Hydrogen Peroxide Produced by Oral Streptococci Induces Macrophage Cell Death

Abstract

Hydrogen peroxide (H₂O₂) produced by members of the mitis group of oral streptococci plays important roles in microbial communities such as oral biofilms. Although the cytotoxicity of H₂O₂ has been widely recognized, the effects of H₂O₂ produced by oral streptococci on host defense systems remain unknown. In the present study, we investigated the effect of H₂O₂ produced by Streptococcus oralis on human macrophage cell death. Infection by S. oralis was found to stimulate cell death of a THP-1 human macrophage cell line at multiplicities of infection greater than 100. Catalase, an enzyme that catalyzes the decomposition of H₂O₂, inhibited the cytotoxic effect of S. oralis. S. oralis deletion mutants lacking the spxB gene, which encodes pyruvate oxidase, and are therefore deficient in H₂O₂ production, showed reduced cytotoxicity toward THP-1 macrophages. Furthermore, H₂O₂ alone was capable of inducing cell death. The cytotoxic effect seemed to be independent of inflammatory responses, because H₂O₂ was not a potent stimulator of tumor necrosis factor-α production in macrophages. These results indicate that streptococcal H₂O₂ plays a role as a cytotoxin, and is implicated in the cell death of infected human macrophages.

Introduction

Members of the oral mitis group of streptococci are causative agents of oral biofilm, dental plaque, and infective endocarditis [1], [2], [3], [4]. Streptococcus oralis, Streptococcus sanguinis, and Streptococcus gordonii are members of the mitis group of oral streptococci and primary colonizers of the human oral cavity [1], [2], [3], [4]. These oral streptococcal species are known to produce hydrogen peroxide (H₂O₂) [1], [2], [5], [6], with the H₂O₂ produced playing important roles in microbial communities such as oral biofilms [6], [7]. S. sanguinis and S. gordonii have been reported to produce H₂O₂ at concentrations sufficient to reduce the growth of many oral bacteria, including the cariogenic Streptococcus mutans [7]. H₂O₂ also stimulates the release of bacterial DNA, which appears to support oral biofilm formation and facilitate gene exchange amongst bacteria [8].

The oral mitis group of streptococci is known to cause a variety of infectious complications, including bacteremia and infective endocarditis [9], [10], [11], [12]. Studies by the United Kingdom’s Health Protection Agency have shown that the rate of bacteremia caused by the mitis group of streptococci is comparable to that of group A or group B streptococci [13]. Furthermore, epidemiological studies have shown the presence of these streptococcal species in heart valve and atheromatous plaque clinical specimens [14], [15], [16].

Macrophages and monocytes are major contributors to host immune responses against bacterial infections. Although oral streptococcal species are known to cause bloodstream infections and infectious endocarditis, their pathogenicity toward macrophages is not well understood. We previously found that S. sanguinis induces foam cell formation and macrophage cell death, and that its cytotoxicity is likely to be associated with reactive oxygen species [17]. Further study suggested that the macrophage cell death is related to H₂O₂ production by the streptococcal species. Although the cytotoxicity of H₂O₂ has been widely recognized, the effects of H₂O₂ produced by oral streptococci on host defense systems remain unknown.

In the present study, we investigated whether H₂O₂ produced by the oral mitis group of streptococci is implicated in infected human macrophage cell death.

Materials and Methods

Bacterial strains and culture conditions

S. oralis ATCC35037, a type strain originally isolated from human mouth [18], was obtained from the Japan Collection of Microorganisms at the RIKEN Bioresource Center (Tsukuba, Japan). S. mutans MT8148 and Streptococcus salivarius HHT were selected from the stock culture collection in the Department of Oral and Molecular Microbiology, Osaka University Graduate School of Dentistry. S. sanguinis SK36 was provided by Dr. M. Killian (Aarhus University, Denmark). These bacteria were cultured in Brain Heart Infusion (BHI) broth (Becton Dickinson, Sparks, MD, USA). Escherichia coli strain XL10-gold (Stratagene, La Jolla, CA, USA) was grown in Luria-Bertani broth.

Cell culture

The human monocyte cell line THP-1 cells were purchased from RIKEN Bioresource Center and cultured in RPMI1640 medium (Invitrogen, Carlsbad, CA, USA) supplemented with 5% fetal bovine serum (FBS) (Invitrogen) (5% FBS RPMI1640), penicillin (100 U/ml), and streptomycin (100 µg/ml) at 37°C in a 5% CO₂ atmosphere. Differentiated THP-1 macrophages were prepared by treating THP-1 cells with 100 nM phorbol myristate acetate (PMA) (Sigma Aldrich, St. Louis, MO, USA) for 2 days.

Cell death of macrophages

Differentiated THP-1 macrophages (2×10⁵ cells in 5% FBS RPMI1640) were infected with viable streptococcal strains at a multiplicity of infection (MOI) of 50, 100, or 200, in the absence of antibiotics, for 2 h. Cells were washed with phosphate buffered saline (PBS, pH 7.2) to remove extracellular non-adherent bacteria, and cultured for 18 h in fresh medium containing antibiotics. Macrophages were then stained with 0.2% trypan blue (Sigma Aldrich) in PBS. After incubation at room temperature for 5 min, the numbers of viable and dead cells were counted using a microscope (Nikon TMS-F, Nikon, Tokyo, Japan).

Cell death induced by H₂O₂ was determined similarly. Differentiated THP-1 macrophages were cultured in the presence of 1, 5, or 10 mM H₂O₂ (Nacalai Tesque, Kyoto, Japan) for 18 h, and viability was determined by trypan blue staining.

Effect of catalase on cell viability

Prior to infection, 10 or 100 U/ml of catalase (Sigma-Aldrich) was added to the cultures of differentiated THP-1 macrophages, and cells were then infected with viable S. oralis strains (MOI; 50, 100, or 200) for 2 h. Cells were washed with PBS, and cultured in fresh medium containing catalase and antibiotics for 18 h. Viability was determined as described above.

Construction of spxB-deficient mutant

The DNA sequence for the pyruvate oxidase gene (SpxB) of S. oralis ATCC35037 (SMSK23_0092) was obtained from Gene Bank (accession number NZ_AEDW01000001). The spxB locus was deleted using a temperature-sensitive suicide vector pSET4s [19], as reported previously [20], [21], [22]. For construction of the spxB deletion mutant, spxKO-F1 and spxKO-R1 primers (Table S1) were utilized for PCR amplification of the upstream flanking sequence of the spxB gene. The downstream flanking sequence of the spxB gene was amplified using primers spxKO-F2 and spxKO-R2 (Table S1). By using the 2 generated PCR products containing complementary ends, overlap PCR was performed with the primers spxKO-F1 and spxKO-R2. The overlap PCR product was digested with EcoRI and BamHI and cloned into the pSET4s vector via EcoRI/BamHI sites. The resultant plasmid pSET4s-spxBKO was transfected into S. oralis ATCC35037 by electroporation. Transformants were grown at 28°C and selected on BHI agar plates containing spectinomycin (100 µg/ml). Single-crossover mutants were obtained by culturing the cells on agar plates with spectinomycin at 37°C, and double-crossover mutants were generated by repeated passaging on agar plates with no antibiotic at 28°C. Finally, spectinomycin-sensitive colonies were tested for deletion of the spxB gene by PCR using primers spx-inside-F/-R and spx-outside-F/-R (Table S1). The S. oralis glucosyltransferase (gtfR) gene was used as a positive control (Table S1) [23]. During the course of the double-crossover, both the spxB-deletion mutant (spxB KO) and the revertant mutant (spxB Rev), which possesses the wild-type allele, were generated from the same ancestor. To rule out the effects of secondary mutations that may have arisen during mutagenesis, a revertant strain was used as a control. Original strain of S. oralis ATCC35037 was used as a wild type (WT) strain.

Hydrogen peroxide measurement

H₂O₂ in S. oralis culture media was quantitatively determined using a hydrogen peroxide colorimetric detection kit (ENZO Life Science, Plymouth Meeting, PA, USA). S. oralis WT, spxB KO, and spxB Rev strains were cultured in BHI broth or RPMI1640 medium supplemented with 5% FBS for 18 h. Culture supernatants were diluted 50-fold in PBS, and H₂O₂ concentrations were then determined according to the manufacturer’s instructions. Our preliminary experiments suggested that, without sufficient dilution, both BHI broth and RPMI1640 medium interfered with the colorimetric reaction of the kit.

Fluorescence microscopy

Differentiated THP-1 cells were cultured on gelatin-coated chamber slides (Asahi Glass, Tokyo, Japan). The macrophages were exposed to S. oralis WT, spxB KO, and spxB Rev strains at an MOI of 200 for 2 h, washed with PBS to remove extracellular bacteria, and cultured for an additional 18 h. The cells were washed with PBS, and then stained by Live/Dead Staining Kit (PromoCell, Heiderberg, Germany). Stained cells were analyzed using an LSM 510 confocal laser microscope (Carl Zeiss, Oberkochen, Germany). Ethidium homodimer III (EthD-III) (red fluorescence) stained the nuclear DNA of dead THP-1 cells, while calcein AM (green fluorescence) stained live cells. THP-1 cells treated with H₂O₂ were stained and observed in a similar manner.

TNF-α assay

Differentiated THP-1 macrophages were infected with viable S. oralis WT, spxB KO, and spxB Rev strains (MOI; 50, 100 or 200) in the absence of antibiotics for 2 h. Cells were washed with PBS to remove extracellular bacteria, and cultured in fresh medium containing antibiotics for an additional 18 h. Cells were also subject to different concentrations of H₂O₂ (1, 5, and 10 mM). The amount of tumor necrosis factor-α (TNF-α) in culture supernatants was measured using ELISA kits (Thermo Scientific, Waltham, MA, USA) according to the manufacturer’s instructions.

Statistical analysis

Statistical analyses were performed using QuickCalcs software (GraphPad Software, La Jolla, CA, USA). Experimental data are expressed as the mean ± SD of triplicate samples. Statistical differences were examined using independent Student’s t-test, with p<0.05 considered to indicate statistical significance.

Results

S. oralis induces cell death of THP-1 macrophages

We previously reported that infection with S. sanguinis induces THP-1 macrophage cell death, with reactive oxygen species apparently contributing to this process. [17]. In the present study, we first examined whether other oral streptococcal species also induce macrophage cell death. Differentiated THP-1 macrophages were exposed to viable oral streptococcal strains, S. mutans MT8148, S. salivarius HHT, and S. oralis ATCC35037. Macrophages were then stained with trypan blue to determine their viability (Figure 1). At an MOI of more than 100, viable S. oralis induced cell death of macrophages at a level comparable to S. sanguinis [17]. Exposure to S. mutans or S. salivarius showed little effects on the viability of the macrophages even at MOIs of 200. During infection at an MOI of over 500, all tested streptococci steadily induced cell death (data not shown). This was likely due to acidification of culture medium and/or accumulation of cytotoxic products such as formic and acetic acids [1], [2], [24].

Figure 1. THP-1 macrophage cell death induced by oral streptococci.

Differentiated THP-1 macrophages were infected with viable S. mutans MT8148, S. salivarius HHT, and S. oralis ATCC35037 for 2 h; washed with PBS to remove non-adherent extracellular bacteria; and cultured in fresh medium containing antibiotics for 18 h. As a control, macrophages were also infected with S. sanguinis SK36 [17]. Macrophage viability was determined by a trypan blue dye exclusion method. Data are shown as the mean ± SD of triplicate samples. *p<0.05 as compared with untreated control (None).

It is well known that S. oralis and S. sanguinis produce H₂O₂, whereas S. mutans and S. salivarius do not [1], [2]. Because reactive oxygen species were previously shown to contribute to cell death of macrophages [17], we investigated the effect of catalase, an H₂O₂-decomposing enzyme, on S. oralis-induced cell death. Exogenously added catalase was shown to reduce cell death in macrophages infected with S. oralis ATCC35037 (Figure 2), suggesting that H₂O₂ is involved in this process.

Figure 2. Effect of catalase on macrophage cell death.

Prior to infection, either 10 or 100 U/ml of catalase was added to cultures of differentiated THP-1 macrophages, and cells were then infected with viable S. oralis ATCC35037 (MOI: 50, 100, or 200) for 2 h. Cells were washed with PBS and cultured in fresh medium containing catalase and antibiotics for 18 h. Viability was determined by a trypan blue dye-exclusion method. Data are shown as the mean ± SD of triplicate samples. *p<0.05 as compared with untreated control (None). **p<0.05 as compared with the cells infected at the same MOI without catalase.

Construction of spxB deficient mutant

Pyruvate oxidase has been reported as being essential for H₂O₂ production in the mitis group of streptococci [5], [6], [25]. Therefore, we constructed a deletion mutant of the pyruvate oxidase gene, spxB, via allelic exchange by using a temperature-sensitive shuttle vector (Figure 3A). Deletion of the spxB gene in the mutant was verified by PCR (data not shown). Decreased production of H₂O₂ by the deletion mutant (spxB KO) was confirmed both in BHI broth and RPMI1640 medium containing 5% FBS at 37°C in a 5% CO₂ atmosphere (Figure 3B). The production of H₂O₂ by the spxB revertant mutant (spxB Rev) was similar to that of a wild type (WT) strain. The mutant strains grew at rates comparable to those of the WT strain (data not shown).

Figure 3. Construction of S. oralis spxB deletion mutant.

(A) Black arrow indicates the gene encoding pyruvate oxidase (SMSK23_0092 spxB). A targeted deletion mutant lacking this region was constructed by allelic exchange using the temperature-sensitive shuttle vector pSET4s. (B) S. oralis ATCC35037 wild-type (WT), spxB-deletion mutant (KO), or reverse mutant (Rev) was cultured in BHI broth or 5% RPMI1640 medium at 37°C for 18 h in a 5% CO₂ atmosphere. Concentrations of H₂O₂ in culture supernatants were quantitatively determined using a hydrogen peroxide colorimetric detection kit. Data are shown as the mean ± SD of triplicate samples. *p<0.05 as compared with concentration of wild-type strain.

Contribution of H₂O₂ produced by S. oralis to macrophage cell death

In order to evaluate the contribution of H₂O₂ produced by S. oralis to macrophage cell death, differentiated THP-1 cells were exposed to S. oralis WT strain, spxB KO mutant, and spxB Rev mutant. Macrophages were then stained with trypan blue to determine their viability (Figure 4, left). At an MOI of 200, macrophages infected with S. oralis WT and spxB Rev strains were found dead, whereas most of the spxB KO-infected cells were still viable. Live/Dead fluorescence staining also revealed reduced cell death of macrophages infected with spxB KO mutant (Figure 4, right).

Figure 4. Microscopic images of macrophage cell death.

THP-1 macrophages were infected with S. oralis wild-type strain (WT), mutant strain defective in H₂O₂ production (spxB KO), or reverse mutant strain (spxB Rev) for 2 h, washed with PBS, and cultured in fresh medium containing antibiotics for 18 h. Macrophages were stained with trypan blue and Live/Dead cell staining kit. EthD-III (red fluorescence) stained the nuclear DNA of dead THP-1 cells, while calcein AM (green fluorescence) stained live cells. Bar, 50 μm.

S. oralis WT and spxB Rev strains induced THP-1 macrophage cell death in a dose-dependent manner (Figure 5). On the other hand, spxB KO mutants had a reduced cytotoxic effect, even at an MOI of 200, indicating that H₂O₂ produced by S. oralis contributes to the induction of macrophage cell death.

Figure 5. Deletion of spxB gene reduces S. oralis cytotoxicity.

THP-1 macrophages were infected with S. oralis wild-type strain (WT), mutant strain defective in H₂O₂ production (spxB KO), or reverse mutant (spxB Rev) for 2 h, washed with PBS, and cultured in fresh medium containing antibiotics for 18 h. Macrophage viability was determined by a trypan blue dye exclusion method. Data are shown as the mean ± SD of triplicate samples. *p<0.05 as compared with untreated control (None). **p<0.05 as compared with the cells infected with WT at the same MOI.

To confirm that H₂O₂ is, in itself, sufficient to induce cell death, THP-1 macrophages were incubated with H₂O₂ alone. As shown in Figure 6, the addition of H₂O₂ to THP-1 cell cultures induced cell death in a dose-dependent manner.

Figure 6. Cell death induced by H2O2.

(A) Differentiated THP-1 macrophages were cultured in the presence of 1, 5, or 10 mM H₂O₂ for 18 h, and their viability was determined by the trypan blue staining. Data are shown as the mean ± SD of triplicate samples. *p<0.05. (B) Macrophages treated with 10 mM H₂O₂ were stained with Live/Dead cell staining kit. EthD-III (red fluorescence) stained the nuclear DNA of dead cells, while calcein AM (green fluorescence) stained live cells. Bar, 50 µm.

Effect of H₂O₂ on TNF-α production in THP-1 macrophages

It is widely recognized that microbial stimulation induces cytokine production in macrophages. Infection with viable S. oralis WT strain induced the production of an inflammatory cytokine, TNF-α (Figure 7). The amount of TNF-α in macrophage culture supernatants increased in a dose-dependent manner. No significant differences in cytokine production between macrophages infected with either WT or spxB Rev strains and those infected with spxB KO mutants were observed. Furthermore, H₂O₂ on its own had a limited stimulatory effect on TNF-α production (Figure 7). These results suggest that H₂O₂ is not essential to TNF-α production in S. oralis-infected macrophages.

Figure 7. Induction of TNF-α by S. oralis spxB KO mutant.

Differentiated THP-1 macrophages were infected with viable S. oralis strains for 2 h, and then washed and cultured for additional 18 h. Other cultures were stimulated by exposure to H₂O₂. The release of TNF-α was determined using an ELISA kit. Data are shown as the mean ± SD of triplicate samples. *p<0.05 as compared with untreated control (None).

Discussion

In our previous study, we showed that S. sanguinis, a member of the oral mitis group of streptococci, induces macrophage cell death [17]. Since S. sanguinis have no established cytotoxins [1], [2], this finding was unexpected. Here, we confirmed that infection with viable S. oralis, another member of oral mitis group, also induced THP-1 macrophage cell death. The most important finding in this study was that streptococci-derived H₂O₂ exhibited cytotoxicity to macrophages.

The oral mitis group of streptococci can give rise to a variety of infectious complications, including bacteremia and infective endocarditis [10], [11], [12], [13]. These bacteria frequently enter the bloodstream following trauma to oral tissues, and then colonize to heart valve surfaces [2], [9], [10], [11]. The mitis group of oral streptococci is the most common cause of native valve endocarditis in humans, accounting for over 30% of cases [9], [10], [11]. The cytotoxicity and tissue-damaging effects of streptococcal H₂O₂ may be factors of bacterial pathogenicity. It is likely that the cytotoxic effect of H₂O₂ enables bacteria to escape from macrophage phagocytosis, and thus contribute to the onset of bacteremia and infectious endocarditis. Furthermore, dead macrophages are reportedly involved in atherosclerosis plaque development [26]. In infective endocarditis, oral streptococci in the bloodstream are entrapped in the platelet-fibrin matrix of cardiovascular tissue vegetations [12], [14]. In such infected lesions, H₂O₂ produced by the streptococci might damage host tissues and allow the bacteria to evade host defense mechanisms. Thus, streptococcal H₂O₂ should be considered as a cytotoxin, and H₂O₂-producing enzymes could be potent targets of the treatments of infections by mitis group of streptococci. Although the H₂O₂ generated by damaged mitochondria is known to induce cell death in various ways [27], our study using an spxB KO mutant strongly suggested that H₂O₂ of bacterial origin plays a major role in macrophage cell death.

Several investigations into Streptococcus pneumoniae, a pathogenic member of the mitis group of streptococci, have reported that bacterial H₂O₂ production is a factor of bacterial pathogenicity. H₂O₂ is suggested as contributing to pneumococcal lung and blood infections in experimental animals [25]. Another study showed that H₂O₂ produced by S. pneumoniae induces microglial and neuronal apoptosis in vitro, and infection with a pneumococcal spxB KO mutant reduces the severity of experimental pneumococcal meningitis [28]. Bioluminescent imaging in infected mice has shown that SpxB contributes to prolonged nasopharyngeal colonization of S. pneumoniae [29]. These studies also indicate that H₂O₂ plays a role as a bacterial cytotoxin. It is therefore conceivable that the H₂O₂ produced by oral streptococci contributes to their virulence. In fact, Stinson et al. [30] reported that the addition of catalase protected endothelial cells from cell death induced by S. gordonii, suggesting that the H₂O₂ produced by the bacteria may contribute to cell death.

The molecular mechanisms underlying streptococcal H₂O₂-mediated cell death are not well understood. H₂O₂ is widely employed as a general-purpose disinfectant. Cell membranes are permeable to H₂O₂, which causes toxicity via oxygen formation, lipid peroxidation, and damage to proteins and nucleic acids [31]. Our previous study using S. sanguinis [17] showed that this cell death is independent of caspase-1 activation. Braun et al. [28] have suggested that pneumococcal H₂O₂ induces apoptosis through release of apoptosis-inducing factor (AIF) from mitochondria in human microglia cells. However, their study showed that the cholesterol-dependent cytolysin, i.e., pneumolysin plays a more important role in induction of microglia cell apoptosis.

Macrophages are known to produce various inflammatory mediators, including cytokines, in response to bacterial components such as lipopolysaccharide and peptidoglycan [32]. Oxidative stress has been implicated in the pathogenesis of a number of inflammatory diseases, including stroke and sepsis [33]. Since streptococcal H₂O₂ contributes to macrophage cell death, it became interesting to clarify whether H₂O₂ stimulates inflammatory responses such as cytokine production. The present study showed that H₂O₂ is not required for TNF-α production in macrophages (Figure 7). Therefore, H₂O₂-mediated cell death seems to be independent of the inflammatory responses of macrophages infected with oral streptococci.

Taken together, our results support the possibility that H₂O₂ plays a significant role in the cell death of macrophages infected with the oral mitis group of streptococci, and suggest a general role for H₂O₂ as a cytotoxin. The contribution of streptococcal H₂O₂ to the pathogenesis of infective endocarditis will be a topic of special interest for future study.

Supporting Information

Table S1

PCR Primers used in this study.

(PDF)

Funding Statement

This work was supported in part by Grants-in-Aid for Scientific Research (B) (#24390410) and (C) (#23593027), and Grants-in Aid for Challenging Exploratory Research (#24659812) from the Japan Society for the Promotion of Science (http://www.jsps.go.jp/j-grantsinaid/). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.