Palatal gingival tissues were isolated from around ligatured implants and were homogenized in lysis buffer using a tissue homogenizer. Total RNA was extracted using PureLink® RNA Mini Kit (Ambion). cDNA was synthesized using the SuperScript III Reversed Transcriptase kit (Invitrogen) according to the manufacturer’s protocol. The mRNA expression of TNF-α and RANKL in gingival was determined by real-time quantitative PCR (RT-qPCR) using LightCycler® SYBR Green I master and LightCycler® 480 Instrument system (Roche). GAPDH gene was used as an internal control. The sequences of primers are listed as follows: TNF-α forward 5′-CACAGAAAGCATGATCCGCGACGT-3′; TNF-α reverse 5′-CGGCAGAGAGGAGGTTGACTTTCT-3′; RANKL forward 5′-GGGTGTGTACAAGACCC-3′; RANKL reverse 5′-CATGTGCCACTGAGAACCTTGAA-3″ GAPDH forward 5′-CCCCAGCAAGGACACTGAGCAA-3′; GAPDH reverse 5′-GTGGGTGCAGCGAACTTTATTGATG-3′.

All the maxilla collected were fixed in 4% formaldehyde overnight and then went through decalcification in 10% EDTA for 3 weeks at 4 °C with shaking. Five-micrometer-thick sections were produced in the mesial-distal plane for hematoxylin and eosin (H&E) and tartrate-resistant acid phosphatase (TRAP) staining. An acid phosphatase kit (catalog number 387A, Sigma) was used for TRAP staining. After 30 min staining and 1 min counterstain with hematoxylin, TRAP-positive cells with three or more nuclei were considered to be osteoclasts. A region of interest (ROI) was defined in peri-implantitis samples as a 1.5 × 1-mm rectangular area aligned with the central long axis of the implant and covered the whole length of the implant. TRAP-positive cell numbers within the ROI were quantified manually by ImageJ. For H&E staining, images were analyzed by ImageJ after being captured by a digital camera. The numbers of inflammatory cells from the implant supportive tissue on each slide were counted at a magnification of × 40, and the average numbers were calculated.

Results were presented as mean ± SD. Unpaired Student’s t test was used to analyze differences between any two groups of data sets. Results with p < 0.05 are considered statistically significant.