The Gαi-GIV binding interface is a druggable protein-protein interaction

Abstract

Heterotrimeric G proteins are usually activated by the guanine-nucleotide exchange factor (GEF) activity of GPCRs. However, some non-receptor proteins are also GEFs. GIV (a.k.a Girdin) was the first non-receptor protein for which the GEF activity was ascribed to a well-defined protein sequence that directly binds Gαi. GIV expression promotes metastasis and disruption of its binding to Gαi blunts the pro-metastatic behavior of cancer cells. Although this suggests that inhibition of the Gαi-GIV interaction is a promising therapeutic strategy, protein-protein interactions (PPIs) are considered poorly “druggable” targets requiring case-by-case validation. Here, we set out to investigate whether Gαi-GIV is a druggable PPI. We tested a collection of >1,000 compounds on the Gαi-GIV PPI by in silico ligand screening and separately by a chemical high-throughput screening (HTS) assay. Two hits, ATA and NF023, obtained in both screens were confirmed in secondary HTS and low-throughput assays. The binding site of NF023, identified by NMR spectroscopy and biochemical assays, overlaps with the Gαi-GIV interface. Importantly, NF023 did not disrupt Gαi-Gβγ binding, indicating its specificity toward Gαi-GIV. This work establishes the Gαi-GIV PPI as a druggable target and sets the conceptual and technical framework for the discovery of novel inhibitors of this PPI.