Surface modification of a porous polyurethane through a culture of human osteoblasts and an electromagnetic bioreactor

Technology and Health Care 15 (2007) 33–45 33IOS Press

Surface modification of a porouspolyurethane through a culture of humanosteoblasts and an electromagnetic bioreactor

L. Fassinaa,∗, L. Visaib, M.G. Cusella De Angelisc, F. Benazzod and G. MagenesaaDipartimento di Informatica e Sistemistica, University of Pavia, Pavia, ItalybDipartimento di Biochimica, University of Pavia, Pavia, ItalycDipartimento di Medicina Sperimentale, University of Pavia, Pavia, ItalydDipartimento SMEC, IRCCS San Matteo, University of Pavia, Pavia, Italy

Abstract. There is increasing interest in new biomaterials and new culture methods for bone tissue engineering, in order toproduce,in vitro, living constructs able to integrate in the surrounding tissue. Using an electromagnetic bioreactor (magneticfield intensity, 2 mT; frequency, 75 Hz), we investigated the effects of electromagnetic stimulation on SAOS-2 human osteoblastsseeded onto a porous polyurethane. In comparison with control conditions, the electromagnetic stimulation caused higher cellproliferation, increased surface coating with decorin and type-I collagen, and higher calcium deposition. The immunolocalizationof decorin and type-I collagen showed their colocalization in the cell-rich areas. The use of an electromagnetic bioreactor aimedat obtaining the surface modification of the porous polyurethane in terms of cell colonization and coating with calcified matrix.The superficially modified biomaterial could be used, in clinical applications, as an implant for bone repair.

Keywords: Electromagnetic stimulation, osteoblast, cell proliferation, extracellular matrix, decorin, type-I collagen, calcium,surface modification, biomimetics

1. Introduction

One of the key challenges in bone tissue engineering is the development of new biomaterials and newculture methods to provide living constructs that possess the ability to integrate in the surrounding tissue.

The culture method is of great importance. Static culture environments suffer from limited diffusionand often result in inhomogeneous cell and extracellular matrix distribution. In order to overcomethe drawbacks associated with static culture systems, several bioreactors have been designed: thehigh-aspect-ratio vessel rotating bioreactor and the flow perfusion bioreactor, for instance [5,25]. Theideal bioreactor supplies suitable levels of oxygen, nutrients, cytokines, growth factors, and physicalstimulation, in order to populate, with bone cells and their extracellular matrix, the volume of the bonegraft substitute.

Nevertheless, an easier alternative way could be the surface modification of the bulk material: Castnerand Ratner defined the “biocompatible surfaces” as the surfaces with the characters of a “clean, fresh

∗Address for correspondence: Lorenzo Fassina, Ph.D., University of Pavia, Dipartimento di Informatica e Sistemistica, ViaFerrata 1, 27100 Pavia, Italy. Tel.: +390382985352; Fax: +390382985373; E-mail: [email protected].

0928-7329/07/$17.00 2007 – IOS Press and the authors. All rights reserved

34 L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts

wound” [8,22]. In order to obtain biocompatible surfaces, the non-specific protein adsorption has beeninhibited [16,17] or a biomimetic strategy has been developed [10,18,23], for instance. In the presentwork we have showed a biomimetic strategy that consisted in the surface modification of a biomaterialwith proliferated bone cells and their calcified extracellular matrix producedin loco.

In previous work, using a perfusion bioreactor to stimulate SAOS-2 osteoblasts, we produced calcifiedmatrix inside a hydrophobic and non-biodegradable polyurethane [11]. In this study, using the samepolyurethane and an “electromagnetic” bioreactor, we attempted to populate the biomaterial surface withcalcified matrix and osteoblasts, of which cell function can be electromagnetically modulated in termsof proliferation and differentiation [1,9,15,19].

Using this approach, a scaffold, coated by human bone proteins and calcium minerals, could be used,in clinical applications, as an implant for bone repair [4]. In addition, the superficially living constructcould be implanted together with the insertion of a vascular pedicle [3].

2. Materials and methods

2.1. Cells

The human osteosarcoma cell line SAOS-2 was obtained from the American Type Culture Collection(HTB85, ATCC). The cells were cultured in McCoy’s 5A modified medium with L-glutamine andHEPES (Cambrex Bio Science), supplemented with 15% fetal bovine serum, 2% sodium pyruvate, 1%antibiotics, 10−8 M dexamethasone, and 10 mMβ-glycerophosphate (Sigma-Aldrich). Ascorbic acid,another osteogenic supplement, is a component of McCoy’s 5A modified medium. The cells werecultured at 37◦C with 5% CO2, routinely trypsinized after confluency, counted, and seeded onto theporous polyurethane scaffolds.

2.2. Three-dimensional polyurethane foam

The three-dimensional polyurethane foam was synthesized and characterized as previously de-scribed [11].

Briefly, the crosslinked polyurethane foam was synthesized by a one-step bulk polymerization, usinga polyether-polyol mixture (component A, ElastoCoat, Elastogran) with polymeric MDI (component B,BASF), using 0.001% (w/wA) Fe-acetyl-acetonate as catalyst and 2% (w/wA) water as expanding agent.Water and Fe-acetyl-acetonate were added to the weighed amount of the component A, and mixed witha mechanical stirrer at 2,000 rpm for 40 s. The appropriate quantity of the component B was added,and the reaction mixture was stirred for 90 s and poured in a custom-made polymethylmethacrylatemold in order to allow the expansion of the foam into a fixed volume. The mold was firmly closedand the expanding reaction was allowed to take place at room temperature by CO2 production fromwater/isocyanate reaction. The polyurethane foam was extracted from the mold after 24 h and post-curedat room temperature for 7 days.

In order to remove potentially noxious molecules characterized by low molecular weight, i.e. by-products and unreacted matter, the foam was purified by an immersion in absolute ethanol for 48 h atroom temperature.

Cylindrical specimens (diameter, 15 mm; height, 10 mm) were cut from the foam with a manual die formorphological and mechanical characterization. Density was evaluated according to the UNI EN ISO845 standard practice. Porosity, expressed as the percentage of open pores, was evaluated according to

L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts 35

the EN ISO 4590 standard practice, with a custom-made picnometer that measured the gas-impenetrablevolume compared with the total volume of a rigid cellular plastic. The average pore size was determinedaccording to the ASTM D 3376-94 standard practice. A compressive mechanical test was performedon 5 cylindrical specimens with an Instron 4200 instrument at a cross-head rate of 1 mm/min, both indry conditions and after 7 days of treatment in distilled water (the tests were performed maintaining thespecimens in water at 37◦C for all the test time). Compressive moduli Edry and Ewet, in dry and in wetconditions, respectively, were obtained from stress-strain curve elaboration.

Cell culture scaffolds (diameter, 15 mm; height, 2 mm) were cut from the foam with a manual die.

2.3. Cell seeding

The scaffolds were sterilized by ethylene oxide at 38◦C for 8 h at 65% relative humidity. After a24-h aeration in order to remove residual ethylene oxide, the scaffolds were press-fitted inside the twoculture systems: the “static” culture system, that is, a standard 24-well-plate far from the electromagneticbioreactor, and the “dynamic” culture system, that is, a standard 24-well-plate inside the electromagneticbioreactor. A cell suspension of 1× 106 cells in 400µl was added onto the top of each scaffold and, after0.5 h, 1 ml of culture medium was added to cover the scaffolds. Cells were allowed to attach overnight,and then the static culture continued in the well-plate far from the electromagnetic bioreactor, which wasturned on.

2.4. Electromagnetic bioreactor

An electromagnetic bioreactor was built. It consisted of a carrying structure custom-machined in atube of polymethylmethacrylate: the windowed tube carried a well-plate and two solenoids, the planesof whom were parallel (Fig. 1). In this experimental setup the magnetic field and the induced electricfield were perpendicular and parallel to the scaffold surfaces, respectively.

The surfaces of the polyurethane scaffolds were 5 cm distant from each solenoid plane, and thesolenoids were powered by a Biostim SPT pulse generator (Igea), a generator of Pulsed ElectromagneticFields (PEMFs).

Given the position of the solenoids and the characteristics of the pulse generator, the electromagneticstimulation had the following parameters: intensity of the magnetic field equal to 2± 0.2 mT, amplitudeof the induced electric tension equal to 5± 1 mV, signal frequency of 75± 2 Hz, and pulse duration ofabout 1.3 ms.

In vivo experiments demonstrated that a continuous exposure to a pulsed electromagnetic field, similarto that used in this study, stimulates the bone repair in the healing process of transcortical holes in adulthorses [7].

The magnetic field was measured with a Hall Effect transverse gaussmeter-probe and a gaussmeter(Laboratorio Elettrofisico), the induced electric tension was measured with a standard coil-probe, andthe temporal pattern of the electromagnetic signal was evaluated by a digital oscilloscope.

The electromagnetic bioreactor was placed into a standard cell culture incubator with an environmentof 37◦C and 5% CO2. The dynamic culture was stimulated by the PEMF 24 h per day for a total of 22days. The culture medium was changed on days 4, 7, 10, 13, 16, and 19.

2.5. Standard well-plate culture

The static culture was placed into a different incubator, where the PEMF stimulation was not detectable.The culture medium was changed on days 4, 7, 10, 13, 16, and 19.

36 L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts

Fig. 1. Electromagnetic bioreactor; the windowed tube carried a well-plate and two solenoids, the planes of whom were parallel:in this experimental setup the magnetic field and the induced electric field were perpendicular and parallel to the culture surfaces,respectively.

2.6. Light microscopy analysis

Scaffolds were fixed with 4% (w/v) paraformaldehyde solution in 0.1 M phosphate buffer (pH= 7.4)for 8 h at room temperature, washed with phosphate buffer, dehydrated in a gradient ethanol series upto 100%, and embedded in paraffin. The histological sections, with 10µm thickness, were stained withhematoxylin/eosin.

2.7. Light microscopy analysis for calcium detection

The histological sections were stained with the von Kossa method (BioOptica). This staining substitutescalcified matrix Ca2+ ions with Ag atoms, revealing the calcified matrix as brown regions: the ratiobetween brown region area and total image area was measured by ImageJ (http://rsb.info.nih.gov/ij) [11].The normalized results are expressed as %/(cell× scaffold).

2.8. DNA content

Cells were lysed by a freeze-thaw method in sterile deionized distilled water. The released DNAcontent was evaluated with a fluorometric DNA quantification kit (PicoGreen, Molecular Probes). ADNA standard curve, obtained from a known amount of osteoblasts, was used to express the results ascell number per scaffold.

L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts 37

2.9. Scanning electron microscopy (SEM) analysis

Scaffolds were fixed with 2.5% (v/v) glutaraldehyde solution in 0.1 M Na-cacodylate buffer (pH=7.2) for 1 h at 4◦C, washed with Na-cacodylate buffer, and then dehydrated at room temperature ina gradient ethanol series up to 100%. The samples were kept in 100% ethanol for 15 min, and thencritical point-dried with CO2. The specimens were mounted on aluminum stubs, sputter-coated with gold(degree of purity equal to 99%), and then observed with a Leica Cambridge Stereoscan 440 microscopeat 8 kV.

2.10. Set of rabbit polyclonal antisera

L.W. Fisher (http://csdb.nidcr.nih.gov/csdb/antisera.htm, National Institutes of Health, National In-stitute of Dental and Craniofacial Research, Craniofacial and Skeletal Diseases Branch, Matrix Bio-chemistry Unit, Bethesda, MD) presented us, generously, with the following rabbit polyclonal antibodyimmunoglobulins G: anti-decorin and anti-type-I collagen.

2.11. Immunohistochemistry

Cultured scaffolds were fixed with 4% (w/v) paraformaldehyde solution in 0.1 M phosphate buffer(pH = 7.4) for 8 h at room temperature, washed with phosphate buffer, dehydrated in a gradient ethanolseries up to 100%, and embedded in paraffin. The sections, with 10µm thickness, were cut orthogonallyto the scaffold axis, deparaffinized, rehydrated, and processed for the detection of decorin and type-Icollagen by the avidin-biotin-peroxidase method. L.W. Fisher’s anti-decorin and anti-type-I collagenantisera were used as primary antibody with a dilution equal to 1:4000 and 1:2000, respectively.

2.12. Extraction of the extracellular matrix proteins from the cultured scaffolds and ELISA assay

At the end of the culture period, the cultured scaffolds were washed extensively with sterile PBS(137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 1.4 mM KH2PO4, pH = 7.4) in order to remove theculture medium, and then incubated for 24 h at 37◦C with 1 ml of sterile sample buffer (1.5 M Tris-HCl,60% [w/v] sucrose, 0.8% [w/v] Na-dodecyl-sulphate, pH= 8.0).

At the end of the incubation period, the sample buffer aliquots were removed, and then the scaffoldswere centrifuged at 4000 rpm for 15 min in order to collect the sample buffer entrapped into the pores.The total protein concentration, both in the static and in the dynamic systems, was evaluated by the BCAProtein Assay Kit (Pierce Biotechnology). Total protein concentration was 1900± 105µg/ml in thestatic culture, whereas 2800± 141µg/ml in the dynamic culture withp < 0.05. Unseeded sterilizedpolyurethane scaffolds were incubated for 24 h at 37◦C with 1 ml of sterile sample buffer, and no proteincontent was detected.

The calibration curves to measure decorin and type-I collagen were performed by an ELISA assay.L.W. Fisher’s anti-decorin and anti-type-I collagen antisera were used. The normalized results areexpressed as fg/(cell× scaffold).

2.13. Statistics

Results are expressed as mean± standard deviation. In order to compare the results between static anddynamic systems, one-way analysis of variance (ANOVA) withpost hoc Bonferroni test was applied,electing a significance level of 0.05.

38 L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts

Fig. 2. SEM image of an unseeded polyurethane scaffold, 18× magnification.

Fig. 3. SEM image of an unseeded polyurethane scaffold, 240× magnification.

3. Results

3.1. Polyurethane foam characterization

The foam had the following characteristics: density equal to 0.098± 0.002 g/cm3, open porosity of80%± 2%, average pore diameter equal to 624µm, compressive moduli Edry, equal to 9.82± 0.40 MPa,and Ewet, equal to 4.44± 0.31 MPa, in dry and in wet condition, respectively.

The high porosity value was related to a low density. By SEM observation the foam morphologyappeared rather uniform with smooth pore surfaces (Figs 2 and 3). The pores appeared spherical andinterconnected. As expected, the polyurethane foam exhibited a higher compressive modulus in the drycondition than in the wet.

L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts 39

Fig. 4. Histological section of the static culture, hematoxylin/eosin staining, 400× magnification (asterisk near cell, arrow=polyurethane).

3.2. Electromagnetic culture of SAOS-2 osteoblasts

The cells were seeded onto the surface of polyurethane scaffolds, and then cultured in an electromag-netic bioreactor for 22 days. This culture system permitted the study of SAOS-2 cells as they proliferatedand produced a calcified extracellular matrix in an electromagnetically active environment.

We compared the cell-matrix distribution, calcified matrix production, and microscope images betweenthe two culture systems.

3.3. Light microscopy and SEM analyses

Statically and dynamically cultured scaffolds showed the presence of cells and matrix only insidethe pores of the upper surface, that is, at the peripheral depths of the culture scaffolds. Nevertheless,the images revealed that the electromagnetic stimulation improved the cell distribution on the scaffoldsurface.

Statically cultured cells were few and essentially organized in a monolayer with a thin discontinuousextracellular matrix (Figs 4 and 5). The electromagnetic stimulation induced a 3D modeling of thecell-matrix organization: several cells homogeneously coated the polyurethane surface in a multilayerwith a highly developed matrix; the volume of the surface pores was tending to be filled by cell-matrixclusters growing from the pore bottom (Figs 6 and 7).

The preceding observations were confirmed by the measure of the DNA content after 22 days ofculture. In the static culture the cell number per scaffold grew to 18.7× 106 ± 8.2 × 104, in thedynamic culture to 36.3× 106 ± 8.4 × 104 with p < 0.05. Since the DNA may remain entrappedin the calcified matrix, an underestimation of the culture cellularity is possible; in addition, since theelectromagnetic stimulation increased the calcification of the matrix (Section 3.4), the underestimationof the DNA content should be higher in the dynamic culture than in the static.

3.4. Calcified matrix deposition

The relative amount of the calcium contained in the scaffolds was quantified in order to evaluate thematrix calcification (Figs 8 and 9). The ratio between the brown area and the total image area was 0.11

40 L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts

Fig. 5. SEM image of the static culture, 300× magnification (asterisk near cell, arrow= polyurethane).

Fig. 6. Histological section of the dynamic culture, hematoxylin/eosin staining, 400× magnification (asterisk near cell, arrow= polyurethane).

× 10−6 ± 0.03× 10−6 %/(cell× scaffold) in the static culture, whereas 0.28× 10−6 ± 0.10× 10−6

%/(cell × scaffold) in the dynamic: the electromagnetic stimulation increased the matrix calcificationaround 2.5-fold withp < 0.05.

3.5. Immunohistochemistry

The immunolocalization of the matrix proteins showed an intense intracellular and extracellular stainingof the cell-rich areas in both culture systems. In comparison with the static culture, we observed highly

L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts 41

Fig. 7. SEM image of the dynamic culture, 300× magnification (asterisk near cell, arrow= polyurethane).

Fig. 8. Histological section of the static culture, von Kossa staining, 400× magnification (asterisk near cell, brown= calcifiedextracellular matrix, arrow= polyurethane).

developed matrix in the dynamic culture (Figs 10 and 11).

3.6. Surface coating with matrix proteins

In order to evaluate the polyurethane surface coating with matrix proteins, an ELISA assay of theextracted matrix was performed. In comparison with the static culture, the electromagnetic stimulationgreatly increased the coating with bone proteins (Table 1). These measures were in accordance with the

42 L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts

Fig. 9. Histological section of the dynamic culture, von Kossa staining, 400× magnification (asterisk near cell, brown=calcified extracellular matrix, arrow= polyurethane).

Fig. 10. Immunolocalization of type-I collagen in the static culture, 400× magnification (asterisk near cell, brown= type-Icollagen, arrow= polyurethane); the images related to decorin were similar.

preceding microscope images of highly developed matrix.

4. Discussion

The aim of this study was to modify the surface of a biomaterial with calcified matrix and osteoblasts,using an electromagnetic bioreactor, to make it more suitable to bone repair. In order to enhance thecoating of the bulk biomaterial, an electromagnetic wave was applied to the seeded scaffolds because acontinuous exposure to a pulsed electromagnetic field can stimulate the bone repair [7].

The electromagnetic stimulation increased the cell proliferation around 2-fold. This result is consistentwith the rise in proliferation [6], in growth factor [1,12] and prostaglandin [20] secretion, and in growth

L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts 43

Table 1Protein coating of cultured scaffolds

Static culture, Stat, [fg/(cell× scaffold)] Dynamic culture, Dyn, [fg/(cell× scaffold)] Dyn/StatDecorin 94.55± 2.43 124.13± 12.53 1.3Type-I collagen 969.08± 186.13 9722.20± 408.40 10.0

Fig. 11. Immunolocalization of type-I collagen in the dynamic culture, 400× magnification (asterisk near cell, brown= type-Icollagen, arrow= polyurethane); the images related to decorin were similar.

factor receptor number [14] in response to an electromagnetic wave.Aaron and Ciombor reported a significant increase in the extracellular matrix synthesis when the

osteoblast-like cells were subjected to an electromagnetic wave [2]. Heermeier showed the electro-magnetic upregulation of type-I collagen [15]. According to the preceding studies the electromagneticbioreactor caused a significant increase in the extracellular matrix synthesis: in comparison with thestatic culture the coating with type-I collagen was enhanced around 10-fold, whereas the coating withdecorin around 1.3-fold. Type-I collagen is the most important and abundant structural protein of thebone matrix, and decorin is an important proteoglycan considered a key regulator for the assembly andthe function of many extracellular matrix proteins [24]. The immunolocalization of type-I collagen anddecorin showed their colocalization in the cell-rich areas. Considering the structural role of decorinand type-I collagen, we observed a concordant 2.5-fold increase in the matrix calcification inside theelectromagnetic bioreactor.

The electromagnetic stimulation raises the net Ca2+ flux in human osteoblast-like cells [13], and,according to Pavalko’s signaling model [21], the increase of the cytosolic Ca2+ concentration is thestarting point of signaling pathways targeting specific bone matrix genes. Considering the model ofPavalko, the analysis of the coating specific for decorin and type-I collagen revealed, concordantly, thatthe application of the electromagnetic wave caused, in comparison with the static culture, an increase inthe gene expression.

In this study, using an electromagnetic bioreactor, we enhanced the biomaterial coating with calcifiedextracellular matrix, that is, we followed a particular biomimetic strategy where the seeded cells built abiocompatible surface.

The use of a cell line showed the potential of the culture method, nevertheless, a better result could beobtained with autologous bone marrow stromal cells instead of SAOS-2 osteoblasts for total immuno-

44 L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts

compatibility with the patient. The cultured “self-surface” could be used fresh, that is, rich in autologouscells and calcified matrix, or after sterilization with ethylene oxide, that is, rich only in autologouscalcified matrix in order to handle a simpler storable tissue-engineering product for bone repair.

Acknowledgments

The authors thank Prof. L.W. Fisher, Prof. P. Speziale, Prof. A. Icaro-Cornaglia, Dr. S. Setti, Dr. R.Cadossi, Dr. S. Fare, Dr. P. Petrini, Mr. D. Picenoni, and Mr. A. Mortara for support. The Biostim SPTpulse generator was a gift from Igea (Carpi, Italy). This work was supported by Fondazione CariploGrant (2004) to Prof. F. Benazzo, by PRIN Grant (2004) from Italian Ministry of Education, Universityand Research to Dr. L. Visai, by FAR Grants (2004) from the University of Pavia to Prof. F. Benazzo,Prof. M.G. Cusella De Angelis, and to Prof. G. Magenes. The paper is dedicated to the artist NorahJones.

References

[1] R.K. Aaron, B.D. Boyan, D.M. Ciombor, Z. Schwartz and B.J. Simon, Stimulation of growth factor synthesis by electricand electromagnetic fields,Clin Orthop Relat Res (2004), 30–37.

[2] R.K. Aaron and D.M. Ciombor, Acceleration of experimental endochondral ossification by biophysical stimulation ofthe progenitor cell pool,J Orthop Res 14 (1996), 582–589.

[3] S. Akita, N. Tamai, A. Myoui, M. Nishikawa, T. Kaito, K. Takaoka and H. Yoshikawa, Capillary vessel networkintegration by inserting a vascular pedicle enhances bone formation in tissue-engineered bone using interconnectedporous hydroxyapatite ceramics,Tissue Eng 10 (2004), 789–795.

[4] H.C. Anderson, P.R. Reynolds, H.H. Hsu, L. Missana, K. Masuhara, P.E. Moylan and H.I. Roach, Selective synthesis ofbone morphogenetic proteins-1, -3, -4 and bone sialoprotein may be important for osteoinduction by Saos-2 cells,J BoneMiner Metab 20 (2002), 73–82.

[5] G.N. Bancroft, V.I. Sikavitsas, J. van den Dolder, T.L. Sheffield, C.G. Ambrose, J.A. Jansen and A.G. Mikos, Fluidflow increases mineralized matrix deposition in 3D perfusion culture of marrow stromal osteoblasts in a dose-dependentmanner,Proc Natl Acad Sci USA 99 (2002), 12600–12605.

[6] T. Bodamyali, B. Bhatt, F.J. Hughes, V.R. Winrow, J.M. Kanczler, B. Simon, J. Abbott, D.R. Blake and C.R. Stevens,Pulsed electromagnetic fields simultaneously induce osteogenesis and upregulate transcription of bone morphogeneticproteins 2 and 4 in rat osteoblastsin vitro, Biochem Biophys Res Commun 250 (1998), 458–461.

[7] V. Cane, P. Botti and S. Soana, Pulsed magnetic fields improve osteoblast activity during the repair of an experimentalosseous defect,J Orthop Res 11 (1993), 664–670.

[8] D.G. Castner and B.D. Ratner, Biomedical surface science: Foundations to frontiers,Surf Sci 500 (2002), 28–60.[9] M. De Mattei, A. Caruso, G.C. Traina, F. Pezzetti, T. Baroni and V. Sollazzo, Correlation between pulsed electromagnetic

fields exposure time and cell proliferation increase in human osteosarcoma cell lines and human normal osteoblast cellsin vitro, Bioelectromagnetics 20 (1999), 177–182.

[10] Y.R. Duan, Z.R. Zhang, C.Y. Wang, J.Y. Chen and X.D. Zhang, Dynamic study of calcium phosphate formation onporous HA/TCP ceramics,J Mater Sci Mater Med 15 (2004), 1205–1211.

[11] L. Fassina, L. Visai, L. Asti, F. Benazzo, P. Speziale, M.C. Tanzi and G. Magenes, Calcified matrix production by SAOS-2cells inside a polyurethane porous scaffold, using a perfusion bioreactor,Tissue Eng 11 (2005), 685–700.

[12] R.J. Fitzsimmons and D.J. Baylink, Growth factors and electromagnetic fields in bone,Clin Plast Surg 21 (1994),401–406.

[13] R.J. Fitzsimmons, J.T. Ryaby, F.P. Magee and D.J. Baylink, Combined magnetic fields increased net calcium flux in bonecells,Calcif Tissue Int 55 (1994), 376–380.

[14] R.J. Fitzsimmons, J.T. Ryaby, F.P. Magee and D.J. Baylink, IGF-II receptor number is increased in TE-85 osteosarcomacells by combined magnetic fields,J Bone Miner Res 10 (1995), 812–819.

[15] K. Heermeier, M. Spanner, J. Trager, R. Gradinger, P.G. Strauss, W. Kraus and J. Schmidt, Effects of extremely lowfrequency electromagnetic field (EMF) on collagen type I mRNA expression and extracellular matrix synthesis of humanosteoblastic cells,Bioelectromagnetics 19 (1998), 222–231.

[16] N.B. Holland, Y. Qiu, M. Ruegsegger and R.E. Marchant, Biomimetic engineering of non-adhesive glycocalyx-likesurfaces using oligosaccharide surfactant polymers,Nature 392 (1998), 799–801.

L. Fassina et al. / Surface modification of a porous polyurethane through a culture of human osteoblasts 45

[17] Y. Iwasaki, S. Sawada, N. Nakabayashi, G. Khang, H.B. Lee and K. Ishihara, The effect of the chemical structureof the phospholipid polymer on fibronectin adsorption and fibroblast adhesion on the gradient phospholipid surface,Biomaterials 20 (1999), 2185–2191.

[18] A. Krout, H.B. Wen, E. Hippensteel and P. Li, A hybrid coating of biomimetic apatite and osteocalcin,J Biomed MaterRes A 73 (2005), 377–387.

[19] C.H. Lohmann, Z. Schwartz, Y. Liu, H. Guerkov, D.D. Dean, B. Simon and B.D. Boyan, Pulsed electromagnetic fieldstimulation of MG63 osteoblast-like cells affects differentiation and local factor production,J Orthop Res 18 (2000),637–646.

[20] C.H. Lohmann, Z. Schwartz, Y. Liu, Z. Li, B.J. Simon, V.L. Sylvia, D.D. Dean, L.F. Bonewald, H.J. Donahue and B.D.Boyan, Pulsed electromagnetic fields affect phenotype and connexin 43 protein expression in MLO-Y4 osteocyte-likecells and ROS 17/2.8 osteoblast-like cells,J Orthop Res 21 (2003), 326–334.

[21] F.M. Pavalko, S.M. Norvell, D.B. Burr, C.H. Turner, R.L. Duncan and J.P. Bidwell, A model for mechanotransduction inbone cells: the load-bearing mechanosomes,J Cell Biochem 88 (2003), 104–112.

[22] B.D. Ratner and S.J. Bryant, Biomaterials: where we have been and where we are going,Annu Rev Biomed Eng 6 (2004),41–75.

[23] C. Sanchez, H. Arribart and M.M. Guille, Biomimetism and bioinspiration as tools for the design of innovative materialsand systems,Nat Mater 4 (2005), 277–288.

[24] P. Sini, A. Denti, M.E. Tira and C. Balduini, Role of decorin onin vitro fibrillogenesis of type I collagen,Glycoconj J 14(1997), 871–874.

[25] X. Yu, E.A. Botchwey, E.M. Levine, S.R. Pollack and C.T. Laurencin, Bioreactor-based bone tissue engineering: theinfluence of dynamic flow on osteoblast phenotypic expression and matrix mineralization,Proc Natl Acad Sci USA 101(2004), 11203–11208.