Psychedelic compound DOI reverses synaptic plasticity in rat brain’s claustrum

Researchers from the University of Kentucky have shown that the synthetic psychedelic compound – 2,5-dimethoxy-4-iodoamphetamine – can flip the direction of synaptic change in a key brain circuit, turning long-term depression into long-term potentiation in neurons linking the claustrum and anterior cingulate cortex. The findings suggest a possible cellular basis for the long-lasting therapeutic effects reported after psychedelic treatment for depression, anxiety, and substance use disorders

Psychedelic compounds have attracted attention for their potential to treat psychiatric disorders, including depression, anxiety, and substance use disorder. Clinicians have reported lasting benefits after only one or a few sessions which implies that these drugs may trigger long-lived changes in brain circuits.

One region of particular interest is the claustrum (CLA) which is a thin sheet of grey matter that sits between the insula and the striatum and connects with many cortical areas. The CLA expresses serotonin 2A and 2C receptors – 5-HT2A/5-HT2C – at very high density, yet little work has tested how psychedelics affect synaptic plasticity there.

In a study from team at the University of Kentucky, College of Medicine, Department of Neuroscience, in Lexington, they examined whether the psychedelic 2,5-dimethoxy-4-iodoamphetamine (DOI), a strong 5-HT2A receptor agonist, changes long-term synaptic plasticity in CLA neurons that project to the anterior cingulate cortex (ACC), a region involved in cognitive control and affect.

The authors recorded from these projection neurons in acute brain slices from male rats. They used a standard spike-timing protocol: an electrical stimulus to presynaptic inputs to elicit an excitatory post-synaptic potential (EPSP), followed 10 ms later by a current pulse in the recorded neuron to evoke a single action potential.

Repeating this ‘pre-then-post’ pairing usually increases synaptic strength – long-term potentiation (LTP) – in many brain areas, consistent with Hebb’s rule that near-simultaneous pre- and post-synaptic activity tends to strengthen connections.

Surprisingly, under control conditions the CLA-ACC synapses weakened rather than strengthened. The same pre-then-post protocol produced long-term depression (LTD): EPSPs fell and remained low for the remainder of the recording.

To check that this effect was specific to the stimulated synapses, the team measured spontaneous EPSPs that arise from ongoing network activity. Their frequency, size, and duration did not change, which argues that the LTD was restricted to the pathway the stimulating electrode activated, not a general drop in release probability across the cell’s inputs.

The key result emerged when the authors bath-applied DOI at 10 µM. With DOI present, the identical pre-then-post protocol now produced robust and sustained LTP. Relative to control, effect sizes were large at both early and late time points. Spontaneous EPSPs again stayed stable, which indicates that DOI did not globally increase excitatory drive but instead shifted the plasticity rule at the activated synapses from depression to potentiation. In short, DOI reversed the polarity of long-term plasticity in this CLA-to-ACC circuit.

The authors also probed intrinsic electrical properties because changes to a neuron’s own excitability can shape plasticity outcomes. Resting membrane potential, input resistance and capacitance were similar with, and without, DOI. However, DOI altered the shape of single action potentials: spikes became broader, with longer rise and decay times, and the afterdepolarisation potential (ADP) became smaller.

A broader spike often points to modulation of potassium currents that repolarise the membrane and help terminate the action potential. Many serotonin receptor effects converge on potassium channels – for example, M-type currents carried by Kv7 channels – so the waveform changes align with known 5-HT actions on membrane conductance.

These findings matter for two reasons. First, they provide a cellular mechanism that could help explain why psychedelics can deliver durable clinical benefit. If a psychedelic can flip an anti-Hebbian LTD rule into an LTP rule in a circuit that regulates salience, control and affect, then experience-dependent strengthening of specific pathways may follow the drug session and consolidate therapeutic learning. Secondly, they support the idea that the CLA is a key node in psychedelic action.

The cortico-claustro-cortical model proposes that the CLA filters and coordinates cortical excitability across networks. If psychedelics shift synaptic weights within claustro-cortical loops, that could contribute both to the acute subjective effects and to longer-term restructuring of network dynamics.

The results also fit within a broader framework for how neuromodulators shape brain plasticity. The authors discuss the concept of an ‘eligibility trace’: a transient tag at an active synapse that requires a neuromodulatory signal to convert into persistent LTP or LTD.

DOI, by activating 5-HT2A receptors, may supply or amplify that permissive signal so that a pre-then-post pattern which normally yields LTD now yields LTP. This mechanism could intersect with evidence that psychedelics reopen critical-period-like plasticity windows in adult animals, making circuits more malleable for a period of time.

The extracellular stimulation may have recruited both excitatory and inhibitory axons, so inhibitory inputs could contribute to the apparent polarity switch. Serotonin is known to enhance some GABAA-mediated currents, and interaction between 5-HT receptors and potassium channels can further complicate net effects on inhibition. Dissecting excitatory and inhibitory contributions with receptor-specific blockers or optogenetic input selection would clarify the mechanism.

Timing rules also warrant mapping: the authors used a fixed 10 ms pre-then-post interval, but serotonin can widen or shift spike-timing windows. It would be useful to chart how different delays, including post-then-pre, behave with and without DOI.

Finally, translation to behaviour will require attention to context. Human psychedelic effects depend on set and setting, and animal studies show that prior experience can gate plasticity. Those factors may interact with the cellular rules described here to determine which synapses tag and consolidate after a dosing session.

The study provides the first direct evidence that a psychedelic can reverse the sign of spike-timing-dependent plasticity in a claustro-cortical pathway. Under control conditions, pre-then-post pairings produced LTD at CLA-ACC synapses; with DOI, the same pairings produced LTP. DOI also broadened and slowed action potentials in a way consistent with potassium channel modulation. Together, these results support a model in which psychedelics induce rapid and lasting changes to synaptic strength in serotonin receptor-rich circuits such as the claustrum – a plausible cellular basis for their enduring therapeutic effects.

The research team for this paper were Dr. Tanner L. Anderson, medical student Artin Asadipooya and Dr. Pavel I. Ortinski.

For further reading please visit: 10.1523/ENEURO.0047-25.2025