GLP-1 drugs work. They also make a lot of people miserable. Roughly half of patients on semaglutide or tirzepatide will tell you about the nausea, the constipation, the food noise being replaced by a different kind of food noise — the kind where the thought of dinner makes your stomach turn. A meaningful chunk of users discontinue inside the first year, and the standard incretin-class drugs come with a tax most marketing decks gloss over: a substantial fraction of the weight lost is lean mass, not just fat.
So when a Stanford lab quietly published a paper in Nature in early 2025 describing a tiny endogenous human peptide that suppresses appetite by ~50% in animals without activating the GLP-1 receptor, without triggering nausea pathways, and without stripping muscle — the metabolic-research world paid attention. By 2026 that peptide, BRP, has become the most-cited "what's next after GLP-1" candidate in the field.
Here's what's actually known, what's hype, and what to watch.
The TL;DR
- BRP stands for BRINP2-Related Peptide.
- It's a 12-amino-acid sequence (THRILRRLFNLC) cleaved out of a larger ~78-kDa secreted parent protein, BRINP2, by the prohormone convertase enzyme PCSK1.
- It was identified not by classical screening but by a computational pipeline the Stanford group built called Peptide Predictor, which scans the human proteome for cryptic peptide fragments that prohormone convertases would naturally release.
- In cell assays, BRP triggered neuronal activity ~10× higher than the GLP-1 hit it was being benchmarked against.
- In lean and obese mice and Yucatan minipigs, a single dose before a meal cut food intake by up to 50% within an hour. Daily dosing in obese mice for 14 days produced ~3 g of weight loss — almost entirely fat.
- It works through a brain pathway distinct from GLP-1, leptin, and the melanocortin (MC4R) system. Translation: it's a brand-new appetite-control mechanism, not just another incretin.
- It is preclinical. No human data exists yet. The team's spinout, Merrifield Therapeutics, is the vehicle for moving it forward.
How They Found It (And Why That's Half the Story)
Most peptide drugs in clinical use today were discovered by working backward from a known hormone — GLP-1 from a Gila monster venom analog, insulin from pancreatic extracts, oxytocin from posterior pituitary lysates. The Stanford approach inverts that.
Their hypothesis: the human genome encodes hundreds of large precursor proteins that get cleaved by prohormone convertases (PCSK1, PCSK2, furin) into smaller bioactive peptides — and most of those fragments have never been characterized. The classical hormones we know about are the tip of the iceberg.
So the team — led by graduate student Laetitia Coassolo under principal investigator Katrin J. Svensson — built Peptide Predictor, a computational tool that:
- Scans every human secreted/membrane protein,
- Identifies the canonical PCSK1/3 cleavage sites,
- Predicts the resulting peptide fragments,
- Filters for sequences with biophysical properties consistent with bioactivity (length, charge, structural propensity).
Out of that pipeline came 2,683 candidate peptides drawn from 373 parent proteins. Most had never appeared in the literature. The team then ran functional screens — primarily a hypothalamic-neuron activation assay measuring c-Fos induction (a marker of neural firing) — to find the hits.
BRP was the standout.
Why this matters beyond BRP itself: Peptide Predictor is a platform. If it produced one strong hit, the realistic question is how many more anti-obesity, anti-anxiety, or pro-cognitive peptides are sitting in those other 2,682 candidates. Several biotechs in 2026 are now openly building "cryptic peptide" discovery stacks of their own.
What BRP Actually Does in Animals
The mouse and minipig data are the meat of the Nature paper. The numbers worth knowing:
| Effect | Result | Comparator |
|---|---|---|
| Acute food-intake suppression (single IP dose, lean mice) | ~50% reduction within 1 hour | Comparable magnitude to high-dose GLP-1 agonist |
| 14-day daily dosing in diet-induced obese mice | ~3 g weight loss, predominantly fat mass | Lean-mass preservation distinct from GLP-1 class |
| Yucatan minipig (closer to human GI/eating patterns) | Significant reduction in voluntary food intake | First non-rodent confirmation |
| Glucose tolerance | Improved | Independent of insulin secretion changes |
| Behavioral aversion (CTA, conditioned-taste-aversion) | None detected | GLP-1 class typically shows aversion at therapeutic doses |
| Pica (kaolin clay consumption — a rodent nausea proxy) | Not induced | Strong contrast to liraglutide/semaglutide controls |
The lean-mass-sparing point is the one most likely to matter clinically. With semaglutide and tirzepatide, depending on the trial roughly 30–40% of total weight lost is lean tissue — bone, muscle, organ mass. That's tolerable for someone with a BMI of 38 starting therapy. It's a much harder sell for the majority of GLP-1 users in 2026, who are now being prescribed for milder obesity or aesthetic indications and don't have lean mass to spare. If BRP's preclinical lean-sparing profile holds in humans, it changes the calculus.
The Mechanism: A New Lever in the Hypothalamus
This is the part the press releases all skipped over but it's where the biology gets interesting.
GLP-1 drugs work through a circuit that runs partly through the hypothalamus and partly through the brainstem area postrema — the same vomiting center that anti-emetics target. That's not a coincidence. The aversion and nausea aren't side effects in the colloquial sense; they are built into the mechanism. You cannot fully decouple GLP-1R activation from area-postrema activation in the doses needed for therapeutic weight loss.
BRP appears to bypass that entirely.
The Stanford team's tracing experiments localized BRP's central effect to arcuate-nucleus POMC neurons in the hypothalamus. POMC neurons are a known appetite-suppressing population — they're upstream of the melanocortin system — but BRP activates them through a cAMP → PKA → CREB → FOS signaling cascade that is independent of MC4R, leptin signaling, and the GLP-1 receptor. Knocking out any one of those didn't blunt BRP's effect.
Critically, BRP did not light up the area postrema in c-Fos mapping. That's the proposed reason for the no-nausea profile.
The receptor BRP binds to remains unknown as of the published paper. The team is pursuing that in follow-up work; an orphan GPCR is the leading hypothesis. From a drug-development standpoint, that's both exciting (clean new target) and risky (you can't deeply optimize a molecule until you know what it's hitting).
What This Is Not
Three caveats that matter for anyone tempted to over-index on the headlines:
1. There is no human data. None. Zero. The press cycle in 2025–2026 has compressed a preclinical paper into language ("rivals Ozempic") that implies clinical equivalence. That comparison is biological, not clinical. Plenty of peptides have looked just as good in mice and failed in Phase 1 — degradation, immunogenicity, and central penetration are all open questions for a 12-mer.
2. Half-life looks short. Native peptides this size are typically cleared in minutes. The published data uses frequent injections; a usable human therapeutic will almost certainly require stabilization chemistry — PEGylation, lipidation, D-amino-acid substitutions, or a fusion-protein scaffold — to get a once-weekly profile. Whether those modifications preserve the clean side-effect profile is unknown.
3. "Natural" is mostly a press hook. BRP is endogenous to humans, which is genuinely interesting biology, but every drug you take ends up at a non-natural concentration in a non-natural location. The pharmacology, not the origin, will determine the safety profile.
How BRP Compares to the Current Field
| Compound | Class | Weight loss (clinical) | Nausea / aversion | Lean-mass impact | Status |
|---|---|---|---|---|---|
| Semaglutide (Wegovy) | GLP-1 mono-agonist | ~15% at 68 wk | Substantial | Significant lean loss | Approved |
| Tirzepatide (Zepbound) | GLP-1 / GIP dual | ~22% at 72 wk | Substantial | Significant lean loss | Approved |
| Retatrutide | GLP-1 / GIP / glucagon triple | up to ~28.7% (Phase 3) | Substantial | Some lean loss | Phase 3 readouts in 2026 |
| CagriSema | Amylin + GLP-1 | ~22.7% at 68 wk | Reduced vs. sema-mono | Slightly better | NDA filed; FDA decision in 2026 |
| MariTide | GIP antagonist + GLP-1 agonist | Phase 2 ~20% | Mixed | TBD | Phase 3 |
| BRP | Non-incretin (novel POMC pathway) | No human data | None in animals | Lean-mass-sparing in animals | Preclinical |
The honest framing is: BRP is not yet a competitor to anything. It's a complementary mechanism that, if it survives translation, could either be combined with a GLP-1 (the way amylin is in CagriSema) or stand alone for patients who can't tolerate the incretin class.
Merrifield Therapeutics and the Path Forward
Svensson and Coassolo are listed as inventors on the BRP-related patents and co-founded Merrifield Therapeutics to develop the molecule. The company is named after Bruce Merrifield — the Nobel-winning chemist who invented solid-phase peptide synthesis, the technology that made every modern peptide drug commercially feasible.
As of 2026, Merrifield has not publicly disclosed a clinical-trial start date. The realistic path is:
- Lead optimization of BRP into a stabilized analog (likely 18–24 months from publication).
- IND-enabling toxicology (6–12 months).
- Phase 1 first-in-human — earliest plausible window 2027, more likely 2028.
- Phase 2 efficacy in obesity — 2029+.
That's roughly the same timeline as where retatrutide was in 2021, for scale.
What Researchers Are Watching in 2026
A few specific open questions the field is tracking:
- Receptor identity. Whichever receptor BRP binds will be a new drug target in its own right and immediately attract competitor programs.
- Translatability of the no-nausea profile. Pigs are a better model for human satiety than mice, and the minipig data was clean — but the area postrema in humans is more complex than in either species.
- Combination potential. Can BRP plus a GLP-1 produce additive weight loss? Many predict yes, given the entirely separate pathway.
- Effect on food noise / cravings. GLP-1 drugs do something to reward circuits that has surprised the field (the addiction-research crossover). BRP's central mechanism is different enough that whether it has any effect on craving and reward — beneficial or otherwise — is genuinely unknown.
- Other Peptide Predictor hits. Several other candidates from the same screen are in early characterization. The Svensson lab and at least two well-funded competitor labs are working through the list.
Where BRP Sits in the Bigger Story
The 2020s have been the GLP-1 decade. The 2030s look more likely to be defined by mechanistic diversity — combining incretins with amylin (CagriSema), antagonizing GIP instead of agonizing it (MariTide), pulling in glucagon (Retatrutide), and now potentially layering in a non-incretin appetite peptide that hits the brain through a different door.
BRP's biggest contribution may not be BRP itself. It's the proof-of-concept that AI-driven prohormone-cleavage prediction — searching the human proteome for peptides we already make but haven't named — can yield drug-quality hits. If that approach is general, the next decade of metabolic and CNS peptide discovery could look very different from the last one.
Worth following. Worth being calm about. Worth not buying anything sold as "BRP" online — what's circulating in the gray market in 2026 is, at best, unverified custom synthesis with no animal-grade purity data, and at worst, mislabeled.
This article is for research and educational purposes only. BRP is a preclinical investigational compound. There are no FDA-approved BRP products and no published human safety or efficacy data. Information here should not be interpreted as medical advice or as endorsement of any product. Consult a qualified healthcare professional for personal medical decisions.