ABSTRACT
Labile [4Fe-4S]2+ clusters found at the active sites of many dehydratases are susceptible to damage by univalent oxidants that convert the clusters to an inactive [3Fe-4S]1+ form. Bacteria repair damaged clusters in a process that does not require de novo protein synthesis or the Isc and Suf cluster assembly pathways. The current study investigates the participation of the bacterial frataxin ortholog CyaY and the YggX protein, which are proposed to play roles in iron trafficking and iron-sulfur cluster repair. Previous reports found that individual mutations in cyaY or yggX were not associated with phenotypic changes in Escherichia coli and Salmonella enterica serovar Typhimurium, suggesting that CyaY and YggX might have functionally redundant roles. However, we have found that individual mutations in cyaY or yggX confer enhanced susceptibility to hydrogen peroxide in Salmonella enterica serovar Typhimurium. In addition, inactivation of the stm3944 open reading frame, which is located immediately upstream of cyaY and which encodes a putative inner membrane protein, dramatically enhances the hydrogen peroxide sensitivity of a cyaY mutant. Overexpression of STM3944 reduces the elevated intracellular free iron levels observed in an S. Typhimurium fur mutant and also reduces the total cellular iron content under conditions of iron overload, suggesting that the stm3944-encoded protein may mediate iron efflux. Mutations in cyaY and yggX have different effects on the activities of the iron-sulfur cluster-containing aconitase, serine deaminase, and NADH dehydrogenase I enzymes of S. Typhimurium under basal conditions or following recovery from oxidative stress. In addition, cyaY and yggX mutations have additive effects on 6-phosphogluconate dehydratase-dependent growth during nitrosative stress, and a cyaY mutation reduces Salmonella virulence in mice. Collectively, these results indicate that CyaY and YggX play distinct supporting roles in iron-sulfur cluster biosynthesis and the repair of labile clusters damaged by univalent oxidants. Salmonella experiences oxidative and nitrosative stress within host phagocytes, and CyaY-dependent maintenance of labile iron-sulfur clusters appears to be important for Salmonella virulence.
INTRODUCTION
Iron-sulfur clusters are ubiquitous functionally versatile prosthetic groups that participate in a myriad of essential cellular processes ranging from electron transfer to transcriptional regulation (1). In bacteria, iron-sulfur cluster biogenesis is mediated by products of the iscSUA-hscBA-fdx-iscX gene cluster. IscS is a pyridoxal phosphate-dependent cysteine desulfurase that mobilizes elemental sulfur from cysteine for cluster assembly (2). IscU and IscA act as scaffold proteins for the assembly of nascent clusters, while the HscA and HscB chaperones facilitate cluster release and subsequent transfer to acceptor apoproteins (3). An alternative role for IscA in the recruitment of intracellular iron for iron-sulfur cluster synthesis has also been proposed (4). FdxA, a member of the [2Fe-2S] subgroup of the ferredoxin protein family, has an ill-defined redox function in the assembly of iron-sulfur clusters (5).
A second system for iron-sulfur cluster assembly, called the Suf system, is encoded by the sufABCDSE genes in Eubacteria, Archaea, plants, and parasites. In bacteria that contain both the Isc and Suf systems, the isc operon is proposed to be responsible for the housekeeping Fe-S cluster assembly system, while the suf operon provides a backup system specifically adapted to conditions of iron starvation or oxidative stress (6).
Iron-sulfur cluster-containing enzymes that employ [4Fe-4S]2+ clusters as active-site catalysts, including hydroxy acid dehydratases, are primary targets of oxidative stress (7). The solvent-exposed clusters act as Lewis acids in abstracting a hydroxide ion from the bound substrate. Hydroxy acid dehydratases are highly susceptible to oxidative inactivation by oxidants such as superoxide (O2·−) and hydrogen peroxide (H2O2) (8, 9), which abstract a single electron from the cluster to convert it to a [4Fe-4S]3+ form. This form of the cluster is unstable and spontaneously releases the substrate-binding Fe atom as Fe2+, leaving a catalytically inactive residual [3Fe-4S]1+ cluster. A secondary consequence of cluster damage is that the released Fe2+ can bind to DNA and catalyze the formation of hydroxyl radicals that cause DNA damage. To mitigate these effects, bacteria have evolved mechanisms to repair oxidized iron-sulfur clusters. Escherichia coli cells containing elevated superoxide concentrations progressively lose labile dehydratase activity, which is reversed upon a shift to anaerobiosis (10, 11). While it has previously been demonstrated that the process of cluster repair following oxidative damage is distinct from that of Isc/Suf-mediated de novo assembly, the precise mechanism of repair is yet to be determined. The prevailing view is that cluster repair involves conversion of a [3Fe-4S]1+ cluster back to the [4Fe-4S]2+ active state, involving univalent reduction to [3Fe-4S]0 followed by remetalation with Fe2+ (10). The potential toxicity of the Fe2+ likely requires the presence of a chaperone protein to traffic Fe2+ and facilitate its transfer to [3Fe-4S]0.
A protein implicated in Fe-S cluster assembly is CyaY (12, 13). CyaY is the bacterial ortholog of frataxin, a small (210-amino-acid) mitochondrial Fe-binding protein. Frataxin displays remarkable evolutionary conservation, with homologs in mammals, worms, yeast, and Gram-negative bacteria. This ubiquity suggests a critical conserved function. Indeed, reduced frataxin expression resulting from homozygosity for a large GAA repeat expansion in the first intron of the FXN gene leads to a progressive neurodegenerative disease called Friedreich's ataxia. Friedreich's ataxia is clinically manifested by limb and gait ataxia, hypertrophic cardiomyopathy, and diabetes mellitus. Yeast frataxin-knockout models, as well as histopathological investigation of patients with Friedreich's ataxia, have shown that frataxin deficiency leads to iron accumulation in mitochondria and diminished activity of various iron-sulfur proteins, including aconitase (14). On the basis of these observations, a role for frataxin in iron-sulfur cluster assembly has been proposed. The phylogenetic co-occurrence of the frataxin gene and its orthologs, including cyaY, with the hscA and hscB genes of the Isc operon in all sequenced genomes lends support to this proposal (15). Furthermore, a role for frataxin as an iron chaperone protein that prevents oxidant-induced inactivation of aconitase and facilitates reactivation has been reported (16).
The CyaY protein of Salmonella enterica serovar Typhimurium shares 27% identity with human frataxin. The S. Typhimurium cyaY (stm3943) gene is 70 bp downstream of the stm3944 open reading frame (ORF), which encodes a putative inner membrane protein with four transmembrane helices in the same protein family as the nickel-iron dehydrogenase small subunit. The sequence of stm3944 also includes a signature metal-binding CXXC sequence motif found in a wide variety of metal-binding proteins, including ferredoxins, metallothioneins, and metal efflux proteins. This genomic organization is observed only in Salmonella enterica. Although distant homologs of stm3944 (≤35% identity) are found in other bacteria, it is uncertain whether the encoded proteins are functionally related.
While the role of STM3944 is yet to be determined, CyaY has been demonstrated to sequester redox-active free iron and alleviate iron-mediated production of hydroxyl radicals in vitro in the presence of hydrogen peroxide, suggesting that CyaY may function as an iron chaperone under conditions of oxidative stress (4). In support of a role as an iron chaperone, frataxin/CyaY has been shown to bind iron with a relatively low affinity (Kd [dissociation constant], 55 μM for human frataxin and 3.8 μM for E. coli CyaY) via the carboxyl groups of conserved aspartate and glutamate residues (17–19). Studies have demonstrated that CyaY binds to the IscS desulfurase and can also participate in a ternary complex with IscU and IscX, leading to the suggestion that CyaY might sense iron levels and regulate Fe-S cluster biosynthesis (20–22). Complementation of iron-sulfur cluster enzyme and heme deficiencies of a yeast frataxin mutant by mitochondrion-targeted E. coli CyaY suggests at least a partial conservation of function between eukaryotic frataxins and CyaY (23). However, despite the profound phenotypes associated with frataxin deficiency in yeasts and humans, an E. coli cyaY mutant reportedly lacked any apparent phenotype (24). Similarly, studies in Salmonella have failed to identify an overt phenotype for Salmonella enterica carrying a cyaY mutation, other than a modest decrease in the activity of iron-sulfur cluster-containing enzymes and a slight sensitivity to cobalt (13, 25). This has been attributed to functional redundancy between CyaY and other bacterial proteins, including ApbC and YggX, which have proposed roles in iron-sulfur cluster synthesis, protection, or repair (26). YggX, a member of the soxRS regulon (27), confers resistance to oxidative stress and is proposed to have a role in the trafficking of iron to iron-sulfur cluster biosynthetic enzymes and repair systems. YggX has been shown to weakly bind iron (28) and confer resistance to the superoxide-generating compound paraquat (27), decrease the spontaneous mutation frequency, and suppress the defects in iron-sulfur cluster enzyme activity observed in apbC, abpE, and gshA mutants (29, 30).
In this study, we investigated the roles of CyaY, YggX, and STM3944 in the in vivo assembly and repair of iron-sulfur clusters and Salmonella virulence. Our results are consistent with a model in which CyaY is critical for iron-sulfur cluster repair during oxidative stress and functions in concert with YggX in the assembly of iron-sulfur clusters, and STM3944 facilitates the extrusion of adventitious free iron. YggX plays a distinct role in the repair of iron-sulfur clusters following severe oxidative stress and likely has an additional function in ameliorating oxidative stress.
MATERIALS AND METHODS
Materials, bacterial strains, and growth conditions.Bacterial strains, plasmids, and primers are listed in Table 1. Unless otherwise stated, all chemicals were purchased from Sigma-Aldrich (St. Louis, MO). All reagents were made in Millipore Super-Q water (≤0.01 ppm) to reduce levels of iron contamination. The λ-Red recombinase method (31) was employed for the construction of deletion mutations in the cyaY, yggX, stm3944, iscA, and edd genes. Briefly, PCR amplification was performed using primers that included 40 to 60-nucleotide (nt) extensions homologous to regions within the gene to be inactivated and 20-nt priming sequences corresponding to pKD3 or pKD4 templates. Construction of the fur mutant has been described previously (32). Mutations were transduced into wild-type Salmonella enterica serovar Typhimurium ATCC 14028s with bacteriophage P22. Mutant construction was verified by PCR using gene-specific primers. For complementation of the cyaY and yggX mutants, the ribosome binding sites and complete coding regions were PCR amplified prior to cloning into the SacI and SphI sites of a pBAD18 vector carrying a chloramphenicol or kanamycin resistance marker. The stm3944 coding region and ribosomal binding site were introduced into a pBAD18-Cm vector (33) for expression in a fur mutant strain. Plasmid pBAD18-Cm::stm3944 was digested with SacI and SphI, and the ∼500-bp fragment containing stm3944 was gel purified. This fragment was ligated into pBAD18-Kan (33) digested with SacI and SphI to generate pJK714. Construction of pJK707 was as follows: the upstream promoter region of stm3944 was joined to the coding region of cyaY by splicing by overlap extension (SOEing) PCR (34). Equal amounts of PCR template generated from PCR products using primer sets JKP497-cyaYA1/JKP499-cyaYB and JKP500-cyaYC/JKP501-cyaYD and genomic DNA (gDNA) from ATCC 14028s as the template were PCR amplified using primer set JKP497-cyaYA1/JKP501-cyaYD. The 1,071-bp fragment was digested with BamHI and HindIII, ligated into low-copy-number vector pRB3 (35), and confirmed by DNA sequencing. For construction of pJK710, PCR amplification using primers JKP505-yggXD and JKP506-yggXF and gDNA from strain ATCC 14028s as the template generated a 637-bp product containing −361 bp upstream and the coding region of yggX. The PCR fragment was digested with BamHI and HindIII, ligated into pRB3, and confirmed by DNA sequencing. All strains were routinely cultured by shaking at 250 rpm and 37°C in Luria-Bertani (LB) medium supplemented with chloramphenicol (40 μg ml−1), kanamycin (50 μg ml−1), or ampicillin (100 μg ml−1), as indicated below.
Bacterial strains, plasmids, and primers
Whole-cell EPR spectroscopy.A whole-cell electron paramagnetic resonance (EPR) spectroscopy protocol was adapted from the protocol of Woodmansee and Imlay (36). Cells were cultured aerobically in 250 ml LB to an optical density at 600 nm (OD600) of ∼1.0 before harvesting by centrifugation. The pellet was resuspended in 20 mM Tris-HCl, pH 7.4, containing 20 mM desferrioxamine and incubated for 10 min at 37°C. The cells were then centrifuged, washed with cold 20 mM Tris-HCl, pH 7.4, and resuspended in a final volume of 0.5 ml 20 mM Tris-HCl, pH 7.4, containing 10% glycerol. The resuspended cells were loaded into a 3-mm quartz EPR tube (Wilmad-Labglass, Buena, NJ) and immediately frozen in liquid N2. Samples were stored in liquid N2 until EPR measurements were performed. The EPR signals were measured with a Bruker EMX X-band spectrometer (Rheinstetten, Germany). The EPR parameters used were as follows: center field, 1,564 G; sweep width, 500 G; receiver gain, 105; modulation frequency, 100 kHz; modulation amplitude, 10 G; resolution, 2,048 points; number of scans, 10. Iron levels were quantified by normalizing the amplitude of the iron signal of the samples to that of iron standards, and internal concentrations were calculated using the intracellular volume.
ICP-MS.Strains EF434 and EF435 were cultured aerobically in 30 ml LB to an OD600 of approximately 0.3 before protein expression was induced with 0.2% arabinose for 60 min. Following induction, FeSO4 (in 2 mM ascorbate) was added to the culture medium to a final concentration of 1 mM. Cultures were grown for an additional 30 min before a final OD600 was recorded, and cells from a 25-ml culture volume were harvested by centrifugation at 4°C. Cell pellets were washed twice with 25 ml 1 mM EDTA and once with 1 ml ultrapure double-distilled H2O, followed by a high-speed spin. Once all supernatant was removed, the pellets were resuspended in 1 ml trace metal-grade nitric acid and incubated in an 85°C water bath for 45 min. Remaining cell debris was removed by centrifugation at 21,000 × g for 15 min, and the nitric acid solution was diluted 1:10 in Milli-Q-purified water. Inductively coupled plasma mass spectrometry (ICP-MS) analysis was conducted by the Environmental Health Laboratory and Trace Organics Analysis Center at the University of Washington.
H2O2 and Sper/NO susceptibility assays.To measure growth in the presence of H2O2, overnight cultures were diluted 1,000-fold in LB containing 1 mM H2O2 (Fisher Chemical, Fairlawn, NJ). To measure growth in the presence of nitric oxide (NO·), overnight cultures were diluted to a final OD600 of ∼0.001 in M9 minimal medium with 0.2% gluconate containing 750 μM spermine NONOate (Sper/NO; EMD Chemicals, Billerica, MA). Aerobic growth at 37°C was monitored with a Labsystems Bioscreen C microplate reader (Helsinki, Finland). For hydrogen peroxide killing assays, overnight cultures were diluted 1:100 into fresh LB medium and grown at 37°C to an OD600 of ∼0.5 and then treated with a final concentration of 10 mM hydrogen peroxide for 2 h with aeration at 37°C. Aliquots were serially diluted in phosphate-buffered saline (PBS) and then plated onto LB medium for enumeration of the CFU. Percent survival was determined by comparison of the number of surviving cells to the number of cells at time zero.
Catalase activity assays.Catalase activity was determined for cultures grown in LB at 37°C for 18 to 20 h essentially as described previously (37). The average specific activity of catalase (μmol of H2O2 decomposed per min per mg total protein) was calculated from three biological replicates for each strain and subsequently normalized relative to that for the wild type. P values were determined using a two-tailed t test.
qPCR.Three independent cultures were grown in LB at 37°C to an OD600 of ∼1.0. RNA was isolated using an RNeasy minikit (Qiagen, Valencia, CA). First-strand cDNA was synthesized from 500 ng total RNA using RevertAID reverse transcriptase (Thermo Fisher Scientific, Waltham, MA) per the manufacturer's protocol. The primers used for quantitative PCR (qPCR) are listed in Table 1. qPCR assays were performed on cDNA using a SYBR green master mix recipe (38) with a CFX96 real-time system (Bio-Rad, Hercules, CA). Target transcript levels were normalized to rpoD levels. P values were calculated using a two-tailed t test.
Measurement of iron-sulfur cluster enzyme activity and repair.For aconitase assays, 300-ml LB cultures were grown to an OD600 of ∼1.0 before a 50-ml aliquot was harvested for determination of basal aconitase activity. Spectinomycin (500 μg ml−1) was added to the remaining culture to inhibit protein synthesis, immediately followed by addition of 4 mM H2O2 and 1 mM the cell-impermeant iron chelator diethylenetriamine pentaacetic acid (DTPA) (39). After 15 min, 40,000 U catalase were added to terminate the H2O2 stress, and 50-ml aliquots were harvested at designated time points for determination of aconitase reactivation. All assay solutions were equilibrated in an anaerobic chamber for 24 h prior to use. Sample preparation and spectrometry were performed under normal atmospheric conditions. Cell pellets were washed with 50 mM Tris-HCl, pH 7.4, and extracts were prepared by sonication in 50 mM Tris-HCl, pH 7.4, containing 2 mM sodium citrate. Aconitase activity was assayed immediately, as previously described (40). To compare the activities of aconitase, serine deaminase, and NADH dehydrogenase I, 300-ml LB cultures were grown to an OD600 of ∼1.0, and 50-ml aliquots were harvested for determination of basal activity. Cultures were then treated with 0.5 mM, 4 mM, or 8 mM H2O2 for 15 min. The H2O2 stress was terminated by the addition of 40,000 U catalase, and 50-ml aliquots were harvested for analysis. For determination of serine deaminase activity, cells were harvested by centrifugation (20 min, 4°C, 8,000 × g), washed with 100 mM Tris-HCl, pH 8.0, and then resuspended in the same buffer before sonication. The crude extract was clarified by centrifugation (14,000 × g for 30 min), and the supernatant was immediately assayed for serine deaminase activity. Serine deaminase activity was assayed spectrophotometrically in a coupled reaction with l-lactate dehydrogenase by monitoring the decrease in the NADH concentration at 340 nm. For measurement of NADH dehydrogenase I activity, cells were harvested, washed, and resuspended in 50 mM potassium phosphate buffer (pH 7.8) prior to sonication. The cell extract was then clarified by centrifugation, and NADH dehydrogenase I activity was measured by monitoring the oxidation of deamino-NADH at 340 nm. The protein concentration of all cell extracts was measured by the bicinchoninic acid protein assay (Pierce, Rockford, IL).
In vitro reactivation of aconitase.Cells were grown as described above for the aconitase assays but with the following modifications: no DTPA was added to the culture, and cells were harvested immediately following the addition of catalase. Aconitase activity was assayed in catalase-treated cell lysates, and lysates were supplemented with 5 mM dithiothreitol (DTT) or 5 mM DTT and 500 μM FeSO4. Lysates were incubated at 25°C during reactivation. Activity was normalized with respect to the total protein concentration.
Mouse virulence assays.The animal experiments used in this study were approved by the University of Washington Institutional Animal Care and Use Committee. Seven-week-old female C3H/HeN (Nramp1+ Ityr) mice (Charles River Laboratories, Wilmington, MA) were used to measure the virulence of the S. Typhimurium strains as previously described (32). Mice were injected intraperitoneally with 1,000 to 2,000 CFU from cultures grown overnight in LB and diluted in PBS. Inocula were verified by quantitative plating. Five mice were included per experimental group. Mice were monitored daily for signs of illness, and moribund animals were euthanized. In parallel experiments, female C3H/HeN mice were inoculated intraperitoneally with 2,000 CFU of a 1:1 mixture of the wild-type strain (ATCC 14028s) and an isogenic cyaY, yggX, or iscA mutant strain (n = 5 for each mixture). Eight days later, the mice were euthanized, spleens and livers were harvested and homogenized in PBS, and homogenates were plated onto LB agar for determination of the numbers of CFU. Colonies were screened on antibiotic-containing medium to determine the ratio of mutant/wild type. The competitive index is the ratio of (mutant/wild type)output to (mutant/wild type)input. Statistical significance was determined using the Wilcoxon rank-sum test.
RESULTS
CyaY and YggX play distinct roles in Salmonella Typhimurium resistance to oxidative stress.Frataxin-deficient yeast cells display increased mitochondrial iron accumulation and hypersusceptibility to oxidative stress (14), and cultured fibroblasts from patients with Friedreich's ataxia exhibit increased sensitivity to oxidative stress (41). These observations suggest that frataxin deficiency disrupts mitochondrial iron homeostasis and leads to reactive oxygen species-dependent cytotoxicity. We sought to determine whether the frataxin homolog CyaY similarly affords protection against oxidative stress in bacteria. An S. Typhimurium mutant with deletion of the cyaY gene was constructed, and the growth of wild-type and cyaY mutant strains was compared in the presence of hydrogen peroxide. While both wild-type and cyaY mutant S. Typhimurium strains grew equally well in the absence of H2O2, a cyaY mutant exhibited a prolonged lag phase relative to that of the wild type in the presence of H2O2 (Fig. 1). A mutation in ORF stm3944, located just upstream of cyaY (stm3943), did not enhance H2O2 sensitivity compared to that of the wild type. However, a combination of cyaY and stm3944 mutations produced a marked increase in the lag phase that was more pronounced than that observed in a mutant with a mutation in cyaY alone (Fig. 1A).
Relative H2O2 resistance of wild-type and mutant S. Typhimurium strains. Wild-type, cyaY (A, B, and C), stm3944 (A and C), stm3944 cyaY (A and C), and yggX and cyaY yggX (B and C) mutant strains were cultured in the presence of H2O2. (A and B) Strains were inoculated at a density of 106 CFU ml−1 in fresh LB with (open symbols) or without (closed symbols) 1 mM H2O2. Growth was monitored in a Bioscreen C microplate reader with agitation at 37°C. Data shown are means ± standard deviations from at least three independent experiments. (C) Strains were grown in LB at 37°C to an OD600 of ∼0.5 and then treated with 10 mM H2O2 for 2 h with aeration before percent survival was determined. The medians are indicated by horizontal bars. P values were calculated by comparison of the results for the mutants to those for the wild type using the Wilcoxon rank-sum test. Survival T = 2 h, survival at 2 h; *, significant difference between Salmonella cyaY (P < 0.05) and stm3944, stm3944 cyaY, yggX, and cyaY yggX (P < 0.005) mutants.
To address the possibility that YggX compensation might mask phenotypes associated with a cyaY mutation, the H2O2 susceptibility of an S. Typhimurium yggX mutant was also measured (Fig. 1B). The yggX mutant displayed an increased lag phase similar to that of a cyaY mutant strain (Fig. 1B). The H2O2 susceptibility of both cyaY and yggX mutant strains was decreased by introducing a complementing plasmid in trans (see Fig. S1 in the supplemental material). The introduction of a cyaY mutation into a yggX mutant strain increased the H2O2-induced lag phase compared to that for single mutants, suggesting that cyaY and yggX play independent roles in Salmonella resistance to oxidative stress (Fig. 1B).
The enhanced H2O2 susceptibility observed in cyaY, yggX, stm3944 cyaY, and cyaY yggX mutant strains was not attributable to altered catalase activity since the catalase activities of these mutant strains were found to be approximately equivalent to that of the wild type (Fig. S2 in the supplemental material). The catalase activity of a cyaY yggX double mutant was reduced by ∼20% compared to that of the wild type, but there was no significant difference between the double mutant and the cyaY or yggX single mutants (see Fig. S2 in the supplemental material). Furthermore, H2O2 killing assays showed that all mutant strains had significantly reduced survival compared to the wild type, and the cyaY yggX double mutant displayed enhanced sensitivity compared to the cyaY or yggX single mutants (Fig. 1C). These observations are also consistent with distinct roles of cyaY and yggX in oxidative stress resistance.
As increased susceptibility to exogenous H2O2 can be observed as a consequence of elevated intracellular free iron levels (32, 42), the intracellular free iron concentration of a nonstressed cyaY mutant was measured as desferrioxamine-chelatable iron by EPR spectroscopy but was found to be similar to that of the wild type (not shown). Moreover, the steady-state total cellular iron content, measured by inductively coupled plasma atomic emission spectroscopy, was also comparable in cyaY mutant and wild-type strains (not shown). This suggested that the effects of oxidative stress on free iron levels in a cyaY mutant might be only transient and further led us to examine whether stm3944 might ameliorate elevations in free iron.
The stm3944 gene is predicted to code for an inner membrane protein with homology to efflux pumps of the major facilitator superfamily, including the copper-translocating P-type ATPase and the Mn2+/Fe2+ transporter of the Nramp family (KEGG Sequence Similarity Database). The enhanced sensitivity of an stm3944 cyaY double mutant to H2O2 stress suggested that STM3944 might participate in the maintenance of intracellular iron homeostasis. To test this hypothesis, intracellular free iron concentrations were measured in the wild-type strain, the cyaY and stm3944 single mutant strains, and the stm3944 cyaY double mutant strain. Each of these strains exhibited free iron levels similar to the level of the wild type under nonstress conditions (not shown). However, the overexpression of STM3944 in trans on the pBAD18-Cm plasmid resulted in a significant reduction in intracellular free iron levels in a fur mutant strain (Fig. 2A and B). S. Typhimurium fur mutants exhibit elevated free iron levels as a consequence of constitutive iron uptake together with defective iron storage (32). The observed reduction in free iron is consistent with a role for STM3944 in the maintenance of iron homeostasis under stress conditions when free iron levels are elevated, possibly via iron efflux. To further investigate a possible role for STM3944 in iron efflux, cells expressing the FeoAB iron uptake system from a plasmid were grown in medium supplemented with 1 mM FeSO4 to increase intracellular iron concentrations (43). Under these conditions, total cellular iron was reduced by approximately 10% when STM3944 was coexpressed from a plasmid, in comparison to that in cells containing empty vector (not shown). The difference in total iron was statistically significant (P < 0.05) by the Wilcoxon signed-rank test. A reduction in both free and total iron is consistent with a role for STM3944 in efflux of free iron from the cell, although iron efflux may not be the sole mechanism by which elevated free intracellular iron levels are ameliorated during stress conditions.
Intracellular free iron concentrations of an S. Typhimurium fur mutant overexpressing STM3944. (A) Strains JV119 and JV120 were grown in LB to an OD600 of ∼1.0, and desferrioxamine-chelatable free iron concentrations were determined by EPR spectroscopy. Representative EPR spectra for an S. Typhimurium fur mutant containing the pBAD-Cm vector (JV119) or pBAD-Cm::stm3944 (JV120) normalized to cell density are shown. (B) Quantitation of intracellular free iron concentrations. Data are means ± 1 standard deviation from three independent experiments. *, concentration significantly different from that of the wild type (P = 0.01, Student's t test).
CyaY and YggX play distinct roles in the de novo biosynthesis and maintenance of iron-sulfur clusters.Biogenesis of iron-sulfur clusters requires the concerted delivery of iron and sulfur to scaffold proteins. While it is well established that the sulfur in iron-sulfur clusters is derived from l-cysteine in an IscS-mediated process, the specific iron donor for the iron-sulfur cluster assembly remains uncertain. A previous study has suggested that IscA and SufA may redundantly mediate iron recruitment and delivery for iron-sulfur cluster assembly in E. coli (44). In vitro, CyaY can function as an iron donor to IscU in support of iron-sulfur cluster assembly (12). YggX represents another candidate for participation in this process. Thus, the respective contributions of CyaY and YggX to in vivo iron-sulfur cluster assembly were felt to warrant further investigation. The activities of three iron-sulfur cluster-containing enzymes, aconitase and serine deaminase, which contain oxygen-labile iron-sulfur clusters, and NADH dehydrogenase I, which contains multiple oxygen-stable iron-sulfur clusters, were assayed in wild-type and cyaY and yggX mutant strains (Fig. 3A). An isogenic iscA mutant, which has previously been shown to exhibit diminished activity of diverse iron-sulfur proteins in E. coli (10), was included as a positive control.
Iron-sulfur cluster activity in wild-type and mutant S. Typhimurium strains under basal conditions and following H2O2 stress. (A) Basal iron-sulfur cluster enzyme activity was assayed in wild-type and cyaY, yggX, cyaY yggX, and iscA mutant strains. Aconitase (Acn), serine deaminase (Sda) and NADH dehydrogenase I (NdhI) activities were measured in mutant strains grown in LB to an OD600 of ∼1.0 and normalized to wild-type activity. Data shown are means ± 1 standard deviation from at least three independent experiments. *, activity significantly different from that of the wild type (P < 0.05, Student's t test); #, activity significantly different from that of the wild-type, cyaY mutant, or yggX mutant strain (P < 0.05, Student's t test). (B) Relative Sda, Acn, and NdhI activity was assayed following 15 min exposure to 0.5, 4, or 8 mM H2O2. For each assay, activity was normalized to the enzymatic activity in the absence of H2O2, which is displayed as 100%.
The results showed that the basal serine deaminase activity in a cyaY mutant was only 35% of that of the wild type, while the activities of aconitase and NADH dehydrogenase I were comparable in cyaY mutant and wild-type strains (Fig. 3A). In contrast, a yggX mutant displayed a modest 33% reduction in NADH dehydrogenase I activity, whereas the activities of aconitase and serine deaminase in a yggX mutant were similar to those in the wild type (Fig. 3A). In comparison to the activities in the wild type and cyaY or yggX single mutants, a cyaY yggX double mutant exhibited a reduction in basal serine deaminase and NADH dehydrogenase I activities, while the levels of aconitase activity were unaffected (Fig. 3A). An iscA mutant showed reduced activities of aconitase, serine deaminase, and NADH dehydrogenase I, although the degree to which these enzymes were affected differed significantly, with aconitase being the enzyme that was the least affected (Fig. 3A). This is in accord with previous observations in E. coli (10).
As the assays of aconitase and serine deaminase activity included the total activities of aconitase isoenzymes AcnA and AcnB and serine deaminase isoenzymes SdaA, SdaB, TdcG, and DsdA, we examined the transcript levels corresponding to each isoenzyme in wild-type and mutant strains by qPCR (see Fig. S3 in the supplemental material). The levels of the nuoA transcript in wild-type and mutant strains were also measured. The levels of the acnA and acnB transcripts were essentially the same in the mutant strains as in the wild type (Fig. S3 in the supplemental material). The levels of acnA in the cyaY and cyaY yggX mutant strains were 2-fold less than the level in the wild type, but this had little effect on overall aconitase activity (Fig. 3A). Transcript levels for the serine deaminases were not significantly different in the mutant strains and the wild type, with the exception of the level of the tdcG transcript. In the cyaY yggX and iscA mutant strains, tdcG transcript levels were reduced by 4.7- and 11-fold, respectively (see Fig. S3 in the supplemental material), which may account in part for the decreased enzymatic activity observed in these strains (Fig. 3A). However, differences in tdcG transcript levels cannot explain the observed defect in serine deaminase activity in the cyaY mutant (Fig. 3A; see Fig. S3 in the supplemental material). No significant differences in nuoA transcript levels were observed among the strains tested. Hence, with the possible exception of tdcG, we did not observe reductions in transcript levels to account for the reduced enzyme activity observed in the mutant strains (Fig. 3A).
The observed reduction in basal activity of the oxygen-sensitive serine deaminase in a cyaY mutant, but not that of aconitase or NADH dehydrogenase I, led us to test the relative susceptibilities of these enzymes to oxidative stress imposed by various concentrations of H2O2 (Fig. 3B). These studies showed that 4 mM H2O2 destroys approximately 80% of serine deaminase activity, whereas it destroys 60% of aconitase activity, and NADH dehydrogenase I was completely resistant to H2O2 concentrations up to 8 mM. Thus, serine deaminase appears to be the most H2O2 sensitive of these enzyme activities, followed by aconitase and NADH dehydrogenase I, in that order.
Together, these results suggest that even in the presence of YggX, a functional cyaY gene is crucial to maintain the basal activity of enzymes that contain highly oxygen-labile iron-sulfur clusters, such as serine deaminase, but dispensable for the activity of stable iron-sulfur cluster-containing enzymes, such as NADH dehydrogenase I. In contrast, YggX is not required for basal iron-sulfur cluster enzyme activity except in the absence of CyaY, in which case it is required for full NADH dehydrogenase I activity. Unlike aconitase and serine deaminase, which each possess one iron-sulfur cluster, NADH dehydrogenase I contains up to eight iron-sulfur clusters, a feature that may render it particularly dependent on YggX. The dependence of iron-sulfur cluster-containing dehydratases on the presence of CyaY or YggX for basal activity parallels their dependence on IscA (Fig. 3A), with aconitase being the enzyme least affected by an iscA mutation (10) and also the least affected by cyaY and yggX mutations (Fig. 3A). Hence, CyaY and YggX appear to play overlapping but distinct roles in the de novo assembly and maintenance of iron-sulfur clusters.
CyaY and YggX play distinct roles in the in vivo repair of oxidized aconitase.The specific requirement for CyaY to maintain the basal activity of an oxygen-labile iron-sulfur cluster led us to hypothesize that CyaY mediates the repair of oxidatively damaged iron-sulfur clusters. Indeed, a role for human frataxin in facilitating the in vitro transfer of Fe2+ to the [3Fe-4S]1+ cluster of aconitase, resulting in the reactivation of oxidized enzyme, has been demonstrated (16). To determine whether CyaY plays an analogous function in bacteria, in vivo reactivation of aconitase activity was assayed in wild-type and cyaY-deficient mutant strains following a 15-min exposure to H2O2. Brief exposure to physiological oxidants like peroxynitrite, superoxide, and H2O2 has been demonstrated to lead to the rapid in vivo inactivation of iron-sulfur cluster enzymes, including aconitase (10). Data from the same study have indicated that moderate doses of oxidants do not fully degrade clusters, leaving a [3Fe-4S]1+ cluster that may be converted back to an active [4Fe-4S]2+ cluster by a repair process that involves its univalent reduction followed by metalation with Fe2+ (10). Indeed, we found that after cells were treated with a 15-min exposure to H2O2, aconitase activity could be restored in vitro by the addition of iron and DTT, suggesting that the [3Fe-4S]1+ cluster repair observed in vivo in intact cells resulted from remetalation rather than from synthesis of new [4Fe-4S]2+ clusters (see Fig. S4 in the supplemental material).
Following routine aerobic growth, basal levels of aconitase activity were comparable in wild-type and cyaY mutant cells. A 15-min exposure to H2O2 abrogated >90% of aconitase activity in both wild-type and cyaY mutant bacteria. However, within 10 min of terminating the oxidative stress, wild-type bacteria had recovered approximately 55% of their original aconitase activity (Fig. 4A), while a cyaY mutant required 30 min to recover 50% of its original activity (Fig. 4A). Moreover, the presence of pBAD18-Cm::cyaY in the cyaY mutant, but not an empty vector, restored aconitase reactivation kinetics to levels comparable to wild-type levels (Fig. 4B).
Reactivation of aconitase activity following damage by H2O2. Strains grown in LB to an OD600 of ∼1.0 were treated with 4 mM H2O2 for 15 min. Spectinomycin and the iron chelator DTPA were included to block new protein synthesis and to inhibit iron transport from the extracellular medium, respectively. Catalase was added (at 0 min) to terminate the H2O2 stress, and aconitase reactivation was measured at the indicated time intervals after the termination of H2O2 stress. Activities were normalized to the activity of the untreated control (prior to H2O2 exposure), which was set to 100%. (A) Aconitase reactivation in wild-type S. Typhimurium, the cyaY mutant, the yggX mutant, and the cyaY yggX mutant. (B) Aconitase reactivation in wild-type S. Typhimurium, the cyaY mutant containing empty vector pBAD-Cm, and the cyaY mutant containing pBAD-Cm::cyaY. (C) Aconitase reactivation in wild-type S. Typhimurium, the yggX mutant containing empty vector pBAD-Kan, and the yggX mutant containing pBAD-Kan::yggX. Data shown are means ± 1 standard deviation from three independent experiments. *, reactivation levels significantly different from that for the wild type (P < 0.05, Student's t test); #, reactivation significantly different from that for the wild-type, cyaY mutant, and yggX mutant strains (P < 0.05, Student's t test).
To examine whether CyaY and YggX participate in iron-sulfur cluster repair, the kinetics of aconitase repair was compared in isogenic wild-type and yggX mutant S. Typhimurium strains. Like a cyaY mutant, a yggX mutant was impaired in reactivation of oxidized aconitase compared to that for the wild type, requiring 30 min to recover 50% of its basal aconitase activity (Fig. 4A). Furthermore, the observed defect in aconitase repair could be complemented by the provision of yggX in trans on pBAD18-Kan::yggX (Fig. 4C). Combining cyaY and yggX mutations led to a further decrease in the rate of aconitase reactivation; a cyaY yggX double mutant repaired only 25% of its initial aconitase activity in 30 min (Fig. 4A). Collectively, these data indicate that CyaY and YggX exert additive effects in the repair of oxidized aconitase iron-sulfur clusters.
CyaY and YggX are important for 6-GPD-dependent growth during nitrosative stress.Gram-negative bacteria utilize the Entner-Doudoroff (ED) pathway when grown on gluconate as a sole carbon source (45). The loss of activity of 6-phosphogluconate dehydratase (6-GPD), a metabolic enzyme essential for growth on gluconate via the ED pathway, can result from damage to its [4Fe-4S] cluster, which is O2-labile (6, 46). A modest growth disadvantage was seen in Salmonella carrying single or double cyaY and yggX mutations cultured in M9 minimal medium with 0.2% gluconate (Fig. 5). This suggested the involvement of CyaY and YggX in maintaining 6-phosphogluconate dehydratase activity. The growth defects of cyaY, yggX, and cyaY yggX mutant strains in M9 minimal medium-gluconate were exacerbated by treatment with the NO· donor spermine NONOate (Sper/NO) (Fig. 5). The cyaY yggX double mutant had a more severe growth defect during nitrosative stress than cyaY or yggX single mutants, further indicating that CyaY and YggX make distinct contributions to [Fe-S] cluster repair and maintenance (Fig. 5).
Relative nitric oxide resistance of wild-type and mutant S. Typhimurium strains. Growth of the wild type and the cyaY, yggX, and cyaY yggX mutant strains was monitored in the presence of the NO· donor Sper/NO. Strains were inoculated at a density of 106 CFU ml−1 in M9 minimal medium plus 0.2% gluconate with (open symbols) or without (closed symbols) 750 μM Sper/NO. Growth was monitored in a Bioscreen C microplate reader with agitation at 37°C. Data shown are means ± 1 standard deviation from at least three independent experiments.
Since gluconate can also be metabolized via the pentose phosphate pathway, we constructed an edd (6-GPD) mutant and monitored growth in M9 minimal medium-gluconate with or without the addition of Sper/NO to confirm that inhibition of 6-GDP activity by NO· was responsible for the observed growth defects. In the absence of edd, Salmonella was able to grow in M9 minimal medium-gluconate via the pentose phosphate pathway, although the growth rate of the mutant was lower than that of the wild type (Fig. S5 in the supplemental material). The growth of an edd mutant was unaffected by the addition of Sper/NO. However, wild-type Salmonella was inhibited for growth in the presence of Sper/NO, suggesting that 6-GPD is, in fact, inactivated by NO· (see Fig. S5 in the supplemental material).
CyaY is essential for S. Typhimurium virulence during systemic infection of mice.Salmonella must withstand damage to iron-sulfur cluster-containing proteins mediated by host phagocyte-derived reactive oxygen and nitrogen species (47). To investigate the importance of iron-sulfur cluster assembly and repair in Salmonella pathogenesis, the virulence of cyaY, yggX, and iscA strains was evaluated following intraperitoneal inoculation of mice. Two thousand CFU of each strain was inoculated intraperitoneally into 7-week-old C3H/HeN mice, after which the mice were monitored over a 4-week period (Fig. 6A). All three mutant strains appeared to be attenuated for virulence in mice, but only the results for the cyaA and iscA mutants achieved statistical significance (Fig. 6A). In parallel experiments (Fig. 6B), the cyaY and iscA mutants exhibited a statistically significant (P < 0.05) competitive disadvantage in both the liver (competitive indices, 0.19 and 0.41, respectively) and the spleen (competitive indices, 0.41 and 0.39, respectively). The yggX mutant displayed a significant, albeit minor, competitive disadvantage only in the liver (competitive index, 0.72). The requirement of CyaY for Salmonella virulence suggests that iron-sulfur clusters are an important target of host immunity.
Virulence of wild-type S. Typhimurium and isogenic mutant derivatives in a murine model of systemic infection. (A) Seven-week-old C3H/HeN mice (n = 10) were injected intraperitoneally with 1 × 103 to 2 × 103 CFU of wild-type or isogenic mutant strains. Mice were monitored daily for signs of illness, and moribund animals were euthanized. Data are representative of those from at least two or more reproducible, independent experiments. P values were calculated using the log-rank Mantel-Cox test. (B) Seven-week-old C3H/HeN mice (n = 5) were injected intraperitoneally with 2 × 103 total CFU of a 1:1 mixture of wild-type and isogenic mutant bacteria. All animals were euthanized at 8 days postinfection, and the competitive index of the remaining mutant bacteria to the remaining wild-type bacteria was calculated for both livers and spleens. The cyaA, yggX, and iscA mutants each displayed a statistically significant (P < 0.05) competitive disadvantage in the liver compared to the wild type, though the yggX defect (confidence interval = 0.72) was minor compared to the cyaY (confidence interval = 0.19) and iscA (confidence interval = 0.41) mutations. In the spleen, only cyaY and iscA mutants displayed statistically significant (P < 0.05) competitive disadvantages of 0.41 and 0.39, respectively. P values were calculated by the Wilcoxon rank-sum test.
DISCUSSION
Iron-sulfur clusters play a critical role in a wide range of biochemical processes but also render cells susceptible to oxidative damage. Although the mechanisms and molecular participants in de novo iron-sulfur cluster biosynthesis are now relatively well documented, the processes and mediators of iron-sulfur cluster repair remain poorly defined.
The conservation of the iron-binding protein CyaY in all kingdoms of life is indicative of a critical function (13), yet a cyaY mutation was reported to confer no apparent phenotype in E. coli (24), and similar studies comparing wild-type and cyaY mutant Salmonella strains also failed to identify a significant function. Only the introduction of a cyaY mutation into strains defective in other loci implicated in iron-sulfur cluster metabolism resulted in impaired activity of iron-sulfur cluster-containing enzymes, namely, succinate dehydrogenase and NADH dehydrogenase I, along with increased sensitivity to the redox-cycling compound paraquat (13).
Here we provide evidence that CyaY plays an essential role in the maintenance and repair of oxidatively labile iron-sulfur clusters. Moreover, we corroborate previous studies in Salmonella in demonstrating that stable iron-sulfur clusters require both CyaY and YggX for full activity (13), suggesting that CyaY and YggX perform related but distinct functions in iron-sulfur cluster assembly. The importance of CyaY and YggX is particularly apparent in the assembly of NADH dehydrogenase I, which contains a large number of iron-sulfur clusters.
Reactive oxygen species generated during aerobic metabolism damage labile enzymatic iron-sulfur clusters, resulting in their inactivation and the release of iron that in turn catalyzes Fenton chemistry to generate DNA-damaging hydroxyl radicals (42, 48). The steady-state concentrations of hydrogen peroxide and superoxide in E. coli are estimated to be 20 nM (49) and 20 to 40 pM (11), respectively. Moreover, 10 to 17% of E. coli aconitase has been shown to be inactive at any given time as a consequence of inactivation by superoxide (11). Thus, even under routine aerobic growth conditions, the steady-state activity of enzymes containing labile iron-sulfur clusters represents a dynamic balance between their oxidative inactivation and repair. Our observations suggest a model wherein CyaY functions as a chaperone to provide a constant source of iron to support the turnover and repair of enzymes damaged as a result of oxidative stress (Fig. 7). Relative to aconitase, a greater percentage of serine deaminase is expected to be inactive owing to its extreme sensitivity to oxidative damage, and this is reflected by the profoundly reduced serine deaminase activity observed in a cyaY mutant (Fig. 3A and B). Human frataxin has previously been demonstrated to facilitate aconitase reactivation in vitro by transferring Fe2+ to the [3Fe-4S]1+ cluster of aconitase (16). Data from the present study show that CyaY is required for the efficient recovery of aconitase activity in living cells following exposure to oxidative stress conditions and thereby suggest that a role in iron-sulfur cluster repair is conserved in bacterial and human frataxins.
Hypothetical model of interplay between CyaY, YggX, and STM3944 in iron-sulfur cluster repair and protection against iron-induced toxicity. Under conditions of mild and severe oxidative stress, prior to the recruitment of the SoxRS-dependent YggX protein, damage to labile iron-sulfur clusters requires continuous turnover by CyaY, which is shown in complex with the cysteine desulfurase IscS. Extrusion of iron released from oxidized clusters is facilitated by the STM3944 protein. Once recruited, YggX may serve as a sink for iron released from damaged iron-sulfur clusters and promote the repair of damaged clusters independently of CyaY.
Other supporting evidence of a proposed chaperone role for CyaY includes its weak Fe2+ binding (17). Facile iron binding and displacement would allow the formation of transient complexes that may be a necessary feature for CyaY to serve as a carrier protein for iron that is ultimately destined for apoproteins. The reported attenuating effect of CyaY on hydroxyl radical production in vitro (17) and the inability of cyaY mutant S. Typhimurium to efficiently repair iron-sulfur clusters (Fig. 4) help to account for the enhanced susceptibility of a cyaY mutant in the presence of hydrogen peroxide (Fig. 1). The structure of CyaY indicates that it binds iron via the carboxyl groups of conserved aspartate and glutamate residues. Iron binding by carboxyl groups is unaffected by hydrogen peroxide, thereby allowing CyaY to bind iron under conditions of oxidative stress (4). Accumulating evidence indicates that both bacterial CyaY and mammalian frataxin interact with the cysteine desulfurase IscS but exert different effects on iron-sulfur cluster biosynthesis (12, 21, 50). The effect of a cyaY mutation on the kinetics of cluster repair in the current study is similar to that observed for an iscS mutation in E. coli (10) and may indicate that a CyaY-IscS complex participates in both cluster assembly and repair.
A repair function of CyaY would be predicted to be particularly important during host-pathogen interactions in vivo, when host-derived nitric oxide (NO·), reactive oxygen species, or congeners formed by their interaction can damage iron-sulfur clusters (7, 42, 51). During nitrosative stress, growth dependent on the iron-sulfur cluster-containing enzyme 6-phosphogluconate dehydratase (Edd) was reduced in cyaY and yggX mutant Salmonella strains (Fig. 5). Furthermore, the attenuated virulence of a cyaY mutant strain in a murine model of systemic Salmonella infection may be indicative of damage to microbial iron-sulfur cluster-containing enzymes by mediators of host innate immunity. Alternatively, it is conceivable that CyaY plays additional physiological roles in Salmonella metabolism, for example, the donation of iron to ferrochelatase in the final step of heme biosynthesis, as has been described for human frataxin (19), and such activities may also contribute to roles of CyaY during infection.
The present study demonstrates that under severe oxidative stress conditions, YggX, like CyaY, can serve independently to support the repair of labile iron-sulfur clusters, although YggX may have an additional function as a sink for free adventitious iron leached from damaged clusters, which can be retained and utilized for subsequent cluster repair (Fig. 7). The susceptibility of yggX mutant S. Typhimurium to H2O2 observed in this study, together with the observation that YggX sequesters Fe2+ to limit DNA damage by hydroxyl radicals in vitro, is in agreement with this proposal. Indeed, iron leached from oxidized iron-sulfur clusters into the cytosol after a peroxynitrite challenge has been shown to be rapidly sequestered by an undefined scavenging system in E. coli (7). A yggX mutant may be somewhat less attenuated for virulence during murine Salmonella infection than a cyaY mutant (Fig. 6), suggesting that the function of YggX might be supplanted by CyaY and/or other unidentified proteins. In E. coli, the di-iron protein YtfE has been shown to be required not only for the maintenance of the basal iron-sulfur cluster-containing enzyme activity but also for the protection and repair of iron-sulfur clusters under conditions of oxidative and nitrosative stress (52). These observations have led to the proposal that YtfE is involved in the process of Fe2+ recruitment and integration into damaged clusters (52). A ytfE homolog is also present in the S. Typhimurium genome. Whether the protein encoded by ytfE can serve a similar iron-sulfur cluster repair function during oxidative stress in Salmonella remains to be determined, although the markedly attenuated repair of oxidatively damaged aconitase observed in a cyaY yggX double mutant (Fig. 3A) suggests that ytfE cannot fully compensate for the absence of CyaY and YggX.
The mitigation of oxidative stress by STM3944 is apparent only in the absence of CyaY, as evidenced by the enhanced susceptibility of a cyaY stm3944 double mutant to oxidative stress relative to that of stm3944 or cyaY single mutants. We postulate that under conditions of oxidative stress and in the absence of CyaY, damage to iron-sulfur clusters and their inefficient repair lead to a transient rise in intracellular free iron. Adventitious free iron in the cytosol can accelerate hydroxyl radical production and subsequent DNA damage and cell death. Iron released from damaged iron-sulfur clusters thus poses a significant threat, particularly prior to the SoxRS-mediated upregulation of YggX synthesis. STM3944 may help to restrict intracellular free iron levels by promoting iron efflux under such conditions (Fig. 7). The ability of STM3944 overexpression to lower free iron levels in a fur mutant strain (Fig. 2) and to reduce total cellular iron during conditions of iron overload supports a role for STM3944 in the extrusion of free iron.
In summary, this study has elucidated distinct roles of at least two proteins, CyaY and YggX, in the efficient repair of labile iron-sulfur clusters following oxidative stress. Furthermore, our observations provide insight into the role of a previously uncharacterized ORF, stm3944, in iron efflux, providing a model for the coordinated trafficking of iron to support iron-sulfur cluster repair processes while limiting the deleterious consequences of elevated intracellular free iron concentrations. The attenuated virulence of CyaY-deficient Salmonella during systemic infection suggests that iron-sulfur cluster maintenance and repair are essential for pathogenic bacteria subjected to reactive oxygen and nitrogen species generated by host phagocytes. Frataxins appear to share critical conserved functions in iron-sulfur cluster metabolism from bacteria to humans, underscoring the fundamental role of iron in biochemical processes.
ACKNOWLEDGMENTS
We are grateful to Pierre Moënne-Loccoz (Oregon Health and Science University) for his useful discussions.
This work was supported by grants to F.C.F. from the National Institutes of Health (AI39557, AI77629).
FOOTNOTES
- Received 16 August 2013.
- Returned for modification 15 September 2013.
- Accepted 22 December 2013.
- Accepted manuscript posted online 13 January 2014.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.01022-13.
- Copyright © 2014, American Society for Microbiology. All Rights Reserved.
