Genomic imprinting and the epigenetic control of genome function: regulation, redundancy and resilience

Lead Research Organisation: University of Cambridge
Department Name: Genetics

Abstract

We are interested in understanding mechanisms of imprinted expression. What makes an imprinted gene be expressed from one of the chromosomes in a pair rather than both of them? The process causing only the paternally inherited or the maternally inherited copy to be expressed, is an 'epigenetic' one. Epigenetics means 'on top of genetics' and the epigenetic state at an imprinted gene is manifested as chemical modifications that sit on the DNA at only one parental chromosome and not the other causing the gene located there to be expressed on one of the chromosomes and kept off on the other. Our goal is to generate new knowledge about what the two parental chromosomes look like at an imprinted domain and how that influences one parental copy being on and the other being off.

In our first aim we will compare the chromosome inherited from dad, with the chromosome inherited from mum, at an imprinted domain. This will allow us to identify if the two parentally inherited chromosomes are packaged differently from each other and in turn, whether this is a cause or a consequence of the expression or repression of the genes in the domain being investigated. In our second aim, we study a very important imprinted gene predominantly in the brain, to try and understand more about human diseases that are associated with abnormalities in this gene. Finally, we have recently discovered that the preimplantation embryo is able to 'repair' lost DNA methylation. This resilience to methylation loss has implications for offspring health and well being. In our third aim, we will characterise the sequences where methylation can be restored and the machinery that does the restoration. In particular, using the mouse as a model, we will investigate whether this process is compromised in obese mothers or aged mothers leading to abnormal methylomes in their offspring with implications for their life long health.

Technical Summary

Gene expression is co-ordinately controlled by intricate communication between different genomic elements in each cell type. The Dlk1/Gtl2 imprinted region is an excellent model to study the relationship between epigenetics, topology, and gene expression because the two parental copies behave differently from each other within the same cell. Imprinting of the domain is achieved via the tightly regulated, hierarchical activity of epigenetically controlled genomic elements. Comparing the two parentally inherited alleles enables us to identify interactions that are important for expression and repression. In the first aim, using the Dlk1-Gtl2 domain as a model, we dissect the hierarchical relationships between regulatory elements and their epigenetic states, differential transcription from the two parental chromosomes, the function of a long non-coding RNA in these interactions, and the formation and function of the boundaries of a parental origin-specific topological domain.

Imprinted genes can exhibit sophisticated patterns of tissue-specific imprinting. In particular, Dlk1 shows selective absence of imprinting in the postnatal neurogenic niche and this switch from canonical imprinting to a biallelic state provides a dosage that is crucial for postnatal neurogenesis. The mechanisms regulating this are not understood. In the second aim we focus on this biomedically important gene, and its non-imprinted paralogue (Dlk2) in the regulation of pre and postnatal neural development, assessing functional reduncancy, novel concepts in regulatory cross-talk between genes and mechanisms of dosage control.

Finally. we have identified a novel methylation surveillance/repair system acting very early after fertilisation. In the third aim we will characterise this system and determine whether it is compromised within the milieu of older oocytes, or in eggs from females exposed to a high fat diet with implications for intergenerational inheritance of faulty methylation.

Publications

10 25 50