Cellular processes of Ciona atrial siphon tube invagination.

(A) Representative images of atrial siphon morphogenesis in Ciona embryos from 13.5 hpf to 18 hpf. The white arrow indicates the central cell. Scale bar: 10 μm. (B) Measurement parameters of the Ciona atrial siphon. The cell undergoing the most prominent apical constriction at the center of invagination was defined as the center cell (0). The adjacent cells on the left and right were defined sequentially as −1, −2, −3, −4 and +1, +2, +3, +4, respectively. (C) Quantification of the invagination depth in the atrial siphon of Ciona embryos. Red lines indicate linear regression fits of invagination depth during the initial (13.5-16 hpf, k = 0.2617) and accelerated (16-18 hpf, k = 2.7920) stages. n = 20. (D, E) Quantification of the center cell height and apical-to-basal area ratio at the atrial siphon of Ciona embryos. The blue-shaded region represents the initial stage, while the orange-shaded region indicates the accelerated stage. N = 20.

Bidirectional redistribution of actomyosin between apical and lateral domains during atrial siphon tube invagination.

(A) Representative images of Ciona embryos stained for active myosin II (anti-pS19 MRLC, red) and F-actin (green) at different stages of atrial siphon invagination. Scale bar: 10 μm. (B) Normalized fluorescence intensity of active myosin II (anti-pS19 MRLC) and F-actin at the apical and lateral regions of center cells during different stages of atrial siphon invagination (basal level set to 1). The blue-shaded region represents the initial stage, while the orange-shaded region indicates the accelerated stage. **p < 0.01, ***p < 0.001, ****p < 0.0001, n = 20. (C) Schematic model illustrating the mechanical forces driving atrial siphon primordium invagination. Brown arrows indicate the direction of contractile forces; red areas depict active myosin II localization.

Modulation of atrial siphon invagination by overexpression of myosin mutants

(A) Representative images of Ciona atrial siphon primordium expressing wild-type and mutant MRLC constructs. Scale bar: 10 μm. (B) Quantification of invagination depth and the center cell height in MRLC (T18ES19E), MRLC (T18AS19A), and MRLC groups. The blue-shaded region represents the initial stage, while the orange-shaded region indicates the accelerated stage. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Disruption of contractile forces during rapid invagination of the Ciona atrial siphon using an optogenetic system

(A) Schematic diagram depicting the structure and mechanism of the MLCP-BcLOV4 system. The PP1C::MYPT169::BcLOV4::mCherry::NES fusion protein is initially dispersed in the cytoplasm. Upon exposure to blue light, BcLOV4 undergoes a conformational change, allowing it to interact electrostatically with the plasma membrane. This leads to the recruitment of the PP1C and MYPT169 components, subunits of myosin light chain phosphatase (MLCP), to the membrane, where they reduce myosin activity. (B) Representative images of developmental progression in the MLCP-BcLOV4 expression group exposed to blue light for 1 h, and in the dark control group maintained in darkness for 1 h. Scale bar: 10 μm. (C, D) Quantification of invagination depth and the center cell height in the MLCP-BcLOV4 expression group, the dark control group and MLCP control group (Figure 4—figure supplement 1). *p < 0.05, ***p < 0.001, ****p < 0.0001.

Vertex model simulations of siphon invagination.

(A) Schematic of cell-based vertex model. The epithelial tissue is represented by polygonal cells connected in a circular arrangement, surrounding inner supporting cells. The tissue is partitioned into an active region (consisting of cells with indices −3 to +3) with localized active forces, and an inactive region without actomyosin intensity. The effective energy U takes into cell area constraint (modulus KA), passive cortical contraction (coefficient KC), tissue surface bending (stiffness KB), tissue area constraint (modulus ), and active line tensions on apical and lateral edges (coefficient km). (B) Temporal evolution of actomyosin intensities at apical, lateral, and basal domains of the center cell in the model, parameterized based on experimentally measured F-actin distribution. (C) Evolution curves of invagination depth (orange) and center cell height (blue) in simulations. Dots represent corresponding experimentally measured values. (D) Snapshots at specific timepoints during simulation progression (bottom), compared with corresponding experimental images (up). Colors on cell edges represent the normalized actomyosin intensity.

Mechanics of siphon invagination revealed by vertex model simulations.

(A, B) Invagination depth (A) and center cell height (B) regulated by changes in actomyosin intensities under the scaling factor α at both apical and lateral regions of all active cells. The data points are from the mutant experiments in Figure 3B. (C) Contributions of apical and lateral contractility. Simulations lacking lateral (left) or apical (right) activity result in shallower invagination or outward evagination, respectively, which highlight the distinct roles of apical constriction in initiating apico-basal imbalance and lateral contraction in driving deep folding. (D) Effects of tissue size and bending stiffness. With the active region size held constant, the global tissue curvature decreases as the total cell number increases. For N= 108, only the upper half of the tissue at KB = 0.0001 is shown. For N = 324, a portion of the tissue is shown at KB = 0.0001, and only the region near the active region is shown at KB = 1.

Quantification of F-actin intensity and intercellular distance during Ciona atrial siphon morphogenesis.

(A) Normalized F-actin intensity at the apical and lateral regions of center cells during Ciona atrial siphon morphogenesis (basal level set to 1). (B) Quantification of the linear distance between the −3/−4 and +3/+4 cell junctions at the apical or basal surface in the atrial siphon of Ciona embryos. The blue-shaded region represents the initial stage, while the orange-shaded region indicates the accelerated stage. Representative images are shown in Figure 1A. n = 20.

EdU and TUNEL staining during Ciona atrial siphon morphogenesis.

(A) Representative images of EdU staining at 14-15 hpf. Orange: EdU-positive nuclei indicating cell proliferation. Blue: DAPI. No EdU signal was detected in the atrial siphon primordium (white dashed outline). Scale bar: 10 μm. n = 10. (B) Representative images of TUNEL staining at 15 hpf. (B1) Positive control: DNase I pretreatment (20 U/mL, 10 min) induced DNA fragmentation. (B2) Negative control: staining performed without terminal deoxynucleotidyl transferase (TdT) enzyme. (B3) Experimental group: no detectable TUNEL signal in the atrial siphon primordium (white dashed outline). Scale bar: 10 μm. n = 10.

MRCL control group in the optogenetic experiment

(A) Schematic diagram depicting the structure and mechanism of the MLCP control system. The PP1C::MYPT169::mCherry::NES fusion protein remained diffuse in the cytoplasm under light exposure and failed to function. (B) Representative images of developmental progression in the MLCP control group exposed to blue light for 1h. Scale bar: 10 μm.

Optogenetic inhibition effectively reduces the activity of lateral myosin.

(A) Representative images of active myosin II (anti-pS19 MRLC, red) and F-actin (green) staining in the MLCP-BcLOV4-expressing embryos after 1 h of light exposure or dark control. Scale bar: 10 μm. (B) Quantification of normalized active myosin II intensity at the apical and lateral domains of the center cell. **p < 0.01.

Entire head morphology of Ciona embryo.

(A, B) Bright-field images showing the entire head of Ciona embryos at 13 hpf and 18 hpf, illustrating how the tissue geometry becomes flatter during growth. Black arrow indicates the position of atrial siphon. Scale bar: 100 μm.

Key resource table

Primer sequences used in this study