SEARCH

SEARCH BY CITATION

Keywords:

  • growth/differentiation factor 5;
  • connexin 43;
  • chondrogenesis;
  • heart development;
  • spine development;
  • tendon development

Abstract

  1. Top of page
  2. Abstract
  3. RESULTS
  4. DISCUSSION
  5. LITERATURE CITED

Growth/differentiation factor 5 (GDF5) regulates connexin expression and enhances embryonic chondrogenesis in a gap junction-dependent manner, suggesting that GDF5 action on developmental skeletogenesis is coordinated with gap junction activities. The results shown here demonstrate concordance between the mRNA expression profiles of GDF5 and the gap junction gene, Cx43, in the mouse embryonic limb, spine, and heart, consistent with coordinated functions for these gene products during developmental organogenesis. Anat Rec Part A 275A:1117–1121, 2003. © 2003 Wiley-Liss, Inc.

Growth/differentiation factor 5 (GDF5) is a key regulator of limb skeletal development (Storm et al., 1994; Storm and Kingsley, 1996). GDF5 mRNA expression has been localized to the condensing mesenchyme of the limb and in future joint spaces, directly correlating with the position, number, and timing of developing joints (Chang et al., 1994; Storm and Kingsley, 1996, 1999; Francis-West et al., 1999). Natural mutations in GDF5 result in the brachypod (bp) phenotype in mice (which consists of a normal axial skeleton and skull, with severely malformed limbs, joint fusions, absent skeletal elements, and bone spurs) and supporting a role for GDF5 in proper skeletal development and joint formation (Gruneberg and Lee, 1973; Storm et al., 1994; Storm and Kingsley, 1996; Clark et al., 2001; Mikic et al., 2001, 2002; Shum et al., 2003). In the first published description of the anatomy of the bp mouse, Gruneberg and Lee (1973) hypothesized that limb malformations resulted from decreased cellular ability to undergo condensation. In vitro, bp limb mesenchymal cells flatten and are unable to adhere to the tissue culture plate, which suggests that these cells are unable to adhere to and/or communicate with one another (Elmer and Selleck, 1975; Owens and Solursh, 1982). In further support of this hypothesis, we recently demonstrated that GDF5 regulates connexin gene expression, and that the requirement of gap junction communication in GDF5 stimulates cellular condensation during limb mesenchymal chondrogenesis (Chatterjee et al., 2003; Coleman and Tuan, 2003a).

Gap junctions are transmembrane channels made up of subunits known as connexins (Cx). Specifically, during limb patterning, gap junction-mediated interactions are critical factors in cellular synchronization and communication (Allen et al., 1990; Coelho and Kosher, 1991a, b; Laird et al., 1992; Dealy et al., 1994; Green et al., 1994; Meyer et al., 1997). It has been demonstrated that gap junctions play a role during precartilage mesenchymal condensation by compacting cells, presumably to aid in cellular communication and organization (Zimmermann et al., 1982). Connexin 43 (Cx43) has been implicated in early cartilage development. It has been hypothesized that Cx43-containing gap junctions aid in the organization and compartmentalization of tissues, and permit the transmission of instructive cues during patterning and differentiation (Ruangvoravat and Lo, 1992). Cx43 mRNA expression is observed in the condensing limb mesenchyme, but decreases as cells differentiate into chondrocytes, eventually localizing to the perichondrium. This suggests that Cx43-containing gap junctions play a role in compartmentalizing the differentiating chondrogenic core from adjacent nonskeletogenic tissues (Dealy et al., 1994; Green et al., 1994; Meyer et al., 1997; Lecanda et al., 1998; Donahue, 2000; Lecanda et al., 2000; Levin 2002).

To further assess the possible functional linkage between GDF5 and Cx43 during developmental chondrogenesis, we profiled GDF5 and Cx43 mRNA expression during organogenesis in mice to ascertain whether there is a similar spatiotemporal profile during the development of embryonic structures. CD1 mice (Charles River Laboratories, North Franklin, CT) were mated and staged. At 12.5, 13.5, 14.5, and 15.5 days postcoitus (dpc), the females were killed. Embryos were collected, fixed overnight in 4% paraformaldehyde, and processed for paraffin embedding and in situ hybridization according to the method of Wawersik and Epstein (2000). The protocol was approved by the institutional animal care and use committees of Thomas Jefferson University and the National Institutes of Health. For analysis of gene expression, [35S]-labeled antisense and sense probes were generated using cDNAs encoding the 3′UTR of mouse Cx43 (Ruangvoravat and Lo, 1992) or a fragment encoding the first 445 base pairs of the mouse GDF5 cDNA (kindly provided by Genetics Institute, Cambridge, MA). All transcription reactions were carried out in the presence of [35S]-UTP (NEN Life Science Products, Boston, MA). Sections were viewed using dark- and bright-field optics on a Jenaval microscope (Jena, Germany). Images were acquired utilizing a Kontron Progress 3012 digital camera (Munich, Germany)

Our results revealed similar expression patterns in the developing limb, tendon, spine, and heart, consistent with the hypothesis that GDF5 and Cx43-containing gap junctions act in conjunction with each other during organogenesis. In addition, we demonstrated for the first time GDF5 mRNA expression in the heart and the cartilage primordium of the neural arch.

RESULTS

  1. Top of page
  2. Abstract
  3. RESULTS
  4. DISCUSSION
  5. LITERATURE CITED

Cx43 and GDF5 Expression in the Developing Limb

GDF5 mRNA expression was localized to the condensing digit and long bones of the hindlimb at 12.5 dpc (data not shown). Cx43 transcripts were previously documented in the condensing mesenchyme of the limb at a similar stage (Meyer et al., 1997). GDF5 and Cx43 transcripts were both localized to the perichondral regions of the forelimb of a 13.5 dpc embryo, with GDF5 localization persisting to 14.5 dpc (Fig. 1A and C, and data not shown). GDF5 transcripts were also localized to the presumptive elbow joint (Fig. 1A) and hip joint surrounding the femoral head from 12.5 to 14.5 dpc (data not shown).

thumbnail image

Figure 1. Developmental expression of GDF5 and Cx43 mRNA in mouse embryonic limb, tendon, spine, and heart. Sections of mouse embryos at 12.5 (I and J), 13.5 (C, D, and K–P), 14.5 (A, B, E, and F), and 15.5 (G and H) dpc were probed with GDF5 and Cx43 antisense probes as indicated. Dark-field images and corresponding Mallory-stained closely adjacent sections are presented. In the developing limb (A–H), GDF5 transcripts localized to the elbow joint (A). Cx43 transcripts were observed in the perichondrium of the long bones (C). GDF5 transcripts localized to the tendons of the hip (E), just as Cx43 transcripts hybridized in the dermis and tendons of the foot (G). In the developing spine, GDF5 transcripts localized to the developing cartilage primordium of the neural arch (I). Cx43 transcripts were seen in the lining of the neural tube and surrounding the condensing vertebrae (K). In the developing heart, GDF5 transcripts localized to both the ventricle and the atrium of the heart (M). Additionally, as reported previously (Chang et al., 1994; Storm et al., 1994; Storm and Kingsley, 1996; Francis-West et al., 1999), GDF5 transcripts were detected in the shoulder and elbow joints and the perichondral regions of the digit. Cx43 mRNA was detected in the ventricle of the heart (O). Abbreviations: a, atrium; c, cartilage primordium of the neural arch; d, digit; de, dermis; e, elbow; f, femur; lu, lungs; n, neural tube; r/u, radius/ulna; te, tendon; v, vertebrae. Scale bar is 500 μm in all panels except E–H, where the scale bar is 5 mm.

Download figure to PowerPoint

The expression of GDF5 also mirrored that of Cx43 during tendon development. GDF5 mRNA was observed in the tendon of the hip in a 14.5 dpc embryo (Fig. 1E). Cx43 mRNA transcripts were detected in the dermis and surrounding the tendons in each toe of a 15.5 dpc mouse (Fig. 1G).

Cx43 and GDF5 Expression in the Developing Spine

Cx43 and GDF5 transcript expression patterns were similar during vertebral development. GDF5 mRNA signal was observed in the cartilage primordium of the neural arch from 12.5 dpc to 14.5 dpc (Fig. 1I and data not shown). Cx43 was observed in the spine of a 13.5 dpc embryo, specifically associated with the internal and external lining of the neural tube, as well as surrounding the condensing vertebrae (Fig. 1K). GDF5 and Cx43 transcripts are thus expressed in closely adjacent structures (the cartilage primordium of the neural arch and the vertebrae, respectively).

GDF5 and Cx43 Expression in the Heart at 13.5 dpc

A transverse section of a 13.5 dpc embryo revealed GDF5 expression in the heart (Fig. 1M). This pattern was observed only at 13.5 dpc of development. GDF5 transcripts were also localized in the expected pattern in the developing limb, including the shoulder and elbow joints, and the perichondrium of the radius/ulna and digits. Cx43 transcripts (Fig. 1O) were detected in the heart ventricle in a spatiotemporal expression pattern similar to that of GDF5.

DISCUSSION

  1. Top of page
  2. Abstract
  3. RESULTS
  4. DISCUSSION
  5. LITERATURE CITED

In this study, we profiled GDF5 and Cx43 mRNA expression during mouse organogenesis. The results showed a large degree of similarity in their spatiotemporal expression profiles during the development of embryonic structures, including the developing limb, tendon, spine, and heart, consistent with the hypothesis that GDF5 and Cx43-containing gap junctions act in conjunction with one another during organogenesis.

In the developing appendicular skeleton, expression of both GDF5 mRNA and Cx43 transcripts localizes to the condensing mesenchyme and the perichondrium, and in the interdigital mesenchyme (Meyer et al., 1997). This similarity in expression profiles supports our recent findings that GDF5 action depends on functional gap junction communication (specifically Cx43-containing junctions) during the condensation phase of endochondral ossification (Coleman and Tuan, 2003a). In vivo, gap junction-mediated communication at this stage may initiate and/or propagate the synchronization of cells in the condensing mesenchyme to initiate cellular differentiation, as well as segregate the maturing chondrocytes of the long bones from the adjacent mesenchyme by creating a cell communication barrier along the perichondral regions. However, GDF5 is expressed in the developing joint, where Cx43 is absent. It has been hypothesized that GDF5 influences chondrocyte maturation during joint development, possibly through a different pathway from that involved in precartilage condensation (Francis-West et al., 1999; Coleman and Tuan, 2003b).

Cx43 and GDF5 share a similar spatiotemporal profile of mRNA expression during tendon development in the limb. Supporting a role for GDF5 in tendon formation, ectopic GDF5 protein was previously demonstrated to stimulate neotendon growth (Wolfman et al., 1997). Additionally, studies in the bp mouse have also shown that in the absence of active GDF5, tendon placement is altered and the overall strength of the tendons is reduced (Gruneberg and Lee, 1973; Mikic et al., 2001). Therefore, we propose that GDF5 influences tendon development through a mechanism similar to that observed in chondrogenesis, including Cx43-containing gap junctions, to organize cellular populations and regulate their placement and maturation during development.

GDF5 and Cx43 are also coexpressed in immediately adjacent tissues during vertebral development. Cx43 expression is specifically localized to the lining of the neural tube and surrounding the condensing, differentiating vertebrae. GDF5 is expressed in regularly spaced intervals adjacent to the condensing vertebrae, localizing to the cartilage primordium of the neural arch. Although both Cx43 and GDF5 are expressed in the same general region at similar times, their expression is actually localized to two distinct structures. GDF5 in the cartilage primordium of the neural arch may function by activating cellular responses, including the expression of Cx43, to maintain the segmented vertebral structure. A similar mechanism of action during joint development has been proposed for GDF5 (Hartmann and Tabin, 2001; Coleman and Tuan, 2003b). Additionally, GDF5 may induce the expression of connexins other than Cx43 in the cartilage primordium of the neural arch, thereby stimulating a similar mechanism of cellular communication through gap junctions leading to the differentiation of the neural arch. The expression profiles of other connexin genes in this region have not been explored in this study.

GDF5 mRNA expression is observed in the developing heart, specifically at 13.5 dpc. Cx43 expression during cardiac development begins at 9.5 dpc and continues throughout organogenesis (Ruangvoravat and Lo, 1992); therefore, it is most likely regulated by a factor(s) other than GDF5. However, it is very interesting that GDF5, which is known for its chondro-stimulatory activity, is expressed during cardiac development. During endochondral ossification in the limb, GDF5 enhances cellular adhesion, proliferation, and maturation. It is possible that during cardiac development GDF5 plays a similar role in influencing cell adhesion and proliferation. Alternatively, GDF5 may regulate cellular synchronization by influencing gap junction-mediated cellular communication or cellular organization by regulating the expression of other connexins. This interesting finding warrants further exploration in the future.

In conclusion, we have demonstrated similar spatiotemporal developmental mRNA expression patterns of Cx43 and GDF5 in the developing limb, tendon, spine, and heart. These findings are consistent with the hypothesis that GDF5 influences cellular communication and the resultant synchronization and organization of tissues during differentiation and organogenesis, and does so through Cx43-containing gap junctions.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. RESULTS
  4. DISCUSSION
  5. LITERATURE CITED