Tumors exhibit altered metabolism compared to non-transformed tissues (DeBerardinis and Chandel, 2016). For example, animal limbs transformed with oncogenic viruses exhibit increased glucose uptake and lactate secretion relative to unaffected limbs (Cori and Cori, 1925). Some aspects of tumor metabolism, including higher rates of glucose fermentation to lactate, are cell-intrinsic features that are retained when cancer cells are isolated from tumors (Koppenol et al., 2011). Indeed, numerous studies have delineated how cell-intrinsic factors such as oncogenic lesions or epigenetic state alter cellular metabolism, causing phenotypes such as increased glycolysis (Nagarajan et al., 2016).
However, beyond cell-intrinsic alterations, tumors have modified tissue architecture and an altered tissue microenvironment; these cell-extrinsic factors can also impact the metabolism of tumors (Muir et al., 2018). For instance, the metabolic utilization of both glucose and the amino acid glutamine differs between cells growing in culture and murine tumor models (Davidson et al., 2016; Muir et al., 2017; Sellers et al., 2015; Tardito et al., 2015). Further, the metabolic enzymes that cancer cells rely upon for proliferation in culture are different than those that the same cells utilize to support growth and survival in tumors (Alvarez et al., 2017; Possemato et al., 2011; Yau et al., 2017). Thus, cancer cell metabolism is influenced by microenvironmental cues.
Numerous microenvironmental factors affect cancer cell metabolism (Anastasiou, 2017; Bi et al., 2018; Muir et al., 2018; Wolpaw and Dang, 2018), including the presence of stromal cells (Lyssiotis and Kimmelman, 2017; Morandi et al., 2016), tumor acidity (Corbet and Feron, 2017; Persi et al., 2018), extracellular matrix properties (DelNero et al., 2018; Tung et al., 2015) and tumor nutrient levels (Muir and Vander Heiden, 2018). In particular, environmental nutrient availability is an important regulator of cancer cell metabolism (Cantor et al., 2017; Muir et al., 2017; Schug et al., 2015; Tardito et al., 2015). Nutrient differences between standard cell culture and animal tumor models can drive substantial changes in cancer cell metabolism that alter the response of cancer cells to metabolically targeted drugs (Cantor et al., 2017; Gui et al., 2016; Muir et al., 2017; Palm et al., 2015), such that drugs that inhibit proliferation of cancer cells in culture fail to exhibit efficacy in tumors derived from the same cells (Biancur et al., 2017; Davidson et al., 2016). Thus, determining the concentrations of nutrients in the tumor microenvironment is important to understand and therapeutically target cancer cell metabolism.
Tumors and tissues are supplied with nutrients through the vasculature, which filters a nutrient rich fluid from the circulation into the interstitial space of a tissue or tumor (Wiig and Swartz, 2012). The interstitial fluid (IF) then perfuses through the tissue or tumor, exchanging nutrients and wastes with cells. IF is then drained from the tissue or tumor via capillaries and the lymphatic system. Thus, cells in tissues or tumors are not necessarily directly exposed to the nutrients in circulating plasma, but instead are exposed to IF nutrient levels. For healthy organs, IF nutrient levels may be similar to circulating nutrient levels, as these tissues are well vascularized, allowing rapid metabolic exchange with the plasma. Indeed, the IF glucose concentration in healthy skin is very similar to that of circulating plasma (Lönnroth et al., 1987). In contrast to the functional vessels found in normal tissues, tumors commonly have an abnormal vasculature with few vessels transporting blood (Fukumura et al., 2010). This may lead to reduced nutrient delivery and waste exchange between tumor cells and the circulation. Thus, tumor interstitial fluid (TIF) is thought to be nutrient depleted compared to either normal tissues or the circulation, and have increased concentrations of metabolic waste products. Indeed glucose levels in the TIF of some tumors are lower than in the circulation, while lactate levels are higher (Burgess and Sylven, 1962; Gullino et al., 1964; Ho et al., 2015). However, despite the importance of nutrient availability in regulating tumor metabolism and drug sensitivity, TIF nutrients beyond glucose and lactate have not been comprehensively measured, and the factors that determine TIF composition have not been characterized.
We sought to systematically measure absolute nutrient concentrations in plasma and TIF. To do so, we have utilized a quantitative mass spectrometry-based approach using both external standards and stable isotope dilution of a library of carbon labeled metabolites to determine the concentration of >118 nutrients in biological fluids. We applied this technique to measure nutrient levels in plasma and TIF isolated from autochthonous and transplantation models of murine lung (LUAD) and pancreatic adenocarcinomas (PDAC) driven by activation of Kras and deletion of p53. Interestingly, we found that anatomical location and tumor tissue of origin are both major determinants of TIF nutrient composition. Dietary changes are also reflected in TIF nutrient levels, while introduction of a Keap1 mutation into LUAD cells had a smaller effect on the composition of the metabolic tumor microenvironment. Collectively, these experiments elucidate the microenvironmental constraints placed upon tumor metabolism by TIF nutrient levels and provide insight into the factors that dictate tumor nutrient availability.
