Mitochondrial DNA Haplogroup N9a Negatively Correlates with Incidence of Hepatocellular Carcinoma in Northern China

Abstract

Mitochondrial DNA (mtDNA) haplogroups are associated with various types of cancer; however, the molecular mechanisms by which mtDNA haplogroups affect primary hepatocellular carcinoma (HCC) are not known. In this study, we carried out a case-control study on 388 HCC patients and 511 geographically matched asymptomatic control subjects in northern China. We found that mtDNA haplogroup N9a and its diagnostic SNP, m.16257C > A, negatively correlated with the incidence of HCC in northern China (odds ratio [OR] 0.290, 95% confidence interval [CI] 0.123-0.685, p = 0.005), particularly in patients with infection of hepatitis B/C virus (HBV/HCV) (for haplogroup N9a: OR 0.213, 95% CI 0.077-0.590, p = 0.003; for m.16257C > A: OR 0.262, 95% CI 0.107-0.643, p = 0.003). However, mtDNA haplogroup N9a is not associated with clinical characteristics of HCC including serum alpha-fetoprotein (AFP) level and tumor size. In addition, cytoplasmic hybrid (cybrid) cells with N9a haplogroup (N9a10a and N9a1) had transcriptome profiles distinct from those with non-N9a (B5, D4, and D5) haplogroups. Gene set enrichment analysis (GSEA) showed that metabolic activity varied significantly between N9a and non-N9a haplogroups. Moreover, cells with haplogroup N9a negatively correlated with cell division and multiple liver cancer pathways compared with non-N9a cells. Although it is still unclear how N9a affects the aforementioned GSEA pathways, our data suggest that mtDNA haplogroup N9a is negatively correlated with the incidence and progression of HCC in northern China.

Keywords: GSEA; cybrid; gene set enrichment analysis; haplogroup; hepatocellular carcinoma; mitochondrial DNA; mitochondrial SNP; mtDNA; mtSNP.

Figure 1

The PCA of HCC Patients and Controls, as well as Other Reported Han Populations across China, Based on mtDNA Haplogroup Frequencies HCC patients and controls are from Zhengzhou, Henan, are shown with yellow and green fonts, the previously reported populations from Henan are shown with red fonts, and the previously reported populations from northern and southern China are shown with blue and purple fonts.

Figure 2

The Association between Macro-Haplogroup M/N and Age at Onset HCC p value was estimated by Kaplan-Meier method.

Figure 3

Haplogroup N9a Have Lower Tumorigenic Activity than Non-N9a (B5a1, D5a2, D4g2) Cybrids (A) Cell proliferation curve of N9a and non-N9a cybrids (n = 3). (B) Colony formation assay of N9a and non-N9a cybrids (n = 3). Data are presented by means ± SEM, and ***p < 0.001.

Figure 4

N9a and Non-N9a (B5a1, D5a2, D4g2) Cybrids Feature Distinct Transcriptome Profiles (A) Heatmap of RNA-seq data showing the differentially expressed genes (DEGs) between N9a and non-N9a cybrids (p < 0.05, |log₂FC| > 1). (B) GSEA enrichment using KEGG database showing metabolic pathways are differently activated between N9a and non-N9a cybrids. (C and D) Heatmap showing transcriptional changes of nuclear-encoded (C) and mitochondrially encoded (D) OXPHOS subunits in N9a and non-N9a cybrids. (E) GSEA enrichment of cell-cycle related events in N9a and non-N9a cybrids using KEGG, Reactome, and GCP from MsigDB. (F) GSEA enrichment of liver cancer related events in N9a and non-N9a cybrids using GCP.