Neuropsychiatric Drug Biosensors in Organelles, Cells, Biofluids, & Behaving Animals

Citation

Muthusamy, Anand Kumar (2024) Neuropsychiatric Drug Biosensors in Organelles, Cells, Biofluids, & Behaving Animals. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/1043-8k76. https://resolver.caltech.edu/CaltechTHESIS:06042024-004047037

Abstract

Biology is distinguished in its several levels of spatial organization—from molecular to whole body—that give rise to coherent, goal-related behaviors. Cutting across these layers are definable circuits, each with its own dynamics. These systems should be studied in their natural context to preserve their structure and function. However, quantitative measurements typically demand invasive apparatuses or infrequent, ex vivo measurements. The advent of genetically encoded fluorescent biosensors solves this problem by detecting molecules in situ, read by a microscope or implantable optical probe. These biosensors typically are fusions of a conformational switch and a fluorescent protein. A naturally occurring protein that binds the target of interest typically provides an initial scaffold. However, several molecules, particularly various human-made drugs, do not have similar naturally occurring cognate conformational switches in nature robust enough for this approach.

This work develops and applies the first genetically encoded drug biosensors in cellular and behavioral assays addressing substance abuse disorders. We term these biosensors intensity-based drug-sensing fluorescent reporters or “iDrugSnFRs.” These biosensors are based on a choline-binding periplasmic binding protein (PBP), OpuBC, interrupting a circularly permuted green fluorescent protein (GFP). This work reports a method of optimizing this construct toward the detection of several classes of neural drugs, including nicotinic, SSRIs, ketamine family drugs, and opioids.

The opportunity for continuous monitoring is particularly prominent in brain-body-behavior relationships. For example, a core tenet of behavioral neuropharmacology is the existence of some stereotyped relationship between the time course of a drug and behavioral outcomes such as opioid use disorder. Interindividual variability in pharmacokinetics (PK) complicates the problem of optimized opioid dosing, especially outside the clinic. The problem of personalizing pharmacokinetics is severe in substance use disorders: the patient must receive opioid levels that relieve pain, minimize tolerance and other side effects, and remain within a therapeutic window to maximize adherence. That ideal window is a “moving target” due to tolerance, changes in metabolism, and stressors.

As an end-to-end study with a preclinical model, the final chapter reports the development of iOpioidSnFRs and their application to the continuous monitoring of fentanyl alongside a computer vision routine to quantify behavior. The fentanyl sensor, iFentanylSnFR2.0, was expressed in the ventral tegmental area of mice and reported [fentanyl] vs. time. This recording is the longest continuous measurement of the brain [drug] alongside behavior (4 hours). We found a stereotypic, repetitive motor pattern that tracked the entire fentanyl time course (2-3 hours) despite variable PK across individuals. This result challenges current models of cellular desensitization and acute tolerance timescales. In a separate experiment, we investigated if this stereotypical pattern impaired mice in a survival task where mice forage for water through a labyrinth maze. Like in the open arena, mice in the maze exhibited circling/stalling for approximately 3 h, to the complete exclusion of successful foraging. Critically, this paradigm offers a normative definition of a deficit, as mice should have a baseline level of successful foraging to survive. We introduce this task to the substance use disorder field as an additional metric for the deficits caused by opioid administration.

Finally, this work demonstrates the utility of iOpioidSnFRs in diagnostic tests owing to their suitable aqueous solubility, dynamic range, sensitivity, selectivity, kinetics, and stability after lyophilization. Plate reader assays using iFentanylSnFR2.0, iS-methadoneSnFR, iTapentadolSnFR, and iLevorphanolSnFR provided quantitation across the pharmacologically relevant concentration ranges. These biosensors were also used in a simulated field test using readily available parts: dark box, blue LED strips, band pass filter, and a cellphone camera. This test could be used to determine the presence of a health hazard in the environment (e.g., fentanyl) or determine the exposure level in a person. These results encourage diagnostic and continuous monitoring approaches to personalizing opioid regimens.

Thesis Files