Parkinson’s Disease: A Learning Disorder?

347600-mouse-rota-rod-treadmillOccasionally, drugs produce beneficial mysteries – effects that are useful to physicians despite being largely unexplained. Levodopa (L-dopa), the most commonly-used treatment for the symptoms of Parkinson’s disease, is meant to replace dopamine, the neurotransmitter lost as the disease progresses to its most severe stages. Clinicians recognize that the benefical effect builds up slowly over weeks despite the same dose of medication. In addition, after a patient’s L-dopa is stopped, the relief of a patient’s motor symptoms can persist partially for weeks, long past the time it takes to clear the drug entirely from one’s system. Though this effect has a name – the “long duration response,” or LDR – nobody’s quite sure what causes it, though physicians are happy to put it to use in patients.

Adding to the mystery is the fact that animal models of Parkinson’s disease, usually involving chemical brain lesions, have consistently failed to replicate the long duration response. But in a bit of serendipity, a University of Chicago laboratory studying the role of dopamine in the learning of motor skills may have unintentionally found the LDR in a mouse. Published this month in the Annals of Neurology, the finding could dramatically shift both the theory and treatment of Parkinson’s.

Jeff Beeler, a postdoctoral researcher in the laboratory of Xiaoxi Zhuang, associate professor of neurobiology, wanted to study a difficult problem: how can one untangle motor performance from motor learning? The deficits seen with Parkinson’s disease, such as muscle tremors and akinesia (the inability to initiate movement), implicate dopamine as an important part of the brain’s motor control. But is it simply that one needs dopamine to move, or do you need dopamine to learn how to move? It’s a hard question to experimentally test, Beeler said.

“If you can’t perform something, in a sense you can’t learn it,” Beeler said. “On the other hand, if you can’t learn it you can’t perform it. It’s a chicken and egg thing.”

So Beeler began working with a unique mouse strain, called aphakia mice, with a naturally-occurring genetic mutation. The mice were first known for being blind (aphakia means the loss of the lens of the eye), but were later found to have a severe, 90 percent depletion of dopamine neurons – similar to the loss seen in advanced Parkinson’s patients. But strangely, these mice exhibit only very subtle motor deficits.

“It’s a good model of particular aspects of Parkinson’s disease,” said Un Jung Kang, professor of neurology and director of the University of Chicago Movement Disorders Center and another author of the paper. “The simple way of saying it is that this is a model of very mild Parkinson’s disease that reflects a lot of endogenous compensation. They basically have a very minor deficit in terms of motor performance.”

Those mild symptoms allowed Beeler and his colleagues to test the role of dopamine in motor learning vs. performance, something that was impossible in the more severely-affected toxic Parkinson’s models – “Their performance is so bad, there’s nothing left to learn and to retain,” Kang said.

The aphakia mice were given L-dopa and run on a rotarod test, a slowly rotating cylinder that the mice must learn to navigate like a competitor in a log-rolling contest (seen in the above video from a different lab). An aphakia mouse given a control saline injection typically fell off the rotarod in seconds, and didn’t show much improvement with repeated trials. But with L-dopa, the aphakia mice steadily improved their rotarod skills, until on the 5th day of training they were able to stay on for a couple of minutes, similar to normal mice.

This effect reproduced the slow build up of beneficial effects in patients. But then the team cut off L-dopa treatment, and continued testing the mice on the rotarod. Surprisingly, the dopamine-depleted mice initially remained adept rod-rollers after discontinuation of L-dopa – an animal version of the long duration response.

But while modeling an unusual drug effect in humans for the first time was exciting, additional experiments went further, raising questions that could potentially redefine the central philosophy of treating Parkinson’s disease. The lingering rotarod performance even in the absence of L-dopa suggested that dopamine was involved in learning the task in addition to performing it. But did a mouse’s performance eventually drop off because it “forgot” how to stay on the rotarod in the absence of dopamine?

Beeler found that the animals’ performance only declined after they repeated the rotarod task without L-dopa – performing another task (running on a treadmill) without L-dopa had no effect on subsequent rotarod trials. That result suggests that mice without dopamine were still learning, but were acquiring the incorrect motor actions needed to stay on the rotarod.

“In the absence of dopamine, they’re essentially learning not to do things they should do,” Beeler said. “The decline is dependent upon further experience with that task. You can give them a similar task, and that doesn’t make a difference. They have to do the task in order to get worse at it.”

Extrapolated to human Parkinson’s patients, the result suggests that the akinesia associated with the disease may be partially due to this sort of aberrant learning, in addition to a loss of dopamine’s direct effects on motor function. Such a theory – which needs much more testing, the researchers stressed – could affect the way Parkinson’s patients are treated.

“Clinically, if dopamine is necessary for learning, it changes the philosophy of treating these symptoms,” Kang said. “The usual philosophy is that we’re just providing a symptomatic treatment by dopamine; therefore, you usually take a conservative approach of waiting until they’re more bothered by the symptoms before you treat. But if it’s affecting your learning, and that lasts a long time, we should be treating more aggressively as soon as we know the patient has a deficit in dopamine.”

The research also suggests new therapeutic targets that can be used as an alternative to L-dopa. The Zhuang laboratory is studying novel approaches to maintaining correct motor learning while avoiding unwelcome side effects of conventional therapies. Like the above theory, it’s a hypothesis that needs further testing in animal models and in humans, but it’s nevertheless a promising clinical lead born, almost accidentally, from animal research.

“I think this is a great example of clinical and basic research going back and forth and synergizing with each other,” Kang said. “Now we can go back to patients to test specifically the role of dopamine in motor learning and their ability to do everyday activities.”


Beeler JA, Cao ZF, Kheirbek MA, Ding Y, Koranda J, Murakami M, Kang UJ, & Zhuang X (2010). Dopamine-dependent motor learning: insight into levodopa’s long-duration response. Annals of neurology, 67 (5), 639-47 PMID: 20437561

About Rob Mitchum (526 Articles)

Rob Mitchum is communications manager at the Computation Institute, a joint initiative between The University of Chicago and Argonne National Laboratory.

%d bloggers like this: