Impact of maternal diet on the epigenome and potentially modifiable effects on offspring health

Our experiences in early life can have life-long effects on our health and wellbeing. For example, in a rural population in The Gambia in West Africa we have observed that children born in the rainy season are 6 times more likely to die between the ages of 15 and 65 than those born in the dry season. In fact there is mounting evidence that detrimental influences on lifelong health can stretch right back to the early stages of embryonic development. This underlines the importance of research into the underlying mechanisms, so that the processes linking environmental exposures to negative outcomes can be understood, and hopefully corrected.

One such possible mechanism involves a process known as methylation, which is one type of 'epigenetic' modification of the genome. Methylation requires a defined set of nutrients including folic acid and B-vitamins, both to provide the necessary chemical compounds, known as methyl groups, and to undertake the necessary metabolic conversions. Animal experiments have previously shown that supplementing the diets of female mice with these nutrients before they conceive has a profound effect on their offspring's appearance (e.g. changing their coat colour) and that these changes were associated with higher levels of methylation on their DNA. Until now it was unknown whether similar effects on offspring methylation occur in humans, but our group recently presented first-in-human evidence that they do. We have since followed up this work by looking at patterns of methylation across the genome. We found evidence of unusual or 'disrupted' methylation patterns associated with both maternal nutrient status and season of conception in certain types of genes, and notably in one gene (VTRNA2-1), where disrupted methylation has previously been linked with some forms of cancer and also with negative effects on the immune system.

With this grant we hope to extend this work in a number of ways. Firstly, we want to characterise these patterns of disruption more precisely by looking at a larger number of infants; interrogating key regions of the epigenome at high resolution using more advanced technologies; and looking for methylation effects all the year round. We hope this will provide further clues about the mechanisms underlying epigenomic disruption. Secondly, we want to investigate the effects of disrupted methylation in VTRNA2-1. Our Gambian research centre is a particularly good place to do this as we are able to link an individual's epigenetic information with medical records and other demographic data, and we can also conduct detailed laboratory investigations on blood cells in individuals known to have abberant methylation. These functional studies are an important part of the chain linking epigenetic effects to real, adverse health outcomes in people. Finally, with the help of advanced computer modelling, we will identify the specific combination of MD-related nutrients that may be causing the observed patterns of disrupted methylation. We will then develop a nutritional supplement to correct the observed suboptimal nutrient profile, and we will test its effectiveness in a randomised controlled trial. If effective, in future work we would seek to assess the effect on offspring methylation of giving this supplement to mothers-to-be. The hope is that the patterns of disrupted methylation previously observed in infants conceived at certain times of the year would then be prevented. In the longer term, we hope that the work described here will inform strategies for pre-conceptional supplementation in mothers that will lead directly to improved outcomes for infant growth and development, with life-long benefits for health and wellbeing.

By exploiting a natural experiment in which rural Gambians consume different diets in dry and rainy seasons, we have recently obtained first-in-human evidence that maternal nutrient status at conception leads to sustained systemic alterations in DNA methylation of offspring. We further showed that conception in the dry season, with measured deficiencies in maternal methyl donor (MD) nutrients, preferentially disrupts the methylation of imprinted genes, allele-specifically expressed genes and metastable epialleles, leading to an excess of both hyper- and hypo-methylation and implying there will be variable developmental phenotypes. The gene with strongest evidence for disruption (VTRNA2-1) plays a central role in regulating RNA-dependent protein kinase (PKR); a pivotal antiviral protein with known tumour suppressor activity. We here seek funds to extend this work with the following aims: a) to fine map the patterns of epigenomic disruption within sensitive gene classes using refined methods (targeted capture and next generation bisulphite sequencing) on banked samples from a much larger cohort, to more precisely pinpoint conception months (and related exposures) associated with minimal and maximal disruption, and identify the most disrupted genes; b) to use a 'recall-by-epigenotype' design to study the functional consequences in T-cells and monocytes of disrupted methylation at VTRNA2-1 (expression by Northern blotting, PKR activation by Western blotting, and assessment of phosphorylation of eIF2a); c) to use in silico modelling of our MD metabolome data to identify key nutrient imbalances leading to disrupted methylation, use these insights to design supplements to optimise methylation pathways, and pilot these in non-pregnant women in anticipation of a large future RCT of pre-conceptional supplementation to prevent disrupted methylation. This work may have far-reaching implications for populations worldwide and across numerous physiological and disease endpoints.

We are optimistic that our proposed research can have a very high level of impact with potential implications for future children worldwide. Our current analyses (summarised in the Case for Support) strongly suggest that we have serendipitously discovered a link between pre- and peri-conceptional environmental exposures (with diet being the leading candidate) and instability in the normal processes of DNA methylation leading to epigenetic errors. Such errors are known to lead to many pathologic syndromes arising from developmental defects and these are likely to be the tip of an iceberg. If, as our data suggest, a mother's nutritional status at conception can cause a permanent disruption of methylation at differentially methylated regions in imprinted genes, allele-specifically expressed genes and metastable epialleles, then there is a very high likelihood that these could be linked to a wide range of 'disordered' phenotypes. The fact that our most susceptible candidate gene is a tumour suppressor gene that also plays a central role in regulating immunometabolism provides an indication of the potential significance of our work. The implications are that by correcting a woman's methyl donor status prior to conception we could greatly reduce epigenetic instability and avoid a large proportion of methylation errors. It is conceivable that such errors contribute to a wide range of avoidable pathologies, sub-clinical syndromes and disease susceptibility traits. For instance, imprinted genes play an active role in fetal and placental development and imprinting errors are already known to cause a host of developmental abnormalities. We speculate that they may influence traits such as intra-uterine growth retardation (IUGR) and pre-term birth (PTB) that, whilst less severe for the individual, represent enormous burdens of disease on a worldwide basis.

Who will benefit from this research? a) Parents worldwide who are planning a family. Note that folic acid is already widely promoted and there are many supplement formulations marketed to mothers-to-be, but our data suggest that folic acid alone is not sufficient and other supplements are purely empirical and lack an evidence base; b) Governments and their health systems. Although is is premature to make such claims it is not inconceivable that reducing epigenetic errors by improving mothers' preconceptional nutritional status could reduce the prevalence of a wide range of diseases that affect the health, wealth and well-being of populations.

How will they benefit from this research? Our ultimate vision is to be able to formulate appropriate foods and/or supplements to optimise the metabolic machinery necessary for appropriate DNA methylation during early embryonic and fetal development. These could be promoted to mothers-to-be through multiple routes; either commercially or non-commercial. There is a strong chance that a 'one-size-fits-all' solution may not work and that supplements would need to be targeted by country, population groups, ethnicity, etc or even formulated for each woman. Our in silico modelling of nutrient effects on methyl donor pathways will be important in inserting this question and in designing a range of supplements to suit different population groups.