Ultra-fine Micronized Human Acellular Dermal Matrix Promotes Diabetic Wound Healing in a Mouse Model

Abstract

Background/Aim

Diabetic wounds are characterized by chronic inflammation, impaired angiogenesis, and abnormal extracellular matrix (ECM) remodeling, which contribute to delayed wound healing and non-healing ulcers. Because ECM degradation and impaired matrix repair are central features of diabetic wound pathology, we evaluated an ultra-fine micronized human acellular dermal matrix (UFM-hADM; mean particle size approximately 70 µm) as an injectable ECM scaffold for diabetic wound repair.

Materials and Methods

Public transcriptomic datasets from human diabetic foot ulcers were reanalyzed to assess ECM remodeling- and growth factor-related gene expression patterns and enriched wound repair-associated pathways. Full-thickness excisional wounds were then created on the dorsum of streptozotocin (STZ)-induced diabetic mice and treated with perilesional microinjections of UFM-hADM or saline. Wound closure was assessed by digital planimetry. Wound tissues were analyzed by histology, immunohistochemistry, and quantitative reverse transcription-polymerase chain reaction.

Results

Human diabetic foot ulcer datasets showed increased expression of several matrix metalloproteinase family members, altered ECM- and growth factor-related transcripts, and enrichment of tissue remodeling-associated pathways. In STZ-induced diabetic mice, UFM-hADM significantly improved wound closure compared with saline treatment, with near-complete epithelial coverage and more organized granulation tissue by Day 10. UFM-hADM-treated wounds showed increased expression of Col1a1, Col3a1, Tgfb1, and Egf, whereas Mmp1 and Mmp3 expressions were reduced. Immunohistochemistry showed increased CD31-positive microvessel density and α-smooth muscle actin-positive myofibroblast staining.

Conclusion

Human diabetic foot ulcer transcriptomic profiles support ECM remodeling and growth factor-related signaling as disease-associated features of chronic diabetic wounds. UFM-hADM improved wound repair in STZ-induced diabetic mice, with increased angiogenesis, reduced MMP expression, altered ECM-related gene expression, and more organized granulation tissue formation. These findings support further evaluation of UFM-hADM as an injectable human-derived ECM scaffold for chronic diabetic wounds.

Introduction

Diabetes mellitus is frequently associated with impaired cutaneous repair, including delayed healing of surgical wounds, traumatic injuries, pressure-related wounds, and lower-extremity ulcers (1, 2). Although diabetic foot ulcers are the most clinically recognized manifestation, impaired healing under diabetic conditions is not limited to one wound type but reflects broader defects in tissue repair (3, 4). In diabetic wounds, the transition from inflammation to tissue formation is delayed, neovascularization is reduced, and extracellular matrix (ECM) turnover is disrupted (2, 4, 5). These changes impair granulation tissue formation and limit re-epithelialization. Therefore, effective diabetic wound therapies should address not only wound coverage but also local matrix and vascular microenvironment.

Cutaneous wound repair requires coordinated interactions among keratinocytes, fibroblasts, endothelial cells, and inflammatory cells within a provisional ECM. During normal repair, this matrix supports cell migration, regulates growth-factor availability, and is gradually replaced by newly synthesized collagen-rich tissue (6, 7). In diabetic wounds, this process is disturbed by persistent inflammation, reduced fibroblast activity, impaired angiogenesis, and an imbalance between ECM synthesis and degradation (4, 5, 8-10). Persistent matrix metalloproteinase (MMP) activity can shift the wound microenvironment toward excessive matrix degradation, leading to the breakdown of newly formed extracellular matrix and repair-associated cytokines, which may impair granulation tissue formation and subsequent epithelial advancement (11). These features support the use of scaffold-based approaches that can supplement the local matrix environment and support cellular events required for tissue repair.

Regenerative approaches for diabetic wounds are increasingly focused on improving the wound-bed environment rather than simply covering the wound surface (12, 13). Biologic ECM scaffolds are of interest because they can provide a provisional three-dimensional matrix, facilitate host-cell infiltration, and present ECM-derived cues that support angiogenesis, epithelial migration, and matrix remodeling (14-16). Human acellular dermal matrix (ADM) is a representative ECM scaffold used in wound management. Because decellularized dermis preserves major dermal ECM components, including collagen and elastin, it can provide structural support and may help retain endogenous growth factors within the wound bed. Clinical studies have reported improved healing outcomes with ADM-based products in complex acute and chronic wounds, including diabetic foot ulcers (17, 18). However, conventional sheet or mesh-type ADM products require close contact with the wound surface and may be difficult to apply to irregular, undermined, or deep wounds. Incomplete removal of cellular remnants or donor-derived components is also an important consideration for the biocompatibility of decellularized matrices (19, 20).

To improve delivery to irregular or difficult-to-access wound beds, particulate and micronized ADM formulations have been developed as paste-type or injectable matrices that can be applied to cavities, undermined margins, and uneven wound surfaces (21-24). However, some micronized or paste-type ADM products contain relatively coarse or heterogeneous particles, which may limit dispersion through narrow-gauge needles and reduce matrix-tissue contact within the perilesional wound bed (25-27). UFM-hADM was developed as an ultra-fine injectable form of human ADM to address these practical limitations while retaining the biological features of human dermal ECM. Human dermis is decellularized, purified, and mechanically micronized to a mean particle size of approximately 70 µm. This ultra-fine injectable format is intended to allow local microinjection around the wound margin while retaining the ECM features of human dermis, but its relevance to the ECM abnormalities seen in diabetic wounds and its effect on diabetic wound repair remain to be defined.

In this study, we first reanalyzed public transcriptomic datasets from human diabetic foot ulcers to examine ECM remodeling- and growth factor-related gene expression patterns. We then evaluated whether perilesional UFM-hADM microinjection improves wound repair in a streptozotocin-induced diabetic mouse model of full-thickness excisional skin injury. Wound closure, re-epithelialization, granulation tissue organization, neovascularization, and ECM remodeling were assessed by digital planimetry, histology, immunohistochemistry, and gene-expression analysis of collagen-, growth factor-, and MMP-related markers. This study evaluated UFM-hADM as an injectable human-derived ECM scaffold for diabetic wound repair.

Materials and Methods

Animals. All experiments used male BALB/c mice (8-10 weeks old, 26-29 g body weight) obtained from RaonBio (Yongin, Republic of Korea). Mice were housed under standard laboratory conditions with controlled temperature and humidity and with ad libitum access to water and food. All animal procedures conformed to institutional and national guidelines and were approved by the Institutional Animal Care and Use Committee (IACUC) of Hulux (Approval No. Hulux-2025-02-002). Mice were acclimatized for at least 7 days before randomization to control (saline) and UFM-hADM groups (n=5/group).

Induction of diabetes. Type 1 diabetes was induced by intraperitoneal injection of STZ 50 mg/kg (S0130; Sigma-Aldrich, St. Louis, MO, USA) for 5 consecutive days. Disease development was monitored for 1 week. Diabetes was confirmed in animals with non-fasting blood glucose levels of 250 mg/dl or higher, measured twice weekly.

Wound creation and treatment. After confirmation of diabetes, mice were anesthetized with isoflurane (DaonPharm, Seongnam, Republic of Korea), the dorsum was shaved, and a single 8-mm full-thickness excisional wound was created. UFM-hADM (LCI Bio Co., Ltd., Namyangju, Republic of Korea) or sterile 0.9% saline was administered perilesionally on Days 2, 5, and 7 after wounding as 10 intradermal/subcutaneous deposits of 10 µl distributed evenly around the wound margin (total 100 µl per wound) using 29-31-gauge needles. Control animals received equal volumes of saline on the same schedule.

Ultra-fine micronized human acellular dermal matrix. UFM-hADM (LCI Bio Co., Ltd.) is an ultra-fine micronized human acellular dermal matrix supplied as an injectable suspension. The product is manufactured from donated human dermis by a multi-step decellularization process that removes cellular components and nucleic acids, followed by mechanical micronization. The manufacturer reports a mean particle size of approximately 70 µm and exhaustive decellularization to minimize residual cellular components and to improve suspension stability and biocompatibility.

Macroscopic wound analysis. Wound healing was documented by digital photography on Days 0, 4, 7, and 10. Standardized images were captured at a fixed distance and lighting with a scale marker. The wound area was traced and measured using ImageJ software (National Institutes of Health, Bethesda, MD, USA), and the wound closure rate was calculated as a percentage of the initial Day-0 wound area, as closure rate (%) = [(ADay 0 − ADay x)/ADay 0] × 100.

Hematoxylin and eosin staining. On Day 10, wounds together with 2-3 mm of surrounding margin were excised, fixed in 4% paraformaldehyde, embedded in paraffin, and sectioned at 5 µm. Sections were stained with hematoxylin and eosin (ab245880; Abcam, Cambridge, UK) and imaged with an NSS-6 digital slide scanner (Nexcope, Ningbo, China). Quantitative endpoints, measured on wound-center sections by blind observers, included epithelial tongue length and residual epithelial gap distance. Five sections per group were analyzed.

Masson’s trichrome staining. Paraffin-embedded skin sections (5 µm) were stained with Masson’s trichrome (KH07007; BQC, Asturias, Spain) to evaluate collagen content and granulation tissue. Granulation tissue thickness was quantified within a standardized wound-bed region of interest using ImageJ under identical image-acquisition settings. Five independent non-overlapping sections per group were analyzed, and images were acquired with the NSS-6 digital slide scanner.

Immunohistochemistry. Paraffin-embedded sections were deparaffinized in xylene and rehydrated through a graded ethanol series. Heat-induced epitope retrieval was performed in Tris-EDTA buffer (pH 9.0) for 20 min. Sections were blocked and incubated overnight at 4°C with anti-CD31 (1:2,000, EPR17259; Abcam) and anti-α-smooth muscle actin (α-SMA; 1:500, 19245; Cell Signaling Technology, Danvers, MA, USA). Signals were developed using horseradish peroxidase-conjugated secondary antibodies and 3,3’-diaminobenzidine (GK500710; Gene Tech, Shanghai, China), followed by hematoxylin counterstaining. Images were acquired with the NSS-6 digital slide scanner. The number of CD31-positive microvessels per high-power field (HPF) and the percentage of α-SMA-positive area were quantified by two blinded observers using three randomly selected HPFs per section.

Quantitative reverse transcription-polymerase chain reaction. Total RNA was extracted from Day-7 wound tissue using RNAiso (Takara, Shiga, Japan) with DNase treatment, and cDNA was synthesized from 500-1,000 ng RNA using a Takara reverse-transcription kit. Quantitative PCR was performed with SYBR Green Master Mix (TOYOBO, Osaka, Japan) on a Synergy HTX reader (BioTek, Winooski, VT, USA) in 10-µl reactions (0.3 µM primers; approximately 5-20 ng cDNA equivalent). Cycling conditions were 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 60 s; melt-curve analysis confirmed single products. Target genes were Col1a1, Col3a1, Mmp1, Mmp3, Tgf-b1, Egf, and Vegf, with actin-beta (Actb) as the reference gene. Primer sequences are listed in Table I. Relative expression was calculated by the 2^−ΔΔCt method using the control group as calibrator, and reactions were run in technical triplicate.

Table I. Primer sequences used for quantitative reverse transcriptionpolymerase chain reaction.

Transcriptomic dataset reanalysis. Public human diabetic foot ulcer (DFU) transcriptomic datasets were retrieved from the National Center for Biotechnology Information Gene Expression Omnibus. GSE80178 (non-diabetic control skin versus DFU) was used to examine ECM- and growth-factor-related gene expression, and GSE199939 (DFU versus control skin) was used for gene set enrichment analysis (GSEA). Differential-expression and functional-enrichment analyses, including Gene Ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) terms, were performed in R with multiple-testing correction.

Statistical analysis. All data are presented as the mean±standard deviation, with individual points representing biological replicates (n=5/group unless otherwise specified). Comparisons of wound closure rate between the saline and UFM-hADM groups were performed using a two-tailed unpaired t-test. Single-endpoint histomorphometric, immunohistochemical, and qRT-PCR comparisons were analyzed using a two-tailed unpaired t-test when assumptions of normality and homogeneity of variance were satisfied; otherwise, the Mann-Whitney U-test or one-way analysis of variance (ANOVA) was applied as appropriate. Analyses were performed using GraphPad Prism (GraphPad Software, Boston, MA, USA).

Results

Human diabetic ulcers show altered ECM remodeling and growth factor-related transcriptomic patterns. Human diabetic foot ulcers represent a chronic wound environment in which tissue breakdown persists despite ongoing repair attempts. Because this failure is closely linked to ECM degradation, protease activity, and altered growth factor signaling, we first reanalyzed public DFU transcriptomic datasets to define the matrix-related abnormalities present in patient-derived wounds (9, 10, 14). We focused on ECM structural genes, matrix-degrading enzymes, and TGF/EGFR-related transcripts because these pathways are directly relevant to the proposed function of UFM-hADM as an injectable dermal ECM scaffold. In GSE80178, a focused heatmap showed broad differences between DFU and control skin in transcripts associated with ECM remodeling and growth factor-related pathways (Figure 1A). Several collagen and structural genes, including COL1A2, COL4A5, COL14A1, and ACTA2, showed altered expression in DFU tissue. In contrast, multiple MMP family members, including MMP1, MMP2, MMP3, MMP7, MMP9, MMP10, and MMP12, were relatively increased in DFU tissue. This pattern is consistent with previous reports describing increased collagenase and stromelysin activity in chronic and diabetic ulcers and supports the presence of a protease-rich wound environment (9, 10). Components of the TGF and EGFR axes, including TGFA, TGFB1, TGFB2, TGFBR family members, and EGFR, also showed differential expression, suggesting altered growth factor-related signaling in DFU tissue.

Figure 1.

Public RNA-seq datasets highlight dysregulated ECM remodeling and signaling pathways in DFU. (A) Heatmap of selected ECM-, MMP-, and TGF/EGFR-related genes in non-diabetic control skin (Ctrl) and DFU tissue from GSE80178, showing coordinated changes in collagen genes (COL1A2, COL4A5, COL14A1), the contractile marker ACTA2, multiple MMPs (MMP1, MMP2, MMP3, MMP7, MMP9, MMP10, MMP12), and components of the TGF/EGFR axis (TGFA, TGFB1, TGFB2, TGFBR2, TGFBR3, TGFBR3L, TGFBRAP1, EGFR). Expression values are presented as row-wise z-scores (log2 scale). (B) Representative GSEA plots from GSE199939 for Gene Ontology terms enriched in DFU versus control, including connective tissue replacement and cellular-component categories related to epithelial and cortex-associated structures. (C) GSEA plot for the KEGG calcium signaling pathway in GSE199939, showing significant enrichment in DFU. ECM: Extracellular matrix; MMP: matrix metalloproteinase; TGF: transforming growth factor; EGFR: epidermal growth factor receptor; GSEA: Gene set enrichment analysis; KEGG: Kyoto Encyclopedia of Genes and Genomes; ES: enrichment score; DFU: diabetic foot ulcers.

We further analyzed an independent dataset, GSE199939, using gene set enrichment analysis (Figure 1B and 1C). DFU samples showed enrichment of selected Gene Ontology terms related to connective tissue replacement and epithelial or structural organization, together with enrichment of the KEGG calcium signaling pathway. These findings indicate that human DFU tissue is associated with altered ECM remodeling, matrix degradation, and growth factor-related signaling. Based on this disease-associated ECM remodeling profile, we next evaluated UFM-hADM as an injectable human-derived ECM scaffold in a diabetic mouse wound model.

UFM-hADM improves wound closure and histological repair in STZ-induced diabetic mice. To evaluate whether UFM-hADM improves wound repair under diabetic conditions, full-thickness dorsal excisional wounds were generated in STZ-induced diabetic mice (Figure 2A). All mice maintained non-fasting blood glucose levels of 250 mg/dl or higher after STZ administration, confirming the diabetic state before wound creation (Figure 2B). After generation of an 8-mm circular wound, mice received perilesional injections of saline or UFM-hADM on Days 2, 5, and 7 after wounding, and wound closure was monitored by digital planimetry until Day 10 (Figure 2C and 2D).

Figure 2.

UFM-hADM accelerates wound closure and tissue repair in an STZ-induced diabetic mouse model. (A) Experimental scheme: after 7 days of acclimation, mice received intraperitoneal STZ (50 mg/kg, once daily for 5 days); diabetes was confirmed by non-fasting blood glucose of 250 mg/dl or higher, followed by creation of an 8-mm full-thickness dorsal wound and perilesional injection of saline (control) or UFM-hADM (10 μl at 10 sites) on Days 2, 5, and 7 after wounding, with biopsy on Day 10. (B) Serial blood glucose levels in the control and UFM-hADM groups; the dashed line indicates 250 mg/dl. (C) Representative wound photographs and color-segmented images on Days 0, 4, 7, and 10. (D) Quantification of wound closure rate (%) at Days 4, 7, and 10. (E) Representative Day-10 hematoxylin and eosin-stained sections with quantification of epithelial tongue length and epithelial gap distance. (F) Representative Day-10 Masson’s trichrome-stained sections with quantification of granulation tissue thickness. Data are presented as the mean±standard deviation (n=5/group). *p<0.05, **p<0.01, ***p<0.0001 versus control (unpaired t-test). Scale bars: 2 mm (whole-wound overviews) and 400 μm (higher-magnification views). UFM-hADM: Ultra-fine micronized human acellular dermal matrix; STZ: streptozotocin; H&E: hematoxylin-eosin; MT: Masson’s trichrome.

Compared to saline-treated controls, UFM-hADM-treated wounds showed a faster reduction in wound area throughout the observation period. Quantitative analysis confirmed significantly higher wound closure rates in the UFM-hADM group on Days 4, 7, and 10 (Figure 2D), indicating that perilesional UFM-hADM microinjection improved macroscopic wound closure in this diabetic mouse model. Histological analysis on Day 10 further supported the wound closure results (Figure 2E and 2F). Hematoxylin and eosin-stained sections showed longer epithelial tongues and a smaller residual epithelial gap in UFM-hADM-treated wounds than in saline-treated controls, consistent with more advanced re-epithelialization (Figure 2E). Masson’s trichrome staining showed that saline-treated wounds retained a relatively thick granulation tissue layer with loosely arranged collagen, whereas UFM-hADM-treated wounds exhibited a thinner and more compact wound-bed structure beneath the neo-epidermis. Quantification confirmed increased epithelial tongue length, reduced epithelial gap distance, and reduced granulation tissue thickness in the UFM-hADM group (Figure 2E and 2F). These findings indicate that UFM-hADM improved early wound closure and was associated with more advanced epithelial and dermal repair by Day 10.

UFM-hADM alters ECM remodeling and growth factor-related gene expression in diabetic wounds. Because human DFU transcriptomic analysis showed altered ECM remodeling and growth factor-related signaling, we next examined whether UFM-hADM affected related molecular markers in diabetic mouse wounds. Day-7 wound tissues were analyzed by quantitative RT-PCR (Figure 3). Compared with saline-treated controls, UFM-hADM-treated wounds showed significantly higher Col1a1 expression, while Col3a1 showed a non-significant increasing trend (Figure 3A). Conversely, Mmp3 expression was significantly reduced in the UFM-hADM group, and Mmp1 showed a modest decreasing trend. Growth factor-related analysis showed that Tgfb1 and Egf expression were significantly increased after UFM-hADM treatment, whereas Vegf showed a non-significant upward trend at this time point (Figure 3B). Overall, UFM-hADM treatment was associated with increased collagen- and growth factor-related markers and reduced expression of selected matrix-degrading enzymes. This expression pattern was consistent with the improved re-epithelialization and more compact granulation tissue architecture observed histologically.

Figure 3.

UFM-hADM promotes collagen synthesis and suppresses MMP expression while enhancing pro-healing growth factors. (A) Relative mRNA expression of COL1A1, COL3A1, MMP1, and MMP3 in Day-7 wound tissue from saline control and UFM-hADM-treated STZ-induced diabetic mice, normalized to β-actin. UFM-hADM significantly increased COL1A1 and tended to increase COL3A1, while MMP3 was significantly reduced and MMP1 showed a decreasing trend compared with controls. (B) Relative mRNA expression of TGF-β1, EGF, and VEGF at Day 7, showing significant upregulation of TGF-β1 and EGF in the UFM-hADM group and a non-significant increase in VEGF. Data are presented as the mean±standard deviation (n=5/group). *p<0.05, **p<0.01, ***p<0.001 versus control (unpaired t-test). UFM-hADM: Ultra-fine micronized human acellular dermal matrix; STZ: streptozotocin; COL1A1: collagen type I alpha 1 chain; COL3A1: collagen type III alpha 1 chain; MMP: matrix metalloproteinase; TGF-β1: transforming growth factor beta 1; EGF: epidermal growth factor; VEGF: vascular endothelial growth factor; ns: not significant.

UFM-hADM increases CD31-positive microvessels and α-SMA-positive stromal responses in diabetic wounds. Because impaired neovascularization is a major feature of diabetic wound healing, and Vegf expression tended to increase after UFM-hADM treatment, we assessed vascular and stromal remodeling in Day-10 wound tissue by immunostaining for CD31 and α-SMA (Figure 4). Saline-treated diabetic wounds showed relatively few CD31-positive microvessels within the granulation tissue. In contrast, UFM-hADM-treated wounds exhibited a greater number of CD31-positive microvascular structures distributed across the wound area (Figure 4A). Quantitative analysis confirmed a significant increase in CD31-positive microvessels per high-power field in the UFM-hADM group compared with saline controls (Figure 4B).

Figure 4.

UFM-hADM increases neovascularization and myofibroblast activity in diabetic wounds. (A) Representative Day-10 wound sections stained for CD31 (endothelial cell marker) and α-SMA (myofibroblast marker) in saline control and UFM-hADM-treated STZ-induced diabetic mice. Arrows indicate CD31-positive microvessels and α-SMA-positive myofibroblast layers within the granulation tissue. (B) Quantitative analysis of the number of CD31-positive microvessels and the α-SMA-positive area (%) in each group. Data are presented as the mean±standard deviation (n=5/group). *p<0.05, **p<0.01 versus control (unpaired t-test). Scale bars: 100 μm. UFM-hADM: Ultra-fine micronized human acellular dermal matrix; STZ: Streptozotocin; CD31: cluster of differentiation 31; α-SMA: alpha-smooth muscle actin.

α-SMA staining also differed between groups. In saline-treated wounds, α-SMA-positive staining was limited and sparsely distributed within the granulation tissue. In contrast, UFM-hADM-treated wounds showed a broader and more organized α-SMA-positive stromal pattern along the regenerating wound area (Figure 4A). Quantification demonstrated a significant increase in α-SMA-positive area in the UFM-hADM group (Figure 4B). Together, these findings indicate that UFM-hADM treatment was associated with increased microvascular density and α-SMA-positive stromal remodeling in diabetic wounds.

Discussion

In this study, public transcriptomic datasets from human diabetic foot ulcers showed a matrix-degrading wound profile, with increased expression of several MMP family members and altered ECM- and growth factor-related transcripts compared with non-diabetic control skin. These findings support the relevance of altered ECM remodeling and proteolytic imbalance as disease-associated features of human diabetic foot ulcers (21, 28). We evaluated UFM-hADM as an injectable human-derived ECM scaffold in an STZ-induced diabetic mouse wound model. Perilesional UFM-hADM microinjection improved wound closure, re-epithelialization, granulation tissue organization, and CD31-positive microvessel formation. These findings indicate that the effect of UFM-hADM was not limited to macroscopic wound closure but was also associated with changes in epithelial repair, matrix organization, and neovascularization. Taken together, these observations in an STZ-induced excisional wound model, an established setting for evaluating impaired repair under diabetic conditions and local matrix-based interventions (8, 22), suggest that UFM-hADM can locally counter the ECM and proteolytic imbalances that characterize human diabetic ulcers.

A central barrier in diabetic wound repair is the instability of the local ECM environment, in which excessive proteolytic activity may degrade newly deposited matrix, reduce growth-factor retention, and compromise the formation of a stable wound bed (9, 10, 14, 23). In line with this concept, the human DFU datasets analyzed in this study showed broad alterations in ECM remodeling- and growth factor-related transcripts. In the mouse model, UFM-hADM-treated wounds showed significantly increased Col1a1 expression and reduced Mmp3 expression, with a non-significant increase in Col3a1 and a decreasing trend in Mmp1. Notably, MMP activity is not uniformly detrimental during repair, since Mmp9-dependent signaling can promote keratinocyte migration and re-epithelialization (29); the reduced Mmp3 expression observed here is therefore better interpreted as a shift toward proteolytic balance than as a global suppression of matrix turnover. These transcriptional changes were accompanied by a more compact Masson’s trichrome-stained wound-bed architecture. Together, these findings suggest that UFM-hADM treatment was associated with increased matrix-related gene expression and reduced expression of selected matrix-degrading enzymes within the diabetic wound environment. This interpretation is consistent with wound-bed preparation principles, which emphasize the need to reduce persistent proteolytic stress and support a matrix structure permissive for cell migration, granulation tissue organization, and epithelial advancement (6, 13, 15). These parallel changes across both axes suggest that UFM-hADM may engage the same mechanistic pathway targeted by clinical wound-bed preparation strategies, supporting its biological rationale in the diabetic wound environment.

The growth factor-related and vascular findings further support the involvement of tissue repair responses after UFM-hADM treatment. UFM-hADM increased Tgf-b1 and Egf expression, both of which are involved in fibroblast activity, granulation tissue formation, and epithelial repair, while VEGF showed a non-significant increasing trend at the examined time point. Consistent with these molecular changes, UFM-hADM-treated wounds showed increased CD31-positive microvessel density and a broader α-SMA-positive stromal response (30). These findings are consistent with improved vascular and stromal remodeling in UFM-hADM-treated diabetic wounds. They are also in line with known properties of ECM scaffolds. ECM-associated proteoglycans and glycosaminoglycans can bind and protect heparan sulfate-binding growth factors, including VEGF and FGF, thereby helping maintain growth factor availability within protease-rich wound environments (14, 23, 31). Although the present study did not directly measure growth factor binding or vascular maturation, the parallel changes in growth factor-related transcripts, CD31-positive microvessels, and α-SMA-positive stromal staining support the possibility that UFM-hADM treatment improved local conditions for vascular and matrix remodeling.

Perilesional microinjection may be useful for diabetic wounds because irregular margins, undermined areas, and fragile wound tissue can limit the contact and retention achieved by sheet-type matrices or surface-applied materials (32, 33). Topically applied formulations can also accelerate rodent skin wound healing (34), but their efficacy depends on sustained surface contact, which is difficult to maintain across irregular and undermined diabetic wounds. In this context, the ultra-fine particulate format of UFM-hADM provides a practical delivery advantage. The mean particle size of approximately 70 µm was compatible with administration through fine-gauge needles, enabling circumferential microinjection around the wound margin and potentially improving local matrix distribution within the perilesional tissue plane. A finer and more uniform injectable suspension may increase the matrix-tissue interface and facilitate exposure of resident and infiltrating cells to dermal ECM components (25, 26, 35). Together with the observed improvements in wound closure, re-epithelialization, ECM-related gene expression, and CD31-positive microvessel density, these delivery-related features support UFM-hADM as a feasible locally delivered injectable ECM scaffold for diabetic wound repair.

These findings should be interpreted in the context of several experimental considerations. This study used an STZ-induced mouse model, which reflects important aspects of impaired diabetic wound repair but does not fully reproduce the chronicity, microbial burden, ischemia, neuropathy, and repeated mechanical stress observed in human diabetic foot ulcers. The analysis was also limited to early wound healing endpoints; therefore, long-term durability, scar quality, and recurrence were not examined. Although UFM-hADM treatment was associated with changes in ECM-, MMP-, growth factor-, and vascular-related markers, this study did not directly test causal mechanisms such as growth factor binding, cell-specific stromal responses, or vascular maturation. In addition, comparative studies with conventional sheet-type ADM or other injectable/micronized matrix formulations will be needed to define the relative advantage of the ultra-fine injectable format. Nevertheless, the combined human transcriptomic and diabetic mouse wound data support further evaluation of UFM-hADM as a local ECM-based therapy for diabetic wounds.

This study used an STZ-induced mouse model, which reflects important aspects of impaired diabetic wound repair but does not fully reproduce the chronicity, microbial burden, ischemia, neuropathy, and repeated mechanical stress seen in human diabetic foot ulcers. The analysis was limited to early wound healing endpoints, and long-term durability, scar quality, and recurrence were not examined. Time-course analyses of rodent wound models have shown that histological and transcriptional repair markers continue to evolve beyond the early phase (36), so a single terminal time point cannot capture the later remodeling trajectory. Although UFM-hADM treatment was associated with altered ECM-, MMP-, growth factor-, and vascular markers, the study did not directly test causal mechanisms such as growth factor binding, macrophage polarization, fibroblast activation, or vascular maturation. The study was also conducted on a relatively small scale, and the observed differences, although consistent across the macroscopic, histological, immunohistochemical, and transcriptional readouts examined, should be regarded as preliminary. Larger-scale studies will therefore be required to confirm these effects and to establish their magnitude and reproducibility. Finally, comparative studies against conventional sheet-type ADM, coarser micronized ADM, hydrogel-based carriers, or standard wound dressings will be needed to define the relative advantage of the ultra-fine injectable format. Despite these limitations, the combined human transcriptomic and diabetic mouse wound data support further evaluation of UFM-hADM as a local ECM-based therapy for diabetic wounds.

Conclusion

Perilesional UFM-hADM microinjection improved diabetic wound repair in an STZ-induced mouse model, as shown by faster wound closure, enhanced re-epithelialization, more compact granulation tissue, and increased CD31-positive microvessel density. These tissue-level changes were accompanied by increased collagen- and growth factor-related gene expression and reduced Mmp3 expression. Together with the human DFU transcriptomic findings showing altered ECM remodeling and growth factor-related signaling, these results support further evaluation of UFM-hADM as a locally delivered, human-derived injectable ECM scaffold for diabetic wounds.

Conflicts of Interest

Authors B.S., K.-W.J., D.K., and S.L. are employed by Hulux, and Author Y.J.L. is employed by Live Cell Instrument. UFM-hADM is a commercial product associated with the institutions involved in this study. The Authors declare that these affiliations did not influence the design, conduct, analysis, or reporting of the study, and that they have no other conflicts of interest in relation to this work.

Authors’ Contributions

Conceptualization: Y.J.L., S.-M.L., and B.S.; Methodology: B.S., K.-W.J., D.K., and S.L.; Investigation: H.J., B.S.; Verification of animal experiments and assistance with histological staining: J.Y.K.; Resources: Y.J.L.; Writing - original draft: H.J., B.S., and S.-M.L.; Writing - review and editing: H.J. and S.-M.L.; Supervision: Y.J.L. and S.-M.L. All Authors read and approved the final version of the manuscript.

Acknowledgements

The Authors thank the staff of the Department of R&D, Hulux, for technical assistance with animal procedures and histological processing.

Funding

This work was supported by the Technology Innovation Program (RS-2025-02311948, Development of Multi-Organ Chip Culture Automation and Multimodal Analysis System for Non-Clinical Drug Evaluation) funded by the Ministry of Trade, Industry & Resources (MOTIR, Korea).

Artificial Intelligence (AI) Disclosure

During the preparation of this manuscript, a large language model (ChatGPT, OpenAI) was used solely for language editing and stylistic improvements in select paragraphs. No sections involving the generation, analysis, or interpretation of research data were produced by generative AI. All scientific content was created and verified by the Authors. Furthermore, no figures or visual data were generated or modified using generative AI or machine learning-based image enhancement tools.