Memory Molecules

Neurobiologists have uncovered a pathway that stores information from the brain's temporary memory onto its "hard disk." The findings, published in the April issue of Nature Neuroscience, might lead to new drugs to improve memory function in amnesia and neurodegenerative diseases.

Although memory is still in large part a mystery, scientists think it's a two-step process: Information first enters a temporary, short-term storage that works by selectively strengthening connections between some neurons. To make impressions last, however, brain cells have to run a specific genetic program and synthesize new proteins. Experiments in mice suggested an important role in long-term memory for CREB, one of many so-called transcription factors, proteins that switch genes on and off. Anatomical studies also suggest that the hippocampus, an arc-shaped structure buried deep inside the brain, is essential for learning and memory. So neuroscientists Mark Bear and Cristina Alberini at Brown University in Providence, Rhode Island, pulled those two lines of research together: They wondered what exactly the role of CREB in the hippocampus might be.

Their team trained rats to avoid a dark chamber by giving them a mild electroshock when they entered. Normal rats remembered their lesson for several days after undergoing the experiment a single time; rats in which the team had severed the fornix, a bundle of nerves that sends input to the hippocampus, also learned to stay away from the dark--but for a few hours only. When the team compared brain slices from the two groups of rats, they discovered that in the intact animals, more CREB molecules carried a phosphate group--a crucial step for their activation. Signals coming in through the fornix are apparently required for the activation of CREB and for long-term memories, says Bear. "If we can identify these signals, we might be able to restore [them], suggesting a possible treatment to improve cognitive function."

Howard Eichenbaum, a cognitive neuroscientist at Boston University, hails the study as "a breakthrough. It associates the molecular machinery with the role of the hippocampus in memory formation. That link hasn't been made previously," he says. But he is puzzled about the mechanism whereby nerve stimuli would attach phosphate groups to CREB. "It's unclear to me how a nerve impulse could activate gene expression in a general way," he says.