Scholarlink

<p>The treatment of sepsis is challenging due to unclear mechanisms. Propionate is increasingly seen as critical to sepsis pathophysiology by bridging gut microbiota and immunity, but the mechanisms remain unclear. Our study analysed differences in propionate metabolism in peripheral blood mononuclear cells from septic patients and healthy controls using single-cell RNA-seq (scRNA-seq) data. Differentially expressed genes (DEGs) analysis, pathway enrichment, transcription factor (TF) prediction, intercellular communication, and trajectory inference were used to explore the role of propionate metabolism in sepsis. We constructed a sepsis diagnostic model using LASSO and machine learning (XGBoost, CatBoost, NGBoost) with bulk RNA-seq data. scRNA-seq analysis revealed that propionate metabolism was highest in plasma cells (PCs), which can be classified into high and low metabolism groups, identifying 9,155 DEGs. High propionate metabolism was associated with metabolism such as short-chain fatty acids, while low metabolism was related to negative regulation of wound healing. The DoRothEA regulator algorithm showed TFs such as IRF4, ARID3A, FOXO4, and ATF2 were activated in high propionate metabolism subgroups, whereas NR5A1, BCL6, and CDX2 were activated in low subgroups. Cell-cell communication revealed that both groups interacted primarily with B cells and neutrophils, with the high propionate metabolism PCs showing more significant interactions. The receptor-ligand pairs primarily involved were VEGFA-FLT1 and VEGFB-FLT1, and the high propionate metabolism PCs and B cells might interact through BMP8B-BMPR2. Trajectory analysis indicated differentiation from B cells, first to low, then high propionate metabolism PCs. Finally, the LASSO algorithm identified 13 key genes, with the CatBoost model achieving perfect diagnostic performance (AUC = 1.000). These 13 key genes were validated through in vitro experiments. Collectively, these findings suggest that propionic acid metabolism may be a potential target for diagnosing and treating sepsis, offering new insights into its pathophysiology.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Identifying propionate metabolism-related genes as biomarkers of sepsis development and therapeutic targets

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月8日

查看全文
  • Lechen Yang,
  • Weifeng Shang,
  • Dongjie Chen,
  • Hang Qian,
  • Sheng Zhang,
  • Xiaojun Pan,
  • Sisi Huang,
  • Jiao Liu,
  • Dechang Chen

摘要

The treatment of sepsis is challenging due to unclear mechanisms. Propionate is increasingly seen as critical to sepsis pathophysiology by bridging gut microbiota and immunity, but the mechanisms remain unclear. Our study analysed differences in propionate metabolism in peripheral blood mononuclear cells from septic patients and healthy controls using single-cell RNA-seq (scRNA-seq) data. Differentially expressed genes (DEGs) analysis, pathway enrichment, transcription factor (TF) prediction, intercellular communication, and trajectory inference were used to explore the role of propionate metabolism in sepsis. We constructed a sepsis diagnostic model using LASSO and machine learning (XGBoost, CatBoost, NGBoost) with bulk RNA-seq data. scRNA-seq analysis revealed that propionate metabolism was highest in plasma cells (PCs), which can be classified into high and low metabolism groups, identifying 9,155 DEGs. High propionate metabolism was associated with metabolism such as short-chain fatty acids, while low metabolism was related to negative regulation of wound healing. The DoRothEA regulator algorithm showed TFs such as IRF4, ARID3A, FOXO4, and ATF2 were activated in high propionate metabolism subgroups, whereas NR5A1, BCL6, and CDX2 were activated in low subgroups. Cell-cell communication revealed that both groups interacted primarily with B cells and neutrophils, with the high propionate metabolism PCs showing more significant interactions. The receptor-ligand pairs primarily involved were VEGFA-FLT1 and VEGFB-FLT1, and the high propionate metabolism PCs and B cells might interact through BMP8B-BMPR2. Trajectory analysis indicated differentiation from B cells, first to low, then high propionate metabolism PCs. Finally, the LASSO algorithm identified 13 key genes, with the CatBoost model achieving perfect diagnostic performance (AUC = 1.000). These 13 key genes were validated through in vitro experiments. Collectively, these findings suggest that propionic acid metabolism may be a potential target for diagnosing and treating sepsis, offering new insights into its pathophysiology.