A force balance can explain local and global cell movements during early zebrafish development

Abstract

Embryonic morphogenesis takes place via a series of dramatic collective cell movements. The mechanisms that coordinate these intricate structural transformations across an entire organism are not well understood. In this study, we used gentle mechanical deformation of developing zebrafish embryos to probe the role of physical forces in generating long-range intercellular coordination during epiboly, the process in which the blastoderm spreads over the yolk cell. Geometric distortion of the embryo resulted in nonuniform blastoderm migration and realignment of the anterior-posterior (AP) axis, as defined by the locations at which the head and tail form, toward the new long axis of the embryo and away from the initial animal-vegetal axis defined by the starting location of the blastoderm. We found that local alterations in the rate of blastoderm migration correlated with the local geometry of the embryo. Chemical disruption of the contractile ring of actin and myosin immediately vegetal to the blastoderm margin via Ca(2+) reduction or treatment with blebbistatin restored uniform migration and eliminated AP axis reorientation in mechanically deformed embryos; it also resulted in cellular disorganization at the blastoderm margin. Our results support a model in which tension generated by the contractile actomyosin ring coordinates epiboly on both the organismal and cellular scales. Our observations likewise suggest that the AP axis is distinct from the initial animal-vegetal axis in zebrafish.

Figure 1

Mechanical deformation probes the role of CCT in zebrafish epiboly. (A) (i) Schematic representation of epiboly. The blastoderm mass (green) begins at the animal pole of the yolk (yellow). Over time the blastoderm spreads over the yolk (ii (50% epiboly)) until it converges at the vegetal pole (iii). An actomyosin contractile band (red) forms just vegetal of the blastoderm margin. (B) Schematic depiction of directional compression, showing 1.4% w/v agarose (dark blue) and 0.8% w/v low-melting-point agarose (light blue). (i and ii) Agarose overlays are applied directionally to generate a shear direction. (iii) Embryos mounted in agarose are initially symmetric. (iv) Addition of an agarose overlay gently compresses the embryos, causing them to become asymmetric and introducing a new long axis at an angle relative to the AV axis. The shear direction in (iv) is rotated 90° into the page from (i) and (ii). Scale bar, 250 μm.

Figure 2

Development of a compressed embryo from the four-cell stage to the conclusion of epiboly shows alterations in the local rate of epiboly and reorientation of the embryonic AP axis. (A–F) Compressed embryo in the four-cell stage (A), immediately after the MBT (B), at 30% epiboly (C), at 70% epiboly (D), and immediately after blastopore closure (E). The trailing, right edge migrates farther than the leading, left edge. Red triangles indicate the initial AV axis orientation and blue triangles the AP axis. (F) Light blue lines connect the left and right edges of the blastoderm margin at successive time points. Faster migration along the right side of the embryo causes the center of the blastoderm margin (dark blue line) to rotate relative to the initial AV axis (black dashed line). (G–I) Uncompressed embryo at the one-cell stage (G), at 50% epiboly (H; note that both edges have progressed equal distances at this time), and immediately after blastopore closure (I). Red triangles represent the initial AV axis and blue triangles the AP axis at the conclusion of epiboly. Note that in the uncompressed embryos, the AV and AP axes are aligned. Scale bar, 250 μm. (See also Movie S1.)

Figure 3

Differential BM migration can be quantified via a simple force balance. (A) The free body diagram of forces at each edge of the blastoderm margin. Red arrows indicate a constant driving tangential force, F_Tan. Blue arrows indicate the force generated by CCT. The vectors F_Tan and F_CCT form the angle θ. (See also Fig. S4 and Movies S2 and S3.) (B) The projection of F_CCT (blue arrows) adds to F_Tan on the right edge and counteracts F_Tan on the left edge, creating a larger net force at the right edge of the blastoderm margin (purple arrows). (C) Experimental data shown in blue are Δv and Δcosθ calculated at 30 min intervals between 30% and 80% epiboly in 16 embryos. A seven-point moving average is shown in green to indicate the overall trend. A linear fit of Δv versus Δcosθ is shown in red (R² – 0.12), which yields a slope of 23 ± 6 nm s⁻¹ (p-value < 0.001 equivalent to 0 nm s⁻¹) and intercept of 0.30 ± 1.56 nm s⁻¹, with uncertainties determined based on the least-squares fit of the raw data (blue) to a line (red). E2(+): E2 medium including 1 mM Ca²⁺.

Figure 4

Disrupting CCT eliminates differential blastoderm migration and eliminates axis realignment. (A) An asymmetric embryo grown in reduced, 62.5 μM Ca²⁺ media. (i) The initial blastoderm mass defines the AV axis (red arrows). (ii) The location of blastopore closure and eventual AP axis (blue arrows) coincide with the initial AV axis (red arrows). (iii) The migration pattern of the blastoderm shows nearly uniform migration (red parallel lines) over the course of epiboly. Consequently, migration of the center of the blastoderm margin (magenta transverse line) lies very near the AV axis (dashed black line; cf. Fig. 1C). (B) (i and ii) Quantification of the CCT force which is equivalent to the slope of the linear fit. The slope of the linear fit is statistically equivalent to 0 when CCT is disrupted by lowering Ca²⁺ to 62.5 μM (i; E2(−), −4 ± 5 nm s⁻¹, R² – 0.005, p-value = 0.82) or by introducing 13.8 μM blebbistatin (ii; Bleb, 3 ± 5 nm s⁻¹, R² – 0.003, p-value = 0.56). (iii) F_CCT/c_Drag is statistically significantly lower (^∗∗∗p < 0.001) in embryos grown in E2(−) media or with introduced blebbistatin compared to embryos grown in E2(+) media. Error bars represent the standard deviation of the slope of the fit as described in Fig. 3C. (iv) Overall rates of epiboly in E2(+), E2(−), and E2 with 13.8 μM blebbistatin. The overall rate of epiboly in E2(−) is statistically equivalent to that observed in E2(+), whereas introduction of blebbistatin severely delays epiboly. The rates of epiboly reflect the mean ± SD of the average rate of epiboly over multiple embryos (16 for E2(+), 15 for E2(−), and 11 for Bleb). (See also Fig. S6, Movies S4 and S5, and Table S1 for more detailed fit parameters.)

Figure 5

Reduction of CCT in E2(−) media reduces the coordination of neighboring cells at the blastoderm margin. (A and B) View of the blastoderm margin in TMR-phalloidin-stained embryos fixed at 60% epiboly after developing in E2(+) (A) or E2(−) (B). Cell-to-cell coordination is markedly reduced in low Ca²⁺ media. The apparent striations in (A) and (B) result from finite slice spacing in the confocal images (see also Fig. S2 and Movie S6). (C) The distribution of cell-cell correlation vectors shows a statistically significant increase in margin roughness for embryos grown in E2(−). Red line, mean; box, first and third quartiles; whiskers, 2.7 standard deviations. (D) Schematic of tension along the blastoderm margin. Tension (blue arrows) generates a net restorative force (green arrow) against local perturbations. Scale bars, 40 μm.