Detection of pleiotropy through a Phenome-wide association study (PheWAS) of epidemiologic data as part of the Environmental Architecture for Genes Linked to Environment (EAGLE) study.

We performed a Phenome-wide association study (PheWAS) utilizing diverse genotypic and phenotypic data existing across multiple populations in the National Health and Nutrition Examination Surveys (NHANES), conducted by the Centers for Disease Control and Prevention (CDC), and accessed by the Epidemiological Architecture for Genes Linked to Environment (EAGLE) study. We calculated comprehensive tests of association in Genetic NHANES using 80 SNPs and 1,008 phenotypes (grouped into 184 phenotype classes), stratified by race-ethnicity. Genetic NHANES includes three surveys (NHANES III, 1999-2000, and 2001-2002) and three race-ethnicities: non-Hispanic whites (n = 6,634), non-Hispanic blacks (n = 3,458), and Mexican Americans (n = 3,950). We identified 69 PheWAS associations replicating across surveys for the same SNP, phenotype-class, direction of effect, and race-ethnicity at p<0.01, allele frequency >0.01, and sample size >200. Of these 69 PheWAS associations, 39 replicated previously reported SNP-phenotype associations, 9 were related to previously reported associations, and 21 were novel associations. Fourteen results had the same direction of effect across more than one race-ethnicity: one result was novel, 11 replicated previously reported associations, and two were related to previously reported results. Thirteen SNPs showed evidence of pleiotropy. We further explored results with gene-based biological networks, contrasting the direction of effect for pleiotropic associations across phenotypes. One PheWAS result was ABCG2 missense SNP rs2231142, associated with uric acid levels in both non-Hispanic whites and Mexican Americans, protoporphyrin levels in non-Hispanic whites and Mexican Americans, and blood pressure levels in Mexican Americans. Another example was SNP rs1800588 near LIPC, significantly associated with the novel phenotypes of folate levels (Mexican Americans), vitamin E levels (non-Hispanic whites) and triglyceride levels (non-Hispanic whites), and replication for cholesterol levels. The results of this PheWAS show the utility of this approach for exposing more of the complex genetic architecture underlying multiple traits, through generating novel hypotheses for future research.

Lipid trait-associated genetic variation is associated with gallstone disease in the diverse Third National Health and Nutrition Examination Survey (NHANES III).

BACKGROUND: Gallstone disease is one of the most common digestive disorders, affecting more than 30 million Americans. Previous twin studies suggest a heritability of 25% for gallstone formation. To date, one genome-wide association study (GWAS) has been performed in a population of European-descent. Several candidate gene studies have been performed in various populations, but most have been inconclusive. Given that gallstones consist of up to 80% cholesterol, we hypothesized that common genetic variants associated with high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and triglycerides (TG) would also be associated with gallstone risk. METHODS: To test this hypothesis, the Epidemiologic Architecture for Genes Linked to Environment (EAGLE) study as part of the Population Architecture using Genomics and Epidemiology (PAGE) study performed tests of association between 49 GWAS-identified lipid trait SNPs and gallstone disease in non-Hispanic whites (446 cases and 1,962 controls), non-Hispanic blacks (179 cases and 1,540 controls), and Mexican Americans (227 cases and 1,478 controls) ascertained for the population-based Third National Health and Nutrition Examination Survey (NHANES III). RESULTS: At a liberal significance threshold of 0.05, five, four, and four SNP(s) were associated with disease risk in non-Hispanic whites, non-Hispanic blacks, and Mexican Americans, respectively. No one SNP was associated with gallstone disease risk in all three racial/ethnic groups. The most significant association was observed for ABCG5 rs6756629 in non-Hispanic whites [odds ratio (OR) = 1.89; 95% confidence interval (CI) = 1.44-2.49; p = 0.0001). ABCG5 rs6756629 is in strong linkage disequilibrium with rs11887534 (D19H), a variant previously associated with gallstone disease risk in populations of European-descent. CONCLUSIONS: We replicated a previously associated variant for gallstone disease risk in non-Hispanic whites. Further discovery and fine-mapping efforts in diverse populations are needed to fully describe the genetic architecture of gallstone disease risk in humans.

Effect of HIV-1 infection on human DNA yield from saliva.

PURPOSE: Saliva is a good source of DNA for genomic research, and leukocytes are a predominant source of DNA in human saliva. Advanced human immunodeficiency virus (HIV)-type 1 infection disrupts tonsillar architecture and depletes tonsillar lymphocytes. We tested whether HIV-1 infection reduces extracted human DNA yield from saliva. METHODS: Approximately 2 mL of expectorated saliva was collected from HIV-infected adults during routine primary care clinic visits and from healthy, HIV-negative controls. Human DNA was manually extracted and was specifically quantified by assaying for the RNAse P gene. RESULTS: Seventy-five individuals were studied, including 25 HIV-infected adults with <200 CD4+ T cells/mm(3) (i.e., acquired immunodeficiency syndrome), 25 with >200 CD4+ T cells/mm3, and 25 HIV-negative controls. Overall DNA yield was 64.7 microg [29.0-139.7 microg] (median [interquartile range]). Yields were comparable among HIV-infected individuals with lower CD4+ T cell counts (74.3 microg [39.4-151.4 microg]), higher CD4+ T cell counts (63.9 microg [29.2-172.1 microg]), and HIV-negative controls (61.4 microg [28.4-123.4 microg]) (p > .05). CONCLUSION: Infection with HIV-1 does not reduce human DNA yield from saliva. Expectorated saliva should provide sufficient extracted native DNA for genomic studies in HIV-infected individuals.

Whole-genome amplification of oral rinse self-collected DNA in a population-based case-control study of breast cancer.

The availability of large amounts of genomic DNA (gDNA) is the limiting factor for many of the molecular biology assays in genetic epidemiologic studies. Whole-genome amplification using multiple displacement amplification is used to amplify a representative sample of gDNA from small amounts of gDNA to optimize gDNA yield. We collected oral rinse DNA samples through the mail from 3,377 women enrolled in a population-based U.S. breast cancer case-control study and did whole-genome amplification by multiple displacement amplification. Genotyping was done for 66 single nucleotide polymorphisms (SNP) in 18 candidate susceptibility genes using amplified DNA with genomic replicates included for quality control. The concordance rates (percentages of agreement) in 95 quality control replicates of gDNA and amplified DNA for 66 SNPs ranged from 88% to 100% (median, 97%). The average allelic error rate was 0.9%. However, in further analyses based on the full control series (n = 1,492), >60% of the SNPs failed tests for Hardy-Weinberg equilibrium (P < 0.05), with evidence of heterozygote loss in the great majority. Even eliminating the 9% of samples with lower quality or input DNA, tests for Hardy-Weinberg equilibrium indicated persistent allele bias in nearly a third of the SNPs. Whole-genome amplification may introduce substantial allele amplification bias in gDNA collected using a common protocol in population-based epidemiologic studies.

Potential differences in breast cancer risk factors based on CYP1A1 MspI and African-American-specific genotypes.

OBJECTIVES: Recent studies show that an Mspl polymorphism in the 3'-noncoding region of the CYP1A1 gene is associated with breast cancer in African-American women but not in Caucasian women. In addition, an African-American-specific (AAS) polymorphism is located in intron 7 of this gene. We hypothesized that the AAS polymorphism may partially account for this race-specific association and that different environmental risk factor profiles are a function of genotype status. We studied both CYP1A1 polymorphisms to determine if African-American women with these variants have breast cancer risk factor profiles that are different from those of other African-American women. METHODS: A case-control analysis was conducted. Cases were 304 African-American patients pathologically diagnosed with breast cancer from 1995 to 1998 who lived in three Tennessee counties. Controls were 305 African-American women without breast cancer, selected through random-digit dialing and frequency matched to cases by age and county. Information on risk factors was collected through telephone interviews. Tumor tissue samples were collected for CYP1A1 genotyping. There were 215 and 188 cases with the Mspl and AAS polymorphisms measured respectively. RESULTS: Our study results suggest that some risk factors for breast cancer are dependent upon CYP1A1 genotype. Specifically, low intakes of folate, methionine, vitamin C, and vitamin E appear to increase the risk of breast cancer in individuals with the AAS variant: the odds ratio (OR) estimates and 95% confidence intervals were 2.10 and 0.99-4.44 for folate, 1.96 and 0.91-4.23 for methionine, 2.13 and 1.00-4.53 for vitamin C, and 2.43 and 1.12-5.25 for vitamin E. Such associations are stronger for tumors with both AAS and MspI polymorphisms: the OR estimates increased to >6.00 for all these variables except for vitamin C. CONCLUSIONS: This study found that methyl-deficient diets and antioxidant vitamins may be related to the risk of breast cancer as a function of the Mspl and AAS genotpyes. Our results are preliminary because of a small number of cases with polymorphisms at both sites, but they indicate the need for large-scale epidemiologic studies of both African-American and Caucasian women that include genotype information from controls with more detailed information on risk factors.

Phenotype-Driven Plasma Biobanking Strategies and Methods.

Biobank development and integration with clinical data from electronic medical record (EMR) databases have enabled recent strides in genomic research and personalized medicine. BioVU, Vanderbilt's DNA biorepository linked to de-identified clinical EMRs, has proven fruitful in its capacity to extensively appeal to numerous areas of biomedical and clinical research, supporting the discovery of genotype-phenotype interactions. Expanding on experiences in BioVU creation and development, we have recently embarked on a parallel effort to collect plasma in addition to DNA from blood specimens leftover after routine clinical testing at Vanderbilt. This initiative offers expanded utility of BioVU by combining proteomic and metabolomic approaches with genomics and/or clinical outcomes, widening the breadth for potential research and subsequent future impact on clinical care. Here, we describe the considerations and components involved in implementing a plasma biobank program from a feasibility assessment through pilot sample collection.

A multi-investigator/institutional DNA bank for AIDS-related human genetic studies: AACTG Protocol A5128.

BACKGROUND: An understanding of the relationships among allelic variability and clinical outcomes will be critical if HIV-infected patients are to benefit from the explosion in knowledge in human genomics. Human DNA banks must allow future analyses while addressing confidentiality, ethical, and regulatory issues. METHOD: A multidisciplinary group of clinical investigators, ethicists, data managers, regulatory specialists, and community representatives developed Adult AIDS Clinical Trials Group (AACTG) Protocol A5128. Participants in past or present AACTG clinical trials may contribute DNA. Extraction from whole blood is performed at a central laboratory, where participants' unique identifiers are replaced by randomly assigned identifiers prior to DNA storage. To identify genotype-phenotype relationships, genetic assay results can be temporarily linked to clinical trials data. RESULTS: Institutional review boards in 21 states and Puerto Rico have approved Protocol A5128, and accrual is ongoing. Of the first 4,247 enrollees, 82% are male, 56% are white, 26% are African American, and 15% are Hispanic. Because participants may participate in multiple AACTG protocols, these represent 11,424 cases in 324 different AACTG studies and substudies, with at least 100 participants from 24 different studies. Studies exploring specific genotype-phenotype relationships are underway. CONCLUSION: The AACTG DNA bank will be an important resource for genomic discovery relevant to HIV therapy.

Rare variant APOC3 R19X is associated with cardio-protective profiles in a diverse population-based survey as part of the Epidemiologic Architecture for Genes Linked to Environment Study.

BACKGROUND: A founder mutation was recently discovered and described as conferring favorable lipid profiles and reduced subclinical atherosclerotic disease in a Pennsylvania Amish population. Preliminary data have suggested that this null mutation APOC3 R19X (rs76353203) is rare in the general population. METHODS AND RESULTS: To better describe the frequency and lipid profile in the general population, we as part of the Population Architecture using Genomics and Epidemiology I Study and the Epidemiological Architecture for Genes Linked to Environment Study genotyped rs76353203 in 1113 Amish participants from Ohio and Indiana and 19 613 participants from the National Health and Nutrition Examination Surveys (NHANES III, 1999 to 2002, and 2007 to 2008). We found no carriers among the Ohio and Indiana Amish. Of the 19 613 NHANES participants, we identified 31 participants carrying the 19X allele, for an overall allele frequency of 0.08%. Among fasting adults, the 19X allele was associated with lower triglycerides (n=7603; ß=-71.20; P=0.007) and higher high-density lipoprotein cholesterol (n=8891; ß=15.65; P=0.0002) and, although not significant, lower low-density lipoprotein cholesterol (n=6502; ß= -4.85; P=0.68) after adjustment for age, sex, and race/ethnicity. On average, 19X allele participants had approximately half the triglyceride levels (geometric means, 51.3 to 69.7 versus 134.6 to 141.3 mg/dL), >20% higher high-density lipoprotein cholesterol levels (geometric means, 56.8 to 74.4 versus 50.38 to 53.36 mg/dL), and lower low-density lipoprotein cholesterol levels (geometric means, 104.5 to 128.6 versus 116.1 to 125.7 mg/dL) compared with noncarrier participants. CONCLUSIONS: These data demonstrate that APOC3 19X exists in the general US population in multiple racial/ethnic groups and is associated with cardio-protective lipid profiles.