THE IMPACT OF ONE-CARBON METABOLISM POLYMORPHIC VARIANTS ON CANCER RISK AND SURVIVAL RATE IN PRIMARY LIVER AND COLON NEOPLASMS ACCORDING TO GLOBAL DNA METHYLATION AND HYDROXYMETHYLATION STATUS: ROLE OF EPIGENETIC MODULATION FROM THE RESULTS OF A CLINICAL STUDY

One-carbon metabolism is a complex pathway involved both in nucleotide synthesis and in biological methylation reactions including that of DNA methylation, the main epigenetic mechanism in humans. An aberrant DNA methylation is a common feature in cancer disease. Several polymorphic gene variants within one-carbon metabolism have been related to cancer risk, but whether the link between one-carbon metabolism variants and cancer risk/survival rate may occur through an altered DNA methylation is still poorly investigated and yet unclear. In addition to DNA methylation, hydroxymethylation of DNA has been recently described as a novel, potentially useful epigenetic mark. Primary liver cancers, i.e. hepatocellular carcinoma (HCC) and cholangiocarcinoma (CC) as well as colon cancer are highly prevalent but epigenetically poorly characterized, so far. The main aims of the present project were to define a possible role for DNA methylation in cancer, and specifically: 1. to determine whether polymorphic variants of one- carbon metabolism genes may influence the risk of human primary liver cancers and colon cancer through DNA methylation, defined as methylcytosine (mCyt) levels measured in peripheral blood mononuclear cells (PBMCs) DNA; 2. to verify whether DNA methylation status measured in PBMCs of subjects affected by HCC, CC and colon cancer may correlate with clinical outcomes and survival rate, therefore defining a possible epigenetic marker of disease; 3. to further characterize the epigenetic layout of HCC and CC by evaluating the global DNA mCyt and hydroxymethylcytosine (hmCyt) content in neoplastic and homologous non-neoplastic tissues. Material and Methods: Genotyping for BHMT 716A>G, DHFR 19bp ins/del, MTHFD1 1958G>A, MTHFR 677C>T, MTR 2756A>G, MTRR 66A>G, RFC1 80G>A, SHMT1 1420C>T, TCII 776C>G and TS 2rpt-3rpt was performed in 102 cancer patients and 363 cancer-free subjects. Methylcytosine and hmCyt were measured by an LC/MS/MS method in PBMCs DNA of all subjects and in neoplastic and homologous non-neoplastic tissues of 47 primary liver cancer patients undergoing curative surgery. A follow-up examination was performed on all the patients 48-60 months after the surgical procedure. Results: 1. The MTHFD1 1958AA genotype was significantly less frequent among cancer patients as compared to controls (p=0.007) and related to 63% reduction of the overall cancer risk (OR=0.37, p=0.003) and 75% of risk for colon cancer (OR=0.25, p=0.006). The MTHFD1 1958GG was significantly more frequent among cancer patients (p=0.007) and associated to lower DNA methylation as compared to MTHFD1 1958 allele A carriers (p=0.048). 2. The RFC1 80AA polymorphic variant was associated in primary liver cancer with a significantly reduced survival rate as compared to GG and GA (p=0.005) at a follow-up period of 60 months. When the mCyt levels were stratified as either high (>5.34%) or low (≤5.34%) according to the mCyt median value, the combination of AA genotype and low mCyt led to a significantly worse survival as compared to the two genotype groups carrying the G allele considered as a whole (p<0.0001); no difference in survival was found between G carriers and AA in association with high mCyt (p=0.919). Moreover, a higher mortality risk was associated with the concomitant presence of the AA genotype and low mCyt, as compared to G carrier subjects (OR=8.35, p=0.001). 3. In primary liver cancer, levels of mCyt ≥5.59% in PBMCs were associated to a significantly higher life expectancy as compared to mCyt<5.59% (p=0.034) at a follow-up period of 48 months. Methylcytosine content in HCC neoplastic tissue was notably lower than in CC tissues (3.97% vs. 5.26% respectively, p<0.0001). Significantly reduced mCyt levels were observed in HCC neoplastic as compared to non-neoplastic tissue (3.97% vs. 4.82% mCyt, p<0.0001) but such difference was not found in liver tissue of patients affected by CC. Hydroxymethylation was significantly decreased in HCC neoplastic vs. non-neoplastic liver tissue (0.044 vs. 0.128, p<0.0001), as well as in CC neoplastic tissue vs. both non-neoplastic liver and gall bladder (0.030 vs. 0.124, p=0.026; 0.030 vs. 0.123, p=0.006, respectively). Conclusions: Our results suggest a possible association between one-carbon metabolism genes polymorphic variants and DNA methylation. This association suggests that DNA methylation may be the underlying link between polymorphic gene variants and primary liver and colon cancers, both in terms of risk and life expectancy. This finding may situate PBMCs DNA methylation as a possible biomarker for those types of cancer disease. Moreover, since DNA methylation is a reversible phenomenon that relies on one-carbon metabolism for provision of methyl groups, adequate strategies to optimize sources of such compounds to maintain DNA methylation within adequate levels for disease prevention may be considered in the framework to modulate this epigenetic mechanism. The results of this study demonstrate, furthermore, that a significantly lower degree of DNA hypomethylation characterize HCC from CC tissue, whereas DNA hypo-hydroxymethylation characterizes both HCC and CC neoplastic tissues as compared to the non-neoplastic. Further validation of the present novel data is required to assess whether this genetic-epigenetic biomarkers may indeed be a reliable and useful for clinical application as a prognostic tool for patients with primary liver and colon cancers.