Sensitizer‐Induced Basophils Accelerate Skin Re‐Epithelialization via IL‐4/IL‐13‐Mediated Macrophage Polarization

ABSTRACT

Background

Atopic dermatitis (AD) patients exhibit a paradox of impaired skin barrier with intense itching, yet often demonstrate rapid re‐epithelialization after scratching. While basophils are key effector cells in allergic inflammation, their role in the subsequent tissue repair remains unexplored.

Objective

We investigated whether basophils, recruited during sensitized skin responses, contribute to wound healing.

Methods

Using single‐cell RNA sequencing, flow cytometry, and immunofluorescence, we mapped basophil infiltration and activation in sensitizer (oxazolone)‐induced skin injury models. We employed genetic (Mcpt8 ^CT/+ R26 ^DTA/+ ) and antibody‐mediated (anti‐FcεRI) basophil depletion, as well as basophil‐specific Il4/Il13 knockout mice (Il4/13 ^fl/fl Mcpt8 ^CT/+) to define their functional contribution. Macrophage polarization was assessed by flow cytometry, RT‐qPCR, and immunohistochemistry.

Results

We identified a significant enrichment of basophils in wounded skin, peaking at day 3–6 post‐injury. Sensitizer‐challenge enhanced basophil recruitment and accelerated wound closure, angiogenesis, and re‐epithelialization. Depletion of basophils severely impaired these repair processes. Mechanistically, basophils were the predominant source of IL‐4 and IL‐13 in the early wound microenvironment. These cytokines were essential for driving macrophage polarization toward a pro‐repair M2 phenotype. Loss of IL4/IL13 specifically in basophils phenocopied the healing defects observed in basophil‐deficient mice, and this could be rescued by local cytokine administration.

Conclusion

Our study uncovers a novel pro‐repair function of basophils in sensitizer‐exposed skin. Beyond their well‐known role in provoking itch and inflammation, basophils are critical for initiating type 2 immune‐mediated tissue regeneration via IL‐4/IL‐13‐dependent macrophage reprogramming. This axis represents a promising therapeutic target for chronic wounds, particularly in the context of allergic skin disorders.

Oxazolone exposure drives rapid basophil recruitment to wounded skin, accelerating re‐epithelialization, angiogenesis, and tissue regeneration. Basophils are the dominant early source of IL‐4 and IL‐13 in the wound microenvironment. Basophil‐derived IL‐4/IL‐13 promote macrophage polarization toward a pro‐repair M2 phenotype, enabling efficient wound healing. Baso, basophil; EtOH, ethanol; IL, interleukin; Mcpt8, mast cell protease 8; mDC, mature dendritic cell; Mo, monocyte; МФ, macrophage; Neu, neutrophil; OXA, oxazolone.

1. Introduction

Patients with atopic dermatitis (AD) experience a chronic itch–scratch cycle that perpetuates epidermal barrier disruption and cutaneous inflammation and is therefore generally expected to impair efficient tissue repair [1, 2, 3]. Consistent with this view, most mechanistic studies in AD emphasize defective barrier function and dysregulated inflammatory responses [3, 4]. Nevertheless, in routine clinical practice, scratch‐induced epidermal injury in AD most commonly manifests as superficial excoriations that re‐epithelialize rather than progressing to non‐healing or ulcerative lesions. These observations raise the possibility that, despite persistent inflammation, the atopic skin microenvironment retains context‐dependent mechanisms that permit effective epithelial repair after injury.

Understanding this apparent disconnect requires consideration of the multifaceted roles of type 2 immune responses in the skin. Type 2 cytokines, particularly interleukin (IL)‐4 and IL‐13, are central drivers of AD pathogenesis and contribute to barrier dysfunction and chronic inflammation [3, 4]. Basophils, as rapid responders and early sources of IL‐4 and IL‐13, are consistently enriched in AD lesions and play a key role in orchestrating type 2 inflammation [5, 6]. Notably, basophils in AD may play dual roles—promoting barrier dysfunction while also contributing to immune regulation and resolution of inflammation—as suggested by Pellefigues et al. [5]. In particular, IL‐4 and IL‐13 promote alternative (M2) macrophage polarization, a phenotype closely associated with the resolution of inflammation and tissue regeneration [7].

These findings suggest that type 2 immune responses in AD may exert context‐dependent effects, contributing to both pathology and repair. Given that basophils are a primary early source of IL‐4/IL‐13 in allergic settings [5, 6], we hypothesized that basophils might serve a dual, beneficial function in atopic skin by initiating a pro‐repair program following injury. To test this, we employed a sensitizer‐induced skin injury model that recapitulates robust basophil recruitment and type 2 immune activation, while partially modeling features of chronic AD pathogenesis [8]. Using this approach, we sought to determine whether basophils are recruited to sites of epidermal injury and whether they facilitate wound repair through IL‐4/IL‐13–dependent macrophage polarization.

2. Materials and Methods

2.1. Animals and Basophil Depletion Strategies

Mcpt8 ^CT/+ (Mcpt8 ^CT/+ , C57Bl/6 background, Cat. No. NM‐KI‐200006), Il4/Il13‐flox (Il4/Il13 ^fl/fl , C57Bl/6 background, Cat. No. NM‐CKO‐200169) mice were purchased from Shanghai Model Organisms Center Inc. Rosa26 ^EGFP‐DTA (R26 ^DTA/+ , C57Bl/6 background, Cat. No. C001334) were purchased from Cyagen Biosciences Inc. Mcpt8 ^CT/+ R26 ^DTA/+ , Il4/Il13 ^fl/fl Mcpt8 ^CT/+ were generated and maintained at the animal facility of Xi'an Jiaotong University Medical School. Cre‐ERT2 animals were treated for 5 consecutive days with 2.5 mg tamoxifen dissolved in corn oil i.p. (Sigma‐Aldrich) to induce Cre recombinase activity.

To interrogate the functional contribution of basophils, we employed complementary genetic and antibody‐based approaches that achieve functional basophil depletion within this experimental context [9, 10, 11]. For long‐term ear wound‐healing experiments, tamoxifen was re‐administered continuously for three consecutive days starting on post‐injury day 10 to maintain effective recombination and sustained basophil depletion throughout the observation period. For antibody‐mediated basophil depletion, 5 μg anti‐FcεRI antibody (eBiosciences, clone MAR‐1) was injected i.v. on pre‐injury days −2, −1, and 0, and repeated on post‐injury days 3, 6, and 8 in order to maintain reduced levels. All animal experiments were approved by the Research Ethics Committee of the Second Affiliated Hospital of Xi'an Jiaotong University (approval number: XJTUAE2023‐1018) and were conducted in accordance with institutional guidelines for animal care and use. The validation of the specificity of the basophil‐targeting models was performed, and the details are in the Supporting Files.

2.2. Full‐Thickness Skin Wound and Ear Pinna Injury

Mice 8–10 weeks old were used in the experiments. For ear wounds, the skin was prepared using alcohol wipes. Punch wounds 2 mm in diameter were created using a thumb‐type metal ear punch (Fisherbrand). Topical applications of 5 μL anhydrous ethanol or 5% oxazolone (OXA) were administered to the left and right ears of each mouse on post‐injury days 0, 1, and 2 to induce the sensitized skin [12]. For dorsal wounds, hair was removed from the entire dorsum using an electric shaver followed by depilatory cream. The dorsal skin was then prepped using three sequential alternating swabs of betadine and 70% ethanol. Using sharp surgical scissors, a circular full‐thickness skin lesion (10 mm in diameter) was inflicted on the dorsal region of the mice. For topical applications, mice were treated topically with recombinant IL‐4 and/or IL‐13 (10 μg/mouse/day, MedChemExpress, China).

The other detailed methods are described in detail in the Supporting Methods.

3. Results

3.1. Basophils Accumulate at Wound Sites Following Injury

To analyze the change of immune cell composition after skin wound, we performed a single‐cell RNA sequencing (scRNA‐seq) dataset anlysis of skin leukocytes collected on day 4 after injury [13]. After batch correction and integration, we identified 15 immune clusters, including monocytes and monocyte‐derived macrophages (M/MdM), conventional dendritic cells (cDC), M2 macrophages (M2 Mac), neutrophils (Neu), dendritic epidermal T cells (DETCs), dermal γδ T17 cells (dγδT17), NK cells, CD4⁺ T cells, Treg cells, mature dendritic cells (mDC), monocytes (Mono), Langerhans cells (LC), basophils (Baso), plasma DC (pDC), Mast cells (Figure 1A), according to our previously reported gene markers [12]. Dotplot and marker gene lists validated the cell types (Figure S1A and Table S1). Since AD has enriched basophils [14] and our previous research showed enriched basophils after OXA treatment [12], to preliminarily investigate the role of basophils in wound healing, we also reanalyzed our previously published scRNA‐seq dataset of OXA‐treated skin with the same cell identities maintained [12] (Figure 1B).

FIGURE 1.

Basophils accumulate in the skin during wound healing. (A) UMAP plot of single‐cell RNA‐sequencing (scRNA‐seq) data showing immune cell clustering in wounded skin tissue. (B) UMAP plot of scRNA‐seq data showing immune cell clusters in oxazolone (OXA)‐sensitized skin tissue. (C) Bar plot comparing the relative proportions of immune cell subsets between wounded and unwounded skin. (D) Bar plot comparing the relative proportions of immune cell subsets between OXA and ethanol (ETOH) control skin. (E) Representative flow cytometry plots identifying basophils in skin wounds. (F) Kinetics of basophil accumulation from day 0 to day 8 post‐injury, shown as both percentage of CD45⁺ cells and absolute cell counts (n = 6–9 per group for d0‐d8), one‐way ANOVA following unpaired two‐tailed Student's t test. (E) qPCR analysis of Mcpt8 expression in skin at indicated time points (n = 8 per group); unpaired t test. (F) Immunofluorescence staining of basophils (MCPT8⁺ cells) in skin tissue at day 0 and day 3 post‐injury, bar = 100 μm; quantification of basophil numbers is shown in the right panel (n = 6 per group for d0 and d3), unpaired t test. All data were obtained from three independent experiments and shown as means ± SEM. *p < 0.05, **p < 0.01, ****p < 0.0001.

By analyzing the changes in cell proportions between the wounded and unwounded groups, we observed that Mono, Baso, Mast cells, and pDC were specifically present in the wounded group (Figure 1C), demonstrating their potential involvement in the wound response. In contrast, upon OXA‐induced skin sensitization, basophils became the most dominant immune population, displaying the most pronounced fold‐change increase (Figure 1D). Importantly, basophils uniquely expressed high levels of IL‐4 and IL‐13 (Figure S1B), whereas other expanded populations lacked a comparable type 2 cytokine signature, providing a functional rationale for subsequent mechanistic analyses. We therefore hypothesized that basophils may play an underappreciated but functionally important role in modulating the wound‐healing process.

To further investigate whether basophils are recruited during wound healing, we used a full‐thickness skin wound model and examined basophil infiltration over time by flow cytometry. Basophils were identified as CD45⁺Lin⁻CD117⁻CD49b⁺IgE⁺ cells [15]. Flow cytometry analysis revealed that both the proportion and number of basophils significantly increased in injured skin after wounding, peaking at day 6 post‐injury (Figure 1E,F). RT‐qPCR showed elevated Mcpt8 expression (Figure 1G). Immunofluorescence staining further demonstrated increased MCPT8⁺ cells in the wound area (Figure 1H). Since multiple cytokines (IL‐18, IL‐33, thymic stromal lymphopoietin (TSLP)) mediate basophil homing and activation, triggering chemokines (CCL2 and CCL5) production and release [16], we also checked whether these genes are upregulated. Our qPCR data showed significant upregulation of these cytokines in injured skin on day 3 post‐injury (Figure S1C), which potentially act as supplementary stimuli for basophil recruitment and activation during wound healing [17]. These results demonstrate that basophils are actively recruited and functionally activated during skin wound.

3.2. Basophil Accumulation Correlates With Accelerated Wound Closure

Given the close developmental and functional association between mast cells and basophils, we also focused our previous analysis on these two cell types (Figure 1B). UMAP visualization showed that basophils and mast cells clustered in close proximity (Figure S2A). Furthermore, feature plots confirmed their distinct transcriptional identities, with mast cells marked by Mcpt4 ⁺ and Kit ⁺ expression, and basophils by Mcpt8 ⁺ and Itgam ⁺ (Figure S2B). Quantitative analysis revealed a significant enrichment of basophils, but not mast cells, in OXA‐treated skin (Figure S2C). Gene Ontology (GO) analysis of basophil‐specific upregulated genes showed enrichment in wound healing‐related pathways (Figure S2D), demonstrating a potential role for basophils in promoting tissue repair.

To functionally validate this association, we preferred the ear punch model and employed OXA to enhance the wound‐recruiting capacity of basophils. We induced basophil infiltration by topically applying high‐concentration OXA (5%) to mouse ears wound sites on days 0, 1, and 3 after ear punch, thereby promoting local accumulation while minimizing chronic inflammation (Figure 2A). Flow cytometry confirmed a selective increase in basophils, with no comparable recruitment of mast cells (Figure S2E–I). Macroscopic examination revealed faster wound closure in OXA‐treated right ears compared to ethanol‐treated contralateral controls (Figure 2B). Ear thickness measurements indicated peak inflammation at day 4, followed by progressive reduction and wound closure (Figure 2C,D). Quantitative assessment of ear hole area and closure rate further confirmed that OXA‐treated ears exhibited smaller residual wound areas and faster healing kinetics (Figure 2D,E). By day 21, OXA‐treated wounds demonstrated significantly improved closure relative to controls (Figure 2F). Histological analysis corroborated these findings, showing reduced wound gap diameters (Figure 2G,H) and enhanced tissue regeneration at the wound margins at day 21 (Figure 2I). Notably, OXA treatment also promoted hair follicle regeneration within the healed tissue (Figure 2J).

FIGURE 2.

OXA enhances wound healing in both ear punch and dorsal excision models. (A) Schematic of the ear punch wound healing model with OXA or EtOH treatment. (B) Representative images of ear wounds from day 0 to day 21 under OXA and EtOH treatment. (C–E) Quantification of ear thickness (C), wound area (D), and percent wound closure (E) over time in OXA‐ and EtOH‐treated mice (n = 7 per group); one‐way ANOVA following unpaired two‐tailed Student's t test. (F) Comparison of residual ear hole area at day 21 between groups (n = 7 per group); unpaired t test. (G) H&E staining of ear tissue at day 21 showing histological differences in tissue regeneration; 4× higher magnification. (H–J) Quantification of ear hole diameter (H), extension length of regenerated tissue (I), and number of regenerated hair follicles (J) in each group (n = 4 per group); unpaired t test. (K) Schematic of dorsal full‐thickness wound healing model with OXA or EtOH treatment. (L) Representative images of dorsal wounds during healing. (M) Quantification of wound area and percent closure over time (n = 6 per group); one‐way ANOVA following unpaired two‐tailed Student's t test. (N–O) H&E staining at day 8 (N), wound gap size (O), and regenerated tissue length (O) in dorsal wounds under each condition (n = 5 per group); unpaired t test. All data were obtained from three independent experiments and shown as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

To determine whether elevated circulating basophils contribute to wound‐site accumulation and accelerate healing, we employed a modified approach. To avoid the confounding effects of direct OXA application on open wounds, including exaggerated inflammation and scratching, we treated mouse abdominal skin with OXA after inflicting dorsal skin wounds (Figure 2K). Based on our previous findings that basophil recruitment to injured skin peaked at day 6 (Figure 1C,D), flow cytometry was performed on day 6 to quantify the increase in circulating basophils and those infiltrating the injured dorsal skin after OXA treatment of the abdominal skin. Flow cytometry revealed that OXA treatment significantly elevated both the proportion and absolute number of circulating basophils, which corresponded with increased basophil infiltration at wound sites (Figure S2J–M). Consistent with earlier observations, OXA‐treated mice showed enhanced macroscopic wound closure compared to ethanol‐treated controls (Figure 2L,M). Histological analysis confirmed smaller wound gaps and improved re‐epithelialization in the OXA group (Figure 2N,O). Collectively, these findings demonstrate that OXA‐induced basophil accumulation, both locally and systemically, facilitates wound healing.

3.3. Basophils Are Functionally Required for Efficient Wound Healing

To elucidate the functional contribution of basophils to OXA‐enhanced wound healing, we generated basophil‐deficient mice (Mcpt8 ^CT/+ R26 ^DTA/+ ) [11]. Immunofluorescence, PCR, and flow cytometry confirmed effective basophil ablation in this model (Figure S3A–D), while the numbers and distribution of immune cell populations remained unaffected (Figure S4). In the absence of OXA sensitization, basophil‐deficient mice exhibited ear punch wound‐healing kinetics comparable to those of control mice, likely reflecting the limited size of the circular wounds, which may be insufficient to recruit substantial numbers of basophils under steady‐state conditions. In contrast, under OXA treatment, basophil‐deficient mice displayed significantly delayed ear wound closure compared with control animals (Figure 3A,B). Consistent with this finding, measurement of ear thickness showed that OXA treatment alone did not ameliorate the increased swelling observed in basophil‐deficient mice (Figure 3C). By day 21 post‐injury, OXA‐treated basophil‐deficient mice retained significantly larger wound areas than their control counterparts (Figure 3D). Histological analyses further revealed enlarged wound diameters and impaired tissue regeneration at the wound margins in basophil‐deficient mice (Figure 3E–G).

FIGURE 3.

Basophil depletion delays wound healing. (A) Representative images of ear punch wounds in basophil‐deficient and control mice treated with OXA or EtOH. (B, C) Quantification of ear hole area and wound closure percentage (B), and ear thickness (C) across treatment groups (n = 7 per group). (D) Residual ear hole area at day 21 post‐injury (n = 7 per group). (E) H&E staining of ear tissue at day 21 post‐injury; 4× higher magnification. (F, G) Quantification of ear hole diameter (F) and regenerated tissue length (G) (n = 7 per group). (H) Representative images of dorsal wound healing in basophil‐deficient and control mice. (I) Quantification of dorsal wound area and percent closure over time (n = 7 per group). (J, K) H&E staining (J) and regenerated tissue length (K) of dorsal wounds at day 3, 6, and 8 (n = 4–5 per group). (L, M) H&E staining (L) and wound gap measurements (M) at day 10 post‐injury (n = 5 per group). All above statistic tests are one‐way ANOVA following unpaired two‐tailed Student's t test. (N–R) Immunohistochemical and histological analyses at day 10 post‐injury, including Masson's trichrome, Ki67, α‐SMA, and CD31 staining; quantification of collagen deposition (O), proliferative activity (P), myofibroblast activation (Q), and neovascularization (R) (n = 5 per group); unpaired t test. All data were obtained from three independent experiments and shown as means ± SEM. *p < 0.05, **p < 0.01, and ***p < 0.001.

To determine whether basophil accumulation at wound sites represents bystander infiltration or active functional involvement, we generated dorsal skin wounds in both basophil‐deficient mice and wild‐type mice treated with the basophil‐depleting MAR‐1 antibody (anti‐FcεRI) [10]. To ensure complete depletion, MAR‐1 antibodies were administered three times before and three times after wounding (Figure S5A). As expected, anti‐FcεRI treatment resulted in near‐complete elimination of basophils from both peripheral blood and wounded skin (Figure S5B,C). Although FcεRI is also expressed on mast cells and certain DC subsets, we observed no significant differences at wound sites between MAR‐1–treated and isotype control groups (Figure S5D,E).

Both genetic basophil‐deficient mice and MAR‐1–treated mice exhibited significantly impaired wound healing, characterized by larger wound areas and delayed closure (Figure 3H,I; Figure S6A). Histological evaluation revealed delayed re‐epithelialization and increased wound gaps in both models (Figure 3J–M; Figure S6B). Immunohistochemical analyses further showed reduced angiogenesis (CD31), diminished myofibroblast activation (α‐SMA), decreased cellular proliferation (Ki67⁺), and no significant change in collagen deposition—factors all closely linked to effective wound repair (Figure 3N–R; Figure S6C,D).

Together, these findings demonstrate that basophils are indispensable for efficient wound healing, contributing to critical regenerative processes such as angiogenesis, fibroblast activation, and epithelial proliferation.

3.4. Basophil Depletion Alters Macrophage Polarization During Wound Healing

To explore the mechanisms by which basophils facilitate wound repair, we examined the immune cell landscape in wounded skin from basophil‐sufficient and ‐deficient mice. We focused on neutrophils, monocytes/macrophages, dendritic cells, B cells, and T cells, the key immune populations orchestrating tissue healing. Flow cytometry with detailed subpopulation gating (Figure S7A,B) revealed that there was a significant reduction in Ly6C^hi monocytes and alternatively activated (M2) macrophages (Figure 4A). In contrast, no notable differences were observed in dendritic cells, neutrophils, total monocyte/macrophage numbers, classically activated (M1) macrophages, B cells, or T cells (Figure 4A).

FIGURE 4.

Basophil deficiency alters immune cell composition and macrophage polarization in wounded skin at day 6 post‐injury. (A) Flow cytometry analysis comparing frequencies of neutrophils, dendritic cells (DCs), total monocytes/macrophages, Ly6C^hi monocytes, CD4⁺ T cells, CD8⁺ T cells, B cells, M1 macrophages, and M2 macrophages in wounded skin of basophil‐deficient and control mice (n = 4–5 per group, unpaired t test). (B) RT‐qPCR analysis showing relative expression changes of M1‐associated and M2‐associated marker genes in wound tissue (n = 5–6 per group, unpaired t test). (C) Immunofluorescence staining of M1 and M2 macrophage markers in wounded skin from basophil‐deficient and control mice, bar = 25 μm; right panel quantifies differences in marker expression (n = 4 per group; unpaired t test). All data were obtained from three independent experiments and shown as means ± SEM. *p < 0.05 and **p < 0.01.

Consistent with these findings, transcriptomic analysis of wounded skin in basophil‐deficient mice revealed a marked shift toward a pro‐inflammatory macrophage phenotype. M1‐associated genes—including Cd86, Nos2 (Inos), and Il1b—were significantly upregulated, while M2 markers such as Cd206 (Mrc1), Arg1, and Il10 were downregulated (Figure 4B) [18]. Immunofluorescence staining further corroborated these shifts: F4/80⁺CD206⁺ M2 macrophages were markedly decreased, whereas F4/80⁺CD86⁺ M1 macrophages were increased in the absence of basophils (Figure 4C).

These data demonstrate that basophils play a critical immunoregulatory role during the wound healing process by promoting a macrophage polarization bias toward the pro‐repair M2 phenotype. Notably, this altered macrophage landscape is consistent with the scRNA‐seq data, which identified M2 macrophages as the predominant subtype present in the wound microenvironment (Figure 1B). Thus, in the absence of basophils, the balance shifts further toward a sustained pro‐inflammatory state, potentially impeding resolution and tissue regeneration.

3.5. Basophil‐Derived IL‐4 and IL‐13 Are Key Regulators of Wound Healing

Since cytokines are critical mediators of wound healing [19], we sought to elucidate the molecular mechanisms by which basophils regulate monocyte/macrophage responses during skin repair. To this end, we analyzed the local expression of key cytokines (TNF, IL‐1β, IL‐6, IL‐4, IL‐13, IL‐10, and AREG) at day 4 post‐injury [19]. Initial analysis of wound tissue at day 4 revealed that Il1b, Il6, Il4, and Il13 were significantly upregulated within basophils (Figure S8A). Following treatment with OXA, we observed sustained enrichment of Il1b, Il6, Il4, and Il13 in basophils (Figure S8B), demonstrating that basophil‐derived IL‐1β, IL‐6, IL‐4, and IL‐13 may play pivotal roles in orchestrating the wound healing process.

We then examined the mRNA expression of these cytokines following basophil depletion. Despite all four cytokines (Il1b, Il6, Il4, and Il13) being upregulated during the course of wound healing, only Il4 and Il13 were significantly reduced in basophil‐deficient mice compared to controls (Figure 5A). Consistently, ELISA of wound lysates confirmed a marked decrease in the protein levels of both cytokines (Figure 5B), indicating that the production of IL‐4 and IL‐13 is largely dependent on the presence of basophils within the wound microenvironment.

FIGURE 5.

Basophil‐derived IL‐4 and IL‐13 promote wound healing. (A) Quantitative PCR analysis of Il1b, Il6, Il4, and Il13 mRNA expression levels in wounded skin (n = 8 per group); unpaired t test. (B) ELISA analysis showing IL‐4 and IL‐13 protein levels in wound tissue (n = 5 per group); unpaired t test. (C) Representative images of wound healing in basophil‐deficient mice receiving PBS or recombinant IL‐4, IL‐13, or IL‐4+IL‐13. (D) Quantification of wound area and healing rate in basophil‐deficient mice following IL‐4 or IL‐13 administration (n = 6 per group); one‐way ANOVA following unpaired two‐tailed Student's t test. (E) H&E staining of day 10 wound sections in basophil‐deficient mice with or without IL‐4/IL‐13 treatment; 4× higher magnification. (F) Measurement of wound gap width on day 10 (n = 5 per group); one‐way ANOVA following unpaired two‐tailed Student's t test. (G) Quantification of regenerated tissue length in wound beds on day 10 after cytokine treatment (n = 5 per group); one‐way ANOVA following unpaired two‐tailed Student's t test. All data were obtained from three independent experiments and shown as means ± SEM. *p < 0.05, **p < 0.01, and ***p < 0.001.

To functionally validate the role of IL‐4 and IL‐13 in tissue repair, we applied recombinant cytokines directly to the wounds of basophil‐deficient mice. Treatment with either IL‐4 or IL‐13 alone significantly accelerated wound closure compared to both vehicle‐treated (PBS), highlighting their essential roles in promoting tissue regeneration (Figure 5C,D). Notably, combined administration of IL‐4 and IL‐13 did not produce an additive effect, demonstrating functional redundancy or convergence on a shared downstream signaling pathway (Figure 5C,D). Histological analysis further confirmed restoration of tissue architecture: wounds in cytokine‐treated mice exhibited significantly smaller diameters and nearly complete epithelial regeneration at the margins, comparable to vehicle‐treated (PBS) basophil‐deficient mice (Figure 5E–G). Together, these findings identify IL‐4 and IL‐13 as key basophil‐derived mediators that facilitate effective wound healing.

3.6. Basophil‐Derived IL‐4/IL‐13 Are Essential for Macrophage Polarization and Tissue Repair

To directly assess the functional importance of basophil‐derived IL‐4 and IL‐13 in regulating macrophage polarization and promoting wound repair, we generated mice with basophil‐specific deletion of both IL‐4 and IL‐13 (Il4/13 ^fl/fl Mcpt8 ^CT/+). PCR confirmed LoxP sites insertion, showing bands of 452 and 376 bp in Il4/13 ^fl/fl Mcpt8 ^CT/+ and Il4/13 ^fl/fl mice, as well as the presence of CreER^T2 gene at 539 bp in Il4/13 ^fl/fl Mcpt8 ^CT/+ mice (Figure S9A). Flow cytometric analyses further demonstrated that this genetic manipulation did not significantly alter the overall abundance of major immune cell populations when compared with control mice, indicating that basophil‐specific cytokine deletion did not broadly perturb immune cell composition (Figure S9B). Consistent with this, ELISA measurements revealed a marked reduction in IL‐4 and IL‐13 protein levels at wound sites in these mice, validating effective depletion of basophil‐derived cytokines in vivo (Figure S9C).

We first evaluated the impact of basophil‐derived IL‐4/IL‐13 on OXA‐mediated ear wound healing. In the absence of OXA treatment, wound healing kinetics were comparable between Il4/13 ^fl/fl Mcpt8 ^CT/+ and Il4/13 ^fl/fl mice, indicating a relatively baseline course of repair (Figure 6A left). However, under OXA treatment, which promotes basophil infiltration, mice lacking IL‐4/IL‐13 specifically in basophils displayed significantly delayed wound closure compared to controls (Figure 6A,B). Ear thickness measurements showed that OXA‐induced inflammation was not alleviated by cytokine deletion, demonstrating that the impaired healing was not due to changes in the inflammatory burden per se (Figure 6C). By day 21 post‐injury, OXA‐treated basophil‐specific Il4/Il13‐deficient mice exhibited significantly larger residual wound areas, accompanied by increased wound diameters and defective tissue regeneration at the wound margins (Figure 6D–G). These findings closely mirror those observed in basophil‐deficient mice, underscoring that IL‐4/IL‐13 are the principal effector molecules through which basophils facilitate wound repair.

FIGURE 6.

Basophil‐derived IL‐4 and IL‐13 are essential for efficient wound healing. (A) Representative images of ear wound healing in basophil‐specific Il4/Il13 knockout and Il4/Il13 flox control mice treated with OXA or EtOH. (B) Quantification of ear hole area (left) and percent wound closure (right) over time in both groups (n = 6 per group); one‐way ANOVA following unpaired two‐tailed Student's t test. (C) Ear thickness measurements comparing knockout and control mice (n = 6 per group); one‐way ANOVA following unpaired two‐tailed Student's t test. (D) Residual ear hole area at day 21 post‐injury (n = 6 per group); one‐way ANOVA following unpaired two‐tailed Student's t test. (E) H&E staining of ear tissue at day 21 showing histological differences; 4× higher magnification. (F, G) Quantification of ear hole diameter (F) and regenerated tissue length (G) at day 21 (n = 5 per group); one‐way ANOVA following unpaired two‐tailed Student's t test. (H) Representative images of dorsal wound healing in knockout and control mice. (I) Dorsal wound area and percent closure over time (n = 6 per group); one‐way ANOVA following unpaired two‐tailed Student's t test. (J) H&E staining of dorsal wounds at day 10 post‐injury. (K, L) Quantification of regenerated tissue length (K) and wound gap size (L) at day 10 (n = 5 per group); unpaired t test. (M) Flow cytometry analysis of macrophage subsets (CD86⁺ M1 and CD206⁺ M2) in dorsal wounds; right panels show frequencies and absolute numbers as percentages of CD45⁺ immune cells (n = 6 per group); unpaired t test. All data were obtained from three independent experiments and shown as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

To generalize these findings, we next assessed dorsal full‐thickness skin wound healing in basophil‐specific Il4/Il13‐deficient mice. Consistent with the ear wound model, these mice exhibited significantly impaired wound closure, delayed re‐epithelialization, and larger wound gaps relative to controls (Figure 6H–L). Immunohistochemical analysis further revealed hallmarks of impaired tissue regeneration, including reduced angiogenesis (CD31), diminished myofibroblast activation (α‐SMA), and lower proliferative activity (Ki67⁺), whereas collagen deposition remained unchanged (Figure S10A,B). These findings demonstrate that basophil‐derived IL‐4/IL‐13 is crucial for coordinating multiple aspects of the regenerative response.

Since our previous data indicated that basophils influence macrophage polarization (Figure 4A), we next investigated whether basophil‐derived IL‐4 and IL‐13 contribute to this process in vivo. Using CellChat analysis of both the wound and OXA‐treated datasets, we found that OXA administration markedly enhanced interactions between basophils and macrophages, whereas such interactions were minimal in the absence of OXA treatment (Figure S10C,D). This demonstrates that OXA‐induced inflammation strengthens the crosstalk between basophils and macrophages.

To further assess macrophage polarization under these conditions, we analyzed macrophage subsets in wounded tissue. Flow cytometry revealed a significant reduction in CD206⁺ M2 macrophages—both in proportion and absolute number—while CD86⁺ M1 macrophages remained unchanged (Figure 6M). RT‐qPCR analysis showed a downward trend in M2‐associated gene expression (Arg1, Mrc1, Il10), accompanied by a mild, though not statistically significant, increase in M1‐associated markers (Cd86, Inos, Il1b) (Figure S11A). Notably, total monocyte/macrophage counts and the proportion of Ly6C^hi monocytes were unaffected (Figure S11B), indicating that basophil‐derived cytokines selectively regulate macrophage polarization rather than recruitment.

Collectively, these findings establish basophil‐derived IL‐4 and IL‐13 as critical mediators that promote wound healing by regulating macrophage polarization toward a pro‐repair M2 phenotype. The fact that genetic deletion of these cytokines specifically in basophils largely phenocopied complete basophil ablation highlights the non‐redundant and indispensable role of the basophil–IL‐4/IL‐13 axis in coordinating immune‐mediated tissue regeneration.

4. Discussion

The biological effects of IL‐4 and IL‐13 in the skin are highly context‐ and time‐dependent. Sustained exposure to these cytokines in chronic inflammatory settings such as human AD drives epidermal barrier dysfunction and persistent inflammation, consistent with the clinical efficacy of IL‐4Rα blockade. In contrast, our data support a model in which transient, localized IL‐4/IL‐13 production during the early phase of wound repair licenses macrophage polarization toward a reparative phenotype, thereby accelerating re‐epithelialization. These effects are mechanistically and temporally separable from the chronic pathogenic type 2 signaling observed in AD and underscore the importance of temporal resolution when interpreting cytokine function in skin biology.

Our study unveils a novel and unexpected function for basophils in the skin: orchestrating tissue repair following injury in a type 2–skewed skin sensitization context within an allergen‐sensitized microenvironment. Specifically, basophils rapidly accumulate at wound sites in sensitized skin and serve as an early source of IL‐4 and IL‐13, which in turn promote macrophage polarization toward an M2‐like reparative state. This cascade accelerates wound closure, enhances angiogenesis, and supports tissue regeneration. Rather than representing a true clinical paradox, our findings provide a mechanistic framework for understanding the apparent dissociation between chronic barrier dysfunction and preserved re‐epithelialization capacity following superficial skin injury in type 2‐inflamed settings.

Traditionally, basophils have been viewed primarily as amplifiers of allergic inflammation. Our findings substantially expand this functional repertoire by positioning basophils as key regulators of type 2 immunity–mediated tissue repair. This aligns with emerging concepts in immunology that classify certain immune responses as “type 2 reparative inflammation”, which can be beneficial in contexts of tissue damage [10, 20, 21]. Importantly, our findings suggest that the functional outcome of basophil infiltration is context‐dependent: in the absence of injury, they contribute to itch and inflammation, but upon tissue damage, they rapidly switch to a pro‐repair program. This duality highlights the complexity of basophil biology and echoes observations by Pellefigues and colleagues [5], who reported context‐dependent functions for basophils in allergic skin inflammation, contributing to both inflammatory exacerbation and resolution. Our findings thus expand the functional repertoire of basophils beyond allergic inflammation, positioning them as key regulators of tissue repair.

Although other immune populations, including mast cells and group 2 innate lymphoid cells, are capable of producing IL‐4 and IL‐13 in allergic skin inflammation [22, 23, 24], our temporal and transcriptional analyses indicate that basophils represent the predominant early source of these cytokines during acute wound repair in sensitized skin. Moreover, basophils exhibited the most pronounced relative expansion following oxazolone sensitization, providing a functional rationale for their central role in initiating the reparative macrophage response observed in this system.

From a translational perspective, our findings suggest that the basophil–IL‐4/IL‐13–macrophage axis may represent a therapeutic target to promote healing in chronic wounds, which are characterized by persistent inflammation and failure to progress to the reparative phase. This is particularly relevant for chronic ulcers in patients with diabetic or venous disease. Topical administration of lL‐4 or lL‐13, or strategies to safely modulate local basophil function, could represent a novel therapeutic strategy. However, caution is warranted, as unchecked type 2 signaling can also promote fibrosis [25]. Future studies should aim to identify the precise triggers that switch basophils from a pathogenic to a reparative state, allowing for more targeted interventions.

Our study has several limitations. The oxazolone‐based model reflects allergen‐driven skin sensitization rather than human AD, which is IgE‐mediated and chronic in nature. Consequently, our findings should not be interpreted as directly modeling AD pathogenesis. While the oxazolone system is well‐suited for interrogating basophil recruitment and function, it does not fully recapitulate the cellular and clinical complexity of human AD. In addition, although we employed multiple genetic and antibody‐based strategies to target basophils, absolute cell‐type specificity cannot be assumed. Instead, our conclusions rely on convergent transcriptional, cellular, and functional evidence supporting a predominant role for basophils within this experimental context. Furthermore, while we identified macrophages as the primary target, IL‐4 and IL‐13 likely act on other stromal cells such as fibroblasts and keratinocytes to promote repair, a nuance worth exploring further.

In conclusion, we have identified a previously unrecognized pro‐repair function for basophils in allergen‐sensitized skin. This work challenges the conventional view of basophils solely as agents of disease and repositions them as potential mediators of beneficial tissue regeneration in the context of type 2 inflammation. Harnessing this axis may offer new hope for treating debilitating chronic wounds.

Author Contributions

Yale Liu and Yumin Xia conceptualized and supervised the study. Yale Liu, Yufei Zhang, Xueting Peng, Kaixuan Ren, Shiran Kang, Zihan Xue, Yazhuo Li, Zhu Yan, Rongfang Feng, Min Gao, Qin Chen, Xiaoying Ning, and Fan Bai contributed to laboratory data collection. Yale Liu, Yufei Zhang, Xueting Peng, Kaixuan Ren, and Shiran Kang contributed to formal data analysis. Yale Liu and Yumin Xia curated the data. All authors contributed to the interpretation of the data. Yale Liu, Yumin Xia, and Liesu Meng drafted the original draft. All authors were involved in writing, review, and editing. All authors approved the manuscript before submission. All authors had final responsibility for the decision to submit for publication.

Funding

This work was supported by the National Natural Science Foundation of China (Grants 82373476 and 82173445).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: all70279‐sup‐0001‐FigureS1‐S11.docx.

TABLE S1: Cluster marker genes for the Seurat object shown in Figure 1A.

Table S2: all70279‐sup‐0014‐TableS2.docx.

Table S3: all70279‐sup‐0015‐TableS3.docx.

Data Availability Statement

The scRNA‐seq data are from GSE142471 (skin wound data) and GSE149121 (oxazolone‐treated data), the code used for the scRNA‐seq data can be found on GitHub (https://github.com/Yale73/Wound‐and‐oxazolone‐treated‐scRNA‐seq‐data‐analysis/). All other data are available in the main text or the Supporting Information.