2.1

Participants

Participants were recruited among partially edentate individuals refereed to the prosthodontics clinic of the University of São Paulo (Ribeirão Preto, Brazil). Potential participants were invited to participate in the study if they: (a) were at least 18 years old; (b) were indicative of a cemented-retained single-unit implant-supported restoration in the anterior maxilla; (c) were indicative of a cemented-retained single-unit implant-supported restoration in the posterior maxilla/mandible; (d) possessed the contra-lateral and antagonists teeth; (e) have had no treatment or professional cleaning in the previous 3 months. Additional exclusion criteria were pregnancy, lactation, periodontal or antibiotic therapy in the previous 3 months, and any systemic condition which could influence the course of periodontal status. The study was approved by the local ethics committee (CAAE 0066.0.138.000-10) and all the experiments were undertaken with the informed and written consent of each subject according to ethical principles.

2.2

Experimental design

Twenty healthy individuals, 17 women and 3 men (mean-age 45.5 years), fulfilled the study’s criteria. The clinical parameter probing depth was selected as a primary variable in determining the progression of peri-implant disease and also for determining the sample size ( N ). To compare two independent groups (Titanium and Zirconia) with repeated measures performed on the three proposed times (baseline, 3 months, and 6 months), and considering a standard deviation of 1.26 among individuals and 0.89 in the intra-individual analysis (values estimated from the deviations observed in the applied literature), the sample size provided a statistical power (power of study) equal to 84% for the factor “group” and 96% for the factor “time” and “group × time” interaction, with a significance level of 5%, and magnitude of effect (effect size) of 0.79 for the factor “group”, and 1.12 for the “time” factor and “group x time” interaction. Participants were enrolled into 2 groups of 10 participants each, according to the investigated abutments and manufacturers’ recommendations: the first group (8 women and 2 men, mean-age 47 years) comprehended individuals who received a two-part dental implant with a morse taper connection (Ankylos C/X, Dentsply, USA) in the anterior area of maxilla and were rehabilitated using esthetic pre-machined zirconia abutments (Ankylos Cercon Balance, Dentsply, USA) and the second group (9 women and 1 men, mean-age 48 years) comprehended individuals who received a two-part dental implant with a morse taper connection (Ankylos C/X, Dentsply, USA) in the posterior area of maxilla/mandible and were rehabilitated using pre-machined titanium abutments (Ankylos Regular Abutment C/X, Dentsply, USA). In both groups, implants were placed at the level of the alveolar bone crest followed by the insertion of transmucosal healing abutments. The healing abutments were different lengths so that the occlusal surface ended 1.0 mm above the gingival marginal level. After 75 days (following manufacturer’ instructions), all the participants were rehabilitated with a total ceramic (in the anterior area) or a metalloceramic (in the posterior area) single unit cemented-retained restoration. All surgical and prosthetic procedures were performed by the same clinician.

2.3

Follow-up and data collection

Microbiological samples and clinical data collection were conducted at baseline–implant loading (T ₀ ), 3-month (T ₁ ) and 6-month (T ₂ ) after the intervention.

2.3.1

Clinical parameters assessment

At each evaluation time point a complete peri-implantar/periodontal examination using an electronic periodontal probe (FP32, Florida Probe, USA) was conducted to record clinical parameters (probing depth, clinical attachment level and bleeding on probing). Each site (mesial, medial and distal in buccal and palatal/lingual aspects) was probed twice to reduce the potential error in probing angulation and the final measurement for probing depth and clinical attachment level was a mean of the 2 evaluations.

Radiographic examinations were conducted at each time point to assess the marginal bone level around dental implants. Intraoral radiographs were obtained using the same film (Kodak E-Speed, Kodak, São José dos Campos, São Paulo, Brazil) and X-ray apparatus (Timex 70E, Gnatus, Ribeirão Preto, SP, Brazil). A paralleling technique was performed by means of an oral device which allowed setting the film on the same position for all the radiographs taken from each participant. The radiographs were digitized using a computerized scanning and images were evaluated using the software Image J Tool (Version 3.00 for Windows, University of Texas Health Sciences Center, USA). The vertical and horizontal marginal bone loss was measured using the most coronal point of the implant as the reference point and the lowest point of marginal bone around the implant as the bone level. Bone loss was measured on the mesial and distal sides of the implants.

2.3.2

Microbiological sampling

Supragingival biofilm samples from the selected prostheses/contra-lateral teeth were recovered with sterile microbrushes. Subgingival biofilm samples from peri-implantar sulci of implants and from periodontal sulci of their contra-lateral teeth were collected with sterile paper points. Each subgingival sample was a pool of 6 paper points exposed for 30 s in the peri-implant or periodontal sulcus, 3 in the buccal and 3 in the palatal/lingual aspects, respectively in mesial, medial and distal positions. Biofilm samples from the internal parts of the implants and abutments surfaces and screws were collected with sterile microbrushes after isolation of the implant site.

Microbiological samples were recorded as follows: Implant-related sites—(1) supragingival biofilm of prosthesis, (2) subgingival biofilm of peri-implant sulcus, (3) biofilm from the internal parts of implant, and (4) biofilm from implant abutment surfaces; Dental-related sites—(1) supragingival biofilm of tooth crown, (2) subgingival biofilm of periodontal sulcus. Samples were stored at −20 °C prior to laboratorial processing.

2.4

Molecular detection of microorganisms

2.4.1

DNA checkerboard hybridization analysis

DNA Checkerboard hybridization procedures were performed as described by do Nascimento et al. . Thirty-eight bacterial species, including putative periodontal pathogens ( Aggregatibacter actinomycetemcomitans a and b, Bacteroides fragillis, Capynocitophaga gingivalis, Campylobacter rectus, Escherichia coli, Eikenella corrodens, Enterococcus faecalis, Fusobacterium nucleatum, Fusobacterium periodonticum, Kleibsella Pneumoniae, Lactobacillus casei, Mycoplasma Salivarium, Neisseria mucosa, Porphyromonas aeruginosa, Peptostreptococcus Anaerobios, Porphyromonas endodontalis, Porphyromonas gingivalis, Prevotella intermedia, Prevotella melaninogenica, Parvimonas micra, Prevotella nigrencens, Pseudomonas putida, Staphylococcus aureus, Streptococcus constelatus, Streptococcus gordonii, Streptococcus mitis, Solobacterium moreei, Streptococcus mutans, Streptococcus oralis, Streptococcus parasanguinis, Staphylococcus pasteuri, Streptococcus salivarius, Streptococcus sanguinis, Streptococcus sobrinus, Treponema denticola, Tanerella forsythia and Veillonella parvula ) and 5 Candida species frequently found harboring the oral microbiota of healthy and diseased individuals were investigated ( C. albicans, C. dublinienses, C. glabrata, C. krusei and C. tropicalis ). Descriptive statistics was used to analyze the frequency of positive signals. The normality of data was assessed using a normality probability plot and Kolmogorov-Smirnov goodness of fit test. The Friedman two-way analysis of variance by rank test for original-level data was used to compare quantitation of each species for repeated measures design. When applicable, post hoc analysis was performed by Wilcoxon test corrected by Dunn method. Differences were considered significant when p < 0.05. The SPSS 17.0.0 statistical software (SPSS Inc., Chicago, IL, USA) was used for data analysis.

2.4.2

Pyrosequencing of PCR-amplified 16S rDNA

DNA from clinical samples was extracted according to Smith et al. and further purified using the QiAmp DNA mini kits (Qiagen, CA, USA). For multiplex sequencing a total of 18 forward primers, each one with a different 10-mer nucleotide barcode, was employed ( Table A.1 ). 16S rDNA Amplification was performed in 25 μl reactions using the Fast Start High Fidelity PCR System from Roche (USA) and 5–250 ng of DNA. PCR cycles were as follows: initial denaturation step of 95 °C, for 2 min; 35 cycles of 95 °C, for 30 s, 55 °C, for 30 s and 72 °C, for 60 s; final extension, at 72 °C, for 4 min. Amplicons were purified using Agencourt AMPure XP kit (Beckman Coulter, USA) and the quality of the purified amplicons was checked using the 2100 Bioanalyzer (Agilent, USA). The amplified libraries were quantified by fluorometry using QuantiFluor® dsDNA System (Promega, USA) and equal amounts (10 ¹⁰ molecules) of 18 different samples were pooled together for multiplex sequencing. Since 36 samples were to be sequenced, we performed 2 independent rounds of amplification, sample pooling and multiplex sequencing. emPCR was performed using the GS Junior emPCR Kit-Lib-L (454/Roche). Sequencing was executed according to the manufacturer’s instructions using the Genome Sequencing Junior System (454/Roche).

2.4.3

Sequencing analysis and taxonomic assignment

Initially low quality sequences, short fragments and sequences derived from polyclonal amplification generated during the emPCR step were removed. Subsequently sequences were demultiplexed and the data was deposited and analyzed in the Metagenomics RAST server (MG-RAST, version 3) . Reads were clustered at 97% identity, and the longest sequence was picked as the cluster representative. A BLAT search for the cluster representative was then performed against the ITS, Greengenes, Silva LSU, M5RNA, RDP and Silva SSU databases. A minimum cutoff of 98% and maximum e-Value Cutoff of 1 × 10 ⁻⁵ were applied, which were sufficient for species identification. A minimum alignment length cutoff of 200 bp was used for analysis. Phylogenetic tree and heatmaps of taxonomic composition were generated using the gplots library with relative frequencies per sample.