Distinguishing Biochemical Noise From Population Variability in Drosophila Development

Abstract

There are two predominant approaches in modern thought about reproducibility in organismal development. In the “robustness" approach, a developing embryo is faced with very noisy input signals that it must “filter out" in order to develop reproducibly, while in the “precisionist" approach, signals are reproducible from the start of development onwards. Comparison of data supporting either view is much hampered by the confounding of different sources of variability in the “robustness" literature. Here, the distinction between the potential effects of biochemical noise and genetic or environmental variation on total variability is rigorously explained, and it is shown how failure to control for genetic or environmental (population) variability can confound attempted measurements of biochemical noise. Based on experiments resolving the contributions of biological noise and population variability to left/right asymmetry in wings of inbred and wildtype Drosophila, these different sources of variation are attempted to be resolved over time from the total variability in the positions of Even-skipped domains (Eve stripes) in the Drosophila anteroposterior patterning system.

Despite the fact that the methods used are found to be insufficient to distinguish between the sources of variability in this system, it is determined that Eve stripes are reproducible to the maximal biologically-relevant degree in both the inbred and wildtype line studied and over all time, thus supporting the “precisionist" view and opening the way to using inbred lines as a control for biochemical noise in Drosophila development.