Glycerol metabolism contributes to competition by oral streptococci through production of hydrogen peroxide

Abstract

As a biological byproduct from both humans and microbes, glycerol's contribution to microbial homeostasis in the oral cavity remains understudied. In this study, we examined glycerol metabolism by Streptococcus sanguinis, a commensal associated with oral health. Genetic mutants of glucose-PTS enzyme II (manL), glycerol metabolism (glp and dha pathways), and transcriptional regulators were characterized with regard to glycerol catabolism, growth, production of hydrogen peroxide (H₂O₂), transcription, and competition with Streptococcus mutans. Biochemical assays identified the glp pathway as a novel source for H₂O₂ production by S. sanguinis that is independent of pyruvate oxidase (SpxB). Genetic analysis indicated that the glp pathway requires glycerol and a transcriptional regulator, GlpR, for expression and is negatively regulated by PTS, but not the catabolite control protein, CcpA. Conversely, deletion of either manL or ccpA increased the expression of spxB and a second, H₂O₂-non-producing glycerol metabolic pathway (dha), indicative of a mode of regulation consistent with conventional carbon catabolite repression (CCR). In a plate-based antagonism assay and competition assays performed with planktonic and biofilm-grown cells, glycerol greatly benefited the competitive fitness of S. sanguinis against S. mutans. The glp pathway appears to be conserved in several commensal streptococci and actively expressed in caries-free plaque samples. Our study suggests that glycerol metabolism plays a more significant role in the ecology of the oral cavity than previously understood. Commensal streptococci, though not able to use glycerol as a sole carbohydrate source for growth, benefit from the catabolism of glycerol through production of both ATP and H₂O₂.

Importance: Glycerol is an abundant carbohydrate in the oral cavity. However, little is understood regarding the metabolism of glycerol by commensal streptococci, some of the most abundant oral bacteria. This was in part because most streptococci cannot grow on glycerol as the sole carbon source. In this study, we show that Streptococcus sanguinis, a commensal associated with dental health, can degrade glycerol for persistence and competition through two pathways, one of which generates hydrogen peroxide at levels capable of inhibiting Streptococcus mutans. Preliminary studies suggest that several additional commensal streptococci are also able to catabolize glycerol, and glycerol-related genes are actively expressed in human dental plaque samples. Our findings reveal the potential of glycerol to significantly impact microbial homeostasis, which warrants further exploration.

Keywords: Streptococcus sanguinis; competition; dental caries; glycerol metabolism; hydrogen peroxide; phosphotransferase system (pts).

Fig 1

Diagram depicting genes required for glycerol metabolism in several lactic acid bacteria. Genes annotated for the glycerol oxidation pathway (glp) are presented in blue, whereas those for the glycerol dehydrogenation pathway (dha) are in green. Putative transcription regulators associated with these genes are depicted in orange. Also denoted are catabolite response elements (cre) located in the intergenic regions, a formate acetyltransferase pfl2, and a transaldolase-like protein mipB.

Fig 2

Growth curves measured in FMC medium modified to contain (A) glycerol; (B) glucose; (C) glucose and glycerol; and (D) glucose, glycerol, and 5 µg/mL catalase. Strains were cultured to the mid-exponential phase (OD₆₀₀ = 0.5) in brain–heart infusion (BHI), before being diluted 100-fold into 200 µL of defined medium (FMC) and loaded into a 96-well plate and covered with 60 µL mineral oil. OD₆₀₀ was recorded using the Biotek Synergy 2 once every hour for 24 hours. The incubation was carried out at aerobic conditions at 37°C. Each sample was represented by at least four biological replicates, and error bars denote standard deviations.

Fig 3

Glycerol induced in the manL mutant increased release of eDNA and H₂O₂. The bacteria were cultivated in FMC constituted with 20 mM glucose and 0 or 5 mM of glycerol for 24 hours in an aerobic atmosphere with 5% CO₂. The supernatant of each culture was harvested by centrifugation, and (A) the relative levels of extracellular DNA were measured by reacting with a fluorescent DNA dye, and (B) H₂O₂ concentrations using a biochemical reaction and a standard curve. At least three biological replicates were included in each sample, and the results were normalized against the cell density represented by OD₆₀₀ of each bacterial culture. The bars represent the mean of the measurements, and error bars denote the standard deviations. Asterisks indicate statistical significance assessed by two-way ANOVA followed by Tukey’s multiple comparisons test (*, P < 0.05; **, P < 0.01; ***, P < 0.001; and ****, P < 0.0001).

Fig 4

Prussian blue (PB) plate assay for measurement of H₂O₂ release. Overnight cultures of bacteria in brain–heart infusion (BHI) were washed and dropped onto PB plates containing 20 mM glucose, 40 mM glycerol, or 20 mM glucose plus 5 mM glycerol. After another day of incubation in an ambient incubator maintained with 5% CO₂, (A) the plates were photographed and (B) the width of the blue zone was measured to represent the relative amounts of H₂O₂ being released. At least three biological replicates were used for each sample, and the average size of the PB zone was used to plot the bar graph, with error bars representing standard deviations and asterisks denoting statistical significance relative to results of the parent strain SK36 assayed on the same carbohydrate (unless specified otherwise) (Student’s t-test; **, P < 0.01; ***, P < 0.001; and ****, P < 0.0001).

Fig 5

Growth curves constructed using S. sanguinis strains SK36 and its mutant derivatives (A, B, C), S. gordonii strain DL1 (D), S. oralis strain BCC11 (E), and S. mitis strain BCC36 (F). The media were based on FMC that was modified to contain 20 mM galactose, with or without 5 mM glycerol, and 5 µg/mL catalase (C). Each strain was represented by at least three biological replicates, with their means and standard deviations (error bars) in OD₆₀₀ values being presented.

Fig 6

Deletion of manL enhanced the antagonism of S. mutans in the presence of glycerol. To avoid contaminating carbohydrates, FMC was used as the base medium and was modified to contain 20 mM glucose, 20 mM glucose and 5 mM glycerol, 40 mM glycerol, 20 mM galactose, or 20 mM galactose and 5 mM glycerol. Each S. sanguinis strain was inoculated on the agar plates first. After overnight incubation, a fresh BHI culture of S. mutans UA159 (SMU) was inoculated to the immediate right of the colony, followed by another day of incubation before photography. Each interaction was tested at least twice on separate days, with a representative set of results being presented here.

Fig 7

Glycerol greatly benefits S. sanguinis (SSA) against S. mutans (SMU) in a planktonic competition assay. Exponentially growing BHI cultures of S. sanguinis SK36, ΔglpK, ΔgldA, and ΔdhaKL were each mixed with equal amounts of S. mutans strain UA159, inoculated at a 100-fold dilution rate, and allowed to grow overnight in TY (A) or FMC (B) medium supplemented with 20 mM glucose, 40 mM glycerol, or a combination of 20 mM glucose and 5 mM glycerol. The CFU of each competing species before and after the competition was enumerated to calculate the competition indices, with values greater than 1 indicative of an advantage for S. sanguinis. Each strain was represented by four separate cultures, from which the means and standard deviations (error bars) were calculated and presented. Statistical analysis was carried out using two-way ANOVA, followed by Tukey’s multiple comparisons (*, P < 0.05; **, P < 0.01; ***, P < 0.001; and ****, P < 0.0001).

Fig 8

Glycerol promotes S. sanguinis (SSA) competition against S. mutans (SMU) in a two-species biofilm in a glpK-dependent manner. SK36 and its isogenic mutants ΔglpK, ΔdhaKL, and ΔgldA were each mixed with an equal volume of S. mutans UA159 culture and inoculated, at a 1:100 ratio, into a BMGS medium held in a 24-well plate with a hydroxyapatite disk and kept in an aerobic atmosphere with 5% CO₂. After 24 hours of incubation, the culture supernatant was replaced with BM base, BM containing 18 mM glucose (Glc), or BM with 36 mM glycerol (Gly). After another day of incubation, the biofilm was washed and harvested by sonication, followed by serial dilution and CFU enumeration. Each strain was represented by four separate cultures and each condition four biofilm samples. The average CFU and standard deviation (error bars) of each species were used to plot the graph and for statistics (Students’ t-test; *, P < 0.05).

Fig 9

Glycerol metabolic pathway of S. sanguinis shows elevated expression in caries-free dental plaque samples. Metatranscriptomic analysis was performed on 70 plaque samples, 34 caries-free (PF) and 36 caries-active [11 from enamel sites (PE) and 25 from dentin sites (PD)]. Scatter plots show fold changes in transcript abundance in (A) caries-free vs enamel lesions and (B) caries-free vs dentin lesions. The red lines denote the value of 1, and the blue lines show the median of the population. The red circle represents the “super-pathway of glycerol degradation to 1,3-propanediol,” and the orange diamond represents the “CDP-diacylglycerol biosynthesis pathway.”