Dental caries is a non-communicable biofilm-mediated disease affecting both crown and exposed root surfaces in the primary and permanent dentitions. The carious process involves interactions between the biofilm formed on the tooth surface, sugars, salivary and genetic factors. Based on more than 100 years of research, there is unequivocal evidence that dietary fermentable carbohydrates (sugars and starch) play a key role in caries initiation and progression. In this context, sucrose deserves special attention; apart from being rapidly converted into acids it is also synthesized into extracellular glucans, fructans and intracellular storage compounds. According to the ecological plaque hypothesis, it is generally accepted that sucrose exposure is fueling and driving the stable and diverse symbiotic oral biofilm to a dysbiotic form with a reduced diversity and overgrowth of acidogenic and acid-tolerating species. Such typical bacterial profiles have demonstrated in subjects with different stages of caries in cross-sectional settings but the timing associated with a sugar provocations is less known. Moreover, the use of probiotic bacteria in adjunct to regular oral care to support biofilm diversity and prevent dental caries has gained momentum in recent years. It has been demonstrated that probiotic supplements can increase salivary pH, and reduce the counts of salivary S. mutans, thereby exert a stabilizing effect on the oral microbiota. In this context, it is of interest to explore whether or not the use of beneficial bacteria can counteract a sugar-driven shift in the salivary microbiota. Another question of interest is to study if the oral biofilm has a colonization memory similar to that of the gastro-intestinal tract and the suggested study design could possibly enlighten this area of research.

Objective

The aim of this study is to investigate the bacterial profiles in whole saliva before and after a 2-week sugar provocation with and without probiotic supplements in healthy young adults. A second aim is to study the bacterial profiles 3 weeks after the termination of the sugar stress. The null hypothesis are:

1. Sucrose-induced stress does not affect the bacterial profiles in whole saliva compared with stress induced by a non-fermentable pentitol (xylitol)
2. The intake of lactobacilli-derived probiotic bacteria does not affect the sugar-induced changes in the bacterial profiles in whole saliva when compared with placebo
3. The bacterial profiles in saliva do not differ from baseline 3 weeks after the termination of the intervention.

Material The study group will consist of 80 young adults with uncompromised oral health that volunteer after informed consent. The inclusion criteria are i) over 18 years with more than 20 own natural teeth, ii) no chronic systemic diseases affecting salivary functions, iii) no medication except for contraceptives, and iv) being a non-smoker. The project will be submitted for ethical approval.

Study design The study will employ a randomized triple-blind, placebo controlled design with four parallel arms as shown in Figure 1. The group allocation will be arranged through computer-generated envelopes and concealed for all parts (subjects, investigators, laboratory staff). After the baseline registration (Day 0), the participants are asked to rinse their mouth with 10 mL of a sucrose- or xylitol-containing solution every second hour (7-8 times per day) during 14 days (day 14). The solution should be "swished" around and between the teeth for at least 30 seconds and then spitted out. No after-rinsing with water is allowed. In addition, the subjects are provided with tablets containing either probiotic lactobacilli or placebo for daily use during the 14-day rinsing period. The participants will be thoroughly instructed to maintain their normal eating and drinking habits as well as their daily oral hygiene procedures.

Intervention - probiotics and placebo The active and placebo lozenges for the 14-day intervention period will be distributed to the participants after baseline examination and sampling. The active lozenges contain two strains of probiotic bacteria; Lactobacillus rhamnosus PB01 (DSM 14869) and Lactobacillus curvatus EB10 (DSM 32307) at a concentration of 109 colony forming units each. The tablets have previously been used in a randomized controlled trial with gingival inflammation and microbial endpoints (Keller et al., 2018). The full composition is shown in Appendix X. The placebo lozenges have an identical composition, except for the bacteria, with the same size and taste. The participants are instructed to let one tablet slowly dissolve in the mouth two times per day (morning, evening) after regular tooth brushing. They are also required to keep a daily logbook covering the sugar rinses and tablet intakes and to return non-consumed tablet at the first follow-up. The sugar rinses will be prepared and distributed in non-marked bottles and the participants will be supplied with a dose spoon indicating 10 mL. The participants will also be given a standardized fluoride toothpaste for use during the entire study. An independent monitor at Copenhagen University will guarantee the allocation concealment.

Clinical examination and saliva samplings At baseline, a clinical examination is performed by one experienced and calibrated dentist. No radiographs are exposed. The following variables are registered: a) caries experience, expressed as decayed, missed and filled teeth (DMFT), b) the amount of supragingival plaque (s-PI), c) the level of gingival inflammation (s-GI), and d) bleeding-on-probing (s-BOP). The PI, GI and BOP will be scored in a simplified manner on six pre-selected teeth and re-evaluated at the follow-ups after 14 and 35 days. Stimulated whole saliva samples (approximately 1.0 mL) will be collected at baseline and after 14 and 35 days in a standardized way by one trained investigator. All samples are stored at -80°C until further processing.

Laboratory procedures The DNA will be extracted using a Pathogen_Universal_200 protocol (Roche) in accordance with the manufacturer's guidelines at the Institute for Inflammatory Research, Rigshospitalet, København, Danmark. The profiles of the salivary microbiome will be analysed with Next Generation Sequencing (NGS) and qPCR as preciously described (Keller et al., 2018).

Endpoints Salivary bacterial profiles will be analyzed by means of the Human Oral Microbe Identification using Next Generation Sequencing (HOMINGS) in co-operation with the Costerton Center at University of Copenhagen. Secondary endpoints are simplified plaque index (PI), gingival index (GI) and bleeding-on-probing (BOP) as scored after 15 and 35 days.

Statistical methods All data will be checked for normality and processed with the IBM-SPSS software. Clinical data are compared using the Freidman test with Dunn's comparison. Relative abundance of bacterial DNA reads are compared between groups of samples at genus and species level using the Kruskal-Wallis and Mann-Whitney tests with Benjamini-Hochberg correction for multiple dependent analyses. P-values less than 0.05 will be considered statistically significant.

Inclusion Criteria:

• more than 20 own natural teeth
• no chronic systemic diseases affecting salivary functions
• no medication except for contraceptives
• non-smoker

Exclusion Criteria:

• treatment requiring oral diseases i.e. dental caries, gingivitis and periodontitis
• systemic antibiotics within the latest 3 months

Computerized randomization followed by pseudo anonymication