DEV000216 Supplementary Material

Files in this Data Supplement:

• Supplemental Figure S1 -

  Fig. S1. Comparison of the expression patterns of caudal Wnts and their relationship to FGF and Retinoid pathways. Fgf8 (A,A′) transcripts overlap with Wnt3a (B,B′) and Wnt5a (C,C′) expression in the caudal neural plate/stem zone, primitive streak and caudal presomitic mesoderm, while, by contrast, Wnt8c (D,D′) expression also extends more rostrally into the preneural tube/transition zone (Hollyday et al., 1995; Hume and Dodd, 1993). Lef1 (E,E′) is detected in the primitive streak as well as the preneural tube and flanking rostral presomitic mesoderm (Schmidt et al., 2004). Importantly, Wnt8c transcripts disappear as the neural tube forms at the level of the most recently formed somites which coincides with rising levels of Retinoid signalling as shown by the onset of the retinoid synthesizing enzyme Raldh2 (F,F′) in the rostral presomitic mesoderm and expression of the retinoic acid receptor RARβ (G,G′) in the forming neural tube and (after a longer period of in situ development) in the neighbouring mesoderm (Cui et al., 2003). Finally, Wnt8c does not overlap with NeuroM (H,H′), a marker of newly differentiating neurons, which is only observed later in the neural tube flanked by somites (Diez del Corral et al., 2002; Roztocil et al., 1997). Based on these observations, canonical Wnt signalling downstream of Wnt8c is a good candidate to act as a relay of FGF signalling to control differentiation onset in the extending neural axis. Transverse sections (white lines in A-H). Arrowheads indicate the last somitic boundary formed. Scale bars: 100 μm in A for A-H; and 50 μm in A′ (for A′-H′).

• Supplemental Figure S2 -

  Fig. S2. Fgf8 is lost more rapidly from caudal presomitic mesoderm than from the caudal neural plate and Retinoic acid abolishes Fgf8 in both tissues. In some embryos, it is apparent that Fgf8 persists longer in the preneural tube than in the presomitic mesoderm and this difference in timing of FGF decline in the two tissues may help to determine the spatial and temporal onset of Raldh2 expression as this depends on; (a) loss of FGF signalling in the mesoderm, and (b) transduction of Wnt signals that persist in the neighbouring neuroepithelium, but which depend on FGF signalling. To investigate this apparent difference, sets of four explants were derived from each embryo, two caudal neural plate (CNP) and two caudal presomitic mesoderm (CPM) then cultured and processed for Fgf8 expression in parallel (A). In each set, one CNP and one CPM were cultured in control media while their contralateral pairs were cultured in the presence of 9-cis RA or the RA agonist TTNPB. In the table summarising the results of explant sets cultured 8 hours (green outline) or 24 hours (purple outline), fractions represent the number of explants with a positive signal out of the total assessed, and the percentage indicates the average extent of Fgf8 expression (B). After 8 hours culture, Fgf8 transcripts are robustly detected in caudal neural plate explants (C), while paired caudal presomitic mesoderm explants weakly express this gene (D). By 24 hours, although caudal neural plate explants still clearly express Fgf8 (E), caudal presomitic mesoderm explant pairs have completely lost their transcripts (F). Fgf8 transcripts are attenuated by RA signalling in caudal neural plate explant pairs (G) and caudal presomitic mesoderm (H) after 8 hours. Strikingly by 24 hours, while Fgf8 transcripts are already lost in the caudal presomitic mesoderm, they are still present in the caudal neural plate explants, but are completely lost on exposure to retinoids (I,J). RA can therefore attenuate Fgf8 levels efficiently in both neural and mesodermal tissue. Percentages in C-J correspond to the extent of the Fgf8 in situ signal scored for each explant in these two particular explants sets (C,D,G,H, green outline) and (E,F,I,J, purple outline). Scale bar: 50 μm in C for C-J.