a The Clock and Wavefront model: antagonistic gradient of Fgf8 (originating from the posterior PSM, green) and RA (originating from the somites, violet) define a wavefront which interacts with a particular phase of the segmentation clock (in the PSM, red) to generate somites at periodic times and positions. b Kymograph of somitogenesis from 7 to 20 somites. The tail elongates at a constant rate V_tail while the PSM shrinks at a roughly constant rate V_PSM resulting in a somite wavefront propagating⁵ (in the lab frame) at a rate V_front = V_tail − V_PSM.

a Time lapse observations of the Erk activity domain in DREKA embryos following exposure to a low (50 μM) concentration of SU5402 (an inhibitor of the Fgf pathway). The time span between images is 4 min. (the period of the clock in this experiment is about 45 min.) The yellow lines point to the boundary of Erk activity which periodically jumps by about 50 μm except in the first period after administration of SU5402 where the jump is larger (labeled as SU). Scale bar: 50 μm. b–e Normal somitogenetic development in DREKA embryos (control). b Erk activity in DREKA embryos at 14 somite stage: dark nuclei point to high Erk activity. c The false color images represent the Erk activity averaged over a 30 μm perpendicular stack of images. Notice the low activity area in the pre-pattern region (shown by the bracket) at the anterior PSM. d In situ hybridization (ISH) against xirp2a (a marker of somite boundaries) in 30 hpf embryos. e Immuno-Histochemical (IHC) staining against Mesp2a, a determinant of the last somite boundary (arrow). f–i Somitogenetic development in DREKA embryos in presence of a high (200 μM) concentration of SU5402 from the 10 somite stage (arrow in (h)). Somitogenesis is impaired past 13 somite stage (indicated by a * in h) with unclear somite boundaries. f, g The Erk activity is repressed (low) throughout the PSM. h ISH against xirp2a in 30 hpf embryos. i The determinant of the last somite boundary (Mesp2a), shown by an arrow is absent past 13 somite stage in i but present in e. Scale bar: 50 μm.

a–d Control: normal development. a Erk activity in DREKA embryos at 14 somite stage: dark nuclei point to high Erk activity. b The false color images represent the Erk activity averaged over a 30 μm perpendicular stack of images. c ISH against xirp2a (a marker of somite boundaries) in 30 hpf embryos. d IHC against Mesp2a, a marker of the last somite boundary at 14 somites (arrow). e–h Somitogenetic development in DREKA embryos incubated from one-cell stage in 10 μM DEAB in presence of 1 μM RA from the ten somite stage (arrow in g). e, f The Erk activity is high throughout the PSM. g ISH against xirp2a in 30hpf embryos. Somitogenesis is impaired past 13 somites (indicated by a *) with unclear somite boundaries. h IHC against Mesp2a show that it is not expressed past 13 somite stage (arrows in d and h). Scale bar: 50 μm.

a The transgene uas:fgf8-T2A-cfp (under control of Gal4-ERT) and its marker Eos. b The two characteristic phenotypes observed at 24 hpf upon activation of the transgene. c Erk domain of activity visualized by IHC against phosphorylated Erk (pErk) at ten somites in non-activated embryos (left) or in embryos in which the transgene was activated from bud stage (right): the uniformly expressed exogenous Fgf8 enlarges the domain of activity of Erk to the whole embryo, but doesn’t alter the Erk activity gradient, Fig. S4. d Time variation of total fgf8 mRNA (RTqPCR data) in presence of an exogenous source of Fgf8. At time t = 0 (bud stage) the concentration of fgf8 is contributed by the endogenous one (continuous line: best linear fit y = 0.11 x + 1 with r² = 0.98). From then on, the endogenous concentration decreases (see Fig. 5) while the exogenous concentration increases: it typically doubles the initial endogenous concentration in 9 h (about 20 somite stage). Inset: gel displaying the increase in exogenous fgf8 mRNA at various times post activation versus a reference gene (β-actin). e Rates of PSM shrinkage (V_PSM), tail growth (V_tail), and wavefront velocity (V_front) in embryos (n = 17) in which an exogenous source of Fgf8 was turned on. The various rates are only mildly affected by the increase in Fgf8 due to the exogenous source.

Negative ln(<[fgf8m]>) (=ln(2) δCt values from the RTqPCR data) at different somitic stages (blue ○: whole embryo; red ◊ : PSM only). Red line: linear best fit (y = ax + b) with a = 0.094 ± 0.013 (Χ²= 22; DF = 36), corresponding to a an exponential decrease of <[fgf8m]> with a time constant τ = 1/a ≅ 11 somitic periods, Eq. (1). Green ◊: -ΔPSM/λ with λ = 260 μm and the PSM shrinkage, ΔPSM, taken from Fig. 4e. From the value of λ and τ we deduce a mean velocity: V_PSM = λ/τ ≅ 24 μm/somite.

Phosphorylated Erk (pErk) distribution in WT embryos grown at different temperatures (20, 24, 28, and 32 °C) at 4 different stages of somitogenesis (7, 11, 15, or 19 somites stages) visualized by fluorescence following Immunohistochemistry staining (IHC). Even though embryos at 20 °C grow about three time slower than embryos at 32 °C, at similar stages embryos grown at different temperatures are similar and display similar pErk domain.

The δCt of genes (her1, her7, hes6, deltaC, deltaD, notch1a, myog, mespa, and mespb) known to play a role in somitogenesis relative to reference genes (rpl13a or β-actin) are plotted as a function of somite stage. Notice that as a function of somite stage the data taken at different temperatures collapse on the same curve. For some genes for which δCt appears to vary linearly with somite stage (exponentially decreasing concentrations) we display the best fit. For others the continuous curve is just a guide to the eye. The increasing concentration of myog (decreasing δCt values) is consistent with its role as a differentiation factor during myogenesis which follows somite formation.