Cooperation of polarized cell intercalations drives convergence and extension of presomitic mesoderm during zebrafish gastrulation

Abstract

During vertebrate gastrulation, convergence and extension (C&E) movements narrow and lengthen the embryonic tissues, respectively. In zebrafish, regional differences of C&E movements have been observed; however, the underlying cell behaviors are poorly understood. Using time-lapse analyses and computational modeling, we demonstrate that C&E of the medial presomitic mesoderm is achieved by cooperation of planar and radial cell intercalations. Radial intercalations preferentially separate anterior and posterior neighbors to promote extension. In knypek;trilobite noncanonical Wnt mutants, the frequencies of cell intercalations are altered and the anteroposterior bias of radial intercalations is lost. This provides evidence for noncanonical Wnt signaling polarizing cell movements between different mesodermal cell layers. We further show using fluorescent fusion proteins that during dorsal mesoderm C&E, the noncanonical Wnt component Prickle localizes at the anterior cell edge, whereas Dishevelled is enriched posteriorly. Asymmetrical localization of Prickle and Dishevelled to the opposite cell edges in zebrafish gastrula parallels their distribution in fly, and suggests that noncanonical Wnt signaling defines distinct anterior and posterior cell properties to bias cell intercalations.

Figure 1.

Patterns of C&E movements in the axial and presomitic mesoderm. (A and D) Lateral views of live embryos at the one-somite stage (10.5 hpf). The regions where the time-lapse recordings were taken are indicated by brackets. (B and E) Snapshots of PIV analyses of C&E movements at the one-somite stage, showing the tissue displacements within a 9-min time window. Dorsal views, anterior to the top. Arrows indicate the direction of cell movements, whereas the length of the arrows represents the movement speed. (C and F) Same images as shown in B and E, except that different movement speeds are presented in a color-coded manner. Purple lines delineate the axial mesoderm. (G–J) Snapshots of the analyzed medial PSM cell population at the beginning and the end of the time-lapse recordings. Dorsal views, anterior to the top. Red arrows illustrate the maximum length and width of the analyzed cell population when the time lapse started. (K) Quantification of the tissue shape changes (Materials and methods). 10 embryos of each genotype were analyzed. Error bars represent the standard error. *, P < 0.05; **, P < 0.005, mutant vs. WT. A, anterior; P, posterior; Axial, axial mesoderm. Bars: (A and D) 100 μm; (B, C, E, F, and G–J) 20 μm.

Figure 2.

C&E movements of the medial PSM entail multiple cell intercalation behaviors. (A) Schematic of the analyzed cell population in WT. Dorsal views. Cells leaving the analyzed layer during the recording are shown in blue; cells entering the layer are shown in red. Six cells are assigned numbers to illustrate the significant neighbor exchanges. (B–E) Schematics and frequencies of planar medial intercalation, coordinated radial/medial intercalation, radial AP intercalation, and radial ML intercalation, respectively. Medial is to the left. (F) Frequencies of the three types of radial intercalations listed in C–E in WT, kny;tri double mutants, and embryos injected with synthetic RNA encoding Xdd1. (B–F) Error bars represent the standard error. A, anterior; M, medial. *, P < 0.05; **, P < 0.005, mutant vs. WT.

Figure 3.

Computational modeling of cell intercalation behaviors during the medial PSM C&E. (A) Tissue shape changes predicted by the control simulation. Each blue block represents a single cell. (B) Quantitative comparison of the C&E rates calculated from the WT time-lapse analyses and the control simulations. Error bars represent the standard error. P > 0.3, simulated results vs. in vivo observation. (C–F) Comparison of the results from the control simulations and the simulations in which the cell behaviors were modified based on the kny;tri double-mutant data in one of the four ways: reducing the total amount of intercalations (C); increasing radial ML intercalation (D); reducing planar medial, radial AP, and coordinated radial/medial intercalations (E); and reducing planar medial, radial AP, and coordinated radial/medial intercalations, while increasing radial ML intercalation (F). 20 independent simulations were performed for the control and each modification. Error bars represent the standard error. A, anterior; M, medial. *, P < 0.05; **, P < 0.005; ***, P < 0.0005.

Figure 4.

Subcellular localization of GFP-Pk during C&E. (A, and D–G) Confocal images of the dorsal mesoderm in live embryos expressing GFP-Pk and membrane-RFP (memRFP) at the tailbud stage (10 hpf). Dashed lines illustrate the boundary between the axial and presomitic mesoderm. (B and C) GFP-Pk localization in the dorsal mesoderm of the same WT embryo at 8 hpf (B) and 10 hpf (C). (H) Quantification of GFP-Pk localization at the tailbud stage in WT and embryos deficient in noncanonical Wnt signaling (Materials and methods). (A–G) Dorsal views. A, anterior; P, posterior; Axial, axial mesoderm. Bar, 20 μm.

Figure 5.

Subcellular localization of GFP-Dsh in the axial mesoderm during C&E. (A–D) Confocal images of the axial mesoderm in live embryos expressing GFP-Dsh at the tailbud stage (10 hpf). (A′–D′) Confocal images of the same embryos as shown in A–D, but colabeled with memRFP. Histograms of the fluorescent intensity of the selected cells (*) are shown in the right panels. Green lines represent the fluorescent intensity of GFP-Dsh and the red lines represent the intensity of memRFP. (E, F, E′, and F′) GFP-Dsh localization in the axial mesoderm of the same WT embryo. Arrows show that GFP-Dsh was mainly localized in the cytoplasm at 75% epiboly (8 hpf) (E and E′), but became restricted to the posterior cell membrane at the tailbud stage (F and F′). (G) Quantification of GFP-Dsh localization in the axial mesoderm at the tailbud stage (Materials and methods). (A–F and A′–F′) Dorsal views. A, anterior; P, posterior. Bar, 20 μm.