A PBP 2 mutant devoid of the transpeptidase domain abolishes spermine-β-lactam synergy in Staphylococcus aureus Mu50

Abstract

Exogenous spermine was reported to enhance the killing of methicillin-resistant Staphylococcus aureus (MRSA) by β-lactams through a strong synergistic effect of unknown nature. Spermine alone also exerts an antimicrobial activity against S. aureus in a pH-dependent manner. MIC measurements revealed stronger effects of spermine under alkaline conditions, suggesting the nucleophilic property of spermine instead of its positive charge as the cause of adverse effects. A spontaneous suppressor mutant (MuM) of MRSA Mu50 was selected for spermine resistance and conferred complete abolishment of spermine-β-lactam synergy. In comparison to that in Mu50, the spermine MIC in MuM remained constant (64 mM) at pH 6 to 8; however, MuM, a heat-sensitive mutant, also grew in a very narrow pH range. Furthermore, MuM acquired a unique phenotype of vancomycin-spermine synergy. Genome resequencing revealed a 7-bp deletion in pbpB, which results in a truncated penicillin-binding protein 2 (PBP 2) without the transpeptidase domain at the C terminus while the N-terminal transglycosidase domain remains intact. The results of fluorescent Bocillin labeling experiments confirmed the presence of this defective PBP 2 in MuM. All the aforementioned phenotypes of MuM were reverted to those of Mu50 after complementation by the wild-type pbpB carried on a recombinant plasmid. The anticipated changes in cell wall metabolism and composition in MuM were evidenced by observations that the cell wall of MuM was more susceptible to enzyme hydrolysis and that MuM exhibited a lower level of autolytic activities. Pleiotropic alterations in gene expression were revealed by microarray analysis, suggesting a remarkable flexibility of MuM to circumvent cell wall damage by triggering adaptations that are complex but completely different from that of the cell wall stress stimulon. In summary, these results reveal phenotypic changes and transcriptome adaptations in a unique pbpB mutant and provide evidence to support the idea that exogenous spermine may perturb normal cell wall formation through its interactions with PBP 2.

Fig 1

Growth inhibition by exogenous spermine. Cells from a fresh overnight seed culture were inoculated into prewarmed LB broth (pH 8.0). Spermine (10 mM) was added at the indicated time points after inoculation, and OD₆₀₀ was monitored at 1-h intervals.

Fig 2

pH-dependent effects of spermine alone (A) or in combination with oxacillin (B). MICs of spermine or oxacillin for Mu50 were determined at different pHs. For measurements of oxacillin MICs, 0.25 times the MIC (for pH 8.0 to 9.5) or 0.5 mM (for pH 6.0 to 7.5) of spermine was included.

Fig 3

ESI-MS analysis. All compounds were dissolved in ammonium acetate buffer (5 mM), and mass spectra were acquired in positive-ionization mode. (A) Spermine (Spm); (B) oxacillin (Oxa); (C) equimolar mixture of spermine and oxacillin at pH 7; (D) equimolar mixture of spermine and oxacillin at pH 9.

Fig 4

Growth behaviors of Mu50 and MuM. Growth of the cells in LB broth at 37°C with aeration was monitored by OD measurements. (A) Growth in the presence or absence of 0.4% glucose; (B) growth of MuM in LB buffered with 20 mM Tris-HCl at the indicated pH; (C) growth of Mu50 in LB buffered with 20 mM Tris-HCl at the indicated pH.

Fig 5

Identification of the pbpB lesion in spermine-resistant mutant MuM. (A) Sequence alignment of the pbpB genes from Mu50 and MuM at the mutation site. Also shown are the location of the 7-bp deletion (nt 1366 to 1372) in the pbpB gene of 2,184 bp, the amino acid sequence of the truncated PBP 2 of MuM, and the sizes and locations of the transglycosidase and transpeptidase domains of PBP 2. (B) The absence of PBP 2 in MuM and the presence of other penicillin-binding proteins were revealed by Bocillin labeling (top panel) or immunoblotting (bottom panel).

Fig 6

Analysis of cell wall hydrolysis. (A) In vitro susceptibility of cell walls. Crude cell walls of Mu50 or MuM were subjected to degradation by autolytic enzymes from S. aureus COL. Preparations of these samples are described in Materials and Methods. The experiments were reproduced several times, and data from a single representative experiment are shown. (B) Triton X-100-induced autolysis of Mu50 and MuM cells. Values indicate the average of two independent experiments. (C) Zymographic analysis of the autolysins in SDS cell extracts from Mu50 and MuM. Samples were analyzed in a 10% resolving gel containing heat-killed RN4220 cells. Lytic activity was detected by incubation of the gel with gentle agitation at 37°C in renaturing buffer. Equivalent amounts of protein samples were loaded. (D) Protein profiles in culture supernatants. Extracellular proteins from Mu50 and MuM were prepared as described in Materials and Methods. The gel was stained with Coomassie blue to detect the total secreted protein profile, and the protein band of autolysin was identified by MALDI-TOF MS-MS.