Identification and phylogenetic analysis of oral Veillonella species isolated from the saliva of Japanese children.

Background: As the most frequent infectious disease among children worldwide, dental caries have a strong relationship with oral hygiene status, specifically in the development of infection. However, the study regarding the identification and distribution of oral Veillonella are limited. The oral Veillonella community may affected by the differences in geographical location, age, diet, lifestyle, socio-economic status and oral hygiene status. Here, we studied the oral hygiene status by examining the composition and proportion of oral Veillonella species in saliva of Japanese children. Methods: Microbial samples collected from 15 Japanese children divided into three oral hygiene groups were cultured under anaerobic conditions after homogenization and dilution, and inoculated onto brain heart infusion and selective medium Veillonella agar. Genomic DNA was extracted from each isolate. Veillonella species were detected by one-step PCR using rpoB species-specific primers. To analyse the phylogenetic properties of the unknown Veillonella strains, PCR amplification and sequence analysis of rpoB were conducted for 10 representative strains. Results: Although V. rogosae was found as the predominant species among all groups, its prevalence was significantly lower in the children with poor oral hygiene than in those with good oral hygiene. V. parvula was the prevalent species in the poor oral hygiene group. Approximately 10% of the isolated Veillonella strains were not classified to any established species; the phylogenetic analysis showed that they were most closely related to V. infantium Conclusions: This study demonstrates that the composition and proportion of oral Veillonella species in the saliva of Japanese children is correlated with different oral hygiene status. Changes in detection ratios of V. parvula and V. rogosae can be useful indicators of oral hygiene status. Furthermore, new strains closely related to V. infantium were isolated from the saliva of Japanese children.

the identification and distribution of oral are limited. The oral Veillonella community may affected by the differences in geographical Veillonella location, age, diet, lifestyle, socio-economic status and oral hygiene status. Here, we studied the oral hygiene status by examining the composition and proportion of oral species in saliva of Japanese children. Veillonella Microbial samples collected from 15 Japanese children divided Methods: into three oral hygiene groups were cultured under anaerobic conditions after homogenization and dilution, and inoculated onto brain heart infusion and selective medium agar. Genomic DNA was extracted from Veillonella each isolate. species were detected by one-step PCR using Veillonella species-specific primers. To analyse the phylogenetic properties of rpoB the unknown strains, PCR amplification and sequence analysis Veillonella of were conducted for 10 representative strains. rpoB Although was found as the predominant species Results: V. rogosae among all groups, its prevalence was significantly lower in the children with poor oral hygiene than in those with good oral hygiene.
was the V. parvula prevalent species in the poor oral hygiene group. Approximately 10% of the isolated strains were not classified to any established species; Veillonella the phylogenetic analysis showed that they were most closely related to V. infantium This study demonstrates that the composition and proportion Conclusions: of oral species in the saliva of Japanese children is correlated Veillonella The oral biofilm comprises multiple bacterial species and develops as a result of adhesion of pioneer bacterial species to adsorption of salivary proteins and glycoproteins on the enamel surface. These biofilms are not formed by random simultaneous colonization, but rather by selective, reproducible, and sequential colonization 1,2 . Oral biofilms are a source of bacteria that cause oral infections, for instance dental caries and periodontal disease, and they sometimes lead to or worsen systemic diseases 3 .
Saliva is an acknowledged pool of biological markers that range from biochemical molecules changes such as DNA, RNA, and proteins, to those in microbiota structural composition 4 . Furthermore, saliva has an important role in oral biofilm development and maintenance. Recently, metagenomic analysis from saliva samples of Thai children demonstrated that Streptococcus and Veillonella were the predominant bacterial genera in the samples, and the proportion of Streptococcus decreased, while that of Veillonella increased in the children with poor oral hygiene status 5 .
The genus Veillonella consist of multiple gram-negative bacterial species, obligate anaerobic, non-motile, non-spore forming, small cocci belonging to the family Veillonellaceae 6 .
No Veillonella species ferment carbohydrates or amino acids, except for V. criceti, V. ratti, and V. seminalis. The metabolic end products of Veillonella species from trypticase-glucoseyeast extract are mainly acetic acid and propionic acid 6 . Veillonella species are present as commensal organisms in the oral cavity, intestinal tract and genitouritary and respiratory systems of humans and animals. Previous studies have reported that Veillonella species are rare causative organisms of meningitis, endocarditis, bacteraemia, discitis, vertebral osteomyelitis, and prosthetic joint infection [7][8][9] . Generally, Veillonella species are known to be resistant to tetracycline and sensitive to penicillin and ampicillin. However, some Veillonella strains resistant to both penicillin and ampicillin have recently emerged 10 .
There are 14 species reported to belong to genus Veillonella including V. infantium which was assign as a novel species in 2018 11 . Of the 14 documented species, V. atypica, V. denticariosi, V. parvula, V. rogosae, V. dispar, V. infantium, and V. tobetsuensis have been found in human saliva or on tongue or dental biofilms 12-17 . Periasamy and Kolenbrander reported that oral Veillonella species are an early colonizer during the formation of oral biofilm, along with Streptococcus species, which were reported as initial colonizers in developing multispecies communities of oral biofilm 18 . Therefore, it is important to determine the role of oral Veillonella species in formation of oral biofilm to improve the prevention and treatment of oral infectious diseases.
Veillonella strains are relatively easy to identify at the genus level, but remain difficult to identify at the species level, since there are no useful phenotypic or biochemical examinations to distinguish them 19 . To resolve this problem, Mashima et al. established a novel one-step PCR method with species-specific primer sets based on the variable region of the rpoB gene sequences of oral Veillonella species 12 . Additionally, 1,442 Veillonella strains isolated from the saliva of 107 Thai children were identified by this method as V. dispar, V. parvula, V. rogosae, V. atypica, V. denticariosi, and V. tobetsuensis in our previous study 20 . In that study, V. parvula was significantly more prevalent in the poor oral hygiene, and the detection rate of oral Veillonella species in the saliva was indicative of the oral hygiene status of Thai children 20 . Additionally, another study suggested that several novel Veillonella species may inhabit the human oral cavity 21 . However, the study regarding the identification and distribution of oral Veillonella are limited. The oral Veillonella community may affected by the differences in geographical location, age, diet, lifestyle, socio-economic status and oral hygiene status.
Therefore, in this study, we examined composition and proportion of oral Veillonella species in saliva of Japanese children with different oral hygiene status.
Furthermore, we determined the phylogenetic position of the unknown Veillonella strains evaluated by the genus-specific PCR primer set as members of the genus Veillonella with a phylogenetic tree.

Subjects
The 15 children selected to take part in the study were 6 boys and 9 girls, aged 4 to 14 years old. Participants were recruited in-person during appointments at the Dental Hospital, Health Sciences University of Hokkaido. The subjects who had a history of immunosuppression or systemic diseases (e.g. leukemia, hepatitis), or any conditions requiring antibiotic monitoring or treatment procedures (e.g. heart conditions, bone fractures), or those with mucosal lesions, previous chemotherapy, radiation therapy, or medications that can reduce the salivary flow, and those that underwent treatment with antimicrobials within the previous three months were excluded from this study.
Subjects of this study were divided into three groups based on their evaluation by the Simplified Oral Hygiene Index (OHI-s) into good, moderate, and poor oral hygiene groups, according to the criteria of Greene and Vermillion 22 . Owing to the small number of children with poor hygiene (n=5), it was decided that 5 children would be chosen for each group. The good oral hygiene group (OHI-S score: 0-1.2) was composed of two males and three females. The moderate group (OHI-S score: 1.3-3.0) was composed of 3 males and 2 females. The poor group (OHI-S score: 3.1-6.0) was composed of 1 male and 4 females.

Sample collection
The saliva samples were collected at the Dental Hospital, Health Sciences University of Hokkaido, Japan, over a period between 2016 and 2017. Saliva was stimulated by paraffin chewing for 1 min and was then collected into sterile plastic tubes, and transferred to an anaerobic box (Hirasawa Works, Inc., Osaka, Japan)

Amendments from Version 4
Addition of grant number Information. It was necessary to our university.
Any further responses from the reviewers can be found at the end of the article containing 10% H 2 , 85% N 2 , 5% CO 2 . These samples (1 ml each) were transferred to 1.5-ml Eppendorf tubes, then homogenized for 1 min with a BioMasher ® II (Nippi, Incoporated Protein Engineering Office, Tokyo, Japan).

DNA extraction
The genomic DNA was extracted from the isolated bacterial cells by using Insta Gene Matrix Kit (Bio-Rad Laboratories, Hercules, CA, USA). The DNA concentration determination was based on fluorescence by using a Qubit 3.0 Fluorometer. (Invitrogen, Carlsbad, CA, USA), according to the manufacturer's protocol. Additionally, genomic DNA extracted from the standard strains stated above was used as positive control for PCR.

Identification of Veillonella species
For the identification of Veillonella species at the genus level, a genus-specific PCR primer pair, Veill-rpoBF and Veill-rpoBR, were used according to the protocols described by Arif et al. and Mashima et al. 12,13,24 . Strains confirmed by PCR as members of genus Veillonella were then subject to the one-step PCR method with the species-specific primers sets ATYR, DENR, DISR, PARR, ROGR, TOBR, and VF, performed according to the method reported by Mashima et al., for identification at species level 12 .
The PCR products were applied to a 2.0% agarose gel, and after electrophoresis, the gel was stained with SYBR ® Safe DNA gel stain (Invitrogen).

Phylogenetic analysis of unknown strains
For phylogenetic analysis of unknown strains, genomic DNA was also extracted from bacterial cells of unknown Veillonella strains showing positive PCR reaction with the genus-specific primer, but negative with the species-specific primer sets. In addition, PCR-based amplification and sequence analysis of rpoB were performed using the previously described primers for genus Veillonella rpoB-forward (5'-GTA ACA AAG GTG TCG TTT CTC G-3') and rpoB-reverse (5'-GCA CCR TCA AAT ACA GGT GTA GC-3') 24 .
The PCR product contained DNA fragments were purified by using QIAquick ® Gel Extraction Kit (Qiagen, Hilden, NW, Germany), according to the manufacturer's instructions. The DNA concentration after purification was determined based on fluorescence using a Qubit ® 3.0 Fluorometer dsDNA HS Assay Kit (Invitrogen life Technologies, Carlsbad, CA, USA. The PCR reaction was performed with 15-20 ng/µl of DNA template for cycle sequence. Purified DNA from PCR was sequenced with an BigDye ® Terminator v1.1 Cycle Sequencing kit (Thermo Fisher, Waltham, MA, USA), BigDye ® Terminator 5X Sequencing Buffer (Thermo Fisher, Waltham, MA, USA), single primer 1 pM and PCR product in a final volume of 20 µl. Cycle sequencing of the purified DNA was as follows: preheating at 96°C for 1 minutes; followed by 25 cycles of denaturation at 96°C for 10 seconds and annealing with extension at 60°C for 4 minutes 12 . Furthermore, the sequencing of PCR products were purified by using Centri-Sep column (Princeton Separations, Adelphia, NJ, USA), according to the manufacture's instruction and resolved for the sequencing analysis.
DNA sequences were determined using an ABI PRISM 310 Genetic Analyzer (Applied Biosystem) and were aligned and connected using SEQMAN Pro from the LASERGENE program (DNASTAR). The programs MEGALIGN, which includes CLUSTALW and NJPlot were used to compare sequences and to reconstruct an evolutionary tree by the neighbour-joining method 25 . Confidence intervals were also assessed by CLUSTALW with bootstrap analysis. Furthermore, pairwise similarity values were determined with MEGALIGN in the LASERGENE program. The rpoB sequences of the unknown Veillonella strains were aligned against the sequence of the established Veillonella species retrieved from GenBank. Unipro UGENE could be use as free alternative for both sequencing and pairwise similarity values.

Ethical considerations
All subjects and their parents were made aware of the objectives and procedures of the study and parents of participants provided written informed consent. This study was conducted with the approval of The Ethics Committee of the Health Sciences University of Hokkaido, Japan, under process number of 2016-015

Colony numbers
The average number of colony forming units (CFU)/ml of all bacteria on BHI agar increased with decreased oral hygiene: 1.38E+08, 2.2E+08 and 4.48E+09 in the good, moderate and poor groups, respectively. Raw CFU data are available on Figshare 26 .
As shown in the results, V. rogosae was found as the predominant species in the saliva samples of all oral hygiene groups. However, V. denticariosi and V. tobetsuensis were not found in all oral hygiene groups (Table 1-Table 3). Figure 1 shows the per cent ratio of the total number of strains of each species to the total number of Veillonella isolates from saliva samples of the good, moderate, and poor oral hygiene groups.

Strain characteristics
In this study, 179 strains were identified as Veillonella strains, and 162 strains were identified as V. atypica, V. dispar, V. parvula or V. rogosae. However, 17 (9.5%) of 179 strains failed to be classified as any of the known oral Veillonella species, thus, they were defined as unknown species. Of the 17 unknown Veillonella strains 10 (S3-1, S9-1, S10-1, S15-1, S17-1, S21-1, S25-2, S28-1, S29-1 and S30-1) were selected as representative strains from different hygiene groups for phylogenetic analysis. After determination of the rpoB sequences of these 10 strains, these sequences were aligned toward to the sequences of Veillonella type strains retrieved from GenBank. The evolutionary tree produced by analysing the type strains of the 14 Veillonella Table 1. Ratio of the number of isolates of each species to the total number of Veillonella isolate in saliva from the good oral hygiene group. The colony-forming units (CFU) of all anaerobic bacteria on brain heart infusion agar and Veillonella strains on Veillonella agar (detection limit <0.1% of the total count). The total of Veillonella isolates identified by the Veillonella genus-specific PCR primer. Individual species as a percentage of the number of isolates identified by one-step PCR with the species-specific primer sets for each subject (n = 5) from saliva of the good oral hygiene group.

Subject
Total    According to this data, the most closely related species was V. infantium, although the 10 unknown strains formed three clusters. The DNA sequence similarity of the 10 unknown Veillonella strains to V infantium JCM 31738 T (LC191258) ranged from 97.1 to 99.7%.

Discussion
It was previously reported that a higher number of anaerobic bacteria was detected on BHI agar in saliva from Thai children with poor oral hygiene than those with good and moderate oral hygiene 20 . This prior study demonstrated that oral Veillonella isolates were detected at a twofold higher frequency in the saliva of Thai children with poor rather than good or moderate oral hygiene 20 . Here, it was demonstrated that the number of anaerobic bacteria on BHI agar and Veillonella species on the selective medium increased in saliva of Japanese children with worsening oral hygiene status. Therefore, the detection level of anaerobic bacterial strains and oral Veillonella strains in saliva from Japanese children with good, moderate and poor oral hygiene status was similar to that from Thai children.
Using the Illumina MiSeq platform, Mashima et al. demonstrated that Streptococcus and Veillonella species were the predominant bacterial species in the saliva microbiome of Thai children, but that the proportion of Streptococcus decreased while that of Veillonella increased with poor oral hygiene status 5 . They also found that Veillonella species were detected predominantly in the tongue microbiome of Thai children with poor oral hygiene status compared to those with good or moderate oral hygiene status 5 . Taken together with the results of the present study, it is possible that Veillonella species could be a biomarker of oral hygiene status for Thai and Japanese children.
This study demonstrated that V. rogosae was the predominant species detected in all groups of Japanese children (Figure 1 Conversely, the detection rate of V. parvula in the poor (31.7%) oral hygiene was significantly higher than that in the good (10.5%) and moderate (12.3%) oral hygiene groups, in this study with Japanese children. This result is conformed with data from another study, in which V. parvula was most frequently detected in saliva of Thai children with poor oral hygiene status 20 . Previously, it was also reported that V. parvula was frequently detected in periodontal pockets 27 and active carious-lesions 28 . These data suggest that oral cavities with poor hygiene status are suitable environments for V. parvula.
In this study, 179 strains were isolated members of the genus Veillonella from saliva of 15 Japanese children, V. denticariosi and V. tobetsuensis were not found in any samples. In the case of saliva samples from Thai children, the detection rate of V. denticariosi (0.4%) and V. tobetsuensis (1.7%) were very low. In this study 1,609 Veillonella strains were isolated from 107 Thai children 20 . Similarly, it was reported that V. denticariosi was not detected in any of the tongue biofilms of Thai children 12 , and V. denticariosi was detected in tongue biofilm of only one young Japanese adult 17 . Therefore, V. denticariosi may be the least prevalent oral Veillonella species in the saliva and tongue microbiome. On the other hand, V. tobetsuensis, was not detected in saliva from Thai children with good oral hygiene status. However, the detection rate of V. tobetsuensis was 14.3% and 17.8% in the saliva of Thai children with moderate and poor oral hygiene, respectively (20). Similarly, it was demonstrated that the prevalence of V. tobetsuensis ranged from 7.6% to 20.0% in tongue biofilm samples from Japanese adults 16 . Therefore, these data suggest that V. tobetsuensis may be potential to co-occur with other Veillonella species in saliva and tongue biofilms.
In the present study with saliva samples of Japanese children, 17 (9.5%) of 179 strains confirmed as member of genus Veillonella were not belong to any established Veillonella species as unknown species. Theodorea et al., also reported that 167 (10.4%) of 1,609 Veillonella isolates from saliva of Thai children could not be assigned to any species of the genus Veillonella 20 . Furthermore, it was reported that 43 (9.7%) of the 442 Veillonella isolates from periodontal pockets and gingival sulcus could not be identified as any of the known Veillonella species. These data may indicate that other novel Veillonella species inhabit human oral cavities, although only six species were detected as oral Veillonella species up to this point. Further phylogenetic studies are needed to evaluate the possibilities of novel Veillonella species.
In 2018, Mashima et al. 11 proposed V. infantium as a novel species isolated from saliva of Thai children, representing a seventh oral Veillonella species. Therefore, for phylogenetic analysis of the unknown Veillonella strains isolated in this study, the rpoB sequences of type strains of the established Veillonella species, including V. infantium JCM 31738 T , were examined. Consequently, 10 unknown Veillonella strains analysed in this study formed three clusters distinct from V. dispar, the most closely related species was V. infantium. Further studies are required to assign these strains most accurately.
In conclusions, this is the first study to identify oral Veillonella at the species level in the saliva of Japanese children divided into three oral hygiene status groups: good, moderate and poor group. Although V. denticariosi and V. tobetsuensis were not found in any groups in this study because of small number of subjects, the distribution and frequency of V. atypica, V. dispar, V. parvula and V. rogosae, were mostly the same as those in the saliva from Thai children divided into the aforementioned oral hygiene status groups. Additionally, the results of this study demonstrate that changes in the ratio of some Veillonella species, such as an increase of V. parvula and decrease of V. rogosae in those with poor oral hygiene, can be a useful indicator of oral hygiene status, as with results obtained in the study of saliva taken from Thai children. The present study also showed that approximately 10% of the isolated Veillonella strains were not classified to any Veillonella species, and that they will be assigned to V. infantium or novel Veillonella species after further studies.

Takuichi Sato
Division of Clinical Chemistry, Department of Medical Technology, Niigata University Graduate School of Health Sciences, Niigata, Japan The authors have described the following in the Response, but I am unable to find the differences in the abstract (and/or introduction).
"We have changed the wording of the relevant in the abstract to express the reason why did we want to know the detection rate and distribution of oral species in Japan. The reasons are the study Veillonella regarding the identification and distribution of oral are limited; also the oral Veillonella Veillonella community may affected by the differences in geographical location, age, diet, lifestyle, socio-economic status and oral hygiene status." No competing interests were disclosed. Competing Interests: Reviewer Expertise: Oral Microbiology I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Are sufficient details of methods and analysis provided to allow replication by others? Yes
If applicable, is the statistical analysis and its interpretation appropriate? Not applicable Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed.

Competing Interests:
Reviewer Expertise: Oral Microbiology I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
Author Response 24 May 2019 , Faculty of Dentistry, Universitas Indonesia, Jakarta 10430, Indonesia, Jakarta, Ariadna Djais Indonesia Abstract and Introduction: I think that the manuscript, in the Abstract and Introduction, needs to express rationales of this study in more detail, such as, why did the authors want to know the detection rate and distribution of oral species in saliva of Japanese children? Why did the authors want to compare the Veillonella results with those of Thai children? Response: Thank you very much for this direction. We have changed the wording of the relevant in the abstract to express the reason why did we want to know the detection rate and distribution of oral species in Japan. The reasons are the study regarding the identification and Veillonella distribution of oral are limited; also the oral community may affected by the Veillonella Veillonella differences in geographical location, age, diet, lifestyle, socio-economic status and oral hygiene status. Figure 1 and Tables 1-3: I think that the data in Figure 1 and Tables 1-3 are overlapping, and Tables 1-3 could be deleted from the manuscript, because the authors mainly stated the total and the mean (proportions) of the isolates of species in the Results and Discussion. Veillonella Response: Thanks for raising this important point. However, we consider to state the Table 1-3 as Ratio of the number of isolates of each species from each subject. Figure 1 was indicated to express the Total isolated number of each species Veillonella from each group of oral hygiene.

1.
of oral species in the saliva of Japanese children compared to previous studies in Thailand. Veillonella The manuscript is certainly well written but I have some major concerns on the data analysis: The number of samples in this study was fewer (15 children) compared to the number of samples in the previous study from Thailand (107 children). It is suggested that the analysis uses proportions so that it can describe the oral species according to the number of samples. Veillonella To find out the comparison of the results of this study with the results of a previous study from Thailand, a correlation analysis is suggested.
It is also necessary to add the results of a correlation analysis between the results of 's Veillonella oral identification and oral health status.
Minor comment: Some references used are older than 10 years. I suggest to use current references from at least the last 10 years.

If applicable, is the statistical analysis and its interpretation appropriate? No
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed.

Competing Interests:
Reviewer Expertise: Area of my expertise are oral biology and immunology I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

Ariadna Djais Indonesia
This study is good for identification and phylogenetic analysis of oral Veillonella species isolated 1.

1.
This study is good for identification and phylogenetic analysis of oral Veillonella species isolated from saliva samples of children in Japan. The aim of this study was to analyze the composition and proportion of oral species in the saliva of Japanese children compared to previous Veillonella studies in Thailand. Response: Thank you for your valuable comment. The aim of this study was focus to analyze the composition and proportion of oral species in the saliva of Japanese children. Veillonella Furthermore, we compared the results with some reports such as in Thailand. The revise version has been submitted. The manuscript is certainly well written but I have some major concerns on the data analysis: The number of samples in this study was fewer (15 children) compared to the number of samples in the previous study from Thailand (107 children). It is suggested that the analysis uses proportions so that it can describe the oral species according to the number Veillonella of samples. Response: These children visited to the Dental Hospital, Health Sciences University of Hokkaido for dental examination, over a period between 2016 and 2017. Based on the evaluation by the Simplified Oral Hygiene Index, they were divided into three groups, good, moderate and poor. But, among many children, only five children had poor oral hygiene status. Therefore, we used these five children as poor group. And we have selected five children from good and moderate groups, without distinction. As the results, the total subject : 15 subjects.
To find out the comparison of the results of this study with the results of a previous study from Thailand, a correlation analysis is suggested. It is also necessary to add the results of a correlation analysis between the results of Veillonella's oral identification and oral health status. Response: Thank you for another good point at No. 2 and No.3. The research of oral Veillonella were very limited. The analysis was use in this study : descriptive analysis with mean value to determine the distribution and proportion of oral , then we compare the characteristic of Veillonella oral veillonella distribution with the previous study. Minor comment: Some references used are older than 10 years. I suggest to use current references from at least the last 10 years.
Response: As we mentioned before, that the research of oral were very limited. Veillonella Therefore, some of references were used older that 10 years. Thank you for taking the time and energy to help us improve this manuscript. We hope that you will find these revision rise to your expectation.
No competing interests were disclosed.

Competing Interests:
The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias You can publish traditional articles, null/negative results, case reports, data notes and more The peer review process is transparent and collaborative