The role of non-coding RNAs in epigenetic regulation of gene expression

Although we now have the entire genomic sequence for humans and several other organisms we still know little about how genes are actually controlled. Traditionally genes are defined as the genetic units that encode proteins and have been considered to be the only meaningful parts of the genome. About 98% of the genomic DNA is non-coding sequence and has often been referred to as 'junk' DNA, containing meaningless 'spacer' sequences. Genes are transcribed into mRNA molecules, which transport genetic information from the gene to the cytoplasm of cells where the genetic code is translated into a functional protein molecule with enzymatic or structural roles. We normally think of transcription of protein-coding genes as the sole purpose of the genome, but in fact recent research shows that gene transcription makes up only a very small percentage of the transcriptional activity of the genome. In fact the vast majority of transcribed genomic regions do not have coding potential. Many of these non-coding or intergenic transcripts are highly unstable or rare and their function is being investigated. Other non-coding transcripts stably accumulate and appear to function in regulating gene expression over wide areas of the genome. For example the Xist RNA is transcribed from the X chromosome. Female cells have two X chromosomes whereas males have one X and one Y chromosome. This doubling in the amount of X chromosome genes in female cells is a potential problem that could lead to an imbalance in the amount of hundreds of gene products in female cells. However, the Xist RNA has evolved to even the score. The Xist RNA appears to coat one of the X chromosomes in female cells leading to the complete inactivation of nearly all genes on that X chromosome. Little is known about how Xist achieves this feat. More recently other non-coding RNAs have been discovered which appear to be functional molecules. The Air RNA is a large non-coding RNA that appears to be necessary to silence a small cluster of imprinted genes. Most imprinted genes are involved in growth control and are unusual because unlike most genes, expression of an imprinted gene is dependent on who you inherited it from. We all have two copies of every gene in our genome, one from our mother and one from our father. Normally both genes are expressed but imprinted genes are either expressed or silenced depending on which parent they came from. Our work and the work of others has suggested that the Air RNA silences a cluster of genes by spreading over them and coating them, possibly in a very similar way to Xist control of the X chromosome. Though Xist was once thought to be an oddity, it now appears that other functional non-coding RNAs may operate in a similar way. The fact that Air operates over a fairly small region of the genome compared to Xist which covers a vast area encompassing an entire chromosome, makes Air function more amenable to investigation. We will attempt to identify the sites of Air interaction with the imprinted gene cluster that it controls. Another non-coding RNA that also appears to be functional and control a cluster of imprinted genes is the Kcnq1ot1 RNA. We will perform similar experiments on this RNA and the Xist to obtain evidence on their mechanisms of action. It is highly likely that many more functional RNAs exist that play important but, still unappreciated roles in the regulation of gene expression. The three RNAs mentioned in this proposal regulate hundreds of genes between them. These experiments will provide important insights into the regulation of the genome with a significant impact on human health.

Non-coding RNAs are emergent players in an expanding number of roles in epigenetic regulation of gene expression, though in many cases their precise function remains enigmatic. The pre-eminent example of one class of functional RNAs is the non-coding Xist RNA which is essential for random X chromosome inactivation in mammalian female cells. Xist is a 17 kb RNA encoding by the Xist gene on the X chromosome. In undifferentiated ES cells Xist gene expression is biallelic. During the first few days of ES cell differentiation Xist expression is downregulated on the future active X chromosome and Xist RNA from the future inactive X is upregulated. Xist proceeds to 'coat' or 'paint' the future inactive X leading to the recruitment of chromatin modifying complexes, which epigenetically silence gene expression from the entire chromosome. Little is known about the identity of these complexes and even less is known about how Xist spreads over the entire chromosome or which specific sequences it interacts with. Two other non-coding RNAs that have recently gained precedence are the Air and Kcnq1ot1 (Lit1) RNAs. Both RNAs are expressed from clusters of imprinted genes. Air is 108 kb in length and transcribed in the antisense direction from an intronic promoter within the Igf2r gene (insulin-like growth factor II receptor) on mouse chromosome 17. Similarly Kcnq1ot1 is approximately 100 kb in length and is transcribed in the antisense direction from an intron of the Kcnq1 gene (Potassium voltage-gated channel subfamily KQT member 1) on mouse chromosome 7. Both are expressed from the paternal allele only and appear to be necessary to silence several other imprinted genes in their respective clusters. In this project we will characterize the behaviour of the Air and Kcnq1ot1 RNAs in tissues in which they are required for imprinted expression of their neighboring genes and in tissues without imprinted expression of genes in their clusters. Our preliminary data suggests that Air may silence imprinted genes in cis via a spreading or coating mechanism, similar to Xist in X inactivation. The relatively small size of the genomic region over which Air exerts its influence compared to Xist, makes it an ideal system to study possible interactions at the molecular level. We will attempt to identify specific genomic sequences through which Air and Kcnq1ot1 exert their control over imprinted gene clusters. We will perform similar experiments on the Xist RNA to identify its sites of interaction with regions of the inactive X chromosome in female mouse cells. Together these three non-coding RNAs epigenetically regulate hundreds of genes. Given that it is now recognized that non-coding transcripts represent the clear majority of the transcribed regions of mammalian genomes, it is likely that non-coding RNA control is a much more widespread phenomenon in epigenetic regulation than previously appreciated.