Informing plasmid compatibility with bacterial hosts using protein-protein interaction data.

Area of Science:

• Microbiology
• Computational Biology
• Genetics

Background:

• Plasmids are key vectors for antimicrobial resistance (AMR) and virulence genes.
• In vitro screening for plasmid-host compatibility is time-consuming.
• Computational methods can predict plasmid-host interactions.

Purpose of the Study:

• To computationally analyze protein-protein interactions (PPIs) for predicting plasmid-host compatibility.
• To identify potential host cells compatible with significant AMR-carrying plasmids like pOXA-48.
• To understand the role of PPIs in plasmid establishment and AMR dissemination.

Main Methods:

• Analysis of protein-protein interactions (PPIs) between plasmid and host cell proteins.
• Screening of large bacterial datasets (n=4363) for interaction patterns.
• Identification of specific plasmid-host interactions, including those with pOXA-48 and plasmids common in Escherichia coli.

Main Results:

• Identified significant excesses of PPIs for eight key plasmids, including pOXA-48, across numerous bacterial species.
• Found 23 species with high interaction rates for four blaOXA-48-positive plasmids.
• Discovered 48 species with high interaction rates for plasmids prevalent in Escherichia coli, and a link between a plasmid and the pil operon enhancing adhesion.

Conclusions:

• Excessive PPIs indicate potential host-plasmid compatibility, a critical factor in AMR spread.
• Computational PPI analysis offers a scalable approach to predict plasmid-host compatibility.
• Understanding these interactions is vital for developing strategies to control AMR.