Every science writer loves a good challenge to dogma. I wish I had been in the working world in the spring of 1992, when one such intellectual overhaul happened in neuroscience. The dogma: Neurons, unlike most of the body’s cells, can’t be replenished. You’re born with just 100 billion of them and you better use them wisely. The challenge: Samuel Weiss and Brent Reynolds reported in Science that brain tissue taken from adult mice could be chemically coaxed into making new neurons.
“It left us speechless,” Weiss told the New York Times. Everybody else was pretty stunned, too. Over the next six years, other researchers confirmed that this so-called neurogenesis happens in the adult hippocampus of many animals, including tree shrews, marmosets, Old World monkeys and people. Today, more than two decades since the splashy Science report, adult neurogenesis is a bona fide subfield, with hundreds of labs studying it around the world.
But after all this time, researchers still don’t really know what it’s for. Studies have uncovered a wide variety of environmental stimuli — what you might think of as inputs — that affect neurogenesis in the dentate gyrus, a part of the hippocampus. Running and antidepressants can ramp up neurogenesis, for example, while stress, social isolation, sleep deprivation and aging can shut it down. Scientists have also looked at the outputs of neurogenesis, showing that a boost of new neurons may be important for exploratory behavior and certain kinds of learning, such as navigating a new space. But how do the inputs lead to the outputs?
“I like to think of the dentate as an association machine,” says Sam Pleasure, a neuroscientist at the University of California, San Francisco. All day long, he says, we’re walking around the world trying to associate various sensations and emotions — big dog with fangs, small screaming toddler, perilous traffic intersection — so that we can remember them later. “All these stimuli are happening and converge on this circuit, and they somehow affect how new neurons are recruited into the circuit, and that ends up coming out as the ability to form new memories.” But how it all works on the molecular level is a black box.
Two papers published today in Cell Stem Cell open that box a little bit. They identify molecular inhibitors — what Pleasure calls “wet blankets” — that turn off neurogenesis in certain contexts.
Both studies concern a famous biological network, called Wnt, that’s crucial for the development of the nervous system, including setting up the dentate gyrus. In 2005, researchers showed that Wnt signals also drive neurogenesis in adulthood. Networks like Wnt are super-complicated, though, and controlled by dozens of molecular interactions.
In one of the new reports, scientists from Johns Hopkins University focused on a protein called ‘secreted frizzled-related protein 3′, or SFRP3, which litters the dentate gyrus. The protein can bind to Wnt outside of the cell and thus block Wnt signaling. The study found that mice lacking SFRP3 have increased neurogenesis. But more exciting, the study showed that when normal mice exercise (voluntary running on a wheel), the level of SFRP3 in their brains goes down and neurogenesis goes up. In other words, SFRP3 is a neurogenesis wet blanket that can be pulled off with exercise.
The second study focused on a protein called Dickkopf-1, or DKK1, which has an interesting history. In the first description of it, in 1998, a German team led by Christof Niehrs showed that in tadpoles, over-expressing DKK1 leads to multiple heads, whereas knocking it out leads to no head at all. (These findings explain its name: In German, dick means fat and kopf means head.) Over the next decade, researchers showed that DKK1 is also a Wnt inhibitor and that, like SFRP3, its expression throws a brake on neurogenesis.
The new work uncovers a piece of the DKK1 mechanism that could have big clinical implications for Alzheimer’s and other disorders that involve age-related dementia. Niehrs and collaborators showed that in mice, DKK1 expression increases with age, leading to a decrease in neurogenesis. Mice engineered to lack DKK1 in neural stem cells, meanwhile, show some amazing behaviors in old age. At 18 months old, normal mice are well into senescence and all that comes with it, including memory loss and anxiety. But old DKK1 mutants perform tasks of working memory and memory consolidation just as well as younger mice, and have abnormally low anxiety to boot.
“For us this was very astonishing,” says lead investigator Ana Martin-Villalba, head of Molecular Neurobiology at the German Cancer Research Center in Heidelberg. The findings suggest, she says, that drugs that block DKK1 might work against age-related dementia. (Some DKK1 blockers are already being tested in clinical trials for osteoporosis, but they can’t cross the blood-brain barrier.)
But before going forward in the clinic, Martin-Villalba wants to answer a much more fundamental question: What’s the point of DKK1 in the first place? “If you don’t have it and you are happier and you have less cognitive decline, why should you have it?” she says. “At the moment I cannot tell you why.”
Pleasure, who wasn’t involved in either study, agrees that many of these ‘why’ questions need to be answered before thinking about neurogenesis therapies. Why does neurogenesis decrease with age? Why does the brain have any brakes on neurogenesis, and why so many different brakes? These mechanistic studies “are filling in small black boxes,” he says, figuring out the messy middle part between inputs and outputs. “We’re nowhere close to finishing the middle, but these are the reasonable steps.”
Not as sexy as knocking down scientific dogma, maybe, but thrilling all the same.