A pavlovian learning experiment in lowly sea slugs has provided new clues to how we remember our childhood. The findings, reported in tomorrow's Science, are the strongest evidence yet for the long-suspected role of long-term potentiation (LTP), a physiological process that jacks up the response of certain neurons to incoming signals.
When a neuron equipped for LTP receives two signals in rapid succession, the second one triggers the LTP process, in which special channels open up and allow calcium ions to flow into the neuron. The calcium triggers biochemical reactions that alter the neuron's sensitivity so that a repeat of that signal will produce a heightened response. LTP seemed to be a logical way to encode memories, because it provides a means for neurons to associate simultaneous events, and it reinforces the neurons' response to those stimuli when they occur again. Researchers have worked for decades to understand LTP and to establish its link to memory, but they were hampered by the fact that LTP had only been shown to exist in mammals.
In search of a simpler test, David Glanzman's team at the University of California, Los Angeles turned to sea slugs. Four years ago they had discovered LTP in neurons from the slugs, which can also learn to associate a light touch on their siphon with an electrical shock to their tail. Normally sea slugs don't respond to the touch, but after conditioning, the touch makes them withdraw their siphon in anticipation of a shock.
Grad student Geoffrey Murphy took a closer look at this pavlovian learning in partially dissected sea slugs. To condition the neurons, he simultaneously stimulated the sensory neuron that registers touch on the siphon, and an incoming nerve from the tail, both of which communicate directly with the motor neuron that moves the siphon. The motor neuron, which had shown little response to stimulation of the sensory neuron alone prior to the simultaneous conditioning, nearly leapt out of its shoes in response to a tweak of the sensory neuron after the conditioning--suggesting that the neuron had learned to associate the firing of the sensory neuron with the more noxious signals from the tail nerve. Proof that the learning was due to LTP came when the researchers showed that it could be blocked by a chemical that stops the calcium influx necessary for LTP.
The finding is exciting because it provides the best link yet between LTP and learning, says neuroscientist Dan Johnston of Baylor College of Medicine in Houston. "It is one thing to believe it and another to have it demonstrated," he says. What's more, adds Yale neuroscientist Tom Brown, it provides a simple system in which to study LTP, and shows that LTP must have evolved early as a mechanism for memory.