The pancreas could make its own GLP-1, permanently
In laboratories far from the clinic, scientists are attempting something quietly radical: teaching the body to heal itself by rewriting the instructions inside its own cells. Researchers have developed gene therapy techniques that reprogram pancreatic cells to produce GLP-1 — the hormone that blockbuster drugs like Ozempic merely mimic — raising the possibility that a single intervention could one day replace a lifetime of injections. The work, still confined to animal models and early tissue studies, sits at the intersection of humanity's oldest medical ambition and its newest molecular tools: not to treat illness endlessly, but to render treatment unnecessary.
- The global obesity crisis, fueled by ultra-processed foods and sedentary lives, has made GLP-1 drugs like Ozempic enormously popular — yet their side effects and need for indefinite use leave millions searching for something more lasting.
- Two biotech companies are now racing down divergent paths: Fractyl Health uses a modified virus to deliver DNA directly into pancreatic cells, while RenBio employs electrical pulses to slip genetic instructions into muscle tissue — both coaxing the body into producing its own GLP-1.
- Mouse studies are turning heads, with animals losing 15 to 20 percent of their body weight and resisting fat gain even on high-fat diets — results that, in the cautious language of science, are described as striking.
- Experts are urging restraint: reprogramming the pancreas in humans could be irreversible, and the long-term consequences of sustained GLP-1 production on a vital organ remain entirely unknown.
- Human trials remain years away, leaving this moment as one of genuine scientific promise shadowed by the honest acknowledgment that proof of concept in mice is a long road from a safe and available therapy.
Ozempic has become a cultural shorthand for weight loss — a weekly injection that suppresses appetite and steadies blood sugar, now used by millions far beyond its original diabetic population. It works, but it demands ongoing commitment: refills, needles, side effects. Researchers are now asking whether the pancreas itself could be taught to do what the drug does.
Two biotech companies are pursuing that question through gene therapy. Fractyl Health's approach, called Rejuva, uses a modified virus to ferry DNA instructions into pancreatic cells, enabling them to manufacture GLP-1 independently. In mouse studies, treated animals lost roughly 20 percent of their body weight in three weeks, and normal-weight mice fed a high-fat diet afterward neither gained weight nor developed blood sugar problems. The company's CEO argues that local delivery to the pancreas would require only a very low viral dose, potentially improving safety.
RenBio takes a different route entirely — mixing DNA into a saline solution, injecting it into muscle, then using electrical pulses to open cell membranes and allow the genetic material inside. Mice on this protocol lost around 15 percent of their body weight and held that loss for at least a year, with improved blood sugar control alongside.
The scientific community is watching carefully, and carefully is the operative word. Gene therapy of this kind, if it translates to humans, would likely be permanent — pancreatic cells, once reprogrammed, would produce GLP-1 indefinitely. What that means for the organ over decades, and whether unforeseen complications might emerge, remains unknown. The pancreas governs blood sugar regulation, and altering it — even toward a beneficial end — carries risks that only human trials can begin to illuminate. Those trials are still years away.
What exists today is a compelling proof of concept and a genuine advance in understanding how cellular function can be rewritten. What does not yet exist is any certainty that this will work safely in people, or when it might reach them. The research is real; the distance between here and a clinic is also real.
Ozempic has become shorthand for weight loss in recent years—a weekly injection that suppresses appetite and steadies blood sugar, borrowed from diabetes treatment and repurposed by millions seeking to shed pounds. The drug works. It also comes with side effects, and it requires discipline: another appointment, another needle, another prescription to refill. But what if your pancreas could make the hormone itself?
Researchers say they've found a way to reprogram pancreatic cells to produce GLP-1, the very hormone that Ozempic mimics, without the need for ongoing injections. The work is early—mostly confined to lab studies of human tissue and mouse models—but the implications are substantial enough that multiple biotech companies are racing to develop the approach. If it works in humans, it could fundamentally change how we treat obesity and diabetes, shifting from a lifetime of medication to a single intervention that teaches the body to heal itself.
The context matters. Obesity has become the leading malnutrition problem among young people worldwide, surpassing undernourishment. Ultra-processed foods and sedentary living have created a public health crisis that GLP-1 drugs have helped address, but imperfectly. These medications are effective at producing weight loss—sometimes dramatic weight loss—yet they carry their own complications and require indefinite use. The appeal of a permanent solution is obvious.
Fractyl Health, a biotech company, is pursuing what amounts to genetic reprogramming. Their approach, called Rejuva, uses a modified virus as a delivery vehicle. The virus carries DNA instructions that, once inserted into pancreatic cells, teach those cells to manufacture GLP-1 on their own. In mouse studies, the results were striking: roughly 20 percent weight loss in three weeks. Normal-weight mice given the treatment and then fed a high-fat diet neither gained weight nor developed blood sugar problems. Harith Rajagopalan, the company's cofounder and CEO, told CNN the approach outperformed semaglutide itself. Because the treatment is delivered locally to the pancreas, Rajagopalan argued, only a very low dose of the viral vector would be needed, which he suggested could improve safety in human patients.
A competitor, RenBio, is taking a different path. Rather than using a virus, they mix DNA instructions into a saline solution and inject it directly into muscle tissue. Electrical pulses then open the cells' membranes, allowing the genetic material to enter. The cells then use their own machinery to manufacture GLP-1. In mice, this approach produced roughly 15 percent weight loss—still substantially more than placebo—and the animals maintained that loss for at least a year while also showing improved blood sugar control.
But experts are sounding caution. Gene therapy of this kind, if it works in humans, would likely be irreversible. Once the pancreatic cells are reprogrammed, they would continue producing GLP-1 indefinitely. The long-term consequences of that—what happens to the pancreas over decades, whether sustained GLP-1 production might cause unforeseen complications—remain unknown. The pancreas is responsible for blood sugar regulation, and tampering with it, even in service of a beneficial outcome, carries risks that won't be clear until human trials are conducted. Those trials are still years away.
What we have now is proof of concept in animals and a genuine scientific breakthrough in understanding how to reprogram cellular function. What we don't have is any certainty about whether this will work safely in people, or when it might become available. The research is compelling, but it is also, by any honest measure, still in its infancy.
Citações Notáveis
It outperformed semaglutide— Harith Rajagopalan, cofounder and CEO of Fractyl Health, on their Rejuva approach in mouse models
A Conversa do Hearth Outra perspectiva sobre a história
So the pancreas can be taught to make its own GLP-1? How does that actually work?
A virus or electrical pulse delivers genetic instructions into pancreatic cells. Those cells then read the instructions and start manufacturing GLP-1 on their own, the same hormone that Ozempic mimics.
And in the animal studies, what happened?
Mice lost 15 to 20 percent of their body weight within weeks or maintained that loss for a year. Their blood sugar improved. They didn't regain the weight even on a high-fat diet.
That sounds almost too good. What's the catch?
The treatment would likely be permanent. Once you reprogram those cells, they keep making GLP-1 for life. We don't know what that means for the pancreas over decades, or what other effects might emerge.
So it's irreversible?
Yes. And the pancreas controls blood sugar regulation. Altering it, even for a good reason, is a significant intervention. We need human trials to understand the real risks.
How far away is that?
Years, probably. Right now we have compelling animal data and the science is sound, but human safety and efficacy studies take time. This is still very early.