Histone methylation defects as a mechanism for the long-term consequences to offspring of maternal undernutrition.

It is now well established that malnutrition, or other forms of stress in pregnancy, is sometimes associated with the development of heart disease or diabetes in adult life. The mechanism by which this happens is not clear. One suggestion for which there has been some supporting experimental data is that certain genes in the foetus are modified by addition of a methyl group ? so called DNA methylation. Alternatively, we propose, similar modifications may occur to the proteins around which DNA is wound, and this too can alter the way in which genes are expressed. In this project we will use new large scale DNA sequencing technology to identify all the genes that are modified by DNA methylation and all those that are modified by methylation of associated proteins. This pattern of modifications across the genome in early life should reveal the key mechanisms that lead to the phenomenon of foetal programming. Understanding these mechanisms will ultimately allow us to predict the presence of such changes and perhaps to reverse them.

Whilst the phenomenon of foetal programming and its potential contribution to human disease is well recognised and has been extensively and carefully described, our understanding of the underlying mechanisms remains poor. Many researchers have investigated the alterations in expression of specific genes of interest in various models and in some cases have identified DNA methylation as a potential mechanism that could lead to long-term shifts in gene expression. Findings on deficiency of one carbon donors (e.g. folic acid) appear, in part, support these conclusions. We argue that DNA methylation may at best provide only a partial explanation for programmed changes, and that a stronger candidate mechanism would be changes in histone methylation catalysed by polycomb and trithorax complexes. These long-term alterations in histone methylation are most likely to effect genes involved in development and cellular differentiation processes, and may be equally susceptible to one carbon metabolism deficiency. We therefore propose to use the rat maternal low protein diet model to investigate the relative effects of programming and folic acid rescue on histone H3-lysine 4 and 27 trimethylation as well as DNA methylation across the entire genome. This will be achieved by a chromatin immunoprecipitation and parallel DNA sequencing (ChIP-Seq) approach with confirmatory real time PCR analysis of differentially methylated regions. This work will define the role of each of these potential mechanisms and should enable a significant advance in the understanding, prediction and potentially the avoidance of programmed pathologies.