High salivary levels of JP2 genotype of Aggregatibacter actinomycetemcomitans is associated with clinical attachment loss in Moroccan adolescents

Abstract It has previously been shown that the presence of Aggregatibacter actinomycetemcomitans in subgingival plaque is significantly associated with increased risk for clinical attachment loss. The highly leukotoxic JP2 genotype of this bacterium is frequently detected in adolescents with aggressive forms of periodontitis. The aims of the study were to quantify the levels of JP2 and non‐JP2 genotypes of A. actinomycetemcomitans in saliva of Moroccan adolescents with the JP2 genotype earlier detected in the subgingival plaque. The salivary concentrations of inflammatory proteins were quantified and linked to the clinical parameters and microbial findings. Finally, a mouth rinse with leukotoxin‐neutralizing effect was administrated and its effect on the levels the biomarkers and A. actinomycetemcomitans examined. The study population consisted of 22 adolescents that previously were found to be positive for the JP2 genotype in subgingival plaque. Periodontal registration and sampling of stimulated saliva was performed at baseline. A mouth rinse (active/placebo) was administrated, and saliva sampling repeated after 2 and 4 weeks rinse. The salivary levels of JP2 and non‐JP2 were analyzed by quantitative PCR and inflammatory proteins by ELISA. Both the JP2 and the non‐JP2 genotype were detected in all individuals with significantly higher levels of the non‐JP2. Enhanced levels of the JP2 genotype of A. actinomycetemcomitans was significantly correlated to the presence of attachment loss (≥3 mm). Salivary concentrations of inflammatory biomarkers did not correlate to periodontal condition or levels of A. actinomycetemcomitans. The use of active or placebo leukotoxin‐neutralizing mouth rinse did not significantly interfered with the levels of these biomarkers. Saliva is an excellent source for detection of A. actinomycetemcomitans on individual basis, and high levels of the JP2 genotype were significantly associated with the presence of clinical attachment loss.


| INTRODUCTION
Periodontal clinical attachment loss (CAL) is highly prevalent in adolescents in North-and West-Africa (Haubek et al., 2001(Haubek et al., , 2008Haubek, Ennibi, Poulsen, Benzarti, & Baelum, 2004;Höglund Åberg, Kwamin, Claesson, Johansson, & Haubek, 2012;Kissa et al., 2016). High prevalence of a highly leukotoxic genotype (JP2) of the periopathogen Aggregatibacter actinomycetemcomitans has contributed to explain the high numbers of aggressive periodontitis (AgP) in these geographic regions (Haubek et al., 2008;Höglund Åberg et al., 2014). Longitudinal studies have shown that periodontally healthy carriers of the JP2 genotype of A. actinomycetemcomitans in the subgingival plaque have a significantly increased risk to develop CAL compared with individuals without this bacterium (Haubek et al., 2008;Höglund Åberg et al., 2014). The leukotoxin that is highly expressed by the JP2 genotype of A. actinomycetemcomitans protects the bacteria from phagocytic killing by neutrophils and induces a pro-inflammatory cell death in macrophages (Johansson, 2011). The leukotoxin-affected macrophages release substantial amounts of bioactive interleukin-1β (IL-1β) by activation of the inflammasome complex in the cytosol (Kelk et al., 2011). This cytokine is a highly active pro-inflammatory protein that is used as a target in the treatment of many inflammatory-induced degenerative disorders (Dinarello, Simon, & van der Meer, 2012).
In aggressive forms of periodontitis, CAL occurs at a much faster rate than in chronic periodontitis and the disease is associated with the presence of specific pathogens (Könönen & Müller, 2014). The presence of A. actinomycetemcomitans in the subgingival plaque is highly linked to AgP, especially the highly leukotoxic genotypes described previously (Ennibi, Benrachadi, Bouziane, Haubek, & Poulsen, 2012;Haubek et al., 2001Haubek et al., , 2008Höglund Åberg et al., 2014;Johansson, Claesson, Höglund Åberg, Haubek, & Oscarsson, 2017). The management of AgP remains a challenge for the clinicians due to lack of standardized protocols for effective disease control (Akrivopouloua, Green, Donos, Nair, & Ready, 2017). The traditional management of this disease consists on mechanical debridement with or without surgery, and generally supplemented with antimicrobial drugs (Deas & Mealey, 2010;Eick et al., 2018). There is a great need of new more specific treatment strategies as well as new tools for an early identification of risk individuals. Targeting the etiological factors before disease onset might be the optimal preventive strategy for aggressive forms of periodontitis.
In this infection-induced inflammatory disease, both bacterial virulence factors and host-related inflammatory proteins have the potential to be targets for therapeutic strategies.
Colonization of A. actinomycetemcomitans in humans takes place in early childhood, acquired from close relatives and localized in the oral mucosa (Könönen & Müller, 2014). The bacterium can be translocated from the initial oral colonization site to subgingival crevices and has to compete with other bacteria in the periodontal niche. Establishment of persistent colonization of A. actinomycetemcomitans in subgingival crevices may lead to periodontal destruction and thereby development of periodontitis in susceptible individuals (Fine, Kaplan, Kachlany, & Schreiner, 2006). This colonization pattern of the bacterium indicates that saliva can be a useful source for microbial sampling in order to identify healthy risk individuals carrying periodontal pathogens.
Plants and other natural products have been extensively used in the management of oral infections and improvement of oral health in many populations (Akkaoui & Ennibi, 2017;Chinsembu, 2016). Among several tested herbal plants, we found that extract of leaves or twigs from Psidium guajava efficiently neutralizes the activity of A. actinomycetemcomitans leukotoxin (Kwamin, Gref, Haubek, & Johansson, 2012). Virulence blocking might be a tool for preventive strategies for aggressive forms of periodontitis associated with the presence of highly leukotoxic A. actinomycetemcomitans (Ben Lagha, LeBel, & Grenier, 2018;.
The aims of the present study were to quantitatively examine the presence of JP2 and non-JP2 genotypes of A. actinomycetemcomitans in saliva of healthy as well as periodontally diseased Moroccan adolescents.
Second, it was to quantify inflammatory proteins in saliva in relation to the number of A. actinomycetemcomitans and the presence of CAL.
Finally, we analyzed effect of leukotoxin-neutralizing mouth rinse on salivary levels of A. actinomycetemcomitans and inflammatory proteins.

| Ethical considerations
Ethical clearance for this study has been approved by the Ethical committee of Mohammed V University, Rabat, Morocco (N/R 67/16).
Signed consents were received from the parents or the guardians of the participants before they entered the study. Subjects identified with AgP were informed of their status and referred for treatment to the Dental School of Mohammed V University, Rabat, Morocco.

| Study population
The study population comprised 22 Moroccan pupils from different secondary schools in Rabat. All individuals have previously been shown to be positive for the presence of the JP2 genotype of A. actinomycetemcomitans in subgingival plaque, when analyzed at the age of 8-9 years by conventional PCR (Jensen, Ennibi, Ismaili, Poulsen, & Haubek, 2016). The examined adolescents consisted of 13 boys and nine girls with a mean age of 13.5 ± 0.7 years. All examined students had a bad oral hygiene. Indeed, they all showed light to heavy plaque deposits on tooth surfaces, and all had mild gingivitis. Spontaneous bleeding was reported by 8.7% of the student, and 78.2% reported having gingival bleeding when they brushed their teeth. In a previous study on oral hygiene in Moroccan school children and their mothers, authors noticed that the use of a toothbrush was low in children (68%); and clinical evaluation of plaque index was high 56.2. In the same study, bleeding index was 13.3 and the average decayed missing filled teeth index was 6.5 ± 2.6 (Assimi, Tajmouti, & Ennibi, 2016).

| Collection of saliva samples
Saliva from each participant was collected before the clinical examination, as well as after 2 and 4 weeks of rinsing with the test products.
The participants were asked to chew a paraffin tablet (Ivoclar Vivadent AB, Solna, Sweden) for 1 min. Subsequently, saliva (>1 ml) was collected in a plastic beverage cup. One milliliter of the saliva was transferred into a sterile plastic tube with screw cap and stored in a freezer (−20°C) until quantification of A. actinomycetemcomitans and inflammatory proteins.

| Clinical examination and documentation
The examination included measurement of the probing pocket depth and the distance from the free gingival margin to the cementoenamel junction at the buccal aspect of the mesial and distal surfaces of all fully erupted permanent teeth by means of a calibrated periodontal probe (Rönvig®, Kohler, Germany). CAL was defined as the distance from the cemento-enamel junction to the bottom of the periodontal pocket or crevice and was calculated as the difference between the two measurements described above. Individuals were regarded as being periodontally diseased, if one or more sites had periodontal attachment loss of 3 mm or more, and as periodontally healthy if no such sites were found. After the first examination (Day 0), all received a toothbrush, fluoride toothpaste, and the mouth rinse (either the placebo or the active solution). At the end of the clinical study, the students were given a questionnaire regarding treatment satisfaction. The students were asked questions about the taste of the product and gingival bleeding before and after using the mouth rinse.

| Production and administration of active and placebo mouth rinse
Guava leaves were collected in Ghana by Dr F. Kwamin and transported with a courier to Umeå University, Umeå, Sweden.
Guava leaves at a concentration of 250 g/L water were boiled for 10 min before the leaves were removed by filtration and cleared from debris by centrifugation. The guava raw extract was mixed in 0.2% sodium fluoride in water at a concentration of 5% of the total volume (active mouth rinse). The placebo mouth rinse consisted of only 0.2% sodium fluoride in water and was packed and labeled identically with the active mouth rinse. Each participant was administrated randomly 2 × 60 ml mouth rinse of either active or placebo mixture. All individuals were instructed to rinse 10 ml twice per day (morning and evening) during a 4-week period. The manufacturing of the guava mouth rinse clinical material, that is, active test product and placebo test product, has been performed according to a standard close to cGMP (Camber consulting AB, Bromma, Sweden).
Leukotoxin-neutralizing activity and microbial purity of the test products were controlled before the onset of the clinical trial. Both the examiner and the participants were blinded to the allocated mouthwash sample.

| DNA isolation
Stimulated saliva was collected from 22 individuals, three samples per donor, and mixed 1:1 with DNA preservation buffer (Norgen Biotek Corporation, Thorold, Canada). The Viral DNA extraction kit (DiaSorin AB, Dublin, Ireland) was used for the DNA isolation, and for the procedure, an automated extraction instrument was used (Liaison® IXT, Diasorin AB, Ireland). DNA was extracted from 550 μl of the sample mixture and eluted in a volume of 100 μl. Standard suspensions of the JP2 genotype (HK1651) and non-JP2 genotype (D7s), respectively, (10 8 -10 1 cells/ml), prepared in A. actinomyctemcomitans-free saliva, were treated as described above. The samples and the standard solutions were stored at +4°C until use.
The samples and the standards solution (10 8 -10 1 cells/ml) were analyzed in duplicates by using a Corbett Research Rotor-Gene 6000 Rotary Analyze instrument (QIAGEN, Valencia, CA, USA).
The two qPCR-based methods for quantification of A. actinomycetemcomitans in the saliva samples, in accordance with Kirakodu (total Aa) and Yoshida (non JP2 + JP2), respectively, were compared.

| Quantification of inflammatory proteins by ELISA
The inflammatory-associated proteins IL-1β, MMP-8, sCD14, and sICAM-1 were quantified by commercial ELISA kit available from RnD systems (Abingdon, UK). The saliva samples were thawed and an aliquot centrifuged in 10,000g for 10 min at 4°C. The supernatants were stored on ice, and the specific proteins in the solutions were quantified by the different ELISA kits in accordance to the manufacturer's protocol (RnD systems).

| Statistical analyses
Data analyses were performed using SPSS 24.0 (SPSS Inc., Chicago, IL, USA) and STATA 8.0 (StataCorp LP., College Station, Texas, USA). In the statistical analyses, the primary outcome was salivary levels of A. actinomycetemcomitans (JP2 or non-JP2 genotypes), inflammatory biomarkers, and CAL ≥ 3 mm in one or more sites at the subject level.
The quantitative variables were expressed as medians and quartiles.
Logistic regression analysis was used to determine the association of the dependent variables and CAL ≥ 3 mm, respectively. Mean values of three independent observations from each individual were used in the calculations. Differences between the effect of the two different blends of mouth rinse on the levels of salivary biomarkers and A. actinomycetemcomitans were evaluated by Mann Whitney U test.
Confidence interval and P values were based on the Wald statistics.
A level of significance of 5% (P < 0.05) was used for the analyses.

| RESULTS
Clinical registrations of the study population showed that four out of the 22 individuals had ≥1 site of AL ≥ 3 mm (Table 4). The salivary concentrations of the inflammatory proteins IL-1β, MMP-8, sCD14, and sICAM-1 were compared between the two groups of periodontally healthy or diseased individuals and showed that there were no significant differences between the two groups ( Figure 5).
The use of active leukotoxin-neutralizing mouth rinse during the 4-week test period did not significantly change the salivary levels of A. actinomycetemcomitans or inflammatory proteins compared with the group that was administrated the placebo mouth rinse (Table 5). Despite these negative results, 69.6% of the students stated that gingival bleeding was improved after using the mouth rinse. This self-reported improvement was independent of if the student used active or placebo mouth rinse. However, students' statements are subjective, and variables like plaque index and bleeding index should be evaluated in a future study on a larger population to help on having objective conclusions.

| DISCUSSION
Salivary analyses on the young (12-15 years old) study population in the present study showed that A. actinomycetemcomitans of both the non-JP2 and the JP2 genotypes could be detected in all individuals.
The presence of the JP2-genotype in saliva of all individuals was not surprising; based on that, the inclusion criteria for the studied adolescents was a previous detection of this genotype in samples from subgingival plaque. Despite this inclusion criterion, the salivary concentration of non-JP2 genotype was much higher than that of the JP2 genotype in the majority of the samples. In four of the 22 individuals, CAL of ≥3 mm could be detected and these individuals were classified as periodontally diseased, in line with previous investigations (Haubek et al., 2008). When results from the periodontally diseased individuals were separated from that of the healthy, we found that the JP2 genotype was detected at high levels in saliva from the periodontally diseased individuals. This further strengthens the role of the JP2 genotype as a disease-associated risk factor in this population, in line with previous studies (Haubek, 2010;Tsai et al., 2018). High levels of JP2 genotype A. actinomycetemcomitans in saliva may be a risk factor for disease onset but could also be a result of release of bacteria from diseased pockets. It has previously been shown that the presence of periodontal pathogens in subgingival plaque correlate with that found in saliva; however, these data were semiquantitative without specific quantification (Haririan et al., 2014;Kageyama et al., 2017).
Saliva has previously been used as a source for quantification of host inflammatory proteins associated to infection or inflammation (Bostanci & Bao, 2017). In the present study, we found that out of Levels of Aggregatibacter actinomycetemcomitans in saliva collected from Moroccan adolescents. Correlation of results from two different qPCR-based methods is shown. JP2 and non JP2 genotypespecific sequences within the leukotoxin promotor region (x-axis) and sequences within ltxA (y-axis) are targeted the periodontitis-associated biomarkers that were quantified, no one of them were linked to disease. We have previously showed that the JP2 genotype is an efficient inducer of IL-1β secretion from human macrophages (Kelk, Claesson, Chen, Sjöstedt, & Johansson, 2008).
We can conclude from the results in the present study that saliva can be a suitable source for identification of carrier ship for both JP2 and non-JP2 genotypes of A. actinomycetemcomitans. High levels of the JP2 genotype of A. actinomycetemcomitans were correlated to the occurrence of CAL; however, it should be noted that the study population was small. Finally, with the present test protocol with 4 weeks use of leukotoxin-neutralizing mouth rinse, we could not see any significant effects on the salivary levels of A. actinomycetemcomitans or inflammatory proteins.