Companies seek a second obesity treatment revolution—in pill form
Smaller, more easily made drugs that target GLP-1 could treat obesity and diabetes with fewer side effects
15 Apr 2025
6:15 PM ET
Small molecules such as Pfizer’s discontinued candidate danuglipron bind to the same site on the glucagonlike peptide-1 receptor as the much larger, natural GLP-1 hormone.Zhang et al., Molecular Cell, 80:3 (2020)
Drugs that mimic glucagonlike peptide-1 (GLP-1), such as semaglutide—marketed as Ozempic or Wegovy—have revolutionized the treatment of obesity and type 2 diabetes, but they have major drawbacks. “[They] are expensive to manufacture, they have to be refrigerated, and they often have to be injected because they cannot go through the gastrointestinal tract without being degraded,” explains Alejandra Tomas, a cell biologist at Imperial College London who studies the cellular receptor GLP-1 drugs target. That’s all because they consist of peptides, or long chains of amino acids.
A small-molecule version of the therapy, on the other hand, could be given as a daily pill and would be much cheaper to produce. Companies including Eli Lilly, Pfizer, and Roche have launched clinical trials of such compounds. Results from Lilly’s first phase 3 trial of its oral drug are expected later this year. But Pfizer announced this week it was halting development of its candidate after signs of liver injury in a trial participant.
The candidates furthest along in development activate the same receptors as peptide drugs do, in much the same way. But several firms are exploring more innovative small molecules that target different sites on those receptors—and could lead to even more effective treatments with fewer side effects.
“In the next 4 or 5 years, this field will mature and more patients ultimately should be able to get these medicines,” says Kyle Sloop, a molecular biologist at Lilly Research Laboratories.
By mimicking a natural hormone, semaglutide and other drugs in its class help regulate blood sugar by increasing insulin secretion from the pancreas in response to glucose, and suppress appetite by slowing down digestion. The first generation of peptide drugs were essentially copies of GLP-1, with modifications to prevent the peptide from quickly degrading once in the body. Novo Nordisk first won U.S. approval for semaglutide to treat type 2 diabetes in 2017. It needed to be injected, but in 2019 the company added a pill form, which includes an absorption-enhancing ingredient that allows the peptide to penetrate the stomach wall. However, it requires a high dose and has to be taken while fasting, with minimal liquid.
A small molecule, besides being cheaper, is more easily absorbed. But unlike with hormone-mimicking peptides, “you’re essentially starting from scratch” to find an effective compound, says Nick Traggis, founder and CEO of Ambrosia Biosciences.
Typically, companies seeking leads for a small-molecule drug test thousands of compounds in an automated process to find ones that bind to a target receptor in the lab. But because the GLP-1 receptor has a large area of interface with the natural hormone, it’s hard to know which part of this real estate a much smaller compound needs to hit to be effective.
A breakthrough came in 2017, when researchers used cryogenic electron microscopy to work out the 3D molecular structure of the GLP-1 receptor. Researchers can now use computational modeling to simulate how molecules interact with the receptor, Sloop says. That lets them predict the structures most likely to be a fit. “We don’t have to make every single molecule” to test its promise.
Lilly’s leading oral drug candidate, orforglipron, binds to the same area or “pocket” within the GLP-1 receptor as the natural hormone, Sloop and colleagues reported in December 2024. But this smaller compound attaches via different amino acids within the pocket. “Those interactions are very efficient,” and are enough to activate the receptor, Sloop says.
In 2023, the company published results from a phase 2 trial, in which people taking orforglipron once a day showed average body weight reductions of up to 14.7%, roughly comparable to results from semaglutide.
Pfizer’s candidate, danuglipron, emerged after screening 2.8 million compounds against a GLP-1 receptor to which another compound was already bound outside the target site, changing its shape and rendering it more sensitive. This so-called positive allosteric modulator allowed the company to identify some weakly binding molecules it could then modify and improve. It’s an unusual “but smart” approach, says David Gloriam, an expert in computational drug design from the University of Copenhagen.
But in a 14 April announcement, Pfizer revealed it was dropping danuglipron after one person in a dose-optimization study “experienced potential drug-induced liver injury which resolved after discontinuation of danuglipron.” The company said it would continue to pursue a small-molecule treatment for obesity.
A handful of case studies from patients on injectable treatments suggest liver damage could be a very rare side effect of GLP-1 drugs, though they are also under study for potential benefits in certain types of fatty liver disease. “The issue Pfizer saw … seems to be specific” to danuglipron, Traggis says. Ambrosia, which is also developing a small-molecule drug, “remain[s] incredibly bullish on the long-term benefits and safety” of oral GLP-1 therapies, he adds.
Ambrosia and other companies are looking outside the receptor site targeted by the GLP-1 hormone, seeking allosteric sites where a molecule could bind to trigger the receptor’s signaling. Ideally, a drug would preferentially activate a cell’s insulin-regulating pathway but avoid a second type of signaling that has been linked to side effects seen with the current peptide drugs, such as nausea, vomiting, and diarrhea. “Small molecules targeting novel binding sites might potentially offer improved tolerability,” suggests Jeff Finer, CEO of Septerna Therapeutics, “although this remains to be determined.”
Septerna has developed a technique that places the GLP-1 receptor in an artificial, membranelike structure, allowing researchers to see its binding and signaling behavior in action. The technique has enabled them to identify a new binding site—and find promising small molecules that can bind to it.
Like peptide drugmakers, teams pursuing small molecules hope to supercharge the treatments’ effectiveness by targeting not only GLP-1 receptors, but also others involved in glucose metabolism, such as glucagon receptors and the glucose-dependent insulinotropic polypeptide (GIP) receptor. Lilly’s peptide drug, tirzepatide (marketed as Zepbound or Mounjaro), which targets both GLP-1 and GIP receptors, reduced weight more than semaglutide in a head-to-head trial; it won U.S. approval in 2022 for diabetes and in 2023 for weight loss. Other multireceptor-targeting drugs are in clinical trials.
Developing a small molecule that hits multiple receptors is “technically challenging,” Sloop says, “but there is evidence that it can happen.” Septerna’s novel binding pocket may help identify such molecules, Finer says, because 80% to 90% of the pocket’s genetic sequence is shared among GLP-1, GIP, and the glucagon receptor. But there remains uncertainty about how these additional receptors work and how they should be drugged—for example, there’s a debate about whether it’s more beneficial to activate GIP receptors or block them. “As good as we all are at chemistry, we’re all still learning the biology,” Traggis says.
At Ambrosia, Traggis and his colleagues are developing separate small-molecule drugs for each receptor. They plan to develop these into multiple pills, “where you can then tailor the dosing to each person,” a strategy that may help minimize side effects, Traggis says.
In the next 5 years, some researchers predict small-molecule drugs that are cheaper and easier to administer than their peptide cousins will have a shot at similar blockbuster status. But Finer suspects these options will coexist with the injectables, not supplant them. “The enormity of the patient population affected by obesity and diabetes—estimated to be more than 800 million people worldwide—suggests there’s room for multiple therapeutic approaches.”