The word probiotics comes from the Greek words “pro” and “bios”, meaning ‘for life’. The term probiotics was first mentioned by Lilley and Stillwell in 1965 and was defined as the opposite of antibiotics [1]. “Organisms and substances which contribute to intestinal microbial balance” was a later definition from Parker [2]. Although this definition is still used today, this interpretation included also antibiotics, which is not commonly used today. An advancement was the interpretation from Fuller in 1989, because the importance of living cells was emphasised and the word “substances” was eliminated to avoid confusion. “A live microbial feed supplement which beneficially affects the host animal by improving its intestinal microbial balance”[3]. Ever since, the definition has been reworked several times and it was for instance broadened by Vrese and Schrezenmeir, who included other target location than the gut [4]. An often cited definition is the one in the WHO/FAO report from 2006, wherein probiotics were defined as: “Live microorganisms which when administered in adequate amounts confer a health benefit on the host” [5].

Probiotics are mainly used in dairy products, but they are increasingly becoming available as dietary supplements and pharmaceutical preparations. A large variety of microorganisms are classified as probiotics at present. Holzapfel [6] listed strains used as probiotics. Lactic acid bacteria are of most importance in food and nutrition, especially lactobacilli and the genus Bifidobacterium, whereas non-lactic acid bacteria are mainly used in pharmaceutical preparations. Different Lactobacillus species are also found in the oral cavity and in the saliva, the most common of which are: L. fermentum, L. rhamnosus, L. salivarius, L. casei, L. acidophilus and L. plantarum [7-9]. Three of these species are used in dairy products, but yet it is not clear, if they are detected due to frequent consumption of dairy products leading to temporary colonisation, or if the oral environment is the natural habitat [10].

If probiotics are taken orally, they are considered to be safe. A theoretical concern is that these viable bacteria move into the bloodstream and cause systemic infections. Lactobacillus rhamnosus GG (LGG) became available in Finland in 1990. Salimen et al. analysed blood samples in the period between 1990-2000 and did not find an increase related to the increasing probiotic use of LGG-containing commercial dairy products [11]. One case has been published that was caused by probiotic lactobacilli. A L. rhamnosus GG strain was isolated from a liver abscess in 19994. Some cases of Endocarditis [12] or L. rhamnosus GG leading to bacteraemia [13] are described, but it is estimated, that only 0.05%-0.4% of these cases are due to lactobacilli and bifidobacterial [14]. These lactobacilli and bifidobacteria are thought to be part of the commensal microbiota, instead of administered probiotics. Most cases of probiotic bacteraemia responded well to appropriate antibiotic therapy. Common risk factors to Lactobacillus bacteraemia are immunosuppression, prior hospitalisation and previous antibiotic treatment [15]. Even in HIV-infected patients, the use of the probiotic L. rhamnosus GG was well tolerated and no adverse effects, like bacteraemic outbreaks, were observed10. The consumption of 1 × 10¹⁰ CFU/day was safe and well tolerated in a HIV-positive population [16]. Antibiotic resistance of lactobacilli can be another safety issue. Probiotics should not carry transmissible antibiotic resistance genes [17]. Lactobacilli are naturally resistant to some antibiotics, as shown by Charteris [18], but this resistance is usually not of a transmissible type. Non-transmissible antibiotic resistance is usually of no safety concern. Lactobacillus plantarum CCUG 43738 displayed atypical phenotypic resistance to tetracycline and minocycline, from a plasmid-located tet(S) gene. This acquired antibiotic resistance is undesirable for potential probiotics [19]. There are additional reports of acquired antibiotic resistance [20-21]. The transferal of antimicrobial resistance genes between probiotics and closely related opportunistic bacteria may take place [10]. More studies are needed, because the development of resistance to antibiotic drugs is a global issue. The US Food and Drug Administration (FDA) designation ‘Generally Recognized as Safe’ (GRAS) has been applied to several Lactobacillus strains [22] being a strong indication for safety in human consumption.

Strain selection

The selection of appropriate probiotic strains is a key issue. In contrast to the often used dietary lactobacilli, indigenous bacteria offer an advantage, because they are well adjusted to the human oral ecology. Consequently, the use of orally derived probiotics is recommended. Furthermore, the binding of LAB to mucus might be necessary to show positive effects, if administered to the oral cavity. Carbohydrate-protein interactions probably play a key role in the adhesion of these proteins to mucin-bound oligosaccharides [23]. Presently, most reports on mucus binding of probiotics are in the gastrointestinal field, because this is the main field of application. Several adhesion promoting proteins in Lactobacillus spp. are described and summarised by Van Tassell et al. [23] and by Vélez et al. [24]. The most studied mucus-targeting bacterial adhesin is mucus-binding protein (MUB), which is produced by L. reuteri [25]. An adhesion-promoting protein that mediated L. fermentum 104R binding to both small intestinal porcine mucus and porcine gastric mucin is described by Royas et al. [26]. Moreover, some Lactobacillus strains adhere to Caco-2 cell cultures, which was observed by Tuomola and Salimen [27]. A mucus adhesion promoting protein, MapA, mediates the adhesion of L. reuteri to human intestinal epithelial cells [28]. In L. plantarum WCFS1 a lectin-like mannose-specific adhesin that interacts with the host intestinal tract has also been described [29]. It is expected, that bacteria, intended as oral probiotics, should adhere to and colonise surfaces of the oral cavity [30]. Yli-Knuuttila et al. investigated whether L. rhamnosus GG could be detected in the oral cavity after discontinuation of administration of a probiotic product. This strain could only temporally be detected, but did not colonise the oral cavity. The authors concluded, that colonisation with this strain is improbable, but possible in some cases30. In a clinical study, the application of a chewable tablet of L. reuteri ATCC 55730 (108 CFU/tablet) for 2 weeks did not lead to permanent colonisation of the oral cavity [31]. The consumption of probiotic yoghurt-containing lactobacilli did also not lead to their installation in the oral cavity [32]. In contrast, salivary Lactobacillus counts increased after consumption of a product containing seven different Lactobacillus strains [33]. L. plantarum 299v (Lp299v) and L. paracasei showed survival in saliva for 24 h. The L. paracasei strain bound well to a surface mimicking human dental enamel, because over 20% of the added bacteria were bound but, for Lp299v the binding was poor with below 5% [34]. Lactobacillus GG, incorporated into yoghurt, was consumed twice a day for 7 d. After discontinuation of consumption, the bacterium was recovered in saliva samples, and after 2 weeks the strain was still present in 8 out of 9 subjects [35].

A final aspect to consider is lactic acid production by lactobacilli, because this process has been implicated in enamel demineralisation and dental caries in the mouth. The ability to produce acid varies significantly between strains and if they belong to the homofermentative or heterofermentative group [36]. Some strains, such as L. plantarum 299v can not only ferment carbohydrates (e.g. glucose and fructose) but also sugar alcohols (e.g. mannitol or sorbitol) leading to a decrease in pH in-vitro. In contrast Keller and Twetman found no evidence for an increased plaque acidity for L. plantarum 299v with fructose or xylitol [37]. Strains with low metabolic activity are favourable, for example L. paracasei [38].

Periodontal disease

Three main factors considered for plaque-related periodontal inflammation are: a susceptible host, which is hard to address; second the presence of pathogenic species and the reduction or even absence of ‘beneficial bacteria’. Today, periodontal therapies aim to reduce the bacterial threat and are based on mechanical subgingival debridement plus improvement of oral hygiene [39]. The subgingivalmicrobiota then shifts towards a less pathogenic composition, meaning a higher proportion of Gram-positive aerobic species and fewer periodontopathogens [40]. Unfortunately, this shift is only temporary. The use of antibiotics or antiseptics for a certain time does not improve the long-term effect of periodontal therapy [41]. The use of probiotics in this field offers the possibility to restore the reduced numbers of beneficial bacteria to prevent and treat plaque-related periodontal disease. On the one hand the inhibition of specific pathogens is a possible mechanism and on the other hand the host response could be affected by probiotics.

To date, the following effects have been shown for probiotics in the field of periodontal disease. L. salivarius tablets reduced 5 periodontopathic bacteria in subgingival plaque after 4 weeks and the levels tended to be lower up to 8 weeks compared with placebo [42]. In a clinical trial with L. reuteri containing tablets the number of periodontal pathogens in the subgingivalmicrobiota was reduced, but no significant clinical impact could be shown [43]. A Lactobacillusmicrobiota inhibited growth of S. mutans, P. gingivalis and P. intermedia [44]. Furthermore, antimicrobial effects against P. gingivalis have been described [45-46]. Co-aggregation activity of probiotic strains with oral pathogens (Porphyromonasendodontalis, T. forsythia, Eubacteriumsaphenum, Filifactoralocis or P. gingivalis) can be another mode of action [44,45,47]. Competition with peridontopathogens for the uptake of nutrients could also improve oral health [48]. Some strains produce biosurfactants, which are able to prevent the adhesion of pathogens [49]. L. salivarius TI 2711 administered as tablets did not significantly change the total number of bacteria in the saliva, or the number of mutans-streptococci and lactobacilli. Only for black-pigmented anaerobic rods, which include most periodontopathic bacteria, a significant decrease was found [50]. A promising strain is also L. reuteriProdentis, because not only antimicrobial, but also inflammatory, effects were found along with inhibition of plaque formation [51]. The reduction of pro-inflammatory cytokines (TNF-α, IL-8) in gingival crevicular fluid can be regarded proof of concept for the application of Lactobacillus reuteri in a chewing gum. The modulating effect of short-term intake of probiotics on the oral immune response is dose dependent [52]. An anti-inflammatory effect of Lactobacillus brevis, administered as lozenge to patients with chronic periodontitis, was shown by a significant decrease in the amount of nitrite/nitrate, prostaglandin E2 and matrix metalloproteinase in saliva [53]. Different microbial species show anti-inflammatory activity [46]. Although an anti-inflammatory effect of probiotic milk drink containing L. caseiShirota was shown in a clinical trial, increased plaque accumulation occurred, which is probably linked to the high sugar content [54]. The same product increased the plaque index and papilla bleeding index compared to baseline, whereas the amount of matrix metalloproteinase-3 was reduced [55]. By using a probiotic treatment (L. reuteri) also clinical effects were found, like a significant reduction in the gingival index and plaque scores [56]. Smokers showed a significantly greater improvement in plaque scores in a clinical trial using freeze-dried probiotic tablets of L. salivarius [57]. Even heat-killed L. plantarum decreased the depth of periodontal pockets in a randomised, double-blind, placebo-controlled clinical trial [58].

Caries

Dental caries is on the one hand associated with S. mutans, but also with lactobacilli. Indeed, L. plantarum and L. paracasei are among the species found in adult and childhood caries [59]. Various Lactobacillus spp. have numerical importance in carious dentine [60] and are often found in caries lesions [61]. L. salivarius LS 1952R induced dental caries in rats [62]. In an in-vitro model the combination of LGG and S. mutans was more cariogenic than monospeciesbiofilms of S. mutans. Therefore, LGG did not reduce the caries activity, but contributed to the caries process [63]. Another finding in this direction is that after restoration of caries, S. mutans and lactobacilli are reduced in saliva [64]. By now the role of lactobacilli in caries is not absolutely sure. It has been observed, that lactobacilli from caries-free people exert more effective inhibition of mutans-streptococci than lactobacilli isolated from caries-active subjects in-vitro [65]. L. plantarum 299v and other Lactobacillus strains coaggregate in-vitro with selected oral streptococci. Coaggregation is an important factor in the development of biofilms and dental plaque [66].

Several studies have been published that hint towards an anti-caries effect of lactobacilli. L. casei ATCC 11578 is able to prevent and even decolonise the adhesion of S. mutans to saliva-coated hydroxyapatite by modifying the protein composition of the salivary pellicle [67]. In a clinical study using L. paracasei GMNL-33 a significant reduction of salivary S. mutans was detected after 4 weeks [68]. L. paracasei showed maximum interference activity against S. mutans in-vitro, as reported by Simark-Mattsson et al. [65]. Lozenges containing L. brevis CD2 taken by school children with a high caries risk, resulted in a reduction in plaque acidogenicity, salivary mutans-streptococci and bleeding on probing [69]. L. reuteri significantly inhibited thegrowth of S. mutans, if applied as yoghurt [70], via straws and tablets [71,72] or as chewing gum [73]. A combination of L. sporogens, L. bifidum, L. bulgaricus, L. thermophilus, L. acidophilus, L. casei and L. rhamnosus either as capsule or as liquid form was tested. The counts of lactobacilli in the saliva increased significantly, but S. mutans was not significantly influenced. Unfortunately, no follow-up was performed after the trial [33]. Lactobacilli (including L. plantarum 299v, L. paracasei, L. reuteri, L. acidophilus and L. rhamnosus) showed co-aggregation with mutans-streptococci in-vitro and inhibited the clinical mutans-streptococci [74]. In a randomised, double-blind, placebo-controlled intervention study in 594 children, milk containing LGG reduced dental caries and the counts of mutans-streptococci after 7 months. Consequently, the risk of caries was significantly reduced [75]. In another double-blind, placebo-controlled trial of milk supplemented with fluoride and/or L. rhamnosus the effect on primary root caries lesions was investigated. The use of milk only supplemented with lactobacilli showed a positive effect, as well as the fluoride, but the beneficial effect was strongest by combining both [76].

Halitosis

There is currently limited evidence to support the use of lactobacilli in halitosis. After taking L. salivarius WB21 and xylitol in tablet form daily, the scores of an organoleptic test and bleeding on probing significantly decreased after 4 weeks [77]. In a randomised double-blind, placebo-controlled cross-over trail with 25 adults a probiotic chewing gum with L. reuteri DSM 17938 and L. reuteri ATCC PTA 5289 or placebo was used. After a treatment duration of 14 days the organoleptic scores were significantly lower in the group with probiotic chewing gum [78].

Candidiasis

For candidiasis there is also little evidence available. In a randomised, double-blind, placebo-controlled study the consumption of probiotic cheese (L. rhamnosus GG plus Propionibacteriumfreudenreichiisusp.Shermanii JS) led to reduced C. albicans counts [79]. However, in a 3 week intervention for which subjects ate 5 × 15 g cheese per day containing LGG and L. rhamnosus LC 705, no significant difference in the effect between the probiotic and control cheese on salivary Candida counts was found [80]. The administration of L. casei and Bifidobacteriumbreve for 20 d significantly reduced Candida prevalence and the level of anti-Candida immunoglobulin A [81]. Lozenges with L. reuteriProdentis are available on the market and reduce the prevalence of oral Candida in fragile elderly people [82]. Finally, L. plantarum and L. paracasei showed some co-aggregation with C. albicans [83].

The cited studies are often limited due to small sample size, lack of appropriate randomisation, blinding, duration of intervention plus follow-up and study set-up. Even fewer studies make inter-group comparison with true placebo or a negative control. The tested population is very heterogeneous and the probiotic doses applied differ up to 2 log scales. Furthermore, the vehicles used to administer the probiotics also clearly differ and might influence the outcome of the study. Typical means of probiotic administration are yoghurt, cheese [79], lozenges and tablets [72], capsules or liquids33. Besides, it is expected to be difficult to induce a microbiological shift or a clinical probiotic effect in a matured oral microbiological environment. The probiotics will have difficulties in colonising the mouth. A pre-treatment, to reduce the levels of indigenous microbiota, might be useful and necessary. The authors of a meta-analysis concluded, that the evidence supporting the use of probiotics to prevent or treat caries and periodontal disease is insufficient at present. In spite of the lack of evidence, they also state, that no adverse effects were reported, so there is no strong argument against using such a treatment [84].

Probiotic effects are strain-specific and cannot be transferred to other subspecies, or even other strains without any tests. Therefore, clear nomenclature of the evaluated strains is crucial for allociation of effects and for the judgement of efficiency. Overall, lactobacilli seem to be a promising way forward regarding restoration of periodontal health, especially in gingivitis, but better designed clinical trials in larger populations are needed. Moreover, combining different probiotic strains could lead to synergistic effects. This research field is in a very early state and enormous efforts are necessary to advance this promising strategy. An application in caries is questionable, as long as the impact of lactobacilli on the cause and progression of the disease is not clarified.

Anja Hoffmann reports Personal fees from Symrise AG and Rolf Daniels reports Grants from Symrise AG, both outside the submitted work. In addition, Rolf Daniels and Anja Hoffmann have a patent EP 15 184849.6 pending.