Identification of Highly Potent Protein Kinase C‐Related Kinase 1 Inhibitors by Virtual Screening, Binding Free Energy Rescoring, and in vitro Testing

Abstract

Despite the considerable interest in protein kinase C-related kinase 1 (PRK1) as a target in cancer research, there is still a lack of PRK1 inhibitors with suitable selectivity profiles and physicochemical properties. To identify new PRK1 inhibitors we applied a virtual screening approach, which combines ensemble docking, minimization of the protein-ligand complex, binding free energy calculations, and application of quantitative structure-activity relationship (QSAR) models for predicting in vitro activity. The developed approach was then applied in a prospective manner to screen available libraries of kinase inhibitors from Selleck and GlaxoSmithKline (GSK). Compounds that showed favorable prediction were then tested in vitro for PRK1 inhibition. Some of the hits were found to inhibit PRK1 in the low-nanomolar range. Three in vitro hits were additionally tested in a mass-spectrometry-based cellular kinase profiling assay to examine selectivity. Our findings show that nanomolar and drug-like inhibitors can be identified by the virtual screening approach presented herein. The identified inhibitors are valuable tools for gaining a better understanding of PRK1 inhibition, and the identified hits can serve as starting points for further chemical optimization.

Visualizing protein–ligand binding with chemical energy-wise decomposition (CHEWD): application to ligand binding in the kallikrein-8 S1 Site

“…Calculation of binding energies as a screening tool is highlighted by numerous implementations in medicinal chemistry and drug design [6][7][8]59]. Decomposing the binding energy into individual residue contributions can provide useful insight into the nature of the protein-ligand complex.…”

“…Structure-based drug design (SBDD) has benefited significantly over the years from developments in the field of proteomics, comparative genomics and high-end computing [1][2][3][4][5]. Binding affinity is an important metric in the assessment of new drug targets and their optimization [5][6][7][8]. The ability to calculate binding affinities reliably in silico is advantageous, reducing the need for expensive experiments [9].…”

Application of Computer Modeling to Drug Discovery: Case Study of PRK1 Kinase Inhibitors as Potential Drugs in Prostate Cancer Treatment

“…In spite of the importance of PRK1 in the targeted therapy of cancer, only a few known inhibitors have been identified. The known PKC inhibitors (Figures 1 and 2) include staurosporine (PubChem CID: 44259) and its analogue (Ro-318220; PubChem CID: 5083), bisindolylmaleimide I (BIM I; PubChem CID: 2396), and lestaurtinib (PubChem CID: 126565), as well as the nonselective Akt inhibitor GSK-690693 (PubChem CID: 16725726) and Pfizer's JAK nonselective inhibitor CP-690550 (tofacitinib; PubChem CID: 9926791) [17,18]. In a previous work, several novel PRK1 inhibitors were identified containing different scaffolds, using a homology model, with varying potencies [19].…”

“…The data set DS1 includes 28 active (Figures 1 and 2 [18]) and 300 inactive compounds. First, the active compounds were docked toward the ensemble of the three PRK1 structures using the previously described docking methods.…”

“…It was observed, from docking, that the isoquinoline derivatives are able to form an H bond between the N atom of the isoquinoline motif and the NH backbone of Ser704 located in the hinge region. The isoquinoline motif occupied the adenine binding site; meanwhile, the substituents were located in the sugar-binding site, which is surrounded by several hydrophilic residues (e.g., Asp708, Asp750 Comparison of experimental pIC50 versus predicted pIC50 values calculated by QSAR_3 for the 26 training set molecules [18].…”