Computational Discovery of Potent Nucleoprotein Inhibitors for Influenza A Virus: Validation Through QM/MM Analysis and Experimental Binding Assays

Zixiao Liu, Jialin Guo, Chao Zhang, Yukun Ding, Shiyang Sun, Bin-Rong Yao, Xing Cheng, Xiaoping Liu, Chun Hu, Junhai XiaoMolecules 2025 · DOI

TL;DR

This paper presents a computational approach to discover new inhibitors for the influenza A virus (IAV) nucleoprotein (NP). By combining virtual screening of millions of compounds with advanced molecular simulations and experimental validation, they identified three potent inhibitors (compounds 8, 13, and 14) with promising antiviral potential.

What Problem Does This Paper Address?

Influenza A virus (IAV) remains a significant global health threat, with existing antiviral treatments facing challenges like drug resistance. The nucleoprotein (NP) of IAV is a highly conserved viral protein essential for replication and lacks a human homolog, making it an attractive target for developing new, broad-spectrum antivirals. However, identifying effective NP inhibitors is complex. This study aims to address this by developing and validating a computational workflow to efficiently discover and characterize novel NP inhibitors.

Methods

The researchers employed a multi-stage computational strategy:

  • Virtual Screening: They screened over 10 million compounds from large chemical libraries using Schrödinger's Glide module, employing High-Throughput Virtual Screening (HTVS), Standard Precision (SP), and Extra Precision (XP) docking protocols. This initial step aimed to identify compounds with favorable binding to the NP protein.
  • Molecular Mechanics/Generalized Born Surface Area (MM-GBSA): Promising candidates from virtual screening underwent MM-GBSA calculations to estimate their binding free energies, further refining the selection process.
  • Molecular Dynamics (MD) Simulations: 100 ns MD simulations were performed on the top candidate complexes to assess the stability of their interactions with the NP protein and understand their dynamic behavior.
  • Alanine Scanning Mutagenesis: This experimental technique was used to confirm the functional importance of key residues identified in computational analyses for ligand binding.
  • Quantum Mechanics/Molecular Mechanics (QM/MM) Analysis: This method was used to analyze electronic properties, such as HOMO-LUMO distributions, to understand pi-pi stacking and electrostatic interactions crucial for drug-receptor binding.
  • Surface Plasmon Resonance (SPR) Assays: This experimental technique was used to validate the binding affinities of the prioritized compounds with the NP target in real-time, measuring equilibrium dissociation constants ($K_D$).

Datasets

The study utilized a large virtual compound library, primarily from the ChemDiv and Chembridge libraries, containing over 10 million molecules. The protein target was the Influenza A virus nucleoprotein, specifically using the crystal structure from PDB ID: 5B7B.

Key Insights and Ideas

  • NP as a Druggable Target: The study reinforces the idea that NP is a highly conserved and essential viral protein, making it an excellent target for developing broad-spectrum antivirals with a reduced likelihood of resistance.
  • Integrated Computational Workflow: The combination of virtual screening, MM-GBSA, MD simulations, and QM/MM analysis proved effective in narrowing down a vast chemical space to a few promising candidates. This integrated approach accelerates drug discovery.
  • Key Residues for Binding: The research identified critical residues (TYR289, LEU306, ASN309, and ARG384) that consistently participate in stable interactions with the NP protein across different ligands, providing valuable insights for future drug design.
  • Mechanism of Action: The study elucidated the binding mechanisms, including hydrogen bonds, hydrophobic interactions, and pi-pi stacking, which are crucial for understanding how these inhibitors function.

Key Results

  • Identification of Potent Inhibitors: The study successfully identified three novel NP inhibitors: compounds 8, 13, and 14.
  • Experimental Validation: SPR assays confirmed superior target engagement for compounds 8, 13, and 14, with $K_D$ values of $7.85 \times 10^{-5}$ M, $3.82 \times 10^{-5}$ M, and $6.97 \times 10^{-5}$ M, respectively. While these are not sub-micromolar, they represent significant binding affinities and indicate novel structural scaffolds.
  • Computational Performance: The virtual screening process successfully reduced the initial pool of millions of compounds to 16 prioritized lead molecules.
  • MD Simulation Stability: MD simulations indicated that compounds 13 and 14 exhibited greater conformational stability compared to compound 8 and the benchmark nucleozin, suggesting more robust binding.

Limitations and Future Work

  • $K_D$ Values: While promising, the reported $K_D$ values are in the micromolar range, suggesting room for optimization to achieve sub-micromolar or nanomolar affinities for enhanced potency.
  • In Vivo Efficacy: The study focused on in vitro and computational validation. Future work should involve optimizing these compounds through residue-directed modifications and advancing them to in vivo efficacy studies to combat seasonal and pandemic influenza threats.
  • Broader Spectrum: The authors suggest that these inhibitors exhibit potential for broad-spectrum activity against influenza A and B viruses, which warrants further investigation.

Summary updated

Abstract

This study employed an integrated computational approach to discover novel nucleoprotein (NP) inhibitors for influenza A virus (IAV). Beginning with virtual screening of over 10 million compounds using Schrödinger's Glide module (HTVS, SP, XP docking), the workflow identified promising candidates with favorable binding energies. Subsequent molecular mechanics/generalized born surface area (MM-GBSA) calculations and 100 ns molecular dynamics (MD) simulations prioritized 16 compounds for experimental validation. Surface plasmon resonance (SPR) assays revealed that compounds <b>8</b>, <b>13</b>, and <b>14</b> demonstrated superior target engagement, showing equilibrium dissociation constants (<i>K</i><sub>D</sub>) of 7.85 × 10<sup>-5</sup> M, 3.82 × 10<sup>-5</sup> M, and 6.97 × 10<sup>-5</sup> M, respectively. Molecular dynamics, alanine scanning mutagenesis, and quantum mechanics/molecular mechanics (QM/MM) analysis were conducted to analyze the binding modes, providing a reference for the design of subsequent compounds. These findings validate the efficacy of structure-based virtual screening in identifying high-affinity NP inhibitors and provide insights for the development of broad-spectrum anti-influenza therapeutics.
BibTeX
@inproceedings{liu2025computational,
  author = {Zixiao Liu and Jialin Guo and Chao Zhang and Yukun Ding and Shiyang Sun and Bin-Rong Yao and Xing Cheng and Xiaoping Liu and Chun Hu and Junhai Xiao},
  title = {{Computational Discovery of Potent Nucleoprotein Inhibitors for Influenza A Virus: Validation Through QM/MM Analysis and Experimental Binding Assays}},
  booktitle = {Molecules},
  year = {2025},
  doi = {10.3390/molecules30193960},
}
Figure