Cell movement during chick primitive streak formation
Manli Chuai, Wei Zeng, Xuesong Yang, Veronika Boychenko, James A. Glazier and Cornelis J. Weijer
Tiens, tiens, il n’y a pas que moi qui voit seulement « two counter-rotating vortices ». Il y en a d’autres.
Looking forward to see L2/R2.
Cell movement during primitive streak initiation involves two counter-rotating vortices
Previously we have shown using optical flow image processing techniques and by following the trajectories of multiple groups of DiI labelled cells in prestreak chick embryos that the tissue movements occurring during streak formation are detectable from pre-streak stages onwards and are organised in two counter rotating cell flow patterns that meet at the site of streak formation (Cui et al., 2005). To be able to analyse these cell movements at the individual cell level, we developed an electroporation protocol that allowed us to transfect scattered cells in the early chick–embryo epiblast cells with GFP expression constructs. Our electroporation protocol typically results in the transfection of several hundred to thousands of cells in the area pellucida and area opaca in a mosaic pattern (Figs. 1A, B, C). Sectioning of the early transfected embryos confirmed that the majority of the cells (>90%) transfected were found in the epiblast. (Figs. 1D, E). GFP fluorescence of expression constructs under the control of the CMV promoter becomes visible 2–4 h after transfection and the fluorescence intensity/cell normally increases during the next 5–10 h of incubation. This labelling method allowed us to track the movement and division of individual cells for up to 24 h of development. Every track shown in Figs. 2C, F, I is the movement trajectory of an individual GFP-expressing cell, over a 4-h period. These cell tracking experiments confirmed the existence of the two counter-rotating cell flows in the epiblast, which merged at the site for the future streak primitive streak initiation (Fig. 2). The flows become visible well before any optically-dense structure in the streak are detected. Cells overlaying Koller’s sickle move into the streak and are being replaced by cells from more antero-lateral positions. Cells just anterior to the Sickle move forward and sideways into the area destined to become the neural plate (Hatada and Stern, 1994). The cells in the centres of both counter-rotating flows move relatively little. These centres do not coincide with any known morphological structures, suggesting that they are not special signalling centres. Speeds of cell movement vary from 0 to 2 μm/min, with the highest speeds at the periphery of the vortices. These data confirm our previous observations using DiI labelling (Cui et al., 2005), with better spatial coverage and at cellular resolution.
Gastrulation in amniotes begins with extensive re-arrangements of cells in the epiblast resulting in the formation of the primitive streak. We have developed a transfection method that enables us to transfect randomly distributed epiblast cells in the Stage XI–XIII chick blastoderms with GFP fusion proteins. This allows us to use time-lapse microscopy for detailed analysis of the movements and proliferation of epiblast cells during streak formation. Cells in the posterior two thirds of the embryo move in two striking counter-rotating flows that meet at the site of streak formation at the posterior end of the embryo. Cells divide during this rotational movement with a cell cycle time of 6–7 h. Daughter cells remain together, forming small clusters and as result of the flow patterns line up in the streak. Expression of the cyclin-dependent kinase inhibitor, P21/Waf inhibits cell division and severely limits embryo growth, but does not inhibit streak formation or associated flows. To investigate the role off cell–cell intercalation in streak formation we have inhibited the Wnt planar-polarity signalling pathway by expression of a dominant negative Wnt11 and a Dishevelled mutant Xdd1. Both treatments do not result in an inhibition of streak formation, but both severely affect extension of the embryo in later development. Likewise inhibition of myosin II which as been shown to drive cell–cell intercalation during Drosophila germ band extension, has no effect on streak formation, but also effectively blocks elongation after regression has started. These experiments make it unlikely that streak formation involves known cell–cell intercalation mechanisms. Expression of a dominant negative FGFR1c receptor construct as well as the soluble extracellular domain of the FGFR1c receptor both effectively block the cell movements associated with streak formation and mesoderm differentiation, showing the importance of FGF signalling in these processes.