Peroxisome Proliferator-Activated Receptor-γ2 Polymorphism Pro12Ala Is Associated With Nephropathy in Type 2 Diabetes: The Berlin Diabetes Mellitus (BeDiaM) Study

Patients who had type 2 diabetes (n = 445) and were of Caucasian ethnicity were recruited from three diabetes and two dialysis centers as well as two general hospitals in Berlin within the framework of the Berlin Diabetes Mellitus (BeDiaM) Study. Classification of type 2 diabetes relied on the revised diagnostic criteria recommended by the American Diabetes Association (17). The diagnosis of nephropathy was based on repeated observation of UAE of >20 μg/min in nonoliguric patients or on the presence of ESRF necessitating renal replacement therapy (dialysis or transplantation). Macroalbuminuria was defined as UAE >200 μg/min. Other causes of increased albumin excretion were excluded by appropriate clinical and technical investigation. In agreement with recently published World Health Organization criteria (18), hypertension was defined as a systolic pressure of ≥140 mmHg or diastolic pressure of ≥90 mmHg on at least two separate occasions or as the necessity to prescribe antihypertensive drugs, irrespective of the blood pressure level. An investigator unaware of the patient’s genotype searched the medical records for history of CAD, stroke, or retinopathy. The diagnosis of CAD required a history of myocardial infarction, coronary angioplasty or bypass surgery, coronary lesions on angiography, or treatment with nitrates or a clinical history of angina pectoris. Retinopathy was defined as stages III or IV on funduscopy or as history of laser therapy. Serum creatinine, serum lipids, and glycosylated hemoglobin (HbA_1c) were determined by standard laboratory techniques.

The study protocol was approved by the institutional Ethics Committee. All subjects gave informed consent before their enrollment in the study.

We used SAS version 8.1 (SAS Institute, Cary, NC) for database management and statistical analysis. Measurements with a skewed distribution were normalized by logarithmic transformation. Comparisons of means and proportions were performed with the standard normal Z test and Fisher’s exact test, respectively. Significant covariables were traced by stepwise linear or logistic regression. We used analysis of covariance to compare adjusted continuous measurements between genotypes. The presence of mild renal dysfunction in relation to genotype was studied by multiple logistic regression.

Genomic DNA was prepared from peripheral blood using a DNA-selective preparation method (Qiagen, Hilden, Germany). The region encompassing the Pro12Ala polymorphic site was amplified by PCR as previously described (16). PCR products were digested using the restriction enzyme BstU-1 and visualized on ultraviolet-transilluminated ethidium bromide-stained agarose gels.

The clinical characteristics of 220 men and 225 women with type 2 diabetes are presented in Table 1. On average, women were older than men and had a higher serum creatinine concentration. Women were more frequently hypertensive but in fewer instances proceeded to ESRF.

The genotypic distribution of the PPAR-γ2 Pro12Ala gene polymorphism was in Hardy-Weinberg equilibrium (P = 0.42), and the Ala12 allele frequency in our diabetes population was 0.14 (Table 2).

Among our patients with type 2 diabetes, the 118 Ala12 carriers showed a significantly lower UAE than the 327 noncarriers (17.1 μg/min [95% confidence interval (CI) 13.5–21.8 μg/min] vs. 25.8 μg/min [95% CI 21.9–30.4 μg/min]; P = 0.01) (data not shown). With adjustment for sex, age, BMI, hypertension, hyperlipidemia, duration of diabetes, and treatment with insulin or oral antidiabetic drugs, the percentage difference was 32% (95% CI 8–50%; P = 0.01) (Fig. 1). Furthermore, the percentage difference in UAE between Ala12 allele carriers and noncarriers increased (P = 0.003) with longer duration of diabetes (Fig. 1). With similar adjustments as mentioned above, the percentage difference was 54% in patients with diabetes and a disease history from 10 to 19 years (95% CI 20–74%; P = 0.008) and 70% in those who had diabetes for ≥20 years (95% CI 24–88%; P = 0.01). In contrast, among those patients in whom diabetes was diagnosed within the past 10 years, there was no difference in UAE with respect to the different Pro12Ala genotypes (P > 0.99) (Fig. 1).

We also calculated the odds ratio of having proteinuria coded as a dichotomous trait (0, 1) with adjustments for the same covariables as in the continuous analysis. In all 445 patients, the odds ratio was 0.66 for Ala12 allele carriers compared with noncarriers (95% CI 0.42–1.04; P = 0.07) (Table 3), which just failed to reach statistical significance because of a significant interaction (P = 0.003) (Fig. 2) between the Pro12Ala polymorphism and the duration of diabetes. Indeed, the odds ratio decreased from 1.22 (95% CI 0.64–2.32; P = 0.54) to 0.11 (95% CI 0.02–0.55; P = 0.007) in patients in whom the disease history was <10 years compared with those who had diabetes for ≥20 years.

There was no significant association of the PPAR-γ2 Pro12Ala polymorphism with any other measured phenotype in the BeDiaM Study, including ESRF or diabetic retinopathy, before and after allowing for the duration of diabetes.

The main finding of the present study was that among patients with type 2 diabetes, PPAR-γ2 Ala12 allele carriers had significantly lower UAE and presented less frequently with macroalbuminuria. This suggests a role of genetic variation in the PPAR-γ2 gene in the development of diabetes-related nephropathy. The strength of the association increased with longer duration of diabetes, a phenotype that is likely to represent a surrogate of environmental factors that over time modulate the genetic impact on diabetic nephropathy. The reason that we were not able to show an association between ESRF and Pro12Ala in our patients with type 2 diabetes might be related to the fact that ESRF is a more complex phenotype that is determined by a plethora of different pathophysiological as well as genetic factors. However, patients with type 2 diabetes have an excessive cardiovascular risk, leading to a dramatically reduced survival before ESRF develops (6,19), therefore introducing a bias toward low ESRF rates that hampers the unraveling of the underlying genetic risk factors. In the development of diabetic nephropathy, morphologic changes such as thickening of the glomerular basement membrane, glomerular hypertrophy, mesangial expansion, and modest expansion of the tubulointerstitium play a role. All of these factors lead to microalbuminuria as a diagnostic hallmark of diabetic nephropathy, preceding the development of overt proteinuria (20).

The risk of developing diabetic nephropathy is likely to be genetically determined, because not all patients share the same renal outcome even when obviously presenting with the same degree of metabolic dysregulation. In this regard, the prevalence of microalbuminuria has been reported to be significantly higher in Mexican-American than in non-Hispanic white diabetic subjects (21,22), and higher UAEs have been observed in offspring of patients with type 2 diabetes (23). Strojek et al. (8) recently reported higher UAE rates in healthy offspring of diabetic parents with nephropathy versus those without nephropathy. Furthermore, the risk of developing diabetic nephropathy has been linked to different chromosomes, including chromosome 3 (10), on which the gene encoding PPAR-γ2 is located.

Recently developed compounds such as TZDs act as potent PPAR-γ ligands and have been shown to ameliorate microalbuminuria in patients with early diabetic nephropathy (15). Nakamura et al. (24) recently observed that TZDs reduce UAE in microalbuminuric patients as well as the serum concentration of type IV collagen, whereas glibenclamide showed no such effects. Other investigators reported that TZDs lower microalbuminuria, possibly by inhibiting platelet-derived growth factor-stimulated cell growth and angiotensin II-induced plasminogen-activator inhibitor (PAI)-1 expression, both of which are known to contribute to the development of nephropathy (25).

The mechanisms by which PPAR-γ2 and variation in its gene might actually protect against diabetic nephropathy remain to be elucidated. Even if glomerular expression of PPAR-γ has been reported (26,27), the major sites of PPAR-γ2 expression are adipocytes (28). Adipocytes generate and secrete PAI-1 (29), angiotensin II, and type IV collagen (30). In this respect, it is conceivable that a reduced level of PPAR-γ2 activation in preadipocytes, leading to reduced adipocyte differentiation, may be responsible for reduced levels of PAI-1, angiotensin II, and type IV collagen in patients who carry the Ala12 allele of PPAR-γ2, possibly resulting in renal protection.

In agreement with several other association studies (31,32, 33), on the PPAR-γ2 Pro12 allele and type 2 diabetes, our results speak in favor of a protective effect of the Ala12 allele on diabetes-associated phenotypes, i.e., nephropathy. The molecular mechanism by which the Ala12 carrying PPAR-γ2 protein might achieve these beneficial actions, which largely resemble the specific effects of TZDs on diabetic nephropathy, remains to be elucidated.

In conclusion, among patients with type 2 diabetes, PPAR-γ2 Ala12 allele carriers have significantly lower UAE and develop overt proteinuria less frequently. These observations suggest a protective effect of the Ala12 allele in relation to diabetic nephropathy. Whether TZDs as PPAR-γ ligands might be more effective in patients with type 2 diabetes who carry the PPAR-γ2 Pro12 risk allele remains to be investigated in future clinical studies.

This study was supported in part by a research grant from the Universitätsklinikum Benjamin Franklin, Berlin (FSP-EB 1999/2000). J.-G.W. was supported by the bilateral scientific and technical collaboration between the People’s Republic of China and Flanders, Belgium (BIL 98/ 15). S.-M.H. is a participant in the grants of the Deutsche Forschungsgemeinschaft: “Graduierten-Kolleg 426/2-00, Molekularbiologische Grundlagen der Therapie,” and “ Graduierten-Kolleg 754, Myokardiale Genexpression und Funktion, Myokardhypertrophie.”

We are grateful to Dr. Elke Austenat of the Diabetes Nachtklinik Tempelhof and Prof. Dr. Wolfgang Pommer of the Humbold Krankenhaus, Berlin, Germany, for support in the recruitment of patients. We thank Brigitte Egbers and Klaus Schlotter for expert technical assistance in genotyping.