Tumor Microenvironment Influences Metabolism of Cancer Cells

June 6, 2016, by Cosimo Commisso

When we think about tumor heterogeneity, we often imagine variation that originates from genetic and/or epigenetic mechanisms. However, there is increasing evidence that variation exists at the level of cellular metabolism and that this intratumoral heterogeneity could be important to tumor progression. The vast majority of RAS-driven tumors are notoriously aggressive and do not respond to chemotherapies, and it is conceivable that the metabolic heterogeneity found within these tumors could be contributing to their aggressiveness or to chemoresistance. Variations in metabolic wiring could allow some cancers cells within the tumor to be better positioned to survive specific stresses within the harsh tumor microenvironment. Moreover, metabolic adaptation to these stresses might also impact signaling pathways that regulate macropinocytosis and autophagy, both critical routes of nutrient supply in RAS-driven tumors^{1, 2}.

In rapidly proliferating RAS-mutated pancreatic cancer cells, cellular metabolism comes in at least two broad flavors – cells that are glycolytic and those that are lipogenic³. Interestingly, the glycolytic cells, which have elevated levels of various components of glycolysis, are associated with more mesenchymal properties, while lipogenic cells that are enriched for various lipid metabolites are associated with a classical, more epithelial cell type^{3, 4}. This is a great example of how metabolic heterogeneity within RAS-driven tumors might be contributing to tumor aggressiveness and/or invasion. In my lab we have observed that in addition to the differential usage of various metabolic pathways, RAS-mutated pancreatic cancers also exhibit intratumoral variation in how they obtain nutrients, with some cells depending on nutrient transporters and other cells relying on protein scavenging pathways, such as macropinocytosis. Altogether, it looks like it will be imperative to take into account this heterogeneity when designing novel therapeutic strategies for this disease.

What could be driving this metabolic heterogeneity in RAS-driven tumors? One possibility is stresses within the tumor microenvironment, such as nutrient depletion or hypoxia. In poorly vascularized tumors, one can imagine that nutrient or oxygen consumption by the tumor cells might outpace supply, leading to nutrient stress or hypoxic conditions. Indeed, we found that pancreatic tumors are depleted for non-essential amino acids such as glutamine, cysteine, serine and asparagine, which tumors use as nutrients to drive their growth⁵. It is unclear whether such nutrient deficiencies are found throughout the tumor or are more regional in nature; however, it is possible that as nutrients wane, RAS-mutant cells are able to kick in compensatory mechanisms such as macropinocytosis or autophagy. When and how protein scavenging is activated in these tumors is a vigorous area of research for my lab, and we are cognizant that the overall tumor landscape needs to be considered. An example of how tumor architecture might affect metabolic heterogeneity comes from non-small cell lung cancer (NSCLC). NSCLCs, which can harbor mutations in KRAS or EGFR, display metabolic heterogeneity within and between tumors⁶. Strikingly, hypovascularized regions of these tumors employ glucose as a key nutrient, while highly perfused tumor regions seem to rely more heavily on fuels other than glucose⁶. Therefore, proximity to nutrient or oxygen supply could be an important factor that influences metabolic heterogeneity within tumors.

Can we exploit metabolic dependencies to develop novel therapies for RAS-mutant tumors? To target so-called metabolic Achilles’ heels, we need to have a much deeper understanding of not only cancer metabolism, but also the extent to which metabolic variation occurs within tumors. What we have already discovered could be the tip of the iceberg, especially when you consider the vast capacity for RAS-driven tumors to change and adapt to their environments. Targeting tumor metabolism will require a multipronged approach, probably including multiple metabolism-based strategies, together with conventional antineoplastics. We will need to consider not only how metabolism in tumor cells differs from quiescent cells, but also the degree to which metabolic heterogeneity and adaptation contribute to tumor progression. With the more widespread use of 3-dimensional organoid and organotypic models, we are now better positioned to understand how tumor cells interact with and adapt to their complex and stress-inducing environments.

About the Author

Cosimo Commisso is an Assistant Professor in the Cancer Metabolism and Signaling Networks Program at Sanford Burnham Prebys NCI-Designated Cancer Center in La Jolla, California. In 2013, while a postdoctoral fellow in Dafna Bar-Sagi’s laboratory at NYU, Dr. Commisso led a team that discovered how macropinocytosis supports tumor growth by serving as an amino acid supply route in RAS-driven cancers. His lab is focused on understanding the molecular basis of RAS-induced macropinocytosis, and the impact of metabolic heterogeneity on tumorigenesis and targeted cancer therapies.