Novel Development of Predictive Feature Fingerprints to Identify Chemistry-Based Features for the Effective Drug Design of SARS-CoV-2 Target Antagonists and Inhibitors Using Machine Learning.

A unique approach to bioactivity and chemical data curation coupled with random forest analyses has led to a series of target-specific and cross-validated predictive feature fingerprints (PFF) that have high predictability across multiple therapeutic targets and disease stages involved in the severe acute respiratory syndrome due to coronavirus 2 (SARS-CoV-2)-induced COVID-19 pandemic, which include plasma kallikrein, human immunodeficiency virus (HIV)-protease, nonstructural protein (NSP)5, NSP12, Janus kinase (JAK) family, and AT-1. The approach was highly accurate in determining the matched target for the different compound sets and suggests that the models could be used for virtual screening of target-specific compound libraries. The curation-modeling process was successfully applied to a SARS-CoV-2 phenotypic screen and could be used for predictive bioactivity estimation and prioritization for clinical trial selection; virtual screening of drug libraries for the repurposing of drug molecules; and analysis and direction of proprietary data sets.

Clustering, Spatial Distribution, and Phosphorylation of Discoidin Domain Receptors 1 and 2 in Response to Soluble Collagen I.

Discoidin domain receptors (DDR1 and DDR2) are receptor tyrosine kinases that signal in response to collagen. We had previously shown that collagen binding leads to clustering of DDR1b, a process partly mediated by its extracellular domain (ECD). In this study, we investigated (i) the impact of the oligomeric state of DDR2 ECD on collagen binding and fibrillogenesis, (ii) the effect of collagen binding on DDR2 clustering, and (iii) the spatial distribution and phosphorylation status of DDR1b and DDR2 after collagen stimulation. Studies were conducted using purified recombinant DDR2 ECD proteins in monomeric, dimeric or oligomeric state, and MC3T3-E1 cells expressing full-length DDR2-GFP or DDR1b-YFP. We show that the oligomeric form of DDR2 ECD displayed enhanced binding to collagen and inhibition of fibrillogenesis. Using atomic force and fluorescence microscopy, we demonstrate that unlike DDR1b, DDR2 ECD and DDR2-GFP do not undergo collagen-induced receptor clustering. However, after prolonged collagen stimulation, both DDR1b-YFP and DDR2-GFP formed filamentous structures consistent with spatial re-distribution of DDRs in cells. Immunocytochemistry revealed that while DDR1b clusters co-localized with non-fibrillar collagen, DDR1b/DDR2 filamentous structures associated with collagen fibrils. Antibodies against a tyrosine phosphorylation site in the intracellular juxtamembrane region of DDR1b displayed positive signals in both DDR1b clusters and filamentous structures. However, only the filamentous structures of both DDR1b and DDR2 co-localized with antibodies directed against tyrosine phosphorylation sites within the receptor kinase domain. Our results uncover key differences and similarities in the clustering abilities and spatial distribution of DDR1b and DDR2 and their impact on receptor phosphorylation.