Deciphering the Coevolutionary Dynamics of L2 β-Lactamases via Deep Learning.

Area of Science:

• Biochemistry and Molecular Biology
• Computational Biology
• Microbiology

Background:

• L2 β-lactamases, serine-based class A enzymes from Stenotrophomonas maltophilia, are critical mediators of antimicrobial resistance (AMR).
• Despite their significance, research on L2 β-lactamases remains limited.
• Understanding their evolutionary dynamics is crucial for developing effective AMR strategies.

Purpose of the Study:

• To investigate the coevolutionary dynamics of L2 β-lactamases.
• To explore conformational changes and correlations within the L2 β-lactamase family and other class A β-lactamases.
• To identify the role of hydrophobic nodes and binding site residues in enzyme function.

Main Methods:

• Adaptive sampling molecular dynamics simulations.
• Deep learning methods, including convolutional variational autoencoders and BindSiteS-CNN.
• Comparative analysis of L2 β-lactamases with SME-1 and KPC-2.

Main Results:

• Comprehensive insights into the dynamic behavior and evolutionary trajectory of β-lactamases were obtained.
• The study elucidated the potential role of specific residues and hydrophobic interactions in enzyme function.
• A theoretical framework for understanding β-lactamase evolution under environmental pressure was established.

Conclusions:

• The convergence of computational approaches provides a robust understanding of L2 β-lactamase evolution.
• This research offers a promising strategy for drug development against AMR.
• The findings lay the groundwork for future studies on combating antimicrobial resistance through enzyme-targeted therapies.