Psilocybin’s acute disruption of the default mode network (DMN) correlates with antidepressant response, but the link to lasting structural change remains indirect. We hypothesize that psilocybin potentiates glymphatic clearance during subsequent sleep, removing extracellular metabolites that otherwise inhibit BDNF/mTOR‑driven synaptogenesis. This clearance creates a permissive microenvironment for dendritic spine growth and synaptic density increases observed in rodent cortex and hippocampus, thereby stabilizing the DMN decoupling seen in functional imaging. Specifically, 5‑HT2A receptor activation on astrocytes triggers intracellular calcium waves that upregulate aquaporin‑4 polarity, boosting perivascular CSF‑interstitial fluid exchange. Enhanced glymphatic flow reduces amyloid‑β, lactate, and inflammatory cytokines, which are known to suppress mTOR signaling and spine formation. Consequently, the BDNF surge induced by psilocybin can act on nascent synapses without opposition, leading to persistent (+10% spine density) and (+6‑9% synaptic density) changes that outlast the drug’s pharmacological presence.

To test this, we propose a randomized, double‑blind, placebo‑controlled trial in treatment‑resistant depression (N=80) receiving a single 25 mg oral psilocybin dose or identical placebo. Participants will undergo simultaneous multimodal imaging at baseline, 6 h, 24 h, 72 h, 1 week, and 1 month: (1) resting‑state fMRI to quantify DMN within‑network connectivity and cross‑network integration; (2) dynamic contrast‑enhanced MRI with intrathecal gadobutrol as a glymphatic tracer to measure perivascular clearance rates; (3) PET with [11C]‑BDNF analog (if available) or CSF BDNF ELISA to track neurotrophic signaling; and (4) MR spectroscopy for lactate and glutamate as metabolic indices of glymphatic efficacy. A subgroup will receive acetazolamide (a known glymphatic inhibitor) before sleep on the first night post‑dose to pharmacologically suppress clearance. The primary prediction is that psilocybin will increase glymphatic influx/efflux by ≥20% during the first sleep episode relative to placebo, and this increase will mediate ≥30% of the observed DMN decoupling at 1 week (mediation analysis). Furthermore, blocking glymphatic flow with acetazolamide will abolish the psilocybin‑induced rise in spine‑density proxy (measured via synaptic vesicle glycoprotein 2A PET) and prevent the long‑term reduction in DMN connectivity, despite equivalent acute 5‑HT2A occupancy (verified by [11C]‑ketanserin PET). Failure to observe these relationships would falsify the hypothesis that sleep‑dependent glymphatic activation is a necessary conduit for psilocybin‑driven structural plasticity and sustained clinical benefit.