Longitudinal analysis of Drosophila larval development as a model for human disease

Abstract

Drosophila larval development has been well characterized descriptively and through endpoint analyses at the population level, but longitudinal studies across developmental timescales and at the individual level remain underexplored. In this thesis, we use the Auxodrome, a high-throughput platform for longitudinal whole-organism imaging of Drosophila development, to analyze the period between hatching and the first larval molt. Using segmentation-based downstream analysis pipelines, we reconstruct full larval body outlines during the first instar (L1), track larval trajectories, and quantify growth across time. These analyses show that L1 larvae exhibit an approximately linear growth trend and reveal movement signatures that can be used to identify molting events. This framework is then applied to a gain-of-function RAS pathway mutant, the MEK variant F53S, and its matched PAM1 control. Compared with controls, F53S larvae show reduced total displacement during the first hour after hatching and substantially longer dwelling bouts, indicating altered early larval movement dynamics. Finally, we test whether local egg density affects hatching time by comparing sparse and crowded conditions differing by 5-fold and 47-fold. Across these experiments, hatching-time distributions and hatch rates were highly similar, providing no evidence that egg density measurably alters hatching timing under the conditions tested. Together, these results establish a longitudinal framework for quantifying early Drosophila development and show its usefulness for studying subtle developmental phenotypes relevant to human disease.