Engineering dynamic gates in binding pocket of penicillin G acylase to selectively degrade bacterial signaling molecules

Abstract

The rapid increase in antibiotic-resistant bacteria and the inability to provide new generations of potent antimicrobials necessitates the search for new, unconventional solutions. Methods based on targeting bacterial communication induced by signaling molecules, known as quorum sensing, are gaining increasing interest. Quorum quenching (QQ), as the process of interrupting this communication is called, can be achieved by enzymatic degradation of signaling molecules and represents a promising solution as it limits the expression of genes responsible for virulence, biofilm formation, and drug resistance. It is also believed to circumvent common resistance mechanisms. Therefore, enzymes with QQ activity represent potential next-generation antimicrobial agents for use in medicine, industry, and other areas of life. This work focuses on a biotechnologically optimized penicillin G acylase from Escherichia coli (ecPGA), for which primary QQ activity for smaller signaling molecules was recently confirmed. Herein, we introduced triple-point mutations within the binding pocket by an ensemble-based design aimed at modulating the pocket structure and the dynamics of its entrance gates. Next, we proposed a computational workflow to select promising combinations for further modeling. We selected three candidates for experimental evaluation using molecular dynamics simulations of the constructs with six different, biologically relevant signaling molecules. These comprised (i) the VAF variant with enhanced activity towards the medium-sized ligands like the signaling molecule of Burkholderia cenocepacia, C08-HSL (N-octanoyl-L-homoserine lactone); (ii) the YAF variant preferring longer substrates like the signaling compound of pathogenic Vibrio species, C10-HSL (N-decanoyl-L-homoserine lactone); and finally (iii) the MSF variant with improved efficacy for the longest substrate, C12-3O-HSL (N-3-oxo-dodecanoyl-L-homoserine lactone), the signaling molecule of Pseudomonas aeruginosa. In-depth analyses of these engineered variants revealed modulated topology and dynamics of the binding pockets. While we could consistently expand the pockets in these variants, the reactive binding of longer substrates became limited, due to either overly promoted dynamics of the pocket in the VAF variant or an overstabilized pocket in the MSF variant. In summary, we demonstrated the designability of ecPGA for improved QQ and provided insights into the role of adequately modulated pocket dynamics for the activity. Such knowledge, together with the methodology developed for filtering and scoring large datasets of potential variants that reflect the outcomes of our biochemical assays, may provide a suitable toolbox for future exploration and design of tailored QQ acylases toward particular signaling molecules, making them viable antimicrobial agents.