For the 35 million American adults who are battling obesity, the age-old advice to “improve diet and exercise” often falls short. And surgical interventions that reduce stomach size—although effective—can prove risky.

But there may be another alternative: flooding the body with a protein that makes an individual prefer low-fat food and feel full longer, and that activates neurons responsible for regulating the body’s energy intake. Such a simple fix may sound too good to be true, but the approach has shown promise in experiments with mice, rats and monkeys. If the therapy can be improved and shown to work well in humans, it could be an entirely new way to help battle obesity, diabetes and related conditions. Currently there are a few federally approved weight loss and insulin control agents on the market, but they appear to work along different pathways than the new therapy does. They can also cause serious side effects, and are still not as effective as surgical weight-loss interventions.

In new work spearheaded by pharmaceutical company Amgen, researchers took advantage of a protein linked with multiple metabolic disorders to try to make thinner, healthier animals. In multiple species, lean creatures appear to have naturally higher concentrations of the protein GDF15 than their stouter counterparts, the Amgen team noted. So they tried to boost levels of the compound in obese animals via gene therapy—giving rodents injections of a form of the gene that would cause their bodies to produce more of the protein than they would have otherwise. In the short-term that seemed to help make the animals healthier, but their bodies cleared the substance too quickly for it to produce a lasting effect. So the researchers decided to bypass the gene therapy approach and engineer two stable, longer-lasting forms of GDF15 that they injected directly into the animals.

Weekly injections of either of these engineered molecules slimmed down rodents and monkeys without causing any apparent serious side effects, the scientists report Wednesday in Science Translational Medicine. After about a month of weekly treatments with one of the two engineered forms of GDF15, treated rodents sometimes lost as much as 17 or 24 percent of their body weight, depending on what molecule they received, says senior author Murielle Véniant, who studies metabolic disorders as a lead researcher at Amgen. The treatment also helped monkeys shave pounds. In the most extreme cases, the treated monkeys lost 5 or 10 percent of their body weight during that same time period, she says. Meanwhile, untreated animals continued to gain weight. “We were very surprised to see improvements in body weight, glucose and insulin levels as well as triglycerides,” she adds.

Exactly how these weekly GDF15 treatments work remains murky. As a whole, the findings suggest the proteins act on the gut–brain axis—the collection of two-way nerve and chemical pathways that communicate between the digestive tract and the brain—and that GDF15 might be a clinical target for metabolic disorders. The specifics have yet to be ironed out, but the Amgen team has uncovered a few clues in their rodent studies: They found treated animals had increased activation of certain neurons in the brain that detect blood sugar, and this may have helped them sense when it was time to stop eating. The treated rodents’ stomachs also took longer to empty out—suggesting they may have felt full longer and that GDF15 may block the transmission of signals from the vagus nerve, which connects the brain and digestive tract. Finally they found animals receiving the therapy seemed to prefer a lower-fat diet than those that did not—although the researchers cannot yet explain why taste preferences changed.

Some experts were cautiously optimistic about the findings. “This group really seemed to go a long way toward identifying what may be an interesting new approach to treatment, and that’s fantastic,” says Paul Kenny, chair of the Department of Neuroscience at Mount Sinai Health System, who was not involved with the Amgen work. This is exciting, he adds, “but the key will be [determining] if it’s safe in humans.” If this work makes it into human trials, he notes, two areas to watch will be whether there is any significant muscle mass loss with treatment—because treated animals lost some lean mass alongside their fat—and if the therapy causes any serious side effects such as depression or mood changes, which have been concerns with some other proposed weight-loss products. Véniant declined to comment on Amgen’s next steps or potential timeline for clinical trials in humans but says the company plans to continue its work in this area. Clearly, she says, “GDF15 is a very interesting protein.”