| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Comments |
Steno Diabetes Center (L.T.D., K.B.-J., T.H., O.P.), DK-2820 Gentofte, Denmark; Danish Epidemiology Science Center at the Institute of Preventive Medicine, Copenhagen University Hospital (T.I.A.S.), DK-1399 Copenhagen, Denmark; Center of Preventive Medicine, Glostrup University Hospital (T.D.), DK-2600 Glostrup, Denmark; and Roskilde County Hospital (T.A.), DK-4000 Roskilde, Denmark
Address all correspondence and requests for reprints to: Dr. Louise T. Dalgaard, Steno Diabetes Center, Niels Steensens Vej 2, DK-2820 Gentofte, Denmark.
Abstract
Variability of the uncoupling protein 3 (UCP3) promoter
has been associated with increased body mass index (BMI) and altered
lipid profiles. Here we tested the hypothesis that variation of the
UCP3 promoter is associated with either juvenile or maturity-onset
obesity or body weight change over a 26-yr follow-up among Danish
subjects. Mutation screening of approximately 1 kb 5' upstream of the
UCP3 gene revealed one previously described -55
C
T variant. The frequency of the polymorphism was evaluated by
restriction fragment length polymorphism analysis in four groups of
subjects: 1) a group of 744 obese Danish men who at the draft board
examinations had a body mass index (BMI) of at least 31
kg/m2, 2) a randomly selected control group consisting of
857 draftees, 3) 258 middle-aged subjects, and 4) 409 60-yr-old
subjects. The frequency of the T allele was 26.0% (95% confidence
interval, 23.8–28.2%) among the obese draftees and 26.9%
(24.8–29.0%) in the control group (P = 0.6). The
variant was not associated with BMI at a young age or with weight gain
after a 26-yr follow-up. The frequency of the T allele was 29.5%
(25.6–33.4%) in the middle-aged group and 25.8% (22.8–28.8%) among
the 60-yr-old subjects. The polymorphism was not associated with
increased BMI or percent body fat in these 2 groups. It is concluded
that this variant does not play a major role in the development of
common obesity among Danish subjects.
OBESITY MAY RESULT from interactions between different factors controlling the environment, eating behavior, and energy expenditure. The inherited predisposition for obesity is, with rare monogenic exceptions, considered to be due to polymorphisms in several genes acting synergistically to dispose for obesity (1). Segregation analyses show that the most likely model contains two major recessive loci affecting variation in body weight in the normal range (2, 3, 4). Thus, the genetics behind common obesity is likely to be dominated by several polymorphisms each with moderate effect. Therefore, large study populations are needed to identify such polymorphisms.
Uncoupling protein 3 (UCP3) belongs to a family of mitochondrial transporters that could uncouple the oxidative phosphorylation by increasing the proton leak of the inner mitochondrial membrane (5), which would lead to heat production rather than energy storage. Decreased function or expression of UCP3 could reduce energy expenditure and increase the propensity to store energy as fat (6). Several studies have pointed to a role of UCP3 in the regulation of whole body energy homeostasis. In mice quantitative trait loci for obesity coincide with the genomic location of the UCP2-UCP3 gene cluster on chromosome 7 (7, 8), and in French Canadian subjects linkage between UCP2-UCP3 on chromosome 11q13 and resting metabolic rate has been reported (9). In UCP2 a 3'-untranslated insertion was associated with increased body mass index (BMI) in Pima Indians above 45 yr of age (10), but this was not observed in Danish Caucasian subjects (11). Transgenic mice overexpressing UCP3 in muscle tissue are lean and resistant to diet-induced obesity (12). The mice also exhibit lower fasting plasma glucose, total serum cholesterol, and serum insulin levels and an increased glucose clearance rate. Furthermore, the expression of UCP3 messenger ribonucleic acid (mRNA) in humans and rodents is positively correlated with plasma levels of free fatty acids (FFA) (13, 14). UCP3 might therefore be involved in the regulation of lipids as metabolic substrates (15).
The C/C genotype of a polymorphism in the UCP3 promotor
(-55 CT) is associated with increased expression of UCP3
mRNA in skeletal muscle of Pima Indians (16). Other
investigators have reported that the -55 T/T genotype is associated
with increased BMI and interacts negatively with physical activity
(17). Recently, it was shown that the -55 T/T genotype
was also associated with an atherogenic lipid profile in French
Caucasians, and that the T/T genotype conferred decreased risk of type
2 diabetes (18).
The major objectives of the present study were: 1) to examine the 5' region of the UCP3 gene for variability, 2) to evaluate whether variants could be associated with juvenile-onset obesity or with increased BMI after a 26-yr follow-up, and 3) to identify the possible effects of promoter variants on the fasting levels of serum lipids, plasma glucose, and serum insulin in two population-based samples.
Subjects and Methods
Subjects
The primary mutation analysis was performed in 60 obese subjects (20 men and 40 women), recruited at Steno Diabetes Center. The mean BMI was 39.5 kg/m2 (range, 31.6–52.9). Type 2 diabetes was present in 44 subjects, and mean age at diagnosis was 51 yr (range, 21–73). The treatments of type 2 diabetic patients were as follows: biguanide (2), sulfonylurea (13), biguanide and sulfonylurea (9), insulin (3), and diet (17).
The association study was performed in 2 study groups selected from a population of young Danish males, who at the mean age of 20 yr (range, 18–31) were examined at the draft board where height and weight were measured. The 791 subjects with a BMI of at least 31 kg/m2 were selected to comprise the obese group. As a random control group, 915 subjects were selected at random as every 200th draftee. During the Copenhagen City Heart Study Program the subjects were reexamined between 1992–1994 after a mean time interval between examinations of 26 yr (range, 15–50) (19). A lean control group was selected from the random control group as the 353 individuals with a BMI less than 25 kg/m2 at reexamination. A morbidly obese group was defined as the 156 individuals with a BMI greater than 40 kg/m2 at reexamination. Genomic DNA was obtained from blood samples drawn at the second examination.
Further investigations were performed in a population-based sample of 258 middle-aged (mean age, 53; range, 30–88 yr) men and women from the Copenhagen area and in a population-based sample of 409 60-yr-old men and women, which constitutes a random subset of an age-specific cohort, also from the Copenhagen area. All study participants were Danish Caucasians by self-identification. All subjects underwent a 75-g oral glucose tolerance test, and height, weight, and fat percentage (measured with bioimpedance) were recorded. Diabetes was diagnosed according to WHO 1998 guidelines (20).
Methods
Blood samples for measurement of serum levels of insulin, cholesterol, high density lipoprotein cholesterol, triglycerides, and plasma glucose and FFA were drawn after a 12-h overnight fast. Serum tri- glycerides, total serum cholesterol, serum high density lipoprotein cholesterol, and plasma FFA were analyzed with enzymatic colorimetric methods (GPO-PAP and CHOD-PAP, Roche Molecular Biochemicals, Mannheim, Germany; NEFA C, Wako, Neuss, Germany). The plasma glucose concentration was analyzed by the glucose oxidase method (Granutest, Merck & Co., Inc., Darmstadt, Germany), and serum-specific insulin was measured by enzyme-linked immunosorbent assay (insulin kit K6219, DAKO Corp., Ely, UK). Informed consent was obtained from all subjects before participation. The study was approved by the ethical committee of Copenhagen and was in accordance with the principles of the Declaration of Helsinki II.
The promoter sequence of UCP3 was divided into six
overlapping segments for single strand conformation polymorphism and
heteroduplex analysis. Primer sequences are listed in Table 1
. PCR amplification was carried out as
described previously (11) with MgCl2
concentration, dimethylsulfoxide content, and annealing temperature as
described in Table 1. Single strand conformation polymorphism was
performed as described previously (21). The -55 C/T
polymorphism was genotyped using primers RG-F and RG-R, followed by
digestion with SmaI. Restriction fragments were resolved on
a 3% agarose gel and visualized with ethidium bromide staining.
|
Data analysis was restricted to 744 of 791 obese draftees and
857 of 915 of the control draftees due to lack of PCR product. For the
groups of middle-aged and 60-yr-old subjects, 220 and 301
glucose-tolerant subjects, respectively, entered the analysis of
fasting serum lipid variables. 
2 analysis was
performed to test for differences in allele frequencies between groups.
A general linear model, with age and BMI (when appropriate) as
covariates, and genotype and gender as fixed factors, was used when
variables (or transformed variables) were normally distributed.
Otherwise the Mann-Whitney or Kruskal-Wallis rank sum test was applied.
Data are given as the mean (SD) or median (interquartile
range). P < 0.05 was considered significant. Analyses
were performed using the Statistical Package for Social Science
(SPSS, Inc., Chicago, IL) version 10.
Results
Mutation analysis of 1098 bp upstream sequence of the
UCP3 gene encompassing the two transcription initiation
sites revealed one prevalent CT polymorphism situated 55 nucleotides
5' of the second transcription initiation site (22) and
527 nucleotides 3' of the other reported transcription initiation site
(23). Interestingly, the polymorphism is situated only 6
bp upstream from the TATA box of the second transcription initiation
site and 4 bp from of a site involved in UCP3 responsiveness to
retinoic acid (24).
The allelic frequency of the T allele of the -55 C/T polymorphism in
the UCP3 gene was 26.0% (95% confidence interval,
23.8–28.2%) among the obese subjects and 26.9% (24.8–29.0%) among
the random control subjects (P = 0.6). In the lean
group, selected as the 353 subjects with a BMI less than 25
kg/m2 at reexamination at the age of 46 yr, the
frequency of the T allele was 25.2% (22.0–28.4%; P =
0.7 compared with the obese group). The allele frequency in the
morbidly obese group, defined as the 156 individuals with a BMI greater
than 40 kg/m2 at reexamination, the frequency of
the T allele was 27.6% (22.6–32.5%; P = 0.3 compared
with the lean group). The genotypes were in Hardy-Weinberg equilibrium.
The BMI at mean ages of 20 and 46 yr was similar in subjects carrying
the C/C genotype, the C/T geno- type, and the T/T genotype within
the obese and random control groups, respectively (Table 2). There was a tendency among the obese
group for the C/C carriers to have a higher BMI, which did not reach
statistical significance. There was no difference between genotypes
when considering the increase in BMI per yr between examinations (Table 2).
|
) or among
the group of 60-yr-old subjects (Table 4). Gender was included as a covariate in
the analyses; however, if the two genders were analyzed separately, the
results were similar (data not shown). There was no relationship
between the variant and altered fasting serum lipid profiles or serum
insulin among the glucose-tolerant middle-aged and 60-yr-old subjects
(Tables 3 and 4). The level of plasma FFA was slightly higher among T/T
homozygotic subjects (Table 3), but this finding did not reach
statistical significance. Fasting plasma glucose was higher for -55
C/C carriers in the group of middle-aged glucose-tolerant subjects
(Table 3), but in the group of 60-yr-old subjects there was no
association between genotype and fasting plasma glucose (Table 4).
|
|
The identified polymorphism (-55 CT) has previously been
reported (16, 17). We investigated the impact of this
variant on juvenile-onset obesity and on weight gain over a 26-yr
follow-up in two samples of Danish men. Furthermore, the possible
effect of the polymorphism on BMI and percent body fat was investigated
in two other groups of middle-aged to elderly Danish men and women of
normal body weight. However, the variant was not associated with
increased BMI, body fat or altered fasting serum lipid levels in any of
the four investigated groups consisting of Danish Caucasian
subjects.
In Pima Indians, the -55 C/C subjects had lower expression of UCP3 mRNA in skeletal muscle (16). However, no difference in genotype frequencies was observed between obese and lean subjects in Pima Indians. In French subjects, however, the T/T carriers belonged to the most obese subgroup (17) and had significantly increased total serum cholesterol together with decreased risk of developing type 2 diabetes (18). The power of the present study of men recruited at the draft board to detect a difference in BMI of only 1 kg/m2 is above 95%, and the difference in BMI observed in French subjects was 6 kg/m2 (17). The effect of the promoter variant could be gender specific, but we did not detect any influence of gender on BMI, body fat content, or fasting serum lipid levels. It is therefore unlikely that the -55 C/T variant has an effect on BMI, body fat content, or changed fasting serum lipid profiles among Danish subjects. The slightly higher fasting plasma FFA levels in T/T subjects are, although not significant, interesting keeping in mind the relationship with plasma FFA levels and the expression level of UCP3 mRNA in skeletal muscle (13, 14, 15). We observed that in the group of middle-aged glucose-tolerant subjects the C/C carriers had higher fasting plasma glucose. However, this was not confirmed in the otherwise comparable group of 60-yr-old glucose-tolerant subjects, and we must therefore regard this as an incidental finding.
The absence of a functional UCP3 gene does not cause obesity in mice (25, 26), although this finding does not exclude that the gene has importance for energy homeostasis, as compensatory mechanisms might be activated. Interestingly, up-regulation of the UCP3 gene in mouse muscle proved to make a lean phenotype (11). In Mexican-American subjects, linkage studies show that variation in the UCP2-UCP3 gene cluster does not have a major impact on obesity (27). In this study we have with greater than 95% power excluded the present variant to have a major influence on BMI (>1 kg/m2) in the normal Danish population. The -55 C/T polymorphism might not be a functional variant, and the discrepancy between published data could thus be explained by variation in the extent of linkage disequilibrium between different ethnic groups. Further study of the genetic variation in the UCP2/UCP3 region together with in vitro promoter studies is necessary to answer this question.
Acknowledgments
We thank Annemette Forman, Lene Aabo, and Bente Mottlau for technical assistance; and Grete Lademann for secretarial support.
Footnotes
1 This work was supported by the University of Copenhagen, the Danish
Medical Research Council, EEC Grant BMH4-CT98-3084, the Danish Heart
Foundation, the Foundation of Director Jacob Madsen and wife Olga
Madsen, and the Foundation of King Christian the Xth. 
Received September 20, 2000.
Revised November 14, 2000.
Accepted November 20, 2000.
References
This article has been cited by other articles:
 |
|
 |
|
  A. F M van Abeelen, M. de Krom, J. Hendriks, D. E Grobbee, R. A H Adan, and Y. T van der Schouw Variations in the uncoupling protein-3 gene are associated with specific obesity phenotypes Eur. J. Endocrinol., May 1, 2008; 158(5): 669 - 676. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Rudofsky Jr, A. Schroedter, A. Schlotterer, O. E. Voron'ko, M. Schlimme, J. Tafel, B. H. Isermann, P. M. Humpert, M. Morcos, A. Bierhaus, et al. Functional Polymorphisms of UCP2 and UCP3 Are Associated With a Reduced Prevalence of Diabetic Neuropathy in Patients With Type 1 Diabetes Diabetes Care, January 1, 2006; 29(1): 89 - 94. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-J. Liu, P.-Y. Liu, J. Long, Y. Lu, L. Elze, R. R. Recker, and H.-W. Deng Linkage and association analyses of the UCP3 gene with obesity phenotypes in Caucasian families Physiol Genomics, July 14, 2005; 22(2): 197 - 203. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Schrauwen and M. Hesselink UCP2 and UCP3 in muscle controlling body metabolism J. Exp. Biol., August 1, 2002; 205(15): 2275 - 2285. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Argyropoulos and M.-E. Harper Molecular Biology of Thermoregulation: Invited Review: Uncoupling proteins and thermoregulation J Appl Physiol, May 1, 2002; 92(5): 2187 - 2198. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
