Ketamine's Real Address: Mu-Opioid Receptors on Somatostatin Interneurons — and a Map for the Next Generation of Antidepressants

For a decade we have prescribed ketamine while admitting we did not know how it worked. NMDA blockade was the textbook story; opioid involvement was the inconvenient footnote. This Cell paper from Weill Cornell, Stanford and Johns Hopkins forces the footnote into the headline — and then uses it as a starting point for rational drug design rather than philosophical hand-wringing. The result is the most concrete mechanism-to-molecule pipeline psychiatry has produced this decade.

What the data shows

Munguba and colleagues used conditional knockouts, two-photon imaging and photopharmacology to dissect ketamine action in the mouse mPFC. When MORs were deleted specifically from SstINs, ketamine still bound NMDA receptors but lost its behavioural antidepressant signature. That is a strong dissociation — the molecule was present, the canonical target was engaged, the effect was gone. Chronic restraint stress produced bouton-level hypertrophy of SstINs and a corresponding suppression of layer 5 pyramidal firing. Ketamine — and selective MOR engagement on SstINs — normalised both. From there the team did something the field has been asking for since the SAINT and Spravato approvals: they used the cell-type-specific transcriptome to ask what other receptors live on this interneuron, and can we hit them. The screen produced a small panel of validated GPCRs whose engagement reproduced the behavioural rescue. Co-targeting two of them synergistically widened the therapeutic window — antidepressant-like effects without the locomotor and reinforcing signatures that have kept ketamine and psilocybin tethered to clinics with monitoring infrastructure.

For your practice

Three implications, in order of clinical proximity. First, the persistent debate over whether ketamine "is an opioid" stops being a culture war and becomes a circuit description: ketamine's antidepressant action runs through mu-opioid signalling in a defined inhibitory cell population, not through systemic opioidergic euphoria. Naltrexone-blunting studies in patients now have a mechanistic anchor — and so does the conversation with patients on opioid agonist therapy or in early recovery, where co-prescription warrants explicit risk-benefit framing. Second, the SstIN — pyramidal disinhibition motif joins a small but growing catalogue of cell-resolution depression circuits; if you teach trainees the monoamine hypothesis as gospel, this paper is your update. Third, the drug pipeline this work seeds — non-opioid, non-dissociative GPCR agonists targeting an SstIN-defined ensemble — is what the next decade of antidepressant development will probably look like. Patients asking "is there something coming after ketamine" now have a real answer that is neither hype nor handwave.