-BIOINFORMATIC AND COMPUTATIONAL ANALYSIS OF CELL ENVELOPE PROTEASES FROM LACTIC BACTERIA

CAROL, JUAN J.; BUSTOS, ANA Y.; LEDESMA, ANA E.

Congreso; XVII Congreso Argentino de Microbiología General; 2022

Lactic acid bacteria (LAB) are microorganisms with high nutritional requirements, since they requirecarbohydrates, vitamins, purines, pyrimidines, trace elements, peptides and amino acids to develop.The later are obtained from a complex proteolytic system composed by extracellular proteases, cellenvelope proteases (CEPs), membrane transporters and intracellular peptidases. In the case of LABcapable of growing in dairy products, casein and its derivatives are the main source of nitrogen. In this sense, CEPs have a serine protease domain capable of hydrolyse casein and obtaining the necessary nutrients for the cell. Besides, most of CEPs have numerous domains with variable structural complexity; however, many of their functions are still completely unknown.In previous work, we isolated two strains of Limosilactobacillus plantarum and Lentilactobacillus parabuchneri with high proteolytic activity from artisanal goat cheese. The objective of this work was to predict affinity and types of interactions between the CEPs from Limosilactobacillus plantarum and Lentilactobacillus parabuchneri and casein derivatives, by modelling and molecular docking studies. The construction of 3D models for both proteinases, together with the α and β derivatives of casein of goat origin, was carried out through the AlphaFold2 server and taking into account the pLDDT values, from primary sequences, in FASTA format, available in the UniProt database. Next, these models were validated, using stereochemical parameters, provided by the PROCHECK and VERIFY3D servers by loading their respective files in “.pdb” format. Then, using the Chimera 1.16 program, each of these models could be visualized as well as the percentage of secondary structure, distinguishing between the α-helix, βsheet, turns and disordered structure motifs. the multiple alignment of sequences by both serine protease domains, through the BLAST tool, allowed to know the percentage of identity with other bacterial proteins also available in databases. Finally, by means of the AutoDock 4.2 program, the predictive calculations of interaction between the macromolecules were carried out to know the free energies of stabilization of the union complexes formed and thus infer the preference of the substrate against the casein derivatives. The results showed that the casein binding site is close to the catalytic site. This interaction allows us to confirm the proteolytic activity of these two proteins, where the proteases cleave the casein structure between the PHE and THR residues. In addition, analysis of interaction pairs between residues revealed how hydrophobic and van der Waals interactions enhance ligand binding to the receptor. These results allow us to deepen our knowledge of these enzymes in order to enhance their applications.