Hsp90 is an essential chaperone that guards proteome integrity and amounts to 2% of cellular protein. We now find that Hsp90 also has the ability to directly interact with and deform membranes via an evolutionarily conserved amphipathic helix. Using a new cell-free system and in vivo measurements, we show this amphipathic helix allows exosome release by promoting the fusion of multivesicular bodies (MVBs) with the plasma membrane. We dissect the relationship between Hsp90 conformation and membrane-deforming function and show that mutations and drugs that stabilize the open Hsp90 dimer expose the helix and allow MVB fusion, while these effects are blocked by the closed state. Hence, we structurally separated the Hsp90 membrane-deforming function from its well-characterized chaperone activity, and we show that this previously unrecognized function is required for exosome release.

This article is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND). Usage and distribution for commercial purposes as well as any distribution of modified material requires written permission.

Herein, we describe the synthesis and structure–activity relationships of cyclic peptides designed to target heat shock protein 90 (Hsp90). Generating 19 compounds and evaluating their binding affinity reveals that increasing electrostatic interactions allows the compounds to bind more effectively with Hsp90 compared to the lead structure. Exchanging specific residues for lysine improves binding affinity for Hsp90, indicating some residues are not critical for interacting with the target, whereas others are essential. Replacing l- for d-amino acids produced compounds with decreased binding affinity compared to the parent structure, confirming the importance of conformation and identifying key residues most important for binding. Thus, a specific conformation and electrostatic interactions are required in order for these inhibitors to bind to Hsp90.

Macrocycles have several advantages over small-molecule drugs when it comes to addressing specific protein-protein interactions as therapeutic targets. Herein we report the synthesis of seven new cyclic peptide molecules and their biological activity. These macrocycles were designed to understand how moving an N-methyl moiety around the peptide backbone impacts biological activity. Because the lead non-methylated structure inhibits the oncogenic regulator heat-shock protein 90 (Hsp90), two of the most potent analogues were evaluated for their Hsp90 inhibitory activity. We show that incorporating an N-methyl moiety controls the conformation of the macrocycle, which dramatically impacts cytotoxicity and binding affinity for Hsp90. Thus, the placement of an N-methylated amino acid within a macrocycle generates an unpredictable change to the compound's conformation and hence biological activity.

Protein–protein interactions (PPIs) regulate all signaling pathways for cellular function. Developing molecules that modulate PPIs through the interface of their protein surfaces has been a significant challenge and there has been little success controlling PPIs through standard molecular library screening approaches. PPIs control the cell’s protein-folding machinery, and this machinery relies on a multiprotein complex formed with heat shock protein 70 (Hsp70). Described is the design, synthesis, and biological evaluation of molecules aimed to regulate the interaction between two proteins that are critical to the protein-folding machinery: heat shock protein 70 (Hsp70) and cochaperone heat shock organizing protein (HOP). We report the first class of compounds that directly regulate these two protein–protein interactions and inhibit protein folding events.