January 1, 2008;
Coordination of cell polarity during Xenopus gastrulation.
Cell polarity is an essential feature of animal cells contributing to morphogenesis. During Xenopus gastrulation, it is known that chordamesoderm
cells are polarized and intercalate each other allowing anterior
elongation of the embryo
proper by convergent extension (CE). Although it is well known that the cellular protrusions at both ends of polarized cells exert tractive force for intercalation and that PCP pathway is known to be essential for the cell polarity, little is known about what triggers the cell polarization and what the polarization causes to control intracellular events enabling the intercalation that leads to the CE. In our research, we used EB3
(end-binding 3), a member of +TIPs that bind to the plus end of microtubule
(MT), to visualize the intracellular polarity of chordamesoderm
cells during CE to investigate the trigger of the establishment of cell polarity. We found that EB3
movement is polarized in chordamesoderm
cells and that the notochord
boundary plays an essential role in generating the cell polarity. This polarity was generated before the change of cell morphology and the polarized movement of EB3
cells was also observed near the boundary between the chordamesoderm tissue
and naïve ectoderm tissue
or lateral mesoderm
tissues induced by a low concentration of nodal
mRNA. These suggest that definitive tissue
separation established by the distinct levels of nodal
signaling is essential for the chordamesodermal cells to acquire mediolateral cell polarity.
[+] show captions
Figure 1. Tracking of EB3-GFP movement shows microtubule growth in Xenopus explants.
(a) In an animal cap cell, EB3-GFP molecules show radially symmetrical movement toward the rim of the cell in the rose diagram (a'). (b) In a chordamesoderm cell, the movement was mostly bidirectional toward both ends of the cell (b'). (c, d) Magnification of the membrane region (white box) of a and b shows that EB3 moved toward the cytoplasmic membrane and disappeared after attaching to it in an animal cap cell (c), while EB3 moved along the membrane in a dorsal mesoderm cell. Arrows show the direction of the EB3 movement, and dotted lines indicate the cytoplasmic membrane. (d). (e) Z section of chordamesoderm cell reveals that most of the EB3 comets are observed in the plane, which is 1.2 from glass dish. In the planes, which are 4.2 and 7.2 from glass dish, the yolk granule were located in the center region of the cell and EB3 comets were located only near the cytoplasmic membrane (white arrow heads). Scale bars, 20 . Each scale of the rose diagrams corresponds to 10%.
Figure 2. Notochord-somite boundary and/or extra-notochord tissue attracts EB3 movement.
(a) A relatively late notochord cell (Stage 13~) that was proximal to the boundary. The cell was divided into three regions (left: L, center: C, right: R) of equal horizontal (mediolateral) length for tracking the EB3 comets to clarify the direction of their movement. The EB3 movements were highly biased and mostly rightward in the center portion (a'). (b) A late notochord cell (Stage 13~) that was distant from the boundary. No significant bias of the EB3 movement was observed (b'). (c) APC-GFP localization in the notochord in Keller explant. (d) APC-GFP localization in the chordamesoderm cells near the boundary in Keller explants. The localization was restricted to the one cell end pointing toward the boundary. (e) APC-GFP localization in the chordamesoderm cells located in the center of the notochord. Note that the localization was diffuse and unfocused. The dotted line indicates the notochord-somite boundary, and arrowheads indicate the localization of APC-GFP. Scale bars, 20 . Each scale of the rose diagrams corresponds to 10%.
Figure 3. EB3 shows polarized movement independent of cell morphology in the early phases.
(a) Time course of cell-shape change from 0' to 180' in Keller explants. The dashed lines outline the shape of the cells. (b) Rose diagrams of EB3 movement in the indicated cells. Each cell was divided into left and right portions at the center of the horizontal (mediolateral) cell length, and the EB3 movements in each portion were plotted as rose diagrams. (L) Left and (R) right. (c) Correlation between cell shape (length/width) and bias of EB3 movement (rightward EB3; −45to +45. Note the biased rightward movement without significant cell-shape change. Scale bars, 20 . Each scale of the rose diagrams corresponds to 10%.
Figure 4. A heterogeneous combination of tissues can confer planar cell polarity.
(a) The single culture of animal cap expressing a high concentration of Nodal mRNA (12550 pg: NAC). (b) Combined culture of uninjected animal cap (AC) and NAC. Homogenous combined cultures of NAC (c) and AC (d). The EB3 in the cells in panels a respectively, was tracked. (a'') Rose diagrams of the EB3 movements in cells a. (e) RT-PCR (reverse transcription-polymerase chain reaction) analysis of low concentration of Nodal-mRNA (255 pg) injected animal caps. Animal caps into which a relatively high concentration of Nodal mRNA (75 pg: SN-AC) was injected expressed the chordamesoderm and somite marker, Xnot and XmyoDa, papc, and those receiving a relatively low concentration of Nodal mRNA (37.50 pg: S-AC) expressed only the somite marker, XmyoD. (f) RT-PCR analysis of high concentration of Nodal-mRNA (150 pg00 pg: NAC, H-NAC). (g) The angle of the cell body in relation to the border in NAC (150 pg) combined with explants treated with different Nodal levels (05 pg of mRNA). Cell alignment perpendicular to the boundary, representing the mediolateral polarity, was observed when the combination was heterogeneous and loss when the combination became more homogeneous (37.55 pg). (per: perpendicular, para: parallel, ran: random) (h) The ratio of perpendicularly aligned cells in relation to the border was altered by difference of nodal signaling level. The cells in H-NAC (250 pg) tended to align perpendicular to the border with S-AC (37.50 pg). Scale bars, 20 mm (a). Each scale of the rose diagrams corresponds to 10%.
Figure S3.(a) Fibrillin antibody staining of a Keller explant at Stage 12. Fibrillin was localized to the cytoplasm (puncta). (b) Stage 14. The fibrous structure of fibrillin was formed. (c) Stage 16. The dotted line shows the notochord-somite boundary. An enrichment of fibrillin accumulation was observed around the notochord. (a'') Fibrillin antibody staining of nocodazole treated Keller explants at Stage 12~16. The fibrillin accumulation around the notochord was inhibited by nocodazole, an inhibitor of MT polymerization.
Figure S2. Smad 2 is an intracellular component of activin/nodal signaling, thus smad2-expressing AC induced chordamesoderm tissue cell autonomously. RT-PCR analysis showed that AC receiving over 250 pg smad2 mRNA expressed the chordamesoderm marker. When the AC was co-cultured with Smad2-AC, EB3 in the AC cells moved randomly, unlike that in the AC cells co-cultured with Activin expressing AC, which moved toward the boundary in a mirror image. These results suggest that the boundary between two distinct tissues has information to attract the EB3 movement; moreover, the attracting cue is specific for chordamesoderm tissue, because the EB3 in the AC moved randomly near the tissue boundary.
Figure S4. To investigate the relevance of the PCP pathway in the biased EB3 movement, a dominant-negative form of Dishevelled (D2) was co-injected with activin mRNA. The EB3 in D2-AAC moved toward the boundary with AC. This result shows that the biased EB3 movement is independent of the PCP pathway.