Proposal of metagenomic-origin LRA-5 as a precursor of active β-lactamases through Tyr69Gln and Val166Glu amino acid substitutions: a functional and structural analysis.

Wild-type LRA-5, recovered from Alaskan soil samples, shares no more than 33% amino acid sequence identity with enzymes from pathogens like PER β-lactamases. Recombinant E. coli expressing wild-type LRA-5 and its engineered variants LRA-5Y69Q and LRA-5V166E showed MIC values equivalent to control strains. However, LRA-5Y69Q/V166E displayed MICs above the resistant breakpoint for some β-lactams. Kinetic parameters correlated with the MICs, showing that the catalytic efficiency of LRA-5Y69Q/V166E was comparable to those from class A β-lactamases, such as CTX-M-15, PER-2, and KPC-2. LRA-5Y69Q/V166E exhibited kcat/Km values up to 11,000-fold higher compared to wild-type LRA-5, which is associated with the presence of Glu166. The X-ray crystallographic structure of wild-type LRA-5 (1.80 Å; PDB 8EO5) shows that the lack of both Glu166 and a deacylation water molecule contributes to a biologically insignificant activity. Interactions observed between LRA-5 and ceftazidime (2.35 Å; PDB 8EO6) show structural conservation with other β-lactamases. In contrast, the crystallographic structure of LRA-5Y69Q/V166E (2.15 Å; PDB 8EO7) bears a deacylation water molecule that is associated with the increase in catalytic activity compared to the wild-type variant. Circular dichroism results confirm that amino acid substitutions in LRA-5 do not affect the overall content of the secondary/tertiary structures. Evidence suggests that alternative evolutionary paths could have occurred for β-lactamases like LRA-5, produced by environmental microorganisms: (i) proteins having similar structural features than active β-lactamases may accumulate a small number of mutations (e.g., Y69Q/V166E) to yield active enzymes and (ii) the β-lactamase fold may have lost key residues in the absence of antibiotics.