Integrated Quantitative Phosphoproteomics and Cell-based Functional Screening Reveals Specific Pathological Cardiac...

Cardiac hypertrophic signaling cascades resulting in heart failure diseases are mediated by protein phosphorylation. Recent developments in mass spectrometry-based phosphoproteomics have led to the identification of thousands of differentially phosphorylated proteins and their phosphorylation sites. However, functional studies of these differentially phosphorylated proteins have not been conducted in a large-scale or high-throughput manner due to a lack of methods capable of revealing the functional relevance of each phosphorylation site. In this study, an integrated approach combining quantitative phosphoproteomics and cell-based functional screening using phosphorylation competition peptides was developed. A pathological cardiac hypertrophy model, junctate-1 transgenic mice and control mice, were analyzed using label-free quantitative phosphoproteomics to identify differentially phosphorylated proteins and sites. A cell-based functional assay system measuring hypertrophic cell growth of neonatal rat ventricle cardiomyocytes (NRVMs) following phenylephrine treatment was applied, and changes in phosphorylation of individual differentially phosphorylated sites were induced by incorporation of phosphorylation competition peptides conjugated with cell-penetrating peptides. Cell-based functional screening against 18 selected phosphorylation sites identified three phosphorylation sites (Ser-98, Ser-179 of Ldb3, and Ser-1146 of palladin) displaying near-complete inhibition of cardiac hypertrophic growth of NRVMs. Changes in phosphorylation levels of Ser-98 and Ser-179 in Ldb3 were further confirmed in NRVMs and other pathological/physiological hypertrophy models, including transverse aortic constriction and swimming models, using site-specific phospho-antibodies. Our integrated approach can be used to identify functionally important phosphorylation sites among differentially phosphorylated sites, and unlike conventional approaches, it is easily applicable for large-scale and/or high-throughput analyses.