Molecular Regulators of Innervation and Patterning in the Developing Chicken Inner Ear
Normal hearing and balance relies on the detection of sound, orientation and acceleration by sensory hair cells (HCs) located in the inner ear. Once sound is detected, that information must be transmitted to the brain by sensory neurons. Damage to the HCs and/or neurons in the auditory or vestibular organs of the inner ear can result in hearing loss or balance disorders. In mammals, these disorders can be permanent, as HCs do not regenerate after damage. While hearing aids and cochlear implants can restore some ability to hear, there are currently no molecular therapies for hearing loss. By examining genes involved in HC development and innervation, basic science can identify candidate genes for potential molecular therapies. This dissertation focuses on molecular regulators involved in establishing and/or maintaining innervation in the chicken inner ear during embryonic development.
The basilar papilla (BP) is the auditory sensory organ in the chicken and is homologous to the mammalian organ of Corti (oC). The BP houses two types of sensory HCs – tall HCs and short HCs. On the neural side of the BP, tall HC receive primarily afferent innervation (neural-side identity). On the abneural side, short HC receive primarily efferent innervation (abneural-side identity). The patterning of these two identities along the radial axis is dependent upon the precise spatiotemporal expression of certain genes during embryonic development. One such gene is Wingless/integrated (Wnt)9a.
Previous work has shown that Wnt9ais expressed on the neural edge of the BP and is likely secreted in a gradient across the prosensory domain during crucial time points when proliferation, differentiation, and innervation are occurring. When Wnt9a was overexpressed, we observed an increase in the width of the BP as well as an expansion of the neural-side identity, likely at the expense of the abneural-side identity. RNA sequencing of Wnt9a-overexpressing and control BPs identified genes involved in the Wnt signaling pathway, cytoskeletal remodeling, and axon guidance signaling that were differentially expressed. This dissertation focuses on axon guidance genes, specifically those involved in Slit/Robo (Roundabout), Contactin (Cntn), and Semaphorin (Sema) signaling, that were differentially expressed in this RNA sequencing data set.
Slits typically act as repulsive cues for neurites expressing Robo receptors. RNA sequencing data indicates that Slit2transcripts increased by 1.2 fold when Wnt9a was overexpressed. When examining Slit2 spatial expression pattern in Wnt9a-overexpressing BPs, we did not observe an upregulation of Slit2 but rather an expansion of the Slit2-expression domain that is likely due to increased proliferation in response to Wnt9a. To better understand the role of Slit/Robo signaling in the developing BP, we examined the radial expression patterns of Slit2, Robo1, and Robo2. Slit2 is expressed on the anterior and posterior walls of the cochlear duct (CD). Robo1and Robo2 had graded expression in the prosensory domain of the BP, highest on the abneural side. Robo1is also present in the auditory ganglion. While only a small population of cochleovestibular ganglion neurites have been previously shown to respond to Slits, Slit-Robo has also been shown to activate TCF transcription factor by non-canonically activating β-catenin through Abl kinase. We examined Abl kinase-activated b-catenin in Slit2-and Wnt9a-overexpressing BPs but did not observe a change in phosphorylated b-catenin. We also overexpressed a dominant-negative Robo1. In some dominant-negative Robo1 overexpressing ears, we observed a reduction in ganglion size; however, this affect did not reliably replicate. These data suggests that Slit-Robo signaling could be involved in neuroblast delamination and/or migration.
RNA sequencing results indicate that Contactin 6Cntn6 transcripts increased by 1.5 fold when Wnt9a was overexpressed. Contactins are cell adhesion molecules that have been previously shown to impact neurite outgrowth and innervation. In the auditory field, clinical studies have also shown that patients diagnosed with autism who also have mutations in Cntn5 and Cntn6 are more likely to exhibit increased sensitivity to sound. Based on RNA sequencing in the embryonic day (E)6 chicken ear, Cntn6 has low levels of expression in controls. We attempted to examine the spatial expression of Cntn6 but found that in situ hybridization is not sensitive enough to detect low levels of Cntn6 in control or Wnt9a-overexpressing BPs.
Class III Semaphorinsecreted ligands are known to repel neurites expressing Neuropilin (Nrp) and/or Plexin (Plxn) receptors. Sema3D and Nrp2 were downregulated in the presence of exogenous Wnt9a; however, the spatial expression of these transcripts didnot support their role in establishing or maintaining radial innervation patterns. There is, however, a growing body of literature supporting that Sema signaling also has alternative roles in development such as synaptogenesis, boundary formation, and vasculogenesis. To evaluate these options during inner ear development, we used in situ hybridization or immunohistochemistry to map the expression of Sema3D, Sema3F, Nrp1, Nrp2, and PlxnA1 in the chicken inner ear from E5 to E10. The resulting expression patterns in either the otic epithelium or its surrounding mesenchyme suggest that Sema signaling could be involved in each of the varied functions reported for other tissues. Sema3Dexpression flanking the sensory tissue in vestibular organs suggests that it may repel Nrp2- and PlxnA1-expressing neurites of the vestibular ganglion away from nonsensory epithelia, thus channeling them into the sensory domains at E5-E8. Expression of Sema signaling genes in the sensory hair cells of both the auditory and vestibular organs on E8–E10 may implicate Sema signaling in synaptogenesis. In the nonsensory regions of the cochlea, Sema3Din the future tegmentum vasculosum opposes Nrp1 and PlxnA1 in the future cuboidal cells; the abutment of ligand and receptors in adjacent domains may enforce or maintain the boundary between them. In the mesenchyme, Nrp1 colocalized with capillary-rich tissue. Sema3D immediately flanks this Nrp1-expressing tissue, suggesting a role in endothelial cell migration towards the inner ear. In summary, Sema signaling may play multiple roles in the developing inner ear.
To better understand innervation patterns in the avian BP, we also examined the developing efferent innervation patterns from E11 to E17 using NeuroVue lipophilic tracer dye. Our data suggest that efferents have already begun to penetrate the sensory epithelium at E11 and that efferents arrive to the ipsilateral BP earlier than the contralateral BP. By E12, many efferents appear to send back branches out to short HCs. At E15, many efferents appear to have reached the abneural edge of the BP, are innervating the hyaline cells, and are projecting apically.
In summary, this work suggests that Slit and Sema signaling are not involved in establishing radial innervation patterns but may have alternative roles in inner ear development. Additionally, while efferents appear to arrive to the ipsilateral BP sooner than the contralateral BP, both ears send projections across the radial axis and back branch around the same time.