Synergistic pairing of Lactobacillus helveticus Stellar Lh-170 and Lactiplantibacillus plantarum Stellar Lp-199 suppresses Streptococcus mutans and disrupts cariogenic biofilms

ABSTRACT

Background

Streptococcus mutans exists within a biofilm ecosystem where neighboring microbes influence each other’s viability. Probiotic strains with complementary antagonistic properties may enhance suppression beyond individual strains. An evaluation of probiotic combinations assessed synergistic biofilm and pathogen suppression.

Methods

Nineteen probiotic strains were screened for hydrogen peroxide production and antagonism of Streptococcus mutans; five candidates were further tested in two-strain combinations for synergistic antagonism of S. mutans and hydroxyapatite adhesion. An artificial saliva culture model assessed both the inhibition and disruption of S. mutans biofilms alone and with Candida albicans, with findings validated on human molars.

Results

Of the 19 strains assessed, Lactobacillus helveticus Stellar^TM Lh-170 and Lactiplantibacillus plantarum Stellar^TM Lp-199 exhibited the strongest synergistic antagonism against all tested S. mutans strains (p < 0.001). Both strains rapidly produced H₂O₂ and adhered to hydroxyapatite. When co-incubated at a 10:1 ratio with S. mutans, the combination reduced biofilm biomass by 62% (n.s.) and significantly reduced S. mutans viability below detectable CFU limits (~10⁷-fold reduction). When co-incubated with S. mutans and C. albicans together, dose-dependent reductions in the viability of both pathogens were observed. When pre-established S. mutans biofilms were treated with the two-strain combination, the number of viable S. mutans CFUs was reduced by >  10,000-fold in vitro and 3,135-fold in extracted human molars after 24 h (p < 0.0001).

Conclusions

Synergistic strain combinations effectively modulate cariogenic pathogens in vitro, exceeding benchmark strains. These findings support the clinical evaluation of the Lactobacillus helveticus Stellar Lh-170 and Lactiplantibacillus plantarum Stellar Lp-199 combination for prophylactic caries risk management.

KEY MESSAGES

Lactobacillus helveticus Stellar Lh-170 and Lactiplantibacillus plantarum Stellar Lp-199 synergistically inhibit S. mutans growth and biofilm formation.

Introduction

Dental caries begins with enamel demineralization and, if untreated, progresses irreversibly into the dentin, where management is limited to restorative procedures or tooth extraction. A surgeon general review in 2000 underscored the links between oral health and systemic health and emphasized persistent inequalities in access to proper dental care for all age, ethnic, and socioeconomic groups [1]. Despite decades of preventative efforts, data indicate that little progress has been made and caries remain highly prevalent [2]. In Canada, caries-related conditions account for 4 million lost workdays and 2 million lost school days annually, contributing to an estimated $12.8 billion dental care burden, second only to cardiovascular disorders in health care costs [3]. Globally, tooth decay is the most prevalent of 291 diseases recognized by the World Health Organization, affecting nearly all adults [4]. This highlights the need for scalable adjunct therapies to the existing prophylactic paradigm of brushing and flossing. Some suggested approaches include bioactive remineralization strategies (e.g. fluoride and calcium–phosphate technologies), biofilm-control, non-cariogenic sugar substitutes (notably xylitol, with mixed clinical evidence), and probiotics aimed at shifting the oral microbiome away from acidogenic, cariogenic communities rather than broadly sterilizing plaque [5–7].

Dental caries is now widely understood as a biofilm-mediated disease. The mutans group of Streptococci – most notable is Streptococcus mutans – are the keystone pathogens that produce exopolysaccharides from dietary sucrose that facilitate tooth adhesion and the formation of acidogenic biofilms (plaque) on the enamel hydroxyapatite [8,9]. Within the biofilms, a diverse microbiome ecology sustains carbohydrate fermentation that generates the localized acidic microenvironment, overcoming the buffering capacity of saliva to shift the community to a cariogenic state that becomes self-reinforcing without intervention. This dysbiotic ecology is further amplified by cross-kingdom interactions, as Candida albicans co-aggregates with S. mutans to form mixed-species biofilms that enhance biomass accumulation, acidogenicity, and biofilm stability, particularly in early childhood caries (ECC) [10,11]. For these reasons, conventional antimicrobial interventions are often short-lived in the context of a self-reinforcing plaque biofilm that rapidly re-establishes cariogenic dominance. In contrast, approaches that shift microbial ecology by selectively suppressing cariogenic taxa, disrupting biofilm structure, and enabling stable colonization by health-associated microbes, may prolong the therapeutic effects of brushing and flossing by stabilizing a non-cariogenic ecosystem for longer periods. Probiotic strategies designed to integrate into oral biofilms and competitively antagonize Streptococcus mutans represent a biologically aligned and potentially more durable approach to caries control [12,13].

Probiotic therapies have been explored as an all-natural means of regulating S. mutans and the broader microbiota, with most clinical and preclinical studies focusing on single strains [14–24]. However, the mechanisms of action are usually unspecified or, when studied, found to be highly strain-specific, including co-aggregation with S. mutans, the secretion of antimicrobial bacteriocins or hydrogen peroxide, and competitive colonization of plaque niches [25–27]. This mechanistic diversity suggests that combining strains with complementary activities may yield synergistic suppression of cariogenic biofilms compared to single strains alone. We therefore hypothesized that consortia of strains with anti-S. mutans effects may act synergistically to inhibit or reduce the formation of pathogenic biofilms in a clinically relevant in vitro environment.

Thus, we set out to systematically identify the relative synergistic capacity of probiotic strains. We first screened individual candidate probiotic strains for their anti-caries properties, including S. mutans antagonism, hydrogen peroxide production, and hydroxyapatite adhesion. We then assessed strains with anti-caries properties for multi-strain synergistic antagonistic effects against S. mutans isolates. Overall, the results illustrated that the combination of Lactiplantibacillus plantarum (formerly Lactobacillus plantarum) and Lactobacillus helveticus strains displays synergistic S. mutans antagonism, with strains L. plantarum Stellar™ Lp-199 and L. helveticus Stellar™ Lh-170 exhibiting the greatest combined potency. This combination was further evaluated for in vitro S. mutans and C. albicans biofilm preventive and disruptive abilities in comparison to S. salivarius DSM 14865 as a benchmark anti-caries strain [22]. Validation of the findings through in vitro biofilm models suggests that the consortia could be an effective microbiological adjuvant for prophylactic caries management.

Materials and methods

Cells, media, and growth conditions

Stocks of all lactobacilli and bifidobacteria were maintained on de Man-Rogosa-Sharpe (MRS; Difco) agar. Strains of S. mutans (including ATCC 25175 – a strain isolated from carious dentin) were maintained on brain‒heart infusion with 1% yeast extract (BHYE; Difco) agar. Streptococcus salivarius and Streptococcus thermophilus strains were maintained on CABK 12 agar [28]. C. albicans TIMM 1768, a strain known to induce oral candidiasis, was maintained on Sabouraud dextrose agar (SDA; Difco). Overnight broth cultures of all probiotics and C. albicans were prepared in YEPD (1% yeast extract; 2% peptone; 2% dextrose). Overnight cultures of S. mutans were prepared in BHYE. All of the above strains were cultured for 18–22 h under microaerophilic conditions (37°C, 5% CO₂). Selective plating for S. mutans was carried out on mitis salivarius-bacitracin (0.2 U/ml) agar (MSB; Difco) supplemented with 20% sucrose, for 48 h under microaerophilic conditions (37°C, 5% CO₂), as this method was found to reliably differentiate S. mutans from other streptococci and enrich the naturally harsh MSB medium. To mimic the physiological conditions of the oral cavity, hydroxyapatite adhesion assays and all biofilm assays were performed in basal medium mucin (BMM) artificial saliva prepared according to Wong and Sissons (2.5 g partially purified pig gastric mucin, 5 g yeast extract, 5 g protease peptone, 2.5  mg haemin, 1 mg menadione, 33.5  mmol KCl, 1 mmol arginine, 1 mmol urea, mixed in 1 L of Milli-Q water and autoclaved) [29]. For biofilm prevention and disruption assays, BMM was supplemented with 2% sucrose and 1% fetal bovine serum (Gibco), as these ratios promoted the most robust biofilm formation by S. mutans and C. albicans.

Dental plaque from two consenting adults was collected using sterile probes, spread directly onto MSB + 20% sucrose agar, and incubated for 48 hours at 37°C, 5% CO₂. Individual colonies were then quadrant streaked onto BHYE agar and incubated again for 24 h at 37°C, 5% CO₂. Genotyping was undertaken on the isolated colonies to confirm their species. Bacterial DNA was extracted using InstaGene Matrix (BioRad) according to the manufacturer’s instructions. All PCR reagents were obtained from Invitrogen. A total of 30  ng of DNA template was added to a master mix of 15  mM MgCl₂, 200  mM dNTPs, 10  µM MUT-F primer, 10  µM MUT-R primer and 5 U of Taq DNA polymerase in 1X PCR buffer for a 50 µL total reaction volume. The genotyping primers used were MUT-F 5′-ACTACACTTTCGGGTGGCTTGG-3′ and MUT-R 5′-CAGTATAAGCGCCAGTTTCATC-3′, specific to a 517-bp DNA region of the gtfB extracellular glucosyltransferase of S. mutans [30]. The thermocycling conditions were as follows: 95°C for 1  min; cycling 95°C for 30 s, 59°C for 30 s, 72°C for 1  min for 30 cycles, followed by 72°C for 5  min. The products were separated on a 1.5% agarose gel with 0.05% ethidium bromide.

Hydrogen peroxide production assay

To test the candidate probiotics for hydrogen peroxide production, a methodology similar to Kang et al. was employed [31]. Standard agar growth media (MRS, BHYE, CABK12) were modified by the addition of 0.25  mg/mL 3,3',5,5'-tetramethylbenzidine (TMB) and 0.1  ng/mL peroxidase. TMB was first dissolved in dimethyl sulfoxide (DMSO) (0.125 g/mL) and peroxidase in water (5 mg/mL), then filter sterilized and added to autoclaved liquid agar before plate preparation. Single colonies were streaked directly onto TMB/peroxidase and standard MRS/BHYE/CABK12 control plates before incubation for 72 h at 37°C, 5% CO₂. The colonies and surrounding agar were assessed for color every 24 h and compared to control plates. A total of three biological replicates were completed.

Measurement of deferred antagonistic activity

Deferred antagonism assays were performed as previously described by Tagg and Bannister [32]. Overnight suspensions, or 1:1 mixtures of overnight suspensions of Lactobacilli, Bifidobacteria, and Streptococci probiotics, were swabbed onto brain heart infusion (BHI; Difco) plates supplemented with 0.1% (w/v) CaCO₃ in a measured 1 cm wide edge-to-edge streak using sterile cotton swabs. The plates were incubated for 48 h at 37°C, 5% CO₂. Afterward, bacterial growth was removed from the surface with a glass microscope slide, and the plates were sterilized with chloroform vapor for 20 mins. Overnight suspensions of S. mutans were then swabbed perpendicular to the original growth streak, before incubating at 37°C, 5% CO₂ for 48 h. The zone of inhibition (ZOI) was calculated as the sum of the distances of growth inhibition on either side of the original probiotic streak. One-way ANOVAs followed by Bonferroni post hoc tests were used to compare means. Synergism was calculated as a synergistic quotient (SQ): ZOI of the combination of the strains, divided by the sum of the ZOIs of each individual strain, such that SQ > 1 indicates synergism [33]. The results represent at least four biological replicates for each condition. All statistics throughout were performed using Prism software (GraphPad, v6).

Hydroxyapatite adhesion assay

Overnight suspensions of probiotics and S. mutans were resuspended to 1 × 10⁶ CFU/ml in BMM. 2 mL of each suspension was added to independent wells of a 24-well polystyrene plate in triplicate, and sterile hydroxyapatite discs measuring 7 mm diameter × 1.8 mm thickness were placed in the center of each well (Clarkson Chromatography Products; Pennsylvania, USA). The plates were incubated at 37 °C, 5% CO₂ for either 6 h or 24 h before the discs were carefully removed and dipped twice in 500 µL of phosphate-buffered saline (PBS; Gibco) to remove non-adherent cells. Discs were then dropped in 1  mL of PBS in a 1.5  mL Eppendorf tube, scraped with a cell-scraper, and sonicated for 15 minutes before enumeration by ‘drop plating’: 10-fold serial dilutions were made, and three 10 µL drops of each dilution were pipetted onto the appropriate agar for each bacterium before incubating for 24 h at 37°C, 5% CO₂. The dilutions that yielded between 5 and 50 colonies were counted, triplicates were averaged, and CFU per disc was determined. Statistical comparisons between groups were made using a two-way ANOVA with Tukey’s multiple comparison.

Inhibition of biofilm formation and pathogen viability in vitro

Overnight cultures of pathogens (S. mutans and C. albicans) were resuspended to concentrations of 1 × 10⁷ CFU/mL in the BMM and probiotics to 1 × 10⁸ CFU/mL in the BMM. Regardless of test group, the concentrations of pathogens and probiotics per well were held constant: 1 × 10⁶ CFU total pathogens and 1 × 10⁷ CFU total probiotics per well of a 96-well polystyrene plate, with a constant 200 µL volume. Thus, for assessing the prevention of mono-species S. mutans biofilms, 100 µL of S. mutans was added, and for dual-species S. mutans/C. albicans biofilms, 50 µL of each was added. Likewise, 100 µL of each probiotic was inoculated per well for tests of individual strains, and when Lh-170 and Lp-199 were combined, 50 µL of each was added. When assessing the dose-dependent effects, pathogens were inoculated at 10 [6] total CFU, while total probiotic CFUs were administered in 10-fold dilutions from 10 [6] to 10 [3] per well. All test groups were prepared in duplicate. Following incubation for 24 h at 37°C, 5% CO₂, the supernatant was carefully removed without disturbing the biofilm. The biofilms were gently washed twice with 200 µL of PBS to remove non-adherent cells, and then 200  µL of fresh BMM was applied before incubating the plates again for 24 h at 37 °C, 5% CO₂. Finally, the supernatant was removed, and the biofilms were gently washed twice with 200 µL of PBS before biofilm prevention was assessed in one of two ways:

• a)Crystal violet staining as per Abe et al. (1994) [34] – biofilms were fixed with 70% ethanol, stained with 0.5% (w/v) crystal violet, and OD₆₀₀ readings were taken. The results represent the means ± SEM of three biological replicates with three technical replicates each. Statistical comparisons between treatment groups and untreated controls were made using a two-way ANOVA with Tukey’s multiple comparison test.
• b)Drop plating on selective media – a final 250 µL of PBS was added to each well. Cell scrapers were used to detach biofilms from the wells, and a wide-bore P1000 pipette was used to transfer all material to labeled 1.5 mL Eppendorf tubes before sonicating for 15 min. Ten-fold serial dilutions of the sonicated samples were prepared in PBS before plating on selective media (MSB + 20% sucrose for S. mutans; SDA for C. albicans), and incubating for 24 h at 37°C, 5% CO₂. Enumerations of two technical replicates were used to calculate the mean viable S. mutans or C. albicans CFUs from the biofilms within each biological replicate. The results represent the mean ± SEM of at least three biological replicates. Statistical comparisons between treatment groups and untreated controls were made using a two-way ANOVA with Tukey’s multiple comparison test.

Disruption of pre-formed S. mutans biofilms in culture

Similar to above, 1 × 10⁶ CFU total of pathogens (S. mutans alone or in combination with C. albicans) were inoculated into the wells of a 96-well polystyrene plate, with a total volume of 200 µL of BMM. The plates were incubated for 24 h at 37°C, 5% CO₂ to allow biofilms to establish, before the supernatant was carefully removed. The biofilms were gently washed twice with 200 µL of PBS to remove non-adherent cells. A total of 1 × 10⁷ CFUs of probiotics (DSM 14865 or Lh-170 and Lp-199) suspended in 200 µL of BMM, were added to each well except the untreated control (200 µL of BMM only). All test groups were prepared in duplicate. The samples were incubated at 37°C, 5% CO₂ for 10  min, 6  h, 9  h, 12  h, 24  h, 36  h and 48  h to assess the rate of biofilm disruption. At each endpoint, the samples were washed twice with 200 µL of PBS to remove non-adherent cells and resuspended in 250 µL of fresh PBS. Cell scrapers were used to detach biofilms from the wells, and a wide-bore P1000 pipette was used to transfer all material to labeled 1.5 mL Eppendorf tubes before sonicating for 15 min. Ten-fold serial dilutions of the sonicated samples were prepared in PBS before plating on selective media. Enumerations of two technical replicates were used to calculate the mean viable S. mutans or C. albicans CFUs from the biofilms within each biological replicate. Drop plating on MRS was used to evaluate the ability of the Lh-170 and Lp-199 combination to penetrate the pre-formed biofilms. The results represent the means ± SEM of at least three biological replicates. Statistical comparisons between treatment groups and untreated controls were made using a two-way ANOVA with Dunnett’s multiple comparison test.

Disruption of pre-formed S. mutans biofilms on extracted human molars

Extracted tooth specimens (permanent 1^st and 2^nd molars) were obtained from the Interface Center for Oral and Maxillofacial Surgery (London, Ontario, Canada) after approval by the Fanshawe College Research Ethics Board (Protocol: 17-02-02-1). The teeth were cleaned of excess tissue via mechanical scraping, sectioned into halves of approximately equal size with a diamond bur, and submerged in 13% sodium hypochlorite for 24 h. Each half was rinsed twice in 5 ml of sterile PBS and then stored in fresh PBS until used.

One-half of a molar was incubated with 1 × 10⁶ CFU of S. mutans in 6  ml of BMM for 24 h at 37°C, 5% CO₂, on a rotary stand set to ~100 rpm to promote biofilm formation. Using sterile forceps, each tooth was then gently dipped in 6 mL of PBS twice to wash away non-adherent cells. A total of 1 × 10⁷ CFUs of probiotics (DSM 14865 or Lh-170 and Lp-199) suspended in 6 mL of BMM were added to each tooth sample except the untreated control (6 mL of BMM only). The samples were incubated at 37°C, 5% CO₂ on a rotary stand for 10  min, 6  h, 9  h, 12  h, 24  h, 36  h, and 48  h to assess the rate of biofilm disruption. At each endpoint, samples were gently dipped in 6 mL of PBS twice to remove non-adherent cells, then transferred to a 15 mL Falcon tube containing 6 mL of PBS, where they were sonicated for 30 min. Ten-fold serial dilutions of the sonicated samples were prepared in PBS before plating on selective media. Enumerations of two technical replicates were used to calculate the mean viable S. mutans CFUs from the biofilms within each biological replicate. Drop plating on MRS was used to evaluate the ability of the Lh-170 and Lp-199 combination to penetrate the pre-formed biofilms. The final results represent the means ± SEM of at least three biological replicates. Statistical comparisons between treatment groups and untreated controls were made using a two-way ANOVA with Dunnett’s multiple comparison test.

Results

Specific strains of L. helveticus and L. plantarum are highly synergistic in their secreted antagonism of S. mutans

Since hydrogen peroxide is a common interbacterial antagonistic metabolite, and S. mutans is susceptible to oxidative stress, we screened candidate probiotic strains for its production as a potential anti-cariogenic mechanism. Most Lactobacillus strains (10 of 11), both Streptococcus thermophilus strains, and all tested Bifidobacterium strains produced detectable hydrogen peroxide, whereas S. salivarius strains did not (Table 1).

Table 1.

Hydrogen peroxide production by probiotic strain.

Production within 24 hours	Production within 48 hours	Production within 72 hours	No production by 72 hours
S. thermophilus R0083	L. acidophilus La-14 L. reuteri Lru-1038 L. salivarius Ls-33	S. thermophilus St-21 L. rhamnosus GG L. plantarum Stellar Lp-185 L. rhamnosus Lr-32 L. casei Lc-11 B. lactis Bl-04 B. breve Bb-03 B. bifidum R0071	S. salivarius BAA-1024 S. salivarius DSM 14865 L. paracasei Lpc-37
L. plantarum Stellar Lp-199
L. helveticus Stellar Lh-172
L. helveticus Stellar Lh-170
B. longum R0175

S. mutans ATCC 25175 was used as the initial benchmark for qualifying deferred antagonism. Only five strains (S. salivarius BAA-1024, S. salivarius DSM 14865, L. plantarum Stellar Lp-199, L. helveticus Stellar™ Lh-172, and L. helveticus Stellar^TM Lh-170) from Table 1 produced any inhibitory effects on ATCC 25175 within a 1 cm wide streak (Supplemental Table 1) and were therefore investigated for quantitative synergies against ATCC 25175 and four other S. mutans isolates by measuring the additive ZOIs external to the 1 cm streak in all possible two-strain combinations (Table 2). In general, S. mutans ATCC 25175 was more resistant to antagonistic factors secreted by the probiotics compared to fresh isolates – no individual probiotics displayed S. mutans ATCC 25175 inhibition adjacent to the 1 cm-wide probiotic streak, and few combinations did, whereas each isolate was inhibited by at least 3 individual probiotics and multiple probiotic combinations (Table 2). As single strains, L. helveticus Stellar Lh-170 and L. helveticus Stellar Lh-172 displayed the widest range of inhibition.

Table 2.

Determination of synergistic deferred antagonism against five isolates of S. mutans.

	S. mutans 25175			S. mutans 1			S. mutans 2			S. mutans 3			S. mutans 4
Strain or Combination	ZOI (mm)	SQ		ZOI (mm)	SQ		ZOI (mm)	SQ		ZOI (mm)	SQ		ZOI (mm)	SQ
BAA-1024	0.00	N/A		11.5	N/A		6.50	N/A		0.00	N/A		1.5	N/A
DSM 14865	0.00	N/A		0.75	N/A		0.00	N/A		0.75	N/A		0.00	N/A
Stellar Lp-199	0.00	N/A		1.14	N/A		1.14	N/A		3.57	N/A		5.71	N/A
Stellar Lh-172	0.00	N/A		1.67	N/A		5.33	N/A		6.67	N/A		5.67	N/A
Stellar Lh-170	0.00	N/A		4.71	N/A		4.00	N/A		1.00	N/A		0.00	N/A
BAA-1024 + DSM 14865	0.00	0.00		1.00	0.08		0.50	0.08		1.25	1.67		0.00	0.00
BAA-1024 + Stellar Lh-170	0.00	0.00		6.75	0.42		1.00	0.10		4.50	1.26		0.00	0.00
BAA-1024 + Stellar Lh-172	0.00	0.00		3.25	0.25		0.00	0.00		2.75	0.41		0.00	0.00
BAA-1024 + Stellar Lp-199	0.00	0.00		3.00	0.24		1.50	0.20		2.67	2.67		0.00	0.00
DSM 14865 + Stellar Lh-170	0.00	0.00		5.00	0.92		6.00	1.50		9.50	2.20		2.50	0.44
DSM 14865 + Stellar Lh-172	0.00	0.00		4.25	1.76		3.25	0.61		4.50	0.61		0.00	0.00
DSM 14865 + Stellar Lp-199	5.75	∞		11.00	5.81		10.50	9.19		10.25	5.86		8.50	∞
Stellar Lh-170 + Stellar Lh-172	0.00	0.00		6.50	1.02		1.75	0.19		5.00	0.49		4.25	0.37
Stellar Lh-170 + Stellar Lp-199	13.00‡	∞		20.14‡	3.44		20.29‡	3.94		19.29‡	4.22		23.86‡	4.18
Stellar Lh-172 + Stellar Lp-199	0.00	0.00		9.25†	3.29		9.75*	1.51		13.75†	1.79		9.25*	1.63

* = p<0.05; † = p<0.01; ‡ = p<0.001.

SQ = synergistic coefficient (probiotic combination ZOI / sum of individual probiotic ZOIs).

N/A = not applicable.

Note: ZOIs are measured externally to the initial 1-cm-wide streak of probiotics.

The average ZOIs were often greater when two probiotic strains were combined. Together, L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 resulted in the largest ZOIs against all five strains of S. mutans. The ZOIs of this combination were significantly larger than the sum of the individual ZOIs against all S. mutans strains (p < 0.001), indicating synergistic inhibition (Table 2). The synergistic quotients were calculated to be  ∞ , 3.44, 3.94, 4.22 and 4.18 against S. mutans ATCC 25175, and isolates 1–4, respectively, where a quotient greater than 1 indicates a level of combined antagonistic activity that is greater than the sum of each probiotic’s individual antagonistic activity. The pairing of L. plantarum Stellar Lp-199 with L. helveticus Stellar Lh-172 also resulted in synergism against all isolates (p < 0.05); and with S. salivarius DSM 14865 resulted in non-significant synergism against all five S. mutans strains. However, neither of these combinations displayed ZOIs as large as the L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 combination.

Select strains adhere to hydroxyapatite as a model of the tooth surface

Durable prophylactic antagonism is contingent on probiotic localization with S. mutans at tooth surfaces, the primary site of cariogenic biofilm formation. Thus, we evaluated the ability of L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 to adhere to hydroxyapatite in the BMM, in comparison to S. salivarius DSM 14865 and S. mutans 25175 (Figure 1). After 6 h, L. helveticus Stellar Lh-170 displayed the weakest adhesion (1.2 × 10⁵ CFU/disc), and S. mutans displayed the strongest (5.9 × 10⁵ CFU/disc); however, none of the differences were significant. After 24 h, all bacteria trended towards increased adherence relative to that at 6 h, but only L. plantarum Stellar Lp-199 was significantly increased (p < 0.001), with values of 1.06 × 10⁷ versus 3.9 × 10⁵ CFU/disc. L. plantarum Stellar Lp-199 also displayed significantly increased adherence relative to all other tested bacteria, including S. mutans, at 24 hours (p < 0.001).

Figure 1.

Adhesion to hydroxyapatite. 2x10⁶ CFU of each bacterium was inoculated onto a 7mm x 1.8mm hydroxyapatite disc in a 24-well plate and incubated for 6 or 24 h. Non-adherent cells were removed and adherent cells were quantified by drop plating. The results represent the means ± SEM of three biological replicates with three technical replicates each (*** = p < 0.001).

L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 prevent biofilm formation

Given that S. mutans’ role as a key in biofilm generation, and our demonstrated synergistic antagonism of S. mutans by the L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 combination, we sought to evaluate their joint ability to prevent S. mutans biofilm formation compared to S. salivarius DSM 14865 as a benchmark anti-caries strain [22]. Using crystal violet staining, both Lh-170 and the Lh-170/Lp-199 combination reduced biofilm formation by approximately 62%, whereas Lp-199 reduced biofilm formation by 25% and that of DSM 14865 by only 5% (Figure 2A), but none of the group’s differences were statistically significant from each other or the control. Using selective plating for S. mutans, all the groups significantly reduced S. mutans viability relative to the control by at least 99% (p < 0.05; Figure 2B). No treatment group was significantly better than another. However, no viable S. mutans cells were recoverable after co-incubation with the Lh-170/Lp-199 combination, representing a > 10,000,000-fold reduction.

Figure 2.

Inhibition of S. mutans biofilm formation by probiotics. 10⁶ CFU of S. mutans and 10⁷ total CFU of probiotics were co-incubated for 24 h on 96-well polystyrene plates. Non-adherent cells were removed, and adherent cells were quantified by: (A) crystal violet staining; (B) selective drop plating for S. mutans. The results represent the means ± SEM of at least three biological replicates. Significant reductions in S. mutans viability were observed for all treatment groups compared to the control (* = p < 0.05).

Recently, analysis of dental plaques from children with dental caries revealed that S. mutans and C. albicans commonly form mixed-species biofilms, and that the co-occurrence of these pathogens may increase susceptibility to dental caries [11,35]. Previous research by our group has revealed the ability of probiotics to prevent C. albicans biofilm formation [36]. Thus, we assessed the ability of S. salivarius DSM 14865 and the L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 combination to prevent mixed-species biofilm formation by S. mutans and C. albicans when co-incubated (Figure 3A/3C). We also tested the dose-dependency of the effect by progressively decreasing the initial probiotic inoculum by factors of 10 (Figure 3B/3D). At an initial inoculum of 10 [7] CFU – 10x greater than the pathogens – DSM 14865 did not significantly reduce S. mutans viability (3.76-fold reduction) but did significantly reduce C. albicans viability by 14-fold (p < 0.01). Comparatively, the Lh-170/Lp-199 combination reduced S. mutans and C. albicans viability by 6182-fold and 231-fold, respectively (p < 0.01). At a 1:1 ratio with pathogens, the Lh-170/Lp-199 combination still yielded a statistically significant 17.5-fold reduction (p < 0.05) in viable S. mutans. Interestingly, when the ratio was skewed 10:1 in favor of pathogens, the combination still significantly reduced S. mutans viability by 7.7-fold (p < 0.05; Figure 3B).

Figure 3.

Inhibition of S. mutans/C. albicans mixed species biofilm formation by probiotics. 5x10⁵ CFU of S. mutans and C. albicans, and 10⁷ total CFU of probiotics (A/C) or 10³-10⁶ total CFU of probiotics (B/D) were co-incubated for 24 hours on 96-well polystyrene plates. Non-adherent cells were removed, and adherent cells were quantified by selective drop plating for S. mutans on MSB agar or C. albicans on SDA. Results represent the means ± SEM of at least three biological replicates. The L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 combination displayed the greatest level of mixed-species biofilm prevention for both S. mutans and C. albicans at a 10⁷ inoculum (A/C) and still significantly reduced S. mutans viability when outnumbered 10:1 by pathogens (B) (* = p < 0.05).

L. helveticus stellar Lh-170 and L. plantarum stellar Lp-199 disrupt pre-formed biofilms both in culture and on human molars

Building on our hydroxyapatite adhesion data and our in vitro findings showing that the two-strain probiotic combination best inhibits S. mutans biofilms in culture, we sought to directly visualize these combined effects in an ex vivo tooth-surface model for translatable relevance. In a one-off demonstration (Supplemental Figure 1), one half of a molar was incubated with either 1 × 10⁶ CFU of S. mutans alone, or in combination with 5 × 10⁶CFU of each of L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 for 24 h. Non-adherent cells were removed, and fresh media were added for an additional 24-hour incubation before SEM imaging to assess biofilm formation. biofilm accumulation on the tooth surface of control teeth, but not probiotic-inoculated teeth. The apparent inhibition was consistent with our culture-based results (Figures 2 and 3).

With apparent near-total S. mutans inhibition in culture and exemplified on donor tooth surfaces, and to render this observation into biologically relevant data, we next assessed whether the L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 combination or S. salivarius DSM 14865 harbored any ability to disrupt pre-formed S. mutans biofilms. S. mutans were inoculated into 96-well plates (1 × 10⁶ CFU) and incubated for 24 hours to establish biofilms. Probiotics were then added to each well (1 × 10⁷ CFU) to assess the efficacy and rate of biofilm disruption over a 48-hour time course. After a 9-hour exposure, the Lh-170/Lp-199 combination significantly reduced the number of viable S. mutans biofilm cells by 30-fold (p < 0.01), whereas DSM 14865 yielded a 1.7-fold reduction. After 12 h, the Lh-170/Lp-199 combination significantly reduced the number of viable S. mutans biofilm cells by 85-fold (p < 0.0001), while DSM 14865 provided a statistically significant 6-fold reduction (p < 0.001). After 24 h, no viable S. mutans cells were recoverable when treated with the Lh-170/Lp-199 combination (p < 0.0001), while DSM 14865 only reduced the number of viable S. mutans biofilm cells by 6-fold. DSM 14865 achieved a 41-fold reduction in S. mutans viability by 48 h (Figure 4A). Enumeration by drop plating of treated biofilm samples revealed that the Lh-170/Lp-199 combination reduced S. mutans cell viability, at least one strain of lactobacilli remained viable and integrated into the pre-existing biofilms (Figure 4B). When a comparable experiment was carried out on extracted human half-molars, similar results were achieved (Figure 4C/D). After 9, 12, 24, 36 and 48 h, the Lh-170/Lp-199 combination significantly reduced the number of viable S. mutans biofilm cells by 32-fold (p < 0.01), 54-fold (p < 0.0001), 3,135-fold (p < 0.0001), 131,836-fold (p < 0.01), and 140,756-fold (p < 0.05), respectively. DSM 14865 provided statistically significant 5-fold (p < 0.01), 10-fold (p < 0.0001), 9.5-fold (p < 0.001), 8-fold (p < 0.01), and 57-fold (p < 0.05) reductions over the same time periods (Figure 4C). Viable lactobacilli were recovered from biofilms on the teeth (Figure 4D).

Figure 4.

Disruption of pre-formed S. mutans biofilms by probiotics. 10⁶ CFU of S. mutans were incubated in BMM for 24 hours on 96-well polystyrene plates (A/B) or on one half of a human molar (C/D). The nonadherent cells were removed, and 1 x 10⁷ CFU of probiotics were added to each well before incubating for 10–48 h. Non-adherent cells were removed, and adherent cells were quantified by selective drop plating for S. mutans (A/C) or lactobacilli (B/D). The results represent the means ± SEM of at least three biological replicates. The L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 combination performed better than S. salivarius DSM 14865 at disrupting S. mutans biofilms on 96-well plates and teeth.

Discussion

In this study, we demonstrated that combining select probiotic strains can produce synergistic suppression of Streptococcus mutans and interfere with the development and persistence of S. mutans within cariogenic biofilms in vitro. Screening for diffusible antagonism identified several candidate strains with combinations of L. helveticus and L. plantarum excelling. The specific pairing of L. helveticus Stellar Lh-170 with L. plantarum Stellar Lp-199 yielded synergistic inhibition of S. mutans, exceeding the activity of either strain alone, and outperforming the commonly studied S. salivarius DSM 14865 as a benchmark anti-cariogenic strain. Importantly, this combination proved effective in biologically relevant contexts, including adhesion to hydroxyapatite, prevention of S. mutans and C. albicans biofilm formation, and the disruption of established S. mutans biofilms on human molars. The ability of at least one member of this probiotic combination to attach to biofilms while progressively reducing S. mutans viability supports a theoretical mechanism in which sustained antagonism is achieved through biofilm localization and continued secretion of inhibitory factors, rather than simply transient antimicrobial exposure. The in vitro antagonistic experiments support a probiotic approach to caries prevention that targets biofilm resilience and microbial community dynamics.

Through various mechanisms, several bacterial strains have demonstrated their abilities to inhibit dental caries-related pathogens. Notably, S. salivarius DSM 14865 has been shown both in vitro and in clinical studies to express anti-S. mutans activity [22]. Additionally, Lactobacillus rhamnosus GG has displayed preliminary short-term and long-term anti-dental caries effects [16,37]. Here, we demonstrate that several strains of probiotic bacteria exceed these two probiotics with respect to their S. mutans antagonism. L. rhamnosus GG did not display antagonism against any of the five strains of S. mutans used in this study. Interestingly, several probiotics—including L. plantarum Stellar Lp-199 and S. salivarius BAA-1024—displayed greater antagonism than DSM 14865 for at least one strain of S. mutans, while both L. helveticus strains studied (Stellar Lh-170 and Stellar Lh-172) performed with greater or equal antagonism against all five S. mutans strains. Compared to all individual probiotics and two-strain combinations, the combination of Lh-170 and Lp-199 yielded the strongest antagonism against all S. mutans strains, and the ZOIs were at least two-fold larger than the additive ZOIs of Stellar Lh-170 and Stellar Lp-199 alone. This action is particularly impressive against the S. mutans 25175 type strain that is exceptionally pathogenic (having been extracted from a carious lesion) and was not inhibited beyond the initial 1 cm streak by any individual strains (Table 2) [38].

It is unknown what metabolic mechanisms are responsible for the observed synergistic action between Lh-170 and Lp-199. Some strains of L. helveticus, secrete bacteriocins such as helveticin J, and helveticin V-1829. Likewise, strains of L. plantarum produce a variety of bacteriocins and other antimicrobial substances encoded at the pln locus, whose secretion is regulated through a quorum-sensing mechanism [39–41]. Although the production of specific antimicrobial factors by Lp-199 has not been verified, their expression is thought to be relatively common among L. plantarum strains [42]. Thus, antimicrobial peptides secreted by these two probiotic strains may form a particularly effective synergistic combination that promotes complementary lethality against various S. mutans strains. The non-streptococcal strains that exhibited S. mutans antagonism were amongst the strongest H₂O₂ producers; however, quick production of H₂O₂ was not sufficient to cause antagonism, as B. longum R0175 rapidly produced H₂O₂ but yielded no S. mutans inhibition. H₂O₂ is evidently not necessary for antagonism, as S. salivarius BAA-1024 and DSM 14865 are H₂O₂ non-producers, and therefore may exert their anti-S. mutans activity through secretion of specific bacteriocins [43]. Paradoxically, S. mutans is able to both produce and degrade H₂O₂, but is nevertheless susceptible to it in concentrations above its degradation threshold [44]. Although the mechanism remains unknown, the ability of both Lh-170 and Lp-199 to adhere to hydroxyapatite surfaces, and potentially S. mutans biofilms, likely allows these strains to exert their S. mutans antagonism in a localized manner.

Not only is biofilm formation on hydroxyapatite a means of increasing antibiotic and immune evasion by S. mutans, but it is also essential for dental caries disease pathogenesis. Additionally, recent studies have revealed a relationship between C. albicans – the biofilm-forming agent behind oral candidiasis – S. mutans, and dental caries biofilm development [10,11,35]. Lactobacilli have been shown to inhibit biofilm formation by both of these species individually [36,45,46]. Thus, we tested the L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 combination for its ability to prevent S. mutans biofilms alone or in combination with C. albicans. With respect to S. mutans mono-species biofilm formation, the Lh-170 and Lp-199 combination yielded a 62% (n.s.) reduction in biofilm biomass, while significantly reducing the viability of S. mutans by ~10,000,000-fold, compared to a 5% (n.s.) reduction in biomass and ~10,000-fold reduction in S. mutans viability associated with DSM 14865. In mixed-species biofilms, the Lh-170 and Lp-199 combination resulted in significant > 6000-fold and > 200-fold reductions in S. mutans and C. albicans viability, respectively, compared to 3.76 (n.s.) and a significant 14-fold reduction by DSM 14865. The reduction in S. mutans viability is consistent with some studies that show inhibition when incubated with probiotic candidates, such as L. plantarum K41 (98.4% reduction) and W. cibaria CMU (~95% reduction) [46,47]. The reduction in C. albicans viability is consistent with our previously published data [36].

A strength of our experimental design was the evaluation of probiotic disruption of pre-formed S. mutans biofilms in culture and on human teeth in vitro. After 24 h, the Lh-170 and Lp-199 combination significantly reduced S. mutans viability within pre-formed biofilms by over 10,000-fold on 96-well plates, and by 3,135-fold on teeth. In both experiments, at least one of the lactobacilli was able to adhere and persist in the biofilms, which is congruent with our data that they adhered well to hydroxyapatite. In contrast, DSM 14865 reduced the number of viable S. mutans biofilm cells by 6-fold on 96-well plates and 9.5-fold on teeth in the same timeframe. It is well known that quorum sensing in S. mutans will naturally dissipate biofilms at high cell densities, limiting the scope and reliability of 96-well plate models, especially at longer timepoints [48]. This could explain the sharp decrease in adherent S. mutans cells after 24 h on polystyrene, even in the control group. We posit that our tooth model circumvents this issue, as reflected in the data (Figure 4A vs. 4C), owing to a larger volume of solution and surface area for binding, delaying the effects of quorum sensing on CFU counts. Irrespective of which model of pre-formed biofilms we used, the Lh-170 and Lp-199 combination proved superior to DSM 14865 at reducing S. mutans viability.

Clinical studies have already demonstrated mediation of dental health and dental caries risks in children and adults using single-strain probiotic products. Compared to placebo, consumption of milk supplemented with L. rhamnosus LB21 and fluoride slowed the decayed-missing-filled-teeth score progression by 75% in children aged 1–5 over a 21-month intervention and yielded additional health benefits, such as decreased incidence of otitis media and decreased need for antibiotic therapy [17]. A similar prevention study in children utilizing milk supplemented with L. rhamnosus GG reduced caries incidence and lowered the mutans streptococci counts [16]. Short-term administration of Bifidobacterium lactis Bb-12, Bifidobacterium animalis subsp. lactis DN-173 010, or L. reuteri ATCC 55730, through various delivery systems have also shown reductions in salivary mutans streptococci in adults [14,23,24]. Finally, the treatment of dental caries-active children with lozenges containing S. salivarius DSM 14865 reduced S. mutans counts and plaque scores [22]. Based on our preliminary in vitro findings, we suggest that the L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 combination may be effective for the management of dental caries-causing microbes, including S. mutans and C. albicans, and should be evaluated clinically.

Owing to the acidogenic and aciduric nature of lactobacilli, acidification of saliva could negatively contribute to caries progression. A recent double-blind placebo-controlled clinical trial of the effects of a probiotic lozenge containing L. helveticus Lh-170, L. plantarum Lp-199, and a remineralization agent, on enamel and salivary parameters revealed no significant differences in salivary buffering capacity or volume, but significant increases in the salivary flow rate and pH (p < 0.05), and a 32.6% improvement in the DIAGNOdent enamel health scores (p < 0.01) after 56 days compared to placebo [49,50]. While we did not measure the effect of probiotics on pH in our in vitro studies, these in vivo results are reassuring that they do not contribute to drastically altered pH nor a cariogenic nature of these specific strains. Salivary microbiological studies would strengthen the clinical evidence for their efficacy and lack of harm.

Conclusions

Our results indicate that the combination of specific L. helveticus and L. plantarum strains produces synergistic antagonism against S. mutans and C. albicans, with the pairing of L. helveticus Stellar Lh-170 and L. plantarum Stellar Lp-199 yielding the strongest and most consistent effect. Beyond synergistic inhibition of S. mutans growth, this combination effectively prevented biofilm formation and disrupted established S. mutans biofilms, performing as well or better than S. salivarius DSM 14865. Further clinical evaluation of the two-strain combination may clarify its in vivo capacity to reduce the cariogenic microbial burden and promote conditions favorable to enamel remineralization.

Funding Statement

Funding was provided by the Natural Sciences and Engineering Research Council of Canada (NSERC), awarded to PC and Stellar Biome Inc. as an ARD grant. MS was supported by a FedDev Ontario Graduate Enterprise Internship salary grant. GPT5 was used for light language improvement in some places. MS is supported by the Texas Children’s Hospital Research Scholar Program. The experiments were conceived and designed by MS, KJ, PC and JB. MS, KJ, and ZK conducted the experiments, generated, analyzed, and interpreted the data. The original manuscript was drafted by KJ and MS. PC, JB, KJ and MS edited the manuscript. All authors reviewed and approved the final manuscript.

Disclosure statement

This study was conducted in collaboration with Stellar Biome, Inc., which holds related intellectual property and to which some authors are professionally affiliated. No authors receive personal financial benefit from this intellectual property. These relationships are disclosed in accordance with journal policy.

Ethical approval

Extracted tooth specimens (permanent 1^st and 2^nd molars) were obtained from the Interface Center for Oral and Maxillofacial Surgery (London, Ontario, Canada) after approval by the Fanshawe College Research Ethics Board (Protocol: 17-02-02-1).

Supplementary material

Supplemental data for this article can be accessed at https://doi.org/10.1080/20002297.2026.2723648.