Generation and characterization of ... - Wiley Online Library

Generation and Characterization of Osteochondral Grafts With Human Nasal Chondrocytes Marina Barandun,1 Lukas Daniel Iselin,1 Francesco Santini,2 Michele Pansini,2 Celeste Scotti,3 Daniel Baumhoer,4 Oliver Bieri,2 Ueli Studler,2 Dieter Wirz,5,6 Martin Haug,1 Marcel Jakob,1 Dirk Johannes Schaefer,1 Ivan Martin,1 Andrea Barbero1* 1

Departments of Surgery and of Biomedicine, Basel University Hospital, University of Basel, Basel, Switzerland, 2Department of Radiology, Clinic of Radiology and Nuclear Medicine, University of Basel Hospital, Basel, Switzerland, 3I.R.C.C.S., Istituto Ortopedico Galeazzi, Milano, Italy, 4 Bone Tumor Reference Center at the Institute of Pathology, Basel University Hospital, Basel, Switzerland, 5Laboratory for Biomechanics and Biocalorimetry, Biozentrum– Pharmazentrum, University of Basel, Basel, Switzerland, 6Orthomerian, Gotthelfstrasse 105 4054, Basel, Switzerland Received 12 December 2014; accepted 8 February 2015 Published online 28 May 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jor.22865

ABSTRACT: We investigated whether nasal chondrocytes (NC) can be used to generate composite constructs with properties necessary for the repair of osteochondral (OC) lesions, namely maturation, integration and capacity to recover from inflammatory burst. OC grafts were fabricated by combining engineered cartilage tissues (generated by culturing NC or articular chondrocytes – AC – onto 1 1 Chondro-Gide matrices) with devitalized spongiosa cylinders (Tutobone ). OC tissues were then exposed to IL-1b for three days and cultured for additional 2 weeks in the absence of IL-1b. Cartilage maturation extent was assessed (immune) histologically, biochemically and by delayed gadolinium-enhanced magnetic resonance imaging of cartilage (dGEMRIC) while cartilage/bone integration was assessed using a peel-off mechanical test. The use of NC as compared to AC allowed for more efficient cartilage matrix accumulation and superior integration of the cartilage/bone layers. dGEMRIC and biochemical analyzes of the OC constructs showed a reduced glycosaminoglycan (GAG) contents upon IL-1b administration. Cartilaginous matrix contents and integration forces returned to baseline up on withdrawal of IL-1b. By having a cartilage layer well developed and strongly integrated to the subchondral layer, OC tissues generated with NC may successfully engraft in an inflammatory post-surgery joint environment. ß 2015 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 33:1111–1119, 2015. Keywords: tissue engineering; osteochondral lesion; chondrocytes; dGEMRIC; chondrogenic differentiation

Trauma and disease of joints often result in damages of the articular cartilage surface and the underlying subchondral bone. Such osteochondral (OC) defects do not spontaneously heal, leading to progressive articular surface damage and, ultimately, to osteoarthritic degenerative changes.1–4 Despite various therapeutic approaches have been developed to treat OC defects, none of them have proved yet to ensure long-lasting articular surface regeneration. Autologous or allogeneic osteochondral transplants are often considered for the treatment of OC lesions but such techniques are potentially limited by the availability of material, donor site morbidity and risk of infection.5 The fabrication of engineered OC composites, using autologous cells and suitable three-dimensional (3D) scaffolds, would provide the possibility to repair the damaged articular joint by grafting biologically and biomechanically competent tissues, obtained with minimal donor site morbidity. OC grafts have been mainly generated using mesenchymal stem/stromal cells (MSC) or articular chondrocytes (AC).6 However the use of these cell sources for the fabrication of OC as graft material to treat articular defects would be hampered by the phenotypic instability of the cartilage tissue formed by MSC7 or the large inter-donor

Marina Barandun, Lukas Daniel Iselin contributed equally to this work. Grant sponsor: Swiss National Science Foundation; Grant number: 310030-126965.1. Correspondence to: Andrea Barbero (T: þ41 61 265 2379; F: þ41 61 265 3990; E-mail: [email protected]) # 2015 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.

variability in the cartilage forming capacity of AC.8 To bypass the aforementioned critical issues, a more reproducible chondrogenic cell source should be identified. As compared to AC, nasal chondrocytes (NC) were shown to have a higher capacity to generate functional cartilaginous tissues, with lower donor-related dependency,9–11 and to respond similarly to AC to physical forces resembling joint loading.12 Moreover, in a recent study,13 we have demonstrated the molecular compatibility of NC at an articular site, verified by pre-clinical studies in goats and by an ongoing clinical trial at the University-Hospital Basel (http://clinicaltrials.gov Identifier: NCT01605201). In addition, the use of autologous NC for generation of OC grafts would ensure minimal donor site morbidity, especially when compared to the use of autologous AC.12,14 It has, however, not been explored yet whether NC can be used to generate composite constructs with properties enabling their utilization for the repair of OC lesions. With this goal in mind, we performed the following investigations. First, OC tissues generated by combining NC or AC with biomaterials currently in clinical use15 were compared for the amount of cartilage matrix accumulated in the chondral layer and for the extent of cartilage/bone integration. We then investigated how these properties were modulated by a shortterm exposure to interleukin (IL)-1b, mimicking the initial inflammatory implantation site, and following a recovery time. Finally, with the perspective to monitor quantitatively the maturation of the grafts in a non-invasive manner, we assessed whether delayed gadolinium-enhanced magnetic resonance imaging of JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2015

1111

1112

BARANDUN ET AL.

cartilage (dGEMRIC) allowed an accurate estimation of the GAG contents in the cartilage layers of the OC composite.

METHODS Cartilage Biopsies, Chondrocytes Isolation and Expansion For the preliminary experiment aimed at comparing the properties of osteochondral (OC) composites generated with articular chondrocytes (AC) and nasal chondrocytes (NC), matched healthy articular and nasal cartilage biopsies were harvested, respectively, from full-thickness biopsies of the femoral condyle and from the nasal septum of three cadaver donors (two males and one female; mean age 68 years; range 58–79 years). The donor (male, 58 year) from which AC had the highest chondrogenic potential, tested by means of a micromass culture system,8 was chosen for this first experiment (during such screening, noteworthy, NC from all the three donors displayed similarly high chondrogenic capacities). For the other experiments performed to investigate the responses of the OC composites to IL-1b stimulation, chondrocytes were collected from the nasal septum of four patients (three males and one female; mean age: 35 years; range 35–41 years) undergoing rhinoplasty surgery at the University Hospital Basel. Articular cartilage specimens collected from the femoral condyles of two additional cadavers (male, age 56 years, and female, age 78 years) were used as control samples for the dGEMRIC study. All tissues were harvested following informed consent and local ethical committee approval. Cartilage tissues were immediately used for the generation of OC composites as described below or minced into small pieces and digested with 0.15% (w/v) type II collagenase (355 U/mg; Worthington Biochemical Corp., Lakewood, NJ, 10 ml solution/g tissue) for 22 h. The isolated AC and NC were expanded for two passages (corresponding to 7.8 0.9 and 8.9 1.2 population doublings, respectively) with Dulbecco’s Eagle’s Medium (DMEM) containing 4.5 mg/ml D-glucose, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 100 mM HEPES buffer, 100 U/ml penicillin, 100 mg/ml streptomycin and 0.29 mg/ml L-glutamate (basic medium) supplemented with 10% (v/v) of fetal bovine serum (FBS), 1 ng/ml of Transforming Growth Factor b1 (TGFb-1), and 5 ng/ml of Fibroblast Growth Factor 2 (all from R&D Systems), as previously described.16 Generation of Osteochondral Samples Osteochondral tissues were generated according to a previously established protocol15 using collagen type I/III matrices 1 (8-mm-diameter, 1.5-mm-thick disks, Chondro-Gide , Geistlich Pharma AG), and devitalized bovine cancellous bone 1 blocks (8-mm-diameter, 10-mm-thick cylinders, Tutobone , Tutogen Medical GmbH) respectively for the cartilage and bone layers. Chondrocytes were detached from culture flasks and seeded onto the matrix at a density of 4 106 cells/cm3. Cell-seeded matrices were then culture with basic medium supplemented with 5% (v/v) FBS, 0.1 mM ascorbic acid 1 (Sigma), 10 mg/ml Insulin (Actrapid , Novo Nordisk) (Chondrogenic medium, CM) (R&D Systems) in a humidified 37 ˚C/ 5% CO2 incubator. After 3 day’s culture, 20 ml/construct of a 1 40 mg/ml fibrinogen solution (Tisseel , Baxter) was added on the seeded surfaces of the constructs, and the constructs were immediately placed on the top of the bony cylinders, 1 previously wetted in 8 U/ml thrombin solution (Tisseel , Baxter). After 20 min of polymerization, the combined conJOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2015

structs were transferred to culture dishes and cultured in CM with medium changes twice a week. NC based OC constructs were cultured either in CM containing 10 ng/ml of TGFb-1 or in CM without TGFb-1. AC based constructs were cultured only in the presence of TGFb-1, based on preliminary results demonstrating poor extracellular matrix deposition by AC cultured in the absence of TGFb-1. After 4 weeks of culture, constructs were exposed to 1 ng/ml human recombinant IL-1b (Sigma Chemical, St. Louis, MO) for 3 days and then assessed as described below or cultured for additional 2 weeks in the absence of IL-1b. For the dGEMRIC study, additional OC composites were generated by gluing native cartilage tissues or unseeded collagen sponges previously 1 shaped with an 8 mm punch biopsy to the Tutobone blocks using the aforementioned protocol. Peel-Off Test for the Quantification of the Mechanical Integration Between the Chondral and Bone Layer of OC Constructs The peel-off test was performed as previously described15,17 on a small-scale mechanical testing machine (MTS Synergie 100, MTS Systems, Inc., Eden Prairie, MN) equipped with a 2 N load cell. The bone layer of the OC construct was fixed with a modified screw clamp. The edge of the chondral layer was grasped with a small surgical clamp attached to the crosshead of the testing machine. The crosshead was raised at 1 mm/s, and force and displacement were recorded until full detachment of the chondral layer occurred. Parameters assessed were maximum peel-off force and total peel energy, normalized to the interfacial areas of the samples. dGEMRIC Quantification of Glycosaminoglycan Contents in the Chondral Layer of OC Constructs OC tissues were placed onto wells preformed in a customized silicone mold (5 constructs/mold) (Fig.1A) which was inserted in a 50 ml tube. Approximately 25 ml of DMEM was added in order to cover all the remaining volume of the tube. A non-invasive estimation of the GAG concentration in the samples was performed by means of Magnetic Resonance Imaging (MRI) using the delayed gadolinium-enhanced magnetic resonance imaging of cartilage (dGEMRIC) method.18 This method consists of the acquisition of quantitative T1 maps of the tissue, both natively and after the addition of Gadolinium to the culture medium. The concentration of GAG can be derived from the T1 values using the Donnan equilibrium theory.19,20 The samples were scanned in a human whole-body MRI machine at a field strength of 3 Tesla (Magnetom Prisma, Siemens Healthcare, Erlangen, Germany) using a 12-channel wrist coil as a signal receiving element. T1 quantification was performed with a variable flip angle three-dimensional spoiled gradient echo sequence21 with flip angles of 4˚ and 15˚, TR/TE 5.2/1.8 ms, resolution 0.6 0.6 0.6 mm3, Fieldof-view 150 37 34 mm3, sagittal orientation, number of averages 32, resulting in a total scan time of 56 min. After the first measurements the DMEM was removed from the tube containing the OC grafts and replaced with new DMEM additioned with the contrast agent gadopentetate dimeglumine (Gd-DTPA, Magnevist1, Bayer Schering AG, Berlin, Germany, 1 mM). Constructs were incubated for 4 h at 37 ˚C before performing the second MRI measurements. The second MRI experiment was performed with the aforementioned protocol. T1 maps were calculated offline

NASAL CHONDROCYTES FOR OSTEOCHONDRAL TISSUE FABRICATION

1113

Figure 1. Comparison on the capacity of articular chondrocytes (AC) and nasal chondrocytes (NC) to form osteochondral (OC) composite. (A) Representative macroscopic appearances and Safranin O staining of OC composites generated by culturing AC and NC seeded Chondro-Gide1 matrices on the top of Tutobone1 blocks in chondrogenic medium without (TGF) or with (þTGF) Transforming Growth Factor ß-1. Dotted line delimitate the thickness of the chondral layers. The inserts are high magnification images of chondral layers. Scale bar ¼ 1 mm. (B) Peak force and total energy needed to detach the two layers of the OC composites measured by peel-off test. (C) Quantification of glycosaminoglycans (GAG) normalized for the wet weight accumulated in the chondral layers delaminated following peel-off test of OC composites. Values in B and C are mean SD of measurements obtained from 3 different constructs generated from AC and NC of a single donor. *p < 0.05 from AC based constructs.

JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2015

1114

BARANDUN ET AL.

from the acquired images and average pre- and post-contrast T1 values were obtained for each sample by drawing a region of interest in a slice passing through the middle of the samples. The corresponding concentration of Gd-DTPA ([Gd]) was calculated from the T1 values by assuming a relaxivity (R) of 4.59 mM1 s1, and GAG concentration was estimated using the following formulas: 1 1 1 R T1post T1pre pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ½Nabath ½Gdtissue ½Nabath ½Gdtissue pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi FCD ¼ ½Gdbath ½Gdtissue 502:5 ½GAGtissue ¼ FCD ð2Þ ½Gd ¼

The concentration of sodium was assumed to be 154 mM, T1post and T1pre are the T1 values measured after and before contrast agent addition to the culture medium, and FCD is the tissue fixed charge density.

ANALYTICAL ASSAYS Biochemistry Chondral and bone layers, delaminated during peel-off tests of OC constructs or harvested after the dGEMRIC analysis were digested for 16 h at 56 ˚C with 1 or 2 ml protease K (1 mg/ml protease K in 50 mM Tris with 1 mM EDTA, 1 mM iodoacetamide and 10 mg/ml pepstatin-A, respectively). GAG amounts were measured spectrophotometrically after reaction with dimethylmethylene blue, with chondroitin sulfate as a standard.22 GAG contents were reported as GAG/wet weight tissues. DNA was measured spectrofluorometrically using the CyQuant cell proliferation assay Kit (Molecular Probes, Eugene, OR), with calf thymus DNA as a standard.23 GAG contents were reported as GAG/wet weight tissue or GAG/DNA. Histology and Immunohistochemistry The OC constructs were fixed in 4% formalin and decalcified with 7% EDTA solution (Sigma). The EDTA decalcification was supported by ultrasound and carried out at a constant temperature of 30 ˚C. The solution was changed twice a day for a total time of 3– 4 days. Decalcified samples were then mbedded in paraffin. Sections (5 mm thick) were stained with Safranin-O and Toluidine blue for GAG. Sections were also processed for immunohistochemistry using antibodies against type I collagen (Quartett Immunodiagnostika und Biotechnologie GmbH, Berlin, Germany) and type II collagen (MP Biomedicals, Illkirch, France) as previously described.15 Statistical Analysis Statistical evaluation was performed using SPSS software version 7.5 software (SPSS, Sigma Stat, Erkrath, Germany). Values are presented as mean standard deviation (SD). Differences between experimental groups were assessed by the Mann—Whitney test. Correlations between dGEMRIC- and biochemically measured GAG were assessed using two-tailed Pearson’s tests. P values JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2015

less than 0.05 were considered to indicate statistically significant differences or correlations.

RESULTS Characterization of OC Grafts Generated with AC and NC 1 A 3 days’ preculture of NC- or AC on Chondro-Gide membrane did not result in evident matrix contraction, thus making relatively simple a precise shape matching of the cartilage layers with the upper bone layers. The cartilage layers of the OC composites generated by NC were glossier and thicker than those generated by AC, especially in the constructs cultured in the absence of TGFb-1 (Fig. 1A). The thickness of the chondral layers reflected the amount of GAG accumulated in these compartments. Histological analysis indicated a stronger Safranin O staining in the cartilage layers of NC-constructs compared to ACconstructs (especially when the CM was not supplemented with TGFb-1) and a trend toward a reduction of Safranin O staining from the edges to the center of the tissue samples. Importantly, the cartilaginous matrix was not only confined to the chondral layers but also extended to the cartilage/bone interfaces (Fig. 1A). The peel-off test indicated that integration forces were significantly higher in OC constructs generated by NC (in CM without TGFb-1) as compared to those generated by AC (1.7- and 1.6-fold respectively for peak force and total energy of peel off) (Fig. 1B). Consistent with the histological staining intensity and pattern, GAG amounts accumulated in the chondral layer of NC based OC constructs were remarkably and significantly higher than those measured in the AC based OC constructs (up to twofold) (Fig.1C). Overall the results of these initial investigations indicated that, as compared to AC, NC produced larger amounts of cartilaginous matrix and therefore generated more developed and more efficiently integrated OC grafts. The performance of NC was superior in chondrogenic medium not containing TGFb-1. Thus, for the next of our investigations described below, grafts generated with NC in medium without TGFb-1 were used. Characterization of the Responses of OC Grafts to IL-1b Exposure to IL-1b resulted in a loss of cartilaginous matrix as evidenced by a reduction in the intensity of staining for GAG (Safranin O and Toluidine blue) and type II collagen in IL-1b treated (IL-1) versus untreated (ctr) samples. Following IL-1b withdrawal, NC started to accumulate new cartilaginous matrix, so that after additional 2 weeks of culture in CM, the intensity of staining for GAG and type II collagen in the recovery groups approached that of the control groups (Fig. 2A). Interestingly, the intensity of staining for type I collagen was faint in the cartilage layers of the OC but became stronger at the cartilage/bone interfaces (Fig. 2A), in agreement with the capacity of


1115

Figure 2. Characterization of the responses of NC based OC composites to interleukin-1b (IL-1b. (A) Representative Tolouid blue (upper images), Safranin O, type II collagen (bottom left and right images respectively, corresponding to the white dotted area) and type I collagen staining of NC based OC composites cultured for a total of 4 weeks in chondrogenic medium without TGFb-1. Constructs were cultured for the last 3 days with (IL-1) or without (ctr) IL-1b and then maintained for an additional 2 weeks without IL-1b (recovery). Black dotted lines indicate the bone/cartilage interface. c ¼ chondral layer, b ¼ bone layer *cell occlusive layer of the Chondro-Gide1 membrane. Scale bar ¼ 100 mm. (B) Peak force and total energy needed to detach the two layers of the OC composites measured by peel-off test normalized to ctr without IL-1b treatment. (C) Quantification of glycosaminoglycans (GAG) and DNA normalized for the wet weight accumulated in the chondral layers delaminated following peel-off test of OC composites. Values in B and C are mean SD of measurements obtained from three experiments with NC from three different donors (with a least 3 replicates/ experimental group). *p < 0.05 from ctr. JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2015

1116

BARANDUN ET AL.

Figure 3. Correlation between GAG contents measured with delayed gadolinium-enhanced magnetic resonance imaging of cartilage (dGEMRIC) or measured biochemically (A) OC composites were fixed onto a customized silicone mold (5 constructs/mold) for the dGEMRIC analyses. OC constructs were generated by gluing a disk of native articular cartilage (I, III and V) or NC seeded ChondroGide1 matrices (II and IV) on the top of Tutobone1 blocks. (B) Post-contrast T1 map of the samples used for the drawing of ROIs and the calculation of GAG concentration through dGEMRIC. (C) Values of GAG concentration and GAG contents measured by dGEMRIC in OC or biochemically in the corresponding harvested chondral layers. OC composites were generated using native articular cartilage tissues (Native) or engineered cartilage using NC and expose or not to IL-1b[Engineered (IL-1) and Engineered (ctr)] or Chondro-Gide1 membrane (Collagen sponge). Values are mean SD of measurements obtained from two experiments with NC from one donor and native articular cartilage from two cadavers (with 3 replicates/experimental group). *p < 0.05 from Native, ˚p < 0.05 from ctr. n.d.: below the limit of detection. (D) Correlation between dGEMRIC- and biochemically measured GAG.

the NC to adapt their properties according to the environment to which they are exposed.7 In parallel to the modulation of cartilaginous matrix components, integration forces significantly decreased following IL-1b exposure (1.5- and 1.7-fold respectively for peak force and total energy of integration). However, they remained more than threefold higher than those previously measured in acellular OC constructs generated using fibrin glue [15] and after the recovery JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2015

time were similar to those measured in ctr samples (Fig. 2B, Table 2). Biochemical results showed that GAG amounts accumulated in the chondral layers of ctr samples were significant higher than those of IL-1 samples (1.5-fold) but similar to those of recovered samples. In contrast, DNA contents were comparable in the different groups (Fig. 2C, Table 3). GAG and DNA amounts accumulated in the bone layers did not


Table 1. Glycosaminoglycan (GAG) and DNA contents on the bone layer of osteochondral scaffolds (data are mean standard deviation of three replicate specimens from one donor

ctr IL-1 Recovery

GAG/construct (g)

DNA/construct (g)

250 25 185 32 275 28

11.1 2.4 10.5 3.2 12.1 4.2

significantly differ among the experimental groups (Table 1). Correlation Between dGEMRIC and Biochemically Measured GAG Contents The MRI scans of the samples yielded T1 maps with good homogeneity and the cartilaginous layers of the constructs were identifiable on top of the bone scaffolds as features with acceptable contrast (Fig. 3B). Trends in dGEMRIC and biochemical analyzes of the OC constructs indicated lower GAG contents in the chondral layers composed of engineered cartilage as compared to native cartilage and reduced GAG contents following IL-1b. However, a statistically significant difference in the GAG contents between ctr and IL-1 groups was observed only in the biochemically measured GAG (Fig.1C). We observed a positive (R2 ¼ 0.814) and statistically significant (p < 0.01) correlation between the biochemically- and dGMERICmeasured GAG contents (Fig.1D).

DISCUSSION In this study we demonstrated that the use of nasal chondrocytes (NC) as opposed to articular chondro-

1117

cytes (AC) resulted in the generation of osteochondral (OC) composites with a chondral layers more developed and more strongly integrated with the corresponding bone layers. NC-based OC constructs recovered well after IL-1b treatment by re-establishing cartilage matrix contents and cartilage/bone integration forces to levels measured before exposure to this inflammatory cytokine. The dGEMRIC method allowed to accurately estimate the GAG contents in the cartilage layers of the OC composites with a clinical compliant and non-destructive procedure. Expanded human NC have been previously shown to better re-differentiate than AC when cultured in three-dimensional pellets or on different scaffolds.11,12,16 Here, we showed that, when comparing the two cell types, NC can accumulate more abundant cartilage matrix even when combined with a bone substrate. In addition, the extracellular matrix (ECM) produced by NC resulted in a stronger cartilage/bone integration. Interestingly, these two properties that are essential for the fabrication of OC constructs starting from separate layers, were further enhanced when culturing NC in medium without TGFb-1. The response of NC-based cartilaginous tissue to IL-1b has been analyzed before.16 In the current study, we extended the investigation to the effect of this strong inflammatory factor on the cartilage/bone integration. Our results show that ECM in the cartilage/bone interface and consequently integration forces were transiently reduced after IL-1b exposure but promptly restored following a relatively short recovery time. This suggests that OC composites generated with NC could have favorable chances to successfully engraft into the joint and regenerate the osteochondral surface. Our protocol consisting of static seeding and

Table 2. Peak force and total energy needed to detach the two layers of the OC composites measured by peel-off test normalized to ctr without IL-1b treatment (data are mean standard deviation of 3–5 replicate specimens from three experiments with NC from three different donors) Peak force

Donor 1 Donor 2 Donor 3

Energy

ctr

IL-1

recovery

ctr

IL-1

recovery

1.00 1.00 1.00

0.71 0.14 0.75 0.12 0.57 0.09

0.90 0.21 1.02 0.33 0.96 0.22

1.00 1.00 1.00

0.61 0.16 0.77 0.24 0.43 0.12

0.98 0.25 0.85 0.07 1.03 0.11

Table 3. Glycosaminoglycan (GAG) and DNA contents on the cartilage layer of osteochondral scaffolds normalized to the ctr without IL-1b treatment (data are mean standard deviation of 3–5 replicate specimens from three experiments with NC from three different donors) GAG/wet weight

Donor 1 Donor 2 Donor 3

DNA/wet weight

ctr

IL-1

recovery

ctr

IL-1

recovery

1.00 1.00 1.00

0.67 0.16 0.61 0.35 0.71 0.16

1.22 0.21 1.03 0.18 1.15 0.12

1.00 1.00 1.00

1.12 0.24 1.05 0.23 0.75 0.11

0.97 0.12 1.75 0.14 0.74 0.26


1118

BARANDUN ET AL.

culturing of NC led to a suboptimal cartilage matrix spatial distribution. The implementation of a perfusion bioreactor system24 might overcome such issues. Additionally, under the described experimental conditions, the GAG content in the OC composites was still far from that of native cartilage. Considering the cell source (expanded/dedifferentiated adult chondrocytes) and the relatively short culture time (4 weeks) used in our study, this had to be expected. However, the implantation of rather immature as opposed to fully mature OC grafts could allow for a better integration with adjacent articular cartilage25 and result in enhanced osteochondral repair.26 In our study the cartilaginous layer was combined to a cell-free bone scaffold. Ongoing studies in our lab are aimed at investigating whether the presence of mesenchymal progenitor cells (naturally occurring in the bleeding subchondral bone during implantation) may alter the functional stability of the NC-derived osteochondral grafts. dGEMRIC was used in the current study to monitor in a non-destructive manner the GAG contents of the cartilage layers in in vitro fabricated OC tissues. To our knowledge this is the first time that engineered OC composite tissues were assessed with dGEMRIC. In a previous study27 dGEMRIC was used to investigate GAG distribution within engineered cartilaginous tissue generated with rabbit chondrocytes. In their study, the authors showed that dGEMRIC allowed not only to detect differences in GAG concentration of tissues with different maturation stages (i.e., samples culture for 3 or 6 weeks) but also to reveal variation in GAG spatial distribution in the samples. In our study the dGEMRIC-measured GAG loss in response to IL1b treatment was not statistically significant, even though it has been previously shown that this technique allowed to quantify a decrease of GAG concentration in cartilage samples following IL-1b exposure.19 However, the results of Bashir et al. cannot directly be compared to those described here due to the large differences in the experimental conditions among the two studies. These include the type of tissue analyzed (native cartilage specimens from young calves as opposed to human based engineered cartilaginous tissues), the protocol of IL-1b exposure (20 ng/ml of IL-1b over a 6 days vs. 10 ng/ml of IL-1b for 3 days), and the field strength of the magnet used (8.45 Tesla vs. 3 Tesla). Significant differences in GAG concentrations between ctr and IL-1 samples could be found using a high resolution imaging and increasing the sample size under investigation. In conclusion, we demonstrated that NC can be used to engineer OC composites consisting of a relatively mature cartilage layer, efficiently integrated with the bone scaffold layer through the extracellular matrix produced by the cells and with properties compatible with their engraftment in an inflammatory post-surgery joint environment. The described model JOURNAL OF ORTHOPAEDIC RESEARCH AUGUST 2015

is based on (i) a highly chondrogenic source of cells which can be harvested under limited donor site morbidity and safely be used for the repair of articular cartilage defects,13 (ii) biomaterials currently in clinical use, (iii) a medium free of growth factors and (iv) non-invasive monitoring of quality. These features may facilitate embedding within a regulatory framework of combined advanced therapy medicinal products for a pilot clinical test.

AUTHORS’ CONTRIBUTIONS MB, LDI, FS have drafted of the paper, acquired, analyzed and interrelated the data; MP, CS, DB, DW have acquired, analyzed and interpreted the data; OB, US, MH, MJ, DJS, IM have critically revised the manuscript; AB: has designed the study and critically revised the manuscript. All authors have read and approved the final submitted manuscript.

ACKNOWLEDGEMENTS We are grateful to Dr. med. M. Mumme for the harvest of cartilage biopsies, and to Beat G€ opfert, MEng, EMBA, for constructing the specimen holder used in the peel-off test. We also thank M. Eloy (Geistlich 1 Pharma AG) for providing Chondro-Gide and to S. 1 Bischofberger for providing Tutobone . This work was financed by the Swiss National Science Foundation (SNF Project No 310030-126965.1).

REFERENCES 1. Gelber AC,Hochberg MC, Mead LA, et al. 2000. Joint injury in young adults and risk for subsequent knee and hip osteoarthritis. Ann Intern Med 133:321–328. 2. Prakash D, Learmonth D. 2002. Natural progression of osteo-chondral defect in the femoral condyle. Knee 9:7–10. 3. Furman BD, Olson SA, Guilak F. 2006. The development of posttraumatic arthritis after articular fracture. J Orthop Trauma 20:719–725. 4. Ding C, Cicuttini F, Jones G. 2007. Tibial subchondral bone size and knee cartilage defects: relevance to knee osteoarthritis. Osteoarthritis Cartilage 15:479–486. 5. Gomoll AH, Filardo G, de Girolamo L, et al. 2012. Surgical treatment for early osteoarthritis. Part I: cartilage repair procedures. Knee Surg Sports Traumatol Arthrosc 20:450–466. 6. Martin I, Miot S, Barbero A, et al. 2007. Osteochondral tissue engineering. J Biomech 40:750–765. 7. Pelttari K, Winter A, Steck E, et al. 2006. Premature induction of hypertrophy during in vitro chondrogenesis of human mesenchymal stem cells correlates with calcification and vascular invasion after ectopic transplantation in SCID mice. Arthritis Rheum 54:3254–3266. 8. Barbero A, Grogan S, Sch€ afer D, et al. 2004. Age related changes in human articular chondrocyte yield, proliferation and post-expansion chondrogenic capacity. Osteoarthritis Cartilage 12:476–484. 9. Rotter N, Bonassar LJ, Tobias G, et al. 2002. Age dependence of biochemical and biomechanical properties of tissueengineered human septal cartilage. Biomaterials 23: 3087–3094. 10. Kafienah W, Jakob M, Demarteau O, et al. 2002. Threedimensional tissue engineering of hyaline cartilage: comparison of adult nasal and articular chondrocytes. Tissue Eng 8:817–826. 11. Tay AG, Farhadi J, Suetterlin R, et al. 2004. Cell yield, proliferation, and postexpansion differentiation capacity of


12.

13.

14.

15.

16.

17.

18.

19.

human ear, nasal, and rib chondrocytes. Tissue Eng 10: 762–770. Candrian C, Vonwil D, Barbero A, et al. 2008. Engineered cartilage generated by nasal chondrocytes is responsive to physical forces resembling joint loading. Arthritis Rheum 58:197–208. Pelttari K, Pippenger B, Mumme M, et al. 2014. Adult human neural crest-derived cells for articular cartilage repair. Sci Transl Med 6:251ra 119. Lee CR, Grodzinsky AJ, Hsu HP, et al. 2000. Effects of harvest and selected cartilage repair procedures on the physical and biochemical properties of articular cartilage in the canine knee. J Orthop Res 18:790–799. Scotti C, Wirz D, Wolf F, et al. 2010. Engineering human cell-based, functionally integrated osteochondral grafts by biological bonding of engineered cartilage tissues to bony scaffolds. Biomaterials 31:2252–2259. Scotti C, Osmokrovic A, Wolf F, et al. 2012. Response of human engineered cartilage based on articular or nasal chondrocytes to interleukin-1b and low oxygen. Tissue Eng Part A 18:362–372. B€ urgin V, Daniels AU, Francioli S, et al. 2014. 90˚ Peel off Tests of Tissue Engineered Osteochondral Constructs: A New Method to Determine the Osteochondral Integration. Int J Tissue Engineering 2014:1–6. Bashir A, Gray ML, Hartke J, et al. 1999. Nondestructive imaging of human cartilage glycosaminoglycan concentration by MRI. Magn Reson Med 41:857–865. Bashir A, Gray ML, Burstein D. 1996. Gd-DTPA2- as a measure of cartilage degradation. Magn Reson Med 36: 665–673.

1119

20. Zheng S, Xia Y. 2010. The impact of the relaxivity definition on the quantitative measurement of glycosaminoglycans in cartilage by the MRI dGEMRIC method. Magn Reson Med 63:25–32. 21. Deoni SCL, Rutt BK, Peters TM. 2003. Rapid combined T1 and T2 mapping using gradient recalled acquisition in the steady state. Magn Reson Med 49:515–526. 22. Barbosa I, Garcia S, Barbier-Chassefiere V, et al. 2003. Improved and simple micro assay for sulfated glycosaminoglycans quantification in biological extracts and its use in skin and muscle tissue studies. Glycobiology 13:647–653. 23. Jakob M, Demarteau O, Sch€ afer D, et al. 2001. Specific growth factors during the expansion and redifferentiation of adult human articular chondrocytes enhance chondrogenesis and cartilaginous tissue formation in vitro. J Cell Biochem 81:368–377. 24. Wendt D, Marsano A, Jakob M, et al. 2003. Oscillating perfusion of cell suspensions through three-dimensional scaffolds enhances cell seeding efficiency and uniformity. Biotechnol Bioeng 84:205–214. 25. Khan IM, Gilbert SJ, Singhrao SK, et al. 2008. Cartilage integration: evaluation of the reasons for failure of integration during cartilage repair. A review. Eur Cell Mater 16:26–39. 26. Miot S, Brehm W, Dickinson S, et al. 2012. Influence of in vitro maturation of engineered cartilage on the outcome of osteochondral repair in a goat model. Eur Cell Mater 23:222–236. 27. Henderson JH, Welter JF, Mansour JM, et al. 2007. Cartilage tissue engineering for laryngotracheal reconstruction: comparison of chondrocytes from three anatomic locations in the rabbit. Tissue Eng 13:843–853.
