February 1, 2011;
PDGF-A controls mesoderm cell orientation and radial intercalation during Xenopus gastrulation.
Radial intercalation is a common, yet poorly understood, morphogenetic process in the developing embryo
. By analyzing cell rearrangement in the prechordal mesoderm
during Xenopus gastrulation, we have identified a mechanism for radial intercalation. It involves cell orientation in response to a long-range signal mediated by platelet-derived growth factor (PDGF-A
) and directional intercellular migration. When PDGF-A
signaling is inhibited, prechordal mesoderm
cells fail to orient towards the ectoderm
, the endogenous source of PDGF-A
, and no longer migrate towards it. Consequently, the prechordal mesoderm
fails to spread during gastrulation. Orientation and directional migration can be rescued specifically by the expression of a short splicing isoform of PDGF-A
, but not by a long matrix-binding isoform, consistent with a requirement for long-range signaling.
[+] show captions
Fig. 1. Dorsal mesoderm during gastrulation. (A,B) Diagrams of (A) stage 10 (early gastrulation) and (B) stage 12 (late gastrulation); ventral Xbra expression is in blue. (C) Mesoderm regions at mid and late gastrulation, traced from SEM pictures. Scale bar: 120 μm. (D) SEM micrograph of sagittally fractured stage 11 embryo. Purple, chordamesoderm; yellow, PCM; orange, LEM. The percentage of the total mesoderm length occupied by each region is indicated. White box, transition between PCM and LEM. (E-G) High-magnification SEM images of (E) LEM, (F) PCM and (G) chordamesoderm from sagittally fractured stage 11 embryo; broken lines indicate boundaries of mesoderm regions. Cells extend protrusions in (white arrowheads) and out (white arrows) of the plane of the image. Red arrowhead in E indicates a cell with oblique orientation near the leading edge. Curved arrow in G indicates the direction of mesoderm internalization. Red arrowhead in G indicates a cell oriented perpendicular to the BCR at the anterior border of the region. (E′-G′) SEM images of (E′) LEM, (F′) PCM and (G′) chordamesoderm at stage 11, showing the surface previously in contact with the BCR. (E′) Red arrowheads indicate cells extending protrusions animally in a shingle arrangement. (F′) Red arrowheads indicate non-oriented cellular protrusions. (G′) Red arrowheads indicate bipolar cells in mediolateral orientation. (E′-G′) In situ hybridization for (E′) Cer, (F′) Gsc and (G′) Xbra; broken lines and arrowheads show region boundaries. (H) The angle (θ) between protrusion and the BCR as a measure of cell orientation. (I-K) Rose diagrams representing the percentage of cells extending protrusions (I,J), or the long axis (K) oriented with respect to BCR. 0° is towards the BCR, 90° toward the animal pole; n, number of cells; BCR, blastocoel roof; BPL, blastopore lip; BC, blastocoel; PCM, prechordal mesoderm; EN, endoderm; AC, archenteron; LEM, leading edge mesendoderm; Chorda, chordamesoderm.
Fig. 2. PDGF-A inhibition disrupts PCM radial intercalation. (A) The PCM thins significantly (P<0.0001, n=10 embryos/stage) between stages 11 and 12; thinning is reduced upon AG1296 treatment (P<0.0001, n=10 embryos). (B) The change in layer index of PCM over time and following AG1296 treatment at stage 11. (C) The PCM layer index is significantly higher at stage 12 following AG1296 treatment (P=0.0013, n=13 embryos) or at stage 11 in PDGF-A MO-injected embryos (P<0.0001, n=12 embryos); chordamesoderm is unaffected (P=0.0886, n=13 embryos). (A-C) Error bars represent s.d., asterisks indicate significant results. (D,E) Orientation of chordamesoderm cells: BCR, 0°; n=number of cells. No significant difference between (D) controls and (E) AG1296-treated embryos (P=0.9842). (F-H′) Sagittally fractured stage 12 embryos treated with DMSO (F-H) or 10 μM AG1296 (F′-H′) 2 hours prior to fixation; broken lines indicate region boundaries. (F) Red arrowheads indicate PCM cells in a single layer in DMSO controls. (F′) Red arrowheads indicate superficial PCM cells in contact with the BCR; red arrows indicate a second layer of cells in AG1296-treated embryos. (G,G′) Red arrowheads indicate LEM cells oriented perpendicular to the BCR in controls (G) and obliquely oriented LEM cells in AG1296-treated embryos (G′). (H,H′) Red arrowheads indicate chordamesoderm cells extending protrusions towardsthe BCR; red arrows indicate cells extending protrusions toward the archenteron epithelium. BCR, blastocoel roof; EN, endoderm; AC, archenteron.
Fig. 4. sf-PDGF-A is required for radial orientation of PCM cells. (A) A sagittally fractured stage 11 embryo injected with 60 ng PDGF-A morpholino; cells are oriented parallel to the BCR (red arrowheads). The broken yellow lines indicate region boundaries. (B-D′,F-H′) Orientation of PCM cells (B-D,F-H) and LEM cells (B′-D′,F′-H′) from embryos injected with (B,B′) 60 ng PDGF-A morpholino, (C,C′) 1.6 ng of dominant-negative PDGF-A 1308 RNA, (D,D′) 400 pg of dominant-negative PDGFR-37 RNA, (F,F′) 60 ng PDGF-A MO + 800 pg sf-PDGF-A RNA, (G,G′) 60 ng PDGF-A MO + 800 pg lf-PDGF-A RNA or (H,H′) 800 pg sf-PDGF-A RNA. (E) Sagittally fractured stage 11 embryo co-injected with PDGF-A MO and sf-PDGF-A mRNA. Cells extend protrusions towards the BCR in (red arrowheads) and out (red arrows) of the plane of the image; the broken yellow line indicates the region boundary. (I,J) In situ hybridization for (I) Xbra and (J) Gsc from embryos injected with 60 ng PDGF-A MO. (K,K′) Orientation of cells from embryos injected with 400 pg of sf-PDGF-A mRNA in the mesoderm. n, number of cells per region; 0° and 90° correspond to BCR and animal pole, respectively. EN, endoderm; PCM, prechordal mesoderm; LEM, leading edge mesendoderm.
Fig. S1. Mapping of normal marker gene expression domains onto scanning electron micrographs. Average length and width dimensions of Xbra (chordamesoderm) and Gsc (PCM) expression domains were measured from in situ hybridization images. The lengths of the expression domains varied (see standard deviations); thus, their average lengths were determined, and the percentage of total mesoderm length occupied by each domain was used to map domains onto the SEM pictures. Standard deviations were of the order of one to two cell diameters. The widths of domains varied little, and average values were used directly. The domains thus defined corresponded well to anatomical boundaries between ectoderm and mesoderm (Brachet�s cleft) on one side, and large endoderm cells or the archenteron epithelium on the other. No unambiguous markers were available to distinguish between LEM and deep endoderm, and no distinct boundary was visible. For practical purposes, we extended the PCM-endoderm boundary anteriorly in parallel to the BCR, probably underestimating the width of the LEM. The table shows average length, width and average proportion values with standard deviations for the different regions of the normal mesoderm. This method was used for all experiments, in connection with respective in situ hybridization data. Purple, chordamesoderm (xBra); yellow, pre-chordal mesoderm (Gsc); orange, leading edge mesoderm (non-Gsc, Cer expressing).