OBJECTIVE: The effects of alpha-2A (A2A)-, beta-2 (B2)- and beta-3 (B3)-adrenergic receptor (ADR) gene polymorphisms on adiposity, fat distribution and plasma insulin and leptin changes in response to long-term overfeeding were explored.
METHODS: Twenty four men (mean (±s.d.) age 21±2 y) who constituted 12 pairs of identical twins ate a 4.2 MJ/day energy surplus, 6 days a week, for a period of 100 days. Total body fat was assessed by hydrodensitometry and total subcutaneous fat by the sum of eight skinfolds. Abdominal fat areas were measured by computerized tomography (CT). Plasma glucose and insulin during fasting and in response to an oral glucose tolerance test (OGTT) were assayed. The insulin and glucose areas were computed using the trapezoidal method. Plasma leptin was measured with an enzyme-linked immunosorbent assay. The ADR polymorphisms were identified by PCR or Southern blot technique.
RESULTS: The ADRB2 Gln27Gln genotype (n=10) was associated with a larger gain (percentage change) in weight (P<0.001) and total subcutaneous (P<0.005) fat than the Glu27Glu/Gln27Glu genotype (n=14). In addition, overfeeding induced greater increases in the insulin areas under the curve during the OGTT and the fasting plasma level of leptin (P<0.01 and <0.03, respectively) among Gln27Gln than in the Glu27Glu/Gln27Glu subjects. The body composition and metabolic changes among the ADRB2 BanI 3.7/3.4 kb subjects (n=10) were similar to those of Gln27Gln subjects. ADRA2A DraI (n=4) 6.3/6.3 kb subjects experienced a decrease (−8%) while 6.7/6.3 kb subjects (n=20) registered an increase (+10%; P=0.017) of OGTT glucose area after the 100-day caloric surplus. The four carriers of the ADRB3 variant (Trp64Arg) experienced the same magnitude of changes as the 20 homozygotes for the Trp allele. In general, comparisons based on the 24 subjects considered as unrelated men and the mean values for each of the 12 pairs yielded similar results.
CONCLUSION: The ADRB2 Gln27Gln subjects gained more weight and total subcutaneous fatness and also experienced a greater increase in insulin resistance than Glu27Glu/Gln27Glu subjects with exposure to long-term overfeeding. Similar differences were observed between carriers and non-carriers of the ADRB2 3.7/3.4 kb BanI variant. Genetic variation at the ADRB2 locus could thus be one of the factors responsible for the large inter-individual differences observed in the response to long-term alterations in energy balance and should be further investigated.
There are individual differences in the response to various dietary changes. However, the genetic basis of the differences in responsiveness to changing energy balance conditions is largely unknown. Genes encoding proteins that are key regulators of lipid oxidation rates and energy expenditure are likely involved in the modulation of the response to environmental pressures causing weight gain. Human adipocytes express adrenergic receptors (ADR) which are able to stimulate (B1, B2 and B3)1 or inhibit (A2)2 lipolysis. The ADR genes can thus be involved in the responsiveness to dietary changes. Indeed, ADRA2A and ADRB2 mutations have been reported to be associated with adiposity in some3,4,5,6 but not all7,8,9 studies and the association of ADRB3 gene with obesity remains controversial.10
The aim of this study was to explore the role of polymorphisms in the ADRA2A, ADRB2 and ADRB3 genes in the changes observed in response to a long-term (100 days) overfeeding protocol conducted with 12 pairs of monozygotic twins. Despite the fact that only 24 young men (or 12 pairs of twins) were challenged by the overfeeding protocol, the data can be useful in the effort to identify genes and mutations that should be tested in future studies based on large sample sizes.
The specific aims, study design and methodology of this overfeeding protocol have been described elsewhere.11 Briefly, 24 sedentary young men (mean (±s.d.), age 21.0±2.0 y and mean body mass index (BMI) 19.7±2.0 kg/m2), which constituted 12 pairs of healthy identical twins were studied. Here the twins are considered as 24 subjects and as 12 pairs. The men were housed in a closed section of a dormitory on the campus of Laval University. Each man stayed in the unit for 120 consecutive days: 14 days for the assessment of baseline daily energy intake, 3 days for testing before the period of overfeeding, 100 days for the period of overfeeding, and 3 days for testing after the period of overfeeding. During overfeeding, the men were fed a diet containing 4.2 MJ (1000 kcal) per day above their measured baseline energy intake, 6 days a week, for 100 days. On the seventh day of each week, they were given the baseline amount of calories.
The BMI was calculated as body weight (in kilograms) divided by height (in meters squared). Body density was determined by underwater weighing, and percentage body fat was calculated with a standard equation.11 The skinfold thickness was measured at eight sites (biceps, triceps, front midthigh, medial calf, subscapular, suprailiac, abdominal and midaxillary) according to the procedures recommended at the Airlie Conference.12 Abdominal computerized tomography (CT) scanning was performed before and after the overfeeding period with a Siemens Somatom DRH scanner (Erlangen, Germany) as reported earlier.11
On the first two days of the 3-day test period before the overfeeding period, blood samples were obtained in the morning after an overnight fast for the determination of plasma glucose and insulin concentrations. On day 2, a 75 g oral glucose tolerance test (OGTT) was performed. Blood samples were obtained every 15 min during the first hour and every 30 min during the next 2 h for the determination of plasma glucose and insulin. The same measurements were repeated after the overfeeding period. Plasma glucose and insulin were assayed as described earlier13 and plasma leptin was measured with an enzyme-linked immunosorbent assay specific for the human peptide.14
The ADRB2 Arg16Gly and Gln27Glu as well as ADRB3 Trp64Arg polymorphisms and PCR technique by which they were uncovered were described earlier.15 The ADRA2A DraI and ADRB2 BanI polymorphisms were analyzed by Southern blot analysis technique.15
Differences in phenotype changes (percentage) between genotypes at a given marker were assessed by a t-test. Percentage changes were calculated from individual scores. Analyses were performed both with the 24 subjects considered as unrelated persons and the phenotype mean of each of the 12 pairs. Statistical analyses were performed with the SAS statistical package (SAS Institute, Cary, NC).
Changes in body weight, body composition and fat distribution with overfeeding in this study of identical twins have been reported previously.11
No differences in response to overfeeding were found between ADRB2 Arg16Gly genotypes. The changes with overfeeding in body weight, fat mass, plasma leptin, insulin area under the curve during the OGTT and measures of subcutaneous fat and CT-abdominal adiposity for the ADRB2 Gln27Glu genotypes are shown in Table 1. Because of the low number of Glu27 homozygotes (n=2), they were pooled together with the Gln27Glu subjects in all analyses. Overfeeding induced greater increases in body weight (P<0.001) in the Glu27Glu/Gln27Glu (n=14) than in the Gln27Gln (n=10) subjects. In addition, total subcutaneous fat (P<0.005) and plasma levels of leptin (P<0.03) increased more in the Glu27Glu/Gln27Glu than in the Gln27Gln subjects (Table 1). The Glu27Glu/Gln27Glu genotype also experienced a greater increase in insulin area under the curve during the OGTT (P<0.01) than the Gln27Gln genotype. The overfeeding-induced differences for percentage changes in fasting plasma insulin, OGTT glucose areas or abdominal visceral fat between the genotypes were not significant (data not shown). Before overfeeding, only plasma leptin levels (P<0.05) were significantly different between the genotypes. However, after overfeeding, Gln27Gln subjects showed higher levels of fat mass (P<0.005), plasma leptin (P<0.005), total subcutaneous (P<0.03) and abdominal fat (P<0.01) than Glu27Glu/Gln27Glu subjects.
Table 2 shows the changes with overfeeding for the same phenotypes for the two ADRB2 BanI genotypes. No homozygotes for the 3.7 kb allele were observed. Overfeeding induced a greater increase in body weight (P=0.021) in the 3.7/3.4 kb (n=10) than in the 3.4/3.4 kb (n=14) subjects. Moreover, plasma levels of leptin (P<0.001) and insulin areas under the curve during the OGTT (P=0.031) increased more in the 3.7/3.4 kb than in the 3.4/3.4 kb subjects (Table 2). The 3.7/3.4 kb genotype also experienced a greater increase in total subcutaneous fat (P<0.001) and abdominal fat (P=0.047) than the 3.4/3.4 kb genotype in response to overfeeding. Although the changes in response to overfeeding were significant, none of the phenotypes showed significant differences between the two BanI genotypes before or after overfeeding.
The number of carriers for the rare alleles of the ADRA2A DraI (n=4) and ADRB3 Trp64Arg (n=4) polymorphisms were low and no differences in response to overfeeding were found between genotypes at these two genes, except in one case. The ADRA2A DraI 6.3/6.3 kb subjects (n=4) experienced a decrease (−8%) and the 6.7/6.3 kb subjects (n=20) an increase (+10%; P=0.017) in OGTT glucose area. Moreover, the OGTT insulin area tended to increase less in 6.3/6.3 kb individuals (P=0.7; not shown).
The present data indicate that in response to overfeeding the ADRB2 Gln27Gln men registered a slightly higher gain in body weight but a considerably higher increase in total subcutaneous fat, as assessed from the sum of eight skinfolds, compared to the Glu27Glu/Gln27Glu individuals. Moreover, the insulin area under the curve during the OGTT and plasma leptin levels increased more in subjects with the Gln27Gln genotypes. The within-pair resemblance in weight gain was slightly higher in the Glu27Glu and Gln27Glu twin pairs (mean intra-pair difference of 1.5 kg) than in the Gln27Gln pairs (2.4 kg). The body composition and metabolic changes in response to overfeeding among the ADRB2 BanI 3.7/3.4 kb subjects were similar to those of the Gln27Gln subjects.
It has been previously reported that the effects of the ADRB2 Gln27Glu polymorphism are different between men and women4 such that there is a positive association between Glu27 and obesity in females,3,4 but a negative association between Glu27 and obesity in males.4,16 Our findings are in accordance with the latter. Thus, in the present study, male carriers of the Glu allele experienced less weight and total subcutaneous fat gain as well as a lesser increase in OGTT insulin area compared to Glu allele non-carriers. Additionally, in our series, the BanI 3.7/3.4 kb variant of the ADRB2 gene was associated with the response to overfeeding in a manner similar to the Gln27Gln variant. Some earlier studies have suggested that, although not causing a change in the sequence of the receptor molecule, the mutation detected with BanI could be related to the ADRB2 function.17 The mutation is located at nucleic acid residue 52314 and is in linkage disequilibrium with codon 1615 and 27 polymorphisms.15,18 Therefore, the effects of the ADRB2 BanI polymorphism identified here could be mediated by its linkage disequilibrium with the codon 27 polymorphism, as the codon 16 variant was not associated with the response to overfeeding in these young men. Interestingly, the codon 16, but not codon 27, polymorphism has been reported to be functionally significant in females, in which it influences the sensitivity of the B2-adrenoceptor to agonists.3 However, this has not been studied in males. We speculate that the Gln27Glu variant could have an effect on ADRB2 function in men by lowering the B2-receptor-mediated lipolytic rates. A lower lipolytic rate could lead to a higher propensity to store subcutaneous fat and, secondarily, to higher plasma leptin levels.
The present study also shows that the ADRB2 Gln27Gln and BanI 3.7/3.4 kb genotypes were both associated with a higher rise in insulin levels during the OGTT following exposure to overfeeding. The latter provides support for the concept that the variants may predispose, perhaps as a consequence of body fat gain, to insulin resistance under conditions of caloric affluence. Again, in contrast with the Glu27Glu males of this study, the ADRB2 Glu27Glu genotype was previously reported to have an increasing effect on plasma insulin levels in women.3
In the current study, there were a limited number of subjects with the rare alleles for the ADRA2A or ADRB3 markers. However, the ADRA2A DraI 6.3/6.3 kb variant was associated with lower increases in glucose (significant) and insulin (tendency) areas during the OGTT. This is in accordance with an earlier finding from our laboratory of an association of this genotype with lower plasma insulin concentrations and possibly higher insulin sensitivity in obese subjects.15
In conclusion, the Gln27Gln subjects of the ADRB2 Gln27Glu polymorphism experienced more accumulation of subcutaneous fat as well as a greater increase in insulin resistance compared to the Glu27Glu/Gln27Glu subjects in response to long-term overfeeding. Similar effects were observed for the 3.7/3.4 kb BanI cases. Since both mutations in the ADRB2 exhibit the same associations with the adiposity, leptin and insulin responses to the overfeeding protocol, one can speculate that the Gln27Glu mutation is more likely to cause the differential responses than the BanI variant. Genetic variation at the ADRB2 locus could thus be one of the factors responsible for the inter-individual differences observed in response to chronic positive energy balance. It must be kept in mind that the sample size for this study was small. Nonetheless, it provides useful information on the genes and pathways that should be further explored.
We are indebted to Jacques Bouillon, Suzie Hamel, Brigitte Zément, Maryse Lebrun, Martine Marcotte, Monique Chagnon, Josée Lapointe, Henri Bessette, Gilles Bouchard and Serge Carbonneau for their contributions to this study. Gratitude is expressed to Dr A Nadeau and the staff of the Diabetes Research Unit for the glucose and insulin assays. We thank Dr L Arthur Campfield for the leptin assay. Special thanks go to Guy Fournier and Dr Germain Thériault for their role in the management of the study and to Claude Leblanc for his statistical support. Supported in part by a grant (DK 34624) from the National Institutes of Health and the Finnish Heart Foundation. C Bouchard is supported in part by the George A Bray Chair in Nutrition.