Gene Expression in the limbless Mutant: Polarized Gene Expression in the Absence of Shh and an AER

Transcription

1 DEVELOPMENTAL BIOLOGY 179, (1996) ARTICLE NO Gene Expression in the limbless Mutant: Polarized Gene Expression in the Absence of Shh and an AER Selina Noramly,*, Jacqueline Pisenti, Ursula Abbott, and Bruce Morgan*,1 *Cutaneous Biology Research Center, Massachusetts General Hospital, Harvard Medical School, Charlestown, Massachusetts 02129; Department of Avian Sciences, University of California, Davis, California 95616; and Graduate School of Medical Sciences, Cornell University, New York, New York Limb buds in the limbless chick begin to form normally but fail to form an AER and ultimately degenerate. Wnt7a and LMX-1, which are restricted to the dorsal half of a normal limb bud, are expressed throughout the ectoderm and mesenchyme of the mutant buds, respectively. Engrailed-1, normally expressed in ventral limb ectoderm, is not expressed in the limbless bud. This defect precedes the normal period of AER formation and no localized expression of genes normally found in the AER is observed in limbless buds. Consistent with the lack of molecular specialization of an AER, Shh and BMP-2 are not expressed in the ZPA of the mutant bud. Despite the lack of Shh, FGF-4, or BMP-2 expression, the hoxd genes are expressed at low levels in the posterior mesenchyme of the bud. Forced expression of Shh fails to rescue the positive feedback loop between the AER and the ZPA and does not lead to distal outgrowth. However, Shh does maintain the part of the bud formed during pre-aer stages. These results support the importance of the dorsal/ventral boundary in the initiation of AER formation and imply that Shh is not required for the initial activation of polarized hoxd gene expression during limb development Academic Press, Inc. INTRODUCTION The apical ectodermal ridge (AER) and zone of polarizing activity (ZPA) are two mutually dependent signaling cen- ters required for the normal formation of the limb bud. Sonic hedgehog (Shh) has been identified as an important mediator of polarizing activity based on its expression in the ZPA and its ability to exert polarizing activity when ectopically expressed at the anterior margin of the wing bud (Riddle et al., 1993). In a similar fashion, FGFs have been identified as important mediators of AER function based on expression in the ridge and the fact that the ridge can be functionally replaced by an exogenous source of FGF-2, -4, or -8 during the intermediate and late stages of limb devel- opment (Niswander et al., 1993; Fallon et al., 1994). FGF-2 and -8 are expressed in limb ectoderm prior to the formation of the AER as well as in the ridge itself (Savage et al., 1993; 1 To whom correspondence should be addressed. bmor gan@cbrc.mgh.harvard.edu. Heikinheimo et al., 1994; Crossley and Martin, 1995). FGF- 4 expression is first detected during the formation of the ridge and is restricted to it as development procedes. The AER and the ZPA are mutually dependent on each other for maintenance of FGF-4 and Shh expression (Laufer et al., 1994; Niswander et al., 1994). Furthermore, signals from both centers are required for the induction of many down- stream genes thought to mediate the morphogenetic re- sponse to polarizing activity. Hoxd-11, -12, and -13 as well as BMP-2 and -7 are all induced in the anterior limb bud in response to ectopic Shh, but only if FGF-4 or an AER is also present (Laufer et al., 1994; Niswander et al., 1994; B. A. Morgan, unpublished). In order to gain further insight into AER and ZPA function, we have undertaken an analysis of gene expression in mutant embryos with aberrant AERs. Here we describe a molecular analysis of the limbless embryo which lacks wings and legs but is otherwise apparently normal and viable (Prahlad et al., 1979). Leg and wing buds initiate properly in this mutant, but an AER fails to form normally. Subsequently the nascent bud degenerates and no limbs form /96 $18.00 Copyright 1996 by Academic Press, Inc. All rights of reproduction in any form reserved. 339

2 340 Noramly et al. The defect in this embryo appears to be confined to the ectoderm (Carrington and Fallon, 1988). Mutant ectoderm recombined with wild-type limb mesenchyme does not form an AER, while wild-type ectoderm grafted on mutant mesenchyme is capable of forming a normal ridge. Several genes normally expressed in the AER including msx-2 and IGF-1 are not observed in the limb bud ectoderm of the limbless mutant (Coehlo et al., 1991; Dealy and Kosher, 1996). We have undertaken a more extensive examination of gene expression during the period of limb formation in the limbless mutant to determine whether any of the genes whose expression is associated with ridge function are locally expressed in the absence of morphological differentiation. We have also extended this analysis to genes which presage AER formation and find that defects in gene expression are apparent in the limbless bud prior to the period when a ridge would normally form. Furthermore, none of the genes whose expression is associated with ridge formation or function are expressed in a spatially restricted pattern. Hence the mutant limb bud reflects development in the complete absence of AER function. Surgical removal of the AER has been used to infer genetic regulatory hierarchies involved in limb development. AER extirpation leads to profound changes in the patterns of gene expression in the limb. A number of genes expressed at the time of AER removal are downregulated, while others normally expressed at later stages fail to be induced (Ros et al., 1992; Hyamizu et al., 1994; Laufer et al., 1994; Niswander et al., 1994; Riddle et al., 1995; Vogel et al., 1995; Yang and Niswander, 1995). These experiments have also been combined with the forced expression of Shh on the anterior margin of the limb to show that the induction of BMP-2, -7, and hoxd-11 through 13 requires both Shh and FGF (Laufer et al., 1994; Niswander et al., 1994). From these experiments as well as normal patterns of gene expression during development, it has been inferred that Shh and FGF are required for the induction of these genes on the posterior margin of the normal limb bud. Although the limbless bud ultimately degenerates, it persists through intermediate stages of development. It is therefore possible to evaluate gene expression in the mutant limb buds that have been free of AER influence throughout their development in the absence of surgical perturbations. In the complete absence of ridge function, several of the genes thought to be depen- dent on the AER for expression, including Shh, are not ob- served. However, despite the lack of FGF-4 or Shh, we observe posteriorly restricted expression of the hoxd genes. embryos were identified by the characteristic defects of the limb buds including the absence of an AER and a flattened morphology of the wing bud. These defects are first apparent in the wing bud at a time when the leg bud is morphologically normal. Phenotypically normal siblings (presumably including limbless heterozygotes) and embryos from a wild-type white leghorn flock (SPAFAS, CT) showed indistinguishable patterns of gene expression with all probes employed in this study. Whole mount in situ hybridizations were performed as described (Riddle et al., 1993) with the exception that the proteinase K treatment was reduced to 5 min with a 3 mg/ml solution of proteinase K. Digoxygenin-labeled riboprobes were as described: Hoxd (Nelson et al., 1996), Shh (Riddle et al., 1993), BMP-2 (Laufer et al., 1994), FGF-4 (Niswander et al., 1994), TCF-1 (Gastrop et al., 1992), Cek- 3 (Patstone et al., 1993), ptc (Marigo et al., 1996), Wnt7a, LMX-1 (Riddle et al., 1995). Degenerate PCR was used to isolate chick copies of BMP-2, -4, and -7 and engrailed-1. These probes were used to isolate full-length cdnas from a chick limb bud libraries. Shh hedgehog injections were performed essentially as described (Morgan and Fekete, 1996). A infectious units per milliliter of stock of RCASBP(A) Shh (Riddle et al., 1993) was injected into the primordium of the right leg bud at stage 14. Embryos were harvested at stage 26 and infection was assessed by whole mount in situ hybridization. RESULTS Markers of the Dorsal/Ventral Axis Are Abnormally Expressed The position of the presumptive ridge is marked in the stage 17 limb bud by the interface of Wnt7a-expressing cells on the dorsal surface of the limb and engrailed-expressing cells on the ventral surface. Wnt7a is clearly restricted to the dorsal ectoderm of the stage 17/18 limb bud of the wildtype embryo. However, in the limbless mutant Wnt7a is expressed in all limb ectoderm including the ventral surface of the bud at all stages examined (stages 18 23, Fig. 1 and data not shown). In a complementary fashion, engrailed-1 is normally expressed in the ventral ectoderm by stage 18 and continues to be expressed throughout limb bud develop- ment (Davis et al., 1991). In the limbless mutant, engrailed- 1 is not detected in the limb ectoderm at any stage. Consis- tent with these observations, LMX-1 is also abnormally ex- pressed. LMX-1 expression in the dorsal mesenchyme of a normal limb is induced by Wnt7a and is sufficient to impose dorsal character on expressing tissue (Riddle et al., 1995; Vogel et al., 1995). In the limbless mutant, LMX-1 is ex- pressed throughout the limb bud. MATERIALS AND METHODS The limbless line maintained at U.C. Davis was derived by out- crossing limbless chickens to the UCD line 003 to develop a uniform genetic background. This strain has been maintained independently of the Wisconsin flock for over 15 years. Homozygous limb- less embryos derived from heterozygous matings were harvested from stage 17 (Hamburger and Hamilton, 1951) onward. Mutant Genes Normally Expressed in the AER The growth factors FGF-4, BMP-2, BMP-4, and BMP-7, as well as the transcription factors TCF-1 and Msx-2 and the FGF receptor CEK-3 are all expressed in the AER from the earliest stages of its formation (Niswander et al., 1992; Francis et al., 1994; Francis-West et al., 1995; Gastrop et al., 1992; Coehlo et al., 1991; Patstone et al., 1993; and data

3 Molecular Analysis of limbless 341 not shown). None of these genes were specifically expressed in the apical ectoderm of the limbless buds between stages 17 and 24 (Fig. 2 and data not shown). Wild-type embryos processed in the same tube served as positive controls for detection of these genes in each experiment. In most cases, expression of the gene elsewhere in the embryo served as an internal control as well. Thus the failure of ridge formation is accompanied by a lack of molecular specialization as well. Genes Thought to Be Dependent on the Ridge for Expression The maintenance of Shh expression in the ZPA is depen- dent on FGF from the AER (Niswander et al., 1994; Laufer et al., 1994) and it has been postulated that the initiation of Shh expression is dependent on FGF from the ridge (Crossley and Martin, 1995). Consistent with these hypotheses, Shh expres- sion was not observed in the limb buds of limbless embryos. Thirty embryos ranging from stage 17 through stage 24 were examined. At younger stages, expression of Shh in the floor plate of the spinal chord served as a positive control for these experiments. Wild-type embryos exhibiting similar staining in the floor plate showed robust expression of Shh in the limb buds (Fig. 3 and data not shown). Shh represses patched protein activity which in turn represses transcription of the patched gene (ptc). The ptc transcription therefore serves as a sensitive indicator of Shh activity even when very low levels of Shh protein are produced (Goodrich et al., 1996; Marigo et al., 1996). The induction of ptc transcription by Shh does not require FGF and should occur despite the lack of FGF-4 in the limbless bud if Shh is present (Marigo et al., 1996). No localized transcription of patched was observed in the limbless buds (Fig. 3). Our inabil- ity to detect either Shh or ptc transcripts argues that Shh is not produced in the mutant limb bud. In the wild-type limb, hoxd-11 and Shh are expressed syn- chronously at the posterior margin of the limb bud (Laufer et al., 1994). Hoxd-12 and -13 are expressed subsequently in progressively restricted posterior domains. The fact that exoge- nous Shh is capable of inducing these genes in the anterior margin of the limb in a similar spatial and temporal progression led to the proposal that their expression in the posterior limb is dependent on Shh (Nelson et al., 1996). However, despite the lack of Shh in the limbless limb bud, all three genes are expressed in posteriorly restricted domains (Fig. 3). The level of expression of the hoxd genes is consistently lower in the mutant limb buds than that observed in wild-type controls, but the pattern of expression is very similar to that of wild type at early stages. Because the mutant limb buds begin to degenerate by stage 20, the complex maturation of hoxd gene expression patterns observed during the second and third phase of limb bud outgrowth cannot be assessed in the mutant. In a similar fashion, the normal spatial and temporal patterns of BMP-2 and BMP-7 expression as well as their induction in the anterior limb bud by Shh suggested that they were also dependent on Shh for polarized expression (Francis et al., 1994; Laufer et al., 1994; Francis-West et al., 1995, and data not shown). Bmp-2 and -7 were absent from the ectoderm and mes- enchyme of the mutant buds (Fig. 2 and data not shown). Forced Expression of Shh Does Not Rescue the Mutant Defect Sonic hedgehog and FGF-4 are known to be involved in a positive feedback loop which is required for the mainte- FIG. 1. The dorsal ventral boundary of the limb is abnormal in limbless buds. Yellow arrows mark the dorsal/ventral axis of the limb bud. Wnt7a is expressed in the ectoderm across the dorsal ventral axis of the limbless wing bud (left) but is restricted to the dorsal surface of the limb in a wild-type embryo (right). This restriction precedes the formation of the AER. Engrailed-1 (En) is not detected in the ectoderm of the mutant limb bud (left) even when the detection reaction is overexposed. Expression in ectoderm of the ventral trunk serves as a positive control for the detection of ectodermal expression. In the wild-type limb bud, (right), en-1 is expressed in the ventral ectoderm of the bud. LMX-1 is expressed across the dorsal/ventral axis of the mutant limb bud (left), but is restricted to the dorsal mesenchyme in the wild-type bud (right). Lower left panel shows a broken limbless bud revealing expression of LMX-1 across the D/V axis (ll, limbless; w.t., wild type). FIG. 2. Apical expression of ridge markers is not observed in the limbless bud. FGF-4 is not expressed in the ectoderm of the limbless bud (left) but is prominently expressed in the AER of a wild-type embryo (right). Expression in the body wall is readily detected in both embryos. BMP-2 and BMP-7 are absent from the limbless bud in either ectoderm or mesenchyme (left) but are expressed at high levels in the AER and posterior mesenchyme of a wild-type bud (right). Cek-3 is expressed in the AER of the wild-type limb but is not detected in the mutant limb bud. TCF-1 is not expressed in the limbless bud, but is expressed in the AER and progress zone of the wild-type bud. FIG. 3. Hoxd genes are expressed in a polarized fashion in the absence of Shh. Shh is not observed in the limb buds of a mutant embryo (ll), although expression in the floor plate and gut is normal. Shh expression is robust in the limbs of a wild-type (w.t.) embryo that has similar levels of signal in the floor plate and gut (left). Ptc expression is not detected in the limb bud of the mutant (left) even when overdeveloped, but expression elsewhere in the embryo is normal. Patched is highly expressed in the ZPA of a normal limb bud (right, w.t.). Hoxd-11 is normally expressed in the posterior half of the limb bud (w.t.). Hoxd-11, -12, and -13 are expressed at low levels in the posterior mesenchyme of the mutant limb buds (ll). Expression of hoxd-13 is particularly weak and only detectable in some embryos. FIG. 4. Forced expression of Shh in the mutant limb bud fails to rescue distal outgrowth. Expression of Shh in the infected embryos was used to confirm infection (arrows). The infected buds were consistently larger than the contralateral uninfected buds at stage 27. However, at this stage, wild-type siblings have a distally complete limb. In the embryo shown at right, expression of Shh outside the leg bud had no effect (red arrow), while expression of Shh in the wing bud of the same embryo (black arrow), or the leg bud of a sibling (yellow arrow) led to maintenance of the limb tissue.

4 342 Noramly et al.

5 343

6 344 Noramly et al. nance of the AER (Laufer et al., 1994; Niswander et al., 1994). The lack of Shh expression suggested that the defect in limbless could result from the failure to generate a signal necessary for Shh expression or maintenance which in turn could prevent maintenance of a nascent AER. To test this possibility, we infected the right leg bud primordia of limbless embryos with a retrovirus encoding the Shh protein. Eight mutant embryos were infected with the virus. Although all showed some additional growth of the infected limb bud when compared with the uninfected limbs, none formed a substantially complete limb (Fig. 4). It appears that the expression of Shh can maintain the part of the bud which forms during early development in a limbless mutant. However, it cannot sustain outgrowth and patterning of the mutant limb. DISCUSSION The limbless mutation offers a unique opportunity to study the role of the AER in limb bud formation without resorting to surgical manipulations which may have unintended effects on development. Our analysis of genes expressed in the ridge suggests that ridge induction is completely blocked. Of the genes normally expressed in the AER which were surveyed, none are expressed in the ectoderm in a spatially restricted pattern that might suggest the initial stages of ridge formation. During normal limb bud development, BMP-2 and msx-2 are expressed in the forming ridge prior to overt differentiation of the structure. The lack of localized expression of these genes demonstrates that AER formation is blocked at a very early stage. This conclusion is supported by changes in the expression of genes which presage the position of the ridge and precede the expression of ridge genes. Neither Wnt7a nor engrailed- 1 are strictly required for the formation of the apical ectodermal ridge in the mouse. Mice homozygous for null mutations for either gene form an AER at the correct place and time, although the function and maintenance of the ridge may be defective in both cases (Parr and McMahon, 1995; Wurst et al., 1994). Nevertheless, the expression of these genes serve as markers of the dorsal/ventral boundary which appears to be important in the formation of the apical ectodermal ridge (Cohn et al., 1995; Crossley et al., 1996). As judged by the expression of both genes, this boundary is defective in the limbless mutant. The expression of Wnt7a across the D/V axis of the limb and the lack of engrailed expression imply that this boundary is not properly formed in limbless buds. Wnt7a is not normally expressed in the ridge, and its expression throughout the ectoderm could be thought to contribute to the failure to form the AER. However, this is unlikely to be the case. The localized expression of BMP-2 and msx-2 precedes the repression of Wnt7a in the forming ridge (B.A.M., unpublished data). The lack of localized expression of these genes in the ectoderm suggests that the formation of the AER is blocked before any requirement for the repression of Wnt7a expression. Shh is expressed in the posterior mesenchyme of the limb bud at the time that hoxd-11 gene expression is first observed in the posterior margin of the limb (Laufer et al., 1995). This fact coupled with the observation that exoge- nously supplied Shh is capable of inducing the ectopic expression of hoxd genes when applied to the anterior margin of the limb led to the hypothesis that Shh is the endogenous inducer of their expression in the limb bud (Riddle et al., 1993; Nelson et al., 1996). The lack of Shh expression in the limbless limb bud implies that this is not the case. Although Shh may be involved in the maintenance or amplification of expression, the initial expression of these genes in the posterior half of the limb bud appears to be independent of Shh activity. Although we cannot formally exclude the possibility that a low level of Shh expression is responsible for this restricted hoxd gene expression, extended development of the color reaction gave no indication of Shh expression and suggests that RNA levels must be reduced at least 50-fold. Furthermore, no transcription of the patched gene was observed. Based on these observations, we conclude that the initial polarized expression of the hoxd genes is not dependent on Shh expression but may instead be dependent on the same signals that localize Shh expression to the posterior margin. Analysis of Shh mutants will be required to confirm this conclusion. The lack of BMP-2 and BMP-7 expression in the mutant limb bud is consistent with the proposal that these genes are downstream of Shh, although we cannot rule out that their activation is depen- dent directly on the signals which normally activate Shh in the posterior limb. Duprez and co-workers (1996) have proposed that these BMPs are required for activation of hoxd genes in the distal limb at later stages but are not required for the initial stage of hox gene activation in the limb. Gene expression in the limbless mutant supports this proposal. When genes are involved in a positive feedback loop, it is difficult to determine whether the inability to detect expression reflects a true failure to initiate gene expression or merely a failure to amplify that initial level of expression. We have attempted to distinguish between these possibilities by forcing the expression of Shh in the limbless bud to determine whether this would establish the feedback loop necessary to amplify and maintain ridge gene expression and thereby rescue the limb. The infected limbs do not degenerate, but they fail to undergo significant distal out- growth. It appears that expression of Shh in the limb can maintain the bud formed during the early stages of develop- ment, but is not able to induce the FGFs required for distal outgrowth and patterning. The study of mutant chick embryos provides a powerful complement to embryological manipulations performed on wild-type embryos in the study of limb development. Analysis of the limbless mutant supports the hypothesis that the juncture of dorsal and ventral tissue is important for the generation of ectoderm competent to form an AER. The limbless mutant seems to be completely devoid of any AER function and does not express genes associated with the ZPA either. Despite this fact, the hoxd genes are induced

7 Molecular Analysis of limbless 345 in a polarized fashion in the mutant limb bud. Analysis of Goodrich, L. V., Johnson, R. L., Milenkovic, L., McMahon, J. A., additional mutant strains will continue to refine our undersignaling and Scott, M. P. (1996). Conservation of the hedgehog/patched standing of the genetic hierarchies which orchestrate limb pathway from flies to mice: induction of a mouse patched gene by Hedgehog. Genes Dev. 10, development. Hamburger, V., and Hamilton, H. L. (1951). A series of normal stages in the development of the chick embryo. J. Exp. Morphol. ACKNOWLEDGMENTS 88, Hayamizu, T. F., Wanek, N., Taylor, G., Trevino, C., Shi, C., Anderson, R., Gardiner, D. M., Muneoka, K., and Bryant, S. (1994). We thank H. K. Song, R. Riddle, C. Tabin, and L. Niswander for Regeneration of HoxD expression domains during pattern regulaprobes and P. Goetinck for comments on the manuscript. This tion in chick wing buds. Dev. Biol. 61, work was supported by a grant from Shisheido Corp. (B.A.M.) and Heikinheimo, M., Lawshe, A., Shackleford, G., Wisol, D., and Macfunds from the University of California Genetics Resource Conser- Arthur, C. (1994). FGF-8 expression in the post-gastrulation vation Program (U.A.). mouse suggests roles in the development of the face, limbs, and central nervous system. Mech. Dev. 48, Laufer, E., Nelson, C. E., Johnson, R. L., Morgan, B. A., and Tabin, C. (1994). Sonic hedgehog amd Fgf-4 act through a signaling cas- REFERENCES cade and feedback loop to integrate growth and patterning of the developing limb bud. Cell 79, Carrington, J. L., and Fallon, J. F. (1988). Initial limb budding is Marigo, V., Scott, M. P., Johnson, R. L., Goodrich, L. V., and Tabin, independent of apical ectodermal ridge activity: Evidence from C. J. (1996). Conservation in hedgehog signaling: Induction of a a limbless mutant. Development 104, chicken patched homolog by Sonic hedgehog in the developing Coelho, C. N. D., Krabbenhoft, K., Upholt, W. B., Fallon, J. F., and limb. Development 122, Kosher, R. A. (1991a). Altered expression of the chicken homeo- Morgan, B., and Fekete, D. (1996). Manipulating gene expression box-containing genes GHox-7 and GHox-8 in the limb buds of with replication competent retroviral vectors. In Methods in Limbless mutant chick embryos. Development 113, Avian Embryology (M. Bronner Frasier, Ed.), Methods in Cell Cohn, M. J., Izpisua-Belmonte, J. C., Abud, H., and Heath, J. K. Biology, Vol. 51, Academic Press, San Diego. (1995). Fibroblast growth factors induce additional limb develop- Nelson, C., Morgan, B., Burke, A., Laufer, E., Murtaugh, C., Di ment from the flank of chick embryos. Cell 80, Mambro, E., Gonzales, E., Tessarollo, L., Parada, L., and Tabin, Crossley, P., and Martin, G. R. (1995). The mouse FGF-8 gene en- C. (1996). Analysis of Hox gene expression in the chick limb. codes a family of polypepetides and is expressed in regions that Development 122, direct outgrowth and patterning in the developing embryo. De- Niswander, L., Jeffrey, S., Martin, G. R., and Tickle, C. (1994). A velopment 121, positive feedback loop coordinates growth and patterning in the Crossley, P. H., Minowada, G., MacArthur, C. A., and Martin, vertebrate limb. Nature 371, G. R. (1996). Roles for FGF8 in the induction, initiation, and Niswander, L., and Martin, G. R. (1992). Fgf-4 expression during maintenance of chick limb development. Cell 84, gastrulation, myogenesis, limb and tooth development in the Davis, C., Holmyard, H., Millen, K., and Joyner, A. L. (1991). Exammouse. Development 114, ining pattern formation in mouse chicken and frog embryos with an En-specific antiserum. Development 111, Niswander, L., Tickle, C., Vogel, A., Booth, I., and Martin, G. R. Dealy, C. N., and Kosher, R. A. (1996). IGF-I, insulin and FGFs (1993). FGF-4 replaces the apical ectodermal ridge and directs induce out growth of the limb buds of amelic mutant chick emoutgrowth and patterning of the limb. Cell 75, bryos. Development 122, Parr, B. A., and McMahon, A. P. (1995). Dorsalizing signal Wnt-7a Duprez, D., Kostakopoulo, K., Francis-West, P., Tickle, C., and required for normal polarity of D-V and A-P axes of mouse limb. Brickel, P. (1996). Activation of FGF-4 and HoxD gene expression Nature 374, by BMP-2 expressing cells in the developing chick limb. Develop- Parr, B. A., Shea, M. J., Vassileva, G., and McMahon, A. P. (1993). ment 122, Mouse Wnt genes exhibit discrete domains of expression in the Fallon, J., Lopez, A., Ros, M., Savage, M., Olwin, B., and Simandl, early embryonic CNS and limb buds. Development 119, 247 B. (1994). FGF-2: Apical ectodermal ridge growth signal for chick 261. limb development. Science 264, Patstone, G., Pasquale, E. B., and Maher, P. A. (1993). Different Francis, P. H., Richardson, M. K., Brickell, P. M., and Tickle, C. members of the fibroblast growth factor receptor family are spe- (1994). Bone morphogenetic proteins and a signaling pathway cific to distinct cell types in the developing chicken embryo. that controls patterning in the developing chick limb. Develop- Dev. Biol. 155, ment 120, Prahlad, R., Skala, G., Jones, D., and Briles, W. (1979). Limbless: A Francis-West, P. H., Robertson, K. E., Ede, D. A., Rodriquez, C., new genetic mutant in the chick. J. Exp. Zool. 209, Izpisua-Belmonte, J. C., Houston, B., Burt, D. W., Gribbin, C., Riddle, R. D., Ensini, M., Nelson, C., Tsuchida, T., Jessell, T. M., Brickell, P. M., and Tickle, C. (1995). Expression of genes encodby and Tabin, C. (1995). Induction of the LIM homeobox gene Lmx1 ing bone morphogenetic proteins and sonic hedgehog in talpid WNT7a establishes dorsoventral pattern in the vertebrate (ta3) limb buds: Their relationships in the signaling cascade in- limb. Cell 83, volved in limb patterning. Dev. Dynamics 203, Riddle, R. D., Johnson, R. L., Laufer, E., and Tabin, C. (1993). Sonic Gastrop, J., Hoevenagel, R., Young, J. R., and Clevers, H. C. (1992). hedgehog mediates the polarizing activity of the ZPA. Cell 75, A common ancestor of the mammalian transcription factors TCF-1 and TCF-1 alpha/lef-1 expressed in chicken T cells. Eur. Robert, B., Lyons, G., Simandl, B. K., Kuroiwa, A., and Buckingham, J. Immunol. 22, M. (1991). The apical ectodermal ridge regulates Hox-7 and Hox-

8 346 Noramly et al. 8 gene expression in developing chick limb buds. Genes. Dev. 5, (1995). Dorsal cell fate specified by chick Lmx1 during vertebrate limb development. Nature 378, Ros, M. A., Lyons, G., Kosher, R. A., Upholt, W. B., Coelho, Wurst, W., Auerbach, A. B., and Joyner, A. L. (1994). Multiple devel- C. N. D., and Fallon, U. F. (1992). Apical ridge dependent and opmental defects in Engrailed-1 mutant mice: an early mid-hindindependent mesodermal domains of GHox-7 and GHox-8 ex- brain deletion and patterning defects in forelimbs and sternum. pression in chick limb buds. Development 116, Development 120, Savage, M., Hart, C., Riley, B., Sasse, J., Olwin, B., and Fallon, J. Yang, Y., and Niswander, L. (1995). Interaction between the signal- (1993). The distribution of FGF-2 suggests it has a role in chick ing molecules WNT7a and SHH during vertebrate limb developlimb bud growth. Dev. Dyn. 198, ment: Dorsal signals regulate anteroposterior patterning. Cell 80, Vogel, A., Roberts-Clarke, D., and Niswander, L. (1995). Effect of FGF on gene expression in chick limb bud cells in vivo and in vitro. Dev. Biol. 171, Received for publication July 22, 1996 Vogel, A., Rodriguez, C., Warnken, W., and Izipsua-Belmonte, J. C. Accepted August 7, 1996