Reply to “Heritable germline epimutation is not the same as transgenerational epigenetic inheritance”

Leung et al. reply:

Regarding the heritable germline (soma-wide) methylation of MSH2 (ref. 1) and MLH1 (ref. 2) genes observed in the families reported by us and others, we agree that although they may represent an example of transgenerational epigenetic inheritance, the possibility of an underlying genetic change that causes the heritable methylation cannot be excluded. Whitelaw et al. specified the strict requirements necessary to document transgenerational epigenetic inheritance in non-human organisms; these include experimental conditions that can never be recapitulated in human studies. Interestingly, the mode of inheritance in two new families with germline MLH1 methylation recently reported by Ward and colleagues² is in sharp contrast to our reported family with MSH2 methylation. As Martin et al. pointed out, the new results² provide stronger evidence in support of transgenerational epigenetic inheritance for MLH1. However, the results may still be explained by low penetrance or by the presence of an unlinked modifier. Thus, it is almost impossible to unequivocally document a case for transgenerational epigenetic inheritance in humans even if such a phenomenon exists. Despite the difficulties in distinguishing between primary and secondary epimutations, our study¹ as well as those by Ward and colleagues^2,3 have demonstrated the diverse spectrum of heritable germline methylation, showing, for instance, a range of propensities for transmission to offspring and variation in the degree of mosaicism in tissue distribution of the methylated alleles. The existence of these forms of heritable methylation (either due to incomplete erasure in germ cells or subsequent re-establishment in somatic development), as distinguished from heritable germline mutation, and their role in the causation of the most common hereditary cancer syndrome in humans, deserves recognition. Specifically, the problems they create in genetic diagnosis and their possible role in the causation of other forms of disease that may mimic polygenic or complex traits warrant further study. Also, a common unique feature for these heritable epimutations is the presence of an intact underlying gene that may make modification of the epigenetic states possible as a therapeutic strategy. For example, it has been shown in animal studies that maternal diet during pregnancy may modulate methylation status at epigenetically labile promoter regions in offspring⁴. Last, development of a unified language that can be easily understood by both geneticists and epigeneticists may be necessary to classify these phenomena.