The “natural Ozempic” headline that slid into my feed this month carried a date I couldn’t make sense of. ScienceDaily posted it on November 4, 2026, but the Stanford Medicine release it was adapted from went out on March 5, 2025, the same day the paper appeared in Nature. A mouse-and-minipig study more than a year and a half old is circulating as fresh news, right as a gray market in injectable “research peptides” is booming, and the retellings barely mention that the molecule has never been tested in a person.
The pitch is easy to love: a 12-amino-acid peptide called BRP, which the human body already makes, that cuts appetite the way semaglutide does without the queasiness. I wanted to know what it actually did, in which animals, and how far it still sits from anything a doctor could prescribe you. So I went back to the paper.
Many appetite hormones start out as larger precursor proteins that an enzyme, prohormone convertase 1/3, snips at specific sites. GLP-1, the gut hormone semaglutide imitates, is one of those snipped-out pieces. Katrin Svensson’s Stanford lab built software to predict what else that enzyme cuts that nobody had ever studied. According to the paper, the program worked through 2,082 secreted human proteins, flagged 373 likely precursors and predicted 2,683 peptides, and the team made 100 of them to test on cells. One fragment, cut from a protein called BRINP2, pushed the activity marker Fos up more than tenfold in a neuron-like cell line. GLP-1, tested in pancreatic beta cells, managed about threefold.
In lean mice, a single injection of BRP cut food intake by as much as 50 percent, and in minipigs it halved intake over the first hour, roughly matching liraglutide. Obese mice dosed daily for 14 days, ten per group, ended up about 4 grams lighter than untreated mice, and the authors write that “the reduction in body weight was entirely owing to loss of fat mass,” with lean mass unchanged. Liraglutide, for the record, took weight off faster at the start.
Something else is listening
I expected BRP to be one more way of pressing the GLP-1 button, and that’s where I stopped skimming. When the team blocked the GLP-1 receptor with the drug exendin(9–39), GLP-1’s effect on appetite vanished and BRP’s held. BRP still cut eating and weight in mice that make no leptin, and in mice missing the melanocortin-4 receptor, the appetite switch behind a good share of inherited obesity. Those are the routes appetite researchers check first, and BRP doesn’t need any of them.
Wait, if it isn’t using any of the usual doors, which neurons is it even talking to? The team mapped which brain cells switched on after an injection. BRP activated neurons in the dorsomedial, preoptic, tuberal and arcuate regions of the hypothalamus and produced no activation in the hindbrain or brainstem. In the arcuate nucleus, the cells that lit up weren’t the POMC neurons that dominate the textbook diagrams. Something else is listening. The receptor hasn’t been found, and the authors write that identifying it “will be essential” to understanding how BRP works.
The molecule isn’t just a lab construct, either. The team measured BRP in human cerebrospinal fluid at roughly 700 picomolar to 3 nanomolar, so it’s already circulating in the fluid around your brain. The paper doesn’t say what it does there on an ordinary day.
The no-nausea claim is what sold the headline, and the animal evidence behind it is unusually thorough. BRP didn’t give mice a conditioned taste aversion, the learned avoidance animals develop after something makes them sick, while lithium chloride did. In rats it didn’t trigger pica, the clay-eating that cisplatin reliably caused in the same experiment. Gastric emptying and body temperature didn’t budge, and open-field tests showed no sign of anxiety. For an appetite suppressor, that’s an impressive run of null results.
Now set that beside what people on semaglutide go through. In STEP 1, the 1,961-person, 68-week trial funded by Novo Nordisk, nausea and diarrhea were the most common adverse events, and 4.5 percent of people on the drug stopped treatment because of gastrointestinal effects, against 0.8 percent on placebo. A body-composition substudy of that trial found fat mass fell 19.3 percent, but lean mass, which counts organs and water along with muscle, fell 9.7 percent as well. A better drug would have to fix both of those problems.
| Group | Value (percent) |
|---|---|
| Semaglutide | 4.5 |
| Placebo | 0.8 |
The “natural Ozempic” label still overreaches. BRP was never tested against Ozempic: the comparators were liraglutide, an older daily GLP-1 drug, and native GLP-1. The semaglutide line comes from the press release, where Svensson said, “Nothing we’ve tested before has compared to semaglutide’s ability to decrease appetite and body weight.” She was describing her lab’s earlier candidates, and no head-to-head experiment backs the comparison.
The pharmacology shows how far this is from a medicine. In the 14-day obesity study the mice got 5 milligrams of BRP per kilogram against 100 micrograms per kilogram of liraglutide, fifty times as much by weight. BRP peaked in the blood about a minute after injection and had a plasma half-life of under 10 minutes, and brain levels ran about a tenth of blood levels. A pair of adjacent arginines in the peptide looks like a cleavage site, and the fragments the team found in plasma were inactive. The lean-mass result also comes from two weeks in mice, while semaglutide’s lean-mass loss was measured over 68 weeks in people.
Semaglutide has been down this road. It began as a natural gut hormone that chemists modified until a single shot lasts a week. The release says the Stanford researchers are “investigating how to help the peptide’s effects last longer in the body.” If that work succeeds, whatever reaches a clinic may well be an engineered cousin of BRP, which would make it about as natural as Ozempic.
Who owns the next step
The paper’s competing-interests statement runs two sentences. Svensson and lead author Laetitia Coassolo are named inventors on patents covering BRP peptides for metabolic disorders. Svensson is also a cofounder of Merrifield Therapeutics, and the release says she started a company to run human trials, quoting her: “We are very eager to learn if it is safe and effective in humans.” When I searched ClinicalTrials.gov on November 10, 2026, it listed no study mentioning BRINP2 and none with Merrifield as a sponsor.
Then look at who paid for the hunt. The acknowledgements list four NIH grants, American Heart Association awards, Stanford’s SPARK translational program, Stanford Bio-X and a string of foundation fellowships. That’s the standard arrangement in academic drug discovery, where public and philanthropic money funds the search and the commercial upside travels through patents and startups. Keep it in mind the next time an obesity molecule gets sold to you as a gift from nature.
The risk closest to home comes in a vial. An August 2026 Stanford Medicine piece on the peptide boom notes that BRP hasn’t been tested in humans, and reports that molecules Svensson discovered and tested “so far, only in mice” have shown up on internet-based company websites. The piece doesn’t say BRP is among them. Svensson’s own warning there is blunt: “I work with these experimental peptides myself. I know how potent some of them can be.” A recycled “natural Ozempic” headline is free marketing for those sellers, and the paper gives a buyer neither a human dose nor any human safety data.
I find BRP exciting, a signal already present in human spinal fluid that works through wiring nobody had mapped, with a receptor still waiting to be named. I’m not putting anything labeled BRP into my body until a registered human trial reports a dose people can tolerate.
Sources
- Coassolo et al., Nature (2025), full text via PMC
- Stanford Medicine news release, March 5, 2025 (EurekAlert)
- Wilding et al., STEP 1, New England Journal of Medicine (2021)
- STEP 1 body-composition analysis, Journal of the Endocrine Society (2021)
- Stanford Medicine Insights, Peptides: What does the science say? (August 10, 2026)
- ScienceDaily, November 4, 2026