|
Fig. 1. deltaXF/TdXF expression perturbs gastrulation. (A) Schematic outline of the domain structure of the prototypical fibrillin and the injection constructs: deletion XF (deltaXF), twice-deleted XF(TdXF) (similar to deltaXF but with a partial EGF repeat removed) and two carboxy-terminal deletions of TdXF, TdXF-Protease (TdXF-P) and TdXF-Repeats (TdXF-R). Shown are calcium binding EGF-like repeats (gray ovals), EGF-like repeats (black ovals), 8-cysteine repeats (black rectangles), hybrid repeats between EGF-like and 8-cysteine repeats (gray rectangles) unique sequences (lines and gray box), N-cad signal sequences (black oval and pentagon—representing two distinct fusion sites between signal sequence and XF sequence) and furin processing site (arrow). (B) Percent of embryos exhibiting open blastopores is plotted with SEM for presumptive dorsal vs. ventral mesoderm injection of deltaXF (D, n = 56; V, n = 49), TdXF (D, n = 77; V, n = 80), TdXF-P (D, n = 71; V, n = 71), TdXF-R (D, n = 75; V, n = 59), pre-prolactin (PPL) (D, n = 47; V, n = 44) or uninjected embryos (n = 195). (C) DeltaXF dorsally injected late gastrula embryos exhibiting failure of blastopore closure and partial exogastrulation. Scale bars are 0.5 mm in panels C–H. (D) The same embryos as in panel C, but visualizing co-injected GFP to reveal the dorsal targeting of these injections. (E) Dorsal deltaXF-injected St 22 embryo exhibiting open blastopore. (F) deltaXF-injected embryo stained for a notochord marker (Tor70), which reveals notochord (N) ringing the yolk plug. (G and H) Frames from movies of developing embryos injected with (G) TdXF or (H) PPL with elapsed time indicated. YP = yolk plug; n = notochord; NF = neural folds.
|
|
Fig. 2. Morphology and XF fibrils in deltaXF-expressing explants. (A–D) Keller sandwiches made from embryos injected with increasing doses of deltaXF, allowed to extend and stained for notochord (N). Scale bar in panel B is 0.2 mm. (A) Control sandwich from uninjected embryos (representative example shown, n = 5), (B–D) 0.5 ng, 1 ng and 2 ng deltaXF RNA per dorsal blastomere, respectively (n = 3, 3 and 5, respectively). (E–H) Explants stained for native XF immunoreactivity (E and G), and both native fibrils (red) and GFP (green), indicating the region of delXF expression, with yellow being the combined signal (F and H). Arrowhead in panel G indicates bifurcation of explant, and bracket indicates region shown in panel H. Scale bars are 0.1 mm in panels E–J. (I) XF fibrils around notochord of an uninjected open face explant and (J) XF fibrils in notochord region of explant globally expressing deltaXF.
|
|
Fig. 3. Effect of deltaXF expression and CB1 peptide on cell behavior. (A and B) Genetic mosaic transplantation analysis of deltaXF-expressing cells reveals that deltaXF expression perturbs notochordal cell behaviors. (A) Embryos injected with either GFP and deltaXF or GFP and pre-prolactin were used to make DIMZ explants, and two small groups of expressing cells, one from each type of injected embryo, were transplanted above the dorsal lip of a single host embryo at midgastrulation. After healing, the DIMZ of this host embryo was explanted into culture and the labeled control and experimentally injected cells were observed during extension. (B) A single frame from a movie showing morphology of GFP labeled, transplanted control cells (injected with pre-prolactin + GFP; right of arrows) or experimentally injected (deltaXF + GFP; left of arrows) cells in an extending explant. Cells that exhibit an elongate morphology fit the profile of an MIB-expressing cell, and most control cells have length-to-width ratios greater than 2, in contrast to deltaXF-expressing cells in the same explant. In addition, note that the experimentally injected graft has maintained its cohesion during extension of the host explant, in contrast to the control injected graft which has been broken up through the process of intercalation to generate intermixed labeled and unlabeled cells. This serves as a positive control for this experiment, indicating that the grafted cells can participate in morphogenesis. (C) CB-1 peptide does not affect notochordal cell behavior. Time-lapse frames visualizing notochordal cell behavior in an axial explant undergoing morphogenesis to which CB-1 peptide is added at 50 μg/ml at frame t = 0. In contrast to deltaXF expression above, notochordal cell behavior is not affected by acute application of the CB-1 peptide.
|
|
Fig. 4. Effect of CB-1 peptide on development. (A) CB-1-injected embryos exhibit marked axis shortening and open blastopores. (B) Sibling embryos to those in panel A injected with scrambled peptide develop normally. Scale bars in panels A and B are 1 mm. (C–F) Embryos injected with 300 ng CB-1 peptide exhibiting graded levels of gastrulation arrest that correlates with fibrillin antibody staining. Scale bars for panels C–F is 0.5 mm. (C) An injected embryo exhibiting strong gastrulation arrest exhibits little fibrillin staining and a large open blastopore and will develop into an embryo such as shown in panel A. (D) An embryo that is progressively better at closing its blastopore exhibits progressively more fibrillin staining. The majority of injected embryos appear like those in panel C or D. (E) Embryos that do close their blastopore can exhibit fibrillin staining defects, such as the somitic staining on only one side seen in dorsal view here (side lacking fibrillin staining indicated by arrow). (F) A few injected embryos exhibit normal fibrillin staining but still exhibit a slightly abnormal morphology. (G) Time-lapse sequence showing failure of blastopore closure in an injected embryo; compare to Fig. 1H for normal closure. Scale bar is 0.5 mm. AF = axial fibrillin staining; YP = yolk plug.
|
|
Fig. 5. Functionally defining the CB-1 domain responsible of perturbing gastrulation. (A) Percent of gastrulation arrested embryos is plotted with SEM from at least three experiments for uninjected embryos (n = 273), or injected with CB-1 peptide (n = 205), scrambled peptide (n = 272), deletion CB1 peptide (n = 84), CB1M1S (n = 68), CB1M2S (n = 76), CB1M4S (n = 58), CB1M6S (n = 74). CB1M6W (n = 72), CB1 + 10 (n = 139), CB1M10 (n = 50). (B) Table of sequences of injected peptides. Mutations from CB-1 sequences are shown in bold. (C) Injection of an mRNA encoding a deletion mutant of the TXF-P construct that lacks a functional CB-1 domain does not perturb gastrulation. Domain structure key is as in Fig. 1. Injection of TXF-P on the dorsal (n = 40) but not ventral (n = 43) side of the embryo causes gastrulation arrest. In contrast, expressing an eight amino acid deletion variant, TXF-delCB1, does not cause gastrulation arrest when injected dorsally (n = 42) or ventrally (n = 34). (D) The gastrulation arrest phenotype of the CB1 peptide is heparin sensitive. Failure of blastopore closure is plotted for uninjected embryos (n = 75), embryos injected with 30 nl of 200 μg/ml heparin (n = 66), injected with 30 nl of 10 mg/ml CB-1 (300 ng) peptide (n = 80) or injected with CB-1 peptide (300 ng) preincubated with 200 μg/ml heparin (n = 81).
|
|
Fig. 6. XF morpholino perturbs gastrulation. Reducing XF translation in embryos blocks the normal progression of gastrulation. (A) Percent of gastrulation arrested embryos is plotted with SEM for XF morpholino (n = 93), control MO (n = 88) or uninjected embryos (n = 94). (B) Western blot for XF reveals XF morpholino injection results in a partial knockdown of XF protein levels, when normalized against a loading control (LC). (C) Embryos injected with 10 μM 5-misprime XF control morpholino develop normally. (D and E) Embryos injected with 10 μM XF MO exhibit open blastopores. (F) Embryos unilaterally injected with 10 μM 5-misprime control morpholino do not bend. (G) Embryos injected with 10 μM XF morpholino bend towards the injected side. H = head; YP = yolk plug.
|
|
Fig. 7. Model for function of Xenopus fibrillin in convergence and extension, and consequences for morphogenesis of perturbing XF globally and unilaterally. All panels show an open face DIMZ explant before and after autonomous extension. For each, extension is downward (posterior) with yellow representing notochord and red lines indicating the location of the presumptive notochordal–somitic boundary (upper panels), or the localization of XF in the mature, matrix-rich notochordal–somitic boundary (lower panels). (A) In normal open face DIMZ explants cells (open circles) are induced to become bilaterally protrusive and elongate (exhibit MIB), bind to the XF containing notochord–somite boundary, where they exhibit local inhibition of protrusion. These cells then converge towards the midline, dragging the boundary with them and causing directed extension of the notochord (shown in the extended explant). In this model XF at the boundary functions in the process of directing extension posteriorly and may function in regulating cell motility at the boundary. (B) In explants in which XF deposition at the boundary is globally perturbed, either by expressing deltaXF widely or injecting the CB-1 peptide, fibrillin fibril localization to both of the bilateral boundaries is disrupted. This leads to twisted notochords as extension is no longer globally directed in the posterior direction, and in whole embryo experiments leads to failure of blastopore closure. (C) In explants in which XF is compromised unilaterally bending occurs, again suggesting that XF fibrils normally function to direct extension posteriorly.
|
