Déborah Bourc’his: the art of methylation

What exactly is DNA methylation? At what point does this phenomenon arise? 

It involves chemical modifications that take place to regulate gene expression without changing the DNA sequence. Groups known as “methyl groups” position themselves on the DNA, which changes the way in which the cell reads the genome. DNA methylation represses the genes, in a stable but potentially reversible manner. We talk about epigenetic marks. DNA methylation plays a crucial role in early development in particular, as the embryo is forming all the cell types. We also know that tumorous cells often have abnormal methylation profiles, related to their abnormal cell identity.

In my team, we are studying how via their gametes, parents transmit not only their genetic heritage, but epigenetic information that will influence the way in which their offspring’s genes are expressed and, as a result, their physiology. Disturbances to this parental epigenetic memory have serious pathological consequences in terms of development and viability.

Tell us about your recent discoveries

Our results demonstrate the importance of DNA methylation at two separate moments in reproduction: in the sperm cells for the publication in Science, prior to fertilisation, and after fertilisation for the publication in Nature Genetics, in the very early embryo stages.

We believed that all the enzymes responsible for controlling DNA methylation, 4 of them, had already been identified in mammals. Each of these enzymes is essential in itself, as demonstrated by pathologies resulting in mutations in DNA methyltransferase, in mice or in humans.  By chance, following the spontaneous occurrence of a mutation in a colony of mice, we have just revealed a new methyltransferase, DNMT3C, in collaboration with Florian Guillou, from the INRA in Tours, and Yann Hérault, from the IGBMC in Strasbourg (CNRS/Inserm). The level of specialisation of DNMT3C is very surprising: it acts only during foetal spermatogenesis, it methylates only transposons and selectively recognises the most active and thus potentially the most dangerous. Transposons are types of parasitic sequences endogenous to our genome, whose activity can lead to “risky” rearrangements and thus alter hereditary material.

Without DNMT3C the life of mice is not affected, but without methylation the transposons are massively reactivated during spermatogenesis: no sperm is produced and the male mice are completely sterile. Our work shows that DNMT3C is an essential protector of fertility.

Another surprise: DNMT3C is an innovation specific to rodents, which appeared around 46 million years ago. Now we want to understand why this enzyme is so specialised and which alternative mechanisms exist in humans, who don’t have DNMT3C, to block the action of transposons in sperm cells. This discovery also raises a major question in terms of public health. How can we extrapolate studies on reproduction in rodents, such as for example the harmful effects of endocrine disruptors, given that the epigenetic mechanisms regulating fertility present differences that are not negligible, under the effects of DNMT3C or not.

Our second discovery highlights the long-term function of a temporary parental imprint, a phenomenon that we had previously demonstrated but without any idea of its physiological role.

What is the parental imprint?

At the time of fertilisation, the oocyte and the sperm cell have drastically different methylation profiles in mammals, which creates a methylation asymmetry on the paternal and maternal chromosomes in the newly-formed embryo. During the early days of development, these parental methylation profiles are reprogrammed. However, a handful of genes keep this parental methylation asymmetry in their memory throughout their life, which leads to their expression from a single chromosome copy. Although there are few of them (around one hundred), the imprinted genes have a determining role in foetal, neurone and metabolic development.

We have identified a new form of parental imprint which is atypical since it is temporary: it lasts only for the very early days of development following fertilisation (4 in rodents and 7 in humans) then it disappears. By studying the Zdbf2 gene in mice, we showed that the temporary imprint of this gene, although labile, irreversibly programmes the height of individuals. It triggers a type of switch, which will act a lot later on the hypothalamo-hypophyseal portal system, which is at the centre of multiple hormone regulation pathways, including that of the growth hormone. Embryos that do not have this temporary imprint develop completely normally, but after birth mice experience delayed growth and remain smaller throughout their lives. This is one of the few examples of early epigenetic programming, in the embryo, of features that will appear a lot later in the adult years. Now we are seeking to find out whether the temporary imprint of Zdbf2 could also play a role in determining the height of humans. Some external factors such as the mother’s nutrition or the use of medically-assisted reproduction methods could also alter this early programming.