Establishment of a new molecular subtyping and prognostic signature with m6A/m5C/m1A/m7G regulatory genes for hepatocellular carcinoma

Ting Liu; Yang Wang; Zhizhao Li; Lei Sun; Kun Yang; Jiamin Chen; Xiaoyi Han; Liming Qi; Xingang Zhou; Peng Wang

doi:10.1016/j.heliyon.2023.e21285

Establishment of a new molecular subtyping and prognostic signature with m6A/m5C/m1A/m7G regulatory genes for hepatocellular carcinoma

Heliyon. 2023 Oct 30;9(11):e21285. doi: 10.1016/j.heliyon.2023.e21285. eCollection 2023 Nov.

Authors

Ting Liu¹, Yang Wang², Zhizhao Li³, Lei Sun¹, Kun Yang¹, Jiamin Chen¹, Xiaoyi Han¹, Liming Qi¹, Xingang Zhou¹, Peng Wang¹

Affiliations

¹ Department of Pathology, Beijing Ditan Hospital, Capital Medical University, No. 8 Jing Shun East Street, Chaoyang District, Beijing 100015, People's Republic of China.
² Department of General Surgical Department, Beijing Ditan Hospital, Capital Medical University, No. 8 Jing Shun East Street, Chaoyang District, Beijing 100015, People's Republic of China.
³ Department of Cardiovasology, Beijing Ditan Hospital, Capital Medical University, No. 8 Jing Shun East Street, Chaoyang District, Beijing 100015, China.

Abstract

Background: RNA modification, including m6A, m5C, m1A, and m7G, participated in tumor progress. Therefore, the purpose of the present study was to explore the role of m6A/m5C/m1A/m7G regulatory genes in the prognosis and tumor microenvironment (TME) for hepatocellular carcinoma (HCC).

Methods: 71 m6A/m5C/m1A/m7G regulatory genes expression for HCC was detected, differentially expressed genes were screened, and molecular forms were classified by unsupervised consensus clustering. Cox regression and the Least Absolute Shrinkage and Selection Operator (LASSO) analysis were applied to establish a prognostic signature. Time-dependent receiver operating characteristic (ROC) curves were evaluated for clinical effectiveness and accuracy of the prognostic hazard model. In cluster subtypes and risk models, the differences in prognosis, immune cell infiltration, immune checkpoint, immunotherapy, and drug sensitivity between different subtypes were evaluated.

Results: HCC patients were classified into two clusters (cluster 1 and cluster 2) according to the expression of 71 m6A/m5C/m1A/m7G regulatory genes. Cluster 1 had a poor prognosis and different immune cell infiltration. Cluster 1 had higher immune checkpoint expression and TIDE score than cluster 2. Subsequently, we construct a five-gene prognostic model of m6A/m5C/m1A/m7G regulatory genes (YTHDF2, YTHDF1,YBX1, TRMT61A, TRMT10C). The Kaplan-Meier and ROC curve analysis showed that the prognostic signature exhibited good predictability. The risk score was considered an independent poor prognostic index. The high-risk group had higher immune checkpoint expression and higher TIDE scores. 5-Fluorouracil, docetaxel, doxorubicin, etoposide, gemcitabine, paclitaxel, sorafenib, and vinblastine were more suitable for high-risk patients. ECM receptor interaction, cell cycle, and Leishmania infection were enriched in the high-risk group.

Conclusion: The clustering subgroups and prognostic model of m6A/m5C/m1A/m7G regulatory genes were linked with bad prognosis and TME for HCC, and had the potential to be a novel tool to evaluate the outcomes of HCC patients.

Keywords: Hepatocellular carcinoma; Molecular typing; Prognostic genes; Tumor microenvironment; m6A/m5C/m1A/m7G regulatory genes.