A Prognostic Autophagy-Related Gene Pair Signature and Small-Molecule Drugs for Hepatocellular Carcinoma.

Zebing Song,Shaoqiang Li,Guopei Zhang,Yang Yu

doi:10.3389/fgene.2021.689801

Abstract

Dysregulation of autophagy-related genes (ARGs) is related to the prognosis of cancers. However, the aberrant expression of ARGs signature in the prognosis of hepatocellular carcinoma (HCC) remain unclear. Using The Cancer Genome Atlas and the International Cancer Genome Consortium database, 188 common autophagy-related gene pairs (ARGPs) were identified. Through univariate, least absolute shrinkage and selection operator analysis, and multivariate Cox regression analysis, a prognostic signature of the training set was constructed on the basis of 6 ARGPs. Further analysis revealed that the ARGP based signature performed more accurately in overall survival (OS) prediction compared to other published gene signatures. In addition, a high risk of HCC was closely related to CTLA4 upregulation, LC3 downregulation, low-response to axitinib, rapamycin, temsirolimus, docetaxel, metformin, and high-response to bleomycin. Univariate Cox and multivariate Cox analysis revealed that the risk score was an independent prognostic factor for HCC. These results were internally validated in the test and TCGA sets and externally validated in the ICGC set. A nomogram, consisting of the risk score and the TNM stage, performed well when compared to an ideal nomogram. In conclusion, a 6-ARGP-based prognostic signature was identified and validated as an effective predictor of OS of patients with HCC. Furthermore, we recognized six small-molecule drugs, which may be potentially effective in treating HCC.

Highlights

Hepatocellular carcinoma (HCC) is one of the most common liver malignancies worldwide, with increasing rates of morbidity and mortality annually (Bray et al, 2018)
Another study revealed that the activation of autophagy can promote metastasis through the upregulation of MCT1 via activating Wnt/β-catenin signaling in hepatocellular carcinoma (HCC) cells (Fan et al, 2018)
Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis revealed that the primary pathways of these autophagy-related genes (ARGs) were “autophagy-animal,” “IL-17 signaling pathway,” “PI3K-Akt signaling pathway,” and “mTOR signaling pathway,” which were primarily correlated with autophagy, immune process, and carcinogenicity (Figures 2E,F)

Summary

INTRODUCTION

Hepatocellular carcinoma (HCC) is one of the most common liver malignancies worldwide, with increasing rates of morbidity and mortality annually (Bray et al, 2018). Transcription profiling RNA data, along with the HCC clinical data were downloaded from TCGA1 (Liu J. et al, 2018), and were used to identify differentially expressed ARGs. 370 patients with complete survival data were identified from TCGA and randomized into a training set (n = 185) and test set (n = 185); These two sets were used to develop and internally validate the HCC prognostic signature. With Cox regression coefficients of the ARGPs. The formula was as followed: risk score = Cox regression coefficient of ARGPi ∗ expression value of gene ARGPi. The prognostic signature was evaluated by utilizing the training set and validated using the test set, TCGA set, and ICGC set. The receiver operating characteristic (ROC) curve was plotted and the area under the curve (AUC) was calculated to ensure that the prognostic signature prediction efficacy can be estimated Both univariate and multivariate Cox analysis were conducted with the clinicopathologic features and risk score to explore the HCC prognostic factors. The survival ROC R package was used to calculate the AUC of the survival ROC curve

RESULTS

DISCUSSION

CONCLUSION

DATA AVAILABILITY STATEMENT