Cell migration, intercalation and growth regulate mammalian cochlear extension

Abstract

Developmental remodeling of the sensory epithelium of the cochlea is required for the formation of an elongated, tonotopically organized auditory organ, but the cellular processes that mediate these events are largely unknown. We used both morphological assessments of cellular rearrangements and time-lapse imaging to visualize cochlear remodeling in mouse. Analysis of cell redistribution showed that the cochlea extends through a combination of radial intercalation and cell growth. Live imaging demonstrated that concomitant cellular intercalation results in a brief period of epithelial convergence, although subsequent changes in cell size lead to medial-lateral spreading. Supporting cells, which retain contact with the basement membrane, exhibit biased protrusive activity and directed movement along the axis of extension. By contrast, hair cells lose contact with the basement membrane, but contribute to continued outgrowth through increased cell size. Regulation of cellular protrusions, movement and intercalation within the cochlea all require myosin II. These results establish, for the first time, many of the cellular processes that drive the distribution of sensory cells along the tonotopic axis of the cochlea.

Keywords: Cochlear development; Convergent extension; Inner ear; Live imaging; Mouse; Radial intercalation.

Fig. 1.

Extension of the cochlear duct. (A,C,E) Single-plane confocal images of the base of the mouse cochlear duct at the indicated time points and orientations, with axes shown in C,C′ (bm, basement membrane). 3D renderings are shown in A″,C″,E″. PrCs and their derivatives, namely HCs and SCs, are marked with anti-CDKN1B (red) and anti-SOX2 (blue), together with F-actin (green). Scale bars: 10 μm. (B,D,F) yz optical projections of Sox2^CreERT2; R26R^tdTomato cochleae with tdTomato-filled PrCs, HCs and SCs (magenta) and F-actin (green). Scale bars: 5 μm. (A-A″) At E14, PrCs are grouped in a pseudostratified band. (B) E14 PrCs have morphologies consistent with undifferentiated, pseudostratified epithelial cells. (C-D) At E16, the basal region of the cochlear duct has developed the basic cellular pattern of the OC, with rows of developing inner and outer HCs evident. (E-F) At P0, differentiation of the OC is evident and HCs have developed stereociliary bundles. Distinct SC phenotypes, such as pillar cells and Deiters' cells, can be identified based on tdTomato labeling. (G) Quantification of cell number within a section of the epithelium extending along 100 µm of the axis of extension [base-to-apex (B-A)] indicates a significant decrease of ∼42% between E14 and E16 (P<0.01). By contrast, the decrease between E16 and P0 is only 15% (not significant). n=5 samples for each age. (H) To quantify cellular stratification, the number of cells within a 10×10 µm column extending between the basement membrane and the luminal surface (BM-L, A″) was determined for each time point. Stratification significantly decreases from an average of 9.33 cells at E14 to 5.04 cells at E16 and 3.21 cells at P0. P<0.01 for all pairwise comparisons; n=5. (I) As PrCs develop into HCs or SCs, there are significant increases in cell volume. Number of cells: E14, 19; E16, 13 HCs, 40 SCs; P0, 6 HCs, 23 SCs. P<0.01 for all comparisons except P0 HCs versus SCs, P<0.05. (J) To illustrate the effects of changes in cell volume on cochlear extension and growth, the epithelial volume of 100 PrCs and/or their derivatives was determined at each time point. Results indicate significant increases (P<0.01; n=5) between each time point, demonstrating that cell growth contributes to cochlear extension.

Fig. 2.

OC cells migrate towards the cochlear apex. (A) z-stack projection images from an Atoh1^Cre*PR; R26R^tdTomato cochlear explant established at E13 and imaged at 2 DIV (E15 equivalent). Indicated time points (h:min) are relative to the beginning of the time-lapse. Lengthening and slight narrowing of the population of labeled cells are apparent as they move in the apical direction. Scale bar: 50 µm. (B) Migration tracks for pseudocolored individual cells from an Atoh1^Cre*PR; R26R^ZsGreen 16 h time-lapse video started at 1 DIV (E14). A single explant was imaged near the mid-apex (B) and at the base (B′) of the cochlea. Cells at the mid-apex move steadily in the apical direction; cells at the base show less consistent migratory paths. Some cells are nearly stationary (arrows). Scale bar: 20 µm. (C) Displacement rates (µm/h) for individual cells located at the mid-apex are significantly greater than those at the base (P<0.001). (D) Average straightness index of cells at apex versus base, indicating that cells in the mid-apex migrate in a straighter line (P<0.001). (C,D) n=3 explants, 18 cells. (E) Average displacement of migrating cells on the indicated embryonic day equivalents, taken from time-lapse sequences of 10-16 h. The rate of cell movement gradually decreases with developmental age. All differences are statistically significant (P<0.01), except E14 versus E16, and E15 versus E16. n=(age, explants, cells): E14, 7, 41; E15, 9, 68; E16, 8, 75; E17, 10, 66; E18, 6, 30. (F) Average displacement of cells per day when imaged once every 24 h. Early (E14-E16) values are all significantly different (P<0.01) from late (E17-P0) values. n=(age, explants, cells): E14, 4, 14; E15, 5, 15; E16, 7, 24; E17, 4, 14; E18, 4, 14. (E,F) Data shown are averages±s.e.m.

Fig. 3.

CE occurs during cochlear outgrowth. (A,B) z-stack projections from an Atoh1^Cre*PR; R26R^ZsGreen time-lapse begun at E14 equivalent, showing CE measurements. The distance between cells along the x-axis (base-to-apex) increases, whereas distance along the y-axis (medial-lateral) decreases. Scale bar: 20 µm in B for A,D,E. (C) Quantification of average change in distance between nearby pairs of cells over 6 h. Convergence (negative ΔY) and extension (positive ΔX) are observed at E14-E15. Extension continues until E17, but little convergence is observed beyond E15. Data shown are averages±s.e.m., and statistically different values are indicated (P<0.01). n=(age, explants, cell pairs): E14, 8, 26; E15, 8, 28; E16, 5, 46; E17, 5, 45. (D,E) Cells align into rows in an E15 Atoh1^Cre*PR; R26R^tdTomato cochlear explant. A rosette of labeled HCs (D, numbered) rearranges into the first and second rows of OHCs over 6 h (E). (F) The luminal surface of a phalloidin-stained E14 cochlear epithelium, overlaid with depictions of the cell vertices. Vertices of three cells are marked with a pink dot, and those containing four or more cells with a green dot. Scale bar: 5 µm. (G) The percentage of cell vertices containing four or more cells decreases with developmental age. Values are significantly different (P<0.01). n=(age, sample, cell vertices): E14, 3, 740; E16, 4, 1170; E18, 3, 657. (H) The orientation of T1 transition summed nematics at E15, with axes indicated. Each line represents the average orientation of four-cell (T1) vertices resulting in neighbor exchange in a single time-lapse frame. The lengths of the lines along the nematic norm reflect how uniformly ordered the orientations of the exchanges are. A horizontal line orientation indicates productive neighbor exchange leading to extension along the base-to-apex axis, but the majority of the T1 nematics are oriented at ∼45°, between the base-to-apex and medial-lateral axes. Productive T1 neighbor exchange and summed nematic is depicted below the plot. n=3 explants, 143 frames, 1238 cells. (I) Still images of a 2 µm luminal surface z-stack confocal projection from an E14 R26R^mT-mG cochlear explant, with four cells pseudocolored and numbered. In 1 h, cells 2 and 4 intercalate between cells 1 and 3. A common vertex forms, then resolves as cells 2 and 4 intercalate perpendicular to the base-to-apex elongation axis. Scale bar: 5 µm. (J-L) Luminal surface views of the OC at the indicated ages. Spatial relationships for cell-cell contacts were determined as shown (magenta lines). An overlay of spatial relationships for multiple cells in the OHC region is shown for each age. Orientations of cell-cell contacts at E14 and E16 are similar. At P0 cellular patterning has changed significantly. Rose diagrams illustrate the percentage of cells with contacts at the indicated positions. n=(age, samples, cells, contacts) E14, 3, 45, 271; E16, 3, 45, 179; P0, 3, 45, 123.

Fig. 4.

Migrating cochlear epithelial cells generate cellular protrusions. (A,B) Images from a Sox2^CreERT2; R26R^mT-mG E15 explant, labeled with membrane EGFP. (A) xy (luminal) view; (B) xz (luminal-basal) view. Cellular protrusions are present on both SCs and HCs. SC protrusions extend primarily along the basement membrane (white arrows). HCs generate thin basal projections (green arrows). Putative cell types are labeled in B. Scale bar: 10 µm. See Movie 5. (C) Duration of cellular protrusions in HCs and SCs. The average lifespan of an SC protrusion is longer than that of a similar protrusion in an HC (P<0.001). (D) SCs display nearly three times as many protrusions over time than HCs (P<0.001). (C,D) n=5 explants, 23 cells, 1365 protrusions, 1261 frames (1 cell/frame). (E) Distribution of angles of protrusions in xy, relative to cell centroid, in the indicated cell types at the indicated time points. Analysis of protrusions in PrCs at E14 indicates a non-uniform distribution (P<0.05), with a bias towards the medial side of the epithelium. At E15, SC protrusions also show a non-uniform distribution (P<0.005), skewed 45° towards the axis of extension. Separation of E15 SCs based on position within the developing OC [lateral supporting cells (LSCs) are in the OHC region and medial supporting cells (MSCs) are in the inner hair cell (IHC) region] indicates that the distributions differ (P<0.005), with more protrusions in each region towards the midline. At E16, no bias in the distribution of protrusions was observed in SCs. E15 or E16 HCs did show a bias in protrusions towards the medial side (P<0.005), but these cells do not contact the basement membrane. n=(age, cells, protrusions): E14, 13, 537; E15, 23, 1365; E16, 16, 1239.

Fig. 5.

MyoII is required for cell migration and intercalation. (A) xz projections from a time-lapse (Movie 6) of a tdTomato-labeled E15 explant treated with blebbistatin for 3 h. During MyoII inhibition, a presumptive HC extends a basal protrusion (arrows), which retracts when blebbistatin is diluted. Scale bar: 10 µm. (B) Time-lapse images from an Atoh1^Cre*PR; R26R^{tdTomato/mT-mG} E16 explant labeled with membrane tdTomato and Cre-induced cytoplasmic tdTomato. Blebbistatin (10 µM) slows cell movement and disrupts cellular organization. Four putative HCs (numbered) become misaligned during MyoII inhibition. See Movie 7. Scale bar: 10 µm. (C) Cell displacement per hour in blebbistatin-treated explants. Each line represents the displacement of a single cell. The gray box indicates the duration of the 10 µM blebbistatin treatment. After 3 h, blebbistatin was diluted to 1 µM, and the rate of cell movement gradually increased. (D) Measurements of luminal surface area of PrCs from time-lapse sequences, showing an increase in surface area before and after 2.5 h of 10 µM blebbistatin treatment (P<0.05). (E) Cell height of the PrCs in D, measured along the z-axis of xz projections, decreases after 2.5 h of MyoII inhibition (P<0.01). (D,E) n=3 explants, 10 cells. (F) Control cochlear explant established at E13 and maintained for 3 DIV, stained with phalloidin. Cell vertices containing three (magenta) or four or more cells (green) are indicated. Scale bar: 10 µm. (G) Similar image as in F from an explant treated with 10 µM blebbistatin, with fewer four-cell vertices. (H) Blebbistatin treatment leads to a decrease in the percentage of vertices containing four or more cells (P<0.001). Number of vertices: 1218, control; 1471, blebbistatin. (I) The luminal surface area of both HCs and SCs increases in response to MyoII inhibition (P<0.001). Number of cells: 133, control; 157, blebbistatin. (H,I) Four explants per condition.

Fig. 6.

Model of cell movement and rearrangement of OC cells during cochlear development. At E14, PrCs are highly pseudostratified in a relatively wide array. Cells intercalate at the luminal surface (arrows), and radially intercalate through the basement membrane to luminal axis. Presumptive HCs (light green) have thin protrusions toward the basement membrane, and overall cell movement is towards the apex. At E16, movement toward the apex continues, but differentiating HCs (green) no longer contact the basement membrane. Some four-cell junctions are still present, and stratification of the epithelium has decreased. By P0, cellular patterning is mostly complete, creating the longer, thinner OC with a mosaic of HCs and SCs.