Background

Multiple stress signals and primary cardiomyopathies induce the activation of cardiac Ca^{2+/Calmodulin-dependent kinase (CaMKII), thereby promoting heart failure and arrhythmia. In contrast, class IIa histone deacetylases (HDACs) are known to be cardio-protective. CaMKII physically interacts with HDAC4 to inactivate class IIa HDACs, thus mediating disease-causing transcriptional processes in cardiomyocytes. We aim to specifically disrupt this signaling pathway with CaMKII-HDAC4 protein-protein interaction inhibitors (CHPPIIs) to selectively blunt the responsiveness of HDAC4 to CaMKII without affecting the essential physiological functions of these two enzymes. We hypothesize that this target-specific drug approach has a high therapeutic potential with less on-target side effects than upstream receptor blockade or enzymatic CaMKII inhibition.}

Objective

The aim of this study was (i) to extensively validate the target in vivo via a genetic drug-mimicking mouse model and (ii) to develop small molecule CHPPIIs that effectively and specifically disrupt the CaMKII-HDAC4 interaction

Methods and Results

To conduct in vivo proof-of-principle studies, we engineered CaMKII-resistant HDAC4 knock-in (KI) mice, in which one amino acid (Arg-598) of HDAC4 is mutated to Phe. Subjecting these KI mice to models of acquired heart disease (transverse aortic constriction, TAC) as well as crossing to models of genetic heart disease (RBM20 cardiomyopathy) we observed marked protection from adverse cardiac remodeling and contractile dysfunction (ejection fraction after TAC 52% in KI vs. 28% in WT; and 43% in KI/RBM20 KO vs. 34% in RBM20 KO). Unbiased multi-organ in vivo phenotyping revealed that the HDAC4 KI mice (in contrast to mice lacking CaMKII globally) develop no adverse effects in other organ systems, indicating that specifically inhibiting the CaMKII-HDAC4 interaction is a safe therapeutic approach.

To identify CHPPII hit compounds we developed a high-throughput homogeneous time-resolved fluorescence (HTRF)-based assay to screen a compound library of 560k small molecules. For extensive compound profiling we established multiple assays for analysis of PPII specificity, in vitro kinase activity, biophysical properties and cellular activity. Based on activity in the assay cascade and drug metabolism and pharmacokinetics (DMPK) properties, one chemical scaffold was prioritized for hit-to-lead development. First rounds of medicinal chemistry-based optimization enabled us to develop improved CHPPIIs with in vitro IC_{50<300nM and specific activity against the CaMKII-HDAC4 interaction, but no activity against global CaMKII enzymatic function in cardiomyocytes. The frontrunner compounds are currently undergoing proof-of-concept testing in engineered human myocardium (EHM) models generated from WT and RBM20 mutant human iPSC-derived cardiomyocytes and are further optimized for in vivo application in preclinical models.}

Conclusion

Our in vivo experiments on CHPPII-mimicking HDAC4 KI mice revealed that specific inhibition of the CaMKII-HDAC4 interaction is an efficient and safe strategy to treat heart disease. Via high-throughput small molecule screen, extensive compound profiling and medicinal chemistry-based optimization we developed compounds that effectively and specifically disrupt the CaMKII-HDAC4-binding in non-cellular and cellular assays, representing candidates for further optimization to allow clinical application of a new drug class.