Vital information on the function of mitochondria in cancer cells could lead to new approaches in cancer treatment

Investigators from the UCLA Jonsson Comprehensive Cancer Center utilized PET (positron emission tomography) alongside electron microscopy in generating 3D ultra-resolution maps of mitochondrial networks in genetically engineered mice with lung tumors. Using deep learning, an AI technique, the researchers categorized the tumors based on factors such as mitochondrial activity. This quantified the mitochondrial architecture throughout the tumor.

Examining two major subtypes of NSCLC (non-small cell lung cancer) - adenocarcinomas and squamous-cell carcinomas - they observed distinct subpopulations of mitochondrial networks within these tumors. They have discovered how the mitochondria regularly organize themselves with lipid droplets to form subcellular structures that aid metabolism and mitochondrial activity.

In human cancer cell samples, the authors are expecting that the mitochondrial populations will not be mutually exclusive to their respective tumor subtype. They assumed that there would be a spectrum of activity.

The generated maps enabled the scientists to create a structural and functional atlas of lung tumors, which gave them crucial insight into how tumor cells organize themselves in response to the high metabolic demands of tumor growth. Regarding the metabolic flux of lung tumors, the study also revealed how the nutrient preference can be determined by the compartmentalization of the mitochondria with other organelles, either relying on glucose or free fatty acids.