Making Nuanced Memories

Mice aren’t known for their skill with complicated memory tricks, but they can usually recall their last meal. Once they happen upon food in a laboratory maze, they are pretty good at remembering the location from one trial to the next. In one recent study, though, half the mice got too confused to find their snacks.

All the mice in two groups tested could remember the location of a new reward stashed in a vastly different place from an earlier one. But one group had trouble when the payoff lay just slightly off from a previous spot. Those mice had a good excuse, though: Their brains were incapable of creating new nerve cells, or neurons, in a region important for memory.

In the late ’90s, scientists stunned the research world with the discovery that human adults aren’t stuck with only the neurons they’re born with — an idea long entrenched in neuroscience dogma. In fact, adult brains are getting fresh batches of nerve cells every day. Since that revelation, researchers have been trying to answer a nagging question: What are the new neurons good for?

While it’s now widely accepted that new cells are appearing in a part of the brain that codes and packages memories, the precise function of these newborn brain cells remains unclear. Many researchers are now convinced that new cells are indeed vital for recording memories, but not all forms of memory — just those that tend to get jumbled with other similar ones (such as what you had for lunch yesterday or where you parked your car).

Besides investigating what the cells do, researchers are discovering ways that people could encourage fresh neurons to grow, through diet and lifestyle. One day soon, medical science might offer ways to enhance memory and protect the brain from erosion that comes with age — a goal so fundamental to human existence that the ancient Greeks even worshiped a goddess of memory.

“We are closer to understanding how memories are truly formed and stored in the brain,” says Craig Stark, director of the Center for the Neurobiology of Learning and Memory at the University of California, Irvine. “If we want to try to help get better memories, we’d darn well better know how the system works.”

Enduring birth

New nerve cells in the brain arise from the hippocampus, a sea horse–shaped relay of cells important for learning and memory. In particular, the hippocampus encodes and prepares new memories for storage, then dispatches them to different parts of the brain. In 1998, scientists reported evidence that the human hippocampus is not only a depot for memories, but also a birthplace for neurons — thousands each month. (The olfactory bulb, a brain structure involved in odor perception, also gets new cells via neurogenesis, the formation of new neurons.) Even after more experiments supported the finding, many scientists were slow to accept such a revolutionary idea.

These adult-born nerve cells are now revealing that the science of memory is more complex than ever imagined. Despite the common perception that the nervous system has one central filing cabinet for memories, the ability to remember is a function spread throughout the brain, with events recorded in different ways depending on the kind of information (a phone number or the technique of a golf swing) and how often it’s retrieved (just for the next five minutes or every Saturday).

While the hippocampus and the network around it are vital for saving new experiences, the nursery for nerve cells is restricted to a raisin-sized region of the hippocampus called the dentate gyrus. At any given time, about 3 to 5 percent of the cells in the dentate gyrus are in some stage of growth, says Fred Gage, a neuroscientist at the Salk Institute for Biological Studies in La Jolla, Calif.

New nerve cells begin their lives as what are called “neuronal progenitor cells.” A population of these progenitor cells permanently resides in the dentate. Once they divide and start to mature, new nerve cells do not immediately resemble adult neurons and aren’t connected to the brain’s neural network. During their first month, the cells mature into a kind of highly excitable teenage phase and begin to send feelers into the surrounding brain tissue. At about 16 days, the cells elongate and start to look more like nerve cells in the rest of the brain.

After a couple more weeks, the newly minted cells take on the appearance of mature nerve cells, and after two months, they are indistinguishable from neurons elsewhere in the brain. Once the cells finish maturing, they integrate into the rest of the hippocampus, where they remain for a lifetime. “Most of the dentate gyrus is formed after birth,” Gage says. “A lot of it is formed in the first four years of life. That’s when you’re getting your baseline of memories. Then a low level of neurogenesis persists.”

Memory lane

In part, this continuous crop of fresh cells may keep new memories from disrupting old ones, like increasing the storage capacity on your hard drive. But scientists including Gage believe that the main purpose of these adult-born nerve cells is to encode a kind of memory called pattern separation, which is necessary for the accuracy of memories because it keeps similar experiences from overlapping.

“New neurons are helping to distinguish between events that are close to each other,” Gage says. Let’s say someone offers you a banana. The dentate gyrus records the fact that you’ve just seen a yellow tropical fruit. “It’s like a bar code,” he says. “You put the bar code of the banana into the dentate. It’s coded with lots of information.” When you see another banana, the dentate will determine whether it’s the same one. But if the next fruit is an apple, the dentate doesn’t get excited, Gage says, because the difference is so big you know it’s not déjà vu.

New neurons don’t bother with vastly different information, instead working to separate locations and images that are similar, like one departure gate from the next at the airport. In 2009, Gage and his colleagues described in Science some of the strongest evidence for the theory, with the food-finding experiment comparing how two groups of mice learned to navigate a maze to find a reward. One group had a normal dentate gyrus, and the other was incapable of neurogenesis because researchers had knocked out the dentate with X-rays.

New nerve cells may help to distinguish not only among related locations, but also events occurring close in time. A computer program written by Gage and colleagues to mimic the neural network of the dentate gyrus suggests that the adolescent cells respond easily to a new stimulus. But as those cells mature, they lose their hair trigger for reacting to new events, the team reported in 2009 in Neuron.

Those cells that fire together during their youth are forever linked, Gage proposes, sequencing and connecting events in time when the memories finally crystallize in the brain. That’s one reason, he says, why people often bring up associated memories when trying to recall something (what was I doing when I put down my keys?).

For memory to be accurate, the brain doesn’t record just an image but the entire context, says Raymond Kesner, a psychology professor at the University of Utah in Salt Lake City. “If you try to remember a story, time and place will always be important.”

Work by Kesner and his colleagues has helped scientists understand exactly what kinds of patterns undergo separation (similarities in distance, time or sequence), and how the dentate gyrus works in concert with the rest of the hippocampus. For example, in experiments described in 2008 in Hippocampus, Kesner’s team demonstrated that rats with a disabled dentate gyrus had difficulty remembering the locations of objects, along with the order they were presented, when the distance between the two objects in the test was small. But the rats had no difficulty when the objects were far apart. His experiments support the idea that neurons in the dentate gyrus are vital for remembering the order in which the brain encounters similar objects in space — like the locations of landmarks passed on a road trip.

Much of the evidence for the dentate gyrus’s involvement in pattern separation comes from experiments done in animals and with theoretical models. But Stark and his team have also demonstrated the effect in people. Volunteers were shown images while undergoing functional MRI scans that detected brain activity. First, the participants saw a series of new images. Not surprisingly, the dentate gyrus appeared active — the brain was recording the new images. In another round of the experiment, people were shown a series in which many of the images were unique, but only subtly so (such as a picture of a jack-o’-lantern, followed by a picture of another jack-o’-lantern with a differently carved face). When people could detect the difference, their brains lit up as if the images were completely new. But when people mistook a new image for a previous one, the dentate gyrus did not stir, Stark’s group reported in 2008 in Science.

Using similar tests and technology, Stark and others are now trying to understand why short-term memories become more difficult to capture as people get older, even among adults who remain mentally sharp into their later years. In 2010, in Hippocampus, Stark and his colleagues described evidence of a sluggishness in new neurons that arise in the dentate gyrus of aged brains. In older tissue, the newborn nerve cells appear to require greater contrasts among images and experiences before reacting and capturing a memory. As people age, Stark says, “we seem to be less good about details and specifics.”

Brain food

But researchers are also exploring ways to keep newborn neurons of old age as numerous and eager as those formed in younger years. First, scientists are identifying influences that, in animal studies, appear to decrease neurogenesis in the dentate — including stress, alcohol consumption and, according to a study in Neuroscience Letters in 2010, a high-fat diet. In addition to the enemies of neurogenesis, researchers have identified habits that protect and nurture all vintages of brain cells, including physical activity and some common components of plants.

“Exercise is the strongest neurogenenic stimulus I know of,” says Henriette van Praag of the Neuroplasticity and Behavioral Unit at the National Institute on Aging in Bethesda, Md. Van Praag and her colleagues discovered the role of exercise almost by accident in 1999, when they were testing whether learning increased the rate of neurogenesis. Learning itself didn’t make a difference, but surprisingly, one group of mice did do better: those whose cages had an exercise wheel, and were included in the experiment as one of the comparison groups. “Not only did the exercising animals have more new neurons, they also learned better in a maze,” she says.

The finding launched a series of investigations into the influence of exercise on neurogenesis and learning, with later research finding that even elderly mice (by rodent measures) could improve neurogenesis and memory function by regularly running. Supporting the link with neurogenesis, van Praag and colleagues reported last year in the Proceedings of the National Academy of Sciences that mice that exercise regularly perform better in pattern separation tests. In the experiment, the mice would receive a reward if they touched a certain icon on a computer screen. When the images were placed close together, the sedentary mice had to make almost twice as many attempts to find the payoff as the active mice did. Another 2010 study in mice, described in the journal Cancer Research, found that exercise, by encouraging neurogenesis, might help maintain memory function during a kind of brain irradiation often used in the treatment of tumors.

While exercise encourages new nerve cells, a healthy diet may help keep more of those cells, and even mature cells, in top form. Research is steadily revealing compounds in fruits, vegetables and herbs that appear to enhance the survival of neurons new and old. Among the apparent brain foods: the omega-3 fatty acids found in fish such as salmon and sardines, flavonoids found in non-green vegetables and berries, and curcumin, a common component of curry. Van Praag has found that epicatechin, in green tea and chocolate, doesn’t promote the birth of neurons directly but does encourage existing neurons to sprout more connections to neighbors, improving memory. The effect is particularly strong when combined with exercise.

It’s still unknown whether people can eat and exercise their way to better neurogenesis, but researchers including van Praag hope to have an answer sometime during the next decade. While the Greeks may have sought help from the spiritual realm, the true route to better memory may lie in the choices people make every day.