Integrative In Silico and Experimental Characterization of Endolysin LysPALS22: Structural Diversity, Ligand Binding Affinity, and Heterologous Expression.

Endolysins, phage-derived enzymes capable of lysing bacterial cell walls, hold significant promise as novel antimicrobials against resistant Gram-positive and Gram-negative pathogens. In this study, we undertook an integrative approach combining extensive in silico analyses and experimental validation to characterize the novel endolysin LysPALS22. Initially, sixteen endolysin sequences were selected based on documented lytic activity and enzymatic diversity, and subjected to multiple sequence alignment and phylogenetic analysis, which revealed highly conserved catalytic and binding domains, particularly localized to the N-terminal region, underscoring their functional importance. Building upon these sequence insights, we generated three-dimensional structural models using Swiss-Model, EBI-EMBL, and AlphaFold Colab, where comparative evaluation via Ramachandran plots and ERRAT scores identified the Swiss-Model prediction as the highest quality structure, featuring over 90% residues in favored conformations and superior atomic interaction profiles. Leveraging this validated model, molecular docking studies were conducted in PyRx with AutoDock Vina, performing blind docking of key peptidoglycan-derived ligands such as N-Acetylmuramic Acid-L-Alanine, which exhibited the strongest binding affinity (-7.3 kcal/mol), with stable hydrogen bonding to catalytic residues ASP46 and TYR61, indicating precise substrate recognition. Visualization of docking poses using Discovery Studio further confirmed critical hydrophobic and polar interactions stabilizing ligand binding. Subsequent molecular dynamics simulations validated the stability of the LysPALS22-NAM-LA complex, showing minimal structural fluctuations, persistent hydrogen bonding, and favorable interaction energies throughout the 100 ns trajectory. Parallel to computational analyses, LysPALS22 was heterologously expressed in