Pharmacological YAP activation promotes regenerative repair of cutaneous wounds

Significance

Millions of Americans suffer from chronic wounds, including diabetic foot ulcers, venous leg ulcers, and pressure ulcers. Despite this persistent and growing medical need, therapeutics that actively accelerate wound closure are limited. Here, we demonstrate that pharmacological activation of the transcriptional coactivator YAP, a key regulator of organ size and cellular proliferation in animals, accelerates wound closure in human and pig models of wound healing. This work may serve as the basis for a novel therapeutic paradigm in treating chronic wounds.

Abstract

Chronic cutaneous wounds remain a persistent unmet medical need that decreases life expectancy and quality of life. Here, we report that topical application of PY-60, a small-molecule activator of the transcriptional coactivator Yes-associated protein (YAP), promotes regenerative repair of cutaneous wounds in pig and human models. Pharmacological YAP activation enacts a reversible pro-proliferative transcriptional program in keratinocytes and dermal cells that results in accelerated re-epithelization and regranulation of the wound bed. These results demonstrate that transient topical administration of a YAP activating agent may represent a generalizable therapeutic approach to treating cutaneous wounds.

Chronic wounds remain a significant public health issue, as an estimated 6.5 million individuals in the USA suffer from a chronic wound (1). Among them, diabetic foot ulcers (DFUs) are the most common, affecting 15 to 25% of individuals with diabetes during their lifetime (2). Venous leg ulcers and pressure ulcers are also common, as an estimated 1 to 2% of the general population collectively suffer from these conditions (3). The 5-y prognosis for DFUs is worse than many cancers, as they can have a significant impact on quality of life, lead to amputation, and are associated with high healthcare costs (3). Current standards of care for treating chronic wounds like DFUs remain debridement of the wound and offloading (4). Also common in treatment regimens is the use of amniotic membranes, which have been shown to aid in repair but do not actively promote wound closure (5).

Mammalian wounds heal in a characteristic sequence, progressing through stages of hemostasis, inflammation, proliferation, and remodeling, which are essential for effective resolution of injury (6). In the context of a chronic wound like a DFU; however, persistent inflammation from infection often prevents successful execution of the proliferative phase (6). The proliferative phase involves both regranulation, the deposition of extracellular matrix from proliferative dermal fibroblasts, as well as re-epithelialization, the process of active wound closure carried out by rapidly cycling epidermal keratinocytes (6). Because proliferative activity is decreased in the context of a chronic wound, early therapeutic efforts focused on developing recombinant growth factors to encourage proliferation of these cellular populations (7). However, chronic wounds also express high levels of degradative proteases, which has hampered the use of topical biologics (8). Despite this challenge, topical recombinant human platelet-derived growth factor (PDGF, becalpermin, called Regranex) was approved more than 20 y ago to help promote regranulation by stimulating dermal fibroblast proliferation (9). Unfortunately, Regranex only displays modest clinical efficacy and is not widely used by physicians (8).

A means to overcome limitations with use of topical growth factors might involve transient activation of a cell intrinsic pro-proliferative transcriptional program to accelerate the proliferative phase of wound healing. Yes-associated protein 1 (YAP1, called YAP throughout), originally identified from genetics screens of organ overgrowth in Drosophila, is a transcriptional coactivator that enacts a pro-proliferative transcriptional response in stem, progenitor, and parenchymal cells (10, 11). The interplay of YAP and the repressive Hippo pathway ultimately controls organ size and ectopic growth in several tissues (12, 13). Recent work has demonstrated that YAP activation is essential for wound healing, as its loss dramatically decreases proliferative wound repair in mice (14). Likewise, overexpression of YAP has been shown to promote a marked expansion of the epidermal layer in transgenic mice, although an effect on accelerating wound closure was not evaluated in this context (15). Recently, we reported the identification of PY-60 (SI Appendix, Fig. S1A), a small-molecule activator of YAP-driven transcriptional activity (EC₅₀ = 1.6 µM) that acts by inhibiting a central scaffolding protein in the Hippo pathway, Annexin A2 (16). PY-60 inhibits the ability of ANXA2 to shepherd YAP to the plasma membrane where it receives inhibitory phosphorylation, resulting in the titratable ability of PY-60 to expand cells ex vivo from 500 nM to 10 µM in a YAP-dependent manner (16). Additionally, topical application of PY-60 to the skin of naïve mice promotes the YAP-dependent expansion of epidermal keratinocytes, enabling this cellular population to overcome endogenous organ size restrictions in the adult animal (16).

Here, we hypothesized that the capacity of PY-60 to promote ectopic growth of keratinocytes might be exploited to accelerate cutaneous wound repair. Herein, we demonstrate that gel formulated PY-60 promotes a robust expansion of both epidermal and dermal cellular populations, accelerating the rate and quality of wound closure in pig and human models. Together, this works puts forward a promising initial preclinical data package suggestive that transient activation of YAP might be an effective means to heal chronic wounds via a regenerative mechanism.

Results

Pharmacological YAP Activation Promotes Proliferation and Stemness in Keratinocytes.

Human wound healing is predominated by a pro-proliferative response in epidermal keratinocytes, which is ultimately responsible for closure of wounds (17). As such, we first sought to determine whether PY-60 treatment might activate a YAP-driven transcriptional program in primary adult human epidermal keratinocytes (HEKa, ATCC). As anticipated, early passage HEKa cells displayed strong sensitivity to PY-60, as 24-h treatment (10 µM) resulted in the ~100-fold induction of the YAP responsive transcript ANKRD1 (Ankyrin repeat domain-containing protein 1, SI Appendix, Fig. S1B). RNA-sequencing experiments with PY-60 (10 µM, 24 h) additionally indicated that compound treatment induced a robust and selective YAP-driven transcriptional program in HEKa cells, as gene set enrichment analysis (GSEA) of core YAP target genes and YAP occupied loci indicated a robust enrichment in PY-60-treated samples (Fig. 1 A and B and SI Appendix, Fig. S1 C and D). Interestingly, a gene set previously associated with proliferation of keratinocytes was markedly upregulated in PY-60 samples (SI Appendix, Fig. S1E). In contrast, dimethyl sulfoxide (DMSO)-treated samples were more enriched for gene sets previously associated with the differentiation and apoptosis of keratinocytes (SI Appendix, Fig. S1 F and G), suggesting PY-60 likely promotes proliferation of this cellular population while decreasing their propensity for terminal differentiation and apoptosis.

Fig. 1.

Pharmacological YAP activator PY-60 promotes expansion, stemness, and the capacity for wound closure in primary human keratinocytes. (A) Gene set enrichment analysis plots depicting core YAP-dependent gene sets (MSigDB ID: M2845 and M2871) upregulated in response to 24-h treatment with PY-60 (10 µM) of primary human keratinocytes. (B) Heatmap depicting leading edge genes upregulated in the Cordenonsi et al gene set in A. (C) Quantification of nuclei of human keratinocytes treated with the indicated concentrations of PY-60 for 3 d in low serum medium. (D) Number of migrated keratinocyte nuclei per well after 24 h of treatment with PY-60 (1 µM). (E) Relative closure and representative images of wound length of primary human keratinocytes treated for 24 h with the indicated concentrations of PY-60 in Ibidi wounding chambers deprived of serum. (F) Relative number (Left) and diameter right of keratinocyte spheroids formed in matrigel after 72-h treatment with 1 µM PY-60 or PPARd agonist GW0742 (n = 3, mean and SEM; t test; *P < 0.05, **P < 0.005, ***P < 0.0005; Scale bar, 100 µm).

We next evaluated whether pharmacological YAP activation might increase the proliferation and wound healing capacity of HEKa cells. Indeed, PY-60 treatment was found to dose dependently increase the proliferative capacity of keratinocytes, increasing the number of HEKa cells that grew in monolayer conditions by approximately 1.7-fold over a 3-d period (Fig. 1C). Consistent with an increased migratory capacity, PY-60 treatment (1 µM) also enabled HEKa cells to transverse a 5-µm transwell membrane (Boyden Chamber Assay, Millipore), increasing the number of migrated cells by ~sevenfold (Fig. 1D). We next assessed the in vitro wound healing capacity of HEKa cells treated with PY-60 using silicon wounding chambers (Ibidi), an assay which mimics an in vitro scratch assay but provides a more precise “wound” upon removal of the silicon insert. Benchmarking to a small molecule agonist (GW0742) of peroxisome proliferator-activated receptor delta (PPARd), a mechanism which was been previously shown to promote keratinocyte hyperplasia and wound healing capacity in rodents (18, 19), we found that PY-60 treatment (1 µM) induced a similar if not greater response at promoting wound closure to that of an optimal concentration of GW0742 (1 µM), with higher concentrations of PY-60 (5 µM) enabling full wound closure over this time period (Fig. 1E). Lastly, we found that PY-60 treatment (1 µM) enabled the culture of HEKa cells as Matrigel (Corning) embedded spheroids, increasing the number and size of spheroids that could be grown over a 72-h period (Fig. 1F). In contrast, GW0742 treatment (1 µM) was unable to promote spheroid growth in these conditions. Together, these results suggested that pharmacological YAP activation by PY-60 promotes a pro-proliferative transcriptional program in keratinocytes, which may derive from increasing the stemness of this population, as has been shown with activating YAP in other differentiated cell types (20, 21).

PY-60 Promotes Proliferative Wound Repair in Pigs.

We recently reported that topical administration of PY-60 increases epidermal thickness in naïve adult mice, a result derived from augmented YAP-dependent keratinocyte proliferation in situ (16). In these studies, PY-60 was applied as a 10 mg/mL solution in acetone, a vehicle which likely did not provide maximal delivery of compound. From experiments involving increasing durations of PY-60 exposure to HEK293A cells (10 µM, 2 to 24 h), we found that at least 16 h of continuous exposure of compound was required for maximal induction of a YAP transcriptional response, as read out by measuring the levels of transcripts ANKRD1 and CYR61 by RT-qPCR at 24 h (SI Appendix, Fig. S2). As such, we hypothesized that a vehicle displaying persistent residence at the site of administration was likely essential to induce an optimal pharmacodynamic response in animal studies.

Screening several combinations of excipients ultimately allowed for the identification of a stable, gel-based formulation of PY-60 (propylene glycol: 26%, Labrasol ALF: 17.5%, PEG 400: 46.5%, Povidone K-90: 8.5%, PY-60: 0.15 or 1.5%) that is compatible with topical dosing. From experiments in which PY-60 was delivered to adult naïve mice as a gel (10 mg/mL or 1.5 wt%) or as a solution of either acetone or DMSO (10 mg/mL each) once daily for 10 d, gel-formulated material (10 mg/mL) was found to induce roughly a fourfold increase of epidermal thickness in contrast to either of the solvent-based formulations, which resulted in an approximate twofold thickening at this time point (SI Appendix, Fig. S3A). Notably, gel formulated PY-60 was found to double epidermal thickness at a 5-d timepoint (SI Appendix, Fig. S3A), suggesting that gel-based delivery of the compound can likely accelerate the rate and magnitude of epidermal thickening. Importantly, uniform histological staining for immunopositivity of the epidermal layer for keratinocyte marker Cytokeratin 14 confirmed the expansion of the epidermis corresponded to an increase in keratinocyte numbers (SI Appendix, Fig. S3 B and C).

Despite some structural similarities, the skin of mice differs considerably from that of humans, as cutaneous wounds in the mouse heal rapidly by a physiological process dominated by wound contraction rather than by proliferative repair (22). We therefore elected to evaluate the response to PY-60 in the skin of pigs, as this species more closely mimics human skin, both in that it is much thicker like human skin and that wound healing occurs principally by proliferative expansion of epidermal and dermal layers (23). Using the Yucatan mini pig as a model species, we evaluated the histological effects of topical application of gel formulated PY-60 (0.15% or 1.5%) once daily for 10 d (Fig. 2A). As anticipated, PY-60 administration significantly increased the thickness of the epidermal layer by ~50% (~40 µm for vehicle vs. ~60 µm for PY-60-treated animals; Fig. 2B), a result derived from increasing the number of keratinocytes (Fig. 2 C and D). We confirmed that increased expansion of keratinocytes was derived from increased proliferation, as PY-60 treatment was found to increase the percentage and number of KI-67-positive keratinocytes in the epidermal layer (Fig. 2 E and F and SI Appendix, Fig. S4B). In addition to the epidermis, we observed a marked increase in the thickness and the number of cells in the dermal layer (Fig. 2 G and H and SI Appendix, Fig. S4C), indicating that PY-60 can also penetrate and enact pro-proliferative transcriptional program in this compartment. This result is consistent with qPCR analysis of full thickness skin punches at study end, which indicated that PY-60 enacted a lasting transcriptional response, as determined by increased levels of transcripts Ccn1 and Ccn2, pig equivalents of CYR61 and CTGF respectively (SI Appendix, Fig. S4A).

Fig. 2.

PY-60 promotes proliferative wound healing in pigs. (A) Schematic depicting a pharmacodynamic study in which Yucatan mini-pigs are administered gel formulated PY-60 topically once daily (0.15 wt% gel). Quantification of epidermal thickness (B) and keratinocyte numbers (C) from representative H&E-stained histological sections (D) at study end. Quantification of percent (E) and total number (F) of KI-67-positive keratinocytes per imaging field from stained histological sections taken at study end. Quantification of dermal thickness (G) and dermal nuclei (H) at study end (Scale bar, 50 µm; n = 5; *P < 0.05, **P < 0.005, t test). (I) Schematic depicting a wound healing study in which Yucatan mini pigs were wounded with partial or full thickness wounds and then dosed topically with PY-60 (0.15 wt%) or Regranex (commercial pharmaceutical grade material) every other day over 8 d. Percent wound contraction measurements derived from images of partial thickness (J) or full thickness (K) wounds at the indicated time points from animals treated with topical PY-60 (0.15 wt%) or Regranex. Representative Masson’s trichrome stained histological sections (L) and quantification of epidermal thickness (M) of wounds obtained from study end (Scale bar, 200 µm; n = 5; **P < 0.005, t test).

We next sought to determine if the ability of PY-60 to expand cells in the skin of naïve animals might result in accelerated cutaneous wound healing. We therefore evaluated the potential of PY-60 (0.15 wt% gel) or Regranex (becalpermin, 0.01% recombinant rhPDGF) to accelerate wound closure in Yucatan mini pigs subjected to partial thickness (removal of the epidermal layer by dermatome) or full thickness wounds (removal of epidermal and dermal layers by surgical resection) over the course of 8 d. Every other day treatment of PY-60 was found to markedly accelerate the closure of both partial and full thickness wounds, as measured by the decrease in cutaneous relative wound area over the study period (Fig. 2 J and K). This contrasts with Regranex (also dosed every other day), which modestly accelerated wound closure in full thickness wounds but displayed no effect in partial thickness wounds. Histological assessment of partial thickness wounds at study end indicated that both PY-60 and Regranex thickened the epidermal layer; however, PY-60 treatment was found increase epidermal thickness by roughly twofold compared to Regranex (Fig. 2 L and M). These results together suggest PY-60 accelerates cutaneous wound repair in the pig, a result likely driven by increased regranulation and re-epithelization of the wound bed.

YAP Activation Promotes Wound Closure and a More Youthful Phenotype in Human Skin.

We next determined whether PY-60 might display a similar regenerative phenotype in human skin by evaluating the potential of PY-60 to accelerate wound closure in human skin equivalents (EpiDermFT, Mattek). These engineered tissues consist of primary normal adult human keratinocytes and dermal fibroblasts, matured in culture such that they possess functional a stratum corneum, epidermis, and dermis at a similar thickness to that of human skin (24). From experiments involving pre-wounded (3 mm biopsy punch from supplier) skin equivalents, PY-60 treatment (0.5 µM) was found to accelerate wound closure to a greater degree than the addition of 1% fetal bovine serum (FBS) or PDGF (10 ng/mL) over 72 h (Fig. 3A). PY-60 was found to promote thickening of the epidermal layer of the wound at study end, a result derived by increasing the total number and the percentage of proliferative keratinocytes in the epidermis (Fig. 3 B–E). Notably PY-60 induced a similar, if not superior, induction of proliferative expansion of the epidermis compared to supplementation with serum or PDGF (Fig. 3 B–E).

Fig. 3.

PY-60 promotes wound closure in human skin equivalents and Induces a more youthful phenotype in middle-aged human skin explants. Morphological measurements of wound diameter (A; mm), epidermal wound thickness (B; µm), representative histological images of KI-67 staining (C; Scale bar, 100 µm), quantification of epidermal nuclei numbers (D, keratinocyte content), and quantification of percent KI-67-positive epidermal cells (E) from wounded human skin equivalents exposed to the indicated treatments for 3 d (PY-60, 0.2 or 0.5 µM; 1% FBS; 10 ng/mL recombinant human PDGF; n = 3; mean and SEM). Representative Masson’s trichrome stained histological sections (F), epidermal thickness measurements (µm, G), and Rete pegs per mm (H) obtained from a 7-d treatment of primary human skin explants with PY-60 (5 or 20 µM), ATRA (10 µM) or 1% FBS (n = 4, mean and SEM; t test).

We next evaluated the effect of PY-60 treatment on explanted primary human skin. We were able to establish conditions in which 1 cm diameter biopsy punches of human skin derived from surgically discarded skin could be maintained for more than 1 wk in tissue culture. From experiments involving 1 cm full thickness biopsy punches derived from abdominal skin of a 41-y–old Caucasian female, we found that PY-60 treatment (5 and 20 µM in DMSO supplemented to the culture medium) was found to dose dependently increase the thickness of the epidermis of this tissue by more than twofold over this period, as assessed by quantifying Masson’s trichrome stained histological sections at study end (Fig. 3 F and G; 7 d of treatment). A notable observation from this donor was a decreased incidence of Rete pegs, dermal epithelial extensions that are predominant in young skin and help maintain integrity in response to shear stress. Interestingly, PY-60 was found to also increase the number of Rete pegs in these tissues (Fig. 3H). The addition of 1% FBS to cultures as a comparator additionally resulted in increased epidermal thickness and the number of Rete pegs but to a lesser magnitude (Fig. 3 F–H). We also compared this effect to all trans retinoic acid (ATRA, also called Retinol), which is used extensively in the cosmetic industry. ATRA treatment (10 µM) resulted in a desquamating effect, decreasing the thickness of the epidermal layer, and displayed no statistical effect in modulating the number of Rete pegs over this period, although there was a visible effect at increasing the intensity of dermal collagen (Fig. 3 F–H).

Cutaneous Proliferation Induced by YAP Activation Is Reversible.

We next sought to understand if the proliferative effect of PY-60 on the skin of mice might be reversible. Accordingly, we performed an experiment in naïve adult mice (12 wk old) in which cohorts were dosed topically with vehicle or PY-60 (0.15% or 1.5% gel) once daily for 10 d. At this time point, half the cohort was then killed while the remaining animals continued without treatment for an additional 20 d (Fig. 4A). As anticipated, mice treated with either dose of PY-60 for 10 d displayed a robust thickening of the epidermis, a result derived from increased proliferation as measured by percent KI-67-positive keratinocytes (Fig. 4 B–D). Mice treated with PY-60 also displayed a robust expansion in the number of quantifiable dermal nuclei at day 10 of the study (Fig. 4E), suggesting a strong effect on dermal fibroblast expansion like that observed in the pig. Interestingly, mice that had been treated with compound for 10 d and then allowed to recover for an additional 20 d displayed a phenotype that was statistically indistinguishable from that of vehicle-treated animals in all measurements (epidermal thickness, percent KI-67-positive keratinocytes, and dermal nuclei numbers; Fig. 4 B–E) indicating that the pro-proliferative effects of compound had been reversed upon removal of the test agent.

Fig. 4.

The proliferative effects of topical YAP activation by PY-60 are reversible. (A) Schematic depicting the design of a mouse study in which naïve mice were topically administered PY-60 (0.15 or 1.5 loading percent) or vehicle for 10 d and then evaluated for reversal of phenotype. (B) Representative images of H&E-stained histological sections at the indicated timepoints. Quantification of epidermal thickness (C), KI-67-positive cells (D), and dermal nuclei (E) from histological sections at the indicated time points (n = 3, mean and SEM; t test; **P < 0.005, ***P < 0.0005, NS = not statistically significant; scale bar, 100 µm).

Lastly, we performed several initial preclinical profiling experiments with PY-60 to understand whether this chemotype might be suited as a potential therapeutic agent for promoting wound repair. We first measured the intrinsic clearance of intravenously administered PY-60 (5 mg/kg), evaluating the plasma levels of compound over a 24-h period in the mouse (T_1/2 = 0.46 h; AUC_0-24 = 623 ng*h/mL; SI Appendix, Fig. S5A). PY-60 demonstrates fast clearance (133 mL/min/kg) exceeding that of mouse hepatic blood flow, suggesting that it is rapidly cleared by first pass metabolism and does not display sufficient exposure to activate YAP in internal organs. We then assessed if the compound might display intrinsic phototoxicity. We found that PY-60 did not display measurable cytotoxicity to 3T3 fibroblasts at concentrations up to 100 µM in neutral red uptake assays either in the presence or absence of UV irradiation (SI Appendix, Fig. S5B). Likewise, from in vitro experiments measuring reactive oxygen species (ROS) production, unlike the known phototoxic agent quinine, PY-60 (50 µM) did not promote the generation of superoxide anion or singlet oxygen in response to solar irradiation (SI Appendix, Fig. S5C). Together, these data are suggestive that PY-60 might serve as a potential therapeutic development candidate for activating YAP topically with minimal systemic exposure.

Discussion

Here, we explored whether a pharmacological activator of YAP, PY-60, might promote a pro-proliferative wound healing response in mammalian tissues. First, from in vitro experiments aimed at defining its effects in normal adult human keratinocytes (HEKas), we found that PY-60 induced a robust YAP driven transcriptional response associated with decreased terminal keratinocyte differentiation and apoptosis. This pro-proliferative effect accelerated wound healing and increased migratory capacity of HEKa cells in vitro. Interestingly, PY-60 also enabled keratinocytes to grow as spheroids, a property associated with increased stemness. This contrasts with PPARd agonism, which could increase wound healing but not promote spheroid growth. These results suggest that YAP activation may result in the transient acquisition of a stem cell-like phenotype or potentially expand existing keratinocyte precursors. Relatedly, transient YAP activation has been shown to reprogram terminally multiple differentiated cell types to proliferative progenitors capable of ex vivo expansion (20). The extent to which PY-60 might act similarly by inducing or modulating the activity of keratinocyte precursor populations will be of high interest to the field in future work.

To enable the targeted tissue delivery of PY-60 for wound healing, we developed a gel-based topical formulation, which enabled up to 1.5 weight percent loading of PY-60. Gel-formulated compound promoted the expansion of epidermal keratinocytes and dermal fibroblast populations in naïve mini pigs. The capacity to promote proliferation of both epidermal and dermal cellular populations enabled accelerated wound closure in mini pig models of partial and full thickness wounds and resulted in thickened dermal and epidermal skin layers of the wound at study end. Because PY-60 promotes proliferation of both epidermal and dermal cellular populations, these results are suggestive that pharmacological YAP activation likely increases both regranulation of the wound bed as well as re-epithelialization, leading to accelerated wound closure. This contrasts with Regranex (topical recombinant PDGF) which only accelerates regranulation by promoting the expansion of dermal fibroblasts (7). Additionally, the ability to induce thicker and more resilient wounds has been associated with a higher “wound quality” that is less susceptible to tearing or reinjury (2). As such, agents like PY-60 may not only be useful in accelerating the rate of wound closure but may also affect the potential incidence of reinjury in at risk populations.

Although we have demonstrated a proliferative effect on both epidermal keratinocytes and dermal fibroblasts, effective wound healing involves the concerted effort of several other cellular populations including endothelial cells to revascularize wounded tissue as well as neurons to re-innervate the various layers of the skin (6). Likewise, chronic wounds display increased infiltration of activated inflammatory cells (25). Lastly, a body of literature suggests that hair follicles themselves play a role in wound healing and remodeling of the skin by manipulating the effect of nearby cellular populations (26). YAP has been shown to control several stages of hair follicle development and is highly expressed in proliferative basal cellular populations during the anagenic phase of the hair cycle (21). The extent to which PY-60 might manipulate these and other cellular populations will necessarily be the work of future studies (21). Indeed, longitudinally tracking the transcriptional activities in all skin cell types using single-cell RNA sequencing will undoubtedly provide a means to investigate what additional roles pharmacological YAP activation may have in promoting wound healing.

Interestingly, recent work has shown that pharmacological inhibition of YAP results in wound healing in mice without permanent scarring (27, 28). This phenotype has been attributed to YAP enacting a proscarring transcriptional program in a subset of dermal fibroblasts that decreases the capacity for epithelial cell types to efficiently repopulate the wound. In contrast, others studying the spiny mouse (Acoymys cahirinus), a mammal that heals without scarring, have shown that YAP activation in dermal fibroblasts is central to the ability of this species to close dermal wounds without fibrosis (29). Intriguingly, this group demonstrated that strong YAP activity in the postwound setting followed by a period of little to no YAP-driven transcription several weeks later is essential for this phenotype (29). These data along with the data presented here suggest a complex relationship by which the magnitude and/or duration of YAP transcriptional activity in several cell types dictate the phenotypic outcome of a given wound. Future time course studies using genetics and pharmacological tools like PY-60 in several species will help address these apparent discrepancies.

As noted above (27, 28), excessive YAP activity in dermal fibroblasts may result in a phenotype of “over healing” in which increased scar tissue inhibits new hair follicle growth and efficient epidermal repair. This scenario contrasts with that of a chronic wound, in which inflammation and other factors result in an “under healing” phenotype with a decreased proliferative phase. As such, augmenting the level of YAP may be of therapeutic benefit to rebalance the potential of a chronic wound to heal. Indeed, the key goal of treating chronic wounds is promoting wound closure as quickly as possible, and not necessarily whether the affected area appears aesthetically normal. Future studies involving topical PY-60 treatment in animal models of chronic wounds, like those of diabetic ulcers in db/db mice or in streptozotocin-treated pigs, will help shed light on whether YAP activation may be of therapeutic utility in healing chronic wounds.

In addition to chronic wounds, YAP activation may be of potential therapeutic utility in accelerating the rate of acute wound healing, especially in clinical settings like physical trauma or burns in which time to healing is also critical to patient survival. Indeed, one of the major life-threatening complications that occurs in the postburn setting is acute fluid loss due to lost barrier function (30, 31). As such, accelerating the rate of re-epithelialization would likely be of high medical benefit, forestalling the potential for lethal fluid loss and ensuing infection. Future studies aimed at evaluating PY-60 in additional acute wound healing scenarios will clarify if pharmacological YAP activation will be of benefit in this patient segment.

Beyond primary keratinocytes, we have shown that PY-60 treatment promotes wound closure in human skin equivalents, engineered human skin that displays similar thickness and morphology to that of primary human tissue. Here, PY-60 was also found to promote wound closure due to increased proliferation of epidermal keratinocytes, a result suggesting that PY-60’s effect in animals can most likely be recapitulated in humans. Consistent with this notion, we have shown that PY-60 induced a pro-proliferative phenotype in primary human skin tissue explants from a middle-aged door, increasing epidermal thickness by more than twofold over the course of a week in culture. In addition to wound healing, the ability of PY-60 to rapidly thicken the skin of aged individuals may be of high therapeutic utility in treating senile purpura, a common condition involving easily bruised skin due to decreased epidermal thickness of aged skin (32). In addition to thickening the epidermal layer, PY-60 was also found to increase the amount of dermal collagen and the number of Rete pegs in primary human skin explants. Key characteristics which differentiate older skin from that of youthful skin are thinning of the epidermis, decreased dermal collagen, and fewer Rete pegs per unit length (33). The observation that PY-60 positively reverses these three phenotypes raises the compelling idea that transient topical YAP activation may display utility in aesthetic dermatology, potentially increasing the youthful character and appearance of aged skin to a greater degree than retinoids, which are which are used prevalently in the skin care industry and only show modest effects in our explant model.

A key concern with the use of pro-proliferative regenerative therapies is the potential for inducing proliferative disease. To begin to address this issue, we have performed a study in mice which demonstrated that the pro-proliferative effects of PY-60 in the skin of mice are reversible within 20 d, a result that is consistent with the rate of epidermal turnover in mice of 8 to 10 d (34). This result indicates that PY-60 treatment does not obligately result in runaway proliferation. Instead, these results are suggestive that the transient remodeling of the epidermal and dermal layers imparted by compound treatment is not permanent and that likely additional intrinsic tissue resident mechanisms exist for maintaining a set organ size. To the extent that the Hippo pathway may be involved in signaling to or helping to maintain such “set points” will be of keen interest in future work.

Lastly, we have shown that PY-60 has high intrinsic clearance from plasma, exceeding that of hepatic blood flow. As such, we believe that fast clearance molecules applied locally to the site of injury have the potential to be useful regenerative therapeutics with minimized risk for inducing proliferative disease. Future studies aimed at understanding the long-term safety of PY-60 and similar molecules will necessarily be required before therapeutic trials. Collectively, however, the work presented here is suggestive that topical application of a pro-proliferative YAP activating agent may display utility in treating cutaneous wounds and other dermatological diseases associated with aging.

Methods

Cell Culture.

Primary human epidermal keratinocytes (normal human adult, HEKa) were obtained from American Type Culture Collection (PCS-200-011, ATCC) and cultured in dermal basal medium supplemented with a Keratinocyte growth kit (PCS-200-030 and PCS-200-040, also from ATCC). 293A cells were from Thermo Fisher and were maintained in Dulbecco’s Modified Eagle Medium (DMEM) (10-013-CMR, Corning) supplemented with 10% FBS (Gibco) and 1% penicillin-streptomycin solution (Penicillin–Streptomycin 10,000 U/mL, Gibco). Cells were used at passage 2 to 3 post-thawing and routinely evaluated for contaminating mycobacterium using an in-house testing service.

Chemicals.

Quinine, GW0742, and ATRA were from Cayman Chemical. Topical PY-60 was formulated (1.5 or 0.15 weight percent) as a gel (propylene glycol: 26%, Labrasol ALF: 17.5%, PEG 400: 46.5%, Povidone K-90: 8.5%) at Aragen Life Sciences. The methods to synthesize the batch of PY-60 used for the gel formulation used in this work is provided in the SI Appendix.

Transcriptional Profiling.

Primary human keratinocytes were plated at 5 × 10⁵ per well in six-well plates in 2 mL of keratinocyte growth medium as above. The next day, cells were treated with PY-60 (10 µM) or DMSO for an additional 24 h, at which time cells were trypsinized, collected by 3-min centrifugation at 300 g, and their total RNA isolated using an RNeasy kit (Qiagen). Abundance of ANKRD1 was assessed by qRT-PCR (SYBR green reagents) after generation of cDNAs using a reverse transcription kit (Super Script III First Strand Synthesis system). Abundance was estimated using the delta Ct method with primers targeting TUBG1. For RNA sequencing, purified RNA samples were sent to BGI Genomics (Shenzhen, China) and analyzed via proprietary DNBSEQ technology.

Time under Compound Treatment Assay.

293A cells were plated in 2 mL of growth medium in six-well plates (5 × 10⁵ cells per well). They were incubated for 48 h at 37 °C at which time their medium was changed to growth medium containing PY-60 (10 µM). After incubation for the indicated time interval, cellular monolayers were washed three times with phosphate-buffered saline (PBS, Corning) and then three times over a 15-min period with growth medium to ensure maximal compound washout. Cells were then incubated in growth medium for the remainder of the washout period until they were collected by trypsinization and the abundance of ANKRD1 measured using the methodology above.

Keratinocyte Proliferation Assay.

HEKa cells were plated at low density (10⁴ cells per well) in 1 mL of HEKa medium in 24-well tissue culture plates. Cells were allowed to attach overnight at 37 °C at which time they were treated with the indicated concentrations of PY-60 using 1,000× DMSO stocks of the compound. Cells were incubated for an additional 72 h and were then fixed using 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) for 15 min. Wells then washed three times with PBS and were then exposed to a PBS solution containing 2 µg/mL Hoechst 33342 dye for 1 h in the dark at room temperature (RT). Wells were washed again three times with PBS and a representative image per well with 100× magnification was captured using Eclipse Ti inverted microscope (Nikon). Images were processed using an automated image analysis macro within Image J (FIJI application, NIH) for quantifying Hoechst-positive nuclei and data were expressed as cells per imaging field.

Keratinocyte In Vitro Wound Closure Assay.

To enable uniform distance separation between cellular monolayers in culture and thereby mimic wounded tissue, silicon inserts from Ibidi (80209) were firmly secured onto the surface of each well of a 24-well tissue culture plate using sterile forceps. To each chamber of the insert was dispensed 5 × 10⁴ HEKa cells in 80 µL of growth medium. After 24 h of incubation at 37 °C and the cells had achieved full confluency within the chamber and the silicon insert was removed using sterile forceps thereby mimicking a wound. One milliliter of medium with or without the supplementation of 0.5% FBS was then added to each well at which time the indicated compound was added as 1,000× dilution from a DMSO stock. Cultures were incubated for 24 h at 37 °C and then fixed with 4% PFA in PBS for 10 min. Cells were permeabilized and stained with Phalloidin-488 (Invitrogen, 1:500) in PBS containing 1% FBS and 0.1% Triton X-100 for 1 h in the dark. The wells were washed three times with PBS and then imaged using an Eclipse TI microscope (Nikon). Measurements were made in Image J (FIJI application, NIH) and data were expressed as relative wound closure based on images taken from freshly wounded wells.

Histological Staining.

Paraffin sections were dewaxed in xylene and rehydrated through gradually from ethanol to water. Antigen retrieval was performed by microwaving sections in 0.01 M citrate buffer, pH 6.0, for 20 min at 500 W. For Ki-67 staining, tissue sections were incubated with CAS-Block™ buffer (Invitrogen) at RT for 5 min. Sections were then exposed to KI-67 monoclonal antibody (SolA15, Invitrogen) (1:100 in CAS-Block™) at RT for 1 h. Sections were then washed three times with dH2O for 5 min and incubated with secondary antibody (ImmPRESS®-AP Goat Anti-Rat IgG Polymer Detection Kit, Alkaline Phosphatase from Vector Laboratories) at RT for 15 min. The tissue sections were washed three times for 5 min with dH2O and were then exposed to working solution (ImmPACT® Vector® Red Substrate Kit, Alkaline Phosphatase, Vector Laboratories) at RT until the desired staining intensity had developed. For K14 staining, the tissue sections were incubated with 3% hydrogen peroxide at RT for 10 min. Next, sections were washed twice with dH2O and once with TBST (1x Tris Buffered Saline with Tween 20) for 5 min each. The sections were blocked with 1× Animal-Free Blocking Solution (#15019, Cell Signaling Technology) at RT for 1 h and then incubated overnight at 4 °C with anti-Cytokeratin 14 antibody SP53 (ab119695, Abcam; 1:100 in SignalStain® Antibody Diluent #8112 from Cell Signaling Technology). After decanting the antibody solution, tissue sections were washed with TBST three times for 5 min each and then incubated with SignalStain® Boost Detection Reagent (HRP, Rabbit #8114, Cell Signaling Technology) in a humidified chamber at RT for 30 min. Sections were washed an additional three times with TBST for 5 min each and then incubated with the working solution (SignalStain® DAB Substrate Kit #8059, Cell Signaling Technology) at RT until the desired staining intensity had been achieved. All histological sections from a given study were exposed to the same staining and developing solutions to mitigate any differences in intensity.

Pharmacodynamic Assessment in Pigs.

Pharmacodynamic studies in pigs were performed at Absorption Systems. One female Yucatan mini pig (30 kg, S&S Farms) was used for the study. The dorsal region of the animal was shaved 24 h prior to dosing. Fifteen 3 × 3 cm sections of the dorsal region were demarcated with indelible ink and dosing groups were randomly distributed such that no region was enriched for a given treatment. Animals were dosed once daily with 500 µL of gel-formulated material by hand for 10 d. After each application of test article, the area was covered with Tegaderm and Elastikon to prevent test article from moving. Twenty-four hours after the last dose, the animal was sacrificed by the intravenous administration of 150 mg/kg Beuthanasia (VetOne). The test sites were wiped clean using a 70% EtOH aqueous solution and one full thickness biopsy punch containing all dermal layers to the muscle fascia was collected per animal. Staining of histological slides with hematoxylin and eosin (H&E) was carried out the Sanford Burnham Prebys histological core. KI-67 staining was achieved as noted above 15 epidermal thickness were performed per testing site using Aperio histologial viewing software and expressed as an average measurement in microns per site. Fifteen representative snapshots of dermal layers per wound were used to quantify dermal nuclei numbers using a custom Image J macro. The abundance of CTGF and CYR61 was measured using the methodology above.

Wound Healing Study in Pigs.

Wound healing studies in pigs were performed at Absorption Systems. Four eight-month-old female Yucatan mini pigs (34 to 40 kg, S&S Farms) were used in the study. Each animal had their dorsal regions shaved 24 h prior to wounding. Under anesthesia, each animal received either 12 partial thickness or 12 full thickness wounds (3 × 3 cm area per each). Partial thickness wounds were achieved using a dermatome achieving a total wound depth of 3 to 5 mm (epidermal layer and some dermal component). Full thickness wounds were achieved via surgical scalpel such that all layers of the skin were removed leaving exposed muscle fascia. Each wound received 500 µL of formulated gel or Regranex (Smith & Nephew) every other day under anesthesia. Upon application of test article, several layers of Tegaderm, Telfa pads, and Elastikon were used to bandage wounds and contain the volume of test article within the wound. At study end, animals were sacrificed as above and the entire wound was excised via surgical scalpel. Staining of histological slides with hematoxylin and eosin (H&E) was carried out the Sanford Burnham Prebys histological core. Quantification of epidermal and dermal thickness was achieved using the methodology above.

Human Skin Equivalents.

Human skin equivalents were obtained from Mattek prewounded bearing 3 mm full thickness wounds plugged with plastic stoppers. Upon receipt from supplier, stoppers were removed and equivalents in plastic housings were transferred to fresh six-well plates, washed once with PBS, and then 3 mL of DMEM/F12 medium (Corning) containing 1% pen-strep were added to each well. After 2 h of acclimation at 37 °C, medium was decanted and replaced with 3 mL of DMEM/F12 medium (Corning) containing 1% pen-strep containing the indicated test agent. The volume of medium in wells was sufficient to bathe the dermis but did not contact the epidermal layer, which was exposed to ambient air. After 3 d of incubation at 37 °C, skin equivalents were imaged using an inverted Eclipse Ti microscope (Nikon) before being removed from plastic housings and placed into 4% PFA in PBS overnight before delivering to the Sanford Burnham Prebus Histology core for mounting and H&E staining. KI-67 staining was performed as above.

Human Skin Explant Experiments.

Skin explants were obtained from BioIVT (S02521). One-centimeter–diameter biopsy punch sections were generated from abdominal skin discarded after abdominoplasty procedure from a consenting 41-y-old Caucasian female. Fresh explants preserved for 24 h in DMEM were randomly assigned to six-well dishes and oriented such that the hypodermal face of the section directly contacted the plastic surface of the well. Two milliliters of DMEM/F12 medium containing 1% pen-strep was administered to each well, which enabled survival of the skin section without submerging the epidermal surface of the skin, leaving it in contact with ambient air. Compounds and serum were delivered to the medium for 7 d at which time the entire section was submerged in 4% PFA before sending to the Sanford Burnham Prebys Histology core for mounting and staining with Masson’s trichrome according to in-house procedures. Epidermal thickness was determined as above. Entire sections were evaluated manually for the number of rete pegs (indentations greater than 10 µm) and rete peg number was expressed per unit length.

Neutral Red Uptake Assay.

Neutral red uptake assays were performed at Pharmaron Inc. Briefly, BALB/C 3T3 mouse fibroblasts were exposed to a dose response of compounds as delivered as 2X aqueous solutions in 96-well plates for an hour before exposure to UV irradiation (5 J/cm²). A replicate plate was incubated in the dark. After an additional 30-min incubation, plates were exposed to neutral red solution, destained, and dye uptake assessed by absorbance readings at 540 nm. Photo irritation factor (PIF) measurements were calculated as the fold change in half maximal uptake values of irradiated as compared to non-irradiated samples. PIFs of greater than 5 correspond to a photo-toxic response whereas compounds with PIFs between 2 and 5 are considered equivocal. Values less than 2 are typically considered non-phototoxic. Chlorpromazine (PIF = 71.23) and Tamoxifen (PIF = 1.33) were used as positive and negative controls for the assay respectively.

In Vitro ROS Formation Assays.

To measure the propensity to form ROS, two absorbance-based assays were employed to detect either superoxide anion production or singlet oxygen. To detect superoxide, 50 µM of test compound was incubated in 20 mM sodium phosphate buffer pH 7.8 containing 50 µM nitroblue tetrazolium (NBT method). Absorbance readings were taken at 540 nm before and after 4 h of exposure to solar radiation. To detect singlet oxygen, 50 µM of test compound was incubated in 20 mM sodium phosphate buffer pH 7.8 containing 50 µM p-nitrosodimethylaniline (RNO) and 50 µM imidazole (RNO and imidazole method). Absorbance readings were recorded at 450 nm before and after 4 h of exposure to solar radiation. Reactions were performed in parallel using clear UV-compatible 96-well plates in a volume of 200 µL per well with three technical replicates. Absorbance measurements were captured using SpectraMax 250 plate reader (Molecular Devices) and were expressed as a thousand times the change in absorbance as per convention.

Mouse Pharmacokinetics.

Mouse pharmacokinetics were performed at Pharmaron Inc. Briefly, PY-60 was formulated at 2.5 mg/mL in 60% PEG300 and 50% D5W as a clear solution. Five milligram per kilogram of material was delivered via tail vein injection to fasted male CD-1 mice. Quantification of compound concentration was performed via liquid chromatography tandem mass spectrometry (API 4000 instrument, Sciex) derived from first generating a standard curve of PY-60. Pharmacokinetic parameters were calculated using an IV-noncompartmental fit model using Phoenix WinNolin version 6.3 software.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

RNA sequencing data have been deposited in National Center for Biotechnology Information Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE234332) (35).