Mammalian skin evolution: a reevaluation


Dr P. Maderson, Department of Biology, Brooklyn College of City University of New York, Brooklyn, NY 11210, USA
Tel.: +1 718 951 5706
Fax: +1 718 951 4659


Abstract:  A 1972 model for the evolutionary origin of hair suggested a primary mechanoreceptor role improving behavioral thermoregulation contributed to the success of late Paleozoic mammal-like reptiles. An insulatory role appeared secondarily subsequent to protohair multiplication. That model is updated in light of new data on (a) palaeoecology of mammalian ancestors; (b) involvement of HRPs in keratinization; (c) lipogenic lamellar bodies that form the barrier to cutaneous water loss; and (d) growth factors involved in hair follicle embryogenesis and turnover. It is now proposed that multiplication of sensory protohairs caused by mutations in patterning genes initially protected the delicate barrier tissues and eventually produced the minimal morphology necessary for an insulatory pelage. The latter permitted Mesozoic mammals to occupy the nocturnal niche ‘in the shadow of dinosaurs’. When the giant reptiles became extinct, mammals underwent rapid radiation and reemerged as the dominant terrestrial vertebrates.


Throughout the past 210 Myr, a time-traveler interested in mammalian or avian skin would see few differences from modern species. During the 142 Myr of the mid-late Mesozoic era (Fig. 1) (1), early mammals scurried around at night ‘in the shadow of dinosaurs’ (2). By day, giant reptiles ruled the warm continents and oceans while early birds flew above. The functional properties of fully evolved pelages and plumages, respectively, enabled mammals and birds to ‘bide their time’ waiting the demise of dinosaurs due to meteor impact at the end of the Cretaceous 65 Myr ago. Rocks of the 122 Myr preceding the Jurassic document the early history of amniote vertebrates (reptiles, mammals and birds), wherein we seek clues to skin evolution.

Figure 1.

Vertebrate history in a geological context. Asterisks mark groups known only as fossils. Lineages and taxa mentioned in the text shown in various fonts. After 200 Myr of aquatic life, the first Tetrapods (four limbed) evolved from fish ancestors. These Amphibians radiated in the Coal swamps of the Mississippian and Pennsylvanian periods giving rise to (→) the ancestors of modern amphibians (frogs, toads and salamanders) and the Anthracosaurs– amniote ancestors. Basal amniotes soon → organisms included generically in either the Theropsid lineage comprising Synapsid reptiles and mammals or the Sauropsid lineage comprising turtles and all living reptiles, the denizens of Jurassic Park, and birds. Data from (1).

Two features distinguish amniotes from amphibians (frogs, salamanders, etc.) and fish: their reproduction does not require water and they survive the dehydrating effects of life on land (1).

Mammals and birds independently evolved endothermy (a physiologically regulated, constant, body temperature) from their ectothermic (colloquially ‘cold-blooded’) reptilian ancestors (1). Reconstructing the evolution of physiological features involves evaluating fossils in light of data concerning soft tissue form and function in living species. Insulatory hairs and feathers are important to endothermy but how did they evolve? Because the fossil record for proto-mammals is chronologically more extensive and more detailed than for proto-birds (2,3), it might seem easier to trace the evolution of hair than of feathers. For several technical reasons this is not so. Here, two problems concern us. First, although the movie ‘Jurassic Park’ made dinosaurs (and their crocodilian and avian cousins) familiar to non-specialists, few know anything of mammalian forebears (synapsid reptiles) (2,3) who emerged from Palaeozoic coal swamps 300+ Myr ago, ruled the earth for 120 Myr and became extinct when their descendents were temporarily ‘side-lined’ by the dinosaurs. Second, although we are certain that hair evolved minimally 200 Myr BP, we lack any direct evidence.

Lacking fossil evidence, surviving proto-mammals, or any intermediates between hair and any known epidermal derivative, how do we address the problem of mammalian skin origins? While ubiquity of the insulatory role in extant taxa seems to imply an early relation to the origin of endothermy, basic physics show that initial changes in a reptilian integument could not have been directly related to such (4). Can we identify a plausible primary role for changes in a scaled skin that could lead eventually to ‘hair-like structures’?

A model for the origin of mammalian hair

On the assumption that the success of synapsid reptiles aeons before the ‘Age of Dinosaurs’ reflected some metabolic advantage, Maderson (5) offered a model proposing that: ‘… hairs arose from highly specialized sensory appendages of mechanoreceptor function which facilitated thermoregulatory behavior in early synapsids. …[A later chance mutation multiplied morphogenetic fields responsible for original, sparsely arranged ‘protohairs’ and produced a ‘protopelage’ whose properties were the subject of subsequent selection in an insulatory context although sensory function remained important]'.

Because of absence of relevant data this model ignored, or addressed inadequately, several questions: 1) In what environments did amniote skin evolution occur? 2) When and how did amniotes acquire a barrier to cutaneous water loss (CWL)? 3) Why did early synapsids reduce and lose their scales? 4) How did multiplication of hypothesized protohairs occur and what initial selective advantages accrued therefrom? These shortcomings have been ameliorated by data and interpretations from several different research programs and a new appreciation of complexity of form-function interplay in skin: we now see that scales in reptiles and epidermal appendages in mammals and birds (hairs and feathers) have important roles of mechanical protection at both organismic and tissue levels (6).

New data ameliorate deficiencies of the original model and reveal the unaddressed questions to be inter-related

Palaeontologists who reconstruct the ecology of extinct species suggest that while amniotes originated on islands in coal swamps 300 Myr BP, perhaps in pursuit of insects for food, ‘truly terrestrial’ ecosystems wherein synapsids faced potentially desiccating and mechanically abrasive environmental factors may not have emerged until 250 Myr BP (7).

Two conclusions relevant to mammalian skin origins emerge from studies of skin structure and development in reptiles and birds (4,8,9). The classic notion of ‘hard’ and ‘soft’ keratins/keratinized tissues has been ‘refined’ by the realization that corneous tissues may be based on either α- or β-proteins. A primordial histidine-rich protein (HRP) originated as a feature of amniote α-keratogenesis later expressed as keratohyalin granules (KHGs) in mammals and keratohyalin-like granules (KHLGs) in reptiles and birds.

Another amniote epidermal feature is the association of lipogenic lamellar bodies (LBs) with α-keratogenic tissues reducing CWL (10). However, different barrier cytologies in at least three lineages imply independent acquisition (4,10), so that perhaps parental coping with CWL constrained emergence of ‘full terrestrialism’– not reproductive mode (1)!

A plausible explanation for scale reduction and loss in early synapsids is that Theropsid amniotes [‘Sail’ and ‘Mammal-like’ reptiles plus mammals (Fig. 1)] pursued a different strategy to cope with environmental abrasion of their skin than that seen in other reptiles (Sauropsids) although, interestingly, birds ‘copied’ this strategy when feathers evolved (4,6). The nature of, and basis for, the hypothesized theropsid strategy can be summarized. A mutation involving a molecular trigger involved in patterning produced the initial multiplication of mechanosensory ‘protohairs’ and the resultant ‘protopelage’ had an initial selective advantage in that it was the final step in providing a tissue to protect the barrier to CWL. Later, as endothermy was gradually perfected in early mammals, perhaps a Jurassic event when they occupied the nocturnal niche (11), an insulatory boundary was further improved by multiplication of non-tactile hairs.

New molecules, organelles and tissues in the evolution of form and function in vertebrate skin – a synthesis


Mammalian skin evolution is best understood against a background of two basic facts. First, vertebrate skin is geometrically patterned whether comprising scales or appendages (5). Second, several ‘new’ molecules, organelles and therefore cytodifferentiative events producing tissue differences have characterized epidermal history (Fig. 2) (4).

Figure 2.

Evolutionary history of some molecules, organelles and tissues that characterize the skin of tetrapod vertebrates. Arrows on branching cladogram show origins of major lineages, emphasizing Theropsid Amniotes (Mammals and their reptilian antecedents). Groups mentioned in text are shown in the same font as in Figure 1. Asterisks indicate those known only as fossils. ‘Dinosaurs’*# include bird ancestors. Major skin changes indicated by numbered boxes. [1] Epidermal covering of patterned scales comprises mucogenic cells containing 70A tonofilaments indicating α-keratogenic potential. [2] Primitively, tetrapods bore scales covered by a stratified, squamous epidermis whose cells possessed an envelope and showed reduced mucogenesis compared with ancestral fish. [3] Modern amphibians have lost patterned scales. [4?] In some Paleozoic amphibia or perhaps basal amniotes (4), major changes in epidermal structure facilitated a more terrestrial life-style: further reduction in mucogenicity accompanied α-keratogenesis enhanced by filaggrin derived from a primordial Histidine Rich Protein ( pHRP) and keratinocytes formed lipogenic lamellar bodies. [5] Theropsid amniotes soon reduced and lost patterned scales and epidermal α-keratogenesis involved keratohyalin granules. [6] Spatially patterned epidermal appendages –‘mechanosensory protohairs’– facilitated behavioral thermoregulation (5). [7] Multiplication of protohairs caused by changes in patterning molecules initially protected delicate barrier tissues and formed a ‘protopelage’ whose insulatory properties permitted refinement of mammalian endothermy. [8] In reptilian sauropsids, β-keratogenic epidermal tissues covering scales protect the subjacent α-keratogenic tissues that house the barrier to CWL. Vertical alternation between α- and β-keratogenesis involves keratohyalin-like granules (4). [9] In birds, replacement of scales by epidermal appendages – feathers – protects barrier tissues and facilitates both flight and avian endothermy.

Events in early tetrapods

Throughout evolution the overall mucogenicity of vertebrate epidermis has been steadily reduced and in amniotes it is expressed only in oral, genital and anal mucus membranes.

In amphibians, enhancement of α-keratogenicity ( primitively represented in all fish by 70A tonofilaments) by a ‘cell envelope’ (marginal layer containing involucrin) permitted formation of a stratified, cornified epidermis (12,13). It is no barrier to CWL so that most species are confined to ‘wet and therefore mechanically lubricating’ environments (1).

In Palaeozoic, anthracosaurian amphibians, or ‘Basal Amniotes’ (Figs 1 and 2), α- keratinization was enhanced by filaggrin derived from a primordial Histidine Rich Protein (pHRP) (14). Lipogenic lamellar bodies (LBs) might have provided some barrier to CWL, but the ‘delicate’ epidermal tissues covering scales were susceptible to abrasion so that organisms were still restricted to a ‘wet’ aquatic environment (4).

Two different strategies in amniotes

In Sauropsid Amniotes, mechanical protection derives from the β-keratogenic epidermal tissues covering reptilian scales, or avian feathers (4,6). Sauropsid HRP, represented by KHLGs, is involved in vertical alternation between α- and β-keratogenesis (4,9). The barrier to CWL is housed in α-keratogenic tissues that are protected by overlying β-keratogenic tissues. Their various degrees of efficacy permitted exploitation of desiccating and abrasive terrestrial niches.

Early Theropsid Amniotes evolved a skin structurally and functionally similar to that of modern toads, living amphibians whose reduced epidermal mucogenicity and cornification (4) permits them to walk over dusty driveways and hide in garages! Theropsids lost scales because their α-keratogenic epidermis was somewhat toughened by mammalian-type HRP (KH), and the derived hydrophilic filaggrin facilitates epidermal flexibility (14). Their barrier tissues, whose minimal efficacy is suggested by the environments they inhabited, remained susceptible to abrasion unless ‘lubricated’ regularly. These organisms almost certainly practiced ‘behavioral thermoregulation’ as do many living frogs and toads, and this behavior was facilitated by the spatially patterned sensory protohairs (5).

In late Theropsid Amniotes, multiplication of protohairs, caused perhaps by a mutation leading to up-regulation of a patterning trigger such as β-catenin (15), provided the necessary enhanced mechanical protection for the thin stratum corneum. Once the included barrier tissues were less susceptible to abrasion, exploration of ‘true terrestrial’ niches became possible.

In basal Mammals, the ‘protopelage’– whose constituent units might have been strengthened by trichohyalin (14), to form ‘true hairs’– had dual roles. First, an increased density of hairs further improved mechanical protection of inter-follicular barrier tissues (4,6). Second, this density also permitted an insulatory function that could have been further enhanced by other changes in the patterning mechanisms (17).

Concluding comments

Hypothetical models of this type do not solve evolutionary questions and are not intended to do so. Their value lies in the fact that continual incorporation of new data from a variety of biologic disciplines into such models, and/or presentation of new models to supplant them enhances the probability that new research programs addressing empirically refutable hypotheses will emerge from such interdisciplinary considerations.