Interleukins decrease epithelial protein and gene expression by K14 + CD105+ equine hoof progenitor cells Authors: Qingqiu Yang, Mandi J Lopez Published: 2026 ABSTRACT AbstractBackgroundProgenitor cell epithelial to mesenchymal transition (EMT) could impact healing in the epidermal-dermal junction of the equine hoof stratum internum (SI). The hypothesis of this study was that K14 + CD105+ equine hoof SI progenitor cells assume a mesenchymal phenotype in the presence of inflammatory interleukins in vitro.MethodsK14 + CD105+ progenitor cell percentages isolated from proliferative scarred (fibrous) and healthy SI were quantified. Decellularized matrix was prepared from and K14 + CD105+ cells localized in healthy SI. Gene expression (CD44, CD105, E-cadherin, N-cadherin, β-catenin, K1, K10, p63, TGF-β1, -β2, -β3) and E- and N-cadherin+ cell percentages were determined in continuously cultured cells. Revitalized K14 + CD105+ cells were cultured in stromal medium with or without interleukins 2 and 6 (IL medium) or on decellularized SI matrix in stromal medium. After 21 days, cell gene expression (K1, K10, p63, β-catenin, TGF-β1, -β2, -β3), E-cadherin + and K14 + cell percentages, and medium TGF-β1 levels were measured.ResultsMost cells from healthy and fibrous SI were K14 + CD105+, and K14 and CD105 antigens were present on cells in situ and in vitro. K14 + CD105+ equine hoof SI progenitor cells maintained the immunophenotype over multiple cell passages in vitro. E-cadherin+ cell percentages were higher and K1 expression lower in K14 + CD105+ versus unsorted cells and more K14 + CD105+ cells were E- versus N-cadherin+. K14 + CD105+ cells cultured on matrix had the highest E-cadherin+ cell percentage and p63 and K10 gene expression. Cell TGF-β2 and β-catenin expression was highest with stromal medium, and cell TGFβ-1 and TGF β-3 expression was lowest with IL medium and on matrix, respectively. Cell TGF-β2, TGF-β3 and β-catenin expression was lower on matrix versus in stromal medium. Cells in IL medium had higher TGF-β3 expression than those on matrix. TGF- β1 levels were lower in IL versus stromal medium.ConclusionsThe cell genetic and antigen profiles suggest that inflammatory interleukins drive mesenchymal differentiation while healthy matrix supports epidermal differentiation. K14 + CD105+ progenitor cells from the SI epidermal-dermal niche provide a platform to elucidate progenitor cell EMT capabilities and restore normal tissue healing. These findings have important implications for improving treatment strategies for injuries and disease of the epidermal-dermal junction.Keywords: Epithelial-mesenchymal transition, Ectoderm, Fingernail, Horse, Laminitis, Matrical, Stem cell Background Epidermal scarring resulting from injury or disease can alter both appearance and function. Impaired function can reduce quality of life and contributes to more than $39 billion in annual medical costs the United States [1–6]. The human fingernail and horse hoof are specialized epidermal appendages with unique attachments to the underlying bone. Specifically, the matrical nail mesenchyme, a modified dermis located beneath the nail bed epithelium, is connected to the third phalanx [7–9]. In humans, the epidermis is tightly attached to dermal ridges in a tongue and groove fashion [10]. The horse hoof is a modified epidermis covering the third phalanx. It has an epidermal-dermal interface, the stratum internum (SI), which contains interdigitating epidermal and dermal lamellae [11] that connect the third phalanx to the hard outer hoof capsule [12]. There is significant focus on optimizing and accelerating skin healing with minimal scar formation using various animal models of healthy skin [13–15]. However, there remains a critical need for animal models that replicate abnormal epidermal tissue repair [16]. Abnormal cytokine or hormonal cascades can result in various amounts of detachment between epidermal and dermal lamellae of the equine hoof SI [12, 17–19]. Successful restoration of the healthy tissue architecture is rarely, if ever, achieved when significant detachment occurs, often due to exuberant fibrous tissue formation in the damaged SI [20, 21]. Similarly, nail bed injuries are common, difficult to manage, and may lead to permanent nail disfigurement and compromised function, frequently due to fibrous scar tissue [22–24]. Epidermal-dermal interactions are necessary for hoof and nail production, and unique progenitor cells reside in the epidermal-dermal interface of both human fingernails and equine hooves [9, 25–27]. Progenitor cells in the hoof SI and human fingernail bed appear to have the capacity to rapidly differentiate into distinct lineages, likely to maintain the tissue interface [25, 26, 28, 29]. Understanding the progenitor cells that maintain and repair the unique epidermal-dermal boundary will significantly advance therapeutic strategies to restore tissue damaged by disease or injury. Epithelial to mesenchymal transition (EMT) broadly refers to the process by which epithelial cells lose ectodermal intercellular adhesions, migrate, and support new tissue generation as mesenchymal-like cells [30–33]. It includes induction of transcription factors that alter gene expression to mediate the change from an epithelial to mesenchymal phenotype [34]. Transforming growth factor-β (TGF-β) and Wnt–β-catenin cell signaling pathways contribute to the altered transcription factor expression characteristic of EMT [34, 35]. The form of EMT associated with wound healing and tissue regeneration typically ends when the initiating inflammation resolves [30, 31]. If inflammation persists, it can lead to fibroproliferative wound healing like that in hypertrophic scarring and disorganized fibrous tissue in the hoof SI [36–39]. It is possible that protracted or dysregulated progenitor cell differentiation toward a mesenchymal versus epithelial phenotype that resembles EMT may play a role in abnormal epithelial tissue healing in the presence of inflammatory cytokines [40, 41]. Neotissue generation by endogenous and/or exogenous progenitor cells is facilitated by healthy tissue matrix, and abnormal tissue microarchitecture contributes to hypertrophic scarring [42, 43]. Tissue decellularization makes it possible to maintain microstructural architecture while removing immunogenic cells and infectious agents that can limit the use of auto- and allografts [44, 45]. The potential to generate the skin epidermal-dermal junction in vitro and in vivo with decellularized tissue matrix and cells is established [46, 47]. However, there is comparatively limited study of the unique ectodermal-dermal interface that is vital to keratinized nail structure and function. Study of progenitor cell-tissue matrix constructs under in vitro conditions that replicate natural conditions of inhibited or altered epidermal tissue healing could provide a consistent, repeatable mechanism to target and implement new therapies without inherent variability of animal models. The goal of this study was to establish a novel, in vitro model by which to evaluate EMT characteristics in progenitor cells that reside in a unique epidermal-dermal interface like that between the human fingernail bed and equine distal phalanx. To accomplish this goal, progenitor cells that expressed both epithelial, keratin (K) 14, and mesenchymal, cluster of differentiation (CD) 105, antigens were isolated from healthy and proliferative scarred (fibrous) equine hoof SI tissue. The cells were thoroughly characterized to confirm reproducibility of the isolation and culture conditions. Subsequently, gene and protein expression by cells from healthy tissue were assessed under standard conditions with and without decellularized matrix or inflammatory interleukins. The tested hypothesis was that K14 + CD105+ equine hoof SI progenitor cells assume a mesenchymal phenotype in the presence of inflammatory interleukins in vitro. Overall, this unique and vital in vitro model represents a platform to elucidate progenitor cell EMT capabilities and has important implications for improving treatment strategies for epidermal-dermal junction injuries and disease. Methods Study design The work in this manuscript has been reported in line with the ARRIVE guidelines 2.0. No horses were euthanized for this study. Healthy (H) and fibrous (F) SI tissue was harvested from 22 equine forelimb hooves (n = 11 H, n = 11 F) immediately postmortem after horses, 15 mares and 7 geldings, were humanely euthanized by a licensed veterinarian for health reasons unrelated to this investigation (Fig. 1). Specimens were harvested in the same order that donors met established inclusion criteria. The inclusion criteria for healthy SI donors included: (1) Light breed; (2) 450–550 kg; (3) 5 years of age or older; (4) mare or gelding. An additional criterion of a confirmed diagnosis of Obel grade II or greater laminitis for a minimum of three months was required for fibrous SI donors; the inclusion criteria were otherwise identical to healthy SI donors. Tissue was processed for progenitor cell isolation (n = 11, H; n = 11, F), evaluated with immunohistochemistry (n = 3, H; n = 2, F) for in situ localization of K14 and CD105, or decellularized (n = 5, H). Percentages of K14 + CD105+ cells in passages 0–2 of isolates from both tissue conditions were determined. Cell isolates from healthy tissue were cultured up to P1 when K14 + CD105+ cells were sorted from one half of each heterogeneous cell isolate. One half of each K14 + CD105+ sorted population was cryopreserved. Continuously cultured cell assessments included CD44, CD105, E-cadherin, N-cadherin, β-catenin, K1, K10, p63, and TGF-β1, -2, and − 3 gene expression (n = 5, H) of unsorted and sorted P3 cells, and percentages of E-cadherin + and N-cadherin+ cells within the same cell groups determined with flow cytometry. Cryopreserved K14 + CD105+ cells (n = 5, H) were revitalized and cultured to P2 in stromal medium. The immunophenotype of subcultured cells was confirmed with immunocytochemical double labeling with K14 and CD105 antibodies (P3, n = 2, H). Passage (P) 3 cells were also cultured in stromal medium alone, stromal medium with IL-2 and 6, or on decellularized SI matrix in stromal medium for 21 days. Gene expression, K1, K10, p63, β-catenin, TGF-β1, -2, and − 3, and percentages of E-cadherin + and K14 + cells were determined as indicated above (n = 6, H). An ELISA was used to quantify TGF-β1 protein levels in culture medium (n = 5, H). Study design schematic. The isolation and characterization of equine SI K14 + CD105+ progenitor cells and the evaluation of environmental conditions on cell gene and protein expression are depicted Stratum internum harvest Tissue was collected as previously described [25]. Briefly, the outside surface of each disarticulated hoof was cleansed with 10% sodium hypochlorite and 0.1% chlorhexidine before tissue collection. Stratum internum on the dorsal most aspect of the hoof was exposed by elevating a 5 cm section of hoof wall between two full thickness cuts created with a sterile blade on a reciprocating saw. Pieces of SI (1 cm x 1 cm) from the center of the exposed section were sharply excised. Tissue for progenitor cell isolation was lightly rinsed with 70% ethanol followed by submersion in 0.1% chlorhexidine for 3 min, and then it was thoroughly rinsed with and subsequently maintained in phosphate buffered saline (PBS, Hyclone, Logan, UT) containing 1% antibiotic-antimycotic solution (MP Biomedical, Irvine, CA) for no more than 10 min. Immunohistochemistry Harvested SI tissue was embedded in optimal cutting temperature (OCT, Sakura Finetek Inc., Torrance, CA) medium and solidified at -80o C followed by cryostat (Leica® CM1850, Sarasota, FL) sectioning (5 μm). Sections on glass slides (poly L-lysine coated, Sigma Aldrich, St. Louis, MO) were were permeabilized with 0.5% Triton-x100 for 20 min and then blocked with 10% goat serum (Gibco®, Grand Island, NY) diluted in PBS for 1 h at room temperature. Slides were then incubated with K14 (monoclonal, mouse anti-human, 0.2 mg/ml, # MA5-11599, Fisher Scientific, Rockford, IL) and CD105 (polyclonal, rabbit anti-human, 0.2 mg/ml, #PA5-16895, Invitrogen, Rockford, IL) primary antibodies for 12 h at 4o C and rinsed with tris-buffered saline (TBS) containing 0.025% triton followed by incubation with labeled secondary antibodies (Alexa Fluor™ 488, polyclonal, donkey anti-mouse, 2 mg/ml, #A-21202; Alexa Fluor™ 594, chicken anti-rabbit, 2 mg/ml, #A21442, Thermo Fisher Scientific, Eugene, Oregon) for 2 h at room temperature protected from light. Slides were rinsed again with TBS and then nuclei were labeled with 4, 6-diamidino-2-phenylindole (DAPI, Thermo Fisher Scientific, Waltham, MA) for 10 min at room temperature in darkness. To measure autofluorescence and verify the specificity of the secondary antibody, both unstained sections and those treated only with the secondary antibody were assessed. Images were generated with a fluorescent microscope (DM 4500b, Leica) equipped with a digital camera (DFC 480, Leica). Primary cell isolation and expansion As reported previously [25], a portion of the SI tissue from each hoof was minced into 5 mm x 5 mm pieces on a sterile petri dish (Nunc™ IVF, Thermo Fisher Scientific) and then digested with 0.1% type-1 collagenase (300U/ml, Worthington Biochemical Corporation, Lakewood, NJ) in Dulbecco’s modified Eagle’s medium with Ham’s F12 nutrient mixture (DMEM-Ham’s F12, Hyclone) containing 1% bovine serum albumin (BSA, Sigma Aldrich) at 37o C for 2 h with 3-dimensional agitation (5.5×g, GyroTwister™ GX-1000, Labnet International, Inc., Edison, NJ). The mixture was then sequentially filtered through 100 μm and 70 μm filters (Thermo Fisher Scientific) and subsequently centrifuged for 5 min at 250×g. The resulting stromal vascular pellet was resuspended in stromal medium (1% antibiotic-antimycotic solution (MP Biomedical) and 10% characterized fetal bovine serum (FBS, VWR Life Science, Radnor, PA) in DMEM-Ham’s F12 culture medium). Cells were seeded on 10 cm culture dishes (Thermo Fisher Scientific) at a density of 5 × 103 cells/cm2 after which they were maintained under standard culture conditions (37 °C, 5% CO2). Medium was refreshed after the first 12 h and then every 2 days until 80% cell confluence when cells were passaged following 0.5% trypsin (Hyclone, Logan, UT) detachment. K14+CD105+ cell quantification and sorting Cells were incubated at 4 °C with antibodies (l µl /1 × 106 cells) against CD105-PE (phycoerythrin, mouse, 0.2 ug/ul, #12-1057-42, eBioscience, San Diego, CA) and K14-dylight 633 (mouse, 0.2 mg/ml, #MA5-11599, Thermo Fisher Scientific) for 45 min in darkness. Double positive cells were quantified or isolated with a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA). Sorted cells were cryopreserved or cultured as described below. Unlabeled cells served as a control for autofluorescence. Progenitor cell cryopreservation Passage 1 progenitor cells were suspended in cryopreservation medium (10% dimethyl sulfoxide (DMSO), 10% DMEM-Ham’s F12, 80% FBS) in a microvial (Thermo Fisher Scientific), and slowly cooled (CoolCell™ LX, BioCision, Tewksbury, MA) to -80 °C at which they were maintained in an ultralow freezer for 7 days [28]. Microvials were thawed for 5 to 10 min at room temperature and then centrifuged at 250×g. The resulting cell pellet was washed with PBS, centrifuged again, and resuspended in stromal medium. Cells were seeded in 10 cm culture dishes and cultured as described above. Immunocytochemistry Revitalized K14 + CD105+ progenitor cells from above were seeded (5 × 103 cells/cm2) in six-well culture plates (Fisher Scientific) and cultured until 80% confluent. Cells were then washed 3 times with PBS and fixed with 4% paraformaldehyde overnight at 4 °C. Following a PBS rinse, cells were permeabilized with 0.5% Triton-x100 for 20 min and then blocked with 10% goat serum (Gibco®) for 1 h at room temperature. Cells were subsequently incubated with antibodies against CD105 (rabbit, 0.2 mg/ml, #PA5-16895, Invitrogen) and K14 (monoclonal, mouse anti-human, 0.2 mg/ml, #MA5-11599, Fisher Scientific) for 2 h at 37o C followed by Alexa Fluor® 594 chicken anti-rabbit IgG (2 mg/ml, #1917947, Thermo Fisher Scientific) and Alexa Fluor® 488 donkey anti-mouse IgG (2 mg/ml, #1975519, Thermo Fisher Scientific) secondary antibodies for 1 h at 37 °C in darkness. Nuclei were stained with DAPI for 10 min in darkness. Images were captured as described above. K14 + CD105+ cell culture in interleukin medium Revitalized K14 + CD105+ progenitor cells were subcultured to P2 in stromal medium. Passage 3 cells were added to T75 culture flasks at a cell density of 5 × 103 cells/cm2 and cultured in stromal medium alone or stromal medium containing 0.125 ug/ml each of IL-2 and IL-6 (PeproTech, Inc., Rocky Hill, NJ) under standard culture conditions (5% CO2, 37 °C) for 21 days. Medium was refreshed every 2–3 days. Stratum internum scaffold - K14 + CD105+ construct preparation and culture Stratum internum tissue (5 × 5 × 5 mm) was decellularized with 0.1% EDTA for 1 h and then with 0.5% SDS as previously described [28]. After decellularization, samples were soaked in PBS for 1 h and then lyophilized in a freeze dryer (-55o C, VirTis® BenchTop™, Stanford, TX) for 12 h. Lyophilized tissues were sterilized with ethylene oxide. A total of 3 × 106 revitalized P3 K14 + CD105+ cells, subcultured as described above, were suspended in 100 ul of stromal medium. They were added to SI matrix by pipetting to give a cell density of 2.4 × 104/mm3. Constructs were immersed in stromal medium in 12 well plates (Thermo Fisher Scientific) and cultured for a total of 21 days under standard culture conditions (5% CO2, 37 °C). Medium was changed every 2–3 days. Culture medium TGF-β1 quantification A commercially available equine TGF-β1 kit (Genorise Scientific, Inc., Glen Mills, PA) was used to quantify the protein in cell culture medium after a total of 21 days of culture and 3 days after the last medium change according to the manufacturer’s instructions. Protein was quantified based on a standard curve (0–1 × 103 pg/ml) of TGF-β1 absorption at 450 nm and 540 nm as determined with a plate reader (BioTek®, Winooski, VT). Target gene expression Equine specific primers were designed for CD44, CD105, K1, K10, p63, E-cadherin, N-cadherin, TGF-β1, TGF-β2, TGF-β3, and β-catenin (Table 1). Total cell mRNA was extracted (RNAqueous Micro Kit, Ambion, Inc., Austin, TX) and quantified (NanoDrop, Montchanin, DE). Target gene mRNA levels were determined using cDNA (QuantiTect reverse transcription kit, Qiagen Sciences, German Town, MD) and SYBR green (Qiagen) RT-PCR (QuantStudio 5™, Thermo Fisher Scientific). The ΔCT (threshold cycle) values were determined relative to the reference gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as CTtarget-CTreference. Table 1Primer sequencesGeneForwardReverseAccession No./Product Length (bp)GAPDHCAATGACCCCTTCATTGACCGAAGATGGTGATGGCCTTTCNM_001163856/133K1CAACAAGCGGACAAATGCCGGACCTCAGCGATAATGNM_001346199/243K10GCTCGCCTACCTGAAGAAGCTCCAGACCTTGAACAGTGNM_001346076/292P63ATCAACCCTCAGCAGCGATGAACAGCCCAACCTCGXM_023623496/235β cateninCAGGTGGTGCTGAATAAGGGTTGTGGAGAGTTGTAATGGXM_023619816/291E cadherinCATTGTGTACTCCATCCTCAGCTGTACTTAAGCCCTCACCTTGAXM_023637192/175N cadherinTGCCGGCACCATGTTAACAAGCCAATTGGCAGGATCAGATXM_023647607/98TGF β1AGCCTGAGGCGGACTACTGCCGGTCTCCACAGTCTTATXM_005596086/89TGF β2ATGCCAACTTCTGTGCTGGAACACACAGCAAGGAGAAGCAXM_005608026/120TGF β3TGGAACCAACCTGTTCCGAGGCGTTGTTTGGCGATATGCTXM_001492687/127CD44GTGTGGACAGAAGAAGAAGCTCAATCTTCATGTCCACATTCTGCXM_023653788/197CD105GACGCCAATCACAACATACATCCACATAGGACGCTACGACXM_003364144/158 Primer sequences Flow cytometry – E-cadherin, N-cadherin, K14 Cell aliquots (106 cells) were suspended in 800 ul of 1% BSA containing 0.2% of sodium azide. Labeled, monoclonal antibodies, E-cadherin-488 (rat, 0.5 mg/ml, #53-3249-82, Thermo Fisher Scientific), N-cadherin-PE (mouse, 50 mg/ml, #12-3259-42, Thermo Fisher Scientific), or K14-dylight 633 (mouse, 0.2 mg/ml, #MA5-11599, Thermo Fisher Scientific) were added to cell suspensions and incubated for 45 min at 4o C in darkness. Following a PBS wash, and resuspension in the same solution, percentages of cells expressing each protein were quantified with a FACSCalibur flow cytometer and Cell Quest Pro software as described above. Statistical analysis All results are expressed as mean ± standard error of the mean (SEM). Statistical analyses were performed using commercially available software (Prism, v 10.5.0, GraphPad Software, San Diego, CA, USA). After confirming data normality with the Shapiro-Wilk test, a one-way or two-way ANOVA followed by Tukey’s post-hoc tests were applied for multiple-group comparisons (p<0.05). Specific comparisons included K14 + CD105+ cell percentages among tissue sources within passages, antigen expression and gene expression levels between continuously cultured K14 + CD105+ and heterogeneous progenitor cells, and antigen, protein and gene expression by revitalized K14 + CD105+ cells among culture conditions. Results Progenitor cells isolated from the epidermal-dermal interface of the equine hoof SI have epithelial and mesenchymal antigens and express mesenchymal, EMT, and epithelial genes (Figs. 2 and 3). Double positive cell percentages at multiple passages and localization of K14 and CD105 antigens in equine hoof stratum internum tissue and progenitor cell isolates. The majority of cells in primary cell isolates from healthy and fibrous stratum internum are K14 + CD105+ (A), and basal cells of the secondary epidermal lamella express both antigens (B-D). The double positive phenotype is maintained after cryopreservation, revitalization and culture (E-H). Percentages (mean ± SEM) of K14 + CD105+ cells from healthy and fibrous stratum internum tissue at cell passages (P) 0–2 (A). Columns with different superscripts within passages are significantly different (p < 0.05). Fluorescent photomicrographs of healthy equine stratum internum sections labeled with antibodies against K14 (arrow, green, B) or CD105 (arrow, red, C) and merged images (arrow, yellow, D). Scale bar, 27 μm. PEL = primary epidermal laminae; SEL = secondary epidermal laminae. Fluorescent photomicrographs of revitalized P3 K14 + CD105+ progenitor cells from healthy stratum internum labeled with antibodies against K14 (green, E) or CD105 (red, F) antigens separately and in a merged image (yellow, G). Cells with nuclear stain only (H, DAPI, blue). Scale bar, 100 μm Evaluation of heterogenous and K14 + CD105+ cell isolate E- and N- cadherin antigen and mesenchymal, EMT, and epidermal gene expression. Percentages of cells expressing E-cadherin and N-cadherin antigens vary among continuously cultured P3 heterogeneous and K14 + CD105+ isolates and within the K14 + CD105+ immunophenotype, respectively. Gene expression is similar between isolates except for K1 which is significantly higher in heterogeneous cells. Percentage (mean ± SEM) of cells expressing E- or N-cadherin in continuously cultured P3 heterogeneous and K14 + CD105+ equine hoof stratum internum progenitor cell isolates (A) and levels of mesenchymal, EMT, and epithelial gene expression (ΔCT) in the same cell isolates. Columns with different symbols are significantly different between cell isolates for a given cadherin, and columns with different cell superscripts are significantly different between cadherins within a cell isolate (A) or between cell isolates for a given gene (B, p < 0.05). Reference gene: GAPDH. CD = cluster of differentiation; EMT = epithelial to mesenchymal transition; GAPDH = glyceraldehyde-3-phosphate dehydrogenase; K = keratin; TGF = Transforming growth factor The majority of progenitor cell isolates from healthy and fibrous equine SI tissue had epithelial and mesenchymal antigens. The antigens were present in cells along the basement membrane between epidermal and dermal tissue and on cultured cells. Heterogeneous and K14 + CD105+ cells were positive for both epidermal and mesenchymal cadherins and cultured cells expressed mesenchymal, EMT, and epithelial genes. Specifically, approximately 80–90% of cells from healthy SI tissue and about 80% from fibrous SI tissue were K14 + CD105+ in passages 0–2, and there was a significantly higher percentage of K14 + CD105+cells from healthy SI tissue at P0 (Fig. 2A). Immunohistochemical labeling confirmed localization of K14 and CD105 antigens to basal cells of the secondary epidermal lamella (Fig. 2B-D). Immunocytochemistry confirmed localization of both antigens to previously cryopreserved (revitalized) P3 K14 + CD105+ cells from healthy SI (Fig. 2E-H). Morphology of the cultured cells varied between spindle-shaped, polygonal, and stellate. Continuously cultured heterogeneous and K14 + CD105+ P3 cells from N tissue expressed E-cadherin and N-cadherin proteins (Fig. 3A). The percentage of E-cadherin expressing cells was higher in K14 + CD105+, 63.7 ± 5.7%, versus heterogeneous, 38.0 ± 6.3%, cell isolates, and higher than those expressing N-cadherin, 41.3 ± 7.8%, within K14 + CD105+ cells. The K1 expression was significantly higher in continuously cultured heterogeneous versus K14 + CD105+, progenitor cells from healthy tissue, and the other epithelial genes, E-cadherin, P63 and K10, had similar expression levels. Expression levels of K10 were the highest among epithelial genes in both cell isolates (Fig. 3B). The mesenchymal gene expression, CD105 and CD44, were similar between isolates as were EMT genes, N-cadherin, E-cadherin, β-catenin, and TGFβ-1, -2, and − 3. Expression levels of N-cadherin and TGFβ-1 in both cell isolates were negligible as were K1 expression levels in heterogeneous cells. Interleukins decrease epithelial protein, antigen, and gene expression by K14 + CD105+ progenitor cells from the epidermal-dermal interface of the equine hoof. Levels of culture medium TGF- β1 protein were highest in stromal medium, 403.1 ± 45.0 pg/ml, followed by culture with matrix, 249.6 ± 42.4 pg/ml, and then IL medium culture, 214.6 ± 30.4, pg/ml, and the levels in stromal medium were significantly higher than those in IL medium (Fig. 4A). The percentage of E-cadherin+ cells was lowest in revitalized K14 + CD105+ cells cultured with interleukins, 20.9 ± 6.8%, higher in those cultured in stromal medium, 50.1 ± 4.9%, and highest in cells cultured on matrix, 72.1 ± 1.7% (Fig. 4B). There was no difference in K14+ cell percentages among culture conditions, all of which were above 95%. The EMT gene expression was higher than epithelial gene expression in the revitalized cells under all culture conditions (Fig. 4C). Epithelial gene, K1, p63, and K10, expression levels were higher in revitalized K14 + CD105+ cells cultured on healthy matrix than cells cultured in IL medium. The K1 expression levels were also lower in cells cultured in IL medium than stromal medium. In general, K10 expression levels were higher than the other epithelial genes; they were higher with healthy matrix compared to in stromal or IL medium. The P63 expression levels were lower in IL versus stromal medium. Among EMT genes, TGF-β2 and β-catenin, had higher expression levels in stromal medium versus in IL medium or with matrix. The TGFβ-1 expression levels were lower in IL medium compared to the other culture conditions while the TGF β-3 levels were significantly lower with matrix versus stromal or IL medium. Assessment of the effects of interleukins and healthy matrix on TGF-β1 medium levels, cell E-cadherin and K14 antigen expression and epithelial and EMT gene expression. TGF-β1 medium levels are highest in cultures of K14 + CD105+ cells in stromal medium. Culture in stromal medium containing interleukins decreases the percentage of P3 K14 + CD105+ E-cadherin expressing cells while culture on healthy stratum internum matrix increases the percentage. Revitalized P3 K14 + CD105+ cell culture in stromal medium containing interleukins 2 and 6 decreases epithelial gene expression, and culture on healthy matrix in stromal medium increases epithelial gene expression and decreases EMT gene expression. Culture medium TGF- β1 levels (A, mean ± SEM) in the medium, percentages of E-cadherin + and K14 + cells (B, mean ± SEM) and expression levels of TGF- β1, -2, -3, β-catenin, K1, K10, and p63 (C, ΔCT, mean ± SEM) of revitalized P3 K14 + CD105+ cells from healthy stratum internum tissue after 21 days of culture in stromal medium alone (stromal) or with interleukins 2 and 6 (IL medium) or on decellularized SI matrix (healthy matrix). Reference gene: GAPDH. Significant differences among culture conditions are indicated by different superscripts above columns (A) and for each antigen (B) and gene (C, p < 0.05). CD = cluster of differentiation; EMT = epithelial to mesenchymal transition; GAPDH = glyceraldehyde-3-phosphate dehydrogenase; K = keratin; TGF = Transforming growth factor Discussion The results of this study confirm that K14 + CD105+ progenitor cells are a prominent immunophenotype within the equine hoof epidermal-dermal interface, and that the cells maintain a dual epithelial-mesenchymal phenotype in vitro over several cell passages and after cryopreservation. The presence of the cells in healthy hoof SI tissue as well as in proliferative, fibrous SI tissue that often replaces healthy tissue with laminitis confirms the persistence of progenitor cells in damaged tissue. This suggests that tissue restoration is possible, but that local conditions may alter progenitor cell behavior. Based on E-cadherin positive cells, gene expression, and protein expression, the inflammatory cytokines, IL-2 and 6, appear to support K14 + CD105+ progenitor cell mesenchymal characteristics in contrast to healthy matrix that seems to support an epithelial cell phenotype. Based on these findings, the hypothesis is accepted. The outcomes of this study may provide a unique and reproducible model for study of progenitor cell EMT to develop targeted therapies and prevent the consequences of abnormal tissue healing in the unique epidermal-dermal interface. Progenitor cells in the epidermal-dermal interfaces of the fingernail bed and equine hoof contribute to homeostasis and repair of the tissues [10, 48]. Investigation of progenitor cells with epithelial and mesenchymal characteristics provides guidance for complex tissue engineering [49]. Double positive cells may represent an intermediate or transitional state from which the cells can rapidly mature into multiple lineages [50, 51]. However, there is relatively little information about double positive cells, their potential for EMT, or the impact of environmental conditions on the transition process. Cell EMT is a component of healthy and pathologic tissue growth and repair, and it may represent a form of transdifferentiation, when a cell at an advanced stage of maturity assumes a different tissue phenotype [31, 52, 53]. The potential for progenitor cells to utilize various cell signaling pathways to alter phenotype and function is a natural extension of documented progenitor cell behavior [54, 55]. It is reasonable to assume that the numerous, parallel, and overlapping cell EMT signaling pathways are unique among progenitor cells and parent tissues. A major role of the TGF-β signaling pathway in EMT of many cell types is well established, and the pathway can occur downstream or synergistically with other cell signaling pathways [56]. The precise role of the pathway on the effects of IL-2 and 6 on the observed behaviors of the K14 + CD105+ cells was not examined in this study. Nonetheless, the results establish EMT potential in the progenitor cells, and the cells and cell-matrix constructs could be a platform for elucidating normal and abnormal cell signaling for strategic therapeutic intervention. The finding that cytokines in culture medium appears to increase the mesenchymal characteristics of double positive progenitor cells is consistent with current knowledge that inflammatory proteins alter progenitor cell behavior, induce EMT during skin fibrogenesis and carcinogenesis, and accentuate EMT of dysregulated airway epithelial repair [40, 57, 58]. The information in this study additionally highlights that two interleukins known to contribute to tissue damage in equine hooves and other tissues appear to drive progenitor cells away from an epithelial and toward a mesenchymal phenotype at the genetic and protein levels. This preliminary information may explain the proliferative fibrous tissue that often follows and contributes to tissue damage in epithelial-mesenchymal interfaces within the equine hoof and fingernail [22, 59]. It is possible that native progenitor cells and damaged tissue matrix contribute to a stage of incomplete EMT confirmed in human keloid tissue that results in a predominance of fibrous proliferation [4, 6, 37]. Further studies to determine variable influences of different cytokine levels and combinations will be instrumental to understanding their influence. It is feasible that high dose, local or regional anti-inflammatories that target specific inflammatory mediators of abnormal EMT may be more effective that systemic administration, especially when circulation is compromised by trauma or disease [60–62]. Progenitor cell integration into synthetic or organic matrix is a highly studied topic with significant focus on customized scaffolds that optimize intended cell behavior [63, 64]. Decellularized matrices like those from gastrointestinal, cardiac, and nerve tissues provide templates for de novo tissue generation [65–67]. To date, there has been little study of the interface tissue matrix evaluated in this study. Matrix with exogenous progenitor cells often improves cell recruitment and neotissue production [68]. Drug and cell delivery can be a challenge in interface tissues, especially those damaged by trauma, chronic inflammation, and fibrous scar tissue, all of which compromise vasculature. Based on the results of this study, it is plausible that addition of decellularized matrix with or without cells near the germinal tissue center in the nail bed or hoof could help guide tissue healing toward an epithelial versus mesenchymal phenotype. Previous studies confirmed the potential to access such areas via percutaneous routes in numerous tissues, including the hoof [26, 62, 69, 70]. Work is necessary to confirm the potential therapeutic value of exogenous matrix with and without cells in healthy and abnormal interface tissue. However, additional testing, including evaluation of K14 + CD105+ cell behavior on 3D matrix in the presence of ILs, is necessary prior to initiating preclinical or clinical trials. The results of this in vitro study are specific to the conditions under which they were achieved with acknowledged limitations. The culture conditions evaluations were performed with revitalized cells. As such, results may differ in continuously cultured cells [28]. The cytokine media did not replicate complex and dynamic inflammatory conditions and should not be equated to inflamed tissue. The stromal medium used in this study that is specifically designed for mesenchymal progenitor cells and the long duration of the study, 21 days, might have reduced detection of transient EMT gene and protein expression. Effects of IL medium and matrix were considered separately to quantify the independent effects of each on cell behavior. The medium was selected intentionally for comparative purposes, and the effects of interleukins and matrix compared to stromal medium alone were apparent. However, the lack of an evaluation of matrix in IL medium is a limitation of this study. Results will likely vary with distinct culture conditions, especially with medium designed to maintain epithelial versus mesenchymal progenitor cells and with IL medium and matrix combined. Further delineation and refinement of the matrix components that attract and support cell recruitment and integration will significantly advance knowledge about the process in SI and other interface tissues. Despite these acknowledged limitations, the results of this study provide new and important insight about progenitor cell biology in a unique tissue junction. Conclusions The majority of progenitor cells harvested from the equine hoof SI, an epidermal-dermal interface similar to that of the fingernail, express dual tissue lineage antigens, K14 + and CD105+. The cells can be isolated from healthy and abnormal tissue and maintain the immunophenotype over multiple passages in vitro and after cryopreservation. The genetic and protein profile of the K14 + CD105+ cell population tends toward a mesenchymal phenotype in culture medium with cytokines and toward an epithelial phenotype on decellularized tissue matrix. The work is among the first to identify, isolate and test progenitor cells from a unique epidermal-dermal niche and establish a platform for future studies to elucidate and capitalize upon progenitor cell EMT capabilities to obviate consequences of abnormal tissue healing. Acknowledgements Special thanks to Marilyn Dietrich, Michael Keowen, Catherine Takawira, Frank M. Andrews, and Xiao Niu for technical assistance with the work in this manuscript. Abbreviations Bovine serum albumin Cluster of differentiation Threshold cycle 4, 6-diamidino-2-phenylindole Epithelial to mesenchymal transition Fibrous Glyceraldehyde-3-phosphate dehydrogenase Healthy Interleukin Keratin Passage Phosphate buffered saline Phycoerythrin Standard error of the mean Stratum internum Tris-buffered saline Transforming growth factor Wingless/Int Author contributions QY contributed to study design, performed cell culture and outcome assessments, performed initial data reduction, analysis, and interpretation, and was a major contributor to writing the manuscript. MJL contributed to study design, supervised cell culture and outcome assessments, assisted with data reduction, analysis and interpretation, and was a major contributor to writing the manuscript. Funding This study was supported in part by the Tynewald Foundation, Louisiana State University Equine Health & Sports Performance Program and US Department of Agriculture 1433 Animal Health and Disease Research Program. Availability of data and materials. The raw data for this study are available from the corresponding author upon reasonable request. Data availability The raw data for this study are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The raw data for this study are available from the corresponding author upon reasonable request.