The increased morbidity related to obesity may be largely accounted for by the presence of several risk factors defining the metabolic syndrome. The National Cholesterol Education Program (NCEP)1-Adult Treatment Panel III (ATPIII) has proposed to characterize the metabolic syndrome as the presence of three or more of the following components: abdominal obesity, high circulating concentrations of triglycerides, hypoalphalipoproteinemia, hypertension, and high fasting plasma glucose levels (1). Knowledge of factors determining the severity of the syndrome remains largely incomplete, but there is strong evidence for genetic susceptibilities in this pathophysiology (2).
Adipose tissue type 1 11β-hydroxysteroid dehydrogenase (11β-HSD1), which generates hormonally active cortisol from inactive cortisone, has been shown to play a central role in adipocyte differentiation and abdominal obesity-related metabolic complications (3). An accumulation of abdominal fat in the visceral compartment is observed in patients with Cushing's syndrome, suggesting that glucocorticoids play a potentially key role in the pathogenesis of visceral obesity (4). Moreover, transgenic mice overexpressing 11β-HSD1 in adipose tissue developed visceral obesity and an insulin-resistant state and hyperlipidemia, leading the investigators to suggest that 11β-HSD1 may be a common molecular etiology for visceral obesity and the metabolic syndrome (5). Thus, the gene coding for this enzyme represents a promising candidate for the metabolic syndrome. The aim of this study was to investigate whether genetic variants in the human 11β-HSD1 gene are associated with the metabolic syndrome among French-Canadian men.
Subjects’ characteristics are shown in Table 1. The molecular screening of the 11β-HSD1 gene revealed the presence of four variants (Table 2). Three of these variants were located in intronic regions (g.4478T>G, g.10733G>C, and g.4437-4438insA). One sequence variant was identified in exon 6. Based on the predicted amino acid sequence, this polymorphism is silent and does not alter the glycine at codon 248 (c.744G>C or G248G). Screening of the promoter region did not reveal any sequence variations. The relative allele frequency of these variants is shown in Table 2. Intronic variants identified were tested for linkage disequilibrium. The two mutations in intron 3 (g.4478T>G and g.4437-4438insA) were in perfect linkage disequilibrium (r2 = 1.0, p < 0.0001). Moreover, the single-nucleotide polymorphism (SNP) in intron 4 (g.10733G>C) was also in linkage disequilibrium with the SNP and the insertion in intron 3 (r2 = 0.83, p < 0.0001 for both).
|Variables||Mean ± SD||Range|
|Age (years)||42.8 ± 7.9||22.0 to 63.0|
|BMI (kg/m2)||29.6 ± 4.3||18.7 to 41.0|
|Waist circumference (cm)||101.4 ± 11.4||67.0 to 127.7|
|Total cholesterol (mM)||5.19 ± 0.76||3.57 to 7.26|
|Low-density lipoprotein-cholesterol (mM)||3.44 ± 0.74||1.45 to 5.26|
|High-density lipoprotein-cholesterol (mM)||0.92 ± 0.21||0.49 to 1.87|
|Total cholesterol/high-density lipoprotein-cholesterol||5.89 ± 1.44||2.47 to 10.80|
|Triglycerides (mM)||2.22 ± 1.10||0.45 to 7.56|
|Diastolic blood pressure (mm Hg)||81.93 ± 7.24||63.0 to 100.0|
|Systolic blood pressure (mm Hg)||122.30 ± 9.58||95.0 to 157.0|
|Sequence variant||Frequency of the rare allele|
|Intron 3 g.4478T>G||0.14|
|Intron 3 g.4437–4438insA||0.28|
|Intron 4 g.10733G>C||0.21|
|Exon 6 c.744G>C||0.005|
We then verified whether these variants were associated with components of the metabolic syndrome. Due to the very low frequency of the c.744G>C SNP (0.005 for 744C allele), association studies were not pursued further with this variant. Indices of obesity and plasma insulin and glucose levels in the fasting state and in response to an oral glucose tolerance test were not associated with any of the genetic variations identified (Table 3). For plasma lipoprotein/lipid levels, a significant difference was observed in plasma apolipoprotein B (apo B) levels. Indeed, carriers of the 10733C allele had plasma apo B levels of 1.13 ± 0.19, whereas G10733/G10733 homozygotes had plasma apo B levels of 1.07 ± 0.23 (p = 0.04). Similar results were observed for carriers of the 4478G and 4438insA alleles (p = 0.06 for both). This was the only association observed. Similar results were observed after adjustment for the effect of age, age and BMI, or age and visceral adipose tissue accumulation assessed by computed tomography (data not shown). We subsequently compared the frequency of the rare allele of each variant in subjects with and without the metabolic syndrome using criteria defined by the NCEP-ATPIII (1). The frequency of each rare allele (10733C, 4437-4438insA, and 4478G) in men with fewer than three components of the metabolic syndrome was similar to that in men with three or more components of the metabolic syndrome (χ2 = 0.01, p = 0.92; χ2 = 0.01, p = 0.91; χ2 = 0.01, p = 0.91, respectively).
|Variables||4478 T/T||4478 G carriers||Wild type (w/w)||4437–4438insA carriers||10733 G/G||10733 C carriers|
|BMI (kg/m2)||29.5 ± 4.3 (139)||29.6 ± 4.2 (78)||29.5 ± 4.3 (139)||29.6 ± 4.2 (78)||29.5 ± 4.4 (129)||29.7 ± 4.2 (88)|
|Waist circumference (cm)||101.3 ± 11.6 (138)||101.6 ± 11.0 (78)||101.3 ± 11.6 (138)||101.6 ± 11.0 (78)||101.0 ± 11.6 (128)||102.0 ± 11.1 (88)|
|VAT accumulation (cm3)||164.7 ± 63.0 (138)||176.1 ± 65.0 (76)||164.7 ± 63.0 (138)||176.1 ± 65.0 (76)||161.95 ± 59.89 (128)||178.87 ± 68.30 (86)|
|High-density lipoprotein-cholesterol (mM)||0.93 ± 0.22 (138)||0.90 ± 0.17 (76)||0.93 ± 0.22 (138)||0.90 ± 0.17 (76)||0.94 ± 0.21 (128)||0.89 ± 0.20 (86)|
|Triglycerides (mM)||2.19 ± 1.10 (138)||2.29 ± 1.11 (76)||2.19 ± 1.10 (138)||2.29 ± 1.11 (76)||2.15 ± 1.09 (128)||2.34 ± 1.13 (86)|
|Total apo B (g/L)||1.07 ± 0.23 (136)||1.13 ± 0.19 (75)||1.07 ± 0.23 (136)||1.13 ± 0.19 (75)||1.07 ± 0.23 (126)||1.13 ± 0.19 (85)*|
|Systolic blood pressure (mm Hg)||121.9 ± 9.6 (130)||123.1 ± 9.6 (70)||121.9 ± 9.6 (130)||123.1 ± 9.6 (70)||121.7 ± 9.9 (120)||123.2 ± 9.1 (80)|
|Diastolic blood pressure (mm Hg)||82.2 ± 7.5 (130)||81.5 ± 6.7 (70)||82.2 ± 7.5 (130)||81.5 ± 6.7 (70)||82.1 ± 7.6 (120)||81.7 ± 6.7 (80)|
|Fasting glucose (mM)||5.47 ± 0.56 (138)||5.51 ± 0.59 (78)||5.47 ± 0.56 (138)||5.51 ± 0.59 (78)||5.46 ± 0.55 (128)||5.51 ± 0.60 (88)|
|Fasting insulin (pM)||104.3 ± 78.9 (138)||102.6 ± 64.3 (78)||104.3 ± 78.9 (138)||102.6 ± 64.3 (78)||99.6 ± 63.9 (128)||109.6 ± 86.3 (88)|
|HOMA IR||3.64 ± 2.98 (138)||3.56 ± 2.63 (78)||3.64 ± 2.98 (138)||3.56 ± 2.63 (78)||3.49 ± 2.65 (128)||3.79 ± 3.14 (88)|
Because the gene encoding the 11β-HSD1 has been shown to be involved in abdominal obesity and its related complications, it was relevant to search for molecular variations that might have an impact on features of the metabolic syndrome. The present study addressed this question, and we report the identification of four 11β-HSD1 variants. Two of these sequence variants were SNPs in intronic regions of the 11β-HSD1 gene (g.4478T>G and g.10733G>C). One adenine insertion (g.4437-4438insA) was identified in intron 3. The present study also reports one sequence variation in exon 6 that results in a nucleotide substitution (c.744G>C). However, this variant does not alter the amino acid sequence. The allele frequency of each polymorphism was determined in a sample of French-Canadian men. The genotype frequency for all variants identified did not deviate from Hardy-Weinberg predictions. To ascertain the contribution of this gene to the development of the metabolic syndrome, we tested linkage disequilibrium of the three intronic variants and performed association studies. The three SNPs were in linkage disequilibrium. Further analyses did not reveal any association among these mutations and components of the metabolic syndrome as defined by the NCEP-ATPIII, suggesting that these variants are unlikely to be the molecular defect explaining the observed phenotype. Among all variables tested, apo B level was the only one to be significantly different between genotype groups. Power analyses revealed that at an α-level of 0.05, plasma apo B was the variable with the highest power to detect this association. At this point, we cannot exclude the possibility that the lack of association with other variables is the reflection of a low power rather than an absence of association.
The insertion in intron 3 has been previously reported by Gelernter-Yaniv et al. (6). They observed in a study sample of healthy overweight and normal weight children from different racial backgrounds that this variant was associated with greater body mass and with altered body composition and insulin resistance. Differences in the genetic background of these populations may explain inconsistencies observed between the present study and the study by Gelernter-Yaniv et al. As mentioned by the authors, it is also possible that this association may be due to linkage disequilibrium with another mutation in a gene located in the vicinity of the 11β-HSD1 gene, which is not found in the French-Canadian population. In addition, the lack of association might be attributable to the low number of subjects on which the present study is based, although Gelernter-Yaniv et al. have reported a positive association in a relatively similar sample size (263 vs. 217) (6). Caramelli et al. have also identified two variants in 12 obese and nonobese Italian women. One of them is the insertion in intron 3, and the second is an 11-bp deletion in intron 1 (7). These variants were found in either abdominally obese subjects or control subjects, suggesting that these molecular defects were not involved in the etiology of visceral obesity. In the present study, we did not observe the deletion in intron 1, suggesting that this variant may be specific to the Italian population. The 11β-HSD1 gene has been screened previously in a woman characterized by central obesity and other hormonal dysfunctions (8). The authors were not able to associate genetic variations in the 11β-HSD1 gene with these metabolic alterations. Both variants identified in intron 3 have been previously described in individuals with cortisone reductase deficiency (9). Transcriptional activity assays have demonstrated 2.5 times lower 11β-HSD1 activity for the two mutated HSD11B1 constructs. However, the authors suggested that these mutations were not the molecular defect explaining the cortisone reductase deficiency because 25% of unaffected controls were heterozygous, and 3% were homozygous for these polymorphisms. They also identified two mutations in the hexose-6-phosphate dehydrogenase gene (H6PD 620ins29bp621 and H6PD R453Q) and showed epistasis with variants identified in the 11β-HSD1 gene to cause the cortisone reductase deficiency. At this point, we cannot exclude the possibility that mutations in the hexose-6-phosphate dehydrogenase gene might interact with variants identified in 11β-HSD1 to influence the development of the metabolic syndrome.
In conclusion, results obtained in the present study come at a time of great interest in the role of 11β-HSD1 in central obesity and its complications. This study did not reveal the presence of genetic variants in the 11β-HSD1 gene that could help understand the impact of local excess glucocorticoids in the metabolic syndrome among French-Canadian men. However, further studies involving 11β-HSD1 activity measurements and cortisol/cortisone metabolites are clearly needed to confirm results of the present study. In addition, these data do not exclude the possibility that genetic variations in the 11β-HSD1 gene may be associated with susceptibility to abdominal obesity and related metabolic complications in other populations. The identification of additional genetic variants in the 11β-HSD1 gene provides helpful tools for further investigations of association with the metabolic syndrome to confirm these results.