ARTICLE IN PRESS teration observed in our conditions. Moreover, the present limited three points particles at 100 ug/mL in DMEM. Results are presented in the Fig. 4. very reproducible. The histological assessment was made in two times. l From 3 to 15 days of culture, we observed a regular increase of the ncluding the synovial lining su cell growth for both types of cell. The cell proliferation was not field. The synod embrane thickness and the villous hyperplasia were scored. affected by the presence of alumina or zirconia particles. Indeed, no hereafter, using 40x power field we examined separately the synovial lining cell significant differences were observed in the cell proliferation synovial tissue. Results presented were the between cultures with or without particles. The study of the cell bserved in five continuous selected helds from the proximal region of the synovial viability by blue trypan exclusion confirmed this result(data not ssue near the patella. In the synovial lining cell layer, we scored the cell layer shown). In order to assess the cell ability to produce ECM, all DMEM perplasia(number of cell layers)and the cell hypertrophy(o= normal flat cells, cuboidal cells, 2= cylindrical cytoplasmic extensions) In the sub-synovial supernatants were collected for CIH and FN assays Results are gion, the cell infiltration was evaluated by scoring the amount of presented in the Fig. 4c-f. Both CIH or FN productions and cell istiocytes cells(O=no or very low. 1=moderate and 2= high amount) and growth increased in comparable way. Compared to the control aggregates). Finally, vascularisation was scored (0=f culture (without particles). cells cultured in the presence of 1= moderate increase and 2 alumina or zirconia at concentration of 100 ug/mL did not affect these two proteins synthesis Results were similar for fibroblasts and osteoblasts from 3 to 15 days of culture. 3.2.2. Particles at 1000 ug/mL In order to evaluate the influence of the particles concentration on (r)and Spearman (r) correlation tests using GraphPad Prism version 4.02 software cell functions, we reproduced accurately all this experiment with ten for Macintosh(GraphPad Software, San Diego, CA, USA). Statistical significance was fold more particles(1000 ug/mL) Due to this particles amount and considered at p<0.05. All in vitro experiments were run in triplicate and repeated at compared to the cell number and the cell culture surface, we quickly observed an important particle layer that recovered all cells. This extreme condition for an in vitro study. 3. Results However, fibroblasts and osteoblasts growth were not really affected by this particles amount. Results are presented in the Fig. 5. Compared 3. 1. In vitro cytocompatibility of ceramic discs with control culture, we observed just a light increase of the fibroblast growth when cultured in the presence of alumina aPA(p 0.05) and Compared to cells cultured on plastic (reference substrate rep- zirconia Tz-0(p <0.05)after 15 days. This result was associated with sented by an empty well), fibroblasts and osteoblasts proliferation an increase of the fibronectin synthesis in the pi e of both decreased on both rough and polished ceramics discs tested in this alumina or zirconia particles at the same culture period (respectively study at all time periods even if the differences were not always p<0.01 and p<0.001 compared with control culture). However, no tatistically significantFig3 a-b). The proliferation of fibroblasts significant variation of collagen synthesis was detected In the pres- was significantly decreased on rough discs after 3 and 6 day of ence of the same particle amount, osteoblasts growth was also not ulture compared to the control(Fig. 3a, P<0.01). For the osteo- affected by the presence of alumina or zirconia after 3, 6 and 10 days of blasts, the proliferation was significantly lower on rough ceramics culture. It was just slightly reduced at day 15(p<0.01)for both discs after 3, 6 and 10 days of culture(respectively p<0.01, 0.05 and alumina and zirconia. At this time, this variation was associated with p<0.05 and p <0.01)than control( Fig. 3b). The cell functions were Regarding the results of cell growth and ECM synthesis in our study assessed by the assays of two important proteins of the extra-cellular conditions, alumina and zirconia particles showed no cytotoxiceffect. matrix produced by fibroblasts and osteoblasts: human type I collagen and fibronectin(Fig. 3c-f). Compared to respective control 3. 3. Cell morphology in the presence of ceramic particles cultures, the levels of the two proteins were generally weaker but not significantly decreased for both fibroblasts and osteoblasts. Two The present study aimed at investigating the influence of exceptions were observed, only for cells cultured on rough ceramic. particles on osteoblasts and fibroblasts morphology and spreading. The first one concerned the production of CIH by fibroblasts after 10 Just after addition of alumina and zirconia in the cell culture media, days( Fig 3c, p<0.01)and the second concerned the production of we observed a slight particle aggregation in the cell culture fibronectin by osteoblasts after 6 days of culture( Fig. 3f, p<0.05). medium according to the granulometry studies discussed in para- Moreover, although not statistically significant, the production of the graph 2.1. As with the cells cultured without particles, cells extra-cellular matrix(ECM) was always greater in the presence of morphology and spreading appeared normal when cultured in the the polished ceramic than with the rough one For respective cells presence of alumina or zirconia for both fibroblasts and osteoblasts and substrates, these variations of the production of the ECM fol- up to 15 days of contact. After incubation with both alumina lowed the evolution of the proliferation zirconia powders, fibroblasts and osteoblasts quickly endocytosed an important proportion of particles. This internalization seems to 3. 2. In vitro cytocompatibility of ceramic particles start after a 2 h d appears clearly after 24 h(data not shown). Particles could be seen within and around the cells(the In order to assess the effect of the ceramic constituents, we also particles appears in black and are indicated by arrows in Fig. 6). The performed in vitro studies with particles. Fibroblasts and osteo- intracellular localization was especially peri-nuclear. lasts were cultured in the presence of alumina APA or zirconia tzO articles at two different concentrations(100 and 1000 ug/mL) 3.4. In vivo biocompatibility of alumina and zirconia particles luring 15 days. At 3, 6, 10 and 15 days, cell growth, CIH and FN synthesis were measured The rats well tolerated the particles or physiological fluid injection and ambulated without any pain or locomotion difficulties. Animals 3.2.1. Particles at 100 ug/mL showed no inflammation nor infection signs in all groups and all The first step of this study was the assessment of fibroblasts and along the 6 weeks of the study. These observations were confirmed by osteoblasts proliferation in the presence of alumina or zirconia autopsies. During dissection, the macroscopic examination revealed lease cite this article in press as: Roualdes O, et al, In vitro and in vivo evaluation of an., Biomaterials(2009), doi: 10.1016/ j biomaterials 2009. 11.107alteration observed in our conditions. Moreover, the present limited three points scale was very reproducible. The histological assessment was made in two times. At first, we performed a general histological evaluation of the whole synovial tissue including the synovial lining surface and the underlining stroma using 5 power field. The synovial membrane thickness and the villous hyperplasia were scored. Thereafter, using 40 power field we examined separately the synovial lining cell layer and the sub-synovial tissue. Results presented were the mean of the scores observed in five continuous selected fields from the proximal region of the synovial tissue near the patella. In the synovial lining cell layer, we scored the cell layer hyperplasia (number of cell layers) and the cell hypertrophy (0 ¼ normal flat cells, 1 ¼ cuboı¨dal cells, 2 ¼ cylindrical  cytoplasmic extensions). In the sub-synovial region, the cell infiltration was evaluated by scoring the amount of macrophages/ histiocytes cells (0 ¼ no or very low, 1 ¼ moderate and 2 ¼ high amount) and lymphocytes (0 ¼ no or very low, 1 ¼ moderate/diffuse infiltration and 2 ¼ high infiltration or aggregates). Finally, vascularisation was scored (0 ¼ normal, 1 ¼ moderate increase and 2 ¼ high increase). 2.8. Statistical analysis Cell proliferation, human fibronectin and type I collagen levels and in vivo data were analyzed statistically using a standard unparametric analysis of variance techniques (ANOVA Kruskall walis) following Dun’s multiple range test or Pearson (r0 ) and Spearman (r) correlation tests using GraphPad Prism version 4.02 software for Macintosh (GraphPad Software, San Diego, CA, USA). Statistical significance was considered at p < 0.05. All in vitro experiments were run in triplicate and repeated at least 3 times. Data presented on graphs or in the text are means  SEM. 3. Results 3.1. In vitro cytocompatibility of ceramic discs Compared to cells cultured on plastic (reference substrate rep￾resented by an empty well), fibroblasts and osteoblasts proliferation decreased on both rough and polished ceramics discs tested in this study at all time periods even if the differences were not always statistically significant (Fig. 3 a–b). The proliferation of fibroblasts was significantly decreased on rough discs after 3 and 6 day of culture compared to the control (Fig. 3a, p < 0.01). For the osteo￾blasts, the proliferation was significantly lower on rough ceramics discs after 3, 6 and 10 days of culture (respectively p < 0.01, 0.05 and p < 0.001) and on polished discs after 3 and 10 days (respectively p < 0.05 and p < 0.01) than control (Fig. 3b). The cell functions were assessed by the assays of two important proteins of the extra-cellular matrix produced by fibroblasts and osteoblasts: human type I collagen and fibronectin (Fig. 3c–f). Compared to respective control cultures, the levels of the two proteins were generally weaker but not significantly decreased for both fibroblasts and osteoblasts. Two exceptions were observed, only for cells cultured on rough ceramic. The first one concerned the production of CIH by fibroblasts after 10 days (Fig. 3c, p < 0.01) and the second concerned the production of fibronectin by osteoblasts after 6 days of culture (Fig. 3f, p < 0.05). Moreover, although not statistically significant, the production of the extra-cellular matrix (ECM) was always greater in the presence of the polished ceramic than with the rough one. For respective cells and substrates, these variations of the production of the ECM fol￾lowed the evolution of the proliferation. 3.2. In vitro cytocompatibility of ceramic particles In order to assess the effect of the ceramic constituents, we also performed in vitro studies with particles. Fibroblasts and osteo￾blasts were cultured in the presence of alumina APA or zirconia TZ0 particles at two different concentrations (100 and 1000 mg/mL) during 15 days. At 3, 6, 10 and 15 days, cell growth, CIH and FN synthesis were measured. 3.2.1. Particles at 100 mg/mL The first step of this study was the assessment of fibroblasts and osteoblasts proliferation in the presence of alumina or zirconia particles at 100 mg/mL in DMEM. Results are presented in the Fig. 4. From 3 to 15 days of culture, we observed a regular increase of the cell growth for both types of cell. The cell proliferation was not affected by the presence of alumina or zirconia particles. Indeed, no significant differences were observed in the cell proliferation between cultures with or without particles. The study of the cell viability by blue trypan exclusion confirmed this result (data not shown). In order to assess the cell ability to produce ECM, all DMEM supernatants were collected for CIH and FN assays. Results are presented in the Fig. 4c–f. Both CIH or FN productions and cell growth increased in comparable way. Compared to the control culture (without particles), cells cultured in the presence of alumina or zirconia at concentration of 100 mg/mL did not affect these two proteins synthesis. Results were similar for fibroblasts and osteoblasts from 3 to 15 days of culture. 3.2.2. Particles at 1000 mg/mL In order to evaluate the influence of the particles concentration on cell functions, we reproduced accurately all this experiment with ten fold more particles (1000 mg/mL). Due to this particles amount and compared to the cell number and the cell culture surface, we quickly observed an important particle layer that recovered all cells. This concentration was an extreme condition for an in vitro study. However, fibroblasts and osteoblasts growth were not really affected by this particles amount. Results are presented in the Fig. 5. Compared with control culture, we observed just a light increase of the fibroblast growth when cultured in the presence of alumina APA (p < 0.05) and zirconia TZ-0 (p < 0.05) after 15 days. This result was associated with an increase of the fibronectin synthesis in the presence of both alumina or zirconia particles at the same culture period (respectively p < 0.01 and p < 0.001 compared with control culture). However, no significant variation of collagen synthesis was detected. In the pres￾ence of the same particle amount, osteoblasts growth was also not affected by the presence of alumina or zirconia after 3, 6 and 10 days of culture. It was just slightly reduced at day 15 (p < 0.01) for both alumina and zirconia. At this time, this variation was associated with a light but not significant decrease of CIH synthesis and increase of FN. Regarding the results of cell growth and ECM synthesis in our study conditions, alumina and zirconia particles showed no cytotoxic effect. 3.3. Cell morphology in the presence of ceramic particles The present study aimed at investigating the influence of particles on osteoblasts and fibroblasts morphology and spreading. Just after addition of alumina and zirconia in the cell culture media, we observed a slight particle aggregation in the cell culture medium according to the granulometry studies discussed in para￾graph 2.1. As with the cells cultured without particles, cells morphology and spreading appeared normal when cultured in the presence of alumina or zirconia for both fibroblasts and osteoblasts, up to 15 days of contact. After incubation with both alumina and zirconia powders, fibroblasts and osteoblasts quickly endocytosed an important proportion of particles. This internalization seems to start after a 2 h contact, and appears clearly after 24 h (data not shown). Particles could be seen within and around the cells (the particles appears in black and are indicated by arrows in Fig. 6). The intracellular localization was especially peri-nuclear. 3.4. In vivo biocompatibility of alumina and zirconia particles The rats well tolerated the particles or physiological fluid injection and ambulated without any pain or locomotion difficulties. Animals showed no inflammation nor infection signs in all groups and all along the 6 weeks of the study. These observations were confirmed by autopsies. During dissection, the macroscopic examination revealed 4 O. Roualdes et al. / Biomaterials xxx (2009) 1–12 ARTICLE IN PRESS Please cite this article in press as: Roualdes O, et al., In vitro and in vivo evaluation of an..., Biomaterials (2009), doi:10.1016/ j.biomaterials.2009.11.107