br In this study we report the oral microbiota
In this study, we report the oral microbiota of Japanese oral cancer patients compared with that in non-cancer subjects, using 16S rRNA amplicon sequencing of saliva samples. This study aimed to clarify the relationship between oral microbiota and oral cancer occurrence in Japan.
2. Materials and methods
2.1. Sample collection
In total, 140 salivary samples taken from patients between 2016 and 2018 were included in this study; this was approved by the Research Ethics Committee at Southern TOHOKU General Hospital (Approval number: 216e3), Tsurumi University (Approval number: 1523) and the National Institute of Biomedical Innovation, Health and Nutrition (Approval number: 167). Prior to sample collection, written informed consent was obtained from all patients. Patients recruited from the Southern TOHOKU General Hospital were clas-sified into two groups: control (80 noncancer individuals) and OSCC (60 patients). The control group was defined as individuals without any diagnosed mucosal diseases and other cancers and these individuals were over the age of 40. The diagnosis of a healthy oral cavity was made after a thorough clinical examination. All di-agnoses of OSCC were confirmed by biopsy and pathological find-ings. Participants who did not follow the instructions or had low saliva secretion capacity were excluded. Participants who were undergoing chemo radiotherapy or had received Lithocholic Acid treat-ment within 28 days were excluded. The participants were asked to chew gum for 5 min and stimulated saliva samples were collected into sterile plastic tubes. The samples were stored at 80 C until use. The participants completed a questionnaire regarding sex, age, drinking, and smoking habits as well as denture use.
DNA was extracted according to a previously described protocol . Homogenization of the suspended salivary samples (200 ml) was performed using beads with 300 ml lysis buffer (No. 10, Kurabo Industries Ltd., Osaka, Japan) and 0.5 g of 0.1-mm glass beads in 2- ml vials. The mixture was mechanically disrupted using a Cell destroyer PS1000 (Bio Medical Science, Tokyo, Japan) at 4260 rpm for 50 s at room temperature (20e25 C). All samples were centrifuged at 12,000 g for 5 min at room temperature. The su-pernatant was collected and mixed with 150 ml lysis buffer and 150 ml proteinase K buffer (No. 2, Kurabo Industries Ltd) containing 0.4 mg/ml proteinase K. DNA was extracted using a Gene Prep Star PI-80X device (Kurabo Industries Ltd). The extracted DNA was determined using a NanoDrop Spectrophotometer ND-1000 (Thermo Fisher Scientific Inc., DE, USA), and the samples were stored at 30 C until use.
16S rRNA sequencing was performed as described previously . The V3eV4 region of the 16S rRNA gene was amplified from salivary DNA samples using previously published primers. The re-action process was performed at 95 C for 3 min, followed by 25 cycles at 95 C for 30 s, 55 C for 30 s, and 68 C for 1 min, with a final extension at 68 C for 5 min. PCR products were purified with 20 ml of Agencourt AMPure XP (Beckman Coulter, Inc., CA, USA) in accordance with the manufacturer's protocol and eluted into 50 ml of 10 mM TriseHCl, pH 8.5. For DNA library preparation, Illumina adapters were attached to the PCR products using Illumina MiSeq Nextera kit set A (Illumina Inc., CA, USA). 16S rRNA gene sequencing of the PCR products was performed using Illumina MiSeq (Illumina) in accordance with the manufacturer's instructions.
2.4. Bioinformatics analysis
FASTQ files were obtained after Illumina pair-end 16S rRNA gene amplicon sequencing, and the operational taxonomic units (OTUs) classification and diversity analyses were performed using QIIME version 1.9.1  according to previously described methods . The sequences were clustered into OTUs by an open-reference OTU picking process against the SILVA 128 reference database  at 97% similarity using USEARCH . The OTUs were classified taxonomically up to genus level using the SILVA 128 reference database .
2.5. Statistical analysis
The resulting data were exported as BIOM files and imported to R (version 3.5.0). Diversity analysis was performed using the “phyloseq” R-package . a-diversity indexes (OTU observed, Chao 1 Index, Shannon Index, and Simpson Index) were calculated by the estimate_richness function and the b-diversity index, calcu-lated by unweighted UniFrac distance and weighted UniFrac dis-tance, was generated using the unifrac function in the “phyloseq” R-package. Oral bacterial community structure similarity of each sample was calculated using principal coordinate analysis (dudi.pco function in “ade4” R-package). The dominant bacteria from phylum to genus level were defined as the mean of the distribution of bacterial composition with at least 0.05%. Student's t-test was used to compare the dominant oral bacterial community (larger than 1%) between the OSCC and controls (t.test function in “stats” R-pack-age). To compare the relative abundances of taxa between OSCC samples of different T-stage or N-stage, we used the Wilcoxon rank sum test (wilcox.test function in “stats” R-package) and Pearson correlation analysis (cor function in “stats” R-package), respectively. Logistic regression analysis by the stepwise-selected model was employed to identify the impact factor for cancer incidence (glm function and step function in “stats” R-package).