A bio-orthogonal ion channel that only a synthetic key can open
The Pharmacologically Selective Actuator Module (PSAM) is a chimeric ligand-gated ion channel built by fusing the extracellular ligand-binding domain of a mutated nicotinic acetylcholine receptor (α7 nAChR) to the transmembrane pore domain of a serotonin type-3 receptor (5-HT3, cation-selective) or a glycine receptor (GlyR, chloride-selective). Point mutations in the binding pocket abolish sensitivity to acetylcholine while creating a new pocket shaped only for a synthetic small molecule. The result is a channel with no natural ligand — an orphan receptor waiting for a bespoke key.
Ligand-gated ion channels (Cys-loop receptor superfamily) are built from two functionally separable modules: an extracellular ligand-binding domain (LBD) that recognizes a chemical cue, and a transmembrane pore domain that conducts ions once the LBD signals a conformational change. Because these domains are structurally modular across the Cys-loop family — nicotinic receptors, 5-HT3, GABA-A, glycine receptors all share the same fold — they can be swapped between family members while preserving gating function.
PSAM starts from the α7 nicotinic acetylcholine receptor LBD and introduces mutations at the acetylcholine-binding pocket (notably around the aromatic box residues) that eliminate responsiveness to ACh and all known nicotinic agonists/antagonists, while opening a new pocket complementary to a synthetic phenyl-based small molecule (PSEM). The mutated LBD is then genetically fused to the pore domain of 5-HT3A (permeable to Na⁺/K⁺ — excitatory) or GlyR (permeable to Cl⁻ — inhibitory), producing PSAM-5-HT3 or PSAM-GlyR channels respectively.
Because the ligand-binding pocket was engineered de novo through iterative mutagenesis and pharmacological counter-screening against native receptors, PSAM channels show no detectable activation by acetylcholine, serotonin, glycine, or any panel of ~30 endogenous neuroactive compounds tested in the original characterization.
PSAM channels are delivered to neurons using the same viral toolkit as other chemogenetic and optogenetic actuators: adeno-associated virus (AAV) carrying a Cre-dependent (DIO/FLEX) expression cassette, injected stereotactically into a target brain region of a Cre-driver mouse line. This restricts PSAM expression to a genetically defined neuron population — for example, AgRP neurons in the hypothalamus, or a specific interneuron class in cortex.
Once expressed, PSAM channels traffic to the plasma membrane like any Cys-loop receptor, assembling as pentamers and distributing across the somatodendritic surface. Because the channel is otherwise silent, its mere presence does not perturb baseline neuronal physiology — a critical property for a clean "on/off" experimental actuator that does not confound the circuit it is meant to probe.
PSEMs (Pharmacologically Selective Effector Molecules) are small synthetic agonists — PSEM89S and the higher-potency uPSEM792/uPSEM817 — engineered in parallel with the PSAM binding pocket. Delivered systemically (IP injection or oral gavage), they distribute through the body and cross the blood-brain barrier, but bind productively only where a PSAM channel happens to be expressed. Everywhere else, the molecule is pharmacologically inert.
PSEM design followed the same logic as the channel: start from a scaffold related to nicotinic ligands, then iteratively modify both the small molecule and the receptor pocket in tandem until a matched, mutually exclusive pair emerges. PSEM89S — a quaternary ammonium phenylurea derivative — was selected from a combinatorial chemistry series screened against panels of PSAM point mutants, converging on a ligand that activates only the co-evolved channel and nothing else in the genome.
The second-generation ligand uPSEM792 (and its analog uPSEM817) improved potency roughly 10- to 100-fold over PSEM89S while preserving orthogonality, allowing effective in vivo dosing in the sub-milligram-per-kilogram range and reducing the risk of off-target pharmacology from high-dose administration.
Because PSEM has no endogenous binding partner, its systemic pharmacokinetics are governed purely by distribution, blood-brain-barrier penetration, and clearance — not by target engagement outside the engineered tissue. After IP injection, PSEM reaches brain concentrations sufficient for channel activation within minutes, comparable to CNO dosing for DREADDs, but the downstream effect it triggers is mechanistically distinct once it arrives.
This clean pharmacokinetic profile is what makes PSAM/PSEM attractive not only as a research tool but as a template for engineered cell-therapy control systems: a systemically dosed small molecule that only "talks" to a receptor deliberately installed by gene therapy, with no chatter into native physiology.
Because PSEM89S and uPSEM792 are structurally unrelated to clozapine-N-oxide (CNO) and do not engage muscarinic DREADDs — and CNO does not activate PSAM channels — the two systems can be dosed in the same animal on the same day without crosstalk.
The defining mechanistic advantage of PSAM over GPCR-based chemogenetics (DREADDs) is speed. PSEM binding to the extracellular domain directly and allosterically forces the fused pore open — a single conformational relay with no intermediate signaling cascade. Ion flux begins within tens of milliseconds of ligand binding, versus the minutes-to-tens-of-minutes latency required for a Gq/Gi-coupled DREADD to accumulate second messengers and modulate excitability.
DREADDs (Designer Receptors Exclusively Activated by Designer Drugs) are engineered G-protein-coupled receptors: CNO binding triggers a heterotrimeric G-protein cycle (Gq for hM3Dq activation, Gi/o for hM4Di silencing), which then modulates downstream effectors — phospholipase C and IP3/Ca²⁺ release for Gq, or adenylyl cyclase inhibition and GIRK channel opening for Gi. Each of these enzymatic steps takes time to amplify and propagate, producing effects that build over minutes and persist for tens of minutes to hours.
PSAM channels skip this entire cascade. PSEM binding at the extracellular LBD is mechanically coupled, through the same allosteric transduction pathway Cys-loop receptors use for their natural ligands, directly to the pore gate. The channel opens as a single conformational event — the identical mechanism by which nicotinic or 5-HT3 receptors respond to acetylcholine or serotonin in tens of milliseconds during fast synaptic transmission.
Because PSAM retains the native Cys-loop receptor gating mechanism, its activation kinetics resemble fast ionotropic neurotransmission rather than neuromodulation — millisecond-scale onset that DREADDs, bound by the biochemistry of G-protein cycling, cannot physically achieve.
The same PSAM ligand-binding domain can be fused to different pore domains to select the sign of the effect:
• PSAM-5-HT3 (cation-permeable, Na⁺/K⁺): depolarizes the neuron on PSEM binding — an excitatory actuator, functionally analogous to hM3Dq but ionotropic.
• PSAM-GlyR (chloride-permeable): hyperpolarizes or shunts the neuron on PSEM binding — an inhibitory actuator, functionally analogous to hM4Di but ionotropic and fast enough to silence firing within a single behavioral event rather than over a sustained pharmacological window.
This modularity means a single PSEM ligand can drive opposite physiological outcomes purely by which pore-domain PSAM construct was virally delivered — the pharmacology stays identical, only the downstream ion selectivity changes.
Once the PSAM pore opens, the resulting change in membrane conductance produces a neuronal response with onset kinetics that approach optogenetic tools, while retaining the systemic, cell-type-restricted, hands-off delivery that makes chemogenetics practical for long, unrestrained behavioral experiments where an optical fiber is impossible or undesirable.
Optogenetics achieves millisecond-precision control of neural activity but requires an implanted optical fiber and is intrinsically limited to tethered or head-fixed preparations, or bulky wireless hardware. Chemogenetics has traditionally traded that speed away in exchange for a systemically administered, fiber-free actuator — DREADDs let an animal move freely in a large arena or a naturalistic environment for hours, but the onset lag makes it unsuitable for precisely timing activity to fast behavioral events.
PSAM/PSEM narrows this gap: because gating is ionotropic, the physiological response to a systemically delivered drug can begin within roughly a second of the drug reaching the target tissue and saturating local PSAM channels — fast enough to be used around discrete behavioral epochs, not just as a slow, hours-long state switch.
Because PSEM concentration in tissue is dose-dependent and channel open probability scales with occupancy, PSAM/PSEM provides graded, titratable control: low doses produce partial channel activation and moderate changes in firing, while higher doses drive near-maximal channel occupancy and strong, sustained depolarization or hyperpolarization. This dose-response relationship allows an experimenter to tune the magnitude of circuit perturbation to the question being asked, rather than working with a strictly binary on/off actuator — while the channel returns to baseline as PSEM clears, without the need for genetic or optical reversal.
The single property that elevates PSAM/PSEM from "a faster DREADD" to a distinct tool class is complete pharmacological orthogonality to the entire DREADD/CNO system and to native neurotransmission. Because neither ligand nor receptor overlaps, a researcher can express PSAM channels in one neural population and DREADDs in a second, independent population within the same animal, and drive each circuit separately — and even simultaneously — with two different, mutually inert drugs.
Most interesting behaviors are not the output of one circuit but the interaction of several — an excitatory drive population and an inhibitory gating population, or two parallel pathways that converge on a shared downstream target. Manipulating only one such population at a time cannot reveal how they interact; a researcher needs to activate or silence two genetically distinct circuits independently, and ideally in the same experimental session, to test necessity, sufficiency, and interaction.
PSAM/PSEM was deliberately engineered to be orthogonal to the existing DREADD/CNO system for exactly this purpose. A Cre line can direct PSAM to one cell type and a Flp line can direct hM3Dq/hM4Di DREADDs to a second, intersecting or parallel cell type in the same animal; PSEM and CNO are then administered on independent schedules, each engaging only its own receptor, to activate or silence each population on demand without any pharmacological interference between the two manipulations.
Multiplexed chemogenetics with PSAM/PSEM and DREADD/CNO has been used to independently drive and suppress opposing arms of feeding and arousal circuitry in the same mouse, distinguishing circuit elements that classical single-actuator approaches could not separate.
The broader trend this technology exemplifies is the assembly of a toolkit of mutually orthogonal actuators — PSAM/PSEM, DREADD/CNO (and newer CNO-free variants such as deschloroclozapine), and optogenetic opsins — that can be layered within a single experimental animal, each addressing a genetically and pharmacologically distinct target with no crosstalk. Combined with intersectional genetics (Cre/Flp/Dre driver combinations), this makes it possible to dissect circuits with a level of independent, multi-channel control that no single actuator technology could provide alone: fast ionotropic drive from PSAM, slower modulatory tone from DREADDs, and millisecond optical precision from opsins, all addressable independently in the same brain.