ARTHRITIS & RHEUMATISM Vol. 63, No. 11, November 2011, pp 3417–3427 DOI 10.1002/art.30543 © 2011, American College of Rheumatology
Fibroblast Growth Factor 2 and Transforming Growth Factor 1 Induce Precocious Maturation of Articular Cartilage Ilyas M. Khan,1 Sam L. Evans,1 Robert D. Young,1 Emma J. Blain,1 Andrew J. Quantock,1 Nick Avery,2 and Charles W. Archer1 the treatment of diseased cartilage, through phenotype modulation of osteoarthritic chondrocytes in order to stimulate growth and maturation of cartilage repair tissue.
Objective. We have discovered that a combination of fibroblast growth factor 2 and transforming growth factor 1 induce profound morphologic changes in immature articular cartilage. The purpose of this study was to test the hypothesis that these changes represent accelerated postnatal maturation. Methods. Histochemical and biochemical assays were used to confirm the nature of the morphologic changes that accompany growth factor stimulation of immature bovine articular cartilage explants in serumfree culture medium. Growth factor–induced apoptosis, cellular proliferation, and changes in the collagen network were also quantitatively analyzed. Results. Growth factor stimulation resulted in rapid resorption from the basal aspect of immature cartilage explants that was simultaneously opposed by cellular proliferation from the apical aspect driven from a pool of chondroprogenitor cells we have previously described. Maturation-dependent changes in tissue stiffness, collagen crosslinking, and collagen fibril architecture as well as differentiation of the extracellular matrix into distinct pericellular, territorial, and interterritorial domains were all present in growth factor– stimulated cartilage samples and absent in control samples. Conclusion. Our data demonstrate that it is possible to significantly enhance the maturation of cartilage tissue using specific growth factor stimulation. This may have applications in transplantation therapy or in
Articular cartilage is a highly specialized connective tissue that lines the surfaces of bones in synovial joints and provides a low-friction, hydrodynamic, loadbearing surface for articulation about the joint. It is composed principally of large aggregating proteoglycans that are constrained by a collagen network and maintained at a low cell-to–extracellular matrix (ECM) ratio by resident chondrocytes (1). At birth, articular cartilage is relatively amorphous; chondrocytes are present at high density and are randomly distributed throughout the depth of the tissue (2). In a series of developmental transitions that lead to sexual maturity and adulthood, articular cartilage undergoes significant structural, compositional, and morphologic changes in response to the biomechanical demands placed on the tissue (3,4). In the months following birth, the articular cartilage develops a distinct zonal stratification that is a hallmark of maturation. Chondrocytes become arranged in 4 layers: a thin superficial zone with flattened cells, a transitional layer with rounded cells, a thicker deep layer where chondrocytes are aligned in columns, and a calcified zone, which interdigitates between the hyaline cartilage and the subchondral bone plate (2). Concomitant with these morphologic changes is a reorganization of the collagen network, leading to alignment of chondrocytes in the directional plane of collagen fibrils (5). Furthermore, the ECM is concentrically organized from the cell outward into pericellular, territorial, and interterritorial matrices, each of which has specific biochemical and biomechanical functions (6). While the mechanism of postnatal maturation of articular cartilage is hypothesized to be a gradual pro-
Supported by Arthritis Research UK (grant 18871). 1 Ilyas M. Khan, PhD, Sam L. Evans, PhD, Robert D. Young, PhD, Emma J. Blain, PhD, Andrew J. Quantock, PhD, Charles W. Archer, PhD: Cardiff University, Cardiff, UK; 2Nick Avery, PhD: University of Bristol, Bristol, UK. Address correspondence to Ilyas M. Khan, PhD, School of Biosciences, Cardiff University, Museum Avenue, Cardiff CF10 3AX, UK. E-mail:
[email protected]. Submitted for publication November 6, 2010; accepted in revised form July 5, 2011. 3417
3418
KHAN ET AL
cess of resorption from below and appositional growth from the surface of the tissue (4) directed by tissuespecific stem cells (7,8), identification of the biochemical stimuli that induce these progressive changes has remained elusive. Identification of these stimuli is important in two main respects. First, it would allow molecular dissection of the pathways that regulate this important developmental transition. Second, from a practical perspective, the ability to initiate or accelerate cartilage maturation would have significant clinical relevance. Implantation of chondrocytes (9) or mesenchymal stem cells (10) into cartilage defects generates, in the first instance, repair tissue that resembles in some respects immature cartilage (i.e., a random distribution of chondrocytes within an undifferentiated ECM) (11). Over time, measured in years, the repair tissue can in some cases mature to form hyaline cartilage and demonstrate zonal stratification (11). The ability to stimulate or accelerate maturation of implanted cells or the ability to precondition engineered cartilage in vitro would be a useful attribute for surgeons and bioengineers. In this study, we found that a combination of fibroblast growth factor 2 (FGF-2) and transforming growth factor 1 (TGF1) induces profound morphologic changes in immature articular cartilage consistent with a highly accelerated maturational response. Growth factor stimulation induced apoptosis and resorption from the basal aspect and cellular proliferation in surface chondrocytes. Most significantly, we found evidence of collagen remodeling, which manifested itself in changes in collagen crosslinking that led to increased mechanical stiffness and enhanced the formation of pericellular matrix around chondrocytes. MATERIALS AND METHODS In vitro culture and analysis of articular cartilage explants. Using 6-mm biopsy punches (Stiefel), cartilage explants were excised under sterile conditions from the lateral aspect of the medial condyle of the metacarpophalangeal (MCP) joints of immature male bovine calves and grouped in adjacent pairs for experimental analysis. A thin layer of subchondral bone was present on the base of each explant. Freshly explanted tissue was cultured in chemically defined medium, consisting of high-glucose Dulbecco’s modified Eagle’s medium (DMEM), 10 mM HEPES, pH 7.5, 50 g/ml of gentamicin, 50 g/ml of sodium ascorbate, and supplemented with insulin–transferrin–selenium (ITS; Sigma). Explants were cultured for 21 days with medium changes every fourth day. Analysis of sulfated glycosaminoglycan content was performed by dimethylmethylene blue assay of papaindigested extracts as previously described (12). Quantification
of collagen content was performed using an assay that measures the hydroxyproline of acid-hydrolyzed tissue as previously described (13). The DNA content of papain-digested explant extracts was determined using Hoechst dye. In situ and gel zymography. Explants were fixed in periodate–lysine–paraformaldehyde fixative (2% paraformaldehyde, 75 mM lysine, and 10 mM sodium periodate, pH 7.4) for 12 hours at 4°C and then processed as normal for histologic assessment. Eight-micron sections from explants were then overlaid with 100 g/ml DQ gelatin in 50 mM Tris/CaCl2 and incubated for 8 hours at 37°C in a humidified chamber. The gelatinolytic activity was observed as green labeling by fluorescence microscopy (Olympus BX61 fluorescence microscope). Negative controls were incubated without DQ gelatin. The expression and activation status of matrix metalloproteinase 2 (MMP-2) and MMP-9 were assessed using gelatin zymography as previously described (14). Relative quantities of proteolytic enzymes were analyzed by scanning densitometry (UMAX Systems) and NIH Image software (National Institute of Health; online at http://rsb.info.nih.gov/nih-image/) and normalized to the wet weight of the cartilage explants. TUNEL labeling for apoptosis. TUNEL assay for apoptosis was performed using a FragEL DNA fragmentation detection kit (Calbiochem) following the recommended protocol. Bromodeoxyuridine (BrdU) incorporation assays. Twenty-four hours prior to the end of incubation, BrdU was added to the culture medium to a concentration of 10 M. Sections were incubated for 1 hour in 1N HCl and then briefly in 0.1M borate buffer, pH 8.0, after which they were incubated for 24 hours at 4°C with 5 g/ml of antibody G3G4 (DSHB). Goat anti-mouse IgG Alexa Fluor 594 conjugate (Invitrogen) was used to identify bound primary antibodies. In vitro growth of superficial zone chondrocytes. The surface zone of articular cartilage from the MCP joints of 7-day-old male bovine calves was removed surgically, and the chondrocytes were extracted enzymatically by sequential digestion in Pronase and collagenase (8). Chondrocytes were cultured for 7 days in standard, chemically defined medium in the presence or absence of growth factors and then subjected to MTT assay for cell viability. Analysis of collagen crosslinks. Samples for collagen crosslink analyses were prepared as described elsewhere (15). Crosslink analysis was achieved using a Pickering sodium-form cation-exchange column (ARC Sciences), Dionex ICS 3000 hardware, a Pickering post-column reactor, and Dionex Ultimate 3000 variable-wavelength detector set at 570 nm. Crosslinks were verified using a chromatography program developed in-house and by comparison with authenticated standards also prepared in-house. Quantification used published values for leucine equivalence (16). Immunofluorescence detection of epitopes. Explants were flash frozen in n-hexane, and 8-m cryosections were prepared. Sections were incubated with rat antiperlecan or rabbit active caspase 3 primary antibody (ab44937 and ab13847, respectively; Abcam) at a concentration of 5 g/ml. Alexa Fluor–conjugated secondary antibodies were used to visualize the location of primary antibodies. Negative controls used preimmune sera and/or rat IgG.
ACCELERATING MATURATION IN BOVINE CARTILAGE EXPLANTS
Preparation of cartilage explants for electron microscopy. Explants were removed from culture, rinsed in serumfree medium, and fixed for 3 hours at room temperature in 2.5% glutaraldehyde/2% paraformaldehyde in 0.1M sodium cacodylate buffer, pH 7.3. After brief storage in buffer, explants were sliced into blocks of ⬃1 ⫻ 0.5 ⫻ 0.5 mm, postfixed in 1% aqueous osmium tetroxide, followed by 0.5% uranyl acetate, each for 1 hour, then dehydrated in ethanol and embedded in Araldite resin (Agar Scientific). Sections were contrasted with uranyl acetate and lead citrate before examination with a JEOL 1010 transmission electron microscope equipped with a Gatan Orius SC1000 CCD camera. Mechanical strength testing of cartilage explants. Young’s modulus [EY] for cartilage explants was determined by indentation testing using a 0.5 mm–diameter flat-ended cylindrical indenter. Specimens were tested using a Losenhausen servohydraulic testing machine with an MTS FlexTest GT controller and a 5N load cell (Interface). The samples were loaded at a constant speed of 0.1 mm/second. Forcedisplacement curves allowed the aggregate Young’s modulus of the cartilage samples to be determined, using the analytical solution described by Hayes et al (17), assuming the initial deformation was isochoric with a Poisson ratio of 0.5. Quantitative polymerase chain reaction (qPCR). Cartilage explants were frozen in liquid nitrogen and then homogenized in the presence of 1 ml of frozen TRI Reagent (Sigma) using a Mikro-Dismembrator U (B Braun Biotech), and RNA was extracted using RNeasy RNA extraction columns (Qiagen). The qPCR analysis was performed using GoTaq qPCR Master Mix (Promega), 12.5 ng of complementary DNA, and 0.3 mM forward and reverse primers (primer sequences available upon request from the author). Reactions were performed with a Stratagene Mx3000 real-time PCR analyzer (Agilent Technologies) using the following thermal cycling program: 95°C for 10 minutes for 1 cycle and then 95°C for 30 seconds, 55°C for 60 seconds, and 72°C for 30 seconds for 40 cycles. The data shown are the ratio of the concentration of the gene of interest (in nanograms) to 18S ribosomal RNA (in nanograms). Statistical analysis. Results are presented as mean ⫾ SD. All data sets were checked for normal distribution using the Shapiro-Wilk test and for homogeneity of variances using Levene’s test prior to parametric analysis. For analysis of 2 groups, we used Student’s paired-sample 1-tailed or 2-tailed t-test, and for multiple groups, a one-way analysis of variance test. Data sets were analyzed using PASW Statistics 18, release version 18.0.0 (SPSS).
RESULTS Induction of maturational changes in immature articular cartilage by FGF-2 and TGF1. Bovine articular cartilage derived from the perinatal MCP joint has a cellular and extracellular matrix morphology that has been described as immature (4). The tissue is thick (⬃2–3 mm), and chondrocytes within it are organized
3419
Figure 1. Postnatal developmental maturation of articular cartilage. Histologic sections of immature (7 days old) and mature (over 18 months old) cartilage samples taken from the metacarpophalangeal joint of male bovine calves were labeled with Safranin O to highlight proteoglycan content. Postnatal maturation results in a reduction of cartilage height, a decrease in cellular density, and changes in collagen fibril architecture (4,5). The bar at the left of each section delineates the superficial (top; open), transitional (middle; shaded), and deep (bottom; darkly shaded) zones of the cartilage samples.
isotropically compared to those in mature articular cartilage (Figure 1). We observed that when immature articular cartilage explants from the MCP joint were cultured ex vivo for 21 days in a chemically defined basal culture medium in the combined presence of growth factors FGF-2 (100 ng/ml) and TGF1 (10 ng/ml), there were significant changes in explant morphology (Figure 2A). We observed a 52% reduction (P ⬍ 0.01) in the height of explants cultured with growth factors compared to explants cultured in basal medium alone (mean ⫾ SD 2,725 ⫾ 327 m versus 1,435 ⫾ 292 m, respectively) (data available upon request from the author). Notably, the larger, more hypertrophic cells of the deep zone of control explants were absent in growth factor–treated explants. Biochemical analysis of FGF-2/TGF 1– stimulated explants showed that there was an ⬃30% decrease in glycosaminoglycan content (P ⬍ 0.05), but
3420
Figure 2. Induction of morphologic remodeling of immature articular cartilage by fibroblast growth factor 2 (FGF-2) and transforming growth factor 1 (TGF1). A, Paired cartilage explants were cultured for 21 days in insulin–transferrin–selenium (ITS) medium alone (control) or in ITS containing FGF-2 (100 ng/ml) plus TGF1 (10 ng/ml). Sections were stained with 0.1% toluidine blue. B, In situ zymography was performed on control (ITS) and growth factor–stimulated explants that had been overlaid with DQ gelatin. Highly fluorescent chondrocyte labeling can be observed projecting above the main resorption front in growth factor–treated explants (vertical bracket), whereas in control explants, the resorption front is smoother and the chondrocyte labeling less intense. Bar ⫽ 100 m. C, Polyacrylamide gel electrophoresis gelatin zymography of culture medium from control and growth factor–stimulated explants is shown at the top. A band for active matrix metalloproteinase 9 (MMP-9) (arrowhead) was observed only in the growth factor–treated samples. Densitometric analyses of the zymograms for levels of proMMP-9 and proMMP-2 are shown at the bottom. Values are the mean ⫾ SD of 4 samples per group.
no change in the hydroxyproline content normalized to the tissue wet weight as compared to unstimulated control explants (data available upon request from the author). Analysis of the DNA content normalized to the explant wet weight showed a decrease of ⬃25% (mean ⫾ SD 0.28 ⫾ 0.02 versus 0.21 ⫾ 0.03 g DNA per
KHAN ET AL
mg wet weight) in growth factor stimulated explants compared to controls (data available upon request from the author). The reductions in tissue height, proportion of proteoglycan, and the DNA content are hallmarks of articular cartilage maturation, and based on these preliminary observations, we hypothesized that FGF-2 and TGF1 accelerate postnatal maturation of articular cartilage. Induction of precocious resorption by growth factor stimulation. To visualize the active process of resorption, we overlaid DQ gelatin, a fluorescenceactivated collagenase substrate, over periodate–lysine– paraformaldehyde–fixed sections of explants (Figure 2B). A smooth boundary between calcified cartilage and deep-zone hyaline cartilage was delineated by enzymatically activated fluorescent DQ gelatin in control explants. In contrast, the same boundary was broken and ragged in growth factor–treated explants, with highly fluorescent labeling surrounding groups of cells in the hyaline portion of the cartilage. Polyacrylamide gel electrophoresis gelatin zymography of culture medium samples showed increased activity of proMMP-9 (10.7fold; P ⬍ 0.013) and proMMP-2 (2.25-fold: P ⬍ 0.026) in growth factor–treated explant medium compared to controls (Figure 2C). Basal expression of type X collagen was maintained in control and growth factor–treated explants, and tartrate-resistant acid phosphatase activity was not present within the deep zones of control or growth factor–treated explants, thus indicating a possible epiphyseal origin for the resorptive activity (data available upon request from the author). High-power microscopy also revealed the presence of condensed nuclei in the deep zone of growth factor–stimulated explants, providing partial evidence of apoptosis. TUNEL analysis (Figure 3A) revealed that apoptosis was significantly increased by 6-fold (P ⬍ 0.01) in the deep and middle zones of growth factor–treated explants as compared to control explants, where apoptosis was sparse and sporadic (mean ⫾ SD ratio of TUNEL-positive cells to total cells 0.186 ⫾ 0.01 versus 0.03 ⫾ 0.007) (Figure 3B). Of the TUNEL-negative chondrocytes in the upper deep zone of growth factor– stimulated explants, many were immunopositive for labeling with antibodies specific for the active isoform (17 kd) of caspase 3 (Figure 3C) and were therefore in the pathway of apoptosis. Caspase 3–positive chondrocytes were confined to the margins of the deep and calcified zones in control cartilage samples. Induction of a switch in the polarity of growth by growth factor stimulation. The surface of FGF-2/ TGF1–stimulated cartilage explants was densely populated with chondrocytes as compared to that of control
ACCELERATING MATURATION IN BOVINE CARTILAGE EXPLANTS
3421
Figure 3. Induction of apoptosis in immature articular cartilage by fibroblast growth factor 2 (FGF-2) and transforming growth factor 1 (TGF1). A, TUNEL analysis of immature cartilage explants treated with insulin–transferrin–selenium (ITS) (control) or with ITS containing FGF-2 plus TGF1. Chondrocytes within the upper and lower deep zone of growth factor–treated articular cartilage were susceptible to apoptosis (green and orange). Nuclei were counterstained with propidium iodide (red). Bars ⫽ 100 m. B, Quantification of the extent of apoptosis in control and growth factor–treated cartilage explants. Values are the mean ⫾ SD of 4 samples per group. C, Anti–active caspase 3 antibody staining of control and growth factor–treated articular cartilage samples. Not all of the chondrocytes in the upper deep zone of growth factor–stimulated cartilage samples were TUNEL-positive, but the majority of these cells were positive for labeling with anti–active caspase 3 antibodies, whereas chondrocytes from the same region of control cartilage were negative for labeling. Bars 20 m.
explants (Figure 4A). We performed cell counts to quantify this and observed a 39% increase in cellular density in treated explants (mean ⫾ SD 234.38 ⫾ 33.10 cells per microscopic field versus 169.13 ⫾ 34.06 cells per microscopic field) (Figure 4B). To eliminate the possibility that the increase in cellular density in the surface zone may have been due to dehydration or loss of ECM, we performed BrdU incorporation assays to directly visualize dividing chondrocytes (Figure 4C). In unstimulated explants, BrdU incorporation was present sporadically in the superficial layer but was more prominent within the deeper interstitial layer. Upon growth factor stimulation, the polarity of cell division was switched, whereby the majority of chondrocytes displaying BrdU incorporation were localized to the superficial layer. In addition, in vitro analysis of cellular viability of chondrocytes isolated from the surface zone and cultured in the presence or absence of FGF-2, TGF1, or FGF-2 plus TGF1 for 7 days (data available upon request from the author), showed that the combination of growth factors synergistically induced cellular proliferation by ⬃36% (P ⬍ 0.05) over that found in the absence or presence of either growth factor alone (mean ⫾ SD percentage cell viability 100.0 ⫾ 9.5% in control cells, 104.0 ⫾ 13.5% in FGF-
2–treated cells, 105.9 ⫾ 7.4% in TGF1-treated cells, and 135.8 ⫾ 17.9% in FGF-2 plus TGF1–treated cells). Next, we performed quantitative PCR analysis of groups of proteins that are differentially regulated during the postnatal maturational process in articular cartilage (18) (Table 1). We saw an expected decrease in type IIB and type IX collagen gene expression, a 6-fold increase in type I collagen expression, no significant change in aggrecan or ADAMTS-5 expression, but a 20-fold increase in ADAMTS-4 transcript levels. Interstitial collagenase MMP-1 was significantly up-regulated in growth factor–treated explants, again an expected outcome, as was the up-regulation of MMPs 2, 9, and 13. Gene expression levels of tissue inhibitor of metalloproteinases 1 (TIMP-1), TIMP-2, and TIMP-3 were also up-regulated, with the greatest increase occurring for TIMP-1 (⬃64-fold). The transcript levels of cartilage oligomeric matrix protein (COMP), a marker of abnormal turnover in articular cartilage, decreased ⬃50-fold in our system, indicating that the catabolic changes we observed are distinct from those observed during pathologic transitions, where COMP transcript levels rise (19). Maturational changes accompanied by remodeling of collagen. One of the hallmarks of articular cartilage maturation is an increase in collagen crosslinking,
3422
KHAN ET AL
Table 1. Quantitative polymerase chain reaction analysis of gene expression in control and growth factor–treated immature bovine cartilage explants* ITS/FGF-2/ TGF1
ITS COL1A1 COL2A1 COL9A1 COMP Aggrecan ADAMTS-4 ADAMTS-5 MMP-1 MMP-2 MMP-9 MMP-13 TIMP-1 TIMP-2 TIMP-3
⫺3
⫺3
4.22 ⫾ 1.7 7.06⫺2 ⫾ 6.1⫺3 3.80⫺3 ⫾ 3.5⫺4 3.00⫺3 ⫾ 6.1⫺4 1.25⫺2 ⫾ 1.4⫺3 2.56⫺5 ⫾ 5.7⫺6 1.00⫺3 ⫾ 6.0⫺5 1.03⫺7 ⫾ 4.4⫺9 9.00⫺4 ⫾ 9.0⫺5 2.26⫺7 ⫾ 8.5⫺8 2.00⫺4 ⫾ 4.0⫺5 3.10⫺3 ⫾ 2.2⫺4 1.10⫺3 ⫾ 6.0⫺5 6.00⫺4 ⫾ 4.0⫺5
⫺2
⫺2
2.60 ⫾ 1.2 7.40⫺3 ⫾ 2.9⫺3 3.00⫺4 ⫾ 1.2⫺4 1.00⫺4 ⫾ 1.0⫺5 8.20⫺3 ⫾ 3.7⫺3 5.23⫺4 ⫾ 1.2⫺4 2.30⫺3 ⫾ 9.1⫺4 3.82⫺4 ⫾ 1.4⫺4 1.09⫺2 ⫾ 4.0⫺3 3.67⫺5 ⫾ 1.3⫺5 1.10⫺3 ⫾ 3.6⫺4 1.99⫺1 ⫾ 6.2⫺2 1.01⫺2 ⫾ 3.0⫺3 2.80⫺3 ⫾ 7.8⫺4
Fold difference
P
6.3 0.11 0.07 0.02 0.4 20.4 2.3 3,708 12 162 5.5 64 9.2 4.7
0.04 0.0005 0.001 0.009 NS 0.01 NS 0.01 0.04 0.03 0.04 0.025 0.028 0.038
* Values are the mean ⫾ SD ratio of product (in nanograms) normalized to 18S ribosomal RNA (in nanograms) for each gene derived from cDNA generated from growth factor–stimulated explants (insulin– transferrin–selenium [ITS]/fibroblast growth factor 2 [FGF-2]/ transforming growth factor 1 [TGF1]) versus control explants (ITS) (n ⫽ 4 per group). Expression of types II and IX collagen decreased ⬃9-fold and 14-fold, respectively, and there was no significant difference in aggrecan gene synthesis between growth factor–treated and untreated explants. COMP ⫽ cartilage oligomeric matrix protein; NS ⫽ not significant; MMP-1 ⫽ matrix metalloproteinase 1; TIMP1 ⫽ tissue inhibitor of metalloproteinases 1.
Figure 4. Induction of appositional growth in immature articular cartilage by fibroblast growth factor 2 (FGF-2) and transforming growth factor 1 (TGF1). A, Cellularity in the surface zone of cartilage explants treated with insulin–transferrin–selenium (ITS) (control) or with ITS containing FGF-2 plus TGF1. Fluorescence microscopy of DAPIstained (blue) nuclei shows increased cellularity in the surface zone of growth factor–stimulated explants. Arrows indicate doublet cells, which are absent in untreated explants. Bars ⫽ 50 m. B, Quantification of cell numbers in paired untreated and growth factor–stimulated cartilage explants per unit area. Values are the mean ⫾ SD of 8 samples per group. C, Analysis of bromodeoxyuridine (BrdU) incorporation into control and growth factor–stimulated paired immature cartilage explants. Two asymmetric growth zones (differential interference contrast images merged with fluorescent BrdU-positive nuclei [red]), interstitial (lower vertical bracket) and appositional (upper vertical bracket), can be seen in the control section. The polarity of these zones switches following growth factor treatment. Bars ⫽ 50 m.
specifically an increase in the ratio of mature trivalent lysylpyridinoline and hydroxylysylpyridinoline crosslinks to immature divalent hydroxylysinoketonorleucine (HLKNL) crosslinks in collagen fibrils (20), which is catalyzed by the enzyme lysyl oxidase. Using highperformance liquid chromatography, we analyzed the quantity of immature and mature collagen crosslinks formed in cartilage explants. A significant reduction in the number of immature (HLKNL) collagen crosslinks was found in growth factor–stimulated explants, which translated as a significantly greater ratio (⬃68%) of mature to immature collagen crosslinks as compared to that in control explants (Figure 5A). We hypothesized that the increase in mature collagen crosslinks in growth factor–stimulated explants would increase their stiffness, and this was analyzed using indentation testing (Figure 5B). Growth factor– treated explants displayed average values for Young’s modulus (EY) that were 229% greater than those in control explants (mean ⫾ SD 4.98 ⫾ 0.45 MPa versus 2.17 ⫾ 0.60 MPa). Using electron microscopy, we tried to visualize possible changes in ECM remodeling in explants (Figure 5C). In control articular cartilage, surface-zone cells had no discernible pericellular matrix, and membranous processes projected into the ECM; however, in growth factor–stimulated explants, chondrocytes along the surface of the explant were more
ACCELERATING MATURATION IN BOVINE CARTILAGE EXPLANTS
3423
Figure 5. Promotion of collagen crosslinking, increased mechanical strength, and pericellular coat formation by growth factor stimulation of immature articular cartilage. A, Amino acid analysis of immature divalent hydroxylysinoketonorleucine (HLKNL) and mature trivalent lysylpyridinoline (LP) or hydroxylysylpyridinoline (HP) collagen crosslinks. Growth factor treatment of immature explants induced a significant reduction in immature crosslinks, coupled with an increase in the ratio of mature to immature crosslinks. Results are expressed as moles per mole of collagen. Values are the mean ⫾ SD of 5 samples per group. B, Analysis of Young’s modulus of cartilage explants treated with insulin–transferrin–selenium (ITS) (control) or with ITS containing fibroblast growth factor 2 (FGF-2) plus transforming growth factor 1 (TGF1) following 21 days of culture. Values are the mean ⫾ SEM of 5 samples per group. C, Electron microscopy of surface-zone chondrocytes in growth factor–stimulated explants. Growth factor–treated chondrocytes developed pericellular (thickened rim of cell), territorial (area surrounding arrowheads), and interterritorial (ⴱ) matrices. The chondron and pericellular coat are undefined in chondrocytes of control explants. Arrows indicate membrane contacts with the extracellular matrix. Original magnification ⫻ 1,200. D, Perlecan antibody labeling of control and growth factor–stimulated frozen tissue sections of cartilage explants. Arrowheads denote the major distributions of perlecan labeling in the extracellular matrix. Original magnification ⫻ 400.
rounded, and a pericellular thickening around the cells was apparent, as well as differentiation of the surrounding ECM into distinct territorial and interterritorial compartments. In addition, the ECM was significantly more electron-dense, indicating increased density of collagen fibrils. We used antibodies directed against perlecan (21) to further probe ECM remodeling and differentiation. Perlecan is diffusely localized to the interterritorial and territorial ECM in articular cartilage obtained from perinatal immature tissue and is found predominantly in the pericellular matrix surrounding chondrocytes in mature cartilage (22). In control cartilage explants, perlecan labeling was found throughout the ECM and in chondrocytes; however, in growth factor–stimulated explants a distinct pericellular ring of labeling surrounding individual chondrocytes was apparent (Figure 5D). DISCUSSION Postnatal articular cartilage maturation is a key developmental process that adapts synovial joints to
increasing biomechanical stress and strain during growth (23–25). In rabbits, the major changes that lead to structural maturity of articular cartilage occur within 3 months of birth, with an early rapid phase of growth and remodeling—the end of which is coincident with sexual maturity—leading onto slower growth and skeletal maturity at ⬃8 months of age (4). The process of cartilage maturation involves resorption and growth concomitant with changes in the collagen fibril network. The findings of this study demonstrate that continuous exposure to FGF-2 and TGF1 induce prolific morphologic and molecular changes in immature cartilage that are consistent with accelerated postnatal maturation. FGF-2 has been shown to be mitogenic for articular chondrocytes (26); it stimulates and maintains their differentiation capacity in vitro (27,28) and has been reported to improve the regeneration of cartilage defects in vivo (29–32). TGF1 is also a developmentally regulated cytokine that is expressed early in limb development (33), and its expression progressively increases postnatally in articular cartilage (34). Studies have
3424
shown that TGF1 stimulates proteoglycan synthesis during in vitro organ culture of articular cartilage (35), but that its effects are largely homeostatic (36). The combination of FGF-2 and TGF1 has previously been shown to increase cellularity and the capacity of articular chondrocytes to redifferentiate during in vitro pellet culture (37). We observed that the combination of FGF-2 and TGF1 induced profound morphologic changes in immature bovine articular cartilage during organ culture. There was a significant decrease in height that was attributed to resorption. We hypothesize that the latter process is primarily driven by epiphysealderived chondrocytes that form a continuum—the chondroepiphyseal growth plate—with articular chondrocytes in immature articular cartilage. In situ gelatin zymography identified fluorescence-labeled deep-zone chondrocytes as the primary sites of gelatinolytic activity, while the findings of tartrate-resistant acid phosphatase assays for osteoclast activity were negative. Hypertrophic type X collagen–positive chondrocytes in the deep zone of growth factor–treated immature cartilage explants demonstrated elevated interstitial collagenase activity, and gene expression levels of MMP-1 and MMP-13 were also significantly up-regulated in these cartilage explants. We also observed coordinated expression of gelatinases MMP-2 and MMP-9, important enzymes in ECM remodeling. The presence of apoptosis in deep-zone chondrocytes also suggests that this process is related to terminal differentiation of chondrocytes (38). Autodegradation, as we observed in this system, is not of itself unique. It has been observed in Meckel’s cartilage (39). In vivo, this type of resorptive process is generally coordinated with osteoclast- and osteoblast-driven bone formation. The reason deep-zone chondrocytes in immature articular cartilage are affected disproportionately by apoptosis may be due to their phenotype, which is endochondral in origin (40). While the role of apoptosis during mineralization in endochondral ossification is a subject of controversy (41), an increase in the rate of apoptosis has been shown to directly correlate with an increase in the rate of mineralization of cartilage (42). Removal of the endochondral portion of the chondroepiphyseal growth plate through apoptosis and resorption may also indirectly stimulate maturational responses in the surface growth plate. Our preliminary analysis of the gene expression levels of important modulators of endochondral ossification in growth factor–stimulated cartilage, such as Indian hedgehog and parathyroid hormone–related protein (PTHrP), show 4-fold and 3-fold decreases, respec-
KHAN ET AL
tively (data not shown). The generation of a PTHrP knockin mouse (43) has shown that postnatally, there are 2 bands of PTHrP expression (40), one that lies at the junction of the chondroepiphyseal growth plate and one that is found in joint surface chondrocytes and is contingent upon mechanical loading (44). PTHrP signaling suppresses terminal differentiation of endochondral chondrocytes (45); the reduction of PTHrP gene expression through apoptosis of chondrocytes at the chondroepiphyseal junction, coupled with the absence of mechanical loading in our system may act to accelerate resorptive and apoptotic effects. Therefore, as well as initiating maturation, mechanical loading of articular cartilage also actively regulates this process through multiple signaling pathways. Thus, in the absence of mechanical loading, the effect of FGF-2 and TGF1 on immature cartilage explants in vitro can only be seen as a caricature of the process in vivo. Following growth factor stimulation, there was also a switch in the polarity of growth from interstitial to appositional, consistent with previous hypotheses regarding the mechanisms of postnatal growth of articular cartilage (4,7). Cellular proliferation was highly activated in the superficial zone, where we have previously identified stem cells that direct the differentiation and growth of articular chondrocytes (8). Also, FGF-2 and TGF1 exert a synergistic effect on cellular proliferation, implying that some of the effects we observed are not simply the sum of the activities of both factors. A previous gene expression study of postnatal articular cartilage development has described increases in MMP-1 and TIMP-2 and decreases in type IIB and type IX cartilage as markers of maturation (18). In this study, we also observed similar changes in gene expression. However, this previous study used immature and mature cartilage samples and not a tissue that was actively undergoing maturation, and therefore, changes in the expression of MMP-2, MMP-9, TIMP-1, and TIMP-3, as we have observed, while crucial during the process of maturation, are not necessarily required following the attainment of this state, and are therefore missed when using endpoint analysis of development states. With growth and maturity, articular cartilage exhibits greater load-bearing capacity and increased compressive modulus, which is itself a sensitive indicator of biomechanical maturation (46). Previous studies have shown that immature bovine articular cartilage in organ culture for 3 weeks loses up to 50% of its equilibrium modulus (46). In contrast, following stimulation with FGF-2 plus TGF1, we saw a 229% increase in stiffness
ACCELERATING MATURATION IN BOVINE CARTILAGE EXPLANTS
compared to unstimulated control explants. In mature cartilage, the tensile and compressive properties are dependent upon the composition and organization of the collagen fibril network. As maturation proceeds in cartilage, the ratio of mature trivalent pyridinoline collagen crosslinks to immature divalent ketoimine crosslinks increases (20), and this probably occurs in part as a response to reduced turnover of collagen in the tissue (20). We observed a significant decrease in divalent collagen crosslinking in FGF-2/TGF1–stimulated cartilage, which was attributable to the conversion of 2 divalent immature crosslinks to generate a single, mature pyridinoline trivalent linkage between collagen fibrils (47). The decrease in divalent, reducible crosslinks in collagen is a characteristic of articular cartilage maturation, whereas in humans, they virtually disappear at the time of skeletal maturity (20). At the ultrastructural level, we saw profound changes in the organization of collagen fibrils in FGF2/TGF1–stimulated explants, leading to the differentiation of the ECM into pericellular, territorial, and interterritorial matrices, a distinction that was absent in unstimulated control cartilage. Differentiation of the ECM was most apparent at the surface of cartilage explants, within the region of cellular proliferation. The pericellular matrix surrounding chondrocytes regulates the biophysical, biomechanical, and biochemical environmental interactions between the cell and the ECM and is a feature of mature articular chondrocytes (6,48). The generation of this microanatomic structure in our model of cartilage maturation was further confirmed through the redistribution of perlecan protein to newly formed pericellular matrices, matching patterns of labeling seen during the transition from immature to mature articular cartilage (22). Many of the maturational changes in articular cartilage are thought to be initiated through dynamic loading of joint tissues following birth (23). Exercise deprivation of foals during this critical period of growth delays site-specific developmental maturation of joints, and subsequent exercise does not fully compensate for this deficit (23). We hypothesize that in our model, where cartilage explants remain unloaded, high levels of FGF-2, a known mechanotransducer (49), compensate for the lack of dynamic loading in the initiation of maturational processes. In conclusion, our data demonstrate that maturational processes are precociously induced in immature articular cartilage through specific growth factor stimulation. Clinically, in joints repaired with autologous chondrocytes or mesenchymal stem cells, postsurgical transient supplementation with FGF-2 in combination
3425
with TGF1 may be expedient in 2 principal respects: first, to initiate maturational processes and second, to accelerate the developmental transition to a biomechanically robust tissue. A final consequence of our data that requires exploration concerns the potential effect of FGF-2 and TGF1 on osteoarthritic cartilage. A fraction of chondrocytes within osteoarthritic tissue express extracellular molecules typical of an immature cartilage phenotype, such as type IIA procollagen (50). Notionally, these chondrocytes may permit phenotype modulation through specific growth factor stimulation, such that they may be able to induce growth, repair, and maturation of diseased cartilage. ACKNOWLEDGMENTS Drs. Khan and Archer would like to acknowledge the funding received from Arthritis Research UK for the completion of these studies. Dr. Khan is deeply indebted to Professor Vic Duance for his help and encouragement in the course of these studies. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Khan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Khan, Archer. Acquisition of data. Khan, Evans, Young, Blain, Quantock, Avery. Analysis and interpretation of data. Khan, Archer.
REFERENCES 1. Muir H. The chondrocyte, architect of cartilage: biomechanics, structure, function and molecular biology of cartilage matrix macromolecules. Bioessays 1995;17:1039–48. 2. Hunziker EB. Articular cartilage repair: basic science and clinical progress. A review of the current status and prospects. Osteoarthritis Cartilage 2002;10:432–63. 3. Williams GM, Klisch SM, Sah RL. Bioengineering cartilage growth, maturation, and form. Pediatr Res 2008;63:527–34. 4. Hunziker EB, Kapfinger E, Geiss J. The structural architecture of adult mammalian articular cartilage evolves by a synchronized process of tissue resorption and neoformation during postnatal development. Osteoarthritis Cartilage 2007;15:403–13. 5. Benninghoff A. Form und bau der Geleknorpel in ihren Bezeihungen zur Funktion. Z Zellforsch Mikrosk Anat Anz 1925;2: 783–862. 6. Guilak F, Alexopoulos LG, Upton ML, Youn I, Choi JB, Cao L, et al. The pericellular matrix as a transducer of biomechanical and biochemical signals in articular cartilage. Ann N Y Acad Sci 2006;1068:498–512. 7. Hayes AJ, MacPherson S, Morrison H, Dowthwaite G, Archer CW. The development of articular cartilage: evidence for an appositional growth mechanism. Anat Embryol (Berl) 2001;203: 469–79. 8. Dowthwaite GP, Bishop JC, Redman SN, Khan IM, Rooney P,
3426
9.
10.
11.
12.
13.
14.
15. 16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
Evans DJ, et al. The surface of articular cartilage contains a progenitor cell population. J Cell Sci 2004;117:889–97. Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O, Peterson L. Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med 1994;331: 889–95. Wakitani S, Goto T, Pineda SJ, Young RG, Mansour JM, Caplan AI, et al. Mesenchymal cell-based repair of large, full-thickness defects of articular cartilage. J Bone Joint Surg Am 1994;76: 579–92. Roberts S, McCall IW, Darby AJ, Menage J, Evans H, Harrison PE, et al. Autologous chondrocyte implantation for cartilage repair: monitoring its success by magnetic resonance imaging and histology. Arthritis Res Ther 2003;5:R60–73. Farndale RW, Buttle DJ, Barrett AJ. Improved quantitation and discrimination of sulphated glycosaminoglycans by use of dimethylmethylene blue. Biochim Biophys Acta 1986;883:173–7. Creemers LB, Jansen DC, van Veen-Reurings A, van den Bos T, Everts V. Microassay for the assessment of low levels of hydroxyproline. Biotechniques 1997;22:656–8. Blain EJ, Gilbert SJ, Wardale RJ, Capper SJ, Mason DJ, Duance VC. Up-regulation of matrix metalloproteinase expression and activation following cyclical compressive loading of articular cartilage in vitro. Arch Biochem Biophys 2001;396:49–55. Avery NC, Sims TJ, Bailey AJ. quantitative determination of collagen cross-links. Methods Mol Biol 2009;522:103–21. Avery NC, Sims TJ, Warkup C, Bailey AJ. Collagen cross-linking in porcine M. logissimus lumborum: absence of a relationship with variation in texture at pork weight. Meat Science 1996;42:355–69. Hayes WC, Keer LM, Herrmann G, Mockros LF. A mathematical analysis for indentation tests of articular cartilage. J Biomech 1972;5:541–51. Mienaltowski MJ, Huang L, Stromberg AJ, MacLeod JN. Differential gene expression associated with postnatal equine articular cartilage maturation. BMC Musculoskelet Disord 2008;9:149. Salminen H, Perala M, Lorenzo P, Saxne T, Heinegard D, Saamanen AM, et al. Up-regulation of cartilage oligomeric matrix protein at the onset of articular cartilage degeneration in a transgenic mouse model of osteoarthritis. Arthritis Rheum 2000; 43:1742–8. Eyre DR, Dickson IR, Van Ness K. Collagen cross-linking in human bone and articular cartilage: age-related changes in the content of mature hydroxypyridinium residues. Biochem J 1988; 252:495–500. Gomes R, Kirn-Safran C, Farach-Carson MC, Carson DD. Perlecan: an important component of the cartilage pericellular matrix. J Musculoskelet Neuronal Interact 2002;2:511–6. Melrose J, Roughley P, Knox S, Smith S, Lord M, Whitelock J. The structure, location, and function of perlecan, a prominent pericellular proteoglycan of fetal, postnatal, and mature hyaline cartilages. J Biol Chem 2006;281:36905–14. Brama PA, TeKoppele JM, Bank RA, Barneveld A, van Weeren PR. Development of biochemical heterogeneity of articular cartilage: influences of age and exercise. Equine Vet J 2002;34:265–9. Asanbaeva A, Tam J, Schumacher BL, Klisch SM, Masuda K, Sah RL. Articular cartilage tensile integrity: modulation by matrix depletion is maturation-dependent. Arch Biochem Biophys 2008; 474:175–82. Julkunen P, Harjula T, Iivarinen J, Marjanen J, Seppanen K, Narhi T, et al. Biomechanical, biochemical and structural correlations in immature and mature rabbit articular cartilage. Osteoarthritis Cartilage 2009;17:1628–38. Sah RL, Chen AC, Grodzinsky AJ, Trippel SB. Differential effects of bFGF and IGF-I on matrix metabolism in calf and adult bovine cartilage explants. Arch Biochem Biophys 1994;308:137–47. Mandl EW, Jahr H, Koevoet JL, van Leeuwen JP, Weinans H,
KHAN ET AL
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38. 39.
40.
41.
42.
43.
44.
Verhaar JA, et al. Fibroblast growth factor-2 in serum-free medium is a potent mitogen and reduces dedifferentiation of human ear chondrocytes in monolayer culture. Matrix Biol 2004; 23:231–41. Schofield JN, Wolpert L. Effect of TGF-1, TGF-2, and bFGF on chick cartilage and muscle cell differentiation. Exp Cell Res 1990;191:144–8. Jentzsch KD, Wellmitz G, Heder G, Petzold E, Buntrock P, Oehme P. A bovine brain fraction with fibroblast growth factor activity inducing articular cartilage regeneration in vivo. Acta Biol Med Ger 1980;39:967–71. Chuma H, Mizuta H, Kudo S, Takagi K, Hiraki Y. One day exposure to FGF-2 was sufficient for the regenerative repair of full-thickness defects of articular cartilage in rabbits. Osteoarthritis Cartilage 2004;12:834–42. Yamamoto T, Wakitani S, Imoto K, Hattori T, Nakaya H, Saito M, et al. Fibroblast growth factor-2 promotes the repair of partial thickness defects of articular cartilage in immature rabbits but not in mature rabbits. Osteoarthritis Cartilage 2004;12:636–41. Kaul G, Cucchiarini M, Arntzen D, Zurakowski D, Menger MD, Kohn D, et al. Local stimulation of articular cartilage repair by transplantation of encapsulated chondrocytes overexpressing human fibroblast growth factor 2 (FGF-2) in vivo. J Gene Med 2006;8:100–11. Heine U, Munoz EF, Flanders KC, Ellingsworth LR, Lam HY, Thompson NL, et al. Role of transforming growth factor- in the development of the mouse embryo. J Cell Biol 1987;105:2861–76. Moroco JR, Hinton R, Buschang P, Milam SB, Iacopino AM. Type II collagen and TGF-s in developing and aging porcine mandibular condylar cartilage: immunohistochemical studies. Cell Tissue Res 1997;289:119–24. Morales TI, Roberts AB. Transforming growth factor  regulates the metabolism of proteoglycans in bovine cartilage organ cultures. J Biol Chem 1988;263:12828–31. Asanbaeva A, Masuda K, Thonar EJ, Klisch SM, Sah RL. Regulation of immature cartilage growth by IGF-I, TGF-1, BMP-7, and PDGF-AB: role of metabolic balance between fixed charge and collagen network. Biomech Model Mechanobiol 2008;7: 263–76. Jakob M, Demarteau O, Schafer D, Hintermann B, Dick W, Heberer M, et al. Specific growth factors during the expansion and redifferentiation of adult human articular chondrocytes enhance chondrogenesis and cartilaginous tissue formation in vitro. J Cell Biochem 2001;81:368–77. Gibson G. Active role of chondrocyte apoptosis in endochondral ossification. Microsc Res Tech 1998;43:191–204. Ishizeki K, Nawa T. Further evidence for secretion of matrix metalloproteinase-1 by Meckel’s chondrocytes during degradation of the extracellular matrix. Tissue Cell 2000;32:207–15. Onyekwelu I, Goldring MB, Hidaka C. Chondrogenesis, joint formation, and articular cartilage regeneration. J Cell Biochem 2009;107:383–92. Bianco P, Cancedda FD, Riminucci M, Cancedda R. Bone formation via cartilage models: the “borderline” chondrocyte. Matrix Biol 1998;17:185–92. Roy R, Kudryashov V, Binderman I, Boskey AL. The role of apoptosis in mineralizing murine versus avian micromass culture systems. J Cell Biochem. 2010;111:653–8. Chen X, Macica CM, Dreyer BE, Hammond VE, Hens JR, Philbrick WM, et al. Initial characterization of PTH-related protein gene-driven lacZ expression in the mouse. J Bone Miner Res 2006;21:113–23. Broadus AE, Macica C, Chen X. The PTHrP functional domain is at the gates of endochondral bones. Ann N Y Acad Sci 2007;1116: 65–81.
ACCELERATING MATURATION IN BOVINE CARTILAGE EXPLANTS
45. Vortkamp A, Lee K, Lanske B, Segre GV, Kronenberg HM, Tabin CJ. Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 1996;273:613–22. 46. Williamson AK, Masuda K, Thonar EJ, Sah RL. Growth of immature articular cartilage in vitro: correlated variation in tensile biomechanical and collagen network properties. Tissue Eng 2003; 9:625–34. 47. Eyre DR. Crosslink maturation in bone collagen. New York: Elsevier; 1981. 48. Poole CA, Wotton SF, Duance VC. Localization of type IX
3427
collagen in chondrons isolated from porcine articular cartilage and rat chondrosarcoma. Histochem J 1988;20:567–74. 49. Vincent TL, McLean CJ, Full LE, Peston D, Saklatvala J. FGF-2 is bound to perlecan in the pericellular matrix of articular cartilage, where it acts as a chondrocyte mechanotransducer. Osteoarthritis Cartilage 2007;15:752–63. 50. Aigner T, Zhu Y, Chansky HH, Matsen FA III, Maloney WJ, Sandell LJ. Reexpression of type IIA procollagen by adult articular chondrocytes in osteoarthritic cartilage. Arthritis Rheum 1999; 42:1443–50.
