During Xenopus gastrulation, platelet-derived growth factor (PDGF) receptor-α is expressed in involuting marginal zone cells which migrate over ectodermal cells expressing PDGF-A. To investigate the role of PDGF signalling during this process, we have generated a novel point mutant of PDGF receptor-α analogous to the W37 mutation of c-kit. This molecule is a specific, potent, dominant inhibitor of PDGF signalling in vivo. Injection of RNA encoding this protein into Xenopus embryos prevents closure of the blastopore, leads to abnormal gastrulation and a loss of anterior structures. Convergent extension is not inhibited in these embryos, but rather, involuting mesodermal cells fail to adhere to the overlying ectoderm. PDGF may therefore be required for mesodermal cell-substratum interaction.

Gastrulation involves a complex, co-ordinated series of cell movements that give rise to the body plan in all vertebrates. In the frog, Xenopus laevis, this process begins with the involution of cells on the dorsal side of the embryo. Apical constriction of the outermost epithelial cells gives rise to the blastopore lip and precedes the involution of mesodermal cells from the marginal zone (for review, see Keller, 1991). As these mesodermal cells involute, they migrate over the inner surface of the blasotcoel roof toward the future anterior end of the embryo. The migration and fate of these involuting cells are presumed to be controlled by extracellular signals, although these are as yet poorly understood.

A candidate signalling molecule that may be involved in gastrulation is platelet-derived growth factor A (PDGF-A, Jones et al., 1993). We now show by in situ hybridisation that PDGF-A mRNA is present in the blastocoel roof of the early gastrula stage Xenopus embryo. This tissue forms the substratum for the migration of marginal zone cells which express the receptor for PDGF-A, PDGFR-α.

In order to assess the role of PDGF-A during Xenopus gastrulation, we have generated a novel, dominant negative mutant of its receptor, PDGFR-α. This mutant is based on the W37 mutation of the homologous receptor tyrosine kinase c-kit (Nocka et al., 1990; Reith et al., 1990). The amino acid residue affected in W37 is conserved throughout the PDGF receptor family and is therefore thought to be essential for function. Indeed, a mutation of c-fms corresponding to W37 c-kit abolishes the kinase activity of this receptor. More importantly, the W-like fms mutants can act in a trans-dominant manner to prevent tyrosine phosphorylation of the wild-type receptor and to block colony stimulating factor-1-dependent transformation of Rat-2 cells (Reith et al., 1993).

Here we demonstrate that a point mutant of Xenopus PDGFR-α analogous to the W37 mutation of c-kit lacks detectable kinase activity and acts in a dominant negative manner to inhibit signalling by the wild-type receptor in vivo. We have used this mutant receptor to disrupt PDGF-A signalling in the Xenopus embryo and show that PDGF is necessary for normal gastrulation but is not required for mesoderm induction. We further demonstrate that defects arise due to the inappropriate behaviour of mesodermal cells and discuss possible mechanisms that may lead to these abnormalities.

Embryos

Xenopus embryos were fertilized in vitro and chemically dejellied using 2% cysteine-HCl, pH 7.8, then maintained in 10% Marc’s modified Ringer’s (0.1× MMR, Peng, 1991) at temperatures between 13 and 20°C until they had reached the appropriate stages (according to Nieuwkoop and Faber, 1967) for injection or analysis. Microinjection was performed in a solution of 3% Ficoll in 1× MMR using bevelled, glass capillary needles. Xenopus animal poles were explanted at stage 8 and maintained in 75% MMR/0.1% bovine serum albumin until control embryos had reached the required stage. Growth factors (50 ng/ml PDGF-AA, Gibco-BRL, 500 ng/ml bFGF, Amgen and 2 U/ml βA activin, from transfected COS cell supernatant) were added directly to this medium immediately after explants were excised. Exogastrulae were induced by manually removing the vitelline membrane of embryos at stage 8 and transferring to 1× MMR in dishes coated with 1% agarose in 1× MMR. β-galactosidase activity was detected by standard procedures using the chromogenic substrate X-gal.

In situ hybridisation

For radioactive in situ hybridisation experiments, albino embryos were fixed in 4% paraformaldehyde in 70% PBS for 1 hour at 4 °C prior to embedding in paraffin. 8 μm sections were mounted on glass slides and hybridised to 35S-labeled RNA probes as described (O’Keefe et al., 1991). Serial sections were hybridised to sense and antisense cRNA probes for Xenopus PDGF-A (Mercola et al., 1988) and PDGFR-α (Jones et al., 1993). In no case was a hybridisation signal detected with sense probes. After washing, slides were coated with Kodak NTB-2 emulsion and exposed for 2 weeks at 4 °C. For whole-mount in situ hybridisation, pigmented embryos were fixed in 4% formaldehyde, dehydrated in methanol and then processed as described (Ho et al., 1994). Digoxigenin-labelled cRNA probes for goosecoid and brachyury (Amaya et al., 1993) were the kind gift of Pascal Stein. Embryos were dehydrated in methanol and in some cases bleached in 10% hydrogen peroxide and cleared in benzyl alcohol:benzyl benzoate (BA:BB, 1:2) prior to photography.

Site-directed mutagenesis

A cDNA clone encoding the full-length Xenopus PDGFR-α (Jones et al., 1993) was subcloned as a SacI/BglII fragment into pALTER-1 (Promega) and mutagenised by single-stranded, site-directed mutagenesis according to manufacturer’s instructions. A control frameshift mutation was generated in the full-length receptor by digestion with KpnI, blunting and re-ligation, causing premature termination in the extracellular domain of the protein at amino acid 159.

DNA cloning/RNA production

The wild-type, frameshift and mutagenised Xenopus PDGFR-α were subcloned into pGHE2, a vector containing a poly(A) tail and 5′ and 3′ untranslated sequences from Xenopus β-globin (a gift from the late Peter Hess, Harvard Medical School) such that sense RNA transcripts were produced by using T7 RNA polymerase. Capped mRNA was made using the mMessage Machine system (Ambion) according to manufacturer’s instructions. β-galactosidase mRNA containing a nuclear localisation signal was produced from the SP64nuc-βgal plasmid (Vize and Melton, 1991).

Phosphotyrosine detection

COS cells were passaged at 1×106 cells per 100 mm dish, 24 hours before DEAE-Dextran transfection. Cells were transfected with either 10 μg pMT2, 2 μg or 5 μg Xenopus PDGFR-α in pMT2, 10 μg PDGFR-37 in pMT2, 5 μg Xenopus PDGFR-α plus 5 μg PDGFR-37 or 2 μg Xenopus PDGFR-α plus 8 μg PDGFR-37. The total amount of DNA was adjusted to 10 μg with pMT2 where necessary. After 24 hours, cells were transferred to DMEM containing 0.5% calf serum for a further 24 hours. Cells were treated with 20 ng/ml PDGF-BB (Gibco-BRL) for 10 minutes, washed in ice-cold PBS, then lysed as described previously (Chen and Blenis, 1990). Insoluble material was removed by pelleting in a microcentrifuge for 10 minutes at 13,000 revs/minute. Approximately 25 μg of total cellular protein was fractionated by SDS-PAGE, then transferred to nitrocellulose membranes. Blots were probed with 4G10 anti-phosphotyrosine antibody (UBI) and proteins detected via ECL (Amersham).

Transformation assays

NIH-3T3 cells were passaged onto 60 mm dishes at 5×105 cells per dish, 24 hours before transfection. Cells were transfected using DEAE-Dextran with either 20 μg vector alone (pMT2, Genetics Institute), 2 μg v-sis and 18 μg pMT2, or 2 μg v-sis and 18 μg Xenopus PDGFR-37 in pMT2. In all cases, cells were co-transfected with 2 μg of pHM24 (the kind gift of Dr J. Dougherty, UMDNJ), a plasmid that confers hygromycin resistance (Mikawa et al., 1991). After 48 hours, each transfected dish was passaged onto three 60 mm dishes and grown in DMEM containing 0.5% calf serum and 100 μg/ml hygromycin for a further 12 days. Cells were fixed in methanol and colonies visualised by staining with Giemsa stain (Sigma).

Histology

Embryos were fixed in 4% paraformaldehyde in 70% PBS, dehydrated via an ethanol series then transferred to butanol. They were infiltrated with paraffin wax then embedded and sectioned at 8–10 μm. Sectioned tissue was dewaxed in xylene, rehydrated and stained with Feulgen, light green and Orange G by the method of Cooke (Cooke, 1979). Stained tissue was dehydrated, mounted in Permount and viewed using an inverted microscope.

RT-PCR

RNA prepared from animal pole explants was reverse transcribed using the Stratascript RT-PCR kit (Stratagene) according to manufacturer’s instructions. cDNA was amplified by PCR in the presence of [32P]dATP using primers specific for EF1-α and brachyury as described (Wilson and Melton, 1994). Amplified cDNA products were detected by autoradiography following electrophoresis on 5% non-denaturing polyacrylamide gels.

Distribution of PDGF-A and PDGFR-α in Xenopus embryos

Using in situ hybridisation we have determined that PDGF-A mRNA is distributed uniformly throughout the inner layers of ectodermal cells of the blastocoel roof, but not in the superficial layer of early gastrula stage Xenopus embryos (Fig. 1A). PDGFR-α mRNA is localised to the equatorial region of the embryo in cells of the marginal zone (Fig. 1B). During gastrulation, cells of the marginal zone begin to involute at the dorsal side of the embryo and move over the inner surface of the blastocoel roof towards the future anterior end of the embryo (reviewed in Keller, 1991). These cells express PDGFR-α and first come into contact with cells expressing PDGF-A shortly after the onset of gastrulation. This complementary pattern of ligand and receptor expression persists throughout gastrulation (Fig. 1C-F) and suggests that signalling between these cell populations may play an important role in controlling cellular movement or differentiation during this process.

Fig. 1.

In situ hybridisation of PDGF A and PDGFRα mRNA during the gastrula stages of development. Serial sections of gastrula staged embryos were probed with 35S-labelled antisense PDGF-A and PDGFR-α cRNA probes. PDGF A is expressed in the presumptive ectoderm (A,C,E) and PDGFR-α is expressed in the presumptive mesoderm (B,D,F). (A,B) Sagittal sections through the dorsal midline of a stage 10.25 embryo. (C,D) Parasagittal and (E,F) horizontal sections through a stage 12 embryo. An, animal pole; Vg vegetal pole; D, dorsal; V ventral; A, anterior; P, posterior; bl. blastocoel; a, archenteron. Scale bar, 100 μm.

Fig. 1.

In situ hybridisation of PDGF A and PDGFRα mRNA during the gastrula stages of development. Serial sections of gastrula staged embryos were probed with 35S-labelled antisense PDGF-A and PDGFR-α cRNA probes. PDGF A is expressed in the presumptive ectoderm (A,C,E) and PDGFR-α is expressed in the presumptive mesoderm (B,D,F). (A,B) Sagittal sections through the dorsal midline of a stage 10.25 embryo. (C,D) Parasagittal and (E,F) horizontal sections through a stage 12 embryo. An, animal pole; Vg vegetal pole; D, dorsal; V ventral; A, anterior; P, posterior; bl. blastocoel; a, archenteron. Scale bar, 100 μm.

Construction of a dominant negative mutant of PDGFR-α

To examine the function of PDGF-A during Xenopus development, we engineered a dominant negative mutant of its receptor corresponding to the W37 mutation of c-kit (Nocka et al., 1990; Reith et al., 1990). This mutation (illustrated schematically in Fig. 2) was named PDGFR-37. As expected, PDGFR-37 was not tyrosine phosphorylated in response to PDGF and prevented the tyrosine phosphorylation of wild-type PDGFR-α in a dose-dependent manner in COS cells (Fig. 3A). COS cells transfected with vector alone showed ligand-dependent phosphorylation of a band of ∼170×103Mr after stimulation with PDGF, corresponding to endogenous receptor. Cells transfected with Xenopus PDGFR-α showed an additional tyrosine-phosphorylated band migrating at ∼180×103Mr. This upper band was not seen in COS cells transfected with ten times the amount of PDGFR-37 DNA. Furthermore, when COS cells were co-transfected with wildtype PDGFR-α and increasing amounts of PDGFR-37, there was a corresponding decline in the amount of tyrosine phosphorylation of wild-type Xenopus PDGFR-α. A four-fold excess of PDGFR-37 DNA was sufficient to abolish all detectable tyrosine phosphorylation of Xenopus PDGFR-α. PDGFR-37 also blocked the v-sis-dependent transformation of NIH-3T3 cells (Fig. 3B). While PDGFR-α was able to induce elongation of Xenopus animal pole explants (an indication of mesoderm induction, Symes and Smith, 1987), PDGFR-37 was not (Fig. 3C). However, PDGFR-37 did not interfere with activin- or FGF-induced explant elongation. A molecular analysis of these growth factor-treated animal pole explants correlated with the observed morphological response (Fig. 4). Thus, the mesodermal gene brachyury (Smith et al., 1991b) was not expressed in untreated animal pole explants from frameshift-RNA-injected embryos. Treatment of explants with either FGF or activin caused induction of brachyury expression, while PDGF-A did not. Explants from PDGFR-37-injected embryos responded identically to frameshift-RNA-injected embryos, showing robust induction of brachyury when treated with activin or FGF but not when treated with PDGF-A. Together, the above experiments demonstrate that PDGFR-37 is inactive as a kinase and is able specifically to inhibit PDGFR-α signalling in a trans-dominant manner. However, PDGFR-37 does not interfere with the induction of the mesodermal marker, brachyury, by FGF or activin. We therefore used this molecule to disrupt PDGF-A signalling in intact Xenopus embryos.

Fig. 2.

Schematic representation of PDGFR-37 mutation. PDGFR-α consists of five immunoglobulin-like repeats in the extracellular domain, a single transmembrane domain (dark stippling) and a split catalytic domain (hatching) with intervening kinase insert. The nucleotide and predicted amino acid sequence of Xenopus PDGFR-α in the region adjacent to the first catalytic domain are shown at top, and the corresponding region in PDGFR-37 below. The sequence of the mutagenesis oligonucleotide is shown in bold type.

Fig. 2.

Schematic representation of PDGFR-37 mutation. PDGFR-α consists of five immunoglobulin-like repeats in the extracellular domain, a single transmembrane domain (dark stippling) and a split catalytic domain (hatching) with intervening kinase insert. The nucleotide and predicted amino acid sequence of Xenopus PDGFR-α in the region adjacent to the first catalytic domain are shown at top, and the corresponding region in PDGFR-37 below. The sequence of the mutagenesis oligonucleotide is shown in bold type.

Fig. 3.

PDGFR-37 lacks tyrosine kinase activity and acts as a dominant inhibitor. (A) Quiescent, transiently transfected COS cells were treated with PDGF (+) or control medium (−). Tyrosine phosphorylated proteins were detected following SDS-PAGE. Endogenous PDGF receptor migrates at ∼170×103Mr (left pointing arrowhead). Transfected Xenopus PDGF receptor migrates at ∼180×103Mr (right pointing arrowhead). Phosphorylated endogenous receptor was detected in PDGF-treated COS cells transfected with 10 μg of an empty expression vector (pMT2) and in non-transfected NIH-3T3 cells (NIH-3T3). COS cells transfected with 2 μg or 5 μg of Xenopus wild-type PDGFR-α (PDGFR2, 5) contained a small amount of phosphorylated receptor in untreated cells (PDGFR2−, PDGFR5−) which increased on addition of PDGF (PDGFR2+, PDGFR5+). Endogenous phosphorylated PDGF receptor was also detected in PDGF-treated COS cells transfected with 10 μg of PDGFR-37, but no phosphorylation of the transfected mutant receptor was detected (PDGFR-37). Co-transfection of an equal amount of PDGFR-37 and PDGFR-α (5 μg of each) resulted in a reduction of Xenopus receptor phosphorylation in response to PDGF (compare 1:1+ to PDGFR5+). A four-fold excess of PDGFR-37 to PDGFR-α reduced this phosphorylation below the level of detection (compare 4:1+ to PDGFR2+). NIH-3T3 cells act as a control for PDGF stimulation and detection of phosphotyrosine. (B) NIH-3T3 cells were transfected with either 20 μg of pMT2 (pMT2), 2 μg of v-sis plus 18 μg of pMT2 (v-sis) or 2 μg of v-sis plus 18 μg of PDGFR-37 (v-sis PDGFR-37). Cells were passaged into medium containing 0.5% calf serum and colony formation assessed 14 days after transfection. Note that v-sis-dependent colony formation was inhibited by co-transfection of PDGFR-37. (C) Xenopus embryos were injected in the animal pole region of both blastomeres at the 2-cell stage with RNA encoding either wild-type PDGFR-α (PDGFR-wt), PDGFR-37 or the PDGFR-α frameshift. Animal pole explants were made when embryos reached stage 8 and placed into either 75% MMR alone (untreated), or 75% MMR containing either 50 ng/ml PDGF-AA (PDGF), 500 ng/ml bFGF or 2 U/ml activin. Explants were photographed when sibling embryos had reached stage 16-17. Explants injected with the PDGFR-α frameshift control RNA did not elongate in the presence or absence of PDGF, but did elongate in the presence of activin or FGF. Explants from PDGFR-α-injected embryos elongated in the presence of activin, FGF or PDGF, but not in the absence of PDGF. Explants from PDGFR-37-injected embryos did not elongate in the absence or presence of PDGF, but elongated in the presence of activin or FGF, demonstrating that PDGFR-37 does not inhibit FGF or activin-mediated signalling.

Fig. 3.

PDGFR-37 lacks tyrosine kinase activity and acts as a dominant inhibitor. (A) Quiescent, transiently transfected COS cells were treated with PDGF (+) or control medium (−). Tyrosine phosphorylated proteins were detected following SDS-PAGE. Endogenous PDGF receptor migrates at ∼170×103Mr (left pointing arrowhead). Transfected Xenopus PDGF receptor migrates at ∼180×103Mr (right pointing arrowhead). Phosphorylated endogenous receptor was detected in PDGF-treated COS cells transfected with 10 μg of an empty expression vector (pMT2) and in non-transfected NIH-3T3 cells (NIH-3T3). COS cells transfected with 2 μg or 5 μg of Xenopus wild-type PDGFR-α (PDGFR2, 5) contained a small amount of phosphorylated receptor in untreated cells (PDGFR2−, PDGFR5−) which increased on addition of PDGF (PDGFR2+, PDGFR5+). Endogenous phosphorylated PDGF receptor was also detected in PDGF-treated COS cells transfected with 10 μg of PDGFR-37, but no phosphorylation of the transfected mutant receptor was detected (PDGFR-37). Co-transfection of an equal amount of PDGFR-37 and PDGFR-α (5 μg of each) resulted in a reduction of Xenopus receptor phosphorylation in response to PDGF (compare 1:1+ to PDGFR5+). A four-fold excess of PDGFR-37 to PDGFR-α reduced this phosphorylation below the level of detection (compare 4:1+ to PDGFR2+). NIH-3T3 cells act as a control for PDGF stimulation and detection of phosphotyrosine. (B) NIH-3T3 cells were transfected with either 20 μg of pMT2 (pMT2), 2 μg of v-sis plus 18 μg of pMT2 (v-sis) or 2 μg of v-sis plus 18 μg of PDGFR-37 (v-sis PDGFR-37). Cells were passaged into medium containing 0.5% calf serum and colony formation assessed 14 days after transfection. Note that v-sis-dependent colony formation was inhibited by co-transfection of PDGFR-37. (C) Xenopus embryos were injected in the animal pole region of both blastomeres at the 2-cell stage with RNA encoding either wild-type PDGFR-α (PDGFR-wt), PDGFR-37 or the PDGFR-α frameshift. Animal pole explants were made when embryos reached stage 8 and placed into either 75% MMR alone (untreated), or 75% MMR containing either 50 ng/ml PDGF-AA (PDGF), 500 ng/ml bFGF or 2 U/ml activin. Explants were photographed when sibling embryos had reached stage 16-17. Explants injected with the PDGFR-α frameshift control RNA did not elongate in the presence or absence of PDGF, but did elongate in the presence of activin or FGF. Explants from PDGFR-α-injected embryos elongated in the presence of activin, FGF or PDGF, but not in the absence of PDGF. Explants from PDGFR-37-injected embryos did not elongate in the absence or presence of PDGF, but elongated in the presence of activin or FGF, demonstrating that PDGFR-37 does not inhibit FGF or activin-mediated signalling.

Fig. 4.

PDGFR-37 does not block the induction of brachyury by FGF or activin. Xenopus embryos were injected in the animal pole region of both blastomeres at the 2-cell stage with RNA encoding either PDGFR-37 or the PDGFR-α frameshift. Animal pole explants were made when embryos reached stage 8 and placed into either 75% MMR alone (untreated), or 75% MMR containing either 50 ng/ml PDGF-AA (PDGF), 500 ng/ml bFGF or 2 U/ml activin. Explants were frozen when sibling embryos had reached stage 13–14, then processed to extract RNA followed by RT-PCR with primers specific for EF1-α and brachyury (Xbra). Brachyury RNA was not detected in untreated animal pole explants from PDGFR-α frameshift-injected embryos, was strongly induced by FGF or activin, but not by PDGF. Explants form PDGFR-37-injected embryos displayed an identical response to all these growth factor additions.

Fig. 4.

PDGFR-37 does not block the induction of brachyury by FGF or activin. Xenopus embryos were injected in the animal pole region of both blastomeres at the 2-cell stage with RNA encoding either PDGFR-37 or the PDGFR-α frameshift. Animal pole explants were made when embryos reached stage 8 and placed into either 75% MMR alone (untreated), or 75% MMR containing either 50 ng/ml PDGF-AA (PDGF), 500 ng/ml bFGF or 2 U/ml activin. Explants were frozen when sibling embryos had reached stage 13–14, then processed to extract RNA followed by RT-PCR with primers specific for EF1-α and brachyury (Xbra). Brachyury RNA was not detected in untreated animal pole explants from PDGFR-α frameshift-injected embryos, was strongly induced by FGF or activin, but not by PDGF. Explants form PDGFR-37-injected embryos displayed an identical response to all these growth factor additions.

PDGFR-37 causes defects in Xenopus development

Synthetic RNA encoding PDGFR-37 was injected into the lateral, equatorial region of both blastomeres of 2-cell Xenopus embryos. Development of these embryos was monitored over several days until the late tadpole stage. PDGFR-37 produced profound developmental defects in Xenopus embryos in a dose-dependent manner. At the highest doses (1-2 ng RNA per embryo), embryos did not survive beyond stage 8-9. At slightly lower doses (0.2-0.5 ng RNA), embryos began to display defects at stage 6–7 around the presumptive injection site. In these cases, the injected blastomeres were delayed in their division, so that they were larger than their neighbours and the corresponding cells of sibling control embryos. The majority of these embryos did not survive beyond stage 10. Injection of similar or greater amounts of PDGFR-α frameshift RNA (or a truncated PDGFR-α) did not produce any noticeable defects in Xenopus embryos, suggesting that these effects are specific to PDGFR-37 (data not shown).

At lower doses of PDGFR-37 RNA (50–200 pg), embryos appeared externally normal throughout the blastula and early- to mid-gastrula stages. There was no discernible difference between PDGFR-37, PDGFR-α frameshift RNA-injected, or non-injected embryos in blastomere size or in dorsal-lip formation. Defects in the development of these PDGFR-37-injected embryos became apparent by stage 11 when blastopore closure failed to progress and the deep endoderm began to evaginate (Fig. 5A,B). We have concentrated on analysing the defects produced at this lowest dose of RNA.

Fig. 5.

PDGFR-37 causes defects in Xenopus gastrulation. Xenopus embryos were injected with 50–200 pg of mRNA encoding PDGFR-37 or a PDGFR-α frameshift mutation in the lateral region of both blastomeres at the 2-cell stage. Defects became apparent between stages 11 and 12. Vegetal view of stage 12 embryos injected with PDGFR-α frameshift (A) or PDGFR-37 RNA (B). Note failure of blastopore closure in B. Appearance of embryos at stage 27 (C). The PDGFR-α frameshift-injected embryo (upper) looks normal (lateral view, anterior to left) while the PDGFR-37-injected embryo has reduced head structures and the neural folds have not closed (dorsal view, anterior to left). (D) PDGFR-α frameshift-injected embryo at stage 40. (E) PDGFR-37-injected embryos at stage 40 have poorly developed heads (anterior at left), open neural folds and shortened anterior-posterior axis. The defects caused by PDGFR-37 are rescued by wild-type PDGFR-α. Embryos were injected with 140 pg of PDGFR-α frameshift RNA which did not cause any defects (F). Embryos were also unaffected by 70 pg of PDGFR-α (G). Injection of 70 pg of PDGFR-37 caused defects in 44% of embryos (H) which were reversed by co-injection of 70 pg of PDGFR-37 plus 70 pg of PDGFR-α (I).

Fig. 5.

PDGFR-37 causes defects in Xenopus gastrulation. Xenopus embryos were injected with 50–200 pg of mRNA encoding PDGFR-37 or a PDGFR-α frameshift mutation in the lateral region of both blastomeres at the 2-cell stage. Defects became apparent between stages 11 and 12. Vegetal view of stage 12 embryos injected with PDGFR-α frameshift (A) or PDGFR-37 RNA (B). Note failure of blastopore closure in B. Appearance of embryos at stage 27 (C). The PDGFR-α frameshift-injected embryo (upper) looks normal (lateral view, anterior to left) while the PDGFR-37-injected embryo has reduced head structures and the neural folds have not closed (dorsal view, anterior to left). (D) PDGFR-α frameshift-injected embryo at stage 40. (E) PDGFR-37-injected embryos at stage 40 have poorly developed heads (anterior at left), open neural folds and shortened anterior-posterior axis. The defects caused by PDGFR-37 are rescued by wild-type PDGFR-α. Embryos were injected with 140 pg of PDGFR-α frameshift RNA which did not cause any defects (F). Embryos were also unaffected by 70 pg of PDGFR-α (G). Injection of 70 pg of PDGFR-37 caused defects in 44% of embryos (H) which were reversed by co-injection of 70 pg of PDGFR-37 plus 70 pg of PDGFR-α (I).

As development proceeded, the yolk plug continued to evaginate, resulting in a mass of endodermal tissue outside the embryo. The neural tubes of these embryos failed to fuse along the dorsal midline, presumably because of this evaginated endodermal tissue. From stage 18 onwards, many embryos displayed a ‘split tail’ phenotype, outwardly similar to that observed in embryos in which FGF signalling is disrupted (Amaya et al., 1991). In contrast to these embryos, however, the head structures of PDGFR-37-injected embryos were also abnormal (Fig. 5D,E). Heads were usually narrower than control embryos and the eyes smaller, fused or absent. In the most severe cases, no discernible head structures formed (Fig. 5E). It was apparent that embryos contained some mesoderm because somites (often poorly organised) were discernible. The heart appeared of normal shape in PDGFR-37-injected embryos, although it was often reduced in size. In very rare cases, it was absent; this phenotype was often associated with the most severely affected embryos that had no recognisable head structures.

To demonstrate that the defects that we observe on PDGFR-37 injection are specific to an inhibition of PDGF signalling, we show that these defects can be rescued by co-injection of wild-type Xenopus PDGFR-α. (Results of three experiments are summarised in Table 1 and embryos from experiment three are shown in Fig. 5F-I.)

Table 1.

Wild-type PDGFR-α rescues PDGFR-37 induced defects

Wild-type PDGFR-α rescues PDGFR-37 induced defects
Wild-type PDGFR-α rescues PDGFR-37 induced defects

Histology of late embryos

The phenotype observed in PDGFR-37-injected embryos correlates with a disruption of mesodermal structures, particularly those at the anterior of the embryo. This could be due to either a failure to form mesoderm, or alternatively, a failure of mesoderm to move anteriorwards. We examined PDGFR-37-injected embryos for the presence of mesoderm by histochemical staining at stage 33-34. PDGFR-37-injected embryos did contain mesodermal tissues including notochord and muscle (Fig. 6) although they were severely disorganised, as was the neural tube. The presence of mesoderm in these embryos is not surprising, as PDGFR-37 does not prevent mesoderm induction in animal pole explants (Figs 3C, 4), nor does PDGF itself induce mesoderm (Ruiz i Altaba and Melton, 1989, Fig. 4). Mesodermal tissues were found to be missing only from particular regions of the embryo (e.g. anterior notochord, Fig. 6B,G; dorsal somitic muscle, Fig. 6G,H) suggesting that the anterior and anteriolateral mesoderm were most severely affected. Thus, mesoderm induction was not inhibited by PDGFR-37, but the mesoderm was incorrectly organised. This spatial arrangement of ectoderm, mesoderm and endoderm is established during gastrulation. We have therefore begun to address whether PDGFR-37 may disrupt patterning or cell movements in the early Xenopus embryo.

Fig. 6.

Histological analysis of PDGFR-37-injected embryos at stage 33–34. Transverse sections of PDGFR-α frameshift (A,C) and PDGFR-37-injected (B,D) embryos through the level of the eye (A, B) and otic vesicle (C,D). Note absence of notochord (no) in B, and misshapen neural tube (nt). The notochord and neural tube are abnormal in D. Sagittal sections of PDGFR-α frameshift (E,F) and PDGFR-37 (G,H)-injected embryos. The notochord forms a straight, well organised structure in control embryos but is misshapen or missing from certain regions of PDGFR-37-injected embryos. Muscle is largely absent from the dorsal and anterior regions of PDGFR-37-injected embryos and is often poorly organised when present. Large masses of endodermal tissue (endo) are exposed on the dorsal side of PDGFR-37-injected embryos (G). All embryos are oriented with dorsal at the top and anterior to the left (E-H). In this experiment, 40 of 64 embryos appeared abnormal, 13 appeared normal and 11 were dead. Seven abnormal embryos with a recognisable anterior-posterior axis were entirely sectioned and all displayed similar phenotypes.

Fig. 6.

Histological analysis of PDGFR-37-injected embryos at stage 33–34. Transverse sections of PDGFR-α frameshift (A,C) and PDGFR-37-injected (B,D) embryos through the level of the eye (A, B) and otic vesicle (C,D). Note absence of notochord (no) in B, and misshapen neural tube (nt). The notochord and neural tube are abnormal in D. Sagittal sections of PDGFR-α frameshift (E,F) and PDGFR-37 (G,H)-injected embryos. The notochord forms a straight, well organised structure in control embryos but is misshapen or missing from certain regions of PDGFR-37-injected embryos. Muscle is largely absent from the dorsal and anterior regions of PDGFR-37-injected embryos and is often poorly organised when present. Large masses of endodermal tissue (endo) are exposed on the dorsal side of PDGFR-37-injected embryos (G). All embryos are oriented with dorsal at the top and anterior to the left (E-H). In this experiment, 40 of 64 embryos appeared abnormal, 13 appeared normal and 11 were dead. Seven abnormal embryos with a recognisable anterior-posterior axis were entirely sectioned and all displayed similar phenotypes.

PDGFR-37 does not disrupt mesodermal patterning

To examine whether PDGF is required for the expression of early mesodermal markers in gastrula-stage embryos, we examined the spatial distribution of brachyury and goosecoid transcripts in PDGFR-37-injected embryos by in situ hybridisation. goosecoid is expressed in the organiser region of the stage 10 embryo and is a marker of the anterior mesoderm (Cho et al., 1991) while brachyury is expressed throughout the mesoderm (prior to involution) at stage 11 (Smith et al., 1991b; Amaya et al., 1993). We found that the expression of both of these genes was unaffected by the injection of PDGFR-37 RNA into the lateral, equatorial region of one (brachyury) or both (goosecoid) blastomeres at the 2-cell stage, even though the anterior and lateral mesoderm is affected (Fig. 7). Thus, it appears that PDGFR-37 does not interfere with the induction of mesoderm, nor with its initial patterning. We therefore examined the effect of PDGFR-37 on cell movement in the gastrula stage Xenopus embryo.

Fig. 7.

PDGFR-37 does not affect expression of goosecoid or brachyury in whole embryos. Embryos injected with PDGFR-α frameshift (A,C) or PDGFR-37 (B,D) RNA were processed for whole-mount in situ hybridisation with cRNA probes for goosecoid at stage 10-10.25 (A,B) or brachyury at stage 11 (C,D). goosecoid expression was seen in the organiser region of control embryos, directly above the dorsal lip. The embryos in A and B are viewed from the vegetal side with the dorsal side of the embryo toward the top. Embryos in C and D have been cleared and are viewed form the vegetal side to show the circumferential expression pattern of brachyury.

Fig. 7.

PDGFR-37 does not affect expression of goosecoid or brachyury in whole embryos. Embryos injected with PDGFR-α frameshift (A,C) or PDGFR-37 (B,D) RNA were processed for whole-mount in situ hybridisation with cRNA probes for goosecoid at stage 10-10.25 (A,B) or brachyury at stage 11 (C,D). goosecoid expression was seen in the organiser region of control embryos, directly above the dorsal lip. The embryos in A and B are viewed from the vegetal side with the dorsal side of the embryo toward the top. Embryos in C and D have been cleared and are viewed form the vegetal side to show the circumferential expression pattern of brachyury.

PDGFR-37 does not inhibit convergence and extension

The major mechanical driving force for Xenopus gastrulation is provided by the convergence and extension that results from intercalatory movements of marginal zone cells (for review, see Keller, 1991). This force causes the inward movement of prospective mesodermal cells and also leads to elongation of the embryo along the anterior-posterior axis. When Xenopus embryos are placed in a hypertonic solution during gastrulation, osmotic pressure in the blastocoel prevents involution and leads to the formation of an exogastrula (Holtfreter, 1933). The marginal zone tissue that has failed to involute is still capable of undergoing convergence and extension movements, which causes extensive elongation of the exogastrula (reviewed in Keller, 1991). However, marginal zone cells do not contact ectoderm in these exogastrulae, thus allowing the study of convergent extension movements largely in the absence of marginal zone-ectoderm interactions. To test whether PDGFR-37 was inhibiting convergent extension, we injected Xenopus embryos at the 2-cell stage with RNA encoding PDGFR-α frameshift or PDGFR-37. Embryos were cultured until stage 8, when their vitelline membranes were removed manually. They were then transferred to a hypertonic salt solution and allowed to develop until sibling embryos had reached the late tadpole stage. Exogastrulation occurred to a similar extent in both control and PDGFR-37-injected embryos (data not shown and Fig. 9). We therefore conclude that the gastrulation defects caused by PDGFR-37 injection are not caused by an inhibition of convergence and extension movements of marginal zone cells.

Fig. 9.

PDGFR-37-expressing mesoderm is not affected in exogastrulae. Embryos were co-injected with β-galactosidase and either PDGFR-α frameshift (A) or PDGFR-37 (B) RNA on the presumptive dorsal side of both blastomeres at the 2-cell stage and were induced to exogastrulate. Exogastrulae were stained as whole mounts with X-gal. As before, convergent extension was unaffected in PDGFR-37-injected embryos. The ectoderm (pigmented tissue at right of figure) is separated from the endoderm (large, yolky cells at left of figure) by the convergently extended mesoderm. X-gal-stained cells extend throughout this mesoderm, with the X-gal staining concentrated in the nuclei of these cells (in contrast to X-gal-stained cells in the blastocoel in Fig. 8B).

Fig. 9.

PDGFR-37-expressing mesoderm is not affected in exogastrulae. Embryos were co-injected with β-galactosidase and either PDGFR-α frameshift (A) or PDGFR-37 (B) RNA on the presumptive dorsal side of both blastomeres at the 2-cell stage and were induced to exogastrulate. Exogastrulae were stained as whole mounts with X-gal. As before, convergent extension was unaffected in PDGFR-37-injected embryos. The ectoderm (pigmented tissue at right of figure) is separated from the endoderm (large, yolky cells at left of figure) by the convergently extended mesoderm. X-gal-stained cells extend throughout this mesoderm, with the X-gal staining concentrated in the nuclei of these cells (in contrast to X-gal-stained cells in the blastocoel in Fig. 8B).

Lineage tracing of involuting mesodermal cells

In addition to cell intercalation movements in the marginal zone, migration of mesodermal cells over the interior surface of the blastocoel is also known to be crucial for normal gastrulation (for review, see Keller, 1991). We have therefore examined the fate of mesodermal cells expressing PDGFR-37 during Xenopus gastrulation by lineage tracing with the histochemical marker, β-galactosidase containing a nuclear localisation signal. Embryos were injected with β-galactosidase RNA plus either PDGFR-37 or PDGFR-α frameshift RNA in both blastomeres on the presumptive dorsal side at the 2-cell stage. Limited diffusion of this injected RNA means that only a small proportion of the daughter cells of each blastomere are labelled. Embryos were analysed beginning at the blastula stage (stage 8) until late gastrula (stage 12). Fixed embryos were stained with a chromogenic substrate (X-gal) and examined as whole mounts. Prior to the onset of gastrulation, β-galactosidase activity was localised to the nuclei of cells in the marginal zone and animal cap of the dorsal side of the embryo. Staining was identical in both control and PDGFR-37-injected embryos until after the onset of gastrulation when large, X-gal-stained cells were observed loosely atached to the floor of the blastocoel in PDGFR-37-injected embryos (data not shown).

We examined these X-gal-stained embryos in more detail by sectioning at various times during development. As with embryos viewed as whole mounts, X-gal-stained cells were present in the ectoderm of the animal cap and in the marginal zone, but not in the blastocoel, prior to the onset of gastrulation (data not shown). As gastrulation proceeded, X-gal-stained cells were seen migrating with the head mesoderm and throughout the involuting marginal zone (IMZ) of control embryos (Fig. 8A). In PDGFR-37-injected embryos, the extent of mesoderm migration over the blastocoel roof was delayed compared to control embryos and few X-gal-stained cells were present in the anterior mesoderm. However, X-gal-stained cells were present throughout marginal zone tissue which had not involuted (Fig. 8B). In addition, X-galstained cells were found on the floor of the blastocoel cavity. X-gal staining was nuclear in all cells of control embryos and in mesodermal cells that had not involuted in PDGFR-37-injected embryos (compare figure 8A to 8B, and within 8B the cells extending throughout the mesoderm prior to involution on the dorsal side of the embryo to those in the blastocoel). However, X-gal staining extended throughout the cytoplasm of cells from PDGFR-37-injected embryos after involution, suggesting that these cells were dead or dying.

Fig. 8.

PDGFR-37 disrupts the migration of involuting mesoderm. RNA encoding β-galactosidase (containing a nuclear localisation signal) was co-injected into the equatorial region of the presumptive dorsal side of both blastomeres of the 2-cell embryo with either PDGFR-α frameshift (A) or PDGFR-37 RNA (B). Embryos are positioned with the animal pole at the top and the dorsal side to the right. Extensive involution and migration of X-gal-stained dorsal mesoderm cells occurred in PDGFR-α frameshift-injected embryos (A). PDGFR-37-injected embryos had X-gal-stained cells in the blastocoel and in the marginal zone prior to involution (B). Few X-gal-stained cells were present in the anterior mesoderm. Note also, the poor demarcation of involuting mesoderm in PDGFR-37-injected embryos and the failure of the mesoderm to extend across the blastocoel roof on the dorsal, but not the ventral, side of these embryos.

Fig. 8.

PDGFR-37 disrupts the migration of involuting mesoderm. RNA encoding β-galactosidase (containing a nuclear localisation signal) was co-injected into the equatorial region of the presumptive dorsal side of both blastomeres of the 2-cell embryo with either PDGFR-α frameshift (A) or PDGFR-37 RNA (B). Embryos are positioned with the animal pole at the top and the dorsal side to the right. Extensive involution and migration of X-gal-stained dorsal mesoderm cells occurred in PDGFR-α frameshift-injected embryos (A). PDGFR-37-injected embryos had X-gal-stained cells in the blastocoel and in the marginal zone prior to involution (B). Few X-gal-stained cells were present in the anterior mesoderm. Note also, the poor demarcation of involuting mesoderm in PDGFR-37-injected embryos and the failure of the mesoderm to extend across the blastocoel roof on the dorsal, but not the ventral, side of these embryos.

The loose cells within the blastocoel do not die as a result of toxicity from injection of RNA (as controls are normal) but rather, this effect is specific to PDGFR-37 injection and then, only to the subset of mesodermal cells that have involuted. We therefore examined the fate of these cells in exogastrulae, by co-injecting β-galactosidase RNA plus either PDGFR-37 or PDGFR-α frameshift RNA in both blastomeres on the pre-sumptive dorsal side at the 2-cell stage. We found that, in contrast to embryos grown in 10% MMR, PDGFR-37-injected cells in these exogastrulae were able to contribute to the mesoderm of these embryos and appeared healthy, with nuclear-localised X-gal staining (Fig. 9). Furthermore, dissection of these embryos did not reveal any loose X-gal-stained cells. The relevance of these findings to the role of PDGF-A during Xenopus gastrulation is discussed below.

The Xenopus embryo provides an excellent model for the study of gastrulation. Much has been learned at the cellular level (for review, see Keller, 1991) but considerably less is known about the molecular mechanisms that govern cell behaviour during this process. Morphogenetic movements during gastrulation bring mesodermal cells of the IMZ into apposition with the ectoderm as the latter population moves towards the anterior of the embryo. Subsequent development depends on the precise movement of these cells. Thus, it seems likely that this process is tightly regulated, possibly by signals emanating from the ectoderm. The presence of PDGF-A and PDGFR-α in the ectoderm and mesoderm, respectively, of the gastrula stage Xenopus embryo suggests that PDGF signalling may contribute to the regulation of this process.

In order to investigate the requirement for PDGF-A signalling during gastrulation, we have generated a dominant negative mutant of PDGFR-α that can inhibit signalling by the wild-type receptor. This approach has proven effective in inhibiting FGF and activin signalling in Xenopus embryos using receptors lacking the intracellular catalytic domain (Amaya et al., 1991; Hemmati-Brivanlou and Melton, 1992). However, a similar truncation of PDGFR-α generated in our laboratory does not produce any defects in Xenopus development when injected into the 2-cell embryo (unpublished results). We suspect that the inability of this truncated PDGFR-α to perturb Xenopus development is due to a combination of two factors: first, the relatively low efficiency of truncation mutants – a 50- to 100-fold excess of mutant to wild-type receptor is required to abolish signalling of FGF, activin and PDGF receptors (Amaya et al., 1991; Ueno et al., 1991; Hemmati-Brivanlou and Melton, 1992; Ueno et al., 1993); second, that PDGFR-α is more abundant than either the FGF or activin receptors in the early Xenopus embryo.

We have generated a potent inhibitor of PDGF receptor activity (PDGFR-37) by mutating a conserved glutamate residue to a lysine. Analogous point mutations in other PDGF receptor family members abolish detectable kinase activity (Nocka et al., 1990; Reith et al., 1990, 1991, 1993) and produce biological effects when expressed in a 1:1 ratio with wild-type receptor (Nocka et al., 1990; Reith et al., 1990). We find that PDGFR-37 transiently transfected into COS cells is not detectably phosphorylated on tyrosine residues in response to PDGF, while the phosphorylation of wild-type PDGFR-α is readily detected. Furthermore, a four-fold excess of PDGFR-37 DNA prevents Xenopus PDGFR-α phosphorylation in co-transfection experiments. PDGFR-37 is also able to block the v-sis-dependent transformation of NIH 3T3 cells. These cells express both PDGFR-α and PDGFR-β, both of which bind PDGF-B (reviewed in Heldin, 1992), and are transformed by overexpression of growth factor. Thus, the latter experiment demonstrates that Xenopus PDGFR-37 is able to inhibit both mouse PDGFR-α and PDGFR-β. Finally, the most compelling evidence that PDGFR-37 acts as an inhibitory molecule is the finding that its effect on Xenopus development can be antagonised by the co-expression of wild-type PDGFR-α.

Additional evidence that PDGFR-37 lacks kinase activity is provided by experiments from animal pole explants. Such explants are usually unable to form mesoderm in response to PDGF (Ruiz i Altaba and Melton, 1989). However, we show that animal pole explants from PDGFR-α-injected embryos undergo elongation (Fig. 3C) and express brachyury (data not shown) when PDGF-A is added to the medium, suggesting that mesoderm induction in Xenopus ectoderm does not require specific growth factors but can be mediated by a number of receptor kinases that activate ras (Whitman and Melton, 1992). This conclusion is supported by the finding that mesoderm is induced in animal pole explants expressing transforming growth factor-β (TGF-β) type II receptor which have been treated with TGF-β1 (Bhushan et al., 1994). TGF-β1 is normally unable to induce mesoderm in this assay. In contrast to PDGFR-α, PDGFR-37 does not cause the elongation or the induction of brachyury in animal pole explants However, neither activin- or FGF-induced elongation nor brachyury induction is prevented by PDGFR-37, demonstrating that this block is specific to PDGF-mediated signalling.

Injection of PDGFR-37 into Xenopus embryos produces severe developmental defects, which we have demonstrated are specifically caused by inhibition of PDGF-mediated signalling. Although PDGFR-37 is capable of interacting with both PDGFR-α and PDGFR-β, it is unlikely that the observed phenotype is due to a perturbation of PDGFR-β signalling in the embryo. PDGFR-β transcripts have not been detected in gastrula stage Xenopus embryos (M. M., unpublished results) while we have shown that PDGFR-α is expressed in the affected tissue. In the mouse, the expression patterns of PDGFR-α and PDGFR-β do not overlap (Soriano, 1994) and PDGFR-α is expressed in the primitive streak mesoderm (Orr-Urtreger and Lonai, 1992; Palmieri et al., 1992) which is analogous to the involuting marginal zone of Xenopus. Furthermore, mice with null mutations for either PDGF-B or PDGFR-β gastrulate normally and display defects associated with abnormal kidney and cardiovascular development (Levéen et al., 1994; Soriano, 1994). Taken together, these findings strongly suggest that PDGFR-37 induced defects can be ascribed to an inhibition of PDGFR-α signalling in vivo.

The defects seen in PDGFR-37-injected embryos are characterised by a loss of mesodermal structures, particularly in anterior and dorsal tissues. We have shown that such embryos make mesoderm, but this mesoderm is not properly organised in dorsal regions (Fig. 6). PDGFR-37 injection also produces embryos with narrow heads, which is probably due to a deficiency of head mesoderm. Eyes in these narrow-headed embryos are often joined in the middle, or fused into a single eye. The defects observed in ectodermal derivatives such as the eyes and the neural tube are likely to arise as a secondary effect of mesodermal malformation. The notochord in particular is known to be essential for correct patterning and differentiation of neural tissue in vertebrates (reviewed in Harland, 1994) and is most notably abnormal or absent from anterior regions of PDGFR-37-injected embryos.

Defects caused by lateral injection of PDGFR-37 are more pronounced on the dorsal side of the embryo even though PDGFR-α is distributed evenly throughout the marginal zone. This may be because cells of the dorsal marginal zone have further to migrate than cells from the ventral marginal zone (reviewed in Keller, 1991), thus accentuating defects in cell movement. We did not observe any difference in the expression pattern of goosecoid (which marks the dorsal, anterior mesoderm in the early gastrula) in PDGFR-37-injected embryos, nor in the expression pattern of the general mesodermal marker, brachyury. These findings, together with the presence of mesodermal cell types in the embryos (Fig. 6) and the inability of PDGFR-37 to block elongation or brachyury expression in response to activin or FGF in animal caps (Fig. 4), suggest that PDGFR-37 does not interfere with mesodermal induction or patterning. We cannot rule out the possibility, however, that PDGF signalling lies downstream of goosecoid or brachyury, or in a separate pathway. Similarly, PDGFR-37 may affect the expression pattern of other early mesodermal markers that we have not examined.

The inability of the dorsal midline to fuse appears to be due to physical obstruction by the endoderm, a consequence of incomplete gastrulation. Embryos were able to initiate gastrulation but they did not progress to a closed blastopore. At the same time, PDGFR-37-injected embryos were able to undergo convergent extension movements when induced to exogastrulate (as were PDGFR-37-injected animal pole explants treated with activin). These results indicate that PDGF does not control this process.

During gastrulation, the first cells to involute are those that constitute the head mesoderm. They do not move as a result of convergent extension of the marginal zone, but rather, migrate over the surface of the blastocoel ahead of the mesodermal mantle (Keller, 1986; Keller and Winklbauer, 1992; Winklbauer, 1994). The head mesoderm is often severely affected in PDGFR-37-injected embryos, even when the RNA is injected laterally. It was not surprising therefore, to find abnormalities in the migration of the head mesoderm at the earliest times following the onset of gastrulation when RNA was injected on the future dorsal side of the embryo. When compared to controls, the anterior-most mesoderm of PDGFR-37-injected embryos did not migrate as a cohesive sheet of cells, nor was it flattened and in close apposition to the inner surface of the blastocoel roof. β-galactosidase co-injected as a lineage label was localised to the nucleus of cells in the ectoderm and noninvoluting marginal zone (NIMZ), but appeared more diffuse and cytoplasmic in cells that had involuted, suggesting that these cells were dead or dying. It is intriguing that PDGFR-37-injected mesodermal cells in exogastrulae do not die, but instead contribute to mesodermal structures. This suggests to us that cells in the IMZ become PDGF dependent only after they have involuted.

It is not yet clear when or why involuting mesodermal cells die in the absence of PDGF signalling. The cause is not general toxicity due to RNA injection, as cells in the ectoderm and pre-involuted mesoderm expressing β-galactosidase appear healthy. PDGF elicits a wide variety of responses in cells in vitro, so many possibilities exist as to its function during Xenopus gastrulation. PDGF was first identified as a mitogen (Kohler and Lipton, 1974; Ross et al., 1974) but has since been shown to modulate differentiation and migration (Noble et al., 1988), to act as a survival factor (Barres et al., 1993) and to regulate chemotaxis (Grotendorst et al., 1982; Kundra et al., 1994). To distinguish between these possibilities will require further analysis of PDGFR-37-expressing cells both in vivo and in vitro. The ability to analyse the spreading, adhesion and migration of Xenopus marginal zone cells in culture (Symes et al., 1994) will facilitate a better understanding of the role that PDGF-A plays during gastrulation.

It has been suggested that PDGF also plays an important role in the gastrulation of other vertebrates. In the mouse, the patterns of PDGF-A and PDGFR-α expression are similar to those in the gastrula stage Xenopus embryo. PDGF-A is expressed in the embryonic ectoderm, while PDGFR-α is present in the adjacent layer of mesoderm (Orr-Urtreger and Lonai, 1992; Palmieri et al., 1992). Mice homozygous for Patch (Ph) lack the gene for PDGFR-α (Smith et al., 1991a; Stephenson et al., 1991) and die in utero at various stages of gestation including at the period around gastrulation (Grüneberg and Truslove, 1960). Ph/Ph mice that survive beyond this point display specific defects in heart and cranial development. These features have been ascribed to abnormalities in the neural crest, a prominent site of PDGFR-α expression in older embryos (Morrison-Graham et al., 1992; Orr-Urtreger et al., 1992; Schatteman et al., 1992). PDGFR-α is found in many mesenchymal tissues at various stages of development while PDGF-A is often present in adjacent ectodermal or endodermal layers (Orr-Urtreger and Lonai, 1992; Palmieri et al., 1992; Ho et al., 1994). We therefore propose that PDGF-A acts to promote the migration of mesenchymal cells throughout vertebrate development.

We thank our colleagues Mike Dunn, Nanette Nascone and Brenda Williams for helpful comments on the manuscript and Dr Lorraine Gudas (Cornell University Medical College) for helpful discussions and support. We thank Carole LaBonne, Kevin Lustig and Pascal Stein for their generous gifts of reagents. This work was supported by a BBSRC postdoctoral fellowship to P. A., a Cancer Research Institute/F. M. Kirby foundation fellowship to M. M. C., NIH grant 1HF32HD07561 to L. H. and grants from the NIH (RO1 HD28460), March of Dimes Birth Defects Foundation and American Heart Association (Grant-in-Aid, 94014850) to M. M.

Amaya
,
E.
,
Musci
,
T. J.
and
Kirschner
,
M.
(
1991
).
Expression of a dominant negative mutant of the FGF receptor disrupts mesoderm formation in Xenopus embryos
.
Cell
66
,
257
270
.
Amaya
,
E.
,
Stein
,
P. A.
,
Musci
,
T. J.
and
Kirschner
,
M. W.
(
1993
).
FGF signalling in the early specification of mesoderm in Xenopus
.
Development
118
,
477
487
.
Barres
,
B. A.
,
Schmid
,
R.
,
Sendnter
,
M.
and
Raff
,
M. C.
(
1993
).
Multiple extracellular signals are required for long-term oligodendrocyte survival
.
Development
118
,
283
295
.
Bhushan
,
A.
,
Lin
,
H. Y.
,
Lodish
,
H. F.
and
Kintner
,
C. R.
(
1994
).
The transforming growth factor β type II receptor can replace the activin type II receptor in inducing mesoderm
.
Mol. Cell. Biol
.
14
,
4280
4285
.
Chen
,
R.-H.
and
Blenis
,
J.
(
1990
).
Identification of Xenopus S6 protein kinase homologs (pp90rsk) in somatic cells: phosphorylation and activation during initiation of cell proliferation
.
Mol. Cell. Biol
.
10
,
3204
3215
.
Cho
,
K. W. Y.
,
Blumberg
,
B.
,
Steinbeisser
,
H.
and
De Robertis
,
E. M.
(
1991
).
Molecular nature of Spemann’s organizer: the role of the Xenopus homeobox gene goosecoid
.
Cell
67
,
1111
1120
.
Cooke
,
J.
(
1979
).
Cell number in relation to primary pattern formation in the embryo of Xenopus laevis. I: The cell cycle during new pattern formation in response to implanted organisers
.
J. Embryol. Exp. Morph
.
51
,
165
182
.
Grotendorst
,
G. R.
,
Chang
,
T.
,
Seppä
,
H. E. J.
,
Kleinman
,
H. K.
and
Martin
,
G. R.
(
1982
).
Platelet-derived growth factor is a chemoattractant for vascular smooth muscle cells
.
J. Cell. Physiol
.
113
,
261
266
.
Grüneberg
,
H.
and
Truslove
,
G. M.
(
1960
).
Two closely linked genes in the mouse
.
Genet. Res., Camb
.
1
,
69
90
.
Harland
,
R. M.
(
1994
).
Neural induction in Xenopus
.
Curr. Opin. Gen. Dev
.
4
,
543
549
.
Heldin
,
C.-H.
(
1992
).
Structural and functional studies on platelet-derived growth factor
.
EMBO J
.
11
,
4251
4259
.
Hemmati-Brivanlou
,
A.
and
Melton
,
D. A.
(
1992
).
A truncated activin receptor inhibits mesoderm induction and formation of axial structures in Xenopus embryos
.
Nature
359
,
609
614
.
Ho
,
L.
,
Symes
,
K.
,
Yordán
,
C.
,
Gudas
,
L. J.
and
Mercola
,
M.
(
1994
).
Localization of PDGF A and PDGFRa mRNA in Xenopus embryos suggests signalling from neural ectoderm and pharyngeal endoderm to neural crest cells
.
Mechanisms of Development
48
,
165
174
.
Holtfreter
,
J.
(
1933
).
Die totale exogastrulation eine selbstablosung ektoderm von entomesoderm
.
Arch. Entwigslungmech. Org
.
129
,
669
793
.
Jones
,
S. D.
,
Ho
,
L.
,
Smith
,
J. C.
,
Yordan
,
C.
,
Stiles
,
C. D.
and
Mercola
,
M.
(
1993
).
The Xenopus platelet-derived growth factor α receptor: cDNA cloning and demonstration that mesoderm induction establishes the lineage-specific pattern of ligand and receptor expression
.
Developmental Genetics
14
,
185
193
.
Keller
,
R.
(
1986
).
The cellular basis of amphibian gastrulation
.
In The Cellular Basis of Morphogenesis
(ed.
L.W.
Browder
) pp
241
387
.
Plenum Press
,
New York
.
Keller
,
R.
(
1991
).
Early embryonic development of Xenopus laevis
.
In Xenopus laevis: Practical uses in cell and molecular biology
(ed.
B.K.
Kay
) pp
61
113
.
Academic Press
,
San Diego
.
Keller
,
R.
and
Winklbauer
,
R.
(
1992
).
The cellular basis of amphibian gastrulation
.
Curr. Top. Dev. Biol
.
27
,
39
89
.
Kohler
,
N.
and
Lipton
,
A.
(
1974
).
Platelets as a source of fibroblast growth-promoting activity
.
Exp. Cell Res
.
87
,
297
301
.
Kundra
,
V.
,
Escobedo
,
J. A.
,
Kazlauskas
,
A.
,
Kim
,
H. K.
,
Rhee
,
S. G.
,
Williams
,
L. T.
and
Zetter
,
B. R.
(
1994
).
Regulation of chemotaxis by the platelet-derived growth factor receptor-β
.
Nature
367
,
474
476
.
Levéen
,
P.
,
Pekny
,
M.
,
Gebre-Medhin
,
S.
,
Swolin
,
B.
,
Larsson
,
E.
and
Betsholtz
,
C.
(
1994
).
Mice deficient for PDGF B show renal, cardiovascular and hematological abnormalities
.
Genes & Dev
.
8
,
1875
1887
.
Mercola
,
M.
,
Melton
,
D. A.
and
Stiles
,
C. D.
(
1988
).
Platelet-derived growth factor A chain is maternally encoded in Xenopus embryos
.
Science
241
,
1223
1225
.
Mikawa
,
T.
,
Fischman
,
D.
,
Dougherty
,
J. P.
and
Brown
,
A. M. C.
(
1991
).
In vivo analysis of a new lacz retrovirus vector suitable for cell lineage marking in avian and other species
.
Exp. Cell Res
.
195
,
516
523
.
Morrison-Graham
,
K.
,
Schatteman
,
G. C.
,
Bork
,
T.
,
Bowen-Pope
,
D. F.
and
Weston
,
J. A.
(
1992
).
A PDGF receptor mutation in the mouse (Patch) perturbs the development of a non-neuronal subset of neural crest-derived cells
.
Development
115
,
133
142
.
Nieuwkoop
,
P. D.
and
Faber
,
J.
(
1967
).
Normal table of Xenopus laevis (Daudin)
. 2nd ed.
North-Holland Publishing Company
,
Amsterdam
.
Noble
,
M.
,
Murray
,
K.
,
Stroobant
,
P.
,
Waterfield
,
M. D.
and
Riddle
,
P.
(
1988
).
Platelet-derived growth factor promotes division and motility and inhibits premature differentiation of the oligodendrocyte/type-2 astrocyte progenitor cell
.
Nature
333
,
560
562
.
Nocka
,
K.
,
Tan
,
J. C.
,
Chiu
,
E.
,
Chu
,
T. Y.
,
Ray
,
P.
,
Traktman
,
P.
and
Besmer
,
P.
(
1990
).
Molecular bases of dominant negative and loss of function mutations at the murine c-kit/white spotting locus: W37, Wv, W41 and W
.
EMBO J
.
9
,
1805
1813
.
O’Keefe
,
H. P.
,
Melton
,
D. A.
,
Ferreiro
,
B.
and
Kintner
,
C.
(
1991
).
In situ hybridization
.
In Xenopus laevis: Practical uses in cell and molecular biology
(ed.
B.K.
Kay
and
H.B.
Peng
) pp
444
463
.
Academic Press
,
San Diego
.
Orr-Urtreger
,
A.
,
Bedford
,
M. T.
,
Do
,
M.-S.
,
Eisenbach
,
L.
and
Lonai
,
P.
(
1992
).
Developmental expression of the α receptor for platelet-derived growth factor, which is deleted in the embryonic lethal Patch mutation
.
Development
115
,
289
303
.
Orr-Urtreger
,
A.
and
Lonai
,
P.
(
1992
).
Platelet-derived growth factor-A and its receptor are expressed in separate, but adjacent cell layers of the mouse embryo
.
Development
115
,
1045
1058
.
Palmieri
,
S.
,
Payne
,
J.
,
Stiles
,
C. D.
,
Biggers
,
J.
and
Mercola
,
M.
(
1992
).
Expression of mouse PDGF-A and PDGF α receptor genes during pre- and post-implantation development: Evidence for a developmental shift from an autocrine to a paracrine mode of action
.
Mechanisms of Development
39
,
181
191
.
Peng
,
H. B.
(
1991
).
Solutions and Protocols
.
In Xenopus laevis: Practical uses in cell and molecular biology
(ed.
B.K.
Kay
and
H.B.
Peng
) pp
657
662
.
Academic Press
,
San Diego
.
Reith
,
A. D.
,
Ellis
,
C.
,
Lyman
,
S. D.
,
Anderson
,
D. M.
,
Williams
,
D. E.
,
Bernstein
,
A.
and
Pawson
,
T.
(
1991
).
Signal transduction by normal isoforms and W mutant variants of the kit receptor tyrosine kinase
.
EMBO J
.
10
,
2451
2459
.
Reith
,
A. D.
,
Ellis
,
C.
,
Maroc
,
N.
,
Pawson
,
T.
,
Bernstein
,
A.
and
Dubreuil
,
P.
(
1993
).
‘W’ mutant forms of the Fms receptor tyrosine kinase act in a dominant manner to suppress CSF-1 dependent cellular transformation
.
Oncogene
8
,
45
53
.
Reith
,
A. D.
,
Rottapel
,
R.
,
Giddens
,
E.
,
Brady
,
C.
,
Forrester
,
L.
and
Bernstein
,
A.
(
1990
).
W mutant mice with mild or severe developmental defects contain distinct point mutations in the kinase domain of the c-kit receptor
.
Genes Dev
.
4
,
390
400
.
Ross
,
R.
,
Glomset
,
J. A.
,
Kariya
,
B.
and
Harker
,
L.
(
1974
).
A platelet-dependent serum factor that stimulates the proliferation of arterial smooth muscle cells in vitro
.
Proc. Natl. Acad. Sci. USA
71
,
1207
1210
.
Ruiz i Altaba
,
A.
and
Melton
,
D. A.
(
1989
).
Interaction between peptide growth factors and homeobox genes in the establishment of anteroposterior polarity in frog embryos
.
Nature
341
,
33
38
.
Schatteman
,
G. C.
,
Morrison-Graham
,
K.
,
Van Koppen
,
A.
,
Weston
,
J. A.
and
Bowen-Pope
,
D. F.
(
1992
).
Regulation and role of PDGF receptor α-subunit expression during embryogenesis
.
Development
115
,
123
131
.
Smith
,
E. A.
,
Seldin
,
M. F.
,
Martinez
,
L.
,
Watson
,
M. L.
,
Choudhury
,
G. G.
,
Lalley
,
P. A.
,
Pierce
,
J.
,
Aaronson
,
S.
,
Barker
,
J.
,
Naylor
,
S. L.
and
Sakaguchi
,
A. Y.
(
1991a
).
Mouse platelet-derived growth factor receptor α gene is deleted in W19H and patch mutations on chromosome 5
.
Proc. Natl. Acad. Sci. USA
88
,
4811
4815
.
Smith
,
J. C.
,
Price
,
B. M. J.
,
Green
,
J. B. A.
,
Weigel
,
D.
and
Herrmann
,
B. G.
(
1991b
).
Expression of a Xenopus homolog of Brachyury (T) is an immediate-early response to mesoderm induction
.
Cell
67
,
79
87
.
Soriano
,
P.
(
1994
).
Abnormal kidney development and hematological disorders in PDGF β-receptor mutant mice
.
Genes & Dev
.
8
,
1888
1896
.
Stephenson
,
D. A.
,
Mercola
,
M.
,
Anderson
,
E.
,
Wang
,
C.
,
Stiles
,
C. D.
,
Bowen-Pope
,
D. F.
and
Chapman
,
V. M.
(
1991
).
Platelet-derived growth factor receptor α-subunit (Pdgfra) is deleted in the mouse patch (Ph) mutation
.
Proc. Natl. Acad. Sci. USA
88
,
6
10
.
Symes
,
K.
and
Smith
,
J. C.
(
1987
).
Gastrulation movements provide an early marker of mesoderm induction in Xenopus laevis
.
Development
101
,
339
349
.
Symes
,
K.
,
Yordán
,
C.
and
Mercola
,
M.
(
1994
).
Morphological differences in Xenopus embryonic mesodermal cells are specified as an early response to distinct threshold concentrations of activin
.
Development
120
,
2339
2346
.
Ueno
,
H.
,
Colbert
,
H.
,
Escobedo
,
J. A.
and
Williams
,
L. T.
(
1991
).
Inhibition of PDGF β receptor signal transduction by coexpression of a truncated receptor
.
Science
252
,
844
848
.
Ueno
,
H.
,
Escobedo
,
J. A.
and
Williams
,
L. T.
(
1993
).
Dominant-negative mutations of platelet-derived growth factor (PDGF) receptors: Inhibition of receptor function by ligand-dependent formation of heterodimers between PDGF α- and β-receptors
.
J. Biol. Chem
.
268
,
22814
22819
.
Vize
,
P. D.
and
Melton
,
D. A.
(
1991
).
Assays for gene function in developing Xenopus embryos
.
In Xenopus laevis: Practical Uses in Cell and Molecular Biology
(ed.
B.K.
Kay
and
H.B.
Peng
) pp
367
387
.
Academic Press, Inc
.,
San Diego
.
Whitman
,
M.
and
Melton
,
D. A.
(
1992
).
Involvement of p21ras in Xenopus mesoderm induction
.
Nature
357
,
252
254
.
Wilson
,
P. A.
and
Melton
,
D. A.
(
1994
).
Mesodermal patterning by an inducer gradient depends on secondary cell-cell communication
.
Curr. Biol
.
4
,
676
686
.
Winklbauer
,
R.
(
1994
).
Mesoderm cell migration in the vertebrate gastrula
.
Seminars in Developmental Biology
5
,
91
99
.