Tirzepatide is a synthetic peptide of 39 residues, amidated at its C-terminus, that activates two receptors rather than one: the glucose-dependent insulinotropic polypeptide (GIP) receptor (GIPR) and the glucagon-like peptide-1 (GLP-1) receptor (GLP-1R) [1, 2]. It first appeared in the literature under the development code LY3298176, described as a fatty-acid-modified peptide with dual GIP and GLP-1 receptor agonist activity and characterized in signaling and functional assays in cell lines expressing recombinant or endogenous incretin receptors [2]. That first report presents it as a GIP-based dual incretin receptor agonist [2]; a later structure–activity review instead describes its residues as drawn mainly from GLP-1, GIP, and semaglutide, with a few unique to the molecule [1]. Both descriptions are in the published record, and they emphasize different provenances for the sequence.

The interesting thing about the molecule, from a bench point of view, is that it is not simply a GLP-1 agonist with GIP activity bolted on. The published pharmacology says it engages the two receptors with different signatures — closely resembling the native hormone at one of them and diverging sharply from the native hormone at the other [3, 4]. This guide sets out what has been measured, in which systems, where the accounts disagree, and where the record is empty.

Identity and characterization

The values below are the ones Echo Peptides publishes for this compound. They are the numbers a receiving laboratory should reconcile against an incoming certificate of analysis before anything else happens to the vial.

The sequence differs from the endogenous incretins at a number of positions, and one structure–activity review is given over to analyzing those discrepancies [1]. Two positions carry α-aminoisobutyric acid (Aib), a non-proteinogenic residue [1, 2]. At position 2 it occupies a site cleaved by dipeptidyl peptidase-4, a substitution reported to prevent that cleavage [1]; at position 13 its contribution is less settled — the same review suggests it lowers GLP-1 activity while leaving GIP activity intact, and says in the same breath that the hypothesis lacks supporting evidence [1]. The C-terminal region is prolonged with a sequence taken from the C-terminus of exenatide [1]. And a C20 fatty diacid is conjugated through a spacer to the lysine at position 20 [1, 2]. That last modification is the same design principle as the C16 chain on liraglutide and the C18 chain on semaglutide: the acyl group binds serum albumin, a strategy used to extend half-life [1].

Those features also bear on how the molecule is made. A 2025 study in the Journal of the American Chemical Society took the total synthesis of tirzepatide as one of two demonstrations for a set of cyclic dipeptide-derived hydroxylamine building blocks, which form canonical amino acids directly under modified α-ketoacid–hydroxylamine (KAHA) ligation conditions; the authors note that the peptide contains Aib residues and a fatty acid side chain modification, and present the work as extending the method to the practical synthesis of challenging peptide targets [5]. That is a reason to read purity and identity data on incoming material closely rather than assume a routine synthesis.

One reconciliation step is easy to skip and worth not skipping. A certificate of analysis is issued against a specific quantity of material, and a certificate published alongside a listing is not automatically a certificate for the vial size in front of you; where that is the case on this site, the product page says so explicitly rather than letting the document imply coverage it does not have. Check the mass the certificate was issued against, check whether it carries a batch identifier at all, and reconcile the formula and molecular weight above against whatever the document reports before treating either as confirmed.

The two receptors

GLP-1R and GIPR are class B1 G protein-coupled receptors. In pancreatic beta cells, activation of either raises cyclic AMP and increases glucose-dependent insulin secretion [6]. Because the two receptors are closely related and are co-expressed in the same cells, the consequences of engaging both at once are hard to isolate in vivo, and the downstream signaling pathways of GLP-1R–GIPR co-agonists are, in the words of one review of beta-cell incretin signaling, not fully elucidated [6].

The word doing quiet work in that sentence is glucose-dependent. The insulin secretion these receptors potentiate is conditional on ambient glucose [6], which means a beta-cell or islet assay run at one glucose concentration can report a very different picture from the same assay run at another. Any potency figure quoted for this compound in an insulin-secretion readout is a figure for a particular glucose condition, and comparisons across published studies should establish that the conditions match before treating the numbers as commensurable.

The GIP arm carries an unresolved puzzle of its own. Both activation and inhibition of GIPR, in combination with GLP-1R activation, have been pursued pharmacologically and have been reported to give similar outcomes — a point made directly in a 2024 characterization of GIPR genetic variants, which names tirzepatide as the co-agonist case and a separate agent combining GIPR antagonism with GLP-1R agonism as the opposing one [7]. None of the work cited here reconciles the two. That is worth holding in mind before treating any single mechanistic result on this compound as settled.

What "imbalanced and biased" means here

The most detailed pharmacological description of tirzepatide comes from a 2020 study that established a method for calculating receptor occupancy at the exposures modeled in that work, and concluded the compound engages GIPR to a greater degree than GLP-1R — an imbalanced mechanism of action [3]. In signaling assays from the same work, tirzepatide mimicked the actions of native GIP at GIPR, but at GLP-1R it showed bias favoring cAMP generation over β-arrestin recruitment, coincident with a weaker ability to drive GLP-1R internalization than GLP-1 itself [3].

Structural work from the same group supports the same asymmetry. Cryogenic electron microscopy combined with molecular dynamics simulations found that although tirzepatide closely resembles native GIP in how it activates GIPR, it differs markedly from GLP-1 in its activation of GLP-1R, resulting in less agonist-induced receptor desensitization; the authors framed the mode of action of a multireceptor agonist as determined by fatty acid modification combined with amino acid sequence [4].

An independent group reached a compatible conclusion by different means. Using bioluminescence resonance energy transfer and live-cell HILO microscopy, they reported that tirzepatide and a second dual agonist preserved maximal cAMP production despite only partial Gαs recruitment, paralleled by diminished ligand-induced receptor internalization at both target receptors, relative to native and acylated GLP-1 agonists [8]. Their reading was that the structure of GLP-1/GIP dual agonists may confer a biased agonism that modulates receptor trafficking as well as intracellular signaling [8].

A caution on all three: "bias" is a description of how a ligand behaves in a particular assay against a particular reference, and its apparent magnitude depends on the readout, the cell background, and receptor expression level. These are convergent results in engineered and primary systems, not a measured property of the molecule in isolation.

Where the accounts do not line up

β-arrestin is the point of divergence. In primary islets, β-arrestin1 was found to limit the insulin response to GLP-1 but not to GIP or tirzepatide, which the authors offered as a suggestion that the compound's biased agonism enhances insulin secretion [3]. A 2023 study of β-arrestin 2 (ARRB2) in mouse and human islets found something that does not sit neatly alongside that: GIP-potentiated insulin secretion needs ARRB2 in both species, the GIPR–ARRB2 axis mediating GIP-induced F-actin depolymerization rather than cAMP/PKA or ERK signaling — while tirzepatide did not require ARRB2 for its potentiation of insulin secretion [9].

Read together, and without putting a conclusion in anyone's mouth: if tirzepatide mimics native GIP at GIPR in signaling assays [3], yet native GIP and tirzepatide differ in whether ARRB2 is required for insulin potentiation in islets [9], then the mimicry is not uniform across every readout. Which readout matters depends on what is being asked.

The 2024 variant study pushes in a third direction. Pharmacological characterization of 47 GIPR variants, combined with burden testing and mouse work, indicated that β-arrestin recruitment is necessary for GIPR function: variants impaired in both Gs signaling and β-arrestin 2 recruitment contributed to lower adiposity-related traits, whereas variants impaired in Gs signaling alone did not, and genetic ablation of β-arrestin 2 impaired cAMP production and decreased GIP efficacy on glucose control in male mice [7]. A compound described as biased away from β-arrestin at one receptor, acting on a system where β-arrestin appears required at the other, is not a contradiction — but it is not a resolved picture either.

Adipocytes and rodent models

Two lines of preclinical work address where the GIPR component acts. In obese mice, tirzepatide improved insulin sensitivity to a greater extent than GLP-1R agonism; comparing wild-type with Glp-1r-null mice, and so removing GLP-1R-driven changes in body mass, the compound still improved insulin sensitivity by enhancing glucose disposal in white adipose tissue [10]. A long-acting GIPR agonist tested alongside it was likewise found to enhance insulin sensitivity by augmenting glucose disposal in that tissue, which is the observation that ties the effect to the GIPR arm rather than to the GLP-1R arm [10]. The effect was associated with reduced circulating branched-chain amino acids and ketoacids, and with upregulation of catabolic genes in brown adipose tissue — an association the authors reported as such [10].

Cellular work in human adipocytes and mouse models found the GIPR contribution to be conditional on insulin state: GIPR agonism enhanced insulin signaling, augmented glucose uptake, and increased conversion of glucose to glycerol cooperatively with insulin, whereas in the absence of insulin, GIPR agonists increased lipolysis [11]. The model the authors proposed is a differential one — long-acting GIPR agonists modulating fed and fasted adipose tissue function in opposite directions [11]. Note that several of the experiments in that work used a long-acting GIPR agonist rather than tirzepatide itself, which is the appropriate design for isolating the GIPR arm but means the read-across to the dual agonist is an inference.

Where the design came from

The unimolecular dual incretin concept predates this molecule. A 2013 report described a peptide derived from an intermixed GLP-1 and GIP sequence with potent, balanced co-agonism at both receptors, showing synergy in reducing fat mass in obese rodents, whereas a selective GIP receptor agonist demonstrated negligible weight-lowering efficacy; duration of action was refined through site-specific lipidation or PEGylation [12]. Tirzepatide's own preclinical characterization followed the same shape: activation of both receptors in vitro, then glucose-dependent insulin secretion and improved glucose tolerance in mice through both receptors, with chronic administration reducing body mass and food intake more than a GLP-1 receptor agonist did [2].

One measurement that came back negative

A comparative study of brain uptake in male mice measured four incretin receptor agonists with radiolabeled tracers and multiple-time regression analysis. Three of the four had fast uptake rates within the first hour. Tirzepatide did not appear to cross the blood-brain barrier in that window; like albumin, it entered slowly over six hours, which the authors said was presumably via extracellular pathways [13]. That is one comparative study in one species, it says nothing about distribution to tissues other than brain, and "presumably" is the authors' word, not a demonstrated route. It is also the kind of negative result that tends to disappear from summaries, which is why it is here.

What is not established

Several things are worth stating plainly. The β-arrestin results above come from three different systems — engineered cell lines, primary mouse and human islets, and variant panels with mouse genetics — and none of the papers cited here reconciles them. Much of the mechanistic literature originates with the developing company or with close collaborators, and the relevant competing-interest statements are disclosed on the papers themselves; that does not invalidate the work, but a reader should know it. Whether GIPR agonism and GIPR antagonism can both produce the same downstream result, and why, remains open [7].

The occupancy argument deserves its own caveat. The claim that the compound engages GIPR more than GLP-1R rests on a calculation method introduced in the same paper that reports the imbalance [3]. Occupancy is inferred from binding parameters and exposure, not observed directly, so the conclusion travels only as far as the assumptions behind it. The signaling asymmetry that accompanies it has separate support — structural work from the same group [4], and trafficking data from an independent laboratory [8] — but neither of those studies calculates occupancy, so neither of them tests the occupancy figure itself.

There is also a practical point about potency comparisons. The C20 diacid is there to bind albumin [1], so the free concentration of the peptide in a serum-containing medium is not the nominal concentration, and potency values measured in different media are not automatically comparable. That is a matter of experimental design rather than a claim about the compound.

Handling and storage

Echo Peptides supplies tirzepatide as a lyophilized white powder, specified for refrigerated storage at 2–8 °C and reconstitution in bacteriostatic water or a comparable sterile buffer. Confirm identity and purity against the certificate of analysis for the vial in hand and record the reconstitution volume, since the acylated peptide's behavior in serum-containing media makes the nominal concentration a poor stand-in for the free one.

This compound is supplied for laboratory, in-vitro research use only. It is not for human or veterinary use, nor for any diagnostic or therapeutic application, and nothing above describes such use. Each result summarized above is presented as a description of the system it was obtained in.