Zhou Wenchao, Li Hui, Zhang Juan, Liu Changsheng, Liu Dan, Chen Xupeng, Ouyang Jing, Zeng Tian, Peng Shuang, Ouyang Fan, Long Yunzhu, Li Yukun
Department of Assisted Reproductive Centre, Xiangya Hospital Zhuzhou Central South University, Central South University, Zhuzhou, China.
Department of Gynecology, The Second Affiliated Hospital, Hunan Province Key Laboratory of Tumor Cellular & Molecular Pathology, Cancer Research Institute, Hengyang Medical School, University of South China, Hengyang, China.
Ann Med. 2025 Dec;57(1):2477301. doi: 10.1080/07853890.2025.2477301. Epub 2025 Mar 12.
Butyrate may inhibit SARS-CoV-2 replication and affect the development of COVID-19. However, there have been no systematic comprehensive analyses of the role of butyrate metabolism-related genes (BMRGs) in COVID-19.
We performed differential expression analysis of BMRGs in the brain, liver and pancreas of COVID-19 patients and controls in GSE157852 and GSE151803. The differentially expressed genes (DEGs) and module genes between COVID-19 patients and healthy controls in GSE171110 were screened through 'limma' and 'WGANA' R package, respectively, followed by an intersection with BMRGs via 'ggvenn' R package. Six machine learning algorithms were employed to determine the best model for identifying biomarkers, and receiver operating characteristic (ROC) curves were plotted to evaluate the diagnostic value of the biomarkers in COVID-19. Moreover, the differences in immune-infiltrating cells between the COVID-19 and control groups were compared using CIBERSORT. The differences in immune cells and expression levels of biomarkers in immune cells among different tissues were analysed using GSE171668.
The BMRGs were the most different in the brain between the COVID-19 and control groups, including 21 upregulated and 16 downregulated genes. Five important common BMRGs were screened as biomarkers for COVID-19 using XGBoost, namely CCNB1, CCNA2, BRCA1, HBB and HSPA5, with increased diagnostic performance. Enrichment analysis revealed that these five genes were related to the cell cycle, cell proliferation and cell senescence. The infiltrating abundance of 12 immune cells was different between the COVID-19 and control groups. Finally, the expression levels of HSPA5, BRCA1 and HBB were higher in annotated cells than in CCNB1 and CCNA2, and there were four different types of immune cells in the liver, heart, lungs and kidneys.
These five genes may be potential biomarkers of butyrate metabolism in COVID-19 patients. These findings provide a direction for further studies on the molecular mechanisms underlying COVID-19.
丁酸盐可能抑制新型冠状病毒（SARS-CoV-2）复制并影响2019冠状病毒病（COVID-19）的发展。然而，尚未对丁酸盐代谢相关基因（BMRGs）在COVID-19中的作用进行系统全面的分析。
我们在GSE157852和GSE151803中对COVID-19患者和对照的脑、肝和胰腺中的BMRGs进行差异表达分析。分别通过“limma”和“WGANA”R包筛选GSE171110中COVID-19患者和健康对照之间的差异表达基因（DEGs）和模块基因，随后通过“ggvenn”R包与BMRGs进行交集分析。采用六种机器学习算法确定识别生物标志物的最佳模型，并绘制受试者工作特征（ROC）曲线以评估生物标志物在COVID-19中的诊断价值。此外，使用CIBERSORT比较COVID-19组和对照组之间免疫浸润细胞的差异。使用GSE171668分析不同组织中免疫细胞和免疫细胞中生物标志物表达水平的差异。
COVID-19组和对照组之间，BMRGs在脑中差异最大，包括21个上调基因和16个下调基因。使用XGBoost筛选出五个重要的常见BMRGs作为COVID-19的生物标志物，即细胞周期蛋白B1（CCNB1）、细胞周期蛋白A2（CCNA2）、乳腺癌1号基因（BRCA1）、血红蛋白β（HBB）和热休克蛋白家族A（Hsp70）成员5（HSPA5），其诊断性能有所提高。富集分析表明，这五个基因与细胞周期、细胞增殖和细胞衰老有关。COVID-19组和对照组之间12种免疫细胞的浸润丰度不同。最后，在注释细胞中，HSPA5、BRCA1和HBB的表达水平高于CCNB1和CCNA2，并且在肝、心、肺和肾中有四种不同类型的免疫细胞。
这五个基因可能是COVID-19患者丁酸盐代谢的潜在生物标志物。这些发现为进一步研究COVID-19的分子机制提供了方向。
