Brain appetite-suppressing peptide (BRP)
A naturally occurring peptide discovered by AI that activates hunger-controlling brain cells to suppress appetite and reduce body fat; experimental, not yet an approved drug.
A researcher, an agent, or an algorithm wrote down the sequence and picked a target to hit.
An AI model like OpenFold3 or AlphaFold built a 3D structure and scored how well it fits the binding site.
A second contributor repeated the computation on their own hardware and the scores matched.
A chemistry service or a researcher ordered the sequence, it was manufactured, and mass spectrometry confirmed the right molecule was produced.
A binding or activity measurement confirmed that it actually does what the computer predicted — or didn't.
Snapshot
Class: Non-incretin POMC-activating endogenous hypothalamic peptide fragment
Evidence tier: Animal-only evidence
Status: Unapproved investigational peptide; no human trials initiated as of April 2026
Best-supported effect: Appetite suppression and fat-selective weight loss in rodent and minipig models (preclinical only)
Main caveat: No human dosing has occurred; the receptor mediating BRP's effect on POMC neurons has not been identified; all efficacy and safety data are from a single preclinical publication by one research group
What this is
BRP (BRINP2-Related Peptide) is a 12-amino-acid peptide with the sequence THRILRRLFNLC, corresponding to residues 386–397 of the human BRINP2 protein (BMP and retinoic acid inducible neural-specific 2). It is cleaved from its 78-kDa secreted precursor by prohormone convertases at flanking KK and KR recognition sites, and is detectable endogenously in human cerebrospinal fluid at approximately 700 pM to 3 nM — establishing it as a naturally occurring signaling peptide rather than a purely synthetic construct.
BRP was computationally identified by Laetitia Coassolo and colleagues in Katrin Svensson's laboratory at Stanford Medicine, using an AI-driven prohormone-cleavage prediction pipeline ("Peptide Predictor"), and published in Nature in March 2025. In animal studies, it activates pro-opiomelanocortin (POMC) neurons in the arcuate hypothalamus through a mechanism that is independent of leptin, the GLP-1 receptor, and the melanocortin 4 receptor — placing it in a pharmacological class distinct from all currently approved anti-obesity drugs. All published pharmacological data are preclinical; no human dosing has been reported.
Key structural note: Bioactivity requires C-terminal amidation (–NH₂). The non-amidated form is inactive. The minimal active sequence is LRRLFNLC (residues 5–12); L8 is essential (L8A mutation completely abolishes activity). In vivo half-life is less than 10 minutes; BRP is cleaved at R6–R7 into THRIL and LFNLC fragments, both inactive.
Evidence map
| Evidence layer | Grade | What it supports |
|---|---|---|
| Human | Not present | No human pharmacological administration has occurred; endogenous BRP measured in human CSF establishes biological presence but not therapeutic effect |
| Animal | Moderate | Acute food intake reduction (up to 50% within one hour) in lean mice and minipigs; fat-selective weight loss and improved glucose/insulin tolerance in obese mice after 14-day daily dosing; no nausea, muscle loss, or behavioral changes observed |
| In vitro | None identified | No cell assay or binding assay data identified |
| Computational | Present / discovery-context | Identified via Peptide Predictor pipeline screening ~2,683 candidate prohormone cleavage products from ~20,000 human protein-coding genes; computational step supported discovery, not binding or activity validation |
| Mechanism | Plausible | POMC/cAMP-PKA-CREB-FOS cascade characterized in preclinical work; GLP-1R, leptin, and MC4R independence demonstrated; Gαs-coupling inferred; specific upstream GPCR identity unresolved |
All animal evidence originates from a single primary publication (Nature, March 2025) by the Svensson laboratory at Stanford. Independent replication has not been reported as of April 2026.
Claim check
| Claim | Verdict | Evidence layer | Confidence |
|---|---|---|---|
| Appetite suppression and food intake reduction | Supported (animal) | Animal — rodent and minipig acute models | Medium — single publication; no independent replication |
| Fat-selective weight loss without muscle loss | Supported (animal) | Animal — obese mouse 14-day study | Medium — single publication; 14-day duration only |
| Anti-obesity effect is GLP-1-independent and MC4R-independent | Supported (animal) | Animal — mechanistic pharmacology in primary publication | Medium — demonstrated in preclinical models |
| Improved glucose and insulin tolerance | Supported (animal) | Animal — obese mouse model, secondary endpoint | Medium — secondary endpoint; single publication |
| Absence of nausea, GI disturbance, or muscle loss | Supported (animal, short-protocol) | Animal — preclinical observation window only | Low — short preclinical studies only; human tolerability unknown |
| Human efficacy for appetite suppression or weight loss | Not established | Human — no human dosing data | High confidence in "not established" |
| Safety profile comparable to or better than GLP-1 agonists in humans | Not established | Human — no human data | High confidence in "not established" |
| "Natural Ozempic" equivalence with semaglutide | Not established | Animal / mechanistic | High — mechanism is GLP-1-independent; shared downstream output does not establish clinical equivalence |
Experimental exposure
This section reports exposure used in animal experiments. It does not establish human dosing.
| Context | System | Dose | Duration | Endpoint | Notes |
|---|---|---|---|---|---|
| Dose-response (mice) | Lean mice, subcutaneous | 0.5 mg/kg | Acute | No effect | Sub-threshold |
| Dose-response (mice) | Lean mice, subcutaneous | 5 mg/kg | Acute | Potent food intake reduction | Active dose |
| Dose-response (mice) | Lean mice, subcutaneous | 20 mg/kg | Acute | Maximum suppression, ~3-hour window | Ceiling dose |
| Minipig model | Lean minipigs, IM | 2 mg/kg | Acute | ~50% food intake reduction within 1h; equieffective to ~5 mg/kg GLP-1(7-37) | Single injection |
| Chronic weight loss | Obese mice, subcutaneous | 5 mg/kg/day | 14 days | ~4 g fat-mass reduction; controls +3 g; improved glucose/insulin tolerance | Short duration; obese mouse only |
No approved human formulation, dose, or dosing schedule exists. No human pharmacokinetic data exists.
Preclinical safety signals
| Signal | System | Notes |
|---|---|---|
| No nausea or conditioned taste aversion | Mice and minipigs | Favorable preclinical signal; does not establish human tolerability |
| No gastric emptying delay | Mice | Favorable; short duration |
| No muscle loss observed | Obese mice — 14-day study | Short duration; human musculoskeletal effects unstudied |
| No movement, water intake, or anxiety-like behavioral changes | Mice | Standard preclinical battery |
| Long-term POMC activation effects | Not characterized | Receptor downregulation and compensatory activation unknown |
| Receptor off-target profile | Not assessable | GPCR identity unknown; systematic profiling not possible |
| Human pharmacokinetics (half-life <10 min in vivo) | Animal only | Rapid degradation at R6-R7; major formulation challenge for human use |
| Reproductive and developmental toxicology | Not established | No data in any species |
Regulatory status
| Region | Status | Notes |
|---|---|---|
| US (FDA) | Not approved; no IND publicly disclosed | No IND filed as of April 2026; research-chemical sales not an authorized human-use channel |
| EU (EMA) | Not authorized | Unapproved investigational peptide |
| UK, Canada, Australia, Japan | Not authorized | No authorization by any major regulatory authority |
| WADA | Not specifically listed | WADA S2 category plausibly covers BRP; no formal ruling issued |
Patent protections in place for BRINP2-derived peptide compositions for obesity therapy. Svensson co-founded Merrifield Therapeutics to advance BRP toward human trials.
Mechanism
BRP is cleaved from the secreted BRINP2 protein by prohormone convertases, which recognize KK and KR flanking motifs surrounding residues 386–397 of the 78-kDa precursor. The released 12-residue fragment (THRILRRLFNLC) is detectable endogenously in human cerebrospinal fluid at approximately 700 pM to 3 nM. C-terminal amidation (–NH₂) is required for bioactivity.
In the arcuate hypothalamus, BRP selectively activates POMC neurons via a Gαs-coupled orphan GPCR (molecular identity unresolved). Downstream cascade: Gαs → cAMP → PKA → CREB phosphorylation → FOS expression → neuronal activation. BRP triggers approximately 10× greater hypothalamic neuronal activity than full-length BRINP2. CNS FOS activation mapping shows activity in arcuate POMC neurons, DMH, POA, tuberal nucleus, and arcuate POMC-negative neurons — a broader hypothalamic engagement than initial POMC-centric framing suggests.
Mechanistic independence: BRP's anti-obesity effect is independent of leptin signaling, the GLP-1 receptor, and the melanocortin 4 receptor. The absence of GI-tract engagement is the proposed mechanistic explanation for the absence of nausea in animal models.
Primary mechanistic gap: The GPCR identity is unknown. Until deorphanized, systematic off-target profiling, structure-activity optimization, and drug-drug interaction assessment are materially constrained.
Chemistry
| Field | Value |
|---|---|
| Amino-acid sequence | THRILRRLFNLC |
| Length | 12 amino acids |
| Topology | Linear |
| C-terminal modification | Amidation (–NH₂) required for bioactivity; non-amidated form inactive |
| Minimal active sequence | LRRLFNLC (residues 5–12) |
| Essential residue | L8 (L8A mutation abolishes activity) |
| In vivo half-life | <10 minutes |
| Primary degradation site | R6–R7 cleavage → THRIL + LFNLC (both fragments inactive) |
| Brain penetration | ~10% of plasma concentration |
| Parent protein | BRINP2 (78-kDa secreted, human) |
| Parent protein residues | 386–397 |
| Cleavage sites | Flanking KK (N-terminal) and KR (C-terminal) |
| Cleavage enzyme | Prohormone convertases (PCSK1) |
| Endogenous CSF concentration | ~700 pM to ~3 nM (human) |
| Synthesis (research use) | Solid-phase synthesis, >90% HPLC purity, C-terminal NH₂ amide (e.g. GenScript) |
| Sequence verified | Yes |
Open questions
- Receptor identification: The Gαs-coupled GPCR that BRP binds on hypothalamic POMC neurons has not been identified. Deorphanization is the highest-priority mechanistic gap before human development can be fully designed.
- Human efficacy translation: No human pharmacological administration has occurred. The rodent and minipig results must translate across species.
- Human pharmacokinetics and formulation: In vivo half-life is <10 minutes in animal models; human PK is entirely unstudied. Practical dosing will require reformulation (lipidation, PEGylation, albumin binder, or depot formulation).
- Durability of weight loss: The 14-day rodent study does not characterize long-term efficacy, plateau, or rebound.
- Tolerance and receptor desensitization: Chronic BRP dosing effects on receptor sensitivity are uncharacterized.
- Comparative efficacy vs approved agents: No head-to-head comparison with GLP-1 agonists, tirzepatide, or setmelanotide in any species.
- Independent replication: All published evidence derives from one paper by one research group.
- Reproductive and developmental safety: No data in any species.
Research directions for this peptide, selected from the current sources — hypotheses you can explore and model. None of it is proven yet; tap any one to see the full thinking.
Which receptor in the brain does this peptide actually grab onto to trigger weight loss?
If BRP turns out to work through a receptor called GPR101, researchers would know exactly what to optimize the drug against, and could check in advance whether it might interact badly with other medicines. Right now that unknown is the single biggest obstacle to moving this peptide toward human trials.
Does BRP burn fat by turning up the body's internal furnace, rather than just telling you to eat less?
If the fat-burning effect runs through a brain area that controls body heat rather than hunger signals alone, it could explain why BRP targets fat without touching muscle. That distinction would matter a lot for people who want to lose fat without losing strength, and it would tell drug developers which delivery route the medicine actually needs to take to work.
Can a single small change stop the body from destroying this peptide before it can do its job?
BRP currently falls apart in the body within about ten minutes. If flipping one amino acid from its natural form to its mirror image is enough to block that breakdown while the active part of the molecule still works, it would be a simple, low-risk way to make the peptide last long enough to be a real drug, without rebuilding it from scratch.
Could a modification borrowed from existing diabetes drugs make BRP last long enough for once-daily dosing?
Every approved injectable weight-loss peptide on the market today works in part because it hitches a ride on a blood protein called albumin, which dramatically slows its clearance. If the same trick works for BRP, it could shift from something that vanishes in ten minutes to something a patient takes once a day under the skin, which is the baseline requirement for a practical anti-obesity medicine.
Can adding BRP to a semaglutide-type drug protect the muscle mass that those drugs tend to erode?
With current GLP-1 drugs like semaglutide, roughly a quarter of the weight lost is muscle, not fat, which is a real concern especially for older adults who are already at risk of weakness. Because BRP works through a completely different pathway that appears to spare muscle, combining the two could, if the hypothesis holds, deliver strong fat loss without compounding that muscle drain. That would be a meaningful clinical advance for people with obesity who cannot afford to lose lean mass.
Does this peptide only take its working shape once it meets the receptor it needs to activate?
Many peptide drugs fall apart or lose potency because they are unstructured and easy to degrade in the bloodstream. If BRP folds into its active shape only at the moment of receptor contact, then engineering a version that arrives pre-folded could make it bind faster and tighter without changing what it binds to. That could lead to a more potent drug using a smaller dose.
▸full evidence table2 metrics
| metric | value | tool |
|---|---|---|
| ipTM | 0.764664888381958 | boltz2 |
| ranking score | 0.764664888381958 | boltz2 |
▸3-letter notation
▸recipeboltz2 2.0
| parameter | value |
|---|---|
| model | boltz2 2.0 |
| weights | — |
| hardware | — |
| mlx version | — |
| python | — |
| random seed | 1 |
| msa strategy | none_monomer |
| runtime | — |
| predicted by | — |
| predicted at | 2026-05-10 |
▸citationbibtex
@peptide{pep11012,
sequence = {THRILRRLFNLC},
target = {brp-orphan},
author = {peptidemodel},
year = {2026},
status = {bioassayed}
}