Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 May 26:8:e9201.
doi: 10.7717/peerj.9201. eCollection 2020.

Comprehensive analysis reveals a metabolic ten-gene signature in hepatocellular carcinoma

Zhipeng Zhu #  1 , Lulu Li #  1 , Jiuhua Xu  2 , Weipeng Ye  2 , Borong Chen  1 , Junjie Zeng  1 , Zhengjie Huang  1   2
Affiliations

Comprehensive analysis reveals a metabolic ten-gene signature in hepatocellular carcinoma

Zhipeng Zhu et al. PeerJ. .

Abstract

Background: Due to the complicated molecular and cellular heterogeneity in hepatocellular carcinoma (HCC), the morbidity and mortality still remains high level in the world. However, the number of novel metabolic biomarkers and prognostic models could be applied to predict the survival of HCC patients is still small. In this study, we constructed a metabolic gene signature by systematically analyzing the data from The Cancer Genome Atlas (TCGA), Gene Expression Omnibus (GEO) and International Cancer Genome Consortium (ICGC).

Methods: Differentially expressed genes (DEGs) between tumors and paired non-tumor samples of 50 patients from TCGA dataset were calculated for subsequent analysis. Univariate cox proportional hazard regression and LASSO analysis were performed to construct a gene signature. The Kaplan-Meier analysis, time-dependent receiver operating characteristic (ROC), Univariate and Multivariate Cox regression analysis, stratification analysis were used to assess the prognostic value of the gene signature. Furthermore, the reliability and validity were validated in four types of testing cohorts. Moreover, the diagnostic capability of the gene signature was investigated to further explore the clinical significance. Finally, Go enrichment analysis and Gene Set Enrichment Analysis (GSEA) have been performed to reveal the different biological processes and signaling pathways which were active in high risk or low risk group.

Results: Ten prognostic genes were identified and a gene signature were constructed to predict overall survival (OS). The gene signature has demonstrated an excellent ability for predicting survival prognosis. Univariate and Multivariate analysis revealed the gene signature was an independent prognostic factor. Furthermore, stratification analysis indicated the model was a clinically and statistically significant for all subgroups. Moreover, the gene signature demonstrated a high diagnostic capability in differentiating normal tissue and HCC. Finally, several significant biological processes and pathways have been identified to provide new insights into the development of HCC.

Conclusion: The study have identified ten metabolic prognostic genes and developed a prognostic gene signature to provide more powerful prognostic information and improve the survival prediction for HCC.

Keywords: Bioinformatics; Biomarker; Diagnosis; Gene signature; Hepatocellular carcinoma; Metabolism; Prognosis; Survival.

PubMed Disclaimer

Conflict of interest statement

The authors declare that they have no competing interests.

Figures

Figure 1
Figure 1. The schematic workflow of the study.
Figure 2
Figure 2. Heatmap and Volcano plot of metabolism-related DEGs.
(A) Heatmap of metabolism-related DEGs. Red indicates that the gene expression is relatively high, green indicates that the gene expression is relatively low, and white indicates no significant changes in gene expression (FDR < 0.05, absolute log FC > 1.5). A total of 10 prognostic genes were marked using red arrow. (B) Volcano plot of metabolism-related DEGs. The red points represent high expression genes, the green points represent low expression genes, the black points represent genes with no significant difference (FDR < 0.05, absolute log FC > 1.5). A total of 10 prognostic genes were marked using black arrow.
Figure 3
Figure 3. Expression of the ten genes in low- and high-risk groups of training cohort, internal testing cohort, GSE14520 testing cohort and ICGC testing cohort.
(A) Training cohort. (B) Internal testing cohort. (C) GSE14520 testing cohort. (D) ICGC testing cohort.
Figure 4
Figure 4. Gene signature performance analysis using training cohort.
(A) Distribution of 10‐gene‐based risk scores, patient survival durations, gene expression levels. (B) One-year ROC curve analyses of gene signature and clinical parameters. (C) Three-year ROC curve analyses of gene signature and clinical parameters. (D) Five-year ROC curve analyses of gene signature and clinical parameters. (E) Kaplan–Meier curves of OS based on gene signature. (F) Prognostic value detection of the gene signature via univariate survival-related analysis. (G) Prognostic value detection of the gene signature via multivariate survival-related analysis.
Figure 5
Figure 5. Gene signature performance analysis using ICGC testing cohort.
(A) Distribution of 10‐gene‐based risk scores, patient survival durations, gene expression levels. (B) One-year ROC curve analyses of gene signature and clinical parameters. (C) Three-year ROC curve analyses of gene signature and clinical parameters. (D) Five-year ROC curve analyses of gene signature and clinical parameters. (E) Kaplan–Meier curves of OS based on gene signature. (F) Prognostic value detection of the gene signature via univariate survival-related analysis. (G) Prognostic value detection of the gene signature via multivariate survival-related analysis.
Figure 6
Figure 6. Gene signature performance analysis using GSE14520 testing cohort.
(A) Distribution of 10‐gene‐based risk scores, patient survival durations, gene expression levels. (B) One-year ROC curve analyses of gene signature and clinical parameters. (C) Three-year ROC curve analyses of gene signature and clinical parameters. (D) Five-year ROC curve analyses of gene signature and clinical parameters. (E) Kaplan–Meier curves of OS based on gene signature. (F) Pognostic value detection of the gene signature via univariate survival-related analysis. (G) Prognostic value detection of the gene signature via multivariate survival-related analysis.
Figure 7
Figure 7. Gene signature performance analysis using internal testing cohort.
(A) Distribution of 10‐gene‐based risk scores, patient survival durations, gene expression levels. (B) One-year ROC curve analyses of gene signature and clinical parameters. (C) Three-year ROC curve analyses of gene signature and clinical parameters. (D) Five-year ROC curve analyses of gene signature and clinical parameters. (E) Kaplan–Meier curves of OS based on gene signature. (F) Prognostic value detection of the gene signature via univariate survival-related analysis. (G) Prognostic value detection of the gene signature via multivariate survival-related analysis.
Figure 8
Figure 8. Gene signature performance analysis using entire testing cohort.
(A) Distribution of 10‐gene‐based risk scores, patient survival durations, gene expression levels. (B) One-year ROC curve analyses of gene signature and clinical parameters. (C) Three-year ROC curve analyses of gene signature and clinical parameters. (D) Five-year ROC curve analyses of gene signature and clinical parameters. (E) Kaplan–Meier curves of OS based on gene signature. (F) Prognostic value detection of the gene signature via univariate survival-related analysis. (G) Prognostic value detection of the gene signature via multivariate survival-related analysis.
Figure 9
Figure 9. The predictive performance of the gene signature on OS in different subgroups stratified by clinical parameters.
(A) Training cohort. (B) Internal testing cohort. (C) ICGC testing cohort. (D) GSE14520 testing cohort. (E) Entire testing cohort.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Abdelzaher E, Mostafa MF. Lysophosphatidylcholine acyltransferase 1 (LPCAT1) upregulation in breast carcinoma contributes to tumor progression and predicts early tumor recurrence. Tumour Biology. 2015;36(7):5473–5483. doi: 10.1007/s13277-015-3214-8. - DOI - PubMed
    1. Aoki T, Weber G. Carbamoyl phosphate synthetase (glutamine-hydrolyzing): increased activity in cancer cells. Science. 1981;212(4493):463–465. doi: 10.1126/science.7209543. - DOI - PubMed
    1. Arner ES, Holmgren A. The thioredoxin system in cancer. Seminars in Cancer Biology. 2006;16(6):420–426. doi: 10.1016/j.semcancer.2006.10.009. - DOI - PubMed
    1. Aye Y, Li M, Long MJ, Weiss RS. Ribonucleotide reductase and cancer: biological mechanisms and targeted therapies. Oncogene. 2015;34(16):2011–2021. doi: 10.1038/onc.2014.155. - DOI - PubMed
    1. Barajas JM, Reyes R, Guerrero MJ, Jacob ST, Motiwala T, Ghoshal K. The role of miR-122 in the dysregulation of glucose-6-phosphate dehydrogenase (G6PD) expression in hepatocellular cancer. Scientific Reports. 2018;8(1):9105. doi: 10.1038/s41598-018-27358-5. - DOI - PMC - PubMed

Grants and funding

This work was supported by Xiamen Scientific and Technological Plan (No. 3502Z20194005, 3502Z20184020). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

LinkOut - more resources