Expression of beta-keratin mRNAs and proline uptake in epidermal cells of growing scales and pad lamellae of gecko lizards

Abstract

Beta-keratins form a large part of the proteins contained in the hard beta layer of reptilian scales. The expression of genes encoding glycine–proline-rich beta-keratins in normal and regenerating epidermis of two species of gecko lizards has been studied by in situ hybridization. The probes localize mRNAs in differentiating oberhautchen and beta cells of growing scales and in modified scales, termed pad lamellae, on the digits of gecko lizards. In situ localization at the ultrastructural level shows clusters of gold particles in the cytoplasm among beta-keratin filaments of oberhautchen and beta cells. They are also present in the differentiating elongation or setae of oberhautchen cells present in pad lamellae. Setae allow geckos to adhere and climb vertical surfaces. Oberhautchen and beta cells also incorporate tritiated proline. The fine localization of the beta-keratin mRNAs and the uptake of proline confirms the biomolecular data that identified glycine–proline-rich beta-keratin in differentiating beta cells of gecko epidermis. The present study also shows the presence of differentiating and metabolically active cells in both inner and outer oberhautchen/beta cells at the base of the outer setae localized at the tip of pad lamellae. The addition of new beta and alpha cells to the corneous layer near the tip of the outer setae explains the anterior movement of the setae along the apical free-margin of pad lamellae. The rapid replacement of setae ensures the continuous usage of the gecko's adhesive devices, the pad lamellae, during most of their active life.

Introduction

The epidermis of geckoes and other lizards comprises an outer epidermal generation that is sloughed during moulting, while an inner epidermal generation forms beneath to replace the lost outer generation (Maderson, 1966, 1970; Hiller, 1972). As in the epidermis of other lizards and snakes, the epidermal generation of scales in geckoes comprises six epidermal layers called oberhautchen, beta, mesos, alpha, lacunar and clear layer (Maderson, 1985; Maderson et al. 1998). Oberhautchen and beta layers synthesize beta-keratin while the other layers produce mainly alpha-keratin. Beta-keratin is the hard form of the corneous material of reptilian scales, while alpha-keratin forms the softer part of the corneous layer of scales. The last layer to be formed in the outer generation, the clear layer, is joined by desmosomes with the first layer of the inner generation, the oberhautchen. This temporary union between epidermal generations forms the shedding complex, but desmosomes are eventually degraded and the two layers detach one from another, resulting in shedding.

Another characteristic of the oberhautchen layer of the epidermis of normal (generalized) body scales is the ability to produce 1–3-µm-long spinules 1–2 µm in diameter. On the digital pads of geckoes and of other lizards, a modified type of scales used by geckoes to climb vertical surfaces, the spinulae of the oberhautchen layer grow over 30 µm long, forming long bristles termed setae (Maderson, 1964, 1970; Hiller, 1972; Peterson & Williams, 1981; Alibardi, 1997; Rosenberg et al. 1998). Setae also form in the specialized caudal scales of some geckoes, and these provide caudal prehension (Bauer, 1998). During their outward growth, setae are enwrapped by the cytoplasm of clear cells, to which they are joined by cell junctions. However, the latter are degraded before shedding so that the cornified clear layer is lost with the other epidermal layers during moulting.

Setae are largely composed of beta-keratins of 10–18 kDa but also contain alpha-keratins and some minor proteins (Thorpe & Giddings, 1981; Alibardi, 2003b; Alibardi & Toni, 2005; Rizzo et al. 2006). Beta-keratins contribute to the resistance of setae and form long bundles orientated along the main axis of these elongated bristles. The molecular characteristics of some beta-keratins in gecko setae have recently been determined, including the nucleotide and deduced amino acid sequences (Dalla Valle et al. 2007). This study indicates that multiple genes, probably arranged in a head-to-tail linear sequence, code for basic glycine–proline–serine-rich proteins with 169–191 amino acids and with a molecular weight varying between 16 and 18 kDa. By using RT-PCR, the study has shown that both normal scales and pad lamellae of the gecko Hemidactylus turcicus expressed mRNAs for these proteins. However, the specific tissue localization in pad lamellae was not analysed.

In the present study we studied the tissue expression of genes for the sequenced glycine–proline–serine-rich proteins at both light and electron microscopic level in the differentiating epidermis of geckoes. The aim of the study was to determine the cellular localization of the sequenced mRNAs in relation to beta-keratin synthesis. We utilized the epidermis of regenerating tails and pad lamellae in geckoes in order to obtain the epidermis during a differentiating phase when beta-keratin cells are produced. The present study contributes to our knowledge of cell distribution and organization of hard corneous proteins of known composition and sequence in the epidermis of reptiles. The detailed knowledge of these proteins will allow the eventual comparison of their sequences and genomic organization, which in turn will lead to determination of the phylogenetic relationship of beta-keratins with hard keratins in other amniotes.

Materials and methods

Tissue collection and fixation

Three adults of the gecko Hemidactylus turcicus and four adults of the larger species of the gecko Tarentola mauritanica were used in the present in situ hybridization study. The tail was amputated by pulling it to induce its release by autotomy, a natural mechanism of auto-amputation of the tail in geckos. The amputated animals were left to regenerate their tails at 25–33 °C, following which 3–5 mm of skin was collected from the new tail when it measured between 0.6 and 1.2 cm in length. The epidermis during this period of 3–4 weeks post-amputation begins to regenerate scales in which the differentiation of new epidermal layers, including the beta layer, is triggered (Alibardi, 2003).

Three other individuals of H. turcicus received an intraperitoneal injection of tritiated proline (4–5 µCi g⁻¹ body weight) and were killed 4 h later as previously detailed (Alibardi & Toni, 2005).

Tissues from regenerating tails or from normal digits (including digit pad lamellae) were fixed in 2.5% glurataldehyde in 0.12 m phosphate buffer for 5–6 h at 4 °C, post-fixed in 2% OsO₄ for 1 h and then stained in 2% uranile acetate for 2 h. After dehydration tissues were embedded in Durcupan resin at 60 °C for 24–36 h. Tissues were sectioned using an ultramicrotome and thin sections of 40–80 nm were collected with copper grids. Sections on grids were routinely stained with uranyl acetate and lead citrate for observation under a transmission electron microscope. Other samples containing similar tissues were fixed for 5 h in freshly made 4% paraformaldehyde solution in 0.1 m phosphate buffer, pH 7.4, dehydrated in ethanol and embedded in either wax or in Bioacryl resin at 0–4 °C under UV light (Scala et al. 1992).

Some tissues were sectioned with a rotary microtome at 6–7 µm for in situ hybridization staining. Other samples were sectioned using an ultramicrotome in order to collect 1–4-µm semithin sections for light microscopy and immunocytochemistry. Other 40–90-nm thin sections were collected on nickel grids for ultrastructural immunocytochemical and in situ hybridization study.

Tissues injected with tritiated proline were fixed in glutaraldehyde as above, sectioned and coated with an Ilford Nuclear Emulsion (K5) for 5–9 weeks. Sections were developed in Kodak D19 and fixed in Ilford fixer, washed in water and lightly stained with 0.5% toluidine blue or left unstained for grain detection (autoradiography).

Ultrastructural immunocytochemistry

Non-specific antigenic sites were blocked by incubating the sections attached to grids for 10 min with 5% normal goat serum in 0.05 m Tris-buffered saline at pH 7.6 with 2% cold water fish gelatin. The sections were incubated overnight at 4 °C in the primary antibody while the negative controls were incubated with pre-immune sera or buffer only.

An antibody against chick scale beta-keratins was used, called beta-1 keratin (produced in rabbit against a chick scale beta-keratin, a generous gift of Dr R. H. Sawyer, University of South Carolina, Columbia, USA; see Sawyer et al. 2000). The primary antibody was diluted 1 : 200 in buffer and grids with the sections were incubated overnight at 0–4 °C. Grids were rinsed, and incubated for 1 h at room temperature with the above Tris buffer containing a 0.1% TritonX-100 and 1% cold-water fish gelatin (to block non-specific antigenic sites). Controls were incubated with the buffer only or the preserum. After rinsing in buffer, grids were incubated for 1 h at room temperature with the secondary anti-rabbit antibody conjugated to 10-nm gold particles (Sigma or Chemicon, Italy) at 1 : 40 dilution in the same buffer. After rinsing in the buffer, grids were washed in distilled water, stained for 7 min in 2% uranyl acetate, rinsed in buffer, then in distilled water, and dried for observation under an electron microscope.

Light and ultrastructural in situ hybridization

Digoxigenin-labelled probes were prepared as previously reported (Dalla Valle et al. 2007) using recombinant plasmids containing the whole coding region of beta-keratins from the geckos T. mauritanica and H. turcicus (Fig. 1). The plasmids were linearized by restriction cleavage and used as a template. The cRNA transcripts were digoxigenin-labelled by in vitro transcription using a DIG RNA labelling kit (Roche Diagnostics, Milan, Italy) and T7 and SP6 polymerase.

Fig. 1.

Nucleotide sequence of the cDNA for beta-keratins isolated from the epidermis of T. mauritanica and H. turcicus in phase of tail regeneration (bars, 5 mm). The nucleotide primers (Dalla Valle et al. 2005, 2007) used to amplify the complete coding region are indicated in yellow while the in-frame start and stop codons are indicated in red. The digoxigenin probe was done on the whole coding region of both species (between ATG and TAA) of which the coded name and Accession Number are indicated.

In situ hybridization using the the beta-keratin probes was conducted as previously reported (see details in Dalla Valle et al. 2005, 2007) on tissues of both T. mauritanica (light microscopy) and H. turcicus (light and transmission electron microscopy).

Briefly, this involved fixing 2–3-mm-long pieces of regenerating skin in 4% paraformaldehyde as above, and embedding them in paraffin. Using a rotary microtome, 5–7-µm sections were collected on slides and left on a hot plate at about 40 °C to dry for about 6 h. Sections were dewaxed, hydrated and incubated at room temperature for 10 min with Proteinase-K (10 µg mL⁻¹) in phosphate-buffered saline (PBS), preincubated in hybridizing solution [50% formammide, 4× saline citrate solution (SSC), 0.1% Tween-20, 50 mg mL⁻¹ tRNA, 100 g mL⁻¹ 50 mg mL⁻¹, 10 mm EDTA, 50 mm mL⁻¹ heparin, 0.5% Blocking reagent from Roche Diagnostics] at 60 °C for 45 min, and hybridized overnight at 60 °C using the antisense-cRNA probe. Controls used sense RNA probes or omitted the probes from the hybridization solution. Hybridization was done by incubating sections at 60 °C with 1–2 ng mL⁻¹ of digoxigenin-labelled probe. Sections were then rinsed for 10 min each change with increasing stringent standard SSC of decreasing concentration (2×, 0.5×, 0.2×, 0.1× SSC) and finally rinsed in phosphate-saline-Tween buffer (PBT buffer). In order to reveal the hybridization digoxygenated complex, sections were incubated for 3 h at room temperature with an anti-digoxigenin alkaline phosphatase-conjugated antibody (Roche Diagnostics) diluted 1 : 500 in PBT buffer. Detection utilized PBT buffer containing 4-nitroblue tetrazolium chloride (NBT) and 5-bromo-4 chloro-3-indolyl phosphate (BCIP) as substrates (Roche Diagnostics).

The technique above was used on sections collected on nickel grids for the ultrastructural study on scales and pad lamellae of H. turcicus with the following modifications. Thin sections of tissues embedded in Bioacryl resin were incubated withouth proteinase-K treatment. Hybridization was carried out for 4–5 h at 60 °C using the antisense RNA probe. As control, a sense probe, used at the same concentration as the antisense probe, was utilized. Grids with the attached sections were rinsed in the above post-hybridizing solutions of increasing stringency, and then grids were incubated at room temperature for 20 min in 0.05 mTris, pH 7.6, with 2% bovine serum albumin. This was followed by incubation for 3 h at room temperature or overnight at 0–4 °C in the primary antibody (mouse anti-digoxigenin, Roche Diagnostics) diluted at 1 : 500 in the buffer. After repeated rinsing in the buffer, grids were incubated with the secondary antibody (anti-mouse IgG conjugated with gold particles of 5 nm in diameter) for 1 h at room temperature. After rinsing in buffer and in distilled water, grids were stained for 5–7 min in 2% uranyl acetate, dried and observed under an Hitachi 600 electron microscope.

Results

Epidermis of normal scales and pad lamellae

The normal epidermis of gecko scales in resting phase consisted of a thin corneous layer over a multistratified, living epidermis of 2–4 cell layers above the germinal layer (Fig. 2). The ultrastructure of the epidermis of the outer scale surface showed the short spinulae of the oberhautchen layer merged with a pale beta layer, and underneath the intermediate thin mesos region, followed by compacted and still differentiating alpha cells. The latter were located above undifferentiated keratinocytes, whose tonofilaments were aggregated to form a dense corneous mass. Mature alpha cells appeared more or less electron-dense. In addition, pale vesicles, representing lipid droplets or vesicles extracted by the solvent used in the process of embedding, were seen among the tonofilaments.

Fig. 2.

Electron micrograph of the epidermis of outer scale of H. turcicus in post-shedding phase (inset, the arrow indicates the corneous layer covering the scale; bar 10 µm). Externally oberhautchen spinulae are seen (s) followed by the merged oberhautchen and beta layer (oβ), the thin mesos region (m), mature alpha layer (α), presumptive alpha cells (pα), suprabasal (sb), and basal epidermal cells (b). Bar, 2 µm.

In the digit pads of both species, the scales were modified and numerous long setae were seen in the outer part of the corneous layer (Figs 2 and 3). The detailed cytological analysis of epidermal layers of pad lamellae showed that the beta layer merged with the oberhautchen cells (bearing the setae) forming a thin corneous layer sustaining the setae. The latter varied in length, from short (a few µm) near the hinge region of the lamella to long setae (over 30 µm) at the tip of the lamella. In some samples it was possible to detect the formation of the two epidermal generations, with both outer setae (to be replaced at shedding) and differentiating inner setae (Figs 3A and 4A). The latter derived from the elongation of the apical part of the polygonal-shaped cells in the inner oberhautchen layer, which were followed beneath by various layers of differentiating beta cells (Figs 3B and 4B). Two or more setae were apparently produced from each oberhautchen cell. The outer setae occupied most of the scale surface in post-shedding pad lamellae or in early renewal epidermis. In the latter stage, the inner setal generation was at the beginning of differentiation and short setae were visible (Fig. 3A). As the inner generation was in mid or late differentiation, the length of setae, in particular, appeared similar to that of outer setae, while the latter appeared mainly localized toward the tip of pad lamellae (Figs 4A and 5A). It appeared that the setae had moved toward the tip to make space for those of the inner generation. In fact, the advanced outer setae rested upon a corneous outgrowth by the tip of pad lamellae, the free margin (Fig. 5A). The accumulation of folded corneous material was seen on the inner side of the pad lamella, and a similar accumulation of corneous layers was not observed in normal scales.

Fig. 3.

Longitudinal section of pad lamella of T. mauritanica in renewal phase stained with toluidine blue (A,B) and ultrastructural detail of a forming seta (C). (A) The two epidermal generations are visible (the arrow indicates the sectioning detachment of the outer generation). Bar, 20 µm. (B) Detail of inner seta produced from oberhautchen cells (arrows) and embedded (arrowheads) in the cytoplasm of clear cells. Differentiating beta-cells derive from the basal layer. Bar, 15 µm. (C) Ultrastructural detail of a forming seta of the inner oberhautchen with beta-1-immunolabelled keratin bundles (arrows). The surrounding cytoplasm of clear cells is immunonegative. Bar, 250 nm. Legends: ba, basal (germinal) layer; be, differentiating cells of the beta layer; bs, base of seta; cl, clear cell cytoplasm; de, dermis; fα, forming outer alpha layer; fβ, forming beta layer; fs, forming seta; i, thinner epidermis of inner scale surface; is, forming inner setae; ob, oberhautchen; os, outer setae; t, tip of the lamella.

Fig. 4.

Longitudinal section of apical part of a pad lamella of H. turcicus in renewal phase stained with toluidine blue (A,B) with an ultrastructural detail of forming seta (C). (A) The two epidermal generations are visible (due to processing, the outer setae with the oberhautchen layer are detached from the following layers). Arrows indicate the outer clear layer and arrowheads indicate cells of the inner oberhautchen layer. Bar, 15 µm. (B) Detail of inner setae, which are derived from oberhautchen cells (arrowheads). The setae reach cells of the clear layer (arrows). Bar, 15 µm. (C) Ultrastructural detail of forming seta (fs) of inner oberhautchen, which shows beta-1-immunolabelled keratin bundles (arrows). The surrounding cytoplasm of clear cells (cl) is immunonegative. Bar, 250 nm. Legends: be, differentiating cells of the beta layer; cl, clear cell cytoplasm; de, dermis; e, epidermis; fα, forming outer alpha layer; i, thinner epidermis of inner scale surface; is, forming inner setae; ob, oberhautchen; os, outer setae; t, tip of the lamella.

Fig. 5.

Histological section of pad lamella and normal scale in a digit (A) and autoradiographic sections after tritiated proline injection (B–F) in H. turcicus. (A) Tip of the lamella shows the corneous free margin (arrow) that sustains the distal outer setae. Residual corneous layers cover the inner side of the pad (arrowhead). The inner setae have grown to almost identical dimension of outer setae, which are concentrated by the free margin at the tip of the scale. Bar, 20 µm. (B) Silver grains mainly localized over the forming inner setae and oberhautchen in early renewal stage (stage 3, see Maderson et al. 1998). Long exposure (9 weeks). Bar, 20 µm. (C) Silver grains are mainly concentrated over outer oberhautchen/beta layer by the tip of scale (arrow) in mid renewal stage (stage 4). Also, strong labelling is present especially by the forming tip (arrowhead) of the inner oberhautchen of the inner generation. Bar, 20 µm. (D) Another section at stage 4 showing the distal location of outer setae and the labelling in the distal outer (arrow) and inner (arrowhead) oberhautchen/beta layer. Also, differentiating beta cells are labelled (double arrowhead). Bar, 20 µm. (E) Apical localization of labelling (arrow) in a pad lamella kept with a short autoradiographic exposure (5 weeks). Bar, 20 µm. (F) Detail of silver grains distribution over oberhaucthen and beta cells of the inner generation (short exposure, 5 weeks). Bar, 10 µm. (G) Apical part of a forming seta showing separated bundles of beta-keratin (arrows) among non-keratinized cytoplasm. Bar, 1 µm. (H) Detail of bundles of beta-keratin (arrows) forming the spatular termination at the apex of a seta tip surrounded by clear cell cytoplasm (arrowheads indicate the cell membrane). Bar, 250 nm. (I) Tip of seta. Apical branching of beta-keratin bundles (arrows) among which is present the cytoplasm of clear cells (separated by still visible cell membranes (arrowheads). Bar, 250 nm. Legends: be, differentiating beta cells; c, cytoplasm of clear cell; de, dermis; nk, non-keratinized cytoplasm; ns, normal scale; is, inner setae (forming); ob, oberhautchen; os, outer setae; s, setae; t, tip of pad lamella. Dashes underline the basal layer of the epidermis.

The forming setae of the inner generation contacted the cells in the clear layer of the outer epidermal generation and were surrounded by the cytoplasm of clear cells over their entire length. After immunostaining with the Beta-1 antibody only the beta layer and the oberhautchen with its spinulae or setae were immunoreactive (Figs 3C and 4C). The gold particles labelled the long keratin bundles, which were incorporated within the growing inner setae.

Autoradiographic observations on pad lamellae and setae apical branching

Setae were variably labelled (silver grains) with tritiated proline according to their stage of maturation. Long and mature setae were unlabelled in both the outer (mature) and the inner (almost mature) setae generation (Fig. 5B,C) whereas at the beginning of their formation in the inner generation, growing setae were labelled (Fig. 5B–D). In addition, the oberhautchen/beta layer on which the setae rest, in both the inner and especially the outer generation in contact with the outer setae, appeared labelled (Fig. 2C–D). The higher intensity of labelling was often observed as being localized in areas toward the tip of the pad lamellae (Fig. 5C–E). Not only was the tip highly labelled, however; the oberhautchen/beta layer, localized at the beginning of the inner scale surface beneath the tip, also appeared intensely labelled, suggesting a metabolic connection with the oberhautchen present in the tip of the lamella.

The differentiating and hypertrophic oberhautchen cells and the beta cells beneath the inner generation were also seen to be incorporating tritiated proline (Fig. 5F).

Our ultrastructural examination of these tissues allowed us to study the process of branching of the apical part of setae with their spatulae. The apical part of growing setae of the inner generation contained rims of non-keratinized cytoplasm among the narrow terminals made of beta-keratin bundles (Fig. 5G). The latter formed curved bundles of 100 nm in diameter or less (forming spatulae), apparently surrounded by the cytoplasm of clear cells (Fig. 5H,I). The apical branching appeared to be derived from the isolation of terminal bundles of beta-keratin present in setae after the cytoplasm of clear cells disappears by degeneration of detachment at shedding.

Regenerating scales

Both the normal and the neogenic scales of regenerating tail produced a specific immunolabelled beta layer using the Beta-1 antibody (data not shown; see Alibardi, 2003). As previously reported (Liu & Maneely, 1969; Alibardi & Toni, 2006), the formation of scales in the regenerating tail epidermis of animals 3–5 weeks post-amputation occurred in a proximo-distal gradient (Figs 6 and 7A). The whole sequence of scale morphogenesis was seen moving from the apical areas of the regenerating tail to proximal areas near the tail stump (Figs 6A,B and 7A). Within this period, neogenic scales histologically similar to normal scales were formed in proximal regions, near the stump of the original tail (Fig. 6C). These scales appeared imbricate, like those of normal epidermis, although they were of smaller size. Further details of scale neogenesis in geckos skin have been previously reported (Dalla Valle et al. 2007).

Fig. 6.

Regenerating tail at 5 weeks post-amputation in T. mauritanica (A) with the general histology of regenerating scales (B,C) embedded in wax and stained with toluidine blue. In D–G are sections of regenerating scales of H. turcicus at 4 weeks post-amputation, embedded in Bioacryl resin and stained with toluidine blue. (A) Regenerating tail at 5 weeks post-amputation. Bar, 3 mm. (B) Longitudinal section of regenerating tail of about 4 weeks post-amputation with regenerating scales (arrows). Bar, 100 µm. (C) Completely regenerated scales of proximal regions of regenerating tail 5 weeks post-amputation. The arrow indicates shedding of the superficial wound epidermis. Bar, 40 µm. (D) Epidermal pegs penetrating into the dermis and covered by lacunar and wound epidermis in apical areas of regenerating tail. The middle region (arrows) shows elongated cells representing clear and oberhautchen cells. Bar, 15 µm. (E) Elongation of pegs localized closer to normal epidermis. Differentiating clear and oberhautchen cells (arrows), and beta cells (arrowheads) are seen in the middle of pegs. Bar, 15 µm. (F) Further stage of differentiation in proximal scales with packed line (arrow) indicating the spinules of the oberhautchen layer. Two to three layers of elongated beta cells are forming (arrowheads) on the outer scale surface (the inner side is on the right). The clear layer contacts externally the wound epidermis. Bar, 20 µm. (G) Neo-regenerated scales among two hinge regions. The compact oberhautchen layer (arrow) in the outer scale surface delimits the scale outline from the wound epidermis. Beta cells (arrowheads) are only seen in the outer scale surface not in the inner surface where only the oberahautchen is present. Bar, 15 µm. Legends: c, clear cells; de, dermis; h, hinge region; i, inner scale surface/side; la, lacunar tissue; mu, regenerated muscle bundles; n, normal tail (stump); o, outer scale surface/side; pg; epidermal peg; r, regenerating tail; sc, scale; t, tail tip; w, wound epidermis.

Fig. 7.

Schematic representation of proximal (left) and distal (right) scale neogenesis in regenerating tail of T. mautitanica (A) and relative in situ hybridization using alkaline-phosphatase detection (B–E). (A) Differentiating beta cells (arrows) form a compact beta layer in mature, proximal neogenic (regenerated) scales. (B) Forming neogenic scales with expression of mRNA for beta-keratin (arrows) in forming beta layer. Bar, 20 µm. (C) Detail of labelled cells (arrows) in the middle of regenerating scales. Bar, 10 µm. (D) Sense control showing no reactive cells (arrows). Bar, 10 µm. (E) Detail of the shedding complex in regenerating scales with the negative clear layer (arrowheads) contacting the highly reactive oberhautchen cells (arrow). Bar, 10 µm. Legends: aR, antisense cRNA probe; de, dermis; i, inner scale surface; l, lacunar tissue; o, outer scale surface; p, epidermal peg; s, sloughing wound epithelium; Rs, sense cRNA probe; t, tail tip; w, wound epidermis. Dashes underline the basal layer of the epidermis.

Regenerating scales are derived from the invagination of epidermal pegs inside the dermis of apical regions of the regenerating tail (Fig. 6A). These pegs were covered by a multilayered epidermis that cornified on the surface, as indicated by a corneous wound epidermis. Differentiating cells beneath the wound epidermis (spinosus cells) formed the so-called lacunar tissue (Maderson, 1985). Pegs had grown deeply into the dermis and their central layers had begun to differentiate into elongated cells (Fig. 6E). The pale cells localized vertically in the middle of pegs formed the clear layer of the neogenic scale, and were followed by other narrow cells of the oberhautchen layer (Fig. 6E,F). In pegs localized more proximally, near the tail stump, oberhautchen cells had formed a dense line and were followed by fusiform/flat cells of the differentiating beta layer (Fig. 6E,F). The latter accumulated only in the forming outer surface of the regenerated scale, which coincided with the right side of the peg as presented in Fig. 6(D–G). The inner scale surface of the (preceding) scale was therefore derived from the left side (rostral) of a peg. The new scales were now evident beneath the wound epidermis following the dense line made by the oberhautchen spinulae (Fig. 6C,F,G). Neogenic scales are therefore derived from the splitting of successive pegs into an outer (caudal part of the peg) and inner (rostral part of the same peg) scale surface with the loss of the wound epidermis that had covered the pegs (Fig. 6G).

The covering wound epidermis first begins to detach and is eventually sloughed from the underlying neogenic scales (Fig. 6C). The sequence of epidermal differentiation in regenerating scales occurs as during renewal of normal scales and of embryonic scales. As a result, numerous differentiating oberhautchen and beta cells were found in regenerating scales, and they served appropriately for the in situ hybridization study.

In situ hybridization

Alkaline-phosphatase-tagged RNA probes from T. mauritanica beta-keratins were localized in regenerating scales, revealed by a reddish-blue colour (Fig. 1) essentially present in the forming oberhautchen layer and in the differentiating beta cells localized in the middle part of the epidermal pegs (Fig. 7A–E). The clear layer and the remaining epidermal layers, and non-epidermal tissues, were all negative. In the latter the differentiating oberhautchen and beta layers remained unstained (Fig. 6D).

In addition, the fluorescent anti-DIG labelling probes from H. turcicus (Fig. 1) indicated the presence of positive cells in the forming oberhautchen and beta-layer within neogenic scales of T. mauritanica (interspecies cross-hybridization, Fig. 8A–D). The fluorescent tag also detected messengers for beta-keratin in differentiating oberhautchen and beta cells of the inner generation of setae in H. turcicus (Fig. 8E,F). Moreover some positive signal was also seen in cells of the beta layer at the base of setae of the outer generation (Fig. 8G) supporting the proline labelling previously described in this layer. Sense controls were negative (Fig. 8H,I).

Fig. 8.

In situ hybridization and fluorescence detection using FITC of regenerating scales of T. mauritanica at 4 weeks post-amputation (A–D) and pad lamellae of H. turcicus (E–I). (A) Neogenic scales with forming beta layer (arrows) expressing mRNAs for beta keratin. Bar, 20 µm. (B) Detail of the positive cells (arrows) of forming oberhautchen/beta cells of regenerating scales. Bar, 10 µm. (C) Sense control showing lack of reactivity in the beta layer (arrow). Bar, 10 µm. (D) Detail of reactive beta cells (arrow) in elongated peg. Bar, 10 µm. (E) Overview of pad lamella with reactive base (arrow) of outer setae, oberhautchen (arrowheads) and beta cells of the inner generation. Bar, 20 µm. (F) Detail of reactive oberhautchen cells (arrows) in early stage of differentiation. Bar, 10 µm. (G) Detail of reactive cells (arrow) at the base of outer setae. Bar, 10 µm. (H) Sense control showing lack of reactivity in oberhautchen (arrowheads) and beta cells. Bar, 10 µm. (I) Detail of sense control of oberhautchen (arrows) and beta cells. Bar, 10 µm. Legends: aR, antisense cRNA probe; de, dermis; s, setae; sR, sense cRNA probe (control). Dashes underline the basal layer of the epidermis.

The ultrastructural in situ hybridization study on normal scales and pad lamellae with differentiating inner spinulae or setae confirmed the above light microscopic in situ localization of beta-keratins. Sparse clusters of gold granules (3–7) were among keratin material or associated with the less dense beta-keratin filaments of oberhautchen and beta cells (Fig. 9A,B). Sparse clusters of gold particles were also present among beta-keratin bundles or associated with the long filaments of beta-keratin seen inside differentiating setae at the beginning of their growth (Fig. 9C). Diffuse and occasional cluster labelling was present in the surrounding cytoplasm of clear cells.

Fig. 9.

Ultrastructural in situ hybridization with antisense cRNA probes (aR) on tissues of H. turcicus. (A) Detail of oberhautchen cell containing clusters of gold particles (arrow) in the pale cytoplasm or associated with denser keratin bundles (arrowhead). Bar, 250 nm. (B) Detail of elongated clusters of gold particles among keratin filaments of setae (arrowheads). Bar, 200 nm. (C) Detail of beta-keratin bundles in setae (s) among which clusters of gold particles are seen (arrows). Bar, 250 nm.

In nuclei of oberhautchen and beta cells diffuse labelling was observed over euchromatin, but clusters were less commonly seen than in the cytoplasm (Fig. 10A). Diffuse gold particles, not forming clusters, were seen in the sense controls of the oberhautchen and beta cells (Fig. 10B). In other cell layers of the epidermis and in fibroblasts of the dermis, sparse and randomly distributed gold granules were seen. No labelling was seen in the other cells (muscle, blood, endothelium; data not shown).

Fig. 10.

Ultrastructural in situ hybridization with antisense (A) and sense (B) cRNA probes on tissues of H. turcicus. (A) Part of the nucleous of an oberhautchen cell with diffuse antisense cRNA probe labelling (arrowhead) over chromatin and close (arrow) to a perinuclear beta-keratin filament. Bar, 150 nm. (B) Sense control cRNA probe of an area at the base of setae between the cytoplasm of a clear cell (cl). No clusters of gold particles are seen. Bar, 200 nm. Legends: aR, antisense cRNA probe; bs, base of setae; β-fil, beta-keratin filament; cl, cytoplasm of clear cells; n, nucleous (outlined by dashes); s, setae; sR, sense cRNA probe (control).

Discussion

Expression of beta-keratins

Beta-keratins are small proteins present in the hard corneous layers of reptilian scales. Their synthesis decreases or completely disappears in soft layers (Wyld & Brush, 1979, 1983; Sawyer et al. 2000; Alibardi & Toni, 2006a, b). Despite some information on the protein pattern of tissue expression, molecular weight, and on the percentage of amino acid composition of some beta-keratins, molecular studies have only recently begun to determine the amino acid sequences and gene structure of beta-keratins (Dalla Valle et al. 2005, 2007). These latter studies will make possible the comparison of reptilian corneous proteins with those present in the skin of birds and mammals and thus will increase our understanding of their molecular evolution (Alibardi et al. 2006).

The present molecular technique is an essential partner to sequencing studies, showing the precise cellular site of expression of beta-keratins in the epidermis of geckoes. Moreover, proline uptake indicates the synthesis of beta-keratins of 14–24 kDa (Alibardi & Toni, 2005), which can then be precisely located in the epidermis by using radiolabelled proline autoradiography. This study utilized both regenerating scales of the tail skin and the renewing epidermis of the pad lamellae of geckoes: in these tissues the formation of epidermal layers follows a similar programme of differentiation as that during physiological regeneration (the shedding cycle) in normal scales (Maderson et al. 1998; Alibardi, 2003).

The study confirms that the messengers coding for the glycine–proline–serine-protein are specifically present in the cytoplasm of cells of the oberhautchen and beta cells, and more precisely among the beta-keratin filaments of these cells. Growing setae of the gecko also contain the messengers that appear associated with the surface of these keratin filaments. This indicates that in the beta cells of the gecko epidermis, newly synthesized protein is rapidly polymerized on the surface of beta-keratin filaments, leading to an increase in the diameter of the long bundles of beta-keratin filaments producing setae growth (Alibardi, 1997, 2003; Alibardi & Toni, 2005; Rizzo et al. 2006). Therefore, the synthesis and aggregation of beta-keratin monomers for the elongation of the setae occurs not only at the base of the setae but also inside the setae themselves as long as ribosomes and living cytoplasm remain before their complete cornification. Finally, the nuclear localization to euchromatin may indicate that the primary transcripts are being recognized by the probe. This is in agreement with molecular studies on gecko beta-keratin that have indicated lack of introns in genes coding for these proteins (Dalla Valle et al. 2007).

The sequenced glycine–proline–serine-rich protein is only one of the beta-keratin components of beta-filaments of gecko setae. Other beta-keratins, alpha-keratins and associated proteins of still unknown composition are also present. Previous biochemical studies have shown the presence of high-sulfur and low-sulfur keratins in gecko epidermis (Thorpe & Giddings, 1981). Therefore, beta-keratins, in addition to their mechanical protection of reptilian scales, also form the resistant elongations of the scale surface, the setae, that permit adhesion and climbing in geckoes.

The present observations support similar results obtained for oberhaucthen and beta cells of scales in the lizard Podarcis sicula and for the epidermis of other reptiles for which the mRNAs coding for glycine–proline–serine-rich proteins have been localized (Dalla Valle et al. 2005; Alibardi et al. 2006; our unpublished observations). The expression of messengers for these proteins appears ubiquitous in beta cells of the epidermis of reptiles.

Setae growth, apical branching and replacement during moulting

The specialized scales of pad lamellae maintain a relatively flexible corneous layer which is needed for maintaining the elasticity and adaptation of these scansorial organs to the substrate (Lillywhite & Maderson, 1968; Peterson & Williams, 1981; Bauer, 1998; Rosenberg et al. 1998).

Wearing of setae or their reciprocal adhesion due to the extensive usage of these adhesive devices produces a decrease of adhesion efficiency. This requires a rapid change of the outer with an inner generation of setae to maintain the best adhesion efficiency. The frequency of shedding in normal scales of geckoes, including pad lamellae, is about 2 weeks or less at high (summer) temperatures (Maderson, 1985; L.A., personal observations). Therefore, it is conceivable that the corneous surface of digital pads may act as a ‘rolling-track’ where the outer setae move toward the pad tip to make room for the next setae, and are shed from the apical part of the lamella (Fig. 11). This indicates that even the outer epidermal generation is not completely mature before it is shed. Shedding is a phenomenon of delayed keratinocyte differentiation, which has been described particularly in snake epidermis (Maderson, 1985; Maderson et al. 1998).

Fig. 11.

Schematic drawing of gecko digit pads (A) with the detail (square) of the shedding cycle of a pad lamella (B–F, see text for more details) and details of setae formations (B1, C1, C2, D1, E1, F1). After shedding the lamella is in resting stage with one generation of setae (B). Setae are mature with an apical branching (B1). A new generation begins to form underneath the old (outer) setae (C). Each seta is derived from the incorporation of beta-keratin material and other corneous material (in red) from the oberhautchen cell (violet in C1) and is surrounded by the cytoplasm of a clear cell (pale yellow). More material added at the base of the growing seta (blue in C2) determines its length. The tip of the scale moves forward (right) and determines a progressive shifting of the outer setae in mid (D) and late (E) renewal to form the free margin (elongating tip). The apical branching of elongating seta takes place (in E1) while more material incorporated at the base of the setae (in orange) allows a further elongation. Finally, near shedding (F) the apical branching of elongating setae is almost complete and the clear cell begins to separate (double arrows) from the oberhautchen of the inner generation of setae (F1). The apical separation of terminal beta-keratin bundles probably occurs by degeneration of the non-keratinized cytoplasm present among terminal beta-keratin bundles. Opposed terminal spatulae remain as they are keratinized structures. Fully differentiated inner setae eventually replace the outer setae at shedding and the moult is lost.

As they grow gecko setae branch apically (see details of setae in Fig. 11C1, C2, D1, E1, F1). From the present TEM observations multiple spatular endings appear to be derived from the initial, fine carving of the tip of setae from the cytoplasm of clear cells in conjunction with the disappearance of the non-corneous cytoplasm present among terminal bundles of beta-keratin in the setae. Therefore, it seems that a species-specific mechanism of cytoplasm degeneration among terminal beta-keratin bundles of setae is probably responsible for apical branching of gecko setae (as illustrated in Fig. 11E1, F1, B1).

The movement of outer setae toward the tip of pad lamellae has been previously shown for the lizard Anolis carolinensis (Lillywhite & Maderson, 1968). The latter study hypothesized that setae moved apically by two processes: (1) epidermal lengthening of the oberhautchen/beta layer sustaining the setae, and (2) dermal retraction of the pad lamella that leaved an elongated oberaucthen/beta layer to sustain the setae. A combination of the two processes could also be implicated in this process. The present autoradiographic and in situ hybridization observations indicate that indeed the addition of newly synthesized beta-keratin still occurs in oberhautchen/beta cells at the base of the outer setae before shedding (Fig. 11). These cells would be produced mainly near the tip of the pad lamella and would determine the movement of previously deposited cornified cells toward the free margin. In the inner surface of pad lamellae, this forward movement determines the accumulation of the extra-corneous material above the corneous layer (arrowhead in Fig. 5A).