Abstract Rosacea is a chronic inflammatory skin disorder linked to the antimicrobial peptide LL-37 and immune cells. STING, a key DNA-sensing adaptor, initiates innate immune responses, with excessive activation contributing to inflammation. This study investigates LAPTM5, a STING-interacting protein, and its role in rosacea. We observe elevated nuclear DNA fragmentation within the dermal lesions of rosacea patients and LL-37-induced rosacea-like mice. LAPTM5 and STING levels are upregulated in macrophages within rosacea lesions and LL-37-induced models, along with STING hyperactivation. LAPTM5 knockdown in macrophages reduces STING protein levels, signaling, and inflammatory responses under DMXAA and HT-DNA stimulation. LAPTM5 associates with STING and represses its K48- and K63-linked polyubiquitination, preventing proteasomal and lysosomal degradation, thereby maintaining STING stability at homeostasis and after activation. Both STING antagonist H-151 and LAPTM5 knockdown alleviate LL-37-induced rosacea-like phenotypes. These findings highlight LAPTM5 as a STING stabilizer, aggravating STING-driven inflammation in rosacea, offering insights for potential treatments. Subject terms: Chronic inflammation, Acne vulgaris __________________________________________________________________ LAPTM5 stabilizes STING to exacerbate LL-37-induced inflammation in rosacea, revealing a potential therapeutic target for modulating STING-mediated immune responses in inflammatory skin diseases. Introduction Rosacea is a prevalent inflammatory skin disease characterized by facial flushing, transient or persistent erythema, telangiectasia, papules, pustules, and glandular hyperplasia^[46]1–[47]3. Rosacea significantly impacts the physical and mental health of affected individuals, with a significant proportion of patients suffering from anxiety, depression, and suicidal tendencies^[48]4. The limited efficacy of current therapies underscores the urgent need to further elucidate the pathogenesis of rosacea. The etiology of rosacea is closely associated with ultraviolet radiation, microbial infection, physical injury, and stress, which lead to the production of abundant cathelicidin LL-37 and macrophage infiltration in skin lesions. LL-37 promotes innate immune response and exacerbates inflammation under high concentrations or specific conditions^[49]5–[50]8. Macrophage-mediated innate immune responses play a role in the inflammation seen in rosacea^[51]9–[52]11. Histopathological examination reveals a profound macrophage infiltration in the dermis of various subtypes of rosacea^[53]12. The TLR2/KLK5/cathelicidin pathway in macrophages is necessary for rosacea inflammation and constitutes a potential therapeutic target^[54]13. For example, carvedilol can effectively inhibit the expression of TLR2 in macrophages, thereby suppressing inflammatory factors and alleviating skin inflammation both in vivo and in vitro^[55]14. Moreover, LL-37 enhances the activation of the NLRP3 inflammasome in macrophages primed with lipopolysaccharide, contributing to the development of rosacea-like skin inflammation^[56]7. Nevertheless, the molecular mechanisms of macrophage activation induced by LL-37 remain largely undefined. The STING (stimulator of interferon genes) signaling pathway plays critical roles in the innate immune response, recognizing cytoplasmic DNA including microbial DNA and self-DNA from damaged mitochondria or nucleus^[57]15. The enzyme cGAS (cyclic guanosine monophosphate-adenosine monophosphate synthase) synthesizes the secondary messenger cGAMP, which binds to STING and promotes its trafficking to Golgi apparatus. Then, STING recruits TBK1 to activate IRF3 and NF-κB, leading to the production of type I interferons and pro-inflammatory cytokines, respectively^[58]15,[59]16. It has been shown that the function of STING is finely controlled through different post-translational modifications, including polyubiquitination and deubiquitination, which influences STING stability and immune homeostasis^[60]17–[61]19. STING hyperactivation is associated with various skin inflammatory diseases, such as lupus erythematosus, psoriasis, dermatomyositis, and acne^[62]20–[63]23. It has been demonstrated that targeted inhibition of STING can effectively alleviate skin inflammation^[64]21,[65]24. Nevertheless, the role of the STING signaling pathway in the pathogenesis of rosacea remains uncertain. LAPTM5 (lysosomal-associated protein transmembrane 5) is a member of the endosome/lysosome protein family, involved in regulating protein homeostasis and inflammatory signaling^[66]25. It is predominantly expressed in lymphoid and myeloid cells and plays a crucial role in immune response regulation^[67]25–[68]27. The PY and UIM motifs of LAPTM5 interact with various substrates, facilitating the sorting of proteins from the Golgi to lysosomes, thus contributing to intracellular substrate transport and lysosomal stability^[69]28. In macrophages, LAPTM5 enhances toll-like receptor- and tumor necrosis factor receptor-mediated NF-κB signaling, thereby promoting the production of pro-inflammatory cytokines^[70]27. Nonetheless, the specific role of LAPTM5 in regulating STING-mediated inflammation remains to be elucidated. In this study, we provide evidence that the expressions of LAPTM5 and STING are increased in macrophages in the lesional skin of rosacea patients, as well as in LL-37-induced mouse and cell models, accompanied by hyperactivation of STING signaling pathway. Furthermore, we reveal the mechanisms that LAPTM5 positively regulates the STING-mediated immune response by stabilizing STING protein both at homeostasis and after activation. Notably, both the STING antagonist H-151 and LAPTM5 knockdown effectively alleviate LL-37-induced rosacea-like phenotypes, highlighting potential therapeutic strategies for rosacea. Results STING signaling is hyperactivated in skin lesions of rosacea We first analyzed publicly available gene array data from the NCBI GEO database (Accession number: [71]GSE65914)^[72]29. Hierarchical clustering analysis revealed a distinct upregulation of mRNA levels in genes associated with both the STING signaling pathway and monocyte/macrophage markers. This upregulation was observed in skin samples from different rosacea subtypes (ETR, PPR, and PhR) compared to those from healthy controls (HC) (Fig. [73]1A). Gene Set Enrichment Analysis (GSEA) revealed that the regulation of macrophage activation was enriched in rosacea, together with regulation of type I interferon-mediated signaling pathway, in which STING expression was highlighted as a key component (Fig. [74]1B, C). Further investigations into the dataset revealed a positive correlation between STING and monocyte/macrophage markers, interferon-stimulated genes (ISGs), and inflammatory factors in patients with rosacea (Fig. [75]S1A). Figures [76]1D and [77]1E showed the significant upregulation of STING and CD68 (a macrophage marker) expression in skin samples from rosacea patients compared to HC, while Fig. [78]1F illustrated their positive correlation in rosacea. Consistently, analysis of another gene array dataset from the Genome Sequence Archive (GSA) repository (Accession number: HRA000378) corroborated our findings^[79]30, demonstrating the upregulation of both macrophage-related and STING-related genes in rosacea skin samples compared to those in HC samples (Fig. [80]S1B). Similarly, a dataset from the NCBI GEO database (Accession number: [81]GSE147950) revealed comparable expression patterns in skin samples from LL-37-induced rosacea-like mice (Fig. [82]S1C). Fig. 1. Macrophage- and STING-related genes are upregulated in lesional skin from rosacea patients. [83]Fig. 1 [84]Open in a new tab A Heatmap of differentially regulated genes in skin biopsies from healthy controls (HC, n = 20) and patients with erythematotelangiectatic rosacea (ETR, n = 14), papulopustular rosacea (PPR, n = 12), and phymatous rosacea (PhR, n = 12). The expression values were acquired from NCBI GEO (accession number: [85]GSE65914). Blue color denotes low FPKM expression; red, high FPKM expression. B Top-ranked enriched KEGG terms in genes that were differentially regulated in the comparison (rosacea vs HC) revealed by gene set enrichment analysis (GSEA). Regulation of macrophage activation and regulation of type I interferon-mediated signaling pathway were highlighted in red box. C GSEA analysis shows enrichment for the regulation of type I interferon-mediated signaling pathway in rosacea samples, with STING expression as a key component. NES, normal enrichment score; FDR, false discovery rate; FDR value < 0.05 was considered significant. Significance was calculated by permutation test. D, E Analysis of the mRNA expression level of STING (D) and CD68 (E) in HC, ETR, PPR and PhR skin samples. Data represent the mean ± SEM, **p < 0.01, ***p < 0.001, and ****p < 0.0001. One-way ANOVA with Bonferroni’s post hoc test was used. F Correlation between STING expression and macrophage marker CD68 in rosacea samples (n = 38). Spearman’s correlation coefficient was used for the correlation analysis (two-tailed). G Immunohistochemistry staining of CD86, pSTING and STING on skin sections from rosacea and HC. n = 5–8 for each group. Scale bar: 100 μm. Higher magnified images of boxed areas are shown at the right of lower magnified images for each group. H–J Histograms depicting the Integrated Optical Density (IOD) to Area ratio in the immunohistochemical analysis of CD86 (H), STING (I) and pSTING (J) in skin sections from rosacea patients and HC. All results are representative of at least three independent experiments. K, L Representative immunofluorescence staining of dsDNA (green) in skin tissues from rosacea patients and HC (K). Arrowheads in magnified images indicate dsDNA fragments foci located outside of DAPI (blue). The percentage of dsDNA-positive cells was calculated as the ratio of dsDNA to DAPI in three randomly selected fields for each sample (L). n = 5 for each group. Scale bar: 100 µm. M, N Detection of nuclear DNA fragmentation (red) by TUNEL staining (M) and quantification of the percentage of TUNEL-positive cells in skin tissues from rosacea patients and HC (N). Nuclei are stained with DAPI (blue). For each sample, three regions were randomly selected, and the percentage of TUNEL-positive cells was calculated over DAPI-positive cells. n = 5 for each group. Scale bars: 100 μm. Data represent the mean ± SEM, **p < 0.01, ***p < 0.001. Two-tailed unpaired Student’s t-test was used. The results of immunohistochemistry staining showed higher expression of CD86, an M1 macrophage marker, in the skin samples from rosacea patients than in those from HC (Fig. [86]1G, H). We further observed that the protein levels of phosphorylated and total STING were markedly increased in the lesional skin of rosacea patients compared to those in HC (Fig. [87]1G, I, J), indicating that STING is hyperactivated in the skin lesions of rosacea. The release of dsDNA may serve as a trigger to promote STING activation^[88]15. It has been reported that bacterial DNA can lead to pathogenic type I IFN-driven inflammation in rosacea, while DNA damage-induced nuclear DNA leakage could contribute to STING-dependent inflammation in psoriasis^[89]31,[90]32. In this study, dsDNA staining showed that elevated levels of DNA fragments were detected in the skin lesions of rosacea patients compared to those in HC (Fig. [91]1K, L). The results of TUNEL staining revealed severe nuclear DNA fragmentation within the dermal lesions of rosacea patients (Fig. [92]1M, N). Collectively, these data indicate that STING hyperactivation may be involved in the pathogenesis of rosacea. The STING antagonist H-151 ameliorates inflammation in LL-37-induced rosacea-like mice It has been widely manifested that cathelicidin LL-37 is upregulated in the skin lesions of rosacea patients, and that intradermal injection of high concentrations of LL-37 into mouse the back skin promotes rosacea-like features^[93]33. We further intraperitoneally injected H-151 (a STING antagonist) to assess its inhibitory effect on rosacea-like dermatitis in mice (Fig. [94]2A). As expected, our data showed that intradermal injection of LL-37 led to prominent rosacea-like phenotypes, including erythema, edema, and telangiectasia, which were significantly alleviated by the intraperitoneal injection of H-151 (Fig. [95]2B, C). The average area of redness and the severity score were significantly reduced after the injection of H-151 (Fig. [96]2D, E). Moreover, histological staining demonstrated that H-151 reduced the number of inflammatory cells infiltrating the dermis (Fig. [97]2F, G). Immunohistochemical staining revealed a substantial increase in the expression of inflammatory factors, including IFN-β, CXCL10, TNF-α, IL-6, and IL-1β, in lesional tissues from LL-37-treated mice compared to controls. In contrast, H-151 treatment notably decreased the levels of these inflammatory factors (Fig. [98]2H–J). Thus, these findings reveal the anti-inflammatory effect of H-151 on rosacea-like dermatitis in mice by dampening STING signaling pathway. Fig. 2. The STING antagonist H-151 ameliorates inflammation in LL-37-induced rosacea-like mice. [99]Fig. 2 [100]Open in a new tab A Schematic diagram of administration of LL-37 and H-151 in mice. LL-37 (320 μM in 40 μl PBS) or PBS was injected intradermally into BALB/c mice four times at 12-h intervals to induce rosacea-like dermatitis; meanwhile, H-151 (3.75 mM in 200 µl of 10% Tween-80 in PBS) or control vehicle was injected intraperitoneally three times at 24-h intervals. n = 6 for each group. Illustration elements (syringe and mouse) were obtained from Figdraw 2.0. B Representative images of the macroscopic appearance of mouse skin in different groups. Images were taken 12 h after the last injection of LL-37. C Representative images of the polarized light dermoscopy of mouse skin in different groups. D, E The severity of the rosacea-like phenotypes was evaluated based on the area of redness (D) and the severity score (E). n = 6 for each group. F HE staining of the skin samples from the four groups. Below panels, magnified pictures of boxed areas. Scale bar: 100 μm. G Quantitative results of HE staining for dermal infiltrating inflammatory cells. n = 6 for each group. Data represent the mean ± SEM. H IHC staining of IFN-β and CXCL-10 in the dorsal skin tissues from different groups. Below panels, magnified pictures of boxed areas. Scale bar: 100 μm. I IHC staining of TNF-α, IL-6 and IL-1β in the dorsal skin tissues from different groups. Below panels, magnified pictures of boxed areas. Scale bar: 100 μm. J Quantification of IHC staining intensity (IOD/Area) for IFN-β, CXCL-10, TNF-α, IL-6, and IL-1β in skin sections from mice in different groups. n = 6 for each group. Data represent the mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. One-way ANOVA with Bonferroni’s post hoc test was used. STING signaling is hyperactivated in macrophages in LL-37-induced rosacea-like mice While our analysis of the aforementioned dataset demonstrated a positive correlation between STING and macrophages in rosacea patients, we sought to determine whether STING was hyperactivated in macrophages in LL-37-induced rosacea-like mice. Immunohistochemical analysis revealed that LL-37 injection led to increased expression of F4/80 and STING proteins in the lesional skin. Additionally, STING phosphorylation was elevated in the model group (Fig. [101]3A–D). Consistently, Western blot analysis confirmed an elevation in STING protein levels in the model group (Figs. [102]3E and S2A). Fig. 3. STING signaling is hyperactivated in macrophages in LL-37-induced rosacea-like mice. [103]Fig. 3 [104]Open in a new tab A IHC staining of F4/80, pSTING and STING in skin sections from control and LL-37-induced mice. Scale bar: 100 μm. n = 6 for each group. B–D Quantification of staining intensity (IOD/Area) for F4/80 (B), pSTING (C), and STING (D) in skin lesions. E Immunoblotting analysis of STING in skin lysates from control and LL-37-induced mice. F Immunofluorescence staining of F4/80 (red) and STING (green) in skin sections of LL-37- or H-151-treated mice. The magnified images show detailed co-localization of F4/80 and STING. Nuclei are stained with DAPI (blue). Scale bars: 100 µm. G–I Quantification of relative fluorescence intensity for STING (G) and F4/80 (H) in the immunofluorescent images of LL-37- or H-151-treated mice. STING and F4/80 colocalization (I), expressed as the percentage of double-positive cells relative to total F4/80-positive cells in three randomly selected fields. n = 6 for each group. J, K Representative immunofluorescence images showing dsDNA (green) in skin tissues from LL-37-treated mice and the controls (J). Nuclei are stained with DAPI (blue). The merged and zoom-in images highlight the presence of dsDNA fragments outside the nuclei (arrows) in the LL-37-treated group. The percentage of dsDNA-positive cells was calculated as the ratio of dsDNA to DAPI in three randomly selected fields for each sample (K). Scale bar: 50 µm. n = 6 for each group. L, M Detection of nuclear DNA fragmentation (red) by TUNEL staining (L) and quantification of the percentage of TUNEL-positive cells in skin tissues from LL-37-treated mice and the controls (M). Nuclei are stained with DAPI (blue). The percentage of TUNEL-positive cells was calculated over DAPI-positive cells in three randomly selected fields for each sample. Scale bar: 100 µm. n = 6 for each group. Data represent the mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Two-tailed unpaired Student’s t-test or One-way ANOVA with Bonferroni’s post hoc test was used. Next, we conducted immunofluorescence staining to observe the co-localization of F4/80 with STING. As shown in Fig. [105]3F, lesional skin tissues from LL-37-treated mice exhibited a pronounced increase in the protein levels of both F4/80 and STING, accompanied by augmented co-localization. By contrast, this upregulation was significantly reversed in the group treated with both LL-37 and H-151 (Fig. [106]3G–I). Previous studies have shown that clodronate liposomes could deplete macrophages and alleviate LL-37-induced skin inflammation in rosacea-like models^[107]11. We investigated the effect of macrophage depletion on STING expression within the lesions of LL-37-induced rosacea-like dermatitis mice. Mice were intraperitoneally injected with clodronate liposomes prior to LL-37 administration. Compared to the PBS group, the liposome control group (LL-37 + PBS-lipo) showed significantly higher F4/80 (red) expression in the lesion area (Fig. [108]S2B, C). In contrast, the clodronate liposome group (LL-37 + Clo-lipo) showed markedly reduced F4/80 expression, indicating effective macrophage depletion. Immunofluorescence analysis revealed significantly elevated STING (green) expression in the liposome control group (LL-37 + PBS-lipo) compared to the PBS group. Conversely, STING expression was markedly reduced in the clodronate liposome group (LL-37 + Clo-lipo), suggesting that macrophage depletion could reduce STING expression in the skin lesions (Fig. [109]S2D, E). Therefore, macrophages are a major source of the increased expression of STING in rosacea-like dermatitis mouse. We sought to ascertain if there was an increase in DNA fragments in rosacea-like dermatitis lesions. The dsDNA staining showed abundant DNA fragments in the lesional tissues from LL-37-treated mice (Fig. [110]3J, K). The TUNEL assay exhibited increased nuclear DNA fragmentation in the dermis of LL-37-treated mice compared to the controls (Fig. [111]3L, M). Collectively, these findings imply that hyperactivation of STING in macrophages contributes to the inflammatory response in LL-37-induced rosacea-like dermatitis in mice. LAPTM5 is upregulated and associated with the activation in macrophages in rosacea To identify key regulators of STING in rosacea, we performed Weighted Gene Co-expression Network Analysis (WGCNA) on 10,422 genes from the [112]GSE65914 dataset after batch effect correction. WGCNA grouped genes into 15 core modules based on topological overlap and dynamic tree cutting (Fig. [113]S3A). A correlation heatmap showed that the STING-containing cyan module was strongly associated with rosacea (r = 0.70, p = 2.5e-5; Fig. [114]S3B). KEGG analysis indicated that this module was enriched in immune and inflammatory pathways (Fig. [115]S3C). Applying thresholds of log2FC > 1.3 and adjusted p < 0.05, we identified 180 upregulated genes within the cyan module. Based on the Motani group’s proximity-dependent biotinylation screen of STING interactors in macrophages^[116]34, we further filtered candidates using a DMXAA/DMSO ratio >1.3 and p < 0.01, yielding 330 potential interactors. Venn diagram analysis identified two overlapping genes: LAPTM5 and TMEM176B (Fig. [117]S3D). Analysis of the [118]GSE65914 dataset revealed a significant upregulation of LAPTM5 mRNA in skin samples from ETR, PPR, and PhR patients compared to healthy controls (Fig. [119]4A), which was further validated using another dataset (HRA000378, Fig. [120]4B). Similarly, Laptm5 mRNA was elevated in skin lesions of the LL-37-induced mouse model ([121]GSE147950, Fig. [122]4C). Further correlation analysis ([123]GSE65914) demonstrated a positive association between LAPTM5 expression and macrophage markers, ISGs, and inflammatory factors (Fig. [124]S3E). Notably, LAPTM5 was strongly correlated with CD68 (r = 0.764, p < 0.001, Fig. [125]4D) and STING (r = 0.876, p < 0.001, Fig. [126]4E). These results suggest LAPTM5 may play a role in regulating the STING signaling pathway in the inflammatory response of rosacea. Fig. 4. LAPTM5 is upregulated and associated with the activation of the STING signaling pathway in macrophages in rosacea. [127]Fig. 4 [128]Open in a new tab A–C Analysis of the expression levels of LAPTM5 in skin samples from rosacea patients ([129]GSE65914 and HRA000378) and LL-37-induced rosacea-like mice ([130]GSE147950). D, E Scatter plots showing positive correlation of LAPTM5 expression in rosacea skin ([131]GSE65914) and macrophage marker CD68 (D) and STING (E). F, G IHC of LAPTM5 in skin tissues from healthy controls (HC) and rosacea patients (F). Quantification of staining intensity (IOD/Area) for LAPTM5 in the IHC images of rosacea patients and HC (G). n = 5 for each group. H–J IHC of LAPTM5 in skin tissues from control and LL-37-induced rosacea-like mice (H). Quantification of staining intensity (IOD/Area) for LAPTM5 in the IHC images of LL-37-induced rosacea-like mice and control (I). n = 6 for each group. J Immunoblotting analysis of LAPTM5 protein expression in skin tissue lysates from control and LL-37-induced rosacea-like mice. K Immunofluorescence staining of F4/80 (red) and LAPTM5 (green) in skin sections from LL-37- or H-151-treated mice. Nuclei are stained with DAPI (blue). The merged and zoom-in images highlight co-localization of F4/80 and LAPTM5. Scale bars: 100 µm. L–N Quantification of relative fluorescence intensity for LAPTM5 (L) and F4/80 (M) in skin sections from LL-37- or H-151-treated mice. LAPTM5 and F4/80 colocalization (N), expressed as the percentage of double-positive cells relative to total F4/80-positive cells in three randomly selected fields for each sample. Scale bar: 100 µm, n = 6 for each group. O Immunoblot analysis of STING and LAPTM5 expression in skin tissue lysates from mice treated with LL-37, H-151, or LL-37 plus H-151. P, Q Immunoblot analysis of pSTING, STING, pIRF3, IRF3, pNF-κB p65, NF-κB p65, and LAPTM5 in cell lysates from RAW264.7 and L929 cells exposed to increasing concentrations of LL-37 (0, 1, 2, 4, and 8 µM) for 24 h. Data represent the mean ± SEM, **p < 0.01, ***p < 0.001, ****p < 0.0001. Two-tailed unpaired Student’s t-test or One-way ANOVA with Bonferroni’s post hoc test was used. All results are representative of at least three independent experiments. To confirm LAPTM5 upregulation in rosacea, immunohistochemistry showed elevated protein levels in patient skin compared to healthy controls (Fig. [132]4F, G), consistent with increased LAPTM5 expression in LL-37-treated mouse skin, as confirmed by both immunohistochemistry and immunoblotting (Figs. [133]4H–J, S4A). Immunofluorescence further revealed enhanced LAPTM5 and CD68 expression in rosacea lesions, with increased fluorescence intensity and a higher proportion of LAPTM5⁺CD68⁺ double-positive cells (Fig. [134]S4D–G). Similarly, LL-37-treated mice showed elevated LAPTM5 and F4/80 levels with enhanced colocalization (Fig. [135]4K). Notably, immunoblotting showed selective suppression of STING but not LAPTM5 (Figs. [136]4O, S4B–C). We then examined the effect of macrophage depletion on LAPTM5 expression in the lesions of the LL-37-induced mouse model using immunofluorescence analysis. Compared to the PBS group, the liposome control group (LL-37 + PBS-lipo) showed significantly enhanced LAPTM5 fluorescence signals in the skin lesions. In the clodronate liposome group (LL-37 + Clo-lipo), LAPTM5 expression was significantly reduced compared to the liposome control group (LL-37 + PBS-lipo) (Fig. [137]S5A, B). These results suggest that macrophage depletion with clodronate liposomes significantly reduces LAPTM5 expression, indicating its involvement in macrophages in LL-37-induced rosacea-like dermatitis mouse. Previous studies have established LL-37-induced mouse and cell models as a widely utilized approach for investigating rosacea pathogenesis. LL-37 treatment activates inflammatory signaling, leading to immune responses^[138]35,[139]36. Higher doses of LL-37 can induce nuclear DNA damage and fragmentation, and thus promote the release of nucleic acids^[140]37,[141]38. Since we have demonstrated the exacerbation of nuclear DNA damage and the elevation of dsDNA content, we supposed that LL-37 treatment could activate downstream signaling of STING in cellular models. Consistent with this hypothesis, RAW264.7 and L929 cells treated with LL-37 exhibited significant upregulation of LAPTM5 and STING proteins, accompanied by phosphorylation of STING, TBK1, IRF3, and NF-κB p65, indicative of activated STING signaling (Fig. [142]4P, Q). Collectively, these findings highlight the potential importance of LAPTM5 upregulation coupled with STING pathway activation in macrophages, contributing to the inflammatory response observed in rosacea. LAPTM5 positively regulates STING-mediated inflammatory response The possible role of LAPTM5 in STING-mediated inflammatory signaling was evaluated in vitro cellular experiments. We first assessed the protein levels of LAPTM5 and STING in RAW264.7 and L929 cells, which are murine macrophage and fibroblast cell lines, respectively. In line with the findings of the previous study, LAPTM5 was highly expressed in RAW264.7 cells, whereas STING showed high expression in both L929 and RAW264.7 cells (Fig. [143]S5C). We then knocked down the expression of LAPTM5 by infecting RAW264.7 cells, with lentivirus expressing shRNA against LAPTM5, while overexpressing LAPTM5 in L929 cells. We observed that LAPTM5 knockdown substantially inhibited the mRNA expressions of Ifnb, Ifit1, Tnfα, Il6, and Cxcl10 in response to both HT-DNA and DMXAA (Figs. [144]5A and S6A). On the contrary, overexpression of LAPTM5 increased the expression of the Ifnb, Ifit1, Tnfα, Il6, and Cxcl10 mRNAs induced by HT-DNA and DMXAA in L929 cells (Figs. [145]5B and S6B). Fig. 5. LAPTM5 positively regulates STING-mediated inflammatory response. [146]Fig. 5 [147]Open in a new tab A qRT-PCR analysis of the mRNA levels of Ifnb, Ifit1, Tnfα, Il6, and Cxcl10 in stable LAPTM5-knockdown RAW264.7 cells following a 3-hour stimulation with either HT-DNA (2 μg/ml). B qRT-PCR analysis of the mRNA levels of Ifnb, Ifit1, Tnfα, Il6, and Cxcl10 in stable LAPTM5-overexpressing L929 cells following a 3-hour stimulation with either HT-DNA (2 μg/ml). C Immunoblot analysis of the indicated proteins (pSTING, STING, pTBK1, TBK1, pIRF3, IRF3, pNFκB p65, NFκB p65, and LAPTM5) in stable LAPTM5-knockdown RAW264.7 cells untreated or treated with HT-DNA (2 μg/ml) for the indicated times. D–F Quantitative analysis of the data in (C), D ratio of p-STING to total STING, E p-STING signal intensity relative to the LAPTM5-knockdown 1-hour time point, and F area under the curve (AUC) of p-STING signal intensity. G Immunoblot analysis of the indicated proteins in stable LAPTM5-overexpressing L929 cells untreated or treated with HT-DNA (2 μg/ml) for the indicated times. H–J Quantitative analysis of the data in (G), H ratio of p-STING to total STING, I p-STING signal intensity relative to the control 1-hour time point, and J area under the curve (AUC) of p-STING signal intensity. Data are representative of three independent experiments. Data represent the mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Two-tailed unpaired Student’s t-test, Two-Way ANOVA was used. All results are representative of two independent experiments with similar results. To investigate the regulatory role of LAPTM5 in the STING signaling pathway, RAW264.7 and L929 cells were stimulated with HT-DNA at different time points to assess its effects on the phosphorylation of STING and downstream signaling proteins. Initially, we measured the amplitude and kinetics of STING activation. The activation kinetics, evaluated as the ratio of p-STING to total STING, were comparable between control shRNA and LAPTM5-knockdown RAW264.7 cells, with both peaking at 3 h (Fig. [148]5C, quantitated in Fig. [149]5D). However, LAPTM5 knockdown significantly reduced p-STING levels and the amplitude of the p-STING response (Fig. [150]5E), as indicated by a lower area under the curve (AUC) compared to controls (Fig. [151]5F). Correspondingly, total STING protein levels were reduced in LAPTM5-knockdown cells, indicating a smaller pool of activable STING and an impaired capacity for STING-mediated signaling. This reduction also led to significantly diminished phosphorylation of downstream signaling proteins, including TBK1, IRF3, and NF-κB p65 (Fig. [152]5C). In contrast, LAPTM5 overexpression in L929 cells markedly enhanced p-STING levels and prolonged STING activation following stimulation (Fig. [153]5G). This was accompanied by increased phosphorylation of TBK1, IRF3, and NF-κB p65, along with elevated total STING protein levels, which were sustained for longer durations compared to control cells. Quantitative analysis revealed that LAPTM5 overexpression resulted in a significantly higher AUC for p-STING, highlighting the robust enhancement of STING signaling (Fig. [154]5H–J). These results suggest that LAPTM5 promotes an increase in the size of the activable STING pool, thereby enhancing STING signaling by elevating STING protein levels, sustaining its phosphorylation, and amplifying downstream signaling pathways. To corroborate these findings, RAW264.7 and L929 cells were stimulated with DMXAA, yielding results consistent with those observed with HT-DNA (Fig. [155]S6C, D). Additionally, to further validate the role of LAPTM5, L929 cells were treated with cGAMP, a well-characterized secondary messenger in the STING pathway. LAPTM5 overexpression in cGAMP-treated cells led to a consistent upregulation of STING phosphorylation and total protein levels, accompanied by increased phosphorylation of downstream proteins TBK1, IRF3, and NF-κB p65 (Fig. [156]S7A). Taken together, these suggest that LAPTM5 plays a positive regulatory role in STING-mediated immune response. LAPTM5 interacts with STING We first employed computational analysis of protein–protein interactions to explore whether LAPTM5 could interact with STING. According to the results of computational simulations, pose 2 ranked first with a PIPER pose score of −930.222 and a PIPER pose energy of −1774.039 kcal/mol. This free energy was lower than that of other poses, indicating that this interaction pose was the most stable binding scenario among 70,000 calculations involving the two proteins. 3D stereoscopic images revealed that LAPTM5 (yellow) and STING (blue) could bind together stably (Fig. [157]6A). Specifically, LAPTM5’s GLU176 and STING’s GLN334 formed a hydrogen bond, LAPTM5’s PHE74 and STING’s HIS94 established a π-π bond and 11 van der Waals interactions, and LAPTM5’s GLU42 and STING’s LYS150 formed a hydrogen bond and a salt bridge. Consequently, protein interaction analysis suggests that LAPTM5 may interact with STING. Fig. 6. LAPTM5 interacts with STING. [158]Fig. 6 [159]Open in a new tab A Molecular docking model of LAPTM5 (yellow) and STING (blue). The left and right zoomed-in views show detailed interaction sites. B Co-immunoprecipitation assay of endogenous STING and LAPTM5 in RAW264.7 cells untreated or treated with DMXAA (50 μg/ml) for indicated times. Anti-STING antibody was used for immunoprecipitation with IgG as a negative control. Western blotting identified endogenous LAPTM5, TBK1, and pSTING. C, D Co-immunoprecipitation assay of STING and LAPTM5 in L929 cells stably overexpressing Flag-tagged LAPTM5 untreated or treated with cGAMP (2 μM) for indicated times with indicated antibodies. Anti-STING antibody (C) and anti-Flag antibody (D) were used for immunoprecipitation. E Co-immunoprecipitation assay of STING and LAPTM5 in HEK293 cells transfected with plasmids encoding Myc-STING and Flag-LAPTM5. Forty-eight hours post-transfection of indicated plasmids, the cells were stimulated with cGAMP (2 μM) for indicated times and cell extracts were immunoprecipitated with anti-Myc antibody. F Immunofluorescence analysis of LAPTM5 and STING colocalization in RAW264.7 cells treated with or without DMXAA stimulation. Cells were fixed and underwent immunofluorescent staining for STING (green) and LAPTM5 (red). A scale bar of 10 μm is included. The boxed area is magnified in the right panel. The upper right panel displays the pixel intensity profile along the white dashed line. G Pearson correlation coefficients were determined using ImageJ software to quantify colocalization (n ≥ 16, n represents cell numbers). Data are presented as boxplots with min/max and median, **p < 0.01. Two-tailed unpaired Student’s t-test was used. All results are representative of two experiments with similar results. Subsequently, using co-immunoprecipitation of endogenous proteins in RAW264.7 cells, we observed that LAPTM5 was associated with STING in unstimulated cells, and this association apparently increased after DMXAA stimulation (Fig. [160]6B). Moreover, we found that cGAMP stimulation consistently enhanced the interaction of FLAG-tagged LAPTM5 with endogenous STING in L929 cells (Fig. [161]6C, D). Expectedly, TBK1 was also co-immunoprecipitated with the STING-LAPTM5 complex after DMXAA or cGAMP stimulation (Fig. [162]6B, D). It was interesting to note that LAPTM5 overexpression markedly enhanced the interaction between STING and TBK1, together with the phosphorylation of TBK1 (Fig. [163]6E). Confocal microscopy further confirmed that LAPTM5 and STING were well colocalized in the cytoplasm in RAW264.7 cells after DMXAA stimulation (Fig. [164]6F). The Pearson’s correlation coefficient between LAPTM5 and STING revealed the significant increase in their co-localization after DMXAA stimulation (Fig. [165]6G). In summary, these results suggest that LAPTM5 associates with STING at the resting state, and their association largely depends on STING activation, which in turn is strengthened by the association. LAPTM5 maintains stability of STING In our aforementioned results, we preliminarily observed the positive effects of LAPTM5 on regulation of STING protein expression, thereby enhancing the downstream signaling and inflammatory response. We subsequently explored the role of LAPTM5 in modulating the stability of STING protein. Knockdown of LAPTM5 in RAW264.7 cells led to downregulation of endogenous STING, which was in contrast upregulated in LAPTM5-overexpressing L929 cells (Fig. [166]7A, B). Meanwhile, we observed a gradual upregulation of STING protein expression in RAW264.7 cells infected with increasing amounts of LAPTM5-expressing lentivirus, both before and after DMXAA stimulation (Fig. [167]7C). Furthermore, we treated cells with cycloheximide (CHX) to suppress protein synthesis and found that STING protein was rapidly degraded within 12 h in cells containing the FLAG-vector, whereas it was stable in LAPTM5-overexpressing cells with a half-life greater than 12 h (Fig. [168]7D, E). Collectively, these results suggest that LAPTM5 maintains the stability of STING protein. Fig. 7. LAPTM5 maintains stability of STING. [169]Fig. 7 [170]Open in a new tab A Immunoblot analysis of STING protein expression in RAW264.7 cells infected with lentivirus expressing control or LAPTM5 shRNA (1#, 2#, 3#). B Immunoblot analysis of STING protein expression in L929 cells infected with lentivirus expressing Flag or Flag-LAPTM5 at multiplicities of infection (MOI) of 3 and 10. C Immunoblot analysis of STING protein expression in RAW264.7 cells infected with lentivirus expressing Flag or Flag-LAPTM5 at the indicated MOI, and then untreated or treated with DMXAA (50 μg/ ml) for 2 h. D, E Immunoblot analysis of STING protein expression in stable LAPTM5-overexpressing L929 cells untreated or treated with cycloheximide (CHX, 100 µg/ml) for indicated times. Densitometry quantification of protein bands is shown below (E). F Immunoblot analysis of cells transfected with control or LAPTM5 siRNA for 36 h and then left treated with DMXAA alone or in combination with lysosome inhibitor Bafilomycin A1 (BafA1, 50 nM) or proteasome inhibitor MG132 (10 µM). G Immunoprecipitation and immunoblot analysis of polyubiquitin-conjugated STING in HEK293 cells co-transfected with Myc-STING, Flag-LAPTM5, or HA-ubiquitin (Ub). H Immunoprecipitation and immunoblot analysis of K48- or K63-linked polyubiquitination of STING in HEK293 cells co-transfected with indicated plasmids. I Immunoprecipitation and Immunoblot analysis of K63-linked polyubiquitination of STING in HEK293 cells co-transfected with indicated plasmids for 48 h and then left untreated or treated with HT-DNA for indicated times. Data are representative of two independent experiments with similar results. It has been previously demonstrated that STING is targeted for degradation by both proteasomes and lysosomes^[171]16. To further elucidate the protein degradation pathway involved in the decreased STING protein expression in LAPTM5-knockdown cells, we treated RAW264.7 cells with MG132 (a proteasome inhibitor) and BafA1 (a lysosome inhibitor) after transfection with or without siRNA of LAPTM5 (si-LAPTM5). The results showed that both BafA1 and MG132 restored STING protein levels in LAPTM5-knockdown cells in the absence of DMXAA, suggesting that LAPTM5 prevents STING degradation through both lysosomal and proteasomal pathways at the resting state (Fig. [172]7F). In contrast, only BafA1 restored STING protein levels in LAPTM5-knockdown cells in the presence of DMXAA, suggesting that lysosomes play an important role in STING degradation following ligand stimulation (Fig. [173]7F). It has been proven that STING degradation is primarily associated with polyubiquitination, with K48- and K63-linked ubiquitination being the most studied. K48-linked ubiquitinated STING promotes its degradation by proteasomes, whereas K63-linked ubiquitination of STING is essential for its lysosomal degradation^[174]19,[175]39. We next investigated whether LAPTM5 could suppress STING ubiquitination to prevent its degradation. In the overexpression system, STING polyubiquitination was induced by UB-WT, UB-K48, and UB-K63, while LAPTM5 markedly repressed the polyubiquitination of STING, including K48- and K63-linked ubiquitination (Fig. [176]7G, H). Moreover, UB-K63-mediated ubiquitination of STING in HT-DNA-transfected cells was also significantly attenuated by LAPTM5 (Fig. [177]7I). Taken together, our results indicate that LAPTM5 facilitates STING deubiquitination, thereby preventing its degradation mediated by the lysosomal and proteasomal pathways. AAV-mediated knockdown of LAPTM5 reduces STING activation and mitigates LL-37-induced rosacea-like inflammation To further explore the possibility that LAPTM5 is a therapeutic target, adeno-associated virus serotype 9 (AAV9) expressing Laptm5 shRNA was intradermally injected into mice to knock down its expression in the dermal skin of mice. Twenty-one days after AAV injection, the mice were exposed to LL-37 injection to induce the rosacea-like dermatitis (Fig. [178]8A). The effect of LAPTM5 knockdown was validated by immunoblot (Fig. S7B), and we used Laptm5 shRNA1 in all the subsequent experiments. We found that knockdown of LAPTM5 markedly reduced LL-37-induced rosacea-like inflammation, as evidenced by a decrease in area and severity of redness and a reduction in the number of dermis-infiltrating cells (Fig. [179]8B–G). Furthermore, knockdown of LAPTM5 resulted in reduced protein levels of STING and its phosphorylation (Fig. [180]8H–J), as well as the indicated inflammatory factors (Fig. [181]8K–P). Our results demonstrate that LAPTM5 is a potential therapeutic target for protecting mice against STING-mediated rosacea. Fig. 8. AAV-mediated knockdown of LAPTM5 reduces STING activation and attenuates LL-37-induced rosacea-like inflammation. [182]Fig. 8 [183]Open in a new tab A Schematic diagram of intradermal injection of AAV-shLaptm5 21 days before LL-37 injection in BALB/c mice. n = 6 for each group. Illustration elements (syringe and mouse) were obtained from Figdraw 2.0. B Representative images of the macroscopic appearance of mouse skin in different groups. Images were taken 12 h after the last injection of LL-37. C Representative images of the polarized light dermoscopy of mouse skin in different groups. D, E The severity of the rosacea-like phenotypes was evaluated based on the area of redness (D) and the severity score (E). F HE staining of the lesional skin sections. Below panels, magnified pictures of boxed areas. Scale bar: 100 μm. G Quantitative results of HE staining for dermal infiltrating inflammatory cells. H IHC staining of pSTING and STING in skin sections. Right panels, magnified pictures of boxed areas. Scale bar: 100 μm. I, J Quantification of IHC staining intensity (IOD/Area) for pSTING (I) and STING (J) in skin sections. K IHC staining of TNF-α, IL-6, IL-1β, IFN-β, CXCL-10 and in skin sections. Right panels, magnified pictures of boxed areas. Scale bar: 100 μm. L–P Quantification of IHC staining intensity (IOD/Area) for TNF-α (L), IL-6 (M), IL-1β (N), IFN-β (O), and CXCL-10 (P) in skin sections. n = 6 for each group. All data represent the mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001. Two-tailed unpaired Student’s t-test was used. Discussion The present study demonstrates that both LAPTM5 and STING are upregulated in macrophages in the lesional skin from rosacea patients as well as in LL-37-induced mouse and cellular models, accompanied by the hyperactivation of STING signaling. LAPTM5 positively regulates STING-mediated inflammatory response by maintaining the stability of STING protein, indicating the possible modulation of STING signaling in the pathogenesis of rosacea. The findings of the study are summarized in Fig. [184]9. Fig. 9. Graphical abstract: LAPTM5 exacerbates STING-mediated inflammation induced by LL-37 through stabilizing STING in rosacea. [185]Fig. 9 [186]Open in a new tab In rosacea, elevated LL-37 induces nuclear damage in lesional skin, leading to the release of nuclear DNA fragments into the cytoplasm and even the extracellular environment. These DNA fragments activate macrophage cGAS-STING signaling pathway and trigger the production of type I interferons and pro-inflammatory cytokines. LAPTM5 is upregulated by LL-37 in macrophages and interacts with STING, inhibiting its K48- and K63-linked polyubiquitination. By preventing both proteasomal and lysosomal degradation, LAPTM5 maintains STING protein stability under both resting and activated conditions, thereby enhancing STING signaling and exacerbating inflammation in rosacea. Illustration elements (human face and skin) were obtained from Figdraw 2.0. Although rosacea is a multifactorial human disease, intradermal injection of LL-37 effectively recapitulates several key pathological features observed in patients, including persistent erythema, edema, and telangiectasia, along with elevated expression of type I interferon (IFN-β) and various pro-inflammatory mediators (e.g., CXCL10, TNF-α, IL-6, IL-1β). These phenotypes have been validated in multiple studies using both clinical samples and experimental models^[187]7,[188]31,[189]33. Type I IFN expression has been reported to increase significantly during acute flare-ups in rosacea patients and is subsequently sustained by other pro-inflammatory cytokines contributing to chronic inflammation^[190]31. While LL-37–DNA complexes potently activate pDCs through TLR9 to produce IFN-α during acute flare-ups^[191]31,[192]40, our findings reveal that macrophage-derived STING activation leads to sustained production of type I IFNs and pro-inflammatory cytokines, likely contributing to the chronic inflammatory milieu in rosacea skin. pDCs and macrophages appear to play complementary roles in the initiation and maintenance of type I IFN responses in rosacea; however, this requires further investigation. STING has been shown to activate IRF3 and NF-κB pathways, resulting in the induction of type I interferons and pro-inflammatory cytokines^[193]39. Notably, STING hyperactivation has been implicated in the development of multiple inflammatory skin disorders^[194]23,[195]32,[196]41. We analyzed publicly available gene expression datasets and found that mRNA levels of STING signaling pathway–related genes were upregulated in the skin lesions of both rosacea patients and rosacea-like mouse models. Pathway enrichment analysis further revealed significant macrophage activation in lesional skin, consistent with findings from previous studies^[197]10,[198]11, but also highlighted the type I IFN signaling pathway, where STING expression serves as a key component in its activation. The protein expression of STING and pSTING was significantly increased in rosacea lesions, indicating hyperactivation of the STING signaling pathway in rosacea. Consistent with this, analysis of publicly available datasets revealed elevated STING mRNA levels in the lesional skin of rosacea patients and in LL-37–treated mouse models. This observation suggests that, in addition to post-translational regulation, STING may also be transcriptionally upregulated under chronic inflammatory conditions. Although the precise mechanisms driving STING transcriptional upregulation in rosacea remain unclear, we speculate that persistent upstream stimuli (e.g., cytokines, oxidative stress, or microbial products) may activate transcription factors such as IRF3, NF-κB, or STATs, which in turn promote STING gene expression^[199]30,[200]33. Further mechanistic studies are still needed to elucidate the upstream regulators of STING transcription under rosacea conditions. Since H-151 is extensively used to inhibit STING activity and thus blocks STING-mediated inflammatory responses in several experimental mouse models in vivo^[201]20,[202]21, we further validated the inhibitory impact of H-151 on LL-37-induced rosacea-like dermatitis. Firstly, H-151 significantly reduced STING protein levels in macrophages. Secondly, H-151 treatment significantly alleviated rosacea-like phenotypes, including erythema and inflammatory cell infiltration. Thirdly, our analysis of the inflammatory milieu in the rosacea-like mice revealed that the increased levels of inflammatory factors in lesional skin were markedly reversed by H-151 treatment. Together, these findings underscore the pivotal role of the STING signaling pathway in rosacea pathogenesis and support its potential as a therapeutic target. The progression of rosacea is closely related to the LL-37-mediated innate immune responses, characterized by significant macrophage infiltration and activation in the skin lesions^[203]7,[204]11. Rosacea patients exhibited a noteworthy correlation between STING mRNA expression and macrophage markers, suggesting the potential activation of STING signaling within macrophages. The expression of CD86, a marker of M1-type macrophages, was significantly elevated in the skin lesions of rosacea patients. In LL-37-induced rosacea-like skin lesions in mice, the expression and co-localization of F4/80 (a macrophage marker) and STING were significantly increased compared to the control group. Macrophage depletion in the mouse model led to a marked reduction in STING expression in skin lesions. These findings suggest that the STING signaling pathway is hyperactivated in macrophages within rosacea skin lesions. Although our study focused on macrophages due to their pronounced STING activation, we acknowledge that STING is also activated in other cell types, such as vascular endothelial cells, which are highly relevant to rosacea pathogenesis. To better understand the broader cellular context of STING signaling, future studies will incorporate single-cell RNA sequencing and multiplex immunofluorescence to assess STING expression across different cell populations in lesional skin. LAPTM5 was previously reported to act upstream of multiple signaling pathways as a positive inflammatory regulator in macrophages^[205]27. In our study, public transcriptomic datasets revealed increased LAPTM5 mRNA expression in rosacea lesions and LL-37-induced mouse skin, which we further validated at the protein level in patient tissues, mouse models, and in vitro cell models. Immunofluorescence analysis showed that LAPTM5 was predominantly expressed in dermal macrophages, and macrophage depletion led to a marked reduction in LAPTM5 and STING expression in lesional skin, indicating that macrophages are the principal source of LAPTM5. While its expression in other immune cell subsets cannot be excluded, future investigations employing single-cell RNA sequencing or lineage-tracing approaches will be required to fully characterize LAPTM5-expressing cell populations in rosacea. The upstream signals driving LAPTM5 upregulation remain to be fully defined. LL-37 is known to activate TFEB, a master regulator of lysosomal biogenesis. In addition, STING activation can promote TFEB nuclear translocation independently of type I IFN signaling^[206]42–[207]44. Based on these observations, we hypothesize that LAPTM5 may be transcriptionally induced as part of a STING-TFEB-CLEAR regulatory axis in response to inflammatory stress. This model links LAPTM5 to lysosomal-autophagic adaptation under inflammatory conditions and will be explored in future studies. LL-37-mediated innate immune responses are closely associated with the progression of rosacea^[208]33. It has been shown to activate inflammasomes in murine macrophages, thereby inducing rosacea-like inflammatory responses in mouse skin^[209]7. LL-37, an antimicrobial peptide produced by epidermal and immune cells, possesses both antimicrobial activity and immunomodulatory functions^[210]45. At higher concentrations, LL-37 can also induce nuclear DNA damage and fragmentation, leading to the release of nuclear DNA fragments^[211]46,[212]47. Increased TUNEL-positive staining in rosacea lesions and LL-37-treated mouse skin reflects widespread DNA fragmentation. Consistent with these findings, our results showed a significant increase in extranuclear dsDNA fragments in these lesional tissues. Although TUNEL does not distinguish nuclear from cytoplasmic localization, it supports the notion that apoptosis contributes to the pool of cytosolic DNA capable of activating STING. To estimate cytosolic DNA, we quantified extranuclear dsDNA by measuring signal intensity outside DAPI⁺ nuclear regions. While this approach minimizes nuclear signal interference, it does not allow precise determination of the DNA’s cellular origin, highlighting the need for more refined methods such as subcellular fractionation or high-resolution imaging in future studies. Future studies involving topical antibiotic pretreatment in the LL-37 mouse model may help clarify whether bacterial DNA is required for STING activation, or if endogenous host DNA is sufficient to trigger the cytosolic DNA sensing pathway. This would further distinguish STING-mediated inflammation from the previously described TLR9–pDC–IFN-α axis in rosacea. Moreover, we found that high-dose LL-37 not only activated STING and its downstream targets in cellular models, but also upregulated both STING and LAPTM5 protein levels. Together with our in vivo findings, these results support a model in which excess LL-37 promotes LAPTM5 expression and STING pathway activation, thereby exacerbating the inflammatory response in rosacea. It is well established that LAPTM5 functions as a membrane-associated protein localized to intracellular vesicles and traffics from the Golgi apparatus to late endosomes and lysosomes^[213]14,[214]48. In contrast, STING resides in the endoplasmic reticulum (ER) at baseline and translocates to the Golgi, endosomes, and lysosomes upon stimulation^[215]39. Notably, STING undergoes dynamic trafficking even under resting conditions, rather than remaining confined to the ER^[216]49. Several proteins, including iRhom2, STEEP, TOLLIP, and MYSM1, have been shown to constitutively interact with STING and play critical roles in regulating STING signaling^[217]17,[218]18,[219]50,[220]51. Our in vitro experiments demonstrated that LAPTM5 positively regulates the STING-mediated immune response by maintaining STING protein stability. Specifically, loss of LAPTM5 led to reduced STING protein levels in macrophages, accompanied by impaired activation of STING signaling and decreased production of inflammatory mediators following DMXAA or HT-DNA stimulation. Conversely, LAPTM5 overexpression increased baseline STING protein levels and prolonged its expression after stimulation in fibroblasts. In parallel with this stabilizing effect, autophagy also serves as an important counter-regulatory mechanism to limit STING activation in macrophages. Recent studies suggest that autophagy in macrophages plays a vital role in maintaining immune homeostasis during chronic inflammation^[221]52. Beyond its role in inducing inflammatory cytokines, cGAS–STING activation also triggers noncanonical autophagy^[222]53, facilitating the clearance of cytosolic DNA and pathogens, and thereby limiting excessive pathway activation^[223]44. Autophagy further modulates macrophage polarization, promoting an anti-inflammatory M2 phenotype^[224]54,[225]55. In rosacea, STING activation in macrophages may induce autophagy as a negative feedback mechanism to prevent overstimulation by promoting STING degradation via the autophagy–lysosome pathway. Disruption of this regulatory loop could exacerbate STING-driven inflammation. Further studies are needed to clarify the specific autophagy-related molecules involved in this process under rosacea conditions. Computational protein–protein interaction simulations, co-immunoprecipitation, and confocal microscopy analyses suggested that LAPTM5 can interact and colocalize with STING, with this interaction enhanced upon ligand stimulation. Our current data from co-immunoprecipitation and subcellular localization analyses support the formation of a functional LAPTM5-STING complex. These findings indicate that LAPTM5 modulates STING signaling through direct interaction, although the exact binding interface remains to be defined. Future studies will focus on identifying the key residues mediating this interaction through mutational analysis. Although the precise membrane compartment in which the LAPTM5-STING complex forms remains unclear, our colocalization data suggest that the interaction likely occurs at a post-Golgi stage, such as on late endosomes or at membrane contact sites during STING trafficking. This is supported by the known localization of LAPTM5 to endolysosomal membranes and the dynamic trafficking of STING from the ER to lysosomes upon activation. Further investigation using high-resolution imaging or subcellular fractionation will be necessary to determine the precise site of interaction. Previous studies have shown that blocking STING degradation - such as through autophagy inhibition or lysosomal blockade-leads to enhanced STING aggregation and prolonged activation of TBK1 and IRF3^[226]56,[227]57. While LAPTM5 appears to facilitate the interaction between STING and TBK1, it remains unclear whether it directly promotes STING phosphorylation. Future studies will investigate the dynamic phosphorylation status of STING and its contribution to signal amplification. Furthermore, STING protein levels in LAPTM5-knockdown cells were restored by BafA1 treatment to levels comparable to control cells, both under resting conditions and following DMXAA stimulation. In contrast, MG132 treatment restored STING protein only under resting conditions. These findings suggest that LAPTM5 may protect STING from both proteasomal and lysosomal degradation pathways. The regulation of STING protein stability is closely associated with its proteasomal and lysosomal degradation pathways, which are respectively subjected to stringent regulation by K48- and K63-linked polyubiquitination^[228]16,[229]58. In the resting state, nascent STING protein is ubiquitinated and targeted for proteasomal degradation^[230]59. Several proteins have been reported to promote STING-mediated innate immune response by removing K48-linked polyubiquitin chains from STING to inhibit its proteasomal degradation^[231]60–[232]62. On the other hand, STING undergoes K63-linked polyubiquitination after activation, and this ubiquitination is essential for its lysosomal degradation. For instance, the autophagy receptor p62 targets K63-linked ubiquitinated STING for degradation within lysosomal compartments^[233]63. Additionally, ESCRT proteins selectively identify K63-linked ubiquitinated STING during its transit from the Golgi to recycling endosomes, subsequently targeting it for lysosomal degradation^[234]56,[235]64,[236]65. Despite the importance of lysosomes in mediating the degradation of activated STING, resting-state STING may also be transported to lysosomes for degradation^[237]17,[238]57. Our study demonstrates that aberrant expression of LAPTM5 inhibits both K48- and K63-linked polyubiquitination of STING, thereby maintaining its protein stability under both homeostatic and activated conditions and expanding the intracellular STING pool. Although STING ubiquitination was not directly assessed in patient tissues, this remains an important question. Future studies will aim to validate this mechanism in lesional samples using optimized ubiquitin-enrichment protocols, and to explore the involvement of deubiquitinases in the inhibitory effect of LAPTM5. Our in vitro results demonstrate that LAPTM5 enhances STING signaling not only in macrophages but also in non-hematopoietic cells, such as fibroblasts and epithelial-derived cells, suggesting a broader regulatory role for LAPTM5 in STING-mediated inflammation. However, since these findings were derived from established cell lines, they may not fully reflect the physiological conditions in vivo. To address this limitation, future studies should examine LAPTM5-mediated STING regulation in primary immune cells or patient-derived macrophages to confirm its relevance in the context of rosacea. In addition, generating a macrophage-specific LAPTM5 knockout mouse model will be essential to definitively determine its cell-type-specific function in vivo and will be a major focus of our future research. While STING activation is implicated in multiple inflammatory skin diseases, including lupus, psoriasis, and acne vulgaris, our findings identify LAPTM5 as a previously unrecognized regulator of STING-mediated inflammation in rosacea. Whether it plays a similar role in other disease contexts warrants further investigation. Among the resident dermal innate immune cells, STING exerts pro-inflammatory effects not only in macrophages but also in dendritic cells and mast cells, all of which are implicated in the progression of rosacea^[239]66,[240]67. Herein, AAV containing shRNA targeting LAPTM5 was delivered to dermal skin and found to attenuate the phenotypes of LL-37-induced rosacea-like dermatitis, inhibit the production of inflammatory factors, and reduce the protein levels of STING and its phosphorylation. These data demonstrate that regulatory effect of LAPTM5 on the STING-mediated innate immune response is not limited to macrophages, and LAPTM5 is a potential therapeutic target for protecting mice against STING-mediated inflammation in rosacea. Conclusion In summary, the present study highlights the essential role of LL-37-induced LAPTM5 overexpression and STING hyperactivation in macrophages in the pathogenesis of rosacea. In vitro, LAPTM5 associates with STING, suppresses its ubiquitination, and prevents both proteasomal and lysosomal degradation of STING. Our findings suggest that LAPTM5 functions as a STING stabilizer both at homeostasis and after activation, thereby positively regulating the STING-mediated inflammation. Moreover, both the STING antagonist H151 and LAPTM5 knockdown effectively ameliorate LL-37-induced rosacea-like phenotypes, holding potential for treatment of rosacea. Methods Human skin biopsies Skin biopsies from rosacea patients and healthy controls were collected from the Department of Dermatology at The First Hospital of China Medical University. The Ethics Committee of The First Hospital of China Medical University approved the study (approval number: [2024]-202), which adhered to the principles of the Declaration of Helsinki. Written informed consents were obtained from all participants. The information regarding participants is presented in Supplementary Table [241]S1. Bioinformatic analysis Transcriptome datasets were obtained from the National Center for Biotechnology Information Gene Expression Omnibus ([242]GSE65914 and [243]GSE147950) and the Genome Sequence Archive (HRA000378)^[244]29,[245]30. Statistical and bioinformatic analyses were conducted using R software. Weighted gene co-expression network analysis (WGCNA) was performed on the [246]GSE65914 dataset to construct a gene co-expression network using genes with the top 30% variance. A soft-thresholding power (β) of 8 (R² = 0.78) was selected, and modules with similar expression profiles were merged at a threshold of 0.25. Cluster dendrograms and module-trait correlation heatmaps were subsequently generated. Differential expression analysis was performed using the limma package. Hierarchical clustering heatmaps were visualized using the ComplexHeatmap package, while KEGG pathway enrichment analysis was conducted for genes within modules of interest. Correlations between primary and secondary variables were assessed and visualized using lollipop plots generated by the ggplot2 package. Mouse model Female BALB/c mice (7–8 weeks old, 18–20 g) were purchased from Changsheng Biotechnology Co., Ltd. All experiments were conducted under specific pathogen-free conditions and approved by the Animal Ethics Committee of China Medical University (IRB number: CMUKT2023002). Before experiments, mice were acclimated for one week with free access to food and water, and dorsal fur was shaved 24 h prior to injections. For the H-151 intervention experiment, mice were randomly divided into four groups (n = 6 per group): (A) PBS + Vehicle, (B) PBS + H-151, (C) LL-37 + Vehicle, and (D) LL-37 + H-151. Groups C and D received intradermal injections of LL-37 (40 μL, 320 μM, Selleck) into designated dorsal areas every 12 h, for a total of four administrations. Groups B and D received intraperitoneal injections of H-151 (200 μL, 3.75 mM, Selleck) once daily for three consecutive days. Control groups received equivalent volumes of respective vehicle solutions at corresponding time points. At 12 h after the final LL-37 administration, dorsal skin erythema was photographed using a digital camera and dermatoscope. Redness severity was scored on a scale from 1 to 5, with 5 indicating the highest intensity. The erythematous area was quantified using ImageJ software. For the clodronate liposome intervention, mice were randomly divided into three groups (n = 6 per group): (A) PBS, (B) LL-37 + PBS-lipo, (C) LL-37 + Clo-lipo. On Day 1, Group C received intraperitoneal injection of 200 μL clodronate liposome (Clo-lipo) suspension, followed by an additional 100 μL injection on Day 3. At the same time points, Group B received equivalent volumes of PBS-lipo. On Day 3, Groups B and C received intradermal injections of LL-37 to induce rosacea-like dermatitis, while Group A received an equivalent volume of PBS solution. Twelve hours after the final treatment, mice were anesthetized with isoflurane and euthanized. Skin tissues from erythematous regions were harvested for further analyses. Histological analysis Mouse dorsal skin biopsies were fixed in 4% paraformaldehyde for 48 h and then embedded in paraffin. Specimens were sectioned into 5 μm slices. Following deparaffinization and rehydration, sections were stained with hematoxylin and eosin (H&E) staining. The pathological changes were analyzed under a microscope (Olympus) and imaging was performed using cell Sens Standard software (Olympus). Histopathological analysis involved quantifying the infiltrating cells in the dermis. For each sample, three non-overlapping areas were randomly selected from three sections. Immunofluorescence staining Mouse skin tissues were embedded in OCT compound (SAKURA Tissue-Tek), sectioned to an 8 µm thickness, and placed on glass slides. Cells were cultured within 24-well plates. Both tissues and cells were fixed with 4% paraformaldehyde for 15 min at room temperature, and then washed with phosphate-buffered saline (PBS) for three times. Non-specific binding was blocked by incubating 5% Bovine Serum Albumin (BSA) and 0.3% Triton X-100 for one hour at room temperature. The samples were incubated overnight at 4 °C with primary antibodies. Then, the samples were incubated with different secondary antibodies for 2 h at room temperature. Antibody information is listed in Supplementary Table [247]S5. The sections were washed with PBS for three times and then stained with 4′,6-diamidino-2-phenylindole (DAPI, Solarbio) for 15 min. Images were captured by a confocal laser scanning microscope (Olympus). Three fields were selected from each sample section for analysis. Immunohistochemistry Human and mouse skin samples were fixed in 4% paraformaldehyde, embedded in paraffin, and sectioned into 5-μm-thick slices. These sections were deparaffinized in xylene, rehydrated through graded ethanol series, and subjected to antigen retrieval using heated citrate buffer (pH 6.0) or Tris/EDTA buffer (pH 9.0). After blocking endogenous peroxidase with 3% H[2]O[2], the sections were treated with BSA for 30 min. The sections were then incubated overnight at 4 °C in a humidified chamber with a variety of primary antibodies (Antibodies information were listed in Supplementary Table [248]S5). Negative control sections were similarly processed but without primary antibodies. After incubation, they were washed with PBS and incubated with secondary antibodies from MXB Biotechnologies at 37 °C for 30 min. The staining results were photographed using an Olympus optical microscope. For each sample, three non-overlapping areas were randomly selected from three different sections. The integrated optical density (IOD) of F4/80, LAPTM5, pSTING, STING, IL-6, TNF-α, CXCL10, Interferon-β, IL-1β, and CD86 staining per view was evaluated using Image-Pro Plus (IPP) software. Cell culture and treatment Murine macrophage cells (RAW264.7), murine fibroblast cells (L929), and human embryonic kidney cells (HEK293) were obtained from the National Collection of Authenticated Cell Cultures (Shanghai, China). They were maintained in Dulbecco’s modified Eagle’s medium (DMEM, Hyclone) with 10% fetal bovine serum (FBS, Biological Industries) and 1% penicillin-streptomycin (Biological Industries) in a 5% CO[2] environment at 37 °C. For LL-37 treatment, RAW264.7 and L929 cells were placed into six-well plates and exposed to increasing doses (1, 2, 4, and 8 µM) of LL-37 for 24 h. For STING pathway activation, herring testis DNA (HT-DNA, Sigma-Aldrich), 2’,3’-cGAMP (cGAMP, Selleck), and 5,6-dimethylxanthenone-4-acetic acid (DMXAA, Selleck) were used at final concentrations of 2, 5, and 50 µg/ml for indicated time, respectively. HT-DNA or cGAMP was transfected into cells using Lipofectamine 2000 (Invitrogen) following the manufacturer’s instructions. For protein degradation pathway inhibition, at a confluency of 70%, cells were starved overnight with the incubation of 50 nM BafA1 and 10 µM MG132 (MedChemExpress). siRNA, plasmid, lentivirus construction and transfection Mouse Laptm5 siRNA and control siRNA were designed and produced by Sangon Biological Technology. Expression plasmids for Flag-tagged LAPTM5, Myc-tagged STING, HA-tagged Ubiquitin and its mutants, and LAPTM5 overexpression or interference lentiviruses used in this study were purchased from Genechem Co., Ltd. All the plasmids were identified by DNA sequencing. Transfection was carried out using jetPRIME (Polyplus) or Lipofectamine 2000 (Invitrogen) as per the manufacturer’s guidelines. The siRNA sequences were in Table [249]S2. Adenovirus associated virus (AAV) An adeno-associated virus serotype 9 (AAV9) vector carrying short-hairpin RNA (shRNA) targeting LAPTM5 was designed and produced by Hanbio Biotechnology. The Laptm5 shRNA sequences were synthesized and cloned into the pHBAAV-U6-MCS vector (shRNA sequences were in Table [250]S3). An intradermal injection of AAV suspension (1.0 × 10^11 v.g/ml) was administered to a 1 cm × 1 cm area of dorsal skin on 5-week-old female BALB/c mice (Changsheng Biotechnology). The viral suspension was administered using a 1 ml syringe, with five to seven injections of approximately 20 µl each per mouse. Twenty-one days after AAV administration, the effect of Laptm5 knockdown was assessed using immunoblot analysis. Rosacea-like dermatitis was induced by LL-37 injection as previously described. Real-time quantitative PCR (RT-qPCR) Total RNA extraction from cells was performed using TRIzol Reagent (Invitrogen), and RNA quality was assessed via a NanoDrop spectrophotometer (Thermo Scientific). mRNA was reverse-transcribed into cDNA using PrimeScript RT reagent Kit with gDNA Eraser (Takara) following manufacturer’s protocol. Real-time PCR was conducted using TB Green Premix Ex Taq II (Takara) on an Applied Biosystems 7300 Real-Time PCR System (Thermo Fisher Scientific). Each sample was analyzed in triplicate, and relative gene expression was quantified using the delta CT method. GAPDH served as the internal control gene, and fold change was normalized against the control group. Cycling conditions were: 95 °C for 10 s, followed by 45 cycles of 95 °C for 5 s and 60 °C for 30 s. The mouse-specific primer sequences (Takara) were in Table [251]S4. Immunoblotting The mouse skin biopsies and cultured cells were rinsed with cold phosphate-buffered saline (PBS) and subsequently lysed using RIPA lysis buffer (Beyotime) supplemented with a protease and phosphatase inhibitor cocktail (Selleck). Protein concentrations were determined using the bicinchoninic acid (BCA) protein assay kit (Beyotime) according to the manufacturer’s protocol. Equivalent protein amounts (20–50 μg per lane) from each sample were resolved by 10–12% SDS-PAGE (Bio-Rad) and transferred to polyvinylidene fluoride (PVDF) membranes (Bio-Rad). The membranes were blocked with 5% non-fat milk in TBS containing 0.1% Tween 20 (TBST) for 2 h at room temperature. Subsequently, they were incubated overnight at 4 °C with primary antibodies diluted in 5% BSA in TBST. After incubating with primary antibody, the membranes were incubated with TBST-diluted secondary antibodies conjugated with horseradish peroxidase (HRP) for 50 min at room temperature. The immunoreactive bands were visualized using chemiluminescent substrate (Beyotime) on Tanon 5200 Multi system. The antibody information was listed in Supplementary Table [252]S5. Co-immunoprecipitation For co-immunoprecipitation (CoIP), whole-cell lysates were prepared from the cells using IP lysis buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Nonidet P-40) supplemented with protease inhibitor cocktail. A small aliquot of lysates was retained as input. Proteins (0.5–2 mg) were precleared with 35 µl of Protein A/G PLUS Agarose beads (Santa Cruz) at 4 °C for 30 min. The supernatants were subsequently incubated with control IgG (Cell Signaling Technology) or the indicated antibodies (0.5–2 µg) and 35 µl of the agarose beads overnight at 4 °C. Following three washes with 1 mL of IP lysis buffer, the IP complexes were eluted by boiling them in 2 × sodium dodecyl sulfate (SDS) sample buffer for 5 min. Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) analysis The TUNEL assay was conducted on paraffin-embedded or frozen skin sections using a TUNEL Apoptosis Assay Kit (Beyotime). Skin sections (8 μm) were fixed in 4% paraformaldehyde and rinsed three times with PBS. They were blocked for 30–60 min using a PBS buffer supplemented with 1% bovine serum albumin (BSA) and 0.3% Triton X-100. The slides were subsequently incubated with 20 µg/ml proteinase K at 37 °C for 20 min. After three PBS washes, they were exposed to a combination of terminal deoxynucleotidyl transferase enzyme and Cy3-labeled dUTP, and incubated at 37 °C in the dark for one hour. Nuclear DNA damage was assessed using confocal microscopy and quantified using ImageJ software. Protein–protein docking The crystal structures of LAPTM5 and STING proteins were obtained from the Research Collaboratory for Bioinformatics Database ([253]https://www.rcsb.org/pdb). Subsequently, the protein crystals were prepared using the protein preparation wizard module in Schrodinger Maestro 13.5 software (Schrodinger Inc.), which included protein preprocessing, regeneration of the native ligand states, H-bond assignment optimization, protein energy minimization, and removal of waters. The processed LAPTM5 and STING proteins were then subjected to a protein–protein interaction simulation using the protein–protein docking module. The settings included 70,000 ligand rotations and up to 30 returned poses. A lower score indicated reduced binding free energy, implying greater binding stability. Protein–protein interaction complex with the lowest interaction score was marked in different colors for different chains, and a surface was added to display a 3D stereoscopic display. Additionally, the Protein Interaction Analysis module was utilized to identify the specific binding regions between LAPTM5 and STING. CHX chase assay L929 cells were infected with the LAPTM5-expressing lentivirus and the control lentivirus, followed by seeding into six-well plates for overnight incubation. Cells were subsequently treated with 100 µg/ml CHX (Selleck) for 4, 8, and 12 h. Cell lysates were collected at the specified time intervals, then examined for STING protein via immunoblot as previously described. Statistics and reproducibility All statistical analyses were performed using GraphPad Prism 8.0.2 software. Data are presented as mean ± standard error of the mean (SEM) unless otherwise stated. Differences between two groups were assessed using a two-tailed unpaired Student’s t-test, and comparisons among more than two groups were made using one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test. A p-value < 0.05 was considered statistically significant. Statistical significance is indicated as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. Correlation analysis was conducted using Pearson’s r-test. Sample sizes (n) represent the number of independent biological replicates, as detailed in the figure legends. All experiments were repeated independently at least three times with similar results. Ethical approval The collection of clinical materials was approved by The Ethics Committee of The First Hospital of China Medical University (approval number: [2024]-202), following the Declaration of Helsinki. All ethical regulations relevant to human research participants were followed. Animal experiments were approved by the Animal Ethics Committee of China Medical University (IRB No. CMUKT2023002). We have complied with all relevant ethical regulations for animal use. Supplementary information [254]Transparent Peer Review file^ (387.5KB, pdf) [255]Supplementary information^ (5.2MB, pdf) [256]42003_2025_8861_MOESM3_ESM.docx^ (21.1KB, docx) Description of Additional Supplementary Files [257]Supplementary Data^ (69.1KB, xlsx) [258]nr-reporting-summary^ (2.4MB, pdf) Acknowledgements