Identification of Veillonella Species in the Tongue Biofilm by Using a Novel One-Step Polymerase Chain Reaction Method

Six Veillonella species have been frequently isolated from human oral cavities including infectious sites. Recently, it was reported that diet, smoking, and possibly socioeconomic status can influence the bacterial profile in oral cavities. In addition, oral hygiene habits may also influence oral microbiota in terms of both numbers and diversity of microorganisms. In this study, the identification of Veillonella species in tongue biofilms of Thai children, divided into three groups dependent on their status of oral hygiene. For this, we used a novel one-step PCR method with species-specific primer sets based on sequences of the rpoB gene. As shown in the results, the number of isolates of Veillonella species was 101 strains from only 10 of 89 subjects. However, the total number of bacteria was high for all subjects. Since it was reported in previous studies that Veillonella species were easy to isolate in human tongue biofilms at high numbers, the results obtained in this study may suggest country- or age-specific differences. Moreover, Veillonella species were detected predominantly in subjects who had poor oral hygiene compared to those with good or moderate oral hygiene. From these results, there is a possibility that Veillonella species may be an index of oral hygiene status. Furthermore, V. rogosae was a predominant species in tongue biofilms of Thai children, whereas V. parvula and V. denticariosi were not isolated at all. These characteristics of the distribution and frequency of Veillonella species are similar to those reported in previous studies. Although further studies are needed in other countries, in this study, a successful novel one-step PCR method was established to detect Veillonella species in human oral cavities easily and effectively. Furthermore, this is the first report investigating the distribution and frequency of Veillonella species in tongue biofilms of Thai children.


Introduction
In this study, the distribution and frequency of Veillonella species in tongue biofilms of children in Thailand, in association with the level of oral hygiene, were examined. We then compared the results to previous reports of our group and others regarding the identification of Veillonella species in tongue biofilm. Moreover, the effective novel one-step PCR method using species-specific primer sets designed from the rpo B gene (encoding the β subunit of bacterial RNA polymerase) was established in this study.

Ethics Statement
This study received approval from the Ethics Committee, Mahidol University, Bangkok, Thailand under process number MU-DT/PY-IRB 2015/DT028, and samples were collected in March 2015. The participants and their parents were made aware of the objectives and procedures of the study and agreed to participate by providing written, informed consent.

Subjects
The 89 children selected consisted of 41 males and 48 females (age range: 7 to 15 years). Children with a history of immunosuppression or systemic diseases (e.g. diabetes and HIV), those using medications that reduce saliva flow, and those under treatment with antimicrobials in the previous three months, were excluded from the study. The participants were evaluated by the Simplified Oral Hygiene Index (OHI-S) according to the criteria of Greene &Vermillion [38] and divided into three groups. The first group (Good oral hygiene) was composed of 29 children (10 males and 19 females) with OHI-S scores of 0-1.

Sample Collection
Tongue biofilm samples were collected at Dental Hospital, Mahidol University Faculty of Dentistry using sterile cotton wool swabs to swab the tongue 5 times each per subject. Samples in reduced transport fluids were transported in an anaerobic box (HIRASAWA WORKS Inc.) containing 80% N 2 , 10% CO 2 , and 10% H 2 . The samples were immediately placed in 1 mL of sterile saline (< 1 hour from the time of collection). Samples were homogenized for 3 min with a BioMasher 1 II(Nippi, Incorporated Protein Engineering Office, Tokyo, Japan) to disperse the biofilm and were serially diluted 10-fold with sterile saline from 10 −3 to 10 −8 .
The total number of bacteria in the samples was determined by counting the total number of colonies on BHI agar, while the number of Veillonella was determined by counting the total number of typical Veillonella colonies on Veillonella agar. Bacterial cells of typical Veillonella colonies were confirmed by light microscopy after Gram staining.

Bacterial Strains
V. atypica ATCC 17744 T , V. denticariosi JCM 15641 T , V. dispar ATCC 17748 T , V. parvula ATCC 10790 T , V. rogosae JCM 15642 T and V. tobetsuensis ATCC BAA-2400 (= JCM 17976 T ) were used as control oral Veillonella to confirm the specificity of the primer sets. They were cultured on BHI agar at 37°C in the anaerobic box for 5 days.

DNA Extraction
Genomic DNA was extracted from individual bacterial cells using an InstaGene Matrix Kit (Bio-Rad) and the DNA concentration was determined based on fluorescence using a Qubit 1 3.0 Fluorometer (Invitrogen life technologies) according to the manufacturer's instructions.

Design of Primer Sets
According to the DNA sequence of a conserved region of the partial sequence of the rpo B gene, species-specific primer sets for six Veillonella species were designed by the standard manual method [40]. The accession numbers of the rpo B gene sequences of reference strains from the GenBank database (Database ID: BA123456) were EF185159 for V. atypica, EF185162 for V. denticariosi, EF185161 for V. dispar, EF185158 for V. parvula, EF211831 for V. rogosae, and AB698646 for V. tobetsuensis. Pairwise similarities among six Veillonella species were estimated by MEGALINE including CLUSTAL W in the LASERGENE program (DNASTAR).

PCR Protocol
PCR at the species level was performed using 1 μL of template DNA, 1 μL of each primer (10 pmol/mL), 17 μL of PCR grade water, and 25 μL of master mix from an AmpliTaq Gold 1 360 Master Mix (Applied Biosystems). When using specific primer sets for six Veillonella species, PCR mixtures were subjected to preheating at 94°C for 15 min; followed by 30 cycles of 92°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 30 s; with a final extension at 72°C for 5 min. The PCR products were applied to a 3.0% agarose gel. In the case of PCR at the genus level, the PCR protocol was performed in accordance with the protocols described by Arif et al. and Beighton et al. [6,41]. After electrophoresis, the gel was stained with SYBR 1 Safe DNA gel stain (Invitrogen™).

Clonal Analysis of Unknown Strains
Unknown strains of genus Veillonella (PCR reaction was positive with genus-specific primers, but was negative with species-specific primer sets in this study) were examined clonal analysis. The representative 11 strains in the 40 unknown strains were chosen in this study.
DNA was extracted from the individual bacterial cells isolated on Veillonella agar by using the InstaGene Matrix kit (Bio-Rad), according to the manufacturer's instructions. PCR-based amplification and partial sequence analyses of rpoB were performed using previously described specific primers for genus Veilloenlla [41,42]. The sequence determined with an ABI PRISM 310 Genetic Analyzer were aligned with each other and connected by using SEQMAN IIof the LASERGENE program (DNASTAR). The programs MEGALIGN including CLUSTALW and NJPlot were used to compare sequences and the reconstruct the evolutionary tree by the neighbour-joining method. Also, confidence intervals were assessed by CLUSTAL W with bootstrap analysis. In particular, pairwise similarity values were determined with MEGALIGN in the LASERGENE program. The rpoB (559 nt) partial sequences of 11 strains were aligned against the sequences of the representative strains retrieved from GenBank.

Species-specific Primer Sets for Oral Veillonella
A similarity search of the rpo B gene of six Veillonella species revealed a high degree of polymorphism in the region of position 2500 to 3100 in all oral Veillonella tested. The similarities among partial sequences of the rpo B gene regions of these six Veillonella species were found to be from 73.7 to 90.9% (Table 1).

Specificity and Sensitivity of the Primer Sets
To achieve faster and more convenient identification of oral Veillonella, all six reverse primers (ATYR, DENR, DISR, PARR, ROGR, and TOBR) were examined for use with VF in single PCR mixture. PCR was performed using DNA templates from the type strain of six Veillonella species as described in Materials and Methods. The electrophoretically detected PCR amplicons were shown as species-specific products (Fig 2). The molecular weights of the PCR products were also identical to the theoretical values, which were 396 base pairs (bp) for V. atypica, 594 bp for V. denticariosi, 257 bp for V. dispar, 311 bp for V. parvula, 510 bp for V. rogosae and 447 bp for V. tobetsuensis (Fig 2).
Each of the six PCR products were mixed and applied to the agarose gel to use as a molecular marker (Fig 2) for reference when six Veillonella species were identified from tongue samples, through electrophoresis. These results demonstrated that six Veillonella species could be specifically and effectively identified by PCR using a mixture of primer sets, ATYR, DENR, DISR, PARR, ROGR, TOBR and VF. To examine the sensitivity of these primers, PCR was performed with 0.01pg-10 ng of template DNA from the type strains of six Veillonella species (Fig 3). 1 pg of V. atypica and V. denticariosi DNA, 0.1-0.01 ng of V. dispar DNA, 0.01 ng of V. parvula and V. tobetsuensis DNA, and 0.01 ng-1 pg of V. rogosae DNA could be detected using these primer sets (Fig 3).

Detection of Oral Veillonella at the Species Level from Tongue Samples
Oral Veillonella were detected in only 10 subjects from 89 subjects from all oral hygiene groups as shown in Tables 2-4. However, the tongue biofilm swabs did indeed yield high numbers of bacterial colonies on BHI agar. Mean (SE) colony-forming units (CFU/mL) per swab were 2.2 (0.08) x 10 6 with a median of 9.0 x 10 5 in the first group (Good oral hygiene), Table 2, 5.6 (0.02) x 10 5 with a median of 9.0 x 10 4 in the second group (Moderate oral hygiene), Table 3, and 1.4 (0.04) x 10 6 with a median of 6.5 x 10 5 in the third group (Poor oral hygiene), Table 4. A total of 101 strains were isolated from Veillonella agar in all groups. Veillonella species were predominantly detected in subjects who had poor oral hygiene compared to those with good or moderate oral hygiene (Tables 2-4). Typical Veillonella colonies on the Veillonella agar were 2-4 mm in diameter, regular and slightly domed in shape and were entirely white; they were composed of small, gram-negative coccal cells, mainly existing as single cells but with some short chains visible. The detection limit was dependent on the number of bacteria in the sample and was < 0.1% of the total colony count.
Using species-specific primer sets, 61 of the 101 isolates were identified as either V. atypica, V. dispar, V. rogosae, or V. tobetsuensis (Tables 2-4). Among them, V. rogosae was detected as the predominant species through all groups. However, V. parvula and V. denticariosi were not isolated from any subjects. V. dispar was isolated mainly from subjects who had good or moderate oral hygiene. In addition, of the 101 strains isolated in this study, zero or multiple PCR products were detected using species-specific primer sets with DNA from 40 strains isolated from eight subjects (Tables 2-4). These 40 strains made distinctive PCR products with Veillonella genus primers (Data not shown).

Clonal Analysis of Unknown Strains
The 11 strains formed distinct taxa with robust bootstrap support (100 and 95.9%) in the rpoB tree (Fig 4) within the genus Veillonella. The rpoB partial sequence similarity among the 11 strains was 92.4-100%. The mean rpoB sequence similarity among the 11 strains belonged to V. dispar and the most closely related species was 98.3%.

Discussion
Veillonella species are relatively easily identified at the genus level, but species-level identification of the members of this genus is difficult because of a lack of conventional phenotypic and biochemical tests [5,[42][43][44]. Molecular methods based on 16S ribosomal DNA (rDNA) gene sequencing, including PCR-random fragment length polymorphism analysis, have been used to identify Veillonella strains at the species level [44,45]. However, recent studies have shown that identification of the members of this genus using 16S rDNA gene sequencing is not reliable [6,7,44,46]. To overcome this problem, sequence analysis of housekeeping genes, including  rpoB, dnaK, and gyrB, have been used to define Veillonella species [5][6][7]46]. This approach has enabled clear discrimination between species that are well defined based on DNA-DNA homology but poorly defined based on 16S rDNA. V. denticariosi, V. rogosae, and V. tobetsuensis have been proposed as novel species of oral Veillonella based on the results of the molecular analysis using rpoB or dnaK gene sequencing in conjunction with sequence analysis of the 16S r DNA gene [5][6][7]. Beighton et al. [11] successfully used the rpoB gene sequence, rather than 16S rDNA gene sequence, to discriminate between 253 Veillonella isolates from human Total number of anaerobic bacterial colony counts, total number of Veillonella species counts and the number of isolates from each subject identified using species-specific primer sets. CFU: colony-forming unit; detection limit <0.1% of the total count. Individual species as a percentage of the number of isolates from each subject identified using species-specific primer sets. tongues of healthy adults at the species level. However, it is time consuming to obtain sequence data of many strains, and laborious to identify them at the species level. Igarashi et al. [47] reported the successful design of species-specific primer sets for five species of oral Veillonella (V. atypica, V.denticariosi, V. dispar, V. parvula, and V. rogosae) based on a highly variable region (positions 2500-3100) of the rpoB gene. They distinguished between five species of oral Veillonella at the species level by using PCR products generated by a two-step PCR method with species-specific primer sets. This proved to be easier than using Total number of anaerobic bacterial colony counts, total number of Veillonella species counts and the number of isolates from each subject identified using species-specific primer sets. CFU: colony-forming unit; detection limit <0.1% of the total count. Individual species as a percentage of the number of isolates from each subject identified using species-specific primer sets. doi:10.1371/journal.pone.0157516.t003 Identification of Veillonella spp. by Using Novel PCR Method complete rpoB gene sequences. In addition, the distribution and frequency of five species of oral Veillonella in tongue biofilms of healthy adults were analyzed by PCR with these specificprimer sets as performed in our previous study [12]. Furthermore, species-specific primer pairs for V. tobetsuensis were designed using the sequence of the dnaK gene in the region displaying a high degree of polymorphism (positions 424-1048) [48]. In addition, the distribution and frequency of six species of oral Veillonella in the periodontal pockets were recently clarified by PCR using these primer sets and those of our previous study [21]. Thus, our previous study   demonstrated that these species-specific primer sets were effective in identifying oral Veillonella at the species level, more so than sequence analysis of housekeeping genes. However, these previous procedures used three steps (a two-step PCR with species-specific primer sets for five Veillonella species and PCR with species-specific primer pair for V. tobetsuensis) to define all species of oral Veillonella. This study describes the establishment of the first successful one-step PCR method, using species-specific primer sets for the highly variable region of the rpoB gene of six species of oral Veillonella (positions 2,500-3,100). This one-step PCR method with species-specific primer sets is easier and more effective than previous PCR methods with species-specific primer sets and pairs for oral Veillonella described above. For faster and more convenient identification of the six Veillonella species, our pilot study used a multiplex PCR method and was tested in many conditions, simultaneously using all species-specific primer sets and DNA templates of six Veillonella species in a single PCR mixture. Only the PCR product of VF and DISR, specific for V. dispar, could not be confirmed (Data not shown). Although the reason was not clarified, it was suspected to be due to the sensitivity of DISR. The specific primer set for V. dispar, VF and DISR, showed lower sensitivity than the other sets (Fig 3). Therefore, a one-step PCR method, with species-specific primer sets for six species of oral Veillonella, was developed in this study.
According to the results of Tables 2-4, Veillonella species had a tendency to be detected in subjects who had poor oral hygiene compared to those with good or moderate oral hygiene. However, the CFU count of all bacteria in tongue biofilms had no relationship with oral hygiene status. It was suggested that the appearance of Veillonella species in tongue biofilm of Thai children would be one of the indexes for deteriorating oral hygiene.
In our previous study, the distribution and frequency of oral Veillonella at a species level in the tongue biofilm of Japanese healthy young adults (Twenty-seven subjects: 12 males and 15 females, age: 20s) were investigated [12]. In that study, all subjects had Veillonella species (more than 0.6 x 10 6 CFU/mL), and the number of isolates was 416 from all subjects. Tongue biofilm samples from these subjects could be divided into two groups based on the distribution and frequency of five species of oral Veillonella. In one group, V. rogosae was predominant, while the other group consisted of mainly V. dispar and V. atypica. Beighton et al. [11] also reported the predominant cultivable Veillonella species in tongue biofilm of healthy adults (eleven subjects: the ratio of gender and range of age were not shown) in the UK. They also demonstrated that Veillonella species were isolated from all subjects, and the total number of Veillonella isolates was 253. In addition, they reported that the predominant species in tongue biofilms were V. atypica, V. dispar, and V. rogosae. In this study, only 10 of 89 subjects had Veillonella species, and the total number of isolates was 101 from 10 subjects. The ratio of Veillonella isolates was clearly low compared to these previous reports. As described in the introduction, it was reported that Veillonella species have a central role in biofilm formation as the early colonizer [32]. The results herein may be due to country-or age-specific differences, in which case, other oral bacterial species may have central roles as the early colonizer to form the biofilm and to cause various oral infectious diseases. As the source of these differences was not clarified here (for example, whether it is due to country-or age-specific differences), further studies are needed to investigate the distribution and frequency of Veillonella species from tongue biofilms of children in other countries including Japan.
V. rogosae was also isolated as the predominant species in tongue biofilms in this study. In addition, V. parvula and V. denticariosi were not isolated at all. Similarly, Beighton et al. [11] reported that V. parvula was isolated from only one subject while V. denticariosi was not isolated from any subjects in their study. In our previous study [12], V. parvula was isolated from four subjects and V. denticariosi was isolated from only one subject. Taken together, it appears that the distribution and frequency of Veillonella species in tongue biofilm have generational and universal commonality. In this study, V. tobetsuensis was isolated from only one subject. In addition, in our previous studies [12,48], 12 strains of V. tobetsuensis were isolated from tongue biofilms of 5 of 27 subjects. Here, it was also considered that V. tobetsuensis is a minor species in the tongue biofilm. There are likely to be several factors that influence these differences in distribution and frequency of Veillonella species in tongue biofilms.
PCR products could not be identified in 40 strains, using species-specific primer sets for oral Veillonella species, but these 40 strains produced PCR products with the Veillonella genusspecific primer set used in the present study. This confirms that these 40 strains were members of the genus Veillonella. Although these 40 strains could not be classified as any of the established oral Veillonella, these results suggest the possibility that other Veillonella species also inhabit the human oral cavity.
In addition, according to the results of clonal analysis, although the representative 11 strains in the 40 unknown strains showed similar sequences with V. dispar, they formed distinct taxa with robust bootstrap support (100 and 95.9%) in the rpoB tree (Fig 4). These 40 isolates may be members of novel oral Veillonella species. Furthermore, there are at least two novel Veillonella species in the 40 unknown strains based on the evolutionary tree of rpoB gene (Fig 4). However, the sequence analysis of other housekeeping genes, such as 16S rDNA, dnaK and gyrB genes, are required to propose the novel species of genus Veillonella [5][6][7][44][45][46]. At present time of writing this manuscript, the sequence analysis of other housekeeping genes of these representative 11 unknown strains are in the process.
In this study, a novel one-step PCR method with species-specific primer sets for oral Veillonella was established using a partial sequence of the rpoB gene. Moreover, this is the first report of the use of species-specific primer sets for determining the distribution and frequency of oral Veillonella in tongue biofilm of Thai children. However, as described above, further studies are needed to discuss the country-or age-specific differences in the distribution and frequency of oral Veillonella. The study of the distribution and frequency of oral Veillonella by using the tongue biofilms in other countries including Japan will be investigated in near future.