Machine learning creates opportunities for new personalized therapies
Researchers at the University of Michigan Rogel Cancer Center have developed a computational platform capable of predicting novel specific metabolic targets in ovarian cancer, suggesting opportunities to develop personalized therapies for constitutionally informed patients genetics of their tumours. The study appeared in Natural metabolism.
Cancerous mutations occur frequently in ovarian cancer, giving cells a growth advantage that contributes to the aggressiveness of the disease. But sometimes deletions of certain genes can occur alongside these mutations and make the cells vulnerable to treatment. Yet cancer cells grow so well because paralogous genes can compensate for this loss of function and continue to promote tumor formation.
Deepak Nagrath, Ph.D., an associate professor of biomedical engineering who led this study, wanted to know more about these compensatory genes as they relate to metabolism. “When a gene is deleted, the metabolic genes, which allow cancer cells to grow, are also deleted. The theory is that vulnerabilities emerge in cancer cell metabolism due to specific genetic alterations.
When genes that regulate metabolic function are deleted, cancer cells essentially rewire their metabolism to come up with a backup plan. Using a method that integrates complex metabolic modeling, machine learning and optimization theory in cell line and mouse models, the team discovered an unexpected function of a cancer enzyme of the ovary, MTHFD2. This was specific to ovarian cancer cells with impaired mitochondria, due to a common deletion of UQCR11. This led to a critical imbalance of an essential metabolite, NAD+, in the mitochondria.
The algorithm predicted that MTHFD2 surprisingly reversed its role in delivering NAD+ into cells. This created a vulnerability that could be targeted to selectively kill cancer cells while minimally affecting healthy cells.
“Personalized therapies like this are becoming a growing opportunity to improve the effectiveness of first-line cancer treatments,” said Abhinav Achreja, Ph.D., researcher and first author of this study. “There are multiple approaches to discovering personalized targets for cancer, and multiple platforms predict targets based on big data analytics. Our platform makes predictions by considering metabolic functionality and mechanism, increasing the chance of success when translating to the clinic.
Article quoted: “Identification of Lethal Collateral Metabolic Targets Reveals MTHFD2 Paralogue Dependence in Ovarian Cancer”, Natural metabolism. DOI: 10.1038/s42255-022-00636-3
This research was supported by funding from the National Cancer Institute, the Office of the Director of the National Institutes of Health, the Precision Health Scholars Award from the University of Michigan, and the Forbes Scholar Award from the Forbes Institute of Cancer Discovery.