Discovery of a cryptic pocket in the AI-predicted structure of PPM1D phosphatase explains the binding site and potency of its allosteric inhibitors

Abstract

Virtual screening is a widely used tool for drug discovery, but its predictive power can vary dramatically depending on how much structural data is available. In the best case, crystal structures of a ligand-bound protein can help find more potent ligands. However, virtual screens tend to be less predictive when only ligand-free crystal structures are available, and even less predictive if a homology model or other predicted structure must be used. Here, we explore the possibility that this situation can be improved by better accounting for protein dynamics, as simulations started from a single structure have a reasonable chance of sampling nearby structures that are more compatible with ligand binding. As a specific example, we consider the cancer drug target PPM1D/Wip1 phosphatase, a protein that lacks crystal structures. High-throughput screens have led to the discovery of several allosteric inhibitors of PPM1D, but their binding mode remains unknown. To enable further drug discovery efforts, we assessed the predictive power of an AlphaFold-predicted structure of PPM1D and a Markov state model (MSM) built from molecular dynamics simulations initiated from that structure. Our simulations reveal a cryptic pocket at the interface between two important structural elements, the flap and hinge regions. Using deep learning to predict the pose quality of each docked compound for the active site and cryptic pocket suggests that the inhibitors strongly prefer binding to the cryptic pocket, consistent with their allosteric effect. The predicted affinities for the dynamically uncovered cryptic pocket also recapitulate the relative potencies of the compounds (τb = 0.70) better than the predicted affinities for the static AlphaFold-predicted structure (τb = 0.42). Taken together, these results suggest that targeting the cryptic pocket is a good strategy for drugging PPM1D and, more generally, that conformations selected from simulation can improve virtual screening when limited structural data is available.

Exploring Conformational Landscapes and Cryptic Binding Pockets in Distinct Functional States of the SARS-CoV-2 Omicron BA.1 and BA.2 Trimers: Mutation-Induced Modulation of Protein Dynamics and Network-Guided Prediction of Variant-Specific Allosteric Binding Sites

“…We employed conformational ensembles of the S-BA.1 and S-BA.2 trimers in combination with template-free and highly efficient P2Rank and PASSerRank approaches to describe the available spectrum of potential binding sites and compute the residue-based pocket propensities in the ensembles of the S proteins. We combined these robust tools for the enumeration of potential pockets with a network-based weighing of the pocket probabilities to enable pocket ranking based on allosteric function which provides an adaptation and extension of the reversed allosteric communication strategy [81][82][83][84][85][86]. The central result of this analysis is the discovery of variant-specific differences in the distribution of binding sites in the BA.1 and BA.2 trimers, suggesting that small variations could lead to different preferences in the allocation of druggable sites (Figures 6 and 7).…”

“…By combining MD simulations, Markov state models (MSMs), and a novel MSM-based approach to aggregating docking results across structural ensembles, a recent study identified cryptic pockets targeted by allosteric myosin-II; this functional pocket is always closed in ligand-free experimental structures [84]. The cryptic binding sites were uncovered in AlphaFold-predicted ensembles of PPM1D phosphatase, and a neural network trained to evaluate the quality of docked poses predicted that this site is the most likely binding mode for the allosteric inhibitors of PPM1D [85].…”

A Reversible Spherical Geometric Conversion of Protein Backbone Structure Coordinate Matrix to Three Independent Vectors of ρ, θ and φ

“…Protein structure prediction is the computational task of determining the three-dimensional structure of a protein from its amino acid sequence. This process is crucial for understanding protein function, interactions, and designing novel therapeutics [1][2][3][4][5][6][7][8]. By definition, accurate prediction of protein structure from its amino acid sequence is a formidable challenge in computational structural biology, with profound implications for understanding biological function and designing novel therapeutics [9][10][11][12].…”

Toward physics‐based precision medicine: Exploiting protein dynamics to design new therapeutics and interpret variants

“…Moreover, many proteins must be written off as undruggable because their structures lack pockets where an inhibitor has the potential to bind tightly enough to serve as a valuable drug (Borrel et al, 2015 ; Cox et al, 2014 ; Hopkins & Groom, 2002 ). Finally, current computational drug design methods struggle to quantitatively predict protein–ligand binding affinities, suggesting there is a fatal flaw in the single structure assumption (Jones et al, 2021 ; Meller, de Oliveira, et al, 2023 ).…”