Background
Probiotic bacteria, defined as "live microorganisms which when administered in adequate amounts confer a health benefit on the host" (FAO/WHO 2001), are suggested to play a role in the maintenance of oral health
1
2
. Such health promoting bacteria are added to different commercial dairy products such as milk, cheese and yogurt as well as chewing gums and fruit drinks. Possible actions of probiotic bacteria in the oral environment are competition of binding sites, production of antimicrobial substances and activation and regulation of the immune response
3
. Bacterial antagonism may occur when growth of one bacterial species is hampered by components produced by another species. Lactic acid bacteria produce antimicrobial components
4
5
and some have the ability to produce hydrogen peroxide (H2O2) that can be toxic to organisms producing little or no H2O2-scavenging enzymes.
Molecular analyses of the oral microbiota in preschool children have shown that Streptococcus mutans is significantly associated with early childhood caries
6
. Candida albicans is a persistent member of the oral microbiota in children with caries
7
with a substantial growth response to sucrose exposure
8
. C. albicans produce organic acids like pyruvate and acetate and are considered to have a significant contribution to caries pathogenisis
9
. Lactobacilli play a significant role in the oral ecosystem and can be linked with oral disease as well as oral health
10
. Since the discovery by Polonskaya
11
that L. acidophilus inhibits growth of certain streptococci in vitro, clinical studies have confirmed that probiotic lactobacilli can reduce the counts of salivary mutans streptococci after ingestion of L. rhamnosus GG
12
13
and L. reuteri
14
15
16
. Furthermore, naturally occurring Lactobacillus species, including L. paracasei, L. plantarum and L. rhamnosus, may inhibit growth of laboratory strains of mutans streptococci as well as subject's autologous mutans streptococci in vitro
17
. Hatakka et al
18
found that a cheese containing a mixture of probiotic bacteria decreased the salivary count of C. albicans in a randomized controlled trial among elderly.
The aim of the present study was to investigate the ability of a selection of lactobacilli strains, used in commercially available probiotic products, to inhibit growth of mutans streptococci and C. albicans in vitro. The null hypothesis tested was that none of the lactobacilli strains would differ significantly from the other.
Methods
Lactobacilli strains and cultivation
Eight strains of probiotic lactobacilli (L. plantarum 299v, L. plantarum 931, L. rhamnosus GG ATCC 53103, L. rhamnosus LB21, L. paracasei F19 and L. reuteri PTA 5289, L. reuteri ATCC 55730 and L. acidophilus La5) used in different probiotic products were selected (Table 1). The bacteria were provided by the different producers in pure forms (frozen suspensions or lyophilized) except for L. acidophilus La5 that was isolated from A-fil® (Arla Ltd, Stockholm, Sweden). The strains were characterized by the API 50 CH system (BioMérieux® SA, Marcy-l 'Etoile, France) to confirm their identity. The bacteria were initially cultured for 16-20 h on MRS agar (de Man, Rogosa, Sharpe, Oxid, Hampshire, England). A distinct colony of each bacterium was then transferred to 4.5 ml MRS broth for further 16-20 h of incubation.
<p>Table 1</p>Probiotic lactobacilli strains, producer and products used in the present experiments.
Strain
Producer
Product
L. plantarum 299v
Probi AB, Lund, Sweden
Fruit drink
L. plantarum 931
Essum, Umeå, Sweden
Fruit drink
L. rhamnosus GG ATCC 53103
Valio Ltd, Helsinki, Finland
Yogurt
L. rhamnosus LB21
Essum, Umeå, Sweden
Yogurt
L. paracasei F19
Arla Ltd, Stockholm, Sweden
Yogurt, Porridge, Gruel
L. reuteri PTA 5289
BioGaia, Stockholm, Sweden
Chewing gum, Tablets
L. reuteri ATCC 55730
BioGaia, Stockholm, Sweden
Chewing gum, Gruel, Drops, Tablets
L. acidophilus La5
Arla Ltd, Stockholm, Sweden
Fermented milk
Mutans streptococci, candida strains and cultivation
Five strains of mutans streptococci (MS) including both laboratory reference strains (S. mutans NCTC 10449, S. mutans Ingbritt, and S. sobrinus OMZ176) and clinical isolates (S. mutans P1:27 and S. mutans P2:29) were cultured on blood agar plates (Columbia Blood Agar Base, Alpha BioScience, Baltimore, USA) supplemented with 5% horse blood during 16-20 hours and on the following day, pure colonies of each bacterium were transferred to 2 ml Todd-Hewitt broth (Oxid, Hampshire, England) and incubated for another 16-20 hours. Five strains of Candida albicans including both reference strains (C. albicans ATCC 28366, C. albicans ATCC 10231) and clinical isolates (C. albicans 1957, C. albicans 3339 and C. albicans GDM8) were cultured on Difco™ Sabourad Maltose Agar (Becton, Dickinson and Complany, Sparks, USA) over night and the following day, pure colonies of each yeast were transferred to 2 ml Difco™ Sabourad Maltose broth and incubated for 16-20 hours. The C. albicans strains were characterized by the API Candida (BioMérieux® SA, Marcy-l'Etoile, France). Culturing of lactobacilli and mutans streptococci were performed in anaerobic atmosphere (10% H2, 5% CO2 and 85% N2) in an anaerobe chamber at 37°C, while culturing of C. albicans was performed under aerobic conditions at 37°C.
Agar overlay interference tests
The broth cultures of lactobacilli were serial diluted in tenfold steps. The optical density (OD) was measured at 630 nm using a spectrophotometer at dilution 10-1 (Ultrospec 100 pro, Visible Spectrophotometer, Biochrom Ltd. Cambridge, England). Undiluted suspension and cell suspensions corresponding to approximately 109, 107, 105, and 103 CFU/ml were used in the inhibition experiments. One ml of each lactobacilli suspension was added into 24 ml molten sterile MRS agar (ca 45°C) and plates were casted. When the agar was set, the plates were incubated at 37°C over night in anaerobic atmosphere. The next day, a second agar layer of 23 ml molted M17 agar supplemented with 10% sterile filtered lactose (May and Baker, Dagenham, England) was casted on top of the MRS agar with grown lactobacilli. The plates were allowed to dry for 3 hours at room temperature. Broth cultures of MS grown for 16-20 hours in Todd-Hewitt broth were diluted in the same medium and the OD was measured at 500 nm and adjusted to 0.2. The suspensions of MS were stamped on the plates with Steer's replicator (CMI-Promex ICN, Pedricktown, USA). The plates were left at room temperature for 1 hour to dry and were subsequently incubated over night at 37°C in the anaerobe chamber. To test the susceptibility of C. albicans, 24 ml Difco™ Sabourad Maltose Agar (SAB) was casted on top of the MRS agar with grown lactobacilli. Broth cultures of C. albicans grown for 16-20 h in Difco™ Sabourad Maltose Broth were diluted in the same medium and the OD was measured at 500 nm and adjusted to 0.2. The suspensions were then stamped on the plates in the same way as MS (see above). The plates inoculated with C. albicans were incubated in air at 37°C. As controls, the MS and C. albicans strains were stamped onto agar plates without lactobacilli within the first agar layer.
All assays were made in duplicate and each experiment was repeated on three different occasions. The results of the agar overlay tests were categorized as follows according to Simark-Mattsson et al.
17
: Score 0 = complete inhibition (no visible colonies), Score 1 = slight inhibition (at least one visible colony but definitely smaller amounts then in the control plate), and Score 2 = no inhibition (the same growth as on the control plate). The evaluation of the plates was performed by two independent observers, in case of disagreement a consensus was reached after discussion.
pH-measurements
As an estimation of the lactobacilli acid production, the surface pH of the MRS plates with M17 and SAB agar was determined before and after the final incubation of each lactobacilli strain but without mutans streptococci or candida. A pH-meter (Seven Easy pH, Mettler-Toledoo GmbH, Schwerzenbach, Switzerland) equipped with a flat electrode was used and the mean of two measurements was calculated.
Statistical method
The data were processed with the SPSS software (version 17.0, Chicago Ill, USA) and subjected to chi-square tests. A p-value < 0.05 was considered as statistically significant.
Results
Characterization of lactobacilli and Candida albicans
Both L. plantarum strains as well as the L. acidophilus La5, L. paracasei F19 strains were identified by the API 50 CH system without difficulty. The two L. reuteri strains gave a biochemical profile interpreted as L. fermentum. None of the L. rhamnosus strains were fully identified by the system. All five candida strains were completely identified as C. albicans by the API Candida assay.
Growth inhibition of mutans streptococci
The results of the growth inhibition assay are summarized in Table 2. At concentrations ranging from 109 to 105 CFU/ml, all lactobacilli strains inhibited the growth of the MS strains completely with the exception of L. acidophilus La5 that executed only a slight inhibition of some strains at concentrations corresponding to 107 and 105 CFU/ml. L. acidophilus La5 had a statistically significantly weaker inhibition capacity in comparison with the other probiotic strains (p < 0.05). At the lowest cell concentration (103 CFU/ml), only L. plantarum 299v and L. plantarum 931 displayed a total growth inhibition while a slight inhibition was seen for all five MS strains by L. rhamnosus LB21, L. paracasei F19, L. reuteri PTA 5289 and L. reuteri ATCC 55730. L. rhamnosus GG ATCC 53103 diluted to 103 CFU/ml inhibited the growth slightly for three of the five MS strains (S. mutans NCTC 10449, S. sobrinus OMZ176 and S. mutans P1:27) while low concentrations of L. acidophilus La5 did not affect MS growth. In general, no significant differences in growth inhibition were observed between the reference strains and the oral isolates.
<p>Table 2</p>Growth inhibition of mutans streptococci strains by selected probiotic lactobacilli at different cell concentrations.
Mutans streptococci (inhibition score)
Lactobacilli
CFU/ml
Ingbritt
NCTC
10449
OMZ176
P1:27
P2:29
L. plantarum 299v
103-109
0
0
0
0
0
L. plantarum
931
103-109
0
0
0
0
0
L. rhamnosus GG
105-109
0
0
0
0
0
ATCC 53103
103
0
1
1
1
0
L. rhamnosus
105-109
0
0
0
0
0
LB21
103
1
1
1
1
1
L. paracasei
105-109
0
0
0
0
0
F19
103
1
1
1
1
1
L. reuteri
105-109
0
0
0
0
0
PTA5289
103
1
1
1
1
1
L. reuteri
105-109
0
0
0
0
0
ATCC 55730
103
1
1
1
1
1
L. acidophilus
109
0
0
0
0
0
La5 *
107
0
1
1
0
1
105
1
1
1
1
1
103
2
2
2
2
2
Inhibition scores; 0 = total inhibition; 1 = slight inhibition; 2 = no inhibition.
* p < 0.05, chi-square test.
Growth inhibition of Candida albicans
All the tested lactobacilli strains reduced candida growth but the effect was generally weaker than for mutans streptococci. Since an identical inhibition pattern was displayed for all candida strains, the results of the assays for only one of them are displayed in Table 3. At concentrations 109 and 107 CFU/ml, all lactobacilli except L. acidophilus La5 and L. reuteri PTA 5289 inhibited the five candida strains completely. At 105 CFU/ml, L. rhamnosus LB21, L. rhamnosus GG ATCC 53103, L. paracasei F19 and L. reuteri PTA 5289 displayed a slight inhibition, L. acidophilus La5 showed no inhibition at all while L. plantarum 299v, L. plantarum 931 and L. reuteri ATCC 55730 executed a total inhibition. At the lowest cell concentration, no inhibition was recorded except for the two L. plantarum strains.
<p>Table 3</p>Growth inhibition of Candida albicans ATCC 28366 by selected probiotic lactobacilli at different cell concentrations.
Strain
Cell concentration (CFU/ml)
10
9
10
7
10
5
10
3
L. plantarum 299v
0
0
0
1
L. plantarum 931
0
0
0
1
L. rhamnosus GG ATCC 53103
0
0
1
2
L. rhamnosus LB21
0
0
1
2
L. paracasei F19
0
0
1
2
L. reuteri PTA 5289
1
1
1
2
L. reuteri ATCC 55730
0
0
0
2
L. acidophilus La5
1
1
2
2
Inhibition scores; 0 = total inhibition; 1 = slight inhibition; 2 = no inhibition.
pH-measurements
The initial pH on the surface of the M17 and SAB agar plates were 6.8 and 5.8, respectively. After incubation, the pH decreased in all plates inoculated with lactobacilli and varied between 3.7 and 5.3 in the plates containing 103 CFU/ml (Table 4). L. acidophilus La5 displayed the highest pH values while the lowest values were recorded for the L. plantarum strains.
<p>Table 4</p>Final pH on the surface of M17 and SAB agar plates after incubation with selected probiotic lactobacilli (103 CFU/ml).
Strain
M17 agar
SAB agar
L. plantarum 299v
4.4
3.7
L. plantarum 931
4.4
3.7
L. rhamnosus GG ATCC 53103
4.6
4.0
L. rhamnosus LB21
4.5
4.1
L. paracasei F19
4.8
4.4
L. reuteri PTA 5289
5.1
4.0
L. reuteri ATCC 55730
5.1
3.9
L. acidophilus La5
5.2
5.3
Discussion
The predominant strategy for caries prevention relies on influencing the re- and demineralisation processes, mainly by using fluorides but also other factors involved in the caries process may be targeted
19
. One alternative could be to promote colonization of caries inhibiting bacteria and in this aspect, probiotic bacteria constitute a novel concept that needs further exploration. This in vitro study was performed to compare the effect of eight commercial probiotic lactobacilli strains on growth inhibition of mutans streptococci and C. albicans. We selected these particular lactobacilli strains because they are found in dairy products, fruit drinks, drops, gruels, chewing gums and tablets on the market. The test panel of mutans streptococci and yeast contained both reference strains and clinical isolates. Candida albicans is the most commonly isolated candida species in the oral cavity and constitute up to 80% of the clinical isolates
20
. Furthermore, candida has been ascribed a significant predictive value for dental caries in children
21
. The majority of the lactobacilli strains were fully identified with biochemical analysis using the API 50 CH system while a satisfactory identification of the L. reuteri and L. rhamnosus strains would require molecular genetic methods
22
.
The agar overlay method was considered suitable for the research question since the method can test the inhibitory effect on multiple strains on a single plate. The method has earlier been used to demonstrate the in vitro growth inhibition of different isolates of lactobacilli on pathogenic bacteria in the female urogenital region
23
and to study the inhibitory effect of alpha-hemolytic streptococci on the otitis media pathogens Streptococcus pneumoniae, Haemophilus influenzae and Moraxella catarrhalis
24
. Kõll et al
25
used the deferred antagonism method to test the inhibitory capacity against mutans streptococci and C. albicans. In that method the lactobacilli are stab-inoculated on the surface of the bottom agar and pathogen suspensions are poured over the macrocolonies of lactobacilli. After incubation the width of the inhibition zones are measured. The main advantage with the deferred antagonism method is the more exact outcome while the agar overlay method enables assessment of different cell concentrations. Thus, the methods may be regarded as complements to each other.
With the exception of L. acidophilus, the present probiotic strains displayed strong inhibitory capacities against both the reference strains and the oral mutans streptococci isolates and these results were mainly in agreement with the recent observations of Simark-Mattsson and coworkers
17
. The findings concerning the candida strains were however novel and showed an even larger variation between the lactobacilli strains and only three of the eight tested strains exhibited strong inhibitory capacity. These results were in disagreement to Kõll et al
25
who failed to demonstrate inhibition of Candida albicans 048 (wild-type). This could due to the method used in the experiments or the fact that the tested candida strain were more resistant then our reference strains and clinical isolates. The production of bacteriocins can be different in different systems and the ability to diffuse through the agar. Since significant growth inhibiting differences between the probiotic lactobacilli were noted, the null hypothesis was rejected. Similar strain-dependant differences have previously been observed concerning the metabolic capacity to form acids from dietary sugars that differs significantly between various probiotic strains
26
27
.
The final pH in the medium has been suggested to be an important factor for growth inhibition, either directly or due to the production of bacteriocins at low pH
28
. We found a similar tendency in the present experiments; the lactobacilli strains that caused the lowest pH after incubation were also most effective in inhibiting MS and candida growth. The weakest inhibition of both mutans streptococci and C.albicans was displayed by L acidophilus La5 which also had the weakest acid production in both systems. In the agar overlay method the lactobacilli is grown inside the MRS-agar and it is possible that this particular system not is optimal for L. acidophilus La5. The inhibitory potential may be different, for better or for worse, in other test system or in vivo. However, also some bacteria with fairly weak acid production, such as L. reuteri ATCC 55730, proved to be effective against both mutans streptococci and candida. This indicates that other inhibitory substances also may be involved in the process with H2O2 being among the primary metabolites with inhibitory capacity against microbial pathogens
29
. The importance of H2O2 production for health can be illustrated in bacterial vaginosis, since lack of lactobacilli with such production seems to be associated with an altered composition of the vaginal flora resulting in a massive anaerobic overgrowth
30
. Tano et al.
31
concluded that the inhibitory effect of alpha-hemolytic streptococci on otitis media pathogens most likely was due to their H2O2 production. The antimicrobial glycerol derivative reuterin is another example of a growth inhibitory substance produced by L. reuteri
32
that may be involved in the reduction of mutans streptococci in saliva
14
15
as well as beneficial influences on gingivitis
33
and inflammatory mediators in gingival crevicular fluid
34
. Although any conclusion from in vitro studies must be drawn with caution, our findings provide some further support to the hypothesis that probiotic lactobacilli may influence the homeostasis and microbial profile in the oral cavity.
Authors' contributions
PH contributed with the laboratory work, the acquisition, analysis and interpretation of data and took part in drafting of the manuscript. MH contributed with the design, laboratory work, analysis and interpretation of data and took part in drafting of the manuscript. ST contributed with drafting of the manuscript and revised it critically. CSB contributed with analysis and interpretation of data, drafting of the manuscript and revised it critically. All authors read and approved the final manuscript.