背景：系统性红斑狼疮疮（Systemic lupus erythematosus, SLE）是一种以育龄期女性发病为主的自身免疫性疾病，临床表现异质性大。CD4⁺ T细胞活化和分化通过分泌细胞因子及辅助B细胞分泌抗体参与自身免疫反应，SLE的CD4⁺ T细胞存在多个细胞亚群表型及功能的异常，包括Th1、Th2、Th17、Treg、Tfh等。多项研究表明SLE CD4⁺ T细胞还存在线粒体超级化及功能障碍。线粒体自噬为健康细胞清除受损线粒体的一种代谢方式，经典的线粒体自噬通路由PINK1-Parkin介导，其在SLE CD4⁺ T细胞的研究尚属空白。

目的：（1）探究SLE患者CD4⁺ T细胞及其亚群中线粒体自噬的状态及其主要分子PINK1、Parkin的表达，线粒体自噬与疾疾病活动度SLEDAI评分的相关性；探究PINK1缺陷对T细胞功能影响的机制；（2）探究PINK1对小鼠CD4⁺ T效应细胞亚型分化的影响；（3）探究线粒体自噬缺陷对SLE模型小鼠疾病表型及免疫细胞的影响，及其与I型干扰素的相关性。

方法：（1）本研究首先通过透射电镜、实时荧光定量 PCR（Real-time quantitative PCR，RT-qPCR）和蛋白免疫印迹（Western blot，WB）技术首先确定正常健康对照者（Healthy control，HC）和SLE患者CD4⁺ T细胞线粒体自噬情况。进一步使用线粒体自噬特异性探针分析CD4⁺ T细胞及其各亚群线粒体自噬及与SLEDAI的相关性，通过转染siPINK1后对CD4⁺ T细胞表型及T细胞系Jurkat细胞状态的影响。（2）分选野生型和PINK1敲除鼠体内的naïve CD4⁺T细胞进行T细胞亚群的体外诱导分化，同时过继转移至RAG2敲除鼠体内构建T细胞免疫工具模型即T细胞肠炎模型观察体内T细胞诱导分化差异。（3）对PINK1和Parkin全敲小鼠使用TLR7激动剂咪喹莫特（Imiquimod，IMQ）进行耳部涂抹构建狼疮小鼠模型观察临床表型差异，通过酶联免疫吸附试验（Enzyme Linked Immunosorbent Assay，ELISA）、免疫组织化学（Immunohistochemistry，IHC）、流式细胞术（Flowcytometry，FCM）检测血清学、组织病理学和免疫学的差异。

结果：（1）与HC相比，SLE患者CD4⁺ T细胞线粒体皱缩，线粒体DNA减少，PINK1和Parkin在RNA和蛋白水平均降低，进一步分析细胞亚群的线粒体自噬发现Treg细胞的线粒体自噬水平降低最为明显（HC vs SLE：45512.8000 ± 3293.87231 vs 26576.8462 ± 2864.59111，P < 0.0001）。干扰PINK1后体外诱导分化小鼠Treg细胞数量无明显影响，而人Treg细胞数量可降低（siNC vs siPINK1：11.68 ±    1.729 vs 9.160 ± 1.293，P < 0.05）。干扰Jurkat细胞PINK1后Treg细胞功能相关p-Akt通路受损，细胞早期凋亡增加，但可被糖酵解抑制剂拯救。（2）PINK1缺失的naïve CD4细胞在体外诱导分化为Th1细胞的比例更高（WT vs KO：9.040 ± 3.771 vs 33.90 ± 7.063, P ＜ 0.05），且与T细胞肠炎的体内诱导结果一致(肠系膜淋巴结WT vs KO：5.481 ± 0.8418 vs 15.62 ± 1.270，P ＜ 0.001），线粒体自噬激动剂尿石素A体外处理狼疮小鼠MRL/lpr小鼠的CD4⁺ T细胞观察到Th1细胞比例降低。（3）小鼠经8周IMQ诱导红斑狼疮模型，Parkin敲除（Knock out，KO）鼠以自身抗体增多（WT vs KO：14836.60 ± 1761.31 vs 36629.2 ± 7113.08, P ＜ 0.01）和肾脏损害为主，PINK1敲除鼠以皮肤损害为主，但都与组织中I型IFN相关蛋白STING表达增加相关。Parkin KO鼠和PINK1 KO鼠在I型IFN分泌细胞即浆样树突状细胞存在不一致的变化趋势（淋巴结WT vs Parkin KO：3.27 ± 0.5567 vs 5.037± 0.3268，P < 0.05；WT vs PINK1 KO：1.490 ± 0.2796 vs 1.250 ± 0.2774，P ＞ 0.05）。

结论：（1）SLE患者CD4⁺ T细胞存在PINK1依赖的线粒体自噬障碍，且Treg细胞表现最为显著；PINK1可影响人Treg的表型、p-Akt通路及早期凋亡。糖酵解抑制剂可能为有效的拮抗剂。（2）PINK1 KO的naïve T细胞体内体外诱导分化结果提示促进了Th1细胞的分化。（3）PINK1和Parkin敲除鼠经TLR7激动剂IMQ诱导狼疮样模型后临床表现不一致，提示线粒体自噬相关蛋白具有细胞及组织异质性，靶向线粒体自噬治疗SLE还需要更多的具体到组织和基因的研究数据。

Background: Systemic lupus erythematosus (SLE) is a kind of autoimmune disease mainly in women of reproductive age. After activation and differentiation, CD4⁺ T cells participate in the immune response by secreting cytokines and helping B cells to secret antibodies. CD4⁺ T cells in SLE have abnormalities in the phenotype and function of several cell subsets, including Th1, Th2, Th17, Treg, Tfh, etc. Several studies have shown that SLE CD4⁺ T cells have mitochondrial hyperpolarization and dysfunction compared to healthy controls. Mitophagy is a pathway for healthy cells to clear damaged mitochondria, and the classical mitophagy is mediated by PINK1-Parkin pathway. It is worth noting that mitophagy has not been well-studied in SLE CD4⁺ T cells.

Objectives: (1) To investigate the status of mitophagy and the expression of its major molecules PINK1 and Parkin in SLE CD4⁺ T subsets, and the correlation between mitophagy and the SLEDAI score of disease activity; to explore the mechanism of PINK1 deficiency on T cell function; (2) To investigate the effect of PINK1 on the differentiation of effector T cells; (3) To investigate whether defective mitophagy will impact the clinical phenotype and immune cells of SLE mouse model and the correlation with type I interferon (IFN).

Methods: (1) In this study, we first determined the mitophagy of healthy control (HC) and SLE total CD4⁺ T cells by transmission electron microscopy (TEM), real-time quantitative PCR (RT-qPCR) and western immunoblot (WB) techniques. Mitophagy-specific probes were used to analyze mitophagy in CD4⁺ T cells and its various subsets and further, correlation with SLEDAI. We further transfected siPINK1 to explore the phenotype change of CD4⁺ T cells and the state of Jurkat cells. (2) In vitro induction of T cell subpopulation differentiation was performed by sorting naive CD4⁺ T cells from wild-type and PINK1 knockout (KO) mice, while T cell colitis mouse model in RAG2 knockout mice was constructed through cell transfer to observe the difference of in vivo T cell differentiation. (3) Topical TLR7 agonist IMQ treatment of WT, PINK1 KO and Parkin KO mice to construct a lupus-like mouse model to observe clinical phenotypic differences at the tissue level and cellular level by ELISA, immunohistochemistry (IHC), and flowcytometry (FCM).

Results: (1) Compared to HC, CD4⁺ T cells from SLE patients showed mitochondrial crinkling, reduced mitochondrial DNA, and reduced PINK1 and Parkin at both RNA level and protein level. Further analysis in cell subpopulations revealed the most significant reduction in mitophagy level in Treg cells (HC vs SLE：45512.8000 ± 3293.87231 vs 26576.8462 ± 2864.59111，P < 0.0001). Interference of PINK1 had no significant effect on the differentiation of mouse Treg cells in vitro, but reduced the differentiation of human Treg cells (siNC vs siPINK1：11.68 ±      1.729 vs 9.160 ± 1.293，P < 0.05). After PINK1 interference in Jurkat cells, p-Akt pathway, a critical pathway for Treg function, was impaired and early apoptotic cells were increased, but they could be rescued by glycolysis inhibitor. (2) Higher percentage of differentiation into Th1 cells were observed in PINK1-deficient CD4⁺ T cells in vitro (WT vs KO：9.040 ± 3.771   vs 33.90 ± 7.063, P ＜0.05) and was consistent with the induction of T-cell colitis in vivo(WT vs KO：5.481 ± 0.8418 vs 15.62 ± 1.270，P ＜0.001), and a reduced percentage of Th1 cells was observed in CD4⁺ T cells in lupus mice treated with the mitochondrial autophagy agonist urolithin A in vitro. (3) In the IMQ induced lupus mouse model, increased autoantibodies (WT vs KO：14836.60 ± 1761.31 vs 36629.2 ± 7113.08, P ＜0.01) and kidney damage were predominantly observed in Parkin knockout mice while skin damage was more significant in PINK1 knockout mice, but all of these changes were associated with increased expression of the type I IFN-related protein STING in the tissues. At the cellular level, Parkin KO and PINK1 KO mice showed opposite trends in type I IFN-secreting cells, plasmacytoid dendritic cells (WT vs Parkin KO：3.27 ± 0.5567 vs 5.037 ± 0.3268，P < 0.05；WT vs PINK1 KO：1.490 ± 0.2796 vs 1.250 ± 0.2774，P ＞ 0.05).

Conclusions: (1) The deficiency of PINK1-dependent mitophagy was present in CD4⁺T cells of SLE patients, and especially, Treg cells. PINK1 can affect the phenotype, p-Akt pathway and early apoptosis of human Treg, and glycolysis inhibitor may be an effective antagonist. (2)T cell differentiation in vitro and in vivo suggested that PINK1 KO results in promoted Th1 cell differentiation. (3) The inconsistent clinical phenotypes of PINK1 KO and Parkin KO mice after induction of lupus-like model by IMQ suggests that those two mitophagy-related proteins have cellular and tissue heterogeneity, and regulating mitophagy to treat SLE requires more data of tissue- and gene-specific studies.