背景：

系统性红斑狼疮（Systemic Lupus Erythematosus, SLE）是一种高度异质性的系统性自身免疫病，其特征为免疫系统对自身抗原的异常识别，进而引发多器官损伤^[1]。作为全球重大健康负担，SLE每年影响约341万人^[2]。目前研究认为SLE发病与免疫系统紊乱密切相关，但现有理论尚无法完全解释其临床异质性。

皮肤作为机体抵御外界环境的第一道屏障，在维持组织稳态中发挥关键作用^[3-6]。超过75%的SLE患者会出现皮肤病变，其中约30%以皮肤症状为首发表现。值得注意的是，高达92%的患者存在光敏现象，紫外线暴露可诱发或加重病情^[7]。机制研究显示，角质形成细胞功能异常是SLE发病进程中的重要环节，包括IFNκ的过度分泌、Hippo通路异常活化等^[8-10]。SLE患者皮损处普遍存在角质形成细胞功能障碍，表现为炎症反应增强和I型干扰素（IFN）信号过度激活^[11-14]。甚至无明显皮损的狼疮患者皮肤也呈现IFN水平升高，提示角质形成细胞可能参与疾病早期发生^[15]。这些发现共同提示皮肤免疫微环境紊乱参与SLE的发生发展。

我们前期对SLE患者皮肤病变组织的转录组学分析发现，过氧化物酶体增殖物激活受体（Peroxisome Proliferator-Activated Receptor, PPAR）信号通路在皮损组织中呈现显著下调。其中，PPARγ作为调控脂质代谢的核心转录因子，近年来在SLE发病机制研究中备受关注，但现有研究多聚焦于其在免疫细胞中的调节功能^[16]。虽然体外实验证实PPARγ激动剂可抑制SLE患者外周免疫反应^{[17, 18]}，但小鼠模型显示系统性炎症发生后，PPARγ的激活并不能有效延缓疾病进展^{[16, 19]}。这一现象在临床试验中得到验证：口服吡格列酮（PPARγ激动剂）未能显著改善患者病情^[20]，提示PPARγ的活化受阻可能并非SLE关键环节。而在皮肤中，PPARγ的作用存在争议：一方面，PPARγ缺失会加剧秃毛SKH-1小鼠的皮肤炎症和肿瘤发生^[21]；另一方面，研究显示角质形成细胞PPARγ对皮肤稳态维持和创伤修复影响有限^[22-24]。因此，角质形成细胞中PPARγ在SLE中的功能仍有待研究。

目的：

研究角质形成细胞PPARγ在SLE疾病发生和发展中扮演的角色，进一步揭示其对SLE疾病进程的影响机制，为寻找SLE新的治疗靶点提供实验依据。

方法：

检测SLE患者皮损的转录组学数据，分析包括PPAR通路在内的相关通路的改变，并通过免疫组化验证PPARγ蛋白含量的改变。进一步将 SLE 皮损组织与健康皮肤、疾病对照（特应性皮炎及银屑病）进行比较，以明确PPARγ在SLE患者角质形成细胞中的含量变化及其细胞定位特征。

构建角质形成细胞Pparg条件性敲除小鼠（Pparg^fl/flKrt5^creERT2/+），并在双耳腹侧及背侧皮肤涂抹 4-羟基他莫昔芬（4-OHT）以诱导局部皮肤角质形成细胞中 Pparg 的敲除。随后，观察小鼠是否出现狼疮样皮损、血清自身抗体升高、尿蛋白增加等 SLE 样表型。

完善角质形成细胞Pparg条件性敲除小鼠皮损的单细胞测序，解析皮损组织内各类细胞的比例变化，并评估角质形成细胞IFN通路的激活状态。体外构建PPARγ敲除/敲降、过表达及敲除后回补的角质形成细胞，并在有/无UVB照射条件下，比较IFN通路的激活程度，以探讨PPARγ对角质形成细胞IFN信号的调控。

采用免疫荧光染色和蛋白共沉淀实验，验证PPARγ是否与干扰素调节因子3（IRF3）存在直接相互作用。体外构建 IRF3 敲除以及 PPARγ/IRF3双敲除的角质形成细胞，并在有/无 UVB 照射条件下比较IFN通路的激活情况。通过染色质免疫沉淀实验验证PPARγ敲除前后的角质形成细胞中IRF3与IFNβ启动子区域的结合能力。通过分子对接分析确定了PPARγ与IRF3的潜在结合位点。在 PPARγ cDNA 中的这些位点引入了突变，构建结合位点突变的 PPARγ过表达质粒并分别转染至角质形成细胞中，以评估PPARγ结合IRF3对 UVB 照射后角质形成细胞内IFN通路的影响。在体内实验方面，构建角质形成细胞特异性Pparg和Irf3双基因条件性敲除小鼠，并与角质形成细胞Pparg条件性敲除小鼠对比，以解析IRF3在PPARγ介导的SLE样表型调控中的作用。

对比分析 SLE 患者与角质形成细胞Pparg条件性敲除小鼠皮损的单细胞测序数据，评估皮损内树突状细胞的激活状态，并探讨其与角质形成细胞IFN通路激活的相关性。在体外实验中，利用接受UVB照射的PPARγ敲除/敲降、过表达及敲除后回补PPARγ的角质形成细胞，与单核细胞来源的树突状细胞（moDC）进行非接触共培养，并通过流式细胞术检测CD83和CD86的表达水平，以评估 moDC 的激活状态。

构建Pparg^fl/+Krt5^creERT2/+Itgax^DTR/+小鼠，在诱导角质形成细胞 Pparg 条件性敲除的同时通过皮内注射白喉毒素选择性耗竭皮内树突状细胞（DC），以观察 SLE 样表型的变化。进一步，从Pparg条件性敲除小鼠的外周血及皮损组织中分选DC，并分别与脾脏来源的CD4⁺ naive T 细胞共培养，以评估DC对CD4⁺ naive T细胞的促分化作用。

在诱导小鼠角质形成细胞Pparg条件性敲除的同时，于其背部皮肤涂抹 FITC 溶液，并比较皮肤引流淋巴结中皮肤来源 FITC⁺ DC的比例变化，以评估皮肤DC的迁移能力。进一步，对比Pparg条件性敲除小鼠与对照小鼠的引流淋巴结大小及重量，并采用流式细胞术分析T细胞亚型比例的变化，以评估角质形成细胞Pparg缺失对皮肤引流淋巴结T细胞增殖分化的影响。

结果：

1. 在 SLE 患者的皮损组织中，PPAR信号通路显著下调。与健康皮肤相比，PPARγ 蛋白水平在 SLE、特应性皮炎和银屑病患者的皮损中均显著下降。然而，在特应性皮炎和银屑病的皮损中，PPARγ信号的减弱主要集中于棘层和颗粒层的角质形成细胞，而基底层角质形成细胞的PPARγ信号依然较为保留。相较之下，在 SLE 患者的皮损中，大多数基底层角质形成细胞的PPARγ蛋白明显减少。

2. 在小鼠双耳皮肤诱导角质形成细胞特异性Pparg条件性敲除后，观察到小鼠出现狼疮样皮损，伴随自身抗体水平显著升高、尿蛋白增加。HE 染色显示皮损及肾脏组织中存在显著的淋巴细胞浸润，免疫荧光检测揭示皮肤基底膜带及肾小球内IgG沉积明显增强。此外，敲除小鼠的脾脏体积显著增大，流式细胞术分析显示Th1、Th2、Th17、Tfh、Treg及浆细胞的比例均显著升高，提示系统性免疫激活。

3. 单细胞测序数据提示小鼠角质形成细胞Pparg条件性敲除后皮损部位角质形成细胞中的IFN通路明显激活。体外敲除或敲降角质形成细胞的PPARγ可加剧UVB照射引发的IFN通路激活，而过表达或在敲除后回补PPARγ的角质形成细胞则可以抑制UVB诱导的IFN通路激活。这些结果表明，角质形成细胞中的PPARγ在调控IFN信号传导中发挥重要作用。

免疫荧光染色明确在UVB照射后PPARγ与IRF3存在细胞核空间共定位；Co-IP实验明确 PPARγ与IRF3存在直接相互作用。IRF3 敲除以及 PPARγ/IRF3 双敲除的角质形成细胞均可抑制UVB 诱导的IFN通路的激活，ChIP实验发现敲除PPARγ的角质形成细胞中IRF3与IFNβ启动子区域的结合显著增强。将突变了IRF3 结合位点的 PPARγ过表达质粒转染至角质形成细胞后无法抑制UVB诱导的IFN通路的激活，提示PPARγ与IRF3蛋白相互结合以抑制其转录活性。而当PPARγ表达降低时，PPARγ与IRF3结合减少，IRF3与IFNβ 启动子区域结合增强，从而诱导IFN通路的激活。体内实验证明角质形成细胞条件性敲除 Pparg 的基础上再敲除Irf3，双敲除小鼠的皮损、尿蛋白、自身抗体、肾脏IgG沉积、脾脏重量、表皮IFNβ水平均显著缓解。此外，在角质形成细胞条件性敲除 Pparg的小鼠皮损内注射IFNAR1抗体，小鼠的皮损、尿蛋白、自身抗体也均显著缓解。

SLE 患者与角质形成细胞Pparg条件性敲除小鼠皮损内DC均显著激活，且其激活状态与角质形成细胞IFN通路激活呈显著相关。体外共培养实验发现接受 UVB 照射的PPARγ敲除/敲降角质形成细胞可以显著上调moDC的CD83、CD86的表达，促进moDC的成熟。

耗竭皮内DC可以缓解角质形成细胞Pparg条件性敲除小鼠的SLE 样表型。角质形成细胞Pparg条件性敲除小鼠外周血来源的DC对脾脏来源的 CD4⁺ naive T 细胞没有明显的促分化作用，而其皮损来源的DC可以显著促进脾脏来源的CD4⁺ naive T 细胞向Th1、Th2、Th17、Tfh分化。

角质形成细胞Pparg条件性敲除的小鼠，其引流淋巴结显著增大变重，其中 Th1、Th2、Tfh比例显著升高。在这些小鼠背部皮肤涂抹FITC溶液，引流淋巴结内FITC⁺ DC比例显著升高，提示皮肤来源的DC向引流淋巴结的迁移，介导T细胞分化。

结论：

在SLE患者皮损的角质形成细胞中PPARγ表达显著降低，在小鼠中模拟这一现象，敲除双耳皮肤角质形成细胞内的Pparg可快速诱发SLE样表型。伴随PPARγ减少，IRF3在I型IFN位点占据增加，促进IFN分泌并激活局部DC。这些被激活的DC随后迁移至引流淋巴结，以非MHC II依赖方式激活CD4⁺ naive T细胞并促进其分化为效应T细胞，最终导致疾病发生。研究表明角质形成细胞减少或缺失可能是SLE致病驱动因素，强调了皮肤免疫在SLE发病中的关键作用。

Background:

Systemic lupus erythematosus (SLE) is a chronic autoimmune rheumatic disease characterized by generalized loss of immune tolerance and inflammation in numerous organs^[1]. This disease is a major global health concern, affecting approximately 3.41 million individuals each year^[2]. Current research has attributed SLE onset primarily to immune system disorders; however, existing theories are insufficient to fully explain the origins and clinical symptoms of SLE, for example, skin involvement in SLE.

The skin is the primary barrier against environmental assault and maintains tissue homeostasis^[3-6]. Cutaneous manifestations are evident in more than 75% of patients with SLE^[7]. Indeed, skin involvement can sometimes be the initial indication of the disease^[7]. Photosensitivity, which exacerbates or triggers the disease under sunlight or ultraviolet radiation exposure, is observed in up to 92% of SLE patients. Mechanistic studies on this phenomenon have focused on keratinocyte dysfunctions, such as the overproduction of epidermal-derived interferon kappa or abnormal activation of the Hippo signaling pathway^[8-10]. Moreover, there is substantial evidence indicating that keratinocyte dysfunction is prevalent in the skin lesions of SLE patients and is characterized by an enhanced inflammatory response and hypersensitive IFN responses^[11-14]. Even the clinically normal-appearing skin of lupus patients shows elevated levels of type I interferons, highlighting the important role of keratinocytes in the early stages of the disease^[15]. These findings collectively underscore the potential significance of abnormal skin cell function in the pathogenesis of SLE.

Our preliminary transcriptomic analysis of skin lesions from SLE patients revealed significant downregulation of the peroxisome proliferator-activated receptor (PPAR) signaling pathway in affected tissues. PPARγ, as a master transcriptional regulator of lipid metabolism, has garnered increasing attention in SLE pathogenesis research. However, current studies have predominantly focused on its regulatory functions in immune cells^[16-19]. While in vitro studies suggest that PPARγ agonists can suppress immune responses in the peripheral blood of SLE patients^[17], findings from in vivo mouse experiments indicate that PPARγ activation following the onset of systemic inflammation in SLE patients does not substantially hinder disease progression^{[16, 19]}. This finding was further substantiated by clinical trials in SLE patients in which oral administration of pioglitazone failed to significantly improve disease severity^[20], suggesting that the importance of PPARγ activity in SLE is limited. In addition, the role of PPARγ in skin-related studies is controversial. PPARγ depletion exacerbates local inflammation and tumor development in SKH-1 hairless albino mice^[21]. However, other studies reported that PPARγ in keratinocytes does not play an important role in skin homeostasis or wound healing^[22-24]. These contradictory findings have prevented the elucidation of the function of PPARγ in keratinocytes within the context of SLE. Elucidating the role of skin homeostasis in SLE, particularly the common decrease in PPARγ in the keratinocytes of SLE skin lesions, represents a crucial step in revisiting and comprehensively understanding SLE on the basis of the clinical manifestations observed in patients.

Objective:

This study aims to investigate the role of keratinocyte PPARγ in the onset and progression of SLE, further elucidating the mechanisms through which it influences the disease process, and providing experimental evidence for the identification of novel therapeutic targets for SLE.

Methods:

1. Analyze the alterations in the PPAR pathway based on transcriptomic data from skin lesions of SLE patients, and validate the changes in protein expression levels through immunohistochemistry. Additionally, compare SLE skin lesions with healthy skin, atopic dermatitis, and psoriasis controls to clarify the expression changes of PPARγ in keratinocytes and its tissue localization characteristics.

2. Construct keratinocyte-specific Pparg conditional knockout mice (Pparg^fl/flKrt5^creERT2/+), and apply 4-hydroxytamoxifen to the ventral and dorsal skin of both ears to induce local keratinocyte-specific knockout of Pparg. Subsequently, observe whether the mice develop lupus-like skin lesions, elevated serum autoantibodies, and increased urinary protein, resembling SLE-like phenotypes.

3. Perform single-cell sequencing analysis of skin lesions in keratinocyte-specific Pparg conditional knockout mice to assess the changes in the proportions of various cell types within the lesions and evaluate the activation status of the keratinocyte interferon pathway. In vitro, construct keratinocyte models with PPARγ knockout/knockdown, overexpression, and PPARγ reconstitution after knockout, and compare the activation levels of the interferon pathway under UVB exposure or non-exposure conditions to investigate the regulatory role of PPARγ in keratinocyte-mediated interferon signaling.

4. Use immunofluorescence staining and co-immunoprecipitation (Co-IP) experiments to verify whether PPARγ directly interacts with interferon regulatory factor 3 (IRF3). In vitro, construct keratinocyte models with IRF3 knockout and PPARγ/IRF3 double knockout, and compare the activation of the interferon pathway under UVB exposure or non-exposure conditions. Molecular docking analysis was used to identify potential binding sites on IRF3. Mutations were introduced at these sites in the PPARγ cDNA, and a plasmid overexpressing the mutated PPARγ was constructed and transfected into keratinocytes to assess how alterations in the binding ability of PPARγ to IRF3 affect UVB-induced activation of the interferon pathway. In in vivo experiments, keratinocyte-specific Pparg and Irf3 double gene conditional knockout mice were generated to investigate the effects of combined Pparg and Irf3 deletion on the development of SLE-like phenotypes, including skin lesions, autoantibody production, and renal damage, as well as their impact on the interferon signaling pathway.

5. Comparative analysis of single-cell sequencing data from skin lesions of SLE patients and keratinocyte-specific Pparg conditional knockout mice was performed to evaluate the activation status of dendritic cells within the lesions and explore its correlation with IFN pathway activation in keratinocytes. In vitro experiments were conducted using UVB-irradiated PPARγ-knockout, knockdown, overexpressing, or PPARγ-reconstituted keratinocytes co-cultured (non-contact) with monocyte-derived dendritic cells (moDCs). Flow cytometry was employed to detect CD83 and CD86 expression levels to assess moDC activation status.

6. Pparg^fl/+Krt5^creERT2/+Itgax^DTR/+ mice were generated to achieve keratinocyte-specific Pparg conditional knockout while selectively depleting dermal dendritic cells (DCs) through intradermal diphtheria toxin injection, thereby observing changes in SLE-like phenotypes. Furthermore, DCs were sorted from peripheral blood and skin lesions of Pparg conditional knockout mice and co-cultured with spleen-derived CD4⁺ naive T cells respectively to evaluate the pro-differentiation effects of DCs on CD4⁺ naive T cells.7. Keratinocyte-specific Pparg conditional knockout mice, with FITC solution applied to the dorsal skin, showed a significant enlargement and weight increase in their draining lymph nodes, along with a marked elevation in the proportion of FITC+ DCs. This suggests enhanced migration of skin-derived DCs to the draining lymph nodes. Flow cytometry analysis revealed a significant increase in the proportions of Th1, Th2, and Tfh cells within the draining lymph nodes.

Results:

1. In the skin lesions of SLE patients, the PPAR signaling pathway is significantly downregulated. Compared to healthy skin, the protein level of PPARγ is markedly reduced in the skin lesions of patients with SLE, atopic dermatitis, and psoriasis. However, in the skin lesions of atopic dermatitis and psoriasis, the attenuation of PPARγ signaling is primarily concentrated in the keratinocytes of the spinous and granular layers, whereas PPARγ signaling in the basal layer keratinocytes remains relatively preserved. In contrast, SLE skin lesions show a more widespread reduction in PPARγ expression across all epidermal layers, including the basal layer. This suggests that the modulation of PPARγ signaling may differ between these dermatological conditions, potentially influencing the pathogenesis and immune responses involved.

2. After inducing keratinocyte-specific Pparg conditional knockout in the skin of both ears of mice, lupus-like skin lesions were observed, accompanied by a significant increase in autoantibody levels and elevated urinary protein. Hematoxylin and eosin (HE) staining revealed significant lymphocytic infiltration in both the skin lesions and renal tissue. Immunofluorescence detection showed enhanced deposition of IgG in the skin basement membrane zone and glomeruli. Additionally, the spleen of the knockout mice exhibited notable changes in its immune cell composition, with an increased number of activated T and B cells, suggesting a systemic autoimmune response.

3. Single-cell sequencing data indicate that after keratinocyte-specific Pparg conditional knockout in mice, the interferon pathway is significantly activated in the keratinocytes at the site of skin lesions. In vitro, PPARγ knockout or knockdown in keratinocytes exacerbates UVB-induced activation of the interferon pathway, while overexpression of PPARγ or reconstitution of PPARγ after knockout in keratinocytes suppresses UVB-induced interferon pathway activation. These results suggest that PPARγ in keratinocytes plays a crucial role in modulating the activation of the interferon pathway.

4. Immunofluorescence staining clearly demonstrated nuclear co-localization of PPARγ and IRF3 following UVB exposure. Co-immunoprecipitation (Co-IP) experiments confirmed a direct interaction between PPARγ and IRF3. Both IRF3 knockout and PPARγ/IRF3 double knockout keratinocytes were able to suppress UVB-induced activation of the interferon pathway. ChIP experiments revealed that in PPARγ knockout keratinocytes, IRF3 binds to the IFNβ promoter region, suggesting that PPARγ modulates the interaction between IRF3 and the IFNβ promoter, thereby influencing interferon pathway activation. In vivo experiments demonstrated that further knockout of Irf3 in keratinocyte-specific Pparg conditional knockout mice significantly alleviated skin lesions, urinary protein levels, autoantibodies, renal IgG deposition, spleen weight, and epidermal IFNβ levels. Furthermore, in keratinocyte-specific Pparg conditional knockout mice, intradermal injection of IFNAR1 antibody also led to significant alleviation of skin lesions, urinary protein, and autoantibodies.

5. Immunofluorescence staining clearly demonstrated nuclear co-localization of PPARγ and IRF3 following UVB exposure. Co-immunoprecipitation experiments confirmed a direct interaction between PPARγ and IRF3. Both IRF3 knockout and PPARγ/IRF3 double knockout keratinocytes were able to suppress UVB-induced activation of the interferon pathway. ChIP experiments revealed that in PPARγ knockout keratinocytes, IRF3 binds to the IFNβ promoter region, suggesting that PPARγ modulates the interaction between IRF3 and the IFNβ promoter, thereby influencing interferon pathway activation. In vivo experiments demonstrated that further knockout of Irf3 in keratinocyte-specific Pparg conditional knockout mice significantly alleviated skin lesions, urinary protein levels, autoantibodies, renal IgG deposition, spleen weight, and epidermal IFNβ levels. Furthermore, in keratinocyte-specific Pparg conditional knockout mice, intradermal injection of IFNAR1 antibody also led to significant alleviation of skin lesions, urinary protein, and autoantibodies.

6. Depletion of dermal DCs can alleviate the SLE-like phenotypes in keratinocyte-specific Pparg conditional knockout mice. DCs from the peripheral blood of these knockout mice do not show a significant effect on the differentiation of spleen-derived CD4⁺ naive T cells. However, DCs derived from skin lesions significantly promote the differentiation of spleen-derived CD4⁺ naive T cells into Th1, Th2, Th17, and other subsets, suggesting a localized immune response in the skin lesions that influences T cell differentiation.

7. In keratinocyte-specific Pparg conditional knockout mice, application of FITC solution to the dorsal skin resulted in significant enlargement and increased weight of the draining lymph nodes, along with a marked elevation in the proportion of FITC+ DCs. This suggests enhanced migration of skin-derived DCs to the draining lymph nodes. Flow cytometry analysis revealed a significant increase in the proportions of Th1, Th2, and Tfh cells in the draining lymph nodes.

Conclusions:

We found that the level of PPARγ is substantially reduced in the skin lesions of patients, and replicating this reduction in mice led to rapid disease onset with multiple hallmarks of SLE. As PPARγ decreases in keratinocytes, which is accompanied by increased occupancy of interferon regulatory factor 3 at the IFN locus, DCs are recruited to the epidermis and are activated by keratinocyte-secreted type I interferon. These activated DCs migrate to local draining lymph nodes, where they activate CD4⁺ T cells in a non-MHC II-dependent manner, promoting their differentiation into effector T cells and thus contributing to disease onset. Our study revealed that the dysregulation of keratinocytes can be a pathogenic driver of SLE and also emphasize the pivotal role of skin immunity in the onset of SLE.