Remodeling of the metabolome during early frog development

Abstract

A rapid series of synchronous cell divisions initiates embryogenesis in many animal species, including the frog Xenopus laevis. After many of these cleavage cycles, the nuclear to cytoplasmic ratio increases sufficiently to somehow cause cell cycles to elongate and become asynchronous at the mid-blastula transition (MBT). We have discovered that an unanticipated remodeling of core metabolic pathways occurs during the cleavage cycles and the MBT in X. laevis, as evidenced by widespread changes in metabolite abundance. While many of the changes in metabolite abundance were consistently observed, it was also evident that different female frogs laid eggs with different levels of at least some metabolites. Metabolite tracing with heavy isotopes demonstrated that alanine is consumed to generate energy for the early embryo. dATP pools were found to decline during the MBT and we have confirmed that maternal pools of dNTPs are functionally exhausted at the onset of the MBT. Our results support an alternative hypothesis that the cell cycle lengthening at the MBT is triggered not by a limiting maternal protein, as is usually proposed, but by a decline in dNTP pools brought about by the exponentially increasing demands of DNA synthesis.

Figure 1. Remodeling of the metabolome during early X.laevis development.

(A) Dendrogram and heat map showing the wide-spread remodeling of the metabolome in early development. On three different days, three different male/female pairs were bred in vitro to obtain clutches of unfertilized eggs (time = 0) and early embryos at various times post-fertilization (pf) at18°C. Single eggs or embryos were rapidly quenched, extracted, and subjected to metabolomic analysis. Metabolite levels were expressed relative to the average value for that metabolite in the egg samples. The ratios were calculated separately for the three different clutches. The metabolites (rows) were then clustered hierarchically and ratios plotted on a color scale. For each time point n = 3–4. (B) The timing of key events in early X.laevis development at 18°C. Groups of 5 embryos from the same mating were harvested at the indicated time points post-fertilization (pf) and extracts prepared for western blotting. The Nieuwkoop-Faber (N-F) stage of embryos was recorded . Y15 on Cdc2 phosphorylation rose 8 hours pf indicating that cell cycles had begun to slow and embryos were entering the MBT. Bottle cells appeared on embryos 13 hours pf indicating the start of gastrulation. (C–E) Plots of individual metabolite abundance relative to the value measured in eggs during early development. For each single egg or embryo, each metabolite abundance was divided by the average egg value for that metabolite. This ratio was then log₂ transformed (n = 3-4± 1SD). These data are the same data presented in (A). The three clutches were plotted separately (black: clutch 1, green: clutch 2, blue: clutch 3). (F) Illustration of the embryos harvested at the indicated hours post fertilization. The corresponding Nieuwkoop-Faber stages are also indicated.

Figure 2. Metabolic fluxes that were traced in early embryos.

Metabolites whose abundance increases in early development are shown in red, those whose abundance decreases are in green, those with no observable change are in black, and metabolites that were not measured are in grey. Enzymes are in italics. AT: aminotransferase, PDH: pyruvate dehydrogenase, PC: pyruvate carboxylase, CS: citrate synthase, MDH: malate dehydrogenase.

Figure 3. Alanine consumption generates energy in early embryos.

(A) ¹⁵N-alanine injected into early embryos is quickly consumed with the heavy nitrogen being transferred to glutamate and aspartate. Glutamine, valine and proline accumulate ¹⁵N later in the time course. 1 nmol of ¹⁵N-alanine was injected into 2-cell embryos 2 hours post-fertilization (pf). For the first time point, embryos were quenched as quickly as possible following the injection (∼3 min). The proportion of each of the measurable isotopic forms, relative to the total abundance of that metabolite, was calculated for each embryo at each time point. The isotopic proportions in the 3–4 individual embryos measured at each time point were averaged and plotted. Error bars are ± 1SD. The abundance of each isotopic form was corrected to account for the natural abundance of ¹³C. (B) The heavy carbons from U-¹³C-alanine injected into early embryos are transferred to TCA cycle intermediates. Injections and analysis are as for (A). (C) U-¹³C-aspartate injected into early embryos almost immediately equilibrates with the malate pool and then moves into the rest of the TCA cycle. Injections and analysis are as for (A), except that 0.5 nmol of U-¹³C-aspartate was injected. Unlike (A) and (B), the first time point (t = 0) is for uninjected 2-cell embryos, and the second time point (t = 3 min) is for embryos that were quenched as quickly as possible following injection.

Figure 4. Evidence suggesting that dATP limitation triggers the MBT.

(A,B) Plots of dATP, dTTP, and dCTP abundance during early development. dGTP was not measurable in part due to interference from ATP, which is much more abundant. The MBT occurs 8 hours post-fertilization (pf), see Figure 1B. Results plotted as for Figure 1C. Error bars are ± 1SD. (C) The two subunits of ribonucleotide reductase (Rrm1, Rrm2) do not change in abundance during early development. Western blotting of unfertilized eggs and embryos from two different females frogs was quantified (bottom panel). (D) Time lapse imaging of the MBT in unperturbed embryos (top panel) and embryos incubated in 20 mM hydroxyurea (HU, bottom panel). The animal pole of single embryos was imaged and the timing of each cytokinesis recorded. The time interval between each synchronous group of cell divisions is noted. The two embryos are descended from different frogs on different days. There was notable variation in the time interval between cleavage events in otherwise unperturbed, imaged embryos (data not shown), so the longer intervals in HU treated embryo are not necessarily due to HU toxicity. (E) Hydroxyurea causes hyperphosphorylation of Y15 on Cdc2 beginning at the MBT, 7.5 hours after fertilization. Embryos were incubated at 18°C for 3 hours and then the temperature slowly raised to 23°C, accounting for the slightly more rapid entry into the MBT than in Figure 1B. * indicates a non-specific band.