基于孟德尔随机化探讨免疫细胞与系统性红斑狼疮的因果关系

Analysis of the Causal Relationship between Immune Cells and Systemic Lupus Erythematosus Based on Mendelian Randomization

摘要: 目的：本研究旨在通过严谨的孟德尔随机化(MR)分析框架，探究多种免疫细胞表型与系统性红斑狼疮(SLE)发病风险的因果关联。基于大规模全基因组关联研究(GWAS)汇总数据，研究首先筛选出24个强效遗传工具变量(SNPs，F统计量 = 19.54~3159.29)，并整合逆方差加权法(IVW)、加权中位数法(WM)和MR-Egger回归进行多模型因果推断。通过异质性检验(Cochran’s Q)、水平多效性检验(MR-PRESSO/MR-Egger截距)及留一法敏感性分析确保结果可靠性，最终以IVW法为核心识别出24种与SLE风险显著相关的免疫表型。结果：18种免疫表型可显著降低SLE风险(OR < 1)，其中CD3-淋巴细胞比例的保护效应最强(风险降低34.7%，OR = 0.653)；6种表型增加风险(OR > 1)，以CD4+细胞上的CD45表达危害最大(风险增加31.5%，OR = 1.315)。关键结果包括：HLA DR+ T细胞(OR = 0.859~0.908)、CD25激活型Treg (OR = 0.923)等免疫调控表型的保护作用，以及BAFF-R记忆B细胞(OR = 1.089)、CD11b髓系细胞(OR = 1.136~1.148)等促炎表型的风险效应。结论：特定免疫细胞表型(如Treg功能标志物、HLA DR表达谱及髓系活化状态)是SLE潜在干预靶点，为解析免疫机制和开发靶向策略提供因果证据。

Abstract: Objective: This study aimed to investigate the causal association between various immune cell phenotypes and the risk of systemic lupus erythematosus (SLE) using a rigorous Mendelian randomization (MR) framework. Leveraging large-scale genome-wide association study (GWAS) summary data, we identified 24 strong genetic instruments (SNPs; F-statistic = 19.54~3159.29). Causal inference was performed by integrating inverse-variance weighted (IVW), weighted median (WM), and MR-Egger regression methods. Robustness was ensured through heterogeneity testing (Cochran’s Q), horizontal pleiotropy assessment (MR-PRESSO/MR-Egger intercept), and leave-one-out sensitivity analysis. The IVW method identified 24 immune phenotypes significantly associated with SLE risk. Results: The analysis revealed that 18 immune phenotypes significantly decreased SLE risk (OR < 1), with the proportion of CD3- lymphocytes exhibiting the strongest protective effect (34.7% risk reduction; OR = 0.653). Conversely, 6 phenotypes increased risk (OR > 1), with CD45 expression on CD4+ cells showing the greatest hazard (31.5% risk increase; OR = 1.315). Key findings included the protective effects of immune regulatory phenotypes such as HLA DR+ T cells (OR = 0.859~0.908) and activated CD25+ Tregs (OR = 0.923), alongside the risk effects of pro-inflammatory phenotypes like BAFF-R+ memory B cells (OR = 1.089) and CD11b+ myeloid cells (OR = 1.136~1.148). Conclusion: Specific immune cell phenotypes (e.g., Treg functional markers, HLA DR expression profiles, and myeloid activation states) represent potential intervention targets for SLE. This study provides causal evidence for elucidating underlying immune mechanisms and developing targeted therapeutic strategies.