hide
Free keywords:
single-cell RNA-sequencing
brain organoids
technical artifacts
Markov chain Monte Carlo
transcriptional cascades
Abstract:
Over the past decade, there has been a rapid expansion in the development and utilization of brain organoid models, enabling three-dimensional in vivo-like views of fundamental neurodevelopmental features of corticogenesis in health and disease. Nonetheless, the methods used for generating cortical organoid fates exhibit widespread heterogeneity across different cell lines. Here, we show that a combination of dual SMAD and WNT inhibition (Triple-i protocol) establishes a robust cortical identity in brain organoids, while other widely used derivation protocols are inconsistent with respect to regional specification. In order to measure this heterogeneity, we employ single-cell RNA-sequencing (scRNA-Seq), enabling the sampling of the gene expression profiles of thousands of cells in an individual sample. However, in order to draw meaningful conclusions from scRNA-Seq data, technical artifacts must be identified and removed. In this thesis, we present a method to detect one such artifact, empty droplets that do not contain a cell and consist mainly of free-floating mRNA in the sample. Furthermore, from their expression profiles, cells can be ordered along a developmental trajectory which recapitulates the progression of cells as they differentiate. Based on this ordering, we model gene expression using a Bayesian inference approach in order to measure transcriptional dynamics within differentiating cells. This enables the ordering of genes along transcriptional cascades, statistical testing for differences in gene expression changes, and measuring potential regulatory gene interactions. We apply this approach to differentiating cortical neural stem cells into cortical neurons via an intermediate progenitor cell type in brain organoids to provide a detailed characterization of the endogenous molecular processes underlying neurogenesis.
