Artificial intelligence identifies anti-aging drug candidates targeting 'zombie' cells

Researchers from Integrated Biosciences, a biotechnology company that combines synthetic biology and machine learning to combat aging, have published a new study in the May issue of Nature Aging that shows the power of artificial intelligence (AI) to find novel senolytic compounds, a class of small molecules that are actively researched for their capacity to inhibit age-related processes like fibrosis, inflammation, and cancer.

In a paper titled "Discovering small-molecule senolytics with deep neural networks," researchers from MIT, the Broad Institute of MIT and Harvard, and the Massachusetts Institute of Technology (MIT) describe how AI-guided screening of more than 800,000 compounds led to the discovery of three drug candidates with similar efficacy and superior medicinal chemistry properties to those of senolytics currently being studied.

According to Felix Wong, Ph.D., co-founder of Integrated Biosciences and first author of the study, "This research result is a significant milestone for both longevity research and the application of artificial intelligence to drug discovery." These results show that, in comparison to even the most promising examples of their kind now under investigation, it is possible to explore chemical space in silico and produce a number of potential anti-aging drugs that are more likely to be successful in clinical trials.

Senolytics are substances that specifically cause apoptosis, or programmed cell death, in senescent, non-dividing cells. Senescent cells, a sign of aging, have been linked to a variety of age-related illnesses and ailments, including Alzheimer's disease, diabetes, cancer, and cardiovascular disease. Despite encouraging clinical outcomes, the majority of senolytic drugs have been discovered so far have limited bioavailability and negative side effects. The goal of Integrated Biosciences, which was established in 2022, is to use artificial intelligence, synthetic biology, and other next-generation methods to overcome these challenges, focus on more underappreciated signs of aging, and accelerate anti-aging medication development more generally.

Finding therapeutic approaches that selectively eliminate these cells from the body in a manner akin to how antibiotics kill bacteria without hurting host cells is one of the most promising ways to cure age-related disorders. Satotaka Omori, Ph.D., Head of Aging Biology at Integrated Biosciences and joint first author of the study, stated the compounds "display high selectivity as well as the favorable medicinal chemistry properties needed to yield a successful drug." We think the substances found using our technology will have better chances in clinical trials and eventually assist restore health to aged people.

Researchers at Integrated Biosciences trained deep neural networks on data produced via experiments to forecast the senolytic action of any chemical in their latest study. From a chemical space of over 800,000 molecules, they found three very selective and powerful senolytic drugs using their AI model. In tests for hemolysis and genotoxicity, all three were shown to have good toxicity profiles and to have chemical characteristics indicative of excellent oral bioavailability.

According to structural and biochemical investigations, all three substances bind Bcl-2, a protein that controls apoptosis and is also a target for chemotherapy. One of the compounds was tested in experiments on 80-week-old mice, which are about equivalent to 80-year-old people. It was discovered that the chemical removed senescent cells and decreased expression of senescence-associated genes in the kidneys.

The founding chair of the Integrated Biosciences Scientific Advisory Board, James J. Collins, Ph.D., said that this work "illustrates how AI can be used to bring medicine a step closer to therapies that address aging, one of the fundamental challenges in biology." In 2020, a team led by Dr. Collins, senior author on the Nature Aging publication, found the first antibiotic recognized by machine learning.

"Integrated Biosciences is expanding on the fundamental work that my academic group has been doing for the last eight to ten years, which has demonstrated that we can tune cellular stress responses utilizing systems and synthetic biology. its study stands out in the area of drug discovery and will significantly advance the field of longevity research thanks to its experimental tour de force and the outstanding platform that generated it.