Integrins regulate epithelial cell shape by controlling the architecture and mechanical properties of basal actomyosin networks

Abstract

Forces generated by the actomyosin cytoskeleton are key contributors to many morphogenetic processes. The actomyosin cytoskeleton organises in different types of networks depending on intracellular signals and on cell-cell and cell-extracellular matrix (ECM) interactions. However, actomyosin networks are not static and transitions between them have been proposed to drive morphogenesis. Still, little is known about the mechanisms that regulate the dynamics of actomyosin networks during morphogenesis. This work uses the Drosophila follicular epithelium, real-time imaging, laser ablation and quantitative analysis to study the role of integrins on the regulation of basal actomyosin networks organisation and dynamics and the potential contribution of this role to cell shape. We find that elimination of integrins from follicle cells impairs F-actin recruitment to basal medial actomyosin stress fibers. The available F-actin redistributes to the so-called whip-like structures, present at tricellular junctions, and into a new type of actin-rich protrusions that emanate from the basal cortex and project towards the medial region. These F-actin protrusions are dynamic and changes in total protrusion area correlate with periodic cycles of basal myosin accumulation and constriction pulses of the cell membrane. Finally, we find that follicle cells lacking integrin function show increased membrane tension and reduced basal surface. Furthermore, the actin-rich protrusions are responsible for these phenotypes as their elimination in integrin mutant follicle cells rescues both tension and basal surface defects. We thus propose that the role of integrins as regulators of stress fibers plays a key role on controlling epithelial cell shape, as integrin disruption promotes reorganisation into other types of actomyosin networks, in a manner that interferes with proper expansion of epithelial basal surfaces.

Fig 1. Integrins regulate whip-like structures and stress fibers formation.

(A) Schematic drawing of a S8 egg chamber illustrating the different types of actin organisations found on the basal side of FCs. (B-E’) Basal surface view of mosaic S7 (B, B’), S8 (C, C’), S9 (D, D’) and S10 (E, E’) egg chambers containing mys FC clones, stained for anti-GFP (green) and Rhodamine Phalloidin to detect F-actin (red). (B, B’) mys FCs (GFP-negative) contain more whip-like structures (arrow in B’) than control FCs (GFP-positive, arrowhead in B’). (C-E’) Stress fiber number diminishes progressively from S8-10 in mys FCs. (F) Quantification of the number of whip-like structures at the leading edge of S7 control and mys migrating FCs. (G) Quantification of the number of actin fibers per μm in S9 control and mys FCs. (H) Quantification of relative F-actin intensity in stress fibers in S9 control and mys FCs. The statistical significance of differences was assessed with a t-test, *** P value < 0.0001. All error bars indicate s. e. Scale bars, 5 μm. The dotted white and red circles indicate area occupied by clones of mutant cells.

Fig 2. Loss of integrins results in reorganisation of the basal actin cytoskeleton.

(A-A”) Confocal micrographs of mys FC clones in living egg chambers expressing LifeActin-YFP (red) and the membrane marker Resille-GFP (green). mys FCs (GFP-negative) form abnormal actin-rich protrusions. (B-D) Time-lapse series of one representative mys FC (GFP-negative) labelled with Sqh-mCherry (red) and Resille-GFP (green). (D) The total area occupied by projections at different time points is coloured in orange. In blue, outline of the basal surface of the cell. (E, F) Simultaneous quantification of basal myosin changes and % of total basal surface occupied by projections (E) or total basal area (F) in one representative mys FC. Scale bars, 5 μm.

Fig 3. Loss of integrins in FCs results in increased membrane tension.

(A, B) Images of life S9 wild type (A) and mosaic egg chambers containing mys FC clones (GFP-negative) (B), expressing Resille-GFP, before and after single-cell bonds are ablated. Red bar and arrows indicate the point of ablation and the vertexes displaced, respectively. (C) Quantification of initial velocity of vertex displacement and (D) vertex displacement over time of the indicated ablated bonds. The statistical significance of differences was assessed with a t-test, *** P value < 0.0001. All error bars indicate s. e. Scale bars, 5 μm.

Fig 4. mys mutant FCs show defective basal surface expansion.

Basal (A) and apical (B) surface views of a S10A mosaic egg chamber containing mys FC clones (GFP-negative) and expressing the cell membrane marker Resille-GFP, stained with anti-GFP. (C) Lateral view of a mosaic FE stained with anti-GFP (green) and Rodamine Phalloidin to detect F-actin (red). (D, E, F) Box plots of the basal surface (D), apical surface (E) and height (F) of control (green) and mys (grey) S10 FCs. (G) Box plot of the basal area of control (green) and mys (grey) FCs at different stages of oogenesis. (H). Box plot of the nuclear size of control (green) and mys (grey) S10 FCs. The statistical significance of differences was assessed with a t-test, *** P value < 0.0001. Scale bars, 10μm.

Fig 5. Integrins regulate basal surface area by controlling F-actin levels at cell edges.

(A, A’, B, B’) Basal surface view of S10 mosaic FE containing mys FC clones (GFP-negative) and (B, B’) expressing an abi RNAi (tj>abiRNAi), stained for anti-GFP (green) and Rhodamine Phalloidin to detect F-actin (red). (C) Quantification of relative F-actin intensities along boundaries between cells of the indicated genotypes. (D) Quantification of the basal area of S10 FCs of the designated genotypes. (E) Quantification of initial velocity of vertex displacement of the indicated ablated cell bonds. The statistical significance of differences was assessed with a t-test, *** P value < 0.0001. All error bars indicate s. e. Scale bars, 5 μm.

Fig 6. Elimination of integrin in FCs disrupts cytoskeletal organisation in neighbouring control cells.

(A, B) Basal surface view of S10A (A, A’) and S10B (B, B’) mosaic follicular epithelia containing mys FC clones (GFP-negative), stained with anti-GFP (green) and Rhodamine Phalloidin to detect F-actin (red). (C) Lateral view of a S10 mosaic egg chamber stained with anti-GFP (green), Rhodamine Phalloidin (red) and anti-Dlg (basolateral polarity marker Discs large, blue). (D) 3D reconstruction of mys FCs and surrounding control cells. Arrows in C and D point to the basal surface of a mutant FC. (E, E’) Basal surface view of a S10B mosaic FE containing mys FC clones (GFP-negative). Yellow arrows and asterisks mark control FCs contacting control and mys FCs, respectively. (F-G’) Confocal images of live S10B mosaic egg chambers containing mys (F) or GFP (G) clones and expressing the cell membrane marker Resille-GFP. Images were taken with a 40 minutes interval. Scale bars, 5μm.