Advances in biocatalytic synthesis of pharmaceutical

intermediates
Sven Panke1 and Marcel Wubbolts2
Biocatalytic synthesis of enantiomerically pure compounds for
pharmaceutical intermediates is gaining momentum. This is the
result of advances in genomics, screening and evolution
technologies leading to the increased availability of new and
robust biocatalysts suited for industrial-scale application, and
is stimulated by an increased demand for catalysts that are able
to address the increased complexity of active pharmaceutical
ingredients. The vast majority of biotransformation reactions for
the manufacturing of optically active pharmaceutical
intermediates are still based on enantioselective ketone
reductions and enantiospecific hydrolyses. This review aims to
point at alternative reaction types and integrated multienzymatic steps that are emerging in large-scale applications.
Addresses
1
Institute for Process Engineering, ETH Zurich, Sonneggstrasse 3,
ML K 23, CH-8092 Zurich, Switzerland
2
DSM Research, Advanced Synthesis Catalysis & Development,
P.O. box 18, 6160 MD Geleen, The Netherlands
Corresponding author: Wubbolts, M (marcel.wubbolts@dsm.com)

Current Opinion in Chemical Biology 2005, 9:188194

This review comes from a themed issue on
Biocatalysis and biotransformation
Edited by Dan E Robertson and Uwe T Bornscheuer

1367-5931/$ see front matter

# 2005 Elsevier Ltd. All rights reserved.
DOI 10.1016/j.cbpa.2005.02.007

Introduction
The molecular complexity of active pharmaceutical
ingredients (APIs) is rising and is characterized by a
growing number of functional groups and chiral centers.
To synthesize an advanced intermediate (AI) to assemble
such a complex API requires catalysts that are highly
selective and that operate under conditions compatible
with frequently sensitive functional groups. Enzymes, as
highly selective catalysts operating under ambient conditions, are particularly suited for this purpose, and
numerous industrial applications have been reviewed
recently [1,25].
A long-known but illustrative example of an API with
multiple chiral centers created by biocatalytic means is
Omapatrilat, an antihypertensive drug that acts both as an
inhibitor of angiotensin converting enzyme and neutral
endopeptidase [5]. Omapatrilat contains four chiral
Current Opinion in Chemical Biology 2005, 9:188194

centers as well as several sensitive functional groups

(Figure 1). The molecule can be built-up from alternatively (S)-6-hydroxynorleucine (1) or (S)-allysine ethylene acetal (2), and D-phenylalanine (3) as key
intermediates. The latter is accessible via a dynamic
kinetic resolution using D-hydantoinase and D-N-carbamoylase technology. An elegant route to (S)-6-hydroxynorleucine from the racemate has been developed that
makes use of D-amino acid oxidase to oxidize the
(R)-enantiomer to the corresponding 2-keto-6-hydroxyhexanoic acid, which undergoes reductive amination by
glutamate dehydrogenase with cofactor regeneration by
glucose dehydrogenase. Routes to (2) include aminopeptidase resolution of the amino acid amide, but asymmetric
synthesis by reductive amination of the corresponding
ketoacid acetal by phenylalanine dehydrogenase with
formate dehydrogenase to regenerate NADH has been
implemented [5]. Even the closure of the rings to the
thiazepine (5) has been investigated. L-lysine-e-aminotransferase is able to catalyze the conversion of the
(S)-homocystein-(S)-lysine dipeptide (4) to the corresponding aldehyde, which closes to the ring under acidic
conditions, and the third chiral center is thereby configured correctly.
The Omapatrilat case illustrates the use of enzymes and
multi-enzyme systems to address both chirality and sensitive functional groups in the synthesis of a complex API.
In the following text, we discuss other examples and
novel technological means to enable the use of enzymes
in a wider range of applications in the pharmaceutical
area.

Biocatalytic routes to HMG-CoA reductase

inhibitors
Another prominent example of the versatility of biocatalysis for manufacturing a complex pharmaceutical
ingredient is the stereoselective synthesis of the 3,5dihydroxyhexanoate side chain of 3-hydroxy-3-methylglutaryl CoA (HMG-CoA) reductase inhibitors (statins).
These cholesterol-lowering drugs represent a multibillion-dollar business, and the market demand has led to
the development of numerous routes to synthesize the
side chain of statins. The need for high stereochemical
quality at both carbon centers of the molecule (> 99.5%
enantiomeric excess (ee), > 99% diastereomeric excess
(de)) has stimulated the development of biocatalysis
research, and at least six different routes have been
implemented that involve a biocatalytic step [6].
Figure 2 illustrates the chemistry basis of these routes,
which provide access to the desired 3,5-dihydroxyhexwww.sciencedirect.com

Advances in biocatalytic synthesis of pharmaceutical intermediates Panke and Wubbolts 189

Figure 1

NH2
S

OH

OH

HO
O

O
N
H

HS

O
O

COOH

NH2

O
Omapatrilat

NH2
3

OH
O
1. D-amino acid oxidase
Catalase
2. Glutamate DH, NH3
Glucose DH, glucose

NH2
OH

NH2
OH
99% ee

HO

HO

O
O

Phenylalanine DH
Formate DH, formate

NH2

OH

OH

98% ee

O
SH

H2N

1. Lysine--AT, KG
Glutamate oxidase
2. Acid ring closure

H
N

H
5

70% yield

RHN
RHN
O

COOH

COOH

Current Opinion in Chemical Biology

Several biocatalytic steps that have been developed by BMS for the synthesis of Omapatrilat. aKG, a-ketoglutarate; AT, aminotransferase;
DH, dehydrogenase.

anoate either via consecutive steps that make use of chiral

induction (routes a to d) and those that introduce the
chirality of both positions in one concerted step (routes
e to g).

N3 or CN to reopen the oxirane [7]. The latter nucleophile results in (S)-4-cyano-3-hydroxybutanoate ethyl
ester, which is a more advanced intermediate towards
Atorvastatin (Figure 2b) [8].

An initial route, which was developed by Kaneka, is based

on enzymatic reduction of 4-chloro-3-ketobutanoate
esters to (S)-4-chloro-3-hydroxybutanoate (> 99% ee)
using Candida magnoliae ketone reductase in combination
with Bacillus megaterium glucose dehydrogenase to regenerate the NADPH (Figure 2a). As a first improvement,
the kinetics of both these enzymes have been enhanced
with gene shuffling by Codexis. In addition, a haloalcohol
dehalogenase step was added that forms the epoxide from
the (S)-4-chloro-3-hydroxybutanoate ethyl ester, and the
same enzyme can accept alternative nucleophiles such as

A more elegant route to (S)-4-cyano-3-hydroxybutanoic

acid is the asymmetric hydrolysis of meso compound
hydroxyglutaronitrile (Figure 2c), which is easily made
from epichlorohydrin. The nitrilases that were investigated previously to carry out this reaction were inappropriate in terms of productivity and selectivity, but
Diversa was able to find both proR and proS selective
enzymes in an environmental library screen of over 200
new nitrilases [9]. Subsequent evolution of the nitrilases
to tolerate higher substrate concentrations increased the
versatility dramatically, resulting in a process that gives a

www.sciencedirect.com

Current Opinion in Chemical Biology 2005, 9:188194

190 Biocatalysis and biotransformation

Figure 2

F
OH

OH

O
OH
OH

OH

O
6

R
OR'

Atorvastatin
NHPh

(a)

rec. C. magnolia ADH

GDH, glucose

OH

Cl

Cl

99% ee

OEt
O

(b)

OEt
1. ADH
2. Dehalogenase, HCN

OH

Cl

OEt
Nitrilase

OH
NC

96% ee

NC

OEt

(c)

OH

CN

O
96% ee

NC
OH

H3CO(O)C
O

(d)

1. -Chymotrypsin
2. PLE

OH

O
98% ee

HO

(e)

HO

OEt

A. calcoaceticus ADH
FDH, formate

BnO

OH

O
97% ee
OEt

DERA

OH

Cl
(g)

OH
BnO

OEt
(f)

OEt

OH

Cl
O

DERA S238D

OH

OH

O
[]D=72
(35% yield)

+
N3

97% de

N3
Current Opinion in Chemical Biology

Several published routes towards (ad) early and (eg) advanced intermediates towards HMGCoA reductase inhibitors (statins), such as
atorvastatin [6]. GDH, B. megaterium glucose dehydrogenase. ADH, alcohol dehydrogenase; FDH, formate dehydrogenase; PLE, pig liver
esterase; rec, recombinant.

96% yield and 98.5% ee at 3 M substrate concentration

[10]. Another elegant route that starts from a prochiral
compound and that utilizes cheaply accessible hydrolases
has been developed by Ciba. Here, 3-methoxycarbonyloxy-pentanedioic acid diethyl ester is hydrolysed by
Current Opinion in Chemical Biology 2005, 9:188194

a-chymotrypsin, a common protease, to (R)-3-methoxycarbonyloxy-pentanedioic acid monoethyl ester at 98% ee

and 94% yield [11]. The methoxycarbonyl group is
subsequently removed by pig liver esterase, without
hydrolysis of the ethyl ester (Figure 2d).
www.sciencedirect.com

Advances in biocatalytic synthesis of pharmaceutical intermediates Panke and Wubbolts 191

A disadvantage of all routes described above is that

subsequent chemical or biocatalytic steps are needed
to obtain the (3R,5S)-dihydroxyhexanoate statin intermediate. Ideally, one would like to obtain both chiral
centers in a single step. Ketoreductases that reside in
Acinetobacter calcoaceticus were described almost a decade
ago by the group of Patel et al., which carried out the
double reduction of ethyl 6-benzyloxy-3,5-dioxohexanoate to the (3R,5S)-diol at 85% yield and 97% ee (Figure 2e)
[6].
Recently, a deoxyribose-5-phosphate aldolase (DERA)
process for the (3R,5S)-dihydroxy-hexanoate statin intermediate was described (Figure 2f). It is based on the early
observations by Wong et al. that DERA is, at high enzyme
loadings, able to catalyse, as a low yielding side reaction,
the aldol condensation of chloroacetaldehyde with two
molecules of acetaldehyde. The resulting (3R,5S)-6chloro-3,5-dihydroxyhexanal is stabilized as a hemiacetal, and by optimising process conditions, a high
yielding process was developed at DSM [6,12]. Independently, a similar process employing alternative enzymes to the originally used DERA enzyme from Escherichia
coli was developed by the Wong group and Diversa
[13,14]. Lastly, the DERA enzyme was improved by
a Ser238Asp mutation, to accept azidoproprionaldehyde
as an alternative to chloroacetaldehyde, enabling the
synthesis route to 7-azido-(3R,5S)-dihydroxy-heptanal

as an advanced intermediate to Atorvastatin [14,15]

(Figure 2g).

Improvements in pyruvate decarboxylase

for ephedrine production
In the (R)-phenylacetycarbinol (R-PAC) process, a thiamine diphosphate-dependent pyruvate decarboxylase
(PDC) activates acetaldehyde, which is carboligated to
externally provided benzaldehyde (Figure 3). This classical biotransformation process suffers from several drawbacks: it relies on living yeast cells to provide pyruvate,
and the viability of the cells and therefore the pyruvate
supply is decreased by the added benzaldehyde.
Finally, the process leads to prominent by-products,
mostly benzyl alcohol, formed by different reducing
systems in the yeast [16].
An alternative process is based on isolated or enriched
PDC, which in the absence of living cells allows increased
supply of benzaldehyde even as a second phase (solubility
in aqueous buffer is 40 mM). Using an emulsion of
benzaldehyde in aqueous buffer with PDC from Rhizopus
javanicus, product concentrations of up to ca. 50 g/l were
initially obtained, where living yeast-based processes
yield 2030 g/l of R-PAC. The addition of a water-immiscible organic solvent led to a decrease in substrate and
product concentrations in the aqueous phase, leading to
reduced aldehyde-inactivation. Indeed, a final product

Figure 3

OH

OH

7
HN

()-Ephedrine
O

(a)

Yeast (PDC)
CO2

N+
HO

HO

OH

H+

S
O

OH
10

8
(b)
O

Z. mobilis
PDC W392M

O
OH

8
O
Current Opinion in Chemical Biology

Carboligation of pyruvate or acetaldehyde and benzaldehyde to R-PAC (7) by (a) thiamine diphosphate-dependent pyruvate decarboxylases.
An activated acetaldehyde-cofactor intermediate (8) is formed that can give rise to by-products acetaldehyde (9) and acetoin (10). (b) Z. mobilis
PDC W392M is more specific and accepts acetaldehyde.
www.sciencedirect.com

Current Opinion in Chemical Biology 2005, 9:188194

192 Biocatalysis and biotransformation

concentration of 100 g/l in the organic phase was achieved

with Candida utiliz PDC when using octanol or nonanol as
a solvent [17]. An improved octanol/water system allowed
even higher product concentrations of 167 g/l in the
organic phase and 93% yield on benzaldehyde. As the
kinetic parameters of the deactivation process have been
elucidated recently [18], the stage is set for a further
process improvement of this already remarkably productive process.
The yeast PDC system has three significant drawbacks:
first, pyruvate is rather expensive (although recent
advances in metabolic engineering might provide a solution [19,20]); second, the decarboxylation of pyruvate
leads to acidification of the medium, gas formation and
potentially to undesired liquid/aqueous interfaces in an

industrial process; finally, the decarboxylation is faster

than carboligation, leading to an accumulation of acetaldehyde in the mixture and inactivation of the enzyme.
These points have been addressed by switching from
yeast-derived PDCs to bacterial PDCs. The latter
enzymes are capable of using acetaldehyde rather than
pyruvate and thus prevent the first two problems. In
particular, the PDC of Zymomonas mobilis has proven
useful, although the enzyme is significantly less active
than the bakers yeast enzyme and also more sensitive to
acetaldehyde. Mutagenesis yielded variants that were
improved in stability and carboligation activity, and for
Z. mobilis mutant enzyme PDCW392M a two-liquid
phase approach was investigated. The highest R-PAC
concentrations up to 17 g/l were obtained in the
presence of alcohols as cosolvent and acetaldehyde as a

Figure 4

CO2Et

HO

AcHN

COOH

HO
NH2

OH

GS4104

OH

(a)

O
HO
HO

(b)

COOH

HO
rec. E. coli

OH
OH

HO
OH

OH
O
HO
HO

(c)

OH
OH

rec. A. mediterranei

HO

B. pumilus / rec. E. coli

HO

COOH

NH2

OH
O
HO
HO

(d)

COOH
rec. E. coli

OH
OH

HO
OH

OH
O
HO
HO

OH
OH

COOH
rec. E. coli
OH
OH
Current Opinion in Chemical Biology

Engineering the central metabolism and shikimic acid pathway of E. coli and other organisms to allow the accumulation of (a) shikimic acid and
(b) aminoshikimic acid as potential starting material for Tamiflu. (c,d) The trans-diols 3,4- and 2,3-trans-CHD are derived from the shikimate or
chorismate pathway as well.
Current Opinion in Chemical Biology 2005, 9:188194

www.sciencedirect.com

Advances in biocatalytic synthesis of pharmaceutical intermediates Panke and Wubbolts 193

substrate [18]. Despite the lower productivity and yield,

the advantage of using acetaldehyde as a cheaply
available substrate is substantial, and considerable progress in process development is to be expected in the near
future. To this end, a high-throughput compatible chromogenic assay to detect R-PAC by reduction of 2,3,5triphenyltetrazolium has become available that enables
directed evolution of the Z. mobilis enzyme to address the
shortcomings described [21].

Fermentative production of (non-natural)

intermediates
Rather than using single enzyme reactions on a chemically produced intermediate, such pharmaceutically relevant intermediates could, in selected cases, be made by
fermentation. A very prominent example here is the
production of shikimic acid as an intermediate for
Roches neuraminidase inhibitor Tamiflu1 (Roche) with
engineered E. coli strains (Figure 4). Shikimic acid is a
regular intermediate in aromatic amino acid synthesis,
which makes fermentative production a very appealing
option. Several strategies have been tried and have led to
very promising titers up to 84 g/l achieved by Frost et al.
(Figure 4a), and the different strategies have recently
been reviewed [22].
Recently, the Frost group has been able to produce
amino-shikimic acid, a potential more-downstream intermediate to Tamiflu1 and other advanced chiral structures, by engineering Amycolatopsis mediteranei or by
combining recombinant E. coli in a fermentation with
Bacillus pumilus (Figure 4b) [23]. The same group also
published a revolutionary new entry point to the shikimic
acid pathway, thereby circumventing the limitations in
phosphoenolpyruvate flux towards D-arabino-heptulosonate-7-phosphate (DAHP) synthase caused by the phosphotransferase system. To this end, KDPG aldolase was
improved by directed evolution to more efficiently condensate pyruvate with erythrose-4-phosphate, thereby
creating a by-pass route to DAHP [24].
Utilizing an engineered shikimic acid pathway, two promising chiral scaffolds for diversity oriented synthesis of
new active compounds have been derived from E. colis
metabolism. Thus, the non-natural compounds (S,S)-2,3dihydroxy-2,3-dihydrobenzoic acid (2,3-trans-CHD,
Figure 4d) and (R,R)-3,4-dihydroxy-3,4-dihydrobenzoic
acid (3,4-trans-CHD, Figure 4c), have been made available in preparative amounts by manipulating the chorismate metabolism in E. coli. For 2,3-trans-CHD, it was
enough to overexpress the entB and C genes, converting
chorismate to 2,3-trans-CHD, and block subsequent steps
that allowed accumulation of 2,3-trans-CHD to 4.6 g/l
[25] (Figure 4). For 3,4-trans-CHD, five genes involved
in directing chorismate into different pathways were
deleted and entB was overexpressed to obtain 0.8 g/l
[26]. These building blocks are now available on the
www.sciencedirect.com

kilogram scale and have been used, for example, for

the synthesis of the plant growth inhibitor streptol and
carbohydrate mimetics.
Another example of using metabolic engineering to make
non-naturally occurring fermentation products is the use
of E. coli for fermentative production of unnatural L-aamino acids, where the strain was deregulated to produce
sufficient amounts of O-acetylserine for the O-acetylserine-sulfhydrylase, and the unnatural part of the product
was fed externally or at a later stage. The accessible
products included an intermediate for an HIV-protease
inhibitor [27].
The quoted examples illustrate the importance of E. coli
as an organism for pathway design, which goes beyond
that of merely serving as the model system for metabolic
engineering it used to be.

Conclusions
Biocatalysis has more to offer than providing a solution
when the chemistry doesnt work. The examples given in
this review go beyond more established enzyme classes
such as the hydrolases and ketone reductases, and illustrate the enormous potential of enzymatic catalysis in
single-step reactions, combined enzyme reactions and
metabolic engineering.

References and recommended reading

Papers of particular interest, published within the annual period of
review, have been highlighted as:
 of special interest
 of outstanding interest
Breuer M, Ditrich K, Habicher T, Hauer B, Kesseler M, Stu rmer R,
Zelinski T: Industrial methods for the production of optically
active intermediates. Angew Chem Int Ed Engl 2004, 43:788-824.
Comprehensive review on industrial manufacturing of chiral intermediates
using classical resolution approaches, biocatalysis as well as chemical
catalysis.

1.


2.

Panke S, Held M, Wubbolts MG: Trends and innovations in

industrial biocatalysis for the production of fine chemicals.
Curr Opin Biotechnol 2004, 15:272-279.

3.

Shaw NM, Robbins KT, Kiener A: Biocatalytic approaches for

the large-scale production of asymmetric synthons. In
Asymmetric Catalysis on Industrial Scale. Edited by Blaser HU,
Schmidt ES. Wiley-VCH; 2004:105-115.

4.

Yasbeck DR: Challenges in the development of an efficient

enzymatic process in the pharmaceutical industry.
Tetrah Asymmetry 2004, 15:2757-2763.

5.

Patel R, Hanson R, Goswami A, Nanduri V, Banerjee A,

Donovan M-J, Goldberg S, Johnston R, Brzozowski D, Tully T
et al.: Enzymatic synthesis of chiral intermediates for
pharmaceuticals. J Ind Microbiol Biotechnol 2003,
30:252-259.

Mu ller M: Chemoenzymatic synthesis of building blocks

for statin side chains. Angew Chem Int Ed Engl 2005,
44:362-365.
Excellent review on chemoenzymatic routes towards HMGCoA reductase inhibitors.

6.


7.

Lutje Spelberg JH, Tang L, Kellogg RM, Janssen DB: Enzymatic

dynamic kinetic resolution of epihalohydrins. Tetrahedron
Asymmetry 2004, 15:1095-1102.
Current Opinion in Chemical Biology 2005, 9:188194

194 Biocatalysis and biotransformation

8.

Davis SC, Grate JH, Gray DR, Gruber JM, Huisman GW, Ma SK,
Newman LM, Sheldon R, Wang LA: Halohydrin dehalogenases
and method for production of 4-cyano-3-hydroxybutyric acid
esters and amides. WO 2004/015132, (Codexis Inc.).

9.

Robertson DE, Chaplin JA, DeSantis G, Podar M, Madden M, Chi

E, Richardson T, Milan A, Miller M, Weiner DP et al.: Exploring
nitrilase sequence space for enantioselective catalysis. Appl
Environ Microbiol 2004, 70:2429-2436.

10. Desantis G, Wong K, Farwell B, Chatman K, Zhu Z, Tomlinson G,

 Huang H, Tan X, Bibbs L, Chen P et al.: Creation of a productive,
highly enantioselective nitrilase through gene site saturation
mutagenesis (GSSM). J Am Chem Soc 2003, 125:11476-11477.
Illustrating the power of directed evolution to adapt enzymes to conditions that are required in industrial processes (high selectivity, productivity and concentrations).
11. O hrlein R, Baisch G: Chemo-enzymatic approach to statin side
chain building blocks. Adv Synth Catal 2003, 345:713-715.
Deracemization with cheaply available enzymes to catalyze the synthesis
of statin side chains.
12. Kierkels JGT, Mink D, Panke S, Lommen FAM, Heemskerk D:
Process for the preparation of 2,4-dideoxyhexoses and
therapeutic uses thereof. WO 2003/006656, (DSM).
13. Greenberg WA, Varvak A, Hanson SR, Wong K, Huang H, Chen P,
Burk MJ: Development of an efficient, scalable, aldolase
catalyzed process for enantioselective synthesis of statin
intermediates. Proc Natl Acad Sci USA 2004, 101:5788-5793.
14. DeSantis G, Liu J, Clark DP, Heine A, Wilson IA, Wong C-H:
 Structure-based mutagenesis approaches towards
expanding the substrate specificity of D-2-deoxyribose-5phosphate aldolase. Bioorg Med Chem 2003, 11:43-52.
Use of high-resolution 3D structure information to construct mutants of an
enzyme that can accommodate a new substrate. Potential new route to
atorvastatin (see also [15]).
15. Liu J, Hsu C-C, Wong C-H: Sequential aldol condensation
catalyzed by DERA mutant Ser238Asp and a formal total
synthesis of atorvastatin. Tetrahedron Lett 2004, 45:2439-2441.
16. Mandwal AK, Tripathi CKM, Trivedi PD, Joshi AK, Agarwal SC,
Bihari V: Production of L-phenylacetyl carbinol by immobilized
cells of Saccharomyces cerevisiae. Biotechnol Lett 2004,
26:217-221.
17. Rosche B, Sandford V, Breuer M, Hauer B, Rogers PL: Enhanced
production of R-phenylacetylcarbinol (R-PAC) through
enzymatic biotransformation. J Mol Catal B Enzym 2002,
19-20:109-115.

Current Opinion in Chemical Biology 2005, 9:188194

18. Rosche B, Breuer M, Hauer B, Rogers PL: Biphasic aqueous/


organic biotransformation of acetaldehyde and benzaldehyde
by Zymomonas mobilis pyruvate decarboxylase.
Biotechnol Bioeng 2004, 86:788-794.
Excellent example of how even a well-established process can be
improved by a change of enzyme system and process development.
19. Zelic B, Gostovic S, Vuorilehto K, Vasic-Racki D, Takors R:
Process strategies to enhance pyruvate production with
recombinant Escherichia coli: from repetitive fed-batch to in
situ product recovery with fully integrated electrodialysis.
Biotechnol Bioeng 2004, 85:638-646.
20. Causey TB, Shanmugam KT, Yomano LP, Ingram LO:
Engineering Escherichia coli for efficent conversion of
glucose to pyruvate. Proc Natl Acad Sci USA 2004,
101:2235-2240.
21. Breuer M, Pohl M, Hauer B, Lingen B: High-throughput assay
of (R)-phenylacetylcarbinol synthesized by pyruvate
decarboxylase. Anal Bioanal Chem 2002, 374:1069-1073.
22. Kra mer M, Bongaerts J, Bovenberg R, Kremer S, Mu ller U, Orf S,
Wubbolts M, Raeven L: Metabolic engineering for microbial
production of shikimic acid. Metab Eng 2003, 5:277-283.
23. Guo J, Frost JW: Synthesis of aminoshikimic acid. Org Lett
 2004, 6:1585-1588.
From solving the enigma on the biosynthesis of aminoshikimic acid as
present in natural products towards a means to produce the molecule.
24. Ran N, Draths K, Frost JW: Creation of a shikimate pathway
 variant. J Am Chem Soc 2004, 126:6856-6857.
First example of engineering a new link between central carbon metabolism and the shikimic pathway, aimed at solving limitations in PEP flux
due to the PTS sugar transport system.
25. Franke D, Lorbach V, Esser S, Dose C, Sprenger GA, Halfar M,
Tho mmes J, Mu ller R, Takors R, Mu ller M: (S,S)-2,3-dihydroxy2,3-dihydrobenzoic acid: microbial access with engineered
cells of Escherichia coli and application as starting material in
natural product synthesis. Chem Eur J 2003, 9:4188-4196.
26. Franke D, Sprenger GA, Mu ller M: Easy access to (R,R)-3,4dihydroxy-3,4-dihydrobenzoic acid with engineered strains of
Escherichia coli. ChemBioChem 2003, 4:775-777.
27. Maier THP: Semisynthetic production of unnatural L-a-amino
 acids by metabolic engineering of the cysteine-biosynthetic
pathway. Nat Biotechnol 2003, 21:422-427.
Both by metabolically engineering the cysteine pathway of E. coli and by
feeding precursors to the fermentation process, new routes to intermediates of potential pharmaceutical interest were obtained.

www.sciencedirect.com