Table 1 DNA methylation and HCC
Study objectives | Methods | Type of sample | Genes identified | Key findings | Implications of study | Refs. | |
|---|---|---|---|---|---|---|---|
Hypo-methylated | Hyper-methylated | ||||||
To investigate the global methylation profile of purified single hepatocytes | Illumina Infinium Human Methylation27 BeadChip, COBRA and bisulfite sequencing | Single hepatocytes isolated from HBV-positive HCC (HBHC) tissues | – | EMILIN2, WNK2, TM6SF1, TLX3, HIST1H4F, TRIM58, GRASP | Hepatocyte methylation profiles can be classified into 3 groups based on hepatocyte origin: HCC, adjacent tissue and normal liver. 7 novel genes were found to be aberrantly methylated in HBHC | These genes can be potential novel biomarkers of HBHC once validated in larger clinical cohorts | Tao et al. [16] |
To identify genes hypermethylated in HCC that can be detected in plasma DNA for early diagnosis | Illumina Infinium Human Methylation27 BeadChip | 62 paired HCC tumor and NAT | CCL20, AKT3, SCGB1D1, WFDC6, PAX4, GCET2, CD300E, CD1B, FLJ00060, MNDA | DAB2IP, BMP4, ZFP41, SPDY1, CDKN2A, TSPYL5, CDKL2, ZNF154, ZNF540, CCDC37 | 684 CpG sites significantly hypermethylated in HCC tissues. 5 of these genes (CDKL2, CDKN2A, HIST1H3G, STEAP4, ZNF154) had detectable hypermethylated DNA in plasma of up to 63% of patients | Measuring DNA methylation from patient plasma is feasible. Panel of methylated genes identified can be potential biomarkers for early diagnosis | Shen et al. [17] |
To study aberrant DNA methylation in HCC using higher resolution genome-wide analysis | Illumina HumanMethylation 450 BeadChip | 27 HCC and 20 NAT | NFATC1 | BMP4, CDKN2A, GSTP1 | Greater global hypomethylation patterns observed in HCC compared to NAT, with higher frequency of events occurring in promoter CpG islands than CpG shores and shelves | Allows deeper understanding of differential methylation patterns in various gene regulatory regions | Song et al. [18] |
To identify tumor suppressor genes silenced by DNA methylation in HCC | Illumina Infinium Human Methylation27 BeadChip, combined with microarray analysis of gene re-expression studies | 71 primary HCC tissues, 8 non-diseased normal tissues, 4 HCC cell lines | – | ACTL6B, C19orf30, DGKI, DLX1, ELOVL4, LDHB, LRAT, MLF1, NEFH, PPM1 N, PRPH, SLC8A2, SMPD3 | Identified 13 candidate tumor suppressor genes; NEFH and SMPD3 were functionally validated in vitro and in vivo. Low levels of SMPD3 were associated with early HCC recurrence after curative surgery in an independent patient cohort | SMPD3 identified to be a potent tumor suppressor gene and could be an independent prognostic factor for early recurrence of HCC | Revill et al. [19] |
To investigate novel genome-wide aberrant DNA methylation patterns in HCV-related HCC | Illumina Infinium HumanMethylation 450 BeadChip | 66 pairs of HCC tumor and NAT | Â | Identified 500 significant differentially methylated CpG sites that can distinguish HCC from NAT. Within NAT tissues, 228 CpG sites were identified to be significantly associated with HCV infection | Further functional studies would help to identify markers among the large subset of CpG sites/genes found to correlate with HCV infection, liver cirrhosis or HCC to aid in diagnosis and treatment | Shen et al. [20] | |
To investigate the genome-wide DNA methylation profile and identify stochastic epigenetic mutations (SEMs) in HCC | Illumina Infinium HumanMethylation 450 BeadChip | 69 pairs of HCC tumor and NAT | AJAP1, ADARB2, PTPRN2, SDK1 (hypermethylated at promoter level with concomitant hypomethylation at gene body level) | HCC tissues showed increased number of SEMs as compared to NAT. From a subset of SEMs unique to tumor tissues, 4 epigenetically-regulated genes that could be involved in HCC tumorigenesis were identified | Methylation and SEM profiles of HCC and adjacent non-cancerous liver tissues are highly different, allowing for the identification of important driver epimutations in HCC | Gentilini et al. [22] | |
To examine the effects of epigenetic alterations and features on the HCC genome architecture | Whole-genome bisulfite, whole-genome shotgun, long read and virus-capture sequencing approaches | 373 HCC cases | NA | Somatic mutations occur preferentially in both highly methylated as well as hypomethylated regions in the liver cancer genome. HBV integration sites occur more frequently in inactive chromatin regions | Epigenetic features greatly influence the mutational processes in HCC. Understanding the mechanisms behind the interdependency between genetic, viral and epigenetic alterations in HCC can help in identifying epigenetic driver events | Hama et al. [23] | |