Phosphoglucomutase of Yersinia pestis is required for autoaggregation and polymyxin B resistance

Abstract

Yersinia pestis, the causative agent of plague, autoaggregates within a few minutes of cessation of shaking when grown at 28 degrees C. To identify the autoaggregation factor of Y. pestis, we performed mariner-based transposon mutagenesis. Autoaggregation-defective mutants from three different pools were identified, each with a transposon insertion at a different position within the gene encoding phosphoglucomutase (pgmA; y1258). Targeted deletion of pgmA in Y. pestis KIM5 also resulted in loss of autoaggregation. Given the previously defined role for phosphoglucomutase in antimicrobial peptide resistance in other organisms, we tested the KIM5 DeltapgmA mutant for antimicrobial peptide sensitivity. The DeltapgmA mutant displayed >1,000-fold increased sensitivity to polymyxin B compared to the parental Y. pestis strain, KIM5. This sensitivity is not due to changes in lipopolysaccharide (LPS) since the LPSs from both Y. pestis KIM5 and the DeltapgmA mutant are identical based on a comparison of their structures by mass spectrometry (MS), tandem MS, and nuclear magnetic resonance analyses. Furthermore, the ability of polymyxin B to neutralize LPS toxicity was identical for LPS purified from both KIM5 and the DeltapgmA mutant. Our results indicate that increased polymyxin B sensitivity of the DeltapgmA mutant is due to changes in surface structures other than LPS. Experiments with mice via the intravenous and intranasal routes did not demonstrate any virulence defect for the DeltapgmA mutant, nor was flea colonization or blockage affected. Our findings suggest that the activity of PgmA results in modification and/or elaboration of a surface component of Y. pestis responsible for autoaggregation and polymyxin B resistance.

FIG. 1.

PgmA is required for Y. pestis autoaggregation. (A to D) Overnight cultures grown at 28°C were transferred to a test tube and allowed to sit, without shaking, for autoaggregation at room temperature. The OD₆₂₀ values of the tubes were recorded at the indicated time points. The results are from duplicate assays of three independent experiments (n = 6). (E) Expression levels of PgmA-HA or PgmA-HA catalytic mutants (S160A and R521S) were determined by Western blotting using an anti-HA tag antibody directed against the C terminus of the hybrid protein. (F) Autoaggregation of identical numbers of Y. pestis KIM5 and the ΔpgmA mutant after a 45-min incubation at room temperature. Defects in autoaggregation were restored upon complementation with plasmid-encoded PgmA.

FIG. 2.

PGM and PMM activities of the KIM5 ΔpgmA mutant. PGM assays were recorded 5 min after addition of α-glucose-1-phosphate as the assay reached saturation. PMM assays were recorded at 60 min after the addition of α-mannose-1-phosphate when the reaction had reached saturation. The ΔpgmA mutant was complemented with PgmA expressed from plasmid pMMB207 or HA-tagged wild-type PgmA or catalytic mutants of PgmA expressed from the plasmid pMMB208.

FIG. 3.

DOC-PAGE analysis of purified Y. pestis LPS samples by the phenol-chloroform extraction method. For each strain 0.1, 0.2, and 0.5 μg of LPS were loaded. Lanes 2 to 4, KIM5; lanes 5 to 7; KIM5 ΔpgmA. In lane 1, 2 μg of LPS from Salmonella enterica serovar Minnesota served as a control for an O-antigen-containing LPS molecule.

FIG. 4.

Composition analysis of the LPS from Y. pestis KIM5 (A) and a Y. pestis ΔpgmA mutant (B). The analysis was performed by preparation and gas chromatography (GC)-MS analysis of trimethylsilyl methyl glycosides. The analysis reveals the glycosyl residues as well as the fatty acids as fatty acid methyl esters. The quantification values for each residue are shown in Table 4.

FIG. 5.

The MALDI-TOF mass spectra of the lipid A from Y. pestis KIM5 (A) and its ΔpgmA mutant (B). Both spectra were acquired in the negative reflector mode, and observed ions are (M-H)⁻ deprotonated and monosodiated forms of lipid A (Table 5 gives details). Representative structures of the three identified lipid A molecules are provided: hexa-acylated, penta-acylated, and tetra-acylated (C).

FIG. 6.

Neutralization of LPS toxicity by polymyxin B. RAW264.7 murine macrophages were incubated with 200 ng/ml LPS and 10-fold increasing concentrations of polymyxin B. NO₂⁻ was detected in culture supernatants by a Griess reaction assay after 20 h of incubation. Each point represents averages and standard deviations of two independent experiments, with triplicate assays in each experiment.