Embryonic cell fate decisions rely on transcription programs that are coordinated in both space and time. Temporal coordination is particularly important in the rapidly dividing C. elegans embryo, where developmental regulators need to be transcribed quickly enough to instruct fate decisions before the next cell cycle. This requires embryonic transcription to be very accurate. But transcription is known to be noisy, occurring in bursts followed by periods of quiescence. Thus, how transcriptional noise is prevented or reduced to allow for developmental robustness is unclear. Several early zygotic regulators in C. elegans are transcribed at very high rates. Analysis of single molecule RNA FISH (smFISH) data shows that the mRNA levels of the endoderm regulator
end-3 increase from zero to 600 transcripts in one cell within a 15-minute cell cycle. We hypothesize that high transcription rates of critical embryonic regulators can promote developmental robustness. To determine whether such high rates occur for other genes, we used single-cell RNA sequencing (scRNA-seq) data from 84,625 embryonic cells to measure maximum transcript levels at single cell resolution. We identified 20 candidate genes with high transcript abundance above a set threshold. We are using smFISH to validate the high expression of these genes and to measure the noise associated with their transcription. The expression of one candidate gene,
ceh-51, a muscle-specific transcription factor is similar to
end-3, producing 600 transcripts per cell. To estimate transcriptional bursting during
ceh-51 expression, we targeted smFISH probes to
ceh-51 introns to label sites of nascent transcription. Surprisingly, we detect transient accumulation of nuclear intron-containing transcripts within the nucleus outside of transcriptionally active sites. This suggests that transcription and splicing may be decoupled for
ceh-51, which we are exploring further. scRNA-seq identifies genes with high transcript abundance in single cells, but does not directly measure transcription rates. Hence, we are combining the metabolic labeling of mRNAs with scRNA-seq to detect recent transcription events and quantify genome-wide transcription rates during embryonic development. Our approach and methodologies can be applied to identify universal mechanisms that support rapid and temporally precise transcription.