UCSF

Neurological and psychiatric disorders (such as depression, neurodegeneration, and drug addiction) are among the greatest challenges faced by modern medicine. Though they are some of the most prevalent and debilitating illnesses in the world, current treatment options leave much to be desired due to limited efficacy or intolerable side effects. Our understanding of the nervous system is still rudimentary, preventing the development of new drugs that precisely target the pathogenic mechanisms of disease. Instead, most neuroactive drugs in clinical use were either developed decades ago or are chemical derivatives of these legacy drugs. Typical target-based approaches to drug screening have had limited success in psychiatric drug discovery due to inadequate mechanistic understanding and the multifactorial nature of the disorders they aim to treat. As a result, neuroactive drug development has stagnated while pharmaceutical companies focus their resources on less risky endeavors. How can we overcome these challenges to accelerate psychoactive drug discovery in the future? One approach is phenotypic drug screening based on observations of animal behavior. This strategy has the potential to uncover behaviorally active compounds with unprecedented mechanisms but is limited by the difficulty of efficiently measuring animal behaviors at scale. This dissertation documents contributions to phenotypic drug screening utilizing the zebrafish model organism, including the identification of meaningful behavioral phenotypes which are accessible in a large-scale context and new methods to automate the quantification of these behaviors. In Chapter 1, the influence of neuroactive drugs on zebrafish buoyancy is investigated using a computational method to approximate the depth of larvae in a 96-well plate. This approach is easily scaled up, allowing us to rapidly screen thousands of molecules to discover new ligands for adrenergic and serotonergic receptors which alter behavior in vivo. In Chapter 2, we explore phenotypes related to paradoxical excitation during early-stage general anesthesia – a complex and poorly understood behavioral state that would be very difficult to study in systems other than intact animals. By identifying compounds from a behavioral screen that phenocopy the response of zebrafish to anesthetics like etomidate and propofol, we uncovered new GABAA receptor modulators with potential anesthetic activity in vivo. In summary, the work included in this dissertation provides new methods to quantify zebrafish behavior on a high-throughput scale and demonstrates their use by identifying several new behavior-modifying small molecules.