Allosteric Regulation of Hsp90α’s Activity by Small Molecules Targeting the Middle Domain of the Chaperone

Abstract

Hsp90 is a target for anti-cancer drug development. Both the conformational events tuned by ATP/ ADP and co-chaperones and the chaperoning cycle timing are required for Hsp90's fully functional display. Interfering with either one of the conformational events or the cycle timing will down-regulate Hsp90's function. In this manuscript, non-covalent allosteric modulators (SOMCL-16-171 and SOMCL-16-175) targeting Hsp90a's middle domain (Hsp90M) were developed for the first time. Multiple techniques were then applied to characterize the interactions between two active compounds and Hsp90a. Two loops and one a-helix (F349-N360, K443-E451, and D372-G387) in Hsp90M were identified responsible for the recognition of SOMCL-16-171 and SOMCL-16-175. Meanwhile, the binding of SOMCL-16-171 and SOMCL-16-175 to Hsp90M was demonstrated to allosterically modulate the structure and function of Hsp90a's N-terminal domain. Finally, cellular assays were conducted to evaluate the cellular activity of SOMCL-16-175, and the results indicate that SOMCL-16-175 destabilizes Hsp90's client proteins and reduces cell viability.

Applications of Solution NMR in Drug Discovery

“…This ratio is important because a larger compound-target ratio may mask binding due to the strong signal from the free compound [ 86 , 89 ]. As that shown in Figure 2 a, active compound SOMCL-16-171 exhibited substantial line broadening and signal shifting in the recorded CPMG spectrum upon its binding to Hsp82, which is a yeast homologue of human Hsp90 [ 90 ]. Saturation transfer difference (STD) spectroscopy is another commonly used ligand-observed NMR method in the first round of fragment hit screening and validation [ 85 , 91 ].…”

“…The STD experiment is sensitive to off-rate kinetic constant of the binding process [ 85 ]; generally, an appropriate range of the dissociation constant between the target and ligand for STD experiment is about 1 mM to 0.1 µM [ 85 , 92 ]. As shown in Figure 2 b, fragment hit 1-E6, which is a weak binder to the middle domain of Hsp90 (Hsp90M), exhibited positive STD signals upon the presence of Hsp90M [ 90 ]. In our group, CPMG and STD were jointly applied to discover fragment hits targeting BRD4, Hsp90M (Hsp90α’s middle domain), PDEδ, USP7.…”

“…In our group, CPMG and STD were jointly applied to discover fragment hits targeting BRD4, Hsp90M (Hsp90α’s middle domain), PDEδ, USP7. Multiple fragment hits with new scaffolds against the aforementioned protein targets were identified [ 37 , 70 , 90 ]. Among four target proteins, Hsp90M (Hsp90α’s middle domain) is the so-called ‘’undruggable’’ protein.…”

“…Then, the chemical shift perturbation values of several residues in Hsp90M that exhibited substantial chemical shift changes were applied to calculate the binding affinity of SOMCL-16-175 to Hsp90M according to Equation (1) [ 110 ]. A zoomed-in view of the titration spectra for residue L328 in Hsp90M was presented for illustration purpose, and the binding affinity of SOMCL-16-175 to Hsp90M was determined (Kd = 804 ± 24 µM) accordingly [ 90 ]. For the binder with a moderate binding affinity (Kd in micromolar range) to the target protein, an intermediate exchange between the target in the apo state and the target in the ligand bound state is expected, and the gradually shifted and attenuated chemical shifts for the perturbed residues in the target protein will be observed with the addition of increased amounts of binder.…”

“…We found that two loops and one α-helix (F349-N360, K443-E451, and D372-G387) in Hsp90M are responsible for the recognition of SOMCL-16-171 and SOMCL-16-175. Meanwhile, the binding of SOMCL-16-171 and SOMCL-16-175 to Hsp90M allosterically modulates the structure and function of Hsp90α’s N-terminal domain, which consequently up-regulates Hsp90α’s ATPase activity [ 90 ].…”