Hepatocyte growth factor in biosheets promotes autonomous regeneration of cutaneous tissue after transplantation onto a full-thickness skin defect

Abstract

Biosheets and biotubes are collagenous connective tissue membranes that surround unabsorbable molds embedded in living tissues, and can regenerate damaged tissues in various organs following in-situ transplantation. However, the mechanisms underlying tissue regeneration are still unclear. In this study, we analyzed the histological features and cellular composition in the biosheets, and observed dramatic time-dependent changes. We transplanted biosheets into full-thickness wounds in a mouse model, and evaluated skin regeneration through histological analysis in comparison with subcutaneous fascia transplantation. Our findings confirm that biosheets can accelerate epithelialization of skin wounds of mouse more effectively than fascia. To elucidate the mechanism underlying skin regeneration by biosheets, we compared the protein expression in the biosheets and subcutaneous fascia. Proteomics analysis revealed significant differences in the protein profiles of the biosheets and fascia, with marked elevation of hepatocyte growth factor (HGF) in the biosheets. Addition of HGF in situ accelerated angiogenesis and epithelial regeneration in the fascia-transplanted full-thickness wounds in mice. On the other hand, neutralization of HGF impaired the wound healing process following biosheet transplantation. In conclusion, our findings suggest that HGF mediates biosheet-mediated regeneration of cutaneous tissues by contributing to re-epithelialization and angiogenesis.

Graphical abstract

Highlights

• •Biosheets have been investigated as scaffolds for cutaneous tissue regeneration.
• •Biosheets accelerated cutaneous regeneration in a mouse model.
• •Proteomic analysis of biosheets showed elevation of HGF, PDGF-β, TGF-β, and Cdk1.
• •Neutralization of HGF in biosheets by anti-HGF antibodies impaired cutaneous regeneration.

1. Introduction

Regenerative medicine can effectively treat intractable diseases by replacing injured or congenitally malfunctional organs, and may prevent life-long disability. Foreign bodies implanted in tissues are encapsulated in a provisional matrix of fibrin, cytokines and chemokines, which attracts and activates inflammatory cells such as polymorphonuclear leukocytes, mast cells and macrophages that degrade or phagocytose the foreign bodies. The foreign body giant cells resulting from the fusion of these macrophages and fibroblasts eventually form a fibrous capsule [1,2]. Based on this principle, the in-body tissue architecture (iBTA) method was established for fabricating a biomaterial in a living body for tissue regeneration [3]. Briefly, a mold made of unabsorbable materials such as silicone and acryl is implanted into a subcutaneous pouch. Within one month of implantation, the mold is capsulized with a membrane composed of collagenous connective tissue (Fig. 1A). The major advantage of iBTA is that the shape or thickness of the capsule membranes can be easily controlled by tuning the shape of the molds. Connective tissue membranes of various shapes have been developed using iBTA, such as tubular connective tissues or biotubes [3,4], tri-leaflet tissues or biovalves [5], and sheet-shaped tissues or biosheets [6,7]. Biotubes were first applied to in-situ blood vessel engineering in a rabbit model. The biotubes anastomosed to rabbit carotid arteries regenerated blood vessel wall tissues, including epithelium, collagen, elastic fibers and smooth muscles [8]. Several types of biotubes have been developed in recent years, such as long biotubes [9], small-caliber biotubes [10,11], and thick-walled biotubes [12,13]. The long-term pliability of these biotubes in blood vessels has been confirmed in animal models [11,14]. In addition, there are some recent case reports of applying biotubes in human subjects [15,16].

Fig. 1.

Preparation of biosheets (A) Schematic illustration of in-body tissue architecture (iBTA). (B) Implantation of silicon molds into a dorsal subcutaneous pouch of mouse and fabrication of biosheet. (C) Representative images of HE-stained full-thickness dorsal skin showing the different layers. (D) Implantation of the biosheet mold into the fascia layer. The rectangle with dashed lines indicates the position of the mold. (E) Representative images of the superficial sides of 1w-, 2w-, and 4w-biosheets with HE and Masson Trichrome (MT) staining. Dense layers and sparse layers are indicated by blue and black arrows respectively. (F, G) The thickness of the dense layer (F) and sparse layer (G) on the superficial side of 1w-, 2w-, and 4w-biosheets. (H) Representative images of the deep sides of the biosheets with HE and MT staining. (I, J) The thickness of the dense layer (I) and sparse layer (J) on the deep side of 1w-, 2w-, and 4w-biosheets. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Biotubes or biosheets can also induce tissue regeneration in the cornea [7], esophagus [17,18], bladder [19], abdominal wall [20,21], etc. We successfully treated congenital tracheal stenosis in a rabbit model through autologous transplantation of biosheets to a rectangular tracheal defect. The cartilage and epithelium of the injured trachea were regenerated on the biosheet 4 weeks post-implantation [6]. In addition, we have also engineered tubular trachea using a silicon mold incorporating rings of collagen sponges that were seeded with chondrocytes. A biotube with articular rings, called BIO-AIR-TUBE, was formed after embedding this mold in the dorsal subcutaneous pouches of rabbits [22]. Finally, we used biosheets to repair damaged diaphragm in a rabbit model, and detected autonomous regeneration of striated muscular tissues 3 months post-implantation [23]. Altogether, connective tissue membranes made with iBTA can broaden the applicability of regenerative therapy to various organs or diseases, as well as pediatric surgery. Biotubes and biosheets can be easily fabricated without specialized techniques or expensive materials, and have a low risk of immunological reactions because they are derived from autologous or allogeneic tissues.

The skin is the largest organ of the human body and serves essential physiological functions, such as maintenance of hydration, barrier protection against chemicals and pathogens, excretion, and thermoregulation. Therefore, certain skin disorders, such as severe burns, chronic ulcers, or genetic conditions, can lead to severe infections, excessive fluid loss, impaired thermoregulation [24]. Tissue engineering can be used to treat various skin conditions, including burns, injuries, surgical scars [24,25], and age-related changes [26]. Furthermore, regeneration therapy has also been applied to pediatric patients for treating congenital cutaneous diseases such as giant nevi [25] and junctional epidermolysis bullosa [27]. Several biomaterials have been developed for wound dressings [28,29], and have demonstrated their ability to promote wound healing and cutaneous tissue regeneration [30,31]. Biosheets are also applicable for cutaneous tissue engineering. Maeta et al. reported that biosheets accelerated wound healing in a canine skin defect model [32]. In addition, biosheets have been shown to promote wound healing in patients with severe diabetic foot ulcers [33].

However, the mechanisms underlying cutaneous tissue regeneration in the biosheets are unclear. Maeta et al. found that biosheets expressed Nanog and contained cells positive for SSEA and CD105, which are markers for mesenchymal stem/stromal cells (MSCs). They also identified several cytokines in the biosheets, including vascular endothelial growth factor (VEGF), transforming growth factor (TGF), and interleukin (IL)-4 [32]. However, the functional relevance of these cells and cytokines in cutaneous tissue regeneration have not been adequately evaluated. It is necessary to elucidate the mechanisms of biosheet-mediated skin regeneration in order to develop biosheets with enhanced regenerative capacity, and accelerate their clinical application in skin regeneration.

In this study, we aimed to identify factors involved in biosheet-mediated skin regeneration to clarify the underlying mechanisms. We first analyzed the histological and cellular characteristics of biosheets. Previous studies have demonstrated that biosheets can support the regeneration of cartilage [6] and muscular tissues [18,20] can be regenerated using biosheets, and several reports have suggested the possible presence of MSCs within biosheets [32,34]. Therefore, we hypothesized that tissue regeneration in the biosheets is driven by MSCs [35]. For mechanistic studies, we established a novel mouse model of biosheets-mediated skin regeneration, and found that the biosheets accelerated regeneration of cutaneous tissues in full-thickness skin wounds. To identify factors involved in skin regeneration, we performed a comparative proteomic analysis of biosheets and fascia, and validated the function of the identified factor in biosheet-mediated regeneration of cutaneous tissues.

2. Results

2.1. Structural characteristics of the biosheets

We analyzed the temporal histological changes around the silicone molds to monitor the formation of biosheets. In addition, biosheets harvested 1, 2, and 4 weeks after implantation were compared with the subcutaneous fascia [36] (Fig. 1B–J). As biosheets were derived from fascia following mold implantation, we hypothesized that the factors associated with tissue regeneration could be identified by comparing the biosheets with fascia. Prior to mold implantation, the fascia consisted of sparse fibrous tissue and cells (Fig. 1C). One week after implantation, cells accumulated on the inner surface of the biosheets and gradually increased in number (Fig. 1E–H). We also detected giant cells with multiple nuclei (Fig. S1), which were most likely formed to phagocytose or degrade the mold. Two histologically distinct layers were visible in the wall of the biosheet (Fig. 1D, E, H), which emerged on the superficial as well as the deep side of the implanted molds. The thin, deeply stained inner layer was composed of densely packed cells and collagen fibers, and gradually thickened over time (Fig. 1E, F, H, I). In contrast, the outer layer had sparse cells and fibers that stained lightly, and its thickness remained unchanged throughout the observation period (Fig. 1E–G, H, J). We defined the continuous tissue from the dense layer to the sparse layer as a biosheet (Fig. 1D).

The structural features of the biosheets were analyzed by transmission electron microscopy (TEM). Prior to implantation, the fascia consisted of bundled collagenous fibers and scattered cells such as fibroblasts (Fig. 2A). However, the shape, density, and alignment of cells and collagen fibers underwent significant changes during the formation of biosheets. Furthermore, the number of fibroblasts, macrophages and neovascular vessels increased following implantation (Fig. 2B–E), which is consistent with the immune response to foreign bodies. In line with the findings of hematoxylin-eosin (HE) and Masson Trichome (MT) staining (Fig. 1E–H, S1), TEM analysis also showed significant accumulation of cells on the inner surface of the biosheets (Fig. 2B–ii, C, D-iii, E-i), including that of macrophages with numerous lysosomes (Fig. 2D–iii). The cells adhered to each other and formed a layered structure over time, which was particularly noticeable in the 4w-biosheets (Fig. 2C–E-i). In addition, cells with multiple nuclei were also detected in 4w-biohseets (Fig. 2B–D-iii), which may correspond to foreign body giant cells seen in the histological examination (Fig. S1). These findings are indicative of phagocytosis around the implanted mold.

Fig. 2.

TEM analysis of biosheets

(A) Representative transmission electron micrographs of the fascia. (B) Superficial side of the biosheets. B-i: overall images, B-ii: low-magnification images. The broken lines in (B-i) indicate the border between the dense layer and the sparse layer. (C) Deep side of the biosheets. C-i: overall images, C-ii: low-magnification images. The blue arrows indicate dense layers and the red arrows indicate the sparse layer. (D, E) High-magnification images of each layer on the superficial side (D) and deep side (E) of biosheets. D-ⅰ: sparse layer, D-ⅱ: dense layer, D-ⅲ: inner surface, E-ⅰ: inner surface, E-ii: dense layer. The white arrowheads indicate fibroblasts and black arrowheads indicate macrophages. The broken lines in (D-iii) and (E-i) indicate the surface of biosheets. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

The TEM images also distinguished the dense layer and sparse layer on the superficial side of the biosheets (Fig. 2B). The dense layer had tightly packed cells, including fibroblasts and macrophages (Fig. 2D–ii), and collagenous fibers aligned in the intracellular spaces (Fig. 2B). Although the density of the fibers and cells appeared to be consistent across the 1w-, 2w-, and 4w-biosheets, the thickness of this layer gradually increased with time (Fig. 2B). The sparse layer had fewer collagenous fibers and cells (fibroblasts and macrophages) compared to that in the dense layer. Furthermore, the fibroblasts in the 2w-biosheets and 4w-biosheets were flatter and had longer processes that were more parallel to the surface of the skin or the mold compared to the cells in 1w-biosheets. Likewise, the density of the collagenous fibers was also significantly higher in the 2w- and 4w-biosheets compared to that in 1w-biosheets (Fig. 2B–ii, D-i).

In contrast to the findings of HE and MT staining, TEM analysis could only identify the dense layer and not the sparse layer on the deep side of the biosheets (Fig. 2C). The different procedures used to prepare the specimens for optical and electron microscopy may have influenced the microscopic structures. Neither the density nor the thickness of the dense layer changed with time, although the fibroblasts became flatter and aligned more parallel to the skin surface (Fig. 2C–E-ii).

2.2. Analysis of cell surface markers and multipotency of stromal cells in biosheets

To explore the cellular landscape of the biosheets, we digested the biosheets and fascia using collagenase, and analyzed the number and composition of cell types. In particular, we isolated and characterized the stromal cells from biosheets, and characterized them based on surface marker expression and multipotent differentiation.

The number of cells in the biosheets was ten-fold higher than that in the fascia, and the highest number of cells was detected in 1w-biosheets (Fig. S2A). The surface markers and differentiation capacity of the stromal cells isolated from biosheets and fascia were subsequently analyzed. The expression of MSC-specific markers, including Sca-1, CD29, CD44, CD73, CD90, CD105 and CD146, were examined on the CD31⁻CD45⁻ stromal cells by flow cytometry [[37], [38], [39], [40], [41]]. Although cell populations positive for Sca-1, CD90, CD29 or CD44 (83.5%, 82.5%, 91.4%, 38% respectively) were detected, almost all stromal cells were negative for CD73, CD105 and CD146 (Fig. S2B). We identified a population of stromal cells that were positive for Sca-1, CD90, CD29 and CD44, and these quadruple-positive cells were considered potential MSC candidates (Fig. 3A). The percentage of CD31⁻CD45⁻ stromal cells appeared to be higher in the fascia than in the biosheets, but the difference did not reach statistical significance (Fig. S2C). The percentage of quadruple-positive cells in the CD31⁻CD45⁻ stromal cells in the 1w-, 2w- and 4w-biosheets were 64.2 ± 0.8%, 57.5 ± 2.3% and 37.6 ± 3% respectively, compared to the significantly lower 4.8 ± 2.3% in the fascia (p < 0.001; Fig. 3B).

Fig. 3.

Analysis of surface markers and multipotency of stromal cells in biosheets

(A) Representative flow cytometry profiles of cells isolated from 2w-biosheets and identification of a population of quadruple-positive cells (P7). (B) The ratio of quadruple-positive cells (P7) to stromal cells (P5) in the biosheets and fascia. (C-E) Representative images of alkaline phosphatase (ALP) staining and Alizarin red staining (C), oil red staining (D), Toluidine blue (TB) staining and Safranine O (SO) staining (E) of stromal cells isolated from biosheets and subjected to osteogenic, adipogenic and chondrogenic differentiation. (F-H) Expression levels of osteogenic (F), adipogenic (G), and chondrogenic (H) marker genes. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

To analyze the multipotency of the stromal cells in biosheets, we sorted the CD31⁻CD45⁻fraction from the biosheets by flow cytometry, and induced their differentiation to the osteoblast, adipocyte and chondroblast lineages. Alkaline phosphatase (ALP) staining and Alizarin red staining showed the presence of osteoblastic cells and calcium deposition in the stromal cells cultured in the osteogenic differentiation medium (Fig. 3C). Likewise, Oil red staining also confirmed the presence of adipocyte-like cells with lipid droplets following adipogenic differentiation (Fig. 3D). Although the pellets cultured in the differentiation medium showed deeper Toluidine blue and Safranine O staining compared to those cultured in the control medium, the typical structure of cartilage could not be identified (Fig. 3E). The expression levels of the different lineage marker genes were elevated after culturing the cells in the respective differentiation media (Fig. 3F–H). Taken together, these findings may indicate the presence of MSC-like cells, although their presence could not be confirmed in this study.

2.3. Establishment of a mouse model of skin regeneration using biosheets

To explore the mechanism of cutaneous tissue regeneration after biosheet transplantation, we established a mouse model of skin regeneration using biosheets or fascia. Briefly, a patch of 2w-biosheet or fascia derived from an EGFP (enhanced green fluorescent protein) transgenic mouse were transplanted into full-thickness skin wounds (6 mm in diameter) on the dorsum of wild-type mice (Fig. 4A). The grafts and wounds were covered with polyurethane foam dressing to prevent them from drying (Fig. 4B). In this experiment, biosheets and fascia derived from EGFP transgenic mouse were used for GFP (green fluorescent protein) immunohistochemical staining to determine whether biosheet- or fascia-derived cells could survive within the wound and potentially contribute to cutaneous tissue regeneration. In addition, we used 2w-biosheets rather than 1w- or 4w-biosheets based on the surmise that earlier-phase biosheets would be more suitable for tissue regeneration, as foreign body reactions are more active during the early phase, resulting in abundant cells and growth factors involved in tissue repair. Consistent with this hypothesis, the number of Sca-1⁺CD90⁺CD29⁺CD44⁺ quadruple-positive cells identified in flowcytometry analysis was higher in 1w- and 2w-biosheets than in the 4w-biosheets (Fig. 3B). However, 1w-biosheets were too thin and fragile for stable handling and transplantation into the wounds, whereas histological analysis revealed that 2w- and 4w-biosheets had thicker and more rigid structures (Fig. 1, Fig. 2). Therefore, considering the balance between regenerative potential and mechanical stability for stable handling, the 2w-biosheets were selected for this experiment. Cutaneous regeneration was analyzed 7- and 14-days post-wounding. On day 7, the unclosed wound areas were not s significantly different between the two groups (26.9 ± 3.1% vs 36.3 ± 4.4%, p = 0.16). However, almost all biosheet-treated wounds were completely closed on day 14, while wounds treated with fascia were only partly closed (0.36 ± 0.23% vs 22.2 ± 6.9%, p = 0.03) (Fig. 4C and D). In addition, histological analysis showed formation of thick tissues in the biosheet-treated wounds on day 7, along with epithelial regeneration and the presence of dense subcutaneous tissues at the wound margins. On the other hand, only sparse and thin tissues were observed in the wounds treated with fascia, and epithelial regeneration was minimal (Fig. S3). On day 14 post-wounding, the surface of the biosheet-treated wounds was completely covered with regenerated epithelium, and thick subcutaneous tissues with dense collagen fibers and closely arranged cells were visible. However, fascia-treated wounds showed non-epithelialized areas in the middle, and the underlying tissues were thin with few collagen fibers and cells (Fig. 4E). Furthermore, we detected GFP-positive biosheet-derived cells and vWF (von Willebrand factor)-positive capillary blood vessels in the subcutaneous tissues of biosheet-treated wounds, while fascia-treated wounds exhibited neither GFP-positive fascia-derived cells nor angiogenesis (Fig. 4F).

Fig. 4.

Establishment of a mouse model of skin regeneration using biosheets

(A) Schematic representation of the mouse model of skin regeneration. Briefly, 2w-biosheets or facia derived from EGFP transgenic mice were transplanted into full thickness skin wounds in C57BL/6J mouse. (B) The postoperative dressing on the wounds. (C) Macroscopic images of the wounds on days 7 and 14 post-wounding. (D) The unclosed wound area on days 7 and 14 post-wounding following implantation of biosheets or fascia. (E) Representative images of HE-stained dorsal skin sections on day 14 post-wounding. (F) High-magnification views of biosheet and fascia from the images in (E), and images of MT staining and immunostaining for vWF and GFP. The arrowheads indicate the regenerated epithelium.

Taken together, these findings suggest that biosheets can effectively regenerate the cutaneous epithelium compared to fascia. The implanted biosheets promoted wound epithelialization by inducing formation of granulation tissue, and activated neovascularization in the subcutaneous tissues. As GFP-positive cells were detected in the biosheet-treated wounds, we speculated that biosheet-derived cells and proteins contribute to cutaneous tissue regeneration. In contrast, GFP-positive cells were absent in fascia-treated wounds, indicating that fascia-derived cells did not survive and likely did not play a role in regeneration. Given that murine proteins and cells are well-characterized, we used biosheets derived from mice on a mouse model to explore the mechanism of biosheets-induced tissue regeneration.

2.4. HGF levels increased significantly during the formation of biosheets

We also performed proteomics analysis of the biosheets and fascia to identify proteins involved in biosheet-driven tissue regeneration. A total of 9253 proteins were identified in the biosheets and fascia, and their proteomic signatures showed substantial differences (Fig. 5A and B). As shown in Figs. 5C and 1915 proteins were significantly upregulated (fold change >2 and p < 0.05) and 1099 proteins were downregulated (fold change <0.5 and p < 0.5) in the 2w-biosheets compared to the fascia. Likewise, the levels of 1373 proteins increased and that of 1057 proteins decreased in the 4w-biosheets (Fig. 5D). We identified a cluster of 2979 proteins that was elevated in both 2w- and 4w-biosheets compared to the fascia (Fig. 5A). Gene ontology (GO) analysis of proteins in this cluster revealed significant enrichment of terms related to RNA, translation and ribosome, suggesting activation of protein synthesis in the biosheets (Fig. S4A–C). In KEGG pathway analysis, pathways related to leukocyte, phagocytosis, leukocyte and TNF signaling pathway were highly ranked, which indicates that biosheets may be fabricated through foreign body reaction. Pathways that promote angiogenesis, such as “proteoglycans in cancer,” the mTOR signaling pathway, and the HIF-1 signaling pathway, were also enriched (Fig. S4D). Upstream analysis is an analytical approach used to identify upstream regulators that may explain the observed changes in protein expression [42]. The key proteins regulating the cluster of 2979 proteins were further identified through this upstream analysis. As shown in Fig. 5E, several cytokines associated with the anti-inflammatory effect (IL-10, TGF-β) [43,44], and those associated with cellular proliferation, migration and angiogenesis (PDFG-β, HGF) [45] were among the top candidates. Some of these top-ranked upstream factors were detected in the original liquid chromatography–mass spectrometry (LC–MS) proteomic dataset, of which Cdk1 exhibited the highest fold change compared with fascia (2w-biosheets: fold change, 54.8; p = 0.0001; 4w-biosheets: fold change, 26.0; p = 0.0005). Cdk1 is a key regulator of the cell cycle and plays an essential role in cell proliferation [46]. Therefore, its marked upregulation may reflect active proliferation of fibroblasts and inflammatory cells during biosheet formation. HGF was also notably elevated in biosheets, showing the second highest fold change compared to the fascia (2w-biosheets: fold change, 17.7; p = 0.0025; 4w-biosheets: fold change, 15.5; p = 0.0036) (Fig. 5C–E). Given its role in tissue regeneration, as well as the therapeutic potential in regenerative medicine [[47], [48], [49], [50], [51], [52], [53], [54]], the role of HGF in biosheets was examined further. First, the increase in HGF levels in the biosheets relative to the fascia was validated by ELISA. As shown in Fig. 5F, HGF levels were almost 10-fold higher in the biosheets compared to the fascia. The distribution of HGF in the biosheets was also confirmed by immunostaining (Fig. 6; Fig. S5). An HGF-positive layer was detected on the inner surface of the biosheets, which was thin in the 1w-biosheets but gradually thickened with time (Fig. 6A, B, D, E). In addition, HGF-positive cells were also present in the outer sparse layer of the biosheets, especially on the deeper side (Fig. 6D and E; Fig. S5A and B). Furthermore, the number of HGF-positive cells were significantly higher in the biosheets than in the fascia, wherein almost all cells were negative for HGF (Fig. 6B–F; Fig. S5C). The H scores of HGF immunostaining were also significantly higher in the biosheets than in the fascia (Fig. 6C). The inner surface of biosheets were also probed for the HGF receptor MET, and MET-positive cells were detected in the outer loose layer (Fig. 6G and H). In contrast, MET-positive tissues nor cells could not be detected in the fascia (Fig. 6I). Taken together, HGF levels were significantly higher in the biosheets compared to the fascia, indicating a potential role in tissue regeneration.

Fig. 5.

Proteomic analysis of biosheets and analysis of upstream factors

(A) Heatmap showing the differentially expressed proteins among fascia, 2w-biosheets and 4w-biosheets. (B) Heatmap of Pearson correlation between fascia, 2w-biosheets and 4w-biosheets. (C, D) Volcano plots depicting protein variance between 2w-biosheets and fascia (C) and between 4w-biosheets and fascia (D). E) The top 20 proteins in the biosheets identified by upstream analysis among 2979 proteins clustered in (A). Asterisks indicate proteins that were detected in the original LC–MS proteomic dataset, and their fold changes (FC) and p values (p) relative to fascia are shown. (F) HGF levels in the fascia and biosheets.

Fig. 6.

Expression of HGF and MET in biosheets and the regenerated cutaneous tissue

(A) Representative image of HGF immunostaining of 2w-biosheet. The mold is indicated by a rectangle with a dashed line. (B, C) The number of HGF-positive cells (B) and H scores of HGF immunostaining (C) in each biosheet and fascia. (D-F) Representative images of HGF immunostaining in the (D) superficial side and (E) deep side of the biosheets and (F) fascia. Arrowheads indicate HGF-positive layers. (G-I) Representative images of MET immunostaining in the (G) superficial side and (H) deep side of the biosheets and (I) fascia. Arrowheads indicate MET-positive layers and arrows indicate MET positive cells. (J) Representative images of HGF immunostaining in the dorsal skin of mice on day 14 post-wounding. (K, L) High-magnification views of blue boxed areas in (J). (K) HGF and (L) MET immunostaining. The areas marked with dashed lines indicate HGF or MET-positive tissues. (M) The number of HGF-positive cells in the dorsal skin of mice on day 7 and day 14 post-wounding. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

2.5. HGF is a key factor of skin regeneration using biosheets

To confirm the role of HGF in skin regeneration, we analyzed its expression in the specimens of cutaneous tissues resected from the mouse model. HGF-positive tissues were only detected in the biosheet-treated wounds and not in the fascia-treated wounds on day 7, although HGF-positive cells were observed in both wounds (Fig. S6A and B). Furthermore, only MET-positive cells were detected at this timepoint, and both types of wound tissues were negative for MET (Fig. S6C). On day 14 post-wounding, subcutaneous tissues underlying the regenerated epithelium in the biosheet-treated wounds showed more intense staining for HGF compared to the day 7 specimens (Fig. 6J and K). Furthermore, these tissues were also positive for MET (Fig. 6L). On the other hand, the fascia-treated wounds were negative for both HGF and MET even on day 14 post-wounding (Fig. 6J–L). The number of HGF-positive cells was significantly higher in the biosheet-treated wounds on day 7 and showed a slight increase toward day 14 post-wounding, although the difference was not statistically significant (Fig. 6M). These results suggested that HGF-positive cells and tissues emerge following biosheet implantation, and accelerate cutaneous tissue regeneration.

To determine whether HGF supplementation can promote cutaneous wound healing after transplantation of biosheet, we administrated recombinant murine HGF (rmHGF) to fascia-transplanted wounds on day 0, and analyzed the indicators of skin regeneration (Fig. 7A). On day 14 post-wounding, the unclosed area was significantly smaller in the HGF-treated wounds compared to the control wounds (19.3 ± 2.9% vs 35.8 ± 5%, p = 0.02; Fig. 7B and C). Deeply stained subcutaneous tissues appeared at the margin of the HGF-treated wounds on day 14 (Fig. 7E and F), which was similar to that observed in the biosheet-treated wounds. The area of deeply stained subcutaneous tissues was significantly increased in the HGF-treated wounds (Fig. 7D). Moreover, vWF-positive neovascular vessels were also observed in areas corresponding to the deeply stained subcutaneous tissues (Fig. 7G), but not detected outside this region, suggesting that HGF promoted neovascularization. Thus, local administration of HGF had a similar therapeutic effect on full-thickness skin wounds as the implanted biosheets.

Fig. 7.

HGF is a key factor driving skin regeneration by biosheets

(A-F) Administration of recombinant murine HGF in the skin regeneration model.

(A) Schematic illustration of administering recombinant HGF in the mouse model of skin regeneration. (B) Macroscopic images of the wounds on day 14 post-wounding. Broken lines indicate the area of unclosed wounds. (C) The unclosed wound area on day 14 post-wounding. (D) Deeply stained areas (surrounded by yellow broken lines in (E)) on day 14 post-wounding. (E) Representative images of HE-stained dorsal skin tissues on day 14 post-wounding. Yellow broken lines indicate the deeply stained subcutaneous tissues. (F, G) High-magnification views of (E) HE staining and (F) vWF immunostaining. Arrowheads indicate vWF-positive capillary blood vessels.

(H-M) In-situ neutralization of HGF in the skin regeneration model using anti-HGF antibody.

(H) Schematic illustration of the experiment. (I) Macroscopic images of the HE-stained wounds on day 7. Black arrowheads indicate the regions corresponding to epithelial regeneration and yellow broken lines indicate the deeply stained subcutaneous tissues. (J, K) High-magnification views of (J) HE staining and (K) vWF immunostaining. Arrowheads indicate neovascular vessels. (L) The length of the regenerated epithelium on day 7 post-wounding as measured using histological images. (M) Deeply stained areas (surrounded by yellow broken lines in (I)) on day 7 post-wounding. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

To further determine whether inactivation of HGF in the biosheets would delay epithelialization, we treated the transplanted biosheets with anti-HGF antibody. In situ administration of the anti-HGF antibody decreased the number of phosphorylated MET-positive cells in the biosheets, indicating that this antibody can suppress the physiological effects of HGF in the skin regeneration model (Fig. S7A–C). As shown in Fig. 7H, the anti-HGF antibody was locally injected into the implanted biosheets on days 0, 1, 3 and 5 post-wounding in order to neutralize the HGF that was continuously secreted from the biosheets. In addition, to prevent absorption of the antibody into the polyurethane foam dressing, the wounds were covered only with film dressing. Given that the wounds treated with either anti-HGF antibody or control IgG had completely epithelialized by day 14, we compared skin regeneration between the two groups on day 7 (Fig. 7H and I). The surface of the wounds had dried in the absence of polyurethane foam dressing, which made it difficult to identify the epithelialized area macroscopically. Therefore, wound closure and epithelialization were evaluated by histological measurement of the length of regenerated epithelium using HE staining, according to previously reported methods [53,55,56], and were further confirmed by immunostaining for pan-cytokeratin (CK). The regenerated epithelium was defined as the stratified epithelial structure (Fig. S7D and E). The regenerated epithelium at the wound margin was significantly shorter in wounds treated with anti-HGF antibody compared to that in the control IgG-treated wounds (Fig. 7I–L). Moreover, HGF neutralization impaired the formation of a dense underlying tissue at the wound margins after biosheet transplantation (Fig. 7I–M), and the subcutaneous tissue was relatively sparse. The site of vWF-positive neovascularization was located within these dense underling tissues and was absent outside the deeply stained area, suggesting that anti-HGF antibody also reduced angiogenesis in the subcutaneous tissues (Fig. 7J and K). Thus, HGF blockade inhibited epithelialization and neovascularization of wounds after biosheet implantation, indicating that HGF is a key factor inducing skin regeneration by the biosheets.

3. Discussion

In this study, we demonstrated that biosheets promoted rapid cutaneous tissue regeneration in full-thickness wounds in mice, and identified the cellular and molecular factors that likely contribute to biosheet-mediated tissue regeneration. Compared to the fascia, the biosheets surrounding the implanted molds showed drastic changes in the alignment of collagen fibers and fibroblasts, which may influence cellular migration and contribute to tissue regeneration. Furthermore, we also identified a distinct population of stromal cells, termed quadruple-positive cells, that expressed four MSC-specific markers: Sca-1, CD90, CD29, and CD44. Although their multipotency could not be confirmed in this study, these findings raise the possibility that MSCs migrate to the biosheets after mold implantation and contribute to tissue regeneration. Proteomic analysis revealed significant changes in the expression of various proteins between the biosheets and fascia, and HGF was identified as a key upstream factor of the upregulated proteins. In addition, HGF-positive layers and cells were detected in the biosheets. In the mouse model of skin regeneration, both HGF and MET were upregulated in the subcutaneous tissues at the biosheet implantation site. The MET-positive cells in the subcutaneous tissues were likely derived from the recipient mouse, since their number appeared to be greater than that of biosheet-derived GFP-positive cells on day 14 post-wounding. These findings suggested that HGF present in the biosheets promoted migration of MET-positive cells to the wound during the healing process. Moreover, the presence of vWF-positive neovascular vessels in this subcutaneous region further indicated that biosheet-derived HGF promoted regeneration indirectly by inducing angiogenesis, as well as directly by activating the epithelial cells. Furthermore, in situ administration of recombinant murine HGF enhanced angiogenesis and regeneration, while antibody-mediated neutralization of HGF impaired the healing process.

HGF is secreted by mesenchymal cells and activates the PI3K-Akt, Ras-MAPK, and STA3 pathways in epithelial cells [57]. It promotes cellular proliferation and migration, survival, and angiogenesis [47], and also exerts anti-inflammatory and anti-fibrotic effects [48]. The therapeutic effects of HGF in tissue regeneration have been demonstrated in animal models and human subjects. Intramuscular injection of a plasmid expressing human HGF into a rabbit model of hindlimb ischemia promoted angiogenesis and blood flow [49]. In randomized clinical trials on patients with critical limb ischemia, HGF-expressing plasmid reduced ischemic pain and increased the rate of complete ulcer healing [50]. Furthermore, intrathecal administration of HGF-expressing viral vector or recombinant HGF in animal models of spinal cord injury reduced the area of damage by promoting neuronal survival, angiogenesis and axonal regeneration, resulting in enhanced functional recovery [51]. A phase I/II clinical trial showed that intrathecal administration of HGF improved the recovery of motor function after spinal cord injury [52]. HGF has also been implicated in re-epithelialization and neovascularization during wound healing [53,54]; therefore, we focused on HGF as a potential factor contributing to biosheet-mediated cutaneous tissue regeneration. In our mouse model of skin regeneration with biosheets, HGF likely induced cutaneous regeneration by promoting angiogenesis in the subcutaneous tissue, along with epithelialization through its pro-proliferative, anti-inflammatory and anti-fibrotic effects.

In addition, cellular factors are also considered to play critical roles in biosheet-mediated cutaneous tissue regeneration. The presence of biosheet-derived GFP-positive cells on day 14 post-wounding (Fig. 4F) may suggested their involvement in skin regeneration. In this study, we identified the population of quadruple-positive stromal cells in the biosheets that expressed four surface markers -Sca-1, CD90, CD29 and CD44⁻which raise the possibility of the presence of MSCs within the biosheets. MSCs differentiate into cells comprising mesenchymal tissues such as bone, cartilage, adipose tissue, muscle and so on [35]. However, studies on animal models have shown that MSCs do not necessarily engraft or replace tissues after transplantation, but act on certain cell types in a paracrine manner by secreting growth factors, cytokines and extracellular vesicles, or interact directly with the cells through tunneling nanotubes. Furthermore, MSCs promote tissue repair and regeneration after injury and in fibrotic diseases by modulating immune cells, promoting cell survival, and inhibiting inflammation and fibrosis [[58], [59], [60]]. Clinical studies have confirmed that MSCs can alleviate immune-related disorders by regulating the function of immune cells [58,59,61]. Intravenous infusion of bone marrow-derived MSCs in patients with acute graft-versus-host disease improved the rates of complete and partial response, and increased the survival rate as well [62]. Furthermore, therapeutic effects of MSCs have been demonstrated in patients with Crohn's disease [63], diabetes [64,65], and multiple sclerosis [66].

However, the quadruple-positive cells identified in this study were negative for several MSC markers such as CD73 and CD105. In addition, only a small proportion of the stromal cells in biosheets differentiated into the osteogenic and adipogenic lineages in vitro, and chondrogenic differentiation were histologically incomplete. Therefore, we could not ascertain whether the same population of cells could differentiate into all three lineages. The presence of MSCs in the biosheets should be confirmed using more specific markers, and the multipotency of the isolated cells will also have to be ascertained. To further assess involvement of the isolated MSCs in biosheet-mediated cutaneous regeneration, a conditional knockout mouse model could be established to selectively remove MSCs from the biosheets. However, given the technical complexity of this approach, the association between MSC density within biosheets and skin regeneration in a mouse wound model could be evaluated as an alternative. Biosheets with varying MSC densities can be fabricated using different mold materials, and by varying the duration after mold implantation.

Furthermore, MSCs are one of the main cell types responsible for HGF production and secretion. Given that cells containing HGF within the cytoplasm were detected in biosheets (Fig. 6D; Fig. S5), and both HGF-positive cells and tissues persisted in the wound until day 14 post-wounding (Fig. 6J–H; Fig. S6), we surmised that some cell types within biosheets continuously produce and secrete HGF, thereby promoting cutaneous tissue regeneration. HGF is secreted by MSCs in a paracrine fashion [67], and their therapeutic effects against spinal cord injury [68], lung injury [69], and myocardial infarction [70] in animal models were reinforced by administering HGF. In addition, neutrophils also release substantial amounts of HGF in the blood and bone marrow of patients with inflammatory response [71,72]. In addition, certain types of neutrophils have been shown to promote tumor progression [73] or liver regeneration [74] through HGF secretion. Given the possibility of the presence of MSC-like cells in the biosheets, and the fact that biosheets are generated by an inflammatory reaction against foreign bodies, it is reasonable to surmise that MSCs and neutrophils are the likely sources of HGF in the biosheets. These hypotheses should be validated through future cytological and histological analyses. For example, specific markers of MSCs and neutrophils (e.g., Ly6G and CD11b) [75,76] will first be identified in the biosheets, and the cells will be isolated by fluorescence-activated cell sorting. Thereafter, HGF mRNA and protein levels in the cell lysates and culture supernatants will be analyzed by RT-PCR and protein assays. Furthermore, the HGF-expressing cells within the biosheets will be identified by combining in situ hybridization for HGF mRNA with immunostaining for MSC or neutrophil markers.

The putative mechanism of tissue regeneration mediated by the biosheets, as suggested by the present study, is illustrated in Fig. 8. Although the presence of MSCs has not yet been confirmed, they are included in the illustration as a hypothetical component to represent future research perspectives. Briefly, the MSCs and HGF in the biosheets are released into the vicinity immediately after implantation, and inhibit inflammation and fibrosis that can prevent tissue regeneration (Fig. 8A). HGF then promotes neovascularization and neurogenesis at the margin of the biosheets (Fig. 8B), which creates a local environment suitable for tissue regeneration. Finally, the stem cells in surrounding tissues migrate to the biosheet, proliferate in response to HGF (Fig. 8C), and then differentiate into multiple lineages that reconstitute the injured tissues (Fig. 8D).

Fig. 8.

Schematic illustration of the mechanism of tissue regeneration using biosheets

(A) The early phase of wound healing after implantation of biosheets. The MSCs and HGF in the biosheets inhibit inflammation and fibrosis. (B) Neurogenesis and neovascularization occur at the margin of the biosheet. (C) MSCs migrate to the biosheet from the surrounding tissues, and undergo proliferation and differentiation. (D) The damaged tissues are reconstructed by the generation of new tissues.

A greater understanding of these processes can broaden the application of iBTA in the field of regenerative medicine. For instance, the proteins and cells in the biosheets underwent changes in a time-dependent manner after implantation of the molds. Similarly, the material and surface features of these molds can affect protein adsorption, migration or adhesion of inflammatory cells, encapsulation, and fibrosis [1]. It was also reported that thickness and mechanical strength differ depending on the material of the molds [3]. Furthermore, the iBTA molds reported so far have been embedded in subcutaneous tissues. It stands to reason that embedding these molds in other tissues, such as the muscles or peritoneal cavity, will alter the cellular and protein composition of the biosheets. Therefore, identifying the factors related to tissue regeneration under different conditions can aid in developing the most appropriate method to fabricate biosheets and optimize the regeneration process. We also found that tissue regeneration using biosheets can be reinforced by supplementing pro-regenerative factors such as HGF. Further identification of the factors and mechanisms involved in tissue- or organ-specific regeneration will help optimize the conditions of fabricating biosheets depending on the organ or disease.

The limitation of this study is that we could only validate the relationship between HGF and skin regeneration, and were not able to ascertain the involvement of MSCs. Tissue regeneration in biosheets is regulated by diverse factors, which may differ depending on the organ. In fact, several proteins other than HGF which may contribute to tissue regeneration, such as TGF-β, PDGF-β and Cdk1, were also identified as possible upstream regulators of the proteins upregulated in biosheets. In addition, biosheets-mediated neovascularization after wounding was evaluated solely by vWF immunostaining. However, the expression of vWF and other endothelial markers, such as CD31 and CD34, varies depending on the organ and the maturation status of blood vessels [77,78]. In this study, there were sparse vWF-positive blood vessels in the full-thickness skin wound specimens, and none were detected outside the deeply stained areas. This observation may reflect a potential bias of vWF staining toward vessels actively secreting vWF, which could be associated with endothelial maturity. Therefore, to detect early-stage neovascularization more accurately, it is necessary to stain for additional endothelial markers. Although we attempted immunostaining for CD31, we were unable to optimize a reliable staining protocol in the present study. Taken together, further experiments are necessary to fully understand the mechanisms underlying the regenerative effects of biosheets.

4. Conclusions

HGF is one of the factors involved in biosheet-mediated cutaneous tissue regeneration. HGF-producing cells with cytoplasmic HGF immunoreactivity are present in the biosheets and may contribute to the regenerative effect. We also identified additional factors that may be related to tissue regeneration in biosheets, such as quadruple-positive cells and various differentially expressed proteins. Our study provides new insights into the potential clinical applications of iBTA.

5. Materials and methods

5.1. Animals

Six weeks old C57BL/6J mice and C57BL6-Tg (CAG-EGFP) mice were obtained from SLC Inc. (Shizuoka, Japan). All experiments were approved by the Animal Care and Use Committee of The University of Tokyo. The mice were anesthetized before and during the surgery through isoflurane inhalation (Maruishi Pharmaceutical Co. Ltd., Osaka, Japan).

5.2. Preparation of biosheets and harvest of fascia

The molds for biosheets were prepared by trimming a silicon sheet (thickness 1.5 mm) into 1 cm² pieces. Although biosheet formation varies depending on the material and surface characteristics [1,3], silicone has been used in previous studies related to iBTA, including our own [4,6,7,16,23]. In addition, silicone is widely used in medical devices such as central venous catheters, urethral catheters, ventriculoperitoneal shunts, and breast implants, and its biocompatibility is well established [79,80]. A median longitudinal incision was made on the dorsal skin of mice, and the subcutaneous layer was dissected to make two pouches on both sides of the incision. The silicone plate was inserted into each subcutaneous pouch, and the skin incision was sutured and closed with 5-0 Nylon. The biosheets surrounding the molds were harvested 1, 2 and 4 weeks after implantation (Fig. 1B), and designated as 1w-, 2w, and 4w-biosheets respectively. The full-thickness dorsal skin was resected en bloc, and the fascia layer was carefully dissected and separated from the panniculus carnosus using a stereomicroscope.

5.3. Flow cytometry

The fascia and biosheets harvested at different time points were minced and digested with 0.1% collagenase (#032-22364, FUJIFILM, Tokyo, Japan) at 37 °C for 30 min. The tissue homogenates were filtered through a 100 μm cell strainer and the dissociated cells were collected in 50 ml conical tubes. The cell suspension was centrifuged at 150 g for 5 min and the supernatant was removed. The pellet was resuspended in Cell Staining Buffer (#420201, BioLegend, San Diego, CA, USA), and stained with PE/Cy7-conjugated anti-CD31 (#25-0311-82, eBioscience, Waltham, MA, USA), PE/Cy7-conjugated anti-CD45 (#25-0351-82, eBioscience), APC/Cy7-conjugated anti-Sca-1 (#10825, BioLegend), FITC-conjugated anti-CD90 (#553004, BD Biosciences, Franklin Lakes, NJ, USA), PE-conjugated anti-CD29 (#12-0291-82, eBioscience), APC-conjugated anti-CD44 (#103011, BioLegend), APC-conjugated anti-CD73 (#127209, BioLegend), APC-conjugated anti-CD105 (#17-1051-82, eBioscience) and FITC-conjugated anti-CD146 (#11-1469-42, eBioscience) antibodies at 4 °C for 20 min. After counterstaining with DAPI (#564907, BD Biosciences), the cells were analyzed by flow cytometry with BD Aria™ Fusion (BD Biosciences). Stromal cells were defined as negative for both CD31 and CD45. Flow cytometric analysis was performed using three biological replicates.

5.4. Isolation and differentiation of stromal cells

Cells were isolated from 2w-biosheets and stained anti-CD31 and anti-CD45 antibodies as described previously. The CD31-/CD45-stromal cells were sorted using BD Aria™ Fusion (BD Biosciences), and suspended in Dulbecco's modified Eagle's medium (DMEM) supplemented with 20% fetal bovine serum (FBS; #10270-106, Gibco, Waltham, MA, USA), 20 ng/ml fibroblast growth factor (FGF; A16321, Kaken Pharmaceutical, Tokyo, Japan), and 1% P/S. The cells were seeded in 10 cm culture dishes at the density of 1 × 10⁵ cells/plate (1.8 × 10³ cells/cm²) and incubated at 37 °C with 5% CO₂. The medium was changed every 3-4 days, and the cells were dissociated in 1x trypsin (T4174, Sigma-Aldrich, St. Louis, MO, USA) once 100% confluent.

For osteogenic and adipogenic differentiation, the stromal cells were seeded into 24-well plates coated with fibronectin (C-43060, PromoCell, Heidelberg, Germany) at the density of 2 × 10⁴ cells/well (1.1 × 10⁴ cells/cm²) in MSC Growth Medium 2 (C-28009, PromoCell). Once the cells reached 80%-100% confluence, the medium were replaced with MSC Osteogenic Differentiation Medium (C-28009, PromoCell) or MSC Adipogenic Differentiation medium (C-28016, PromoCell). The medium was changed every 2 to 3 days. After 10 days of differentiation, the cells were fixed with 10% formaldehyde. Osteogenic differentiation was detected by staining with alkaline phosphatase (ALP) and Alizarin red, and Oil red staining was performed to assess adipogenic differentiation.

To induce chondrogenic differentiation, 1 × 10⁵ cells were seeded in U-bottomed 96 well plates in complete DMEM (supplemented with 10% FBS). The cells formed pellets after 3 days of culture, and the medium was replaced with MSC Chondrogenic Differentiation Medium (C-28012, PromoCell). The medium was changed every 3 days. After 21 days of induction, the pellets were embedded in Tissue-Tek OCT compound (Sakura Finetechnical Co Ltd, Tokyo, Japan) and cut into 5 μm-thick sections. Chondrogenic differentiation was analyzed by staining with Toluidine blue and Safranine O.

5.5. Mouse model of skin regeneration with biosheets

The dorsal skin of six weeks old C57BL/6J mice was depilated using a hair clipper and depilatory cream. Two nitril-butadiene rubber (NBR) rings (internal diameter 10 mm) were fixed to the dorsal skin on both sides of dorsum midline, and 6 mm full-thickness wound was created in the center of each NBR ring by excising all skin layers from the epithelium to the subcutaneous fascia, including the panniculus carnosus, using a punch biopsy instrument. Fascia and 2w-biosheet were resected from the C57BL6-Tg (CAG-EGFP) mice, immediately trimmed into circles of diameter 8 mm. As the biosheets were thin and crumpled easily during handling, they were trimmed using punch biopsy instruments before removal of the silicone molds, and the biosheets on the deep side of the molds were handled with care. The fascia or 2w-bosheet was placed onto each wound and fixed using 7-0 unabsorbable suture (Fig. 4A). The 2w-biosheet was placed with its outer surface in contact with the wound surface. The wounds were covered with polyurethane foam dressing (Hydrocyte plus, Smith & Nephew KK, Tokyo, Japan) measuring 8 mm in diameter to prevent the grafts and wounds from drying. The body of each mouse was wrapped with film dressing (Airwall fu-wa-ri, Kyowa Ltd, Tokyo, Japan; Fig. 4B). The dressings were removed 7 and 14 days after wounding, and macroscopic and histological analyses of skin regeneration were performed on five wounds transplanted with 2w-biosheet or fascia at each time point. The wounds were photographed using a TG-6 digital camera (Olympus, Tokyo, Japan). Macroscopic re-epithelialization was subsequently evaluated. The mice were euthanized, and full-thickness dorsal wall from the epithelium to the muscles including the wound sites were resected and processed for histological analysis.

5.6. Real-time quantitative PCR

Total RNA was extracted from the differentiated cells using RNeasy Mini kit (#74104, QIAGEN, Hilden, Germany) and reverse transcribed to cDNA using reverse transcriptase (SuperScript™ III q-Strand Synthesis SuperMix for qRT-PCR, #11752050, Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. QRT-PCR was performed using FAST SYBR® Green Master Mix (#4385612, Applied Biosystems, Waltham, MA, USA) in A 7500 Fast Real-Time PCR System (Applied Biosystems). The primer sets are shown in Table 1. GAPDH was used as the internal control. RT-PCR was performed using three biological replicates, and each sample was analyzed in triplicate.

Table 1.

Primer sequences used in this study.

Gene name		Sequence
Col2a1	FW	5′- TTGAGACAGCACGACGTGGAG -3′
	RV	5′- AGCCAFFTTGCCATCGCCATA -3′
Bglap	FW	5′- GGGCAATAAGGTAGTGAACAG -3′
	RV	5′- GCAFCACAFFTCCTAAATAGT -3′
Adipoq	FW	5′- TGTTCCTCTTAATCCTGCCCA -3′
	RV	5′- CCAACCTGCACAAGTTCCCTT -3′
Gapdh	FW	5′- CCACTAACATCAAATGGGGTGAGG -3′
	RV	5′- TACTTGGCAGGTTTCTCCAGGC -3′

5.7. Enzyme-linked immunosorbent assay (ELISA)

Fascia, 2w-biosheet and 4w-biosheet were resected from C57BL/6J mice, snap frozen, and preserved at −80 °C. The thawed samples were minced and lysed in Cell Lysis Buffer 2 (#895347, R&D systems, Minneapolis, MN, USA) at room temperature for 30 min. After centrifuging at 600 g for 10 min, the supernatant was collected and the protein content was measured using a BCA assay kit (T9300A, TaKaRa, Tokyo, Japan). HGF levels in the samples were measured using Quantikine ELISA Mouse/Rat HGF Immunoassay kit (MHG00, R&D Systems) according to the manufacturer's instructions. Three biological replicates were used for ELISA, and each sample was analyzed in triplicate.

5.8. Proteomics analysis

Proteomic analysis was performed on 2w-biosheets, 4w-biosheets and fascia (n = 5). The samples were mixed with 0.1% trifluoroacetic acid (TFA) in acetonitrile, and homogenized with a 5 mm zirconia bead (TOMY SEIKO, Tokyo Japan) using a Tissue Lyser (Qiagen). The bead was removed, and the homogenates were centrifuged at 15,000×g for 15 min at 4 °C. After discarding the supernatant, the precipitate was dissolved in 4% sodium dodecyl sulfate and 100 mM Tris-HCl, pH 8 using a sonicator in a water bath (Bioruptor II, CosmoBio, Tokyo Japan). The protein content in the extracts was determined using a BCA protein assay kit (Thermo Fisher Scientific) and adjusted to 500 ng/μL with 4% sodium dodecyl sulfate and 100 mM Tris-HCl, pH 8. Aliquots of 40 μL were treated with 20 mM tris(2-carboxyethyl)phosphine) at 80 °C for 10 min and alkylated with 35 mM iodoacetamide in the dark at room temperature for 30 min. The alkylated samples were cleaned up and digested by the SP3 method [81,82] using hydrophilic (CAT# 45152105050250) and hydrophobic (CAT# 65152105050250) Sera-Mag SpeedBead carboxylate-modified magnetic beads (Cytiva, Marlborough, MA, USA). The beads were combined at 1:1 (v/v) ratio, washed twice with distilled water, and reconstituted in distilled water at the concentration of 10 μg/μL. The alkylated protein samples were each mixed with 20 μL of the reconstituted beads, and 99.5% ethanol was added to a final concentration of 75%. After mixing for 20 min, the supernatant was discarded, and the beads were washed twice with 80% ethanol. The beads were then resuspended in 100 μL 50 mM Tris-HCl (pH 8.0) with 1 μg trypsin/Lys-C Mix (Promega, Madison, WI, USA) and mixed gently at 37 °C for 16 h to digest the proteins. The digested samples were acidified with 20 μL 5% TFA, sonicated using Bioruptor II, and then desalted using SDB-STAGE tip (GL Sciences, Tokyo, Japan) according to the manufacturer's protocol. After drying in a centrifugal evaporator (miVac Duo concentrator, Genevac, Ipswich, UK), the samples were redissolved in 2% ACN in 0.1% TFA. The peptide concentration was determined using a Lunatic instrument (Unchained Labs, Pleasanton, CA, USA) and the samples were transferred to a hydrophilic MS vial (AMR, Tokyo, Japan).

The digested peptides were directly injected onto an Aurora column (C18, 75 μm ID, 25 cm length, 1.6 μm beads, IonOpticks, Victoria, Australia) at 60 °C and then separated with a 120-min gradient (A = 0.1% formic acid in water, B = 0.1% formic acid in 80% ACN; 0 min 2% B, 108 min 28%, 114 min 65% B, 120 min 65% B) at the flow rate of 200 nL/min using an UltiMate 3000 RSLCnano LC system. The peptides eluted from the column were analyzed by overlapping window DIA-MS [83,84] using an Orbitrap Exploris 480 with an InSpIon system [85]. The MS1 spectra were collected in the range of m/z 495–745 at the resolution of 15,000 to set an AGC target of 3 × 10⁶ and a maximum injection time of “Auto”. The MS2 spectra were collected at m/z 200-1,800 at the resolution of 60,000 to set an AGC target of 3 × 10⁶, maximum injection time of “Auto”, and normalized collision energies of 22%, 26% and 30%. The isolation width for MS2 was set to 4 Da, and overlapping window patterns at m/z 500–740 were used for window placements optimized via Scaffold DIA v3.2.1 (Proteome Software, Portland, OR, USA).

The MS files were analyzed with human spectral libraries using Scaffold DIA v3.2.1. The Mus musculus spectral libraries were generated from the UniProt protein sequence database (UniProt ID UP000000589, reviewed, canonical, 21,986 entries, downloaded in November 2021) using Prosit [86,87]. The Scaffold DIA search parameters were set as follows: experimental data search enzyme, trypsin; maximum missed cleavage sites, 1; precursor mass tolerance, 10 ppm; fragment mass tolerance, 10 ppm; static modification, cysteine carbamidomethylation. The protein identification threshold for both peptides and proteins was false discovery rate <1%. The proteins and peptides were quantified using Scaffold DIA.

Statistical analysis was performed using the Perseus software 1.6.15.0 (https://maxquant.net/perseus/) [88]. Quantified protein intensities were log2-transformed and filtered so that each protein had a minimum of 70% valid values in at least one group. Proteins exhibiting significant differences with p < 0.05 (ANOVA test) were hierarchically clustered, and the cluster upregulated relative to the fascia was selected. GO and KEGG pathway analyses were performed using Enricher tool. Upstream analysis of the proteins from the selected cluster was performed using the TRASPATH and TRANSFAC databases in the geneXplain platform (GeneXplain GmbH, Wolfenbüttel, Germany).

5.9. HGF application in the skin regeneration model

To evaluate the effect of HGF on the regeneration model, fascia resected from the C57BL6J mice was transplanted onto a full-thickness dorsal wound of C57BL6J mice according to Section 5.3. Fifty nanograms recombinant murine HGF (315-23, PeproTech, Cranbury, NJ, USA) was diluted in 10 μL PBS, and administrated to the fascia-transplanted wound immediately after transplantation. Sterile 10 μL PBS was administrated as a control. Fourteen days after wounding, the regeneration of cutaneous tissue was compared macroscopically and histologically between the HGF group (n = 7) and control group (n = 7).

5.10. Preparation of anti-HGF antibody

Recombinant rat HGF was purified from the culture supernatant of Chinese hamster ovary cells as previously described [89,90]. Female Japanese White rabbits (SLC Inc.) were immunized with rat HGF. The antiserum titer was measured and monitored by enzyme-linked immunosorbent assay and the IgGs from anti-HGF serum were purified using Protein A-Sepharose. The anti-HGF antibody neutralizes murine and rat HGF but not human HGF [91].

5.11. Antibody-mediated neutralization of HGF in biosheets

To test antibody-mediated neutralization of HGF in the biosheets, the anti-HGF antibody was administrated to the 2w-biosheets in-situ. Silicon molds were implanted into mice as described, and the biosheets were injected with the antibody 2 weeks later. Briefly, a median longitudinal incision was made on the dorsal skin of the mice and the biosheets were exposed, and 10 μg anti-HGF IgG diluted in 50 μL PBS was injected around the biosheets using 30G needle. As a control, 10 μg rabbit control IgG (15006, Sigma-Aldrich) diluted in 50 μL PBS was injected in the same manner. The skin incision was sutured and closed with 5-0 Nylon again. The biosheets were harvested 24 h later, and HGF neutralization was determined by immunostaining for phosphorylated MET. Three biosheets per group were analyzed.

5.12. Effect of HGF neutralization the skin regeneration model

Full-thickness skin wounds were induced on the dorsum of C57BL/6J mice, and 2w-biosheets resected from C57BL/6J mice were transplanted as described on day 0 of wounding. To neutralize HGF in the biosheets, 100 μg anti-HGF IgG diluted in 50 μL PBS was administrated locally on days 0, 1, 3 and 5 post-wounding. Rabbit IgG (15006, Sigma-Aldrich) diluted in 50 μL PBS was administrated as a control. The wounds were covered with film dressing (Airwall fu-wa-ri) instead of polyurethane foam dressing to prevent absorption of the antibodies. Tissue regeneration in the anti-HGF IgG (n = 5) and rabbit control IgG groups (n = 5) were compared 7 days post-wounding.

5.13. Histological analysis and immunohistochemistry (IHC)

The biosheets and dorsal wounds were harvested by hollowing out the dorsum. All specimens were cut craniocaudally, embedded in Tissue-Tek Oct compound 4583 (Sakura Finetechnical Co Ltd, Tokyo, Japan), and stored at −80 °C. The blocks were cut in the sagittal plane into 7 μm-thick sections, and then stained with hematoxylin and eosin (HE) or Masson Trichrome (MT) solutions.

For IHC, the sections were treated with 3% hydrogen peroxide for 7 min at room temperature, and then with 5% bovine serum albumin for 30 min to block non-specific staining. The sections were incubated overnight with rabbit anti-HGF (DF6326, 1:200, Affinity Bioscience, Cincinnati, OH, USA), rabbit anti-c-Met (718000, 1:50, Invitrogen, Waltham, MA, USA), and rabbit anti-phosphorylated Met (ab5662, 1:200, Abcam PLC, Cambridge, MA, USA) antibodies at 4 °C. Rabbit IgG (NI01, Merck KGaA, Darmstadt, Germany) was used as the isotype control, and diluted to the same concentration as the primary antibodies. The sections were then incubated with biotinylated goat anti-rabbit IgG at room temperature for 30 min, followed by the ABC reagent (VECTASTAIN® Elite® ABC-HRP Kit, PK-6101, Vector laboratories, Burlingame, CA, USA) at room temperature for 30 min. The color was developed with DAB using the ImmPACT® DAB Substrate Kit (SK-4105, Vector laboratories), and the stained sections were observed under a microscope (BX-51, Olympus Co., Tokyo, Japan). For immunofluorescence staining, the sections were incubated with Alexa Fluor488 goat anti-rabbit IgG (ab150077, 1:200; Abcam PLC) or Alexa Fluor594 goat anti-rabbit IgG (ab150080, 1:200, Abcam PLC) following incubation with the primary antibodies, and the nuclei were stained with DAPI (Thermo Fisher Scientific). Images were acquired on a fluorescence microscope (BZ-9000, Keyence Co., Osaka, Japan).

5.14. TEM analysis

The biosheets were harvested by hollowing out the dorsum, and were fixed in a 0.1 M cacodylate buffer (pH 7.4) containing 2% glutaraldehyde and 4% paraformaldehyde at 4 °C for 3 h. The specimens were rinsed in the same buffer and post-fixed in 2% osmium tetroxide at 4 °C for 3 h. Dehydrated specimens were embedded in epoxy resin (Epon 812), and ultrathin sections (70 to 80 nm) were cut on an LKB-8800 ultramicrotome (GE Healthcare) using a diamond knife, and mounted on 200-mesh copper grids. Sections were stained with 2% uranyl acetate and lead citrate, and observed with a JEM-2000EX transmission electron microscope (JEOL) operating at 100 kV.

5.15. Image analysis

The digital images were analyzed using the Quick Grain Standard software (Inotech, Hiroshima, Japan) to measure the wound area and the length of epithelialization. The number of positively stained cells were manually counted. For H-score evaluation, at least 200 cells were manually assessed per specimen across 5 high-power fields.

5.16. Statistical analysis

Statistical analysis was performed using R Studio software (version 1.2.5033). All results were presented as mean ± standard error. Multiple groups were compared by one-way analysis of variance (ANOVA) followed by Bonferroni's post hoc test. Two groups were compared by Student's t-test. P value < 0.05 was considered statistically significant.

Funding

The article processing charge was partially funded by the University of Tokyo Library System.

CRediT authorship contribution statement

Keisuke Suzuki: Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Validation, Visualization, Writing – original draft. Hiroko Komura: Conceptualization, Investigation, Methodology, Validation. Ryo Konno: Formal analysis, Investigation, Resources, Visualization. Yusuke Kawashima: Formal analysis, Resources, Writing – review & editing. Eiichiro Watanabe: Conceptualization, Methodology. Hiroki Sato: Methodology, Resources, Writing – review & editing. Kunio Matsumoto: Methodology, Resources, Supervision, Writing – review & editing. Ryoko Inaki: Investigation, Methodology. Sanshiro Kanazawa: Investigation, Methodology. Yukiyo Asawa: Methodology, Validation. Atsuhiko Hikita: Resources, Supervision, Writing – review & editing. Kazuto Hoshi: Resources, Supervision. Jun Fujishiro: Resources, Supervision. Makoto Komura: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Makoto Komura reports a relationship with SheepMedical Co., Ltd. that includes: funding grants. Yukiyo Asawa reports a relationship with SheepMedical Co., Ltd. that includes:. Atsuhiko Hikita reports a relationship with Fujisoft Inc., CPC Corporation, Kyowa Co., Ltd., Kanto Chemical Co., Inc., Nichirei Corp., Kohjin Bio Co., Ltd. that includes: funding grants. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Data availability

Data will be made available on request.