Listeria monocytogenes requires phosphotransferase systems to facilitate intracellular growth and virulence

Abstract

The metabolism of bacterial pathogens is exquisitely evolved to support virulence in the nutrient-limiting host. Many bacterial pathogens utilize bipartite metabolism to support intracellular growth by splitting carbon utilization between two carbon sources and dividing flux to distinct metabolic needs. For example, previous studies suggest that the professional cytosolic pathogen Listeria monocytogenes (L. monocytogenes) utilizes glycerol and hexose phosphates (e.g., Glucose-6-Phosphate) as catabolic and anabolic carbon sources in the host cytosol, respectively. However, the role of this putative bipartite metabolism in L. monocytogenes virulence has not been fully assessed. Here, we demonstrate that when L. monocytogenes is unable to consume either glycerol (ΔglpD/ΔgolD), hexose phosphates (ΔuhpT), or both (ΔglpD/ΔgolD/ΔuhpT), it is still able to grow in the host cytosol and is 10- to 100-fold attenuated in vivo suggesting that L. monocytogenes consumes alternative carbon source(s) in the host. An in vitro metabolic screen using BioLog's phenotypic microarrays unexpectedly demonstrated that WT and PrfA* (G145S) L. monocytogenes, a strain with constitutive virulence gene expression, use phosphotransferase system (PTS) mediated carbon sources. These findings contrast with the existing metabolic model that cytosolic L. monocytogenes expressing PrfA does not use PTS mediated carbon sources. We next demonstrate that two independent and universal phosphocarrier proteins (PtsI [EI] and PtsH [HPr]), essential for the function of all PTS, are critical for intracellular growth and virulence in vivo. Constitutive virulence gene expression using a PrfA* (G145S) allele in ΔglpD/ΔgolD/ΔuhpT and ΔptsI failed to rescue in vivo virulence defects suggesting phenotypes are due to metabolic disruption and not virulence gene regulation. Finally, in vivo attenuation of ΔptsI and ΔptsH was additive to ΔglpD/ΔgolD/ΔuhpT, suggesting that hexose phosphates and glycerol and PTS mediated carbon source are relevant metabolites. Taken together, these studies indicate that PTS are critical virulence factors for the cytosolic growth and virulence of L. monocytogenes.

Fig 1. Mutants defective for consumption of glycerol (ΔglpD/ΔgolD) and/or hexose phosphates (ΔuhpT) require alternative carbon sources to support growth in a defined medium.

(A) Simplified model of glycerol being imported and funneled into two parallel glycerol utilization pathways (glpD and golD) for entry into central catabolic glycolysis. Hexose phosphate imported and funneled into the anabolic pentose phosphate pathway. Indicated strains were grown in LSM at 37°C, shaking at 250 r.p.m. with the addition of 55mM glucose (B) or carbon equivalent amounts of glycerol (C), hexose phosphates (+10mM glutathione) (D), and glycerol and hexose phosphates (+10mM glutathione) (E). OD₆₀₀ was monitored every 15 minutes for 24 hours. Data represents average of three technical replicates from one representative of three biological replicates.

Fig 2. Mutants defective for metabolism of glycerol (ΔglpD/ΔgolD) and/or hexose phosphates (ΔuhpT) are readily able to grow in the host cytosol and maintain virulence.

(A) Intracellular growth of WT, ΔglpD/ΔgolD, ΔuhpT, ΔglpD/ΔgolD/ΔuhpT was determined in BMDMs following infection at an MOI of 0.2. Growth curves are representative of at least three independent experiments. Error bars represent the standard deviation of the means of technical triplicates within the representative experiment. (B) L2 fibroblasts were infected with indicated L. monocytogenes strains at an MOI of 0.5 and were examined for plaque formation 4 days post infection. Assays were performed in biological triplicate and data displayed is the median and SEM of a strain’s plaque size relative to WT in one of three representative biological replicates. (C) Bacterial burdens from the spleen and liver were enumerated at 48 hours post-intravenous infection with 1x10⁵ bacteria. Data are representative of results from two separate experiments. Horizontal dashed lines represent the limits of detection, and the bars associated with the individual strains represents the mean and SEM of the group.

Fig 3. Carbon metabolite respiration of WT, ΔglpD/ΔgolD/ΔuhpT, and PrfA* L. monocytogenes is globally similar, identified using Biolog’s Phenotypic Microarrays.

Clustered heatmaps indicating level of tetrazolium dye color change as measured by OD₄₉₀ at 48 hours in response to ΔglpD/ΔgolD/ΔuhpT (Top), WT (Middle), and PrfA* (Bottom) respiration of carbon metabolites (PM1 & PM2A) at 37°C stationary. Each bar indicates the average of 3 biologic replicates. Samples were normalized to readings of a α-D-glucose control (~1 on scale and labeled) and sorted based on cluster analysis. Select differentially used metabolites from Tables 1 and 2 of hexose phosphates, glycerol, and lactose are labeled above.

Fig 4. ΔptsI mutants are impaired for intramacrophage growth and virulence, with more decreased virulence in a ΔglpD/ΔgolD/ΔuhpT background and can be trans-complemented with ptsI over expression for intracellular growth and virulence.

(A) PTS mediated free sugar import and phosphorylation by phosphocarrier protein phospho-cycling from the terminal conversion of phosphoenol-pyruvate (PEP) to pyruvate. (B) Intracellular growth of WT, ΔglpD/ΔgolD/ΔuhpT, ΔptsI, and ΔglpD/ΔgolD/ΔuhpT/ΔptsI was determined in BMDMs following infection at an MOI of 0.2. Growth curves are representative of at least three independent experiments. Error bars represent the standard deviation of the means of technical triplicates within the representative experiment. (C) L2 fibroblasts were infected with indicated L. monocytogenes strains at an MOI of 0.5 and were examined for plaque formation 4 days post infection. Assays were performed in biological triplicate and data displayed is the median and SEM of a strain’s plaque size relative to WT in one of three representative biological replicates. (D) Bacterial burdens from the spleen and liver were enumerated at 48 hours post-intravenous infection with 1x10⁵ bacteria. Data are representative of results from two separate experiments. Horizontal dashed lines represent the limits of detection, and the bars associated with the individual strains represents the mean and SEM of the group.

Fig 5. ΔglpD/ΔgolD/ΔuhpT and ΔptsI mutants are not rescued by constitutively active PrfA and show basally higher levels of virulence protein activity.

(A) Bacterial burdens from the spleen and liver were enumerated at 48 hours post-intravenous infection with 1x10⁵ bacteria. Data are representative of results from two separate experiments. Horizontal dashed lines represent the limits of detection, and the bars associated with the individual strains represents the mean and SEM of the group. (B) Median and SEMs of percent hemolysis interpolated at OD600 of 0.025 from biological triplicate for displayed strains. (C) Representative toxin dose-response curves from a single biological replicate used for interpolation and generation of continuous percent hemolysis data as a function of OD600.

Fig 6. ΔptsH mutants are more attenuated for intracellular growth and virulence than in ΔptsI in all backgrounds.

(A) Intracellular growth of wild-type, ΔglpD/ΔgolD/ΔuhpT, ΔptsH, and ΔglpD/ΔgolD/ΔuhpT/ΔptsH was determined in BMDMs following infection at an MOI of 0.2. Growth curves are representative of at least three independent experiments. Error bars represent the standard deviation of the means of technical triplicates within the representative experiment. (B) Bacterial burdens from the spleen and liver were enumerated at 48 hours post-intravenous infection with 1x10⁵ bacteria. Data are representative of results from two separate experiments. Horizontal dashed lines represent the limits of detection, and the bars associated with the individual strains represents the mean and SEM of the group.