Fu Kang, Su Junzhe, Zhou Yiming, Chen Xiaotong, Hu Xiao
Department of Hepatobiliary Pancreatic Surgery, The Affiliated Hospital of Qingdao University, Qingdao, China.
Front Pharmacol. 2024 Nov 21;15:1498031. doi: 10.3389/fphar.2024.1498031. eCollection 2024.
Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal malignancy with poor prognosis. Epigenetic dysregulation plays a crucial role in PDAC progression, but its comprehensive landscape and clinical implications remain unclear.
We integrated single-cell RNA sequencing, bulk RNA sequencing, and clinical data from multiple public databases. Single-cell analysis was performed using Seurat and hdWGCNA packages to reveal cell heterogeneity and epigenetic features. Weighted gene co-expression network analysis (WGCNA) identified key epigenetic modules. A machine learning-based prognostic model was constructed using multiple algorithms, including Lasso and Random Survival Forest. We further analyzed mutations, immune microenvironment, and drug sensitivity associated with the epigenetic risk score.
Single-cell analysis revealed distinct epigenetic patterns across different cell types in PDAC. WGCNA identified key modules associated with histone modifications and DNA methylation. Our machine learning model, based on 17 epigenetic genes, showed robust prognostic value (AUC >0.7 for 1-, 3-, and 5-year survival) and outperformed existing models. High-risk patients exhibited distinct mutation patterns, including higher frequencies of KRAS and TP53 mutations. Low-risk patients showed higher immune and stromal scores, with increased infiltration of CD8 T cells and M2 macrophages. Drug sensitivity analysis revealed differential responses to various therapeutic agents between high- and low-risk groups, with low-risk patients showing higher sensitivity to EGFR and MEK inhibitors.
Our study provides a comprehensive landscape of epigenetic regulation in PDAC at single-cell resolution and establishes a robust epigenetics-based prognostic model. The integration of epigenetic features with mutation profiles, immune microenvironment, and drug sensitivity offers new insights into PDAC heterogeneity and potential therapeutic strategies. These findings pave the way for personalized medicine in PDAC management and highlight the importance of epigenetic regulation in cancer research.
胰腺导管腺癌(PDAC)是一种预后很差的高致死性恶性肿瘤。表观遗传失调在PDAC进展中起关键作用,但其全面情况及临床意义仍不清楚。
我们整合了来自多个公共数据库的单细胞RNA测序、批量RNA测序及临床数据。使用Seurat和hdWGCNA软件包进行单细胞分析以揭示细胞异质性和表观遗传特征。加权基因共表达网络分析(WGCNA)确定了关键的表观遗传模块。使用包括套索法和随机生存森林法在内的多种算法构建了基于机器学习的预后模型。我们进一步分析了与表观遗传风险评分相关的突变、免疫微环境和药物敏感性。
单细胞分析揭示了PDAC中不同细胞类型的独特表观遗传模式。WGCNA确定了与组蛋白修饰和DNA甲基化相关的关键模块。我们基于17个表观遗传基因的机器学习模型显示出强大的预后价值(1年、3年和5年生存率的AUC>0.7),且优于现有模型。高风险患者表现出独特的突变模式,包括KRAS和TP53突变的频率更高。低风险患者显示出更高的免疫和基质评分,CD8 T细胞和M2巨噬细胞浸润增加。药物敏感性分析显示高风险组和低风险组对各种治疗药物的反应不同,低风险患者对EGFR和MEK抑制剂表现出更高的敏感性。
我们的研究以单细胞分辨率提供了PDAC表观遗传调控的全面情况,并建立了一个强大的基于表观遗传学的预后模型。将表观遗传特征与突变谱、免疫微环境和药物敏感性相结合,为PDAC异质性和潜在治疗策略提供了新见解。这些发现为PDAC治疗中的个性化医疗铺平了道路,并突出了表观遗传调控在癌症研究中的重要性。