Advances and obstacles in 4D-automated tracking technologies: assessing cell dynamics in Trisomy 21 using brain assembloids

OA Version
Citation
Abstract
This thesis investigates the challenges and current limitations of existing automated four-dimensional (4D) tracking technologies in analyzing migration patterns of cells, such as neurons and oligodendrocytes, within brain organoids. Brain organoids are three-dimensional cell aggregates which recapitulate developmental processes and structural features of the developing human brain better than two-dimensional cell culture systems. The goal of this thesis is to establish an analysis pipeline that enables automated tracking of cell migratory behavior in 4D-live cell imaging experiments with human brain organoids. Ultimately, the analysis framework developed here will be used to compare cell migration between control organoids and organoid models of neurodevelopmental disorders, such as Trisomy 21 (Down Syndrome), which involve alterations in cell migration and motility. To establish the analysis framework, I used datasets provided by the Dr. Zeldich lab that were recorded on a spinning disk confocal microscope equipped with a closed incubation chamber. Datasets consist of 4D-image volumes with organoids containing neurons and oligodendrocytes expressing enhanced green fluorescent protein. Z-stacks at 10-x magnification were acquired every twenty minutes to cover the entire organoid; recordings were performed for eighteen hours. Organoids were either derived from patients with Trisomy 21 (Down Syndrome) or represented isogenic controls. First, I developed a robust MATLAB-based pipeline that converts and pre-processes the raw imaging data. Using these pre-processed data from a subset of recordings, we will use a manual cell tracking approach to obtain quantitative metrics, including distance traveled and speed of individual cells, serving as “ground-truth” measurements. These data will be used to test for Trisomy 21-related changes in cell migratory behavior. Second, I explored several options to automatically extract cell migration data from our preprocessed datasets. Therefore, I evaluated the effectiveness of existing open-source and commercial high-end tools for segmenting and tracking dynamic cellular structures. Here, it became apparent that currently available tools struggle with accurately distinguishing continuous cellular structures from noise across multiple Z-slices and time points, and they lack the versatility in handling the heterogeneity present in our datasets, which feature both compact cell bodies and elongated, filament-like cell processes of variable length and diameter. In conclusion, this work underscores the need for advanced computational methods to accurately discern the intricate patterns of cell migration, addressing the unique challenges posed by the density and mobility of cellular components in developmental neurobiology. Solving these issues, either by adapting existing tools or developing entirely new softwares should enable more accurate and comprehensive investigations into cell migration dynamics in organoid models.
Description
2025
License