Genomic context as well as sequence of both psr and penicillin-binding protein 5 contributes to β-lactam resistance...

Penicillin-binding protein 5 (PBP5) of Enterococcus faecium (Efm) is vital for ampicillin resistance (AMP-R). We previously designated three forms of PBP5, namely, PBP5-S in Efm clade B strains [ampicillin susceptible (AMP-S)], PBP5-S/R (AMP-S or R), and PBP5-R (AMP-R) in clade A strains. Here, pbp5 deletion resulted in a marked reduction in AMP minimum inhibitory concentrations (MICs) to 0.01-0.09 µg/mL for clade B and 0.12-0.19 µg/mL for clade A strains; in situ complementation restored parental AMP MICs. Using D344SRF (lacking ftsW/psr/pbp5), constructs with ftsW_A/psr_A (from a clade A1 strain) cloned upstream of pbp5-S and pbp5-S/R alleles resulted in modest increases in MICs to 3-8 µg/mL, while high MICs (>64 µg/mL) were seen using pbp5 from A1 strains. Next, using ftsW ± psr from clade B and clade A/B and B/A hybrid constructs, the presence of psr_B, even alone or in trans, resulted in much lower AMP MICs (3-8 µg/mL) than when psr_A was present (MICs >64 µg/mL). qRT PCR showed relatively greater pbp5 expression (P = 0.007) with pbp5 cloned downstream of clade A1 ftsW/psr (MIC >128 µg/mL) vs when cloned downstream of clade B ftsW/psr (MIC 4-16 µg/mL), consistent with results in western blots. In conclusion, we report the effect of clade A vs B psr on AMP MICs as well as the impact of pbp5 alleles from different clades. While previously, Psr was not thought to contribute to AMP MICs in Efm, our results showed that the presence of psr_B resulted in a major decrease in Efm AMP MICs.

The findings of this study shed light on ampicillin resistance in Enterococcus faecium clade A strains. They underscore the significance of alterations in the amino acid sequence of penicillin-binding protein 5 (PBP5) and the pivotal role of the psr region in PBP5 expression and ampicillin resistance. Notably, the presence of a full-length psrB leads to reduced PBP5 expression and lower minimum inhibitory concentrations (MICs) of ampicillin compared to the presence of a shorter psrA, regardless of the pbp5 allele involved. Additionally, clade B E. faecium strains exhibit lower AMP MICs when both psr alleles from clades A and B are present, although it is important to consider other distinctions between clade A and B strains that may contribute to this effect. It is intriguing to note that the divergence between clade A and clade B E. faecium and the subsequent evolution of heightened AMP MICs in hospital-associated strains appear to coincide with changes in Pbp5 and psr. These changes in psr may have resulted in an inactive Psr, facilitating increased PBP5 expression and greater ampicillin resistance. These results raise the possibility that a mimicker of PsrB, if one could be designed, might be able to lower MICs of ampicillin-resistant E. faecium, thus potentially resorting ampicillin to our therapeutic armamentarium for this species.