Acta Biomaterialia 6 (2010) 2484–2493

Contents lists available at ScienceDirect

Acta Biomaterialia

journal homepage: www.elsevier .com/locate /ac tabiomat

Candidate bone-tissue-engineered product based on human-bone-derived cellsand polyurethane scaffold q

Piotr Wozniak a, Monika Bil b, Joanna Ryszkowska b, Piotr Wychowanski c, Edyta Wróbel a, Anna Ratajska d,Gra _zyna Hoser e, Jacek Przybylski a, Krzysztof J. Kurzydłowski b, Małgorzata Lewandowska-Szumieł a,*

a Department of Biophysics and Human Physiology, Medical University of Warsaw, Chałubinskiego 5, 02-004 Warsaw, Polandb Faculty of Materials Science and Engineering, Warsaw University of Technology, Wołoska 141, 02-507 Warsaw, Polandc Department of Dental Surgery, Medical University of Warsaw, Nowogrodzka 59, 00-006 Warsaw, Polandd Department of Pathologic Anatomy, Medical University of Warsaw, Chałubinskiego 5, 02-004 Warsaw, Polande Department of Clinical Cytology, Medical Centre of Postgraduate Education, Marymoncka 99/103, 01-813 Warsaw, Poland

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

Article history:Received 10 January 2009Received in revised form 10 October 2009Accepted 13 October 2009Available online 4 November 2009

Keywords:Bone tissue engineeringTissue-engineered productHuman primary osteoblastsPolyurethane scaffoldsBioreactor

1742-7061/$ - see front matter � 2010 Published bydoi:10.1016/j.actbio.2009.10.022

q Research presented at the E-MRS 2008 SymposiumEngineering: Materials and Processing Methods, organand Prof. D.W. Hutmacher.

* Corresponding author. Tel.: +48 22 628 63 34; faxE-mail address: malgorzata.lewandowska-szumi

dowska-Szumieł).

Biodegradable polyurethanes (PURs) have recently been investigated as candidate materials for boneregenerative medicine. There are promising reports documenting the biocompatibility of selected PURsin vivo and the tolerance of certain cells toward PURs in vitro – potentially to be used as scaffolds for tis-sue-engineered products (TEPs). The aim of the present study was to take a step forward and create a TEPusing human osteogenic cells and a polyurethane scaffold, and to evaluate the quality of the obtained TEPin vivo. Human-bone-derived cells (HBDCs) were seeded and cultured on polyurethane scaffolds in a bio-reactor for 14 days. The TEP examination in vitro was based on the evaluation of cell number, cell phe-notype and cell distribution within the scaffold. TEPs and control samples (scaffolds without cells)were implanted subcutaneously into SCID mice for 4 and 13 weeks. Explants harvested from the animalswere examined using histological and immunohistochemical methods. They were also tested in mechan-ical trials. It was found that dynamic conditions for cell seeding and culture enable homogeneous distri-bution, maintaining the proliferative potential and osteogenic phenotype of the HBDCs cultured onpolyurethane scaffolds. It was also found that HBDCs implanted as a component of TEP survived and kepttheir ability to produce the specific human bone extracellular matrix, which resulted in higher mechan-ical properties of the harvested explants when preseeded with HBDCs. The whole system, including theinvestigated PUR scaffold and the method of human cell seeding and culture, is recommended as a can-didate bone TEP.

� 2010 Published by Elsevier Ltd. on behalf of Acta Materialia Inc.

1. Introduction products [4,5]. This opened a new field for PUR applications in

Polyurethanes (PURs) have been used as materials for the prep-aration of medical devices, mainly for cardiovascular applications,since the 1960s [1,2]. They were expected to be biostable; how-ever, their long-term in vivo use was found to be accompaniedby molecular instability in a biologically active environment[1,3]. In consequence, extensive research on understanding PURdegradation was performed, and resulted not only in the enhance-ment of PUR biostability for in vivo long-term applications, but alsoin obtaining a new class of PURs, i.e. biodegradable ones, designedto undergo hydrolytic or enzymatic degradation to non-cytotoxic

Elsevier Ltd. on behalf of Acta Mat

on New Scaffolds for Tissueized by Dr. W. Swieszkowski

: +48 22 628 78 46.el@wum.edu.pl (M. Lewan-

regenerative medicine, where safe degradability that allowsreplacement of the implanted material by the host tissue is ex-pected to be achieved. In particular, application of biodegradablePURs in bone reconstruction has been postulated [6]. It may be ta-ken into account as an alternative to the polymers of the poly(a-hydroxy acid) group, which are widely investigated as scaffoldsfor tissue (including bone) engineering. These polymers undergobulk degradation, with their mass reduction occurring much moreslowly than the decrease of their molecular weight, which isaccompanied by the autocatalytic effect of the acidic end groups.In consequence, the disintegration of the scaffold does not proceedgradually, but may be explosive, which in turn results in a suddendrop in pH at the implantation site. This is not the case with PURs.What is more, due the segmented structure of PURs, a wide rangeof physical and mechanical properties may be obtained, makingPURs an attractive alternative to the aliphatic polyesters. Besides,the elastomeric character of PURs facilitates surgical handling aswell as filling the tissue cavity at the implantation site, which addi-

erialia Inc.

P. Wozniak et al. / Acta Biomaterialia 6 (2010) 2484–2493 2485

tionally makes their use in bone reconstructive and regenerativesurgery encouraging.

Promising results of resorbable PUR experimental implantationto iliac crest defects in sheep have been reported by Gorna andGogolewski [7]. On the other hand, good tolerance of osteogeniccells toward PURs and PUR-based biodegradable scaffolds wasshown in the culture of rabbit- [8], mouse- [9] and rat- [10] bonemarrow stromal cells (BMSCs), and only recently also in human cellculture, i.e. MG-63 osteosarcoma cell line and mesenchymal cellsobtained from human bone marrow [11]. Altogether, good resultsof experimental intrabone implantation of degradable PURs, aswell as promising observations of osteogenic cells behavior inthe PUR scaffolds in vitro, call for the next step, i.e. experimentalimplantation of the hybrid system consisting of human osteogeniccells in a biodegradable PUR scaffold. In this work we decided tomake an attempt to create a bone-tissue-engineered product com-posed of human osteogenic cells (human-bone-derived cells,HBDCs) and biodegradable polyurethane scaffolds, and to verifythe survival and the functioning of cells transplanted within thedeveloped system.

2. Materials and methods

2.1. HBDC culture

Human-bone-derived cells were isolated from the trabecularbone tissue chips which were harvested from the bottom distalpart of the long tight bone during the standard procedure of theknee joint alloplasty (and would otherwise be discarded). Materialfrom two female patients, aged 65 and 64, was used. Cells har-vested from different donors were cultured separately. Theyshowed similar proliferation capacity as verified in a parallelobservation (not shown). For each experiment, cells from one do-nor were used. All procedures were approved by the Local EthicsCommittee of the Warsaw Medical University (Decision No. KB/74/2005) and the donors provided informed consent.

The procedure of HBDC isolation was based on the protocols de-scribed by Gallagher et al. [12], with modifications [13]. Briefly,soft tissue found on the extracted bone pieces was removed byscraping, and the bone was cut into small, about 1 mm-size, pieces.They were then washed with PBS (containing Ca2+ and Mg2+ions;Gibco) and subsequently digested overnight using a collagenaseXI S (200 U ml�1) (Sigma–Aldrich). After that, samples werewashed with PBS again and placed in a 150 ml cell culture flask(20–30 bone pieces per flask) (Costar-Nunc), containing standardculture medium based on Dulbecco’s Modified Eagle Medium(DMEM) (Gibco) supplemented with 10% (final concentration) fetalbovine serum (FBS) (Gibco), L-glutamine (1% solution in medium –based on stock solution (100�), Gibco), 1% antibiotic–antimycoticmixture (containing 10,000 units of penicillin (base), 10 mg ofstreptomycin (base), and 25 lg of amphotericin B ml�1 – GibcoBRL, Paisley, UK), and 100 lM L-ascorbic acid 2-phosphate (Sigma).They were cultured in a CO2 incubator (at 37 �C in 5% CO2, humid-ified atmosphere). The medium was changed every 7 days. Within2–4 weeks, cell migration from the bone chips occurred. Afterreaching a confluence, the HBDCs were detached by means of col-lagenase and trypsin digestion, and used in experiments.

2.2. In vitro 3D scaffold preparation

The three-dimensional porous scaffold was manufactured froma PUR2PCL530 degradable polyurethane, which was synthesizedusing 4,40-methylenebis (cyclohexyl isocyanate) (HMDI), 530 Damolecular weight polycaprolactone diol (PCL diol) and ethyleneglycol (EG). The molar ratio of PCL:HMDI:EG was 1:2:1. This type

of polyurethane was selected as the most suitable from six typesof degradable polyurethanes which differed regarding: the molec-ular weight of the PCL diol used for their synthesis; molar ratio ofPCL:HMDI:EG; and the molecular weight (Mw) of the polycaprolac-tone diol. The synthesis details and material characteristics havebeen described elsewhere [14]. The most appropriate material formaking a 3D support for HBDCs was selected on the basis of cellviability and osteogenic potential, reflected by alkaline phospha-tase activity determined in our previous experiments [15–17].

Porous polyurethane structures were fabricated by a polymercoagulation/salt particulate leaching method. The typical experi-mental procedure is outlined as follows. The polyurethanes weredissolved in 1-methyl-2-pyrrolidone at the concentration of 15%(w/w). NaCl particles were hand-sieved with a stacked stainlesssteel fraction of NaCl crystals in the range of 200–300 lm. Themass ratio of the polymer to NaCl particles was 1:5. The poly-mer/salt/solvent mixture was poured into a Teflon mould (6-mmdiameter) and immersed in distilled water for 2 days, while precip-itation of the polymer and leaching of the salt particles were takingplace. The obtained porous polyurethane samples were dried un-der vacuum at 37 �C, and then sterilized using ionizing irradiation(25 kGy). It was shown in the previous study that the sterilizationof the investigated PUR under the applied conditions does notcause any toxic consequences toward HBDCs in vitro [18]. The scaf-folds’ open porosity, measured via mercury intrusion porosimetry,was about 70%, and the average macro-pore size was 203 lm, asevaluated by a quantitative analysis of the SEM images. The porousstructure of the obtained scaffolds is described in detail in our pre-vious publication [14]. The dimensions of the samples were:d = 6 mm and h = 4 mm.

2.3. Fabrication of TEP based on HBDCs and PUR scaffold

2.3.1. Seeding and culture of HBDCs on polyurethane scaffolds in theSpinner Basket� bioreactor

Prior to cell seeding, PUR scaffolds were pre-wetted under vac-uum conditions in a specially designed apparatus connected to avacuum pump. Then, samples were placed in 24-well tissue culturepolystyrene (TCPS) plates, covered with culture medium and incu-bated for �24 h in the standard conditions. After that, the scaffoldswere transferred to the bioreactor, and placed in the basket (seeFig. 1). In the next step, 480 ml of fresh standard culture mediummixed with the cell suspension was added. The cell number wasadjusted to around 3.0 � 105 per scaffold. The same number of cellswas seeded statically on the bottom of a 24-well plate, and servedas the control. The bioreactor and the plate were then transferredto the incubator (37 �C in 5% CO2). At all times Spinner Basket�

was kept on the magnetic stirrer (speed �60 rpm). Both cultures,under the dynamic (cells on PUR scaffolds; C+ or TEP samples)and the static conditions (control cells on TCPS), lasted 14 days.On the 7th day the culture medium was exchanged.

2.3.2. Evaluation of cell viability/numberIn order to examine the viability/number of HBDCs cultured on

PUR scaffolds (C+ or TEP samples), on the 7th and 14th day of cul-ture, the XTT test was performed. XTT assay is based on the capac-ity of metabolically active cells to convert the reaction substrate,that is, yellow tetrazolium salt (2,3-bis(2-methoxy-4-nitro-5-sulf-ophenyl)-5-((phenylamino)carbonyl)-2H-tetrazolium hydroxide),into a product, i.e. orange formazan dye. The reaction substrate,XTT (Sigma) solution in DMEM (1 mg ml�1), was first mixed withthe PMS (Sigma) solution in PBS (1.53 mg ml–1), delivering theworking solution XTT/PMS (1 ml/5 ll). The assay was performedin the 24-well culture plates where the HBDCs cultured on exam-ined PUR scaffold were placed, and where the control cells werecultured on the bottom of the wells. The tested cell cultures were

Fig. 1. Internal schema and fluid flow of the Spinner Basket� bioreactor. Modifiedfrom the device operating manual (http://www.nbsc.com/product_list.aspx?cat_id=10).

2486 P. Wozniak et al. / Acta Biomaterialia 6 (2010) 2484–2493

washed twice with PBS, and then XTT/PMS working solution wasadded. The incubation lasted 4 h in 37 �C. To allow an effectivepenetration of the XTT/PMS solution within the pores of the scaf-fold as well as effective transport of the reaction product, everyhour of the incubation, plates were shaken in a horizontal shakerfor 15 min. The final product of the reaction was measured bythe ELISA reader (FLUOstar OPTIMA – BMG LABTECH) at 450 nm.The results for PUR samples were compared to those for cells cul-tured on TCPS and are presented as the absorbance level.

The results of the XTT assay are presented as means ± standarddeviation from four absorbance readings. To verify the differencesin the cell number between the examined time points, a nonpara-metric Mann–Whitney test was used. The differences were consid-ered statistically significant at p < 0.05. The analysis was performedusing GraphPad Prism software.

2.3.3. Evaluation of cell distribution within the PUR scaffoldsCells presence in the scaffold pores was observed in a scanning

electron microscope (HITACHI S-2600N) after 7 and 14 days of dy-namic culture.

HBDCs distribution within the 3D PUR scaffolds was verified byan observation in a fluorescence microscope at UV light (NIKONECLIPSE TE 2000-U microscope, excitation: 355 nm; emission:460 nm). Cell nuclei were visualized using DNA-specific Hoe-chst33342 dye (Sigma) at a concentration of 1 mg ml�1 in 96% eth-anol. The nuclei localization was registered in the upper andbottom part of the scaffold as well as on the surface of the cross-section, i.e. on the surface uncovered by the cutting of the speci-men parallel to the base at the level of half of the specimen hight.At each time point, 2–4 samples were examined.

2.3.4. Evaluation of cell phenotypeIn order to confirm the osteoblastic phenotype of HBDCs pres-

ent in the obtained TEP, the osteogenic markers gene expressionwas investigated by means of reverse transcription-polymerasechain reaction (RT-PCR). The analysis was performed for cells ob-tained from each donor separately. The material, isolated fromthe TEP samples, cultured in the bioreactor for 14 days (TEP, day‘‘14”), was tested, as well as a reference one, i.e. mRNA, isolatedfrom HBDCs cultured on TCPS (control, day ‘‘14”). Additionally,the markers’ expression was evaluated in the material isolated

from the sub-population of the HBDCs on the day of cell seeding(control, day ‘‘0”).

The total mRNA was isolated using Rneasy Micro Kit (Qiagen).The gene expression for the alkaline phosphatase (ALP), and thecollagen type I (COLL I), as well as for the osteocalcine (OC) wasinvestigated – Titan One Tube RT-PCR Kit (Roche) was used. Theproducer protocol was followed (for details see www.roche-ap-plied-science.com/pack-insert/1939823a.pdf).

The sequence of the oligonucleotydes used in the RT-PCR wasdetermined based on the literature review [19] (see Table 1). Theoligocucleotydes for COLL I and OC genes were synthetized at theDNA Sequencing and Synthesis Laboratory of the Biochemistryand Biophysics Institute, Polish Academy of Science, while the oli-gonucleotydes for ALP and GAPDH genes were provided by OperonTechnologies Inc.

RT-PCR, composed of the reverse transcription (RT) reaction andthe polymerase chain reaction (PCR), was performed using a one-tube technique in the Bio-Rad thermocycler. The products of thePCR were cooled and stored at 4 �C until used in electrophoresis.

2.4. In vivo examination of TEP

In the in vivo study, TEP samples based on the PUR scaffoldsseeded with HBDCs (marked as C+ samples = seeded with cells)were subcutaneously implanted into the experimental animals.The same PUR scaffolds – pre-wetted for 24 h prior to implantationin a complete, i.e. serum containing culture medium – but notseeded with HBDCs, were used as the control (marked as C� sam-ples = not seeded with cells).

2.4.1. ImplantationDue to the implantation of the xenogenic material (HBDCs),

immunodeficient SCID mice were used. The SCID mice were bredand maintained under defined flora conditions at the Departmentof Clinical Cytology. The breeding of mice was started fromCB17-Icr Hnd Hsd SCID animals obtained from Harlan Farms, UK.The animals were controlled systematically for the absence of Band T lymphocytes. They were exposed to cyclophosphamide, sub-cutaneously at 3 mg per animal, over two consecutive days beforethe experiment. This procedure is necessary for the depletion of NKcell activity in animals before human cell implantation.

Ten- to twelve-week-old female animals were used in thestudy. The surgical procedure was conducted under general anes-thesia. Nembutal (pentobarbital sodium) was administered intra-peritoneally (20 lg g�1). Two symmetrical 5 mm skin dorsal cutswere performed. The line of incision proceeded longitudinally,3 mm away from the mouse spine, starting just under the lowerlimb. Afterwards, two pouches were prepared subcutaneously un-der the dorsal muscle layer by blunt dissection. Then, scaffoldswere inserted into each pocket. Two returned sutures were appliedper pocket. The stitches were taken out after 1 week.

The experiment was performed twice. Each time, 22 TEP (C+)samples and 22 control (C–) samples were used. Each mouse wasimplanted with one C+ and one C� sample. Four and 13 weeksafter the implantation, the animals were sacrificed and both typesof the grafts (C+ and C�) were removed and proceeded for histo-logical evaluation (frozen samples) or mechanical propertiesexamination (paraformaldehyde fixed samples).

2.4.2. Evaluation of the tissue-explants quality2.4.2.1. Histological evaluation of the explants. Directly after theywere removed from animal tissues, the samples designated for his-tology (6 samples of each type) were frozen using Tissue FreezingMedium (OCT, Jung), and then stored at �80 �C until they wereused for histological section preparation.

Table 1The sequence of the oligonucleotydes used in the RT-PCR.

Primers for Oligonucleotides sequence Primers annealing temperature (�C) PCR product size (bp) Reference

GAPDH 50-TCAAGGAAGCTACGGGCA- 60.3 250 Operon Technologies Inc.50-TGGCAGAAATTACACACACACAC-

ALP 50-ATGTGGACTACCTATTGGGTCTC- 56.0 354 Operon Technologies Inc.50-GGGCCAGACCAAAGATAGAG-

Coll 1 50-AGGGCTCCAACGAGATCGAGATCCG- 62.9 223 [19]50-TACAGGAAGCAGACAGGGCCAACGTCG-

OC 50-TGAGAGCTCTCACACTCCTCGCCCTATTGG- 62.9 168 [19]50-GCTCCCAGCCATCGATACAGGTAGCGC-

P. Wozniak et al. / Acta Biomaterialia 6 (2010) 2484–2493 2487

The sections (7 lm thickness) were prepared from frozen sec-tions without embedding and were stained using histologicalmethods (Hematoxylin and Eosin staining (HE)), as well as immu-nohistochemical techniques (human specific anti-collagen type Iand human specific anti-osteopontine antibodies conjugated withFITC – Abcam). To confirm cell positions in the sections, cell nucleiwere stained with DAPI fluorescence dye (Santa CruzBiotechnology).

2.4.2.2. SEM observation of the explants’ sections. Three explants ofeach type, i.e. C+ and C� after both observation periods wereembedded in Paraplast� X-tra (BioChemika, Fluka) under vacuum,and they were cut by a microtome (Leica RM2165) at �10 �C. In or-der to avoid artefacts resulting from crumbling of the thin sections,cross-sections (of the middle part of each sample) were observedinstead. The SEM images of the osmium-coated surfaces were ob-tained in a back-scattered electron-imaging mode (BSE) in the TM-1000 Hitachi microscope.

2.4.2.3. Evaluation of the mechanical properties of the explants. Dir-ectly after they were removed from tissues, the samples desig-nated for the mechanical trial (8 samples of each type) werefixed using of 4% paraformaldehyde solution. After 24 h incubationat 4 �C, samples were washed for 2 h under tap water, and thenstored at 4 �C in 70% ethanol solution, until they were used inthe compression test.

The mechanical properties of the explants were examined andcompared with the reference samples – PUR scaffolds not seededwith HBDCs and not implanted into the SCID mice. The referencesamples were sterilized, fixed and stored exactly in the same manner

Fig. 2. The results of the XTT assays performed after 7 and 14 days of culture on the TEP (level: means ± standard deviation (SD) (n = 4). Asterisks show a significant difference be

as the explanted ones – in order to eliminate the influence of the ster-ilization, storage and fixation procedure on the results. Ultimately,the mechanical strength of five experimental groups was evaluated:R (reference PUR scaffolds); C+ 4 weeks (TEP, 4-week implantationperiod); C� 4 weeks (control scaffolds, 4-week implantation peri-od); C+ 13 weeks (TEP, 13-week implantation period); C� 13 weeks(control scaffolds, 13-week implantation period).

All samples were hydrated and then evaluated in the staticcompression test (MTS Qtest™/10 device; 0–1000 N sensor; sensorv = 0.7 mm min�1).

To compare the compression strength between the experimen-tal groups, the stress level measured at a 15% deformation was ta-ken into account.

The results of the compression test are presented asmeans ± standard deviation from eight measurements. To verifythe differences of the mechanical strength between the examinedgroups, nonparametric ANOVA (Kruskal–Willis test with Dunn’spost hoc test) and nonparametric Mann–Whitney tests were per-formed. The differences were considered statistically significantat p < 0.05. The analysis was performed using GraphPad Prismsoftware.

3. Results

3.1. In vitro fabrication of TEP

3.1.1. Evaluation of cell viability/numberThe influence of dynamic seeding and culture conditions on the

viability/number of HBDCs cultured on the PUR scaffolds was eval-uated on the basis of the results of the XTT assay. The test was per-

C+) and control (cells on TCPS) samples. The results are presented as the absorbancetween values for experimental groups at *p < 0.05 (Mann–Whitney test).

Fig. 3. Fluorescence microscopy pictures showing the HBDCs nuclei on the PURscaffolds after the 14-day culture. Cells are visualized on the upper (A) and bottom(B) surface of the scaffolds, as well as on the middle cross-section (C). Bar represents500 lm.

2488 P. Wozniak et al. / Acta Biomaterialia 6 (2010) 2484–2493

formed on the 7th and 14th day of culture. The results obtained forcells cultured on PUR scaffolds in the bioreactor were compared tothe control results obtained in a routine culture under static condi-tions (Fig. 2). Statistical analysis showed that in the case of bothexperimental groups, there was a significant increase of the XTTabsorbance level for the 14-day-cultured samples by about 70%as compared to the 7-day-cultured samples (nonparametricMann–Whitney test; p < 0.05). It was also found that at both thetime points, the XTT absorbance level for TEP (C+) samples waslower than for the control cells by about 35–40%.

3.1.2. Evaluation of cell distribution within the PUR scaffoldsThe cell distribution within the 3D PUR scaffolds was evaluated

by means of fluorescence microscopy after 7 and 14 days of cultureand was found to be uniform – numerous cells were found on boththe upper and bottom surface of the scaffolds, as well as on themiddle cross-section (Fig. 3). The number of the detectable nucleiincreased for the 14-day-cultured samples, as compared to the 7-day-cultured samples (data not shown). The scaffold pores werefound to be filled with HBDCs and the extracellular matrix after7 days of observation and – more densely – after 14 days, whichis shown in comparison to the unseeded scaffold in SEM observa-tion (Fig. 4).

3.1.3. Evaluation of cell phenotypeThe HBDC phenotype was evaluated on the basis of three osteo-

blastic markers’ gene expression, i.e.: ALP, COLL I and OC. RT-PCRproducts were visualized on the agarose gel after the electrophore-sis (Fig. 5). Results obtained for the material originating from bothdonors (analysed separately) were similar (results for one of thetwo are shown in Fig. 5).

In the case of HBDCs cultured on PUR scaffolds in the bioreactor(TEP day ‘‘14”), as well as for the control cells (Control day ‘‘0” andControl day ‘‘14”), the expression of all three markers wasconfirmed.

3.1.4. In vivo examination of TEPAll animals survived the implantation procedure, and both after

4 weeks and 13 weeks the implants were found to be well incorpo-rated in host tissues. Neither macroscopic signs of inflammationnor implants resorption were observed.

3.1.4.1. Histological evaluation of the explants. Histological picturesof the explants harvested after 4 weeks are not shown due to theirunsatisfactory quality, which resulted from crumbling of the sam-ples during the cutting. Such behavior of unembedded samples ischaracteristic for materials of high porosity, which indicates thepoor integration of the samples with the tissues.

In the case of the explants harvested after 13 weeks, differencesbetween ‘‘C+” and ‘‘C�” samples were found (HE staining, Fig. 6). InC+ samples there were numerous cells within the openings of thescaffold (PUR) and around the scaffold. The cells were fusiform(fibroblasts), oval and round in shape. Some fusiform cells resem-bled fibrocytes. Occasionally, multinucleated giant cells were visi-ble. In C� samples the number of cells seemed to be less numerous,and the pores’ infiltration by tissue was much poorer as comparedto C+, which probably resulted from the tissue falling out duringsectioning. It happened with both C + and C� specimens, but wasmore common in the C� group, thus indicating a poorer cohesionof tissue with the material.

Independently from the group, there were no adverse effectslike necrosis or inflammation (e.g. neutrophils, lymphocytes wereabsent) after the 13-week implantation period.

Immunohistochemical observations revealed the presence ofproteins of human origin, i.e. human collagen type I and humanosteopontin, in the ‘‘C+” explants harvested from mice after the13-week observation in vivo (Fig. 7), while the staining was nega-tive in the ‘‘C�” group.

3.1.4.2. SEM observation of the explants’ sections. It was found thatboth the scaffolds preseeded and non-preseeded with HBDCs wereentirely filled with tissue 4 weeks, as well as 13 weeks, afterimplantation (example shown in Fig. 8). At the majority of ana-lysed surfaces empty pores were not found at all. Singular emptypores were met sporadically on one-quarter of the investigatedsurfaces.

Fig. 4. SEM picture of the investigated PUR2PCL530 scaffold in vitro: in its initial form (A), 7 days after cell seeding (B), and 14 days after cell seeding. In the lower part of thefigure (C1, C2, C3) cells with extracellular matrix organized on the surface of scaffolds, 14 days after cell seeding (shown in (C) at a magnification of 40�), are shown underhigher magnification (600� – i.e. 15� higher compared to the upper pictures).

Fig. 5. RT-PCR products of the osteoblastic markers (ALP, COLL I, OC) for threeexperimental groups: TEP day ‘‘14” – cells cultured on PUR scaffold for 14 days inthe bioreactor; control day ‘‘14” – control cells cultured for 14 days on TCPS; controlday ‘‘0” – control cells on the day of cell seeding.

Fig. 6. Histological sections of the 13-week ‘‘C�” (A) and ‘‘C+” (B) explants – HEstaining. Multinucleated giant cells (shown under higher magnification at the upperinset) and fibrocytes (shown under higher magnification at the lower inset) arevisible.

P. Wozniak et al. / Acta Biomaterialia 6 (2010) 2484–2493 2489

3.1.4.3. Evaluation of the mechanical properties of the explants. Theresults of the compression test of the examined samples are pre-sented in Fig. 9. The results for five experimental groups were ana-lysed: R (reference PUR scaffolds); C+ 4 weeks (TEP, 4-weekimplantation period); C� 4 weeks (control scaffolds, 4-weekimplantation period); C+ 13 weeks (TEP, 13-week implantationperiod); C� 13 weeks (control scaffolds, 13-week implantationperiod). On the basis of a statistical analysis, significant differencesbetween the mechanical strength of the examined samples werefound. The highest results were obtained for the ‘‘C+ 13 weeks”group. There were no significant differences between the strengthof the ‘‘R”, ‘‘C+ 4 weeks” and the ‘‘C� 4 weeks” samples.

4. Discussion

In the presented study, the PUR-based TEP for bone tissuereplacement and regeneration has been manufactured. Since itwas our goal to obtain the product which might be potentially ap-plied in human clinics, the TEP was prepared in a manner whichmight be eventually recommended for practical use. For that rea-son human cells were used, although working with animal cellsfacilitates TEP examination in vivo, particularly by opening thepossibility of intrabone implantation to animal tissues. However,

the animal-cells-based TEP does not appropriately reflect the prop-erties of the human-cells-based one. This is because of different

Fig. 7. Human specific proteins’ antibodies (conjugated with FITC) staining of histological sections of the 13-week ‘‘C+” explants. Pictures on the left (A–D) relate to thehuman specific anti-osteopontine antibodies staining, and present the same area in various projections, to show where the antibody specific staining is located. Pictures onthe right (E–H) relate to the human specific anti-collagen type I antibodies staining, and are shown in the sequence analogous to the one on the left of the figure. (A) and (E)are light inverted microscope pictures – scaffold is labelled with ‘‘s”, tissue with ‘‘t”; (B) and (F) are fluorescence pictures in the wavelength specific for FITC ((B) – humanspecific anti-osteopontine antibodies, (F) – human anti-collagen type I specific antibodies); (C) and (G) are fluorescence pictures in UV light (DAPI nuclei staining), (D) and (H)present all the above pictures ((B–D) and (F–H), respectively) merged to facilitate the co-localization. Bar: 100 lm.

2490 P. Wozniak et al. / Acta Biomaterialia 6 (2010) 2484–2493

metabolism; proliferation rate and cell size of osteoblasts of animalvs. those of human origin. These in turn affect the cell amount to beused, the duration of the TEP preparation in vitro, and the cell dis-tribution within the scaffold (which may depend on the relationbetween cell size and the dimension of the scaffold’s pores andinterconnections). The price to be paid when human cells are usedin experiments is the necessity of using immunodeficient animalsfor experimental implantation. SCID mice were used for the in vivoobservation of the obtained TEP, as this is a commonly used systemfor investigation of human-cell-based TEPs [20–26].

The characteristics of the obtained TEP was performed in vitroprior to implantation. Combining the XTT assay results, microscopy

observations, and the evaluation of the expression of genes forosteoblastic markers, it seems that dynamic seeding and cultureconditions enable a homogeneous distribution, maintaining theproliferative potential and osteogenic phenotype of the HBDCs cul-tured on three-dimensional polyurethane scaffolds. In particular, itwas found that the proliferation rate of HBDCs cultured on the PURscaffolds in the bioreactor was comparable to the control cells cul-tured on TCPS (see Fig. 2). The lower value of absorbance in the XTTassay on PUR scaffolds, as compared to the control, might be re-lated to the lower cell seeding efficiency in the bioreactor. This isbecause cells are put into the bioreactor as a suspension in500 ml of medium, and they are expected to seed at the surface

Fig. 8. SEM picture of the uncovered surface from the middle part of the ‘‘C+”explant harvested from a mouse after 13 weeks – representative for all investigatedsamples, independently from the observation period and the preseeding withHBDCs. The scaffold (indicated by arrows) is entirely filled with tissue.

P. Wozniak et al. / Acta Biomaterialia 6 (2010) 2484–2493 2491

of the relatively small samples located in the upper part of the bio-reactor, whereas under static conditions, the same number of cellsper sample (but in 200 ll of culture medium) is seeded on the sur-face of the samples which fill the whole wells of the plate. Obvi-ously, culture conditions on the bottom of a culture dish and onthe surface of the pores within a 3D scaffold are different, butthe comparable proliferation rate in both systems confirms thatthe investigated scaffold provides an appropriate environmentfor HBDCs in a dynamic culture in vitro. The growing cell density,accompanied by the presence of extracellular matrix, was also con-firmed quantitatively by means of SEM observations of the scaf-folds (see Fig. 4). Not only cell survival, but also cell distributionin the scaffold, is a key factor in the characteristics of TEP obtainedin vitro. As was shown in the fluorescence microscopic examina-tion, HBDC distribution in the examined scaffolds was superb.

Fig. 9. The mechanical strength of the explants and the reference samples (R) measured istrain level measured at a 15% deformation of the sample, as a mean ± standard deviationon the graph. Asterisks show a significant difference between the values for the experim

Numerous cells were observed not only at the upper and the bot-tom side of the samples, but also at the entire surface of the cross-section acquired from the central part of the sample (see Fig. 3).This must be partly due to the dynamic culture system, whichwe decided to apply based on the observations from our previousstudy, where a significant benefit of applying such a system overthe standard static culture was observed [27]. Improvement of celldistribution due to the application of a bioreactor is a known phe-nomenon [28–33]. Our work documents that the applied SpinnerBasket� is good enough for a uniform distribution and a cultureof HBDCs in the investigated scaffolds. Obviously, cell distributionand extracellular matrix production also depend on the scaffoldarchitecture, which is, additionally, of critical importance for thetissue ingrowth after implantation in vivo.

The fate of the produced PUR-based TEP after implantationin vivo was investigated by means of subcutaneous implantationto immunodeficient SCID mice. It was found that both scaffoldsseeded with cells prior to implantation, and scaffolds which wereimplanted without cells, were infiltrated by the host tissue aftersubcutaneous implantation to SCID mice, which was confirmedby SEM observations of explants harvested both after 4 weeksand after 13 weeks (as an example, see Fig. 8). It confirms theappropriate pore size and interconnections in the scaffold architec-ture. However, tissue integration with the material was not equallygood. The decohesion resulting from the cutting of the frozen ex-plants harvested 13 weeks after implantation was mostly visibleon sections obtained from the samples which were not seededwith human cells prior to implantation – it is shown in Fig. 6. Inthe case of implants containing HBDCs, the immunoexpression ofthe human collagen type I and human osteopontin was revealed13 weeks after implantation. The presence of proteins of humanorigin within such implants harvested from mice tissue confirmsthat HBDCs implanted into SCID mice, as a component of TEP, sur-vive and keep their ability to synthesize and secrete the specificcomponents of the human bone extracellular matrix. This corre-sponds with the results of the mechanical test. The strength ofthe 4-week explants (both C+ and C�) was comparable to thoseof the reference samples (R – not seeded with HBDCs and not im-

n the compression test. The results for the experimental groups are presented as the(SD) (n = 8). Only significant differences between the examined groups are marked

ental groups at: *p < 0.05 and **p < 0.01 (Kruskal–Wallis and Mann–Whitney tests).

2492 P. Wozniak et al. / Acta Biomaterialia 6 (2010) 2484–2493

planted into animals), which indicates that the tissue organizedwithin the implants during this period must have been immatureand not able to bear the load. On the contrary, the mechanicalstrength of the 13-week explants was significantly higher com-pared to the reference samples, as well as to the 4-week explants.Moreover, the strength of the C+ explants was significantly higherthan for the C� samples (see Fig. 9), although the mineralized bonetrabeculas were not found in the investigated material, probablydue to the insufficient calcium ion availability in the subcutaneoustissue. This indicates, however, that the C+ explants must consistnot only of mouse tissue infiltrating the scaffold, but also of the tis-sue produced by the transplanted human cells. In view of the po-tential application of PUR-based bone TEPs containing humanosteogenic cells, this is a promising result. As mentioned in theintroduction, host tissue tolerance towards degradable PURs wasshown in various animal models [7,9,34–38]. Also in our experi-ments, no adverse cell or tissue reaction against the investigatedPUR2PCL530 material was observed both in vitro and in vivo. Sinceno signs of material loss were noticed after the 13-week observa-tion in mice, we cannot judge about the reaction to the degradationproducts in vivo. However, the degradability of the investigatedPUR was studied in vitro in our previous study [39]. The expositionof the investigated PUR to a simulated body fluid (SBF) resulted inthe reduction of molecular weight around 52% after 180 days. Thedegradation products – added to the culture of MG-63 cells in theform of concentrated extracts, obtained by placing the PUR sam-ples in separate cell culture medium – did not induce a cytotoxicreaction [39]. The degradation of PURs is in general relatively slow.However, this is not necessarily a disadvantage. First of all, there isno risk of unexpected, rapid degradation, which may occur due tothe so-called ‘‘skin effect” in the case of PLLA or PGLA, and sec-ondly, it seems to respond well to the real needs in many clinicalapplications. Regarding two possible strategies described byHutmacher [40] – i.e. TEPs with advanced scaffold resorption priorto implantation, or TEPs with scaffolds retaining the physical prop-erties of the scaffold for at least 6 months – it seems that slowlydegradable PURs, based on non-cytotoxic aliphatic diisocyanatehard segments, and PCL soft segments, perfectly match therequirements of the second approach.

Altogether, the investigated PUR2PCL530 scaffold can be rec-ommended as candidate for bone tissue engineering. The abilitiesof HBDCs to keep their osteogenic phenotype as verified here byRT-PCR (Fig. 5) are not of least value. Only a few observations ofhuman osteogenic cells in polyurethanes scaffolds have been re-ported until now. Fassina et al. reported the fabrication of hybridstructures composed of 3D PUR scaffolds and human osteogeniccells under physical stimulation – yet, they did not use cells fromprimary culture, but from the osteosarcoma cell line (SAOS-2)[41,42]. Only recently, Zanetta showed the abilities of two typesof PUR foams to support cell proliferation and the differentiationof MSCs from human bone marrow into osteoblasts in vitro [11].Since the observed cells were reduced nearer to the core of thescaffold in his system, he postulated the application of a bioreac-tor for further studies. Indeed, as shown in our experiments, thereis a benefit from using dynamic conditions for human osteogeniccell seeding and culture in PUR scaffolds, as discussed above.Apart from the evidence of the HBDC tolerance towards theinvestigated PUR2PCL530 scaffold, and the efficiency of the ap-plied dynamic cell culture system in obtaining a hybrid implant,to the best of our knowledge, the present work is the firstdescription of the fabrication and experimental implantation ofa bone-tissue-engineered product based on human primaryosteoblastic cells and biodegradable PUR scaffolds. The results ob-tained provide evidence of the added value brought by the pres-ence of the osteogenic cells in the PUR scaffold over the implantswithout cells.

5. Conclusions

Dynamic cell seeding and cell culture conditions enable homo-geneous distribution, maintaining the proliferative potential andosteogenic phenotype of human-bone-derived cells cultured onthe selected three-dimensional polyurethane scaffold.

Human-bone-derived cells implanted into the experimental ani-mals as a component of a polyurethane-based tissue-engineeredproduct survive and keep their ability to synthesize and secretethe specific components of the bone tissue extracellular matrix.

The quality of the polyurethane-based implants was found to bebetter when seeded with human-bone-derived cells – which con-firms the potential usefulness of PUR-based implants with HBDCsfor bone tissue engineering.

Acknowledgments

The authors would like to thank Tomasz Brynk for performingthe compression tests.

The authors also appreciate the technical help of Anna Pod-bielska related to histology examination.

This scientific work was financially supported by the Ministry ofScience and Higher Education, Grant R13 01901, and by the Medi-cal University of Warsaw (NZME/W2/08).

Appendix A. Figures with essential colour discrimination

Certain figures in this article, particularly Figs. 1, 3 and 5–7, aredifficult to interpret in black and white. The full colour images canbe found in the on-line version, at doi: 10.1016/j.actbio.2009.10.022).

References

[1] Guelcher SA. Polyurethanes. In: Hollinger JO, editor. An introduction tobiomaterials, vol. 11. New York: CRC Press; 2005. 576 pp..

[2] Zdrahala RJ, Zdrahala IJ. Biomedical applications of polyurethanes: a review ofpast promises, present realities, and a vibrant future. J Biomater Appl1999;14(1):67–90.

[3] Lamba NMK, Woodhouse KA, Cooper SL. Polyurethanes in biomedicalapplications. Washinghton, DC: CRC Press; 1998. 288 pp..

[4] Zdrahala RJ. Small caliber vascular grafts. Part ii: polyurethanes revisited. JBiomater Appl 1996;11(1):37–61.

[5] Santerre JP, Woodhouse K, Laroche G, Labow RS. Understanding thebiodegradation of polyurethanes: from classical implants to tissueengineering materials. Biomaterials 2005;26(35):7457–70.

[6] Gorna K, Gogolewski S. Biodegradable porous polyurethane scaffolds for tissuerepair and regeneration. J Biomed Mater Res A 2006;79(1):128–38.

[7] Gogolewski S, Gorna K. Biodegradable polyurethane cancellous bone graftsubstitutes in the treatment of iliac crest defects. J Biomed Mater Res A2007;80(1):94–101.

[8] Zhang JY, Beckman EJ, Piesco NP, Agarwal S. A new peptide-based urethanepolymer: synthesis, biodegradation, and potential to support cell growthin vitro. Biomaterials 2000;21(12):1247–58.

[9] Ganta SR et al. Vascularization and tissue infiltration of a biodegradablepolyurethane matrix. J Biomed Mater Res A 2003;64(2):242–8.

[10] Kavlock KD, Pechar TW, Hollinger JO, Guelcher SA, Goldstein AS. Synthesis andcharacterization of segmented poly(esterurethane urea) elastomers for bonetissue engineering. Acta Biomater 2007;3(4):475–84.

[11] Zanetta M, Quirici N, Demarosi F, Tanzi MC, Rimondini L, Fare S. Ability ofpolyurethane foams to support cell proliferation and the differentiation ofmscs into osteoblasts. Acta Biomater 2009;5(4):1126–36.

[12] Gallagher JA, Gundle R, Beresford JN. Isolation and culture of bone-forming cells(osteoblasts) from humane bone. In: Jones GE, editor. Methods in molecularmedicine: human cell culture protocols. Totowa, New Jersey: Humana Press; 1996.

[13] Kudelska-Mazur D, Lewandowska-Szumiel M, Mazur M, Komender J.Osteogenic cell contact with biomaterials influences phenotype expression.Cell Tissue Bank 2005;6(1):55–64.

[14] Bil M, Ryszkowska J, Kurzydlowski KJ. Effect of polyurethane composition andthe fabrication process on scaffold properties. J Mater Sci 2009;44:1469–76.

[15] Bil M, Ryszkowska J, Wozniak P, Kurzydłowski KJ, Lewandowska-Szumieł M.Optimization of the structure of polyurethanes for bone tissue engineeringapplications. Acta Biomater 2010;6:2501–10.

[16] Wozniak P, Bil M, Chroscicka A, Olkowski R, Ryszkowska J, Lewandowska-Szumiel M. Biocompatibility of polyurethanes in a human bone derived cells

P. Wozniak et al. / Acta Biomaterialia 6 (2010) 2484–2493 2493

culture in vitro. In: 21st European conference on biomaterials 2007, Brighton,UK: oral presentation.

[17] Ryszkowska J, Bil M, Wozniak P, Lewandowska-Szumiel M, Kurzydlowski KJ.Influence of catalyst type on biocompatibility of polyurethanes. Mater SciForum 2006;514–516:887–91.

[18] Przybytniak G, Kornacka E, Ryszkowska J, Bil M, Rafalski A, Wozniak P, et al.Influence of radiation sterilization on poly(ester urethanes) designed formedical applications. Nukleonika 2006;51(Suppl. 1):121–8.

[19] Carmeliet G, Nys G, Bouillon R. Microgravity reduces the differentiation ofhuman osteoblastic mg-63 cells. J Bone Miner Res 1997;12(5):786–94.

[20] De Kok IJ, Hicok KC, Padilla RJ, Young RG, Cooper LF. Effect of vitamin dpretreatment of human mesenchymal stem cells on ectopic bone formation. JOral Implantol 2006;32(3):103–9.

[21] Kasten P et al. Influence of platelet-rich plasma on osteogenic differentiation ofmesenchymal stem cells and ectopic bone formation in calcium phosphateceramics. Cells Tissues Organs 2006;183(2):68–79.

[22] Kasten P, Beyen I, Niemeyer P, Luginbuhl R, Bohner M, Richter W. Porosity andpore size of beta-tricalcium phosphate scaffold can influence proteinproduction and osteogenic differentiation of human mesenchymal stemcells: an in vitro and in vivo study. Acta Biomater 2008;4(6):1904–15.

[23] Morgan SM, Tilley S, Perera S, Ellis MJ, Kanczler J, Chaudhuri JB, Oreffo RO.Expansion of human bone marrow stromal cells on poly-(DL-lactide-co-glycolide) (pdl lga) hollow fibres designed for use in skeletal tissueengineering. Biomaterials 2007;28(35):5332–43.

[24] Steffens L, Wenger A, Stark GB, Finkenzeller G. In vivo engineering of a humanvasculature for bone tissue engineering applications. J Cell Mol Med 2008.

[25] Thorrez L et al. Growth, differentiation, transplantation and survival of humanskeletal myofibers on biodegradable scaffolds. Biomaterials 2008;29(1):75–84.

[26] Vats A et al. Chondrogenic differentiation of human embryonic stem cells: theeffect of the micro-environment. Tissue Eng 2006;12(6):1687–97.

[27] Wozniak P, Chroscicka A, Olkowski R, Lewandowska-Szumiel M. Cell seedingtechnique – the key factor affecting 3-d cell culture in vitro. Eng Biomater2005;47-53:190–2.

[28] Martin I, Wendt D, Heberer M. The role of bioreactors in tissue engineering.Trends Biotechnol 2004;22(2):80–6.

[29] Sikavitsas VI, Bancroft GN, Mikos AG. Formation of three-dimensional cell/polymer constructs for bone tissue engineering in a spinner flask and arotating wall vessel bioreactor. J Biomed Mater Res 2002;62(1):136–48.

[30] Freed LE, Vunjak-Novakovic G. Tissue engineering bioreactors. In: Vacanti JP,Vacanti JP, editors. Principles of tissue engineering. London: Academic Press;2000. p. 143–56.

[31] Qi X, Liu JG, Chang Y, Xu XX. Comparative study on seeding methods of humanbone marrow stromal cells in bone tissue engineering. Chin Med J (Engl)2004;117(4):576–80.

[32] Janssen FW, Oostra J, Oorschot A, van Blitterswijk CA. A perfusion bioreactorsystem capable of producing clinically relevant volumes of tissue-engineeredbone: in vivo bone formation showing proof of concept. Biomaterials2006;27(3):315–23.

[33] Stiehler M, Bunger C, Baatrup A, Lind M, Kassem M, Mygind T. Effect ofdynamic 3-d culture on proliferation, distribution, and osteogenicdifferentiation of human mesenchymal stem cells. J Biomed Mater Res A 2008.

[34] Klompmaker J, Jansen HW, Veth RP, de Groot JH, Nijenhuis AJ, Pennings AJ.Porous polymer implant for repair of meniscal lesions: a preliminary study indogs. Biomaterials 1991;12(9):810–6.

[35] Klompmaker J, Jansen HW, Veth RP, Nielsen HK, de Groot JH, Pennings AJ, et al.Meniscal repair by fibrocartilage? An experimental study in the dog. J OrthopRes 1992;10(3):359–70.

[36] de Groot JH, de Vrijer R, Pennings AJ, Klompmaker J, Veth RP, Jansen HW. Useof porous polyurethanes for meniscal reconstruction and meniscal prostheses.Biomaterials 1996;17(2):163–73.

[37] Farso Nielsen F, Karring T, Gogolewski S. Biodegradable guide for boneregeneration. Polyurethane membranes tested in rabbit radius defects. ActaOrthop Scand 1992;63(1):66–9.

[38] Gogolewski S, Gorna K, Turner AS. Regeneration of bicortical defects in the iliaccrest of estrogen-deficient sheep, using new biodegradable polyurethane bonegraft substitutes. J Biomed Mater Res A 2006;77(4):802–10.

[39] Bil M. Polyurethane support for bone tissue culture in vitro. PhD Thesis,Warsaw University of Technology; 2009.

[40] Hutmacher DW. Scaffolds in tissue engineering bone and cartilage.Biomaterials 2000;21(24):2529–43.

[41] Fassina L, Visai L, Asti L, Benazzo F, Speziale P, Tanzi MC, et al. Calcified matrixproduction by saos-2 cells inside a polyurethane porous scaffold, using aperfusion bioreactor. Tissue Eng 2005;11(5-6):685–700.

[42] Fassina L et al. Effects of electromagnetic stimulation on calcified matrixproduction by saos-2 cells over a polyurethane porous scaffold. Tissue Eng2006;12(7):1985–99.