Bacteria species involved in peri-implantitis investigated using PCR method in this study are listed in Table 1. Bacterial strains were obtained from American Type Culture Collection (ATCC; USA), Culture Collection University of Gothenburg (CCUG; Sweden) and Japan Collection of Microorganisms (JCM; Japan). The following bacterial strains were used in the present study: Eubacterium sulci ATCC 35585, Eubacterium saphenum ATCC 49989, Eubacterium limosum JCM 6421, Eubacterium nodatum JCM 14550, Eubacterium branchy ATCC 33089, Eubacterium yurii ATCC 43714, Eubacterium infirmum ATCC 700433, Eubacterium minutum ATCC 700079, Atopobium minutum ATCC 33267, Atopobium deltae CCUG 65171, Atopobium rimae ATCC 49626, Atopobium fossor ATCC 43386, Atopobium parvulum ATCC 33793, Atopobium vaginae ATCC BAA-55, Gemella morbillorum JCM 12968, Streptococcus mutans JCM 5705, Filifactor alocis ATCC 35896, Fretibacterum fastidiosum JCM 16858, Fusobacterium nucleatum JCM 8532 and Slackia exigua JCM 11022. Anaerobic bacteria strains, i.e. genera Eubacterium, Atopobium, Gemella, Filifactor, Fretibacterium, Fusobacterium and Slackia exigua, were maintained by cultivating them on Fastidious Anaerobe Agar (FA; Neogen Co., Lansing, MI, USA) with sterile defibrinated sheep blood. These organisms were cultured at 37˚C for 48 h under anaerobic conditions with a gas pack system (AnaeroPack; Mitsubishi Gas Chemical Co., Inc., Tokyo, Japan). S. mutans strains were maintained by cultivating them on Bact^TM Brain Heart Infusion (BHI; Becton, Dickinson and Co., Sparks, MD, USA) and 1.5% agar (BHI agar). The organism was cultured at 37˚C overnight in an atmosphere of 5% CO₂ in a CO₂ incubator (MCO-18AIC; Sanyo Electric Co., Tokyo, Japan).

Design of species-specific primers for some bacteria species involved in peri-implantitis, i.e. E. sulci, E. saphenum, E. limosum, E. nodatum, E. branchy, E. yurii, E. infirmum, E. minutum, A. minutum, A. deltae, A. rimae, A. fossor, A. parvulum, A. vaginae, G. morbillorum, S. mutans, F. alocis, F. fastidiosum, F. nucleatum and S. exigua was performed as described previously [20] . Briefly, the 16S rRNA gene sequences of E. sulci (accession no. AJ006963), E. saphenum (U65987), E. limosum (M59120), E. nodatum (Z36274), E. branchy (Z36272), E. yurii (GU269551), E. infirmum (Z36273), E. minutum (AB020885), A. minutum (M59059), A. deltae (KF537630), A. rimae (AB540986), A. fossor (AB015945), A. parvulum (AB558168), A. vaginae (Y17195), G. morbillorum (LC096237), S. mutans (AJ243965), F. alocis (AJ006962), F. fastidiosum (GQ149247), F. nucleatum (M58683) and S. exigua (AF101240) were obtained from the DNA Data Bank of Japan (DDBJ; https://www.ddbj.nig.ac.jp/services/index.html, Mishima, Japan), and a multiple sequence alignment analysis was performed with the CLUSTAL W program.

A PCR method for detecting bacteria species involved in peri-implantitis, using the designed primers in this study and the previous reported primers [21] - [26] that were listed in Table 1, was developed as follows. Bacterial cells which were cultured on FA or BHI agar were suspended at a density of 1.0 McFarland standard (approximately 10⁷ CFU in 1 ml of sterile distilled water). A total of 5.6 μl of the suspension was then used as a PCR template. The detection limit for PCR was assessed by serially diluting known numbers of bacterial cells in sterile distilled water and then subjecting each suspension to PCR. The PCR mixture contained 2 μM of each primer, 10 μl of 2 × MightyAmp Buffer Ver. 2 (Takara Bio Inc., Shiga, Japan), 0.4 μl of MightyAmp DNA Polymerase (Takara), and 3.6 μl of the template in a final volume of 20 μl. PCR was performed in a DNA thermal cycler (Applied Biosystems 2720 Thermal Cycler; Applied Biosystems, Carlsbad, CA). PCR conditions included an initial denaturation step at 98˚C for 2 min, followed by 30 cycles consisting of 98˚C for 10 s and 68˚C for 1 min. PCR products were analyzed by 2.0% agarose gel electrophoresis before being visualized by electrophoresis in 1 × Tris-borate-EDTA on a 2% agarose gel stained with ethidium bromide. A 100-bp DNA ladder (Takara Biomed, Shiga, Japan) was used as a molecular size marker.

Sixty patients attending Nihon University Hospital, School of Dentistry at Matsudo, participated in the present study. They were divided into two subject groups: periodontally healthy patient with implants (HI) and patients with peri-implantitis (PI) groups. Thirty HI and thirty PI subjects were selected by inclusion criteria for peri-implantitis as follows: patients who underwent dental implantation treatments between 2015 and 2019; patients with at least one dental implant for more than half a year; according to the Guidelines of Periodontology, PI was defined as bleeding of probing (BOP) and/or probing pocket depth (PPD) ≥ 4 mm, accompanied by bone tissue loss under the first thread of the implant (i.e. bone absorption ≥ 2 mm). HI was defined as PPD ≤ 3 mm, and the absence of BOP, pus discharge, and bone absorption. Exclusion criteria were as follows: patients with systematic diseases; patients receiving periodontal therapy within six months; taking immunosuppressive agents or antibiotics; the long-term use of contraceptive drugs; pregnant women. Peri-implant sulcus fluid (PISF) samples were collected using endodontic paper points from all subjects and placed in a sterile microcentrifuge tube containing 50 μl of Tris-HCl buffer (0.05 M, pH 7.2). Samples were dispersed by sonication for 30 s in an ice bath (50 W, 20 kHz, Astrason^® System model XL 2020, NY., USA). The detection frequencies of bacteria species involved in peri-implantitis in each PISF sample were determined using a PCR method. The present study was conducted in accordance with the principles of the Declaration of Helsinki, and was approved by the Ethics Committee of Nihon University School of Dentistry at Matsudo, Japan (EC 18-033).

The detection frequencies of bacteria species involved in peri-implantitis in both groups were compared using Fisher’s exact test. Values of p < 0.05 were considered significant. Moreover, bacteria species involved in peri-implantitis were divided into three hierarchies (Hierarchy I: p < 0.01; Hierarchy II: p < 0.05; Hierarchy III: p > 0.05).

The clinical parameters of HI and PI groups are shown in Table 2. The average ages and PPDs of HI and PI groups were 56 (range: 37 - 68) and 58 (range: 44 - 71), and 2.32 mm and 7.78 mm, respectively. The detection frequencies of bacteria species involved in peri-implantitis in PISF samples obtained from the two groups are shown in Table 1. Also, Table 3 shows the hierarchy of bacteria species involved in peri-implantitis. In the PI group, detection frequencies of Corynebacterium durum (p = 0.0001), F. fastidiosum (p = 0.0002) and Slackia exigua (p = 0.0004) were significantly higher than those of the HI group, and were grouped into Hierarchy I (p < 0.01). Following those organisms, Tannerella forsythia (p = 0.003), E. nodatum (p = 0.005), Porphyromonas endodontitis (p = 0.015), F. alocis (p = 0.015), E. yurii (p = 0.024), E. infirmum (p = 0.024) and Campylobacter rectus (p = 0.037) were grouped into Hierarchy II (p < 0.05). On the other hand, Porphyromonas gingivalis (p = 0.158) and Treponema denticola (p = 0.054) belonging to red complex were frequently detected from some of the HI group, and were grouped into Hierarchy III (p > 0.05).