Measuring Gene Expression Noise A Window to the Molecular Mechanisms of Transcription

Abstract

Gene expression is known to be a highly stochastic process. Despite the noisiness of gene transcription, developmental outcomes in multicellular organisms are highly reproducible. Quantitative study of transcriptional noise will shed light on how biological organisms cope with these high noise levels. Additionally, noise provides a window into the microscopic process of transcription. As such, explaining the causes of transcriptional noise may give us insight into the mechanisms of transcription. Here, we investigate transcriptional noise in a multicellular system by observing the transcriptional dynamics of two identical genes within the same nucleus in embryos of the fruit y Drosophila melanogaster. Sites of nascent transcript formation in living embryos were observed using a synthetic enhancer. Observing transcription in two identical copies of the same gene allowed us to discriminate between intrinsic and correlated transcriptional noise. We found that both intrinsic noise and correlated noise contribute signi cantly to the variability seen in transcription, and that the relative strength of these two sources of noise varies as a function of position within the embryo. We modeled the correlated noise as a consequence of the propagation of input noise through the gene's input-output function, and we were able to extract microscopic parameters such as burst size and frequency from the intrinsic noise. These models were then tested by changing the enhancer architecture to modify the input-output function. These results demonstrate that noise can be e ectively characterized in vivo in a multicellular context, and open the door to further quantitative characterization of transcriptional activity and noise at the single cell level in living multicellular organisms.