April 1, 2004;
Specification of the otic placode depends on Sox9 function in Xenopus.
The vertebrate inner ear
develops from a thickening of the embryonic ectoderm
, adjacent to the hindbrain
, known as the otic placode
. All components of the inner ear
derive from the embryonic otic placode
. Sox proteins form a large class of transcriptional regulators implicated in the control of a variety of developmental processes. One member of this family, Sox9
, is expressed in the developing inner ear
, but little is known about the early function of Sox9
in this tissue
. We report the functional analysis of Sox9
during development of Xenopus inner ear
otic expression is initiated shortly after gastrulation in the sensory layer of the ectoderm
, in a bilateral patch of cells immediately adjacent to the cranial neural crest
. In the otic placode
colocalizes with Pax8
one of the earliest gene expressed in response to otic placode
inducing signals. Depletion of Sox9
protein in whole embryos using morpholino antisense oligonucleotides causes a dramatic loss of the early otic placode
. Later in embryogenesis, Sox9
morpholino-injected embryos lack a morphologically recognizable otic vesicle
and fail to express late otic markers (Tbx2
) that normally exhibit regionalized expression pattern throughout the otocyst
. Using a hormone inducible inhibitory mutant of Sox9
, we demonstrate that Sox9
function is required for otic placode
specification but not for its subsequent patterning. We propose that Sox9
is one of the key regulators of inner ear
specification in Xenopus.
[+] show captions
Fig. 1. Sox9 whole-mount in situ hybridization. (A) Sox9 expression at early neurula stage (stage 3), dorsolateral view, anterior towards the right. Sox9 is detected in two major domains, the neural crest-forming regions (green arrows) and the prospective otic placode (red arrows). (B) Sox9 expression is greatly increased in the neural crest (green arrows) and in the prospective otic placode (red arrows) at stage 14. Lateral view, anterior towards right. (C) For comparison, Pax8 expression at the neurula stage is shown. Pax8 is detected in the prospective pronephros (green arrow) and the otic placode (red arrow). Lateral view, anterior towards the right. (D) Cross-section through the prospective otic placode of a stage 14 embryo, Sox9 is detected in the sensory layer of the ectoderm (op, red arrowheads) before any sign of thickening of the ectoderm. (E) At the tailbud stage, Sox9 persists in the invaginating otic placode (red arrow). Lateral view, anterior towards the right. (F) Serial cross-sections of a tailbud stage embryo; Sox9 is expressed throughout the invaginating otic cup. (G) Stage 30 embryo, lateral view, anterior towards the right; (H) corresponding section. Sox9 expression is detected in the dorsal region of the otic vesicle (ov, red arrows). (I-M) In stage-matched embryos, Sox9 is first detected in the presumptive otic ectoderm (red arrows) lateral to the prospective neural crest (green arrows) at stage 12.5 (J,K). (K) Higher magnification view of the embryo in J. Sox9 expression appears prior to that of Pax8 (L,M), which is initiated in the presumptive otic ectoderm (red arrow) and pronephros (green arrow) at stage 13 (N). Dorsal views, anterior towards the top (I,L). Lateral views, anterior towards the right (J,M,N).
Fig. 2. Inhibition of Wnt or Fgf signaling pathway blocks Sox9 expression in the otic placode/vesicle. (A) Sox9 expression is lost in the otic placode (upper panels, stage 17) of embryos injected with 1ng of GSK3β mRNA (arrows, injected side). Sox9 expression is reduced in the otic vesicle (lower panels, stage 23) of embryos injected with 1 ng of dominant-negative FgfR (XFD) mRNA (arrows, injected side). For comparison, normal expression of Sox9 is shown on the uninjected side of these embryos (arrows, uninjected side) at stage 17 and stage 23. As Fgf signaling is also required for development of posterior structures, XFD/FgfR-injected present a characteristic open blastopore (lower panels). All panels are lateral views, anterior towards the right (left panels) or anterior towards the left (right panels). (B) Transverse section through an embryo injected with GSK3β (upper panel) or XFD/FgfR (lower panel) mRNA. In both cases the embryos present a reduced otic vesicle on the injected side (left side, arrows). br, brain; no, notochord; ov, otic vesicle.
Fig. 3. Sox9 depletion prevents formation of the otic placode. (A) At stage 22, Pax8 and Tbx2 expression is eliminated upon injection of 10 ng of Sox9-mo in one blastomere at the two-cell stage (left side). For comparison, the otic expression of Pax8 and Tbx2 is unaffected (arrows) in embryos injected with 10 ng of Co-mo or Sox9-mis oligos. Sox9-mo phenotype can be rescued by injection of 2 ng of mouse Sox9 mRNA (Sox9-mo+mSox9). RNA encoding the lineage tracerβ -galactosidase was coinjected with the morpholino oligo to identify the injected side (red staining, left side in all panels). (B) Quantification of Tbx2 and Pax8 in situ hybridization results. The numbers at the top of each bar indicate the number of cases analyzed.
Fig. 4. Sox9 depletion prevents expression of late otic vesicle markers. (A) Schematic representation of the expression pattern (purple color) of four markers genes expressed in the epithelium of the otic vesicle at stage 35. Tbx2 is expressed through out the entire otic vesicle. Bmp4 expression is restricted to the anterior and posterior region of the otic vesicle. Wnt3a and Otx2 are expressed at the dorsal and ventral aspect of the otic vesicle, respectively. d, dorsal; v, ventral; a, anterior; p, posterior. (B) Tbx2, Bmp4, Wnt3a and Otx2 expression is lost in the otic vesicle of embryos injected with 2 ng of Sox9-mo in one blastomere at the eight-cell stage (arrows, injected side). RNA encoding the lineage tracer β-galactosidase was co-injected with the oligo AS to identify the injected side (red staining). For comparison, normal expression of Tbx2, Bmp4, Wnt3a and Otx2 is shown on the uninjected side of these embryos (arrows, control side). (C,D) Transverse section through an embryo injected with 2 ng of Sox9-mo. On the injected side (left side, arrow), the otic vesicle fails to form. br, brain; no, notochord; ov, otic vesicle.
Fig. 6. Sox9 is required to specify the otic placode. (A) Experimental procedure. Embryos at the two-cell stage are injected in one blastomere with 1 ng of Sox9δC-GR mRNA. Embryos are subsequently incubated with dexamethasone (+Dex) at different time points during development (stage 6, 11 and 15), and fixed at stage 22 for detection of Pax8 and Tbx2 by whole-mount in situ hybridization. (B) Addition of dexamethasone at the blastula (stage 6) and gastrula (stage 11) stages results in a severe reduction of Pax8 and Tbx2 expression domain in the otic placode. However, Pax8 and Tbx2 otic expression remain unaffected in embryos treated with dexamethasone at stage 15 (arrows). Upper panels are dorsal views, anterior towards the top. Lower panels are anterior views. RNA encoding the lineage tracer β-galactosidase was co-injected to identify the injected side (red staining). In all panels, embryos are injected on the left side. (C) Quantification of the in situ hybridization results. The number of cases analyzed for each time point is indicated.
Fig. 7. Sox9 is not required to pattern the otic vesicle. (A) Experimental procedure. Embryos at the two-cell stage are injected in one blastomere with 1 ng of Sox9δC-GR mRNA. Embryos are subsequently incubated with dexamethasone (+Dex) at different time points during development (stage 11, 15 and 20), and fixed at stage 35 for detection of Wnt3a and Otx2 by whole-mount in situ hybridization. (B) While addition of dexamethasone at stage 11 blocks Wnt3a and Otx2 expression on the injected side, addition of dexamethasone at the stage 15 or stage 20 does not affect the regionalized expression of Wnt3a and Otx2 (arrows). RNA encoding the lineage tracer β-galactosidase was co-injected to identify the injected side (red staining), the right side for all embryos.
Fig. 8. Dlx3 depletion blocks the otic expression of Sox9. (A) Injection of 30 ng of Dlx3-mo in one animal ventral blastomere at the two-cell stage reduces Sox9 otic expression (injected side, arrow). For comparison, the control side shows the normal expression of Sox9 in the presumptive otic placode (arrow). Pax8 expression was not significantly inhibited by this procedure. Lateral views of stage 14 embryos, anterior towards the right (left panel). (B) Dlx3-mo-injected embryos form normal otic vesicle at stage 35 (arrows). br, brain; no, notochord; ov, otic vesicle.