Impact of ROS-Induced Damage of TCA Cycle Enzymes on Metabolism and Virulence of Salmonella enterica serovar Typhimurium

Abstract

Salmonella enterica serovar Typhimurium (STM) is exposed to reactive oxygen species (ROS) originating from aerobic respiration, antibiotic treatment, and the oxidative burst occurring inside the Salmonella-containing vacuole (SCV) within host cells. ROS damage cellular compounds, thereby impairing bacterial viability and inducing cell death. Proteins containing iron-sulfur (Fe-S) clusters are particularly sensitive and become non-functional upon oxidation. Comprising five enzymes with Fe-S clusters, the TCA cycle is a pathway most sensitive toward ROS. To test the impact of ROS-mediated metabolic perturbations on bacterial physiology, we analyzed the proteomic and metabolic profile of STM deficient in both cytosolic superoxide dismutases (ΔsodAB). Incapable of detoxifying superoxide anions (SOA), endogenously generated SOA accumulate during growth. ΔsodAB showed reduced abundance of aconitases, leading to a metabolic profile similar to that of an aconitase-deficient strain (ΔacnAB). Furthermore, we determined a decreased expression of acnA in STM ΔsodAB. While intracellular proliferation in RAW264.7 macrophages and survival of methyl viologen treatment were not reduced for STM ΔacnAB, proteomic profiling revealed enhanced stress response. We conclude that ROS-mediated reduced expression and damage of aconitase does not impair bacterial viability or virulence, but might increase ROS amounts in STM, which reinforces the bactericidal effects of antibiotic treatment and immune responses of the host.

Keywords: aconitase; iron–sulfur cluster damage; metabolomics; oxidative stress; superoxide dismutase.

FIGURE 1

Deletion of both cytosolic superoxide dismutases impacts metabolite concentrations and enzyme abundance of TCA cycle (A), glycolysis (B), and pentose phosphate pathway (C). STM WT and STM ΔsodAB strains were grown aerobically in LB broth for 18.5 h at 37°C, harvested and proteomes or metabolomes were extracted for quantitative profiling using LC-MS^E or GC-MS, respectively. Data represent means of at least three biological replicates for metabolomics or proteomics analyses. Data were adjusted for multiple hypothesis testing and only significant differences with p < 0.05 or lower are displayed color-encoded. Oval symbols indicate relative changes in enzyme amounts detected for STM ΔsodAB compared to WT, whereas squares reflect relative changes in amounts of metabolites. Gray symbols are indifferent in enzyme or metabolite amounts.

FIGURE 2

Metabolic profile of sodAB-deficient STM compared to WT. STM WT and STM ΔsodAB were cultured aerobically in LB broth for 18.5 h at 37°C, before cells were harvested and metabolites for subsequent GC-MS analyses extracted. Metabolite levels were normalized to WT levels and means of four biological replicates are shown. The color code indicates relative metabolite concentration in STM ΔsodAB compared to WT (red = decreased; green = increased). Gray color show indifferent metabolite concentrations between STM ΔsodAB and WT (Student’s t-test, p < 0.05).

FIGURE 3

Comparison of the metabolic profiles of STM ΔsodAB and mutant strains defective in aconitases or fumarases and succinate dehydrogenase. STM WT and STM ΔsodAB, STM ΔacnAB, STM ΔsdhCDAB and STM ΔfumABC were cultured aerobically in LB broth for 18.5 h at 37°C. Cells were harvested and metabolites extracted for subsequent GC-MS analyses. Metabolite levels were normalized to WT levels and means of at least four biological replicates are shown. Background colors indicate the relative concentration of the respective metabolite in the mutant strains compared to WT (red = decreased; green = increased). Gray background indicates indifferent concentrations between STM mutant and WT strains (Student’s t-test, p < 0.05).

FIGURE 4

Quantification of relative levels of AcnA in STM WT and ΔsodAB. STM WT, ΔsodAB and ΔacnAB were cultured aerobically in LB broth for 3.5, 8, 14, and 18.5 h at 37°C, cells were harvested and lysed. Proteins were separated using SDS-PAGE on 10% gels and blotted onto nitrocellulose membrane. Blots were incubated with anti-AcnA serum and goat-anti-rabbit-HRP antibody, followed by detection using an ECL kit and Chemidoc system. Blots were stripped and incubated with anti-DnaK antibody and goat-anti-mouse-HRP antibody and subsequent detection. (A) Western blot bands of AcnA and DnaK in the respective strains grown for various culture times. (B) Quantification of AcnA signals (DnaK normalized) in the respective strains. Depicted values are representative for two biological replicates.

FIGURE 5

acnA expression in STM WT and ΔsodAB. STM strains were cultured aerobically in LB broth for (A) 3.5, 8, 14, or 18.5 h, or (B) 3 h at 37°C, followed by treatment with MV for 30 min. RNA was extracted using hot-phenol method, followed by DNase I digest, cDNA synthesis and qPCR. (A) acnA expression (16S rRNA normalized) in STM WT and ΔsodAB over time (WT = 1). (B) acnA expression (16S rRNA normalized) in STM WT without (=1) or with MV treatment. Depicted values present results from one of two biological replicates.

FIGURE 6

Phagocytosis and intracellular replication of STM ΔsodAB and mutant strains with defects in Fe-S cluster-containing TCA cycle enzymes. RAW264.7 macrophages were activated with 5 ng/ml interferon-γ 24 h prior infection. STM strains were grown aerobically o/n in LB broth at 37°C and used for infection at a MOI of 1. Infection was synchronized by centrifugation for 5 min. After infection for 25 min, non-internalized bacteria were removed by washing and remaining extracellular bacteria were killed by gentamicin treatment (1 h at 100 μg/ml, followed by 10 μg/ml for the remaining time). Host cells were lysed 2 and 16 h p.i. with 0.1% Triton X-100 in PBS and lysates were plated onto MH agar plates to determine the CFU of intracellular STM. (A) Depicted are CFU/ml obtained at 2 and 16 h p.i. (B) X-fold replication was determined as ratio of CFU at 2 and 16 h p.i. One experiment representative for three biological replicates is shown. Statistical analyses were performed by Student’s t-test and significances are indicated as follows: ^∗p < 0.05; n.s., not significant.

FIGURE 7

Differentially abundant proteins in ΔacnAB- or ΔsodAB-deficient STM compared to WT. Detected proteins were classified according to Gene ontology (biological process) ‘gene expression,’ ‘metabolic process’ and ‘response to stress.’ Depicted are numbers of proteins with significantly altered levels in mutant strains compared to WT. Statistical analyses were performed with Student’s t-test, p < 0.05.

FIGURE 8

Model summarizing effects of aconitase or superoxide dismutase deletion on STM physiology. Deletion of aconitases disrupts the TCA cycle, leading to citrate and, by an unknown mechanism, isocitrate accumulation. Citrate acts as chelator of iron, among other ions, which reacts in Fenton and Haber-Weiss reactions to HR. Increased ROS concentration induce stress regulons, minimizing effects of ROS damage. Deletion of cytosolic superoxide dismutases leads to ROS stress by accumulating SOA. SOA dismutate in part to HPO. Both kinds of ROS damage Fe–S clusters and form with liberated iron ions HR. Although stress response proteins are induced, continuous oxidative stress overwhelms the capacity of ROS-detoxifying enzymes. Beside reduced expression of acnA, ROS attack Fe–S clusters, including those of aconitases, mediating citrate accumulation and downstream effects similar to those observed in STM ΔacnAB. In STM ΔsodAB ROS from both sources damage macromolecules like DNA, which reduces bacterial viability.