A Study of Transcriptional Noise in the Early Embryogenesis of Drosophila melanogaster

Abstract

Despite the noisiness inherent to gene expression, biological organisms are able to develop with very high reproducibility. In particular, the early embryogenesis of Drosophila melanogaster has been shown to occur with very high levels of precision. Studies of noise in biological systems have found that transcriptional noise can be decomposed into intrinsic and extrinsic components. The former are due to the variability in the embryo's cellular environment while the latter are caused by the inherent stochasticity of the biochemical reactions of transcription. Quantifying and disentangling these various components of noise can thus allow us to gain various insights into both the changing environment of the cell as well as the molecular mechanisms of transcription.

In this study, we looked at the transcriptional noise in the expression of two alleles of the hunchback-P2 gene. We found that, in this out of steady-state system, removing the extrinsic component of noise allowed us to probe the temporal correlations between the expression of alleles in the same nucleus. From there, we showed that initiation mechanisms were only independent if the system performed temporal averaging. However, at an instantaneous level, we observed that initiation mechanisms are dynamically changing and that transcriptional decisions have to be continuously made and renewed throughout the embryo's development.

Moreover, we found that alleles displayed high levels of temporal correlation in their expression patterns. We were able to quantitatively determine the time scales of these correlations as well as other relevant time scales of our system. Finally, we demonstrated that physical proximity between alleles increased their correlations, indicating that cross-allele transcriptional mechanisms were regulated at a local level rather than on a global cellular level.

These elements converge to indicate that local factors are of great importance for determining transcriptional noise between the expression of alleles of the hunchback-P2 gene. These results illustrate how live-imaging data can be used to characterize noise in an in vivo multi-cellular context, and opens the door for future quantitative explorations of noise in such systems.