加权基因共表达网络分析全氟和多氟烷基化合物对人间充质干细胞的毒性靶点

More than 3,000 per- and polyfluoroalkyl substances (PFASs) have been put into the global market, and the environmental effects and health risks of a large number of new substitutes remain unknown, which has brought great challenges to the supervision of PFASs. In order to reveal the biological processes and sensitive target genes interfered by PFASs and new substitutes, human bone mesenchymal stem cells (hBMSCs) were exposed to chlorinated polyfluoroalkyl ether sulfonates (Cl - PFESAs), perfluorooctane sulfonate (PFOS), perfluorohexane sulfonate (PFHxS) and perfluorooctanoic acid (PFOA), and the changes in the cell gene expression profiles were compared by using the weighted gene co - expression network analysis (WGCNA) method. WGCNA calculates the degree of association among 5,004 genes using topological overlap and divides them into 14 gene modules distinguished by colors in combination with the dynamic cut tree method, and the genes within each module except the gray module are highly correlated. The analysis of the correlation between gene modules and cell samples found that the gene expression profiles of the control group cells were positively correlated with the Blue and Greenyellow modules, the PFOS - exposed group was significantly negatively correlated with the Blue module, and the Cl - PFESAs - exposed group was significantly negatively correlated with the Greenyellow module, indicating that PFASs exposure caused significant changes in the gene expression patterns of the Blue and Greenyellow modules in the cells. The biological processes significantly enriched in the Blue and Greenyellow modules were mainly cholesterol biosynthesis and negative regulation of bone marrow leukocyte differentiation, suggesting that immune regulation and lipid metabolism may be the common potential targets of various PFASs acting on hBMSCs. The hub genes were screened according to the connectivity of the module gene co - expression network, and it was found that the DHCR24, SQLE and EBP genes related to lipid metabolism may be the sensitive target genes for PFASs to affect lipid metabolism. Further using the hBMSCs in vitro model to study the related toxic effects and mechanisms of PFASs is helpful to establish biomarkers for the toxicity prediction and screening of such chemicals and to provide scientific basis and new methods for their safety evaluation and supervision.

超过3 000种全氟多氟化合物(per-and polyfluoroalkyl substances, PFASs)已被投入全球市场,大量新型替代品的环境效应与健康风险仍然未知,为PFASs的监管带来了极大的挑战。为揭示PFASs及新型替代品干扰的生物学过程和敏感靶基因,将人骨髓间充质干细胞(human bone mesenchymal stem cells, hBMSCs)暴露于氯代多氟醚基磺酸(chlorinated polyfluoroalkyl ether sulfonates, Cl-PFESAs)、全氟辛烷磺酸(perfluorooctane sulfonate, PFOS)、全氟己烷磺酸(perfluorohexane sulfonate, PFHxS)和全氟辛酸(perfluorooctanoic acid, PFOA),利用加权基因共表达网络分析(weighted gene co-expression network analysis, WGCNA)方法比较细胞基因表达谱的变化。WGCNA使用拓扑重叠计算5 004个基因间的关联程度,并结合动态剪切树法将其划分为14个通过颜色区分的基因模块,除灰色模块外各模块内的基因高度相关。基因模块与细胞样本相关性分析发现,对照组细胞基因表达谱与Blue和Greenyellow模块正相关,PFOS暴露组与Blue模块显著负相关,Cl-PFESAs暴露组与Greenyellow模块显著负相关,表明PFASs暴露使细胞中Blue和Greenyellow模块基因表达模式发生较大变化。Blue和Greenyellow模块显著富集的生物学过程主要为胆固醇生物合成和骨髓白细胞分化负调控,提示免疫调控和脂质代谢可能是多种PFASs作用于hBMSCs的共同潜在靶点。根据模块基因共表达网络的连通性筛选枢纽基因,发现与脂质代谢相关的DHCR24、SQLE和EBP基因可能是PFASs影响脂质代谢的敏感作用靶基因。进一步利用hBMSCs体外模型研究PFASs的相关毒性作用和机制,有助于建立该类化学物毒性预测和筛选的生物标志物,为其安全性评价和监管提供科学依据和新方法。