PhD Student Texas A&M University Houston, Texas, United States
Introduction: Cancer metabolism is a complex interplay of various metabolic processes that support the growth and survival of cancer cells. This model presents a comprehensive approach by integrating the roles of master gene regulators AMPK, HIF-1, and MYC with key metabolic pathways involving glucose, fatty acids, and glutamine. Apart from glucose, fatty acids and glutamine have been identified as significant metabolic ingredients for tumorigenesis and cancer progression. The model identifies five distinct metabolic phenotypes in cancer cells: highly catabolic with intense oxidative processes; highly anabolic with pronounced reductive activities; two hybrid phenotypes, one combining catabolic and anabolic activities, and the other, relying on glutamine oxidation with high glucose reduction; and finally, an extra phenotype indicating reduced overall metabolic activity. We validated the results utilizing RNA-seq data from tumor patients; this framework not only enhances our understanding of cancer metabolism but also suggests potential therapeutic targets by manipulating these metabolic pathways. Understanding cancer metabolism thus provides critical insights into various hallmarks of cancer, such as metastasis and immune suppression.
Materials and
Methods: The methodology of this study involves a mathematical model that simulates the temporal dynamics of regulatory proteins and metabolites in cancer metabolism. The model uses shifted Hill functions to represent the competition between the various metabolic processes considered: glucose oxidation, glycolysis, reductive glucose metabolism, fatty acid oxidation (FAO), and glutamine metabolism. The model incorporates thresholds for glucose uptake rate, acetyl-CoA utilization rate for mitochondrial respiration, and glutamine uptake rate to simulate restrictions on metabolic rates. The model also includes the regulatory roles of proteins like HIF-1 and AMPK in up-regulating or down-regulating specific pathways. The model represents the dynamics of glucose oxidation rate, the glycolysis rate, the reductive glucose metabolic rate, the FAO rate, and the reductive fatty acid metabolism. It also captures the ATP production rates from different metabolic sources and ATP consumption rates for various metabolic activities. The net total production rate of ATP is calculated considering all these factors. Shifted Hill function parameters including the level of the regulator, the threshold, the fold-change, and the Hill coefficient. It also outlines competitive regulation of noxROS by AMPK and HIF-1 through specific functions. RNA-Seq data and clinical data from several cancer samples were downloaded from The Cancer Genome Atlas (TCGA) via cBioPortal. These data were subsequently analyzed, focusing on the downstream genes associated with the various metabolic processes examined in our model. This approach allowed for a comprehensive exploration of the metabolic landscape in HCC.
Results, Conclusions, and Discussions: The model identifies different metabolic phenotypes that can be acquire by cancer cells based on MYC levels. We first investigated two MYC expression levels, denoted as low and high. Our analysis revealed that under low MYC expression, cells primarily exhibit two metabolic states, labeled as "W" and "O". However, with MYC overexpression, cells can now also manifest a hybrid metabolic state termed "W/O". Principal component analysis (PCA) and k-means clustering were applied to the MYC Low vs High datasets obtained after stable state solutions from 1000 sets of randomized parameters, revealing five corresponding clusters. The clusters were visualized along the first two principal components and the AMPK vs HIF-1 axis. The study also validated the predicted characterization of different metabolic phenotypes using a previously defined glutamine metabolism gene signature (GMGS). The GMGS was applied to 45 breast cancer samples and 45 corresponding adjacent normal tissue samples. The results suggest that the “W” state has a more efficient uptake of glutamine as it exhibits higher expression of glutamine transporter genes. Systematic analysis of the main genes involved in glutamine metabolism was conducted using RNA-seq data from three types of tumor patient samples. The analysis found that the tumor samples in the “O” and “W/O” states exhibit higher expression of glutamine oxidation genes relative to tumor samples in the “W” state. The “W” state exhibits high glutamine uptake and anabolic processes involving glutamine, and the “W/O” state exhibits both high glutamine oxidation and high anabolic activity involving glutamine, all of which is consistent with the model predictions. Finally, further analysis of the survival of HCC patients for the identified clusters showed that Hyb2 phenotypes had the worst survival than the rest of the groups. This study provides a robust framework for understanding the complex dynamics of cancer metabolism and opens new avenues for the development of targeted therapies. The integration of genetic regulation with metabolic pathways, coupled with the use of a mathematical model, allows for a detailed characterization of the metabolic phenotypes of cancer cells.
Acknowledgements (Optional): JTG is a CPRIT Scholar in Cancer Research and is supported by CPRIT grant (RR210080). JVC acknowledges CONAHCYT for financial support granted for PhD Studies (CVU 637952).
