Regulation of the Intracellular Free Iron Pool by Dpr Provides Oxygen Tolerance to Streptococcus mutans

Dpr is an iron-binding protein required for oxygen tolerancein Streptococcus mutans. We previously proposed that Dpr couldconfer oxygen tolerance to the bacterium by sequestering intracellularfree iron ions that catalyze generation of highly toxic radicals(Y. Yamamoto, M. Higuchi, L. B. Poole, and Y. Kamio, J. Bacteriol.182:3740-3747, 2000; Y. Yamamoto, L. B. Poole, R. R. Hantgan,and Y. Kamio, J. Bacteriol. 184:2931-2939, 2002). Here, we examinedthe intracellular free iron status of wild-type (WT) and dprmutant strains of S. mutans, before and after exposure to air,by using electron spin resonance spectrometry. Under anaerobicconditions, free iron ion concentrations of WT and dpr strainswere 225.9 ± 2.6 and 333.0 ± 61.3 µM, respectively.Exposure of WT cells to air for 1 h induced Dpr expression andreduced intracellular free iron ion concentrations to 22.5 ±5.3 µM; under these conditions, dpr mutant cells maintainedintracellular iron concentration at 230.3 ± 28.8 µM.A decrease in cell viability and genomic DNA degradation wasobserved in the dpr mutant exposed to air. These data indicatethat regulation of the intracellular free iron pool by Dpr isrequired for oxygen tolerance in S. mutans.

INTRODUCTION

Top

Introduction

Iron is an essential and beneficial nutrient for most organismsbut has toxic properties in the presence of oxygen (4, 15, 31).Iron ions stimulate the generation of highly reactive and toxicoxygen species such as hydroxyl radicals (11, 15). In vitroexperiments have shown that Fe(II) catalyzes nonenzymatic hydroxylradical formation from hydrogen peroxide via the Fenton reaction,whereas hydrogen peroxide remained intact in the absence ofiron ions at physiological pH (11). Intracellular iron sourcefor the Fenton reaction is considered to be a free iron poolin cells (15, 17). Tight regulation of iron metabolism, especiallythe intracellular free iron pool, is therefore regarded as adetermining factor for survival of an organism in air (11, 15,17, 31).

Methods to investigate the intracellular free iron pool in intactcells were recently developed, and the presence of several factorsaffecting the free iron status was reported in both prokaryotesand eukaryotes (17, 19). It has been reported that, in Escherichiacoli and Saccharomyces cerevisiae, accumulation of intracellularsuperoxide, owing to a superoxide dismutase deficiency, increasesthe level of free iron pool by releasing iron ions from proteinscontaining iron-sulfur clusters (19, 29). In an E. coli furmutant, aberrant regulation of iron uptake was associated withan increase in the level of free iron (19). In mammalian cells,repression of ferritin H subunit expression increased the levelof intracellular free iron (17). In all reported cases, an increasein the free iron pool correlated with an increase in oxidativestress (17, 19, 29, 31).

Streptococcus mutans, a principal causative agent of human dentalcaries, cannot synthesize heme and lacks both a respiratorychain and catalase, which are required for elimination of hydrogenperoxide in most aerobic organisms. However, S. mutans growsunder aerobic conditions and induces several antioxidant proteinswhen cells are exposed to air (12, 13, 24, 27, 35-37). We previouslyidentified dpr (for dps-like peroxide resistance) as a potentialperoxide resistance gene from S. mutans. Studies of a seriesof dpr-deficient strains led us to conclude that dpr plays avital role in aerobic survival of S. mutans (36). Our furtherstudies on the purified dpr gene product showed that Dpr formsferritin-like spherical dodecamers and binds up to 480 ironatoms per complex (37). Primary amino acid sequence homologiesindicate that Dpr is a member of the Dps (for DNA-binding proteinfrom starved cells) (3) protein family (36). Dps is a nonspecificDNA-binding protein that is induced by oxidative or nutrientstress in E. coli (3). Stable Dps-DNA complex formation is believedto protect DNA from hydrogen peroxide action (3, 21, 33). However,in the case of S. mutans, Dpr could not bind DNA (37). We thereforeproposed another mode of cell protection from oxidative stressby Dpr, based on its sequestration of intracellular iron ions.We demonstrated in vitro that Dpr prevents iron-dependent hydroxylradical formation (36). At almost the same time, Zhao et al.reported iron-binding and iron-detoxifying properties of E.coli Dps (38). It was also reported that some Dps family proteinshaving iron binding, but not DNA-binding ability, were involvedin oxidative stress resistance (8, 16, 28). The crystal structureof Dps family proteins, including Streptococcus suis Dpr homologue,revealed a ferritin-like structure of the proteins, indicatingthat this class of proteins could incorporate iron ions as ferritindoes (8, 10, 14, 18, 38). Taken together with our data on Dprproperties, it was suggested that Dps family proteins couldaffect the cellular free iron ion status, thereby conferringoxygen tolerance. In the present study, we measured the intracellularfree iron pool of wild-type (WT) and dpr strains of S. mutansand clarified the role of Dpr in regulating the intracellularfree iron pool and on bacterial survival in air.

MATERIALS AND METHODS

Top

Materials and Methods

Strains, media, and growth conditions. S. mutans GS-5 (WT strain) and DES (dpr-deficient mutant) (36)were used in the present study. Cells were prepared for analysisby electron spin resonance (ESR) spectrometry as follows. A10-ml preculture of S. mutans, prepared in Todd-Hewitt broth(THB; Difco Laboratories, Detroit, Mich.) under anaerobic conditions(in an anaerobic glove box [Hirasawa Works, Tokyo, Japan] inan atmosphere of 80% nitrogen, 10% carbon dioxide, and 10% hydrogen),was added to a 300-ml culture in the same medium. The culturewas incubated at 37°C for 3.5 h (A₆₆₀ = $~$ 0.8) under anaerobicconditions. At this time, part of the culture (100 ml) was removedand used as the zero time sample. The remaining culture (200ml) was centrifuged at 7,800 x g for 5 min, resuspended in thesame volume of fresh THB medium, transferred to 500-ml flasks,and then incubated at 37°C with shaking (120 cycles/min).After 30 min of incubation, 100 ml of the culture was removedas the 30-min sample. The rest of the culture was incubatedfor another 30 min and used as the 60-min sample.

ESR spectrometry sample preparation. Portions (100 ml) of the cultures described above were centrifugedat 7,800 x g for 5 min. Pellets were resuspended in 5 ml ofTHB medium with or without 20 mM deferoxamine (Sigma) and thenincubated at 37°C with shaking (170 cycles/min) for 10 minunder aerobic conditions. Cells were collected by centrifugationat 7,800 x g for 5 min, washed with ice-cold 20 mM Tris-HClbuffer at pH 7.0, and resuspended in 0.3 ml of the same buffercontaining 10% (vol/vol) glycerol. An aliquot of each samplewas taken to measure the optical density at 660 nm. Then, 200µl of each cell suspension was transferred to a quartzESR tube, immediately frozen, and stored at –80°Cuntil ESR measurements were carried out.

ESR spectrometry. ESR spectra were recorded on an RE-3X ESR spectrometer (JEOL,Ltd., Tokyo, Japan). Samples were maintained at –196°Cby using a finger Dewar vessel filled with liquid nitrogen.Experimental conditions used for low-temperature Fe(III) electronparamagnetic resonance (EPR) were as follows: center field,250 mT; sweep width, 150 mT (250 mT for wider sweep); frequency,9.21 GHz; microwave power, 5 mW; modulation amplitude, 1 mT;modulation frequency, 100 kHz; receiver gain, 1x100; sweep time,4 min; and time constant, 0.03 s. The g value was calculatedby using the standard formula g = hv/ßH, where h isPlanck's constant, v is the frequency, ß is the Bohrmagneton, and H is the external magnetic field at resonance.

Calculation of intracellular free iron concentration. The double-integrated intensities of the g = 4.3 signal of eachsample were converted to intracellular free iron ion concentrationsas follows. The amount of deferoxamine-Fe(III) in the ESR samplewas quantified by using the EPR signals of deferoxamine-Fe(III)of known concentrations. First, 1 ml of cell suspension (opticaldensity at 600 nm of 1.0) was calculated to contain 0.58 µlof intracellular water volume, based on (i) the reported internalwater content in S. mutans cells of 1.6 µl per mg (dryweight) (25) and (ii) the fact that 1 ml of cell suspension(A₆₆₀ = 1.0) contained 0.365 ± 0.034 mg (dry weight).We used this value, along with the ESR signal from an externalFe(III) standard and the optical density of the ESR sample,to quantify intracellular free iron concentrations.

Measurement of total iron. S. mutans cells were collected by centrifugation at 7,800 xg for 10 min. Cells were washed once with phosphate-bufferedsaline (pH 7.0) and twice with Milli-Q water (Millipore Corp.,Tokyo, Japan). Washed cells were resuspended in 1 ml of Milli-Qwater and then transferred to a Teflon container. Water wasremoved from cells by incubation at 90°C for 20 h, and thebacterial dry weight was measured. Next, 2 ml of concentratednitric acid (Ultratrace analysis grade; Wako Pure Chemical Industries,Osaka, Japan) and 0.2 ml of concentrated perchloric acid (UltrapureAA-100; Tama Chemicals, Kanagawa, Japan) were added to about100 mg of dried bacterial cells in a Teflon container, and thecells were dissolved into liquid by microwave treatment as describedpreviously (23). After the cells were dissolved, the containerswere heated on a hot plate at 160°C to near dryness andthen dissolved in 5 ml of 5% nitric acid solution for analysisby atomic absorption spectrometry with an atomic absorptionspectrometer (170-30; Hitachi, Tokyo, Japan). The iron contentand bacterial cell dry weight of samples, coupled with the reportedinternal water content in S. mutans cells of 1.6 µl permg (dry weight) (25), allowed us to quantify the total ironconcentration in the cell.

Monitoring survival, genomic DNA degradation, and expression of Dpr. For viable cell determinations, culture dilutions were platedon solid THB medium supplemented with 500 U of bovine livercatalase (Sigma). After 48 h of incubation in an anaerobic boxat 37°C, the CFU were counted. Genomic DNA of S. mutanswas prepared as described previously (35), with some modifications.Cells were treated with both mutanolysin (200 U/ml; Sigma) andacromopeptidase (1,000 U/ml; Wako) for 15 min at 37°C in10 mM Tris-HCl buffer (pH 8.0) containing 1 mM EDTA prior tolysis by sodium dodecyl sulfate. DNA samples (500 ng) were electrophoresedon a 1% Tris-acetate agarose gel and then visualized by ethidiumbromide staining. For Western blot analyses, cell lysates wereprepared as described previously (36) and separated by sodiumdodecyl sulfate-15% polyacrylamide gel electrophoresis. Proteinbands corresponding to Dpr were identified as described by usinganti-Dpr antibody (37).

RESULTS AND DISCUSSION

Top

Results and Discussion

Pool size of intracellular free iron in WT and dpr-deficient mutant of S. mutans. The intracellular free iron pool was measured by ESR spectrometryessentially according to the method of Woodmansee and Imlay(34). Anaerobically grown S. mutans cells were treated withdeferoxamine, a cell-permeable iron chelator, that converts"free iron" to the Fe(III) species. This species gives an intenseESR signal at g = 4.3 (34). Representative low-temperature Fe(III)ESR spectra of whole S. mutans cells are shown in Fig. 1. Astrong signal at g = 4.3 was obtained for deferoxamine-treatedS. mutans cells. The Fe(III) ESR signal at g = 4.3 is characteristicof ferric iron in a high spin complex (34). In contrast, theESR spectrum of iron-loaded Dpr, which was prepared as previouslydescribed (37), gave a very broad ESR signal at a g value ofca. 2 but did not give an ESR signal at a g value of 4.3 (datanot shown). Thus, the signal at g = 4.3, obtained for S. mutanscells treated with deferoxamine, is distinguishable from thesignal corresponding to Dpr-bound iron.

View larger version (16K):
[in this window]
[in a new window]
FIG. 1. ESR results of nontreated and deferoxamine-treated WT S. mutans cells. Anaerobically grown S. mutans WT cells (late exponential phase) were used to obtain whole-cell ESR spectra. ESR spectra of S. mutans cells not treated (a) or treated (b) with deferoxamine are shown.

We used ESR spectrometry to compare intracellular free ironconcentrations in WT and dpr strains before and after exposureto air (Fig. 2). Under anaerobic conditions, similar intensitiesof signal at g = 4.3 were detected for both strains. Upon exposureto air, the intensity of signal obtained for WT cells decreased.In contrast, the intensity of signal for the dpr-deficient cellsremained high (Fig. 2).

View larger version (17K):
[in this window]
[in a new window]
FIG. 2. ESR results of WT and dpr S. mutans cells. ESR spectra of deferoxamine-treated WT and dpr mutant strains before (0 min) or after (30 and 60 min) exposure to air are shown. The zero time samples were processed in the presence of 50 µg of chloramphenicol/ml to prevent protein synthesis during preparation. ESR results typical of those obtained in three independent experiments performed for each strain and condition tested are shown.

The double-integrated intensities of ESR signals were then convertedto intracellular free iron ion concentrations (see Materialsand Methods). Under anaerobic conditions, WT and dpr strainshad intracellular free iron concentrations of 225.9 ±2.6 and 333.0 ± 61.3 µM, respectively (Fig. 3A and B).After cell exposure to air for 1 h, the values of theWT strain dropped to 22.5 ± 5.3 µM; in contrast,values for the dpr mutant remained high, at 230.3 ± 28.8µM (Fig. 3A and B). Total iron content in cells, determinedby atomic absorption spectrometry, is also shown (Fig. 3A and B).In contrast to the rapid decrease of intracellular freeiron in the WT strain, total iron content decreased by only13% after 1 h of exposure to air, indicating that the decreaseof free iron concentrations did not result from the change oftotal iron content but rather from the change of iron statusinside the cell. Western blot analysis of Dpr expression underthe different growth conditions showed that Dpr synthesis wasinduced by exposing cells to air (Fig. 3C). In WT cells, thedecrease in intracellular free iron concentrations correlatedwith increased Dpr expression (Fig. 3). In order to examinewhether Dpr synthesis is really responsible for attenuatingintracellular free iron concentrations, chloramphenicol (50µg/ml) was added to the WT culture prior to aerobic incubation.After 1 h of incubation in air, cells were analyzed by Westernblotting and ESR spectrometry. Chloramphenicol treatment ofWT cells prevented Dpr expression (data not shown) and resultedin higher intracellular free iron concentrations (468.5 ±15.6 µM). These results strongly indicate that, in S.mutans, Dpr incorporates free iron ions in vivo and contributesto lowering the intracellular free iron ion concentration.

View larger version (21K):
[in this window]
[in a new window]
FIG. 3. Intracellular free iron ion concentrations of WT and dpr strains of S. mutans. Total iron content ( ${square}$ ) and intracellular free iron ion concentrations ( ${blacksquare}$ ) of the S. mutans WT strain (A) and dpr mutant (B) before (0 min) or after (30 and 60 min) exposure of cells to air are shown. The total iron content and intracellular free iron concentrations were calculated as described in Materials and Methods. The results are the means ± standard deviations for triplicate determinations. (C) Expression of Dpr in S. mutans WT strain upon exposure to air. Expression of Dpr was analyzed by immunoblotting with Dpr-specific antibody for detection. Each lane was loaded with 1.25 µg of protein of the corresponding extract. A result typical of three independent experiments is shown.

In the dpr mutant, total iron amounts decreased more rapidlythan in the WT strain after exposure of the cells to air (Fig.3A and B). After 1 h of incubation of the dpr mutant in air,both total and free iron concentrations decreased by ca. 30%.This result implies that decreased iron uptake in the dpr mutant,due to cell death as described below, might reduce the intracellularfree iron concentrations in the dpr mutant by reducing the totaliron content.

Effects of high intracellular free iron concentrations on growth, survival, and DNA degradation of S. mutans. In the presence of oxygen, excess amounts of intracellular freeiron ions may catalyze the generation of reactive oxygen speciesthat degrade cellular components and cause cell death (15, 31).We explored the effects of high intracellular free iron concentrationson growth and survival of the dpr mutant (Fig. 4A). Cell densitiesof both WT and dpr mutant strains slightly increased duringthe incubation period in air. However, the number of dpr mutantCFU decreased 100-fold after 1 h of exposure of the cells toair, whereas the number of WT strain CFU remained constant (Fig.4A). A main target of oxygen-induced cellular damage is DNA(15, 31). The effect of aeration on genomic DNA extracted fromWT and dpr mutant strains was examined by gel electrophoresis(Fig. 4B). Under anaerobic conditions, no significant differencesin electrophoretic mobility were observed between genomic DNAfrom WT and dpr mutant strains. After exposure to air, however,marked degradation of DNA was observed in the dpr mutant extract(Fig. 4B). DNA integrity and cell survival of the dpr mutantwere restored by the addition of catalase or deferoxamine, eachof which removes a substrate for the Fenton reaction, to growthmedium during cell exposure to air (Fig. 5). These results stronglyindicate that iron-mediated generation of hydroxyl radicalsvia the Fenton reaction degraded cellular components such asDNA and caused cell death in the dpr mutant.

View larger version (30K):
[in this window]
[in a new window]
FIG. 4. Growth, survival, and genomic DNA degradation of S. mutans WT and dpr strains before or after cell exposure to air. (A) Growth of WT strain ( ${square}$ ) and dpr mutant (•) were monitored by measuring optical density at 660 nm. Respective CFU values are shown as white and black bars. The results presented are the means ± standard deviations for triplicate determinations. (B) Genomic DNAs extracted from WT and dpr mutant strains before or after exposure to air were analyzed by gel electrophoresis. Genomic DNAs of the dpr mutant at 0 min (lane 1), WT strain at 0 min (lane 2), dpr mutant after 60 min (lane 3), and WT strain after 60 min (lane 4) are shown. A result typical of three independent experiments is shown.

View larger version (34K):
[in this window]
[in a new window]
FIG. 5. Effects of catalase and deferoxamine on survival and DNA degradation of dpr mutant. Either catalase or deferoxamine was added to an anaerobic culture of the dpr strain, which was further incubated at 37°C with aeration by shaking (120 cycles/min) for 60 min. CFU (A) and electrophoretic profiles of genomic DNA (B) were analyzed. dpr mutant (lane 1), dpr mutant supplemented with 100 or 1,000 U of bovine liver catalase (lanes 2 and 3)/ml and dpr mutant supplemented with 0.1 or 1 mM deferoxamine (lanes 4 and 5) are shown. In panel A, the results are the means ± standard deviations for triplicate determinations. In panel B, results typical of three independent experiments are shown.

The present study provided the first direct evidence that Dpsfamily proteins can affect the intracellular free iron pool.Pulliainen et al. recently identified a Dpr homologue in Streptococcussuis as an iron-binding protein and demonstrated, by introducingthe site-mutated alleles of Dpr into an S. suis dpr mutant,that the iron-binding ability of the protein is required forhydrogen peroxide resistance (28). The result is in agreementwith our present study with S. mutans. It is now evident thatferritin-like iron binding is a common feature of Dps familyproteins (6, 8, 16, 28, 30, 32, 37, 38). It is therefore notsurprising that Dps family proteins could potentially participatein the regulation and detoxification of intracellular free ironpool, as we described here. The presence of Dpr and Dps familyproteins may prevent an iron-catalyzed Fenton reaction and,as a consequence, play an important role in oxygen tolerance,especially in catalase-deficient bacteria such as S. mutansand S. suis.

The addition of catalase to the medium restored the survivalof the dpr mutant under air (Fig. 5), indicating the presenceof hydrogen peroxide in the cells under this condition. Severallactic acid bacteria are known to accumulate hydrogen peroxidein the medium via the action of hydrogen peroxide-generatingenzymes such as pyruvate oxidase and NADH oxidase (7, 22, 26).S. mutans lacks pyruvate oxidase but has hydrogen peroxide-formingNADH oxidase (Nox-1) (2, 13). Although Nox-1 functions as acomponent of bicomponents, peroxidase and the AhpC component(27), the expression of only Nox-1 (absence of AhpC) under conditionsdescribed previously (13) might allow the bacterium to producehydrogen peroxide.

An interesting finding of the present study is that S. mutanscells contained significant amounts of intracellular iron (Fig.3), particularly since lactic acid bacteria including streptococciare believed to require little or no iron for growth (5). Thetotal iron contents in S. mutans (from 0.005 to 0.008% in dryweight) were some 2.5- to 4-fold less than that in E. coli grownin rich medium (1). Although the iron requirement in S. mutansreportedly depends on growth conditions (20), iron assimilationcould facilitate metabolism, e.g., for amino acid biosynthesisutilizing the iron-containing protein aconitase (9) or potentiallyfor activating iron-requiring ribonucleotide reductases identifiedin the genome sequence (2).

ACKNOWLEDGMENTS

We are grateful to A. Gruss for critical reading of the manuscript.

This study was supported in part by grants-in-aid for scientificresearch 12876016 and 14656030 from the Japan Society for thePromotion of Science (JSPS), the Nagase Science and TechnologyFoundation, and the Noda Institute for Scientific Research Foundation.Y.Y. was the recipient of a predoctoral fellowship from JSPS.

FOOTNOTES

* Corresponding author. Mailing address: Department of Microbial Biotechnology, Graduate School of Agricultural Science, Tohoku University, Aoba-ku, Sendai 981-8555, Japan. Phone: 81-22-717-8779. Fax: 81-22-717-8780. E-mail: ykamio{at}biochem.tohoku.ac.jp.

Present address: Unité de Recherches Laitièreset Génétique Appliquée—URLGA, InstitutNational de la Recherche Agronomique, Domaine de Vilvert, Jouyen Josas, France.

REFERENCES

Top