SEARCH

SEARCH BY CITATION

Keywords:

  • evolution;
  • hair ultrastructure;
  • marsupial;
  • monotreme;
  • rodent tail skin;
  • trichohyalin immunocytochemistry

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The inner root sheath (IRS) allows the exit of hairs through the epidermal surface. The fine structure of monotreme and marsupial IRS and trichohyalin is not known. Using electron microscopy and immunocytochemistry, the localization of trichohyalin and transglutaminase have been studied in monotreme and marsupial hairs, and compared with trichohyalin localization in placental hairs. Trichohyalin in all mammalian species studied here is recognized by a polyclonal antibody against sheep trichohyalin. This generalized immunoreactivity suggests that common epitopes are present in trichohyalin across mammals. In differentiating IRS cells, trichohyalin granules of variable dimensions are composed of an immunolabelled amorphous matrix associated with a network of 10–12-nm-thick keratin filaments. Transglutaminase labelling is present among keratin bundles and trichohyalin granules, and in condensed nuclei of terminally differentiating cells of the inner root sheath. The IRS in monotreme hairs is multistratified but lacks a distinguishable Henle layer. Cornification of IRS determines the sculpturing of the fibre cuticle and later shedding from the follicle for the exit of the hair fibre on the epidermal surface. It is hypothesized that the stratification of IRS in Henle, Huxley and IRS cuticle layers is derived from a simpler organization, like that present in the IRS of monotremes. The IRS is regarded as a localized shedding/sloughing layer needed for the exit of hairs without injury to the epidermis. The formation of the IRS during the evolution of mammalian epidermis allowed the physiological exit of hairs produced inside the skin. The peculiar morphogenesis of hairs in possible primitive skins, such as those of the monotremes (mammals with some reptilian characteristics) or the tails of some rodents (a scaled skin), may elucidate the evolution of hairs. In monotreme and rodent tail skin, the dermal papilla remains localized on the proximal side of the hair peg and forms a hair placode with bilateral symmetry. The papilla is progressively surrounded by the down-growing hair peg until a dermal papilla with radial symmetry is formed. It is speculated that the progressive reduction of the extended dermal papilla of reptilian scales into small and deep papillae of therapsid reptiles produced hairs in mammals.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Hairs are characteristic epidermal appendages of the skin of mammals that comprise different shells of elongated cells that differentiate into a central hair fibre (medulla, cortex and fibre cuticle), surrounded by an inner root sheath (IRS, divided into a sheath cuticle, Huxley and Henle layers) and an outer root sheath (ORS) (Montagna & Parakkal, 1974; Orwin, 1979; Rogers et al. 1999). Hairs posses variable shape, dimension and internal structure. During embryogenesis, hairs derive from the induction of a specialized group of dermal fibrobasts (dermal papilla) on a receptive epidermis (epidermal placode) (reviewed by Hardy, 1992; Jahoda et al. 1992; Philpot & Paus, 1999; Millar, 2002). The dermal papilla also re-activates epidermal cells of the ORS (especially in the bulge region) that re-form hair matrix cells for the regeneration of a new hair after shedding of the old hair (reviewed by Stenn et al. 1999; Stenn & Paus, 2001).

It is assumed that hairs, as a typical mammalian characteristic, present a similar morphogenesis, structure and modality of growth in all subclasses, monotremes, marsupials and placentals (Sokolov, 1982). It is believed that whereas monotremes diversified from the remaining mammals during the Triassic, marsupials represent a group of mammals that diversified from placentals much later, probably during the Cretaceous, so that overall they show closer affinities with placentals than with monotremes (Stonehouse, 1977; McFarland et al. 1979).

Monotremes have numerous reptilian characteristics, and studies of their skin may reveal useful characteristics pertaining to the evolution of hairs. Monotreme hairs conform to the general structure of those of the remaining mammals, although some peculiarities of their morphogenesis and structure have been reported (Poulton, 1894; Spencer & Sweet, 1899; Wildman & Manby, 1938; Griffiths, 1978; Sokolov, 1982). Among these peculiarities, neither dermal condensation nor hair placode have been described during the initial stages of hair morphogenesis. The hair placode is formed on a side of the hair peg and soon becomes connected to a dermal papilla. The hair peg is formed as an open tube on the epidermal surface. The IRS is not clearly distinguished into three layers (Henle, Huxley and IRS cuticle) and appears to be a continuum that is not clearly distinguishable from the corneous layer of the epidermis. The upper coat hairs (primaries) have a more or less expanded and flat apical portion, and are termed shield hairs, an uncommon characteristic in hairs of other mammals.

Studies performed on marsupial hairs have revelaed a substantial uniformity in structure and morphogenesis with hairs of placentals (Gibbs, 1938; Bolliger & Hardy, 1944; Lyne, 1957, 1970; Lyne et al. 1970; Barbour, 1977; Mykytowyctz & Nay, 1987; Alibardi & Maderson, 2003). Although the basic histology of monotreme and marsupial hairs is known, no ultrastructural analysis has been performed, in particular on the fine structure of the IRS in comparison with that of placental hairs. It is believed that the main functions of the IRS are to mould the cuticle pattern, to hold the hair inside the follicle and to direct it toward the surface of the epidermis before it is sloughed (Orwin, 1979; Woods & Orwin, 1982; Rogers et al. 1999). It is the disappearance of the IRS around the hair fibre in the upper part of the hair canal that permits the physiological exit of the shaft on the epidermal surface without pricking the epidermis from beneath. In the present study it is emphasized that the IRS represents a shedding/sloughing layer connected with the physiological process of detachment between the hair fibre and the epidermis near the upper part of the hair canal.

Cells of the IRS of placental hairs are organized into IRS-cuticle, Huxley and Henle layers. These cells contain trichohyalin granules, which largely comprise a 190–220-kDa protein (depending on the species) rich in arginine that is called trichohyalin (Rothnagel & Rogers, 1986; Hamilton et al. 1993; Rogers et al. 1999). The latter protein determines the characteristic cornification of the IRS, which is different from the process of cornification of both hair fibre and ORS/epidermis (Orwin, 1979; Rogers 1985; O'Guin & Manabe, 1991; Rogers et al. 1991; Fietz et al. 1993; Manabe & O'Guin, 1994). Trichohyalin seems to be an interkeratin-associated protein, which is initially localized in trichohyalin granules and is rapidly mixed to specialized keratins within IRS cells, and determines the cornification of IRS (Taresa et al. 1997; Langbein et al. 2002).

The enzyme peptidyl-arginine deiminase catalyses the deamination of the lateral chains of peptidyl-arginine present in trichohyalin-forming citrulline residues (Rogers et al. 1999). Trichohyalin is also a substrate of hair transglutaminase (TGase), which cross-links the protein into a dense network of parallel, intermediate-like filaments by the formation of N-ɛ(γ-glutamyl) lysine isopetide bonds (Rogers et al. 1999). During terminal differentiation of IRS cells, keratin filaments connected with trichohyalin granules organize themselves into parallel rows, and citrulline-rich trichohyalin (which is more soluble than its arginine-rich precursor) is dispersed among them (Rogers et al. 1991; Taresa et al. 1997; Langbein et al. 2002; Steinert et al. 2003). Finally, trichohyalin is cross-linked by the transglutaminase present in maturing IRS cells. The transformation of arginine into citrulline residues determines the susceptibility of trichohyalin to degradation by lytic enzymes, which are produced in terminally differentiating IRS cells, in cells of the ORS, and from sebaceous gland secretions (O'Guin & Manabe, 1991; Rogers et al. 1999).

No information is available on the distribution of trichohyalin in monotremes and marsupials, on the fine structure of their IRS and on the chemical similarity of their trichohyalin to that of placentals (which might provide an indication of conserved common epitopes in different species of mammals). The goals of the present study are: (1) to describe for the first time the ultrastructure of the IRS in active bulbs (anagen) of monotreme and marsupial hairs; (2) to detect a possible cross-reactivity of their trichohyalin with that of placentals; (3) to compare the cytological and immunocytochemical characteristics of IRS in monotreme and marsupials with those of placentals. Emphasis of the present study is given to the stratification of IRS layers, its trichohyalin-based modality of cornification in relation to shedding of IRS cuticle from the hair fibre cuticle; (4) and finally the morphogenesis of hairs and IRS differentiation in monotremes (considered primitive) is compared with that of the scales of rodent tail skin, also considered to be primitive (Spearman, 1964; Alibardi & Maderson, 2003). This comparison is used to formulate an hypothesis on hair evolution from the skin of reptilian ancestors.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The present study was conducted on small skin samples collected from different species of Australian monotremes and marsupials: platypus (Ornithorynchus anatinus, n = 6, foot skin); echidna (Tachyglossus aculeatus, n = 2, belly skin); red kangaroo (Macropus rufus, diprotodonts, n = 1, dorsal skin), eastern possum (Trichosurus vulpecula, diprotodonts, n = 2, ear skin), wombat (Vonbatus ursinus, diprotodonts, n = 1, dorsal skin), stripe-faced dunnart (Sminthopsis macroura, polyprotodonts, n = 2, ventral skin). For comparison of trichohyalin immunoreactivity, skin samples from some placental mammals were also used: mole (Talpa europaea, insectivores, n = 1, tail and ventral skin), rat (Rattus norvegicus, rodents, n = 9, ventral and tail skin), mouse (Mus musculus, rodents, n = 9, ventral and tail skin), rabbit (Oryctolagus cuniculus, lagomorphs, n = 2, ear and ventral skin), cat (Felis catus, carnivores, n = 1, ventral skin) and humans (Homo sapiens, primates, n = 2, limb skin). Skin biopses were collected according to principles of ethical animal care and handling.

Developing hairs were studied in tail and ventral skin of at 17–18 days (mouse, n = 6; rat, n = 4) and 19–20 days (mouse, n = 4; rat, n = 6) of embryonic life. Other tail-skin samples were collected at 0–1 day (mouse, n = 3; rat, n = 3), 3–4 days (mouse, n = 4; rat, n = 4) and 7–8 days (mouse, n = 4; rat, n = 5) postpartum. The animals were killed by decapitation.

Small pieces of skin (2 × 4 mm) were collected and immediately fixed. For the ultrastructural study, pieces were fixed for 6–8 h in 2.5% glutaraldehyde in phosphate buffer 0.1 m at pH 7.4, post-fixed in 2% osmium tetroxyde, dehydrated and embedded in Spurr's resin. For immunocytochemistry, tissues were fixed for 4–5 h in 3% paraformaldehyde in 0.1 m phosphate buffer at pH 7.4 or in Carnoy's fixative (9 parts 80% ethanol + 1 part acetic acid), dehydrated and embedded in Lowcryl KM4 resin under ultraviolet light polymerization at 0–4 °C.

Sections 1–3 µm in thickness were obtained from the Spurr's-embedded tissues using an ultramicrotome. For the microscopical study, sections were stained with 0.5% toludine blue. Thin sections (40–90 nm thick) from selected areas of the skin were collected on copper or nickel grids, and were stained with uranyl acetate and lead citrate for the electron microscopic study. These sections contained hairs sectioned in cross, oblique or longitudinal sections. Hairs were studied at different levels, from the bulb above the matrix to the sebaceous gland. These sections showed the IRS at different stages of differentiation. In the present study, larger hairs (presumably primary or guard hairs) in all specimens were studied using an electron microscope. Hairs were studied during the apparent anagen stage, but telogen hairs were also seen; owing to the limited amount of samples available, either the age of animals or the period of hair growth cycles were not known. It was considered that large hairs in the anagen stage, coupled with references from the few literature sources (cited above), were representative enough to provide the initial information on ultrastructural features of large hairs in these species.

Other 2–5-µm-thick sections were collected from tissues embedded in Lowcryl K4M, and they were attached to gelatine-coated slides for the immunocytochemical study. The polyclonal antibody against sheep-trichohyalin was produced in rabbits (see Rothnagel & Rogers, 1986). It was generously supplied by Dr G. Rogers (University of Adelaide, Australia), and recognizes rat, sheep and human trichohyalin, but its cross-reactivity to trichohyalin of other species is unknown (Rothnagel & Rogers, 1986; O'Guin & Manabe, 1991). A rabbit polyclonal antibody against guinea-pig TGase was purchased from Abcam (ab421, Abcam Ltd, Cambridge, UK; kindly supplied by Dr G. Gargiulo, University of Bologna, Italy).

For light microscopy immunocytochemistry, tissues were pre-incubated for 20–30 min in 5% normal goat serum with 2% BSA in 0.05 m Tris/HCl buffer at pH 7.6 (to reduce background binding). The pre-incubation solution was replaced by the incubation solution (the Tris/HCl buffer with no goat serum but containing the primary antibody at a dilution of 1 : 500 for trichohyalin). The incubation lasted overnight at 0–4 °C for both tests and their control sections (in which the primary antibody was omitted). After several rinses in the Tris/HCl buffer, the sections were incubated in 2% BSA in 0.05 m Tris/HCl buffer at pH 7.6 for 1 h at room temperature containing 1 : 50 of anti-rabbit-IgG FITC-conjugated secondary antibodies. After rinsing, sections were mounted in Fluoromount (EM Sciences, USA) and observed under a Zeiss epifluorescence microscope equipped with a fluorescein filter.

For immunoelectron microscopy, 40–90-nm-thick sections were collected on nickel grids. Non-specific binding sites were blocked by incubating for 10–20 min in 1% cold-water fish gelatine or 2% BSA in Tris buffer as above, and immunostained overnight at 0–4 °C with the anti-trichohyalin (1 : 500) or anti-transglutaminase (1 : 150) antibodies (the primary antibody was omitted in controls). After three rinses with the buffer, an anti-rabbit IgG conjugated to 10-nm gold particles was used as the secondary antibody (Sigma, USA, or Biocell, UK). After a light staining with uranyl acetate, the sections were observed under a CM-100 Philips electron microscope.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Fine structure and immunoreactivity for trichohyalin in monotreme hairs

Cross-sectioned hair bulbs of platypus and echidna beneath the hair cone revealed undifferentiated layers of cells (of the future IRS and hair fibre) inside the external epithelium of the ORS (Fig. 1a). Above the hair cone, the hair was made of a partially cellularized cortex surrounded by concentric layers of flat cells forming the IRS and ORS. In more differentiated hairs, the IRS appeared cornified and thinner than the ORS, and it comprised 4–6 layers of very flat cells (Fig. 1b,c). A granulated layer externally contacted the IRS. This corresponds to the companion layer described in placental hairs (Fig. 1c).

image

Figure 1. Light (a–c) and ultrastructural (d–f) figures of monotreme hairs. a, cross-section of undifferentiating platypus foot hair, showing the external layers of the ORS and of the forming IRS. 770×. b, cross-section of differentiated hair of echidna belly epidermis with the cornified IRS (double arrowhead) containing few pale immature cells (arrowhead). 400×. c, longitudinal section of echidna hair with the multistratified IRS and cuticle (arrowhead) facing the hair fibre cuticle (arrow). The companion layer contains granular inclusions (double arrowhead). 450×. d, polygonal-shaped IRS cells with trichohyalin granules (double arrowheads) of differentiating platypus hair. The arrowhead indicates the basal membrane of the ORS. 5400×. e, detail of the stratification of the IRS (arrowheads indicate cell membranes) in a differentiated hair of platypus foot. The companion layer (double arrowhead) contains numerous non-trichohyalin granules (double arrowheads) and dark keratin bundles. 7500×. f, fibrous parallel network (arrow) in mature IRS of echindna hair as shown in c. Dense round granules (arrowheads) are present in cells of the companion layer. 4300×. CP, companion layer; H, hair fibre; I, IRS; IC, cuticle of the IRS; O, ORS.

Download figure to PowerPoint

The ultrastructural examination of the 4–5 outer layers of the sectioned hair showed that they differentiated into the ORS, and the remaining 4–6 inner layers formed the IRS. The latter appeared polygonal and paler than the external, flat ORS cells. IRS cells began to accumulate round trichohyalin granules (Fig. 1d). In more differentiated hairs, all stratified cells of the IRS contained a rich fibrous network, including the innermost cuticle cells, which formed a serrated layer with that of hair cuticle cells (Fig. 1e,f). A distinct, external Henle layer was not visible in large and smaller hairs, and the external cornified cells were joined to the granulated companion layer. In the latter, numerous dark keratin bundles were present together with dense granules with an average range of diameters between 0.05 and 0.5 µm. Trichohyalin granules, round to irregular in shape, were formed by a dense matrix among areas of the cytoplasm of IRS cells, which were often associated with keratin filaments.

The IRS was immunofluorescent using antibodies to trichohyalin (Fig. 2a). Only trichohyaline granules with variable diameter (0.2–0.7 µm) within immature or incompletely cornified IRS cells were intensely immunoloabelled, whereas the granules in the companion layer were not labelled (Fig. 2b–d). Only a few granules in the companion layer were immunoreactive.

image

Figure 2. Light (a) and ultrastructural (b–d) immunolocalization of trichohyalin in ventral echidna hairs. a, cross-section similar to Fig. 1(b), showing immunoreactivity in the IRS (arrow). 340×. b, immunolabelled small granule (arrow) within the cytoplasm of a differentiating IRS cell. No labelling is seen in the surrounding cytoplasm and in the mature cells (arrowhead). 52 000×. c, detail of 10–12-nm irregular immunolabelled filaments (arrowhead). The surrounding filaments with parallel orientation are unlabelled (arrow). 110 000×. d, large and small immunolabelled trichohyalin granules (arrows) among the immunonegative keratinized matrix. 42 400×. H, hair fibre; K, keratin filaments/bundles; O, ORS; TH, trichohyalin immunolabelling.

Download figure to PowerPoint

The ultrastructural study showed that gold particles were associated with a network of 10–12-nm-thick dense filaments, but the labelling was not exclusive of these filaments. Instead, the labelling appeared close to the filaments or was associated with their periphery more than over the filaments. Controls were unlabelled.

The electron-pale keratin bundles around trichohyalin granules were immunonegative. A higher magnification study on the pale keratin bundles showed that 10–12-nm filaments were also present but that they were less electron-dense and less easily distinguishable than those within trichohyalin granules. In some sections, the 10–12-nm-thick filaments were orientated in parallel rows within maturing IRS cells but the labelling was nearly absent (Fig. 2c). Trichohyalin granules appeared as areas among keratin bundles where the 10–12-nm filaments formed a labelled network that was not seen in most of the amorphous mass of keratin (Fig. 2d).

Fine structure and immunocytochemistry for trichohyalin of marsupial hairs

In the different marsupials studied here, hairs were either medullated or non-medullated (Fig. 3a,b). The ultrastructure of large hairs in the follicle (medullated, primary hairs) was studied in the wombat and red kangaroo (Fig. 3c). Cortical cells accumulated long keratin bundles whereas medullary cells and IRS cells accumulated 0.05–3-µm trichohyalin granules (Fig. 3c). Keratin bundles, with circular orientation in cross-sectioned hairs, were more frequent in the companion layer than in cells of the ORS, and numerous desmosomes joined this layer with the layer of Henle. In addition, in some sections cytoplasmic elongation of Huxley cells was seen among Henle cells, directly in contact with companion cells (Fig. 4b, ultrastructural data not shown). In kangaroo larger hairs, up to two layers of Huxley cells were commonly present in their cross-section. In the large hairs of the wombat up to 4–5 layers of Huxley cells were normally seen in their larger section. It was not specifically studied whether the number of layers varied along the whole length of hairs above their follicle. When the cortex, cuticle and IRS were cornified, the medulla remained visible inside the hair shaft (Fig. 3d).

image

Figure 3. Structure and immunocytochemistry of marsupial hairs. a, general structure and scalation of hair fibre cuticle (arrowhead) near the bulb of a kangaroo hair. 410×. b, cross-section of kangaroo hair with differentiating IRS, cortex and medulla (arrow). 450×. c, ultrastructural detail of layers of the IRS of kangaroo hair. Dense trichohyalin granules (arrow) within Huxley layer and keratin material in cuticle cell (arrowhead). 10 600×. d, two cross-sectioned possum hairs with completely cornified IRS around the hair fibre. The arrow indicates the medulla. 640×. e, trichohyalin-immunofluorescent IRS of opossum hair. The central hair fibre and ORS are immunonegative. 700×. f, cross-sectioned hairs of dunnart skin showing the trichohyalin-immunofluorescent IRS and medulla. 470×. U, hair bulb; C, cortex; CP, companion layer; H, hair fibre; HE, Henle layer; HU, Huxley layer; I, IRS; IC, cuticle of the IRS; ME, hair medulla; O, ORS; TH, trichohyalin immunolabelling.

Download figure to PowerPoint

image

Figure 4. a, immunogold-labelled trichohyalin granules of Huxley cells of kangaroo hair with associated keratin filaments. 45 900×. b, detail of coarse filaments (arrowhead) of immunolabelled trichohyalin granules. The arrow points to the immunonegative parallel filaments of cornified cytoplasm in Henle layer. 73 800×. HE, Henle layer; K, keratin filaments/bundles; T, trichohyalin granules; TH, trichohyalin immunolabelling.

Download figure to PowerPoint

The immunoreactivity for trichohyalin in various species of marsupials (diprotodonts such as kangaroo, wombat and possum, and polyprotodons such as the dunnart) was specifically present in the IRS or the medulla, when present, of large and small hairs (Fig. 3e,f). Numerous trichohyalin granules of different dimension were seen in the Huxley layer, and appeared as round organelles consisting of a network of 10–12-nm-thick filaments among a pale matrix, surrounded by bundles of keratin (Fig. 4a,b). The granules were immunolabelled with the trichohyalin antibody whereas the surrounding keratin-like bundles were unlabelled. Analysis at high magnification showed that the labelling was associated with but did not overlap the 10–12-nm-thick filaments. Gold particles were mostly present at the border of or over the 10–12-nm filaments and less frequently in the low electron-dense spaces among the filaments (Fig. 4b). Larger granules or coalescing granules were present in the maturing Huxley cells. As in monotreme hairs, in hairs of the kangaroo, trichohyalin granules appeared as immunolabelled areas where a network of 10–12-nm-thick filaments were seen among amorphous keratin bundles.

In mature cells of the Henle layer, at higher levels of the hair and in cells of the Huxley layer, immunolabelling for trichohyalin disappeared. In these mature cells a regular meshwork of parallel 10–12-nm-thick filaments alternated with less electron-dense material (Fig. 4b).

Ultrastructural localization of transglutaminase in monotreme and marsupial IRS

A sparse or a completely negative immunolabelling was seen among the irregularly orientated 10–12-nm-thick, and parallel filaments of trichohyalin in echidna and kangaroo IRS (Fig. 5a). No TGase-immunolabelling was present in fully cornified IRC in both species. The gold-labelling for transglutaminases was more concentrated over the heterochromatine of nuclei of maturing IRS cells in both echidna and kangaroo hairs (Fig. 5b). Gold particles were absent or sparse over trichohyalin granules in both echidna and kangaroo IRS. Gold particles were more concentrated on the cornifing cytoplasm of mature Huxley and Henle cells in kangaroo hairs (Fig. 5c).

image

Figure 5. Transglutaminase immunolabelling in echidna (a) and red kangaroo (b,c) IRS. a, diffuse labelling among 10–12-nm-thick parallel filaments (arrows) in maturing cell. The arrowhead points to a darker, more compact cell with almost no gold labelling. 46 900×. b, immunolabelled nucleus of Henle layer. 70 000×. c, diffuse immunolabelling in corneous material within maturing Huxley cell and immunonegative small trichohyalin granules (arrows). 41 200×. K, keratin filaments/bundles; NU, nucleus; TG, transglutaminase immunolabelling.

Download figure to PowerPoint

Fine structure of forming hairs in scales of rodent tail

At 17–18 days of embryonic life in mouse and at 18–19 days in rat, the tail tip was made of a multistratified epithelium with a multilayered granular layer. In more proximal regions small epidermal pegs in the basal layer were seen, which represent hair placodes (Fig. 6a,b). In the latter, cells became elongated and more toluidinophilic (basophilic), but no dermal condensation or dermal papillae were present. Dermal cells were perpendicularly orientated beneath both hair placodes and non-placode epidermis. Dermal papillae became clearly visible during initial hair peg formation, but were localized only on the proximal (rostral) and lowermost side of pegs (Fig. 6c,d). The columnar hair placode epidermis, located on the proximal side of the down-growing peg, was associated with a proximally located dermal papilla. The papilla was progressively surrounded by the peg and became centrally located inside the base of the peg (Fig. 6e). The connections between cells of the dermal papilla and those of the connective sheath surrounding the hair peg were evident in both longitudinal and serial cross-sections (Fig. 6f–i). The latter confirmed that the dermal papilla was formed ventrally and proximally to the hair placode (Fig. 6g–i). At 19–20 days of intrauterine embryonic life, a similar sequence of hair morphogenesis was still seen moving from proximal to distal areas of the tail.

image

Figure 6. Hair morphogenesis in mouse tail skin of 17–18 days of embryonic life. a, undulation of epidermis near the tail tip with hair placodes (arrowheads). 550×. b, detail of an asymmetric hair placode (arrow) at the centre of future scale borders (arrowheads). 550×. c, down-growth of the hair placode with associated dermal papilla (arrowheads) located on the proximal side of the hair peg in correspondence with the future hinge region of the scale (arrow). The remaining fibroblasts maintain the perpendicular orientation (double arrowhead). 480×. d, detail of an asymmetric elongated peg with proximal hair placode (arrow) and penentrating papilla (arrowhead). 490×. e, two hair pegs in correspondence with the future hinge regions of scales (arrows) with proximally penetrating mesenchymal cells to form radiate papillae (arrowheads). 460×. f, detail of a more tangential section of hair peg to show the continuity between papilla cells and those of the connective sheath (arrowheads). 400×. g, cross-section of pedunculated hair peg with surrounding flat fibroblast of the connective sheath (arrowheads). 500×. h, cross-sectioned hair peg surrounded by cells of the connective sheath (arrowhead). 470×. i, a more proximal cross-section of the hair peg of the previous figure shows the appearance of the dermal papilla (arrow) on the ventral side of the remaining hair peg, surrounded by the connective sheath (arrowhead). 450×. E, epidermis; GR, granular layer; P, hair peg.

Download figure to PowerPoint

At 0–1 day postpartum, whereas longer hair follicles were present in proximal regions, in more apical parts of the tail slanted hair pegs were still developing. In the latter, pale fibroblasts of the dermal papillae appeared to contact laterally the rostral side of the hair peg (Fig. 7a–c). The epidermis showed a well-developed stratum granulosum. The latter was present also at 4 days postpartum near the tip of the tail, where hairs were elongated and dermal papillae were deeply penetrated into hair pegs forming typically radial papillae (Fig. 7d). In proximal areas of the tail skin at 4 days postpartum, hairs were well developed and hair bulbs were shaped as an enlarged, glandular-like flask from which the hair fibre emerged (Fig. 7e). In the epidermis of forming scales the granular layer was reduced but still present. The granular layer was partially or completely absent in the epidermis of scales (between hair bulbs) at 7–8 days postpartum in both species (Fig. 7f).

image

Figure 7. Toluidine-blue-stained sections of forming tail scale in the rat at 0–1 (a–c), 2 (d), 4 (e) and 7 (f) days postnatal. a, three elongating and obliquely orientated hair pegs with forming dermal papillae localized on the proximal side of pegs (arrows). The arrowhead points to the granular layer. 290×. b, detail of the hair placode (arrow) and dermal papilla (arrowhead). The double arrowhead points to the granular layer. 490×. c, other detail showing the posterior penetration of the dermal papilla surrounded by the hair placode (arrows). 490×. d, elongated hair pegs with deep dermal papilla (arrowhead) while the granular layer is still well developed (double arrowhead). 280×. e, developed hair bulb with medullated hair (arrows) while the epidermis of the forming scale still retains a granular layer (arrowhead). 180×. f, the epidermis of the forming scale between two successive hairs (arrows) has lost most of the granular layer (arrowhead). 280×. BU, hair bulb; E, epidermis; FS, forming scale; P, hair peg.

Download figure to PowerPoint

The ultrastructural analysis of mouse skin at 17–18 days of embryonic life showed that the basal portion of elongated, electron-dense and ribosome-rich cells of the hair placodes was contacted by mesenchymal elongations (Fig. 8a,b). In these areas, the dense lamella beneath the epidermis showed numerous interruptions, and epidermal elongations were seen to cross or deform the lamella, while collagen fibrils decreased and an amorphous extracellular matrix was present (Fig. 8b,c). In non-placode areas of the skin, collagen fibrils were more numerous (Fig. 8d), and a dense lamella with collagen fibrils was generally present. These ultrastructural features were also seen in cells of the dermal papilla and epidermal cells of the hair placode in more advanced stages of hair morphogenesis. Numerous collagen fibrils surrounded the hair peg in contact with elongated fibroblasts of the connective sheath (Fig. 9a).

image

Figure 8. Ultrastructural details of the dermis under hair pegs in embryonic mouse of 18 days. a, asymmetric hair placode (double arrowhead) with elongated electron-dense cells surrounded by paler epidermal cells. Mesenchymal cells are perpendicularly orientated (arrowhead). 1300×. b, detail of perpendicularly orientated mesenchymal cell contacting the hair placode. Few collagen fibrils and a discontinuous dense lamella (arrowhead) in correspondence with a hemidesmosome (double arrowhead) are seen. 63 000×. c, cytoplasmic elongation (arrowhead) of epidermal cell crossing the desolving dense lamella (double arrowheads) to contact mesenchymal elongations. 42 700×. d, collagen fibrils (arrowhead) are more numerous in non-placode areas. 32 300×. CL, collagen fibrils; D, dermal cells; E, epidermis; M, mesenchymal cell.

Download figure to PowerPoint

image

Figure 9. a, ultrastructural detail of the numerous collagen fibrils (double arrowheads) connecting cells of the connective sheath to the basement membrane of epidermal cells of the hair peg of mouse skin at about 18 days of embryonic life. 12 000×. b, two hair pegs from ventral epidermis in embryonic rat of 18 days, showing a radial papilla (double arrowhead). 450×. c, bent hair bulb (double arrowhead) of a 7–8 day postpartum mouse tail. A group of fibroblasts (arrowhead) are external to the bulb while the IRS (arrow) is partially cornified. 250×. c, immunofluorescence for trichohyalin in IRS and medulla of longitudinally sectioned ventral hairs of adult rat. 350×. d, immunofluorescence for trichohyalin in IRS and medulla of longitudinally sectioned ventral hairs of adult rat. 350×. e, immunofluoresent IRS and medulla cells in cross-sectioned ear rabbit hairs. 400×. E, epidermis; F, fibroblasts; I, IRS; ME, medulla; P, hair peg; TH, trichohyalin immunolabelling.

Download figure to PowerPoint

Analysis of the developing skin in the ventral areas (belly) of rat and mouse at 17–18 days of embryonic development revealed that hairs were at a more advanced stage (hair pegs) than in the distal parts of the tail. Dermal papillae at this stage were already radial (Fig. 9b). The search for earlier stages (up to the hair germ or plug stage) showed that ventral hairs developed like those located in the dorsal skin, and that the hair germ has an epidermal placode with a more centrally located dermal papilla. At 7–8 days postpartum, the hair bulb formed an angle, part of the dermal papilla was extrabulbar, the hair fibre was surrounded by an IRS and hairs among scales were generally medullated (Fig. 9c). The structure of the IRS showed a monostratified Henle, 1–2 stratified Huxley layers and the cuticle.

In adult hairs (body and tail), the immunolabelling for trichohyalin granules (0.2–2 µm in diameter) was intense and specifically localized in the IRS and, when present, in the medulla of all species of placental mammals studied in the present survey, including rat and rabbit (Fig. 9d,e). In tail hairs of adult rats, ultrastructural immunocytochemical analysis of the cornified Henle layer showed that labelling was diffuse (Fig. 10a). Gold particles were mainly located among the parallel filaments, and less frequently were associated with these filaments. In the Huxley layer (not yet cornified), intensely immunolabelled trichohyalin granules were either bounded or unbounded by a fibrous lining (Fig. 10a). The high-magnification analysis of trichohyalin granules showed some coarse, but irregular and ramified 8–10-nm-thick filaments forming an alveolate pattern (Fig. 10b). The gold particles were associated with but did not overlap these filaments. In controls no gold particles were seen.

image

Figure 10. a, trichohyalin immunolabelling in maturing IRS of rat tail hairs. The Henle layer is occupied by parallel keratin filaments with diffuse immunolabelling. Immunolabelled trichohyalin granules (arrowheads) are present in the Huxley layer, and are associated with scarce keratin bundles. 38 300×. b, detail of rat trichohyalin granule containing irregular coarse filaments (arrowhead) within a pale matrix. 90 000×. HE, Henle layer; HU, Huxley layer; k, keratin filaments/bundles; TH, trichohyalin immunolabelling.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

IRS structure in hairs

The present study has shown that the cytology and stratification of the IRS of marsupial and placental large hairs, presumably primary hairs, is essentially the same.

In contrast, the large hairs of monotremes (primaries) present a unique structural organization, not seen in hairs of marsupials and placentals. In monotreme primary hairs, all cells of the IRS have a similar appearance and cornify at the same time so that a specific Henle layer is absent. This undifferentiated stratification of the IRS inside the hair canal resembles that of the corneous layer of the epidermis (Alibardi & Maderson, 2003). Although our samples were limited to few areas of the body (ventral, tail and digits), our observations are supported by those of previous studies on different areas of the body of monotremes (Poulton, 1894; Spencer & Sweet, 1899).

The variation in the number of layers of the Huxley layer in placental–marsupial hairs or in the undifferentiated IRS of monotreme hairs is probably related to the papilla and hair size and shape. It is known that large hairs often have a thicker Huxley layer than that of thinner hairs (Priestley & Rudall, 1965; Straile, 1965; Montagna & Parakkal, 1974; Orwin, 1979). The Huxley and companion layers participate in the formation of the slippage plane along which the hair fibres move to exit through the epidermal surface (Straile, 1965, 1968; Stenn & Paus, 2001; Langbein et al. 2002). Cells of the Huxley layer that are located closer to the hair fibre seem to move upward at a faster rate than those of the companion layer (Straile, 1965, 1968). The latter was considered as the innermost layer of the ORS (Orwin, 1979), but its characteristics suggest that it is a specific compartment derived from the hair matrix, or even the external layer of the IRS (Winter et al. 1998; Langbein et al. 2002). The companion layer is an integral part of the slippage plane along which the hair fibres move externally (Stenn & Paus, 2001).

The upward movement of the hair fibre probably produces a drag on the less cornified cells of the Huxley layer, which are sandwiched between the IRS cuticle and the already cornified Henle layer. This dragging may determine the formation of a stratified Huxley layer because cells disrupted by the hair fibre movement can overlap, forming irregularly stratified planes. The Huxley layer probably functions as a cushion of various thickness between the hair fibre and the ORS.

The stratification of the Huxley layer may derive from the variation in the production of the cells from the matrix in the hair follicle, which determines a variation in the thickness of the Huxley layer along the IRS (‘complementation’; see Priestley & Rudall, 1965; Straile, 1965; Orwin, 1979). This produces a thicker Huxley layer (see Fig. 9c in rabbit hairs) where the hair fibre is thinner and maintains a uniform diameter of the hair follicle even when the hair fibre has an irregular shape. The mechanism by which the hair matrix can produce a variable number of cells for the formation of the Huxley layer is unknown.

IRS and the trichohyalin modality of cornification

Trichohyalin is the main protein of the cornified matrix of IRS and medulla of placental hairs (Rogers, 1985; Rothnagel & Rogers, 1986; O'Guin & Manabe, 1991; Rogers et al. 1991, 1999). The present study shows for the first time that trichohyalin of monotreme and marsupial hairs shares a cross-reactivity with that of placental mammals, and that the process of IRS cornification based on trichohyalin is alike in all mammals. Therefore, the antibody (Rothnagel & Rogers, 1986) recognizes epitope/s that are conserved across monotreme, marsupial and placental species. Probably this epitope has been present since the beginning of the molecular evolution of trichohyaline in therapsids, before the splitting between prothotherians and eutherians in the early Jurassic or even in the Triassic (McFarland et al. 1979).

The presence of trichohyalin-immunolabelling in few granules of cells external to the last cornified IRS layer (the companion layer) suggests that in hair follicles of monotremes the control of trichohyalin synthesis is less precise than in follicles of marsupial and placental hairs. It is also possible that in this transitional layer between IRS and ORS the synthesis of trichohyalin is rapidly but not completely suppressed. The present immunocytochemical study indicates the non-specialization of the more external part of the IRS of monotreme hairs. Whereas in marsupial and placental hairs only cells of the IRS produced trichohyalin, the presence of this protein has not been described in their companion layer. However, recent studies (Winter et al. 1998; Langbein et al. 2002) have shown that the characteristics of the companion layer suggest that it represents a different compartment of the hair follicle, perhaps even the more external layer of the IRS. No trichohyalin expression in these cells has been found (Rothnagel & Roop, 1995; Mahony et al. 1999).

The meshwork of 10–12-nm-thick filaments within the trichohyalin granules of monotreme and marsupial hairs is similar to that of placental hairs (Rothnagel & Rogers, 1986; Rogers et al. 1999; present study). These filaments probably correspond to the specific keratins found in IRS (Bawden et al. 2001; Langbein et al. 2002). During the process of cornification of the IRS in all mammalian species, the filaments organize themselves into parallel rows and the immunoreactivity for trichohyalin is lost. This general process probably occurs after trichohyalin has been dispersed among the filaments, as noted in the present study, and then cross-linked by the transglutaminase present in maturing IRS, or in desolving trichohyalin granules within the corneous material. Therefore, the dispersion of trichohyalin from the initial accumulating granules of IRS is a universal process in any mammalian hair, which progressively determines hardening of IRS cells.

The nuclear localization of transglutaminase immunolabelling is probably related to the formation of isopeptide bonds in the chromatin of terminal differentiating hair cells (Haake & Polakowska, 1993; Philpot & Paus, 1999). The transglutaminase labelling probably indicates terminal differentiation for IRS cells before full cornification of the IRS: this process also occurs in all mammalian species.

Cornification-based on trichohyaline and transglutaminase produces an IRS with a different consistency with respect to that of the hair fibre, based on special trichocytic keratins, on high sulphur proteins and disulphuric bonds formed by the action of sulphydril oxidases (Marshall et al. 1991; Langbein et al. 1999, 2001; Rogers et al. 1999; Hashimoto et al. 2000). The different consistency of the IRS vs. that of the cuticle of the hair fibre determines the detachment of the hair fibre from the IRS along the serrated interface of the fibre cuticle and the IRS cuticle. The two cuticles separate from each other after cell junctions are degraded by enzymes and by sebaceous secretions (Orwin, 1979; Stenn & Paus, 2001). Trichohyaline in IRS is enzymatically degraded in the sloughing zone of a hair (near the exit of the duct of the sebaceous gland) while sulphur-rich proteins of the cuticle are resistant to these lysis enzymes (Montagna & Parakkal, 1974; Orwin, 1979; O'Guin & Manabe, 1991; O'Guin et al. 1992).

The IRS also cornifies based on trichohyaline, producing a corneous material different from that of the corneous layer of the epidermis, whose cornification is based on filaggrin–keratin interactions (Resing & Dale, 1991). The different consistency between the IRS and corneous layer of the epidermis may determine shedding of the IRS from the corneous layer of the ORS epidermis in the upper hair canal. Degradation of the IRS allows the exit of the hair fibre through the skin surface without perforation of the epidermis. The latter process might otherwise lead to water loss, microbe penetration and inflammatory reactions. Microbes and cell fragments in the hair canal are removed by the outward movement of the scaled surface of the hair fibre cuticle. The serrated interlocking between the two cuticles resembles a ‘shedding layer’ formed within the epidermis, in which two cell strata separate from each another. This shedding layer provides the physiological exit of hairs. The IRS functions as an intra-epidermal shedding layer: by this process the continuity of the epidermis is maintained, and dehydratation, infection, and inflammation are avoided.

IRS represents the shedding layer of mammalian epidermis

The establishment of an IRS was probably essential during evolution of hairs. The IRS shows some analogies with shedding layers in amphibian, reptilian and avian epidermis (Alibardi, 2003). Shedding layers in vertebrate epidermis allow internal epidermal structures (e.g. scales, hairs, feathers) to exit on the surface without inducing water loss, inflammation and microbe penetration (Maderson et al. 1998). Shedding layers are formed by opposing surfaces within the epidermis where different modalities of cornification occur, and along which cell junctions are enzymatically degraded.

The complex stratification of the IRS and the retarded cornification of cells of Huxley and companion layers have some functional analogies with the shedding layer of lizard scales (Maderson et al. 1998; Alibardi, 2003). In the latter, a particular epidermal tissue, the lacunar layer, delays its cornification and becomes multistratified. The lacunar layer, as with the Huxley layer, is sandwiched between the cornified layers of scale epidermis, and is located in a ‘slippage plane’ along which scales grow laterally. It may function as a plastic cushion during distension of the new growing epidermis beneath the old epidermis prior to moult. Beneath the lacunar layer, a specialized shedding layer is formed.

Although shedding complexes have evolved independently along different lineages of amniotes (reptiles and mammals), it is relevant that intra-epidermal shedding layers are formed when new structures produced beneath the epidermis (scales or hairs) have to exit to the skin surface. By this mechanism, the continuity of the epidermis is maintained and infection/inflammation is avoided.

Comparative analysis of the formation of shedding layers in the epidermis of different vertebrates, together with the study on hair morphogenesis in primitive skins (monotremes and rodent tail), has produced a hypothesis on the biological role of the IRS and on possible trends followed during hair evolution (Spearman, 1964; Maderson, 1972b; see Fig. 11). In contrast to reptilian scales that possess an extended/laminar surface, hairs are compacted and narrow cylinders of corneous material. Hairs derive from morphogenetic processes that move epidermal cells (hair pegs) inside the dermis. The hypothesis considers the IRS as a tissue whose primary role was to allow the shedding/sloughing of the hair fibre after it was generated within the epidermis. With the lengthening of hair follicles into the dermis, and the need for hairs to elongate in order to exit on the skin surface, the IRS acquired roles of holding, guiding and addressing the hair fibre toward the epidermal surface.

image

Figure 11. Schematic drawing of hypothesized trends in hair evolution from a synapsid-scaled integument (see explanation in Discussion). 1a, 2a, 3a, 4a, 5a, 7a, 8a are dorsal views showing the reduction of the area of dermo-epidermal complexes (mesenchymal cells are rounded and the associated epidermis is stippled; the other areas are occupied by flat fibroblasts). The remaining figures represent longitudinal sections. Mesenchymal cells (round circles) and basement membrane + epidermis (thick black line) form dermo-epidermal complexes. 1-1a represent the original reptilian scales from which dermo-epidermal complexes have migrated toward the posterior (2-4) or anterior (5-8) part of the original scale. 9-9c represent monotreme/rodent hair development. 10-10d represent marsupial–placental hair development. The variations in shape of the dermal papilla + hair placode are tridimensionally illustrated in 11-14 (without mesenchymal cells for simplification), where 11a-14a represent their respective longitudinal sections. DEC, dermo-epidermal complex; DP, dermal papilla; F, flat fibroblasts; HI, hinge region; HP, hair placode; IRS, inner root sheath; M, mesenchymal cell; O, outer scale surface; P, hair peg; PH, proto-hair; PP, proximal, lenticular-shaped papilla (with bilateral symmetry in tridimensional view to form a flat/cone-shaped hair fibre); RH, round hair (see arrowed cross-section); RP, radial dermal papilla (with radial symmetry in tridimensional view to form a cylindrical hair fibre); SH, shield hair (see arrowed cross-section); SL, shedding layer. Dashed lines represent progressive reduction areas around the dermo-epidermal complexes.

Download figure to PowerPoint

IRS in relation to speculations on hair evolution

The integument of therapsid reptiles (mammalian-like reptiles) during the early Permian may have progressively reduced scales, so that the skin became smooth (Spearman, 1964; Findlay 1968; Maderson, 1972a, 2002). The resulting epidermis was probably lacking a granular layer but had a variably thick corneous layer. The progressive accumulation of histidin-rich proteins among bundles of keratin might have produced some microscopically visible keratohyalin granules that transformed the transitional layer into the granular layer (Spearman, 1964; Maderson, 2002; Alibardi, 2003; Alibardi & Maderson, 2003).

The flexible epidermis of therapsids and their loose dermis allowed the formation of epidermal papillae, a form of epidermal invagination that preceded the stage of hair morphogenesis or gland invagination (Alibardi & Maderson, 2003). A primitive skin patterned into scales explains the derivation of regularly distributed epidermal appendages, such as hairs (Fig. 11, 1-1a). In reptilian scales, most of the outer scale surface possibly resulted from the association of columnar epidermal cells with a special mesenchyme to form a ‘dermo-epidermal complex’ (DEC). Through the basement membrane the dermis in this region exchanges morphoregulatory molecules, which influence proliferation and differentiation of the associated epidermal cells.

In therapsids, the progressive posterior (Fig. 11, 2-4) or anterior (Fig. 11, 5-8) migration of DECs into the dermis and their reduction into smaller units determined the progressive loss of scales. DECs might have concentrated into deep (Fig. 11, 3-4) or superficial (Fig. 11, 6-7) hinge regions. The reduction of DECs might have induced the formation of dermal condensations and of a dermal papilla in deep (Fig. 11, 4) or superficial (Fig. 11, 7) areas of the skin. We favour the anterior migration of DECs (Fig. 11, 5-8), and the superficial formation of dermal papilla. This is suggested by hair morphogenesis in rodent tail (present study), and in the skin of monotremes (Poulton, 1894; Spencer & Sweet, 1899) (Fig. 11, 9). The anterior migration of DECs into the proximal side of a superficial hinge region may explain the backward-tilted orientation of hairs. The formation of a superficially located DEC (the hair placode) in the rostral side of hair pegs in rodents and monotreme skin, and the movement of the hair placode around the papilla supports this hypothesis (Fig. 11, 9-9c).

In therapsids, DECs were constructed of dermal papillae, which probably induced cell proliferation at the base of epidermal pegs. The result was the formation of rod-like outgrowths, and eventually hairs (Fig. 11, sequences 2-2c, 3-3c, 4-4c, 5-8c). The primitive hairs might have been produced from hinge regions: unfortunately no fossilized hairs have been found so far to support or deny this hypothesis. The presence of an open hair peg during morphogenesis of monotreme hairs (Poulton, 1894) supports the hypothesis of a closure and down-growth of the primitive hinge region among synapsid scales (Fig. 11, g-g2-g3). The observations by Poulton were, however, not confirmed by a later study (Spencer & Sweet, 1899).

A specific mechanism of epidermal shedding around the hair fibre was necessary to avoid the pricking of the epidermis, which might have induced inflammatory reactions and microbe invasion. The initial IRS was possibly made by a stratified corneous layer generated around the hair bulb, but no differentiation into Huxley and Henle layers was yet present, as in monotreme hairs (Poulton, 1894; Spencer & Sweet, 1899; present observations). IRS cells become cornified using a protein (trichohyalin) that produces a different corneous material with respect to that of the epidermis. Trichohyalin probably derived from molecular modifications of the ancient profilaggrin (Fietz et al. 1993). Profilaggrin and trichohyalin genes are closely localized in the same chromosomal ‘epidermal differentiation complex’, a cluster of genes coding for several epidermal proteins, which share some common sequences (Fietz et al. 1993; Taresa et al. 1997). Trichohyalin and filaggrin coexist in some epidermal areas or in mixed granules, suggesting some coexpression (O'Guin & Manabe, 1991; Hamilton et al. 1993; Manabe & O'Guin, 1994). Filaggrin and trichohyalin might have been initially mixed in the primitive, perhaps tubular, hair canal (Figs 11, 3b, 4 and 8-8c). IRS degradation favoured shedding along the serrated, faceted surfaces of the two cuticles, and from the corneous layer of the epidermis (O'Guin & Manabe, 1991).

The last step toward hair formation was the transformation of the proximal, lenticular-shaped papilla (as with that present in hair pegs of monotremes and rodent tail skin) into the typical centrally located, radial papilla of mammalian hairs (Fig. 11, h-h5, i-i5). In monotremes, but also in some eutherian hairs, the distal-most hair shaft is flat and forms a shield-hair while the proximal portion forms the typical cylindrical hair (Poulton, 1894; Spencer & Sweet, 1899; Wildman & Manby, 1938; Sokolov, 1982; see Fig. 11, 9-9c, sequence 11–14). The shield hair has a bilateral symmetry, perhaps derived from a papilla with a bilateral symmetry (Fig. 11, 11-11a). Variations in the shape of the papilla and associated hair placode are illustrated in Fig. 11, 11-14, where the tridimensional aspect (Fig. 11, 14) is shown together with the respective longitudinal sections (Fig. 11, 11a-14a).

Initially, the hair placode is localized on the proximal side of the hair peg, has a lenticular shape and a bilateral symmetry (see the axes of symmetry in Fig. 11, 11). Cell multiplication in the hair placode determines its expansion to surround the dermal papilla, which is eventually ‘invaginated’ into the hair peg. From the proliferation of cells of a bilaterally symmetric hair placode, a coniform and flat hair fibre is initially produced (Fig. 11, 12-13). The forming hair fibre is rapidly turned into a cylindrical rod as the papilla assumes a radial symmetry (Fig. 11, 13-14). When the radial papilla is formed, the hair fibre changes from an ovoidal/flat shape to a cylindrical shape (Fig. 11, 14-14a). This process probably produces a shield-hair, a type of guard hair present in monotremes (Poulton, 1894; Wildman & Manby, 1938). It is possible that shield-hairs were also common in theromorphs, where a similar morphogenetic mechanism of papilla transformation might have occurred. Further studies on hair morphogenesis in monotremes, coupled with the study of fossilized synapsid skin, will produce more information on the evolution of hairs.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This study was partially supported by a University of Bologna and from a Borsa Ghigi Award from the Academy of Science in Bologna, and by self-support. Tissues were kindly supplied by Drs S. Nicholls (University of Tasmania), J. Joss (Macquarie University), T. Grant and M. Beal (University of New South Wales), and F. Geiser (University of New England, Armidale). Rodent material was kindly provided by Mr M. Ghidotti (University of Padua, Italy) and Dr E. Ciani (University of Bologna, Italy). Figure 11 was skilfully drawn in Corel Draw by Mr C. Friso and Mr R. Mazzaro (University of Padua, Italy). Mr J. Crosby (University of Auckland, NZ) commented on an earlier version of the manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Alibardi L (2003) Adaptation to land: the skin of reptiles in comparison to that of amphibians and endotherm amniotes. J. Exp. Zool. (Mol. Dev. Evol.) 298B, 1241.
  • Alibardi L, Maderson PFA (2003) Distribution of keratin and associated proteins in the epidermis of monotreme, marsupial, and placental mammals. J. Morphol. 258, 4966.
  • Barbour RA (1977) Anatomy of marsupials. In The Biology of Marsupials (eds StonehouseB, GilmoreD), pp. 237246. Baltimore: University Park Press.
  • Bawden CS, McLaughlan C, Nesci A, Rogers G (2001) A unique type I keratin intermediate filament gene family is abundantly expressed in the inner root sheaths of sheep and human hair follicles. J. Invest. Dermatol. 116, 157166.
  • Bolliger A, Hardy MH (1944) The sternal integument of Trichosurus vulpecula. J. Royal. Soc. NSW 78, 122133.
  • Fietz MJ, Presland RB, Rogers GE (1993) Analysis of sheep trichohyalin gene: potential structural and calcium-binding roles of trichohyalin in the hair follicle. J. Cell Biol. 121, 855865.
  • Findlay GH (1970) The role of the skin in the origin of mammals. South Afr. J. Sci. 66, 277283.
  • Gibbs HF (1938) A study of the development of the skin and hair of the Australian opossum, Trichurus vulpecula. Proc. Zool. Soc. Lond. (Series B) 108, 611648.
  • Griffiths M (1978) The Biology of Monotremes. New York: Academic Press.
  • Haake AR, Polakowska RR (1993) Cell death by apoptosis in epidermal biology. J. Invest. Dermatol. 101, 107112.
  • Hamilton EH, Payne RE, O'Keefe EJ (1993) Trichohyalin: presence in the granular layer and stratum corneum of normal human epidermis. J. Invest. Dermatol. 96, 666672.
  • Hardy MH (1992) The secret life of the hair follicle. Trends Genet. 8, 5561.
  • Hashimoto Y, Suga Y, Matsuba S, et al. (2000) Immunohistochemical localization of sulfhydryl oxidase correlates with disulfide crosslinking in the upper epidermis of rat skin. Arch. Dermatol. Res. 292, 570572.
  • Jahoda CAB, Mauger A, Bard S, Sengel P (1992) Changes in fibronectin, laminin and type IV collagen distribution relate to basement membrane restructuring during the rat vibrissa follicle hair growth cycle. J. Anat. 181, 4760.
  • Langbein L, Rogers MA, Winter H, et al. (1999) The catalog of human hair keratins. I. Expression of nine type I members in the hair follicle. J. Biol. Chem. 274, 1987419884.
  • Langbein L, Rogers MA, Winter H, et al. (2001) The catalog of human hair keratins. II. Expression of the sis type II members in the hair follicle and the combined catalog of human type I and type II keratins. J. Biol. Chem. 276, 3512335132.
  • Langbein L, Rogers MA, Praetzel S, Aoki N, Winter H, Schweizer J (2002) A novel epithelial keratin, hK6irs1, is expressed differentially in all layers of the inner root sheath, including specialised Huxley cells (Flugelzellen) of the human hair follicle. J. Invest. Dermatol. 118, 789799.
  • Lyne AG (1957) The development and replacement of pelage hairs in the bandicoot. Parameles nasuta Geoffroy (Marsupialia: Paramelidae). Aust. J. Biol. Sci. 10, 197221.
  • Lyne AG (1970) The development of hair follicles in the marsupial Trichosurus vulpecula. Aust. J. Biol. Sci. 23, 12411253.
  • Lyne AG, Henrickson RC, Hollis DE (1970) Development of the epidermis of the marsupial Trichosurus vulpecula. Aust. J. Biol. Sci. 23, 10671075.
  • Maderson PFA (1972) When? Why? and How? Some speculations on the evolution of the vertebrate integument. Am. Zool. 12, 159171.
  • Maderson PFA, Rabinowitz T, Tandler B, Alibardi L (1998) Ultrastructural contributions to an understanding of the cellular mechanisms involved in lizard skin shedding with comments on the function and evolution of a unique lepidosaurian phenomenon. J. Morphol. 236, 124.
  • Maderson PFA (2002) Mammalian skin evolution: a revaluation. Exp. Dermatol. 12, 15.
  • Mahony D, Karunaratne S, Rothnagel JA (1999) The companion layer and outer root sheath of the anagen follicle. Exp. Dermatol. 8, 329331.
  • Manabe M, O'Guin WM (1994) Existence of trichohyalin-keratohyalin hybrid granules: co-localization of two major intermediate filament-associated proteins in non-follicular epithelia. Differentiation 58, 6575.
  • Marshall RC, Orwin DFG, Gillespie JM (1991) Structure and biochemistry of mammalian hard keratins. Electron Microsc. Res. 4, 4783.
  • McFarland WN, Pough HF, Cade TJ, Heiser JB (1979) Vertebrate Life. New York: Collier Macmillan Publishing Co, Inc.
  • Millar SE (2002) Molecular mechanisms regulating hair follicle development. J. Invest. Dermatol. 118, 216225.
  • Montagna W, Parakkal PF (1974) The Strcture and Function of Skin. the Pilary Apparatus, 3rd edn. New York: Academic Press.
  • Mykytowyctz R, Nay T (1987) Studies of the cutaneous glands and hair follicles of some species of macropodidae. CSIRO Wildl. Res. 9, 200217.
  • O'Guin MW, Manabe M (1991) The role of trichohyalin in hair follicle differentiation and its expression in nonfollicular epithelia. In The Molecular and Structural Biology of Hair (eds Stenn KS, Messenger AG, Baden, HP). Ann. NY Acad. Sci. 642, 5163
  • O'Guin MW, Sun TT, Manabe M (1992) Interaction of trichohyalin with intermediate filaments: three immunologically defined stages of trichohyalin maturation. J. Invest. Dermatol. 98, 2432.
  • Orwin DFG (1979) The cytology and cytochemistry of the wool follicle. Int. Rev. Cytol. 60, 331374.
  • Philpott M, Paus R (1999) Principles of hair follicle morphogenesis. In Molecular Basis of Epithelial Appendages Formation (ed. Chuong, MC), pp. 75109. Georgetown, TX: Landes-Bioscience, USA.
  • Poulton EB (1894) The structure of the bill and hairs of Ornithorhynchus paradoxus. Quart. J. Micr. Sci. 36, 143199.
  • Priestley GC, Rudall KM (1965) Modifications in the Huxley layer associated with changes in fibre diameter and output. In Biology of the Skin and Hair Growth (eds LyneAG, ShortBF), pp. 165170. Sydney: Angus and Robertson.
  • Resing KA, Dale BA (1991) Proteins of keratohyalin. In Physiology, Biochemistry and Molecular Biology of the Skin, Vol. 1, 2nd edned. GoldsmithLA), pp. 148167. New York: Oxford University Press.
  • Rogers GE (1985) Genes for hair and avian keratins. In Intermediate Filaments (eds Wang E, Fishman D, Liem RKH, Sun TT). Ann. NY Acad. Sci. 455, 403425.
  • Rogers GE, Fietz MJ, Fratini A (1991) Trichohyalin and matrix proteins. In The Molecular and Structural Biology of Hair (eds StennKS, MessengerAG, BadenHP). Ann. NY Acad. Sci. 642, 6481.
  • Rogers G, Dunn S, Powell B (1999) Late events and the regulation of keratinocyte differentiation in hair and feather follicles. In Molecular Basis of Epithelial Appendages and Morphogenesis (ed. ChuongCM), pp. 315338. Austin, TX: Molecular Biology Intelligence Unit. RG Landes Co.
  • Rothnagel JA, Rogers GE (1986) Trichohyalin, an intermediate filament-associated protein of the hair follicle. J. Cell Biol. 102, 14191429.
  • Rothnagel JA, Roop DR (1995) Hair follicle companion layer: reacquanting an old friend. J. Invest. Dermatol. 104, 42S43S.
  • Sokolov VE (1982) Mammalian Skin. Berkeley: University of Califrornia Press.
  • Spearman RIC (1964) The evolution of mammalian keratinized structures. In The Mammalian Epidermis and its Derivative (ed. EblingFJ), Symposium n. 12, pp. 6781. London: Zoological Society.
  • Spencer B, Sweet G (1899) The structure and development of the hairs of monotreme and marsupials. Quart. J. Microsc. Sci. 36, 549588.
  • Steinert PM, Parry DAD, Marekov LN (2003) Trichohyalin mechanically strenthens the hair follicle. J. Biol. Chem. 278, 4140941419.
  • Stenn KS, Parimoo S, Prouty SM (1999) Growth of the hair follicle: a cycling and regenerating biological system. In Molecular Basis of Epithelial Appendages Formation (ed. ChuongMC), pp. 111130. Georgetown, TX: Landes Biosciences.
  • Stenn KS, Paus R (2001) Control of hair follicle cycling. Physiol. Rev. 81, 449494.
  • Stonehouse B (1977) Introduction: the Marsupials. In The Biology of Marsupials (eds StonehouseB, GilmoreD), pp. 15. Baltimore: University Park Press.
  • Straile WE (1965) Root sheath-dermal papilla relationships and the control of hair growth. In Biology of the Skin and Hair Growth (eds LyneAG, ShortBF), pp. 3557. Sydney: Angus and Robertson.
  • Straile WE (1968) Possible function of the external root sheath during growth of the hair follicle. J. Exp. Zool. 150, 207224.
  • Taresa E, Marekov LN, Andreoli J, et al. (1997) The fate of trichohyalin. J. Biol. Chem. 272, 2789327901.
  • Wildman AB, Manby J (1938) The hairs of monotremata, with special reference to their cuticular scale pattern. Trans. Royal Soc. Edinburgh 54 (part II), 333356.
  • Winter H, Langbein L, Praetzel S, et al. (1998) A novel human type II cytokeratin, K6hf, specifically expressed in the companion layer of the hair follicle. J. Invest. Dermatol. 111, 955962.
  • Woods JL, Orwin DFG (1982) The cytology of cuticle scale pattern formation in the wool follicle. J. Ultrastruct. Res. 80, 230242.