Disruption of Glycolysis by Nutritional Immunity Activates a Two-Component System That Coordinates a Metabolic and Antihost Response by Staphylococcus aureus

Abstract

During infection, bacteria use two-component signal transduction systems to sense and adapt to the dynamic host environment. Despite critically contributing to infection, the activating signals of most of these regulators remain unknown. This also applies to the Staphylococcus aureus ArlRS two-component system, which contributes to virulence by coordinating the production of toxins, adhesins, and a metabolic response that enables the bacterium to overcome host-imposed manganese starvation. Restricting the availability of essential transition metals, a strategy known as nutritional immunity, constitutes a critical defense against infection. In this work, expression analysis revealed that manganese starvation imposed by the immune effector calprotectin or by the absence of glycolytic substrates activates ArlRS. Manganese starvation imposed by calprotectin also activated the ArlRS system even when glycolytic substrates were present. A combination of metabolomics, mutational analysis, and metabolic feeding experiments revealed that ArlRS is activated by alterations in metabolic flux occurring in the latter half of the glycolytic pathway. Moreover, calprotectin was found to induce expression of staphylococcal leukocidins in an ArlRS-dependent manner. These studies indicated that ArlRS is a metabolic sensor that allows S. aureus to integrate multiple environmental stresses that alter glycolytic flux to coordinate an antihost response and to adapt to manganese starvation. They also established that the latter half of glycolysis represents a checkpoint to monitor metabolic state in S. aureus Altogether, these findings contribute to understanding how invading pathogens, such as S. aureus, adapt to the host during infection and suggest the existence of similar mechanisms in other bacterial species.IMPORTANCE Two-component regulatory systems enable bacteria to adapt to changes in their environment during infection by altering gene expression and coordinating antihost responses. Despite the critical role of two-component systems in bacterial survival and pathogenesis, the activating signals for most of these regulators remain unidentified. This is exemplified by ArlRS, a Staphylococcus aureus global regulator that contributes to virulence and to resisting host-mediated restriction of essential nutrients, such as manganese. In this report, we demonstrate that manganese starvation and the absence of glycolytic substrates activate ArlRS. Further investigations revealed that ArlRS is activated when the latter half of glycolysis is disrupted, suggesting that S. aureus monitors flux through the second half of this pathway. Host-imposed manganese starvation also induced the expression of pore-forming toxins in an ArlRS-dependent manner. Cumulatively, this work reveals that ArlRS acts as a sensor that links nutritional status, cellular metabolism, and virulence regulation.

Keywords: ArlRS; Staphylococcus aureus; calprotectin; glycolysis; manganese; nutritional immunity.

FIG 1

ArlRS is activated in response to reduced levels of intracellular Mn. (A) Wild-type bacteria and a ΔarlR mutant were grown in TSB with various concentrations of CP. ArlRS activity was assessed by measuring the activity of the mgrA P2 promoter using a YFP-reporter construct at t = 6 h. *, P ≤ 0.05 (relative to the same strain without CP by two-way analysis of variance [ANOVA] with Dunnett’s multiple-comparison test). #, P ≤ 0.05 (relative to wild-type [WT] bacteria at the same CP concentration by two-way ANOVA with Sidak’s multiple-comparison test). RFU, relative fluorescence units. (B) Wild-type S. aureus and a ΔmntC ΔmntH mutant were grown in TSB with 480 μg/ml WT CP or the ΔSI, ΔSII, or ΔSIΔSII mutant. ArlRS activity was assessed by measuring the activity of the mgrA P2 promoter using a YFP-reporter construct at t = 9 h. *, P ≤ 0.05 (relative to wild-type CP in the same strain by two-way ANOVA with Dunnett’s multiple-comparison test). (C and D) Wild-type bacteria and a ΔmntC ΔmntH mutant were grown in TSB with increasing concentrations of CP. The activity of the mgrA P2 (C) and mntA (D) promoters was assessed using YFP-reporter plasmids at t = 9 h. *, P ≤ 0.05 (relative to wild-type bacteria at the same CP concentration by two-way ANOVA with Sidak’s multiple-comparison test). #, P ≤ 0.05 (relative to the same strain in the absence of CP by two-way ANOVA with Dunnett’s multiple-comparison test). (E and F) Wild-type S. aureus and the ΔmntC ΔmntH mutant were grown in chemically defined medium (CDM) supplemented with various concentrations of manganese, and activity of the mgrA P2 (E) and mntA (F) promoters was assessed using a YFP-reporter plasmid at t = 9 h. *, P ≤ 0.05 (relative to wild-type bacteria at the same Mn concentration by two-way ANOVA with Sidak’s multiple-comparison test). #, P ≤ 0.05 (relative to the same strain in the presence of 1 μM Mn by two-way ANOVA with Dunnett’s multiple-comparison test). n ≥ 3 (all panels). Error bars indicate standard errors of the means (SEM).

FIG 2

Pyruvate indirectly activates ArlRS. (A) Wild-type S. aureus was grown in yeast complete (YC) medium supplemented with various concentrations of pyruvate (pyr) or glucose (glc). ArlRS activity was assessed by measuring the activity of the mgrA P2 promoter using a YFP-reporter plasmid at t = 6 h. *, P ≤ 0.05 (relative to bacteria grown in the absence of pyruvate by one-way ANOVA with Dunnett’s multiple-comparison test). (B) Wild-type S. aureus was grown in TSB medium in the presence and absence of 240 μg/ml CP, and pyruvate levels were assessed. *, P ≤ 0.05 (relative to bacteria grown without CP by unpaired two-tailed t test). (C) Wild-type S. aureus was grown in YC medium supplemented with 0.25% glucose or 2% pyruvate, and 2-phosphoglycerate levels were assessed. *, P ≤ 0.05 (relative to bacteria grown in YC medium alone by one-way ANOVA with Dunnett’s multiple-comparison test). n ≥ 3 (all panels). Error bars indicate SEM.

FIG 3

The absence of glucose and manganese starvation activate ArlRS. Wild-type S. aureus was grown in TSB (A) or LB (B) supplemented with 0.25% glucose (+glc) or without glucose (no glc) as indicated. ArlRS activity was assessed by measuring the activity of the mgrA P2 promoter using a YFP-reporter plasmid at t = 6 h. *, P ≤ 0.05 (relative to bacteria grown without glucose at the same CP concentration by two-way ANOVA with Sidak’s multiple-comparison test). #, P ≤ 0.05 (relative to bacteria grown in the same medium without CP by two-way ANOVA with Dunnett’s multiple-comparison test). n ≥ 3. Error bars indicate SEM.

FIG 4

Disruption of the latter half of glycolysis activates ArlRS. (A) Simplified representation of glycolysis indicating where fructose and glycerol feed into the pathway. (B) Wild-type S. aureus Newman was grown in TSB without glucose and supplemented with carbon-balanced concentrations of glucose (glc), fructose (fru), glycerol (glyc), or pyruvate (pyr). (B to D) ArlRS activity was assessed by measuring the activity of the mgrA P2 promoter using a YFP-reporter plasmid at t = 6 h. (B) *, P ≤ 0.05 (relative to medium without glucose by one-way ANOVA with Dunnett’s multiple-comparison test). (C and D) Wild-type S. aureus Newman (C) and USA300 JE2 (D) were grown in TSB without glucose and without supplementation (−) or supplemented with carbon-balanced concentrations of glucose (glc), glycerol (glyc), and/or pyruvate (pyr) in the absence and presence of 480 μg/ml CP. *, P ≤ 0.05 (relative to growth in medium without glucose at the same CP concentration by two-way ANOVA with Dunnett’s multiple-comparison test). #, P ≤ 0.05 (relative to growth in the same carbon source without CP by two-way ANOVA with Sidak’s multiple-comparison test). n ≥ 3 (panels B to D). Error bars indicate SEM.

FIG 5

Relative levels of glycolytic intermediates in cells treated with CP or deprived of glucose. S. aureus was grown either in the presence of 120 μg/ml CP or without glucose, and the relative levels of glycolytic intermediates were measured using mass spectrometry as described in Materials and Methods. The level of each metabolite was normalized to the total ion current of the 86 metabolites detected. Each dot represents an independent measurement, and means are indicated by horizontal lines.

FIG 6

Accumulation of 3-PG or of other upstream intermediates activates ArlRS. (A and B) Wild-type S. aureus was grown in TSB without glucose (no glc) supplemented with 0.25% glucose (glc) (A) or in glucose-containing TSB in the absence and presence of 480 μg/ml CP (B), and 2-phosphoglycerate levels were assessed. *, P ≤ 0.05 (relative to bacteria grown in the absence of either glucose [A] or CP [B] by unpaired two-tailed t test). (C) Wild-type, ΔgpmI, and ΔgpmA S. aureus were grown in TSB in the absence and presence of 480 μg/ml CP. ArlRS activity was assessed by measuring the activity of the mgrA P2 promoter using a YFP-reporter plasmid at t = 6 h. *, P ≤ 0.05 (relative to wild-type bacteria at the same CP concentration by two-way ANOVA with Dunnett’s multiple-comparison test). #, P ≤ 0.05 (relative to the same strain in the absence of CP by two-way ANOVA with Sidak’s multiple-comparison test). n ≥ 3 (panels A to C). Error bars indicate SEM.

FIG 7

CP induces ArlRS-dependent expression of pore-forming leukocidins. (A to D) Wild-type and ΔarlR S. aureus USA300 JE2 were grown in TSB without glucose (no glc), supplemented with 0.25% glucose (glc) (A and B) or glucose-containing TSB in the presence of increasing concentrations of CP (C and D). Leukocidin expression was assessed by measuring the activity of the lukE (A and C) and lukS-PVL (B and D) promoters using a Lux-reporter plasmid at t = 6 h. (A to D) *, P ≤ 0.05 (relative to wild-type bacteria under each condition by two-way ANOVA with Sidak’s multiple-comparison test). (C and D) #, P ≤ 0.05 (relative to the same strain in the absence of CP by two-way ANOVA with Dunnett’s multiple-comparison test). n ≥ 3 (panels A to D). Error bars indicate SEM.

FIG 8

Model of ArlRS activation. When glycolytic carbon sources (such as glucose and glycerol) and manganese are abundant in the host environment, ArlRS remains inactive. However, in an environment poor in glycolytic carbon sources or under conditions of host-imposed manganese starvation, both stimuli (which alter flux in the latter half of glycolysis) elicit the activation of ArlRS. Activation of this TCS allows S. aureus to coordinate an antihost response consisting of the production of virulence factors such as leukocidins and results in reprogramming of staphylococcal metabolism in response to nutritional immunity.