| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Department of Internal Medicine I, University of Regensburg, 93042 Regensburg, Germany
Address all correspondence and requests for reprints to: K.-D. Palitzsch, M.D., Klinik und Poliklinik für Innere Medizin I, Franz-Josef-Strauss-Allee 11, 93042 Regensburg, Germany.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
From over 6450 participants in the Diabetomobile Study, a nationwide epidemiologic study on the prevalence of metabolic disorders (carried out from 1993 to 1996 in Germany), 1259 participants were randomly chosen. The 3-BAR genotype status was determined by 3-BAR gene-specific genomic PCR and consecutive restriction fragment length polymorphism analysis.
The frequencies of the different genotypes in the examined cohort were as follows: Trp64/Trp64, 88.3%; Trp64/Arg64, 10.8%; and Arg64/Arg64, 0.8%. No significant differences between the different genotypes were found when comparing age, body mass index, weight, total and high-density lipoprotein (HDL) cholesterol, fasting insulin, HbA1c, and blood pressure; neither did the type 2 diabetes mellitus participants in the different genotype groups differ significantly in terms of age of diabetes onset or HbA1c.
This is the largest population-based study on the Trp64Arg polymorphism reported yet. The Arg64 allele of the 3-BAR gene was found commonly in Germany. In our cohort, no significant associations between the Arg64 allele and metabolic disorders (e.g. obesity, type 2 diabetes mellitus, dyslipidemia, or hypertension) were detected.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In 1995, Walston and co-workers (2) detected a point mutation of a putative candidate, the ß 3-adrenergic receptor (3-BAR) gene. The 3-BAR is a G-protein-coupled transmembrane receptor located in brown and white adipose tissue and playing a role in the regulation of thermogenesis and lipolysis in rodents (3); stimulation of the 3-BAR is discussed to have an antidiabetic effect (4). The detected mutation results in a tryptophan/arginine exchange at position 64 (Trp64Arg polymorphism) of the amino acid chain. In the initial and in following studies, an earlier onset of type 2 diabetes mellitus, a higher capacity to gain weight (5, 6), a high body mass index (BMI) (7), and elevated fasting insulin levels (8) were found in homozygous carriers of the Arg64 allele in different ethnic groups. Kurabayshi et al. (9) also found a significant positive association between the Arg64 allele and reproductive history.
These findings prompted us to examine the prevalence of the Arg64 gene allele in a large population-based cohort in Germany and to look for possible associations with metabolic disorders such as diabetes mellitus, obesity, hypertension, and dyslipidemia.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Subjects in this study had participated in the Diabetomobile Study, an epidemiologic field survey on metabolic disorders in a representative adult German population, carried out from 1993 to 1996. A total of 6450 persons, between 18 and 70 yr of age, were investigated by a physician from our department with a mobile survey unit in representative German cities and rural communities. Participants were either included in a so-called street setting (by randomly choosing streets and house numbers and asking the inhabitants to participate voluntarily in the study) or in a marketplace setting (where all interested bystanders were included). All participants had to be fasting for at least 1 h.
All study subjects gave their informed consent for study participation and further analysis of their blood samples, including all biochemical and genetic studies performed. All participants remained anonymous throughout the survey.
All persons were interviewed, and they responded to a questionnaire on medical history, medication, life-style, and life-quality. Body height and weight were measured in the survey unit with approved medical care instruments, and BMI was computed as weight (kilograms) divided by squared height (meters squared). Resting blood pressure (BP) was measured, after subjects had been in a sitting position for at least 30 min, with a mercury sphygmomanometer. Participants with elevated BP (>140/90 mm Hg) were reexamined before leaving the mobile survey unit.
Biochemical measurements
Blood was drawn from all subjects, and the time of the last meal was recorded. Serum total and HDL cholesterol, glucose, and HbA1c were determined in the mobile survey unit. Total and HDL cholesterol were measured by dry chemistry (DT 60, Kodak, Germany). Intraassay coefficients of variation (CVs) were 4.5% and 4.4%, respectively; interassay CVs were 4.9% and 4.8%, respectively. Hypercholesterolemia was defined as total cholesterol more than 200 mg/dL (following the recommendations of the German Nutrition Society); low HDL cholesterol, as less than 35 mg/dL for men and less than 40 mg/dL for women. Glucose levels were determined with an Accutrend-GlucoseR Glucometer (Boehringer Mannheim, Mannheim, Germany). The intraassay CV was 2.9% at normal range glucose concentrations. HbA1c was measured by an immunoassay (DCA 2000, Bayer AG, Leverkusen, Germany). Intraassay and interassay CVs were 3.1% and 4.4%, respectively. The sensitivity range was from 2.5–14.0%, the correlation coefficient to ion exchange chromatography determined with 45 independent assays 97%.
Insulin and C-peptide levels were measured by an immunoassay (Enzymun Insulin, Boehringer Mannheim, Immulite C-Peptide, DPC, Los Angeles, CA) out of frozen plasma samples in our clinical laboratory. Intraassay CVs were 4.9% and 6.2%, respectively; interassay CVs were 7.0% and 5.9%, respectively.
Diabetes mellitus was assumed: 1) when the HbA1c was elevated (>6%, according to the manufacturer’s instructions); or 2) the proband stated to have a known history of diabetes in the interview and the questionnaire. To differentiate insulin-dependent diabetes mellitus (IDDM) from type 2 diabetes mellitus, we used age, diabetes onset, BMI, and C-peptide levels. IDDM was defined as diabetes onset before 35 yr of age, BMI less than 25 kg/m2, and C-peptide levels less than 1 ng/mL. All other subjects with diabetes were assigned to the type 2 diabetes mellitus group.
All instruments used for anthropometrical and biochemical measurements were calibrated and technically looked after weekly by a technical assistant.
Genotyping
Finally, 1259 participants in the Diabetomobile Study were randomly chosen for genotyping. Genomic DNA was isolated from whole-blood samples by standard procedures (QiaAmp Blood Kit, Qiagen, Hilden, Germany). Genotyping was then performed, as described in (6), by amplification of genomic DNA with a 3-BAR gene specific primer pair and consecutive analysis of restriction fragment length polymorphism after BstOI digestion.
Statistics
The street setting cohort represents the German population between 18 and 70, as compared with sex, age, BMI, and the other measured parameters [verified by comparison with statistical data on the total German population from the German federal statistic office (Statistisches Bundesamt, Wiesbaden, Germany), data not shown]. Data obtained from this cohort were therefore considered representative for the total German population.
The aim of our study was to examine whether statistically significant
correlations between the 3-BAR genotypes and metabolic disorders exist.
The parameters (age, BMI, total and HDL cholesterol, and fasting
insulin) were not normally distributed in our cohort as shown by
Kolmogorov-Smirnov tests. Therefore, subjects were compared according
to their 3-BAR genotype for differences in anthropometric and
biochemical data by two-tailed Mann-Whitney or Kruskal-Wallis tests for
comparison of two or more independent samples and Pearson
2 tests for associations between classified variables.
Data are expressed as means and SD for simplicity. The
significance level (
) was set to 0.05 in all statistical tests.
To examine associations as exactly as possible, tests were first performed with all three genotypes together in a one-way ANOVA setting. When a significance level less than 0.1 was achieved, we examined correlations of the possibly associated parameter to the 3-BAR genotypes in every possible combination, i.e. Trp64/Trp64 vs. Trp64/Arg64, Trp64/Trp64 vs. Arg64/Arg64, and Trp64/Arg64 vs. Arg64/Arg64.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
describes the
characteristics of the cohorts studied. In both study groups, there
were more female than male subjects, with no significant gender
differences among them. Subjects in the street setting were younger
(mean age, 44.2 yr vs. 50.8 yr; P < 0.001),
were less obese (mean BMI, 25.0 vs. 26.5 kg/m2;
P < 0.001), and showed a lower incidence of diabetes
mellitus (7.5 vs. 14.1%, P < 0.001) than
those in the marketplace setting.
|
2 tests, P
= 0.31 and P = 0.23, respectively). Because the street
setting has been proved to be representative for the total population,
we estimate the genotype frequencies in Germany derived from the total
pooled data to be about 88.3% (95% CI, 86.4–90.2) for the
Trp64/Trp64-genotype, about 10.8% (95% CI, 5.7–16.1) for the
Trp64/Arg64-genotype, and about 0.8% (95% CI, 0.3–2.4) for the
Arg64/Arg64-genotype.
Anthropometric and biochemical data of the pooled cohorts are shown for
male and female subjects in Tables 2 and 3, according to sex and the 3-BAR
genotype. The prevalence of the heterozygous genotype was
insignificantly higher in women than in men (11.1 vs.
10.6%). Interestingly, the prevalence of the homozygous Arg64 carriers
was opposite: higher in men than in women (1.4 vs. 0.3%),
but this finding was also not significant.
|
|
The age means were higher in subjects homozygous for the Arg64 allele in both sexes [52.9 yr (men) and 55.5 yr (women) vs. 48.9 yr and 48.0 yr, respectively, in Trp64 homozygotes], but this was not statistically significant. Detailed analysis of the prevalence of the 3-BAR genotypes in different age classes showed that Arg64/Arg64 carriers peak in the second youngest (but also in the oldest) age class, arguing against a genotype-specific age distribution (data not shown).
BMI, body weight
In men, BMI and body weight were similar in all groups. In women, heterozygous subjects had a slightly lower mean BMI (24.0 kg/m2), and Arg64 homozygotes had a higher mean BMI (37.1 kg/m2) than Trp64 homozygotes. These differences were not significant (P = 0.11 for all genotypes, 0.23 for comparison of Trp64 homozygotes with Trp64/Arg64 heterozygotes, and 0.09 for comparison of Trp64 and Arg64 homozygotes).
Figure 1 shows the prevalence of the
Trp64/Arg64 and the Arg64/Arg64 genotype, according to BMI for men and
women. No significant differences were observed between the different
BMI classes in either sex.
|
The mean systolic BP was slightly (but not significantly) elevated in the Arg64/Arg64 group in both sexes, as compared with the other genotypes. Diastolic BP means were lower in Arg64 homozygotes, but this also was not statistically significant.
The analysis of the genotype prevalence in different BP classes
(diastolic 
90 and > 90 mmHg; systolic 140,
141–160, and >160 mmHg) did not show any significant differences
(data not shown).
Cholesterol
Total cholesterol levels did not differ significantly in women. In men, analysis of all three genotypes showed an almost significant difference between cholesterol levels of Trp64 homozygotes and Trp64/Arg64 heterozygotes (P = 0.06). In the exact analysis, after adjustment for BMI and age, we found a trend to higher cholesterol levels in Arg64 heterozygous men with a BMI between 25 and 30 kg/m2, when compared with matched Trp64 homozygotes (means 231.3 vs. 207.0 mg/dL, P = 0.06), but not with matched Arg64 homozygotes. HDL cholesterol levels were similar in both sexes.
When comparing the prevalence of the different genotypes in subjects with normal and with elevated total cholesterol or low HDL cholesterol, respectively, no significant differences were found (data not shown).
Trp64Arg genotypes and diabetes mellitus
Of the total 1259 examined persons, 121 met the criteria for diabetes mellitus. Of these, 97 participants (56 male, 41 female) belonged to the type 2 diabetes mellitus, and 24 (13 male, 11 female) belonged to the IDDM group. Subjects with type 2 diabetes mellitus were significantly older than those with IDDM (mean age 61.5 vs. 45.7 yr, P < 0.001). No significant differences were detected for cholesterol levels, BP, and HbA1c (data not shown).
The characteristics of the type 2 diabetes mellitus subjects, with
respect to the 3-BAR genotype, are given in Table 4 (91.8% were homozygous for Trp64,
2.1% for Arg64; and 6.8% were heterozygous). The 3-BAR genotype
distribution did not differ significantly from the total cohort.
Because only two Arg64 homozygotes were detected, statistical analysis
was performed first for comparison of all 3-BAR genotype groups and
then for comparison of Trp64 homozygotes with pooled heterozygous and
homozygous Arg64 carriers.
|
). |
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the present investigation, the allelic and genotype frequencies of the 3-BAR genotypes in our population-based cohort were similar to those reported earlier in Caucasian populations (3, 6, 7, 10, 11). In Japanese and Pima populations, the Arg64 allele is found much more often, showing clear ethnic differences in 3-BAR genotype distribution. No consistent correlations of the 3-BAR genotypes to age groups could be found, arguing against a certain genotype being a negative selection factor associated with higher mortality or lower life expectancy. This, together with the quite high prevalence rates of the Arg64 allele in all examined populations, confirms the view that both alleles constitute an allelic polymorphism of the 3-BAR receptor gene, rather than the Arg64 allele being a mutation of the so-called correct Trp64 allele.
No significant correlations between the Trp64/Arg64 or the Arg64/Arg64
genotype and the prevalence of type 2 diabetes mellitus could be
detected. This is consistent with many previous reports examining the
frequency of the Arg64 allele in type 2 diabetes mellitus subjects
(5, 6, 7). In our study, no association of 3-BAR genotype and type 2
diabetes mellitus onset age was found, questioning earlier reports (5, 6). Fasting insulin levels in type 2 diabetes mellitus subjects were
not elevated in heterozygous type 2 diabetes mellitus carriers of the
Arg64 allele (Table 4). Because a high fasting insulin level is
believed to result from insulin resistance, these data argue against a
major role for the Arg64 allele in the development of insulin
resistance in type 2 diabetes mellitus in heterozygous carriers.
Interestingly, the two Arg64 homozygotes had elevated insulin levels
after 3 and 4 h of fasting, respectively, but this also was not
significant in comparison with the other 3-BAR genotypes. For the
interpretation of this finding, it is important to notice that
HbA1c levels were also elevated in the diabetic Arg64
homozygotes, when compared with diabetic Trp64 homozygotes and
Trp64/Arg64 heterozygotes. From our data, we cannot answer the question
of whether the worse metabolic control indicated hereby is caused by
the homozygous Arg64 allele, or perhaps to other imaginable reasons
such as different genetic factors, compliance problems, or maybe wrong
therapeutic strategies. In a previously published report by Urhammer
et al. (9), higher fasting serum C-peptide levels and lower
insulin sensitivity were demonstrated in three young healthy Danes with
the Arg64/Arg64 genotype, indicating that the Arg64 allele may
contribute to worsening insulin resistance and thereby also metabolic
control in diabetic homozygous carriers. From our work, we cannot
contradict these results, because dynamic quantifications of insulin
sensitivity cannot be compared with our epidemiologic cross-sectional
data appropriately. Nevertheless, when trying to evaluate the
contribution of the Arg64 allele to the development of diabetes in the
whole population, in our opinion, it is more useful to examine the
association of the 3-BAR genotypes to diabetes mellitus in a
representative cohort and not to insulin sensitivity in a healthy
subgroup. We did not find significantly more Arg64 homozygotes in the
type 2 diabetes mellitus than in the nondiabetic group. In our opinion,
this strongly argues against an important role in diabetes pathogenesis
for the Arg64 allele. Because of the small homozygous subject number,
this question cannot be answered finally, with reasonable statistical
power, from our data.
Some limitations of the study design must be taken into account when interpreting our data. In our setting (an epidemiologic field study on anonymous subjects), some difficulties arise in reliably defining the presence of diabetes mellitus and the differentiation between subjects with IDDM and type 2 diabetes mellitus. The indirect parameters that we used for diabetes diagnosis (HbA1c, personal statement) may lead to bias. It can be assumed that only a few participants with known diabetes (usually diagnosed in Germany by criteria of the World Health Organization or the German Diabetes Association) failed to state this in the interview with the study physician or in the questionnaire and that very few nondiabetics stated wrongly that they had diabetes. In a recent study of Pima Indians, comparing different diabetes diagnosis strategies (12), maximal equivalence with the World Health Organization criterion of a 2-h plasma glucose concentration of more than 200 mg/dL was achieved by setting the HBA1c cutoff-point to 6.1% (equalling the cutoff-point used in our study). A recent metaanalysis (13) found maximal sensitivity and specificity at an HbA1c cutoff-point of 7%. Only 1 case of diabetes was newly diagnosed in a Trp64 homozygote by our criteria; and in 9 cases (including 8 Trp64 homozygotes and 1 Trp64Arg heterozygote), the diabetes type had to be defined in accordance with our type 2 diabetes mellitus/IDDM criteria, because no sufficient data on medical history were obtainable. Taken together, we cannot rule out that a portion of participants may have been assigned erroneously to either the type 2 diabetes mellitus, IDDM, or nondiabetic group. But this portion is very small; and because these systematic errors apply to all participants, irrespectively of their 3-BAR genotype, they are not likely to cause significant mistakes in the comparison of the different 3-BAR genotypes and the conclusions on the pathogenetic relevance of the Trp64Arg-mutation derived from this.
The recently described associations of the Arg64 allele to obesity and hypercholesterolemia could not be confirmed after adjustment for age and sex. Only in Arg64 heterozygous men with a BMI between 25 and 30 kg/m2 could a trend to higher cholesterol levels be detected. The two female Arg64 homozygotes had a higher BMI, but this was not significant in statistical analysis. First, these differences from previous results could be attributed to the different genetic backgrounds in the examined populations. The two studies showing a significant association between the Arg64 allele and BMI were carried out in Japanese and Australian subjects (8, 10). Both studies performed the statistical analysis with lower subject numbers of Arg64 allele carriers than in our survey and with rather heterogenously composed study groups, which could have led to chance statistical effects. One can also speculate that a healthy lifestyle in our relatively lean and healthy total population-based cohort may override modest effects of the Arg64 variant, contributing to the conflicting results. On the other hand, in our study, neither the obese (BMI > 25 kg/m) nor the hypertensive, hyperlipidemic, or diabetic subgroups showed a higher prevalence of the Arg64 allele, arguing against this speculation and against a major role for the Arg64 allele in the development of these disorders in the total population.
Still, although our survey includes data on one of the largest Caucasian groups of Arg64 homozygotes reported yet, we cannot rule out the possibility that the Arg allele in its homozygous state carries some increased risk of obesity because of the low absolute number of Arg64 homozygous women in our cohort. This problem is difficult to solve, given the low prevalence of the Arg64 homozygotes in our population. One can hypothesize that a major genetic factor contributing to the pathogenesis of metabolic disorders probably would be likely to show disadvantageous effects in heterozygotes too. We are not likely to have missed such effects because of our high case numbers.
Some other clinical data have recently been published, questioning the role of the 3-BAR Trp64Arg polymorphism in the pathogenesis of human metabolic disorders (11, 14, 15, 16). A recent metaanalysis by Odawara and co-workers (17) was unable to detect any correlations between the Trp64Arg polymorphism and BMI or frequency of diabetes mellitus. The functional relevance of the Arg64 mutation has been discussed controversially. The point mutation leading to the Trp64Arg polymorphism is located in the first intracellular loop (5), which does not seem to participate in receptor binding or signaling pathways (18). This is consistent to the recent findings by Candelore et al. (19), who could not find any difference in the pharmacological and functional properties of the Arg64 version of the 3-BAR, compared with the Trp64 variant, arguing strongly against a pathophysiological effect of the Arg64 allele. In contrast to this, Pietri-Rouxel et al. (20) observed significant reductions of maximal cAMP accumulation in response to various ß 3-adrenergic agonists in human and hamster cells expressing the Arg64 3-BAR variant, whereas binding inhibition and adenylyl cyclase activation constants remained unchanged. Despite the distinct pharmacological functions of 3-BAR in rodent adipose tissue, mice lacking the 3-BAR showed only a minor increase in adipose tissue (21), indicating that a lack in 3-BAR can be compensated in vivo.
Recently, Elbein and co-workers could not detect any linkage between the 3-BAR gene locus and obesity or type 2 diabetes mellitus (22). The effect of 3-BAR stimulation in humans is still not exactly known, and its exact tissue distribution is controversially discussed (23, 24, 25).
Considering these findings and our data, the present investigation argues against a major role for the Arg64 allele in the development of obesity, type 2 diabetes mellitus, and the metabolic syndrome. Routine genotype analysis of obese patients, therefore, cannot be recommended.
Received October 21, 1997.
Revised January 30, 1998.
Revised April 16, 1998.
Accepted April 24, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. Masuo, T. Katsuya, Y. Fu, H. Rakugi, T. Ogihara, and M. L. Tuck {beta}2- and {beta}3-Adrenergic Receptor Polymorphisms Are Related to the Onset of Weight Gain and Blood Pressure Elevation Over 5 Years Circulation, June 28, 2005; 111(25): 3429 - 3434. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Froguel and P. Boutin Genetics of Pathways Regulating Body Weight in the Development of Obesity in Humans Experimental Biology and Medicine, December 1, 2001; 226(11): 991 - 996. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. P. Erickson and P. E. Graves Genetic Variation in {beta}-Adrenergic Receptors and Their Relationship to Susceptibility for Asthma and Therapeutic Response Drug Metab. Dispos., April 1, 2001; 29(4): 557 - 561. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
