News

Numerous genome-wide association (GWA) studies have shown that single nucleotide polymorphisms (SNPs) in a region of chromosome 8q24 confer an increased susceptibility to both colorectal and prostate cancers. Resequencing of the region suggests that one particular SNP (rs6983267) may be the causal variant; the high-risk G allele of this SNP has a frequency of around 50% in individuals of European decent and 100% in populations of African origin, and confers a relative risk of around 1.2 in heterozygotes and around 1.5 in homozygotes. Like many susceptibility loci revealed through GWA studies, the SNP is located in a so-called gene desert, a region of the genome that contains no genes. Therefore, the biological mechanism underlying these associations has thus far been unclear.

 

Two new studies, published back-to-back in the journal Nature Genetics, use numerous complementary functional genomics approaches to uncover the role of 8q24 in the development of cancer. The first analyses the effect of rs6983267 upon the expression of the closest gene, which lies around 335 kilobases away and encodes the oncogenic transcription factor MYC [Pomerantz M. et al (2009) Nat Genet 41(8);882-4]. The team showed that, although the region containing rs6983267 is largely devoid of transcription, that the DNA itself physically interacts with the MYC gene and its promoter. Moreover, when this region of DNA was linked to a reporter gene, the G allele significantly increased the expression of the gene relative to the T allele. Both of these pieces of evidence, together with an analysis of the epigenetic modifications around 8q24, are consistent with the region acting as an enhancer, i.e. a stretch of DNA that literally enhances the transcription level of a gene (or genes) outside that region, often through binding and co-localising various proteins that are required for transcription.

 

The second study reports that genomic rearrangements only occur around the 8q24 region when the high-risk G-allele of rs6983267 is present [Tuupanen S et al. (2009) Nat Genet 41(8):885-90]. This results in a preferential copy number increase of the region during the development of colorectal cancer. They further showed that rs6983267 is part of a conserved binding site for TCF4, a transcription factor which is activated in most colorectal cancers, and that the G-allele substantially increases the affinity of this protein for the site. Therefore, through a complex set of signalling pathways, rs6983267 probably affects the expression of numerous genes and thus could be directly causal in the development of cancer.

 

Comment: The accompanying editorial suggests that these papers might offer a “general methodology to move from association to function” [Harismendy O & Frazer KA (2009) Nat Genet 41(8):868-9]. Following the identification of genetic variants that are associated with a particular disease (or trait) through GWA studies, functional characterisation of the associated genomic region is undertaken using a two-pronged approach: firstly, a bioinformatics analysis to assess the cross-species sequence conservation of the region and look for any predicted transcription factor binding sites; and secondly, laboratory experimentation to detect epigenetic modifications and assess the accessibility of the DNA in that region, as well as measure the activity of the region using in vitro and in vivo assays. Whilst this method does little to either support or refute the use of such genetic variants in susceptibility testing (see previous news), it will undoubtedly improve our understanding of gene regulation and the causal pathways underlying common disease.