Association analysis of vitamin D-binding protein gene polymorphisms with variations of obesity-related traits in Caucasian nuclear families

Article metrics

Abstract

Background:

Vitamin D-binding protein (DBP) gene is well known for its function on glucose and vitamin D metabolism in human populations. Previous studies suggested that the in vivo DBP level may play a role in the etiology of obesity. However, few studies explored the contribution of DBP gene to the variance of obesity phenotypes.

Objective:

To investigate the relationship of DBP polymorphisms and obesity in Caucasian nuclear families.

Design:

We genotyped 14 single-nucleotide polymorphisms (SNPs) located in and around the DBP gene in 1873 Caucasian subjects from 405 nuclear families. Three obesity-related quantitative phenotypes were investigated, including body mass index (BMI), fat mass and percentage of fat mass (PFM). Single SNPs and haplotypes (three blocks) were tested by family-based association using the FBAT software.

Results:

SNP2 (rs17467825) and its corresponding haplotype GAA (frequency 0.270) in block1 showed strongest associations with PFM (P=0.0011 and 0.0023, respectively). In multivariate test significant association was also observed for SNP2 with fat_mass&BMI&PFM (P=0.0098). Subsequent sex-stratified analyses demonstrated nominal association for SNP2 and haplotype GAA with PFM in the female subgroup.

Conclusion:

Polymorphisms of DBP gene were significantly association with human PFM, especially in female, suggesting the importance of DBP gene in the pathogenesis of human obesity.

Introduction

Obesity is a multifactorial disorder associated with the increased risk of complex diseases such as type 2 diabetes (T2D), hypertension and coronary heart disease.1, 2 In the past two decades, the prevalence of obesity has increased dramatically in both developing and developed countries,3, 4 thus rendering obesity the sixth most important risk factor contributing to the overall burden of human disease worldwide. Family, twin and adoptee studies demonstrated that genetic factors significantly contribute to the etiology of obesity, with heritability ranging from 0.3 to 0.8.5, 6, 7 However, the inheritance pattern of obesity is so complex that more than 100 candidate genes each with small to modest effects may be involved in its etiology.8

Vitamin D-binding protein (DBP), also known as group-specific component (GC), is composed of 13 exons and 12 introns and located in chromosome 4q12–q13.9 DBP was implicated in many important biological functions, such as transport and storage of vitamin D, bone development, and modulation of immune and inflammatory responses.10, 11

It was already known that the alterations in the vitamin D endocrine system were associated with obesity.12, 13 The levels of serum 1,25-dihydroxyvitamin D (1,25(OH)2D) were elevated, whereas 25-hydroxyvitamin D (25(OH)D) levels were reduced in morbidly obese patients.12, 14, 15 In addition, the levels of serum vitamin D metabolites inversely correlate with fat mass12, 16 and body mass index (BMI).16 Therefore, DBP, playing an important role in the vitamin D endocrine system by binding with 25(OH)D and 1,25(OH)2D and regulating their in vivo levels and functions,17, 18 may also influence the obesity susceptibility in the human populations.

DBP was also recognized as one factor associated with insulin resistance that leads to the increased risk of obesity. Pratley et al.19 found that the 4p15–q12 region containing the DBP gene was linked to plasma glucose and insulin concentrations in nondiabetic Pima Indians. In the Caucasian and Japanese populations, the polymorphisms of DBP were found to be associated with the fasting insulin level, an index of insulin resistance.20, 21 Moreover, correlations between the in vivo DBP concentrations with BMI and fat mass were also observed.18

However, the direct genetic studies on the relationship between DBP gene and obesity is still lacking to date. The aim of this study is to investigate whether the DBP gene is associated with obesity phenotypes by studying a set of high-density single nucleotide polymorphisms (SNPs) of DBP in a relatively large sample of Caucasian nuclear families.

Materials and methods

Subjects

The study was approved by the Creighton University Institutional Review Board. The study subjects came from an expanding database created for ongoing studies in the Osteoporosis Research Center (ORC) of Creighton University to search for genes underlying common complex diseases/traits in humans, including osteoporosis, obesity and height, etc. Signed informed consent documents were obtained from all study participants before they entered the study. The study design and exclusion criteria were published before.22 The exclusion criteria may be useful to minimize the influence of environment factors on obesity phenotypes. A total of 405 nuclear families (family sizes ranging from 3 to 12, and the average was 4.62) with 1873 individuals ranging in age from 19 to 88 were analyzed in this study, including 740 parents, 389 male children and 744 female children. All of the participants were US Caucasians of European origin.

Genotyping

Genomic DNA was extracted from whole blood using a commercial isolation kit (Gentra Systems, Minneapolis, MN, USA) following the procedure detailed in the kit. DNA concentration was assessed by a DU530 UV/VIS Spectrophotometer (Beckman Coulter, Inc., Fullerton, CA, USA).

A total of 17 SNPs in and around DBP were selected on the basis of the following criteria: (1) validation status (validated experimentally in human populations), especially in Caucasians, (2) an average density of one SNP per 3 kb, (3) degree of heterozygosity, that is, minor allele frequencies (MAFs) >0.05, (4) functional relevance and importance, (5) reported to dbSNP by various sources. Fourteen SNPs were successfully genotyped using the high-throughput BeadArray SNP genotyping technology of Illumina Inc. (San Diego, CA, USA). The average rate of missing genotype data was 0.05%. The average genotyping error rate estimated through blind sample duplication was <0.01%. The 14 SNPs successfully genotyped in this study were spaced 4 kb apart on average and covered the full length (including up- and down-stream potential regulatory regions) of the DBP gene (Figure 1).

Figure 1
figure1

LD structure of the DBP gene (SNP8 and SNP10 with MAF <0.05 in Table 2 were excluded).

Phenotype measurement

Weight was measured in light indoor clothing, using a calibrated balance beam scale, and height was measured using a calibrated stadiometer. BMI was calculated as body weight (in kilograms) divided by the square of height (in meters). Fat mass was measured by dual-energy X-ray absorptiometry (DXA). Percentage of fat mass (PFM) was calculated as the ratio of fat mass to body weight. The measurement precisions of weight, height and fat mass, as reflected by the coefficient of variation (CV) were 1.2, 0.9 and 2.2%, respectively. The phenotypes were significantly correlated, with the phenotypic correlations being 0.879 between fat mass and BMI, 0.815 between fat mass and PFM, and 0.530 between BMI and PFM, respectively.

Statistical analyses

We used PedCheck23 to check Mendelian consistency of SNP genotype data and removed any inconsistent genotypes. Then the error checking option embedded in Merlin24 was run to identify and discard the genotypes flanking excessive recombinants, thus further reducing genotyping errors. Hardy–Weinberg equilibrium (HWE) for each SNP was tested using the PEDSTATS procedure implemented in Merlin.

The absolute D′ and r2 values for SNPs were calculated and used to measure the extent of pairwise linkage disequilibrium (LD), and LD block structure was examined by the program Haploview V3.2 (http://www.broad.mit.edu/mpg/haploview/).25 A block is identified when at least 95% of SNP pairs in a region meet these criteria for strong LD. The algorithm of integer linear programming (ILP) implemented in PedPhase V2.0 (http://www.cs.ucr.edu/~jili/haplotyping.html) and based on LD assumption were used to recover phase information at each marker locus with great speed and accuracy even in the presence of 20% missing data.26

FBAT statistic implemented in the software package FBAT V1.5527 was used to perform the family-based association tests for single SNPs with three quantitative obesity-related traits. The allelic association examines the transmissions of the interested markers from parents to the affected offspring. We applied the additive genetic model as it turns out to perform well even when the true genetic model is not the additive one. The null hypothesis here is ‘no linkage and no association between the marker and any obesity susceptibility locus’. The multivariate FBAT test based on the generalized estimating equation (GEE) approach were conducted to reduce the power loss associated with multiple testing approaches,28 it allows to test the null hypothesis that the marker locus is not linked to any genetic locus that has an influence on the selected phenotypes.

We performed HBAT to test the haplotype association.29 Similar to FBAT, HBAT provides a test in the family-based design, which is robust to population admixture, phenotype distribution misspecification and ascertainment bias. Empirical global P-values were obtained using 10 000 permutation procedure implemented in HBAT. With three phenotypes and the assumption that the three blocks represent independent information, for reaching the experiment-wide significance level of α=0.05, the approximate corrections were calculated as follows: for univariate testing in the entire sample, the single test threshold was α=0.05/(3*3)=0.0056; for multivariate testing in the entire sample, the single test threshold was α=0.05/3=0.013. In addition, we performed family-based association analyses in sex-stratified manner. To account for multiple testing issues, the single test threshold in the male or female subgroups was α=0.05/(3*3*2)=0.0028.

QTDT program (http://www.sph.umich.edu/csg/abecasis/QTDT/) was used to estimate the genetic contribution of SNPs and haplotypes of DBP gene to obesity phenotypes. Approximate phenotypic variance explained by the significant or suggestive marker was computed as 2p(1−p)a2, where p is the frequency of the studied allele or haplotype and a is the estimate of additive effects.

Regression residuals representing age- and sex-adjusted fat mass, BMI and PFM values were used in all the FBAT analyses in this study. All of the residual data were tested for normality by the Kolmogorov–Smirnov test implemented in the software Minitab (Minitab Inc., State College, PA, USA) before the association tests. No significant deviation from the normal distribution was found for the residual phenotypic data analyzed. Before sex-stratified association analyses, we pursued the interaction analyses between SNPs and gender on phenotypic variation. Significant interactions were observed between each SNP and gender for three obesity-related traits.

Results

The phenotypic characteristics of the study subjects stratified by sex are presented in Table 1. There were significant differences between men and women in terms of the mean values of the phenotypes such as BMI, fat mass and PFM (P<0.001 by t-test). On average, men had larger BMI values than women, whereas women had larger fat mass and PFM values than men.

Table 1 Characteristics of the study subjects analyzed in the 405 nuclear families

Table 2 shows the map locations and marker characteristics for the 14 genotyped SNPs in and around DBP. SNP8 and SNP10 with MAFs <0.05 were abandoned for subsequent analyses due to the lack of statistical power caused by low MAFs. The rare-allele frequencies for the other analyzed SNPs ranged from 5 to 44%. No significant deviation from HWE was found for any analyzed SNP.

Table 2 Information of all the genotyped SNPs

We identified three blocks with high LD (Figure 1). Block1 extended from 3′ UTR to intron6 and included SNPs 1–7. Block2 mainly covered intron1 and included SNPs 9, 11 and 12. Block3 spanned the promoter region and consisted of SNP13 and SNP14. Five tag SNPs were selected by Haploview program, including SNPs 1, 3, 6, 9 and 13. Block1 was represented by SNPs 1, 3 and 6, whereas block2 and block3 were represented by SNP9 and SNP13, respectively. The haplotype structure, diversity and tag SNP are shown in Figure 2.

Figure 2
figure2

The haplotype structure, diversity and tag SNP of DBP gene (SNP8 and SNP10 with MAF <0.05 in Table 2 were excluded).

Significant association results are summarized in Table 3. Among the 12 SNPs, the most significant associations with PFM were observed for SNP2 (rs17467825) in block1 (P=0.0011), as well as for the corresponding haplotype GAA (frequency=0.270) containing SNP2 (P=0.0023). In addition, significant association trends for SNP2 and haplotype GAA regarding other obesity phenotypes like fat mass and BMI were also detected (P=0.0061, 0.0477, 0.0102, 0.0640, respectively). Interestingly, we found SNP2 was associated with fat mass, BMI and PFM simultaneously (P=0.0098) in the multivariate analyses. For SNP1 (rs1491711) and SNP3 (rs705117), suggestive associations were observed with PFM and fat mass, respectively. However, after correcting for multiple testing, only SNP2 (rs17467825) and its corresponding haplotype GAA met the experiment-wide significance criterion for PFM, and SNP2 also reached the experiment-wide significant level for fat_mass&BMI&PFM in the multivariate association test. Moreover, SNP2 and haplotype GAA were found to explain 1.42 and 1.19% of PFM variance by QTDT program.

Table 3 Association results for the DBP polymorphisms, namely SNPs (SNP1, SNP2 and SNP3) and the haplotype GAA in block1, and the empirical global P-values for block1 as a whole marker

In further sex-stratified analyses, SNP2 was also associated with PFM in the female subgroup. Consistent with entire sample, we found the nominally significant association for SNP2 with fat_mass&BMI&PFM (P=0.0457), and for haplotype GAA with fat mass and PFM in female subjects (P=0.0195 and 0.0084), whereas the signals were insignificant in male subjects. However, after accounting for multiple testing, the above results were only borderline significant.

Discussion

We investigated the relationship between common variants of the DBP gene with obesity-related phenotypes, including fat mass, BMI and PFM, in Caucasian nuclear families. Compared with the previous population-based association studies on DBP gene,18, 21 the family-based association design is robust to false-positive/negative errors owing to the potential population stratification.30

Various phenotypes were assessed in this study. BMI is a widely used measure of obesity, owing to its measurement convenience and low cost. However, BMI may not distinguish fat mass from lean mass in some circumstances and thus limited in its utility in terms of adequately and accurately reflecting the obesity status.31 For example, athletes often have a BMI>25 kg/m2, along with only 10–15% body fat.31 Therefore, in addition to BMI, we also used fat mass and PFM as quantitative phenotypes to better understand the relationship of DBP gene and obesity.

In this study, we detected the experiment-wide significant associations for SNP2 and its haplotype GAA with PFM, as well as nominal significant associations with fat mass and BMI. Considering the high correlations among the three obesity-related phenotypes, we also performed the multivariate tests and confirmed the associations between SNP2 and obesity. Our data thus strongly supported that DBP could be important to PFM, an important obesity phenotype. However, the exact biological mechanism of how the detected significant DBP polymorphism influence the obesity phenotypes was unknown, although we hypothesize that SNP2 or its highly correlated polymorphisms may influence the mRNA stability of DBP, as SNP2 is located in the 3′ UTR that usually determines the in vivo mRNA decay rate of a gene.32

Furthermore, the significant associations with obesity were found to be female-specific in our sample as suggested by the sex-stratified analyses. This is not unexpected given the previous example such as the sex-specific genetic association between the estrogen receptor alpha (ESR1) gene and the variation of body fat in female subjects.33 Moreover, it was proved that women were distinctly different to men with the actions of insulin and the susceptibility to develop insulin resistance which was correlated with obesity.34 However, owing to the limited sample size of the male subgroup, whether the genetic effects of SNP2 and its specific haplotype on obesity are indeed only present in female subjects need further substantiation.

In conclusion, on the basis of a large sample of Caucasian nuclear families and the robust statistical approaches, our study provides certain important evidence that the polymorphisms of the DBP gene are significantly associated with the variation of obesity phenotypes in the Caucasian population, especially in Caucasian females. The present work furnished some clues for further molecular studies aimed at identifying the hidden mechanisms of how the DBP gene may influence the obesity status in humans.

References

  1. 1

    Kopelman PG . Obesity as a medical problem. Nature 2000; 404: 635–643.

  2. 2

    Poirier P, Giles TD, Bray GA, Hong Y, Stern JS, Pi-Sunyer FX et al. Obesity and cardiovascular disease: pathophysiology, evaluation, and effect of weight loss. Arterioscler Thromb Vasc Biol 2006; 26: 968–976.

  3. 3

    Flegal KM, Carroll MD, Kuczmarski RJ, Johnson CL . Overweight and obesity in the United States: prevalence and trends, 1960–1994. Int J Obes Relat Metab Disord 1998; 22: 39–47.

  4. 4

    Gallus S, Colombo P, Scarpino V, Zuccaro P, Negri E, Apolone G et al. Overweight and obesity in Italian adults 2004, and an overview of trends since 1983. Eur J Clin Nutr 2006; 60: 1174–1179.

  5. 5

    Liu YJ, Liu PY, Long J, Lu Y, Elze L, Recker RR et al. Linkage and association analyses of the UCP3 gene with obesity phenotypes in Caucasian families. Physiol Genomics 2005; 22: 197–203.

  6. 6

    Sorensen TI, Holst C, Stunkard AJ . Adoption study of environmental modifications of the genetic influences on obesity. Int J Obes Relat Metab Disord 1998; 22: 73–81.

  7. 7

    Stunkard AJ, Foch TT, Hrubec Z . A twin study of human obesity. JAMA 1986; 256: 51–54.

  8. 8

    Perusse L, Rankinen T, Zuberi A, Chagnon YC, Weisnagel SJ, Argyropoulos G et al. The human obesity gene map: the 2004 update. Obes Res 2005; 13: 381–490.

  9. 9

    Witke WF, Gibbs PE, Zielinski R, Yang F, Bowman BH, Dugaiczyk A . Complete structure of the human Gc gene: differences and similarities between members of the albumin gene family. Genomics 1993; 16: 751–754.

  10. 10

    White P, Cooke N . The multifunctional properties and characteristics of vitamin D-binding protein. Trends Endocrinol Metab 2000; 11: 320–327.

  11. 11

    Gomme PT, Bertolini J . Therapeutic potential of vitamin D-binding protein. Trends Biotechnol 2004; 22: 340–345.

  12. 12

    Arunabh S, Pollack S, Yeh J, Aloia JF . Body fat content and 25-hydroxyvitamin D levels in healthy women. J Clin Endocrinol Metab 2003; 88: 157–161.

  13. 13

    Reis AF, Hauache OM, Velho G . Vitamin D endocrine system and the genetic susceptibility to diabetes, obesity and vascular disease. A review of evidence. Diabetes Metab 2005; 31: 318–325.

  14. 14

    Bell NH, Epstein S, Greene A, Shary J, Oexmann MJ, Shaw S . Evidence for alteration of the vitamin D-endocrine system in obese subjects. J Clin Invest 1985; 76: 370–373.

  15. 15

    Buffington C, Walker B, Cowan Jr GS, Scruggs D . Vitamin D deficiency in the morbidly obese. Obes Surg 1993; 3: 421–424.

  16. 16

    Parikh SJ, Edelman M, Uwaifo GI, Freedman RJ, Semega-Janneh M, Reynolds J et al. The relationship between obesity and serum 1, 25-dihydroxy vitamin D concentrations in healthy adults. J Clin Endocrinol Metab 2004; 89: 1196–1199.

  17. 17

    Arnaud J, Constans J . Affinity differences for vitamin D metabolites associated with the genetic isoforms of the human serum carrier protein (DBP). Hum Genet 1993; 92: 183–188.

  18. 18

    Taes YE, Goemaere S, Huang G, Van Pottelbergh I, De Bacquer D, Verhasselt B et al. Vitamin D binding protein, bone status and body composition in community-dwelling elderly men. Bone 2006; 38: 701–707.

  19. 19

    Pratley RE, Thompson DB, Prochazka M, Baier L, Mott D, Ravussin E et al. An autosomal genomic scan for loci linked to prediabetic phenotypes in Pima Indians. J Clin Invest 1998; 101: 1757–1764.

  20. 20

    Iyengar S, Hamman RF, Marshall JA, Majumder PP, Ferrell RE . On the role of vitamin D binding globulin in glucose homeostasis: results from the San Luis Valley Diabetes Study. Genet Epidemiol 1989; 6: 691–698.

  21. 21

    Hirai M, Suzuki S, Hinokio Y, Hirai A, Chiba M, Akai H et al. Variations in vitamin D-binding protein (group-specific component protein) are associated with fasting plasma insulin levels in Japanese with normal glucose tolerance. J Clin Endocrinol Metab 2000; 85: 1951–1953.

  22. 22

    Xiong DH, Liu YZ, Liu PY, Zhao LJ, Deng HW . Association analysis of estrogen receptor alpha gene polymorphisms with cross-sectional geometry of the femoral neck in Caucasian nuclear families. Osteoporos Int 2005; 16: 2113–2122.

  23. 23

    O’Connell JR, Weeks DE . PedCheck: a program for identification of genotype incompatibilities in linkage analysis. Am J Hum Genet 1998; 63: 259–266.

  24. 24

    Abecasis GR, Cherny SS, Cookson WO, Cardon LR . Merlin – rapid analysis of dense genetic maps using sparse gene flow trees. Nat Genet 2002; 30: 97–101.

  25. 25

    Barrett JC, Fry B, Maller J, Daly MJ . Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005; 21: 263–265.

  26. 26

    Li J, Jiang T . Computing the minimum recombinant haplotype configuration from incomplete genotype data on a pedigree by integer linear programming. J Comput Biol 2005; 12: 719–739.

  27. 27

    Horvath S, Xu X, Laird NM . The family based association test method: strategies for studying general genotype–phenotype associations. Eur J Hum Genet 2001; 9: 301–306.

  28. 28

    Lange C, Silverman EK, Xu X, Weiss ST, Laird NM . A multivariate family-based association test using generalized estimating equations: FBAT-GEE. Biostatistics 2003; 4: 195–206.

  29. 29

    Horvath S, Xu X, Lake SL, Silverman EK, Weiss ST, Laird NM . Family-based tests for associating haplotypes with general phenotype data: application to asthma genetics. Genet Epidemiol 2004; 26: 61–69.

  30. 30

    Thomson G . Mapping disease genes: family-based association studies. Am J Hum Genet 1995; 57: 487–498.

  31. 31

    Allison DB, Saunders SE . Obesity in North America. An overview. Med Clin North Am 2000; 84: 305–332, v.

  32. 32

    Fang Y, van Meurs JB, d’Alesio A, Jhamai M, Zhao H, Rivadeneira F et al. Promoter and 3′-untranslated-region haplotypes in the vitamin D receptor gene predispose to osteoporotic fracture: the Rotterdam study. Am J Hum Genet 2005; 77: 807–823.

  33. 33

    Okura T, Koda M, Ando F, Niino N, Ohta S, Shimokata H . Association of polymorphisms in the estrogen receptor alpha gene with body fat distribution. Int J Obes Relat Metab Disord 2003; 27: 1020–1027.

  34. 34

    Mittendorfer B . Insulin resistance: sex matters. Curr Opin Clin Nutr Metab Care 2005; 8: 367–372.

Download references

Acknowledgements

The study was partially supported by grants from NIH (R01 AR050496, K01 AR02170-01, R01 AR45349-01 and R01 GM60402-01A1) and an LB595 grant from the State of Nebraska. This study also benefited from grants from Chinese National Science Foundation, Huo Ying Dong Education Foundation and Ministry of Education of PR China, Hunan Normal University and Xi'an Jiao Tong University.

Author information

Correspondence to H-W Deng.

Rights and permissions

Reprints and Permissions

About this article

Keywords

  • DBP
  • association
  • PFM

Further reading