Long Range Cis-Silencing By Imprinted Macro IncRNAs

Imprinted genes are clustered in the mammalian genome, in most cases associated with a macro long non-coding (lnc) RNA. LncRNAs are non-protein coding transcripts longer than 200 nucleotides that are increasingly implicated in transcriptional regulation, while macro lncRNAs form an inefficiently spliced sub-class where the unspliced RNA is the functional form. In three of four tested cases the imprinted macro lncRNA has been shown to be responsible for silencing of all imprinted genes in a cluster. Imprinted expression can be limited to specific developmental stages and tissues. A distinct class of imprinted genes, that tend to be further away from the regulating macro lncRNA than other imprinted genes, show imprinted expression limited to certain extra- embryonic lineages (EXEL) from the placenta and visceral yolk sac (VYS). This proposal uses regulation of EXEL genes as a model system to investigate long-range cis-silencing by macro lncRNAs. At the Igf2r imprinted cluster the macro lncRNA Airn causes the silencing of Igf2r and two EXEL genes on the paternal allele. I have shown that the active EXEL gene promoters interact with elements within the Airn lncRNA gene body only on the maternal allele where Airn is not expressed. These interacting elements contain VYS- specific marks associated with enhancers. A hypothesis consistent with the published data and my results is the enhancer interference hypothesis, where transcription of the macro lncRNA blocks enhancer-promoter interactions preventing upregulation of EXEL genes. To test this hypothesis, this proposal will first examine the Igf2r cluster, producing a precise map of the enhancers in the Airn region, testing if these enhancers are dependent on Airn, and if they show the expected expression in extra-embryonic tissues in an enhancer assay. In the second part, I will test the predictions of the enhancer interference hypothesis in all other imprinted clusters, taking a genome-wide approach to identify putative enhancers within imprinted macro lncRNAs, and then testing for interactions between these enhancers and EXEL promoters. The outcome of this study will increase understanding of the mechanism of imprinted macro lncRNA silencing, and may indicate a new mechanism how lncRNAs in general can silence genes.

The same set of genetic information is found in every cell in the body, but only a specific subset of genes are turned on in any given tissue. This tissue-specificity is controlled by distant genetic elements called enhancers that interact with the gene to turn it on. Normal genes are transcribed into an RNA message that is then translated by an enzyme machine to assemble a protein. Long non-coding RNAs are unusual genes that do not code for a protein, but may be involved in regulating normal genes.Genes have a maternal and paternally inherited copy, and in most cases both copies are either turned on or off. A small number of genes display an unusual pattern where only one copy is turned on, due to so-called imprinted silencing of the other copy. This imprinted silencing is often controlled by a lncRNA and tissue-specific, indicating that the lncRNA may act by blocking enhancers. To investigate this, first, we generated a comprehensive map of imprinted genes in the mouse and a map of active and repressive marks on the genome, and then compared the two to determine the relationship between active enhancers, imprinted genes and lncRNAs. Second, as a model system we used genetic mutants to test if the lncRNA Airn causes imprinted silencing of a cluster of neighbouring genes by interfering with enhancer action.Imprinted genes occur together in clusters, and our genome-wide approach enabled us to show that the size of these clusters vary dramatically during development. This indicated that that imprinted silencing acts on tissue-specific enhancers, which was supported by a correlation between the size of the imprinted cluster and region with active imprinted enhancers marks. At the Airn lncRNA gene we found chromosome interactions between potential enhancers in the gene and neighbouring imprinted genes only on the active copy where Airn is turned off, indicating that Airn lncRNA may cause imprinted silencing by blocking interactions between enhancers and imprinted genes. To test this we created a large genetic deletion that included the Airn gene and enhancers. This deletion did not affect imprinted genes on the active copy, but led to the loss of imprinted silencing on the normally silent copy, indicating that the Airn lncRNA has to be turned on for silencing to occur, and that there are no essential enhancers for imprinted genes in the Airn gene. We propose that initially the observed chromosome interactions may exist on both copies and enable the Airn lncRNA to find its targets, and that these interactions are then lost on the silent copy due to the deposition of repressive marks.