Mark Wightman is a mind reader. With a superfine fiber and some electronics, he can find out what's going on in your head. And if it's got anything to do with addiction, he's very interested. Especially since he discovered that the brain's "reward system" isn't as simple as we used to think.

Wightman knows this because he does his research right where the action is. He sticks a probe ten times thinner than a hair into the brain and listens electronically as cells "talk" to each other by releasing chemical signals called neurotransmitters.

"There are other methods for detecting chemicals in the brain," says Wightman, professor of chemistry. "But they're too slow to catch the cell's equivalent of words or phrases. It's like playing a tape too fast: All the sound comes at once. Our method works in real time. And we can zero in on a single message."

The "message" Wightman is listening for is the neurotransmitter dopamine, which helps keep our moods in check. Low levels can cause depression or Parkinson's disease. Unusually high levels are associated with schizophrenia and drug use.

Dopamine has long been called the key player in the brain's "reward system." Psychologist James Olds discovered this system by chance in the 1950s, when he put an electric probe into the limbic system of a rat's brain—the part believed to be involved in emotions. When the animal pressed a lever, the probe delivered a tiny shock—just enough to stimulate his brain. The sensation was so pleasurable that the rat pressed the lever over and over, stopping only to sleep.

Later, when researchers looked for the brain chemicals involved in this frenzied self-stimulation, dopamine seemed to be a major player. And because drugs such as cocaine are known to affect dopamine levels, the chemical has been thought to play a role in addiction. In fact, Olds' experiment is still used as a model to study addiction.

Wightman didn't know any of this 20 years ago when he started eavesdropping on cells. His field was analytical chemistry, where scientists' main preoccupation is trying to identify materials or detect their presence. But Wightman's post-doctoral adviser challenged him to learn about the brain using the field's tool of choice, the electrode.

Electrodes make great detectives. Stick one into a beaker of liquid, turn on the power, and electric current starts to flow. So do clues. The amount of current tells you how many molecules are giving or receiving electrons. And you can find out which molecules they are by knowing how much voltage is needed to start the current.

As it happens, dopamine and some other neurotransmitters can give up electrons, so, in theory, an electrode can detect them. But, back then, plunging even the smallest known microelectrode into a brain was like unleashing Godzilla in a sculpture garden. The enormous, clumsy devices—about the size of a large pencil lead—would destroy the brain's delicate network of neurons. Beyond that, microelectrodes weren't sensitive enough to detect the low levels of chemicals in the brain.

"That's why my adviser called it a challenge," Wightman says. "There were some problems to solve before we could collect any data."

The first step—making the electrode smaller—wasn't too hard. Another researcher told Wightman about carbon fibers—superfine strands that, when glued together, make graphite tennis rackets. The fibers had the property Wightman needed: They didn't produce much background current. One of Wightman's students encased a fiber in glass to insulate it and to prevent it from breaking. Voilà! An ultra-microelectrode was born.

Figuring out how to measure the device's tiny currents was another matter. They were so faint that just having fluorescent lights on created enough background noise to drown them out. Wightman's group built superfast electronics to detect the minute flows: They couldn't afford to let a single electron slip by undetected. Then the researchers had to find the signal buried in the noise.

According to what Wightman had learned, the latter wasn't possible. To get stronger signals, he had to make the electronics faster. But every time he did so, he'd get more noise. There was no way around it. Or so it seemed.

Then he met two British physiologists who taught him a technique they had just applied to their own work: They measured the background noise first, then subtracted it electronically from the signal. It worked like a charm.

With his technical problems behind him, Wightman began to test the ultra-microelectrodes (now simply called microelectrodes) using cells taken from the adrenal glands, which sit atop the kidneys. Adrenal cells develop from the same cells that neurons do, so they're similar. The main differences are that adrenal cells are larger and release adrenaline, which is related to dopamine, rather than dopamine itself.

But both neurons and adrenal cells release their chemicals in "packets"—a process called "exocytosis"—and that's what Wightman wanted to see. Back then, it was still a theory. People had seen microscope images that supported the idea, but those pictures showed freeze-dried cells. Nobody had seen exocytosis in live cells. So Wightman's group went to work with their probes.

"We put the cells in a special chamber and used the electrode on them," Wightman says, a conspiratorial glint in his eye. "It was like questioning them under a spotlight. They told us everything."

The theory said that chemical signals didn't leave the cell in a steady stream. Instead, they were collected and packaged inside small spheres. Each sphere rose to the cell's surface and dumped its load outside. Sure enough, Wightman's group saw the level of adrenaline rise and fall repeatedly, each spike indicating that a sphere had just spilled its contents.

"We saw peak after peak," Wightman says. "It was just beautiful. Packets of chemicals were coming out rapid-fire: pow, pow, pow."

One of Wightman's students also noticed a small bump before each spike. It looked like a tiny foothill next to a mountain. But Wightman didn't think the "foot" was real. He wrote it off as some kind of background noise, even though they never found a foot after a spike. That point bothered him a bit: Why would there be noise before a spike but not after?

Still, Wightman didn't think the bumps were worth mentioning when he gave a talk about his unpublished results. But in the audience sat a German physiologist named Erwin Neher, who won a Nobel prize for inventing the patch clamp—a device that measures a cell's electrical activity. Wightman's microelectrode did the same thing with chemical activity, so the senior researcher took an interest in the young scientist's work. He duplicated Wightman's experiments, then published a paper refining the theory of exocytosis—and including an explanation of the foot.

When a chemical-carrying sphere touches the inside of a cell's membrane, the two join together at one spot. A tiny pore opens there, letting a minuscule amount of chemical leak out. That shows up as the foot. As the pore grows, the remaining chemicals escape, creating the spike.

"It was a lot of fun to have a Nobel winner interacting with us and sharing the excitement of discovery," Wightman says. But there's a hint of regret when he adds, "My students were the first ones to see the foot, but I didn't have enough biology training to understand what it meant. Neher did."

Later, Wightman made his own refinement. He realized that chemicals took a hundred times longer than expected to escape. He looked more closely at the spheres and found that they're filled with a kind of jelly, which probably protects the chemicals during storage. After the sphere spills its contents, the gel takes a little time to dissolve, so the chemicals seep out slowly.

As informative as the studies of individual cells were, Wightman knew the real potential of his technique lay in explaining behavior. That's when he began reading the minds of rats. And what he found surprised him.

In early experiments, he and his students learned that dopamine travels farther in the brain than anyone thought. Traditional theory says that neurons which release dopamine are hard-wired, meaning they line up head to tail. One neuron releases dopamine from its synapse—its "tail" end—and the "head" of the next neuron picks it up. Because that was believed to be the chemical's one and only stop, researchers thought information could only be conveyed where such connections were made.

It doesn't happen this way, Wightman says—not for the neurons in charge of dopamine or serotonin, another mood elevator in our brains. Instead, the chemicals ooze out and affect many neurons at once, even distant ones. This suggests that the brain isn't especially orderly. At least, these networks of neurons aren't.

"Just knowing that much—knowing that various types of neurons may be organized differently—tells us more about the brain and how to treat problems when they arise," he says. Next, Wightman's group looked at how cocaine affects the brain. Other researchers had shown that the drug interferes with dopamine "uptake"—that's when a neuron sequesters the chemical it just released. Whatever escapes goes to nearby neurons and activates their receptors, making us feel good. When drugs block uptake, there's a lot of dopamine available for activation, and we get high. When Wightman's group used their probes to track dopamine, they found that uptake was slow. The results seemed to confirm the idea that dopamine is the brain's primary or "continuous" reward.

Then the researchers duplicated the classic self-stimulation experiments, which had previously shown that dopamine release goes hand-in-hand with addictive behavior. Everything seemed to be the same: The rats were trained, and during the experiment the next day, the animals pressed the bar and got pleasurable sensations. But unlike earlier researchers, Wightman's group didn't see much dopamine at all. "The levels were absolutely feeble," he says. "Insignificant, compared to what we expected."

They tried over and over. Nothing. The rats behaved like addicts, but the substance thought responsible was nowhere to be found. When the researchers stimulated dopamine release directly in untrained rats, however, they got tons. Dopamine was there, all right. It just wasn't being released during the test.

"We were stymied," he says.

Then the group lucked out. They trained a rat on a Friday and, by chance, had to wait until Tuesday to conduct the self-stimulation test. The result? Dopamine galore.

"Clearly, dopamine was not responsible for the continuous reward," Wightman says. "The previous rats got plenty of reward without it.

"So why did we see dopamine the last time?" he wonders. "We think the rat had forgotten its training over the long weekend and was relearning the test. We think dopamine's role is not in addiction, but in novelty, in learning." As further evidence, Wightman adds that rats prevented from releasing dopamine couldn't learn the behavior needed for the test.

Other research supports the idea that dopamine isn't the brain's main reward. Scientists at Columbia University found that mice susceptible to cocaine addiction have a genetic problem affecting serotonin, not dopamine. And researchers at Duke University showed that mice can get high without any dopamine at all.

"This means people need to look for the chemical that really is responsible for continuous reward," Wightman says. "Dopamine no longer looks like it."

Now that this work has been published in Nature, that's just what Wightman's group will be looking for. He says he already has a few candidates. He and his students will also be trying to link dopamine to all kinds of "novelty behavior," including being startled.

How does all this apply to humans? Wightman's not sure. He says that he's nowhere near the point of switching from studies of rats to studies of humans. "What would be the point?" he says. "Of course, people really want to get into the human brain. But we know so little about neuroscience that we can't even ask the right questions yet. We couldn't hope to find the right answers." In other words, if we went into the human mind now, we'd be in over our heads.

This research appeared in the March 4, 1999 issue of Nature. It was funded by The National Institute of Drug Abuse and the National Institute of Neurological Disorders and Stroke.