| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Circulation. 2004;110:2644-2650.)
© 2004 American Heart Association, Inc.
Molecular Cardiology |
From the Study Coordinating Centre, Hypertension and Cardiovascular Rehabilitation Unit, Department of Molecular and Cardiovascular Research, University of Leuven, Leuven, Belgium (T.K., J.A.S., L.T., R.F.); Institute of Internal Medicine, Novosibirsk, Russian Federation (T.K., A.R., S.M., Y.N.); Department of Endocrinology and Nephrology, Charité, University of Berlin, Campus Benjamin Franklin, Berlin, Germany (C.K., E.B.); First Cardiac Department, Jagiellonian University, Cracow, Poland (A.O., K.S., K.K.-J.); Department of Clinical and Experimental Medicine, University of Padova, Padova, Italy (V.T., E.C.); Cattedra e Scuola di Nefrologia, Universita Vita e Salute San Raffaele, Milano, Italy (G.B.); Department of Molecular Genetics for Cardiovascular Disease, Institute for Arteriosclerosis Research, University of Münster, Münster, Germany (S.-M.B-H.); and Department of Internal Medicine D (Nephrology and Hypertension), University Clinic Münster, Münster, Germany (E.B.).
Correspondence to J.A. Staessen, Study Coordinating Centre, Laboratory of Hypertension, Hypertension and Cardiovascular Rehabilitation Unit, Department of Molecular and Cardiovascular Research, Campus Gasthuisberg, University of Leuven, Herestraat 49, B-3000 Leuven, Belgium. E-mail jan.staessen{at}med.kuleuven.ac.be
Received December 19, 2003; de novo received April 12, 2004; revision received June 9, 2004; accepted June 10, 2004.
| Abstract |
|---|
|
|
|---|
Methods and Results We randomly recruited 221 nuclear families (384 parents, 431 offspring) in Cracow (Poland), Novosibirsk (Russia), and Mirano (Italy). Echocardiographic LVM was indexed to body surface area, adjusted for covariates, and subjected to multivariate analyses using generalized estimating equations and quantitative transmission disequilibrium tests in a population-based and family-based approach, respectively. For AGTR1 and AGTR2, there was no heterogeneity in the phenotypegenotype relations across populations. LVM index was unrelated to the AGTR1 A1166C polymorphism. In men, in the population- and family-based analyses, the allelic effects of the AGTR2 polymorphism on LVM index differed (P=0.01) according to sodium excretion. In women, this geneenvironment interaction did not reach statistical significance. In untreated men, LVM index (4.2 g/m2 per 100 mmol) and left ventricular internal diameter (0.73 mm/100 mmol) increased (P<0.02) with higher sodium excretion in the presence of the G allele with an opposite tendency in A allele carriers. The ACE D/I polymorphism, together with the ACE genotype-by-sodium interaction term, significantly and independently improved the models relating LVM index to the AGTR2 polymorphism and the AGTR2 genotype-by-sodium interaction.
Conclusions The present findings support the hypothesis that in men the AGTR2 G1675A and the ACE D/I polymorphisms independently influence LVM and that salt intake modulates these genetic effects.
Key Words: angiotensin genes hypertrophy sodium receptors
| Introduction |
|---|
|
|
|---|
Taken together, the above observations raise the possibility that genetic variability in the AT1 and AT2 receptor genes (AGTR1 and AGTR2) might have an impact on left ventricular structure. In the European Project On Genes in Hypertension (EPOGH), we therefore investigated in 3 populations whether LVM was associated with the AGTR1 A1166C and AGTR2 G1675A polymorphisms. Because complex multigenic traits such as LVM can be studied only within their ecogenetic context, our analysis accounted for salt intake, estimated from the 24-hour urinary excretion of sodium and other covariables, including the genegene interactions with the ACE D/I and AGT 532C/T polymorphism.
| Methods |
|---|
|
|
|---|
Three EPOGH centers opted to take part in cardiac phenotyping with echocardiography. They randomly recruited white nuclear families consisting of at least 1 parent and 2 siblings. Age ranged from 18 to 60 years. The overall response rate was 61.3%. Of 976 participants recruited in Cracow (Poland, n=325), Novosibirsk (the Russian Federation, n=301), and Mirano (Italy, n=345), we discarded 161 from analysis. Ninety-three subjects declined the invitation for the ultrasonographic examination. The echocardiogram was of insufficient quality in 25 participants. We excluded 7 subjects with left ventricular dysfunction caused by myocardial infarction (n=4) or a valvular disorder (n=3). DNA of 22 subjects did not amplify, and the genotype of 9 subjects could not be determined with certainty. In addition, we detected 5 cases of inconsistency in mendelian segregation. Thus, the number of subjects statistically analyzed totaled 815.
The blood pressure phenotype was the average of 5 consecutive readings at 1 home visit. Body surface area (BSA) was calculated as body weight (kg)0.425xbody height (cm)0.725x0.007184. Using a standardized questionnaire, observers collected information on each subjects medical history, smoking and drinking habits, and use of medications. The participants collected a 24-hour urine sample in a wide-neck plastic container for the measurement of sodium, potassium, and creatinine. If urinary volume or creatinine excretion was outside published limits,6 the urinary results were discarded.
Echocardiographic Measurements
In each center, one experienced observer performed all echocardiograms as described elsewhere.1 Left ventricular internal diameter (LVID) and interventricular septal (IVST) and posterior wall (PWT) thicknesses were measured at end diastole according to the recommendations of the American Society of Echocardiography.7 For statistical analysis, the online measurements of 3 cardiac cycles were averaged. Studies were also recorded on videotape for quality control within each center and at the Coordinating Office in Leuven, Belgium. We calculated LVM by an anatomically validated formula from the end-diastolic measurements.8 Mean wall thickness (MWT) and left ventricular mass index (LVMI) were defined as (IVST+PWT)/2 and LVM/BSA, respectively. The intraobserver intersession reproducibility coefficient for LVM computed according to Bland and Altmans method9 was 2.5% for Cracow, 2.0% for Novosibirsk, and 2.6% for Mirano.
Determination of Genotypes
Genomic DNA was extracted from peripheral blood. We used previously described techniques to genotype the AGTR1 A1166C,10 ACE D/I,11 and AGT 532C/T10 polymorphisms. DNA fragments that contained the G1675A polymorphism of AGTR2 were amplified by polymerase chain reaction (PCR) with forward 5'-ATTACGTCCCAGCGTCTGAG-3' and reverse 5'-GGCACTAA-GCAAGCTGATTTAT-3' primers. The PCR products were digested by the addition of 5 U Hpy188III restriction enzyme. In the presence of the 1675G allele, the PCR product (255 bp) was cut into 2 fragments 194 and 61 bp long and visualized on ethidium bromidestained 1.5% agarose gels.
Statistical Analysis
We used the SAS software package, version 8.1 (SAS Institute), for database management and most statistical analyses. Comparison of means and proportions relied on the Tukey multiple-means test and a
2 statistic with Bonferroni adjustment for multiple comparisons, respectively.
In population-based analyses, we tested the association of continuous traits with the genotypes by use of generalized estimating equations (GEEs). GEEs allow adjustment for covariates and for the nonindependence of observations within families.12 In the GEE approach, we also tested for heterogeneity across populations and sexes using the appropriate interaction terms with the genotypes. We assessed the gene-gene and genotype-by-sodium interactions through comparison of models by log likelihood ratio tests.
In the family-based analysis, we performed transmission disequilibrium tests for quantitative traits. First, we evaluated the within- and between-family components of phenotypic variance using the orthogonal model as implemented by Abecasis et al13 in the QTDT software (version 2.3; http://www.sph.umich.edu/csg/abecasis/QTDT). The QTDT algorithm does not support the analysis of X-linked genes such as AGTR2. Second, using the approach proposed by Allison, we regressed the quantitative phenotypes of the offspring on their genotypes while controlling for parental genotypes.14 To allow for residual correlation among offspring, we implemented Allisons method using GEEs. Finally, in the PROC LOGISTIC procedure of the SAS package, we modeled the probability of the transmission of the allele of interest from each heterozygous parent as a function of the quantitative phenotype.15
| Results |
|---|
|
|
|---|
5 g/d). Urinary sodium excretion was on average 57 mmol/d higher in Cracow than in Mirano and intermediate in Novosibirsk. As shown in Table 2, LVM was significantly higher in Slavic participants than in Italians. Except for the AGTR1 A1166C genotype in Novosibirsk (P=0.01) and the ACE D/I genotype in Mirano (P=0.02), the within-center frequencies of genotypes (Table 3) complied with Hardy-Weinberg equilibrium (0.12
P
0.97). The AGTR2 A, ACE D, and AGT 532C allele frequencies were lower in Cracow and Novosibirsk than in Mirano (Table 3).
|
|
|
Population-Based Association Study
We adjusted left ventricular phenotypes for center, sex, age, systolic blood pressure, body weight, height, waist-to-hip ratio, use of antihypertensive drugs, and lifestyle factors, including smoking and alcohol consumption in excess of 5 g/d. Models with LVMI as a dependent variable were not adjusted for body weight and height.
Because for AGTR1 and AGTR2 there was no heterogeneity in the phenotypegenotype relations across centers (0.08
P
0.64), we combined all subjects. For the AGTR1 A1166C polymorphism, we did not observe any significant interaction with gender in relation to the left ventricular phenotypes (P>0.56). For the X-linked AGTR2 G1675A polymorphism, we analyzed men and women separately. There was no interaction between the 2 genotypes in men (P=0.11) and women (P=0.33).
In analyses that did not account for a genotype-by-urinary sodium interaction (Table 4), the phenotypegenotype relations did not reach statistical significance (0.10<P<0.94). In all men (n=371; Table 4), there was a significant interaction between the AGTR2 genotype and sodium excretion analyzed as a continuous variable in relation to LVMI (P=0.01) and LVID (P=0.002). Specific adjustments for specific subgroups of antihypertensive drugs did not alter our findings, as reported in Table 4. Further sensitivity analyses, from which we excluded 58 men on antihypertensive treatment, confirmed that in AGTR2 G allele carriers (n=155), LVMI (P=0.02) and LVID (P=0.003) increased with higher sodium excretion by 4.2 g/m2 per 100 mmol and 0.73 mm/100 mmol, respectively. In contrast, in 158 untreated men carrying the AGTR2 A allele, LVMI (P=0.18) and LVID (P=0.05) tended to decrease with higher sodium excretion. Moreover, in untreated men whose sodium excretion was <240 mmol/d (median), LVMI was 5.8 g/m2 higher (P=0.01) in AGTR2 A allele carriers than in their G allele counterparts. Figure 1 illustrates the association between LVMI and the AGTR2 polymorphism according to the median sodium excretion in untreated men (240 mmol/d) and women (180 mmol/d) and in both sexes combined with exclusion of heterozygous women.
|
|
Family-Based Association Study
Our study population (n=815) included 384 parents and 431 offspring. The number of offspring per family amounted to 1 in 28 families, 2 in 170 families, and
3 in 23 families.
In all subjects and in men and women analyzed separately, the QTDT analysis (number of informative offspring [n]=220;
2=0.12; P=0.73), logistic analysis (n=187;
2=1.22; P=0.27), and Allisons approach (n=113; df=2;
2=4.43; P=0.11) did not reveal any association between LVMI and transmission of the AGTR1 C allele.
Allisons approach demonstrated a significant interaction between transmission of the AGTR2 A allele and urinary sodium excretion in relation to LVMI in untreated men (n=91;
2=4.93; P=0.026) but not women (n=72;
2=0.94; P=0.63). Logistic regression (n=91;
2=4.57; P=0.03) confirmed the genotype-by-sodium interaction in male offspring (Figure 2). Overall, the
2 statistics for the combined effects of the AGTR2 polymorphism, sodium excretion, and their interaction in relation to LVMI amounted to 6.02 (P<0.05) and 8.40 (P<0.02) in Allisons approach and logistic regression, respectively.
|
Interactions With Other Renin System Genes
We previously demonstrated that LVMI increased with higher sodium excretion but that ethnicity and genetic variation in the ACE1 and AGT (data submitted for publication) genes modulated this effect. Using the population-based approach, we further explored the interactions between genetic variation in AGTR2, ACE, and AGT while accounting for the influence of sodium (Table 5). In men (P
0.04) and in Slavic women (P=0.02), the ACE D/I polymorphism, together with the ACE genotype-by-sodium interaction term, significantly and independently improved the models relating LVMI to the AGTR2 polymorphism, the AGTR2 genotype-by-sodium interaction, and other covariates (Table 5).
|
| Discussion |
|---|
|
|
|---|
AGTR1 maps to the long arm of chromosome 3. The A1166C variant is located in the 3' untranslated region of the gene. The putative association of the C allele with various cardiovascular phenotypes is usually attributed to a hitherto unconfirmed linkage disequilibrium with a functional polymorphism.16 In keeping with the present observations, most,17,18 albeit not all,19 clinical researchers were unable to demonstrate an association between LVM and the AGTR1 A1166C polymorphism. This might be due to insufficient sample size, the low prevalence of the CC genotype, or nonfunctionality of this genetic variant.
The third exon of the X-linked AGTR2 gene harbors the entire open reading frame of the AT2 receptor.20 Regulatory elements in the promoter area and in the first intron control the transcription of AGTR2.21 The G1675A polymorphism, located in intron 1, is probably functional. The G allele may be associated with enhanced AGTR2 transcription and increased expression of the AT2 receptor. In humans, the AT1 receptor is present in most tissues and mediates virtually all known biological actions of angiotensin II.5 Expression of the AT2 receptor declines rapidly after birth in most tissues except for the brain and adrenal gland. In addition, the AT2 receptor protein can be detected in the adult heart, vasculature, and kidneys.22 To the best of our knowledge, only 2 studies addressed the possible association of LVM with the AGTR2 G1675A polymorphism in humans.23,24 In accordance with our findings, the G allele frequency in these studies varied from 43% to 51%. In 120 men with normal or mildly elevated blood pressure and a mean age of 26 years, Schmieder et al23 found that hypertensive but not normotensive G allele carriers had a lower LVM than A allele carriers because of a reduced wall thickness. Herrmann et al24 observed a higher prevalence of the G allele in 55- to 74-year-old Scottish men (n=336) without ECG left ventricular hypertrophy enrolled in the Glasgow Heart Scan Old (GLAOLD) study compared with those with hypertrophy but could not confirm these findings in the Glasgow Heart Scan (GLAECO) survey.
None of the aforementioned studies on the AGTR2 polymorphism in humans accounted for sodium intake. We consistently noticed in our family- and population-based analyses a significant interaction in men between the AGTR2 G1675A polymorphism and sodium excretion. The AGTR2 genotype-by-sodium interaction was weaker in women than in men, possibly because heterozygous women attenuated this effect or because of the influence of estrogens on the expression of the AT receptors.25 The hypothesis that the AT2-mediated effects of angiotensin II on LVM are modulated by sodium intake is plausible, although the exact mechanism remains to be elucidated. Indeed, in animal experiments, dietary sodium depletion enhanced the expression of the AT2 receptor.26 Moreover, in sodium-depleted rodents, stimulation of the AT2 receptor produced natriuresis,27 whereas the opposite might occur in sodium-replete animals.28 LVM is calculated from LVID and MWT. LVID to some extent reflects the circulating fluid volume, whereas MWT might be more indicative of processes confined to the myocardium itself. We found that the interaction between the G1675A polymorphism and sodium excretion in relation to LVMI was mediated via LVID. This suggests that renal AT2 receptors, through their influence on sodium balance and the circulation plasma volume, might at least partially explain our findings. On the other hand, we also confirmed an independent effect of the ACE D/I genotype on LVMI. As we showed previously,1 this effect was mediated via MWT and was dependent on sodium intake.
LVM is a quantitative trait prone to measurement error. However, we believe that the difference in echocardiographic LVM between Slavic and Italian subjects (94.9 versus 80.4 g/m2) was due to ethnic and environment factors rather than to interobserver variability or measurement error. Indeed, Sega et al29 found in a north Italian population (n=1648; age, 48.2 years) a mean±SD LVMI of 83.4±18.2 g/m2, which is similar to our findings in the Mirano cohort (80.4±19.3 g/m2). Moreover, we recalculated LVMI according to Devereuxs formula in the men enrolled in the Novosibirsk subsample of the MONICA cohort. Again, the mean value of LVMI was similar to the present findings in the men from Novosibirsk (105.3±22.0 versus 105.1±26.4 g/m2, respectively). Our findings were consistent in population- and family-based analyses. We did not find evidence for population stratification in any of the 3 centers in the family-based analysis (data not shown).
In conclusion, the present findings support the hypothesis that in men the AGTR2 G1675A and the ACE D/I polymorphisms independently influence LVM and that salt intake modulates this genetic effect. However, because of the exploratory nature of our analyses, which is inherent to all observational epidemiological studies, our findings need confirmation in further human and experimental studies.
| Appendix |
|---|
|
|
|---|
asowski, L. Gijsbers, T. Nawrot, L. Thijs, Y. Toremans, J.A. Staessen, S. Van Hulle, J.G. Wang, R. Wolfs. Bulgaria (Sofia): C. Nachev, A. Postadjian, E. Prokopova, E. Shipkovenska, K. Vitljanova. Czech Republic (Pilzen): J. Filipovsky, V. Svobodova, M. Ticha. Czech Republic (Prague): O. Beran, L. Golán, T. Grus, J. Pele
ka, Z. Marecková. Italy (Padova): E. Casiglia, A. Pizzioli, V. Tikhonoff. Poland (Cracow): K. Kawecka-Jaszcz, T. Grodzicki, K. Stolarz, B. Wizner, A. Olszanecka, A. Adamkiewicz-Piejko, W. Lubaszewski, J.
yczkowska. Romania (Bucharest): S. Babeanu, D. Jianu, C. Sandu, D. State, M. Udrea. Russian Federation (Novosibirsk): Y. Nikitin, S. Malyutina, T. Kuznetsova, E. Pello, M. Ryabikov, M. Voevoda.
Coordination and Committees
Project coordinator: J.A. Staessen. Scientific coordinator: K. Kawecka-Jaszcz. Steering Committee: S. Babeanu, E. Casiglia, J. Filipovsky, K. Kawecka-Jaszcz, C. Nachev, Y. Nikitin, J. Pele
ka, J.A. Staessen. Data Management Committee: T. Kuznetsova, J.A. Staessen, K. Stolarz, L. Thijs, V. Tikhonoff, J.G. Wang. Publication Committee: E. Casiglia, K. Kawecka-Jaszcz, Y. Nikitin.
Advisory Committee on Molecular Biology: G. Bianchi (Milan), E. Brand (Berlin, Münster), S.M. Brand-Herrmann (Münster), H.A. Struijker-Boudier (Maastricht). EPOGH-EurNetGen liaison: A. Dominiczak (Glasgow), J.A. Staessen (Leuven).
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S Sookoian, T F Gianotti, and C J Pirola Role of the C-344T aldosterone synthase gene variant in left ventricular mass and left ventricular structure-related phenotypes Heart, July 1, 2008; 94(7): 903 - 910. [Abstract] [Full Text] [PDF] |
||||
![]() |
S.-M. Brand-Herrmann Where Do We Go for Atherothrombotic Disease Genetics? Stroke, April 1, 2008; 39(4): 1070 - 1075. [Full Text] [PDF] |
||||
![]() |
E. D. Frohlich The Salt Conundrum: A Hypothesis Hypertension, July 1, 2007; 50(1): 161 - 166. [Full Text] [PDF] |
||||
![]() |
M. Nishikino, T. Matsunaga, K. Yasuda, T. Adachi, T. Moritani, G. Tsujimoto, K. Tsuda, and N. Aoki Genetic Variation in the Renin-Angiotensin System and Autonomic Nervous System Function in Young Healthy Japanese Subjects J. Clin. Endocrinol. Metab., November 1, 2006; 91(11): 4676 - 4681. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Warnecke, P. Mugrauer, D. Surder, J. Erdmann, C. Schubert, and V. Regitz-Zagrosek Intronic ANG II type 2 receptor gene polymorphism 1675 G/A modulates receptor protein expression but not mRNA splicing Am J Physiol Regulatory Integrative Comp Physiol, December 1, 2005; 289(6): R1729 - R1735. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2004 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |