GIP — the incretin that wasn't needed for 50 years
Glucose-dependent insulinotropic polypeptide (GIP) was isolated in 1971 by John Brown in Vancouver — before GLP-1, before exendin-4, before all other incretins. Nevertheless it remained pharmacologically unused for decades. Only in 2022, with tirzepatide, did GIP return to the clinic as the second half of a multi-target design. A story about the hormone that was not the hero of its own discovery.
A substance first known by its wrong name
John Brown isolated in 1971 in Vancouver from pig intestine a polypeptide factor that inhibited histamine- or pentagastrin-stimulated gastric acid secretion. He called the peptide 'gastric inhibitory polypeptide' — hence the acronym GIP. It was a 42-amino-acid peptide from the K-cells of the proximal small intestine. Only years later was it clear: the gastric inhibition was an artefact of an unphysiologically high-dose bioassay. The actual physiological function of GIP is glucose-dependent stimulation of the pancreatic β-cell to insulin secretion. The name was reinterpreted from 'gastric inhibitory' to 'glucose-dependent insulinotropic' polypeptide — the acronym GIP stayed.
Thus GIP was the first identified incretin. GLP-1 was derived and isolated from the proglucagon gene only in the mid-1980s. Had the story gone fairly, GIP would have dominated incretin-based diabetes therapies. It did not.
Why GIP lagged pharmacologically
Three reasons are methodologically central. First: in type-2 diabetes the insulinotropic response of the β-cell to GIP is strongly blunted, while the response to GLP-1 is largely preserved. This observation was documented in the 1990s in classical hyperglycaemic clamp studies (Nauck and Holst). From a pharmacological view it suggested: GIP is therapeutically of little use because the target system does not respond in the indication population. GLP-1 analogs (exenatide from 2005) became the standard incretin therapy; GIP was skipped.
Second: GIP has shown pro-adipogenic effects in some animal models. GIP receptor knockout mice are more resistant to high-fat diet-induced obesity. This observation seemed to suggest that GIP stimulation in obesity patients would be counterproductive — which also spoke against a pharmacological GIP programme. Third: GIP-specific ligands are chemically harder to stabilise than GLP-1 ligands, and the patent landscape was considerably more attractive for GLP-1.
The rehabilitation: what tirzepatide changed
Matthias Tschöp and Richard DiMarchi (see separate tirzepatide multi-target article) formulated from the late 2000s an alternative hypothesis: perhaps the GIP response of the β-cell in diabetes is not intrinsically defective but secondarily suppressed by chronic hyperglycaemia and hyperinsulinaemia. With concurrent GLP-1 activation — which normalises glucose — the GIP receptor sensitivity could be pharmacologically restored. This hypothesis had been untestable for 20 years because no suitable tools existed.
Tirzepatide (LY3298176, FDA approval Mounjaro 2022 for diabetes, Zepbound 2023 for obesity) was the first clinical tool addressing both receptors simultaneously. The SURPASS and SURMOUNT data showed effect increases over GLP-1 monotherapy larger than either component alone would have predicted. The mechanistic explanation is debated in the literature — additive central satiety, GIP modulation of gastrointestinal side effects, GIP-specific lipid-metabolism effects — but the clinical superiority is robustly documented.
„GIP was never gone. It was on the bench for 50 years because we didn't know how to deploy it. Tirzepatide brought it back — as the second half of a team, not as playmaker."
What the GIP story methodologically shows
Three structural lessons from this line. First: pharmacological 'failure' of a substance can be methodological or temporal, not biological. GIP was therapeutically unattractive for 50 years because the tools were missing to deploy it in the right context. A substance classified today as 'non-functional' can become relevant tomorrow with new tools. Second: monogamous therapeutic thinking (one substance, one receptor) often misses the combinatorial reality of endocrinology. Hormones act in networks; testing a hormone that is not active in combination with its natural partner can underestimate its clinical potential.
Third: the GIP pro-adipogenesis knockout data from the 2000s were not wrong but context-dependent. In a setting of acute chronic stimulation of a specific receptor type, pharmacological effects can differ from physiological function. This context-dependency is a property of endocrinological pharmacology repeatedly underestimated in animal-to-human translation.
Open questions
- Are selective GIP antagonists — which produce weight loss in some preclinical models — a complementary therapeutic line to GIP agonists?
- Which other 'historically discarded' incretins (e.g. glicentin, oxyntomodulin, also from the proglucagon gene) deserve rehabilitation?
- How does the GIP response of the β-cell quantitatively change under tirzepatide therapy over time — is sensitivity actually restored?
- What lessons from the GIP rehabilitation story transfer to other 'second-class' hormones (e.g. the forgotten pancreatic polypeptide)?