The chemical secrets of memory

View full image

Some labs are as white and sterile as hospital rooms, but Karyn Frick's looks more like the inside of FAO Schwarz. Everywhere you look, there are buckets crammed with plastic toys in rainbow colors, transforming the lab into one big amusement park for the dozens of female mice in residence. But Frick's not out just to keep her rodent charges occupied; she uses the toys to glean insight into how they learn and remember. She picks up two identical multicolored prisms and puts them in a large white box, a kind of murine playpen. "We put two of these objects in," she explains, "and let the mice explore for 30 seconds." Next comes the crucial step: injecting the mice with estrogen.

Scientists have long known that estrogen is responsible for female sex development in mammals. Frick, a Yale associate professor of psychology, is one of a growing number of researchers who have done experiments showing that the hormone also aids memory in females. (Estrogen is also important for males. Many of the effects of testosterone on memory in males are due to its conversion to estrogen.) If Frick can help to find out how and why estrogen has these effects, her work may lead to new drug therapies for people suffering from Alzheimer's and other memory disorders. "All of us are going to get old someday," she says. "The question is, Is there a type and duration of estrogen treatment that can enhance memory function or reduce its decline?"

Drugs like these could be a particular boon for women who experience memory problems after menopause or later in life and find themselves forgetting not just their keys, but their grandchildren's names—or their children's. Memory loss increases precipitously with advancing age: only 4 percent of people aged 65 to 69 report moderate or severe memory problems, but 36 percent of people 85 and older do. Like nuclear fallout, the impact of memory loss on individuals and families is widespread and seemingly irreversible. Frick has made it her mission to give memory-loss sufferers some hope—to isolate the pathways that govern memory retrieval in order to fortify them against the assaults of age and disease.

To gauge estrogen's effect on her mouse subjects' memory retention, Frick rounded up mice that had looked at the prisms two days before and put them back in the playpen with one of the prisms and a novel toy—a water spigot. "The longer they explore the new object, the more we know they remember the old one," Frick says. "Since they already know about it, they're not that interested in it. But if they don't remember the old object, they'll explore both objects about equally." In a recent study published in the Journal of Neuroscience, Frick demonstrated that mice receiving estrogen injections remembered objects better than did the control group of mice, which had gotten a placebo injection.

Frick's study was one of many that present tantalizing evidence that estrogen can indeed serve as a memory aid. As far back as 1992, Bruce McEwen and colleagues at Rockefeller University found that the number of synapses, or nerve connections, in female rats' brains rise and fall with their estrous cycle. McEwen's studies focused on the hippocampus, a part of the brain that is needed for the formation of many kinds of memories and that is exceptionally vulnerable to aging and Alzheimer's disease. Since then, researchers have probed, and confirmed, the link between estrogen and memory from many different angles.

Yet as promising as the research is, its proponents keep bumping up against formidable obstacles. In 2004, the National Institutes of Health called a halt to its large-scale study of estrogen-replacement therapy because participants had a slightly higher than average risk of stroke—eight extra strokes per year for every 10,000 women—as well as more cases of breast and uterine cancer. Thousands of women flushed their estrogen pills down the toilet, literally or figuratively, when they and their doctors heard the news. The findings were also discouraging for estrogen-based memory therapy. Analysis of the NIH data found no discernible memory improvement. In some women, estrogen use even seemed to carry an increased risk of dementia.

What lies behind this apparent contradiction? Part of the answer may be the age of the NIH subjects. "The population in this study was older—they had all been off natural estrogen for many years before they started the therapy," says Christina L. Williams, a behavioral neuroscientist at Duke. "More-current studies have suggested that when estrogen is replaced right after menopause, it can be very good for cognitive function over the long term."

Frick concurs. "The mean age of the women in this study was mid-sixties, but estrogen may work much better when women are in their fifties. The NIH data support the critical-period theory that suggests there's a limited window in which estrogen therapy can be beneficial."

But most importantly, the seeming conflict between the NIH study and other research merely underlines what scientists already know: that memory is extraordinarily complicated. In 2002, Frick herself had published a paper noting that there were "inconsistencies regarding the ability of estrogen to improve memory in menopausal women." She has probed estrogen's inconsistencies in her research with mice. The range of results she has produced shows, among other things, that estrogen acts differently on mice of different ages; that progesterone can block estrogen's effects but can also have memory-boosting effects of its own; and that mice experiencing different levels and lengths of environmental enrichment (more toys, more places to explore) respond differently to estrogen.

To understand these varied effects, scientists need to find out how estrogen works at the level of the brain cells that encode and retain the information we call memory. As Frick pointed out in her 2002 paper, the inconsistencies make it all the more important to find out what's really going on.