遗传预测的β-羟基丁酸与冠状动脉粥样硬化:孟德尔随机化及APOE相关的人动脉粥样硬化斑块转录组特征
本研究整合孟德尔随机化、跨性状GWAS、ApoE pQTL-MR及共定位分析,结合转录组、机器学习和单细胞分析,探究酮体与冠状动脉粥样硬化的遗传关联及APOE相关分子特征。遗传预测的β-羟基丁酸与冠状动脉粥样硬化风险相关(OR=1.59,95%CI:1.29–1.95,p=8.8×10⁻⁶),而乙酰乙酸和丙酮无显著关联。APOE相关的五基因特征在斑块中富集于巨噬细胞、内皮细胞和成纤维细胞。
要点
- 遗传预测的β-羟基丁酸与冠状动脉粥样硬化风险相关(OR=1.59,95%CI:1.29–1.95,p=8.8×10⁻⁶)。
- 乙酰乙酸和丙酮与冠状动脉粥样硬化无显著关联。
- APOE相关五基因特征在GSE43292中AUC为0.956,GSE100927中AUC为0.974,GSE28829中AUC为0.736。
核心结论
- β-羟基丁酸是与冠状动脉粥样硬化易感性关联最强的酮体,APOE相关的多细胞特征可能有助于分子分层。
研究设计与样本
- 研究整合孟德尔随机化、跨性状GWAS分析、ApoE pQTL-MR和共定位分析,并结合批量转录组分析、机器学习特征构建、免疫浸润分析和单细胞分析。
主要发现
- 单变量孟德尔随机化中,遗传预测的β-羟基丁酸与冠状动脉粥样硬化风险相关(OR=1.59,95%CI:1.29–1.95,p=8.8×10⁻⁶)。
- 相互调整乙酰乙酸和丙酮后,β-羟基丁酸仍显著相关(OR=1.58,95%CI:1.14–2.18,p=0.005),而乙酰乙酸和丙酮均无显著关联。
- ApoE pQTL-MR支持遗传预测的ApoE与冠状动脉粥样硬化相关(OR=1.23,p=0.003)。
- WGCNA和机器学习算法(LASSO、随机森林、SVM-RFE、XGBoost)识别出APOE相关的五基因特征:XAF1、CEP128、TMEM106A、VEPH1和ABCB1。
- 该特征在GSE43292中表观AUC为0.956,GSE100927中AUC为0.974,GSE28829中AUC为0.736;高特征评分与免疫炎症变化相关。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
ABSTRACT Aims Ketone bodies are implicated in cardiovascular metabolism, yet the contributions of β-hydroxybutyrate (BHB), acetoacetate and acetone to coronary atherosclerosis remain unclear. Methods We integrated Mendelian randomization (MR), cross-trait GWAS analysis, ApoE pQTL-MR and colocalization to investigate genetic associations with coronary atherosclerosis, complemented by bulk transcriptomic analysis, machine learning-based signature construction, immune infiltration analysis and single-cell analysis to characterize APOE-associated molecular features in human atherosclerotic plaques. Results Genetically predicted BHB was associated with coronary atherosclerosis risk in univariable MR (OR = 1.59, 95% CI: 1.29–1.95, p = 8.8 × 10−6) and remained significant after mutual adjustment for acetoacetate and acetone (OR = 1.58, 95% CI: 1.14–2.18, p = 0.005), whereas neither acetoacetate nor acetone showed a significant association. Cross-trait gene prioritization highlighted APOE for further investigation. ApoE pQTL-MR supported an association between genetically predicted ApoE and coronary atherosclerosis (OR = 1.23, p = 0.003). Weighted gene co-expression network analysis (WGCNA) and machine-learning algorithms, including LASSO, random forest, SVM-RFE and XGBoost, identified an APOE-associated five-gene signature comprising XAF1, CEP128, TMEM106A, VEPH1 and ABCB1. The signature achieved an apparent AUC of 0.956 in GSE43292 and AUCs of 0.974 in GSE100927 and 0.736 in GSE28829. High signature scores were associated with immune-inflammatory changes. Single-cell analysis showed enrichment of the signature in macrophages, endothelial cells and fibroblasts, while CellChat analysis identified a macrophage-centred pattern of inferred cell–cell communication. Conclusions These findings identify BHB as the ketone body most strongly associated with coronary atherosclerosis susceptibility and define an APOE-associated multicellular signature in human atherosclerotic plaques with potential value for molecular stratification.
引用
- Diabetes, Obesity and Metabolism:Genetically Predicted β‐Hydroxybutyrate and Coronary Atherosclerosis: Mendelian Randomization and APOE‐Associated Transcriptomic Signatures in Human Atherosclerotic Plaques
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。