Requirements for cardioprotective GRK2 inhibition in vitro and in vivo

In the family of G protein-coupled receptor kinases (GRKs), GRK2 and GRK5 are the predominant members expressed in the heart. Both kinases are up-regulated in the failing heart. To date, a major pathophysiological role is established for GRK2 while the impact of GRK5 for heart failure pathogenesis is still under investigation. Although it is widely accepted that GRK2 inhibition is cardioprotective, mechanisms underlying cardioprotection are incompletely understood. Therefore, the major goal of my thesis was the identification of mechanisms required for cardioprotective GRK2-inhibition. In the first part of my thesis, I compared activities of different GRK2 inhibitors in vitro and in vivo. GRK2 was inhibited by three different approaches: (i) the dual-specific GRK2 and Raf/Erk axis inhibitor RKIP (raf kinase inhibitor protein), (ii) the small molecule paroxetine, which was previously established as a cardioprotective ATP-site GRK2 inhibitor by drug repurposing, and (iii) the dominant-negative GRK2-inhibitory mutant, GRK2-K220R. In order to compare substrate specificities of the GRK2 inhibitors, RKIP and paroxetine, I established an in vitro kinase assay. The test system used recombinant GRK2 purified from Sf9 (Spodoptera frugiperda) cells, which were infected with a recombinant GRK2-encoding baculovirus. The GRK2 substrates, phosducin and SRSF1 (serine/arginine-rich splicing factor 1), and the GRK2 inhibitor RKIP were purified from E. coli BL21-DE3 bacteria with T7 RNA polymerase-directed expression of recombinant proteins applying the pET expression system. The in vitro phosphorylation assay detected different modes of GRK2 inhibition by RKIP compared to paroxetine. RKIP only inhibited the GRK2-mediated phosphorylation of phosducin with an IC50 value of 950 nM whereas GRK2-mediated SRSF1 phosphorylation was not significantly decreased by RKIP up to 30 microM. In contrast, the ATP-site inhibitor paroxetine inhibited the phosphorylation of phosducin and SRSF1 with comparable IC50 values in the micromolar range. These different substrate specificities of the GRK2 inhibitors paroxetine and RKIP could be of major relevance in vivo, because paroxetine showed cardioprotection against myocardial infarction-induced heart failure whereas the in vivo function of RKIP is still under investigation. For the investigation of the in vivo role of RKIP, I used transgenic mice with myocardium-specific expression of RKIP under control of the alpha-MHC promoter, which were previously generated by our group. RKIP-expressing mice were compared with transgenic mice with myocardium-specific expression of the dominant-negative GRK2-K220R mutant. Our results showed that a moderately increased RKIP level of 2.8-3.2-fold over the non-transgenic control was sufficient to trigger severe signs of heart failure at an age of 8 months in the FVB and B6 background as evidenced by cardiac dysfunction, cardiac hypertrophy with dilation and cardiac lipid overload. RKIP led to activation of the adipogenic and heart failure-promoting transcription factor Pparg (peroxisome proliferator-activated receptor-gamma) by inhibition of the Raf-Erk axis-mediated serine-273 phosphorylation of Pparg. Whole genome microarray gene expression profiling showed up-regulation of heart failure-related Pparg target genes in RKIP-transgenic hearts with signs of heart failure. In contrast, GRK2 inhibition by transgenic expression of the dominant-negative GRK2-K220R mutant did not activate Pparg. In addition, GRK2-K220R retarded the development of chronic pressure overload-induced cardiac dysfunction imposed by abdominal aortic constriction. From these data it is concluded that an intact Raf-Erk axis is required for cardioprotective GRK2 inhibition. In the second part of my thesis, I investigated the role of GRK5, which could contribute to effects induced by GRK2 inhibition because GRK5 is reportedly up-regulated by GRK2 inhibition. However, the role of GRK5 is still under investigation. To analyse the in vivo role of GRK5, we generated GRK5-transgenic mice with a 2-fold upregulation of GRK5 in heart tissue, which reproduces the GRK5 up-regulation induced by GRK2 inhibition. The phenotype of 6-month-old Tg-GRK5 mice was determined. GRK5-transgenic mice appeared normal. Tg-GRK5 mice did not show any signs of cardiac hypertrophy, i.e. the heart weight to body weight ratio was not significantly different from non-transgenic controls. Histology analysis of hematoxylin-eosin-stained heart sections further documented that moderate GRK5 overexpression did not cause any cardiac abnormalities. Thus, my data show that moderate GRK5 expression had no adverse cardiac effect. With the newly established GRK5-transgenic model, future studies will have to investigate the role of GRK5 under pathophysiological conditions. In summary, my thesis with RKIP-transgenic mice found that an intact Raf-ERK axis is required for cardioprotective GRK2 inhibition. In addition, a new transgenic model was established, which allows to study the impact of compensatory GRK5 up-regulation under conditions of GRK2 inhibition