Segregation of the anterior and lateral cristae in the chicken inner ear.

a-c) Low magnification surface views and transverse line optical reconstructions (x,y) of whole-mount preparations of the anterior part of the chicken otocyst at different stages of development (HH=Hamburger and Hamilton stages), immunostained for Sox2 expression. The anterior (AC) and lateral (LC) cristae segregate sequentially from the pan-sensory domain (PD), which then becomes the utricle (UT). The LC is connected to a second Sox2-expressing domain (stars) that does not give rise to a sensory organ. d-l’’) Stages 1 (d-f’’), 2 (g-i’’) and 3 (j-l’’) of the segregation of the LC, characterized by a progressive down-regulation of Sox2 expression in the domain separating the LC from the PD domain. High magnification views (e-e’’, h-h’’ and k-k’’ correspond to the boxed areas in respectively d-g-j) reveal striking changes in cellular morphology and organization during LC segregation. A specialized boundary domain composed of large cells that progressively align their membranes along the border of the LC (arrowheads in e’-l’’) becomes clearly visible from stage 2 onwards. This is concomitant to the formation of a basal constriction at the interface of the crista and pan-sensory domain, visible from stage 2 (arrowheads) in orthogonal z-reconstructions of the boundary domain (arrowheads in i-i” and l-l”).

Segregation of the anterior crista in the chicken inner ear.

Low and high magnification views of the anterior crista (AC) and pan-sensory domain (PD) at different stages of development (HH=Hamburger and Hamilton stages), immunostained for Sox2 expression and labelled with fluorescent phalloidin. At stage 1 of segregation (a-b’’), the two patches are in contact and few cells have enlarged surfaces at their interface (arrowheads in b’-b’’). At stage 2 (c-d’’), several rows of larger cells are seen at the interface of the PD and prospective AC (arrowheads in d’-d’’). At stage 3 (e-f’’), the two patches are separated by a Sox2-negative domain and several rows of large cells (arrowheads in f’-f’’).

A multicellular basal constriction forms at the interface of the lateral crista and pan-sensory domain.

a-c) Samples stained for F-actin and Sox2 expression at stages 1-3 of LC segregation in the chicken inner ear. Optical sections collected from the surface and basal planes, and orthogonal reconstructions (z). In all images, the pan-sensory domain is on the left and the LC on the right. a) F-actin is enriched at the base of the interface of the LC from stage 1 (arrowheads). b-c) At stage 2-3, the F-actin enrichment becomes very clear along the prospective border of the LC (arrowheads). A basal constriction of the cells (arrows in z) and an upward ‘pull’ of the basal lamina (stars) are also noticeable.

Analysis of the morphological changes during segregation of the lateral crista in the chicken otocyst.

a) Cell skeletons generated from phalloidin stained preparations at three stages of LC segregation. In all images, the interface domain is in the middle, the pan-sensory domain on the left, and the LC on the right. b-c) Colour scale and pooled scatter plots of the surface cell area show that the largest cells are located at the interface of the LC and pan-sensory domain from stage 1. d) Color-coded representation of the cell elongation ratio. e) Visualization of the long axis of the cells showing a progressive elongation between stages 1-3 and a clear alignment of the cells along the border of the LC at stage 3.

Visualization of cell nuclei at the boundaries of segregating sensory organs.

A-a”) Whole-mount surface view of a LC boundary stained for the apical surface marker ZO-1, Sox2, and DAPI. DAPI staining in a” shows no apparent difference in the sizes of cell nuclei between the cells with large surface (arrowheads) at the boundary (brackets) and the neighbouring cells.

Segregation of the anterior and lateral cristae in the heterozygous Lmx1a GFP/+ knock-in mouse inner ear.

Whole-mount surface views and orthogonal reconstructions (x) of the anterior part of the otocysts from Lmx1a GFP/+ mouse at different stage of development, immunostained for Sox2 expression and stained for F-actin. a) At E11.5, the prospective anterior (AC) and lateral (LC) cristae are visible at the anterior edge of the pan-sensory domain (PD). Lmx1a/GFP expression pattern overlaps with that of Sox2 in between the presumptive cristae and at their interface with the PD (b-b’). The border of the Lmx1a-GFP positive territory is jagged, but GFP-positive and negative cells are clearly separated (arrowheads in c’). Cell surfaces appear slightly enlarged at the limit of the Lmx1a/GFP (star in c). d) At E13.5, the AC, LC and UT are separated. Lmx1a/GFP expression is present in Sox2-positive cells at the border of the AC and LC (arrowheads in e-e’) and in prospective non-sensory cells in between the two cristae, which exhibited a clear reduction in Sox2 expression. Cells with enlarged surface areas are clearly recognizable in between the cristae and UT (brackets in f-f’). Lmx1a/GFP is detected in the half of the large cell domain facing the cristae. A basal constriction is present at the precise interface of the Lmx1a-expressing and non-expressing cells (arrows in f-f’). At P0, Lmx1a/GFP expression persisted in a wide non-sensory territory separating the two cristae from the UT (g). There is no overlap between Sox2 and GFP expression at the border of the sensory organs (stars & bracket in h-h’). No basal constriction is observed (bracket in i).

Alignment of cell borders at the interface of Lmx1a-positive and negative cells in the mouse inner ear.

(a-a’) Whole-mount preparations of the interface between the pan-sensory domain (PD) and lateral crista (LC) in an E13.5 Lmx1a GFP/+ mouse otocyst. F-actin labelling reveals a marked alignment of apical cell borders along the boundary of Lmx1a-GFP expression (arrowheads).

Comparison of the boundary domain in Lmx1a GFP/+ and Lmx1a GFP/GFP mouse inner ear.

Whole-mount preparations of the utricle (UT), lateral (LC) and anterior (AC) cristae region, after immunostaining for Sox2 (magenta) and F-actin labelling with phalloidin (white). (a) In the E15 Lmx1a GFP/+ inner ear (a-d), the three sensory patches are segregated and Lmx1a-GFP is present around the utricle and in between the two cristae (star in a). High magnification surface (b-c) and transverse reconstruction (d) of the interface domain (brackets). The boundary of Lmx1a-GFP expression is relatively straight (arrowheads) and Sox2 expression is reduced in the cells immediately adjacent to it and facing the utricle (stars in c). A basal constriction is visible at the border of the Lmx1a-GFP expression domain (arrow in d). In the E15 Lmx1a GFP/GFP inner ear (e-h) GFP expression is reduced at the interface of the UT and the fused AC and LC (stars in e). At higher magnification (f-h), note the presence of GFP-negative cells in between GFP-positive cells at the interface domain (arrows in f-g) but a domain of reduced Sox2 expression (stars in g) is maintained between the fused cristae and the UT. (h) In orthogonal reconstructions, note the absence of a basal constriction at the border of the cristae domain.

Quantitative analysis of the boundary domain in Lmx1aGFP/+ (left column) and Lmx1aGFP/GFP (right column) E15 mouse inner ear.

a) Representative cell skeletons generated from phalloidin stained preparations. In all images, the interface domain is in the middle, the utricle on the left, and the LC/AC domain on the right. b-c) Colour scale representation (from skeletons shown in a) and pooled scatter plots (from 2 Lmx1aGFP/+and 3 Lmx1aGFP/GFP samples) of the surface cell areas. Note the more uniform distribution of the large cells on both sides of the interface domain in the Lmx1a-null inner ear. d-e) Visualization of the long axis (from skeletons shown in a) and pooled elongation ratio (from 2 Lmx1aGFP/+and 3 Lmx1aGFP/GFP samples) of the cells. Note that cells are more aligned with one another and more elongated in the homozygous than in the heterozygous samples.

Expression pattern of phospho-myosin light chain (PMLC) in cross-sections of an E5 chicken inner ear.

a-a”) Punctate PMLC staining is present in the ventral part of the inner ear where a Sox2-positive prosensory domain (star) is localised. b-b”) High magnification views of the boxed area in a. The interface between the strong and weak Sox2 staining domains (arrows), typically corresponding to the boundary domain (bracket), exhibits an noticeable enrichment of PMLC staining at the surface and basal aspects of the epithelium arrowheads).

Effects of the ROCK inhibitor Y-27632 on the segregation of the lateral crista in organotypic cultures of embryonic chicken otocysts.

a-f) Surface and orthogonal views (x, c, f) of organotypic cultures of stage HH26 otocysts maintained for 3 hours in control medium (a-c) or medium supplemented with 20μm Y-27632 (d-f). Panels b-c and e-f are high magnification views of the boxed area in respectively a and d; the position of the large cell domain is indicated by a bracket and an example of multicellular rosette is highlighted (arrowhead in b’). Treatment with Y-27632 induces a loss of multicellular rosettes and abnormal folding of the epithelium (stars in e’-f). g-l) Surface and orthogonal views (x, i, l) of a pair of otocysts from the same HH26 embryo; one otocyst was fixed immediately (g-i) and the other maintained for 24 hours in culture (j-l). Panels h-i and k-l are high magnification views of the boxed area in respectively g and j; the position of the large cell domain is indicated by a bracket. After 24 hours, the lateral crista has progressed in its segregation from the pan-sensory domain. There is a clear reduction in Sox2 expression levels at the edge of the prospective LC and a basal constriction is noticeable (arrows in k-k’). m-r) Surface and orthogonal views (x, o, r) of a pair of otocysts from the same HH26 embryo; one otocyst was cultured 24 hours in control medium (m-o) and the other maintained for 24 hours in 20μm Y-27632 (p-r). Panels n-o and q-r are high magnification views of the boxed area in respectively m and p. Treatment with Y-27632 induces a loss of the typical epithelial cell morphology and large cell domain (compare n’ and q’) and a disruption of sensory organ segregation (stars in p). Abbreviations: LC=lateral crista; AC=anterior crista; PD=pan-sensory domain.

Quantification of the effects of ROCK inhibition on cell morphology and organisation during segregation of the lateral crista.

a) Representative images of phalloidin-stained apical surfaces of epithelial cells from organotypic cultures maintained for 3 and 24 hours in control medium, or medium supplemented with 20μm Y-27632. In all images, the interface domain is in the middle and the lateral crista on the right. b) Colour scale representation (from skeletons shown in a). c) pooled scatter plots (3 samples per condition) of the cell surface areas and d) pooled box plots (bar=median; red dot=mean) of the surface areas of cells located within a +/- standard deviation from the location of the largest cells (the presumptive boundary cells) of each sample; there is a statistically significant difference between 24h control and Y-27632 treated samples (Mann Whitney test W=2801, p=2.2e-16). e-f) Visualization of the long axis (from skeletons shown in b) and pooled elongation ratio of the epithelial cells.

Overexpression of a dominant-negative form of ROCK in the developing chicken inner ear disrupts sensory organ segregation.

a) surface view of otic epithelial cells 24h after in ovo electroporation with the pRCII-GFP expression plasmid. b-d) analysis of cell skeletons of transfected (in green) versus untransfected (UT) cells indicate that overexpression of the RCII-GFP fusion protein induces an increase in apical cell surface area. e) quantification of the between the LC and UT in ears transfected with pRCII-GFP expression plasmid and control RCAS-GFP. f-g’’) surface views of an E6 control otocyst electroporated at E2 with an RCAS-GFP proviral DNA. In this sample, the limit of Sox2 expression (arrowheads in g) matches that of strong GFP fluorescence (arrowheads in g’) and the presumptive tissue boundary in the middle of the large cell domain (bracket in g’’) separating the lateral crista (LC) from the utricle (UT). h-k’’) surface views of two E6 samples electroporated at E2 with RCAS-RCII-GFP and showing abnormal segregation of the LC. In the first example, the interface domain contains poorly aligned Sox2-expressing cells (arrowheads in i). A basal constriction is still visible (arrows in i’ and i’’, transverse reconstructions) but the large cells of the interface domain are not properly aligned (arrowheads in i’’). In the second example (j-k’’), a mosaic pattern of RCII-GFP fluorescence is observed in the interface domain. Cells with low levels of RCII-GFP fluorescence (star in k’) exhibit a more regular alignment of their apical borders (arrowheads in k’’) than transfected cells.

Analysis of cell proliferation during the segregation of the lateral crista in the chicken otocyst.

whole-mount preparations of samples collected one hour after an EdU pulse in ovo at stage 1 (a-b) or stage 2-3 (c-d) of segregation of the LC, immunostained for EdU, Sox2 an labelled with DAPI. Numerous EdU-positive cells (arrows in a-c) and mitotic figures (arrows in b-d) are present at the interface between the LC and PD (dotted line in b-d) at both stages. e) Spatial distribution of the mitotic figures within a 36×72μm region centred on the middle of the boundary domain, for early and late stages of segregation (n=7 embryos for each stage).

Schematic representation of sensory organ segregation.

a) The up-regulation of Lmx1a expression within the pan-sensory domain coincides with the formation of a specialised boundary domain composed of cells with enlarged cell surfaces. A basal constriction and an alignment of cell borders occur at the precise interface of Lmx1a-positive and Lmx1a-negative cells. We propose that it is a transient lineage-restricted boundary, dependent on actomyosin contractility, which separates adjacent pools of sensory progenitors. As the spatial segregation proceeds, the Lmx1a-expressing cells give rise to non-sensory cells separating sensory organs. b) Hypothetical regulatory gene network regulating sensory organ segregation. Notch-dependent lateral induction maintains Sox2 expression and promotes the prosensory fate. Lmx1a antagonizes Notch activity and promotes adoption of a non-sensory fate; it is also required for the proper formation of the boundary domain between the cristae and utricle.