A meal starts a conversation we cannot hear. The intestine detects nutrients, the pancreas responds, and the liver helps regulate energy availability. GLP-1, GIP and glucagon are among the messages connecting these functions. Understanding their differences has opened a major design question: what happens when one molecule can activate several of their receptors?
The story begins before semaglutide, tirzepatide and retatrutide. It leads back to experiments that identified each signal, a surprising observation involving a reptile, and researchers who began bringing different activities into one structure. Following that path makes the field easier to read: behind the abbreviations are molecules, mechanisms and studies addressing different questions.
Two molecules, approximately US$71 billion
The scale is visible beyond the laboratory, too. In 2025, Novo Nordisk and Eli Lilly medicines containing semaglutide and tirzepatide generated approximately US$71 billion in worldwide sales. That figure shows the commercial adoption reached by two peptide designs built on decades of research.
Semaglutide, sold as Ozempic, Wegovy and Rybelsus, recorded combined sales of DKK228,288 million. Tirzepatide, sold as Mounjaro and Zepbound, generated US$36,507 million. Novo Nordisk full-year 2025 results. Eli Lilly full-year 2025 results.
Two molecules. Approximately US$71 billion.
Worldwide sales of Ozempic, Wegovy, Rybelsus, Mounjaro and Zepbound over the full financial year.
Billions of US dollars. Axis starts at zero.
Semaglutide converted from DKK at the annual average of DKK6.62/US$; figures rounded. Editorial calculation using results from Novo Nordisk and Eli Lilly.
Converting semaglutide sales to approximately US$34.5 billion is an editorial calculation: the Danish-krone total is divided by 6.62, Novo Nordisk’s published average 2025 exchange rate. Adding tirzepatide gives a rounded total of US$71 billion. These are worldwide sales of five brands over twelve months, covering diabetes and obesity, not an estimate of the entire GLP-1 market. Annual report and exchange rates.
For perspective, OpenAI disclosed more than US$20 billion in annualized revenue for 2025, while Anthropic reported approximately US$9 billion at that year’s end. An annualized run rate is like taking the speed at one moment and projecting it across a year; annual sales measure the distance actually traveled. The figures illustrate the scale of these therapies alongside two leading AI businesses, using different metrics. They do not establish that GLP-1 is larger than the entire AI sector or describe the current comparison in 2026. OpenAI disclosure. Anthropic disclosure.
For organizations following peptide research, the opportunity lies in understanding the path behind that scale: molecular characterization, reproducible findings, clinical development and authorizations for specific products and indications. That trajectory invites further research questions and capabilities. Future commercialization of new medicines will depend on evidence, approvals and applicable rights in each market. Approved-medicine sales do not represent research-material sales or predict approval of other candidates.
From natural hormones to molecular design
The field developed along several paths. Glucagon emerged from research on pancreatic extracts. GIP was linked to the ability of intestinal signals to stimulate insulin secretion. Later, work on the proglucagon precursor helped identify GLP-1 and its activity.
A precursor is a molecule that can be processed to produce others. In 1987, Ørskov, Holst and colleagues studied proglucagon processing in human pancreatic and intestinal tissues. Their work helped describe how one precursor gives rise to different products depending on the tissue. Original research on proglucagon.
An agonist is a molecule that activates a receptor. Agonists designed for GLP-1, GIP or glucagon draw on knowledge of these natural signals, with modifications aimed at specific activity profiles and durations. A natural hormone and a related pharmacological molecule have their own identities and characteristics.
Think of a receptor as a doorway with a recognition mechanism. Certain molecules fit and initiate a response inside the cell. The analogy explains activation; the actual response also depends on the tissue, signal strength and how long that activity lasts. Designing an agonist means investigating all of those variables.
GLP-1: connecting nutrients with the pancreatic response
GLP-1 stands for glucagon-like peptide-1. In the intestine, it is associated with endocrine cells called L cells. It belongs to the incretins: intestinal signals that help increase the insulin response to nutrients.
In 1987, Svetlana Mojsov, Gordon Weir and Joel Habener showed that an active form called GLP-1(7-37) stimulated insulin release in a perfused rat pancreas. The longer form tested in the same study did not show that activity. Original publication by Mojsov and colleagues.
The difference was in the molecular form. This finding, obtained in a rat organ under experimental conditions, shows why sequence and processing are central to the story. One general name can encompass variants with different behavior.
Later human studies examined its effects on insulin, glucagon and glucose. A 1993 study by Nauck and colleagues investigated GLP-1 in people with type 2 diabetes and documented a reduction in fasting hyperglycemia under the experimental conditions. Insulin stimulation depended on glucose levels. A reduction in glucagon secretion was also observed. Both responses help explain how the signal participates in glucose regulation. Human study from 1993.
Another 1987 study, by Daniel Drucker and colleagues, showed increases in cyclic AMP and insulin gene expression in a rat islet cell line. Cyclic AMP is an intracellular messenger. This experiment provided information about how a peptide signal can connect with the cellular response that produces insulin. Original research on GLP-1 signaling.
GLP-1 in digestion and appetite research
GLP-1 research expanded to the rate at which the stomach delivers its contents to the intestine, known as gastric emptying. Schirra and colleagues evaluated this process alongside pancreatic and glucose responses in healthy volunteers. They observed an initial delay in gastric emptying, greater total insulin release and a lower glucagon response after a meal. Original study on gastric emptying.
Another line of research examined satiety and food intake. In 1998, Flint and colleagues reported greater satiety and lower energy intake during an experimental evaluation of GLP-1 in humans. Original work on satiety.
The conversation had grown beyond a single glucose measurement. The rate of nutrient delivery and signals associated with continued eating mattered too. Picture a distribution center: the amount of material matters, but so do arrival times and how intake is organized. That image helps connect processes these studies measured separately.
GIP: the other incretin and its distinct activity
GIP is now known as glucose-dependent insulinotropic polypeptide. Its historical name, gastric inhibitory polypeptide, reflects an earlier stage of research. In 1973, Dupré, Ross, Watson and Brown published an experiment demonstrating stimulation of insulin secretion by GIP in humans. Original publication from 1973.
GIP is released by K cells in the small intestine in response to nutrients. Its function links what reaches the intestine with the pancreatic response. Glucose dependence matters: activity is interpreted within the metabolic state at that moment.
A 2011 study by Christensen and colleagues evaluated GIP in ten healthy men at different glucose levels. At elevated glucose levels, GIP enhanced insulin secretion. At low glucose levels or during fasting, it increased the glucagon response and had little or no effect on insulin. The authors described it as a glucose-dependent bifunctional regulator. Original work on GIP, insulin and glucagon.
The same messenger produced different responses in different settings. That makes GIP interesting in its own right: understanding a signal requires understanding the state of the system receiving it. Dual agonists bring this question into molecular design by investigating what GIP activation contributes alongside GLP-1.
Glucagon: energy availability and new metabolic questions
Glucagon's history began with the study of pancreatic extracts. In 1923, Kimball and Murlin described an activity capable of increasing blood glucose. Decades later, molecular research made it possible to work with the hormone and analogs of defined sequence. An analog development paper documenting this historical background.
Glucagon is produced mainly by pancreatic alpha cells and acts on the liver to increase glucose availability. Related processes include mobilizing stored glycogen and producing new glucose. This function helps explain its relationship with the balance between energy supply and use. Physiological context in the study by Tan and colleagues.
Interest in activating glucagon within a broader metabolic molecule also arose from experiments on energy expenditure. In 2013, Tan and colleagues studied glucagon and GLP-1 in human volunteers. The combination increased resting energy expenditure during the experiment and attenuated the hyperglycemia observed with glucagon. Original publication on the experimental combination.
The observation is interesting because of the tension between these functions. One signal can increase glucose availability; another can help modulate the metabolic response. Studying how they coexist opens broader questions about energy. This was an experimental result: activity depends on the molecule, its balance of receptor actions and the conditions evaluated.
The Gila monster and a discovery story
One of the field's most striking stories begins with a reptile. In 1992, John Eng and colleagues described exendin-4, a 39-amino-acid peptide isolated from the venom of Heloderma suspectum, known as the Gila monster. Original publication on exendin-4. The following year, Göke and colleagues demonstrated that it acted as a GLP-1 receptor agonist in cellular systems and rat islets. Receptor interaction study.
The twist is where the discovery happened. A molecule from reptile venom made it possible to study a pathway already present in the human body. Observation came first, followed by questions about the receptor and characterization of the activity. Exendin-4 has its own sequence. Semaglutide, tirzepatide and retatrutide have the molecular designs described in their respective publications.
In his conversation with Julian Dorey, Bakri recalls his grandmother's diabetes care with Byetta and how that experience influenced his interest in medicine. The account adds a human dimension: behind the interest in receptors are families who have experienced changes in disease care. It is his personal recollection, not a trial of the medication's outcomes.
Later, both Dorey and Bakri and the Santa Cruz Medicinals conversation highlight the move from one to several receptor activities. The scientific history linked here helps explain that enthusiasm: each design opens a new question about coordinating metabolic signals.
Semaglutide: design around the GLP-1 receptor
With the signals better understood, the question changed: how could that knowledge become a molecule with its own characteristics and measurable outcomes? The following publications show three design paths and three research programs. Each belongs within its own population and duration.
Semaglutide is a GLP-1 analog. Lau and colleagues published its molecular development in 2015 as part of the search for a prolonged activity profile. It is an example of how sequence engineering turns knowledge of a natural signal into a molecule with its own characteristics. Original publication on semaglutide development.
STEP 1, published in 2021, included 1,961 adults with obesity or overweight and a weight-related condition, without diabetes. At 68 weeks, average weight reduction was 14.9% with semaglutide and 2.4% with placebo; both groups received a lifestyle intervention. STEP 1 trial.
SELECT expanded the research to cardiovascular outcomes. In 17,604 adults with established cardiovascular disease and overweight or obesity, without diabetes, the composite outcome of cardiovascular death, nonfatal myocardial infarction or nonfatal stroke occurred in 6.5% with semaglutide and 8.0% with placebo. The study also reported more discontinuations due to adverse events with semaglutide. SELECT trial, published in 2023.
Tirzepatide: activating GIP and GLP-1 together
Tirzepatide was developed as LY3298176, a molecule that activates GIP and GLP-1 receptors. In 2018, Coskun and colleagues published its preclinical characterization and an initial clinical evaluation. The design question was whether combining both activities could contribute additional results. Development publication for LY3298176.
SURMOUNT-1, published in 2022, evaluated tirzepatide over 72 weeks in 2,539 adults with obesity or overweight and a weight-related complication, without diabetes. Average weight reduction ranged from 15.0% to 20.9% in the tirzepatide groups and was 3.1% with placebo. The most common adverse events were gastrointestinal. SURMOUNT-1 trial.
Retatrutide: three receptors in one molecule
Retatrutide activates GIP, GLP-1 and glucagon receptors. Its design allows incretin signaling and glucagon receptor activity to be investigated together. The 2023 phase 2 trial included 338 adults. At 48 weeks, the highest-dose group studied had an average weight reduction of 24.2%, compared with 2.1% with placebo. Gastrointestinal events and dose-related increases in heart rate were reported. Original retatrutide trial.
Retatrutide: weight change at 48 weeks
Phase 2 trial, 338 adults with obesity or overweight and a weight-related condition.
Magnitude of mean body-weight reduction
Bars start at zero. These are groups from the same trial, not a comparison of molecules across different studies. Original publication.
It is tempting to line up the percentages and read them as a ranking. Understanding them means keeping the context: STEP 1, SURMOUNT-1 and the retatrutide trial had different populations, groups and durations. They were not a direct comparison between molecules. This article describes selected historical publications without summarizing the full subsequent clinical program or each compound's current regulatory status.
Receptors: a quick reference
This overview summarizes the profiles described in the molecules' development publications and trials. Activating multiple receptors is part of a specific molecule's design; it does not describe a mixture or a combination protocol.
| Molecule | GLP-1 | GIP | Glucagon |
|---|---|---|---|
| Semaglutide | Yes | No | No |
| Tirzepatide | Yes | Yes | No |
| Retatrutide | Yes | Yes | Yes |
Why this field offers scientific opportunity
The story began with attempts to understand the body's messages and reached molecules that activate several receptors. That change brings together decades of discovery, engineering and measurement. For distributors and organizations, the receptor map offers an initial orientation; publications and the identity of each material supply the depth needed for an informed commercial conversation.
Following the field with enthusiasm means following its questions: what each receptor contributes, how the signals are balanced, and which results can be reproduced in each setting. That is the appeal of this generation of metabolic peptides: designing with an increasingly precise understanding of biology.
The Huberman Lab podcast with Dr. Abud Bakri complements this history with a contemporary conversation. The article is grounded in the linked publications. Clinical results apply to each trial's preparations and populations and do not characterize products intended exclusively for research.