QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online 3 July 2006
doi: 10.1242/dev.02460

This Article Summary Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Alexandre, P. Articles by Wassef, M. Search for Related Content PubMed PubMed Citation Articles by Alexandre, P. Articles by Wassef, M. Social Bookmarking                         What's this?

Positive and negative regulations by FGF8 contribute to midbrain roof plate developmental plasticity

Paula Alexandre^*, Isabelle Bachy, Morgane Marcou and Marion Wassef

Régionalisation Nerveuse CNRS/ENS UMR 8542, Département de Biologie, Ecole normale supérieure, 46 rue d'Ulm, 75005 Paris, France.

View larger version (13K): [in a new window]   Fig. 1. Background on midbrain roof plate properties: schematic representation of previous experiments that disclosed unexpected properties of the RP near the isthmic organizer (IsO). (A-C) For simplicity, although the midbrain RP is not yet differentiated from the neural folds at stage HH10, the experimental manipulation performed at stage HH10 and its outcome on RP patterning 1 or 2 days later are both represented on the same outline of a stage HH10 chick neural tube. (A) Cells from a restricted isthmic domain (red circle isthmic node, Isnode) extend to populate the midline on both sides of the MH junction, marked by an arrow (Millet et al., 1996; Louvi et al., 2003; Alexandre and Wassef, 2003). (B) Interfering with local patterning at the Isnode prevents RP formation on the adjacent midline (Alexandre and Wassef, 2003). (C) A lateral source of FGF8 created by insertion of a FGF8-soaked bead or of an IsO transplant induces RP duplication/bifurcation (Bally-Cuif and Wassef, 1994; Crossley et al., 1996). (D) The RP fails to develop on the dorsal midline of midbrain vesicles whose AP axis has been inverted by a 180° rotation (Marin and Puelles, 1994; Alexandre and Wassef, 2003).

View larger version (61K): [in a new window]   Fig. 2. Bipolar differentiation of the midbrain roof plate. (A-A'') BMP signaling is required for RP differentiation later than stage HH10. Lateral (A), posterior (A') and dorsal (A'') views of the same embryo fixed 2 days after insertion of a noggin-soaked bead in the midbrain and labeled for Gdf7 transcripts. The midbrain RP is mostly missing (A, arrowhead) except for a faint labeling at its caudal (A', arrowhead) and rostral (A'', arrowhead) ends. (B,C) Local and distant effects of FGF8 beads. Lateral views of embryos labeled for Wnt1 2 days after implantation of FGF8 beads. In most cases, the induced RP extends from the dorsal midline to the bead (B, arrowhead). In cases where the FGF8 bead induces locally a small RP segment (C, arrow), a small deflection of the endogenous RP also points in the direction of the bead (C, arrowhead).

View larger version (96K): [in a new window]   Fig. 3. Expression of RP competence factors in rotated dorsal midbrain transplants. Dorsal views of four HH22-23 quail-chick chimeras (A-A'',B,B',C,C',D-D'') that received a 90° rotated dorsal midbrain transplant 3 days earlier at stage HH10, as schematized on the left (black, quail donor). The embryos were treated for the detection of Wnt1, Lmx1b or Gdf7 transcripts, as indicated together with the color of detection (red or purple) at the top of each picture. Species-specific probes were used in A-A'',C-D'' to discriminate between chick and quail Wnt1 transcripts. After using the second probe, the quail transplant was identified by QCPN immunocytochemistry (A'',B',D''). RP-like structures developed in the graft in contact with one or both ends of the host RP in about 25% of the transplants (16/63, A-A''). In more than 70% of cases, Wnt1 remained widely expressed in the transplants (B-D'). Lmx1b was expressed ectopically in part of the Wnt1 expression domain (4/4, B'). Except in the smallest transplants, Wnt1 expression was regulated along the DV axis, leading to its reorganization in a wide band that flanked the host midline (C',D'). In these cases, the transplant straddles the midline (D', see also F, a transverse section through the transplant, double arrowhead) without differentiating a RP (compare with E, a section through the host RP, arrowhead).

View larger version (64K): [in a new window]   Fig. 4. LMX1B and WNT1 slowly induce GDF7 expression. Lmx1b and Wnt1 are competence factors for RP differentiation. At stage HH10-11, Lmx1b (A) and Wnt1 (B) are expressed both on the dorsal midline and widely in the caudal midbrain (between arrowheads). Lateral views (C,E,E') and transverse sections of the midbrain (D-D'',F) of embryos fixed 10 (D-D''), 24 (C) and 34 (E,E',F) hours after coelectroporation of Lmx1b and Egfp expression vectors. Lmx1b (C,E,F) or Wnt1 (D-D'') transcripts are detected in red, Gdf7 transcripts in purple (C,D'',E,F) and EGFP in green (D',E'). Twenty-four hours after electroporation, very few scattered cells express Gdf7 ectopically (C). Gdf7 expression is strongly induced 34 hours after electroporation (E,F). Except in the caudal midbrain (arrowhead in E) and at the midbrain-forebrain junction, the purple staining for GDF7 partly obscures that for Lmx1b (E,F). This fainter induction of GDF7 does not result from a difference in the efficiency of Lmx1b electroporation around the constrictions, as the co-electroporated EGFP reporter is expressed at high levels in the same regions (arrowhead in E').

View larger version (72K): [in a new window]   Fig. 5. FGF8 delays RP maturation. Posterior (A-E) and lateral (F-G) views of HH14-15 (A-C), HH16-17 (D-F') and HH18 (G) chick embryos treated by in situ hybridization for the detection of RP markers. The expression of noggin (A), Id3 (B) and Gdf7 (C) is transiently downregulated in the caudal midbrain at stage HH14-15 (arrowheads). A mature expression pattern of Gdf7 is observed in the caudal midbrain, beginning at stage HH16 (D). Insertion of a FGF8-soaked bead (E, red dot) in the caudal midbrain prolongs Gdf7 downregulation beyond HH17. (F,G) The same delay in Gdf7 expression is observed on the ectopic RP induced by FGF8 beads (red dots in F,G). Embryos fixed 24 (F,F') or 30 (G) hours after FGF8 bead insertion were treated for the detection of Gdf7 in purple and Wnt1 in red. (F,F') Same embryo shown in bright and dark field. At 24 hours, an ectopic preRP is already labeled for Wnt1 (F,F'), but Gdf7 expression lags behind (F). At 30 hours Gdf7 and Wnt1 are co-expressed in the induced RP (G). (H-J) The hypothetical consequences of an imbalance between the positive and negative influences of FGF8 on RP differentiation. The competence factors Lmx1b and Wnt1 (purple) are targets of FGF8 signaling; their expression is sufficient to induce slowly maturing RP markers (see Fig. 4). Conversely, high levels of FGF8 signaling (red) inhibit the expression of RP maturation markers. (H) A balance between these two influences of FGF8 maintains a progress zone of RP maturation (orange). (I) An imbalance in favor of competence factors should result in RP widening, whereas (J) increasing the inhibitory activity of FGF8 at the expense of competence results in the formation of gaps in the developing RP. We propose that the lack of a RP structure in the caudal part of rotated midbrain vesicles (Alexandre and Wassef, 2003; Marin and Puelles, 1994) is the consequence of this last configuration.

View larger version (69K): [in a new window]   Fig. 6. BMP family members and RP autoinduction. Whole embryos treated for the detection of Gdf7 transcripts (A,D,E,G,H) or WNT1 (purple) and TTR (black) (B,C) 3 (A), 5 (B,B',C,C') and 1 (D-I) day(s) after implantation of beads soaked in BMP7, noggin, GDF7 or a mixture of GDF7+FGF8 recombinant proteins (see color code). Dorsal (A-C) or posterior (D,E,G,H) views. (B',C') Transverse sections through B and C, respectively. The inset in B' shows a comparable section through a control embryo. (F,I) Longitudinal sections through the midbrain of embryos similar to D and G, respectively. The sections were immunostained for active caspase 3 (*caspase). (A,B,B') BMP7 beads did not induce RP markers locally. They affected RP patterning on the host midline, inducing the formation of a sheet of cells between the two widened Gdf7 (A) or Wnt1 (B,B') positive RP halves. This cell sheet expressed neither the roof-plate marker WNT1 nor the choroid plexus marker TTR (B,B'). (C,C') Noggin beads induced the formation of a dome-shaped tectum that lacked both WNT1 and TTR expression. (D,E; 13/16) GDF7 beads induced Gdf7 expression in a cell patch isolated from the midline (D) or through widening the midline Gdf7 domain (E). Cell death was not markedly increased by GDF7 overexpression: *caspase immunoreactive cells were barely detectable (F, a single immunoreactive cell at most per section in one out of five embryos). Beads releasing GDF7 and FGF8 together induced a widely scattered expression of Gdf7 (G,H) and impaired the condensation of Gdf7 expression into the characteristic structure of the FGF8-induced ectopic RP. Cell death was increased near the midline (I) in all cases (7/7, at least three sections containing more than four *caspase immunoreactive cells).

View larger version (57K): [in a new window]   Fig. 7. FGF8 regulates the activity of activin, a potent modulator of RP development. (A-C) Dynamic expression of activin B (arrowheads) in the midbrain (A,B) and caudal forebrain (B,C) of HH10, HH12 (A,B, dorsal views) and HH18 (C, lateral view) embryos. (D-H') Overexpression of activin through bead insertion (blue dots) destabilizes the midbrain RP. (D,E) Activin beads induce ectopic LMX1B expression 7 (D) and 23 (E) hours after bead insertion. At 48 hours (F-H'), a thin row of GDF7-expressing cells links the activin bead to the endogenous RP (arrowheads in F,H,H'; H' is a sagittal section through H). Activin also interferes with GDF7 expression in the endogenous caudal midbrain RP (arrows in G,H). (I-P) follistatin expression is modulated by FGF8 signaling. At stage HH11 (I,J), follistatin is expressed in a bilateral domain that straddles the midbrain-forebrain junction and flanks (arrowhead in I) the site of initiation of Gdf7 expression on the dorsal midline of the anterior midbrain (arrowhead in J). (K) Lateral view of a stage HH14 embryo illustrating the subsequent posterior extension of follistatin expression in the midbrain (posterior is leftwards, the arrowhead indicates the midbrain-forebrain junction). In contrast to noggin-soaked beads (Fig. 2A-A''), dorsal follistatin-soaked beads (L, arrowhead) did not impair RP differentiation. Compared with the contralateral control side (arrowheads in M), follistatin expression was down- or upregulated, respectively, near beads soaked in FGF8 (red dot in M) or the FGF8 signaling inhibitor SU5402 (black dot in N). (O,P) Stage HH13-14 heads were separated on the midline into two halves. To remove the endogenous source of FGF8, the isthmic region was removed from one half (as indicated by the broken line in O). The control (O) and ablated (P) halves were cultured side by side on floating membranes for 6 hours and treated together for the detection of follistatin transcripts. The level of expression of follistatin was higher on the ablated side (arrows in O and P).

View larger version (33K): [in a new window]   Fig. 8. Summary and interpretation. The left diagrams indicate the initial condition; the right diagrams indicate how it evolves with time. (A) In the caudal midbrain, FGF8 maintains expression of Lmx1b and Wnt1 in a wide domain that is therefore competent to differentiate into RP (Chizhikov and Millen, 2004b). However, FGF8 prevents RP maturation by inhibiting the transcription of maturation markers or by inducing cell apoptosis. Conversely, short-range homeogenetic mechanisms (arrows), mediated in part by GDF7, induce RP extension. When the system equilibrates, a RP progress zone forms at short distance from the FGF8 source. Our previous studies (Louvi et al., 2003; Alexandre and Wassef, 2003) indicate that growth, but also convergent extension, may increase the distance between the FGF8 source and the progress zone, leading to RP extension posteriorly. Interestingly, although elsewhere in the neural tube RP differentiation is confined to the competent neural folds, FGF8 releases RP progression from the neural fold environment by inducing a RP competent territory. (B) Lateral extension of the RP under the influence of a FGF8-soaked bead depends on similar mechanisms. (C) The same interactions could induce a short local RP segment near FGF8 beads. FGF8 signaling rapidly induces high levels of Lmx1b and Wnt1 before decreasing. Because Gdf7 expression is slowly induced by Lmx1b and Wnt1, a patch of Gdf7-expressing cells may be induced whose extension would follow the FGF8 decreasing gradient or respond to convergent extension. (D) We find that changing the polarity of dorsal midbrain fragments results in the maintenance of high levels of Wnt1 and induction of Lmx1b expression in the transplants. The transplant thus becomes competent for RP homeogenetic induction from the adjacent host RP. Long-range DV regulations probably also contribute to RP induction by reorganizing tissue polarity in the transplant and upregulating dorsally the expression of competence factors. The asterisks indicate the sites where homeogenetic RP induction was initiated.

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

Twitter