|
Fig 1. 4D lattice light sheet image processing.Each slice is composed from two fluorescent channels, the calcein (green) that marks the calcium and the FM4-64 (white) that marks cell membranes. Each time points (frame) is composed of 80â120 slices that are reconstructed to form a 3D image of the volume of the embryos, as demonstrated in S1 Movie. We then project the 3D image into a selected plane and generate movies showing 200â400 consecutive frames spanning about 20â30 minutes.
|
|
Fig 2. Examples of the LLSM 3D images of sea urchin embryos in different developmental stages and treatments.These representative images are 2D projections of the 3D rendered frames of selected datasets, as demonstrated in Fig 1. Calcein staining is marked in green and the FM4-64 membrane staining is marked in white. (A-C), control embryos (DMSO) at the early gastrula stage before spicule formation (A), just after the tri-radiate spicule forms (B) and when the spicule is elongated (C). (D-F) Representative embryos at gastrula stage treated with VEGFR inhibitor, axitinib, do not have spicules. Scale bars are 5μm. 3D movies showing the first 100â200 frames of each of the dataset presented in this figure are provided as S2âS7 Movies. Ecâectoderm, SCâskeletogenic cells, NSM- non-skeletogenic mesoderm, Sâspicule.
|
|
Fig 3. Sequences of time-lapse images demonstrating cellular dynamics in control and under VEGFR inhibition.(A) Time-lapse images of control embryo (S4 Movie) showing the rapid movement of a free mesenchymal cell (marked in arrowhead) compared to the stable position of the skeletogenic cells that are in direct contact with the ectoderm (asterisks). (B) Time lapse images of an embryo grown under VEGFR inhibition demonstrating the active filopodia extension and fusion between two skeletogenic cell clusters. Relative time from the beginning of the movie is shown in seconds at the top of each frame. Scale bar is 10μm. Ecâectoderm, Sâspicule, SCâskeletogenic cell, FâFilopodia.
|
|
Fig 4. Vesicle volume is larger in the skeletogenic cells compared to the ectodermal cells and is significantly larger under VEGFR inhibition.(A-C) An example for the image processing involved in the quantification of vesicle volume in the ectodermal vs. skeletogenic embryonic domains. (A) Raw image rendering of calcium vesicle detection. (B) Demonstration of the automated vesicle detection (segmentation) overlaid on the image in (A). (C) Manual identification of ectodermal region in red, rendered along with raw image frame. (D) Comparison of vesicle sizes in ectodermal and skeletogenic cells in control and VEGFR inhibition. The total number of vesicle measured in the control skeletogenic cells: 815, ectodermal cells: 3719; VEGFR inhibition skeletogenic cells: 1530, ectodermal cells: 8621. Each box plot shows the average (white square), median (middle line), the first and the third quartiles (the 25th and 75th percentiles, edges of boxes) and outliers (black dots). Vesicle volume is significantly higher in the skeletogenic cells compared to the ectodermal cells and VEGFR inhibition significantly increases vesicle volume in both cell types. (Dunn-Sidak test, p<0.0001, exact p-values are given in S1 Dataset).
|
|
Fig 5. Vesicle tracking reveals an active diffusion motion with higher diffusion coefficient and speed in the skeletogenic cells compared to the ectodermal cells.(A, B) Examples for the automated tracking used to quantify vesicle kinetics. (A) Instantaneous speed indicates the distance traveled between sequential frames divided by the time interval between the frame. The magenta line demonstrates this distance for the magenta labeled vesicle. (B) Directionality index is the ratio of maximal displacement (white line) over the total distance traveled (magenta line) within a one-minute time interval. A representative 3D movie of tracking session is provided in S9 Movie. (C-E) Comparison of vesicle motion statistics between control and VEGFR inhibited embryos. The total number of vesicle measured in the control skeletogenic cells: 825, ectodermal cells: 4193; VEGFR inhibition skeletogenic cells: 884, ectodermal cells: 4508. Each box plot shows the average (white square), median (middle line), the first and the third quartiles (the 25th and 75th percentiles, edges of boxes) and outliers (black dots). (C) Vesicle instantaneous speed. (D) Directionality index. (E) Vesicle diffusion coefficient. (Dunn-Sidak test, p<0.0001, exact p-values are given in S1 Dataset). (F) Histogram of Diffusion model fit, 90% of vesicle tracks are well modeled by the standard diffusion model (R2>0.8).
|
|
Fig 6. Vesicle velocity is not directed toward the spicule and vesicle speed is lower near the spicule.(A) An example for the manually identified spicule centerline shown in red with raw image data. (B) Average instantaneous velocity (μm/sec) toward the spicule relative to the average distance from spicule. Each box plot shows the average (white square) median (middle line), the first and the third quartiles (the 25th and 75th percentiles, edges of boxes) and outliers (gray diamonds). (C) Average instantaneous vesicle speed (μm/sec) at increasing average distances from the spicule (μm). The speed at distances 1â2μm are significantly lower than at distances >6μm (Dunn-Sidak test, p<0.001, exact p-values are given in S1 Dataset), with n = 803 vesicles.
|
|
Fig 7. Actin filaments are detected around the spicule and are enriched in the ectodermal cells compared to the skeletogenec cells.(A-F) Representative images showing actin filaments in normal embryos (A-C) and VEGFR inhibited embryos (D-F). Phalloidin was used to stain f-actin (A and D) and 6a9 was used to mark the skeletogenic cells (B and E). Arrow in A marks the spicule, arrowheads in A and D mark the apical side of the ectodermal cells. In C and F we present the overlay of the phalloidin and the skeletogenic marker, with indicated sections enlarged on the right. (G-H) quantification of the phalloidion signal. The number of red pixels (phallodin signal) per marked area was measured (G). The phallodin signal is significantly higher in the ectodermal cells compared to the skeletogenic cells and it unaffected by VEGFR inhibition (Dunn-Sidak test, p<0.0001, exact p-values are given in S1 Dataset). Based on 3 biological replicates where overall n = 24 normal embryos and n = 30 VEGFR inhibited embryos were studied.
|
|
Fig 8. Active myosinII signal is enriched in the ectodermal cells compared to the skeletogenec cells.(A-F) Representative images showing active myosinII (MyoIIP) in normal embryos and VEGFR inhibited embryos. (A, D) active MyosinII in normal and VEGFT inhibited embryos, respectively. Arrowhead indicates the apical side of the ectodermal cells. (B, E) skeletogenic cells marker (6a9) in normal and VEGFR inhibited embryos, respectively. (C, F) overlay of the two markers, with indicated sections enlarged on the right. (G-H) quantification of the active myosinII signal. The number of red pixels (active myosinII) per marked area was measured (G). The active myosinII signal is significantly higher in the ectodermal cells compared to the skeletogenic cells and is unaffected by VEGFR inhibition (Dunn-Sidak test, p<0.0001, exact p-values are given in S1 Dataset). Based on 3 biological replicates where overall n = 27 normal embryos and n = 30 VEGFR inhibited embryos were studied.
|
