SEARCH

SEARCH BY CITATION

Keywords:

  • regeneration;
  • wound healing;
  • lizard;
  • scar-free;
  • amputation;
  • autotomy

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. HISTOLOGY
  7. DISCUSSION
  8. Acknowledgements
  9. LITERATURE CITED
  10. Supporting Information

Many lizards are able to undergo scar-free wound healing and regeneration following loss of the tail. In most instances, lizard tail loss is facilitated by autotomy, an evolved mechanism that permits the tail to be self-detached at pre-existing fracture planes. However, it has also been reported that the tail can regenerate following surgical amputation outside the fracture plane. In this study, we used the leopard gecko, Eublepharis macularius, to investigate and compare wound healing and regeneration following autotomy at a fracture plane and amputation outside the fracture plane. Both forms of tail loss undergo a nearly identical sequence of events leading to scar-free wound healing and regeneration. Early wound healing is characterized by transient myofibroblasts and the formation of a highly proliferative wound epithelium immunoreactive for the wound keratin marker WE6. The new tail forms from what is commonly referred to as a blastema, a mass of proliferating mesenchymal-like cells. Blastema cells express the protease matrix metalloproteinase-9. Apoptosis (demonstrated by activated caspase 3 immunostaining) is largely restricted to isolated cells of the original and regenerating tail tissues, although cell death also occurs within dermal structures at the original-regenerated tissue interface and among clusters of newly formed myocytes. Furthermore, the autotomized tail is unique in demonstrating apoptosis among cells adjacent to the fracture planes. Unlike mammals, transforming growth factor-β3 is not involved in wound healing. We demonstrate that scar-free wound healing and regeneration are intrinsic properties of the tail, unrelated to the location or mode of tail detachment. Anat Rec, 2012. © 2012 Wiley-Periodicals, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. HISTOLOGY
  7. DISCUSSION
  8. Acknowledgements
  9. LITERATURE CITED
  10. Supporting Information

Wound healing is a dynamic morphogenetic event involving the activation of multiple intracellular and extracellular processes, leading to the restoration of tissue integrity and physiological equilibrium (Singer and Clark, 1999; Gurtner et al., 2008). Details of this reparative phenomenon are complex and varied, and influenced by numerous features such as the size, severity and nature of the injury, and the site of the wound (Gurtner et al., 2008). At completion, wound healing yields one of two possible outcomes: (1) the wound site is repaired with fibrotic scar tissue; or (2) the wound site undergoes scar-free wound healing and the original tissue is restored. While scar formation is the most familiar mode of wound healing, at least for mammals, the resulting production of fibrotic tissue alters the structural and functional properties of the wound site (Ferguson and O'Kane, 2004). In contrast, scar-free wound healing, as commonly demonstrated by urodeles, restores tissue structure and, in most instances, function (reviewed in Ferguson and O'Kane, 2004). Furthermore, scar-free healing appears to be a requirement for reparative regeneration (Ferguson and O'Kane, 2004; Metcalfe and Ferguson, 2007; Gurtner et al., 2008; Yokoyama, 2008; Occleston et al., 2010).

Among amniotes, one of the most remarkable examples of scar-free wound healing and regeneration follows tail loss in some lizards. In various species, including most scincids, lacertids, and gekkotans (geckos), removal of a portion of the tail initially results in an open cross-sectional injury with various tissues exposed. This wound is rapidly sealed without the formation of scar tissue, and is eventually replaced by a new tail. The new tail begins from what is commonly referred to as a blastema (e.g., Barber, 1944; Jamison, 1964; Werner, 1967; Bellairs and Bryant, 1985; Alibardi, 2010; McLean and Vickaryous, 2011), a mass of proliferating mesenchymal-like cells that gradually gives rise to the tissues of the replacement appendage. Although widespread, strictly speaking this definition of “blastema” is incomplete. Among other regeneration models, a blastema is more precisely defined as a population of proliferating undifferentiated progenitor cells (e.g., Kragl et al., 2009; Lehoczky et al., 2011) and to date it remains unclear if lizard tail regeneration involves dedifferentiation. However, in the interests of avoiding unnecessary confusion, and in keeping with the existing literature, we use the term blastema while acknowledging that the source of these cells remains unknown.

For most species, tail regeneration is associated with the ability to self-detach or autotomize the tail (Bellairs and Bryant, 1985). Autotomy is an anti-predation strategy wherein the tail is lost at pre-existing planes of weakness known as fracture planes. Each fracture plane is a transversely oriented bilayer of fibrous connective tissue that subdivides sequential segments of skeletal muscle, adipose tissue and (in the majority of taxa) individual vertebrae (Bellairs and Bryant, 1985; McLean and Vickaryous, 2011). As a result, the site of tail loss is intravertebral. Once shed, the site of autotomy undergoes a conserved program of wound healing and regeneration that includes cell proliferation, tissue differentiation and eventually, the re-establishment of a replacement appendage (Bellairs and Bryant, 1985; Alibardi, 2010; McLean and Vickaryous, 2011).

Although clearly adaptive, experimental evidence suggests that the phenomena of autotomy and regeneration are, in fact, independent of one another. For example, several studies have reported tail regeneration taking place following surgical amputation (e.g., Woodland, 1920; Barber, 1944; Werner, 1967; Whimster, 1978). These findings point towards the lizard tail as a valuable but often overlooked amniote model for the study of spontaneous reparative regeneration. In this study, we conducted a comparative investigation of wound healing and regeneration following two forms of tail loss, autotomy at the fracture plane and amputation outside the fracture plane, using the leopard gecko, Eublepharis macularius. We provide a detailed characterization of changes in tissue structure following both forms of tail loss at multiple stages of the regenerative program to test our prediction that scar-free wound healing and regeneration are intrinsic properties of the leopard gecko tail, unrelated to the location (i.e., at or outside the fracture plane) or mode (autotomy vs. amputation) of tail detachment. To document the tissue-level events, we used serial histochemistry and immunohistochemistry, focusing on six proteins of interest known to be associated with wound healing: proliferating cell nuclear antigen (PCNA), the wound keratin WE6, matrix metalloproteinease-9 (MMP-9), α-smooth muscle actin (α-SMA), transforming growth factor-β3 (TGF-β3), and activated caspase 3 (C3).

PCNA is a marker of the synthesis (S) phase of the cell cycle and visualizes proliferating cells in a wide variety of organisms including teleost fish (Santos-Ruiz et al., 2002), mammals (Foley et al., 1993), and reptiles (Handrigan et al., 2010; McLean and Vickaryous, 2011). In leopard geckos, PCNA is expressed by various proliferative cell types in the original tail (e.g., basal epithelial cells) and by mesenchymal-like blastema cells and committed progenitor cells (e.g., chondroprogenitors, myoblasts) throughout the process of tail regeneration (McLean and Vickaryous, 2011).

WE6 is a wound keratin originally isolated from newts following limb amputation (Estrada et al., 1993; Campbell and Crews, 2008). WE6 has since been identified in the wound epithelium of the lizard Podarcis muralis (Alibardi and Toni, 2005). Although the function of WE6 is currently unknown, its expression appears to be restricted exclusively to wound epithelium (Estrada et al., 1993; Alibardi and Toni, 2005; Campbell and Crews, 2008), thus indicating that original epidermis and wound epithelium differ in their structural composition.

MMP-9 is a protease capable of degrading collagen (types IV and V) and gelatin (types I and V) (Cawston, 1996). In other wound repair models (e.g., axolotls, MRL mice) MMP-9 is associated with scar-free wound healing and regeneration (Yang et al., 1999; Gourevitch et al., 2003). Interestingly, hypertrophic scars and keloids are associated with reduced levels of MMP-9, suggesting a potential role for this protein in regulating scar formation (Neely et al., 1999).

α-SMA is a protein characteristic of the smooth muscle contractile apparatus. During wound healing α-SMA is expressed by myofibroblasts following their differentiation from pre-existing fibroblasts (Hinz, 2007). The coordinated efforts of many myofibroblasts facilitate wound contracture (Desmoulière et al., 1993). In addition to its contractile properties, myofibroblasts secrete substantial amount of collagens (e.g., types I and III) into the wound site (Gabbiani, 2003). Furthermore, the persistence of myofibroblasts at the wound site has been associated with excessive scarring pathologies (e.g., keloids) (Desmoulière et al., 2005). As wound closure (i.e., re-epithelialization) is completed wound-associated myofibroblasts undergo apoptosis (Desmoulière et al., 1995). Evidence from the study of urodeles (Kragl et al., 2009) indicates that myofibroblasts are not a source of regenerating skeletal muscle. α-SMA is also found in pericytes and smooth muscle cells of larger blood vessels (Skalli et al., 1989).

TGF-β3 is a cytokine previously documented to participate in scar-free wound healing in mammalian embryos and fetuses (Bullard et al., 2003; Coolen et al., 2010). More specifically, it has been demonstrated that elevated levels of TGF-β3 (and comparatively low levels of TGF-β1) promote scar-free wound healing (Ferguson and O'Kane, 2004).

C3 is the effector of the caspase-mediated apoptotic program and is widely recognized as a definitive marker of programmed cell death (Budihardjo et al., 1999; Krysko et al., 2008). Activation of C3 is known to be required for regeneration in both zebrafish, Danio rerio, and the anuran Xenopus laevis (Chablais and Jazwinska, 2010). In X. laevis, apoptosis is observed within the blastema and wound epithelium of regenerating limbs (Suzuki et al., 2005). When apoptosis is blocked by the use of a C3 inhibitor, cell proliferation ceases and outgrowth of the once regenerating limb is halted (Tseng et al., 2007). Previous investigations on lizards have explored the role of apoptosis during tissue homeostasis, but not regeneration. For example, it is reported that C3 participates in seasonal apoptotic remodeling of the pancreas, spleen, and oviducts in various lizard species (Buono et al., 2006; Assisi et al., 2011) as well as androgen-induced thymic atrophy (Hareramadas and Rai, 2005, 2006).

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. HISTOLOGY
  7. DISCUSSION
  8. Acknowledgements
  9. LITERATURE CITED
  10. Supporting Information

Animal Care

All leopard geckos were captive bred and obtained from a commercial supplier (Global Exotic Pets, Kitchener, Ontario). Animal utilization protocols were approved by the University of Guelph Animal Care Committee (protocol numbers 09R026 and 10R113). Geckos were housed individually in standard rat enclosures in the Central Animal Facility at the University of Guelph, following the husbandry protocol of Vickaryous and McLean (2011; see also McLean and Vickaryous, 2011). Room temperature was maintained at 24–26 degrees Celsius (°C). A heat cable was placed under one end of the enclosure, raising the substrate temperature to 28°C thus establishing a thermal gradient. Geckos were maintained on a 12:12 photoperiod and fed gut-loaded mealworms (larval Tenebrio sp.) dusted with powdered calcium plus vitamin D3 (cholecalciferol) supplement once daily. Each enclosure included two hide boxes and a dish of water (changed twice weekly). Geckos were given a one week acclimatization period prior to any manipulation.

Tail Collection

Autotomy.

Autotomy was achieved using the protocol of McLean and Vickaryous (2011). Leopard geckos were manually restrained and the tails were firmly pinched at a position halfway between the cloaca and tail tip. Autotomy targeted constrictions in the integument (Fig. 1A), corresponding to the positions of fracture planes. The distal portion of the tail is self-detached, whereas the proximal portion is retained. Following this initial autotomy event, individuals were then returned to their enclosures and allowed to regenerate. Representative tails were then induced to autotomize a second time at each of the seven morphological stage of regeneration (McLean and Vickaryous, 2011) and the tissue was collected and processed for histological investigation. To date, we have autotomized and collected tissues from over 100 individuals.

thumbnail image

Figure 1. Gross morphology and osteology of the tail of the leopard gecko, Eublepharis macularius. (A) Dorsal view of the original tail. The original tail contains transverse constrictions that correspond to intravertebral fracture planes. (B) Reconstructed microcomputed tomographic scans of the original tail in lateral view. The arrows indicate intravertebral fracture planes. Gross morphology of site of tail detachment in cranial view following autotomy (C,E,G) and amputation (D,F,H). Stage I tails following autotomy (C) and amputation (D). Stage II tails following autotomy (E) and amputation (F). Note the formation of an exudate clot. Stage III tails following autotomy (G) and amputation (H). Note the formation of a wound epithelium.

Download figure to PowerPoint

Amputation

Individuals were randomly assigned into one of four groups: unaltered controls (no autotomy, no anesthesia, no amputation; 3n), anesthesia controls (no autotomy, no amputation; 4n), autotomy controls (no anesthesia, no amputation; tails collected as described above), and anesthesia plus amputations (no autotomy; 17n). Geckos undergoing amputation first received a preanesthetic intramuscular injection of buprenorphine (0.075 mg/kg), administered into the epaxial muscles of the neck. Buprenorphine is an analgesic in reptiles that has been reported to minimize breath holding (Schumacher and Yelen, 2005; Mosley, 2011). After 10 min, geckos were placed in a plastic anesthetic chamber (11 cm length, 11 cm deep, 7 cm height) connected to an anesthetic machine with an oxygen and isoflurane supply. The anesthetic chamber was set above a heat cable to maintain the preferred temperature range. Oxygen flow rate to the chamber was set at 1 L/min throughout the procedure, during which time the isoflurane concentration was slowly increased from 0% to 5% saturation (0.5% increments every 2–3 min). Anesthetic depth was evaluated by rotating the chamber 180 degrees to determine if the gecko retained a righting reflex. The surgical plane of anesthesia was established when the righting reflex was no longer present (Schumacher and Yelen, 2005). Once the adequate anesthetic depth was achieved, individuals were removed and either placed back into their enclosures for observation (anesthesia controls) or had their tails amputated with a scalpel blade (anesthesia plus amputation group). Amputations targeted positions halfway along the tail length outside the fracture plane constrictions in the integument (Fig. 1A,B). All geckos survived anesthesia and amputation. As described above, representative tails were then induced to autotomize a second time at each of seven morphological stage of regeneration (McLean and Vickaryous, 2011) and the tissue was collected and processed for histological investigation.

Histology

Following autotomy/amputation, tails were fixed with 10% neutral-buffered formalin (NBF) for 20–24 hr then stored in 70% ethanol. Prior to sectioning, tails were decalcified (Surgipath Decalcifier) for 30 min, and processed for paraffin histology (dehydration to 100% ethanol, clearing in xylene, and embedded in paraffin wax; Tissue Prep). Sections were cut at 5 μm and mounted on charged glass slides (Snow Coat X-tra or Fisher Scientific Superfrost Plus) and baked at 37°C overnight.

Histochemistry

Four histochemical protocols were used to investigate tissue structure and composition including Hematoxylin and Eosin (Carson, 1997), Masson's trichrome [connective tissue stain; adapted from Witten and Hall, 2003 (see McLean and Vickaryous, 2011)], Alcian blue (glycosaminoglycan stain; Spicer et al., 1967) and Safranin O (glycosaminoglycan stain; Rosenberg, 1971).

Immunohistochemistry

Immunohistochemistry was used to identify the location and temporal expression of the six proteins of interest (PCNA, WE6, MMP-9, α-SMA, TGF-β3, and C3). Unless otherwise noted, all immunohistochemical protocols were similar (Table 1). Slide mounted serial sections were rehydrated, quenched in 3% H2O2 for 20 min, and rinsed with phosphate-buffered saline (PBS). When necessary (MMP-9, WE6), heat induced epitope retrieval (citrate buffer heated to 90°C) was used to unmask antigenic sites. Sections were then incubated with a 3% normal goat serum (NGS) blocking solution and incubated overnight with primary antibody at 4°C. Primary antibody was omitted from one section on each slide to function as a negative control. Slides were rinsed in PBS (three washes, 2 min each) and then all sections were incubated with biotinylated secondary antibody in a humidity chamber (1 hr). Slides were then rinsed with PBS (three washes, 2 min each) and incubated with horse radish peroxidise-conjugated streptavidin (SA-HRP; 1:200 dilution; Jackson Immuno Research Laboratories, 016-030-084) for 1 hr at room temperature. Following SA-HRP incubation slides were rinsed with PBS (three washes, 2 min each) and stained with 3,3′-diaminobenzidine [DAB; diluted in 5mL d H2O (Peroxidase substrate kit, Vector Laboratories, SK-4100)]. Slides were then rinsed in deionized water, counterstained in Mayers Hematoxylin (1 min, 30 sec for C3), rinsed in deionized water, blued in ammonia water, dehydrated through absolute isopropanol (3 changes, 2 min each), cleared with xylene (3 changes, 2 min each), and coverslipped using cytoseal.

Table 1. Summary of the immunohistochemistry protocols for the six proteins of interest: proliferating cell nuclear antigen (PCNA), matrixmetalloproteinase 9 (MMP-9), WE6, α-smooth muscle actin (α-SMA), transforming growth factor-β3 (TGF-β3), and activated caspase 3 (C3)
 PCNAMMP-9WE6α-SMATGF-β3C3
Antigen RetrievalNone12 min citrate buffer (pH 6.0, 90°C), cool for 20 min in buffer13 min citrate buffer (pH 6.0, 90°C), cool for 20 min in bufferNoneNoneNone
Primary Antibody1:500, Santa Cruz Biotechnology PCNA (FL-261) sc-79071:200, Santa Cruz Biotechnology MMP-9 (H-129) sc-10731:5, Hybridoma Bank WE61:500, Santa Cruz Biotechnology sc-322511:20, Hybridoma Bank, D-β3 act1:500, Calbiochem PC679
Secondary Antibody1:200, Biotinylated Goat anti-rabbit, Jackson Immuno Research Laboratories, product code 111-066-0031:200, Biotinylated Goat anti-rabbit, Jackson Immuno Research Laboratories, product code 111-066-0031:200, Biotinylated Goat anti-mouse IgG Vector Laboratories, catalog BA-92001:200, Biotinylated Goat anti-mouse IgG Vector Laboratories, catalog BA-92001:200, Biotinylated Goat anti-mouse IgG Vector Laboratories, catalog BA-92001:200, Biotinylated Goat anti-rabbit, Jackson Immuno Research Laboratories, product code 111-066-003
       

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. HISTOLOGY
  7. DISCUSSION
  8. Acknowledgements
  9. LITERATURE CITED
  10. Supporting Information

Gross Morphology

To determine if the reparative events following intravertebral autotomy and intervertebal amputation are comparable, we documented wound healing and tissue restoration at each of the seven stages of tail regeneration (McLean and Vickaryous, 2011). For the sake of convenience, these descriptions are organized into separate sections detailing aspects of gross morphology and histology for both post-autotomy and post-amputation wound healing and regeneration. As will be demonstrated, the reparative events following each mode of tail loss were virtually identical from stages III onward; hence, these data are presented together. A summary of the differences observed following the two forms of tail loss is presented in Table 2. It is also worth noting that all geckos survived and rapidly recovered from each form of tail loss. Differences in behavior observed among individual undergoing either autotomy or amputation were minimal, and mostly related to recovery from anesthetic (Supporting Information 1).

Table 2. A comparison of gross morphological and histological differences between post-autotomy and post-amputation tails
 Post-autotomy tailPost-amputation tail
Stage I• Autotomy surface jagged• Amputation surface flat
• Minimal blood loss (less than one drop)• Increase in blood loss (∼two drops)
• Exudate clot localized to distal tip of spinal cord• Exudate clot spans entire amputation surface
• Tail detached at a transverse constriction of the integument• Tail detached between two transverse constrictions of the integument
Stage II• Exudate clot spans entire autotomy surface• Exudate clot rich in red blood cells spans amputation surface
• Vertebral remnant protrudes from autotomy surface• Vertebral remnant is not obvious at amputation surface
• Wound epithelium evenly stratified and planar• Wound epithelium unevenly stratified and undulating
Stages III to VII• Interface of original and regenerated integument occurs at a transverse constriction• Interface of original and regenerated integument occurs between to adjacent transverse constrictions
• Vertebral remnant approximately one-quarter of original vertebra• Vertebral remnant approximately one-half of original vertebra

Original Tail

The original tail is a cone-shaped, distally tapering appendage. In dorsal and lateral views the tail is characterized by a repeated pattern of scalation consisting of large raised tubercles and small imbricating scales. At regular intervals, the scalation is interrupted by transverse constrictions (Fig. 1A). As evidenced by microcomputed tomography (Fig. 1B), the majority of these constrictions correspond to the positions of intravertebral fracture planes. Pigmentation of the tail, as for elsewhere on the body, is highly variable but typically includes some combination of dark colored spots (e.g., black, brown or purple) on a lighter background (typically yellow, white and orange). The ventral surface of the tail has fewer, smaller-sized dark spots.

Stage I post-autotomy (0 to ∼24 hr).

Following tail loss by autotomy, the distal end of the tail resembles an open wound with various tissues exposed including skeletal muscle, bone, adipose tissue, and spinal cord (Fig. 1C). Blood loss is minimal (approximately one drop or less), and is primarily restricted to the area adjacent to the ruptured spinal cord. In profile, the site of autotomy is uneven and almost jagged in appearance, reflecting the zigzag course of the fibrous fracture plane through the tail. Shortly after autotomy, the integument surrounding the autotomy surface begins to retract and partially close over the exposed tissues. Within hours, continued retraction gives the integument a puckered appearance. The proximal half of the autotomized vertebra is distinctly visible protruding from the center of the autotomy surface. By the end of stage I the site of autotomy is capped by a thin exudate clot composed of tissue fluid, blood, and tissue debris.

Stage I post-amputation (0 to ∼24 hr).

The site of tail loss immediately following amputation (Fig. 1D) closely resembles that of a stage I post-autotomy tail, with two major exceptions. First, compared to autotomized tails there is an increase in blood loss following amputation [approximately two drops (∼75μL) instead of one]. As a result, the exudate clot is more prominent and darker in color. The second major difference is that the site of amputation is flat, lacking a jagged appearance and protruding vertebral remnant. Otherwise, amputation and autotomy yield a similar open wound site with exposed skeletal muscle, adipose tissue, and spinal cord.

Stage II post-autotomy (∼1–7 days).

By stage II, exposed tissues at the site of tail loss have undergone noticeable retraction, further exposing the vertebral remnant (Fig. 1E). The exudate clot is completely crystallized and spans the entire autotomy surface.

Stage II post-amputation (∼1–7 days).

The gross morphology of stage II following amputation (Fig. 1F) is similar to that of autotomy, although the exudate clot appears to be thicker and distinctly darker. In addition, the vertebral remnant at the amputation site is minimally exposed or entirely absent.

Stage III post-autotomy (∼7–10 days).

By stage III, the exudate clot and protruding vertebral remnant are ablated, exposing a smooth, taut wound epithelium. Deep to the wound epithelium are various small blood vessels coursing within the regenerating tissue (Fig. 1G). Outgrowth of the blastema is minimal at this stage.

Stage III post-amputation (∼7–10 days).

From stage III onwards, the gross morphology of the tail following amputation is indistinguishable from that of a tail following autotomy (Fig. 1H). Of particular note at this stage, the once thick, blackened exudates clot is lost exposing a smooth, tightly stretched wound epithelium indistinguishable in appearance from that of an autotomized tail.

Stage IV post-autotomy and post-amputation (∼10–15 days).

Stage IV (Fig. 2A,B) marks the beginning of blastema outgrowth, resulting in the formation of a pinkish, dome-like mound of regenerating tissue, with no evidence of pigmentation or discrete scalation. The blastema has a length to diameter ratio of less than 0.5.

thumbnail image

Figure 2. Gross morphology of the regenerating tail of the leopard gecko, Eublepharis macularius, in dorsal view. Stage IV tails following autotomy (A) and amputation (B). This is the first regeneration stage where outward growth of the blastema is visible. Stage V tails following autotomy (C) and amputation (D). Stage VI tails following autotomy (E) and amputation (F). Stage VII tails following autotomy (G) and amputation (H).

Download figure to PowerPoint

Stage V post-autotomy and post-amputation (∼15–21 days).

At stage V (Fig. 2C,D) blastema outgrowth has continued, giving this cone-like structure a length to diameter ratio of greater than 0.5 but less than 1.0. The accumulating mound of new tissue remains unpigmented and without scalation.

Stage VI post-autotomy and post-amputation (∼17–26 days).

At stage VI (Fig. 2E,F) the regenerate outgrowth has become more tapered, with a length to diameter ratio of 1.0. However, the maximum diameter of the regenerate portion of the tail is still less than the minimum diameter of the original tail. Stage VI marks the first appearance of scalation, but the regenerated tissue remains unpigmented.

Stage VII post-autotomy and post-amputation (∼21 days+).

Stage VII (Fig. 2G,H) is distinguished by the initiation of pigmentation and the tapering cone-like appearance of the tail (length to diameter ratio greater than 1.0). In addition, the maximum diameter of the regenerate tail is equivalent to the minimum diameter of original tail and the pattern of scalation has become more prominent. However, the enlarged tubercles and transverse constrictions of the tail are not restored in the regenerate portion of the tail. Instead, scalation is composed of a homogenous pattern of small, imbricating scales.

HISTOLOGY

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. HISTOLOGY
  7. DISCUSSION
  8. Acknowledgements
  9. LITERATURE CITED
  10. Supporting Information

Original Tail

A detailed description of gecko original tail anatomy and histology is available elsewhere (McLean and Vickaryous, 2011); the following is presented as a summary of relevant structural features with emphasis on the site of tail loss. The original tail is a complex appendage with a concentric organization of tissues, including integument, skeletal muscle, adipose and fibrous connective tissue, and a centrally positioned skeletal system (Fig. 3A). The outermost layer of the integument, the epidermis, is a stratified, squamous-keratinized epithelium. As for other reptiles, the epidermis is divisible into several distinct horizons including the stratum basale (germinativum), stratum spinosum, stratum granulosum, and stratum corneum (Fig. 3B). Subadjacent to the epidermis, the dermis consists of the stratum superficiale, a loose connective tissue layer with abundant cells, and the stratum compactum, a deeper layer with fewer cells and more densely organized irregular connective tissue. Deep to the dermis is the hypodermis (Fig. 3C), a loose connective tissue network with numerous adipocytes.

thumbnail image

Figure 3. Tail histology of the leopard gecko, Eublepharis macularius, prior to tail loss. (A–D) Serial histological sections stained with Masson's trichrome and Safranin O (E). (F-Q) Immunohistochemistry using DAB visualization of histological sections. Insets are negative controls. Arrows indicate an example of an immunopositive cell. A: Sagittal section showing the organization of tissues in the original tail (dorsal towards top of page). Higher magnification images of the original epidermis (B), dermis and hypodermis (C), caudal artery and vein (D), and notochord in longitudinal section (E). Black arrowheads indicate the position of perinotochordal cartilage surrounding the notochord. Note the presence of two types of tissue in the notochord, chondroid (cartilage-like) tissue and chordoid (vacuolated cell-rich) tissue. Proliferating cell nuclear antigen (PCNA) immunopositive epidermis (F), fibroblasts (G), endothelial cells (H) and chondrocytes (I). The wound keratin marker WE6 is not expressed by cells of the original epidermis (J). α-smooth muscle actin (SMA) immunopositive blood vessels are present throughout the original tail (K). Matrix metalloproteinase-9 (MMP-9) immunopositive osteoclasts (L), fibroblasts (M), and chondrocytes of the vertebra (N). Transforming growth factor-β3 (TGF-β3) immunopositive fibroblasts (O) and chondrocytes (P). Activated caspase 3 (C3) immunopositive fibroblasts of the dermis (Q). Scale bars: A, 200 μm; B–C, E, J,K, N, 50 μm; D, 100 μm; F–I, L,M, O-Q, 20 μm. Abbreviations: ad, adipose tissue; ca, caudal artery; cv, caudal vein; hd, hypodermis; m, muscle; my, myoseptum; nc, notochord; ncc, notochordal chondroid (cartilage) tissue; it, integument; sb, stratum basale; sc, stratum corneum; sg, stratum granulosum; sp, spinal cord; ss, stratum spinosum; stc, stratum compactum.

Download figure to PowerPoint

Skeletal (striated) muscle is longitudinally organized into zigzag-shaped blocks or myomeres. In cross-section the myomeres are organized into paired epaxial and hypaxial quadrants, each composed of numerous muscle fibers. In parasagittal section, sequential myomeres are connected by myosepta. Fibrous fracture planes pass within these myosepta, giving the autotomized tail a serrated or jagged appearance. Deep to the musculature are four (paired epaxial and hypaxial) bands of white adipose tissue immediately surrounding the vertebral column (Fig. 3A). This perivertebral adipose tissue is unilocular, and contains numerous small blood vessels and nerve fibers.

The skeletal axis of the tail includes a vertebral column plus a persistent notochord (Fig. 3A,E). The notochord is a rod-like organ that passes continuous through the majority of vertebrae beginning in the cervical series and ending within the distal most of the caudal series. Unlike most terrestrial vertebrates, skeletally mature geckos retain a notochord; in most taxa the notochord is a transient organ eventually replaced by cartilage. Histologically, the gecko notochord consists of two tissue types: chordoid tissue, dominated by large vacuolated cells with limited amounts of extracellular matrix; and chondroid tissue (notochordal cartilage), a cartilage-like matrix with smaller chondrocyte-like cells. These tissues are organized in an alternating pattern along the length of the notochord (Fig. 3E). The entire notochord is then encased within a thin layer of chondrocytes in matrix known as the perinotochordal cartilage. It is worth noting that in the tail the location of chondroid tissue matches with the position of the autotomy plane of the vertebra.

Each tail vertebra consists of a vertebral body (centrum), a dorsally positioned neural arch and a pair of laterally directed transverse processes. Most tail vertebrae also develop a freely articulating chevron (Fig. 1B). Unique to geckos (among lizards), the centrum is amphicoelous with a well-developed central canal for passage of the notochord (notochordal canal). In lateral view, the tail centra are subdivided by intravertebral planes: transversely oriented zones of unmineralized connective tissue subdividing the elements into cranial and caudal segments (Fig. 1B). Hematopoietic tissue in marrow cavities is present within each of the cranial and caudal segments of the tail vertebrae.

The spinal cord passes dorsal to the centra and is partially enclosed by the overlying neural arches. Ventral to the spinal cord, within the neural canal, is a small arteriole, the spinal artery. The largest blood vessels of the tail, the caudal artery and caudal vein, are positioned ventral to the vertebral column (Fig. 3D). The caudal artery is distinguished by a series of prominent smooth muscle sphincters, each located immediately proximal to a fibrous fracture plane. Following detachment of the tail, these sphincters contract to minimize blood loss (Bellairs and Bryant, 1985; McLean and Vickaryous, 2011).

A summary of the immunohistochemical data for the original tail, as well as post-autotomy and post-amputation tails is presented in Table 3. In the original tail, PCNA immunostaining is restricted to cells of the stratum basale (of the epidermis; Fig. 3F) and hematopoietic tissue, as well as various fibroblasts (Fig. 3G), endothelial cells (Fig. 3H) and chondrocytes (Fig. 3I). Cells of the original tail do not express WE6 (Fig. 3J). α-SMA expression is restricted to smooth muscle cells in the walls of blood vessels and pericytes (Fig. 3K). Osteoclasts (Fig. 3L) as well as various fibroblasts of the dermis (Fig. 3M) and chondrocytes (Fig. 3N) are immunopositive for MMP-9. Isolated fibroblasts (Fig. 3O) and chondrocytes (Fig. 3P) are also immunopositive for TGF-β3. Throughout the original tails are various isolated cells immunoreactive for C3, including keratinocytes and fibroblasts of the dermis (Fig. 3Q).

Table 3. A summary of protein expression in original and regenerating tail tissues/cells
ProteinOriginal tissue/cellsWound site/regenerating tissue
Stage
IIIIIIIVVVIVII
  1. Abbreviations: PCNA, proliferating cell nuclear antigen; MMP-9, matrixmetalloproteinase 9; α-SMA, α-smooth muscle actin; TGF-β3, transforming growth factor-β3; C3, activated caspase 3.

PCNAEpidermis (stratum basale), hematopoietic cells, endothelial cells, chondroblastsNoneWound epithelium, blastema cellsWound epithelium, blastema cellsWound epithelium, blastema cellsWound epithelium (stratum basale)Wound epithelium (stratum basale), fibroblasts, chondroblastsWound epithelium (stratum basale), fibroblasts, chondrocytes
WE6NoneNoneWound epitheliumWound epitheliumWound epithelium (stratum spinosum)Wound epithelium (stratum spinosum)Wound epithelium (stratum spinosum)Wound epithelium (stratum spinosum)
α-SMABlood vesselsNoneMyofibroblastsMyofibroblasts, blood vesselsBlood vesselsBlood vesselsBlood vesselsBlood vessels
MMP-9Osteoclasts, isolated fibroblasts and chondrocytesOsteoclastsWound epithelium, osteoclastsBlastema cellsBlastema cellsBlastema cells, fibroblastsFibroblastsFibroblasts
C3Isolated cells (mostly keratinocytes and fibroblasts)Perinotochordal cartilage cells (autotomy only)Perinotochordal cartilage cells (autotomy only)Perinotochordal cartilage cells, cells near the epidermal downgrowths, isolated blastema cellsCells nears the Epidermal downgrowths, myocytes, isolated blastema cellsCells nears the Epidermal downgrowths, myocytes, isolated blastema cellsCells near the Epidermal downgrowths, myocytes, isolated blastema cellsIsolated cells
TGF-β3Fibroblasts, chondroblastsNoneNoneNoneNoneNoneFibroblasts, chondrocytesFibroblasts, chondrocytes
Stage I post-autotomy (0 to ∼24 hr).

In stage I, numerous tissues are exposed at the site of autotomy, including skeletal muscle, adipose tissue, the vertebra, and the spinal cord (Fig. 4A). Almost immediately, the integument begins to collapse around the autotomy surface, partially reducing the area of the open wound. As the fracture planes pass within the zigzag-shaped myosepta, detachment of the tail results in a series of deep V-shaped recesses (Fig. 1C). These recesses are initially prominent but gradually become reduced in size as the soft tissues retract from the wound. Blood loss at the autotomy surface is limited as the major blood supply to the tail, the caudal artery, is occluded by contraction of the smooth muscle sphincter bordering the autotomy surface (Fig. 4B). A small amount of leakage from the spinal artery results in the formation of an exudate clot distally adjacent to the spinal cord (Fig. 4C). Concurrent with the formation of this clot, the spinal cord retracts from the site of autotomy. Towards the end of stage I, numerous MMP-9 immunopositive osteoclasts are observed in association with the protruding vertebral remnant (Fig. 4D). At the site of autotomy, PCNA expression is restricted to cells of stratum basale (of the torn original epidermis; Fig. 4E). Although there is some evidence of cells at or adjacent to the site of autotomy undergoing apoptosis (e.g., myofibers and adipocytes), the most notable increase in C3 immunostaining is observed among perinotochordal cartilage cells adjacent to the intravertebral fracture planes (Fig. 4F). In particular, clusters of perinotochordal chondrocytes at intact (i.e., unfractured) fracture planes more proximal to the site of autotomy express C3. Cells and tissues at the site of autotomy do not immunostain for any of WE6, α–SMA or TGF-β3 at this stage.

thumbnail image

Figure 4. Stage I post-autotomy histology of the leopard gecko, Eublepharis macularius. Serial histological sections stained with Masson's trichrome (A–C) and DAB visualized immunohistochemistry with hematoxylin counterstaining (D–F). Insets are negative controls. Arrows indicate an example of an immunopositive cell. A: Sagittal section of the wound site (autotomy surface to the right). B: Higher magnification image of the distal most arterial sphincter of the caudal artery. Note this sphincter is contracted, thus preventing blood loss. C: Higher magnification image of the spinal cord retracting from the autotomy site, capped by an exudate clot. D: Matrix metalloproteinase-9 (MMP-9) immunopositive osteoclasts degrading the vertebral remnants. E: Proliferating cell nuclear antigen (PCNA) immunopositive cells of the wound epithelium. F: Activated caspase 3 (C3) immunopositive chondrocytes of the perinotochordal cartilage adjacent to the fracture plane. Scale bars: A, 200 μm; B,C, 100 μm; D–F, 50 μm. Abbreviations: ad, adipose; cl, exudate clot; it, integument; m, muscle; nc, notochord; sms, smooth muscle sphincter; sp, spinal cord; vt, vertebra; we, wound epithelium.

Download figure to PowerPoint

Stage I post-amputation (0 to ∼24 hr).

As confirmed by serial histology, all amputated tails were severed outside the fracture plane, frequently within the intervertebral joint/disc. Similar to the post-autotomy tail, stage I following amputation is characterized by the exposure of multiple tissues (Fig. 5A). However, amputated tails differ in that there is less skin available to reduce the size of the open wound and there are no obvious muscle recesses as the scalpel blade does not trace the zigzag pattern of the autotomy plane (compare Fig. 1C and D). As a result, the severed muscles directly contact, and often become embedded within, the developing exudate clot. Although blood loss following amputation is minimized by contraction of the distal-most sphincter of the caudal artery (Fig. 5B), it is still greater than that following autotomy. Hence, the exudate clot is thicker with a greater abundance of blood cells (Fig. 5C).

thumbnail image

Figure 5. Stage I post-amputation histology of the leopard gecko, Eublepharis macularius. Serial histological sections stained with Masson's trichrome (A–C) and DAB visualized immunohistochemistry with hematoxylin counterstaining (D–F). Insets are negative controls. Arrows indicate an example of an immunopositive cell. A: Sagittal section of the wound site (amputation surface to the right). B: Higher magnification image of the distal most arterial sphincter of the caudal artery. Note this sphincter is contracted, thus preventing blood loss. C: Higher magnification image of the spinal cord retracting from the autotomy site, capped by a thick exudate clot. D: Proliferating cell nuclear antigen (PCNA) immunopositive cells of the wound epithelium. E: Matrix metalloproteinase-9 (MMP-9) immunopositive osteoclasts degrading the vertebral remnants. F: There is no activated caspase 3 (C3) expression among chondrocytes of the perinotochordal cartilage adjacent to the fracture plane. Scale bars: A, 200 μm; B,C, 100 μm; D–F, 50 μm. Abbreviations: ad, adipose; cl, exudate clot; fp, fracture plane; it, integument; m, muscle; nc, notochord; sms, smooth muscle sphincter; sp, spinal cord; vt, vertebra; we, wound epithelium.

Download figure to PowerPoint

As for the autotomized tails, cell proliferation at the site of amputation (as evidence by PCNA immunostaining) is restricted to basal cells of the collapsing epidermis (Fig. 5D). Furthermore, as stage I continues various MMP-9 immunopositive osteoclasts are observed associated with the amputated vertebra (Fig. 5E). Although isolated cells of the original epidermis and dermis are immunopositive for C3, perinotochordal cells adjacent to the intact fracture planes are not (Fig. 5F). There are also no WE6, α-SMA or TGF-β3 immunostaining cells in regenerating tissue at this stage.

Stage II post-autotomy (∼1–7 days).

Compared with stage I, the wound site at post-autotomy stage II is capped by a thicker exudate clot clearly composed of crystallized tissue fluid, tissue debris and blood cells (Fig. 6A). Deep to the clot, the wound epithelium is beginning to span the autotomy surface (Fig. 6B), although re-epithelialization is not completed until stage III (McLean and Vickaryous, 2011). Cells of the wound epithelium are immunopositive for PCNA (Fig. 6C), WE6 (Fig. 6D), and MMP-9 (Fig. 6E). Remnants of the autotomized vertebra (including neural and transverse processes) remain embedded within the exudate clot (Fig. 6F), associated with numerous osteoclasts. Stage II marks the first appearance of mesenchymal-like blastema cells. These PCNA immunopositive, non-differentiated cells are localized immediately distal to the retracted spinal cord. α–SMA expressing myofibroblasts are abundant throughout the dermis (Fig. 6G). C3 immunostaining is observed among various isolated cells of the original tissue, including clusters of perinotochordal chondrocytes at the fracture planes. No TGF-β3 immunopositive cells are present in the regenerating tissue at this stage.

thumbnail image

Figure 6. Stage II post-autotomy histology of the leopard gecko, Eublepharis macularius. (A,B,F) Serial sections stained with Masson's trichrome. (C–E,G) Immunohistochemistry visualized with DAB and hematoxylin counterstaining. Insets are corresponding negative controls. Arrows indicate an example of an immunopositive cell. A: Sagittal section showing the tissue organization and wound site (autotomy surface to the right). B: Higher magnification image of the wound epithelium developing deep to the exudate clot. Cells of the wound epithelium are immunopositive for proliferating cell nuclear antigen (PCNA) (C), WE6 (D) and matrix metalloproteinase-9 (MMP-9) (E). F: Higher magnification image of the exudate clot at the distal tip of the torn spinal cord. White arrow indicates vertebral remnant embedded in exudate clot. G: α-smooth muscle actin (SMA) immunopositive myofibroblast within original dermis adjacent to the autotomy surface. Scale bars: A, 200 μm; B, 100 μm; C,G, 20 μm; D,E, 50 μm; F, 200 μm. Abbreviations: ad, adipose tissue; cl, exudates clot; it, integument; m, skeletal muscle; nc, notochord; sp, spinal cord; we, wound epithelium.

Download figure to PowerPoint

Stage II post-amputation (∼1–7 days).

Unless otherwise stated, the histology of a stage II post-amputation tail closely matches that of a stage II post-autotomy tail. The amputation wound site is completely sealed by a thick exudate clot (Fig. 7A). This clot is distinctly richer in red blood cells when compared to the stage II post-autotomy tail. The wound epithelium has begun to form, passing deep to the clot. Unlike that of the post-autotomy tails, the post-amputation wound epithelium has a markedly convoluted appearance (Fig. 7B). However, wound epithelial cells demonstrate comparable immunoreactivity for PCNA (Fig. 7C), WE6 (Fig. 7D), and MMP-9 (Fig. 7E). α–SMA immunopositive myofibroblasts are also present in this stage (Fig. 7F). Although remnants of the amputated vertebra do not protrude from the wound site (Fig. 7A), they are associated with MMP-9 immunopositive osteoclasts (Fig. 7G). C3 immunostaining remains restricted to isolated cells of the original tissue and is not observed among perinotochordal chondrocytes. There is no evidence of TGF-β3 immunostaining among the regenerating tissues.

thumbnail image

Figure 7. Stage II post-amputation histology of the leopard gecko, Eublepharis macularius. (A,B) Serial histological sections stained with hematoxylin and eosin. (C–G) Immunohistochemistry visualized with DAB and hematoxylin counterstaining. Insets are corresponding negative controls. Arrows indicate an example of an immunopositive cell. A: Sagittal section demonstrating the tissue organization and wound site (amputation surface to the right). B: Higher magnification images of the amputation surface showing the wound epithelium formed deep to the exudate clot. Cells of the wound epithelium are immunopositive for proliferating cell nuclear antigen (PCNA) (C), WE6 (D), and matrix metalloproteinase-9 (MMP-9) (E). F: α-smooth muscle actin (SMA) immunopositive myofibroblasts within original dermis adjacent to the autotomy surface. G: MMP-9 immunopositive osteoclasts observed adjacent to the vertebral remnant. Scale bars: A, 200 μm; B, 100 μm; C–G, 20 μm. Abbreviations: cl, exudate clot; it, integument; m, skeletal muscle; nc, notochord; sp, spinal cord; we, wound epithelium.

Download figure to PowerPoint

Stage III post-autotomy (∼7–10 days).

Stage III is characterized by the loss of the exudate clot to reveal the completed wound epithelium (Fig. 8A). Although initially only 2–3 cell layers thick, the wound epithelium becomes increasingly stratified throughout this stage. In particular, it becomes thickest at the distal tip or apex of the regenerating tail, forming what is termed the apical epithelial cap (AEC; Alibardi, 2010) (Fig. 8B). The boundary between the regenerating wound epithelium and the original epidermis is marked by prominent involuting structures known as epidermal downgrowths (Fig. 8C). Deep to the wound epithelium is the blastema, an aggregation of proliferating (PCNA immunoreactive) morphologically homogeneous, mesenchymal-like cells (Fig. 8D). These mesenchymal-like cells are weakly immunopositive for MMP-9 (Fig. 8E) and isolated cells are immunoreactive for C3 (Fig. 8F). The blastema also contains α–SMA immunopositive myofibroblasts (Fig. 8G) and blood vessels (Fig. 8H). Cells of the wound epithelium continue to express PCNA (Fig. 8I) and WE6 (Fig. 8J), but not MMP-9 (data not shown). Numerous C3 immunopositive cells are now observed in the dermis between the involuting epidermal downgrowths (Fig. 8K). TGF-β3 immunostaining remains absent from the regenerating tissue at this stage.

thumbnail image

Figure 8. Stage III post-autotomy histology of the leopard gecko, Eublepharis macularius. (A–C) Serial histological sections stained with Masson's trichrome. (D–K) Immunohistochemistry on serial sections using DAB visualization and hematoxylin counterstaining. Insets are corresponding negative controls. Arrows indicate examples of immunopositive cells. A: Sagittal histological section demonstrating the tissue organization and wound site (autotomy surface to the right). This stage is characterized by complete re-epithelialization of the autotomy surface and loss of the exudate clot. B: Higher magnification image of the distal apex of the regenerating tail, where the wound epithelium thickens to form an apical epithelial cap. C: Higher magnification image of an epidermal downgrowth, present at the interface of the original epidermis and wound epithelium. Cells of the blastema are immunopositive for proliferating cell nuclear antigen (PCNA) (D) and matrix metalloproteinase-9 (MMP-9) (E), and isolated cells are undergoing apoptosis, as evidence by C3 expression (F). α-smooth muscle actin (SMA) visualizes myofibroblasts (G) and regenerating blood vessels (H). The wound epithelium remains immunopositive for PCNA (I) and WE6 (J). Clusters of cells within the dermis adjacent to the epidermal downgrowths immunostain with activated caspase 3 (C3) (K). Scale bars: A, 200 μm; B, 10 μm; C,H, 50 μm; D–G, I–K, 20 μm. Abbreviations: ad, adipose tissue; aec, apical epithelial cap; bl, blastema; edg, epidermal downgrowth; it, integument; m, skeletal muscle; nc, notochord; sp, spinal cord; we, wound epithelium.

Download figure to PowerPoint

Stage III post-amputation (∼7–10 days).

The histology of a stage III tail post-amputation is similar to that following tail autotomy (Fig. 9A), although the distal-most vertebral remnant is relatively longer (representing just over half the original element, as compared to only a quarter for the autotomized tails). The wound epithelium is complete, smooth (i.e., no longer undulating) and, as for the post-autotomy tail, has developed an AEC (Fig. 9B) and epidermal downgrowths (Fig. 9C). As for the autotomized tail, the wound epithelium of the stage III amputated tail remains immunopositive for both PCNA and WE6 but not MMP9 (data not shown). Deep to the wound epithelium, cells of the blastema are immunopositive for PCNA (Fig. 9D) and MMP-9 (Fig. 9E), with isolated cells immunoreactive for C3 (Fig. 9F). In addition, the developing blastema includes numerous α-SMA immunopositive pericytes (Fig. 9G) and myofibroblasts (Fig. 9H). Numerous C3 immunopositive cells are observed in the dermis between the involuting epidermal downgrowths and original epidermis (Fig. 9I). No TGF-β3 expression was observed in the regenerating tissue at this stage.

thumbnail image

Figure 9. Stage III post-amputation histology of the leopard gecko, Eublepharis macularius. (A–C) Serial histological sections stained with Masson's trichrome. (D–I) Immunohistochemistry on serial sections using DAB visualization and hematoxylin counterstaining. Insets are corresponding negative controls. Arrows indicate examples of immunopositive cells. A: Sagittal histological section demonstrating the tissue organization and wound site (amputation surface to the right). This stage is characterized by complete re-epithelialization of the autotomy surface and loss of the exudate clot. B: Higher magnification image of at the distal apex of the regenerating tail, where the wound epithelium thickens to form an apical epithelial cap. C: Higher magnification image of an epidermal downgrowth, present at the interface of the original epidermis and wound epithelium. Cells of the blastema are immunopositive for proliferating cell nuclear antigen (PCNA) (D) and matrix metalloproteinase-9 (MMP-9) (E), with isolated cells undergoing apoptosis (F) (as evidenced by activated caspase 3 (C3) immunostaining). α-smooth muscle actin (SMA) visualizes regenerating blood vessels (G) and myofibroblasts (H). Clusters of cells within the dermis adjacent to the epidermal downgrowths immunostain with C3 (I). Scale bars: A, 200μm; B–I, 20μm. Abbreviations: aec, apical epithelial cap; ad, adipose tissue; bl, blastema; edg, epidermal downgrowth; it, integument; m, skeletal muscle; sp, spinal cord; we, wound epithelium.

Download figure to PowerPoint

Stage IV post-autotomy and post-amputation (∼10–15 days).

Characteristic for stage IV, the regenerated tail has begun to emerge as a convex outgrowth from the site of tail loss (Fig. 10A). As there are no histological differences between post-autotomy and post-amputation tails at this or any subsequent stage, the following descriptions apply to both forms of tail loss. The wound epithelium is beginning to keratinize and differs from the original epidermis in that it is thicker (7–12 cell layers thick, compared to 4–7 cell layers; Fig. 10B,C) and lacks pigmentation. Furthermore, there is no dermis deep to the wound epithelium. The regenerating tissue remains dominated by mesenchymal-like cells, although a rich vascular network has become established (Fig. 10D).

thumbnail image

Figure 10. Stage IV histology of the leopard gecko, Eublepharis macularius. Beginning at stage IV both the post-autotomy and post-amputation tails demonstrate a near identical histology. (A–D) Serial histological sections stained with Masson's trichrome. (E–J) Immunohistochemistry visualized with DAB and hematoxylin counterstaining. Insets are corresponding negative controls. Arrows indicate examples of immunopositive cells. A: Sagittal section demonstrating the tissue organization and blastema (regenerate tissue to the right). This stage is characterized by visible outgrowth of the blastema. Higher magnification images of the wound epithelium (B) and original epidermis (C). D: Higher magnification image of the blastema. Note the abundance of mesenchymal-like cells and blood vessels. Cells of the wound epithelium are immunopositive for proliferating cell nuclear antigen (PCNA) (E) and WE6 (F), while cells of the blastema are immunopositive for PCNA (G) and matrix metalloproteinase-9 (MMP-9) (H). This stage is also characterized by clusters of C3 immunoreactive myocytes (I). (K) α-smooth muscle actin (SMA) positive blood vessels within the newly developing tissue. Scale bars: A, 200 μm; B,C, 10 μm; D–J, 20 μm. Abbreviations: ad, adipose tissue; bl, blastema; bv, blood vessel; et, ependymal tube; it, integument; nc, notochord; oe, original tail epidermis; sp, spinal cord; we, wound epithelium.

Download figure to PowerPoint

Immunohistochemistry reveals that cells throughout the wound epithelium continue to proliferate (Fig. 10E) and remain immunopositive for WE6 (Fig. 10F), although this protein is becoming increasingly localized to cells of the stratum spinosum. The blastema remains populated by PCNA (Fig. 10G) and MMP-9 (Fig. 10H) immunopositive cells, along with the occasional cells undergoing apoptosis. Cells immunoreactive for C3 are also localized between the involuting epidermal downgrowths and original dermis and, conspicuously, among clusters of differentiating myocytes (Fig. 10I). At this stage blood vessels remain α–SMA immunopositive (Fig. 10J), but immunoreactive myofibroblasts are no longer present. Regenerating tissue remains immunonegative for TGF-β3.

Stage V post-autotomy and post-amputation (∼15–21 days).

During stage V (Fig. 11A), the wound epithelium becomes increasingly keratinized and begins to form scales. Scalation occurs as the wound epithelium invaginates into the underlying blastema, creating a series of ridges or ingrowths similar to the epidermal pegs of mammalian integument (Fig. 11B). Throughout this stage the dermis remains indistinct. Cells of the wound epithelium continue to proliferate (as evidenced by PCNA; Fig. 11C), but WE6 expression is now diffuse and largely restricted to cells of the stratum spinosum (Fig. 11D). The blastema remains heavily populated by proliferating mesenchymal-like cells and abundant α–SMA immunopositive blood vessels (Fig. 11E). Along with continued skeletal muscle development (Fig. 11F), chondroprogenitors and chondroblasts are observed forming a condensation encircling the ependymal tube. The ECM of this condensation stains positive for Alcian Blue (Fig. 11G), indicating the presence of glycosaminoglycans (GAG). Isolated cells within the blastema, as well as occasional myocytes and cells adjacent to the epidermal downgrowths remain immunopositive for C3 (data not shown).

thumbnail image

Figure 11. Stage V histology of the leopard gecko, Eublepharis macularius. Both post-autotomy and post-amputation tails exhibit a near identical histology. (A,B,F) Serial histological sections stained with Masson's trichrome and (G) Alcian blue. (C–E) Immunohistochemistry on serial sections visualized with DAB and hematoxylin counterstaining. Arrows indicate examples of immunopositive cells. A: Sagittal section of a stage V tail demonstrating regenerate tissue organization (regenerate tissue to the right). At this stage, proximal locations in the wound epithelium demonstrate the first evidence of scalation (B). Cells of the wound epithelium are immunopositive for proliferating cell nuclear antigen (PCNA) (C) and WE6 (D). α-smooth muscle actin (SMA) immunopositive blood vessels are abundant throughout the developing tissue (E). Skeletal muscle differentiation (F) continues and now cartilage formation (G) becomes obvious as a condensation of cells embedded in an Alcian blue-positive extracellular matrix. This condensation will ultimately form the regenerated skeleton. Scale bars: A, 200 μm; B, 50 μm; C–F, 20 μm; G, 100 μm. Abbreviations: cc, cartilage cone; dm, differentiating skeletal muscle; it, integument; m, skeletal muscle; sp, spinal cord; we, wound epithelium.

Download figure to PowerPoint

Stage VI post-autotomy and post-amputation (∼17–26 days).

By stage VI (Fig. 12A) the wound epithelium is no longer structurally distinct from the original epidermis. Both epithelia are stratified, have a similar thickness (approximately 4–7 cell layers), and characterized by large numbers of keratinized squamous cells (Fig. 12B,C). Although PCNA expression in both the regenerated and original epidermis is now similarly localized to cells of the stratum basale, only the newly formed epithelium is immunoreactive for WE6 (Fig. 12D). Stage VI is distinct in marking the first appearance of a fibrous connective tissue-rich dermis (Fig. 12E). Fibroblasts throughout the dermis (and other connective tissues) are immunopositive for MMP-9 (Fig. 12F), PCNA and, for the first time, TGF-β3 (Fig. 12G). TGF-β3 is also expressed by a population of chondrocytes in the regenerate cartilage (Fig. 12H). Chondrocytes of the regenerating cartilage are also immunopositive for PCNA (Fig. 12I). As for previous stages, isolated cells throughout the regenerate tissues, including myocytes and cells adjacent to the epidermal downgrowths, continue to undergo apoptosis.

thumbnail image

Figure 12. Stage VI histology of the leopard gecko, Eublepharis macularius. Both post-autotomy and post-amputation tails demonstrate a near identical histology. (A–C, E) Serial histological sections stained with Masson's trichrome. (D–I) Immunohistochemistry on serial sections visualized with DAB and hematoxylin counterstaining. Insets are negative controls. Arrows indicate examples of immunopositive cells. A: Sagittal section of a stage VI tail demonstrating regenerate tissue organization (regenerate tissue to the right). The wound epithelium (B) has a similar thickness to original epidermis (C). D: WE6 expression is becoming increasingly localized to the stratum spinosum. E: Within the regenerate dermis, nerve fascicles and blood vessels are present. Fibroblasts throughout the regenerating tail are immunopositive for matrix metalloproteinase-9 (MMP-9) (F) and, for the first time, transforming growth factor-β3 (TGF-β3) (G). Chondrocytes of the regenerated cartilaginous cone are also immunoreactive for TGF-β3 (H), as well as proliferating cell nuclear antigen (PCNA) (I). Scale bars: A, 200 μm; B–D,F–I, 20 μm; E, 50 μm. Abbreviations: bv, blood vessel; cc, cartilage cone; it, integument; nf, nerve fiber; oe, original tail epidermis; we, wound epithelium.

Download figure to PowerPoint

Stage VII post-autotomy and post-amputation (∼21+ days).

Stage VII is characterized by the complete differentiation of the epidermis, dermis, skeletal muscle, adipose tissue and cartilage (Fig. 13A). Similar to stage VI, the new epidermis (former wound epithelium) is immunopositive for PCNA (among the cells of the stratum basale; Fig. 13B) and WE6 (within the stratum spinosum; Fig. 13C). In addition, there are now melanocytes in the basal layer of the regenerate epidermis (Fig. 13D). Fibroblasts throughout the regenerated portion of the tail are immunopositive for MMP-9 (Fig. 13E), PCNA (Fig. 13F), and TGF-β3 (Fig. 13G). The regenerate skeleton is composed of a hollow cone of cell-rich and matrix-poor cartilage that attaches directly to the distal most original vertebra (Fig. 13A). This distal-most remnant is just over half the original length in the amputated tails, but only a quarter of the original length in the autotomized tails. Chondrocytes of the regenerated cartilage cone immunostain for PCNA and TGF-β3, especially among the distal-most cells (Fig. 13H). Isolated C3 immunopositive cells are also observed throughout the regenerating tissue.

thumbnail image

Figure 13. Stage VII histology of the leopard gecko, Eublepharis macularius. Both post-autotomy and post-amputation tails demonstrate a near identical histology. (A,D) Masson's trichrome stained serial sections. (B–C,E–H) Immunohistochemistry on serial sections visualized with DAB and hematoxylin counterstaining. Insets are corresponding negative controls. Arrows indicate examples of immunopositive cells. A: Sagittal section through a stage VII tail showing the regenerated tissue organization. Differentiation is now complete. Cells of the stratum basale of the regenerate epidermis are immunopositive for proliferating cell nuclear antigen (PCNA) B: while cells of the stratum spinosum immunostain for WE6 (C). D: At this stage melanocytes (pigment cells) are present within the regenerated dermis. Fibroblasts of the regenerate dermis are immunopositive for matrix metalloproteinase-9 (MMP-9) (E), PCNA (F) and transforming growth factor-β3 (TGF-β3) (G). Chondrocytes also immunostain for TGF-β3 (H). Scale bars: A, 200μm; B,C, 50 μm; D–H, 20 μm. Abbreviations: ad, adipose tissue; bv, blood vessel; cc, cartilage cone; ep, epidermis; it, integument; m, skeletal muscle; mel, melanocyte.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. HISTOLOGY
  7. DISCUSSION
  8. Acknowledgements
  9. LITERATURE CITED
  10. Supporting Information

We investigated wound healing and tail regeneration following two different modes of tail loss: autotomy at a fracture plane and involuntary amputation outside of a fracture plane. Wound healing and regeneration following both forms of tail loss are essentially identical at the level of both gross morphology and histology, with only minor differences relating to the earliest events (Table 2). Following both autotomy and amputation of the tail, the wound site heals without the formation of fibrotic tissue. Our data demonstrates that spontaneous scar-free wound healing and regeneration are intrinsic properties of the leopard gecko tail, and that these phenomena are unrelated to the mechanism (natural or surgical) or location (i.e., at or outside the fracture plane) of tail detachment.

Gross Morphology: Tail Autotomy Versus Tail Amputation

Autotomized and amputated tails follow a matching sequence of gross morphological events, beginning with the development of an exudate clot at the wound site. Almost in parallel a wound epithelium is formed, followed by a blastema. Although the origin of the cells contributing to this outgrowth remains uncertain, they are proliferative and mesenchymal-like in appearance. Ultimately this cellular mass, herein considered to represent a regeneration blastema, gives rise to the replacement tail. Interestingly, the timeframe for wound healing and regeneration for each of the two modes of tail detachment under matching environmental conditions is virtually identical, approximately 30 days.

Only two major gross morphological differences distinguish the wound site of an amputated tail compared to an autotomized tail, and these features are only observed during the earliest stages (stages I and II; see Table 2). The first difference is that amputated tails suffer a greater degree of blood loss, resulting in the formation of a thicker, more blood cell-rich exudate clot at the wound site. The second difference is that the wound site of the amputated tail is almost flat in profile, lacking the muscle recesses and protruding vertebral remnant characteristic of autotomized tails. Once the wound epithelium is complete and the exudate clot is lost (stage III) amputated and autotomized wound sites are identical. However, tails lost by autotomy detach at transverse constrictions of the integument, whereas amputated tails were detached at locations midway between these constrictions. Therefore, the mode of detachment can be retrospectively established externally based on the appearance of the integument.

Histology: Tail Autotomy Versus Tail Amputation

Serial histology verified that amputated tails were cut outside the fracture plane, and that the two forms of tail loss follow a matching sequence of histological events during wound healing and regeneration with only minor differences. In late stage II/early stage III following amputation, the wound epithelium forms an unevenly stratified undulating horizon. By late stage III however, the post-amputation wound epithelium becomes restructured into a smooth and relatively flat covering with a uniform stratification, and is thereafter comparable to that of the post-autotomy tails. During the regenerative stages (stages IV to VII) both forms of tail loss are similar histologically although they do differ in the relative length of the distal-most vertebral remnant. At autotomy, an individual tail vertebra is split midlength at the intravertebral fracture plane. Subsequent ablation by osteoclasts reduces this remnant by half. The regenerating cartilaginous cone attaches to this quartered remnant. In contrast, most amputations occur at the intervertebral joint (between adjacent vertebrae). Subsequent action by osteoclasts ablates this distal-most element to just over half its original length.

As will be discussed, the location (tissue, cell type) and pattern of temporal expression of the six proteins of interest (PCNA, MMP-9, WE6, α-SMA, TGF-β3, and C3) is highly conserved between the two different forms of tail loss.

Scar-Free Wound Healing and Regeneration Post-Tail Amputation

Our findings confirm and enhance previous observations on the ability of (at least some) lizards to undergo scar-free wound healing and regeneration following tail amputation (e.g., Woodland, 1920; Werner, 1967). Of particular interest is the work of Whimster (1978), who first reported on the ability of the leopard gecko to regeneration following tail amputation. In addition, Whimster also demonstrated the importance of the wound epithelium through a series of grafting experiments. Geckos receiving a graft of original integument covering the amputation surface failed to regenerate whereas individuals not receiving a graft did. It is worth noting however, that Whimster did not record if the location of tail loss was at or outside of the fracture plane.

Similar to leopard geckos, tail regeneration has been documented following amputation in the gekkotan taxon Hemidactylus (Woodland, 1920; Werner, 1967) and the lacertid Lacerta (Moffat and Bellairs, 1964). For other lizards, this ability may be variable. For example, a comparative study exploring intravertebral autotomy and intervertebral amputation in five species (Anolis carolinensis, Eumeces fasciatus, Lygostoma laterale, Gerrhonotus multicarinatus, and Sceloporus undulatus) yielded mixed results (Jamison, 1964). In all five taxa, amputation of the tail resulted in a diminished or absent regenerative response compared with species-matched autotomy controls. Curiously, these findings contradict other investigations using A. carolinensis, which reported that post-amputation wound healing and regeneration was similar to that of the post-autotomy tail (Barber, 1944; Maderson and Licht, 1968). To date, these conflicting observations await explanation.

Overall, the sequence of gross morphological and histological events following tail loss in the leopard gecko closely resembles scar-free wound healing and regeneration as described for urodeles (e.g., Iten and Bryant, 1973; Tank et al., 1976; Nye et al., 2003; Lévesque et al., 2010). Following amputation, urodeles develop a wound epithelium, including an apical epithelial cap, thus sealing the injury without tissue fibrosis. Deep to the wound epithelium, a mass of mesenchyaml-like undifferentiated blastema cells begins to accumulate. As blastema cells proliferate, a new conical outgrowth is formed, ultimately giving rise to the new tissues of the regenerate appendage. Also similar to leopard geckos, the regenerative capacity of urodeles is an intrinsic phenomenon, not restricted to a particular location within an appendage. However, unlike the leopard gecko, regeneration in urodeles results in a structurally perfect replacement, indistinguishable from the original. In the leopard gecko the regenerate tail differs in the pattern of scalation, arrangement of skeletal muscle, and the structure of the central nervous system and skeleton (McLean and Vickaryous, 2011). Interestingly, under experimental conditions, regeneration of both the urodele appendage and leopard gecko tail requires a similar timeframe, approximately 30 days (Iten and Bryant, 1973; Tank et al., 1976; McLean and Vickaryous, 2011).

Protein Expression Following Tail Loss

Consistent with the conserved sequence of gross morphological and histological events, immunohistochemical data revealed a conserved pattern of expression of the six proteins of interest following both autotomy and amputation. As evidenced by PCNA expression, wound epithelial and blastema cells are highly proliferative during the early stages of regeneration (stages I to III). As tissues in the blastema begin to differentiate (stages IV and onwards), other cell types are observed to be immunoreactive for PCNA, including ependymal cells, myoblasts, endothelial cells, fibroblasts, and chondroblasts. These findings match those previously reported for the leopard gecko (McLean and Vickaryous, 2011), and are consistent with other data gleaned from other regeneration models (e.g., Santos-Ruiz et al., 2002).

In both urodeles and leopard geckos WE6 is a marker of wound epithelial cells. During stages II and III in the leopard gecko, all cells of the wound epithelium exhibit strong immunostaining for WE6. As outgrowth of the regenerate tail begins (stages IV to VII), WE6 expression becomes irregular and progressively weaker in intensity. WE6 was originally identified in the wound epithelia of red-spotted newts (Notopthalmus viridescens) following limb amputation (Estrada et al, 1993). In red-spotted newts, WE6 expression was upregulated by cells of the wound epithelium, but not in unaltered (control) epidermis. In leopard geckos, WE6 immunostaining also occurs in wound epithelial cells contributing to the epidermal downgrowths following tail amputation and autotomy. Previous work has documented WE6 expression following both tail autotomy and limb amputation in the wall lizard, Podacris muralis (Alibardi and Toni, 2005). WE6 expression was only observed among wound epithelial cells of the regenerating tail, and only once the wound epithelium was complete across the wound site. At the site of limb amputation, scar tissue formed and there was no immunoreactivity for WE6. In leopard geckos, WE6 is expressed prior to and following the complete formation of the wound epithelium across the wound site.

MMP-9, a protease involved in the remodeling of the extracellular matrix, is expressed during scar-free wound healing in the leopard gecko. Cells immunopositive for MMP-9 include osteoclasts, cells of the wound epithelium, and blastema cells. Previous work has documented MMP-9 mRNA expression during limb regeneration in urodeles (Yang et al., 1999). In urodeles, in situ hybridization reveals a biphasic expression of MMP-9 transcripts during limb regeneration. The first phase occurs 2 hr to 2 days (peaking at hour 14) following limb amputation, with MMP-9 expression localized to the developing wound epithelium (Yang et al., 1999). These findings suggest that MMP-9 expression in urodeles prevents basal lamina formation and thus promotes epithelial-mesenchymal interactions. Subsequently, during the second phase (from the medium bud stage until redifferentiation begins), MMP-9 expression is localized to cells at the severed margins of the skeletal elements suggesting a role in skeletal degradation. Data from leopard geckos is consistent with these results. More specifically, the developing wound epithelium and cells associated with the damaged margin of the vertebrae (osteoclasts) are MMP-9 immunopositive. Unlike urodele blastema cells, mesenchymal-like cells in the regenerating leopard gecko tail are also MMP-9 positive. At present, it remains unclear why MMP-9 expressing cells are found in the regenerating leopard gecko tail, but not the blastema of urodeles. One possibility is that urodeles express other MMPs with similar substrate specificity to MMP-9 (e.g., MMP-2) to facilitate similar functional roles (e.g., ECM degradation, activation of ECM bound growth factors).

We used α-SMA expression to document myofibroblasts and pericytes. Immunopositive myofibroblasts are initially abundant during stages II and III but effectively disappear shortly after the wound epithelium is complete. In mammals, myofibroblasts have two main roles during wound healing: contraction to minimize wound size; and collagen secretion to facilitate tissue fibrosis and scar formation (Desmoulière et al., 2005; Hinz, 2007; Wynn, 2008). Interestingly, while myofibroblasts are initially common at the site of tail loss in leopard geckos, collagen deposition at the wound site is limited (as evidenced by trichome histochemistry). Among juvenile axolotls (A. mexicanum), α-SMA expression has been documented following cutaneous excisional wounding (Lévesque et al., 2010). However, while blood vessel walls were clearly α-SMA immunopositive (indicating the presence of pericytes), myofibroblasts appeared to be absent from the wound site. The absence of wound-associated myofibroblasts in urodeles is not fully understood, but points to a fundamental difference in the process of wound healing of this group compared to lizards and mammals.

C3 is the effector of both the intrinsic and extrinsic caspase-mediated apoptotic pathways, and is considered to be a reliable marker of programmed cell death (Lamkanfi et al., 2007). Prior to and throughout tail regeneration, apoptosis occurs within individual cells of the dermis and epidermis. Following both forms of tail loss, C3 is dynamically expressed by mesenchymal-like cells and multiple differentiating cell types. Although mostly limited to isolated cells, apoptosis is most common among cells adjacent to the epidermal downgrowths (stages III to VI) and developing myocytes (stage IV). Overall, post-autotomy and post-amputation tails demonstrate a similar distribution of C3 immunoreactive cells, with one major difference. Unique to the autotomized tails, apoptosis was observed among the perinotochordal chondrocytes surrounding the fracture plane, suggesting a role for programmed cell death during the act of autotomy. Although presently untested, it appears that the application of a relatively broad (i.e., finger width) external force to induce autotomy leads to localized apoptosis of perinotochordal chondrocytes at fracture planes. As a result, the structural integrity of multiple intravertebral fracture planes is compromised, thus facilitating tail detachment. In contrast, this scenario is not initiated when the tail is surgically transected using a scalpel blade.

This investigation is the first to document TGF-β3 expression during lizard tail regeneration. Beginning in stage VI, TGF-β3 was expressed by chondrocytes of the regenerating axial skeleton (the cartilage cone). Subsequently, at stage VII, TGF-β3 immunopositive fibroblasts were also observed in the regenerating dermis. Several previous investigations have demonstrated a role for TGF-β3 in scar-free wound healing among mammals, including embryos and adults (e.g., Shah et al., 1995; Metcalfe and Ferguson, 2007; Ferguson et al., 2009). Thus, it is somewhat unexpected that leopard geckos do not demonstrate evidence of TGF-β3 expression at the wound site until the later stages of regeneration (stages VI and VII). One possibility is that one of other TGF-β isoforms (e.g., TGF-β1 or 2) may play a more prominent role during wound healing. For example, during wound healing following limb amputation in the axolotl (A. mexicanum), TGF-β1 mRNA is initially abundant (Lévesque et al., 2007). Furthermore, application of the TGF-β Type 1 receptor inhibitor SB-431542 (an inhibitor of TGFβ dependent Smad signaling) results in complete abolishment of regeneration (Lévesque et al., 2007). Amputated limbs treated with the inhibitor developed a wound epithelium, albeit slower than controls, but cell proliferation was significantly decreased and no distinct blastema was formed.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. HISTOLOGY
  7. DISCUSSION
  8. Acknowledgements
  9. LITERATURE CITED
  10. Supporting Information

The authors would like to thank H. Coates, K. McLean, S. Payne, E. Gilbert, and J. Vieira for assistance in the lab; Drs. A. Hahnel and A. Viloria-Petit and three anonymous reviewers for the comments that improved this manuscript. In addition, they are grateful to Drs. R. Fisher and K. Kusumi for the invitation to participate in this volume. The monoclonal antibodies WE6, developed by R.A Tassava, and D- B3(act) (TGF- β3), developed by R. Runyan, were obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biology, Iowa City, IA 52242.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. HISTOLOGY
  7. DISCUSSION
  8. Acknowledgements
  9. LITERATURE CITED
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. HISTOLOGY
  7. DISCUSSION
  8. Acknowledgements
  9. LITERATURE CITED
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
AR_22490_sm_SuppInfo.doc25KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.