The current study investigated 70 patients (23 males, 47 females; mean age, 69.5 years; range, 45–92 years) who visited the Department of Oral Health of Hiroshima University Hospital, Hiroshima, Japan from March 2018 to April 2018. A 0–3 mm probing pocket depth was considered to indicate mild or no periodontitis (11). Moderate periodontal disease was defined as the presence of ≥4 and <6 mm probing pocket depth (11) Severe periodontitis was defined as the presence of ≥6 mm pocket (11). All patients exhibited a mild to moderate periodontal status. No patients exhibited symptoms of acute periodontal disease. Regarding toothbrushing frequency, all participants brushed their teeth more than twice daily. None used a tongue scraper to clean the surface area of the tongue. The study design was approved by the Ethics Committee of Hiroshima University, and all participants signed an informed consent agreement. Clinical data were obtained regarding age, sex, smoking status, alcohol consumption and medical history.

Samples were collected from the tongue dorsum using a Orcellex^® Brush (Rovers Medical Devices, Oss, The Netherlands) which allows collection of cells or bacteria from oral epithelial layers. The tongue surface was swabbed prior to tooth brushing using the Orcellex^® 10 times. DNA was extracted and purified using a PureLink™ Microbiome DNA Purification kit (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA). For PCR analysis, 1 µl was used from a total of 50 µl of oral sample DNA.

The bacterial 16S rRNA gene was used to quantify bacterial numbers using real-time PCR. Since 16S rRNA gene copy numbers differ among bacteria, it is difficult to use them in the accurate quantification of bacteria. Many bacteria have 1 to 10 16S rRNA gene copies, while some have >10 (12,13). Here, 16S rRNA gene copy numbers were used from Staphylococcus aureus (5 copies per cell) to quantify approximate oral bacterial numbers.

First, DNA was extracted from Staphylococcus aureus [1.0×10⁸ colony forming units (CFU)/ml] and used for real-time PCR. Serial 10-fold dilutions of the DNA sample (ranging from 10¹ to 10⁸ CFU/ml) were prepared to make a standard curve (Fig. 1). DNA levels were quantified using a CFX connect real-time PCR detection system (Bio-Rad Laboratories, Inc., Hercules, CA, USA) and SYBR-Green PCR Master Mix (Toyobo Life Science, Osaka, Japan), with a reaction mixture containing 1 µl DNA, 9 µl SYBR-Green Mix, and 10 µmol of each oligonucleotide primer pair. Amplifications were performed with initial melting at 95°C for 5 min, followed by 40 cycles of denaturing at 95°C for 30 sec, annealing at 56°C for 30 sec, and extension at 72°C for 1 min. The 16S rRNA primer sequences were forward, 5′-CGTTAGTAATCGTGGATCAGAATG-3′ and reverse, 5′-TGTGACGGGCGGTGTGTA-3′ (14). A standard curve indicating the cycle threshold value versus the 16S rRNA gene was obtained to estimate the bacterial number per sample (15).

For PCR, 1 µl DNA samples (bacterial numbers: 1.0×10³−1.0×10⁶ CFU/ml) were used. Each mixture was amplified with 10X PCR buffer, dNTPs, Taq DNA polymerase (both from Toyobo Life Science) and primers. The following previously reported PCR primer sets were used: For Porphyromonas gingivalis (P. gingivalis) forward, 5′-AGGCAGCTTGCCATACTGCG-3′ and reverse, 5′-ACTGTTAGCAACTACCGATGT-3′; For Tannerella forsythia (T. forsythia) forward, 5′-GCGTATGTAACCTGCCCGCA-3′ and reverse, 5′-TGCTTCAGTGTCAGTTATACCT-3′; and for Treponema denticola (T. denticola) forward, 5′-TAATACCGAATGTGCTCATTTACAT-3′ and reverse, 5′-TCAAAGAAGCATTCCCTCTTCTTCTTA-3′ (16). The PCR program included initial melting at 95°C for 5 min, followed by 35 cycles of 95°C for 30 sec, 56°C for 1 min and 72°C for 30 sec. Following the reaction, 10 µl of the PCR product was electrophoresed on 2% agarose gels with ethidium bromide staining.

Oral examination including periodontal examination [to assess probing depth (PD) and bleeding on probing (BOP)] was performed by dentists at Hiroshima University Hospital. Remaining teeth and denture use were recorded. Oral wetness was evaluated using an oral moisture-checking device (Moisture Checker Mucus^®; Scalar Corporation, Tokyo, Japan) as previously described (17). The oral moisture level was measured at the tongue dorsum 10 mm from the apex linguae and expressed as a percentage, with mean values calculated from three independent measurements; a value of >30% was defined as high moisture level and a value of ≤30% as low moisture level as in previous study (17). The tongue coating was classified as none, light-thin (visibly pink underneath) or heavy-thick (no pink observed). The presence of a light-thin and/or heavy-thick coating was considered positive (18).

Statistical analysis was performed using SPSS software, version 24.0 (IBM Corp, Armonk, NY, USA). Determination of Spearman's rank correlation coefficient was performed to examine the correlation of bacterial number with age and moisture level. The Mann-Whitney U test with Bonferroni correction or the Kruskal-Wallis test was used to evaluate significant differences in bacterial numbers and clinical factors. All Kruskal-Wallis tests indicated no significant difference and therefore post-hoc tests were not applicable. The χ² test or Fisher's exact test was used to evaluate significant differences between positive rates of periodontal bacteria and clinical factors. P<0.05 was considered to indicate statistical significance.

Age was weakly correlated with bacterial numbers on the tongue surface [Spearman's rank correlation coefficient (R)=0.130, P=0.280; Fig. 2]. Meanwhile, the oral bacterial numbers were weakly negatively correlated with moisture levels (R=−0.131, P=0.278; Fig. 3). Table I summarizes the associations between bacterial counts detected by real-time PCR and clinical factors. Oral bacterial numbers appeared to be increased in individuals with fewer than 10 remaining teeth and in denture users; however, bacterial numbers did not differ significantly according to remaining teeth or denture use (Table I). Furthermore, no significant difference was identified between bacterial numbers on the tongue surface and oral hygiene status [BOP and PD (≥4 mm)]. Individuals with a medical history of stroke exhibited increased bacteria compared with those without, but the difference was not significant (Table I).

Red complex bacteria (e.g., T. denticola, T. forsythia and P. gingivalis) serve major roles in periodontal disease etiology (19). Therefore, the present study performed PCR to detect DNA from these bacteria (Fig. 4), to examine the relationship between periodontal bacteria of the tongue dorsum and clinical factors. A total of 31 out of the 70 participants (57.1%) were P. gingivalis-positive (Table II), with the rate appearing higher in people aged 65 years and over (50.0%) compared with in those aged under 65 (30.0%). Additionally, smokers exhibited higher P. gingivalis-positive rates than nonsmokers. Subjects with medical histories of hypertension, diabetes, stroke and heart disease exhibited higher P. gingivalis-positive rates than those without these disorders. Subjects with BOP and ≥4 mm PD exhibited higher P. gingivalis-positive rates (50.0 and 51.1%, respectively) than those with no BOP and <4 mm PD (39.5 and 30.4%, respectively), but these differences were not significant.

In contrast to the results for P. gingivalis, T. forsythia exhibited no apparent differences in positive rates between those with and without factors associated with systemic and oral disease (Table III). All smokers were T. forsythia-positive, but no significant difference was identified in positive rate between smokers and never-smokers.

For T. denticola, men were indicated to exhibit significantly higher positive rates than women (Table IV). T. denticola positivity did not differ between those with and without a medical history of hypertension, diabetes, stroke or heart disease. Interestingly, there appeared to be a significant association between the T. denticola-positive rate and the number of remaining teeth. All individuals with fewer than nine teeth exhibited T. denticola-positivity, while the positive rates in those with 10 or more teeth were <40%. These results indicate that T. denticola on the tongue surface may be associated with tooth loss.