The GIP paradox

Here is a fact that should not sit comfortably together. Tirzepatide, the dual agonist behind some of the largest weight-loss numbers in medicine, works partly by activating the GIP receptor. And maridebart cafraglutide, one of the most watched next-generation candidates, works partly by blocking the very same receptor. Agonist and antagonist, the same target — and both produce substantial weight loss.

Under the intuition that a receptor is a volume knob — more signal, more effect — this is a contradiction. One of them should be turning the effect up and the other turning it down. Instead they point the same way on the scale. Resolving that is a small master class in how receptors actually behave, and it quietly dismantles the idea that drug design is just a matter of pushing harder.

GIP — glucose-dependent insulinotropic polypeptide — is the other incretin, released from gut K-cells after a meal. Like GLP-1, it potentiates glucose-dependent insulin release, which is the straightforward part. Its role in body weight and fat, though, has always been the murkier one: older work tied GIP signaling to fat storage, while more recent work points to central actions on appetite and, intriguingly, on the nausea that limits incretin therapy. GIP is not a hormone with one clean job — and that ambiguity is exactly the room the paradox lives in.

In tirzepatide, GIP agonism rides alongside GLP-1 agonism, and the pairing clearly outperforms GLP-1 alone. The leading explanations are that GIP-receptor activation adds a complementary appetite signal through the brain, and that it may blunt the nausea of GLP-1 agonism — letting the GLP-1 arm be pushed further before tolerability caps it. On this view GIP is a genuine second lever, and switching it on is additive.

Maridebart takes the opposite tack — an anti-GIP-receptor antibody carrying GLP-1 peptides, so it agonizes GLP-1 while antagonizing GIP — and it too drives large weight loss. The rationale draws on GIP’s older association with fat storage (block it, and you remove a pro-storage signal) and on a subtler receptor argument we will come to. The antibody scaffold is its own payoff: a long half-life that supports once-monthly dosing, a cadence injectable peptides can’t easily reach.

Three resolutions are on the table, and they aren’t mutually exclusive:

• Sustained agonism becomes functional antagonism. flood a receptor continuously and it desensitizes and internalizes — down-regulating until it barely responds. A long-acting GIP agonist may end up chronically silencing the receptor, arriving at nearly the same downstream state as an outright blocker. Push it hard enough and blocking it start to rhyme
• GIP does different things in different tissues. its actions in the brain, the pancreas, and adipose aren't the same, so agonism and antagonism can each pick out a different, beneficial subset of effects rather than being simple opposites
• It isn't settled yet. the honest answer is that the field does not fully agree on GIP's contribution, and the two strategies are, in part, a live experiment being run at scale to find out

The first is the most striking, because it means agonist and antagonist need not be enemies at all. If continuous activation ends in a quiet receptor, then the two drugs may be taking different roads to the same downstream destination — reduced net GIP signaling — one by exhausting the receptor, the other by capping it directly.

Step back and the GIP puzzle is the cleanest possible refutation of the “bigger number” model of this field. If two molecules can hit the same receptor in opposite directions and both win, then “more agonism” was never the axis of progress. What matters is net signaling in the right tissues over the right timescale — a far more interesting design target than raw potency, and one that rewards cleverness like maridebart’s antibody trick over brute force.

It also reframes GIP itself. Far from a settled “second incretin,” it is an open question the industry is probing from both ends at once — the same spirit in which the triple agonist found a use for glucagon, the hormone that raises blood sugar. The frontier keeps making the same point in new costumes: a hormone’s meaning isn’t fixed by its label. It depends on the dose, the duration, the tissue, and the company it keeps — and the molecules that read that nuance best are the ones redrawing the edge.