Background
Periodontitis is a major cause of tooth loss and one of the most prevalent chronic infectious inflammatory diseases, affecting up to 46% of the adult population [
1,
2]. The disease share several common features, including genetic association to human leukocyte antigen (HLA) DR4 alleles and smoking as environmental risk factors with rheumatoid arthritis (RA), another chronic inflammatory disease [
3,
4]. One of the potential links between these two diseases is the oral pathogen
Porphyromonas gingivalis (
P. gingivalis), Gram-negative anaerobic bacteria associated with periodontitis. This pathogen was proposed to possibly contribute to the generation of citrullinated antigens and the production of citrullinated protein antibodies (ACPAs), a hallmark of RA. Since then, it has been shown that antibodies to
P. gingivalis are more common in serum from patients with RA compared to matched controls [
5,
6]. In addition, the presence of anti-
P. gingivalis antibodies has also been associated with ACPAs in individuals at increased risk of RA [
7]. Furthermore, the disease-specific ACPAs [
8], purified from RA serum, may cross-react with citrullinated
P. gingivalis enolase [
9] due to the unique property of
P. gingivalis to express citrullinating enzymes, named
P. gingivalis-derived peptidylarginine deiminase (PPAD), which may, together with host’s endogenous PADs, contribute to local citrullination [
4,
10,
11]. Increased expression of human PAD2 and PAD4 enzymes has been observed in gingival tissue of patients with periodontitis and, both human PAD and bacterial enzyme PPAD activities have been reported to be elevated in both RA and non-RA patients with periodontitis [
12,
13].
Aggregatibacter actinomycetemcomitans (
A. actinomycetemcomitans), another Gram-negative periodontal bacteria, has long been associated with chronic and aggressive periodontitis [
14], and has also been detected in gingival crevicular fluid of patients with RA [
15]. Moreover, recently, the pore-forming leukotoxin (LtxA), secreted by
A. actinomycetemcomitans, was shown to be capable of triggering the activation of endogenous PAD in neutrophils through the disruption of the cell membrane and increased cellular calcium influx [
15], suggesting that this pathogen could be another candidate bacteria for gingival citrullination triggering autoimmunity in RA.
Citrullination of proteins is a posttranslational conversion of peptidylarginine to peptidylcitrulline by PAD enzymes that occurs naturally in different physiological processes, as well as in numerous pathological processes, including inflammation and autoimmunity. The periodontitis-associated pathogens
P. gingivalis and
A. actinomycetemcomitans have been reported to be involved in citrullination [
11,
13,
15] but not their effects on citrullination and expression of endogenous PAD enzymes. Therefore, the aim of this study was to investigate the presence of citrullinated proteins and expression of endogenous PADs (PAD2 and PAD4) in relation to the periodontal pathogens
P. gingivalis and
A. actinomycetemcomitans leukotoxin in periodontal tissue of individuals with and without periodontitis as well as in that of patients with RA.
Discussion
Several studies have shown clinical and epidemiological associations between periodontitis and RA, although the strength, temporal relationship and biological explanations to these associations remain unclear [
26,
27]. The main finding in the present study was that both citrullination and expression of endogenous PADs (PAD2 and PAD4), both at mRNA and protein level, are increased in gingival connective tissue of patients with periodontitis compared to periodontally healthy controls independently of the presence of the periodontal pathogen
P.
gingivalis or leukotoxin of
A. actinomycetemcomitans.
Anti-citrullinated protein antibodies are present in about two-thirds of all patients with RA, but rare in the non-RA population [
28]. It has been suggested that immunity towards citrullinated proteins may be triggered in genetically predisposed individuals by an increased expression of citrullinated proteins in the inflamed sites of the body, for example in the lungs of smokers or the periodontium of patients with periodontitis [
12,
29‐
31]. This immunity may subsequently contribute to the development of chronic inflammatory processes in the joint, where citrullination is also present [
22,
31‐
33]. In case of periodontitis, it has been established that periodontal pathogens initiate a local host response in the periodontal pocket, involving recruitment of inflammatory cells, with increased apoptosis and necrosis of neutrophils in the gingival connective tissue, which may contribute to deamination of arginine residues and hypercitrullination of proteins [
34]. In agreement with this, our results showed increased levels of citrullination in gingival connective tissue of patients with periodontitis compared to periodontally healthy controls. On the contrary, in the gingival epithelium, where citrullination occurs as a physiological process [
12], the levels of citrullination did not differ between the periodontitis and non-periodontitis groups.
It has been shown that bacterial PAD enzymes, expressed by the periodontitis-associated bacteria
P.
gingivalis, is capable of citrullination of endogenous bacterial, as well as human proteins, suggesting a role for this pathogen in this aetiological model [
25,
35]. Apart from
P. gingivalis, another periodontal bacteria,
A. actinomycetemcomitans, has also been suggested as a potential trigger of gingival citrullination through the actions of a pore-forming leukotoxins [
15]. Hence, in order to elucidate the involvement of these oral pathogens in generating citrullinated proteins in gingival tissue we have investigated citrullination and expression of human PADs, in relation to the presence of
P. gingivalis and leukotoxins of
A. actinomycetemcomitans and
M. haemolytica, in gingival tissue of patients with periodontitis and periodontally healthy controls. In the current study, the presence of
P.
gingivalis was comparable in gingival tissue samples from patients with periodontitis and periodontally healthy controls, while increased protein citrullination and human PAD expression were observed in periodontitis-affected tissue, which is in line with previous studies [
12,
30,
36,
37]. However, we could not observe a correlation between the presence of
P. gingivalis and citrullinated proteins, or the expression of endogenous PADs (i.e. PAD2 and PAD4), suggesting that the increased citrullination, observed in this limited number of patients, is caused by the actions of host PAD enzymes, rather than
P. gingivalis-produced PAD. Notably, increased expression of human PAD enzymes has been reported previously at the site of inflammation including gingival tissue and is likely to contribute to the increased levels of citrullination [
12,
30,
36,
37]. Our study was not designed to determine whether citrullination in the gingival tissue of patients with periodontitis was caused by human PAD enzymes or enzymatic activity of PPADs was involved. Moreover, Laugisch et al. [
13] have shown that PPAD activity is higher in gingival crevicular fluid samples of both RA and non-RA patients with periodontitis compared to those without periodontitis.
The periodontal bacteria
A. actinomycetemcomitans has been shown to induce hypercitrullination in neutrophils by activating human PAD enzymes through the actions of leukotoxins [
15]. In the current study the presence of leukotoxins was detected in gingival tissue from patients with periodontitis. However, there was no association between the presence of leukotoxins and the expression of citrullinated proteins or human PAD enzymes in gingival biopsies from patients with periodontitis. Thus, our results may suggest that the increased citrullination in the gingival connective tissue in periodontitis is independent of the presence of the periodontal pathogen
P. gingivalis and leukotoxin of
A. actinomycetemcomitans. One possible explanation for this might be that the progression of the polymicrobial disease periodontitis is dependent not only on the quantity of the periodontal pathogens but also on quality, local host response as well as other risk factors such as smoking and genetic predisposition [
38]. Notably, the gram-negative anaerobic periodontal pathogens
P. gingivalis and
A. actinomycetemcomitans are commensal bacteria commonly present in gingival tissue, as also observed in our study. The pathogenicity of periodontitis is initiated via colonization of “keystone pathogens”, such as
P. gingivalis—which even in low numbers—can increase the virulence of the entire community by communicating with other commensal organisms, promoting dysbiosis, the transition to pathogenicity and, thereby, indirectly ensuing the inflammation [
34,
39,
40]. Another explanation might be that different strains of
P. gingivalis are known to differ in their virulence factors [
41], including PAD activity [
42] and the aetiological model linking
P. gingivalis to ACPA-positive RA depends on the bacteria’s ability to citrullinate, which we have not investigated in our study. In addition, other bacteria may also be involved, as the oral microbiome
Cryptobacterium curtum was recently suggested to be involved in the production of autoantigenic citrullinated peptides in RA [
43]. Furthermore, the
P. gingivalis- and
A. actinomycetemcomitans-independent increase of citrullination and PAD expression may also be a consequence of cellular hypoxia, which has been associated with inflammatory disorders such as periodontitis and RA [
44,
45]. This anaerobic microenvironment promotes citrullination and expression of PAD through activation of hypoxia-inducible factor-1α (HIF1α) [
46], a proinflammatory key transcription factor [
45], reported to be induced by hypoxia in human synoviocytes and in gingival fibroblasts [
44,
46].
Our study mainly investigated non-RA study individuals, which may differ from RA patients with regard to oral pathogens, PAD enzymes, and the oral “citrullinome”. Unfortunately, the number of RA patients in our study (n = 4) was too low to make any relevant comparisons, therefore future studies with larger sample size should address potential differences between systemically healthy patients with periodontitis and RA patients with periodontits. Importantly, increased anti-
P. gingivalis antibody levels have been reported in RA (in particular ACPA-positive RA), compared to systemically healthy patients with periodontitis [
47], and increased anti-LtxA antibody levels have been described in RA, in relation to ACPAs and HLA-DR4 alleles [
15]. Thus, the possibility that the co-occurrence of RA and periodontitis might have a common denominator associated with citrullination and anti-citrulline immunity deserves a series of additional studies, directed toward understanding the specificity and the pathogenicity of citrullination and anti-citrulline immunity, in periodontitis and RA. Nevertheless, this is to our knowledge the first study investigating the presence of citrullinated proteins and expression of PAD2 and PAD4 in relation to the oral pathogens
P. gingivalis and leukotoxin of
A. actinomycetemcomitans, in gingival tissue biopsies, however the results should be confirmed in additional studies.
Authors’ contributions
ME, LL, KE and TYL performed immunohistochemistry experiments and NG and TYL performed RT-qPCR analyses. ME, TYL, AIC, KE and LL performed scoring and evaluation of biopsies. MH, APN and AJ—contributed with production of antibodies against P. gingivalis, citrullinated proteins and A. actinomycetemcomitans LtxA, respectively. TYL, AIC, LK, ME, MH, NG and KL—designed and/or contributed the research study. TYL, KE, LL and NG—performed experimental and statistical analysis. TYL, ME, KE, LL—wrote the manuscript and all authors reviewed the manuscript. All authors read and approved the final manuscript.