Comprehensive Pan-Cancer Analysis and the Regulatory Mechanism of ASF1B, a Gene Associated With Thyroid Cancer Prognosis in the Tumor Micro-Environment

Abstract

Background: The incidence of thyroid cancer, whose local recurrence and metastasis lead to death, has always been high and the pathogenesis of papillary thyroid carcinoma (PTC) has not been clearly elucidated. Therefore, the research for more accurate prognosis-related predictive biomarkers is imminent, and a key gene can often be a prognostic marker for multiple tumors.

Methods: Gene expression profiles of various cancers in the TCGA and GTEx databases were downloaded, and genes significantly associated with the prognosis of THCA were identified by combining differential analysis with survival analysis. Then, a series of bioinformatics tools and methods were used to analyze the expression of the gene in each cancer and the correlation of each expression with prognosis, tumor immune microenvironment, immune neoantigens, immune checkpoints, DNA repair genes, and methyltransferases respectively. The possible biological mechanisms were also investigated by GSEA enrichment analysis.

Results: 656 differentially expressed genes were identified from two datasets and 960 DEGs that were associated with disease-free survival in THCA patients were screened via survival analysis. The former and the latter were crossed to obtain 7 key genes, and the gene with the highest risk factor, ASF1B, was selected for this study. Differential analysis of multiple databases showed that ASF1B was commonly and highly expressed in pan-cancer. Survival analysis showed that high ASF1B expression was significantly associated with poor patient prognosis in multiple cancers. In addition, ASF1B expression levels were found to be associated with tumor immune infiltration in THCA, KIRC, LGG, and LIHC, and with tumor microenvironment in BRCA, LUSC, STAD, UCEC, and KIRC. Further analysis of the relationship between ASF1B expression and immune checker gene expression suggested that ASF1B may regulate tumor immune patterns in most tumors by regulating the expression levels of specific immune checker genes. Finally, GSEA enrichment analysis showed that ASF1B high expression was mainly enriched in cell cycle, MTORC1 signaling system, E2F targets, and G2M checkpoints pathways.

Conclusions: ASF1B may be an independent prognostic marker for predicting the prognosis of THCA patients. The pan-cancer analysis suggested that ASF1B may play an important role in the tumor micro-environment and tumor immunity and it has the potential of serving as a predictive biomarker for multiple cancers.

Keywords: anti-silencing function protein 1 homolog B (ASF1B); pan-cancer; prognosis; thyroid cancer; tumor immune micro-environment.

Figure 1

Differentially expressed genes and functional enrichment analysis; (A) volcano plot: grey dots indicate significantly differentially up-regulated genes and orange dots indicate significantly differentially down-regulated genes; (B) Heatmap plot of the DEGs; (C) KEGG enrichment analysis; (D) GO enrichment analysis

Figure 2

Overlap of differentially expressed genes and prognostic genes; (A) top 20 genes associated with DFS (disease-free survival) are shown; (B) overlap of DEGs and prognostic genes; (C) DFS survival curve of ASF1B; (D) DFS survival curve of SEZ6L2; (E) DFS survival curve of GALNT15; (F) DFS survival curve of ITGA2.

Figure 3

Expression of ASF1B in tumors; (A) Expression level of ASF1B in THCA; (B) Expression level of ASF1B in different TNM stages of THCA; (C) Expression level of ASF1B in 27 cancer type. In (A) *** is P < 0.0001.

Figure 4

Univariate survival analysis was used to analyze the relationship between ASF1B expression and survival time in 33 tumors; (A) forest plot showing the relationship between ASF1B expression and OS; (B) forest plot showing the relationship between ASF1B expression and DFS; (C) KM curves of high and low ASF1B expression in 16 tumors significantly associated with OS survival; (D) KM curves of high and low ASF1B expression in 7 tumors significantly associated with DFS.

Figure 5

Correlation of ASF1B with tumor immune infiltration and the tumor microenvironment in THCA and other cancers; (A) Correlation of ASF1B expression with immune cell infiltration in THCA; (B) Correlation of ASF1B expression with immune cell infiltration in KIRC; (C) Correlation of ASF1B expression with immune cell infiltration in LGG; (D) Correlation of ASF1B expression with immune cell infiltration in LIHC; (E) Correlation of ASF1B with the immune score, stromal score, and ESTIMATE score in pan-cancer.

Figure 6

Correlation analysis of ASF1B expression in pan-cancer with immune neoantigens and immune checkpoint genes; (A) Correlation analysis of ASF1B expression in pan-cancer with immune checkpoint gene expression; (B) Correlation analysis of ASF1B expression in 19 tumors with the number of tumor neoantigens. In (A) * is P < 0.05, ** is P < 0.01 and *** is P < 0.001.

Figure 7

ASF1B expression in relation to DNA repair genes and methyltransferases; (A) correlation between ASF1B expression and gene expression levels of five MMRs; (B) correlation between ASF1B expression and expression of four methyltransferases; red: DNMT1, blue: DNMT2, green: DNMT3A, purple: DNMT3B. In (A) * is P < 0.05,** is P < 0.01, *** is P < 0.001.

Figure 8

GSEA analysis of ASF1B; (A) enrichment analysis of ASF1B in KEGG signaling pathway; (B) enrichment analysis of ASF1B in HALLMARK signaling pathway.