In obese mice, a designed peptide that first switches on the GIP receptor and then blocks it drove a 21.6 percent drop in body weight. A near-identical peptide that only blocked the receptor managed 14.5 percent. The gap held with a strong statistical margin (p less than 0.001), and the authors read it as a case for building the on-then-off behavior into a single molecule.
The work, published July 6 in Diabetes, Obesity and Metabolism ↗, is a mouse study, and it lands in the middle of the most unsettled argument in obesity drug design: at the GIP receptor, should a drug turn it on or shut it off?
The paradox the paper is poking at
GIP is a gut hormone. Its receptor, GIPR ↗, sits next to the GLP-1 receptor as a metabolic target, and the two best-known drugs that engage it do opposite things. Tirzepatide ↗, sold as Mounjaro and Zepbound, is a dual agonist: it activates both GLP-1 and GIP receptors. MariTide ↗, Amgen's monthly injection, does the reverse at GIPR, pairing GLP-1 activation with a GIP receptor blockade. Both make people lose weight. Nobody has cleanly shown which approach is better, or why blocking a receptor and activating it should land in the same place.
The Chinese team built three peptides to test the question head-on, each anchored to the same GLP-1 agonist so the only variable was the GIP arm. One activated GIPR. One blocked it. The third was engineered to start as an activator and transition into a blocker, using a GIP fragment (residues 1-30) designed to lose its agonist grip over time. That sequential molecule won, and both the activating and blocking versions improved blood fats, cutting triglycerides and total cholesterol in the DIO mice ↗.
What to hold onto, and what not to
The honest caveat is the loudest one: these are mice, and the peptides are lab constructs, not clinical candidates. A 21.6 percent body-weight drop in a diet-induced-obese mouse does not convert to a human number, and the on-then-off timing that works over a mouse's short study window may behave differently across a weekly human dose. What the paper does add is a mechanistic argument that GIP receptor activation and inhibition are not simply two roads to the same clinic. Sequencing them beat doing either alone.
This is the same receptor peptidemodel has tracked from the other direction. A June study on blocking GIP ↗ found that shutting the receptor down dimmed the after-meal blood rush to the gut, a reminder that GIPR does more than tune the scale. If a single molecule can run the receptor in two modes on a schedule, the design space for the next wave of dual GLP-1 ↗ and GIP drugs gets wider than the current agonist-versus-antagonist split suggests. For now it is a mouse result, and a sharp one.