Inhibitory effect of Lactobacillus salivarius on Streptococcus mutans biofilm formation
Chien-Chen Wu1, Ching-Ting Lin2, Ching-Yi Wu3, Wu-Shun Peng1, Ming-Ju Lee1, and Ying-Chieh Tsai1*
1 Institute of Biochemistry and Molecular Biology, National Yang-Ming University, Taipei, Taiwan, Republic of China
2 School of Chinese Medicine, China Medical University, Taichung, Taiwan, Republic of China
3 Institute of Oral Biology, National Yang-Ming University, Taipei, Taiwan, Republic of China
Running title: L. salivarius inhibit S. mutans biofilm formation
Keywords: probiotic, plaque, exopolysaccharide, glucosyltransferase
* Correspondence: Dr. Ying-Chieh Tsai, Institute of Biochemistry and Molecular Biology, National Yang-Ming University, 155, Section 2, Linong Street, Taipei 11221, Taiwan. Republic of China.
Phone: 886-2-28267000 ext. 5641. FAX: 886-2-28264843. E-mail:
Word count = 6216 (including all tables, legends and references)
Dental caries arises from an imbalance of metabolic activities in dental biofilms developed primarily by Streptococcus mutans. This study was conducted to isolate potential oral probiotics with antagonistic activities against S. mutans biofilm formation from Lactobacillus salivarius, frequently found in human saliva. We analyzed 64 L. salivarius strains and found that two, K35 and K43, significantly inhibited S. mutans biofilm formation with inhibitory activities more pronounced than those of Lactobacillus rhamnosus GG (LGG), a prototypical probiotic that shows anti-caries activity. Scanning electron microscopy showed that S. mutans co-cultured with K35 or K43 resulted in significantly reduced amounts of attached bacteria and network-like structures, typically comprising exopolysaccharide. Spot assay for S. mutans indicated that K35 and K43 strains possessed a stronger bactericidal activity against S. mutans than LGG. Moreover, quantitative real-time PCR showed the expression of genes encoding glucosyltransferases, gtfB, gtfC, and gtfD was reduced when S. mutans were co-cultured with K35 or K43. However, LGG activated the expression of gtfB and gtfC, but did not influence the expression of gtfD in the co-culture. A transwell-based biofilm assay indicated that these lactobacilli inhibited S. mutans biofilm formation in a contact-independent manner. In conclusion, we identified two L. salivarius strains with inhibitory activities on the growth and expression of S. mutans virulence genes to reduce its biofilm formation. This is not a general characteristic of the species, therefore presenting a potential strategy for in vivo alteration of plaque biofilm and caries.
The human mouth harbors approximately 1000 bacterial species, reaching homeostasis in the oral cavity (Dewhirst and others 2010; Lazarevic and others 2010). Oral bacteria are responsible for the two most common bacterial diseases in humans, i.e. dental caries and periodontal disease (Wade 2013). Extensive investigation of dental caries has shown Streptococcus mutans is the major pathogen (Bowen and Koo 2011; Smith and Spatafora 2012). S. mutans does not have a free-living lifestyle. Its natural habitat is the human mouth, specifically found in the dental plaque on the tooth surfaces (Bowen and Koo 2011; Takahashi and Nyvad 2011). Dental plaque is a biofilm that consists of a group of microorganisms embedded in a matrix composed mainly of insoluble polysaccharides. Bacterial or salivary proteins that are associated with bacterial adhesion or aggregation, lipid and nucleic acids can also be found (Bowen and Koo 2011). The major virulent traits of S. mutans include: (i) its acidogenicity that exacerbates the damage to dental hard tissues, (ii) its aciduricity that contributes to its survival in low pH environments or to its out-competition against other oral bacteria and (iii) its ability to synthesize insoluble exopolysaccharide (EPS) from sucrose, which is involved in the initial attachment, colonization and accumulation of dental plaque (Koo and others 2013; Takahashi and Nyvad 2011). S. mutans expresses three glucosyltransferases (Gtfs) to synthesize EPS. GtfB and GtfC produce insoluble EPS, while GtfD forms a soluble, readily metabolized polysaccharide and acts as a primer for GtfB (Bowen and Koo 2011; Koo and others 2013).
Probiotics are defined as living bacteria that, when administered in adequate amounts, confer health benefits to the host (FAO/WHO 2001). Most microorganisms identified to date as probiotics are Gram-positive and belong to the genera Lactobacillus or Bifidobacterium that have been used for centuries because of their benefits to human health (Behnsen and others 2013; Turroni and others 2014). In recent years, the interest in probiotic therapies to prevent and control oral diseases has grown significantly (Tanzer and others 2010; Twetman and Keller 2012; Yanine and others 2013). While the existing animal and clinical trials have apparent limitations and additional studies are required, several reports have demonstrated the therapeutic potential of probiotics as anti-caries treatments (Twetman and Keller 2012). This concept is based on the idea of maintaining or restoring the natural microbiome in oral biofilm via interference and/or inhibition of pathogenic bacteria. Specific modes of bacterial interference include the secretion of anti-microbial substances, such as bacteriocins, in addition to competition for nutrients and adhesion. Protection against oral pathogen-induced inflammation by probiotic stains has also been described (Yanine and others 2013). However, the manner by which probiotics exerts antagonistic activities against oral pathogens remains largely unknown.
In this study, we focused on the identification of potential oral probiotics from Lactobacillus salivarius strains. These specific microorganisms are most frequently found on the tooth surface (26.7%) (Colloca and others 2000), and are also isolated from human saliva (8.3-48%) (Colloca and others 2000; Koll-Klais and others 2005; Nelun Barfod and others 2011) or the surface of tongue (5.6-9.5%) (Ahrne and others 1998; Colloca and others 2000). Since persistent presence of a probiotic strain makes its interference of pathogens readily, L. salivarius are widely studied as candidate probiotics in the oral cavity (Messaoudi and others 2013; Neville and O'Toole 2010; Strahinic and others 2007). In this study, 64 L. salivarius strains were isolated and analyzed for their inhibitory activities on S. mutans biofilm formation in a co-culture model. A prototypical probiotic strain, L. rhamnosus GG (LGG), which has been reported to demonstrate anti-caries activities, was also included. Two L. salivarius stains were thus identified and their antagonistic mechanisms against S. mutans biofilm formation were analyzed.
Bacterial strains, media, and growth conditions
S. mutans type strain ATCC 25175 (serotype c) isolated from carious dentine and the commercial probiotic strain LGG were obtained from the Bioresource Collection and Research Center (BCRC). The 64 L. salivarius strains were isolated from various sources, including fermented vegetables (5 strains), human saliva (49 strains), and breast milk (10 strains). Identification of bacterial species was performed according to 16S rDNA sequencing, as previously described (Chao and others 2013; Chao and others 2008), and both K35 and K43 were isolated form human saliva. Lactobacilli and S. mutans were respectively cultured in deMan, Rogosa, and Sharpe (MRS) and brain heart infusion (BHI) medium (BD Difco) at 37°C for 20 h under anaerobic atmospheric conditions (10% CO2, 10% H2, 80% N2). BHI agar plates containing 1.75 g/ml polymyxin B, 0.3 U/ml bacitracin, and 0.005% crystal violet were used for the selective isolation of S. mutans.
Stationary-phase S. mutans and lactobacilli were respectively adjusted to OD600 of approximately 1. Then, both lactobacilli and S. mutans were diluted 100-fold in BHI medium supplemented with 0.2% sucrose, mixed well, and then inoculated in polystyrene cell culture plates (Corning Inc.). Lactobacilli were also cultivated alone to establish mono-species biofilms. Biofilm cultures were grown in an anaerobic incubator at 37°C for 24 h, and the final pH of the biofilm cultures were measured with a pH meter (Ai-On Industrial Corp.). After the incubation, supernatants were removed from the cultures, and the plates were washed once with deionized water. To determine biofilm mass, wells were stained with 0.1% safranin for 30 min, washed three times with deionized water, and air-dried. The dye was solubilized in 33% acetic acid, and the absorbance at 492 nm was determined using a microtiter plate absorbance reader (Tecan). For the transwell-based biofilm assay, a 24-well Millicell plate with a PET membrane insert (Millipore) was used. The pore size of the membrane was 0.4 μm, since bacteria were unable to traverse filters with pore size less than 0.4 µm (Wu and others 2001). Lactobacilli were inoculated in the upper insert, while S. mutans was inoculated in the bottom.
Scanning electron microscopy (SEM)
Both mono-species and dual-species biofilms were grown on glass slides placed in a 24-well plate for SEM observation. Slides were gently washed with phosphate buffered saline (PBS) once, fixed with PBS containing 2.5% glutaraldehyde and 4% paraformaldehyde, dehydrated in graded ethanol solutions, dried in liquid CO2, and finally sputter-coated with gold before SEM observation (JSM-7600F, JEOL).
Growth inhibition on agar plate
The inhibitory effect of lactobacilli against S. mutans was assessed using an agar diffusion method with minor modifications (Teanpaisan and others 2011). Briefly, BHI agar plates were seeded with overnight-grown S. mutans using an aseptic cotton swab. After S. mutans was swabbed, 3 μl of stationary-phase lactobacilli cultures were dropped onto the BHI plates. The plates were then anaerobically incubated at 37°C for 24 h to generate an inhibitory zone.
Quantitative real-time PCR (qRT-PCR)
Bacterial RNA was prepared and analyzed as previously described (Lin and others 2013). In brief, total RNA was isolated from the lactobacilli S. mutans co-culture using RNeasy midi-columns (QIAGEN) according to the manufacturer’s instructions. The RNA was DNase treated with RNase-free DNase I (MoBioPlus) to eliminate DNA contamination. A 100 ng quantity of RNA was reverse-transcribed with the Transcriptor First Strand cDNA Synthesis Kit (Roche) using random primers. qRT-PCR was performed with a Roche Light- Cycler® 1.5 Instrument using Light Cycler TaqMan Master (Roche) to detect gtfB (primers gtfB-F: 5’-CTT CTG ATC GCG TGG TTG T-3’ and gtfB-R: 5’-AAG GTC GGT AAG CTT GGT TCT-3’; probe: 60), gtfC (primers gtfC-F: 5’-GGC TAA TTC CAA CTA CCG TAT CTT-3’ and gtfC-R: 5’-GGT AAG TGG GGC CTT AGC TC-3’; probe: 82), gtfD (primers gtfD-F: 5’- cca ata ttc cga cag cct atg-3’ and gtfD-R: 5’-tca cca taa taa aga cgt gta att gaa-3’; probe: 56), and 23S rRNA (primers 23S-F: 5’-GCG ATC AGC TGT ATA CCT TGG-3’ and 23S-R: 5’-GAT CGA ACC GCT GAC CTC-3’; probe: 67). Primers and probes were designed for selected target sequences using the Universal Probe Library Assay Design Center (Roche-applied science). Data were analyzed using the real time PCR software of the Roche LightCycler® 1.5 Instrument. Relative gene expression was quantified using the comparative threshold cycle 2-CT method with 23S rRNA of S. mutans as the endogenous reference.
Analytical Profile Index (API) Typing
Sugar utilization of K35 and K43 were investigated using the API 50CHL system (bioMerieux, France) according to the manufacturer’s instructions. The biochemical profile for the strain was identified using the apiwebTM identification software with database (V5.1).
Peripheral blood mononuclear cells assay
Isolation of human peripheral blood mononuclear cells (hPBMCs) from healthy volunteers and the treatment of lactobacilli were performed as previously described (Liu and others 2011) with slight modifications. The hPBMCs were collected and resuspended in RPMI 1640 medium containing 10% FBS, 100 IU/ml penicillin, 0.1 mg/ml streptomycin, 0.25 μg/ml amphotericin, and 1% L-glutamate. hPBMCs (2×105 cells/well) were seeded in 96-well tissue culture plates and treated with 2×106 CFU of heat-killed lactobacilli (hPBMCs: lactobacilli ratio of 1:10) at 37°C and 5% CO2 for 48 h. The use of phytohemagglutinin (PHA, 2 μg/ml), E. coli lipopolysaccharide (LPS, 1 μg/ml), or medium only were performed as control experiments. The cultured plates were centrifuged (1000 g, 10 min, 4°C), and the supernatants were collected for cytokine (IFN-γ and IL-10) determination. The concentrations of IFN-γ and IL-10 were determined using an ELISA procedure according to the manufacturer’s instructions (R&D Systems).
Repetitive sequence-based PCR genomic fingerprinting
Repetitive sequence-based (rep)-PCR analysis is considered a convenient tool for the discrimination of bacterial strains. To discriminate the two L. salivarius strains, three rep-PCR genomic fingerprinting methods, including enterobacterial repetitive intergenic consensus (ERIC)-PCR, BOX-PCR, and (GTG)5-PCR, were used. Purification of the bacterial genomic DNA and PCR were performed as previously described (Chao and others 2008; Mohapatra and others 2007). PCR primers: ERIC1R (5’- ATG TAA GCT CCT GGG GAT TCA C-3’) as well as ERIC2 (5’-AAG TAA GTG ACT GGG GTG AGC G-3’), BOXA1R (5’-CTA CGG CAA GGC GAC GCT GAC G-3’), and (GTG)5 (5’-GTG GTG GTG GTG GTG-3’) were used for the PCR reactions. The PCR products were separated by electrophoresis on a 1% agarose gel.
Experimental results were analyzed for statistical significance with Prism5 software (GraphPad) using one-way ANOVA followed by Bonferroni post hoc correction. A P -value of < 0.01 was used as significant in all cases. The results of the biofilm formation assay, qRT-PCR, and cytokine production assay were confirmed with at least three independent experiments. Each sample was assayed in triplicate and the mean activity and standard deviation are presented.
For isolation of normal human peripheral blood from healthy volunteers, the procedure and the respective consent documents were approved by the Ethics Committee of the National Yang-Ming University, Taipei, Taiwan. All healthy volunteers provided written informed consent.
Co-culture with lactobacilli reduces S. mutans biofilm formation
A total number of 64 L. salivarius strains isolated from fermented vegetables, human saliva, or breast milk, were analyzed to determine if these strains possess potential inhibitory effects on S. mutans biofilm formation. A biofilm formation assay was performed as described with LGG was used as a control since this probiotic strain has been described to have anti-caries activity (Nase and others 2001). We found that S. mutans co-cultured with most of the L. salivarius strains, respectively, resulted in 0-30% reduction of total biofilm mass as compared with that of the S. mutans only (data not shown). However, two particular strains, K35 and K43, which did not form strong biofilms in a mono-species model (Fig. 1A), appeared to cause evident reductions (P<0.01) in biofilm formation, approximately 64.3% and 69.6%, respectively. The control species, LGG, caused an approximately 44% reduction of the biofilm mass (Fig. 1B). Besides, LGG presented relatively strong biofilm forming activity in a mono-species model (Fig. 1A). Heat treatment (80°C for 20 min) of K35 or K43 cultures before the inoculation in the biofilm co-culture resulted in loss of the inhibitory activities. Nevertheless, heat-killed LGG still caused a slight reduction (~24%) in biofilm formation as compared to that with S. mutans only (Fig. 1C).
Co-culture with lactobacilli influence the morphology of S. mutans biofilms
Biofilm formation of the co-cultures, as observed by SEM, is shown in Fig. 2. S. mutans appeared to form a compact and island-like biofilm covered by large amounts of slime or network-like structures, suggested to be EPS. S. mutans co-cultured with K35 or K43 resulted in visibly fewer amounts of bacteria and smaller microcolonies attached to the surface. Moreover, amount of EPS appeared to be decreased in the co-culture with the K35 or K43 strains. LGG, the long bacilli observed in the photo, was found to occupy most of the exterior of the biofilm, and network-like EPS was still observed in the co-culture. Additionally, S. mutans co-cultured with heat-killed LGG also resulted in an island-like biofilm covered with EPS; however, the biofilms appeared to have larger spaces between the bacterial cells, which were not as compact as that of the S. mutans only.
L. salivarius inhibits the growth and expression of gtf of S. mutans
To determine how lactobacilli influence S. mutans biofilm formation, S. mutans growth in the co-culture was analyzed using a spot assay. Biofilm from the lactobacilli and S. mutans co-culture was scraped and rigorously mixed with the planktonic fraction. Subsequently, the bacterial suspension was serially diluted with PBS, and a 5-l aliquot was spotted onto the S. mutans selection plate, which inhibited the growth of lactobacilli. As shown in Fig. 3A, co-culture with the K35 or K43 strains apparently reduced the growth of S. mutans. Compared with K35 and K43, LGG appeared to have a weaker inhibition on the growth of S. mutans, while heat-killed LGG did not exhibit bactericidal activity. To further verify the bactericidal activity, we analyzed the inhibitory activity of lactobacilli against S. mutans growth on agar plates. As shown in Fig. 3B, clear zones with similar diameters, ranging from 18 to 19 mm, surrounding the macro-colonies of strains K35, K43, LGG, and G01, which did not influence biofilm formation in the co-culture experiment, could be observed. This result confirmed that these lactobacilli possess bactericidal activities to S. mutans, possibly through secreted factors.
On the other hand, we used qRT-PCR to analyze the expression of S. mutans gtfB, gtfC, and gtfD genes, which encode glucosyltransferases and play a crucial role in biofilm formation. As shown in Fig. 4, co-culture with the K35 or K43 strains significantly decreased the mRNA levels of gtfB, gtfC, and gtfD of S. mutans, suggesting reduced EPS production. Interestingly, we also found that LGG evidently increased the expression of gtfB and gtfC and did not regulate the expression of gtfD. In addition, heat-killed LGG did not cause an apparent effect. Our results suggest that L. salivarius K35 and K43 strains not only inhibit the growth, but also decrease the expression of glucosyltransferases-encoding genes of S. mutans to reduce its biofilm formation.
L. salivarius reduces S. mutans biofilm formation in a contact-independent manner
To determine if the inhibitory activity of lactobacilli on S. mutans biofilm formation required direct cell-cell interactions, a transwell-based biofilm formation assay was performed. Lactobacilli were inoculated in the upper insert, and they could not enter the bottom of the culture well. As shown in Fig. 5, we found that the presence of lactobacilli in the upper insert apparently reduced the S. mutans biofilm formation in the bottom. Compared with the biofilm formed by only S. mutans, biofilm formation was reduced by approximately 52.5% and 46.3%, when K35 or K43 strains, respectively, were added. LGG caused a modest reduction (~30.2%), while heat-killed LGG did not exhibit an evident influence. Since lactobacilli in the upper insert could not pass through the membrane, their modulation of S. mutans biofilm formation is suggested to be dependent on secretory factors, such as bacteriocins, hydrogen peroxide, and lactic acid, which remain to be investigated. Additionally, the measurement of the final pH values of the biofilm co-cultures revealed that, compared to that of S. mutans only (pH ~4.97), the presence of K35 or K43 resulted in decreased pH values to approximately 4.84 and 4.86, respectively; LGG also caused a slight reduction (pH ~4.92), while heat-killed LGG did not have an evident effect (pH ~4.96) (Fig. 5). The change in pH values seemed obscure, implying the involvement of other factors in this phenomenon, which have yet to be elucidated. Our results showed that K35, K43, and LGG could inhibit S. mutans biofilm formation in a contact-independent manner and suggested that heat-killed LGG reduced biofilm formation via physical interference.