Zinc-dependent substrate-level phosphorylation powers Salmonella growth under nitrosative stress of the innate host response

Abstract

The metabolic processes that enable the replication of intracellular Salmonella under nitrosative stress conditions engendered in the innate response of macrophages are poorly understood. A screen of Salmonella transposon mutants identified the ABC-type high-affinity zinc uptake system ZnuABC as a critical determinant of the adaptation of Salmonella to the nitrosative stress generated by the enzymatic activity of inducible nitric oxide (NO) synthase of mononuclear phagocytic cells. NO limits the virulence of a znuB mutant in an acute murine model of salmonellosis. The ZnuABC transporter is crucial for the glycolytic function of fructose bisphosphate aldolase, thereby fueling growth of Salmonella during nitrosative stress produced in the innate response of macrophages. Our investigations demonstrate that glycolysis mediates resistance of Salmonella to the antimicrobial activity of NO produced in an acute model of infection. The ATP synthesized by substrate-level phosphorylation at the payoff phase of glycolysis and acetate fermentation powers the replication of Salmonella experiencing high levels of nitrosative stress. In contrast, despite its high potential for ATP synthesis, oxidative phosphorylation is a major target of inhibition by NO and contributes little to the antinitrosative defenses of intracellular Salmonella. Our investigations have uncovered a previously unsuspected conjunction between zinc homeostasis, glucose metabolism and cellular energetics in the adaptation of intracellular Salmonella to the reactive nitrogen species synthesized in the innate host response.

Fig 1. Effects of carbon sources on resistance of Salmonella to NO.

(A) Growth of Salmonella treated (+ sNO) or untreated (ctrl) with 750 μM of spermine NONOate in MOPS minimal media supplemented with either glucose (GLC), casamino acids (CAA), or both carbon sources (GLC + CAA) (N = 4, mean ± S.E.M.). (B) Depiction of genes in glycolysis and tricarboxylic acid cycle (TCA) targeted for mutation. Sites of entry of GLC and CAA into glycolysis and TCA are shown. Steps of ATP synthesis by substrate-level phosphorylation and fbaAB genes encoding fructose bisphosphate aldolase are also indicated. Growth of Salmonella harboring mutations in TCA (C) or glycolysis (D) after 20 h of growth in MOPS minimal media supplemented with GLC or CAA (N = 3, mean ± S.E.M.). (E) Growth of wild-type (WT), ΔatpB, and ΔackA Δpta Salmonella in MOPS supplemented with GLC or CAA (N = 4, mean ± S.E.M.). (F) Anaerobic growth of Salmonella in MOPS media supplemented with 50 mM NaNO₃ (N = 4, mean ± S.E.M.). *, p < 0.05; ***; p < 0.001.

Fig 2. Selection of Salmonella transposon mutants by nitrosative stress.

A library of 140,000 Salmonella transposon mutants grown in MOPS minimal media supplemented with either glucose (GLC), casamino acids (CAA), or glucose and casamino acids (GLC + CAA) were challenged with 750 μM of spermine NONOate (sNO), 5 mM S-nitrosoglutathione (GSNO) or 5 mM DETA NONOate (dNO) for 20 h. Genomic DNA was extracted, indexed by PCR, and analyzed by deep-sequencing. Gene mutations that showed differential survival with a false discovery rate < 10% were sorted into overlapping groups by media, and the number of common genes were plotted in Venn diagrams (A, C, and E). Positively or negatively selected genes for each NO donor and media were sorted by the Panther Gene Ontology program into biological process categories (B, D, and F). (G) Heat map of log₂ fold changes (F.C.) of transposons in znuABC genes after NO treatment; ns: not statistically significant.

Fig 3. Salmonella require znuB to recover from nitrosative stress and cause disease.

(A) Growth of wild-type (WT) and ΔznuB Salmonella after challenge with 5 mM diethylenetriamine (DETA) or 5 mM DETA NONOate (dNO) in E salts minimal media supplemented with glucose (EG), casamino acids (ECA), or glucose and casaminmo acids (EGCA). Where indicated, media were supplemented with 5 μM ZnCl₂ (+ZnCl₂) (N = 5 to 10, mean). (B) Intracellular growth of WT and ΔznuB Salmonella after 20 h of culture in J774 macrophage-like cells. Select samples were treated with 500 μM of the iNOS inhibitor aminoguanidine (AG) (N = 16, mean ± S.E.M.). (C) Intracellular survival of Salmonella in periodate-elicited macrophages from C57BL/6 or iNOS^-/- mice (N = 5, mean ± S.D.) * and ***, p < 0.05 and <0.001, respectively, as determined by two-way ANOVA. (D) Survival of C57BL/6 or iNOS deficient (iNOS^-/-) mice infected i.p. with approximately 100 CFU of wild-type (WT) or ΔznuB Salmonella (N = 7–9 mice). p < 0.0001 and 0.001 for C57BL/6 and iNOS^-/-, respectively, as determined by logrank analysis.

Fig 4. Fructose bisphosphatate aldolase content.

(A) Fructose-1,6-bisphosphate aldolase (Fba) activity in Salmonella lysates. Where indicated, the lysates were treated with 100 μM of the zinc chelator TPEN. The Fba activity was normalized to WT samples. N = 12, mean ± S.E.M. (B) Effect of the supplementation of EG minimal media with 5 μM ZnCl₂ on the Fba activity of log phase Salmonella. N = 8, mean ± S.D. (C) Fba activity in Salmonella treated for 5 min with 750 μM spermine NONOate. N = 8, mean ± S.D. *, **, *** p < 0.05, <0.01, <0.001, respectively, as determined by two-way ANOVA. (D) Growth of ΔznuB Salmonella after 16–20 h of culture in J774 cells. Where indicated, ΔznuB Salmonella were complemented with fbaA or fbaB genes heterologously expressed from the pBAD promoter. N = 8–16; mean ± S.D. ***, p < 0.001 as determined by one-way ANOVA.

Fig 5. Glycolysis enhances the recovery of Salmonella from nitrosative stress.

(A) Growth of wild-type (WT) and ΔpfkAB Salmonella in LB broth that was left untreated (ctrl) or treated with 500 μM DETA NONOate (dNO) (N = 5, mean). (B) Intracellular replication of WT and ΔpfkAB Salmonella 20 h after challenge of J774 macrophages. Select samples were treated with 500 μM of the iNOS inhibitor N6-(1-iminoethyl)-L-lysine (L-NIL) (N = 4–8, mean ± S.D.). ***, p < 0.001 as determined by two-way ANOVA. (C) Survival of C57BL/6 or iNOS deficient (iNOS^-/-) mice after i.p. inoculation of 100 CFU of wild-type WT or ΔpfkAB Salmonella. (N = 9–10 mice). p < 0.01 for both C57BL/6 and iNOS^-/-, as determined by logrank analysis. (D) Growth of the indicated Salmonella strains in EG minimal media. Where indicated, the specimens were treated with 750 μM spermine NONOate (+ sNO) (N = 4, mean ± S.E.M.). (E) Replication of Salmonella in J774 macrophages in the presence or absence of L-NIL (N = 4–8, mean ± S.D.). (F) Survival of C57BL/6 and iNOS^-/- mice after i.p. challenge with 100 CFU of WT, ΔatpB, or Δnuo Δndh Salmonella. (N = 5 or 10 mice).

Fig 6. Effects of carbon source on the ATP pool of Salmonella undergoing nitrosative stress.

(A) ATP was visualized after TLC analysis of ³²P-labeled Salmonella grown in MOPS minimal media supplemented with either glucose (GLC) and/or casamino acids (CAA). When indicated, the cultures were treated with 750 μM spermine NONOate (sNO) for 5 min. Images are representative of experiments performed on at least two independent days. (B, C-E) The ATP pools were independently estimated with firefly luciferase. The bacteria were grown in EG (C and E) or LB broth (D). Select bacteria were treated with 750 μM sNO for 5 min (N = 6, mean ± S.E.M.). *, **, *** p < 0.05, 0.01, 0.001, respectively, as determined by one-way or two-way ANOVA. (F) Growth of the indicated Salmonella strains after treatment with 1 mM DETA NONOate (dNO) in EG media. (G) Intracellular growth of Salmonella in J774 cells 20 h after infection. Some of the specimens were treated with 500 μM L-NIL. (H) Intracellular survival of Salmonella in periodate-elicited macrophages from C57BL/6 or iNOS^-/- mice (N = 4–5, mean ± S.D.). ***, p < 0.001 as determined by two-way ANOVA. (I) Survival of C57BL/6 or iNOS^-/- mice after i.p. inoculation of 100 CFU of the indicated Salmonella strains. N = 9–15; * p < 0.05, WT vs. ΔpykAF ΔackA Δpta in C57BL/6 mice as determined by logrank analysis; all other comparisons to WT were not statistically significant.

Fig 7. Model for zinc and energy metabolism in the antinitrosative defenses of Salmonella.

ATP can be generated by substrate-level phosphorylation in glycolysis and acetate fermentation or oxidative phosphorylation in the electron transport chain. Nitric oxide (NO) produced by iNOS preferentially targets oxidative phosphorylation, and thus most ATP used for growth under nitrosative stress is derived from substrate-level phosphorylation. Zn²⁺ acquired through the ABC-type high affinity zinc transport system ZnuABC plays a critical role to the antinitrosative defenses of intracellular Salmonella by enabling fructose bisphosphate aldolase enzymatic activity in glycolysis. The utilization of zinc in multiple cellular processes also adds to the antinitrosative defenses of Salmonella.