AI-guided pipeline for protein-protein interaction drug discovery identifies a SARS-CoV-2 inhibitor.

Protein-protein interactions (PPIs) offer great opportunities to expand the druggable proteome and therapeutically tackle various diseases, but remain challenging targets for drug discovery. Here, we provide a comprehensive pipeline that combines experimental and computational tools to identify and validate PPI targets and perform early-stage drug discovery. We have developed a machine learning approach that prioritizes interactions by analyzing quantitative data from binary PPI assays or AlphaFold-Multimer predictions. Using the quantitative assay LuTHy together with our machine learning algorithm, we identified high-confidence interactions among SARS-CoV-2 proteins for which we predicted three-dimensional structures using AlphaFold-Multimer. We employed VirtualFlow to target the contact interface of the NSP10-NSP16 SARS-CoV-2 methyltransferase complex by ultra-large virtual drug screening. Thereby, we identified a compound that binds to NSP10 and inhibits its interaction with NSP16, while also disrupting the methyltransferase activity of the complex, and SARS-CoV-2 replication. Overall, this pipeline will help to prioritize PPI targets to accelerate the discovery of early-stage drug candidates targeting protein complexes and pathways.

Identification of small molecule antivirals against HTLV-1 by targeting the hDLG1-Tax-1 protein-protein interaction.

Human T-cell leukemia virus type-1 (HTLV-1) is the first pathogenic retrovirus discovered in human. Although HTLV-1-induced diseases are well-characterized and linked to the encoded Tax-1 oncoprotein, there is currently no strategy to target Tax-1 functions with small molecules. Here, we analyzed the binding of Tax-1 to the human homolog of the drosophila discs large tumor suppressor (hDLG1/SAP97), a multi-domain scaffolding protein involved in Tax-1-transformation ability. We have solved the structures of the PDZ binding motif (PBM) of Tax-1 in complex with the PDZ1 and PDZ2 domains of hDLG1 and assessed the binding of 10 million molecules by virtual screening. Among the 19 experimentally confirmed compounds, one systematically inhibited the Tax-1-hDLG1 interaction in different biophysical and cellular assays, as well as HTLV-1 cell-to-cell transmission in a T-cell model. Thus, our work demonstrates that interactions involving Tax-1 PDZ-domains are amenable to small-molecule inhibition, which provides a framework for the design of targeted therapies for HTLV-1-induced diseases.

AI-guided pipeline for protein-protein interaction drug discovery identifies a SARS-CoV-2 inhibitor.

Protein-protein interactions (PPIs) offer great opportunities to expand the druggable proteome and therapeutically tackle various diseases, but remain challenging targets for drug discovery. Here, we provide a comprehensive pipeline that combines experimental and computational tools to identify and validate PPI targets and perform early-stage drug discovery. We have developed a machine learning approach that prioritizes interactions by analyzing quantitative data from binary PPI assays and AlphaFold-Multimer predictions. Using the quantitative assay LuTHy together with our machine learning algorithm, we identified high-confidence interactions among SARS-CoV-2 proteins for which we predicted three-dimensional structures using AlphaFold Multimer. We employed VirtualFlow to target the contact interface of the NSP10-NSP16 SARS-CoV-2 methyltransferase complex by ultra-large virtual drug screening. Thereby, we identified a compound that binds to NSP10 and inhibits its interaction with NSP16, while also disrupting the methyltransferase activity of the complex, and SARS-CoV-2 replication. Overall, this pipeline will help to prioritize PPI targets to accelerate the discovery of early-stage drug candidates targeting protein complexes and pathways.

Structure-activity relationships of agonists for the orphan G protein-coupled receptor GPR27.

GPR27 belongs, with GPR85 and GPR173, to a small subfamily of three receptors called "Super-Conserved Receptors Expressed in the Brain" (SREB). It has been postulated to participate in key physiological processes such as neuronal plasticity, energy metabolism, and pancreatic ß-cell insulin secretion and regulation. Recently, we reported the first selective GPR27 agonist, 2,4-dichloro-N-(4-(N-phenylsulfamoyl)phenyl)benzamide (I, pEC50 6.34, Emax 100%). Here, we describe the synthesis and structure-activity relationships of a series of new derivatives and analogs of I. All products were evaluated for their ability to activate GPR27 in an arrestin recruitment assay. As a result, agonists were identified with a broad range of efficacies including partial and full agonists, showing higher efficacies than the lead compound I. The most potent agonist was 4-chloro-2,5-difluoro-N-(4-(N-phenylsulfamoyl)phenyl)benzamide (7y, pEC50 6.85, Emax 37%), and the agonists with higher efficacies were 4-chloro-2-methyl-N-(4-(N-phenylsulfamoyl)phenyl)benzamide (7p, pEC50 6.04, Emax 123%), and 2-bromo-4-chloro-N-(4-(N-phenylsulfamoyl)phenyl)benzamide (7r, pEC50 5.99, Emax 123%). Docking studies predicted the putative binding site and interactions of agonist 7p with GPR27. Selected potent agonists were found to be soluble and devoid of cellular toxicity within the range of their pharmacological activity. Therefore, they represent important new tools to further characterize the (patho)physiological roles of GPR27.