The Gram-positive foodborne pathogen Listeria monocytogenes is a causative agent of gastroenteritis 1 and in severe cases, listeriosis, which ranges from a mild flu-like illness to meningitis in non-pregnant individuals, or as infection of the foetus in pregnant women 2. Ampicillin, either alone or in combination with gentamicin, remains the gold standard treatment for L. monocytogenes infection 3. Conversely, cephalosporins are usually poorly active and thus not recommended in the treatment of listeriosis 45. This dissociation in susceptibilities between β-lactams has been rationalized by the observation that cephalosporins have poor affinity for PBP3 compared to penicillin, suggesting that PBP3 is the primary lethal target for β-lactams in L. monocytogenes 6. However, at least five different PBPs have been identified in L. monocytogenes based on their ability to bind radioactive penicillin 67. Moreover, genome analysis revealed seven distinct genes with homologies to PBPs in L. monocytogenes 8. With the exception of pbpB, each of the remaining genes has been disrupted by insertional mutagenesis 9 without loss of cell viability, suggesting that they are not critically required for normal synthesis of cell wall peptidoglycan. Based on a previous study by Guinane et al., 9 in which an increase in the activity of cephalosporins was observed for L. monocytogenes disrupted in lmo2229 and lmo0441, we investigated the role of their corresponding proteins (PBP4 and Lmo0441 respectively) in protecting the pathogen against a penicillin (ampicillin, AMP), a carbapenem (meropenem, MEM) and a cephalosporin (cefuroxime, CFX).

The growth rates for wild-type and mutant strains, together with time profiles for the activity of AMP, MEM and CFX at 100 times their MIC were tested over a 24 h period in TSB broth (Fig. 1, left panel). No significant differences in the growth rates were observed between wild type and both mutant strains. However, this effect is only observed after an incubation of 24 h, the ΔPBP4 strain appeared significantly more sensitive than the wild type to both AMP and MEM, a phenotype not observed for the ΔLmo0441 strain. In contrast, both PBP mutant strains appeared less sensitive to CFX, after 24h, when compared to the wild type strain, an unexpected result given their MICs.

To better understand the role of PBP4 and Lmo0441 proteins, full dose-response experiments were run with AMP, MEM and CFX using a wide range of concentrations and are presented in Fig. 1, right panel. Data collected were used to determine the Hill function regression parameters allowing the determination of the relative drug efficacy (E_max) presented in Table 1. An increase of drug efficacy was observed for AMP and MEM against the ΔPBP4 strain (1.64 and 1.58 times the E_max of the parental wild-type strain respectively); an effect which could not have been predicted from MIC measurements; while E_max of AMP and MEM were not affected in the ΔLmo0441 strain. In contrast to the results obtained for AMP and MEM, both PBPs mutant strains were affected by the activity of CFX and in this case, a decrease of drug efficacy was observed with the use of ΔPBP4 and ΔLmo0441 strains (0.63 and 0.53 times the E_max of the wild-type strain respectively).

In conclusion then, the construction of PBP mutant strains ΔPBP4 and ΔLmo441 (Fig. 2) displayed no obvious growth defects in L. monocytogenes, suggesting that the individual PBPs are most likely not essential for growth under normal growth conditions. Furthermore, we demonstrate a role for PBP4 in penicillin and carbapenem activities in L. monocytogenes. Indeed, dose-response experiments over a wide concentration range reveals an increase in sensitivity of the ΔPBP4 strain for both AMP and MEM compared to the parental wild-type strain. This observation, which could not have been inferred from MIC measurements alone, underscores the importance of accurate techniques to predict antibiotic activity. Moreover, this effect appears specific to the ΔPBP4 strain and is not apparent in the ΔLmo0441 mutant. The observed protective effect of PBP4 against penicillin and carbapenem activities may result from decreased competition between PBPs or from the loss of cooperative action with other PBPs for the formation of the cell wall peptidoglycan.

This study also reveals that loss of PBP4 or Lmo0441 proteins allows a decrease of maximal activity of cephalosporin, a class of antibiotics which, due to their poor affinity for the PBP3, are not particularly effective against L. monocytogenes 6. This observation, which runs contrary to what was observed for AMP and MEM may be explained by the fact that PBP3 may increase its relative action in peptydoglycan synthesis when PBP4 or lmo0441 are deleted: conferring resistance to cephalosporin. Thus, even if PBP3 remains the principal target of β-lactam antibiotics, as is the current dictum, PBP4 and Lmo0441 also appear to play an important role in the activity of β-lactam antibiotics.

The authors declare that they have no competing interests.

SVDV, PMT, CH and RDS conceived of the study and drafted the manuscript. SVDV, SC and FVB carried out the work. All authors read the manuscript and approved the final draft.