At a lab meeting of Fred “Rusty” Gage’s group at the Salk Institute for Biological Studies in the mid-1990s, the neuroscientist told his team that he wanted to determine whether new neurons are produced in the brains of adult humans. At the time, adult neurogenesis was well established in rodents, and there had been hints that primate brains also spawned new neurons later in life. But reports of neurogenesis in the adult human brain were sparse and had not been replicated. Moreover, the experiments had relied primarily on autoradiography, which revealed images of cell division but did not follow the fate of new cells, so researchers couldn’t be sure if they really became mature neurons.
Gage’s group, which included clinicians, was familiar with the use of bromodeoxyuridine (BrdU) to monitor the progression of certain cancers. BrdU is an artificial nucleoside that can stand in for thymidine (T) during DNA replication. As cells duplicate their genomes just before they divide, they incorporate BrdU into their DNA. To assess tumor growth, physicians inject the nucleoside substitute into a patient’s bloodstream, then biopsy the tumor and use an antibody to stain for BrdU. The number of BrdU-labeled cells relative to the total number of cells provides an estimate of how quickly the cancer is growing. “If that nucleotide is labeled in such a way that [we] can identify it, you can birthdate individual cells,” Gage says.
Because BrdU goes everywhere in the body, Gage and his colleagues figured that in addition to labeling the patients’ tumors, the artificial base would also label the cells of the brain. If the researchers could get their hands on brain specimens from patients who’d been injected with BrdU, perhaps it would be possible to see new brain cells that had been generated in adults. With a second antibody, they could then screen for cell-type markers to determine if the new cells were mature neurons. “If you can . . . [use] a second antibody to identify the fate of a cell, then that’s pretty definitive,” Gage says.
As Gage’s fellows and postdocs left the lab and got involved in clinical trials involving BrdU injections, they began to keep an eye out for postmortem brain samples that Gage could examine. In 1996, one of them came through. Neurologist Peter Eriksson, who at the time was working at the Sahlgrenska University Hospital in Gothenburg, Sweden, began sending Gage samples from the brains of deceased patients. Every few months, a new sample arrived. And while waiting for the next delivery, Gage and his team were “getting fresh tissue from the coroner’s office to practice staining fresh tissue,” he says, “so that when we got these valuable brains we could see [what] they were doing.”
Soon enough, a clear picture emerged: the human hippocampus, a brain area critical to learning and memory and often the first region damaged in Alzheimer’s patients, showed evidence of adult neurogenesis. Gage’s collaborators in Sweden were getting the same results. Wanting to be absolutely positive, Gage even sent slides to other labs to analyze. In November 1998, the group published its findings, which were featured on the cover of Nature Medicine.1
“When it came out, it caught the fancy of the public as well as the scientific community,” Gage says. “It had a big impact, because it really confirmed [neurogenesis occurs] in humans.”
Fifteen years later, in 2013, the field got its second (and only other) documentation of new neurons being born in the adult human hippocampus—and this time learned that neurogenesis may continue for most of one’s life.2 Neuroscientist Jonas Frisén of the Karolinksa Institute in Stockholm and his colleagues took advantage of the aboveground nuclear bomb tests carried out by US, UK, and Soviet forces during the Cold War. Atmospheric levels of 14C have been declining at a known rate since such testing was banned in 1963, and Frisén’s group was able to date the birth of neurons in the brains of deceased patients by measuring the amount of 14C in the cells’ DNA.
“What we found was that there was surprisingly much neurogenesis in adult humans,” Frisén says—a level comparable to that of a middle-aged mouse, the species in which the vast majority of adult neurogenesis research is done. “There is hippocampal neurogenesis throughout life in humans.”
But many details remain unclear. How do newly generated neurons in adults influence brain function? Do disruptions to hippocampal neurogenesis play roles in cognitive dysfunction, mood disorders, or even psychosis? Are there ways to increase levels of neurogenesis in humans, and might doing so be therapeutic? Researchers are now seeking to answer these and other questions, while documenting the extent and function of adult neurogenesis in mammals.
Breaking the mold
© LAURIE O'KEEFEIn the early 1960s, MIT neurobiologist Joseph Altman used a hypodermic needle to induce lesions in rat brains, while simultaneously injecting tritiated thymidine, a radioactive form of the nucleoside commonly used for tracking DNA synthesis and cell proliferation. He found evidence of new brain cells that had been born at the time of injection, including some neurons and their neuroblast precursors.3
Researchers were immediately skeptical of the results. Long-standing theory held that neurons in the brain that had been damaged or lost could not be replaced; Altman was suggesting the opposite. “They were really good, solid indications, but it was such a strong dogma that neurons couldn’t be generated in the adult brain,” says Frisén. “It wasn’t really until the ’90s, when new techniques came along, [that researchers] showed, yes, indeed, new neurons are added in the rodent brain.”
Those new techniques included BrdU, as well as neuron-specific protein markers and confocal imaging, which together enabled researchers to identify the newly generated cells. Multiple studies subsequently confirmed that neurogenesis occurs in limited regions of the rodent brain, specifically in the olfactory bulb and the dentate gyrus region of the hippocampus. (See illustration.) Research also revealed that the rate of neurogenesis decreases with stress, depression, and anxiety, but increases with exercise and enrichment.
© LAURIE O'KEEFE
“The field grew enormously at this point,” Gage says, and its focus began to shift from whether new neurons were being produced—they were—to whether those cells formed connections with existing networks to become functional—they do. Turns out, “these newly born cells have 5,000 synapses on their dendrites,” Gage says—well within the range of other neurons in the brain.
But would those rodent results hold up in primates? All signs pointed to yes. In March 1998, Princeton University’s Elizabeth Gould and colleagues found evidence of neurogenesis in the dentate gyrus of adult marmoset monkeys—and the researchers determined that the rate of cell proliferation was affected by stress, just as in rodents.4 Six months later, Gage’s group published its findings based on the clinical samples of human brain tissue. “It was a surprise to me, and I think to most people,” Frisén says. And the point was hammered home with the Frisén group’s analysis of 14C in human brain samples.
“The human evidence now unequivocally suggests that the dentate gyrus in humans undergoes turnover in our lifetime,” says Amar Sahay of Harvard University. “It really begs the question what the functions are of these adult-born neurons.”
Young and excitable
The first step in understanding the function of the new neurons in the adult brain was to characterize the cells themselves. In the late 1990s and early 2000s, researchers delved into the cell biology of neurogenesis, characterizing the populations of stem cells that give rise to the new neurons and the factors that dictate the differentiation of the cells. They also documented significant differences in the behavior of young and old neurons in the rodent brain. Most notably, young neurons are a lot more active than the cells of established hippocampal networks, which are largely inhibited.5,6
“For a period of about four or five weeks, while [the newborn neurons] are maturing, they’re hyperexcitable,” says Gage. “They’ll fire at anything, because they’re young, they’re uninhibited, and they’re integrating into the circuit.”
To determine the functional role of the new, hyperactive neurons, researchers began inhibiting or promoting adult neurogenesis in rodents by various means, then testing the animals’ performance in various cognitive tasks. What they found was fairly consistent: the young neurons seemed to play a role in processing new stimuli and in distinguishing them from prior experiences. For example, if a mouse is placed in a new cage and given time to roam, then subjected to a mild shock, it will freeze for about 40 seconds the next time it is placed in that same environment, in anticipation of a shock. It has no such reaction to a second novel environment. But in an enclosure that has some features in common with the first, fear-inducing cage, the mouse freezes for 20 seconds before seemingly surmising that this is not the cage where it received the initial shock. Knock out the mouse’s ability to produce new neurons, however, and it will freeze for the full 40 seconds. The brain is not able to easily distinguish between the enclosures.
COURTESY OF FRED H. GAGEJASON SNYDERCOURTESY OF FRED H. GAGEThis type of assessment is called pattern separation. While some researchers quibble over the term, which is borrowed from computational neuroscience, most who study hippocampal neurogenesis agree that this is a primary role of new neurons in the adult brain. “While probably five or six different labs have been doing this over the last four or five years, basically everybody’s come to the same conclusion,” Gage says.
The basic idea is that, because young neurons are hyperexcitable and are still establishing their connectivity, they are amenable to incorporating information about the environment. If a mouse is placed in a new cage when young neurons are still growing and making connections, they may link up with the networks that encode a memory of the environment. Just a few months ago, researchers in Germany and Argentina published a mouse study demonstrating how, during a critical period of cellular maturation, new neurons’ connections with the entorhinal cortex, the main interface between the hippocampus and the cortex, and with the medial septum change in response to an enriched environment.7
“The rate at which [new neurons] incorporate is dependent upon experience,” Gage says. “It’s amazing. It means that the new neurons are encoding things when they’re young and hyperexcitable that they can use as feature detectors when they’re mature. It’s like development is happening all the time in your brain.”
Adding support to the new neurons’ role in pattern separation, Sahay presented findings at the 2014 Society for Neuroscience conference that neurogenesis spurs circuit changes known as global remapping, in which overlap between the populations of neurons that encode two different inputs is minimized.8 “We have evidence now that enhancing neurogenesis does enhance global remapping in the dentate gyrus,” says Sahay. “It is important because it demonstrates that stimulating neurogenesis is sufficient to improve this very basic encoding mechanism that allows us to keep similar memories separate.”
Pattern separation is likely not the only role of new neurons in the adult hippocampus. Experiments that have suppressed neurogenesis in adult rats have revealed impairments in learning in a variety of other tasks. More broadly, “we think it has to do with the flexibility of learning,” says Gerd Kempermann of the Center for Regenerative Therapies at the Dresden University of Technology in Germany.
Last year, for example, neuroscientist Paul Frankland of the Hospital for Sick Children in Toronto and his colleagues found evidence that newly generated neurons play a role in forgetting, with increased neurogenesis resulting in greater forgetfulness among mice.9 “If you think about what you’ve done today, you can probably remember in a great deal of detail,” he says. “But if you go back a week or if you go back a month, unless something extraordinary happened, you probably won’t remember those everyday details. So there’s a constant sort of wiping of the slate.” New hippocampal neurons may serve as the “wiper,” he says, “cleaning out old information that, with time, becomes less relevant.”
Conversely, Frankland’s team found, suppressing neurogenesis seems to reinforce memories, making them difficult to unlearn. “We think that neurogenesis provides a way, a mechanism of living in the moment, if you like,” he says. “It clears out old memories and helps form new memories.”
Neurogenesis in the clinic
While studying the function of hippocampal neurogenesis in adult humans is logistically much more difficult than studying young neurons in mice, there is reason to believe that much of the rodent work may also apply to people—namely, that adult neurogenesis plays some role in learning and memory, says Kempermann. “Given that [the dentate gyrus] is so highly conserved and that the mechanisms of its function are so similar between the species—and given that neurogenesis is there in humans—I would predict that the general principle is the same.”
And if it’s true that hippocampal neurogenesis does contribute to aspects of learning involved in the contextualization of new information—an ability that is often impaired among people with neurodegenerative diseases—it’s natural to wonder whether promoting neurogenesis could affect the course of Alzheimer’s disease or other human brain disorders. Epidemiological studies have shown that people who lead an active life—known from animal models to increase neurogenesis—are at a reduced risk of developing dementia, and several studies have found reduced hippocampal neurogenesis in mouse models of Alzheimer’s. But researchers have yet to definitively prove whether neurogenesis, or lack thereof, plays a direct role in neurodegenerative disease progression. It may be that neurogenesis has “nothing to do with the pathology itself, but [with] the ability of our brain to cope with it,” says Kempermann.
Either way, the research suggests that “the identification of pro-neurogenic compounds would have a therapeutic impact on cognitive dysfunction, specifically, pattern separation alterations in aging and early stages of Alzheimer’s disease,” notes Harvard’s Sahay. “There’s a growing list of genes that encode secreted factors or other molecules that stimulate neurogenesis. Identifying compounds that harness these pathways—that’s the challenge.”
The birth of new neurons in the adult hippocampus may also influence the development and progression of mood disorders. Several studies have suggested that reduced neurogenesis may be involved in depression, for instance, and have revealed evidence that antidepressants act, in part, by promoting neurogenesis in the hippocampus. When Columbia University’s René Hen and colleagues short-circuited neurogenesis in mice, the animals no longer responded to the antidepressant fluoxetine.10 “It was a very big surprise,” Hen says. “The hippocampus has really been always thought of as critical for learning and memory, and it is, but we still don’t understand well the connection to mood.”
Adult neurogenesis has also been linked to post-traumatic stress disorder (PTSD). While it is perhaps less obvious how young neurons might influence the expression of fear, Sahay says it makes complete sense, given the emerging importance of neurogenesis in distinguishing among similar experiences. “In a way, the hippocampus acts as a gate,” he says, with connections to the amygdala, which is important for processing fear, and the hypothalamus, which triggers the production of stress hormones, among other brain regions. “It determines when [these] other parts of the brain should be brought online.” If new neurons are not being formed in the hippocampus, a person suffering from PTSD may be less able to distinguish a new experience from the traumatic one that is at the root of his disorder, Sahay and his colleagues proposed earlier this year.11 “We think neurogenesis affects the contextual processing, which then dictates the recruitment of stress and fear circuits.”
Of course, the big question is whether researchers might one day be able to harness neurogenesis in a therapeutic capacity. Some scientists, such as Hongjun Song of Johns Hopkins School of Medicine, say yes. “I think the field is moving toward [that],” he says. “[Neurogenesis] is not something de novo that we don’t have at all—that [would be] much harder. Here, we know it happens; we just need to enhance it.”
- P.S. Eriksson et al., “Neurogenesis in the adult human hippocampus,” Nat Med, 4:1313-17, 1998.
- K.L. Spalding et al., “Dynamics of hippocampal neurogenesis in adult humans,” Cell, 153:1219-27, 2013.
- J. Altman, “Are new neurons formed in the brains of adult mammals?” Science, 135:1127-28, 1962.
- E. Gould et al., “Proliferation of granule cell precursors in the dentate gyrus of adult monkeys is diminished by stress,” PNAS, 95:3168 -71, 1998.
- C. Schmidt-Hieber et al., “Enhanced synaptic plasticity in newly generated granule cells of the adult hippocampus,” Nature, 429:184-87, 2004.
- S. Ge et al., “A critical period for enhanced synaptic plasticity in newly generated neurons of the adult brain,” Neuron, 54:559-66, 2007.
- M. Bergami et al., “A critical period for experience-dependent remodeling of adult-born neuron connectivity,” Neuron, 85:710-17, 2015.
- K. McAvoy et al., “Rejuvenating the dentate gyrus with stage-specific expansion of adult-born neurons to enhance memory precision in adulthood and aging,” Soc Neurosci, Abstract DP09.08/DP8, 2014.
- K.G. Akers et al., “Hippocampal neurogenesis regulates forgetting during adulthood and infancy,” Science, 344:598-602, 2014.
- L. Santarelli et al., “Requirement of hippocampal neurogenesis for the behavioral effects of antidepressants,” Science, 301:805-09, 2003.
- A. Besnard, A. Sahay, “Adult hippocampal neurogenesis, fear generalization, and stress,” Neuropsychopharmacology, doi:10.1038/npp.2015.167, 2015.