Decreased expression of p63, a regulator of epidermal stem cells, in the chronic laminitic equine hoof



Reasons for performing study: Abnormal epidermal stem cell regulation may contribute to the pathogenesis of equine chronic laminitis.

Objective: To analyse the involvement of p63, a regulator of epidermal stem cell proliferative potential, in chronic laminitis.

Methods: Epidermal tissues from skin, coronet and lamellae of the dorsal foot were harvested from 5 horses with chronic laminitis and 5 control horses. Tissues were analysed using histopathology, immunofluorescence microscopy and quantitative immunoblotting

Results: Hoof lamellae of laminitic horses had a lower frequency of p63 positive cells than control lamellae, particularly in the distal region. Quantitative immunoblotting confirmed reduced p63 expression in the laminitic distal lamellar region. The decreased p63 expression in laminitic epidermal lamellae was most apparent in the abaxial region adjacent to the hoof wall and highly associated with the formation of terminally differentiated, dysplastic and hyperkeratotic epidermis in this region, whereas lamellae from control horses maintained high p63 expression throughout the axial-abaxial axis.

Conclusions: Expression of p63 in equine skin resembles that reported in other species, including man and rodents, suggesting that p63 can serve as a marker for the proliferative potential of equine epidermal stem cells. p63 expression was significantly lower in the chronic laminitic hoof than in that of control horses, suggesting laminitic hoof epithelium has more limited proliferative potential with a shift towards differentiation. This may reflect reduced activity of epidermal stem cells in laminitic hoof. It is proposed that p63 contributes to the maintenance of hoof lamellae and that misregulation of p63 expression may lead to epidermal dysplasia during lamellar wedge formation.

Potential relevance: This study suggests that loss of epidermal stem cells contributes to the pathogenesis of equine laminitis. Autologous transplantation of p63-positive epidermal stem cells from unaffected regions may have regenerative therapeutic potential for laminitic horses.


Basement membrane




4′,6-Diamidino-2-phenylindole, dilactate


Epidermal cell


Epidermal stem cell


Haematoxylin and eosin


Integrated optical density


Keratin axis




Mural sensitive epidermal and dermal depth


Primary epidermal lamella


Secondary epidermal lamella


Despite the importance of the hoof in equine health, the cell biology of this epidermal epithelium is poorly understood. The hoof most likely shares some aspects of epidermal stem cell (ESC) biology with other epidermal accessory organs, such as scales, feathers and claws, as well as with the skin from which it is derived (Bragulla and Hirschberg 2003). In epidermal tissues studied to date, ESCs located within either the basal cell layer or bulge region of the hair follicle are responsible for the maintenance of tissue homeostasis and repair following injuries (Blanpain and Fuchs 2009). The presence of ESCs is also critical for the success of therapeutic autografting of epidermal tissues (De Luca et al. 2006).

Tissue homeostasis is possible due to the presence of ESCs, which have the capacity both for self-renewal and generation of differentiated cells of the epidermal tissue (Blanpain and Fuchs 2009). In the case of the equine hoof, continuous growth is necessary to replace hoof lost at the ground surface (Daradka and Pollitt 2004). This growth is supported mainly by proliferation of epidermal cells (ECs) of proximal (coronary) regions of the hoof, with very low levels of proliferation in mid-lamellar regions, as demonstrated in ponies using 5-bromo-2′-deoxyuridine (BrdU) incorporation (Daradka and Pollitt 2004).

Chronic laminitis is a painful and debilitating disease of horses characterised by biomechanical compromise or failure of the attachment between the epidermal and dermal lamellae of the foot, resulting in palmar/plantar displacement of the distal phalanx (Hood1999). In contrast to nonlaminitic horses, horses with chronic laminitis often have impaired or abnormal hoof growth and accumulation of dysplastic lamellar ECs resulting in the formation of a ‘lamellar wedge’ that contributes to the pathology of this disease (Collins et al. 2010). The presence of the lamellar wedge and the ease with which nonlaminitic hoof regrows following hoof wall resection (Pollitt and Daradka 2004) indicates that the normally quiescent mid-dorsal lamellar ECs are capable of rapid proliferation upon injury. This suggests that epidermal proliferative potential is maintained within a resident ESC population (Blanpain and Fuchs 2009).

The role of ESCs in hoof growth, homeostasis and laminitis has not previously been examined. The transcription factor p63 is a key, lineage-specific determinant of the proliferative capacity in stem cells of stratified epithelia (Senoo et al. 2007). The p63 gene generates multiple isoforms with distinct biological roles (Yang et al. 1998), of which ΔNp63α plays an essential role in epithelial stem cells (Senoo et al. 2007). The objective of this study was to analyse the involvement of ESCs in the normal equine hoof and in chronic laminitis. Specifically, our aims were to 1) examine the usefulness of p63 as an ESC marker in equine epidermal tissues and 2) compare p63 expression levels in control and chronic laminitic horses. We hypothesised that ESCs are present in the hoof epidermis in order to support hoof homeostasis and to respond to injury and that p63 is expressed in basal ECs, with the greatest frequency in coronary and proximal hoof regions. We also hypothesised that chronic laminitic horses show aberrant p63 expression compared to normal horses. To our knowledge, this is the first report of the localisation of ESCs, detected by p63 expression, to specific regions of the normal and laminitic equine hoof.

Materials and methods


All subjects were chosen based on data and samples collected as part of a laminitis tissue repository (Galantino-Homer et al. 2010) between October 2008 and June 2009. Five laminitic subjects were chosen based on clinical history and distal phalanx displacement. Clinical history for laminitis cases included chronic (>72 h) lameness of Obel grade 1 or greater as diagnosed by a veterinarian and exposure to one or more contributing aetiological factors associated with laminitis. The degree of distal phalanx displacement was measured at necropsy and defined as the distance between the dorsal surface of the distal phalanx and the inner limits of the stratum medium, referred to as mural sensitive epidermal and dermal depth (MSEDD). Distal phalanx displacement in laminitic cases was identified as a distal MSEDD that was ≥2 mm greater than proximal MSEDD. Control subjects were chosen based on clinical history and no evidence of distal phalanx displacement, identified as a distal MSEDD that was not greater than proximal MSEDD. Control horses had no clinical history of laminitis. Musculoskeletal pathologies were present in some control horses; however, horses with nonweightbearing lameness or septic conditions were excluded due to the association of these conditions with an increased risk of laminitis (Parsons et al. 2007). Selection of control subjects was not based solely on histopathology, given that it is not uncommon for clinically normal horses to have mild or focal lesions on the sensitive lamellae (Kawasako et al. 2009) and, conversely, mild or focal histopathological changes alone may not justify a diagnosis of laminitis. Each case was matched to one control for similarity in age and the presence of one or more pituitary adenomas at necropsy; however, breed or sex could not be matched with the samples that were available. Horses were subjected to euthanasia by pentobarbital and phenytoin overdose, according to procedures approved by the University of Pennsylvania's Institutional Animal Care and Use Committee.

Tissue retrieval

Feet were disarticulated at the metacarpophalangeal joint and processed within 2 h of euthanasia. Parasagittal sections (1–2 cm thick) were made on either side of the dorsal midline of the foot using a band saw. Mural sensitive epidermal and dermal depth was measured from the dorsal surface of the distal phalanx to the inner limits of the nonpigmented portion of the stratum medium (i.e. the hoof wall) in proximal, middle and distal lamellar areas. Lamellar and coronary tissues were isolated as previously described (Pollitt 1996) and skin tissue obtained dorsally, 1–10 cm above the coronary band.


Six μm thick sections of formalin-fixed, paraffin-embedded tissue were stained with haematoxylin and eosin (H&E) or periodic acid-Schiff stains and used for histology. Lamellar sections were evaluated and categorised by a board-certified pathologist (J.B.E.; American College of Veterinary Pathologists) based on the following criteria: degree of elongation and distortion of primary and secondary epidermal lamellae (PEL and SEL, respectively); presence of epidermal basal/parabasal hyperplasia, acanthosis and hyperkeratosis; presence of abaxial displacement of the keratinised axis (KA) with or without merger of adjacent KA; type and degree of basement membrane (BM) pathology (separation, splitting and isolated epidermal islands) and degree of dermal pathology (vascularisation, fibrosis, necrosis, inflammation). Histological categorisation was based on the following definitions: normal was defined by no lamellar pathology present, minimal was defined by lamellar changes that were mild but very focal, mild was defined by mild changes that were generalised (multifocal to diffuse), mild to moderate was defined by generalised mild lamellar pathology with focal to multifocal moderate changes and moderate to severe was defined by generalised moderate lamellar pathology with focal to multifocal severe lamellar changes.

Indirect immunofluorescence

Eight µm cryosections of paraformaldehyde-fixed, sucrose-dehydrated tissue were incubated with primary antibodies followed by Alexa Fluor 594 rabbit anti-mouse IgG1 (1:500) and Alexa Fluor 488 goat anti-rabbit IgG1 (1:1000) staining. Subsequently, sections were counterstained with 4′,6-diamidino-2-phenylindole, dilactate (DAPI; 0.5 µg/ml)1. The primary monoclonal antibodies used in this study were mouse anti-p63 (clone 4A4; 1:500)2, rabbit anti-Ki-67 (clone SP6; 1:100)3, mouse anti-keratin-14 (K14) (clone LL002; 1:50)4 and mouse anti-involucrin (clone SY5; 1:100)2. As isotype controls, mouse anti-IgG2a (clone MG2a-53; 1:125)5, rabbit anti-GFP (clone FL; 1:100)6 and mouse IgG3 (clone MG3-35; 1:200)5 were used.

Images were acquired with a digital camera on a fluorescence microscope using a 20x objective. For each image, all ECs positive for p63 and/or Ki-67 were manually counted and expressed as a percentage of the total EC count identified by DAPI staining.


Frozen tissues were pulverised (Bio-pulverizer)7 and proteins extracted in SDS buffer (4% SDS, 120 mmol/l Tris base, 20% glycerol, protease inhibitors (Complete Protease Inhibitor Cocktail Tablets and Pefabloc SC)8, pH 6.8), followed by boiling for 5 min and centrifugation at 13,000 ×g at 4°C for 20 min to collect the supernatant. Total protein concentration was measured using a Microplate BCA Protein Assay Kit9 in triplicate with bovine serum albumin as a standard.

Proteins were separated by SDS-PAGE and transferred to polyvinylidene difluoride membrane. Immunoblots were blocked with 5% fish gelatin and incubated with anti-p63 antibody (1:500) or anti-K14 antibody (1:500), followed by horseradish peroxidase conjugated anti-mouse IgG antibody10 (1:5000) to visualise antigens with enhanced chemiluminescence (Amersham ECL kit)11. Scanned images were analysed using imaging software (Gel-Pro 4.5).12

Statistical analysis

Results are presented as mean ± s.d., unless otherwise stated. Data obtained from immunofluorescence staining were analysed by 2 ANOVA models using statistical analysis software (Intercooled Stata Version 9.2).13 The Shapiro-Wilk test (α= 0.01) was used to confirm normality of variables within each group and tissue type. In ANOVA model 1, the effects of laminitis group (laminitis, control), tissue type (skin, coronary, proximal, middle, distal) and their interaction on the mean percentage of p63 or Ki-67 positive cells for each tissue type was evaluated, with horse nested within laminitis group. In ANOVA model 2, the effects of laminitis group, lamellar tissue type (proximal, middle, distal), section within each lamellar tissue type (axial, central, abaxial) and their interaction on the percentage of p63 or Ki-67 positive cells was evaluated, with horse nested within laminitis group. In these models, tissue type and section within lamellar tissue were treated as within-subject factors, and laminitis group was the between-subject factor. Tukey's highly significant difference analysis was used for pairwise comparisons, and significance was set at P<0.05. Data for coronary tissue of one control horse (C1) were excluded from analysis due to inappropriate tissue collection. Integrated optical density (IOD) of each band on immunoblots was measured and reported relative to the mean IOD for the 5 control horses. Inter-group comparisons of age and IOD were determined by independent t test.


Gross pathology and histopathology

History and gross pathology for individual subjects are summarised in Table 1. Mean ± s.d. age of control horses (14 ± 9 years) was not different from that of the laminitic cases (17 ± 12; P = 0.65). The MSEDD was 6.6 ± 1.1, 5.8 ± 1.3, and 5.3 ± 1.0 mm for proximal, middle, and distal lamellar tissue of control horses, and 6.7 ± 2.5, 9.6 ± 3.5, and 11.9 ± 5.5 mm for laminitic cases. Distal MSEDD for laminitic cases were greater than proximal MSEDD (P = 0.030) and were greater than distal MSEDD of control horses (P = 0.048). Mural sensitive epidermal and dermal depth for one control horse (C1) were not available.

Table 1. Subject details, laminitis history and gross pathology for control horses (C1–C5) and laminitic cases (L1–L5) used in the present study
HorseAge, yearSexBreedLaminitis historyGross pathology
  1. TB = Thoroughbred; WB = Warmblood; STB = Standardbred; MOR = Morgan; LF = left front; RF = right front; LH = left hind.

C129MareTBNo history of laminitisPituitary adenoma
C217GeldingTBNo history of laminitisPituitary adenoma, arthritis
C311GeldingWBNo history of laminitisBladder tumour, navicular fracture in contralateral front foot
C47GeldingTBNo history of laminitis‘Bone chip’ (small osteochondral fracture, presumptive)
C56GeldingWBNo history of laminitisTorn meniscus (stifle)
L133MareMORChronic laminitis for 3 weeksPituitary adenoma, suspected insulin resistance, corticosteroid treatment for recurrent airway obstruction prior to onset of laminitis
L225MareSTBRecurrent chronic laminitis for >1 yearPituitary adenoma, suspected insulin resistance, routine pergolide treatment
L317GeldingTBChronic laminitis for 2 weeksHoof abscess (RF) with laminitis in contralateral front foot (LF)
L46MareTBChronic laminitis for 6 weeksRetained fetal membranes with metritis
L55StalllionTBChronic laminitis for 11 weeksCellulitis in LH hock with laminitis in all other feet

Histopathology for each subject used in the present study is summarised in Table 2. In general, lamellar tissue from laminitic horses displayed more severe elongation and distortion of PELs and SELs compared to control horses. Laminitic horses also showed lamellar epidermal hyperplasia and acanthosis with hyperkeratosis, abaxial displacement of the KA with frequent merger of adjacent abaxial lamellae and BM pathology, including BM separation, splitting, duplication and epidermal islands. Although varying in severity and location among the cases, only laminitic animals showed dermal pathology, including necrosis, inflammation, neovascularisation and fibrosis.

Table 2. Histopathology of hoof lamellae in control horses (C1–C5) and laminitic cases (L1–L5)
HorseLamellar histopathology
  1. PEL/SEL = primary/secondary epidermal lamellae; BM = basement membrane; KA = keratinised axis; axial refers to the inner most portions of the sensitive lamellae (e.g. those closest to dorsal distal phalanx); abaxial refers to the outer most portions of sensitive lamellae (e.g. those just beneath the inner nonpigmented stratum medium).

C1Mild axial PEL and SEL distortion
C2Mild axial PEL and SEL distortion
C3Minimal PEL and SEL distortion
C4Mild PEL and SEL elongation and distortion
L1Moderate-severe PEL and SEL elongation and distortion; moderate-severe PEL hyperplasia, acanthosis, hyperkeratotis; multifocal BM separation; multifocal dermal necrosis with serum lakes (abaxial regions)
L2Mild-moderate PEL and SEL elongation and distortion; mild BM changes with multifocal epidermal islands; mild dermal fibrosis
L3Moderate-severe PEL elongation; moderate-severe PEL hyperplasia, acanthosis and hyperkeratosis; abaxial displacement of primary KA with merger of adjacent abaxial KA; multifocal SEL elongation and distortion; multifocal dermal lamellar necrosis (abaxial regions)
L4Moderate-severe PEL distortion and elongation; moderate-severe PEL hyperplasia and acanthosis; abaxial displacement of KA; multifocal SEL elongation, loss, and distortion; mild splitting/duplication of BM with epidermal islands; mild dermal inflammation, fibrosis and neovascularisation
L5Severe PEL distortion and elongation with merger of adjacent abaxial KA; abaxial PEL hyperplasia with severe orthokeratotic hyperkeratosis; SEL shortened, blunted; mild BM splitting and duplication with multifocal epidermal islands; mild dermal oedema and haemorrhage with moderate inflammation, fibrosis and neovascularisation

Expression of p63 in skin epithelium

Expression of p63 was found in the nuclei of basal and suprabasal ECs in the skin (Fig 1a). The levels of p63 expression within the basal cell layer varied and gradually decreased in the more differentiated cells of the suprabasal layers (Fig 1b). More specifically, p63 was highly expressed in putative ESCs in the basal layer marked by K14 staining (Figs 1b,c). Presumed transit-amplifying cells, immediate derivatives of ESCs, expressed moderate levels of p63, while terminally differentiated cells, marked by a terminal differentiation marker involucrin, were p63-negative (Figs 1b,d). Isotype-specific primary antibodies or secondary antibodies alone revealed no specific signals, suggesting that all of the antibodies that were employed in this study can detect specific target proteins (data not shown).

Figure 1.

Expression of p63 in equine skin epidermis. (a–d) Sections from normal horse skin were stained with antibodies to p63 (a, b), K14 (c) and involucrin (d) followed by Alexa Fluor 594 rabbit anti-mouse IgG antibody. In panels a and b, nuclei were counterstained with DAPI. Boxed area in panel a is enlarged in panel b, showing variable expression levels of p63 along differentiation axis from basal layer toward more differentiated suprabasal layers, as well as amongst basal cells. Arrow indicates an example of putative epidermal stem cell based on high-p63 expression, while solid arrowhead and open arrowhead indicate a transit-amplifying cell (moderately p63-positive) and a terminally differentiated cell (p63-negative), respectively. (e) Immunoblot analysis of p63 in equine skin, coronet, and proximal, middle and distal lamellae. Human epithelial cell line ME180, expressing high levels of the ΔNp63α isoform, was used as a size control. Dashed lines indicate epidermal-dermal border. Scale bars = 40 µm.

In order to confirm isoform specificity of p63 in equine epidermis, immunoblot analysis was performed with protein extracts from equine skin and hoof lamellae tissues along with human cervical carcinoma cell line ME180 that expresses high levels of the ΔNp63α isoform. As shown in Figure 1e, anti-p63 antibody detected a band of the same size (approximately 80 kDa) corresponding to the ΔNp63α isoform, in all equine epidermal samples examined. Control immunoblots with secondary antibody alone revealed no specific signals (data not shown).

Expression of p63 in coronary epithelium and secondary epidermal lamellae

Similar to the skin, the basal cell layer of coronary and lamellar tissues was identified by K14 expression in the cytoplasm of basal ECs (Figs 2b,f,j,n), with less intensity in suprabasal cells of coronary tissues (Fig 2b). In both coronary and lamellar tissues, p63 was expressed in the nuclei of basal and suprabasal ECs, with inter-cell staining variability similar to the pattern in the skin (Figs 2c,d,g,h,k,l,o,p). Interestingly, expression of p63 differed significantly across tissue type (P<0.001), with the percentage of p63 positive ECs greatest in coronary tissue and decreasing distally in lamellar epithelium. It is noteworthy that a subset of cells in the basal layer expressed high levels of p63 while others showed intermediate to low expression of p63 (Figs 2d, h, l, p), suggesting that these high p63-positive cells may function as ESCs with high proliferative capacity.

Figure 2.

Expression of p63 and K14 in coronary epithelium and the secondary epidermal lamellae of the equine foot. Representative images for haematoxylin and eosin (H&E) staining a), e), i), m), K14 staining b), f), j), n) and p63 (red) staining merged with DAPI (blue) counterstaining c), g), k), o) are shown for coronet a–d) and proximal e–h), middle i–l) and distal m–p) lamellar tissues from control horse C2. Panels d, h, l and p are enlarged views of boxed areas in panels c, g, k and o, respectively. White arrows indicate high p63-positive cells, while solid and open arrowheads indicate moderately p63-positive and low-negative cells, respectively. Asterisks indicate nonspecific autofluorescence by keratinised lamellae. e, epidermal epithelium; p, the keratinised axis of the primary epidermal lamellae; s, the secondary epidermal lamellae. Scale bars = 50 µm.

Effects of laminitis on p63 expression

The percentage of p63 positive cells was lower (P<0.05) in laminitic compared to control horses in coronary and lamellar tissues while the skin showed equal frequency of p63-positive cells in control and laminitic horses (Fig 3a). To confirm decreased p63 expression in lamellar tissue of chronic laminitic horses, p63 expression was quantified relative to K14 by quantitative immunoblotting. Relative p63 expression was significantly lower only in distal lamellar tissue of laminitic horses (P = 0.01), with no group differences for skin, coronet, proximal lamellae and middle lamellae (Figs 3b,c).

Figure 3.

Relative quantification of p63 by positive cell number and protein levels in skin, coronet and epidermal lamellae. a) Mean ± s.e.m. percentage of absolute number of p63-positive epidermal cells in skin, coronet and proximal, middle and distal lamellae in control and laminitic horses (n = 5 each). Indirect immunofluorescence images were used to count the number of p63-positive epidermal cells and results are expressed as a percentage of the total epidermal cell count identified by DAPI nuclear staining. Asterisks indicate statistical significance (P<0.05) between control and laminitic horses according to ANOVA with Tukey's HSD. b) p63 protein levels relative to K14 expression. Integrated optical density (IOD) of p63 probed with anti-p63 antibody in immunoblot analysis was normalised to that of K14 and relative optical density within each tissue was plotted. Distal lamellae showed statistical significance between groups (P<0.05) according to independent t test. c) Representative immunoblot analysis used in b) for skin and distal lamellar tissue from a control and laminitic horse. C, control; L, laminitic horse.

Regional differences in p63 expression between control and laminitic horses along the epidermal lamellar axial-abaxial axis were also investigated. As shown in Figure 4, the frequency of p63 positive ECs was decreased specifically within abaxial regions of lamellae in laminitic compared to control horses (Figs 4c,i), whereas p63 expression remained high and more similar to controls in the axial regions (Figs 4a,g). Interestingly, laminitic axial regions frequently contained epidermal islands, not usually observed in control horses (Figs 4d,j; see Discussion). Quantification and statistical analysis revealed that there was a group × section effect (P = 0.028), with the percentage of p63 positive ECs lower in abaxial compared to central and axial sections (P<0.05) in the laminitic group (Fig 4n) while there was no difference between regions for the control group (Fig 4m). The decreased p63 expression in the abaxial region corresponds to the lamellar dysplasia, including epidermal islands, and hyperkeratosis noted in laminitic horses (Table 2) and is consistent with an inverse correlation between p63 expression and EC terminal differentiation.

Figure 4.

Decreased p63 expression along the axial-abaxial primary epidermal lamellae (PEL) axis in laminitic horses. Shown are axial a), d), g), j), central b), e), h), k), and abaxial c), f), i), l) middle lamellar tissues in laminitic horse L5 a–f) and control horse C4 g–l). Panels a–c and g–i show representative images of p63 expression as revealed by Alexa Fluor 594 rabbit anti-mouse IgG secondary antibody. Panels d–f and j–l show haematoxylin and eosin (H&E) staining for corresponding regions. m, n) The frequency of p63-positive epidermal cells in axial, central, and abaxial regions in control horses (m) and laminitic horses (n). In panel N, asterisks indicate differences (P<0.05) between tissue regions according to ANOVA with Tukey's HSD. Note that p63-positive epidermal cells are gradually lost along the PEL axis in laminitic horses, while control horses maintain high levels. p, the keratinised axis of the PEL; h, the hyperplastic, dysplastic, and acanthotic primary epidermal lamellae; s, the secondary epidermal lamellae. Black arrows in panels d and e indicate epidermal islands. Asterisks in panels h and i indicate nonspecific autofluorescence by keratinised lamellae. Scale bars = 100 µm.

Epidermal cell proliferation in relation to p63 expression

Although only a small subset of p63 positive basal ECs was also immunoreactive to the proliferating cell marker Ki-67, all of the cells that were Ki-67 positive displayed concurrent p63 expression (Fig 5). Expression of Ki-67 differed across tissue type (P<0.001), with the percentage of Ki-67 positive ECs greater in coronary tissue than in the skin, middle and distal lamellae (Fig 5c). Overall, there was no statistical difference of Ki-67 positive cell frequency between control and laminitic horses across tissue type (Fig 5c).

Figure 5.

Epidermal cell proliferation in the coronary epithelium of the equine foot. Dual indirect immunofluorescence was performed with antibodies to Ki-67 and p63 and representative images of Ki-67 staining (a) and co-localisation of p63 and Ki-67 (b) are shown for coronary epithelium from control horse C2. A small subset of p63-positive cells in the basal layer was Ki-67 positive as highlighted in insets, where open arrows and solid arrows indicate examples of p63+Ki-67- cells and p63+Ki-67+ cells, respectively. Asterisks in panels a) and b) indicate autofluorescence within the dermis. Scale bar = 100 µm c) Overall proliferation profile of control (n = 5) and laminitic (n = 5) horse skin, coronary epithelium and proximal, middle and distal lamellar tissues. Indirect immunofluorescence images were used to count the number of Ki-67 positive epidermal cells, and results are expressed as mean ± s.e.m. percentage of the total epidermal cell count identified by DAPI nuclear staining. Triplicate images for each tissue within each control horse and laminitis were used for analysis. Asterisk in panel c) indicates that for control horses, the mean for coronet is different (P<0.05) than skin, middle and distal lamellar tissues according to ANOVA with Tukey's HSD.

Along the axial-abaxial differentiation axis within control lamellar tissues, the percentage of Ki-67 positive cells was higher in axial (5.3 ± 1.9%) compared to central (1.3 ± 0.5%) and abaxial (1.1 ± 0.9%) sections (P<0.05). In comparison between control and laminitic horses, the percentage of Ki-67 positive cells was lower (P<0.05) in axial sections of laminitic (2.3 ± 0.8%) compared to control horses, whereas it was statistically similar to control in the central (0.8 ± 0.4%) and abaxial (0.6 ± 0.5%) regions.


The present study characterised putative ESCs within the equine hoof by expression of p63, a transcription factor essential for the maintenance and proliferative potential of ESC populations (Senoo et al. 2007). Results of the present study that support the use of p63 as a valid marker of ESCs in horses include: 1) A pattern of high ΔNp63α expression in a subset of basal ECs of equine skin, coronet and lamellae similar to the pattern observed in other established models, including man, rat and mouse (Yang et al. 1998; Parsa et al. 1999; Senoo et al. 2007) and 2) the greatest frequency of p63 positive ECs was detected in the proximal regions of the hoof where the rate of cell proliferation is highest (Daradka and Pollitt 2004).

While p63 is associated with the potential for proliferation, the protein Ki-67 identifies cells in active phases of the cell cycle and is used as a marker of proliferating cells (Scholzen and Gerdes 2000). In the present study, Ki-67 was colocalised to a subset of p63 positive ECs in equine epidermal tissues, with the greatest frequency in the highly regenerative coronary epithelium, which is consistent with previous studies by BrdU incorporation (Daradka and Pollitt 2004). Ki-67 expression in a subset of p63 positive cells is similar to the expression pattern observed in the skin and stratified epidermal cultures, in which p63 expression is present only within cells that are either proliferating or possess the ability to proliferate (Yang et al. 1998; Senoo et al. 2007). Together, these results suggest that p63 positive cells in the coronet continuously supply epithelium to generate rapidly growing hoof wall, while p63 positive cells in the lamellae function in the maintenance of tissue homeostasis and repair, similar to their function in the skin (Blanpain and Fuchs 2009).

The current model of mature epidermal homeostasis describes 2 discrete populations of progenitor cells required for the maintenance of tissue homeostasis: self-renewing ESCs and their immediate progeny, known as transit amplifying cells, which differentiate after several rounds of cell division (Blanpain and Fuchs 2009). Both types of epidermal progenitor cells of the skin are found in the basal cell layer. Once transit amplifying cells become committed to terminal differentiation, they exit the cell cycle and migrate into the suprabasal cell layers. In agreement with this model, it was found that equine skin and hoof epidermal basal cells vary in p63 expression levels. In several epithelial tissues (normal and pathological) and cell culture systems studied to date, cells with the highest levels of p63 represent the ESC population and p63 expression is inversely correlated with terminal differentiation (Parsa et al. 1999; Kurokawa et al. 2006; Senoo et al. 2007). We propose that the basal ECs of the coronet and lamellae with the highest expression of p63 represent self-renewing ESCs, while the basal and suprabasal ECs expressing lower levels of p63 are transit amplifying cells and p63-negative ECs are nonproliferative, differentiated cells.

The present study utilised 5 laminitic horses at a stage where distal phalanx displacement had already occurred, all with a 2 weeks minimum duration of laminitis prior to euthanasia. Although each case had different apparent contributing aetiological factors and histopathological characteristics of laminitis, our results demonstrate that relative expression of p63 is consistently decreased in lamellae of laminitic compared to control horses, especially in the distal regions. This suggests that laminitic epithelium may have impaired ESC activity and a possible shift from a basal to a more differentiated cell phenotype. We detected no difference in overall percentage of Ki-67 positive cells in laminitic horses compared to control horses, suggesting that aberrant constitutive proliferation may not be present in animals that have developed a gross lamellar wedge. It is possible that some of the increased mass of dysplastic keratinised tissue that constitutes the lamellar wedge of chronic laminitis could result from the aberrant accumulation of terminally differentiated cells. In addition to its role in ESC maintenance, p63 also regulates the expression of genes involved in epithelial adhesion to the extracellular matrix, including basal integrins such as α6β4 (Carroll et al. 2006), desmosomal proteins (Ihrie et al. 2005) and Fras1, an epidermal extracellular matrix protein necessary for basement membrane integrity in mice (Koster et al. 2007). Decreased p63 expression could, therefore, relate to the disruption of basal cell adhesion to the basement membrane, a hallmark of laminitis (Pollitt 1996).

Within lamellar epithelium, there was a more pronounced decrease of p63 positive cells in abaxial lamellar regions of laminitic horses. Localised differences in differentiation along the axial-abaxial axis may also be influenced by biomechanical changes that take place in the laminitic hoof during failure of the suspensory apparatus of the distal phalanx. Increased epidermal proliferation has been described in the rat ear in response to tension, and has been ascribed to mechanotransduction of signals that alter gene expression through deformation of cytoskeletal and integrin proteins (Pietramaggiori et al. 2007). During laminitis, partial dermo-epidermal detachment may create localised increases in tension, stimulating temporal epidermal proliferation and lamellar dysplasia. However, eventual dermal vascular displacement resulting from displacement of the distal phalanx during severe laminitis (Collins et al., 2010) may limit blood supply to epidermal tissues closer to the hoof wall and promote differentiation rather than proliferation. Additionally, local inflammation in response to tissue damage or systemic disease may modulate EC proliferation and differentiation (Jameson and Havran 2007; Lee et al. 2009). Our data suggest that dysplastic differentiation of ECs rather than proliferation contributes to the accumulation of laminitic epithelium, but earlier increases in proliferation could also contribute to the initial accumulation of dysplastic lamellar tissue.

Epidermal islands localised to the axial SEL region were apparent in the chronic laminitic horses used in this study and were also described in horses 7 days after experimental induction of laminitis by oligofructose administration (Van Eps and Pollitt 2009). In the present study, in contrast to the reduced frequency of p63 positive cells in abaxial regions, p63 expression remained high in the axial regions containing epidermal islands in laminitic horses. Epidermal islands expressing high levels of p63 have also been described during human cutaneous wound healing and are believed to be important in re-epithelialisation (Kurokawa et al. 2006). Although it is possible that the epidermal islands observed in laminitic horses represent an attempt to re-epithelialise the defect generated by dermal-epidermal separation, these structures are no longer connected to their respective PEL and have therefore lost the capacity to function as a suspensory apparatus linking the dorsal hoof wall to the distal phalanx (Van Eps and Pollitt 2009). With continued instability, epidermal islands could potentially contribute to pathological lamellar wedge structures that appear to exacerbate the pathologies of chronic laminitis, including osteolytic changes to the adjacent dorsal margin of the distal phalanx (Collins et al. 2010; Engiles 2010).

It was proposed that p63 expression is inversely related to the severity of epidermal dysplasia associated with chronic laminitis and the ability of the horse to regenerate functional epidermal lamellae. However, a larger prospective study quantifying both p63 expression in lamellar biopsy from laminitic horses and the clinical outcomes of these cases would be needed to establish such a model. The accumulation of dysplastic epidermal tissue, derived from the conjunctiva, also contributes to the pathology and decreased function of diseased human and equine corneas (Brooks et al. 2008; Rama et al. 2010). Enormous progress has been made in recent years in the treatment of unilateral human corneal disease and the associated blindness by the autologous transplantation of p63-expressing ESCs derived from the limbus of the unaffected eye (Rama et al. 2010). The demonstration, in this study, that equine lamellar and coronary regions also contain p63-expressing ESCs supports the potential application of autologous ESC-selective culture and regenerative therapy for equine laminitis.

In summary, p63 positive cells are located throughout the coronary and lamellar epithelia, purportedly contributing to normal hoof wall growth and tissue homeostasis through maintenance of a stem cell population. Our finding that expression of p63 in hoof lamellae was significantly lower in association with laminitis suggests reduced activity of ESCs and a shift towards differentiation during chronic laminitis. Although more extensive analysis is needed over a wider spectrum of laminitis aetiologies and stages, our present study clearly suggests that loss of epidermal stem cells is a contributing factor in the pathogenesis of equine laminitis and that autologous transplantation of p63-positive epidermal stem cells from unaffected regions may have regenerative therapeutic potential for laminitic horses.

Conflict of interest

The authors have declared no conflict of interest.

Sources of funding

This study was supported by The Fund for Laminitis Research, University of Pennsylvania School of Veterinary Medicine and the Bernice Barbour Foundation.


We are grateful to Dave Lorom, Ralph Conti, Dave Harris, Nigel Watson, Jill Beech, and Sue Lindborg for contributions to tissue retrieval for this study.

Manufacturers' addresses

1 Invitrogen Molecular Probes, Eugene, Oregon, USA.

2 Sigma-Aldrich, Inc., St. Louis, Missouri, USA.

3 Vector Laboratories, Burlingame, California, USA.

4 Abcam, Cambridge, Massachusetts, USA.

5 Biolegend, San Diego, California, USA.

6 Santa Cruz Biotechnology, Santa Cruz, California, USA.

7 BioSpec Products, Bartlesville, Oklahoma, USA.

8 Roche Diagnostics, Indianapolis, USA.

9 Thermo Scientific, Rockford, Illinois, USA.

10 Calbiochem, Darmstadt, Germany.

11 GE Healthcare, Buckinghamshire, UK.

12 Media Cybernetics: Silver Spring, Maryland, USA.

13 Stata-Corp, College Station, Texas, USA.