|Home | About | Journals | Submit | Contact Us | Français|
Wnt/Planar Cell Polarity (PCP) signaling is critical for proper animal development. While initially identified in Drosophila, this pathway is also essential for the proper development of vertebrates. Zebrafish mutants, defective in the Wnt/PCP pathway, frequently display defects in convergence and extension gastrulation movements and additional later abnormalities including problems with craniofacial cartilage morphogenesis. Although multiple Frizzled (Fzd) homologues, Wnt receptors, were identified in zebrafish, it is unknown which Fzd plays a role in shaping the early larvae head skeleton. In an effort to determine which Frizzleds are involved in this process, we analyzed the expression of five zebrafish frizzled homologues fzd2, 6, 7a, 7b, and 8a from 2–4 days post fertilization (dpf). During the analyzed developmental time points fzd2 and fzd6 are broadly expressed throughout the head, while the expression of fzd7a, 7b and 8a is much more restricted. Closer examination revealed that fzd7b is expressed in the neural crest and the mesodermal core of the pharyngeal arches and in the chondrocytes of newly stacked craniofacial cartilage elements. However, fzd7a is only expressed in the neural crest of the pharyngeal arches and fzd8a is expressed in the pharyngeal endoderm.
Although initially discovered in Drosophila (Gubb and Garcia-Bellido, 1982), the Wnt/Planar Cell Polarity (PCP) pathway has since been proven to be critical in establishing and maintaining epithelial and mesenchymal cell polarity in multiple organisms. In the zebrafish, Danio rerio, the PCP pathway regulates convergence and extension movements and organ development (Solnica-Krezel et al., 1996; Rauch et al., 1997; Heisenberg et al., 2000; Topczewski et al., 2001; Dale et al., 2009). Several large screens in zebrafish have identified a number of mutations that lead to craniofacial defects (Neuhauss et al., 1996; Piotrowski et al., 1996; Schilling et al., 1996). One class of these mutations produces severely shortened jaws and is referred to as the “hammerhead” class (Piotrowski et al., 1996). While these mutants retain their ability to pattern the pharyngeal arch cartilages, the individual cartilage elements within the arches are shortened. When mutated in zebrafish, two genes that participate in the Wnt/PCP pathway, glypican 4 (knypek) (Solnica-Krezel et al., 1996; Topczewski et al., 2001) and wnt5b (pipe tail) (Hammerschmidt et al., 1996; Piotrowski et al., 1996; Rauch et al., 1997), result in “hammerhead” phenotypes. In the chicken, members of the Wnt/PCP pathway are expressed within chondrogenic regions of the developing limbs (Hartmann and Tabin, 2000; Church et al., 2002) and regulate the cell polarity of the chondrocytes (Li and Dudley, 2009). Taken together, these data indicate the possibility that the Wnt/PCP pathway may control the behavior of craniofacial chondrocytes in zebrafish.
To further investigate how the Wnt/PCP pathway controls the morphogenesis of craniofacial cartilages in zebrafish, it is essential to identify potential Wnt/PCP receptors expressed during this process. Development of craniofacial cartilage takes place during several steps including migration of the cranial neural crest to pharyngeal arches, cartilage condensation, and the subsequent morphogenetic processes that shape the final cartilage elements (Schilling and Kimmel, 1997; Knight and Schilling, 2006). We sought to determine which of the twelve currently identified zebrafish Frizzled (Fzd) homologues may function as Wnt receptors to control this process. While the expression of several fzd genes has been previously shown (Kim et al., 1998; El-Messaoudi and Renucci, 2001; Sumanas et al., 2001; Thisse et al., 2001; Sumanas et al., 2002; Ungar and Calvey, 2002; Thisse and Thisse, 2004), the majority of this characterization was done in the early embryonic stages of development before the chondrocytes begin to condense. For our analysis we selected genes that have either been previously linked to Wnt/PCP signaling or have shown expression in the pharyngeal arches at the latest characterized stage. We therefore sought to characterize the expression of the fzd genes during this initial condensation and cartilage elongation and continue the analysis until 4 days post fertilization (dpf) when the formation of the larval craniofacial skeleton is mostly completed.
The initial characterization of fzd2 expression was focused on the first day of development (Sumanas et al., 2001). At 24 hours post fertilization (hpf) the expression was localized to the mesoderm within the tailbud. Thisse et al. (2001) observed expression in the tailbud mesoderm at 19–24 hpf, but also found additional expression domains within the cardiovascular system and the hypocord. In addition, 42–48 hpf embryos displayed expression in the head mesenchyme, otic vesicle, pectoral fin musculature and the pharyngeal arches 3 through 7 (Thisse et al., 2001). When fzd2 morphants were observed, Sumanas et al. (2001) found that they produce shorter embryos with kinked or undulating notocords. This phenotype resembles those observed in wnt5b (pipe tail) mutants. These observations along with the fact that Fzd2 transcripts can be found in the cultured chondrocytes of a two week-old mouse (Xu et al., 2001), led us to investigate the expression of this gene at later developmental stages.
Using in situ hybridization, we observed that fzd2 is expressed throughout the head and pharyngeal arches at 2, 3 and 4 dpf (Fig. 1A–C). Similarly, Fzd2 is expressed in an ubiquitous manner at a high level in the maxillary prominence, lateral nasal prominence and the mandibular prominence in stage-24 chicks (Geetha-Loganathan et al., 2009). In 2 and 3 dpf zebrafish, mRNA is detected within the pectoral fin bud (Fig. 1A–B) and the liver (Fig. 4I). Interestingly, although the level of expression within the pharyngeal arches is greater than other regions of the head in 3 dpf embryos (Fig. 1B), coronal sections revealed that the staining in this region was not within the cartilage but in fact was adjacent to the cartilage (Fig. 3N).
While not previously characterized in zebrafish, mice with a targeted mutation in the Fzd6 gene exhibit a whorled hair phenotype that is quite similar to the bristle and wing hair phenotypes observed in Drosophila fz mutants (Gubb and Garcia-Bellido, 1982; Vinson et al., 1989; Guo et al., 2004; Wang et al., 2006). Later studies have firmly established that Fzd6 controls the polarity of these body hairs as part of the Wnt/PCP pathway (Devenport and Fuchs, 2008). Given that Fzd6 plays a role in the mammalian Wnt/PCP pathway, we sought to determine the expression pattern of the zebrafish homologue. Our in situ results show that fzd6 is expressed throughout the head in 2, 3, and 4 dpf embryos (Fig. 1D–F). Coronal sections of 3 dpf embryos revealed a uniform staining throughout the head including the brain (data not shown). Although fzd6 expression is excluded from the craniofacial cartilage elements, it is expressed at a low level within the mesenchymal cells surrounding the cartilage (Fig. 3O). fzd6 is also expressed within the pectoral fin bud (Fig. 1D, E) and the lateral line of 2, 3, and 4 dpf embryos (Fig. 1D–F).
In Xenopus, fzd7 regulates some aspects of gastrulation movements as part of the Wnt/PCP pathway (Djiane et al., 2000). Overexpression of zebrafish fzd7a and fzd7b disrupts proper gastrulation, however rescue experiments suggest that the fzd7a and fzd7b receptors signal through different noncanonical pathways (Knowlton and Kelly, 2004). The distinctive role of both fzd7 genes is further emphasized by their different expression patterns. The expression of fzd7a was detected in the forebrain, hindbrain, midbrain, spinal cord, in the anterior part of the developing somites and within the lateral mesoderm (El-Messaoudi and Renucci, 2001) while fzd7b was expressed within the central nervous system, somatic and lateral mesoderm, lateral line, pectoral fin bud, ear and the pharyngeal arches (Sumanas et al., 2002; Ungar and Calvey, 2002).
During the studied stages, 2–4 dpf, fzd7a and fzd7b are expressed in the pharyngeal arches (Fig. 1G–L, Fig. 2A, B, D, E, G and H). Both genes are also expressed in the pectoral fin bud between 2 – 3 dpf (Fig 1G–H, J–K, Fig. 4D and E). A few major differences between these two frizzled expression patterns are that fzd7a is strongly expressed in the brain in a specific pattern at 2 dpf, whereas fzd7b is not (Fig. 2A–B). In addition, fzd7b, unlike fzd7a, is expressed within the lateral line (Fig. 1J–L, Fig. 2B, E and Q). Furthermore, fzd7a is expressed at a low level within the heart (data not shown). When examined at a higher magnification (Fig. 2) it is clear that fzd7a is expressed at the dorsal part of the hindbrain from 2 – 4 dpf (Fig. 2A, D and G) and at 3 dpf strong staining of the tectum can be detected (Fig. 2D). A superficial dorsal view of an embryo at 2 dpf clearly shows the tectum, rhombic lip and dorsal rhombomere expression of fzd7a (Fig. 2J). As development proceeds, this expression becomes localized more laterally as seen in (Fig. 2M). This magnification also revealed that fzd7a is expressed in the retina adjacent to the lens (Fig. 2P). Sectioning also revealed that fzd7a is expressed in the sensory hair cells of the ear at 55 hpf and 3 dpf (data not shown).
In Xenopus fzd8 is involved in the regulation of convergent extension movements (Wallingford et al., 2001). Guided by the observations that zebrafish fzd8a is present in the pharyngeal arches before the condensation of the chondrocytes (Kim et al., 1998; Thisse and Thisse, 2004), mouse Fzd8 transcripts are present in cultured chondrocytes from two week-old animals (Xu et al., 2001), and that the noncanonical Wnt pathway is responsible for convergence and extension in zebrafish, we sought to determine if fzd8a is expressed at the critical stages of craniofacial cartilage morphogenesis.
Whole mount in situ revealed that fzd8a is specifically expressed within the pharyngeal arches from 2 – 4 dpf (Fig. 1M–O, Fig. 2C, F, and I). During this period of development fzd8a is also expressed within the anterior and posterior lateral line (Fig. 1M–O, Fig. 2C, F, O, R, and Fig. 4K). At 2 dpf fzd8a is also expressed in the otic vesicle (Fig. 2L), in the sensory hair cells of the ear at 55 hpf and 3 dpf (data not shown), and the olfactory bulbs (Fig. 2R) at 3 dpf. In addition, fzd8a is expressed within the chondrocytes of the pectoral fin bud at 55 hpf (Fig. 4F) and 3 dpf (data not shown) and has a low level of expression in the heart at 3 dpf (data not shown).
Although the whole mount in situ analysis of fzd7a, fzd7b and fzd8a revealed that all three genes are expressed in specific patterns in the pharyngeal arches from 2 – 4 dpf, it was not clear where this staining was with respect to the cartilage elements themselves. To determine the location of this expression within the arches we first looked at the posterior pharyngeal arches at 55 hpf. At this stage the cells within the anterior craniofacial cartilage elements have stacked and the most posterior craniofacial cartilage elements are beginning to condense (Schilling and Kimmel, 1997). To label the neural crest within the pharyngeal arches, we cut sagittal sections of 55 hpf Tg(fli1a:egfp)y1/+ embryos (Fig. 3B). In this transgenic line, in addition to endothelial cells, GFP labels cranial neural crest derivatives (Lawson and Weinstein, 2002) and provides a great reference to compare our other sections with. Sagittal sections revealed that although fzd7a is expressed in the neural crest at this stage, it is excluded from the mesodermal core and the pharyngeal endoderm (Fig. 3D). This expression is similar to Fzd7 expression in the mandibular prominence of stage-24 chicks where it is ubiquitously expressed in most of the tissue but is absent from the mesodermal core (Geetha-Loganathan et al., 2009). Next, we found that fzd7b is expressed within the neural crest and the mesodermal core of the pharyngeal arches. In contrast to fzd7a and fzd7b, zebrafish fzd8a is clearly excluded from the mesodermal core and neural crest and is only present in the pharyngeal endoderm (Fig. 3E and F). The expression of fzd7a is highly similar to that of fli1a:egfp (Fig 3B and D). Interestingly, close examination of transverse sections revealed that stacking chondrocytes at 55 hpf within the symplectic cartilage express fzd7b, but not fzd7a or fzd8a (Fig. 3G–I).
Given that each of these genes had a different expression at the time of cartilage condensation, we wanted to look later in development to see if this pattern of expression was maintained when the chondrocytes had finished their initial morphogenesis movements within each cartilage element at 3 dpf. When sections of the fzd7a, fzd7b and fzd8a 3 dpf in situ hybridizations were examined, it was clear that none of the craniofacial cartilages expressed any of those genes (Fig. 3J–L). The expression that was observed within the pharyngeal and pharyngeal arches in the lateral, dorsal and ventral views of these whole mount in situ hybridizations at 3 dpf was due to the expression within the surrounding tissue of the cartilage elements (Figs. 1 and and2).2). The col2a1a probe strongly stained the palatoquadrate and Meckel’s cartilages at 3 dpf, indicating that the in situ hybridization conditions used for the experiments in this study may be suitable (Fig. 3M). Nevertheless, probe penetration of cartilage at 3 dpf is difficult and it is possible that the fzd genes may be expressed by the chondrocytes but are not detected in the sections of the whole mount in situ hybridizations. To increase probe penetration we attempted to over digest the embryos with proteinase K and collagenase type 1 prior to the whole mount in situ hybridization and performed hybridization on sections, neither of which demonstrated that fzd7a, 7b or 8a are expressed in the cartilage (data not shown).
While the lateral views of the whole mounts provided us with a general view of the different fzd expression patterns, we obtained a more complete picture of fzd7a, 7b, and 8a expression within the zebrafish brain by observing sagittal sections at 55 hpf. These sections illustrated that fzd7a is expressed within the forebrain, midbrain tegmentum, dorsal hindbrain and hypothalamus (Fig. 4A). At this stage fzd7b is also expressed within the tectum, tegmentum, thalamus and the rhombic lip (Fig. 4B). The most complex expression pattern in the brain was observed with fzd8a, and includes strong expression domains in the telencephalon, along the border between the hypothalamus and ventral thalamus as well as in the more rostal hypothalamus, hypophysis (pituitary) and in the midbrain tegmentum (Fig 4C). The expression of fzd8a within the midbrain tegmentum is particularly intensive laterally (compare to Fig. 2O). In addition to the brain, fzd7a and fzd8a are also expressed within the neural tube (Fig. 4G, J and L). Coronal sections at 55 hpf demonstrate that fzd7b is expressed in the swim bladder as well as some mesenchymal tissue surrounding the gut (Fig. 4H). More posteriorly, we found that fzd8a is expressed in the liver, gut, perivasculature and along the ventral borders of the myotomes (Fig. 4J–L).
Our analysis suggests that fzd7a, fzd7b, and fzd8a are good candidates to play the role as receptors of the Wnt/PCP pathway during craniofacial development. To test whether Wnt/PCP directly controls the expression of these receptors we conducted in situ hybridization with fzd7a, fzd7b, and fzd8a probes in gpc4 (knypek), wnt5b (pipe tail), and wnt11 (silberblick) (Piotrowski et al., 1996; Heisenberg et al., 2000) mutants and their wild-type siblings at 3 dpf (Fig. 5). Although the overall length of the mutant embryos was reduced due to abnormal convergence and extension movements, the expression patterns of the three frizzled genes appeared to be normal. This is consistent with the observation that mutations affecting Wnt/PCP impair cell movements but not tissue patterning (Topczewski et al., 2001).
Taken together, the data presented here demonstrate that fzd7a, 7b, and 8a are expressed in a tissue specific manner in the pharyngeal arches during craniofacial cartilage morphogenesis and that these patterns are not disrupted by mutations in gpc4 (knypek), wnt5b (pipe tail) or wnt11 (silberblick). Further experiments will determine what role these frizzled receptor homologues may have during craniofacial development.
AB (ZDB-GENO-960809-7), wnt5bta89 (ZDB-GENE-980526-87), gpc4m818 (ZDB-GENE-011119-1), wnt11tz216 (ZDB-GENE-990603-12) and Tg(fli1a:EGFP)y1/+ (ZDB-GENO-070209-102) (Lawson and Weinstein, 2002) zebrafish lines were maintained as described in Topczewski et al. (2001). Embryos were collected from natural spawning and staged according to morphology described in Kimmel et al. (1995).
fzd7a was originally referred to as fz7 (El-Messaoudi and Renucci, 2001). fzd7b was originally referred to as either fzd7a (Ungar and Calvey, 2002) or fzd7b (Sumanas et al., 2002). To eliminate confusion, we follow the nomenclature standards set forth on ZFIN (http://zfin.org) and refer to these genes as fzd7a (ZDB-GENE-990415-223) and fzd7b (ZDB-GENE-990415-229), respectively.
Anti-sense RNA probes were synthesized using the following cDNA clones: fzd7a (Witzel et al., 2006), fzd7b (accession number BC049397), fzd8a (accession number BC056273, ZDB-GENE-00032803), fzd2 (accession number BC056524, ZDB-GENE-990415-224), and col2a1a (Yan et al., 1995) (ZDB-GENE-980526-192). The zebrafish fzd6 homolog was identified using bioinformatics methods and the antisense RNA probe was synthesized from a cDNA clone (accession number BC065361, ZDB-GENE-040426-1451). Whole mount in situ hybridization was performed as described in (Topczewski et al., 2001). All of the probes were hybridized in 50% formamide hybridization buffer except for col2a1a, which was hybridized in 65% formamide hybridization buffer.
Following the whole mount in situ hybridizations, the embryos were postfixed with 4% paraformaldehyde, mounted in sucrose agarose blocks, and cut into 10 μm sections. The Tg(fli1a:EGFP)y1/+fish were cut into 16 μm sections and fluorescent images were collected with a 510 LSM META confocal microscope.
We would like to thank Carl-Philipp Heisenberg for kindly providing the fzd7a probe. We would also like to thank Chunyue Yin for technical advice and Rodney M. Dale and Elizabeth E. LeClair for critical comments of this manuscript. This work was supported by the National Institutes of Health - NIDCR grants: R01DE016678 to J. Topczewski and F32DE019058 to B. Sisson.
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.