The migration of paraxial and lateral plate mesoderm cells emerging from the late primitive streak is controlled by different Wnt signals

Abstract

Background: Co-ordinated cell movement is a fundamental feature of developing embryos. Massive cell movements occur during vertebrate gastrulation and during the subsequent extension of the embryonic body axis. These are controlled by cell-cell signalling and a number of pathways have been implicated. Here we use long-term video microscopy in chicken embryos to visualize the migration routes and movement behaviour of mesoderm progenitor cells as they emerge from the primitive streak (PS) between HH stages 7 and 10.

Results: We observed distinct cell movement behaviours along the length of the streak and determined that this is position dependent with cells responding to environmental cues. The behaviour of cells was altered by exposing embryos or primitive streak explants to cell pellets expressing Wnt3a and Wnt5a, without affecting cell fates, thus implicating these ligands in the regulation of cell movement behaviour. Interestingly younger embryos were not responsive, suggesting that Wnt3a and Wnt5a are specifically involved in the generation of posterior mesoderm, consistent with existing mouse and zebrafish mutants. To investigate which downstream components are involved mutant forms of dishevelled (dsh) and prickle1 (pk1) were electroporated into the primitive streak. These had differential effects on the behaviour of mesoderm progenitors emerging from anterior or posterior regions of the streak, suggesting that multiple Wnt pathways are involved in controlling cell migration during extension of the body axis in amniote embryos.

Conclusion: We suggest that the distinct behaviours of paraxial and lateral mesoderm precursors are regulated by the opposing actions of Wnt5a and Wnt3a as they leave the primitive streak in neurula stage embryos. Our data suggests that Wnt5a acts via prickle to cause migration of cells from the posterior streak. In the anterior streak, this is antagonised by Wnt3a to generate non-migratory medial mesoderm.

Figure 1

DiI labelling in HH8 embryos reveals behaviour and trajectories of cells from the primitive streak. (A, B, C) Long-term time-lapse imaging of embryos labelled in Hensen's node (A), the anterior (B) or posterior primitive streak (C) with DiI. Embryos were imaged for 20 hours and still pictures are shown at intervals of 2.5 hours. (D, E, F) Transverse sections of embryos labelled in Hensen's node (D), anterior (E) or posterior streak (F). (G) Summary of cell fates. Red represents Hensen's node and notochord (nc), green represents anterior primitive streak cells and paraxial mesoderm (pm) and blue represents posterior primitive streak cells and lateral plate mesoderm (lpm). n, node; nc, notochord; pm, paraxial mesoderm; s, somites. In (A) asterisks indicate the position of the node which regresses during axis extension. In (A, B, C) arrows show the most recently formed somite and arrowheads in (B, C) indicate the position of the most anterior DiI-labelled cells.

Figure 2

Differential expression of Wnt ligands and Wnt pathway components in the late primitive streak. (A) Wnt3a in situ hybridisation, white lines indicate the levels of section in (B) through the anterior streak with staining in the neural plate (C), section through the posterior streak with low levels of Wnt3a transcripts. (D) Wnt5a in situ hybridisation, white lines indicate the levels of section in (E) through the anterior and (F) through the posterior streak indicating widespread expression of Wnt5a in neural plate and mesoderm. (G) Wnt8c in situ hybridisation, white lines indicate the levels of section in (H) through the anterior and (I) through the posterior streak indicating expression of Wnt8c in neural plate and mesoderm. (J) Dishevelled-1 was not expressed in posterior regions of HH9 embryos but (K) dishevelled-3 and (N) prickle-1 showed similar expression in neural folds, lateral plate mesoderm and low levels in the primitive streak White lines indicate the levels of section in L, O (anterior), M and P (posterior) streak of dishevelled-3 and prickle-1. HH stage is indicated on each panel. hn, Hensen's node; m, mesoderm; nf, neural fold; np, neural plate; nt, neural tube; ps, primitive streak; * asterisks indicate position of the node.

Figure 3

Ectopic Wnt3a alters the movement behaviour of posterior primitive streak cells in vivo. (A) Injection of DiI (red) and DiO (green) into the anterior and posterior primitive streak followed by long-term time-lapse video microscopy shows the normal cell movement patterns of paraxial and lateral plate mesoderm progenitors. (B) A pellet of cells expressing Wnt3a adjacent to the anterior primitive streak had no effect on cell movement behaviour. (C) Implanting Wnt3a cells into the posterior streak blocked cell migration and production of lateral plate mesoderm in HH7-9 embryos. (D) Wnt3a cells implanted into the posterior streak at HH6 did not affect cell migration. (E) Graphical illustration of the effects of Wnt cell pellets on primitive streak cell behavior; grey shading indicates normal behavior, black shading indicates altered behavior. Only posterior grafts of Wnt3a altered normal movements in HH7-9 embryos. Asterisks indicate the position of the cell pellet at 0 h (B, C, D). White arrows indicate the most recently formed somite and arrowheads indicate the most anterior DiI and DiO labelled cells that have left the primitive streak.

Figure 4

Posterior grafts of Wnt3a cell pellets do not affect cell fate. (A-C) Expression patterns after grafting Rat-B1a-LNCX control cells. (D-F) Expression patterns after grafting Rat-B1a-Wnt3a cells. (A, D) Brachyury, (B, E) Lef1, (C, F) Wnt8c. nc, notochord; m, mesoderm; psm, presegmented mesoderm; lpm, lateral plate mesoderm.

Figure 5

The behaviour of anterior and posterior primitive streak explants is affected by Wnt3a and Wnt5a. Embryos were electroporated with pCS2⁺GFP and following GFP expression explants were dissected from anterior and posterior primitive streak as indicated in the drawings. Explants were placed on the area opaca of a host embryo (HH4). Labelled cells were imaged overnight and still images are shown at zero and 16 hours. Timelapse sequences were scored as migratory if they contained individual cells migrating away from the explant. χ² analysis was used to determine if changes in explant behaviour were statistically significant. (A) Control explants from a HH5 embryo; cells are migratory. (B) Explants from the anterior primitive streak of a HH8 embryo; cells remained cohesive and few individual cells moved away from the explants. This behavior was not affected by Wnt3a expressing cells, but in response to Wnt5a the explants dispersed and more migrating cells were observed. Anterior explants exposed to both Wnt3a and Wnt5a remained cohesive. (C) Explants from the posterior primitive streak of a HH8 embryo contained many cells migrating away and most of the explant dispersed by the end of imaging. In the presence of Wnt3a expressing cells posterior explants remained cohesive. Wnt5a expressing cells had no effect and migrating cells were observed. Posterior explants exposed to both Wnt3a and Wnt5a remained cohesive. Arrowheads indicate cells that have migrated from explants, circles indicate positions of cell pellets. (D) Summary of explant data. χ² analyses showed that the altered behavior in response to Wnt5a or Wnt3a was statistically significant. Black bars indicate the percentage of explants showing migratory behaviour, grey bars indicate explants with non-migratory behaviour.

Figure 6

Wnt signalling components differentially required for generation of mesoderm from the late primitive streak. Timelapse imaging of embryos electroporated and grafted into unlabelled host embryos with GFP over 20 hours – intervals of 2.5 hours are shown. GFP control cells grafted into the anterior primitive streak(A) or into the posterior primitive streak (D). Cells expressing pCAβ-IRES-GFP-DshΔPDZ/DEP grafted into the anterior streak (B) or into the posterior streak (E). Cells expressing pCAβ-IRES-GFP-Pk1ΔPET/LIM grafted into the anterior streak (C) or into the posterior streak (F). Arrows show the position of the most recently formed somite. Arrowheads indicate the anterior-most GFP positive cells.

Figure 7

Electroporation of pk1ΔPET/LIM affects the migration of primitive streak cells in vivo and in explants. (A) Quantitative representation of in vivo grafting experiments shown in Fig. 6. The normal behaviour of anterior and posterior explants, expressing GFP is represented as light grey shading. Dsh1-ΔPDZ/DEP disrupts normal behaviour anteriorly but not posteriorly. pk1-ΔPET/LIM and pk1 constructs disrupt normal behaviour posteriorly but not anteriorly. (B) Streak explants containing migratory cells are represented with black shading. Exposure of anterior explants to Wnt5a can induce migratory behaviour but this is abolished in explants expressing pk1ΔPET/LIM. The majority of posterior explants expressing pk1-ΔPET/LIM are no longer migratory and this cannot be rescued by exposure to Wnt-5a expressing cells. * asterisk indicates statistically significant results.