Publications

Recent studies reported a marked inverse effect of smoking on serum levels of leptin (an adipocyte derived protein), offering a possible explanation for variation in body weight between smokers and non-smokers. The goal of this study was to examine the genetic architecture of the response to smoking in leptin levels using data from the San Antonio Family Heart Study. We employed a variance decomposition analysis using maximum likelihood methods to model genotype by smoking interactions for leptin levels, examined the impact of the exclusion of smokers in a subsequent linkage analysis, and incorporated the QTL identified in the linkage analysis in a model of genotype by smoking interaction. We found significant evidence (P = 0.001) for a genotype by smoking status interaction for serum leptin levels. In the subsequent linkage analysis with smokers excluded, we obtained a maximum LOD score of 3.1 (P = 0.00008) near D8S1102. Using this QTL in a model of genotype by smoking status interaction, we identified significant evidence for an interaction at this specific locus (P = 0.04). Given these results, we hypothesize that a quantitative trait locus in this vicinity of chromosome 8 may have a differential effect on the expression of leptin in smokers versus non-smokers.

OBJECTIVE: Previous studies have reported modest associations between measures of obesity and the Trp64-Arg variant of the beta3-adrenergic receptor (ADRbeta3) and the Pro12Ala variant of the peronisome proliferator-activated receptor (PPAR)-gamma2. We hypothesized that these single gene variants may mark mutations that act through convergent pathways to produce synergistic effects on obesity.

RESEARCH DESIGN AND METHODS: The sample included 453 subjects from 10 large Mexican-American families participating in the population-based San Antonio Family Heart Study. The effects of each gene variant singly and jointly were estimated as fixed effects using the measured genotype approach framework. Analyses were conditioned on the pedigree structures to account for the correlations among family members. Statistical significance was evaluated by the likelihood ratio test with adjustment for age, sex and diabetes status.

RESULTS: The allele frequencies for the ADRbeta3 Trp64Arg and PPARgamma2 Pro12Ala variants were 18 and 12%, respectively. The ADRbeta3 variant was not significantly associated with any of the obesity-related traits, but subjects with the PPAR-gamma2 variant (n = 98) had significantly higher levels of lasting insulin (P = 0.03), leptin (P = 0.009), and waist circumference (P = 0.03) than those without. Subjects with the gene variants (n = 32) had significantly higher BMI, insulin, and leprtin levels than those with only the PPARgamma2 variant (n = 66) (P for interaction: 0.04, 0.02, and 0.01 for BMI, fasting insulin, and leptin, respectively).

CONCLUSIONS: Our results suggest that epistatic models with genes that have modest individual effects may be useful in understanding the genetic underpinnings of typical obesity in humans.

Hyperinsulinemia predicts the development of type 2 diabetes, and family studies suggest that insulin levels are regulated in part by genes. We conducted a genome-wide scan to detect genes influencing variation in fasting serum insulin concentrations in 391 nondiabetic individuals from 10 large multigenerational families. Approximately 380 microsatellite markers with an average spacing of 10 cM were genotyped in all study subjects. Insulin concentrations measured by radioimmunoassay were transformed by their natural logarithms before analysis. In multipoint analysis, peak evidence for linkage occurred on chromosome 3p approximately 109 cM from pter in the region of 3p14.2-p14.1. The multipoint logarithm of odds (LOD) score was 3.07, occurring in the region flanked by markers D3S1600 and D3S1285 (P value by simulation <0.0001). In a two-point analysis, LOD scores ranged from 0.75 to 2.52 for the nine markers typed in the region spanning 88-143 cM from pter. The fasting insulin resistance index was highly correlated with fasting insulin concentrations in this sample and also provided strong evidence for linkage to this region (LOD = 2.99). There was no evidence in our genome-wide scan for linkage of insulin levels to any other chromosome. These results provide evidence that a gene-influencing variation in insulin concentrations exists on chromosome 3p. Possible candidate genes in this region include GBE1 and ACOX2, which encode enzymes involved in glycogen and fatty acid metabolism, respectively.

Peroxisome proliferator-activated receptors (PPAR) are transcription factors that regulate adipocyte differentiation and gene expression. We tested the hypothesis that the Pro12Ala variant of PPAR-gamma2 is associated with obesity and type 2 diabetes-related traits in 921 subjects from the San Antonio Family Heart Study. Subjects with at least one Ala allele (n=210) had significantly higher body mass index (P=0.015) and waist circumference (P=0.028) and significantly higher levels of serum leptin (P= 0.022) than those without the allele. Further studies will determine whether the Pro12Ala variant itself, or other genetic variation at PPAR-gamma, is responsible for this association with measures of obesity in Mexican Americans.

OBJECTIVE: To investigate whether the region of chromosome 11 (11q13) containing the genes UCP2 and UCP3 could be excluded for linkage with a variety of obesity-related phenotypes in humans.

DESIGN: Exclusion mapping using a variance component approach in extended pedigrees.

SUBJECTS: Four-hundred and fifty eight individuals (195 females, 263 males) distributed in 10 Mexican American families of probands randomly ascertained with respect to any disease state and who are participating in the San Antonio Family Heart Study. Ages range from 18 to 87 (mean age 35 y).

MEASUREMENTS: Serum leptin levels, fat mass (FM), body mass index (BMI), and waist circumference.

RESULTS: We were able to exclude the chromosomal region containing UCP2/UCP3 as having an effect on this set of obesity-related phenotypes at relative effect sizes of 10% or greater (P-values<0.05).

CONCLUSIONS: These results suggest that variation in these genes is unlikely to have a substantial effect on the expression of obesity-related phenotypes in the Mexican American population.

Osteocalcin (OC) is an important constituent of bone that is synthesized by osteoblasts. Serum levels of OC have been used as a biochemical marker of bone turnover. To identify the genes influencing variation in serum OC levels, we conducted a genome-wide scan in 429 individuals comprising 10 large multigenerational families. OC levels were measured by immunoassay, and genetic markers were typed at approximately 10-cM intervals across the genome. Quantitative trait linkage was tested using a multipoint analysis based on variance component methodology, adjusting for the effects of age, sex, and oral contraceptive use. Significance levels for linkage were obtained empirically, by Monte Carlo simulation. The heritability of OC levels in this population was 62 +/- 8%. We detected significant evidence for linkage between a quantitative trait locus influencing serum OC levels and markers on chromosome 16q, and suggestive evidence for linkage of OC levels with markers on chromosome 20q. The multipoint lod scores peaked at 3.35 on chromosome 16 and 2.78 on chromosome 20, corresponding to P values of 0.00004 and 0.00017, respectively. A potential candidate gene for bone formation in the linked region on chromosome 20 is CDMP1, which encodes cartilage-derived morphogenetic protein 1. Future studies should evaluate whether variation in CDMP1 or in other genes in the linked regions on chromosomes 16 and 20 influence the rate of bone turnover.

We previously reported that our genome-scanning initiative had detected a highly significant linkage (log odds ratio = 4.95; P = 9 x 10(-7)) between a quantitative trait locus (QTL) on chromosome 2 and leptin levels in Mexican American families. We now have typed additional microsatellite markers in this region, increasing this log odds ratio score to 7.46 (P = 2 x 10(-9)). This region of chromosome 2 contains a strong positional candidate gene, POMC. The POMC gene codes for POMC, the prohormone from which alphaMSH, ACTH, and beta-endorphin are derived. Studies by others have shown that POMC-derived products are involved in the regulation of appetite and obesity. We have used polymorphisms in POMC to map its location within the 95% confidence interval of the peak for the linkage signal for the QTL. We also constructed POMC haplotypes using these polymorphisms and have found a significant association with normal variation in leptin levels (P = 0.001). We conclude that variation in POMC is associated with normal variation in serum leptin levels, providing further evidence that POMC may be the leptin QTL previously identified in Mexican American families.

Small, dense LDL particles are associated with increased risk of cardiovascular disease. To identify the genes that influence LDL size variation, we performed a genome-wide screen for cholesterol concentrations in 4 LDL size fractions. Samples from 470 members of randomly ascertained families were typed for 331 microsatellite markers spaced at approximately 15 cM intervals. Plasma LDLs were resolved by using nondenaturing gradient gel electrophoresis into 4 fraction sizes (LDL-1, 26.4 to 29.0 nm; LDL-2, 25.5 to 26.4 nm; LDL-3, 24.2 to 25.5 nm; and LDL-4, 21.0 to 24.2 nm) and cholesterol concentrations were estimated by staining with Sudan Black B. Linkage analyses used variance component methods that exploited all of the genotypic and phenotypic information in the large extended pedigrees. In multipoint linkage analyses with quantitative trait loci for the 4 fraction sizes, only LDL-3, a fraction containing small LDL particles, gave peak multipoint log10 odds in favor of linkage (LOD) scores that exceeded 3.0, a nominal criterion for evidence of significant linkage. The highest LOD scores for LDL-3 were found on chromosomes 3 (LOD=4.1), 4 (LOD=4.1), and 6 (LOD=2.9). In oligogenic analyses, the 2-locus LOD score (for chromosomes 3 and 4) increased significantly (P=0.0012) to 6.1, but including the third locus on chromosome 6 did not significantly improve the LOD score (P=0.064). Thus, we have localized 2 major quantitative trait loci that influence variation in cholesterol concentrations of small LDL particles. The 2 quantitative trait loci on chromosomes 3 and 4 are located in regions that contain the genes for apoD and the large subunit of the microsomal triglyceride transfer protein, respectively.

The beta-3 adrenergic receptor (ADRB3) has been implicated as a regulator of energy expenditure, and a polymorphism in codon 64 of this gene (Trp64Arg) has been associated in some studies with obesity and insulin resistance. However, many studies have failed to detect an effect of this variant, and the importance of the Trp64Arg variant in human obesity remains controversial. We performed a quantitative linkage analysis of the ADRB3 and obesity, using 12 markers (including the intragenic Trp64Arg polymorphism) spanning a 57-cM region of chromosome 8. The study population consisted of 470 individuals from 10 large multigenerational families of Mexican-American ancestry residing in San Antonio, TX. In two-point analysis, logarithm of odds (LOD) scores >1.0 were observed for six markers surrounding ADRB3 in a 33-cM region spanned by markers D8S1477 and D8S1136. The multipoint LOD score was 3.21, occurring between markers D8S1121 and ADRB3, approximately 2-3 cM from ADRB3. Adjusting for the presence of the Arg64 allele or excluding from the analysis the 11 individuals homozygous for the Arg64 allele did not reduce the evidence for linkage. A genome scan was conducted at 10 cM map density to detect other loci influencing variation in BMI. Multipoint LOD scores >1.0 were observed in four other regions, including two on chromosome 17, one on chromosome 6q, and one on chromosome 2p. These data suggest that the ADRB3 should continue to be regarded as a strong candidate gene for obesity even though evidence for an effect of the Trp64Arg polymorphism could not be established. It is also possible that a gene closely linked to ADRB3 may influence susceptibility to obesity.