Background
Mitochondria act as the batteries of the cell and are vital to the health of cells. Therefore, it is detrimental to cell health if mitochondria are not working properly. Dysfunctional mitochondria must be removed from cells by a process called mitophagy. This happens continuously in healthy cells but is impaired in diseases like Alzheimer’s, causing an accumulation of damaged mitochondria. It’s thought that if compounds can be used to repair or regulate this process, they could be a potential therapy for Alzheimer’s Disease.
Why is the study important?
This study further supports the theory that defects in the removal of faulty mitochondria (mitophagy) play an important role in Alzheimer’s disease, and restoring this process might be an effective therapeutic approach. The authors designed a new method to use Artificial Intelligence (AI) combined with testing in various cell and animal models to identify potential compounds.
What did the authors do and how did they do it:
By combining various AI methods to scan through a collection of thousands of drugs from traditional Chinese medicine, the authors identified 18 compounds that had the potential to increase the removal of damaged mitochondria. Their approach with AI integrated basic chemical information with more complex data about specific areas of the compounds to predict what activity and action each may have. To narrow down the selection and determine the best compounds, they performed tests in cells, nematode worms, and mice; all with mutations to mimic Alzheimer’s.
The process of damaged mitochondrial removal was quantified in worms and mice alongside behavioural studies to examine whether there were improvements on memory. Furthermore, they investigated whether treating mice with the two lead compounds resulted in a reduction of amyloid beta in the brain. This is one of the proteins that forms clumps in the brain which are a hallmark of Alzheimer’s disease.
Although animal models are important for drug discovery, the results do not always translate into humans. Differences in the brains and bodies of the animals in comparison to humans, and the fact that only certain aspects of the disease are mimicked in the models, mean that compounds in animals may not have the same effect or outcome in humans.
What are the results?
After experimenting in cells and worms, two lead compounds were selected – Keam and Rhap. These compounds were amongst the hits from the initial selection identified by AI. The scientists determined that this methodology had a 44% success rate, much higher than other traditional methods. The compounds were shown to improve memory in worms and mice, as well as reducing the clumping of the harmful proteins implicated in Alzheimer’s – amyloid beta and Tau.
What do the findings mean going forward for people with the disease?
This study is important as the authors have developed a method utilising AI to narrow down a collection of thousands of compounds to a select few which can be tested in animal models. This methodology has identified two compounds that could be taken forward into further testing and clinical trials.
When compared to other drug screening methods, the authors determined that the one described here had a much higher success rate. Additionally, the same method could be used for other neurodegenerative diseases.
Although the compounds have shown promise, it’s possible that increasing the removal of damaged mitochondria could be detrimental, and there are concerns whether Rhap is able to pass from the blood into the brain where it is needed.
Nevertheless, this study presents a new way to combine AI with traditional experimental work, accelerating the discovery of new drugs.