For more than 20 years, DNA methylation has been the only known mechanism for regulating mammalian germline imprinting. Now, researchers at Harvard Medical School and Boston Children’s Hospital have identified a new mechanism for imprinting in mice. The findings highlight that histone H3 lysine 27 trimethylation (H3K27me3) modifications are able to silence unmethylated regions of DNA in mouse embryogenesis. Details of this work were published online today in Nature, in an article entitled “Maternal H3K27me3 Controls DNA Methylation-Independent Genomic Imprinting.”
The activation of imprinted genes triggered by removal of H3K27me3 from developing embryos was the key experiment in the study. Yi Zhang, Ph.D., investigator at Harvard Medical School, Boston Children's Hospital, and Howard Hughes Medical Institute, explained the conceptual advance that this has brought to the field: “Previous studies in mice linking H3K27me2 to imprinting are mostly on the traditionally DNA-methylation-dependent genes, thus our study is the first to reveal [a] DNA-methylation-independent imprinting mechanism.” A total of 76 potential candidate genes imprinted by H3K27me3 were identified, several of which are linked to placental development.
Loss-of-imprinting causes a variety of developmental disorders, but the potential to therapeutically intervene by turning epigenetic modifications on or off is an easier solution than fixing a mutated or missing gene. Dr. Zhang reflects on two of the putative genes imprinted by H3K27me3 in mice: “Currently, imprinting defects in mouse Gab1 and Sfmbt2 genes exhibit placental developmental defects. Whether this function is conserved in humans remains to be seen.”
The team used DNase I sequencing of maternal and paternal pronuclei to map imprinted regions in mouse zygotes and morula embryos and compared these with genome-wide methylation sequencing data. First author Azusa Inoue, Ph.D., said, “Much to our surprise, the imprinted genes we looked at lacked DNA methylation, which told us there must be another mechanism at play.” H3K27me3 was found to coat imprinted regions lacking DNA methylation, but only in the maternal gametes, indicating maternal specificity in H3K27me3-mediated imprinting.
The genes associated with this novel form of imprinting were found using H3K27me3 chromatin immunoprecipitation sequencing (ChIP-seq) datasets that compared paternal allele sites that are devoid of DNA methylation but harbored maternal allele-specific H3K27me3. However, Zhang points out that, in contrast to DNA-methylated imprinting markers, “most of the H3K27me3-mediated imprinting is transient, and only a few are maintained in post-implantation embryos and placenta. We currently do not know why this imprinting is mostly transient and why some of them are maintained.”
Previous studies in mice had shown H3K27me3 present at DNA-methylated sites of imprinting, but the histone modification was not thought to be the causal factor in gene silencing. Although this was the first direct evidence of H3K27me3-mediated imprinting in mammals, there has been correlative evidence in marsupials, and Dr. Zhang acknowledges that the mechanism has been shown to be important in the grain plant maize.
The authors also noted that “since loss-of-imprinting of non-canonical imprinted genes is commonly observed in placentae of all cloned mouse embryos, it will be interesting to determine whether defective H3K27me3-mediated imprinting contributes to abnormal placental overgrowth of cloned animals.” This raises potential intrigue in the wider applications of embryonic development research, namely reproductive technologies.
Going forward, the Harvard group seeks to examine whether the new imprinting mechanism is conserved in humans, Zhang explains. “We are collaborating with an in vitro fertilization (IVF) center to analyze discarded human embryos to see whether the related genes are imprinted in humans.”