Buch S, Schafmayer C, Völzke H, Becker C, Franke A, von Eller-Eberstein H, et al. A genome-wide association scan identifies the hepatic cholesterol transporter ABCG8 as a susceptibility factor for human gallstone disease. Nat Genet 2007;39:995-999. (Reprinted with permission.)
With an overall prevalence of 10-20%, gallstone disease (cholelithiasis) represents one of the most frequent and economically relevant health problems of industrialized countries. We performed an association scan of > 500,000 SNPs in 280 individuals with gallstones and 360 controls. A follow-up study of the 235 most significant SNPs in 1,105 affected individuals and 873 controls replicated the disease association of SNP A-1791411 in ABCG8 (allelic P value P(CCA) = 4.1 × 10(−9)), which was subsequently attributed to coding variant rs11887534 (D19H). Additional replication was achieved in 728 German (P = 2.8 × 10(−7)) and 167 Chilean subjects (P = 0.02). The overall odds ratio for D19H carriership was 2.2 (95% confidence interval: 1.8-2.6, P = 1.4 × 10(−14)) in the full German sample. Association was stronger in subjects with cholesterol gallstones (odds ratio = 3.3), suggesting that His19 might be associated with a more efficient transport of cholesterol into the bile.
We are currently facing the prospect that genomewide association studies performed in large numbers of individuals genotyped for hundreds of thousands of single-nucleotide polymorphisms (SNPs) will identify important disease-associated variations in the human genome. Whereas a considerable number of genes responsible for Mendelian diseases have been discovered during the last decades, unlocking genetic variants leading to complex traits, such as is found in most hepatobiliary diseases, has been shown to be even more challenging. Crucial to the success of such studies has been the collection and storage of DNA from large numbers of clinically well-characterized individuals and the availability of the complete sequence of the human genome. Furthermore, the completion of the HapMap project for documentation of patterns of genomewide variations and linkage disequilibrium,1 and the development of high-throughput laboratory technologies for parallel genotyping of 500,000 or more SNPs have brought us to this fascinating new avenue of genetic research of common diseases. Whereas single SNP association studies have been notoriously difficult to replicate,2 the analysis of haplotypes of candidate genes has provided more robust genetic information and association data.3, 4 All these approaches are based on the assumption that a specific candidate gene is related to the presence or severity of a clinical condition. Therefore, the major strength of the genomewide genetic approach is that it should permit an “agnostic” comparison of allele prevalence between cases and controls, obviating the need for selection of candidate genes. However, current genotyping platforms represent approximately two-thirds of known genetic variants, encompassing about 20,000 genes, but the scale of genomewide association studies continues to grow, such that the number of SNPs will soon exceed a million (Fig. 1).
With this armament in place, Buch and co-workers have now succeeded in identifying an important genetic variant contributing to gallstone disease. They performed a genomewide association scan with 500,000 SNPs in 280 Caucasian individuals with gallstones and 360 matched controls. In follow-up studies, the investigators replicated the disease association of the nonsynonymous SNP rs11887534 in the gene for the cholesterol transporter ABCG8 in 3 further independent cohorts of patients from different genetic backgrounds.5 This replication of the initial finding is a particular strength of the current study and is considered as a conditio sine qua non for the belief in the results of a genomewide association study by some geneticists.6 Notably, the same coding SNP (D19H) in ABCG8 has also been identified by Grünhage et al.7 in a nonparametric linkage analysis of affected sib pairs. This consistency across different genetic study designs underscores the validity of the genetic findings. Because the heterodimer ABCG5/G8 is the principal cholesterol transporter across the canalicular membrane and it has been identified as a susceptibility locus in inbred mice,8 it is apparent to ask why a whole-genome scan was necessary to identify this obvious candidate gene?
One of the largest genomewide association studies to date has been published very recently. In this joint investigation by the Wellcome Trust Case Control Consortium, more than 2,000 cases of 7 major complex diseases (type 1 diabetes, type 2 diabetes, Crohn's disease, coronary artery disease, rheumatoid arthritis, hypertension, bipolar disorder) and 3,000 controls were genotyped for 500,000 SNPs.9 As in the study by Buch et al., in this joint study a number of known and obvious candidate genes were identified (for example, APOE for coronary artery disease, INS for type 1 diabetes and HLA-DRB1 for rheumatoid arthritis). Thus, genomewide association studies are important tools not only to investigate human biology unbiased by previous hypothesis or assumptions, but also to confirm earlier findings from smaller scale studies.
What else can we expect to learn from “hypothesis-free” genomewide association studies for clinical hepatology?
Although the study of Buch et al. demonstrates that the genetic gallstone disease risk is attributable to the variant D19H in the ABCG8 gene, it remains unclear how this genetic variant affects the function of the ABCG5/G8 heterodimer and how the difference in function leads to gallstone disease. Therefore, subsequent (functional) studies are clearly required to address these important questions in order to develop therapeutic strategies addressing ABCG8. The same is true for 2 published large association studies with ≈25,000 SNPs in patients with hepatitis C–induced liver fibrosis.10, 11 Although a number of new genes were identified in these studies which have not been associated with liver fibrosis before, the biological significance of these genes remains to be proven. Notably, the published genomewide association studies also show that variations leading to common diseases are diverse; some alter the coding sequence of genes (for example, the ABCG8 D19H SNP), others lie in noncoding regions, and some are even located within chromosomal regions containing no apparent genes. Therefore, understanding the biological function of disease-associated genomic regions is still a challenging task, and other aspects of genetic diversity, such as copy number variations and deletion polymorphisms,12, 13 have to be taken into account. Furthermore, different from high individual risks associated with single gene disorders, candidate alleles identified in genomewide association studies are likely to be associated with moderate overall disease risks. However, if a genetic risk factor is common in a population, it accounts for a substantial fraction of cases in the given population and might therefore be a suitable candidate for large-scale intervention efforts. Accordingly, the population-attributable fraction of the lithogenic ABCG8 variant has been estimated to be 8%-11%.5, 7
Taken together, genomewide association studies appear as a fascinating technique that opens a new dimension for the search of needles in the haystack. However, the results of such genomewide studies tell us that a particular set of genes is important in the pathogenesis of a clinical condition, but they do not tell us the mechanisms that are responsible. Therefore, subsequent complementary methods are still required to form a conceptual framework with clinical relevance. The coming years will be exciting as more and more genomewide association studies of common benign and malignant diseases of the liver and the biliary system will be published. However, it is equally important for the field that the clinical characterization of patients and controls as well as the interpretation and application of the results will be practiced with the same strictness and enthusiasm as the new genotyping technologies.