Development, Synthesis, and Characterization of G Protein-Coupled Receptor Kinase Inhibitors Using Structure Based Drug Design for the Advancement of Heart Failure Therapeutics

Abstract

In heart failure, the β-adrenergic receptors (βARs) become desensitized and uncoupled from heterotrimeric G proteins. This process is initiated by G protein-coupled receptor kinases (GRKs), some of which are upregulated in the failing heart, making them desirable therapeutic targets. Two structurally similar compounds were identified as GRK2 inhibitors. These compounds are the previously known Rho-associated coiled-coiled kinase inhibitor, GSK180736A and the selective serotonin reuptake inhibitor, paroxetine. Comparison of these two compounds crystallized in the active site with the previously crystallized Takeda101 and Takeda103A inhibitors, which are highly potent and selective, shows that paroxetine and GSK180736A do not extend out into a fourth subsite of the active site as do the Takeda compounds. We hypothesized that building off of our lead compounds into the hydrophobic subsite could improve their potency and selectivity. We initially made modifications to the scaffold of GSK180736A finding that building bulkier appendages into the subsite led to an increase in both potency and selectivity. We also identified a pan GRK inhibitor, 215022, which led to one of the first crystal structures of GRK5. Crystallization of GRK5 with 215022 allowed us to compare the two active sites, revealing that GRK2 has a larger hydrophobic subsite, whereas GRK5 has a narrower hydrophobic subsite, explaining the trend in potency and selectivity we had seen.

These structure activity relationships (SAR) were then investigated with paroxetine, which is a more drug-like scaffold that might result in compounds with higher therapeutic potential. Unfortunately our SAR was not transferable. Further exploration of the SAR of the paroxetine scaffold led to a series of paroxetine derivatives containing small heterocyclic appendages that were advantageous to GRK2 potency. Included in this series is a highly potent and selective GRK2 inhibitor, 258208, with an IC50 of 30 nM against GRK2 and greater than 230-fold selectivity over other GRKs and ROCK1. Crystallization of 258208 with Gβγ·GRK2 revealed an additional three hydrogen bonds were being made via an added amide linked pyrazole. Replacing the benzodioxole of 258208 with an indazole increased GRK2 potency to 8 nM, while maintaining selectivity over GRKs 5 and 1. Comparison of the GRK2-Gβγ crystal complexes of the indazole derivatives to the benzodioxole derivatives showed that the compounds make the same hydrogen bonds but the indazole derivatives form slightly tighter hydrogen bonds in the hinge leading to an overall more closed conformation of the kinase domain. Furthermore, 258208 showed a 100-fold improvement in cardiomyocyte contractility assays over paroxetine, a plasma concentration higher than its IC50 for over seven hours, and is currently under investigation in heart failure mouse models for its ability to exhibit cardioprotective effects.

In addition to development of potent GRK2 inhibitors we began a preliminary study to develop covalent GRK5 inhibitors. GRK5 contains a non-conserved cysteine (Cys474) on the active site tether that can be targeted by molecules binding to the active site of the kinase domain. Our initial compounds included saturated and unsaturated acrylamides so that we could identify if we were achieving covalent inhibition. We synthesized two regioisomers of the acrylamides off of a phenyl ring in the core scaffold. Of the two regioisomers synthesized one seems to be more favorable for general potency while the other gives evidence of binding covalently in the active site. Further analogs and biochemical testing are proposed to advance this series of covalent GRK5 inhibitors.