• Gli3;
  • Bmp;
  • Shh pathway;
  • limb development;
  • cell death;
  • pattern formation


  1. Top of page
  2. Abstract
  7. Acknowledgements

Removal of the posterior wing bud leads to massive apoptosis of the remaining anterior wing bud mesoderm. We show here that this finding correlates with an increase in the level of the repressor form of the Gli3 protein, due to the absence of the Sonic hedgehog (Shh) protein signaling. Therefore, we used the anterior wing bud mesoderm as a model system to analyze the relationship between the repressor form of Gli3 and apoptosis in the developing limb. With increased Gli3R levels, we demonstrate a concomitant increase in Bmp4 expression and signaling in the anterior mesoderm deprived of Shh signaling. Several experimental approaches show that the apoptosis can be prevented by exogenous Noggin, indicating that Bmp signaling mediates it. The analysis of Bmp4 expression in several mouse and chick mutations with defects in either expression or processing of Gli3 indicates a correlation between the level of the repressor form of Gli3 and Bmp4 expression in the distal mesoderm. Our analysis adds new insights into the way Shh differentially controls the processing of Gli3 and how, subsequently, BMP4 expression may mediate cell survival or cell death in the developing limb bud in a position-dependent manner. Developmental Dynamics 231:148–160, 2004. © 2004 Wiley-Liss, Inc.


  1. Top of page
  2. Abstract
  7. Acknowledgements

During limb development, the signaling protein Sonic hedgehog (Shh) is secreted by a specific group of mesodermal cells at the posterior border of the bud called the zone of polarizing activity (ZPA). A multitude of experiments indicates that Shh controls patterning in the anterior–posterior axis of the limb, although the mechanism for this process is not well understood. Shh has been proposed to act as a morphogen, and its ability to diffuse has been demonstrated (Lewis et al., 2001; Zeng et al., 2001). It has also been proposed that Shh acts indirectly through the establishment of a gradient of Bmp2 protein across the anterior posterior axis of the limb (Duprez et al., 1996; Yang et al., 1997). Alternatively, it has been proposed that Shh acts obligatorily through the transcription factor Gli3 (Wang et al., 2000; Litingtung et al., 2002; te Welscher et al., 2002).

The analysis of limb development in the absence of Shh demonstrates that Shh is dispensable for normal development up to the elbow/knee joint. Distal to this joint the limb of the Shh mutant mouse is deficient even though the proximodistal axis can be completely realized, at least in the hindlimb, where a single digit forms (Chiang et al., 2001; Kraus et al., 2001). The oligozeugodactyly (ozd) mutation in the chick is characterized by the absence of Shh signaling specifically in the limbs (Ros et al., 2003). The ozd limbs are comparable to the Shh mutant limbs and likewise, a single digit (digit one) develops in the ozd leg (Ros et al., 2003).

In Drosophila, cubitus interruptus (Ci) is the effector of hedgehog (Hh), the orthologue of Shh (Alexandre et al., 1996; Muller and Basler, 2000; Ingham and McMahon, 2001; reviewed in McMahon, 2000). Ci is a bipotential transcription factor that can activate or repress some of the same target genes. In the absence of Hh signaling, Ci is processed to a short form (Ci75) that is a strong repressor. In the presence of Hh, the full-length form of Ci (Ci155) is modified to constitute a transcriptional activator (Aza-Blanc et al., 1997; Methot and Basler, 2001). In vertebrates, three orthologues of Ci, called Gli1, Gli2, and Gli3, have been identified (Kinzler et al., 1988; Ruppert et al., 1988; Hui et al., 1994; Marigo et al., 1996; Schweitzer et al., 2000). During limb development, Gli1 is expressed in a posterior domain overlapping that of Shh, whereas Gli2 and Gli3 show complementary domains of expression with that of Shh (Hui et al., 1994; Marigo et al., 1996; Schweitzer et al., 2000). A processing of the protein similar to that of Ci has been demonstrated for Gli3, and possibly Gli2, but not Gli1 (Dai et al., 1999; Aza-Blanc et al., 2000; Wang et al., 2000).

While Gli1 mutant mice exhibit no developmental defects, Gli2 and Gli3 mutant mice show, among other deficiencies, a limb phenotype. Gli2 mutants show delay in the ossification of the digits (Mo et al., 1997). Gli3 mutants exhibit severe polydactyly, preferentially of the forelimb (Schimmang et al., 1992; Hui and Joyner, 1993). The differences in the limb phenotypes of the three Gli mutants may indicate a preferential and necessary requirement for Gli3 in limb development.

Of interest, the limbs of the double Gli3-/-;Shh-/- mutant are polydactylous and indistinguishable from that of the single Gli3-/- mutant (Litingtung et al., 2002; te Welscher et al., 2002). Similar to Ci in Drosophila, during vertebrate limb development, Shh signaling prevents the processing of the full-length Gli3 (Gli3-190) to a short form (Gli3-83) that functions as a strong repressor. In both mouse and chick limb buds, the repressor form of Gli3 (therefore, called Gli3R) is present in an anterior–posterior gradient with the highest levels in the anterior part of the limb bud where Shh signaling is at minimum (Wang et al., 2000; Litingtung et al., 2002). The genetic data of the Shh, Gli3, and double-compound mutants indicate that the phenotype in the absence of Shh is caused by an excess in the Gli3R form that suppresses gene expression, cell survival, and distal progression of limb bud development. Thus, the main function of Shh during limb development can be considered to prevent the processing of Gli3 to its repressor form.

Shh also plays an important role in regulating cell death during development (Sanz-Ezquerro and Tickle, 2000). Curiously, while Shh signaling acts as a survival factor in the anterior and middle limb bud cells, it promotes cell death in the posterior limb bud cells. In the absence of Shh, cell death is abnormally increased and could, at least partially, account for the phenotype (Chiang et al., 2001; Ros et al., 2003). The amount of cell death decreases in Shh mutants when a functional Gli3 allele is removed and is restored to normal when both Gli3 alleles are removed, permitting a direct correlation between increasing levels of Gli3R and increased cell death (te Welscher et al., 2002). Thus, a reasonable hypothesis is that increased Gli3R levels are responsible for the increased apoptosis observed in the absence of Shh signaling both in the limb bud and the neural tube as well, although no mechanisms have been proposed (Litingtung and Chiang, 2000; te Welscher et al., 2002).

Here, we used the anterior mesoderm, deprived of Shh signaling, as a system in which to investigate the correlation between Gli3R and cell death. It has been known since the experiments performed in the 1980s by the groups of Hinchliffe and Fallon that the anterior limb bud mesoderm requires signaling from the posterior mesoderm to survive (Todt and Fallon, 1987; Wilson and Hinchliffe, 1987). Deprived of Shh signaling, the anterior mesoderm suffers massive apoptosis that precludes further realization of its normal fate. We reasoned that the level of Gli3R would increase in the isolated anterior mesoderm, making this tissue an excellent system to analyze the relationship between Gli3 and apoptosis. Here, we show that the increase in the level of Gli3R in the anterior mesoderm is accompanied by an increase in Bmp4 expression and signaling. Blocking of Bmp signaling by Noggin prevents cell death. The analysis of Bmp4 expression in several mouse and chick mutants with altered Gli3 expression or processing indicates that it is downstream of Gli3R in the anterior–distal mesoderm. Our results are discussed in the context of Shh's control of cell survival and cell death in the developing limb bud.


  1. Top of page
  2. Abstract
  7. Acknowledgements

Processing of Gli3 to Its Repressor Form Increases in the Isolated Anterior Mesoderm Concomitantly With Excessive Cell Death

It is well known that the anterior half of the developing limb bud depends on the posterior mesoderm to develop and realize its fate (Hinchliffe and Gumpel-Pinot, 1981; Todt and Fallon, 1987; Wilson and Hinchliffe, 1987). When the posterior half of the chick wing bud (therefore, the entire ZPA) is removed, only a truncated humerus or occasionally an elongated element corresponding to a fused hypoplastic humerus and radius develops (Fig. 1A; Hinchliffe and Gumpel-Pinot, 1981; Todt and Fallon, 1987). If the anterior half developed in accordance with its prospective fate the radius and digit 2 should have developed (Bowen et al., 1989; Vargesson et al., 1998). However, if the anterior half of the wing is removed (Hinchliffe and Gumpel-Pinot, 1981; Todt and Fallon, 1987; Wilson and Hinchliffe, 1987) or separated by a barrier (Warren, 1934; Summerbell, 1979), the posterior mesoderm develops its fate forming a humerus, ulna, and digits 3 and 4 (Fig. 1A). Taken together, these data are interpreted to mean that removal of the ZPA brings about anterior cell death. However, the situation may be more complex than that (see Todt and Fallon, 1987).

thumbnail image

Figure 1. Removal of the posterior mesoderm is followed by massive cell death in the anterior mesoderm concomitantly with a twofold increase in the level of Gli3R. A: The anterior mesoderm did not achieve its prospective fate if isolated from the posterior mesoderm. Only a truncated humerus (h) or a rudimentary humerus and radius (r) form. However, the prospective fate of the posterior mesoderm was maintained in the absence of the anterior part. u, ulna; 2 and 3, digits 2 and 3. B: The isolated anterior mesoderm suffers abnormal apoptosis, predominantly following the anterior–distal border as shown by terminal deoxynucleotidyl transferase-mediated deoxyuridinetriphosphate nick end-labeling (TUNEL) at the times indicated. h, hours. C: Western blot with an antibody recognizing a 190-kDa band corresponding to full-length (Gli3-190) and two bands of 83 kDa and 75 kDa, respectively, corresponding to the processed repressor forms of Gli3, which is indicated as Gli3R. The level of the repressor band in the anterior mesoderm was sevenfold higher than in the posterior mesoderm. D: There was a twofold increase in the level of Gli3R in the isolated anterior limb bud as at 18 and 6 hr after the operation. E: The transcription of Gli3 was not appreciably altered in the anterior mesoderm after removal of the posterior half.

Download figure to PowerPoint

Thus, isolated from its posterior counterpart, the anterior mesoderm has a very limited development progressively tapering to form a spike, most likely due to the loss of tissue caused by massive apoptosis. We performed a sequential analysis of cell death using a terminal deoxynucleotidyl transferase-mediated deoxyuridinetriphosphate nick end-labeling (TUNEL) assay that confirmed the pattern of cell death previously described using vital dyes (Fig. 1B; Todt and Fallon, 1987; Wilson and Hinchliffe, 1987). Abnormal scattered dead cells were detected 6 hr after the surgery (not shown). By 12 hr after the operation, exaggerated cell death concentrated mainly along the anterior–distal region (Fig. 1B). Furthermore, increased cell death was also observed proximally. This pattern of cell death persisted at 18 and 24 hr after the operation and is likely the cause of the collapse of the anterior mesoderm. Residual cell death was still visible at the distal tip up to 48 hr (Fig. 1B).

During normal limb development, Shh activity dramatically prevents processing of Gli3 protein (Wang et al., 2000); this process results in reverse gradients of Gli3-190 (highest posteriorly) and Gli3R (highest anteriorly; Litingtung et al., 2002). Because the amount of Gli3R has been correlated with the amount of cell death (Litingtung and Chiang, 2000; te Welscher et al., 2002), we hypothesized that the increase of cell death after posterior mesoderm removal could be mediated by the increase in Gli3R after removal of Shh signaling.

To test this, the form of Gli3 protein in the cells of the isolated anterior limb bud was analyzed by Western blots using an affinity-purified Gli3ab1 antibody (Wang et al., 2000). By using lysates from anterior and posterior halves of stage 20 control wing buds, we found a difference of sevenfold in the amount of Gli3R present in the anterior vs. the posterior half of the wing, comparable with previously published results (Wang et al., 2000; Litingtung et al., 2002). It is striking that the anterior half develops normally in the presence of a considerable amount of Gli3R that is almost undetectable in the posterior half (Fig. 1C). Next, we performed Western blots using the anterior halves of operated wings at 18 hr after posterior mesoderm removal, when cell death was very apparent. The contralateral nonoperated limb was dissected into anterior and posterior halves that were collected separately and used as control. We report a twofold increase in the level of Gli3R in the isolated anterior limb bud cells18 hr after the removal of Shh signaling, when compared with the anterior half of the contralateral unaffected wing (Fig. 1D). This twofold increment in the level of Gli3R was present 6 hr after the operation, as demonstrated by the Western blot performed at this time, the earliest analyzed (Fig. 1D).

Because Shh has been shown to repress Gli3 transcription both in vivo and in vitro (Takahashi et al., 1998; Schweitzer et al., 2000; Wang et al., 2000), we also evaluated possible modifications in the transcription of Gli3 after the removal of the posterior half of the limb bud. We found that the level of Gli3 transcription, as assessed by in situ hybridization, was not appreciably changed (Fig. 1E), confirming that the increase in the level of Gli3R (Fig. 1D) was mainly due to an increase in the processing of Gli3 protein.

Our analyses show that the removal of the endogenous source of Shh has consequences in the processing of Gli3 in the anterior limb bud cells, indicating that Shh itself reaches the anterior half of the limb. Our results also support a role for Gli3R in the observed apoptosis of the anterior mesoderm, as we had hypothesized.

Exogenous SHH Prevents Cell Death and Processing of Gli3 in the Isolated Anterior Mesoderm

The ZPA, presumably acting through Shh, has been shown to improve survival and development of the isolated anterior mesoderm (Wilson and Hinchliffe, 1987). Thus, we next wanted to confirm that this rescue indeed was mediated by preventing Gli3 processing.

To do this, immediately before removing the posterior half of the wing, an SHH-soaked bead was implanted into the anterior mesenchyme at a mid-proximodistal position, close to the level of the transverse cut (see scheme in Fig. 2A). The effect of two doses of SHH (1 and 9 mg/ml) was tested. The sequential analysis of cell death performed after the experiment showed that the administration of exogenous SHH completely prevented cell death in the isolated anterior half-bud (Fig. 2B). The administration of SHH also rescued the phenotype seen after posterior bud removal. The rescue effect of a single bead was partial with truncations at the wrist level (Fig. 2C). With the low dose (1 mg/ml), the majority of specimens developed the radius, while the radius and ulna formed when the higher dose (9 mg/ml) was used. The sequential application of a second SHH-loaded bead 24 hr after the first, permitted the development of the autopod with formation of two or three digits. The higher dose always gave more complete limbs (Fig. 2C). The low Shh dose appears to stabilize the anterior mesoderm, permitting it to form anterior structures. The higher dose stabilizes the anterior mesoderm and also respecifies the mesoderm to form posterior structures.

thumbnail image

Figure 2. Exogenous application of Sonic hedgehog (SHH) permits survival and improved development of the anterior half limb in a dose-dependent manner. A: Scheme showing the operation and the site of placement of the bead. B: Terminal deoxynucleotidyl transferase-mediated deoxyuridinetriphosphate nick end-labeling (TUNEL) assay to show the absence of detectable cell death in the anterior mesoderm provided with exogenous SHH. The asterisk indicates the position of the bead. h, hours. C: Skeletal preparations of limbs developed from anterior wing halves that received a SHH bead as indicated at the top of each specimen. The application of two sequential beads and higher concentration dramatically improved the development of the anterior mesoderm. D: Western blot showing a threefold decrease in the level of Gli3R. E: With the low dose of Shh (1 mg/ml), the decrease in the level of Gli3R was only slightly diminished. F: In situ hybridization with Gli3 showing that the expression is not significantly modified in the intact limb by the application of the SHH bead. The position of the bead is indicated by the asterisk.

Download figure to PowerPoint

Next, we analyzed the state of the Gli3 protein in the Shh-rescued anterior halves (Fig. 2D). Western blots showed that the application of the higher doses of SHH effectively prevented the processing of Gli3 to the repressor form. After application of SHH (9 mg/ml), the level of Gli3R in the anterior mesoderm decreased by 2.3-fold compared with the anterior mesoderm in control conditions (Fig. 2D). However, the application of the low dose of 1mg/ml moderately reduced the level of Gli3R (1.4-fold) compared with the control anterior limb bud (Fig. 2E).

Because of the proposal that Shh controls Gli3 transcription, we also analyzed the expression of Gli3 after application of SHH both in the unaltered limb and after removal of the posterior half. Our results showed no detectable change in Gli3 expression after SHH applications with either concentration (Fig. 2F and not shown). Thus, neither removal of the Shh endogenous signaling nor application of exogenous SHH has detectable effects on Gli3 transcription in the conditions analyzed here. In summary, our results support a correlation between the level of Gli3R and cell death and prompted us to explore the mechanisms by which high levels of Gli3R could cause cell death.

Bmp4 Expression Is Rapidly Up-Regulated in the Anterior Mesoderm After Removal of the Posterior Mesoderm

We next explored possible changes in gene expression that could lead to increased apoptosis. First, we analyzed genes known to be expressed in the anterior mesoderm and related to Shh function.

Bmp4 was a good candidate to explore, because it has a normal domain of expression in the anterior mesoderm and, furthermore, was reported to be negatively regulated by Shh (Francis et al., 1994; Tümpel et al., 2002). We performed a sequential analysis of the expression pattern of Bmp4 after removal of the posterior mesoderm. Of interest, we found that Bmp4 expression appeared clearly up-regulated from 6 hr after the surgery (Fig. 3A). This up-regulation persisted and even increased during subsequent development of the anterior limb bud but was prevented by Shh (Fig. 3A).

thumbnail image

Figure 3. Bmp4 expression and signaling increases in the anterior mesoderm after removal of Sonic hedgehog (Shh) signaling. A: The expression of Bmp4 in the anterior mesoderm notably increased after removal of the posterior half, as sequentially shown from 3 to 36 hr (h) after removal of the posterior mesoderm. Bmp4 up-regulation was prevented by Shh. B–E: Expression of Msx2(B), Tbx3 (C), and Alx4 (D) is also up-regulated in the anterior mesoderm, while Bmp7 (E) expression appears down-regulated.

Download figure to PowerPoint

Bmp signaling has been implicated in the control of cell death in multiple developmental contexts and particularly in limb development (Ganan et al., 1996; Kawakami et al., 1996; Zou and Niswander, 1996; Macias et al., 1997; Merino et al., 1998). Thus, it is important to assess whether the up-regulation in Bmp4 expression is also accompanied by activation of its pathway. For this question, we analyzed the expression of two transcription factors, Msx2 and Tbx3, that had been shown to be downstream of Bmp signaling (Graham et al., 1994; Tümpel et al., 2002). Our results showed that Msx2 expression was up-regulated in parallel to Bmp4 (Fig. 3B). Tbx3 expression was also up-regulated but only at the time its expression was detected in the control anterior mesoderm (Fig. 3C). Tbx3 expression is very weak in the anterior mesoderm at the time the posterior removal is performed (Fig. 3C and Tümpel et al., 2002). This probably indicates that Bmp signaling is not the only signal required for Tbx3 expression. The up-regulation in Msx2 and Tbx3 expression confirms the up-regulation of Bmp4 signaling in the isolated anterior limb bud.

Another gene known to establish a negative regulation with Shh in the anterior mesoderm is Alx4 (Takahashi et al., 1998). We found that the transcription of Alx4 was rapidly up-regulated in the anterior limb bud after removal of the posterior half (Fig. 3D). Alx4 shows a different regulation than Gli3, whose transcription is not modified (Fig. 1E and compare with Fig. 3D).

The up-regulation in Bmp signaling suggested we also analyze the expression of other members of the Bmp family, Bmp2 and Bmp7, that are normally expressed in various parts of the limb bud. Bmp2 is not expressed by the anterior mesoderm during normal limb bud development and its expression remained undetectable after posterior mesoderm removal (not shown). Bmp7 has an anterior domain of expression similar to that of Bmp4 (Fig. 3E). Of interest, the anterior domain of Bmp7 expression was not maintained after the removal of the posterior mesoderm, and only a residual uniform expression was observed (Fig. 3E). We conclude that of these three BMPs, Bmp4 alone was up-regulated after posterior mesoderm removal.

Exogenous Noggin Dramatically Modifies the Pattern of Cell Death in the Isolated Anterior Mesoderm

To examine the implications of Bmp4 in the abnormal cell death in the anterior mesoderm, we assayed the ability of Noggin, a potent BMP antagonist, to prevent it. For this, a Noggin-soaked bead (0.5 mg/ml, R&D Systems) was implanted into the anterior mesoderm of stage 19–20 wing buds immediately before the removal of the posterior mesoderm (scheme in Fig. 4A). This strategy permitted us to examine sequentially the pattern of cell death in the anterior mesoderm under conditions of blocked Bmp signaling (Fig. 4B). The TUNEL assay showed a clear delay in the establishment of the abnormal anterior–distal cell death typical of this experiment, but concomitantly, abnormal cell death was observed around the bead (Fig. 4B). Eighteen hours after the operation and placement of the Noggin bead, cell death was observed around the bead (compare Fig. 4B + 18 hr with Fig. 1B + 18 hr). From 24 hr after the operation, cell death in the anterior–distal mesoderm become progressively intense (Fig. 4B). Accordingly, the limbs that developed were extremely truncated, showing one or two cartilage rods or nodules (Fig. 4C). This pattern of cell death can be interpreted as Noggin preventing the anterior–distal cell death caused by the posterior mesoderm removal, but at the same time, inducing cell death at the central level around the bead. The prevention of cell death by Noggin beads is transitory, only spanning the first 24 hr and could be explained by a transient delivery of Noggin from the bead (cf. Fallon et al., 1994). To check this possibility, we again used Msx2 expression as reporter of Bmp activity (Fig. 4D). We found that the Noggin bead completely blocked Msx2 expression 6 hr after implantation, indicating a complete block of Bmp signaling (Fig. 4D). However, its effect, as measured by Mxs2 expression, progressively declined, with Msx2 appearing by 12 hr (arrowhead in Fig. 4D) and becoming normally expressed by 18 hr (Fig. 4D). Thus, we conclude that the Noggin-bead effect was transitory, and, to permanently block BMP signaling, we sequentially applied Noggin beads at 12-hr intervals (Fig. 4E). By doing this, anterior–distal cell death was prevented; however, it is notable, that the phenotype did not improve (Fig. 4E). The complex interactions between Noggin and the apical ectodermal ridge (Pizette et al., 2001) together with the late noggin effects on chondrogenesis, that require BMP signaling (Brunet et al., 1998) may possibly explain this truncation.

thumbnail image

Figure 4. Overexpression of Noggin prevents cell death in the isolated anterior mesoderm. A: Scheme showing the operation and the site of bead placement. B: Terminal deoxynucleotidyl transferase-mediated deoxyuridinetriphosphate nick end-labeling (TUNEL) assay showing delay in the onset of the abnormal cell death typical of the experiment. However, cell death was concomitantly observed around the applied Noggin bead. h, hours. C: Skeletal pattern in limbs that developed from isolated anterior mesoderm that received exogenous Noggin. The phenotype was not improved. D: Noggin effect from the bead is transitory. The Noggin bead completely abolished Msx2 expression 6 hr after application. A minimal down-regulation was observed after 12 hr (arrow) and Msx2 expression was normal after 18 hr. E: The sequential application of Noggin beads with 12-hr intervals prevented the anterior–distal cell death typical of the experiment but did not rescue the phenotype. F: In intact wing buds, exogenous Noggin also induced cell death in association with the bead and resulted in a short humerus. G,H: Overexpression of Noggin by replication-competent retroviral vector (RCAS) prevented cell death in the isolated anterior mesoderm, permitting the formation of a slender outgrowth.

Download figure to PowerPoint

While it is well known that exogenous application of Bmp causes cell death, to our knowledge, this effect has not been reported previously for Noggin. However, in our experiments, we always observed cell death around the Noggin bead. Because this observation was performed in experimental conditions, we also analyzed whether Noggin would cause cell death when applied into the otherwise undisturbed wing bud. We found that the application of Noggin-loaded beads into the intact wing bud was always accompanied by cell death as shown 16 hr after application (Fig. 4F) and eventually resulted in abnormal development of the humerus that was misshapen and very short (arrow in Fig. 4F).

To overcome the problems of Noggin application in beads such as its transitory effect, and to further check the idea that blocking Bmp signaling could prevent cell death in our system, we decided to overexpress Noggin in the limb mesoderm using a replication-competent retroviral vector (RCAS). By using this method, the level of Noggin expression was expected to be at a steady level during the experiment and similar to endogenous messages. RCAS–Noggin was injected into the presumptive limb field at stage 10–12, and when the embryo reached stage 20, the posterior mesoderm was removed. The anterior limb bud overexpressing Noggin formed a small elongated outgrowth very different in shape from the spike formed in the plain posterior removal experiments (Fig. 4G). Anterior buds with confirmed overexpression of Noggin, as assessed by in situ hybridization to tissue sections, always showed a complete absence of cell death in the TUNEL assay (Fig. 4H). Taken together our results strongly indicate that the cell death that occurs in the isolated anterior mesoderm is mediated by Bmp4 signaling.

Interaction Between Gli3 and Bmp4

The Noggin experiments indicate that Bmp signaling is responsible for the abnormal cell death experienced by the anterior mesoderm in the absence of Shh signaling. Thus, it is possible that Gli3R controls cell death through Bmp4. Indeed, there is evidence of a genetic interaction between Gli3 and Bmp4 (Dunn et al., 1997), the phenotype of the double Gli3;Bmp4 heterozygote being much more severe than either heterozygote alone. However, to date, it has not been clarified whether these two factors act in the same or different pathways.

We decided to analyze Bmp4 expression in different conditions of Gli3 expression and processing. For this analysis, we took advantage of several mouse and chick mutants that alter either Gli3 expression or processing. We selected the Extra toes (XtJ) mouse mutant, which lacks Gli3 expression; the Shh-null mice, where Gli3R predominates (Litingtung et al., 2002); the ozd chick mutant, equivalent to the Shh-null mice at limb level (Ros et al., 2003); and the talpid2 (ta2) chick mutant, where the full-length Gli3-190 predominates (Wang et al., 2000; Litingtung et al., 2002).

If Bmp4 expression were somehow dependent on Gli3, we would expect it to be down-regulated in the XtJ mutant limb. As limb buds start development, Bmp4 expression is seen uniformly across the mutant limb bud, as in wild-type (not shown). At E10.5, Bmp4 expression had evolved into two mesodermal domains, the stronger one at the posterior border, overlapping the domain of Shh expression, and another anterior–distal domain of lesser intensity (Fig. 5A). A careful analysis showed that, in the XtJ mutant forelimb, the anterior–distal domain of Bmp4 expression appeared clearly down-regulated as compared with wild-type (100%, six genotyped mutant limbs analyzed, Fig. 5). This reduction was more intense in the forelimb than in the hindlimb (Fig. 5C,D). The posterior domain of Bmp4 expression remained unaltered in the XtJ mutant limb.

thumbnail image

Figure 5. Expression of Bmp4 in several mutations that affect Gli3 expression and processing. A,B:Bmp4 expression was seen in the anterior and posterior mesoderm and in the apical ectodermal ridge (AER) of the 10.5 dpc forelimb (FL) and hindlimb (HL). C,D: Expression of Bmp4 was reduced in the anterior–distal mesoderm of the Xt/Xt mutant limbs, while the expression in the posterior mesoderm and AER was maintained. E,F: Bmp4 expression in the Shh-null mice occurred uniformly along the distal mesoderm. G: Pattern of Bmp4 expression in a stage 23 normal wing (W). H: Bmp4 expression was clearly up-regulated in the wing bud of the ozd mutant that lack Shh signaling specifically at limb level. I: Talpid2 (ta2) wing bud of a corresponding stage 25 showing down-regulation of Bmp4 expression. Only residual expression was observed in the anterior and posterior borders. WT, wild-type.

Download figure to PowerPoint

In the Shh-null limb, processing of Gli3 is not prevented, resulting in increased repressor form (Litingtung et al., 2002). In these conditions, Bmp4 expression begins normally (not shown) but, at day 10.5, appears clearly up-regulated in the distal mesoderm (Fig. 5E,F). This up-regulation is more intense in the hindlimb than in the forelimb (100%, six mutant limbs analyzed, Fig. 5E,F) and appears as a continuous band that spans from the posterior border into the distal mesoderm. In the Shh mutant limb, particularly in the forelimb, the expression of Bmp4 in the junction of the limb to the body wall appears also strongly increased (arrow in Fig. 5E).

ozd mutant limbs lack Shh activity at limb levels and similar to the mouse Shh-null mice we show increased Bmp4 expression in the distal mesoderm (Fig. 5H, two confirmed mutants analyzed). In direct contrast to the Shh-null mouse, the defect in ta2 is that the Shh pathway is constitutively activated all along the anterior–posterior axis of the limb (Caruccio et al., 1999), and the Gli3-190 is the predominant form detected (Wang et al., 2000; Litingtung et al., 2002). We report that Bmp4 expression was clearly reduced in the distal mesoderm of ta2, although some residual expression is observed at the anterior and posterior border (Fig. 5I).

In summary, the analysis of these mutants strengthens the correlation between the presence of Gli3R and Bmp4 expression in the distal mesoderm. While Gli3 is not required for activation of Bmp4 expression, appropriate proportions of Gli3-190 and Gli3R appear to be required for the normal pattern of expression.

Finally, it is interesting to note here the difference between the mouse and chick pattern of Bmp4 expression. Whereas in the mouse, the stronger expression is seen in the posterior domain, in chick, the expression is stronger in the anterior domain (compare Fig. 5A with Fig. 5G).

It has also been reported that Bmp4 induces Gli3 expression in explants of the neural tube (Meyer and Roelink, 2003). To analyze this possibility in the limb bud, we performed experiments of gain and loss of Bmp signaling function and analyzed subsequent modifications in Gli3 transcription and protein processing. For the loss of function, we used an exogenous application of Noggin that did not appreciably change Gli3 transcription as shown for 6 and 12 hr (Fig. 6A) and was not modified at 24 hr (not shown). An area of decreased or absent Gli3 expression was consistently observed around the applied Noggin bead. Because this area duplicated the zone of intense cell death induced by the bead (compare the Gli3 hybridization with the TUNEL assay in Fig. 6A), we propose that it was due to nonspecific loss of gene expression accompanying cell death. As expected, the application of Noggin had no effect on the processing of Gli3 protein as evidenced by the Western blot performed with extracts from anterior wing buds 6 hr and 17 hr after operation (Fig. 6C and not shown). The amount of Gli3R form was comparable in the anterior half, regardless of the presence of Noggin (Fig. 6C).

thumbnail image

Figure 6. Loss and gain of function of Bmp signaling does not significantly modify Gli3 expression or processing. A: The Noggin bead does not appreciably change Gli3 expression, except the reduction in the area of induced cell death. h, hour; TUNEL, terminal deoxynucleotidyl transferase-mediated deoxyuridinetriphosphate nick end-labeling. B: Scheme of the procedure. C: The Western blot analysis shows that processing of Gli3 was not modified by the blocking of Bmp signaling by Noggin. D: The Bmp4 bead slightly down-regulated Gli3 expression in the distal mesoderm. This down-regulation was not attributable to increased apoptosis as shown in the TUNEL assay. E: Scheme of the procedure. F: Western blot showing a very slightly increase in the processing of Gli3 in presence of Bmp4, both at 6 and 18 hr after application.

Download figure to PowerPoint

For the gain of function experiments, we applied BMP4 beads into the wing bud and subsequently analyzed Gli3 transcription and protein processing. To avoid the BMP4 apoptotic effect, we used a very low dose of Bmp4 (0.01 mg/ml). BMP4 caused a faint down-regulation in Gli3 expression in the distal limb mesoderm (Fig. 6D). This finding was consistently accompanied by a distal indentation in the contour of the limb bud (arrow in Fig. 6D) presumably due to the BMP detrimental effect on the apical ectodermal ridge (AER; cf. Pizette and Niswander, 1999). These two effects cannot be due to cell death, because a parallel examination by TUNEL showed no detectable apoptosis in the distal mesoderm and only some scattered dead cells around the bead (Fig. 6D). The Western blot performed demonstrated that, in the presence of BMP, the processing of Gli3R was not appreciably altered (Fig. 6F).

State of the AER in the Isolated Anterior Mesoderm

The AER is the specialized epithelium rimming the distal tip of the developing limb that directs proximodistal growth through the production of several members of the family of fibroblast growth factors (FGF; reviewed in Martin, 1998). There is a positive feedback loop between the AER/FGFs and the ZPA/Shh (Riddle et al., 1993; Niswander et al., 1994). Shh from the ZPA maintains the AER by the induction of Gremlin (Gre), an antagonist of BMP, that prevents the negative effect of BMPs on the AER (Pizette and Niswander, 1999; Zuniga et al., 1999). Thus, in our experiments, it is of interest to analyze the state of the AER. Our analysis detected down-regulation of Fgf8 expression in the anterior wing bud as soon as 3 hr after removal of the posterior half (Fig. 7A). This finding was accompanied by a considerable flattening of the AER that was clearly observed in sections (Fig. 7B,C). These molecular and morphological alterations persisted at 24 hr (Fig. 7D–F). Curiously, a point of stronger Fgf8 expression was observed at the posterior limit of the AER, reminiscent of what is observed in the Shh and ozd mutant limbs (Chiang et al., 2001; Kraus et al., 2001; Ros et al., 2003). Thus, we interpreted that the alterations in the AER are consistent with what is expected after removal of Shh signaling.

thumbnail image

Figure 7. State of the apical ectodermal ridge (AER) in the isolated anterior mesoderm. A–C: Three hours (h) after removal of the posterior mesoderm, Fgf8 expression is maintained in the AER of the operated wing bud, but the AER appears flattened and narrow. D–F: The same morphological alteration was seen 24 hr after the operation.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  7. Acknowledgements

In this report, we explore the molecular events that accompany apoptosis of anterior wing bud mesoderm after surgical removal of the posterior wing bud, including the ZPA. The data demonstrate that posterior wing bud removal results in increased conversion of Gli3 to its repressor form; this conversion is prevented by exogenous Shh protein. With increased Gli3R, Bmp4 is up-regulated and massive apoptosis ensues; this process is prevented by misexpression of exogenous Noggin. These data establish a sequence of events leading to cell death in anterior mesoderm as a result of loss of Shh signaling from the posterior mesoderm, with Bmp4 as the effector molecule. This finding indicates a necessary relationship for anterior mesoderm survival on posterior mesoderm (Shh) signaling.

Level of Gli3R Form Increases in the Anterior Mesoderm Deprived of Shh Signaling

Shh signaling inhibits processing of Gli3 into Gli3R, a C-terminal truncated repressor (Wang et al., 2000; Litingtung et al., 2002). As a consequence, normal limb development occurs in a decreasing anterior-to-posterior gradient of Gli3R (Wang et al., 2000; Litingtung et al., 2002). Increased levels of Gli3R, such as produced in the absence of Shh signaling, are incompatible with normal development leading to altered gene expression and excessive cell death (Chiang et al., 2001; te Welscher et al., 2002). A positive correlation between the level of Gli3R and the amount of cell death is observed both in the neural tube and in the limb (Litingtung and Chiang, 2000; te Welscher et al., 2002). This correlation is further supported by the study of the Gli3 null mice, in which normal areas of cell death appear very diminished (Aoto et al., 2002). The Gli3-dependent control of apoptosis has been postulated to serve to limit areas of Fgf8 expression that are expanded in the Gli3-/- mutant (Aoto et al., 2002).

Based on these findings, our hypothesis was that the massive apoptosis that occurs in the anterior mesoderm subsequent to removal of Shh signaling was due to increased levels of Gli3R. Consistent with this idea, we show here by Western blot, a twofold increase in the amount of Gli3R in the anterior mesoderm in the absence of Shh. This increment was detected 6 hr after the surgery and persisted at 18 hr when the apoptosis was maximal. We have detected excessive and patterned cell death by 12 hr after the removal of the posterior mesoderm. This finding is some hours earlier than previously reported (Todt and Fallon, 1987; Wilson and Hinchliffe, 1987) and probably due to the better sensitivity of the TUNEL essay used here vs. the vital staining used in the previous studies. Therefore, our results add further support to the correlation between excess of Gli3R and apoptosis.

The increase in the processing of Gli3 observed after removal of the posterior mesoderm confirms that Shh itself migrates, at least to the middle of the bud (Gritli-Linde et al., 2001; Lewis et al., 2001; Zeng et al., 2001). Although processing of Gli3 is cell autonomous, we cannot precisely resolve how far anteriorly this effect spreads, because our analysis was performed in extracts from the entire anterior half of the limb. It is also worth noting that, because the level of Gli3 RNA remains largely unchanged after loss of Shh signaling and that the full-length Gli3 is always only a minor fraction of the total Gli3 present, the loss of Shh signaling may also increase the stability of the Gli3 repressor protein.

Bmp4 Expression Is Up-Regulated in the Anterior Mesoderm Deprived of Shh Signaling

Our analysis also shows up-regulation in the expression of Bmp4 and other anteriorly expressed genes after removal of the posterior mesoderm. In particular, BMP signaling has been shown to mediate apoptosis in different contexts and tissues, including the limb bud (Graham et al., 1994; Kawakami et al., 1996; Yokouchi et al., 1996; Zou and Niswander, 1996; Macias et al., 1997; Guha et al., 2002), and consequently, its possible involvement in the increased apoptosis had to be considered. The up-regulation of Msx2 and Tbx3 expression, two transcription factors known to be responsive to Bmp, demonstrates the activation of the Bmp pathway. Furthermore, forced expression of Noggin in the anterior mesoderm, either by the exogenous application of the protein or by misexpression with the RCAS retroviral vector, prevented cell death, demonstrating that Bmp signaling, in fact, was responsible for it. The next question that arises is whether these two concomitant events, increase in Gli3R and increase in Bmp4 expression, are related events.

A striking unexpected result of our experiments was the observation that, while Noggin prevented cell death in the anterior and distal mesoderm, it induced cell death in the deeper mesoderm around the bead. To our knowledge, an apoptotic effect of the Noggin bead has only been reported when applied at the distal digital tip (Merino et al., 1998). However, the time-course analysis of these experiments indicates that cell death is an indirect consequence of the antichondrogenic effect of Noggin. In early limb buds, Noggin applications have been performed mostly at the anterior border (cf. Tümpel et al., 2002), and this method may be the reason why its apoptotic effect in the central mesoderm has not been identified previously. It is tempting to suggest that the precartilaginous cells in the core of the limb bud are sensitive to the antichondrogenic effect of Noggin; however, this question would require further analysis.

Interaction Between Gli3 and Bmp4

In vertebrates, the genetic interaction between Gli3 and Bmp4 has been pointed out previously (Dunn et al., 1997; Aoto et al., 2002). A clear indication of the interaction between the two genes is the observation that the double heterozygous for Gli3 and Bmp4 displays a polydactyly that is much more severe than that of each heterozygote alone (Dunn et al., 1997).

As a way of analyzing the interaction between Gli3 and Bmp4, we have studied the expression of Bmp4 in conditions of modified expression or processing of Gli3. The results of our study in selected mouse and chick mutants indicate that Bmp4 is positively regulated by the repressor form of Gli3. Of interest, we have identified a moderate but consistent down-regulation of expression in the distal mesoderm of the Gli3 mutant limb that was previously unidentified (Buscher et al., 1997; Dunn et al., 1997).

The question then was how the increase in Gli3R could ultimately result in an increase in Bmp4 expression. The Gli3R form is thought to act as a potent repressor of transcription and has not been demonstrated so far to act as an activator. Thus Gli3R may work by means of repressing a repressor of Bmp4 expression in the distal mesoderm. It is interesting to note that the existence of a repressor of Tbx3 expression in the middle of the limb already has been suggested (Tümpel et al., 2002). The same putative repressor could operate for Bmp4 that controls Tbx3 expression in the anterior mesoderm. Alternatively, it may be possible that, under certain circumstances or interactions with other transcription factors, the Gli3R form can directly activate Bmp4 transcription. Gli3R is likely one of multiple cofactors needed for transcription of Bmp4, and it is also possible that there are different mechanisms of transcription involved in the expression of Bmp proteins.

We have also considered the possibility that Bmp4 signaling could modify Gli3 transcription or processing, particularly because it has been shown that Bmp4 induces Gli3 transcription in explants of the neural tube (Meyer and Roelink, 2003). However, in the limb bud, our experiments of gain and loss of Bmp4 function have failed to detect any obvious regulation of Gli3 expression by Bmp4.

Shh and Cell Death in the Limb Bud

Experiments performed in the chick wing bud established that signaling from the posterior mesoderm, later identified as Shh signaling, was required for survival of the anterior mesoderm (Todt and Fallon, 1987; Wilson and Hinchliffe, 1987). More recently, it was also shown that Shh selectively affects cell death and survival of the limb mesoderm in a position-dependent way (Sanz-Ezquerro and Tickle, 2000). Ectopic Shh promotes survival of the anterior and middle mesoderm, while it promotes cell death of the posterior mesoderm (Sanz-Ezquerro and Tickle, 2000). However, the mechanism of how Shh exerts its effect on cell death remains poorly understood.

Recently, it has been suggested that Ptc1, the receptor of Shh, is a proapoptotic dependence factor (Thibert et al., 2003). Similar to the limb bud, the loss of Shh signaling results in massive apoptosis in the developing neural tube, and the apoptosis decreases in parallel with the removal of the Gli3 alleles (Litingtung and Chiang, 2000), indicating that Gli3 is implicated. Now it has been shown that this cell death can also be rescued by the forced expression of a dominant negative form of Ptc-1. It remains to be explored whether Ptc-1 plays a similar role in the control of apoptosis in the limb bud.

Based on the results presented here, we propose that the mechanism by which Shh could control cell death in the limb bud is by controlling, through Gli activity, the level of Bmp signaling. The idea that Gli proteins, as mediators of Shh activity, play a key role in regulating Bmp expression during development is further supported by the observation that BMP4 and BMP7 promoters respond to Gli1 and Gli3 proteins, in certain cell types (Kawai and Sugiura, 2001). Several observations are also consistent with this proposal. For example, the brain of the Shh-/- mutant exhibits abnormal cell death assumed to be secondary to increased Bmp4 signaling (Ohkubo et al., 2002; Aoto et al., 2002). The same occurs in the Shh-/- and ozd mutant limbs where cell death and Bmp4 expression are both exaggerated (Chiang et al., 2001; Ros et al., 2003; this report). Conversely, in the Gli3 mutant, the decrease in cell death correlates well with a reduction of Bmp expression detectable by in situ hybridization in the brain (Tole et al., 2000) and in the limb (this report). Since this decrease in the limb had not been identified previously (Buscher et al., 1997), the reduction of cell death was attributed to the up-regulation of Gremlin, a potent BMP antagonist, that occurs in the Gli3-null limb (Merino et al., 1999; Aoto et al., 2002; te Welscher et al., 2002). Thus, it is likely that both up-regulation of Gremlin and down-regulation of Bmp4 expression contribute to decreased cell death.

During limb development, Shh's control of cell death by the Gli proteins is complex. Part of this complexity is due to the implication of several Bmps that could respond differentially to different combinations of Gli proteins. In our experiments Bmp4, Bmp2, and Bmp7 respond differently, Bmp4 is uniquely up-regulated in the anterior mesoderm when the level of Gli3R increases. In Drosophila, several studies have shown that Ci, orthologue of Gli3, regulates the expression of dpp, homologue of Bmp4 (Alexandre et al., 1996; Dominguez et al., 1996). Similarly, in the vertebrate limb bud, Bmp2 expression strongly depends on Shh activity (Drossopoulou et al., 2000; and our personal observations). Thus, in the posterior mesoderm, Shh could positively regulate cell death through Bmp2, likely mediated by Gli1 and Gli2 activity. In this context, it is interesting to mention that the double Gli1;Gli2-null limb presents a small posterior outgrowth that could reflect decreased cell death (Park et al., 2000). Conversely, Shh may prevent cell death in the anterior and middle mesoderm of the limb through the negative control of Bmp4 expression, mediated by Gli3R. Viewed in this way, Shh not only functions obligatorily through Gli3 to pattern the limb but also functions through Gli3 to maintain homeostasis in the limb bud mesoderm, balancing proliferation and cell death as pattern emerges.


  1. Top of page
  2. Abstract
  7. Acknowledgements

Chick and Mouse Mutants and Genotyping

Normal hen eggs were obtained from local sources and incubated, opened, and staged as described (Hamburger and Hamilton, 1951; Ros et al., 2000). Mouse control embryos were obtained from CRIFFA (Barcelona, Spain).

Gli3Xt/Xt mice (Jackson allele) and Shh-/- embryos were maintained on a C57BL6 background, and adult mice were genotyped according to (Hui and Joyner, 1993). Chick mutant talpid2 and oligozeugodactyly embryos were obtained from the University of Wisconsin – Madison heterozygous flocks.

Experimental Manipulations

The posterior or anterior half of the wing bud from embryos ranking from stage 19 to 21 was removed in ovo by using a tungsten needle with the tip bent in a right angle. A transverse cut through the middle of the bud was followed by a second cut along the junction of the anterior (or posterior) bud with the body wall. Special care was taken not to tear off the anterior AER when performing the first cut. Removal of the AER was performed as described in (Ros et al., 2000). To analyze the skeletal pattern, the cartilage routine Victoria Blue staining was performed in limbs of 10–11 days of total development.

Applications of Proteins Using Beads as Carriers

Heparin acrylic beads (Sigma, H5263) were soaked for at least 1 hr at room temperature in recombinant mouse SHH (1 and 9 mg/ml), Noggin (0.5 mg/ml from R&D Systems), or recombinant human BMP4 (0.01 mg/ml; a generous gift from Genetics Institute) solutions. The beads were implanted where desired into the undisturbed wing. When required, the posterior half of the limb bud was removed immediately after the placement of the bead.

RCAS–Noggin Infections

The RCAS–Noggin construct was kindly provided by L. Niswander. Transfection, growth, and concentration of viruses were performed as described by Morgan and Fekete (1996). SPAFAS eggs were used for infections.

In Situ Hybridization in Whole-Mounts

Digoxigenin-labeled antisense riboprobes were prepared, and whole-mount in situ hybridization analysis was performed according to standard procedures (Nieto et al., 1996). The probes used were Shh, Fgf8, Bmp4, Gli3, Tbx3, Msx2, Noggin, and Alx4 (kindly provided by C. Tabin, A. Kuroiwa, G. Eichele, and L. Niswander).

Cell Death Analysis

In situ detection of DNA fragmentation was performed using TUNEL with the In Situ Cell Death Detection Kit, Apoptag.

Western Blot

Immunoblot (Western blot) analysis was performed as described (Wang et al., 1996). Dissected limb buds of treated embryos were lysed with ice-cold RIPA buffer, and 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis was used to resolve Gli3-190 from Gli3-83 protein, and the affinity-purified Gli3ab1 antibody was used (Wang et al., 2000). α tubulin was assessed as control for normalization.


  1. Top of page
  2. Abstract
  7. Acknowledgements

We thank Genetics Institute for providing the rhBMP4 protein and Dr. Rolf Zeller for making the Gli3Xt/Xt and Shh-/- mice available. We also thank C. Tabin, A. Kuroiwa, G. Eichele, and L. Niswander for probes and reagents. J.F.F. was funded by the National Institutes of Health. This work was supported by grant FP12002-02946 to M.A.R.


  1. Top of page
  2. Abstract
  7. Acknowledgements
  • Alexandre C, Jacinto A, Ingham PW. 1996. Transcriptional activation of hedgehog target genes in Drosophila is mediated directly by the cubitus interruptus protein, a member of the GLI family of zinc finger DNA-binding proteins. Genes Dev 10: 20032013.
  • Aoto K, Nishimura T, Eto K, Motoyama J. 2002. Mouse GLI3 regulates Fgf8 expression and apoptosis in the developing neural tube, face, and limb bud. Dev Biol 251: 320332.
  • Aza-Blanc P, Ramirez-Weber FA, Laget MP, Schwartz C, Kornberg TB. 1997. Proteolysis that is inhibited by hedgehog targets Cubitus interruptus protein to the nucleus and converts it to a repressor. Cell 89: 10431053.
  • Aza-Blanc P, Lin HY, Ruiz i Altaba A, Kornberg TB. 2000. Expression of the vertebrate Gli proteins in Drosophila reveals a distribution of activator and repressor activities. Development 127: 42934301.
  • Bowen J, Hinchliffe JR, Horder TJ, Reeve AM. 1989. The fate map of the chick forelimb-bud and its bearing on hypothesized developmental control mechanisms. Anat Embryol (Berl) 179: 269283.
  • Brunet LJ, McMahon JA, McMahon AP, Harland RM. 1998. Noggin, cartilage morphogenesis, and joint formation in the mammalian skeleton. Science 280: 14551457.
  • Buscher D, Bosse B, Heymer J, Ruther U. 1997. Evidence for genetic control of Sonic hedgehog by Gli3 in mouse limb development. Mech Dev 62: 175182.
  • Caruccio NC, Martinez-Lopez A, Harris M, Dvorak L, Bitgood J, Simandl BK, Fallon JF. 1999. Constitutive activation of sonic hedgehog signaling in the chicken mutant talpid(2): Shh-independent outgrowth and polarizing activity. Dev Biol 212: 137149.
  • Chiang C, Litingtung Y, Harris MP, Simandl BK, Li Y, Beachy PA, Fallon JF. 2001. Manifestation of the limb prepattern: limb development in the absence of sonic hedgehog function. Dev Biol 236: 421435.
  • Dai P, Akimaru H, Tanaka Y, Maekawa T, Nakafuku M, Ishii S. 1999. Sonic Hedgehog-induced activation of the Gli1 promoter is mediated by GLI3. J Biol Chem 274: 81438152.
  • Dominguez M, Brunner M, Hafen E, Basler K. 1996. Sending and receiving the hedgehog signal: control by the Drosophila Gli protein Cubitus interruptus. Science 272: 16211625.
  • Drossopoulou G, Lewis KE, Sanz-Ezquerro JJ, Nikbakht N, McMahon AP, Hofmann C, Tickle C. 2000. A model for anteroposterior patterning of the vertebrate limb based on sequential long- and short-range Shh signalling and Bmp signalling. Development 127: 13371348.
  • Dunn NR, Winnier GE, Hargett LK, Schrick JJ, Fogo AB, Hogan BL. 1997. Haploinsufficient phenotypes in Bmp4 heterozygous null mice and modification by mutations in Gli3 and Alx4. Dev Biol 188: 235247.
  • Duprez DM, Kostakopoulou K, Francis-West PH, Tickle C, Brickell PM. 1996. Activation of Fgf-4 and HoxD gene expression by BMP-2 expressing cells in the developing chick limb. Development 122: 18211828.
  • Fallon JF, Lopez A, Ros MA, Savage MP, Olwin BB, Simandl BK. 1994. FGF-2: apical ectodermal ridge growth signal for chick limb development. Science 264; 104107.
  • Francis PH, Richardson MK, Brickell PM, Tickle C. 1994. Bone morphogenetic proteins and a signalling pathway that controls patterning in the developing chick limb. Development 120: 209218.
  • Ganan Y, Macias D, Duterque-Coquillaud M, Ros MA, Hurle JM. 1996. Role of TGF beta s and BMPs as signals controlling the position of the digits and the areas of interdigital cell death in the developing chick limb autopod. Development 122: 23492357.
  • Graham A, Francis-West P, Brickell P, Lumsden A. 1994. The signalling molecule BMP4 mediates apoptosis in the rhombencephalic neural crest. Nature 372: 684686.
  • Gritli-Linde A, Lewis P, McMahon AP, Linde A. 2001. The whereabouts of a morphogen: direct evidence for short- and graded long-range activity of hedgehog signaling peptides. Dev Biol 236: 364386.
  • Guha U, Gomes WA, Kobayashi T, Pestell RG, Kessler JA. 2002. In vivo evidence that BMP signaling is necessary for apoptosis in the mouse limb. Dev Biol 249: 108120.
  • Hamburger V, Hamilton HL. 1951. A series of normal stages in the development of the chick embryo. J Morphol 88: 4992.
  • Hinchliffe JR, Gumpel-Pinot M. 1981. Control of maintenance and anteroposterior skeletal differentiation of the anterior mesenchyme of the chick wing bud by its posterior margin (the ZPA). J Embryol Exp Morphol 62: 6382.
  • Hui CC, Joyner AL. 1993. A mouse model of greig cephalopolysyndactyly syndrome: the extra-toesJ mutation contains an intragenic deletion of the Gli3 gene. Nat Genet 3: 241246.
  • Hui CC, Slusarski D, Platt KA, Holmgren R, Joyner AL. 1994. Expression of three mouse homologs of the Drosophila segment polarity gene cubitus interruptus, Gli, Gli-2, and Gli-3, in ectoderm- and mesoderm-derived tissues suggests multiple roles during postimplantation development. Dev Biol 162: 402413.
  • Ingham PW, McMahon AP. 2001. Hedgehog signaling in animal development: paradigms and principles. Genes Dev 15: 30593087.
  • Kawai S, Sugiura T. 2001. Characterization of human bone morphogenetic protein (BMP)-4 and -7 gene promoters: activation of BMP promoters by Gli, a sonic hedgehog mediator. Bone 29: 5461.
  • Kawakami Y, Ishikawa T, Shimabara M, Tanda N, Enomoto-Iwamoto M, Iwamoto M, Kuwana T, Ueki A, Noji S, Nohno T. 1996. BMP signaling during bone pattern determination in the developing limb. Development 122: 35573566.
  • Kinzler KW, Ruppert JM, Bigner SH, Vogelstein B. 1988. The GLI gene is a member of the Kruppel family of zinc finger proteins. Nature 332: 371374.
  • Kraus P, Fraidenraich D, Loomis CA. 2001. Some distal limb structures develop in mice lacking Sonic hedgehog signaling. Mech Dev 100: 4558.
  • Lewis PM, Dunn MP, McMahon JA, Logan M, Martin JF, St-Jacques B, McMahon AP. 2001. Cholesterol modification of sonic hedgehog is required for long-range signaling activity and effective modulation of signaling by Ptc1. Cell 105: 599612.
  • Litingtung Y, Chiang C. 2000. Specification of ventral neuron types is mediated by an antagonistic interaction between Shh and Gli3. Nat Neurosci 3: 979985.
  • Litingtung Y, Dahn RD, Li Y, Fallon JF, Chiang C. 2002. Shh and Gli3 are dispensable for limb skeleton formation but regulate digit number and identity. Nature 418: 979983.
  • Macias D, Ganan Y, Sampath TK, Piedra ME, Ros MA, Hurle JM. 1997. Role of BMP-2 and OP-1 (BMP-7) in programmed cell death and skeletogenesis during chick limb development. Development 124: 11091117.
  • Marigo V, Johnson RL, Vortkamp A, Tabin CJ. 1996. Sonic hedgehog differentially regulates expression of GLI and GLI3 during limb development. Dev Biol 180: 273283.
  • Martin GR. 1998. The roles of FGFs in the early development of vertebrate limbs. Genes Dev 12: 15711586.
  • McMahon AP. 2000. More surprises in the Hedgehog signaling pathway. Cell 100: 185188.
  • Merino R, Ganan Y, Macias D, Economides AN, Sampath KT, Hurle JM. 1998. Morphogenesis of digits in the avian limb is controlled by FGFs, TGFbetas, and noggin through BMP signaling. Dev Biol 200: 3545.
  • Merino R, Rodriguez-Leon J, Macias D, Ganan Y, Economides AN, Hurle JM. 1999. The BMP antagonist Gremlin regulates outgrowth, chondrogenesis and programmed cell death in the developing limb. Development 126: 55155522.
  • Methot N, Basler K. 2001. An absolute requirement for Cubitus interruptus in Hedgehog signaling. Development 128: 733742.
  • Meyer NP, Roelink H. 2003. The amino-terminal region of Gli3 antagonizes the Shh response and acts in dorsoventral fate specification in the developing spinal cord. Dev Biol 257: 343355.
  • Mo R, Freer AM, Zinyk DL, Crackower MA, Michaud J, Heng HH, Chik KW, Shi XM, Tsui LC, Cheng SH, Joyner AL, Hui C. 1997. Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development 124: 113123.
  • Morgan BA, Fekete DM. 1996. Manipulating gene expression with replication-competent retroviruses. Methods Cell Biol 51: 185218.
  • Muller B, Basler K. 2000. The repressor and activator forms of Cubitus interruptus control Hedgehog target genes through common generic gli-binding sites. Development 127: 29993007.
  • Nieto MA, Patel K, Wilkinson DG. 1996. In situ analysis of chick embryos in whole mount and tissue sections. In: Bronner-FraserM, editor. Methods in cell biology. New York: Academic Press. p 219235.
  • Niswander L, Jeffrey S, Martin GR, Tickle C. 1994. A positive feedback loop coordinates growth and patterning in the vertebrate limb. Nature 371: 609612.
  • Ohkubo Y, Chiang C, Rubenstein JL. 2002. Coordinate regulation and synergistic actions of BMP4, SHH and FGF8 in the rostral prosencephalon regulate morphogenesis of the telencephalic and optic vesicles. Neuroscience 111: 117.
  • Park HL, Bai C, Platt KA, Matise MP, Beeghly A, Hui CC, Nakashima M, Joyner AL. 2000. Mouse Gli1 mutants are viable but have defects in SHH signaling in combination with a Gli2 mutation. Development 127: 15931605.
  • Pizette S, Niswander L. 1999. BMPs negatively regulate structure and function of the limb apical ectodermal ridge. Development 126: 883894.
  • Pizette S, Abate-Shen C, Niswander L. 2001. BMP controls proximodistal outgrowth, via induction of the apical ectodermal ridge, and dorsoventral patterning in the vertebrate limb. Development 128: 44634474.
  • Riddle RD, Johnson RL, Laufer E, Tabin C. 1993. Sonic hedgehog mediates the polarizing activity of the ZPA. Cell 75: 14011416.
  • Ros MA, Simandl BK, Clark AW, Fallon JF. 2000. Methods for manipulating the chick limb bud to study gene expression, tissue interactions, and patterning. Methods Mol Biol 137: 245266.
  • Ros MA, Dahn RD, Fernandez-Teran M, Rashka K, Caruccio NC, Hasso SM, Bitgood JJ, Lancman JJ, Fallon JF. 2003. The chick oligozeugodactyly (ozd) mutant lacks sonic hedgehog function in the limb. Development 130: 527537.
  • Ruppert JM, Kinzler KW, Wong AJ, Bigner SH, Kao FT, Law ML, Seuanez HN, O'Brien SJ, Vogelstein B. 1988. The GLI-Kruppel family of human genes. Mol Cell Biol 8: 31043113.
  • Sanz-Ezquerro JJ, Tickle C. 2000. Autoregulation of Shh expression and Shh induction of cell death suggest a mechanism for modulating polarising activity during chick limb development. Development 127: 48114823.
  • Schimmang T, Lemaistre M, Vortkamp A, Ruther U. 1992. Expression of the zinc finger gene Gli3 is affected in the morphogenetic mouse mutant extra-toes (Xt). Development 116: 799804.
  • Schweitzer R, Vogan KJ, Tabin CJ. 2000. Similar expression and regulation of Gli2 and Gli3 in the chick limb bud. Mech Dev 98: 171174.
  • Summerbell D. 1979. The zone of polarizing activity: evidence for a role in normal chick limb morphogenesis. J Embryol Exp Morphol 50: 217233.
  • Takahashi M, Tamura K, Buscher D, Masuya H, Yonei-Tamura S, Matsumoto K, Naitoh-Matsuo M, Takeuchi J, Ogura K, Shiroishi T, Ogura T, Belmonte JC. 1998. The role of Alx-4 in the establishment of anteroposterior polarity during vertebrate limb development. Development 125: 44174425.
  • te Welscher P, Zuniga A, Kuijper S, Drenth T, Goedemans HJ, Meijlink F, Zeller R. 2002. Progression of vertebrate limb development through SHH-mediated counteraction of GLI3. Science 298: 827830.
  • Thibert C, Teillet MA, Lapointe F, Mazelin L, Le Douarin NM, Mehlen P. 2003. Inhibition of neuroepithelial patched-induced apoptosis by sonic hedgehog. Science 301: 843846.
  • Todt WL, Fallon JF. 1987. Posterior apical ectodermal ridge removal in the chick wing bud triggers a series of events resulting in defective anterior pattern formation. Development 101: 501515.
  • Tole S, Ragsdale CW, Grove EA. 2000. Dorsoventral patterning of the telencephalon is disrupted in the mouse mutant extra-toes(J). Dev Biol 217: 254265.
  • Tümpel S, Sanz-Ezquerro JJ, Isaac A, Eblaghie MC, Dobson J, Tickle C. 2002. Regulation of Tbx3 expression by anteroposterior signalling in vertebrate limb development. Dev Biol 250: 251262.
  • Vargesson N, Patel K, Lewis J, Tickle C. 1998. Expression patterns of Notch1, Serrate1, Serrate2 and Delta1 in tissues of the developing chick limb. Mech Dev 77: 197199.
  • Wang B, Mysliwiec T, Krainc D, Jensen RA, Sonoda G, Testa JR, Golemis EA, Kruh GD. 1996. Identification of ArgBP1, an Arg protein tyrosine kinase binding protein that is the human homologue of a CNS-specific Xenopus gene. Oncogene 12: 19211929.
  • Wang B, Fallon JF, Beachy PA. 2000. Hedgehog-regulated processing of Gli3 produces an anterior/posterior repressor gradient in the developing vertebrate limb. Cell 100: 423434.
  • Warren AE. 1934. Experimental studies on the development of the wing in the embryo of Gallus domesticus. Am J Anat 54: 449485.
  • Wilson DJ, Hinchliffe JR. 1987. The effect of the zone of polarizing activity (ZPA) on the anterior half of the chick wing bud. Development 99: 99108.
  • Yang Y, Drossopoulou G, Chuang PT, Duprez D, Marti E, Bumcrot D, Vargesson N, Clarke J, Niswander L, McMahon A, Tickle C. 1997. Relationship between dose, distance and time in Sonic Hedgehog-mediated regulation of anteroposterior polarity in the chick limb. Development 124: 43934404.
  • Yokouchi Y, Sakiyama J, Kameda T, Iba H, Suzuki A, Ueno N, Kuroiwa A. 1996. BMP-2/-4 mediate programmed cell death in chicken limb buds. Development 122: 37253734.
  • Zeng X, Goetz JA, Suber LM, Scott WJ Jr, Schreiner CM, Robbins DJ. 2001. A freely diffusible form of Sonic hedgehog mediates long-range signalling. Nature 411: 716720.
  • Zou H, Niswander L. 1996. Requirement for BMP signaling in interdigital apoptosis and scale formation. Science 272: 738741.
  • Zuniga A, Haramis AP, McMahon AP, Zeller R. 1999. Signal relay by BMP antagonism controls the SHH/FGF4 feedback loop in vertebrate limb buds. Nature 401: 598602.