This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Fullerene C60 films of continuous and micropatterned morphology as substrates foradhesion and growth of bone cells

Lubica Grausova a, Jiri Vacik b, Vladimir Vorlicek c, Vaclav Svorcik d, Petr Slepicka d, Petra Bilkova e,Marta Vandrovcova a, Vera Lisa a, Lucie Bacakova a,⁎a Institute of Physiology, Academy of Sciences of the Czech Republic, Videnska 1083, CZ-142 20 Prague 4-Krc, Czech Republicb Nuclear Physics Institute, Academy of Sciences of the Czech Republic, CZ 250 68 Rez near Prague, Czech Republicc Institute of Physics, Academy of Sciences of the Czech Republic, Na Slovance 2, CZ-182 21 Prague 8, Czech Republicd Department of Solid State Engineering, Institute of Chemical Technology, Technicka 5, CZ-166 28 Prague 6, Czech Republice Institute of Plasma Physics, Academy of Sciences of the Czech Republic, Za Slovankou 3, 182 00 Prague 8, Czech Republic

a b s t r a c ta r t i c l e i n f o

Available online 30 October 2008

Keywords:NanoparticlesBiocompatibilitySurface structureBiomedical application

Fullerenes C60 were deposited on microscopic glass coverslips in the form of continuous layers (thickness505±43 nm or 1090±8 nm) or micropatterned layers using a metallic mask with rectangular openings(128×98 µm, spacing 50 µm). Below the openings, the fullerenes formed bulges 484±5 nm or 1043±57 nm inthickness. Very thin fullerene films were also found below the metallic bars of the grid. The adhesion andproliferation activities of human osteoblast-like MG 63 cells in cultures on the fullerene layers were similar asthat on the control polystyrene dishes and glass coverslips. On the thick patterned layers, the cells grewpreferentially in the grooves among the fullerene bulges. Although these grooves occupied onlyapproximately 41% of the surface, they contained from 80% to 98% of the cells. Immunocytochemistryshowed that the cells on all tested surfaces formed β1 integrin- and talin-containing focal adhesion plaques, aβ-actin cytoskeleton, and contained osteocalcin, a marker of osteogenic cell differentiation. These resultssuggest that fullerene C60 layers can act as good substrates for cell colonization, and could also serve for theconstruction of micropatterned surfaces for guided cell adhesion and growth.

© 2008 Elsevier B.V. All rights reserved.

1. Introduction

Fullerenes, i.e. spheroidal molecules made exclusively of carbonatoms (e.g., C60, C70), display a diverse range of biological activity (for areview, see [1–3]). Their unique hollow cage-like shape and structuralanalogy with clathrin-coated vesicles in cells support the idea of thepotential use of fullerenes as drug or gene delivery agents [4–6].Fullerenes are able to accept and release electrons. When irradiatedwith ultraviolet or visible light, fullerenes can convert molecularoxygen into highly reactive singlet oxygen [7]. Thus, they have thepotential to inflict photodynamic damage on biological systems,including damage to cellular membranes, inhibition of variousenzymes or DNA cleavage [8–11]. This harmful effect can be exploitedfor photodynamic therapy against tumors, viruses and bacteriaresistant to multiple drugs [12,13]. On the other hand, C60 is

considered to be the world's most efficient radical scavenger. This isdue to the relatively large number of conjugated double bonds in thefullerene molecule, which can be attacked by radical species. Thus,fullerenes would be suitable for applications in quenching oxygenradicals, and thus preventing the damage of various tissues andorgans, including the cardiovascular and central nervous systems[1,7,14–19]. In addition, fullerenes emit photoluminescence whichcould be utilized in advanced imaging technologies [3].

In their pristine unmodified state, fullerenes are highly hydro-phobic and water-insoluble. On the other hand, they are relativelyhighly reactive, which enables them to be structurally modified.Fullerenes can form complexes with other atoms and molecules, e.g.metals, nucleic acids, proteins, synthetic polymers as well as othercarbon nanoparticles, e.g. nanotubes. In addition, fullerenes can befunctionalized with various chemical groups, e.g. hydroxyl, aldehy-dic, carbonyl, carboxyl, ester or amine group, as well as amino acidsand peptides. This usually renders them soluble in water andintensifies their interaction with biological systems [4,5,8–11,17–21].

Despite all these exciting findings, relatively little is still knownabout the influence of fullerenes, particularly when arranged intolayers and used for biomaterial coating, on cell-substrate adhesion,

Diamond & Related Materials 18 (2009) 578–586

⁎ Corresponding author. Tel.: +420 2 9644 3743, +420 2 9644 3742; fax: +420 2 96442488, +420 2 4106 2488.

E-mail addresses: grausova@biomed.cas.cz (L. Grausova), vacik@ujf.cas.cz (J. Vacik),vorlicek@fzu.cz (V. Vorlicek), vaclav.svorcik@vscht.cz (V. Svorcik),petr.slepicka@vscht.cz (P. Slepicka), bilkova@ipp.cas.cz (P. Bilkova),vandrovcova@biomed.cas.cz (M. Vandrovcova), lucy@biomed.cas.cz (L. Bacakova).

0925-9635/$ – see front matter © 2008 Elsevier B.V. All rights reserved.doi:10.1016/j.diamond.2008.10.024

Contents lists available at ScienceDirect

Diamond & Related Materials

j ourna l homepage: www.e lsev ie r.com/ locate /d iamond

Author's personal copy

subsequent growth, differentiation and viability of cells, especiallybone-forming cells. In our earlier study, a layer of fullerene C60

molecules, i.e. fullerite, deposited on carbon fibre-reinforced carboncomposites (CFRC), enlarged the cell spreading area in human-osteoblast-like MG 63 cells in cultures on these materials [2], whichwas most probably due to the nanostructure of the material surface.Moreover, it is believed that nanostructured surfaces can promotepreferential adhesion and growth of osteoblasts over other “compe-titive” cell types, including fibroblasts, and thus they can preventfibrous encapsulation and loosening of bone implants. It isconsidered that the underlying mechanism is higher adsorption tothe nanostructured surfaces of vitronectin, an extracellular matrix(ECM) protein preferred by osteoblasts [22,23]. Therefore, it can beexpected that carbon nanoparticles, including fullerenes, may serveas novel building blocks for creating artificial bioinspired nanos-tructured surfaces for bone tissue engineering [24–26]. Proliferationand differentiation of chondrocytes, another cell type important inbone reconstruction, were also promoted by fullerenes. C60

enhanced the production of specific large proteoglycan ECMmolecules, typical for cartilage, in rat embryonic limb bud cells inculture [27]. Interestingly, in a line of human epidermoid carcinomacells, fullerenes protected cells from anoikis, i.e., apoptosis due toadhesion deprivation, by a mechanism supporting the formation offocal adhesion plaques, assembly of the actin cytoskeleton and cellspreading [28]. Enhanced attachment and spreading of platelets wasalso found on a polyurethane surface grafted with fullerene C60

molecules [4].In the present study, the adhesion, growth, viability and maturation

of human osteoblast-like MG 63 cells were investigated in cultures onfullerene C60 films of various thickness and morphology, deposited onmicroscopic glass coverslips. In addition to continuous films, alsomicropatternedfilmswere prepared by deposition of fullerenes througha metallic grid in order to obtain templates for spatially controlledadhesion and directed growth of cells.

2. Material and methods

2.1. Preparation and characterization of fullerene C60 layers

Fullerenes C60 (purity 99.5%, SES Research, USA) were depositedonto microscopic glass coverslips (Menzel Glaser, Germany; diameter12 mm) by evaporation of C60 in the Univex-300 vacuum system(Leybold, Germany) in the following conditions: room temperature ofthe substrates, C60 deposition rate≼1 Å/s, temperature of C60

evaporation in the Knudsen cells about 450 °C, time of depositionup to 50 min [2]. The thickness of the layers increased proportionallyto the temperature in the Knudsen cell and the time of deposition.Four types of layers of different thickness and morphology wereprepared: thin continuous, thick continuous, thin micropatterned andthick micropatterned. The micropatterned layers were created bydeposition of fullerenes through a metallic mask with rectangularholes with an average size of 128±3 µm/98±8 µm (about 12 430 µm2)and 50 µm in distance (Fig. 1).

Raman spectra of the deposited C60 films were measured in theback-scattering geometry at room temperature using a Ramanmicroscope Ramascope 1000 (Renishaw, UK) with a 514.5 nmexcitation wavelength Ar-ion laser [2]. The polarization of thescattered radiation was not analyzed. The thickness of the C60 layerwas measured by atomic force microscopy (AFM, Digital Instru-ments CP II Veeco, U.S.A.). A scratch was made in the layer and itsprofile was measured in contact mode. A Veeco CONT20A-CPscanning probe with spring constant 0.9 N/m was used. Repeatedscanning of the same area of the scratch (5 times) was used forthickness measurement with error max.±5%. In addition, thesurface morphology and root mean square roughness (RMS) wasevaluated by AFM.

The surface wettability of the films was estimated from thecontact angle measured by a static method in a material-waterdroplet system using a reflection goniometer (SEE System, MasarykUniversity, Brno, Czech Republic) [2].

2.2. Cells and culture conditions

To cultivate the cells, the fullerene-coated glass coverslips weresterilized by 70% ethanol for 1 h, inserted into 24-well polystyrenemultidishes (TPP, Switzerland; diameter 15 mm) and seeded withhuman osteoblast-like MG 63 cells (European Collection of CellCultures, Salisbury, UK). Each dish contained 5000 cells (i.e., about2,830 cells/cm2) and 1.5 ml of Dulbecco's modified Eagle's MinimumEssential Medium (DMEM; Sigma, U.S.A., Cat. N° D5648) supplemen-ted with 10% fetal bovine serum (FBS; Sebak GmbH, Aidenbach,Germany) and gentamicin (40 µg/ml, LEK, Ljubljana, Slovenia). Thecells were cultured for 1, 3 and 5 or 7 days at 37 °C in a humidified airatmosphere containing 5% of CO2.

2.3. Evaluation of the cell number, morphology and spatial distributionon the material surface

On days 1, 3, 5 and 7 after seeding, the cells were rinsed withphosphate-buffered saline (PBS; Sigma, U.S.A.), fixed with 70%ethanol (−20 °C, 5 min) and stained with hematoxylin and eosin(Fluka) according to the manufacturer's protocol. The number,shape and distribution of the cells on the material surface (namelytheir relation to the grooves and bulges on the two micropatternedsurfaces) were evaluated on microphotographs taken under an IX-51 microscope, equipped with a DP-70 digital camera (both fromOlympus, Japan).

In addition, on days 5 and 7 after seeding, when the cells werebecoming confluent and overlapping, the cell number was evaluated byautomated counting in a ViCell XR analyzer (Beckman Coulter, U.S.A.)after they had been detached by a trypsin-EDTA solution (Sigma, U.S.A.,Cat. N° T4174) in PBS for 10 min at 37 °C.

The cells on the continuous and micropatterned C60 layers wereevaluated in two separate experiments performed in triplicate. Sincethe adhesion and growth activity of MG 63 cells varied slightly in eachexperiment, internal controls represented by polystyrene culturedishes (TPP, Switzerland) and microscopic glass coverslips (MenzelGlaser, Germany), cleansed with ethanol, were included in eachexperiment [4].

2.4. Growth curves and cell population doubling time

The cell numbers were expressed as cell population densities/cm2

and were used for constructing the growth curves. The cell populationdoubling time (DT) was calculated as DT=(t− to)log 2/log Nt-log Nto,where to and t represent earlier and later time intervals after seeding,respectively, and Nto and Nt represent the number of cells at theseintervals.

2.5. Evaluation of cell viability

On days 1, 3 and 7 after seeding, the cells were rinsed with PBS,and their number and viability was determined by the LIVE/DEADviability/cytotoxicity kit for mammalian cells (Invitrogen, MolecularProbes, U.S.A.) according to the manufacturer's protocol. Briefly, thecells were incubated for 5 to 10 min at room temperature in amixture of two of the following probes: calcein AM, a marker ofesterase activity in living cells, emitting green fluorescence, andethidium homodimer-1, which penetrated into dead cells throughtheir damaged membrane and produced red fluorescence. Live anddead cells were then counted on microphotographs taken under anepifluorescence microscope (Olympus IX-51, DP-70 digital camera,

579L. Grausova et al. / Diamond & Related Materials 18 (2009) 578–586

Author's personal copy

Japan). In addition, on days 5 and 7 after seeding, the dead cellswere detected by the trypan-blue exclusion test performed duringcell counting in the ViCell XR analyzer mentioned above.

2.6. Immunofluorescence staining of β1 integrins, talin, β-actin andosteopontin

On day 3 after seeding, the presence and spatial arrangementof the following molecules in MG 63 cells was evaluated: β1

integrins, i.e., an important group of cell-matrix adhesionreceptors on osteoblasts, supporting their differentiation andtheir sensitivity to the material surface properties [29]; talin, anintegrin-associated protein present in focal adhesion plaques, β-actin, an important component of the cytoplasmatic cytoskeleton;and osteopontin, a non-collagenous calcium-binding extracellularmatrix glycoprotein and marker of osteogenic cell differentiation.The cells were rinsed twice in PBS and fixed with cold 70%ethanol (−20 °C, 5 min), pre-treated with 1% bovine serumalbumin in PBS containing 0.05% Triton X-100 (Sigma, St. Louis,MO, U.S.A.) for 20 min at room temperature, and then incubatedwith the following primary antibodies: mouse monoclonal anti-

human integrin β1, mouse monoclonal anti-human talin (bothfrom Chemicon International Inc., Temecula, CA, U.S.A.; Cat No.MAB1981 and MAB3264, respectively), monoclonal anti-β-actin(clone AC-15, Sigma, St. Louis, MO, U.S.A., Cat. No. A-5441) orrabbit polyclonal anti-human osteopontin (Alexis Biochemicals,Cat No. ALX-210-309). The antibodies were diluted in PBS toconcentrations of 1:200 to 1:400 and applied overnight at 4 °C.After rinsing with PBS, the secondary antibodies, represented bygoat anti-mouse F(ab′)2 fragment of IgG (for samples stained withmouse monoclonal antibodies; dilution 1:1000) or goat anti-rabbitF(ab′)2 fragment of IgG (for samples stained with the rabbitpolyclonal antibody; dilution 1:5000) were added for 1 h at roomtemperature. Both secondary antibodies were conjugated withAlexa Fluor® 488 and purchased from Molecular Probes, Eugene,OR, U.S.A. (Cat. No. A11017 and A11070, respectively). Afterincubation with secondary antibodies, the cells were rinsedtwice in PBS, mounted under microscopic glass coverslips in aGel/Mount permanent fluorescence-preserving aqueous mountingmedium (Biomeda Corporation, Foster City, CA, U.S.A.) andevaluated under an epifluorescence microscope (IX-51, Olympus,Japan) equipped with a digital camera (DP-70, Olympus, Japan).

Fig. 1. Morphology of thick continuous fullerene C60 layers (A, B), metallic grid used for creating micropatterned C60 layers (C) and morphology of thin (D, F) and thick (E)micropatterned C60 layers (F: on a bulge). Olympus IX 50 microscope equipped with DP 70 digital camera (A: bar = 200 µm, D, E: bar = 1 mm); Olympus BX41 with DP 70 digitalcamera (C: bar = 1 mm); atomic force microscope Digital Instruments CP II Veeco (B, F).

580 L. Grausova et al. / Diamond & Related Materials 18 (2009) 578–586

Author's personal copy

2.7. Measurement of cell adhesion area

Cells immunostained against β-actin were used for an evaluation ofthe size of the cell spreading area, i.e. the cell-material contact area. Thecells were photographed using a IX 51 microscope (obj. 20) equippedwith a DP 70 digital camera (Olympus, Japan) in 20 randomly selectedmicroscopic fields (size approx. 1.38×10−3 cm2) for each experimentalgroup. The sizeof the areaprojectedon thematerialwasmeasuredusingAtlas software (Tescan Ltd., Brno, Czech Republic). The cells thatdeveloped intercellular contacts were excluded from the evaluation.For each experimental group, 3 independent samples (containing 35 to113 cells in total) were evaluated.

2.8. Statistical analysis

The quantitative data on the physical properties of the materialwas presented as mean±S.D. (Standard Deviation). The quantitative

data obtained in the cells was presented as mean±S.E.M. (StandardError of the Mean). If two experimental groups were compared,Student's t-test for unpaired data was used, whereas multiplecomparison procedures were performed by the One Way Analysis ofVariance (ANOVA), Student–Newman–Keuls method, using SigmaStatsoftware (Jandel Corp. U.S.A.). P values equal to or less than 0.05 wereconsidered significant.

3. Results

3.1. Physical and chemical properties of the fullerene C60 layers

Raman analysis was performed on micropatterned samples imme-diately after deposition and then after sterilization with ethanol.Immediately after deposition, the Raman spectra showed that thefullerene films were prepared with high quality, confirmed by a highpeak Ag(2) at wavenumber 1468 cm−1, low peaks Hg(7) and Hg(8) and

Fig. 2. Raman analysis of a C60 layer immediately after deposition on glass coverslips (A) and after 1 h-sterilizationwith 70% ethanol (B–E). A–C: thin micropatterned layer; D, E: thickmicropatterned layer. B, D: Gaussian peak fit analysis of the Raman spectra; C, E: Raman spectra below the openings (red line) or bars (blue line) of the metallic lattice. (Forinterpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

581L. Grausova et al. / Diamond & Related Materials 18 (2009) 578–586

Author's personal copy

absence of D (disorder, ~1350 cm−1) and G (graphitic, ~1600 cm−1)bands, which are signs of fragmentation and graphitization of C60,respectively (Fig. 2A). After sterilization with ethanol, the thin micro-patterned fullerene layers were almost intact (Fig. 2B), and a consider-able amount of fullerenes was found not only on sites underlying theopenings of the grid, but also below its metallic part (Fig. 2C). However,in thick micropatterned layers, an analysis of the vibration mode Ag(2)showed that some of the C60 molecules reacted with oxygen orpolymerized, as indicated by the satellite peak A(2) on Fig. 2D. Theproportion of C60 molecules involved in these chemical changes wasdeterminedby the ratioA(2)/A(1), and reached about50%.Moreover, theamount of fullerenes below the metallic bars of the grid was very low,though still detectable (Fig. 2E).

As revealed by AFM, the thickness of the thin and thickcontinuous layers was 505±43 nm and 1090±8 nm, respectively.In both micropatterned layers, the fullerenes formed bulge-likeprominences below the holes of the metallic grid. The thickness ofthese bulges was 484±5 nm and 1043±57 nm in the thin and thickfilm, respectively. A fullerene C60 film was also found below themetallic part of the grid, especially in the thin layer, where its

thickness amounted to 158±5 nm. Therefore, the effective height ofthe fullerene domains in the thin micropatterned layer was only326±5 nm. In the thick micropatterned layers, the thickness of thefullerene film formed below the metallic portion of the grid wasless than 57 nm. The surface of both continuous and micropatternedlayers which contained nanosize clusters of the RMS value from4.97 to 6.32 nm (Fig. 1B and F).

Optical microscopy revealed that the fullerene layers werebrownish in color (Fig. 1A, D and E). The color intensity increasedwith layer thickness, while the transparency of the layers in aconventional light microscope decreased (Fig. 1D and E). Despite ofthis, the cells on both continuous layers were well observable, eventhose native and non-stained (Fig. 3A–D). On thick micropatternedlayers, the bulge-like prominences were relatively dark, and thecontrast between the bulges and grooves was relatively high(Fig. 1E). In addition, it was not possible to focus the cells onbulges and in grooves simultaneously, whereas the fluorescencesignal from both groups of cells was observable. Thus, the presenceand morphology of cells on bulges and in grooves was evaluatedusing fluorescence microscopy (Fig. 3F).

Fig. 3. Morphology of human osteoblast-like MG 63 cells in cultures on a polystyrene dish (A) a microscopic glass coverslip (B), thin continuous (C), thick continuous (D), thinmicropatterned (E) and thick micropatterned (F) fullerene C60 layers. A–D: native cultures; E: a culture stained with hematoxylin and eosin; F: a culture stained with LIVE/DEADviability/cytotoxicity kit (Invitrogen). A–D: day 5 after seeding, E–F: day 7 after seeding. Olympus IX 50 microscope, DP 70 digital camera, obj. 20x, bar=200 µm except E, wherebar=100 µm.

582 L. Grausova et al. / Diamond & Related Materials 18 (2009) 578–586

Author's personal copy

Reflection goniometry showed that all fullerene C60 layerswere relatively highly hydrophobic. The continuous and micro-patterned layers had similar water drop contact angles rangingfrom 95.3±3.1° to 100.6±6.8°.

3.2. Adhesion and proliferation of cells on the fullerene C60 layers

On day 1 after seeding, the cells on both continuous thin and thickfullerene layers adhered at similar numbers (3420±420 cells/cm2 and2880±440 cells/cm2, respectively), whichwas comparable to the valuesfound on standard cell culture substrates, represented by the tissueculture polystyrene dish (3080±290 cells/cm2) and the microscopicglass coverslip (2560±310 cells/cm2). On both micropatterned thin andespecially thick fullerene layers, the average cell population densitiestended to be lower (by 11 to 43%) than both polystyrene and glass, butthese differences were not statistically significant (Fig. 4). However, the

cells on the thick micropatterned layers were less spread and adheredover a significantly smaller cell adhesionarea (983±34µm2 compared to1815±131µm2,1757±16µm2,1878±193µm2and2286±148µm2 in thincontinuous, thick continuous and thin micropatterned layers, respec-tively, and 2286±148 µm2 on microscopic glass coverslips). These cellsoften had a roundish or elongated spindle-like shape, whereas the cellson both the continuous and thin micropatterned fullerene layers weremostly polygonal (Fig. 3).

From days 1 to 5 after seeding, the cells on both continuous C60films, polystyrene and glass, proliferatedwith a similar cell populationdoubling time (Table 1), and on day 5 they reached similar cellpopulation densities (from 91 600±9500 cells/cm2 to 112 700±9800cells/cm2). However, on the thick micropatterned fullerene layers, thedoubling time was significantly longer and the final cell populationdensity (measured on day 7 after seeding) was significantly lowerthan the values obtained on the thin micropatterned films (Fig. 4). Theattenuated growth of the cells on thick micropatterned layers wasprobably due to the presence of relatively high prominences and deep

Table 1Doubling time of human osteoblast-like MG 63 cells in cultures on tissue culturepolystyrene dishes, microscopic glass coverslips, continuous thin and thick fullerene C60

layers (A) and micropatterned thin and thick fullerene C60 layers (B)

Sample/Doubling time I. Polystyrene II. Glass Continuous C60 films

III. Thin IV. Thick

A.Day 1 to 5 19.6±0.5 18.6±0.3 19.1±0.4 19.0±0.6

Sample/Doubling time I. Polystyrene II. Glass Micropatterned C60 films

III. Thin IV. Thick

B.Day 1 to 7 21.7±1.1 III 19.8±0.4 18.5±0.9 I, IV 22.2±0.9 III

Means±S.E.M. Statistical analysis: ANOVA, Student–Newman–Keuls Method. Statisticalsignificance: I, III, IV: p≤0.05 in comparison with the sample of the same number.

Fig. 4. Growth curves of human osteoblast-like MG 63 cells in cultures on polystyrenedishes, microscopic glass coverslips, continuous thin and thick fullerene C60 layers (A)and micropatterned thin and thick fullerene C60 layers (B). Mean±S.E.M. from 9 to 30measurements. Statistical analysis: ANOVA, Student–Newman–Keuls Method. Statis-tical significance: ▲■: p≤0.05 in comparison with the sample of the same symbol.

Fig. 5. Population density of human osteoblast-like MG 63 cells on day 1 (A), 3 (B) and 7(C) after seeding on thin and thick fullerene (Full) C60 layers micropatterned withgrooves and bulges. Mean±S.E.M. from 9 measurements. Statistical analysis: Student t-test for the unpaired data. Statistical significance: ⁎⁎p≤0.01 and ⁎⁎⁎p≤0.001 incomparison with the values on the bulges.

583L. Grausova et al. / Diamond & Related Materials 18 (2009) 578–586

Author's personal copy

grooves on these surfaces. The cells colonized practically exclusivelythe grooves (Fig. 3F), thus they used less space for their proliferation.Although the grooves occupied only 41±1% of the material surface,they contained from 80±4% to 98±1% of the total cells on the materialsurface. The cell population density in the grooves was about 5 to 57times higher than on the bulges, and these differences increased withtime of cultivation (Fig. 5). On the other hand, on the thinmicropatterned films, the cells colonized homogeneously the entiresurface of the sample (Fig. 3E), and the percentage as well as thepopulation density of cells in the grooves and on the bulges weresimilar (Fig. 5).

3.3. Cell viability on the fullerene C60 layers

As revealed by staining the adhered cells with calcein andethidium homodimer using the LIVE/DEAD kit, the percentage ofviable cells on days 1 to 7 after seeding on all fullerene layers ranged

from 80±10% to 100±15%, and was similar to the values obtained onthe polystyrene culture dishes (99±24% to 100±15%), and also themicroscopic glass coverslips (88±21% to 99±26%). The trypan-blueexclusion test, performed on the trypsinized cells while they werebeing counted in the ViCell Analyser, also showed similar proportionsof viable cells on the fullerene layers (82±11% to 89±13%), as well asstandard cell culture surfaces (80±13% to 95±13%). The averageviability of the cells tended to increase with time of cultivation.

3.4. Presence and spatial arrangement of β1 integrins, talin, β-actin andosteopontin in cells on fullerene C60 layers

As revealed by immunofluorescence, MG 63 cells on bothcontinuous and micropatterned fullerene layers were intensivelystained for β1 integrins and talin, i.e. molecules participating incell-substrate adhesion, β-actin, an important component of thecytoplasmic cytoskeleton, as well as osteopontin, a marker of

Fig. 6. Immunofluorescence staining of β1 integrins (A, E, I, M, Q), talin (B, F, J, N, R), β-actin (C, G, K, O, S) and osteopontin (D, H, L, P, T) in human osteoblast-like MG 63 cells on day3after seeding on microscopic glass coverslips (A–D), thin continuous (E–H), thick continuous (I–L) thin micropatterned (M–P) and thick micropatterned (Q–T) fullerene C60 layer.Olympus epifluorescence microscope IX 51, digital camera DP 70, obj. 100×, bar=20 µm.

584 L. Grausova et al. / Diamond & Related Materials 18 (2009) 578–586

Author's personal copy

osteogenic cell differentiation. This staining intensity was similaras in cells on the control polystyrene culture dish and microscopicglass coverslips (Fig. 6). All these molecules (particularly extra-cellular matrix protein osteopontin) were found in fine granulardistribution throughout the cells, often preferentially located inthe perinuclear region. In addition, both β1 integrins and talinalso formed dot- or streak-like focal adhesion plaques, visiblemainly on the cell periphery. Beta1 integrin-containing focaladhesion plaques were particularly well developed and wereoften located on fine long protrusions formed by cells, which wasaccompanied by the formation of a fine mesh-like β-actincytoskeleton. No apparent differences in the staining intensityand distribution of all molecules mentioned here were foundbetween cells growing on thin and thick micropatterned fullerenelayers or in cells in grooves and on bulges.

4. Discussion

This study has shown that fullerene C60 layers acted as goodsubstrates for adhesion, growth and maturation of human osteoblast-like MG 63 cells, comparable to standard cell culture materials, such astissue culture polystyrene and microscopic glass coverslips. Theseresults are consistent with the earlier findings of the supportive effectsof fullerene C60 layers formed on polystyrene culture dishes by theevaporation of methanol from colloidal C60-methanol suspensions, onthe adhesion, proliferation and assembly of an actin cytoskeleton inseveral lines of normal andmalignant breast epithelial cells [3]. GraftingC60 molecules onto polyurethane surfaces also increased the number ofadhered platelets and enhanced their activation [4].

The supportive effects of fullerene C60 layers on cell colonizationcould be explained by their surface nanostructure, mimicking thenanoarchitecture of the natural extracellular matrix (ECM), i.e. aphysiological substrate for cell adhesion [22–26]. In the case of thickmicropatterned layers, the cell colonization could be supported alsoby the presence of oxygen, revealed by Raman spectroscopy. Oxygen-containing groups have repeatedly been found to enhance thecolonization of various materials with cells [30,31]. Both thenanostructure and the oxygen content in our C60 layers may actsynergetically by enhancing the adsorption of cell adhesion-mediatingECM molecules, such as vitronectin, fibronectin and collagen,provided by the serum of the culture medium or synthesized by thecells [22–24,30]. The spatial conformation of these molecules on ourfilms may also be more appropriate for the accessibility of specificsites on these molecules by cell adhesion receptors, e.g. integrins [22–24,30]. As revealed by immunofluorescence staining of theMG 63 cellson our samples, β1 integrins (i.e., receptors for collagen, fibronectinand vitronectin), and talin, an integrin-associated protein, wereassembled in focal adhesion plaques, and these processes wereaccompanied by the formation of β-actin cytoskeleton and thepresence of a considerable amount of osteopontin, a marker ofosteogenic cell differentiation. All these events can be considered assigns of active cell-substrate interaction followed by signal transduc-tion inside the cells. Moreover, β1 integrins have been reported toenhance osteogenic cell differentiation, which is manifested by higherlevels of osteocalcin, higher activity of alkaline phosphatase and amore pronounced response to 1,25-dihydroxyvitamin D3 [29].

Similarly, a water-soluble fullerene derivative, C3-fullero-tris-methanodicarboxyllic acid, supported the formation of focaladhesion plaques, an assembly of an actin cytoskeleton and cellspreading in human epidermoid carcinoma cells exposed to UVlight, probably by its antioxidative action [28]. The free radicalscavenger activity of fullerenes and their derivatives alsoprotected keratinocytes from the apoptosis induced by ultravioletlight B [5], minimized the oxidative damage of small boweltransplants [14], decreased neuronal degeneration and death,prevented focal cerebral ischemia [15,16] and attenuated ische-

mia-reperfusion-induced lung injury [17]. Adducts of the C60

fullerenes with poly(N-vinyl-pyrrolidone) prevented the distur-bance of long-term memory consolidation in rats induced bycycloheximide, an inhibitor of protein synthesis [18]. However,polyhydroxylated fullerene derivatives protected cells againstoxidative stress only at low concentrations (10–50 µM in cellculture media), while at high concentrations (1 and 1.5 mM), theycaused cell death [17]. Hydroxylated fullerenes, or water-solublenano-C60, i.e. a fullerene aggregate that readily forms whenpristine C60 is added to water, induced oxidative damage to thecell membrane and DNA fragmentation in various types of normalor tumor cells, and these effects were more pronounced at higherfullerene concentrations [8,9]. In human vascular endothelial cells,water-soluble hydroxylated fullerenes caused cell death probablyby the activation of ubiquitin-autophagy pathways [11]. Suspen-sions of colloidal C60 fullerenes were genotoxic for humanlymphocytes [9]. These harmful effects of fullerenes C60, usuallypotentiated by the light, especially ultraviolet light, could beutilized for antimicrobial therapy or photodynamic therapyagainst tumor cells [12,13].

Nevertheless, as shown by Raman spectroscopy, the fullerenelayers in this study and also in our earlier study [2], were relativelystable and resistant to sterilization and cell cultivation procedures.Similarly, fullerene molecules grafted onto polyurethane remainedintact after tests of their interaction with platelets [4]. Therefore, itcan be supposed that if the fullerenes were released from our filmsinto the cell culture media (which remains to be further investi-gated), only low and harmless concentrations of these moleculeswere achieved. On the other hand, MG 63 cells on thickmicropatterned fullerene layers adhered over a significantly smallerarea than those on the control glass coverslips. However, this lowerspreading of cells, coupled with their elongated morphology, can beexplained by the preferential growth of cells in relatively narrowgrooves rather than by cytotoxic effects of the fullerene layer. Onthin micropatterned fullerene layers, i.e. surfaces with relativelylow irregularities, the MG 63 cells were normally spread in a similarextent as on the control cell culture substrates. Another factorsupporting lower cell spreading on fullerene films may be therelatively high hydrophobicity of these materials. Highly hydro-phobic surfaces have been often shown to be less adhesive for cells[30,31].

Preferential growth of primary osteoblasts as well as osteo-genic cell lines in grooves, pits and other types of hollows hasbeen observed on various polymeric and metallic materialsapplicable for construction of bone and dental implants [32–35].However, the cells were also able to colonize the prominences onthese surfaces, represented e.g. by ridges or pillars, althoughthese prominences were usually much higher (i.e., several µm oreven tens of µm) than those on the thick micropatternedfullerene layers in our present study (only about 1 µm). Inaddition, bone cells, including MG 63 cells, were also able tospread over pore entrances from 40 to even 100 µm in diameteron the surface of scaffolds constructed for bone tissue engineering[36,37]. Surprisingly, on our fullerene C60 layers, the MG 63 cellswere not able to “climb up” relatively low prominences onlyabout 1 µm in height, even at a relatively late culture interval of7 days after seeding. This might be due to a synergistic action ofhydrophobia and other physicochemical properties of the full-erene bulges less appropriate for cell adhesion, such as their steeprise and the tendency of spherical ball-like fullerene C60

molecules to diffuse out of the prominences towards the grooves[38].

Thus, micropatterned fullerene films could be used as templates forregionally-selective cell adhesion and growth. Moreover, the osteogeniccells growing in hollows have been reported to be more active inphosphorylation of various kinases and transcription factors, signal

585L. Grausova et al. / Diamond & Related Materials 18 (2009) 578–586

Author's personal copy

transduction and differentiation [32–35]. Microstructured fullerenelayers (if they adhere strongly to the underlying materials and areresistant to the release of fullerene molecules) could be used as abioactive coating for bone implants in order to improve their integrationwith the surrounding bone tissue.

5. Conclusion

Fullerenes C60 deposited as continuous films or layers micro-patterned with grooves and bulges clearly supported the adhesion,growth, viability and maturation of human osteoblast-like MG 63cells to a similar extent as standard tissue culture polystyrenedishes and microscopic glass coverslips. In addition, thick micro-patterned layers (i.e., with fullerene prominences 1043±57 nm inheight) promoted regionally-selective adhesion and growth of MG63 cells in the grooves located among the prominences.

Acknowledgements

This study was supported by the Grant Agency of the CzechRepublic (Grants No. 204/06/0225 and 101/06/0226), the Acad. Sci.CR (Grants No. KAN400480701 and IAA400100701), and theMinistry of Education, Youth and Sports of the CR (Program No.6046137302 and LC 06041). We also thank Mrs. Ivana Zajanova(Inst. Physiol., Acad. Sci. CR) for her excellent technical assistance inimmunofluorescence staining. Mr. Robin Healey (Czech TechnicalUniversity, Prague) is gratefully acknowledged for his languagerevision of the manuscript.

References

[1] A. Djordjević, G. Bogdanović, S. Dobrić, J. BUON 11 (2006) 391.[2] L. Bacakova, L. Grausova, J. Vacik, A. Fraczek, S. Blazewicz, A. Kromka, M. Vanecek,

V. Svorcik, Diamond Relat. Mater. 16 (2007) 2133.[3] N. Levi, R.R. Hantgan, M.O. Lively, D.L. Carroll, G.L. Prasad, J. Nanobiotechnol. 4

(2006) 14.[4] J.C. Lin, C.H. Wu, Biomaterials 20 (1999) 1613.[5] C. Fumelli, A. Marconi, S. Salvioli, E. Straface, W. Malorni, A.M. Offidani, R.

Pellicciari, G. Schettini, A. Giannetti, D. Monti, C. Franceschi, C. Pincelli, J. Invest.Dermatol. 115 (2000) 835.

[6] S. Foley, C. Crowley, M. Smaihi, C. Bonfils, B.F. Erlanger, P. Seta, C. Larroque,Biochem. Biophys. Res. Commun. 294 (2002) 116.

[7] D.M. Guldi, M. Prato, Acc. Chem. Res. 33 (2000) 695.

[8] C.M. Sayes, A.M. Gobin, K.D. Ausman, J. Mendez, J.L. West, V.L. Colvin, Biomaterials26 (2005) 7587.

[9] A. Dhawan, J.S. Taurozzi, A.K. Pandey, W. Shan, S.M. Miller, S.A. Hashsham, V.V.Tarabara, Environ. Sci. Technol. 40 (2006) 7394.

[10] A. Isakovic, Z. Markovic, B. Todorovic-Markovic, N. Nikolic, S. Vranjes-Djuric, M.Mirkovic, M. Dramicanin, L. Harhaji, N. Raicevic, Z. Nikolic, V. Trajkovic, Toxicol. Sci.91 (2006) 173.

[11] H. Yamawaki, N. Iwai, Am. J. Physiol., Cell. Physiol. 290 (2006) C1495.[12] Y. Yamakoshi, N. Umezawa, A. Ryu, K. Arakane, N. Miyata, Y. Goda, T. Masumizu, T.

Nagano, J. Am. Chem. Soc. 125 (2003) 12803.[13] Y.J. Tang, J.M. Ashcroft, D. Chen, G. Min, C.H. Kim, B. Murkhejee, C. Larabell, J.D.

Keasling, F.F. Chen, Nano Lett. 7 (2007) 754.[14] H.-S. Lai, Y. Chen, W.-J. Chen, K.-J. Chang, L.-Y. Chiang, Transplant. Proc. 32 (2000)

1272.[15] S.S. Huang, S.K. Tsai, C.L. Chih, L.Y. Chiang, H.M. Hsieh, C.M. Teng, M.C. Tsai, Free

Radic. Biol. Med. 30 (2001) 643.[16] D.Y. Yang, M.F. Wang, I.L. Chen, Y.C. Chan, M.S. Lee, F.C. Cheng, Neurosci. Lett. 311

(2001) 121.[17] Y.-W. Chen, K.C. Hwang, C.-C. Yen, Y.-L. Lai, Am. J. Physiol., Regul. Integr. Comp.

Physiol. 287 (2004) R21.[18] I.Ya. Podolskii, E.V. Kondrateva, I.V. Shcheglov, M.A. Dumpis, L.B. Piotrovskii, Phys.

Solid State 44 (2002) 578.[19] E. Nakamura, H. Isobe, Acc. Chem. Res. 36 (2003) 807.[20] T. Kawase, K. Tanaka, Y. Seirai, N. Shiono, M. Oda, Angew. Chem., Int. Ed. Engl. 42

(2003) 5597.[21] Z. Zhou, D.I. Schuster, S.R. Wilson, J. Org. Chem. 68 (2003) 7612.[22] T.J. Webster, C. Ergun, R.H. Doremus, R.W. Siegel, R. Bizios, J. Biomed. Mater. Res. 51

(2000) 475.[23] R.L. Price, M.C. Waid, K.M. Haberstroh, T.J. Webster, Biomaterials 24 (2003) 1877.[24] K.M. Woo, V.J. Chen, P.X. Ma, J. Biomed. Mater. Res. 67A (2003) 531.[25] J. Tan, W.M. Saltzman, Biomaterials 25 (2004) 3593.[26] G. Wei, P.X. Ma, Biomaterials 25 (2004) 4749.[27] T. Tsuchiya, Y.N. Yamakoshi, N. Miyata, Biochem. Biophys. Res. Commun. 206

(1995) 885.[28] E. Straface, B. Natalini, D. Monti, C. Franceschi, G. Schettini, M. Bisaglia, C. Fumelli,

C. Pincelli, R. Pellicciari, W. Malorni, FEBS Lett. 454 (1999) 335.[29] L. Wang, G. Zhao, R. Olivares-Navarrete, B.F. Bell, M. Wieland, D.L. Cochran, Z.

Schwartz, B.D. Boyan, Biomaterials 27 (2006) 3716.[30] L. Bacakova, K. Walachova, V. Svorcik, V. Hnatowitz, J. Biomater. Sci., Polym. Ed. 12

(2001) 817.[31] R. Mikulikova, S. Moritz, T. Gumpenberger, M. Olbrich, C. Romanin, L. Bacakova, V.

Svorcik, J. Heitz, Biomaterials 26/27 (2005) 5572.[32] J.L. Charest, L.E. Bryant, A.J. Garcia, W.P. King, Biomaterials 25 (2004) 4767.[33] H. Kenar, G.T. Köse, V. Hasirci, Biomaterials 27 (2006) 885.[34] D.W. Hamilton, D.M. Brunette, Biomaterials 28 (2007) 1806.[35] B.J. Papenburg, L. Vogelaar, L.A. Bolhuis-Versteeg, R.G. Lammertink, D. Stamatialis,

M. Wessling, Biomaterials 28 (2007) 1998.[36] V. Karageorgiou, D. Kaplan, Biomaterials 26 (2005) 5474.[37] E. Pamula, L. Bacakova, E. Filova, J. Buczynska, P. Dobrzynski, L. Noskova, L.

Grausova, J. Mater. Sci., Mater. Med. 19 (2007) 425.[38] S. Guo, D.P. Fogarty, P.M. Nagel, S.A. Kandel, Surf. Sci. 601 (2007) 994.

586 L. Grausova et al. / Diamond & Related Materials 18 (2009) 578–586