Dumortier et al., 2012 - Collective mesendoderm migration relies on an intrinsic directionality signal transmitted through cell contacts. Proceedings of the National Academy of Sciences of the United States of America   109(42):16945-16950 Full text @ Proc. Natl. Acad. Sci. USA

Fig. 1 All prechordal plate cells show similar migrating properties. (A) Dorsal (A) or reconstructed lateral view (A′) of a Tg(gsc:GFP) embryo injected with membrane mCherry RNAs. The green dashed line delineates the prechordal plate. (B) Tracks of prechordal plate nuclei in dorsal (B) or lateral (B′) view. One representative embryo, tracked for 18 min. (C) Field of instantaneous speed in the prechordal plate (Materials and Methods). Speed norm is color-coded. Two-dimensional projections on the three planes have been plotted. (DI) Front (D, D′, and G), posterior (E, E′, and H), or inner (F, F′, and I) cells emit cytoplasmic extensions containing actin (DF). Location of the cell is assessed by lateral view reconstruction (D′–F′). These extensions are oriented toward the animal pole (GI). Angles between the axis of the extension and the direction of the animal pole are plotted as histograms, with “0” thus representing an extension pointing to the animal pole. (JO) Control cells (J) or cells injected with the DN-PI3K (K) were transplanted from shield to shield. (L and M) Orientations of the cytoplasmic extensions relative to the animal pole. (N) Frequencies of cytoplasmic extensions. (O) PH-mCherry accumulation at the anterior pole of a cell within the plate [assessed by GFP expression, as the host embryo is Tg(gsc:GFP); green channel not displayed]. Animal pole is to the left.

Fig. 2 Cell–cell contacts are required to orient migration. (A) Diagram of the experimental procedure. (B and C) A single induced plate cell expressing GFP (red in C) was transplanted ahead of the prechordal plate, at the onset of gastrulation. Although it is isolated, this cell keeps expressing goosecoid (B, black arrowhead). Here the embryo was fixed approximately 45 min after transplantation, just before the transplanted cell is contacted by the endogenous plate. Dorsal view is shown, with animal pole to the top. (D) A single induced plate cell expressing Lifeact-mCherry was transplanted ahead of the endogenous prechordal plate, in a Tg(gsc:GFP)embryo. Animal pole is to the left. Arrowhead points to an actin accumulation. Asterisk marks the position of the cell at the onset of the movie (Movie S7). (E and G) Tracks of transplanted cells while they remain isolated (E) or have been contacted by the endogenous plate (G). Position of the cell at the onset of the movie (E) or at the time it is contacted by the plate (G) has been set to position (0,0). (F and H) Orientation of the cytoplasmic extensions relative to the animal pole, when cells are isolated (F) or have been joined by the endogenous plate (H).

Fig. 3 E-cadherin, Dsh/PCP, and Rac1 are required for cell orientation. Cells injected with Lifeact-mCherry RNAs and a control morpholino (A and B), E-cadherin morpholino (C and D), DshDep+ RNAs (E and F), or DN-Rac1 RNAs (G and H) were transplanted from shield to shield in Tg(gsc:GFP) embryos. Transplanted cells are within the prechordal plate (assessed by GFP expression), and contour of the plate is delineated (white line) when visible in the frame. Orientations of their cytoplasmic extensions were measured relative to the animal pole and plotted as histograms (B, D, F, and H) or cumulative plot (J). Frequencies of cytoplasmic extensions are also plotted (I).

Fig. 4 Contact with the endogenous prechordal plate is required to orient migration. (A) Small groups of induced plate cells expressing Lifeact-mCherry were transplanted ahead of the endogenous prechordal plate in a Tg(gsc:GFP) embryo. Animal pole is to the left. Arrowheads point at cytoplasmic extensions. (B and D) Tracks of transplanted cells within isolated groups (B; Movie S8) or when the group has been contacted by the endogenous plate (D). Position of the cell at the onset of the movie (B) or at the time the group is contacted by the plate (D) has been set to position (0,0). (C and E) Orientation of the cytoplasmic extensions relative to the animal pole, when the group is isolated (C) or when it has been joined by the endogenous plate (E). (F) Orientation of cytoplasmic extensions relative to animal pole of cells located on lateral sides of the plate. If protrusion orientation was driven by contact inhibition alone, extensions should be perpendicular to the animal pole direction. Here they are oriented as in the rest of the plate (P > 0.05, n = 88 extensions). (G) Model of collective prechordal plate migration. All plate cells dynamically migrate toward the animal pole. Directionality is obtained by transmission of intrinsic information through cell–cell contacts. When isolated, cells show an autonomous ability to polarize, but they lose directionality.

Fig. S2 Activation of the Nodal pathway in the absence of casanova drives cells to a prechordal plate fate. (A and B) Injection of one cell at the 16-cell stage with RNAs encoding the constitutive form of the Nodal receptor (Tar*) leads cells to take part in endodermal derivatives (arrows) and hatching gland (arrowhead, A). When casanova expression is simultaneously blocked by a morpholino, injected cells take part in the hatching gland (B; n = 26 of 29 embryos). (C–H) Cells injected with Tar* RNAs and casanova morpholino were transplanted to the lateral margin of a host embryo at the onset of gastrulation. During gastrulation, these cells express the prechordal plate marker goosecoid (C and D), and take part in the hatching gland at 24 hours post-fertilization (E; n = 15 of 16 embryos). (F–H) During gastrulation, these cells migrate in group toward the animal pole (F–H; n = 19 of 21 embryos). Labeling their membrane and actin cytoskeleton (Lifeact) shows that they emit actin-rich extensions (white arrowheads), with the same frequency (0.216 extensions·cell1·min1) and orientation (H) as endogenous plate cells (P = 0.43, n = 357 and n = 537 extensions; compare Fig. S2H vs. Fig. 1L). Animal pole is to the top (C and D) or left (F and G).

Fig. S3 Role of RhoA and Cdc42 on cytoplasmic extensions orientation. Cells injected with RNAs encoding Lifeact-mCherry and DN-RhoA (A and B) or DN-Cdc42 (D and E) were transplanted from shield to shield, in Tg(gsc:GFP) embryos. Transplanted cells are within the prechordal plate (assessed by GFP expression), and contour of the plate is delineated (white line). Orientations of their cytoplasmic extensions were measured relative to the animal pole and plotted as histograms (B and E) or cumulative plot (F). Frequencies of cytoplasmic extensions are plotted in C. No difference could be observed between control cells and cells expressing DN-Cdc42 (P = 0.84, Kolmogorov–Smirnov test; n = 537 and n = 269 extensions). Expression of DN-RhoA affected the distribution of extension orientations (P = 6.105, Kolmogorov–Smirnov test; n = 537 and n = 524 extensions), even though it did not affect the average orientation (P = 0.35, Student t test). Animal pole is to the left in A and D.

Acknowledgments:
This image is the copyrighted work of the attributed author or publisher, and ZFIN has permission only to display this image to its users. Additional permissions should be obtained from the applicable author or publisher of the image. Full text @ Proc. Natl. Acad. Sci. USA