Incretin Receptor Pharmacology: GLP-1, GIP, and Glucagon Receptors in Laboratory Research

Two of the research compounds in this catalog — GLP-2 TZ (tirzepatide) and GLP-3 RT (retatrutide) — are described in the literature as agonists at more than one receptor. Their compound pages state which receptors those are. This page explains the receptor system itself: what these receptors are, how a cell responds when one is activated, how a laboratory measures that response, and what it means to design a single peptide that engages two or three of them at once.

Understanding this is useful for a specific and narrow reason. Receptor-level potency figures are the most commonly quoted numbers about this class of peptide, and they are also the most commonly misread. Knowing how an EC₅₀ is generated makes it obvious why two papers can report different values for the same molecule without either being wrong — and why such a figure describes an assay, not an organism.

What “incretin” denotes in the literature

“Incretin” is a classification, not a single molecule. It refers to gut-derived peptide hormones released in response to nutrient intake that then act on receptors elsewhere in the body. Two are conventionally described as the canonical incretins: glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP). Both are endogenous signalling molecules; both were characterised as part of basic endocrine physiology long before any synthetic analog existed.

A point of genealogy that explains some of the receptor cross-talk discussed below: GLP-1 and glucagon are both produced from the same precursor protein, proglucagon, by tissue-specific processing. They are therefore structurally related peptides, which is a large part of why a single engineered sequence can be made to engage both of their receptors. GIP derives from a separate gene and precursor, but its receptor belongs to the same structural family.

Describing endogenous hormone biology is a statement about human and animal physiology as characterised in the scientific literature. It is not a statement about what any research compound does, and nothing in this section should be read that way.

Class B GPCRs and how they signal

The GLP-1, GIP, and glucagon receptors are all class B1 G-protein-coupled receptors, sometimes called the secretin-receptor family. They share an architecture that matters for how agonists work on them: a seven-transmembrane helical bundle typical of all GPCRs, plus a comparatively large extracellular domain that the C-terminal portion of the peptide ligand binds first. The N-terminal portion of the ligand then engages the transmembrane core, and that second interaction is what actually triggers activation.

This two-step binding model is the reason the N-terminus of these peptides is so sensitive to modification: changes at the very first residues can preserve binding while altering activation, or preserve both while altering stability. It is a recurring theme in the design choices described later on this page.

Once activated, all three receptors couple principally to Gα_s, which stimulates adenylyl cyclase and raises intracellular cyclic AMP (cAMP). cAMP in turn activates protein kinase A and other effectors. Downstream of that primary route, these receptors also engage β-arrestin recruitment and receptor internalisation, and can show secondary coupling to other G proteins depending on the cell context.

The practical consequence for a laboratory is that cAMP accumulation is the standard readout for agonism at these receptors. When a paper reports a potency value for a peptide at the GLP-1 receptor, it is almost always a cAMP measurement in cells engineered to express that receptor.

The three receptors, individually

• GLP-1 receptor (GLP1R). A class B1 GPCR whose endogenous ligand is GLP-1. Widely studied in cellular models of incretin signalling, and the most extensively characterised of the three from a structural standpoint — a number of active-state structures have been solved, which is why GLP-1R is often the reference receptor in comparative assay work.
• GIP receptor (GIPR). The receptor for GIP, and the second canonical incretin receptor. Investigated in the context of nutrient-sensing signalling. It is worth noting that the pharmacology literature on GIPR is genuinely unsettled at a mechanistic level — both agonism and antagonism at this receptor have been explored as design strategies in different research programmes, which is unusual and means GIPR results should be read with attention to what was actually being tested.
• Glucagon receptor (GCGR). The receptor for glucagon. Studied in hepatic and adipose-tissue models in relation to energy-substrate signalling. Because glucagon and GLP-1 share the proglucagon precursor, GCGR and GLP1R have appreciable ligand cross-reactivity — a fact that is exploited deliberately in multi-agonist design and that also complicates selectivity claims.

A fourth receptor is worth naming only to rule it out, because the naming convention causes real confusion. Glucagon-like peptide-2 (GLP-2) is a distinct endogenous peptide with its own receptor, GLP2R, and its own separate literature. It is not part of the incretin pair above and is not engaged by the compounds discussed here. The catalog designation “GLP-2 TZ” refers to tirzepatide, a GIP/GLP-1 agonist, and has nothing to do with glucagon-like peptide-2 — a point made on that compound's own page as well. Similarly, there is no “GLP-3” receptor in this literature; “GLP-3 RT” is a catalog label for retatrutide.

How agonism is measured in vitro

A receptor potency figure is the output of a defined experiment, and knowing the shape of that experiment is what makes the number interpretable.

The usual system is a receptor-transfected cell line — a cultured line such as CHO or HEK293 engineered to express the receptor of interest, often in large amounts, so that the response is attributable to that receptor rather than to whatever the cell natively carries. The peptide is applied across a range of concentrations, typically spanning several orders of magnitude, and cAMP accumulation is quantified at each concentration.

Plotting response against log concentration gives a sigmoidal concentration–response curve, from which two distinct parameters are extracted:

• EC₅₀ — the concentration producing half of the maximal response in that assay. This is a measure of potency: how little of the compound is needed. Lower means more potent. For engineered peptides of this class, published values commonly fall in the sub-nanomolar to low-nanomolar range.
• E_max — the maximal response achieved, usually expressed relative to the endogenous ligand. This is a measure of efficacy: how much response the compound can produce at all. A compound can be highly potent yet only partially efficacious, which is a meaningfully different pharmacological profile from a full agonist.

Quoting EC₅₀ without E_max is therefore an incomplete description, and it is a common omission in secondary sources. Binding affinity, often reported as K_i or K_d from a competition or direct-binding assay, is a third and separate quantity: it describes how tightly the ligand occupies the receptor, not what the receptor does afterwards. A ligand can bind tightly and activate weakly.

Mono-, dual-, and triple-agonist design

Classical pharmacology prizes selectivity: one molecule, one target, minimal off-target activity. A significant strand of peptide engineering research inverts that goal deliberately, designing a single sequence intended to engage several related receptors at once. This is sometimes described as intentional polypharmacology, and the incretin receptor family is where it has been pursued most visibly.

The structural precondition is the family relationship described earlier. Because GLP-1, GIP, and glucagon are homologous peptides binding structurally homologous class B1 receptors, it is chemically feasible to construct a hybrid sequence — a backbone drawn from one, with substitutions borrowed from another — that retains measurable activity at more than one receptor. That is not possible for arbitrary receptor pairs; it works here because of shared ancestry.

The resulting descriptive vocabulary is straightforward once the measurement basis is clear:

• Mono-agonist — meaningful activity at one of these receptors.
• Dual agonist — measurable agonism at two. Tirzepatide is characterised this way, at GIPR and GLP1R.
• Triple agonist — measurable agonism at three. Retatrutide is characterised this way, adding GCGR.

Two cautions follow, and both are frequently ignored. First, “dual” and “triple” are statements about which receptors show a response in an assay — they say nothing about the relative potency at each. A peptide can be described as a triple agonist while being orders of magnitude more potent at one receptor than another, and that balance or imbalance across receptors is itself a design variable reported separately. Second, more receptors engaged is not a scale of quality or strength. A triple agonist is not a “stronger” molecule than a dual agonist; it is a molecule with a different receptor coverage profile. Any inference beyond that — particularly any inference about effects in an organism — is unsupported by receptor-level data.

For contrast, it is worth noting that not every multi-component research tool works this way. The CJC-1295 / Ipamorelin blend engages two receptors as well, but does so with two separate peptides acting at two unrelated receptors — a GHRH receptor and the ghrelin/GHS-R1a receptor — rather than one peptide with multi-receptor activity. A blend and a multi-agonist are different concepts that happen to produce a superficially similar description.

Structural strategies behind these peptides

Three recurring modifications appear in engineered peptides of this class. Each is a chemical answer to a specific problem, and each shows up in the identity string on a certificate of analysis.

Aib substitution for enzymatic resistance. The endogenous incretins are rapidly cleaved by the enzyme dipeptidyl peptidase-4 (DPP-4), which removes the first two residues from the N-terminus. Since class B1 receptor activation depends on that N-terminal region, cleavage there abolishes activity. Substituting a non-coded residue at the vulnerable position defeats the enzyme: α-aminoisobutyric acid (Aib) is alanine bearing a second methyl group on the α-carbon, which makes that carbon quaternary and removes the backbone geometry DPP-4 recognises. This is precisely why Aib appears at position 2 in engineered sequences of this family — tirzepatide carries Aib at positions 2 and 13. Aib also stabilises helical structure, which can independently improve receptor engagement.

Fatty-acid conjugation for albumin association. Attaching a lipid — commonly a C20 fatty di-acid, conjugated to a specific lysine side chain — lets the peptide associate reversibly with serum albumin. In preclinical protraction studies this markedly slows clearance. For laboratory handling the relevant consequence is different and more mundane: a lipidated peptide is chemically a surfactant-like molecule, and its solubility and surface behaviour differ from those of an unmodified peptide of the same length. That has implications for how it is dissolved, discussed in diluents for reconstituting research peptides.

C-terminal amidation. Replacing the terminal carboxylic acid with a carboxamide removes a negative charge and can improve stability. It also shifts the molecule's mass by about one dalton relative to the free acid — small, but exactly the kind of difference that matters when checking an observed mass against a theoretical one.

All three are recorded in the sequence notation rather than in prose, and misreading that notation means miscalculating the expected mass. The conventions are set out in Peptide Nomenclature and Sequence Notation.

Reading EC₅₀ values critically

An EC₅₀ is not a physical constant of a molecule. It is a property of a molecule in a particular assay, and several features of that assay move the number:

• Receptor expression level. Transfected lines often overexpress the receptor. High receptor density amplifies the response at any given concentration, which shifts EC₅₀ lower — making a compound look more potent than it would appear in a system with physiological receptor numbers. This is probably the single largest source of between-paper variation.
• Cell background. CHO and HEK293 cells differ in their complement of G proteins, effectors, and phosphodiesterases. The same receptor in a different host does not necessarily give the same numbers.
• Readout and amplification. cAMP sits several steps downstream of receptor activation, and each step amplifies. A proximal readout and a distal one report different apparent potencies for the same event. β-arrestin recruitment assays, being differently coupled, routinely give different values again.
• Assay conditions. Incubation time, temperature, buffer composition, and especially the presence or absence of a phosphodiesterase inhibitor all shift measured cAMP.
• Reference ligand. Because E_max is usually normalised to the endogenous ligand, the choice and quality of that comparator propagates directly into the reported efficacy.

There is also a mundane input that is easy to overlook: the accuracy of the stock concentration. An EC₅₀ is only as good as the assumed concentration of the peptide, and that assumption depends on the lot's net peptide content rather than the gross mass of powder in the vial. A stock prepared from gross mass is systematically more dilute than intended, which shifts every derived potency value. This is not a rounding error — see Net Peptide Content vs. Purity, and How to Read a Peptide COA for where the figure appears on a certificate.

The practical rule that follows: compare potency values within a single study, not across studies. A paper that assays several compounds side by side under one protocol supports internal comparison. Two numbers lifted from two different papers, generated in different cell lines at different receptor densities with different readouts, are not comparable, and treating them as if they were is a common analytical error.

What receptor data does not tell you

This is the most important section on the page, and it is deliberately last.

Everything described above is mechanism measured in an artificial system. A cAMP curve from a receptor-transfected cell line establishes that a molecule can activate a receptor in that dish, at those concentrations, under those conditions. It establishes nothing whatsoever about what happens in an intact organism, because it excludes essentially every variable that would determine that: distribution, metabolism, clearance, receptor density and distribution in real tissues, counter-regulatory signalling, and the entire integrated physiology of a living system.

The gap between the two is not a small extrapolation to be made with care. It is a category difference. Receptor-level potency and organism-level outcome are different kinds of claim supported by different kinds of evidence, and a figure of the first kind provides no support at all for a statement of the second kind.

Accordingly: nothing on this page, and no receptor-level figure reported for any compound in this catalog, constitutes evidence of any physiological, metabolic, clinical, or health-related effect in humans or animals. The compounds referenced here are unapproved research chemicals supplied strictly as reference materials for laboratory investigation. They are not drugs, dietary supplements, or approved therapies of any kind; they are not authorised by the FDA or any comparable authority for the treatment, prevention, or diagnosis of any condition; and they are not for human or veterinary use. This page makes no comparison to, and no claim of equivalence with, any commercial or approved medicine. See What Is Research Use Only (RUO)? for what that designation means and what responsibilities it places on the purchaser.

FAQ

What is the difference between EC50 and Emax? They describe different properties of the same concentration–response curve. EC50 is the concentration producing half the maximal response and measures potency — how little compound is required. Emax is the maximal response achieved, usually normalised to the endogenous ligand, and measures efficacy — how much response the compound can produce at all. A compound can be very potent (low EC50) yet only partially efficacious (Emax well below the reference), which is a distinct pharmacological profile. Reporting EC50 without Emax gives an incomplete description.

Why do published EC50 values for the same peptide differ between papers? Because an EC50 is a property of an assay, not a constant of the molecule. Receptor expression level is the biggest factor — transfected cell lines often overexpress the receptor, which amplifies the response and shifts the value lower. Cell background, the choice of readout and its position downstream of the receptor, incubation conditions, phosphodiesterase inhibition, the reference ligand, and the accuracy of the stock concentration all contribute as well. Potency values should be compared within a single study, not lifted from separate ones.

Does a triple agonist mean something is stronger than a dual agonist? No. The terms describe how many of these related receptors show measurable agonism in an assay, not strength, quality, or magnitude of any effect. A triple agonist may be far more potent at one of its receptors than the others, and that balance across receptors is a separate reported variable. Receptor coverage is a description of mechanism only and supports no inference about effects in an organism.

Is GLP-2 TZ related to glucagon-like peptide-2? No. Glucagon-like peptide-2 is a distinct endogenous peptide with its own receptor, GLP2R, and its own separate literature; it is not one of the two canonical incretins. "GLP-2 TZ" is a catalog designation for tirzepatide, which is characterised as a dual GIP/GLP-1 receptor agonist and does not act at the GLP-2 receptor. There is likewise no "GLP-3" receptor — "GLP-3 RT" is a catalog label for retatrutide.

Why does Aib appear at position 2 in these sequences? Because the enzyme dipeptidyl peptidase-4 (DPP-4) cleaves the first two residues from the N-terminus of the endogenous incretins, and class B1 receptor activation depends on that N-terminal region. Alpha-aminoisobutyric acid is alanine with a second methyl group on the alpha-carbon, which makes that carbon quaternary and removes the backbone geometry the enzyme recognises, so the cleavage site is defeated. It also stabilises helical structure, which can independently aid receptor engagement.

Do receptor-level assay results indicate what a compound does in a person? No, and the distinction is a category difference rather than a small extrapolation. A cAMP curve in a receptor-transfected cell line shows that a molecule can activate a receptor in that dish under those conditions. It excludes distribution, metabolism, clearance, real tissue receptor distribution, counter-regulatory signalling, and integrated physiology. Receptor potency and organism-level outcome are different kinds of claim supported by different kinds of evidence. The compounds referenced here are unapproved research chemicals for laboratory use only and are not for human or veterinary use.

References

• IUPHAR/BPS Guide to Pharmacology — Glucagon receptor family
• NCBI Gene — GLP1R (glucagon-like peptide 1 receptor)
• NCBI Gene — GIPR (gastric inhibitory polypeptide receptor)
• NCBI Gene — GCGR (glucagon receptor)
• Bachem — Quality Control of Amino Acids and Peptides: A Guide