(View plain language abstract)
Pregnant women receive folic acid supplementation to prevent spina bifida and other birth abnormalities in their children. Supplementation is very successful in preventing these defects, however, the long term effects/consequences of folic acid and other B vitamins are not known. The project aimed to determine whether folate intake and its metabolism during early pregnancy were associated with changes in the fetal epigenome through alterations in DNA methylation and whether these changes were associated with changes in fetal outcomes such as birth weight centile.
A number of studies were undertaken which all adopted genome-wide technologies; in specific studies findings were also considered in combination with demographic data, analysis of folic acid intermediaries, together with the influence of particular polymorphic variants of enzymes in these pathways.
We showed that methylation of LINE-1 sequences (a surrogate for genome wide methylation) in fetal cord blood DNA is inversely correlated with one of the folic acid intermediaries, homocysteine. We also noted that there was an association with birth weight centile (BWC). We then employed an unbiased genome wide analysis, initially exploiting a technique termed Methylated DNA Immunoprecipitation (MeDip). Subsequent to these studies, we utilised newly developed Illumina Infinium Bead Arrays that permitted interrogation of more than 27,000 (27K) CpGs in ~14,000 genes. The MeDip showed that methylation in CpG island “shores” was associated with folate and one of its intermediaries, homocysteine. The 27K arrays examined specific gene promoters and provided data where, through unbiased hierarchical clustering analysis, we showed, for specific genes, correlations between methylation and both BWC and once again with LINE1 methylation. This study, in the second year of our project, also included analysis of other variables including, dietary questionnaires (diet-derived folic acid), and biochemical analyses of folate-related intermediaries together with demographic data including, parity, maternal age and BMI.
One of our key project objectives was to recruit a large cohort of maternal and cord blood samples together with detailed biochemical analyses of folate-related metabolites and BWC in relation to the mother’s demographic variables including, smoking history, and other variables (e.g. polymorphisms in key folate metabolising enzymes). Through the use of unbiased whole-genome technologies, we demonstrated an association between LINE-1 and also gene-specific promoter methylation and BWC.
We also investigated the methylation status of key genes in a significantly larger cohort of cord blood DNA samples together with the multiple other variable described above including our analysis polymorphisms in key enzymes with roles in folate metabolism. In this study we examined 129 cord blood samples. Our selection of genes was on the basis of the earlier studies, and included the effects of potential confounding factors (maternal smoking, folate-related metabolite levels, genetic variations in the methylene tetrahydrofolate reductase (MTHFR) gene) and other demographic variables described above. Linear and logistical regression analyses were used to identify relationships between BWC and methylation levels in the context of these confounding factors. We found that (i) gene-specific methylation of cord blood DNA is associated with BWC and (ii) this methylation augments the effect of smoking during pregnancy on BWC.
We also examined cord blood samples from mothers taking anti-epileptic drugs where mothers are given significantly increased folic acid supplementation peri-conceptually. The effect of the increased intake is thought to counteract the folate-reducing effects of these drugs on methyl group availability.
We hypothesised that the beneficial effect of folic acid intake during early pregnancy might be countered by its effect on gene expression through its impact on the “epigenome”. We aimed to determine these effects on the whole genome. The data would allow identification of changes in methylation, one of several key epigenomic changes, where some of the targets have implications for cancer risk. Our aim was to provide recommendations regarding folate supplementation in the context of future cancer risk.
Folate supplementation is recommended during early pregnancy to prevent spina bifida and other neural tube defects. Studies have shown that folate and its intermediaries influence a process termed methylation and this process leads to what is known as epigenetic changes that can change the expression of genes and lead to diseases such as cancer. This has led some in the field to ask the question “folate friend or foe?” It is now possible to determine the effects of folate on the “epigenome” in an unbiased way as technologies are now available that allow more than 30,000 genes to be examined simultaneously. In this way it is possible to identify the targets for change in DNA drawn from the cord blood of babies at birth to identify change.
We used novel techniques to identifying genome-wide methylation with technologies that look at 27,000-450,000 sequences in a single experiment. The samples were selected from low, median and high folic acid cord bloods. The data was examined by bioinformatics techniques to identify novel genes and related to other potential confounding factors.