Fig 2: IL-36 agonizts from keratinocytes induce chondrocyte catabolism and inflammation in synovial fibroblasts.a Immunostaining and quantification of IL-36R (red) with PDPN (green) or F4/80 (green) in synovial tissues from mice aged 3, 10, or 18 months. Scale bar: 20 μm. b Immunostaining and quantification of IL-36R (red) in the articular cartilages of mice aged 3, 10, or 18 months. Scale bar: 20 μm. c Schematic illustrating the induction of mouse keratinocyte senescence with UVB irradiation, and investigation into the effects of Aging SNL on chondrocytes or synovial fibroblasts with mIL-36Ra or IL-36R knockdown. d, e Representative images and quantification of cell wound scratch assay and β-gal staining results. Scale bar, 100 μm (d) qRT–PCR detection of mRNA levels of TNFα, IL-1β, IL-6, IL-8, CCL2, and TGF-β (e) in mice synovial fibroblasts under the indicated stimulation. f, g Representative images and quantification of alcian blue and β-gal staining (f; scale bar, 100 μm for β-gal staining), western blotting results of collagen II, aggrecan, MMP13, MMP3, and IL-36R (g) in mouse chondrocytes under the indicated stimulation conditions. h Schematic illustrating the knockdown of IL-36Ra in mouse keratinocytes, and investigation into the effects of IL-36Ra-KD SNL on chondrocytes or synovial fibroblasts with mIL-36Ra or IL-36R knockdown. i–j Representative images and quantification of cell wound scratch assay and β-gal staining results. Scale bar, 100 μm (i). qRT–PCR detection of the mRNA levels of TNFα, IL-1β, IL-6, IL-8, CCL2, and TGF-β (j) in synovial fibroblasts of mice under the indicated stimulation conditions. k, l Representative images and quantification of alcian blue and β-gal staining (k; scale bar, 100 μm for β-gal staining), and western blotting results of collagen II, aggrecan, MMP13, MMP3, and IL-36R (l) in mouse chondrocytes under indicated stimulation; n = 8 (a, b) mice per group and results from one representative replicate are shown. Experiments were repeated five (d, f, i, k) or three (e, g, j, l) times with similar results. Quantitative data are shown as the mean ± SD. P-values are from two-tailed unpaired t test (a), one-way ANOVA test followed by Tukey’s post hoc test (b, d–f, i–k). n indicates biological replicates; any technical repeats were averaged within each biological sample. Source data are provided as a Source Data file.

Fig 3: Inserting microneedles loaded with mIL-36Ra into the skin attenuates experimental OA in mice.a, b A schematic and image illustrating the insertion of microneedles loaded with mIL-36Ra (MN-mIL-36Ra) into the mouse epidermis layer. c Image of MN-mIL-36Ra under optical microscope and SEM. Scale bar: 100 μm (c). d Drug release in MN-mIL-36Ra. e Schematic of the experimental design to determine the effects of MN-mIL-36Ra on OA progression and pain in Zmpste24fl/fl and ZM-cKO mice, IL-36Rafl/fl and IL-36Ra-cKO mice with surgically-induced DMM OA. f–g Immunostaining (f) and quantification (g) of K10 with IL-36α/β/γ in epidermal skin from ZM-cKO and IL-36Ra-cKO mice with or without MN-mIL-36Ra insertion. Scale bar: 50 μm. h Quantified ELISA results of IL-36Ra and IL-36α/β/γ cytokines in serum from ZM-cKO and IL-36Ra-cKO mice with or without MN-mIL-36Ra insertion for 8 weeks. i–k Safranin-O staining (i), scoring of OA parameters (j), and quantitative analysis of the von Frey threshold (k) in DMM-operated Zmpste24fl/fl and Z24-cKO mice with or without MN-mIL-36Ra insertion. The inset in the images is shown in the magnified images in the bottom row. Scale bar: 50 μm. l–n Safranin-O staining (l), scoring of OA parameters (m), and quantitative analysis of the von Frey threshold (n) in DMM-operated IL-36Rafl/fl and IL-36Ra-cKO mice with or without MN-mIL-36Ra insertion. The inset in the images is shown in the magnified images in the bottom row. Scale bar: 50 μm; n = 6 (f–h) or n = 8 (i–n) mice per group. Results from one representative replicate are shown. Quantitative data are shown as the mean ± SD. P-values are from two-tailed Mann–Whitney U-test (k, n, OARSI, synovial inflammation scores, osteophyte maturity scores in j or m) and two-tailed unpaired t test (g, h, SBP thickness in j or m). n indicates biological replicates; any technical repeats were averaged within each biological sample. Source data are provided as a Source Data file.

Fig 4: IL-36Ra deletion in the epidermis keratinocytes activates proinflammatory signaling in vivo.a, b H&E staining and quantification of epidermal thickness and cell numbers in skin sections from IL-36Rafl/fl and IL-36Ra-cKO mice at 12 weeks of age. Scale bar: 20 μm. c–f Immunostaining (c) and quantification (d) of K10 with IL-36Ra or IL-36α/β/γ in epidermal skin. Quantified ELISA results of IL-36α/β/γ cytokines in serum (e) and quantified RT-qPCR analysis of IL-36α/β/γ mRNA in joint tissues (f) from IL-36Rafl/fl and IL-36Ra-cKO mice at 12 weeks of age. Scale bar: 20 μm. g–i. Volcano plot showing differentially-expressed genes (DEGs) (g) KEGG pathway (h) and GSEA analysis (i) for DRGs demonstrating NF-κB, JAK-STAT and MAPK signaling pathway enrichment from RNA-sequencing of epidermis skin tissues from IL-36Rafl/fl and IL-36Ra-cKO mice. j, k Immunostaining (j) and quantification (k) of p-p65, p-STAT3, p-p38, p-JNK, p-ERK1/2 in epidermis skin tissues from IL-36Rafl/fl and IL-36Ra-cKO mice at 12 weeks of age. Scale bar: 50 μm. l, m Immunostaining(l) and quantification (m) of p-p65, p-STAT3, p-p38, p-JNK, p-ERK1/2 in cartilages and synovium from IL-36Rafl/fl and IL-36Ra-cKO mice at 12 weeks of age. Scale bar: 50 μm. n = 6 (a–d) or 5 (e) or 3 (f) or 8 (j–m) mice per group. Results from one representative replicate are shown. Quantitative data are shown as the mean ± SD. P-values are from a two-tailed unpaired t test. n indicates biological replicates; any technical repeats were averaged within each biological sample. Source data are provided as a Source Data file.

Fig 5: Decreased IL-36Ra in senescent keratinocytes results in increased systemic IL-36 agonizts and is involved in OA.a Heatmap of partial differentially present proteins (> 1.5-fold, P < 0.05) identified by proteomic screen analysis in epidermal skin tissue from young (3 month) and aged (18 month) mice. b, c H&E staining, immunostaining of p16 and lamin B1 (b) and quantification of epidermal thickness and numbers of p16- and lamin B1-positive cells (c) in skin sections from young and aged mice. Scale bar: 20 μm. d Safranin-O staining and OARSI scores in joint tissues from young and aged mice. Scale bar: 50 μm. e, f Immunostaining (e) and quantification (f) of K10 with IL-36Ra or IL-36α/β/γ in epidermal skin from young and aged mice. Scale bar: 20 μm. g Quantified ELISA analysis results of IL-36α/β/γ cytokines in serum from young and aged mice. h Quantified RT-qPCR analysis of IL-36α/β/γ mRNA in joint tissues from young and aged mice. i, j H&E staining, immunostaining of p16 and lamin B1 (i), and quantification of epidermal thickness and numbers of p16- and lamin B1-positive cells (j) in skin sections from young and aged male patients. Scale bar: 50 μm. k Safranin-O staining and OARSI scores in joint sections from young and aged male patients. Scale bar: 100 μm. l, m Immunostaining (l) and quantification (m) of K10 with IL-36Ra or IL-36α/β/γ in epidermal skin from young and aged male patients. Scale bar: 50 μm. n Quantified ELISA analysis results of IL-36α/β/γ cytokines in serum from young and aged male patients. o Elisa correlation analysis of serum and synovial fluid. n = 6 (b–g) or 3(h) mice per group, n = 6 (i–j, l–m) patients per group, or n = 6 (Young) or 9 (Aged OA) (k) patients per group, or n = 5 (n) or n = 15 (o) patients per group. Results from one representative replicate are shown. Quantitative data are shown as mean ± SD. P-values are from two-tailed Mann–Whitney U-test (d, k) and two-tailed unpaired t test (c, f, g, h, j, m, n). R² and exact P-values (for the regression slope) are from two-sided simple linear regression (o). n indicates biological replicates; any technical repeats were averaged within each biological sample. Source data are provided as a Source Data file.