WGCNA reveals a biomarker for cancer-associated fibroblasts to predict prognosis in cervical cancer

• Published in: Journal of the Chinese Medical Association Vol. 87; no. 9; pp. 885 - 897
• Main Authors: Liu, Zao-Ling, Chen, Nan, Li, Rong, Ma, Ying-Jie, Qiayimaerdan, Aerna, Ma, Cai-Ling
• Format: Journal Article
• Language: English
• Published: Netherlands 01.09.2024
• Subjects: Biomarkers, Tumor - genetics; Cancer-Associated Fibroblasts; Female; Humans; Prognosis; Tumor Microenvironment; Uterine Cervical Neoplasms - genetics; Uterine Cervical Neoplasms - mortality; Uterine Cervical Neoplasms - pathology
• ISSN: 1726-4901; 1728-7731; 1728-7731; 1726-4901
• DOI: 10.1097/JCMA.0000000000001129

Cancer-associated fibroblasts (CAFs) are crucial components of the cervical cancer tumor microenvironment, playing a significant role in cervical cancer progression, treatment resistance, and immune evasion, but whether the expression of CAF-related genes can predict clinical outcomes in cervical cancer is still unknown. In this study, we sought to analyze genes associated with CAFs through weighted gene co-expression network analysis (WGCNA) and to create a predictive model for CAFs in cervical cancer. We acquired transcriptome sequencing data and clinical information on cervical cancer patients from the cancer genome atlas (TCGA) and gene expression omnibus (GEO) databases. WGCNA was conducted to identify genes related to CAFs. We developed a prognostic model based on CAF genes in cervical cancer using the least absolute shrinkage and selection operator (LASSO) Cox regression analysis. Single-cell sequencing data analysis and in vivo experiments for validation of hub genes in CAFs. A prognostic model for cervical cancer was developed based on CAF genes including COL4A1 , LAMC1 , RAMP3 , POSTN , and SERPINF1 . Cervical cancer patients were divided into low- and high-risk groups based on the optimal cutoff value. Patients in the high-risk group had a significantly worse prognosis. Single-cell RNA sequencing data revealed that hub genes in the CAFs risk model were expressed mainly in fibroblasts. The real-time fluorescence quantitative polymerase chain reaction (PCR) results revealed a significant difference in the expression levels of COL4A1 , LAMC1 , POSTN , and SERPINF1 between the cancer group and the normal group ( p < 0.05). Consistently, the results of the immunohistochemical tests exhibited notable variations in COL4A1, LAMC1, RAMP3, POSTN, and SERPINF1 expression between the cancer and normal groups ( p < 0.001). The CAF risk model for cervical cancer constructed in this study can be used to predict prognosis, while the CAF hub genes can be utilized as crucial markers for cervical cancer prognosis.