Function of the Pyruvate Oxidase-Lactate Oxidase Cascade in Interspecies Competition between Streptococcus oligofermentans and Streptococcus mutans

Abstract

Complex interspecies interactions occur constantly between oral commensals and the opportunistic pathogen Streptococcus mutans in dental plaque. Previously, we showed that oral commensal Streptococcus oligofermentans possesses multiple enzymes for H₂O₂ production, especially lactate oxidase (Lox), allowing it to out-compete S. mutans. In this study, through extensive biochemical and genetic studies, we identified a pyruvate oxidase (pox) gene in S. oligofermentans. A pox deletion mutant completely lost Pox activity, while ectopically expressed pox restored activity. Pox was determined to produce most of the H₂O₂ in the earlier growth phase and log phase, while Lox mainly contributed to H₂O₂ production in stationary phase. Both pox and lox were expressed throughout the growth phase, while expression of the lox gene increased by about 2.5-fold when cells entered stationary phase. Since lactate accumulation occurred to a large degree in stationary phase, the differential Pox- and Lox-generated H₂O₂ can be attributed to differential gene expression and substrate availability. Interestingly, inactivation of pox causes a dramatic reduction in H₂O₂ production from lactate, suggesting a synergistic action of the two oxidases in converting lactate into H₂O₂. In an in vitro two-species biofilm experiment, the pox mutant of S. oligofermentans failed to inhibit S. mutans even though lox was active. In summary, S. oligofermentans develops a Pox-Lox synergy strategy to maximize its H₂O₂ formation so as to win the interspecies competition.

INTRODUCTION

Humans harbor hundreds to thousands of different indigenous bacterial phylotypes which colonize various surfaces, such as the skin (11), gut (9), vagina (37), and oral cavity (1, 38). These microorganisms usually exist as complex biofilms (6, 12), and species composition in these communities varies in response to environmental changes, such as pH, temperature, nutrition supply (3, 5, 35), and interspecies interaction (15, 23, 46). Though the majority of the inhabitants are beneficial bacteria, some are opportunistic pathogens. When environmental changes favor the growth of opportunistic pathogens, infectious diseases, such as vaginitis (25) and dental caries (13, 31), could develop. Therefore, homeostasis among residential bacteria is critical for maintaining human health (24, 28).

So far, as many as 600 to 800 operational taxonomic units (OTUs) have been found in the human oral cavity (18, 38); some of these OTUs inhabit tooth surfaces and form the multispecies biofilm called dental plaque (20). Dental plaque is a site of dynamic interspecies interaction (34, 36). Oral streptococci are believed to be among the earliest colonizers on tooth surfaces and comprise about 80% of the earliest inhabitants (2, 43). It has been demonstrated that complex interspecies competitions also take place among oral streptococci to contend for the limited space and nutritional resources. Some oral commensal streptococci, such as Streptococcus sanguinis, Streptococcus gordonii, and Streptococcus oligofermentans, are capable of producing hydrogen peroxide (H₂O₂) and bacteriocins which inhibit the overgrowth of the cariogenic pathogen Streptococcus mutans (14, 21, 44, 50, 51); S. mutans, in turn, produces plentiful lactic acid and mutacin to suppress the growth of oral commensal streptococci (31, 40). Once the environmental parameters are favorable to S. mutans growth, e.g., frequent intake of large amounts of carbohydrates by the host, S. mutans could become prevalent in dental plaque, resulting in the eventual development of dental caries (31, 54). Conversely, if beneficial oral commensals, such as S. sanguinis, are predominant over S. mutans, good oral health might be achieved. This antagonistic relationship has been confirmed by epidemiological studies (4). Therefore, a prerequisite for building a healthy oral cavity is to better understand the molecular basis for such relationships.

Frequently isolated from healthy tooth surfaces of the human oral cavity (55), S. oligofermentans was first described in 2003 (52). Phylogenetically it belongs to the mitis group of genus Streptococcus (52). S. oligofermentans has been reported to not only inhibit the growth of S. mutans in a two-species biofilm model (50) but also diminish the capacity of S. mutans to demineralize hydroxyapatite (55). The inhibitory effect of S. oligofermentans on S. mutans is attributable to its abundant H₂O₂ production (50, 51). Previously, we reported that S. oligofermentans possesses two H₂O₂-generating enzymes, including a lactate oxidase (Lox) (50), by which the bacterium catalyzes the formation of pyruvate and H₂O₂ from l-lactate and oxygen and an l-amino acid oxidase that generates H₂O₂ from amino acids and peptone (51). Especially significant, the lactate oxidase enables S. oligofermentans to convert the abundant lactate produced by S. mutans into H₂O₂, which in turn inhibits the growth of S. mutans (50). Despite the fact that lactate tends to accumulate in the latter stages of bacterial growth, significant levels of H₂O₂ have been detected in an early-log-phase culture of S. oligofermentans. Furthermore, H₂O₂ can be detected even in cultures of lox mutants (data not shown). Therefore, other H₂O₂ production pathways may be present. In this study, by using biochemical and genetic approaches, we determined that S. oligofermentans possesses another H₂O₂-forming enzyme, pyruvate oxidase (Pox), and demonstrated that Pox contributed to a large portion of H₂O₂ production during the log phase, while Lox generation of H₂O₂ was mainly in the stationary phase. More importantly, inactivation of Pox also largely abolished lactate-dependent H₂O₂ production, suggesting that a Pox-Lox synergy is required for S. oligofermentans to best engage in interspecies competition. According to the prevalence of pox and lox homologues in the oral streptococci, the ecological impact of this synergy can be predicted.

MATERIALS AND METHODS

Experimental strains and culture conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. All Streptococcus strains were routinely grown in brain heart infusion (BHI) broth (Difco, Detroit, MI) at 37°C as static cultures, whereas plate cultures were incubated in a candle jar. To grow mutant strains, kanamycin (1 mg ml⁻¹) and spectinomycin (800 μg ml⁻¹) were added to BHI medium when necessary. Escherichia coli strains were grown in Luria-Bertani (LB) medium, and when needed, spectinomycin (250 μg ml⁻¹) was supplemented for recombinant selection.

Table 1.

Bacterial strains and plasmids used in this study^a

Strain or plasmid	Relevant characteristics and description	Reference or source
Strains
E. coli DH5α	supE44 Δ(lacZYA-argF)U169 ϕ80dlacZΔ M15 hsdR17 recA1 endA1 gyrA96 thi-1 relA1 luxS	48
S. oligofermentans wild type	AS 1.3089, wild type, Kans Sps	52
S. oligofermentans pox mutant	AS 1.3089 pox::kan, AS 1.3089 with pox deletion	This study
S. oligofermentanspox-com	AS 1.3089 pox::kan pDL278-pox, AS 1.3089 pox mutant with pox complement, Kanr Spr	This study
S. oligofermentans lox mutant	AS 1.3089::pFW5-lox-luc, AS 1.3089 with lox insertion inactivation, Spr	50
S. oligofermentanspox lox double mutant	AS 1.3089 pox::kan pFW5-lox-luc, AS 1.3089 with pox deletion and lox insertion inactivation, Spr Kanr	This study
S. oligofermentans Ppox-luc	AS 1.3089::pFW5-Ppox-luc, luc under AS 1.3089 pox promoter, Spr	This study
S. oligofermentans Plox-luc	AS 1.3089::pFW5-Plox-luc, luc under AS 1.3089 lox promoter, Spr	This study
S. mutans UA140	Wild type	40
Plasmids
pALH124	Kanr	30
pDL278	Spr	27
pDL278-pox	Spr, pDL278 with AS 1.3089 pox gene under its inherent promoter	This study
pFW5-luc	Sprluc	39
pFW5-Ppox-luc	Sprluc, luc under AS 1.3089 pox promoter	This study
pFW5-Plox-luc	Sprluc, luc under AS 1.3089 lox promoter	This study

^aKan, kanamycin; Sp, spectinomycin.

Genetic manipulations.

The genomic DNA of S. oligofermentans was extracted and purified using the method of Marmur with slight modifications (33). Due to the cell wall of S. oligofermentans being resistant to lysozyme breakage and sodium dodecyl sulfate (SDS) treatment, penicillin G (1.25 μg ml⁻¹) was added to the culture at the late exponential phase to make the cell wall development imperfect. The cultures were incubated for another 3 h before being subjected to the DNA extraction procedure. The PCR primers were designed according to the unreleased draft genome of S. oligofermentans 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 by using a Qiagen QIAquick PCR purification kit (Valencia, CA). DNA extracted from agarose gels was purified with a Tiangen TIANgel purification midikit (Beijing, China), and plasmids were extracted and purified with a Tiangen TIANprep plasmid minikit. Restriction enzymes and DNA ligase were purchased from New England BioLabs (Beverly, MA). All procedures were carried out as recommended by the suppliers.

A pyruvate oxidase gene (pox) deletion mutant was constructed by using the PCR ligation method (26). Briefly, two fragments corresponding to about 600 bp of the upstream and downstream sequence, respectively, of the pox gene were amplified by PCR with the following two pairs of primers (restriction sites are underlined, and modified sequences are in italics): poxupF (5′-GGCATTCACTAGCATAACCC-3′)/poxupR (with a BamHI restriction site) (5′-AAGGATCCTTGAGTCATAATGATAAC-3′) and poxdnF (with a BamHI restriction site) (5′-AAGGATCCGGAAGAAGAAGGTTTGC-3′)/poxdnR (5′-CAAGCAATGGACAGACGATAG-3′) using S. oligofermentans genomic DNA as the template. The purified PCR products were digested with BamHI. The nonpolar kanamycin resistance gene cassette was released from plasmid pALH124 (30) 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 the method described previously (53). Transformants were selected on BHI agar plates containing 1 mg ml⁻¹ kanamycin; pox deletion was confirmed by PCR and sequencing.

A pox gene complementary strain was constructed as described below. The pox coding region with its inherent promoter was amplified using a pair of 5′-end-modified primers (restriction sites are underlined, and modified sequences are in italics): Pox-com-up (with an EcoRI restriction site) (5′-AGGAATTCCTTTTCACCTACTATTTTG-3′) and Pox-com-dn (with a SalI restriction site) (5′-ATATGTCGACTTATTTAATTGCGCGTGATTG-3′). The 2,112-bp PCR product was purified and double digested with EcoRI and SalI. After digestion, the purified fragment was inserted into compatible sites on an E. coli-Streptococcus shuttle vector pDL278 (27), which was treated with the same enzymes. Recombinant plasmid pDL278-pox was transformed into the pox mutant; transformants were selected on BHI agar containing kanamycin (1 mg ml⁻¹) and spectinomycin (800 μg ml⁻¹). The positive transformant was further confirmed by PCR and sequencing.

Ppox-luc and Plox-luc reporter strains were constructed by the following procedures. Two pairs of 5′-modified primers were designed and synthesized, PpoxF (with a BamHI restriction site) (5′-AAGGATCCTGTTTTCAATACGTTGAGC-3′)/PpoxR (with an NheI restriction site) (5′-AAGCTAGCGCTGCCAGACCTTTTGTAAG-3′) and PloxF (with a BamHI restriction site) (5′-AAGGATCCCATTGCTTGTTTTGTATG-3′)/PloxR (with an NheI restriction site) (5′-AAGCTAGCAAGACCCACCTAGGAAAATG-3′). The pox and lox promoter regions, each including approximately 600 bp of sequence upstream of the gene start codon, were amplified from chromosome DNA of S. oligofermentans by PCR with the primer pairs PpoxF/PpoxR and PloxF/PloxR, respectively. Purified pox and lox promoter fragments were subsequently digested with NheI and BamHI, gel purified, and ligated to compatible sites on the pFW5-luc (39) vector using DNA ligase. Correct recombinants (pFW5-Ppox-luc and pFW5-Plox-luc) were confirmed by restriction analysis, PCR, and sequencing. Plasmids containing Ppox-luc and Plox-luc fusions were then transformed into the S. oligofermentans wild-type strain. Transformants were selected on BHI agar containing 800 μg ml⁻¹ spectinomycin and confirmed by PCR and luciferase activity.

Hydrogen peroxide determination.

Hydrogen peroxide (H₂O₂) in liquid culture was quantified using a modified method described previously (10, 45). 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 optical density at 510 nm (OD₅₁₀) was measured with a Unico 2100 visible-light spectrophotometer (Shanghai, China). A standard curve was generated with known concentrations of chemical H₂O₂.

Luciferase activity assay.

Twenty-five microliters of 1 mM d-luciferin (Sigma) solution (suspended in 1 mM citrate buffer, pH 6.0) was added to 100-μl samples. Luciferase assays were performed essentially as previously described (53) using a TD 20/20 luminometer (Turner system). The OD₆₀₀ of the sample was read with a 2100 visible light spectrophotometer (Unico). All the measurements were done with duplicate samples, and all experiments were repeated at least three times.

Lactic acid determination.

The lactic acid assay was performed using an EnzyChrome lactate assay kit (ECLC-100; BioAssay Systems) according to the manufacturer's instructions. Culture samples or reaction mixtures were centrifuged at 12,000 × g for 2 min, and the supernatant was diluted to a suitable concentration to perform the assay.

Hydrogen peroxide-generating oxidase activity assay.

Cells were collected and washed twice with 1 ml of phosphate-buffered saline (PBS) by centrifugation at 10,000 × g for 10 min. The pelleted cells were resuspended in 1 ml of the same buffer and then permeabilized by mixing the suspensions with 0.02 volume of toluene-acetone (1:9, vol/vol) with a vortex mixer for 2 min. Pyruvate oxidase activity was determined by assaying the acetyl phosphate production as described previously with a slight modification (29, 49). The reaction mixture consisted of 0.5 ml of the permeabilized cell suspension and 0.5 ml of a solution containing 50 mM potassium phosphate buffer (pH 6.0), 10 μM MgCl₂, 0.2 μM thiamine pyrophosphate (Sigma), 50 mM potassium pyruvate, and 12 μM flavin adenine dinucleotide (FAD; Sigma). The mixtures without the permeabilized cells or potassium pyruvate were used as negative controls. After incubation at 37°C for 20 min with shaking, 1 ml each of 4 M hydroxylamine (pH 6.4) and 0.1 M acetate buffer (pH 5.4) was added, and the mixtures were kept standing for 10 min at room temperature. Then, 1 ml each of 36% HCl, 12% (wt/vol) trichloroacetic acid, and 5% (wt/vol) ferric chloride in 0.1 N HCl was added to the mixtures. After being kept at room temperature for 20 min, the mixtures were centrifuged at 12,000 × g for 5 min. The absorbance of the supernatant was measured at 540 nm. A standard curve was generated with known concentrations of acetyl phosphate.

Lactate oxidase and NADH oxidase activities were assessed by assaying the H₂O₂ production as described previously with a slight modification (45). Two hundred microliters of the permeabilized cells was added to 2 ml of the 0.1 M sodium phosphate buffer (pH 7.0) containing 20 mM sodium l-lactate (Shang Hai SSS Reagent Co., Ltd., China) and 13 mM NADH (Sigma), respectively, and the reaction mixtures without sodium l-lactate or NADH were used as negative controls. The mixtures were incubated at 37°C with shaking for 20 min and then centrifuged at 12,000 × g for 5 min. The supernatant was used for H₂O₂ determination using the method mentioned above.

Interspecies antagonism assay between S. oligofermentans and S. mutans.

The interspecies antagonism experiment was performed as described previously with a slight modification (50). Briefly, overnight BHI cultures of S. mutans and various S. oligofermentans strains were adjusted to the same optical density at 600 nm (∼1.0). The cultures were diluted at a 1:10 ratio into fresh BHI broth plus 0.5% sucrose in a 12-well cell culture plate (Corning) as a mixed-species or single-species culture. The plate was incubated at 37°C as a standing culture for 8 h. After incubation, cells were dispersed by vigorous pipetting. Serial dilutions of cells were then plated on BHI agar, and CFU of S. oligofermentans and S. mutans were counted based on their different colony morphologies; the former is white and slimy and the latter is yellow and dry.

Nucleotide sequence accession number.

The DNA sequence of the pox gene has been deposited in the GenBank database under accession number JQ004920.

RESULTS

Determination of pyruvate oxidase activity and identification of a pox gene in S. oligofermentans.

To find extra H₂O₂ production pathways in S. oligofermentans, the well-known bacterial enzymatic activities for H₂O₂ production were tested; those included pyruvate oxidase and NADH oxidase activities. Using the permeabilized cells as the crude enzyme and pyruvate as substrate, the Pox activity of 1.84 ± 0.13 (mean ± standard deviation) μmol/min/mg protein was measured. NADH oxidase activity was barely detected (data not shown); this is probably because the detection limit (1.90 μM) of the H₂O₂ assay makes the assay incapable of detecting the lower H₂O₂ level generated by NADH oxidase.

Next, by using the pyruvate oxidase gene (SGO_0292) from S. gordonii as a probe, a putative pyruvate oxidase gene (pox; open reading frame [orf] 2213) was hit in the draft genome of S. oligofermentans. orf 2213 showed high degrees of identity (96, 96, and 95%, respectively) with pyruvate oxidase genes from Streptococcus oralis, S. gordonii, and Streptococcus mitis.

To further verify the function of the putative pox gene, a pox deletion mutant was constructed by allelic exchange with a nonpolar kanamycin resistance cassette (kan). Meanwhile, a strain ectopically expressing pox, designated pox-com, was constructed by transforming the shuttle vector pDL278 carrying the pox gene with its inherent promoter into the S. oligofermentans pox mutant. Both pox mutant and pox-com strains were confirmed by PCR and sequencing; they were then subjected to pyruvate oxidase (Pox) assay by measuring acetyl phosphate production as described in Materials and Methods. The results showed that the pox mutant lost almost all the acetyl phosphate production (0.09 ± 0.02 μmol/min/mg protein), while an activity of 2.41 ± 0.03 μmol/min/mg protein was determined for the complementary strain pox-com.

Pox and Lox contribute the majority of hydrogen peroxide production in S. oligofermentans but act in different growth phases.

To determine the relative contributions of Pox and Lox activities to H₂O₂ production during different growth phases, S. oligofermentans wild type and pox and lox mutants were grown in BHI broth aerobically, with their optical density and H₂O₂ production monitored every hour. As shown in Fig. 1A, the growth results showed that the lox mutant displayed a growth profile similar to that of the wild type, while the pox mutant grew significantly faster than the wild type and accumulated more cell mass in the end. Also, as shown in Fig. 1B, the wild-type strain produced H₂O₂ throughout the growth phase, with the highest concentrations of H₂O₂ detected at early log phase and stationary phase. Compared to the H₂O₂ production of the wild type, the pox mutant produced markedly smaller amounts of H₂O₂ overall, which increased slightly after cells entered stationary phase. On the other hand, H₂O₂ production by the lox mutant was only slightly less than that of the wild type during the exponential phase, and it then quickly diminished to the level produced by the pox mutant at the start of the stationary phase. These results indicated that Pox contributes the majority of H₂O₂ production by S. oligofermentans when growing exponentially in BHI, while Lox seems to function primarily in the stationary phase.

Fig 1.

Hydrogen peroxide production by various S. oligofermentans strains growing in BHI cultures. Overnight cultures of S. oligofermentans wild type and various mutant strains were diluted at a 1:40 ratio into fresh BHI medium. Subsequently, cultures were sampled every hour and then measured for optical density at 600 nm and H₂O₂ production upon shaking at 200 rpm/min for 20 min. (A) Growth curve of various S. oligofermentans strains expressed as OD₆₀₀. (B) Hydrogen peroxide production throughout the growth phase. Data are expressed as H₂O₂ concentration (mM) per optical density at 600 nm. ♦, wild type; ▵, lox mutant; □, pox mutant; ●, pox lox double mutant. Data are representative of three independent experiments.

To determine whether other glucose-originated H₂O₂ production pathways besides Pox and Lox were present in S. oligofermentans, a pox lox double mutant was constructed by transformation of the lox mutant (50) with a PCR fragment containing a pox deletion. When tested in the same growth/H₂O₂ production assays as described above, the pox lox double mutant also grew faster than the wild type and reached a higher final cell mass (Fig. 1A), while its H₂O₂ level was extremely low (average, 3 μM) throughout the growth phase (Fig. 1B). Together with the phenotype of the pox and lox mutants, these results indicate that the gene products of pox and lox are the primary producers of glucose-derived H₂O₂ in S. oligofermentans. Therefore, the H₂O₂ from the pox mutant can be taken as the product of Lox, while the lox mutant can be taken as a Pox active strain.

Gene expression level and substrate availability determine Pox and Lox activities for hydrogen peroxide production.

Based on the differential H₂O₂ production mediated by Pox and Lox in different growth periods, the levels of growth phase-related expression were assayed for the two genes. Each of the Ppox-luc and Plox-luc fusions was constructed by the fusion of a DNA fragment containing the pox or lox promoter to a promoterless luciferase gene (luc) on the pFW5-luc plasmid (39); the plasmids pFW5-Ppox-luc and pFW5-Plox-luc were then transformed into S. oligofermentans wild type. Overnight BHI cultures of the reporter strains, S. oligofermentans Ppox-luc and S. oligofermentans Plox-luc, were diluted 1:40 into fresh BHI broth, and the cultures were sampled at 1-h intervals to measure the optical density at 600 nm and luciferase activity as described in Materials and Methods. As shown in Fig. 2A, luciferase activity of the Ppox-luc fusion could be detected from the early growth phase and was maintained at a somewhat-constant level throughout the growth phase, indicating constitutive expression of the pox gene. While the expression of the lox gene was minimal at early log phase, it increased by about 2.5-fold when cells entered stationary phase. To further examine the expression of these two genes at the protein level, the enzyme activities of Pox and Lox were measured throughout the growth period in S. oligofermentans wild type. The results showed that their enzymatic activities increased steadily during the entire growth period (Fig. 2B).

Fig 2.

Expression profiles and enzymatic activities of pyruvate oxidase and lactate oxidase in S. oligofermentans. Cells were collected at different growth phases and then assayed for luciferase, pyruvate oxidase, and lactate oxidase activities as described in Materials and Methods. (A) Luciferase activities of Ppox-luc and Plox-luc fusions throughout the growth phase. Data are expressed as the relative light units (RLU) per optical density at 600 nm. □, Ppox-luc; ♦, Plox-luc. (B) Pyruvate oxidase (□) and lactate oxidase (♦) activities. Results are expressed as the means ± standard deviations of three independent experiments.

Since Pox- and Lox-mediated H₂O₂ production occurred differentially during these growth phases, substrate availability for each enzyme at different growth periods could be predicted. Pyruvate, the substrate for Pox, is the intermediate of glucose fermentation via the Embden-Meyerhof-Parnas (EMP) pathway, while lactate, the substrate of Lox, is the end product of the EMP pathway and accumulates in late log phase. As shown in Fig. 3, lactate was only detected at about 2 to 3 mM in early log phase, while the yield increased to about 6 mM in the mid-log-phase culture and increased quickly to about 18 mM in the stationary-phase culture. Therefore, both gene expression level and substrate availability coordinately influence H₂O₂ generation by Pox or Lox in S. oligofermentans.

Fig 3.

Time course of lactate production by S. oligofermentans. Overnight cultures of S. oligofermentans were diluted at a 1:40 ratio into fresh BHI medium and incubated as a static culture; samples were taken every hour to measure optical density at 600 nm and lactate concentration. ●, growth expressed as OD₆₀₀; □, lactate concentration (mM). Results are expressed as means ± standard deviations of three independent experiments.

Inactivation of Pox dramatically reduces lactate oxidase activity.

Although low H₂O₂ production by Lox in the earlier growth phase might be due to the low lactate level, reduced H₂O₂ production (about 0.50 mM) in the stationary culture of the pox mutant suggests a possible effect of Pox on the regulation or function of Lox. To investigate this possibility, lactate levels were measured for the stationary cultures of the wild-type strain and the pox and lox mutants. As shown in Fig. 4A, lactate yields in the static cultures were similar in all three strains, with 18.3, 18.1, and 18.4 mM in the wild type and the pox and lox mutants, respectively. Surprisingly, upon shaking the cultures for 20 min to maximize lactate oxidase activity, the lactate levels decreased 14% in the wild type (P < 0.05), whereas little change was seen in either the pox or lox mutant. In shaken cultures of the wild-type strain and the pox and lox mutants, the increased H₂O₂ levels were measured at 1.73 mM, 0.31 mM, and 0.33 mM (Fig. 4B), respectively. This indicates that Pox activity could exert an effect on Lox activity.

Fig 4.

Lactate oxidase activities of S. oligofermentans wild type and pox and lox mutants. Lactate concentration (A) and H₂O₂ production (B) were measured in the stationary-phase cultures before (black bar) and after (gray bar) shaking at 200 rpm/min for 20 min. Data are expressed as the means ± standard deviations of three independent experiments. ∗∗∗, data are statistically significant in comparison to values before shaking, as verified by Student's t test (P < 0.05).

To clarify the effect of Pox on Lox, Lox activity in the wild type was compared with that in the pox mutant as described in Materials and Methods. Lactate oxidase activity, i.e., H₂O₂ formation from lactate, in the pox mutant was found to be only 8% of that in the wild-type strain. According to Taniai et al. (49), lactate is oxidized to acetyl phosphate and H₂O₂ by a coupled reaction of Pox and Lox; i.e., lactate is first oxidized by Lox into H₂O₂ and pyruvate, which is then converted into acetyl phosphate and H₂O₂ by Pox. Therefore, inactivation of pox would in theory disrupt the complete conversion of lactate into H₂O₂. Since the Pox activity in the lox mutant (1.95 ± 0.43 μmol/min/mg) was about the same as that in the wild-type strain (1.84 ± 0.14 μmol/min/mg), Lox activity does not seem to be required for the function of Pox.

Pox contributes more in the interspecies competition between S. oligofermentans and S. mutans in biofilm.

As Pox was determined to contribute the most to H₂O₂ production in S. oligofermentans and to be required for Lox activity, the relative roles of Pox and Lox in the interspecies competition were determined in the two-species biofilm model. To do this experiment, overnight BHI cultures of S. oligofermentans wild type, pox mutant, lox mutant, or pox lox double mutant were inoculated with S. mutans pairwise as a mixed-species biofilm, with monocultures of each strain included as controls. After 8 h, the biofilm cells were scraped and spun down. The cell numbers of each strain in mixed or monocultures were counted on BHI agar plates. S. oligofermentans and S. mutans cells were enumerated based on their different colony morphologies; however, when necessary, S. oligofermentans mutants were counted on BHI agar plates supplemented with antibiotics. As shown in Fig. 5A, compared to the number of S. mutans cells in monoculture, S. mutans cells decreased the most when cocultured with S. oligofermentans wild type, followed by the lox mutant, the pox mutant, and lastly, the pox lox double mutant. Meanwhile, the abundance of S. oligofermentans wild-type cells was comparable in both monoculture and coculture with S. mutans, while the numbers of the pox mutant, lox mutant, and pox lox double mutant in the coculture with S. mutans were reduced to about 3, 21, and 6%, respectively, of their levels in monoculture (Fig. 5B). These results suggest that inactivation of Lox and, especially, loss of Pox resulted in diminished inhibition of S. mutans or, conversely, reduced resistance against S. mutans.

Fig 5.

pox and lox mutant strains of S. oligofermentans lost antagonism against S. mutans in the two-species biofilms formed in BHI-sucrose. Overnight cultures of various S. oligofermentans strains and S. mutans UA140 were adjusted to the same optical density at 600 nm (∼1.0) and then were inoculated (1:10) into a 12-well plate containing fresh BHI broth supplemented with 0.5% sucrose to form mono- or mixed-species biofilms. After 8 h of incubation, the biofilms were harvested and cell numbers enumerated by plating. (A) CFU of S. mutans in mixed cultures with various S. oligofermentans strains (bars 1 to 4) and in a monoculture (bar 5). Bars: 1, with wild type; 2, with pox mutant; 3, with lox mutant; 4, with pox lox double mutant. (B) CFU of various S. oligofermentans strains in mixed cultures with S. mutans (black bar) and in monocultures (gray bar). Data are expressed as the means ± standard deviations of three independent experiments.

DISCUSSION

The production of H₂O₂ is thought to be the principal strategy used by oral commensals to win the interspecies competition over S. mutans (21, 44, 50, 51). Enzymes that oral bacteria employ to produce H₂O₂ include NADH oxidase (17, 32), pyruvate oxidase (Pox) (22), lactate oxidase (Lox) (41, 50), and l-amino acid oxidase (LAAO) (51). Oral commensal S. oligofermentans can produce copious amounts of H₂O₂ from lactate through Lox (50) or from l-amino acids via LAAO (51). As demonstrated by its potent ability to inhibit S. mutans growth via release of H₂O₂, S. oligofermentans has great potential to be developed into a probiotic for prevention of dental caries. In this study, we showed that S. oligofermentans develops a smart strategy to maximize its advantage in interspecies competition over S. mutans, using a synergy of two H₂O₂-producing enzymes, pyruvate oxidase and lactate oxidase.

Pyruvate oxidase (Pox) catalyzes the oxidation of pyruvate to produce acetyl phosphate and H₂O₂. Pox has been identified in several lactic acid bacteria (16, 42, 47), including S. sanguinis and S. gordonii (22). Both S. sanguinis and S. gordonii are believed to be among the first colonizers in dental biofilm. They influence biofilm development through coaggregation and competition with other bacterial species (7), with Pox playing a critical role in their interaction with S. mutans (22). Hence, Pox can be predicted to be part of the primary defense mechanism against dental pathogen S. mutans in the initial stage of biofilm formation. In this study, a pox homologue was identified in S. oligofermentans and verified to encode a functional pyruvate oxidase. The relatively constitutive expression of pox throughout the growth (Fig. 2) and the sufficient substrate (pyruvate produced through glycolysis) enable the Pox-generated H₂O₂ to account for the major portion of the H₂O₂ pool (Fig. 1), especially in the earlier growth phase when little lactate is present. Therefore, Pox could be especially important in helping S. oligofermentans to outcompete H₂O₂-susceptible dental pathogens and inhabit the multispecies community as an earlier colonizer.

More importantly, inactivation of Pox causes a significant reduction in the amount of H₂O₂ produced by Lox. Lactate, the bacterial metabolite present in dental plaque, is the main cause of tooth demineralization and caries (31) but can be converted by the Lox-possessing S. oligofermentans into H₂O₂, which inversely inhibits S. mutans growth (50). Therefore, the activity of Lox can exert a special ecological impact in the oral environment. In this study, we found that through the biochemical cascade with Pox, Lox can achieve its maximal potential in converting the cariogenic lactic acid into H₂O₂ (Fig. 4). Affected by pox inactivation on Lox activity, the quantity of H₂O₂ produced by the pox mutant is very low in the stationary phase (Fig. 1B), even if a large amount of lactate accumulates (Fig. 3) and lox is active (Fig. 2); hence, the sum of H₂O₂ produced in pox and lox mutants is much lower than the wild-type level (Fig. 1B). Therefore, active Pox is critical for S. oligofermentans to win the interspecies competition against S. mutans (Fig. 5). Moreover, H₂O₂ production via Pox is coupled to the generation of acetyl phosphate, and the conversion of acetyl phosphate to acetate generates ATP. The Pox-Lox cascade has been shown to play a role in energy recruitment from lactate oxidation in Streptococcus pneumoniae (49). Hence, in addition to maximal H₂O₂ production from lactate, the Pox-Lox interplay might supply extra energy for oral commensals in dental biofilm.

Bioinformatics analysis showed that pox and lox homologues are also present in the genomes of four other streptococci—S. mitis, Streptococcus cristatus, S. oralis, and Streptococcus infantis. They all belong to the mitis group and are prominent members of the oral commensal streptococci (8). However, no pox and lox homologues can be found in the genome of S. mutans. Therefore, the synergistic action between the two oxidases can be crucial for oral streptococci in interspecies competition in plaque biofilm, thus affecting dental health.

Reportedly, the pyruvate oxidase gene is subjected to repression by catabolite control protein A (CcpA) in S. gordonii (57) and S. sanguinis (56). The isogenic ccpA mutant in the two species shows elevated pyruvate oxidase gene expression and H₂O₂ production. Similarly, lox gene expression is under the control of CcpA in Streptococcus pyogenes (19), and no Lox activity is detected in the glucose-containing culture of S. pyogenes (45), whereas in S. oligofermentans, lox expression and, more importantly, Lox activity and H₂O₂ production are readily detected in the glucose-containing BHI culture (Fig. 2). This may help S. oligofermentans win the competition over S. mutans by using Pox and Lox as soon as it colonizes in the dental biofilm. Therefore, further studies are needed to explore the detailed characterization of these two genes, including their interaction and regulation in the context of different carbon sources, as well as oxygen tension.