MDMA works nothing like a psychedelic and nothing like an antidepressant. It is a monoamine releaser, closer in mechanism to amphetamine than to psilocybin. Like amphetamines, it reverses the serotonin, dopamine, and norepinephrine transporters, forcing these neurotransmitters out of neurons rather than simply blocking their reuptake. Serotonin release dominates (Johnson MP et al., 1986, for review: Feduccia and Mithoefer, 2018).
The therapeutic action in PTSD is driven largely by what MDMA does to the amygdala. Neuroimaging shows that MDMA reduces amygdala cerebral blood flow and increases functional connectivity between the amygdala and hippocampus (Carhart-Harris et al., 2015). Separately, MDMA decreases insula and salience network connectivity (Walpola et al., 2017). The amygdala generates the fear signal. The salience network, anchored by the anterior insula and dorsal anterior cingulate, detects internal arousal states and decides what gets attentional priority. MDMA turns both down simultaneously: reduced amygdala blood flow (Carhart-Harris et al., 2015) and decreased salience network connectivity (Walpola et al., 2017) mean the brain is less likely to flag a traumatic memory as an active threat requiring a fight or flight response.
At the same time, increased amygdala–hippocampus connectivity restores the link between the emotional charge of the memory and its spatiotemporal context: the hippocampus stamps the fear signal with "past, not present." In PTSD, this connectivity is reduced (Sripada et al., 2012), which means the amygdala fires but the hippocampus cannot contextualize it. The fear feels present, not past. That is the definition of a flashback. MDMA reverses both sides of the problem: it suppresses the threat signal and reconnects it to context, so the patient can approach the trauma and engage with it in a therapeutic session without the usual terror shutting the process down.
Memory reconsolidation theory offers a mechanism. When a memory is retrieved, it briefly enters a labile, unstable state before being re-stored. If the emotional context during retrieval is less threatening, as MDMA creates, the memory may be reconsolidated with a reduced fear tag. BDNF signaling driven by MDMA, particularly in the amygdala, supports the synaptic remodeling that makes this possible (Young et al., 2015, for review: Sottile and Vida, 2022).
The critical feature of MDMA is that it requires a guide. The drug opens the window. The therapist and patient do the work inside it. Memory reconsolidation under MDMA is directed, not autonomous.
Methylone (TSND-201, Transcend Therapeutics) works the same monoamine machinery. It is a substrate at the serotonin, norepinephrine, and dopamine transporters, driving release of all three. It has no meaningful activity at 5-HT2A receptors (Warner-Schmidt et al., 2025). No hallucinogenic component. No altered state of consciousness. In their IMPACT-1 Phase 2 trial, four oral doses given one week apart produced rapid and durable reductions in CAPS-5 scores in patients with severe PTSD, with statistically significant improvement by Day 10 that held through Day 64 (Jones et al., 2026). Preclinical data show that methylone promotes neurite outgrowth and enhances fear extinction learning (Warner-Schmidt et al., 2025), suggesting the durable clinical effect is driven by structural neuroplasticity rather than by a guided therapeutic experience. In IMPACT-1, methylone was administered without accompanying psychotherapy during dosing, making it a standalone pharmacological treatment. This raises a pharmacologically important distinction within the entactogen class: MDMA opens a window for directed memory reconsolidation, while methylone appears to act through neuroplastic remodeling of fear circuitry without requiring conscious engagement with the traumatic memory during dosing.
Where noribogaine forces memories to the surface through dissociation, MDMA lowers the emotional barrier that keeps patients from approaching them voluntarily.
In Part 3, we examine a third route entirely: psilocybin, which does not access traumatic memories directly at all, but dismantles the cognitive architecture that holds them in place.
References