Regulation of isthmic Fgf8 signal by sprouty2

Fgf8 is expressed in the isthmus and acts as an organizer for the mesencephalon and metencephalon (Crossley and Martin, 1995; Crossley et al., 1996; Martinez et al.,1999; Nakamura,2001; Sato et al.,2001). Sato et al. (Sato et al., 2001) showed that Fgf8a and Fgf8b are expressed in the isthmus, and that Fgf8b could change the fate of the presumptive mesencephalon to that of the metencephalon; that is, the cerebellum differentiated in place of the tectum after Fgf8bmisexpression by electroporation. Very recent work showed that the Fgf8b signal is transduced by the Ras-ERK signaling pathway to organize cerebellar differentiation. If the Ras-ERK signaling pathway is disrupted by misexpression of a dominant negative form of Ras, the presumptive metencephalon changes its property to that of the mesencephalon(Sato and Nakamura, 2004).

sprouty2 is expressed in the isthmus(Chambers and Mason, 2000; Chambers et al., 2000; Zhang et al., 2001; Lin et al., 2002; Liu et al., 2003). Although sprouty is expressed overlapping Fgf8, and could be induced by an Fgf signal, it is suggested that sprouty2 functions as a negative regulator of the Fgf-Ras-ERK signaling pathway(Hacohen et al., 1998; Casci et al., 1999; Kramer et al., 1999; Lin et al., 2002). The negative feedback of the Fgf8 signaling pathway by Sprouty2 is controversial,but is very interesting from the view of Fgf8 signal transduction and regulation.

We carried out misexpression of Fgf8b, sprouty2 and a dominant negative form of sprouty2 (sprouty2-DN). Our results indicate that sprouty2 expression is induced very rapidly by Fgf8b, and that Sprouty2 interferes with ERK activation. We have also shown that sprouty2 misexpression resulted in the fate change of the presumptive metencephalon to the mesencephalon. Misexpression of dominant negative form of Sprouty2 resulted in an anterior shift of the posterior border of the tectum. The results indicate that Fgf8 activates the Ras-ERK signaling pathway to differentiate the cerebellum, and that the hyper- or hypo-signaling of this pathway affects the fate of the brain vesicles. Thus,Sprouty2 may regulate the Fgf8-Ras-ERK signaling pathway for the proper regionalization of the metencephalon and mesencephalon.

The full length sprouty2 cDNA was isolated from E2 chick by RT-PCR, and inserted in pBluescript (pBluescript-Sprouty2). Primers for N- and C-terminal fragments are 5′-GATGTGTTCTAAGCCTGCTGG-3′ and 5′-AGTGCCAAGACCATAGCTGC-3′, respectively. sprouty2-DN was created by substituting alanine for tyrosine 55 of sprouty2(Sasaki et al., 2001). For this, QuikChange Site-Directed Mutagenesis Kit (STRATAGENE) was used. PfuTurbo DNA polymerase^II in the kit replicates both forward and reverse strands of pBluescript-Sprouty2 with primers that contain mutation(5′-GCAACACGAATGAGGCCACAGAGGGACCGACG-3′ and 5′-CGTCGGTCCCTCTGTGGCCTCATTCGTGTT-3′, underlining indicates mutation). sprouty2 and HA-tagged sprouty2-DN(Sprouty2-DN) were inserted into pMiwIII, which has Rous sarcoma virus enhancer and chicken β-actin promoter(Suemori et al., 1990; Wakamatsu et al., 1997;Mastunaga et al., 2001).

Fgf8a and Fgf8b expression vectors were prepared by Sato(Sato et al., 2001). For transfection, in ovo electroporation was carried out at HH 8-9 (stage 8-9)(for details, see Hamburger and Hamilton,1951) as described previously(Funahashi et al., 1999; Nakamura and Funahashi, 2001). GFP expression vector was co-electroporated to check the efficiency of transfection.

Fluorescein-labeled morpholino antisense oligonucleotide against Sprouty2 was applied by electroporation as described previously(Sheng et al., 2003; Sugiyama and Nakamura,2003).

Embryos were fixed in 4% paraformaldehyde in PBS (phosphate buffered saline), and embedded in Technovit (Kulter). Serial 5 μm sections were stained with Hematoxylin and Eosin.

In situ hybridization was carried out according to Wilkinson(Wilkinson, 1992). Probes for Otx2, Gbx2, Fgf8, Wnt1, Lmx1b and mouse Fgf8 are described in Katahira et al. (Katahira et al., 1999), Matsunaga et al.(Matsunaga et al., 2002) and Funahashi et al. (Funahashi et al.,1999). The template for sprouty2 probe, consisting of 129 bp of 5′UTR and 467 bp of 5′ coding region, was subcloned into pBluescript. Digoxigenine (DIG)- or fluorescein isothiocyanate (FITC)-labeled RNA probes were transcribed by T3 or T7 RNA polymerase according to the manufacturer's protocol. Alkaline phosphatase (ALP)-conjugated anti-FITC or anti-DIG antibodies were used for detection, and were colored by Fast Red/Naphtol AS/MX, and nitroblue tetrazolium chloride (NBT) and 5-bromo-4-chloro-3-indolyl-phosphate (BCIP), respectively. For some cases,Fast Red/Naphtol AS/MX was washed away in ethanol.

For immunohistochemistry, anti-HA rabbit polyclonal antibody (Berkeley Antibody Company), anti-neurofilament monoclonal antibody, 3A10 (DSHB), and anti-diphosphorylated ERK antibody (Sigma) were used as primary antibodies. As secondary antibodies, horseradish peroxidase (HRP)-conjugated anti-mouse IgG(Jackson), Cy3-conjugated anti-mouse IgG (Jackson), and biotinylated anti-mouse IgG antibody (VECTOR) were used. HRP was detected with 3,3′-diaminobenzidine (DAB). For detection of biotinylated antibody, the ABC-Elite system (VECTOR) was adopted.

First, we examined the normal expression pattern of Fgf8 and sprouty2. As reported previously, Fgf8 was expressed in the rostral tip of the forebrain and the isthmus at HH 11. sprouty2 was expressed in an overlapping manner to Fgf8 throughout the stages examined (Fig. 1). The expression pattern indicates an intimate relationship between Fgf8and sprouty2. Indeed, it has been reported that the Fgf8 bead could induce sprouty2 expression(Minowada et al., 1999; Chambers et al., 2000).

It has been reported that sprouty2 is induced by Fgf8, but that Sprouty2 functions as a negative regulator of Fgf signaling(Hacohen et al., 1998; Kramer et al., 1999; Casci et al., 1999; Lin et al., 2002). We wanted to see if sprouty2 is really induced by Fgf8, and functions as a negative regulator of the Ras-ERK signaling pathway. It was surprising that sprouty2 was already induced by 3 hours after electroporation of both Fgf8a and Fgf8b expression vector (HH 10). At 3 hours after electroporation of Fgf8a and Fgf8b expression vector(HH 9), Fgf8 misexpression could be seen in large areas of the mesencephalon, metencephalon and myelencephalon(Fig. 2A,C,I,K). sprouty2 was already induced in an overlapping manner to Fgf8a and Fgf8b expression(Fig. 2B,D,J,L; Fgf8a; n=2/3, Fgf8b; n=7/9). Induction of sprouty2 by Fgf8 is very rapid as suggested by Chambers et al. (Chambers et al., 2000),and occurs around 1 hour if we consider that translation product is expressed by 2 hours after electroporation(Funahashi et al., 1999).

At 24 hours after electroporation with Fgf8a expression vector at a concentration of 1 μg/μl (HH 18), Fgf8 misexpression was well discerned. sprouty2 expression was induced in Fgf8 misexpression sites at the lateral side of the mesencephalon(Fig. 2E-H; n=7/8).

At 24 hours after electroporation with Fgf8b expression vector at a concentration of 1 μg/μl (HH 18), misexpression of Fgf8b was widely discernible from the metencephalon to the diencephalon. In addition to the misexpression at the lateral side of the neural tube, as was seen after Fgf8a misexpression, strong line-like expression of Fgf8 was discernible along the roof plate of the mesencephalon(Fig. 2M, n=15/16). sprouty2 was induced in a similar pattern to Fgf8 expression(Fig. 2N), but in a little wider region than that of Fgf8 at the lateral side of the neural tube(Fig. 2O,P). Induction of sprouty2 along the roof plate was also visible(Fig. 2N).

At 24 hours after electroporation with 0.1 μg/μl of expression vector(HH 18), which also changes the fate of the mesencephalon to the metencephalon, strong V-shaped expression of Fgf8 and sprouty2 was discernible (Fig. 2Q,R, n=3/5). Differential hybridization revealed that introduced Fgf8 was hardly detected, but the V-shaped Fgf8expression was the transcripts from the embryonic gene (n=2/2).

We supposed that Fgf8b misexpression in a condition that changes the presumptive mesencephalon to the metencephalic property resulted in new isthmus formation, and that Fgf8 and Sprouty gene misexpression along the roof plate may represent new isthmus, that is, Fgf8 mRNA in the new isthmus may have been transcribed from the embryonic gene. So, we tried to distinguish Fgf8 mRNAs between those transcribed from transfected Fgf8 and those transcribed from the endogenous one. For this purpose, we electroporated mouse Fgf8bexpression vector and hybridized the embryos differentially with chick and mouse Fgf8 probes. Differential hybridization revealed that Fgf8b misexpression along the roof plate is the transcript from the embryonic gene (Fig. 2S, n=5/6). Introduced mouse Fgf8 was observed widely on the lateral part of the mesencephalon, but not near the roof plate.

Since the new Fgf8 expression line was set after Fgf8bmisexpression, we carried out time course analysis of the reconstruction of the isthmus. The posterior limit of the Otx2 expression domain gradually retreated toward the anterior(Fig. 3A; 3, 6 and 9 hours after electroporation, n=2/3, 5/5 and 5/7 respectively). On the contrary, the Gbx2 expression domain extended anteriorly complementary to the Otx2 expression domain after Fgf8bmisexpression (Fig. 3B; 3, 6 and 9 hours after electroporation, n=2/2, 3/4 and 3/3,respectively).

Isthmic Fgf8 expression had been kept until 6 hours after electroporation of Fgf8b expression vector(Fig. 3C, parts a and b, n=4/4), but expression became very weak by 9 hours after electroporation (Fig. 3C, part c, n=5/6), and disappeared by 12 hours after electroporation(n=3/4). New Fgf8 expression line, which had been transcribed from the embryonic gene, appeared near the roof plate in the embryos 24 and 36 hours after electroporation of Fgf8b expression vector (Fig. 2M-S and Fig. 3C, part f, n=12/16). sprouty2 expression was in a similar pattern to Fgf8 expression (3, 6 and 9 hours after electroporation, n=7/9, 2/2 and 3/3, respectively).

We also examined Limx1b expression. Limx1b was shown to repress Fgf8 in a cell-autonomous fashion, but induces Fgf8expression around Limx1b-expressing cells(Matsunaga et al., 2002). Limx1b was induced by misexpressed Fgf8b(Fig. 3D, parts a and b, n=4/4), but Limx1b expression in the mesencephalon became gradually repressed Fig. 3D,parts c-e, n=3/3), and ring-like expression of Limx1bremained in the diencephalon by 36 hours after electroporation(Fig. 3D, parts f and g, n=2/3) (Matsunaga et al.,2002).

It has recently been shown that the Fgf8b signal is transduced by the Ras-ERK signaling pathway to organize the cerebellar differentiation(Sato and Nakamura, 2004). We have shown that sprouty2 is induced very rapidly by Fgf8, but it is indicated that Sprouty2 acts as a negative regulator(Hacohen et al., 1998; Casci et al., 1999; Kramer et al., 1999; Lin et al., 2002). Then we wondered whether Sprouty2 really acts as a negative regulator for the Ras-ERK signaling pathway, and examined the effects of sprouty2 on ERK activity. Activated ERK is phosphorylated at both threonine and tyrosine residues that lie adjacent to each other in the unique TEY sequence(diphosphorylated ERK, dpERK), and could be distinguished by the anti-dpERK antibody (Gabay et al., 1997; Christen and Slack, 1999; Shinya et al., 2001).

Time course analysis of the effects of Sprouty2 and the dominant negative form of Sprouty2 on ERK activity was carried out. In the control side,decrease in the ERK activation zone could be recognized during the time course examined (Fig. 4A-C). Repression of ERK activity was already discernible at 3 hours after electroporation of Sprouty2 expression vector(Fig. 4A, n=3/5). Repression became stronger as time passed, and by 9 hours after electroporation, very strong repression was discernible as assessed by immunohistochemistry with anti-dpERK antibody(Fig. 4C,D, n=6/6).

On the other hand, repression of Sprouty2 activity by antisense morpholino oligonucleotide against Sprouty2 (Fig. 4E, n=2/2) or by sprouty2-DN misexpression(Fig. 4F-H, n=14/15),augmented the activation level of ERK. Although the activation zone of ERK diminished in the control side as an embryo developed(Fig. 4A-C), in the experimental side, the activation zone of ERK remained wide(Fig. 4H).

Co-transfection of sprouty2 and sprouty2-DN showed that Sprouty2-DN canceled the effects of Sprouty2. Co-transfection resulted in slight widening of the ERK activation zone(Fig. 4I). This indicates that Sprouty2-DN really suppress Sprouty2 activity.

The results indicate that Sprouty2 repressed ERK phosphorylation, and that repression of Sprouty2 activity raised the activation level of ERK. Thus, our study has confirmed that Sprouty2 functions as a negative regulator of the Ras-ERK pathway.

Recently, It was shown that Fgf8b could change the fate of the presumptive mesencephalon from the optic tectum to the cerebellum by activating the Ras-ERK signaling pathway(Sato et al., 2001; Liu et al., 2003; Sato and Nakamura, 2004). Disruption of the Ras-ERK pathway by the dominant negative form of Ras resulted in a fate change of the presumptive metencephalon to the tectum(Sato and Nakamura, 2004). We carried out misexpression of sprouty2, wondering whether Sprouty2 could also change the presumptive metencephalon to the mesencephalic property

We could distinguish the cerebellum and the tectum of E12.5 (HH 38) gross morphologically, because cerebellar swelling is characterized by sulci on its surface and the tectal swelling is smooth and larger than the former(Fig. 5A-C; n=7/10). Histologically, the cerebellum is characterized by the external granular layer, while the tectum has its distinct layer formation(Fig. 5D,E,H). After sprouty2 misexpression, the swelling in the metencephalic region on the experimental side did not have sulci, and looked smooth(Fig. 5A,C; n=3/3). Histologically, the swelling did not have an external granular layer, but had the laminar structure that was comparable to the tectum proper(Fig. 5D,F,G). Thus we conclude that the optic tectum differentiated in place of the cerebellum by sprouty2 misexpression. Trochlear nerve trajectory also supports the idea that the ectopic tectum differentiated in the metencephalic region. In normal embryos, the trochlear nucleus occupies the ventral part of the isthmus, from which nerve fibers run dorsally along the posterior margin of the mesencephalon (Fig. 5I). In the embryos at HH 21 after sprouty2 misexpression, nerve fibers that ran along the posterior margin of the ectopic swelling could be detected in addition to the proper trochlear nerve fibers(Fig. 5J).

We looked at the effects of Sprouty2 on mesencephalon- and metencephalon-related gene expression. It has been shown by misexpression study that the mesencephalon-metencephalon boundary is determined by repressive interaction between Otx2 and Gbx2(Broccoli et al., 1999; Millet et al., 1999; Katahira et al., 2000). Otx2 could change the fate of the metencephalic alar plate to the tectum. It was also shown that Fgf8 expression is induced at the boundary of the Otx2 and Gbx2 expression domain overlapping with the Gbx2 domain. At 24 hours after electroporation of sprouty2 (HH 18), Otx2 expression was induced in the metencephalic region (Fig. 6A-C; n=12/16). Gbx2 expression in the metencephalic region was repressed as in the case of Otx2misexpression (Fig. 6E-G; n=7/10). By 48 hours after electroporation (HH 22), regulation of Otx2 and Gbx2 expression may have occurred and their expression became complementary in the metencephalic region(Fig. 6D,H; Otx2; n=6/7, Gbx2; n=3/4), which corresponds well with the morphological change observed at a later stage.

Fgf8 is expressed in the isthmus in a ring-like pattern in normal embryos (Fig. 6K). After sprouty2 misexpression, the Fgf8 expression belt shifted to the caudal region (Fig. 6I,J; n=5/8).

Next, we carried out misexpression of sprouty2-DN, to see if it exerts opposite effects to sprouty2. As in the case of Gbx2misexpression, sprouty2-DN caused a rostral shift of the caudal boundary of the tectum (Fig. 7; n=5/8). The histology and expression pattern of Wnt1 all support the rostral shift of the caudal boundary of the tectum(Fig. 7B,D; n=2/3). Wnt1 is normally expressed at the posterior margin of the mesencephalon in addition to the dorsal midline of the mesencephalon. After sprouty2-DN misexpression, the Wnt1 expression ring shifted rostrally (Fig. 7D).

sprouty2-DN induced Gbx2 expression(Fig. 8D-F; n=12/13). As in the case of Gbx2 misexpression(Katahira et al., 2000), Otx2 expression was repressed(Fig. 8A-C; n=7/8). Otx2 and Gbx2 expression may have been regulated and their expression became complementary by 48 hours after electroporation(Fig. 8C,F; Gbx2; n=5/5, Otx2; n=3/5). Fgf8 expression was induced at the caudal part of the mesencephalon(Fig. 8G,H).

In the present study we have shown that: (1) Sprouty2 interferes with ERK activation, (2) sprouty2 is induced very rapidly by both Fgf8a and Fgf8b, (3) after Fgf8b misexpression that causes fate change of the presumptive mesencephalon to the metencephalon, endogenous Fgf8expression is induced along the roof plate, and Sprouty is also induced overlapping the new Fgf8 expression site, (4) sprouty2misexpression changes the fate of the presumptive metencephalon to the mesencephalon, (5) a dominant negative form of Sprouty2 activates ERK and causes anterior shift of the posterior margin of the tectum.

In normal development, ERK activation was seen at the site of Fgf8 expression, that is, the anterior neural ridge and the isthmus. It has been indicated that Sprouty2 acts as a negative feedback regulator of the growth factor-mediated Ras-ERK signaling pathway(Hacohen et al., 1998; Kramer et al., 1999; Casci et al., 1999; Lin et al., 2002). On the other hand, contrary effects of Sprouty were reported for EGF receptor(EGFR)-mediated signaling (Egan et al.,2002; Wong et al.,2002; Fong et al.,2003; Hall et al.,2003; Rubin et al.,2003; Stang et al.,2004). hSprouty2 sequestrated c-Cbl, which in turn abrogated EGFR ubiquitylation and endocytosis, and consequently sustained EGF-induced ERK signaling. The present study showed that misexpression of Sprouty2 interfered with ERK activation, and Sprouty2-DN augmented ERK activation as assessed by staining with anti-diphosphorylated ERK. Morpholino antisense oligonucleotide against Sprouty2 also augmented activity of ERK. Thus we have concluded that Sprouty2 negatively regulates the Ras-ERK pathway in the isthmus.

sprouty2 was induced by both Fgf8a and Fgf8b. Induction of sprouty2 by Fgf8 is very rapid. sprouty2 induction could be already seen by 3 hours after electroporation of both Fgf8a and Fgf8b corresponding to the Fgf8 misexpression site. If we consider that the translation products could be detectable by 2 hours after electroporation, sprouty2 induction may occur within an hour of Fgf8 expression. The results indicate that the Fgf8b signal is transduced very rapidly via the Ras-ERK signaling pathway, and controls transcription of the downstream gene.

Fgf8b misexpression by electroporation at a concentration of 0.1μg/μl or 1 μg/μl changes the fate of the presumptive mesencephalon to that of the metencephalon (Sato et al.,2001). After 24 hours of electroporation of 1.0 μg/μl of Fgf8b vector, Fgf8 and Sprouty2 expression was observed on the lateral side of the mesencephalon and along the roof plate,while expression along the roof plate did not exist after Fgf8amisexpression. Liu et al. (Liu et al.,2003) suggested that misexpression of Fgf8 along the roof plate represents new isthmus after Fgf8 misexpression, that is, Fgf8 expression along the roof plate is from the embryonic gene. We tried to confirm this notion by electroporating mouse Fgf8bexpression vector, and by hybridization in situ with chick and mouse Fgf8b probes that do not cross hybridize each other. It was revealed that Fgf8 expression along the roof plate is the transcript from the endogenous chick Fgf8, and that Fgf8 mRNA on the lateral side of the mesencephalon was that transcribed from the introduced mouse Fgf8b. Since expression along the roof plate did not exist after Fgf8a misexpression, we supposed that new isthmus is formed along the roof plate, and consequently, mesencephalon may have acquired the metencephalic property after Fgf8b misexpression. After electroporation with 0.1 μg/μl Fgf8b expression vector, which also changes the fate of the presumptive mesencephalon to the metencephalon, new isthmus was set as V-shaped. It also indicates that sprouty2 is induced at the ectopic site very rapidly and negatively regulates Fgf signaling.

It was shown that Fgf8 is induced at the interface of the Otx2 and Gbx2 expression domain, overlapping with Gbx2 expression (Broccoli et al.,1999; Millet et al.,1999; Hidalgo-Sanchez et al., 1999; Katahira et al., 2000; Garda et al., 2001). After Fgf8b misexpression, a new Fgf8 expression line is set at the border of the Otx2 and Gbx2 expression domain. Reconstruction of the Fgf8 expression line may be a result of interaction among Fgf8, Limx1b, Wnt1, Otx2 and Gbx2(Garda et al., 2001; Matsunaga et al., 2002). It was shown that Limx1b represses Fgf8 expression in a cell-autonomous fashion (Matsunaga et al.,2002), Limx1b induces Wnt1 expression, and Wnt1 in turn induces Fgf8 expression. Thus Limx1b induces Fgf8 expression around Limx1b-expressing cells. Although Gbx2 represses Limx1bexpression, Fgf8 induces Limx1b expression(Matsunaga et al., 2002). After Fgf8b misexpression, Limx1b expression was induced in the mesencephalic region in a similar pattern to Fgf8 misexpression at first. Then Limx1b may have repressed endogenous Fgf8 expression in the isthmus. Since Fgf8b induces Gbx2, Limx1b expression in the mesencephalic region may have been repressed by Gbx2, and disappeared by 36 hours after electroporation. Finally, ring-like expression remains rostral to the Gbx2 expression. Now the roof plate has become the interface of Otx2 and Gbx2 expression in the mesencephalic region, Fgf8 expression may have been induced along the roof plate.

As discussed before, Fgf8b could change the property of the presumptive mesencephalon to that of the metencephalon. This raises a question as to how the Fgf8 signal is transduced to organize cerebellar differentiation. This subject was challenged by disrupting the Ras-ERK signaling pathway by misexpression of dominant negative form of Ras(Sato and Nakamura, 2004). Since disruption of the Ras-ERK signaling pathway resulted in differentiation of the optic tectum in place of the cerebellum, and Ras-DN canceled the effects of Fgf8b after co-electroporation of Fgf8b and Ras-DN, it was suggested that Fgf8b activates the Ras-ERK signaling pathway to organize cerebellar differentiation. Very recent study showed that Ras-ERK signaling cascades modulate the activity of Irx2 by phosphorylation for the cerebellar development (Matsumoto et al.,2004).

It was reported that Sprouty2 negatively regulates the Ras-ERK signaling pathway. We have shown that Sprouty2 really repressed ERK phosphorylation. Misexpression of Sprouty2 induced Otx2 expression and repressed Gbx2 expression in the metencephalon, and resulted in the differentiation of tectum in place of the cerebellum. As in the case of the dominant negative form of Ras misexpression, the swelling in the metencephalic region after sprouty2 misexpression did not show cerebellar sulci on its surface but was smooth. Histologically, it did not contain the external granular layer, which is characteristic of the cerebellum, but consisted of the laminar architecture characterisitic of the tectum. Misexpression of sprouty2-DN exerted the reverse effects to sprouty2. Gbx2 was induced in the mesencephalon by Sprouty2-DN, and the obtained results were similar to those after Gbx2 misexpression. On the control side, ERK activation zone in the isthmus narrows from stage 9 to 10 (see Fig. 4A-C,F,G). Repression of Sprouty2 activity by misexpression of Sprouty2-DN or by application of morpholino antisense oligonucleotide,interfered with narrowing of the ERK activation zone (see Fig. 4E-H). The results indicate that the Ras-ERK signal should be weakened in a short period of development, and Sprouty2 may contribute to weakening of the Ras-ERK signaling. Taken together, the results indicate that the Fgf8 signal is so strong that a negative regulator is needed. Once the Fgf8 signal is transduced, it quickly induces its negative regulator, thus the negative feedback loop may regulate the Fgf8 signaling.

Recently, another negative regulator for Fgf signaling, Sef, was reported(Minowada et al., 1999; Fürthauer et al., 2002; Lin et al., 2002). Sef is expressed in an overlapping manner to Fgf8, induced by Fgf8, and functions as a negative regulator for Fgf8 signaling. Sef is also expressed in the isthmus(Fürthauer et al., 2002; Lin et al., 2002). The relationship between Sef and Sprouty must be elucidated, but the Fgf8 signaling plays a crucial role in morphogenesis and is so strong, its signaling may be regulated repeatedly.

In conclusion, Fgf8 activates the Ras-ERK signaling pathway, and very rapidly regulates downstream gene expression. sprouty2 is induced very rapidly by Fgf8-Ras-ERK signaling, and regulates this pathway negatively for the metencephalon to receive appropriate signaling. We have shown that regionalization of the neural tube is disturbed by both hyper- and hypo-signaling.

This work was supported by the grants from the Ministry of Education,Culture, Sports, Science and Technology and from the Mitsubishi Foundation. T.S. is a recipient of a JSPS Research Fellowship for Young Scientists.