Tirzepatide, retatrutide and rational multi-target design — what happens when a peptide binds three receptors at once

The incretin idea, a second time

The incretin hypothesis — that intestinal hormones stimulate postprandial insulin secretion — was set up in the 1960s and 70s. Two central incretins were identified: GIP (glucose-dependent insulinotropic polypeptide, earlier 'gastric inhibitory polypeptide', first characterised in 1971 by John Brown in Vancouver) and GLP-1 (glucagon-like peptide-1, identified in the 1980s from the proglucagon gene). Both are released from enteroendocrine cells of the small intestine, both act glucose-dependently on the pancreatic β-cell, both have very short plasma half-lives.

The GLP-1 line became — as described in the separate articles on exenatide/Gila monster and the Wegovy pivot — a multi-billion class. GIP, in contrast, remained pharmacologically unattended for a long time. There was an open question: in type-2 diabetes the GIP response of the β-cell is blunted — does that mean GIP activation is therapeutically of little use, or does it mean GIP receptor sensitivity would need to be pharmacologically restored? This question remained unresolved for two decades.

Tschöp, DiMarchi and the dual-agonist hypothesis

Matthias Tschöp (today Helmholtz Munich) and Richard DiMarchi (Indiana University) developed from the mid-2000s the hypothesis that simultaneous activation of GLP-1 and GIP receptors could yield synergistic metabolic effects — even if isolated GIP activation shows little in type-2 diabetes. The logic: the pancreatic system is designed for combined GLP-1/GIP stimulation; if one molecule hits both receptors in a physiological ratio, the β-cell response could be more comprehensive than with pure GLP-1 activation.

The medicinal chemistry was demanding. A peptide had to be designed to bind two structurally different receptors with balanced affinity — and remain pharmacokinetically suitable for weekly administration. Eli Lilly took up the concept and developed a series of candidate peptides through the 2010s. The end product was tirzepatide (LY3298176), a 39-amino-acid peptide with GLP-1 receptor affinity similar to semaglutide and approximately five-fold higher GIP receptor affinity — with a C20 fatty acid modification enabling albumin binding and thereby the weekly half-life.

The clinical data — and the open mechanism question

The SURPASS series (type-2 diabetes) and SURMOUNT series (obesity) delivered phase 3 data in 2021-2022. SURMOUNT-1 (2022, NEJM, Jastreboff et al.) showed in 2,539 adults with obesity over 72 weeks an average weight loss of 22.5% on tirzepatide 15 mg vs. 2.4% on placebo. That was a different order of magnitude than semaglutide monotherapy (STEP-1: 14.9%). In 2022 the FDA approved Mounjaro (diabetes), in 2023 Zepbound (obesity).

An important methodological observation: the mechanistic explanation of this effect increase remains debated. One line of the literature argues that GIP receptor activation amplifies the central hunger effect and contextually moderates GLP-1-induced nausea. A second line argues that the additional effect may be explained by the specific affinity ratio or structural stability of the tirzepatide molecule rather than pure dual receptor binding. The SURPASS-CVOT trial (cardiovascular endpoints) and mechanistic studies are still running.

Retatrutide: three receptors in one molecule

Eli Lilly developed in parallel a triple-agonist peptide: retatrutide (LY3437943) simultaneously activates the GLP-1, GIP and glucagon receptors. Glucagon activation is conceptually counter-intuitive — glucagon raises hepatic glucose production, which is undesirable in diabetes. But glucagon receptor activation also raises energy expenditure (thermogenesis), reduces hepatic lipid accumulation and can promote weight loss. The design goal is balanced activation using the positive metabolic effects (thermogenesis, fat burning) without triggering the negative (hyperglycaemia).

Phase 2 data on retatrutide (Jastreboff et al., NEJM 2023) showed weight losses of 24.2% at the highest dose (12 mg) over 48 weeks in obesity — again a step up from tirzepatide. Phase 3 trials have been running since 2023; approval could follow 2026-2027.

What this line methodologically shows

Three structural observations. First: rational multi-target design is possible but non-trivial. It requires deep understanding of individual receptor pharmacology, balanced affinity ratios and half-life tuning — and from the start it is a clinical risk because unexpected interaction effects only become visible in human studies. Second: each additional receptor axis empirically increases effect, but not linearly. The second generation (tirzepatide) is ~50% stronger than the first (semaglutide); the third (retatrutide) is only about 10% stronger than tirzepatide. Marginal returns decline.

Third: multi-target design competes pharmacologically with the multi-substance combination line (e.g. CagriSema). Which of the two strategies wins long-term is open. Regulatorily, a single molecule is simpler (one approval, one PK, one safety database); pharmacologically, a combination is more flexible (dose ratios can be individually adjusted, components can become generic-eligible). The next 5-10 years will show which path prevails.

Open questions

• What is the mechanistic principal contribution of the GIP component to tirzepatide — and how would patients benefit if a selective GIP antagonist were available?
• Do the SURPASS-CVOT and retatrutide-CVOT data confirm cardiovascular benefits or show unexpected signals?
• Where are the marginal returns of multi-target design? If the fourth receptor add-on only yields 5% effect increase, is it still worth it?
• Will the multi-target or the multi-substance line (CagriSema) prevail long-term pharmacologically and commercially?