Versican expression during skeletal/joint morphogenesis and patterning of muscle and nerve in the embryonic mouse limb
Article first published online: 4 JAN 2005
Copyright © 2005 Wiley-Liss, Inc.
The Anatomical Record Part A: Discoveries in Molecular, Cellular, and Evolutionary Biology
Volume 282A, Issue 2, pages 95–105, February 2005
How to Cite
Snow, H. E., Riccio, L. M., Mjaatvedt, C. H., Hoffman, S. and Capehart, A. A. (2005), Versican expression during skeletal/joint morphogenesis and patterning of muscle and nerve in the embryonic mouse limb. Anat. Rec., 282A: 95–105. doi: 10.1002/ar.a.20151
- Issue published online: 24 JAN 2005
- Article first published online: 4 JAN 2005
- Manuscript Accepted: 30 SEP 2004
- Manuscript Received: 16 AUG 2004
- National Institutes of Health. Grant Number: 1R15HD040846-01A1
- limb development;
- joint development;
- muscle migration;
- nerve migration
Versican, an extracellular matrix proteoglycan, has been implicated in limb development and is expressed in precartilage mesenchymal condensations. However, studies have lacked precise spatial and temporal investigation of versican localization during skeletogenesis and its relationship to patterning of muscle and nerve during mammalian limb development. The transgenic mouse line hdf (heart defect), which bears a lacZ reporter construct disrupting Cspg2 encoding versican, allowed ready detection of hdf transgene expression through histochemical analysis. Hdf transgene expression in whole mount heterozygous embryos and localization of versican relative to cartilage, muscle, and nerve tissues in paraffin-embedded limb sections of wild-type embryos from 10.5–14 days postcoitum were evaluated by lacZ histochemistry, immunohistochemistry, and in situ hybridization. Versican was localized within precartilage condensations and nascent cartilages with expression diminishing during maturation of the cartilage model at later time points. Interestingly, versican remained highly expressed in developing synovial joint interzones, suggesting potential function for versican in joint morphogenesis. Isolated myoblasts, incipient skeletal muscle masses, and neurites were not present in areas of strong versican expression within the developing limb. Versican-expressing tissues may reserve space for the future limb skeleton and developing joints and may aid in patterning of muscle and nerve by deterring muscle migration and innervation into these regions. © 2005 Wiley-Liss, Inc.
Limb development has been frequently used as a model for tissue patterning, particularly during skeletogenesis. Although many regulatory genes responsible for skeletal patterning have been identified, less is known about downstream molecules involved in chondrogenesis. Limb chondrogenesis occurs following mesenchymal cell aggregation into precartilage condensations (Solursh et al., 1978, 1982; Solursh, 1984; Hall and Miyake, 1995. Formation of a mesenchymal condensation is related to changes in the type of glycosaminoglycan (GAG) and proteoglycan secreted by prechondrogenic cells. Decreases in the concentration of hyaluronan and an increase in chondroitin sulfate permit precartilage aggregate formation and facilitate chondrogenesis (Hall and Miyake, 1992, 1995).
Several proteoglycans are represented in precartilage condensations, including core proteins bearing heparan sulfate and at least one with chondroitin sulfate (Shinomura et al., 1990). The most prevalent chondroitin sulfate proteoglycan in the precartilage limb bud is versican, a large hyaluronic acid-binding proteoglycan containing several different domains, including a hyaluronate-binding and two chondroitin sulfate attachment domains (GAG-α and -β) (Zimmerman and Rouslahti, 1989; Ito et al., 1995). Versican has been localized in the limb core and other sites undergoing cartilage differentiation in both the chick (Kimata et al., 1986; Zimmerman and Rouslahti, 1989; Shinomura et al., 1990) and mouse (Yamamura et el., 1997; Shibata et al., 2003). PG-M, the avian ortholog of mammalian versican, binds to hyaluronan, fibronectin, and type 1 collagen, all molecules implicated in precartilage aggregation, further supporting versican's participation in early chondrogenic events (Kimata et al., 1986; Yamagata et al., 1986, 1989).
In addition to possible roles in limb cartilage development, versican may play a role in patterning of other cell types, for example, that of muscle and nerve. Schramm and Solursh (1990) found that limb premuscle mass formation coincides with early chondrogenic events and suggested that cell surface or ECM molecules in mesenchymal precartilage condensations may be involved in eliminating myogenic cells from the limb core. Although specific molecules were not implicated, myoblast avoidance of cartilaginous regions in transfilter cultures in vitro suggested that either cell surface or ECM molecules were responsible (Schramm et al., 1994). Landolt et al. (1995) also found that in the chick, versican was localized in the pelvic girdle precursor avoided by extending axons and that axons later migrated into the limb through versican-negative gaps in this region. These studies suggest the possibility that versican may be involved in regulation of skeletal myoblast and axonal migration into the developing mammalian limb. However, support for versican's function in this regard has not been provided by spatial and temporal colocalization with skeletal muscle and nerve.
Recently, the transgenic line hdf (heart defect) was created by random insertion of a lacZ reporter construct into the mouse genome (Zhang et al., 1994; Yamamura et al., 1997). Hdf homozygous mutants exhibited severe segmental heart defects, which resulted in embryonic lethality at approximately 10.5 days postcoitum (dpc). Mjaatvedt et al. (1998) found that the hdf mutation occurred between exons encoding chondroitin sulfate attachment domains of Cspg2 (versican), implicating a critical role for this molecule in embryonic heart development. Hdf heterozygotes, which develop normally and live to reproductive maturity, permitted ready determination of the spatial expression of versican (Yamamura et al., 1997). Initial study of extracardiac hdf transgene expression by Yamamura et al. (1997) showed localization in early chondrogenic regions of the limb core, with diminution in maturing cartilages and a concurrent shift to less differentiated chondrocytes in the epiphyseal region of developing long bones, consistent with wild-type patterns of versican localization (Kimata et al., 1986; Shinomura et al., 1990; Shibata et al., 2003).
Limited studies of versican expression in the developing mammalian limb have been performed (Yamamura et al., 1997; Shibata et al., 2003) and, to our knowledge, none has reported its localization relative to the formation of the complete skeletal/articular template during patterning of skeletal muscle masses and in-growing axons. The present study investigated expression of versican in the developing limb of wild-type and hdf heterozygous embryos during the critical developmental period in which the cartilage template is fully established. We report that versican is expressed in the mammalian limb in a dynamic, developmentally regulated pattern that lends further support for versican function in early chondrogenesis and extends these findings to show that versican expression is also consistent with a role for this molecule in both synovial joint morphogenesis and facilitation of muscle and nerve patterning.
MATERIALS AND METHODS
Mouse embryos were harvested from pregnant females at 10.5, 11, 12, 13, and 14 dpc, following East Carolina University Institutional Animal Care and Use Committee (IACUC) approved guidelines (AUP D182). Thirteen to 15 pregnancies were used for each time point. Embryos were staged (Theiler, 1972) in order to ensure accurate developmental age. Critical stages of early limb development were targeted: 10.5 dpc (stages 16 and 17), initial limb bud outgrowth; 11 dpc (stage 18), initial precartilage condensations defined and axon outgrowth begun into the limb; 12 dpc (stage 20), muscle masses apparent and humeral cartilage established; 13 dpc (stage 21), cartilages present in most skeletal primordia and muscle masses prominent; 14 dpc (stage 22), maturing cartilage model distinguished and muscle masses well developed.
Embryos were harvested and placed directly into ice-cold phosphate-buffered saline (PBS). Embryos used for immunohistochemistry (with or without accompanying lacZ histochemistry) were fixed for 2 hr on ice in Dent's fixative (4:1 methanol:dimethylsulfoxide; DMSO) (Dent et al., 1987) containing 2 mM magnesium chloride and 5 mM EGTA. Samples used solely for lacZ whole mount processing were fixed 2 hr on ice in 2% paraformaldehyde-PBS fixative containing 2 mM magnesium chloride and 5 mM EGTA. Two different fixatives were used because although paraformaldehyde-fixed tissues displayed more intense lacZ staining compared to Dent's fixed tissues, several antibodies utilized did not react with paraformaldehyde-processed samples. Similar patterns of lacZ expression were observed with both fixatives. Dent's fixed embryos were incubated in 70% ethanol for 10 min prior to lacZ histochemistry. Embryos that were not used immediately were stored in 70% ethanol at 4°C.
Genotyping of hdf Mice
DNA from tail clips and embryos was extracted by adaptation of Laird (1991). PCR amplification utilized primers specific for the lacZ portion of the hdf transgene 5′-CGGCCAGGACAGTCGTTTGCCGTCTG-3′ and 5′-CCTGACCATGCAGAGGATGATGCTCG-3′ (Yamamura et al., 1997). PCR reactions consisted of the following cycles: 94°C for 4 min, 36 cycles of 95°C for 30 sec of denaturing, 58°C for 30 sec of annealing, and 72°C for 30 sec of extension, with 72°C final extension for 10 min. Known heterozygous hdf samples were used as positive controls. LacZ histochemistry was also used to identify hdf heterozygous embryos.
LacZ histochemistry was conducted by modification of Zhang et al. (1994). Embryos were washed three times (30 min for each wash) using PBS rinse containing 2 mM magnesium chloride, 0.01% sodium deoxycholate, and 0.02% NP-40. Samples were incubated at 37°C for 12 hr in PBS containing 2 mM magnesium chloride, 0.01% sodium deoxycholate, 0.02% NP-40, 20 mM sodium ferrocyanide, 20 mM sodium ferricyanide, and 0.1% X-gal (Invitrogen). Embryos were rinsed three times with PBS containing 10 mM EDTA (10 min per wash). Whole mount specimens were postfixed in 4% paraformaldehyde-PBS. Samples used for immunohistochemical analysis were rinsed in 70% ethanol for 10 min and postfixed in Dent's fixative. All embryos were postfixed for at least 8 hr. In selected experiments, embryos were dehydrated, embedded in paraffin, and sectioned at 7 μm. All embryos were sectioned transversely relative to the body axis and a minimum of 2–3 embryos at each time point was used to assess transgene expression. Wild-type embryos were used as controls for lacZ histochemistry.
In Situ Hybridization
Samples were fixed in 85% ethanol, 10% formaldehyde, 5% glacial acetic acid, dehydrated, and embedded in paraffin. Limb were sectioned transverse to the body axis at 8 μm and in situ hybridization was performed essentially as described by Breitschopf et al. (1992). Briefly, cDNA prepared by RT-PCR from 10.5 dpc total embryonic RNA was utilized to generate a 348 bp cDNA probe template specific for the Cspg2 exon 6/8 boundary (V1 splice form) using primers 5′-GCGACTGTTGGAGAACTTCAGG-3′ and 5′-TTCAAATGAGTCTGGTAACTCGGG-3′. Digoxygenin-labeled RNA probes were prepared from NcoI or NdeI linearized probe template in pGEMT (Promega) with the DIG RNA Labeling Kit (Roche Molecular Biochemicals) according to the manufacturer's instructions and labeled probe was quantified by dot blot. Hybridizations were performed ON at 55°C in 2 × SSC, 50% formamide, 10% dextran sulfate, 0.02% SDS, and 0.01% yeast tRNA containing either 200 ng/mL antisense and sense (negative control) probes. Specimens were washed thoroughly at 55°C in 1 × SSC, 50% formamide. Sections were incubated 1 hr at RT in blocking buffer (Roche) containing 10% fetal calf serum and 1% sheep serum and 1 hr in 1:1,000 sheep antidigoxigenin-alkaline phosphatase (Roche) in blocking buffer. Following TBS washes, bound probe was visualized using NBT/BCIP substrate (Roche).
Rabbit polyclonal antimouse antibodies were utilized that recognize peptides specific to the chondroitin sulfate attachment domains of versican: GAG-β-GST fusion protein antibody was prepared by standard methods and used at 2 μg/ml; versican GAG-α antibody purchased commercially (Chemicon) was used at 2.5 μg/ml. Biotinylated-peanut agglutinin (PNA; 16.7 μg/ml) and FITC-conjugated strepavidin (5 μg/ml; Vector Labs) were used to detect precartilage condensations (Zimmerman and Thies, 1984). Monoclonal antibodies utilized included antitype 2 collagen (II-II6B3; 1:20) for detection of overt cartilage matrix, antisarcomeric myosin (MF-20; 1:10) for localization of skeletal myoblasts and skeletal muscle masses, and antineurofilament-associated antigen (2H3; 1:20) for recognition of nerve. II-II6B3, MF-20, and 2H3 antibodies were obtained from the Developmental Studies Hybridoma Bank, University of Iowa.
Deparaffinized sections were treated with 10 mM sodium citrate in a pressurized decloaking chamber (BioCare Medical) for antigen retrieval (20 psi; 120°C for 20 min) prior to blocking. Sections to be stained with II-II6B3 were omitted from antigen retrieval due to loss of immunoreactivity following treatment. Immunohistochemical staining with all other antibodies showed similar localization when compared with untreated samples, but overall reactivity was enhanced by antigen retrieval, particularly for MF-20-stained tissues. Subsequent immunohistochemical staining procedures were a modification of Capehart et al. (1999). Briefly, sections were blocked in PBS containing 3% bovine serum albumin and 1% normal goat serum for 1 hr and incubated with primary antibodies ON at 4°C. Double labeling was performed routinely with antiversican in combination with other antibodies. Samples were washed four times (5 min per wash) with PBS and incubated 2 hr at RT with the fluorescein- or rhodamine-conjugated antimouse or -rabbit IgG secondary antibodies (ICN-Cappel) diluted 1:200 in blocking buffer. Samples were washed five times with PBS, postfixed in 80% and 50% ethanols (5 min each), reequilibriated in PBS, and mounted in 10% PBS-90% glycerol containing 100 mg/mL 1,4-diazabicyclo(2,2,2)octane (Sigma). In selected experiments, deparaffinized samples were pretreated with 0.1 U/mL chondroitinase ABC (Sigma) in 50 mM Tris-Cl, 60 mM sodium acetate, pH 8.0 (Linhardt, 1994), for 30 min at 37°C and washed three times with PBS (5 min each) prior to further processing.
Control experiments included omission of primary antibody and use of irrelevant primary antibodies. As primarily the versican GAG-β antibody was utilized for versican localization, additional controls included preincubation of versican GAG-β antibody (2 μg/mL) with versican GAG-β peptide-GST fusion protein (11 μg/mL), and GAG-β antibody (2 μg/mL) with versican GAG-α peptide-GST fusion protein (11.5 μg/mL) to ensure specificity of the GAG-β antibody. Preincubation of PNA lectin with 50 mM galactose to verify specificity of PNA binding was also performed and specimens were uniformly negative (not shown).
Whole mounts were examined using an Olympus SZ-60 stereo dissecting photomicroscope. Sectioned samples were viewed with an Olympus BX40 photomicroscope equipped with epifluorescence optics. Images were captured using SPOT RT software (Diagnostic Instruments) and overlaid using Adobe Photoshop (Adobe Systems). Identification of anatomical structures followed Kaufman (1999).
Major sites of versican expression during limb development were surveyed utilizing the hdf transgenic mouse line, which bears a β-galactosidase (lacZ) reporter transgene insertion between chondroitin attachment domains of versican (Yamamura et al., 1997; Mjaatvedt et al., 1998). In the developing limb, lacZ histochemical staining of whole mount heterozygous hdf embryos displayed weak versican transgene expression beginning at 10.5 dpc in the future pectoral girdle (Fig. 1A). At 11 dpc, increased hdf transgene expression was present in the pectoral girdle and had extended into precartilage mesenchyme of the limb core (stylopod/zeugopod; Fig. 1B), essentially the same pattern described by Yamamura et al. (1997). At 12 dpc, strong versican transgene expression was detected distally in mesenchymal condensations in the autopod (Fig. 1C), with reduced staining in differentiating cartilages of the proximal limb. At 13 and 14 dpc, in spite of reductions in proximal staining (Fig. 1D and E), strong transgene expression persisted in the autopod as described previously (Yamamura et al., 1997). Interestingly, intense hdf transgene expression was also observed in the future elbow region between 11 and 13 dpc with slight reduction in this location at 14 dpc. LacZ staining was also performed at 15 and 16 dpc, but there was no major difference in versican transgene expression patterns relative to 14 dpc (not shown).
In order to ensure that hdf transgene expression reflected accurately wild-type patterns of versican expression, in situ hybridization was performed. LacZ histochemistry and versican in situ hybridization patterns codistributed in the developing limb (Fig. 2A and B). Similar patterns were also detected using antiversican antibodies, further corroborating hdf transgene expression in the wild-type versican pattern. While lacZ reporter histochemistry provided a convenient method for hdf transgene detection that reflected areas of strong versican expression, in our hands it was overall a less sensitive measure of versican localization in tissue sections than either in situ hybridization or immunohistochemistry. In addition, X-gal reacted tissues often elevated nonspecific secondary antibody background in immunostaining experiments. Therefore, versican immunohistochemistry without prior lacZ staining was used to localize versican protein in subsequent experiments. Versican GAG-β and GAG-α peptide antibodies displayed similar staining patterns in the developing skeletal template, yet GAG-β staining was consistently more intense than GAG-α staining (Fig. 2C and D). Consequently, the GAG-β antibody was used for subsequent immunohistochemical detection of versican. Preincubation of GAG-β antibody with GAG-β peptide immunogen yielded no staining (Fig. 2F), while preincubation of GAG-β antibody with GAG-α peptide immunogen showed typical versican immunolocalization (Fig. 2G), confirming specificity of the versican GAG-β antibody. In order to determine whether changes in versican expression during the embryonic stages examined could be attributed to the masking of antibody binding sites by increasing matrix complexity, chondroitinase ABC treatment was also employed prior to further antigen retrieval. Chondroitinase-treated samples showed identical versican localization and only slight enhancement of GAG-β reactivity in 12- (Fig. 2H and I) and 15-dpc samples (not shown), suggesting that dynamic spatial and temporal changes noted in versican immunolocalization were due to developmental regulation.
Versican Expression During Formation of Limb Skeletal Template
In sectioned limb tissue, versican colocalized with PNA in prechondrogenic regions in the pectoral girdle and limb core at 11 dpc (Fig. 3A–C). There was no immunodetectable type 2 collagen-positive cartilage detected at 11 dpc (Fig. 3D). Versican codistributed with PNA at 12 dpc in the future shoulder and elbow joint regions and in precartilage aggregates in the distal limb (autopod; Fig. 3E and F). Versican expression diminished in differentiated cartilage, yet remained along the perichondrial periphery at 12 dpc. Slight overlap of versican expression and type 2 collagen in maturing cartilage was observed in the future shoulder joint between the proximal end of the humerus and distal scapula and in the future elbow joint interzone at the distal end of the humerus in 12 dpc embryonic limbs (Fig. 3G). Versican expression persisted along the periphery of maturing cartilage at 13 dpc (Fig. 3H and I). Versican immunostaining remained very strong in interzone regions of future synovial joints of the elbow (Fig. 3H–M), wrist (Fig. 3H and K), and digits (Fig. 3H and K). Overlap occurred between versican and type 2 collagen-positive cartilage in the coronoid process of the humerus and distal region of the ulna at 13 dpc (Fig. 3M), but versican reactivity was also intense in the type 2 collagen negative region between the articular ends of forming skeletal elements. At 14 dpc, versican reactivity again codistributed with PNA-positive cells in the metatarsophalangeal and interphalangeal joint interzones and persisted in the elbow joint interzone (articular cartilages) and future wrist joint interzone (Fig. 3N–R).
Relationship of Versican Localization to Limb Skeletal Muscle and Nerve Patterning
At 11 dpc, organized premuscle masses were limited to proximal sites of the trunk, corralled by regions of strong versican expression in the precartilage mesenchyme of the pectoral girdle and sclerotome (Fig. 4B). Distal to the shoulder at this stage, organized skeletal muscle masses were not found, but individual myoblasts were observed in versican-negative regions on the dorsal and ventral sides of the versican-positive precartilage limb core (Fig. 4C–E). At 11 dpc, nerves of the brachial plexus localized in areas of little or no versican expression within the pectoral girdle and branched around regions of high versican expression in the precartilage limb core (stylopod and future elbow region; Fig. 4F). Dorsal and ventral muscle masses had begun to organize in the developing limb along a proximal-to-distal gradient by 12 dpc and were located in regions that lacked strong versican expression, skirting the limb core and periphery of differentiating cartilage in the proximal region (Fig. 5A). Nerves had migrated distally and were also localized peripheral to regions of strong versican expression in the shoulder, future elbow, and perichondrial tissues (Fig. 5C). By 13 dpc, muscle masses near the future elbow were found in areas exhibiting weak versican staining, yet more distal premuscle masses were located in regions still devoid of versican expression (Fig. 6A–D). At 14 dpc, well-developed proximal skeletal muscle masses were again located in or bordered by regions of weak versican expression (Fig. 7A), and premuscle masses of the autopod were still excluded from areas of high versican expression (Fig. 7B and C). Nerve fibers traversing the elbow region were again observed in areas with little or no versican reactivity, most often located superficial to versican-positive perichondrial tissue and the joint interzone at 13 dpc (Fig. 6F). Similarly, axons were also found peripheral to versican-positive tissues alongside developing digits and developing interphalangeal joints at 14 dpc (Fig. 7D).
The present study was undertaken in order to examine the relationship between versican expression and development of the cartilaginous skeletal template in conjunction with the patterning of skeletal muscle and nerve during mammalian limb development to ascertain whether in addition to previously proposed roles during early chondrogenesis, versican could function in the organization of muscle and nerve during limb morphogenesis. Multiple isoforms of versican have been described in the developing limb (Kimata et al., 1986; Shinomura et al., 1995), and although the present study did not expressly address splice variant expression, use of a V1-specific probe for in situ hybridization in concert with antibodies directed against GAG-α- and GAG-β-specific peptides revealed that at a minimum, V0 and V1 versican forms were expressed in a similar pattern in the developing skeletal template. Whole mount observations utilized hdf transgene expression to monitor versican during the critical period in which the limb skeletal model is constructed. Detection of the hdf transgene in early chondrogenic tissues agreed with previously reported results (Yamamura et al., 1997) and was reinforced by versican protein immunolocalization, further corroborating use of the hdf transgenic line as a viable model for the study of versican. Interaction of versican with hyaluronan and CD-44 (Kawashima et al., 2000; Evanko et al., 1999) or its disruption of cellular interaction with fibronectin and collagen (Yamagata et al., 1986, 1989) has been suggested to promote mesenchymal aggregation and the assumption of the rounded morphology required for chondrocyte differentiation (Daniels and Solursh, 1991). Consistent with this hypothesis, versican was localized initially in precartilage mesenchymal condensations in the shoulder girdle and proximal limb core with expression progressing along a proximal-distal gradient concomitant with appearance of additional mesenchymal condensations as described previously in chick (Shinomura et al., 1990) and mouse (Yamamura et al., 1997; Shibata et al., 2003). Disappearance of versican from maturing mouse limb cartilages and sustained expression in the perichondrial periphery were also consistent with previous reports (Yamamura et al., 1997; Shibata et al., 2003), lending support to the hypothesis that versican must be downregulated for progression of terminal cartilage differentiation (Zhang et al., 1998). PNA binding in limb precartilage mesenchyme is not sensitive to either testicular hyaluronidase or chondroitinase ABC treatment, demonstrating that PNA does not recognize chondroitin sulfate GAG chains (Aulthouse and Solursh, 1987; Capehart et al., 1997). However, codistribution of antiversican and PNA in this study does not rule out the possibility that PNA recognizes other N- or O-linked galactose-N-acetylgalactosamine disaccharides on the versican core protein.
Interestingly, the present study provides new evidence regarding versican localization in developing synovial joints of the mammalian limb. Yamamura et al. (1997) showed that following disappearance of hdf transgene expression in the diaphysis, strong transgene expression was retained in a transitional zone at the end of the digital template between differentiated chondrocytes and undifferentiated free mesenchyme. Shibata et al. (2003) also noted continued versican localization in the tibial epiphysis following diminution in the diaphysis. Our results extend these findings and show that versican localization between the developing epiphyseal ends of the skeletal template extended well into type 2 collagen-negative areas of future joint interzones, regions involved in formation of the future synovial cavity, capsule, or synovial lining. Versican was present at 12 dpc in the future shoulder and elbow joints with persistent expression through 14 dpc in the elbow, wrist, and digit joint interzones along with weaker expression in the articular cartilage regions. The developing joint interzone is composed of multiple layers, two chondrogenic layers that surround articular cartilages, with a nonchondrogenic layer between consisting of stellate cells and enlarged cellular space (Mitrovic, 1978). Microfibrillar material in the joint space, which is morphologically similar to intercellular spaces of precartilage tissue, was suggested to be mucopolysaccharide (Mitrovic, 1977). Indeed, chondroitin-6-sulfate GAGs have been localized in the joint interzone, while chondroitin-4-sulfate is preferentially localized in the articular cartilage regions (Edwards et al., 1994; Archer et al., 1996), suggestive evidence for the presence of versican, which is enriched in chondroitin-6-sulfate (Yamauchi et al., 1997). Skeletal elements of the limb develop as continuous rods that must be separated in order for the joint to form, a process characterized by chondrocyte conversion to nonchondrogenic flattened cells (Mitrovic 1977, 1978; Shubin and Alberch, 1986; Archer et al., 1994; Francis-West et al., 1999). It has also been reported that cartilage-characteristic type 2 collagen expression diminishes with formation of the joint interzone, while type 1 collagen expression is present (Craig et al., 1987; Lizarraga et al., 2002). Other studies have found that CD44, hyaluronan, and uridine diphosphoglucose (UDPGD) are also present in the interzone (Craig et al., 1990; Archer et al., 1994; Edwards et al., 1994; Pitsillides et al., 1995; Dowthwaite et al., 1998). Hyaluronan and other GAGs may aid in joint formation by causing swelling pressure to protect against secondary fusion by opposing bones (Craig et al., 1990; Archer et al., 1994; Pitsillides et al., 1995). Hyaluronan-binding proteins are also present in the joint interzone and may aid in separation events (Dowthwaite et al., 1998). CD44 along with hyaluronan has been suggested to play a role in changes in fibroblastic cells, which may also be responsible for joint cavitation (Edwards et al., 1994). UDPGD in the joint interzone is integral to sulfated GAG synthesis (Archer et al., 1994; Pitsillides et al., 1995). Versican may function in joint morphogenesis by stabilizing hyaluronan or interactions between hyaluronan and CD44, thus aiding in formation of the synovial cavity. It is also interesting to speculate that the G3 domain of versican, which has been demonstrated to cause dedifferentiation of chondocytes into a flattened fibroblastic-like morphology (Zhang et al., 2001), may alter the shape of interzone cells. While the precise function of versican during embryonic joint formation is not yet clear, recent data have demonstrated versican production by human adult synovial cells with a positive correlation between IL-6 signaling and versican synthesis (Eklund et al., 2002).
Versican Expression and Patterning of Limb Skeletal Muscle and Nerve
Little is known regarding potential roles of versican in patterning of skeletal muscle myoblasts during their migration into the limb or during organization of premuscle masses. Schramm and Solursh (1990) suggested that myogenic cells in the chick limb redistribute from the precartilage core to the periphery. While molecules involved were not identified, this inhibitory influence was due to cell:cell or cell:matrix contact (Schramm et al., 1994). Versican expression along the lateral body wall was also suggested to regulate somitic myoblast migration (Henderson et al., 1997). In the present study, skeletal myoblasts organized in the limb along a proximal-distal gradient that lagged behind development of the cartilaginous template and initial muscle mass formation occurred in regions of little or no versican expression. Myoblast migration involves coordination of multiple events, including integrin-mediated cell-ECM contact (Velleman, 2002), and the antiadhesive properties of versican (Yamagata et al., 1986, 1989; Zhang et al., 1999) may be nonpermissive for sustained myoblast movement, resulting in net myoblast exclusion. This potential action of versican and/or other inhibitory substrates could limit skeletal muscle myoblasts to the limb periphery, facilitating their spatial organization during muscle mass formation. Following completion of the cartilaginous skeletal template, versican was expressed weakly in proximal muscle masses during stages corresponding to myotube fusion, consistent with a previous report of versican in the pericellular matrix surrounding myotubes in the chick (Carrino et al., 1999). Versican associated with maturing skeletal muscle masses may be involved in subdivision of the larger premuscle grouping into the individual skeletal muscles derived from it.
Several reports have provided evidence that versican serves a barrier function to axonal migration, although only limited investigation of versican's relationship to axonal patterning within the developing limb has been undertaken. In the present study, nerve fibers entered the limb flanked by regions of strong versican expression in the pectoral girdle similar to previous observations by Landolt et al. (1995) in the chick. Upon entry into the limb, branches of the brachial plexus diverged around the versican-rich prechondrogenic core. Throughout later limb stages, nerves were not present in regions of strong versican expression as detected with the GAG-β antibody, but instead resided superficial to the perichondral periphery along with developing muscle masses, consistent with axonal avoidance of developing cartilages. Developing joint regions were also avoided by nerve as well as myoblasts, suggesting that continued versican expression in these tissues could perhaps deter their invasion during early phases of joint morphogenesis. Interestingly, only the GAG-α antibody stained nerve fibers entering the limb (not shown), suggesting that the versican V2 splice form was expressed in those tissues, consistent with previous reports in the bovine nervous system where V2 has been demonstrated to regulate axonal growth (Schmalfeldt et al., 1998, 2000).
Previous studies have shown that barrier tissues to axonal migration exhibit alcian blue staining of GAGs, bind PNA, and contain chondroitin-6-sulfate (Tosney and Landmesser, 1985; Oakley and Tosney, 1991; Oakley et al., 1994), all of which describe precartilage regions where versican is present. In the central nervous system, versican expressed by glial cells may inhibit axonal migration and compromise repair of neural insult (Schmalfeldt et al., 1998, 2000; Fawcett and Asher, 1999; Niederost et al., 1999; Jones et al., 2003a, 2003b; Shearer et al., 2003). As with other tissues, it is possible that in the developing limb versican may interact with axon growth cone receptors, having a direct effect on nerve patterning. Alternatively, versican may exert an indirect effect on nerve by inhibiting interaction of axons with migration-permissive molecules, such as fibronectin and laminin (Yamagata et al., 1989; Braunewell et al., 1995), or by enhancing the antiadhesive effects of tenascin (Aspberg et al., 1995; Probstmeier et al., 2000), all of which are expressed in the core of the developing limb (Shinomura et al., 1990). Localization of postmigratory axons bordering regions of versican expression may also regulate potential growth of axon collaterals along the length of the nerve.
In summary, the present study shows that during formation of the limb skeletal template, versican expression is consistent with a role in precartilage aggregation and chondrogenesis. It also extends current knowledge regarding versican by demonstrating localization in the interzone of developing synovial joint tissues. Versican's temporal expression in the limb core precedes that of skeletal muscle mass and neural organization. Moreover, its expression in differentiating cartilage and sustained localization in perichondrium and future joint tissues are consistent with a possible role as a barrier to myoblast and axonal migration. Versican's dynamic localization may reserve space for the skeletal/joint template and deter muscle and nerve migration, which would aid in their ultimate patterning peripheral to the limb skeleton and articulations. Experiments are currently under way to examine the role of versican in synovial joint morphogenesis and skeletal/muscle/nerve patterning during embryonic limb development.
The authors acknowledge technical assistance from Edmond White and Kristen McGinty and the efforts of Jim Crosier and the staff of the Department of Biology Animal Facility at East Carolina University. This work was taken from a thesis by H.E.S. submitted to the faculty of the Department of Biology at East Carolina University in partial fulfillment of the requirements for the degree of master of science. Supported by National Institutes of Health grant 1R15HD040846-01A1 (to A.A.C.).
- 1994. Cellular aspects of the development of diarthrodial joints and articular cartilage. J Anat 184: 447–456. , , .
- 1996. The development of articular cartilage: II, the spatial and temporal patterns of glycosaminoglycans and small leucine-rich proteoglycans. J Anat 189: 23–35. , , , .
- 1995. The versican C-type lectin domain recognizes the adhesion protein tenascin-R. Proc Natl Acad Sci USA 92: 10590–10594. , , .
- 1987. The detection of a precartilage, blastema-specific marker. Dev Biol 120: 377–384. , .
- 1995. Functional involvement of sciatic nerve-derived versican- and decorin-like molecules and other chondroitin sulphate proteoglycans in ECM-mediated cell adhesion and neurite outgrowth. Eur J Neurosci 7: 805–814. , , , , , .
- 1992. In situ hybridization with digoxigenin-labeled probes: sensitive and reliable detection method applied to myelinating rat brain. Acta Neuropathol 84: 581–587. , , , , .
- 1997. Production of a monoclonal antibody by in vitro immunization that recognizes a native chondroitin sulfate epitope in the mebryonic chick limb and heart. J Histochem Cytochem 45: 1567–1581. , , , , .
- 1999. Dynamic expression of a native chondroitin sulfate epitope reveals microheterogeneity of extracellular matrix organization in the embryonic chick heart. Anat Rec 254: 181–195. , , , .
- 1999. Dynamic expression of proteogylcans during chicken skeletal muscle development and maturation. Poultry Sci 78: 769–776. , , .
- 1987. The spatial and temporal pattern of collagens I and II and keratin sulphate in the developing chick metatarsophalangeal joint. Development 99: 383–391. , , .
- 1990. A role for hyaluronan in joint development. J Anat 171: 17–23. , , , .
- 1991. Modulation of chondrogenesis by the cytoskeleton and extracellular matrix. J Cell Sci 100: 249–254. , .
- 1987. Whole mount analyses of cytoskeletal reorganization and function during oogenesis and early embryogenesis in Xenopus. In: SchattenH, SchattenG, editors. The cell biology of fertilization. San Diego, CA: Academic Press. p 63–103. , .
- 1998. An essential role for the interaction between hyaluronan and hyaluronan binding proteins during joint development. J Histochem Cytochem 46: 641–651. , , .
- 1994. The formation of human synovial joint cavities: a possible role for hyaluronan and CD44 in altered interzone cohesion. J Anat 185: 355–367. , , , , , , .
- 2002. Proteoglycan production in disomic and trisomy 7-carrying human synovial cells. Matrix Biol 21: 325–335. , , , , , .
- 1999. Formation of hyaluronan- and versican-rich pericellular matrix is required for proliferation and migration of vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 19: 1004–1013. , , .
- 1999. The glial scar and CNS repair. Brain Res Bull 49: 377–391. , .
- 1999. BMP/GDF-signaling interactions during synovial joint development. Cell Tissue Res 296: 111–119. , , , .
- 1992. The membranous skeleton: the role of cell condensations in vertebrate skeletogenesis. Anat Embryol 186: 107–124. , .
- 1995. Divide, accumulate, differentiate: cell condensation in skeletal development revisited. Int J Dev Biol 39: 881–893. , .
- 1997. Over-expression of the chondroitin sulfate proteoglycan versican is associated with defective neural crest migration in the Pax3 mutant mouse (splotch). Mech Dev 69: 39–51. , , .
- 1995. Multiple forms of mouse PG-M, a large chondroitin sulfate proteoglycan generated by alternative splicing. J Biol Chem 270: 958–965. , , , , .
- 2003a. The chondroitin sulfate proteoglycans neurocan, brevican, phosphocan, and versican are differentially regulated following spinal cord injury. Exp Neurol 182: 399–411. , , .
- 2003b. Axonal regeneration through regions of chondroitin sulfate proteoglycan deposition after spinal cord injury: a balance of permissiveness and inhibition. J Neurosci 23: 9276–9288. , , .
- 1999. The atlas of mouse development. New York: Academic Press. .
- 2000. Binding of a large chondroitin sulfate/dermatan sulfate proteoglycan, versican, to L-selectin, P-selectin, and CD44. J Biol Chem 45: 35448–35456. , , , , , .
- 1986. A large chondroitin sulfate proteoglycan (PG-M) synthesized before chondrogenesis in the limb bud of chick embryo. J Biol Chem 261: 13517–13525. , , , , , , .
- 1991. Simplified mammalian DNA isolation procedure. Nucl Acid Res 19: 4293. .
- 1995. Versican is selectively expressed in embryonic tissues that act as barriers to neural crest migration and axon outgrowth. Development 121: 2303–2312. , , , .
- 1994. Analysis of glycosaminoglycans with polysaccharide lyases. Curr Protocols Mol Biol 3: 1713B. .
- 2002. Studies on the role of Cux1 in regulation of the onset of joint formation in the developing limb. Dev Biol 243: 44–54. , , , .
- 1977. Development of the metatarsophalangeal joint of the chick embryo: morphological, ultrastructural and histochemical studies. Am J Anat 150: 333–348. .
- 1978. Development of the diarthrodial joints in the rat embryo. J Anat 151: 475–486. .
- 1998. The Cspg2 gene, disrupted in the hdf mutant, is required for right cardiac chamber and endocardial cushion formation. Dev Biol 201: 1–11. , , , , .
- 1999. Bovine CNS myelin contains neurite growth-inhibitory activity associated with chondroitin sulfate proteoglycans. J Neurosci 19: 8979–8989. , , , .
- 1991. Peanut agglutinin and chondroitin-6-sulfate are molecular markers for tissues that act as barriers to axon advance in the avian embryo. Dev Biol 147: 187–206. , .
- 1994. Glycoconjugates mark a transient barrier to neural crest migration in the chicken embryo. Development 120: 103–114. , , , .
- 1995. Alterations in hyaluronan synthesis during developing joint cavitation. J Histochem Cytochem 43: 263–273. , , , , .
- 2000. Chondroitin sulfates expressed on oligodendrocytes derived tenascin-R are involved in neural cell recognition: functional implications during CNS development and regeneration. J Neurosci Res 60: 21–36. , , , , .
- 1998. Versican V2 is a major extracellular matrix component of the mature bovine brain. J Biol Chem 273: 15758–15764. , , , .
- 2000. Brain derived versican V2 is a potent inhibitor of axonal growth. J Cell Sci 113: 807–816. , , , , .
- 1990. The formation of premuscle masses during chick wing bud development. Anat Embryol 182: 235–247. , .
- 1994. Role for short-range interactions in the formation of cartilage and muscle masses in transfilter micromass cultures. Dev Biol 163: 467–479. , , .
- 2003. The astrocyte/meningeal cell interface is a barrier to neurite outgrowth which can be overcome by manipulation of inhibitory molecules or axonal signaling pathways. Mol Cell Neurosci 24: 913–925. , , , , , , , .
- 2003. In situ hybridization and immunohistochemistry of versican, aggrecan, and link protein, and histochemistry of hyaluronan in the developing mouse limb bud cartilage. J Anat 203: 425–432. , , , , .
- 1990. The distribution of mesenchyme proteoglycan (PG-M) during wing bud outgrowth. Anat Embryol 181: 227–233. , , , , .
- 1995. The gene structure and organization of mouse PG-M, a large chondroitin sulfate proteoglycan. J Biol Chem 270: 10328–10333. , , , , .
- 1986. A morphogenic approach to the origin and basic organization of the tetrapod limb. Evol Biol 20: 319–387. , .
- 1978. A tissue culture analysis of the steps in limb chondrogenesis. In Vitro 14: 51–61. , , .
- 1982. Two distinct regulatory steps in cartilage differentiation. Dev Biol 94: 311–325. , , , , .
- 1984. Cell-matrix interactions during limb chondrogenesis: matrices and cell differentiation. New York: Alan R. Liss. p 47–60. .
- 1972. The house mouse: development and normal stages from fertilization to 4 weeks of age. New York: Springer Verlag. .
- 1985. Development of the major pathways for neurite outgrowth in the chick hindlimb. Dev Biol 5: 2345–2358. , .
- 2002. Role of the extracellular matrix in muscle growth and development. Am Soc Animal Sci 80: E8–E13. .
- 1986. Chondroitin sulfate proteoglycan (PG-M-like proteoglycan) is involved in the binding of hyaluronic acid to cellular fibronectin. J Biol Chem 15: 13526–13535. , , , , .
- 1989. Regulation of cell-substrate adhesion by proteoglycans immobilized on extracellular substrates. J Biol Chem 264: 8012–8018. , , , , .
- 1997. A heart segmental defect in the anterior-posterior axis of a transgenic mutant mouse. Dev Biol 186: 58–72. , , , .
- 1997. Identification, partial characterization, and distribution of versican and link protein in bovine dental pulp. J Dent Res 76: 1730–1736. , , , , .
- 1994. Ectopic Hoxa-1 induces rhombomere transformation in mouse hindbrain. Development 120: 2431–2442. , , , , , , , , , .
- 1998. The G3 domain of versican inhibits mesenchymal chondrogenesis via the epidermal growth factor-like motifs. J Biol Chem 273: 33054–33063. , , , , .
- 1999. Promotion of chondrocyte proliferation by versican mediated by G1 domain and EGF-like motifs. J Cell Biochem 73: 445–457. , , , , , .
- 2001. Versican modulates embryonic chondrocyte morphology via the epidermal growth factor-like motifs in G3. Exp Cell Res 263: 33–42. , , , , , , , , .
- 1984. Alterations of lectin binding during chondrogenesis in mouse limb buds. Histochemistry 81: 353–361. , .
- 1989. Multiple domains of the large fibroblast proteoglycan, versican. EMBO J 8: 2975–2981. , .