AI May Improve Dementia and Alzheimer's Drug Production

GDJ/Pixabay

A new peer-reviewed study by researchers at The University of Texas at Austin (UT at Austin) shows how artificial intelligence (AI) machine learning can help guide and improve enzyme design for the synthesis of galantamine, a drug used to treat patients with dementia and Alzheimer's disease. This study demonstrates the powerful potential of the combination of AI and directed evolution in the industrial-scale production of important medications in the future.

Galantamine may help improve the cognitive ability of patients diagnosed with dementia or Alzheimer’s by slowing the breakdown of acetylcholine, a neurotransmitter that plays a key role in learning and memory, per the Mayo Clinic. In Alzheimer’s patients, acetylcholine is in lower concentration and function, according to The Interplay of Neurotransmitters in Alzheimer's Disease by Paul Francis. Galantamine, the generic drug for Razadyne™, Reminyl™, and Nivalin®, is a small molecule capable of crossing the blood-brain barrier that has been approved by the U.S. Food and Drug Administration (FDA) to treat mild to moderate Alzheimer’s disease symptoms, according to Boston-based AlzForum. Galantamine occurs naturally in daffodils (Narcissus pseudonarcissus), snowflakes (Leucojum aestivum), and other plants belonging to the family Amaryllidaceae, according to The Alkaloids: Chemistry and Biology by Michael Heinrich.

“Alkaloids produced by the Amaryllidoideae subfamily of flowering plants have great therapeutic promise, including anticancer, fungicidal, antiviral, and acetylcholinesterase inhibition properties,” wrote corresponding author Simon d’Oelsnitz and co-authors Daniel Diaz, Wantae Kim, Daniel Acosta, Tyler Dangerfield, Mason Schechter, Matthew Minus, James Howard, Hannah Do, James Loy, Hal Alper, Y. Jessie Zhang, and Andrew Ellington.

The scientists used directed evolution to create a biosensor for the target alkaloid. In the field of synthetic biology, directed evolution is a laboratory process that accelerates the natural evolution of proteins that speed up chemical reactions called enzymes. American scientist Frances Arnold was one of the recipients of the 2018 Nobel Prize in Chemistry for her pioneering work on the first directed evolution of enzymes in 1993.

According to the UT at Austin researchers, Amaryllidaceae alkaloids are challenging to synthesize, hence they are extracted and purified from daffodils, a process that does not lend itself well to industrial-scale biomanufacturing. “A promising alternative to amaryllidaceae alkaloid extraction from plants is microbial fermentation,” the researchers wrote.

Using an AI model called MutComputeX created by postdoctoral fellow Daniel Diaz at UT's Institute for Foundations of Machine Learning (IFML), the scientists were able to determine the ideal temperature and how to optimize the protein mutations within Escherichia coli (E. coli ), gram-negative rod-shaped bacterium that are part of normal intestinal flora in healthy humans and animals. MutComputeX is a self-supervised convolutional neural network (CNN) that predicts the mutations that would enable E. coli to efficiently make the desired chemical building block to produce galantamine.

The researchers report that with the help of AI, they spotted mutations that enable E. coli to produce the target molecule up to three times more efficiently than processing daffodils.

“Overall, these results highlight the powerful capability of using evolved biosensors for precisely reporting on pathway intermediates while avoiding cross-reactivity with closely related precursor molecules,” the researchers concluded.

Copyright © 2024 Cami Rosso. All rights reserved.