Abstract
Aim: The biocompatibility of human osteoblasts (HOB) and human unrestricted somatic stem cells (USSCs) with membranes (BioGide®, GORE-TEX®, GENTA-FOIL resorb®, RESODONT®, BioMend®, BioMend® Extend™) was evaluated. Materials and Methods: After osteogenic differentiation (dexamethasone, ascorbic acid and β-glycerolphosphate) cells were seeded on membranes. On days 1, 3 and 7, attachment, proliferation, cell vitality, cytotoxicty and cell morphology were analyzed. Results: Cells on BioGide® and RESODONT® exhibited significantly higher attachment (p<0.005) and proliferation (p<0.005). On BioMend® cells showed a significantly higher attachment compared to BioMend® Extend™ (p<0.005), whereas on BioMend® Extend™ cells had significantly higher proliferation (p<0.005). The vitality of cells was significantly better on BioGide® and RESODONT® (p<0.005). There were no significant differences between USSCs and HOBs. Scanning electron microscopy confirmed these results. Conclusion: BioGide® and RESODONT® had the best biocompatibility and are appropriate membranes for use in stem cell-derived regeneration of bone.
Guided bone regeneration (GBR) is a well-established procedure in maxillofacial surgery for treatment of bony defects and for augmentation procedures. The principle of this technique is to inhibit the growth of soft tissue into bony defects by covering them with membranes. Therefore a regeneration of bone is possible (1-3). There are multiple characteristics demanded for barrier membranes. Membranes should be biocompatible, permit nutrition of the cells and vascularization below the membrane, furthermore membranes should also be cell-occlusive to avoid the growth of soft tissue into bony defects. In addition, membranes must, at the same time, offer an adequate rigidity to resist the pressure of the surrounding soft tissue and must be flexible so that they can easily be adapted to the clinical situation (4).
Membranes can be classified according to their origin into native or synthetic membranes and into resorbable and non-resorbable membranes. Disadvantages of synthetic, non-resorbable membranes are their tendency for spontaneous exposure to the oral cavity, with subsequent bacterial colonization (5-6) and the need for a second surgery to remove the membrane. Pihlstrom et al. showed that every flap elevation for periodontal disease leads to a crestal resorption of the alveolar bone (7). Thus a second surgery may diminish the obtained regenerated bone. Native membranes usually consist of porcine, bovine or equine collagen I and III and have considerable advantages compared to synthetic membranes. Collagen supports the formation of a coagulum, is chemotactic for fibroblasts, and supports wound healing. In addition, these membranes are semi-permeable and thus enable the nutrition of the covered tissue (8, 9). The enzymatic biodegradation of the membranes by collagenases can be prolonged by several cross-linking techniques (10, 11), such as ultraviolet light, diphenylphosphorylazide, glutaraldehyde, or hexamethylenediisocyanate (9, 12-14). However, it is well-known that chemical-mediated cross-linking leads to reduced biocompatibility of the membranes (15).
In recent years, there has been a change in research for bony reconstruction. Cell-based tissue engineering approaches are considered to support the regeneration of bone by the transplantation of living osteogenic cells (16). Therefore cells have been cultivated and differentiated towards an osteoblastic linage in vitro by the use of different osteogenic stimuli. It is generally accepted that the formation of minerals in a stem cell culture indicates ostoblastic differentiation of the stem cells (17, 18). Furthermore, it has been shown that addition of dexamethasone, ascorbic acid and β-glycerol phosphate (DAG) to culture medium triggers osteogenic differentiation of various types of stem cells [e.g. mesenchymal stem cells (MSC) and embryonic stem cells (ESC)] (19-22). In a previous study, we showed that osteogenic stimulation with DAG exceeds the stimulation with bone morphogenetic protein-2 (BMP-2) in ESCs (20). The mineral-inductive potential of ESCs was also demonstrated in a previous animal study in rats, and ESCs were able to promote ectopic bone formation in vivo when they were used with de-mineralized bone (23). A few years ago, a promising stem cell source, unrestricted somatic stem cells (USSC), were established by Kögler and colleagues (21). These cells derived from umbilical cord blood and had the ability to develop into mesodermal, endodermal and ectodermal cells. These cluster of differentiation 45 (CD45) and human leukocyte antigen (HLA) class II-negative stem cells display proliferative capacity in vitro without spontaneous differentiation. It is possible to expand the cells to 1015 cells without losing multipotency. In vitro, a differentiation into osteoblasts, chondroblasts, hematopoietic and neural cells is possible using specific stimuli (21). In previous studies, we showed that stimulation with DAG leads to mineralization both in vitro (24) and in vivo (25). Compared to ESCs, USSCs do not suffer any immunological rejection and have fewer ethical and legal restrictions. The use of USSCs is thus much closer to clinical practice than that of ESCs, i.e. hematopoietic cells from cord blood are already used in the therapy of hematopoietic and genetic disorders (26). The aim of this study was to analyze the biocompatibility of different commercially available membranes with osteogenic pre-differentiated USSCs, as a pre-condition for an enhanced bony regeneration by the combination of GBR with tissue engineering approaches. We analyzed the biocompatibility of different membranes with human USSCs in vitro and compared it to the biocompatibility of membranes with human osteoblasts.
Membranes analyzed in this study with their origin, collagen type, cross-linking status, resorptability and the supplying company.
Materials and Methods
Membranes. We included six commercially available, clinically well-established membranes in this study: BioGide®, RESODONT®, GENTA-FOIL resorb®, BioMend®, BioMend® Extend™ and GORE-TEX® (Table I). In order to standardize the dimension of the surface, membranes were trimmed under sterile conditions to a circle with a diameter of 1 cm.
Cultivation of the cells and osteogenic pre-differentiation of the USSCs. USSCs were kindly provided by the José Carreras Stem Cell Bank (Heinrich Heine University of Düsseldorf, Germany). Cells were isolated from cord blood with informed consent from the mother, and cultivated in accordance with a standardized protocol published by Kögler et al. (21). Briefly, Ficoll (Biochrom, Berlin, Germany) gradient centrifugation was used to isolate the mononuclear cell fraction. Cells were plated out at 5-7×106 cells/ml on T25 culture flaks (Costar, Amsterdam, The Netherlands) in low glucose Dulbecco's modify medium (DMEM) (Cambrex, Wiesbaden, Germany), supplemented with 30% Fetal Calfs Serum (FCS), dexamethasone (10−7 M; Sigma-Aldrich, St. Gallen, Switzerland), penicillin (100 U/ml; Grünenthal, Aachen, Germany), streptomycin (100 mg/ml; Hefa-pharm, Werne, Germany) and ultraglutamine (2 mM; Cambrex, Wiesbaden, Germany). Later on in the expansion of the cells, dexamethasone was omitted from the medium. The cells were incubated in a humidified atmosphere at 37°C in 5% CO2. When confluency reached 80%, the cells were split by detachment with 0.25% trypsin (Lonza, Frankfurt, Germany) and re-plating at a ratio of 1:3. The medium was changed every day. Cells were cultivated in a monolayer technique. Osteogenic pre-differentiation was performed by the addition of 0.1 μM dexamethasone, 50 μM ascorbic acid and 10 mM β-glycerolphosphate (all from Sigma-Aldrich) to the normal growth medium, according to previously published work (19, 27-28). Osteogenic pre-differentiation was performed for three days prior to the seeding of the membranes. Comercially available HOBs (PromoCell, Heidelberg, Germany) of three different donors (two female, one male) were used. Cells were plated out at 5 to 7×106 cells/ml on T25 culture flaks (Costar) in low glucose DMEM (Cambrex), supplemented with 30% FCS, penicillin (100 U/ml; Grünenthal) and streptomycin (100 mg/ml; Hefa-pharm). The cells were incubated in a humidified atmosphere at 37°C in 5% CO2. When confluency reached 80%, the cells were split by detachment with 0.25% trypsin (Lonza) and re-plating at a ratio of 1:3. The medium was changed every day. Membranes were placed into 24-well plates and were rehydrated with 610 μl cell-specific medium for 5 min before seeding of cells was performed. Subsequently 19,000 cells suspended in 190 μl medium were placed on each membrane, resulting in a total volume of 800 μl. To minimize the impact of cells not linked to a membrane, specimens were placed into new 24-well plates after 24 h. Specimens were cultured for seven days; media were changed on day four. After 24 h, three and seven days, a fraction of the specimen was collected for analytic purpose. Experiments were repeated three times.
Overall attachment of unrestricted somatic stem cells (USSC) and human osteooblasts (HOB) on different membranes. Cells on BioGide® and RESODONT® exhibited significantly greater attachment compared to the other membranes (p<0.005). Attachment to BioMend® and GORE-TEX® was significantly higher than that to BioMend® ExtendTM and GENTA-FOIL® (p<0.005). Controls exhibited the best attachment (p<0.005).
CyQuant® assay. To assess the attachment and proliferation of cells on the various membranes, a CyQUANT® assay (CyQUANT Cell Proliferation Assay Kit®; Fa. Invitrogen, Karlsruhe, Germany) was performed on days 1, 3 and 7. The CyQUANT® cell proliferation assay is a highly sensitive, fluorescence-based microplate assay, for determining the numbers of cultured cells. The assay employs the CyQuant GR dye, which produces a large fluorescence enhancement upon binding to cellular nucleic acids that can be measured using standard fluorescein excitation (485 nm) and emission (535 nm) wavelengths. Fluorescence emission of the dye nucleic acid complexes correlates linearly with the cell number over a large range using a wide variety of cell types. Under the recommended assay conditions, the readouts of the experiments lay well within the detection limits. The assay was performed as previously described (29).
Vitality of the cells. Cell vitality was analyzed with a CellTiter-Blue® Cell Viability Assay (Promega, Madison, WI, USA) on day 1, 3 and 7 according to the manufacturer's instructions. Only metabolically active vital cells can convert the dye resazurin (blue) into resofurin (pink). Measured fluorescence emission of the dye-conversation (560 nm) correlates with the amount of resofurin and therefore with the vitality of the cells.
Cytotoxity of the membranes. Cytotoxity of membranes was measured on days 3 and 7 with Cyto Tox-ONE™ Homogenous Membran Integrity Assays (Promega), according to the manufacturer's instructions. Due to the fact that dead cells release lactatedehydrogenase (LDH), lactate is converted to pyruvate by the reduction of NAD+ to NADH, and NADH converts resazurin (blue) into resofurin (pink), in such a way that the measured fluorescence emission of the dye conversation (560 nm) correlates with the amount of LDH. This is in direct proportion to the cytotoxity of membranes.
Scanning electron microscopy (SEM). For SEM, membranes were fixed for 8 min with 2.5% glutaraldehyde in 0.1 M PBS (pH=7.3) and subsequently washed in 0.1 M PBS for 30 min three times. Samples were dehydrated in increasing concentrations of acetone (from 50% to 100%, 10% steps). After critical-point drying using CO2 as transitional fluid (Bal-Tec Dryer CPD-030; BAL-TEC GmbH Schalksmuehle, Germany), specimens were sputter-coated with gold (Cressington Sputter 108Auto; Cressington Scientific Instruments Ltd., Watford, UK), and observed in a scanning electron microscope (REM S 3000; Hitachi, Krefeld, Germany) with an acceleration voltage of 15 kV.
Statistical analysis. A software package (SPSS 19.0, SPSS Inc., Chicago, Il, USA) was used for the statistical analysis. Mean values and standard deviations were calculated for each group. Results were considered statistically significant at p<0.05.
Results
There were no signs of any bacterial or fungal contamination of the specimens during the experiment. To gain an overview of the general behavior of osteogenic cells on membranes, we first analyzed the results without any discrimination between USSCs and HOBs. The CyQuant® assay showed a significantly higher attachment of cells both USSCs and HOBs on BioGide® and RESODONT® compared to the other membranes (p<0.005). In addition, attachment was significantly higher on BioMend® and GORE-TEX® compared to BioMend® Extend™ and GENTA-Foil® (p<0.005) (Figure 1). Analysis of the proliferation of cells showed a significantly higher proliferation from day 1 to 3 and from day 1 to 7 on BioGide® and RESODONT®, compared to the other membranes (p<0.005). Proliferation of the cells was significantly higher on GENTA-FOIL® and BioMend® Extend™ compared to BioMend® and GORE-TEX®. Analysis of the vitality of the cells confirmed the findings of the CyQuant® assay. BioGide® and RESODONT® led to significantly superior vitality of cells on the membranes on day 3 and 7 of the experiment compared to other membranes (p<0.005). Analysis of the cytotoxity of membranes showed no significant differences between the membranes, but a statistically significant higher cytotoxity of all membranes compared to the control (cells cultured without any membrane) was detected (p<0.005). In addition, controls showed significantly better attachment, proliferation and vitality of cells compared to the tested membranes.
Overall attachment of unrestricted somatic stem cells (USSC) and human osteooblasts (HOB) to different membranes. Attachment was best to BioGide® and RESODONT® compared to the other membranes (p<0.005). USSCs had similar behavior on different membranes. Controls exhibited the best attachment (p<0.005).
In order to analyze if there were differences between the cell lines regarding the biocompatibility, we studied them separately. We confirmed the above mentioned findings regarding the biocompatibility of the membranes for USSCs and HOBs and found that USSCs and HOBs have a similar behavior on membranes. There were no statistically significant differences between the cell lines regarding the attachment, proliferation, cell vitality and cytotoxicity on the different membranes. Figure 2 shows an example of the results for the attachment of the cells on the different membranes. Scanning electron microscopy after 7 days, confirmed the findings regarding the biocompatibility of the membranes with the cells. Only on BioGide® and RESODONT® was a dense cell layer detectable. Furthermore, both cell lines had a flat appearance, with cytoplasmatic extensions approaching confluency on BioGide® and RESODONT®. On the other membranes, only a few round-shaped cells were observed (Figure 3).
Discussion
The principle of GBR is to inhibit the growth of soft tissue, especially of the gingival epithelium, into bony defects and to generate a space that can be infiltrated by bony cells and fibroblasts (30-31). Therefore defects are covered with membranes which are cell-occlusive and offer an adequate rigidity to resist the pressure of the surrounding soft tissue (32-33). Sculean et al. showed that a combination of GBR with grafting materials in two wall intrabony defects, gave superior histological results in bone repair to barrier membranes alone (34). A combination of osteogenic pre-differentiated USSCs with biomaterials may additionally enhance bony regeneration by GBR. The biocompatibility of membranes or biomaterials with osteogenic differentiated USSCs is a pre-condition for such an approach. In a previous study, we analyzed the biocompatibility of different biomaterials with USSCs and showed that insoluble collagenous bone matrix followed by β-tricalciumphosphate is highly suitable for bone tissue engineering regarding cell attachment and proliferation (35).
Scanning electron microscopy of membranes with unrestricted somatic stem cells (USSC) and human osteooblasts (HOB) after 7 days. On BioGide® and RESODONT®, a dense layer of flat cells with multiple lamellopodia was detected, reflecting a strong cell attachment. On the other membranes, only few round-shaped cells were detectable, reflecting low attachment and proliferation rates. To emphasize the different appearance of the cells on the membranes one cell is stained in red in every image. (Magnification × 500).
In the present study, we analyzed the biocompatibility of different commercially available membranes with osteogenic pre-differentiated USSCs in vitro and compared the biocompatibility of USSCs with HOBs. HOBs were thought to be naturally osteogenic reference cells. We analyzed the biocompatibility in a two-dimensional culture system because it provides a defined environment and thereby cell membrane interactions can be very well analyzed. It is well-known that collagenous membranes are superior to synthetic membranes regarding attachment and proliferation of cells, because collagen plays an active role in coagulum formation and is chemotactic for fibroblasts (8, 36-37). The present study highlights the advantage of collagen membranes. On BioGide®, RESODONT® and BioMend® cells exhibited significantly higher attachment compared to GORE-TEX® (p<0.005). Furthermore, USSCs proliferated significantly better on BioGide®, RESODONT®, GENTA-FOIL® and BioMend® Extend™ compared to GORE-TEX® (p<0.005). However, cells on GORE-TEX® had a significantly superior attachment compared to BioMend® Extend™ and GENTA-FOIL®. Due to the fact that there was no correlation between the origin (porcine, bovine, equine) of the collagen and the results, the origin of the collagen does not appear to influence the biocompatibility of membranes. Interestingly, cells exhibited significant differences regarding attachment and proliferation on BioMend® and BioMend® Extend™. While on BioMend® cells had a significantly higher attachment compared to BioMend® Extend™, it was vice versa regarding the proliferation of the cells. These findings are in concordance with the literature, as Rothamel et al. demonstrated the same findings with human periodontal ligament (PDL) fibroblasts and human osteoblast-like cells (15). Differences regarding proliferation and attachment of cells between BioMend® and BioMend® Extend™ may be caused by different levels of cross-linking of the membranes, e.g. by the use of glutaraldehyde in order to prolong biodegradation. Different work groups correlated increased cross-linking with reduced biocompatibility and increased cytotoxic effects (38-39). Furthermore, Rothammel et al. demonstrated in vivo that cross-linking of bovine and porcine-derived collagen types I and III with glutaraldehyde is associated with reduced tissue integration and vascularization (9). In our study, no significant differences regarding the cytotoxity between the membranes was detected. Interestingly, there were significant differences between the biocompatibility of cells on membranes compared to controls on petri dishes. Controls exhibited significantly better attachment, proliferation and cell vitality compared to the best membrane. These findings are in concordance with the literature (15, 40-41). These differences in biocompatibility (membranes/culture dishes) may be caused by disparities of the surface. Brunette et al. showed that the surface topography influences the attachment of cells (42). Furthermore, it is well-known that fibroblasts attach better on a smooth surface, whereas osteoblast-like cells prefer a rough surface for attachment (43-44). These findings seem to disagree with our results, namely that USSCs and HOBs attached better on smooth petri dishes compared to rough membranes. However, there is an optimum surface roughness to enhance biocompatibility. Park et al. demonstrated that a porosity of 30 to 50 nm boosted attachment and proliferation of MSCs, whereas increasing or decreasing porosity reduced biocompatibility (45). Consequently, the porosity of membranes could affect their biocompatibility, although rough surfaces support the attachment and proliferation of osteoblast-like cells in general. On the other hand, porosity also influences the osteogenic differentiation of cells. Oh et al. showed that a porosity of 70-100 nm leads to an elongation of cells and consequently to osteogenic differentiation via cytoskeletal stress (46). There is a controversial discussion about the optimal size of pores for osteogenic differentiation in combination with enhanced attachment and proliferation. Finally, the selection of pore size is a compromise between good biocompatibility and promotion of osteogenic differentiation. Another marker for the biocompatibility of membranes is the morphology of cells on the membranes. Trylovich et al. showed that flat cells with multiple lamellopodia indicate a strong attachment to the surface the are growing on (47). In contrast, round-shaped cells reflect lower attachment and lower proliferation rates compared to flattened cells (15, 48-49). The results of the present study are in concordance with these studies regarding the correlation between cell morphology and biocompatibility. SEM revealed flat cells with multiple lamellopodia on BioGide® and RESODONT®, whereas on the other membranes, cells exhibited a round morphology. Therefore the results of the SEM confirm the results of the CyQuant® assay. Even if our in vitro study does not reflect the complex in vivo conditions we conclude that osteogenic pre-differentiated USSCs exhibit behavior similar with that of HOBs on membranes. Furthermore, on BioGide® and RESODONT® were most biocompatible. Collagen membranes seem to be advantageous compared to synthetic membranes, whereas chemical cross-linking leads to reduced biocompatibility of the membrane. In addition, we showed neither that the origin of the collagen, nor the type of collagen appear to affect the biocompatibility of the membranes.
Acknowledgments
We thank the group of G. Kögler for providing USSCs. Fabian Langenbach was supported by a Deutsche Forschungsgemeinschaft (DFG) grant (HA3228/2-1).
Footnotes
-
Conflicts of Interest
The Authors declare that there are no conflicts of interest. We confirm that all co-authors have reported any potential conflicts.
- Received September 13, 2012.
- Revision received October 24, 2012.
- Accepted October 25, 2012.
- Copyright © 2013 The Author(s). Published by the International Institute of Anticancer Research.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.