The total protein from cells were lysed with pre-cooled radioimmunoprecipitation (RIPA, Cat no. P0013B, Beyotime Biotechnology Co., Ltd. Shanghai, China) buffer, containing protease inhibitor cocktail (Cat no. P9599, F. Hoffmann-La Roche AG, Basel-Stadt, Switzerland). Then, 30 µg of total protein was separated by SDS-PAGE (Cat no. LK301, Shanghai Epizyme Biomedical Technology Co., Ltd., Shanghai, China) and transferred onto polyvinylidene fluoride membrane (MilliporeSigma, MA, USA). After blocking with 5% non-fat milk for 1 h, membranes were respectively incubated with primary antibodies against Alpha-Smooth Muscle Actin ($\alpha$-SMA) (1:1000, cat no. A7248, ABclonal Technology, Woburn, MA, USA), ASC/TMS1 (1:1000, cat no. A16672, ABclonal Technology, Woburn, MA, USA), IL-1$\beta$ (1:1000, cat no. A25874, ABclonal Technology, Woburn, MA, USA) and Caspase-1 (1:1000, cat no. A25308, ABclonal Technology, Woburn, MA, USA) at 4 °C overnight. The $\beta$-actin (1:10,000, cat no. 20536-1-AP, Proteintech Group, Inc, Rosemont, IL, USA) or $\beta$-Tubulin (1:10,000, cat no. ABL1030, Abbkine Scientific Co., Ltd., Wuhan, Hubei, China) was served as internal control. Then, the corresponding secondary antibodies of HRP-conjugated Goat Anti-Mouse IgG (1:2000, cat no. SA00001-1, Proteintech Group, Inc, Rosemont, IL, USA) or HRP-conjugated Goat Anti-Rabbit IgG (1:2000, cat no. SA00001-2, Proteintech Group, Inc, Rosemont, IL, USA) were added and incubated for 1 h at RT. The signals were visualized and the band intensities were quantified by densitometry using ImageJ Software version 1.6 (National Institutes of Health, Bethesda, MD, USA).

Mouse skin tissues were homogenized in TRIzol (Takara Bio Inc., Kusatsu, Shiga, Japan) using a Bead beater (Biospec), whereas cells were directly resuspended in TRIzol (Takara Bio Inc., Kusatsu, Shiga, Japan). Total RNAs were isolated and reverse transcribed into complementary DNA using the Revert Aid First Strand cDNA Synthesis Kit (Roche, F. Hoffmann-La Roche AG, Basel-Stadt, Switzerland) according to the manufacturer’s instructions. The primers were synthetized as in Table 1 (Sangon Biotech Co., Ltd., Shanghai, China). RT–qPCR was performed in triplicate using SYBR green master mix (Roche, F. Hoffmann-La Roche AG, Basel-Stadt, Switzerland) on a Step One Plus Real-Time PCR System (Applied Biosystem, Thermo Fisher Scientific, Waltham, MA, USA). Samples with a low yield of RNA were predetermined and excluded. Results were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and the comparative $\mathrm{\Delta }$$\mathrm{\Delta }$CT method was used to determine the quantification of gene expression. The information of primers in Table 1.

Total RNA was extracted from mice at 2rd day post-wound, and ribosomal RNA (rRNA), constituting $>$90% of total RNA, was depleted using a standard kit to enrich messenger RNA (mRNA). Eukaryotic mRNA was further enriched via Oligo (dT) magnetic bead selection for polyadenylated transcripts. This polyA-selected mRNA was fragmented ultrasonically. First-strand cDNA synthesis used fragmented mRNA, random hexamer primers, and M-MuLV Reverse Transcriptase. Following RNA template degradation (RNase H), second-strand cDNA synthesis was performed with DNA Polymerase I and dNTPs. Purified double-stranded cDNA (dscDNA) underwent end repair, 3^′ adenylation (A-tailing), and ligation of standard Illumina sequencing adapters. cDNA fragments of ~200 bp were size-selected using AMPure XP beads (Beckman Coulter Life Sciences, Indianapolis, IN, USA), PCR amplified, and purified again with AMPure XP beads to generate the final library.

Stringent quality control was applied: RNA integrity and DNA contamination were assessed by agarose gel electrophoresis; RNA purity (OD260/280, OD260/230) was measured using a NanoPhotometer spectrophotometer (Implen GmbH, Munich, Germany); RNA concentration was precisely quantified using a Qubit 2.0 Fluorometer (Thermo Fisher Scientific, Waltham, MA, USA); and RNA integrity was accurately determined using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA).

Libraries were sequenced on the Illumina NovaSeq platform (Illumina, Inc., San Diego, CA, USA) using high-density flow cells. This system delivers high throughput (yielding up to 6 Tb and 20 billion reads per run), enabling sensitive detection of coding and diverse non-coding RNAs, even in challenging samples. The platform facilitates multiplexed sequencing of up to 400 transcriptomes per run.

Data are presented as Mean $\pm$ SD. One-way ANOVA with Tukey’s post-hoc test was used for comparisons among three groups or more. For repeated measurement data or two-factor factorial design, Two-way ANOVA or Dunnett’s multiple comparison test was used (GraphPad Prism, version 10, GraphPad Software, Inc., San Diego, CA, USA). Significance was set at p 0.05.

Topical application of 1% quercetin significantly enhanced wound healing in the murine excisional model, and topical application of 1% TGF-$\beta$ was used as a positive control, because the factor would regulate multiple phases including inflammation, proliferation, and remodeling during wound healing and stimulate excessive collagen deposition and myofibroblast differentiation during scar formation [29, 30]. By day 6, quercetin-treated wounds achieved significantly smaller area, showing as 0.028 $\pm$ 0.003 cm² closure compared with 0.041 $\pm$ 0.002 cm² in the vehicle control group and 0.034 $\pm$ 0.001 cm² in the TGF-$\beta$ group (Fig. 1A,B, p 0.05). Histological analysis (H&E staining) confirmed reduced inflammatory cell infiltration and accelerated granulation tissue formation in quercetin-treated wounds compared with the other two groups (Fig. 1C), and the epidermis thickness of quercetin treatment was significantly smaller than the control group based on H&E staining (Fig. 1E, p 0.05). In addition, Ki-67 immunohistochemistry showed an increase in proliferating cells within the quercetin-treated wound bed on day 6, localized predominantly to the epidermal margins and dermal papillae (Fig. 1D,F, p 0.05).

Fig. 1.

Quercetin accelerates early wound closure in Mice. (A,B) Photographs of the gross appearance of wound healing in mice and the statistical analysis. (C) HE staining of wound on the 2nd, 5th and 28th day among the three groups. (D) IHC staining of Ki-67 on the 6th day among the three groups. (E) Statistical analysis of epidermis thickness of wound in mice according to HE staining. (F) Quantitative analysis of Ki-67 expression in wound among the three groups. Scale bar = 500 µm or 50 µm. * p 0.05, ** p 0.01, *** p 0.001, and ns, not significant when compared between each group. n = 3~5 in each animal group. The data are shown as the mean $\pm$ SD. One-way ANOVA for (E,F) and Two-way ANOVA for (B) were used for multiple group comparisons. IHC, Immunohistochemistry.

To further elucidate potential targets and signaling pathways involved in quercetin-mediated wound healing, wound tissue was harvested from the vehicle control group and 1% quercetin treated group on the 2nd day, and analyzed with RNA sequencing (RNA-seq). According to previous study, day 2 (approximately 48 hours) post-wounding is a well-suited time point for analyzing the inflammatory response during wound healing, as it captures peak neutrophil and early macrophage activity, along with cytokine and growth factor expression [31]. Considering that the strong anti-inflammatory capability of quercetin, we hypothesized that the promotion of skin healing by quercetin may be related to its anti-inflammatory properties. As shown in Fig. 2A, a total of 21,991 differentially expressed genes were screened out, of which 1157 genes were upregulated and 1254 genes were downregulated, as evidenced by the volcano plot. Furthermore, the enrichment analysis results from Go term, KEGG pathways and Reactome pathways indicated that the differentially expressed genes primarily affected the inflammatory response, such as NLRP3 inflammasome complex, inflammasome-mediated signaling pathway and so on (Fig. 2B–D). Heating map and chord diagram suggested that the inflammation-related genes of NLRP1B, DDX3X, LFI213, IRGM2 and PYCARD were significantly down-regulated in quercetin treated group when compared with the vehicle control group, as depicted in Fig. 2E,F. Taken together, the application of quercetin may regulate the inflammatory response during wound healing.

Fig. 2.

Quercetin regulates the inflammatory response during early wound closure in mice. (A) Volcano plot illustrating the differentially expressed genes between the vehicle control group and the quercetin treated group. (B–D) Enrichment analyses of GO term, KEGG pathways, and Reactome pathways comparing the two groups. Inflammatory response related pathways are highlighted with red boxes in the GO terms. (E,F) Heat map and chord diagram depicting the inner relationships among differentially expressed genes. n = 3 in each animal group.

Based on the transcriptomic data, we further validated that quercetin treatment markedly attenuated inflammation during the early phase of wound repair. The results of immunofluorescence in mice revealed significant reduction in IL-1$\beta$ and caspase-1 in wound lysates on day 2 when compared with the vehicle control group and the TGF-$\beta$ treatment group, as depicted in Fig. 3A,B (p 0.05). To further explore the anti-inflammatory effect, BJ and HaCat cell was used for in vitro study. HaCaT acting as keratinocytes model the epidermal layer, while BJ acting as fibroblasts represented the dermal layer, reflecting their respective roles in skin structure and wound healing. They were treated with varying doses of quercetin and the CCK-8 result suggested the drug was safe at the concentration ranging from 10 µM to 200 µM for HaCat cell, while the safe dosage was 50 µM for BJ cell, as shown in Fig. 3C (p 0.05). In BJ cells, with the safe dose of quercetin at 50 µM, the protein expression of ACS (p 0.05 ), Caspase-1 (p 0.05), and IL-1$\beta$ (p 0.05) were significantly lower than the NC group (Fig. 3D,E). In HaCat cells, within the safe dose ranging from 10 µM to 200 µM, the protein expression of ACS was only lower than the NC group at 200 µM (p 0.05), while Caspase-1 was lower than the NC group at 10 µM (p 0.05), 50 µM (p 0.05), 100 µM (p 0.05), and 200 µM (p 0.05). IL-1$\beta$ was lower than the NC group at 10–200 µM (Fig. 3F,G, p 0.05). Therefore, except for ACS in HaCat cells, quercetin at 50 µM significantly inhibited the expression of inflammatory proteins in both BJ and HaCat cells, although a dose-response trend was observed across all tested concentrations. In addition, the results of qRT-PCR are different with the results of western-blot analysis in Fig. 3H,I. Under high concentration stimulation of quercetin, the inhibition of IL-1$\beta$ protein levels in keratinocytes is mainly due to its inhibition of inflammatory signaling pathways (NF-$\kappa$B, NLRP3) and receptors (such as IL1R1), blocking protein maturation and secretion. And gene expression may increase due to imbalanced transcription factor networks (such as YY1 inhibition release), epigenetic modifications, or negative feedback regulation. At the same time, keratinocytes have specific regulatory mechanisms, and certain stimuli can induce IL-1B gene transcription without protein secretion. Quercetin may amplify this transcription translation decoupling effect. Therefore, quercetin can effectively inhibit the inflammatory response during the skin healing process in vivo and in vitro.

Fig. 3.

Suppression of pro-inflammatory cytokines in quercetin-treated wound healing. (A,B) IF staining and quantitative analyses of CASPASE-1 and IL-1$\beta$ on the 2nd day of wound healing among the three. (C) BJ and HaCat cell vitality analyses when treated with quercetin at gradient increasing concentration. (D–G) Western-blot analysis of ACS, CASPASE-1, IL-1$\beta$ in BJ and HaCat cells when treated with quercetin in a dose-response manner (normalized to internal $\beta$-actin). (H,I) Analysis of gene expression of pro-inflammatory cytokines: IL-1$\beta$, IL-6 and IL-8 in BJ and HaCat cells treated with quercetin. Scale bar = 500 µm. * p 0.05, ** p 0.01, *** p 0.001, **** p 0.0001, and ns, not significant when compared between each group. The data are shown as the mean $\pm$ SD. n = 3~5 for each animal group. Two-way ANOVA with Dunnett multiple comparison test were used for multiple group comparisons. IL, Interleukin; IF, Immunofluorescence; NC, negative control; DAPI, 4^′,6-diamidino-2-phenylindole.

According to the wound healing image on day 28 (Fig. 4A), the positive drug TGF- $\beta$1 can indeed promote early wound healing. However, we found that on day 28, the pigmentation and scar formation at the wound site were significantly greater than those in the quercetin group, moreover even exceeded the control group. The scars in the quercetin group were the lightest, so our future research will investigate whether quercetin inhibits scar formation during tissue remodeling. First, we measured the scar size of the control group, quercetin group, and TGF-$\beta$1 group on day 28. We found that the scar length was reduced in the quercetin group (Fig. 4B, p 0.05). In addition, Masson’s trichrome staining demonstrated that quercetin improved collagen architecture during the remodeling phase (day 28, Fig. 4C). The collagen III/I ratio in quercetin-treated scars was significantly higher than in the control group (Fig. 4D,F, p 0.05). Additionally, $\alpha$-SMA, as a marker of myofibroblasts and scars, showed significantly lower expression in the quercetin-treated group than the control group, which detected by immunofluorescence in the dermis (Fig. 4E,G, p 0.05).

Fig. 4.

Quercetin modulated collagen deposition and improved scar quality. (A,B) Photographs of scar length and statistical analysis among three groups. (C) Masson staining of wound quality on the 28th day among three groups. (D,E) IF staining of Collagen I/III and $\alpha$-SMA in wound on the 28th day among three groups. (F,G) Statistical analysis of Ki-67 and ration between Collagen III/I among three groups. Scale bar = 500 µm or 50 µm. * p 0.05, ** p 0.01, and *** p 0.001 when compared between each group. The data are shown as the mean $\pm$ SD. n = 3~5 for each group. Two-way ANOVA for (B,F,G) were used for multiple group comparisons. $\alpha$-SMA, alpha-smooth muscle actin.

To further investigate the differentiation effect of quercetin on fibroblasts into myofibroblasts, the BJ cells were treated with quercetin only and quercetin with TGF-$\beta$1. As depicted in Fig. 5A–C, low concentration of quercetin (20–50 µM) effectively suppressed TGF-$\beta$1-driven myofibroblast differentiation, showing as decrease in $\alpha$-SMA expression of genes and protein, while the high concentration (100 µM) showed the opposite effect as promoting myofibroblast differentiation (p 0.05). Since HaCaT cells acting as keratinocytes model the epidermal layer, while BJ acting as fibroblasts represented the dermal layer, reflecting their respective roles in skin structure and wound healing. During skin wound healing and scar formation, increased migration of HaCaT cells reflects enhanced epidermal re-epithelialization, while elevated collagen production by BJ fibroblasts indicates dermal matrix remodeling, consistent with their roles in skin repair. A similar trend was found in the scratch wound assay that, in BJ and HaCat cells, quercetin attenuated BJ fibroblast migration, and increased the proliferation of HaCat cells, with the statistical difference, as shown in Fig. 5D–G (p 0.05). These results are consistent with Fig. 1, where quercetin promotes Ki-67 expression in the early stages of wound healing. On the one hand, it promotes wound healing by inducing proliferation, and on the other hand, it alleviates scar formation by inhibiting BJ fibroblast differentiation into myofibroblasts.

Fig. 5.

Quercetin inhibited TGF-$\beta$1-Induced myofibroblast differentiation. (A,B) Western-blot analysis of $\alpha$-SMA expression when TGF-$\beta$1-induced BJ cells are treated with quercetin. (C) Gene expression of ACTA2 in TGF-$\beta$1-induced BJ cells treated with quercetin. (D,E) Scratch wound assay of BJ and HaCat cells treated with quercetin in a dose-response manner. (F,G) Statistical analysis of the scratch wound assay of BJ and HaCat cells. Scale bar = 50 µm. * p 0.05, ** p 0.01, and **** p 0.0001, and ns, not significant when compared between each group. The data are shown as the mean $\pm$ SD. One-way ANOVA for (B,C) or Two-way ANOVA for (F,G) was used for multiple group comparisons. TGF-$\beta$, transforming growth factor beta.