点击显示 收起
1 From the Human Genomics Laboratory, Pennington Biomedical Research Center, Baton Rouge, LA (RJFL, TR, and CB), and the Department of Social and Preventive Medicine, Division of Kinesiology, Laval University, Ste-Foy, Canada (SR, AT, and LP)
2 Supported by grants no. PG-11811, MT-13960, and GR-15187 from the Medical Research Council of Canada and the Canadian Institutes for Health Research; by a postdoctoral fellowship from the American Heart Association, Southeast affiliate (no. 0325355B) (to RJFL); and by the George A Bray Chair in Nutrition (to CB).
3 Reprints not available. Address correspondence to C Bouchard, Pennington Biomedical Research Center, Human Genomics Laboratory, 6400 Perkins Road, Baton Rouge, LA 70808. E-mail: BouchaC{at}pbrc.edu.
See corresponding editorial on page 1.
ABSTRACT
Background: Despite adiponectin's presumed role in fatty acid oxidation and energy homeostasis, little is known about the effect of gene variants on substrate oxidation, energy expenditure, and adiposity-related phenotypes.
Objective: We examined the effects of genetic variation in adiponectin (ADIPOQ) and adiponectin receptors 1 and 2 (ADIPOR1 and ADIPOR2) on resting metabolic rate, respiratory quotient (RQ), and adiposity-related phenotypes.
Design: We studied the associations of ADIPOQ, ADIPOR1, and ADIPOR2 polymorphisms with resting metabolic rate, RQ, and body mass index, percentage body fat, sum of 6 skinfold thicknesses, waist circumference, and total, subcutaneous, and visceral fat in 759 participants in the Québec Family Study.
Results: The ADIPOQ 45TG single-nucleotide polymorphism (SNP) was significantly (P = 0.0002 to 0.04) associated with overall adiposity and abdominal adiposity; the rare homozygotes (G/G) had a leaner phenotype than did the carriers of the common allele. One SNP each in the putative promoter of ADIPOR1 (ie, –3882TC) and ADIPOR2 (ie, IVS1 –1352GA) was associated with RQ (P = 0.03 and 0.04, respectively), and the association was even stronger in nonobese persons (P = 0.02 and 0.003). Carriers of the common alleles (ADIPOR1 T and ADIPOR2 G alleles) had a lower RQ than did the rare homozygotes. A significant genotype-by-genotype interaction (P = 0.0002 to 0.02) was found between SNPs in the promoters of ADIPOQ (–3971AG) and ADIPOR1 (–3882TC). Subjects carrying the minor ADIPOQ allele (G allele) who were rare homozygotes (C/C) for the ADIPOR1 SNP had a higher RQ (P = 0.003) and greater overall (P < 0.03) and abdominal (P < 0.05) adiposity than did persons with other genotype combinations.
Conclusions: Previous findings that the ADIPOQ 45TG variant contributes to overall fatness and abdominal obesity are confirmed. Moreover, variants in the promoter region of both ADIPOR genes contribute to substrate oxidation.
Key Words: Adiponectin • adiponectin receptor • resting metabolic rate • respiratory quotient • obesity • abdominal obesity • adiposity
INTRODUCTION
Adiponectin is a protein that is produced exclusively and abundantly in the adipose tissue and that is present in plasma in relatively high concentrations (1). Unlike the concentrations of most other adipokines, such as tumor necrosis factor , interleukin 6, leptin, and resistin, those of adiponectin are low in patients with obesity and type 2 diabetes (2, 3) and are inversely associated with measures of overall adiposity (2-5) and with insulin resistance, independent of fat mass (5). The administration of adiponectin in animal models of obesity and diabetes results in weight loss, prevention of diet-induced obesity without influencing food intake, reduced tissue triacylglycerol content in muscle and liver, increased fatty acid oxidation in skeletal muscle, and improved insulin sensitivity and glucose tolerance (4, 6, 7). Adiponectin is encoded by the adiponectin gene, ADIPOQ, located on chromosome 3q27 in humans, which coincides with a diabetes susceptibility locus (8). In humans, adiponectin and ADIPOQ gene polymorphisms have been studied extensively for their effects on obesity- and diabetes-related phenotypes (2-5). Despite the presumed role of adiponectin in fatty acid oxidation and energy homeostasis, only 2 studies in humans have evaluatedadiponectin concentrations in relation to substrate oxidation and energy expenditure (EE) (9, 10). In a Finnish study that included 158 offspring of persons with type 2 diabetes, adiponectin concentrations during hyperinsulinemia were related positively to EE and negatively to lipid oxidation rates (9). However, in 75 nondiabetic Pima Indians and 18 nondiabetic whites, no such associations were found with plasma adiponectin concentrations (10).
Much less is known about the recently discovered adiponectin receptors, encoded by the ADIPOR1 and ADIPOR2 genes, that are located at 1p36.13-q41 and 12p13.31, respectively (11). Both receptors are expressed in human skeletal muscle; ADIPOR1 expression is more abundant than ADIPOR2 expression (11, 12). In vitro experiments, in which cultured myocytes were treated with adiponectin, showed that suppression of the expression of the ADIPOR gene greatly reduced fatty acid oxidation and glucose uptake, whereas overexpression of ADIPOR1 increased fatty acid oxidation and glucose uptake beyond the effects of adiponectin (11). ADIPOR gene expression was significantly lower in those Pima Indian subjects with a strong family history of type 2 diabetes than in those with no family history (12). However, the association between ADIPOR expression and insulin sensitivity could not be confirmed by other investigators (13). Common gene variants in the ADIPOR genes have been studied for associations with diabetes-related phenotypes (14-17). However, to the best of our knowledge, no studies have reported on the effect of common ADIPOR1 and ADIPOR2 gene variants on EE or substrate oxidation rates, despite their known roles in lipid oxidation and glucose uptake.
In the current study, we examined whether sequence variations in the ADIPOQ and ADIPOR genes contributed to variation in adiposity, EE, and fasting respiratory quotient (RQ) in participants in the Québec Family Study (QFS).
SUBJECTS AND METHODS
Study population
The aims and design of the QFS cohort have been described previously (18). Briefly, the QFS aimed to study the genetics of obesity and its related phenotypes by applying the most recent advances in genetic epidemiology available: ie, segregation analyses, linkage analyses, and association analyses. The QFS was initiated in 1978 and has consisted of 3 phases; phase 1 was based on 375 families, of which a subsample of 105 families and an additional 74 families were followed 10 y later in phase 2. Most of these families were remeasured 5 y later in phase 3. In the current study, which is based on phase 2 of the QFS, 759 subjects from 183 French-Canadian families living in the Greater Québec City area were included (Table 1).
View this table:
TABLE 1. Descriptive characteristics of participants1
Written informed consent was obtained from all participants. The QFS project was approved by the Medical Ethics Committee of Laval University.
Phenotypes
Height and weight were measured to the nearest 0.1 cm and 0.1 kg, respectively, by using a stadiometer and a balance-beam scale. Body mass index (BMI; in kg/m2) was calculated. Body density, obtained from underwater weighing (19), was converted to percentage body fat (%BF) by using the equation of Siri (20). Waist circumference was measured to the nearest 0.1 cm at the level of noticeable waist narrowing by using an anthropometric fiberglass tape in accordance to the procedures recommended by Lohman et al (21). Six measures of skinfold thicknesses (ie, suprailiac, subscapular, abdominal, medial calf, biceps, and triceps) were obtained on the left side of the body by using a Harpenden skinfold caliper and following the procedures recommended by the International Biological Program (22). We calculated the sum of the 6 skinfold thicknesses (ST6) as an indicator of subcutaneous fatness and the ratio of trunk to extremity [(trunk:extremity) ie, (suprailiac + subscapular + abdominal):(medial calf + biceps + triceps)] as an indicator of preferential deposition of subcutaneous fat on the trunk rather than on extremities. Abdominal fat (total, visceral, and subcutaneous) was measured by using computed tomography scans (Somatom DRH scanner; Siemens, Erlanger, Germany) as described previously (23). Briefly, an abdominal scan was taken between the 4th and 5th lumbar vertebrae while participants were in a supine position with the arms stretched above the head. A radiograph of the skeleton was used as reference to determine the position of the scan to the nearest millimeter. Total and visceral adipose tissue areas were delineated with a graph pen. The attenuation interval used in the quantification of the areas of adipose tissue was between –190 and –30 Hounsfield units. The abdominal visceral fat (AVF) area was defined by drawing a line within the muscle wall surrounding the abdominal cavity. The abdominal subcutaneous fat (ASF) area was calculated by subtracting the AVF area from the abdominal total fat (ATF) area.
Resting metabolic rate (RMR) and RQ at rest were measured simultaneously with the use of a ventilated hood coupled to an open-circuit indirect calorimeter, as described previously (24). Measurements were made early in the morning, after an overnight fast, while participants were resting quietly in a semi-reclined position. Respiratory exchange data from the final 10 min of the 30-min data collection period were used to calculate RMR and RQ. Gas samples were assayed with a zirconia cell oxygen analyzer (Amatek CD-3A; Thermox Instruments Division, Pittsburgh, PA) and an infrared carbon dioxide analyzer (Amatek S-3A; Thermox Instruments Division). Analyzers were calibrated before each test with the use of gases of known percentages of oxygen and carbon dioxide. RMR is expressed in kcal/min EE, whereas RQ is simply the ratio of carbon dioxide produced to oxygen consumed.
Genotypes
Genomic DNA was prepared from permanent lymphoblastoid cells by using proteinase K and the phenol-chloroform technique. DNA was dialyzed 4 times against buffer [10 mmol Tris/L, 1 mmol EDTA/L; pH 8.0 (TE buffer)] for 6 h at 4 °C and precipitated.
We typed 3 SNPs in the ADIPOQ gene (Figure 1 A)—IVS2 276GT (rs1501299), exon 2 (EX2) 45TG (rs2241766) and IVS1 –3971AG (rs822396). The IVS2 276GT and EX2 45TG SNPs are the most extensively studied ADIPOQ SNPs and are located in the second of 2 linkage disequilibrium (LD) blocks, whereas the IVS1 –3971AG SNP is located in the first LD block (25). The IVS2 276GT polymorphism was detected with the BsaMI restriction enzyme (call rate: 98.0%), and the EX2 45TG polymorphism with the SmaI restriction enzyme (call rate: 99.8%). The genomic DNA (20 ng) was amplified in 6-µL mixtures containing dNTPs (30 µmol/L each), Taq DNA polymerase (0.3 U), buffer (15 mmol MgCl2/L), and polymerase chain reaction (PCR) primers (50 nmol/L each). The reaction mixture was denaturated for 5 min at 95 °C, followed by 34–38 PCR amplification cycles. The first 10 cycles consisted of denaturation at 95 °C for 20 s, primer annealing at 57 °C for 1 min, and primer extension at 72 °C for 15 s. The following 24–28 cycles consisted of denaturation at 94 °C for 20 s, primer annealing at 52 °C for 55 s, and primer extension at 72 °C for 15 s. The PCR mixture was then digested by the restriction enzyme for 15 min at 37 °C and then for 20 min at 80 °C. A mini-sequencing assay was performed according to Sun et al (26): dATP/ddNTP mix (1.56 µmol/L each), IRDye tag primers (3.125 nmol/L), Thermosequenase (0.3 U), and 0.6X buffer were added to the PCR mixtures. Subsequently, the mixture was denaturated for 2 min at 95 °C and then subjected to 30 amplification cycles. Each cycle consisted of denaturation at 96 °C for 10 s, primer annealing at 57 °C for 30 s, and primer extension at 72 °C for 15 s.
FIGURE 1.. Location and genotype frequencies of the single-nucleotide polymorphisms in the adiponectin (ADIPOQ) and adiponectin receptor 1 and 2 (ADIPOR1 and ADIPOR2, respectively) genes, according to parental data.
The ADIPOQ IVS1 –3971AG SNP and the SNPs in the ADIPOR1 (ie, rs1539355 and rs2275737) and ADIPOR2 (ie, rs10773982 and rs2058112) genes were typed by using the Acycloprime Fluorescent Polarization SNP detection kit (Perkin Elmer Life Sciences, Inc, Boston, MA), which includes PCR amplification, amplicon purification, and dye-terminator incorporation reaction (call rates for the SNPs: 97.7%, 92.2%, 97.0%, 95.0%, and 98.0%, respectively). The genomic DNA (25 ng) was amplified in 6-µL reaction mixtures containing 0.5 µL 10x PCR buffer (Applied Biosystems), 0.4 µL 25 mmol MgCl2/L (2), 0.1 µL 5 mmol dNTP/L, 0.06 µL PCR primers (10 µmol/L each), 0.02 µL Taq polymerase, and 4.86 µL H2O. The reaction mixture was held at 95 °C for 2 min for Taq enzyme activation and then underwent 35 amplification cycles. Each cycle consisted of denaturation at 95 °C for 10 s, primer annealing at 60 °C for 20 s, and primer extension at 68 °C for 30 s. The reaction mixtures were then incubated at 68 °C for 10 min for final primer extension. Unused primers and unincorporated dNTPs were removed by using clean-up reagent (0.2 µL), clean-up dilution buffer (1.8 µL; Perkin Elmer Life Sciences), and H2O (3 µL) and then incubated at 37 °C for 60 min and at 80 °C for 15 min for enzyme inactivation.
For the primer extension reaction, the purified PCR product was combined with 0.05 µL acyclopol, 2.0 µL 10x reaction buffer, 0.5 µL of the specific dye terminator combination (Perkin Elmer Life Sciences), 0.5 µL SNP primer (10 µmol/L), and 6.95 µL H2O. The reaction mixtures were then incubated at 95 °C for 2 min and the underwent 5–15 cycles, each at 95 °C for 15 s and at 55 °C for 30 s. Fluorescent polarization values were directly measured by using a plate reader (VICTOR2; Perkin Elmer Life Sciences). The 4 SNPs in the ADIPOR gene were selected on the basis of information that was available from the Centre d'Etude du Polymorphisme Humain population (specifically, Utah residents with Northern and Western European ancestry) of the International HapMap Project (27). We selected the SNPs in such way that the linkage disequilibrium between them is low; ie, each SNP independently represents a part of the genetic variation of the gene. Therefore, we used the SNPSpD (Single Nucleotide Polymorphism Spectral Decomposition; Internet: http://genepi.gimr.edu.au/general/daleN/SNPSpD) Web interface that, on the basis of spectral decomposition of matrixes of pairwise LD between SNPs, enables the selection of a subset of SNPs that optimize the information in a genomic region (28).
Statistical analysis
Associations between the ADIPOQ, ADIPOR1, and ADIPOR2 SNPs and the phenotypes were analyzed by using the MIXED model procedure in SAS software (version 8.2; SAS Institute Inc, Cary, NC). Nonindependence among family members was adjusted for by using a "sandwich estimator" that asymptotically yields the same parameter estimates as ordinary least-squares or regression methods, but the standard errors and resulting hypothesis tests are adjusted for the dependencies. The method is general and assumes the same degree of dependency among all members within a family. The general-linear-model procedure was used to analyze associations between the parents, because they are not genetically related. Possible gene-by-gene, sex-by-gene. and BMI (<30 or 30)-by-gene interactions were tested with the MIXED model procedure by including main effects and interaction terms in the same model. Haplotypes were estimated by using the best option that is available in the MERLIN software (version 1.0.1; 29). Associations between haplotypes and phenotypes were performed by using the MIXED model procedure. Only subjects with unambiguous haplotypes were included in the analyses.
RMR and RQ measurements were adjusted for sex, age, fat mass, and fat-free mass. ST6 and trunk:extremity were adjusted for sex, age, and height. Waist circumference, ASF, AVF and ATF were adjusted for sex and age. We log transformed ST6, trunk:extremity, ASF, AVF, and ATF variables before analyses to normalize their skewed distributions. The inverse log-transformed values are given in the tables.
A likelihood ratio test was performed to assess whether the observed genotype frequencies were in Hardy-Weinberg equilibrium. The pairwise LD among the SNPs was assessed by r2 and D (normalized measure of Lewontin) by using the ldmax program as implemented in the GOLD software (8 November 2001 version; 30).
RESULTS
The genotype frequencies for the ADIPOQ, ADIPOR1, and ADIPOR2 polymorphisms are shown in Figure 1. The naming of the SNPs is based on their intronic or exonic nucleotide change, and it reflects their relative position to the start codon. All genotype frequencies were in Hardy-Weinberg equilibrium. The pairwise LD among the ADIPOQ SNPs was generally low; ie, r2 = 0.02 and D = 0.71 between IVS1 –3971AG and EX2 45TG, r2 < 0.01 and D = 0.10 between IVS1 –3971AG and IVS2 276GT, and r2 = 0.09 and D = 0.86 between EX2 45TG and IVS2 276GT. For ADIPOR1 SNPs, the pairwise LD between IVS1 –3882TC and IVS1 –100GT was r2 = 0.34 and D = 0.97; for ADIPOR2 SNPs, the pairwise LD between IVS1 –35361AG and IVS1 –1352GA was r2 = 0.09 and D = 0.99. The phenotypic characteristics of the 759 participants are shown in Table 1.
The G/G homozygotes of the EX2 45TG ADIPOQ gene were significantly leaner and had less abdominal fat than did the carriers of the T allele. On average, the G/G homozygotes had %BF 6 percentage points lower, ST6 28 mm smaller, and ASF and ATF measures 75 cm2 and 103 cm2 lower, respectively, than did the T/T homozygotes (Table 2). None of the ADIPOQ variants had an association with RMR or RQ, and no associations were observed with the ADIPOQ IVS2 276GT variant. Moreover, the ADIPOQ IVS1 –3971AG variant was not directly associated with any of the studied phenotypes. However, significant genotype-by-genotype interactions (P = 0.0002 to 0.02) were observed between the ADIPOQ IVS1 –3971AG variant and the ADIPOR1 IVS1 –3882TC variant (Figure 2). Carriers of the ADIPOQ IVS1 –3971AG G allele who also were C/C homozygotes for the ADIPOR1 –3882TC variant had a significantly higher RQ (>0.03), higher %BF (>3.4%), and greater ST6 (>21 mm) and waist circumference (>4.8 cm) than did subjects with the other genotype combinations.
View this table:
TABLE 2. Associations between the ADIPOQ EX2 45TG polymorphism and the anthropometric measures resting metabolic rate (RMR) and respiratory quotient (RQ)1
FIGURE 2.. Genotype-by-genotype interaction: association between the IVS1 –3971A
In each of the 2 ADIPOR genes, a 5-untranslated region (UTR) variant had a significant association with RQ (Table 3). The rare homozygotes (C/C) of the ADIPOR1 IVS1 –3882T>C variant had a significantly higher RQ than did the carriers of the T allele, and this difference was more pronounced in the nonobese subjects. Moreover, for the ADIPOR2 IVS1 –1352G>A variant, the rare homozygotes (A/A) had a significantly higher RQ than did the G allele carriers. The association was significantly stronger in nonobese than in obese subjects (gene-by-obesity status interaction, P = 0.001). None of the ADIPOR gene variants had a direct association with any of the adiposity measures. The haplotypes did not provide any more information than did the individual markers.
View this table:
TABLE 3. Associations between the ADIPOR1 IVS1 –3882TC and ADIPOR2 IVS1 –1352GA polymorphisms and respiratory quotient (RQ) in nonobese and obese subjects1
DISCUSSION
SNPs in the ADIPOQ gene have frequently been associated with obesity-related phenotypes (31-39). However, not all studies support this hypothesis (40-44). Some studies have found ADIPOR gene DNA sequence variants to be related to type 2 diabetes and insulin resistance (17). However, to the best of our knowledge, no studies have reported associations between ADIPOR SNPs and fat oxidation rate, EE, and other adiposity-related phenotypes.
Our data show that the ADIPOQ 45TG variant is significantly associated with measures of overall adiposity and abdominal fatness and that the minor allele (G/G) homozygotes have a leaner phenotype than do the carriers of the common alleles (T/T and G/T). In addition, we found that one SNP each in the promoter region of ADIPOR1 (IVS1 –3882TC) and ADIPOR2 (IVS1 –1352GA), respectively, was associated with RQ, and this association was even stronger in nonobese persons. Carriers of the common alleles (T allele for ADIPOR1 and G allele for ADIPOR2) had a higher fat oxidation rate (ie, lower RQ) than did the rare homozygotes (C/C and A/A, respectively). A significant genotype-by-genotype interaction was observed between SNPs in the proximal promoters of ADIPOQ (–3971AG) and ADIPOR1, respectively (–3882TC). Subjects carrying the rare ADIPOQ allele (G allele) who also were rare homozygotes (C/C) for the ADIPOR1 variant had a lower fat oxidation rate (higher RQ) and higher measures of overall and abdominal adiposity than did subjects with the other genotype combinations.
The ADIPOQ EX2 45TG, one of the most extensively studied ADIPOQ gene polymorphisms, is a synonymous mutation located in EX2. Because this SNP does not result in an amino acid change, it is unlikely that it has functional implications for protein activity. However, this SNP can be in complete or nearly complete LD with a yet unidentified polymorphism that has functional relevance. We found that the rare homozygotes (G/G) had significantly less overall body fat (%BF and ST6) as well as less abdominal fat (ASF and ATF) than did carriers of the common allele. Our findings have been supported by others (32, 34, 35, 37, 38). In the Swedish Obese Subjects cohort, rare homozygotes had a significantly smaller waist circumference than did the carriers of the common allele (34). Moreover, in a study of 713 Hispanic men and women, G/G homozygotes had a leaner phenotype—ie, lower BMI and smaller waist circumference and ASF—than did carriers of the T allele (37). Two other studies, one in nondiabetic whites (32) and the other in African Americans (35), found the G allele to be related to reduced measures of overweight and abdominal adiposity, but only when the EX2 45TG SNP was part of a haplotype with IVS2 276GT. These findings suggest that the ADIPOQ EX2 45TG variant plays a protective role against the development of obesity. Several recent studies found the G allele to be associated with higher plasma adiponectin concentrations than was the T allele (45-47). This corroborates the findings in the current study, because higher adiponectin concentrations are correlated with lower body mass, BMI, and other obesity-related phenotypes (2-5, 10, 40). Other investigators, however, did not find an association between the EX2 45TG variant and obesity-related phenotypes, which may be due to the fact that they examined the dominant model (G/G + G/T compared with T/T) because of small sample sizes (32, 35, 36, 40-42, 48), rather than a codominant one (T/T compared with G/T compared with G/G), as was used in the current study. Results opposite of those in the current study were reported from a German and a French study. The German study found the G allele to be associated with higher BMI, waist-to-hip ratio, and %BF in those without a family history of diabetes than in those with such a family history (31). In the French study, the G allele was associated with increases in body weight and waist-to-hip ratio during a 3-y follow-up (36). Some investigators reported significant associations between the EX2 45TG – IVS2 276GT haplotype and measures of obesity (32, 39, 49), but we did not find evidence for such an association.
The ADIPOQ IVS2 276GT variant is not located in a coding region, and its functional implications should be investigated further. Despite the lack of direct evidence that this variant affects the expression concentrations of adiponectin, it was found to be associated with adiponectin concentrations in several populations (40, 44, 45, 47, 48, 50). We found no associations between this ADIPOQ gene variant and RMR, RQ, or any of the adiposity-related phenotypes, whereas significant associations with body fat have been reported in African Americans (35) and whites (48).
To the best of our knowledge, only one other study (9) has examined the association between the ADIPOQ EX2 45GT and IVS2 276GT gene variants and EE and RQ. This Finnish study found no associations in 158 offspring of type 2 diabetic patients (9), which is similar to the findings of the current study.
Several recent studies reported that variants in the proximal promoter region of the ADIPOQ gene are associated, either independently or as part of a haplotype, with adiponectin concentrations, type 2 diabetes, and BMI (33, 36, 39, 43, 44, 47). However, we found no association between the ADIPOQ –3971AG variant, located in the 5-UTR, and any of the adiposity-related phenotypes, which is in accordance with other studies (33, 44). However, we did find a significant genotype-by-genotype interaction between the ADIPOQ –3971AG and ADIPOR1 –3882TC variants in relation to RQ and adiposity. The carriers of the G allele for the ADIPOQ variant who also were C/C homozygotes for the ADIPOR1 SNP had a significantly higher RQ and significantly higher measures of overall and abdominal adiposity than did subjects with the other genotype combinations. The association with substrate oxidation rates was independent of body composition because we adjusted RQ for fat mass and fat-free mass. The lower fat oxidation rates and greater adiposity suggest that this rather rare genotype combination results in less efficient adiponectin-adiponectin receptor signaling than do the other genotypes. However, the molecular interactions between adiponectin and its receptors and the importance of the gene promoter regions in these interactions remain to be elucidated.
The adiponectin receptors mediate increases in AMP-activated protein kinase, peroxisome proliferator–activated receptor ligand activities, glucose uptake, and fatty acid oxidation by adiponectin (51). To the best of our knowledge, the current study is the first to report that variants in the 5-UTRs of both ADIPOR1 (IVS1 –3882TC) and ADIPOR2 (IVS1 –1352GA) are associated with RQ and that the rare homozygotes have a lower rate of fat oxidation (ie, higher RQ). However, none of the ADIPOR variants were associated with EE or any of the adiposity-related phenotypes. This finding is in agreement with other studies, which focused on the association between ADIPOR1 and ADIPOR2 SNPs in relation to type 2 diabetes and insulin resistance and reported no association with BMI (15-17). Because this is the first study to show an association between ADIPOR variants and RQ, further studies are warranted in other populations, and functional studies of these ADIPOR gene variants are necessary to clarify their potential roles in substrate oxidation rates and lipid metabolism.
In conclusion, these findings suggest that the ADIPOQ 45TG variation contributes to variation in overall adiposity and abdominal fatness, whereas variants in the proximal promoter regions of ADIPOR1 and ADIPOR2 contribute to variation in relative substrate oxidation rates as evidenced by RQ measurements. In addition, a significant genotype-by-genotype interaction was found between gene variants in the proximal promoters of ADIPOQ and ADIPOR1, and that affects RQ and total body fatness and abdominal adiposity.
ACKNOWLEDGMENTS
We thank Germain Theriault, Guy Fournier, Lucie Allard, Claude Leblanc, and Monique Chagnon, who were involved in the data collection and the management of the data.
RJFL performed the literature review; SR, TR, AT, LP, and CB were responsible for the study design and data collection; RJFL performed the statistical analyses and interpreted the data; RJFL was responsible for drafting the manuscript, and SR, TR, AT, LP and CB contributed to revisions of the manuscript. CB is a member of the Mars Nutrition Research Council. None of the authors had a personal or financial conflict of interest.
REFERENCES
Related articles in AJCN: