Periodontal Disease and Subgingival Microbiota as Contributors for RA Pathogenesis: Modifiable Risk Factors?

Abstract

Purpose of review

Since the early 1900s, the role of periodontal disease in the pathogenesis of rheumatoid arthritis has been a matter of intense research. The last decade has witnessed many advances supporting a link between periodontitis, the presence of specific bacterial species (i.e., Porphyromonas ginigivalis) and their effects in immune response. This review will examine available evidence on the subject.

Recent findings

Epidemiological studies have stressed the commonalities shared by periodontal disease and rheumatoid arthritis. Many groups have focused their attention towards understanding the periodontal microbiota and its alterations in states of health and disease. The presence of circulating antibodies against periodontopathic bacteria and associated inflammatory response has been found in both RA patients and subjects at-risk for disease development. Most recently, the periodontal microbiota of smokers and patients with RA has been elucidated, revealing profound changes in the bacterial communities compared to that of healthy controls. This has led to several small clinical trials of PD treatment as adjuvant for disease-modifying therapy in RA.

Summary

Smoking and periodontal disease are emerging risk factors for the development of RA. Epidemiological, clinical and basic research has further strengthened this association, pointing towards changes in the oral microbiota as possible contributors to systemic inflammation and arthritis.

Introduction

Rheumatoid arthritis (RA) describes a relatively novel term defining one of the most common inflammatory arthropaties. It affects roughly 1% of the Caucasian population and its incidence peaks in middle-age women[1]. Some ethnic groups, particularly those belonging to various Native American tribes, have a strikingly high prevalence of the disease, leading to the speculation that RA is rather a disease of the New World that has been ‘transmitted’ to other populations after the Conquest [2,3]. Genes have been implicated, but the current discoveries by gene-wide association studies (GWAS) can only explain about a quarter of the disease variance[4]. Further, studies in monozygotic twins revealed a concordance rate of up to 15%, suggesting that environmental and/or epigenetic factors are required for clinically evident RA[5–7].

Anti-citrullinated peptide antibodies (ACPAs) are immunoglobulins highly specific for the diagnosis of RA and are markers for worse prognosis and disease progression[8]. Intriguingly, they are detectable in circulation many years prior to the development of clinically classifiable RA (i.e., pre-clinical phase of disease)[9]. During this phase, curiously, the synovial architecture shows no signs of inflammation or immune cell activation[10]. This, coupled with the fact that ACPA and cytokine titers only rise just a few months prior to clinical arthritis[11], has led to many investigators to propose that there must be a ‘second hit’ beyond the mere presence of humoral autoimmunity in order to trigger the synovitis phase of RA.

Many environmental factors have been contemplated in RA pathogenesis over time[12]. Chief among those is tobacco smoking[13]. Recent epidemiological data supports this association, particularly in the context of genetic predisposition and seropositivity[14]. The mechanistic reasons behind this connection are likely multifactorial, but a plausible explanation may relate to the deleterious effects of tobacco and other substances in mucosal sites, particularly in the periodontium and distal airways. In this review we will cover the most current knowledge on smoking as a risk factor for both PD and RA, periodontal disease pathogenesis (and its microbiota changes) and rheumatoid arthritis development.

The Healthy Periodontal Microbiome

The oral cavity harbors billions of bacteria and ranks second in total microbial load when compared to the human gut and other body sites. Over 1000 species or phylotypes have been detected in the oral cavity by our group and others[15,16].

Early observations employing 16S rRNA gene-sequencing evaluated the oral microbiome of clinically healthy individuals[17]. Biofilm samples from subgingival sites with no signs of disease or inflammation (gingival sulci) were obtained. Whereas several species of Streptococcus and Gemella were often found, species associated with periodontal disease, such as Porphyromonas gingivalis, Tannerella forsythia, and Treponema denticola, could not be detected. Other studies have identified these putative periodontal pathogens in healthy gingival tissues, although at low levels[18,19].

Interestingly, the microbiome diversity of the healthy oral cavity is dependent on geographic localization. Examination of healthy gingival sulci has unraveled a distinct predilection of Streptococci at high levels for the vestibular gingival sulci and of Fusobacterium for the lingual gingival sulci[20]. These differences are related to the oxygen conditions, which are lower for the lingual sites, leading to selection for anaerobic species such as Fusobacterium. Conversely, oxygen levels in the vestibular sites are higher, providing a biologically stable environment for aerobes or facultative anerobes such as Streptococci. This fact underscores the need to sample multiple sites when studying the subgingival microbiome.

Other studies using 454-pyrosequencing suggested that the relative abundance of Actinomyces, was higher in health when compared to PD patients[21].

Lastly, a study employing a small number of periodontally healthy subjects confirmed previous studies showing that gram-positive genera such as Streptococcus, Actinomyces, and Granulicatella were significantly enriched in healthy periodontal samples[22]. Moreover, this study looked at the functional potential of the healthy periodontal microbiome. A limited number of pathways were significantly enriched in the periodontally healthy microbiome including pathways for fatty acid biosynthesis, purine metabolism, and glycerol-3-phosphate metabolism. Fatty acids in particular have been shown to have a protective role in periodontal health, suggesting that some of these metabolites are synthesized by the healthy microbiota in an effort to protect against periodontopahic taxa.

Smoking and Periodontal Diseases

Tobacco smoking is an established and modifiable risk factor for periodontal inflammation and destruction[23]. Similarly, smoking has a negative impact on several inflammatory diseases, including rheumatoid arthritis, multiple sclerosis and inflammatory bowel disease[24].

The epidemiology of the effects of tobacco smoking on periodontal health and possible associated mechanisms has been extensively reviewed [25,26] and is beyond the scope of this manuscript. Evidence for the association between smoking and destructive periodontitis is apparent in several cross-sectional studies. The strength of these associations derives from one of the largest studies involving over 1300 subjects[27]. Participants who smoked were at greater risk for experiencing severe bone loss compared to never-smokers. Light and heavy smokers exhibited odds ratios for PD of 3.35 and 7.8, respectively. A more recent study based on 12,239 participants in NHANES III database corroborated earlier findings and showed that current smokers were 4 times more likely to be diagnosed with destructive periodontitis than non-smokers and that a dose-response could be demonstrated[28].

Longitudinal studies have also provided evidence of the relationship of smoking and PD in young and older adults alike[29–32], demonstrating that tooth loss, bone loss and attachment loss significantly increased overtime in smokers when compared to non-smokers. Interestingly, the periodontal health condition in former smokers is similar to that of non-smokers and remains stable over time, suggesting that smoking cessation can reverse at least partially periodontal homeostasis.

Observations of the Effects of Smoking on the Selection of Periodontal Pathogens

Periodontal diseases are chronic polymicrobial infections that lead to local inflammatory reactions mediated by antigenic challenges. Evaluation of how microbes activate the immune system in the periodontal tissues is possible through sampling of subgingival periodontal biofilms. Available data related to the effects of smoking on the selection of periodontal pathogens suggests that smokers may harbor a putative periodontal pathogenic flora. Early observations using culture-dependent methods showed that Porphyromonas gingivalis , Tannerella forsythia and Aggregatibacter actinomycetemcomitans were more prevalent in smokers than in no-smokers[33]. Others have studied the effects of smoking on selected periodontal pathogens in periodontal patients[34]. Using logistic regression analysis, one group revealed that smoking and specific measures for PD (i.e., increased periodontal probing depth) significantly correlated with the detection of A. actinomycetemcomitans, Prevotella intermedia, and Eikenella corrodens. Employing real time PCR, Teixeira et al. have also demonstrated a positive dose-response relationship between smoking and levels of putative periodontal pathogens, most notably P. gingivalis[35,36].

These findings have been somehow controversial, as others have reported no differences between smokers and non-smokers with respect to the detection of periodontal pathogens [37,38]in terms of prevalence [39]and microbial load [34]in the subgingival domain.

A more recent study looked at the subgingival microbiota of current and never-smokers with PD using 16S sequencing [40]. The microbial profile of smoking-associated PD was distinct from that of non-smokers, with significant differences in the prevalence and abundance of multiple organisms, including increases in pro-periodontopathic taxa such as Parvimonas, Fusobacterium, and Treponema. Taken together, these studies suggest that smoking can be associated with a specific microbial signature independent of PD status.

Smoking and Rheumatoid Arthritis

Tobacco smoking is perhaps the best-established environmental risk factor for RA development[41–43].

A recent meta-analysis has corroborated this relationship, particularly in hose patietns that are seropositive and heavy smokers[44].

Population-based case–control studies[43] have shown that smokers have twice the risk of developing seropositive RA when compared with never smokers. This association appears to be dose-dependent on lifetime exposure to smoking.[45,46]

As discussed, compelling evidence has also shown strong gene–environment interactions between smoking and human leukocyte antigen (HLA)-containing shared epitope alleles (HLA-DR SE)[47]. This gene– smoking interaction appears to markedly influences the development of anticitrulline antibody (ACPA)–positive RA[48]. This interaction, however, cannot be demonstrated in patients with ACPA–negative RA. Notably, antibodies against citrullinated peptides are found in sera from patients with aggressive PD without arthritis[49], although a gene-environment interaction between HLA-DR SE genes and chronic PD has not been reported.

Little is known about the effects of smoking cessation on future risk of RA. A recent study in women[50] showed that the risk of RA decreases over time after smoking cessation. However, when former smokers were compared to never smokers, the risk was still statistically significantly higher. Interestingly, investigators have also been able to show effects of smoking cessation on the gut microbiome[51]. Marked shifts in the microbial composition after smoking cessation could be demonstrated with an increase of Firmicutes and Actinobacteria and lower proportions of Bacteroidetes and Proteobacteria. In addition, an increase in microbial diversity could be observed after smoking cessation. This could explain, at least partially, why smoking exacerbates Crohn’s disease. Whether these shifts in the gut microbiome after smoking cessation translates into decreased risk for RA remains to be established.

The Microbiome in Periodontal Disease

A classical study in the field of periodontology looked at bacterial complexes involved in PD, based on microbial analyses of 185 subjects in over 13,000 subgingival plaque samples[52]. The results showed unequivocally that higher levels of P. gingivalis, T. forsythia and T. denticola in dental plaque are strongly associated with periodontal disease. Since then, this group of microorganisms has been termed the ‘red complex bacteria’ and numerous studies have replicated these findings.

The subgingival microbiome of diseased and healthy sites were also evaluated by 16S rRNA sequencing in PD and controls[15]. The relative abundance of 51 of 170 genera and 200 of 746 species were found to be significantly different between healthy and diseased sites. Within the deep sites, several species implicated in the original PD studies exhibited significant increases. This was particularly true for P. gingivalis, Porphyromonas endodontalis, Fusobacterium nucleatum, Prevotella nigrescens, Treponema denticola, Treponema medium and Tannerella forsythia.

Periodontal Disease, Subgingival Microbiota, P. gingivalis and Rheumatoid Arthritis

The concept that PD is implicated in RA pathogenesis is not original. In the early days of the 20^th century, the ‘oral sepsis’ hypothesis was proposed to involve apical infections of the teeth in the etiology of RA[53]. This led to the use of dental extraction as a therapeutic option, which went on for at least four decades until it was eventually deemed not clinically beneficial. A more refined and complex hypothesis was suggested a decade ago by Rosenstein and Weissman[54], who proposed that it was actually the humoral immune response to P. ginigivalis what provided a stimulus for the development of RA. Multiple recent lines of investigation have supported this relationship between PD, P. gingivalis, ACPA and RA[55]. First, several studies have shown an increased prevalence of PD in RA patients, even at the onset of disease[56–59], arguing for either a causal relationship and/or a common etiopathogenic pathway. Second, it has been postulated that ACPA generation is the result of citrullination of arginine residues in human tissues by the enzyme Peptidyl Arginine Deiminase (PAD). This is of particular importance because P. ginigivalis is reportedly the only bacterial species that carries PAD as part of its enzymatic machinery. Notably, P. gingivalis-derived PAD is able to citrullinate human peptides[60], while human citrullinated peptides (such as alpha-enolase) are sufficient for triggering inflammatory arthritis in animal models[61]. Moreover, the presence of antibodies to P. gingivalis has also been correlated with autoimmunity (i.e., circulating ACPA) and clinical RA. This is not only true for patients with established RA[62–64], but also in subjects at-risk for disease development (i.e., first-degree relatives)[65]. However, the demonstrated prevalence of such humoral response has been variable, mostly due to the heterogeneity of the assays employed in the different studies[66].

Finally, only a handful of reports looked at the presence of P. gingivalis (or other microorganisms) in subgingival biofilms of patients with RA. With the use of high-throughput, bacterial DNA sequencing, we have shown that about 55% of new-onset, untreated RA patients carried P.gingivalis, twice as many as healthy controls[57]. This difference did not reach statistical significance and was mostly due to the high prevalence of the advanced forms of PD in the RA groups, rather than secondary to arthritis phenotype. These findings suggest that while both PD and the presence of P. gingivalis may promote autoantibody formation in RA, only PD demonstrate an independent relationship with established seropositive RA (both in the new-onset and the chronic phases of the disease). Utilizing low-throughput PCR technology, Mikuls et al. have recently validated these observations in a large case-control study[67]. A smaller study by Wolff B. et al., reported similar clinical associations in early RA patients and a trend towards P.gingivalis increase[68].

Intersetingly, we and others have identified additional bacterial species that correlate with either RA-related autoantibodies (Anaeroglobus geminatus) or early RA (i.e., Prevotella, Leptrotrichia and Tannerella species). This reinforces the notion that RA may require other antigenic triggers in the context of PD, as recently demonstrated by Abdollahi et. al., using a murine model of PD and RA in which both P. gingivalis and Prevotella nigrescens were able to promote subgingival and synovial inflammation, albeit through different immune mechanisms[69].

Periodontal Disease Treatment as Adjuvant for Disease-Modifying Anti-rheuamtic Drugs (DMARDs)

Scaling and root planning remains the mainstay of therapy for PD. It consists of a non-surgical process aimed at eliminating the etiologic agents of PD, namely dental plaque, its products, and calculus formation. If in fact PD is at least partially responsible for the systemic inflammatory process found in RA, then it follows that scaling and root planning should help ameliorate synovitis and disease outcomes. To date, only a few relatively small studies have looked at this hypothesis. One of the first such trials evaluated disease outcomes in 20 subjects with RA on background DMARDs and 20 receiving DMARDs plus anti-TNF treatment[70]. Each of the groups was randomized to either scaling/root planning or no PD treatment. Patients that received PD treatment showed a significant decrease in DAS28, ESR and serum TNF-α levels compared to those without PD therapy. Curiously, anti-TNF therapy also resulted in a significant improvement in clinical PD parameters. These findings have also been corroborated independently in different geographic regions[71–73]. In a recent prospective study on RA patients about to undergo anti-TNF treatment, it was shown that the presence of persistent PD hampered the clinical response to TNF-blockade since only those patients without established periodontitis achieved good RA outcomes[74].

Taken together, these studies support the notion that PD may affect treatment efficacy in patients with RA and that a dual, combination approach may be warranted to potentially increase the clinical response of various immunosuppressive drugs.

Conclusion

The role of tobacco smoking in PD has been established many decades ago. Emerging epidemiologic and mechanistic work is now revealing that smoking is also a risk factor for the development of RA, particularly in the ACPA-positive, SE-positive subset. An association between tobacco, PD, presence of P. gingivalis and RA has been strengthened by multiple lines of research, particularly with the advent of novel bacterial DNA sequencing technologies that bypass the need for classic microbiologic culture methods. Because smoking and PD are readily modifiable contributing factors to RA pathogenesis, it is plausible that a multidisciplinary approach to treatment may hold the key for achieving better outcomes in more patients with RA. Small-scale data supports the potential utility for such approach. A large, randomized, clinical trial will be therefore required to address this highly relevant question. Incorporating periodontal microbiota analyses and RA/PD outcomes will also be necessary in order to better understand the biologic connection between these two seemingly interdependent disease processes.

Key points.

• Genetic risk alleles in rheumatoid arthritis have been extensively studied, although they can only explain a fraction of the total disease variance.

• Environmental factors implicated in RA include smoking and chronic infections.

• Periodontal diseases and the microbes harboured in the dental plaque, particularly porphyromonas ginigivalis and the elicited humoral immune response, appear to have a pathogenic role.

• The oral microbiota and its local and systemic effects in health and disease are being elucidated.

• A combination therapeutic approach using DMARDs and non-surgical PD treatment may improve RA outcomes.

Abbreviations

RA

Rheumatoid Arthritis

PD

periodontal disease

DMARD

disease-modifying anti-rheumatic