免疫细胞、血浆代谢物和肺癌之间的中介孟德尔随机化分析
Association between immune cells,plasma metabolites and lung cancer:a mediated Mendelian randomization analysis
朱凯丽 1曹梦悦 1苏飞 2侯小明 2令晓玲2
作者信息
- 1. 兰州大学第一临床医学院,兰州 730099
- 2. 兰州大学第一医院肿瘤内科,兰州 730099
- 折叠
摘要
目的 通过中介孟德尔随机化(MR)分析阐明免疫特征与肺癌(LC)发病之间的介导调控机制.方法 从公开的全基因组关联研究(GWAS)数据库中获取731种免疫细胞、1400种血浆代谢物、11 348例LC患者和15 861例对照的汇总统计数据.首先通过全基因组关联研究确定与变量相关的单核苷酸多态性(SNP)作为工具变量,采用逆方差加权方法推断因果关系,同时MR-艾格法、中位数法、简单模式、加权模式作为逆方差加权法的补充方法进行两步、双样本MR分析,并利用敏感性分析来验证可靠的MR结果.结果 最终中介MR分析揭示6种介导关系,从免疫细胞CD4/CD8br到LC的途径可能由4种血浆代谢物介导:5α-雄甾烷-3β,17α-二醇二硫酸盐水平[比值比(OR)=1.193,95％置信区间(CI):1.096～1.298,P＜0.05]、花生四烯酸(20:4n6)/副黄嘌呤水平(OR=1.173,95％CI:1.052～1.309,P＜0.01)促进 LC 发生,中介效应占比分别是-10.90％(P＜0.05)和-9.47％(P＜0.05);糖基-N-二十二烷基-鞘氨醇(d18∶2/22∶0)水平(OR=0.877,95％CI:0.804～0.956,P＜0.05)和 α-酮戊二酸/天冬氨酸水平(OR=0.861,95％CI:0.769～0.965,P＜0.05)抑制LC发生,中介效应占比分别是7.91％和8.80％.此外,鞘磷脂(d18∶1/22∶2、d18∶2/22∶1、d16∶1/24∶2)水平(OR=0.838,95％CI:0.745～0.942,P＜0.05)在 BAFF-R 的过渡性 B 细胞抑制 LC发生中起重要作用,中介效应占比是11.70％(P＜0.05).N-乙酰胆碱水平(OR=1.131,95％CI:1.039～1.232,P＜0.05)在终末分化CD4+T细胞上的CD3的T细胞促进LC发生中起重要作用,中介效应占比是5.85％(P＜0.05).在本研究中,敏感性分析没有异质性或多效性.结论 中介MR分析结果提示血浆代谢物可能介导免疫特征影响LC发生发展的过程.
Abstract
Objective To elucidate the mediating mechanism of association between immune char-acteristics and lung cancer(LC)by mediated Mendelian randomization analysis.Methods Summary sta-tistics of 731 immune cells,1400 plasma metabolites,11,348 LC patients and 15,861 controls were ob-tained from the publicly available Genome-wide association studies(GWAS)database.Genome-wide asso-ciation studies was used to determine the variable-related single nucleotide polymorphism(SNP)as a tool variable,and the inverse variance weighting method was used to infer the causal relationship.At the same time,MR-Eiger method,median method,simple mode and weighted mode were used as supplementary methods of the inverse variance weighting method for two-step and two-sample MR analysis,and the relia-ble MR results were verified by sensitivity analysis.Finally,the process from immune cells to LC mediated by plasma metabolites was determined.Results The causal relationship between 6 immune cells and 25 plasma metabolites on LC was further verified by sensitivity analysis.Finally,mediated MR analysis re-vealed six kinds of mediating relationships.The pathway from CD4/CD8br to LC may be mediated by four kinds of plasma metabolites:5α-androstane-3β,17α-diol disulfate level[odds ratio(OR)=1.193,95％confidence interval(95％CI):1.096-1.298,P＜0.01],arachidonic acid(20:4n6)/paraxanthine levels(OR=1.173,95％CI:1.052-1.309,P＜0.05)played an important role in promoting LC,and Media-ting effects accounted for-10.90％(P＜0.05)and-9.47％(P＜0.05)respectively.Glycosyl-N-docosanol(d18∶2/22∶0)level(OR=0.877,95％CI:0.804-0.956,P＜0.05)and α-ketoglutaric acid/aspartic acid level(OR=0.861,95％CI:0.769-0.965,P＜0.05)play important roles in inhibiting LC,with the mediating effects accounting for 7.91％and 8.80％respectively.The levels of sphingomyelin(d18∶1/22∶2,d18∶2/22∶1,d16∶1/24∶2)(OR=0.838,95％CI:0.745-0.942,P＜0.05)were involved in the development of transitional B cell inhibition LC in BAFF-R.The level of N-acetylcholine(OR=1.131,95％CI:1.039-1.232,P＜0.05)plays an important role in the promotion of LC by T cells with CD3 on CD4+,and the mediating effect accounts for 5.85％(P＜0.05).In this study,there is no hetero-geneity or pleiotropy in sensitivity analysis.Conclusion The results of intermediary MR analysis suggest that plasma metabolites may mediate immune profiling to influence LC development.
关键词
肺癌/免疫细胞/血浆代谢物/孟德尔随机化/中介分析Key words
Lung cancer/Immune cells/Plasma metabolites/Mendelian randomization/Mediation analysis引用本文复制引用
出版年
2024
NETL
NSTL
万方数据