Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Feb 15.
Published in final edited form as: Dev Biol. 2009 Dec 5;338(2):183. doi: 10.1016/j.ydbio.2009.11.034

Fgf8b-containing spliceforms, but not Fgf8a, are essential for Fgf8 function during development of the midbrain and cerebellum

Qiuxia Guo 1, Kairong Li 1, N Abimbola Sunmonu 1, James YH Li 1,*
PMCID: PMC2815264  NIHMSID: NIHMS164281  PMID: 19968985

Summary

The single Fgf8 gene in mice produces eight protein isoforms (Fgf8a–h) with different N-termini by alternative splicing. Gain-of-function studies have demonstrated that Fgf8a and Fgf8b have distinct activities in the developing midbrain and hindbrain (MHB) due to their different binding affinities with FGF receptors. Here we have performed loss-of-function analyses to determine the in vivo requirement for these two Fgf8 spliceforms during MHB development. We showed that deletion of Fgf8b-containing spliceforms (b, d, f and h) leads to loss of multiple key regulatory genes, including Fgf8 itself, in the MHB region. Therefore, specific inactivation of Fgf8b-containing spliceforms, similar to the loss of Fgf8, in MHB progenitors results in deletion of the midbrain, isthmus, and cerebellum. We also created a splice-site mutation abolishing Fgf8a-containing spliceforms (a, c, e, and g). Mice lacking Fgf8a-containing spliceforms exhibit growth retardation and postnatal lethality, and the phenotype is variable in different genetic backgrounds, suggesting that the Fgf8a-containing spliceforms may play a role in modulating the activity of Fgf8. Surprisingly, no discernable defect was detected in the midbrain and cerebellum of Fgf8a-deficient mice. To determine if Fgf17, which is expressed in the MHB region and possesses similar activities to Fgf8a based on gain-of-function studies, may compensate for the loss of Fgf8a, we generated Fgf17 and Fgf8a double mutant mice. Mice lacking both Fgf8a-containing spliceforms and Fgf17 display the same defect in the posterior midbrain and anterior cerebellum as Fgf17 mutant mice. Therefore, Fgf8b-containing spliceforms, but not Fgf8a, are essential for the function of Fgf8 during the development of the midbrain and cerebellum.

Keywords: Fgf8, alternative splicing, midbrain, cerebellum, organizer, signaling

Introduction

Alternative splicing occurs in 40–60% of the genes in humans and mice (Modrek and Lee, 2002). It is believed that alternative splicing significantly increases the complexity of the RNA and protein products expressed by the mammalian genome and allows an additional layer of control of gene function. However, the biological significance of alternative splicing of most mammalian genes is largely unexamined. Studies of the in vivo function of alternative splicing of key developmental regulators should provide important insight into the regulation and function of these molecules during development.

Fibroblast growth factor 8 (Fgf8), which belongs to the fibroblast growth factor (FGF) family, plays a key role in regulating different developmental processes (Itoh and Ornitz, 2008). Deletion of Fgf8 in mice leads to catastrophic consequences during gastrulation, in addition to latter roles in the development of the central nervous system, sensory system, heart, limbs, jaws and teeth (Chi et al., 2003; Frank et al., 2002; Garel et al., 2003; Lewandoski et al., 2000; Moon and Capecchi, 2000; Storm et al., 2003; Sun et al., 2002; Sun et al., 1999; Trumpp et al., 1999). One of the best-studied functions of Fgf8 is in formation of the midbrain and cerebellum, which develop from two adjacent segments of the neural tube called the mesencephalon (mes) and rhombomere 1 (r1), respectively (Wurst and Bally-Cuif, 2001). The dorsal mes gives rise to the superior colliculus and inferior colliculus, which process visual and auditory information, respectively, whereas dorsal r1 forms the cerebellum, which coordinates motor activity. Embryological and genetic studies have demonstrated that the expression of Fgf8 in the mes/r1 junction, known as the isthmic organizer, patterns the mes and r1 and coordinates the growth and differentiation of the midbrain and cerebellum (Sato et al., 2004; Wurst and Bally-Cuif, 2001). Deletion of Fgf8 in the mes/r1 region results in cell death between E8.5 and E9.5, and eventual loss of the entire midbrain, isthmus, and cerebellum by E18.5 (Chi et al., 2003).

The vertebrate Fgf8 gene undergoes alternative splicing to produce eight possible spliceforms (a–h) that encode protein isoforms with different N-termini (Crossley and Martin, 1995; MacArthur et al., 1995; Sato et al., 2001). Previous studies have shown that two of the evolutionarily conserved Fgf8 isoforms, Fgf8a and Fgf8b, possess distinct activities (Sato et al., 2004). In mouse and chick embryos, misexpression of Fgf8a results in overgrowth of the mes and transforms the posterior forebrain into a midbrain fate (Lee et al., 1997; Liu et al., 1999; Sato et al., 2001). By contrast, misexpression of Fgf8b transforms the mes into a cerebellar fate (Liu et al., 1999; Sato et al., 2001). The remarkably different activities of these two Fgf8 isoforms can be exclusively attributed to a single amino acid in Fgf8b, which forms an additional binding site and thus increases affinity for FGFRs (Olsen et al., 2006). Analyses of mouse embryos lacking Fgf8b-containing spliceforms (b, d, f, and h) have demonstrated that these Fgf8 spliceforms are essential for the function of Fgf8 before gastrulation, but that the remaining Fgf8a-containing spliceforms (a, c, e and g) can partially compensate for the loss of Fgf8b-containig spliceforms in promoting mesoderm migration through the primitive streak in a dosage-dependent manner (Guo and Li, 2007). Experiments in Xenopus embryos have shown that Fgf8a plays an important role in the formation of the posterior neural tissue and induction of neural crest (Fletcher et al., 2006; Hong et al., 2008). The above observations suggest that Fgf8a and Fgf8b may work in concert to determine the function of Fgf8 in vertebrate development. However, it remains unknown whether Fgf8a and Fgf8b have different contributions to Fgf8 function during organogenesis in mice.

Fgf17, which is closely related to Fgf8, is expressed in a broad domain around the isthmus between E8.5 and E15.5 (Xu et al., 2000). Gain-of-function studies have demonstrated that Fgf17 has similar activity to Fgf8a (Liu et al., 2003; Olsen et al., 2006). Deletion of Fgf17 leads to partial loss of the posterior midbrain and anterior cerebellum in mice, and removal of one copy of Fgf8 enhances the tissue loss in the anterior cerebellum of Fgf17 null mutant mice (Xu et al., 2000). These observations suggest that there may be synergistic interactions between Fgf8a and Fgf17 in coordinating growth and differentiation of the cerebellum.

To determine the in vivo requirement for Fgf8a and Fgf8b during development of the midbrain and cerebellum, we have examined formation of these brain structures in mice lacking Fgf8b-containing spliceforms in tissue-specific conditional mutant mice. We also created a mutation abolishing Fgf8a-containing spliceforms in mice to determine the requirement for Fgf8a. Finally, by generating double mutant mice that lack Fgf17 and Fgf8a-containing spliceforms, we investigated the genetic interactions between Fgf17 and Fgf8a during development of the midbrain and cerebellum. Our studies demonstrate the distinct requirement for the two sets of Fgf8 spliceforms, containing Fgf8a or Fgf8b, in the overall function of Fgf8.

Materials and Methods

Generation of a mouse line deficient for Fgf8a

Using gene targeting, we introduced a point mutation into the 3′ alternative splice acceptor of Fgf8 exon 1D, changing the sequence from cagCATGTG to caaCATGTG (Fig. 3A). The mutation produces a new restriction endonuclease site for PciI without altering the final amino acid sequence. A selectable marker, neo was placed 400 bp downstream of exon 1D. Targeted embryonic stem cells carrying the mutation, designated as Fgf8Δa-neo, were identified by Southern blot, PCR and restriction analysis, and used to generate germline chimeras. The neo cassette, which was flanked by two loxP sites, was subsequently removed using a germ-line CMV-Cre mouse line (Li et al., 2002), producing a new allele called Fgf8Δa (Fig. 3A).

Figure 3. Mutant mice lacking Fgf8a spliceform are retarded in growth.

Figure 3

Open in a new tab

(A) Schematic representation of the generation of a mutation deleting Fgf8a in mice. (B) RT-PCR analysis of spliceforms of Fgf8 from total RNA of E8.5 embryos. Note that Fgf8a transcripts (asterisk) are missing in Fgf8Δa/Δa embryos, whereas the transcripts of Fgf8b, d, f and h appear slightly more abundant compared with those in the controls. Question mark indicates two unknown splice variants of Fgf8. (C) Fgf8Δa/Δa mice are smaller in size at 6 weeks old. (D) Histogram of the average body weight of Fgf8Δa/Δa mutants and their littermate controls at E18.5 and P24.

Mouse breeding and genotyping

Noon of the day on which the vaginal plug was found was designated as embryonic day 0.5 (E0.5). Embryos between E8.5 and E9.5 were staged according to the number of somites. All mouse lines were maintained on an outbred CD1 genetic background unless stated otherwise (Charles River Lab, Wilmington, MA). To produce C57BL/6J congenic animals, Fgf8+/Δa or Fgf8+/Δb mice were backcrossed with C57BL/6J mice (the Jackson Laboratory, Bar Harbor, ME) for at least seven generations. Genotypes of mice were determined by PCR.

RNA in situ hybridization and histochemistry

For tissue preparation, embryonic brains were fixed by immersion in 4% paraformaldehyde (PFA) at 4°C for 3–16 hours. Postnatal mice were deeply anesthetized and transcardially perfused with 4% PFA. Brains were postfixed in the same fixative overnight at 4°C. For frozen tissue sections, brains were cryoprotected in 30% sucrose in PBS before rapid freezing in OCT compound (TissueTek) and sectioned with freezing microtome. Standard protocols for X-gal staining, in situ hybridization, and immunofluorescence on sections or in whole mount embryos were used (Nagy, 2003). Detailed protocols are available on the Li lab website (http://www.genetics.uchc.edu/lilab/Pages/Protocols.html). Analysis of cell death was performed by Nile Blue Sulfate staining of wholemount embryos as described previously (Chi et al., 2003). Antibodies used in the study are as follows: rabbit anti-TH (Pel-Freez, Rogers, AR), goat anti-5-HT (ImmunoStar, Hudson, WI), and Alexa fluorescent secondary antibodies (Invitrogen, Carlsbad, CA).

Results

Fgf8b spliceform is essential for Fgf8 function during MHB development

We have previously shown that Fgf8Δb/Δb embryos, which lack Fgf8b, d, f and h spliceforms (defined as Fgf8b-containing spliceforms), display severe abnormalities and growth retardation at E9.5 (Guo and Li, 2007). To circumvent the early embryonic lethality of Fgf8Δb/Δb embryos, we combined the Fgf8Δb allele with a conditional Fgf8 mutant allele, Fgf8F (Meyers et al., 1998), and an En1Cre knock-in allele (Kimmel et al., 2000). In embryos carrying the En1Cre allele, the expression of Cre leads to recombination in cells throughout the mes/r1 area between the 3- to 10-somite stage (Li et al., 2002), so that the remaining wild-type Fgf8 allele is deleted and thus only Fgf8a-containing spliceforms (a, c, e, and g) are expressed in the mes/r1 of Fgf8Δb/F; En1+/Cre embryos, designated as Fgf8b conditional knockout (Fgf8b-CKO).

Fgf8Δb/F mice were normal, but Fgf8Δb/F; En1+/Cre mice died at birth. Examination of Fgf8Δb/F; En1+/Cre embryos at E18.5 revealed a nearly complete loss of the midbrain and the cerebellum (Fig. 1B). The phenotype is remarkably similar to that found in Fgf8−/F; En1+/Cre embryos with conditional deletion of Fgf8 (Fig. 1C)(Chi et al., 2003). To compare the extent of MHB tissue loss between Fgf8b-CKO and Fgf8-CKO mice, we introduced a Cre-reporter allele, R26R+ (Soriano, 1999), into these conditional mutant mice so that mes/r1 cells would be permanently marked by β-gal activity. Analysis of β-gal activity by X-gal histochemistry revealed that the entire mes/r1 area was populated by labeled cells in control embryos of genotype Fgf8+/F; En1+/Cre; R26R+/− at E10.5 and E18.5 (Fig. 1D and G). In both Fgf8b-CKO and Fgf8-CKO embryos, the mid-hindbrain domain demarcated by β-gal activity was reduced to a narrow stripe (Fig. 1E, F, H, and I). The residual X-gal positive domain was comparable between Fgf8b-CKO and Fgf8-CKO embryos at E18.5 (Fig. 1H–I). To analyze development of the mes/r1 basal plate, we examined the formation of dopaminergic and serotonergic neurons around the MHB junction (Zervas et al., 2004) by immunofluorescence with antibodies against tyrosine hydroxylase (TH) and 5-hydroxytryptamine (5-HT). As described previously (Chi et al., 2003), dopaminergic neurons were largely depleted in the substantia nigra and ventral tegmental area (SN-VTA) of Fgf8-CKO embryos at E18.5 (Fig. 1L). Furthermore, there was a noticeable reduction in the number of serotonergic neurons in the raphe nuclei (Fig. 1J and L). Similar to Fgf8-CKO embryos, Fgf8b-CKO embryos at E18.5 exhibited an almost complete loss of dopaminergic neurons in the SN-VTA area, and a reduction of serotonergic neurons in raphe nuclei (Fig. 1K). Together, our data show that the Fgf8b-containing spliceforms are essential for Fgf8 activity in MHB development.

Figure 1. Specific deletion of Fgf8b or Fgf8 in the mes/r1 region results in similar loss of midbrain and hindbrain tissues.

Figure 1

Open in a new tab

(A–C) Dorsal views of E18.5 brains of genotypes as indicated at the top. (D–I) X-gal staining of wholemount embryos at E10.5 (D–F) and sagittal sections of E18.5 brains (G–I) of genotypes indicated. Bracket marks the mes/r1 domain. (J–L) Immunofluorescence labeling of dopaminergic (TH) and serotonergic (5-HT) neurons in the ventral midbrain and hindbrain near the MHB junction on sagittal sections of E18.5 brains. The positions J–L are indicated in boxed areas in D–F, respectively. Arrowheads mark the remaining TH-positive neurons in the mutant embryos. Abbreviations: cb, cerebellum; ctx, cerebral cortex; ic, inferior colliculus; mb, midbrain; sc, superior colliculus.

Fgf8b-containing spliceforms are essential for maintaining the genetic cascade in the mes/r1 area

It has been shown that the expression of Wnt1, Gbx2, Fgf17, En1, and Spry2 is lost in the mes/r1 region of Fgf8−/F; En1+/Cre embryos between the 11- and 15-somite stages (Chi et al., 2003). To determine if Fgf8b-containing spliceforms are essential for Fgf8 signaling to maintain the expression of these mes/r1 genes, we performed RNA in situ hybridization on Fgf8Δb/F; En1Cre/+ and control littermates between the 11- and 14-somite stages. Transcripts of Fgf8 were detected in the isthmus of Fgf8Δb/F; En1Cre/+; R26R+/− embryos at the 11-somite stage, but were missing by the 14-somite stages (Fig. 2B and data not shown). At the 14-somite stage, Wnt1 and Gbx2 are normally expressed in two apposing transverse stripes in the posterior mes and anterior r1, respectively (Fig. 2C and E). In addition, Wnt1 is expressed in dorsal midline, except for in r1 region, of the neural tube (Fig. 2C). The transverse stripes of Wnt1 and Gbx2 expression at the mes/r1 border were lost in Fgf8Δb/F; En1Cre/+ embryos at the 14-somite stage (Fig. 2D and F). However, Wnt1 expression in the dorsal midline of the neural plate persisted and was abnormally expanded through presumptive r1 (Fig. 2D). The expression of En1 and Fgf17 was absent or greatly diminished at the mes/r1 region of Fgf8Δb/F; En1Cre/+ embryos at the 13–14 somite stages (Fig. 2H and J). These changes in gene expression in Fgf8b-CKO embryos are remarkably similar to those reported in Fgf8-CKO mutants (Chi et al., 2003). We examined two additional markers for mes/r1. The homeobox gene Dmbx1 is expressed in the mes and diencephalon (Fig. 2K) (Ohtoshi et al., 2002). Spry2, which is a downstream target and feedback inhibitor of Fgf8 signaling, is expressed in the posterior mes and the metencephalon (Fig. 2M)(Liu et al., 2003). No expression of Dmbx1 and Spry2 was detected in mes/r1, whereas their expression in the diencephalon and the posterior metencephalon, respectively, was maintained in Fgf8Δb/F; En1Cre/+ embryos at the 13-somite stages (Fig. 2L and N). As described in Fgf8-CKO embryos (Chi et al., 2003), Nile Blue sulfate staining, which specifically marks dying cells, revealed a significant increase in the number of dead cells in the mes and to a less extent in r1 of Fgf8+/F; En1Cre/+ embryos at the 13-somite stage (Fig. 2P). These data show that inactivation of Fgf8b in mes/r1 results in very similar responses in both gene expression and cell death to those found in the Fgf8-CKO embryos. Our results thus demonstrate that Fgf8b-containing spliceforms are essential for Fgf8 function in maintaining the mes/r1 genetic cascade and the survival of mes/r1 cells.

Figure 2. Fgf8b is required to maintain expression of mes/r1 genes.

Figure 2

Open in a new tab

(A–N) In situ hybridization. Probes and genotypes are indicated at the left and top of the panel, respectively. The number of somites is indicated between each pair of control and mutant embryos. A ribroprobe corresponding to a full-length Fgf8 cDNA is used so that Fgf8 transcription is analyzed even though exon 3 is deleted by Cre-mediated recombination of Fgf8F allele. Arrowhead marks reduced expression of Fgf8 at the isthmus in Fgf8b-CKO embryos; asterisk indicates the absence of gene expression at the isthmus; bracket demarcates r1 region; dashed line circles the mes-r1 region. (O–P) Nile Blue sulfate staining of dying cells. Arrow indicates increased number of dead cells in the mes.

Deletion of Fgf8a-containing spliceforms results in growth retardation and postnatal lethality in mice

Because deletion of Fgf8b leads to loss of the expression of Fgf8a and all other remaining Fgf8 spliceforms in the isthmus by E9.5, we were unable to determine whether Fgf8a plays any role in the development of the midbrain and cerebellum. To address this question, we abolished Fgf8a expression by introducing a point mutation (designated as Fgf8Δa) at the 3′ alternative splice acceptor of Fgf8 exon 1D in mice (Fig. 3A). RT-PCR demonstrated that the transcripts of Fgf8a-containing spliceforms (a, c, e, and g) were missing, while Fgf8b-contaiing spliceforms, which utilize the remaining alternative splice acceptor of the exon 1D, are expressed in Fgf8Δa/Δa embryos at E8.5 (Fig. 3B).

Most Fgf8Δa/Δa mice on a CD1 outbred background were viable. However, adult Fgf8Δa/Δa mice (6-week or older) were smaller in size (Fig. 3C), and the average body weight of Fgf8Δa/Δa mutants was significantly lower than that of wild type or heterozygous littermates at E18.5 and postnatal day 24 (P24) (Fig. 3D). Some Fgf8Δa/Δa mice (14 out of 86 pups) were found dead before weaning, and the number of Fgf8Δa/Δa mice recovered at weaning was significantly lower than expected according to Mendelian inheritance (Table Ia). However, embryos of different genotypes from intercrosses between Fgf8+/Δa mice were recovered at normal ratios (Table Ib), suggesting that the homozygous mutation of Fgf8Δa results in partial lethality after birth. Interestingly, compared with wild-type controls, Fgf8+/Δa mice tended to have a lower body weight (Fig. 3C). On a C57BL/6J genetic background, few Fgf8Δa/Δa mice were recovered, and the number of Fgf8+/Δa mice was significantly reduced (p<0.039, X2 test) (Table Ic). We examined matings between wild-type and Fgf8+/Δa or Fgf8+/Δb mice, and recovered normal numbers of Fgf8+/Δb pups, but Fgf8+/Δa pups were significantly fewer than that predicted by Mendelian ratios (p<0.029, X2 test) (Table Id–e). Since both Fgf8Δa and Fgf8Δb alleles contain the same residual loxP site and are genotyped by the same method (see Materials and Methods), the observed reduction of Fgf8+/Δa mice is unlikely due to an adverse effect of the loxP sequence left in the intron of Fgf8Δa allele, or errors of PCR genotyping. Collectively, our results demonstrate that the deletion of Fgf8a-containing spliceforms leads to growth retardation and postnatal lethality, and the phenotype of Fgf8+/Δa and Fgf8Δa/Δa mice is more severe on a C57BL/6J versus CD1 background.

Table I.

Ratio of progeny of different genotypes

Genotype
Mating WT Het Homo Total Litters Average litter size Chi square test
a CD1 Fgf8+/Δa × Fgf8+/Δa (P0-20) 122 201 (4) 86 (14) 427 39 10.9 0.0340
b Fgf8+/Δa × Fgf8+/Δa (E15.5–18.5) 46 108 51 205 16 12.8 0.659
*0.0389
c C57BL/6J Fgf8+/Δa × Fgf8+/Δa (P0-20) 40 52* 3 (5)** 71 26 3.9 **3.601×10−7
d Fgf8+/Δa × Fgf8+/+ (P0-20) 57 36 (2) 89 17 5.6 0.029
e Fgf8+/Δb × Fgf8+/+(P0-20) 52 53 105 19 5.5 0.922
Open in a new tab

Number of dead pups is shown in parenthesis.

Fgf8a-containing spliceforms are dispensable for the development of the midbrain and cerebellum

To determine if Fgf8a is essential for the development of the mes/r1, we examined the midbrain and cerebellum of Fgf8Δa/Δa mutants at E18.5 and P15. The histology of the superior colliculus, inferior colliculus, and the cerebellum, which are derived from the alar plate of mes/r1, was indistinguishable among Fgf8Δa/Δa, Fgf8+/Δa, and their wild-type littermates (Fig. 4A–D). Immunofluorescence analysis of TH and 5-HT showed no obvious alterations of dopaminergic neurons and serotonergic neurons in the ventral midbrain and pons of Fgf8Δa/Δa embryos at E15.5 (Fig. 4E–F). Furthermore, we performed whole-mount immunohistochemistry using an anti-neurofilament antibody to examine the cranial nerves and their associated ganglia that derived from the MHB region. The formation of the cranial nerves (III–XI), their associated ganglia, and peripheral nerves was indistinguishable between Fgf8Δa/Δa and wild-type embryos at E10.5 and E11.5 (Fig. 4K–L and data not shown). These data collectively suggest that both the alar and basal plates of mes/r1 develop normally without Fgf8a-containing spliceforms.

Figure 4. Normal development of the midbrain and hindbrain in mice lacking Fgf8a.

Figure 4

Open in a new tab

(A–D) Nissl staining of sagittal sections of brains at E18.5 (A–B) and P15 (C–D). The genotypes are indicated at the top. Insets show dorsal views of the midbrain and cerebellum. (E–F) Immunofluorescence of TH and 5-HT positive neurons in the basal plate around the mid-hindbrain junction (dashed line). Arrowhead indicates the junction between the inferior colliculus (ic) and cerebellum (cb). (G–J) X-gal staining of sagittal sections of brains at E16.5 (G–H) and P15 (I–J). The genotypes are indicated at the top of the panel. (K–L) Neurofilament immunohistochemistry of cranial nerves (Roman numerals) in wild-type (K) and Fgf8Δa/Δa embryos at E10.5. Abbreviations: dr, dorsal raphe nuclei; mb, midbrain; pmr, paramedial raphe nuclei; sc, superior colliculus; vta, ventral tegmental area.

We next generated En1Cre/+; Fgf8Δa/F; R26R+/− mice to specifically remove Fgf8a-containing spliceforms in the mes/r1 region. These mutant mice were normal in size, demonstrating that the growth retardation of Fgf8Δa/Δa mutants is caused by a defect outside the MHB region. No difference in the size of the midbrain and the cerebellum, which were marked by X-gal histochemistry, was observed between En1Cre/+; Fgf8Δa/F; R26R+/−Δand En1Cre/+; Fgf8Δa/+; R26R+/− littermates at E16.5 or P15 (Fig. 4G–J).

To examine the patterning and growth of the mes/r1, we analyzed expression of MHB markers in Fgf8Δa/Δa embryos at E8.5, E9.5, and E10.5. Normal expression of Gbx2, En1, Wnt1, and Fgf8 (Fgf8b, d, f and h only) was found in the mes/r1 region between E8.5 and E10.5, suggesting that loss of Fgf8a-containing spliceforms does not affect patterning of the mes/r1 (Fig. 5B and data not shown). Spry1 and Spry2 are expressed in two gradients diminishing from the isthmus in the mes/r region (Fig. 5C, and data not shown), and the expression of these two genes is considered a readout of Fgf8 signaling (Liu et al., 2003). No noticeable difference in the expression of Spry1 and Spry2 was found in the mes/r1 region of Fgf8Δa/Δa embryos at E9.5 and E10.5 (Fig. 5D and data not shown). To determine the size of the mes, we analyzed the expression of two homeobox genes, Meis2 and Dmbx1, which are expressed in the mes and other areas of the neural tube (Fig. 5E and G) (Cecconi et al., 1997; Gogoi et al., 2002; Oulad-Abdelghani et al., 1997). The expression domains of Meis2 and Dmbx1 in the mes were indistinguishable between control and Fgf8Δa/Δa embryos at E10.5 (Fig. 5F and H). Our data collectively demonstrate that deletion of Fgf8a-containing spliceforms does not result in noticeable change in patterning and growth of the mes/r1, and that Fgf8a-containing spliceforms are dispensable for the formation of the midbrain and cerebellum.

Figure 5. Expression of mes/r1 genes is unaffected in the absence of Fgf8a.

Figure 5

Open in a new tab

(A–H) In situ hybridization of whole-mount embryos at E10.5. The probes and genotypes are indicated to the left and the top, respectively. The range (brackets) and the gradient (boxes) of marker expression in the mes or r1 are indistinguishable between wild-type and Fgf8Δa/Δa embryos. Arrow and asterisk indicate the isthmus and the diencephalon, respectively.

Fgf17 dose not interacts with Fgf8a-containing spliceforms in the regulation of MHB development

Previous studies have demonstrated that there are genetic interactions between Fgf8 and Fgf17 (Xu et al., 2000). Furthermore, gain-of-function studies showed that Fgf8a and Fgf17 have similar biological activity (Liu et al., 2003; Olsen et al., 2006). To determine if Fgf17 may compensate for the loss of Fgf8a in Fgf8Δa/Δa mice, we examined the formation of the midbrain and the cerebellum in mutant mice with Fgf8Δa mutation on an Fgf17-null background.

Histology of the midbrain and cerebellum was examined around 5 weeks after birth, when these brain structures have fully matured. As described previously (Xu et al., 2000), the tissue between the inferior colliculus and the anterior cerebellum was significantly reduced in adult Fgf17−/− mice (Fig. 6B). Although there were variations in the size and the pattern of lobules I–III among wild-type mice, lobules I–III were truncated and considerably smaller in size in adult Fgf17−/− mice (Fig. 6B). Significantly, no additional reduction was detected in the inferior colliculus and the lobules I–III in the cerebellar vermis in Fgf8+/Δa; Fgf17−/− (n=11) or Fgf8Δa/Δa; Fgf17−/− mice (n=5) compared with Fgf17−/− mice (Fig. 6C and data not shown). These data demonstrate that Fgf8a and Fgf17 do not play redundant roles in development of the midbrain and cerebellum.

Figure 6. Fgf8a and Fgf17 does not play redundant role in the development of the cerebellum.

Figure 6

Open in a new tab

(A–C) Nissl staining of sagittal sections of the midbrain and cerebellum of indicated genotypes at about 3 weeks of age. The lobules of the vermis are marked by Roman numerals according to Inouye and Oda (Inouye and Oda, 1980). Red arrowheads and arrows indicate the truncated lobules I–II and III, respectively. Abbreviations: ic, inferior colliculus; sc, superior colliculus.

Discussion

Previous studies using a gain-of-function approach have shown that the two Fgf8 splice variants Fgf8a and Fgf8b have distinct inductive activities in mes/r1 fate specification (Lee et al., 1997; Liu et al., 1999; Sato et al., 2001). In chick embryos, expression of Fgf8a and Fgf8b are co-expressed in the isthmus (Sato and Nakamura, 2004). Furthermore, we observed that all eight Fgf8 splice variants are co-expressed in the mes/r1 region of mouse embryos between E8.5 and E16.5 (Li and Li, manuscript in preparation). However, the in vivo function of these Fgf8 spliceforms has not been examined. Using a loss-of-function approach, we have determined the in vivo requirement for two complementary sets of Fgf8 spliceforms containing Fgf8a or Fgf8b in development of the midbrain and cerebellum in mice, and uncovered molecular and cellular processes differentially regulated by each set of spliceforms.

We previously showed that deletion of Fgf8b-containing spliceforms results in a similar but significantly less severe phenotype than that of Fgf8-null mutant embryos during gastrulation (Guo and Li, 2007). Furthermore, in Fgf8Δb/Δb embryos at the early somite stage, region-specific markers, such as Otx2, Gbx2, En1, and Fgf8, are normally expressed in the neuroepithelium. These results demonstrate that Fgf8a-containing spliceforms are able to partially compensate for the loss of Fgf8b, and that Fgf8b-containing spliceforms are not essential for the induction of these mes/r1 genes (Guo and Li, 2007). In this study, we show that remaining Fgf8 transcripts were lost in the mes/r1 region of Fgf8Δb/F; En1+/Cre embryos between the 11- to 14-somite stages (Fig. 2B and data not shown). Thereafter, specific deletion of Fgf8b-containing spliceforms in the mes/r1 area results in a nearly complete deletion of the dorsal midbrain and the cerebellum, similar to the phenotype resulted from specific deletion of Fgf8 in the isthmus. Since Fgf8Δb/F; En1+/Cre embryos have only one functional Fgf8a copy, it is formally possible that two copies of Fgf8a is necessary to compensate for the loss of Fgf8b. However, we found that Fgf8 expression were also lost in the isthmus in rare Fgf8Δb/Δb embryos that had survived to E9.5 (n=3/33), even though they had two copies of Fgf8a (supplementary figure S1). Therefore, our results from deletion of Fgf8b-containing spliceforms throughout the embryos or specifically in the mes/r1 area clearly demonstrate that Fgf8b-containing spliceforms are absolutely necessary to maintain the expression of Fgf8 in the isthmus, and the remaining Fgf8a-containing spliceforms are not sufficient to compensate for the loss of Fgf8b-containing spliceforms.

Given the potent activity of Fgf8a in promoting the growth and differentiation of the midbrain (Lee et al., 1997; Liu et al., 1999; Sato et al., 2001), it is surprising that the midbrain develops normally in the absence of Fgf8a-containing spliceforms (Fig. 5). Our examination has not revealed any discernable abnormalities in the morphology, histology, and cellular specification and differentiation of the midbrain and the cerebellum of Fgf8Δa/Δa and En1+/Cre; Fgf8Δa/F conditional mutant mice. In addition, we found that mice double homozygous for Fgf8Δa and Fgf17-null mutations display the same MHB phenotype as Fgf17-null mice (Fig. 6). These genetic data demonstrate that Fgf17 does not compensate for the loss of Fgf8a in Fgf8Δa/Δa and Fgf8a-CKO mice, though overexpression of Fgf8a or Fgf17 produces a similar phenotype in the mes/r1 area (Liu et al., 2003; Olsen et al., 2006; Xu et al., 2000). Our results demonstrate that Fgf8a-containing spliceforms are largely dispensable for the development of the midbrain and cerebellum.

How do Fgf8b-containing spliceforms achieve the overall activity of Fgf8 in directing development of the midbrain and cerebellum? It has been proposed that the differentiation of the midbrain and cerebellum is controlled by distinct levels of FGF signaling. Although ectopic expression of Fgf8b transforms the mes into a cerebellar fate, low-level expression of Fgf8b results in a similar phenotype as ectopic expression of Fgf8a or Fgf17 (Liu et al., 2003; Sato et al., 2001). Furthermore, a genetic study in which Spry2, an inhibitor of FGF signaling, was misexpressed in the mes/r1 region provide evidence that the regionalization of the embryonic midbrain and cerebellum is regulated by different levels of FGF signaling (Basson et al., 2008). Multiple inhibitors of FGF signaling, including Sprouty proteins (Minowada et al., 1999), Sef (Lin et al., 2002), Flrt1/3 (Haines et al., 2006), and Mpk3 (Klock and Herrmann, 2002), are expressed in the mes/r1 region in vertebrates. Significantly, Spry1 and Spry2 are induced by Fgf8b but not Fgf8a in mes/r1 cells in chick (Liu et al., 2003). Therefore, graded Fgf8b signaling shaped by feedback inhibition or graded Fgf8b protein concentration established by diffusion from the Fgf8-expressing cells in the isthmus may be sufficient to control the growth and patterning of the mes/r1 region. It is worth noting that multiple Fgf8 spliceforms, Fgf8b, d, f and h, are deleted in Fgf8Δb/Δb mutants, and these spliceforms are still expressed in Fgf8Δa/Δa mice. Fgf8d, f and h, whose activity has not been examined in mes/r1 cells, may interact with Fgf8b to constitute the overall function of Fgf8 and contribute to compensate for the loss of Fgf8a in Fgf8Δa/Δa mice.

Deletion of Fgf8a-containing spliceforms, results in growth retardation and partial postnatal lethality (Fig. 3). Our preliminary results also indicate that these mutant mice have reduced fertility (data not shown). Interestingly, all these phenotypes are more severe on the C57BL/6J than the CD1 outbred background. The precise cause of the mutant phenotype and strain-dependent phenotypic variation remains to be determined. Nevertheless, our findings indicate that Fgf8a-containing spliceforms may have a functional role in Fgf8-expressing tissues outside the mes/r1. Abnormalities in the alternative splicing of FGF8 affecting the production of FGF8a or FGF8e (there are no equivalents to Fgf8c and Fgf8g in humans) may contribute to growth retardation or neonatal lethality in humans.

Knockdown experiments using morpholino oligonucleotides have revealed that Fgf8a plays an important role in posterior neural specification in X. laevis and X. tropicalis embryos (Fletcher et al., 2006). However, no discernable abnormality in the central and peripheral nervous system was detected by marker analysis and neurofilament immunohistochemistry in Fgf8Δa/Δa embryos at E8.5 and E10.5 (Fig. 4K, L and data not shown). Therefore, loss of Fgf8a does not produce obvious defects in posterior neural specification in mice. The discrepancy in the mutant phenotype due to the loss of Fgf8a in mice and Xenopus might result from the different experimental approaches or a real difference between mouse and Xenopus Fgf8 isoform function.

Supplementary Material

01. Supplementary Figure S1. Specific loss of Fgf8 expression at the mes/1 border in Fgf8b-deficient embryos.

(A–B) In situ hybridization of Fgf8 transcripts in wild-type (A) and Fgf8Δb/Δb (B) embryos at E9.5. Asterisk indicates the absence of Fgf8 expression in the isthmus. Abbreviations: bc, branchial arches; cp/fn, commissural plate/frontonasal prominence; fl, forelimb bud; inf, infundibulum; is, isthmus; so, somite; tb, tail bud; zli, zona limitans intrathalamica.

Acknowledgments

We are grateful to Drs. Sohyun Ahn, Alexandra Joyner, and Mark Zervas for critical reading of the manuscript. We thank Dr. David Ornitz for providing the Fgf17 mutant mice, Drs. Gail Martin and Alexandra Joyner for providing the Fgf8F and En1Cre mice, and the Gene Targeting and Transgenic Facility (GTTF) at UConn Health Center for the generation of Fgf8Δa mutant mice. We thank Drs. Richard Behringer and Alexandra Joyner for providing probes for RNA in situ hybridization. J. Li is supported by grants from the NIH and March of Dimes foundation.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Basson MA, Echevarria D, Ahn CP, Sudarov A, Joyner AL, Mason IJ, Martinez S, Martin GR. Specific regions within the embryonic midbrain and cerebellum require different levels of FGF signaling during development. Development. 2008;135:889–98. doi: 10.1242/dev.011569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Cecconi F, Proetzel G, Alvarez-Bolado G, Jay D, Gruss P. Expression of Meis2, a Knotted-related murine homeobox gene, indicates a role in the differentiation of the forebrain and the somitic mesoderm. Dev Dyn. 1997;210:184–90. doi: 10.1002/(SICI)1097-0177(199710)210:2<184::AID-AJA10>3.0.CO;2-E. [DOI] [PubMed] [Google Scholar]
  3. Chi CL, Martinez S, Wurst W, Martin GR. The isthmic organizer signal FGF8 is required for cell survival in the prospective midbrain and cerebellum. Development. 2003;130:2633–2644. doi: 10.1242/dev.00487. [DOI] [PubMed] [Google Scholar]
  4. Crossley PH, Martin GR. The mouse Fgf8 gene encodes a family of polypeptides and is expressed in regions that direct outgrowth and patterning in the developing embryo. Development. 1995;121:439–51. doi: 10.1242/dev.121.2.439. [DOI] [PubMed] [Google Scholar]
  5. Crossley PH, Martinez S, Martin GR. Midbrain development induced by FGF8 in the chick embryo. Nature. 1996;380:66–8. doi: 10.1038/380066a0. [DOI] [PubMed] [Google Scholar]
  6. Fletcher RB, Baker JC, Harland RM. FGF8 spliceforms mediate early mesoderm and posterior neural tissue formation in Xenopus. Development. 2006;133:1703–14. doi: 10.1242/dev.02342. [DOI] [PubMed] [Google Scholar]
  7. Frank DU, Fotheringham LK, Brewer JA, Muglia LJ, Tristani-Firouzi M, Capecchi MR, Moon AM. An Fgf8 mouse mutant phenocopies human 22q11 deletion syndrome. Development. 2002;129:4591–603. doi: 10.1242/dev.129.19.4591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Garel S, Huffman KJ, Rubenstein JL. Molecular regionalization of the neocortex is disrupted in Fgf8 hypomorphic mutants. Development. 2003;130:1903–14. doi: 10.1242/dev.00416. [DOI] [PubMed] [Google Scholar]
  9. Gogoi RN, Schubert FR, Martinez-Barbera JP, Acampora D, Simeone A, Lumsden A. The paired-type homeobox gene Dmbx1 marks the midbrain and pretectum. Mech Dev. 2002;114:213–7. doi: 10.1016/s0925-4773(02)00067-9. [DOI] [PubMed] [Google Scholar]
  10. Guo Q, Li JY. Distinct functions of the major Fgf8 spliceform, Fgf8b, before and during mouse gastrulation. Development. 2007;134:2251–60. doi: 10.1242/dev.004929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Haines BP, Wheldon LM, Summerbell D, Heath JK, Rigby PW. Regulated expression of FLRT genes implies a functional role in the regulation of FGF signalling during mouse development. Dev Biol. 2006;297:14–25. doi: 10.1016/j.ydbio.2006.04.004. [DOI] [PubMed] [Google Scholar]
  12. Hong CS, Park BY, Saint-Jeannet JP. Fgf8a induces neural crest indirectly through the activation of Wnt8 in the paraxial mesoderm. Development. 2008;135:3903–10. doi: 10.1242/dev.026229. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Inouye M, Oda SI. Strain-specific variations in the folial pattern of the mouse cerebellum. J Comp Neurol. 1980;190:357–62. doi: 10.1002/cne.901900209. [DOI] [PubMed] [Google Scholar]
  14. Itoh N, Ornitz DM. Functional evolutionary history of the mouse Fgf gene family. Dev Dyn. 2008;237:18–27. doi: 10.1002/dvdy.21388. [DOI] [PubMed] [Google Scholar]
  15. Kimmel RA, Turnbull DH, Blanquet V, Wurst W, Loomis CA, Joyner AL. Two lineage boundaries coordinate vertebrate apical ectodermal ridge formation. Genes Dev. 2000;14:1377–89. [PMC free article] [PubMed] [Google Scholar]
  16. Klock A, Herrmann BG. Cloning and expression of the mouse dual-specificity mitogen-activated protein (MAP) kinase phosphatase Mkp3 during mouse embryogenesis. Mech Dev. 2002;116:243–7. doi: 10.1016/s0925-4773(02)00153-3. [DOI] [PubMed] [Google Scholar]
  17. Lee SM, Danielian PS, Fritzsch B, McMahon AP. Evidence that FGF8 signalling from the midbrain-hindbrain junction regulates growth and polarity in the developing midbrain. Development. 1997;124:959–69. doi: 10.1242/dev.124.5.959. [DOI] [PubMed] [Google Scholar]
  18. Lewandoski M, Sun X, Martin GR. Fgf8 signalling from the AER is essential for normal limb development. Nat Genet. 2000;26:460–3. doi: 10.1038/82609. [DOI] [PubMed] [Google Scholar]
  19. Li JY, Lao Z, Joyner AL. Changing requirements for Gbx2 in development of the cerebellum and maintenance of the mid/hindbrain organizer. Neuron. 2002;36:31–43. doi: 10.1016/s0896-6273(02)00935-2. [DOI] [PubMed] [Google Scholar]
  20. Lin W, Furthauer M, Thisse B, Thisse C, Jing N, Ang SL. Cloning of the mouse Sef gene and comparative analysis of its expression with Fgf8 and Spry2 during embryogenesis. Mech Dev. 2002;113:163–8. doi: 10.1016/s0925-4773(02)00018-7. [DOI] [PubMed] [Google Scholar]
  21. Liu A, Li JY, Bromleigh C, Lao Z, Niswander LA, Joyner AL. FGF17b and FGF18 have different midbrain regulatory properties from FGF8b or activated FGF receptors. Development. 2003;130:6175–85. doi: 10.1242/dev.00845. [DOI] [PubMed] [Google Scholar]
  22. Liu A, Losos K, Joyner AL. FGF8 can activate Gbx2 and transform regions of the rostral mouse brain into a hindbrain fate. Development. 1999;126:4827–38. doi: 10.1242/dev.126.21.4827. [DOI] [PubMed] [Google Scholar]
  23. MacArthur CA, Lawshe A, Xu J, Santos-Ocampo S, Heikinheimo M, Chellaiah AT, Ornitz DM. FGF-8 isoforms activate receptor splice forms that are expressed in mesenchymal regions of mouse development. Development. 1995;121:3603–13. doi: 10.1242/dev.121.11.3603. [DOI] [PubMed] [Google Scholar]
  24. Martinez S, Crossley PH, Cobos I, Rubenstein JL, Martin GR. FGF8 induces formation of an ectopic isthmic organizer and isthmocerebellar development via a repressive effect on Otx2 expression. Development. 1999;126:1189–200. doi: 10.1242/dev.126.6.1189. [DOI] [PubMed] [Google Scholar]
  25. Meyers EN, Lewandoski M, Martin GR. An Fgf8 mutant allelic series generated by Cre- and Flp-mediated recombination. Nat Genet. 1998;18:136–41. doi: 10.1038/ng0298-136. [DOI] [PubMed] [Google Scholar]
  26. Minowada G, Jarvis LA, Chi CL, Neubuser A, Sun X, Hacohen N, Krasnow MA, Martin GR. Vertebrate Sprouty genes are induced by FGF signaling and can cause chondrodysplasia when overexpressed. Development. 1999;126:4465–75. doi: 10.1242/dev.126.20.4465. [DOI] [PubMed] [Google Scholar]
  27. Modrek B, Lee C. A genomic view of alternative splicing. Nat Genet. 2002;30:13–9. doi: 10.1038/ng0102-13. [DOI] [PubMed] [Google Scholar]
  28. Moon AM, Capecchi MR. Fgf8 is required for outgrowth and patterning of the limbs. Nat Genet. 2000;26:455–9. doi: 10.1038/82601. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Nagy A, Gertsenstein M, Vintersten K, Behringer R. Manipulating the Mouse Embryo. Cold Spring Harbor Laboratory Press; Cold Spring Harbor, New York: 2003. [Google Scholar]
  30. Ohtoshi A, Nishijima I, Justice MJ, Behringer RR. Dmbx1, a novel evolutionarily conserved paired-like homeobox gene expressed in the brain of mouse embryos. Mech Dev. 2002;110:241–4. doi: 10.1016/s0925-4773(01)00587-1. [DOI] [PubMed] [Google Scholar]
  31. Olsen SK, Li JY, Bromleigh C, Eliseenkova AV, Ibrahimi OA, Lao Z, Zhang F, Linhardt RJ, Joyner AL, Mohammadi M. Structural basis by which alternative splicing modulates the organizer activity of FGF8 in the brain. Genes Dev. 2006;20:185–198. doi: 10.1101/gad.1365406. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Oulad-Abdelghani M, Chazaud C, Bouillet P, Sapin V, Chambon P, Dolle P. Meis2, a novel mouse Pbx-related homeobox gene induced by retinoic acid during differentiation of P19 embryonal carcinoma cells. Dev Dyn. 1997;210:173–83. doi: 10.1002/(SICI)1097-0177(199710)210:2<173::AID-AJA9>3.0.CO;2-D. [DOI] [PubMed] [Google Scholar]
  33. Sato T, Araki I, Nakamura H. Inductive signal and tissue responsiveness defining the tectum and the cerebellum. Development. 2001;128:2461–9. doi: 10.1242/dev.128.13.2461. [DOI] [PubMed] [Google Scholar]
  34. Sato T, Joyner AL, Nakamura H. How does Fgf signaling from the isthmic organizer induce midbrain and cerebellum development? Dev Growth Differ. 2004;46:487–94. doi: 10.1111/j.1440-169x.2004.00769.x. [DOI] [PubMed] [Google Scholar]
  35. Sato T, Nakamura H. The Fgf8 signal causes cerebellar differentiation by activating the Ras-ERK signaling pathway. Development. 2004;131:4275–85. doi: 10.1242/dev.01281. [DOI] [PubMed] [Google Scholar]
  36. Soriano P. Generalized lacZ expression with the ROSA26 Cre reporter strain [letter] Nat Genet. 1999;21:70–1. doi: 10.1038/5007. [DOI] [PubMed] [Google Scholar]
  37. Storm EE, Rubenstein JL, Martin GR. Dosage of Fgf8 determines whether cell survival is positively or negatively regulated in the developing forebrain. Proc Natl Acad Sci U S A. 2003;100:1757–62. doi: 10.1073/pnas.0337736100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Sun X, Mariani FV, Martin GR. Functions of FGF signalling from the apical ectodermal ridge in limb development. Nature. 2002;418:501–8. doi: 10.1038/nature00902. [DOI] [PubMed] [Google Scholar]
  39. Sun X, Meyers EN, Lewandoski M, Martin GR. Targeted disruption of Fgf8 causes failure of cell migration in the gastrulating mouse embryo. Genes Dev. 1999;13:1834–46. doi: 10.1101/gad.13.14.1834. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Trumpp A, Depew MJ, Rubenstein JL, Bishop JM, Martin GR. Cre-mediated gene inactivation demonstrates that FGF8 is required for cell survival and patterning of the first branchial arch. Genes Dev. 1999;13:3136–48. doi: 10.1101/gad.13.23.3136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Wurst W, Bally-Cuif L. Neural plate patterning: upstream and downstream of the isthmic organizer. Nat Rev Neurosci. 2001;2:99–108. doi: 10.1038/35053516. [DOI] [PubMed] [Google Scholar]
  42. Xu J, Liu Z, Ornitz DM. Temporal and spatial gradients of Fgf8 and Fgf17 regulate proliferation and differentiation of midline cerebellar structures. Development. 2000;127:1833–43. doi: 10.1242/dev.127.9.1833. [DOI] [PubMed] [Google Scholar]
  43. Zervas M, Millet S, Ahn S, Joyner AL. Cell behaviors and genetic lineages of the mesencephalon and rhombomere 1. Neuron. 2004;43:345–57. doi: 10.1016/j.neuron.2004.07.010. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01. Supplementary Figure S1. Specific loss of Fgf8 expression at the mes/1 border in Fgf8b-deficient embryos.

(A–B) In situ hybridization of Fgf8 transcripts in wild-type (A) and Fgf8Δb/Δb (B) embryos at E9.5. Asterisk indicates the absence of Fgf8 expression in the isthmus. Abbreviations: bc, branchial arches; cp/fn, commissural plate/frontonasal prominence; fl, forelimb bud; inf, infundibulum; is, isthmus; so, somite; tb, tail bud; zli, zona limitans intrathalamica.

NIHMS164281-supplement-01.pdf (95.9KB, pdf)

RESOURCES