Drug development for the central nervous system (CNS), especially for psychiatry, has slowed, partially
because we do not know the mechanisms by which some drugs exert their therapeutic or harmful effects. The
project provides data to test the hypothesis that several CNS drugs act, in addition to their acute effects, in a
slower, “inside-out” fashion. The drugs would start by binding to their classical molecular targets, but in
organelles.
By measuring neural drugs, and their target interactions, within organelles of living cells, this project
helps to test inside-out pharmacology. The experiments invent, then exploit, genetically encoded fluorescent
biosensors to measure drugs in organelles. The biosensors are bacterial and archaeal periplasmic binding
proteins (PBPs), fused to circularly permuted green fluorescent protein (cp-GFP). Sub-Approach A is a
solution-based screen of drugs x existing biosensors. The library of 92 compounds includes many orally
available drugs approved for various indications, but emphasizing psychiatry. The collection of 60 purified
biosensor proteins comprises five existing families, which now sense glutamate, dopamine, GABA, and
serotonergic drugs. Sub-Approach B utilizes “directed evolution” to improve the “hits”, toward the goal of
detecting the drugs at pharmacologically appropriate sub-micromolar concentrations. The major tools—site-
saturation mutagenesis, atomic-scale structure, computational docking, and high-through fluorescence
screening--are expected to converge on appropriate biosensors. Sub-Approach C expresses the refined
biosensors in ER and performs live-cell, time-resolved imaging while the drugs are applied extracellularly. We
begin with the simple questions, “does the drug enter the ER, and how quickly?” We then analyze signals
within organelles that also express the classical targets for the drugs. We expect a rich set of data on “kinetic
buffering” of diffusion by binding to the targets within ER, thus revealing drug-receptor interaction within
organelles of live cells. The sub-approach then graduates to mouse preparations, using viral vectors, brain
slices, and two-photon imaging in intact animals will be employed. Sub-Approaches D and E complement each
other. D extends subcellular pharmacokinetics to acidic organelles, including secretory granules and
neurotransmitter vesicles already suspected of accumulating drugs via “acid trapping”. We'll retain the PBP
portions of the biosensors, but employ additional cp-fluorescent proteins, known to function at low pH, and also
modify linkers. The result will become a collection of fluorescent biosensor platforms, each specialized to
perform best within, and targeted to, a class of organelles. Sub-approach E extends the drug biosensor
strategy to new classes of PBPs, and to new classes of drugs. We will retain the cp-fluorescent protein part of
the biosensors, but optimize the new PBPs and linkers.
The transformative overall results will produce at least ten, and as many as 100, biosensors to detect
drugs within organelles, and a clear roadmap for subcellular pharmacokinetics as a robust research tool. Data
could suggest transformative therapeutic strategies for psychiatry, addiction, and neurodegeneration.