Nitric Oxide Disrupts Zinc Homeostasis in Salmonella enterica Serovar Typhimurium

Abstract

Nitric oxide (NO·) produced by mammalian cells exerts antimicrobial actions that result primarily from the modification of protein thiols (S-nitrosylation) and metal centers. A comprehensive approach was used to identify novel targets of NO· in Salmonella enterica serovar Typhimurium (S. Typhimurium). Newly identified targets include zinc metalloproteins required for DNA replication and repair (DnaG, PriA, and TopA), protein synthesis (AlaS and RpmE), and various metabolic activities (ClpX, GloB, MetE, PepA, and QueC). The cytotoxic actions of free zinc are mitigated by the ZntA and ZitB zinc efflux transporters, which are required for S. Typhimurium resistance to zinc overload and nitrosative stress in vitro Zinc efflux also ameliorates NO·-dependent zinc mobilization following internalization by activated macrophages and is required for virulence in NO·-producing mice, demonstrating that host-derived NO· causes zinc stress in intracellular bacteria.IMPORTANCE Nitric oxide (NO·) is produced by macrophages in response to inflammatory stimuli and restricts the growth of intracellular bacteria. Mechanisms of NO·-dependent antimicrobial actions are incompletely understood. Here, we show that zinc metalloproteins are important targets of NO· in Salmonella, including the DNA replication proteins DnaG and PriA, which were hypothesized to be NO· targets in earlier studies. Like iron, zinc is a cofactor for several essential proteins but is toxic at elevated concentrations. This study demonstrates that NO· mobilizes free zinc in Salmonella and that specific efflux transporters ameliorate the cytotoxic effects of free zinc during infection.

Keywords: Salmonella; nitric oxide; pathogenesis; transporters; zinc homeostasis.

FIG 1

Classification of proteins in the S. Typhimurium S-nitrosoproteome. (A) Functional classification of proteins with cysteine residues modified by NO· treatment. A total of 141 modified proteins were identified, and classification is shown as a percentage of the total. (B) Of the proteins modified by NO·, 15 (~10%) were found to be zinc metalloproteins. The metalloproteins are sorted by functional category and listed in numerical order by gene identifier in S. Typhimurium strain 14028s.

FIG 2

Expression of zinc transport systems in S. Typhimurium. qPCR data are presented as a positive fold change of treated compared to untreated cells with zinc efflux systems shown in blue and zinc acquisition systems shown in red. The solid line indicates a fold change of 1 to delineate between upregulation (>1) and downregulation (<1). (A) At 5 min after treatment with the NO· donor diethylamine NONOate (DEANO), expression of all zinc transport systems except zntA (dark blue bar) was modestly upregulated. The expression changes for zntB, zitB, and zupT achieved statistical significance with P values of 0.05, 0.004, and 0.03, respectively. At later time points, there was no significant difference in zinc transporter expression between treated and untreated cells. (B) In the presence of excess zinc (2 mM ZnSO₄), expression of the high-affinity zinc efflux system zntA (dark blue bar) was significantly upregulated (P = 0.04), while expression of the high-affinity acquisition system znuABC (represented by znuC, red bar) was downregulated (P = 0.005). When cultures were treated with 1.5 mM EDTA to chelate zinc in the medium, expression of the znuABC acquisition system increased (P < 0.001). *, statistical significance was determined by one-sample t test compared to theoretical means of 2 for upregulated genes and 0.5 for downregulated genes. Data are the means from 3 (ZnSO₄) and 8 (EDTA) replicates. Error bars represent standard deviations.

FIG 3

ICP-MS analysis of total cellular zinc content in S. Typhimurium following NO· treatment and transcriptional monitoring of NO· sensed by cells. (A) S. Typhimurium cells at an OD₆₀₀ of ≈1 were treated with 2 mM diethylamine NONOate (DEANO), and total cellular zinc was measured at various times posttreatment. By 5 min posttreatment, total cellular zinc had fallen significantly compared to untreated cells, suggesting that zinc is effluxed from the cell following NO· treatment. Zinc levels gradually recovered to baseline levels over the course of 60 min. Statistical significance was determined by unpaired two-tailed t test; * indicates P values of <0.001 and 0.036, respectively. Error bars represent standard deviations. (B) A portion of each culture at each time point was also used to prepare RNA and cDNA for transcriptional analysis. Data are presented as the mean fold change in transcript level compared to untreated cells at each time point. The error bars represent standard deviations. Expression of hmp is regulated by the NO·-sensing NsrR regulator. The high level of expression at 5 min posttreatment indicates that the highest levels of NO· were present during this time period. At later time points, the level of hmp transcript declined significantly, indicating that the amount of NO· declined by 15 min posttreatment and remained low thereafter.

FIG 4

ZntA and ZitB are the primary zinc efflux transporters in S. Typhimurium. (A) A ΔzntA mutant (pink) was impaired for growth, represented by a delayed exit from lag phase, in both 0.125 mM ZnSO₄ and 0.25 mM ZnSO₄, while a ΔzntB mutant (aqua) and a ΔzitB mutant (light green) exhibited growth comparable to wild type (blue) under these conditions. (B) Significance of the growth defect for the ΔzntA mutant in panel A was determined by calculating the mean time required to reach 50% of the maximum final OD₆₀₀ for each strain. (C) A ΔzntB ΔzitB double mutant (purple) exhibited growth comparable to wild type (blue) when exposed to elevated zinc concentrations. A ΔzntA ΔzntB mutant (red) behaved similarly to a ΔzntA mutant, whereas a ΔzntA ΔzitB mutant (green) displayed a more severe growth delay at 0.125 mM ZnSO₄ and was unable to grow at 0.25 mM ZnSO₄. A ΔzntA ΔzntB ΔzitB triple mutant (orange) exhibited growth characteristics identical to those of a ΔzntA ΔzitB double mutant. (D) Significance of the growth defects for the mutants in panel C was determined by calculating the mean time required to reach 50% of the maximum final OD₆₀₀ for each strain. Statistical significance of the growth defects was determined by unpaired two-tailed t test, and an asterisk indicates P values of <0.001 for ΔzntA, ΔzntA ΔzntB, ΔzntA ΔzitB, and ΔzntA ΔzntB ΔzitB mutants compared to wild type at all concentrations. Error bars represent standard deviations.

FIG 5

Zinc efflux by ZntA and ZitB is required for S. Typhimurium resistance to nitrosative stress in vitro. (A) Single zinc efflux mutants (pink, aqua, and light green) and a ΔzntA ΔzntB double mutant (red) were no more sensitive to NO· generated by the donor spermine NONOate (SperNO) than wild-type cells (blue). However, a ΔzntA ΔzitB double mutant (green) exhibited significantly delayed growth, indicating enhanced NO· sensitivity (*, P < 0.0001). (B) The growth defect of the ΔzntA ΔzitB mutant was complemented by expression of either ZntA or ZitB from its native promoter in trans. Statistical significance was determined by comparing the time required to reach 50% of the maximal OD₆₀₀ (dashed line) by unpaired two-tailed t test.

FIG 6

Free intracellular zinc levels increase in a ΔzntA ΔzitB S. Typhimurium mutant during macrophage infection in response to NO· production. (A) Changes in NO· production are shown at the time of infection (0 h) and 13 h postinfection in the presence or absence of the NOS inhibitor l-NMMA. IFN-γ-primed murine macrophages infected with either wild-type (blue) or ΔzntA ΔzitB (green) S. Typhimurium produced significant levels of NO· after 13 h in the absence of the NOS inhibitor l-NMMA (P < 0.001) but not in the presence of 2 mM l-NMMA (ns). (B) Changes in FRET ratio are shown immediately following infection (0 h) and 13 h postinfection. The FRET ratio of the ZapCV5 biosensor increased in ΔzntA ΔzitB mutants (green) isolated from murine macrophages after 13 h (P < 0.001), but not in wild-type S. Typhimurium, indicating that intracellular free zinc levels rise in the efflux-deficient mutant during infection. The increase in FRET was not observed when macrophages were treated with 2 mM l-NMMA to inhibit NO· production (ns). (C) Flow cytometry histograms from one representative experiment. The mean value of each histogram (T₀, gray shape, and T₁₃, blue line) is indicated. Data in panels A and B are presented as the means with error bars representing standard deviations. Statistical significance (*) was determined by one-way analysis of variance.

FIG 7

Virulence of ΔzntA ΔzitB mutant S. Typhimurium is attenuated in NO·-producing mice. Solid lines represent the median competitive index (CI) for each organ. The dotted line represents the expected CI if neither strain has a competitive advantage. (A) In wild-type C3H/HeOuJ mice, a ΔzntA ΔzitB mutant has a significant competitive disadvantage compared to wild type (P = 0.002 by Wilcoxon signed-rank test to a hypothetical median of 1 for both spleen and liver). (B) In C3H/HeOuJ mice that cannot produce NO· due to treatment with 500 µg ml⁻¹ l-N₆-(1-iminoethyl)lysine dihydrochloride (l-NIL), the mutant no longer has a statistically significant disadvantage compared to wild type, and the CIs are significantly different (*, P = 0.007 for spleen and P < 0.001 for liver by Mann-Whitney test) from the CI in untreated mice. A total of 10 animals were tested for each condition. Data points at a CI of 0.01 were at the limit of detection for the assay.

FIG 8

A model of zinc homeostasis in Salmonella Typhimurium. Under conditions of zinc deficiency, zinc is not available to bind to the Zur repressor, leading to expression of the ZnuABC zinc acquisition system. ZupT, whose regulation is uncharacterized, has also been shown to contribute to zinc acquisition. When zinc is abundant, Zur bound to zinc represses ZnuABC expression. In addition, free cytoplasmic zinc binds to the transcriptional activator ZntR to induce expression of the ZntA zinc efflux system. Zinc efflux in S. Typhimurium is also mediated by ZitB. Under conditions of nitrosative stress, S-nitrosylation of cysteine ligands in zinc metalloproteins leads to mobilization of free intracellular zinc. The zinc efflux activities of ZntA and ZitB are required for the resistance of S. Typhimurium to nitrosative stress.