Fig 1.
Summary of the oxidative phase of the pentose phosphate pathway including 6-phosphogluconolactonase (Pgl), that converts 6-P-gluconololactone to gluconate-6-P.
For reference, key glycolysis, TCA cycle, nucleotide and cell wall biosynthetic pathway intermediates are also shown. Fructose-6-P is fluxed from glycolysis to peptidoglycan (PG), wall teichoic acid (WTA) and lipoteichoic acid (LTA) via UDP-N-acetylglucosamine (UDP-GlcNAc) and UDP-N-acetylmuramic acid (UDP-MurNAc). Fosfomycin (FOS) targets MurA which together with MurB is required for the conversion of UDP-GlcNAc to UDP-MurNAc. Oxacillin (OX) targets the transpeptidase activity of the penicillin binding proteins required for PG crosslinking. The putative gluconate shunt involves the export of 6-phosphogluconolactone, which spontaneously degrades to gluconate before being transported into the cell by the gluconate permease GntP and phosphorylated by the gluconate kinase GntK. Schematic made using Biorender.com.
Table 1.
Minimum inhibitory concentrations (μg/ml; % for Congo Red) of strains used in this study to oxacillin (OX), targocil (TG), tunicamycin (TM), fosfomycin (FOS), D-cycloserine (DCS), Congo Red (CR), vancomycin (VAN), amsacrine (AMS), sulfamethoxazole (SMX) and polymyxin B (PMB) in Mueller Hinton Broth (+ 2% NaCl for OX).
Fig 2.
Mutation of pgl increases resistance to oxacillin.
A. Oxacillin MICs of JE2, pgl and the complemented pgl mutant in Mueller Hinton broth with 2% NaCl (MHB), chemically defined media (CDM) and CDM with glucose (CDMG). Note that the Y axis (Oxacillin MIC) is a log2 scale. B and C. Growth of JE2, pgl and pglcomp for 25 hrs at 35°C in CDM (B) and CDMG (C) supplemented with OX 10 μg/ml. Growth (OD600) was measured at 15 min intervals in a Tecan plate reader. Data are the average of 3 independent experiments and error bars represent standard deviation.
Fig 3.
Mutation of pgl reduces cell size and prevents OX-induced cell lysis in CDMG.
A and B. Representative microscopic images of JE2, pgl and pglcomp cells grown in CDMG (A) or CDMG supplemented with OX 0.05 μg/ml (B) and labelled with vancomycin BODIPY FL, which binds to the terminal d-ala-d-ala in the peptidoglycan stem peptide (green, top panel) or WGA Alexa Fluor 594, which binds to GlcNAc and other sugars in the cell envelope (red, bottom panel). C. Average diameter of JE2, pgl and pglcomp cells grown in CDMG or CDMG OX. Images of cells from four biological replicates were acquired using Fv3000 confocal microscope and software, 50 cells measured per biological replicate (200 cells in total) for CDMG and 60 cells in total counted for CDMG OX (due to cell lysis), and the violin plots for the four biological replicates were generated using GraphPad Prism V9. Asterisks indicate statistically significant difference according to using a Kruskal-Wallis test followed by a Dunn’s multiple comparison test. Adjusted p-values * p<0.05, *** p<0.001 and **** p<0.0001 are indicated. D. Extensive lysis of JE2 and pglcomp (but not pgl) in CDMG OX 0.05 μg/ml cultures. Cells were labelled with WGA Alexa Fluor 594 and representative microscopic images are shown.
Fig 4.
PPP activity is impaired in the pgl mutant.
A. JE2, pgl and the complemented pgl mutant (pglcomp) were grown in CDM [1,2-13C]Glucose and fluxes via glycolysis and the pentose phosphate pathway (PPP) were compared as described previously [64]. The M+2 pyruvate is unique to glycolysis and the M+1 pyruvate to PPP. Thus, the M+2/M+1 ratio is indicative of carbon flux through glycolysis relative to PPP. The M+0 pyruvate can arise from different sources including the unlabeled part of the [1,2-13C]Glucose and pyruvogenic amino acids that are consumed alongside glucose. B. Relative levels of M+1 pyruvate indicative of PPP activity and M+2 pyruvate indicative of glycolytic activity in JE2, pgl and pglcomp. C. The M+2/M+1 ratio indicative of pyruvate produced directly from glucose flux through glycolysis in JE2, pgl and pglcomp. Data are the average of three independent experiments and standard deviations are shown. Significant differences were determined using ordinary one-way ANOVA with Dunnett’s multiple comparison using GraphPad Prism V9 and adjusted p-value **** p<0.0001 is indicated.
Fig 5.
Heatmap comparison of cell wall, pentose phosphate pathway (PPP)/glycolysis, TCA cycle, redox, nucleotides and amino acid metabolites in JE2, pgl and pglcomp.
Whole cell metabolomics was performed on JE2, pgl and pglcomp grown in CDMG and CDMG OX 10 μg/ml and the cells collected after 4–5 hours (early exponential phase). Data presented are the average of three biological replicates (2 biological replicates for FAD) analysed using GraphPad Prism V9. Individual metabolite levels that were significantly different using a one-way ANOVA with Turkey’s post-hoc in pgl grown in CDMG, CDMG OX or both are highlighted in bold text. * significant difference in either CDMG or CDMG OX. ** significant difference in both CDMG and CDMG OX.
Fig 6.
Increased OX resistance in the pgl mutant is associated with a ruffled surface morphology, a thicker cell wall and thicker septa between dividing cells.
Transmission electron microscopy at 30,000× (left) and 100,000× (right) magnification was performed on JE2 (A), pgl (B) and pglcomp (C) cells collected from exponential phase cultures grown for 4.5 h in CDMG OX 1 μg/ml normalized to OD600 = 1 in PBS before being fixed and thin sections prepared. Representative cells from each strain are shown. Scale bars represent 500 nm at 30,000× or 100 nm at 100,000× magnification.
Fig 7.
Mutation of vraG restores wild-type OX resistance, but not cell size, in the pgl mutant grown in CDMG.
A and B. Growth of JE2, pgl::Kmr, pglR1, putA, graR, vraG, pgl/putA, pgl/graR, pgl/vraG, pgl/putA/vraG for 25 hrs at 35°C in CDMG (A) and CDMG supplemented with OX 10 μg/ml (B). Growth (OD600) was measured at 15 min intervals in a Tecan plate reader. Data are the average of 3 independent experiments and error bars represent standard deviation. C. Representative microscopic images of JE2, pgl, pglcomp, pgl::Kmr, pglR1, vraG and pgl/vraG cells grown in CDMG and labelled with vancomycin BODIPY FL (green, top panel) or WGA Alexa Fluor 594 (red, bottom panel). D. Average diameter of JE2, pgl::Kmr, pglR1, vraG and pgl/vraG cells grown in CDMG. Images of cells from three biological replicates were acquired using Fv3000 confocal microscope and software, 50 cells measured per biological replicate (150 cells in total) and the violin plots for the three biological replicates were generated using GraphPad Prism V9. Asterisks indicate statistically significant difference according to using a Kruskal-Wallis test followed by a Dunn’s multiple comparison test. Adjusted p-values **** p<0.0001 or ns, not significant are indicated.
Fig 8.
Lipoteichoic acids are reduced and the cell surface is more positively charged in the pgl mutant.
A. Comparison of wheat germ agglutinin (WGA) Alexa Fluor 594 binding to JE2, pgl, pglcomp, pgl::Kmr, vraG, vraF, graR, putA, pgl/vraG, pgl/vraF, pgl/graR, pgl/putA, pgl/putA/vraG and pglR1 cells grown for 4.5 h in CDMG OX using fluorescence microscopy at 594nm excitation/618nm detection. The data are the average of 2 independent experiments and error bars represent standard deviation. Significant differences were determined using a two-way ANOVA with Turkey’s post-hoc analysis. Adjusted p-values **** p<0.0001 or ns, not significant are indicated. B. Comparison of relative wall teichoic acid (WTA) levels in JE2, pgl, pglcomp, vraG, pgl/vraG and tagO (negative control) grown in CDMG and CDMG OX 0.05 mg/ml. Purified WTA samples were separated on 20% native polyacrylamide (PAA) gels before being stained with 0.1% Alcian blue. A representative image from 3 independent biological repeats is shown. C. Comparison of relative lipoteichoic acid (LTA) levels in JE2, pgl, pglcomp, pgl::Kmr, vraG, pgl/vraG, USA300 LAC* and ltaS/gdpP (negative control) grown in CDMG and CDMG OX 0.05 mg/ml (the ltaS/gdpP mutant is OX susceptible and was only grown in CDMG). Extracted LTAs were separated on 15% PAA gels, transferred to a PVDF membrane and probed with LTA antibody (1:5000), followed by HRP-conjugated protein G (1:2000) and colorimetric detection with Opti-4CN substrate kit. Three independent experiments were performed, and a representative blot is shown. D. Mutation of vraG partially restores Congo Red resistance in the pgl mutant. 10-fold serial dilutions of JE2, pgl, pglcomp, vraG and pgl/vraG inoculated onto TSA supplemented with 0.125% Congo Red and grown for 24 h at 37°C. This experiment was repeated three times and a representative plate is shown. E. Comparison of cell surface charge using a cyctochrome c binding assay in JE2, pgl, vraG and pgl/vraG grown in CDMG and CDMG OX 0.05 mg/ml. Positively charged cytochrome c binds more strongly to negatively charged cells. The data are the average of 3 independent experiments and error bars represent standard deviation Significant differences were determined using ordinary one-way ANOVA followed by Turkey’s multiple comparison post-hoc test (* p<0.05, **** p<0.0001).
Fig 9.
Suggested model for VraFG-dependent high-level β-lactam resistance in the MRSA pgl mutant.
A. Illustration of JE2 and pgl cell division during growth in CDMG OX. pgl cells are smaller than wild-type JE2 when grown in CDMG and undergo normal cell division, whereas extensive lysis is evident among wild-type cells. B and C. Illustration of peptidoglycan (PG), wall teichoic acid (WTA) and lipoteichoic acid (LTA) biosynthesis in wild-type JE2 (B) and pgl (C). Mutations in vraF, vraG and to a lesser extent graR reverse the increased OX resistance phenotype of the pgl mutant. Metabolic reprogramming in the pgl mutant increases carbon flux to cell envelope precursors and β-lactam resistance via a mechanism dependent on VraFG/GraRS-controlled regulation of WTA/LTA biosynthesis, export or posttranslational modification. Previous studies have implicated the VraFG/GraRS complex in resistance to cationic antimicrobial peptides and regulation of dltABCD and mprF transcription, and it has also been proposed to play a role in the export of peptidoglycan or teichoic acid precursors or modifying subunits. Reduced levels of LTAs in pgl (C) compared to JE2 (B) may contribute to reduced cell lysis under OX stress. The significantly increased positive charge in the pgl mutant (C) compared to JE2 (B) was partially reversed by the vraG mutation suggesting that reduced levels of LTAs and increased VraFG/GraRS-dependent d-alanylation of WTAs and LTAs in the pgl mutant combine to increase OX resistance. Figure made using Biorender.com.