PerR-Regulated Manganese Ion Uptake Contributes to Oxidative Stress Defense in an Oral Streptococcus

Abstract

Metal homeostasis plays a critical role in antioxidative stress. Streptococcus oligofermentans, an oral commensal facultative anaerobe lacking catalase activity, produces and tolerates abundant H₂O₂, whereas Dpr (an Fe²⁺-chelating protein)-dependent H₂O₂ protection does not confer such high tolerance. Here, we report that inactivation of perR, a peroxide-responsive repressor that regulates zinc and iron homeostasis in Gram-positive bacteria, increased the survival of H₂O₂-pulsed S. oligofermentans 32-fold and elevated cellular manganese 4.5-fold. perR complementation recovered the wild-type phenotype. When grown in 0.1 to 0.25 mM MnCl₂, S. oligofermentans increased survival after H₂O₂ stress 2.5- to 23-fold, and even greater survival was found for the perR mutant, indicating that PerR is involved in Mn²⁺-mediated H₂O₂ resistance in S. oligofermentans. Mutation of mntA could not be obtained in brain heart infusion (BHI) broth (containing ∼0.4 μM Mn²⁺) unless it was supplemented with ≥2.5 μM MnCl₂ and caused 82 to 95% reduction of the cellular Mn²⁺ level, while mntABC overexpression increased cellular Mn²⁺ 2.1- to 4.5-fold. Thus, MntABC was identified as a high-affinity Mn²⁺ transporter in S. oligofermentans. mntA mutation reduced the survival of H₂O₂-pulsed S. oligofermentans 5.7-fold, while mntABC overexpression enhanced H₂O₂-challenged survival 12-fold, indicating that MntABC-mediated Mn²⁺ uptake is pivotal to antioxidative stress in S. oligofermentans. perR mutation or H₂O₂ pulsing upregulated mntABC, while H₂O₂-induced upregulation diminished in the perR mutant. This suggests that perR represses mntABC expression but H₂O₂ can release the suppression. In conclusion, this work demonstrates that PerR regulates manganese homeostasis in S. oligofermentans, which is critical to H₂O₂ stress defenses and may be distributed across all oral streptococci lacking catalase.

INTRODUCTION

Oxidative stress is encountered by all organisms on earth (1, 2), as deleterious reactive oxygen species (ROS), including superoxide ion (O₂⁻), hydroxyl radical (HO·), and hydrogen peroxide (H₂O₂), are generated when molecular oxygen participates in electron transfer reactions or is auto-oxidized by reduced enzyme cofactors (2–4). Therefore, organisms have developed various mechanisms to protect against ROS, such as detoxifying enzymes (5, 6). Aerobes generally employ the superoxide dismutase (SOD)-catalase cascade to eliminate cellular ROS (4–7). Anaerobes use the superoxide reductase (SOR)-dependent oxide decomposition pathway to reduce O₂⁻ to H₂O (8). Intact H₂O₂ is not particularly deleterious to cells, but it can react with cellular Fe²⁺ via Fenton chemistry to form highly toxic HO· (2). Ferritin and miniferritin (Dps family) proteins chelate Fe²⁺ to prevent the deleterious Fenton reaction (9). Moreover, not only is manganese ion (Mn²⁺) a cofactor of SOD (7, 10), but it does not undergo Fenton chemistry (11). Rather, it acts as an inorganic enzyme to dismute O₂⁻ by complexing with small molecules in bacterial cells (11–13). Hence, ROS damage to cells is intimately related to metal ion homeostasis (6).

Expression of antioxidant genes is controlled by redox-sensing transcriptional factors. A Fur family regulator, peroxide-responsive regulator (PerR), is the prototype of the redox-sensing transcriptional repressor found mainly in Gram-positive bacteria (6, 14, 15). Upon sensing cellular H₂O₂, PerR::Fe derepresses the transcription of antioxidant genes, including catalase, alkylhydroperoxide reductase (AhpC/AhpF), heme biosynthesis enzyme, and ferrous and zinc ion homeostatic genes (6). For example, the PerR protein of Bacillus subtilis regulates oxidant-detoxifying enzymes and metal ion homeostatic genes, namely, the Fe²⁺ binding protein MrgA (a Dps family homolog) and the Zn²⁺ transporter ZosA (6, 16, 17). PerR is also involved in Fe²⁺ and Zn²⁺ homeostasis and the oxidative-stress response in Streptococcus pyogenes (18, 19).

Streptococci are facultative anaerobes performing fermentative metabolism. During aerobic growth, they not only convert O₂⁻ to H₂O₂ via SOD, but produce H₂O₂ catalyzed by various oxidases (20–24). However, they do not possess the main H₂O₂-degrading enzyme catalase, and the identified peroxidases, such as AhpC/AhpF and glutathione peroxidase, also contribute little to H₂O₂ resistance in streptococci (20, 25). Thus, streptococci must employ unique mechanisms for H₂O₂ stress defense. Recent studies demonstrate that metal ion homeostasis is pivotal for H₂O₂ defense in streptococci (6). A dps-like nonspecific DNA-binding protein (Dpr), which chelates Fe²⁺, and the Zn²⁺ pump protein PmtA contribute to oxidative-stress resistance in streptococci (18, 26, 27). In addition, inactivation of Mn²⁺ transporter genes in some streptococci diminishes H₂O₂ and O₂⁻ resistance (28–30). Both dpr and pmtA are reported to be regulated by PerR (18, 26); however, how Mn²⁺ uptake is responsive to oxidative stress remains unclear.

Oral streptococci inhabit dental plaque biofilms, where dynamic interspecies interaction occurs and the outcome of competition determines oral health. The oral cavity is an environment under fluctuating oxidative stress. The ability of oral streptococci to defend against oxidants determines whether they can win the interspecies competition and therefore determines oral health. Streptococcus oligofermentans is isolated from noncariogenic dental plaque in humans (31). It generates copious H₂O₂ via multiple pathways (22–24) and inhibits the growth of the dental caries pathogen Streptococcus mutans. Compared to other streptococci, S. oligofermentans has the greatest H₂O₂ tolerance (32), making it a model organism in studying bacterial anti-oxidative-stress mechanisms. Although the Dpr protein plays a role in H₂O₂ resistance in S. oligofermentans (32), the dpr mutant still retains partial H₂O₂ tolerance, indicating that other mechanisms may confer oxidant resistance. Here, by using physiological, biochemical, and genetic approaches, we demonstrate that manganese is important for S. oligofermentans H₂O₂ tolerance and that, in the presence of H₂O₂, PerR releases repression of the manganese transporter mntABC and the Fe²⁺-chelating protein dpr genes.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. All Streptococcus strains were routinely incubated in brain heart infusion (BHI) broth (Difco, Detroit, MI) at 37°C as a static culture under low oxygen levels or anaerobically under 100% N₂. BHI agar (1.5% [wt/vol]) plates were used to select mutant strains and count colonies. Antibiotics (1 mg ml⁻¹ kanamycin and 1 mg ml⁻¹ spectinomycin) were added to the BHI medium when necessary. The Escherichia coli strains (33) were grown in Luria-Bertani (LB) medium at 37°C with shaking and were used for plasmid amplification. When needed, kanamycin (50 μg ml⁻¹) and spectinomycin (250 μg ml⁻¹) were added for recombinant selection.

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Characteristics and descriptiona	Reference or source
E. coli
DH5α	supE44 lacU169 (ϕ80d lacZΔM15) hsdR17 recA1 endA1 gyrA96 thi-1 relA1 luxS	33
S. oligofermentans
Wild type	AS 1.3089; Kans Sps	31
ΔperR	AS 1.3089 perR::Kan; Kanr; AS 1.3089 with perR deletion	This study
ΔmntA	AS 1.3089 mntA::Kan; Kanr; AS 1.3089 with mntA deletion	This study
perR-com	AS 1.3089 perR::Kan pDL278-perR; Kanr Spr ΔperR with perR complement	This study
mntABC-exp	AS 1.3089 pDL278-mntABC; Spr; AS 1.3089 with mntABC ectopic expression	This study
PmntABC::luc	AS 1.3089 pFW5-PmntABC-luc; Spr; AS 1.3089 with PmntABC::luc fusion	This study
ΔperR PmntABC::luc	AS 1.3089 perR::Kan pFW5-PmntABC-luc; Kanr Spr ΔperR with PmntABC::luc fusion	This study
Plasmids
pALH124	Kanr	38
pDL278	Spr	40
pFW5-luc	Spr	41
pDL278-perR	Spr; pDL278 with AS 1.3089 perR gene under its inherent promoter	This study
pDL278-mntABC	Spr; pDL278 with AS 1.3089 mntABC gene under its inherent promoter	This study
pFW5-PmntABC-luc	Spr; pFW5-luc with AS 1.3089 mntABC promoter	This study

^aKan^r, kanamycin resistant; Sp^r, spectinomycin resistant.

DNA manipulation.

Standard recombinant DNA techniques were used for plasmid construction and PCR product ligation. All restriction and ligation enzymes were purchased from New England BioLabs (Beverly, MA). S. oligofermentans genomic DNA was extracted and purified using the method of Marmur (34) with slight modifications (35). All primers (see Table S1 in the supplemental material) were designed according to the complete genome sequence of S. oligofermentans (36) and synthesized by Sangon Company (Shanghai, China). PCR amplifications were performed with KOD-Plus-Neo (Toyobo, Japan), and purification of PCR products was carried out with a Qiagen (Valencia, CA) QIAquick PCR Purification Kit. DNA extracted from agarose gels was purified with a Tiangen (Beijing, China) Tiangel Midi Purification Kit, and plasmids were extracted and purified with a Tiangen (Beijing, China) Tianprep Mini Plasmid Kit.

Construction of mutant strains.

Peroxide-responsive repressor (perR) and metal ABC transporter substrate-binding lipoprotein (mntA) gene deletion mutants were constructed by the PCR ligation method (37). Briefly, two ∼550-bp fragments of the upstream and downstream sequences of the perR and mntA genes, respectively, were amplified by PCR using S. oligofermentans genomic DNA as a template. The purified PCR products were digested with BamHI. The nonpolar kanamycin resistance gene cassette was cut from plasmid pALH124 (38) by digestion with BamHI. All three fragments were purified and mixed at a 1:1:1 molar ratio. A fused fragment was formed by T4 DNA ligase treatment and transformed into the S. oligofermentans wild-type strain using a published method (39). Transformants were selected on BHI (for the perR mutant) or BHI supplemented with 0.1 mM MnCl₂ (for the mntA mutant) agar plates containing 1 mg ml⁻¹ kanamycin. The corresponding gene deletion was confirmed with PCR and sequencing.

perR-complemented and mntABC overexpression strains were constructed as described below. The entire gene fragments, including the coding region and the promoter sequence of perR and mntABC, were amplified from chromosomal DNA of S. oligofermentans by PCR with the primers listed in Table S1 in the supplemental material. Then, 762- and 2,660-bp PCR products were purified and double digested with EcoRI and SalI. After gel purification, the PCR products were inserted into E. coli-Streptococcus shuttle plasmid pDL278 (40), which was cut with the same enzymes. Positive transformants were selected on LB agar plates containing 250 μg ml⁻¹ spectinomycin and identified by PCR and sequencing. The recombinant plasmids pDL278-perR and pDL278-mntABC were then transformed into the S. oligofermentans perR mutant and wild-type strains, respectively. Transformants were selected on BHI agar plates containing 1 mg ml⁻¹ kanamycin and 1 mg ml⁻¹ spectinomycin or 1 mg ml⁻¹ spectinomycin and identified by PCR and sequencing.

Construction of luciferase reporter strains.

For construction of the PmntABC-luc reporter strain, DNA fragments corresponding to a 430-bp sequence upstream of the start codon of the mntABC gene were amplified from chromosomal DNA using PCR with primers listed in Table S1 in the supplemental material. The purified PCR product was double digested with BamHI and NheI, gel purified, and ligated to plasmid pFW5-luc (41), which was digested with the same enzyme. The ligation mixtures were transformed into E. coli DH5α. Positive transformants were identified by PCR and sequencing. The recombinant plasmid pFW5-PmntABC-luc was then transformed into the S. oligofermentans wild-type strain and the perR mutant, and positive transformants were identified by PCR, sequencing, and luciferase activity.

Assay of hydrogen peroxide sensitivity.

S. oligofermentans strains were grown in BHI broth under static or anaerobic conditions at 37°C overnight. The overnight cultures were diluted 1:100 into fresh BHI broth and incubated under static or anaerobic conditions. When the optical density at 600 nm (OD₆₀₀) reached 0.6 to 0.7, cells (1 ml) were removed and harvested by centrifugation. After two washings with phosphate-buffered saline (PBS), the cells were resuspended in 1 ml of fresh BHI broth. Cell aliquots (200 μl) were distributed into 1.5-ml Eppendorf tubes. One aliquot was challenged with 20 mM H₂O₂, and an aliquot without H₂O₂ treatment was used as a control. After incubation at 37°C for 10 min, the cells were collected, washed twice with PBS buffer, and resuspended in 200 μl BHI broth. Cell chains were separated by sonication for 30 s with a UP 200S sonicator (Germany). The samples were then serially 10-fold diluted. Appropriate dilutions were plated on BHI agar, and CFU were counted after 24 h of incubation in a candle jar at 37°C. Survival (percent) was calculated as the ratio of CFU in the H₂O₂-challenged sample to those in controls. Experiments were executed in triplicate, and each was repeated at least three times independently.

Metal content measurement.

Concentrations of iron, manganese, and zinc in static and anaerobic cultures of various S. oligofermentans strains were measured using inductively coupled plasma mass spectrometry (ICP-MS). Overnight BHI cultures of tested strains were diluted 1:50 in fresh BHI broth and incubated at 37°C under static or anaerobic conditions. Mid-log-phase cells were harvested by centrifugation at 13,400 × g for 10 min. The cell pellets were washed twice in PBS with 1 mM EDTA and once in PBS without EDTA and then resuspended in 1 ml of PBS. One hundred microliters of suspension was used to measure the protein concentration with a bicinchoninic acid (BCA) protein analysis kit according to the manufacturer's recommendations. The remaining 900 μl of suspension was collected by centrifugation at 13,400 × g for 10 min. The pelleted bacterial cells were resuspended in 500 μl of nitric acid (ultrapure). After overnight incubation at room temperature, the cell suspension was brought to 1.5 ml with deionized distilled water. Then, metal ions were analyzed by ICP-MS (DRCII; PerkinElmer) at Beijing University Health Science Center. Beryllium, indium, and uranium standard solutions (PerkinElmer; National Institute of Standards and Technology [NIST] certified) were used to calibrate the ICP-MS. Experiments were conducted in triplicate, and each was repeated at least three times. The metal content was expressed in nmol per mg protein.

Luciferase activity measurement.

Twenty-five microliters of 1 mM d-luciferin (Sigma-Aldrich, St. Louis, MO) solution (suspended in 1 mM citrate buffer, pH 6.0) was added to 100-μl samples, and luciferase activity assays were performed as previously described (39) using a TD 20/20 luminometer (Turner Biosystems, Sunnyvale, CA). The sample optical density (OD₆₀₀) was measured with a 2100 visible spectrophotometer (Unico, Shanghai, China) and used to normalize the luciferase activity. All measurements were performed in triplicate, and all experiments were repeated at least three times.

Hydrogen peroxide measurement.

H₂O₂ in liquid culture was measured as described previously (24). Briefly, 650 μl of culture supernatant was added to 600 μl of solution containing 2.5 mM 4-amino-antipyrine (4-amino-2,3-dimethyl-1-phenyl-3-pyrazolin-5-one; Sigma) and 0.17 M phenol. The reaction proceeded for 4 min at room temperature; horseradish peroxidase (Sigma) was then added to a final concentration of 50 mU/ml in 0.2 M potassium phosphate buffer (pH 7.2). After 4 min of incubation at room temperature, the OD₅₁₀ was measured with a Unico (Shanghai, China) 2100 visible-light spectrophotometer. A standard curve was generated with known concentrations of chemical H₂O₂.

Quantitative PCR.

Total RNA was extracted from mid-log-phase (OD₆₀₀ = 0.4 to 0.5) cells using TRIzol reagent (Invitrogen, Carlsbad, CA) as recommended by the suppliers. After quality confirmation with a 1% agarose gel, the RNA was treated with RNase-free DNase (Promega, Madison, WI) and analyzed by PCR for possible chromosomal DNA contamination. cDNA was generated from 2 μg total RNA with random primers using Moloney murine leukemia virus reverse transcriptase (Promega) according to the supplier's instructions and used for quantitative-PCR (qPCR) amplification with the corresponding primers (see Table S1 in the supplemental material). Amplifications were performed with a Mastercycler ep realplex² (Eppendorf, Germany). To estimate copy numbers for a given mRNA, a standard curve of the tested gene was generated by quantitative PCR using 10-fold serially diluted PCR product as the template. The 16S rRNA gene was used as the biomass reference. The copy number of each gene was normalized to the number of 16S rRNA copies. The number of copies of the transcript of each gene per 1,000 16S rRNA copies is shown.

RESULTS

PerR functions as a peroxide-responsive repressor in S. oligofermentans.

To identify perR homologs in S. oligofermentans, the perR gene from B. subtilis was used as a probe to query the complete genome. An open reading frame (I872_05555) annotated as “ferric transport regulator protein” with 38% amino acid identity was hit, and it was noted as PerR. Furthermore, PerR orthologs were found in almost all oral streptococci. PerR from S. oligofermentans had the highest identity (97%) with that of Streptococcus cristatus and had 58 to 87% identity with other oral streptococci. Amino acid sequence alignment of PerR proteins from representative oral streptococci and B. subtilis (Fig. 1) revealed that the essential amino acid residues for metal ion binding (H37, H91, D85, and D104 for manganese or ferric ions; C96, C99, C136, and C139 for zinc ions) in B. subtilis (15) were conserved in all oral streptococcal PerR proteins examined except H93, where an asparagine (N) is found in four streptococcal species, namely, S. oligofermentans, Streptococcus gordonii, Streptococcus sanguinis, and Streptococcus salivarius. Sequence conservation suggests that the oral streptococcal PerR proteins function similarly to that of B. subtilis.

FIG 1.

Sequence alignment of PerR proteins from S. oligofermentans, B. subtilis, and other representative species of oral streptococci. Amino acid sequences of PerR were retrieved from the protein database of NCBI. *, conserved amino acid residue essential for manganese or ferric ion binding in B. subtilis; #, conserved cysteine residue for zinc ion binding. BSU08730, B. subtilis; soi_I872_05555, S. oligofermentans; SGO_0703, S. gordonii; SSA_0686, S. sanguinis; Ssal_01385, S. salivarius; SMU_593, S. mutans. Black shading indicates the homology level is 100%; gray shading indicates the homology level is ≥75%.

To identify the function of PerR in S. oligofermentans, a perR deletion mutant was constructed and verified by PCR and sequencing. Both the perR mutant and the wild-type strain were cultured under anaerobic condition to exclude endogenous H₂O₂ production. Mid-log-phase cultures (OD₆₀₀ = 0.6 to 0.7) were collected and subjected to H₂O₂ pulsing (20 mM) for 10 min. Then, the survival rate was quantified as described in Materials and Methods. As shown in Fig. 2, perR mutant survival (22.05%) was ∼32-fold higher than that of the wild-type strain (0.67%). Furthermore, a perR-complemented strain, perR-com, showed significantly reduced survival (0.34%) compared with the perR mutant and similar to that of the wild-type strain. Thus, PerR has been demonstrated to be a peroxide-responsive repressor in S. oligofermentans.

FIG 2.

Survival of S. oligofermentans wild-type, perR mutant, and perR-complemented strains after H₂O₂ pulsing. The strains were incubated statically (gray bars) and anaerobically (black bars), and mid-log-phase cultures were collected and subjected to 20 mM H₂O₂ pulsing for 10 min. Viable cells were counted based on CFU on a BHI agar plate after serial dilution. Survival was calculated as the ratio of CFU in H₂O₂-pulsed samples to those without H₂O₂ treatment. The data are expressed as means ± standard deviations of three independent experiments. *, data are statistically significant in comparison to values of anaerobically cultured wild-type and perR-com strains as verified by Student's t test (P < 0.01); #, data are statistically significant compared to the values of the respective strains cultured anaerobically, as verified by Student's t test (P < 0.05).

perR deletion causes increased cellular manganese.

Because PerR regulates metal ion homeostasis in B. subtilis (6), its involvement in cellular metal ion balance in S. oligofermentans was examined. The wild-type, perR mutant, and perR-com strains were cultured anaerobically in BHI broth, and mid-log-phase cells (OD₆₀₀ = 0.6 to 0.7) were collected to measure cellular metal by ICP-MS. Blank BHI broth was also included to determine the baseline metal content. As shown in Table 2, Mn²⁺ was 4.5-fold higher in the perR mutant, while its levels were similar in perR-com and the wild-type strain. However, iron and zinc were at similar levels in the three strains. This indicates that PerR specifically downregulates cellular manganese, likely by depressing the expression of Mn²⁺ transporters.

TABLE 2.

Cytoplasmic metal ion concentrations in various S. oligofermentans strains cultured statically or anaerobically

Strain	Metal ion concna
	Static culture			Anaerobic culture
	Mn	Fe	Zn	Mn	Fe	Zn
Wild type	4.13 ± 0.96d	7.77 ± 1.42	2.37 ± 0.49	0.64 ± 0.15	6.33 ± 1.20	7.78 ± 1.50
perR mutant	4.59 ± 0.19	6.97 ± 0.56	2.12 ± 0.01	2.88 ± 0.29d	7.17 ± 1.70	6.21 ± 0.70
perR-com	ND	ND	ND	0.58 ± 0.06	6.96 ± 0.27	7.30 ± 1.40
mntA mutantb	0.44 ± 0.04e	ND	ND	0.33 ± 0.07e	ND	ND
mntABC-exp	8.82 ± 0.21f	ND	ND	2.86 ± 0.81d	ND	ND
Wild type-2.5Mnb	9.02 ± 0.34	ND	ND	1.86 ± 0.33	ND	ND
Wild type-Mnc	ND	ND	ND	6.85 ± 0.09d	ND	ND
perR mutant-Mnc	ND	ND	ND	9.80 ± 0.67g	ND	ND
mntABC-exp-Mnc	ND	ND	ND	15.46 ± 3.6g	ND	ND

^aData are the means ± standard deviations of three independent cultures; metal content is expressed as nmol/mg protein. ND, not determined.

^bStrain grew in BHI broth with the addition of 2.5 μM MnCl₂.

^cStrain grew in BHI broth with 0.1 mM MnCl₂ added.

^dValue with significant difference from the anaerobically incubated wild-type strain (P < 0.05; Student's t test).

^eValue with significant difference from the wild-type strain growing in 2.5 μM MnCl₂ (wild type-2.5Mn) (P < 0.01; Student's t test).

^fValue with significant difference from the statically incubated wild-type strain (P < 0.05; Student's t test).

^gValue with significant difference from the anaerobically incubated wild-type strain growing in 0.1 mM Mn²⁺ (wild type-Mn) (P < 0.01; Student's t test).

Mn²⁺ is important for H₂O₂ tolerance in S. oligofermentans.

To elucidate the role of cellular Mn²⁺ in H₂O₂ tolerance, the S. oligofermentans wild-type strain was cultured anaerobically in BHI broth supplemented with MnCl₂ at different concentrations. Mid-log-phase cultures (OD₆₀₀ = 0.6 to 0.7) were collected and subjected to H₂O₂ pulsing as described above. The survival of S. oligofermentans increased 2.5- and 23-fold upon addition of 0.1 and 0.25 mM MnCl₂, respectively (Fig. 3). In addition, 10.7-fold more manganese was detected in cells grown with 0.1 mM MnCl₂ than in cultures without MnCl₂ supplementation (Table 2). These data suggest that Mn²⁺ plays a significant role in the H₂O₂ tolerance of S. oligofermentans. Furthermore, addition of 0.25 mM MnCl₂ enhanced the shaking growth of S. oligofermentans by ∼35%, similar to that with catalase addition (200 U/ml); this confirms that Mn²⁺ can be an H₂O₂ scavenger in S. oligofermentans.

FIG 3.

Impact of manganese ion on the H₂O₂-stressed survival of S. oligofermentans wild type and perR mutant. S. oligofermentans strains were anaerobically cultured in BHI broth with or without 0.1 and 0.25 mM MnCl₂. After H₂O₂ pulsing, viable cells were counted, and survival was quantified as described for Fig. 2. The data are expressed as means ± standard deviations of three independent experiments. Gray bars, wild-type strain; black bars, perR mutant. *, data are statistically significant in comparison to the values of the wild-type strain growing in BHI broth, as verified by Student's t test (P < 0.05); #, data are statistically significant compared to the values of the wild-type strain growing in the same medium, as verified by Student's t test (P < 0.05).

perR inactivation increases Mn²⁺-assisted H₂O₂ survival of S. oligofermentans.

To find the possible linkage between PerR and Mn²⁺ in H₂O₂ resistance, an S. oligofermentans perR mutant was cultured anaerobically in BHI broth supplemented with 0.1 and 0.25 mM MnCl₂ and then tested for H₂O₂ survival as described above. The results showed that perR inactivation increased H₂O₂ survival by 66% and 54% upon addition of 0.1 and 0.25 mM Mn²⁺, respectively, which was more than the increase in H₂O₂ survival (21%) in BHI broth (Fig. 3). Accordingly, addition of 0.1 mM MnCl₂ increased cellular manganese more in the perR mutant than in the wild-type strain (Table 2). This indicates that PerR regulates manganese ion uptake, which contributes to H₂O₂ resistance of S. oligofermentans.

mntABC encodes a high-affinity Mn²⁺ transporter in S. oligofermentans.

To identify genes responsible for transporting Mn²⁺ in S. oligofermentans, two manganese transporter genes, scaCBA and mntH, which, respectively, belong to the manganese ABC transporter and the eukaryotic Nramp (natural resistance associated macrophage protein) families (42, 43), from S. gordonii were used as probes to query the complete genome of S. oligofermentans. Two targets were found: I872_09645-09655, a gene cluster encoding a putative manganese ABC transporter complex, designated mntABC (I872_09645, metal ABC transporter substrate-binding lipoprotein; I872_09650, ABC transporter membrane-spanning permease–manganese transport; I872_09655, ATP binding protein), and I872_08215, encoding a putative manganese transporter, Nramp, named mntH. The amino acid sequences of MntA, MntB, MntC, and MntH of S. oligofermentans shared 93%, 99%, 89%, and 89% identity with ScaA, ScaB, ScaC, and MntH from S. gordonii, respectively.

To measure the activity of the two manganese transport-related genes in S. oligofermentans, gene expression was measured in wild-type cells cultured statically or anaerobically in BHI broth. qPCR determined that in mid-log-phase cells (OD₆₀₀ = 0.6 to 0.7), mntA expression was higher in all the cultures than mntH expression (Table 3), indicating that mntABC may be the main manganese ion transporter in S. oligofermentans. Both mntA and mntH increased (3.3- and 4.3-fold) their expression in the static culture, suggesting that the oxidative state induces the expression of the Mn²⁺ transporter genes mntABC and mntH.

TABLE 3.

Transcript levels of mntA, mntH, and dpr in the S. oligofermentans wild-type, perR mutant, and perR-complemented strains cultured statically or anaerobically

Gene	Transcript levela
	Wild type		perR mutant		perR-com
	Static	Anaerobic	Static	Anaerobic	Static	Anaerobic
mntA	14.86 ± 3.02	4.49 ± 1.37b	11.88 ± 1.69	25.86 ± 8.13	ND	4.27 ± 0.74b
mntH	0.26 ± 0.07	0.06 ± 0.01c	0.30 ± 0.02	0.13 ± 0.06	ND	ND
dpr	6.17 ± 1.49	0.58 ± 0.15b	15.90 ± 2.81	23.30 ± 8.81	ND	0.06 ± 0.01b,d

^aData are shown as means ± standard deviations of three independent experiments. Gene expression is shown as the means ± standard deviations of copy number/0.001 16S rRNA gene copy. ND, not determined.

^bData are statistically significant compared to the corresponding genes in a statically cultured wild-type strain and a statically or anaerobically cultured perR mutant; Student's t test (P < 0.05).

^cData are statistically significant compared to the corresponding genes in a statically cultured wild-type strain; Student's t test (P < 0.05).

^dData are statistically significant compared to the corresponding genes in an anaerobically cultured wild-type strain; Student's t test (P < 0.05).

Inactivation of mntA could be achieved in BHI culture of S. oligofermentans only when supplemented with an additional ≥2.5 μM Mn²⁺. Since only 418 nM Mn²⁺ was detected in BHI broth, MntABC is predicted to be essential for S. oligofermentans in a low-Mn²⁺ environment. To identify the function of MntABC in the uptake of Mn²⁺, the wild-type strain and the mntA mutant were grown in BHI broth supplemented with 2.5 μM Mn²⁺, and mid-log-phase cells (OD₆₀₀ = 0.6 to 0.7) were collected to measure cellular metal by ICP-MS. Compared to the wild-type strain, Mn²⁺ decreased by 95% and 82% in the statically and anaerobically cultured mntA mutant (Table 2), respectively, confirming that MntABC functions as a high-affinity Mn²⁺ transporter in S. oligofermentans.

Furthermore, 2.1- and 4.5-fold-greater manganese levels were found in a strain in which mntABC is ectopically expressed (mntABC-exp) when it was statically and anaerobically cultured, respectively (Table 2), verifying the metal-transporting function of MntABC in S. oligofermentans.

MntABC inactivation reduces the oxidative-stress tolerance of S. oligofermentans.

To investigate the role of the Mn²⁺ transporter MntABC in H₂O₂ tolerance in S. oligofermentans, both the wild-type strain and the mntA mutant were cultured statically in BHI broth supplemented with 2.5 μM Mn²⁺. Mid-log-phase cells (OD₆₀₀ = 0.6 to 0.7) were collected and subjected to H₂O₂ pulsing as described above. Compared to the wild-type strain, mntA mutation reduced H₂O₂ survival 5.7-fold (Fig. 4), indicating that MntABC-mediated Mn²⁺ uptake plays a role in protecting S. oligofermentans from H₂O₂ attack.

FIG 4.

Role of MntABC-mediated manganese uptake in H₂O₂ stress tolerance in S. oligofermentans. The wild-type strain (1) and mntA mutant (2) were grown statically in BHI broth with addition of 2.5 μM MnCl₂, and the wild-type strain (3) and mntABC-exp (4) were grown anaerobically in BHI broth with supplementation of 0.1 mM MnCl₂. After H₂O₂ pulsing for 10 min, viable cells were counted, and survivors were quantified as described for Fig. 2. The data are expressed as means ± standard deviations of three independent experiments. *, data are statistically significant in comparison to the wild-type strain incubated under the same conditions, as verified by Student's t test (P < 0.01).

To further verify the role of MntABC in H₂O₂ resistance, the mntABC-exp strain was grown in BHI broth supplemented with 0.1 mM MnCl₂ and anaerobically incubated. Cells at mid-log phase were treated with H₂O₂ (20 mM) for 10 min. About 12-fold-higher survival was determined for the mntABC-exp strain than for the wild-type strain, and cellular manganese was significantly increased in the mntABC-exp strain, as well (Table 2).

Inactivation of perR upregulates mntABC and dpr expression.

Increased Mn²⁺ was found in the perR mutant, so correlations among perR, mntABC, and mntH were tested in anaerobically cultured wild-type, perR mutant, and perR-com strains. Using qPCR, 5.8-fold upregulation of mntA was detected in the perR mutant, while similar expression levels were found in perR-com and the wild-type strain. Similarly, dpr, encoding an Fe²⁺-chelating protein, was upregulated 40-fold in the perR mutant, whereas perR complementation greatly reduced dpr expression to as much as ∼10-fold lower than that in the wild-type strain (Table 3). This shows that PerR represses expression of both mntABC and dpr. However, no significant change in mntH expression was detected in the perR mutant (Table 3), indicating that mntH is not subjected to PerR regulation.

Furthermore, a putative PerR-binding sequence “Per box” (AATTAGAAGCATTATAATT) was found in the promoter region of dpr but not in that of mntABC. Electrophoretic mobility shift assay (EMSA) detected PerR's binding only to the dpr promoter (see Fig. S1 in the supplemental material) but not to mntABC (data not shown). This suggests direct regulation of dpr by PerR but indirect regulation of mntABC in S. oligofermentans.

H₂O₂ releases PerR repression of mntABC.

Since an effect of perR mutation on cellular manganese, H₂O₂ survival, and mntABC expression was not observed in statically cultured S. oligofermentans (Tables 2 and 3 and Fig. 2) and abundant H₂O₂ was produced only in statically cultured cells (see Fig. S2 in the supplemental material) while none was detected in anaerobic cultures under the detection limit of 1.90 μM H₂O₂, H₂O₂ may be a substance that interferes with PerR regulation of mntABC. To confirm this, an mntABC luciferase reporter strain was constructed by introducing the integrative plasmid pFW5-PmntABC-luc into the S. oligofermentans wild-type strain and the perR mutant. The constructed S. oligofermentans PmntABC::luc and ΔperR PmntABC::luc strains were cultured anaerobically, and 20 μM H₂O₂, a concentration similar to that produced in statically cultured cells (see Fig. S2 in the supplemental material), was added to the early log-phase cells (OD₆₀₀, ∼0.2). After cultivation for 20, 40, 60, and 80 min, the luciferase activity and OD₆₀₀ were measured. mntABC expression was significantly upregulated in H₂O₂-pulsed cultures (Fig. 5A); however, no significant difference in mntABC expression was found between the H₂O₂-pulsed and nonpulsed perR mutant cells (Fig. 5B). This indicates that H₂O₂-induced mntABC expression depends on the presence of perR. Similar to the qPCR results (Table 3), 9- to 10-fold-elevated mntABC expression was detected in the non-H₂O₂-pulsed perR mutant compared to wild-type cells (Fig. 5), supporting the conclusion that PerR negatively regulates transcription of mntABC.

FIG 5.

H₂O₂ induced mntABC expression in the wild-type strain (A) and perR mutant (B) of S. oligofermentans. Overnight BHI cultures of S. oligofermentans::PmntABC-luc and S. oligofermentans ΔperR::PmntABC-luc were diluted 1:30 in fresh BHI broth and incubated anaerobically. Then, 20 μM H₂O₂ was added to early-log-phase cells (OD₆₀₀ = ∼0.2). Cells were collected at 20, 40, 60, and 80 min. The OD₆₀₀ and luciferase activity (relative light units [RLU]) were measured as described in Materials and Methods. mntABC expression is expressed as RLU/OD₆₀₀ unit. Gray bars, no H₂O₂ addition; black bars, 20 μM H₂O₂ addition. The experiments were repeated 3 times, and the data are expressed as means ± standard deviations of three reads of each independent experiment. *, data are statistically significant in comparison to the values of the wild-type strain growing in BHI broth without H₂O₂ addition at the respective time points, as verified by Student's t test (P < 0.01).

DISCUSSION

The oral commensal S. oligofermentans produces ample H₂O₂ (4.6 mM) via multiple pathways. In particular, lactate oxidase activity enables it to inhibit the caries pathogen S. mutans (22–24). S. oligofermentans is also resistant to higher H₂O₂ levels (5.5 mM) than other bacterial species, although it has no catalase (22). However, the activity of the ferric iron-chelating protein Dpr is insufficient for such higher H₂O₂ tolerance, as demonstrated by our previous work (32). Here, we report that S. oligofermentans contains extraordinarily high concentrations of manganese, which acts synergistically with Dpr to enable greater H₂O₂ tolerance than in other bacteria. Figure 6 depicts a model of anti-oxidative stress in S. oligofermentans. In the presence of endogenous or exogenous H₂O₂, PerR derepressed the expression of the manganese transporter mntABC and ferric iron-chelating protein dpr genes. MntABC facilitates Mn²⁺ internalization. Mn²⁺ can act as a cofactor of SOD or a catalase-like inorganic catalyzer; it can also substitute for Fe²⁺ in protein active sites to reduce protein oxidation. Dpr protein, by trapping Fenton reaction-stimulating Fe²⁺, prevents inert H₂O₂ from converting to highly reactive HO·.

FIG 6.

Diagram depicting the antioxidative mechanisms in S. oligofermentans. Endogenous H₂O₂ produced during aerobic growth or exogenous H₂O₂ releases PerR's inhibition of the Mn²⁺ transporter gene mntABC and the Fe²⁺-chelating protein gene dpr. Mn²⁺ might protect S. oligofermentans from O₂⁻ by activating SOD and from H₂O₂ by unknown mechanisms. Dpr prevents Fenton reactions by chelating Fe²⁺, avoiding the production of toxic HO·.

Fe and Mn are essential trace metal elements for bacteria, although the cellular Fe level is usually higher than that of Mn, e.g., the Mn/Fe ratio of 0.0072 (0.0197 nmol Mn and 2.72 nmol Fe/mg protein) found in E. coli (44). It has been reported that the poor Fenton reagent metal ions Mn²⁺ and Zn²⁺ prevent H₂O₂-derived HO· production (11), and a higher Mn/Fe ratio (0.24) contributes to oxidative-stress and radiation resistance in Deinococcus radiodurans (44). Moreover, the SOD-lacking Lactobacillus plantarum accumulates a high concentration of cellular Mn (20 to 35 mM) to defend against oxidative stress (10, 45). In this work, a high Mn/Fe ratio (0.48) was also detected in statically cultured S. oligofermentans (Table 2), but a high zinc level was not observed. Furthermore, a higher cellular Mn²⁺ level was found in S. oligofermentans (4.13 ± 0.96 nmol/mg protein) than in other streptococci (1.12 ± 0.23 and 1.57 ± 0.14 nmol/mg protein in Streptococcus pneumoniae and S. mutans) in the parallel measurements in this study. Supplementation experiments also confirmed the role of Mn²⁺ in the protection of S. oligofermentans against H₂O₂ (Fig. 3). How Mn²⁺ offers protection against H₂O₂ has been hypothesized. Stadtman et al. predicted a catalase-like activity of the Mn²⁺-bicarbonate complex in scavenging H₂O₂ (46), a finding consistent with the observation in S. oligofermentans that, although high H₂O₂ concentrations are found in the culture media, cellular H₂O₂ is undetectable. The H₂O₂-scavenging activity of Mn²⁺ must be confirmed, as only barely detectable H₂O₂ degradation was found by mixed incubation of MnCl₂ with cell extracts of S. oligofermentans (data not shown).

Mn²⁺ is thought to reduce chemical oxidation of many proteins by substitution for the active Fenton-reactive Fe²⁺ at protein active sites (47). The growth of streptococci is reported to require iron (48), and high cellular iron levels (6 to 8 nmol/mg protein) are found in S. oligofermentans, as well, when cultured in BHI broth (Table 2). This suggests that Mn²⁺ substitution for Fe²⁺ can occur in the oral streptococcus under the current experimental conditions. Released Fe²⁺ might then be chelated by Dpr protein, avoiding Fenton chemistry.

Among the three types of manganese transporters, the manganese ABC transporter is key for bacterial uptake of Mn²⁺ (49). Manganese ABC transporter mutation affects biofilm formation and oxidative-stress resistance of S. gordonii and S. mutans (28, 49, 50), as well as the virulence of a number of bacterial pathogens, including S. pneumoniae, S. pyogenes, and Yersinia pestis (11, 29, 49). The mntA mutant can be obtained only in BHI medium (which contains trace amounts of Mn²⁺ [∼0.4 μM]) supplemented with ≥2.5 μM Mn²⁺. This suggests that MntABC acts as a main Mn²⁺ transporter at low Mn²⁺ levels in S. oligofermentans, while other manganese transporters, such as MntH, might function at relatively high extracellular Mn²⁺ levels. MntABC protects S. oligofermentans, not only from H₂O₂ pulsing (Fig. 4), but also from attack by paraquat, an O₂⁻ producer (data not shown), suggesting that MntABC-transported Mn²⁺ plays an important role in anti-oxidative stress in S. oligofermentans.

In response to H₂O₂ stress, E. coli employs OxyR, a transcriptional activator, to promote expression of the manganese transporter mntH and a Fur family regulator, which controls iron transporter genes (47, 51, 52). Gram-positive bacteria use PerR to repress zinc and iron transporters. In the absence of H₂O₂, PerR::Fe binds to the conservative DNA sequence (Per box) in the promoter regions, while H₂O₂ can oxidize and inactivate PerR::Fe protein, leading to derepression of zinc- and iron-transporting genes (6, 15, 53, 54). PerR in S. oligofermentans has been verified to be a peroxide-responsive repressor, as well (Fig. 2). However, unlike those of B. subtilis and S. pyogenes (6, 18), the PerR protein of S. oligofermentans represses the manganese transporter gene mntABC, in addition to dpr, so enhanced cytoplasmic Mn²⁺ was observed in the perR mutant (Tables 2 and 3 and Fig. 5), and this contributed to H₂O₂ resistance (Fig. 3). In addition, a physiological concentration of H₂O₂-induced mntABC transcription occurs only in the anaerobically cultured wild-type strain, but not in the perR mutant (Fig. 5). This indicates that the PerR protein of S. oligofermentans is involved in H₂O₂-induced mntABC gene expression. Collectively, PerR has been determined to play a pivotal role in sensing the cellular oxidative status and regulating Mn²⁺ and Fe²⁺ homeostasis in S. oligofermentans.

MntR, a member of the DtxR family, has been shown to regulate manganese transporter transcription in response to the cellular manganese concentration (55). This study also found that an MntR ortholog repressed the expression of mntABC in S. oligofermentans; however, mntR expression was not affected by perR mutation (data not shown). These data suggest that mntABC can independently respond to the cell redox status and manganese levels. Detailed regulation by PerR of mntABC and manganese homeostasis is under study in our laboratory through transcriptomics.

perR mutation does not significantly affect the antioxidant phenotype and mntABC expression when S. oligofermentans is statically cultured (Tables 2 and 3 and Fig. 2), suggesting that H₂O₂ produced under low oxygen levels inactivates perR. In accordance with this finding, PerR represses mntABC and dpr much more in anaerobic culture than in statically cultured cells (Table 3), and accordingly, more manganese is measured in the statically cultured wild-type strain (Table 2). More importantly, the survival of statically cultured S. oligofermentans is significantly higher than that in anaerobic culture after high-dose H₂O₂ pulsing (Fig. 2), indicating that antioxidant gene derepression promotes S. oligofermentans resistance to a high level of H₂O₂. Therefore, S. oligofermentans produces H₂O₂, not only as a chemical weapon in interspecies competition, but also as a strategy of “autoimmunity” to defend against more drastic oxidative stress. Orthologs of PerR and MntABC are found in almost all oral streptococci, suggesting that a PerR-regulated Mn²⁺-based antioxidative mechanism can be generally used by these streptococci devoid of catalase. The oral commensals, such as S. sanguinis and S. gordonii, also suppress the growth of S. mutans by producing H₂O₂ (21), and this plays a role in oral health.