In the spring of 1984, the groups of Azim Surani and Davor Solter published landmark papers that reported the phenomenon of imprinting of the genome in mammals (Surani et al.; McGrath & Solter). In both studies highly skilled manipulations were performed to transfer maternal or paternal chromosome sets in a fertilized mouse egg before pronuclear fusion to construct ‘paternal-only’ or ‘maternal-only’ embryos (androgenetic and gynogenetic). The reconstructed ‘monoparental’ embryos could not survive to term. Looking back, whether you need the genomes of both parents for successful embryo development might seem an esoteric question, but there was a need to test reports, subsequently discredited, that viable mice could be produced from eggs ‘parthenogenetically’ activated in the absence of sperm.
Following these studies, as well as others from Bruce Cattanach and colleagues, genomic imprinting became a paradigm in the growing discipline of epigenetics and helped energize a whole body of research in mammalian developmental epigenetics. In the years that followed, and with the identification in 1991 of the first endogenous imprinted genes (defined as parent-of-origin-dependent monoallelic expression), we came to understand the principles governing imprinting: the role of germline DNA methylation; imprinting control regions that can specify imprinting of whole domains of genes in cis; long non-coding RNAs as agents of epigenetic modification; and the existence of long-term epigenetic memory transmitted from one generation to the next. And we enjoyed stimulating discussions about the evolutionary rationale for imprinting: what is the selective advantage of purposefully silencing one of the two alleles of these genes?
“genomic imprinting became a paradigm in the growing discipline of epigenetics”
Although DNA methylation has been known to specify imprinting since the 1990s, the possibility that histone modifications in the germ line could also determine imprinting was only demonstrated recently. In 2017, experiments performed by Yi Zhang’s group, which were every bit as elegant as the Surani and Solter studies and deployed some of the same embryo manipulations, identified genes whose imprinting depends on the repressive histone modification histone 3 lysine 27 trimethylation (H3K27me3) in the oocyte (Inoue et al.). The authors were able to exploit recent advances in genome-wide expression and epigenomic profiling methods to enable them, in a single study, to identify genes imprinted by H3K27me3 and intervene directly in their epigenetic regulation. Although some of the principles of this other mode of imprinting are beginning to be elucidated, there is much more to be discovered, and the question of why it should differ in many key respects — its restriction to the placenta, the properties of the control elements and its conservation — from DNA methylation-dependent imprinting remains.
Surani, M. A., Barton, S. C. & Norris, M. L. Development of reconstituted mouse eggs suggests imprinting of the genome during gametogenesis. Nature 308, 548–550 (1984)
McGrath, J. & Solter, D. Completion of mouse embryogenesis requires both the maternal and paternal genomes. Cell 37, 179–183 (1984)
Inoue, A. et al. Maternal H3K27me3 controls DNA methylation-independent imprinting. Nature 547, 419–424 (2017)
The author declares no competing interests.
About this article
Cite this article
Kelsey, G. Imprints in the history of epigenetics. Nat Rev Mol Cell Biol 21, 566–567 (2020). https://doi.org/10.1038/s41580-020-00289-8