COMMENTARIES

An Intriguing Genetic Association between a Tumor Suppressor Locus, Coronary Artery Disease and Diabetes

Martin Farrall, Department of Cardiovascular Medicine, University of Oxford, The Wellcome Trust Centre for Human Genetics, Roosevelt Drive, Oxford, OX3 7BN, UK, E-mail: Martin.Farrall@cardiov.ox.ac.uk

Cases of coronary artery disease (CAD) often cluster in families which, backed up by concordance studies of monozygous and dizygous twins, supports the hypothesis that CAD is a multifactorial condition with an important heritable component. Researchers have subsequently expended much effort in attempts to map specific variants that influence susceptibility to CAD, mostly through the application of the “candidate gene” paradigm. Recent advances in genotyping technology that facilitate the study of hundreds of thousands of genetic markers allow geneticists to greatly expand the scope of their searches for susceptibility genes. This work came to the boil last year as a pioneering series of genome-wide association (GWA) studies were completed. Three studies of CAD risk were published in quick succession each identifying a striking genetic association to chromosome 9p [1-3]. This result was particularly exciting as the statistical support was both very strong and consistent across several populations belaying any doubts regarding the replicability of the locus. Bioinformatic analysis revealed no candidate genes in biological pathways with an obvious role in atherosclerosis hinting that this assignment might eventually provide novel insights into the pathophysiology of CAD. Moreover, contemporaneous GWA studies of type 2 diabetes (T2D) risk identified a T2D association signal that roughly coincided with the CAD locus [4]; although curiously the strongest associated single nucleotide polymorphism (SNP) for T2D was apparently unassociated with CAD risk.

Our group (www.PROCARDIS.org) is using genome-wide and other association mapping techniques to study CAD susceptibility genes in four European populations and we have recently studied the chromosome 9p locus in our samples [5]. By using an extended panel of SNPs that includes markers reported to show CAD or T2D association (“literature” SNPs) as well as haplotype-tagging SNPs (www.hapmap.org), we were able to convincingly confirm the association with CAD (p < 10^-12). The magnitude of the association (odds ratio = 1.3) was similar in each population (Germany, Italy, Sweden, and UK) and haplotype analysis showed that a “yin-yang” haplotype (two “perfectly” complementary haplotypes) spanning 53 kb could explain all of the susceptibility effect. Although geneticists are usually wary of subgroup analyses (as they are notorious for generating false-positive results), we argued that given the overwhelming statistical strength of the association, that this could be useful here to understand the susceptibility effect in the context of conventional risk factors. As it turned out, we found only small differences in CAD risk for subgroups defined by clinical diagnosis or other risk category (CAD patients suffering myocardial infarction versus those with other CAD diagnoses, co-morbidity with diabetes or hypertension, older versus. younger CAD patients, male versus female patients). Analyses of several quantitative intermediate phenotypes (e.g. LDL-cholesterol) also showed no convincing association. Together these results suggest that the chromosome 9p locus does not act through a pathway that is strongly associated with conventional risk factors. But this conclusion seems at odds with the GWA result showing an overlapping chromosome 9p association with T2D being mindful of the strong pathogenic link between these two diseases. To address this important point, we fitted a statistical model to simultaneously assess CAD and diabetes risk with two genetic predictors: 1) rs2891168, a SNP that falls within the 53 kb CAD-risk haplotype and 2) rs10811661, the strongest “literature” T2D marker which maps outside of the CAD –risk haplotype. This analysis showed that the CAD and diabetes susceptibility effects were apparently statistically independent of each other, leaving us with the awkward conclusion that neighboring genetic variants might influence atherosclerosis and T2D through unrelated (or weakly related) biological pathways!

Our bioinformatic and literature-search analysis uncovered a newly annotated gene, ANRIL (antisense noncoding RNA in the INK4 locus) that includes exons within the 53 kb CAD-risk haplotype [6]. ANRIL, together with its neighbors CDKN2B, CDKN2A, and ARF, is deleted in a French family with multiple melanoma-neural system tumors and collectively are proposed as constituting a candidate tumor suppressor locus [6]. We studied ANRIL using RT-PCR and found it to be expressed in several cell types and tissues subject to atherosclerosis. Of particular note, ANRIL was expressed in abdominal aortic aneurysm samples; others have found a genetic association between the chromosome 9p locus and abdominal and cranial aneurisms [7].

The identification of a convincing CAD susceptibility locus on chromosome 9p confirms the promise of genetic strategies to suggest new insights into the pathophysiology of complex diseases. Such new knowledge may be challenging for those who want to rapidly design definitive experiments to identify functional variants.  At present there are several perplexing questions concerning the respective roles of the genes within the tumor suppressor locus and how they might contribute to CAD (and T2D) susceptibility. As little is known at present about the function of large antisense RNA molecules such as ANRIL, we can only speculate that they contribute to the regulation of cis and trans-acting targets and suspect that considerable experimental effort will be required to make inroads into these possibilities. Finally, the chromosome 9p CAD locus provides support for the common disease/common variant hypothesis as it is both common (allele frequency ~ 50%) and low penetrance (increase in risk per allele ~30%). The simple genetic architecture in Europeans with a frequent CAD-risk haplotype spanning 53 kb defined by dozens of SNPs means that fine genetic mapping will be of limited value; analysis in populations with different ancestries could prove informative.

References

1.    McPherson R, Pertsemlidis A, Kavaslar N, et al. 2007. A common allele on chromosome 9 associated with coronary heart disease. Science 316: 1488-91.
2.    Helgadottir A, Thorleifsson G, Manolescu A, et al. 2007. A common variant on chromosome 9p21 affects the risk of myocardial infarction. Science 316: 1491-93.
3.    Wellcome Trust Case Control Consortium. 2007. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 447: 661-78.
4.    Zeggini E, Weedon MN, Lindgren CM, et al. 2007. Replication of genome-wide association signals in UK samples reveals risk loci for type 2 diabetes. Science 316: 1336-41.
5.    Broadbent HM, Peden JF, Lorkowski S, et al. 2007. Susceptibility to coronary artery disease and diabetes is encoded by distinct, tightly linked, SNPs in the ANRIL locus on chromosome 9p. Hum Mol Genet [Epub ahead of print].
6.    Pasmant E, Laurendeau I, Heron D, Vidaud M, Vidaud D, Bieche I. 2007. Characterization of a germ-line deletion, including the entire INK4/ARF locus, in a melanoma-neural system tumor family: identification of ANRIL, an antisense noncoding RNA whose expression coclusters with ARF. Cancer Res 67: 3963-69.
7.    Helgadottir A, Thorleifsson G, Magnusson KP, et al. 2008. The same sequence variant on 9p21 associates with myocardial infarction, abdominal aortic aneurysm and intracranial aneurysm. Na. Genet 40: 217-24.