The discovery that new neurons are born in the adult brain overturned decades-old dogma in neuroscience. But it also raised a host of questions about what exactly these neurons do (Science, 17 February 2006, p. 938). Now the authors of a new study suggest that the newcomers clear away the remnants of old memories to make room for new ones.
The brain's hippocampus is a bit like a secretary's inbox: Although many memories start out here, they eventually get filed to the neocortex for permanent storage. That's why the famous patient H.M., who had his hippocampus removed in experimental surgery for epilepsy, could remember events prior to his operation despite being unable to form any new memories afterward (Science, 26 June, p. 1634).
To investigate whether newly born neurons play a role in memory transfer to the neocortex, researchers led by Kaoru Inokuchi of the University of Toyama in Japan examined rats and mice with impaired hippocampal neurogenesis. The researchers trained the rodents to associate a particular chamber with a mild shock. Like normal animals, they remembered the association for weeks, freezing up any time they were placed in the chamber. This type of memory usually hangs out in the hippocampus for less than a month: When the researchers injected the brains of normal rodents with a drug that essentially turned off the hippocampus after 28 days, it had little to no effect on their freezing behavior--presumably because the memory had already moved on to the neocortex. But in the rats and mice with impaired neurogenesis, the same treatment substantially reduced freezing behavior, suggesting that the fearful memory had lingered longer than usual in the hippocampus instead of being transferred to the neocortex. A similar set of experiments with mice that exercised on a running wheel--an activity previously shown to boost neurogenesis--bolstered the idea that neurogenesis plays a role in transferring memories. In that case, memories appeared to shift from the hippocampus to the neocortex faster than usual.
Finally, recordings of neural activity indicated that that long-term potentiation, a physiological strengthening of neural connections thought to underlie this type of learning and memory, persists longer than usual in the hippocampus of the neurogenesis-impaired rodents. Altogether, the findings—reported tomorrow in Cell--suggest that new neurons act like an efficient secretary, making sure the physiological traces of old memories are promptly removed from the hippocampus inbox to make room for new ones.
"The authors went through a lot of experiments to prove their case," says Gerd Kempermann, an expert on neurogenesis at the Center for Regenerative Therapies Dresden in Germany. But Kempermann is not quite convinced that the specific job of new neurons is to clear the hippocampus for new information. An alternative explanation, he says, is that new neurons simply enable the hippocampus to work more quickly. "But their conclusion is certainly interesting and great food for thought."