Single-cell combined transcriptome probes prognostic mechanisms of sialylation-related genes in cervical cancer

Introduction: Sialylation has been linked to cervical dysplasia, while its involvement in cervical cancer is uncertain. Hence , the aim of this study was to develop a prognostic model based on sialylation-related characteristics for cervical cancer patients and investigate how sialylation-related genes are altered in cervical cancer via analyses of transcriptome and single-cell RNA sequencing (scRNA-seq) data. Methods: The current work incorporated 4 transcriptome datasets relevant to cervical cancer (including scRNA-seq) and 110 sialylation-related genes (SRGs). Initially, differentially expressed SRGs (DE-SRGs) were discovered by differential expression analysis, among other methods. Subsequently, least absolute shrinkage and selection operator (LASSO) and Cox regression analysis was applied using DE-SRGs to detect prognostic genes and build prognostic models. Next, independent prognosis test was conducted, and a nomogram model was built using clinical characteristics and risk scores. Meanwhile, scRNA-seq was applied to examine the cellular composition and cell-to-cell regulation in cervical cancer vs normal group, and key cells were determined via prognostic genes and their differentiation process was investigated. Finally, the immunological microenvironment, mutant genes, and medication sensitivity were assessed. Clinical samples were taken to assess the expression of prognostic genes by quantitative reverse transcriptase PCR (qRT-PCR). Results: First, we detected 19 DE-SRGs related with sialylation. Three prognostic genes, GALNT12, GCNT4, and NPL, were discovered by LASSO cox regression. A risk model constructed with prognostic genes revealed that patients in high-risk group had a much poorer survival rate than those in group with low risk. Meanwhile, low-risk cervical cancer patients were more likely to respond to immunotherapy and chemotherapy, depending on immunology, tumor microenvironment, and drug sensitivity. ScRNA-seq data suggests that the expression of prognostic genes was higher in key cells, macrophages and fibroblasts, and played a more critical role in cervical cancer. The findings from qRT-PCR demonstrated that GCNT4 and NPL were considerably overexpressed in the cervical cancer group. Discussion: In this research, GALNT12, GCNT4 and NPL were discovered as sialylation-related prognostic genes in cervical cancer, providing novel pathways for detection and treatment.