| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
EDITORIAL |
1 From the Department of Genetics, Southwest Foundation for Biomedical Research, San Antonio, TX
See corresponding article on page 509.
The promise of identifying the functional genetic variants underlying the expression of complex phenotypes, including those associated with common diseases such as obesity, diabetes, and heart disease, has rapidly moved closer to realization since completion of the Human Genome Project. Advances in molecular genetic technology now make large-scale resequencing efforts increasingly viable (1, 2), and these efforts are revealing extensive variation in genes of potential interest to biomedical researchers. With access to this new wealth of genetic information, some of the assumptions that underpinned earlier efforts at gene mapping have begun to be challenged. For most of the past decade, a generally accepted assumption was that variations observed in complex phenotypes were the result of common genetic variants with fairly substantial effects, and whereas that does seem to be the case in some instances—such as that of lipoprotein(a)—we are beginning to see that it is not the universal pattern. Using the extensive genetic information now obtainable through intensive resequencing of candidate genes (often first identified on the basis of linkage results) and applying recent advances in analytic methods, numerous groups have begun to show that, in many cases, the observed effects actually result from the acting in concert of multiple variants within a gene, even though those variants are not likely to display a detectable effect when considered individually. Such results have recently been reported for variants in the factor VII structural gene (F7) that explain a quantitative trait locus for factor VII clotting concentrations (3) as well as promoter variants in the selenoprotein S gene (SELS) that mediates plasma cytokine concentrations (4). These resequencing efforts have also identified another deviation from our expectations: the pattern of linkage disequilbrium that is the foundation of association analysis may not always be as predictable as expected, particularly within genes. In the case of F7, there are 48 common single-nucleotide polymorphisms (SNPs) that behave statistically like 39 independent variants with respect to linkage disequilibrium (3). What this means from a practical standpoint is that identification of the state of one of these SNPs does not offer any information about the state of the others. We have also observed similar situations in other genes such as pro-opiomelanocortin (POMC) and adiponectin (ACDC). As a result of these findings, it is becoming increasingly clear that the analysis of a few variants per gene is not likely to be sufficient to identify potential functional effects, and, indeed, that it could even lead to false-negative conclusions about the potential contribution of a gene.
Since the discovery of adiponectin approximately a decade ago, there has been an explosion of work focused on the potential role of this adipokine in obesity, type 2 diabetes, heart disease, and the metabolic syndrome. Multiple lines of investigation—including epidemiologic studies showing both an inverse relation of circulating adiponectin concentrations to obesity and its comorbidities and a positive relation between adiponectin concentrations and insulin sensitivity and biochemical and mouse transgenic and knockout studies—have established adiponectin as a key player in the integration of processes that affect the cluster of traits and diseases that make up the metabolic syndrome.
Circulating concentrations of adiponectin have been found to be heritable in humans. Genetic linkage studies in white, Pima Indian, Hawaiian Japanese and Chinese, and Amish adults and in Mexican American children have identified multiple loci in the genome that appear to affect circulating concentrations of adiponectin (5, 6). However, none of the studies have found significant evidence of linkage of adiponectin concentrations to the chromosome 3q27 region where the structural gene ACDC (also called APM1 or ADIPOQ) is located. However, 
26 published studies have investigated the possible effect of genetic variation in ACDC itself on adiponectin mRNA and protein concentrations as well as on intermediate phenotypes that lead to the development of these diseases (reviewed in 7). Evidence from these studies in multiple populations shows significant associations between ACDC variants and obesity-related traits and diseases. In this issue of the Journal, Yang et al (8) report evidence of associations between several SNPs in ACDC and various obesity- and diabetes-related phenotypes, this time in an elderly Chinese cohort. This study is unique, because, despite the efforts of the Human Genome Project and the identification of almost 100 polymorphisms in human ACDC that have been cataloged in the dbSNP database (Internet: http://www.ncbi.nlm.nih.gov/SNP/index.html), <20 polymorphisms have been used in genetic epidemiologic candidate gene–association studies, and <10 of these polymorphisms result in nonsynonymous amino acid changes. Very little, if anything, is known about the potential functional significance of any given ACDC polymorphism.
It is now time for investigators to proceed to the types of study designs that will not only identify associations between genetic markers and disease phenotypes but will also help define thephysiologic consequences of the various genetic polymorphisms. There is now a need to associate a particular polymorphism or variant with a phenotype of interest, not just with a genetic locus. Ultimately the question is not whether adiponectin is a relevant player in metabolic disease, because there is ample evidence to support that claim. The pressing questions are 1) how does adiponectin exert its effects, and 2) how do genetic variants within the adiponectin gene influence the gene's ability to exert these effects? In the case of adiponectin, the time has clearly come to move beyond association studies to functional analyses. Whereas the identification of well-supported candidate genes such as adiponectin is a necessary first step in genetic studies, in our efforts to identify the underlying genetic mechanisms contributing to variation in our phenotypes of interest, we would do well to heed the words of Winston Churchill: "Now this is not the end. It is not even the beginning of the end. But it is, perhaps, the end of the beginning."
ACKNOWLEDGMENTS
Neither of the authors had a personal or financial conflict of interest with respect to the study by Yang et al or the authors of that study.
REFERENCES
Related articles in AJCN:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
