'crystallinity '-the water, so to speak, builds a micro-
scopic iceberg around it.
Direct experimental evidence for the idea of the "iceberg" principle of

Frank and Evans was first realized in the X-ray analysis of the small hydropho-
bic protein crambin (46 amino acids, MR = 4,720) which is found in the embry-
onic tissue of seeds from Crambe abyssinica, commonly known as Abyssinian
cabbage. 34-36 Similarly, analysis of the water structure of yB-crystallin, a struc-
tural protein of the eye lens, shows extensive three-dimensional cages of highly
ordered solvent molecules around exposed non-polar groups (such as the pyrro-
lidine ring of proline and the phenyl ring of phenylalanine), screening the
hydrophobic patches. Likewise, the reported crystal structures of the clathrate
hydrates suggest that re-organization of the random hydrogen bonded netw?rk of

water molecules is indeed possible to facilitate the creation of solvent cages to
accomodate molecules with large non-polar groupings (e.g. t-butylamine).
Many may regard the explanation as not totally adequate. The concept has,
nevertheless, provided a useful picture upon which to base qualitative explana-
tions for a number of phenomena which may be placed under the umbrella
description of "hydrophobic effects" or "hydrophobic interactions." The inter-
action of non-polar groups or regions in biopolymers is thought to be a major
determinant in selecting the native conformation of the macromolecule. The
schematic representation of such a "hydrophobic interaction," between exposed
alanyl and leucyl side chains on a protein, was first illustrated figuratively by
Nemethy and Scheraga37 .38 (Fig. lO) and visualizes an approach of the side-chains
until they virtually "touch," with a concomitant reduction of the number of nearest
neighbor water molecules. It is favored because of an entropy effect; highly
ordered water molecules around separate groups are, by virtue of the hydropho-
bic interaction, released into the bulk medium-the process, in relation to water,
is structure breaking. In slightly more colorful language, Perutz39 described, with
seductive simplicity, the key role which "hydrophobic forces" play in protein
folding:
He (Kauzman) suggested that the water molecules" anarchic distaste for the orderly regimen-
tation imposed upon them by the hydrophobic side chains of the protein forces these side chains
to shy away from water and to congregate in the centre of the protein.
There is ample experimental evidence that relatively non-polar molecules

have a favorable net free-energy of interaction with each other in aqueous media
and that these hydropohobic effects are probably the most significant driving force
for non-covalent intermolecular interactions in aqueous media.
Hansch and his collaborators40A ! have employed a constant 1t to estimate
the hydrophobic bonding of various substituent groups in a particular molecule.
They defined 1t as 1t == log P, /PH. where PH is the partition coefficient of a parent
compound between octan-I-ol and water, and P x is that of the derivative X and
represents the free energy of transfer of the substituent from an aqueous to a
lipophilic phase. They studied the adsorption of a number of phenols (-20)
by bovine serum albumin and found that binding depends on the lipophilic char-
acter of the substituent; a linear relationship exists between the logarithm of
the binding constant and 1t. The authors concluded that, since the adsorption of
the phenols onto the protein closely parallels the transfer of the phenols from a
water phase to one of octan-I-ol, the binding is probably promoted in large
measure by the lipophilic groups and that the phenolic group itself may playa
secondary role.
The significance of solvation and "hydrophobic effects" is also vividly illus-
trated by the example of three natural polyphenols-I3-I,2,3,4,6-penta-O-galloyl-
D-glucose (2), and the diastereoisomeric pair castalagin (5) and vescalagin (6)
(a)
alanyl
HN
Me,
reo I
(b)
(c)
I
I Me
M'~l,""l
OC NH
• carbon, etc "hydrogen ® NH group

o water molecules
Figure 10. Schematic representation of the hydrophobic effect resulting from the congregation
of two amino acid side chains (alanyl & leucyl) of a protein (a to b), after Nemethy and
Scheraga. 37 ,38 There is a reduction in the number of nearest neighbor water molecules; water
molecules are "released" to the bulk medium (c) and this is one of the driving forces for the
interaction.
derived from Quercus and Castanea species (Fig. II). The molecule of p-
1,2,3,4,6-penta-O-galloyl-D-glucopyranose has, in its most favored conforma-
tion, the shape of a circular disc. Molecular models clearly reveal that the
periphery of the molecule is hydrophilic, by virtue of the presence of the pheno-
lic groups, whilst the upper and lower faces of the molecule (disc-shape) are, in
contast, largely hydrophobic in character. p-I,2,3,4,6-Penta-O-galloyl-D-glucose
is amorphous and has a limited solubility in water (-1.0 mM at 20°C). Solutions
of higher concentration, obtained by heating to 50-60°C, readily gel on cooling
to ambient temperature. Such gels probably arise from the ability of the polyphe-
nol to form extensive three dimensional lattices. These would result from verti-
cal stacking of molecules (rather as a pile of coins) with their hydrophobic faces
brought in juxtaposition to minimize water contacts and from intermolecular
ASTRINGENCY AND POLY PHENOL PROTEIN INTERACTIONS 307
OH
Figure 11. ~-1,2,3,4,6-Penta-O-gal1oyl
D-glucose: Castalagin and Vescalagin: a C-l , a-OH , Castalagin ( 5 )
structural comparison. C-l, J3-0H, Vescalagin (6)
cross-linking of stacks by hydrogen bonding. The gel is readily disrupted by

addition of solutes such as l-~-octyl-D-glucopyranoside which presumably func-
tions by intercalation of the hydrophobic hydrocarbon chain in the stacks.
The molecule ~-l ,2,3,4,6-penta-O-galloyl-D-glucose (2) has, in principle, a wide
range of conformations available tc it. The molecule binds strongly to proteins
and is an effective "tannin." In partition experiments, when bis-N-butyl ether is
the organic phase, ~-l ,2,3,4,6-penta-O-galloyl-D-glucose partitions exclusively
to the aqueous phase. However when the organic phase is octan-I-ol the value of
K [octan-I-ol/water] is 32.11
In complete contrast, castalagin (5) and vescalagin (6», which are unique
open-chain derivatives of D-glucose, are both soluble in water but are not
extracted therefrom by ethyl acetate and show a K [octan-I-ol I water] of -0.1.
Both molecules in the conventional tests for tannins react poorly, and quantita-
tive studies show that they bind weakly to proteins. I I Although these results
clearly support the contention of Hansch and his colleagues that solvation and
"hydrophobic effects" are the major driving forces favoring phenol (and hence
polyphenol) complexation with proteins, they do not answer the question why
such dramatic differences should exist. These comparative properties are certainly
not ones that would have been readily predicted a priori. Thus, although both
vescalagin and castalagin have formally just 6 hydrogen atoms less than ~-
1,2,3,4,6-penta-O-galloyl-D-glucose, all three molecules have a similar molecu-
lar mass and possess 5 aromatic nuclei and 15 phenolic hydroxyl groups. At the
moment, one can only speculate as to the nature of the solvation shell around
polyphenol molecules such as ~-1,2,3,4,6-penta-O-galloyl-D-glucose, castalagin,
and vescalagin. Presumably there will be significant clusters (a cloud) of water
molecules in the vicinity of the three phenolic groups of each galloyl ester TImc-
tion. The distribution of water molecules around the remainder of each aromatic
ring is, in present circumstances, far more difficult to define.
ASTRINGENCY
The objective evaluation of the taste and flavor of foods and beverages still
depends largely upon sensory perception. A fundamental understanding of the
physiology and chemistry underlying various aspects of taste is still lacking. This
is certainly the case with respect to that quality generally referred to as astrin-
gency which influences the acceptability of many fruits and fruit products, such
as fruit juices and wines. Astringency results from a loss of lubrication in the
mouth. It is normally recognized as a feeling of extreme dryness and constric-
tion, roughness or puckeriness of the palate which takes a significant time to
develop. It is diffuse and not confined to a particular region of the palate. The
word astringent is derived from the Latin ad (to) and stringere (bind); thus astrin-
gency is properly defined as a binding reaction, a sensation of touch. Indeed,
astringents in medicine are recognized as substances that bind to and precipitate
proteins. Typical astringents include the salts of multivalent cations (AI, Cr, Zn,
Pb, Ca, B), dehydrating agents such as alcohol and dimethyl ketone, mineral acids
and natural polyphenols (vegetable tannins). A mucous membrane covers all the
exposed surfaces of the mouth which are moistened by the secretions from the
salivary glands. According to Bate-Smith,42 the primary process whereby astrin-
gency develops is via precipitation of proteins and mucopolysaccharides in the
mucous secretions. Accepting this view (which is still broadly assumed), an
understanding of the mechanism of the astringent response focuses attention auto-
matically upon (i) the molecular basis of the action of salivary proteins which
give rise to lubrication in the mouth and (ii) polyphenols and their interactions
with salivary proteins which result in the loss of lubrication.
In terms of taste, natural polyphenols have two distinctive characteristics-
astringency and bitterness. Distinguishing between these qualities is an important
facet in the training of members of taste panels in the food industry. The data
Table 4. Astringency: polyphenols: threshold values for the perception of an

astringent response (Clapperton, Williamson, and Haslam, unpublished data)
Polyphenol Threshold (mM)
propy I gallate 212 1.2
~-l ,3,6-tri-O-galloyl-D-glucose 620 0.05
~-l ,2,3 ,6-tetra-O-galloyl-D-glucose 788 0.009
~-I ,2,3,4,6-penta-O-galloyl-D-glucose (2) 940 0.009
rugosin D (1) 1,847 0.005
castalagin (5)/vescalagin (6) 936 0.4
(+)-catechin 290 0.3
(-)-epicatechin 290 0.9
(+ )-catechin-3-0-gallate 458 0.1
(0.002)-epigallocatechin-3-0-gallate (4) 474 0.4
procyanidin B-2 578 0.05
chi orogenic acid 353 0.4
shown in Table 4 were measured by an experienced taster from the food indus-
try (Clapperton, Williamson and Haslam, unpublished data). Increasing aliquots
of each polyphenol were taken and a "quasi-binding" curve established for each
sample. From these curves the minimum threshold concentration for the percep-
tion of astringency in the palate was determined.
Even the most cursory examination of the data in Table 4 indicates a broad
correlation of the astringent response of a particular polyphenol with the picture
presented earlier of the relationship between polyphenolic character and capac-
ity to bind to protein (e.g. molecular size, the number and disposition of pheno-
lic nuclei, conformational flexibility and water solubility). Such a correlation
strongly suggests that, in the case of polyphenols, it is (as originally suggested
by Bate-Smith) their binding to salivary proteins which fundamentally underlies
the development of astringency in the mouth.
Saliva is produced by the major salivary glands which empty their secre-
tions into the oral cavity. The macromolecules in saliva consist almost exclusively
of proteins (-1.0-3.Smg./ml.), and amino acid analyses of human salivary pro-
teins have demonstrated the presence of an unusually large amount of proline
(-16-33% of total amino acids). From the work in severallaboratories,43 it is now
clear that saliva contains a unique group of proteins-proline-rich proteins or
PRPs. They can be sub-divided into acidic (APRP), basic (BPRP), and glycosy-
lated (GPRP) proteins. In human parotid saliva, these account for 17%,23%, and
30%, respectively of the total protein. The acidic proline-rich phosphoproteins
have important biological functions related to providing a protective environment
for teeth and may modulate the adhesion of bacteria to oral surfaces. The bio-
logical function(s) of the glycosylated and basic proline-rich proteins are less
clearly defined but most probably include the binding of oral bacteria and
masticatory-lubricating properties. Salivary proline-rich proteins have a repeti-
tive primary structure particularly rich in the amino acids proline, asparagine, glu-
tamine, and glycine. The reason for the large number of isoforms of PRPs is not
entirely clear; some smaller APRPs are thought to be derived by post-translational
cleavage of the larger APRPs, and one or more basic PRPs may similarly arise
by proteolysis of acidic PRPs.
The primary structures of three closely related ISO-residue and three 106-
residue human APRPs, considered to be derived by post-translational cleavage of
the larger APRPs, have been defined. 44 The amino acid sequence of the phos-
w
AOO
•
o~
OH
)IN''
OH
)IN''
H~o
H 0 H 0
• P,prolyl ~ N, asparaginyl ~ Q, glutaminyl
1
00~~00~®~00~.~~~00
.•• ..
•• O@O~O~0000~ ••• ~·
.••••
·~~O~.O~·0.~~00-»
»» •• ~
~ ~~ ~~~~ ~
·0.~O • • ~~··O ••••~.
•. .••• .. .••
0.~·
.~
~·~O.~
®.~
•• ~~··O.0 ••0·0
• • • ~~ •• O~~ • • • • • •
150
~ ®.~
~D,E <9 K,R ~ N,Q
~ A,I,L,V • G
Figure 12. Diagrammatic representation of the amino acid sequence of the human acidic PRP:
phosphoprotein PIF -so The break shown in the amino acid sequence betweeen position 51
and 52 [»»>1 illustrates the amphipathic nature of the molecule, with a strongly acidic N-
terminus (-50 amino acids) and the proline-rich C-terminus (-100 amino acids).
phoprotein PIF-s (Fig. 12) has a strongly hydrophilic "head" segment from the
N-terminus to position 51 [»»>] containing two phosphoserines and fourteen
amino acids with a carboxyl side chain (aspartic, glutamic). Conversely, the
remainder of the amino acid sequence to the C-terminus is rich in the amino
acids proline, glycine, asparagine, and glutamine and, in comparison to the N-
terminus, is relatively apolar and hydrophobic. The protein may be likened to a
detergent molecule with a charged, hydrophilic "head" and a non-polar "tail."
Kauffman et al. 45 similarly reported the isolation of eleven basic proline-
rich proteins from the parotid gland of a single individual. All contain segments
with identical or similar amino acid sequences. Protein IB-7 (59 amino acids)
contains two contiguous 21 residue segments (1-22 and 22-42) that are virtually
identical. These segments are followed by a 12 amino acid sequence which is
identical to residues 2-13 and 23-34 (Fig. 12). The contiguous repeat pattern
1-54 recurs in other human parotid basic proline-rich proteins. Such molecules
may similarly be thought of as amphipathic, but in this case the strongly
hydrophilic side chains (lysine and arginine) are not grouped but distributed
throughout the amino acid sequence (Fig. 13).
Similar precise structural information and hence relative molecular masses
of the GPRPs are not yet available. However, human BPRPs and GPRPs are
encoded by a series of 4 genes; each gene gives rise to a precursor protein which
through proteolytic cleavage gives rise to secreted BPRPs and GPRPs of varying
sizes. The amino acid sequences of some GPRPs are known from genetic studies.
They are similar to those of BPRPs and contain a recognition sequence for N-
glycosylation at asparagine. This similarity suggests that these molecules also are
amphipathic, but, until more is known concerning the distribution of the
hydrophilic sugars on the proline-rich amino acid sequence, direct comparison
with either the structures of the BPRPs or APRPs is not possible.
Pre-eminent amongst the proteins that bind polyphenols most strongly are
the salivary pRPS.12 Basic PRPs complex with and precipitate polyphenols, and
.•••••.
1
0.~
.••• ..
O •• ·0.~ •••• ~ •• ~~.~
.•••• .••• ~.~
·0.~
~
~.
~
.0 .p
O~<90<9
59
~ Q.N
o K,R
Figure 13. Diagrammatic representation of the amino acid sequence of the human basic PRP:
187.
recent observations suggest that both acidic PRPs and glycosylated PRPs are fully
capable of forming soluble complexes with tannins (cf Fig. 1) but ones which
are not readily precipitated by virtue of the strongly solubilizing groups present
elsewhere on the peptides. Complexation is presumed to occur selectively (as with
the BPRPs) in the relatively hydrophobic proline-rich regions of the peptides.
The suggestion that the role of these proteins is as a "first line of defense"
against the detrimental effects of polyphe no Is in the diet of herbivores is a natural
and attractive one. 46 Based on his own and other's work, Bate-Smith42 first suc-
cinctly stated the presumed role of vegetable tannins in plant defense:
From the biological point of view the importance of tannins in plants lies in their effectiveness
as repellants to predators, whether animal or microbial.
In either case, the relevant property is astringency, rendering the tissues

unpalatable by precipitating proteins or by immobilizing enzymes, impeding inva-
sion of the host by the parasite. Twenty five years later it is apposite to ask "What
physical picture now emerges of the nature of the astringent response?" Since
astringency results from the loss oflubrication in the mouth, it is pertinent to con-
sider first how lubrication is itself achieved in the oral cavity. It has long been
held that lubrication is associated with and saliva derives its inherently viscous
nature from the presence (-3.Smg./ml.) of thread-like molecules such as
mucopolysaccharides and the glycosylated PRPs (Fig. 14a). It is presumed that
such molecules adopt extended random coil conformations in solution and, prob-
ably by non-covalent interactions (e.g. hydrogen bonding), form a tribological
layer over the exposed surfaces of the palate (Fig. 14a). The strongly hydrophilic
carbohydrate groups attached to the glycosylated PRPs (Fig. 14c) may be envis-
aged to contribute not only to this overall structural process but also, by virtue
of their approximate orthogonal relationship to the longitudinal peptide chains
(similar to the hairs on the root of a plant), to place additional limits on the lateral
compressibility of the PRPs as flow across the surfaces of the palate occurs. Their
presence in the tribological layer would thus be expected to enhance lubrication
in the palate.
Viscosity is a measure of the rate of energy dissipation in flow and for a
macromolecule depends critically upon its shape and asymmetry, and its hydro-
dynamically effective volume. 47 This last quantity must necessarily include not
only the volume of the anhydrous material but also the solvent that is closely
associated with the macromolecule. One view (developed in association with Pro-
fessor A.J. Ryan, University of Sheffield) of the origins of astringency takes note
of that fact. It is suggested that astringency derives directly from a collapse of
the tribological layer, under the influence of the astringent principles, to a release
of bound solvent (water molecules) into the surrounding medium, and hence to
a loss of lubrication. In the case of natural polyphenols, this collapse would
be engineered by intra- and intermolecular non-covalent cross linking of the
salivary PRPs (Fig.14b), by mechanisms such as those described earlier. Although
(a) surface of oral cavity
triboJogicaJJayer
(b) sUrface of oral cavity
poJyphenoJs
( c ) gJycosyJated PRPs
,gJycosyJ residues
"
Figure 14. Polyphenol: PRPs interactions; Lubrication and Astringency. (a)-PRP molecules
in solution: random coil conformations giving rise to a tribological layer in the palate;
(b )-collapse of the tribological layer and release of water giving rise to astringency; (c)-
glycosylated PRPs.
precipitation of polyphenol: protein complexes may occur, ultimately, this is

not a necessary condition for the generation of the astringent response. The
critical feature is the collapse of the tribological layer in the palate. Since the
whole process is envisaged as a reversible equilibrium (Fig. I), the sensation of
astringency would ultimately be lost as fresh liquids and saliva enter the oral
cavity. Equally it would be expected that the development of the astringent sen-
sation would be a time dependent one. This is not only because saliva is viscous
and there would be a time delay before the necessary intermolecular collisions
occur but also because of the very nature of polyphenol/protein interactions them-
selves.
The principles which underlie molecular recognition phenomena may be
analyzed not only in terms of the composition, structure, and conformation of
the substrates taking part in the complexation reactions but also in terms of three
idealized concepts. 48 "Die-mould" (jig-saw) matching is essentially static with an
exact fit of donor and acceptor molecules together. "Key-IocR' matching is time
dependent, since the key (donor) invariably has to be maneuvered into the lock
(acceptor) to achieve the correct fit. Finally "hand-in-glove" matching of donor
and acceptor molecules is both time dependent and dynamic. Both donor and
acceptor molecules are mobile and flexible and may assume a variety of subtly
different shapes as complexation proceeds. In such situations, it is important
that both substrates are able to sample a variety of different relative orientations
with respect to each other such that ultimately the maximum number of
strong contacts are made between donor and acceptor species. Such associative
reactions frequently exhibit strong cooperative effects. Present evidence, sum-
marized above, suggests that where polyphenol-protein complexation is at its
most effective it is largely of the "hand-in-glove" type where polyphenol and
protein both possess considerable conformational mobility and structures which
permit the possibility of bringing about multi-dentate complexation. If this view
is accepted, this would also predict a delayed time dependent astringent response
in the palate.
Finally, brief comment should be made on the role of low molecular mass
phenolic compounds, such as (-)-epicatechin, (+)-catechin, (-)-epigallocatechin-
3-0-gallate, and chlorogenic acid, in the development of astringency in the palate.
The evidence from taste panels shows that the threshold values for the percep-
tion of these compounds as astringents in the palate is relatively high, cf Table
4. Work such as that of Englehardt et al. 49 suggests nevertheless that where the
concentration of these compounds in a liquor is high then the liquor is generally
perceived as mildly astringent. A priori it seems unlikely that the explanation is
as suggested for polyphenols in Figure 14, namely direct cross-linking of the
PRPs. The binding of simple phenols to proteins is weak (cf Table 3) and the
principal (if not sole) binding site(s) appears to be the o-dihydroxy, o-trihydroxy
"B'phenolic nuclei. The participation of such substrates in direct cross-linking
reactions seems, for steric reasons, also to be doubtful. Accordingly, any expla-
nation must center upon their relatively high concentrations in the liquor(s) and
the hypothesis that under such conditions the equilibrium (Fig. I) between phe-
nolic compound and PRP is driven towards the intermolecular complex. Coating
of the surface (either wholly or partially) of the PRP would lead to a relatively
hydrophobic phenolic layer. In turn, this would lead to aggregation and to the
release of solvent (water) to the bulk medium. It is interesting to note in this
context that McManus et al. 11 observed that simple phenols, such as pyrogallol,
may indeed induce precipitation of globular proteins from solution so long as the
phenol can be maintained at a sufficiently high initial concentration in the aqueous

medium.
POSTSCRIPT
Commenting almost thirty years ago on the state of the art of natural
product chemistry one of its most distinguished practitioners, the late T.A.
Geissman,5o wrote bravely, but percipiently:
Increasingly attention will be paid to the role of organic compounds in plants and less attention
to what they are. Certainly structures are important but the determination of structure in itself
is ceasing to be of much interest or importance, and often turns out to be an exercise in the
manipulative and interpretive skill of the investigator. ... The future of phytochemistry is to
use the chemical information as the starting point for inquiry into questions that lie in the realm
of biology.
The story of the chemistry of plant polyphenols as it has unfolded over the
past three decades amply illustrates the truth of Geissman's shrewd observations.
For, whilst structural problems still remain to be solved, attention has turned
increasingly to the questions which are posed by the presence of these complex
metabolites in plant materials. These range from those of metabolism, through
ecology to the practical and applied problems which ensue in food science and
medicine.
To plagiarize the poet John Donne it is certainly a truism to say that in
nature "no molecule is an Island. entire of itself" Many of the intrinsic proper-
ties of plant polyphenols are related to their propensity to associate with other
molecular species. Most notable are those which occur with proteins, and under-
standing the molecular mechanisms involved remains a major scientific challenge
for the 21 st century. The study of the molecular basis of polyphenol-protein inter-
actions described above is one such contribution. It is not only of intrinsic impor-
tance to the wider questions of molecular recognition but it is also of considerable
practical significance to the foodstuffs and beverages industry where the modifi-
cation and control of the astringent response is often a much sought after objec-
tive. Knowledge in this area is similarly important in the increasing use of plant
materials containing polyphenols as sources of anti-oxidants, and as herbal med-
icines. In such cases, the palate, with its associated PRPs, may indeed be seen as
a potential barrier which such potential medicines must first surmount before they
may be absorbed into the plasma.
REFERENCES
1. HASLAM, E. 1997. Vegetable tannage: Where do the tannins go? J.Soc.Leather Tech and
Chemists 81: 45-51.
2. HASLAM, E. 1998. Practical Polyphenolics. From Structure to Molecular Recognition

and Physiological Action, Cambridge University Press: Cambridge, 422 p.
3. DAVY, H. 1803. An account of some experiments and observations on the constituent
parts of some astringent vegetables and on their operation in tanning (from Phi!.Trans.,
1803). In: Incunabula of Tannin Chemistry (M. Nierenstein, ed.), Edward Arnold: London
(1932), 116 p.
4. HASLAM, E., LILLEY, T.H., WARMINSKI, E., LIAO, H., CAl, Y., MARTIN, R.,
GAFFNEY, S.H., GOULDING, P.N., LUCK, G. 1992. Polyphenol complexation-a
study in molecular recognition. In: Phenolic Compounds in Food and Their Effects on
Health. I. Analysis, Occurrence and Chemistry. (C.-T. Ho, c.Y. Lee, and M.-T. Huang,
eds), A.C.S. Symposium Series 506, American Chem.Soc.: Washington D.C, pp. 8-50.
5. HASLAM, E. 1989. Plant Polyphenols. Vegetable Tannins Re-visited, Cambridge
University Press: Cambridge, 230 p.
6. MCMANUS, lP., DAVIS, K., BEART, lE., GAFFNEY, S.H., LILLEY, T.H., HASLAM,
E. 1985. Polyphenol interactions. Part I. Introduction. Some observations on the
reversible complexation of polyphenols with proteins and polysaccharides.
lChem.Soc.(Perkin Trans.2), 1429-1438.
7. CAl, Y., HASLAM, E. 1994. Plant polyphenols (vegetable tannins): Gallic acid metabo-
lism. Natural Product Reports (Royal Soc.Chem.), II: 41-66.
8. BAXTER, N.l, LILLEY, T.H., HASLAM, E., and WILLIAMSON, M.P. 1997. Multiple
interactions between polyphenols and a salivary proline-rich protein repeat result in
complexation and precipitation. Biochemistry 36: 5566-5577.
9. HATANO, T., HEMINGWAY, R.W 1996. Association of (+)-catechin and catechin-(4a-
8)-catechin with oligopeptides. J.Chem.Soc.(Chemical Communications), 2537-2538.
10. MURRAY, N.J., WILLIAMSON, M.P., LILLEY, T.H., HASLAM, E. 1994. Study of the
interaction between proline-rich proteins and a polyphenol by IH NMR spectroscopy.
Eur.J.Biochem. 219: 923-935.
II. MCMANUS, J.P., DAVIS, K., LILLEY, T.H., HASLAM, E. 1981. The Association of
Proteins with Phenols. J.Chem.Soc.Chemical Communications, 309-311.
12. BUTLER, L.G., HAGERMAN, A.E. 1981. The specificity of proanthocyanidin-protein
interactions. J.Bio!.Chem. 256: 4494-4497.
13. LUCK, G., LIAO, H., MURRAY, N.J., GRIMMER, H.R., WARMINSKI, E.E.,
WILLIAMSON, M.P., LILLEY, T.H., HASLAM, E. 1994. Polyphenols, astringency and
proline-rich proteins. Phytochemistry 37: 357-371.
14. CHARLTON, A.J., BAXTER, N.J., LILLEY, T.H., HASLAM, E., MCDONALD, C.J.,
WILLIAMSON, M.P. 1996. Tannin interactions with a full-length human salivary
proline-rich protein display a stronger affinity than with single proline-rich repeats. FEBS
Letters 382: 289-292.
IS. HASLAM, E. 1996. Natural polyphenols (vegetable tannins) as drugs:Possible modes of
action. 1Nat.Prod. 59: 205-215.
16. PERUTZ, M.E and LEVITT, M. 1988. Aromatic rings act as hydrogen bond acceptors.
1Mol.Biol. 201: 751-754.
17. HUNTER, C.A., HANTON, L.R., PURVIS, D.H. 1992. Structural consequences of a
molecular assembly that is deficient in hydrogen-bond acceptors. 1Chem.Soc.Chemical
Communications, 1134-1136.
18. HAGERMAN, A.E. 1989. Chemistry of tannin-protein complexation. In: Chemistry and
Significance of Condensed Tannins (R.W. Hemingway and I I Karchesy, eds.), Plenum
Press: New York and London, pp. 308-333.
19. HESTRIN, S., FEINGOLD, D.S., SCHRAMM, M. 1955. Enzymes of carbohydrate
metabolism. Hexoside hydrolases. Methods in Enzymo!. 1: 231-257.
20. MEJBAUM-KNATZELLENBOGEN, W, DOBRYSZYCHKA, WM., JAWORSKA, K.,
MORAWIECKA, B. 1959. Regeneration of protein from insoluble protein-tannin com-

pounds. Nature 184: 1799~ 1800.
21. MEJBAUM-KNATZELLENBOGEN, W, DOBRYSZYCHKA, WM. 1962. Immuno-
chemical properties of the serum proteins after regeneration from the protein-tannin
compounds. Nature 193: 1288~1289.
22. GUSTAVSON, K.H. 1956. The Chemistry of Tanning Processes. Academic Press:
London and New York.
23. HULME, A.c., JONES, 1.D., WOOLTORTON, L.S.C. 1965. Use of poly (vinylpyrroli-
done) in the isolation of enzymes from apple peel. Phytochemistry 4: 659--676.
24. GOLDSTEIN, 1.L. and SWAIN, T. 1965. The inhibition of enzymes by tannins. Phyto-
chemistry 4: 185~192.
25. JONES, WT. and MANGAN, 1.L. 1977. Complexes of the condensed tannins of sainfoin
(Onobrychis viciijhlia) with fraction I leaf protein and with submaxillary mucoprotein
and their reversal by polyethylene glycol and pH. 1.Sci.Food Agric. 28: 126~136.
26. JEFFERY, G.A. 1969. Water structure in organic hydrates. Acc.Chem.Res. 2: 344-352.
27. FRANK, H.S. 1970. The structure of ordinary water. Science 169: 635~641.
28. FRANK, H.S., EVANS, M.W 1945 Free volume and entropy in condensed systems. Ill.
Entropy in binary liquid mixtures; partial molal entropy in dilute solutions; stucture and
thermodynamics in aqueous electrolytes. 1.Chem.Phys. 13: 507~532.
29. EISENBERG, D., KAUZMANN, W 1969. Structure and Properties of Water. Oxford
University Press: Oxford, London.
30. ELLIS, S.c., PANKHURST, K.G.A. 1953. The interaction of tanning materials with col-
lagen monolayers. Disc.Faraday Soc. 16: 170~ 179.
31. GUSTAVSON, K.H., HOLM, B. 1952. The reactions of poly ami des with tanning agents.
1.Amer.Leather Trades Chem.Assoc. 47: 700-711.
32. JENCKS, WP. (1969). Catalysis in Chemistry and Enzymology, McGraw-Hill: New York,
393 p.
33. HOFF, 1.E., OH, H.I., ARMSTRONG, G.S., HAFF, L.A. 1980. Hydrophobic Interaction
in Tannin-Protein Complexes. 1.Agric.Food Chern. 28: 394~398.
34. HENDICKSON, WA., TEETER, M.M. 1981. Structure of the hydrophobic protein
crambin determined directly from the anomalous scattering of sulphur. Nature 290:
107~113.
35. TEETER, M.M. 1984. Water structure of a hydrophobic protein at atomic resolution: Pen-
tagon rings of water molecules in crystals of crambin. Proc.NatI.Acad.Sci.U.S.A. 81:
6014~6018.
36. TEETER, M.M., WHITLOW, M.D. 1987. Hydrogen bonding in the high resolution struc-
ture of the protein crambin. Trans.Am.Cryst.Assoc. 22: 75~88.
37. NEMETHY, G., SCHERAGA, H.A. 1962. Structure of water and hydrophobic bonding
in proteins. I. A model for the thermodynamic properties of liquid water. J.Chem.Phys.
36: 3382~3400.
38. NEMETHY, G., SCHERAGA, H.A. 1962. Structure of water and hydrophobic bonding
in proteins. II. A model for the thermodynamic properties of aqueous solutions of hydro-
carbons. 1.Chem.Phys. 36: 3382~3400.
39. PERUTZ, M.F. 1992. What are enzyme structures telling us? Faraday Disc. 93: 1~11.
40. HANSCH, c., FUJITA, T., IWASA, 1. 1964. A new substituent constant, 11:, derived from
partition coefficients. 1.Amer.Chem.Soc. 86: 5175~5180.
41. HANSCH, c., KIEHS, K., LAWRENCE, G.L. 1965. The role of substituents in the
hydrophobic bonding of phenols by serum and mitochondrial proteins. 1.Amer.Chem.Soc.
87: 5770~5773.
42. BATE-SMITH, E.C. 1973. Haem analysis of tannins-the concept of relative astringency.
Phytochemistry, 12: 907~912.
318 E. HASLAM et af.
43. BENNICK, A. (1982). Salivary proline rich proteins. MoI.Cell.Biochem. 45: 83-99.
44. HAY, D.1., BENNICK, A., SCHLESINGER, D.H., MINAGUCHI, K.,
MADAPALLITAM, G., SCHLUCKEBIER, S.K. 1988. The primary structures of six
human salivary acidic proline-rich proteins (PRP-I, PRP-2, PRP-3, PRP-4, PIF-s and
PIF-f). Biochem.1. 255: 15-21.
45. KAUFFMAN, D.L., BENNICK, A., BLUM, M., KELLER, P.1. 1991. Basic proline-rich
proteins from human parotid saliva: Relationships of covalent structures of ten proteins
from a single individual. Biochemistry 30: 351-3355.
46. MCARTHUR, c., SANSON, G.D., BEAL, A.M. 1995. Salivary proline-rich proteins
in mammals: Roles in oral homeostasis and counteracting dietary tannin. J.Chem.Ecol.
21: 663-691.
47. VAN HOLDE, K.E.1971. Physical Biochemistry, Prentice Hall: New Jersey.
48. WILLIAMS, R.J.P. 1988. Lecture, Biochemical Society, Sheffield, April 15 th •
49. ENGLEHARDT, U.H., KUHR, S., DING, Z. 1992. Influence ofcatechins and theaflavins
on the astringent taste of black tea brews. Z. Lebensm. Unters.Forsch. 195: 108-IlI.
50. GEISSMAN, T.A. 1972. In: Recent Advances in Phytochemistry (VC.Runeckles and
J.E.Watkin, eds.), volume 4, p. 303.
Chapter Twelve
PHYTOCHEMISTRY OF BRYOPHYTES
Biologically Active Terpenoids and Aromatic Compounds
from Liverworts
Yoshinori Asakawa
Faculty of Pharmaceutical Sciences

Tokushima Bunri University, Yamashiro-cho
Tokushima 770-8514, Japan
Introduction ................................................... 320

Biological Activity .............................................. 320
Characteristic Scents .......................................... 320
Pungency and Bitterness ....................................... 325
Allergenic Contact Dermatitis .................................. 328
Cytotoxic Activity and Anti-HIV-l and DNA Polymerase ~
Inhibitory Activity .......................................... 329
Antimicrobial and Antifungal Activity ........................... 332
Insect Antifeedant Activity, Mortality, and Nematocidal Activity ....... 332
Superoxide Anion Radical Release Inhibitory Activity ............... 333
5-Lipoxygenase, Calmodulin, Hyaluronidase, and Cyclooxygenase
Inhibitory Activity .......................................... 333
Piscicidal and Plant Growth Inhibitory Activity .................... 335
Neuritic Sprouting Activity .................................... 335
Muscle Relaxing Activity ...................................... 335
Cathepsin L and Cathepsin B Inhibitory Activity .................... 337
Cardiotonic and Vasopressin Antagonist Activity .................... 337
Conclusion ................................................... 339
319
320 YASAKAWA
INTRODUCTION
The bryophytes are placed taxonomically between algae and pteridophytes;

there are about 24,000 species in the world. They are further divided into three
classes, Musci (mosses 14,000 species), Hepaticae (liverworts 6,000 species) and
Anthocerotae (hornworts 300 species). The Hepaticae contain oil bodies which
are easily extracted with organic solvents, while the other two classes do not. A
number of bryophytes have been used in North America, China, and Europe to
cure bums, bruises, external wounds, etc. The mosses and liverworts shown in
Table 1 are medicinal plants and said to possess various biological activities. I- 5
Some bryophytes show characteristic fragrant odors and have an intense
hot and bitter or saccharine-like taste. Generally, bryophytes are not damaged by
insects, snails, slugs, and other small animals. Furthermore, some liverworts
cause intense allergenic contact dermatitis and have allelopathic activity. We
have been interested in the bioactive substances found in bryophytes and have
studied about 800 species of Hepatcicae collected in North and South America,
North and South Africa, India, Pakistan, Nepal, Taiwan, Australia, New Zealand,
Europe, and Japan with respect to their chemistry, pharmacology, and application
as sources of cosmetics and medicinal or agricultural drugs. It has been demon-
strated that most of the Hepaticae contain mainly mono- sesqui- and diterpenoids
and lipophilic aromatic compounds (bibenzyls, bis-bibenzyls, benzoates, cinna-
mates, long chain alkyl phenols, naphthalenes, phthalides, and isocoumarins)
which constitute the oil bodies. The biological activities of liverworts are due to
these substances, and at present over 300 new compounds have been isolated and
their structures elucidated.6-10 The biological characteristics of the terpenoids and
aromatic compounds isolated from the liverworts in our laboratory are: 1) char-
acteristic scents, 2) pungency and bitterness, 3) allergenic contact dermatitis, 4)
cytotoxic activity and anti-HIV-l and DNA polymerase ~ inhibitory activity, 5)
antimicrobial and antifungal activity, 6) insect antifeedant activity, mortality, and
nematocidal activity, 7) superoxide anion radical release inhibitory activity, 8)
5-lipoxygenase, calmodulin, hyaluronidase, and cyclooxygenase inhibitory activ-
ity, 9) piscicidal and plant growth inhibitory activity, 10) neurite sprouting activ-
ity, 11) muscle relaxing activity, 12) cathepsin L and cathepsin B inhibitory
activity, and 13) cardiotonic and vasopressin antagonist activity.
The chemical structures of terpenoids and aromatic compounds found in
several liverworts and some of their biological activities, including characteristic
odors and taste, will be surveyed in this paper.
BIOLOGICAL ACTIVITY
Characteristic Scents
Liverworts emit volatile terpenoids or simple aromatic compounds when
crushed. These are responsible for intense sweet-woody, intense turpentine, sweet-
PHYTOCHEMISTRY OF BRYOPHYTES 321
Table 1. Medicinal bryophytes: Their biological uses and effects!!!

Musci:
Bryum argenteum Antidotal, antipyretic, antirhinitic activity; for bacteriosis
Cratoneuron filicinum For malum cordis (heart disease)
Ditrichum pallidum For convulsions, particularly in infants
Fissidens japonicum Diuretic activity; for growth of hair, burns, and choloplania
(jaundice, icterus)
Funaria hygrometrica For hemostatis, pulmonary tuberculosis, vomitus cruentus
(hematemesis), bruises, and athlete's foot dermatophytosis
(dermatomycosis, dermomycosis)
Haploc/adium catillatum Antidotal, and antipyretic activity; for adenopharyngitis,
pharyngitis, uropathy, mastitis, erysipelas, pneumonia,
urocystitis, and tympanitis
Leptodictyum riparium Antipyretic; for choloplania, and uropathy
Mnium cuspidatum For hematostasis, and nosebleed
Dreas martiana For anodyne (pain), hemostasis, external wounds, epilepsy,
menorrhagia, and neurasthenia (nervosism, nervous
exhaustion)
Philonotis fontana Antipyretic, and antidotal activity; for adenopharyngitis
Plagiopus oederi As a sedative; for epilepsy, apoplexy, and cardiopathy
Polytrichum species Diuretic activity; for growth of hair
Polytrichum commune Antipyretic and antidotal; for hemostasis, cuts, bleeding from
gingivae, hematemesis, and pulmonary tuberculosis
Rhodobryum giganteum Antipyretic, diuretic, and antihypertensive; for sedation,
neurasthenia, psychosis, cuts, cardiopathy, and expansion of
heart blood vessels
Rhodobryum roseum As a sedative; for neurasthenia, and cardiopathy
Taxiphy//um taxirameum Antiphlogistic; for hemostasis, and external wounds
Weissia viridula Antipyretic, and antidotal; for rhinitis
Hepaticae:
Conocephalum conicum Antimicrobial, antifungal, antipyretic, antidotal activity; used to
cure cuts, burns, scalds, fractures, swollen tissue, poisonous
snake bites, and gallstones
Frullania tamarisci Antiseptic activity
Marchantia polymorpha Antipyretic, antihepatic, antidotal, diuretic activity; used to cure
cuts, fractures, poisonous snake bites, burns, scalds, and
open wounds
Reboulia hemisphaerica For blotches, hemostasis, external wounds, and bruises
mossy, fungal-like, carrot-like, mushroomy, or seaweed-like scents. 7•1! In Table 2,

some liverworts possessing such characteristic odors are listed.
Almost all liverworts which smell of mushrooms contain 1-octen-3-ol and
its acetate. A small thalloid liverwort, Asterella species, grown in Pulau Dayang
Bunting island, Malaysia emits an intense unpleasant odor which is due to
skatole.!2 The stink bug smell of Cheilolejeunea pallidus is attributable to
(E)-dec-2-enal, (Z)-dec-2-enal, (E)- and (Z)-pent-2-enals. 13 The characteristic
fragrance of Leptolejeunea elliptica is due to p-ethylanisol, p-ethylphenol, and
322 Y.ASAKAWA
Table 2. Characteristic odors of liverworts 7,11

Species Odor
Asterella species Indole- or skatole-like and higher plant Houttuynia
cordata-like
Bazzania japonica Sweet, balsamic, tree moss-like
B,pompeana Oak moss-like
Cheilolejeunea imbricata Strong milky smell
Chiloscyphus pallidus Intense smell reminiscent of stink bug
Conocephalum conicum Camphoraceous, strong mushroomy, and lactone-like
Conocephalum japonicum Higher plant Houttuynia cordata-like
Frullania davulica Mossy
F tamarisci subsp, tarmarisci Oak moss-like
Jungermannia obobata Carrot-like
(Solenostoma obovata)
Leptolejeunea elliptica Intense and sweet phenol-like
Mixed smell of naphthalene, and dried bonito
Lophocolea heterophylla, Strong and distinctly mossy
L. bidentata Strong and distinctly mossy
L. minor Strong moss-like
Lophozia bicrenata Pleasant odor like cedar oil
Makinoa crispata Rooty, earthy, woody, and amber-like
Mannia fragrans Strong sweet mossy
Odontoschisma denudatum Civet, animal-like
Pellia endiviifolia Dried seaweed-like
Plagiochila sciophila Sweet mossy and woody
(P acanthophylla subsp. japonica)
Porella gracillima Woody-earthy
P vernicosa Malty, earthy
Radula perrottetii Castor-like, animal-like
Takakia lepidozioides Mixed smell of cinnamon and burnt wheat powder
Targionia hypophylla Sweet turpentine
Trichocolea tomentella Sulfur-like
Wiesnerella denudata Strong sweet mushroomy, green, and citrus
p-ethyl phenylacetate, 14 The strong milk-like fragrance of Cheiloejeunea imbri-

cata is due to a mixture of (R)-dodec-2-en-l ,5-olide (1) and (R)-tetradec-2-en-
1,5-olide (2) (See Table 3).14
Bicyclohumulenone (3) isolated from Plagiochila sciophila (P acantho-
phylla subsp. japonica) possesses an aroma reminiscent of a variety of scents
based on a strong woody note, resembling the odor of patchouli, vetiver, cedar
wood, iris, moss, and carnations. Tamariscol (4) from Frullania tamarisci subsp.
tamarisci, F tamarisci subsp. obscura, F nepaiensis, and F asagrayana similarly
possesses a remarkable aroma reminiscent of the woody and powdery green
notes of mosses, hay, costus, violet leaf, and seaweeds. Both compounds are
important in commerce, They are used as perfumes as such or as perfume com-
Table 3. List of compound names in figures

1. (R)-Dodec-2-en-I,S-olide 41. 6-Undecyl catechol
2. (R)- Tetradec-2-en-1 ,5-olide 42. Plagiochiline A 13-octanoate
3. Bicyciohumulenone 43. 14-Hydroxyplagiochiline A 13-2E,4E-
4. Tamariscol dodecadienoate
5. I-Hydroxy-I-(2-methyl-I-propenyl)- 44. Plagiochiline A 13-decanoate
cyclohexane 45. Marsupellone
6. 4-Hydroxy-4-methyl-cyclohex-2-en- 46. Acetoxymarsupellone
I-one 47. Riccardin A
7. Geosmin 48. Riccardin B
8. Grimaldone 49. Perrottetin E
9a. (-)-Polygodial 50. Perrottetin F
9b. (+)-Polygodial (Synthetic) 51. Marchantin A
9c. Warburganal 52. Marchantin A trimethyl ester
10. Diplophyllolide 53. Marchantin A triacetate
11. ent-7a-Hydroxydiplophyllolide 54. Marchantin B
12. Tulipinolide 55. Marchantin D
13. Plagiochiline A 56. Marchantin E
14. Plagiochiline I 57. Paleatin B
IS. Sacculatal 58. Pusilatin A
16. I P-Hydroxysacculatal S9. Pusilatin B
17. I a-Hydroperoxy-4a,5p- 60. Pusilatin C
epoxygermacra-IO(l4), II (13)-dien- 61. Pusilatin D
12,8a-olide 62. Tomentellin
18. I P-Hydroperoxy-4a,Sp- 63. Demethyltomentellin
epoxygermacra-I O( 14), II (13 )-dien- 64. Trichocolein
12,8a-olide 65. Plagiochilide
19. Plagiochilal B 66. Infuscaic acid
20. Furanoplagiochilal 67. Norpinguisone methyl ester
21. Gymnocolin A 68. Norpinguisone
22. Infuscaside A 69. Cyclomyitaylyl 3-caffeate
23. Infuscaside B 70. Bicyciogermacrenal
24. Infuscaside C 71. Herbertenediol
25. Infuscaside D 72. Isocuparene-3,4-diol
26. Infuscaside E 73. Perrottetianal A
27. Anastreptin A 74. Radulanin K
28. Barbilycopodin 7S. Perrottetin A
29. Scapanin A 76. Perrottetin D
30a. (+)-Frullanolide 77. 3-Hydroxy-5-methoxy-4-(3-methyl-2-
30b. (-)-Frullanolide butenyl)bibenzyl
31. 3-Undecylphenol 78. 3,4-Dihydroxy-2-(3-methyl-2-
32. 3-Tridecylphenol butenyl)bibenzyl
33. 3 -Pentadecy Iphenol 79. Radulanin A
34. 3-Heptadecylphenol 80. Radulanin C
35. 6-undecylsalicylic acid 81. Radulanin H
36. 6-Tridecyl salicylic acid 82. 3-hydroxybibenzyl
37. 6-Pentadecyl salicylic acid 83. 3,4'-Dimethoxybibenzy I
38. Potassium 6-undecyl salicylate 84. 3-Methoxy-3',4'-
39. Potassium 6-tridecyl salicylate methylenedioxybibenzyl
40. Potassium 6-pentadecyl salicylate 85. 3,4;3'4'-Dimethylenedioxybibenzyl
324 YASAKAWA
Continued
86. 3,5-Dihydroxy-2-geranylbibenzyl 96. Mastigophorene D
87. 2-Carboxy-3,4-dihydroxy-5-(3-methyl- 97. Plagiochin A
2-butenyl)bibenzyl 98. d- Tubocurarine
88. Labda-12,14-dien-7,8-diol 99. 7',8'-Dehydromarchantin A
89. Isoriccardin C 100. Isomarchantin C
90. Riccardin C 10 1. II-Epiporelladiolide
91. Lunularic acid 102. I I, I 3-Dehydroporelladiolide
92. Isopolygodial 103. Porellaolide
93. Isosacculatal 104. Porelladiolide
94. Mastigophorene A 105. 3a,4a-Epoxyporelladiolide
95. Mastigophorene B
ponents of the powdery ftoral-, oriental bouquet-, fantastic chypre-, fancy

violet-, and white rose-types in various cosmetics. The total 13-step synthesis of
(±)-tamariscol (4) has been accomplished by using commercially available p-
methoxyacetophenone. '5 After it had been shown that both the tertiary alcohol
and the 2-methyl-l-propenyl group attached to the cyclohexane ring oftamariscol
were both necessary for the characteristic scent of 4, thirteen mini-tamariscols
were synthesized by Grignard reactions of 2,7-dimethylcyclohexanone, 2-
methylcyclohexanone, 4-methylcyclohexanone, cyclohexanone, and cyclopen-
tanone with vinylmagnesium bromide, 2-methyl-l-propenylmagnesium bromide
and 2-methyl-2-propenylmagnesium bromide, respectively. I-Hydroxy-l-(2-
methyl-I-propenyl)-cyclohexane (5) had a sweet mossy aroma similar to that of
tamariscol itself.'6
There are three chemo-types of Conocephalum conicum. Types 1, 2, and 3
emit (-)-sabinene, (+)-bomyl acetate, and methyl cinnamate as the major com-
ponents, respectively, which are responsible for the characteristic odor of each
type. 17 Jungermannia obobata contains a tris-normonoterpene ketone, 4-hydroxy-
4-methylcyclohex-2-en-l-one (6) which possesses an intense carrot-like odor. '8 .'9
The strong and distinct mossy odor of Lophocolea heterophylla and L. bidentata
is due to a mixture of (-)-2-methylisobomeoI 20 and geosmin (7) (Tabacchi, R.,
10ulain, D., unpublished results). Geosmin, possessing a strong earthy-musty
odor, has also been found in in vitro cultured Symphyogyna brongniartii. 21 The
strong sweet mossy note of Mannia fragrans is attributable to the cuparene-type
sesquiterpene ketone, grimaldone (8).22 The sweet turpentine-like odor of Tar-
gionia hypophylla is due to a mixture of cis- and trans-pinocarveyl acetates. 23
The strong sweet-mushroomy scent of the ether extract of Wiesnerella
denudata is due to (+)-bomyl acetate and a mixture of the monoterpene hydro-
carbons, a-terpinene, ~-phellandrene, terpinolene, a-pinene, ~-pinene, and cam-
phene. 7 The odor of the steam distillate of W denudata is weaker than that of its
ether extract. The steam distillate contains nerol (14%), neryl acetate (27%), and
y-terpinene (31%), but the content of l-octen-3-ol (7%) and its acetate (2%) is
low (Asakawa, Y., unpublished results).
Pungency and Bitterness
Some genera of the Hepaticae produce intense pungent and bitter sub-
stances which exhibit interesting biological activities described in subsequent sec-
tions. Most North American liverworts contain unpleasant substances, some of
which taste like immature green pea seeds or pepper.24 Porella vernicosa complex
(P arboris-vitae, P fauriei, P gracillima, P obtusata subsp. macroloba, P roellii
and P vernicosa) contain very pungent substances, and Jamesoniella autumnalis
contains an intense bitter principle whose taste resembles that of the leaf of lilac
and Swertia japonica or the root of Gentiana scabra var. orientalis. The strong
hot taste of Porella vernicosa complex is due to (-)-polygodial (9a).6
We reported that two eudesmanolides, diplophyllolide (10), and ent-
7a-hydroxydiplophyll0lide (11), a germacranolide, tulipinolide (12), two 2,3-
secoaromadendrane-type sesquiterpene hemiacetals, plagiochiline A (13), and
plagiochiline I (14), and a sacculatane-type diterpene dialdehyde, sacculatal (15),
possessing an intense pungent taste, had been isolated from some Chiloscyphus,
Wiesnerella, Plagiochila, Pellia, and Trichocoleopsis species, respectively. 6
Further fractionation of the ether extract of Pellia endiviifolia resulted in the iso-
lation of a new pungent 1~-hydroxysacculatal (16), together with several saccu-
latane-type diterpenoids. 25 The hot taste of Pallavicinia levieri lO and Riccardia
robata var. yakushimensis 26 is due to sacculatal (15). Polygodial and sacculatal
have been obtained from cell suspension cultures from each Iiverwort.27.28 Porella
acutifolia subsp. tosana is a pungent stem-leafy liverwort. Its taste is due to the
presence of hydroperoxysesquiterpene lactones, 1a- (17), and 1~-hydroperoxy-
4a,5~-epoxygermacra-IO (l4),II(l3)-dien-12,18a-0Iides (18).29 When one
chews a whole plant of the stem-leafy liverworts, Plagiochila asplenioides, Pfru-
ticosa, P ovalifolia, and P yokogurensis which contain plagiochiline A (13), one
slowly feels a potent hot taste. It is suggested that 13 might be converted into a
pungent unsaturated dialdehyde by human saliva. Enzymatic treatment of 13 with
XH -
H9
~(5)
Y yP (6) (7)
(8)
Figure 1. Characteristic odorants.

326 Y.ASAKAWA
amylase in phosphate buffer or with human saliva produces two strong pungent
2,3-secoaromadendrane-type aldehydes, plagiochilal B (19) whose partial
structure is similar to that of the pungent drimane-type sesquiterpene dialdehyde,
polygodial (9a), and furanoplagiochilal (20).30
Most of the species belonging to the Lophoziaceae produce surprisingly
intense bitter substances. Gymnocolea inflata is persistently bitter and induces
vomiting when one chews a few leaves for several seconds. The earlier review
mentioned that this is due to gymnocolin A (21).6 Jungermannia infusca has an
intense bitter taste. This is due to the presence of the infuscasides A-E (22-26).31
The bitterness of Anastrepta orcadensis, Barbilophozia lycopodioides, and
Scapania undulata are attributable to anastreptin A (27),18,32 barbilycopodin
(28),18,32 and scapanin A (29),33 respectively.
Q:Y
CHO CHO CHO
WHO
,;'I HO
~HO
"
"
; :"'r ~.
(9a) (9b) (ge)
yq
::::,...
R
'0,
~
"DAC
(10) R=H (12) 0

(11) R=OH (13) R=Ac
(14) R=H
R CHO
~HO
(15) R=H
LA (17) (18)
(16) R=OH
OH~ r--
(19) (20)
Figure 2. Pungent sesqui- and diterpenoids.

?
Jik-oAC
W"""
J--:;i
(23) (24)
(25)
y;:C
~
' OA~
'~" :.
a
":. 0
"
(27) (28) (29)
Figure 3. Bitter diterpenoids.

328 Y.ASAKAWA
Allergenic Contact Dermatitis
Frullania species (Hepaticae) are notable as liverworts that cause allergenic

contact dermatitis. The allergy-inducing substances are sesquiterpene lactones,
(+)-frullanolide (30a) and (-)-frullanolide (30b) and their related a-methylene-
y-butyrolactones which have been isolated from Frullania dilatata and F.
tamarisci, respectively.6 F. asagrayana, F. bolanderi, F. eboracensis, F. francis-
cana, F. inflata, F. kunzei, F. nisqualiensis, F. riparia, and the other Fruliania
species which contain sesquiterpenes with a-methylene-y-butyrolactones cause
strong allergenic contact dermatitis as does Schistochila appendiculata. The aller-
gens of the latter are long chain alkylphenols, 3-undecyl- (31), 3-tridecyl (32), 3-
pentadecyl (33), and 3-heptadecyl phenols (34), long chain alkyl salicylic acids,
6-undecyl- (35), 6-tridecyl- (36), 6-pentadecyl salicylates (37), and their potas-
sium salts, potassium 6-undecyl- (38), 6-tridecyl- (39), and 6-pentadecyl salicy-
lates (40) as well as 6-undecyl catechol (41).7 Such dermatitis is similar to that
caused by the long chain alkylphenols of the fruit of Ginkgo biloba and Anacar-
diaceae. Marchantia polymorpha and Metzgeria furcata also cause allergenic
contact dermatitis but their allergens have not been isolated.
(30a)
ct2t (30b) °
(31) R=H
(35) R=C02 H
(38) R=C02 K
11
Q
~I
H
C 17 H33
(34)
13
(32) R=H
(36) R=C02 H
(39) R=C0 2K
15
11
Figure 4. Allergy inducing sesquiterpene lac tones and long chain alkyl phenols.
Cytotoxic Activity and Anti-HIV-l and DNA Polymerase P

Inhibitory Activity
A few eudesmanolides and germacranolide possessing inhibitory activity

against KB cells have been isolated from liverworts. 7 Conocephalum conicum and
Wiesnerella denudata contain guaianolides which exhibited cytotoxic activity
against P-388 lymphocytic leukemia. 7 The crude ether extract (4-20 Ilg/ml) of the
following liverworts showed cytotoxicity against P-388 in vitro (Asakawa, Y,
unpublished results): Bazzania pompeana, Kurzia makinoana, Lophocolea
heterophylla, Makinoa crispata, Marsupella emarginata, Pellia endiviifolia,
Plagiochila jruticosa, P ovalifolia, Porella caespitans, P japonica, P perrotte-
tiana, P vernicosa, and Radula perrottetii. On the other hand, Frullania diversi-
texta, F ericoides, F muscicola, F tamarisci subsp. obscura, Lepidozia vitrea,
Pallavisinia subciliata, Plagiochila sciophila, Supraceanthus semirepandus, and
Trocholejeunea sandvicensis were not active against P-388.
2,3-Secoaromadendrane-type sesquiterpenoids, plagiochiline A (13), pla-
giochiline A 13-octanoate (42), and 12-hydroxyplagiochiline A 13-2E,4E-
dodecadienoate (43) isolated from P ovalifolia showed cytotoxic activity (IDso 3,
O.OS, O.OSllg/ml, respectively) against P_388. 34 Polygodial (9a) isolated from
P vernicosa complex, sacculatal (15) from P endiviifolia, and two 2,3-
secoaromadendrane-type sesquiterpene hemiacetals (42), and plagiochiline A 13-
decanoate (44) from P ovalifolia showed cytotoxic activity (2-4llg/ml) against
melanoma. 35 Sacculatal (15) also showed cytotoxic activity against Lui cell (IC 50
S.7Ilg/ml), KB cell (3.2), LN Cap cell (7.6) and ZR-7S-1 cell (7.6), respectively
(Cordell, G. A., Pezzuto, 1. M., Asakawa, Y, unpublished results). Marsupellone
(45) and acetoxymarsupellone (46) from M. emarginata showed cytotoxicity
(IDso Illg/ml) against P388. 3S .36
Riccardins A (47) and B (48) from Riccardia multifida inhibited KB cells
at a concentration of 10 and 12Ilg/ml, respectively. Many Plagiochila species
and Radula perrottetii contained cytotoxic plagiochiline A (13) (0.28Ilg/ml) and
perrottetin E (49) (12.Sllg/ml) active against KB cell, respectively.
The thalloid liverwort, Marchantia polymorpha, which can cause allergenic
contact dermatitis, shows inhibitory activity against Gram-positive bacteria, and
has diuretic activity. The methanol extract (100 g) of Japanese M. polymorpha
was chromatographed on silica gel and Sephadex LH-20 to give cyclic bis-
bibenzyls, marchantin A (MA) (51, 30 g) and its analogues (MB-G). The yield of
MA (51) is dependent upon Marchantia species. 100 To 120g of pure MA has
been isolated from 2 kg of dried M. paleacea var. diptera. This thalloid liverwort
elaborates not only the marchantin series, marchantin A (51), B (54), D (55) and
E (56), but also the acyclic bis-bibenzyls, perrottetin F (50) and paleatin B (57).
Marchantins A, B, D, paleatin B, and perrottetin F show DNA polymerase ~
inhibitory (IDso 14.4-97.SIlM), cytotoxicity (3.7-20IlM against KB cell), and
anti-HIV-I activity (S.30-23.7Ilg/ml) (Cordell, G. A., Pezzuto, 1. M., Asakawa,
(45) R=H
(42) R'=Octanoyl, R2=H (46) R=OAc
(43) R'=2E,4E-Dodeca-
dienoyl, R2=OH
(44) R'=Decanoyl, R2=H
MeO H
(47) (49) R=H

(48)
(50) R=OH
(51) R'=R2=H (55) (56)

(52) R'=Me, R2=H
(53) R'=Ac,R 2=H
(54) R'=H, R2=OH
H H
(57) (58)
Figure 5. Sesquiterpenoids and cyclic and acyclic bis-bibenzyls possessing various biological
activities.
(59)
(60)
(61)
Me
~
(62)
o Me (63)
MeO ~ I
:::::,.. h
(64)
Figure 5. Continued.
332 Y.ASAKAWA
Y., unpublished results). Marchantin A (51) also shows cytotoxicity (TIC 117)
against P-388. 7 Blasia pusilla produces bis(bibenzyl) dimers, pusilatin A-D
(58-t;1). Pusilatins B (59) and C (60) possess DNA polymerase ~ inhibitory activ-
ity (lC50 13.0 and 5.l6IlM), moderate cytotoxicity against KB cell (ED5o 13.1 and
13.0Ilg/ml) and weak HIV-RT inhibitory activity.37 Trichocolea species produce
prenyl ethers, tomentellin (62), demethyltomentellin (63), and trichocolein (64).38
Compound 62 is the major cytotoxic component of T mollissima, active against
BSC cells at 151lg/disk.38 Compound 63 isolated from T tomentella also showed
the same activity.38
Antimicrobial and Antifungal Activity
Several liverworts, Bazzania species, Conocephalum conicum, Dumortiera

hirsuta, Marchantia polymorpha, Metzgeria furcata, Pellia endiviifolia, Pla-
giochila species, Porella vernicosa complex, P platyphylla, and Radula species
show antimicrobial activity.7 Several, such as Bazzania species, Conocephalum
conicum, Diplophyllum albicans, Lunularia cruciata, Marchantia polymorpha,
Plagiochila species, Porella vernicosa complex, and Radula species, display
antifungal activity. 7
Marchantin A (51) from many Marchantia species, M. chenopoda, M. poly-
morpha, M. paleacea var. diptera, M. piicata, and M. tosana, show antibacterial
activity against Acinetobacter calcoaceticus (MIC 6.25Ilg/ml), Alcaligenes
faecalis (100), Bacillus cereus (12.5), B. megaterium (25), B. subtilis (25),
Cryptococcus neoformans (12.5), Enterobacter cloacae, Escherichia coli, Proteus
mirabilis, Pseudomonas aeruginosa, Salmonela typhimurium (100), and Staphy-
lococcus aureus (3.13-25).7 They also have antifungal activity against Alternaria
kikuchiana, Aspergillus fumigatus (MIC 100 Ilg/ml), A. niger (25-100), Candi-
ada albicans, Microsprorum gypseum, Penicillium chrysogenum (100), Piricu-
laria oryzae (12.5), Rhizoctonia solani (50), Saccharomyces cerevisiae,
Sporothrix schenckii (100), and the dermatophytes Trichophyton mentagrophytes
(3.13) and T rubrum (100).7 The prenyl phenyl ethers (62 and 63) isolated from
Trichocolea mollissima and T tom en tella, respectively were mildly antifungal
against T mentagrophytes. 38 Compound 64 isolated from T lanata showed similar
mild antifungal activity. 38
Insect Antifeedant Activity, Mortality, and Nematocidal Activity

As mentioned earlier, plagiochiline A (13), found in several Plagiochila
species, is a strong antifeedant against the African army worm (Spodoptera
exempta).6 Compound 13 shows nematocidal activity against Caenorphabdiitis
elegans (lllllg/ml) (Asakawa, Y., unpublished results). The pungent sacculatal
(15) kills tick species Panonychus citri. Compound 15, eudesmanolides (10, 11)
from Chiloscyphus polyanthos, and gymnocolin (21) from Gymnocolea infiata
also have antifeedant activity against larvae of Japanese Pieris species. 7 A series
of natural drimanes and related synthetic compounds was tested for antifeedant
activity against aphids. 39 Polygodial (9a) from the Porella vernicosa complex and
warburganal (9c) from the African tree Warburgia ugandensis were the most
active substances. Natural (-)-polygodial (9a) and the synthetic (+)-enantiomer
(9b) showed similar levels of activity as aphid antifeedants. (-)-Polygodial killed
mosquito larvae at a concentration of 40 ppm and had mosquito repellent activ-
ity which is stronger than the commercially available DEET. Plagiochilide (65)
isolated from Plagiochila species killed Nilaparvata lugens (Delphacidae) at
lOOllg/ml (Asakawa, Y., unpublished results).
Superoxide Anion Radical Release Inhibitory Activity
Excess superoxide anion radical (0 2-) in organisms causes various

angiopathies, such as cardiac infarction and arterial sclerosis. Infuscaic acid
(clerod-3, 13(l6)-14-trien-17 -oic acid) (66) from Jungermannia infusca and pla-
giochilal B (19) inhibit the release of superoxide from rabbit PMN at IC so 0.07
and 6.0Ilg/ml, respectively, and from guinea pig peritoneal macrophage induced
by formyl methionyl leucyl phenylalanine (FMLP) at IC so 40llg/ml and 25.0
Ilg/ml, respectively.7 Norpinguisone methyl ester (67) from Porella elegantula
exhibits 50% inhibition of the release of superoxide from the guinea pig
peritoneal macrophage at 351lg/ml. The same activity (IC so 7.5Ilg/ml) has been
found in cyclomyltaylyl-3-caffeate (69) from Bazzaniajaponica. Other sesquiter-
penoids, plagiochilide (65) isolated from Plagiochila fruticosa, P. ovalifolia and
P. yokogurensis, norpinguisone (68) from Porella vernicosa, bicyclogermacrenal
(70) from Conocephalum conicum, herbertenediol (71) and isocuparene-3,4-diol
(72) from Mastigophola diclados, the diterpenoids, infuscaside A (22), and infus-
caside B (23) from Jungermannia in/usca, and perrottetianal A (73) from Porella
perrottetiana also inhibit superoxide release from guinea pig peritoneal macro-
phage (IC so 12.5-50 Ilg/ml). 7 Radulanin K (74) from Radula javanica inhibits the
release of superoxide anion radical from guinea pig macrophage (IC so 6llg/ml). 35
Polygodial (9a) and sacculatal (15) also show superoxide anion radical release
inhibition at 4.01lg/ml from guinea pig peritoneal macrophage (Asakawa, Y.,
unpublished results).
5-Lipoxygenase, Calmodulin, Hyaluronidase, and Cyclooxygenase

Inhibitory Activity
Marchantin A (51) from several Marchantia species showed 5-lipoxygenase

inhibitory activity: (89% at 1O-5 mol, 94% at 10-6 mol, 45% at 10-7mol, 16%
at 10-8 mol) against LTB3 (5S, 12R-dihydroxy-6,8, 10, 14-eicosatetraenoic acid);
(99% at 10-5 mol, 97% at 10-6 mol, 70% at 10-7 mol, 40% at 10-8 mol) against 5-
HETE (5-hydroxy-6,8,11,14-eicosatetraenoic acid); and calmodulin inhibitory
334 YASAKAWA
activity at ID50 1.85 )lg/ml. 7 Perrottetins A (75) and D (76) from Radula perrat-
tetii and prenyl bibenzyls (77-81), also from Radula species, riccardin A (47)
from Riccardia multifida, and marchantins D (55) and E (56) from Marchantia
species had calmodulin inhibitory activity (IDso 2.0-95.0 )lg/ml). 7 The simple
bibenzyls (82-85) from Radula and Frullania species also showed weak calmod-
ulin inhibitory activity (IDso 100 )lg/ml) as did the labdane-type diterpene diol,
labda-12, 14-dien-7 ,8-diol (88) (ID50 82 )lg/ml) isolated from Parella perrotte-
tiana. 7 Perrottetin A (75), prenylbibenzyls (76, 84, 86, 87), marchantins D (55)
and E (56), and riccardin A (47) also inhibited 5-lipoxygenase (76-4% at 10-6
mol).7 The following phenolic compounds showed significant cyc\ooxygenase
inhibitory activity: marchantin A (51) (lC 50 46.4 )lM), marchantin B (54) (55.9),
marchantin E (56) (58.0), paleatin B (57) (45.2), perrottetin D (76) (26.2), radu-
lanin H (81) (39.7), isoriccardin C (89) (50.8), and riccardin C (90) (53.5).40
(67) R 1=Me, R2 =C02Me

(65) (66)
(68) R 1 =R 2 =Me
H
~W
I "
"""
..........
b
~
0",
".,
"''''
".
". .
H (69) (70)
~
(71) (72)
0(7~:~ HO
(74)
Figure 6. Sesqui- and diterpenoids, and bibenzyls possessing superoxide anion radical release
inhibitory activity.
Lunularic acid (91), which is found in almost all liverworts as a minor com-
ponent, has anti-hyaluronidase activity (IC 5o 0.13 nM). This activity is stronger
than that of tranilast (N-3',4' -dimethoxycinnamoylanthranilic acid) which is an
anti-allergenic agent developed in Japan for oral administration. Lunularic acid
(91) has been obtained from hydrangenol-~-glucoside via hydrangenol in good
yield. 4l Perrottetin E (49) exhibited inhibitory activity fWAor thrombin (lC 50
I 811M) which is associated with blood coagulation. 42
Piscicidal and Plant Growth Inhibitory Activity
The strongest piscicides are the pungent (-)-polygodial (9a) from Porella
vernicosa complex and sacculatal (15) from Pellia endiviifolia, Pallavicinia
levieri, Riccardia robata var. yakushimensis, and Trichocoleopsis sacculata.
Killie-fish (Oryzia latipes) is killed within 2 hr by 0.4 ppm solution of9a and 15. 6,7
Sacculatal (15) and 1~-hydroxysacculatal (16) also kill killie-fish within 20 min
at I ppm?5 Killie-fish is also killed within 2 hr by a 0.4 ppm solution of synthetic
pungent (+)-polygodial (9b). Hence, piscicidal activity is not affected by the
chirality of polygodial. Polygodial is also toxic to fresh water bitterlings, which
are killed within 3 min by a 0.4 ppm solution. 7 On the other hand, isopolygodial
(92) from cultured cells of Porella vernicosa and the higher plant Polygonum
hydropiper and isosacculatal (93) from Pellia, Riccardia, and Trichocoleopsis
species show neither piscicidal nor molluscicidal activity even at 1O,000ppm.7
Almost all crude extracts from liverworts which contain bitter or pungent
substances show phytotoxic activity. (-)-Polygodial (9a) inhibits the germination
and root elongation of rice in husk at 100ppm. At a concentration less than
25 ppm, it dramatically promotes root elongation ofrice. 6A3
Neuritic Sprouting Activity
Mastigophorenes A (94), B (95), and D (96) from Mastigophora diclados

exhibit neurotrophic properties at 10-5-10- 7 M, greatly accelerating neuritic sprout-
ing and network formation in the primary neuritic cell culture derived from the
fetal rat hemisphere. 44 Plagiochilal B (19) and plagiochilide (65) from Plagiochila
fruticosa show not only acceleration of neurite sprouting but also enhancement of
choline acetyl transferase activity in a neuronal cell culture of the fetal rat cerebral
hemisphere at 10-5 M.lO Plagiochin A (97) also shows the same activity at 10-6 M.45
Two bitter diterpene glucosides, infuscasideA (22) and B (23), show neurite bundle
formation at 1O-7M (Asakawa, Y., unpublished results).
Muscle Relaxing Activity
Marchantin A (51) and the related cyclic bis-bibenzyls are structurally

similar to bis-bibenzyl-isoquinoline alkaloids such as d-tubocurarine (98) which
336 Y.ASAKAWA
OH OH OH
(77)
OH OR'
(79) R'=R 2=R 3 =H (82) R'=R2=H

(80) R'=R 2=H, R3 =OH (83) R'=Me, R2=OMe
(81) R'=R 3 =H, R2=COOH
OH
OH
OH
OH
(88)
OH
HO
OH
(91)
HO
(90)
Figure 7. Bibenzyls possessing 5-lipoxygenase, calmodulin, cyclooxygenase, and

hyaluronidase inhibitory activity.
are pharmacologically important muscle relaxing active drugs. Surprisingly,

marchantin A (51) and its trimethyl ether (52) also show muscle relaxing activ-
ity.9A6 Nicotine in Ringer solution effects maximum contraction of rectus
abdominis in frogs (RAF) at a concentration of 1O-6 M. After preincubation of
marchantin A trimethyl ether (52) (at a concentration of 2 x 10-7 - 2 x 10-4 M)
in Ringer solution, nicotine (10-8 - 10-4 M) was added. At a concentration of
1O-6 M, the contraction of RAF decreased by about 30%. d-Tubocurarine (98)
exhibits similar effects as does 52 with acetyl choline. 9A6 Although the mecha-
nism of action of marchantin A (51) and its methyl ether (52) in effecting muscle
relaxation is still unknown, it is interesting that these cyclic bis-bibenzyls,
possessing no nitrogen atoms, cause concentration dependent decrease of
contraction of RAF. Marchantin A and its trimethyl ether also had muscle relax-
ing activity in vivo in mice. MM2 calculations indicate that the conformation of
marchantin A and its trimethyl ether and the presence of an ortho hydroxyl group
in 51 and an ortho methoxyl group in 52 contribute to the muscle relaxing activ-
ity.9 Marchantin triacetate (53) and 7',8'-dehydromarchantin A (99) and acyclic
bis(bibenzyls), such as perrottetin E (49) and F (50) did not show any muscle
relaxing activity (A sakawa, Y., unpublished results).
Cathepsin L and Cathepsin B Inhibitory Activity
Cathepsin L is correlated with osteoporosis 47 and allergy.48 We are currently

searching for enzyme inhibitors from natural products to develop chemopreven-
tive drugs for these diseases. The marchantin series showed both cathepsins L
and B have inhibitory activity. Isomarchantin C (100) was the strongest inhibitor
against both enzymes (95% for cathepsin Land 93% for cathepsin B at 10-5M).
Infuscaic acid (66) was also active (63% and 32% at 10-5M) (Asakawa, Y., unpub-
lished results). The crude extract of Porella japonica showed potent inhibition of
cathepsins Band L. Biological activity guided fractionation gave three guaiano-
Iides, II-epiporelladiolide (101), 11,13-dehydroporelladiolide (102), and porel-
laolide (103), together with porelladiolide (104) and its epoxide (105). Only
compound (102) possessed a weak inhibitory activity against cathepsin B (13.4%
at 10-5M) and cathepsin L (24.7% at 10-5M).40
Cardiotonic and Vasopressin Antagonist Activity
Marchantin A (51) shows cardiotonic activity [increased coronary blood

flow (2.5 mllmin. at 0.1 mg)V Prenyl bibenzyl (78) from Radula perrottetii
indicates vasopressin antagonist activity (ID5o 27Ilg/ml). However, 2-
geranylbibenzyl (86) from the same liverwort did not show vasopressin antago-
nist activity. 7
338 YASAKAWA
(92)
(94)
(96)
Figure 8. Sesquiterpenoids, and cyclic bis-bibenzyl possessing neurite sprouting activity.
(99)
Figure 9. d-Tubocurarine and 7',8'-dehydromarchantin A.

CONCLUSION
Most of the compounds isolated from or detected in the Hepaticae are

lipophilic terpenoids (mono-, sesqui-, and diterpenoids) and aromatic com-
pounds, of which only a few nitrogen- or sulfur-containing compounds have been
found. 6,10 It is noteworthy that ca 80% of the sesqui- and diterpenoids found in
liverworts are the enantiomers of those found in higher plants. Mono- and
sesquiterpenoids have not been isolated from the Musci and Anthocerotae. At
present, only 5% of the total bryophytes have been studied chemically. Although
liverworts are a small group of plants, they contain a number of new terpenoids
and phenolic compounds, several of which show interesting biological
activity.
(100)
~
, -..;:
~
R ~
0 0 0
(101 ) (102) (103)
o
(104) (105)
Figure 10. Cyclic bis-bibenzyl, and guaiane-type sesquiterpene lactones possessing cathepsins
Band L inhibitory activity.
340 Y.ASAKAWA
ACKNOWLEDGMENTS
The author thanks Professor N. Katsunuma (TBU), Profs. 1M. Pezzuto, T.

Pengsupaarp and G.A. Cordell (The University of Illinois at Chicago), Otsuka
Pharmaceutical Co. Ltd. and Kirin Brewer Co. Ltd. and Takeda Pharmaceutical
Co. Ltd. for bioassay of the crude extracts and isolated compounds from liver-
worts. A part of this work was supported by a Grant-in-Aid of Cancer Research
from the Ministry of Health and Welfare, Japan.
REFERENCES
1. GARNIER, G., BEZANIGER-BEAUQUESNE, 1., DEBRAUX, G. 1969. Resources

medicinales de la £lore francaise. Vol. I: Vigot Freres Editeurs, Paris, pp. 78-81.
2. SUIRE, C. 1975. Les donnes actuelles sur la chimie des bryophytes. Rev. Bryol. et
Lichenol. 41: 105-262.
3. DING, H. 1982. Zhong guo Yao yun Bao zi Zhi wu. Kexue Jishu Chuban She,
Shanghai. pp. 1-409.
4. Wu, p.c. 1982. Some uses of mosses in China. The Bryol. Times. 13: 5.
5. ANDO, H., MATSUO, A. 1984. Applied bryology. In: Advances in Bryology, W
Schultze-Motel (ed.), Vol. 2: J. Cramer, Vaduz, pp. 133-224.
6. ASAKAWA, Y. 1982. Chemical constituents of Hepaticae. In: Progress in the Chemistry
of Organic Natural Products, W Herz, H. Grisebach and G.W Kirby (eds.), Vol. 42:
Springer, Wien, pp. 1-285.
7. ASAKAWA, Y. 1990. Terpenoids and aromatic compounds with pharmacological
activity from bryophytes. In: Bryophytes: Their Chemistry and Chemical Taxonomy,
D.H. Zinsmeister and R. Mues (eds.), Oxford University Press, Oxford, pp. 369-
410.
8. ASAKAWA, Y. 1990. Biological active substances from bryophytes. In: Bryophyte
Development: Physiology and Biochemistry, R.N. Chopra and S.c. Bhatia (eds.), CRC
Press, Boca Raton, pp. 259-287.
9. ASAKAWA, Y. 1993. Biologically active terpenoids and aromatic compounds from
liverworts and the inedible mushroom Cryptoporus volvatus. In: Bioactive Natural
Products: Detection, Isolation, and Structural Determination, S.M. Colegate and R.J.
Molyneux (eds.), CRC Press, Boca Raton, pp. 319-347.
10. ASAKAWA, Y. 1995. Chemical constituents of bryophytes. In: Progress in the
Chemistry of Organic Natural Products, W Herz, G.W Kirby, R.E. Moore, W Steglich
and C. Tamm, (eds.), Vol. 65: Springer, Wien, pp. 1-562.
II. MIZUTANI, M. 1975. Smell of some bryophytes. Misc. Bryol. et Lichenol. 6: 64.
12. ASAKAWA, Y., TOYOTA, M., TANAKA, H., HASHIMOTO, T., JOULAIN, D. 1995.
Chemical constituents of an unidentified Malaysian liverwort Asterella (?) species. J.
Hattori Bot. Lab. 78: 183-188.
13. TOYOTA, M., ASAKAWA, Y. 1994. Volatile constituents of the liverwort Chiloscyphus
pallidus (Mitt.) Engel & Schuster. Flavour and Fragr. J. 9: 237-240.
14. TOYOTA, M., KOYAMA, H., ASAKAWA, Y. 1997. Volatile components of the liverworts
Archilejeunea olivacea, Cheilolejeunea imbricata and Leptolejeunea elliptica. Phyto-
chemistry 44: 1261-1264.
15. TORI, M., SONO, M., NISHIGAKI, Y., NAKASHIMA, K., ASAKAWA, Y. 1991. Studies
on the liverwort sesquiterpene alcohol tamariscol. Synthesis and absolute configuration.

J. Chern. Soc. Perkin Trans. I: 435-445.
16. SONO, M. 1991. Structure and total synthesis of fragrant component, tamariscol from
Frullania tamarisci and its analogue from Conocephalum conicum, Ph. D. Thesis,
Tokushima Bunri University, I10p.
17. TOYOTA, M., SAITO, T., MATSUNAMI, J., ASAKAWA, Y 1997. A comparative study
on three chemo-types of the liverwort Conocephalum conicum using volatile constituents.
18. CONNOLLY, J.o. 1982. New terpenoids from the Hepaticae. Rev. Latinoamer. Quim.
12: 121-126.
19. CONNOLLY, J.D. 1990. Monoterpenoids and sesqui-terpenoids from the Hepaticae. In:
Bryophytes: Their Chemistry and Chemical Taxonomy, D.H. Zinsmeister and R. Mues,
(eds.), Oxford University Press, Oxford, pp. 41-58.
20. TOYOTA, M., ASAKAWA, Y, FRAHM, J.-P. 1990. Homomono- and sesquiterpenoids
from the liverwort Lophocolea heterophylla. Phytochemistry 29: 2334-2337.
21. SPORLE, J. 1990. Phytochemische Untersuchungen an ausgewahlten Panamaischen
Lebermoosen, Ph. D. Thesis, Universitat des Saarlandes. 90p.
22. HUNECK, S., CONNOLLY, J.D., FREER, A.A., RYCROFT, D.S. 1988. Grimaldone, A
tricyclic sesquiterpenoids from Manniafragrans. Crystal structure analysis. Phytochem-
istry 27: 1405-1407.
23. ASAKAWA, Y, TOYOTA, M., CHEMINAT, A. 1986. Terpenoids from the French liver-
wort Targionia hypophylla. Phytochemistry 25: 2555-2556.
24. INOUE, H. 1978. Koke no Sekai (World of Bryophytes). Idemitsu Ltd., Tokyo, 178p.
25. HASHIMOTO, T., OKUMURA, Y, SUZUKI, K., TAKAOKA, S., KAN, Y, TORI, M.,
ASAKAWA, Y 1995. The absolute structure of new I P-hydroxysacculatane-type diter-
penoids with piscicidal activity from the liverwort Pellia endivii{olia. Chern. Pharm. Bull.
43: 2030-2032.
26. ASAKAWA, Y, HARRISON, LJ., TOYOTA, M. 1985. Occurrence of a potent piscici-
dal diterpenedial in the liverwort Riccardia lobata var. yakushimensis. Phytochemistry
24: 261-262.
27. ONO, K., SAKAMOTO, T., ASAKAWA, Y 1992. Constituents of cell suspension cul-
tures of selected liverworts. Phytochemistry 31: 1249-1250.
28. ONO, K., SAKAMOTO, T., TANAKA, H., ASAKAWA, Y 1995. Sesquiterpenoids from
a cell suspension culture of the liverwort Porella vernicosa Lindb. Flavour and Fragr.
J. 11: 53-56.
29. TOYOTA, M., UEDA, A., ASAKAWA, Y 1991. Sesquiterpenoids from the liverwort
Porella acutifolia subsp. tosana. Phytochemistry 30: 567-573.
30. HASHIMOTO, T., TANAKA, H., ASAKAWA, Y 1994. Stereostructures of pi agio chi line
A and conversion of plagiochiline A and stearoyl velutinal into hot-tasting compounds
by human saliva. Chern. Pharm. Bull. 42: 1542-1544.
31. NAGASHIMA, F., TOYOTA, M., ASAKAWA, Y 1990. Bitter kaurane-type diterpene
glucosides from the liverwort Jungermannia in{usca. Phytochemistry 29: 1619-1623.
32. RYCROFT, D.S. 1990. Some recent NMR studies of diterpenoids from the Hepaticae. In:
Bryophytes: Their Chemistry and Chemical Taxonomy, D.H. Zinsmeister and R. Mues,
(eds.), Oxford University Press, Oxford, pp. \09-119.
33. HUNECK, S., CONNOLLY, J.D., HARRISON, L.J., JOSEPH, R., PHILLIp, WR.,
RYCROFT, D.S., FERGUSON, G., PARVEZ, M. 1986. New labdane diterpenoids from
the liverwort Scapania undulata. J. Chern. Res. (s), 162-163.
34. TOYOTA, M., TANIMURA, K., ASAKAWA, Y 1998. Cytotoxic 2,3-
secoaromadendrane-type sesquiterpenoids from the liverwort Plagiochila ovalifi)lia Mitt.
Planta Medica 64: 462-464.
342 Y ASAKAWA
35. ASAKAWA Y 1994. Highlights in phytochemistry of Hepaticae-biologically active

terpenoids and aromatic compounds. Pure and App!. Chern. 66: 2193-2196.
36. NAGASHIMA, F., OHI, Y, NAGAI, T., TORI, M., ASAKAWA, Y, HUNECK, S. 1993.
Terpenoids from some German and Russian liverworts. Phytochemistry 33: 1445-1448.
37. YOSHIDA, T., HASHIMOTO, T., TAKAOKA, S., KAN, Y, TORI, M., ASAKAWA, Y,
PEZZUTO, 1.M., PENGSUPAARp, T., CORDELL, G.A. 1996. Phenolic constituents of
the liverwort: four novel cyclic bisbibenzyl dimers from Blasia pusilla. Tetrahedron 52:
14487-14500.
38. PERRY, N.B., FOSTER, L.M., LORIMER, S.D., MAY, B.c., WEAVERS, R.T.,
TOYOTA, M., NAKAISHI, E., ASAKAWA, Y 1996. Isoprenyl phenyl ethers from
liverworts of genus Trichocolea: cytotoxic activity, structural corrections, and synthesis.
1. Nat. Prod. 59: 729-733.
39. ASAKAWA, Y, DAWSON, G.W, GRIFFITH, D.C., LALLEMAND, J.-Y, LEY, S.V,
MORI, K., MUDD, A., PEZECHK-LECLAIRE, M., PICKETT, A.1., WATANABE, H.,
WOODCOCK, C.M., ZONG-NING, Z. 1988. Activity of drimane antifeedants and
related compounds against aphids and comparative biological effects and chemical
reactivity of (-)- and {+)-polygodia!. 1. Chern. Eco!., 14: 1845-1855.
40. SCHWARTNER, c., BORS, W, MICHEL, C., FRANCK, u., MULLER-JAKIC, B.,
NENNINGER, A., ASAKAWA, Y, WAGNER, H. 1995. Effect ofmarchantins and related
compounds on 5-lipoxygenase and cyclooxygenase and their antioxidant properties: a
structure activity relationship study. Phytomedicine 2: 113-117.
41. HASHIMOTO, T., TORI, M., ASAKAWA, Y 1988. A highly efficient preparation of
lunularic acid and some biological activities of stilbene and dihydrostilbene derivatives.
42. NAGASHIMA, F., MOMOSAKI, S., WATANABE, Y, TOYOTA, M., HUNECK, S.,
ASAKAWA, Y 1996. Terpenoids and aromatic compounds from six liverworts. Phyto-
chemistry 41: 207-211.
43. ASAKAWA, Y, HUNECK, S., TOYOTA, M., TAKEMOTO, T., SUIRE, C. 1979. Mono-
and sesquiterpenes from Parella arboris-vitae. 1. Hattori Bot. Lab. 46: 163-167.
44. FUKUYAMA, Y, ASAKAWA, Y 1991. Novel neurotrophic isocuparene-type sesquiter-
pene dimers, mastigophorenes A, B, C and D, isolated from the liverwort Mastigophora
diclados.1. Chern. Soc. Perkin Trans. 2737-2741.
45. FUKUYAMA, Y, KODAMA, M. 1996. Search for novel neurotrophic factor-like sub-
stances in natural products. Food & Food Ingredients 1. 45-56.
46. TAIRA, Z., TAKEI, M., ENDO, K., HASHIMOTO, T., SAKIYA, Y, A SAKAWA, Y 1994.
Marchantin A trimethyl ether: Its molecular structure and tubocurarine-like skeletal
muscle relaxation activity. Chern. Pharm. Bull. 42: 52-56.
47. KATSUNUMA, N. 1997. Molecular mechanisms of bone collagen degradation in bone
resorption. 1. Bone Miner. Metab. 15: 1-8.
48. MATSUNAGA, Y, SAIBARA, T., KIDO, H., KATSUNUMA, N. 1993. Participation of
cathepsin B in processing of antigen presentation MHC class II. FEBS. 324: 325-330.
49. HASHIMOTO, T., IRITA, H., YOSHIDA, M., KIKKAWA, A., TOYOTA, M., KOYAMA,
H., MOTOIKE, Y, ASAKAWA, Y 1998. Chemical constituents of the Japanese
liverworts Odontoschisma denudatum, Parella japonica, P. acutifolia subsp. tosana and
Frullania hamatiloha. 1. Hattori Bot. Lab. 84: 309-314.
Chapter Thirteen
BIOLOGICALLY ACTIVE COMPOUNDS FROM

BUDDLEJA SPECIES
Peter 1. Houghton and Abraham Y. Mensah
Pharmacognosy Research Laboratories

Department of Pharmacy
King's College London
Manresa Road
London SW3 6LX, U.K.
Introduction-The Genus Buddleja ................................ 344

Ethnopharmacology of Buddleja ................................... 345
Wound Healing and Related Conditions ........................... 345
Treatment for Liver Disease .................................... 345
Bronchial Complaints ......................................... 345
Diuretic Activity ............................................. 345
Antirheumatic and Analgesic Activities ........................... 347
Sedative/Tranquilizing Effects ................................... 347
Other Medicinal Uses ......................................... 347
Chemical Constituents of Buddleja ................................ 347
Flavonoids .................................................. 348
Phenylethanoids .............................................. 348
Terpenoids .................................................. 351
lridoids ................................................... 351
Sesquiterpenoids and Diterpenoids ............................. 351
Triterpenoids .............................................. 351
Lignans ..................................................... 357
Phenolic Fatty Acid Esters ..................................... 357
Bioactivity Studies on Buddleja Extracts and Constituents .............. 357
Activities Relevant to Wound Healing ............................ 357
Associated Factors ......................................... 357
Studies Using Extracts ...................................... 360
Antimicrobial Activity ....................................... 360
Phytochemicals in Human Health Protection, Nutrition. and Plant Defense, edited by Romeo.
343
344 P. J. HOUGHTON AND A. Y. MENSAH
Antiinflammatory Effects .................................... 361

Antioxidant Effects ......................................... 361
Other Activities Associated with Skin Condition .................. 362
Conclusion ......... . ..................................... 362
Liver-Protectant Properties ..................................... 362
Antifungal Activity ........................................... 363
Analgesic Properties .......................................... 365
Antirheumatic Uses .......................................... 365
Other Activities .............................................. 365
Eye Treatment ............................................. 365
Insect-Plant Interaction ...................................... 366
Antiprotozoal and Molluscicidal Activity ........................ 366
Conclusions ................................................... 366
INTRODUCTION-THE GENUS BUDDLEJA
The genus Buddleja (sometimes written as Buddleia) was named by

Linnaeus after the English botanist Adam BuddIe who lived in the 18th century.
It is the largest genus in a taxon classed either as a separate family, the
Buddlejaceae, or as tribe, the Buddlejeae, within the Loganiaceae.! The former
position is that adopted by modem taxonomists such as Hutchinson. 2
The genus comprises about 100 species distributed throughout the warmer
temperate parts of southern Asia, south and east Africa, and in America from
Texas through to Chile and Argentina. Most species consist of woody bushes
attaining up to 3 m in height with many branches rather than a central trunk.
Some species have been used extensively in horticulture because of their col-
orful flowers which attract insects in large numbers. B. davidii Franchet, which
comes originally from China, is known commonly in England as the Butterfly Bush
because large numbers of these insects feed on the flowers during the summer. In
spite of the fact that it has only been grown in the UK for just over 100 years, it is
regarded as a typical plant of English gardens, and cultivars, having a range of
colors from white to deep red and blue, have been introduced. B. davidii has the
ability to grow in rocky places and has exploited this ecological niche in the bomb-
sites in major cities which were the result of the 1939-45 war, so that it has now
become naturalized in urban areas, on waste ground and on derelict land such as
disused railway lines, and in rural areas in road cuttings and chalk pits. About
twenty other species, as well as some hybrids, are grown as garden plants in the
UK, but the only other common ones are B. globosa Lam. from Chile and B. alterni-
folia Maxim from China, and neither of these has become naturalized.
Interest in the chemistry and biological activity of Buddleja is compara-
tively recent since it is not particularly toxic and its medicinal uses were
not widely known in the traditional North American and European centers of
scientific investigation. A review of the ethnopharmacology of Buddleja was pub-
lished in 1984. 3 This chapter seeks to cover and comment on the developments
BIOLOGICALLY ACTIVE COMPOUNDS FROM BUDDLEJA SPECIES 345
which have occurred since then, many of which have shown that the speculated
chemical basis for the reputed activities is in fact correct.
ETHNOPHARMACOLOGY OF BUDDLEJA
A detailed account from the literature of the ethnopharmacology of Bud-

dleja spp. has already been published. 3 A summary of the major traditional uses
compiled from this paper is given in Table I.
Wound Healing and Related Conditions
A poultice ofleaves of the indigenous Buddleja spp. is used in several dif-

ferent parts of the world to aid the healing of wounds. Ulcers, leprous lesions, and
bums are specifically mentioned in some cases. The underside of the leaves of Bud-
dleja species are commonly covered with a fairly dense mat of stellate covering tri-
chomes, and these may act as a crude felt to retain warmth. Evidence from scientific
studies, discussed below, suggests, however, that water-soluble compounds found
in the leaves may play an active role in the promotion of wound healing.
An aqueous extract of the flowers of some species, notably B. officinalis
Maxim., is used to treat sore eyes and opacity of the cornea. The symptoms for
which they are used, such as inflammation, are likely to arise from factors similar
to those causing skin complaints.
Treatment for Liver Disease
Liver disease is often characterized by jaundice and a general malaise,

which in more severe cases results in a variety of serious toxic effects. An aqueous
infusion of the leaves of Buddleja spp. is the most common form which is taken,
although roots are sometimes employed.
Bronchial Complaints
Preparations for the treatment of catarrh, coughs, bronchitis, and asthma

made from aqueous infusions of leaves are a feature of the medicinal use of Bud-
dleja from most of its endemic locations. In most cases, these are taken orally.
Diuretic Activity
The leaves of some Buddleja species formerly were official in the pharma-
copeias of several Central and South American countries because of their diuretic
activity. Diuresis is achieved by oral administration of a decoction of the leaves,
and occasionally the roots, of the species in question. The effect was ascribed to
the flavonoid and iridoid content of the leaves, but no further work has been
reported to justify or contradict this hypothesis.
Table 1. Ethnopharmacological uses of Buddleja species]
Species Geographical area
WOUND HEALING and related conditions
B. americana L. Mexico
B. asiatica Lour. Nepal, Philippines
B. cordata H.B.K. Mexico
B. davidii Franchet China (Shaanxi Province)
B. diffusa Ruiz et Pav. Peru, Ecuador
B. globosa Lam. Chile
B. incana Ruiz et Pav. Peru, Ecuador
B. officinalis Maxim. China
B. salviifalia Lam. South Africa
B. sessiliflora Kunth Mexico
LIVER AILMENTS
B. americana L. roots Mexico, Guatemala
B. cordata ~-[.B.K. roots Mexico
B. globosa Lam. leaves Chile
B. officinalis Maxim. China
BRONCHIAL CONDITIONS
B. albiflora Hems!. leaves, flowers China
B. cambara Arechav. Brazil
B. curviflora Hook et Am. leaves, flowers China
B. globosa Lam. leaves Chile, Bolivia
B. incana Ruiz et Pav. leaves Peru, Ecuador
B. madagascarensis Lam. leaves Madagascar, Uruguay
B. salviifolia Lam. South Africa
B. saligna Willd. South Africa
ANTIRHEUMATIC EFFECTS
B. americana L. Mexico, Guatemala
B. humboldtiana Roemer & Schultes leaves Texas, USA
B. salviifolia Lam. South Africa
DIURETIC EFFECTS
B. americana L. leaves, roots Mexico,Guatemala
B. humboldtiana Roemer & Schultes leaves Texas, USA
ANALGESIC EFFECTS
B. americana L. leaves, root, bark poultice Mexico, Guatemala,
B. hrasiliensis Jacqu. leaves, roots Brazil
B. diffusa Ruiz et Pav. Bolivia
SEDATIVE/TRANQUILIZING EFFECTS
B. americana L. root. Mexico, Guatemala,
B. madagascarensis Lam. leaves Madagascar
B. quinquenaria Cham. et Schltr. Brazil
Table 2. Miscellaneous minor uses of Buddleja species}

Use Species
Fish poisons B. asiatica leaves (Vietnam)
B. curviflora leaves (China, Japan)
Abortifacient B. asiatica leaves (Nepal, Philippines)
B. americana L. leaves (Mexico)
Antimalarial B. asiatica roots (Philippines)
B. curviflora flowers (China)
Antitumour B. asiatica leaves (Philippines)
B. officinalis Maxim (China)
Urinogenital disinfectant B. americana leaves (Mexico)
B. globosa leaves (Bolivia)
B. incana bark (Peru, Ecuador)
Antirheumatic and Analgesic Activities
Several species of Buddleja are reported to be used to make a poultice to

wrap around joints in the arms or legs to relieve aches and pains due to rheuma-
tism. Any positive effect might be due to analgesic, anti-inflammatory, or some
other related effect. In some cases, a more general analgesic use has been recorded
as shown in Table 1.
Sedative/Tranquilizing Effects
Some of the ethnopharmacological uses of Buddleja species indicate that

they may possess properties associated with depression ofCNS activity since they
are used as hypnotics and sedatives.} No work dealing with this activity has been
reported.
Other Medicinal Uses
Buddleja species are reported to be used for a variety of other medicinal

purposes which are listed in Table 2.
CHEMICAL CONSTITUENTS OF BUDDLEJA
Buddleja species contain the typical chemical characteristics (i.e. iridoids

and phenylethanoids) of the group of dicotyledons having flowers with fused
petals. The occurrence of congeners of these different types, e.g. lignan-
iridoid, phenylpropanoid-iridoid, and phenylpropanoid-flavonoid, is a noteworthy
feature of the reports on Buddleja phytochemistry published in the last few
years. 4 .5.6
348 P. 1. HOUGHTON AND A. Y. MENSAH
R50
R4 R3
Compound R1 R2 R3 R4 Rs
1 Linarin H OCH 3 H H Rutinose
2 Kaempferol H OH OH H H
3 Quercetin OH OH OH H H
4 Acacetin H OCH 3 H H H
5 Luteolin OH OH H H H
6 Luteolin 7-0- OH OH H H Glucose
glucoside
7 6-0H luteolin OH OH H OH H
8 Apigenin-7 -0- H OH H H Glucose
glucoside
9 Rutin H OH O-rutinose H H
10 Scutallerein 7-0- H OH H OH Glucose
glucoside
11 Pectolinarigenin H OCH 3 H OCH 3 H
12 Salvigenin H OCH 3 H OCH 3 CH 3
13 Buddlenoid A 13'0 H OH OH H Cinnamoyl-6-
glucose
14 Buddlenoid B 14'u OCH 3 OH OH H Cinnamoyl-6-
glucose
Figure L Flavonoids isolated from Buddleja species.
Flavonoids
The first flavonoid to be isolated from this genus was called

buddleoflavonoloside and was obtained from B. davidii. 7 This was later shown to
be linarin (1) which had been isolated also from B. officinaiis leaves. 8 Since then,
a variety of flavones and flavonols has been isolated from the flowers and leaves
of several Buddleja species and these are shown in Figure 1.
Pbenyletbanoids
A compound showing a blue fluorescence under UV light 365 nm had been
noted in the early screening procedures for flavonoids carried out by Bate-Smith in
1962. 9 This compound was shown to be orobanchin (15) and, subsequently, several
other related compounds have been isolated from Buddleja species (Figure 2).4
HOy~
I
O~
n~O
~R
~
0 ____O,
00
HO ~~O OH I
CH 3 0 ~ OH
HO
HO OH
R R'
15 OH OH Orobanchin
16 H OH Verbascoside
(Acteoside)
CH 2 0H
17 H
H~~O,- Echinacoside
OH
R R' R"
18 Ac H H
19 H Ac H
20 H H Ac
Figure 2. Phenylethanoids isolated from Buddleja species.

CH20R"
R'O~O ?' OH
o OH ~
c~
HO '" OR
HO
OH
R R' R"
0
21 CH 3
~
H
HO
0
CHaO~
/" 1 '"
22 H H
HO '"
H H
HO~
23 :1 '"
HO
0
CH30~
24 H /'" 1 '"
H
HO '"
0
CH30~ H
25 CH3 /"'1 '"
0
HO '"
CH30~
/"'1 '"
HO '"
26 CH3 H
0
HO~ H
27 CH3 :1 '"
HO
0
CH30~
28 CH3 /"1 '" arabinose
HO '"
0
CH30~
29 /"1 '" apiose
CH3 HO '"
0
30 H HO~ arabinose
:1 '"
HO
Terpenoids
Iridoids. The iridoids of Buddleja belong to the C-9 catalpol and aucubin
group rather than the C-IO type such as loganin. The latter is found characteris-
tically in the Loganiaceae, and the occurrence of the C-9 compounds gives
chemotaxonomic support to the contention that Buddleja should form part of a
separate family rather than be included in the Loganiaceae. Aucubin (31) from
the leaves of B. globosa was the first iridoid to be isolated, and several similar
compounds have been reported since then. 3 The catalpol type of iridoid (34),
where an epoxide group replaces the C6-C7 double bond found in the aucubin
(31) analogues, also occurs, and compounds with ester-linked phenylpropanoid
residues are known for both of these types as well as for the related compound
ajugolY lridoids so far reported (apart from those with a lignan moiety) are
shown in Figure 3.
More recently, a group of compounds consisting of iridoids linked to a
lignan moiety have been isolated from the polar fractions ofthe roots of B. davidii.
(see Fig. 7).
Sesquiterpenoids and Diterpenoids. A series of caryophyllane se-

squiterpenes named buddledins A-E (54-58) was isolated from the roots of B.
davidii.lO,11 These compounds possess piscicidal activity. Two other sesqui-
terpenes described from two Mexican species B. cordata H.B.K. and B.
sessilifiora Kunth. were at first thought to be contain a novel skeleton (59,60) but
were subsequently shown to be identical to the known compounds cyclocol-
orenone (61) and l-hydroxycyclocolorenone (62).12,13 The first diterpene from this
genus has only recently been reported from the stembark and roots of B. globosa
and B. albifiora. It was named buddlejone and has an abietane skeleton with an
interesting structure (63) which, although containing only one carbonyl group,
showed two signals in its 13C NMR spectrum due to its tautomeric nature. 14 Three
similar compounds (64-66) have recenly been isolated from the same source in
our laboratories. 15 All the sesquiterpenes and diterpenes so far isolated are shown
in Figure 4.
Triterpenoids. The presence of saponins in the leaves of Buddleja

species had been known for several decades before any compounds were
isolated and characterized. At first, common sterols and amyrin triterpenoids
were the only compounds found, but more recently several saikosaponin
oleananes (67-71) have been described. 16 Related compounds named mim-
engosides A (72) and B (73) have been isolated from the flowers of B.
officinalis, and mimengoside A has also been obtained from the leaves of B.
madagascarensis .17.18 The structure of these novel compounds is shown in
Figure 5.
RO H
R~
~
CH2~H O"L~~H
R
0
HO~OHOH
~~ 2
R
0"
HO OH OH
3
H Aucubin
31 34 H Catalpol
~"'
¢
OH
p-Methoxycinnamoyl
3
32 aucubin 5 35 Catalposide
o 0, )
o 0"
33 HO~
"" '
I
0~CCH3 ~'
CH 3 -·.0
o
36 p-Methoxycinnamoyl
catalpol 8
9
Buddlejoside A 2
o 0,
RO~
HO
)o! O~O 3 0"
CH20H
HO/ OH OH
R
37
Vanillylajugol
38 Feruloylajugol
\
CHao~OH
39 HO OH
CHa
Buddlejoside A1 9
o
Figure 3. Iridoids isolated from Buddleja species.

c~m
R"O
R'O
0
OR 0
'-"::::
0
CH,~O~
HO OH OH
R R' R"
40 H H C~O~
41 CHaCO H C~O~ Buddlejoside A3
42 H CH3CO CH30~
Buddlejoside ~
43 H CH30~ H Buddlejoside ~
44 CH3COC~O~ H Buddlejoside Ar.

0
45 CHaCO H C~O-o! Buddlejoside A7
46 H CH3CO CH30 -OJ~ /, Buddlejoside As
47 H H CHaO~ Buddlejoside As
CH3
48 CH3CO H CHaO~ Buddlejoside A10

CH30
49 H CH3CO CH30~ Buddlejoside A11
C~9Y
50 H CH3CO CH30 ~ /, Buddlejoside A12.
CH30
51 H CH3CO CH30~ Buddlejoside A13
HO
52 H Buddlejoside A14
53 H Buddlejoside A15
CH, Y
+H- CH,
CH 2
R 57 Buddledin 0
54 OH Buddledin A
55 H Buddledin B
56 OOCCH 3 Buddledin C
58 Buddledin E
~
\CH3
." H CH3
CHC/i!.CH3
?'" 0 o
-:~
CH 3 ...~
H R
CH 3 CH 3 CH 3
R R
59 H 61 H Cycloclorenone
60 OH 62 OH 1- Hydroxycycloclorenone
Figure 4. Sesquiterpenoids and diterpenoids isolated from Buddleja species.

63 Buddlejone
64 Anhydrobuddlejone 65 Anhydrobuddlejol
66 Maytenone
R R'
67 OH H Saikosaponin A
68 OH Rhamnose (1- 4) glucose Buddlejasaponin 1
69 OH Xylose (1- 4) glucose Buddlejasaponin 2
70 OH Xylose Buddlejasaponin 3
71 OH Glucose Buddlejasaponin 4
72 H Rhamnose (1- 4) glucose Mimengoside A
73 Mimengoside B
Figure 5. Triterpenoid saponins isolated from Buddleja species.

Lignans
Lignans and neolignans from B. davidii stem, first isolated in 1984, are
somewhat unusual, as some of them are trimeric congeners (see Fig. 6) and
most lignans are dimeric in nature. 19 Several compounds have also been described
where a lignan unit is combined with an iridoid, but in these cases, the iridoid
component is ajugol rather than aucubin or catalpol (see Fig. 7).
Phenolic Fatty Acid Esters
Investigation of the least polar fraction of a lipophilic extract of B. globosa

bark led to the isolation of fatty acid esters of simple phenolic a1cohol. 20 (Fig. 8).
These compounds are thought to occur usually as units of the complex polymeric
material known as suberin.
BIOACTIVITY STUDIES ON BUDDLEJA EXTRACTS

AND CONSTITUENTS
Activities Relevant to Wound Healing
Associated Factors. Wound healing is a complicated process and

consists of a cascade of events triggered by the causative trauma. Inflammation
and clotting processes to prevent blood loss occur quickly, and these are
followed by granulation and neovascularization which, after a few days in other-
wise healthy patients, are followed by wound contraction accompanied by re-
epithelialization and, in the longer term, tissue remodelling. The healing process
may be delayed by bacterial, fungal, or viral infection of the wound, and such
infection can also lead to systemic infection if the wound is still "open".
A substance or extract reputed to have wound-healing properties may thus
exhibit one or more of a variety of biological activities which would indicate a
basis for its traditional use. These include anti-inflammatory effects, epidermal
cell growth stimulation, stimulation of phagocytosis, and antimicrobial proper-
ties. These properties in tum may be due to one or more types of activity at the
cellular and enzyme level. Anti-inflammatory effects may be due to inhibition of
various stages of leukotriene and prostaglandin synthesis, inhibition of oxidation
processes, reduction of macrophage and complement activity, as well as inhibi-
tion of the proteinases produced as the end result of the combination of the above
processes.
Comparatively little work has been done to investigate whether extracts and
isolated constituents from Buddleja have any such effects, but the circumstantial
evidence from the activity which has been recorded for some of the types of con-
stituents which have been isolated is discussed below.
OCH 3
OH
CHO
HO
74 Balanophonin 75 Syringaresinol
0 OCH 3
HO OCH 3
CH 2 0H
R
76 CHO Buddlenol A
77 CH 2 0H Buddlenol B
OCH 3
R'
OH
OH
0
OCH 3
CH 3 0
0
0
OCH 3
CH 2 0H
R R'
78 H OCHa Buddlenol C
79 OCHa OCHa Buddlenol D
80 H H Buddlenol E
R OCH 3 H
81 OCHa Buddlenol F
OH
Figure 6. Lignans isolated from Buddleja species.

HO
HO
{p 0H
H
~
o
CH3 O~O~H
H~HOH
82
HO
HO
{p 0H
H
~
o
CH 3 O~O~OH
OH H~HOH
84
Figure 7. Lignan-iridoid congeners isolated from Buddleja species.
OH OH
n n
85 20 87 20
86 22 88 22
Figure 8. Phenolic fatty acid esters isolated from Buddleja species.
Studies Using Extracts. An aqueous infusion of the leaves of B. globosa,

used as a wound healing agent in Chile, showed antioxidant properties in a test
which measured the bleaching of preformed colored cations. 21 Recent
in vitro studies carried out in our laboratories have shown that a standardized
aqueous extract of leaves of B. globosa has properties which indicate that its
traditional use may be justified. Preliminary unpublished work shows that the
growth of cultured epidermal cells, indicated by incorporation of neutral red, is
stimulated by concentrations of the leaf extract on the order of 10 Ilg/ml. This is
well within the concentration of the preparation utilized in traditional medicine
in Chile. It should be noted, however, that higher concentrations are cytotoxic,
so the evidence for efficacy in the cell proliferation aspects of wound healing is
equivocal.
Antimicrobial Activity. Several of the types of constituents present in the

leaves of Buddleja spp. may display antimicrobial activity if present in sufficient
concentrations, and these compounds are discussed below.
The aglycones formed by hydrolysis of iridoids are unstable and are often
biologically more active than the parent glycoside. It has been shown that the
unstable aglycone formed on hydrolysis of aucubin (31) has bactericidal activity,
so in situ enzyme hydrolysis of the glycosides present in the leaves may occur in
a poultice of crushed leaves and so exert an antibacterial effect. 22
In addition, any phenolic propanoid iridoid esters present are likely
to exhibit intrinsic antibacterial activity such as demonstrated for the c1osely-
related catalpol compounds in Kigelia pinnata. 23 Verbascoside (16), the major
phenylethanoid from the leaves of B. globosa, has been shown to be the domi-
nant antibacterial compound when tested against Staph. aureus and E. coli. 24 More
recently, a related compound, angoroside A (30) has also been reported from the
same species and been shown to be active against Staph. aureus. 25
Antiinflammatory Effects. The process of inflammation in tissue is a

complex one involving the synthesis of prostaglandins and leukotrienes, the
release of histamine, and increased permeability of peripheral blood vessels
leading to edema, pain, and swelling. Inflammation can be reduced or prevented
by agents which reduce or inhibit the enzyme processes or effects on tissue which
take place. Constituents of plants are known to act in a variety of these ways to
achieve the desired effect.
Flavonoids have often been implicated in the antiinflammatory activity of
plant extracts, and their presence may therefore aid the wound healing process.
The flavonoids of Budd/eja are found principally in the leaves and flowers, and
it is these parts of the plant which are used extensively for wound healing and for
the treatment of sore eyes. The flavone aglycones luteolin (5) and quercetin (3)
have both been detected in Buddleja species. 26 These compounds are potent
inhibitors of histamine release while flavonols such as kaempferol (2) inhibit
5-lipoxygenase which is a key enzyme in prostaglandin synthesisy,28 Both the
aqueous extract of B. cordata leaves and its contained linarin (1) have been shown
to posess anti-inflammatory activity when tested in vivo with rats by using the
carrageenan-induced paw edema test. 29 The triterpenoids so far reported from
Buddleja are saikosaponins, and it is worth noting that similar compounds are
constituents of Bupleurum species, widely used in traditional Chinese medicine,
and they have been shown to have a considerable antiinflammatory effect. 30 If the
Buddleja saikospaonins also have this effect, they could contribute to the wound
healing process.
Antioxidant Effects. Wound-healing can be affected adversely by the pres-

ence of excess oxygen free radicals, so anti-oxidants also may contribute to
Table 3. Anti-oxidant effect of constituents of B. globosa aqueous leaf extract.

(Viability of cultured fibroblast cells measured by absorbance of
incorporated neutral red)
Concentration % protection in cells
Extract/compound Ilg/mL challenged with H2O,
Total extract 5 93.3
Total extract 10 96.3
Verbascoside 16 10 88.0
Echinacoside 17 10 93.2
Linarin 1 10 92.0
Luteolin 5 10 93.8
6-0H luteolin 7 10 91.7
Catalase (positive control) 250i.u.lmL 95.5
wound healing. Catechol f1avonoids such as quercetin (3) are particularly effec-
tive in this context due to chelation with iron, and are important in oxidation
processes and the production of the oxygen free radicals which cause tissue
damage.
Recent studies in our laboratories also have shown that the aqueous extract
of B. globosa leaf displays considerable antioxidant activity. A variety of con-
stituents of this extract were tested as single compounds (Table 3). It can be seen
that both the phenylethanoids such as verbascoside (16) and the f1avonoids such
as linarin (1) display a significant effect.
Other Activities Associated with Skin Condition. It is interesting to note

that two new flavonol derivatives, buddlenoid A (13) and B (14) from B. coriacea
leaves have been shown to have an inhibitory effect on tyrosinase. 3 ! This enzyme
increases skin pigmentation, sometimes noted after inflammatory conditions, so
any inhibitory substance will cause a lightening in color of the skin to which it
is applied.
Triterpenoid glycosides are the major type of saponins, naturally-occurring
detergents, which can be used to cleanse the skin. As discussed above, this type
of compound has been reported from the leaves and flowers of several Buddleja
spp., and so an aqueous extract of these might be expected to remove grease, dirt,
and associated microbial organisms and so have a cleansing effect which would
improve the conditions for wound healing.
Conclusion. It can be seen that Buddleja contains several different types

of compounds which are likely to aid wound healing in several ways. This is not
uncommon in plant extracts used for treatment of a condition where multiple
factors are involved in alleviation.
Liver-Protectant Properties
The reports of the use of Buddleja extracts from widely-different geo-

graphical localities for the treatment of liver disease prompted an investigation
into three of the species used, i.e. B. americana roots, B. globosa leaves, and B.
officinalis leaves. s An in vitro model of cultured hepatocytes was used in which
the amount of the enzyme glutamate-pyruvate transaminase (GPT) released into
the surrounding medium is measured and taken as an indication of the integrity
of the cell membrane. Cell damage was induced by standard toxic agents such
as carbon tetrachloride, galactosamine, and a complement-mediated system. The
reduction in GPT levels, when the extract was added to the cultures, was mea-
sured. All three extracts showed significant protectant effects. Iridoids, f1avonoids,
and phenylethanoids were isolated from the extract of B. globosa leaves used and
were tested in the same system while using glycyrrhizin as a positive control. The
greatest activity was shown by the flavonoid linarin (1) and the phenylethanoids
Table 4. Reduction in GPT release from challenged cultured hepatocytes in the

presence of Buddleja aqueous extracts and constituents of Buddleja globosa leaves 5
GPT level (%) n = 3
Concentration Carbon
Extract! compound mg!mL tetrachloride+ Galactosamine+
Negative control 100 ± 2 100±1
B.americana root 1.0 132 ± 5 43 ± 1**
B. officinalis leaf 1.0 92 ± 3 46 ± 2**
B. globosa leaf 1.0 78 ± 0 94 ± I
p-Methoxycinnamoyl-aucubin 32 0.1 92 ± I 101 ± I
p-Methoxycinnamoyl-catalpol 36 0.1 72 ± 2* 80 ± 5
Catalposide 35 0.1 92 ± I 102±8
Catalpol34 0.1 78 ± 2 102±1
Aucubin 31 0.1 76 ± 3 107 ± 2
Verbascoside 16 0.1 66 ± 1* 70 ± II
Echinacoside 17 0.1 55 ± 6* 34 ± 2**
Linarin 1 0.1 73 ± 6* 72 ± 5*
Glycyrrhizin (positive control) 1.0 47 ± 1** 39 ± 7**
'Hepatotoxic agent used as challenge *p < 0.0 I **p < 0.00 I.
verbascoside (16) and echinacoside (17), while the iridoids had no activity unless
they contained a phenylpropanoid residue (Table 4).
Antifungal Activity
There are no direct ethnobotanical reports of a specific use of Buddleja to

treat fungal infections in humans, plants, or animals although their uses for skin
disorders might imply such activity since fungal infections are a common cause
of skin blemishes. Interest in the possible antifungal properties arose from the
observation that the freshly-dug roots of B. gtobosa and B. davidii had a fairly
strong odor, indicating the release of volatile compounds into the air spaces in
the soil (Houghton, personal observation). Such compounds might be released
and act as defense compounds against soil pathogens such as nematodes and
fungi. Preliminary in vitro tests showed that the chloroform extracts of roots, but
also mature stem bark, of these species possessed antifungal activity. Subsequent
bioactivity-guided fractionation of the chloroform extract of B. globosa mature
stem bark has shown that the most active compound are the sesquiterpenes bud-
dledin A (54) and buddledin B (55). The extracts and compounds were active
against the dermatophytic fungi Epidermophyton jioccosum, Trichophyton inter-
digitate, and r rubrum, but not against the other fungi and yeasts tested. Full
details of activity are shown in Table 5.
Another study has shown that the triterpenoids buddlejasaponin I (68) and
mimengoside A (72) have activity against 9 yeast species. 'R .32
Screening of the essential oil from B. asiatica flowers against nine fungal
w
~
Table 5. Antifungal activity (MIC;llg/mL) of compounds isolated from B. globosa stembark CHCb extract
ORGANISMS*
COMPOUND AN CA EF PN SB SC SD TI TR
Total extract >1,000 >1,000 7.8 >1,000 >1,000 >1,000 >1,000 7.8 7.8
Buddledin A 54 >1,000 >1,000 12.0 >1,000 >1,000 >1,000 >1,000 12.0 12.0
Budd1edin B 55 >1,000 >1,000 12.0 >1,000 >1,000 >1,000 >1,000 12.0 12.0
Buddlejone 63 >1,000 >1,000 750 >1,000 >1,000 >1,000 >1,000 750 750
Anhydrobuddlejone 64 >1,000 >1,000 750 >1,000 >1,000 >1,000 >1,000 750 750
Maytenone 66 >1,000 >1,000 375 >1,000 >1,000 >1,000 >1,000 750 750
Miconazole (positive control) 1.2 2.4 2.4 2.4 2.4 2.4 2.4 2.4 2.4 ;c
~
'Key to organisms. ::c:
AN Aspergillus niger ATCC 16404. SC Saccharomyces cerivisae ATCC 10234. o
CA Candida albicans ATCC 1023l. SO Scytalidium dimidiatum EL 936. c:
EF Epidermophytonjloccosum SJH. TI Trichophyton interdigitale EL 5171.
PN Penicillium notatum ATCC 11654. TR Trichophyton rubrum EL 5095.
SB Scrophulariopsis brevicaulis EL 3839. ~
~
?>
:<
~
~
rn
~
species showed that it possessed a dose-related activity against all of them, and
a 0.5% solution of the oil in Tween 80 was comparable in activity to a 2% resor-
cinol solution. 33
Analgesic Properties
In vivo tests using mice and rats have justified the ethnopharmacological
uses of the Central American species of Buddleja as analgesics. 34 The leaves of
B. cordata and also its principal flavonoid, linarin (1), both showed effects in mice
which indicated suppression of pain. A dose of lOOmg/kg of linarin (1) had a
similar effect to 3 mg/kg morphine sulphate for heat-induced pain and a greater
effect than 100mg/kg of acetylsalicylic acid. However, 100mg/kg extract had
much the same effect as the same dose of linarin (1), and this indicates that the
activity could not be due solely to linarin. This was emphasised even more in a
test for acetic acid-induced abdominal writhing where the ED50 of the extract
(22mg/kg) was lower than that of linarin (89.0mg/kg).34
Antirheumatic Uses
Rheumatism is an inflammatory condition and so the anti-inflammatory
activities of Buddleja constituents mentioned above could contribute to the relief
of rheumatic pain. The in vivo reduction of carrageenan-induced rat paw edema
displayed by the extract of B. cordata leaves and linarin (1) is an indication that
anti-rheumatic activity might exist. 29
Other Activities
Eye Treatment. The flowers of B. officinalis, known as "Mi Mueng Ha",
have enjoyed a considerable reputation in Chinese traditional medicine as a treat-
ment for sore eyes and improving the clearness of the eye. 3 The flowers contain
flavonoids, notably linarin (1), and the triterpenoid saikosaponins known as
mimengosides A and B (72,73) as well as the phenylethanoid acteoside. 8,17 The
triterpenoids may well have antiinflammatory activity (see above) which would
Table 6. Inhibitory activity (lC 50 values in 11M) of

constituents of B. officinalis flowers on
rat lens aldose reductase 35
Compound Activity-IC,o (11M)
Luteolin 5 0.21
Luteolin-7-0-~-D-glucopyranoside 6 0.28
Apigenin 0.58
Linarin 1 0.75
alleviate soreness, but a more interesting finding is the inhibitory activity of con-
stituents of the flowers of B. officinalis on aldose reductase when tested in vitro. 35
Aldose reductase is involved in cataract formation as one of the complications of
diabetes and its inhibition would indicate some lessening in the risk of this occur-
ring. A 70% methanolic extract of the flowers showed high activity and, when the
constituents were isolated and tested, four of the flavonoids showed significant
effects at low concentrations (Table 6).
Insect-Plant Interaction. The attraction of Buddleja flowers against insects

such as butterflies and bees has been mentioned above. No chemical basis has yet
been reported for this, and the composition of the volatile oil produced by the
flowers is not known. The flowers produce large amounts of nectar, and the base
of the inside of the tubular corolla can be seen to have an intense yellow color.
Preliminary studies in our laboratories have shown that this is due to polar
carotenoid compounds rather than flavonoids, and these may well serve as visual
attractants for insects.
Methanol extracts of the leaves of B. coriacea showed tyrosinase inhibitory
effects, and the compounds responsible were shown to be the flavonoids buddle-
noids A (13) and B (14).31 Tyrosinase is a key enzyme in the insect molting
process, and so the presence of inhibitors indicates possible applications of these
compounds as insect control agents.
Antiprotozoal and Molluscicidal Activity. The saponin buddlejasaponin I

(68) has been shown to be active at inhibiting growth of Trichomonas vagina lis
and Leishmania infantum at 12.5Ilg/ml and 251lg/m1.32 The related compound
mimengoside A (72) had activities of 20 /lg/ml and 40llg/ml, respectively against
the same two organisms. ls These two compounds also were found to be mollus-
cicidal against Biomphalaria alexandrina at concentrations of 10 Ilg/ml and 4
Ilg/ml, respectively.ls.32
5. CONCLUSIONS
Sufficient knowledge of the chemistry and biological activity of Buddleja

extracts and constituents is now known to account for the major traditional uses
of this genus. Previously-unknown activities, notably the fungicidal effects, have
come to light and provide new lead compounds.
The reports of compounds consisting of congeners is interesting, as are also
the novel types of terpenoids being isolated from the bark and roots. As more
species are investigated, it seems likely that additional novel, active compounds
belonging to these types will be reported. More work needs to done to investi-
gate any chemical basis for the insect-attractant properties of the flowers of this
genus since this may be of use in an agricultural context.
ACKNOWLEDGMENTS
One of us (A.Y.M.) thanks the Association of Commonwealth Universities

for a Commonwealth Research Fellowship. Financial support for some of the
work carried out at King's College London has also been given by the British
Council and the Royal Society. Our thanks also to the Chelsea Physic Garden and
the Shardong Insitute of Medical Sciences, Peoples' Republic of China, for supply
of plant material.
REFERENCES
I. BRUMMITT, R.K. 1992. Vascular Plant Families and Genera. Royal Botanic Gardens
Kew, p. 510.
2. HUTCHINSON, 1. 1959. The Families of Flowering Plants vol. I. Clarendon Press,
Oxford, p. 373.
3. HOUGHTON, PJ. 1984. Ethnopharmacology of some Buddleja species. 1. Ethnophar-
macology 11: 293~308.
4. YAMAMOTO, A., NITTA, S., MIYASE, 1., UENO, A., WU, L-1. 1993. Phenylethanoid
and lignan-iridoid complex glycosides from roots of Buddleja davidii. Phytochemistry
32: 421---425.
5. HOUGHTON, P1., HIKINO, H. 1989. Antihepatotoxic activity of extracts and con-
stituents of Buddleja species. Planta Medica 55:123~126.
6. MIYASE, 1., AKAHORI, C, KOHSAKA, H., UENO, A. 1991. Acylated iridoid glyco-
sides from Buddleja japonica Hemsl. Chern. Pharm. Bull. 39:2944~2951.
7. YU, H. 1933. Chemical study of Buddleia variabilis. Bull. Soc. Chim. BioI. 15: 482---497.
8. TSENG, K.F., CHANG, S. 1953. Constituents of Buddleja officinalis. Acta Pharmaco-
logica Sinica I: 84~85.
9. BATE-SMITH, E.C 1962. The phenolic constituents of plants and their taxonomic
significance. 1. Linn. Soc. (Bot) 58: 95~173.
10. YOSHIDA, 1., NOBUHARA, 1., UCHIDA, M., OKUDA, 1. 1978. Studies on the con-
stituents of Buddleja species I. Chern. Pharm. Bull. 26: 2535~2542.
II. YOSHIDA, 1., NOBUHARA, 1., FUJII, N., OKUDA, 1. 1978. Studies on the constituents
of Buddleja species II. Chern. Pharm. Bull. 26: 2543~2549.
12. DEVIVAR, A.R., NIETO, D.A., GAVINO, R., PEREZ, A.L. 1995. Isocapnell-9-en-8-one
and 6-alpha-hydroxyisocapnell-9-en-8-one, sesquiterpenes from Buddleia species. Phy-
tochemistry 40: 167~ 170.
13. DEVIVAR, A.R., NIETO, D.A., GAVINO, R., PEREZ, A.L. 1996. Isocapnell-9-en-8-one
and 6-alpha-hydroxyisocapnell-9-en-8-one, sesquiterpenes from Buddleia species. Phy-
tochemistry 42: 1709.
14. HOUGHTON, P.l., WOLDEMARIAM, TZ., CANDAU, M., BARNARDO, A.,
KHEN-ALAFUN, O. AND LI SHANGXIAO 1996 Buddlejone, a diterpene from Bud-
dleja albiflora. Phytochemistry 42: 485---488.
15. HOUGHTON, PJ., MENSAH, A.Y., L1AO, Y.H. 1998 Novel diterpenoids from B.
globosa and B. yunnanensis. Phytochemistry (submitted).
16. YAMAMOTO, A., MIYASE, 1., UENO, A., MAEDA, 1. 1991. Buddleja saponins I-IV,
4 new oleanane-triterpene saponins from the aerial parts of Buddleja japonica Hemsl.
Chern. Pharm. Bull. 39: 2764~2766.
17. DING, N., YAHARA, S., NOHARA, 1. 1992. Structure of mimengoside-A and
368 P. I HOUGHTON AND A. Y. MENSAH
mimengoside-B, new triterpenoid glycosides from Buddlejae Flos produced in China.

Chern. Pharm. Bull. 40: 780-782.
18. EMAM, A.M., MOUSSA, A.M., FAURE, R., FAVEL, A., DELMAS, F., ELIAS, R.,
BALANSARD, G. 1996. Isolation and biological study of a triterpenoid saponin,
mimengoside-A, from the leaves of Buddleja madagascarensis. Planta Medica 62: 92-93.
19. HOUGHTON, PJ. 1984. Lignans and neolignans from Buddleja davidii. Phytochemistry
24: 819-826.
20. HOUGHTON, PJ. 1989. Phenolic fatty acid esters from Buddleja globosa stembark. Phy-
tochemistry 28: 2693-2695.
21. CAMPOS, A.M., LISSI, E.A. 1995. Evaluation of the antioxidant capacity of herbal teas
by a procedure based on the bleaching of ABTS radical cations. Boletin de la Societad
Chilena de Quimica 40: 375-381.
22. ROMBOUT, IE., LINKS, 1. 1956. The chemical nature of the antibacterial substance
present in Aucuba japonica Thunb. Experientia 123: 78-80.
23. AKUNYILI, D.N., HOUGHTON, PJ., RAMAN, A. 1991. Antimicrobial activities of the
stembark of Kigelia pinnata. I Ethnopharmacology 35: 173-177.
24. PARDO, F., PERICH, F., VILLAROEL, L., TORRES, R. 1993. Isolation of verb as cos ide,
an antimicrobial constituent of Buddleja globosa leaves. I Ethnopharmacology 39:
221-222.
25. PARDO, F., PERICH, F., TORRES, R. 1997. A new glycoside with antibacterial activity
isolated from Buddleja globosa. Boletin de la Societad Chilena de Quimica 42: 101-104.
26. HARBORNE, IB., WILLIAMS, C.A. 1971. 6-Hydroxyluteolin and scutellarain as
phyletic markers in higher plants. Phytochemistry 10: 367-368.
27. AMELLAL, M., BRONNER, c., BRIANCON, F., HAAG, M., ANTON, R., LANDRY,
Y. 1985. Inhibition of mast cell histamine release by flavonoids and biflavonoids. Planta
Medica 51: 16-22.
28. ALCARAZ, AJ., HOULT, 1.R.S. 1985. Actions of flavonoids and the novel anti-
inflammatory flavone, hypolaetin-8-g1ycoside, on prostaglandin biosynthesis and inacti-
vation. Biochem. Pharmacol. 31: 1449-1454.
29. MARTINEZ-VASQUEZ, M., APAN, T.O.R., LASTRA, A.L., BYE, R. 1998. A compar-
ative study of the analgesic and anti-inflammatory activities of pectolinarin I isolated
from Cirsium subcoriaceum and linarin isolated from Buddleia cordata. Planta Medica
64: 134-137.
30. CHANG, H.M., BUT, P'P.-H. 1987. Pharmacology and applications of Chinese materia
medica. World Scientific Publishing, Singapore. p. 967.
31. KUBO, I., YOKOKAWA, Y. 1992. Two tyrosinase inhibiting flavonol glycosides from
Buddleia coriacea. Phytochemistry 31: 1075-1077.
32. EMAM, A.M., DIAZ-LANZA, A.M., MATELLANO-FERNANDEZ, L., FAURE, R.,
MOUSSA, A.M., BALANSARD, G. 1997. Biological activities of buddlesaponin iso-
lated from Buddleja madagascarensis and Scrophularia scorodonia. Pharmazie 52:
76-77.
33. GARG, S.G., OSWAL, V.B. 1981. In vitro antifungal activity of the essential oil of Bud-
dleia asiatica Lour. Revista Italiana E.P'P'O.S. 57: 365.
34. MARTINEZ-VASQUEZ, M., APAN, T.O.R., AGUILAR, H., BYE, R. 1996. Analgesic
and antipyretic activities of an aqueous extract and of the flavone linarin of Buddleia
cordata. Planta Medica 62: 137-140.
35. MATSUDA, H., CAl, H., KUBO, M., TOSA, H., IINUMA, M. 1995. Study on anti-
cataract drugs from natural sources 2. Effects of Buddlejae Flos on in vitro aldose-
reductase activity. BioI. Pharm. Bull. 18: 463-466.
Chapter Fourteen
CYANIDE IN FOODS
Biology of Cyanogenic Glucosides and Related Nutritional
Problems
Dirk Selmar
Botanical Institute of the Technical University Braunschweig

Mendelssohnstr. 4
38092 Braunschweig
Germany
Introduction ................................................... 369

Cyanogenesis .................................................. 370
Biosynthesis and Structures ...................................... 371
Translocation .................................................. 373
Functions ..................................................... 375
Ecological Significance ....................................... 375
Evolutionary Aspects ......................................... 376
Significance in Weakly Cyanogenic Plants ........................ 378
Cyanogenic Glucosides in Foods .................................. 380
General Aspects ............................................. 380
Weakly Cyanogenic Plants and HCN Detoxification ................ 381
Strongly Cyanogenic Plants .................................... 382
Cyanogenic Glucosides in Cassava .............................. 383
Projects to Decrease the High HCN-Potential in Cassava Tubers ....... 384
Conclusions ................................................... 387
INTRODUCTION
Plants that are able to liberate significant amounts of HCN are referred to
as cyanogenic. The main source of the cyanide is the so-called cyanogenic glu-
cosides. These compounds consist of a-hydroxynitriles, also called cyanohydrins,
369
370 D. SELMAR
which are stabilized by a sugar. Nearly 3,000 plant species have been reported to
be cyanogenic. 1.2 The HCN liberated from cyanogenic plants is thought to be an
important ecological factor, e.g. in plant defence against herbivores, and must be
clearly distinguished from low levels ofHCN production during ethylene biosyn-
thesis in intact plants. 3 The amounts ofHCN produced during ethylene synthesis
are several magnitudes lower than those concentrations resulting from cyano-
genesis due to tissue disruption.
In addition to several reviews dealing with general aspects of cyanogene-
sis and cyanogenic glycosides,I.4-6 other more specialized ones have been
published. Among these are reviews on the occurrence and distribution of
cyanogenic glycosides/·8 structures,2,7 structural deterrnination,9 function, 10, II
biosynthesis,2,12,13 and occurrence and toxicity in foodstuffs, 14-16 This chapter sum-
marizes briefly the recent progress in the research on cyanogenic glucosides, with
special emphasis on their significance in plant derived foodstuffs and related
nutritional aspects,
It is generally known that cyanide or HCN is toxic, as it inhibits numerous
metabolic processes, The acute toxicity of HCN and cyanide, respectively, is a
consequence of their affinity for various heavy metals, such as iron or copper,
with which they form cyano complexes. Most important is the effect on
cytochromes that results in an efficient inhibition of respiration. In addition,
numerous other metabolic processes are affected (for review see ref, 17). The
lethal dose of cyanide for humans is considered to be about 1 mg per kg body
weight. 18 They are widespread, present in most foodplants, but fortunately their
concentration often is low.
CYANOGENESIS
Cyanogenic glucosides are stable compounds. In the plant, they accumu-

late in the vacuole,19,20 whereas the corresponding hydrolytic enzymes, typically
~-glucosidases, are localized either in protein bodies 21 or in the apoplasmic cell
wall (for review see ref. 22). Only when cyanogenic plants are damaged and cells
are disrupted is HCN liberated. This process, called cyanogenesis, is initiated by
loss of cell integrity, leading to the contact of cyanogenic glycosides with their
hydrolytic enzymes. As a consequence, HCN is liberated.
This post mortem HCN-Iiberation consists of two steps: first, the
cyanogenic glucosides are hydrolyzed by ~-glucosidases (Fig. I). The hydrox-
ynitriles produced are unstable and dissociate in a second step to yield HCN and
a carbonyl compound. In general, this reaction is catalyzed by hydroxynitrile
Iyases (for review see refs. 4 and 24). In addition to numerous plants that accu-
mulate cyanogens at high concentrations, many contain only small amounts. They
may be universal, and the presence or absence may be just a question of sensi-
tivity of detection.
The biochemical properties of ~-glucosidases involved in cyanogenesis
CYANIDE IN FOODS 371
B - glucosidase
glucose
cyanogenic hydroxynitrile
glucoside
<
(cyanohydrin)
hydroxynitrile
lyase
hydroxynitrile HeN
Figure 1. Cyanogenesis in plants.
differ. In plants like A/ocasia or Trig/ochin, they are highly specific for the
cyanogenic glucoside triglochinin that occurs in these plants. 24 .25 In contrast, the
cyanogenic ~-glucosidases of Sorghum or of Prunus have moderate specificity;
in addition to hydrolyzing the cyanogenic glucosides present in the plants,
they are able to hydrolyze various other glucosides. 26 •27 In flax, cassava, and
Hevea, the related cyanogenic ~-glucosidases are nonspecific and are capable of
hydrolyzing a large variety of different glucosides?8-30 Despite their observed dif-
ferences in substrate specificity, all ~-glucosidases share one important feature:
they do not hydrolyze cyanogenic glycosides that contain two sugar moieties. 22
The inability to hydrolyze cyanogenic diglucosides is of importance with regard
to their translocation (see below).
BIOSYNTHESIS AND STRUCTURES
The biosynthesis of cyanogenic glucosides was studied intensively in

the lab of Eric E. Conn, and later on in Birger Moller's lab. The precursors of
cyanogenic glucosides are amino acids. During a complex oxidative parthway,
catalyzed by a multienzyme complex, amino acids are converted to hydroxy-
nitriles, which in a final step are glucosylated to yield the glucosides. As no
intermediates are detectable, this pathway has been considered to be a chanelled
biosynthetic process, meaning that the product of one enzyme is used directly as
the substrate for the next enzyme of the complex without any chance of leaving
the site of biosynthesis. Recently, Birger MolJer and his group elucidated the
entire biosynthetic pathway of dhurrin, the cyanogenic glucoside that occurs in
Sorghum (Fig. 2). Surprisingly, the biosynthesis of the hydroxynitriles is catalyzed
by only two enzymes, namely two multifunctional cytochrome P450s. The first,
P450,yr catalyzes the conversion of tyrosine to an aldoxime. 3 !33 This transforma-
tion includes two hydroxylations and a subsequent decarboxylation. The second,
~COOH W
--l
tv
0, HO~ NH2
_ _ _ _ _ _ _ _ _ _ tyrosine
__~
......
, NADPH - - - - - - - - - - - - - - ------
cytochrom P 450 (
, yr \,
NADP+
, ~COOH
2
~COOH+C02 ~H,
NH O
,,HO ~ I
6H > ,~ HO ~ /
iN,, HN.N~'
O· , ,
NADPH NADP+
HO OH~20 OH,
(E)-p-hydroxyphenyl-
, N-hydroxytyrosine _ _ _ _ _ _ _ _ _ _ _ ~~-~h~d:o~~r~i~e _ _ _ _ _ _ _ ~c~a~o~i~e _ ,;'
,
, cytochrom P 450 ox ,,
HO
D'J
I~ /
N
HO
(Z)-p-hydroxyphenyl-acetaldoxime
G'ucosy'- Glucose
, O2 dO -
transferase -:P'
~ "'-- ~ CN
I
,~ CN ~ ~
~:N I
HO ::::::,... NADPH NADP+ , HoN HO ~ o
p-hydroxyphenyl- , Vi
p-hydroxymandelonitrile dhurrin t11
, acetonitrile , r
a::
;J>
;>::l
Figure 2. Biosynthesis of cyanogenic glucosides. Biochemical reactions involved in dhurrin biosynthesis as outlined by reference 2.
P450 ox> converts this aldoxime to the hydroxynitrile, which then is glucosylated
by a UDP-glucose dependent glucosyltransferase (For review see ref. 2).
Presently, such comprehensive knowledge, including the sequences of the
cytochromes involved, is available only for the Sorghum system. Nevertheless,
due to strong homologies in the properties of the biosynthetic apparatus of other
cyanogenic plants, learned by numerous labeling experiments with 14C-amino
acids, we know that nearly all cyanogenic glucosides are synthesized by an anal-
ogous mechanism. 2 .4 Astonishingly, only six amino acids are used as precursors.
Various structures and their amino acid precursors are listed in Table I. Valine
and isoleucine are the precursors of simple aliphatic cyanogens, such as Iinamarin
or lotaustralin. Branched cyanogenic glucosides, such as heterodendrin, are
derived from leucine. Aromatic cyanogens, like prunasin or dhurrin, are synthe-
sized either from phenylalanine or tyrosine, depending on their hydroxylation. In
addition to these five protein amino acids, cyclopentenyl glycine is a nonprotein
amino acid that serves as precursors for cyanogenic glucosides. A series of
cyclopentenyl glycine derived compounds, such as deidaclin, can be detected in
various members of the Passifloraceae. 6.34
From these basic structures, various derivatives are known, in which either
the aglycone is modified by further hydroxylation, or in which additional sugar
molecules are attached, e.g. amygdalin which corresponds to a prunasin gluco-
side. Overall, about 60 different structures all derived from these six basic struc-
tures, are known (see refs. I, 2, 7).
TRANSLOCATION
Various transport studies have shown that cyanogenic glucosides are

translocated within plants and that monoglucosides are glucosylated to yield
diglucosides before they are translocated. Such diglucosidic transport metabo-
lites are required to protect the compounds against cleavage by apoplastic ~
glucosdase. In the course of long distance transport, all substances must pass
the apoplasm, for instance, during phloem loading,35 or in the course of an
endosperm-cotyledon-passage. The high activity of apoplasmic ~-glucosidases
would split cyanogenic monoglucosides immediately when they entered the
apoplasmic space. In contrast, the diglucosides are resistant against cleavage by
the apoplastic enzymes (for review see refs. 4, 22).
After monoglucosides are glucosylated to diglucosides in source-tissues,
they are translocated via phloem into sink-tissues where they are cleaved. Hydol-
ysis may occur either by a sequential or by a simultaneous mechanism. 36 In the
first case, only the terminal glucose moiety is detached (Fig. 3). The original
cyanogenic monoglucoside, which subsequently is accumulated in the vacuoles
of the sink-tissue, is again produced, resulting in a simple translocation of
cyanogenic glucosides from source to sink by the means of a diglucosidic
374 D. SELMAR
Table 1. Precursors and structures of various cyanogenic glucosides

Precursors Basic structures Derivatives (examples)
/NH2 linustatin = linamarin-6' glucoside
, '"
CH 3 CN
CH3 /C~
~CH COOH ",C,
CH 3 0- glucose
CH 3
valine linamarin
/NH2 neolinustatin = lotaustralin-6'-glucoside

CH3-CH.1! ",CN
CH3-CH2 /C~
'CH COOH "'~0- glucose
'"
CH3
CH3
isoleucine (R)-lotaustralin
(S)-epilotaustralin
CH3 proacacipetalin = heterodendrin-2,3-en

~Ct' ",CN cardiospermin =
CH3 CH, 4-hydroxy-proacacipetalin
O-glucose proacaciberin =
proacacipetal in -6' -arabinoside
leucine (S)-heterodendrin
(R)-epiheterodendrin
taraktophyllin =
4-(S)-hydroxy-deidaclin
taraktophyllin-6' -rhamnoside
gynocardin =
cyclopentenylglycine (R )-deidaclin 4-(S)-5-(R)-tetraphyllin A
(S)-tetraphyllin A
H, ",CN amygdalin = prunasin-6'glucoside
@ ~
O-glucose
holocalin = m- hydroxyprunasin
prunasin-6'-malonate vicianin =
prunasin-6' -arabinoside
phenylalanine (R)-prunasin
(S)-sambunigrin
t:::J
H CN proteacin = p-glycosyloxy-dhurrin
''''
C* dhurrin-6'-glucoside nandinin =
'O-glucose 4' -caffeoyl-p-glycosy loxy-
mandelonitrile
HO
tyrosine (S)-dhurrin
(R)-taxiphyllin
sequential
CH3, /
CN diglucosidase
/C,
CH3 O-glucose
I
O-Glucose glucose
simultaneous
diglucosidase
CH3 , /
CN
/C,
CH3 O-glu~ose
O-glucose gentiobiose
asparagine .....1 - - - - - B-cyano- ......I----c:::

alanine
cysteine
Figure 3_ Sequential and simultaneous hydrolysis of cyanogenic diglucosides.
transport metabolite. In the second case, the two glucose moieties are split off as
gentiobiose, producing hydroxynitriles-which also occurs in post mortem
cyanogenesis-but now under controlled conditions. The HeN produced quanti-
tatively can be refixed by p-cyanoalanine synthase. This enzyme catalyzes the
reaction of cyanide and cysteine. 37 The resulting p-cyanoalanine is hydrolyzed to
asparagine. 38 In this manner, the nitrogen present in cyanogenic glucosides is
incorporated into asparagine and becomes available for the general nitrogen pool.
Simultaneous cleavage of cyanogenic glucosides results in their metabolization
(for details see ref. 4).
To summerize: In source-tissues, cyanogenic (mono)glucosides are mobi-
lized by glucosylation and translocated as diglucosides into sink-tissue where they
are hydrolyzed. Their metabolic fate is determined by the mode of hydrolysis:
Sequential cleavage leads to a simple translocation, whereas simultaneous clevage
initiates metabolization to non-cyanogenic compounds. Depending on the actual
ratio of activities of simultaneous to sequential diglucosidase, the one or the other
process is favored (Fig. 4).
FUNCTIONS
Ecological Significance
Many plants that are able to synthesize cyanogenic glucosides accumulate

them to a certain degree in their vacuoles. Because of the co-occurring hydrolytic
376 D. SELMAR
Imetabolization I Itranslocation I
source-
organs
Figure 4. Translocation and metabolic fate of cyanogenic glucosides.
enzymes, they are able to liberate relatively large amounts ofHCN when injured;
these plants are classified as highly or strongly cyanogenic. In contrast, other
plants that synthesize cyanogenic glucosides contain much smaller concen-
trations, often less than 10 nmol per g fresh weight. These are called weakly
cyanogenic plants.
The HCN liberated upon tissue damage-at least in strongly cyanogenic
plants-respresents an important ecological factor by repelling potential herbi-
vores (see refs 39, 40). Studies on the protective role of cyanogenesis also have
been performed (see refs 10, 11). These investigations have shown in general that
it is not the presence of cyanogenic glucosides but the ability of plants to liber-
ate HCN rapidly that represents the protective function. Surprisingly, in some
cases, the carbonyl compound, for instance, benzaldehyde, also acts as a repel-
lent. 41 If th~ protection is due to an effective cyanogenesis, the presence of the
corresponding hydrolytic enzymes is essential for the ecological function. Both,
the l3-glucosidases, which initialize and enable cyanogenesis, and the hydroxyni-
trile lyase are involved in the defensive strategy. The relatively unstable hydrox-
ynitriles may dissociate spontaneously; nevertheless, effective cyanogenesis
requires these lyases which may accelerate HCN-liberation up to 20-fold. 42
Evolutionary Aspects
The ability of plants to synthesize cyanogenic glucosides corresponds to an
ancient character. This can be deduced from their widespread occurrence through-
out the plant kingdom, as well as from the strong similarities in biosynthetic path-
ways in quite different plant groups. As the significance of cyanogenic glucosides

as repellents depends on effective cyanogenesis, the degradative enzymes must
also be ancient characters.
The ~-glucosidases of many cyanogenic and acyanogenic plants have been
isolated and characterized, and various related cDNA-sequences have been pub-
lished, e.g. those of the linamarase from Trifolium repens,43 and from cassava,44
and that of dhurrinase from Sorghum. 45 All these reveal high homologies to each
other and belong to family AI. The cassava enzyme is glycosylated, having high
mannose-type N-asparagine-linked oligosaccharides. Consistent with this struc-
ture and the extracellular localization of the active enzyme is the identification of
an N-terminal signal peptide. 44
The situation is more complex in regard to hydroxynitrile lyases. At present,
the c-DNAs of hydroxynitrile lyases from five plants have been cloned and
sequenced: Prunus serotina,46 Manihot escuienta,47 Sorghum bicoior,48 Hevea
brasiliensis,49 and Linum usitatissimum. 50 Recently, the first genomic clone of the
mandelonitrile lyase from Prunus serotina was sequenced. 51 The hydroxynitrile
lyase from Sorghum, which does not show strong homologies to the enzymes
from Prunus or Linum, is apparently related to the lyases from Hevea and
Manihot. This sorghum lyase is strongly homologous to carboxypeptidases,
whereas the lyases from Hevea brasiliensis exhibit strong homologies to a rice
protein whose function still is unknown. Obviously, these two groups of hydrox-
ynitrile lyases have different direct ancestors, carboxypeptidases and a protein x
of unknown function, respectively (Fig. 5). Nevertheless, both seem to have devel-
oped from a common, but more ancient, protein bearing a aJ~-hydrolase fold and
Sorghum bico/or
HNL
from
Prunus serotina
Hevea brasiliensis Prunus amygdalus
dehydrogenases
'1/
"alP-hydrolase fold"
'"p1 / a p-motif"
(catalytic triad) (nucleotid binding domain)
Figure 5. Evolutionary relationships of hydroxynitrile lyases according to reference 50.

378 D. SELMAR
a catalytic triad (Fig. 5). The sequence comparison of the hydroxynitrile lyases
from Prunus and Linum points in a similar direction. These hydroxynitrile lyases
reveal only slight homologies to each other. The lyase from Prunus seems to be
homologous to flavin-dependent dehydrogenases, but the lyase from flax is more
homologous to Zn-containing alcohol dehydrogenases. 5o (Fig. 5) Both groups of
dehydrogenases have their origin in a common putative ancestor, sharing a
nucleotide binding sequence and a ~a~-motif. From these data, it can be deduced
that hydroxynitrile lyases have developed convergently at least four times. 50
As mentioned, the ability to synthesize cyanogenic glucosides is an ancient
character. This means that in the course of evolution, plants existed that were
able to synthesize cyanogenic glucosides but which lacked hydroxynitrile lyases
and, thus, the ability to carry out efficient cyanogenesis. As the repellent effect
for potential herbivores depends on rapid cyanogenesis, the significance of
cyanogenic glucosides in these ancient plants must have been due to different
selective pressures.
Significance in Weakly Cyanogenic Plants
In order to obtain information about alternative functions, we have to focus

on weakly cyanogenic plants, whose concentrations of cyanogenic glucosides
are often less than 10 nmol per gram fresh weight, and frequently less than I nmol
per gram fresh weight. There are several possible explanations. In some plants,
cyanogenic glucosides accumulate to high levels only in certain organs or tissues,
whereas in others, only trace amounts are present, e.g., in the strongly cyanogenic
primary leaves of barley, between two- and four ~moles of cyanogens per g fresh
weight are accumulated, whereas, the seeds are only weakly cyanogenic,
often with less than one nmole per g fresh weight. Other examples are shown in
Table 2.
In other plants, all organs and tissues are weakly cyanogenic. Examples of
this are shown in Table 3. Concentrations in these plants are in the picomole
per g fresh weight range. The small amounts of cyanogenic glucosides and HCN
liberated during tissue injuries cannot cause any deterrent effect on herbivores.
Nonetheless, the presence of such low concentrations is inherited, and there is
likely to be an evolutionary advantage for their presence. Presently, little in this
context is known. Nevertheless, by considering known data from a new perspec-
tive, some suggestions for a putative function can be put forward.
Concentrations of cyanogenic glucosides in weakly cyanogenic plants are
in the same range as phytohorrnones and metabolic effectors or mediators. There
are some data that show that small amounts of HCN are able to alter plant devel-
opment. Tanaka and coworkers 61 were able to induce flowering of the short-day
plant Lemna paucicostata under continuous light when small concentrations of
HCN were added to the cultivation medium. Another effect of cyanide, one which
has been known for a long time, is the ability to break domancy of buds and seeds,
Table 2. Content of cyanogenic glucosides in different

organs of various cyanogenic plants
Cyanogenic glucosides
Plant species organs (nmol/g fresh weight)
Hordeum vulgare
Primary leaves 2,000-4,100
Seeds 0.2-13
Manihot esculenta
Leaves 9,000-17,000
Tubers 1,000-8,000
Seeds 10-50
Sorghum bicolor
Primary leaves 21,000-40,000
Seeds 0.1-50
Prunus avium
Seeds 1,850-2,500
Fruits 270-320
Leaves 0.4-3.0
The contents of cyanogenic glucosides in these materials have been
reported in references 52-59.
for example, the spouting of various plants that normally depends on vernaliza-
tion, can also be induced by application of cyanide. 62
Small amounts of HCN also are produced in the course of ethylene pro-
duction (Fig. 6). The biogenetic precursor of this gaseous plant hormone is 1-
aminocyc1opropane-l-carboxylic acid (ACC). ACC is oxidized by a specific
oxidase to produce equimolar quantities of ethylene, HCN, and CO 2 •3.63 Applica-
tion of exogenous ethylene to plants induces endogenous ethylene production.
Thus, exogenous ethylene application also leads to the liberation of small amounts
Table 3. Content of cyanogenic glucosides in various

weakly cyanogenic plants
Concentration
Plant species (nmollg fresh weight)
Carica papaya 0.5-6
Coffea arabica 0.04-3.7
Lactuca sativa 0.05-10
Lens culinaris 0.4-3.2
Oryza sativa 0.03-0.6
Saccharum officinarum 0.003-0.03
Solanum lycopersicum 0.5-6.0
Data on the content of cyanogenic glucosides is /Tom reference 60.
380 D. SELMAR
1-aminocyclopropane
1-carboxylic acid ethylene
(ACC)
Figure 6_ Biosynthesis of ethylene.
of HCN. Consequently, it cannot be excluded that some putative ethylene

effects are, in reality, HCN effects. In order to determine whether the HCN
derived from cyanogenic glucosides present in weakly cyanogenic plants corre-
sponds to a metabolic mediator that is able to influence a plants metabolism and
development, comprehensive research and new ideas are required.
CYANOGENIC GLUCOSIDES IN FOODS
General Aspects
Since cyanogenic glucosides are essentially ubiquitous, and since different
amounts are present in different plants and plant parts, quite different amounts of
cyanide are released upon tissue disruption. In general, during food preparation,
plant tissue is disrupted to different degrees. On the one hand, plant parts may be
consumed as they are, for instance, as fresh vegetables, fruits, and salad greens.
On the other hand, other parts are dried, cooked, ground, or mashed. Because of
these differences in tissue disruption, there are correspondingly strong differences
in the initiation of post-mortem processes such as cyanogenesis.
Depending on the degree of disruption, almost the entire amount of
cyanogenic glucosides originally present in the intact plant may still be present
in some foods, whereas, in others, nearly all of the cyanogenic glucosides are
cleaved, and only the corresponding degradation products are detectable. In sub-
sequent processing steps, the degree of decomposition and evaporation of the
volatile hydroxynitriles and gaseous HCN, respectively, depends on the process-
ing method, especially on the amount of heating. In order to provide solid and
comparable information on the quantity of these compounds in different prod-
ucts, in the following paragraphs, the term HCN-potential is used, which includes
the concentrations of all cyanogens, including the uncleaved cyanogenic gluco-
sides, the hydroxynitriles, free HCN, and cyanide. The data will be given in both
mol and in g of HCN equivalents that would be liberated if all the cyanogenic
compounds were completely dissociated to HCN.
Weakly Cyanogenic Plants and HCN Detoxification
Since the concentration of cyanogens in weakly or medium cyanogenic

plants is relatively low, due to the loss of HeN during processing of foods, the
actual HeN-potential in the corresponding foods becomes even lower. In Table
4, some examples are provided. Depending on the original content and on dif-
ferences in the loss during processing, values vary drastically. Even the highest
concentration shown in Table 4--nearly one thousand Ilg present in one kg cherry
jam or one liter banana juice-still corresponds to a low concentration. These
concentrations pose no risk to our health. Our bodies can deal with even higher
concentrations without problems. Although cyanide is a powerful toxin that
inhibits various metabolic processes, especially cell respiration, this is overcome
by an efficient detoxification mecanism involving the rhodanese system in our
liver. 65
Rhodanese catalyzes the reaction of cyanide with a sulphane sulphur
to form thiocyanate, also called rhodanide (Fig. 7). Whereas thiosulfate is
generally used as a sulphur donor in enzymatic assays, in vivo, various sulphane-
containing anions serve this function. 65 The thiocyanate produced subsequently
is excreted via the urine. By this mechanism, an adult can deal with several
mg of cyanide per day without problems. However, higher doses, 20-50mg,
Table 4. HeN-potential of foods derived from weakly

or medium cyanogenic plants
HCN-potential
(Ilmollkg) (Ilg /kg)
Foods derived from cereal
Wheat flour 0.8 22
White bread 0.4 11
Rye bread 0.1 3
Corn flakes 0.3 8
Popcorn 0.07 2
Rice, parboiled 0.04 I
Beer, pilsener 6.0 160
Fruit products
Peach jam 3.0 83
Cherry jam 36.5 980
Apple juice 2.1 57
Grapefruit juice 0.7 20
Banana juice 27.4 740
Pineapple juice 0.3 8
Strawberry juice 0.2 6
The data on the HeN-potential have been compiled from references 60,
64.
382 D. SELMAR
Irhodanese I
CN" + S-SOl" ~ SCN" + SOl" Figure 7. Detoxification of cyanide by
l
excretion
rhodanese. In mammalian liver cells, in
contrast to the artificial substrate thiosulfate,
which is used in standard enzymatic assays,
via urine various sui fane sulfur anions are used. 65
become critical, and about 60mg HCN per day is lethal. Concentrations
of cyanide in foods derived from weakly cyanogenic plants, all of which are in
the Ilg-range, are no problem for our detoxification mechanism and for our
health.
The limiting factor in the detoxification of cyanide is not the rhodanese
reaction, but sulphur availability and excretion of the resulting thiocyanate. 66
Consequently, people who consume cyanide rich foods, e.g., up to twenty mg
per day, have enhanced levels of thiocyanate in the blood which can create
severe health problems (see below). The small amounts of cyanide present in
foods derived from weakly cyanogenic plants do not significantly enhance the
thiocyante level in our blood and, thus, do not cause the dissorders described
below.
Only in special cases does the small amount of cyanide in foods derived
from weakly or medium cyanogenic plants create toxicological problems. Such
an example is the preparation of whisky and other alcoholic beverages. As men-
tioned, leaves of barley contain relatively high concentrations of cyanogenic
glucosides, mostly derived from leucine. Consequently, in the malt and in the
mash, cyanogens and their degradation products also are present. In the course
of the distillation procedure, resultant HCN reacts with ethanol to yield toxic
carbamates. 67 •68
Strongly Cyanogenic Plants
In contrast to weakly cyanogenic foods, the consumption of those that

contain much higher HCN-concentrations can, indeed, be dangerous. The HCN-
potential of these plants-some of which are used as foodplants-is magnitudes
higher than those concentrations so far described (Table 5), and is sufficiently
high that the consumption ofless than.ten grams of fresh bamboo sprouts or about
two hundred grams of fresh cassava tubers, for instance, could be lethal. It is
evident that such highly cyanogenic plant parts must be detoxified before they
can consumed safely. In principle, such processing is easy. First, compartmenta-
tion must be broken down in order to mix the hydrolytic enzymes with cyanogenic
precursors. Then, after hydrolysis, the produced HCN must be evaporated, for
instance by heating.
Table 5. HCN-potential of strongly cyanogenic food plants

HCN-potential
Plant species Tissue (mmollkg) (mg/kg)
Cassava tubers (max. range) 2-8 (1-25) 50-200 (25--650)
(Manihot esculenta) flour (gari) 0.02-11 0.5-300
Bamboo sprouts up to 300 up to 8,000
(Bambusa vulgaris)
Lima beans seeds 2-75 500--2,000
(Phaseolus lunatus)
Vetch seeds 2-20 50-500
(Vicia sativa)
Flax seeds 10-20 250-500
The HeN-potentials have been reported in references 15, 55, 56, 69, 70.
Cyanogenic Glucosides in Cassava
As toxicological problems related to highly cyanogenic foods are especially

severe with cassava products, this topic will be outlined in detail. Cassava
(Manihot esculenta) is one of the world's most important crops. In the tropics,
more than 400 million people depend on cassava tubers as a staple food. Many
consume as much as one kg of cassava products daily. The strongly cyanogenic
tubers contain high concentrations of linamarin. Consequently, they must be care-
fully detoxified before consumption. A classical method used in Central Africa
is the preparation of gari. By grinding or grating the tubers, compartmentation
is destroyed and cyanogenesis is initiated. In the resulting mash, cyanogenic glu-
cosides are hydrolyzed. Subsequent roasting guarantees complete hydrolysis of
cyanohydrins and evaporation of the prussic acid produced. Provided that heating
is not performed too soon after tissue disruption, and that the ~-glucosidases are
not destroyed before all cyanogenic glucosides are hydrolyzed, safe cassava prod-
ucts are produced in this manner. The residual HCN-potential in the correspond-
ing flour, called gari, in general is below one mg per kg (for reviews see refs. 56,
71,72).
Unfortunately, detoxification often is performed incompletly, due to inade-
quate methods or shortcuts in processing. In some arid regions of Africa, large
amounts of HCN are consumed as a result of eating improperly processed cassava
foods. Consequently, HCN-intoxication and related disorders are widespread.
Nevertheless, acute and lethal HCN intoxication is seldom seen. In contrast,
chronic intoxication is common. As mentioned, it is not only a direct effect of the
HCN that causes problems but also the detoxification product, thiocyanate. The
intake of large amounts of HCN leads to enhanced SCN concentrations in
the blood which, in tum, cause health problems (Table 6). Such disorders are
384 D. SELMAR
Table 6. Health disorders caused by cyanide

Acute cyanide intoxication
Hyperthroidism
endemic goitre
cretenism
Neurological disorders
tropical atactic neuropathy
spastic paraparesis (e.g. konzo)
paralysis
Diabetes (?)
Information compiled from refs. 73-76.
mainly due to the interference of SCN with iodine metabolism. Thus, long term
exposure to significant concentrations of HCN causes or aggravates iodine
deficiency disorders, which are expressed mainly as goiter and cretinism. 73
In addition, various investigations have suggested that the consumption of
cassava products with high residual cyanide content causes neurological and par-
alytic disorders, for instance, konzo and tropical ataxic neuropathy.74.75 In these
cases, however, the correct pathogenic mechanism is unknown, and it is unclear
whether there is a link to cyanide metabolism. There are also some indications
that long-term consumption of increased cyanide concentration in foodstuffs may
induce diabetes. 76 But apart from investigations in Nigeria, in which a correla-
tion between frequent consumption of improperly processed cassava products and
the frequency of aquired diabetes has been observed, no other data are available.
Projects to Decrease the High HCN-Potential in Cassava Tubers

Due to the severe problems caused by improperly detoxified cassava prod-
ucts, there is a strong demand for cassava plants that produce acyanogenic tubers,
or at least tubers with low concentrations of cyanogenic glucosides. Unfortu-
nately, no acyanogenic or even weakly cyanogenic cassava plants with other
required properties have been obtained by classical plant breeding. Some low
cyanogenic varieties are available, but they taste much sweeter than commonly
cultivated varieties, and this seems to be an undesired characteristic. Bitter taste,
which is preferred by most cassava consumers, is not directly attributed to the
presence of cyanogenic glucosides. Bitterness and content of cyanogenic gluco-
sides is generally correlated, 77 however, several sweet varieties that contain
the same or higher concentrations of cyanogens as bitter-tasting varieties are
known,78.79 and some have more. 80 A compound that putatively is responsible for
the bitter taste has been identified as isopropylapiosylglucoside. 81
Generation of cassava plants with weakly cyanogenic tubers should be
approchable by gene technology. Such efforts first should concentrate on the
biosynthetic pathway. Elimination of biosynthesis of cyanogenic glucosides by a

total knockout of the two cytochromes involved should yield acyanogenic cassava
plants, which-apart from their acyanogenic character-should not differ from
the standard varieties. However, due to the lack of chemical protection provided,
these acyanogenic plants may be fed on extensively by generalist herbivores.
Thus, cultivation of these transgenic plants may become difficult.
Transgenic plants also may be generated in which the content of cyanogenic
glucosides can be drastically reduced only in the tubers. This may be achieved by
modifYing the transport of cyanogens within the plant. Although cassava tubers
are able to synthesize some cyanogens,82 most of those present in the tubers are
synthesized in the leaves 83 ,84 and subsequently translocated into the tubers. 85 After
import of the transport metabolite linustatin into cassava tubers, this diglucoside
is hydrolyzed preferentially by a sequential diglucosidase, yielding linamarin,
which accumulates 86 and which causes the undesireble nutritional problems (Fig.
8a). The simultaneous diglucosidase is detectable only in traces, and, thus, metab-
Manihot escu/enta
leaves
(source)
Itranslocation I
transport
via phloem
[ accumulation 1
sequential linamarin
diglucoSid~
tuber
linustatin h-y-d-ro-Iy-s-is'l
rl .
(sink) . asparagine
simultaneo~ acetone / ' "
diglucosidase cyanohydrin I metabolization I
Figure SA. Synthesis, translocation, and degradation of cyanogenic glucosides in Cassava.

386 D. SELMAR
Manihot esculenta
Ibiosynthesis)
~
leaves
(source)
Itranslocation I
transport
"aspired
via phloem metabolism"
sequential
di9lucOSida~
linamarin Iaccumulation I
tuber linustatin Ihydrolysis)
(sink)
simultane~ acetone ' "
diglucosidase cyanohydrin
Figure 8B. Synthesis, translocation, and degradation of cyanogenic glucosides in genetically

modified Cassava plants. Due to the tuber specific overexpression of simultaneous diglucosi-
dase and concomitant suppression of sequential diglucosidase, the major amount of cyanogenic
glucosides imported into the tubers should be metabolized to non-cyanogenic compounds.
olization of the imported cyanogens does not appear to play any role in cassava.
Our goal for producing modified cassava plants is outlined in Figure 8b. Tuber
specific suppression of the sequential diglucosidase and concomitant over-
expression of the simultaneous diglucosidase should change the metabolic
fate of linustatin imported into the tubers, causing it to be metabolized into
non-cyanogenic compounds. The other enzymes required for this process, ~
cyanoalanine synthase and cyanoalanine hydrolase, are present naturally in
cassava tubers of the original varieties. 87 In such plants, the leaves should contain
the same amount of cyanogenic glucosides as the original varieties, but in the
root and tubers, the cyanogenic potential should be reduced significantly. Whether
these target plants indeed will have the desired properties, or whether-for
instance due to the initiation of supplementary synthesis of cyanogenic gluco-
sides in the tubers-the cyanogenic potential will again be high, will only be
clarified by experiments.
CONCLUSIONS
Cyanogenic glucosides are widespread in the plant kindom. In addition to

strongly cyanogenic plants, many others exist in which only small amounts of
cyanogenic glucosides are present. Whereas in strongly cyanogenic plants, the
significance of these natural compounds is determined by repellent effects toward
potential herbivores by the HCN liberated upon tissue disruption, the function of
cyanogenic glucosides in weakly cyanogenic plants is still unknown. The complex
biosynthetic pathway involved in the synthesis of these compounds is complex.
The putative action of the HCN as a metabolic mediator or its involvement in sig-
naling could explain the general occurrence of cyanogenic glucosides.
Since they occur widely and also are present in the plants used for human
nutrition, at least low levels of cyanogenic glucosides and HCN are liberated
during food preparation and are present in most of our foods. With respect to tox-
icity, only foods derived from plants that accumulate relatively high levels of
cyanogenic glucosides are relevant. The small amounts ofHCN present in weakly
cyanogenic plants are detoxified easily by the rhodanese system in our liver, and
thus, do not cause problems for our health.
In contrast, the large amounts of HCN present in foods derived from
strongly cyanogenic plants, e.g. cassava, lima beans, or bamboo, can cause
significant chronic health problems. Our detoxification mechanism is able to
handle relatively high concentrations of HCN; a lethal dose of cyanide is about
I mg HCN per kg bodyweight. Consequently, acute cyanide-intoxication is
seldom observed, however, several chronic HCN-related disorders are known.
Long term intake of enhanced cyanide concentrations leads to increased SCN-
concentrations in the blood, which, in turn, leads to interference with iodine
metabolism, which causes or aggravates iodine deficiency disorders, which are
expressed mainly as goiter and cretinism, as observed in various arid regions in
Africa. In these areas, people consume large amounts ofHCN as a result of eating
improperly processed cassava-derived foods. Other HCN-related neurological
disorders, such as tropical atactic neuropathy or spastic paraparesis, are
widespread.
Cassava is one of the most important food crops in the tropics. Due to the
health problems related to consumption of improperly processed cassava prod-
ucts, there is a strong demand for those plants containing only small amounts of
cyanogenic glucosides in their tubers. Unfortunately, until now, few such vari-
eties have been produced by classical plant breeding. The creation of the needed
cassava plants should be approchable by gene technology, either by a total knock-
out of the pathway for biosynthesis of cyanogenic glucosides, or by prevention
of the transport and accumulation of these natural products in the tubers. In order
to accomplish these projects successfully, further knowledge of physiology
and biochemistry as well as of ecological aspects of cyanogenic glucosides are
required.
388 D. SELMAR
ACKNOWLEDGMENTS
I wish to thank Dr. Dave S. Seigler (University of Illinois, Urbana) for crit-
ical reading the manuscript and helpful suggestions related to the sientific content
as well as for linguistic improvements.
REFERENCES
1. SEIGLER, D.S. 1991. Cyanide and cyanogenic glycosides. pp. 35-77 in Herbivores:
Their Interaction with Secondary Plant Metabolites, Volume I: The Chemical Participants
(G.A. Rosenthal and M.R. Berenbaum, eds.), Academic Press, San Diego.
2. M0LLER, B.L., SEIGLER, D.S. 1998. Biosynthesis of cyanogenic glucosides, cyano-
lipids and related compounds. pp. 563-609 in Plant Amino Acids: Biochemistry and
Biotechnology (B.K. Singh, ed.), Dekker Press, New York.
3. JOHN, P. 1997. Ethylene biosynthesis: The role of I-aminocyclopropane-l-carboxylate
(ACC) oxidase, and its possible evolutionary origin. Physiol. Plantarum 100(3): 583-592.
4. SELMAR, D. 1999. Cyanogenic glycosides, glucosinolates and non-protein amino acids.
in Annual Plant Reviews Volume 2: The Role of Secondary Metabolites and Their
Utilization in Biotechnology (M. Wink, ed.), in press, Sheffield Academic Press.
5. POULTON, IE. 1990. Cyanogenesis in plants. Plant Physiol. 94: 401-405.
6. NAHRSTEDT, A. 1992. The biology of the cyanogenic glycosides: New developments.
pp. 249-269 in Annual Proceedings of the Phytochemical Society of Europe 29: Nitro-
gen Metabolism in Plants (K. Mengel and DJ. Pilbeam, eds.), Oxford University Press,
Oxford.
7. NAHRSTEDT, A. 1987. Recent developments in chemistry, distribution and biology of
the cyanogenic glycosides. pp. 213-234 in Annual Procedings of the Phytochemical
Society of Europe 24: Biologically Active Natural Products (K. Hostettmann, PJ. Lea,
eds.), Oxford University Press, Oxford.
8. HEGNAUER, R. 1986. Chemotaxonomie der Pflanzen Vol. VII, Birkhauser Verlag,
Basel-Stuttgart.
9. BRINKER, A.M., SEIGLER, D.S. 1992. Determination of cyanide and cyanogenic gly-
cosides. pp. 359-381 in Modern Methods of Plant Analysis. New Series, Volume 13:
Plant Toxin Analysis (H.F. Linskens and IF. Jackson, eds.), Springer, Berlin.
10. NAHRSTEDT, A. 1985. Cyanogenic compounds as protecting agents for organisms.
Plant Syst. Evol. 105: 35-47.
11. KAKES, P. 1990. Properties and function of the cyanogenic system in higher plants.
Euphytica 48: 25-43.
12. HALKIER, B.A., SCHELLER, H.V, M0LLER, B.L. 1988. Cyanogenic glucosides: The
biosynthetic pathway and the enzyme system involved. pp. 49-61 in Cyanide Compounds
in Biology (D. Everett, S. Harnett, eds.), Wiley & Sons, Chichester, UK.
13. CONN, E.E. 1988. Biosynthetic relationship among cyanogenic glycosides, glucosino-
lates, and nitro compounds. pp. 143-154 in Biologically Active Natural Products: Poten-
tial Use in Agriculture (G. Cutler, ed.), American Chemical Society, Washington, DC.
14. NAHRSTEDT, A. 1993. Cyanogenesis is foodplants. pp. 107-129 in Annual Proceed-
ings of the Phytochemical Society of Europe: Phytochemistry and Agriculture (T.A. van
Beek, H. Breteier, eds.), Oxford University Press, Oxford.
15. POULTON, IE. 1989. Toxic compounds in foodstuffs: Cyanogens. pp. 38 1-40 I in Food
Proteins. (lE. Kinsella and WG. Soucie, eds.), American Oil Chemists' Society, Cham-
paign, IL.
16. JONES, D.A. 1998. Why are so many food plants cyanogenic? Phytochemistry 47(2):
155-162.
17. SOLOMONSON, L.P. 1981. Cyanide as a metabolic inhibitor. pp. 11-28 in Cyanide in
Biology (B. Vennesland, E.E. Conn, C.J. Knowles, 1 Westley, F. Wissing, eds.), Acade-
mic Press, London.
18. MONTGOMERY, R.D. 1969. Cyanogens. in Toxic Constituents of Plant Foodstuffs (I.E.
Liener, ed.), Academic Press, London.
19. SAUNDERS, lA., CONN, E.E. 1977. Subcellular localization of the cyanogenic gluco-
side in Sorghum by autoradiography. Plant Physiol. 59: 647-652.
20. GRUHNERT, Ch., BIEHL, B., SELMAR, D. 1994. Compartmentation of cyanogenic glu-
cosides and their degrading enzymes. Plant Physiol. 195: 36-42.
21. SWAIN, E., LI, C.P., POULTON, lE. 1992. Tissue and subcellular localization of
enzymes catabolizing (R)-amygdalin in mature Prunus serafina seeds. Plant Physiol.
100(1): 291-300.
22. SELMAR, D. 1993. Apoplastic occurrence of cyanogenic ~-glucosidases and conse-
quences for the metabolism of cyanogenic glucosides. pp. 191-204 in The Biochemistry
and Molecular Biology of ~-Glucosidases (A. Esens, ed.), American Chemical Society,
Washington.
23. POULTON, lE. 1988. Localization and catabolism of cyanogenic glucosides. pp. 67-81
in Cyanide Compounds in Biology (D. Everett, S. Harnett, eds.), Wiley & Sons,
Chichester.
24. HOSEL, W, NAHRSTEDT, A. 1975. Spezifische Glucosidasen fUr das Cyanglucosid
Triglochinin, Reinigung und Charakterisierung von ~-Glucosidasen aus Alocasia macr-
orrhiza, Schott). Hoppe-Seyler's Z. fUr Physiol. Chemie 356: 1265-1275.
25. NAHRSTEDT, A., HOSEL, W, WALTHER, A. 1979. Characterisation of cyanogenic
glucosides and -glucosidases in Triglochin maritima seedlings. Phytochemistry 18:
1137-1141.
26. HOSEL, W, TOBLER, I., EKLUND, S.H., CONN, E.E. 1987. Characterization of~
glucosidases with high specificity for the cyanogenic glucoside dhurrin in Sorghum
hicolor (L.) Moench seedlings. Arch. Biochem. Biophys. 252: 152-162.
27. KUROKI, G.W, POULTON, lE. 1986. Comparison of kinetic and molecular properties
of two forms of amygdalin hydrolase from black cherry (Prunus serafina Ehrh.) seeds.
Arch. Biochem. Biophys. 247: 433-439.
28. FAN, T.WM., CONN, E.E. 1985. Isolation and characterization of two ~-glucosidases
from flax seeds. Arch. Biochem. Biophys. 243: 361-373.
29. YEOH, H.-H. 1989. Kinetic properties of ~-glucosidase from cassava. Phytochemistry
28: 721-724.
30. SELMAR, D., LIEBEREI, R., BIEHL, B., VOIGT, 1 1987. Linamarase in Hevea-a non-
specific ~-glycosidase. Plant Physiol. 83: 557-563.
31. KOCH, B.M., SIBBESEN, 0., HALKIER, B.A., SVENDSEN, 1., MOLLER, B.L. 1995.
The primary sequence of cytochrome P450tyr, the multifunctional N-hydroxylase cat-
alyzing the conversion of L-tyrosine to p-hydroxyphenylacetaldehyde oxime in the
biosynthesis of the cyanogenic glucoside dhurrin in Sorghum hicolor (L.) Moench. Arch.
Biochem. Biophys. 323(1): 177-186.
32. SIBBESEN, 0., KOCH, B., HALKIER, B.A., MOLLER, B.L. 1994. Isolation of the
heme-thiolate enzyme cytochrome P-450-TYR, which catalyzes the committed step in
the biosynthesis of the cyanogenic glucoside dhurrin in Sorghum hicolor (L.) Moench.
Proc. Natl. Acad. Sci. USA 91(21): 9740-9744.
33. SIBBESEN, 0., KOCH, B., HALKIER, B.A., MOLLER, B.L. 1995. Cytochrome P-450-
390 D. SELMAR
TYR is a multifunctional heme-thiolate enzyme catalyzing the conversion of L-tyrosine

to p-hydroxy-phenylacetaldehyde oxime in the biosynthesis of the cyanogenic glucoside
dhurrin in Sorghum bie%r (L.) Moench. 1 BioI. Chern. 270(8): 3506-3511.
34. OLAFSDOTTIR, S., ANDERSEN, lV., JAROSZEWSKI, J.W 1989. Cyanohydrin gly-
cosides of Passifloraceae. Phytochemistry 28(1): 127-132.
35. VAN BEL, A.lE. 1989. The challenge of symplastic phloem loading. Botanica Acta 102:
183-185.
36. KUROKI, G., LIZOTTE, P.A., POULTON, lE. 1984. Catabolism of (R)-amygdalin and
(R)-vicianin by partually purified p-glucosidases from Prunus serotina Ehrh. and
Davallia triehomanoides. Zeit. Naturforsch. 39: 232-239.
37. BLUMENTHAL, G.S., HENDRICKSON, H.R., ABROL, YP., CONN, E.E. 1961.
Cyanide metabolism in higher plants. I BioI. Chern. 243: 5302-5307.
38. CASTRIC, P.A., FARNDEN, K.F., CONN, E.E. 1972. Cyanide metabolism in higher
plants. 5: The formation of asparagine from p-cyanoalanine. Arch. Biochem. Biopys. 152:
62-69.
39. WOODHEAD, S., BERNAYS, E. 1977. Changes in release rates of cyanide in relation
to palatability of Sorghum to insects. Nature 279: 235-236.
40. BERNAYS, E.A., CHAPMAN, R.E, LEATHER, E.M., MCCAFFERY, A.R., MODDER,
WWD. 1977. The Relationship of Zonoeerus variegatus with cassava. Bull. Ent. Res. 67:
391--404.
41. PETERSON, S.c. 1986. Breakdown products of cyanogenesis. Repellency and toxicity
to predatory ants. Naturwiss. 73: 627-628.
42. SELMAR, D., LIEBEREI, R., BIEHL, 8., CONN, E.E. 1989. a-Hydroxynitrile Lyase in
Hevea brasiliensis and its Significance for Rapid Cyanogenesis. Physiol. Plantarum 75:
97-101.
43. OXTOBY, E., DUNN, M.A., PANCORO, A., HUGHES, M.A. 1991. Nucleotide and
derived amino acid sequence of the cyanogenic p-glucosidase (linamarase) from white
clover (Trifolium repens L.). Plant Molec. BioI. 17(2): 209-220.
44. HUGHES, M.A., BROWN, K., PANCORO, A., MURRAY, B.S., OXTOBY, E.,
HUGHES, 1. 1992. A molecular and biochemical analysis of the structure of the
cyanogenic p-glucosidase (linamarase) from cassava (Manihot eseulenta Cranz) Arch.
45. CICEK, M., ESEN, A. 1995. Cloning and sequencing of a cDNA coding for p-
glucosidase (dhurrinase) from Sorghum bieolor (L) Moench. Plant Physiol. 109(4):
1497.
46. CHENG, J.P., POULTON, IE. 1993. Cloning of cDNA of Prunus seratina (R)-(+)-
mandelonitrile lyase and identification of a putative FAD-binding site. Plant Cell Physiol.
34(7): 1139-1143.
47. HUGHES, 1., CARVALHO, F.1.P., HUGHES, M.A. 1994. Purification, characterization,
and cloning of alpha-hydroxynitrile lyase from cassava (Manihot eseulenta Crantz). Arch.
48. WAJANT H., MUNDRY, K.W, PFIZENMAIER, K. 1994. Molecular cloning of hydrox-
ynitrile lyase from Sorghum bieolor (L.). Homologies to serine carboxypeptidases. Plant
Molec. BioI. 26(2): 735-746.
49. HASSLACHER, M., SCHALL, M., HAYN, M., GRIENGL, H., KOHLWEIN, S.D.,
SCHWALB, H. 1996. Molecular cloning of the full-length cDNA of (S)-hydroxynitrile
lyase from Hevea brasiliensis. I BioI. Chern. 271: 5884-5891.
50. TRUMMLER, K., WAJANT, H. 1997. A novel class of hydroxynitrile Iyases. 1 BioI.
Chern. 272(8): 4770--4774.
51. HU, Z., POULTON, IE. 1997. Sequencing, genomic organization, and preliminary pro-
motor analysis of black cherry (R)-( +)-mandelonitrile lyase gene. Plant Physio!. 115(4):
1359-1369.
52. ERE, N., ZINSMEISTER, H.D., LEHMANN, G., MICHELY, D. 1981. Der Blausiiurege-
halt von Getreidearten gemii~igter Klimazonen. Z. Lebensm. Unters. Forsch. 173: 176-
179.
53. MONTGOMERY, R.D. 1969. Cyanogens. pp. 143-157 in Toxic Constituents of Plant
Foodstuffs (I.E. Liener, ed.), Academic Press, London.
54. NARTEY, F. 1968. Studies on cassava, Manihot usitatissimum. Phytochemistry 20:
1311-1314.
55. NARTEY, F. 1981. Cyanogenesis in tropical feeds and foodstuffs. pp. 115-132 in Cyanide
in Biology (B. Vennesland, E.E. Conn, C.1. Knowles, J. Westley, F. Wissing, eds.),
Academic Press, London.
56. OKE, 0.1. 1994. Eliminating cyanogens from cassava through processing: Technology
and tradition. Acta Horticulture 375: 163-174.
57. ZINSMEISTER, H.D., ERE, N. 1981. Cyanogene Glycoside in Getreidearten ver-
schiedener Klimazonen. Lebensmitte1chem. u. gerichtl Chern. 35: 55-59.
58. ZINSMEISTER, H.D., ERE, N., LEHMANN, G. 1980. Der Blausiiuregehalt tropischer
und subtropischer Getreideaten. Z. Lebensm. Unters. Forsch. 171: 170-173.
59. NAHRSTEDT, A. 1971. Zur Cyanogenese von Prunus avium. Phytochemistry II:
3121-3126.
60. LANG, I. 1990 Cyanogene Verbindungen in Nahrungs- Gewiirz- und
Genu~mittelpflanzen sowie in Nahrungs- und Genu~mitteln. Masters thesis, University
of Saarbriicken.
61. TANAKA, 0., CLELAND, c., BEN-TAL, Y. 1983. Effect of ferricyanide, ferrocyanide
and KCN on growth and flowering on the short-day plant Lemna paucicostata. Plant and
Cell Physiology 24(4): 705-711.
62. GASSNER, G., Heuer, W 1927. Praktische Anleitung zum Friihtreiben von Pflanzen
mittels Blausiiure. Verlagsbuchhandlung Paul Parey, Berlin.
63. YANG, S.F., HOFFMANN, N.E. 1984. Ethylene biosynthesis and its regulation in higher
plants. Ann. Rev. P!. Physio!. 35: 155-189.
64. LEHMANN, G., ZINSMEISTER, H.D., ERB, N., NEUNHOEFFER, O. 1979. Uber den
Blausiiuregehalt von Getreiden und Getreideprodukten Z. Erniihrungswiss. 18: 16-22.
65. WESTLEY, J. 1981. Cyanide and sulfane sulfur. pp. 61-76 in Cyanide in Biology
(B. Vennesland, E.E. Conn, C.1. Knowles, J. Westley, F. Wissing, eds.), Academic Press,
London.
66. ROSLING, H. 1994. Measuring effects in humans of dietary cyanide exposure from
cassava. Acta Horticulture 375: 271-283.
67. MACKENZIE, WM., CLYNE, A.H., MACDONALD, L.S. 1990. Ethyl carbamate for-
mation in grain-based spirits: Part II. The identification and determination of cyanide
related species involved in ethyl carbamate formation in scotch whisky. J. Inst. Brewing
96(4): 223-232.
68. COOK, R., MCGAIG, N., MCMILLAN, J.M.B., LUMSDEN, WB. 1990. Ethyl caramate
formation in grain-based spirits: Part III. The primary source. J. lnst. Brewing 96(4):
233-244.
69. BAUDOIN, J.P., BARTHELEMY, J.P., NDUNGO, V 1991. Variability of cyanide con-
tents in the primary and secondary genepools of the lima bean, Phaseolous lunatus 1.
Plant Genetic Resources Newsletter 85: 5-9.
70. SCHlLCHER, H., WILKENS-SAUTER, M. 1986. Quantitative Bestimmung cyanogener
Glycoside in Linum usitatissimum mit Hilfe der HPLC. Fette, Seifen, Anstriche 8:
287-290.
392 D. SELMAR
71. DUFOUR, D.L. 1994. Cassava in Amazonia: Lessons in utilization and safety from native
people. Acta Horticulture 375: 175-182.
72. ESSERS, S. 1995. Removal of cyanogens from cassava roots. CIP-DATA Koninklijke
Bibliothetheek, Den Haag, 131 p.
73. DELANGE, F., EKPECHI, L.O., ROSLING, H. 1994. Cassava cyanogenesis and iodine
deficiency disorders. Acta Horticulture 375: 289-293.
74. TYLLESKAR 1994. The association between cassava and the paralytic disease konzo.
Acta Horticulture 375: 331-339.
75. SPENCER, P 1994. Human consumption of plant material with neuro-toxic potential.
Acta Horticulture 375: 341-348.
76. AKANJI, A.O. 1994. Cassava intake and the risk of diabetes in humans. Acta Horticul-
ture 375: 349-359.
77. SUDARESAN, S., NAMBISAN, B., ESWARI AMMA, e.S. 1987. Bitterness in cassava
in relation to cyanogen content. Indian 1. Agric. Sci. B57: 37-40.
78. BOKANGA, M. 1994. Distribution of cyanogenic potential in cassava germplasm. Acta
Horticulture 375: 117-123.
79. NYE, M.M. 1991. The mis-measure of manioc (Manihot esculenta). Econ. Bot. 45(1):
47-57.
80. PEREIRA, 1.F., SEIGLER, D.S., SPLITTSTOESSER, WE. 1981. Cyanogenesis in sweet
and bitter cultivars of cassava. Hortscience 16: 776-777.
81. KING, N.L.R., BRADBURY, 1.H. 1995. Bitterness of cassava: Identification of a new
apiosyl glucoside and other compounds that affect its bitter-taste. 1. Sci. Food Agric.
68(2): 223-230.
82. DO, L., BOKANGA, M., M0LLER, B.L., HALKER, B.A. 1995. The biosynthesis of
cyanogenic glucosides in roots of cassava. Phytochemistry 39(2): 323-326.
83. BEDIAKO, M.K.B., TAPPER, B.A., PRITCHARD, G.G. 1981. Metabolism, synthetic
site, and translocation of cyanogenic glycosides in cassava. pp. 143-148 in Tropical Root
Crops: Research strategies for the 1980s (E.R. Terry, K.A. Oduro, F. Caveness, eds.),
International Development Research Centre, Ottawa, Canada.
84. MAKAME, M., AKORODA, M.O., HAHM, S.K. 1987. Effects of reciprocal stem grafts
on translocation in cassava. 1. Agric. Sci. Camb. 109: 605-608.
85. DEBRUIJN, G.H. 1973. The cyanogenic character of cassava (Manihot esculenta) pp.
43-48 in Chronic Cassava Toxicity (B. Nestel, R. MacIntyre, eds.), International Devel-
opment Research Centre, Ottawa, Canada.
86. SELMAR, D. 1994. Translocation of cyanogenic glucosides in cassava. Acta Horticul-
ture 375: 61-67.
87. NARTEY, F. 1969. Studies on cassava, Manihot usitatissimum, II. Plant Physiol. 22:
1085-1096.
Chapter Fifteen
PLANT ECOCHEMICALS FROM THE

VIEWPOINT OF PLANT DEFENSE
Junya Mizutani
Plant Ecochemicals Research Center

Eniwa R & BP Center Bldg.
3-1-1 Megumino Kita, Eniwa
Hokkaido 061-1374, Japan
Introduction ................................................... 394

Sesquiterpenoids from Chloranthus Species ......................... 394
Chloranthus japonicus ........................................ 394
Chloranthus serratus ......................................... 397
Lindenanolides from Other Chloranthus Species ................... 403
Oligostilbenes from Carex Species ................................. 403
Carex fedia var. miyabei ....................................... 403
Carex kobomugi ............................................. 405
Carex pumila ............................................... 405
Biogenetic Considerations and Roles in Carex Plants ............... 407
Sesquiterpenoids from Rosa rugosa ................................ 407
Antimicrobial Rugosal A and Related Compounds ................. 407
Other Sesquiterpenes ......................................... 410
Phenoxychromones and a Related Flavone ....................... .412
Glandular Trichome Exudate and Its Defensive Role ............... .412
Phytoalexin of Taraxacum ojJicinale ................................ 413
Lettucenin A ............ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 413
Isoflavonoids and Flavonoids from Iris pseudacorus .................. 415
Isoflavonoids ................................................ 415
Flavonoids .................................................. 416
Conclusion ................................................... 416
394 1. MIZUTANI
INTRODUCTION
Green plants produce a variety of secondary metabolites which play impor-

tant roles in complex interactions among living organisms, such as plant-plant,
plant-microorganism, and plant-insect, in the natural environment. Molisch
(1937) coined the term allelopathy and defined it as all effects that are either
directly or indirectly the results of chemicals transferred from one plant to another
plant. J Plants include algae, fungi, and the various microorganisms as well as
higher plants. Both Waller (1989)2 and Rizvi and Rizvi (1992)3 included plant-
insect and plant-higher animal interactions in the terms allelopathy and allelo-
chemicals. Lovett has been articulate in pointing out that many of the same
plant-produced chemicals that affect associated plants also influence other organ-
isms, and he has proposed expanding the context of allelopathy.4,5 We proposed
the new term "plant ecochemicals.,,6 Plant ecochemicals are those originating
from plants that may play important roles in complex interactions between plants
and other plants, microorganisms, or animals. In this chapter, some plant eco-
chemicals we have worked on are discussed from the viewpoint of plant defense.
SESQUITERPENOIDS FROM CHLORANTHUS SPECIES
Chloranthus japonicus
We first isolated two sesquiterpenoids (shizukanolide and 8,9-

dehydroshizukanolide) from the ether extract of Chloranthus japonicus Sieb.
(Chloranthaceae) by column chromatography.7 Shizukanolide, later called
shizukanolide A (Fig. 1), was obtained as colorless needles {mp 95-96.5°C,
[a]o20 + 200° (c = 0.21, CHCI 3)} in ca. 0.02% yield from the fresh aerial parts,
and its structure was unambiguously established by X-ray crystallography. The
absolute structure was deduced by correlation between shizukanolide A and Iin-
denene, which was found in the root of Lindera strychnifolia, and the configura-
tion of which was determined by Takeda et at. 8 Shizukanolide A was reduced by
DIBAH to give a furano-derivative (lindenene). Physicochemical properties of the
furano-derivative {[a]oJ6 -51.4° (c = 0.11, CHCb)} agreed with those of Iin-
denene ([ a]o -50.1 0, CHCI3). Additionally, a second sesquiterpene lactone was
obtained as colorless prisms {mp 64-65.5°C, [a]oi6 + 59.8° (c = 1.02, CHCI 3 )}
in 0.017% from the roots. Physicochemical data and derivatization proved it to
be 8,9-dehydroshizukanolide, and it was named shizukanolide B (Fig. 1). Chlo-
ranthalactone A had been isolated from C. glaber (Sarcandra glabra) and a plane
structure proposed, which was identical to that of 8,9-dehydroshizukanolide. 9
Although shizukanolide A did not exhibit antimicrobial activity, shizukanolide B
showed antifungal activity. 7
PLANT ECOCHEMICALS FROM THE VIEWPOINT OF PLANT DEFENSE 395
W<0 ~o ~
H
Shizukanolide A
:::-
Shizukanolide B
WQ
Lindenene
y:t(o
Yr< H~
9'" :::- 9'"
0
I #
Gleehomanolid Furanodienone Shizukafuranol
~ y::q ~ 1#
0
. :::-
9'" ~"
HO\\'" ~
OH
Shizukolidol Isofuranodiene
W<0 ¢eto ~o
Shizuka-aeoradienol
~ :::-
~ H ~ H
l ~
"
OAe
Chloranthalaetone C
"OH
Shizukanolide C
'OAe
Shizukanolide D
Figure 1. Structures of sesquiterpenoids isolated from Chloranlhus japoniclIs.
Subsequently, from the aerial parts, glechomanolid, and from the leaves,
furanodienone, shizukafuranol, and shizukolidol (Fig. I) were isolated and
identified. 10.1 I Isofuranodiene, shizukaacoradienol, chloranthalactone C, shizu-
kanolide C, and shizukanolide D also were isolated from the roots and identified
(Fig. 1).10-13 The root of C japoniclls was known as Chinese drug. The antimicro-
bial activity of these materials was investigated. Shizuka-nolide A, glechomano-
lid, and isofuranodiene showed little activity against microbes. In contrast,
shizukanolide B (8,9-dehydroshizukanolide) showed antifungal activity and
caused remarkable growth inhibition offungi of the genera Mucor and Rhizopus. IO
Anti-fungal activities of chloranthalactone C and shizukanolide C against Mucor
griseo-cyanus were compared with that of shizukanolide B. Despite, the presence
of an a~,y8-unsaturated lactone moiety in the molecules of chloranthalactone C
and shizukanolide C, they were found to be substantially inactive. 12
Furanogermacranes such as isofuranodiene are considered to be key inter-
mediates of other furanosesquiterpenes. Glechomanolid also may be a biosyn-
396 1. MIZUTANI
thetic precursor of many sesquiterpene lactones, including shizukanolide A and

its derivatives. Furthermore, furano-sesquiterpenes are regarded as precursors of
sesquiterpene lactones. It is significant that these two key compounds were found
in the same plant (Fig. 2).10
We found shizukaol A, a sesquiterpene dimer, in the ether extracts of the
roots of C. japonicus. The structure was elucidated by I D- and 2D-NMR analy-
ses and by chemical methods. The presence of shizukaol A in the fresh extracts
suggests that the dimer is produced in living tissues. 14 Pyrolysis of shizukaol A
gave shizukanolide B and compound X as shown in Fig. 3. Therefore, shizukaol
A should be synthesized by a Diels-Alder type reaction of shizukanolide Band
compound X. In addition, five novel dimeric sesquiterpenes, shizukaols E, F, G,
()I~
'Y)ppl
Farnesyl pyrophosphate
)I'
)I'
lsofuranodiene Glechomanolid
...................
Shizukanolide A
Lindenene
Wt°
Shizukanolide B
Figure 2. Possible biogenetic pathways of shizukanolides.

OH
~C~M'
OH
----II.~ Compound X
Shizukaol A
~o
Shizukanolide B
Figure 3. Pyrolysis of shizukaol A.
Hand 1, consisting of two lindenane units, were isolated from the roots of C.
japonicus. 1S Their structures were elucidated principally by ID- and 2D-NMR
methods (Fig. 4).
Furthermore, a novel sesquiterpene trimer, trishizukaol A, consisting of
three lindenane units, was isolated from the roots of C. japonicus. The structure
was determined by the abovementioned methods. A structurally related sesqui-
terpene dimer, designated as shizukaol J, was also found in the same plant. 16
Although a number of lindenane dimers have been isolated from the chlorantha-
ceous plants, this is the first example of the isolation of a trimeric lindenane
sesquiterpene. Shizukaol J was determined to be a homo-dimer of compound X.
The plausible biogenetic relationships of trishizukaol A, shizukaol J, and com-
pound X are shown in Fig. S. Subtraction of a proton from C-13 of the compound
X (iindenatriene) and successive double bond migration results in a carbanion at
C-lS which then attacks C-ll' of another molecule of compound X to give the
dimer shizukaol 1. Finally, a Diels-Alder type cycloaddition between the termi-
nal methylene of shizukaol J and the C-lS"/C-4"/C-S"/C-6" diene of the third
molecule of compound X completes the structure of the trimer trishizukaol A
(Fig. S). So far, more than 10 lindenane dimers containing compound X as a struc-
tural unit have been isolated from Chloranthus species. 14-16,19-21 All attempts to
isolate monomeric compound X, however, have failed. Structural variations and
biogenetic correlations of such lindenane dimers suggest that these compounds
are synthesized in plant cells.
Chloranthus serratus
We isolated isofuranodiene, furanodienone, chloranthalactone C, and

shizukanolide C from the root extract of C. serratus (Thunb.) Roem. et Schult.
398 1. MIZUTANI
OH OH
SllIzukaol A
" fO
OH
0 0
HO HO
O~O
~ ~O
o~r{l0
·'UOH
0 o 0
Shizukaol C
Shizukaol B Shimkaol G
OH OH OH
C02Mc
OH
'~o-r
o
0
0
HO
00
o
~ ;0 0
0
·'UOH
Shimkaoll Shizukaol F Shizukaol H
Figure 4. Structures of lindenane dimer shizukaols A-I isolated from Chloranthus species and
plausible biosynthetic correlations.
(Chloranthaceae). These compounds had already been confirmed as constituents

of C. japonicus. The known compounds, acoragermacrone, acolamone, and
zederone were also identified by comparing their spectral properties with those
of reference compounds, respectively. In addition, new sesquiterpene ketones,
neoacolamone and 7a-hydroxyneoacolamone, were isolated from the roots, and
their structures were elucidated on the basis of their physicochemical properties
and chemical reactions (Fig. 6).17 The amount of sesquiterpenes in C. serratus
may reflect two biosynthetic routes both dominant in this plant, one being
isofuranodiene ~ furanodienone ~ zederone and the other, acoragermacrone ~
acolamone ~ neoacolamone ~ 7a-hydroxyneoacolamone (Fig. 7). The biosyn-
Compound X Compound X
(Lindenatriene) (Lindenatriene)
Compound X
Trishizukaol A
Figure 5. Plausible biosynthetic route for the formation oftrishizukaol A and shizukaol J from
compound X (lindenatriene).
thesis oflindenanolides (chloranthalactone C and shizukanolide C) does not seem

as active in C. serratus as in C. japonicus. The latter plant produces several lin-
denanolides as major sesquiterpenes in addition to such germacranes as isofura-
nodiene. Furanodienone was once isolated as a miticidal ingredient. '8 It was
known that roots of C. serratus contain insecticidal substances. We found about
0.1 % of furanodienone in fresh roots of this herb, suggesting this compound may
be partially responsible for the insecticidal activity.
Two more novel compounds, shizukanolides E and F (Fig. 6), were isolated
from the ether extracts of the roots of C. serrata, and their structures were deter-
mined to be hydroxy derivatives of shizukanolide c. 1l
400 1. MIZUTANI
Acoragermacrone Acolamone
M o
Zederone
H~O OH
Neoacolamone 7a-Hydroxyneoacolamone
Shizukanolide E
Shizukanolide F
Figure 6. Structures of sesquiterpenoids isolated from Chloranthus serratus. Isofuranodiene,

furanodienone, chloranthalactone C, and shizukanolide C were also isolated from the root
extract of C. serratus. Their structures are shown in Fig. 1.
The novel dimeric sesquiterpenes, shizukaols B, C and D, which consist

of two lindenane (modified eudesmane) units, were isolated from the roots ofe.
serratus. Their structures were elucidated by ID- and 2D-NMR analyses and by
chemical methods. 19 From a biogenetic point of view, shizukaol A should be con-
verted into compound Y as a key inter-mediate by oxidation at C-13' and hydra-
tion of the 4'115' double bond as shown in Fig. 8. Acetylation of the 15'-hydroxyl
of compound Y affords shizukaol D, whereas further oxidation at C-4' and tigly-
lation of the 15'-hydroxyl gives shizukaol C. Finally, com-pletion of the macro-
cyclization by introducing a succinyl unit to C-13' and to the oxidized y-position
of the tiglyl residue of shizukaol C yields shizukaol B (Fig. 8).
In the assignment of the non-equivalent methylene group of an asymmet-
ric succinate residue in the macrocyclic lactone ring of shizukaols Band F, dis-
crimination of two-bond C-H and three-bond C-H was achieved by means of
heteronuclear l3CCH} NOE spectroscopy.20 Thus, the structures of shizukaols B
and F, respectively, were determined as shown in Figure 4.
A novel cyclic lindenane dimer, cycloshizukaol A, which has a Cz-
symmetric structure, was isolated from the roots of e. serratus. 21 The unique
cyclic structure, including a 12-membered ring, was elucidated principally by ID-
and 2D-NMR methods, and the stereochemistry was disclosed through detect-
ing NOEs between equivalent protons by l3C-coupled HMQC-NOESY.
~ 1?
opp
'
,
Famesyl pyrophosphate
O(y
Acoragermacrone
isofuranodiene
•
C(?i
1?
I J
0
~~o
~
o
Furanodienone 'OAc
Acolamone Chloranthalactone C
Zederone Shizukanolide C
Neoacolamone
9?Y
7a-Hydroxyneoacolamone
Figure 7. Possible biogenetic bathways of the sesquiterpenes in Chloranthus serratus.
Cyc1oshizukaol A could be derived biogenetically from two molecules of com-

pound X (lindenatriene) through a [6 + 6] cyc1oaddition reaction, which is
unprecedented in natural products biosynthesis (Fig. 9). The precursor compound
X (lindenatriene) also serves as a key component of other lindenane dimers.
So far, more than 10 lindenane dimers have been isolated from Chloran-
thus spp. Nine of them seem to be from the same biogenetic route starting from
a Diels-Alder-type cyc1oaddition of shizukanolide B and compound X. An
402 1. MIZUTANI
OH
15' Shizukaol A
OH
/
Compound Y
C0 2 Me
Shizukaol C
OH O~
0 / 011\
C0 2Me
HOo~ofO
o 0
Shizukaol B
Figure 8. Plausible biogenetic pathways of Iindenane dimers in Chloranthus serratus,
increasing number of naturally occurring Diels-Alder reaction products have been

identified and the possibility of an enzyme-catalyzed biosynthesis of such a
molecule has recently been reported. 22 This biogenetic route for the series of lin-
denane dimers also suggests the presence of the biological Diels-Alder reaction,
since consecutive hydroxylation and acylation occur after the cycloaddition.
In addition, four dimers, shizukaols B, F, G, and H, carry an unusal macro-
cyclic ester ring as a pendent side chain composed of y-hydroxytiglic or y-
hydroxysenecioic acid, and succinic or L-malic acid (Fig. 4).
Oil Oil
Compound X
(Lindenatriene)
HO
Oil
Cycloshizukaol A
Compound X
(Lindenatriepe)
Figure 9. Cycloshizukaol A isolated from the roots of Chloranthus serratus, possibly derived
biogenetically from two molecules of compound X (Iindenatriene) through a [6 + 6] cycload-
dition reaction.
Lindenanolides from Other Chloranthus Species

Chloranthaceae is a small family, and only five species are known in
Japan. We examined the distribution of !indenanolides in C. fortunei (A. Gray)
Solms-Laub, C. spicatus (Thunb.) Makino, and Indonesian C. elatior R. Br. by
comparing gas chromatograms of the ether extracts of these plants. 23 In the roots
C. fortunei, chloranthalactone C was contained as a major component and a
small amount of shizuka-acoradienol was also detected, whereas no significant
peak was found in the aerial parts. The roots of C. spicatus also contained
shizukanolide Band chloranthalactone C, as well as isofuranodiene and shizuka-
furanol, all having been found in C. japonicus. The aerial parts of C. spicatus and
C. elatior showed similarGLC patterns, and both species contained shizukano-
!ide B.
OLIGOSTILBENES FROM CAREX SPECIES
Carex fedia var. miyabei
The methanol extract of the roots and rhizome of Carex .redia Nees var.
miyabei (Franchet) T. Koyama (Cyperaceae) shows strong antibacterial activity
against Staphylococcus aureus. We isolated four 3,5,4'-trihydroxystilbene (resver-
atrol) o!igomers, one dimer, one trimer, and two tetramers. 24 Their structures were
determined from spectroscopic evidence and biogenetic consideration. The dimer
was identified to be viniferin, already identified as a phytoalexin of grapevine
404 J. MIZUTANI
leaves. 25 The trimer and teramers were new compounds and were named
miyabenols C (for the trimer), and A and B (for the tetramers). The stereochem-
istry of miyabenols A and B was later determined, 26 and their structures are shown
(Fig. 10).
OH
OH
OH
OH
Miyabenol A Miyabenol B
HO OH
OH
Miyabenol C Kobophenol A
OH
OH
Hopeaphenol
Kobophenol B
Figure 10. Structures of oligostilbenes (trimer and tetramers) isolated from Carex Species.
Carex kobomugi
An antimicrobial novel tetrastilbene named kobophenol A has been isolated
from the roots and rhizome of Carex kobomugi Ohwi (Cyperaceae), together with
previously known oligostilbenes, miyabenol C and E-viniferin, as a major active
principle. 27 ,28 The structure of kobophenol A has been determined on the basis of
spectroscopic and chemical properties (Fig. 10).
Carex pumila
A novel antimicrobial tetrastilbene constituted of two E-viniferin units
was isolated from the subterranean parts of C. pumila Thunb. (Cyperaceae).29 Its
complex polycyclic structure was elucidated by precise analyses and was named
kobophenol B. The relative stereochemistry of kobophenol B was assigned from
the results of NOESY experiments. Kobophenol B is expected to be biosynthe-
sized from two molecules of (-)-E-viniferin, which has been found in the same
plant, via four succesive electron oxidative coupling reactions. Therefore, the
absolute configuration of kobophenol B can be deduced as depicted in Figure 10.
E-Viniferin was first isolated as a phytoalexin of grapevine (Vilis vinifera,
Vitaceae) leaves. 25 We isolated (-)-E-viniferin and related oligostilbenes (Fig. II)
from Carex species. 24.27 However, no stereochemical assignment for E-vineferin
has been made except that its corresponding 7b,8b-dihydro-erivative, gnetin F
(2b),30 has been shown to have a 7aS, 8aS configuration from CD results (Fig.
11). An ether-soluble phenolic fraction obtained from the methanol extract of
the roots and rhizome of C. pumila was chromatographed on a silica gel column
and subsequent silica gel preparative TLC yielded three oligostilbenes, (-)-E-
viniferin, and miyabenols C and A, which were identified by comparing with
authentic samples. 3 !
The relative configuration of (-)-E-viniferin (la) was next determined by
chemical methods. Permethylation of la with dimethyl sulfate gave a pentamethyl
ether derivative (Sa). Dehydrogenation of Sa with DDQ readily afforded a ben-
zofuran derivative (6). Catalytic hydrogenation of 6 gave a cis-dihydrobenzofu-
ran (7). Cis-trans isomerization at the 7a,8a-positions of7 was smoothly catalyzed
by aluminum chloride to yield a corresponding trans isomer (8). The IH-NMR
spectrum of7b,8b-dihydro-derivative (8a), directly obtained from Sa by catalytic
hydrogenation, was identical with that of 8, but apparently different from that
of 7. Therefore, the relative configuration of la was determined to be trans. The
CD spectrum ofa 7b,8b-dihydro-derivative (2a) of (-)-E-viniferin (la) showed a
completely opposite pattern to that of gnetin F (2b). The absolute configuration
of (-)-E-viniferin (la) from C. pumila was thus concluded to be 7aR,8aR
(Fig. 11).
Interestingly, we found that wild grapevine (Vilis coignetiae) produced (+)-
E-viniferin by Cu2+ induction. 31 Therefore, e-viniferin produced by grapevine is
406 1. MIZUTANI
RO
RO
OR OR
R H-7a H-8a R H-7a H-8a
1a [(-)-E-Viniferin]: H f:l a 2a: H f:l a

1b: H a f:l 2b (Gnetin F): H a f:l
5a: Me f:l a 7: Me cis
6: Me dehydro 8: Me trans
Sa: Me f:l a
Figure 11. Structures of E-viniferin and its derivatives.
supposed to be an antipode (gnetin F type) of 1a, whereas Scirpus fluviatilis con-

tains hydroxylated viniferins (scirpusins A and B) as a (±)_form. 32
Later, we isolated another tetrastilbene, hopeaphenol (Fig. 10), from C.
pumila. 33 This known compound possesses C2 symmetry, and some difficulties
arising from the presence of equivalent carbons and protons in the molecule
were encountered in its structure determination. A number of biologically active
natural products having C2 symmetry have been isolated over the last few years.
The structure determination of such symmetric compounds by NMR methods,
however, is problematic because of the difficulty in distinguishing between equiv-
alent nuclei. From an isotopomeric point of view, however, three isotopomers,
12C_12C, 12C_I3C, 13C_13C, are present with respect to the equivalent carbons con-
necting each other. Among them, the 12C_12C species is undetectable and the I3c_
13C species is excluded owing to its low natural abundance. Therefore, only the
12C_I3C isotopomer is detected in 13C_NMR and this species is not symmetrical.
In the H_ 12 C- 13 C_H system, the separate detection of IJ (CH) and 2J (CH) was
achieved by INEPT-LSPD, COLOC, and HMBC methods to demonstrate the
connectivity between the equivalent carbons from the viewpoint of isotopo-
meric asymmetry. This concept was also applied to detect the NOEs between
equivalent protons which cannot be observed by the conventional NMR
methods. 34
Biogenetic Considerations and Roles in Carex Plants

Subterranean parts of Carex kobomugi contain a series of oligostilbenes,
(-)-E-viniferin, miyabenol C and kobophenol A (Fig. 12). In particular, the
tetramer kobophenol A is the most prominent component (0.3% w/w of fresh
plant), just as other Carex species contain tetrastilbenes as their major con-
stituents (miyabenol A: more than 0.1 % in C. fedia var. miyabei and more than
0.2% in C. pumila; kobophenol B: about 0.1 % in C. pumi_Ia).24.27-2931.33 The
biosynthetic route for oligostilbenes in C. kobomugi is presumed to be by suc-
cessive oxidative coupling from (-)-E-viniferin (a dimer) to kobophenol A (a
tetramer) via miyabenol C (a trimer) because of the coexistence ofthe three com-
pounds. A possible mode of coupling is shown in Figure 12. A resveratrol radical
I is coupled with another radical II from the re face to give (-)-E-viniferin. An E-
viniferin radical is then coupled with radical II from the si face to yield miyabenol
C. Finally coupling of a miyabenol C radical with resveratrol radical I from
the re face followed by addition of water gives kobophenol A. Hence, we
estimated the absolute configurations of kobophenol A and miyabenol C to be
7aR,8aR, 7bS,8bS, 7cS,8cS, 7dS,8dR and 7aR,8aR,7bS,8bS, respective-ly.28 Inter-
estingly, major components are different from each other among three Carex
species. Miyabenol A is also produced from resveratrol via (-)-E-viniferin and
miyabenol C by means of successive electron oxidative coupling reactions. The
last step, a coupling reaction of miyabenol C radical and a resveratrol radical
without the addition of water, gives miyabenol A.
Kobophenol A exhibited moderate antibacterial activity against Staphylo-
coccus aureus by the paper disc method. 27 ,28 An extremely high concentration of
kobophenol A in C. kobomugi indicates that it may playa physiologically impor-
tant role in the plant. Similarly, high concentrations of tetramers miyabenol A
and miyabenol B in C. fedia var. miyabei,24 and miyabenol A (more than 0.2%)
kobophenol B (about 0.1 %) and hopea-phenol (about 0.05%) in C. pumila 29 .3 1,33
suggest their roles as defense allelochemicals in Carex plants.
SESQUITERPENOIDS FROM ROSA RUGOSA
Antimicrobial Rugosal A and Related Compounds
In the course of investigating antimicrobial agents in higher plants, we

found that an aqueous extract of damaged leaves of Rosa rugosa Thunb.
(Rosaceae) inhibited growth of microorganisms, and that an antimicrobial prin-
ciple was readily extracted from the diffusates of damaged leaves. 35 Young fresh
leave of R. rugosa (1.3 kg) were bruised by gently hitting them with a wooden
408 1. MIZUTANI
HO
o HO
•
(-)-E- Viniferin
Resveratrol radical I OH
Resveratrol radical II
OH
OH
HO ...
Reveratrol radical II
Miyabenol C
HO
E-Viniferin radical
o
HO
"Resveratrol radical I
HO
Miyabenol C radical
OH
Kobophenol A
Figure 12. Plausible biogenetic route for oligostilbenes in Carex kobomugi.

hammer. After keeping damaged leaves in 71 of water at 25°C for 24hr, the water
diffusates were extracted with ethyl acetate (21). Column chromatography of the
neutral fraction (0.8 g) on Wako-gel C-200 gave colorless needles (119 mg) as the
anti-microbial principle, named rugosal A (Fig. 13). The structure of rugosal A
was deduced by chemical and spectroscopic (UV, IR, MS, and NMR) methods to
be a novel carotane with an a,p-unsaturated aldehyde group, an endo-peroxide
bridge, and an allyl alcohol partial structure. The corresponding carboxylic acid,
named rugosic acid A, was isolated from the uninjured leaves, but rugosic acid
A did not exhibit antifungal activity.
Since rugosal A has an allylic alcohol partial structure, the exciton chiral-
ity method 36 was expected to be applicable to determined the absolute configura-
tion, after being converted into the corresponding benzoate. Rugosal A benzoate,
however, could not be obtained, and a small amount of an hemiacetal benzoate
(derivative I) was produced (Fig. 13). Therefore, derivative I was produced via a
hemiacetal by heating rugosal A in pyridine followed by the treatment of benzoic
anhydride in triethylamine. Since derivative I is stereochemically equivalent to
rugosal A benzoate, the exciton chirality method is applicable. To exclude an
effect of the aldehyde chromophore on the CD spectrum, derivative I was reduced
with an excess amount of NaBH4 to yield the corresponding alcohol (derivative
15
14
CHO COzH
12
13 Rugosa! A
Rugosa! A hemiaceta!
f ~
Rugosic acid A
~a
CHO CHzOH
Rugosa! A benzoate Derivative I Derivative II
Figure] 3. Structures of rugosal A and rugosic acid A, and chemical conversion of rugosal A
to determine its absolute configuration. a, benzoic anhydride/triethylamine; b, NaBH4; c, pyri-
dine/heating. .
410 1. MIZUTANI
II). The CD spectrum of derivative II showed an apparent positive Cotton effect

at 229 nm. Thus, the absolute configuration of the rugosal A was determined to
be Sat C-2 as depicted in Fig. 13. 37
Subsequently, a new sesquiterpene, (7R, lOR)-carota-1 ,4-dienaldehyde was
isolated from R. rugosa leaves, and its structure was elucidated by spectroscopic
and chemical methods. This compound was markedly unstable under air exposure
to give several oxidized derivatives, in which rugosal A was found as the minor
product due to the 1,4-diene structure. The corresponding carboxylic acid, carota-
1,4-dienoic acid was also found in the leaves. 38 In the autoxidation of carota-l,4-
dien-14-al, an intermediate possessing a 1,5-endoperoxy-2-hydroperoxy system
was isolated as a major product. The structures of the intermediate and other
by-products including rugosal A and rugosic acid A were determined. These autox-
idation products indicated the conversion pathway of carota-I,4-dien-14-al
into rugosal A, similar to the radical reaction pathway in the autoxidation of
polyunsaturated fatty acids. 39 Furthermore, the isolation of a 5-hydroperoxide in
the autoxidation of a non-conjugated cyclic diene (using 14-acetoxycarota-I,4-
diene) allowed us to substantiate the aspects for the autoxidation pathway in non-
conjugated 1,4-diene systems. 40 Structures are shown in Fig. 14.
Further investigation of the autoxidation of carotane sesquiterpenes pos-
sessing a non-conjugated 1,4-diene system has been done. 41 Hashidoko (1996)
wrote a preeminent re-view on the phytochemistry of R. rugosa. 42
Other Sesquiterpenes
Bisaborosaol A, a novel sesquiterpene belonging to the bisabolane class,

and the corresponding acid, hamanasic acid A, were isolated from R. rugosa
leaves (Fig. 14).43 The absolute configuration ofbisaborosaol A at C-4 was shown
to be inverted in comparison with bisabolanoids from the Compo sitae.
Bisaborosaol A and hamanasic acid A may be biogenetically related to the
carotanoids which constitute the major sesquiterpenes of R. rugosa leaves
(Fig. 15).
Additionally, five novel bisabolane sesquiterpenes possessing a tetra-
hydrofuran ring or a hydroperoxy group were isolated from R. rugosa leaves
and their structures elucidated by chemical and spectroscopic methods. These
highly oxidized sesquiterpenes were structurally related to bisaborosaol A,
which is the major bisabolanoid of the plant. (Structures not shown, see refs. 42,
44)
During the investigation of minor sesquiterpenoids in R. rugosa leaves,
11 new sesquiterpenes, eight terpene aldehydes, two carotane acids, and an
acoranoid (named rosacore-none, shown in Fig. 14), were isolated and their
structures (not shown) determined by means of chemical and spectroscopic
methods. 42 ,45
13
R= CHO: Carota-I,4-dienaIdehyde An intermediate possessing 14-Acetoxy-5-hydroper-

(Carota-I,4-dien-14-aI) a 1,5-endoperoxy-2-hy- oxycarota-I,4-diene
R= C0 2H: Carota-I,4-dienoic acid droperoxy system
Carota-4( 5),11 (12)-diene Carota-I(IO),4(5)-diene

(Isodaucene) (Daucene) R= Me: Bisaborosaol A
R= H: Hamanasic acid A
14
HO
I
2 on'I:, 2'
?"
3 6' ~ OR
4 5'
Acora-3( 4), 7(8)-diene °

R= Me: 6-Demethoxy-4'-O-methy1capillarisin
R= Hi Acora-3(4),7(J5)-diene
R= 0: Rosacorenone R= H: 6-Demethoxycapillarisin
OH
Apigenin
Figure 14. Structures of isolated compounds from Rosa rugosa leaves and their derivatives.
412 J. MIZUTANI
R= CH3 : cis, trans-Farnesyl R= CHf Bisabolane cation Bisaborosaol A

pyrophosphate R= CH20H or CHO
R= CH20H or CHO
R= CH3: Carotane cation Carota-I,4-dienaldehyde

R= CH2 0H or CHO
Figure 15. Hypothetical biosynthetic relation between bisabolanoids and carotanoids of Rosa
rugosa.
Phenoxychromones and a Related Flavone
Two 2-phenoxychromones, 6-demethoxy-4' -O-methylcapillarisin and 6-

demethoxy-capillarisin, were identified in leaves of R. rugosa. 46 Apigenin was
also found in the leaves. Their structures are shown (Fig. 14). This was the first
report of phenoxychromones outside the Compositae.
Glandular Trichome Exudate and Its Defensive Role

In earlier work, we found that R. rugosa leaves contain a sesquiterpene per-
oxide, rugosal A, and related compounds including rugosic acid A and carota-
1,4-dienaldehyde. Rugosal A has antifungal properties and, because of its high
content in the leaf,47 we proposed that it may playa general defensive role in the
plant. Later, we found that the glandular trichome exudate of the leave is com-
posed mainly of two major carotane sesquiterpenes, rugosal A and rugosic acid
A, and carota-l,4-dienaldehyde is also present. 48 The occurrence of rugosaI A in
the exudate suggests a defensive role of the glandular trichomes against pest
organisms, as the compound has antifeedant activity against tobacco cutworm
larvae. The crude exudate and pure rugosal A were tested for their antifeedant
activity against neonate tobacco cutworm larvae. 48 Both the trichome exudate and
rugosal A retarded feeding of the larvae, and the activity of the crude exudate
was attributed mainly to rugosal A. It is well known that other higher plants (e.g.
Solanaceae and Compositae) also possess glandular trichomes which exude
defensive compounds.
In addition, some other compounds, minor carotanes, non-carotane
(bisabolane and acorane) sesquiterpenes, and 2-phenoxychromones were also
identified as constituents of the exudate. Since it has been suggested that biosyn-
thesis of the carotane sesquiterpenes is highly active in the tip cells of the glan-
dular trichomes, precursor sesquiterpene hydrocarbons were expected to be
present in the exudate. In the investigation ofless polar fractions, some sesquiter-
pene hydrocarbons were detected, and three major sesquiterpene hydrocarbons
[acora-3(4), 7(15)-diene, carota-4(5), II (12)-diene (isodaucene) and carota-I,4-
diene], and two minor [daucene and acora-3(4),7(8)-diene] were identified. 49 The
presence of these carotadienes, which are possibly precursors of the correspond-
ing C-14-oxygenated carotanoids, suggests that the formation of the carotane
skeleton, followed by the regio-specific oxigenation at C-14, occur in the biosyn-
thesis of R. rugosa carotanoids. The structures of these sesquiterpene hydrocar-
bons are also shown in Figure 14.
PHYTOALEXIN OF TARAXACUM OFFICINALE
Lettucenin A
During a survey of antifungal substances in higher plants, we found that

some fungi toxins accumulated in an aqueous 3 mM cupric chloride solution in
which dandelion leaves (Taraxacum officinale Web., Compositae-Lactucoideae,
tribe Lactuceae) were steeped. The main fungitoxin in the diffusate was isolated
and identified as lettucenin A,50 which had been isolated from lettuce (Lactuca
sativa) infected with Pseudomonas cichorii. It is a novel guaianolide phytoalexin
(Fig. 16).51 The fungitoxic effects of lettucenin A on Cladosporium herbarum
were: >6.50 IJ,g/cm2, complete growth inhibition; 1.63 ~ 3.25 IJ,g/cm2, partial
growth inhibition; and O.20~O.81 IJ,g/cm2, slight growth retardation. These results
were confirmed on TLC plates by a duplicate test. 52
Lettucenin A was first reported to be a sesquiterpenoid phytoalexin of
lettuce by Takasugi et al.,51 and found to be produced by dandelion leaves in
response to cupric chloride stress. 50 However, there was no evidence to show
whether or not this compound is active enough in leaves to inhibit the invasion
of pathogens. We have found that lettucenin A works effectively to ward off a
pathogenic fungus in vivo. 53
414 1. MIZUTANI
13
CHO
5'
R 1= OMe, R2= Me: lrilin A
RI= OMe, R2= H: lrilin B
Lettucenin A
R 1= R2= H: lrilin C
RI= OMe, R2= Me, R 3= H: lristectorigenin B R 1= OMe, R2= H: Tectorigenin

RI= R2= H: Genistein
RI= R J = H, R2= Me: 3'-O-Methylorobol
R 1= H, R2= Me: Biochanin A
R 1= OMe, R2= H, R 3= Me: lristectorigenin A
R 1= R2= H, R 3= Me: Pratensein OH
HO
HO
1'# 2' R2 R
I 3' OH 0
R= H: Apigenin
R= OMe: Hispidulin
6'" 4' RJ
RI= OMe, R2= R 3 = H: Ayamenin A

5'
0(R
R 1= R2= R 3= H: Ayamenin B Meo,(!(o,(,\,V
R 1= OMe, R2= OH, R 3= H: 5,7,3'-Trihy-
droxy-6-methoxycoumaronochromone
~OH
RI= R 3 = H, R2= OH: Ayamenin C OH 0
RI= OMe, R2= H, R 3= OH: Ayamenin D
R= H: Alpinone
RI= R2= H, R 3= OH: LupinaIbin A R= OH: 7-0-Methyldihydrokaempferol
R 1= H, R2= OH, R J = OMe: Ayamenin E
Q
~
yY
H0yY0'('\~ #
OH
OH 0
5,7,2'-Trihydroxyflavanone
Figure 16. Structures of compounds inducibly produced by Taraxacum officinale and Iris
pseudacorus leaves treated with cupric chloride.
When a dandelion leaf is infected by a pathogenic fungus, necrotic spots

appear on the leaf. Yellow fluorescent rings around the spots are observed when
the leaf is viewed under UV 365 nm light. The fluorescence is similar to that
exhibited by lettucenin A, and the presence of this compound in these rings was
confirmed by analyzing the ethyl acetate extract of the brown spots by HPLC and
TLC bioautography, using Cladosporium herbarum as a test fungus. 54 The con-
centration of lettucenin A (16.4llg/g fro wt.) was calculated by HPLC analysis of
the ethyl acetate extract, and the amount seemed sufficient to inhibit the growth
of pathogens, since the growth of C. herbarum is completely inhibited at a level
of 6.251lg/cm2 on TLC plates (0.25 mm thickness). The fluorescence of lettucenin
A around a necrotic spot in which the pathogen was assumed to be dead became
faint. In addition, another fluorescent spot was observed under UV light inde-
pendent of any apparent necrosis. This observation suggests that the production
of lettucenin A starts at an early stage of the fungal infection, before the appear-
ance of symptoms, and ends soon after the death of the pathogen. Lettucenin A
thus works effectively to ward off the invasion of pathogens in vivo. Since
lettucenin A is produced by dandelion cell cultures when they are exposed to a
fungal elicitor, we can categorize the compound as a phytoalexin of the dande-
lion. 53 When we stimulated the culture with the elicitor from C. herbarum,
lettucenin A production occurred within I hr and reached a maximum in 2-6 hr.
A similar time course production of lettucenin A took place in cell cultures
stressed with 1 mM CuCI 2 •
Compositae are classified into two subfamilies: Asteroideae (major) and
Lactucoideae (minor). The latter subfamily is composed of 8 tribes, 440 genera,
and ca. 8,200 species, and characterized by the production of latex. T officinale
belongs to the tribe Lactuceae, which consists of 70 genera and 2,300 species.
We examined the production of lettucenin A by some plants in this tribe. 53 Detec-
tion 0; lettucenin A was carried out by TLC auto-biography. Six species of plants
in 4 genera, when stressed with cupric chloride, produced the compound: Tarax-
acum hondoense, Lactuca dentata, L. scariola, Sonchus oleraceus, S. asper, and
Ixeris repens. Four other species, however, did not: Picris hieracioides, Hieracium
aurantiacum, H. umbel/atum, and Hypochoeris radicata. Lettucenin A seems to
be a common compound enabling plants in some genera of the Lactuceae to with-
stand fungal invasion, and might be used as a taxonomic characteristic of the
plants in this tribe.
ISOFLAVONOIDS AND FLAVONOIDS FROM

IRIS PSEUDACORUS
Isoftavonoids
During a survey of natural fungitoxins induced in higher plants, we found

that the leaves of Iris pseudacorus L. (Iridaceae) treated with cupric chloride
416 1. MIZUTANI
produced 16 phenolic compounds, and some of them exhibited fungitoxic activ-

ity.55 The 16 phenolic compounds isolated and identified are 10 simple isoftavones
and six coumaronochromones: 5,2'-di-hydroxy-6,7-dimethoxyisoftavone (irilin
A), 5,7,2'-trihydroxy-6-methoxyisoftavone (irilin B), 5,7,2'-trihydroxyisoftavone
(irilin C), iristectorigenin B, 3' -O-methylorobol, iristectori-genin A, pratensein,
tectorigenin, genistein, biochanin A, 5,7 -dihydroxy-6-methoxy-coumarono-
chromone (ayamenin A), 5,7-dihydroxycoumaronochromone (ayamenin B),
5,7,3'-trihydroxycoumaronochromone (ayamenin C), 5,7,3' -trihydroxy-6-
methoxycoumaro-nochromone, 5,7,4'-trihydroxy-6-methoxycoumarono-
chromone (ayamenin D) and lupinalbin A (Fig. 16). Among these compounds,
irilins A, Band C, and ayamenins A, B, C, and D were new natural isoftavonoids.
It was the first report of stress metabolites from the Iridaceae. Among these
compounds, irilin C exhibited the strongest antifungal activity. When examined
on TLC plate by using C. herbarum as the test fungus, 25 f..lg/cm 2 of irilin C
showed complete inhibition of the growth.
Further investigation of stress metabolites produced by 1. pseudacorus
leaves treated with cupric chloride revealed the presence of one new compound,
5,7,3' -trihydroxy-4' -methoxycoumaronochromone (ayamenin E), along with
ftavones and related compounds, which will be discussed below. 56
Flavonoids
In addition to ayamenin E, two ftavones (apigenin and hispidulin), two

dihydroftavonols (alpinone and 7-0-methyldihydrokaempferol) and 5,7,2'-
trihydroxyftavanone were isolated from the Iridaceae for the first time (Fig. 16).
However, it is not clear if these compounds are stress metabolites or constitutive,
because the amounts found in the stressed leaves were rather small.
CONCLUSION
Higher plants produce characteristic secondary metabolites, such as alka-

loids, terpenoids, and phenolic compounds. Their specific functions are often
unknown. The essential roles that these metabolites play in complex interactions
among living organisms in the natural environment are gradually being unrav-
eled. We have been working on such compounds from the viewpoint of plant
defense. We have found that a wide variety of secondary metabolites are involved
in plant defense mechanisms.
Chloranthus species (Chloranthaceae) produce characteristic sesquiterpene
lindenanolides. Shizukanolide B, which has an a~,'Y8-unsaturated lactone struc-
ture, shows antifungal activity and especially causes remarkable growth inhi-
bition of the fungi of the genera Mucor and Rhizopus. The concentration of
shizukanolide B is ca. l70ppm in the fresh roots of C. japonicus. Novel dimeric
lindenane sesquiterpenes and the trimer are produced in Chloranthus species. The
biological meaning of plants synthesizing such sophisticated compounds has not
been made clear, but it is interesting from the viewpoint of phytochemistry.
Carex species (Cyperaceae) produce characteristic oligostilbenes as
defense allelochemicals. High concentrations of tetrastilbenes suggest they play
important roles in the defense mechanisms of the plants. Tetrastilbenes in Carex
species are biosynthesized from resveratrol (3,5,4'-trihydroxystilbene, monomer)
via (-)-E-viniferin (dimer) and miyabenol C (trimer) by means of successive elec-
tron oxidative coupling reactions. The last step, a coupling reaction of miyabenol
C radical with a resveratrol radical followed or not followed by the addition
of water, gives kobophenol A or miyabenol A, respectively. Interestingly, major
components are different from each other among the three Carex species. A
breakthrough in determining the structure of hopeaphenol (tetramer having C2
symmetry) from an isotopomeric point of view has been done by using NMR
spectroscopy.
A wild rose, Rosa rugosa (Rosaceae), produces a novel carotane sesquiter-
pene, rugosal A, with antifungal and antifeedant activities. Carota-I,4-dien-14-
aI, precursor of rugosal A is biosynthesized in the gandular trichomes of R. rugosa
leaves, and then the precursor is converted into rugosal A by autoxidation.
Taraxacum officinale (Compositae) leaves are induced to produce an anti-
fungal guaianolide, lettucenin A. When dandelion leaf is infected with a patho-
genic fungus, lettucenin A is produced around the necrotic spots in sufficient
amounts to inhibit the growth of pathogens. Iris pseudacorus (Iridaceae) leaves
also can be induced to produce several isoftavonoids and ftavonoids as defense
chemicals.
ACKNOWLEDGMENTS
This work was carried out in the Faculty of Agriculture, Hokkaido Univer-
sity in collaboration with colleagues, research associates, and many graduate and
undergraduate students. Their names are indicated in references. Some of works
have financially supported by Japan Science and Technology Corporation (JST).
REFERENCES
1. MOLISCH, H. 1937. Die Einfluss einer Pflanze auf die Andere-Allelopathie, Gustav
Fischer, lena.
2. WALLER, G.R. (ed.), 1989. Allelochemicals: Role in Agriculture and Forestry, ACS
Symposium Series 330, American Chemical Society, Washington, DC.
3. RIZVI, S.J.H., RIZVI, V. (eds.), 1992. Allelopathy: Basic and Applied Aspects, Chapman
& Hall, London.
418 1. MIZUTANI
4. LOVETT,1.V 1990. Chemicals in crop protection: Is there an alternative? In: Alterna-

tives to the Chemical Control of Weeds. (C. Bassett, L.S. Whitehouse, and 1.A.
Zabkiewizz, eds.), Ministry of Forestry, FRI Bulletin 155, Rotorua, New Zealand, pp.
57-65.
5. LOVETT, 1.V, RYUNTYU, M. 1992. Allelopathy: Broadening the context. In: Allelopa-
thy: Basic and Applied Aspects. (S.1.H. Rizvi and V Rizvi, eds.), Chapman & Hall,
London, pp. 11-19.
6. MIZUTANI,1. 1991. Plant ecochemicals. 1. Pesticide Sci. (in Japanese) 16: 679-686.
7. KAWABATA, 1., TAHARA, S., MIZUTANI, 1., FURUSAKI, A., HASHI- BA, MAT-
SUMOTO, T. 1979. Shizukanolides, two sesquiterpenoids from Chloranthus japonicus
(Chloranthaceae). Agric. BioI. Chern., 42: 885-887.
8. TAKEDA, K., ISHII, H., TOZYO, T., MINATO, H. 1969. Components of the root of
Lindera strychnifolia ViII. Part XVI. Isolation of lindenene showing a new fundamental
sesquiterpene skeleton, and its correlation with lindenene. 1. Chern. Soc., (C) 1969:
1920-1921.
9. UCHIDA, M., KUSANO, G., KONDO, Y., NOZOE, S., TAKEMOTO, T. 1978. Two new
sesquiterpenoids from Chloranthus glaber Makino. Heterocycles 9: 139-144.
10. KAWABATA, 1., TAHARA, S., MIZUTANI, 1. 1981. Isolation and structure elucidation
of four sesquiterpenes from Chloranthus japonicus (Chloranthaceae). Agric. BioI. Chern.
45: 1447-1453.
11. KAWABATA, 1., FUKUSHI, Y., TAHARA, S., MIZUTANI, 1. 1984. Structures of novel
sesquiterpene alcohols from Chloranthus japonicus (Chloranthaceae). Agric. BioI. Chern.
48: 713-717.
12. TAHARA, S., FUKUSHI, Y., KAWABATA, 1., MIZUTANI, 1. 1981. Linde- nanolides
in the root of Chloranthus japonicus (Chloranthaceae). Agric. BioI. Chern. 45: 1511-
1512.
13. KAWABATA, 1., MIZUTANI, 1. 1989. Shizukanolides D, E and F, novellindenanolides
from Chloranthus spp. (Chloranthaceae). Agric. BioI. Chern. 53: 203-207.
14. KAWABATA, 1., FUKUSHI, Y., TAHARA, S., MIZUTANI, 1. 1990. Shizukaol A, a
sesquiterpene dimer from Ch loran thus japanicus. Phytochemistry 29: 2332-2334.
15. KAWABATA, 1., FUKUSHI, E., MIZUTANI, 1. 1995. Sesquiterpene dimers from
Chloranthus japonicus. Phytochemistry 39: 121-125.
16. KAWABATA, 1., FUKUSHI, E., MIZUTANI, 1. 1998. Sesquiterpene dimer and trimer
from Chloranthus japonicus. Phytochemistry 47: 231-235.
17. KAWABATA, 1., FUKUSHI, Y., TAHARA, S., MIZUTANI, 1. 1985. Structures of novel
sesquiterpene ketones from Chloranthus serratus (Chloranthaceae). Agric. BioI. Chern.
49: 1479-1485.
18. MARADUFU, A. 1982. Furanosesquiterpenoids of Commiphora erythraea and C. myrrh.
19. KAWABATA, 1., MIZUTANI, J. 1992. Dimeric sesquiterpenoid esters from Chloranthus
serratus. Phytochemistry 31: 1293-1296.
20. FUKUSHI, E., KAWABATA, J., MIZUTANI, 1. 1995. Discrimination of two methylene
groups in non-symmetric succinate esters in sesquiterpene dimers by heteronuclear
13C{'H} NOE spectroscopy. Magn. Reson. Chern. 33: 909-912.
21. KAWABATA, 1., FUKUSHI, E., MIZUTANI, 1. 1993. Symmetric sesquiterpene dimer
from Chloranthus serratus. Phytochemistry 32: 1347-1349.
22. OIKAWA, H., SUZUKI, Y., NAYA, A., KATAYAMA, K., ICHIHARA, A. 1994. First
direct evidence in biological Diels-Alder reaction of incorporation of diene-dienophile
precursors in the biosynthesis of solanapyrones. 1. Am. Chern. Soc. 116: 3605-3606.
23. KAWABATA, J., MIZUTANI, 1. 1988. Distribution of lindenanolides in the Chloran-
thaceae. Agric. BioI. Chern. 52: 2965-2966.
24. SUZUKI, K., SHIMIZU, 1., KAWABATA, J., MIZUTANI, J. 1987. New 3,5,4'-
trihydroxystilbene (resveratrol) oligomers from Carexfedia Nees var. miyabei (Franchet)
1. Koyama (Cyperaceae). Agric. BioI. Chern. 51: 1003-1008.
25. LANGCAKE, P., PRYCE, R.1. 1977. A new class ofphytoalexins from grapevines. Expe-
rientia 33: 151-152.
26. KAWABATA, J., MISHIMA, M., KURIHARA, H., MIZUTANI, J. 1995. Stereochemi-
stry of two tetrastilbenes from Carex species. Phytochemistry 40: 1507-1510.
27. KAWABATA, J., ICHIKAWA, S., KURIHARA, H., MIZUTANI, J. 1989. Kobophenol
A, a unique tetrastilbene from Carex kobomugi Ohwi (Cyperaceae). Tetrahedron Lett. 30:
3785-3788.
28. KURIHARA, H., KAWABATA, 1., ICHIKAWA, S., MISHIMA, M., MIZUTANI, J. 1991.
Oligostilbenes from Carex kobomugi. Phytochemistry 30: 649-653.
29. KAWABATA, J., MISHIMA, M., KURIHARA, H., MIZUTANI, J. 1991. Kobophenol B,
a tetrastilbene from Carex pumila. Phytochemistry 30: 645-647.
30. LINS, A.P., YOSHIDA, M., GOTTLIEB, O.R., GOTTLIEB, H.E., KUBITZKI, K. 1986.
Plant chemosystematics and phylogeny. Part XXXII. Gnetins in Welwitschia. Bull. Soc.
Chim. Belg. 95: 737-748.
31. KURIHARA, H., KAWABATA, 1., ICHIKAWA, S., MIZUTANI, 1. 1990. H-E-Viniferin
and related oligostilbenes from Carex pumila Thunb. (Cyperaceae). Agric. BioI. Chern.
54: 1097-1099.
32. NAKAJIMA, K., TAGUCHI, H., ENDO, 1., YOSIOKA, 1. 1978. The constituents of
Scirpus fiuviatilis (Torr.) A. Gray. I. The structures of two new hydroxystilbene dimers,
scirpusin A and B. Chern. Pharm. Bull. 26: 3050-3057.
33. KAWABATA, 1., FUKUSHI, E., HARA, M., MIZUTANI, J. 1992. Detection of connec-
tivity between equivalent carbons in a C, molecule using isotopomeric asymmetry:
Identification of hopeaphenol in Carex pumila. Magn. Reson. Chern. 30: 6-10.
34. KAWABATA, 1., FUKUSHI, E., MIZUTANI, J. 1992. 2D 13C-coupled HMQC-ROESY:
A probe for NOEs between equivalent protons. 1. Am. Chern. Soc. 114: 1115-
1117.
35. HASHIDOKO, Y, TAHARA, S., MIZUTANI, J. 1989. Antimicrobial sesquiterpene from
damaged Rosa rugosa leaves. Phytochemistry 28: 425-430.
36. HARADA, N., IWABUCHI, 1., YOKOTA, Y, UDA, H., NAKANISHI, K. 1981. A
chiroptical method for determining the absolute configuration of aJlylic alcohols. 1. Am.
Chern. Soc. 103: 5590-5591.
37. HASHIDOKO, Y, IWAYA, N., TAHARA, S., MIZUTANI, J. 1989. Absolute configura-
tion of rugosa I A, an endoperoxy carotanoid sesquiterpene. Agric. BioI. Chern. 53:
2505-2507.
38. HASHIDOKO, Y, TAHARA, S., MIZUTANI, 1. 1990. Carota-I,4-dienaldehyde, a
sesquiterpene from Rosa rugosa. Phytochemistry 29: 867-872.
39. HASHIDOKO, Y, TAHARA, S., MIZUTANI, J. 1991. Identification of an intermediate
in autoxidation of carota-I,4-dien-14-al into rugosa I A. 1. Chern. Soc. Perkin Trans. I.
1991: 211-214.
40. HASHIDOKO, Y, TAHARA, S., MIZUTANI, 1. 1991. Intermediates in the autoxidation
of a non-conjugated cyclic diene. 1. Chern. Soc., Chern. Commun. 1991: 1185-1186.
41. HASHIDOKO, Y, TAHARA, S., MIZUTANI, J. 1993. Autoxidation study of carotane
sesquiterpenes possessing a non-conjugated 1,4-diene system. J. Chern. Soc. Perkin
Trans. I. 1993: 2351-2356.
42. HASHIDOKO, Y 1996. The phytochemistry of Rosa rugosa. Phytochemistry 43:
535-549.
43. HASHIDOKO, Y, TAHARA, S., MIZUTANI, J. 199 I. Novel bisabolanoids in Rosa
rugosa leaves. Z. Naturforsch. 46c: 349-356.
420 J. MIZUTANI
44. HASHIDOKO, Y, TAHARA, S., IWAYA, N., MIZUTANI, J. 1991. Highly oxygenated
bisabolanoids in Rosa rugosa leaves. Z. Naturforsch. 46c: 357-363.
45. HASHIDOKO, Y, TAHARA, S., MIZUTANI, 1. 1991. Carotanoids and an acoranoid
from Rosa rugosa leaves. Phytochemistry 30: 3729-3739.
46. HASHIDOKO, Y, TAHARA, S., MIZUTANI, 1. 1991. 2-Phenoxychromones and a struc-
turally related flavone from leaves of Rosa rugosa. Phytochemistry 30: 3837-3838.
47. HASHIDOKO, Y, IWAYA, N., TAHARA, S., MIZUTANI, J. 1991. Changes in carotane
sesquiterpenoids in Rosa rugosa leaves. Nippon Nogeikagaku Kaishi (in Japanese) 65:
1483-1488.
48. HASHIDOKO, Y., TAHARA, S., MIZUTANI, 1. 1992. Rugosal A and related carotane
sesquiterpenes in the glandular trichome exudate of Rosa rugosa. Phytochemistry 31:
779-782.
49. HASHIDOKO, Y, TAHARA, S., MIZUTANI, J. 1992. Sesquiterpene hydrocarbons in
glandular trichome exudate of Rosa rugosa leaves. Z. Naturforsch. 47c: 353-359.
50. TAHARA, S., HANAWA, F., HARADA Y, MIZUTANI, J. 1988. A fungitoxin inducibly
produced by dandelion leaves treated with cupric chloride. Agric. BioI. Chern. 52:
2947-2948.
51. TAKASUGI, M., OKINAKA, S., KAT SUI, N., MASAMUNE, T. 1985. Isolation and
structure of lettucenin A, a novel guaianolide phytoalexin from Lactuca sativa var capi-
tata (Compositae). J. Chern. Soc. Chern. Commun. 1985: 621---.
52. TAHARA, S., NAKAHARA, S., MIZUTANI, 1., INGHAM, 1.L. 1984. Fungal transfor-
mation of the antifungal isoflavone luteone. Agric. BioI. Chern. 48: 1471-1477.
53. HANAWA, F., KANAUCHJ, M., TAHARA, S., MIZUTANI, J. 1995. Lettucenin A as a
phytoalexin of dandelion and its elicitatation in dandelion cell cultures. J. Fac. Agric.
Hokkaido Univ. 66: 151-162.
54. HOMANS, A.L., FUCHS, A. 1970. Direct bioautography on thin-layer chromatograms
as a method for detecting fungitoxic substances. 1. Chromatogr. 51: 327-329.
55. HANAWA, F., TAHARA, S., MIZUTANI, J. 1991. Isoflavonoids produced by Iris pseuda-
corus leaves treated with cupric chloride. Phytochemistry 30: 157-163.
56. HANAWA, F., TAHARA, S., MIZUTANI, 1. 1991. Flavonoids produced by Iris pseuda-
corus leave treated with cupric chloride. Phytochemistry 30: 2197-2198.
INDEX
Abies, 239, 244 Alpinone, 414, 416

A. alba, 228 I-AminocycJopropane-l-carboxylic acid
A. firma, 238 (ACC), 379-380
A. grandis, 224, 229 Ammimajus, 169, 172, 175-176; see also
Abortifacient, 347 Bishops weed
Acacia, 220, 235 Amygdalin, 373-374
A. caffra, 258 Anacardiaceae, 328
A. galpinii, 258, 265 Analgesic, 346-347, 363
A. mearnsii, 268 Anastreptin A, 326
A. melanoxylon, 240, 258 Ancistrocladus, 15
A. nilotica, 226 A. abbreviatus, 15
Acer saccharum, 220 A. korupensis, 14-16
Acetogenins, 104, 106, 108, 111, 113, 116, Angelicin, 171
119,121-122 Angoroside A, 360
adjacent bis-THF, 101, 104, 108-109, Annona muricata, 113
121-122 Annonaceae, 93, 100, 104, 106, 108, 110, 113,
asimicin type, 101, 104 116, 124
bullatacin type, 10 I Annonaceous acetogenins, 93, 120-121,
mono-THF, 101, 106, 108, 113, 122 123-124; see also Acetogenins
nonadjacent bis, 106, 108, 116, 121-122 Annonacin, 101
pesticidal, 93 Antherocerotae (Hornworts), 320, 339
trilobacin type, 101 Anti-AIDS, 12,22; see also Anti-HIV
13-Acetyl-9-Dihydrobaccatin III, 34 Anti-allergenic, 335
Aedes aegyptii, 90; see also Yellow fever Antiangiogenic,61
mosquito (YFM) Antibacterial, 67, 360, 403
A.fzelia, 244 Antibiotic, 3, 25, 242
Agatharesinol, 224-225, 236 Antibody-directed enzyme prodrug therapy
Agrobacterium, 166 (ADEPT), 25
A. tumefaciens,90, 124 Anticancer, 3-8,10-11,13,17,22,26,51,
Ajugol,354 53-55,59--61,90, 121
Aldose reductase inhibitory activity, 365 drug discovery, 7, 90
Aldoxime, 373 Anticarcinogenic, 74, 208
Alfalfa, 135-136, 140-144, 146-147, 150, Anticoagulant, 163
152 Anti-estrogenic, 58-59, 61, 137
Alkaloids, 241-242, 416 Antifeedant, 320, 332-333, 412-413, 417; see
aporphine, 242 also Repellents; Herbivore defense;
Alkyl phenols, 320, 328 Plant defense
Alkylated coumarins, 168 Antifungal, 67, 232, 238, 242, 320-321, 332,
Allelochemicals, 394, 407 362-364,366,394-395,409,
Allelopathic activity, 320. 394 416-417; see also Fungitoxic
421
422 INDEX
Anti-herbivore, 186; see also Herbivore Baccatin, 14

defense; Insecticidal; Pesticidal; Plant Baccatin III, 10, 17, 32-33,40, 44
Defense; Repellents Baccatin VI, 34
Anti-HIY, 8, 10, 14-16,20,75, 163-164,208, Bamboo, 382, 387; see also Bambusa vulgaris
320, 329; see also Anti-AIDS Bambusa vulgaris, 383; see also Bamboo
Antihypertensive, 321 Barbi1ycopodin, 326
Anti-inflammatory, 359-361, 365 Barley, 73, 382-383; see also Hordeum vulgare
Antimalarial, 2, 16, 121,347 Bayin, 243
Antimelanoma, 6 Bean, 140, 143, 149; see also Phaseolus
Antimicrobial, 134-136, 208, 241-242, vulgaris
320-321,332,359-360,394-395, Benchtop bioassays, 90, 124
405,407 Benzaldehyde, 376
Antimitotic, 67 N-Benzoyl phenylisoserine, 33, 42, 45
Antimutagenic, 74 Benzoates: 320
Antineoplastic, 32 Bergapten, 176-177
Antioxidant, 61, 68, 72, 74-75,139,151, 315, Bergaptol, 170, 176-178
361 Berlina, 229, 240
Antipathogen, 241; see also Pathogen defense Bibenzyls, 320, 334, 337
Antiproliferative, 58, 163, 165 Bicyclohumulenone, 322
Antiprotozoal, 366 Biflavanoids, 256-258, 265, 267, 270, 273
Antipyretic, 321 Bioassay guided fractionation, 4, 8, 20, 99,
Antirheumatic, 346-347, 365 110,163,337
Antirhinitic, 321 Bioassays, 89-90, 93
Antiruminant, 205 benchtop, 90, 124
Antitubercular, 68 Biochanin A, 136-138, 146,414,416
Antitumor activity, 3-4, 56, 72, 89-90, 93, 100, Biodiversity, 19-24, 26, 90
120-123, 139,208,347 Bioengineering, 83
Antiviral, 15, 22, 67, 208 Bis-bibenzyls, 320, 329-331, 337-339
Apiaceae, 161, 170 Bisabolane sesquiterpenes, 410
Apigenin, 365, 411-412, 414, 416 Bisabolanoids, 412
Apigenin-7-0-glucoside, 348 Bisaborosao1 A, 410-412
Apocynaceae, 135 Bishops weed, 169; see also Ammi majus
Apoptosis, 121, 139 Bitterness, 320, 325-327, 384
Arabidopsis, 83, 136, 144, 170 Bradykinin, 294-296
A. thaliana, 140 Bradyrhizobium, 136
Arabinogalactan, 232 Brassinosteroids, 138
Arachus hypogea, 73, 273 Breast cancer, 13, 32, 53-54, 57-59, 68, 70,
Ara1iaceae, 99 72, 137-139
Aromadendrin 6-C-glucoside, 234 Brine shrimp, 124; see also Artemia salina
Artemesia, 138 Brine shrimp test (BST), 90-92, 94, 97,
A. salina, 90, 92, 124; see also Brine shrimp 99-100, 113; see also Bioassays
Asimicin, 100-101, 120, 123 Bronchial treatment, 345-346
Asimina longifolia, 104 Bryophytes, 319-321, 339
Asimina triloba, 100-10 1, 106 Bryostatin 1, 6, 11, 17
Asiminacin, 121 Budd1edins, 351, 354, 363-364
Astringency, 289-290, 301, 308-309, 312-314 Buddleja, 343-345, 347-352, 354, 356-363,
Aucubin, 351-352, 354, 360, 363 365-366
Aucuparin, 241-242 B. albiflora, 346, 351
Avocado,96 B. alternifolia, 344
Ayamenins, 414, 416 B. americana, 346-347, 362-363
Azadirachta indica, 226, 237 B. asiatica, 346-347, 363
INDEX 423
Buddleja (cont.) Carex (cont.)

B. cordata, 346, 351, 361, 365 C. kobomugi, 405, 407-408
B. davidii, 344, 346, 348, 351, 354, 363 C. pumila, 405-407
B. difJusa, 346 Carota-1,4-dien-14-aldehyde, 410-412, 417
B. globosa, 344, 346-347, 351, 354, Carota-1,4-dienoic acid, 410-411
360-364 Carotane sesquiterpenes, 410-411, 413, 417
B. madagascarensis, 346, 351 Carotanoids,412-413
B. officinalis, 345-348, 362-365 Cassava, 371, 377, 382-387; see also Manihot
B. sessiliflora, 346, 351 esculenta
bioactivity of, 343, 359, 366 Castalagin, 291-292, 305, 307-309
Buddlejaceae, 344 Castanea, 305
Buddlejone, 355, 364 C. sativa, 233
Buddlejoside A, 352-353 Castanospermum australe, 243
Buddlenoids, 348, 362, 366 Catapol, 351-352, 354, 360, 363
Buddlenols, 357 Catechin, 218, 233, 258, 261-262, 269-271,
Bullatacin, 93, 116, 119-123 273,277,281,294,309,314
Bullatacinones, 10 1, 116, 123 Catharanthus roseus, 4, 10
Cathepsin inhibitory activity, 320, 337, 339
Caffeic acids, 168, 188,233 Cell cultures, 34,36,40,44,90,93, 143, 147,
Calanolides, 20, 163-164 170
calanolide A, 12,20-21, 163-164 Cephalotaxus harringtonia, 4
ca1anolide B, 20-21 Chalcone 2'-O-methyltransferase, 147
Calmodulin inhibitory activity, 320, 334, 336 Chalcone reductase (CHR), 139-143, 150
Calophyllum lanigerum, 20, 163 Chalcone synthase (CHS), 134, 139-143,
Calophyllum teysmanii, 20-21 148-150,223,227
Camalexin, 136 Chalcones, 139, 141-142,277
Camphor, 244 Chamaecyparis obtusa, 237
Camptothecin,4-5 Chemotaxonomic,351
Cancer, 32 Chemotherapy, 32
breast, 13, 32, 53-54, 57-59, 68, 70, 72, Chickpea, 144, 146
137-139 Chloranthaceae, 394, 398, 403, 416
chemoprevention, 72, 134, 138-139 Chloranthus, 394, 397-398, 401, 403, 416-417
colon, 56-57, 68, 70, 72 C. glaber, 394
growth inhibition, 72; see also Anticancer C.japonicus, 394-399, 416
lung, 5, 32 C. serratus, 397-403
ovarian, 13-14,32 4/3-Chloroftavan-3-01, 265
prostate, 53, 57-58,68,72, 119, 122, 138 Chlorogenic acid, 191,205,309,314
protective effects against, 51, 55-57, 59--61, Cholesterol reduction, 72, 74
67, 139 Cinnamates, 320
preventive, 70-71, 83, 138 Cinnamic acid, 42-43, 167
skin, 57 Colon cancer, 56-57, 68, 70, 72
stomach, 139 Colophospermum mopane, 258
Carbohydrates, 216, 222, 229, 232, 234, 237, Compositae, 135,412-413,415,417
240, 244-245 Condensed tannins, 186, 205-206, 208, 217,
Carboxypetidases, 377 303-304; see a/so Proanthocyanidins
Carcinogenesis, 51, 54 a-Conidendrin, 219, 233, 243-244
Cardiotonic and vasopressin antagonist activity, Conifery1 alcohol, 76-80
320, 336 Conioids, 236
Cardiovascular disorder protection, 256 Conocepha/um, 322, 324, 329, 332
Carex, 403-405, 407, 417 Contact dermatitis, 320, 328
C. fedia, 403, 407 Coronary heart disease, 68, 70, 74
424 INDEX
Coumarins, 161-167, 172,241; see also Dehydrogerruginol, 224

Furanocoumarins 5-Dehydroshikimic acid, 188-189
acylated, 168 Deidac1in, 373-374
angular, 170, 172 Demethylsuberosin, 169-172
biosynthesis, 161, 168 Dhurrin,371-374
dihydropyrano, 163 Diabetes, 384
hydroxy, 163-165 Diglucosidases, 375-376, 385-386
isocoumarins, 320 Dihydroflavonols, 275-277, 282
polyoxygenated, 168 Dihydrokaempferol, 244
Coumestrol, 135 Dihydropyranocoumrins, 163
Crambe abyssinica, 304 Dihydrorobinetin, 236-237
Crambin, 304 2,5-Dihydroxy benzoic acid, 241
Creosote bush, 72, 75; see also Larrea 5,8-Dihydroxypsoralen, 176, 178
tridentata Diospyros, 229, 240
Crown gall tumors, 90, 94, 124 Diphenolic compounds, 70
Crush resistance, 215, 232 Diplophyllolide, 325
Cryptomeria japonica, 220-222, 224-226, 229, Dirigent proteins, 76-80, 81, 83
236,243 Disease resistance, 136
Cupressaceae, 235 Diterpenoids, 37, 97, 320,325-327,333-334,
Cyanogenesis, 370-371, 375-378, 380, 383 339,351,354
weakly cyanogenic plants, 376, 378-382, Diuretic activity, 321, 329, 345-346
384, 387 DNA polmerase 13 inhibitory, 320, 329, 332
Cyanogenic, 369-371, 376-380, 384 Dolastatin 6, 10-11, 17
Cyanogenic diglucosides, 371, 373, 375-376 Dormancy, 378
Cyanogenic glucosides, 205, 369-376, Douglas fir, 234
378-380,382-388 Drug development, 1,7,9-12, 18-21, 163
biosynthesis of, 371-372, 376, 384-387 Drug discovery, 1,3, 7, 13, 17-21,23,26,30
evolutionary relationships of, 377 Dryobalanops aromatica, 244
health disorders of, 384 Dryobalanops lanceolata, 244
translocation of, 371, 373, 375-376, 385 Duckweed, 90, 124; see also Lemna minor
Cyclo-olivil, 241-242, 244 (R)-Dodec-2-3n-I,5-olide, 322
Cyclocolorenone, 351, 354
Cyclooxygenase inhibitory activity, 320, Echinacoside, 349, 361-363
333-334 Ecochemicals, 393-394
Cyclpentenyl glycine, 373-374 Ellagic acids, 186, 205, 233, 241
Cytochrome P450 reductase, 172 Ellagitannins, 186, 188, 196, 203-204,
Cytochrome P450 oxygenases, 32, 38-39, 58, 206--208, 233, 291; see also
143-144,170,172-173,176,179, Hydrolyzable tannin
371-372 Enterodiol (ED), 51-53, 57-59, 61, 68-71,
Cytostatic activity, 32, 42 73
Cytotoxic, 72, 93, 96--97, 99, 101, 106, Enterolactone (EL), 52, 57-59, 61, 68-73
110-111, 113-114, 119, 121,320, Epicatcehin, 219, 233, 258, 261-262, 269, 271,
329, 332, 360 273,277,297,309,314
Epicatechin-3-0-gallate, 297
Dacrydium, 244 Epifisetinidol-413-01, 258, 260
Daidzein, 135-139, 143, 145-146, 148 Epigallocatechin gallate, 219, 296-298, 309,
Dalbergia, 229 314
Daphnetin, 168 Epipodophyllotoxin, 4, 17
10-Deactylbaccatin III, 11,32-33,37,40 Equol, 137-138
Decay resistance, 215 Esculetin, 162, 165, 168
Defense mechanisms, 219, 245 Estradiol, 53, 58, 137-138
INDEX 425
Estrogenic, 58-59, 61, 139 Furanocournarins, 163, 168, 170-172, 176

Ethnopharmacology, 344-345, 363 angular, 170
Ethylene, 220, 224, 226-227, 229, 236-238, biosynthesis of, 179
245, 370, 379-380 linear, 165; see also Psoralens
Etoposide, 4-5, 10, 17,74-75 psoralens, 165, 169-172, 174-175, 177-178
Eucalyptus, 222, 228 Furanogermacranes, 395, 399
E. hemiphloia, 219 Furanosesquiterpenes, 395-396
E. marginata, 232, 234, 241
E. polyanthemos, 226 Gallic acids, 186-190, 196-197, 199,208,
E. tereticornis, 236 233-235
Eudesmanolides, 325, 332 Gallotannins, 186-188, 190, 196-199,203,
Eudesmin, 219 206-208
Euphorbiaceae, 93, 97, 124 biosynthesis of, 197-203
Extractives, 215-219, 222-223, 225, 227-228, Galloyl esters, 293
230-235,238-240,242,245 1-0-Galloyl-[3-D-glucose, 189; see also
compartmentalization, 239 [3-Glucogallin
Galloylglucoses, 197,205
Fagus, 239 degradation of, 203
F sylvatica, 239-240 Galloyltransferases, 196, 199, 201-202, 207
Ferruginol, 224-225, 236 Genistein, 136-139, 143, 146,414,416
Ferulic acids, 168, 233 Gentisic acid, 241
Fiber, 68, 70, 74, 215, 217, 221 Geosmin, 323
Fisetinidol-(4a->8)-catechin, 268-271 Geranylgeranyl diphosphate (GGPP), 33, 35-36
Flavan-3-0Is, 182,256-257,261-262,264,275, Geranylgeranyl diphosphate synthase, 32, 35, 37
277-279, 293 Germacranes, 395, 399
Flavan-3,4-diols, 257, 265 Germacranolide, 325
Flavones, 136,361,412,416 Gigantetrocin A, 101
isoflavones, 53, 135, 137, 139, 146 Ginko bi/oba, 328
Flavonoids, 134, 136, 138,207,219,223, Glaucine, 241-242
226-227,235,238,241,243,275, (3-Glucogallin, 187, 189-195, 197,201-202;
345, 347-348, 361-362, 365-366, see also 1-0-Galloyl-[3-D-glucose
415-417 [3-Glucosidases, 197,370-371,373,376-377,
biflavanoids, 256-258, 265, 267, 270, 273 383
isoflavonoids, 70, 134-138, 140-141, 144, Glucosinolates, 205
148,151,153,415-417 Glucosyitransferase, 190
polyflavanoids, 256, 268, 278-279, 281 Glyceollin, 135, 151-152
tetraflavanoids, 256, 262 Glycine max, 73
triflavanoids, 256-258 Glycyrrhiza echinata, 141
Flavonols, 243 Gmelina leichardtii, 244
dihydro, 275-277, 282 Gmelinol, 244
Flax, 72, 81-83, 371, 383; see also Linum Gomisin A, 72, 74-75
usitatissimum Grains, 54, 61, 67-68, 70, 72-73, 83
Flaxseed, 54-57, 59-61, 72-74, 80 Grape, 405
Flickering cluster theory, 302 Grima1done, 324
Formononetin, 137, 145-146 Guaianolides, 337, 417
Forsythia intermedia, 74, 76-77, 79-81 Guananbana, 113-114
Fraxinus excelsior, 228 Guttiferae, 163
French paradox, 256 Gymnocolin A, 326, 331
Frullania, 322, 328-329, 334
Frullanolide, 328 Harringtonine, 6
Fungitoxic, 413, 415; see also Antifungal HeN, 369-371, 376, 378-380, 382-384, 387
426 INDEX
HCN potential, 380-384, 386 Intsia bijuga, 243-244

Health-promoting, 209, 256, 283 Iodine metabolism, 384, 387
Heartwood, 215-216, 218, 220-237, 239-243, lpomeanol, 4--6
245 Ipomoea balatas, 4; see also Sweet potatoes
extractives, 215-219, 222-223, 225, 227-228, Iridaceae, 415--417
230-235,238-240,242-243,245 Iridoids, 345, 347, 351-352, 358, 360, 362; see
phenols, 238 also Terpenoids
Hemicellu10ses, 217, 223, 230, 233 Iris pseudacorus, 414--417
Hepaticae, 320, 325, 328, 339; see also lsocoumarins, 320
Liverworts lsoflavone O-methyltransferase (IOMT), 134,
Herbal medicines, 290, 315; see also 146-148
Traditional medicine Isoflavone synthase, 134-135, 143-144,
Herbivore defense, 186, 370, 387; see also 146--147
Anti-herbivore; Insecticidal; Isoflavones, 53, 135, 137, 139, 146
Pesticidal; Plant defense, Repellents Isoflavonoids, 70, 134-138, 140--141, 144, 148,
Heterodendrin, 373-374 151,153,415--417
Hevea, 371, 377; see also Rubber biosynthesis, 133, 139, 150, 152
H. braziliensis, 237, 377 5-Deoxy, 141
Hispidulin, 414, 416 isoflavones, 53, 135, 137, 139, 146
Homoharringtonine, 4--6 pathway, 150
Hordeum vulgare, 379; see also Barley phytoalexins, 140-141, 144-145
Hyaluronidase inhibitory activity, 320, 333, 335 phytoestrogen, 139
Hydrolyzable tannins, 185-187, 189, 191, pterocarpans, 135-136, 141, 144
205-208; see also Ellagitannins; synthesis, 148
Gallotannins Isoformononetin, 145-146
biosynthesis of, 187-203 Isopimpinellin, 176, 178
Hydrophobic effects, 290, 298-300, 303-307
8-Hydroxybergapten, 176 Juglans nigra, 237, 240
4-Hydroxycoumarins, 163-165
7a-Hydroxydiplophyllolide, 325 Kaempferol, 348, 361
5-Hydroxymarmesin, 176--177 Kaempferol 6-C-glucoside, 234
Hydroxymatairesinol, 219, 233, 238, 243-244 Kaempferol-3-rhamnoside, 244
Hydroxynitrile lyases, 370-371, 376-378 Keyakinol, 234
Hydroxynitriles, 369-371, 373, 375, 380 Kino, 219-220,237
Hydroxypsoralens, 170, 176 Kobophenol A, 405, 407, 417
1-~-Hydroxysacculatal, 325, 335 Korupensamines, 16
Hydroxysugiresinol, 225, 236 Kudzu vine, 140; see also Pueraria lobata
5-Hydroxyxanthotoxin, 176, 178
Laccases, 76, 78-80
Iceberg principle, 304 Lamiaceae, 99, 124
Infuscaic acid, 333, 337 Lariciresinol, 77, 81
Infuscasides, 326, 333, 336 Larix leptoiepis, 222, 229
Insect attractant, 366 Larix occidentalis, 232
Insect resistance, 215, 232; see also Larrea tridendata, 75; see also Creosote
Anti-herbivore bush
Insecticidal, 68, 399; see also Antifeedant; Lauraceae, 94, 124
Herbivore defense; Pesticidal; Plant Legumes, 72-74, 136, 140-142, 153
defense Leguminosae, 135
lntegrase, 163-165 Lemna minor, 90, 124; see also Duckweed
lnterflavanyl bond, 260, 262-263, 265, Lettucenin A, 413-415, 417
268-270,273 Leukemia, 4, 6, 93, 123
INDEX 427
Lignans, 51, 53--61, 67--68, 73, 76, 219, 233, Matairesinol, 52, 54, 68-74, 76--77, 81-82,
235,241-242,244,347,351,354, 219,233,244
357-358 Maura pomifera, 239
biosynthetic pathways, 68, 76--83 Mayapple, 72, 74; see also Podophyllum
mammalian, 51-55, 57, 60--61, 68-73, 83 Maytenone, 355, 364
mechanisms, 58-59 Medicago truncatula, 148
metabolism, 74 Medicarpin, 135-136, 146-147
plant, 52-53 Medicinal plants, 320, 345, 347
Lignification, 218, 227 Mekuba seaweed, 73
Lignins, 207, 217-219 Melanoma, 13
Linamarase,377 Melia volkensii, 94
Linamarin, 374, 383, 385-386 Meliaceae, 94, 124
Linarin, 348, 361-365 Mesuaferrea, 237
Lindename dimers, see Sesquiterpenoids 8-Methoxypsoralen, 165
Lindenanolides, 399, 403, 416 Methyl thujate, 244
Lindenine, 394, 396 7-0-Methyldihydrokaempferol, 414, 416
Unum, 80, 377-378; see also Flax Methylesculin, 162, 165
L. usitatissimum, 82, 377 O-Methyltransferase, 146, 150, 166, 168, 176,
Linustatin, 385-386 242
Lipids, 216, 222-223, 237 Metrosideros polymorpha, 221
Lipochitooligosaccharide signal molecules, Mevalonate biosynthetic pathway, 37, 170,
136 224
5-Lipoxygenase inhibitory activity, 163, 165, Michellamine B, 12, 14-16
320, 333-334, 336, 361 Microberlinia brazzavilliensis, 229, 240
Uriodendron tulipfera, 241 Mimengosides, 351, 356, 365-366
Liver disease treatment, 345-346; see also Mixtures, 60, 100, I 16
Liver protectant Miyabeno1s, 404--405, 407--408, 417
Liver protectant, 74 Molecular mechanisms, 315
Liverworts, 319-322, 325, 328--329, 332, 335, Molecular recognition, 290, 315
337, 339 Molluscicidal activity, 335, 366
biological activity of, 320, 325, 329, Monoacylglycerides, 99
332-335, 337 Mononetin, 146
bitterness, 320, 325-327 Monoterpenoids, 37, 233, 235, 320, 324,
odor, 321-322 339
pungenc~ 320, 325-327, 332 Moraceae, 135, 138
Loganiaceae, 344, 351 Musci, 320, 338
Lotaustralin, 373 Muscle relaxing activity, 320, 335, 337
Lung cancer, 5, 32 Myristicaceae, 94
Lunularic acid, 334
Luteolin, 348, 361, 365 NADH oxidase, 121, 124, 151-153
Luteolin-7 -O-glucoside, 348, 361 Naphthalenes, 320
Lymphomas, 6, 13 Naringenin, 143
Naringenin chalcone, 140-141
Malus pumila, 241 Nematocidal, 320, 332-333
Mammalian Iignans, 51-55, 57, 60-61, 68-73, Nematode defense, 363
83 Neurite sprouting activity, 320, 335
Manihot esculenta, 377, 379, 383, 385-386; Nicotine, 337
see also Cassava Nod genes, 134, 136
Marchantia, 328-329, 332-334 Nordihydroguaiaretic acid, 75
Marchantins, 329, 332, 335, 337 Norlignans, 236
Marmesin, 170-172, 174-177 Nothofagus cunninghamii, 239
428 INDEX
Oak, 190--191, 196,207,240; see also Phenylethanoids, 347-350, 360, 362, 365
Quercus Phenylalanine aminomutase, 43-44
I-Octen-3-ol,321 Phenylalanine ammonia lyase, 42-43, 145, 149,
Olea cunninghamii, 244 152-153,227-228
Oleoresin, 234-235 Phenylpropanoids,68, 134, 166,223,347,351,
Oligopeptides, 281, 294 362
Oligosaccharides, 377 pathway, 74-76, 166
Oligostilbenes, 403-405, 407-408, 417 Phorbol,97
Orobanchin, 348-349 Phospholipases, 223, 236
Osthenol, 169-172 Photosensitizing, 165
Ovarian cancer, 13-14,32 Phototoxic, 163
Oxidative process inhibition, 359 Phthalides, 320
Phytoalexins, 134-136, 140--141,144, 148,
P388 leukemia, 6, 100 150--152, 168,238,403,405,413
Pacific Yew tree, 13, 14, 32; see also Taxus Phytoestrogens, 53, 60, 70--71,83, 137-138
brevifolia Phytotoxic, 335
Paclitaxel,4--6, 10, 13-14, 17; see also Taxol Picea abies, 222-223, 238
Palmae,·97 Picea mariana, 224
Parsley, 141, 172, 176 Pinaceae, 100, 135, 235
Parvifloracin, 106 Pinoresinol, 68--69, 74, 76-78, 80--81
Passifloraceae, 373 Pinoresinolliariciresinol reductase, 76, 80--81,
Pathogen defense, 232, 239, 242, 363; see also 83
Antipathogen Pinosylvins,237
Pathogen inhibition, 415, 417 Pinus, 218-220,224,228-229,233,237-238
Pathogenesis-related proteins, 151 P. banksiana, 228
Paw paw, 100, 104, 123; see also Asimina P. canariensis, 220
triloba P. contorta, 241
Pea, 135; see also Pisum sativa P. densifiora, 221
Pellia, 325, 329, 333, 335 P. japonica, 230
Pentagalloylglucose, 186-188, 190-191, P. radiata, 220, 225-228, 236--237, 239
193-194,196--197,199-204, P. resinosa, 217
206-208,291-292,294,296-298, P. strobus, 238
305-309 P. sylvestris, 220--221, 223, 226, 230-231
Pentosephosphate pathway, 227 P. taeda, 237
Perrottetins, 329, 334, 337 Pisatin, 135-136
Persea americana, 96 Piscicidal, 320, 335, 351
Pesticidal, 89-90, 93, 100 Pisum sativum, 147
Pesticide, 97, 106, 121-122 Pithecellobium dulce, 258
Phaseolus vulgaris, 73, 140; see also Bean Plagiochila, 322, 325, 329, 332, 335
I3-Phellandrene,241-242 Plagiochilal B, 326, 333, 335
Phellopterin, 162, 165 Plagiochilines, 326, 332-333, 335
Phenol oxidizing enzymes, 226--227 plagiochiline A, 325, 329, 335
Phenolics, 51, 187, 189,208,226,236,242, Plant defense, 134, 151,215,232,290,312,
245,255,275,291,296-300,303, 376,393-394,407,416-417; see also
306,308,314-315,338,354,360, Anti-herbivore; Herbivore defense;
415-416; see also Polyphenols Pesticidal
alkyl, 320, 328 Plant growth inhibitory activity, 320, 335
diphenolic compounds, 70 Plant lignans, 52-53
fatty acid esters, 354, 359 Podocarpid acid, 244
Phenolic O-methyltransferases, 144 Podocarpus, 244
Phenoxychromones, 412 Podophyllotoxin, 32, 72, 74-75
INDEX 429
Podophyllum emodii, 10 Protein-polyphenol complexation, 197,

Podophyllum peltatum, 4 292-293,295,297,299-301,304,
Polyflavanoids, 256, 268, 278-279, 281 307,311-314; see also Polyphenol-
Polygodial, 325, 329, 333, 335 protein interactions
Polyketide reductase, 141 Proteinase inhibition, 152, 360
Polyketides, 25-26, 141, 166 Proteracacinidins, 262, 264-266
Polyoxygenated coumarins, 168 Protocatechuic acid, 188-189
Polypeptides, 149,227 Prunasisn,373-374
Polyphenol-protein interactions, 289-291, Prunus, 220, 235,237,241, 371. 377-378
294,301,308,314-315; see also P avium, 379
Protein-polyphenol complexation P domestica, 241
hydrophobic effects, 290, 298-300, 303-307 P jamasakura, 241
Polyphenoloxidases, 290 P serotina, 236, 377
Polyphenols, 185,206-208,218-219,221,223, P yedoensis, 226-227, 241
225-227,231-235,237-239,283, Pseudosindora palustris, 218, 236
289-291,293-296,298-301,303, Pseudotsuga japonica, 221
306,308-309,311-315 Pseudotsuga menzies ii, 241, 244
Parella, 322, 325, 329, 332-335, 337 Psoralen 5-monooxygenase, 176
Potato, 168 Psoralen synthase, 172, 175-176
Potato disc bioassay, 90-91, 94, 124 Psoralens, 165, 169-172, 174-178; see also
Pratensein, 414, 416 Furanocoumarins
Prenyl ethers, 332 Pterocarpans, 135-136, 141, 144
Prenyltransferase, 168, 170, 176 Puberulin, 168
O-Prenylumbelliferone, 169-170 Pueraria, 144
Proanthocyanidins, 186,206,219,226,233, P lobata, 140-141, 143-144; see also
241,255-256,262,265,267-269, Kudzu vine
280, 282, 291, 294; see also Pungenc~ 320, 325-326, 335
Condensed tannins Pusilatins, 332
A-type, 271, 273-275 PUVA therapy, 165
B-type, 268, 273 ;\.-Pyrufuran,241-242
dimeric, 278
polymers, 255-256, 279 Quebracho tannins, 258
oligomers, 256, 258, 265, 275, 277-278, 282 Quercetin, 348, 361
synthesis of, 256-265, 275-277, 279, 282 Quercetin 6-C-glucoside, 234
Procyanidins, 260-262, 267 Quercus, 206, 222, 228, 233, 240-241, 305
procyanidin B-1, 269-270, 280 Q. iberica, 227
procyanidin B-3, 262-263, 268, 270, 280-281 Q. infectoria, 199
Profisetinidins, 256-258, 260, 262, 268-270 Q. pedunculata, 191
Proline-rich proteins (PRPs), 294, 296, 298, Q. petraea, 233
301,309-314 Q. robur, 191, 196
Propanoids, 134 Q. rubra, 196
Propenes, 278
Prorobinetinidins, 268 Radish
Prostate cancer, 53, 57-58, 68, 72, 119, 122, Radula, 322, 334, 337
138 Reaction zones, 238, 240-241; see also
Proteases, 163-164 Protection wood
inhibitors, 152, 360 Red wine, 256
Protection wood, 238; see also Reaction zones Repellents, 68, 312, 377-378, 387; see also
Protein kinases, 150, 165-166 Antifeedant; Herbivore defense;
Protein tyrosine kinases (PTK), 93-94, 139, Plant defense
165 Resins, 219, 221-222, 226, 231, 239
430 INDEX
Resin acids, 223, 233 Sernoa repens, 99

Respiration inhibition, 370, 381 Sesame, 72, 74, 82-83; see also Sesamum
Resveratrol, 407-408, 417 indicum
Reverse transcriptase, 163-164 Sesamin, 72, 74-75
Rhizobium nod genes, 136 Sesamolinol,74-75
Rhizobium-legume symbiosis, 136 Sesamum indicum, 72, 74
Rhodanase, 381-382, 387 Sesquiterpene lactones, 325, 328, 339, 394, 396
Rhodanide, 381; see also Thiocyanate Sesquiterpenoids, 233, 320, 326, 329-331,
Rhus, 206-207 333-334,338-339,351,354,
R. succedanea, 237 394-400,407,410,413,416
R. typhina, 189, 191, 196--199,201-203 bisabolane, 410
Riccardins, 329, 334 carotane, 410-411, 413, 417
Robinetin, 236, 243-244 dimers, 396--398,401-402,416
Robinia, 220 furano, 395-396
R.pseudoacacia, 223-224, 226--230, 236, 239 hemiacetals, 325
Robustic acid, 152, 166 ketone, 324, 398
Rosa rugosa, 407, 409-413, 417 lactones, 325-326, 328, 339, 394, 396
Rosaceae, 407, 417 secoaromadendrane type, 325, 329
Rosy periwinkle, 4; see also Catharanthus trimers, 397
roseus Shakes, 243-244
Rubber, 168,237; see also Hevea, H brasiliensis Shikimate pathway, 166, 189,219,223-224,
Rue, 176, 196 235
RugosalA, 407,409-410, 412-413, 417 Shikimic acid, 188-189,219,227,233
Rugosic acid, 409 Shizukanolides, 394-396, 399-401
Rugosin D, 291-292, 309 shizukanolide B, 394-396, 401, 403, 416
Rutin, 348 Shizukaols, 396-400, 402
Rye, 73-74, 82; see also Secale cereale Shorea robusta, 226
Signal transduction pathways, 134, 139, 148,
Sacculatal, 325, 329, 333, 335 151, 153
Saikosaponins, 351, 361, 365 Skatole, 321
Salicylic acids, 327 Skin cancer, 57
Samanea saman, 237 Solanaceae, 413
Saponins, 351, 356, 362-363, 366 Solanum tuberosum, 90, 124
Sapwood, 216-217, 220--222, 224-230, Sorghum, 371, 373, 377
232-234,237-240,242,244-245 S. bicolor, 294, 377, 379
Sativan, 135 Soybean, 83, 134-136, 138, 140-141, 143, 146,
Saw palmetto, 97 152; see also Glycine max
Scapanin A, 326 Squamotacin, 199, 122
Schinopsis, 234, 269 Squamocin, 101
Schizandra chinensis, 72, 74 Sterols, 236
Scopoletin, 168, 241-242 Stilbenes, 227, 235, 238, 243
Secale cereale, 74 antifungal, 238
Secoaromadendrane-type sesquiterterpenoids, oligostilbenes, 403-405, 407-408, 4 I 7
325, 329 tetrastilbenes, 405-406, 4 17
Secoisolariciresinol, 68-74, 76--77, 81-82 Stomach cancer, 139
Secoisolariciresinol dehydrogenase, 77, 81, 83 Structure-activity relationships (SAR), 42,
Sccoisolariciresinol diglucoside (SDG), 52, 121-122, 124, 136, 163,291,293
54-57,59-61,69,72,81-82 Styryllactones, I 10
Secoisolariciresinol glucosyltransferase, 82 Suberin, 359
Sedative/Tranquilizing effects, 346-347 Subterranean clover, 146; see also Trifolium
Sequirin, 225, 236 subterraneum
INDEX 431
Succinate dehydrogenase, 226 Terpenoids (cont.)

Sugiresinol, 224-225, 236 iridoids, 345, 347, 351-352, 358, 360,
Sumac, 191, 196, 207; see also Rhus typhina 362
Superoxide anion radical release inhibitory monoterpenoids, 37, 233, 235, 320, 324,
activity, 320, 333-334 339
Sweet potatoes, 4, 168; see also Ipomoea triterpenoids, 94, 241-242, 351, 356,
batatas 361-363,365
Synergistic interactions, 60, 152 tetraterpenoids, 37
(R)- Tetradec-2-en-1 ,5-olide, 322
Tamariscol, 322-324 Tetraflavanoids, 256, 262
Tamoxifen, 53, 137 Tetrastilbenes, 405-406, 417
Tannases, 205-206 Tetraterpenoids, 37
Tannic acid, 208, 303 Thiocyanate, 381-383
Tannins, 185-186,208,268,291,294,299, Thuja, 241-242
301,303-304,307-308,312 T orientalis, 221
condensed tannins, 186,205-206,208,217, T plicata, 222, 229, 241, 244
303-304; see also Proanthocyanidins Thujaplicins, 235, 241
gallotannins, 186--188, 190, 196-199,203, Tobacco, 152, 168
206--208 Tomentellins, 332
hydrolyzable tannins, 185-187, 189, 191, Topotecan,4-5
205-208; see also Ellagitannins; Tothecin,4
Gallotannins Traditional medicine, 360-361, 365; see also
Taraxacum officinale, 413-415, 417 Herbal medicines
Taxa-4(20), II (12)-dien-5a-ol transfrerase, Transacylase, 40
39 Transgenic plants, 83, 134, 136, 147,
Taxadiene synthase, 32, 35-37 150, 153, 385
Taxifolin 6-C-glucoside, 234 Transition wood, 237
Taxodiaceae, 100 Transition zone, 223, 225-230, 232, 234,
Taxadienyl acetate, 38-39 236--238, 243, 245
Taxoid hydroxy lases, 40 Trichocolea, 329, 332
Taxol, 4-5, 14, 17,32-38,40-45, 123 Trichomes, 345, 412-413, 417
biosynthesis, 31-45 Triflavanoids, 256--258
cell cultures, 36-37 Trifolium repens, 377
Taxotere, 17, 32, 34; see also Taxol Trifolium subterraneum, 146; see also
Taxus, 10, 14,34-35,44 Subterranean clover
cell/tissue culture, 14,34,37 Triglochinin, 371
T baccata, 14,40,225 Trihydroxychalcones, 140-141, 145
T brevifolia, 13-14,32,36-37,42,44-45; Trihydroxycinnamic acid, 188
see also Pacific Yew tree Trihydroxyflavanone, 414, 416
T canadensis, 35-36, 38, 40 3,5,4'-Trihydroxystilbene, 403; see also
T chinesis, 37 Resveratrol
T cuspidata, 40 Trihydroxylated hydrocarbons, 96
Tea, 72, 74, 256, 290 Trilobacin, 101, 121
Tectona grandis, 223, 226, 240 Triterpenoids, 94, 241-242, 351, 356, 361-363,
Teniposide, 5, 10, 17,74-75 365
Terpenoids, 227, 232, 234, 238, 241, 319-320, Tropolones, 235
339,351,366,416; see also Tsuga, 233, 239
Sesquiterpenoids T heterophylla, 243-244
aldehydes, 410 T mertensiana, 244
diterpenoids, 35, 37, 97, 320, 325-327, Tubocurarine, 335, 338
333-334,339,351,354 Tulipinolide, 325
432 INDEX
Tyrosinase inhibitory activity, 362, 366 Weakly cyanogenic plants, 376, 378-382, 384,
387
Ulmus glabra, 241 Wheat, 73, 82-83
Umbelliferone, 161-162, 167-169, 171, 176 Wound healing, 345-346, 359-362
Uniform continuum model, 302 Wounding, 237-238
Urinogenital disinfectant, 347
Xanthotoxin, 178
Vasodilatory effects, 208 Xanthotoxol, 176, 178
Verbascoside, 349, 360-363
Vescalagin, 291-292, 305, 307-309 Yellow fever mosquito (YFM), 90, 97, 99, III,
Vinblastine, 4-5, 10, 32, 122 121, 124
Vincristine, 4-5, 10, 122
Viniferins, 403, 405-408, 417 Zelkova serrata, 218, 234

(Recent Advances in Phytochemistry 33) Gordon M. Cragg - Michael R. Boyd (Auth.) - John T. Romeo (Eds.) - Phytochemicals in Human Health Protection - Nutrition - and Plant Defense-Springer US (1999)

Diunggah oleh

Informasi Dokumen

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

(Recent Advances in Phytochemistry 33) Gordon M. Cragg - Michael R. Boyd (Auth.) - John T. Romeo (Eds.) - Phytochemicals in Human Health Protection - Nutrition - and Plant Defense-Springer US (1999)

Diunggah oleh

Hak Cipta:

Format Tersedia

recent advances in phytochemistry

Recent Volumes in the Series:

Volume 24 Biochemistry of the Mevalonic Acid Pathway to Terpenoids

Volume 25 Modern Phytochemical Methods

Volume 26 Phenolic Metabolism in Plants

Volume 27 Phytochemical Potential of Tropical Plants

Volume 28 Genetic Engineering of Plant Secondary Metabolism

Volume 29 Phytochemistry of Medicinal Plants

Volume 30 Phytochemical Diversity and Redundancy in Ecological

Volume 31 Functionality of Food Phytochemicals

Volume 32 Phytochemical Signals and Plant-Microbe Interactions

Volume 33 Phytochemicals in Human Health Protection, Nutrition, and

Springer Science+Business Media, LLC

Phytochemicals in human health protection, nutrition and plant delense

AII rights reserved

metabolism and development has been proposed. Finally, Mizutani summarizes

I. Natural Product Drug Discovery and Development: The United

2. Taxol Biosynthesis: A Review of Some Determinant Steps . . . . . . . . .. 31

3. Role of Lignans in Carcinogenesis ............................. 51

4. Toward Engineering the Metabolic Pathways of Cancer-Preventing

5. Simple (Bench-Top) Bioassays and the Isolation of New

6. Molecular Controls for Isoflavonoid Biosynthesis in Relation to

7. Medical Potential and Biosynthesis of Plant Coumarins ............. 161

8. Biosynthesis, Biodegradation, and Cellular Localization of

9. The Formation of Heartwood and Its Extractives: An Overview ...... 215

10. Recent Advances in the Chemistry of Proanthocyanidins ............ 255

11. Astringency and Polyphenol Protein Interactions .................. 289

12. Phytochemistry of Bryophytes: Biologically Active Terpenoids and

13. Biologically Active Compounds from Buddleja Species ............. 343

14. Cyanide in Foods: Biology of Cyanogenic Glucosides

Index ....................................................... .421

NATURAL PRODUCT DRUG DISCOVERY AND

Gordon M. Cragg,l* Michael R. Boyd, I Rita Khanna,2

IDevelopmental Therapeutics Program

Office of the Director

* Author to whom enquiries should be addressed.

National Cooperative Drug Discovery Group Program (NCDDG) 18

surface, represent an enormous resource for the discovery of potential chemother-

ANTICANCER AGENTS DERIVED FROM NATURAL

Taxal (Ph = Phenyl)

Figure 1. Commercial plant-derived anticancer drugs,

Figure 2. Plant-derived anticancer drugs in clinical development.

CURRENT STATUS OF THE NCI NATURAL PRODUCTS

qualified organizations in each of the source countries. Botanists and marine

DECISION NETWORK PROCESS

-·IIA------.· liB -III--+IND

Phase I studies are conducted to determine the maximum tolerated dose

NATURAL PRODUCT DRUG DEVELOPMENT: THE SUPPLY

Figure 5. Some natural product-derived anti-AIDS agents discovered by the NCr.

Potential Anti-HIV Agent, Michellamine B

Michellamine B (Fig. 5) was isolated as the main in vitro active anti-HIV

II' 10' 10'

Figure 6, Korupensamines A-D: plant-derived potential antimalarial compounds.

A. korupensis, provides another class of potential medicinal agents from this

COLLABORATION IN DRUG DISCOVERY AND

In the case of organizations wishing to have pure compounds tested in the

Collaboration in Preclinical Development: Rapid Access to

RAID is a new program designed to facilitate translation to the clinic of

National Cooperative Drug Discovery Group (NCDDG) Program

Source Country Collaboration

Drug Discovery: Memorandum of Understanding. As discussed, the

With the increased awareness of genetically-rich source countries to the

Drug Development: The Calanolides. In 1988, an organic extract of the