Inhibiting mechanotransduction prevents scarring and yields regeneration in a large animal model

Shamik Mascharak; Michelle Griffin; Heather E Talbott; Jason L Guo; Jennifer Parker; Annah Grace Morgan; Caleb Valencia; Maxwell Michael Kuhnert; Dayan J Li; Norah E Liang; Rachel M Kratofil; Joseph A Daccache; Ikjot Sidhu; Michael F Davitt; Nicholas Guardino; John M Lu; Darren B Abbas; Nestor M D Deleon; Christopher V Lavin; Sandeep Adem; Anum Khan; Kellen Chen; Dominic Henn; Amanda Spielman; Asha Cotterell; Deena Akras; Mauricio Downer, Jr; Ruth Tevlin; H Peter Lorenz; Geoffrey C Gurtner; Michael Januszyk; Shruti Naik; Derrick C Wan; Michael T Longaker

doi:10.1126/scitranslmed.adt6387

. Author manuscript; available in PMC: 2026 Feb 13.

Published in final edited form as: Sci Transl Med. 2025 Feb 19;17(786):eadt6387. doi: 10.1126/scitranslmed.adt6387

Inhibiting mechanotransduction prevents scarring and yields regeneration in a large animal model

Shamik Mascharak ^1,^2,^†, Michelle Griffin ^1,^†, Heather E Talbott ^1,^2,^†, Jason L Guo ¹, Jennifer Parker ^1,², Annah Grace Morgan ¹, Caleb Valencia ¹, Maxwell Michael Kuhnert ¹, Dayan J Li ¹, Norah E Liang ¹, Rachel M Kratofil ^3,⁴, Joseph A Daccache ^3,⁴, Ikjot Sidhu ^3,^4,⁵, Michael F Davitt ¹, Nicholas Guardino ¹, John M Lu ^1,², Darren B Abbas ¹, Nestor M D Deleon ¹, Christopher V Lavin ¹, Sandeep Adem ¹, Anum Khan ¹, Kellen Chen ¹, Dominic Henn ¹, Amanda Spielman ¹, Asha Cotterell ¹, Deena Akras ¹, Mauricio Downer Jr ¹, Ruth Tevlin ¹, H Peter Lorenz ¹, Geoffrey C Gurtner ¹, Michael Januszyk ¹, Shruti Naik ^3,^4,^6,⁷, Derrick C Wan ^1,^*, Michael T Longaker ^1,^2,^*

PMCID: PMC12893899 NIHMSID: NIHMS2079743 PMID: 39970235

Abstract

Modulating mechanotransduction by inhibiting yes-associated protein (YAP) in mice yields wound regeneration without scarring. However, rodents are loose-skinned and fail to recapitulate key aspects of human wound repair. We sought to elucidate the effects of YAP inhibition in red Duroc pig wounds, the most human-like model of scarring. We show that one-time treatment with verteporfin, a YAP inhibitor, immediately after wounding is sufficient to prevent scarring and to drive wound regeneration in pigs. By performing single-cell RNA sequencing (scRNA-seq) on porcine wounds in conjunction with spatial proteomic analysis, we found perturbations in fibroblast dynamics with verteporfin treatment and the presence of putative pro-regenerative/profibrotic fibroblasts enriched in regenerating/scarring pig wounds, respectively. We also identified differences in enriched myeloid cell subpopulations after treatment and linked this observation to increased elaboration of interleukin-33 (IL-33) in regenerating wounds. Finally, we validated our findings in a xenograft wound model containing human neonatal foreskin engrafted onto nude mice and used scRNA-seq of human wound cells to draw parallels with fibroblast subpopulation dynamics in porcine wounds. Collectively, our findings provide support for the clinical translation of local mechanotransduction inhibitors to prevent human skin scarring, and they clarify a YAP/IL-33 signaling axis in large animal wound regeneration.

INTRODUCTION

Tissue repair occurs on a spectrum, with possible outcomes ranging from regeneration to fibrosis. Although restoration of a tissues prewounding architecture is the ideal healing outcome, most mammalian injuries including human skin wounds heal by deposition of abnormal, fibrotic scar tissue (1). Scarring is detrimental to skin function, because scar tissue cannot recapitulate the barrier, thermoregulatory, or sensory features of unwounded skin and can substantially limit joint mobility and even growth (2–4). However, despite extensive research to date, no drug treatments exist to prevent or treat scarring. Because of the lack of effective targeted therapies to prevent this fibrotic repair outcome, skin scarring (whether after traumatic injury, burns, or surgery) remains a major health burden in the United States and worldwide (2).

Mouse studies have played an instrumental role in establishing mechanisms of scarring, including the central role of tissue mechanical forces and cell mechanotransduction in promoting wound fibrosis. We recently showed that wound mechanical stimuli promote yes-associated protein (YAP)–mediated mechanotransduction signaling in dermal fibroblasts, leading to expression of the homeobox protein engrailed-1 and activation of profibrotic molecular machinery (1). Blocking this signaling pathway using verteporfin, a small-molecule YAP inhibitor, not only prevents scarring but also promotes robust wound regeneration in mice, with recovery of hair follicles (HFs), extracellular matrix (ECM) ultrastructure, and mechanical strength all comparable to those found in unwounded skin (1, 5, 6).

These findings support that verteporfin treatment of wounds could be a promising translational strategy for targeting human scarring. However, the clinical relevance of mouse studies is limited by substantial interspecies biological differences between rodents and humans (7, 8). In particular, the mechanical properties of mouse versus human skin, and their relative propensity for fibrosis, differ dramatically. Mice are loose-skinned and have low resting skin tension; humans and other large mammals, such as pigs, are tight-skinned and have much higher resting skin tension, a property correlated with a greater baseline degree of scarring (9). Thus, mouse research is vital to early discovery, but large animal models are ultimately essential for bridging the translational gap in our understanding of scarring and wound healing. Red Duroc pigs are an established large mammalian model, which, because of their highly human-like skin and scar architecture, mechanical properties, and healing kinetics, are considered the “gold standard” preclinical model of skin scarring (10–12). Here, we investigate the effects of verteporfin treatment on scarring in a red Duroc pig wound model.

RESULTS

YAP inhibition promotes wound regeneration in red Duroc pigs

We applied a well-established model for wound repair and scarring in red Duroc pigs (13), with all surgical procedures performed by advanced surgical residents and board-certified plastic surgeons to maximize translational relevance and minimize technical variability. Full-thickness hexagonal excisional wounds were made on pigs’ dorsa and, before closure, received one-time treatment with either verteporfin, a small molecule that inhibits YAP mechanotransduction activity by blocking YAP’s interaction with its transcriptional coactivators, TEA domain transcription factors (TEADs) (14), or vehicle control (Fig. 1A and fig. S1A). Treatments were administered through local, intradermal injection into the edges and base of the wound. To avoid bias based on location-specific factors that could possibly influence healing outcomes, treatment conditions were randomized by individual wound. Wounds were then closed under tension in a standard layered fashion by bringing the top and bottom edges of the hexagon together. This closure strategy placed the entire length of the wound under similar supraphysiological tension akin to that of closed human surgical wounds, driving a substantial degree of scarring and making this a rigorous model for assessing scar prevention. Wounds, and unwounded skin sites for comparison, were subjected to both longitudinal analyses and end point analysis after euthanasia and harvested at 2, 4, 8, or 16 weeks of healing (including histologic and tissue mechanical analyses; see fig. S1B for a full summary of all experimental animals, treatment/dosing conditions, and analyses for each individual animal).

Grossly, vehicle-treated control wounds formed scars that remained apparent by week 16 (Fig. 1B, top row). In contrast, verteporfin-treated wounds had minimal visible scarring at week 4 and yielded virtually no gross scar by week 16 (Fig. 1B, bottom row). Histologically, on hematoxylin and eosin (H&E) staining, control wounds formed a clear, densely fibrotic “bare area” (Fig. 1C, left column). Verteporfin-treated wounds contained invaginating structures resembling early regenerating HFs as early as 4 weeks after wounding, which were also present in unwounded skin but conspicuously absent from control wounds (Fig. 1, C and D), consistent with the typical fibrotic/scarring wound healing outcome in which secondary elements are not regenerated. Verteporfin-treated wounds also had a greater amount of regenerated subcutaneous fat, more closely resembling the fat content of unwounded skin, compared with a relative lack of subcutaneous fat in control wounds (Fig. 1C). On Visual Analog Scale (VAS) scoring (1 to 10 scale; 1, severe scarring; 10, normal unwounded skin), verteporfin-treated wounds consistently scored four to five points higher than control scars (Fig. 1E; all differences significant, P ≤ 0.013 at all time points), indicating that these wounds had an improved appearance closer to that of unwounded skin.

We next sought to further analyze the presence of neogenic HFs in verteporfin-treated wounds using histology. On closer examination of H&E histology, follicles present within healed verteporfin-treated wounds were histologically distinct from those in surrounding skin. These HFs had smaller amounts of regenerated fat associated with the base of the follicle (Fig. 1C, right column, fourth row, white arrowhead), further supporting that these were truly neogenic and in a distinct phase of follicle development from preexisting HFs. We then performed immunofluorescent (IF) staining for leucine-rich G protein–coupled receptor-5 (LGR5), a conserved marker of HF stem cells (15, 16), and lymphoid enhancer binding factor-1 (LEF1), a WNT regulatory protein expressed in developing and anagen-phase HFs and implicated in HF regeneration (17, 18). There were small epidermal invaginations and condensations present within 8-week verteporfin-treated wounds whose bases exhibited positive staining for LEF1, consistent with early neogenic hair placodes. These were entirely absent from vehicle-treated wounds (fig. S2A). Verteporfin-treated wounds at 16 weeks contained more mature regenerated HFs with stronger bulge-specific LGR5 staining and LEF1 in the apparent dermal papilla (Fig. 1F and fig. S2B). No such structures were observed in any vehicle-treated wound specimens we examined throughout the study. Collectively, this histologic assessment and IF-based validation of HF markers support that verteporfin-treated wounds harbor bona fide regenerated HFs, rather than nonspecific epidermal cyst-like structures, whereas control wounds fail to even initiate such regenerative features.

We previously showed that verteporfin-induced wound regeneration in mice involves reduction of both canonical mechanotransduction signaling (such as YAP/TEAD and integrin-/cell adhesion–related pathways) and Engrailed-1 lineage-positive fibroblasts (EPFs), which are marked by expression of the surface marker dipeptidyl peptidase 4 [DPP4; also known as cluster of differentiation 26 (CD26)] and are known to be the cellular culprits of scarring in mice (19, 20). To interrogate whether a similar mechanism could be at play in verteporfin-treated pig wounds, we performed IF staining for both YAP and DPP4 in control and verteporfin-treated pig wounds and unwounded skin. As expected, DPP4 expression was widespread throughout the dermis of both unwounded skin [consistent with DPP4⁺ EPFs being the majority fibroblast population in uninjured skin (19)] and vehicle-treated wounds but was significantly abrogated in verteporfin-treated wounds (P ≤ 0.046 at all time points), in keeping with our previous findings that verteporfin treatment in mouse wounds led to healing mediated by Engrailed-1 lineage-negative fibroblasts (ENFs; DPP4⁻) with minimal EPFs (1) (Fig. 1G, top row, and fig. S2, C and D). YAP expression was minimal in the unwounded dermis (21) but significantly elevated in control (scarring) wounds (P ≤ 0.011 at all time points), whereas in verteporfin-treated wounds, YAP expression was reduced and instead comparable to that in unwounded skin (Fig. 1G, bottom row, and fig. S2, E and F, left panel), consistent with verteporfin treatment inhibiting mechanotransduction signaling. Further supporting this assessment, the nuclear localization of YAP, a localization process that is known to be mechanically driven (22, 23) and is necessary for YAP to exert its transcriptional effects, was significantly reduced in verteporfin-treated wounds at 2 weeks postwounding (P = 0.00090; fig. S2F, right panel), consistent with not only YAP expression but also YAP activity being reduced in the context of verteporfin treatment.

Whereas verteporfin-treated wounds had subjectively improved appearance compared with control scars (Fig. 1, B to E), visual assessment of scarring is known to be bias-prone and variable between observers. To avoid these pitfalls, we applied our previously developed ECM ultrastructure algorithm that treatment-agnostically analyzes hundreds of features of each ECM fiber and ECM networks from picrosirius red–stained tissue sections (1, 5, 24–26). Visually, control wounds had dense, linearly aligned ECM typical of a scar, whereas verteporfin-treated wounds had less-dense, basketweave-like ECM more consistent with the appearance of unwounded skin ECM (Fig. 1H, top three rows). Quantified fiber features for unwounded skin and wounds at multiple healing time points were then projected onto a two-dimensional (2D) manifold. We observed a gradual separation of unwounded skin and control scar matrix architecture over time consistent with developing wound fibrosis in control wounds; verteporfin-treated wounds, however, had matrix feature intermediate between those of unwounded skin and control scars across all time points and overlapped strongly with unwounded skin at 16 weeks (Fig. 1H, fourth row). Thus, the ECM analysis algorithm showed that verteporfin treatment yielded wound ECM features with objectively greater overlap with those of healthy, unwounded skin and that these conditions were distinct from control scar ECM architecture (1).

Because of its densely fibrotic ECM, scar tissue is stiffer than normal skin, the latter being more elastic and pliable (27). However, as demonstrated by our ECM analysis of scars and skin (Fig. 1H), scar ECM also lacks the interwoven, “basketweave” pattern of ECM fibers that lends unwounded skin its relative tensile strength while maintaining flexibility. Despite containing more and denser collagen than unwounded skin, scars are weaker and less mechanically robust, possessing at most 80% of unwounded skin’s original strength (12). To directly assess whether verteporfin treatment yielded wounds with more favorable (unwounded skin–like) mechanical properties, we used both longitudinal and end point analyses to measure the tissue mechanical properties of verteporfin- and vehicle-treated wounds. First, we used a cutometer, which evaluates tissue’s deformability, to monitor wounds’ relative elasticity over the course of healing. Control wounds consistently decreased in elasticity over time and remained relatively inelastic by 16 weeks (Fig. 1I, red points). By contrast, verteporfin-treated wounds, after an initial decrease, regained elasticity by week 16 and had more unwounded skin–like elastic properties than control wounds (Fig. 1I), suggesting that YAP inhibition at the time of wounding may yield continued regenerative changes throughout the later remodeling stages of wound healing. We next performed tensile testing on harvested wounds and skin (Fig. 1J). Control wounds had significantly lower ultimate tensile strength (P = 0.0050), indicating weaker tissue, but a significantly higher Young’s modulus (P = 0.0073), indicating greater stiffness, by 16 weeks compared with unwounded skin (Fig. 1, K and L, green versus red). In contrast, both ultimate tensile strength and Young’s modulus of verteporfin-treated wounds were comparable to those of unwounded skin by 16 weeks (Fig. 1, K and L, green versus blue), suggesting that these wounds, like unwounded skin, were stronger and more pliant than control scars.

Local treatment with verteporfin is safe and effective for promoting regeneration in a porcine wounding model

Although our initial drug concentration of 2 mg/ml was selected on the basis of optimal dosing identified in mice (1), we wanted to rigorously determine the ideal dose of verteporfin for promoting wound regeneration in a large mammal. We treated porcine wounds as above with either vehicle or one of four concentrations of verteporfin (1, 2, 4, or 8 mg/ml) and then harvested wounds and unwounded skin after 8 weeks (fig. S1B, pigs #1 and #2). Grossly, verteporfin-treated wounds (1 and 2 mg/ml) had minimal apparent scarring by 8 weeks, whereas some scarring was observed at higher doses of verteporfin (4 or 8 mg/ml) (fig. S3A). Blinded VAS scoring supported optimal wound appearance with the highest VAS scores at the concentration of 2 mg/ml (fig. S3B). Instron testing revealed that whereas Young’s modulus values, which reflect tissue stiffness, did not significantly differ between wound conditions at the 8-week harvest time point (P > 0.05 for all dose comparisons), verteporfin-treated wounds (4 and 8 mg/ml) were weaker than other wound conditions, with lower ultimate tensile strength (fig. S3C). Collectively, these results aligned with our previous findings that excessive dosing of verteporfin may yield reduced benefit, possibly because of toxicity effects of the drug dominating at higher concentrations (1), and supported 2 mg/ml as the optimal concentration for verteporfin treatment to promote regeneration of red Duroc pig wounds. We therefore proceeded with all subsequent experiments using a verteporfin concentration of 2 mg/ml.

Regarding safety, we observed no delayed healing or infections in any wounds treated with verteporfin at any concentration. Although verteporfin was locally administered, we sought to establish the extent of systemic dissemination and any systemic effects. We obtained blood samples from five experimental animals (pigs #2 to #6; fig. S1B) at multiple time points: on the day of wounding and treatment after wound closure [postoperative day (POD) 0] and on POD 2, 4, 7, 14, 21, and 28. We then performed liquid chromatography–mass spectrometry (LC-MS) to measure plasma concentrations of verteporfin; example LC-MS curves are shown in fig. S4A (top panel). In all five assayed pigs, the plasma verteporfin concentration peaked on POD 0 and was negligible by POD 2 and beyond (fig. S4A, bottom panel). Uie highest single blood concentration of verteporfin measured was 13.2 ng of drug/ml of plasma; this occurred in the POD 0 sample from pig #5, which received a cumulative total of 10.8 mg of drug intradermally across all wounds. Verteporfin has a well-established clinical safety profile and is approved for systemic intravenous administration in human patients for the treatment of macular degeneration (28). Uie expected plasma concentration for typical clinical use of verteporfin is reported to be approximately 2 μg/ml, which is more than 150 times the maximum plasma concentration observed in any of our animal subjects. Ulis finding is notable given that each of these animals had multiple verteporfin-treated wounds.

Necropsy by an independent, trained veterinary pathologist reported no grossly or histologically apparent lesions in lung, bone marrow, liver, spleen, kidney, gastrointestinal tract, heart, brain, skin, or eyes in any experimental animal (see fig. S4B for representative histology images). Finally, we performed additional studies for local dermal toxicity by injecting unwounded skin sites with verteporfin (2 mg/ml). Histologic analysis of these sites showed no abnormalities or skin changes 2, 4, or 8 weeks after injection (fig. S4C), consistent with verteporfin causing no local skin toxicity.

Transcriptomic profiling defines scarring and regenerating porcine wound fibroblast subpopulations

Unwounded skin samples and wounds treated with either vehicle or verteporfin (2 mg/ml) as described above were harvested at postoperative weeks 2, 4, 8, and 16 and used for single-cell RNA sequencing (scRNA-seq) analysis (Fig. 2A and fig. S1B). In total, we analyzed 75,621 sequenced cells. We identified numerous known injury-responsive cell types, including myeloid and lymphoid hematopoietic lineage/immune cells, epithelial cells, endothelial cells, pericytes, smooth muscle cells, and fibroblasts (Fig. 2B), suggesting that, broadly, our transcriptomic profiling successfully captured all key cell types previously identified in mouse wounds (5, 29). Initially focusing on fibroblasts, a key cell type involved in scarring, we defined five transcriptionally distinct fibroblast subpopulations (Fig. 2B). Given our previous findings that verteporfin primarily exerts its proregenerative effects in mice by acting directly on fibroblasts (5), we were interested in how verteporfin influenced fibroblast dynamics in our large animal wound model. We thus sought to interrogate fibroblast heterogeneity and differences in fibrotic versus regenerating porcine wounds. Consistent with our histological and ECM analyses, unwounded skin and regenerating verteporfin-treated wounds had similar distributions of the five fibroblast clusters, whereas scarring vehicle-treated wounds had a distinct fibroblast subtype profile (Fig. 2C). In particular, fibroblast cluster 1 was enriched in unwounded skin and verteporfin wound samples at all time points relative to control wounds and generally increased in verteporfin-treated wounds over time. In contrast, fibroblast clusters 2 and 4 were enriched in control wounds relative to verteporfin wounds or unwounded skin and increased in control wounds over time. We thus hypothesized that cluster 1 could represent a more pro-regenerative fibroblast subpopulation and that clusters 2 and 4 may represent a comparatively profibrotic/scarring fibroblast subset.

Fig. 2. — (A) Schematic (BioRender) depicting scRNA-seq analysis of vehicle- and verteporfin-treated pig wounds (at postoperative weeks 2, 4, 8, and 16) and unwounded skin. (B) UMAP plot of scRNA-seq data for all sequenced cells, colored by cell type (Seurat cluster; left panel) or experimental condition of origin {vehicle [phosphate-buffered saline (PBS)]–treated wounds (P) or verteporfin-treated wounds (V) from each harvest week (2, 4, 8, or 16), or unwounded skin (UW); right panel}. (C) Bar graphs depicting number of sequenced cells (top panel) or relative proportion of fibroblasts (bottom panel) belonging to each Seurat fibroblast cluster, among Seurat-classified fibroblasts from each experimental condition. (D) Heatmap showing top differentially expressed genes for each fibroblast Seurat cluster (left; fibroblasts from all conditions/time points), with Gene Ontology (GO) pathway analysis for each cluster (right). (E) Violin plots for genes specifically enriched in fibroblast cluster 1 (first row), 2/4 (second row), 3 (third row), or 5 (fourth row) showing those genes’ expression across all fibroblast clusters. (F) Top panels: Pseudotime analysis of all sequenced fibroblasts, with plots colored by relative pseudotime value (top left panel) and Seurat fibroblast subtype (top right panel). Bottom panels: Pseudotime expression analysis of key genes showing their relative expression across the pseudotime progression from fibroblast cluster 1 to clusters 2/4 (top), cluster 1 to cluster 3 (middle), and cluster 1 to cluster 5 (bottom). (G) Heatmap depicting overall enrichment of cell-cell signaling patterns between each cell type based on CellChat analysis of scRNA-seq data. (H) CellChat analysis of cell-cell interactions among all sequenced cell types for indicated intercellular signaling networks/pathways. (I) Top intercellular signaling pathways relatively enriched in vehicle-treated (blue) versus verteporfin-treated (red) wounds at 2 (first column), 4 (second column), 8 (third column), and 16 (fourth column) weeks postwounding, based on CellChat analysis of cell-cell signaling between fibroblasts and all other wound cell types. Signaling pathways in blue and red text highlight key cell-cell signaling pathways most enriched in each wound treatment condition at 2 weeks (labels to left of plot) versus 16 weeks (labels to right of plot).

Gene set enrichment analyses revealed that fibroblast cluster 1 was enriched for regeneration- and development-related pathways, such as Gene Ontology (GO) terms and genes related to angiogenesis and cell migration, as well as ones relating to antigen processing and presentation (Fig. 2D, “Fibro 1”). In contrast, clusters 2 and 4 were enriched for known profibrotic pathways, including GO terms related to focal adhesion (9) and advanced glycation end product (AGE)/AGE receptor (RAGE) signaling (30) (Fig. 2D, “Fibro 2” and “Fibro 4”). Consistent with a scarring, myofibroblast-like identity, cluster 2 and 4 (2/4) fibroblasts also had relatively high expression of smooth muscle cell genes, such as ACTA2 and TAGLN, and related pathways. In our previous study of fibroblast heterogeneity in control and verteporfin-treated mouse wounds, we identified a fibroblast phenotypic trajectory by scRNA-seq characterized by Dpp4, Il6, and Ccl2 expression that was similarly enriched under scarring conditions, suggesting a broadly conserved “mechanofibrotic” phenotype (5). Overall, the transcriptomic profiles of these clusters, in concert with their relative enrichment in unwounded skin/verteporfin-treated wounds versus scarring control wounds, supported the definition of clusters 1 and 2/4 as putatively pro-regenerative and profibrotic fibroblast subpopulations, respectively. We also examined the gene expression profiles of fibroblast clusters 3 and 5. Overall, cluster 3 exhibited enrichment for numerous pathways related to ECM production and organization, as well as known wound healing–related growth factor pathways including phosphoinositide 3-kinase (PI3K)/protein kinase B (AKT) signaling (31, 32) (Fig. 2D, “Fibro 3”). These findings were consistent with fibroblasts actively engaged in producing wound ECM. Cluster 5 uniquely demonstrated enrichment for nerve- and neuron-related terms, possibly consistent with a peripheral nerve sheath fibroblast identity (Fig. 2D, “Fibro 5”).

To facilitate prospective isolation of the fibroblast subtypes identified by scRNA-seq analysis for further study, we sought to identify characteristic markers of each of the fibroblast subsets of interest. Reviewing the clusters’ transcriptomic profiles (Fig. 2E) with a lens toward surface markers with commercially available targeting antibodies, we homed in on three or four genes that were specifically enriched in each of fibroblast clusters. Fibroblast clusters 2/4 exhibited similar patterns of marker expression; given this fact, as well as their similarly profibrotic transcriptional profiles and enrichment in control wounds, we grouped these two clusters together for further analysis. Cluster 1, the putative “regenerative fibroblast cluster,” was marked by high expression of intercellular adhesion molecule 1 (ICAM1), a cell-cell adhesion molecule involved in the inflammatory response, leukocyte recruitment, and injury resolution (33, 34); moesin (MSN), involved in linking the cell membrane to the actin cytoskeleton (35); fatty acid binding protein 5 (FABP5), a lipid chaperone protein primarily expressed in epidermal cells and implicated in epidermal-fibroblast cross-talk (36–38); and interleukin-33 (IL33), a component of the T helper 2 (T_H2) immune response (39) with roles in tissue remodeling and regeneration (40, 41) (Fig. 2E, first row). Marker genes for fibroblast clusters 2 and 4, the putative “pro-fibrotic fibroblast clusters,” included genes with known roles in fibrosis and epithelial-mesenchymal transition (EMT), such as α-smooth muscle actin 2 (ACTA2) (42); oncostatin M receptor (OSMR), a member of the pro-inflammatory IL-6 receptor family (43–45); and sprouty homolog 2 (SPRY2) (46–48), as well as insulin-like growth factor binding protein 2 (IGFBP2), a regulator of insulin-like growth factor signaling with possible roles in cancer progression and autoimmune disease (49–51) (Fig. 2E, second row). Cluster 3 fibroblasts uniquely expressed the ECM constituent proteins periostin (POSTN), decorin (DCN), and lumican (LUM), as well as cell surface antigen fibroblast activation protein (FAP) (Fig. 2E, third row), all of which have known roles in wound healing and connective tissue repair, fibroblast activation, and collagen fibrillogenesis (52–56). Last, cluster 5 fibroblasts expressed multiple genes associated with nerve development and myelination, including cadherin-19 (CDH19) (57), epidermal growth factor–like protein 8 (EGFL8) (58), proteolipid protein 1 (PLP1; the primary component of myelin), and apolipoprotein A1 (APOA1) (59) (Fig. 2E, fourth row).

After identification of these characteristic cluster-defining genes at the transcriptional level, we sought to validate their expression patterns at the protein level in porcine wounds. To do so, we first performed IF staining for markers of clusters 1 and 2/4, the putative pro-regenerative and profibrotic fibroblast clusters. Consistent with our scRNA-seq data showing that cluster 1 fibroblasts were enriched in regenerating wounds, IF staining for ICAM1, MSN, FABP5, and IL33 revealed substantially stronger expression of all of these proteins, and more positive staining fibroblasts for each marker, in the dermis of verteporfin- compared with vehicle-treated wounds (fig. S5, A and B). Conversely, when we examined expression of IGFBP2, SPRY2, and OSMR, products of genes characteristic of putatively profibrotic fibroblast clusters 2/4, we found that their expression was greater in scarring vehicle compared with regenerating verteporfin wounds (fig. S5, C and D). Staining for markers of all fibroblast clusters (1, 2/4, 3, and 5) showed overlap with type 1 collagen (COL-I) staining (fig. S5, A, C, E, and F), supporting that the cells expressing these markers did have a fibroblast identity.

We next applied pseudotime analysis using Monocle 3, a computational package used to construct single-cell trajectories (60), to infer molecular dynamics among different fibroblast subpopulations on the basis of our scRNA-seq data from porcine wound and unwounded skin fibroblasts (Fig. 2F, top left panel). Ulis analysis suggested an overall branching transcriptional trajectory that originated with fibroblasts of cluster 1 (the putative pro-regenerative cluster) and progressed toward either cluster 2/4 (the putative profibrotic cluster) or cluster 3 or 5, with clusters 2/4 being the farthest along the pseudotime trajectory (Fig. 2F, bottom left panel). Pseudotime analysis also revealed increasing expression of known fibrosis-associated genes, including transgelin (TAGLN) (61) and C-C motif chemokine ligand 2 (CCL2) (62), as well as genes whose expression was associated with fibroblast clusters 2/4 such as CCL19 (Fig. 2D), along the trajectory from cluster 1 to clusters 2/4 (Fig. 2F, right panels, top row). This finding supported the pseudotime trajectory representing a transition from a putatively regenerative cluster 1–like to a putatively fibrotic cluster 2/4–like fibroblast fate. The trajectory from cluster 1 to cluster 3, on the other hand, was associated with increasing expression of ECM genes such as collagen type 1α1 chain (COL1A1) along with genes identified as markers of cluster 3, such as POSTN and DCN (Fig. 2F, right panels, middle row), consistent with fibroblast cluster 3’s postulated identity as an actively matrix-producing fibroblast phenotype.

In vitro mechanomodulation reveals mechanoresponsive fibroblast subpopulation dynamics

We next sought to functionally validate our pseudotime prediction of an activation trajectory leading from cluster 1 to cluster 2/4 fibroblasts (Fig. 2F) by determining whether cluster 1 fibroblasts could give rise to cluster 2/4 fibroblasts in vitro and, if so, what cues would induce this transition. On the basis of the characteristic surface markers we had identified for these key clusters (Fig. 2E), we used fluorescence-activated cell sorting (FACS) to isolate ICAM1⁺FABP5⁺MSN⁺ (cluster 1) and OSMR⁺SPRY2⁺IGFBP2⁺ (cluster 2/4) fibroblasts from unwounded red Duroc pig dorsal skin (fig. S6A). Fibroblasts were defined as lineage-negative for nonfibroblast markers (Lin⁻), per a previously published lineage depletion strategy (5). To directly interrogate the effects of substrate mechanics on the phenotype of fibroblasts from each of these clusters, we cultured them in 3D hydrogels, which were then subjected to either controlled mechanical stretching with fixed applied tension (“Stretch”), no tension (“Unstretched”), or tension with mechanotransduction inhibitor (“Verteporfin + Stretch”) (fig. S6B). We then examined expression of activated fibroblast and cluster-specific markers across experimental conditions using FACS, histologic, and molecular analyses.

Cluster 1 fibroblasts in the absence of stretch strongly expressed cluster 1 markers and minimally expressed cluster 2/4 markers (fig. S7, A and B, top row middle column, and C, white bars), whereas cluster 2/4 fibroblasts specifically expressed cluster 2/4 characteristic markers (fig. S7, A and B, bottom row middle column, and D, white bars). When subjected to stretch, cluster 1 cells decreased their expression of cluster 1 markers and instead increased expression of cluster 2/4 markers (fig. S7, A and B, top row left column, and C, purple bars), providing evidence that mechanical tension alone was sufficient to induce conversion of cluster 1 to cluster 2/4 fibroblasts. This conversion was completely blocked in the setting of YAP inhibition, given that cluster 1 fibroblasts subjected to stretch in the presence of verteporfin retained a cluster 1 surface identity and resembled cluster 1 fibroblasts that had not been exposed to stretching (fig. S7, A and B, top row right column, and C, black bars). In both cluster 1 and cluster 2/4 fibroblasts, stretching increased the proportion of cells expressing α-SMA and COL1A1, and secretion of collagen was greater in cluster 2/4 than in cluster 1 cells in both stretched and unstretched conditions (fig. S7, E and F, first and second columns, and G and H, purple versus white bars), consistent with scRNA-seq findings implicating cluster 2/4 as having a more typical scar fibroblast/myofibroblast identity. In contrast, verteporfin effectively abrogated the effects of stretch on a-SMA and COL1A1 expression (fig. S7, E and F, third column, and G and H, black bars).

Given that pseudotime analysis predicts cellular trajectories without directionality, it was also possible that cluster 2/4 fibroblasts may revert to a cluster 1 identity in a context-specific manner with verteporfin treatment. Stretching did not significantly affect the marker expression of FACS-isolated cluster 2/4 fibroblasts (fig. S7, A and B, bottom row left versus middle column, and D, purple versus white bars, P > 0.05). However, when exposed to verteporfin, a small fraction of cluster 2/4 cells began to express cluster 1 markers (fig. S7, A and B, bottom row right column, and D, black bars), suggesting that inhibiting mechanotransduction in cluster 2/4 fibroblasts could potentially revert them to a cluster 1 phenotype. In cluster 2/4 fibroblasts, verteporfin applied to stretched fibroblasts also decreased a-SMA and COL1A1 secretion to be even lower than that in cluster 2/4 fibroblasts that did not experience any stretch (fig. S7, F and H, white versus black bars).

Regeneration- and fibrosis-related cell-cell signaling is differentially enriched in verteporfin versus control wounds

To expand from our focused fibroblast analysis, we next performed CellChat analysis, which infers cell-cell interactions on the basis of known receptor-ligand interactions (63). Analysis of all cells from the scRNA-seq capture suggested a widespread network of intercellular interactions within the dataset (Fig. 2G). CellChat revealed that platelet and endothelial cell adhesion molecule (PECAM) signaling, a pathway well known to be involved in angiogenesis (64), was enriched in interactions between cluster 1 fibroblasts and other cell types, including endothelial cells and pericytes, whereas the other fibroblast clusters exhibited negligible PECAM cell-cell signaling (Fig. 2H, top row, left panel). Collagen signaling was strongest among cell-cell interactions involving fibroblast cluster 3 (Fig. 2H, top row, middle panel), which we had previously implicated as an actively ECM-producing fibroblast subpopulation. The known pro-inflammatory, profibrotic IL-6 signaling network (65) was specifically enriched in cell-cell interactions between cluster 2 fibroblasts and multiple immune cell types, including dendritic cells (DCs), monocytes, and macrophages (Fig. 2H, top row, right panel). Transforming growth factor–β (TGF-b) signaling, also well known to drive fibrosis (66), was also strongly enriched in cell-cell interactions involving cluster 2 fibroblasts (Fig. 2H, bottom row, right panel). Cluster 4 fibroblasts exhibited active cell-cell cross-talk along the bone morphogenetic protein (BMP) signaling network (Fig. 2H, bottom row, middle panel), which has also been implicated in organ fibrosis (67). Finally, cluster 5 fibroblasts (which exhibited a transcriptomic profile consistent with a peripheral nerve–associated fibroblast phenotype) were implicated in laminin signaling, a pathway associated with peripheral nervous system development and Schwann cells (68) (Fig. 2H, bottom row, left panel).

We next used CellChat to examine shifts in cell-cell signaling over time in vehicle- versus verteporfin-treated wounds. Overall, cells from vehicle-treated wounds at early time points were enriched for BMP signaling, a pathway enriched in cluster 4 fibroblasts (Fig. 2H) and implicated in fibrosis (67), and others such as IL1, IL6, CXC motif chemokine ligand (CXCL), and CCL intercellular signaling, all known to be involved with pro-inflammatory signaling in the early stages of scarring particularly for recruiting and activating wound inflammatory cells such as neutrophils and macrophages (69, 70) (Fig. 2I). By 16 weeks, control wounds exhibited greater enrichment for pathways such as desmosome signaling, a process related to cell-cell junctions, which play a key role in mechanotransduction (71), and continued to exhibit strong active cell-cell signaling along pathways known to positively regulate granulocyte/macrophage populations, such as colony-stimulating factor (CSF) (72) (Fig. 2I). In contrast, verteporfin-treated wounds at 2 weeks exhibited active cell-cell signaling along the IL4 axis, an important component of the type 2 immune response and promoter of T_H2 differentiation (73), and the hedgehog (HH) pathway, previously implicated in wound HF neogenesis/regeneration (74, 75) (Fig. 2I). By 16 weeks, verteporfin-treated wounds were most enriched for tenascin signaling, which is essential for wound angiogenesis (76) (Fig. 2H). Verteporfin-treated wounds at 16 weeks also had active IL6 cell-cell signaling (Fig. 2I), which was also seen in early vehicle-treated wounds, highlighting that certain pathways may have distinct roles in wound repair and fibrosis versus regeneration depending on when during the healing process they are activated.

Immune cell transcriptomic analysis reveals distinct inflammatory dynamics in fibrotic and regenerating wounds

Given the intriguing implications of cell-cell cross-talk involving immune cells and pathways, we returned to our scRNA-seq data to home in on the impact of verteporfin on the wound immune milieu. Immune cells were subclustered from total wound samples on the basis of their expression of protein tyrosine phosphatase receptor type C (PTPRC; or CD45) and then further subdivided into myeloid and lymphoid subsets (Fig. 3A, fig. S8A, and table S1). Focusing on myeloid subsets, we performed a kinetic analysis over the course of treatment to discern the compositional dynamics associated with regenerating verteporfin-treated versus fibrotic vehicle-treated wounds. Vehicle-treated fibrotic wounds had a notable enrichment of monocytes and fibrocytes, whereas wound healing macrophages positive for CD9, secreted phosphoprotein 1 (SPP1), and arginase 1 (ARG1) were expanded in verteporfin-treated wounds at the 4-week time point (Fig. 3A). Additionally, wounds treated with verteporfin had a large increase in migratory DCs positive for C-C chemokine receptor type 7 (CCR7), monocyte DCs (moDCs) positive for integrin subunit aM (ITGAM) and CD14, and classical DCs (cDCs) positive for X-C motif chemokine receptor 1 (XCR1) and zinc finger and BTB domain–containing 46 (ZBTB46), at 4 weeks postwounding (Fig. 3, A and B). By contrast, proportions of classical tissue macrophages positive for complement component 1q (C1Q), lymphatic vessel endothelial hyaluronan receptor 1 (LYVE1), and C-X3-C motif chemokine receptor 1 (CX3CR1), and inflammatory moDCs positive for tumor necrosis factor (TNF), did not substantially vary between treatment and control groups. Consistent with the activated DC phenotype in verteporfin-treated wounds, a number of effector lymphocyte subsets, which take their cues from DCs, were enriched in regenerating compared with control wounds. Specifically, verteporfin treatment induced an influx of γδ T cells, innate lymphoid cells 2 (ILC2s), natural killer cells, regulatory T cells (T_regs), and type 17 T cells compared with vehicle-treated wounds at 2 and 4 weeks postwounding (fig. S8, A and B).

Fig. 3. — (A) UMAP plot of scRNA-seq data for all myeloid cells; black dotted region indicates cells transcriptionally classified as dendritic cells (DCs) and macrophages. Quantification of cell number and cell proportion (%) also shown for each sample. UW, unwounded skin; P, vehicle (PBS)–treated wounds; V, verteporfin-treated wounds. (B) Relative proportions of myeloid cell subsets across different treatments and time. (C) Kyoto Encyclopedia of Genes and Genomes (KEGG), GO, and Mouse Genome Informatics-Mammalian Phenotype (MGI-MP) pathway analyses for wound healing macrophages (left panels) and fibrocytes (right panels) at 4 weeks postwounding. (D and E) Pseudotime analysis with the starting node set in the monocyte cluster (red circle), four predicted branch nodes (black circles), and three end nodes in the fibrocyte and two macrophage clusters (gray circles). (D) shows the different cell subset partitions as different colors, and (E) shows the color scale based on pseudotime. (F) Left panels: Representative IF histology showing expression of either ST2 (top row, green) or IL-4 (bottom row, green) in unwounded skin (left column) or wounds treated with vehicle control (middle column) or verteporfin (2 mg/ml) (right column) at 4 weeks, with DAPI nuclear counterstain (blue). Scale bar, 100 μm. Right panels: Quantification of IF staining shown in left panels. ST2: UW versus vehicle P = 0.017, UW versus verteporfin P = 0.0049, and vehicle versus verteporfin P = 0.033. IL-4: UW versus verteporfin P = 0.018, vehicle versus verteporfin P = 0.035. Quantification from n = 3 randomly selected specimens for each time point/condition and P values calculated by Student’s t test following ANOVA with Bonferroni correction.

Macrophages are an indispensable feature of wound repair. Given the differences in the composition of distinct subsets between our treatment and control groups, we next sought to discern the functionality of these cells from their gene expression. Toward this end, we performed pathway analysis on differentially expressed genes [absolute average log₂ fold change abs(avg_log₂FC) > 1] from wound healing macrophages and fibrocytes at the 4-week time point, when we had observed the most marked differences in populations between treatment groups. Wound healing macrophages, which were increased in verteporfin wounds, had enriched expression of genes involved in immune pathways, including neutrophil activation, degranulation, chemotaxis, and phagosome activation, whereas fibrocytes, which were increased in control wounds, expressed genes involved in ECM and collagen organization (Fig. 3C). This marked dichotomy between verteporfin- and vehicle-treated wounds raised the possibility that monocytes may adopt distinct identities associated with different wound healing outcomes. Given the challenges of working with porcine models, particularly the lack of lineage tracing strategies available, we tested this hypothesis computationally. Again, using Monocle 3 (60), we generated a pseudotime trajectory of myeloid lineage cells starting with the monocyte cluster as the starting node. This pseudotime analysis predicted a major branching point from the monocyte cluster, resulting in two opposing fates as end nodes: fibrocytes and macrophages (Fig. 3, D and E). Thus, verteporfin treatment may skew infiltrating monocytes away from a profibrotic fibrocyte state to instead acquire repair macrophage phenotypes and facilitate regeneration.

To bolster our scRNA-seq findings regarding immune infiltrate differences between verteporfin- and vehicle-treated wounds, we also performed immunohistochemical staining for suppression of tumorigenicity 2 (ST2; the IL-33 receptor) and IL-4, two common markers of a T_H2 immune response, in wounds at 4 weeks. Expression of both ST2 and IL-4 was greater in verteporfin-treated wounds (Fig. 3F), supporting our scRNA-seq analyses suggesting that the immune infiltrate in YAP-inhibited wounds skewed toward a T_H2 inflammatory profile (Fig. 2I).

Spatial proteomic analysis supports divergent fibroblast dynamics in scarring and regenerating wounds

Although scRNA-seq is informative regarding wound cell heterogeneity and transcriptomic phenotypes, a shortcoming of single-cell sequencing is the loss of spatial information. To phenotype verteporfin- and vehicle-treated wounds in a spatially informed fashion, we analyzed 16-week porcine wounds and skin using CO-Detection by indexing (CODEX; now known as PhenoCycler). In this assay, a large panel of individual protein markers is sequentially labeled with, and iteratively imaged between, cyclic additions and washouts of dye-labeled oligonucleotide-conjugated antibodies (77); protein staining patterns can then be used to project a manifold of cell-representative clusters (fig. S9A; markers shown in table S2).

Across unwounded skin and verteporfin- and vehicle-treated wounds at 16 weeks, 11 cell clusters were identified by unique CODEX signatures, including multiple fibroblast, epithelial, and immune subclusters, and adipocyte, smooth muscle, endothelial, and vascular mural cell clusters (fig. S9A, middle and right panels, and table S3). After annotation of these cell subsets based on protein expression patterns, we compared the representation of CODEX-defined cell types within verteporfin- versus vehicle-treated wounds at 16 weeks. Overall, compared with unwounded skin and 16-week verteporfin-treated wounds, 16-week vehicle-treated wounds had grossly apparent differences in spatial proteomic signatures (fig. S9B). Independently of our previous fibroblast subclustering based on scRNA-seq, CODEX analysis identified four fibroblast clusters on the basis of protein expression (termed CODEX type 1 to 4 fibroblasts) in 16-week wounded and unwounded samples. The relative proportions of these CODEX-defined fibroblast clusters differed on the basis of wound treatment. There were fewer CODEX type 2 fibroblasts, which were positive for COL-I, fibronectin, CD26, piezo-type mechanosensitive ion channels 1 and 2 (PIEZO1/2), and SPRY2, and a trend toward more CODEX type 4 fibroblasts, which were positive for COL-I and S100 calcium binding protein A4 (S100A4), in verteporfin-treated wounds compared with control wounds at 16 weeks (fig. S9, C and D).

We next sought to assess the correlation between fibroblast subtypes identified in the scRNA-seq and CODEX datasets. On the basis of our CODEX data, we interrogated the spatial protein expression of COL1⁺MSN⁺FABP5⁺ICAM1⁺ cells, corresponding to scRNA-seq cluster 1 fibroblasts, and COL1⁺SPRY2⁺IGFBP2⁺OSMR⁺ cells, corresponding to scRNA-seq cluster 4 fibroblasts (fig. S9E, left panels). The CODEX type 2 fibroblast cluster mapped closely to the putatively profibrotic scRNA-seq fibroblast cluster 4, whereas the CODEX type 4 fibroblast cluster mapped closely to the putatively pro-regenerative scRNA-seq fibroblast cluster 1 (fig. S9E, middle and right panels), consistent with the relative enrichment of CODEX type 2 and type 4 fibroblasts in vehicle- versus verteporfin-treated wounds, respectively.

Given the insights provided by our CODEX analysis of 16-week wounds, we next sought to apply spatial proteomic analysis to analyze dynamic changes in cell populations and spatial interactions over the course of healing. We used the same CODEX profile to generate spatial proteomic profiles of wounds at 4 and 8 weeks of healing and compared these with CODEX data from unwounded skin and 16-week wounds. Protein staining patterns were again used to project a manifold of cell-representative clusters from each time point (Fig. 4A). CODEX revealed a complex cellular milieu of key wound cell types at each time point, including multiple myeloid, fibroblast, epithelial, and endothelial and vascular mural cell types. These cell types varied over time, but key phenotypes were conserved across time points. For example, fibroblast subpopulations expressing markers of the fibroblast cluster 4 previously identified by scRNA-seq (OSMR⁺SPRY⁺; see Fig. 2E) were identified as “fibroblasts 5” and “fibroblasts 3” in 4- and 8-week CODEX samples, respectively. Another conserved fibroblast phenotype expressed markers consistent with mechanical activation, such as PIEZO1/2 and CD26 (78); these were found as “fibroblasts 1,” “fibroblasts 6,” and “fibroblasts 2” in 4-, 8-, and 16-week CODEX samples, respectively. A blandly ECM gene-expressing fibroblast subpopulation that was positive for COL-I and platelet-derived growth factor receptor α (PDGFRa) was found as “fibroblasts 3” and “fibroblasts 4” in 4- and 8-week CODEX samples, respectively. Regarding myeloid cells, 4- and 8-week samples each contained subpopulations with a fibrocyte-like marker profile, which were positive for CD45, CD54, CD68, PDGFRa, and collagen type IV (COL-IV) (79), corresponding to “myeloid cells 1” in week 4 CODEX samples and “myeloid cells 2” in week 8 samples. When we compared estimates of relative abundance of these cell populations between wound conditions at 4 weeks, myeloid cells 1 (the fibrocyte-like population) were markedly (though not at the level of statistical significance; P = 0.195) enriched in vehicle- compared to verteporfin-treated wounds, whereas myeloid cells 2 were significantly (P = 0.011) more prevalent in verteporfin-compared to vehicle-treated wounds (Fig. 4B). These results suggested the possibility of distinct myeloid cell distributions in the context of verteporfin versus control treatment, with early vehicle-treated wounds possibly biased toward greater numbers of the fibrocyte-like myeloid cells 1 subtype versus verteporfin-treated wounds skewing toward the myeloid cells 2 subtype.

Fig. 4. — (A) UMAP plot of CODEX data for all sequenced cells from verteporfin- and vehicle-treated wounds at 4 (left panel), 8 (middle panel), and 16 weeks (right panel). (B) Bar graphs quantifying myeloid cells 1 (top panel) and myeloid cells 2 (bottom panel) as proportions of all cells based on CODEX analysis of unwounded skin (UW) and vehicle- and verteporfin-treated wounds (P and V, respectively) at 4 weeks. (C) Interaction maps showing differential interactions between indicated pairs of conditions at indicated time points. (D) Bar graphs quantifying indicated cell-cell interaction scores at 4 (top row), 8 (middle row), and 16 weeks (bottom row). (E) Top row: Representative images of indicated selected CODEX markers in vehicle- and verteporfin-treated wounds at 4 weeks (left and middle sets of panels, respectively) and unwounded skin (right set of panels). Top right panel of each set is a magnified view corresponding to inset region in white box, left panel. Bottom row: Representative images of identified UMAP CODEX clusters [see (A), left panel] showing their spatial distribution in vehicle- and verteporfin-treated wounds at 4 weeks and unwounded skin. Error bars in (B) and (D) represent variation (SD) among three randomly selected wounds per experimental condition/time point.

To understand how cell-cell communication dynamics at the protein level varied between wound conditions, we compared CODEX-defined cell-cell interaction networks, which are based on spatial colocalization, among unwounded skin, control wounds, and verteporfin-treated wounds at each time point (Fig. 4, C and D, and fig. S10). Vehicle-treated wounds at 4 weeks compared with unwounded skin exhibited an enriched interaction network involving TAGLN/Desmin⁺ smooth muscle cells and multiple fibroblast subtypes, especially fibroblast types 1 to 3 (Fig. 4C, top row, middle column). Compared with verteporfin-treated wounds at the same time point, 4-week control wounds had enriched interactions between fibrocyte-like myeloid cells 1 and multiple fibroblast subpopulations, including fibroblasts 2, which express the mechanical activation markers PIEZO1/2 and CD26, and fibroblasts 5, which are OSMR/SPRY2/IGFBP2⁺ and resemble the putatively pro-fibrotic scRNA-seq fibroblast cluster 4 (Fig. 4C, top row, left column). Verteporfin-treated wounds at 4 weeks compared with unwounded skin had an enriched interaction network involving endothelial cells, pericytes, and α-SMA–expressing fibroblasts 4, as well as generally increased interactions involving myeloid cells 2 (Fig. 4C, top row, right column). Specifically, interactions of CODEX-defined myeloid cells 1 with both fibroblasts 2 and smooth muscle cells were enriched in vehicle- compared with verteporfin-treated wounds, whereas interactions between myeloid cells 2 and fibroblasts 4 were stronger in verteporfin-treated wounds (Fig. 4D, top row, and fig. S10A, left panel). Overall, interaction network analysis supported that in addition to their differential abundance, myeloid cells 1 (the putatively fibrocyte-like cells) and myeloid cells 2 may also exhibit differentially enriched activity in early vehicle- and verteporfin-treated wounds, respectively. These results were consistent with our scRNA-seq findings of increased fibrocytes in control wounds (Fig. 3). Analysis of the later 8- and 16-week time points also revealed increased activity of interaction networks involving epithelial and vascular/endothelial cells in verteporfin-treated compared with control wounds (Fig. 4D, middle and bottom rows, and fig. S10, B and C, left panels), consistent with previous CODEX cell abundance and scRNA-seq findings revealing increased markers of angiogenesis in verteporfin-treated wounds.

Finally, we visually analyzed staining and cell segmentation patterns in unwounded skin and wounds at 4 weeks, the time point at which our previous CODEX analysis had revealed some of the most interesting and dynamic differences between wound conditions (Fig. 4E). In vehicle-treated wounds, substantially greater staining was observed for the EPF/mechanically activated fibroblast marker CD26 in cells throughout the dermis, and numerous cells of the fibrocyte-like myeloid cells 1 identity were observed in close proximity with fibroblasts (Fig. 4E, left set of panels). This finding was consistent with the enriched cell-cell interactions revealed by our network analysis (Fig. 4, C and D). Overall, time course CODEX analysis supported nuanced myeloid and fibroblast differences between vehicle- and verteporfin-treated wounds.

YAP inhibition promotes wound regeneration in a human foreskin xenograft model

Collectively, the results of our red Duroc pig study supported that YAP inhibition altered wound cell dynamics to induce more regenerative healing compared with control wounds. We were encouraged by the phenotype of reduced scarring in verteporfin-treated porcine wounds (Fig. 1). We previously developed a human skin xenograft model to examine healing by human cells in an experimentally tractable system (25). In this model, human foreskin tissue samples are grafted onto CD-1 nude recipient mice (selected to reduce graft rejection). Then, after an engraftment period, full-thickness wounds are created within the graft. We have previously reported that the engraftment process alone does not cause a fibrotic reaction within the foreskin tissue, and that healing of xenograft wounds is accomplished by human cells within the graft rather than mouse cells entering from the surrounding tissue, as indicated by the wound healing with human rather than mouse collagen (25).

We sought to apply this model to determine the effects of local verteporfin treatment on human skin wound healing. We first adapted our xenograft model to mimic our porcine wounding model as closely as possible. Xenograft wounds were injected with either vehicle (control) or verteporfin (20 μl of 1 mg/ml solution) immediately after wounding and underwent primary closure with sutures to create a linearly closed wound. Wounds were harvested 14 days after wounding for histologic analysis (fig. S11A). On the basis of H&E histology, healed vehicle-treated wounds had a densely fibrotic appearance with linearly aligned ECM, whereas verteporfin-treated wounds were minimally fibrotic and instead had normal-appearing architecture (fig. S11B). Consistent with this histologic impression, verteporfin-treated wounds resulted in less apparent linear scars compared with control treatment (fig. S11C). Masson’s trichrome histology further supported that verteporfin-treated wounds were less densely collagenous than control wounds (fig. S11, D and E), consistent with decreased scarring. Further supporting that verteporfin reduced fibrosis and promoted regeneration in these human skin xenograft wounds, picrosirius red staining and ultrastructural analysis revealed that verteporfin-treated wounds had ECM properties that closely overlapped with those of unwounded xenograft skin, whereas vehicle-treated wounds had a nonoverlapping ECM profile consistent with distinct scar architecture (fig. S11, F and G).

Although this xenograft model with primary closure yielded results consistent with our porcine wounding experiments showing reduced scarring with verteporfin, the narrow linear wounds resulting from the primary closure model substantially limited the quantity of healed wound tissue available for analysis. Thus, we sought to validate these results in our typical xenograft wound model without primary closure (25), where wounds are left open to heal by secondary intention, which results in a larger region of healed wound/scar tissue and could thus allow us to more robustly interrogate the effects of verteporfin on human xenograft healing. Established xenografts were wounded, and wounds were treated with intradermal injection of either vehicle or verteporfin; these xenografts were then harvested 14 days after wounding, at which time all wounds were fully re-epithelialized, for histologic analysis (Fig. 5A).

Verteporfin-treated wounds resulted in grossly less-apparent scars compared to control treatment (Fig. 5B, “Unwounded” versus “Vehicle” versus “Verteporfin”). On H&E and Masson’s trichrome histology, healed verteporfin-treated wounds again appeared substantially less fibrotic than vehicle-treated wounds and instead resembled unwounded xenograft skin (Fig. 5, C to E, “Unwounded” versus “Vehicle” versus “Verteporfin”). Quantification of Masson’s trichrome staining confirmed that verteporfin-treated wounds had lower overall collagen density than control wounds and were instead comparable to unwounded xenograft skin (Fig. 5F, “Unwounded” versus “Vehicle” versus “Verteporfin”), consistent with decreased scarring. Further supporting these histologic data, picrosirius red staining showed that vehicle-treated wounds had a dense, linearly aligned collagen structure consistent with scarring, whereas verteporfin-treated wounds had a looser and more randomly aligned “basketweave” structure of collagen that resembled that of unwounded skin (Fig. 5G, “Unwounded” versus “Vehicle” versus “Verteporfin”). Quantitative matrix ultrastructure results were consistent with these visual impressions: The ECM features of verteporfin-treated wounds overlapped with those of unwounded xenograft, whereas vehicle-treated wounds had a distinct, nonoverlapping pattern of matrix architecture (Fig. 5H, “Unwounded” versus “Vehicle” versus “Verteporfin”).

On the basis of histology, the boundaries between surrounding mouse skin, human xenograft skin, and wounded regions within xenografts were readily morphologically appreciable (Fig. 5, D and E). Thie demarcation between mouse and human tissue was confirmed by immunostaining for mouse- and human-specific collagen, which revealed boundaries between mouse and human skin (fig. S12A). Further, by analyzing gross photographs of xenografts over the course of our wounding experiments, we confirmed that contraction of xenografts after the initial engraftment period was minimal overall and did not significantly differ between any wound conditions (fig. S12B). We were therefore confident that we were able to focus ensuing analyses appropriately and specifically on wound tissue within xenografts without erroneously considering, for instance, mouse tissue at the xenograft boundary or having results confounded by differential contraction of xenografts in different experimental conditions.

We next examined expression of known markers of fibrosis and regeneration across xenograft wound conditions. Vehicle-treated wounds contained significantly more COL-I⁺ cells than did verteporfin-treated wounds (P = 0.0033; fig. S12, C and D, “Unwounded” versus “Vehicle” versus “Verteporfin”). Verteporfin-treated wounds had significantly greater staining for CD31, indicating increased wound neovascularization (P = 0.028; fig. S12, E and F, “Unwounded” versus “Vehicle” versus “Verteporfin”). Finally, as was done in pigs (fig. S4C), we injected verteporfin into unwounded xenograft skin to evaluate its dermal toxicity profile. On basic histology, verteporfin-injected skin appeared morphologically comparable to noninjected unwounded xenograft skin (fig. S13), consistent with a favorable dermal safety profile for local administration. Collectively, our xenograft findings supported that verteporfin treatment promoted wound regeneration with substantially reduced fibrosis/scarring in a model of wound healing mediated by human cells.

IL-33 inhibition validates its functional role in healing of verteporfin-treated wounds

Our single-cell transcriptomic and immunohistochemical findings from porcine wounds previously revealed that IL-33 was a key marker of a putatively pro-regenerative fibroblast cluster (scRNA-seq fibroblast cluster 1; Fig. 2E and fig. S5A). In addition, IL-33 has previously been implicated in tissue repair (40, 41, 80) and is known to be involved in the T_H2 immune response (39), components of which we also observed to be enriched in early verteporfin-treated wounds (Fig. 2H). IF staining for IL-33 in xenograft specimens confirmed that IL-33 expression was significantly increased in verteporfin-treated wounds compared with either vehicle-treated wounds or unwounded skin, whereas YAP expression was significantly decreased compared with vehicle-treated wounds (all P ≤ 0.05; fig. S12, G and H, “Unwounded” versus “Vehicle” versus “Verteporfin”). In addition, staining for FABP5, a marker of the putatively pro-regenerative porcine scRNA-seq fibroblast cluster 1, largely colocalized with IL-33 staining in verteporfin-treated wounds (fig. S12G, “Verteporfin”), suggesting a link between these mechanically naive fibroblasts and IL-33 signaling. Thus, we hypothesized that IL-33 could play a role in verteporfin-induced xenograft wound regeneration.

To functionally test this hypothesis, we administered an IL-33 blocking antibody (IL-33i) to vehicle- and verteporfin-treated wounds (Fig. 5A). Wounds were harvested for histologic analysis either 14 days after wounding or on the day of complete wound closure/re-epithelialization, whichever came later. As noted previously, all wounds treated with only verteporfin or only vehicle were fully healed by 14 days. We now observed that although all wounds treated with both vehicle and IL-33i were also fully healed by 14 days, wounds treated with verteporfin and IL-33i frequently exhibited delayed healing or nonhealing. Specifically, of seven total xenograft wounds treated with both verteporfin and IL-33i, two failed to heal by 14 days and eventually developed exuberant infection, necessitating euthanasia of these mice and precluding analysis of healed wound tissue. Thie remaining five wounds eventually healed but did so more slowly than other wound conditions; these were harvested at the time of wound closure, which occurred at, on average, 16 days postwounding.

IF staining revealed that IL-33 expression was abrogated in wounds in the context of IL-33i treatment (fig. S12, G and H). Consistent with the indications of impaired healing in wounds treated with both verteporfin and IL-33i, grossly, these wounds appeared erythematous and poorly epithelialized relative to other wound conditions including verteporfin treatment alone (Fig. 5B). Histologically, they yielded abnormal tissue that appeared minimally healed (Fig. 5, D and E, bottom rows). In contrast, wounds treated with vehicle and IL-33i healed on a normal timeline and with typically fibrotic-appearing scars that resembled vehicle-treated control scars both grossly and histologically (Fig. 5, B to H). On the basis of ECM content, wounds treated with verteporfin and IL-33i were largely scar-like, with collagen content comparable to that of vehicle- or vehicle and IL-33i–treated wounds (Fig. 5, E and F, and fig. S12, C and D) and comparatively poorer neovascularization versus verteporfin-treated wounds (fig. S12, E and F). Picrosirius red staining patterns also appeared similarly scar-like across these three wound conditions (Fig. 5G), and ECM ultrastructure analysis revealed that wounds treated with verteporfin and IL-33i had ECM that overlapped most strongly with vehicle-treated scarring wounds (Fig. 5H). Overall, the results of our IL-33i experiments suggested that IL-33 is necessary for regenerative healing in the context of YAP inhibition.

Human healing after verteporfin treatment parallels cellular responses in regenerating porcine wounds

Finally, we applied scRNA-seq analysis to vehicle- and verteporfin-treated xenograft wounds (Fig. 6A). After alignment against both mouse and human genome, 3217 cells were captured for analysis. To confirm that the cells analyzed were of human origin, we compared mouse and human mitochondrial counts; only human cells, defined by high human mitochondrial counts and low mouse mitochondrial counts, were analyzed (Fig. 6B). Five transcriptionally defined clusters of human fibroblasts were identified from all cells captured (Fig. 6C). These clusters were present in differential distributions depending on wound treatment (Fig. 6D), and each had distinct transcriptomic signatures (Fig. 6E). Verteporfin-treated wounds and unwounded skin showed similar proportions of these human fibroblast subsets (clusters I to V) at day 14 postwounding, whereas relative to verteporfin-treated wounds or unwounded skin, vehicle-treated wounds showed an expansion of clusters III, IV, and V (Fig. 6F). FABP5, a marker of the putatively pro-regenerative porcine scRNA-seq fibroblast cluster 1, was highly expressed in human fibroblast cluster I (Fig. 6G, top panel); these cells, which were positive for DCN and LUM and secreted frizzled related protein 2 (SFRP2), were enriched in unwounded xenograft skin and verteporfin-treated xenograft wounds (Fig. 6F). Similarly, IGFBP2, a marker of the putatively profibrotic porcine scRNA-seq clusters 2/4, was highly expressed in human fibroblast clusters III [which was positive for CCL19, tissue inhibitor of metalloproteinase 1 (TIMP1), and complement component 3 (C3)], IV [positive for melanoma cell adhesion molecule (MCAM), TAGLN, and COL4A1], and V [positive for marker of proliferation Ki-67 (MKI67) and DNA topoisomerase IIα (TOP2A)] (Fig. 6G, bottom panel), all of which were increased in vehicle-treated xenograft wounds (Fig. 6F).

To evaluate whether transcriptionally analogous fibroblast clusters existed in porcine and human skin and wounds, label transfer analysis was performed between the porcine and human xenograft scRNA-seq datasets using cross-species orthologs. Thie regenerative porcine fibroblast cluster 1 showed a similar gene expression profile to human cluster II, which was positive for SRY-box transcription factor 4 (SOX4), whereas scarring porcine cluster 4 mapped onto human cluster IV (Fig. 6H). Supporting these findings, GO pathway analysis revealed that human cluster II was associated with potentially pro-regenerative terms relating to angiogenesis and protein synthesis and turnover, whereas human clusters III and IV were associated with terms related to focal adhesion assembly, smooth muscle contraction, and activator protein-1 (AP-1) signaling, suggesting a more fibrotic, mechanically activated gene expression profile (Fig. 6I). Last, RNA velocity analysis suggested an activation trajectory leading from human fibroblast cluster II to clusters III/IV (Fig. 6J), resembling the analogous trajectory from fibroblast cluster 1 to clusters 2/4 observed in porcine wounds (Fig. 2F). IGFBP2, a marker of fibrotic porcine clusters 2/4, was also found to be highly expressed in human clusters III/IV by scVelo (Fig. 6J).

DISCUSSION

We previously reported that YAP inhibition using one-time treatment with the small-molecule verteporfin was sufficient to diminish scarring and promote wound regeneration in mice (1). Here, we sought to validate our mouse findings in a large animal preclinical model of wound healing in red Duroc pigs, which form scars similar to those observed in humans. Our findings in this paper support our previous work that inhibiting mechanotransduction at the time of wounding is sufficient to drive wounds toward regeneration (1).

The existence of analogous fibroblast clusters identified in our scRNA-seq analysis was supported by spatial proteomic analysis. Although interpreting the results from this approach is challenging because of the relative lack of previous porcine scRNA-seq and CODEX studies, we were encouraged to find pathway enrichment results in pigs consistent with those we previously identified in mice, suggesting that our methodology was successful in detecting conserved transcriptomic differences between regenerative and fibrotic porcine wound repair.

It is intriguing that a single intervention—local verteporfin administration at the time of wounding—was sufficient to induce persistent cellular and gross phenotypic differences as far out as 16 weeks postwounding, particularly given our data showing that (at least systemically) the drug did not persist in animals beyond 2 days postwounding. These prolonged effects of a single verteporfin dose are promising for clinical translation but raise several mechanistic questions meriting future study. First, fibroblasts from both the proregenerative transcriptomic cluster 1 and profibrotic cluster 2/4 are found in unwounded skin as well as wounds, suggesting a lack of truly “regeneration-specific” cellular subpopulations. We hypothesize that the marked differences in the proportions and interactions of these subpopulations in vehicle- versus verteporfin-treated wounds highlight their different regenerative and scarring roles in skin repair. We and others have previously documented that fibroblasts in wound repair and other fibrotic processes, including specifically in pigs, do not necessarily acquire entirely distinct fibroblast states as viewed through scRNA-seq but rather that conserved fibroblast subpopulations expand and contract in response to external stimuli and disease/fibrotic states (5, 81–83).

Activity of these cell subpopulations can also be dynamic during fibrosis beyond their prevalence alone, such as through altered interactions with other cell populations. Our in vitro, time course scRNA-seq and CODEX data support that one-time verteporfin treatment may initiate a cascade of molecular and cellular events. The earliest of these events is fibroblast fate divergence, with enrichment of the fibroblast 1 phenotype that is IL-33–expressing and relatively quiescent in contrast with the fibro 2/4 activated phenotype enriched in scars that is characterized by actomyosin contractility and inflammatory cytokine expression. This divergence within 2 weeks of wounding is perpetuated by distinct cellular dynamics and crosstalk between multiple cell populations as wound healing progresses, ultimately leading to lasting, marked phenotypic differences. Myeloid cells are particularly key, with fibrocyte cross-talk predominating in scars in contrast with tissue macrophages/moDCs/cDCs in regenerating wounds. Thus, a key component of the pro-regenerative niche after verteporfin treatment is likely an altered inflammatory cell profile linked to blockage of early fibroblast fate changes.

Another interesting question raised is the “chicken and egg” relationship between fibroblasts and fibrocytes in scarring wounds: Do myeloid precursors adopt a fibrocyte-like phenotype in response to signals from cluster 2/4 fibroblasts, or are fibrocytes necessary to promote/sustain the cluster 2/4 fibroblast phenotype or suppress cluster 1 fibroblasts? Exploring these questions will require new ways to experimentally target fibrocytes in pigs, given that these represent a relatively new cell population without existing tools for functional manipulation or depletion.

Our findings regarding a functional role for IL-33 in regenerative wound repair shed further light on possible mechanisms of verteporfin-induced wound regeneration. Our scRNA-seq dataset suggested a trend toward a T_H2-like immune infiltrate in verteporfin-treated wounds, including CD9⁺SPP1⁺ARG1⁺ wound healing macrophages, CCR7⁺ migratory DCs, and IL-4 signaling pathway enrichment by CellChat analysis; this stood in contrast to the fibrocyte-dominant inflammatory profile of control wounds, which featured CD45⁺CD34⁺COL1A1⁺ cells and ECM-related pathway enrichment. Given that T_H2 chemokines like IL-33 promote recruitment of alternatively polarized macrophages (84), it is possible that the IL-33–expressing fibroblast cluster 1 we identified on scRNA-seq of porcine wounds could be an important effector population for the divergent inflammatory outcomes of verteporfin versus scarring wounds. Functionally, the severely impaired healing that we observed when IL-33 was blocked in verteporfin-treated wounds (versus the relatively “normal” scarring healing when IL-33 was blocked in vehicle-treated wounds) suggested that IL-33 signaling may represent an alternate healing pathway that is redundant under typical mechanically activated, scarring wound conditions but becomes essential when mechanotransduction is blocked. This role is consistent with IL-33’s known function as an alarmin associated with tissue damage (80) and adds to a growing consensus on a YAP/IL-33 immunologic axis that affects cell fate, function, and behavior within the context of wound healing. A separate study in mice suggested an inverse relationship between YAP activation and IL-33 expression (85), which we now experimentally verify in pig and human xenograft models. However, to date, there has been conflicting evidence on the role of IL-33 in wound repair and regeneration. For example, Il33 knockout mice lacking nuclear IL-33 showed delayed wound healing with up-regulated Ill/Il6 and CXCL cytokines, and SCA-1–positive fibroblasts enriched in Il33 expression were recently shown to be beneficial in the early inflammatory response to wounding (86, 87). On the other hand, Il33 knockout mice showed enhanced resistance to ischemic ulcer formation (88). Although our results are consistent with IL-33 being necessary for wound regeneration in the acute setting, chronic activation of T_H2 immunity and IL-33 signaling is also associated with fibrosis (89, 90), highlighting that particular inflammatory populations may have different effects on tissue repair and fibrosis depending on the particular time course of their activation.

Large animal models represent a critical stage in bridging the translational gap between small animal experimentation and human clinical implementation. Because of differences in skin and wound biology between mice and humans (91), for any promising mouse findings a vital step in “derisking” human clinical trials is validating those findings in a relevant large animal model. Compared to human tissues, porcine tissues have been shown to exhibit similar responses to substances such as growth factors (92), supporting the use of pigs to model human responses to drugs and other interventions (93). Red Duroc pigs specifically are the gold standard animal model for preclinical wound healing studies, because they exhibit scarring similar to human hypertrophic scarring (83, 94), a common form of pathological fibrosis estimated by some studies to affect more than 90% of postsurgical and burn scars (95). We are thus encouraged by our results herein. This work corroborates and extends previous work by our group demonstrating regenerative outcomes in red Duroc pig burns after inhibiting focal adhesion kinase (FAK), another important mechanotransduction factor (83).

Clinical translation is further derisked by our human xenograft results. Although it may seem surprising that conserved fibroblast populations would exist between such disparate wound contexts, we argue that this finding likely reflects a broadly conserved role for mechanotransduction and YAP signaling in fibrosis. For instance, a recent paper by Ma et al. studying systemic sclerosis identified an endothelium-derived fibroblast population, which undergoes endothelial-mesenchymal transition in response to Hippo signaling (96). This cell population appears analogous to our pig scRNA-seq fibroblast cluster 1, and our results suggest that verteporfin may preserve its peri-endothelial identity and prevent it from adopting a profibrotic fate, consistent with the greater vascularity of verteporfin-treated wounds. These authors also identified an α-SMA–expressing, mechanically activated/fibrotic fibroblast phenotype as a predominant contributor to systemic sclerosis fibrosis; this fibroblast subset is analogous to our pig scRNA-seq fibroblast clusters 2/4. Overall, this and similar recent studies highlight the relevance of mechanotransduction signaling across diverse fibroses.

Our study has several limitations. Neonatal foreskin has distinct properties from other human skin types (for instance, it lacks HFs or sweat glands) and denser collagen than most skin at baseline, making demonstrating antifibrotic results more challenging. Thie xenograft model in CD-1 nude mice, compared with human skin in situ, also represents a fundamentally immunocompromised setting. In addition, although the engrafted human skin is more human-like than mouse-like in its intrinsic stiffness, placing it within the setting of the mouse dorsum means that the skin’s environment lacks the full residual strain of tight-skinned animals. Finally, our xenograft experiments also reflect a shorter time course of healing compared with pig studies, although given that we saw substantial molecular/cellular divergence between verteporfin- and vehicle-treated wounds by 2 weeks in pigs, we believe the 14-day time point in xenografts is still robust for analyzing differences between these wound conditions. Despite these limitations, we believe that our xenograft experiments serve as important proof-of-concept work for the potential effects of verteporfin in human wound healing. Ultimately, these results will need to be replicated in human clinical trials, but given the outcomes of our large animal studies thus far and the well-established safety profile of verteporfin, we believe that local verteporfin treatment could be promising for clinical scar prevention. Overall, this study provides translationally robust evidence that local, one-time verteporfin treatment may be a useful approach for the prevention of skin scarring.

MATERIALS AND METHODS

Study design

The objective of this study was to determine the effects and interrogate the cellular mechanism of verteporfin treatment in porcine and human scarring. Our study involved excisional wounds in 12-week-old red Duroc pigs and human foreskin xenografts in CD-1 nude mice and in vitro experiments on porcine skin fibroblasts. All animal experiments were performed under approved Stanford Administrative Panel for Laboratory Animal Care (APLAC) Protocols (#33704, pig experiments; #11048, xenograft experiments). Foreskin samples for xenografts were collected after circumcision at the Lucile Packard Children’s Hospital under a Stanford Institutional Review Board (IRB)–approved protocol (#45219) from male newborn (2 days old or younger) patients with no other inclusion/exclusion criteria. Wound and skin samples and cells were analyzed as described in Results using histology (including standard histological stains and immunohistochemistry), tensile testing (Instron), cutometry, VAS scoring, scRNA-seq (10x Chromium), and CODEX (PhenoCycler) spatial proteomic analysis. All replicates/wounds were randomly assigned to treatment conditions, and both surgeons performing wounding and all individuals performing analysis (including analysis/quantification of immunohistochemistry and VAS) were blinded to the treatment condition of each specimen. Sample size for pig experiments was estimated assuming 80% power and detection of a two-point VAS increase at optimal dose compared with scarring vehicle wounds; sample size for xenograft experiments was calculated from power analysis with predicted difference based on analogous studies in pig. Detailed descriptions of all methods can be found in the Supplementary Materials.

Statistical analysis

Statistical testing was performed in GraphPad Prism v9 unless otherwise stated. For two-group comparisons, unpaired t tests were used. For multigroup analysis, one-way ANOVAs were used with Tukey’s post hoc corrections to compare groups; P ≤ 0.05 conferred statistical significance for all tests.

Supplementary Material

supplemental methods and figures

NIHMS2079743-supplement-supplemental_methods_and_figures.pdf^{(4.4MB, pdf)}

mdar_reproducibility_checklist

NIHMS2079743-supplement-mdar_reproducibility_checklist.xlsx^{(27.1KB, xlsx)}

data_file_s1

NIHMS2079743-supplement-data_file_s1.pdf^{(460.4KB, pdf)}

Acknowledgments:

We acknowledge the Research Animal Facility veterinary team at Stanford University for their support in the porcine experimental studies.

Funding:

This work was supported by the Hagey Laboratory for Pediatric Regenerative Medicine (to M.T.L., D.C.W., and G.C.G.); the Gunn/Olivier Research Fund, Stinehart/Reed Award, Scleroderma Research Foundation, Wu Tsai Human Performance Alliance, and the Pitch and Catherine Johnson Fund (to M.T.L.); and NIH grants U24-DE029463 (to M.T.L., G.C.G., and D.C.W.), R01- GM136659 (to M.T.L.), and R01-DE027346 (to D.C.W. and M.T.L.).

Footnotes

Competing interests: M.T.L., S.M., and H.E.T. are inventors on patent applications US 2022/0313658, “YAP inhibition for wound healing,” and US 2022/0261996, “Systems and methods for analyzing, detecting, and treating fibrotic connective tissue network formation,” both held by Stanford University.

Data and materials availability:

All data associated with this study are present in the paper or Supplementary Materials. This study did not generate new unique reagents. The scRNA-seq data generated during this study have been deposited in a GEO repository (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE219158).

REFERENCES AND NOTES

1.Mascharak S, desJardins-Park HE, Davitt MF, Griffin M, Borrelli MR, Moore AL, Chen K, Duoto B, Chinta M, Foster DS, Shen AH, Januszyk M, Kwon SH, Wernig G, Wan DC, Lorenz HP, Gurtner GC, Longaker MT, Preventing Engrailed-1 activation in fibroblasts yields wound regeneration without scarring. Science 372, eaba23 74 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Griffin MF, desJardins-Park HE, Mascharak S, Borrelli MR, Longaker MT, Understanding the impact of fibroblast heterogeneity on skin fibrosis. Dis. Model. Mech. 13, dmm044164 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Plikus MV, Wang X, Sinha S, Forte E, Thompson SM, Herzog EL, Driskell RR, Rosenthal N, Biernaskie J, Horsley V, Fibroblasts: Origins, definitions, and functions in health and disease. Cell 184, 3852–3872 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Talbott HE, Mascharak S, Griffin M, Wan DC, Longaker MT, Wound healing, fibroblast heterogeneity, and fibrosis. Cell Stem Cell 29, 1161–1180 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Mascharak S, Talbott HE, Januszyk M, Griffin M, Chen K, Davitt MF, Demeter J, Henn D, Bonham CA, Foster DS, Mooney N, Cheng R, Jackson PK, Wan DC, Gurtner GC, Longaker MT, Multi-omic analysis reveals divergent molecular events in scarring and regenerative wound healing. Cell Stem Cell 29, 315–327.e6 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Eming SA, Martin P, Tomic-Canic M, Wound repair and regeneration: Mechanisms, signaling, and translation. Sci. Transl. Med. 6, 265sr266 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Mascharak S, desJardins-Park HE, Longaker MT, Fibroblast heterogeneity in wound healing: Hurdles to clinical translation. Trends Mol. Med. 26, 1101–1106 (2020). [DOI] [PubMed] [Google Scholar]
8.Zomer HD, Trentin AG, Skin wound healing in humans and mice: Challenges in translational research. J. Dermatol. Sci. 90, 3–12 (2018). [DOI] [PubMed] [Google Scholar]
9.Mascharak S, desJardins-Park HE, Davitt MF, Guardino NJ, Gurtner GC, Wan DC, Longaker MT, Modulating cellular responses to mechanical forces to promote wound regeneration. Adv. Wound Care 11, 479–495 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Zhu KQ, Engrav LH, Gibran NS, Cole JK, Matsumura H, Piepkorn M, Isik FF, Carrougher GJ, Muangman PM, Yunusov MY, Yang TM, The female, red Duroc pig as an animal model of hypertrophic scarring and the potential role of the cones of skin. Burns 29, 649–664 (2003). [DOI] [PubMed] [Google Scholar]
11.Ramos ML, Gragnani A, Ferreira LM, Is there an ideal animal model to study hypertrophic scarring? J. Burn Care Res. 29, 363–368 (2008). [DOI] [PubMed] [Google Scholar]
12.desJardins-Park HE, Gurtner GC, Wan DC, Longaker MT, From chronic wounds to scarring: The growing healthcare burden of under- and over-healing wounds. Adv. Wound Care 11, 496–510 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Gurtner GC, Dauskardt RH, Wong VW, Bhatt KA, Wu K, Vial IN, Padois K, Korman JM, Longaker MT, Improving cutaneous scar formation by controlling the mechanical environment: Large animal and phase I studies. Ann. Surg. 254, 217–225 (2011). [DOI] [PubMed] [Google Scholar]
14.Wang C, Zhu X, Feng W, Yu Y, Jeong K, Guo W, Lu Y, Mills GB, Verteporfin inhibits YAP function through up-regulating 14–3-3o sequestering YAP in the cytoplasm. Am. J. Cancer Res. 6, 27–37 (2016). [PMC free article] [PubMed] [Google Scholar]
15.Polkoff KM, Gupta NK, Green AJ, Murphy Y, Chung J, Gleason KL, Simpson SG, Walker DM, Collins B, Piedrahita JA, LGR5 is a conserved marker of hair follicle stem cells in multiple species and is present early and throughout follicle morphogenesis. Sci. Rep. 12, 9104 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Jaks V, Barker N, Kasper M, van Es JH, Snippert HJ, Clevers H, Toftgård R, Lgr5 marks cycling, yet long-lived, hair follicle stem cells. Nat. Genet. 40, 1291–1299 (2008). [DOI] [PubMed] [Google Scholar]
17.Zhang Y, Yu J, Shi C, Huang Y, Wang Y, Yang T, Yang J, Lef1 contributes to the differentiation of bulge stem cells by nuclear translocation and cross-talk with the Notch signaling pathway. Int. J. Med. Sci. 10, 738–746 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Phan QM, Fine GM, Salz L, Herrera GG, Wildman B, Driskell IM, Driskell RR, Lef1 expression in fibroblasts maintains developmental potential in adult skin to regenerate wounds. eLife 9, e60066 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Rinkevich Y, Walmsley GG, Hu MS, Maan ZN, Newman AM, Drukker M, Januszyk M, Krampitz GW, Gurtner GC, Lorenz HP, Weissman IL, Longaker MT, Identification and isolation of a dermal lineage with intrinsic fibrogenic potential. Science 348, aaa2151 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Mah W, Jiang G, Olver D, Gallant-Behm C, Wiebe C, Hart DA, Koivisto L, Larjava H, Hakkinen L, Elevated CD26 expression by skin fibroblasts distinguishes a profibrotic phenotype involved in scar formation compared to gingival fibroblasts. Am. J. Pathol. 187, 1717–1735 (2017). [DOI] [PubMed] [Google Scholar]
21.Lee MJ, Byun MR, Furutani-Seiki M, Hong JH, Jung HS, YAP and TAZ regulate skin wound healing. J. Invest. Dermatol. 134, 518–525 (2014). [DOI] [PubMed] [Google Scholar]
22.Elosegui-Artola A, Andreu I, Beedle AEM, Lezamiz A, Uroz M, Kosmalska AJ, Oria R, Kechagia JZ, Rico-Lastres P, Le Roux AL, Shanahan CM, Trepat X, Navajas D, Garcia-Manyes S, Roca-Cusachs P, Force triggers YAP nuclear entry by regulating transport across nuclear pores. Cell 171, 1397–1410.e14 (2017). [DOI] [PubMed] [Google Scholar]
23.Zhang C, Zhu H, Ren X, Gao B, Cheng B, Liu S, Sha B, Li Z, Zhang Z, Lv Y, Wang H, Guo H, Lu TJ, Xu F, Genin GM, Lin M, Mechanics-driven nuclear localization of YAP can be reversed by N-cadherin ligation in mesenchymal stem cells. Nat. Commun. 12, 6229 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Mascharak S, Guo JL, Foster DS, Khan A, Davitt MF, Nguyen AT, Burcham AR, Chinta MS, Guardino NJ, Griffin M, Lopez DM, Miller E, Januszyk M, Raghavan SS, Longacre TA, Delitto DJ, Norton JA, Longaker MT, Desmoplastic stromal signatures predict patient outcomes in pancreatic ductal adenocarcinoma. Cell. Rep. Med. 4, 101248 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Abbas DB, Griffin M, Fahy EJ, Spielman AF, Guardino NJ, Pu A, Lintel H, Lorenz HP, Longaker MT, Wan DC, Establishing a xenograft model with CD-1 nude mice to study human skin wound repair. Plast. Reconstr. Surg. 153, 121–128 (2024). [DOI] [PubMed] [Google Scholar]
26.Mack KL, Talbott HE, Griffin MF, Parker JBL, Guardino NJ, Spielman AF, Davitt MF, Mascharak S, Downer M, Morgan A, Valencia C, Akras D, Berger MJ, Wan DC, Fraser HB, Longaker MT, Allele-specific expression reveals genetic drivers of tissue regeneration in mice. Cell Stem Cell 30, 1368–1381.e6 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Corr DT, Hart DA, Biomechanics of scar tissue and uninjured skin. Adv. Wound Care 2, 37–43 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Verteporfin in Photodynamic Therapy Study Group, Verteporfin therapy of subfoveal choroidal neovascularization in age-related macular degeneration: Two-year results of a randomized clinical trial including lesions with occult with no classic choroidal neovascularization--verteporfin in photodynamic therapy report 2. Am. J. Ophthalmol. 131, 541–560 (2001). [DOI] [PubMed] [Google Scholar]
29.Foster DS, Januszyk M, Yost KE, Chinta MS, Gulati GS, Nguyen AT, Burcham AR, Salhotra A, Ransom RC, Henn D, Chen K, Mascharak S, Tolentino K, Titan AL, Jones RE, da Silva O, Leavitt WT, Marshall CD, Desjardins-Park HE, Hu MS, Wan DC, Wernig G, Wagh D, Coller J, Norton JA, Gurtner GC, Newman AM, Chang HY, Longaker MT, Integrated spatial multiomics reveals fibroblast fate during tissue repair. Proc. Natl. Acad. Sci. U.S.A. 118, e2110025118 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Zhao J, Randive R, Stewart JA, Molecular mechanisms of AGE/RAGE-mediated fibrosis in the diabetic heart. World J. Diabetes 5, 860–867 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Li G, Li YY, Sun JE, Lin WH, Zhou RX, ILK-PI3K/AKT pathway participates in cutaneous wound contraction by regulating fibroblast migration and differentiation to myofibroblast. Lab. Invest. 96, 741–751 (2016). [DOI] [PubMed] [Google Scholar]
32.Teng Y, Fan Y, Ma J, Lu W, Liu N, Chen Y, Pan W, Tao X, The PI3K/Akt pathway: Emerging roles in skin homeostasis and a group of non-malignant skin disorders. Cells 10, 1219 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Bui TM, Wiesolek HL, Sumagin R, ICAM-1: A master regulator of cellular responses in inflammation, injury resolution, and tumorigenesis. J. Leukoc. Biol. 108, 787–799 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Rudziak P, Ellis CG, Kowalewska PM, Role and molecular mechanisms of pericytes in regulation of leukocyte diapedesis in inflamed tissues. Mediators Inflamm. 2019, 4123605 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Michie KA, Bermeister A, Robertson NO, Goodchild SC, Curmi PMG, Two sides of the coin: Ezrin/radixin/moesin and merlin control membrane structure and contact inhibition. Int. J. Mol. Sci. 20, 1996 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Carbonetti G, Wilpshaar T, Kroonen J, Studholme K, Converso C, d’Oelsnitz S, Kaczocha M, FABP5 coordinates lipid signaling that promotes prostate cancer metastasis. Sci. Rep. 9, 18944 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Ogawa E, Owada Y, Ikawa S, Adachi Y, Egawa T, Nemoto K, Suzuki K, Hishinuma T, Kawashima H, Kondo H, Muto M, Aiba S, Okuyama R, Epidermal FABP (FABP5) regulates keratinocyte differentiation by 13(S)-HODE-mediated activation of the NF-κB signaling pathway. J. Invest. Dermatol. 131, 604–612 (2011). [DOI] [PubMed] [Google Scholar]
38.Song J, Zhang H, Wang Z, Xu W, Zhong L, Cao J, Yang J, Tian Y, Yu D, Ji J, Zhang S, The role of FABP5 in radiation-induced human skin fibrosis. Radiat. Res. 189, 177–186 (2018). [DOI] [PubMed] [Google Scholar]
39.Murakami-Satsutani N, Ito T, Nakanishi T, Inagaki N, Tanaka A, Vien PT, Kibata K, Inaba M, Nomura S, IL-33 promotes the induction and maintenance of Th2 immune responses by enhancing the function of OX40 ligand. Allergol. Int. 63, 443–455 (2014). [DOI] [PubMed] [Google Scholar]
40.Wen Y, Emontzpohl C, Xu L, Atkins CL, Jeong JM, Yang Y, Kim K, Wu C, Akira S, Ju C, Interleukin-33 facilitates liver regeneration through serotonin-involved gut-liver axis. Hepatology 77, 1580–1592 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Valero-Pacheco N, Tang EK, Massri N, Loia R, Chemerinski A, Wu T, Begum S, El-Naccache DW, Gause WC, Arora R, Douglas NC, Beaulieu AM, Maternal IL-33 critically regulates tissue remodeling and type 2 immune responses in the uterus during early pregnancy in mice. Proc. Natl. Acad. Sci. U.S.A. 119, e2123267119 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Rockey DC, Weymouth N, Shi Z, Smooth muscle a actin (Acta2) and myofibroblast function during hepatic wound healing. PLOS ONE 8, e77166 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Sharanek A, Burban A, Laaper M, Heckel E, Joyal JS, Soleimani VD, Jahani-Asl A, OSMR controls glioma stem cell respiration and confers resistance of glioblastoma to ionizing radiation. Nat. Commun. 11,4116 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Stawski L, Trojanowska M, Oncostatin M and its role in fibrosis. Connect. Tissue Res. 60, 40–49 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Matsuda M, Tsurusaki S, Miyata N, Saijou E, Okochi H, Miyajima A, Tanaka M, Oncostatin M causes liver fibrosis by regulating cooperation between hepatic stellate cells and macrophages in mice. Hepatology 67, 296–312 (2018). [DOI] [PubMed] [Google Scholar]
46.Zhang Q, Wei T, Shim K, Wright K, Xu K, Palka-Hamblin HL, Jurkevich A, Khare S, Atypical role of sprouty in colorectal cancer: Sprouty repression inhibits epithelial-mesenchymal transition. Oncogene 35, 3151–3162 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Tan X, Zhu Y, Chen C, Chen X, Qin Y, Qu B, Luo L, Lin H, Wu M, Chen W, Liu Y, Sprouty2 suppresses epithelial-mesenchymal transition of human lens epithelial cells through blockade of Smad2 and ERK1/2 pathways. PLOS ONE 11, e0159275 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Saini M, Verma A, Mathew SJ, SPRY2 is a novel MET interactor that regulates metastatic potential and differentiation in rhabdomyosarcoma. Cell Death Dis. 9, 237 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Li T, Forbes ME, Fuller GN, Li J, Yang X, Zhang W, IGFBP2: Integrative hub of developmental and oncogenic signaling network. Oncogene 39, 2243–2257 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Ding H, Wu T, Insulin-like growth factor binding proteins in autoimmune diseases. Front. Endocrinol. 9, 499 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Guiot J, Bondue B, Henket M, Corhay JL, Louis R, Raised serum levels of IGFBP-1 and IGFBP-2 in idiopathic pulmonary fibrosis. BMC Pulm. Med. 16, 86 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Hamilton DW, Functional role of periostin in development and wound repair: Implications for connective tissue disease. J. Cell. Commun. Signal 2, 9–17 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Zhang W, Ge Y, Cheng Q, Zhang Q, Fang L, Zheng J, Decorin is a pivotal effector in the extracellular matrix and tumour microenvironment. Oncotarget 9, 5480–5491 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Liu XJ, Kong FZ, Wang YH, Zheng JH, Wan WD, Deng CL, Mao GY, Li J, Yang XM, Zhang YL, Zhang XL, Yang SL, Zhang ZG, Lumican accelerates wound healing by enhancing a2p1 integrin-mediated fibroblast contractility. PLOS ONE 8, e67124 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Karamanou K, Perrot G, Maquart FX, Brezillon S, Lumican as a multivalent effector in wound healing. Adv. Drug Deliv. Rev. 129, 344–351 (2018). [DOI] [PubMed] [Google Scholar]
56.Yang AT, Kim YO, Yan XZ, Abe H, Aslam M, Park KS, Zhao XY, Jia JD, Klein T, You H, Schuppan D, Fibroblast activation protein activates macrophages and promotes parenchymal liver inflammation and fibrosis. Cell. Mol. Gastroenterol. Hepatol. 15, 841–867 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Zorniak M, Clark PA, Kuo JS, Myelin-forming cell-specific cadherin-19 is a marker for minimally infiltrative glioblastoma stem-like cells. J. Neurosurg. 122, 69–77 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Weiss T, Taschner-Mandl S, Janker L, Bileck A, Rifatbegovic F, Kromp F, Sorger H, Kauer MO, Frech C, Windhager R, Gerner C, Ambros PF, Ambros IM, Schwann cell plasticity regulates neuroblastic tumor cell differentiation via epidermal growth factor-like protein 8. Nat. Commun. 12, 1624 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
59.LeBlanc AC, Foldvari M, Spencer DF, Breckenridge WC, Fenwick RG, Williams DL, Mezei C, The apolipoprotein A-I gene is actively expressed in the rapidly myelinating avian peripheral nerve. J. Cell Biol. 109, 1245–1256 (1989). [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Trapnell C, Cacchiarelli D, Grimsby J, Pokharel P, Li S, Morse M, Lennon NJ, Livak KJ, Mikkelsen TS, Rinn JL, The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat. Biotechnol. 32, 381–386 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Yu H, Königshoff M, Jayachandran A, Handley D, Seeger W, Kaminski N, Eickelberg O, Transgelin is a direct target of TGF-β/Smad3-dependent epithelial cell migration in lung fibrosis. FASEB J. 22, 1778–1789 (2008). [DOI] [PubMed] [Google Scholar]
62.Sahin H, Wasmuth HE, Chemokines in tissue fibrosis. Biochim. Biophys. Acta 1832, 1041–1048 (2013). [DOI] [PubMed] [Google Scholar]
63.Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, Myung P, Plikus MV, Nie Q, Inference and analysis of cell-cell communication using CellChat. Nat. Commun. 12, 1088 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Cao G, O'Brien CD, Zhou Z, Sanders SM, Greenbaum JN, Makrigiannakis A, DeLisser HM, Involvement of human PECAM-1 in angiogenesis and in vitro endothelial cell migration. Am. J. Physiol. Cell Physiol. 282, C1181–C1190 (2002). [DOI] [PubMed] [Google Scholar]
65.Johnson BZ, Stevenson AW, Prêle CM, Fear MW, Wood FM, The role of IL-6 in skin fibrosis and cutaneous wound healing. Biomedicine 8, 101 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Meng XM, Nikolic-Paterson DJ, Lan HY, TGF-β: The master regulator of fibrosis. Nat. Rev. Nephrol. 12, 325–338 (2016). [DOI] [PubMed] [Google Scholar]
67.Breitkopf-Heinlein K, Meyer C, König C, Gaitantzi H, Addante A, Thomas M, Wiercinska E, Cai C, Li Q, Wan F, Hellerbrand C, Valous NA, Hahnel M, Ehlting C, Bode JG, Müller-Bohl S, Klingmüller U, Altenöder J, Ilkavets I, Goumans MJ, Hawinkels LJ, Lee SJ, Wieland M, Mogler C, Ebert MP, Herrera B, Augustin H, Sánchez A, Dooley S, Ten Dijke P, BMP-9 interferes with liver regeneration and promotes liver fibrosis. Gut 66, 939–954 (2017). [DOI] [PubMed] [Google Scholar]
68.Yu WM, Yu H, Chen ZL, Laminins in peripheral nerve development and muscular dystrophy. Mol. Neurobiol. 35, 288–297 (2007). [DOI] [PubMed] [Google Scholar]
69.Ridiandries A, Tan JTM, Bursill CA, The role of chemokines in wound healing. Int. J. Mol. Sci. 19, 3217 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Zhu Z, Ding J, Tredget EE, The molecular basis of hypertrophic scars. Burns Trauma 4, 2 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Angulo-Urarte A, van der Wal T, Huveneers S, Cell-cell junctions as sensors and transducers of mechanical forces. Biochim. Biophys. Acta Biomembr. 1862, 183316 (2020). [DOI] [PubMed] [Google Scholar]
72.Jones CV, Ricardo SD, Macrophages and CSF-1: Implications for development and beyond. Organogenesis 9, 249–260 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
73.Zhu J, T helper 2 (Th2) cell differentiation, type 2 innate lymphoid cell (ILC2) development and regulation of interleukin-4 (IL-4) and IL-13 production. Cytokine 75, 14–24 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
74.Lim CH, Sun Q, Ratti K, Lee SH, Zheng Y, Takeo M, Lee W, Rabbani P, Plikus MV, Cain JE, Wang DH, Watkins DN, Millar S, Taketo MM, Myung P, Cotsarelis G, Ito M, Hedgehog stimulates hair follicle neogenesis by creating inductive dermis during murine skin wound healing. Nat. Commun. 9, 4903 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
75.Frech S, Forsthuber A, Korosec A, Lipp K, Kozumov V, Lichtenberger BM, Hedgehog signaling in papillary fibroblasts is essential for hair follicle regeneration during wound healing. J. Invest. Dermatol. 142, 1737–1748.e5 (2022). [DOI] [PubMed] [Google Scholar]
76.Wang Y, Wang G, Liu H, Tenascin-C: A key regulator in angiogenesis during wound healing. Biomolecules 12, 1689 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
77.Black S, Phillips D, Hickey JW, Kennedy-Darling J, Venkataraaman VG, Samusik N, Goltsev Y, Schurch CM, Nolan GP, CODEX multiplexed tissue imaging with DNA-conjugated antibodies. Nat. Protoc. 16, 3802–3835 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
78.Liu H, Hu J, Zheng Q, Feng X, Zhan F, Wang X, Xu G, Hua F, Piezo1 channels as force sensors in mechanical force-related chronic inflammation. Front. Immunol. 13, 816149 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
79.Chesney J, Bacher M, Bender A, Bucala R, The peripheral blood fibrocyte is a potent antigen-presenting cell capable of priming naive T cells in situ. Proc. Natl. Acad. Sci. U.S.A. 94, 6307–6312 (1997). [DOI] [PMC free article] [PubMed] [Google Scholar]
80.Molofsky AB, Savage AK, Locksley RM, Interleukin-33 in tissue homeostasis, injury, and inflammation. Immunity 42, 1005–1019 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
81.Abbasi S, Sinha S, Labit E, Rosin NL, Yoon G, Rahmani W, Jaffer A, Sharma N, Hagner A, Shah P, Arora R, Yoon J, Islam A, Uchida A, Chang CK, Stratton JA, Scott RW, Rossi FMV, Underhill TM, Biernaskie J, Distinct regulatory programs control the latent regenerative potential of dermal fibroblasts during wound healing. Cell Stem Cell 27, 396–412.e6 (2020). [DOI] [PubMed] [Google Scholar]
82.Foster DS, Januszyk M, Delitto D, Yost KE, Griffin M, Guo J, Guardino N, Delitto AE, Chinta M, Burcham AR, Nguyen AT, Bauer-Rowe KE, Titan AL, Salhotra A, Jones RE, da Silva O, Lindsay HG, Berry CE, Chen K, Henn D, Mascharak S, Talbott HE, Kim A, Nosrati F, Sivaraj D, Ransom RC, Matthews M, Khan A, Wagh D, Coller J, Gurtner GC, Wan DC, Wapnir IL, Chang HY, Norton JA, Longaker MT, Multiomic analysis reveals conservation of cancer-associated fibroblast phenotypes across species and tissue of origin. Cancer Cell 40, 1392–1406.e7 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
83.Chen K, Kwon SH, Henn D, Kuehlmann BA, Tevlin R, Bonham CA, Griffin M, Trotsyuk AA, Borrelli MR, Noishiki C, Padmanabhan J, Barrera JA, Maan ZN, Dohi T, Mays CJ, Greco AH, Sivaraj D, Lin JQ, Fehlmann T, Mermin-Bunnell AM, Mittal S, Hu MS, Zamaleeva AI, Keller A, Rajadas J, Longaker MT, Januszyk M, Gurtner GC, Disrupting biological sensors of force promotes tissue regeneration in large organisms. Nat. Commun. 12, 5256 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
84.Gordon S, Alternative activation of macrophages. Nat. Rev. Immunol. 3, 23–35 (2003). [DOI] [PubMed] [Google Scholar]
85.Gao Y, Luo C, Rui T, Fan Y, Yao Y, Shen H, Gao C, Wang T, Wang H, Chen X, Zhang J, Li D, Xia C, Li LL, Wang Z, Zhang M, Chen X, Tao L, Autophagy inhibition facilitates wound closure partially dependent on the YAP/IL-33 signaling in a mouse model of skin wound healing. FASEB J. 35, e21920 (2021). [DOI] [PubMed] [Google Scholar]
86.Oshio T, Komine M, Tsuda H, Tominaga SI, Saito H, Nakae S, Ohtsuki M, Nuclear expression of IL-33 in epidermal keratinocytes promotes wound healing in mice. J. Dermatol. Sci. 85, 106–114 (2017). [DOI] [PubMed] [Google Scholar]
87.Amuso VM, Haas MR, Cooper PO, Chatterjee R, Hafiz S, Salameh S, Gohel C, Mazumder MF, Josephson V, Kleb SS, Khorsandi K, Horvath A, Rahnavard A, Shook BA, Fibroblast-mediated macrophage recruitment supports acute wound healing. J. Invest. Dermatol. 10.1016/j.jid.2024.10.609 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
88.Jin M, Komine M, Tsuda H, Sashikawa-Kimura M, Nakae S, Motegi SI, Ohtsuki M, Interleukin-33 deficiency protects the skin from ulcer formation in an ischemia-reperfusion-induced decubitus mouse model. Exp. Dermatol. 33, e70014 (2024). [DOI] [PubMed] [Google Scholar]
89.Kotsiou OS, Gourgoulianis KI, Zarogiannis SG, IL-33/ST2 axis in organ fibrosis. Front. Immunol. 9, 2432 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
90.Barron L, Wynn TA, Fibrosis is regulated by Th2 and Th17 responses and by dynamic interactions between fibroblasts and macrophages. Am. J. Physiol. Gastrointest. Liver Physiol. 300, G723–G728 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
91.desJardins-Park HE, Chinta MS, Foster DS, Borrelli MR, Shen AH, Wan DC, Longaker MT, Fibroblast heterogeneity in and its implications for plastic and reconstructive surgery: A basic science review. Plast. Reconstr. Surg. Glob. Open 8, e2927 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
92.Swindle MM, Makin A, Herron AJ, Clubb FJ, Frazier KS, Swine as models in biomedical research and toxicology testing. Vet. Pathol. 49, 344–356 (2012). [DOI] [PubMed] [Google Scholar]
93.Masson-Meyers DS, Andrade TAM, Caetano GF, Guimaraes FR, Leite MN, Leite SN, Frade MAC, Experimental models and methods for cutaneous wound healing assessment. Int. J. Exp. Pathol. 101,21–37 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
94.Wong VW, Sorkin M, Glotzbach JP, Longaker MT, Gurtner GC, Surgical approaches to create murine models of human wound healing. J. Biomed. Biotechnol. 2011,969618 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
95.Rabello FB, Souza CD, Farina Junior JA, Update on hypertrophic scar treatment. Clinics 69, 565–573 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
96.Ma F, Tsou PS, Gharaee-Kermani M, Plazyo O, Xing X, Kirma J, Wasikowski R, Hile GA, Harms PW, Jiang Y, Xing E, Nakamura M, Ochocki D, Brodie WD, Pillai S, Maverakis E, Pellegrini M, Modlin RL, Varga J, Tsoi LC, Lafyatis R, Kahlenberg JM, Billi AC, Khanna D, Gudjonsson JE, Systems-based identification of the Hippo pathway for promoting fibrotic mesenchymal differentiation in systemic sclerosis. Nat. Commun. 15, 210 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
97.Ruifrok AC, Johnston DA, Quantification of histochemical staining by color deconvolution. Anal. Quant. Cytol. Histol. 23, 291–299 (2001). [PubMed] [Google Scholar]
98.Chen EY, Tan CM, Kou Y, Duan Q, Wang Z, Meirelles GV, Clark NR, Ma'ayan A, Enrichr: Interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinformatics 14, 128 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
99.Ouyang JF, Kamaraj US, Cao EY, Rackham OJL, ShinyCell: Simple and sharable visualization of single-cell gene expression data. Bioinformatics 37, 3374–3376 (2021). [DOI] [PubMed] [Google Scholar]
100.Bergen V, Lange M, Peidli S, Wolf FA, Theis FJ, Generalizing RNA velocity to transient cell states through dynamical modeling. Nat. Biotechnol. 38, 1408–1414 (2020). [DOI] [PubMed] [Google Scholar]
101.Chen K, Vigliotti A, Bacca M, McMeeking RM, Deshpande VS, Holmes JW, Role of boundary conditions in determining cell alignment in response to stretch. Proc. Natl. Acad. Sci. U.S.A. 115, 986–991 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplemental methods and figures

NIHMS2079743-supplement-supplemental_methods_and_figures.pdf^{(4.4MB, pdf)}

mdar_reproducibility_checklist

NIHMS2079743-supplement-mdar_reproducibility_checklist.xlsx^{(27.1KB, xlsx)}

data_file_s1

NIHMS2079743-supplement-data_file_s1.pdf^{(460.4KB, pdf)}

Data Availability Statement

[R1] 1.Mascharak S, desJardins-Park HE, Davitt MF, Griffin M, Borrelli MR, Moore AL, Chen K, Duoto B, Chinta M, Foster DS, Shen AH, Januszyk M, Kwon SH, Wernig G, Wan DC, Lorenz HP, Gurtner GC, Longaker MT, Preventing Engrailed-1 activation in fibroblasts yields wound regeneration without scarring. Science 372, eaba23 74 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R2] 2.Griffin MF, desJardins-Park HE, Mascharak S, Borrelli MR, Longaker MT, Understanding the impact of fibroblast heterogeneity on skin fibrosis. Dis. Model. Mech. 13, dmm044164 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R3] 3.Plikus MV, Wang X, Sinha S, Forte E, Thompson SM, Herzog EL, Driskell RR, Rosenthal N, Biernaskie J, Horsley V, Fibroblasts: Origins, definitions, and functions in health and disease. Cell 184, 3852–3872 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R4] 4.Talbott HE, Mascharak S, Griffin M, Wan DC, Longaker MT, Wound healing, fibroblast heterogeneity, and fibrosis. Cell Stem Cell 29, 1161–1180 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R5] 5.Mascharak S, Talbott HE, Januszyk M, Griffin M, Chen K, Davitt MF, Demeter J, Henn D, Bonham CA, Foster DS, Mooney N, Cheng R, Jackson PK, Wan DC, Gurtner GC, Longaker MT, Multi-omic analysis reveals divergent molecular events in scarring and regenerative wound healing. Cell Stem Cell 29, 315–327.e6 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] 6.Eming SA, Martin P, Tomic-Canic M, Wound repair and regeneration: Mechanisms, signaling, and translation. Sci. Transl. Med. 6, 265sr266 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R7] 7.Mascharak S, desJardins-Park HE, Longaker MT, Fibroblast heterogeneity in wound healing: Hurdles to clinical translation. Trends Mol. Med. 26, 1101–1106 (2020). [DOI] [PubMed] [Google Scholar]

[R8] 8.Zomer HD, Trentin AG, Skin wound healing in humans and mice: Challenges in translational research. J. Dermatol. Sci. 90, 3–12 (2018). [DOI] [PubMed] [Google Scholar]

[R9] 9.Mascharak S, desJardins-Park HE, Davitt MF, Guardino NJ, Gurtner GC, Wan DC, Longaker MT, Modulating cellular responses to mechanical forces to promote wound regeneration. Adv. Wound Care 11, 479–495 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R10] 10.Zhu KQ, Engrav LH, Gibran NS, Cole JK, Matsumura H, Piepkorn M, Isik FF, Carrougher GJ, Muangman PM, Yunusov MY, Yang TM, The female, red Duroc pig as an animal model of hypertrophic scarring and the potential role of the cones of skin. Burns 29, 649–664 (2003). [DOI] [PubMed] [Google Scholar]

[R11] 11.Ramos ML, Gragnani A, Ferreira LM, Is there an ideal animal model to study hypertrophic scarring? J. Burn Care Res. 29, 363–368 (2008). [DOI] [PubMed] [Google Scholar]

[R12] 12.desJardins-Park HE, Gurtner GC, Wan DC, Longaker MT, From chronic wounds to scarring: The growing healthcare burden of under- and over-healing wounds. Adv. Wound Care 11, 496–510 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R13] 13.Gurtner GC, Dauskardt RH, Wong VW, Bhatt KA, Wu K, Vial IN, Padois K, Korman JM, Longaker MT, Improving cutaneous scar formation by controlling the mechanical environment: Large animal and phase I studies. Ann. Surg. 254, 217–225 (2011). [DOI] [PubMed] [Google Scholar]

[R14] 14.Wang C, Zhu X, Feng W, Yu Y, Jeong K, Guo W, Lu Y, Mills GB, Verteporfin inhibits YAP function through up-regulating 14–3-3o sequestering YAP in the cytoplasm. Am. J. Cancer Res. 6, 27–37 (2016). [PMC free article] [PubMed] [Google Scholar]

[R15] 15.Polkoff KM, Gupta NK, Green AJ, Murphy Y, Chung J, Gleason KL, Simpson SG, Walker DM, Collins B, Piedrahita JA, LGR5 is a conserved marker of hair follicle stem cells in multiple species and is present early and throughout follicle morphogenesis. Sci. Rep. 12, 9104 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] 16.Jaks V, Barker N, Kasper M, van Es JH, Snippert HJ, Clevers H, Toftgård R, Lgr5 marks cycling, yet long-lived, hair follicle stem cells. Nat. Genet. 40, 1291–1299 (2008). [DOI] [PubMed] [Google Scholar]

[R17] 17.Zhang Y, Yu J, Shi C, Huang Y, Wang Y, Yang T, Yang J, Lef1 contributes to the differentiation of bulge stem cells by nuclear translocation and cross-talk with the Notch signaling pathway. Int. J. Med. Sci. 10, 738–746 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R18] 18.Phan QM, Fine GM, Salz L, Herrera GG, Wildman B, Driskell IM, Driskell RR, Lef1 expression in fibroblasts maintains developmental potential in adult skin to regenerate wounds. eLife 9, e60066 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] 19.Rinkevich Y, Walmsley GG, Hu MS, Maan ZN, Newman AM, Drukker M, Januszyk M, Krampitz GW, Gurtner GC, Lorenz HP, Weissman IL, Longaker MT, Identification and isolation of a dermal lineage with intrinsic fibrogenic potential. Science 348, aaa2151 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] 20.Mah W, Jiang G, Olver D, Gallant-Behm C, Wiebe C, Hart DA, Koivisto L, Larjava H, Hakkinen L, Elevated CD26 expression by skin fibroblasts distinguishes a profibrotic phenotype involved in scar formation compared to gingival fibroblasts. Am. J. Pathol. 187, 1717–1735 (2017). [DOI] [PubMed] [Google Scholar]

[R21] 21.Lee MJ, Byun MR, Furutani-Seiki M, Hong JH, Jung HS, YAP and TAZ regulate skin wound healing. J. Invest. Dermatol. 134, 518–525 (2014). [DOI] [PubMed] [Google Scholar]

[R22] 22.Elosegui-Artola A, Andreu I, Beedle AEM, Lezamiz A, Uroz M, Kosmalska AJ, Oria R, Kechagia JZ, Rico-Lastres P, Le Roux AL, Shanahan CM, Trepat X, Navajas D, Garcia-Manyes S, Roca-Cusachs P, Force triggers YAP nuclear entry by regulating transport across nuclear pores. Cell 171, 1397–1410.e14 (2017). [DOI] [PubMed] [Google Scholar]

[R23] 23.Zhang C, Zhu H, Ren X, Gao B, Cheng B, Liu S, Sha B, Li Z, Zhang Z, Lv Y, Wang H, Guo H, Lu TJ, Xu F, Genin GM, Lin M, Mechanics-driven nuclear localization of YAP can be reversed by N-cadherin ligation in mesenchymal stem cells. Nat. Commun. 12, 6229 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R24] 24.Mascharak S, Guo JL, Foster DS, Khan A, Davitt MF, Nguyen AT, Burcham AR, Chinta MS, Guardino NJ, Griffin M, Lopez DM, Miller E, Januszyk M, Raghavan SS, Longacre TA, Delitto DJ, Norton JA, Longaker MT, Desmoplastic stromal signatures predict patient outcomes in pancreatic ductal adenocarcinoma. Cell. Rep. Med. 4, 101248 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R25] 25.Abbas DB, Griffin M, Fahy EJ, Spielman AF, Guardino NJ, Pu A, Lintel H, Lorenz HP, Longaker MT, Wan DC, Establishing a xenograft model with CD-1 nude mice to study human skin wound repair. Plast. Reconstr. Surg. 153, 121–128 (2024). [DOI] [PubMed] [Google Scholar]

[R26] 26.Mack KL, Talbott HE, Griffin MF, Parker JBL, Guardino NJ, Spielman AF, Davitt MF, Mascharak S, Downer M, Morgan A, Valencia C, Akras D, Berger MJ, Wan DC, Fraser HB, Longaker MT, Allele-specific expression reveals genetic drivers of tissue regeneration in mice. Cell Stem Cell 30, 1368–1381.e6 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R27] 27.Corr DT, Hart DA, Biomechanics of scar tissue and uninjured skin. Adv. Wound Care 2, 37–43 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R28] 28.Verteporfin in Photodynamic Therapy Study Group, Verteporfin therapy of subfoveal choroidal neovascularization in age-related macular degeneration: Two-year results of a randomized clinical trial including lesions with occult with no classic choroidal neovascularization--verteporfin in photodynamic therapy report 2. Am. J. Ophthalmol. 131, 541–560 (2001). [DOI] [PubMed] [Google Scholar]

[R29] 29.Foster DS, Januszyk M, Yost KE, Chinta MS, Gulati GS, Nguyen AT, Burcham AR, Salhotra A, Ransom RC, Henn D, Chen K, Mascharak S, Tolentino K, Titan AL, Jones RE, da Silva O, Leavitt WT, Marshall CD, Desjardins-Park HE, Hu MS, Wan DC, Wernig G, Wagh D, Coller J, Norton JA, Gurtner GC, Newman AM, Chang HY, Longaker MT, Integrated spatial multiomics reveals fibroblast fate during tissue repair. Proc. Natl. Acad. Sci. U.S.A. 118, e2110025118 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.Zhao J, Randive R, Stewart JA, Molecular mechanisms of AGE/RAGE-mediated fibrosis in the diabetic heart. World J. Diabetes 5, 860–867 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R31] 31.Li G, Li YY, Sun JE, Lin WH, Zhou RX, ILK-PI3K/AKT pathway participates in cutaneous wound contraction by regulating fibroblast migration and differentiation to myofibroblast. Lab. Invest. 96, 741–751 (2016). [DOI] [PubMed] [Google Scholar]

[R32] 32.Teng Y, Fan Y, Ma J, Lu W, Liu N, Chen Y, Pan W, Tao X, The PI3K/Akt pathway: Emerging roles in skin homeostasis and a group of non-malignant skin disorders. Cells 10, 1219 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] 33.Bui TM, Wiesolek HL, Sumagin R, ICAM-1: A master regulator of cellular responses in inflammation, injury resolution, and tumorigenesis. J. Leukoc. Biol. 108, 787–799 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R34] 34.Rudziak P, Ellis CG, Kowalewska PM, Role and molecular mechanisms of pericytes in regulation of leukocyte diapedesis in inflamed tissues. Mediators Inflamm. 2019, 4123605 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] 35.Michie KA, Bermeister A, Robertson NO, Goodchild SC, Curmi PMG, Two sides of the coin: Ezrin/radixin/moesin and merlin control membrane structure and contact inhibition. Int. J. Mol. Sci. 20, 1996 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R36] 36.Carbonetti G, Wilpshaar T, Kroonen J, Studholme K, Converso C, d’Oelsnitz S, Kaczocha M, FABP5 coordinates lipid signaling that promotes prostate cancer metastasis. Sci. Rep. 9, 18944 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R37] 37.Ogawa E, Owada Y, Ikawa S, Adachi Y, Egawa T, Nemoto K, Suzuki K, Hishinuma T, Kawashima H, Kondo H, Muto M, Aiba S, Okuyama R, Epidermal FABP (FABP5) regulates keratinocyte differentiation by 13(S)-HODE-mediated activation of the NF-κB signaling pathway. J. Invest. Dermatol. 131, 604–612 (2011). [DOI] [PubMed] [Google Scholar]

[R38] 38.Song J, Zhang H, Wang Z, Xu W, Zhong L, Cao J, Yang J, Tian Y, Yu D, Ji J, Zhang S, The role of FABP5 in radiation-induced human skin fibrosis. Radiat. Res. 189, 177–186 (2018). [DOI] [PubMed] [Google Scholar]

[R39] 39.Murakami-Satsutani N, Ito T, Nakanishi T, Inagaki N, Tanaka A, Vien PT, Kibata K, Inaba M, Nomura S, IL-33 promotes the induction and maintenance of Th2 immune responses by enhancing the function of OX40 ligand. Allergol. Int. 63, 443–455 (2014). [DOI] [PubMed] [Google Scholar]

[R40] 40.Wen Y, Emontzpohl C, Xu L, Atkins CL, Jeong JM, Yang Y, Kim K, Wu C, Akira S, Ju C, Interleukin-33 facilitates liver regeneration through serotonin-involved gut-liver axis. Hepatology 77, 1580–1592 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R41] 41.Valero-Pacheco N, Tang EK, Massri N, Loia R, Chemerinski A, Wu T, Begum S, El-Naccache DW, Gause WC, Arora R, Douglas NC, Beaulieu AM, Maternal IL-33 critically regulates tissue remodeling and type 2 immune responses in the uterus during early pregnancy in mice. Proc. Natl. Acad. Sci. U.S.A. 119, e2123267119 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R42] 42.Rockey DC, Weymouth N, Shi Z, Smooth muscle a actin (Acta2) and myofibroblast function during hepatic wound healing. PLOS ONE 8, e77166 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] 43.Sharanek A, Burban A, Laaper M, Heckel E, Joyal JS, Soleimani VD, Jahani-Asl A, OSMR controls glioma stem cell respiration and confers resistance of glioblastoma to ionizing radiation. Nat. Commun. 11,4116 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R44] 44.Stawski L, Trojanowska M, Oncostatin M and its role in fibrosis. Connect. Tissue Res. 60, 40–49 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R45] 45.Matsuda M, Tsurusaki S, Miyata N, Saijou E, Okochi H, Miyajima A, Tanaka M, Oncostatin M causes liver fibrosis by regulating cooperation between hepatic stellate cells and macrophages in mice. Hepatology 67, 296–312 (2018). [DOI] [PubMed] [Google Scholar]

[R46] 46.Zhang Q, Wei T, Shim K, Wright K, Xu K, Palka-Hamblin HL, Jurkevich A, Khare S, Atypical role of sprouty in colorectal cancer: Sprouty repression inhibits epithelial-mesenchymal transition. Oncogene 35, 3151–3162 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R47] 47.Tan X, Zhu Y, Chen C, Chen X, Qin Y, Qu B, Luo L, Lin H, Wu M, Chen W, Liu Y, Sprouty2 suppresses epithelial-mesenchymal transition of human lens epithelial cells through blockade of Smad2 and ERK1/2 pathways. PLOS ONE 11, e0159275 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R48] 48.Saini M, Verma A, Mathew SJ, SPRY2 is a novel MET interactor that regulates metastatic potential and differentiation in rhabdomyosarcoma. Cell Death Dis. 9, 237 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R49] 49.Li T, Forbes ME, Fuller GN, Li J, Yang X, Zhang W, IGFBP2: Integrative hub of developmental and oncogenic signaling network. Oncogene 39, 2243–2257 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R50] 50.Ding H, Wu T, Insulin-like growth factor binding proteins in autoimmune diseases. Front. Endocrinol. 9, 499 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R51] 51.Guiot J, Bondue B, Henket M, Corhay JL, Louis R, Raised serum levels of IGFBP-1 and IGFBP-2 in idiopathic pulmonary fibrosis. BMC Pulm. Med. 16, 86 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R52] 52.Hamilton DW, Functional role of periostin in development and wound repair: Implications for connective tissue disease. J. Cell. Commun. Signal 2, 9–17 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R53] 53.Zhang W, Ge Y, Cheng Q, Zhang Q, Fang L, Zheng J, Decorin is a pivotal effector in the extracellular matrix and tumour microenvironment. Oncotarget 9, 5480–5491 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R54] 54.Liu XJ, Kong FZ, Wang YH, Zheng JH, Wan WD, Deng CL, Mao GY, Li J, Yang XM, Zhang YL, Zhang XL, Yang SL, Zhang ZG, Lumican accelerates wound healing by enhancing a2p1 integrin-mediated fibroblast contractility. PLOS ONE 8, e67124 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R55] 55.Karamanou K, Perrot G, Maquart FX, Brezillon S, Lumican as a multivalent effector in wound healing. Adv. Drug Deliv. Rev. 129, 344–351 (2018). [DOI] [PubMed] [Google Scholar]

[R56] 56.Yang AT, Kim YO, Yan XZ, Abe H, Aslam M, Park KS, Zhao XY, Jia JD, Klein T, You H, Schuppan D, Fibroblast activation protein activates macrophages and promotes parenchymal liver inflammation and fibrosis. Cell. Mol. Gastroenterol. Hepatol. 15, 841–867 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R57] 57.Zorniak M, Clark PA, Kuo JS, Myelin-forming cell-specific cadherin-19 is a marker for minimally infiltrative glioblastoma stem-like cells. J. Neurosurg. 122, 69–77 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R58] 58.Weiss T, Taschner-Mandl S, Janker L, Bileck A, Rifatbegovic F, Kromp F, Sorger H, Kauer MO, Frech C, Windhager R, Gerner C, Ambros PF, Ambros IM, Schwann cell plasticity regulates neuroblastic tumor cell differentiation via epidermal growth factor-like protein 8. Nat. Commun. 12, 1624 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R59] 59.LeBlanc AC, Foldvari M, Spencer DF, Breckenridge WC, Fenwick RG, Williams DL, Mezei C, The apolipoprotein A-I gene is actively expressed in the rapidly myelinating avian peripheral nerve. J. Cell Biol. 109, 1245–1256 (1989). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R60] 60.Trapnell C, Cacchiarelli D, Grimsby J, Pokharel P, Li S, Morse M, Lennon NJ, Livak KJ, Mikkelsen TS, Rinn JL, The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat. Biotechnol. 32, 381–386 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R61] 61.Yu H, Königshoff M, Jayachandran A, Handley D, Seeger W, Kaminski N, Eickelberg O, Transgelin is a direct target of TGF-β/Smad3-dependent epithelial cell migration in lung fibrosis. FASEB J. 22, 1778–1789 (2008). [DOI] [PubMed] [Google Scholar]

[R62] 62.Sahin H, Wasmuth HE, Chemokines in tissue fibrosis. Biochim. Biophys. Acta 1832, 1041–1048 (2013). [DOI] [PubMed] [Google Scholar]

[R63] 63.Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, Myung P, Plikus MV, Nie Q, Inference and analysis of cell-cell communication using CellChat. Nat. Commun. 12, 1088 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R64] 64.Cao G, O'Brien CD, Zhou Z, Sanders SM, Greenbaum JN, Makrigiannakis A, DeLisser HM, Involvement of human PECAM-1 in angiogenesis and in vitro endothelial cell migration. Am. J. Physiol. Cell Physiol. 282, C1181–C1190 (2002). [DOI] [PubMed] [Google Scholar]

[R65] 65.Johnson BZ, Stevenson AW, Prêle CM, Fear MW, Wood FM, The role of IL-6 in skin fibrosis and cutaneous wound healing. Biomedicine 8, 101 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R66] 66.Meng XM, Nikolic-Paterson DJ, Lan HY, TGF-β: The master regulator of fibrosis. Nat. Rev. Nephrol. 12, 325–338 (2016). [DOI] [PubMed] [Google Scholar]

[R67] 67.Breitkopf-Heinlein K, Meyer C, König C, Gaitantzi H, Addante A, Thomas M, Wiercinska E, Cai C, Li Q, Wan F, Hellerbrand C, Valous NA, Hahnel M, Ehlting C, Bode JG, Müller-Bohl S, Klingmüller U, Altenöder J, Ilkavets I, Goumans MJ, Hawinkels LJ, Lee SJ, Wieland M, Mogler C, Ebert MP, Herrera B, Augustin H, Sánchez A, Dooley S, Ten Dijke P, BMP-9 interferes with liver regeneration and promotes liver fibrosis. Gut 66, 939–954 (2017). [DOI] [PubMed] [Google Scholar]

[R68] 68.Yu WM, Yu H, Chen ZL, Laminins in peripheral nerve development and muscular dystrophy. Mol. Neurobiol. 35, 288–297 (2007). [DOI] [PubMed] [Google Scholar]

[R69] 69.Ridiandries A, Tan JTM, Bursill CA, The role of chemokines in wound healing. Int. J. Mol. Sci. 19, 3217 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R70] 70.Zhu Z, Ding J, Tredget EE, The molecular basis of hypertrophic scars. Burns Trauma 4, 2 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R71] 71.Angulo-Urarte A, van der Wal T, Huveneers S, Cell-cell junctions as sensors and transducers of mechanical forces. Biochim. Biophys. Acta Biomembr. 1862, 183316 (2020). [DOI] [PubMed] [Google Scholar]

[R72] 72.Jones CV, Ricardo SD, Macrophages and CSF-1: Implications for development and beyond. Organogenesis 9, 249–260 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R73] 73.Zhu J, T helper 2 (Th2) cell differentiation, type 2 innate lymphoid cell (ILC2) development and regulation of interleukin-4 (IL-4) and IL-13 production. Cytokine 75, 14–24 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R74] 74.Lim CH, Sun Q, Ratti K, Lee SH, Zheng Y, Takeo M, Lee W, Rabbani P, Plikus MV, Cain JE, Wang DH, Watkins DN, Millar S, Taketo MM, Myung P, Cotsarelis G, Ito M, Hedgehog stimulates hair follicle neogenesis by creating inductive dermis during murine skin wound healing. Nat. Commun. 9, 4903 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R75] 75.Frech S, Forsthuber A, Korosec A, Lipp K, Kozumov V, Lichtenberger BM, Hedgehog signaling in papillary fibroblasts is essential for hair follicle regeneration during wound healing. J. Invest. Dermatol. 142, 1737–1748.e5 (2022). [DOI] [PubMed] [Google Scholar]

[R76] 76.Wang Y, Wang G, Liu H, Tenascin-C: A key regulator in angiogenesis during wound healing. Biomolecules 12, 1689 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R77] 77.Black S, Phillips D, Hickey JW, Kennedy-Darling J, Venkataraaman VG, Samusik N, Goltsev Y, Schurch CM, Nolan GP, CODEX multiplexed tissue imaging with DNA-conjugated antibodies. Nat. Protoc. 16, 3802–3835 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R78] 78.Liu H, Hu J, Zheng Q, Feng X, Zhan F, Wang X, Xu G, Hua F, Piezo1 channels as force sensors in mechanical force-related chronic inflammation. Front. Immunol. 13, 816149 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R79] 79.Chesney J, Bacher M, Bender A, Bucala R, The peripheral blood fibrocyte is a potent antigen-presenting cell capable of priming naive T cells in situ. Proc. Natl. Acad. Sci. U.S.A. 94, 6307–6312 (1997). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R80] 80.Molofsky AB, Savage AK, Locksley RM, Interleukin-33 in tissue homeostasis, injury, and inflammation. Immunity 42, 1005–1019 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R81] 81.Abbasi S, Sinha S, Labit E, Rosin NL, Yoon G, Rahmani W, Jaffer A, Sharma N, Hagner A, Shah P, Arora R, Yoon J, Islam A, Uchida A, Chang CK, Stratton JA, Scott RW, Rossi FMV, Underhill TM, Biernaskie J, Distinct regulatory programs control the latent regenerative potential of dermal fibroblasts during wound healing. Cell Stem Cell 27, 396–412.e6 (2020). [DOI] [PubMed] [Google Scholar]

[R82] 82.Foster DS, Januszyk M, Delitto D, Yost KE, Griffin M, Guo J, Guardino N, Delitto AE, Chinta M, Burcham AR, Nguyen AT, Bauer-Rowe KE, Titan AL, Salhotra A, Jones RE, da Silva O, Lindsay HG, Berry CE, Chen K, Henn D, Mascharak S, Talbott HE, Kim A, Nosrati F, Sivaraj D, Ransom RC, Matthews M, Khan A, Wagh D, Coller J, Gurtner GC, Wan DC, Wapnir IL, Chang HY, Norton JA, Longaker MT, Multiomic analysis reveals conservation of cancer-associated fibroblast phenotypes across species and tissue of origin. Cancer Cell 40, 1392–1406.e7 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R83] 83.Chen K, Kwon SH, Henn D, Kuehlmann BA, Tevlin R, Bonham CA, Griffin M, Trotsyuk AA, Borrelli MR, Noishiki C, Padmanabhan J, Barrera JA, Maan ZN, Dohi T, Mays CJ, Greco AH, Sivaraj D, Lin JQ, Fehlmann T, Mermin-Bunnell AM, Mittal S, Hu MS, Zamaleeva AI, Keller A, Rajadas J, Longaker MT, Januszyk M, Gurtner GC, Disrupting biological sensors of force promotes tissue regeneration in large organisms. Nat. Commun. 12, 5256 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R84] 84.Gordon S, Alternative activation of macrophages. Nat. Rev. Immunol. 3, 23–35 (2003). [DOI] [PubMed] [Google Scholar]

[R85] 85.Gao Y, Luo C, Rui T, Fan Y, Yao Y, Shen H, Gao C, Wang T, Wang H, Chen X, Zhang J, Li D, Xia C, Li LL, Wang Z, Zhang M, Chen X, Tao L, Autophagy inhibition facilitates wound closure partially dependent on the YAP/IL-33 signaling in a mouse model of skin wound healing. FASEB J. 35, e21920 (2021). [DOI] [PubMed] [Google Scholar]

[R86] 86.Oshio T, Komine M, Tsuda H, Tominaga SI, Saito H, Nakae S, Ohtsuki M, Nuclear expression of IL-33 in epidermal keratinocytes promotes wound healing in mice. J. Dermatol. Sci. 85, 106–114 (2017). [DOI] [PubMed] [Google Scholar]

[R87] 87.Amuso VM, Haas MR, Cooper PO, Chatterjee R, Hafiz S, Salameh S, Gohel C, Mazumder MF, Josephson V, Kleb SS, Khorsandi K, Horvath A, Rahnavard A, Shook BA, Fibroblast-mediated macrophage recruitment supports acute wound healing. J. Invest. Dermatol. 10.1016/j.jid.2024.10.609 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R88] 88.Jin M, Komine M, Tsuda H, Sashikawa-Kimura M, Nakae S, Motegi SI, Ohtsuki M, Interleukin-33 deficiency protects the skin from ulcer formation in an ischemia-reperfusion-induced decubitus mouse model. Exp. Dermatol. 33, e70014 (2024). [DOI] [PubMed] [Google Scholar]

[R89] 89.Kotsiou OS, Gourgoulianis KI, Zarogiannis SG, IL-33/ST2 axis in organ fibrosis. Front. Immunol. 9, 2432 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R90] 90.Barron L, Wynn TA, Fibrosis is regulated by Th2 and Th17 responses and by dynamic interactions between fibroblasts and macrophages. Am. J. Physiol. Gastrointest. Liver Physiol. 300, G723–G728 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R91] 91.desJardins-Park HE, Chinta MS, Foster DS, Borrelli MR, Shen AH, Wan DC, Longaker MT, Fibroblast heterogeneity in and its implications for plastic and reconstructive surgery: A basic science review. Plast. Reconstr. Surg. Glob. Open 8, e2927 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R92] 92.Swindle MM, Makin A, Herron AJ, Clubb FJ, Frazier KS, Swine as models in biomedical research and toxicology testing. Vet. Pathol. 49, 344–356 (2012). [DOI] [PubMed] [Google Scholar]

[R93] 93.Masson-Meyers DS, Andrade TAM, Caetano GF, Guimaraes FR, Leite MN, Leite SN, Frade MAC, Experimental models and methods for cutaneous wound healing assessment. Int. J. Exp. Pathol. 101,21–37 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R94] 94.Wong VW, Sorkin M, Glotzbach JP, Longaker MT, Gurtner GC, Surgical approaches to create murine models of human wound healing. J. Biomed. Biotechnol. 2011,969618 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R95] 95.Rabello FB, Souza CD, Farina Junior JA, Update on hypertrophic scar treatment. Clinics 69, 565–573 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R96] 96.Ma F, Tsou PS, Gharaee-Kermani M, Plazyo O, Xing X, Kirma J, Wasikowski R, Hile GA, Harms PW, Jiang Y, Xing E, Nakamura M, Ochocki D, Brodie WD, Pillai S, Maverakis E, Pellegrini M, Modlin RL, Varga J, Tsoi LC, Lafyatis R, Kahlenberg JM, Billi AC, Khanna D, Gudjonsson JE, Systems-based identification of the Hippo pathway for promoting fibrotic mesenchymal differentiation in systemic sclerosis. Nat. Commun. 15, 210 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R97] 97.Ruifrok AC, Johnston DA, Quantification of histochemical staining by color deconvolution. Anal. Quant. Cytol. Histol. 23, 291–299 (2001). [PubMed] [Google Scholar]

[R98] 98.Chen EY, Tan CM, Kou Y, Duan Q, Wang Z, Meirelles GV, Clark NR, Ma'ayan A, Enrichr: Interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinformatics 14, 128 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R99] 99.Ouyang JF, Kamaraj US, Cao EY, Rackham OJL, ShinyCell: Simple and sharable visualization of single-cell gene expression data. Bioinformatics 37, 3374–3376 (2021). [DOI] [PubMed] [Google Scholar]

[R100] 100.Bergen V, Lange M, Peidli S, Wolf FA, Theis FJ, Generalizing RNA velocity to transient cell states through dynamical modeling. Nat. Biotechnol. 38, 1408–1414 (2020). [DOI] [PubMed] [Google Scholar]

[R101] 101.Chen K, Vigliotti A, Bacca M, McMeeking RM, Deshpande VS, Holmes JW, Role of boundary conditions in determining cell alignment in response to stretch. Proc. Natl. Acad. Sci. U.S.A. 115, 986–991 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Inhibiting mechanotransduction prevents scarring and yields regeneration in a large animal model

Shamik Mascharak

Michelle Griffin

Heather E Talbott

Jason L Guo

Jennifer Parker

Annah Grace Morgan

Caleb Valencia

Maxwell Michael Kuhnert

Dayan J Li

Norah E Liang

Rachel M Kratofil

Joseph A Daccache

Ikjot Sidhu

Michael F Davitt

Nicholas Guardino

John M Lu

Darren B Abbas

Nestor M D Deleon

Christopher V Lavin

Sandeep Adem

Anum Khan

Kellen Chen

Dominic Henn

Amanda Spielman

Asha Cotterell

Deena Akras

Mauricio Downer Jr

Ruth Tevlin

H Peter Lorenz

Geoffrey C Gurtner

Michael Januszyk

Shruti Naik

Derrick C Wan

Michael T Longaker

Abstract

INTRODUCTION

RESULTS

YAP inhibition promotes wound regeneration in red Duroc pigs

Fig. 1. Gross, histologic, and tissue mechanical analysis of the effects of verteporfin treatment on wounds.

Local treatment with verteporfin is safe and effective for promoting regeneration in a porcine wounding model

Transcriptomic profiling defines scarring and regenerating porcine wound fibroblast subpopulations

Fig. 2. Single-cell transcriptomic and fibroblast subtype analyses of porcine wounds.

In vitro mechanomodulation reveals mechanoresponsive fibroblast subpopulation dynamics

Regeneration- and fibrosis-related cell-cell signaling is differentially enriched in verteporfin versus control wounds

Immune cell transcriptomic analysis reveals distinct inflammatory dynamics in fibrotic and regenerating wounds

Fig. 3. Single-cell transcriptomic and myeloid subtype analyses of porcine wounds.

Spatial proteomic analysis supports divergent fibroblast dynamics in scarring and regenerating wounds

Fig. 4. Time course CODEX analysis of healing porcine wounds.

YAP inhibition promotes wound regeneration in a human foreskin xenograft model

Fig. 5. YAP and IL-33 inhibition in a human skin xenograft excisional wound model without closure.

IL-33 inhibition validates its functional role in healing of verteporfin-treated wounds

Human healing after verteporfin treatment parallels cellular responses in regenerating porcine wounds

Fig. 6. Single-cell transcriptomic analysis of YAP inhibition in a human skin xenograft wound model.

DISCUSSION

MATERIALS AND METHODS

Study design

Statistical analysis

Supplementary Material

Acknowledgments:

Funding:

Footnotes

Data and materials availability:

REFERENCES AND NOTES

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases