Background

Early embryo development in many species begins with a cleavage stage, in which the single-celled zygote undergoes an iconic series of rapid, synchronous cell divisions without much mass increase or transcriptional activity. The end of this cleavage phase—the midblastula transition (MBT)—is associated with two other phenomena: the cell cycle slows, and zygotic genome activation (ZGA) begins.

The MBT, with its attendant changes in cell cycle length and transcription, is well-established to be controlled by the ratio of nuclear (i.e. DNA) to cytoplasmic material, or “N/C ratio”—for example, removal of cytoplasm accelerates MBT timing, while haploid embryos have delayed MBT. However, as cell cycle length and ZGA are themselves also highly interdependent (for example, prolonging the cell cycle with cell cycle inhibitors can induce premature ZGA), it has proven challenging to disentangle how exactly the N/C ratio affects either of these processes individually.

Such an investigation would require both high temporal resolution (to distinguish initial changes from secondary effects) and spatial information (as many early genes have specific spatial patterning). Moreover, ploidy manipulations—a common method of increasing DNA amounts in vivo—add the additional confounding factor of increased template availability.

A previous report by these authors (Amodeo et al., 2015) provided an ingenious demonstration of N/C-ratio-dependent transcription above the threshold of approximately 50 ng DNA/µL cytoplasm. However, this study had not identified which gene transcript(s) specifically show N/C-ratio-dependence; it was also an in vitro assay controlling N/C ratio by adding purified sperm chromatin to cytosolic extracts of unfertilized Xenopus eggs.

 

In this study, the authors address many of the aforementioned technical limitations by using the fluorescence-based MS2-MCP system in Drosophila to monitor the transcription of several candidate genes in vivo. This system first involves the insertion of multiple MS2 repeat DNA sequences into the gene of interest. When transcribed, the MS2 sequences form mRNA stem loops, which are bound by the MCP-GFP protein to activate fluorescence locally, rapidly (with <30 s temporal resolution), and in a manner proportional to transcription levels.

In order to distinguish effects of N/C ratio and cell cycle length on transcription, they performed measurements in several genetic mutants: WT diploids, haploid mutants (which have reduced N/C ratio—but, potentially confoundingly, also shorter cell cycles), and short-cell-cycle diploid mutants (to isolate effects of short cell cycle duration). This study particularly focused on characterizing the transcription of some genes previously found to have N/C-ratio-dependent expression, and some that did not—such as knirps (kni) and snail (sna), respectively.

Results

The first findings were unexpected: regardless of their prior classification, all studied genes were found to have similar transcriptional deficits in haploid mutants. Transcription of the N/C-independent gene sna and the N/C-ratio-dependent genes kni and gt were all delayed by one cell division cycle in haploid embryos relative to WT diploids, suggesting that the N/C ratio may have a greater effect on transcription than previously anticipated.

The authors note that maximum fluorescence amplitude (maximum transcriptional activity per cell) of several genes was reduced in haploids compared to WT, but they attribute this to the shorter cell cycle and consequent early termination of transcription in haploids—transcriptional activity never reaches its steady state maxima. In support of this interpretation, they find no difference in maximum transcriptional activity between haploids and short cell cycle diploid mutants. They thus conclude that the N/C ratio affects cell cycle duration and thereby has an indirect effect on the transcriptional output of all genes.

After finding that all tested genes showed no difference in transcriptional output per nucleus between haploids and WT, the authors looked at probability of transcriptional activation, and found that the proportion of nuclei actively transcribing each gene also remained constant throughout cell division cycles 11–15 in both haploids and WT.

 

One exception stood out, however. Unlike other genes tested, the cell cycle regulator frs displayed rapid switchlike activation at a very specific stage: it was expressed by 16% of WT nuclei in cycle 12, then 80% in the subsequent cycle 13. This increase was delayed by one cell cycle in haploids, occurring between cycles 13 & 14. Crucially, however, this occurred at the normal time (cycles 12–13) in short-cell-cycle diploid mutants, indicating that the delay in activation was not due to cell cycle duration defects, and supporting the authors’ conclusion that frs activation is regulated by ploidy—i.e. directly by the N/C ratio.

Questions for the authors:

• What proportion of genes would you speculate to be directly N/C-ratio-dependent?
• What components of frs transcription regulation might you expect to be involved in N/C ratio sensing?
• Have you tried any other methods of changing N/C ratio? The triumvirate of factors (N/C ratio, cell cycle length, transcription activation) are all so interdependent that it seems like it would be hard to conclude from just one perturbation that N/C ratio is the reason for the delay of frs activation in haploids!
• Do you expect that switchlike genes (similar to frs) would be likely to be N/C-ratio-dependent? In other words, would a good way to find directly N/C-ratio-dependent genes be to look for zygotic genes that show such transcriptional profiles?

References

Amodeo, A.A., Jukam, D., Straight, A.F., & Skotheim, J.M. (2015). Histone titration against the genome sets the DNA-to-cytoplasm threshold for the Xenopus midblastula transition. Proceedings of the National Academy of Sciences of the United States of America, 112(10), E1086-95. https://doi.org/10.1073/pnas.1413990112

 

doi: https://doi.org/10.1242/prelights.14492

Read preprint (No Ratings Yet)