Table_2_Construction of a novel choline metabolism-related signature to predict prognosis, immune landscape, and chemotherapy response in colon adenocarcinoma.xlsx

BackgroundColon adenocarcinoma (COAD) is a common digestive system malignancy with high mortality and poor prognosis. Accumulating evidence indicates that choline metabolism is closely related to tumorigenesis and development. However, the efficacy of choline metabolism-related signature in predicting patient prognosis, immune microenvironment and chemotherapy response has not been fully clarified.
MethodsCholine metabolism-related differentially expressed genes (DEGs) between normal and COAD tissues were screened using datasets from The Cancer Genome Atlas (TCGA), Kyoto Encyclopedia of Genes and Genomes (KEGG), AmiGO2 and Reactome Pathway databases. Two choline metabolism-related genes (CHKB and PEMT) were identified by univariate and multivariate Cox regression analyses. TCGA-COAD was the training cohort, and GSE17536 was the validation cohort. Patients in the high- and low-risk groups were distinguished according to the optimal cutoff value of the risk score. A nomogram was used to assess the prognostic accuracy of the choline metabolism-related signature. Calibration curves, decision curve analysis (DCA), and clinical impact curve (CIC) were used to improve the clinical applicability of the prognostic signature. Gene Ontology (GO) and KEGG pathway enrichment analyses of DEGs in the high- and low-risk groups were performed. KEGG cluster analysis was conducted by the KOBAS-i database. The distribution and expression of CHKB and PEMT in various types of immune cells were analyzed based on single-cell RNA sequencing (scRNA-seq). The CIBERSORT and ESTIMATE algorithms evaluated tumor immune cell infiltration in the high- and low-risk groups. Evaluation of the half maximal inhibitory concentration (IC50) of common chemotherapeutic drugs based on the choline metabolism-related signature was performed. Small molecule compounds were predicted using the Connectivity Map (CMap) database. Molecular docking is used to simulate the binding conformation of small molecule compounds and key targets. By immunohistochemistry (IHC), Western blot, quantitative reverse transcription-polymerase chain reaction (qRT-PCR) experiments, the expression levels of CHKB and PEMT in human, mouse, and cell lines were detected.
ResultsWe constructed and validated a choline metabolism-related signature containing two genes (CHKB and PEMT). The overall survival (OS) of patients in the high-risk group was significantly worse than that of patients in the low-risk group. The nomogram could effectively and accurately predict the OS of COAD patients at 1, 3, and 5 years. The DCA curve and CIC demonstrate the clinical utility of the nomogram. scRNA-seq showed that CHKB was mainly distributed in endothelial cells, while PEMT was mainly distributed in CD4+ T cells and CD8+ T cells. In addition, multiple types of immune cells expressing CHKB and PEMT differed significantly. There were significant differences in the immune microenvironment, immune checkpoint expression and chemotherapy response between the two risk groups. In addition, we screened five potential small molecule drugs that targeted treatment for COAD. Finally, the results of IHC, Western blot, and qRT-PCR consistently showed that the expression of CHKB in human, mouse, and cell lines was elevated in normal samples, while PMET showed the opposite trend.
ConclusionIn conclusion, we constructed a choline metabolism-related signature in COAD and revealed its potential application value in predicting the prognosis, immune microenvironment, and chemotherapy response of patients, which may lay an important theoretical basis for future personalized precision therapy.

背景：结肠腺癌（Colon adenocarcinoma, COAD）是一类常见的消化系统恶性肿瘤，具有高致死率与不良预后。日益积累的研究证据表明，胆碱代谢与肿瘤的发生发展密切相关。然而，胆碱代谢相关基因特征在预测患者预后、免疫微环境及化疗响应方面的效能尚未完全阐明。
方法：通过从癌症基因组图谱（The Cancer Genome Atlas, TCGA）、京都基因与基因组百科全书（Kyoto Encyclopedia of Genes and Genomes, KEGG）、AmiGO2及Reactome通路数据库中获取数据集，筛选正常组织与结肠腺癌组织间的胆碱代谢相关差异表达基因（differentially expressed genes, DEGs）。通过单变量与多变量Cox回归分析，鉴定得到2个胆碱代谢相关基因：CHKB与PEMT。以TCGA-COAD数据集作为训练队列，GSE17536数据集作为验证队列。根据风险评分的最优截断值，将患者划分为高风险组与低风险组。采用列线图（nomogram）评估该胆碱代谢相关基因特征的预后精准度。通过校准曲线、决策曲线分析（decision curve analysis, DCA）及临床影响曲线（clinical impact curve, CIC）提升该预后特征的临床实用性。对高、低风险组中的差异表达基因进行基因本体（Gene Ontology, GO）与KEGG通路富集分析，并通过KOBAS-i数据库完成KEGG聚类分析。基于单细胞RNA测序（single-cell RNA sequencing, scRNA-seq）数据，分析CHKB与PEMT在各类免疫细胞中的分布与表达情况。采用CIBERSORT与ESTIMATE算法评估高、低风险组的肿瘤免疫细胞浸润情况。基于该胆碱代谢相关基因特征，对常见化疗药物的半最大效应浓度（half maximal inhibitory concentration, IC50）进行评估。通过连通图（Connectivity Map, CMap）数据库预测潜在小分子化合物，并采用分子对接技术模拟小分子化合物与关键靶点的结合构象。通过免疫组织化学（immunohistochemistry, IHC）、蛋白质印迹（Western blot）、定量反转录聚合酶链反应（quantitative reverse transcription-polymerase chain reaction, qRT-PCR）实验，检测CHKB与PEMT在人类、小鼠组织及细胞系中的表达水平。
结果：本研究构建并验证了一种包含CHKB与PEMT两个基因的胆碱代谢相关基因特征。高风险组患者的总生存期（overall survival, OS）显著差于低风险组患者。该列线图可有效且精准地预测结肠腺癌患者1年、3年及5年的总生存期。决策曲线与临床影响曲线证实了该列线图的临床应用价值。单细胞RNA测序结果显示，CHKB主要分布于内皮细胞，而PEMT主要分布于CD4+ T细胞与CD8+ T细胞。此外，表达CHKB与PEMT的多种免疫细胞群体存在显著差异。两组风险患者的免疫微环境、免疫检查点表达水平以及化疗响应均存在显著差异。此外，本研究筛选得到5种可用于结肠腺癌靶向治疗的潜在小分子药物。最后，免疫组织化学、蛋白质印迹及qRT-PCR的实验结果一致显示：在人类、小鼠组织及细胞系中，正常样本内CHKB的表达水平升高，而PEMT的表达趋势则与之相反。
结论：综上，本研究构建了结肠腺癌中胆碱代谢相关的基因特征，并揭示了其在预测患者预后、免疫微环境与化疗响应方面的潜在应用价值，可为未来的个性化精准治疗奠定重要的理论基础。