Science Watch

Neuroscientists are beginning to understand the motivational workings of dopamine--the common neurotransmitter whose absence lowers drive and hurts motor control. Some of the most recent findings come from a team at Seattle's Howard Hughes Medical Institute, whose research suggests that, at least in mice, reward learning is quite possible without dopamine. Such insights may one day help everyone from teachers motivating their students to doctors treating patients with Parkinson's disease to counselors treating drug addiction: Dopamine's reach is that broad.

To reach their finding, the Seattle team used genetically altered knockout mice to tease out the impact of dopamine on subtly different aspects of the motivational system: wanting, liking and learning. By using technology to, in effect, turn off dopamine production and see what happened, they discovered that mice don't need dopamine to connect behavior with rewards or to find the rewards satisfying. The research appears in February's Behavioral Neuroscience (Vol. 119, No. 1).

Neuroscientist Kent Berridge, PhD, of the University of Michigan, says it appears that "dopamine is only needed to use already learned information to generate successful motivated performance." Translation: Dopamine promotes what we think of as "wanting."

By comparing the behavior of mice bred with mutations that inhibit dopamine production with the behavior of normal mice, the Seattle team may have helped clarify dopamine's historically ambiguous role. Especially for diseases linked to dopamine deficiencies, such as Parkinson's and schizophrenia, knowing how and whether one can motivate patients could mean a lot for clinical care.

In addition, "Separating motivation components is a popular and important approach to understanding motivation in the context of addiction," says Mark Kristal, PhD, a behavioral neuroscientist at the University at Buffalo of the State University of New York. In that case, psychologists want to know how to suppress the drug motivation of the addicted.

'What's my motivation?'

From the outside, it's hard to tell what most motivates an animal to seek a reward: the pleasure of the reward itself (roughly, liking), the satisfaction of getting it (wanting), or the acquired association between behavior and reward (learning).

"Wanting and liking are what some philosophers of mind have called 'folk psychological' terms about how the mind is organized," explains Jon Horvitz, PhD, a neuroscientist at Boston College. Although he doubts real brains have clearly demarked scripts for "wanting" or "liking," he says it helps to draw some rough distinctions to enable research into dopamine's behavioral pathway.

The Seattle researchers--graduate student Siobhan Robinson, undergraduate Suzanne Sandstrom, psychologist Victor Denenberg, PhD, and biochemist Richard Palmiter, PhD--chose a knockout approach to get a fair comparison between behavior with and without dopamine. Then, they threw caffeine into the mix to compensate for the motor lethargy but not the cognitive deficits caused by low dopamine.

Dopamine appears to be involved both in goal-directed and motor behavior. On the inside, dopamine-producing neurons extend into neighboring motivational and motor parts of the brain. And on the outside, when scientists block dopamine release, rewards such as food, sex and cocaine stop reinforcing behavior. But what does this mean: Do we stop liking them? Wanting them? Or learning that they're good? Once scientists know, they might be able to devise better therapeutic manipulations using dopamine or to design interventions that bypass the dopamine system.

Genetic engineering, says Berridge, author of a same-issue commentary on the study, "gives a completely independent way of asking the question" because experimenters can control--in a clean, noninvasive way--the relevant aspect of a subject's physiology. In knockout breeding, scientists remove or "knock out" a specific gene in an embryonic stem cell before it divides into many new different types of cells. When the resulting animal breeds, it passes down altered genes. Rodents breed so quickly that in short order, scientists can use genetically altered animals to show what a gene does by virtue of what does or doesn't happen in its absence.

Dopamine-deficient (DD) mice lack the enzyme needed to convert the amino acid tyrosine into levodopa, or L-dopa. Once L-dopa is formed, another enzyme they still have converts L-dopa into dopamine. A shot of L-dopa "rescues" DD mice, which will otherwise perish from starvation. Thus, Robinson explains, "The real beauty of the DD mouse is that the experimenter can control whether dopamine is present in the body by simply giving a shot of L-dopa."

Show time

The team's first experiment compared normal mice with lethargic DD mice whose L-dopa shots were converted into dopamine, which got them moving. The scientists trained the mice to run a T-maze with mouse chow at the end of the left or right arm of the upper bar. They compared how eight L-dopa-treated DD mice and nine control mice behaved, reasoning that liking is signaled by how much mice eat, whereas "wanting" is signaled by them chomping down sooner. The learning part comes through efficient running of the maze.

Treated with L-dopa, the knockout mice learned to run the T-maze just like normal mice and ate about as much about as quickly. With dopamine restored, they appeared to like and want the rewards as much as control mice. When the researchers established that DD mice with dopamine perform the task as well as control mice, they had set the stage for the next, more critical experiment. In that, they tested whether mice like, learn about and yearn for rewards without any dopamine in the brain. Would the lab equivalent of a cup of coffee get them going?

The researchers injected 25 DD mice with saline solution, L-dopa or caffeine, the latter of which stimulates locomotor activity through a nondopamine system. Then they measured how fast the mice in each group learned the T-maze. Then the researchers gave all the mice L-dopa and rechecked their learning.

At first the caffeinated DD mice didn't appear to learn much, but in the study's second phase they learned the task much quicker than would typical first-timers. Predictably, the saline-treated DD mice didn't do much of anything in the first phase and had novice learning times in the second one, and the L-dopa treated DD mice maxed out in their learning the first time around.

Thus it appeared that the caffeinated DD mice learned something during the first phase--and they learned it without dopamine. The authors thus conclude that normal reward learning does not depend on dopamine. This finding, coupled with previous findings that wanting does depend on dopamine, creates a fuller picture of motivation.

Still, Kristal cautions, "The mechanism for locomotor and motivational activation with caffeine may be separate from that for dopamine, and caffeine may alter the rate of dopamine metabolism--thereby confounding the results." Berridge agrees that the use of caffeine in DD mice "may muddy the picture a bit."

The plot thickens

New technologies raise new questions. For example, says Berridge, "It's always possible that [in the DD knockout mice], brain development produced some compensation. Maybe these mice have brains that can learn without dopamine." That's why, he says, neuroscientists, including those at the University of Washington lab, are trying to develop "inducible knockouts" in which the mouse could develop normally and then scientists could knock out a gene in later tests. For now, he says the study demonstrates that "brains without dopamine can still learn normally about rewards--at least, if they have caffeine activating them via a separate nondopamine biochemical pathway."

As another example, Kristal notes that the researchers didn't know whether the control mice--littermates with one of the two alleles (gene variations) needed for a functioning gene--behaved the same as normal mice from normal litters. If they behaved differently, that could confound the results and undermine the study's validity.

"Sometimes it almost seems that the correct answer to 'what does dopamine do?' might mostly be 'to confuse neuroscientists,'" says Berridge.

Horvitz adds, "The functional organization of the brain may or may not correspond well to categories such as liking and wanting. I think [the Seattle researchers] mean that dopamine is a player in neural circuitry that serves to vigorously mobilize behavior toward a particular goal object, which in humans, at least, is often accompanied by what we describe as 'wanting.'

"However," he continues, "it's unlikely that a particular neurotransmitter will correspond perfectly to a specific psychological construct such as wanting--or liking for that matter."

Palmiter agrees, saying, "It is very difficult to extrapolate from our studies with mice to humans, especially because our DD mice have much less dopamine than people with even severe Parkinson's disease." However, lead author Siobhan Robinson suggests that, "Perhaps caffeine can be used as a substitute for L-dopa during behavioral therapy with Parkinson's patients. To avoid the motor abnormalities induced by L-dopa during training, patients might learn new tasks without it that they'd be able to perform when on their daily L-dopa regimen."

Palmiter adds that it's also hard to immediately transfer the findings to everyday motivation, because "if dopamine levels were so low that motivation was affected, there would be many other Parkinson's-like symptoms." It is clear that research is needed to more fully understand these results and to begin to think about implications for relapsing drug addition, which is thought to result from over- (not under-) activity of the dopamine system, observe researchers in the area.

Still, Robinson likes to speculate about crafty real-world manipulation of natural dopamine mechanisms. One idea she has: "Designing classroom activities that may increase dopamine signaling, such as unexpected rewards along the way, may enhance the desire to perform well during and after learning. This could lead to better performance of learned tasks."

Rachel Adelson is a science writer in Raleigh, N.C.