Candidate Binding Sites for Allosteric Inhibition of the SARS-CoV-2 Main Protease from the Analysis of Large-Scale Molecular Dynamics Simulations

Matteo Carli, Giulia Sormani, Álex Rodríguez, Alessandro LaioThe Journal of Physical Chemistry Letters 2020 · DOI

TL;DR

This paper uses a novel, high-dimensional analysis of a long molecular dynamics simulation to identify stable, yet transient, conformations of the SARS-CoV-2 main protease. By characterizing these "metastable states," the researchers propose specific binding sites for allosteric inhibitors, offering a new strategy to combat the virus by indirectly blocking the enzyme's active site.

What Problem Does This Paper Address?

The SARS-CoV-2 main protease (Mpro) is a critical enzyme for viral replication, making it a prime target for antiviral drug development. While directly blocking the active site with inhibitors is a common strategy, it can lead to resistance and cross-reactivity with human proteases. Allosteric inhibition, which targets sites away from the active site to indirectly alter its function, offers a promising alternative. However, identifying these allosteric sites and understanding the protein dynamics that enable them is challenging. This paper addresses the need for a method to systematically explore the complex conformational landscape of Mpro and pinpoint potential allosteric binding sites that can be exploited for drug design.

Methods

The researchers employed a sophisticated, non-parametric data analysis approach to dissect a 100 µs molecular dynamics (MD) trajectory of the SARS-CoV-2 main protease. This method allows for the characterization of the protein's free energy landscape across hundreds of variables simultaneously, going beyond traditional approaches that focus on a limited number of degrees of freedom.

Key techniques include:

  • High-Dimensional Free Energy Landscape Analysis: Instead of relying on predefined collective variables, their approach analyzes the landscape in spaces defined by all backbone dihedral angles and by the formation/breaking of residue contacts. This allows for the unbiased discovery of stable, low-energy conformations.
  • Identification of Metastable States: Local minima in the free energy landscape are identified as "metastable states" – conformations that the protein visits and remains in for significant durations (hundreds of nanoseconds).
  • Contact Map Analysis: The formation and breaking of specific residue-residue contacts are tracked. These contacts are crucial for understanding how different parts of the protein interact and influence each other's dynamics.
  • Solvent Accessible Surface Area (SASA) and Pocket Doorway Area (PDA) Calculations: These metrics are used to quantify the accessibility of the catalytic dyad and the overall opening of the active site pocket, respectively.
  • Druggability Assessment: Standard software tools (PockDrug, fpocket, MDpocket) were used to evaluate the potential of identified sites for drug binding.
  • Sequence Conservation Analysis: Residues involved in proposed binding sites were analyzed for their conservation across related protein families to assess the potential for broad efficacy and reduced resistance.

Datasets

The primary dataset used is a 100 µs molecular dynamics (MD) trajectory of the SARS-CoV-2 main protease. This trajectory was generated by D. E. Shaw Research. No new benchmarks were introduced; rather, the study leverages existing simulation data.

Key Insights and Ideas

  • Ergodicity is Not Reached: Even with a 100 µs simulation, the two monomers of the Mpro dimer explored distinct conformational spaces, suggesting that full ergodicity (exploring all possible conformations) is not achieved within this timescale. This highlights the complexity of protein dynamics.
  • Metastable States as Drug Targets: The identification of stable, yet transient, conformations (metastable states) provides concrete structural snapshots that can be targeted for allosteric inhibition. Stabilizing these states with a binder could effectively "trap" the enzyme in a less active form.
  • Allosteric Sites Beyond the Active Site: The analysis revealed potential allosteric binding sites located far from the catalytic dyad, particularly in the interface region between domains II and III (the "distal pocket"). This expands the search space for drug discovery.
  • Contact Patterns Dictate Accessibility: Specific patterns of residue-residue contacts, especially those involving flexible loops around the active site, directly correlate with the accessibility of the catalytic dyad. Manipulating these contacts offers a route to allosteric control.
  • Conservation Suggests Robustness: The high conservation of residues involved in the proposed binding sites across related coronaviruses suggests that allosteric inhibitors targeting these sites could have broad efficacy and be less prone to resistance mutations.

Key Results

  • Identification of 18 Metastable States: The analysis identified 18 distinct metastable states for the SARS-CoV-2 main protease, characterized by unique conformational features and contact patterns.
  • Characterization of Key Loops: The study detailed the dynamics of flexible loops (left flap and linker loop) that control access to the active site, linking their conformations to specific contact formations.
  • Proposal of Three Allosteric Binding Sites: Based on the analysis, the authors propose three specific regions as potential targets for allosteric inhibition:
  • Left Flap Region: Targeting contacts that influence the opening and closing of the left flap.
  • Upper Flap Region: Targeting contacts that affect the positioning of the upper flap relative to the catalytic dyad.
  • Distal Pocket Region: Targeting a set of five specific contacts in the interface between domains II and III, which indirectly influences the active site accessibility.
  • Evidence for Allosteric Influence: The study demonstrated that specific contact patterns in the distal pocket region significantly influence the pocket doorway area (PDA) and solvent-accessible surface area (SASA) of the catalytic dyad, providing evidence for allosteric effects.
  • High Druggability Scores: The proposed target sites were found to have high druggability scores according to computational tools.

Limitations and Future Work

  • Simulation Timescale: While 100 µs is a substantial simulation time, the authors acknowledge that full ergodicity was not reached, meaning some metastable states might have been missed or not fully sampled.
  • Experimental Validation: The proposed binding sites and allosteric mechanisms require experimental validation through methods like site-directed mutagenesis and binding assays.
  • Dynamic Docking: The authors suggest that dynamical docking simulations would be the next step to assess the proposed targets from a more accurate biochemical standpoint.
  • Broader Applicability: The authors emphasize that their analysis approach is general and can be applied to other protein targets for the discovery of allosteric inhibitors.

This work provides a powerful framework for understanding complex protein dynamics and offers concrete, data-driven hypotheses for the development of novel antiviral therapies against SARS-CoV-2.

Summary updated

Abstract

We analyzed a 100 μs MD trajectory of the SARS-CoV-2 main protease by a non-parametric data analysis approach which allows characterizing a free energy landscape as a simultaneous function of hundreds of variables. We identified several conformations that, when visited by the dynamics, are stable for several hundred nanoseconds. We explicitly characterize and describe these metastable states. In some of these configurations, the catalytic dyad is less accessible. Stabilizing them by a suitable binder could lead to an inhibition of the enzymatic activity. In our analysis we keep track of relevant contacts between residues which are selectively broken or formed in the states. Some of these contacts are formed by residues which are far from the catalytic dyad and are accessible to the solvent. Based on this analysis we propose some relevant contact patterns and three possible binding sites which could be targeted to achieve allosteric inhibition.
BibTeX
@article{carli2020candidate,
  author = {Matteo Carli and Giulia Sormani and Álex Rodríguez and Alessandro Laio},
  title = {{Candidate Binding Sites for Allosteric Inhibition of the SARS-CoV-2 Main Protease from the Analysis of Large-Scale Molecular Dynamics Simulations}},
  journal = {The Journal of Physical Chemistry Letters},
  year = {2020},
  doi = {10.1021/acs.jpclett.0c03182},
}
Figure