Automated image analysis for tracking cargo transport in axons

Abstract

The dynamics of cargo movement in axons encodes crucial information about the underlying regulatory mechanisms of the axonal transport process in neurons, a central problem in understanding many neurodegenerative diseases. Quantitative analysis of cargo dynamics in axons usually includes three steps: (1) acquiring time-lapse image series, (2) localizing individual cargos at each time step, and (3) constructing dynamic trajectories for kinetic analysis. Currently, the later two steps are usually carried out with substantial human intervention. This article presents a method of automatic image analysis aiming for constructing cargo trajectories with higher data processing throughput, better spatial resolution, and minimal human intervention. The method is based on novel applications of several algorithms including 2D kymograph construction, seed points detection, trajectory curve tracing, back-projection to extract spatial information, and position refining using a 2D Gaussian fitting. This method is sufficiently robust for usage on images with low signal-to-noise ratio, such as those from single molecule experiments. The method was experimentally validated by tracking the axonal transport of quantum dot and DiI fluorophore-labeled vesicles in dorsal root ganglia neurons.