For three years the obesity-drug industry has been running a strange experiment in public. Eli Lilly’s tirzepatide — sold as Mounjaro and Zepbound — activates a receptor called GIPR. Amgen’s MariTide blocks the same receptor. Both pair that with GLP-1 activation, and both produce weight loss. Two companies pushing the same switch in opposite directions, both getting results, and nobody able to say why.
A team at the Institute of Metabolic Science at the University of Cambridge now says the answer is location. In a study published in Nature Metabolism, first-authored by Dr Jo Lewis with Fiona Gribble and Frank Reimann as senior authors, the researchers report that GIPR agonists and GIPR antagonists work through entirely different regions of the brain.
Turning the receptor on works through the brainstem, where it suppresses appetite directly. Turning it off works through the hypothalamus, where GIPR appears to function as a brake — limiting how strongly the brainstem responds to signals that the body is full. Block it, and the brake comes off. Same target, two addresses, two mechanisms, one outcome.
How they pinned it down
The design was a knockout study in mice. The team built animals missing GIPR in the brainstem, another group missing it in the hypothalamus, and left a third group unmodified as controls. Then they dosed all three with combinations of a GIPR agonist, a GIPR antagonist and a GLP-1 drug, and tracked food intake, body weight, fat mass, blood-sugar control and brain activity.
If a drug stops working when you delete the receptor from one region, that region is where the drug was acting. Agonists lost their effect without brainstem GIPR. Antagonists lost theirs without hypothalamic GIPR. The logic is clean and the answer is unambiguous, which is rarer in metabolic research than the press releases suggest.
The team also reported that blocking GIPR appeared to strengthen the effect of drugs targeting the amylin receptor — another class currently moving through clinical development. The work was funded by the UK Medical Research Council and Wellcome.
Our take: This is a mouse study, and mouse metabolism has embarrassed this field before. Nobody should change a prescription, a diet or an opinion about their own body on the strength of it. What it changes is the industry’s map. Until now, the agonist-versus-antagonist split looked like two companies making incompatible bets on the same biology, and one of them was presumably wrong. If the Cambridge result holds in humans, neither is wrong — they are hitting different circuits, which is exactly the precondition for stacking them. That reframes the next generation of obesity drugs from a winner-take-all race into a combination problem.
Why the mechanism matters commercially
MariTide — maridebart cafraglutide — is in phase 3, and its phase 2 data came with a well-documented tolerability problem: high discontinuation and vomiting rates that forced Amgen to revise its phase 3 dosing plan. Side effects in this drug class are substantially a brainstem story; the brainstem is also where nausea lives. A result that says the antagonist works through the hypothalamus instead is, at minimum, an interesting thing for a medicinal chemist to know.
As Lewis put it, understanding which brain circuits respond to these medications could help design drugs that “produce more weight loss with fewer side effects.” The broader point the Cambridge group is making is that these are not gut drugs that happen to reach the brain. They are brain drugs, acting on specific, now-identifiable circuits.
What to watch
- Human replication. Regional brain mechanism in mice is a hypothesis about humans, not a finding in them. Imaging and receptor-occupancy work is the next step, and it is slow.
- MariTide phase 3 tolerability. Amgen’s revised dosing schedule is the real-world test of whether the antagonist route genuinely carries a different side-effect profile.
- Combination trials. If GIPR antagonism boosts both GLP-1 and amylin drugs through a separate circuit, expect the number of combination programmes to climb.
- The nausea question. Watch whether anyone can decouple appetite suppression from the brainstem side effects. That, not another percentage point of weight loss, is what would widen this market.
The honest summary: a genuine mechanistic advance, in mice, with real implications for how the next round of drugs gets designed and none whatsoever for what you should do this week.