Collaboration to Design New Medicines for Levodopa-induced Dyskinesia
Nearly all people with Parkinson’s receive levodopa to treat the movement difficulties caused by the body’s loss of its own dopamine. Cells metabolize levodopa, a precursor molecule, into dopamine, which helps individuals regain normal movement.
Prolonged use of levodopa, however, leads to its own set of motor symptoms, known as levodopa-induced dyskinesia. This includes symptoms such as muscle twitches and other erratic movements, and fatigue.
At the moment, this condition is largely treated by adjusting a patient’s levodopa dose and prescribing medications such as Gocovri (amantadine), which, while effective, often provide only short-term relief and can carry their own side effects, such as dizziness, fainting, and hallucinations.
In the search for better alternatives, the University of Montreal (UdeM) is leading a collaboration that includes the university’s own Institute for Research in Immunology and Cancer (IRIC); IRICoR, a pan-Canadian drug discovery and research commercialization center; and Valence Discovery, an artificial intelligence drug discovery firm.
The research of Daniel Levesque, PhD, a UdeM pharmacy professor leading the effort, focuses on the Nur77/RXR nuclear receptor complex, which appears to be directly involved in levodopa-induced dyskinesia. Animal studies have shown that blocking this receptor or otherwise preventing its activity alleviates symptoms.
Researchers in the project will use Valence’s machine learning platform (machine learning refers to algorithms that adapt and improve automatically, based on experience) to identify chemical structures likely to contribute to highly selective compounds targeting Nur77/RXR.
“We’re thrilled to be working with Dr. Levesque and the world-class team at IRIC, who have an extensive track record of collaborating with leading industry partners,” said Daniel Cohen, CEO of Valence Discovery.
In contrast to many machine learning algorithms, Valence’s “few-shot” drug discovery platform can design drugs based on relatively small amounts of starting data, the company states. This might prove useful in cases where little historical data exists to help guide drug discovery, or where biochemical drug-finding assays cannot be performed at large scale.
“We are extremely pleased to have Valence’s support on this important drug discovery program, and are confident that our joint efforts will significantly accelerate our path to identifying novel compounds that can treat levodopa-induced dyskinesia, a serious side effect of the most common treatment for Parkinson’s disease,” said Levesque.