Comprehensive analysis reveals signal and molecular mechanism of mitochondrial energy metabolism pathway in pancreatic cancer

Abstract

Pancreatic cancer (PAAD) is one of the most malignant tumors with the worst prognosis. The abnormalities in the mitochondrial energy metabolism pathway are intimately correlated with the occurrence and progression of cancer. For the diagnosis and treatment of pancreatic cancer, abnormal genes in the mitochondrial energy metabolism system may offer new targets and biomarkers. In this study, we compared the dysregulated mitochondrial energy metabolism-associated pathways in PAAD based on pancreatic cancer samples in the Cancer Genome Atlas (TCGA) database and normal pancreas samples from the Genotype Tissue Expression project (GTEx) database. Then identified 32 core genes of mitochondrial energy metabolism pathway-related genes (MMRG) were based on the gene set enrichment analysis (GSEA). We found most of these genes were altered among different clinical characteristic groups, and showed significant prognostic value and association with immune infiltration, suggesting critical roles of MMRG involve tumor genesis of PAAD. Therefore, we constructed a four-gene (LDHA, ALDH3B1, ALDH3A1, and ADH6) prognostic biomarker after eliminating redundant factors, and confirming its efficiency and independence. Further analysis indicated the potential therapeutic compounds based on the mitochondrial energy metabolism-associated prognostic biomarker. All of the above analyses dissected the critical role of mitochondrial energy metabolism signaling in pancreatic cancer and gave a better understanding of the clinical intervention of PAAD.

Keywords: gene set enrichment analysis; immunotherapy; mitochondria; pancreatic cancer; prognosis model.

FIGURE 1

Analysis of MMRG gene function and core gene mutation status in pancreatic cancer. (A) Disordered mitochondrial energy metabolism-related pathways in pancreatic cancer; (B) Heat map of core genes in significant gene concentration; (C) Mutation waterfall diagram of MMRG gene set; (D) CNV mutation frequency distribution map.

FIGURE 2

Expression difference of MMRG core gene of clinical characteristics in tumor and normal samples. (A–G) The expression difference of MMRG core gene set of gender, grade, history of chronic pancreatitis, radiotherapy, stage, age, and smoking history, respectively.

FIGURE 3

Survival analysis of MMRG core gene in pancreatic cancer patients. (A–F) Survival analysis of LDHA, ALDH3B1, LDHAL6B, PKM, ALDH3A1, and PGAM4, respectively (p < 0.05).

FIGURE 4

MMRG core gene immune microenvironment and protein-protein interaction (PPI) network. (A) The correlation between differential expression of MMRG and tumor immune cell infiltration (positive in red and negative in blue); (B) MMRG core gene protein-protein interaction (PPI) network. Color of the gene indicates the differential expression (log2FC) and the node size indicates the degree which also displayed by the barplot leftside; (C) Spearman correlation of MMRG core gene.

FIGURE 5

Prognostic analysis of MMRG core gene signature. (A–E) KM, ROC, risk score, survival time distribution, and expression of prognostic factors of MMRG core gene signature.

FIGURE 6

Risk assessment model construction and correlation of clinical characteristics with different groups. (A) Forest map of univariable and multivariate cox factors in the training set; (B) Forest map of univariable and multivariate cox factors in the validated set; (C–H) Risk scores in different clinical characteristics (age, history of chronic pancreatitis, history of diabetes, history of alcohol documented, grade, and gender, respectively).

FIGURE 7

Difference of immune infiltration and prognosis of immunotherapy between high and low-risk scoring groups. (A) ESTIMATEScore, ImmuneScore, StromalScore, TumorPurity of high and low score groups; (B) Immune infiltration score of high-risk groups; (C) KM curve of the high or low-risk groups; (D) Objective response rate in high- and low-risk groups; (E) Risk score distribution in different clinical response groups.

FIGURE 8

Drug resistance analysis and potential therapeutic compounds. (A) Top 6 drugs whose IC50 positively correlated with the risk score; (B) Top 6 drugs whose IC50 negatively correlated with the risk score; (C, D) IC50 distribution difference of positive and negative related drugs between high and low score groups.