Serotonin 5-HT2B Receptor Blockade Prevents Reactive Oxygen Species–Induced Cardiac Hypertrophy in Mice
We established previously that 5-HT2B receptors are involved in cardiac hypertrophy through the regulation of hypertrophic cytokines in cardiac fibroblasts. Moreover, the generation of reactive oxygen species and tumor necrosis factor-α through the activation of reduced nicotinamide-adenine dinucleotide phosphate [NAD(P)H] oxidase has been implicated in cardiac hypertrophy. In this study, we investigated whether 5-HT2B receptors could be involved in the development of cardiac hypertrophy associated with superoxide anion production. Therefore, we measured the effects of serotonergic 5-HT2B receptor blockade on left-ventricular superoxide anion generation in 2 established pharmacological models of cardiac hypertrophy, ie, angiotensin II and isoproterenol infusions in mice. Angiotensin II infusion for 14 days increased superoxide anion concentration (+32%), NAD(P)H oxidase maximal activity (+84%), and p47phox NAD(P)H oxidase subunit expression in the left ventricle together with hypertension (+37 mm Hg) and cardiac hypertrophy (+17% for heart weight:body weight). The 5-HT2B receptor blockade by a selective antagonist (SB215505) prevented the increase in cardiac superoxide generation and hypertrophy. Similarly, infusion for 5 days of isoproterenol increased left-ventricular NAD(P)H oxidase activity (+48%) and cardiac hypertrophy (+31%) that were prevented by the 5-HT2B receptor blockade. Finally, in the primary culture of left-ventricular cardiac fibroblasts, angiotensin II and isoproterenol stimulated NAD(P)H oxidase activity. This activation was prevented by SB215505. These findings suggest that the 5-HT2B receptor may represent a new target to reduce cardiac hypertrophy and oxidative stress. Its blockade affects both angiotensin II and β-adrenergic trophic responses without significant hemodynamic alteration.
Pathological left ventricular hypertrophy has been associated with increased production of reactive oxygen species (ROS). Recent works have shown that antioxidants inhibit in vitro cardiomyocyte hypertrophy and in vivo aortic banding-induced left ventricular hypertrophy.1 The cardiac ROS production can be triggered by factors such as angiotensin II (Ang II), catecholamines, endothelin-1, tumor necrosis factor-α (TNF-α), and serotonin but also by mechanical stretch. Among these factors, those that signal through the Gq/PLC pathway seem to play a crucial role in the initiation and maintenance of cardiac hypertrophy and are known to stimulate the cardiac ROS generation through the phagocyte-type reduced nicotinamide-adenine dinucleotide phosphate [NAD(P)H] oxidase.2 The gp91phox-containing NAD(P)H oxidase plays a pivotal role in the response to Ang II via a pathway involving protein kinase C, c-Src, and phosphatidylinositol 3-kinase.3 NAD(P)H oxidase reduces oxygen O2, leading to the formation of superoxide anion (O2·−), which can be either dismutated spontaneously to hydrogen peroxide or in a reaction involving superoxide dismutase (SOD).
The 5-HT2B receptor (5-HT2BR) is a Gq/G11 protein-coupled receptor that has been shown to be functionally coupled to ROS synthesis through NAD(P)H oxidase stimulation in a neuroectodermal cell line (1C11).4 Interestingly, 5-HT2BRs appear to control the TNF-α shedding in the extracellular space via NAD(P)H oxidase–dependent TNF-α–converting enzyme activation. This 5-HT2BR–dependent NAD(P)H oxidase activation could contribute to the previously described effect of 5-HT2BR blockade on TNF-α release by ventricular fibroblasts after isoproterenol (ISO) stimulation. We established that 5-HT2BRs are essential for ISO-induced cardiac hypertrophy and are involved in the regulation of hypertrophic cytokines, interleukin-6, interleukin-1β, and TNF-α production by cardiac fibroblasts.5 We hypothesized that 5-HT2BR blockade could affect NAD(P)H-oxidase function and, therefore, participate in the modulation of hypertrophic pathways implying this key enzyme.
The aim of the present study was to determine whether 5-HT2BRs could participate in O2·− generation during the course of pharmacological-induced cardiac hypertrophy by ISO and Ang II. Its roles in the regulation of the oxidative balance between NAD(P)H oxidase and SOD activities and in the O2·− production by left-ventricular fibroblasts have been investigated.
Studies were performed in 129S1/ScImJ mice in accordance with the Canadian Council for Animal Care Guidelines and monitored by the ethical committee for experimental research from the Clinical Research Institute of Montreal Canada (No. 2004-11).
Induction of Cardiac Hypertrophy by ISO and Ang II
Mice were infused by either vehicle, ISO (30 mg · kg−1 · d−1), or Ang II (0.2 mg · kg−1 · d−1; please see the detailed procedures available in the online data supplement at http://hyper.ahajournals. org). The following drugs were tested on cardiovascular responses to ISO and/or Ang II: the 5-HT2BR antagonist SB215505 (1 mg · kg−1 · d−1), the antagonist of β-adrenergic receptors propranolol (5 mg · kg−1 · d−1), or the NAD(P)H oxidase inhibitor apocynin (1.5 mmol/L, drinking water).
Heart rate and systolic arterial pressure were recorded by the tail-cuff method (Visitech), and transthoracic echocardiograms were performed in 2% isoflurane anesthetized mice, as described previously.5 After euthanasia, the heart was weighed and quickly frozen after separation of right and left ventricles and atria. In some experiments, the abdominal aorta was sampled.
Basal and NAD(P)H-stimulated O2·− ventricular productions were measured twice for each ventricle using the lucigenin-enhanced chemoluminescence method as described previously6 and related to milligrams of tissue. For Ang II–treated mice, the aortic O2− production was also measured. Approximately 2 to 3 mg of tissue sample were placed in a glass vial containing 2 mL of a lucigenin solution (5 μmol/L). After the measurement of basal luminescence, 10−4 mol/L of NAD(P)H were added to the same vial to evaluate the maximal O2− production by NAD(P)H oxidase.
SOD Activity Measurements
SOD enzymatic activity was measured according to the hematoxylin method of Chattopadhyay et al.7 The enzymatic reaction was assessed with 10 μg of proteins and 50 μmol/L of hematoxylin with a UV-visible recording spectrophotometer (Shimadzu Corp).
gp47-phox Ventricular Expressions
Ventricular tissue was crushed in liquid nitrogen. Lysis and Western blots were performed as described previously6 with 35 μg of proteins loaded on gels (antibodies from Santa Cruz Biotechnology, gp47-phox rabbit [H-195] sc-14015).
Adult Cardiac Fibroblasts Primary Culture
Ten- to 12-week-old mouse left ventricle fibroblasts were cultured as described previously5 and transferred to serum-free medium before a 24-hour pharmacological stimulation. Cells were treated either with serum-free culture medium only (controls) or added to Ang II (10−7 mol/L) or ISO (10−5 mol/L) alone or simultaneously with the 5-HT2BR antagonist, SB215505 (10−7 mol/L). After treatments, cells were washed in oxygenated Krebs-Hepes buffer. The O2·− production was measured with the lucigenin method and adjusted by the protein concentration (counts per minute per microgram) of the samples.
Data Analysis and Statistics
Data are expressed as means±SEMs. Statistical comparisons were made by ANOVA followed by the Bonferroni’s method with the GraphPad Prism program (GraphPad Software). A value of P<0.05 was considered significant.
Effect of the 5-HT2BR Antagonist, SB215505, on Chronic Ang II–Induced Cardiovascular Alterations
The Ang II infusion induced cardiac hypertrophy as assessed by echocardiographic measurement of the left ventricle mass: body weight ratio (+52% over controls; Figure 1A) and direct measurement of the heart weight:body weight ratio (+17% versus controls; Table). A simultaneous 37-mm Hg increase in blood pressure (Figure 1B) and no significant heart rate change (595±14 bpm in Ang II versus 605±22 bpm in controls; P>0.05) were observed. In response to the afterload increase, the fractional shortening was also increased (Table). SB215505 reduced the Ang II–induced left ventricular hypertrophy (Table and Figure 1A) without affecting the increased blood pressure that remained elevated by 40 mm Hg over controls (Figure 1B). The heart rate was unaffected by the SB215505 treatment (584±17 bpm; P>0.05 versus controls), and the fractional shortening was still increased compared with controls (Table).
Effect of the 5-HT2BR Antagonist, SB215505, on Chronic ISO-Induced Cardiovascular Alterations
ISO induced a cardiac hypertrophy measured by echocardiography (+117% over controls for left ventricle mass:body weight ratio; Figure 1A) and direct measurement of the heart weight:body weight ratio (+31%; Table). This hypertrophy was associated with left ventricular dilatation, as shown by the increase in end-diastolic diameter (+16%; Table and Figure 2) and tachycardia (+17%; P<0.05; Figures 1C and 2⇓). No significant changes in blood pressure (115±7 mm Hg in controls versus 114±6 mm Hg in ISO; P>0.05) and cardiac contractility (fractional shortening) were detected. The cardiac output was preserved (Table). To confirm that these effects were mediated by β-adrenergic receptors, mice were simultaneously treated by ISO and the nonselective β-adrenergic antagonist propranolol. This compound reduced cardiac hypertrophy (Table) and tachycardia induced by ISO (744±38 bpm in ISO versus 619±24 bpm in ISO+propranolol; P<0.05) but also slightly reduced the cardiac output. SB215505 prevented the left ventricular hypertrophy caused by ISO (Figure 1A and Table) and the cardiac dilatation, the end-diastolic diameter (end-diastolic diameter) being similar to controls (Table and Figure 2). This prevention of cardiac hypertrophy was obtained without cardiodepression (Table) or effect on the blood pressure (119±6 mm Hg in ISO+SB215505 versus 115±7 mm Hg in controls; P>0.05). We also demonstrated that the cardiac alterations induced by ISO were dependent on NAD(P)H oxidase activation, because the NAD(P)H oxidase inhibitor, apocynin, prevented ISO-induced hypertrophy (Table). Experiments on Nox2−/− mice showed that the NAD(P)H oxidases involved did not include the Nox2 subunit (please see the data supplement).
Involvement of the 5-HT2BR and NAD(P)H Oxidases on Cardiac O2·− Production in Ang II/ISO-Induced Cardiovascular Alterations
Ang II increased the basal O2·− generation, as well as the NADPH-stimulated NAD(P)H oxidase activity in the left ventricle (Figure 3A and 3B). SB215505 completely normalized O2·− concentrations in the left ventricle. Ang II induced an increase of the p47phox subunit expression, which was not affected by the simultaneous SB215505 treatment (Figure 4). In the aorta, Ang II also increased basal (623±54 cpm · mg−1 in controls to 1145±88 cpm · mg−1 in Ang II; P<0.05) and NAD(P)H oxidase-mediated (11 8076±23 709 cpm · mg−1 in controls to 146 749±10 559 cpm · mg−1; P<0.05) O2·− concentration, but SB215505 was without effect on this Ang II response (820±107 cpm · mg−1 in basal and 181 449±19 596 cpm · mg−1 after NAD(P)H; P>0.05 versus Ang II alone). To verify that the SB215505 effects were not related to a nonspecific antioxidant property, we showed in an in vitro based assay that, in contrast to glutathione and ascorbic acid, it was unable to reduce the level of spontaneous hematoxylin oxidation (please see the data supplement). Moreover, mice subcutaneously infused with 1 mg.kg−1 SB215505 alone (5 or 14 days), did not demonstrate any decrease in basal or NAD(P)H-stimulated O2·− ventricular concentration (please see the data supplement). In conclusion, the pharmacological blockade of 5-HT2BRs abolished the NAD(P)H oxidase overactivation and, therefore, the O2·− generation induced by Ang II in the left ventricle.
In contrast to Ang II, the chronic ISO infusion did not modify the basal concentration of O2·− in the left ventricle (Figure 3A). However, when the maximal activity of NAD(P)H oxidases was tested by the addition of a saturating NAD(P)H concentration (100 μmol/L), we observed a 48% increase in the left ventricular O2·− production as compared with control animals (Figure 3B). SB215505 prevented the ISO-induced NAD(P)H oxidase activation without significant change in the basal O2·− concentration after ISO (Figure 3). To analyze whether the increase in left-ventricular NAD(P)H oxidase activity was because of a change in NAD(P)H complex expression, the p47phox subunit expression was assessed by Western blots. ISO induced a 20% (P<0.05) increase in left ventricular p47phox subunit expression (Figure 4). Similar to Ang II+SB215505-treated mice, this overexpression was not affected by SB215505 (Figure 4). Taken together, these data show that Ang II and ISO induced an increase in left ventricular O2·− generation in association with an increase in the p47phox NAD(P)H oxidase subunit expression.
Effect of SB215505 on SOD Activity After Ang II and ISO Stimulations
To explain the difference between ISO and Ang II in basal O2·−, we postulated a different counterregulatory mechanism by SOD: the enzyme that triggers the dismutation of O2·− to oxygen peroxide and is therefore involved in the O2·− clearance. ISO induced a 54% increase in the activity of SOD (Figure 5) that was prevented by the simultaneous administration of SB215505, indicating that the reducing effect of this drug on left ventricular O2·− concentration is rather attributable to a regulation of NAD(P)H oxidase activity than to an increased SOD activity. In contrast to ISO, the SOD activity was not significantly modified by Ang II. Therefore, the increased NAD(P)H oxidase–mediated O2·− concentration by Ang II was not limited by a simultaneous augmentation of SOD activity, as observed during ISO stimulation. Similarly, when SB215505 was simultaneously administered with Ang II, the SOD activity was not affected.
Effect of 5-HT2BR Blockade on O2·− Production and NAD(P)H Oxidase Activity in Primary Culture of Left Ventricular Fibroblasts
We analyzed the role of the 5-HT2BR on NAD(P)H oxidase activity after a 24-hour stimulation with either Ang II or ISO in left-ventricular fibroblasts. Ang II and ISO did not affect the basal O2·− production (data not shown) but induced, respectively, a 34% and a 42% increase in NAD(P)H oxidase–mediated O2·− concentration compared with controls (Figure 6). These increases were completely prevented by the simultaneous SB215505 treatment, indicating that 5-HT2BRs can regulate the NAD(P)H oxidase activity induced by these agonists in cardiac fibroblasts (Figure 6).
5-HT2BR Blockade Prevents O2·−-Mediated Ang II–Induced Cardiac Hypertrophy
Ang II is recognized as a cardiac hypertrophic factor in vitro, as well as in vivo,8 and gp91phox/Nox2-mediated ROS production in cardiomyocytes contributes to this response.9 We tested whether the blockade of 5-HT2BRs could reduce Ang II-induced hypertrophy and cardiac ROS generation in mice. Ang II induced a left ventricular hypertrophy associated with an increase in O2·− production but no increase of SOD-mediated protection. The blockade of 5-HT2BRs by SB215505 completely prevented the increased left ventricular O2·− generation in basal and stimulated conditions. The reduction of the NAD(P)H oxidase activity by 5-HT2BR blockade does not take place at the expression level, because p47phox-subunit was still overexpressed. This drug was selected because of its higher selectivity (×10) for 5-HT2BRs compared with 5-HT2C.10 Moreover, this drug is a potent antagonist, and we have shown previously5 that only 5-HT2A and 5-HT2B receptors are expressed in the heart. Interestingly, this antihypertrophic action was achieved without reduction of arterial blood pressure, attesting to the load-independent cardiac antihypertrophic effect of SB215505. In the aorta, Ang II also increased basal and NAD(P)H oxidase–mediated O2·− concentration, which was insensitive to SB215505, supporting the notion that Ang II type 1 and 5-HT2BRs are not interacting in aorta. Also, SB215505 neither affected the basal O2·− concentration nor the NAD(P)H oxidase maximal activity in control animals (please see the data supplement), and experiments using hematoxylin autooxidation did not reveal any direct chemical antioxidative properties of the compound. Therefore, SB215505 is not a nonspecific antioxidant and does not directly inhibit the NAD(P)H oxidase complex. These data demonstrate that the 5-HT2BR is a key regulator of Ang II–mediated cardiac hypertrophy and NAD(P)H oxidase activity.
Concerns were raised regarding the capacity of lucigenin to self-generate O2·− through redox cycling. This problem was addressed by using a low concentration of lucigenin (5 μmol/L) as reported by Skatchkov et al11 and confirmed by Spin-trapping studies. No variation in luminescence was detected when comparing a blank vial filled with a lucigenin solution with or without 100 μmol/L of NAD(P)H.
5-HT2BR Blockade Prevents O2.−-Mediated ISO-Induced Cardiac Hypertrophy
We show that the subchronic β-adrenergic activation induces a geometric remodeling of the left ventricle as attested by dilatation. This alteration is completely prevented by the 5-HT2BR antagonist without major hemodynamic adverse effects, ie, no reduction of blood pressure or cardiac contractility. In the present work, a slight reduction of the ISO-induced tachycardia (−9.8%) was observed. This reduction was probably not a consequence of a pharmacological effect of this compound on the sinus node, because this drug never reduced the heart rate in any other group but could rather be associated with the prevention of cardiac remodeling.
The role of ventricular oxidative stress after β-adrenoceptor stimulation is still a matter of debate. A recent study has shown that the initial steps of the cardiac hypertrophy after ISO infusion involve O2·− production,12 but the cellular origin of O2·− is unclear. An in vitro study suggested that the stimulation of β-adrenergic receptors located on cardiomyocytes induces hypertrophy, which is not mediated by increased ROS production.13 In the present work, the NAD(P)H oxidase selective inhibitor, apocynin, prevented the ISO-induced cardiac hypertrophy, suggesting that NAD(P)H oxidase–dependent oxidative stress is involved in this in vivo model. β-Adrenergic receptor stimulation is likely involved, because the nonselective β-blocker propranolol had a similar effect to apocynin. Moreover, we observed an increase of NAD(P)H oxidase maximal activity after ISO infusion. Rathore et al14 have shown that the increased cardiac oxidative stress induced by ISO was counteracted by a simultaneous increase in SOD activity in rats. We reproduced the same result in mice. The physiological relevance of this activation is not yet understood but, 2 aspects could be considered: a deviation of O2·− to hydrogen peroxide, the last being involved in hypertrophy, or a compensatory mechanism to O2·− increase that would not be sufficient to prevent cardiac hypertrophy. These questions were addressed recently in the study by Cabassi et al15 performed on prehypertensive spontaneously hypertensive rats. These animals exhibit an overactivity of the sympathetic nervous system together with increased oxidative stress status and cardiomyocyte hypertrophy. In this model, the SOD mimetic, hydroxytetramethyl piperidinoxyl, was unable to prevent cardiac hypertrophy, indicating that “pushing” O2·− to hydrogen peroxide neither prevented nor amplified the hypertrophic phenotype. Therefore, in our experimental conditions, the SOD overactivation is probably not sufficient to suppress enough O2·− production and prevent ventricular hypertrophy. After treatment with a 5-HT2BR antagonist, the SOD activity returned to control values, showing that the reduction of O2·− concentrations was not because of an increased rate of degradation but rather because of a reduction in the production of O2·−.
The 5-HT2BR Regulates NAD(P)H Oxidase Activation in Cardiac Fibroblasts
In mice knockout for the Nox2 subtype of the gp91phox catalytic subunit of the NAD(P)H oxidase complex (Nox2−/−), ISO induced cardiac hypertrophic responses that were prevented by a treatment with the NAD(P)H oxidase inhibitor apocynin (please see http://hyper.ahajournals.org). In cardiac myocytes, the gp91phox/Nox2 is the prominent isoform, whereas Nox1 and Nox4 are expressed at lower levels.12 In Nox2−/− mice, stimulation with ISO induces Nox1 overexpression (please see the data supplement). Other Nox isoforms are also known to be expressed in human cardiac fibroblasts, whereas Nox2 is barely detectable.16 Therefore, considering the following: (1) apocynin prevents cardiac hypertrophy induced by ISO infusion in Nox2−/− mice; (2) non-Nox2/gp91phox homologues are expressed by fibroblasts; (3) the suppression of Nox2, mainly expressed in cardiomyocytes, cannot suppress ISO-mediated cardiac hypertrophy; and (4) the stimulation of β-adrenergic receptors located on cardiomyocytes induces a ROS-independent hypertrophy,17 it is likely that noncardiomyocytes play an important role in the cardiac hypertrophic responses to β-adrenergic activation.
Our previous study5 indicated that fibroblasts constitute a target of hypertrophic responses to β-adrenergic activation, because 5-HT2BR blockers diminished the release of hypertrophic cytokines. We tested the effect of 5-HT2BR blockade in O2·− production by cardiac fibroblasts during 24-hour stimulations with Ang II or ISO. Ang II and ISO increased the maximal NAD(P)H oxidase–mediated O2·− production, and this effect was prevented by SB215505. This result indicates a role of this receptor in ISO and Ang II-induced NAD(P)H oxidase activation in these cells. This effect was obtained without stimulation of the receptor by its natural agonist, because in our experimental conditions the serotonin concentration in cell culture medium was found <1 nM at the end of the experiments. Therefore, SB215505, through its interaction with the 5-HT2BR, seems to interfere with a crosstalk among β-adrenergic, 5-HT2B, and Ang II receptors.
The involvement of 5-HT2BRs in the development of cardiac hypertrophy is a new finding with potential value for the treatment or prevention of the disease. In the present study, the blockade of this receptor can prevent the increase of NAD(P)H oxidase activity and the development of cardiac hypertrophy induced by Ang II type 1 or β-adrenergic receptors, but several questions remain to be answered. Little is known about the basal activity of the 5-HT2BR and the possible agonists involved in its stimulation (serotonin) or regulation in the context of the development of cardiac hypertrophy. Also, the signaling factors responsible for the interactions among Ang II type 1, β-adrenergic, and 5-HT2BRs must still be identified. A major impact of this work was to demonstrate for the first time that the Gq-coupled 5-HT2BR blockade can modify both Ang II type 1 and β-adrenergic oxidant and hypertrophic responses. The mechanism that triggers this cross-regulation will have to be investigated.
We thank Diane Papin and Jo-Anne Le Guerrier for helpful technical support.
Sources of Funding
This research was supported by the Canadian Institute for Health Research (Ottawa, Canada), Université Pierre et Marie Curie (Paris, France), Université Louis Pasteur (Strasbourg, France), Fondation de France (France), Fondation pour la Recherche Médicale (Paris, France), Association pour la Recherche contre le Cancer (Villejuif, France), and Agence Nationale pour la Recherche (France). L.M.’s team is an Equipe Fondation pour la Recherche Médicale (Paris, France).
The first 2 authors contributed equally to this work.
- Received November 21, 2007.
- Revision received December 10, 2007.
- Accepted June 3, 2008.
Harrison DG, Cai H, Landmesser U, Griendling KK. Interactions of angiotensin II with NAD(P)H oxidase, oxidant stress and cardiovascular disease. J Renin Angiotensin Aldosterone Syst. 2003; 4: 51–61.
Pietri M, Schneider B, Mouillet-Richard S, Ermonval M, Mutel V, Launay JM, Kellermann O. Reactive oxygen species-dependent TNF-alpha converting enzyme activation through stimulation of 5-HT2B and alpha1D autoreceptors in neuronal cells. FASEB J. 2005; 19: 1078–1087.
Jaffre F, Callebert J, Sarre A, Etienne N, Nebigil CG, Launay JM, Maroteaux L, Monassier L. Involvement of the serotonin 5-HT2B receptor in cardiac hypertrophy linked to sympathetic stimulation: control of interleukin-6, interleukin-1beta, and tumor necrosis factor-alpha cytokine production by ventricular fibroblasts. Circulation. 2004; 110: 969–974.
Bendall JK, Cave AC, Heymes C, Gall N, Shah AM. Pivotal role of a gp91(phox)-containing NADPH oxidase in angiotensin II-induced cardiac hypertrophy in mice. Circulation. 2002; 105: 293–296.
Hingtgen SD, Tian X, Yang J, Dunlay SM, Peek AS, Wu Y, Sharma RV, Engelhardt JF, Davisson RL. Nox2-containing NADPH oxidase and Akt activation play a key role in angiotensin II-induced cardiomyocyte hypertrophy. Physiol Genomics. 2006; 26: 180–191.
Cussac D, Newman-Tancredi A, Quentric Y, Carpentier N, Poissonnet G, Parmentier JG, Goldstein S, Millan MJ. Characterization of phospholipase C activity at h5-HT2C compared with h5-HT2B receptors: influence of novel ligands upon membrane-bound levels of [3H]phosphatidylinositols. Naunyn Schmiedebergs Arch Pharmacol. 2002; 365: 242–252.
Zhang GX, Kimura S, Nishiyama A, Shokoji T, Rahman M, Yao L, Nagai Y, Fujisawa Y, Miyatake A, Abe Y. Cardiac oxidative stress in acute and chronic isoproterenol-infused rats. Cardiovasc Res. 2005; 65: 230–238.
Cabassi A, Dancelli S, Pattoneri P, Tirabassi G, Quartieri F, Moschini L, Cavazzini S, Maestri R, Lagrasta C, Graiani G, Corradi D, Parenti E, Tedeschi S, Cremaschi E, Coghi P, Vinci S, Fiaccadori E, Borghetti A. Characterization of myocardial hypertrophy in prehypertensive spontaneously hypertensive rats: interaction between adrenergic and nitrosative pathways. J Hypertens. 2007; 25: 1719–1730.
Cucoranu I, Clempus R, Dikalova A, Phelan PJ, Ariyan S, Dikalov S, Sorescu D. NAD(P)H oxidase 4 mediates transforming growth factor-beta1-induced differentiation of cardiac fibroblasts into myofibroblasts. Circ Res. 2005; 97: 900–907.