After five to ten years on levodopa, the gold-standard drug for Parkinson’s disease, up to 80 percent of patients develop a cruel trade-off: the medication that restores their ability to move also begins triggering involuntary, sometimes disabling, writhing and jerking known as dyskinesia. A study from Northwestern University’s Feinberg School of Medicine, described in an institutional release as published in Science Advances in May 2026, now offers a clearer explanation for why that happens and a potential path to stopping it. (Note: the actual DOI or direct link to the Science Advances paper has not been provided in available sources; the details below are drawn from the institutional summary.)
The research team, led by neuroscientist D. James Surmeier, traced the problem to a specific circuit mechanism in the striatum, the brain region that coordinates voluntary movement. Their findings suggest that selectively blocking certain acetylcholine receptors on striatal neurons could tame dyskinesia without undermining the motor benefits that make levodopa indispensable for the roughly 10 million people living with Parkinson’s worldwide.
What the Northwestern team found
The conventional view of levodopa is straightforward: the drug replenishes dopamine lost as Parkinson’s destroys dopamine-producing neurons. But the Northwestern data paint a more complicated picture. Rather than simply topping off a depleted chemical, levodopa appears to set off a cascade of maladaptive changes in the striatum’s spiny projection neurons, the cells that relay motor commands to the rest of the brain.
Surmeier’s group found that heightened acetylcholine activity in these circuits causes the neurons to “learn” erratic firing patterns over time. Surmeier has called this process “aberrant learning,” a term that reframes dyskinesia not as a chemical overdose but as a circuit-level rewiring problem. In plain terms, the brain’s movement-control system is being taught the wrong lessons by its own signaling chemistry.
The proposed fix is precise. By blocking selected acetylcholine receptors specifically on spiny projection neurons, the researchers showed in preclinical experiments that the faulty plasticity driving dyskinesia could be dampened while the therapeutic motor improvements from levodopa remained intact. That cell-type and circuit-specific approach matters because older anticholinergic drugs, which suppress acetylcholine signaling broadly across the brain, have long been limited by side effects including cognitive fog and dry mouth.
If the strategy holds up in further testing, it could mean that levodopa therapy becomes safer and more sustainable over the long haul, sparing patients the agonizing choice between stiffness and uncontrollable movement.
How this builds on earlier circuit research
The new findings did not emerge from a vacuum. They sit atop a decade of work that has steadily sharpened the field’s understanding of what goes wrong in the striatum during both Parkinson’s disease and long-term levodopa therapy.
Earlier research co-authored by Surmeier established that pathway-specific plasticity changes occur in distinct populations of striatal projection neurons across parkinsonism and levodopa-induced dyskinesia. That 2014 study, published in Neuron, showed that different groups of these neurons respond in opposing ways to dopamine loss and replacement, supporting the idea that dyskinesia arises from imbalanced learning rules in specific circuits rather than a uniform flood of dopamine.
A separate 2019 study in Cell Reports documented neural activity patterns that track the onset and severity of dyskinesia. By recording from defined neuronal populations, that work linked abnormal firing and synchronization in the striatum to the expression of involuntary movements, grounding the “circuit dysfunction” theory in observable, measurable data.
Molecular profiling has added yet another layer. Analyses of gene-expression shifts in spiny projection neurons under dyskinetic conditions have revealed broad changes in the cellular programs governing synaptic strength and receptor composition. Chronic levodopa, it turns out, progressively reshapes striatal signaling at every level, from individual genes to whole-circuit dynamics.
What sets the new Northwestern work apart is its focus on acetylcholine as the key instructor of that reshaping. If cholinergic signaling is what teaches striatal neurons to fire erratically, then modulating that signal offers a concrete, druggable intervention point rather than a vague call for “more research.”
What remains uncertain
Promising as the framework is, several gaps limit how far the findings can be extended today.
First, the evidence is preclinical. The experiments relied on laboratory animal models, not human volunteers. Many mechanisms that look compelling in rodents fail to translate into effective or safe therapies in people. The human striatum is larger and more complex, and patients with Parkinson’s disease often carry comorbidities and medication histories that controlled experiments cannot fully replicate. Whether blocking specific acetylcholine receptors in human striatal circuits would reduce dyskinesia without impairing cognition, mood, or autonomic function is unknown. Dose, timing, and delivery route could all shift the balance between benefit and risk.
Second, the cholinergic mechanism is not the only explanation for levodopa-induced dyskinesia. A well-documented line of research has shown that serotonergic terminals can contribute to unregulated dopamine release in the dopamine-depleted striatum. In this view, serotonin neurons absorb levodopa, convert it to dopamine, and release it in erratic pulses, independently fueling involuntary movements. How this serotonin-driven process interacts with the cholinergic aberrant learning proposed by Surmeier’s group has not been worked out.
It is plausible that multiple mechanisms operate in parallel, each adding to the risk and severity of dyskinesia. If so, a therapy aimed solely at acetylcholine receptors might only partially alleviate symptoms, or might work best in combination with drugs that stabilize dopamine release from serotonin terminals. Conversely, interfering with one pathway could unmask problems in another. Disentangling these interactions will require experiments that manipulate both systems together, ideally in models that closely mimic long-term levodopa treatment in patients.
No timelines for clinical trials or drug development milestones have been disclosed. For patients currently managing dyskinesia, the research should be viewed as a mechanistic advance, not an imminent change in care. Existing strategies, including dose fractionation, extended-release levodopa formulations, and amantadine (the only FDA-approved medication specifically indicated for levodopa-induced dyskinesia), remain the primary tools available in the clinic.
Where the science goes from here
The strongest support for the aberrant learning framework comes from peer-reviewed research in established journals and the convergence of plasticity, activity, and molecular data pointing toward the same striatal circuits. The earlier work on pathway-specific plasticity and the demonstration of dyskinesia-linked activity correlates provide a solid foundation. Together with molecular profiling studies, they make a credible case that chronic levodopa reshapes how striatal circuits encode motor commands.
Still, mechanistic clarity does not guarantee therapeutic success, especially in a condition as heterogeneous as Parkinson’s disease, where age, disease duration, and genetic background all influence treatment response. The serotonergic contribution to dyskinesia, documented through studies of dopamine release from serotonin neurons, is a reminder that no single pathway has yet been shown to account for every case.
For now, the Northwestern team’s work is best understood as a sharper lens on levodopa’s long-term effects. By framing dyskinesia as aberrant learning within specific striatal circuits shaped by acetylcholine, it opens concrete avenues for targeted drug development and more nuanced preclinical models. Whether those avenues will lead to medications that meaningfully improve daily life for people with Parkinson’s disease remains an open question, one that only carefully designed clinical studies can answer.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
