Formation of cartilage matrix proteins by BMP-transfected murine mesenchymal stem cells encapsulated in a novel class of alginates

Biomaterials 23 (2002) 2003–2013

Formation of cartilage matrix proteins by BMP-transfected murine mesenchymal stem cells encapsulated in a novel class of alginates a . . b, M. Webera, A. Steinertb, A. Jorka, A. Dimmlerc, F. Thurmer , N. Schutze C. Hendrichb, U. Zimmermanna,* Department of Biotechnology, University of Wurzburg, Am Hubland, Biozentrum, D-97074 Wurzburg, Germany . . Department of Orthopedic Surgery, University of Wurzburg, Konig-Ludwig-Haus, D-97074 Wurzburg, Germany . . . c Department of Pathology and Anatomy, University of Erlangen, Krankenhausstr. 8-10, D-91054 Erlangen, Germany a

b

Accepted 25 September 2001

Abstract Proliferation and differentiation of wild-type, BMP-2 and BMP-4 transfected cells of C3H10T1/2, a mouse mesenchymal stem cell line that can differentiate into chondrocytes, were studied under monolayer (2D-) and encapsulation (3D-) conditions. Cells were encapsulated in a novel class of alginate. The alginate was of clinical grade (CG) because of complete removal of mitogenic and cytotoxic contaminants by chemical means. Compared to commercial alginates used so far for encapsulation it was characterized by ultra-high viscosity (UHV; viscosity of a 0.1% w/v solution of about 20 cP). In contrast to monolayer cultures, proliferation of cells was prevented when the cells were encapsulated in UHV/CG alginate at the same suspension density. As revealed by immunohistochemistry and quantitative RT-PCR, transfected and wild-type monolayer cells showed synthesis of type I collagen after transfer into differentiation medium, while culture in an alginate scaffold resulted in an upregulation of type II collagen and other hyaline cartilage proteins. BMP-4 transfected cells produced considerably more type II collagen than BMP-2 transfected and wild-type cells. BMP-4 transfected cells were also characterized by type I collagen production up to Day 10 and exhibited transient alkaline phosphatase activity levels that were much higher than the peak values observed for the other two cell lines. The coincidence of the ALP peak values with downregulation of type I collagen in BMP-4 transfected cells suggested that C3H10T1/2 cells differentiate into chondrocytes via a chondroprogenitor-like cell. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Cartilage; Stem cells; Chondrocytes; Alginate; Real time RT-PCR; BMP

1. Introduction Recently, many efforts have been undertaken to repair articular cartilage lesions by transplantation of autologous chondrocytes cultured and expanded under in vitro conditions. A major problem of human cells in culture is the phenomenon of dedifferentiation. Thus, after a few days in monolayer cultures, chondrocytes change their appearance to a fibroblast-like morphology and lose their biochemical and functional properties. Instead of hyaline cartilage-specific type II collagen and *Corresponding author. Tel.: +49-931-888-4508; fax: +49-931-8884509. E-mail address: [email protected] (U. Zimmermann).

other hyaline-specific glycosaminoglycans, type I collagen is mainly synthesized [1]. Matrix molecules are released into the medium and do not aggregate to an appropriate extracellular matrix. However, redifferentiation and formation of an extracellular matrix occurs when the cells are encapsulated in a threedimensional matrix. This is essential for the successful generation of tissue-engineered cartilage. Among the polymers tested so far, alginate has been and will continue to be one of the most important scaffold materials. Alginate cross-linked with Ca2+ or Ba2+ has been used successfully to encapsulate cells and to maintain their function in tissue culture [2–4]. Alginate is biodegradable. It is degraded by enzymatic pathways to its two monomeric subunits, mannuronic and guluronic acids [5–7]. Furthermore, the ability

0142-9612/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 2 - 9 6 1 2 ( 0 1 ) 0 0 3 2 9 - 5

2004

M. Weber et al. / Biomaterials 23 (2002) 2003–2013

to mold Ca2+ alginate gels to produce defined shapes creates the opportunity to fabricate patient-designed cartilage transplants. These and other characteristics are advantageous for a polymer support matrix of cell and tissue transplantation. However, when commercial alginates (cross-linked with Ca2+ or with Ba2+) were implanted into rodents or other animals, heavy foreign body reactions were observed within a few weeks (for review articles, see [2]). The reason for this is that commercial alginate contains at least 10–20 mitogenic and cytotoxic impurities [8–10]. Biocompatibility is not only an obligatory requirement for in vitro construction of transplantable vital tissue structures with cell carriers and for granting medical approval, but also for elimination of interferences of toxic contaminants with the biochemical and biophysical signals that regulate the development of the highly complex extracellular matrix [11]. Therefore, in this communication we used ultra-high viscosity alginate of clinical grade cross-linked with divalent cations as a scaffold matrix. This alginate was obtained by extraction and several purification steps from fresh algal material that removed the impurities without degradation of the polymeric guluronic/mannuronic acid chains [12]. When implanted in rodents or baboons, this product did not evoke any significant foreign body reaction [9]. Similarly, transplantation of encapsulated parathyroid allogeneic tissue segments in rodents showed a proper function (i.e., parathormon release associated with normocalcemia) over more than 1 yr (Bohrer et al., manuscript in preparation). Despite these successes, the question remains whether encapsulation of suspended cells (such as chondrocytes) survives and functions satisfactorily in a matrix made up of ultra-high viscosity alginate cross-linked with divalent cations (for a detailed discussion of the problems, see [13]). For the first approach, we employed this alginate to the encapsulation of C3H10T1/2, a mouse mesenchymal stem cell line. This permanent cell line, when cultured in a favorable environment, can undergo chondrogenesis [14,15]. According to Caplan [16], stem cells are superior for tissue engineering than chondrocytes for several reasons. Among other things, they have the decisive advantage over primary cells that high cell numbers can be produced in a reproducible manner [17]. In addition, C3H10T1/2 cell lines that stably transfected with human BMP-2 and BMP-4 vectors have also been established [14,18]. BMPs (bone morphogenetic proteins) are members of the transforming growth factor-b (TGF-b) superfamily and are involved in the development of cartilage and bone [19–22]. Thus, the study of encapsulated BMP-transfected and wild-type cells allows to elucidate the interplay between the alginate matrix and the differentiation processes of these cells in the absence of any mitogenic and cytotoxic impurities.

2. Materials and methods 2.1. Cell culture Puromycin-resistent, BMP-2 and BMP-4 transfected and expressing C3H10T1/2 cells were obtained from G. Gross (Gesellschaft fur . Biotechnologische Forschung, Braunschweig, Germany). C3H10T1/2 is a murine mesenchymal progenitor cell line permanently transfected with cDNAs encoding the human bone morphogenetic proteins BMP-2 and BMP-4 [14]. Transfected and non-transfected cells were routinely grown in 650 ml polysterene tissue culture flasks (Greiner, Nurtingen, . Germany). The culture medium consisted of Dulbecco’s modified Eagle’s medium (DMEM, Sigma, Deisendorf, Germany) supplemented with 2 mm l-glutamine, 10% (v/v) fetal calf serum (FCS, PAA Laboratories GmbH, Linz, Austria), 50 mm 2-mercaptoethanol (Sigma, Deisendorf, Germany) and 100 U/ml/100 mg/ml penicillin/ streptomycine (Biochrom, Berlin, Germany). The culture medium of the transfected cells contained additional 5 mg/ml puromycin (Sigma, Deisendorf, Germany). Cultures were incubated in a humidified incubator at 371C and 5% CO2. After 3 days, the cells were nearly confluent. Cells were harvested by treatment of the monolayer with 4 ml trypsin/EDTA (0.5 g/l trypsin, 0.2 g/l EDTA dissolved in phosphate buffered saline, PBS) for 2 min at 371C. After centrifugation at 170 g for 10 min, the pellet was washed with PBS and resuspended in 10 ml PBS. The number of cells was determined electronically by using the Casys cell analyzer (Sch.arfe, Reutlingen, Germany). The cells were centrifuged again and resuspended in culture medium at a suspension density of 1
106 cells/ml. For proliferation studies cells were placed in standard 24-well polysterene tissue culture dishes. For immunohistochemistry the cells were seeded on CELLocatess (Eppendorf, Hamburg, Germany). After 3 days culture at 371C and 5% CO2 the medium was replaced by differentiation medium. This medium consisted of a culture medium to which 10 mm b-glycerophosphate (Sigma, Deisendorf, Germany) and 50 mg/ml ascorbic acid (Sigma, Deisendorf, Germany) were added. Every 3 or 4 days, half of the differentiation medium was replaced by fresh medium. Experiments were performed over a period of 17 days. Every 2–4 days cell proliferation and differentiation stage were examined. 2.2. Alginate encapsulation Alginate was extracted from the inner stipes of the kelp Laminaria pallida and purified according to the protocol of Hillg.artner [12] (see also [23]). The ultrahigh viscosity alginate was of clinical grade (termed UHV/CG alginate; [9,10]). Before use the alginate was sterilized and then dissolved in sterile, endotoxin-free

M. Weber et al. / Biomaterials 23 (2002) 2003–2013

0.9% NaCl solution at a concentration of 0.5% w/v. For encapsulation confluent cells were resuspended in the alginate solution after trypsinization. The suspension density was adjusted to 1
106 cells/ml. Homogeneous beads were formed by using a two-channel droplet generator [24]. For cross-linking the alginate/cell mixture was dropped into a 20 mm CaCl2 solution containing 115 mm NaCl buffered at pH 7.0 with 5 mm histidine. About 14000–16000 beads of a mean diameter of 500 mm were obtained from 1 ml alginate input solution. The beads (each containing approximately 50 cells) were allowed to polymerize in this solution for 10 min before three consecutive washes with 0.9% NaCl followed. The beads were transferred into standard 10 cm petri dishes containing 25 ml culture medium and were then cultured as described above. The viability of the encapsulated cells was verified by propidium iodide (PI) staining [25]. To this end, 10 ml PI solution (2.5 mg/ml) were added to 1 ml bead suspension. The cells were viewed under a fluorescence microscope (Axiophot, Zeiss, Oberkochen, Germany). Dead cells exhibited a bright red fluorescence due to binding of PI to nucleic acids. For determination of the proliferation rate, cells were released from the capsules by treatment with 50 mm EDTA and then counted electronically (see above). 2.3. Alkaline phosphatase assay The alkaline phosphatase (ALP) activity in monolayer cultures and in beads was assayed regularly by measuring the release of p-nitrophenol from p-nitrophenylphosphate. In order to release the cells from the alginate matrix the beads were washed 3 times with PBS and were then incubated for 2 min in a 50 mm EDTAsolution (pH 7). Recovered cells and trypsinized monolayer cells, respectively, were counted in a Neubauer-hemacytometer. The cells were then centrifuged and resuspended in 500 ml p-nitrophenylphosphate solution (Sigma Fast Assay, Sigma, Deisendorf, Germany) for 1 h at 371C. The reaction was stopped by the addition of 500 ml of a 50 mm NaOH solution. 200 ml samples were analyzed by using an ELISA-reader (Thermomax microplate reader, Molecular Devices, Menlo Park, USA) at a wavelength of 405 nm. ALP activity was quantified after calibration with p-nitrophenol and given as ng/min/mg protein/cell.

2005

ELISA-reader at a wavelength of 570 nm, using bovine albumin as a standard. 2.5. Extraction and analysis of mRNA by real time RT-PCR 2.5.1. RNA-isolation Total cellular RNA was extracted from frozen cell samples (1
105–107 cells) by using the RNeasy Kits (Qiagen, Hilden, Germany). The RNA was eluated in 30 ml RNase-free water. The kit included DNase treatment, thus minimizing the contamination of the RNA by genomic DNA. 2.5.2. Reverse transcription First-strand cDNA was synthesized by Superscript II reverse transcriptase (GIBCO, Eggenstein, Germany). Four microliters of RNA-eluate was added to 0.25 mg dT15 primer (TIB-Biomol, Berlin, Germany) and 25 ng random hexamers (Promega, Heidelberg, Germany) to a final volume of 6 ml. Annealing was allowed to proceed for 10 min at 651C. 100 units of RT enzyme, 1
transcription buffer (GIBCO, Eggenstein, Germany), 10 mm DTT (GIBCO, Eggenstein, Germany) and 0.5 mm deoxynucleotide triphosphate (Roth, Karlsruhe, Germany) were added to the RNAprimer mix to form a total reaction volume of 10 ml. The reaction was allowed to proceed for 70 min at 421C, followed by 5 min at 951C to inactivate the enzyme.

2.4. Protein determination

2.5.3. Real time PCR Real time PCR (TaqMan-PCR, ABI Prism 7700 Sequence Detection System, Perkin Elmer Applied Biosystems, Foster City, USA) is well established as a fast and sensitive method for precise quantitation [27,28]. If sufficient sequence information was available, we designed intron-spanning primer-probe systems to quantify target and housekeeping gene mRNAs, i.e. the target amplicon of the amplified product fits into two joining exons of the specific cDNA-sequence corresponding to the mRNA of interest with the double fluorescence (FAM, TAMRA) marked probe binding on the exon–exon-junction site. The sequences of primers and probes for type I and type II collagen and for the housekeeping gene GAPDH as an internal calibration were as follows (in brackets final concentration in PCR reactions):

Protein was determined by the method of Bradford [26]. To this end, monolayer cells and cells released from the alginate beads were treated with 20% (v/v) NaOH dissolved in PBS at 801C for 1 h. Then 20% (v/v) HCl and the four-fold volume of the Bradford reagent (Sigma, Deisendorf, Germany) were added. The optical density of the samples was measured by using an

Murine type I collagen (mCol1): Forward primer: 50 -CCGGCTCCTGCTCCTCTTA-30 (100 nm), Reverse primer: 50 -CAGATACAGATCAAGCATACC TCGG-30 (300 nm), Probe: 50 -FAM- CCAAGAAGACATCCCTGAAGTC AGCTGCA -TAMRA-30 (100 nm).

2006

M. Weber et al. / Biomaterials 23 (2002) 2003–2013

Murine type II collagen (mCol2): Forward primer: 50 -GGGCGAGGGCAACAGC-30 (300 nm), Reverse primer: 50 -CGGTACTCGATGACGGTCTTG -30 (900 nm), Probe: 50 -FAM-CCTGAAGGATGGCTGCACGAAACACA -TAMRA-30 (100 nm) Murine GAPDH (mGAPDH): Forward primer: 50 -TGGAAGATGGTGATGGGCTT -30 (100 nm), Reverse primer: 50 -ACATGTTCCAGTATGACTCCA CTCA-30 (300 nm), Probe: 50 -FAM-TTCTCGGCCTTGACTGTGCCGT TG -TAMRA-30 (100 nm). Cleavage of the double marked sequence-specific probe by nuclease activity releases the reporter dye resulting in an emission increase of respective wavelength, which was monitored continuously. The signal is normalized to an internal reference (DRn ) and the system sets the threshold cycle Ct, when DRn becomes equal to ten standard deviations of the baseline. The Ct value makes it possible to quantify the input target number and by using an internal standard of known concentration, it is possible to calculate the amount of target cDNA in each sample. For relative quantitation as used here, we normalized the amount of target gene to the housekeeping gene GAPDH. After cDNA synthesis distilled water was added to a final volume of 200 ml. The 25 ml total PCR volume consisted of 5 ml of this cDNA mix, 0.3 units Taq polymerase, 1
amplification buffer, 1.2 mm ROX (Eurogentec, Seraing, Belgium each), 5 mm MgCl2, 200 mm deoxynucleotide triphosphate (Roth, Karlsruhe, Germany), primers (MWG-Biotech, Ebersberg, Germany) and probe (Eurogentec, Seraing, Belgium) (for concentrations see above). The PCR reaction was then performed under the following conditions: denaturation at 951C for 4 min, then 40 amplification cycles with denaturation at 951C for 15 s followed by annealing and extension at 651C for 1 min. We always used samples of distilled water and RNA-solution without reverse transcription as negative controls to avoid unspecific signals through amplification of genomic DNA. 2.6. Immunohistochemistry Beads were washed twice with PBS and embedded in Tissue Teks (Sakura, Zoeterwoude, Netherland) frozen in liquid nitrogen. 50 mm thick sections were cut, air dried overnight and washed 3 times with 0.5 mol/l Tris(hydroxymethyl–aminoethan)-solution. Monolayer cells were cultivated on CELLocates, fixed in acetone and dried for 15 min at room temperature. For the

detection of type II collagen, some sections were digested for 4 min by pepsin (Dako, Als, Denmark; 8 g/l in 0.1 n HCl). For the detection of chondroitin-4and -6-sulfate some sections were digested for 10 min by chondroitinase ABC (Sigma, Deisendorf, Germany; 5 U/ml in distilled water). After digestion, the sections were washed 3 times with Tris-solution and then treated with a 1% H2O2 solution. Each section was incubated overnight at 41C in a wet chamber filled with a solution containing specific primary antibodies against type I, type II collagen, and chondroitin-4/6-sulfate (1:50 rabbit antimouse type I collagen polyclonal antiserum; 1:100 mouse antihuman type II collagen monoclonal antibody; 1:100 mouse antichondroitin-4-sulfate monoclonal antibody; 1:100 mouse antichondroitin-6-sulfate monoclonal antibody; all from Chemicon International Inc., Temecula, USA). Sections were rinsed with Trissolution extensively. In case of type II collagen chondroitin-4/6-sulfate slides were incubated for 30 min with rabbit antimouse IgG (Dako, Als, Denmark) at room temperature. Then after further rinsing with Tris-solution goat antirabbit IgG (HRPO Conjugate, Caltag Laboratories, Burlingame, USA) was added for 30 min. In case of type I collagen staining the sections were first treated with mouse antirabbit IgG followed by goat antimouse immunoglobuline (Dako, Als, Denmark) treatment. After several washes, antibody binding was visualized by using the diaminobenzidine method (DAB kit, Sigma, Deisendorf, Germany). The reaction was stopped by immersing the sections in water. Slices were incubated for 5 min in Mayers haemalaun solution (Merck, Darmstadt, Germany) for staining of the cells. After 10 min in water, the probes were embedded in Kaisers glycerinegelatine (Merck, Darmstadt, Germany). To rule out unspecific binding of the antibodies, samples were incubated with the second and third antibody but not with the primary one.

3. Results 3.1. Proliferation and ALP activity BMP-transfected and non-transfected C3H10T1/2 cells assumed a fibroblastic morphology when cultured in monolayer cultures over 17 days, even after addition of differentiation medium. They proliferated moderately. Independent of the cell line, the cell number increased from 1
106 cells/ml on Day 0 to 3– 4
106 cells/ml on Day 17 (Fig. 1). The alkaline phosphatase (ALP) activity was low and remained constant within the limits of accuracy (about 1 pg/min/ cell corresponding to 1.5–2.5 pg/min/mg protein/cell; Fig. 2). The protein concentration also did not change (about 50 pg/cell). In contrast, three-dimensionally cultured cells (about 90% as revealed by PI staining)

2007

M. Weber et al. / Biomaterials 23 (2002) 2003–2013 6

5x10

6

cells/ml

4x10

6

3x10

6

2x10

6

1x10

0 0

2

4

6

8

10

12

14

16

18

time [d] Fig. 1. Typical proliferation curves of wild-type (squares), BMP-2 (triangles) and BMP-4 (circles) transfected C3H10T1/2 cells in monolayer (open symbols) and alginate microcapsules (closed symbols) cultures. The initial suspension density was 1
106 cells/ml. Encapsulation was performed by using a 0.5% w/v UHV/CG alginate solution that was cross-linked with 20 mm Ca2+. On Day 3, the culture medium was replaced by differentiation medium. The number of viable cells was counted electronically on a regular basis after treatment of the monolayer cells with a trypsin/EDTA solution and after release of the encapsulated cells by EDTA treatment, respectively. The measurements represent the mean (7SD) of at least three experiments. 10

(A)

8

wild-type

6 4 2 0

ALP [pg/min/cell]

changed to a spherical appearance, i.e. a configuration that has been correlated with a more differentiated phenotype [29]. They showed no increase in cell count (Fig. 1), but an increase in ALP activity. For the wildtype cells ALP activity was gradually upregulated after Day 4, whereas BMP-2 and BMP-4 transfected cells showed an initial increase in the ALP level after Day 6 (Fig. 2). The peak value in ALP activity was reached between Days 8 and 10 in order to decrease then again increase to the original value on Day 13. The highest level for ALP activity was found for BMP-4 transfected cells (6.6 pg/min/cell corresponding to 7.08 pg/min/mg protein/cell). BMP-2 cells and nontransfected cells exhibited significantly smaller values (4.6 pg/min/cell corresponding to 3.6 pg/min/mg protein/cell and 3.5 pg/min/cell corresponding to 2.25 pg/ min/mg protein/cell, respectively). Total protein content per encapsulated cell was comparable to that of the monolayer cells and did not change during culture.

10

(B)

BMP-2

8 6 4 2 0 10

(C)

BMP-4

8 6 4 2

3.2. Type I and type II collagen mRNAS

0 0

Quantitative RT-PCR was used to analyze mRNAs for type I and type II collagen in monolayer and encapsulated cells. Type II collagen is a specific component of hyaline cartilage, while type I collagen is a typical marker for non-hyaline cartilage and is classically taken as a marker for dedifferentiated chondrocytes in culture [30]. It is also detected in chondroprogenitor cells [31]. Representative experi-

2

4

6

8

10

12

14

16

18

time [d] Fig. 2. Typical time dependence of alkaline phosphatase (ALP) activity of wild-type (A), BMP-2 (B) and BMP-4 (C) transfected C3H10T1/2 cells in monolayer (open symbols) and alginate microcapsule (closed symbols) cultures. Encapsulation, culture conditions and replacement of culture medium by differentiation medium were performed according to the experimental conditions described in Fig. 1. The measurements represent the mean (7SD) of at least three experiments. For further details, see text.

2008

M. Weber et al. / Biomaterials 23 (2002) 2003–2013

also obtained for encapsulated BMP-2 cells (Fig. 3). In contrast, encapsulated BMP-4 cells (Fig. 3B) showed an immediate upregulation in type II collagen mRNA after encapsulation. After Day 6, a decrease was observed. On Day 10, when type I collagen mRNA level assumed a peak value, a minimum value was reached. However, with the ongoing culture, type II collagen mRNA increased markedly again. On Day 17, its level exceeded the value of Day 6.

molecules / 1000 molecules GAPDH

ments are shown in Fig. 3. The level of type I collagen mRNA of the wild-type was low for monolayer cells throughout the entire culture period. After encapsulation, it increased only slightly on Day 6. Evaluation of type I collagen mRNA in transfected cells also yielded low values for the monolayer cells, but increased levels were found in encapsulated cells after Day 8 (Fig. 3A). Corresponding to the ALP activity, peak values were reached on Day 10. At this time, the level of this matrix gene was significantly higher in BMP-4 cells than in BMP-2 cells. With ongoing culture, the level of type I collagen mRNA decreased in both transfected cell lines and reached the level of the monolayer cells on Day 17 (Fig. 3A). The values of type II collagen mRNA for the transfected and non-transfected monolayer cells were of the same order of magnitude as the corresponding values of type I collagen mRNA. Similar results were

3.3. Cartilage matrix proteins Inspection of sections from encapsulated BMP-4 cells showed the occurrence of immunoreactivity when treated with type II collagen antibodies after Day 8. With increasing time of culture, the intensity of the staining of type II collagen increased. A typical result for cells taken at Day 17 is shown in Fig. 4A. The

50

(A) 40

30

20

10

0 4

6

8

10

12

14

16

18

14

16

18

molecules / 1000 molecules GAPDH

time [days] 18 16

(B)

14 12 10 8 6 4 2 0 4

6

8

10

12

time [days] Fig. 3. Real time RT-PCR for type I collagen (A) and type II collagen (B) mRNA expression in monolayer (open symbols) and microcapsule cultures (closed symbols) of wild-type (squares), BMP-2 (triangles) and BMP-4 (circles) transfected C3H10T1/2 cells. The mRNA-values are presented as numbers of molecules per 1000 molecules GAPDH. For experimental details see Materials and Methods and legend to Fig. 1. The measurements represent the mean (7SD) of at least three independent experiments.

M. Weber et al. / Biomaterials 23 (2002) 2003–2013

2009

Fig. 4. Immunolocalization of matrix proteins in alginate-encapsulated (A-D) and monolayer-cultured (E-F) C3H10T1/2 cells transfected with cDNA encoding for BMP-4. Immunostaining was performed 17 days after culture in growth and differentiation medium, respectively, according to the experimental conditions described in Fig. 1. For dissolution of the alginate matrix and preparation of microcapsule sections (A-D) as well as for the treatment of the cells before staining with the appropriate specific antibodies, see Materials and Methods. Note that cell clustering occurred upon dissolution of the alginate matrix (A-D). Positive immunostaining for type II collagen (A and F), chondroitin-6-sulfate (B), chondroitin-4-sulfate (D) and type I collagen (C and E), respectively, is indicated by the brown-colored fibril network (arrows) in the vicinity of the blue-colored cells (stained with hematoxylin). For further details, see text. All panels were reproduced at the same magnification (bar=20 mm).

aggregation of cells seen in the figure occurred through the degradation of the alginate matrix by the peroxidase reaction. It is evident that the central cells were surrounded completely by a tight network of collagen fibers whereas less collagen production obviously was observed in the periphery cell ring of the cluster. The central areas also showed positive, but slightly weaker, reactivity with chondroitin-6-sulfate antibodies (Fig. 4B). While chondroitin-6-sulfate is a specific component of hyaline cartilage, chondroitin-4-sulfate is also found in other tissues. This matrix component was detectable after Day 1 (Fig. 4D). Formation of type X collagen apparently did not occur (data not shown).

However, as shown in Fig. 4C, immunostaining revealed small amounts of type I collagen (Fig. 4C). In contrast, intense immunoreactivity was found for type I collagen in monolayer cells (Fig. 4E). Positive type II collagen was also demonstrated (Fig. 4F), but the reaction was rather weak compared to encapsulated cells. Chondroitin-4-sulfate and chondroitin-6-sulfate could not be detected in monolayer cells. Similar immunoreactivities as described for BMP-4 transfected cells were observed for BMP-2 transfected and wild-type cells (Fig. 5). Data were also taken at Day 17. Comparison of Fig. 5 with Fig. 4 shows that there were some notable differences in the immunoreaction of

2010

M. Weber et al. / Biomaterials 23 (2002) 2003–2013

chondroitin-4-sulfate could be demonstrated after Day 1 throughout the entire culture period. Encapsulated wildtype cells displayed an intermediate type II collagen immunoreactivity between BMP-4 and BMP-2 transfected cells (Fig. 5E). In monolayer cells, chondroitin-4sulfate was first detected on Day 10, while type II collagen was completely absent.

4. Discussion

Fig. 5. Immunolocalization of matrix proteins in alginate-encapsulated (A-B,E-F) and monolayer-cultured (C-D,G-H) C3H10T1/2 cells transfected with cDNA encoding for BMP2 (A-D) and the nontransfected wild-type cells (E-H). Immnostaining was made as described in the legend of Fig. 4. Positive staining for type II collagen (A, C, E and G) and type I collagen (B, D, F and H), respectively, is indicated by the brown-colored fibril network (arrows) in the vicinity of the blue-colored cells (stained with hematoxylin). For further details, see text. All panels were reproduced at the same magnification (bar=20 mm).

these cells (listed in Table 1). Encapsulated BMP-2 transfected cells were surrounded by a thin rim of matrix rich in type II collagen after 17 days of cultivation (Fig. 5A), but no type I collagen could be detected (Fig. 5B). Monolayer cells showed no immunoreactivity for both type I and type II collagen (Fig. 5C and D), but

The findings presented here for the wild-type and for BMP-transfected C3H10T1/2 cells demonstrate that encapsulation in a three-dimensional matrix made up of a cross-linked ultra-high viscosity alginate prevents cell proliferation, but promotes cell differentiation. While BMP-transfected and wild-type cells exhibited a fibroblast-like shape during monolayer tissue culture and exhibited proliferation accompanied by type I collagen production, encapsulated cells changed their phenotype to spherical, did not proliferate and were characterized by upregulation of type II collagen and downregulation of type I collagen. This, along with other evidence (see below), strongly suggests that the cells were undergoing chondrogenic differentiation upon encapsulation. Round morphology and synthesis of a matrix rich in type II collagen and other hyaline cartilage specific glycosaminoglycans have also been found for chondrocytes when encapsulated in Ca2+alginate [6,17,32–34]. In contrast to chondrocytes [6,32,34,35], however, proliferation of encapsulated C3H10T1/2 cells could not be observed. Proliferation of chondrocytes in an alginate matrix, although at a significantly lower rate than for cells plated onto tissue culture plastics, may be due to the presence of mitogenic and cytotoxic contaminants in the commercial alginate currently used by most of the laboratories (see also below). A further advantage of the clinical grade alginate used in this study is also its much higher viscosity compared to that of commercial alginate. A 0.1% (w/v) solution of UHV/ CG alginate exhibits a viscosity of about 20 cP, whereas the viscosity of a corresponding solution of commercial alginate is 1–5 cP [9]. Thus, in contrast to commercial alginates low UHV/CG alginate concentrations can be used to produce capsules. The mechanical strength of these capsules is significantly higher than that made up of low-viscosity alginate. It is hypothesized that the tight polymeric network prevented not only shape changes, but also cell division due to mechanical restrictions. The large amount of bound water within the capsules as revealed by electrorotation measurements [10,36] and diffusion limitations of oxygen within the capsule interior [37–40] may be additional factors that could be responsible for the suppression of cell proliferation in UHV/CG alginate and, in turn, for the upregulation of

2011

M. Weber et al. / Biomaterials 23 (2002) 2003–2013

Table 1 The immunoreactivity of different matrix markers for the investigated cell lines, in monolayer and encapsulated in Ca2+ cross-linked UHV/CG alginate after 17 days of culture, respectivelya

Encapsulated wild-type Encapsulated BMP-2 Encapsulated BMP-4 Wild-type monolayer BMP-2 monolayer BMP-4 monolayer a

Collagen I

Collagen II

Chondroitin-4-sulfate

Chondroitin-6-sulfate

+  + ++  ++

++ + +++   +

+++ +++ +++ + + 

++ ++ ++   

The initial cell densiy was 1
106 cells/ml. Key: , not detectable; +, detectable; ++, high yield; +++, very high yield.

cartilage matrix proteins. Support for the assumption that the features of the alginate capsule prevent cell growth and division was also obtained by experiments with encapsulated human chondrocytes. When using the same experimental encapsulation protocol, as described here, significant proliferation of chondrocytes could generally not be detected over 3 months (unpublished data). However, it should be noted that chondrocytes of two patients encapsulated in UHV/CG alginate showed a considerable proliferation rate associated with the formation of a collagen matrix inside and outside the capsules [23]. This indicates that not only matrix-related, but also cell-related factors (such as differentiation state, density, cytokines, etc.) and thus the source of chondrocytes play an important role. These findings corroborate the view mentioned further above (see also [41]) that established culture cell lines are more useful tools than primary cells to study proliferation and extracellular matrix production, respectively, in three-dimensional hydrogels under appropriate, well-defined conditions. This is a prerequisite for the development of optimum alginate-based chondrocyte transplants. Similarly as the wild-type cells, BMP-transfected cells showed only synthesis of hyaline matrix proteins when the cells were cultured in the alginate scaffold. It is likely that the tight network and the non-physiological aqueous surrounding facilitated the accumulation of these growth factors in the neighborhood of the encapsulated cells. Thus, the critical level for BMP reactivity can apparently be reached more easily compared to monolayer cultures. It is well known from studies with the MC615 chondrocyte cell line [42] that the effect of BMP-2 and BMP-4 on the level of type II collagen mRNA increased with increasing concentration of these BMPs. However, the mRNA level gradually declined as the concentrations of the BMPs increased over 100 ng/ml. Cytotoxic effects were even observed in cultures treated with concentrations over 200 ng/ml. The time dependence of accumulation of BMP-2 and BMP-4 under the conditions of this study is most likely affected by cellular andFrelated with thisFmatrix factors. Thus, in the light of the concentrationdependent effects of BMPs on type II collagen gene

expression, it is not surprising thatFin contrast to the finding of Valcourt [42]Fboth transfected cell lines displayed quantitatively different expressions of cartilage marker proteins with the ongoing culture. Immunohistochemistry and analysis of the mRNA levels have presented strong evidence that encapsulated BMP-4 transfected cells produced considerably more type II collagen than encapsulated BMP-2 transfected cells. In addition to type II collagen, BMP-4 transfected cells also produced type I collagen, whereas this protein could not be detected in BMP-2 transfected cells (in agreement with the results of Yamaguchi [15]). BMP-4 transfected cells also exhibited much higher ALP activity levels than BMP-2 (and the wild-type) cells, even though the time course of the development and the subsequent decline of ALP activity was qualitatively similar for the three cell lines. In contrast to the cartilage-specific marker type II collagen, ALP is thought to be involved in the mineralization process and is thus presumably a marker of osteogenesis [19,43], even though its exact function is still unclear [44]. In this context, it is interesting to note that preliminary experiments with cells encapsulated in commercial alginate did not show any increase in ALP activity. In encapsulated BMP-4 transfected cells, the ALP activity peak value coincided with the decline of type I collagen production and, in turn, with a new secretion of type II collagen. This finding suggests that C3H10T1/2 cells differentiate to chondrocytes via a chondroprogenitorlike cell. Type X collagen could not be detected. This indicates that the cells did not differentiate into hypertrophic chondrocytes [31]. It is well known [15,31] that chondroprogenitor cells and not chondrocytes produce type I collagen. The transient upregulation in ALP activity found here for encapsulated cells is not reported in the literature for this cell line. Monolayer cells showed only low and constant ALP levels over the entire culture period (see also [14]). The reason for upregulation of ALP activity in the alginate matrix is not known. However, it is quite likely that the features of the ultra-high viscosity, clinical-grade alginate matrix played an important role in the development of ALP activity.

2012

M. Weber et al. / Biomaterials 23 (2002) 2003–2013

Thus, we can conclude that UHV/CG alginate is a suitable scaffold for initiation of differentiation processes in mesenchymal progenitor cell lines and chondrocytes. Preliminary rat studies in which C3H10T1/2-BMP-4 (and autologous mesenchymal stem cells) encapsulated in UHV/CG alginate were transplanted in cartilage defects strongly corroborate this view.

Acknowledgements The authors are very grateful to J. Pfeuffer, P. Amersbach and S. Schmitt for their excellent technical assistance. This work was supported by grants of the Bundesministerium fur . Bildung, Wissenschaft und Forschung (FKZ: 0311716) and the Graduiertenkolleg GRK 64/2 to U. Z. and grants from the Freistaat Bayern (No. 4c-1999) to C.H.

References [1] Marijnissen WJCM, van Osch GJVM, Aigner J, VerwoerdVerhoef HL, Verhaar JAN. Tissue-engineered cartilage using serially passaged articular chondrocytes. Chondrocytes in alginate, combined in vivo with a synthetic (E210) or biologic biodegradable carrier (DBM). Biomaterials 2000;21:571–80. [2] Kuhtreiber . WM, Lanza RP, Chick WL. Cell encapsulation technology and therapeutics. Boston, MA: Birkh.auser, 1999. [3] Selden C, Khalil M, Hodgson H. Three dimensional culture upregulates extracellular matrix protein expression in human liver cell linesFa step towards mimicking the liver in vivo? Int J Artif Organs 2000;23:774–81. [4] Hasse C, Bohrer T, Barth P, Stinner B, Cohen R, Cramer H, Zimmermann U, Rothmund M. Parathyroid xenotransplantation without immunosuppression in experimental hypoparathyroidism: long-term in vivo function following microencapsulation with a clinically suitable alginate. World J Surg 2000; 24:1361–6. [5] Atala A, Kim W, Paige KT, Vacanti CA, Retik AB. Endoscopic treatment of vesicoureteral reflux with a chondrocyte-alginate suspension. J Urol 1994;152:641–3. [6] Paige KT, Cima LG, Yaremchuk MJ, Vacanti JP, Vacanti CA. Injectable cartilage. Plast Reconstr Surg 1995;96:1390–8. [7] Paige KT, Cima LG, Yaremchuk MJ, Schloo BL, Vacanti JP, Vacanti CA. De novo cartilage generation using calcium alginatechondrocyte constructs. Plast Reconstr Surg 1996;97:168–78. . G, Federlin K, Hannig K, Kowalski M, [8] Zimmermann U, Klock Bretzel RG, Horcher A, Entenmann H, Sieber U, Zekorn T. Production of mitogen-contamination free alginates with variable ratios of mannuronic acid to guluronic acid by free flow electrophoresis. Electrophoresis 1992;13:269–74. [9] Zimmermann U, Mimietz S, Zimmermann H, Hillg.artner M, Schneider H, Ludwig J, Hasse C, Haase A, Rothmund M, Fuhr G. Hydrogel-based non-autologous cell and tissue therapy. Biotechniques 2000;29:564–81. [10] Zimmermann U, Thurmer . F, Jork A, Weber M, Mimietz S, Hillg.artner M, Brunnenmeier F, Zimmermann H, Westphal I, . U, Haase A, Steinert A, Hendrich C. A novel Fuhr G, Noth class of amitogenic alginate microcapsules for long-term immunoisolated tranplantation. Ann N Y Acad Sci 2001; 944:199–216.

[11] Sittinger M, Bujia J, Rotter N, Reitzel D, Minuth WW, Burmester GR. Tissue engineering and autologous transplant formation: practical approaches with resorbable biomaterials and new cell culture techniques. Biomaterials 1996;17: 237–42. [12] Hillg.artner M, Zimmermann H, Mimietz S, Jork A, Thurmer . F, . U, Hasse C, Haase A, Fuhr G, Rothmund M, Schneider H, Noth Zimmermann U. Immunoisolation of transplants by entrapment in 19F-labelled alginate gels: production, biocompatibility, stability, and long-term monitoring of functional integrity. Mat-wiss u Werkstofftech 1999;30:783–92. [13] Hortelano G, Stockley T. Implantable microcapsules for gene therapy for hemophilia. In: Kuhtreiber . WM, Lanza RP, Chick WL, editors. Cell encapsulation technology and therapeutics. Boston, MA: Birkh.auser, 1999. p. 321–9. . [14] Ahrens M, Ankenbauer T, Schroder D, Hollnagel A, Mayer H, Gross G. Expression of human bone morphogenetic proteins-2 or -4 in murine mesenchymal progenitor C3H10T1/2 cells induces differentiation into distinct mesenchymal cell lineages. DNA Cell Biol 1993;12:871–80. [15] Yamaguchi A. Regulation of differentiation pathway of skeletal mesenchymal cells in cell lines by transforming growth factor-b superfamily. Semin Cell Biol 1995;6:165–73. [16] Caplan AI, Goldberg VM. Principles of tissue engineered regeneration of skeletal tissues. Clin Orthop 1999;367(Suppl): 12–6. [17] Lemare F, Steimberg N, Le Griel C, Demignot S, Adolphe M. Dedifferentiated chondrocytes cultured in alginate beads: restoration of the differentiated phenotype and of the metabolic responses to interleukin-1b. J Cell Physiol 1998; 176:303–13. [18] Wang EA, Israel DI, Kelly S, Luxenberg DP. Bone morphogenetic protein-2 causes commitment and differentiation in C3H10T1/2 and 3T3 cells. Growth Factors 1993;9:57–71. [19] Sakano S, Murata Y, Miura T, Iwata H, Sato K, Matsui N, Seo H. Collagen and alkaline phosphatase gene expression during bone morphogenetic protein (BMP)-induced cartilage and bone differentiation. Clin Orthop 1993;292:337–44. [20] Sellers RS, Peluso D, Morris EA. The effect of recombinant human bone morphogenetic protein-2 (rhBMP-2) on the healing of full-thickness defects of articular cartilage. J Bone Jt Surg Am 1997;79:1452–63. [21] Groeneveld EHJ, Burger EH. Bone morphogenetic proteins in human bone regeneration. Eur J Endocrinol 2000;142:9–21. [22] Kramer J, Hegert C, Guan K, Wobus AM, Muller . PK, Rohwedel J. Embryonic stem cell-derived chondrogenic differentiation in vitro: activation by BMP-2 and BMP-4. Mech Dev 2000;92: 193–205. [23] Jork A, Thurmer . F, Cramer H, Zimmermann G, Gessner P, H.amel K, Hofmann G, Kuttler B, Hahn HJ, Josimovic-Alasevic O, Fritsch KG, Zimmermann U. Biocompatible alginate from freshly collected Laminaria pallida for implantation. Appl Microbiol Biotechnol 2000;53:224–9. . G, Pfeffermann A, Ryser C, Grohn . P, Kuttler B, Hahn HJ, [24] Klock Zimmermann U. Biocompatibility of mannuronic acid-rich alginates. Biomaterials 1997;18:707–13. . [25] Djuzenova CS, Sukhorukov VL, Klock G, Arnold WM, Zimmermann U. Effect of electric field pulses on the viability and on the membrane-bound immunoglobulins of LPS-activated murine B-lymphocytes: correlation with the cell cycle. Cytometry 1994;15:35–45. [26] Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248–54. [27] Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitative PCR. Genome Res 1996;6:986–94.

M. Weber et al. / Biomaterials 23 (2002) 2003–2013 [28] Gibson UE, Heid CA, Williams PM. A novel method for real time quantitative RT-PCR. Genome Res 1996;6:995–1001. [29] H.auselmann HJ, Aydelotte MB, Schumacher BL, Kuettner KE, Gitelis SH, Thonar EJ. Synthesis and turnover of proteoglycans by human and bovine adult articular chondrocytes cultured in alginate beads. Matrix 1992;12:116–29. [30] Mayne R, Vail MS, Mayne PM, Miller EJ. Changes in type of collagen synthesized as clones of chick chondrocytes grow and eventually lose division capacity. Proc Natl Acad Sci USA 1976;73:1674–8. [31] Binette F, McQuaid DP, Haudenschild DR, Yaeger PC, McPherson JM, Tubo R. Expression of a stable articular cartilage phenotype without evidence of hypertrophy by adult human articular chondrocytes in vitro. J Orthop Res 1998;16:207–16. [32] Ramdi H, Legay C, Lievremont M. Influence of matricial molecules on growth and differentiation of entrapped chondrocytes. Exp Cell Res 1993;207:449–54. [33] Cancedda R, Descalzi Cancedda F, Castagnola P. Chondrocyte differentiation. Int Rev Cytol 1995;159:265–358. [34] Beekman B, Verzijl N, Bank RA, von der Mark K, TeKoppele JM. Synthesis of collagen by bovine chondrocytes cultured in alginate; posttranslational modifications and cell–matrix interaction. Exp Cell Res 1997;237:135–41. [35] Guo JF, Jourdian GW, MacCallum DK. Culture and growth characteristics of chondrocytes encapsulated in alginate beads. Connect Tissue Res 1989;19:277–97. [36] Esch M, Sukhorukov VL, Kurschner . M, Zimmermann U. Dielectric properties of alginate beads and bound water relaxation studied by electrorotation. Biopolymers 1999;50:227–37. [37] Rajpurohit R, Koch CJ, Tao Z, Teixeira CM, Shapiro IM. Adaptation of chondrocytes to low oxygen tension: relationship

[38]

[39]

[40]

[41]

[42]

[43]

[44]

2013

between hypoxia and cellular metabolism. J Cell Physiol 1996;168:424–32. O’Driscoll SW, Fitzsimmons JS, Commisso CN. Role of oxygen tension during cartilage formation by periosteum. J Orthop Res 1997;15:682–7. Domm C, Fay J, Schunke . M, Kurz B. Die Redifferenzierung von dedifferenzierten Gelenkknorpelzellen in Alginatkultur. Einfluss von intermittierendem hydrostatischen Druck und niedrigem Sauerstoffpartialdruck. Orthop.ade 2000;29:91–9. Grimshaw MJ, Mason RM. Bovine articular chondrocyte function in vitro depends upon oxygen tension. Osteoarthritis Cartilage 2000;8:386–92. Kokenyesi R, Tan L, Robbins JR, Goldring MB. Proteoglycan production by immortalized human chondrocyte cell lines cultured under conditions that promote expression of the differentiated phenotype. Arch Biochem Biophys 2000;383: 79–90. Valcourt U, Ronzi"ere MC, Winkler P, Rosen V, Herbage D, Mallein-Gerin F. Different effects of bone morphogenetic proteins 2, 4, 12, and 13 on the expression of cartilage and bone markers in the MC615 chondrocyte cell line. Exp Cell Res 1999;251:264–74. de Bernard B, Bianco P, Bonucci E, Costantini M, Lunazzi GC, Martinuzzi P, Modricky C, Moro L, Panfili E, Pollesello P, Stagni N, Vittur F. Biochemical and immunohistochemical evidence that in cartilage an alkaline phosphatase is a Ca2+-binding glycoprotein. J Cell Biol 1986;103:1615–23. Kergosien N, Sautier J-M, Forest N. Gene and protein expression during differentiation and matrix mineralization in a chondrocyte cell culture system. Calcif Tissue Int 1998; 62:114–21.