Project Summary
Millions of precise cellular decisions underlie the accurate development of a brain. Small molecules can
impinge on these decisions by binding to proteins with critical roles in lineage trajectories. They can tune
circuits by boosting or depleting specific types of neurons or their connections. A “dictionary” of chemical
compounds and their effects on the brain would be an invaluable resource for classifying environmental
neurotoxins and discovering treatments for neurodevelopmental disorders. Yet, we lack efficient approaches to
quantify these effects.
We propose a multi-pronged approach to screen chemicals for their neurodevelopmental roles, integrating
whole-organism screening, protein engineering, and computational drug discovery. The larval zebrafish will be
our testing ground for these molecules and new tools. First, we will rely on improved versions of our validated
pipeline for high-throughput analysis of behavior, brain activity, and brain structure. We leverage recently
developed methods for whole-brain activity mapping in freely swimming animals, building upon our screens of
genetic mutants for genes that increase risk of neurodevelopmental disease. Our efforts will focus on autism
and intellectual disability: assessing environmental contaminants indicated to increase rates of these disorders
and screening libraries of drugs in clinical development to accelerate treatments. Second, to surpass the
throughput limitations of whole animal screening, we will engineer a tool to convert information about the
presence and activity of specific neuron types into a DNA readout. Larvae exposed to compounds in 96-well
plates will be multiplexed with next-generation sequencing. Finally, the frontier of drug discovery is
computational. There are billions of potentially synthesizable molecules, and it is inconceivable to test even a
small percentage experimentally. We are developing a new algorithm for computational-based prediction of
chemical-protein interactions that integrates a knowledge-based approach with the physical energy potential of
the Rosetta modeling program. Our ongoing studies of the basic biology of the genes linked to
neurodevelopmental disorders will yield the protein targets for modeling.
With each proposed technological advance, we will increase the number of molecules we can analyze by an
order of magnitude. The power of our strategy lies in casting a wide net to define molecules with diverse yet
specific means of manipulating neural development and activity. We will push forward the prevention,
understanding, and treatment of neurodevelopmental disorders by considering both the chemicals that can
drive these disorders and the ones that can reverse phenotypes to treat them.