Cardiac aldosterone overexpression prevents harmful effects of diabetes in the mouse heart by preserving capillary density

Transcription

1 The FASEB Journal Research Communication Cardiac aldosterone overexpression prevents harmful effects of diabetes in the mouse heart by preserving capillary density Smail Messaoudi,*, Paul Milliez,*, Jane-Lise Samuel,*,, and Claude Delcayre*,,1 *INSERM U942, Hopital Lariboisiere, Paris, France; Paris-Diderot University, Paris, France; Cardiology Department, Cote de Nacre University Hospital, Caen, France; and Lariboisiere Hospital, Assistance Publique-Hôpitaux de Paris, Paris, France ABSTRACT Recent reports showed an unexpected worsening of endothelial function by aldosterone antagonism in diabetic patients, suggesting that aldosterone could interfere with the detrimental consequences of diabetes on microvasculature and thus on cardiac function. To test this hypothesis, diabetes (D) was induced using streptozotocin in transgenic (Tg) male mice overexpressing aldosterone-synthase in the heart and in wild-type (Wt) mice. Eight weeks after streptozotocin injection, impairment of left ventricular systolic function, measured by echocardiography (fractional shortening), was accompanied by a decrease in capillaries/cardiomyocyte ratio ( 20%) and VEGFa expression ( 40%) in Wt-D mice compared with normoglycemic littermates. Furthermore, Wt-D mice demonstrated an increase in superoxide production ( 100%) and protein carbonylation ( 33%), hallmarks of oxidative stress. Except for a slight increase in protein carbonylation, all of these diabetes-associated cardiac alterations were undetectable in Tg-D mice. Fibrosis was induced similarly in both diabetic groups. Eplerenone (an aldosterone antagonist) abolished all of the effects of aldosterone-synthase overexpression but had no effect in Wt-D mice. Thus, aldosterone prevents systolic dysfunction through a mineralocorticoid receptor-dependent mechanism that may include preventing VEGFa down-regulation and maintaining capillary density. Understanding how aldosterone prevents VEGFa down-regulation in experimental diabetes could be important to define new strategies targeting the prevention of a decrease in capillary density. Messaoudi, S., Milliez, P., Samuel, J.-L., Delcayre, C. Cardiac aldosterone overexpression prevents harmful effects of diabetes in the mouse heart by preserving capillary density. FASEB J. 23, (2009) Key Words: hormone VEGFa oxidative stress streptozotocin transgenic animal Much less appreciated and more controversial is the notion that a specific diabetic cardiomyopathy referring to myocardial disease in diabetic patients that cannot be ascribed to other cardiovascular risk factors (4) may also contribute to the increased incidence of heart failure in patients with diabetes. Although the existence of such a clinical entity remains controversial, accumulating data from experimental, epidemiological, and clinical studies have shown that diabetes results in cardiac functional and structural changes in the absence of any specific cardiovascular risk factors (systemic hypertension, coronary artery disease, obesity or hypercholesterolemia) (5). Diverse pathogenic mechanisms contribute to the development of such functional and structural changes, including microvascular abnormalities, metabolic disturbances and fibrosis (5) among others. The microvascular abnormalities concern both the structure and the function of vessels. A recent study demonstrated a progressive reduction of the microvasculature in relation to the duration of diabetes. The decrease in capillary density was accompanied by decreased myocardial perfusion, a direct indicator of myocardial ischemia, and by progressive left ventricular dysfunction (6). Aldosterone plays an important pathophysiological role in cardiovascular disease (7). However, although mineralocorticoid receptor (MR) blockers (aldosterone antagonists) decrease mortality in heart failure, rare harmful effects on vasculature have been reported. Indeed, MR blockade by spironolactone worsens endothelial dysfunction of diabetic patients without heart failure (8). Experimentally, spironolactone blocks the enhancement of ischemia-induced neovascularization by aldosterone (9). These studies strongly suggest that aldosterone is not always deleterious to vascular function or structure in some well-defined pathological contexts such as uncomplicated diabetes or ischemia. To test the hypothesis that aldosterone could interfere with the detrimental consequences of diabetes on Diabetes is an independent risk factor for the development of heart failure (1). Conventional wisdom holds that diabetes promotes heart failure through accelerated atherosclerosis and hypertension (2, 3). 1 Correspondence: INSERM U942, Hopital Lariboisiere, 41 Blvd. de la chapelle, Paris, France. claude. delcayre@inserm.fr doi: /fj /09/ FASEB

2 microvasculature and thus on cardiac function, diabetes was induced in a transgenic strain of mice with cardiac-specific overexpression of aldosterone synthase (10). The functional consequences on the heart of such cardiac hyperaldosteronism under this diabetic condition were investigated by echocardiography, while capillary density was assessed by immunohistology. Furthermore, as microvasculature integrity is related to oxidative stress, production of reactive oxygen species and its consequences were studied by DHE staining and immunolabeling of carbonyl groups. MATERIALS AND METHODS Transgenic mice To study the cardiac-specific effects of aldosterone, we used transgenic (Tg) mice that overexpress the aldosterone synthase gene specifically in the heart (10). Briefly, the aldosterone synthase gene is driven by the -myosin heavy chain promoter, and its expression is thus restricted to cardiomyocytes (11). The transgene is detected in the heart of Tg mice, but not in kidney, brain, or skeletal muscle (unpublished data). In Tg mice, cardiac aldosterone concentration is increased by 1.7-fold relative to wild-type (Wt) mice, whereas the plasma aldosterone concentration is normal. As a consequence, blood pressure is normal and no renal or myocardial alterations are detectable. The only detectable alteration is coronary dysfunction in Tg male mice secondary to down-regulation of BKCa channels in coronary smooth muscle cells (12). Diabetic mouse model All experiments were performed in accordance with the European Community guidelines for the care and use of laboratory animals (no ). To induce diabetes, 16 wk-old Tg male mice with cardiac-specific overexpression of aldosterone synthase and their Wt littermates were injected intraperitoneally daily for 5 d with streptozotocin (STZ) (40 mg/kg/d) (Sigma-Aldrich, Saint Louis, MO, USA) dissolved in sodium citrate buffer. Three days after the last injection, whole-blood glucose was monitored using the Euroflash monitor (LifeScan, Milpitas, CA, USA). Streptozotocintreated mice with blood glucose concentration higher than 15 mm were considered diabetic. Eplerenone treatment Eplerenone (Tocris, Bristol, UK) is the most specific MR blocker. It was incorporated in food by SAFE (Augy, France). Treatment with eplerenone (150 mg/kg/d) began 2 wk before the injection of streptozotocin to be sure that the MR was fully blocked at the beginning of diabetes. Eplerenone treatment was maintained during the duration of diabetes (8 wk). Echocardiography Transthoracic echocardiography was performed in a blinded fashion using a GE Vivid 7 machine (General Electric Company, Fairfield, CT, USA) equipped with an 8- to 14-MHz linear transducer. Briefly, as described previously (13), cardiac dimensions, as well as the fractional shortening (FS) were measured in the parasternal long-axis view in M-mode. All recordings were performed on anesthetized animals with isoflurane (0.75%). Systolic dysfunction was defined as FS 36%. This cutoff was derived from the average FS (39.3%) of 5- to 6- mo-old Wt control mice minus 2 sd (1.55%). Blood pressure measurement Systolic blood pressure was measured monthly in unanesthetized mice by the tail-cuff method (BP-2000; Visitech Systems, Apec, NC, USA). Anatomical examination and tissue preparation After lethal anesthesia, body and heart weights were recorded, as well as tibia length. As the body weight of diabetic mice decreased, heart weight variations were expressed relative to tibia length to have a parameter independent of body weight variation. Hearts were included in Tissue-Tek (Sakura, Tokyo, Japan) and frozen in liquid nitrogen-cooled isopentane for immunohistological and biochemical studies. All samples were stored at 80 C until use. Histological examination and morphometry Equatorial cryostat sections (7 m) of the ventricles (LV and RV) were performed for histology and immunolabeling. Immunohistochemistry for caveolin1 and vinculin and analysis of capillary density Double immunolabeling on cryostat sections allowed the identification of endothelial cells with Caveolin 1 antibody from Santa Cruz Biotechnology (Santa Cruz, CA, USA) and cardiomyocytes by vinculin antibody from Sigma-Aldrich, as described previously (14). Left ventricular fields in which cross sections of capillaries and cardiomyocytes were clearly detectable (subendocardial area) were randomly recorded using a Leica camera equipped with a fluorescent Leica DMR (Leica Microsystems, Rueil Malmaison, France). A minimum of 6 fields/section was recorded at 20. The number of capillaries and cardiomyocytes were determined using IPLab software (BD Biosciences, San Jose, CA, USA) by a masked observer. The capillary density was defined as the number of capillaries per myocyte. Determination of fibrosis Cardiac cryostat sections were stained with the collagenspecific Sirius red stain (0.5% in saturated picric acid). The ratio of interstitial collagen surface-to-total ventricular surface was determined with IPLab software. Quantification of fibrosis was performed in a masked fashion on at least 3 areas/ heart. Superoxide anion detection in the heart Dihydroethidium (DHE; Sigma-Aldrich) staining was used to evaluate the in situ levels of superoxide anion in the myocardium. Cardiac cryostat sections were incubated with DHE (37 M) for 30 min in a dark humidified chamber. Acquisition of fluorescent images of ethidium bromide was as described above in the immunohistochemistry section. Sections of control and diabetic animals were analyzed in parallel with strictly identical imaging parameters. The stained area was measured with IPLab software and expressed as a percentage of total image area. BENEFICIAL ALDOSTERONE EFFECT IN DIABETIC HEART 2177

3 Protein study Proteins were extracted as described previously (14) and quantified with a Qubit (Invitrogen, Carlsbad, CA, USA). Supernatants were stored at 80 C. Protein oxidation To detect carbonyl groups, we used the OxyBlot Kit as described by the manufacturer (Millipore, Billerica, MA, USA). Samples were then dotted on an activated PVDF membrane using a 96-well dot-blot apparatus. The membrane was immunoblotted with a rabbit anti-dnp antibody (1:150, Millipore), Chemiluminescent signal was produced using ECL solution (Amersham, Chalfont St. Giles, UK) and detected with a LAS-3000 luminescent image analyzer (Fuji, Courbevoie, France). Relative densitometry was determined using the computer software Multi Gauge V2.3 (Fuji). Western blot analysis Proteins (20 g) were denatured 7 min at 99 C, separated by SDS-PAGE electrophoresis (nonreducing conditions), and electrotransferred onto nitrocellulose membranes using an Iblot (Invitrogen). An antibody against VEGFa (1:500; Santa Cruz Biotechnology) was used for immunoblotting. Actin was used as a protein-loading control and was detected with an antibody directed against total actin (1:1000; Sigma-Aldrich). Chemiluminescent signal was detected as described in protein oxidation section. Statistical analysis Results are expressed as means se. Comparison between groups was performed using analysis of variance (ANOVA), followed by multiple comparisons using unpaired Student s t test. P 0.05 was considered as statistically significant. RESULTS Biological parameters Basal glycemia did not differ between Wt and Tg control groups. Glycemia was doubled in Wt and Tg mice the first week after injection of streptozotocin (data not shown, P ) and further increased 2-fold 7 wk later, without any difference between the diabetic groups. Eight weeks of diabetes induced a decrease of body (P 0.001) and heart weights (P 0.05) in both Wt and Tg mice. In contrast, the systolic blood pressure was not altered by diabetes (Table 1). Echocardiographic measurements and systolic function Echocardiographic parameters were similar between Wt and Tg control groups at baseline. Four weeks after diabetes induction, there was a significant decrease in both septal and posterior wall thicknesses in diabetic groups compared with controls (P 0.005) (Table 1), which is consistent with the cardiac atrophy observed at autopsy. Left ventricular systolic function (fractional shortening) was moderately but significantly decreased in the Wt diabetic (Wt-D) group compared to the Wt group (P ). In contrast, systolic function was unchanged in the Tg diabetic (Tg-D) group compared to the Tg group. Similar morphological and functional data were observed after 8 wk of diabetes (Fig. 1). Cardiac remodeling: fibrosis The diabetic heart is a remodeled heart (15). We thus studied fibrosis, a well-recognized hallmark of cardiac remodeling. Histological analysis indicated that diabetes induced cardiac interstitial fibrosis to a similar extent in both Wt and Tg diabetic groups (P 0.05 vs. respective controls) (Fig. 2). Systolic function was thus unaltered in Tg-D mice despite the development of fibrosis. Capillary density The diabetic heart is characterized by a severe decrease in capillary density. As expected, diabetes induced a significant decrease in capillary density in the Wt-D group: 19%in4wk(P 0.005) (Fig. 3). In contrast, TABLE 1. General characteristics of control and diabetic mice following 8 wk of diabetes Wt Tg Variable Control Diabetic Control Diabetic Biological characteristic Blood glucose (mmol/dl) * BW (g) * HW(mg) * HW/TL (mg/mm) * Echocardiographic measurement IVSd (mm) * LVIDd (mm) LVPWd (mm) * Values are means se.*p 0.05 vs. Wt control; P 0.05 vs. Tg control. BW, body weight; HW, heart weight; TL, tibia length; IVSd, interventricular septum thickness in diastole; LVIDd, left ventricular internal diameter in diastole; LVPWd, left ventricular posterior wall thickness in diastole Vol. 23 July 2009 The FASEB Journal MESSAOUDI ET AL.

4 groups, which is an indicator of the oxidation status of proteins. As expected, cardiac proteins exhibited the highest level of carbonylation in the Wt-D group (P 0.05 vs. Wt and Tg-D), although Tg-D cardiac proteins displayed a moderate increase of carbonylation level (P 0.05 vs. Tg) (Fig. 4B). This slight increase could be due to other locally produced reactive oxygen species (ROS) such as hydroxyl radicals (16). All of these results, obtained with two independent methods, led us to conclude that oxidative stress was reduced in Tg-D mice in comparison with Wt-D mice. Figure 1. Left ventricular systolic function in control and diabetic Wt and Tg mice. Time-dependent changes of systolic function, represented by fractional shortening, in control and diabetic mice. Values are expressed as means se; n 8 15 animals/group. *P 0.05 vs. Wt control; P 0.05 vs. Tg-D. capillary density was unchanged in the Tg-D group. Four weeks later (8 wk of diabetes), there was a stabilization of capillary density in the Wt-D group ( 19%, P vs. Wt mice) and a sustained preservation of capillary density in the Tg-D group. Notably, at this time, there was a modest but significant (P 0.05) difference in the capillary density of Wt ( ) and Tg mice ( ). Oxidative stress We studied several parameters of oxidative stress, as it is associated with decreases in capillary density. First, we analyzed superoxide anion production by DHE staining. As shown in Fig. 4A, DHE labeling was much stronger in the myocardium of Wt-D group than in all other groups. In addition, the oxidative stress was evaluated through the immunodetection of carbonyl VEGFa expression To investigate the potential mechanisms by which aldosterone prevented the reduction in capillary density, we analyzed the cardiac expression of VEGFa. As the capillary density was decreased as soon as 4 wk after diabetes induction, we focused our investigations on the first and fourth weeks of diabetes. No change in VEGFa protein level was found at the first week. In contrast, a decrease of 40% in VEGFa expression was noted in the Wt-D group (P 0.005) at the fourth week. This decrease was not observed in the Tg-D group (Fig. 5). Similar results were observed after 8 wk of diabetes (data not shown). Mineralocorticoid receptor blockade The classical actions of aldosterone are mediated through the mineralocorticoid receptor (MR), although it has been demonstrated that aldosterone can act independently of this receptor (17). To gain further insights into the mechanisms involved, we reasoned that if the activation of the MR were responsible for the beneficial effects of aldosterone, the selective blockade of this receptor should Figure 2. Fibrosis following 8 wk of diabetes. Sirius red staining of myocardial sections of diabetic and control mice (left) and quantification of interstitial collagen (right). Values are expressed as means se; n 5 6 animals/group. *P 0.05 vs. Wt; P 0.05 vs. Tg. Scale bar 100 m. BENEFICIAL ALDOSTERONE EFFECT IN DIABETIC HEART 2179

5 Figure 3. Capillary density quantification following 8 wk of diabetes. A) Immunostaining of cardiomyocytes (antivinculin antibody) and capillaries (anticaveolin1 antibody) in myocardium sections. Scale bar 100 m. B) Time-dependent changes of capillary density in control and diabetic mice. Values are expressed as means se; n 5 8 animals/group. *P 0.05 vs. Wt; P 0.05 vs. Tg-D; P 0.05 vs. Wt. abolish the protection conferred by aldosterone. Thus, animals were treated by eplerenone (the most selective MR blocker) 2 wk before the injection of streptozotocin and throughout the 8 wk of diabetes. In Wt and Tg controls, eplerenone treatment had no effects on the systolic function, but morphometric analysis showed that eplerenone normalized the capillary density of Tg control mice (Table 2). Notably, eplerenone treatment did not affect streptozotocininduced diabetes since both eplerenone-treated Wt-D and Tg-D mice developed a diabetic phenotype similar to that of nontreated mice (hyperglycemia, weight loss, heart atrophy; data not shown). Eplerenone-treated Wt-D mice had mild systolic dysfunction and a reduction in capillary density similar to those observed in untreated Wt-D mice. In contrast, while Tg-D mice were protected from systolic dysfunction and capillary density reduction, eplerenone-treated Tg-D mice had both a decreased fractional shortening (P 0.05) and a reduction of capillary density (P 0.001) (Fig. 6, A, B). Finally, we observed a down-regulation of cardiac VEGFa protein expression in both eplerenone-treated Wt-D ( 40%, P 0.05) and Tg-D mice ( 21%, P 0.05) (Fig. 6C). Thus, the prevention of systolic dysfunction, decrease in capillary density, and VEGFa protein down-regulation in Tg-D mice was suppressed by eplerenone treatment. No changes were observed following eplerenone treatment in the Wt-D mice. DISCUSSION This study demonstrates that overexpression of aldosterone synthase in cardiomyocytes prevents the development of systolic dysfunction in streptozotocin-induced diabetic conditions. This protective effect may be linked to the preservation of myocardial VEGFa expression and capillary density Vol. 23 July 2009 The FASEB Journal MESSAOUDI ET AL.

6 Figure 4. Oxidative stress in control and diabetic Wt and Tg myocardium following 8 wk of diabetes. A) Illustration (top) and quantification (bottom) of superoxide anion production in the myocardium by DHE staining. Scale bar 100 m. B) Protein carbonylation immunodetection by oxyblot. Values are expressed as means se; n 5 9 animals/group. *P 0.05 vs. Wt; P 0.05 vs. Tg; P 0.05 vs. Tg-D. The Tg mouse model Our work aimed to test the hypothesis that aldosterone may interfere with the deleterious effects of diabetes on the cardiac microvasculature and function. To carry out this study, it was important to use an animal model that permitted the distinction between the specific effects of aldosterone on the heart from those on other organs. A systemic infusion of aldosterone would therefore not have been appropriate, since it would have led to an increased aldosterone concentration in the whole organism, with a series of associated systemic effects (this point is discussed below). In this context, the Tg mice are a more appropriate model. Indeed, since aldosterone synthase overexpression is restricted to cardiomyocytes, the aldosterone concentration is increased in heart only, whereas it is normal in plasma. The Tg mice may, therefore, be considered to be a sophisticated minipump model delivering aldosterone in the heart only, as opposed to the systemic distribution of classical osmotic minipumps. Notably, the transgenic expression had no antidiabetic effect per se, since Tg-D mice developed hyperglycemia, polyuria, hypercreatinemia (data not shown), and weight loss, as did Wt-D mice. Diabetes induction was thus equally effective in both Wt and Tg mice. Interestingly, the coronary dysfunction of Tg mice did not promote and/or worsen the systolic dysfunction in diabetic conditions. In the same way, a decreased coronary reserve without systolic dysfunction was also reported in diabetic patients (18). VEGFa expression and capillary density Figure 5. VEGFa expression following 4 wk of diabetes. Western blot analysis of VEGFa (top) and quantification of VEGFa to actin ratio (bottom). Values are expressed as means se; n 4 6 animals/group. *P 0.05 vs. Wt; P 0.05 vs. Tg-D. Diabetes mellitus is associated with macrovascular and microvascular abnormalities. Yoon et al. (6) demonstrated that decreased capillary density and myocardial perfusion in the diabetic heart leads to cardiac dysfunction, with VEGFa protein down-regulation being the seminal event to all these features. Our results obtained in the Wt-D mice are in complete agreement with these findings. Interestingly, the diabetes-induced VEGFa down-regulation was prevented in the Tg-D hearts. This result is in agreement with other studies indicating that aldosterone induces VEGFa expression both in vitro in tubular cells (19) and in vivo in skeletal muscle (9). VEGFa plays an important role in angiogenesis (20) and transgenic mice lacking the two major myocardial VEGFa isoforms exhibited impaired myocardial angio- BENEFICIAL ALDOSTERONE EFFECT IN DIABETIC HEART 2181

7 TABLE 2. Effect of eplerenone on systolic function and capillary density of control mice (Wt and Tg) Wt Tg Variable Eple Eple Eple Eple FS (%) Capillaries/cardiomyocyte * Values are means se. Eple, eplerone. *P 0.05 vs. Wt -eple. P 0.05 vs. Tg-eple. genesis and subsequently developed ischemic cardiomyopathy and severe LV dysfunction (21). In our model, the preservation of VEGFa expression is associated with the maintenance of a normal capillary density and systolic function in Tg-D mice. In contrast, in eplerenone-treated Tg-D mice, the VEGFa was downregulated, the capillary bed was less dense and the fractional shortening was decreased. These results demonstrate that aldosterone activation of the mineralocorticoid receptor plays a key role in the cardiac phenotype observed in the Tg-D mice, but they also support the concept of a possible relationship between VEGFa expression, the microvasculature, and cardiac contractile function, independently of environmental conditions. Aldosterone through a MR-dependent mechanism counteracts an important pathophysiological cascade of events leading to the perturbation of microvascular homeostasis in the diabetic myocardium. Oxidative stress and fibrosis In addition to VEGFa down-regulation and a decrease in capillary density, diabetes results in the production of multiple forms of ROS (22). Interestingly, oxidative stress was reduced in the Tg diabetic myocardium in comparison with Wt-D hearts. The mechanisms responsible for the decreased oxidative stress in Tg-D hearts were not explored. However, it is likely that maintenance of capillary density was involved. Indeed, capillary density is a determinant of myocardial perfusion, and thus of myocardial hypoxia and oxidative stress. Nevertheless, it cannot be excluded that aldosterone in Tg-D mice had a direct effect on myocardial superoxide production. Further studies are thus needed to clarify this point. Another important point of this study is that despite having normal myocardial VEGFa expression, capillary density, systolic function, and ROS levels, Tg-D mice exhibited structural alterations. Both Wt-D and Tg-D mice developed fibrosis. Diabetes-induced fibrosis results from complex mechanisms (23), and it is likely that aldosterone effects in Tg-D mice are insufficient to modify this process. In experimental models, fibrosis appears to be more related to diastolic than to systolic dysfunction. Indeed, reversal of cardiac fibrosis attenuates increased diastolic stiffness without normalizing cardiac contractility in streptozotocin-induced diabetic rats (24). Moreover, the restoration of capillary density and systolic function is not accompanied by a decrease of cardiac fibrosis (6). Our results confirm these previous findings since Tg-D have a normal systolic function despite the presence of fibrosis. Figure 6. Effects of mineralocorticoid receptor blockade by eplerenone on heart function, capillary density, and VEGFa protein expression after 8 wk of diabetes. A) Fractional shortening in Wt and Tg diabetic mice treated with eplerenone. B) Capillary density in Wt and Tg diabetic mice treated with eplerenone. C) Western blot analysis of VEGFa (top) and quantification of VEGFa to actin ratio in Wt and Tg diabetic mice treated with eplerenone (bottom). Values are expressed as means se; n 4 6 animals/group. eple, eplerone. P 0.05 vs. Wt eple; # P 0.05 vs. Tg eple, **P 0.05 vs. Tg-D eple Vol. 23 July 2009 The FASEB Journal MESSAOUDI ET AL.

8 The Tg-D phenotype: possible explanations In the context of the current knowledge of aldosterone effects in heart failure, our results may seem surprising. Indeed, clinical studies demonstrate beneficial effects of MR blockade in patients with heart failure (25, 26). On the basis of this evidence, one may think that blockade of the MR action in such circumstances is beneficial in terms of both disease progression and outcomes. However, it is probably not so simple. Indeed, Beggah et al. (27) report that a cardiomyocytespecific knock-down of the MR in mice leads paradoxically to severe heart failure and cardiac fibrosis. Although the mechanisms responsible for this surprising phenotype are still not established, this study remarkably demonstrates that an inappropriate blockade of the MR may have disastrous cardiac consequences. Furthermore, understanding the role played by the MR in cardiovascular pathophysiology is complicated not only by its apparently important role in cardiac physiology, but also by the uncertainty regarding the nature of its ligand. Indeed, contrary to vascular cells (endothelial and smooth muscle cells) which express the 11 -hydroxysteroid dehydrogenase Type 2 (11 -HSD2), cardiomyocytes do not express this enzyme, which determines MR selectivity. Thus, several authors suggest that the cardiovascular damage classically attributed to aldosterone, such as fibrosis, cannot be mediated by aldosterone and are actually driven by inappropriate glucocorticoids mediated MR activation (28, 29). Our study avoids this ambiguity, since the phenotype observed in our Tg-D mice is clearly the consequence of an aldosterone-mediated MR activation. Another important point to keep in mind is that the patients enrolled in the RALES and EPHESUS studies likely had elevated plasma levels of aldosterone. Indeed, an escape of aldosterone in patients with chronic heart failure under ACE inhibitor treatment is well documented (30, 31), and an increase of renin angiotensin aldosterone system activity is found in postinfarction (32 35). Moreover, circulating aldosterone levels are also strongly increased in the experimental models of aldosterone-induced vascular oxidative stress and cardiovascular fibrosis (36 38). In contrast, our previously published results in Tg mice have demonstrated that a slight increase in cardiac aldosterone in a context of normal circulating aldosterone concentration has no deleterious effects on the cardiac function or structure. Furthermore, the coronary dysfunction of this strain is not due to an increase in oxidative stress, as it is usually observed in experimental models with high circulating aldosterone levels (39), but to a subtle down-regulation of the BKCa channels of the coronary smooth muscle cells (12). It seems thus that circulating aldosterone levels have to be high to induce cardiovascular damage. This reasoning seems to be also applicable in clinical conditions. Indeed, a recent study of Stewart and colleagues demonstrates that in asymptomatic patients with moderate to severe aortic stenosis and a normal plasma level of aldosterone, eplerenone does not slow the onset of LV systolic or diastolic dysfunction, decrease LV mass, or reduce progression of valve stenosis (40). In clinical or experimental diabetes, plasma levels of aldosterone are normal (41) or low (42). To our knowledge, no beneficial effects of aldosterone antagonists on cardiac function have been reported in diabetic patients (the diabetic patients enrolled in the EPHESUS study had postinfarction LV dysfunction, and had thus likely elevated plasma levels of aldosterone). On the contrary, a worsening of endothelial function has been reported in type 2 diabetic patients without heart failure treated with spironolactone (8). In experimental models of diabetes, MR blockade does not improve systolic dysfunction. Indeed, spironolactone failed to normalize cardiac contractility (24) or to improve the contractile function of the working heart of streptozotocin diabetic rats (43). Our results in Wt-D mice confirm and extend these findings to another MR blocker, eplerenone. Taken together, all these data strongly suggest that aldosterone does not participate in the deterioration of the systolic function in diabetes, at least before the development of heart failure. Limitations One of the major limitations of the present study is that the patients enrolled in the study of Davies et al. (8) suffer from type 2 diabetes. This may represent interference between insulin resistance and/or hyperinsulinemia and the MR receptor. The use of an insulinopenic model, such as streptozotocin-induced diabetes, could thus seem poorly relevant. However, it has been demonstrated that mice with restricted knockout of the insulin receptor, but normal insulinemia, have reduced cardiac capillary density under stress (44). This suggests that the vascular defects observed in diabetes result from impaired insulin signaling, likely in conjunction with stressful conditions such as hyperglycemia, rather than from hyperinsulinemia. It is thus possible that the effects of spironolactone on vasculature in the Davies study (8) result from an interference between insulin resistance and the MR receptor. As the streptozotocin model combines hyperglycemia and dramatically impaired insulin signaling, this model has thus been chosen to carry out our study. The use of this model presents, however, some limitations. First, interpreting results in this model may be complicated by nonspecific toxicity of streptozotocin. To mitigate the possible confounding effects of streptozotocin cytotoxicity, we adopted multiple low-dose streptozotocin injections according to the recommendations of the Animal Models of Diabetic Complications Consortium (AMDCC) (45). Second, although streptozotocin-induced diabetes is now a well-recognized and standardized model for type 1 diabetes and diabetes-associated cardiac alterations (46), it does not represent the more prevalent type 2 diabetes in humans. However, one might expect that similar results to ours might be obtained in an BENEFICIAL ALDOSTERONE EFFECT IN DIABETIC HEART 2183

9 experimental model of type 2 diabetes. Indeed, studies in experimental models of type 1 and type 2 diabetes have highlighted similarities in the cardiac phenotypes that are associated with insulin resistance and obesity and those with insulin deficiency (47). This suggests some common mechanisms that lead to specific alterations. This hypothesis needs, however, to be tested. CONCLUSION In conclusion, this study strengthens the concept of a central role played by decreased VEGFa levels in diabetes-induced capillary density decrease. Moreover, our work demonstrates that aldosterone prevents the development of systolic dysfunction under diabetic conditions. This beneficial effect is MR dependent and may include the preservation of VEGFa myocardial expression and capillary network integrity. The mechanisms by which aldosterone prevents VEGFa down-regulation in the diabetic myocardium remain unknown and merit further studies to be elucidated. Understanding these mechanisms could be of importance to design and define new strategies targeting the prevention of capillary density decrease. The authors thank Dr Annabel Chen-Tournoux for editorial assistance. This work was supported by the Groupe de Reflexion pour la Recherche Cardiovasculaire, the European Section of Aldosterone Council, The Groupe Paris Nord Circulation, the Fondation de France, the CNRS, and IN- SERM. REFERENCES 1. Kannel, W. B., Hjortland, M., and Castelli, W. P. (1974) Role of diabetes in congestive heart failure: the Framingham study. Am. J. Cardiol. 34, Beckman, J. A., Creager, M. A., and Libby, P. (2002) Diabetes and atherosclerosis: epidemiology, pathophysiology, and management. JAMA 287, Reaven, G. M., Lithell, H., and Landsberg, L. (1996) Hypertension and associated metabolic abnormalities the role of insulin resistance and the sympathoadrenal system. N. Engl. J. Med. 334, Rubler, S., Dlugash, J., Yuceoglu, Y. Z., Kumral, T., Branwood, A. W., and Grishman, A. (1972) New type of cardiomyopathy associated with diabetic glomerulosclerosis. Am. J. Cardiol. 30, Fang, Z. Y., Prins, J. B., and Marwick, T. H. (2004) Diabetic cardiomyopathy: evidence, mechanisms, and therapeutic implications. Endocr. Rev. 25, Yoon, Y. S., Uchida, S., Masuo, O., Cejna, M., Park, J. S., Gwon, H. C., Kirchmair, R., Bahlman, F., Walter, D., Curry, C., Hanley, A., Isner, J. M., and Losordo, D. W. (2005) Progressive attenuation of myocardial vascular endothelial growth factor expression is a seminal event in diabetic cardiomyopathy: restoration of microvascular homeostasis and recovery of cardiac function in diabetic cardiomyopathy after replenishment of local vascular endothelial growth factor. Circulation 111, Delcayre, C., and Swynghedauw, B. (2002) Molecular mechanisms of myocardial remodeling. The role of aldosterone. J. Mol. Cell. Cardiol. 34, Davies, J. I., Band, M., Morris, A., and Struthers, A. D. (2004) Spironolactone impairs endothelial function and heart rate variability in patients with type 2 diabetes. Diabetologia 47, Michel, F., Ambroisine, M. L., Duriez, M., Delcayre, C., Levy, B. I., and Silvestre, J. S. (2004) Aldosterone enhances ischemiainduced neovascularization through angiotensin II-dependent pathway. Circulation 109, Garnier, A., Bendall, J. K., Fuchs, S., Escoubet, B., Rochais, F., Hoerter, J., Nehme, J., Ambroisine, M. L., De Angelis, N., Morineau, G., d Estienne, P., Fischmeister, R., Heymes, C., Pinet, F., and Delcayre, C. (2004) Cardiac-specific increase in aldosterone production induces coronary dysfunction in aldosterone synthase-transgenic mice. Circulation 110, Subramaniam, A., Jones, W. K., Gulick, J., Wert, S., Neumann, J., and Robbins, J. (1991) Tissue-specific regulation of the alphamyosin heavy chain gene promoter in transgenic mice. J. Biol. Chem. 266, Ambroisine, M. L., Favre, J., Oliviero, P., Rodriguez, C., Gao, J., Thuillez, C., Samuel, J. L., Richard, V., and Delcayre, C. (2007) Aldosterone-induced coronary dysfunction in transgenic mice involves the calcium-activated potassium (BKCa) channels of vascular smooth muscle cells. Circulation 116, Milliez, P., Deangelis, N., Rucker-Martin, C., Leenhardt, A., Vicaut, E., Robidel, E., Beaufils, P., Delcayre, C., and Hatem, S. N. (2005) Spironolactone reduces fibrosis of dilated atria during heart failure in rats with myocardial infarction. Eur. Heart. J. 26, Ratajczak, P., Oliviero, P., Marotte, F., Kolar, F., Ostadal, B., and Samuel, J. L. (2005) Expression and localization of caveolins during postnatal development in rat heart: implication of thyroid hormone. J. Appl. Physiol. 99, Dhalla, N. S., Liu, X., Panagia, V., and Takeda, N. (1998) Subcellular remodeling and heart dysfunction in chronic diabetes. Cardiovasc. Res. 40, Dalle-Donne, I., Aldini, G., Carini, M., Colombo, R., Rossi, R., and Milzani, A. (2006) Protein carbonylation, cellular dysfunction, and disease progression. J. Cell. Mol. Med. 10, Uhrenholt, T. R., Schjerning, J., Rasmussen, L. E., Hansen, P. B., Norregaard, R., Jensen, B. L., and Skott, O. (2004) Rapid non-genomic effects of aldosterone on rodent vascular function. Acta Physiol. Scand. 181, Strauer, B. E., Motz, W., Vogt, M., and Schwartzkopff, B. (1997) Impaired coronary flow reserve in NIDDM: a possible role for diabetic cardiopathy in humans. Diabetes 46(Suppl. 2), S119 S Lai, L., Pen, A., Hu, Y., Ma, J., Chen, J., Hao, C. M., Gu, Y., and Lin, S. (2007) Aldosterone upregulates vascular endothelial growth factor expression in mouse cortical collecting duct epithelial cells through classic mineralocorticoid receptor. Life Sci. 81, Tammela, T., Enholm, B., Alitalo, K., and Paavonen, K. (2005) The biology of vascular endothelial growth factors. Cardiovasc. Res. 65, Carmeliet, P., Ng, Y. S., Nuyens, D., Theilmeier, G., Brusselmans, K., Cornelissen, I., Ehler, E., Kakkar, V. V., Stalmans, I., Mattot, V., Perriard, J. C., Dewerchin, M., Flameng, W., Nagy, A., Lupu, F., Moons, L., Collen, D., D Amore, P. A., and Shima, D. T. (1999) Impaired myocardial angiogenesis and ischemic cardiomyopathy in mice lacking the vascular endothelial growth factor isoforms VEGF164 and VEGF188. Nat. Med. 5, Fiordaliso, F., Cuccovillo, I., Bianchi, R., Bai, A., Doni, M., Salio, M., De Angelis, N., Ghezzi, P., Latini, R., and Masson, S. (2006) Cardiovascular oxidative stress is reduced by an ACE inhibitor in a rat model of streptozotocin-induced diabetes. Life Sci. 79, Asbun, J., and Villarreal, F. J. (2006) The pathogenesis of myocardial fibrosis in the setting of diabetic cardiomyopathy. J. Am. Coll. Cardiol. 47, Miric, G., Dallemagne, C., Endre, Z., Margolin, S., Taylor, S. M., and Brown, L. (2001) Reversal of cardiac and renal fibrosis by pirfenidone and spironolactone in streptozotocin-diabetic rats. Br. J. Pharmacol. 133, Pitt, B., Zannad, F., Remme, W. J., Cody, R., Castaigne, A., Perez, A., Palensky, J., and Wittes, J. (1999) The effect of spironolactone on morbidity and mortality in patients with severe heart failure. Randomized Aldactone Evaluation Study Investigators. N. Engl. J. Med. 341, Pitt, B., Remme, W., Zannad, F., Neaton, J., Martinez, F., Roniker, B., Bittman, R., Hurley, S., Kleiman, J., and Gatlin, M. (2003) Eplerenone, a selective aldosterone blocker, in patients 2184 Vol. 23 July 2009 The FASEB Journal MESSAOUDI ET AL.

10 with left ventricular dysfunction after myocardial infarction. N. Engl. J. Med. 348, Beggah, A. T., Escoubet, B., Puttini, S., Cailmail, S., Delage, V., Ouvrard-Pascaud, A., Bocchi, B., Peuchmaur, M., Delcayre, C., Farman, N., and Jaisser, F. (2002) Reversible cardiac fibrosis and heart failure induced by conditional expression of an antisense mrna of the mineralocorticoid receptor in cardiomyocytes. Proc. Natl. Acad. Sci. U. S. A. 99, Young, M., and Funder, J. W. (2004) Eplerenone, but not steroid withdrawal, reverses cardiac fibrosis in deoxycorticosterone/salt-treated rats. Endocrinology 145, Nagata, K., Obata, K., Xu, J., Ichihara, S., Noda, A., Kimata, H., Kato, T., Izawa, H., Murohara, T., and Yokota, M. (2006) Mineralocorticoid receptor antagonism attenuates cardiac hypertrophy and failure in low-aldosterone hypertensive rats. Hypertension 47, Pitt, B. (1995) Escape of aldosterone production in patients with left ventricular dysfunction treated with an angiotensin converting enzyme inhibitor: implications for therapy. Cardiovasc. Drugs Ther. 9, Staessen, J., Lijnen, P., Fagard, R., Verschueren, L. J., and Amery, A. (1981) Rise of plasma aldosterone during long-term captopril treatment. N. Engl. J. Med. 304, Fraccarollo, D., Galuppo, P., Hildemann, S., Christ, M., Ertl, G., and Bauersachs, J. (2003) Additive improvement of left ventricular remodeling and neurohormonal activation by aldosterone receptor blockade with eplerenone and ACE inhibition in rats with myocardial infarction. J. Am. Coll. Cardiol. 42, Michel, J. B., Lattion, A. L., Salzmann, J. L., Cerol, M. L., Philippe, M., Camilleri, J. P., and Corvol, P. (1988) Hormonal and cardiac effects of converting enzyme inhibition in rat myocardial infarction. Circ. Res. 62, Pinto, Y. M., de Smet, B. G., van Gilst, W. H., Scholtens, E., Monnink, S., de Graeff, P. A., and Wesseling, H. (1993) Selective and time-related activation of the cardiac renin-angiotensin system after experimental heart failure: relation to ventricular function and morphology. Cardiovasc. Res. 27, Schunkert, H., Tang, S. S., Litwin, S. E., Diamant, D., Riegger, G., Dzau, V. J., and Ingelfinger, J. R. (1993) Regulation of intrarenal and circulating renin-angiotensin systems in severe heart failure in the rat. Cardiovasc. Res. 27, Sun, Y., Ahokas, R. A., Bhattacharya, S. K., Gerling, I. C., Carbone, L. D., and Weber, K. T. (2006) Oxidative stress in aldosteronism. Cardiovasc. Res. 71, Rocha, R., Rudolph, A. E., Frierdich, G. E., Nachowiak, D. A., Kekec, B. K., Blomme, E. A., McMahon, E. G., and Delyani, J. A. (2002) Aldosterone induces a vascular inflammatory phenotype in the rat heart. Am. J. Physiol. Heart Circ. Physiol. 283, H1802 H Schiffrin, E. L. (2006) Effects of aldosterone on the vasculature. Hypertension 47, Virdis, A., Neves, M. F., Amiri, F., Viel, E., Touyz, R. M., and Schiffrin, E. L. (2002) Spironolactone improves angiotensininduced vascular changes and oxidative stress. Hypertension 40, Stewart, R. A., Kerr, A. J., Cowan, B. R., Young, A. A., Occleshaw, C., Richards, A. M., Edwards, C., Whalley, G. A., Freidlander, D., Williams, M., Doughty, R. N., Zeng, I., and White, H. D. (2008) A randomized trial of the aldosterone-receptor antagonist eplerenone in asymptomatic moderate-severe aortic stenosis. Am. Heart J. 156, De Azevedo, M. J., Ramos, O. L., and Gross, J. L. (1995) Renin-aldosterone axis in normoalbuminuric insulin-dependent diabetes mellitus patients with glomerular hyperfiltration. Diabetes Res. Clin. Pract. 27, Cronin, C. C., Barry, D., Crowley, B., and Ferriss, J. B. (1995) Reduced plasma aldosterone concentrations in randomly selected patients with insulin-dependent diabetes mellitus. Diabet. Med. 12, Verma, S., Yuen, V. G., Badiwala, M., Anderson, T. J., and McNeill, J. H. (2003) Working heart function in diabetes is not improved by spironolactone treatment. Can. J. Physiol. Pharmacol. 81, McQueen, A. P., Zhang, D., Hu, P., Swenson, L., Yang, Y., Zaha, V. G., Hoffman, J. L., Yun, U. J., Chakrabarti, G., Wang, Z., Albertine, K. H., Abel, E. D., and Litwin, S. E. (2005) Contractile dysfunction in hypertrophied hearts with deficient insulin receptor signaling: possible role of reduced capillary density. J. Mol. Cell. Cardiol. 39, Breyer, M. D., Bottinger, E., Brosius, F. C., 3rd, Coffman, T. M., Harris, R. C., Heilig, C. W., and Sharma, K. (2005) Mouse models of diabetic nephropathy. J. Am. Soc. Nephrol. 16, Hsueh, W., Abel, E. D., Breslow, J. L., Maeda, N., Davis, R. C., Fisher, E. A., Dansky, H., McClain, D. A., McIndoe, R., Wassef, M. K., Rabadan-Diehl, C., and Goldberg, I. J. (2007) Recipes for creating animal models of diabetic cardiovascular disease. Circ. Res. 100, Poornima, I. G., Parikh, P., and Shannon, R. P. (2006) Diabetic cardiomyopathy: the search for a unifying hypothesis. Circ. Res. 98, Received for publication November 6, Accepted for publication February 12, BENEFICIAL ALDOSTERONE EFFECT IN DIABETIC HEART 2185