Tirzepatide and retatrutide are both 39-amino-acid lipidated peptides from Eli Lilly and Company, built on a GIP-derived backbone. The core difference is receptor coverage. Tirzepatide activates two receptors (GIPR and GLP-1R), while retatrutide activates those two plus the glucagon receptor (GCGR). Tirzepatide is FDA-approved as Mounjaro and Zepbound; retatrutide is investigational and not approved anywhere.
The two sequences are identical at 30 of 39 positions, so the comparison is a useful case study in how a handful of substitutions can add a third receptor to a peptide's profile. This article sets out the differences in structure, receptor pharmacology, pharmacokinetic data and regulatory status, then covers what each means for in-vitro work. It does not compare the two for use in people. In the Anhydrolabs catalog, tirzepatide is listed as GLP-2 TRZ and retatrutide as GLP-3 RT.
Side-by-side summary
| Property | Tirzepatide | Retatrutide |
|---|---|---|
| Developer code | LY3298176 | LY3437943 |
| Receptors activated | GIPR, GLP-1R | GIPR, GLP-1R, GCGR |
| Main chain length | 39 amino acids | 39 amino acids |
| Backbone origin | GIP sequence | GIP sequence |
| Non-coded residues | Aib at 2 and 13 | Aib at 2, α-methyl-L-leucine at 13, a further non-coded residue at 20 |
| Lipidated residue | Lys20 | Lys17 |
| Side chain | C20 fatty diacid (1,20-eicosanedioic acid) through a linker | Fatty diacid through a linker |
| C-terminus | Amide | Not stated in the sources below |
| Molecular formula | C225H348N48O68 | C221H342N46O68 |
| Molecular weight | 4813.53 Da (FDA label) | 4731.34 g/mol (KEGG DRUG) |
| CAS number | 2023788-19-2 | 2381089-83-2 |
| Reported half-life in humans | About 5 to 6 days (FDA label) | About 6 days (Phase 1b) |
| Status | FDA-approved (Mounjaro, 2022; Zepbound, 2023) | Investigational, in Phase 3 |
What "dual" and "triple" agonism mean
GLP-1, GIP and glucagon are peptide hormones of the same family, and their receptors are class B G protein-coupled receptors that signal mainly through Gs and cyclic AMP. The hormones share an N-terminal region that enters the receptor's transmembrane core and a helical region that rests on the receptor's extracellular domain. Because the sequences are related, a designed peptide can hold features that each receptor recognizes.
- GLP-1 receptor (GLP1R, UniProt P43220). The receptor for GLP-1, a proglucagon product released from intestinal enteroendocrine cells.
- GIP receptor (GIPR, UniProt P48546). The receptor for glucose-dependent insulinotropic polypeptide, released from intestinal K cells.
- Glucagon receptor (GCGR, UniProt P47871). The receptor for glucagon, the pancreatic alpha-cell hormone.
Tirzepatide is a unimolecular dual agonist at GIPR and GLP-1R. Retatrutide adds GCGR agonism to the same scaffold. Semaglutide, for reference, is a selective GLP-1R agonist built on the GLP-1 sequence rather than on GIP. For background on the class, see our overview of what retatrutide is.
A note on catalog names: the "GLP-2" and "GLP-3" in the catalog codes count receptor targets, following common shorthand. They do not refer to glucagon-like peptide-2, which is a separate proglucagon-derived hormone with its own receptor. Neither compound acts on the GLP-2 receptor.
Sequence comparison
Here are the one-letter sequences as listed in KEGG DRUG, with X marking non-coded residues:
| Sequence (positions 1 to 39) | |
|---|---|
| Tirzepatide | YXEGTFTSDYSIXLDKIAQKAFVQWLIAGGPSSGAPPPS |
| Retatrutide | YXQGTFTSDYSILLDKKAQXAFIEYLLEGGPSSGAPPPS |
The two differ at nine positions:
| Position | Tirzepatide | Retatrutide |
|---|---|---|
| 3 | Glu | Gln |
| 13 | Aib | α-methyl-L-leucine |
| 17 | Ile | Lys (lipidated) |
| 20 | Lys (lipidated) | Non-coded residue |
| 23 | Val | Ile |
| 24 | Gln | Glu |
| 25 | Trp | Tyr |
| 27 | Ile | Leu |
| 28 | Ala | Glu |
A few points stand out:
- Position 3. Native glucagon carries Gln at position 3, and native GIP and GLP-1 carry Glu. Retatrutide's Gln3 matches glucagon, which is consistent with a design aimed at GCGR activity.
- The lipidation site moved. Tirzepatide attaches its fatty diacid at Lys20, while retatrutide attaches it at Lys17. The cryo-EM work on retatrutide describes a salt bridge from Lys17 to GLP-1R that is absent in GIPR, so this residue does more than carry the lipid.
- Shared C-terminal tail. Both end in GGPSSGAPPPS, the same eleven residues that end exenatide, the GLP-1R agonist originally identified in the Gila monster, Heloderma suspectum. This segment is not present in native GIP, which is 42 residues long.
- No tryptophan in retatrutide. Trp25 in tirzepatide becomes Tyr25 in retatrutide. For analysts, this changes the UV absorbance profile at 280 nm, so extinction coefficients are not interchangeable between the two.
Receptor pharmacology in numbers
The discovery papers for the two compounds report data from assays run on different occasions, so treat cross-paper comparisons as approximate.
Tirzepatide. In the 2018 Molecular Metabolism discovery paper, tirzepatide bound GIPR with a Ki of 0.135 nM and GLP-1R with a Ki of 4.23 nM. Its GIPR affinity was comparable to native GIP, while its GLP-1R affinity was about 5-fold weaker than native GLP-1. In cAMP assays, the EC50 values were 0.0224 nM at GIPR and 0.934 nM at GLP-1R: similar to GIP at its receptor and about 13-fold weaker than GLP-1 at GLP-1R. Read together, the numbers show potent activity at both receptors, with relatively more at GIPR.
Retatrutide. The 2024 Cell Discovery structural paper summarizes retatrutide as about 8.9 times more potent than native GIP at GIPR, and about 0.4 and 0.3 times as potent as native GLP-1 and glucagon at GLP-1R and GCGR. The discovery paper describes GCGR and GLP-1R activity as balanced, with more activity at GIPR.
Both molecules therefore lean toward GIPR, and both give up some GLP-1R potency relative to the native hormone. Retatrutide's distinguishing feature is measurable GCGR agonism, which tirzepatide does not have.
Structural biology
Cryo-EM structures have been published for both compounds bound to their receptors with Gs. For retatrutide, the Cell Discovery study reports structures with GLP-1R (2.68 Å), GIPR (3.26 Å) and GCGR (2.84 Å). It compares them with earlier complexes deposited in the Protein Data Bank, including tirzepatide bound to GLP-1R (PDB 7FIM) and to GIPR (PDB 7FIY). The authors note that retatrutide and a glucagon-based triple agonist were built from different parent backbones, GIP and glucagon, yet reach all three receptors by keeping residues conserved across the family.
Half-life and metabolism
Both peptides carry a fatty diacid that binds serum albumin, and this binding is the main reason their half-lives are measured in days.
- Tirzepatide. The Zepbound label gives an elimination half-life of approximately 5 to 6 days and plasma albumin binding of 99%. Metabolism is by proteolytic cleavage of the backbone, beta-oxidation of the C20 fatty diacid, and amide hydrolysis. The 2018 Phase 1 data reported a mean half-life of about 5 days (116.7 hours).
- Retatrutide. The Phase 1b trial in adults with type 2 diabetes reported a half-life of approximately 6 days, with exposure proportional across the range studied.
For in-vitro work, the albumin binding matters more than the half-life. Adding serum or albumin to an assay buffer lowers the apparent potency of both peptides, because a fraction of each is bound. If you compare them, keep albumin conditions identical and report them.
Regulatory status
| Tirzepatide | Retatrutide | |
|---|---|---|
| FDA status | Approved | Not approved |
| Brands | Mounjaro (NDA 215866, approved May 2022); Zepbound (NDA 217806, approved November 2023) | None |
| Labeled indications | Mounjaro: glycemic control in type 2 diabetes, adjunct to diet and exercise. Zepbound: chronic weight management in adults with obesity or overweight with a weight-related condition, and moderate to severe obstructive sleep apnea in adults with obesity | None |
| Development stage | Marketed | Phase 3 (TRIUMPH program) |
A Phase 3 head-to-head trial, TRIUMPH-5 (NCT06662383), is registered to compare retatrutide with tirzepatide in adults with obesity. We report its existence, not its findings.
Choosing between them for in-vitro work
The choice depends on the question, not on which molecule is newer.
- Receptor deconvolution. To separate GCGR-dependent from GCGR-independent effects in a cell system, run the two side by side. The difference between them is largely the glucagon receptor component, though the nine sequence changes also shift GIPR and GLP-1R potency, so a selective GCGR antagonist or a GCGR-knockout line is a useful third arm.
- Reference compound. Tirzepatide has a public FDA label with a stated formula, weight and structure. That makes it an easier reference standard for method development.
- Combined ligands. The catalog also lists a retatrutide and cagrilintide blend for assays that pair a glucagon-family agonist with an amylin analog.
Both compounds are supplied as lyophilized powders in vacuum-sealed vials and as 10-vial kits. Handle them the same way: keep the sealed powder at −20 °C or colder, bring the vial to room temperature before opening, reconstitute in a slightly basic aqueous buffer, use low-binding plastics, and freeze single-use aliquots. Our guide to reconstituting peptides has the concentration arithmetic, and peptide purity explains the HPLC and mass spectrometry figures on a certificate. Both products are for laboratory research only; see research use.