Author + information
- Paola Stella, MD*,* (, )
- Giada Bigatti, MD†,
- Laura Tizzoni, BBSC*,
- Cristina Barlassina, BBSC†,
- Chiara Lanzani, MD*,
- Giuseppe Bianchi, MD* and
- Daniele Cusi, MD†
- ↵*Reprint requests and correspondence:
Dr. Paola Stella, Division of Nephrology, Dialysis and Hypertension, University “Vita e Salute” San Raffaele, Via Olgettina 60, 20132 Milan, Italy.
Objectives The aim of our study was to evaluate the relationship between aldosterone synthase gene polymorphism and cardiac dimensions in essential hypertension.
Background Higher aldosterone synthase messenger ribonucleic acid levels in the human heart are accompanied by increased intracardiac aldosterone production, a phenomenon that is associated with cardiac fibrosis and hypertrophy. Recent evidence suggests that a polymorphism (−344C/T) in the promoter region of the aldosterone synthase gene is associated with increased constitutive aldosterone production.
Methods Relationships between M-mode echocardiographic cardiac dimensions and aldosterone synthase −344C/T polymorphism were studied in 210 never-treated, middle-aged patients (age 41.6 ± 1.4 years) affected by mild to moderate essential hypertension. Among these patients, 48 had the genotype −C344C, 97 had −C344T, and 65 had −T344T. Patients in the three groups were similar in terms of age, gender, body mass index, and blood pressure.
Results Left ventricular (LV) mass and thickness were positively correlated with the number of T alleles: LV mass (CC, CT, and TT, respectively: 168 ± 6.9, 179 ± 5.2, and 193 ± 6.9 g; p = 0.03), LV septal thickness (0.99 ± 0.02, 1.03 ± 0.02, and 11.08 ± 0.03 cm, p = 0.04), PWT (0.93 ± 0.03, 0.95 ± 0.01, and 1.03 ± 0.02 cm; p = 0.002), and relative wall thickness (38.3 ± 1.2%, 38.8 ± 0.8%, and 42.8 ± 1.1%; p = 0.004). This trend was confirmed by linear regression, suggesting a “major gene” behavior for the T allele. Multiple regression analysis showed that this effect was independent of anthropometric and clinical factors, including adrenal aldosterone.
Conclusions Our data suggest that −344C/T polymorphism affects LV mass and thickness in essential hypertension, independent of adrenal aldosterone. A role for intracardiac aldosterone synthesis is hypothesized.
Left ventricular hypertrophy (LVH) is an important predictor of cardiovascular morbidity and mortality in hypertensive patients, as well as in the general population (1,2). It is influenced by many variables, such as age, obesity, dietary habits, pressure fluid overloads (3), and pharmacologic treatment (4). Moreover, it has been previously demonstrated that some vasoactive substances, such as angiotensin II and aldosterone, acting as growth factors, may be associated with the development of LVH (5). In the past few years, particular attention has been focused on aldosterone, for the following reasons: 1) myocardial and vascular fibrosis were shown to be induced by this hormone (6–8); 2) the Randomized ALdactone Evaluation Study (RALES) has demonstrated that the combination of aldosterone antagonists and angiotensin-converting enzyme inhibitors can reduce the mortality and morbidity from cardiac heart failure; this effect does not appear to be due to the action of spironolactone on sodium and potassium renal handling, but implies a direct cardioprotective action of this drug (9); 3) intracardiac synthesis of aldosterone was found in isolated rat hearts (10); 4) increased aldosterone synthase (CYP11B2) expression was recently detected in biopsies from failing human hearts, along with a direct relationship between myocardial fibrosis and 11CYPB2 messenger ribonucleic acid expression (11); and 5) increased aldosterone production in the left ventricle (LV) of hypertensive patients, compared with those with normotension, has recently been demonstrated (12).
The −344C/T polymorphism falls in the promoter region of the aldosterone synthase gene and is associated with increased constitutive aldosterone production: in particular, the T allele is associated with increased plasma aldosterone levels (13). Several studies of the association between this polymorphism and left ventricular mass (LVM) have been published (14–17), with controversial results. This is not surprising, as LVM is not only modulated by environmental factors, but is also under polygenic control. If not adequately minimized, environmental influences and genetic background may also modify or mask genotype–phenotype relationships. As a consequence, only highly homogeneous populations of patients sharing similar environmental factors should be considered in analyses of this association.
To provide further insight into the control of LV structure by CYP11B2 −344C/T polymorphism in essential hypertensive patients, we investigated a highly homogeneous population of only Caucasian middle-aged patients with mild to moderate hypertension recently diagnosed and never treated. To our knowledge, this is the first study to follow these restrictive selection criteria. After this selection, a positive association between the number of −344T alleles and both LV thickness and mass was found.
To minimize environmental confounding factors, inclusion criteria were very restrictive. Patients were required: 1) to be of the same race (all Caucasian); 2) to have only essential hypertension without collateral diseases with cardiovascular implications; 3) to have not undergone pharmacologic treatment for hypertension both before and during the study; 4) to be not obese (body mass index of 30 kg/m2); and 5) to have normal sodium intake (24-h sodium excretion of 300 mEq).
Thus, 210 (172 males and 38 females) never-treated Caucasians (all Italians) with mild to moderate essential hypertension were recruited into the study. All patients were middle-aged and had a relatively short duration of hypertension.
Blood pressure measurements
Hypertension was diagnosed by the contemporary presence of: 1) diastolic blood pressure (BP) ≥85 mm Hg by daytime ambulatory BP monitoring (Spacelabs 90207, Redmond, Washington); and 2) diastolic BP ≥90 mm Hg during office visits. Office BP was measured with a mercury sphygmomanometer; for each patient, all measurements were made by the same doctor and during the same period of the day (8 to 10 am). To minimize the “white-coat” hypertensive effect, 10 office BP measurements were recorded after the patient had been clinically examined and after 10 min of rest in the supine position. The last four measurements were averaged, and this average was used for the diagnosis. The study protocol was approved by the Ethics Committee of the San Raffaele Hospital.
Serum and urine electrolytes and creatinine were measured by standard methods (ion-selective electrode and autoanalyzer), and plasma renin activity (PRA) and plasma aldosterone with a radioimmunoassay (DiaSorin, Vercelli, Italy; and MedicalSystem SpA, Genova, Italy). To have an estimate of the angiotensin II–independent rate of aldosterone production, the aldosterone/PRA ratio was computed for each patient.
Measurement of LVH
Patients underwent mono- and bi-dimensional echocardiography using a Hewlett-Packard imaging system (Sonos 2500 model, Palo Alto, California). All measurements were performed by the same investigator (Dr. Stella), who was blinded to the genotypes of the patients.
Left ventricular internal diameters, septal thickness (ST), and posterior wall thickness (PWT) were measured at end diastole and end systole, according to the guidelines of the American Society of Echocardiography (ASE) (18). Variation coefficients for septal and posterior wall end-diastolic diameters, septal and posterior end-systolic diameters, and LVM were 3%, 6%, 4%, 3%, and 5%, respectively. Technical error was calculated as the square root of the ratio between the sum of squares of observations and the number of observations/averages of observations, as reported by Liu et al. (19).
Left ventricular mass was calculated at end diastole by applying the Devereux correction to the ASE-cube LVM formula (20,21). Relative wall thickness (RWT) was calculated as 2 · posterior wall/end-diastolic diameter ratio, as reported by Ganau et al. (22).
Only patients with normal LV systolic function were included, whereas patients with both wall motion and cardiac valve abnormalities were excluded from the study.
Genomic DNA was isolated from 3 ml whole blood following a standard procedure (23), with minor modifications.
Allelic discrimination of the −344C/T CYP11B2 polymorphism was performed using a 5′ nuclease assay on an ABI Prism 7700 apparatus (Perkin Elmer, Norwalk, Connecticut) (24). Forward and reverse primers and −344C and −344T probes used in the TaqMan assay (Applied Biosystems, Foster City, California) were: 5′-CTAAATCTGTGGTATAAAAATAAAGTCTATTAAAAGA; 5′-TTTCTCCAGGGCTGAGAGGA; 5′VIC-AAGGCCCCCTCTCATCTCACGATA-TAMRA; and 5′FAM-CAAGGCTCCCTCTCATCTCACGATAAG-TAMRA, respectively. Per 25 ml, polymerase chain reaction fluid contained 50 ng DNA, 300 nmol primers, 70 nmol FAM probe, and 50 nmol VIC probe. Amplification conditions were 50°C for 2 min and 95°C for 10 min, followed by 40 cycles at 95°C for 15 min and 62°C for 1 min.
All data are given as the mean value ± SEM. Differences between groups were analyzed with one- or two-way analysis of variance or analysis of co-variance, as specified in the Results section. Deviation from Hardy-Weinberg equilibrium, gender ratios, and the presence of LVH were analyzed with the chi-square test. Continuous variables were analyzed with simple and multiple regression. Analysis was performed with SPSS version 10 for the MacOS statistical software on an iMAC personal computer.
The frequencies of the −344C and −344T alleles were 46% and 54%, respectively. Genotype frequencies for the −344C/T polymorphism are reported in Table 1. The frequencies of the observed genotypes were not different from those expected from the allele frequencies and Hardy-Weinberg equilibrium (p = 0.83). There were no significant differences among the genotypes in terms of gender distribution, body size, BP levels, age, and duration of hypertension. Among anthropometric and biochemical variables, serum urate significantly decreased with an increase of T alleles (p = 0.02). Plasma aldosterone and PRA were similar among the three groups, whereas the aldosterone/PRA ratio was positively associated (p = 0.03) with the number of T alleles (Table 1). Neither plasma aldosterone, nor PRA, nor the aldosterone/PRA ratio was correlated with cardiac dimensions.
Table 2summarizes the univariate analysis of variance for cardiac dimensions according to the three genotypes and shows that the CYP11B2 genotype was significantly associated with LVM, ST, PWT, and RWT. Patients with the TT genotype had the greatest values for these parameters, whereas patients with the CT genotype had values greater than did those with the CC genotype.
Moreover, coefficients of linear regression of CYP11B2 and ST, PWT, RWT, LVM, and aldosterone/PRA on the number of T alleles were all statistically significant, suggesting that the T allele had an “additive effect” on these parameters (Table 3).
To reduce the potential for bias due to confounding with environmental factors, two-step multiple regression analyses were performed. Septal thickness, PWT, RWT, and LVM were considered one by one as the dependent variable in two different regression models. In the first model, the independent variables were age, gender, height, weight, BP, heart rate, duration of hypertension, 24-h urinary sodium output, and aldosterone/PRA ratio. In the second model, the CYP11B2 genotype (defined as the number of T alleles) was added to the equation, retaining all the variables considered in first step. The regression coefficient associated with the CYP11B2 genotype was statistically significant for all the dependent variables. Specifically, the genotype effect explained 4.6%, 12.8%, 8.5%, and 4.9% of the total variability for ST, PWT, RWT, and LVM, respectively. However, the statistical significance of the whole model was reached for ST, PWT, and LVM, but not for RWT (Table 4).
Our study demonstrates a positive relationship between the aldosterone synthase polymorphism CYP11B2 and LV mass, wall thicknesses, and RWT. Simple regression of CYP11B2 polymorphism with cardiac variables shows an “additive effect” of the T allele (e.g., the greater the number of T alleles, the greater the value) for ST, PWT, RWT, and LVM, suggesting a “major gene” behavior for CYP11B2 polymorphism. Multiple regression analysis showed that this effect is independent of anthropometric and clinical factors, including adrenal aldosterone. In particular, CYP11B2 alone explained 4.6%, 12.8%, 8.5%, and 4.9% of total variability for ST, PWT, RWT, and LVM, respectively.
This is presently the first study to show a direct correlation between CYP11B2 polymorphism and LVM in hypertension. A previous report by Kupari et al. (14)reached apparently opposite results, as patients bearing the CC genotype showed a greater LVM. However, their study subjects were normotensive, whereas we studied hypertensive patients. We believe this difference cannot be disregarded, as different pathologic mechanisms may be responsible for the development of LVH in these two populations. In line with this hypothesis, LVM depended on the end-diastolic diameter in Kupari's patients (CC patients showed a greater end-diastolic diameter, although there were no differences in ST and PWT among the three genotypes), whereas in our patients, LVM depended on ST and PWT (TT patients showed a greater ST and PWT, although there were no differences in end-diastolic diameter among the three genotypes).
As far as we know, association studies comparing CYP11B2 polymorphism to LVM resulted in controversial results, even considering different hypertensive populations (15–17). In our opinion, when relaxed selection criteria are used when choosing subjects for analysis, the resulting background noise may mask a possible genotype–phenotype relationship. Consequently, contrasting results are not surprising (15–17)if patients considered for the individual studies differed in terms of race (15), presence of underlying disease, age, or washout period (16), to obtain a complete regression of cardiac changes pharmacologically induced (4). For these reasons, restrictive inclusion criteria, such as those requiring similar race, age, body dimensions, duration and severity of hypertension, and dietary habits, as well as no previous pharmacologic treatment, may facilitate the elucidation of genotype–phenotype relations, as in our study. To our knowledge, the only previous study (17)published with similarly restrictive recruitment criteria showed a trend similar to ours. Delles et al. (17)analyzed a small group of very young hypertensives matched to normotensives. They found greater values of cardiac diameters and an inadequate level of aldosterone secretion after sodium loading in hypertensive patients bearing the CC genotype, as compared with the other genotypes. No genotype differences were seen among normotensives. However, when the Delles' data are examined in depth, a trend similar to that found in our study is observed for LV septal and posterior wall thickening. In both of these studies, these parameters increased with the number of T alleles among hypertensive patients, although statistical significance was not reached. A lack of statistical significance in the Delles' study may be explained by both the reduced size of their sample and a different stage of hypertension (inasmuch as their patients were younger than the patients included in our study). Indeed, it is well known that each phase of the development of hypertension is characterized by different hemodynamic changes inducing different echocardiographic patterns (25).
To further strengthen the interpretation of the results from association studies, it is important to establish a plausible biologic explanation for a statistical relationship between the marker genotypes and the intermediate phenotypes related to the disease, trying to clarify the pathophysiologic meaning of the observed associations.
In the present study, a positive association between CYP11B2 and aldosterone/PRA was found, suggesting a possible regulatory role of adrenal aldosterone secretion for this polymorphism, as previously reported by some investigators (26,27), though with some contrasting results (28). The relationship between plasma aldosterone and LVH and fibrosis is well known (6–8). However, in our patients, no associations between cardiac dimensions and PRA, plasma aldosterone, or the aldosterone/PRA ratio were found, suggesting that LVH and wall thickening are not dependent on adrenal aldosterone production in these patients. Recent observations underlie the role for aldosterone's intracardiac synthesis in the pathogenesis of hypertrophy and fibrosis. This synthesis is regulated by an intracardiac aldosterone synthase, which seems modulated by CYP11B2 polymorphism (10–12). Thus, although indirectly, our data seem to support the hypothesis of a primary role of intracardiac aldosterone secretion that CYP11B2 modulated in the development of LVM, whereas adrenal aldosterone may not be that relevant. It is possible to hypothesize that CYP11B2 regulates aldosterone synthesis in different organs. In particular, at the adrenal level, it could regulate body volumes by controlling aldosterone secretion, whereas at the cardiac level, it could directly contribute to the increase of LVM and thicknesses by regulating aldosterone's paracrine production. Although cardiac aldosterone production has been found to be increased in hypertensives compared with normotensives (12), only measured plasmatic aldosterone (i.e., adrenal production) was measured in our patients. Thus, additional studies on cardiac aldosterone are needed to confirm the hypothesis suggested by this association study.
The present study demonstrates an association between CYP11B2 polymorphism and LVH. The association seems to be independent of adrenal aldosterone production and suggests a role for cardiac aldosterone production.
☆ This work was supported, in part, by grants from the Ministero Università e Ricerca Scientifica of Italy (Grant Cofin 2000/2002 MM06A92341-001 to Dr. Bianchi and MM06A92341-005 to Dr. Cusi) and by the grant FIRST “Special Project on Quantitative Evaluation of Tissue ACE in Hypertension.”
- blood pressure
- left ventricle/ventricular
- left ventricular hypertrophy
- left ventricular mass
- plasma renin activity
- posterior wall thickness
- relative wall thickness
- septal thickness
- American College of Cardiology Foundation
- Devereux R.B.,
- Roman M.J.
- Kaplan N.M.
- Brilla C.G.,
- Pick R.,
- Tan L.B.,
- et al.
- Yamamoto N.,
- Yasue H.,
- Mizuno Y.,
- et al.
- Kupari M.,
- Hautanen A.,
- Lankinen L.,
- et al.
- Schunkert H.,
- Hengstenberg C.,
- Holmer S.R.,
- et al.
- Delles C.,
- Erdmann J.,
- Jacobi J.,
- et al.
- Sahn D.J.,
- De Maria A.,
- Kisslo J.,
- et al.
- Devereux R.B.,
- Reichek N.
- Ganau A.,
- Devereux R.B.,
- Roman M.J.,
- et al.
- Sambrook J.,
- Fritsch E.F.,
- Maniatis T.
- Guyton A.C.
- Pojoga L.,
- Gautier S.,
- Blanc H.,
- et al.