The control point by which chondrocytes take the decision between the cartilage differentiation program or the joint formation program is unknown. Here, we have investigated the effect of α5β1 integrin inhibitors and bone morphogenetic protein (BMP) on joint formation. Blocking ofα5β1 integrin by specific antibodies or RGD peptide(arginine-glycine-aspartic acid) induced inhibition of pre-hypertrophic chondrocyte differentiation and ectopic joint formation between proliferating chondrocytes and hypertrophic chondrocytes. Ectopic joint expressed Wnt14,Gdf5, chordin, autotaxin, type I collagen and CD44, while expression of Indian hedgehog and type II collagen was downregulated in cartilage. Expression of these interzone markers confirmed that the new structure is a new joint being formed. In the presence of BMP7, inhibition of α5β1 integrin function still induced the formation of the ectopic joint between proliferating chondrocytes and hypertrophic chondrocytes. By contrast,misexpression of α5β1 integrin resulted in fusion of joints and formation of pre-hypertrophic chondrocytes. These facts indicate that the decision of which cell fate to make pre-joint or pre-hypertrophic is made on the basis of the presence or absence of α5β1 integrin on chondrocytes.

Joint formation constitutes a key step in the morphogenesis of the appendicular skeleton of vertebrates. Long bones are initially formed as an elongated chondrogenic condensation. Next in development, joints are formed at specific regions of the condensing cartilages by a complex process, which involves dedifferentiation of the chondrogenic tissue and apoptosis. The first morphological evidence of joint development is the formation of the interzone,characterized by flattened, fibroblastic-like chondrocytes and apoptosis. Complete separation of skeletal elements occurs through cavitation, thereby forming the joint capsule (Francis-West et al., 1999a). At the molecular level, joint formation correlates with downregulation of type II collagen, and with the appearance of specific regions with the expression of bone morphogenetic protein (BMP) family members, such as Gdf5, Gdf6, Bmp7 and Bmp2(Francis-West et al., 1999a; Storm and Kingsley, 1999; Merino et al., 1999) and Wnt14 (Wnt9a – Mouse Genome Informatics)(Hartmann and Tabin, 2001). In spite of the specific expression of BMP family members in developing joints,the function of these molecules remains obscure. Overexpression experiments with Bmp2, Bmp4, Bmp7, and Gdf5 genes or knockout of BMP-antagonist Noggin cause intense growth of the cartilage and inhibition of joint formation (Duprez et al., 1996; Storm and Kinsley,1999; Brunet et al.,1998). By contrast, deletion of Gdf5 or Gdf6genes in single or double knockout results in fusion of joints(Settle et al., 2003). By contrast to BMPs, Wnt14 (Wnt9a – Mouse Genome Informatics), a member of the Wnt gene family(Katoh, 2002; Quian et al.,2003), was identified as a primary inducer of joints(Hartmann and Tabin, 2001). Other signaling molecules may also be involved in joint formation. Indian hedgehog (IHH) is a major regulator of cartilage differentiation(Vortkamp et al., 1996; St-Jacques et al., 1999; Minina et al., 2001);furthermore, sonic hedgehog (Shh) or Ihh misexpression inhibits joint formation (Merino et al.,1999).

The important structural function of the extracellular matrix (ECM) in the mature cartilage suggests that the components of the ECM and integrins(receptors for ECM proteins) (Bokel and Brown, 2002; Hynes,2002) exert a major role in cartilage differentiation. However,experimental evidence supporting this idea is scarce. Alterations in ECM components such as collagens (Li et al.,1995; Barbieri et al.,2003), link protein (Wai et al., 1998) or aggrecan(Watanabe and Yamada, 1999)are responsible for some structural changes in cartilage. Also chondrocyte differentiation occurs in the absence of β1 integrins, but these cells are unable to form columns in growth plates. These chondrocytes show reduction of proliferation (Aszodi et al.,2003; Fässler and Meyer,1995). In addition, mutant mice lacking α1, α2,α6, αv, β3, or β5 integrins do not show skeletal malformations (Bouvard et al.,2001), suggesting that other integrins control skeletal development or that functional redundancy may occur.

Since the role of integrins on skeletal development is poorly understood,and much less for joint formation, we investigated the effects ofα5β1 integrin inhibitors and BMP on joint formation in the embryonic mouse limb culture system. Blocking β1 and α5 integrins with specific antibodies or RGD peptides (arginine-glycine-aspartatic acid) in the developing mouse zeugopod led to inhibition of pre-hypertrophic chondrocyte differentiation and ectopic joint formation between proliferating chondrocytes and hypertrophic chondrocytes. Moreover, ectopic joint resulted in expression of Wnt14, Gdf5, chordin, autotaxin and type I collagen and CD44, while expression of IHH and type II collagen was downregulated. Also, misexpression of α5β1 integrin in chick embryos resulted in joint fusion and formation of pre-hypertrophic chondrocytes. In addition, we investigated the effects of BMPs on ectopic joint formation induced by the inhibition of α5β1 integrin. This is the first report on the involvement of α5β1 integrin in the control point by which chondrocytes take the decision between the cartilage differentiation program and the joint formation program.

Reagents

The following antibodies were used: anti-β1 integrin monoclonal antibody (HMβ1), anti-α5 integrin (HMα5) and anti-α5 integrin (MFR5) from Pharmingen; anti-type I collagen and anti-CD44 from Chemicon; anti-type II collagen (CIIC1) and anti-human α5β1 integrin (BIIG2) from Hybridoma Bank; and anti-type X collagen from Calbiochem. The RGE (GRGESP) and RGD (GRGDNP) peptides were obtained from GIBCO BRL. BMP7 and SHH were from R&D Systems.

Microinjections and mouse limb organ cultures

Complete embryos at embryonic day (E) 14.5 were injected at the wrist with 2 μl of antibody solution or just with PBS using a pneumatic microinjector model PV830 (World Precision Instruments) equipped with glass microneedles. Then embryonic forelimb organ cultures were prepared as described by Vortkamp et al. (1996) and Lanske et al. (1996). Briefly, limbs were placed on a Nucleopore filter. Filter and limb were then placed on top of a stainless steel grid (Plain weave mesh 0.38 mm; Goodfellow) inside a well of a 24-well tissue culture plate containing BGJb medium with 1% bovine serum albumin. The grid keeps the limb at the liquid/air interface. Limb cultures were then kept for 4 days at 37°C in a 5% CO2 incubator. For some experiments, cultures were supplemented with 1 μg/ml BMP7 or 400 ng/ml SHH.

Skeletal preparation and tissue staining

Cultured limbs were fixed overnight in 96% ethanol and stained as previously described (Otto et al.,1997).

For tissue staining, limbs were fixed in 4% paraformaldehyde at 4°C overnight, dehydrated and embedded in Paraplast by standard techniques. Tissue sections were then stained with Weigert's acid iron chloride haematoxilin and then stained with Fast Green. Subsequently, sections were stained with Safranin O.

cDNA probes and in situ hybridization

For in-situ hybridizations the following cDNA probes were used: mouse Gdf5 (Storm and Kingsley,1999); Wnt14 (Katoh,2002); chordin (Bachiller et al., 2003). For chicken embryos Wnt14, Gdf5 and Ihh cDNA probes were used. Fragments of autotaxin (accession number NM015744), Cd44 (accession number XM283773) and aggrecan(accession number L07049) mouse genes were obtained by RT-PCR. First-strand cDNA was synthesized with a First-strand cDNA Kit (Roche Applied Science) and 1 μg of RNA of mouse embryo at E12. The following primers (5′ to 3′) were used: autotaxin 5′ primer,5′-CAGCAAGTCGAATTAAGAGG-3′ and 3′ primer 5′-GGCCAGCGTATACAGATTAG-3′ (corresponding to region 127-692); Cd44 5′ primer 5′-AAGACTTGAACAGGACAGGA-3′ and 3′ primer 5′-AGAGATGCCAAGATGATGAG-3′ (corresponding to region 1747-2248); aggrecan 5′ primer 5′-GAGGAGCCATACACATCTTC-3′ and 3′ primer 5′-CACTGAGGTCCTCTACTCCA-3′ (corresponding to region 2506-3040). PCRs were performed in a total volume of 25 μl using Taq DNA polymerase(Invitrogen). The cycling conditions were 15 seconds at 94°C for denaturation, 30 seconds at 55°C for annealing, 1 minute at 72°C for elongation, and then 30 minutes at 72°C after the last cycle (35 cycles). The PCR products were cloned into pGEM T-easy (Promega). The authenticity of the fragments was confirmed by dideoxy sequencing.

Digoxigenin 11 UTP-labeled single-stranded RNA probes were prepared using a DIG RNA labeling kit (Roche Applied Science) according to the manufacturer's instructions. Tissue sections on slides were treated with 1 μg/ml proteinase K (preincubated at 37°C) for 5 minutes at room temperature,then washed with PBT and incubated with hybridization buffer at 65°C for 15 minutes, after which the corresponding Digoxigenin-labeled probe was added and incubated overnight at 65°C. Next morning, sections were washed and incubated with FITC-labeled anti-digoxigenin antibodies at 4°C overnight. Next day, sections were washed and observed under a Nikon Fluorescence Microscope Eclipse E600, and photographed with a Nikon digital camera Coolpix 995 (Nikon Inc).

Immunofluorescence and TUNEL

For immunohistochemical analysis, slides were incubated overnight with antibodies at 4°C. Next morning, slides were incubated with the corresponding FITC-labeled secondary antibody for 2 hours at 37°C and observed under a Nikon Fluorescence Microscope Eclipse E600, and photographed with a digital camera Nikon Coolpix 995 (Nikon Inc.). Apoptosis was detected by the TUNEL assay using an in-situ cell death detection Kit, TMR (Roche Applied Science) according to the manufacturer's instructions.

Electroporation of α5β1 integrin cDNA into chick embryos.

Fertilized White Leghorn chicken eggs (ALPES, Puebla Mexico) were windowed and staged according to Hamburger and Hamilton(1951). Non-linearized full-length human α5 chain of integrins cDNA clone (6 μg/μl;pECE+α5) and non-linearized full-length human β1 chain of integrins cDNA clone (6 μg/μl; pECE+β) were injected into the autopod of the hindlimb at HH stage 27 (Hamburger and Hamilton, 1951) with a microneedle. DNA was mixed with Chinese ink and mineral oil (1.5 μl of pECE+α5 plus 1.5 μl of pECE+β1 plus 0.5 μl of Chinese ink and 0.5 μl of mineral oil) and approximately 0.5-1.0 μl of mix were microinjected. Electroporation was performed at 28 volts with 10 pulses of 70 mseconds. After electroporation, chick embryos were returned to the incubator and analyzed by skeletal staining, histology,immunofluorescence and in-situ hybridization.

Inhibition of β1 integrins induces ectopic joint formation and ectopic expression of Wnt14 and Gdf5

Interaction of integrins with ECM during cartilage differentiation was evaluated by the use of a limb explant culture system(Minina et al., 2001). To determine whether integrins of the β1 family play a role on cartilage differentiation, we microinjected a specific blocking monoclonal antibody(HMβ1) against the β1 chain of integrins into the wrist region of a mouse embryo (E14.5) forelimbs in organotypic cultures. Control limbs treated with an irrelevant IgG showed normal morphology of all skeletal elements(Fig. 1A). By contrast,treatment of the forelimbs with HMβ1 resulted in the loss of some skeletal elements such as phalanges and carpals(Fig. 1E). A gap perpendicular to the long axis of the bone in the skeletal elements was observed at distal level in radio and ulna, suggesting the formation of an ectopic joint-like element (Fig. 1E, box). This was observed with a frequency of 62% (25 out of 40 experiments). The idea that the new structure was a joint was confirmed by histological analysis of tissue sections at the zeugopod region, which revealed a change in cell morphology similar to that observed in the interzone of developing joints(Fig. 1B,F). To evaluate whether these morphological and histological changes correspond with the formation of an ectopic joint, we analyzed the expression of Wnt14and Gdf5. Both early molecular markers were expressed in the ectopic joint of treated limbs (Fig. 1C-D,G-H). Tissue sections of zeugopod region in limbs previously treated with anti-β1 integrin HMβ1 showed reduced red staining,suggesting that cartilage was altered (Fig. 1F).

Fig. 1.

Blocking β1 integrin induces formation of an ectopic joint. Forelimbs of E14.5 mouse embryos cultured for 4 days were injected at the wrist with 10μg of an irrelevant IgG (A-D) or with 10 μg of an anti-β1 integrin mAb (E-H). (A,E) Complete limb stained with Alcian Blue/Alizarin Red to show skeletal elements. Inset shows the area in the red square. (B,F) Histological section of zeugopod region stained with Safranin O and Fast Green to show cartilage in red. (C,G) In-situ hybridizations for Wnt14. (D,H)In-situ hybridizations for Gdf5. (I) Histological sections of zeugopod region in limbs injected with 2 μg RGE peptide stained with Safranin O and Fast Green. (J) Histological sections of zeugopod region in limbs injected with 2 μg RGD peptide stained with Safranin O and Fast Green. (K,L) TUNEL assay for apoptosis, forelimbs injected at the wrist with 10 μg of an irrelevant IgG (K) or with 10 μg of an anti-β1 integrin mAb (L).

Fig. 1.

Blocking β1 integrin induces formation of an ectopic joint. Forelimbs of E14.5 mouse embryos cultured for 4 days were injected at the wrist with 10μg of an irrelevant IgG (A-D) or with 10 μg of an anti-β1 integrin mAb (E-H). (A,E) Complete limb stained with Alcian Blue/Alizarin Red to show skeletal elements. Inset shows the area in the red square. (B,F) Histological section of zeugopod region stained with Safranin O and Fast Green to show cartilage in red. (C,G) In-situ hybridizations for Wnt14. (D,H)In-situ hybridizations for Gdf5. (I) Histological sections of zeugopod region in limbs injected with 2 μg RGE peptide stained with Safranin O and Fast Green. (J) Histological sections of zeugopod region in limbs injected with 2 μg RGD peptide stained with Safranin O and Fast Green. (K,L) TUNEL assay for apoptosis, forelimbs injected at the wrist with 10 μg of an irrelevant IgG (K) or with 10 μg of an anti-β1 integrin mAb (L).

It is well known that several β1 integrins mediate their interaction with ECM proteins through the RGD sequence in their ligands(Hynes, 2002). The effect produced by anti-β1 integrin antibodies could be due to the blockage of integrin–ligand interactions, or to direct activation by binding of the antibody to the integrin at a site different from the ligand-binding site. To further explore whether this effect was caused by blockage of integrin–ligand binding, we injected RGD peptides in the wrist of developing limbs instead of the HMβ1. Treatment with the RGD peptide alone also resulted in the formation of an ectopic joint, while the control peptide RGE had no effect (Fig. 1I-J). In addition, a dramatic reduction in cartilage staining was also observed. In contrast, treatment with an RGE control peptide had no effect on the morphology of skeletal elements (not shown). These data suggested that β1 integrins needed to interact with a ligand in the extracellular matrix through an RGD sequence in order to maintain the normal structure of the cartilage. Apoptotic cells were evident in the ectopic joint that appeared in limbs treated with HMβ1(Fig. 1K-L) or RGD treatment(data not shown).

Inhibition of α5β1 integrins induces molecular changes characteristic of joint formation

Because one of the main integrins that bind RGD ligands is α5β1,we explore whether inhibition of integrin function by blocking the α5 subunit would have the same effect as the anti-β1 antibodies or the RGD peptide. Microinjection of the anti-α5 mAb HMα5 induced the formation of an ectopic joint (Fig. 2A). Another anti-mAb, MFR5, also gave origin to an ectopic joint(data not shown). Together, these data support the idea that α5β1 integrin binds to an RGD-contained ligand in the ECM of a developing limb in order to prevent the appearance of joints.

Fig. 2.

Blocking α5β1 integrin induces molecular changes characteristic of joint formation. Forelimbs of E14.5 mouse embryos cultured for 4 days were injected at the wrist with 10 μg of an anti-α5 integrin mAb(A-G,I,K,M,O). Control injected with 10 μg of an irrelevant IgG (H,J,L,N).(A) Histological sections of zeugopod region stained with Safranin O and Fast Green to show cartilage in red. In-situ hybridizations for Wnt14 (B), Gdf5 (C), chordin (D), autotaxin (E). Immunofluorescence for CD44(F), type III collagen (G), type II collagen (H-I), type I collagen (J-K),type X collagen (L-M) and Indian hedgehog (N-O). Immunofluorescence staining for IHH and for type I, II, III and X collagen is green; cell nuclei were stained with propidium iodide (red).

Fig. 2.

Blocking α5β1 integrin induces molecular changes characteristic of joint formation. Forelimbs of E14.5 mouse embryos cultured for 4 days were injected at the wrist with 10 μg of an anti-α5 integrin mAb(A-G,I,K,M,O). Control injected with 10 μg of an irrelevant IgG (H,J,L,N).(A) Histological sections of zeugopod region stained with Safranin O and Fast Green to show cartilage in red. In-situ hybridizations for Wnt14 (B), Gdf5 (C), chordin (D), autotaxin (E). Immunofluorescence for CD44(F), type III collagen (G), type II collagen (H-I), type I collagen (J-K),type X collagen (L-M) and Indian hedgehog (N-O). Immunofluorescence staining for IHH and for type I, II, III and X collagen is green; cell nuclei were stained with propidium iodide (red).

To confirm that integrin function inhibition by blockage of theα5β1 integrin induces molecular changes characteristic of joint formation, we analyzed molecular markers of the interzone. Since Wnt14 is able to induce prechondrogenic cells into the joint program(Hartmann and Tabin, 2001), we evaluated the expression of other molecular markers such as Gdf5,chordin, autotaxin, CD44 and type III Collagen in the ectopic joint induced by blocking α5β1 integrin (Fig. 2). Results showed that Wnt14(Fig. 2B), Gdf5(Fig. 2C), autotaxin(Fig. 2D), chordin(Fig. 2E), CD44 protein(Fig. 2F) or Cd44 gene (data not shown) and type III collagen (Fig. 2G) were expressed in the ectopic joint. To determine whether formation of ectopic joints is associated with downregulation of molecular markers of cartilage differentiation, we evaluated type II, I and X collagens,aggrecan and IHH expression. We observed changes in the ECM of the ectopic joint induced by blocking α5β1 integrins. Type II collagen, a typical marker of cartilage ECM, was downregulated(Fig. 2I) along the ulna of developing limbs, while type I collagen was upregulated, showing stronger expression in the newly formed joint (Fig. 2K). In control limbs treated with an irrelevant antibody, these changes were not observed (Fig. 2H and 2J). Moreover, we observed that aggrecan expression was downregulated at the level of the ectopic joint (data not shown), while type X collagen was observed only at the distal level of the ectopic joint,corresponding to hypertrophic cartilage as observed in controls(Fig. 2L-M). Similarly, the anti-α5 integrin HMα5 induced inhibition of IHH expression(Fig. 2O), while the irrelevant control antibody did not reduce the expression of IHH(Fig. 2N). Together, these data further confirm that the new structure is a new joint, in which chondrocytes dedifferentiate, changing their cellular fate. Likewise, by IHH and type X collagen analysis the ectopic joint was found at the boundary between proliferating chondrocytes and hypertrophic cartilage.

α5β1 integrin is downregulated during joint formation

Because our data suggested that α5β1 integrin ligation to an RGD ligand in the developing limb contributed to prevent the appearance of a new joint, we explored the pattern of expression of α5β1 during joint formation. The whole autopod of a mouse embryo at E14.5 was stained for theα5 integrin subunit. The α5 integrin was found in the all-forming digits of the hand (Fig. 3A). At the zone where the joints of the fingers are being formed, α5 integrin expression was not observed (Fig. 3A). Closer examination of the interzone confirmed that α5 integrin was not present in the forming joint(Fig. 3B). In addition, Wnt14 was expressed in the interzone where α5 integrin was not expressed (Fig. 3C). In the more advanced skeletal elements such the ulna, α5 integrin expression was evident in pre-hypertrophic chondrocytes and colocalized with IHH but not with type X collagen (Fig. 3D-F). In addition, α5 integrin was seen in less intensity in proliferating chondrocytes, and was evident in perichondrium and joints formed (Fig. 3D-F).

Fig. 3.

Pattern expression of integrins during joint formation. Forelimbs of E14 mice were fixed and tissue sections prepared as described in Materials and methods. Sections at the level of autopod were stained for immunofluorescence with polyclonal antibodies against the α5 integrin subunit. (A) Section of the forming fingers stained for α5 integrins. (B,C) Consecutive sections from a joint in the developing fingers showing immunofluorescence staining for α5 integrin (B) and in-situ hybridization for Wnt14 (C). Serial sections of the ulna stained for α5 integrin(D), type X collagen (E) and IHH (F).

Fig. 3.

Pattern expression of integrins during joint formation. Forelimbs of E14 mice were fixed and tissue sections prepared as described in Materials and methods. Sections at the level of autopod were stained for immunofluorescence with polyclonal antibodies against the α5 integrin subunit. (A) Section of the forming fingers stained for α5 integrins. (B,C) Consecutive sections from a joint in the developing fingers showing immunofluorescence staining for α5 integrin (B) and in-situ hybridization for Wnt14 (C). Serial sections of the ulna stained for α5 integrin(D), type X collagen (E) and IHH (F).

BMPs enhance formation of the ectopic joint induced by inhibition ofα5β1 integrin

BMPs are known to inhibit joint formation and to enhance cartilage differentiation (Duprez et al.,1996; Storm and Kinsley,1999; Brunet et al.,1998), so we evaluated whether the addition of BMP could rescue the appearance of the ectopic joint that was induced by blockingα5β1 integrins. Mouse embryo forelimbs at E14.5 were injected at the wrist with anti-integrin antibodies and then cultured in the presence of BMP7 for 4 days. Limbs injected with an irrelevant antibody and incubated with BMP7 presented a normal distribution of skeletal elements(Fig. 4A), although they seemed thicker that those in control limbs (Fig. 1A). Unexpectedly, we found that BMP in the culture milieu did not revert the effect of specific blocking antibody to α5β1 integrin. Instead, BMP7 enhanced the formation of an ectopic joint at the distal region of radio and ulna in 87% (35 out of 40) of the experiments. BMP7 also led to loss of some skeletal elements such as phalanges and carpals(Fig. 4D) compared with controls (Fig. 4A). Histological analysis of this section showed two areas of cartilage staining,one more intensely stained toward the distal portion and a second one less stained toward the proximal portion (Fig. 4E). The last area resembles hypertrophic chondrocytes. These different areas are not observed in control limbs treated with BMP7 and an irrelevant antibody (Fig. 4B). In addition, the level of expression of Wnt14, a molecular marker for joints, was upregulated in the area of this new structure(Fig. 4F), whereas no Wnt14 could be detected in the control limbs(Fig. 4C). RGD-peptide treatment and BMP induced the formation of an ectopic joint(Fig. 4G). Moreover, an increase in apoptotic cells was evident in the ectopic joint treated with HMα5 compared with BMP7 alone (Fig. 4H-I). These results showed that the new structure formed after injection of the anti-α5 mAb was indeed a new joint, and BMP7 was not able to inhibit the formation of this ectopic joint after integrin blockage.

Fig. 4.

BMP enhances formation of the ectopic joint induced by inhibition ofα5β1 integrin. Forelimbs of E14.5 mouse embryos cultured for 4 days in the presence of 1 μg/ml BMP7 were injected at the wrist with 10 μg of an irrelevant IgG (A,B,C,H,J,L,N) or with 10 μg of an anti-α5 integrin mAb (D,E,F,I,K,M,O). (A,D) Complete limb stained with Alcian Blue/Alizarin Red to show skeletal elements. (B,E) Histological sections stained with Safranin O and Fast Green to show cartilage in red. (C,F) In-situ hybridization for Wnt14. (G) Histological sections of limbs injected with 2 μg RGD peptide stained with Safranin O and Fast Green to show cartilage in red. (H,I) TUNEL assay for apoptosis. (J,K) Immunofluorescence staining for type II collagen (green). (L,M) Immunofluorescence staining for type I collagen (green). (N,O) Immunofluorescence staining for Indian hedgehog(green). In all cases cell nuclei are stained with propidium iodide (red).

Fig. 4.

BMP enhances formation of the ectopic joint induced by inhibition ofα5β1 integrin. Forelimbs of E14.5 mouse embryos cultured for 4 days in the presence of 1 μg/ml BMP7 were injected at the wrist with 10 μg of an irrelevant IgG (A,B,C,H,J,L,N) or with 10 μg of an anti-α5 integrin mAb (D,E,F,I,K,M,O). (A,D) Complete limb stained with Alcian Blue/Alizarin Red to show skeletal elements. (B,E) Histological sections stained with Safranin O and Fast Green to show cartilage in red. (C,F) In-situ hybridization for Wnt14. (G) Histological sections of limbs injected with 2 μg RGD peptide stained with Safranin O and Fast Green to show cartilage in red. (H,I) TUNEL assay for apoptosis. (J,K) Immunofluorescence staining for type II collagen (green). (L,M) Immunofluorescence staining for type I collagen (green). (N,O) Immunofluorescence staining for Indian hedgehog(green). In all cases cell nuclei are stained with propidium iodide (red).

We showed above that type II collagen is downregulated after integrin blockage by monoclonal antibodies. Although BMP7 did not prevent the formation of the ectopic joint, it was possible that the distribution of other molecular markers of cartilage could be different. BMP7 alone caused an increase in type II collagen expression in the cartilage (compare Fig. 2H and Fig. 4J). However, treatment of anti-α5 integrin in the presence of BMP7 again caused a downregulation of type II collagen (Fig. 4K). In addition, the expression of type I collagen was upregulated after integrin blockage even in the presence of BMP7 (Fig. 4M), compared with BMP7 alone(Fig. 4L). Moreover, expression of IHH was increased by BMP7 alone (compare Fig. 2N and Fig. 4N). After treatment of anti-α5 integrin, inhibition of pre-hypertrophic chondrocyte differentiation was observed, as evaluated by expression of IHH(Fig. 4O). All these data together indicate that BMP7 enhances ectopic joint formation between proliferating chondrocytes and hypertrophic chondrocytes.

α5β1 integrin induces joint fusion and differentiation of pre-hypertrophic cells

Our results showed that inhibition of integrin α5β1 induces ectopic joints, so we evaluated whether misexpression of human α5β1 integrin inhibits joint formation. Full-length cDNA clones for human integrinα5 and β1 chain were electroporated into the autopod of embryonic chick legs at HH stage 27. After incubating chick embryos for 5 days, we found that misexpression of α5β1 integrin inhibited joint formation(Fig. 5A-B). Histological analysis of fingers confirmed continuity in the cartilage tissue between the phalanges (Fig. 5C-D). To determine that misexpression of α5β1 integrin occurred in the chick leg, we evaluated it by immunofluorescence with a monoclonal antibody specific for human α5β1 integrin, which does not crossreact with chick. The results showed expression of human α5β1 integrin(Fig. 5E-F). Since ectopic joint formation by inhibition of α5β1 integrin coincides with inhibition of pre-hypertrophic chondrocytes, we evaluated whether misexpression of α5β1 integrin up-regulates IHH, molecular marker of pre-hypertrophic chondrocytes. Results showed that expression of humanα5β1 integrin in the fingers colocalized with IHH(Fig. 5E-H). Wnt14 was not expressed under these conditions (Fig. 5I-J). In conclusion, misexpression of α5β1 integrin induces joint fusion and this phenotype correlates with differentiation of pre-hypertrophic chondrocytes.

Fig. 5.

Misexpression of α5β1 integrins induces fusion of joints and differentiation of pre-hypertrophic cells. Chick legs of HH stage 27 were electroporated with full-length cDNA α5β1 integrins in the region of the autopod and incubated for 5 days (B,D,F,H,J), as controls,contralateral legs were used (A,C,E,G,I). Chick legs were fixed in paraformaldehyde and photographed to observe phenotype induced by misexpression of human α5β1 integrin (A,B). Serial section of the legs shown in A and B were prepared for paraffin inclusion and used for histology (C,D), immunofluorescence for human α5β1 integrin (E,F),IHH (G,H) and in-situ hybridization for Wnt14 (I,J). (K-N) Forelimbs of E14.5 mouse embryos cultured for 4 days in the presence of 1 μg/ml BMP7 were injected at the wrist with 10 μg of an anti-α5 integrin mAb. Cultures were added with nothing (K,L), or sonic hedgehog 400 ng/ml (M,N). Histological sections were stained with Safranin O and Fast Green (K,M), or processed for in-situ hybridization for Wnt14 (L,N).

Fig. 5.

Misexpression of α5β1 integrins induces fusion of joints and differentiation of pre-hypertrophic cells. Chick legs of HH stage 27 were electroporated with full-length cDNA α5β1 integrins in the region of the autopod and incubated for 5 days (B,D,F,H,J), as controls,contralateral legs were used (A,C,E,G,I). Chick legs were fixed in paraformaldehyde and photographed to observe phenotype induced by misexpression of human α5β1 integrin (A,B). Serial section of the legs shown in A and B were prepared for paraffin inclusion and used for histology (C,D), immunofluorescence for human α5β1 integrin (E,F),IHH (G,H) and in-situ hybridization for Wnt14 (I,J). (K-N) Forelimbs of E14.5 mouse embryos cultured for 4 days in the presence of 1 μg/ml BMP7 were injected at the wrist with 10 μg of an anti-α5 integrin mAb. Cultures were added with nothing (K,L), or sonic hedgehog 400 ng/ml (M,N). Histological sections were stained with Safranin O and Fast Green (K,M), or processed for in-situ hybridization for Wnt14 (L,N).

Indian hedgehog inhibits the formation of the ectopic joint

Treatment with anti-α5 integrin downregulated expression of Indian hedgehog and misexpression of human α5β1 integrin upregulated IHH,so we evaluated whether Hedgehog signaling was required to prevent the formation of the ectopic joint. To do this, forelimbs injected with the anti-α5 integrin antibody and cultured in the presence of BMP7 were treated with sonic hedgehog (SHH), a protein that acts similarly to IHH(Vortkamp et al., 1996). Under these conditions, the appearance of the ectopic joint was blocked and pre-hypertrophic chondrocytes were present(Fig. 5K,M). In addition,expression of Wnt14 was inhibited(Fig. 5L,N). Similar results were obtained in cultures without BMP7 (data not shown).

In the present study we provide evidence for the first time thatα5β1 integrin ligation to an RGD ligand is involved in the control of joint formation and cartilage differentiation. Blocking of α5β1 integrin function resulted in the induction of an ectopic joint that expresses Wnt14, the earliest joint inducer(Hartmann and Tabin, 2001),and other specific markers of joints such as Gdf5, autotaxin, chordin and CD44 (Edwards et al.,1994; Bächner et al.,1999; Francis-West et al.,1999b; Merino et al.,1999; Storm and Kingsley,1999). This finding indicates that interactions between cells and ECM participate in the control of chondrocyte differentiation and joint formation. Furthermore, our results also suggest that joint formation and growth plate development are highly integrated processes.

A remarkable finding of our study, of interest to the understanding of the joint formation mechanisms, is the structural differences between normal and ectopic joints induced by integrin blockage. Normal joints are flanked by growth plates with proliferating chondrocytes closest to the articular ends,and hypertrophic chondrocytes further from the joints. By contrast, ectopic joints are induced between proliferating chondrocytes and hypertrophic chondrocytes. Possibly, when the cartilage condensations, which are entirely made up of proliferating chondrocytes, reach a particular size, for a reason that is still totally unknown, the cells in the center of the element exit the cell cycle (Karaplis et al.,1994; Schipani et al.,1997; Rossi et al.,2002). When they do, there are two alternative choices they can make. When the skeletal pattern is first being established, the condensations that reach this critical size initiate joint formation in their center. Later,once the skeletal pattern is established, when the skeletal elements again reach a critical size, the cells in the center, instead of becoming pre-joint cells, adopt the alternative fate of becoming pre-hypertrophic cells, leading down the path to hypertrophy and eventual ossification. So, the decision of which cell fate to adopt when exiting the cell cycle, pre-joint or pre-hypertrophic, could be explained on the basis of the presence or absence of α5β1 integrin in the perichondrial cells and/or postmitotic chondrocytes.

Findings of this study suggest that the perichondrium may also exert an important role in joint formation. Supporting this interpretation, Aszodi et al. (Aszodi et al., 2003) have observed that joints develop normally in β1 integrin-deficient mice under the control of the type II collagen promoter. It must be taken into account that type II collagen is expressed in all chondrocytes but not in perichondrium. So, the strong expression of α5β1 integrin that we observed in the perichondrium of skeletal elements beginning pre-hypertrophic differentiation is not affected by the absence of β1 integrin in type II collagen-expressing cells in β1 integrin-deficient mice. On this basis,the effect of α5β1 integrin on ectopic joint formation and inhibition of pre-hypertrophic cells might be mediated by the perichondrium. Moreover, rescue experiments of this study with SHH protein revealed that hedgehog signaling inhibits ectopic joint formation. Since the perichondrium expressed Patched, which is considered a major target of IHH signaling, it is likely that the effect of ectopic SHH was mediated by the perichondrium. Also, a direct effect of SHH on chondrocytes may explain inhibition of the ectopic joint (Long et al., 2001).

In accordance with the above-described hypothesis, the ectopic joint induced by blocking α5β1 integrin displayed the morphology of a gap perpendicular to the long axis of the bone. This joint pattern differs from that observed by Hartmann and Tabin(2001) in experiments misexpressing Wnt14. The joint induced by misexpression of Wnt14 is extended through the whole cartilage rudiment and is most probably due to instructive signals delivered by Wnt14 that induce downstream genes of the joint pathway in all cartilage cells misexpressing this gene. It is well established that the same integrin is capable of inducing very different cell responses depending on the cell type that expresses them. For example, α5β1 integrin binding to fibronectin in fibroblasts leads to cell survival(Aplin et al., 1999); in mammary cells, it leads to milk production(Schmidhauser et al., 1990);and in leukocytes, it leads to endothelial cell migration or cytokine production (Rosales and Juliano,1995). Thus, although the pattern of α5β1 integrin expression in the cartilage of the developing limb is rather dispersed, we do not necessarily expect as diffuse an effect as the one described after Wnt14 misexpression. The presence of α5β1 integrin in the perichondrium might influence the decision of which cell fate is adopted,either pre-joint or pre-hypertrophic, by proliferating chondrocytes when they exit the cell cycle in a local horizontal population of cells differentiating in concert, since blocking integrins produces joints perpendicular to the long axis of the skeletal element. In this sense, integrins modulate the response to other signals as suggested by Aplin et al.(Aplin et al., 1998).

An additional finding of our study is that the response of chondrocytes to BMP signaling is regulated by integrin. In normal developing limbs, BMP treatments cause intense chondrogenesis and inhibition of joint formation(Duprez et al., 1996; Macias et al., 1997; Brunet et al., 1998; Merino et al., 1999). By contrast, in our experiments, addition of BMP to the culture medium inhibited differentiation of pre-hypertrophic cells and enhanced formation of ectopic joints. This finding is in agreement with the normal expression of several bmp genes in the developing joints(Macias et al., 1997), as it would be difficult to explain why bmp genes are expressed in the joints if they block joint formation(Duprez et al., 1996; Brunet et al., 1998; Merino et al., 1999; Storm and Kingsley, 1999). According to our results, in the absence of α5β1 integrin BMP signaling can direct proliferating chondrocytes to the joint program. On the contrary, in the presence of α5β1 integrin, BMP signaling directs proliferating chondrocytes to the pre-hypertrophic program expressing Ihh (Macias et al.,1997). Whether this modulation of BMP response is mediated through the perichondrium, or whether it is a direct effect of BMPs on the chondrocytes, remains to be clarified.

We thank David Kingsley, Eddy de Robertis, Cliff Tabin and Juan Hurlé for donating the mouse Gdf5 and chordin and chicken Wnt14, Ihh and Gdf5 probes. Full-length human α5 chain of integrins cDNA clone and full-length human β1 chain of integrins cDNA clone were donated by Erkki Ruoshlati. The monoclonal antibody CIIC1 developed by R. Holmdahl and K. Rubin, and monoclonal antibody BIIG2 developed by C. Damski, were obtained from the Developmental Studies Hybridoma Bank, under the auspices of the NICHD and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, IA 52242. We also thank Marcia Bustamante Zepeda, Gerardo Arrellín Rosas, Georgina Diaz Herrera and Lucía Brito for technical assistance, and Marian Ros and Juan Hurlé for their helpful discussions and comments on the manuscript. Also we give thanks to the anonymous referees, because expert comments helped to improve this work. The authors thank Alberto J. Rios Flores, David Cruz Sánchez and René F. Abarca Buis for their help in the electroporation of chick embryos. Isabel Pérez Montfort corrected the English version of the manuscript. This work was supported by grants 34334-N (to J.C.-M.) and 36407-M(to C.R.) from Consejo Nacional de Ciencia y Tecnología (CONACyT),Mexico, and IN216701 (to J.C.-M.) from DGAPA, Universidad Nacional Autonoma de Mexico (UNAM). D.G.-C. was the recipient of a scholarship from CONACYT and UNAM and was supported by grant 202358 from PAEP, UNAM.

Aplin, A. E., Howe, A. and Juliano, R. L.(
1998
). Signal transduction and signal modulation by cell adhesion receptors: the role of integrins, cadherins, immunoglobulin-cell adhesion molecules, and selectins.
Pharm. Rev.
50
,
197
-263.
Aplin, A. E., Short, S. M. and Juliano, R. L.(
1999
). Anchorage-dependent regulation of the mitogen-activated protein kinase cascade by growth factors is supported by a variety of integrin alpha chains.
J. Biol. Chem.
274
,
31223
-31228.
Aszodi, A., Hunzinker, E., Brakebusch, C. and Fässler,R. (
2003
). β1 integrins regulate chondrocyte rotation,G1 progression, and cytokinesis.
Genes Dev.
17
,
2465
-2479.
Bachiller, D., Klingensmith, J., Shneyder, N., Tran, U.,Anderson, R.,Rossant, J. and De Robertis, E. M.(
2003
). The role of chordin/Bmp signals in mammalian pharyngeal development and DiGeorge syndrome.
Development
130
,
3567
-3578.
Bächner, D., Ahrens, M., Betat, N., Schroder, D. and Gross,G. (
1999
). Developmental expression analysis of murine autotaxin (ATX).
Mech. Dev.
84
,
121
-125.
Barbieri, O., Astigiano, S., Morini, M., Tavella, S., Schito,A., Corsi, A.,Di Martino, D., Bianco, P., Cancedda, R. and Garofalo,S. (
2003
). Depletion of cartilage collagen fibrils in mice carrying a dominant negative Col2a1 transgene affects chondrocyte differentiation.
Am. J. Physiol. Cell Physiol.
285
,
C1504
-C1512.
Bokel, C. and Brown, N. (
2002
). Integrins in development: moving on, responding to, and sticking to the extracellular matrix.
Dev. Cell
3
,
311
-321.
Bouvard, D., Brakebush, C., Gustafsson, E., Aszodi, A.,Bengtsson, T.,Berna, A. and Fässler, R. (
2001
). Functional consequences of integrin gene mutations in mice.
Circ. Res.
89
,
211
-223.
Brunet, L., McMahon, J., McMahon, A. and Harland, R.(
1998
). Noggin, cartilage morphogenesis, and joint formation in the mammalian skeleton.
Science
280
,
1455
-1457.
Duprez, D., Bell, E. J., Richardson, M. K., Archer, C. W.,Wolpert, L.,Brickell, P. M. and Francis-West, P. H.(
1996
). Overexpression of BMP-2 and BMP-4 alters the size and shape of developing skeletal elements in the chick limb.
Mech. Dev.
57
,
145
-157.
Edwards, J. C., Wilkinson, L. S., Jones, H. M., Soothill, P.,Henderson, K.J., Worrall, J. G. and Pitsillides, A. A.(
1994
). The formation of human synovial joint cavities: A possible role for hyaluronan and CD44 in altered interzone cohesion.
J. Anat.
185
,
355
-367.
Fässler, R. and Meyer, M. (
1995
). Consequences of lack of beta 1 integrin gene expression in mice.
Genes Dev.
9
,
1896
-1908.
Francis-West, P., Abdelfattah, A., Chen, P., Allen, C., Parish,J., Ladher,R., Allen, S., MacPherson, S., Luyten, F. and Archer,C. (
1999a
). Mechanisms of GDF-5 action during skeletal development.
Development
126
,
1305
-1315.
Francis-West, P., Parish, J., Lee, K. and Archer, C. W.(
1999b
). BMP/GDF-signalling interactions during synovial joint development.
Cell Tissue Res.
296
,
111
-119.
Hamburger, V. and Hamilton, H. L. (
1951
). A series of normal stages in the development of the chick embryo.
J. Exp. Morphol.
88
49
-92.
Hartmann C. and Tabin C. (
2001
). Wnt-14 plays a pivotal role in inducing synovial joint formation in the developing appendicular skeleton.
Cell
104
,
341
-351.
Hynes, R. (
2002
). Integrins: bidirectional,allosteric signaling machines.
Cell
110
,
673
-687.
Karaplis, A. C., Luz, A., Glowacki, J., Bronson, R. T.,Tybulewicz, V. L.,Kronenberg, H. M. and Mulligan, R. C.(
1994
). Lethal skeletal dysplasia from targeted disruption of the parathyroid hormone-related peptide gene.
Genes Dev.
8
,
277
-289.
Katoh, M. (
2002
). Molecular cloning and expression of mouse Wnt14, and structural comparison between mouse Wnt14-Wnt3a gene cluster and human WNT14-WNT3A gene cluster.
Int. J. Mol. Med.
9
,
221
-227.
Lanske, B., Karaplis, A., Lee K., Luz, A., Vortkamp, A., Pirro,A.,Karperien, M., Defize, L., Ho, C., Mulligan R. et al.(
1996
). PTH/PTHrP receptor in early development and Indian hedgehog-regulated bone growth.
Science
273
,
663
-666.
Li, S., Prockop, D., Helminen, H., Fässler, R.,Lapvetelainen, T., Kiraly,K., Peltarri, A., Arokoski, J., Lui, H.,Arita, M. et al. (
1995
). Transgenic mice with targeted inactivation of the Col2 alpha 1 gene for collagen II develop a skeleton with membranous and periosteal bone but no endochondral bone.
Genes Dev.
9
,
2821
-2830.
Long, F., Zhang, X. M., Karp, S., Yang, Y. and McMahon, A. P. (
2001
). Genetic manipulation of hedgehog signaling in the endochondral skeleton reveals a direct role in the regulation of chondrocyte proliferation.
Development
128
,
5099
-5108.
Macias, D., Gañan, Y., Sampath, T. K., Piedra, M. E.,Ros, M. A. andHurle, J. M. (
1997
). Role of BMP-2 and OP-1 (BMP-7) in programmed cell death and skeletogenesis during chick limb development.
Development
124
,
1109
-1117.
Merino, R., Macias, D., Gañan Y., Economides, A., Wang,X., Wu, Q.,Stahl, N., Sampath, K., Varona, P. and Hurle, J.(
1999
). Expression and function of Gdf-5 during digit skeletogenesis in the embryonic chick leg bud.
Dev. Biol.
206
,
33
-45.
Minina, E., Wenzel, H. M., Kreschel, C., Karp, S., Gaffield, W.,McMahon,A. and Vortkamp A. (
2001
). BMP and Ihh/PTHrP signaling interact to coordinate chondrocyte proliferation and differentiation.
Development
128
,
4523
-4534.
Otto, F., Thornell, A. P., Cromptonm T., Denzel, A., Gilmour, K. C.,Rosewell, I. R., Stamp, G., Beddington, R., Mundlos, S., Olsen, B. et al. (
1997
). Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development.
Cell
89
,
765
-771.
Qian, J., Jiang, Z., Li, M., Heaphy, P., Liu, Y. H. and Shackleford, G. M. (
2003
). Mouse Wnt9b transforming activity,tissue-specific expression, and evolution.
Genomics
81
,
34
-46.
Rosales, C. and Juliano, R. L. (
1995
). Signal transduction by cell adhesion receptors in leukocytes.
J. Leukocyte Biol.
57
,
189
-198.
Rossi, F., MacLean, H. E., Yuan, W., Francis, R. O., Semenova,E., Lin,C. S., Kronenberg, H. M. and Cobrinik, D.(
2002
). p107 and p130 Coordinately regulate proliferation, Cbfa1 expression, and hypertrophic differentiation during endochondral bone development.
Dev. Biol.
247
,
271
-285.
Schipani, E., Lanske, B., Hunzelman, J., Luz, A., Kovacs, C. S.,Lee, K.,Pirro, A., Kronenberg, H. M. and Juppner, H.(
1997
). Targeted expression of constitutively active receptors for parathyroid hormone and parathyroid hormone-related peptide delays endochondral bone formation and rescues mice that lack parathyroid hormone-related peptide.
Proc. Natl. Acad. Sci. USA
94
,
13689
-13694.
Schmidhauser, C., Bissell, M. J., Myers, C. A. and Casperson, G. F. (
1990
). Extracellular matrix and hormones transcriptionally regulate bovine b-casein 5′ sequences in stably transfected mouse mammary cells.
Proc. Natl. Acad. Sci. USA
87
,
9118
-9122.
Settle, S., Rountree, R., Sinha, A., Thacker, A., Higgins, K. and Kingsley,D. (
2003
). Multiple joint and skeletal patterning defects caused by single and double mutations in the mouse Gdf6 and Gdf5 genes.
Dev. Biol.
254
,
116
-130.
Shakibaei, M. (
1998
). Inhibition of chondrogenesis by integrin antibody in vitro.
Exp. Cell Res.
240
,
95
-106.
St-Jacques, B., Hammerschmidt, M. and McMahon, A.(
1999
). Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation.
Genes Dev.
13
,
2072
-2086.
Storm, E. and Kingsley, D. (
1999
). GDF5 Coordinates bone and joint formation during digit development.
Dev. Biol.
209
,
11
-27.
Vortkamp, A., Kaechoong, L., Lanske, B., Segre, G., Kronenberg,H. andTabin, C. (
1996
). Regulation of rate of cartilage differentiation by Indian Hedgehog and PTH-Related Protein.
Science
273
,
613
-622.
Wai, A., Ling, N., Watanabe, H., Yamada, Y., Tam, P. and Chea,K. (
1998
). Disrupted expression of matrix genes in the growth plate of mouse cartilage matrix deficiency (cmd) mutant.
Dev. Genet.
22
,
349
-358.
Watanabe, H. and Yamada, Y. (
1999
). Mice lacking link protein develop dwarfism and craniofacial abnormalities.
Nat. Genet.
21
,
225
-229.