Histone deacetylase inhibition by Entinostat for the prevention of electrical and structural remodeling in heart failure

Ventricular tachyarrhythmias are an important predictor of morbidity and mortality in patients with chronic heart failure [1]. Moreover, sustained tachyarrhythmia can lead to ventricular remodeling impairing left ventricular ejection fraction thereby causing a tachycardiomyopathy. Despite a significant progress in pharmacological and interventional antiarrhythmic therapies, treatment options for patients suffering from heart failure are still limited. The development of heart failure is accompanied by complex changes in cardiac electrophysiology and functional properties of cardiomyocytes and fibroblasts, most of which are still poorly characterized. Given the multitude of cell types and signaling cascades that are involved in the process of ventricular remodeling, effective therapeutic strategies would ideally simultaneously impact several different remodeling pathways in the diverse cell types involved [2]. Histone deacetylases (HDACs) may represent such targets. HDACs play a central role in epigenetic regulation of transcription and cell proliferation [3, 4] and in the context of a whole host of key biological processes, including pathological remodeling of the heart [5]. Reversible histone acetylation provides a central mechanism for controlling gene expression and cellular signaling events in multiple cells types, e.g., myocytes, fibroblasts and immune cells [6, 7]. Over the last decade, gene expression and deletion studies have revealed important functions of HDAC in adaptive responses of myocardial tissue to different types of injury, such as hypertrophy, ischemia, cell death, remodeling and fibrosis [5, 8]. Moreover, the development of specific small molecule inhibitors of HDACs allows for distinct studies of pharmacological interventions of HDAC function. HDAC inhibitors with a longer biological half time such as the class I HDAC inhibitor Entinostat (MS-275) are currently being used in clinical studies in patients with several types of hematological malignancies [9, 10] and may represent a significant therapeutic advance.

Recent evidence implies that several HDAC inhibitors could also be valuable tools to blunt pathological cardiac remodeling in the settings of pressure overload and ischemia/reperfusion [11–13]. In a recent study, Entinostat was shown to suppress hypoxia-dependent pulmonary hypertension and RV hypertrophy [14]. Here, we describe electrophysiologic effects of Entinostat in vivo in an experimental model of heart failure. However, due to the inherent biological differences between mouse and human hearts, study results in mice may have only a limited relevance to humans. Instead, the rabbit heart shares some physiological commonalities with the human heart, e.g. a similar ion channel expression pattern and similarities in myocardial action potential [15], and has therefore been successfully used for electrophysiological studies [16]. Using an in vivo model of heart failure, we studied the ability of the small molecule class I HDAC inhibitor Entinostat to prevent proarrhythmic effects of heart failure and structural remodeling associated with the development of heart failure.

The current study provides insight into the pharmacological effects of the class I HDAC inhibitor Entinostat on electrical and structural remodeling in heart failure. Class I HDACs are considered to mediate maladaptive pathways in response to cardiomyocyte stress and are assumed to have a negative impact on cardiac function. Accordingly, HDAC inhibitors have been shown to diminish or even reverse the cardiac remodeling process in several models of cardiomyopathy. In vivo, Entinostat treatment could prevent the development of cardiac fibrosis in response to rapid ventricular stimulation.

Rapid ventricular stimulation caused a significant elevation of monophasic action potential duration compared to sham hearts while treatment with Entinostat could partially abolish this effect. In failing hearts, Entinostat could partially restore the reduction of Cav1.3 and KCNH2 expression.

While our model of pacing induced heart failure did not result in the development of marked cardiac hypertrophy, different HDAC inhibitors have been shown to be able to reduce cardiac hypertrophy. Specifically, the pan HDAC inhibitor TSA blunted cardiac hypertrophy and improved systolic function in a pressure-overload transverse aortic constriction model in mice [7] and TSA was even capable of reversing pre-existing hypertrophy [19]. In a rat infarct model, the HDAC inhibitors valproic acid (VPA) reduced cardiomyocyte hypertrophy and collagen deposition of the infarcted left ventricle and preserved systolic function [20]. However, a preservation of cardiac systolic function could not be demonstrated in our model. This observation can possibly be explained by the specific conditions of rapid ventricular pacing that subjects the heart to a different type of stress compared to other models of heart failure.

In fact, pressure overload induced heart failure is characterized by an adaptive increase in the ratio of LV wall thickness to chamber radius thereby partly compensating the increase in afterload against which the myocardium shortens during systole. The resulting concentric hypertrophy is associated with profound changes in gene expression pattern resembling that of fetal hearts [21]. Histone deacetylase inhibitors have shown to be able to reverse some of these gene expression changes suggesting that they may have a therapeutic potential in cardiac hypertrophy and heart failure. Interestingly, KCNH2 and Cav1.3 protein expression levels could be partially restored by Entinostat in failing, stimulated hearts. On the contrary, heart failure in our model is caused by tachycardia in combination with asynchronic LV contraction. In this model, rapid pacing typically leads to LV chamber dilatation and reduced wall thickness. Thus, the different pathomechanisms leading to heart failure in both models might explain why Entinostat did not efficiently preserve LV function during rapid pacing despite its antiproliferative and antifibrotic effects.

In our model, long term rapid ventricular pacing of rabbit hearts caused a significant increase of MAP duration, more EAD and more non sustained polymorphic ventricular tachycardias compared to sham hearts [16]. Classically, these effects have been attributed to a down-regulation of potassium channels, resulting in a statistically noticeable prolongation of the QT-intervals and in extended monophasic action potentials as a consequence of a prolonged repolarization process [22]. Like in the context of heart failure, an excessive prolongation of action potential duration leading to EAD [23] and an increase of spatial dispersion of repolarization can act as a substrate for the occurrence of proarrhythmia [24]. Here, we show that treatment with Entinostat could partially prevent these effects by prevention of MAP prolongation observed in failing hearts. Consistently, Entinostat was able to partially restore KCNH2 ion channel expression downregulated in rabbit hearts with pacing induced heart failure. Thus, Entinostat likely contributes to a reduction of spatial dispersion of repolarization and causes a stabilization of repolarization and prevention of ventricular tachyarrhythmias. Correspondingly, EAD were reduced by half after administration of Entinostat in failing hearts, and that may account for the lack of proarrhythmia in failing hearts treated with Entinostat. However, in non-paced, structurally normal hearts, Entinostat paradoxically displayed a weak tendency to prolong QT-interval in association with more EADs and more non-sustained ventricular tachyarrhythmias. Therefore, the pharmaceutical effect of Entinostat in healthy hearts with regard to QT interval duration and potassium channel activity seems to be complex. In clinical trials, the HDAC inhibitors Vorinostat and Romidepsin have been demonstrated to cause a prolongation of the QT-interval [25]. In patients with metastatic neuroendocrine tumors, the HDAC inhibitor Depsipeptide FR901228 has been associated with QT-interval prolongation and ventricular tachycardias [26]. In a model of ischemia, however, Entinostat was able to confer protection from myocardial damage by ischemia/reperfusion. Interestingly, this effect was dependent on mitochondridal HDAC1 [27].

Myocardial potassium channel expression is subject of a complex regulation on the transcriptional, post-transcriptional, post-translational and epigenetic level [28]. Potassium channels play an important role in action potential duration and shape. The potassium channels Kv1.4, Kv1.5 and KCNH2 are important molecular components of I_to, I_Kur and I_Kr, respectively [29]. Since epigenetic mechanisms determining cardiac potassium and calcium channel activity have not yet been well characterized, one might speculate that the effects of class I HDAC inhibition by Entinostat on potassium and calcium channel activity critically depend on transcription factors and key regulators of potassium channel expression and activity that are induced or repressed as a result of the development of cardiac failure induced by rapid ventricular pacing. Data on ERG expression in failing hearts has been inconsistent, few studies report a reduction of ERG expression in heart failure [30]. Li et al. did not detect significant differences in mRNA ERG expression in the left ventricles among rabbits with heart failure [31]. In our study, the downregulation of the ERG ion channel KCNH2 in failing rabbit hearts might have contributed to the observed prolongation of MAPs. In contrast to our results, the expression of I_to subunit Kv1.4 mRNA was significantly down-regulated in rabbits with heart failure [31]. Changes in Kv1.5 expression have been variable related to differences in underlying heart disease [32]. In fact, a different expression of selected potassium channel species (Kv1.4 and Kv1.5) could not be confirmed in our study (Fig. 3). This finding is probably related to different stimulation frequency and duration. In addition, other mechanisms including posttranslational channel protein modification may play a role. Hence, further studies will be necessary to clarify these issues. Interestingly, HDAC inhibitors showed several cardiac side effects including atrial fibrillation and QT prolongation in different clinical trials.

Finally, histological analysis of tissue specimen of rabbit hearts revealed that Entinostat could prevent the development of cardiac fibrosis in response to rapid ventricular stimulation. Several studies in animal models imply that HDAC inhibition have beneficial effects on disease states associated with pathological fibrosis. Class I HDAC inhibition retards fibrosis formation in acute myocardial infarction settings and suppresses angiotensin II-mediated cardiac fibrosis [33]. In general, fibrosis has significant consequences for cardiac function, since progressive increases in extracellular matrix synthesis and deposition result in increased mechanical stiffness and contribute to systolic dysfunction. Depending on the stage of heart failure, myofibroblasts are highly localized to sites of injury where synthesis and deposition of collagen promotes scar formation and fibrosis [34]. Specifically, progressive cardiac fibrosis fosters the development of malignant arrhythmias based on local reentry. Consequently, the prevention of cardiac fibrosis might therefore additionally protect the heart from the occurrence of arrhythmias. Thus, Entinostat seems to be a promising candidate for the prevention of electrical and structural remodeling in heart failure.