A combination of Notch signaling, preferential adhesion and endocytosis induces a slow mode of cell intercalation in the Drosophila retina

Abstract

Movement of epithelial cells in a tissue occurs through neighbor exchange and drives tissue shape changes. It requires intercellular junction remodeling, a process typically powered by the contractile actomyosin cytoskeleton. This has been investigated mainly in homogeneous epithelia, where intercalation takes minutes. However, in some tissues, intercalation involves different cell types and can take hours. Whether slow and fast intercalation share the same mechanisms remains to be examined. To address this issue, we used the fly eye, where the cone cells exchange neighbors over ∼10 h to shape the lens. We uncovered three pathways regulating this slow mode of cell intercalation. First, we found a limited requirement for MyosinII. In this case, mathematical modeling predicts an adhesion-dominant intercalation mechanism. Genetic experiments support this prediction, revealing a role for adhesion through the Nephrin proteins Roughest and Hibris. Second, we found that cone cell intercalation is regulated by the Notch pathway. Third, we show that endocytosis is required for membrane removal and Notch activation. Taken together, our work indicates that adhesion, endocytosis and Notch can direct slow cell intercalation during tissue morphogenesis.

Keywords: Adherens junction; Adhesion; Cell intercalation; Epithelia; Nephrins; Notch.

Fig. 1.

Cone cell intercalation and retinal cell growth trajectories. (A) The arrangement of cells in the ommatidium. A, anterior; CC, cone cell; Eq, equatorial; IOC, interommatidial cell; P, posterior; Pl, polar; PPC, primary pigment cell. (B) Stages of CC intercalation. (C) Average relative length (L-L0) of CC and PPC AJs during ommatidium development. Time 0 is the middle of the four-way vertex stage (n=13 ommatidia from two retinas). The different colored lines in the graph refer to the cell boundaries depicted in the schematic. (D) Confocal sections taken from a time-lapse movie of ommatidium development, with AJs labeled with endogenous Ecad::GFP. IOCs are outlined in red and numbered through subsequent frames. Tertiary pigment cells are labeled in blue. Dashed purple lines indicate the inter CC AJs and yellow outlines indicate the bristle cell complexes. (E) Average apical area of CCs over time (n=4 ommatidia). Vertical lines demarcate the stages of CC intercalation. (F) Average CC cluster axis ratio over time relative to the middle of the four-way vertex stage (n=14 ommatidia). (G) Average lengths of CC cluster axes over time relative to the middle of the four-way vertex stage (n=14 ommatidia). (H) Average cross-correlation of the rate of change in the length of the central CC AJ (shown in blue in the schematic), with the adjoining CC AJs (shown in red in the schematic). Correlation coefficient: r=-0.54±0.11 (mean±s.d.) at a time lag of 0 (n=13 ommatidia). (I) Average cross-correlation of rate of change in the length of the central CC AJ (shown in blue in the schematic) with the CC-PPC AJ (shown in red in the schematic) (n=13 ommatidia). Scale bars: 5 μm. Error bars indicate s.d. in F-I.

Fig. 2.

MyoII expression and dynamics in cone cell intercalation. (A-C) sqh^AX3;sqh-sqh::GFP/+;Ecad::Tomato/+ flies showing localization of MyoII (Sqh::GFP) at the junction shrinkage (A-A″), four-way vertex (B-B″) and junction elongation (C-C″) stages of cone cell (CC) intercalation. (D) Quantification at the AJ shrinkage stage of Sqh::GFP intensity on shrinking A/P cone cell AJs (red) compared with adjoining AJs (green) paired by ommatidium (n=514 ommatidia). Student's t-test: ***P<0.0001. Data are mean±s.e.m. (E,F′) Polar histograms showing directions of MyoII flow vectors calculated by PIV. (E) A and (E′) P cone cells during AJ shrinkage phase. (F) A and (F′) P cone cells during AJ elongation phase. Scale bars: 5 μm.

Fig. 3.

Modeling the contribution of MyoII contractility and adhesion to cone cell intercalation. (A-A″″) Quantification at the AJ elongation stage of (A) Sqh::GFP, (A′) Ecad::GFP, (A″) Ncad staining, (A‴) Hbs::GFP intensity and (A″″) Rst::GFP intensity on each AJ type normalized to the average of the cone cell-cone cell-side AJ (shown in green). CC, cone cell; IOC, interommatidial cell; PPC, primary pigment cell. Data are mean±s.e.m. (B,B′) Initial recoil velocity of ablation of (B) PPC-PPC AJs and (B′) IOC-IOC AJs at each stage of ommatidial development. For PPC-PPC AJs: one-way ANOVA n.s. P=0.784, n=29, 20 and 24 AJs for early, mid and late stages of development, respectively. For IOC-IOC AJs: one-way ANOVA P<0.0001, Tukey's post hoc: early-mid n.s. P=0.97; early-late P=0.0001; mid-late P=0.0018. n=11, 8 and 38 AJs for early, mid and late stages of development, respectively. Data are mean±s.e.m. (C) Each cell-cell boundary included in the vertex model is color coded following the code used in A,B. The bonds representing the cytosolic contractile actomyosin meshworks are represented as dashed magenta lines, with CCs highlighted in blue, PPCs in green and IOCs in gray. The parallel spring schematic represents the tension structure for adhesion and myosin contribution in each cell-cell contact. (D) Heatmap demonstrating the state of intercalation as a function of cytosolic contractile actomyosin meshwork strength (as a fraction of base tension level) (x-axis) and the range of contributions from adhesion and myosin intensity measurements (y-axis). Spring schematics represent the weight of each adhesion and MyoII in the calculation of tension values for each row. Ommatidia schematics represent the strength of the cytosolic mesh. See Materials and Methods and Table S1 for details. Green represents stable intercalation (E″), red represents failed intercalation (E) and yellow represents a stable four-way junction forming a rosette (E′). Gray points have unstable ommatidia geometry. (E-E″) Simulation snapshots where the tension values cannot drive or stabilize the intercalation (E), where a stable four-way junction is formed (E′) and where a stable intercalation occurs (E″).

Fig. 4.

The RhoA-MyoII pathway is largely dispensable for cone cell intercalation. (A) UAS-MyoII^DN::YFP expressed under control of pros-Gal4. Arm (A), MyoIIDN::YFP (A′) and merged panel (A″). (B) Progression of cone cell intercalation at 29°C when MyoII^DN is expressed in cone cells (n=4 retinas, 2212 ommatidia). (C) Progression of cone cell intercalation for UAS-zipIR expressed under control of pros-Gal4 alongside matched controls expressing UAS-CD8::mCherry, raised at 29°C (pros-Gal4; zipIR: n=6 retinas; controls: n=4 retinas). ‘Other’ category contains any cone cell orientations that do not fit into the other categories (e.g. shift in position of the primary pigment cell junctions relative to the cone cells). (D) GFP-positive cells (circled using a dashed yellow line) express the RhoA^N19 transgene (green), Ecad::GFP (gray) (D′) and merged panels (D″). White arrows point to the newly extended AJ between Pl and Eq cone cells. (E) GFP-positive cells (circled using a dashed yellow line, n=15) express a RhoA dsRNAi transgene. White arrows point to the newly extended AJ between the Pl and Eq cone cells. (F) Wild type and (F′) spaGal4;;prosGal80 genotype expressing RhoA dsRNAi in primary pigment cells. Ommatidia marked with an asterisk show a shorter AJ between the Pl and Eq cone cells compared with wild type. The length of these AJs is quantified in F″. For the spaGal4; prosGal80/RhoA dsRNAi and spaGal4; prosGal80/+ genotypes, three retinas each were used for quantification. AJ n=80 and control n=65. Data are means. Scale bars: 5 μm.

Fig. 5.

Rst and Hbs regulate cone cell intercalation. (A-F‴) Confocal projection through the cone cells (CCs) showing (A-C‴) Hbs::GFP and (D-F‴) Rst::GFP at (A-A‴,D-D‴) AJ shrinking stage, (B-B‴,E-E‴) four-way vertex stage and (C-C‴,F-F‴) AJ elongation stage. In A″,B″,C″,D″,E″,F″ the Ice lookup table was used to visualize variation in levels along the CC-primary pigment cell AJ. A reduction in intensity is seen around the Pl and Eq CCs. A‴,B‴,C‴,D‴,E‴ and F‴ are merged images of Arm (red) and the GFP channel (green). (G) Schematic depicting where Hbs and Rst colocalize. (H) Representative wild-type, control ommatidium from a hbsIR mosaic retina. (H′) Eq cell expressing hbsIR (red) and stalled at the four-way vertex. (H″) P cone cell expressing hbsIR and stalled at the shrinking stage. (H‴) Anterior CC expressing hbsIR and showing a cell-sorting phenotype. (I) Quantification of cone cell intercalation in hbsIR mosaic ommatidia. dsRNAi-expressing cells are in red. (J) Pl cone cell mutant for rst⁶ (lacking GFP) stalled at the four-way vertex. (J′) Anterior CC mutant for rst⁶ (lacking GFP) undergoes normal intercalation. (J″) Primary pigment cells mutant for rst⁶ (lacking GFP) fail to shrink the A/P CC AJ. (K-M) Quantification of CC intercalation in mosaic ommatidia. dsRNAi for hbs and rst⁶ mutant cells are in red. Scale bars: 5 μm (A-E); 2 μm (H,J).

Fig. 6.

Notch signaling controls cone cell intercalation. (A,A′) Timecourse of NiGFP expression (gray) during cone cell intercalation. Cell membranes are labeled with PH::ChFP (purple). (B,B′) Timecourse of Dl::GFP expression (gray) during cone cell intercalation. Cone cells are outlined using a dashed orange line. Cell membranes are labeled with PH::ChFP (purple) in B and with Baz::ChFP in B′. (C,C′) Timecourse of Neur::GFP expression (gray) during cone cell intercalation. Cell membranes are labeled with Baz::ChFP (purple). (D) Representative NiGFP signal (gray) in cone cell nuclei, also labeled using an RFP-nls reporter (red). (E) Quantification of the nuclear signal for Notch in cone cells over time. There is a decrease in the Notch signal in the P cell as intercalation takes place. Data are mean±s.d. (F) Representative staining of the Notch target gene mδ-GFP (gray). Cone cell nuclei are labeled using an RFP-nls reporter (purple) and are circled using colored dashed lines with one specific color attributed to each quartet. (G) UAS-Mam^DN expressed under control of prosGal4. Yellow asterisks indicate failed intercalation. (H) Progression of cone cell intercalation for retinas expressing UAS-Mam^DN under control of prosGal4 (n=5 retinas, 3433 ommatidia) and for control wild-type flies raised at 25°C (n=3 retinas, 1909 ommatidia). (I) Number of reversions between the different stages of cone cell intercalation in wild-type compared with UAS-Mam^DN/pros-Gal4 retinas (wild type, n=8 ommatidia; UAS-Mam^DN/pros-Gal, n=4 ommatidia). Data are mean±s.d. (J) Single cells expressing UAS-Mam^DN marked by presence of RFP (red). Top arrowhead indicates an example of cone cells at a four-way vertex stage when the Eq cell is affected, and the bottom arrowhead indicates a quartet with a A/P AJ when the Eq and Pl cone cells are affected. (K) Quantification of the percentage of ommatidia with each AJ type (A/P, four-way vertex and Eq/Pl) when different combinations of cone cells express UAS-Mam^DN. n numbers are shown for each bar of the graph. Scale bars: 5 μm in A-C′; 10 μm in G,J.

Fig. 7.

Endocytosis plays a role in all steps of cone cell intercalation. (A,B,D-F) Retinas expressing UAS-shibire^ts under the control of prosGal4 stained for Arm. Flies were raised at 25°C and then transferred to 31°C at (A) 20% APF and (B) 24% APF, and incubated overnight. Flies were transiently transferred to a restrictive temperature for 4 h at (D) 20% APF, (E) 24% APF and (F) 28% APF. (C) Progression of cone cell intercalation in B (n=4 retinas, 1442 ommatidia). (G) Progression of cone cell intercalation in D-F (n=7, 6 and 6 retinas, respectively; 3352, 3100 and 2797 ommatidia, respectively). (H) Stills taken from a movie of retina expressing UAS-shibire^ts under the control of prosGal4 with Ecad::GFP to label the AJs. (I) Quantification of the percentage of ommatidia with each AJ type (A/P, four-way vertex and Eq/Pl) when different combinations of cone cells express UAS-Brd^R. n is indicated for each bar of the graph. Scale bars: 10 μm in A,B,D-F; 5 μm in H.