Neuroscientist David Linden is excited about his latest research findings, even though the experiments' results echo his 4-year-old son's tendency to answer questions with a resounding no, no, no.

But in science, sometimes figuring out what isn't happening can be very telling, says Linden, whose report appears in the Nov. 13 issue of the Proceedings of the National Academy of Sciences. "It may sound dull because the answers are all no, but it's actually exciting, because there's just one possibility left, and we've already seen evidence for it," he says.

Linden is studying what brain cells do to retain a new memory. Scientists have shown that learning happens when a brain cell gets stimulated in a way that reduces its ability to respond to a particular brain messenger called glutamate.

"We have the 'wiring diagram' for some simple forms of motor learning, so we know how memories get stored in the circuit because we know which cells receive the stimulation and how they pass it along," explains Linden. "Now we're using that wiring diagram to look at the details."

Linden is studying strange, sea fan-shaped brain cells called Purkinje (pronounced per-KIN-jee) cells. These odd cells are found only in the cerebellum, a part of the brain involved in coordinating and learning muscle movement patterns.

"Purkinje cells are very unusual," says Linden, professor of neuroscience and director of the Graduate Program in Neuroscience at the School of Medicine. "They are very flat. They are enormous, but unlike other big neurons, they are inhibitory. They receive more connections than other types of neurons, and they fire 50 times per second even when you're sleeping."

The Purkinje cells are involved in simple motor learning processes similar to that in Pavlov's experiments with dogs. But while Pavlov conditioned a dog to salivate upon hearing a bell, Linden is considering a learning pathway that causes the eye to blink in response to a sound rather than a puff of air.

It turns out that the signals from the sound and the puff of air come together at Purkinje cells. Because the signals cross, eventually simple motor learning links the sound and the blink, says Linden. When the signals from the sound and the puff are received in quick succession and many times over, the action causes the temporary neuron shutdown that underlies learning.

Linden can mimic this shutdown, known as "long-term synaptic depression," by growing a Purkinje cell with a second type of brain cell that normally conveys the signal from the sound. With this two-cell system, he's examined a particular question surrounding long-term depression in the cerebellum.

"During long-term synaptic depression, the Purkinje cell doesn't respond to glutamate to the usual extent, but we haven't known for sure the reason behind that reduced response, " Linden says.

Once scientists showed that glutamate levels were constant, the question became, Which of four possible causes for the reduced response is really responsible for long-term depression of the neuron?

To excite a neuron, glutamate binds to its receptor on the surface of the cell. That binding, in turn, opens a channel in the cell's membrane that lets various charged atoms, or ions, through. During long-term depression of a neuron, the reduced response to glutamate could happen at four different points, Linden says.

Last year, Linden and his colleagues showed that one reason is that there are fewer docking points, or receptors, for glutamate on the surface of neurons during long-term depression. But three other possibilities still existed.

One possibility is that the receptor could lose its taste for glutamate and bind it less efficiently; another, the channel opened by glutamate could let ions through more slowly than normal, or it could be open for a shorter amount of time. Linden's new study rules out these three possibilities, leaving reduced receptor number as the only contributor, he reports in what he calls his "Jacob paper" for its similarities to his young son.

"The answers are no, no, and no," Linden says. "We suspected that the time of the ion channels' opening didn't change, but the rate of ion flow or the efficiency of glutamate binding were reasonable possibilities. The experiments, however, show that none of these three are present in Purkinje cells during long-term synaptic depression."

Now that the only contributor to the lowered response to glutamate is receptor number, at least in his laboratory model, Linden and others hope to create a mouse that can't reduce the number of glutamate receptors on its Purkinje cells. If the mouse can't learn to blink in response to the sound, they'll know they've found a key step in storing simple motor memories.

The study was funded by the United States Public Health Service and by the Develbiss Fund. Linden is the only author on the paper.