Notch-mediated lateral induction is necessary to maintain vestibular prosensory identity during inner ear development

Abstract

The five vestibular organs of the inner ear derive from patches of prosensory cells that express the transcription factor SOX2 and the Notch ligand JAG1. Previous work suggests that JAG1-mediated Notch signaling is both necessary and sufficient for prosensory formation and that the separation of developing prosensory patches is regulated by LMX1a, which antagonizes Notch signaling. We used an inner ear-specific deletion of the Rbpjκ gene in which Notch signaling is progressively lost from the inner ear to show that Notch signaling, is continuously required for the maintenance of prosensory fate. Loss of Notch signaling in prosensory patches causes them to shrink and ultimately disappear. We show this loss of prosensory fate is not due to cell death, but rather to the conversion of prosensory tissue into non-sensory tissue that expresses LMX1a. Notch signaling is therefore likely to stabilize, rather than induce prosensory fate.

Keywords: Development; Inner ear; Notch; Otic; Sensory; Vestibular.

Fig. 1.. The complete loss of Notch signaling via the deletion of Rbpjκ causes severe truncations of the vestibular system.

A: Paint-filling series of Rbpjκ cKO inner ears and their WT littermates from E10.5 to E13.5. Asterisks denote truncations of the semicircular canals of the vestibular system. Scale bars = 125μm for E10.5 and 250μm for E11.5 - E13.5. B: Paint-fillings for E14.5 Notch1 cKO inner ears and WT littermate. Scale bar = 350μm C: Immunofluorescent staining for RBPJk (green) and DAPI (blue) in E9.5 Rbpjκ cKO and WT otocysts. White bracket denotes otic epithelium with RBPJk expression present, yellow bracket denotes otic epithelium without RBPJk expression. White arrows indicate RBPJk-positive staining in the neural tube of Rbpjκ cKO and WT embryos. Scale bar = 50μm.

Fig. 2.. Loss of Notch signaling results in a reduction and eventual loss of prosensory markers for the cristae and maculae of the inner ear.

A: RNA in situ hybridization for Bmp4 in Rbpjκ cKO otocyst sections from E10.5 to E11.5. Black arrows indicate Bmp4-expressing cristae prosensory domains, black arrowheads indicate Bmp4-expressing cochlear prosensory domains, red arrows indicate Bmp4 reduction or absence in Rbpjκ cKO cristae prosensory domains. B: RNA in situ hybridization for Lfng in Rbpjκ cKO otocyst sections from E9.5 to E11.5. Black arrows indicate Lfng expressing macular prosensory domains, black arrowheads indicate Lfng expressing cochlear prosensory domains, red arrows indicate Lfng reduction or absence in RBPJ cKO macular prosensory domains.

Fig. 3.. Deletion of Rbpjκat the otocyst stage does not result in a significant increase in the size of the VIII/VIIth ganglion during neurogenesis.

A: Whole mount immunofluorescence for βIII tubulin (Tuj1 antibody) in E9.5 Rbpjκ cKO and WT and littermate cleared with Scale and imaged using LightSheet microscopy. The Vth, VIIth, VIIIth and IXth ganglion are indicated with labels. B: Volume of Tuj1-labeled VIII/VIIth ganglia determined by surface mapping ganglion. C: Graph of data in box-plot format, bars represent maximum and minimum recorded values, with one value plotted as an outlier in WT graph. Mean represented by X, WT = 2.1 thousand millimeters³, Rbpjκ cKO = 2.4 thousand milimeters³. Student’s two-tailed t-test was performed to determine statistical significance. P = 0.19, n = 10 for WT and Rbpjκ cKO. Scale bars in A and B = 100μm.

Fig. 4.. Loss of Rbpjκ results in a loss of some of the key components of lateral induction and vestibular prosensory character.

A: Immunofluorescence staining for JAG1 (red) in Rbpjκ cKO and WT otic sections counterstained with DAPI (blue) from E9.5 to E11.5. White brackets indicate otic epithelium expressing JAG1 (E9.5) and JAG1 expression in vestibular prosensory domains (E10.5-E11.5), yellow brackets indicate vestibular otic epithelium that has lost Jag1 expression. White arrows indicate JAG1-expressing cochlear prosensory domains, yellow arrows indicate cochlear prosensory domains not expressing JAG1. B: Immunofluorescence staining for SOX2 (green) in Rbpjκ cKO and WT otic sections counterstained with DAPI (blue) from E9.5 to E11.5. White brackets indicate otic epithelium expressing SOX2 (E9.5) and SOX2 expression in vestibular prosensory domains (E10.5-E11.5), yellow brackets indicate vestibular otic epithelium that has lost SOX2 expression. White arrows indicate SOX2-expressing cochlear prosensory domains, yellow arrows indicate cochlear prosensory domains with reduced SOX2. Scale bars = 50μm.

Fig. 5.. The loss of vestibular prosensory domains does not appear to be due to significant decrease in proliferation or an increase in apoptosis.

A: Activated caspase 3 (green) immunofluorescence staining in E10.5 Rbpjκ cKO and WT otic sections counterstained with DAPI (blue). White arrows indicate typical instances of apoptosis detected in otic sections. B: Phosphorylated Histone H3 (green) immunofluorescence staining in E10.5 Rbpjκ cKO and WT otic sections counterstained with DAPI (blue). White arrows indicate examples of proliferating cells in the otic epithelium. Sections in A and B are adjacent serial sections from the same otocyst of a WT and mutant animal. C: Cell counts for proliferating cells for otocyst divided into quadrants. The 25% mark represents the anterior-most quarter of otic sections with 100% representing the posterior-most quarter. Bars indicate standard error of the mean. For 25% P = 0.08, for 50% P = 0.06, for 75% P = 0.34, for 100% P = 0.30, n = 5 WT and n = 6 cKO. D: Cell counts for proliferation in aggregate per otocyst. Bars indicate standard error of the mean, P = 0.07, n = 5 WT and n = 6 cKO. Student’s two-tailed t-test was used for statistical analysis in C and D.

Fig. 6.. Loss of lateral induction by deletion of Rbpjκ results in a loss of vestibular prosensory domains and a corresponding spread of non-sensory character.

A: SOX2 and LMX1A immunofluorescence staining on E10.5 Rbpjκ cKO and WT littermate otic sections with. Scale Bars = 100μm B: SOX2 and LMX1A immunofluorescence staining on E11.5 Rbpjκ cKO and WT littermate otic sections. At both ages, the domains of SOX2 and LMX1A are mutually exclusive in wild type embryos. However, in Rbpjκ cKO embryos, SOX2 staining is absent from the lateral region of the otocyst which is now a thin epithelium. LMX1A is beginning to be up-regulated in these regions. Scale bars = 100μm.

Fig. 7.. JAG1-mediated Notch signaling is required to maintain prosensory domains at the expense of non-sensory character.

A: Mating scheme to conditionally delete Jag1 in prosensory cells. Sox2-CreERT2 mice carrying two conditional alleles of Jag1 (Jag1^flox) were mated with mice doubly homozygous for the Jag1^flox allele and the Ai9 ROSA-tdTomato Cre reporter. B: Pregnant females received a dose of tamoxifen at a) E8.5 to initiate recombination and the mice were sacrificed at E10.5 for analysis, or b) E9.5 with sacrifice at E11.5. C: SOX2 and LMX1A immunofluorescence staining of E10.5 Jag1 cKO and WT otic sections with tdTomato reporter activity showing cells in which Jag1 was deleted. White brackets designate vestibular prosensory domains with typical SOX2 expression and corresponding absence of LMX1A expression. Yellow brackets denote vestibular prosensory domains in Jag1 conditional mutants. Tissue in which Jag1 was conditionally deleted are marked by the tdTomato reporter. In the mutant (tdTomato+) regions, SOX2 is staining is patchy and reduced, and LMX1A is has begun to spread into the tdTomato + domain. D: SOX2 and LMX1A immuno fluorescence staining of E11.5 Jag1 cKO and WT otic sections with the tdTomato reporter. In mutant embryos, the expression of SOX2 staining is again reduced, with LMX1A spreading into the former prosensory area (yellow brackets). Scale bars = 100μm.