CCN2/CTGF (CCN family 2/connective tissue growth factor) is a multi-cellular protein with a broad range of activities. It modulates many cellular functions, including proliferation, migration, adhesion and extracellular matrix production, and it is thus involved in many biological and pathological processes. In particular, CCN2/CTGF is essential for normal skeletal development. To identify CCN2/CTGF-interactive proteins capable of modulating its action in cartilage, we carried out a yeast two-hybrid screening using CCN2/CTGF peptide as a bait and a cDNA library from a chondrocytic cell line, HCS-2/8. In the present paper, we report the identification of aggrecan, which is a major proteoglycan of the extracellular matrix in cartilage, as a CCN2/CTGF-binding protein. Among the four domains of CCN2/CTGF, the IGFBP [IGF (insulin-like growth factor)-binding protein-like] and/or VWC (von Willebrand factor type C) domains had a direct interaction with aggrecan in a yeast two-hybrid assay. The results of a solid-phase-binding assay using aggrecan-coated plates also showed binding to recombinant CCN2/CTGF in a dose-dependent manner. rIGFBP (recombinant IGFBP) and rVWC (recombinant VWC) module peptides had stronger binding to aggrecan compared with rTSP1 (recombinant thrombospondin type 1 repeat) and rCT (recombinant C-terminal cystine knot) module peptides. SPR (surface plasmon resonance) analysis showed the direct interaction between the CCN2/CTGF and aggrecan, and ectopically overexpressed CCN2/CTGF and AgG3 (G3 domain of aggrecan) confirmed their binding In vivo. Indirect immunofluorescence analysis indicated that CCN2/CTGF was extracellularly co-localized with aggrecan on HCS-2/8 cells. The rIGFBP–rVWC peptide effectively enhanced the production and release of aggrecan compared with the rTSP–rCT peptide in chondrocytes. These results indicate that CCN2/CTGF binds to aggrecan through its N-terminal IGFBP and VWC modules, and this binding may be related to the CCN2/CTGF-enhanced production and secretion of aggrecan by chondrocytes.
- CCN family 2/connective tissue growth factor (CCN2/CTGF)
- extracellular matrix
- surface plasmon resonance (SPR)
- yeast two-hybrid
CCN2/CTGF (CCN family 2/connective tissue growth factor) is a 36–39 kDa polypeptide and a member of the multi-functional matricellular factor family, which consists of CCN1/Cyr61/Cef10, CCN2/CTGF/Fisp12/Hcs24, CCN3/Nov, CCN4/WISP-1/ELM-1, CCN5/WISP-2/CTGF-L/rCOP-1 and CCN6/WISP-3 . CCN2/CTGF is specifically expressed at high levels in developing cartilage, but low levels in adult tissues; and its overexpression in soft tissues is linked to several pathological disorders, such as fibrosis, because of its strong enhancing effects on the production of ECMs (extracellular matrices) [1–4]. CCN2/CTGF acts on various types of cells, such as fibroblasts, endothelial cells, vascular smooth muscle cells, osteoblasts and chondrocytes [1–4], and promotes cellular functions, including proliferation and differentiation of chondrocytes [5,6], osteoblasts  and endothelial cells [8,9]. Little is known, however, about the transmembrane receptors or signalling pathways that are able to mediate CCN2/CTGF-derived signals in these cells.
CCN2/CTGF consists of four modules: IGFBP [IGF (insulin-like growth factor)-binding protein-like], VWC (von Willebrand factor type C), TSP1 (thrombospondin type 1 repeat) and CT (C-terminal cystine knot). Each of these modules has different binding partners. It has been shown that the CT domain of CCN2/CTGF binds to integrin α5β1 and promotes adhesion and migration of pancreatic stellate cells . This domain also interacts directly with fibronectin and enhances cell adhesion of chondrocytes through integrin α5β1 . Furthermore, it induces adhesion of hepatic cells by direct binding to the integrin receptor αvβ3 and to heparan sulfate proteoglycan through its C-terminal heparin-binding domain . There are also reports indicating that CCN2/CTGF signalling occurs through LRP-1 (low-density lipoprotein receptor-related protein-1) [13,14], an endocytotic receptor, and TrkA [14,15]. Other studies show that CCN2/CTGF directly binds to BMP-4 (bone morphogenetic protein-4) and TGF-β (transforming growth factor-β) through its VWC module and prevents or enhances their binding to their own receptors . Moreover, CCN2/CTGF binds VEGF (vascular endothelial growth factor) and inhibits its angiogenic effect . Because of these characteristics, CCN2/CTGF has been attracting the interest of researchers as a novel type of factor that may be called a ‘signal conductor’.
In order to identify extracellular or cell-surface targets that modulate CCN2/CTGF action on specific target cells, we carried out a yeast two-hybrid assay using a cDNA library from a human chondrocytic cell line, HCS-2/8, which highly expresses CCN2/CTGF. As a result, aggrecan was found as a binding protein to CCN2/CTGF. Aggrecan is the major proteoglycan of cartilage and provides a high degree of physical strength and elasticity to this tissue due to osmotic swelling pressure caused by the highly hydrated sulfated GAG (glycosaminoglycan) side chains . Both of aggrecan and CCN2/CTGF exist abundantly in the pre-hypertrophic zone of the growth plate in cartilage, and CCN2/CTGF enhances the expression of aggrecan [19,20]. However the interaction between CCN2/CTGF and aggrecan had been unclear.
In the present study, we show for the first time that (i) CCN2/CTGF binds via its IGFBP and VWC domains to the G3 domain of aggrecan, (ii) CCN2/CTGF co-localizes with aggrecan in ECMs of HCS-2/8 cells, and (iii) from transient-transfection analysis of full-length, N-terminal-half or C-terminal-half molecules of CCN2/CTGF in HCS-2/8 cells, the production and release of aggrecan is enhanced by overexpression of the N-terminal IGFBP and VWC domains of CCN2/CTGF, but not by that of the C-terminal TSP1 and CT domains.
MATERIALS AND METHODS
Yeast two-hybrid cDNA library screening
Yeast two-hybrid cDNA library screening was performed as described previously . Briefly, full-length CCN2/CTGF cDNA was expressed as a GAL4BD (GAL4 DNA-binding domain)-fusion protein. cDNA library genes from HCS-2/8 cells were expressed as GAL4AD (GAL4 transactivation domain)-fusion proteins for a two-hybrid assay in yeast AH109 cells, and the cells were screened on selection medium [SD/−Ade/−His/−Leu/−Trp (synthetic dropout medium lacking Ade, His, Leu and Trp)]. Positive clones were analysed by DNA sequencing. For the analysis of the aggrecan-binding domain in CCN2/CTGF, expression plasmids of CCN2/CTGF fragments, which were expressed as GAL4BD-fusion proteins, and AgG3 (G3 domain of aggrecan) cDNA, which was expressed as a GAL4AD-fusion protein, were used to transform yeast cells. The binding was monitored by growth in SD/−Ade/−His/−Leu/−Trp selection medium.
Expression and purification of full-length and truncated rhCCN2/CTGFs (recombinant human CCN2/CTGFs)
To clone hCCN2/CTGF, we amplified the cDNA fragment coding for hCCN2/CTGF without a signal peptide by PCR using the 5′ primer (5′-GGC CGA ATT CCC AGA ACT GCA GCG GGC CGT GCC GGT GCC CG-3′) containing an EcoRI site and the 3′ primer (5′-ACG GAG ATC TTT AAT GAT GAT GAT GAT GAT GTG CCA TGT CTC CGT ACA TCT TCC TGT AGT-3′), which encodes a His6 tag (underlined) at the C-terminus of CCN2/CTGF and contains a BglII site after the stop codon. The PCR fragments were cloned into pT7-flag-1 vector at EcoRI/BglII sites, with the result that the expressed CCN2/CTGF carries a FLAG tag at its N-terminus and a His6 tag at its C-terminus. The recombinant plasmid was sequenced to ensure the absence of mutations. For expression of rhCCN2/CTGF, Escherichia coli Rosetta™2(DE3)pLysS (Novagen) was transformed with pT7-flag-1/hctgf, and cultured at 25 °C for 4 h before adding 0.05 mM IPTG (isopropyl β-D-thiogalactopyranoside). After another 18 h in culture, E. coli were harvested by centrifugation, and frozen and thawed once to disrupt the cells. The cells were then resuspended in lysis buffer [0.5 M NaCl, 0.05 M Tris/HCl (pH 8.0), 1% Triton, 1 mM PMSF and 1 μM pepstatin A], sonicated, and centrifuged to remove cell debris. For purification of the rhCCN2/CTGF protein, 1 ml of Ni-NTA (Ni2+-nitrilotriacetate)–agarose gel (Qiagen) was added to the supernatant to adsorb the His-tagged proteins. After gentle shaking for 1 h at 4 °C, the agarose beads were washed with lysis buffer and wash buffer [20 mM Tris/HCl (pH 8.0), 10 mM KCl, 0.5 mM PMSF and 1 μM pepstatinA]. The beads were loaded on to a column and washed. The protein was then eluted with 20 ml of 500 mM imidazole elution buffer [500 mM imidazole, 20 mM Tris/HCl (pH 8.0), 10 mM KCl, 0.5 mM PMSF and 1 μM pepstatin A]. Thereafter 0.2 ml of anti-FLAG® M2 affinity gel freezer-safe (Sigma) suspended in 40 ml of TBS [20 mM Tris/HCl (pH 8.0) and 10 mM KCl] containing 0.5 mM PMSF and 1 μM pepstatin A was mixed with 10 ml of the eluate. The mixture was subsequently incubated for 2 h with gentle rotation to capture the FLAG-fusion protein. After the column had been washed with TBS, 0.1 ml of elution buffer containing 400 μg/ml FLAG peptide (Sigma) was added to the gel to elute the bound protein. The purity and quantity of the rhCCN2/CTGF protein was checked by Coomassie Brilliant Blue staining and Western blotting using an anti-FLAG® M2 monoclonal antibody (Sigma) and an HRP (horseradish peroxidase)-conjugated anti-His (C-terminal) antibody (Invitrogen). Expression and purification of truncated rhCCN2/CTGF fragments were carried out accordingly.
In some experiments, rhCCN2/CTGF (BioVendor Laboratory Medicine) was also used.
Expression and purification of recombinant AgG3
To clone recombinant AgG3, we amplified the cDNA fragment by PCR using the sense primer (5′-GCG CGA ATT CAA CAG CCA CCT CCC CAA CAG ATG CTT CC-3′) containing an EcoRI site, the antisense primer (5′-ATC GTC TAG AGT CAG TGG GCT GTG CTG GGG CGG CTC CT-3′) containing a BglII site and pGADT7/aggrecan isolated from the yeast two-hybrid cDNA library screening was used as a template. The PCR fragments were cloned into pT7-flag-1 vector at EcoRI/BglII sites, giving rise to an AgG3 cDNA with a FLAG tag at its N-terminus. The recombinant plasmid was sequenced to ensure the absence of mutations. Recombinant AgG3 was expressed in Rosetta™2(DE3)pLysS and purified with an anti-FLAG affinity gel similar to that for rhCCN2/CTGF.
Wells of an ELISA plate were coated with various concentrations of aggrecan from bovine articular cartilage (Sigma) in 50 mM NaHCO3 buffer (pH 9.6) at 4 °C overnight. After the wells had been washed with wash buffer [50 mM Tris/HCl (pH 7.4), 150 mM NaCl and 0.05% Tween 20] and blocked with 100 μl of binding buffer [50 mM Tris/HCl (pH 7.4), 150 mM NaCl, 2% (w/v) BSA and 0.05% Tween 20] for 2 h at 37 °C, 50 μl of His6-tagged CCN2/CTGF, which had been diluted with binding buffer, was added to the wells, and incubation was conducted for 2 h at 37 °C. The wells were washed with wash buffer and then incubated with 50 μl of 1:1000 dilution of the HRP-conjugated anti-His (C-terminal) antibody (Invitrogen). Bound HRP was monitored using TMB (3,3′,5,5′-tetramethylbenzidine) peroxidase substrate (Sigma).
Cell culture and mammalian expression vectors
HCS-2/8 human chondrocytic cells  and Cos7 cells were cultured in DMEM (Dulbecco's modified minimum essential medium) (Nissui Pharmaceutical) containing 10% (v/v) fetal bovine serum (Biotrade) .
Transient transfection experiments were performed by using FuGENE6™ (Roche) as described previously . The full-length CCN2/CTGF (full) expression plasmid pcDNA3.1(−)-ctgf has been described previously . The IGFBP–VWC module expression plasmid pcDNA3.1(−)-iv was generated by replacing the PmaCI/AflII fragment encoding the VWC, TSP1 and CT domains with a PCR product encoding the VWC domain only, to which a downstream flanking AflII site had been added via PCR. Prior to the construction of the TSP1–CT module expression plasmid pcDNA3.1(−)-tc, the EcoRV/HindIII fragment containing the full-length CCN2/CTGF coding region was replaced with a PCR product encoding the signal peptide of CCN2/CTGF in pcDNA3.1(−)-ctgf, which was amplified with sense (5′-GGA TAT CGC GTG CCA ACC ATG A-3′) and antisense (5′-GGG AAG CTT GAA TTC CGG CCC GCT GCA GT-3′) primers. The resultant plasmid contained a CCN2/CTGF signal peptide sequence with a unique EcoRI site downstream. Into this plasmid, a PCR product containing the TSP1–CT encoding region, which had been obtained with sense (5′-GGA ATT CAT GCC GCG GAT TAG AGC CAA CTG CCT GGT CCA-3′) and antisense (5′-TAG AAG GCA CAG TCG AGG-3′) primers, was inserted between unique EcoRI and HindIII sites to yield pcDNA3.1(−)-tc.
pFlag-CMV2/AgG3-His which expresses the His6-tagged AgG3 was constructed by inserting the G3 domain of aggrecan cDNA fragment prepared by PCR using pGADT7/aggrecan as a template and primers: pFlag-G3-5′ (5′-GAA GAG AAA GAG AAA GAG CGC GAA TCA AAC AGC CAC CTC CCC AAC AGA AGA TGC TTC C-3′) and pFlag-G3-3′ (5′-ATC GAG ATC TTC AAT GAT GAT GAT GAT GAT GAT GGT GGG CTG TGC TGG GGC GGC TCC T-3′). The His6-tag sequences at the C-terminus of CCN2/CTGF are underlined, and a BglII site was continued after the stop codon.
SPR (surface plasmon resonance) measurements
rhCCN2/CTGF (BioVendor Laboratory Medicine) was diluted to 50 μg/ml with 10 mM sodium acetate buffer (pH 4.0) and immobilized on to CM5 sensor chips (GE Healthcare) according to standard amine coupling procedures. Purified aggrecan (Sigma) diluted with HBS-P [Hepes-buffered saline containing surfactant P; 10 mM Hepes/HCl, 150 mM NaCl and 0.005% surfactant P-20 (pH 7.5) at 25 °C] to concentrations of 12.5, 25, 50 and 100 μg/ml was injected into the flow cells. To examine the effect of Ca2+ on the interaction of CCN2/CTGF with aggrecan, we carried out experiments in the presence or absence of CaCl2 (2 mM). For affinity measurements, binding and dissociation were monitored with a Biacore X (GE Healthcare). The sensorgrams were corrected by subtracting the signal of the reference cell. The data were fitted using the BIAevaluation software version 4.1 (GE Healthcare). Binding data were globally fitted to the 1:1 Langmuir binding model which assumes binding homogeneity. Values for the statistical closeness of fit, χ2, were always below 10, indicating that the simple 1:1 model of interaction correctly described the experimental data.
Plasmid expressing HA (haemagglutinin)-tagged CCN2/CTGF (CCN2/CTGF–HA) and/or that expressing His6-tagged AgG3 (AgG3–His) were transfected into Cos7 cells. The cells were harvested in lysis buffer [50 mM Tris/HCl (pH 7.4), 150 mM NaCl, 1% Triton X-100 and 0.5% protease inhibitor cocktail (Sigma)] after 24 h of transfection, and the complex of CCN2/CTGF–HA and AgG3–His was incubated with an anti-HA antibody (Covance). The CCN2/CTGF–AgG3–antibody complex was precipitated with Protein G–Sepharose (Amersham Biosciences). After washing five times with wash buffer [50 mM Tris/HCl (pH 7.4), 150 mM NaCl and 0.1% Triton X-100], precipitated proteins were analysed by Western blotting using an anti-His antibody (Bethyl Laboratories).
Immunofluorescence staining of extracellular proteins
HCS-2/8 cells were grown on glass slides for 6 h in culture medium and transfected with a plasmid expressing FLAG and CCN2/CTGF–HA. After incubation for 42 h, the cells were fixed with 4% formaldehyde. Extracellular immunofluorescence staining was done by using an anti-aggrecan antibody (R&D Systems) and anti-HA antibody (Clontech), followed by incubation with Alexa Fluor® 568 goat anti-(rabbit IgG) antibody (Molecular Probes) and Alexa Fluor® 488 goat anti-(mouse IgG) antibody (Molecular Probes), as described previously [23,24]. After mounting, protein localization was observed by using a confocal laser scanning microscope (Bio-Rad Laboratories).
Overexpression of CCN2/CTGF fragments and immunoblotting of aggrecan
HCS-2/8 cells were transfected with pcDNA3.1(−)-ctgf, pcDNA3.1(−)-iv, pcDNA3.1(−)-tc or pcDNA3.1(−), as a control, and incubated for 48 h. Subsequently, the culture medium was collected and centrifuged, and the supernatants were used for immunoblotting assays. The cell layers were washed with PBS and lysed in 50 mM TBS and 0.5 M NaCl containing 1% Nonidet P40 and 1% Triton-X 100 for 30 min on ice, sonicated twice for 10 s and centrifuged to remove cell debris. The supernatant from the culture medium or lysate was mixed with an equal volume of 2× SDS sample buffer [4% (w/v) SDS, 20% (v/v) glycerol, 0.1% Bromophenol Blue and 0.125 M Tris/HCl (pH 6.8)] without 2-mercaptoethanol and denatured by boiling for 5 min. Samples were separated by SDS/PAGE and proteins were transferred on to PVDF membranes. The membrane was probed with a monoclonal anti-(human aggrecan) antibody (R&D Systems), and signals were detected by using an ECL® (enhanced chemiluminescence) system (GE Healthcare).
Identification of the interaction of aggrecan with CCN2/CTGF through IGFBP and VWC domains
To identify CCN2/CTGF-binding proteins, we carried out a GAL4-based yeast two-hybrid screening assay using the human chondrocytic cell line HCS-2/8 and a CCN2/CTGF cDNA fragment (amino acid residues 27–349) as a bait. As a result, the aggrecan 1 gene product was identified as one of the CCN2/CTGF-binding proteins among 50000 individual clones. DNA sequencing of this clone revealed that the CLD (C-type lectin-like domain) and CBP (complement-binding protein-like) domain, which were located in the C-terminal AgG3 (Figure 1A), were sufficient for binding to CCN2/CTGF. To identify the aggrecan-interacting sites on the CCN2/CTGF molecule, we co-expressed truncated forms of CCN2/CTGF, as shown in Figures 1B(a)–1B(i), together with AgG3 in AH109 cells. The full-length CCN2/CTGF [CCN2full in Figure 1B(a)] and shorter fragments containing the combination of IGFBP and VWC domain [IGFBP–VWC; Figures 1B(e) and 1B(f)] showed strong binding to aggrecan, whereas relatively weaker binding was observed with the VWC–TSP1–CT [Figure 1B([b)], CT [Figure 1B(d)] and VWC [Figure 1B(h)] domains. Other fragments did not show any interaction (Figure 1B).
Binding of aggrecan to CCN2/CTGF and its modules confirmed by a solid-phase-binding assay
To confirm the binding of CCN2/CTGF to aggrecan by means of a solid-phase-binding assay, we generated full-length rhCCN2/CTGF and its four domain proteins (Figure 2A), as described in the Materials and methods section. Solid-phasebinding assays using increasing amounts of full-length CCN2/CTGF showed that CCN2/CTGF bound to aggrecan-coated plates in a dose-dependent manner (Figure 2B).
To estimate the aggrecan-binding domain of CCN2/CTGF and to evaluate the binding strength of the domain(s), we coated microtitre dishes with aggrecan and incubated them with increasing amounts of rhCCN2/CTGF fragments, adjusted to be of equal amounts. Detection of bound CCN2/CTGF fragments with the anti-His antibody indicated that the IGFBP and VWC domain proteins had strong binding to aggrecan with equal or even higher binding activity than full-length CCN2/CTGF, whereas the TSP1 or CT domain proteins had lower binding (Figure 2C). These findings support the results from the yeast two-hybrid assay (Figure 1B). Furthermore, we generated a recombinant protein having the CLD and CBP domain of aggrecan (AgG3) and confirmed the binding of CCN2/CTGF to AgG3 in the solid-phase-binding assay (Figure 2D). Since the binding of CLD to other proteins is often influenced by Ca2+ [25,26], we examined the effect of Ca2+ on the binding of aggrecan to CCN2/CTGF. As shown in Figure 2(E), the average of absorbance in wells incubated in the presence of 1 mM CaCl2 buffer was decreased to 85% of that for wells without CaCl2 buffer.
Determination of the binding affinities by SPR technology
The binding affinity of CCN2/CTGF for aggrecan was determined by SPR analysis. Kinetic measurements using different concentrations of aggrecan (12.5, 25, 50 and 100 μg/ml) in running buffer without Ca2+ yielded a Kd (dissociation constant) of 1.19×10−8M for CCN2/CTGF (Figure 3A). To detect the effect of Ca2+ on the binding of aggrecan to CCN2/CTGF, we carried out an SPR assay in running buffer with 2 mM CaCl2. As the response of aggrecan in the presence of Ca2+ was too low to analyse the kinetics, we compared their response curves. As shown in Figure 3(B), the number of resonance units for aggrecan to immobilized CCN2/CTGF was decreased by approx. 40%.
In vivo interaction of CCN2/CTGF and AgG3 using co-immunoprecipitation assays
We performed co-immunoprecipitation assays to confirm the interaction of CCN2/CTGF and AgG3 In vivo. Cos7 cells were transfected with pFlag-CMV2/CCN2-HA and/or pFlag-CMV2/AgG3-His. After cultivation for 24 h, cell lysates were collected and HA-tagged proteins were immunoprecipitated using anti-HA monoclonal antibodies. Precipitated proteins were analysed by Western blotting using the anti-His antibody. As shown in Figure 4(A), AgG3–His was effectively co-immunoprecipitated with the anti-HA monoclonal antibody from cells expressing both CCN2/CTGF–HA and AgG3–His. These results indicated the binding of CCN2/CTGF to AgG3 In vivo.
CCN2/CTGF and aggrecan co-localized in ECMs formed by HCS-2/8 cells
To verify the In vivo interaction of CCN2/CTGF with the full-length aggrecan, we carried out indirect immunofluorescent staining of HCS-2/8 chondrocytic cells transiently overexpressing CCN2/CTGF–HA by using anti-aggrecan and anti-HA antibodies. As shown in Figure 4(B), both CCN2/CTGF (red) and aggrecan (green) were co-localized in the ECMs formed by HCS-2/8 cells.
N-terminal fragments of CCN2/CTGF are indispensable for the enhancement of aggrecan production
Previously, Nakanishi et al.  reported that overexpression of CCN2/CTGF in chondrocytes increased the expression of aggrecan core protein and enhanced the synthesis of sulfated proteoglycan. In the present study, we propose a connection between the enhancement of aggrecan synthesis by CCN2/CTGF and the binding of the N-terminal part of CCN2/CTGF to aggrecan. In support of this hypothesis, we overexpressed the N-terminal (IGFBP–VWC), C-terminal (TSP1–CT) or full-length form of CCN2/CTGF in HCS-2/8 cells. Cell lysates and culture supernatants were analysed for aggrecan synthesis and secretion by immunoblotting using an anti-(aggrecan core protein) antibody. The N-terminal peptide of CCN2/CTGF, as well as full-length CCN2/CTGF, enhanced aggrecan synthesis and deposition in the cells (Cell lysate; Figure 5) and aggrecan secretion into the culture medium (Media; Figure 5). In contrast, the C-terminal CCN2/CTGF fragment TSP1–CT did not show enhanced aggrecan production/deposition in the cells or secretion of it into the culture medium compared with cells transfected with a control vector (Figure 5).
CCN2/CTGF, which consists of four domains (IGFBP, VWC, TSP1 and CT domains), regulates the proliferation and differentiation of chondrocytes as well as the production of ECM in cartilage. CCN2/CTGF stimulates aggrecan synthesis in HCS-2/8 chondrocytic cells and primary rabbit chondrocytes . In the present study, for the first time, we provide experimental evidence that CCN2/CTGF binds to aggrecan through its N-terminal domains (IGFBP and VWC domains). Furthermore, we have shown that the N-terminal half of CCN2/CTGF (IGFBP–VWC domain) strongly enhances aggrecan synthesis and secretion by HCS-2/8 chondrocytes. The same domain has been reported to interact with IGF , TGF-β and BMP-4 ; thus this N-terminal part of CCN2/CTGF may be important in transducing cell-surface signalling into the cells.
Previously, we have shown that the C-terminal CT domain of CCN2/CTGF binds to ECM proteins, such as fibronectin, and enhances cell adhesion to fibronectin . In light of other reports on the binding of the CT domain to cell-surface components, such as integrin αMβ2 , integrin αVβ3  or heparan sulfate proteoglycan , the CT domain may play a role in the binding of CCN2/CTGF to the cell surface and in modulating cell–cell contacts. In contrast with other matrix proteins, however, aggrecan, did not show any ability to bind to the CT domain of CCN2/CTGF. Furthermore, unlike fibronectin, aggrecan inhibits adhesion of chondrocytes to culture dishes (results not shown), but the addition of CCN2/CTGF did not affect the inhibition of cell adhesion by aggrecan (results not shown). These findings suggest that the binding of CCN2/CTGF to aggrecan is not concerned with cell adhesion.
Aggrecan consists of a 250 kDa core protein which contains three globular domains (G1, G2 and G3) and a GAG-attachment domain to which multiple chondroitin sulfate and keratan sulfate chains are attached. In cartilage, aggrecan is anchored to hyaluronic acid and stabilized by linking the protein through the C-terminal G1 and G2 domains . The N-terminal G3 domain was reported to bind to several components of ECMs, such as fibulin-1, tenascin or thrombospondin (matricellular proteins), and these interactions were considered to contribute to the constitution of a correct matrix network [25,26,30]. In the present study, we have shown that CCN2/CTGF binds to the aggrecan core protein (Figure 2D and 4A) and, for the binding, only the G3 domain of aggrecan, containing a lectin-like domain, was sufficient. Our present study indicates that the binding of CCN2/CTGF to aggrecan may contribute to the physiological distribution of CCN2/CTGF in the growth plate of cartilage. The role of the CCN2/CTGF–aggrecan interaction is still unclear, but it is possible that the embedding of CCN2/CTGF in the aggrecan-rich cartilage matrix prevents its degradation by proteases. The accumulation of CCN2/CTGF adjacent to the cell surface could enhance the effect of CCN2/CTGF on chondrocytes. In support of this hypothesis, there is a report showing that aggrecan-deficient mice (cmd/cmd) develop short limbs and dwarfism [31–33]. These pathogenic features in cmd/cmd mice may result from the loss of the CCN2/CTGF interaction with aggrecan. Similarly, CCN2/CTGF-deficient mice having immature bone formation along with fragile bones and cartilage may be the result of a lack of aggrecan accumulation in cartilage .
The CLD in the G3 domain has been reported to bind many extracellular components in a Ca2+-dependent manner [25,26]. Binding of CCN2/CTGF to aggrecan, however, was decreased in the presence of Ca2+ (Figures 2E and 3B) and was not influenced by EDTA (Figure 2E). These findings suggest that the interaction between CCN2/CTGF and G3 domain may not occur through the CLD, and that binding of matricellular proteins to CLD in the presence of Ca2+ may abolish G3 binding to CCN2/CTGF.
Our present observations that the N-terminal half of CCN2/CTGF interacts with aggrecan and, at the same time, is involved in the induction of aggrecan synthesis sheds new light on the role of CCN2/CTGF as a modifier of ECM structure and composition. The binding involves AgG3 and is negatively influenced by Ca2+. These findings indicate a new relationship between aggrecan and CCN2/CTGF, which may contribute to the maintenance of ECM in cartilage.
This work was supported by the programs Grants-in-Aid for Scientific Research (C) (to T. H.) and (S) (to M. T.) from the Japan Society for the Promotion of Science, and Grant-in-Aid for Exploratory Research (to M. T.) of the Ministry of Education, Culture, Sports, Science, and Technology of Japan.
We are grateful to all members of the Department of Biochemistry and Molecular Dentistry, Okayama University Graduate School of Medicine, Dentistry, and Pharmaceutical Sciences for helpful discussions and technical assistance. We thank Yuki Nonami-Nakata and Yoko Tada for their secretarial assistance.
Abbreviations: Ade, adenine; AgG3, G3 domain of aggrecan; BMP-4, bone morphogenetic protein-4; CBP, complement-binding protein-like; CCN2/CTGF, CCN family 2/connective tissue growth factor; CLD, C-type lectin-like domain; CT, C-terminal cystine-knot; GAL4AD, GAL4 transactivation domain; GAL4BD, GAL4 DNA-binding domain; ECM, extracellular matrix; GAG, glycosaminoglycan; HA, haemagglutinin; HBS-P, Hepes-buffered saline containing surfactant P; HRP, horseradish peroxidase; IGF, insulin-like growth factor; IGFBP domain, IGF-binding protein-like; Ni-NTA, Ni2+-nitrilotriacetate; (r)hCCN2/CTGF, (recombinant) human CCN2/CTGF; SPR, surface plasmon resonance; TGF-β, transforming growth factor-β; TMB, 3,3′,5,5′-tetramethylbenzidine; TSP1, thrombospondin type 1 repeat; VWC, von Willebrand factor type C
- © The Authors Journal compilation © 2009 Biochemical Society