Receptor-regulated and inhibitory Smads are critical in regulating transforming growth factorβ–mediated Meckel's cartilage development

Authors

  • Yoshihiro Ito,

    1. Center for Craniofacial Molecular Biology, School of Dentistry, University of Southern California, Los Angeles, California
    Search for more papers by this author
  • Pablo Bringas Jr,

    1. Center for Craniofacial Molecular Biology, School of Dentistry, University of Southern California, Los Angeles, California
    Search for more papers by this author
  • Ali Mogharei,

    1. Center for Craniofacial Molecular Biology, School of Dentistry, University of Southern California, Los Angeles, California
    Search for more papers by this author
  • Jingsong Zhao,

    1. Center for Craniofacial Molecular Biology, School of Dentistry, University of Southern California, Los Angeles, California
    Search for more papers by this author
  • Chuxia Deng,

    1. Genetics of Development and Disease Branch, NIDDK, National Institutes of Health, Bethesda, Maryland
    Search for more papers by this author
  • Yang Chai

    Corresponding author
    1. Center for Craniofacial Molecular Biology, School of Dentistry, University of Southern California, Los Angeles, California
    • Center for Craniofacial Molecular Biology, University of Southern California, 2250 Alcazar Street, CSA 103, Los Angeles, California 90033
    Search for more papers by this author

Abstract

The proper development of Meckel's cartilage is critical for craniofacial skeletogenesis, because it serves as the primordium for the formation of mandible, malleus, incus, and sphenomandibular ligament. Cranial neural crest (CNC) cells contribute significantly to the formation of Meckel's cartilage. Members of the transforming growth factor beta (TGF-β) family control the proliferation and differentiation of CNC cells during craniofacial skeletogenesis. TGF-β signaling is transduced from the cell membrane to the nucleus by means of specific type I and type II receptors and phosphorylated Smad proteins. Here we demonstrate that application of TGF-β promotes chondrogenesis by specifically increasing proliferation of CNC-derived chondrocytes and production of extracellular matrix. To understand the molecular regulation of TGF-β signaling, we have examined the biological function of both TGF-β receptor-regulated and inhibitory Smads during Meckel's cartilage development. The expression patterns of Smad2, 3, and 7 are identical to the ones of endogenous TGF-β and its cognate receptors during Meckel's cartilage development, establishing the potential that these intracellular signaling Smads may regulate TGF-β-mediated chondrogenesis. Functional haploinsufficiency of Smad2 delays TGF-β–mediated Meckel's cartilage development. Overproduction of Smad7 severely inhibits Meckel's cartilage formation, indicating a negative feedback on TGF-β signaling by inhibitory Smad is critical in orchestrating TGF-β–mediated gene regulation during embryonic chondrogenesis. The effectiveness of TGF-β signaling is highly sensitive to the level of Smad gene expression. © 2002 Wiley-Liss, Inc.

INTRODUCTION

Transforming growth factor-β (TGF-β) plays a pivotal role in regulating the proliferation and differentiation of chondrocyte precursors (Seyedin et al., 1987; Centrella et al., 1988; Sporn and Roberts, 1988). During craniofacial development, TGF-β ligands are expressed in a time- and tissue-specific manner and may regulate the contribution of cranial neural crest (CNC) cells during Meckel's cartilage development (Chai et al., 1994, 1998). However, regulation of intracellular TGF-β signaling during craniofacial skeletogenesis has not been well defined. Recent studies have shown that Smad proteins relay TGF-β signaling from the cell membrane to the nucleus. Extracellular TGF-β signals are transduced by means of membrane-bound TGF-β type II and type I receptors, which phosphorylate intracellular Smads. Activated Smad molecules regulate the expression of transcription factors and affect the transcriptional status of target genes (Massagué, 1998).

Different members of the Smad family have distinct signaling functions. Smad1, 2, 3, 5, and 8 interact with and are phosphorylated by specific type I serine/threonine kinase receptors, and thereby act in a pathway-restricted manner. In particular, Smad2 and Smad3 are phosphorylated and translocated to the nucleus after stimulation by TGF-β (Eppert et al., 1996; Zhang et al., 1996; Nakao et al., 1997a) or activin (Chen et al., 1996), whereas Smad1, 5, and 8 are activated after BMP stimulation (Hoodless et al., 1996; Liu et al., 1996; Kretzschmar et al., 1997; Suzuki et al., 1997). All these receptor-regulated Smads (R-Smads) are associated with either perichondrium or mature chondrocytes, implicating their active roles in mediating TGF-β or BMP-regulated chondrogenesis (Dick et al., 1998; Flanders et al., 2001). Specifically, the expression of Smad2 or Smad3 is more prominent in the maturing chondrocytes, indicating that TGF-β signaling actively regulates the proliferation of chondrocytes and extracellular matrix formation. Meanwhile, the expression of Smad1 or Smad5 is localized within the perichondrium, suggesting the activity of BMP is critical for differentiation of chondrocyte progenitors (Flanders et al., 2001).

Smad6 and Smad7 diverge structurally from other members of the Smad family (Hayashi et al., 1997; Imamura et al., 1997; Nakao et al., 1997a,b; Topper et al., 1997), and they function as inhibitors of TGF-β, activin, and BMP signaling. In particular, Smad6 appears to inhibit BMP signaling, whereas Smad7 associates stably with TGF-β receptor complex and inhibits TGF-β-mediated phosphorylation of Smad2 and Smad3. Because transcription of the inhibitory Smad gene is induced by stimulation of TGF-β (Nakao et al., 1997b; von Gersdorff et al., 2000), inhibitory Smads may produce autoregulatory negative feedback in signal transduction of the TGF-β superfamily.

Studies using targeted disruptions of Smad genes have revealed important biological functions of these intracellular signaling molecules. However, because of the early embryonic lethality associated with the Smad null mutation, such as Smad2-/-, it has not been possible to investigate the biological function of TGF-β–signaling Smad during organogenesis. Meanwhile, heterozygous loss of Smad2 has resulted in craniofacial malformation, suggesting the level of Smad expression may be critical in regulating TGF-β–mediated craniofacial development (Nomura and Li, 1998, Ito et al., 2001).

Here, we have demonstrated the dynamic distribution of CNC and non-CNC cells during Meckel's cartilage development and the biological function of TGF-β signaling in selectively regulating the proliferation of CNC-derived chondrocytes. Furthermore, we have investigated the function of TGF-β signaling Smads (Smad2 and Smad7) and provided the first biological evidence that Smad2 and Smad7 have important and opposing regulatory roles during Meckel's cartilage development. The regulatory function of TGF-β signaling Smad is highly sensitive to its expression level.

RESULTS

TGF-β Promotes CNC-Derived Chondrocytes Proliferation During Meckel's Cartilage Development

The development of Meckel's cartilage initiates at the molar tooth bud region with formation of bilaterally symmetrical cartilage rods. By labeling the progenies of cranial neural crest (CNC) cell using Wnt1-Cre and R26R transgenic model, we demonstrate that the aggregated CNC-derived cell mass is responsible for the initiation of chondrogenesis (Fig. 1A). After this initiation, Meckel's cartilage forms by extending both anteriorly and posteriorly with CNC cells (dark blue) at the chondrogenic front (Fig. 1B–D) to develop into a “wishbone-like” structure (Fig. 1E,F) that serves as the template for mandibular development. As embryogenesis continues, CNC-derived cells are mixed with non-CNC derived cells (pink) within Meckel's cartilage (Chai et al., 2000). To date, there is still debate regarding the origin of these non-CNC cells. Previous studies have suggested that ventrally emigrating neural tube (VENT) cells contribute significantly to the formation of Meckel's cartilage (Sohal et al., 1999). However, a recent study suggests that that there is no evidence of VENT cells from the mid- and hindbrain during chick embryogenesis (Yaneza et al., 2002). Further studies are necessary to address the origin of non–CNC-derived cells during mammalian Meckel's cartilage development. From the standpoint of analyzing the fate of CNC cells, it is conceivable, although unlikely, that these Wnt1-expressing cells represent only a subpopulation of CNC cells. Based on the observation that progeny of Wnt1-expressing cells are found in all predicted CNC-derived structures and there is virtually no ectopic lacZ expression, we conclude that Wnt1-Cre–mediated lacZ expression is associated with at least a significant population of CNC-derived cells during craniofacial development. Indeed, as shown in previous studies using the avian model, both CNC and non-CNC cells play a significant role during mouse craniofacial chondrogenesis (Noden, 1983; Couly et al., 1993; Sohal et al., 1999; Chai et al., 2000). Significantly, our study demonstrates for the first time that CNC-derived prechondrocytes are always present at the chondrogenic front and may help to establish the pattern of Meckel's cartilage.

Figure 1.

Contribution of cranial neural crest (CNC) and non-CNC cells during Meckel's cartilage development in cultured embryonic day (E) 11, 42-44 somite pairs) mandibular explants. A: At the initiation of Meckel's cartilage formation (E11+2 days in vitro), an aggregated CNC-derived cell (in blue, β-gal positive) mass begins to form Meckel's cartilage at the molar tooth bud region. B,C: At E11+2 and 3, Meckel's cartilage continues to form by anterior and posterior extension with CNC-derived cells at the chondrogenic front. Non-CNC cells (in pink) are mixed with CNC-derived cells. Arrows point to the regions where Smad7 adenovirus was injected (see Fig. 6). D: At E11+4, CNC-derived cells form posterior medial extension to guide the formation of middle ear bones. E: At E11+6, a “wishbone-like” Meckel's cartilage is formed with the bilateral inward curvature (ic) at the rostral region. F: At E11+9, mandible (double arrow) begins to form by using Meckel's cartilage as its template and contains both CNC- and non–CNC-derived elements. A cross-section of the rostral region shows large number of CNC-derived cells at the chondrogenic front (box) of Meckel's cartilage (mc), whereas there are mixtures of CNC- and non–CNC-derived cells in the middle segment (horizontal line) of Meckel's cartilage (mc).

Endogenous TGF-β signaling regulates the proliferation and differentiation of CNC-derived prechondroblast and control chondrogenesis (Chai et al., 1994). Here, we tested the effect of TGF-β on Meckel's cartilage development by focal application of TGF-β. Beads bearing 10 μg/ml TGF-β1 stimulated chondrogenesis and increased the thickness of Meckel's cartilage by 80–100% (n = 36, Fig. 2B,D), when compared with the bovine serum albumin (BSA)-treated controls (n = 40, Fig. 2A,C). To investigate what has contributed to the increase of Meckel's cartilage thickness, we examined the number of chondrocytes per 100,000 μm2 by randomly selecting five nonoverlapping sections adjacent to either control or TGF-β bearing bead at three time points after bead implantation. There was no difference in the number of chondrocytes per unit area 8 hr after either BSA or TGF-β bead treatment (Fig. 2E). Twenty-four hours after bead implantation, however, there was significant increase (P < 0.05) in the number of chondrocytes adjacent to TGF-β bearing bead when compared with the controls (Fig. 2E). The increase in number of chondrocytes by TGF-β treatment continued at 48 hr after bead implantation.

Figure 2.

Transforming growth factor beta (TGF-β) -promoted chondrogenesis. A: Beads bearing 0.1% bovine serum albumin (BSA) were implanted (arrow) in embryonic day (E) 11+2 mandibular explant and cultured for additional 7 days. B: Beads bearing TGF-β1 at 10 μg/ml were implanted (arrow) in E11+2 mandibular explants and cultured for additional 7 days. C: Whole-mount Alcian blue staining of Meckel's cartilage showed no alteration of chondrogenesis with BSA bead treatment. D: TGF-β1 bead promoted chondrogenesis by increasing the thickness of Meckel's cartilage. E: Cell count analysis indicated that TGF-β1 significantly increased the number of chondrocytes per unit area at 24 and 48 hr after implantation of TGF-β beads.

Next, we investigated whether CNC- or non–CNC-derived chondrocytes were affected by increased TGF-β signaling. By using Wnt1-Cre/R26R transgenic mouse mandibular explants, we have learned that application of TGF-β significantly increased the number of CNC-derived chondrocytes (Fig. 3B) in Meckel's cartilage when compared with the ones in the control group (Fig. 3A). The TGF-β signaling-mediated increase in the number of chondrocytes is due to a significant increase (P < 0.05) in the proliferation rate (27%) of CNC-derived chondrocytes (Fig. 3D) compared with the one (11%) in the control (Fig. 3C). Because we did not see any difference in the proliferation rate of non–CNC-derived chondrocytes between TGF-β treated and the control samples, we argue that TGF-β mediated increase in proliferation of CNC-derived chondrocytes is at least partially responsible for promoting Meckel's cartilage development. Meanwhile, we have also compared extracellular matrix formation in Meckel's cartilage between control and TGF-β bead-treated mandibular explants. TGF-β stimulated type II collagen formation within maturing chondrocytes (Fig. 3F) and type I collagen formation in perichondrium (Fig. 3H) of Meckel's cartilage. Thus, TGF-β promotes chondrogenesis by enhancing prechondrocyte differentiation, increasing chondrocyte proliferation and extracellular matrix formation in Meckel's cartilage. Next, to have a better understanding of the biological significance of intracellular TGF-β signaling mediators, we explored the function of both positive and negative TGF-β signaling Smads during Meckel's cartilage development.

Figure 3.

Overexpression of transforming growth factor (TGF-β) affected the proliferation of cranial neural crest (CNC) -derived chondrocytes and extracellular matrix formation. A: Cross-section of Meckel's cartilage (mc) of cultured Wnt1-Cre/R26R transgenic mandibular explant showed both CNC-derived (arrow, β-gal positive) and non–CNC-derived (pink) chondrocytes. The perichondrium (double arrow) contained mainly CNC-derived prechondrocytes. Bead (b) bearing 0.1% bovine serum albumin (BSA) was implanted adjacent to the forming Meckel's cartilage at E11+2. Mandibular explant was harvested 24 hr after bead implantation. B: Bead (b) bearing TGF-β1 (10 μg/ml) significantly increased the number of CNC-derived prechondrocytes (double arrow) in perichondrium and chondrocytes (arrow) in Meckel's cartilage (mc). C: A cross-section of Meckel's cartilage treated with BSA beads showed that very few chondrocytes were undergoing cell proliferation (arrow). D: There was a significant increase in proliferative activity of chondrocytes (arrow) and prechondroblasts (double arrow) when Meckel's cartilage was treated with TGF-β–bearing beads. E: BSA bead-treated Meckel's cartilage showed moderate amount of type II collagen (arrow). F: TGF-β1–treated Meckel's cartilage showed an increased production of type II collagen (arrow). G: BSA-treated Meckel's cartilage showed low-level type I collagen within perichondrium (arrow). H: TGF-β1–treated Meckel's cartilage significantly increased the amount of type I collagen within perichondrium (arrow).

Both Receptor-Regulated (Positive) and Inhibitory (Negative) Smads Are Associated With Meckel's Cartilage Development

Although both TGF-β- and BMP-specific Smads are present in cartilage, different pathway-specific Smads may have different functions in regulating chondrogenesis (Flanders et al., 2001). Here we focus on TGF-β signaling Smads. Both TGF-β receptor regulated and inhibitory Smads were expressed before and during formation of Meckel's cartilage (Fig. 4). In mandibular explants cultured for 6 days (E11+6), Smad2 was expressed in Meckel's cartilage and its perichondrium (Fig. 4A), whereas Smad7 was present predominately in perichondrium and less in maturing chondrocytes (Fig. 4B). The expression of Smad3 significantly overlapped with Smad2 expression (data not shown). At E11+9, phosphorylated Smad2 (PS2) was widely expressed in nuclei of maturing chondrocytes (Fig. 4C, arrow), indicating an active role of TGF-β signaling in regulating chondrocyte maturation. Smad7 expression was associated with Meckel's cartilage and its perichondrium (Fig. 4D). The in vivo distribution of Smad2 and Smad7 appeared to be comparable to the in vitro observation (data not shown). Collectively, our Smads expression analysis suggests that TGF-β signaling Smads are activated and may play important regulatory roles during Meckel's cartilage development.

Figure 4.

Both transforming growth factor beta (TGF-β) receptor-regulated and inhibitory Smads were expressed during Meckel's cartilage formation. Orange-red color denotes positive expression. A: At embryonic day (E) 11+6, Samd2 was expressed within Meckel's cartilage (mc) and its perichondrium (double arrow). B: Smad7 was expressed in Meckel's cartilage and more intensely localized within perichondrium (double arrow). C: At E11+9, phosphorylated Smad2 was localized within the nuclei of chondrocytes (arrow). D: Smad7 was expressed in chondrocytes (arrow) and perichondrium (double arrow).

Meckel's Cartilage Development Is Sensitive to the Expression Level of Receptor-Regulated and Inhibitory Smads

Targeted mutations of both Smad2 and Smad3 have been previously described (Nomura and Li, 1998; Yang et al., 1999). Smad3 homozygous mutant mice have impaired mucosal immunity and diminished T-cell responsiveness to TGF-β, but do not present any obvious developmental defects. Smad2 homozygous mutant embryos, however, fail to form an organized egg cylinder and lack mesoderm. Interestingly, some of the Smad2 heterozygous embryos have severely defective mandibles or eyes, suggesting that the dosage of Smad2 is critical for normal craniofacial development.

To investigate the functional role of TGF-β receptor regulated Smads (Smad2 and 3), E11 mandibular explants isolated from Smad2+/-, Smad3+/-, or Smad2+/- and Smad3+/- (Smad2+/-/Smad3+/- compound heterozygous mutant) mouse embryos were cultured in serumless, chemically defined medium for 9 days. All mandibular explants from Smad3+/- (n = 76) mouse embryos formed normal “wishbone-like” Meckel's cartilage, just as seen in the explants from wild-type littermate controls (Fig. 5A). In cultured mandibular explants obtained from Smad2+/- (n = 52) or Smad2+/-/Smad3+/- (n = 23) mouse embryos, however, approximately 40% of the Meckel's cartilage were significantly thinner and showed developmental delay (Fig. 5B) that represents an earlier stage of Meckel's cartilage development (Fig. 1E). Hence, functional haploinsufficiency of Smad2, but not Smad3, resulted in alteration of Meckel's cartilage development, indicating that the dosage of Smad2 is critical for TGF-β signaling during craniofacial morphogenesis. Additionally, the formation of mandible (Fig. 5B, double arrow) was delayed and displaced comparing with the one in wild-type sample (Fig. 5A, double arrow). The delay and disposition of the initial mandible formation might be due to malformation of Meckel's cartilage (primordium for mandibular development) and may account for malformation of mandible as reported in Smad2 heterozygous mutant mice (Nomura and Li, 1998). There was a direct correlation between the expression level of Smad2 and the delay in Meckel's cartilage formation. The affected Meckel's cartilage had a significant reduction (usually less than half) of Smad2 expression (Fig. 5C) when compared with the controls. And the Smad2 heterozygous mutant mandibular explants with at least half (or more than half) of the normal Smad2 expression level did not show any alteration of Meckel's cartilage development. Exogenous TGF-β1 was added into Smad2 heterozygous mutant mandibular explants culture medium and was unable to rescue the proper development of Meckel's cartilage (data not shown), indicating that adequate level of Smad2 expression was critical in regulating TGF-β–mediated chondrogenesis.

Figure 5.

Smad2 haploinsufficiency delayed Meckel's cartilage development. A: Embryonic day (E) 11 wild-type mandibular explant formed “wishbone-like” Meckel's cartilage after 9 days of culture in serumless chemically defined medium. The formation of mandible (double arrow) started on the lateral side of Meckel's cartilage. B: E11 mandibular explant from Smad2+/- or Smad2+/-/Smad3+/- mutant embryo showed a delayed Meckel's cartilage formation after 9 days in culture. The cartilage resembles the one seen in E11+6 days cultured mandibular explant (see Fig. 1E). The formation of mandible (double arrow) was delayed and displaced, which may account for the malformation of mandible as reported in Smad2 heterozygous mutant. C: Western analysis showed the expression level of Smad2 and Smad3 in wild-type and heterozygous mutant embryos.

Smad7 has been shown to have a distinct expression pattern that is different from TGF-β receptor regulated Smad during endochondral bone formation, suggesting this negative feedback may be important in regulating the function of TGF-β during skeletogenesis (Sakou et al., 1999). We designed experiments to test the regulatory function of Smad7 during Meckel's cartilage formation by introducing adenovirus (AdSmad7) to overexpress Smad7 in cultured mandibular explants. This AdSmad7 has been used successfully to block the phosphorylation of Smad2 upon TGF-β stimulation in various studies, demonstrating its effectiveness as a negative TGF-β signaling regulator (Nakao et al., 1999; Zhao et al., 2000). Our previous study has shown that adenoviral vectors are very effective in altering TGF-β type II receptor gene expression in cultured mandibular explants (Chai et al., 1999).

In this study, AdSmad7 was microinjected into the anterior region (rostral) of mandibular explants at the site of Meckel's cartilage formation 2 days after the initial onset of chondrogenesis (Fig. 1B). Mandibular explants injected with AdSmad7 showed a significant increase in Smad7 expression (4- to 5-fold, P < 0.05), as demonstrated by competitive polymerase chain reaction (PCR) (Fig. 6A), whereas mandibular explants treated with AdSmad6 did not affect the expression of Smad7. To examine the specificity of AdSmad7, we also examined the endogenous Smad3 and Smad2 expression, which was not affected (Fig. 6A, data not shown). Thus AdSmad7 was able to overexpress Smad7 within mandibular explants but did not affect the expression of other Smads. Next, to evaluate the effectiveness of injected viral construct to express the gene of interest at the site of chondrogenesis, adenovirus carrying the LacZ gene was injected into mandibular explants. These explants showed normal Meckel's cartilage formation (Fig. 6B) and efficient LacZ expression (inferred by β-gal expression) in Meckel's cartilage (Fig. 6B insert), indicating the transgene expression within prechondroblasts and chondrocytes. Overexpression of Smad7 at the anterior part of mandibular explant severely inhibited the formation of the rostral segment (arrow) of Meckel's cartilage (Fig. 6C). Thus inhibition of TGF-β signaling retarded the Meckel's cartilage formation, suggesting that changing the level of inhibitory Smad expression may prevent TGF-β's function to promote chondrogenesis.

Figure 6.

Overexpression of Smad7 inhibited Meckel's cartilage development. A: Competitive polymerase chain reaction showed that AdSmad7 (Smad7 adenovirus) injected into mandibular explants resulted in the specific overexpression of Smad7, whereas control virus (empty vector) and Smad6 adenovirus did not affect the expression of Smad7. Furthermore, other endogenous Smad (e.g., Smad3) expression was not affected by AdSmad7 treatment, indicating the specificity of this viral construct. B: Control virus-treated mandibular explant showed proper formation of Meckel's cartilage. LacZ expressing virus showed efficient β-gal expression within Meckel's cartilage (mc) and its perichondrium (arrow), indicating the effectiveness of virus treatment (insert). C: AdSmad7 injected into the rostral portion (arrow) of the explant severely inhibited the formation of Meckel's cartilage.

DISCUSSION

Mandibular development relies on the proper formation of Meckel's cartilage, which serves as the primordium of mandible and guides its early morphogenesis. TGF-β and its cognate receptors are important regulators during mandibular morphogenesis (Pelton et al., 1990; Chai et al., 1994, 1999). Targeted disruption of TGF-β signaling (such as TGF-β2 null mutant) results in multiple developmental defects, including a small mandible with its angle missing. Significantly, many of the affected tissues have neural crest-derived components and the TGF-β2 null mutant phenotypes simulate neural crest deficiencies (Sanford et al., 1997). By using the two-component genetic system (Wnt1-Cre/R26R), we demonstrate that CNC cells contribute significantly to the formation of Meckel's cartilage and ultimately to the formation of mandible. The presence of aggregated CNC cells at the chondrogenic front further indicates the significant role of these cells in initiation and successful completion of chondrogenesis and supports the concept that a critical cell mass plays a pivotal role for the initiation of skeletal development (Hall and Miyake, 2000). Our in vitro mandibular organ culture experiments demonstrate that addition of exogenous TGF-β1 promotes chondrogenesis by selectively increasing the proliferation of CNC-derived chondrocytes and producing additional extracellular matrix. Collectively, these experiments have revealed an important biological function of TGF-β signaling in regulating Meckel's cartilage development.

Members of the TGF-β superfamily are critical for skeletogenesis. For example, BMP4 induces Sox9 expression and promotes ectopic cartilage formation in cultured mandibular explants (Semba et al., 2000). The ability of BMP4 to induce chondrogenesis is position-dependent, suggesting that only a subpopulation of ectomesenchymal cells (most likely CNC-derived) within the first branchial arch are responsive to BMP4-mediated chondrogenesis. Other members of the BMP family are also expressed in the mandibular ectoderm with an overlapping pattern, thus, suggesting functional redundancy in BMP signaling-mediated chondrogenesis (for review, see Mina, 2001). Likewise, the expression pattern of all TGF-β isoforms also suggests functional redundancy. Because both TGF-β and BMP signaling pathways share a common downstream effector, it is quite possible that there is constant cross-talk between these two signaling events during Meckel's cartilage development.

Both receptor-regulated and inhibitory Smads are present within hyaline cartilage of the developing ribs of mouse embryo, indicating the possible regulatory functions of these Smads during chondrogenesis (Flanders et al., 2001). Both Smad2 and Smad3 are mainly associated with maturing chondrocytes, indicating that the TGF-β receptor-regulated Smads are critical for proliferation and terminal differentiation of chondrocytes. The dominant presence of Smad7 within perichondrium indicates that inhibition of TGF-β signaling may control the differentiation of prechondroblast into chondrocyte and help to set the boundary of Meckel's cartilage.

Targeted disruptions of Smad genes have revealed important biological functions of these intracellular signaling molecules. Many of the Smad null mutations, however, resulted in early embryonic lethality, indicating the biological importance of the Smad protein in regulating TGF-β signaling, although preventing the functional analysis of Smad molecule during organogenesis (Nomura and Li, 1998; Weinstein et al., 1998; Yang et al., 1999; and see review by Weinstein et al., 2000). By culturing mandibular explants from Smad mutant mouse embryos in a serumless, chemically defined medium, we were able to evaluate the biological function of Smads in regulating TGF-β–mediated Meckel's cartilage formation in vitro.

By using Smad2+/- and Smad3+/- heterozygous and Smad2+/-/Smad3+/- compound heterozygous mutant mouse embryos, we demonstrate that Smad2, but not Smad3, plays an important role in regulating TGF-β –mediated Meckel's cartilage development. Significantly, our results indicate that the dosage of Smad2 gene expression is critical for TGF-β signaling. In mandibular explants with Smad2 expression level below the threshold of being able to successfully transmit TGF-β signaling, the development of Meckel's cartilage is retarded. Haploinsufficiency of Smad2 has been shown to result in craniofacial malformations, affects the inhibitory effect of TGF-β on proliferation of enamel organ epithelial cells, and inhibits mandibular molar tooth development (Nomura and Li, 1998; Ito et al., 2001; Ferguson et al., 2001). Taken together, our study provides important evidence that a malformed mandibular primordium-the Meckel's cartilage, as the result of insufficient level of Smad2 expression, may be responsible for the mandibular malformation in Smad2 heterozygous mutant mice.

There was no alteration of Meckel's cartilage development in cultured Smad3+/- mandibular explants. Because of the high homology between Smad2 and Smad3 (>95%), it has been speculated that the shorter form of Smad2 (alternatively spliced Smad2 without exon 3) may function as Smad3 in transducing TGF-β signaling, thus can compensate for the Smad3 null mutation (Yagi et al., 1999). Recently, Smad3 has been shown to play a more important role in TGF-β–mediated pathogenetic events and be less involved during embryonic development (Ashcroft et al., 1999). Further studies are necessary to investigate the possible functional uniqueness of Smad3 in regulating TGF-β signaling.

The expression of Smad7 can be rapidly induced by addition of exogenous TGF-β, indicating the negative feedback on TGF-β signaling during embryogenesis (Stopa et al., 2000). This negative regulation on TGF-β signaling is likely achieved by the competitive binding between Smad2, 3, and Smad7 to the TGF-β type I receptor (Hayashi et al., 1997; Nakao et al., 1997b; Wrana, 2000; Massagué et al., 2000). Smad7 can bind onto the GS domain on type I receptor and prevent the phosphorylation of Smad2 upon activation by TGF-β ligand, thus functions as a negative regulator for TGF-β signaling (Nakao et al., 1999; Stopa et al., 2000). Interestingly, a recent study has shown that elevated Smad7 expression (e.g., induced by IFNγ) promotes Smad7–Smurf2 complex formation and, consequently, increases TGF-β receptor turnover (Kavsak et al., 2000). Here, we demonstrate that overexpression of Smad7 inhibits the morphogenesis of Meckel's cartilage, indicating the negative feedback on TGF-β signaling by Smad7 may play a critical role during organogenesis. Theoretically, targeted mutation of Smad7 may help us to have a better understanding of the biological function of this TGF-β signaling negative regulator.

Clearly, TGF-β signaling is mediated by both receptor-regulated and inhibitory Smads, and is tightly controlled temporally and spatially through multiple mechanisms at the extracellular, cell membrane, cytoplasmic, and nuclear levels (Massagué, 1998; Miyazono, 2000; Wrana, 2000; Massagué et al., 2000). The present study demonstrates that TGF-β promotes chondrogenesis by promoting the proliferation of CNC-derived chondrocytes. The TGF-β signaling is mediated by intracellular Smads in a dose-dependent manner. Functional haploinsufficiency of Smad2 causes delayed Meckel's cartilage development, which may be responsible for the mandibular defect seen in Smad2 heterozygous mutant mice. Smad7, however, may be critical for the confinement of TGF-β–mediated Meckel's cartilage development during craniofacial skeletogenesis. It is likely that Smads, together with other transcriptional activators/repressors, orchestrate TGF-β–mediated gene regulation to precisely control embryonic organogenesis. And this regulatory process is highly sensitive to the level of Smad expression.

EXPERIMENTAL PROCEDURES

Two-Component Genetic System for Marking the Progeny of CNC Cells

Both Wnt1-Cre transgenic line and R26R conditional reporter allele have been described previously (Danielian et al., 1998; Soriano, 1999). Mating Wnt1-Cre and R26R mice generated transgenic mice with progenies of neural crest cells labeled with β-gal. Cryostat sectioning and detection of β-galactosidase (LacZ) activities were done as previously described (Chai et al., 2000).

Organ Culture of Wild-Type, Smad2+/-, Smad3+/-, and Smad2+/-/Smad3+/- Mouse Mandibular Explants

All mouse embryos used in this study were maintained on C57BL6/J background. Timed-pregnant mice were sacrificed on postcoital day 11 (E11, 42–44 somite pairs). Embryos were staged according to the external developmental characteristics as described by Theiler (1989). The first branchial arch explants (eight per dish) were cultured for periods up to 9 days, according to the standard methods (Chai et al., 1994, 1998). Genotypes of Smad2 and Smad3 mutant mouse embryos were determined by PCR as previously described (Weinstein et al., 1998; Yang et al., 1999).

Preparation and Introduction of TGF-β Beads

Affi-gel blue beads (Bio-Rad), diameter 50–80 μm, were used. The beads were washed in phosphate buffered saline (PBS) and then incubated for 1 hr at room temperature in 10 μg/ml TGF-β1 (R&D). Control beads were incubated in 0.1% BSA. TGF-β1 or BSA containing beads was placed adjacent to the forming Meckel's cartilage (near the future molar region) 2 days after the start of mandibular organ culture. Then mandibular explants were collected at 8, 24, 48 hr, and 7 days after beads placement for chondrogenesis analysis.

Morphologic and Statistical Analysis

To evaluate the morphologic effect on chondrogenesis induced by implanted TGF-β beads, the number of chondrocytes per unit area was counted in randomly selected sections of Meckel's cartilage. Five sections were chosen adjacent to the implanted bead from each mandibular explant. Five mandibular explants (with a total of approximately 100,000 chondrocytes) were examined from each experimental group, and the number of chondrocytes per unit area was determined. ANOVA was applied to test for statistical significance in the variation of number of chondrocytes among 8-, 24-, and 48-hr samples. A P value of less than 0.05 was considered as statistically significant.

Whole-Mount Staining of Meckel's Cartilage

The three-dimensional architecture of Meckel's cartilage was examined by using a modified whole-mount Alcian blue staining, which stains the chondroitin 4- and 6-sulfate components in the cartilage (Chai et al., 1994).

Western Analysis

The total protein concentration in each mandibular sample was determined by comparing with BSA standards. Seventy-five micrograms of total protein from each sample was loaded in each well on a 12% polyacrylamide gel. Western analysis was done as previously described (Chai et al., 1999).

Smad7, Smad6 Adenovirus, and Their Administration Into Mouse Mandibular Explants

Replication-deficient recombinant adenoviruses expressing full-length Smad7 (AdSmad7) and Smad6 (AdSmad6) under the control of a murine cytomegalovirus (CMV) promoter (gift of Dr. K. Miyazono) were injected into mandibular explants to overexpress Smad7 and Smad6, respectively. Within the adenovector, FLAG-epitope sequence was fused to the N-terminus of Smad7 or Smad6 cDNA. AdSmad7 or AdSmad6 (1 × 1010 pfu/ml) was purified from human 293 cells as previously described (Zhao et al., 2000).

E11 mandibular explants were cultured in serumless, chemically defined medium for 2 days before the introduction of Smad adenovirus. AdSmad7, AdSmad6, or a control vector containing no exogenous gene in PBS was microinjected into mandibular explants at the site of chondrogenesis. High titer adenovirus (1 × 1010 pfu/ml) in a volume of 40 to 50 nl was injected into each mandibular explant. Then, the explants were cultured for another 7 days and harvested at E11+9 days for competitive PCR and whole-mount Meckel's cartilage staining.

Competitive PCR

Competitive PCR has been shown to successfully measure the change in mRNA expression level of mandibular explants after adenovirus treatment (Chai et al., 1999). Primers were designed based on murine Smad7 and Smad6 cDNA sequence and have been demonstrated to detect variation of Smad7 and Smad6 expression level after adenovirus treatment during lung morphogenesis (Zhao et al., 2000). Competitive PCR was performed according to the standard procedures (Chai et al., 1999; Zhao et al., 2000).

Immunohistochemistry

Sectioned immunohistochemistry was accomplished by following standard procedures (Chai et al., 1999). In particular, adjacent sections through the same Meckel's cartilage were mounted onto the same slide so that some of these sections were stained with anti-Smad2 antibody (Santa Cruz Biotechnology, CA), whereas the others were stained with anti-Smad7 (Dr. S. Souchelnytskyi) to examine the colocalization of different Smads. Dr. C-H. Heldin kindly provided the anti-phosphorylated-Smad2 antibody, which has been used successfully to identify the activated form of Smad2 (Nakao et al., 1999; Ito et al., 2001). Positive staining was indicated by orange-red coloration. The slides were counterstained with hematoxylin.

Evaluation of DNA Synthesis Activity During Meckel's Cartilage Formation

DNA synthesis activity within Meckel's cartilage was monitored in cultured mandibular explants with implanted beads. Twenty-four hours after the placement of beads, mandibular explants were incubated in serumless chemically defined medium containing BrdU (5-bromo-2′-deoxy-uridine, Sigma) at 100 μM for 2 hr at 37°C. After 2-hr culture with BrdU, the mandibular explants were harvested and fixed for immunostaining (Chai et al., 1999). BrdU-labeled cells and the total number of cells within Meckel's cartilage were counted from seven randomly selected sections per mandibular explant. Three mandibular explants were evaluated from each experimental group.

Acknowledgements

We thank Dr. Carl Heldin and Dr. Serhiy Souchelnytskyi for providing the various Smad antibodies and Dr. Kohei Miyazono for Smad7 and Smad6 adenoviruses. Mr. Julian Chen's artwork (Fig. 1) is greatly appreciated. Y.C. received funding from the National Institute of Dental and Craniofacial Research, NIH.

Ancillary