Map of 2D Gastruloid Morphospace

This plot was generated from a high-throughput drug screen conducted on a stem cell model of human gastrulation, the 2D gastruloid. In this screen, 210 compounds were used to perturb BMP4-initiated gastruloids, and the resulting cell fate decisions were captured with single-cell resolution. To analyze the data, we employed t-SNE (t-Distributed Stochastic Neighbor Embedding) for dimensionality reduction, followed by a custom clustering algorithm to identify distinct morphological phenotypes.

Each color corresponds to a different phenotype, with red, pink, and purple clusters representing failure modes of gastrulation. The red cluster is characterized by BRA overexpression, the pink by loss of radial symmetry and excessive cell death, and the purple by the absence of BRA expression. Notably, many known teratogens, such as Isotretinoin, are found in the purple cluster. We further tested additional compounds from this cluster, such as Nirogacestat and Ulixertinib, using a zebrafish teratogenicity assay, where both were identified as teratogens. This underscores the platform's potential for high-throughput, human-specific teratogen screening.

The remaining clusters represent variations in canonical cell patterning, where cells organized into radially patterned germ layers. Within this region, we observed that cell density predominantly influences the width of the mesoderm (BRA) band, with higher densities leading to a narrower band.

By hovering over each point on the plot, you can view the corresponding colony image and the drug treatment applied. You can also sort the plot by drug type, cell density, drug target, and drug pathway with the buttons on the right. Drug target and pathway information are based on the information provided by Selleck Chem (Stem Cell Signaling Compound Library, Cat#: L2100). This interactive feature provides a deeper understanding of the morphological phenotypes and their relationship to the various perturbations, offering insights into how different compounds influence cell fate decisions and tissue patterning in the 2D gastruloid.

 

Highlighted Publications

Tracing Information Flow from Erk to Target Gene Induction Reveals Mechanisms of Dynamic and Combinatorial Control. Maxwell Z. Wilson, Pavithran T. Ravindran, Wendell A. Lim, Jared E. Toettcher. Molecular Cell. (Download PDF)

 

Affiliated Researchers

Associate Professor
Combines tools from Biology, Engineering, and Physics to understand the cell’s perceptual field.
Postdoctoral Researcher
Joey's research focuses on the development of high-throughput engineering platforms which offer precise and tunable control over individual developmental parameters (i.e., cell number, cell type, local mechanics, etc.) in 2D and 3D stem-cell based models of embryonic development. By establishing the relationships between individual developmental parameters and their associated developmental phenotypes, he hopes to develop a more comprehensive understanding of enigmatic developmental processes, such as primate gastrulation. Specifically, Joey hopes to help answer the question of why so many primate embryos fail during gastrulation. The insights from his research can lead to improved treatments for infertility and enhanced screening criteria for quality embryos for in vitro fertilization.
Graduate Student
Naomi studies how Erk and Wnt signaling dynamics instruct stem cell fate decisions during early embryonic development.