Defining the Roles of Pyruvate Oxidation, TCA Cycle, and Mannitol Metabolism in Methicillin-Resistant Staphylococcus aureus Catheter-Associated Urinary Tract Infection

Abstract

Methicillin-resistant Staphylococcus aureus (MRSA) is an important cause of complicated urinary tract infection (UTI) associated with the use of indwelling urinary catheters. Previous reports have revealed host and pathogen effectors critical for MRSA uropathogenesis. Here, we sought to determine the significance of specific metabolic pathways during MRSA UTI. First, we identified four mutants from the Nebraska transposon mutant library in the MRSA JE2 background that grew normally in rich medium but displayed significantly reduced growth in pooled human urine (HU). This prompted us to transduce the uropathogenic MRSA 1369 strain with the transposon mutants in sucD and fumC (tricarboxylic acid [TCA] cycle), mtlD (mannitol metabolism), and lpdA (pyruvate oxidation). Notably, sucD, fumC, and mtlD were also significantly upregulated in the MRSA 1369 strain upon exposure to HU. Compared to the WT, the MRSA 1369 lpdA mutant was significantly defective for (i) growth in HU, and (ii) colonization of the urinary tract and dissemination to the kidneys and the spleen in the mouse model of catheter-associated UTI (CAUTI), which may be attributed to its increased membrane hydrophobicity and higher susceptibility to killing by human blood. In contrast to their counterparts in the JE2 background, the sucD, fumC, and mtlD mutants in the MRSA 1369 background grew normally in HU; however, they displayed significant fitness defects in the CAUTI mouse model. Overall, identification of novel metabolic pathways important for the urinary fitness and survival of MRSA can be used for the development of novel therapeutics. IMPORTANCE While Staphylococcus aureus has historically not been considered a uropathogen, S. aureus urinary tract infection (UTI) is clinically significant in certain patient populations, including those with chronic indwelling urinary catheters. Moreover, most S. aureus strains causing catheter-associated UTI (CAUTI) are methicillin-resistant S. aureus (MRSA). MRSA is difficult to treat due to limited treatment options and the potential to deteriorate into life-threatening bacteremia, urosepsis, and shock. In this study, we found that pathways involved in pyruvate oxidation, TCA cycle, and mannitol metabolism are important for MRSA fitness and survival in the urinary tract. Improved understanding of the metabolic needs of MRSA in the urinary tract may help us develop novel inhibitors of MRSA metabolism that can be used to treat MRSA-CAUTI more effectively.

Keywords: Krebs cycle; MRSA; TCA cycle; mannitol metabolism; metabolism; methicillin-resistant Staphylococcus aureus; pyruvate metabolism; urinary tract infection.

FIG 1

Growth curves of MRSA 1369 WT and select mutants. MRSA 1369 WT and mutant strains were grown at 37°C, static, for 24 h in either nutrient-rich BHI (A and C) or pooled human urine (B and D), and CFU per milliliter were determined by dilution plating on TSA at specific time points. The growth of WT and mutant MRSA strains are presented as growth curves showing average CFU per milliliter ± standard error of the mean (A and B) and as average doubling time ± standard error of mean (C and D). The data are from 2 biological replicates for BHI and from 3 to 4 biological replicates for human urine; each biological replicate had 2 technical replicates. The doubling time for each mutant was compared with the WT using unpaired t test. ***, P ≤ 0.001 compared to WT control.

FIG 2

Schematic showing the steps in metabolic pathways catalyzed by fumC, sucD, mtlD, and lpdA.

FIG 3

Competition between WT and mutant MRSA 1369 strains for in vitro growth in pooled human urine (HU). WT MRSA 1369 was separately cocultivated with each gene mutant at a 1:1 ratio in pooled HU at 37°C, static. The inoculum CFU as well as the CFU recovered at 4 h and 24 h were enumerated by dilution plating. The competitive index (CI) for each mutant strain at 4 h (A) and 24 h (B) was calculated as CI = (WT recovered/mutant recovered)/(WT inoculum/mutant inoculum). CI values are shown as scatterplots with each point representing a technical replicate and median as the measure of central tendency. A CI of 1, shown as a dotted line, represents that WT and knockout (KO) are equally competitive, a CI of <1 denotes that the KO has a competitive advantage over WT, and a CI of >1 indicates that the WT has a competitive advantage over the KO. Data from 3 or 4 biological replicates, each with 2 technical replicates, are shown. Data were compared against a theoretical median of 1 using one sample t test and Wilcoxon test. **, P ≤ 0.01; *, P ≤ 0.05.

FIG 4

In vivo competition examining the fitness of MRSA 1369 mutants compared to the WT. C57BL/6 female mice were catheterized and then inoculated transurethrally with a 1:1 mixture of MRSA 1369 WT and either sucD (A and B), fumC (C and D), mtlD (E and F), or lpdA (G and H) mutants. At 24 hpi, WT and mutant CFU burden in bladder, kidneys, and catheter were determined as shown in panels B, D, F, and H. Competitive index (CI) values for the urinary bladder, kidneys, and catheter from each mouse are presented as scatter diagrams with median as the measure of central tendency as shown in panels A, C, E, and G. A CI of 1, shown as a dotted line, represents that WT and mutant are equally competitive, a CI of <1 denotes that the mutant has a competitive advantage over WT, and a CI of >1 indicates that the WT has a competitive advantage over mutant. The CI data were compared against a theoretical median of 1 using one sample t test and Wilcoxon test, and the organ burden data were compared using Wilcoxon matched-pairs signed-rank test; ***, P ≤ 0.001; **, P ≤ 0.01; *, P ≤ 0.05; ns, not significant.

FIG 5

Urinary pathogenesis of lpdA in an in vivo mouse model of CAUTI. C57BL/6 female mice were inoculated with WT or lpdA. At 24 hpi, WT and ΔlpdA CFU in the spleen, bladder, kidneys, and catheter were determined. The data are presented as scatter diagrams showing organ burden from an individual mouse and median as the central tendency. The dotted line represents a limit of detection (LOD) of 10 CFU/mL. Data from 8 mice per group from ≥2 biological replicates were compared using Mann-Whitney U test; ***, P ≤ 0.001; **, P ≤ 0.01; *, P ≤ 0.05; ns, not significant.

FIG 6

mtlD and lpdA susceptibility to H₂O₂, cell surface hydrophobicity, and in vitro killing by whole blood. WT, mtlD, and lpdA were exposed to 15 mM H₂O₂ for 4 h in either BHI (A) or pooled human urine (HU) (B). The inoculum (0 h) and surviving CFU after 4-h-long exposure to H₂O₂ were determined. Additionally, following 2 h exposure to urine, WT and lpdA were compared for cell surface hydrophobicity(C) and percent killing in human blood (D). Scatterplots show individual technical replicates from 3 biological replicates with histograms representing the average. The data for each mutant were compared with the WT (at the 1-h or 4-h time point for percent killing in human blood) using unpaired t test; **, P ≤ 0.01; *, P ≤ 0.05; ns, not significant.