It was a mind-boggling observation. Hook up the circulatory systems of a young mouse and an old one, and the elderly animal seems to be rejuvenated. Since 2005, a handful of research labs have been hotly pursuing the molecules responsible for this effect, first found in the 1950s, hoping to harness them to slow or reverse aging in people. One in particular stood out: a protein found in young blood known as GDF11. In several high-profile papers, two of them published last year in Science, a Harvard University team reported that the protein declines in older animals and that replacing it rebuilds muscles, brain, and the heart. But work described this week by a team at the Novartis research center challenges GDF11’s rejuvenating powers.
The Novartis group does not question that young blood renews old mice. But they say the Harvard group’s explanation is wrong. Their paper, published online today in Cell Metabolism, casts doubt on the assays used in the earlier research and suggests that GDF11 actually inhibits muscle regeneration. “The whole premise is incorrect,” says stem cell researcher Michael Rudnicki of the Ottawa Hospital Research Institute, who co-wrote a commentary accompanying the paper. Others are more cautious, but agree that the new work undermines part of the original GDF11 claim. “GDF11 does not go down with age,” says Thomas Rando, a biologist at Stanford University in Palo Alto, California.
Harvard stem cell biologist Amy Wagers, who led much of the original work, says the Novartis data on GDF11 levels are not persuasive. And although they “appear to conflict with” her group’s, “we are actually very excited to see the Novartis data,” she says. “We remain convinced that at least one form of GDF11 declines in blood with age and that maintaining GDF11 levels in an appropriate physiological range is essential for muscle health.”
Wagers began exploring the many ways in which joining the circulatory systems of mice—a procedure known as parabiosis—affects aging as a postdoc more than a decade ago, with Rando and others. In 2013, her group and cardiologist Richard Lee’s lab at Brigham and Women’s Hospital in Boston reported in Cell that levels of GDF11 in the blood fell as mice aged and that, like young-old parabiosis, restoring GDF11 through injections partially reversed age-related thickening of the heart. In Science last year, she and collaborators, including Lee and Harvard neural stem cell researcher Lee Rubin, reported that GDF11 also nourished blood vessel and neuron growth in old mice’s brains. In a second Science paper, they reported that GDF11 spurred healing from a muscle injury in older mice. Aged mice receiving GDF11 did better on strength and running tests.
Some experts were flummoxed by the muscle paper, because GDF11 is a close cousin of myostatin, a well-studied protein that controls muscle growth. Animals and people lacking myostatin develop huge, bulging muscles; too much of it hinders muscle regeneration. How, then, could a very similar protein have the opposite effect?
Among the skeptics was David Glass of Novartis Institutes for BioMedical Research in Cambridge, Massachusetts, who helped develop a myostatin-blocking drug for muscular atrophy. When his group tested GDF11 levels in rats with both of the assays Wagers had used, a proteomics assay and a commercial antibody, they could not distinguish between GDF11 and myostatin. Using more specific tests, they found that GDF11 levels actually trend upward with age in rat and human blood and GDF11 mRNA levels rise in rat muscle with age.
The Novartis group also tested GDF11’s effects on muscle regeneration. When they treated a young mouse with GDF11 and damaged its leg muscle with snake venom toxin, a common experiment, they found that muscle regeneration was impaired. “The bottom line is that this molecule [GDF11] seems to be harmful to muscle,” Glass says.
Wagers sticks by her data, noting that her group’s Science paper also found a drop in GDF11 with age using a different antibody that distinguished GDF11 from myostatin. And she says the Glass team’s injury experiment cannot be compared to hers because they used young animals and a dose of GDF11 three times higher. (Glass did this in part because he did not see any effect in old mice at the dose Wagers used.) The signaling pathway in which GDF11 lies “is notoriously dose-sensitive,” and low and high doses can have different or opposite effects, she says. Moreover, she says, the Novartis team’s muscle regeneration test was not comparable to hers—the Harvard team made the injury by freezing tissue, which is less likely than a toxin to kill muscle stem cells needed for regeneration.
Wagers says new data from her group will show that “there is a very compelling biological explanation for the apparent discrepancies.” One of her collaborators, Lee, agrees. But Rubin is more cautious: “Obviously, this report has to be taken seriously.” Although the Novartis result does not challenge a second claimed benefit of GDF11, to the brain, “we’re designing a series of experiments to convince ourselves that what we see in the brain is real,” says Rubin, who led that study.
Others say that even if the new finding is correct, it may not contradict at least some of the benefits Wagers and others reported, says molecular biologist Se-Jin Lee of Johns Hopkins University in Baltimore, Maryland, who studies myostatin. He notes that GDF11’s effects in the body are likely complex and haven’t been studied in detail. “There’s still a lot to be sorted out.”
*Correction, 21 May, 3:29 p.m.: An earlier version of this article incorrectly stated that a 2014 Science paper reported that mice injected with GDF11 had an improved sense of smell.