Artificial intelligence discovers potential plant extracts for obesity treatment
In an innovative study utilising artificial intelligence (AI), researchers have identified two plant-based compounds with the potential to act as GLP-1 agonist weight loss medications. This significant finding will be presented at the European Congress on Obesity (ECO 2024), set to take place in Venice from 12th to 15th May, 2024.
GLP-1 (Glucagon-like peptide-1) receptor agonists, such as semaglutide and tirzepatide, have proven highly effective in aiding weight loss. These agents work by emulating the effects of the GLP-1 hormone, which interacts with receptors in cells to diminish appetite and hunger sensations, decelerate gastric emptying, and enhance satiety following meals.
Despite their efficacy, the quest for alternatives is imperative, according to Elena Murcia from the Structural Bioinformatics and High-Performance Computing Research Group (BIO-HPC) & Eating Disorders Research Unit at the Catholic University of Murcia (UCAM), Spain. She notes the presence of side effects associated with current GLP-1 agonists, including gastrointestinal discomfort and mental health fluctuations, such as anxiety and irritability. Moreover, discontinuation of these treatments often leads to weight regain.
A significant limitation of current GLP-1 agonists, which are peptide-based, is their susceptibility to degradation by stomach enzymes, necessitating administration via injection rather than oral intake.
The search for non-peptide alternatives has recently identified two promising synthetic compounds, TTOAD2 and orforglipron. However, Murcia and her team are driven to discover natural counterparts that might offer similar benefits with potentially fewer side effects and simpler administration methods.
Employing high-performance AI techniques, the team embarked on a quest to identify non-peptide, plant-derived compounds capable of activating the GLP-1 receptor. Their investigation began with a virtual screening of over 10,000 compounds, aiming to pinpoint those that could bind to the GLP-1 receptor. Subsequent AI analyses assessed the similarity of these bonds to the natural interaction between the GLP-1 hormone and its receptor. This led to the selection of 100 compounds for further visual scrutiny to evaluate their interaction with crucial receptor residues.
A mathematical approach involving Venn diagrams helped distil the search to 65 potential GLP-1R agonists, among which “Compound A” and “Compound B” showed strong binding affinity to critical receptor sites, akin to the synthetic compounds TTOAD2 and orforglipron.
These compounds, derived from commonly known plants with historically recognised metabolic benefits, are currently under laboratory examination. Pending patent approvals, further specifics about these plants and compounds remain confidential, with aspirations for future pill-form administration.
Murcia highlights the nascent stage of developing these natural source-derived GLP-1 agonists. Should the AI-predicted efficacy be validated through in vitro studies and subsequent clinical trials, these compounds could offer new therapeutic avenues for obesity management.
She further emphasises the advantages of computer-based research methodologies, including cost and time efficiency, the capacity for rapid large-scale data analysis, experimental design flexibility, and pre-emptive ethical and safety risk assessments. These simulations exploit AI capabilities to tackle complex challenges, providing invaluable preliminary insights in the quest for novel pharmaceutical solutions.
