Titanium (Ti) and zirconia (Zi) gave the best results in terms of HGK behavior.
PICNs exhibited intermediate results between Ti–Zi and lithium disilicate glass-ceramic.
No monomer release, nor indirect cytotoxicity, was found for PICNs.
Good results for PICNs could be explained by their innovative polymerization mode.
Clinical impact of these results needs to be investigated on bone-level implants.
Biocompatibility of polymer-infiltrated-ceramic-network (PICN) materials, a new class of CAD–CAM composites, is poorly explored in the literature, in particular, no data are available regarding Human Gingival Keratinocytes (HGK). The first objective of this study was to evaluate the in vitro biocompatibility of PICNs with HGKs in comparison with other materials typically used for implant prostheses. The second objective was to correlate results with PICN monomer release and indirect cytotoxicity.
HGK attachment, proliferation and spreading on PICN, grade V titanium (Ti), yttrium zirconia (Zi), lithium disilicate glass-ceramic (eM) and polytetrafluoroethylene (negative control) discs were evaluated using a specific insert-based culture system. For PICN and eM samples, monomer release in the culture medium was quantified by high performance liquid chromatography and indirect cytotoxicity tests were performed.
Ti and Zi exhibited the best results regarding HGK viability, number and coverage. eM showed inferior results while PICN showed statistically similar results to eM but also to Ti regarding cell number and to Ti and Zi regarding cell viability. No monomer release from PICN discs was found, nor indirect cytotoxicity, as for eM.
The results confirmed the excellent behavior of Ti and Zi with gingival cells. Even if polymer based, PICN materials exhibited intermediate results between Ti–Zi and eM. These promising results could notably be explained by PICN high temperature–high pressure (HT–HP) innovative polymerization mode, as confirmed by the absence of monomer release and indirect cytotoxicity.
Biocompatibility of prosthesis materials with soft tissues constitutes a critical parameter, particularly regarding implant prostheses and the creation of a hermetic barrier between the bone tissue surrounding the implant and the external oral environment. This biological seal, called “biological width”, is located just above the bone level and is composed of 2 distinct parts: a connective tissue attachment and, above it, an epithelial tissue attachment (junctional epithelium) ( Fig. 1 ). On “bone level” implants the biological width is located on the transgingival part of the prosthesis material, which should promote attachment and proliferation of gingival cells. In particular, the junctional epithelium forms the first defense line against bacterial penetration and consequently protects from bone resorption and gingival recession , playing a key role in the long-term success of implant prostheses from both a biological and esthetic point of view. This epithelial tissue attachment is reputed to reach ±2 mm in length, while the connective tissue is ±1 or 2 mm long, but clinically longer epithelial attachment can be observed in the case of thick gingiva: this epithelial ring forms a strong seal around the transgingival part of the implant prosthesis ( Fig. 2 ).
Recent advances in computer-aided design and computer-aided manufacturing (CAD–CAM) materials have introduced composites in addition to metal alloys, mainly titanium, and ceramic materials, such as yttria–tetragonal-zirconia-polycrystal (Y–TZP) or glass-ceramics reinforced with lithium disilicate (Li 2 Si 2 O 5 ), to manufacture implant abutments and single-unit prostheses. One of the main advantages of composites in comparison with ceramics is their machinability: they are more adapted to CAD–CAM processes since they are not as brittle, exhibit a higher damage tolerance and resistance to marginal chipping, can be milled more rapidly and in very low thickness . In comparison with artisanal dental composite materials, current CAD–CAM composites are characterized by their (1) microstructure: dispersed filler materials, as artisanal composites, or polymer-infiltrated-ceramic-network materials (PICNs); (2) polymerization mode: high temperature (HT, used in dispersed filler CAD–CAM composites) or a combination of high temperature and high pressure (HT–HP, used in PICNs), instead of light as in artisanal composites; (3) composition: urethane dimethacrylate (UDMA) instead of bisphenol A–glycidyl methacrylate (Bis–GMA) in artisanal composites. PICNs differ significantly from other composite materials with dispersed fillers classically incorporated by mixing, since they are the result of the infiltration of a pre-sintered glass-ceramic scaffold with urethane dimethacrylate (UDMA) and triethylene glycol dimethacrylate (TEGDMA) monomers, which are secondarily polymerized. The characteristic ceramic network forms a real skeleton of interconnected particles, which is able to promote resistance to breakdown phenomena and gives the material an elastic modulus between dentin and enamel , while dispersed filler CAD–CAM composites exhibit an elasticity modulus lower than dentin. Most importantly, the patented HT–HP polymerization mode (300 MPa, 180 °C) significantly increases the degree of conversion of monomers (up to 96% for experimental PICNs) and, consequently, improves material mechanical properties and decreases its toxicity compared to light-cured dental composites . Indeed, HT–HP UDMA was shown to exhibit very low monomer release (388 times less) in comparison with light-cured UDMA . Recently, Tassin et al. have studied direct and indirect cytotoxicity of three different PICNs with oral stem cells and have not reported any cytotoxic effect or influence on cell proliferation, extracellular matrix synthesis, morphology or inflammatory response, while the light-cured direct composite showed direct cytotoxicity . The authors introduced the hypothesis of the influence of the polymerization mode and degree of conversion of monomers, which were previously reported to influence toxicity and cell response to explain their results .
A recent previous study was dedicated to the biocompatibility of PICN materials with human gingival fibroblasts (HGFs) compared to different implant prosthesis materials, performing cell culture on material discs. The results confirmed that titanium and zirconia could be considered to be the gold standard materials in terms of HGF behavior (attachment, proliferation and spreading), while lithium disilicate glass-ceramic gave intermediate results between the gold standard and the negative control. Despite the presence of polymers and their hydrophobicity, PICN exhibited comparable results to lithium disilicate glass-ceramic: these findings were also explained by the probable low monomer release. Yet, to the authors’ knowledge, there is no study in the literature related to HGK behavior on PICNs, while in general there are less data on HGK than on HGF behavior on prosthesis materials. Many studies focused on the biocompatibility properties of titanium and Y–TZP, several authors having shown a good in vitro cytocompatibility of zirconia and titanium with HGKs in terms of attachment and proliferation. A few publications have directly compared zirconia and titanium and these studies suffer from hardly comparable results since material roughness and evaluation times varied considerably from one study to the other. If in two studies the materials gave comparable results in terms of keratinocytes attachment and proliferation , one study has shown a better initial HGK attachment on titanium and another a better proliferation on zirconia . Regarding lithium-disilicate glass ceramics, HGK behavior on this material has been poorly studied . One study reported slightly better properties for lithium disilicate glass-ceramic than for zirconia in terms of HGK attachment and proliferation , however, another study revealed a small cytotoxic effect of lithium disilicate glass-ceramic on HGK, with a decrease in viability and migration .
Consequently, following up on a previous study concerning HGFs, the first objective of the present work was to evaluate the biocompatibility of PICNs in comparison with metallic and ceramic materials used for dental implant prostheses, evaluating HGK attachment, proliferation and spreading. The second objective was to correlate results with PICN UDMA release and indirect cytotoxicity, notably in comparison with lithium disilicate glass-ceramics.
Materials and methods
Discs of experimental PICN (PICN, n = 40) and commercial grade V titanium (Procera, Nobel Biocare, Göteborg, Sweden) (Ti, n = 34), Y–TZP (Procera, Nobel Biocare) (Zi, n = 34), lithium disilicate glass-ceramic (IPS e.max Press, Ivoclar Vivadent, Schaan, Liechtenstein) (eM, n = 40), and polytetrafluoroethylene (Sirris, Seraing, Belgium) (PTFE, negative control, n = 34), 13.5 mm in diameter, were produced. The experimental PICN (MaJEB sprl, Liège, Belgium) was made by infiltration of a pre-sintered glass-ceramic scaffold (Vita Mark II, Vita Zahnfabrik, Bad Säckingen, Germany) with a monomer (UDMA), and was polymerized under HP/HT (300 MPa, 180 °C) conditions. The PICN discs were cut out of a block and were rounded by polishing. Ti and Zi discs were manufactured by milling with the CAD–CAM Procera process (Nobel Biocare). eM samples were pressed into molds by following the manufacturer’s instructions.
All specimens were sequentially ground with 800-grit, 2400-grit and 4000-grit silicon carbide discs (Planopol 3 ® , Struers, Ballerup, Denmark) to a 1 ± 0.02 mm thickness. They were ultrasonically cleaned in a detergent (Elmasonic One product, Singen, Germany and RBS T105 ® 2%, Chemical Products, Brussels, Belgium) for 10 min, rinsed with phosphate-buffered saline solution (PBS, Lonza Group Ltd., Basel, Switzerland), rinsed 4 times with purified water (Milli-Q ® system, Merck Millipore Corporation, Billerica, Massachusetts, United States) and then ultrasonically cleaned in ethanol 80%. Finally, the discs were sterilized by UV exposure for 30 min per face.
The surface of all samples was characterized in a previous study by X-ray photoelectron spectroscopy, contact angle measurement, profilometry and scanning electron microscopy .
Cell culture and seeding
All manipulations were made under sterile conditions (L2 laminar flow hood). The cells were cultured under the condition of 5% CO 2 and 37 °C (Binder incubator, Tuttlingen, Germany). Primary cultures of Human Gingival Keratinocytes were established using the explant technique. The gingival tissue samples were obtained from healthy adult patients undergoing dental surgery at the oral and maxillofacial surgery department at the University Hospital of Liège. For tissue harvest, informed consent was obtained from the patients and the protocol was approved by the Ethics Committee of the University Hospital of Liège, Belgium (October 8, 2013). Explants were immediately stored at 4 °C in a transport medium composed of Dulbecco’s Modified Eagle Medium (DMEM high glucose, GlutaMAX™ Supplement, pyruvate, Gibco, Invitrogen, Paisley, UK) supplemented with 10% fetal bovine serum (FBS, Gibco), 500 μg/ml gentamicin (Gibco), 200 IU/ml penicillin-200 μg/ml streptomycin (Lonza, Verviers, Belgium) and 25 μg/ml fungizone (Gibco). The epithelial portion of the explant was separated from the conjunctival part using a scalpel. The epithelial portion was placed in 0.25% trypsin without EDTA (Gibco) at 4 °C overnight. The sample was then scrubed with a scalpel in a Petri dish and cells were harvested with FBS (Gibco). The suspension was centrifuged at 200 g for 5 min, the supernatant was removed and the pellet resuspended in a complete medium composed of Keratinocyte Serum Free Medium (KSFM Supplement: Human Recombinant Epidermal Growth Factor (EGF 1-53) & Bovine Pituitary Extract (BPE)) (Gibco) supplemented with 200 IU/ml penicillin-200 μg/ml streptomycin (Lonza). The solution was then filtered in a 100 μm filter (cell strainer BD Faclcon TM, Dutscher, Brumath, France) and the filter was rinsed with additional medium. Wells of a 12-well plate culture were filled with 2 ml of this solution and the plate was placed in the incubator. The cultures were kept for 2–4 weeks in the incubator until near confluence. Cells were washed with phosphate-buffered saline (PBS) solution without calcium and magnesium (DPBS, Lonza) and detachment was performed with 0.05% trypsin–EDTA solution (Gibco). Cells were expanded in T75 flasks by passing a 1:3 split. Stocks from different patients were stored in liquid nitrogen (complete medium—dimethylsulfoxide 10%). All experiments were conducted at the 3rd to 6th passage.
Cell seeding was performed following a method described previously , using specific cylindrical inserts, which ensure an efficient seal around the discs (Insert-Based System for Rigid biomaterial, IBS-R). These cylindrical inserts (2 mm thick, 15.5 mm high and 15.5 mm in diameter) were machined from medical-grade PTFE (Fluteck P2000, Sirris, Liège, Belgium) (Hardinge Model HLV-H Precision, Hardinge Inc., Elmira, NY, United States). In the thickness of the bottom of the cylinder, a 1 mm high and 1 mm thick bevel was milled in order to clip it to a disk sample with a light friction ( Fig. 3 ). 135 “disk-insert” systems were assembled: 3 (technical triplicate) for each material (PICN, Ti, Zi, eM, PTFE) and for each incubation time (24 h, 48 h and 72 h), the test being triplicated with cells coming from 3 different patients (biological triplicate). The systems were distributed in 24-well plate cultures and 2.10 4 cells in 1 ml of culture medium were seeded on each disk. After the different incubation times, the PTFE inserts were removed from the discs.
Cell viability assay
The living cell population on the different materials was evaluated by a viability test (MTS assay, Promega, Madison, USA) directly on discs at 24 h, 48 h and 72 h. Cellular viability of 100% was attributed to the HGK grown in control polystyrene (PS) wells without discs. Cellular viability was quantified by a colorimetric assay using 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (inner salt MTS, Promega, Madison, USA). MTS solutions were prepared according to the manufacturer’s instructions. Discs were rinsed with 1 ml of DMEM/F-12 (Gibco) and 1 ml of fresh DMEM/F-12 with MTS solution (10%) was applied. The plates were incubated at 37 °C in the absence of light for 45 min. The plates were then shaken for 15 s. 200 μl of supernatant was removed from each well and placed in 96 well microplates. The absorbance of supernatant aliquots was read at 492 nm using the Powerwave X microplate spectrophotometer (Biotek instrument Inc., Winooski, VT, USA) and the viability was calculated and normalized from the absorbance of control samples taken as 100% (PS).
After the MTS assay, discs were rinsed with PBS and the cells were fixed with 4% paraformaldehyde (Sigma-Aldrich) at room temperature for 20 min. An immuno-staining was then performed and the number of cells, their covering on discs and their morphology were determined from microscopic fluorescent images. Cell permeabilization was performed with 0.5% Triton X-100 (Sigma) at 4 °C for 20 min. Blocking was performed with 1% BSA (Sigma) in PBS at 37 °C for 1 h. Vinculin was stained with mouse anti-vinculin (Sigma) and rabbit anti-mouse alexa fluor 568 (Life Technologies, Carlsbad, California, United States), the incubation was performed in 0.05% Tween-20 (Sigma) in PBS solution with corresponding dilutions (1:200), at 37 °C for 1 h for mouse anti-vinculin and at room temperature for 30 min for rabbit anti-mouse alexa fluor 568. Actin was stained by incubating in 0.05% Tween-20 (Sigma) in PBS solution with 1:40 dilution of Alexa fluor 488-labeled phalloidin (Life Technologies) at 37 °C for 1 h. A nuclear stain dye DAPI (4′,6-diamidino-2-phenylindole dihydrochloride) (D8417-1MG, Sigma)/PBS 1:5000 was added and allowed to set in at room temperature for 10 min. The stained sample surface was imaged with an IX81 optical inverted microscope equipped with a UPlanFL objective at 10× magnification and with an XCite-iris IX fluorescence unit and a C-BUN-F-XC50 charge-coupled-device camera (Olympus Optical Co., Ltd). An image analysis software “CellSens” (Olympus) was used to quantify the number of cells and their coverage on discs. 350 pictures (10× objective) per sample were performed to allow the analysis of the entire disk surface. Cell spreading was calculated by dividing the covered surface by the number of cells. A qualitative analysis of cellular morphology on discs was conducted with the fluorescent actin and DAPI staining at the three time points with the 10× objective and with the fluorescent actin, DAPI and vinculin staining at 72 h with a UPLSAPO 20× oil objective (Olympus). It must be noted that the observed activated form of vinculin, which mediates cell adhesion, appears at the cell periphery in well-organized focal adhesion plaques while the vinculin present in the cytosol is the inactivated form, which does not contribute to adhesion .
For each variable, the data were treated by a variance analysis with respect to fixed effects (materials, time and interactions between them) and random effects. Materials were compared two by two by multiple comparisons of Scheffe. To highlight the part of the variation due to the subject and due to the disk specimen, the variance components (VC) were calculated by time and material. The level of significance was set at 5% (p < 0.05). The calculations were carried out by means of the software SAS version 9.4 (SAS Institute, Cary, NC, the USA).
Analysis of monomer release using ultra high-performance liquid chromatography coupled to mass tandem spectrometry (UPLC–MS/MS)
The UDMA release from PICN samples was quantified by mass spectrometry. Because they do not contain monomers, eM samples were used as negative controls. 3 PICN and 3 eM discs were incubated at 37 °C for 72 h with 1 ml of complete medium used for HGK culture (KSFM). 20 μl of a 10 mg/l internal standard solution (benzophenone-3 d5 98% chemical purity from C/D/N Isotopes Inc., Quebec, Canada) and 1 ml of Milli-Q water (>18 MΩ) were added to 0.5 ml of KSFM, and extracted twice with 2 ml of ethyl acetate. Combined organic phases were evaporated until dryness under a gentle stream of nitrogen at 30 °C, and reconstituted in 100 μl of a 70:30 (v:v) water–acetonitrile solution prior analysis by UPLC–MS/MS. The Acquity UPLC system (Waters, Milford, MA, USA) was equipped with an Acquity HSST3 column (2.1 × 100 mm, 1.8 μm) from Waters, maintained at 35 °C. The injection volume was 10 μl. The mobile phases A and B consisted, respectively in water acidified by 0.1% of formic acid and acetonitrile. The gradient started at a constant flow of 0.4 ml/min with 70% A held for 1 min, then linearly decreased to 10% A in 9 min, held for 2 min, returned to the initial conditions in 0.5 min and finally maintained for 2.5 min prior to the next injection. The Quattro Premier XE mass spectrometer (Waters) operated in positive electrospray mode with the capillary and cone voltages set at 3.2 kV and 30 V respectively. The source and desolvatation temperatures were set at 120 °C and 350 °C, with nitrogen used as cone and desolvatation gas at a flow of 50 l/h and 800 l/h. The collision gas was argon (99.99997%, Air Liquide, Liege, Belgium) settled at a flow of 0.15 ml/min. Two transitions for UDMA were monitored (493.2 > 407.2 and 113.2 > 69.3) in multiple reactions monitoring (MRM) to check ion ratios and ensure proper compound identification. The transition 234.16 > 151.2 was selected for the internal standard (benzophenone-3 d5). The collision energy was set at 30 V for both UDMA transitions and 21 V for benzophenone-3 d5. The quantification was performed using a matrix-matched calibration curve built by spiking KSFM with increasing levels of UDMA to obtain 8 points from 0.5 to 100 ng/ml. The limit of quantification was set at 0.5 ng/ml. Moreover, a home-made quality control was also analyzed, consisting in KSFM sample spiked at 50 ng/ml.
Indirect cell contact cytotoxicity assay
L929 mouse fibroblasts were grown in DMEM (Gibco) supplemented with 10% FBS (Gibco) and 1% penicillin–streptomycin (Lonza) under standard cell culture conditions (37 °C and 5% CO 2 ). The cytotoxicity of PICN and eM materials was tested as follows using a 24-well plate: 6 wells were filled with 1 ml of DMEM complete medium covering a PICN disk (DMEM + PICN), 6 with 1 ml of DMEM complete medium covering an eM disk (DMEM + eM), and 6 with 1 ml of DMEM complete medium without a disk (DMEM). The plate was placed in the incubator. The conditioned DMEM was collected after 24 h for half of the wells, and after 72 h for the other half. For the cytotoxicity assay, L929 fibroblasts were seeded into 96-well plates at 2000 cells/well and incubated for 24 h to allow cell adhesion (n = 21 wells). The medium was then replaced by 200 μl of conditioned DMEM from the 3 different groups (n = 18 wells). Moreover, three wells served as a control (Control), with cells exposed to non-conditioned (fresh) DMEM complete medium. After an incubation of 72 h, cell viability was evaluated using MTS assays (Promega), as described previously. Cell viability was calculated as a percentage of the control group. The data were treated by a variance analysis with respect to fixed and random effects. The level of significance was set at 5% (p < 0.05).