Cytocompatibility of Polyurethane foams as ... - SAGE Journals

Journal of Applied Biomaterials & Biomechanics 2003; 1: 58-66

Cytocompatibility of polyurethane foams as biointegrable matrices for the preparation of scaffolds for bone reconstruction M.C. TANZI1, S. FARÈ1, P. PETRINI1, A. TANINI2, E. PISCITELLI2, S. ZECCHI-ORLANDINI3, M.L. BRANDI2 Bioengineering Department, Politecnico di Milano, Milano - Italy Departments of Internal Medicine, University of Firenze 3 Department of Anatomy, Histology and Forensic Medicine, University of Firenze, Firenze - Italy 1 2

ABSTRACT: This work reports preliminary results on the development of biointegrable scaffolds, composed of biostable 3D polymer matrices and bioabsorbable inorganic salts, to be used for cell anchorage in bone regeneration. Three crosslinked polyurethane foams (PUFs), prepared by one-step bulk polymerisation from a polyether-polyol mixture, polymeric MDI and water as expanding agent, were tested for their ability to promote adhesion and growth of bone-derived cells. The open porosity of these foams ranged from 16 to 31% with an average pore size of 470÷600 µm, compressive strength (at 10% ε) of 0.28÷0.38 MPa and elastic moduli of 4.88÷6.61 MPa. The human osteosarcoma line Saos-2, and primary cultures of normal human articular chondrocytes and bone marrow-derived (HBM) stromal cells were used for in vitro cytocompatibility tests. For cell adhesion and proliferation analysis, DNA synthesis was evaluated by 3H-thymidine uptake. Osteoblastic differentiation of Saos-2 adherent cells was determined by measuring the enzymatic activity of alkaline phosphatase (ALP). All cell types were able to adhere to all tested PUFs and to synthesize DNA. At 48 hr culture, HBM stromal cells showed the maximal rate of adhesion with the highest rate of proliferation onto PUFs with the largest pore size, whereas both chondrocytes and Saos-2 appeared to adhere preferentially onto foams exhibiting the highest percentage of open porosity. Up to 8 days in culture Saos-2 cells were able to proliferate into all PUFs, with a time-dependent increase of DNA synthesis and ALP activity. At SEM, the morphology of cells adherent to PUF pores was spread with cytoplasmatic extroflessions, indicating a good metabolic activation. These results demonstrate a good cytocompatibility of the proposed 3D matrices, suggesting that their use in the preparation of composite scaffolds is worth further investigation. (Journal of Applied Biomaterials & Biomechanics 2003; 1: 58-66) KEY WORDS: Macroporous matrices, Biointegration, Bone reconstruction, Saos-2, Human chondrocytes, HBM stromal cells Received 15/01/03; Revised 29/01/03; Accepted 03/02/03

Bone reconstruction is essential to provide functional integrity to the patient’s skeleton in a variety of clinical situations. As autologous tissue has a limited availability, and allografts may exhibit incompatibility, the need for alternative bone graft materials is of primary interest. Various synthetic and bi-

ological materials have been proposed, including hydroxyapatite and calcium phosphates, demineralized bone matrix, collagens, bioactive glasses, polymers and composites (1). In recent years, the use of 3D polymer matrices has attracted the interest of various groups for their potential in tissue regeneration as they can be designed and fabricated into structures that have sufficient porosity for dif-

1722-6899/058-09$04.50/0

© Società Italiana Biomateriali

INTRODUCTION

Tanzi et al

fusion of nutrients and clearance of wastes, they present increased area for cell anchorage and growth, adequate mechanical stability and appropriate surface properties to allow the expression of normal cell phenotype (2). Biodegradable polymers are promising, as in principle they can be designed to degrade in vivo in a controlled manner over a predetermined implantation period, allowing replacement by newly regenerated tissues. Biodegradable polymers that have been investigated for bone regeneration include; poly(α-hydroxyesters)(3-5), polydioxanone (6), poly(orthoesters) (7), polyanhidrides (8) and polyurethanes (9, 10). However, the proper equilibrium between the rate of degradation in vivo and tissue regeneration has not yet been attained. Moreover, there is a great concern regarding the degradation mechanism, which includes the local concentration of monomers and oligomers and the production of undegraded particulates. The amount of acidic monomers released during degradation may reduce the local pH, which in turn accelerates the degradation rate of the polymer (11) and induces an inflammatory reaction. The formation of undegraded microparticles due to the different degradation rate between the crystalline and amorphous regions of the polymer may also elicit foreign body inflammatory reactions such as bone resorption (12). An alternative to biodegradation is biointegration. Biointegration can be achieved with “biostable” polymer scaffolds that offer wider possibilities of design. Polyurethanes (PU) represent a good choice for their versatility and the possibility to obtain a wide range of structures and properties, differently from biodegradable polyesters. Our research approach attempts to develop innovative scaffolds for bone reconstruction made of biointegrable polyurethane foams, which can be used as matrices in 3D composites containing osteoconductive bioabsorbable inorganic salts (calcium phosphates). These scaffolds should provide in the short-medium term the essential mechanical support for cell anchorage and new bone deposi-

tion. In the long term, the synthetic support will be engulfed by the bone and, possibly, slowly degraded into small quantities of non toxic degradation products that would be rapidly eliminated. In this case, undesired effects such as the activation of local aspecific inflammatory response that occurs in the case of polylactide and polyglycolide will be avoided. The first step of this research consisted in the preparation of crosslinked polyurethane foams (PUFs) with controlled porosity and density, and in testing their cytocompatibility and ability to promote the adhesion and growth of bone-derived cells.

MATERIALS AND METHODS Composition of the PU foams Cross-linked polyurethane foams (PUFs) were prepared by reacting in a purposely designed mold a polyether-polyol mixture (component A, Elastogran, Italy) with polymeric MDI (component B, B141, BASF), using Fe-acetyl-acetonate (FeAA) or dibutyl-tin-dilaurate (DBTDL) as catalysts (0.001% w/wA), and water as expanding agent (2% w/wA). Component A is identified by a hydroxyl number (nOH) of 223.0 ± 1.0 mgKOH/g (DIN 53240) and does not contain a premixed catalyst; component B is identified by a –NCO content of 22.9 ± 0.2% (ASTM D 1638-74). By varying the catalyst and the ratio between OH groups in component A, and NCO groups in component B, as shown in Table I, different chemical compositions were obtained.

Synthesis of the foams The PU foams were obtained by a one-step bulk polymerization. Water and catalyst were added to the weighed amount of component A, and mixed with a mechanical stirrer at 2000 rpm for 40 sec. The appropriate quantity of component B was then added, the reaction mixture was stirred for 90 sec, and poured in a custom-made polymethylmetacry-

TABLE I - IDENTIFICATION AND FORMULATION OF THE PU FOAMS Foam

A/B(*) (eqOH/eqNCO)

% H2O (% w/wA)

Catalyst (0.001% w/wA)

1200f 1210f 1210d

1.00/1.00 1.00/1.10 1.00/1.10

2 2 2

FeAA FeAA DBTDL

(*) calculated not considering the eqNCO involved in the reaction with water

59

Cytocompatibility of polyurethane foams

late (PMMA) mold to allow the expansion of the foams into a fixed volume. The mold was firmly closed with a PMMA cover by means of stainless steel screws and the expanding reaction was allowed to take place at room temperature by CO2 production from water/isocyanate reaction. Each foam was extracted from the mold after 24 hr, and post-cured at room temperature for 7 days. The foams were purified by a 48 hr immersion in absolute ethanol, at room temperature.

Characterization of the foams The obtained foams were characterized for their density, porosity, and pore size. Density and porosity analyses were made on cylindrical specimens (Ø 15 mm, h 10 mm), cut from each foam with a die. Density was analyzed according to the standard practice for the determination of the apparent volumic mass of cellular plastics and rubbers (EN ISO 845). After conditioning 24 hr at 25°C, the specimens (n=5) were weighed and measured according to the standard practice for linear dimensions of cellular plastics and rubbers (EN ISO 1923). Porosity, expressed as percentage of open pores, was evaluated according to the EN ISO 4590 standard practice, with a custom-made picnometer. This device measures the volume impenetrable to a gas, compared to the total volume of a rigid cellular plastic. The analyses were performed on three specimens for each foam, after conditioning for 24 hr at 25°C. The average cell size was analyzed according to the standard test method for pore size of rigid cellular plastics (ASTM D 3576-94). Rectangular specimens 0.2-0.3 mm thick were cut perpendicularly to the foam growth plane and observed with a stereo microscope (WILD M8) at 50X magnification. By the stereo microscope, preferential cell orientation and foam disuniformity were also observed at 12X and 25X magnification. Compression tests were performed on cylindrical samples (n = 5, Ø 15 mm, h 10 mm) with an Instron model 4200 instrument, at a cross-head rate of 1 mm/min. Tangent modulus (E) and stress at 10% deformation (σ10%) were drawn from stress/strain curve elaboration. The ethanol extracts were analyzed by size exclusion-high performance liquid chromatography (SEC-HPLC). The analyses were performed at 40°C with a Waters LC system equipped with a refractive index detector, using dimethylformamide (DMF) as eluent at a flow rate of 0.8 ml/min, and a set of three Styragel columns (HR-3, HR-4, HR-5). Calibration was made with PMMA standard samples. 60

Cell cultures The human osteosarcoma cell line Saos-2 (9) was obtained from the American Type Culture Collection (ATCC, HTB85, Rockville, MD, USA), and grown in Coon’s modified Ham’s F12 medium (Gibco, Grand Island, NY, USA) supplemented with 10% Fetal Calf Serum (FCS, Gibco, Grand Island, NY, USA), L-glutamine (2 mM), penicillin (100 IU/ml), and streptomycin (100 mg/ml). The cells were grown at 37°C in humidified atmosphere of 5% CO2 in air. Human articular cartilage cells (chondrocytes) and human bone marrow (HBM) stromal cells were obtained from healthy normal volunteers. A written declaration of consent was obtained from each volunteer. Fragments of the articular cartilage were incubated in Coon’s modified Ham’s F12 medium containing 0.125% trypsin and 0.2% collagenase type IV, at 37 °C in atmosphere of 5% CO2 in air. After 2hr-enzymatic digestion, cell aggregates were mechanically dispersed with a Pasteur pipette and centrifuged. The resulting pellet was plated in a tissue culture dish (Ø 100 mm) and grown in Coon’s modified Ham’s F12 medium supplemented with 10% FCS, L-glutamine (2 mM), penicillin (100 IU/ml), and streptomycin (100 µg/ml). A suspension of HBM stromal cells was plated in plastic culture dishes at various concentrations (from 2x105 to 2x107 cells/dish) and incubated for 7 days in humidified atmosphere of 5% CO2 in air. The adherent cells were washed with phosphate buffered saline (PBS, pH 7.4), and cultured in Coon’s modified Ham’s F12 medium containing 10% Nu-serum and 1% ultroser-G, penicillin (100 IU/ml) and streptomycin (100 µg/ml). For cytocompatibility tests, disks (Ø 15mm, h 2 mm) were cut from the PU foams and immersed in absolute ethanol for 3 hours, washed three times with PBS and maintained overnight in PBS. PUF disks were then fitted into the bottom of the wells (1 disk/well) of 24-multiwell plates (Falcon, Lincoln Park, NJ, USA) and Saos-2, chondrocytes, and HBM stromal cells (5x104 cells/well), respectively, were seeded in growth medium onto the top of each disk and allowed to adhere and proliferate. 3

H-thymidine uptake

To assess cell proliferation onto the PU foams, DNA synthesis was evaluated by 3H-thymidine (Sigma, St Louis, MO, USA) uptake. After seeding, all cell types were allowed to adhere to PUF disks for 48 hr. Saos-2 cells were continued in culture up to eight days. Cells were maintained overnight in serum

Tanzi et al

free medium and then 3H-thymidine (2 µCi/ml) was added during the last 4 hours of culture. The medium was discarded from each well and the cells were washed twice with PBS. Thrichloroacetic acid 5% (TCA) was added for 15 minutes at 4°C and replaced twice. After removing the supernatant, 0.5 N NaOH (500 µl/well) was added. The cells were maintained for 30 min in a humidified stove at 8090 °C. 0.5 N HCl (500 µl/well) was added, and aliquots of the supernatant were analyzed by liquid scintillation spectroscopy (Betamatic, Kontron Instruments, Milan, Italy). All the experiments were carried out at least in triplicate.

ture for 1 hr), quickly dehydrated in ascending ethanol series, dried in air, and sputter-coated with 10% gold-palladium. Observations were performed with a Cambridge Stereoscan 100 SEM at 15 keV, with a tilt angle varying from 0° to 40°.

Statistical analyses All the tests were performed at least in triplicate. One-way ANOVA was used to assess the statistical significance of the data.

RESULTS Alkaline phosphatase (ALP) activity Physical properties of PU foams Osteoblastic differentiation was determined by measuring the enzymatic activity of ALP. This was determined with a diagnostic kit (Sigma, St Louis, MO, USA) according to the supplier’s instructions. Saos-2 cells (5x105 cells/ml) were plated on PUF disks in 24 multiwell plates and allowed to grow for two, four and eight days. After this time Nonidet 1% in PBS (500 µl/disk) was added for 30 min and aliquots (100 µl) of lysate were tested for ALP activity. ALP activity was determined as the rate of conversion of p-nitrophenyl phosphate to p-nitrophenol and inorganic phosphate. When made alkaline, p-nitrophenol is converted to a yellow complex readily measured at 400-420 nm by a spectrophotometer. The intensity of colour measured as optical density (OD) is proportional to the phosphatase activity. Results were expressed as alkaline phosphatase units corresponding to this reading from a calibration curve. A phosphatase unit is defined as the amount of enzyme activity that will liberate 1mM of p-nitrophenol per hour under the test condition described in literature (14).

As shown in Table II, higher porosity values of the foams are related to lower density values, whereas there is no evident relationship between chemical composition and foam density. At the stereo microscope, the morphology of all the three foams appeared rather uniform, with no preferential cell orientation (Fig. 1). Compressive parameters of PUFs are reported in Table II, the highest values corresponding to the lowest percentage of open porosity. All the three PUFs showed the typical stress/strain behavior of elasto-plastic foams with an initial linear-elastic regime followed by a plateau of roughly constant stress corresponding to the pores collapse. HPLC analyses of the foam extracts indicated the presence of low molecular weight products that were efficiently extracted by ethanol. As shown in Table III, the molecular weight of these substances is much lower than that of the polyol mixture reagent (Component A) and can be attributed to by-products or unreacted matter and therefore may be recognised as potentially noxious.

Scanning electron microscopy (SEM)

Cell adhesion and DNA synthesis

The morphology of adherent chondrocytes, HBM stromal cells and Saos-2 cells was evaluated at SEM. The samples were fixed in glutaraldehyde (2% in 0.1 M phosphate buffer, pH 7.2, at room tempera-

As shown in Figure 2 a-c, all cell types were able to adhere and synthesize DNA. HBM stromal cells showed the highest rate of cell adhesion. Notably, these cells adhered better to the PU foam with

TABLE II - PHYSICAL AND COMPRESSIVE PROPERTIES OF THE PU FOAMS Foam

Density (g/cm3)

Open porosity (%)

Average cell size (µm)

E (MPa)

σ10% (MPa)

1200f 1210f 1210d

0.134 ± 0.003 0.166 ± 0.004 0.112 ± 0.005

31 ± 2 16 ± 3 25 ± 2

609 522 477

4.88 ± 0.39 6.31 ± 0.47 6.00 ± 0.88

0.28 ± 0.01 0.38 ± 0.03 0.35 ± 0.04

61

Cytocompatibility of polyurethane foams

a

b

Fig. 2 - 3H-thymidine uptake (cpm/well) of a) human bone marrow (HBM) stromal cells, b) human chondrocytes, c) Saos-2 cells grown for 48 hr onto the three different PU foams. All cell types were able to adhere and synthesize DNA. HBM showed the highest rate of cell adhesion, particularly on PU foam with greater pore size (1200f). Both chondrocytes and Saos-2 adhered tightly to the foams exhibiting the highest percentage of open porosity (1200f and 1210d). Data are normalised vs the respective control foams (with no added cells).

c Fig. 1 - Stereo microscope images of the PU foams: a) 1200f; b) 1210f; c) 1210d.

62

greater pore size. This could be attributed to their greater size (up to 200 mm when adherent and spread) if compared to chondrocytes and Saos-2. In fact, the 1210d foam, with an average pore size