Construction and Evaluation of a Risk Score Model for Lymph Node Metastasis-Associated Circadian Clock Genes in Esophageal Squamous Carcinoma

Abstract

Background: Studies suggested that circadian clock genes (CCGs) in human esophageal squamous carcinoma (ESCC) samples are dysregulated. However, the relevance of CCGs to lymph node metastasis (LNM) and prognosis of ESCC remains unclear. Methods: The differentially expressed genes (DEGs) between normal and ESCC samples in The Cancer Genome Atlas database (TCGA) database were intersected with the genes associated with LNM (LNMGs) in ESCC samples and 300 CCGs to obtain the differentially expressed LNM-associated CCGs (DE-LNM-CCGs). The risk model was constructed by Cox regression analysis in the TCGA-ESCC training set, and the accuracy of the risk model was verified by risk profile and overall survival profile. Furthermore, differences of 23 immune cells, 13 immune functions, and immune checkpoint molecules between the high- and low-risk groups were assessed using the single-sample gene set enrichment analysis (ssGSEA) algorithm. Gene set enrichment analysis (GSEA) was conducted to investigate the functional differences between low- and high-risk groups. Finally, we validated the mRNA expression levels of prognostic model genes by quantitative real-time polymerase chain reaction (qRT-PCR). Results: A total of six DE-LNM-CCGs were identified in TCGA-ESCC. TP53 and NAGLU were selected by Cox regression analysis to construct the risk model. Risk profile plots, overall survival plots, and validation results of the risk model in the validation set indicated that the constructed risk model was reliable. The result of ssGSEA showed that the percentages of activated B cells, activated dendritic cells, effector memory CD8 T cells, immune function in neutrophils, plasmacytoid dendritic cells, T cell co-inhibition, and Type 17 T helper cells were different between the high- and low-risk groups. In addition, the expression of CD274, PDCD1, TNFRSF18, and TNFRSF9 was dysregulated between the high- and low-risk groups. GSEA revealed that the high-risk group was associated with cell differentiation, oxidative phosphorylation, and steroid biosynthesis pathways, while the low-risk group was associated with chromosome, ECM–receptor interaction, and other pathways. Finally, qRT-PCR results showed that the mRNA expression levels of two prognostic genes were consistent with TCGA. Conclusion: In conclusion, the risk model constructed based on TP53 and NAGLU could accurately predict the prognosis.