When it comes to the neurobiology of memory, the hippocampus typically gets most of the credit. But although this brain region is crucial for recording new memories, like the name of someone you just met at a bar, people with a damaged hippocampus can still recall memories from days of old. Many neuroscientists believe this is because lasting memories get shifted to the cerebral cortex for permanent storage. Little is known about how this might happen, but a study in today's issue of Science provides some clues.
The new study, by neuroscientists Kaori Takehara-Nishiuchi and Bruce McNaughton, then at the University of Arizona, Tucson, builds on a 2003 study on memory by Takehara-Nishiuchi. She and colleagues trained rats to blink when they heard a tone signaling a mild shock to the eyelid. When the researchers removed the hippocampus a day after the training, rats no longer remembered to blink when they heard the tone, but removing the hippocampus 4 weeks after the training session had no effect: The rats still blinked. Conversely, removing the medial prefrontal cortex (mPFC) a day after the training caused no memory lapses, but removing this region 4 weeks later made the rats forget all about the tone. The findings suggested that whereas the hippocampus is essential for short-term memory storage, the mPFC is critical for long-term storage.
To investigate further, Takehara-Nishiuchi used hair-thin electrodes to record the activity of individual neurons in the mPFC of rats. This time the animals received more complex training. Over the course of 2 months, the rats heard the tone in two different contexts--in a square enclosure, where the tone always preceded a shock; or in a circular enclosure, where the tone and shock occurred randomly. By 2 weeks, rats had learned to blink when they heard the tone inside the square enclosure, where it reliably predicted a shock. But they ignored the tone inside the circular enclosure, where it did not. At the same time, about a quarter of the rat's mPFC neurons fired at higher rates when the tone sounded inside the square enclosure. This selective firing developed gradually over the first 2 weeks of training and persisted for the rest of the training period. The time course of this change in neural firing is similar to the time course of memory consolidation in the mPFC suggested by Takehara-Nishiuchi's 2003 study.
The current findings bolster a prevalent hypothesis about how long-term memory storage might work, says Bruno Bontempi, a neuroscientist at the Université de Bordeaux in Talence, France. The study provides electrophysiological evidence that "memories are progressively laid down in the cortex as they mature over time and become independent of the hippocampus," he says. A lingering question, Bontempi says, is exactly how memories become encoded in the mPFC. Research with both rodents and people has suggested that new memories are reactivated during sleep (Science, 9 March 2007, p. 1360). One possibility, Bontempi says, is that this memory reactivation has a role in changing neural firing patterns and stabilizing memories in the cortex.
)
)
)
)
)
)
)
)
