See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/226888918

Skin derivatives in vertebrate ontogeny and

phylogeny

Article in Biology Bulletin January 2009

DOI: 10.1134/S1062359009020101

CITATIONS READS

6 256

1 author:

Olga Fedorovna Chernova

Severtsov Institute of Ecology and Evolution
151 PUBLICATIONS 190 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Microstructure of mammal hairs View project

Evolutionary aspects of polymorphism of the Vertebrate skin derivatives View project

All content following this page was uploaded by Olga Fedorovna Chernova on 22 December 2014.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
ISSN 1062-3590, Biology Bulletin, 2009, Vol. 36, No. 2, pp. 175183. Pleiades Publishing, Inc., 2009.
Original Russian Text O.F. Chernova, 2009, published in Izvestiya Akademii Nauk, Seriya Biologicheskaya, 2009, No. 2, pp. 218227.

CONFERENCE
PROCEEDINGS

Skin Derivatives in Vertebrate Ontogeny and Phylogeny

O. F. Chernova
Severtsov Institute of Ecology and Evolution, Russian Academy of Sciences,
Leninskii pr. 33, Moscow, 109117 Russia
e-mail: chernova@sevin.ru
Received June 24, 2008

AbstractThe skin of vertebrates has numerous and diverse derivatives, either located within the epithelial sheet
itself (glands) or extending above its surface (teeth, scales, feathers, hairs, etc.). Many of them have a modular struc-
ture and constitute structuralfunctional units. Ontogenetically, all skin derivatives are of ectomesodermal origin, and
their morphogenesis is subject to metabolic control, heterochronies (divergence in the timing of origination and
development), and regulation by means of tissue interactions and molecular signaling via similar pathways. The
diversification (origination of morphological diversity) of skin derivatives within the same morphological type is
explained by the development of new generations of ectomesodermal structures separated by heterochronies and
regulated by changes in the gradients of molecular signaling pathways under the influence of environmental factors.
Evolutionary relationships between the majority of skin derivatives are obscure, except for teeth and glands asso-
ciated with sensory organs that have evolved together with these organs. Apparently, many vertebrate skin derivatives
(scales, feathers, hairs, and glands) originated as novelties at nodal stages of phylogeny and subsequently evolved
convergently or in parallel.
DOI: 10.1134/S1062359009020101

INTRODUCTION Attempts to discuss the problem with regard to compar-

The problem of origination and diversification of
skin derivatives in vertebrates is still a challenge to
zoologists. The main difficulty in solving it is that, in
contrast to the situation with other organ systems (mus-
culoskeletal, nervous, excretory, etc.), it is hardly pos-
sible to trace evolutionary relationships between the
derivatives characteristic of representatives of different
classes. Moreover, numerous cases of parallelism and
convergence make identification of the general trend of
their evolution even more difficult. The discussion of
the problem is accompanied by the advancement of
numerous hypotheses and conjectures that cannot yet
clarify the evolutionary scenario concerning the origi-
nation, morphogenesis, diversification, and homology
or historical divergence of these structures. This is due
to the scarcity of relevant paleontological, embryologi-
cal, and comparative morphological data. Views on
evolutionary transformations of skin derivatives based
on their morphology are largely contradictory, since
they are not supported by factual data on the initial and
transitional forms of these derivatives, induction of
their morphogenesis, differences in the biochemical
composition of tentative homologous structures, etc.
Many of them have the typical drawback noted by
Matveev (1932): they are based on the analysis of mor-
phological structures in a direct consecutive series of
extant forms, but this approach does not allow their
evolutionary relationships to be revealed, since com-
parisons are made not between ancestors and descen-
dants but rather between definitive forms representing
terminals of lateral branches of the phylogenetic tree.
ative morphological, biochemical, molecular biologi-
cal, and tissue engineering data show that such a com-
prehensive approach holds much promise for solving it
(Chuong and Homberger, 2003; Wu et al., 2004; Cher-
nova, 2006, 2008a, 2008b; Fuchs, 2007).
This article deals with only a few aspects of the prob-
lem, with attention being paid to (1) skin derivatives as
modular structures, (2) ways to resolve antagonism
between the glandular and supporting functions of the
skin, (3) morphogenesis of skin derivatives and hetero-
chronies, (4) convergence and parallelism as trends in the
evolution of skin derivatives, (5) hypotheses concerning
evolutionary relationships between the skin derivatives,
and (6) substantiation of the generative concept of the
origination and diversification of skin derivatives. The
analysis is based on the most recent results of research in
comparative anatomy, developmental biology, molecular
biology, and tissue engineering, as well as on our generali-
zations of empirical data and original long-term studies
on the morphology of feathers, hairs, and glands in more
than 100 bird species and 400 mammal species (Sokolov
et al., 1988a, 1994; Sokolov and Chernova, 1994a,
1994b, 2001; Chernova, 1996, 2006, 2008a, 2008b;
Chernova and Tselikova, 2004; Shabadash and Cher-
nova, 2006; etc.).
The integument epithelium of chordates underwent
major transformations in the course of phylogeny,
changing from a monolayer sheet of uniform nonkerati-
nized cells capable of secretion (in Acrania and Tunicata)
to a multilayer stratified epithelium, either nonkerati-
nized (in cyclostomes and fishes, with a few exceptions)

175
176 CHERNOVA
or keratinized partially (in amphibians) or completely (in
amniotes). The skin is an exclusive evolutionary acquire-
ment of vertebrates, and it should be discriminated from
the integuments of other animals. The skin consists of
two tissues differing in origin, namely, the epithelial and
connective tissue layers separated by a basal membrane.
The surface epidermis is of ectodermal origin, while the
underlying derma is a mesodermal derivative developing
on account of the somite dermatome. These layers exert
reciprocal influences on each other when forming differ-
ent derivatives during ontogeny. The emergence of verte-
brates on land, the main historical event in their phylog-
eny, was reflected in the development of two basic types
of the cover epithelium, essentially aquatic and terres-
trial, with the transitional amphibiontic or ichthyoid type
(Chernova, 2008b).
Skin derivatives are numerous and diverse. They can
be divided into appendages, or structures extending
above the skin surface, and glands, which lie within the
skin. However, this division is obviously tentative. For
example, hair and feather follicles lie in the dermal or
even hypodermal layer, with only the keratinous shaft
of these structures protruding outward. Glands lie
within the skin but often have accompanying skin or
keratinous protrusions. The same applies to keratiniza-
tion. Scales, feathers, and hairs are keratinous struc-
tures, whereas glands are not keratinized. However, the
follicles of feathers and hairs are not subject to kerati-
nization and remain capable of regeneration, while the
ducts of many glands are keratinized.
The presence of glands is a characteristic feature of
the skin of vertebrates and chordates in general.
Although the skin of reptiles and birds is traditionally
considered nonglandular, it does contain specialized
structures analogous to glands. Skin glands are capable
of production and release of secretions synthesized by
glandulocytes (glandular cells). They are convention-
ally separated into common glands (CGs) and specific
glands (SGs), which differ in numbers, topography,
structure, the composition and amount of secretion, and
biological significance.
Some trends in the evolutionary development of
integuments can be noted in the phylogeny of verte-
brates. They are as follows: (1) gradual thickening of
the cover epithelium and its differentiation into special
cells, layers, and structures; (2) maintenance of similar
structure and omnipotence of the germinative epider-
mal layer, which gives rise to all other skin structures,
including glands; (3) similar differentiation of glandu-
lar structures into CGs and SGs in all vertebrate classes;
(4) gradual improvement of morphophysiological coor-
dination between the structure of the secretory gland
region and the mode of its secretory activity; (5) multi-
functionality of the skin cover and its glandular appara-
tus, with their different functions being in a dynamical
equilibrium and offering examples of antagonism, sub-
stitution, or mutual complementation; (6) functional
analogy between the glands of different vertebrates,
which is due to their various morphological modifica-
tions; (7) specific features of skin morphology and
functions acquired as adaptations to environmental
conditions; and (8) the absence of evolutionary succes-
sion between skin glands that have evolved in different
classes of Vertebrates.
Skin derivatives as modular structures. Skin deriva-
tives may be considered in terms of methodological
approaches aimed at systematizing data on the morpho-
logical diversity of structures in a complex organ, such
as structural histology (Savostiyanov, 2005), the con-
cept of structuralfunctional units (Khrushchev and
Brodsky, 1961; Terskikh et al., 2003), or a theory of
modular structure of the organisms and their organs
(Marfenin, 1999; Notov, 1999; Savostyanov, 2005).
Feathers, nails, hairs, horns, as well as fish scales and
animal epithelial tissues are classified as modular
objects. The reason for identifying structuralfunc-
tional units within an intricately functioning organ is
that such an organ is always heterogeneous and the
properties of its compartments may vary widely (but
within limits of the norm), whereas the properties of its
structuralfunctional unit are much more stable and
different variations do not affect the principles of its
organization (unless in pathological cases). For exam-
ple, we consider that a structuralfunctional unit of a
hepatoid SG is its individual lobe with a system of
secretory ducts, accessory structures (the hair, hair fun-
nel, and cistern in which the secretion accumulates and
matures), and associated apocrine and sebaceous
glands (together with the hepatoid SG, they form com-
bined lobes and ducts). This complex provides for the
synthesis, maturation, and release of the complete
hepatoid gland secretion, supplementing it with addi-
tional protein or glycoprotein, lipids, and melanin.
The hair follicle complex is also an example of
structural-functional unit. In addition to the follicle
itself, it includes the connective tissue envelope, dermal
papilla, glands, muscles, vessels, nerves, epidermal hair
funnel, and hair shaft.
Glands also demonstrate the modular type of orga-
nization, which is known to pertain to systems with a
modular structure, sustainable growth, and cyclic mor-
phogenesis. Glandular fields may be classified as dis-
crete modular objects, since they consist of CGs
equivalent, interchangeable elements showing rela-
tively loose connections with one another. On the other
hand, SGs consisting of indispensable, strongly inter-
connected elements and firmly integrated compact
organs are examples of unitary objects.
Ways to resolve antagonism between the glandular
and supporting functions of the skin. Skin keratiniza-
tion accompanied by degeneration of epithelial cells
(keratinocytes and corneocytes) and the secretory
capacity of glandulocytes are in antagonism to each
other. This antagonism can be resolved by means of (1)
spatial segregation or (2) heterochronies of anlagen and
their derivatives in the course of development or (3)
BIOLOGY BULLETIN Vol. 36 No. 2 2009
 SKIN DERIVATIVES IN VERTEBRATE ONTOGENY AND PHYLOGENY
combination of both functions at the cell level, i.e., in
the same epithelial cell. In the first case, glands develop
in the areas that are neither keratinized nor covered
with skeletal structures or descend to the derma or
under the skin. Thus, glandulocytes are isolated from
corneocytes (in most glands of mammals). In the sec-
ond case, glands originate and develop prior to the
onset of keratinization. In the third case, epithelial cells
acquire the ability to combine the glandular function
with keratinization (i.e., with the supporting function).
Such a combination is observed, for example, in sebok-
eratinocytes of birds or in certain epidermal layers of
reptiles.
The problem of antagonism between the secretory
and supporting functions can also be considered on the
premise that keratinization is a secretory process
involving the synthesis and secretion of specific protein
(keratin) accompanied by cell degeneration. In such a
case, all integuments are regarded as secretory organs
with different types of secretory activities, which com-
plement rather than oppose each other. The epidermal
corneous layer is actively shed off. In many mammal
species, SGs have epidermal thickenings (callosities)
whose corneous layer is shed off to be accumulated in
the saccular gland cavity. Thus, it becomes an essential
component of the secretions of many glands, obviously
enriching them with additional odoriferous substances.
Morphogenesis of skin derivatives and heterochro-
nies. Self-organization and tissue interactions provide a
basis for morphogenesis of different organs, including
ectodermal organs (Beloussov, 1987; Cherdantsev,
2003; Terskikh et al., 2007). In interactions of the cover
epithelium and mesenchyme, a major role is played by
somatic stem cells and signal proteins inducing their
proliferation (Shh, FGF, Wnt, notch, BMP, etc.), which
are common to all skin derivatives. This role is well
studied for the morphogenesis of feathers (Wu et al.,
2004), hairs (Fuchs, 2007), sebaceous CGs, and mam-
mary SGs (Fuchs, 2007; Terskikh et al., 2007). It has
been shown that minor shifts in the genetic develop-
mental program result in significant structural changes
such as conversion of hair anlagen into CGs in tissue
culture, the growth of hairs in the mouth, or develop-
ment of teeth in chicks.
What are the features of similarity and difference in
the morphogenesis of skin derivatives that can be
revealed by comparative analysis? The features com-
mon to all amniotes are (1) segregation of special skin
fields in which these derivatives arise (pterylae in birds,
vibrissa pads, food pads, glandular fields, etc.); (2) het-
erogenization of initially homogeneous skin fields
through the formation of dermal condensations subja-
cent to placodes (epithelial cell clusters); and (3) actual
morphogenesis of skin derivatives, which proceeds by
different scenarios but always on the basis of epithe-
lialmesenchymal interactions.
Reciprocal epithelialmesenchymal interactions are
of great significance for the development of skin deriv-
BIOLOGY BULLETIN Vol. 36 No. 2 2009
177
atives, with the influence of mesenchyme on the forma-
tion of ectodermal structures being either positive (in
the morphogenesis of skin appendages) or negative (in
gland morphogenesis). For example, Schmalhausen
(1964) refers to the inhibitory effect of mesenchyme on
the development of paired cement SGs in tadpoles. In
the norm, these glands develop on the ventral head sur-
face only in the area where the ectoderm is separated
from the mesenchyme by the anterior gut wall. This
effect is confirmed by the fact that the whole ectoder-
mal explant can transform into a sucker in the absence
of contact with the mesenchyme.
Feathers and scales arise via mesenchyme-induced
evagination of the epithelium and mesenchyme. If hairs
and all glands had arisen only via invagination of the
epithelium and mesenchyme, this would have been one
more perfect example of alternative bifurcation of
developmental programs (Oster and Albrech 1982). It
should be noted that, in the English literature, the term
invagination is applied to the morphogenesis of
appendages such as hairs and glands, which is incorrect
in our opinion, insofar as this term refers to the forma-
tion of an embryonic layer or organ by its infolding into
the subsiding lower layer. In fact, actual epidermal
invagination occurs only in morphogenesis of saccular
SGs, in which the central cavity is formed due to invag-
ination of the epidermis into the subsiding dermal layer,
followed by the growth of glandular parenchyma. For
example, saccular poison and luminescent glands of
fishes, most SGs of reptiles, and various saccular
glands of mammals are formed in this way. Moreover,
these saccular organs do not necessarily have a duct that
opens on the body surface, as in most SGs of amniotes:
there also are glands confined within the dermal layer,
as is the case with saccular photophores of fishes. In
many cases, such saccular SGs are capable of protrud-
ing above the skin surface or even everting, releasing or
ejecting an odorous secretion.
Invagination occurs at later stages of morphogenesis
of the feather follicle: the epidermis invaginates around
the feather anlage to form the sheath of the cylindrical
follicle. Feather morphogenesis involves specific pro-
cesses such as the displacement of the growth zone
from the apical end of skin outgrowth down to the sides
of the follicle (which rarely takes place in the skin),
localization of several stem cell populations, and
branching of the vane via apoptosis of marginal plate
cells rather than via differential cell proliferation
(Lucas and Stettenheim, 1972). This is why feather his-
togenesis is referred to as reverse branching morpho-
genesis, in contrast to simple branching morphogene-
sis characteristic of, e.g., the lungs and mammary
glands (Lucas and Stettenheim, 1972; Wu et al., 2004).
The dorsal and ventral sides of the feather produced by
the intermediate epidermal layer of the cylindrical fol-
licle are not homologous to the corresponding sides of
the scale, which are derived from the upper epidermal
layers. This contradicts the hypothesis concerning the
origin of feathers from scales of the ancestral forms
178 CHERNOVA
(Chernova, 2008a). In the formation of feathers, the
upper epidermal layers produce only a corneous sheath
that is shed when the feather unfolds.
The molecular mechanisms of feather morphogene-
sis have been studied in detail. Different signal proteins
operate on each stage of feather development, regulating
morphogenesis. A detailed model of gene influence on
feather development (the transgenic feather protocol)
has been proposed, which hypothesizes that the diversity
of feathers depends on the orientation and slope of
molecular gradients that condition the topological struc-
ture of the feather epithelium. Thus, the relationship
between the activity of signal proteins and the shape and
novel function of the organ has been revealed (Widelitz
et al., 2003, 2007; Wu et al, 2004; Yue et al., 2006). For
example, the inductive activity of the dermal papilla cre-
ates a gradient of the signal protein Wnt3, which deter-
mines a helical pattern of barb ridges and guides the
symmetrical branching of the vane. The Wnt gradient
accounts for the transformation of bilaterally symmetric
contour feathers into radially symmetric down feathers.
The Noggin protein increases the degree of feather
branching, while BMPs increase the size of the feather
shaft (Yu et al., 2004). The ratio of antagonistic noggin
and BMP4 plays an essential role in feather branching,
determining the number and size of barbs; BMP4 regu-
lates the formation of the feather shaft and the interlock-
ing of barbs; Shh induces apoptosis of the marginal plate
epithelium, which leads to the formation of barbs (Yu et
al., 2002; Chang et al., 2004a, 2004b). In general, the
order of appearance of signal molecules in feather
morphogenesis is as follows: FGF4, BMP4 Shh,
Wnt-7a notch-1, Serrate-l, Delta-1 Msx-1,
Msx-2 Hox, NCAM. Thus, topological transforma-
tions of the feather from a cylindrical corneous to a hier-
archically branching structure are controlled by molecu-
lar mechanisms, which have apparently provided the
possibility of development of different feather forms
both in ontogeny and in phylogeny.
The morphogenesis of CGs proceeds via the
ingrowth of the epithelium into the dermal layer
(Sokolov et al., 1988b), and no invagination takes place.
In Anamnia, the morphogenesis of glandulocytes is a
process of proliferation and differentiation of basal-layer
epithelial cells or of other glandular cells tightly attached
to the basal membrane. In the lamprey, for example, the
sequence of events is as follows: undifferentiated
cells glomerular supporting-glandular cells
protein glands protein secretion. In amniotes, the
CG lumen starts to form in lower secretory regions,
along with an increase in the mass and secretory activity
of glandular parenchyma; consequently, the central cav-
ity and ducts are formed, with the excretory duct opening
on the surface. Therefore, CG structures differentiate in
the outward direction, and this concerns both alveolar
and tubular glands. It should be noted that sebaceous and
hepatoid glands of mammals during the postnatal period
undergo continuous processes such as, the formation of
secondary ducts and cavities due to cell death in the glan-
dular parenchyma and cyclic regeneration of glandular
epithelium on account of stem cells of the epidermis and
hair follicle.
One more pathway of gland morphogenesis, alterna-
tive to the ingrowth of the epidermal anlage into the
derma, is the ingrowth of mesenchyme into the epider-
mis. For example, CG anlagen in salamanders appear at
early stages of larval development as spherical cell
buds in the epidermis. Such a bud does not submerge
into the dermal layer; conversely, the mesenchyme
actively penetrates the epidermis and surrounds the bud,
isolating it from supporting epidermal cells (Sokolov
et al., 1988a). Further differentiation of myocytes and
glandulocytes takes place in this isolated structure, and
only then does the gland sink into the corium. This devel-
opmental scenario principally differs from the above
process of gland formation in the majority of amniotes,
in which the proliferating and differentiating epidermal
bud actively grows into the corium, where its further
morphogenesis takes place. Common to these alternative
processes is early isolation of glandular structures from
supporting cells, which is achieved either with the aid of
mesodermal envelopes or due to the bud submergence
into the corium. Both mechanisms provide for spatial
separation of glandular and supporting integumental
structures.
In mammals, CG morphogenesis is associated with
hair morphogenesis (except for initially independent
eccrine and sebaceous glands): both sebaceous and
apocrine glands are derivatives of the hair follicle. The
hair follicle itself is formed via the ingrowth of the epi-
thelium into the mesenchyme (Sokolov et al., 1988b).
Processes characteristic of different stages of hair mor-
phogenesis are as follows: (1) aggregation, prolifera-
tion, and differentiation of epithelial and mesenchymal
cells in the follicle anlage region; (2) the intimate inter-
action of the epithelium and mesenchyme, which can
even come in direct contact with each other, being not
separated by a basal membrane; hairs fail to form with-
out the involvement of mesenchyme; (3) the appear-
ance of first keratinization centers in the hair anlage
rather than in the epidermis; 4) the growth of the hair
bulb in two directions simultaneously: its further
ingrowth deeper into the derma is accompanied by the
upward movement of differentiating cells; and (5)
simultaneous formation of the anlagen of sebaceous,
apocrine, and hepatoid glands (on account of the epi-
thelium of the follicle anlage), the bulge zone, and then
the hair erector muscle (on account of mesenchymal
cells). The bulge zone is not only the site of muscle
attachment, but it also accommodates a population of
hair follicle stem cells accounting for the formation of
a new hair follicle and participating in regeneration of
the epidermis (Terskikh et al., 2007; Fuchs, 2007). One
more stem cell population is hidden deep in the hair
bulb matrix, above and around the dermal papilla. Sub-
sequent growth and differentiation of the hair follicle
lead to the upward extension of the hair shaft and its
BIOLOGY BULLETIN Vol. 36 No. 2 2009
 SKIN DERIVATIVES IN VERTEBRATE ONTOGENY AND PHYLOGENY
eventual emergence on the surface through a special
cleft in the epidermis, above the hair follicle.
Molecular mechanisms of the morphogenesis of
hairs and their sebaceous glands are known (Fuchs,
2007). Thus, the formation of a placode, its further
growth, and differentiation of hairs into different mor-
phological types are under control of the Wnt/BMPs
signal protein gradient with the involvement of other
regulators such as the Wnt/noggin gradient, Shh, Lef1,
and Lef2. For example, mice devoid of Lef1 have only
guard hairs, while other hair types are absent. All these
processes are subject to mesenchymal induction. For
example, the derma has an effect on the size and density
of follicles, which is confirmed by molecular differ-
ences between dermal cells from different body
regions. The Sonic hedgehog gene (Shh), identified
among genes expressed early in the placode, plays a
key role in condensation of dermal cells into dermal
papillae, which are an essential component of the hair
follicle. The dermal papilla is the source of signal for
the placode ingrowth leading to the formation of the
hair follicle, which subsequently matures under the
effect of further epithelialmesenchymal interactions.
Stem cells in the bulge zone become active during the
anagen phase, when the dermal papilla reaches this
region and apparently induces dormant stem cells.
However, these processes differ from those of embry-
onic hair anlage formation.
Precursor cells of sebaceous glands appear at later
stages of intrauterine development (in mice), but the
glands themselves begin to function only after birth.
The processes of sebocyte differentiation are also sub-
ject to the effect of signal proteins, such as c-myc and
the adipogenic transcriptional factor PPAR-gamma. In
analyzing the periodicity of gland functioning, the
question arises as to the source of new glandulocyte
generations. It has been proved in many cases that they
originate from nondifferentiated somatic stem cells
located in the germinative layer of the epidermis, which
is additional evidence for the potential secretory func-
tion of epithelial cells. In the sebaceous glands of mam-
mals, for example, the glandular parenchyma is
restored due to epidermal stem cell migration along
duct walls to the glandular acini, where these cells dif-
ferentiate into glandulocytes. Stem cells take part in
restoring the functioning of sebaceous glands after the
death of sebocytes: as shown in tissue culture, they
form a layer of nondifferentiated cells that are precur-
sors of sebocytes. Stem cells from the bulge zone also
participate in regeneration of sebaceous glands during
skin wound healing.
Convergence and parallelism in the evolution of skin
derivatives. The evolution of skin derivatives in verte-
brate phylogeny apparently followed parallel and con-
vergent pathways against the background of adaptive
radiation of taxa. It proceeded due to quantitative and
qualitative morphological transformations based on the
capacity of the germinative layer (stem somatic cells) of
BIOLOGY BULLETIN Vol. 36 No. 2 2009
179
the cover epithelium for proliferation and differentiation,
under conditions of unity and antagonism of the support-
ing and glandular functions of the integument.
Analogies among skin derivatives are readily dis-
cernible. For example, many SGs in representatives of
phylogenetically segregated taxa have similar structure,
which is especially true of signal SGs involved in scent
communication or of SGs whose secretions protect sen-
sory organs and mucous membranes. Skin glands are
the derivatives of germ layers and function as heteroge-
neous organs consisting of compartments derivative
from the epithelial, connective, nervous, blood, and
muscle tissues. These integrated SG are present in all
classes of Chordata, except for Acrania.
The feather follicles of birds have no CGs, which is
explained by the need for reducing the weight of integ-
uments; the impossibility of feather lubrication from
the follicular funnel opening; the development of the
uropygial gland specialized for taking care of feathers,
with body aerodynamics remaining unimpaired (due to
a certain position of the gravity center); and the hydro-
phobic properties of feather keratin itself. Nevertheless,
sebocytes of the outer peripheral (sebaceous) part of
the uropygial gland are very similar to those of mam-
malian sebaceous glands in both morphological and
biochemical features. This appears to result from con-
vergence or may be an organotypic trait of sebaceous
glandulocytes.
Historically segregated structures such as hairs and
feathers can also provide an example of convergent
similarity. It is known that mammals and birds repre-
sent two autonomous phylogenetic branches. Among
vertebrates, only they are homoiothermal. The outer
coats in both classes, fur in mammals and plumage in
birds, serve to maintain body homeostasis. Hence,
questions logically arise as to the degree of similarity in
structural adaptations of hair and feather coats provid-
ing for their heat-insulating properties as well as in
morphological structures accounting for performance
and interrelation of their different functions. In addition
to structural similarity of hair and feather coats (strati-
fication, distinct differentiation, regeneration capacity,
production of different generations, cyclicity of growth,
periodic molting, etc.), their main structuralfunctional
units are similar as microanatomical corneous struc-
tures (similar development of air cavities, three-layered
structure, etc.) (Chernova 2005a, 2005b, 2006, 2008a).
Polymorphism of hairs and feathers is due to the exist-
ence of their several generations produced by the folli-
cles during ontogeny, while specificity of their topogra-
phy and structure is accounted for by molecular mech-
anisms and other factors, including developmental
changes in metabolism. From this standpoint, hair and
feather follicles can be defined as dynamic systems
with sufficiently high plasticity (allowing changes in
the morphological type of a hair or feather) that are sub-
ject to age-related and seasonal changes. Hair and
180 CHERNOVA
feather follicles are similar in many respects, but this
similarity is a result of convergence.
Similarity due to convergent evolution, in contrast to
that due to parallelism, is always superficial and con-
cerns only a few structural traits of the convergent
groups (Schmalhausen, 1969); in our case, this is the
architecture of feathers and hairs. The analogy in their
architecture can probably be classified as homoplasy
(the term introduced by Lankester (1870); cited from
Blyakher, 1976), i.e., structural similarity that have
arisen independently but on a common morphological
basis. In our case, such a basis is the epithelialmesen-
chymal interaction. This follows from the unity of both
molecular-genetic mechanisms of developmental regu-
lation and life cycles of hairs and feathers. This phe-
nomenon provides an example of convergent analogy
(Dogel, 1938) or homotypy (Beklemishev, 1964),
i.e., of constructional similarity between the structures
under comparison in the absence of genetic succession
between the corresponding taxa.
Hypotheses concerning evolutionary interrelations
of skin derivatives. As long ago as 1898, Gegenbauer
(see Gabe and Saint-irons, 1965) provided an argument
against the hypothesis of evolutionary connection
between the glands of all three classes of amniotes. He
argued that sebaceous glands of mammals are closely
associated with hair follicles and, consequently, they
are an evolutionary novelty, as is the follicle itself.
Moreover, their association with hair follicles is main-
tained throughout ontogeny. However, Severtsov
(1949) and Matveev (1949), when considering the evo-
lution of hairs, attach no evolutionary significance to
the skin glands and do not regard their origination as an
aromorphosis. In our opinion, it is reasonable to regard
as aromorphosis not only the origination of hairs but
also the formation of the glandhair complex as a struc-
turalfunctional unit, insofar as CGs of mammals orig-
inate from the hair follicle and are closely associated
with this structure.
No direct evolutionary connections between the
glands of vertebrates belonging to different classes can
be traced, and their structure is not subject to morphologi-
cal complication (especially in the case of SGs) even
upon transition from lower Chordata to Anamnia and
then to Amniota. Structural complication of SGs does
not depend on the taxonomic rank of a given group. Even
hagfishes have well-developed mucous SGs with a com-
plex structure. Very intricate, integrated bacterial and
luminescent SGs are widespread among fishes.
Common to all vertebrate taxa (and Chordata in
general) is that they have the germinative layer of epithe-
lial cells, whose proliferation and differentiation account
for the production of all glandulocytes. Quay (1972)
used this fact as a starting point for identifying morpho-
genetic lineages of secretory epidermal cells and formu-
lating the following conclusions. The evolutionary path-
way starts from the central morphogenetic lineage of epi-
dermal stem cells, which subsequently separates into two
large branches of mucous and protein glandulocytes of
Anamnia, with various modifications, and then into
sebaceous and sweat glands of Amniota. The branch of
mucous glands terminates in the class Amphibia
(although in amniotes they are preserved only in mucous
membranes, epithelial cells remain capable of mucous
metaplasia), and the protein gland branch terminates in
the class Reptilia, after giving rise to corneous epidermal
cells. These cells provided a source for the development
of glandular structures in Amniota. The lateral branch of
enamel-secreting protein glandulocytes (ameloblasts
forming the teeth) extends from the ancestors of carti-
laginous fishes to mammals inclusive. In this attempt at
phylogenetic reconstruction, the above author makes a
typical error, since the phylogeny is typically erroneous,
as comparisons are made between recent forms repre-
senting lateral branches of the phylogenetic tree rather
than direct ancestors and descendants.
Other reconstructions, such as attempts to explain
the polymorphism of homotypic holocrine glands in
reptiles, are also unsuccessful, since their authors
ignore the fact that morphological diversification and
complication by no means correspond to phylogenetic
relationships between taxa (Gabe and Saint-Girons,
1965; Sokolov et al., 1994; Chernova, 2008b).
Lactiferous (mammary) SGs of mammals origi-
nated from apocrine CGs of hairs aggregated into a
glandular field (monotremes still have such a field, and
even hair anlagen are preserved in marsupials). Their
initial function (moistening the eggs laid) changed to
that of nourishing the offspring. This was accompanied
by the growth of glandular parenchyma and shaping of
the glandular field into an alveolartubular structure
with or without a nipple (in the latter case, the secretion
trickles down the tufts of hairs). Analogs of lactiferous
glands might be present in cynodonts. Signaling path-
ways that regulate mammary gland anlage formation
and development in mice were identified. In particular,
this concerns the Wnt/-catenin gradient (Veltmaat et al.,
2003): its blocking in transgenic mice resulted in the
absence of mammary glands, hairs, and teeth.
Hypotheses concerning the origin of feathers and
hairs were considered in detail in our previous publica-
tions (Chernova and Tselikova, 2004; Chernova,
2008a). To resolve the widely discussed question of the
evolutionary connection between the reptile scale and
the bird feather, three main approaches has been pro-
posed: (1) the feather originated from the scale that
rolled into a tube and then submerged into the derma,
with consequent formation of a follicle; (2) the feather
evolved through the stage of protofeather, a simple
tubular corneous outgrowth; and (3) the feather follicle
as an evolutionary novelty originated first; in due time,
it produced a featherlike structure that developed by
way of differentiation of a down feather into an asym-
metric quill feather, or vice versa.
Hypotheses concerning the origin of hairs are
numerous, but it appears most likely that the hair folli-
BIOLOGY BULLETIN Vol. 36 No. 2 2009
 SKIN DERIVATIVES IN VERTEBRATE ONTOGENY AND PHYLOGENY
cle originated independently, as a novelty, and that the
formation of vibrissae and hairs in phylogeny followed
parallel pathways, with vibrissae being more ancient.
Since evolutionary connections of skin derivatives
are obscure, some authors chose an alternative
approach and constructed a hierarchical series of struc-
tural complication of these derivatives without consid-
ering their evolutionary connections (Wu et al., 2004).
In this series, asymmetric quill feathers of birds are
regarded as the peak of structural complication. Such
an approach, however, does not solve the problem of
the evolution of skin appendages.
Substantiation of the generative concept of the ori-
gin and diversification of skin derivatives. The idea of
autonomous and independent origin of different skin
derivatives is not new. For example, a scheme of their
independent origin based on ectodermalmesodermal
interactions, proliferation, and/or aggregation of epithe-
lial and mesenchymal cells into the anlage producing
different structures in representatives of different classes
is presented in the book Hymanis Comparative Verte-
brate Anatomy (1979). This idea is supported not only by
morphological data and heterochronies revealed in the
origination of different skin derivatives (in both ontog-
eny and phylogeny) but also by the results of molecular
biological studies. It is in full agreement with the theory
of phylembryogenesis by Severtsov (1949), who con-
nected evolutionary transformations of organs to the ear-
liest stages of embryogenesis.
On the basis of molecular biological studies on the
morphogenesis of skin appendages, a hypothetical
model of their development was proposed, which
shows that all appendages are variations on the same
theme (Chuong and Homberger, 2003), since the
molecular signaling pathways involved in regulation of
their morphogenesis are similar (Chuong et al., 2000;
Wu et al., 2004; Fuchs, 2007). The same is true of
derivatives such as glands (Chernova, 2008b). The
model suggests that de novo formation of derivatives
(appendages and glands) from the cover epithelium and
underlying mesenchyme and their diversification in the
course of evolution took place under the effects of dif-
ferent factors: induction of the epithelial anlage by the
mesenchyme, reciprocal interactions of these tissues
under molecular-genetic control, environmental influ-
ences on molecular processes, and heterochronies of
morphogenesis. Differences in signaling pathways and
ratios of signaling proteins in the period of morphogen-
esis as well as predetermined potencies of the epithe-
lium and mesenchyme account for variations of epithe-
lial organs.
Experimental studies on cellular and molecular
mechanisms of morphogenesis by methods of tissue
engineering have priority in modern science. In particu-
lar, they help in understanding the mechanisms of forma-
tion and evolution of skin derivatives (Matveev, 1932,
1949; Quay, 1972; Chuong and Homberger, 2003; Wu et
al., 2004). The cover epithelium as a tissue structure and
BIOLOGY BULLETIN Vol. 36 No. 2 2009
181
its germinative layer have genetically determined and
potentially different capacities for producing certain
structures within the limits of a given tissue type (epithe-
lial or loose connective tissue).
On the basis of most recent data, Wu et al. (2004)
proposed the concept of morpho-regulation, which
implies that morphogenetic processes can be modu-
lated by morphological regulators that lead to changes
of morphological phenotypes in development and in
evolution. In the evolution of vertebrates, variations in
the number and size of skin derivatives could provide a
basis for the development of a spectrum of different
animal phenotypes (i.e., for adaptive radiation). Such
variations provide the material for selection and, upon
certain environmental changes, may prove to be an
advancement.
The analysis of recent data on the vertebrate skin
derivatives and the results of our long-term studies on
skin glands, hairs, and feathers provide evidence for the
concept of their independent evolutionary origin, which
can be formulated as follows.
In the evolution of vertebrates, different skin deriv-
atives originated at nodal phylogenetic stages as novel
ectomesodermal formations and developed in different
classes independently, in parallel or convergent ways,
depending on specific features of tissue interactions,
metabolic changes, and molecular-genetic regulation of
morphogenesis for the maintenance of homeostasis.
Diversification of structures within the same morpho-
type was due not to their evolution from the simplest to
a more complicated form (or vice versa) but rather to
origination of new generations of ectomesodermal
structures separated by heterochronies and regulated by
changes in the gradients of molecular signaling path-
ways under the influence of environmental factors. Skin
derivatives as tissue structures are homologous to each
other, since they originate from the epithelium and
mesenchyme, but the same derivatives as organs are not
connected by evolutionary succession and have not
been derived from each other. Thus, there is no use in
searching for the pathways of their transformation and
constructing evolutionary schemes of transitions
between these structures, since all of them are novel
formations. Scales, hairs, feathers and glands origi-
nated as independent integumental organs at nodal phy-
logenetic stages owing to multipotency of epithelial
and mesenchymal stem cells (within the limits imposed
by the properties of epithelial and mesenchymal tissues
and their interactions). The independent origin of these
appendages can account for the absence of distinct
intermediate forms between them. Teeth appear to be
an exception, since ameloblasts are homologous in all
vertebrates. The same probably applies to the glands
associated with sensory organs that have evolved
together with these organs (for example, the Harderian
gland in amniotes).
182 CHERNOVA
REFERENCES
Beklemishev, V.N., Osnovy sravnitelnoi anatomii bespoz-
vonochnykh (Foundations of Comparative Anatomy of Inver-
tebrates), vol. 1: Promorfologiya (Promorphology), Mos-
cow: Nauka, 1964.
Beloussov L.V. Biologicheskiy morfogenez (Biological Mor-
phogenesis), Moscow: Mosk. Gos. Univ., 1987.
Blyakher, L.Ya., Problemy morfologii zhivotnykh (Problems
in Animal Morphology), Moscow: Nauka, 1976.
Chang, Ch.-H., Jiang, T.-X., Lin, C.-M., et al., Distinctive
Phenotypes Generated by WNT Member Misexpression dur-
ing Hierarchical Morphogenesis of Dermis, Skin Regions
and Individual Feathers, Mech. Dev., 2004a, vol. 121,
pp. 157171.
Chang, Ch.-H., Yu, M., Wu, P., et al., Sculpting Skin Append-
ages Out of Epidermal Layers via Temporally and Spatially
Regulated Apoptotic Events, J. Invest. Dermatol., 2004b,
vol. 122, pp. 13481355.
Cherdantsev, V.G., Morfogenez i evolyutsiya (Morphogene-
sis and Evolution), Moscow: KMK, 2003.
Chernova, O.F., Glands of Integuments in the Evolution of
Chordata: Morphofunctional Analysis, Extended Abstract of
Doctoral (Biol.) Dissertation, Moscow: Severtsov Inst. Ecol.
Evol., Russ. Acad. Sci., 1996.
Chernova, O.F., Polymorphism of the Architectonics of
Definitive Contour Feathers, Dokl. Akad. Nauk, 2005a,
vol. 405, no. 2, pp. 280285.
Chernova, O.F., One More Example of Morphological Con-
vergence: Similarity between the Architectonics of Feather
and Hair, Dokl. Akad. Nauk, 2005b, vol. 405, no. 3, pp. 425
429.
Chernova, O.F., Evolutionary Aspects of Hair Polymor-
phism, Izv. Akad. Nauk, Ser. Biol., 2006, no. 1, pp. 5262.
Chernova, O.F., The Problem of the Origin of Skin Derivates
in the Evolution of Amniotes. Skin Appendages: Scales,
Feathers, and Hairs, Zh. Obshch. Biol., 2008a, no. 2, pp. 130
151.
Chernova, O.F., Glands of Integuments of Chordates in
Terms of A.N. Severtsovs Morphobiological Theory of Evo-
lution, Tr. Zool. Muz. Mosk. Gos. Univ., 2008b, vol. 49,
pp. 279342.
Chernova, O.F. and Tselikova, T.N., Atlas volos mlekopitay-
ushchikh (Atlas of Mammal Hairs), Moscow: KMK, 2004.
Chuong, Ch.-M. and Homberger, D.G., Development and
Evolution of the Amniotic Integument: Current Landscape
and Future Horizon, J. Exp. Zool. B: Mol. Dev. Evol., 2003,
vol. 298, no. 1, pp. 111.
Chuong, Ch.-M., Chodankar, R., Widelitz, R.B., and Jiang, T.-X.,
Evo-Devo of Feathers and Scales: Building Complex Epithe-
lial Appendages, Curr. Opin. Genet. Dev., 2000, vol. 10,
pp. 449456.
Dogel, V.A., Sravnitelnaya anatomiya bespozvonochnykh
(Comparative Anatomy of Invertebrates), Leningrad: Uch-
pedgiz, 1938.
Fuchs, E., Scratching the Surface of Skin Development,
Nature, 2007, vol. 445, no. 7130, pp. 834842.
Gabe, M. and Saint-Girons, H. Contribution a la morpholo-
gie compare du cloaque et des glandes pidermoides de la
region cloacale chez les Lpidosauriens, Mem. Mus. Nat.
Hist. Paris, Ser. A: Zool., 1965, vol. 33, pp. 15292.
Hymanis Comparative Vertebrate Anatomy, Wake, M.H.,
Ed., Chicago: Univ. of Chicago Press, 1979.
Khrushchov, G.K. and Brodsky, V.Ya., Organ and Cell: Some
Problems in Cytology and Histology, Izv. Akad. Nauk, Ser.
Biol., 1961, vol. 52, no. 2, pp. 181207.
Lucas, A.M. and Stettenheim, P.R., Avian Anatomy: Integu-
ment, in Agriculture Handbook, parts 1 and 2, Washington,
DC: US Dept. Agricult., 1972, no. 362.
Marfenin N.N. The Concept of Modular Organization in Devel-
opment, Zh. Obshch. Biol., 1999, vol. 60, no. 1, pp. 617.
Matveev, B.S., On the Evolution of Vertebrate Integuments
through Embryonic Changes, Zool. Zh., 1932, vol. 24, no. 2,
pp. 3954.
Matveev, B.S., On the Origin of Scaly Cover and Hairs in
Mammals, Zool. Zh., 1949, vol. 28, no. 1, pp. 5970.
Notov, A.A., On the Specificity of Functional Organization
and Individual Development in Modular Objects, Zh.
Obshch. Biol., 1999, vol. 60, no. 1, pp. 6079.
Oster, G. and Albrech, P. Evolution and Bifurcation of Devel-
opmental Programs, Evolution, 1982, vol. 36, pp. 444459.
Quay, W., Integument and the Environment: Glandular Com-
position, Function, and Evolution, Am. Zool., 1972, vol. 12,
no. 1, pp. 95108.
Savostiyanov, G.A., Osnovy strukturnoi gistologii (Founda-
tions of Structural Histology), St. Petersburg: Nauka, 2005.
Severtsov, A.N., Morfologicheskie zakonomernosti evolyutsii
(Morphological Laws of Evolution), Moscow: Akad. Nauk
SSSR, 1949, vol. 5.
Shabadash, S.A. and Chernova, O.F., Gepatoidnye zhelezy
mlekopitayushchikh (Hepatoid Glands of Mammals), Mos-
cow: KMK, 2006.
Schmalhausen, I.I., Proiskhozhdenie nazemnykh pozvonoch-
nykh (The Origin of Terrestrial Vertebrates), Moscow:
Nauka, 1964.
Schmalhausen, I.I., Problemy darvinizma (Problems in Dar-
winism), Leningrad: Nauka, 1969.
Sokolov, V.E. and Chernova, O.F., Kozhnye zhelezy pervich-
novodnykh khordovykh. Obzor issledovaniy (Skin Glands of
Essentially Aquatic Chordates: A Review of Studies), Mos-
cow: Nauka, 1994a.
Sokolov, V.E. and Chernova, O.F., Skin Glands of Birds:
Morphology and Functioning, in Struktura, rost i nekotorye
aspekty gormonalnoi regulyatsii razvitiya spetsificheskikh
zhelez (Structure, Growth, and Some Aspects of Hormonal
Regulation of the Development of Specific Glands), Mos-
cow: Nauka, 1994b.
Sokolov, V.E. and Chernova, O.F., Kozhnye zhelezy mlekopi-
tayushchikh (Skin Glands of Mammals), Moscow: GEOS,
2001.
Sokolov, V.E., Kotova, E.L., and Chernova, O.F., Kozhnye
zhelezy reptilii. Obzor issledovanii (Skin Glands of Reptiles:
A Review of Studies), Moscow: Nauka, 1994.
Sokolov, V.E., Chernova, O.F., Solomatina, N.I., Chemical
Communication in Amphibians, in Etologicheskie i ekolog-
icheskie aspekty khimicheskoi kommunikatsii nazemnykh
pozvonochnykh (Ethological and Ecological Aspects of
Chemical Communication in Terrestrial Vertebrates), Itogi
Nauki Tekh., Ser. Zool. Pozv., Moscow: VINITI, 1988a,
pp. 591.
Sokolov, V.E., Skurat, L.N., Stepanova, L.V., et al., Rukovod-
stvo po izucheniyu kozhnogo pokrova mlekopitayushchikh (A
BIOLOGY BULLETIN Vol. 36 No. 2 2009
 SKIN DERIVATIVES IN VERTEBRATE ONTOGENY AND PHYLOGENY
Manual of Studying the Mammal Skin), Moscow: Nauka,
1988b.
Terskikh, V.V., Vasilev, A.V., and Vorotelyak, E.Ya., Struc-
turalFunctional Units of Epidermis, Izv. Akad. Nauk, Ser.
Biol., 2003, no. 6, pp. 645649.
Terskikh, V.V., Vasilev, A.V., and Vorotelyak, E.Ya., Stem
Cell Niches, Izv. Akad. Nauk, Ser. Biol., 2007, no. 3, pp. 261
272.
Veltmaat, J.M., Mailleux, A.A., Thiery, J.P., et al., Mouse
Embryonic Mammogenesis As a Model for the Molecular
Regulation of Pattern Formation, Differentiation, 2003,
vol. 71, pp. 117.
Widelitz, R.B., Jiang, T.-X., Yu, V., et al., Molecular Biology
of Feather Morphogenesis: A Testable Model for Evo-Devo
Research, J. Exp. Zool., B: Mol. Dev. Evol., 2003, vol. 298,
no. 1, pp. 109122.
BIOLOGY BULLETIN Vol. 36 No. 2 2009
View publication stats
183
Widelitz, R.B., Veltmaat, J.M., Mayer, J.A., et al., Mammary
Glands and Feathers: Comparing Two Skin Appendages
Which Help Define Novel Classes During Vertebrate Evolu-
tion, Semin. Cell. Dev. Biol., 2007, vol. 18, no. 2, pp. 255
266.
Wu, P., Hou, L., Plikus, M., et al., Evo-Devo of Amniote
Integuments and Appendages, Int. J. Dev. Biol., 2004,
vol. 48, pp. 249270.
Yu, M., Wu, P., Widelitz, R.B., and Chuong, Ch.-M., The
Morphogenesis of Feathers, Nature, 2002, vol. 420,
nos. 69116917, pp. 308312.
Yu, M., Yue, Zh., Wu, P., et al., The Developmental Biology
of Feather Follicles, Int. J. Dev. Biol., 2004, vol. 48, pp. 181
191.
Yue, Zh., Jiang, T.-X., Widelitz, R.B., and Chuong, Ch.-M.,
Wnt3a Gradient Converts Radial Bilateral Feather Symmetry
Via Topological Arrangement of Epithelia, Proc. Natl. Acad.
Sci. USA, 2006, vol. 103, no. 4, pp. 951955.