SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Research Design and Methods
  4. Acknowledgment
  5. Disclosure
  6. REFERENCES

Conflicting results have been reported regarding the effect of the peroxisome proliferator-activated receptor-γ−2 (PPARγ2) Pro12Ala polymorphism, (singly or in combination with the silent C1431T polymorphism) on BMI. Gender-based dimorphism has been evidenced for genes that affect BMI, but few and conflicting data are available regarding PPARγ2. We sought to investigate whether the Pro12Ala interacts with gender in modulating BMI in 566 nondiabetic unrelated white subjects (men:women = 211:355, age 36.59 ± 11.85; BMI 25.36 ± 4.53). In the whole study population, BMI, fasting glucose and insulin levels, and lipid profile were similar in Ala12 carriers (i.e., XA) and Pro/Pro homozygous subjects. Among the men, but not among the women, X/Ala individuals showed higher BMI (25.9 ± 3.6 vs. 28.2 ± 4.9, P = 0.006) and risk of obesity (odds ratio = 2.85, 95% confidence interval = 1.07–7.62). A significant gene-gender interaction in modulating BMI was observed (P = 0.039). Among the men, but not among the women, those carrying Ala-T haplotype (i.e., containing both Ala12 and T1431 variants) showed the highest BMI (haplo-score = 3.72, P = 0.0014). Our data indicate that in whites from Italy the PPARγ2 Pro12Ala polymorphism interacts with gender in modulating BMI, thereby replicating some, but not all, earlier data obtained in different populations. Whether the PPARγ2-gender interaction is a general phenomenon across different populations, is still an open question, the answer to which requires additional, specifically designed, studies.

Because of its increasing prevalence and its devastating adverse outcomes, obesity represents a worldwide public health problem (1). Although environmental factors play a central role, there is considerable evidence that genes are also important in the pathogenesis of obesity and related traits (2). Among reported genes, the one that has been most widely studied is the peroxisome proliferator-activated receptor-γ−2 (PPARγ2) gene, a nuclear hormone receptor that is known to play a fundamental role in adipogenesis and insulin sensitization (3). The PPARγ2 gene's common Pro12Ala missense polymorphism has (either singly or in haplotype combination with the silent C1431T polymorphism) been widely investigated for its possible involvement in the modulation of body weight. While several studies have evidenced the association of the Ala12 variant with increased BMI (4,5,6), others have either failed to confirm this finding (7,8,9), or have shown an association of this variant with a lowering of BMI (10). A recent meta-analysis carried out on samples from nondiabetic individuals, has clearly shown that the Pro12Ala polymorphism has no significant and unequivocal effect on BMI, given the high degree of heterogeneity across the samples studied (11). The results of this study led the authors to conclude that, in phenotype-genotype association studies of complex metabolic traits, “stratification of analyses by as many factors as possible” may lead to a better understanding of the genetic background.

Gender-based dimorphism has been repeatedly reported for several genes that affect BMI and related traits (12), but only few and conflicting reports are available regarding the PPARγ2 gene (6,13,14). The aim of this study was to investigate whether common variants at the PPARγ2 locus (i.e., Pro12Ala and C1431T) modulate BMI and related traits differentially in men and women.

For this purpose, a cohort of 566 nondiabetic unrelated individuals were genotyped for both Pro12Ala and C1431T polymorphisms. The proportion of observed genotypes for both polymorphisms in the study sample did not significantly deviate from Hardy-Weinberg equilibrium (χ2 = 0.50, P = 0.47 and χ2 = 0.34; P = 0.55, respectively). Because of the very low number of Ala/Ala individuals (n = 1), data from Pro/Ala and Ala/Ala (here named as X/Ala) individuals were pooled and analyzed together. No difference was observed in BMI, waist circumference, fasting levels of glucose, insulin, triglycerides, and high-density lipoprotein (HDL)-cholesterol, metabolic syndrome (MS) score, prevalence of MS, and homeostasis model assessment of insulin resistance (HOMA-IR) values between the X/Ala and Pro/Pro individuals (Table 1). A significant gene-gender interaction was observed in the modulation of BMI levels (P = 0.039). A gene-gender interaction was observed also in the modulation of waist circumference (P = 0.07), insulin, HOMA-IR, and HDL-cholesterol (P = 0.03 for all). These interactions were no longer significant after adjusting for BMI (P = 0.97, 0.20, 0.21, and 0.13, respectively). In fact, when the men (n = 211) and women (n = 355) were considered separately, X/Ala individuals showed a higher BMI as compared to Pro/Pro carriers among the men but not among the women (Table 2). Among the women, we had an 80% power (P = 0.05) to detect a BMI difference of 2.3 kg/m2 across the two genotype groups; we cannot therefore exclude the possibility that a weaker association may exist among these subjects. Among the men (but not among the women, data not shown) the X/Ala genotype was more frequent in the obese (i.e., BMI ≥ 30 kg/m2) than in the nonobese individuals (7/29, 24.1% vs. 18/182, 9.9%, respectively), thereby indicating a higher risk of obesity for X/Ala individuals than for Pro/Pro carriers: odds ratio = 2.85 (95% confidence interval = 1.07–7.62). Although not reaching formal statistical significance with the sample size used in the study, there was a tendency toward differences across genotype groups for waist circumference, insulin, HOMA-IR, HDL-cholesterol, and prevalence of MS among the men but not among the women (Table 2). The statistical differences in all these variables across the two genotype groups among the men showed substantial changes after adjusting for BMI (Table 2), thereby suggesting that they were dependent mainly on adiposity differences.

Table 1.  Clinical features of the whole sample across different PPARγ2 genotype groups
inline image
Table 2.  Clinical features across different PPARγ2 genotype groups in males and females
inline image

Because of the very low number of T/T genotypes at C1431T polymorphism, data from C/T and T/T (i.e., XT) individuals were pooled and analyzed together. In the whole population, no difference was observed in any variable tested across the two genotype groups (Table 1). Among the men, but not among the women, XT individuals showed higher levels of BMI, but not of waist circumference and HOMA-IR. No interaction between C1431T polymorphism and gender was observed in the modulation of these traits. In our population, Pro12Ala and C1431T polymorphisms were in linkage disequilibrium (D′ = 0.661 and r2 = 0.343). Computational phase inference indicated 4 possible haplotypes (namely, Pro-C, Pro-T, Ala-C, and Ala-T) whose frequencies of occurrence were similar in men (0.90, 0.04, 0.03, and 0.03, respectively) and women (0.91, 0.03, 0.02, and 0.04, respectively). Among the men, significant global P values were obtained for BMI and waist circumference (0.001 and 0.008 respectively), but not for glucose (0.975), insulin (0.232), triglycerides (0.427), HDL-cholesterol (0.351), HOMA-IR (0.260), and MS (0.136). Among the four possible haplotypes, the Ala-T showed the highest levels of these parameters (Table 3). The association with waist circumference was no longer significant after adjusting for BMI (global P value = 0.439). In contrast, among the women no significant global P values were obtained for BMI (0.270), waist circumference (0.324), glucose (0.244), insulin (0.401), triglycerides (0.822), HDL-cholesterol (0.345), HOMA-IR (0.332), and MS (0.308).

Table 3.  Haplotypes data of clinical variables with a significant global P value among males
inline image

Overall, the data relating to haplotypes suggest that, in men, the deleterious effect of the Ala12 variant on BMI requires the simultaneous presence of the T1431 allele.

In summary, our data indicate that the PPARγ2 Pro12Ala polymorphism interacts with gender in modulating BMI, with the deleterious effect of the Ala12 variant being appreciable only among men. The lack of effect of the Ala12 variant in modulating BMI among women replicates data obtained in subjects of European descent from Spain (13) and Brazil (14) but not those in Finns (6), who are descended from not only Europeans but also Asians (15). Therefore, whether the PPARγ2-gender interaction is a general phenomenon across different populations is still an open question, the answer to which requires additional studies to be carried out. A gender-specific genetic effect on BMI modulation has been reported earlier in several animal models (16) as well as in humans (12). Whether the gender specificity we observed is caused by the interaction of PPARγ2 with either sex-linked genes and/or sexual hormone effects is not known. It is also possible that our finding simply reflects the inability of BMI to adequately capture specific aspects of adiposity because of the different fat distribution patterns seen in the two genders. The dissection of the genetic architecture of obesity has been difficult because it is polygenic, with several genes being involved, each contributing a modest effect (2). Also gene-environment and gene-gene interactions are likely to be on the stage (2), making the scenario even more difficult to unravel. In this context, it has been recently proposed that association studies in complex metabolic traits have to be analyzed by stratifying for as many factors as possible (11), thereby allowing testing for interaction between variables. Although they are entirely speculative, our results are in line with such an approach, which is likely to help achieve a better understanding of the pathogenic role played by genes of interest in relation to obesity.

Research Design and Methods

  1. Top of page
  2. Abstract
  3. Research Design and Methods
  4. Acknowledgment
  5. Disclosure
  6. REFERENCES

Selection criteria and features of the samples have been published earlier (17). Briefly, the studied sample comprised 566 nondiabetic unrelated white subjects (211 men and 355 women; BMI 25.36 ± 4.53; age 36.59 ± 11.85, fasting plasma glucose <6.9 mmol/l, not taking medications that are known to interfere with glucose and lipid metabolism), who were residents of the Gargano region (East Coast, Italy). The study was approved by local Ethical Committees and performed according to the Helsinki Declaration. All subjects gave their written informed consent, and were examined between 8:00 and 9:00 am after an overnight fast. Waist circumference was measured at the level midway between the lowest rib margin and iliac crest to the nearest 0.5 cm. Plasma glucose, serum insulin, and lipid profile (total cholesterol, HDL-cholesterol, and triglycerides) were measured using commercially available kits. An individual score relating to MS was assigned, ranging from 0 (no features of MS), to 5 (all features of MS), in accordance with the Adult Treatment Panel III criteria (18). Genotyping of both Pro12Ala and C1431T polymorphisms was performed as described earlier (19,20). Genotyping quality was checked by directly sequencing 10% of randomly selected genotyped samples. The agreement rate of resequenced samples was >99%.

Values are reported as mean values ± s.d. Comparisons between groups were tested using unpaired Student's t-test. Mean values, after adjusting for covariates, were evaluated using the analysis of covariance test. χ2-Test was used for testing for association between obesity and genotype and for Hardy-Weinberg equilibrium. In order to model the effects of different genotypes on the obesity risk, multivariate logistic regression analysis was used and the result was expressed in terms of odds ratio (95% confidence interval). Gene-gender interaction was tested using General Linear Model analysis including gender, genotype and gender × genotype (in terms of multiplicative interaction) in the model. Variables that do not show normal distribution, such as triglycerides, HDL-cholesterol, insulin, and HOMA-IR, were logarithmically transformed before analysis. All the analyses were performed using the SPSS software program, version 12.0 for Windows (Chicago, IL). Comparisons among genotype groups were all adjusted for age (and gender, when performed in whole population). Pairwise disequilibria measures (D′ and r2) were calculated between polymorphisms using software programs PHASE (21) version 2.0. PHASE was also used for reconstructing individual haplotypes. Haplotype association analyses were performed using Haplo Stats (22), which gives the significance of haplotype-phenotype association in terms of Haplo-scores. In our specific context, positive or negative Haplo-scores indicate that carriers of a given haplotype have higher or lower BMI (or waist) levels, respectively, as compared to all other individuals.

Associations were considered significant if both the global (i.e., in reference with the model comprising all possible haplotypes) and haplotype-specific P values were < 0.05.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Research Design and Methods
  4. Acknowledgment
  5. Disclosure
  6. REFERENCES

This work was partly supported by Italian Ministry of Health (Ricerca Corrente 2006 e 2007 to S.P.).

Disclosure

  1. Top of page
  2. Abstract
  3. Research Design and Methods
  4. Acknowledgment
  5. Disclosure
  6. REFERENCES

The authors declared no conflict of interest.

REFERENCES

  1. Top of page
  2. Abstract
  3. Research Design and Methods
  4. Acknowledgment
  5. Disclosure
  6. REFERENCES
  • 1
    WHO World Health Organization. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. No. 894. WHO Technical Series: World Health Organization, Geneva, Switzerland, 2000.
  • 2
    Bell CG, Walley AJ, Froguel P. The genetics of human obesity. Nat Rev Genet 2005; 6: 221634.
  • 3
    Rosen ED, Spiegelman BM. Rosen. PPARgamma: a nuclear regulator of metabolism, differentiation, and cell growth. J Biol Chem 2001; 276: 3773137734.
  • 4
    Beamer BA, Yen CJ, Andersen RE et al. Association of the Pro12Ala variant in the peroxisome proliferator-activated receptor gamma2 gene with obesity in two Caucasian populations. Diabetes 1998; 47: 18061808.
  • 5
    Meirhaeghe A, Fajas L, Helbecque N et al. Impact of the peroxisome proliferator activated receptor gamma2 Pro12Ala polymorphism on adiposity, lipids and non insulin-dependent diabetes mellitus. Int J Obes Relat Metab Disord 2000; 24: 195199.
  • 6
    Valve R, Sivenius K, Miettinen R et al. Two polymorphisms in the peroxisome proliferator-activated receptor-gamma gene are associated with severe overweight among obese women. J Clin Endocrinol Metab 1999; 84: 37083712.
  • 7
    Hamann A, Münzberg H, Buttron P et al. Missense variants in the human peroxisome proliferator-activated receptor-gamma2 gene in lean and obese subjects. Eur J Endocrinol 1999; 141: 9092.
  • 8
    Swarbrick MM, Chapman CM, McQuillan BM et al. A Pro12Ala polymorphism in the human peroxisome proliferator-activated receptor-gamma 2 is associated with combined hyperlipidaemia in obesity. Eur J Endocrinol 2001; 144: 277282.
  • 9
    Mori Y, Kim-Motoyama H, Katakura T et al. Effect of the Pro12Ala variant of the human peroxisome proliferatoractivated receptor gamma 2 gene on adiposity, fat distribution, and insulin sensitivity in Japanese men. Biochem Biophys Res Commun 1998; 251: 195198.
  • 10
    Deeb SS, Fajas L, Nemoto M et al. A Pro12Ala substitution in PPARgamma2 associated with decreased receptor activity, lower body mass index and improved insulin sensitivity. Nat Genet 1998; 20: 284287.
  • 11
    Tonjes A, Scholz M, Loeffler M, Stumvoll M. Association of Pro12Ala polymorphism in peroxisome proliferator-activated receptor gamma with Pre-diabetic phenotypes: meta-analysis of 57 studies on nondiabetic individuals. Diabetes Care 2006; 29: 24892497.
  • 12
    Weiss LA, Pan L, Abney M, Ober C. The sex-specific genetic architecture of quantitative traits in humans. Nat Genet 2006; 38: 218222.
  • 13
    González Sánchez JL, Serrano Ríos M, Fernández Perez C, Laakso M, Larrad MT Martínez. Effect of the Pro12Ala polymorphism of the peroxisome proliferator-activated receptor g-2 gene on adiposity, insulin sensitivity and lipid profile in the Spanish population. Eur J Endocrinol 2002; 147: 495501.
  • 14
    Mattevi VS, Zembrzuski VM, Hutz MH. Effects of a PPARG gene variant on obesity characteristics in Brazil. Braz J Med Biol Res 2007; 40: 927932.
  • 15
    Kittles R, Perola M, Peltonen L et al. Dual origins of Finns revealed by Y chromosome haplotype variation. Am J Hum Genet 1998; 62: 11711179.
  • 16
    Klaman LD, Boss O, Peroni OD et al. Increased energy expenditure decreased adiposity and tissue-specific insulin sensitivity in protein tyrosine phosphatase 1B- deficient mice. Mol Cell Biol 2000; 20: 54795489.
  • 17
    Prudente S, Chandalia M, Morini E et al. The Q121/Q121 genotype of ENPP1/PC-1 is associated with lower BMI in non-diabetic whites. Obesity (Silver Spring) 2007; 15: 14.
  • 18
    Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults: Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 2001; 285: 24862497.
  • 19
    Baratta R, Di Paola R, Spampinato D et al. Evidence for genetic epistasis in human insulin resistance: the combined effect of PC-1 (K121Q) and PPARgamma2 (P12A) polymorphisms. J Mol Med 2003; 81: 718723.
  • 20
    Meirhaeghe A, Fajas L, Helbecque N et al. A genetic polymorphism of the peroxisome proliferator-activated receptor γ gene influences plasma leptin levels in obese humans. Hum Mol Genet 1998; 7: 435440.
  • 21
    Stephens M, Donnelly P. A comparison of bayesian methods for haplotype reconstruction from population genotype data. Am J Hum Genet 2003; 73: 11621169.
  • 22
    Schaid DJ, Rowland CM, Tines DE, Jacobson RM, Poland GA. Score tests for association between traits and haplotypes when linkage phase is ambiguous. Am J Hum Genet 2002; 70: 425434.