Polymorphism of the 3′-Untranslated Region of the Leptin Receptor Gene, but Not the Adiponectin SNP45 Polymorphism, Predicts Type 2 Diabetes: A population-based study

Anthropometric and clinical parameters were measured as described elsewhere (12). At baseline and follow-up, case subjects were classified as having IGT or type 2 diabetes, according to American Diabetes Association criteria (13). The present analysis is based on the 1,485 subjects in whom DNA was available. In 1993, a first follow-up screening was performed; in 1997, 1,141 subjects completed a second follow-up examination. The protocol was approved by the American British Cowdray Hospital in Mexico City, and all the subjects gave informed consent.

DNA was isolated from peripheral blood leukocytes (14), and a fragment in the LEPR of 114 (D/D) or 119 (I/I) bp was generated by PCR using the forward primer 5′-CATGCCTCAATTCCAA AC-3′ and reverse primer 5′-GAGAGAA CAAACAGACAACATT-3′. Amplification was performed using 200 ng DNA and 1 μmol/l of each primer using a GeneAmp PCR-System 2400 (Perkin-Elmer, Norwalk, CT). Then, amplicons were analyzed by single-strand conformation polymorphism, performing an electrophoresis on 6% polyacrylamide gel with 10% glycerol at room temperature and silver stained. Fragments showing variations were reamplified by PCR and electrophoresed onto 3.5% agarose gels. A difference of 5 bp between the carriers of the insertion allele (D/I or I/I) and carriers of the deletion allele (D/D) was revealed.

A DNA fragment of 264 bp was generated by PCR using the primer 5′-TGT GTGTGTGGGGTCTGTCT-3′ and 5′-TTCTCACCCTTCTCACCAGG-3′ for SNP45-APM1. PCR was performed using 100 ng DNA and 0.2 μmol/l of each primer. Then, PCR products were digested with BspH I enzyme and electrophoresed on 6% polyacrylamide gel and silver stained. Wild-type allele (T/T) was digested in fragments of 158 and 106 bp, whereas mutated alleles (G/G) were not digested (264 bp).

Logistic regression was carried out by standard methods; results are expressed as the odds ratio (OR) with 95% CIs. Genetic data are presented according to both genotype and allele frequency. χ² analysis was used to test for Hardy-Weinberg equilibrium within the normal glucose tolerance (NGT), IGT, and type 2 diabetic groups.

The distribution of the I/D-LEPR genotypes was in Hardy-Weinberg equilibrium in all groups (χ² = 0.17, P = NS). The I/I genotype tended to decrease across IGT and type 2 diabetes (as did the allele frequency), but the trend fell short of statistical significance (Table 1). The SNP45-APM1 genotypes also were in Hardy-Weinberg equilibrium in all three groups (χ² = 0.4, P = NS). No difference was found in either genotype distribution or allele frequency across glucose tolerance status or for obesity and body weight.

Over 7 years of follow-up, 95 subjects progressed to IGT and 107 developed type 2 diabetes. The I allele was significantly less expressed in subjects who progressed to type 2 diabetes than in subjects who remained normal glucose tolerant (χ² = 5.8, P = 0.016). In contrast, no association was found between the SNP45-APM1 and incident IGT or type 2 diabetes (Table 1). For type 2 diabetes, incidence was twice as high in DD (1.6% per year) than II (0.8% per year).

In a multiple logistic regression model adjusting for sex, age, BMI, waist circumference, blood pressure, triglyceride, HDL cholesterol, and 2-h plasma insulin concentrations, the presence of the I allele of LEPR was a significant predictor of the combined outcome of type 2 diabetes or IGT (OR 0.66 [95% CI 0.45–0.96]) together with BMI (1.32 [1.16–1.50] for every three BMI units), systolic blood pressure (1.17 [1.05–1.29] for every 10 mmHg), 2-h plasma insulin (1.06 [1.04–1.09] for every 60 pmol/l), or HDL cholesterol 0.87 [0.79- 0.97] for every 5 mg/dl). In the same model, the SNP45-APM1 polymorphism was unrelated to worsening glucose tolerance.

The main findings of the present study were that the 1) I allele of the 3′UTR-LEPR gene signaled a reduced risk of development of diabetes, 2) the SNP45-APM1 did not predict incident diabetes, and 3) no effect of either LEPR or APM1 polymorphism on obesity and body weight was found in this population.

Several polymorphisms in LEPR have been identified (15–18), and recently, two studies carried out in obese subjects reported that heterozygous carriers of the I/D-LEPR had lower serum insulin concentrations than DD carriers (2,3); this finding has been interpreted to indicate that defects of LEPR could play a role in the etiology of diabetes rather than obesity (2). In agreement with this suggestion, Lakka et al. (4) described an association between I/D-LEPR and risk of type 2 diabetes in a prospective, case-control study in healthy men.

In our general population, the association between I/D-LEPR and prevalent IGT or type 2 diabetes did not reach statistical significance, and we did not find significant associations with other features of the metabolic syndrome; however, the I allele did predict incident diabetes independently of confounders, in agreement with data from Lakka et al. (4).

The mechanism underlying the association between this polymorphism and the risk of diabetes is not clear. Leptin receptors are present on pancreatic β-cells, whereby leptin could modulate glucose-induced insulin secretion (15,16). It has been suggested that I/D-LEPR could affect RNA stability and, possibly, the rate of degradation and/or translation of mRNA (3,17). However, direct data on the effects of the polymorphism on leptin action or insulin secretion are not available.

Recently, SNP45 and -276, in exon 2 and intron 2 of APM1, respectively, have been found to be associated with type 2 diabetes in Japanese patients (9), but this association has not been confirmed in French or Swedish populations (18,19). We therefore evaluated the role of SNP45-APM1 on the susceptibility to diabetes and/or the linkage with I/D-LEPR. The current results were negative on the prediction of diabetes. However, it is important to note that in complex traits such as diabetes, the absence of association does not necessarily indicate a lack of effect because the susceptibility to disease may be conferred by combinations of variants rather than a single SNP.