# Computational Identification of Potential Novel Allosteric IHF Inhibitors Using QSAR Modeling to Inhibit Plasmid-Mediated Antibiotic Resistance

## TL;DR

This study uses computational methods to identify potential drugs that could block plasmid-mediated antibiotic resistance by targeting a protein called IHF.

## Contribution

The first computational study proposing allosteric inhibition of IHF as a strategy to prevent antibiotic resistance gene transfer.

## Key findings

- A 3D-QSAR model showed strong predictive performance (R2 = 0.90) for anti-plasmid compounds.
- Allosteric site binding showed more favorable energies and stable interactions compared to the DNA-binding site.
- Allosteric ligands induced conformational changes in key residues, potentially disrupting DNA recognition and replication.

## Abstract

The rapid spread of antibiotic resistance through plasmid-mediated conjugation remains a primary global health concern. Despite its critical role in horizontal gene transfer, no approved drugs currently target this process, leaving a critical therapeutic gap. Integration Host Factor (IHF), a DNA-binding protein essential for plasmid replication and mobilization, emerges as a promising yet underexplored target for anti-conjugation strategies. This work aimed to develop a predictive computational model and identify small molecules that disrupt IHF function, thereby reducing plasmid transfer and limiting resistance gene dissemination. A curated dataset of 65 compounds with reported anti-plasmid activity was analyzed using a 3D-QSAR model based on algebraic descriptors computed with QuBiLS-MIDAS. The model was validated through leave-one-out cross-validation (Q2 = 0.82), Tropsha’s criteria, and Y-scrambling. Representative compounds were selected via pharmacophore clustering and evaluated through molecular docking at both the DNA-binding site and a predicted allosteric pocket of IHF. The most promising complexes underwent 200 ns molecular dynamics simulations to assess stability and interaction patterns. The QSAR model demonstrated strong predictive performance (R2 = 0.90). Docking simulations revealed more favorable binding energies at the allosteric site (up to −12.15 kcal/mol) compared to the DNA-binding site. Molecular dynamics confirmed the stability of these interactions, with allosteric complexes showing lower RMSD fluctuations and consistent binding energy profiles. Dynamic cross-correlation analysis revealed that allosteric ligand binding induces conformational changes in key catalytic residues, including Pro65, Pro61, and Leu66. These alterations may compromise DNA recognition and disrupt the initiation of replication. To our knowledge, this is the first computational study proposing allosteric inhibition of IHF as an anti-conjugation strategy. These findings provide a foundation for experimental validation and the development of novel agents to prevent horizontal gene transfer, offering a promising approach to restoring antibiotic efficacy against multidrug-resistant pathogens.