Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Mar 10.
Published in final edited form as: Dev Biol. 2024 Sep 25;517:157–167. doi: 10.1016/j.ydbio.2024.09.009

Pax2a, Pax5 and Cdh1-β-catenin, but not Wnt, protect sensory hair cells from destabilizing effects of fgf signaling on cell adhesion

Whitney Roberson 1, Jorden N Holland 1, Bruce B Riley 1,*
PMCID: PMC12970481  NIHMSID: NIHMS2145453  PMID: 39332789

Abstract

During inner ear development, specification of sensory epithelia requires dynamic regulation of Fgf signaling. In zebrafish, high levels of Fgf are necessary and sufficient to specify the utricular/vestibular macula, whereas the saccular/auditory macula requires a discreet lower level of Fgf. Transcription factors Pax2a and Pax5 act downstream of Fgf to help specify utricular identity, loss of which leads to sporadic extrusion of hair cells from the utricular macula. The mechanism for utricular instability is not clear but is potentially related to reduced expression of cdh1/Ecad caused by disruption of pax2a. Here we find that utricular hair cells in pax2−/− and pax5−/− mutants gradually lose adhesive contact with the macula, leading to ejection of intact hair cells from either the basal or apical surface. The phenotype is far more severe in pax2a−/− mutants and is progressive, resulting in loss of large swaths of the utricular hair cells by 82 hpf. Instability is caused by elevated Fgf signaling in the utricle, as modest reduction of Fgf signaling with a low dose of SU5402 prevents hair cell loss in pax2a−/− mutants. Misexpression of cdh1/Ecad in pax2a−/− mutants partially rescues pax2a−/− mutants. Elevating β-catenin levels by treatment with BIO, or misexpression of a mutant form of β-catenin lacking transcriptional activity but retaining cell adhesion function, fully rescues pax2a−/− mutants. In contrast, Wnt signaling is not required for utricular stability. Thus, Pax2/5 factors serve to counteract the destabilizing effects of elevated Fgf signaling needed to specify utricular identity.

Keywords: Sensory epithelia, Utricular macula, Saccular macula, β-catenin, Fgf-MAPK, Wnt

1. Introduction

The vertebrate inner ear is a highly conserved organ system containing sensory epithelia that mediate the senses of balance (vestibular function) or hearing. In zebrafish embryos, the otic vesicle, the embryonic precursor of the inner ear, already possesses two incipient sensory epithelia, or maculae, at 24 hpf. The anterior/utricular macula is essential for vestibular function and the posterior/saccular macula is the primary endorgan for hearing in zebrafish (Riley and Moorman, 2000; Kwak et al., 2006; Schuck and Smith, 2009; Breitzler et al., 2020; Lau and Vasconcelos, 2023). Inverse anteroposterior gradients of Fgf and Hh appear to provide critical symmetry-breaking signals required for diversification of nascent sensory maculae. Specifically, high Fgf levels are necessary and sufficient for anterior/utricular specification whereas low Fgf and high Hh are essential for posterior/saccular specification (Sapède and Pujades, 2010; Hammond et al., 2003; Hammond and Whitfield, 2011; Hartwell et al., 2019). Genes activated downstream of these signals include pax5 in the utricular macula and pou3f3b in the saccular macula (Tan et al., 2023). Although the role of pou3f3b remains unknown, pax5 is critical for proper maintenance and function of the utricular macula (Kwak et al., 2006). Knockdown of pax5 causes sporadic loss of utricular hair cells and disrupts vestibular function, whereas saccular development and auditory function remain normal. Pax5 is a member of the Pax2/5/8 subfamily of transcription factors and likely reinforces the function of Pax2a in the utricular macula. pax2a is initially expressed in all otic cells in response to Fgf signaling during otic induction (Phillips et al., 2001; Léger and Brand, 2002; Maroon et al., 2002; Liu et al., 2003). As the otic vesicle forms, pax2a is restricted to the medial wall, including developing sensory maculae (Tan et al., 2022). In addition, elevated Fgf signaling in the utricular macula leads to pronounced upregulation and accumulation of Pax2a protein in utricular hair cell nuclei (Riley et al., 1999). The overlap of Pax2a plus elevated Fgf subsequently induces expression of pax5 in the nascent utricular primordium (Kwak et al., 2006; Tan et al., 2022). In pax2a−/− mutants, pax5 is not expressed and utricular hair cells are sporadically lost, similar to the phenotype caused by pax5 knockdown (Pfeffer et al., 1998; Kwak et al., 2006). The mechanism of hair cell loss remains unknown, but hair cell loss is not observed in fgf3−/− mutants, which also fail to express pax5 (Kwak et al., 2006). Thus, the requirement for pax5 is context dependent and suggests the possibility that pax2a and pax5 are required to protect hair cells from the destabilizing effects of elevated Fgf signaling. Although Fgf is critical for ongoing sensory development (Millimaki et al., 2007), activation of receptor tyrosine kinases (RTKs) can have a general destabilizing effect on epithelia by directly phosphorylating β-catenin, displacing it from junctional complexes and causing loss of cell adhesion (Fang et al., 2007; Lee et al., 2010; Krejci et al., 2012; Weng et al., 2019; reviewed by Lilien and Balsamo, 2005).

Here we have examined the time course and nature of hair cell loss in zebrafish pax2a−/− mutants and tested various mechanisms that influence hair cell stability. We confirm that partially blocking Fgf signaling stabilizes hair cells in pax2a−/− mutants, as does misexpression of cdh1/E-cadherin or a truncated form of β-catenin that retains cell adhesion functions but lacks all transcriptional activity. In contrast, Wnt signaling is not required for hair cell maintenance. These findings support the hypothesis that pax2a is required to protect hair cells exposed to elevated Fgf by reinforcing cell adhesion.

2. Materials & methods

2.1. Fish strains and developmental conditions

Wild-type fish were derived from the AB line (Eugene). Embryos were incubated at 28.5 °C in fish water containing methylene blue and staged as previously described (Kimmel et al., 1995). PTU (1-phenyl 2-thiourea, 0.3 mg/ml, Sigma P-7629) was added to fish water at 24 hpf to prevent melanin formation. pax2aTu29a is thought to be a null allele (Brand et al., 1996; Lun and Brand, 1998), and homozygotes are referred to herein as pax2a−/− mutants. Transgenic line TG(pou4f3: GAP-GFP)s356t (Xiao et al., 2005) was used to visualize mature hair cells and is referred to herein as brn3c:gfp. Transgenic line TG(hsp70:dkk1)w32 (Stoick-Cooper et al., 2007) was used to misexpress Wnt antagonist Dkk1. Transgenic lines TG(hsp70:chd1)x73 and TG(hsp70:β-catenin-dm)x80, and mutant line pax5x84, were generated herein (described below).

2.2. pax5 targeting and genotyping

To generate a pax5 mutant allele, we targeted exon3 with TALENs with the following sequences: Left arm CATCAGGCCAGGGGTAAT, right arm AAAGGTTGCTACACC. Wild-type embryos were injected with TALENs at the one-cell stage and raised to adulthood. Outcross progeny were screened by PCR to detect loss of a BamHI site in exon3. Pooled DNA samples were PCR-amplified using forward primer 5′-CTTCTTGACTGTGAACTGAGC-3′, and reverse primer 5′-CTGTTGATTGAGCTGACGC-3’. Subsequent digestion with BamHI cuts the wild-type amplicon into 257 bp and 151 bp fragments, whereas mutant indels that alter the BamHI site prevent cutting. Upon recovery of mutant allele pax5x84, sequencing revealed a 4 base deletion near the beginning of exon 3, leading to a frameshift and premature stop at the end of exon 4.

2.3. Transgene production and genotyping

To misexpress Cdh1/Ecad and a mutant form of β-catenin, we used the Tol2 retrotransposon kit (Addgene) (Kwan et al., 2007; Villefranc et al., 2007) to generate heat shock inducible transgenic lines, TG(hsp70: chd1)x73 and TG(hsp70:β-catenin-dm)x80. Wild-type embryos were injected at the one-cell stage with Tol2 vectors plus transposase mRNA. After reaching adulthood, transgenic founders were screened by outcrossing to wild-type fish. Outcross progeny were heat shocked at 24 hpf, fixed at 25 hpf, and stained by wholemount in situ hybridization using antisense riboprobe for transgenic mRNA. This revealed germline transmission frequencies and ensured robust transgene activation. Founders with high germline transmission were outcrossed to establish stable transgenic lines.

For PCR genotyping, forward primer corresponding to hsp70 promoter sequence was used for all heat shock-inducible transgenes: 5′-GACGAGGTGTTTATTCGCTCT-3’. Transgene-specific reverse primers were as follows: TG(hsp70:β-catenin-dm) 5′-CATCTTCTTCCTCAGGGTTGC-3’; TG(hsp70:chd1) 5′-CCACTCGAAAAGACCTGAAAAAG-3’; TG(hsp70:dkk1) 5′- GGAATGCAAACACCATTGCTGC-3’.

2.4. Gene misexpression and drug treatments

Heat shock-inducible transgenes were activated by incubating embryos in a water bath at 39 °C for 1 h. For drug treatments, embryos were incubated in 1% DMSO with 5 μM SU5402 (Sigma SML0443), 5 μM BIO (Sigma B1686), or 10 μM IWR-1 (Sigma I0161) beginning at 48 hpf.

2.5. In situ hybridization, immunolocalization and cell death assays

Wholemount in situ hybridization was performed as previously described (Jowett and Yan, 1996; Riley et al., 1999; Phillips et al., 2001). Immunolocalization was performed using primary monoclonal antibodies for Cdh1 (GeneTex no. GTX125890, 1:250) and Otoferlin (HCS-1, Developmental Studies Hybridoma Bank, 1:100), followed by secondary monoclonal antibody Alexa Fluor 546 goat anti-mouse IgG (Invitrogen A-11003), 1:100). To stain dying cells, dechorionated embryos were incubated for 1 h on agarose-coated plates containing fish water with 1 μg/ml acridine orange (Invitrogen A1301), washed briefly and immediately viewed by fluorescence microscopy.

2.6. Hair cell quantitation

The total numbers of hair cells was counted from Photoshop images of the apical surface of brn3c:Gfp + sensory epithelia. Utricular and saccular maculae were fused in a variable fraction of pax2a−/− mutants, depending upon the cross. In experiments in which most pax2a−/− mutants exhibited fused maculae, we determined the total number of hair cells rather than the number of hair cells in each macula. Gaps in the regular pattern of hair cells were readily identified by visual inspection and outlined using the lasso marque tool in Photoshop. The number of pixels contained within the outlined area (shown in the Histogram display) was mathematically converted to μm2 based on image resolution (pixels per cm). Examples are illustrated in Supplemental Fig. S1.

2.7. Reproducibility and statistics

All quantitative data are displayed as scatter plots showing individual data points, means and standard deviations. To evaluate statistical significance, t-tests were used when comparing two groups, or one-way ANOVA with Tukey post-hoc HSD for experiments with three or more groups. All data points represent one ear per embryo.

3. Results

3.1. Hair cell loss in pax2a−/− mutants

To compare sensory development in wild-type and pax2a−/− embryos, we used a transgenic line, brn3c:gfp, to track hair cell accumulation through 82 hpf, the latest stage through which all mutant embryos survive (Fig. 1A–J). At 82 hpf in wild-type embryos, there are just over 40 hair cells each in the utricular and saccular maculae (Fig. 1M–O). In pax2a−/− mutants, reduced delta gene expression causes a modest lateral inhibition defect, leading to over-production of hair cells (Riley et al., 1999; Tan et al., 2022). However, pax2a−/− mutants exhibit sporadic loss of hair cells in the utricular macula beginning around 36 hpf (not shown). Notable gaps appear in the pattern of utricular hair cells by 72 hpf, and the gaps continue to expand over time (Fig. 1F–H). By 82 hpf, the mean area of the largest gap in utricular hair cells expands to over 700 μm2 in pax2a−/− mutants, compared to only ~50 μm2 in wild-type embryos (p < .0001, Fig. 1N). The utricular maculae is often fused at its posterior edge with the saccular maculae in a variable percentage (50%–90%) of pax2a−/− mutants, making it difficult to unambiguously determine the number of hair cells in each macula. However, in a typical cross in which half of pax2a−/− mutants exhibited well delineated maculae, the mean number of utricular hair cells at 82 hpf was reduced by 35%, a highly significant difference (p < .0001, Fig. 1M). Additionally, the morphology of the utricular macula becomes highly variable and disordered in pax2a−/− mutants, unlike the highly reproducible morphology seen in wild-type embryos (Fig. 1K and L). In contrast to the utricle, the mean number of saccular hair cells in pax2a−/− mutants was not significantly different from wild-type embryos (p = .216, Fig. 1O). The pattern of saccular hair cells is less organized in pax2a−/− mutants (Fig. 1I and J), which in some specimens appear as gaps in the anterior region of the saccular macula, but the mean gap size is not significantly different from control embryos (p = .09, Fig. 1P). Thus, hair cell loss is mostly localized to the utricular macula in pax2a−/− mutants.

Fig. 1. Utricular instability in pax2a−/− mutants.

Fig. 1.

Open in a new tab

(A–J) Dorsal lateral views of utricular and saccular maculae in wild-type embryos and pax2a−/− mutants at the indicated stages. (K–L) Dorsolateral views of utricular maculae in multiple wild-type embryos (K) and pax2a−/− mutants (L) at 82 hpf, highlighting marked variation in the latter. (P–Q) Scatter plots showing the mean and standard deviation of the number of utricular hair cells (P) and area of the largest gap between utricular hair cells (Q) in wild-type embryos and pax2a−/− mutants at 82 hpf. (R–S) Scatter plots showing the mean and standard deviation of the number of saccular hair cells (R) and area of the largest gap between saccular hair cells (S) in wild-type embryos and pax2a−/− mutants at 82 hpf. Asterisks indicate significant differences. Scale bars in (A) and (K)), 40 μm.

To clarify the nature of hair cell loss, we tracked hair cell morphology in live pax2a−/− mutants over time. This confirmed that the first signs of utricular hair cell loss were seen at 36 hpf, and the frequency of such events peaked between 48 and 72 hpf, affecting 34% (24/71) of pax2a−/− embryos at any time during this period. In the majority (~75%) of embryos exhibiting hair cell loss, hair cells appeared to be extruded through the basal side of the sensory epithelium (Fig. 2A and B). These hair cells often appeared stretched and distorted, with their basolateral surfaces extending beneath the sensory epithelium while retaining apical attachments, and often persisted in this state for several hours before losing apical attachment and passing beneath the basal surface. In 25% of specimens, hair cells were observed to be ejected into the lumen of the otic vesicle (Fig. 2C). In addition, most mutant embryos exhibited small fluorescent globules in the lumen, likely debris from fragmented hair cells (not shown). To examine whether hair cell extrusion involves loss of cell adhesion, embryos were stained for E-cadherin (Cdh1) at 48 hpf. In wild-type embryos, Cadherin protein is maximally expressed at the apical ends of hair cells, with a lower level of expression along the basolateral surface (Fig. 2D). However, pax2a−/− mutants show almost no detectable Cdh1 staining in the utricle at 48 hpf (Fig. 2E), likely contributing to hair cell extrusion. As hair cells are extruded, they continue to express Otoferlin and brn3c:Gfp, and maintain general apical-basal/epithelial morphology (Fig. 2F and G). Thus, extruded hair cells do not dedifferentiate, nor do they transition to a mesenchymal morphology. Additionally, hair cells do no exhibit typical morphological signs of apoptosis (rounding and blebbing) as they are extruded. Nevertheless, to examine whether hair cell extrusion involves cell death, embryos were stained with acridine orange (AO) at 48 hpf. No wild-type embryos exhibited AO staining in the utricle at 48 hpf (Fig. 2H). Similarly, the majority (51/55) of pax2a−/− mutants were unlabeled (Fig. 2I), although a small minority (4/55) of mutants exhibited AO-stained cells in the utricular macula (Fig. 2J). Thus, the low incidence of cell death cannot account for the majority of hair cell extrusion. Taken together, these data suggest that hair cell extrusion primarily reflects progressive loss of adhesion, beginning with the basolateral surface and culminating with loss of apical attachments. It is unclear why only some hair cells are lost, but macular expression of several other cdh genes remains unaltered in pax2a−/− mutants (Kantarci et al., 2016), possibly providing variable compensation for loss of Cdh1 expression.

Fig. 2. Hair cell loss in pax2a−/− mutants.

Fig. 2.

Open in a new tab

(A, B) Lateral views of the utricular macula in a live pax2a−/− mutant at 52 hpf and 57.5 hpf showing hair cells being ejected from the basal side of the macula (white arrows), one of which persists for the duration of observation. The ratio of pax2a−/− embryos showing such basal extrusion is shown in the lower left corner (A). The otolith (otl) is indicated. (C) Lateral view of the utricular macula in a live pax2a−/− mutant at 52 hpf showing a pair of hair cells that were ejected into the lumen (white arrow) apparently still attached to the otolith (otl), and the ratio of pax2a−/− mutants exhibiting such apical extrusion is shown in the lower left corner. (D, E) Lateral views of the utricular macula showing expression of Cdh1 protein and brn3c:Gfp in a wild-type embryo (D) and pax2a−/− mutant (E). Positions of otoliths (otl) and the apical (ap) and basolateral membranes (blm) of hair cells are indicated. (F, G) Lateral view of utricular hair cells in pax2a−/− mutants stained with anti-Otoferlin antibody (left) and brn3c:Gfp (right). Hair cells being ejected from the basal side of the macula are marked (white arrows). (H–L) Lateral view of the utricular macula in live embryos at 48 hpf stained with acridine orange. The ratio of embryos showing the depicted phenotype is shown in the lower left corner. Apical and basal surfaces of the utricle (utr) are outlined and otoliths are marked (otl). The white arrow in (J) marks 2 dying cells in the utricular macula of a pax2a−/− mutant. Scale bars in (C), (D), (F), (G) and (J), 40 μm.

3.2. Reducing fgf signaling stabilizes sensory epithelia

High RTK-MAPK activity can destabilize epithelia by directly phosphorylating β-catenin, causing dissociation from junctional complexes and subsequent loss of cell adhesion (Fang et al., 2007; Lee et al., 2010; Krejci et al., 2012; Weng et al., 2019; reviewed by Lilien and Balsamo, 2005). Such a destabilizing effect could place a significant burden on the utricular macula, which experiences much higher Fgf signaling than does the saccular macula. We therefore hypothesized that elevated Fgf signaling could contribute directly to loss of utricular hair cells in pax2a−/− mutants, as suggested in a previous study (Kwak et al., 2006). To test this idea further, we treated embryos with a low dose of SU5402 to reduce, but not eliminate Fgf signaling (Fig. 3A–D). Treatment with 5 μM SU5402 from 48 hpf is sufficient to eliminate etv5b expression within 2 h (Supplemental Figs. S2A and B) and reduced the total number of hair cells in wild-type embryos, reflecting a requirement for Fgf in ongoing hair cell development (Fig. 3E). However, the number of hair cells was not significantly different between SU5402-treated wild-type and pax2a−/− mutant embryos (p = .218). Moreover, SU5402 treatment eliminated the hair cell loss phenotype, as no gaps appeared in the pax2a−/− mutant maculae (Fig. 3F). Thus, pax2a is required to stabilize utricular hair cells only when Fgf signaling remains high.

Fig. 3. Rescue of pax2a−/− mutants by reducing Fgf signaling.

Fig. 3.

Open in a new tab

(A–D) Dorsolateral views of utricular AND SACCULAR maculae at 82 hpf in a wild-type embryo (A) and pax2a−/− mutant (B) treated with 1% DMSO, and a wild-type embryo (C) and pax2a−/− mutant (D) treated with 5 μm su5402 from 48 hpf. (E–F) Scatter plots showing the mean and standard deviation of total hair cells (E) and area of the largest gap between utricular hair cells (F) in wild-type and pax2a−/− mutant embryos at 82 hpf. Asterisks indicate significant differences. Scale bar in (A), 40 μm.

3.3. Misexpression of Cdh1 partially rescues pax2a−/− mutants

We showed previously that expression levels of cdh1/Ecad are reduced in the otic vesicle in pax2a−/− mutants (Kantarci et al., 2016), which we reasoned could also contribute to hair cell loss in the utricular macula. We therefore tested whether misexpression of cdh1 from a heat shock inducible transgene (Supplemental Figs. S3A and B) could prevent hair cell loss in pax2a−/− mutants. Embryos were heat shocked for 1 h at 39 °C, which elevates transgenic protein levels for 6–12 h, depending on stability of the protein (Sweet et al., 2011; Vemaraju et al., 2012; Kantarci et al., 2016; 2020). Heat shocking hs:cdh1 every 12 h was not sufficient to alter the pax2a−/− mutant phenotype (not shown). We therefore activated hs:cdh1 every 8-to-10 h (at 48, 56, 62, and 72 hpf) to help maintain transgenic Cdh1 protein levels (Fig. 4A–D). This regimen increased the number of hair cells in both wild-type and pax2a−/− mutant embryos. However, the number of hair cells in pax2a−/−; hs: cdh1 embryos was significantly lower than in hs:cdh1+ embryos (p < .0001, Fig. 4E). The mean size of the hair cell gap in pax2a−/− mutants did not change appreciably following activation of hs:cdh1, but the range of data increased dramatically (Fig. 4F). Notably, over 40% (5/12) of pax2a−/−; hs:cdh1 embryos exhibited only small gaps that were comparable to wild-type and hs:cdh1 embryos, in contrast to large gaps in non-transgenic pax2a−/− mutants that fell far beyond the normal range. These data suggest that misexpression of cdh1 provides partial rescue in some pax2a−/− mutants.

Fig. 4. Partial rescue of pax2a−/− mutants by misexpressing cdh1.

Fig. 4.

Open in a new tab

(A–D) Dorsolateral views of utricular and saccular maculae in wild-type (A), pax2a−/− mutant (B), hs:cdh1(C) and pax2a−/− mutant with hs:cdh1 at 82 hpf. Embryos were serially heat shocked for 1 h, 39 °C, at 48, 56, 62, and 72 hpf, before fixing at 82 hpf. (E–F) Scatter plots showing the mean and standard deviation of total hair cells (E) and area of the largest gap between utricular hair cells (F) in wild-type and pax2a−/− mutant embryos at 82 hpf. Asterisks indicate significant differences. Scale bar in (A), 40 μm.

3.4. Elevation of β-catenin/cell adhesion rescues pax2a−/− mutants

To affect cell adhesion in another way, we elevated steady state levels of β-catenin by treating embryos with BIO, an inhibitor of GSK3β (Fig. 5A–D). pax2a−/− mutants treated with BIO from 48 hpf produce significantly more hair cells than wild-type embryos treated with BIO (Fig. 5G, p = .027), possibly because of retention of the excess hair cells produced by lateral inhibition defects in pax2a−/− mutants (Riley et al., 1999; Tan et al., 2022). Additionally, the gap in utricular hair cells was eliminated in all pax2a−/− mutants treated with BIO (difference compared to wild-type treated with BIO p = .85, Fig. 5C, D, H). These results could reflect the ability of β-catenin to reinforce junctional complexes or, alternatively, it could reflect a transcriptional response to elevated Wnt signaling. To distinguish between these possibilities, we misexpressed β-catenin-dm (Supplemental Fig. S3), a truncated mutant form of β-catenin lacking transcriptional activity but retaining full cell adhesion function (Valenta et al., 2011). For this experiment, embryos were heat shocked twice, at 48 hpf and 60 hpf (Fig. 5E and F). With this regimen, activation of hs:βcat-dm had no effect on wild-type embryos (Fig. 5E–I, J). In pax2a−/− mutants, however, activation of hs:βcat-dm increased the mean number of hair cells to wild-type levels (Fig. 5I) and eliminated gaps in the utricular macula in all specimens (Fig. 5F–J). These data show that augmenting the presumptive role of β-catenin in cell adhesion is sufficient to prevent hair cell loss in pax2a−/− mutants.

Fig. 5.

Fig. 5.

Open in a new tab

Rescue of pax2a−/− mutants by modulating β-catenin. (A–F) Dorsolateral views of utricular and saccular maculae in wild-type (A) and pax2a−/− mutant (B) treated with 1% DMSO), wild-type (C) and pax2a−/− mutant (D) treated with 5 μM BIO, hs:βcat-dm (E) and pax2a−/− with hs:βcat-dm (F) at 82 hpf. Drug treatments (A–D) were intiated at 48 hpf. Transgenic embryos (E, F) were serially heat shocked for 1 h, 39 °C, at 48 and 60 hpf before fixing at 82 hpf. (G–H) Scatter plots showing the mean and standard deviation of total hair cells (G) and area of the largest gap between utricular hair cells (H) in wild-type and pax2a−/− mutant embryos at 82 hpf. Embryos were treated with DMSO or BIO as indicated along the bottom. (I–J) Scatter plots showing the mean and standard deviation of the number of utricular hair cells (I) and area of the largest gap between utricular hair cells (J) in embryos with the indicated genotypes heat shocked at 48 and 60 hpf before fixing at 82 hpf. Asterisks indicate significant differences. Scale bar in (A), 40 μm.

3.5. Blocking Wnt signaling does not affect utricular hair cell stability

To directly address whether Wnt signaling is required for hair cell maintenance, we performed heat shock experiments to misexpress the secreted Wnt antagonist Dkk1 (Fig. 6A–D, I–L). Misexpression of dkk1 effectively blocked expression of the Wnt-target gene axin2 (Supplemental Figs. S2C and D) but had no effect on utricular or saccular maculae in wild-type embryos (Fig. 6A–C, I, K, Q–T). Similarly, activation of hs:dkk1 in pax2a−/− mutant embryos did not significantly alter the mean number of utricular hair cells (p = .09) or the mean gap size in the utricular macula relative to non-transgenic pax2a−/− mutants (p = .08) (Fig. 6B–D, Q, R). However, activation of hs:dkk1 did lead to formation of significant loss of hair cells in the saccular macula in pax2a−/− mutants (Fig. 6L–S, T). Similar results were obtained by treating embryos with an inhibitor of canonical Wnt signaling, IWR-1, beginning at 48 hpf. IWR-1 treatment strongly reduced axin2 expression (Supplemental Figs. S2E and F) but had no effect on wild-type embryos (Fig. 6E–G, M, O, U–X) nor did it alter utricular deficits in pax2a−/− mutant embryos (Fig. 6F–H, U, V). However, IWR-1 did cause significant loss of hair cells in the saccular macula in pax2a−/− mutants (Fig. 6N–P, W, X). Thus, Wnt signaling is not required for hair cell accumulation or maintenance in wild-type embryos, nor is it required in the utricular macula of pax2a−/− mutants. However, Wnt signaling appears to be necessary for maintenance of saccular hair cells specifically in pax2a−/− mutants, the implications of which are discussed below (see Discussion).

Fig. 6. Impact of blocking Wnt signaling.

Fig. 6.

Open in a new tab

(A–P) Sensory epithelia at 82 hpf in the utricle (A–H) and saccule (I–P) under conditions and backgrounds indicated across the top. (Q–T) Scatter plots showing the mean and standard deviation of the total number of utricular and saccular hair cells (Q, S) and the area of the largest gap between utricular and saccular hair cells (R, T) at 82 hpf in heat shocked embryos with the indicated genotypes. (U–X) Scatter plots showing the mean and standard deviation of the total number of utricular and saccular hair cells (U, W) and the area of the largest gap between utricular and saccular hair cells (V, W) at 82 hpf. For heat shock experiments, embryos were serially heat shocked for 1 h, 39 °C, at 48, 56, 62 and 72 hpf. Treatment with 1% DMSO and/or 10 μM IWR-1 was initiated at 48 hpf. Asterisks indicate significant differences; ns, not significant. Scale bar in (A), 40 μm.

3.6. Loss of pax5 causes transient loss of hair cells in the utricle

We previously reported that morpholino knockdown of pax5 leads to sporadic loss of utricular hair cells at 48 hpf (Kwak et al., 2006). To evaluate pax5 function more thoroughly and at later stages, we generated a knockout allele, pax5x84, characterized by 4-base deletion near the beginning of exon 3, causing a frameshift and a premature stop at the end of exon 4 (Fig. 7A). This truncation is predicted to eliminate the transactivation domain and most of the DNA binding domain, likely a strong hypomorph or null allele. pax5x84/x84 homozygotes, hereafter referred to as pax5−/− mutants, show no gross morphological defects but die around 10 dpf. Monitoring brn3c:gfp + hair cells between 48 and 72 hpf revealed that 38% (10/26) of pax5−/− mutants exhibit loss of utricular hair cells (Fig. 7I), consistent with the phenotype seen in pax5 morphants (Kwak et al., 2006). Hair cells appeared to be extruded from the basal side of the sensory epithelium and were distorted as in pax2a−/− mutants. About 10% (4/39) of pax5 ± heterozygotes also showed hair cells extrusion from the basal surface (Fig. 7G). These data suggest that Pax5 contributes to stabilization of the utricular macula in a dosage-dependent manner. However, despite their early hair cell loss, pax5−/− and pax5 ± embryos produce a normal number of utricular and saccular hair cells at 82 hpf, and no large gaps appear in the pattern of utricular hair cells (Fig. 7B–D, J, K). Not surprisingly, saccular development was also normal in pax5 mutants at 82 hpf (Fig. 7L). Thus, hair cell loss in pax5-deficient embryos is neither progressive nor permanent, possibly reflecting ongoing regeneration and repair processes that are inoperative or insufficient in pax2a−/− mutants.

Fig. 7. Moderate utricular instability in pax5−/− mutants.

Fig. 7.

Open in a new tab

(A) A TALEN-induced deletion in exon 3 of pax5 causes a frameshift that eliminates the transactivation domain and most of the DNA-binding domain of Pax5. (B–D) Dorsolateral views of the utricular macula at 82 hpf in embryos with the indicated genotypes. Macular integrity appears normal in pax5-deficient embryos. (E–I) Lateral views of the utricular macula at 48 hpf in embryos with the indicated genotypes. Basally-extruded hair cells (white arrows) are observed in pax5+/(G) and pax5−/− mutants (I). Ratios of embryos exhibiting the represented phenotypes are shown in the lower left corner of each panel. (J–L) Scatter plots showing the mean and standard deviation in the number of hair cells in the utricle (J) and saccule (L), and the area of the largest gap between utricular hair cells in wild-type and pax5-deficient embryos at 82 hpf. No significant differences were detected. Scale bars in (B) and (E), 40 μm.

4. Discussion

Our findings support a model in which pax2a and pax5 provide two coordinated functions downstream of Fgf (Fig. 8). First, pax2a helps promote sensory specification (Millimaki et al., 2007) and, together with pax5, specifies utricular identity (Kwak et al., 2006; Tan et al., 2023). Second, pax2a and pax5 cooperate to stabilize epithelial integrity in the utricular macula. Shoring up stability is likely critical in the utricle because high Fgf-MAPK signaling can destabilize epithelial integrity through direct phosphorylation of β-catenin, leading to dissociation of junctional complexes (Fang et al., 2007; Lee et al., 2010; Krejci et al., 2012; Weng et al., 2019; reviewed by Lilien and Balsamo, 2005). Accordingly, both pax2a−/− and pax5−/− mutants exhibit shedding of utricular hair cells. The phenotype is far more severe in pax2a−/− mutants, likely because pax5 is not expressed in this background (Pfeffer et al., 1998; Kwak et al., 2006; Tan et al., 2023), compounding functional impairment. Loss of utricular hair cells in pax2a−/− mutants occurs progressively starting at around 36 hpf, resulting in formation of large gaps in the pattern of hair cells by 82 hpf. Hair cells in pax2a−/− mutants retain their general apical-basal/epithelial shape as they are extruded and do not show morphological signs of apoptosis or take up acridine orange, showing that they are not being extruded because they are dead or dying. Instead, extruding hair cells appear to lose adhesion first along the basolateral surface, retaining apical attachments for longer periods. Such hair cells often become stretched and distorted as they are extruded, eventually losing apical attachments and passing beneath the sensory epithelium. The near absence of Cdh1 in the utricle in pax2a−/− mutants further supports defective cell adhesion. Partially blocking Fgf signaling with a low dose of SU5402 prevents hair cell loss in pax2a−/− mutants. Reinforcing adhesion by misexpression of cdh1 partially rescues hair cell loss, and elevating β-catenin by treating embryos with BIO fully rescues hair cells in pax2a−/− mutants. It is likely that misexpression of cdh1 provides less efficient rescue because, without a corresponding increase in β-catenin, Cadherin is not trafficked properly from the ER (Chen et al., 1999; Wahl et al., 2003; Curtis et al., 2008). Finally, misexpression of β-catenin-dm also efficiently rescues hair cells in pax2a−/− mutants. This modified β-catenin lacks transcriptional activity but retains full activity in junctional complexes (Valenta et al., 2011), further supporting an important role for cell adhesion. The fate of extruded hair cells remains uncertain, but they appear to be relatively short-lived since they do not accumulate appreciably after leaving the sensory epithelium. Such cells could be cleared by sudden/rapid apoptosis, or they could be engulfed by phagocytes and digested.

Fig. 8. Model and summary of the roles of Pax2a and Pax5.

Fig. 8.

Open in a new tab

Elevated Fgf signaling is necessary and sufficient for specification of utricular identity and acts partly through induction of pax2a and pax5. However, Fgf-Ras-MAPK signaling potentially destabilizes epithelial structure by directly phosphorylating β-catenin, causing dissociation of junctional complexes. Pax2a is critical for maintaining proper expression of cdh1/Ecad in otic cells, a function likely shared by Pax5, helping to offset the destabilizing effects of Fgf. Experimentally elevating expression of cdh1/Ecad or β-catenin restores hair cell stability to pax2a−/− mutants.

Despite the importance of β-catenin for hair cell maintenance, Wnt signaling is not normally required. Blocking Wnt signaling by misexpressing dkk1 or by treating embryos with IWR-1does not alter hair cell stability in wild-type embryos, nor does it enhance loss of utricular hair cells in pax2a−/− mutants. On the other hand, blocking Wnt substantially destabilizes the saccular macula in pax2a−/− mutants. It is possible that blocking Wnt reduces the level of available β-catenin, exacerbating the effects of reduced cdh1/Ecadherin expression seen in pax2a−/− mutants (Kantarci et al., 2016). This could substantially weaken cell adhesion in the saccular macula, whereas the utricular macula is already so compromised in pax2a−/− mutants that reducing β-catenin causes little additional damage.

Pax2/5/8 factors appear to serve a broad role in promoting cell adhesion in specialized epithelia. Pax8 directly binds the promoter of NCAM and strongly activates its transcription (Holst et al., 1994), and Pax5 directly activates transcription of Cdh1/Ecad and Integrin α5/Itga5 (Benzina et al., 2024). During development of the otic placode, otic cells converge from broader territory and coalesce to form a thickened epithelium (Bhat and Riley, 2011; Maulding et al., 2014). In zebrafish embryos lacking pax2a, pax2b and pax8, otic cells initially coalesce into a small placode but later disaggregate and move apart, reflecting partial loss of adhesion (Mackereth et al., 2005). Similarly, knockdown of Pax2 in chick embryos impairs invagination of the otic placode by reducing expression of adhesion molecules Cdh2/N-cadherin and NCAM (Christophorou et al., 2010). In the developing pronephros, Pax2 and Pax8 are required to promote epithelialization of mesenchyme to form the pronephric duct (Bouchard et al., 2010). Cultures of nephric duct tissue from Pax2 mutant mice exhibit progressive structural disorganization, including dynamic cell rearrangements and partial loss of apicobasal polarity of individual cells, with some cells converting to mesenchymal morphology and migrating away (Soofi et al., 2012). During eye development, closure of the optic fissure involves dramatic shifts in cell adhesion leading to local epithelial destabilization followed by restabilization (Hardy and Ringer, 2023). This process requires Pax2 and proper regulation of Fgf, which affect cell fate patterning and requisite cell adhesion. Accordingly, disruption of Fgf, Pax2, Cdh2/Ncad or α-catenin all lead to coloboma, a persistent gap in the retinal epithelium (Sanyanusin et al., 1995; Torres et al., 1996; Macdonald et al., 1997; Masai et al., 2003; Nakayama et al., 2008; Chen et al., 2012, 2013; Cai et al., 2013; Lusk and Kwan, 2021). Thus, Fgf and Pax2/5/8 factors appear to act together in diverse organ systems to coregulate cell fate specification and enhanced cell adhesion.

Supplementary Material

Suppl. Fig. S1
Suppl. Fig. S2
Suppl. Fig. S3

FUNDING

This work was supported by the National Institutes of Health - NIDCD grant R01-DC03806.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.ydbio.2024.09.009.

Footnotes

CRediT authorship contribution statement

Whitney Roberson: Writing – original draft, Investigation, Formal analysis, Data curation. Jorden N. Holland: Investigation, Formal analysis. Bruce B. Riley: Writing – review & editing, Supervision, Project administration, Methodology, Funding acquisition, Formal analysis, Conceptualization.

Data availability

Data will be made available on request.

References

  1. Benzina S, Beauregard A-P, Guerrette R, Jean S, Faye MD, Laflamme M, Maicas E, Crapoulet N, Ouellette RJ, Robichaud GA, 2024. Pax-5 is a potent regulatore of E-cadherin and breast cancer malignant processes. Oncotarget 8 (7), 12052–12066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bhat N, Riley BB, 2011. Integrin-α5 coordinates assembly of posterior cranial placodes in zebrafish and enhances Fgf-dependent regulation of otic/epibranchial cells. PLoS One 6 (12), e27778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bouchard M, de Caprona D, Busslinger M, Xu P, Fritzsch B, 2010. Pax2 and Pax8 cooperate in mouse inner ear morphogenesis and innervation. BMC Dev. Biol. 10, 89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Brand M, Heisenberg C-P, Jiang Y-J, Beuchle D, Lun K, Furutani-Seiki M, Granato M, Haffter P, Hammerschmidt M, Kane D, Kelsh R, Mullins M, Odenthal J, van Eeden FJM, Nüsslein-Valhard C, 1996. Mutations in zebrafish affecting the formation of the boundary between midbrain and hindbrain. Development 123, 179–190. [DOI] [PubMed] [Google Scholar]
  5. Breitzler L, Lau IH, Fonesca PJ, Vasconcelos RO, 2020. Noise-induced hearing loss in zebrafish: investgating structural and functional inner ear damage and recover. Hearing Res 391, 107592. [DOI] [PubMed] [Google Scholar]
  6. Cai Z, Tao C, Li H, Ladher R, Gotoh N, Feng G-S, Wang F, Zhang X, 2013. Deficient FGF signaling causes optic nerve dysgenesis and ocular coloboma. Development 140, 2711–2723. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chen S, Lewis B, Moran A, Xie T, 2012. Cadherin-mediated cell adhesion is critical for the closing of the mouse optic fissure. PLoS One 7 (12), e51705. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chen S, Gaudenz K, Paulson A, Guo F, Trimble R, Peak A, Seidel C, Deng C, Furuta Y, Xie T, 2013. Defective FGF signaling causes coloboma formation and disrupts retinal neurogenesis. Cell Res. 23, 254–273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen Y-T, Stewart DB, Nelson WJ, 1999. Coupling assembly of the E-cadherin/β-catenin complex to the efficient endoplasmic reticulum exit and basal-lateral membrane targeting of E-cadherin in polarized MDCK cells. J. Cell Biol. 144 (4), 687–699. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Christophorou NAD, Mende M, Lleras-Forero L, Grocott T, Streit A, 2010. Pax2 coordinates epithelial morphogenesis and cell fate in the inner ear. Dev. Biol. 345, 180–190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Curtis MW, Johnson KR, Wheelock MJ, 2008. E-cadherin/catenin complexes are formed contranslationally in the endoplasmic reticulum/Golgi compartments. Cell comm. And Adhesion 15, 365–378. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fang D, Hawke D, Zheng Y, Xia Y, Meisenhelder J, Nika H, Mills GB, Kobayashi R, Hunter T, Lu Z, 2007. Phosphorylation of β-catenin by AKT promotes β-catenin transcriptional activity. J. Biol. Chem. 282 (15), 11221–11229. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hammond KL, Loynes HE, Folarin AA, Smith J, Whitfield TT, 2003. Hedgehog signaling is required for correct anteroposterior patterning of the zebrafish otic vesicle. Development 130, 1403–1417. [DOI] [PubMed] [Google Scholar]
  14. Hammond KL, Whitfield TT, 2011. Fgf and Hh signalling act on a symmetrical pre-pattern to specify anterior and posterior identity in the zebrafish otic placode and vesicle. Development 138, 3977–3987. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hardy H, Ringer J, 2023. Cell adhesion marker expression dynamics during fusion of the optic fissure. Gene Expr. Patterns 50, 119344. [DOI] [PubMed] [Google Scholar]
  16. Hartwell RD, England SJ, Monk NA, van Hateren NJ, Baxendale S, Marzo M, Lewis KE, Whitfield TT, 2019. Anteroposterior patterning of the zebrafish ear through Fgf- and Hh-dependent regulation of hmx3a expression. PLoS Genet. 15 (4), e1008051. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Holst BD, Goomer RS, Wood IC, Edelman GM, Jones FS, 1994. Binding and activation of the promoter for the neural cell adhesion molecule by Pax-8. J. Biol. Chem. 269 (35), 22245–22252. [PubMed] [Google Scholar]
  18. Kantarci H, Gerberding A, Riley BB, 2016. Spemann organizer gene Goosecoid promotes delamination of neuroblasts from the otic vesicle. Proc. Natl. Acad. Sci (USA) 113 (34), E6840–E6848. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kimmel CB, Ballard WM, Kimmel SR, Ullman B, Schilling TF, 1995. Stages of embryonic development of the zebrafish. Dev. Dynam. 203, 253–310. [DOI] [PubMed] [Google Scholar]
  20. Krejci P, Aklian A, Kaucka M, Sevcikova E, Prochazkova J, Masek JK, Mikolka P, Pospisilova T, Spoustova T, Weis M, Paznekas WA, Wolf JH, Gutkind JS, Wilcox WR, Kozubik A, Jabs EW, Bryja V, Salazar L, Vesela I, Balek L, 2012. /Receptor tyrosine kinases activate canonical WNT/β-catenin signaling via MAP kinase/LRP6 pathway and direct β-catenin phosphorylation. PLoS One 7 (4), e35826. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kwak S-J, Vemaraju S, Moorman SJ, Zeddies D, Popper AN, Riley BB, 2006. Zebrafish pax5 regulates development of the utricular macula and vestibular function. Dev. Dynam. 235, 3026–3038. [DOI] [PubMed] [Google Scholar]
  22. Kwan KM, Fujimoto E, Grabher C, Mangum BD, Hardy ME, Campbell DS, Parant JM, Yost JF, Kanki JP, Chien C-B, 2007. The Tol2kit: a multisite Gateway-based construction kit for Tol2 transposon transgenesis constructs. Dev. Dynam. 236, 3088–3099. [DOI] [PubMed] [Google Scholar]
  23. Lau IH, Vasconcelos RO, 2023. Noise-induced damage in the zebrafish inner ear endorgans: evidence for higher acoustic sensitivity of saccular and lagenar hair cells. J. Exp. Biol. 226, jeb245992. [DOI] [PubMed] [Google Scholar]
  24. Lee C-H, Hung H-W, Hung P-H, Shieh Y-S, 2010. Epidermal growth factor receptor regulates β-catenin location, stability, and transcriptional activity in oral cancer. Mol. Cancer 9, 64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Léger S, Brand M, 2002. Fgf8 and Fgf3 are required for zebrafish ear placode induction, maintenance and inner ear patterning. Mech. Dev. 119, 91–108. [DOI] [PubMed] [Google Scholar]
  26. Lilien J, Balsamo J, 2005. The regulation of cadherin-mediated adhesion by tyrosine phosphorylation/dephosphorylation of β-catenin. Curr. Op. Cell Biol 17, 459–465. [DOI] [PubMed] [Google Scholar]
  27. Liu D, Chu H, Maves L, Yan Y-L, Morcos PA, Postlethwait JH, Westerfield M, 2003. Fgf3 and Fgf8 dependent and independent transcription factors are required for otic placode specification. Development 130, 2213–2224. [DOI] [PubMed] [Google Scholar]
  28. Lun K, Brand M, 1998. A series of no isthmus (noi) alleles of the zebrafish pax2.1 gene reveals multiple signaling events in development of the midbrain-hindbrain boundary. Development 125, 3049–3082. [DOI] [PubMed] [Google Scholar]
  29. Lusk S, Kwan KM, 2021. Pax2a, but not pax2b, influences cell survival and periocular mesenchyme localization to facilitate zebrafish optic fissure closure. Dev. Dyn. 251 (4), 625–644. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Macdonald R, Scholes J, Strähle U, Brennan C, Holder N, Brand M, Wilson SW, 1997. The Pax protein Noi is required for commissural axon pathway formation in the rostral forebrain. Development 124, 2397–2408. [DOI] [PubMed] [Google Scholar]
  31. Mackereth MD, Kwak S-J, Fritz A, Riley BB, 2005. Zebrafish pax8 is required for otic placode induction and plays a redundant role with Pax2 genes in the maintenance of the otic placode. Development 132, 371–382. [DOI] [PubMed] [Google Scholar]
  32. Maroon H, Walshe J, Mahmood R, Kiefer P, Dickson C, Mason I, 2002. Fgf3 and Fgf8 are required together for formation of the otic placode and vesicle. Development 129, 2099–2108. [DOI] [PubMed] [Google Scholar]
  33. Masai I, Lele Z, Yamaguchi M, Komori A, Nakata A, Nishiwaki Y, Wada H, Tanaka H, Nojima Y, Matthias H, Wilson SW, Okamoto H, 2003. N-cadherin mediates retinal lamination, maintenance of forebrain compartments and patterning of retinal neurites. Development 130, 2479–2494. [DOI] [PubMed] [Google Scholar]
  34. Maulding K, Padanad MS, Dong J, Riley BB, 2014. Mesodermal Fgf10b cooperates with other Fibroblast Growth Factors during induction of the otic and epibranchial placodes in zebrafish. Dev. Dynam. 243, 1275–1285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Millimaki BB, Sweet EM, Dhason MS, Riley BB, 2007. Zebrafish atoh1 genes: classic proneural activity in the inner ear and regulation by Fgf and Notch. Development 134, 295–305. [DOI] [PubMed] [Google Scholar]
  36. Nakayama Y, Miyake A, Nakagawa Y, Mido T, Yoshikawa M, Konishi M, Itoh N, 2008. Fgf19 is required for zebrafish lens and retina development. Dev. Biol. 313, 752–766. [DOI] [PubMed] [Google Scholar]
  37. Pfeffer PL, Gerster T, Lun K, Brand M, Busslinger M, 1998. Characterization of three novel members of the zebrafish Pax2/5/8 family: dependency of Pax5 and Pax8 expression on the Pax2.1 (noi) function. Development 125, 3063–3074. [DOI] [PubMed] [Google Scholar]
  38. Phillips BT, Bolding K, Riley BB, 2001. Zebrafish fgf3 and fgf8 encode redundant functions required for otic placode induction. Dev. Biol. 235, 351–365. [DOI] [PubMed] [Google Scholar]
  39. Riley BB, Moorman SJ, 2000. Development of utricular otoliths, but not saccular otoliths, is necessary for vestibular function and survival. J. Neurobiol. 43, 329–337. [DOI] [PubMed] [Google Scholar]
  40. Riley BB, Chiang M-Y, Farmer L, Heck R, 1999. The delta gene of zebrafish mediates lateral inhibition of hair cells in the inner ear and is regulated by pax2.1. Development 126, 5669–5678. [DOI] [PubMed] [Google Scholar]
  41. Sanyanusin P, Schimmenti LA, McNoe LA, Pierpont MEM, Sullivan MJ, Dobyns WB, Eccles MR, 1995. Mutation of the PAX2 gene in a family with optic nerve colobomas, renal anomalies and vesicoureteral reflux. Nature 9, 358–364. [DOI] [PubMed] [Google Scholar]
  42. Sapède D, Pujades C, 2010. Hedgehog signaling governs the development of otic sensory epithelium and its associated innervation in zebrafish. J. Neurosci. 30 (10), 3612–3623. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Schuck JB, Smith ME, 2009. Cell proliferation follows acoustically-induced hair cell bundle loss in the zebrafish saccule. Hearing Res 253, 67, 67. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Soofi A, Levitan I, Dressler GR, 2012. Two novel EGFP insertion alleles reveal unique aspects of Pax2 function in embryonic and adult kidneys. Dev. Biol. 365, 241–250. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Stoick-Cooper CL, Weidinger G, Riehle KJ, Hubbert C, Major MB, Fausto N, Moon RT, 2007. Distinct Wnt signaling pathways have opposing roles in appendage regeneration. Development 134, 479–489. [DOI] [PubMed] [Google Scholar]
  46. Sweet EM, Vemaraju S, Riley BB, 2011. Sox2 and Fgf interact with Atoh1 to promote sensory competence throughout the zebrafish inner ear. Dev. Biol. 258, 113–121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Tan AL, Mohanty S, Guo J, Lekven AC, Riley BB, 2022. Pax2a, Sp5a and Sp5l act downstream of Fgf and Wnt to coordinate sensory-neural patterning in the inner ear. Dev. Biol. 492, 135–153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Tan AL, Christensen SE, Baker AK, Riley BB, 2023. Fgf, Hh, and pax2a differentially regulate expression of pax5 and pou3f3b in vestbular and audidory maculae in the zebrafish otic vesicle. Dev. Dynam. 252, 1269–1279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Torres M, Gómez-Pardo E, Gruss P, 1996. Pax2 contributes to inner ear patterning and optic nerve trajectory. Development 122, 3381–3391. [DOI] [PubMed] [Google Scholar]
  50. Valenta T, Gay M, Steiner S, Drganova K, Zemke M, Hoffmans R, Cinelli P, Aguet M, Sommer Basler, K., 2011. Probing transcription-specific outputs of β-catenin in vivo. Genes Dev. 25, 2631–2643. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Vemaraju S, Kantarci H, Padanad MS, Riley BB, 2012. A spatial and temporal gradient of Fgf differentially regulates distinct stages of neural development in the zebrafish inner ear. PLoS Genet. 8 (11), e1003068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Villefranc JA, Amigo J, adn Lawson ND, 2007. Gateway compatible vectors ffor analysis of gene function in the zebrafish. Dev. Dynam. 236, 3077–3087. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Wahl JK, Kim YJ, Cullen JM, Johnson KR, Wheelock MJ, 2003. N-cadherin-catenin complexes form prior to cleavage of the proregion and transport to the plasma membrane. J. Biol. Chem. 278 (19), 17269–17276. [DOI] [PubMed] [Google Scholar]
  54. Weng J, Yu L, Chen Z, Su H, Yu S, Zhang Y, Lei X, Chen L, Cui Y, Huang Q, Jiang Y, Guo X, 2019. β-catenin phosphorylation at Y645 and Y142 is crucial for high mobility group box-1 protein-induced pulmonary vascular hyperpermeability. J. Mol. Cell. Cardiology 127, 174–184. [DOI] [PubMed] [Google Scholar]
  55. Xiao T, Roeser T, Staub W, Baier W, 2005. A GFP-based genetic screen reveals mutations that disrupt the architecture of the zebrafish retinotectal projection. Development 132, 2955–2967. [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Suppl. Fig. S1
NIHMS2145453-supplement-Suppl__Fig__S1.jpg (525.7KB, jpg)
Suppl. Fig. S2
NIHMS2145453-supplement-Suppl__Fig__S2.jpg (518KB, jpg)
Suppl. Fig. S3
NIHMS2145453-supplement-Suppl__Fig__S3.jpg (648.9KB, jpg)

Data Availability Statement

Data will be made available on request.

RESOURCES