Epigenetic priming in chronic liver disease impacts the transcriptional and genetic landscapes of hepatocellular carcinoma

Abstract

Hepatocellular carcinomas (HCCs) usually arise from chronic liver disease (CLD). Precancerous cells in chronically inflamed environments may be 'epigenetically primed', sensitising them to oncogenic transformation. We investigated whether epigenetic priming in CLD may affect HCC outcomes by influencing the genomic and transcriptomic landscapes of HCC. Analysis of DNA methylation arrays from 10 paired CLD-HCC identified 339 shared dysregulated CpG sites and 18 shared differentially methylated regions compared with healthy livers. These regions were associated with dysregulated expression of genes with relevance in HCC, including ubiquitin D (UBD), cytochrome P450 family 2 subfamily C member 19 (CYP2C19) and O-6-methylguanine-DNA methyltransferase (MGMT). Methylation changes were recapitulated in an independent cohort of nine paired CLD-HCC. High CLD methylation score, defined using the 124 dysregulated CpGs in CLD and HCC in both cohorts, was associated with poor survival, increased somatic genetic alterations and TP53 mutations in two independent HCC cohorts. Oncogenic transcriptional and methylation dysregulation is evident in CLD and compounded in HCC. Epigenetic priming in CLD sculpts the transcriptional landscape of HCC and creates an environment favouring the acquisition of genetic alterations, suggesting that the extent of epigenetic priming in CLD could influence disease outcome.

Keywords: chronic liver disease; epigenetic priming; hepatocellular carcinoma; methylation.

Fig. 1

Oncogenic transcriptional alterations in CLD are compounded in HCC. (A) PC analysis of 15 healthy liver and 10 paired CLD and HCC samples. (B) Venn diagrams of down‐ and upregulated genes in CLD and HCC compared with normal (|log₂FC| > 1.5 and q < 0.05). (C) T statistics (absolute values) of 1675 DE genes in both CLD and HCC compared to normal livers. DE: |log2FC| > 1.5, q < 0.05. P computed from paired Wilcoxon tests. (D) Hallmark pathways with significant enrichment in CLD, HCC or both from gene set enrichment analysis are shown (GSEA; P < 0.05). CLD: chronic liver disease; HCC: Hepatocellular carcinoma; DE: differentially expressed.

Fig. 2

DNA methylation alterations in HCC are detectable in CLD. (A) PC analysis of 12 healthy liver samples, and 10 paired CLD and HCC samples, showing PC1 and PC3. (B, C) Differential methylation analysis [−log₁₀(q) against log₂ fold change (M value)] comparing CLD (B) and HCC (C) with healthy livers. Significant CpG sites (|log₂ fold change| > 1.5, q < 0.05) are coloured according to the legend. (D) T statistics (absolute values) of 339 differentially methylated CpG sites in both CLD and HCC compared with normal livers. DM: |log2FC| > 1.5, q < 0.05. P computed from paired Wilcoxon test. (E) Distribution of differentially methylated CpG sites (DMPs) according to their genomic features, detected in 10 CLD and 10 HCC compared with 12 normal livers. (F) Venn diagram showing intersection of DMRs called in CLD and HCC samples compared with normal livers. PC: Principal component; CLD: chronic liver disease; HCC: Hepatocellular carcinoma; DM: differentially methylated; DMP: differentially methylated probes; DMR: differentially methylated region.

Fig. 3

Gene expression and DNA methylation changes define the transition from CLD to HCC. (A) Heatmaps of methylation and gene expression at DE genes associated with DMRs (|mean change in methylation B value| > 0.15 and FDR < 0.05). Dot plots showing mean change in B value (methylation) and log₂ fold change (gene expression) between CLD (n = 10) and normal (n = 12) and HCC (n = 10) and normal (n = 12). (B) cg07554771 methylation and MGMT expression from 10 paired CLD (n = 10) and HCC samples (n = 10). P values computed from limma (methylation) and deseq2 (gene expression). (C) Representative immunohistochemistry images from two paired CLD and HCC biopsies stained with anti‐MGMT antibody and MGMT protein expression IHC scores from 12 paired CLD and HCC samples (paired Wilcoxon’s test). Scale bar 20 and 100 µm. DMR: differentially methylated region; CLD: chronic liver disease; HCC: Hepatocellular carcinoma.

Fig. 4

DNA methylation alterations in CLD and HCC are conserved across cohorts. (A) PC analysis of 12 healthy liver samples, and nine paired CLD and HCC samples from validation cohort, based on B values from the 51 CpG sites in the 18 DMRs associated with DE genes in the discovery cohort. (B) Venn diagrams showing overlap between DM CpG sites in CLD and HCC in the discovery and validation cohorts, and the overlap between CLD‐HCC DM CpG sites across cohorts. (C) Heatmap of methylation across normal, CLD and HCC samples at 8 DE genes associated with DMRs in HCC and CLD samples across both cohorts. Dot plot on right shows mean difference in B values between CLD and normal and HCC and normal. (D) Track plots showing B values at two CLD‐HCC overlapping DMRs in UBD in validation and discovery cohorts. DNAse hypersensitivity sites are denoted in the bottom tracks with ENSEMBL gene annotation. CLD: chronic liver disease; HCC: Hepatocellular carcinoma; DMR: differentially methylated region; DE: differentially expressed; DM: differentially methylated; DMR: differentially methylated region.

Fig. 5

Implications for epigenetic priming in CLD on HCC. (A) The approach used to select CpG sites for elastic net regression using TCGA data. (B, C) Kaplan–Meier plots of survival probability in the TCGA test set (B, n = 111) and in an external validation cohort (C, n = 246), stratified into High/Low CLDme score. Logrank P value adjusted for disease stage and history. (D) Number of somatic mutations in TCGA samples stratified by CLDme score (163 high, 142 low). Violin plots show distributions of mutations per sample. Boxplot shows the mean and IQR. Whiskers show the range of the data up to 1.5 × IQR. Samples outside this range are plotted as points. P value computed from Wilcoxon’s test. (E) Barplot showing percentage of samples with TP53 mutation types, stratified by CLDme (178 CLDme high, 183 CLDme low). (F) Extent of copy‐number alterations in TCGA samples stratified by CLDme score (178 CLDme high, 183 CLDme low). Violin plots, boxplots and statistics as for D. TCGA: The Cancer Genome Atlas; CLDme: chronic liver disease methylation; IQR: inter quartile range.