PPARγ Pro12Ala and ACE ID polymorphisms are associated with BMI and fat distribution, but not metabolic syndrome

Epidemiological studies have demonstrated that glucose intolerance, hypertension, and abdominal obesity often coexist in the same individuals increasing the risk for coronary heath disease (CHD). In 1988, Reaven focused on this cluster of cardiovascular risk factors, named it as "syndrome X", and proposed insulin resistance as the possible common etiological factor [1]. The relationship between insulin resistance and metabolic risk factors is not fully understood and appears complex. Most persons with insulin resistance have abdominal obesity, and dysfunctional adipose tissue might be the common link in determining this cluster of risk factors, currently named as metabolic syndrome (MetS).

Among the clinical criteria proposed for the diagnosis of MetS, the one provided by the National Cholesterol Education Program-Adult Treatment Panel III (NCEP-ATPIII) is easily applicable, and identifies patients with different combinations of hypertension, atherogenic dyslipidemia, impaired glucose homeostasis, and visceral obesity [2]. It is difficult to dissect the possible contribution of each MetS component to CHD risk; however, there is little doubt that in individuals with MetS the CHD risk is increased, and that insulin resistance enhances the risk of developing type 2 diabetes.

The pathogenesis of MetS has been associated with the effect of a genetic predisposition [3] in combination with environmental factors. Considering the central role of adipose tissue in MetS, different adipocyte related genes have been studied as possible candidates in MetS, including peroxisome proliferator-activated receptor-γ (PPARγ), and renin-angiotensin system related genes. PPARγ transcriptionally regulates the expression of genes involved in cell metabolism; endogenous ligands are thought to bind PPARγ and promote downstream gene target transcription [4, 5]. From human PPARγ gene two distinct isoforms of mRNA and protein, PPARγ1 and PPARγ2, are spliced [6]. PPARγ 1 is abundantly expressed in different tissues including adipose tissue and macrophages [7], whereas PPARγ 2 expression is restricted only to adipose tissue. In vivo and in vitro studies demonstrated that PPARγ has a critical role in regulating adipocyte differentiation and lipid accumulation [8, 9]; furthermore, its has been implicated in the regulation of lipid homeostasis and insulin sensitivity [10]. Not surprisingly, PPARγ has been identified as the target for thiazolidinediones [11], drugs that improve insulin sensitivity [10]. The role of PPARγ in affecting insulin action has been confirmed by several genetic studies on PPARγ gene polymorphisms; among these, the Pro12Ala mutation in PPARγ 2 (Pro12Ala) is the most common [12]. The Ala12 variant has been associated with decreased body mass index (BMI) [13–16], increased BMI [17, 18], and increased risk of obesity [19]. Different meta-analysis on Pro12Ala mutation and type 2 diabetes indicates that Ala12 carriers have a reduced risk of diabetes compared to Pro12 carriers [20, 21]; nevertheless, the association between Pro12Ala and type 2 diabetes seems to be modulated by genetic and environmental factors, dietary lipids [22], and intrauterine condition [23].

Recently, the Angiotensin Converting Enzyme (ACE) gene also received substantial attention as possible candidate for diabetes [24], hypertension [25], cardiovascular disease [26], and diabetic nephropathy [27]. Recent data suggest that the Renin Angiotensin System might be involved in the pathophysiology of obesity and associated hypertension [28]; therefore, ACE gene might be a good candidate for MetS. The most common ACE gene polymorphism is the insertion-deletion (I/D) located in intron 16; the D allele has been associated with higher levels of circulating ACE [29, 30]. Moreover PPARs modulate the renin-angiotensin-aldosterone system (RAAS), by transcriptional control of renin, angiotensinogen, angiotensin converting enzyme (ACE) and angiotensin II receptor 1 (AT-R1) [31], linking biologically the Renin Angiotensin System with the PPARs. Despite data on the association between ACE gene and MetS or hypertension are not conclusive, it appears interesting to evaluate the possible interaction of these two common polymorphisms.

In the present study we investigated, in a cohort of Italian adult individuals, the possible association between PPARγ2 Pro12Ala and ACE I/D polymorphisms and MetS components; furthermore, we searched for a possible interaction between these two common genetic variants.

Compared with Pro/Pro homozygotes (PP), PPARγ2 Ala carriers were characterized by higher BMI (p: 0.015) and higher fat-mass (p:0.016), but lower systolic blood pressure (p:0.033). No significant differences in waist circumference and other metabolic features were observed. Subjects carrying different ACE polymorphisms did not differ in terms of BMI, waist circumference, blood pressure, fasting glucose-insulin levels, and lipid profile.

The combined effect of PPARγ-ACE genotypes on BMI, fat-mass and waist circumference is shown in Figure 1. In individuals homozygous for ACE D allele, those bearing a PPARγ Ala allele (XA/DD) showed higher BMI (p:0.002) and higher fat-mass (p:0.002).

Figure 2 reports the two linear regressions for BMI vs waist circumference according to PPARg2 Pro12Ala genotypes (PP and XA). BMI correlated with waist circumference in both group; however, the slope tend to be lower in XA compared with PP individuals (Slope XA 1,817 ± 0,2275, slope PP 1,975 ± 0,07910, p:0.53).

By multivariate linear regression analysis we were able to demonstrate that the PPARγ2 Pro12Ala polymorphism (as categorical variable: PP:0; XA:1) was significantly associated with waist circumference (β coefficient: -2.73; p: 0.04) independent of BMI (β coefficient: 2.03; p: 0.001), and gender (β coefficient: 8.75; p: 0.001) (p for model: 0.001).

We evaluated the possible association between PPARγ pro12Ala and ACE I/D gene polymorphisms and MetS, fat-mass, and fat distribution. As previously reported [17, 19, 35], we found that carriers of the Ala allele had higher BMI. Although some studies [13–16] have shown Pro12Ala to be associated with lower BMI, our results coincide with meta-analysis conducted by Masud et al. [18] that found Ala allele association with higher BMI, specifically in subject with mean BMI > 27, similar to our population BMI. We also found a significant gene-gender interaction with BMI and fat-mass significantly higher in women but not men. This is in contrast with the data recently reported by Morini et al. [36], in a population with the same genetic background. We cannot exclude that the lack of association in men might be due to the limited number of males and the consequent lack of statistical power.

Despite having higher BMI, XA subjects showed similar waist, fasting glucose, and fasting insulin compared to PP subjects. Moreover XA individuals showed lower systolic blood pressure. It appears from our data that the increase in BMI associated with PPARγ2 Ala genotype was not associated with a worse metabolic profile. We speculate that Ala genotype might reduce central fat distribution thus protecting from metabolic complication associated with increased BMI. This would be in good agreement with a previous report in a Spanish population [37]. This hypothesis would be supported by the finding that the regression between BMI and waist had a lower slope in XA individuals; despite not being significant. This result suggests that, for a given BMI, subjects carrying the Ala allele might have less visceral fat. Moreover, multiple regression analysis showed that Pro12Ala polymorphism was an independent predictor of waist circumference.

Several human and animal studies support a central role of PPARγ in adipose tissue distribution [38, 39]. Ala genotype is considered to reduce PPARγ transcriptional activity; this reduced activity has been considered pivotal in reducing BMI and, ultimately, in the increasing insulin sensitivity in subjects with Ala allele [15, 22]. However, the relationship between PPARγ activity and insulin sensitivity is far from being linear, and does not necessarily pass through quantitative modification of fat mass. Mechanisms regulating energy balance and metabolic profile involve complex interactions between genetic, environmental and behavioral factors. Genotype may predispose to site-specific adiposity, culminating in a state of energy imbalance. PPARγ appears to be a key regulator of energy balance, with polymorphisms on the PPARγ gene linked to obesity and effects on body composition [40–42]. Humans harboring loss of function PPARγ mutation (P467L) show a pattern of lipodystrophy and are insulin resistant [39]. Qualitative adipose tissue modification and altered adipose tissue distribution are key features in determining metabolic abnormalities in these subjects. In this view, our data support the idea that the increased BMI found in subjects with Ala allele may not be detrimental, similarly to what happens in TZD treated patients showing metabolic improvements despite increased body weight. Indeed, XA individuals had higher BMI but lower systolic blood pressure and similar lipid profile compared to PP subjects.

We couldn't find any difference between subjects with different ACE I/D polymorphism. Studies published so far reported both an association with MetS/hypertension or lack of association [43, 44]. This inconstancy might depend from ethnic background, type of population (diabetic or no diabetic subjects), and diagnostic criteria chosen to define MetS. The NCEP ATPIII criteria are operatively simple to use. Nevertheless, individuals can be classified as hypertensive or diabetic only on the basis of the active use of antidiabetic or antihypertensive drugs, including ACE inhibitors. These drugs might be a confounder when analyzing the possible effect of ACE gene polymorphism. When we selected only pharmacotherapy naives subjects, we found a significantly lower prevalence of MetS in subjects carrying the ACE II compared to DD genotype.

We also found a significant gene interaction between ACE I/D and PPARγ2 Pro12Ala polymorphism on BMI and fat-mass, with subjects XA/DD being the ones with higher values. This is not surprising as recent studies showed a role of ACE gene in regulating body weight; indeed, an ACE-/- knock out mouse had reduced fat-mass and improved insulin sensitivity [45]. Moreover, Eisenmann et al. recently reported that children harboring the DD polymorphism have an higher BMI compared to II carriers [46]. From a mechanistic point of view, ACE DD subjects have higher levels of plasma ACE, and therefore potentially higher level of Angiotensin II which has been identified as a trophic factor in the differentiation of preadipocytes to mature adipocytes [47].

In conclusion we found that, despite a lack of association between Pro12Ala and ACE I/D polymorphism and MetS diagnosis, subjects carrying a PPARγ2 Ala allele have higher BMI and fat-mass but do not have a worse metabolic profile, probably because a more favorable distribution of adipose tissue. Pro12Ala polymorphism effect on BMI is not constant across gender, with significant association only found in woman. It exists a gene interaction between the two common polymorphism Pro12Ala and ACE I/D on BMI and fat mass. Further studies are needed to assess the real contribution of the PPARγPro12Ala polymorphism in determining adipose tissue distribution.