Over the last decade, understanding of the development and progression of oral diseases as well as their prevention and treatment has continuously improved (Petersen, 2003; Touger-Decker and Mobley, 2013). In addition to personal oral hygiene and the availability of dental care, nutrition has been increasingly acknowledged to be an important factor affecting health status of the oral cavity (Touger-Decker and Mobley, 2013). Certain nutritional patterns have been associated with an increased risk of developing oral diseases, such as consumption of sugar-rich food and drinks, whereas the consumption of protein-containing foods such as meat, cheese or fish, as well as use of chewing gum have been associated with a decreased risk (Touger-Decker and Mobley, 2013). The oral microbial community has recently come into focus as an important aspect of oral health care as the type of bacteria present in the oral cavity has a major impact on dental health status (Darveau, 2010; Dentino et al., 2013). 16S rRNA gene sequencing of the human oral bacterial community has shown that dental plaque consists of over 500 different species, dominated by the phyla Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria, Spirochaetes, and Fusobacteria (Wade, 2013). The commensal microbiota plays an important role in maintaining oral and systemic health.

The commensal oral microbiota inhibits colonization by opportunistic pathogens, a phenomenon known as colonization resistance (Wade, 2013). Because all surfaces of the mouth are colonized by commensals, there are few binding sites available for pathogens. Some health-associated bacteria have been shown to be antagonistic to oral pathogens. For example, Streptococcus salivarius strain K12 produces a bacteriocin which inhibits the growth of gram-negative species that are associated with periodontitis and halitosis in vitro (Wade, 2013). Socransky et al. (1998) showed that the presence of specific bacterial species in human plaque samples form reoccurring complexes that may be important in blocking pathogen colonization. An important step in understanding the microbiology of periodontitis was the observation that during the development of disease, a shift in those microbial complexes occurs (Darveau, 2010). While the microbiota present at healthy sites are mainly comprised of gram-positive species (Streptococcus spp.), at diseased sites different bacteria occur, such as gram-negative, anaerobic species including PG, Tannerella forsythia, Prevotella ssp., and Treponema denticola (Darveau, 2010; da Silva-Boghossian et al., 2011).

Lipopolysaccharides from the outer membranes of bacteria can activate TLRs, thus triggering an innate immune response (Hans and Hans, 2011). The LPS of the oral pathogens PG and EC differ in their chemical structure and immune-modulating properties (Tamura et al., 1992). Recent studies in our group have shown that the immune response to PG- and EC-LPS in in vitro challenged human gingival fibroblasts, as well as in immune cells such as monocytes and macrophages, is characterized by a 20–200-fold increase of inflammatory chemokines and cytokines within 6 h (Ehrnhöfer-Ressler et al., 2013; Walker et al., 2013; Soares et al., 2014; Schueller et al., 2015). While elevated levels of cytokines and chemokines have also been detected in diseased sites of patients suffering from gingivitis and severe periodontitis, the innate immune response to plaque bacteria in healthy tissues is normally delicately balanced (i.e., host homeostasis) to maintain healthy tissues and low levels of inflammatory markers (Bartold and Narayanan, 2006; Darveau, 2010). In periodontitis, bacterial communities are shifted and pathways of host homeostasis are altered, leading to an inflammatory infiltrate containing migrating macrophages and activated fibroblasts and endothelial cells as well as increased levels of inflammatory markers (Bartold and Narayanan, 2006). The chemokine IL-8 has important functions in oral health by facilitating the transit of activated immune cells into and through gingival tissues, and promoting immune-cell adhesion, tissue remodeling, and angiogenesis (Campbell et al., 2013). IL-8 is elevated in the saliva of oral cancer patients (720 vs. 250 pg/ml for control individuals) and has been proposed as a biomarker for the development of oral cavity and oropharyngeal squamous cell carcinoma (St. John et al., 2004). In patients with severe periodontitis, 4000–5500 pg/μl IL-8 have been detected in crevicular fluid at non-diseased sites (Goutoudi et al., 2012).

While considerable research has been done on the differences in inflammatory markers and bacterial communities for defining stages of oral disease progression, so far, no phylotypes or indicator indicator OTUs have been defined in oral health that precede the first clinical symptoms of oral disease. As most studies on periodontal health and disease have been done in a comparative way, in this study we deliberately only included orally healthy individuals to examine pre-clinical differences. Also, clinical studies mostly use tissues from tooth extractions or crevicular fluid for determining the inflammatory status, whereas in this study, we determined the release of basal IL-8 over 6 h from GECs in ex vivo incubations. Additionally, the impact of 6 h ex vivo stimulation of GECs with two different kinds of LPS (PG and EC) on IL-8 levels was assessed. The aim of this study was to link basal release levels of IL-8, nutrition, personal oral hygiene and the bacterial community in individuals with healthy gums, and to identify indicator OTUs for an immune response to external bacterial stimuli, established by measurement of IL-8 release from GECs. Our findings present new insights into the differences in IL-8 levels and the oral microbiota in healthy individuals and the response of GECs to external bacterial stimuli, which will be useful for future studies of oral health. This is, to our knowledge, the first study identifying indicator OTUs among oral bacteria associated with high basal release levels of IL-8 from GECs.

All chemicals and reagents were purchased from Sigma-Aldrich (Vienna, Austria), unless otherwise stated.

This study aimed at establishing associations of nutrition, oral hygiene habits and basal release of IL-8 from GECs with the oral bacterial community to identify indicator OTUs that might predict onset of oral pathologies. Basal IL-8 levels obtained in this study were neither comparable to levels found in crevicular fluid of patients with periodontitis (over 90,000-fold higher), nor to levels obtained by stimulated salivary measurement (250 pg/ml) in healthy individuals, which is to be expected due to impaired oral health and different sampling method as compared to our study (Navazesh and Christensen, 1982; Goutoudi et al., 2012). Comparable to this study, release levels of about 30–50 pg/ml IL-8 were found in primary sterile cultures of fibroblasts and periodontal ligament cells from tooth resections in healthy patients (Scheres et al., 2010). The basal levels of IL-8 in this study were found to positively correlate with pH (Table 2) but no significant association between pH and the bacterial community composition was found. Nevertheless, the positive correlation of the pH (values between 5.2 and 8.5) could be related to acidic or basic bacterial and cellular products released into the medium. It has to be considered that although the incubations were done without addition of antibiotics, the culture medium contains buffering salts, thus possibly biasing pH measurements.

In our study, the bacterial community of the mouth was dominated by the phyla Firmicutes, followed by Proteobacteria, Actinobacteria, Bacteroidetes, Fusobacteria and Candidate division TM7, which is consistent with the oral “core” microbiota reported in previous studies (Zaura et al., 2009; Lazarevic et al., 2010). Nutrition and oral hygiene habits can influence the oral microbiota composition and inflammatory state. Considering the latter, the presence of plant extracts in toothpaste showed an effect on basal IL-8 release (Table 3). This is in accordance with literature, as many medicinal plants that are used in toothpastes are anti-inflammatory, anti-bacterial, anti-fungal or astringent (Hotwani et al., 2014). The present analysis found a positive correlation of white and red wine consumption with basal IL-8 levels, and a significant difference between non-white wine drinkers and white wine drinkers (Tables 2, 3). Studies on alcohol consumption and the risk of periodontitis have reported variable findings. Susin et al. (2015) reported that a higher alcohol intake related to the incidence of periodontitis in women, whereas no such correlations could be shown in men (n = 1115). On the other hand, Kongstad et al. (2008) found a negative association of alcohol consumption and clinical signs of oral disease in men but not in women (n = 1521). In the present study, the separate analysis of gender subdivided by alcohol consumption would results in very small number of participants, which do not allow for such an analysis (n = 1–3 for some groups). Interestingly, red wine consumption was also positively correlated with IL-8 levels, as red wine is especially rich in polyphenols, has been shown to attenuate inflammatory markers in gastric inflammation as well as colonic epithelial cells (Mayo et al., 2000; Nunes et al., 2012). Considering our results of healthy participants, it might be hypothesized that moderate consumption of alcohol activates immune pathways in indirectly exposed oral cells, leading to higher basal release levels of IL-8. In our study, associations could be found between the composition of the oral microbiota and the use of chewing gum as well as oral hygiene habits like flossing. It has been demonstrated that the use of xylitol gum decreased the level of Streptococcus mutans but did not change the salivary microbial composition (Söderling et al., 2015), and that flossing habits reduce gingival inflammation by removing interproximal plaque (Berchier et al., 2008).

Moreover, we analyzed associations between the composition of the oral microbiota and consumption of different foods and drinks for which an impact on the inflammatory status has been discussed, such as meat, sweets, most commonly consumed alcoholic beverages and tea. Tea polyphenolic catechins, in particular epigallocatechin gallate and epicatechin gallate, can inhibit the growth of a wide range of gram-positive and gram-negative bacterial species with moderate potency. These molecules may be useful in the control of common oral infections, such as dental caries and periodontal disease (Taylor et al., 2005). In particular, catechin gallates inhibit growth and adherence to oral epithelial cells of P. gingivalis (Sakanaka et al., 1996) and Prevotella (Hirasawa et al., 2002).

As many oral bacteria possess glucose transport systems, they are capable of utilizing dietary carbohydrates and some sugar alcohols. Incorporated carbohydrates and sugar alcohols are subjected to glycolysis via metabolic reactions catalyzed by constitutive and inducible enzymes. Polysaccharides can be hydrolysed into oligosaccharides, disaccharides, and monosaccharides by host and bacterial glycosidases. For example, host α-amylase hydrolyses cooked starch into carbohydrates, which can then be incorporated and metabolized by oral bacteria (Takahashi, 2015). Interestingly, there was a strong association between oral microbiota, meat consumption and basal IL-8 levels (Figure 3). The association between oral microbiota and meat may be explained by the degradation activity of proteins into peptides and amino acids by bacterial proteases and peptidases in the oral cavity (Takahashi, 2015).

In established in vitro oral and immune cell lines, levels of inflammatory markers are up-regulated after stimulation with LPS from various bacterial strains (Ehrnhöfer-Ressler et al., 2013; Walker et al., 2013; Schueller et al., 2015). This is also true for primary sterile cultures of oral fibroblasts (Nagaoka et al., 1996). In our study, using freshly sampled epithelial cells and not using antibiotics, this kind of up-regulation was not established for IL-8 after 6 h of stimulation with pgLPS. Muthukuru et al. (2005) found that in healthy tissues, the expression of TLRs is much lower than in periodontal disease, which would account for the absence of stimulation by their agonist LPS. Additionally, Scheres and Crielaard (2012) demonstrated that primary sterile gingival fibroblast cultures were desensitized to live PG after repeated contact with the pathogen. In this study, cells were cultured directly after sampling from the mouth, with live plaque bacteria still present. It is therefore very likely that GECs exhibited tolerance to EC- and PG-LPS, as observed in the experiments by Scheres and Crielaard (2012). The significant down-regulation of IL-8 release after stimulation of GECs with a non-oral-derived bacterium (EC) only occurred in the high IL-8 group on relative levels. Muthukuru et al. (2005) demonstrated that a repeated challenge with LPS from EC results in a significant down-regulation of TLRs on the cell-surface, thus impeding or reducing activation of IL-8 transcription and subsequent protein release. No association could be found between basal IL-8 levels and response to LPS challenge (p = 0.065). However, the indicator OTU analysis yielded very interesting results (Figure 4). The indicator OTUs that we identified are typical members of the oral cavity according to the Human Oral Microbiome Database (HOMD) (Chen et al., 2010) and some have also been associated with periodontal disease or other inflammatory pathologies. For example, Capnocytophaga ochracea is present in the dental plaque biofilm of patients with periodontitis and it has been reported in several cases of periodontal disease (Hosohama-Saito et al., 2016). Propionibacterium propionicum was found in root canal-treated teeth with apical periodontitis (Siqueira and Rocas, 2009), Eggerthia catenaformis was recently isolated in a patient with a dental abscess (Kordjian et al., 2015) and Actinomyces dentalis was isolated from pus of a human dental abscess (Hall et al., 2005). In particular, the genus Actinomyces houses several long-established human pathogens. Some species cause the inflammatory disease actinomycosis, whereas others are associated with various non-specific inflammatory processes, or may play a role in the development of dental plaque and subsequent caries or periodontal diseases (Hall et al., 2005). Likewise, Selenomonas species dominated the diseased sites of subjects with generalized aggressive periodontitis. Selenomonas sputigena was the most frequently detected bacterial species whereas other species of Selenomonas were often present in high levels, including Selenomonas noxia, Selenomonas oral clone EW076, Selenomonas oral clone EW084 and Selenomonas oral clone CS002. Gonçalves et al. (2012) revealed that Selenomonas phylotypes were frequently part of the subgingival microbiota of generalized aggressive periodontitis patients. Also Drescher et al. (2010) analyzed the topography of the subgingival biofilm by fluorescence in situ hybridization (FISH) and electron microscopy in subjects with generalized aggressive periodontitis, chronic periodontitis and periodontitis-resistant subject and revealed that Selenomonas spp. appeared in large numbers in all parts of the collected biofilms and seemed to make a relevant contribution to their structural organization.

In our study, we found an association between basal IL8 levels and microbiota, suggesting a link between oral bacteria and inflammatory state. Moreover, a link between nutrition, personal oral hygiene with oral microbiota and IL8 levels was also observed. The elevated abundance of some bacterial OTUs in high IL-8 responders, indicating a higher inflammatory activity GEC, may be causally linked to stimulation by bacterial LPS. The indicator OTUs that we identified in healthy subjects may hold promise as biomarkers for inflammatory status of the oral cavity as well as for risk of onset of oral pathologies. The validation of these biomarkers will require further studies with larger cohorts and designs where cohorts are followed until some develop disease so that their prognostic value can be evaluated.

All subjects gave written informed consent in accordance with the Declaration of Helsinki. As samples were anonymized and provided by healthy individuals, ethics committee approval was not required according the regulations of the University of Vienna.

KS designed the study, performed experiments and data analysis, and drafted the manuscript. AR performed microbiota experiments and data analysis and drafted the manuscript. SP performed experiments and data analysis. DB performed microbiota data analysis and contributed to manuscript writing. VS designed the study and contributed to manuscript writing.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.