Prevalence of Porphyromonas gingivalis Four rag Locus Genotypes in Patients of Orthodontic Gingivitis and Periodontitis

Porphyromonas gingivalis is considered as a major etiological agent in periodontal diseases and implied to result in gingival inflammation under orthodontic appliance. rag locus is a pathogenicity island found in Porphyromonas gingivalis. Four rag locus variants are different in pathogenicity of Porphyromonas gingivalis. Moreover, there are different racial and geographic differences in distribution of rag locus genotypes. In this study, we assessed the prevalence of Porphyromonas gingivalis and rag locus genotypes in 102 gingival crevicular fluid samples from 57 cases of gingivitis patients with orthodontic appliances, 25 cases of periodontitis patients and 20 cases of periodontally healthy people through a 16S rRNA-based PCR and a multiplex PCR. The correlations between Porphyromona.gingivalis/rag locus and clinical indices were analyzed. The prevalence of Porphyromonas gingivalis and rag locus genes in periodontitis group was the highest among three groups and higher in orthodontic gingivitis than healthy people (p<0.01). An obviously positive correlation was observed between the prevalence of Porphyromonas gingivalis/rag locus and gingival index. rag-3 and rag-4 were the predominant genotypes in the patients of orthodontic gingivitis and mild-to-moderate periodontitis in Shandong. Porphyromonas.gingivalis carrying rag-1 has the strong virulence and could be associated with severe periodontitis.


Introduction
Malocclusion is one of the most common oral-maxillofacial diseases that bring some negative effects on facial aesthetics, oral physical function and health as well. In China, morbidity of malocclusion in teenagers is as high as 67.82% [1]. Orthodontic treatment is currently the preferred and most common method for the reason of solving above the problems, but it also holds some potentials of harming teeth and periodontal tissues due to plaque accumulation and gingival inflammation that are induced by the changes of oral internal environment after wearing fixed orthodontic appliance, therefore lead to changing of host physiology and the composition of the oral microflora [2,3]. The inflammatory reaction of gingival tissue can often be detected in patients wearing fixed orthodontic appliances. The overall morbidity of gingivitis was higher as 56.8% and 34.4% in adolescent group and adult group respectively during fixed orthodontic treatment in China [4]. Primary pathogenic microorganisms strongly implicated in gingival inflammation and posterior periodontal destruction, such as Porphyromonas gingivalis, Prevotella intermedia/nigrescens, Aggregatibacter actinomycetemcomitans, Tannerella forsythia, Treponema denticola, and Fusobacterium species have been found elevated in patients after bracket placement [5][6][7]. Our previous research showed that the percentage of bacilli, esp. Porphyromonas gingivalis, Aggregatibacter actinomycetemcomitans, Fusobacterium nucleatum increased significantly after wearing ortho-dontic appliance and the increase of those pathogens was significantly related with the development of gingivitis in orthodontic treatment [8][9].
Porphyromonas gingivalis is a gram-negative oral anaerobe and considered as a major etiological agent in periodontal diseases by producing a number of virulence factors and extracellular proteases, such as lipopolysaccharide, capsule, gingipain, fimbria and so on, resulting in the destruction of periodontal tissues [10][11][12][13]. The pathogenicity of P.gingivalis has been investigated in a variety of experimental animal models, such as rat [14][15], mouse [16][17], rabbit [18], drosophila [19], and cell models [20][21][22][23], showing complicated mechanisms of P.gingivalis-host interactions in development of periodontal diseases. Three gingipains referred to be the important virulence factors, Arg-x-specific proteinase and adhesins (RgpA), Arg-x-specific proteinase (RgpB), and a Lys-xspecific proteinase and adhesins (Kgp) have been well known and studied in details with properties of activating and/or degrading a wide range of host proteins through different mechanisms [20][21][22][23].
Pathogenicity island is a large unstable chromosome DNA region encoding virulence determinants of pathogenic bacteria and was first described in human pathogens of the species Escherichia coli by Hacker et al. [24]. The island has been also detected in other pathogens, such as Klebsiella pneumoniae, Enterobacter aerogenes, and Citrobacter koseri isolates with high conservative property among these species and is in association with the yersiniabactin determinant [25]. In 1999, Curtis et al [26] found a novel pathogenicity island in a proportion of P.gingivalis strains named rag locus and it was more frequently detected in deep periodontal pockets in periodontal patients. It was reported that the rag locus of P.gingivalis might arise from Bacteriodes via horizontal gene transfer and encodes RagA and RagB. RagA is a 115-kDa TonB-dependent outer membrane receptor, and RagB is a 55-kDa lipoprotein constituting an immunodominant outer antigen. Both proteins of RagA and RagB constitute a membrane transporter system. Further study demonstrated that four rag locus variants with different pathogenicity were detected from clinical isolates of P.gingivalis [26][27][28]. A significant correlation was observed between prevalence of rag-1 allele and a highly virulent phenotype in a murine model of soft tissue destruction [29]. In addition, there are different racial and geographic differences in distribution of rag locus genotypes [28]. Wang et al. [30] investigated the distribution of rag genotypes in chronic periodontitis patients in Northeast of China and found P.gingivalis carrying rag-1, rag-3 was more predominant in chronic periodontitis so that might be associated with the development of pediodontitis.
There have been reports about the association between prevalence of P.gingivalis and gingival inflammation during orthodontic treatment [31][32][33][34][35]. Our previous research showed that the prevalence of P. gingivalis was totally higher as 40.62% two months after orthodontics detected by using traditional anaerobic culture [9]. While there have been no investigations about the correlation between P.gingivalis rag locus and periodontal health status in orthodontic gingivitis patients. Therefore, we assessed the prevalence of P.gingivalis and rag locus genotypes in gingival crevicular fluid samples from the gingivitis lesions of orthodontic patients and compared them with periodontitis patients as well as periodontally healthy people who showed healthy periodontal tissues before wearing orthodontic appliances.

Subjects
The study subjects consisted of three groups who visited Jinan stomatological hospital for orthodontic or periodontitis treatment from 2010 to 2011. Of three groups, orthodontic group (OG) included 57 patients, 38 females and 19 males, aged between 10 and 30 years (mean 16.3) who got gingival inflammation during orthodontic treatment; control group (CG) contained 20 periodontally healthy people, 12 females and 8 males, aged between 10 and 30 years (mean 16.05) before orthodontic treatment; periodontitis group (PG) was composed of 25 periodontitis patients, 10 females and 15 males, aged from 20 to 60 years (mean 25) who came to hospital for periodontitis treatment. The patients who are having any systemic diseases, antibiotics therapy within the last 3 months and pregnant or lactating females were excluded.

Ethics statement
This work was approved by the Medical Ethics Committee of the Jinan Stomatological Hospital. All patients or their parents gave their verbal followed by written informed consent before the examination was performed. The relevant regulations and institutional polices were followed strictly.

Bacteria strains
The reference strains of P.gingivalis ATCC33277, F.nucleatum ATCC25586 and A.actinomycetemcomitans ATCC29522 were from the West-China Dental School of Sichun University and P.gingivalis

Clinical examination
We selected the gingivitis or periodontal sites that exhibited the deepest pocket depth of every subject. The clinical parameters included gingival index (GI), plaque index (PI), sulcus bleeding index (SBI) and probing depth (PD) of each person were examined and recorded. All clinical examinations were carried out by the same dentist.

Sample collection and DNA extraction
Gingival crevicular fluid (GCF) was obtained from the two deepest periodontal pockets in the maxilla according to Rüdin et al [36]. In brief, before collecting, saline solution was used to rinse out food debris and then each site was cleaned by cotton rolls. Sterile filter paper strips were placed for 30 seconds into the packet until a minimum of resistance. The paper points were placed into a sterile microcentrifuge tube containing 0.5 ml of 16PBS immediately. The tubes were mixed thoroughly and stored at 220uC until analyzed. The bacterial DNA was extracted by the boiling method [37]. In short, a 10 ml aliquot of each stored sample was added to 10 ml of 2 6lysis buffer (2 mM EDTA, 1% X-100). The mixture was boiled for 10 minutes and then placed on ice. The supernatant was used as the template for PCR amplification.

The 16S rRNA-based PCR and multiplex PCR amplification
The 16S rRNA gene specific primers were used to determine the prevalence of P.gingivalis in GCF, while four different rag locus variants primers were utilized to amplify the rag locus variants genes from GCF samples containing P.gingivalis. The 16S rRNAbased PCR was first performed on DNA extracts from GCF samples using primers of 16S rRNA-F (59-AGG CAG CTT GCC ATA CTG CG-39) and 16S rRNA-R (59-ACT GTT AGC AAC TAC CGA TGT-39) that amplify a 404-bp region of the 16S rRNA gene [38]. The specificity of this PCR was confirmed by sequencing and amplifying P.gingivalis ATCC33277, W83, as well as unrelated pathogens F.nucleatum ATCC25586 and A.actinomycetemcomitans ATCC29522. Then the multiplex PCR was utilized to amplify rag locus genes from the positive P.gingivalis samples.

Statistical analysis
The differences in the prevalence of rag locus genes were analyzed using the Chi square test. The Spearman Correlation Test was utilized to analyze the correlation between prevalence of P.gingivalis/rag locus genes and clinical indices in three research groups. All statistical analyses were done by using a statistical software package (SPSS for Windows 13.0).

Detection and confirmation of 16S rRNA-based PCR for P.gingivalis
The reference stains were first amplified by the 16S rRNAbased PCR to evaluate the specificity of it. The positive product appeared only from P.gingivalis ATCC33277 and W83, not from F.nucleatum ATCC25586 and A.actinomycetemcomitans ATCC29522.
Sixty-five P.gingivalis was detected in 65 (63.73%) cases of GCF samples from 102 cases of three groups, thirty-five (61.40%) from orthodontic group (OG), Seven (35%) from control group (CG), and 23 (92%) from periodontitis group (PG). Prevalence of  P.gingivalis was found significantly different in three groups: P.gingivalis was the highest prevalence in PG (P,0.01) and higher level in OG than CG (P,0.05) ( Table 2). Of 65 P.gingivalis positive samples, 10 were randomly sequenced by a 3730 DNA sequencer in Invitrogen Company (Invitrogen, Shanghai) and confirmed the validity of the 16S rRNA-based PCR for clinical GCF samples (date were not showed). The correlation of patients' age and prevalence of P.gingivalis was analyzed. The average age of patients with positive P.gingivalis in GCF was 25.54 years, while patients with negative P.gingivalis was 16.19 years, and there was statistical difference between the patients' ages of both positive and negative P.gingivalis (P,0.05).

Multiplex-PCR amplification of rag locus genes
Multiplex-PCR was firstly used to detect rag locus genes from the high virulent P.gingivalis W83 and low virulent P.gingivalis ATCC33277. It showed that rag-1 gene was amplified from P.gingivalis W83 and rag-4 was from ATCC33277, which were consistent with previous documents [40,41].
There were 52 (80%) positive rag locus genes detected from the 65 GCF samples which were P.gingivalis positive, twenty-nine (82.86%) from those in OG, 22 (95.65%) from those in PG, only one case (14.29%) from healthy people. The prevalence of rag locus was significantly higher in PG and OG than in CG (P,0.01), while no significant difference between PG and OG (P.0.05) (Fig. 1).

Correlation of rag locus genes and clinical indices
It was noticed that clinical indices were all higher in OG and PG than CG (P,0.05). The prevalence of rag locus genes, except rag-2, elevated directly with increases of GI values in both OG and PG. The significant positive correlation between rag locus and GI was showed by using Spearman Correlation Test (P,0.01) (Fig. 2). However, there were no statistical difference between rag locus and PD/PI/SBI, while we detected the increase of rag locus in OG and PG with higher GI and higher PD/PI/SBI. rag-1 often appeared from the deeper periodontal pocket with higher PD/PI/SBI values Figure 2. Correlation of rag locus genes and gingival indices (GI).** P,0.01 between GI 2 and GI 0; * P,0.05 between GI 1 and GI 0, # P,0.05 between GI 2 and GI 1 in PG (Chi-squared test). doi:10.1371/journal.pone.0061028.g002 Table 3. Prevalence of rag locus under different gingival index (GI) among three groups.
Groups GI (n) rag locus(%) rag-1 (%) rag-2 (%) rag-3 (%) rag-4 (%) and 10/14 of rag-1 positive cases accompanied with rag-3 and/or rag-4. Interestingly, an exception was found that one case of rag-1 gene was from a patient of GI 0 lever in orthodontic group, so further research would be done for the possible variation of rag-1 gene. The prevalence of four rag locus genes among three groups showed in table 3.

Discussion
P.gingivalis has been known to be a risk factor for periodontal diseases though the exact roles of it in the initiation and progression of the oral diseases remain unclear. Mouse model tests have indicated difference in the virulence of P.gingivalis with and without rag locus [28,29]. Shi et al. mutated rag locus genes in P.gingivalis by using an allele replacement strategy and clearly showed that inactivation of the rag locus reduced the virulence of P.gingivalis in a mouse model of soft tissue destruction [29]. In a collection of 168 isolates of P.gingivalis from western European countries, including the Netherlands, Romania, Sweden, the United Kindom, Kenya and Germany, Hall et al. found different prevalence and geographic differences in distribution of four rag alleles [28]. In this case, we detected prevalence of P.gingivalis and rag locus genotypes in local patients of orthodontic gingivitis, periodontitis and periodontally healthy people to evaluate the distribution of P.gingivalis and predominant genotypes of rag locus in different periodontal health statuses, then further deduce the pathogenicity of P.gingivalis carrying different rag locus during orthodontic treatment.
In periodontal disease, gingival crevicular fluid (GCF) is an inflammatory exudate. GCF contains substances from supra-and subgingival located bacteria. Analysis of microflora in GCF becomes more and more important in diagnosis and therapy of periodontal diseases. There are a large number of periodontopathic bacteria including P.gingivalis in GCF [42]. Considering the method of filter paper strips is recommended to collect the microflora from GCF for microbiological analysis in dental practice [35]. With sterile filter paper strips, we collected GCF samples divided into orthodontic gingivitis, periodontally healthy control and periodontitis groups from 102 patients to investigate the prevalence of P.gingivalis in above three groups. The occurrence of P.gingivalis was 61.40% in OG, 35% in CG, and 92% in PG, respectively. There was a statistically higher prevalence of P.gingivalis in PG followed by OG than CG. Furthermore, a significantly strong positive correlation was observed between the prevalence of P.gingivalis and GI by Spearman Correlation Test (P,0.01), which was consistent with previous reports [31][32][33]43]. While once wearing fixed appliance, oral hygiene will turns bad if teeth cleaning can not be paid special attention, as a result, dental plaque will accumulate and gingival inflammation will happen. By then, one side, anaerobic environment will be created due to swollen gum, deeper gingival sulcus, and pseudo periodontal pocket; on the other side, the gum will be susceptible to bleeding and therefore it will be more conductive for periodontal anaerobic P.gingivalis to survive. P.gingivalis may play a similar role in orthodontic gingivitis and periodontitis.
P.gingivalis has been reported to be related with adult periodontitis [10,12]. In this study, we analyzed correlation of patients' age and occurrence of P.gingivalis and found the age of both P.gingivalis positive and negative was statistically different; implying the prevalence of P.gingivalis may increase as patients'age increased.
In order to further explore whether the P.gingivalis rag locus was associated with gingival inflammation under orthodontic appliance,we detected the distribution of rag locus in three groups and discovered 52 (80%) of positive rag locus genes from 65 P.gingivalispositive GCFs. The prevalence of rag locus was higher in those of PG and OG than those of CG. The P.gingivalis without rag locus was mostly detected from periodontally healthy control and GI 0 level OG patients, demonstrating that they present the avirulent or weak virulence genotype of P.gingivalis. A clear positive correlation was indicated between the gingival index and rag locus genes, implying rag locus genes may play a pathogenic and similar role in the development of gum inflammation during orthodontic in comparison with periodontitis.
The prevalence of rag-3 (27 cases) was the most detected followed by rag-4 (20 cases) and rag-1 (14 cases); the lowest occurrence was rag-2 (10 cases) with lower GI and PD/PI/SBI values, showing the rag-3 and rag-4 locus genes might be the predominant genotypes in the patients of orthodontic gingivitis and mild-to-moderate periodontitis in the populations of Shandong region. Besides, we found rag-1 was detected from 14 cases, mostly with higher GI and PD/PI/SBI, and often combined with rag-3 and/or rag-4, suggesting the P.gingivalis carrying rag-1 is the strong virulent genotype and can be closely associated with severe periodontitis, which is consistent with Hanley et al [44].
In summary, P.gingivalis carrying rag-3, rag-4 locus is one of the risk factors that are responsible for gingivitis during orthodontic treatment. Thus monitoring P.gingivalis is highly recommended following the placement of orthodontic appliances. In addition, appropriate oral hygiene is necessary to reduce invasion of pathogens and exerts a beneficial effect to oral tissues.

Author Contributions
Conceived and designed the experiments: YL SX. Performed the experiments: YZ LW YG. Analyzed the data: YZ SX. Contributed reagents/materials/analysis tools: YZ SX. Wrote the paper: YL SX.