- Email: [email protected]
Three-dimensional reconstruction of confocal laser microscopy images to study the behaviour of osteoblastic cells grown on biomaterials
Biomaterials 23 (2002) 397–406
Three-dimensional reconstruction of confocal laser microscopy images to study the behaviour of osteoblastic cells grown on biomaterials P.A. Ramires*, A. Giuffrida, E. Milella PASTIS-CNRSM, Biomaterials Unit, S.S. 7 Appia, Km 714, 72100 Brindisi, Italy Received 9 August 2000; accepted 7 March 2001
Abstract The adhesion, spreading and cytoskeletal organization of osteoblastic cells seeded onto titanium and titania/hydroxyapatite composite coating (TiO2/HA) were studied using images acquired by confocal laser scanning microscopy. The fluorescence staining technique was employed to visualize actin cytoskeletal organization of cells. 2-D images were exhaustive when the cells were seeded at low density (in the first 24 h of incubation), but they were less clear when the cells proliferated and appeared stacked. Since the shareware software were not satisfactory, a new 3-D image reconstruction was developed using ordinary software and a model was obtained directly from the optical section set, in order to achieve a more realistic and faithful vision of morphological structures and to evaluate the behaviour of bone cells grown on materials. The results showed that the cells grown on titanium conform to the irregular substrate surfaces maximizing the contact between the cell membrane and the substrate and proliferate disposing close to each other. On the contrary, the osteoblasts seeded onto TiO2/HA coating develop clusters where the cells aggregated extending processes in order to establish intercellular connections. Cell aggregation is an early and critical event leading to cell differentiation and mineralization process and could be a first signal of the tendency of TiO2/HA coating to stimulate cell differentiation. r 2001 Elsevier Science Ltd. All rights reserved. Keywords: 3-D image reconstruction; Confocal laser scanning microscopy; Osteoblastic cells; Titanium; titania/hydroxyapatite coating
1. Introduction The knowledge of interaction mechanisms between cells and biomaterials is particularly relevant for the performance of an implanted material and for the development of more effective devices. Cells attachment to substrates is the first step in the process of cell-surface interactions and affects subsequent cellular and tissue responses. In orthopedic and dental field the early interaction between osteoblastic cells and their substrates affects the later mineralization process and determine the longterm stability of the implanted prostheses. In fact, critical to the success of implants is the development of a stable direct connection between bone and surface implant, which must be structural and functional *Corresponding author. Fax: +390-831-507-379. E-mail address: [email protected] (P.A. Ramires).
without any intervening soft or fibrous tissue (osseointegration). The tissue response to an implant involves physical factors, depending on the implant design, surface topography, hydrophobicity, surface charge density, surface free energy and chemical factors associated with the composition and structure of the material [1–4]. These substrate characteristics may directly influence cell adhesion, spreading and signaling, events that regulate a wide variety of biological functions, including cell growth, cell migration, cell differentiation, extracellular matrix synthesis and tissue morphogenesis [5–7]. Previous reports have described the differential attachment kinetics, spreading characteristics, migration, growth and signaling cascades of various cell types, including osteoblasts, on tissue culture polystyrene, glass and orthopedic alloys [3,7–10]. Moreover, researches are focused to study the highly specialized adhesion structures, namely focal adhesion [10–13].
0142-9612/02/$ - see front matter r 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 2 - 9 6 1 2 ( 0 1 ) 0 0 1 1 8 - 1
398
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
The confocal laser scanning microscopy (CLSM) and immunofluorescence staining technique have been widely used to study cell morphology, to analyze the intracellular location of various cell components, to study the migration and the spreading of cells [14], the cytoskeleton and the focal adhesions [12,15–17] and to study apoptosis [18,19]. We have studied the behaviour of human MG63 osteoblast-like cells seeded onto titanium and titania/ hydroxyapatite composite coating (TiO2/HA) after 1 day of incubation by means of 2-D analysis carried out on CSLM images. The fluorescence staining technique was employed to visualize actin cytoskeletal organization of cells. After long time of incubation the number of cells onto the materials increases and the 2-D images do not permit to analyze exactly the cell morphology, the cell–cell interactions and the contacts with the substrate. A 3-D image reconstruction made by means of a shareware software was not satisfactory. Therefore, a new 3-D representation technique was developed using ordinary software and a model was obtained directly from the optical section set, in order to achieve a more realistic and faithful vision of morphological structures and to study the behaviour of bone cells grown on materials.
(DMEM, Biowhittaker, Belgium), containing penicillin/ streptomycin (100/100 IU), amphotericin B (2.5 mg/ml) and gentamycin (100 mg/ml), supplemented with 10% foetal calf serum (Euroclone Ltd, UK) and kept at 371C in an atmosphere of 5% CO2 and 99% humidity. Media were changed every 3 days. The cells were seeded onto the substrates at a cell density of 0.5 104 cells/cm2. On every sample, placed in a 6-multiwell plate (Corning, USA), 100 ml of a cell suspension was applied carefully and the cells were allowed to attach for 2 h to the underlying substrate, then 3.5 ml of culture medium was also added. 2.3. Staining
2. Materials and methods
After 1, 2 and 4 days of incubation the samples were fixed with 4% formaldehyde in phosphate buffer saline, permeabilized with 20 mm digitonin (Molecular Probes, Netherlands) for 5 min at room temperature and stained with 5 mg/ml phalloidin-FITC (Sigma, USA) for 1 h in order to visualize the F-actin filaments. In some cases, after a treatment with 1 mg/ml ribonuclease A (Sigma), samples were double labelled with phalloidin-FITC and 1 mg/ml propidium iodide (Molecular Probes) for cytoskeleton and nuclear staining respectively. Then, the samples were mounted under glass coverslips with SlowFades solution (Molecular Probes) to prevent quenching.
2.1. Substrates
2.4. Confocal laser scanning microscopy
Commercially pure titanium (Goodfellow, Germany), cut into pieces 20 10 mm2 size, was used as substrate without pretreatment, ultrasonically rinsed in acetone for 20 min, in 70% ethanol solution for 20 min and then in distilled water for 15 min. Moreover, we prepared a composite coating constituted of a titania (TiO2) matrix encapsulating hydroxyapatite (HA) by the sol–gel process described elsewhere [20]. In brief, a titania sol, prepared with titanium isopropoxide, acetyl acetone, nitric acid, n-propane alcohol and distilled water, was mixed with a solution of HA in anhydrous ethanol. The coatings were obtained by the dip technique. The surface features of samples were examined by scanning electron microscopy (SEM) with a Philips XL40 LaB6 at a 20 kV acceleration voltage, equipped with an energy dispersive spectrometer (EDS) EDAX DX4i. The roughness (Ra) of the samples surface was evaluated by means of a profilometer Alpha-step 200 (Tencor Instruments, USA). The measurements were carried out on different regions of the samples surface.
The samples were examined with a BioRad MRC-600 CLSM (BioRad Microscience Ltd, UK), equipped with an ion argon laser source and two photomultiplier tubes. Depth projection micrographs were visualized by 20 up to 40 horizontal image sections throughout the samples (serial optical section set), using a z-step of 0.5 mm and a 60x objective.
2.2. Cell culture Human MG63 osteoblast-like cells (ATCC, USA) were cultured in Dulbecco Modified Eagle’s Medium
2.5. 2-D image analysis The 2-D image analysis was performed with a software program (NIH Image 1.60, Bethesda, MD, USA) and the cell morphological features were automatically evidenced by thresholding. At least eight microscopic fields per sample were randomly acquired in the central and peripheral regions. The following geometrical features were determined for each object element: (1) cell area (expressed in mm2), (2) cell perimeter (mm) and (3) lengths of the major and minor axes of the ellipse with the same gravity center. These cell parameters were used to evaluate: (4) the circularity index (CI), calculated using the following formula: CI ¼ 4pA=P2 , where A is the cell area and P the cell perimeter. CI can range from 1 (a perfect circle) to 0
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
(a straight line); (5) the 2-D polarity factor, defined as the ratio between the lengths of major and minor axes of an ellipse having the same gravity center (a circular object would have a polarity factor of unity). 2.6. 3-D model Four steps were carried out using ordinary software in order to * *
* * * * *
optimize the contrast of cellular elements eliminate the background without losing the resolution identify the threshold level draw the object contours convert the files for 3-D reconstruction stack the images respecting their spatial position join the images with a capping grid and present them as a wireframe.
Step 1: The individual images of serial optical sections set were transformed in gray scale and elaborated by Adobe Photoshops 5.0 software. After the optimization of the image contrast, the background was eliminated. Then, the transitions of areas with major brightness were identified and the contours of the cells were drawn for each color channel, obtaining an effect similar to the lines in a contour map. In order to evaluate the chromatic levels, a threshold level in the range 0–255 was inserted after several preliminary tests. Step 2: Each processed image was converted into a drawing interchange format (dxf) vectorial image by the Corels OCR-Trace 8.0 software. Then it was converted in black and white and the parameters were opportunely set in order to obtain an exact transformation. Step 3: The images in dxf format were imported in Autodesk AutoCADs Release 14.0 software and the cell contour was recognized. The images taken at different focal planes were staked with a z-step correspondent to the micrograph depth. Step 4: Images were joined with a capping grid and presented as a wireframe. In some cases, the plane distance was amplified for a better details visualization.
399
3. Results 3.1. Substrate characterization SEM of titanium (Fig. 1a) revealed the presence of regular parallel grooves on the surface. SEM image of the TiO2/HA coating (Fig. 1b) showed a different surface topography; the coating appeared rough and characterized by numerous aggregations of particles and cracks. EDS analysis revealed that the surface was constituted by a titania matrix and by HA particles. The measurements of roughness (Ra) showed considerable differences being the average roughness of titanium and TiO2/HA coating of 0.6470.10 mm and 1.8070.33 mm, respectively. 3.2. 2-D image analysis After 24 h of incubation, only individual cells were examined because the cell–cell interactions can alter cell morphology. For a longer time, in fact, the number of
2.7. Statistical analysis Data, results of three replications of experiments, were expressed as mean values7standard deviation for each group of samples. After the assessment of significant differences by one-way variance analysis (ANOVA), differences among groups were established with t-test analysis by a two population comparison. Statistical significance was considered at a probability po0:05.
Fig. 1. SEM images of titanium (a) and TiO2/HA coating surface (b). Magnification 400 .
400
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
Table 1 Histomorphometric analysis determined on 2-D images after 24 h of incubationa Samples
Total cell area (mm2)
Circularity index
2-D polarity factor
Titanium TiO2/HA coating
1858.137314 850.907196***
0.2670.05 0.2770.13
2.5570.60 3.6071.41
a
Statistical analysis: difference from titanium; ***po0:001.
cells increases and the 2-D images obtained do not permit an exact evaluation of cellular features. In Table 1 are reported the cellular parameters measured by histomorphometric analysis and determined for at least eight microscopic fields per sample, using the sum of images corresponding to different focal planes obtained by CLSM. The osteoblasts grown on titanium appeared to cover more surface area of substrate with respect to TiO2/HA coating. An evaluation of the cell adhesion can be done calculating the CI that is strictly correlated to the cell shape. The results suggested that the cells are in contact with both kinds of substrate. The 2-D polarity factor, quantifying the asymmetry of cells spreading, indicated that the osteoblasts are polarized. Fig. 2a and b show the representative images of MG63 cultured onto titanium and TiO2/HA coating, respectively. The osteoblastic cells exhibited differences in cell morphology. The cells grown onto titanium were flatted and widespread and were oriented and elongated along the parallel grooves of the substrate. Moreover, they showed a regular and polygonal morphology with few cellular processes. The cells seeded onto TiO2/HA coating were not oriented and their small cellular area indicated that they were less spread than cells grown on titanium. Round and elongated cells, and cells with multiple cellular extensions were observed. Also the cytoskeletal organization of osteoblastic cells differed significantly with the substrate. The cells seeded onto titanium showed well-formed cortical filaments, whereas the osteoblasts grown on TiO2/HA coating showed no discernible actin filaments.
Fig. 2. Representative CLSM images of human MG63 osteoblast-like cells stained with phalloidin-FITC and cultured onto titanium (a) and TiO2/HA coating (b) after 1 day of incubation.
3.3. 3-D reconstruction A faithful 3-D representation of osteoblasts grown onto titanium after 2 days of incubation was realized by a direct extraction of the cellular features from single images of a stack (Fig. 3a) and compared with the corresponding 2-D image (Fig. 3b). The 3-D model allowed to recognize the relation among cells and the interaction between cells and substrate. Moreover, the 3D representation allowed to change the point of observation, rotating and dissecting the object, so obtaining further information that cannot be detected in 2-D images. In fact, the isometric image projection
showed very close connections between cell boundary. Moreover, a cellular process emitted from one cell was not in contact with the substrate. This behaviour is confirmed by the 2-D images (Fig. 3c) perpendicular to the section plane in a stack (slices), obtained using NIH Image software. In the middle of the slices, weak contacts between cells and substrate were visible. In Fig. 4 are reported the osteoblastic cells seeded onto titanium after 4 days of incubation. The cells proliferated disposing close to each other and showing adhesion on substrate. Weak contacts with substrate along the cell boundary were evident in Fig. 4c.
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
401
Fig. 3. (a) 3-D reconstruction of images obtained from a serial optical sections set by CLSM of MG63 osteoblasts grown onto titanium after 2 days of incubation. The arrow indicates a cellular process not in contact with the substrate. The rotation angles of the XY plane are reported. (b) 2-D image of osteoblastic cells stained with phalloidin-FITC obtained by CLSM. (c) 2-D images perpendicular to the section plane in a stack.
The osteoblasts seeded onto TiO2/HA coating showed a different behaviour after 2 days of incubation. Small cell clusters were observed where the cells aggregated (Fig. 5) and the dynamic of cellular interactions was pointed out by a cellular process which embraced the other cell association. After 4 days of incubation the cell number increased (Fig. 6) and a cell with multiple cellular extensions at one end laid on the other losing the contact with the substrate as confirmed by the slice (Fig. 6c). The 3-D reconstruction of osteoblastic cells reported in Fig. 5 is presented as a wireframe in Fig. 7; it showed
the real architecture of cellular elements and their spatial arrangement.
4. Discussion A complete description of the preparation and characterisation of the TiO2/HA coatings has been already reported elsewhere [20]. The analyses performed with SEM, X-ray diffraction, X-ray photoelectron spectroscopy and the adhesion tests of the coatings revealed that the samples were covered with an uniform,
402
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
Fig. 4. (a) 3-D reconstruction of images obtained from a serial optical sections set by CLSM of MG63 osteoblasts grown onto titanium after 4 days of incubation. The rotation angles of the XY plane are reported. (b) 2-D image of osteoblastic cells stained with phalloidin-FITC obtained by CLSM. (c) 2-D images perpendicular to the section plane in a stack.
clean and adherent coating, with a well defined thickness and phase composition. MG63 osteoblast-like cells adhere on titanium with different spreading characteristics with respect to TiO2/ HA coating. The histomorphometric analysis showed that osteoblasts grown on titanium undergo relatively more spreading. No significant differences were found regarding CI values and 2-D polarity factor, suggesting that the cells have a trend to attach on both the material surfaces and are polarized. The morphological observations indicated that the cells seeded onto titanium seem to conform to the
irregular material surface, maximizing the contact between the cell membrane and substrate and appear elongated and oriented along the titanium grooves, as described in literature [12,21]. In some cases, 3-D images revealed the presence of weak contacts among cells and substrate. The cells onto titanium showed actin stress fiber systems, a regular shape with few cellular processes and a compact morphology. The osteoblasts seeded onto TiO2/HA coating exhibited an elongated morphology with less organized actin filament systems, and also round cells were seen. Cells with multiple cellular extensions were present. Probably, the osteoblasts did not follow exactly the
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
403
Fig. 5. (a) 3-D reconstruction of images obtained from a serial optical sections set by CLSM of MG63 osteoblasts grown onto TiO2/HA coating after 2 days of incubation. The rotation angles of the XY plane are reported. (b) 2-D image of osteoblastic cells stained with phalloidin-FITC obtained by CLSM. (c) 2-D images perpendicular to the section plane in a stack.
substrate topography and tried to form stable anchorage sites extending multiple cellular processes. The significant difference of osteoblasts adhesion and spreading on the analyzed materials depends on different chemical composition and topography of substrates. The titanium had the surface covered by regular parallel grooves, whereas the TiO2/HA coating appeared rough with aggregations of particles. Several studies have reported that the roughness of the material is able to increase the efficiency of osteoblastic cell
adhesion with respect to smooth surfaces [1,5,22]. Also the spatial arrangement of osteoblastic cells cultured on biomaterials is considered directly related to the topography of the substrate. The distribution of osteoblastic cells on smooth materials such as a glass has been described as random, whereas the distribution on the ceramic substrate as clusters, due to the porosity of ceramic materials which reduces the surface available and thus facilitates the formation of cell cluster [23]. Cell aggregation is an early and critical event leading to cell
404
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
Fig. 6. (a) 3-D reconstruction of images obtained from a serial optical sections set by CLSM of MG63 osteoblasts grown onto TiO2/HA coating after 4 days of incubation. The rotation angles of the XY plane are reported. (b) 2-D image of osteoblastic cells stained with phalloidin-FITC obtained by CLSM. (c) 2-D images perpendicular to the section plane in a stack.
differentiation and mineralization process. In fact, the osteoblastic cells proliferate into multilayers foci forming 3-D structures known as ‘‘nodules’’ [24,25], where mineralization of extra-cellular matrix occurs. Considering the chemical composition of the substrates, in previous studies [26,27] it has been observed that the TiO2/HA coating resulted to be bioactive thanks to the presence of hydroxyl groups detected on the surface, that promote the calcium and phosphate precipitation and improve the interactions with osteoblastic cells. TiO2/HA coating cultured with human MG63 osteoblast-like cell line and primary rat osteoblasts induced the cell differentiation stimulating the expression of some peculiar biochemical markers such
Fig. 7. 3-D reconstruction of osteoblastic cells reported in Fig. 5 presented as a wireframe.
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
as alkaline phosphatase activity, collagen and osteocalcin production. The cell distribution as clusters showed by 3-D images reconstruction could be a first signal of the tendency of TiO2/HA coating to stimulate cell differentiation. The use of the 3-D model developed allowed to study the cellular aggregation process and to describe the spatial arrangement and the spreading depending on surface properties of the substrate. Further studies are in progress to analyze focal contacts distribution of osteoblastic cells on different substrates in order to study the adhesion points. Moreover, in the long-term studies it will be necessary to evaluate the ability of the biomaterials to promote cell mineralization.
5. Conclusion A complete description of the proliferation of aggregated cells grown on a substrate is possible only assembling together the information obtained by histomorphometric analysis, 2-D images and 3-D image reconstruction. The cells grown on titanium conform to the irregular substrate surfaces maximizing the contact between the cell membrane and the substrate and proliferate disposing close to each other. On the contrary, the osteoblasts seeded onto TiO2/HA coating develop clusters where the cells aggregated extending processes in order to establish intercellular connections. In previous studies, TiO2/HA coating resulted to be bioactive and improved the interactions with osteobastic cells. Therefore, it is possible to suppose that TiO2/HA coating induces cell differentiation. The 3-D representation technique developed, using some common and easily available software and realizing a model directly from the optical section set, has allowed to obtain a more realistic and faithful vision of morphological structures and to evaluate the spatial arrangement of cells grown on materials even in the cases of cell distribution as clusters.
Acknowledgements The authors would like to thank Eng. F. Cosentino, Ms. A. Romito, Mrs. L. Capodieci, Mr. A. Cappello of PASTIS-CNRSM and Dr. S. Tundo of Lecce University for technical assistance and SEM analysis. The research was carried out in the frame of the Innovation Project P4 funded by the Ministry for the University and Scientific and Technological Research (MURST).
405
References [1] Lincks J, Boyan BD, Blanchard CR, Lohmann CH, Liu Y, Cochran DL, Dean DD, Schwartz Z. Response of MG63 osteoblast-like cells to titanium and titanium alloy is dependent on surface roughness and composition. Biomaterials 1998; 19:2219–32. [2] Sinha RK, Morris F, Shah AK, Tuan RS. Surface composition of ortopaedic implant metals regulates cell attachment, spreading, and cytoskeletal organization of primary human osteoblasts in vitro. Clin Orthop 1994;305:258–72. [3] Vrouwenvelder WCA, Groot CG, de Groot K. Histological and biochemical evaluation of osteoblasts cultured on bioactive glass, hydroxylapatite, titanium alloy, and stainless steel. J Biomed Mater Res 1993;27:465–75. [4] Webb K, Hlady V, Tresco PA. Relative importance of surface wettability and charged functional groups on NIH 3T3 fibroblast attachment, spreading and cytoskeletal organization. J Biomed Mater Res 1998;41:422–30. [5] Boyan BD, Sylvia VL, Liu Y, Sagun R, Cochran DL, Lohmann CH, Dean DD, Schwartz Z. Surface roughness mediates its effects on osteoblasts via protein kinase A and phospholipase A2. Biomaterials 1999;20:2305–10. [6] Groessner-Schreiber B, Tuan RS. Enhanced extracellular matrix production and mineralization by osteoblasts cultured on titanium surfaces in vitro. J Cell Sci 1992;101:209–17. [7] Bagambisa FB, Joos U. Preliminary studies on the phenomenological behaviour of osteoblasts cultured on hydroxyapatite ceramics. Biomaterials 1990;11:50–6. [8] Malik MA, Puleo DA, Bizios R, Doremus RH. Osteoblasts on hydroxyapatite, alumina and bone surfaces in vitro: morphology during the first 2 h of attachment. Biomaterials 1992;13(2): 123–8. [9] Oliva A, Salerno A, Locardi B, Riccio V, Della Ragione F, Iardino P, Zappia V. Behaviour of human osteoblasts cultured on bioactive glass coatings. Biomaterials 1998;19:1019–25. [10] H.akkinen L, Yli-Urpo A, Heino J, Larjava H. Attachment and spreading of human gingival fibroblasts on potentially bioative glasses in vitro. J Biomed Mater Res 1988;22:1043–59. [11] Shah AK, Sinha RK, Hickok NJ, Tuan RS. High-resolution morphometric analysis of human osteoblastic cell adhesion on clinically relevant orthopaedic alloys. Bone 1999;24(5):499–506. [12] Walboomers XF, Croes HJ, Ginsel LA, Jansen JA. Growth behavior of fibroblasts on microgrooved polystyrene. Biomaterials 1998;19(20):1861–8. [13] van Kooten TG, Klein CL, Wagner M, Kirkpatrick CJ. Focal adhesions and assessment of cytotoxicity. J Biomed Mater Res 1999;46:33–43. [14] Chang CC, Lieberman SM, Moghe PV. Leukocyte spreading behaviour on vascular biomaterial surfaces: consequences of chemoattractant stimulation. Biomaterials 1999;20(3):273–81. [15] Friedl P, Z.ancher KS, Bro. cker EB. Cell migration strategies in 3-D extracellular matrix: differences in morphology, cell matrix interactions, and integrin function. Microsc Res Tech 1998;43(5):369–78. [16] den Braber ET, Jansen HV, de Boer MJ, Croes HJ, Elwenspoek M, Ginsel LA, Jansen JA. Scanning electron microscopic, transmission electron microscopic, and confocal laser scanning microscopic observation of fibroblasts cultured on microgrooves surfaces of bulk titanium substrata. J Biomed Mater Res 1998;40(3):425–33. [17] Ishaug-Riley SL, Crane-Kruger GM, Yaszemski MJ, Mikos AG. Three-dimensional culture of rat calvarian osteoblasts in porous biodegradable polymers. Biomaterials 1998;19:1405–12. [18] Tsuchida T, Yoshimura K, Shirayama Y, Kawamoto K. Colcemid-induced apoptosis of cultured human glioma: electron
406
[19]
[20]
[21] [22]
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406 microscopic and confocal laser microscopic observation of cells sorted in different phases of cell cycle. Cytometry 1998;31(4): 295–9. Zucker RM, Hunter ES, Rogers JM. Apoptosis and morphology in mouse embryos by confocal laser scanning microscopy. Methods 1999;18(4):473–80. Milella E, Cosentino F, Licciulli A, Massaro C. Preparation and characterization of titania/hydroxyapatite composite coatings obtained by sol–gel process. Biomaterials 2001;22(11):1425–31. Brunette DM. The effect of implant surface topography on the behavior of cells. Int J Oral Maxillofac Implants 1988;3:231–46. Martin JY, Schwartz Z, Hummert TW, Schraub DM, Simpson J, Lankford J, Dean DD, Cochran DL, Boyan BD. Effect of titanium surface roughness on proliferation, differentiation, and protein synthesis of human osteoblast-like cells (MG63). J Biomed Mater Res 1995;29:389–401.
[23] Dubois JC, Souchier C, Couble ML, Exbrayat P, Lissac M. An image analysis method for the study of cell adhesion to biomaterials. Biomaterials 1999;20:1841–9. [24] Beresford JN, Graves SE, Smoothy CA. Formation of mineralized nodules by bone derived cells in vitro: a model of bone formation? Am J Med Genet 1993;45:163–78. [25] Ahmad M, McCarthy M, Gronowicz G. An in vitro model for mineralization of human osteoblast-like cells on implant materials. Biomaterials 1999;20:211–20. [26] Ramires PA, Romito A, Cosentino F, Milella E. The influence of titania/hydroxyapatite composite coatings on in vitro osteoblasts behaviour. Biomaterials 2001;22(12):1467–74. [27] Ramires PA, Torricelli P, Cosentino F, Giavaresi G, Giardino R, Milella E. In vitro response of primary rat osteoblasts to titania/ hydroxyapatite coatings compared with transformed human osteoblast-like cells. J Mater Sci Mater Med, accepted.
Three-dimensional reconstruction of confocal laser microscopy images to study the behaviour of osteoblastic cells grown on biomaterials P.A. Ramires*, A. Giuffrida, E. Milella PASTIS-CNRSM, Biomaterials Unit, S.S. 7 Appia, Km 714, 72100 Brindisi, Italy Received 9 August 2000; accepted 7 March 2001
Abstract The adhesion, spreading and cytoskeletal organization of osteoblastic cells seeded onto titanium and titania/hydroxyapatite composite coating (TiO2/HA) were studied using images acquired by confocal laser scanning microscopy. The fluorescence staining technique was employed to visualize actin cytoskeletal organization of cells. 2-D images were exhaustive when the cells were seeded at low density (in the first 24 h of incubation), but they were less clear when the cells proliferated and appeared stacked. Since the shareware software were not satisfactory, a new 3-D image reconstruction was developed using ordinary software and a model was obtained directly from the optical section set, in order to achieve a more realistic and faithful vision of morphological structures and to evaluate the behaviour of bone cells grown on materials. The results showed that the cells grown on titanium conform to the irregular substrate surfaces maximizing the contact between the cell membrane and the substrate and proliferate disposing close to each other. On the contrary, the osteoblasts seeded onto TiO2/HA coating develop clusters where the cells aggregated extending processes in order to establish intercellular connections. Cell aggregation is an early and critical event leading to cell differentiation and mineralization process and could be a first signal of the tendency of TiO2/HA coating to stimulate cell differentiation. r 2001 Elsevier Science Ltd. All rights reserved. Keywords: 3-D image reconstruction; Confocal laser scanning microscopy; Osteoblastic cells; Titanium; titania/hydroxyapatite coating
1. Introduction The knowledge of interaction mechanisms between cells and biomaterials is particularly relevant for the performance of an implanted material and for the development of more effective devices. Cells attachment to substrates is the first step in the process of cell-surface interactions and affects subsequent cellular and tissue responses. In orthopedic and dental field the early interaction between osteoblastic cells and their substrates affects the later mineralization process and determine the longterm stability of the implanted prostheses. In fact, critical to the success of implants is the development of a stable direct connection between bone and surface implant, which must be structural and functional *Corresponding author. Fax: +390-831-507-379. E-mail address: [email protected] (P.A. Ramires).
without any intervening soft or fibrous tissue (osseointegration). The tissue response to an implant involves physical factors, depending on the implant design, surface topography, hydrophobicity, surface charge density, surface free energy and chemical factors associated with the composition and structure of the material [1–4]. These substrate characteristics may directly influence cell adhesion, spreading and signaling, events that regulate a wide variety of biological functions, including cell growth, cell migration, cell differentiation, extracellular matrix synthesis and tissue morphogenesis [5–7]. Previous reports have described the differential attachment kinetics, spreading characteristics, migration, growth and signaling cascades of various cell types, including osteoblasts, on tissue culture polystyrene, glass and orthopedic alloys [3,7–10]. Moreover, researches are focused to study the highly specialized adhesion structures, namely focal adhesion [10–13].
0142-9612/02/$ - see front matter r 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 2 - 9 6 1 2 ( 0 1 ) 0 0 1 1 8 - 1
398
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
The confocal laser scanning microscopy (CLSM) and immunofluorescence staining technique have been widely used to study cell morphology, to analyze the intracellular location of various cell components, to study the migration and the spreading of cells [14], the cytoskeleton and the focal adhesions [12,15–17] and to study apoptosis [18,19]. We have studied the behaviour of human MG63 osteoblast-like cells seeded onto titanium and titania/ hydroxyapatite composite coating (TiO2/HA) after 1 day of incubation by means of 2-D analysis carried out on CSLM images. The fluorescence staining technique was employed to visualize actin cytoskeletal organization of cells. After long time of incubation the number of cells onto the materials increases and the 2-D images do not permit to analyze exactly the cell morphology, the cell–cell interactions and the contacts with the substrate. A 3-D image reconstruction made by means of a shareware software was not satisfactory. Therefore, a new 3-D representation technique was developed using ordinary software and a model was obtained directly from the optical section set, in order to achieve a more realistic and faithful vision of morphological structures and to study the behaviour of bone cells grown on materials.
(DMEM, Biowhittaker, Belgium), containing penicillin/ streptomycin (100/100 IU), amphotericin B (2.5 mg/ml) and gentamycin (100 mg/ml), supplemented with 10% foetal calf serum (Euroclone Ltd, UK) and kept at 371C in an atmosphere of 5% CO2 and 99% humidity. Media were changed every 3 days. The cells were seeded onto the substrates at a cell density of 0.5 104 cells/cm2. On every sample, placed in a 6-multiwell plate (Corning, USA), 100 ml of a cell suspension was applied carefully and the cells were allowed to attach for 2 h to the underlying substrate, then 3.5 ml of culture medium was also added. 2.3. Staining
2. Materials and methods
After 1, 2 and 4 days of incubation the samples were fixed with 4% formaldehyde in phosphate buffer saline, permeabilized with 20 mm digitonin (Molecular Probes, Netherlands) for 5 min at room temperature and stained with 5 mg/ml phalloidin-FITC (Sigma, USA) for 1 h in order to visualize the F-actin filaments. In some cases, after a treatment with 1 mg/ml ribonuclease A (Sigma), samples were double labelled with phalloidin-FITC and 1 mg/ml propidium iodide (Molecular Probes) for cytoskeleton and nuclear staining respectively. Then, the samples were mounted under glass coverslips with SlowFades solution (Molecular Probes) to prevent quenching.
2.1. Substrates
2.4. Confocal laser scanning microscopy
Commercially pure titanium (Goodfellow, Germany), cut into pieces 20 10 mm2 size, was used as substrate without pretreatment, ultrasonically rinsed in acetone for 20 min, in 70% ethanol solution for 20 min and then in distilled water for 15 min. Moreover, we prepared a composite coating constituted of a titania (TiO2) matrix encapsulating hydroxyapatite (HA) by the sol–gel process described elsewhere [20]. In brief, a titania sol, prepared with titanium isopropoxide, acetyl acetone, nitric acid, n-propane alcohol and distilled water, was mixed with a solution of HA in anhydrous ethanol. The coatings were obtained by the dip technique. The surface features of samples were examined by scanning electron microscopy (SEM) with a Philips XL40 LaB6 at a 20 kV acceleration voltage, equipped with an energy dispersive spectrometer (EDS) EDAX DX4i. The roughness (Ra) of the samples surface was evaluated by means of a profilometer Alpha-step 200 (Tencor Instruments, USA). The measurements were carried out on different regions of the samples surface.
The samples were examined with a BioRad MRC-600 CLSM (BioRad Microscience Ltd, UK), equipped with an ion argon laser source and two photomultiplier tubes. Depth projection micrographs were visualized by 20 up to 40 horizontal image sections throughout the samples (serial optical section set), using a z-step of 0.5 mm and a 60x objective.
2.2. Cell culture Human MG63 osteoblast-like cells (ATCC, USA) were cultured in Dulbecco Modified Eagle’s Medium
2.5. 2-D image analysis The 2-D image analysis was performed with a software program (NIH Image 1.60, Bethesda, MD, USA) and the cell morphological features were automatically evidenced by thresholding. At least eight microscopic fields per sample were randomly acquired in the central and peripheral regions. The following geometrical features were determined for each object element: (1) cell area (expressed in mm2), (2) cell perimeter (mm) and (3) lengths of the major and minor axes of the ellipse with the same gravity center. These cell parameters were used to evaluate: (4) the circularity index (CI), calculated using the following formula: CI ¼ 4pA=P2 , where A is the cell area and P the cell perimeter. CI can range from 1 (a perfect circle) to 0
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
(a straight line); (5) the 2-D polarity factor, defined as the ratio between the lengths of major and minor axes of an ellipse having the same gravity center (a circular object would have a polarity factor of unity). 2.6. 3-D model Four steps were carried out using ordinary software in order to * *
* * * * *
optimize the contrast of cellular elements eliminate the background without losing the resolution identify the threshold level draw the object contours convert the files for 3-D reconstruction stack the images respecting their spatial position join the images with a capping grid and present them as a wireframe.
Step 1: The individual images of serial optical sections set were transformed in gray scale and elaborated by Adobe Photoshops 5.0 software. After the optimization of the image contrast, the background was eliminated. Then, the transitions of areas with major brightness were identified and the contours of the cells were drawn for each color channel, obtaining an effect similar to the lines in a contour map. In order to evaluate the chromatic levels, a threshold level in the range 0–255 was inserted after several preliminary tests. Step 2: Each processed image was converted into a drawing interchange format (dxf) vectorial image by the Corels OCR-Trace 8.0 software. Then it was converted in black and white and the parameters were opportunely set in order to obtain an exact transformation. Step 3: The images in dxf format were imported in Autodesk AutoCADs Release 14.0 software and the cell contour was recognized. The images taken at different focal planes were staked with a z-step correspondent to the micrograph depth. Step 4: Images were joined with a capping grid and presented as a wireframe. In some cases, the plane distance was amplified for a better details visualization.
399
3. Results 3.1. Substrate characterization SEM of titanium (Fig. 1a) revealed the presence of regular parallel grooves on the surface. SEM image of the TiO2/HA coating (Fig. 1b) showed a different surface topography; the coating appeared rough and characterized by numerous aggregations of particles and cracks. EDS analysis revealed that the surface was constituted by a titania matrix and by HA particles. The measurements of roughness (Ra) showed considerable differences being the average roughness of titanium and TiO2/HA coating of 0.6470.10 mm and 1.8070.33 mm, respectively. 3.2. 2-D image analysis After 24 h of incubation, only individual cells were examined because the cell–cell interactions can alter cell morphology. For a longer time, in fact, the number of
2.7. Statistical analysis Data, results of three replications of experiments, were expressed as mean values7standard deviation for each group of samples. After the assessment of significant differences by one-way variance analysis (ANOVA), differences among groups were established with t-test analysis by a two population comparison. Statistical significance was considered at a probability po0:05.
Fig. 1. SEM images of titanium (a) and TiO2/HA coating surface (b). Magnification 400 .
400
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
Table 1 Histomorphometric analysis determined on 2-D images after 24 h of incubationa Samples
Total cell area (mm2)
Circularity index
2-D polarity factor
Titanium TiO2/HA coating
1858.137314 850.907196***
0.2670.05 0.2770.13
2.5570.60 3.6071.41
a
Statistical analysis: difference from titanium; ***po0:001.
cells increases and the 2-D images obtained do not permit an exact evaluation of cellular features. In Table 1 are reported the cellular parameters measured by histomorphometric analysis and determined for at least eight microscopic fields per sample, using the sum of images corresponding to different focal planes obtained by CLSM. The osteoblasts grown on titanium appeared to cover more surface area of substrate with respect to TiO2/HA coating. An evaluation of the cell adhesion can be done calculating the CI that is strictly correlated to the cell shape. The results suggested that the cells are in contact with both kinds of substrate. The 2-D polarity factor, quantifying the asymmetry of cells spreading, indicated that the osteoblasts are polarized. Fig. 2a and b show the representative images of MG63 cultured onto titanium and TiO2/HA coating, respectively. The osteoblastic cells exhibited differences in cell morphology. The cells grown onto titanium were flatted and widespread and were oriented and elongated along the parallel grooves of the substrate. Moreover, they showed a regular and polygonal morphology with few cellular processes. The cells seeded onto TiO2/HA coating were not oriented and their small cellular area indicated that they were less spread than cells grown on titanium. Round and elongated cells, and cells with multiple cellular extensions were observed. Also the cytoskeletal organization of osteoblastic cells differed significantly with the substrate. The cells seeded onto titanium showed well-formed cortical filaments, whereas the osteoblasts grown on TiO2/HA coating showed no discernible actin filaments.
Fig. 2. Representative CLSM images of human MG63 osteoblast-like cells stained with phalloidin-FITC and cultured onto titanium (a) and TiO2/HA coating (b) after 1 day of incubation.
3.3. 3-D reconstruction A faithful 3-D representation of osteoblasts grown onto titanium after 2 days of incubation was realized by a direct extraction of the cellular features from single images of a stack (Fig. 3a) and compared with the corresponding 2-D image (Fig. 3b). The 3-D model allowed to recognize the relation among cells and the interaction between cells and substrate. Moreover, the 3D representation allowed to change the point of observation, rotating and dissecting the object, so obtaining further information that cannot be detected in 2-D images. In fact, the isometric image projection
showed very close connections between cell boundary. Moreover, a cellular process emitted from one cell was not in contact with the substrate. This behaviour is confirmed by the 2-D images (Fig. 3c) perpendicular to the section plane in a stack (slices), obtained using NIH Image software. In the middle of the slices, weak contacts between cells and substrate were visible. In Fig. 4 are reported the osteoblastic cells seeded onto titanium after 4 days of incubation. The cells proliferated disposing close to each other and showing adhesion on substrate. Weak contacts with substrate along the cell boundary were evident in Fig. 4c.
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
401
Fig. 3. (a) 3-D reconstruction of images obtained from a serial optical sections set by CLSM of MG63 osteoblasts grown onto titanium after 2 days of incubation. The arrow indicates a cellular process not in contact with the substrate. The rotation angles of the XY plane are reported. (b) 2-D image of osteoblastic cells stained with phalloidin-FITC obtained by CLSM. (c) 2-D images perpendicular to the section plane in a stack.
The osteoblasts seeded onto TiO2/HA coating showed a different behaviour after 2 days of incubation. Small cell clusters were observed where the cells aggregated (Fig. 5) and the dynamic of cellular interactions was pointed out by a cellular process which embraced the other cell association. After 4 days of incubation the cell number increased (Fig. 6) and a cell with multiple cellular extensions at one end laid on the other losing the contact with the substrate as confirmed by the slice (Fig. 6c). The 3-D reconstruction of osteoblastic cells reported in Fig. 5 is presented as a wireframe in Fig. 7; it showed
the real architecture of cellular elements and their spatial arrangement.
4. Discussion A complete description of the preparation and characterisation of the TiO2/HA coatings has been already reported elsewhere [20]. The analyses performed with SEM, X-ray diffraction, X-ray photoelectron spectroscopy and the adhesion tests of the coatings revealed that the samples were covered with an uniform,
402
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
Fig. 4. (a) 3-D reconstruction of images obtained from a serial optical sections set by CLSM of MG63 osteoblasts grown onto titanium after 4 days of incubation. The rotation angles of the XY plane are reported. (b) 2-D image of osteoblastic cells stained with phalloidin-FITC obtained by CLSM. (c) 2-D images perpendicular to the section plane in a stack.
clean and adherent coating, with a well defined thickness and phase composition. MG63 osteoblast-like cells adhere on titanium with different spreading characteristics with respect to TiO2/ HA coating. The histomorphometric analysis showed that osteoblasts grown on titanium undergo relatively more spreading. No significant differences were found regarding CI values and 2-D polarity factor, suggesting that the cells have a trend to attach on both the material surfaces and are polarized. The morphological observations indicated that the cells seeded onto titanium seem to conform to the
irregular material surface, maximizing the contact between the cell membrane and substrate and appear elongated and oriented along the titanium grooves, as described in literature [12,21]. In some cases, 3-D images revealed the presence of weak contacts among cells and substrate. The cells onto titanium showed actin stress fiber systems, a regular shape with few cellular processes and a compact morphology. The osteoblasts seeded onto TiO2/HA coating exhibited an elongated morphology with less organized actin filament systems, and also round cells were seen. Cells with multiple cellular extensions were present. Probably, the osteoblasts did not follow exactly the
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
403
Fig. 5. (a) 3-D reconstruction of images obtained from a serial optical sections set by CLSM of MG63 osteoblasts grown onto TiO2/HA coating after 2 days of incubation. The rotation angles of the XY plane are reported. (b) 2-D image of osteoblastic cells stained with phalloidin-FITC obtained by CLSM. (c) 2-D images perpendicular to the section plane in a stack.
substrate topography and tried to form stable anchorage sites extending multiple cellular processes. The significant difference of osteoblasts adhesion and spreading on the analyzed materials depends on different chemical composition and topography of substrates. The titanium had the surface covered by regular parallel grooves, whereas the TiO2/HA coating appeared rough with aggregations of particles. Several studies have reported that the roughness of the material is able to increase the efficiency of osteoblastic cell
adhesion with respect to smooth surfaces [1,5,22]. Also the spatial arrangement of osteoblastic cells cultured on biomaterials is considered directly related to the topography of the substrate. The distribution of osteoblastic cells on smooth materials such as a glass has been described as random, whereas the distribution on the ceramic substrate as clusters, due to the porosity of ceramic materials which reduces the surface available and thus facilitates the formation of cell cluster [23]. Cell aggregation is an early and critical event leading to cell
404
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
Fig. 6. (a) 3-D reconstruction of images obtained from a serial optical sections set by CLSM of MG63 osteoblasts grown onto TiO2/HA coating after 4 days of incubation. The rotation angles of the XY plane are reported. (b) 2-D image of osteoblastic cells stained with phalloidin-FITC obtained by CLSM. (c) 2-D images perpendicular to the section plane in a stack.
differentiation and mineralization process. In fact, the osteoblastic cells proliferate into multilayers foci forming 3-D structures known as ‘‘nodules’’ [24,25], where mineralization of extra-cellular matrix occurs. Considering the chemical composition of the substrates, in previous studies [26,27] it has been observed that the TiO2/HA coating resulted to be bioactive thanks to the presence of hydroxyl groups detected on the surface, that promote the calcium and phosphate precipitation and improve the interactions with osteoblastic cells. TiO2/HA coating cultured with human MG63 osteoblast-like cell line and primary rat osteoblasts induced the cell differentiation stimulating the expression of some peculiar biochemical markers such
Fig. 7. 3-D reconstruction of osteoblastic cells reported in Fig. 5 presented as a wireframe.
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406
as alkaline phosphatase activity, collagen and osteocalcin production. The cell distribution as clusters showed by 3-D images reconstruction could be a first signal of the tendency of TiO2/HA coating to stimulate cell differentiation. The use of the 3-D model developed allowed to study the cellular aggregation process and to describe the spatial arrangement and the spreading depending on surface properties of the substrate. Further studies are in progress to analyze focal contacts distribution of osteoblastic cells on different substrates in order to study the adhesion points. Moreover, in the long-term studies it will be necessary to evaluate the ability of the biomaterials to promote cell mineralization.
5. Conclusion A complete description of the proliferation of aggregated cells grown on a substrate is possible only assembling together the information obtained by histomorphometric analysis, 2-D images and 3-D image reconstruction. The cells grown on titanium conform to the irregular substrate surfaces maximizing the contact between the cell membrane and the substrate and proliferate disposing close to each other. On the contrary, the osteoblasts seeded onto TiO2/HA coating develop clusters where the cells aggregated extending processes in order to establish intercellular connections. In previous studies, TiO2/HA coating resulted to be bioactive and improved the interactions with osteobastic cells. Therefore, it is possible to suppose that TiO2/HA coating induces cell differentiation. The 3-D representation technique developed, using some common and easily available software and realizing a model directly from the optical section set, has allowed to obtain a more realistic and faithful vision of morphological structures and to evaluate the spatial arrangement of cells grown on materials even in the cases of cell distribution as clusters.
Acknowledgements The authors would like to thank Eng. F. Cosentino, Ms. A. Romito, Mrs. L. Capodieci, Mr. A. Cappello of PASTIS-CNRSM and Dr. S. Tundo of Lecce University for technical assistance and SEM analysis. The research was carried out in the frame of the Innovation Project P4 funded by the Ministry for the University and Scientific and Technological Research (MURST).
405
References [1] Lincks J, Boyan BD, Blanchard CR, Lohmann CH, Liu Y, Cochran DL, Dean DD, Schwartz Z. Response of MG63 osteoblast-like cells to titanium and titanium alloy is dependent on surface roughness and composition. Biomaterials 1998; 19:2219–32. [2] Sinha RK, Morris F, Shah AK, Tuan RS. Surface composition of ortopaedic implant metals regulates cell attachment, spreading, and cytoskeletal organization of primary human osteoblasts in vitro. Clin Orthop 1994;305:258–72. [3] Vrouwenvelder WCA, Groot CG, de Groot K. Histological and biochemical evaluation of osteoblasts cultured on bioactive glass, hydroxylapatite, titanium alloy, and stainless steel. J Biomed Mater Res 1993;27:465–75. [4] Webb K, Hlady V, Tresco PA. Relative importance of surface wettability and charged functional groups on NIH 3T3 fibroblast attachment, spreading and cytoskeletal organization. J Biomed Mater Res 1998;41:422–30. [5] Boyan BD, Sylvia VL, Liu Y, Sagun R, Cochran DL, Lohmann CH, Dean DD, Schwartz Z. Surface roughness mediates its effects on osteoblasts via protein kinase A and phospholipase A2. Biomaterials 1999;20:2305–10. [6] Groessner-Schreiber B, Tuan RS. Enhanced extracellular matrix production and mineralization by osteoblasts cultured on titanium surfaces in vitro. J Cell Sci 1992;101:209–17. [7] Bagambisa FB, Joos U. Preliminary studies on the phenomenological behaviour of osteoblasts cultured on hydroxyapatite ceramics. Biomaterials 1990;11:50–6. [8] Malik MA, Puleo DA, Bizios R, Doremus RH. Osteoblasts on hydroxyapatite, alumina and bone surfaces in vitro: morphology during the first 2 h of attachment. Biomaterials 1992;13(2): 123–8. [9] Oliva A, Salerno A, Locardi B, Riccio V, Della Ragione F, Iardino P, Zappia V. Behaviour of human osteoblasts cultured on bioactive glass coatings. Biomaterials 1998;19:1019–25. [10] H.akkinen L, Yli-Urpo A, Heino J, Larjava H. Attachment and spreading of human gingival fibroblasts on potentially bioative glasses in vitro. J Biomed Mater Res 1988;22:1043–59. [11] Shah AK, Sinha RK, Hickok NJ, Tuan RS. High-resolution morphometric analysis of human osteoblastic cell adhesion on clinically relevant orthopaedic alloys. Bone 1999;24(5):499–506. [12] Walboomers XF, Croes HJ, Ginsel LA, Jansen JA. Growth behavior of fibroblasts on microgrooved polystyrene. Biomaterials 1998;19(20):1861–8. [13] van Kooten TG, Klein CL, Wagner M, Kirkpatrick CJ. Focal adhesions and assessment of cytotoxicity. J Biomed Mater Res 1999;46:33–43. [14] Chang CC, Lieberman SM, Moghe PV. Leukocyte spreading behaviour on vascular biomaterial surfaces: consequences of chemoattractant stimulation. Biomaterials 1999;20(3):273–81. [15] Friedl P, Z.ancher KS, Bro. cker EB. Cell migration strategies in 3-D extracellular matrix: differences in morphology, cell matrix interactions, and integrin function. Microsc Res Tech 1998;43(5):369–78. [16] den Braber ET, Jansen HV, de Boer MJ, Croes HJ, Elwenspoek M, Ginsel LA, Jansen JA. Scanning electron microscopic, transmission electron microscopic, and confocal laser scanning microscopic observation of fibroblasts cultured on microgrooves surfaces of bulk titanium substrata. J Biomed Mater Res 1998;40(3):425–33. [17] Ishaug-Riley SL, Crane-Kruger GM, Yaszemski MJ, Mikos AG. Three-dimensional culture of rat calvarian osteoblasts in porous biodegradable polymers. Biomaterials 1998;19:1405–12. [18] Tsuchida T, Yoshimura K, Shirayama Y, Kawamoto K. Colcemid-induced apoptosis of cultured human glioma: electron
406
[19]
[20]
[21] [22]
P.A. Ramires et al. / Biomaterials 23 (2002) 397–406 microscopic and confocal laser microscopic observation of cells sorted in different phases of cell cycle. Cytometry 1998;31(4): 295–9. Zucker RM, Hunter ES, Rogers JM. Apoptosis and morphology in mouse embryos by confocal laser scanning microscopy. Methods 1999;18(4):473–80. Milella E, Cosentino F, Licciulli A, Massaro C. Preparation and characterization of titania/hydroxyapatite composite coatings obtained by sol–gel process. Biomaterials 2001;22(11):1425–31. Brunette DM. The effect of implant surface topography on the behavior of cells. Int J Oral Maxillofac Implants 1988;3:231–46. Martin JY, Schwartz Z, Hummert TW, Schraub DM, Simpson J, Lankford J, Dean DD, Cochran DL, Boyan BD. Effect of titanium surface roughness on proliferation, differentiation, and protein synthesis of human osteoblast-like cells (MG63). J Biomed Mater Res 1995;29:389–401.
[23] Dubois JC, Souchier C, Couble ML, Exbrayat P, Lissac M. An image analysis method for the study of cell adhesion to biomaterials. Biomaterials 1999;20:1841–9. [24] Beresford JN, Graves SE, Smoothy CA. Formation of mineralized nodules by bone derived cells in vitro: a model of bone formation? Am J Med Genet 1993;45:163–78. [25] Ahmad M, McCarthy M, Gronowicz G. An in vitro model for mineralization of human osteoblast-like cells on implant materials. Biomaterials 1999;20:211–20. [26] Ramires PA, Romito A, Cosentino F, Milella E. The influence of titania/hydroxyapatite composite coatings on in vitro osteoblasts behaviour. Biomaterials 2001;22(12):1467–74. [27] Ramires PA, Torricelli P, Cosentino F, Giavaresi G, Giardino R, Milella E. In vitro response of primary rat osteoblasts to titania/ hydroxyapatite coatings compared with transformed human osteoblast-like cells. J Mater Sci Mater Med, accepted.