Skip to content
PeptideHormone

Why tirzepatide isn’t two drugs in one

The usual account of tirzepatide is that semaglutide hits one receptor, tirzepatide hits two, and two beat one — a story the head-to-head weight-loss result seems to confirm. The pharmacology is stranger. Tirzepatide is built on the GIP sequence, leans more on the GIP receptor than the GLP-1 receptor, and at the GLP-1 receptor it signals in a biased way that native GLP-1 does not. An imbalanced, biased single molecule, not a tidy pairing of two drugs. What “imbalanced” and “biased” mean, why the GIP half still earns its place, and the one thing the trial cannot tell you.

9 min read · reviewed November 2026

The explanation everyone uses

The standard account of tirzepatide fits in a sentence. Semaglutide activates one receptor, tirzepatide activates two, and two is more than one. The head-to-head result seems to settle it. In adults with obesity and without diabetes, tirzepatide produced a mean weight reduction of 20.2% over 72 weeks against 13.7% for semaglutide (Aronne et al., NEJM, 2025). The trial detail is on the comparison page.

The arithmetic is tidy and the picture behind it is of a GLP-1 drug with a GIP drug bolted on, each doing its own job at full strength. That picture is wrong in two specific ways, both documented by the people who made the molecule. Neither changes the trial result. Both change what the result means.

First correction: it is mostly a GIP molecule

Start with what tirzepatide is made from. It was not built by adding GIP activity to a GLP-1 peptide. It is a 39-amino-acid chain based on the sequence of GIP, modified so that it also fits the GLP-1 receptor, with a fatty-acid chain attached for once-weekly dosing (Coskun et al., 2018).

The heritage shows in how it behaves. When the developers worked out how much of each receptor the drug occupies at the doses that work in people, the answer was lopsided: tirzepatide engages the GIP receptor to a greater degree than the GLP-1 receptor. They called it an imbalanced agonist, and put the word in the title of their paper (Willard et al., 2020). So the mental model of two equal halves is off before the biology even starts. The molecule leans on GIP.

Second correction: the GLP-1 half is a different shape of signal

The second correction is subtler and more interesting. A receptor is not a light switch. When GLP-1 binds its receptor, two things happen: the receptor turns on a G-protein that raises the messenger cAMP, which drives insulin release, and it also recruits a protein called beta-arrestin, which switches the receptor off and drags it inside the cell to be recycled. The first is the signal; the second is the shut-off.

Tirzepatide does not pull both levers equally. At the GLP-1 receptor it favors the cAMP arm over beta-arrestin, and as a result it triggers less of the internalization that native GLP-1 causes (Willard et al., 2020). That is what pharmacologists mean by biased agonism: same receptor, different mix of downstream signals. The trafficking studies tell the same story, with tirzepatide keeping more receptor at the cell surface than matched GLP-1 agonists (Novikoff et al., 2021), and the structural work shows why: tirzepatide closely mimics GIP at the GIP receptor but sits in the GLP-1 receptor in a way that resists desensitization (Sun et al., PNAS, 2022).

Why would keeping a receptor at the surface matter? Because a receptor that is pulled inside stops responding. Work on engineered GLP-1 agonists found that compounds which avoid beta-arrestin and stay on the membrane can produce more sustained insulin release, and in mice did so with less nausea (Jones et al., 2018). Bias is not a quirk of tirzepatide; it is a design lever the whole field is now pulling. The point for reading the drug is that its GLP-1 activity is not a copy of GLP-1 turned down a notch. It is a different shape of signal.

GLP-1 receptorone receptor, two signalsG-protein → cAMPthe insulin signal · tirzepatide favors thisβ-arrestin → internalizethe shut-off · tirzepatide drives this lessbiased toward the top arm
Native GLP-1 drives both arms. Tirzepatide leans on the cAMP arm and triggers less of the beta-arrestin shut-off, keeping more receptor at the cell surface. That is “biased” agonism.

Does the GIP half earn its place?

If the molecule leans on GIP, the obvious question is whether GIP contributes anything, since the role of the GIP receptor in metabolism was contested for years. Three lines of evidence say it does, with the usual caveat about where each one comes from.

In human pancreatic islets, blocking the GIP receptor lowered the insulin response to tirzepatide, direct evidence that the GIP arm is doing work in human tissue (El et al., Nat Metab, 2023). A clamp study in people traced the drug’s glucose-lowering to concurrent gains in beta-cell function and insulin sensitivity (Heise et al., 2022). And in mice, GIP receptor activity improved insulin sensitivity even when weight was held constant (Samms et al., 2021), while GIP signaling in the brain was needed for the extra weight loss of GIP/GLP-1 co-agonism over GLP-1 alone (Zhang et al., 2021). There is even a tolerability angle: in animals, GIP receptor agonism blunted the nausea that GLP-1 agonism causes (Borner et al., 2021), which would let a person tolerate more of the drug.

The genuinely strange part

The GIP receptor is the one target where both activating and blocking it have produced weight loss alongside GLP-1 activity. A selective GIP receptor agonist caused modest weight loss on its own in an early human study (Roell et al., 2026), yet GIP receptor blockers also drive weight loss in the same combination. One proposed reconciliation is that constant agonism eventually desensitizes the receptor in fat tissue, ending up functionally like blockade (Killion et al., 2020). This catalog takes the puzzle up on its own in the GIP paradox.

What the trial does and does not show

Return to the 20.2% versus 13.7%. It is a real, prospective, head-to-head result, and it establishes that this molecule produced more weight loss than that one at their respective maximum tolerated doses. What it cannot do is tell you why.

The trial compared two finished drugs, not two mechanisms. It did not include a version of tirzepatide with the GIP activity removed, or one with the GLP-1 bias undone, so it cannot attribute the margin to the second receptor, to the imbalance, to the biased GLP-1 signaling, or simply to tirzepatide reaching a higher effective dose. Each of those is a live hypothesis. The clean “two receptors beat one” story is the one thing the trial was not designed to prove, however reasonable it sounds. That gap between a result and its mechanism is the recurring lesson of mechanism is not proof, running here in the opposite direction: a solid clinical result whose mechanistic explanation is still an inference.

The simple picture of tirzepatide compared with what the pharmacology actually shows.
FeatureThe simple pictureWhat the pharmacology shows
BackboneA GLP-1 drug with GIP addedBuilt on the GIP sequence
BalanceEqual activity at both receptorsEngages the GIP receptor more than the GLP-1 receptor
At the GIP receptorActs like GIPActs like GIP
At the GLP-1 receptorActs like GLP-1Favors cAMP over beta-arrestin; less receptor internalization
Why it beat semaglutideTwo receptors are better than oneNot isolated by any trial
The head-to-head trial lives in the last row: a measured difference in weight loss, with the molecular reason for it still unproven.

Reading tirzepatide honestly

None of this diminishes the drug. It makes it more interesting. The flat “two is better than one” summary hides the actual engineering, which is a GIP-based molecule carrying a deliberately reshaped GLP-1 signal, tuned by which pathways it favors and how long it stays on the receptor.

  • Not a GLP-1 drug with GIP added. Tirzepatide is built on the GIP sequence and modified to also hit the GLP-1 receptor, not the reverse.
  • Imbalanced. At clinical doses it occupies the GIP receptor more than the GLP-1 receptor. The two halves are not equal.
  • Biased. At the GLP-1 receptor it favors cAMP signaling over beta-arrestin, so it internalizes the receptor less than native GLP-1 does.
  • The GIP half contributes. Human islets, clamp studies and mouse work all point to real GIP-receptor contributions, including a possible easing of nausea — though some of this is preclinical.
  • The trial is a result, not a mechanism. SURMOUNT-5 shows tirzepatide produced more weight loss than semaglutide. It does not isolate which molecular feature caused the difference.

Educational reference on mechanism, summarized from public scientific literature and clinical-trial disclosures and simplified in places. Not individualized medical advice or a prescription. Where a dose or regimen appears it is reported from its source as research literacy — so you can read the evidence yourself — not an instruction to dose. Specific compounds and trials are named to explain the science; verify any claim against the linked primary sources.