Strategic Partner(s)

Bryan Roth: Inside the DARPA Bet on a Non-Psychedelic Psychedelic

Psilocybin is in Phase 3 trials for depression, and the early signal is striking — a single dose followed by months of remission. The therapeutic mechanism, however, is inseparable from the psychedelic experience itself, and not every patient can or should take a psychedelic. This raises one of the most consequential questions in current GPCR pharmacology: can biased signaling at the serotonin 2A receptor decouple the therapeutic effect from the trip?

In this conversation, Dr. Bryan Roth describes the DARPA-funded effort his lab is leading to answer that question through structure-guided drug discovery, biased ligand design, and high-throughput functional assays. The discussion ranges across four decades of receptor pharmacology — from the era when the field debated whether opioid receptors were even proteins, to the cryo-EM-driven structure-function loops that now turn over in three weeks.

Dr. Roth is unusually candid about what drug discovery against GPCRs actually requires, why most hypotheses still fail at the indication rather than the chemistry, and why a relative's psychotic break when he was five years old set the trajectory that eventually led him here.

About the Guest

Dr. Bryan Roth is the Michael Hooker Distinguished Professor of Pharmacology at the University of North Carolina Chapel Hill School of Medicine. He earned his MD and biochemistry PhD at St. Louis University and trained in pharmacology at NIH and in molecular biology and psychiatry at Stanford. His lab works on serotonin, dopamine, and opioid receptor structure, function, and drug discovery, with a particular focus on neuropsychiatric disease. He invented DREADD technology, leads the NIH Psychoactive Drug Screening Program, and has contributed to structural and pharmacological characterization of dozens of GPCRs.

Scientific Themes of the Conversation

Biased agonism at the serotonin 2A receptor and the pursuit of a non-psychedelic psychedelic
Structure-guided GPCR drug discovery in the cryo-EM era
Why drug discovery bottlenecks are biological, not chemical
Polypharmacology for complex neuropsychiatric disease
Functional assay design — BRET, Tango, and what each captures
The historical evolution of receptor pharmacology, from radioligand binding to atomic-resolution structure

Key Insights from the Conversation

The bottleneck in GPCR drug discovery is the indication, not the chemistry. Dr. Roth is direct about this: with sufficient medicinal chemistry resources and modern assays, you can make a biased ligand for almost any pathway. What you usually can't do is know which pathway, in which disease, will actually be therapeutic. The hypothesis is what fails.

Every GPCR is probably good for something — and we mostly don't know what. This is a quiet conviction Dr. Roth has held for years. It reframes the field's task from finding good targets to understanding the biology well enough to connect targets to indications, a problem he believes will take decades more.

A single-target drug may be the wrong goal for complex disease. Clozapine, the most effective antipsychotic, hits at least fifty targets. For schizophrenia and other neuropsychiatric conditions, the genetic landscape may simply be too complex for one-receptor drugs to win. The future, Dr. Roth argues, belongs to designed polypharmacology.

The DARPA non-psychedelic psychedelic project is real, early, and moving fast. The premise: psilocybin's therapeutic effect on depression is now well documented, but it is locked to the psychedelic experience. The bet is that biased ligands at the serotonin 2A receptor can separate the two. Six weeks in, structures are turning over and chemistry is iterating.

Sometimes a discovery validates itself overnight in ways you didn't sanction. Dr. Roth's lab identified salvinorin A as a kappa opioid receptor agonist. The internet seller who supplied the compound disagreed, took naloxone, smoked salvia to disprove it, and emailed back the next morning conceding the point. A reminder of what scientific certainty sometimes looks like outside the lab.

The "you'll never make it" verdict can be wrong, repeatedly. A famous scientist at Stanford and an NIH program officer both told Dr. Roth, in different years, to give up basic science. He didn't. The arc from a lab "about as big as your car" at Case Western to leading a thirty-person GPCR lab at UNC is worth sitting with — particularly for early-career scientists who feel the same verdict closing in.

BRET captures the initial event; second messengers can mislead you. For analytical work on biased ligands, Dr. Roth's lab leans heavily on BRET-based G protein and arrestin assays because they're relatively insensitive to receptor reserve and capture coupling close to the receptor. Tango and calcium assays are reserved for screening, where false negatives are the real cost.

Episode Timeline

00:00 Introduction
02:03 Witnessing a psychotic break at age five
05:00 The visiting professor's lecture that sealed it: receptors
06:30 When the field debated whether opioid receptors were even proteins
11:31 The "dumb luck" of starting on serotonin receptors at NIH
16:43 Studying functional selectivity for the Navy in 1987
22:50 "You're never going to make it as a basic scientist"
28:13 The lab the size of a car at Case Western
35:26 The DARPA project: a non-psychedelic psychedelic for depression
38:21 "I am not uncertain" — the data Bryan won't talk about yet
47:13 Why BRET captures what second-messenger assays miss
51:01 Why most GPCR drug hypotheses fail — and it isn't the chemistry
55:38 Fifty targets, one drug: clozapine and complex disease
01:03:06 The kappa receptor validated in a human trial overnight
01:06:00 Solving the lab's first GPCR structure
01:08:33 Advice for junior scientists who don't have ten Nature papers

Timestamps were generated using AI for readability and may shift slightly relative to the final edited cut.

Selected Quotes

"Every GPCR is good for something, basically. If we make a drug against any GPCR, it's going to be good for some disorder. We don't know what that is. That's the problem."

"The reason that so many things fail is not because a chemist didn't do their work, but because the hypothesis was basically wrong. The idea was, if we make a drug against this target, it will be useful in this disease. And that just may be the wrong hypothesis. It typically is the wrong hypothesis."

"I was too stupid to listen to him, basically, is what I tell people."

"Find the thing that you're really, really passionate about. Find that. And don't pay so much attention to the outcome. It's the doing that's more important."

About this episode

Dr. Bryan Roth is the Michael Hooker Distinguished Professor of Pharmacology at the University of North Carolina Chapel Hill School of Medicine. After receiving his MD and Ph.D. in Biochemistry from St. Louis University in 1983, he subsequently trained in pharmacology (NIH), molecular biology, and psychiatry at Stanford. Bryan leads a $26.9 Million project to create better psychiatric medications, among other things.

But, it wasn’t always this way. Bryan got fascinated with receptors after attending the lecture of a visiting professor on neurotransmission while in college in Montana, where he is from. He was determined to work on psychiatric disorders and persevered even when he was told several times he would never make it as a basic scientist and would never publish anything important. Join me and learn more about Bryan and his work.

Dr. Bryan Roth on the web