A 2025 Nature paper from Katrin Svensson's lab at Stanford ↗ introduced BRP, a 12-amino-acid peptide cleaved from the larger BRINP2 protein by endogenous prohormone convertases, that suppresses food intake by activating POMC neurons in the arcuate hypothalamus through a leptin-, GLP-1R-, and MC4R-independent pathway. The peptide reduced food intake by approximately 50% within one hour of injection in minipigs and produced 3-4 grams of fat-mass loss over 14 days in obese mice while controls gained 3 grams. The discovery method is the part of the story the platform's audience should hold most carefully: an AI-driven scan of more than 2,600 predicted prohormone cleavage products across the human peptidome, the Peptide Predictor tool from the Svensson group.
The biology. BRP (BRINP2-related peptide, sequence THRILRRLFNLC) is a 12-amino-acid fragment cleaved from BRINP2, a parent protein previously not annotated as a hormone-precursor source. The Svensson group identified BRP by computationally predicting which proteins in the human secreted peptidome would be substrates for prohormone convertases (the enzyme family that cleaves insulin, glucagon, and dozens of other hormone precursors into bioactive peptides). They then synthesized predicted cleavage products and screened them for hypothalamic neuronal activity. BRP triggered approximately ten-fold greater POMC neuronal activity than the full-length BRINP2 parent protein, validating the cleaved-fragment hypothesis at the mechanistic level. The peptide was subsequently detected in human cerebrospinal fluid and mouse brain tissue, confirming endogenous production rather than synthetic-only existence.
The mechanism. POMC (proopiomelanocortin) neurons in the arcuate nucleus of the hypothalamus are the central node for satiety signaling. They integrate metabolic information from leptin, ghrelin, insulin, and other peripheral signals and project widely to brainstem and cortical regions to drive food-intake suppression. BRP activates POMC neurons via a cAMP-PKA-CREB-FOS signaling cascade. The activation is independent of leptin, GLP-1R, and MC4R, the three classical satiety-signaling axes. This independence is what makes BRP a non-incretin candidate: it does not work through the same receptor families that GLP-1 receptor agonists, leptin analogues, or melanocortin receptor agonists engage. BRP appears to engage a previously uncharacterized receptor or pathway that the Svensson group has not yet identified.
The data. In minipigs, intraperitoneal injection of BRP reduced food intake by approximately 50% within one hour, with the effect persisting for several hours. In obese mice, 14 days of BRP dosing produced 3-4 grams of fat-mass reduction, while control mice on the same diet gained 3 grams. Adverse-effect screening showed no nausea, no GI distress, no anxiety changes, and crucially no muscle-mass loss. The clean preclinical profile is the differentiator against the GLP-1 class, where GI tolerability burden drives meaningful patient discontinuation and lean-mass loss has become a chronic concern (the news section has covered both extensively in recent obesity-drug coverage).
The translation hurdle. BRP's pharmacokinetic profile is a substantial barrier to clinical development. Peak blood concentration occurs within one minute of intraperitoneal injection in animals, with a very short half-life. Any practical clinical formulation will require lipidation, PEGylation, albumin-binder fusion, or Fc-fusion to extend exposure to clinically useful intervals. Each of these strategies has been deployed successfully in other peptide therapeutics (GLP-1 lipidation in semaglutide ↗, GLP-1 Fc-fusion in dulaglutide ↗), so the engineering problem is tractable, but the optimized derivative may have different pharmacology and a different safety profile than the native peptide. Mouse and minipig data also have a long history of failing to translate to human obesity outcomes. The mechanism-class novelty cuts both ways: no precedent class also means no human safety class precedent.
The discovery method matters more than this molecule. The Peptide Predictor approach scanned more than 2,600 predicted prohormone cleavage products across the human peptidome to identify candidates with bioactivity. BRP is the lead validation of that approach. The same pipeline can in principle generate additional bioactive peptide leads from other underannotated protein precursors, expanding the addressable peptide drug-discovery substrate well beyond the dozens of currently characterized hormone-derived peptides. As the section's coverage of the TransCODE peptideins paper ↗ on May 8 showed, the human peptidome's drug-target landscape is meaningfully larger than annotation has historically captured. BRP and the Peptide Predictor approach together demonstrate the discovery substrate can be mined productively.
The commercial context. Coassolo and Svensson are inventors on BRP-related patents (WO2024030214). Svensson is co-founder of Merrifield Therapeutics, which is commercializing BRP for metabolic disorders. The company is in early stages and has not disclosed clinical-development timelines. Funding for the underlying research came from NIH (R01DK125260, P30DK116074, K99AR081618, GM113854), the American Heart Association, Stanford Bio-X, the Wu Tsai Human Performance Alliance, and the Carlsberg Foundation. The Stanford-led collaboration included UC Berkeley, the University of Minnesota, and the University of British Columbia. Coverage in Nature Reviews Drug Discovery ↗ and Stanford Medicine ↗ followed the original publication.
The platform reference. A peptidemodel card for BRINP2-Related Peptide (THRILRRLFNLC, 12 aa) is not yet live; the platform's existing entries under the "BRP" abbreviation refer to a different molecule (a 17-aa BDNF-mimetic). A dedicated card for the Svensson peptide is in the works and will host the sequence, mechanism summary, and primary-literature links for researchers wanting to model derivatives or follow downstream work.
What this is not. A clinical product. BRP has not entered human trials. The PK challenges and the standard preclinical-to-clinical translation gap apply with full force. The "Ozempic ↗ rival" framing in the popular press treatment of the 2025 Nature paper overstated the proximity to clinical impact. What it is is the strongest mechanistically-novel non-incretin obesity-peptide candidate published in years, with a clean preclinical safety profile and a discovery method that may itself be more important than the lead molecule.
What to watch. Whether Merrifield Therapeutics advances a lipidated or fusion-format BRP derivative into IND-enabling work. Whether subsequent papers identify the BRP receptor (currently unknown). Whether the Peptide Predictor pipeline generates additional bioactive lead peptides from other underannotated parent proteins. Each is the kind of follow-up that would convert BRP from a mechanism-of-the-month curiosity into a class-defining new branch of obesity-drug pharmacology.