Metabolic requirements for cancer cell proliferation

Abstract

Background: The study of cancer metabolism has been largely dedicated to exploring the hypothesis that oncogenic transformation rewires cellular metabolism to sustain elevated rates of growth and division. Intense examination of tumors and cancer cell lines has confirmed that many cancer-associated metabolic phenotypes allow robust growth and survival; however, little attention has been given to explicitly identifying the biochemical requirements for cell proliferation in a rigorous manner in the context of cancer metabolism.

Results: Using a well-studied hybridoma line as a model, we comprehensively and quantitatively enumerate the metabolic requirements for generating new biomass in mammalian cells; this indicated a large biosynthetic requirement for ATP, NADPH, NAD(+), acetyl-CoA, and amino acids. Extension of this approach to serine/glycine and glutamine metabolic pathways suggested lower limits on serine and glycine catabolism to supply one-carbon unit synthesis and significant availability of glutamine-derived carbon for biosynthesis resulting from nitrogen demands alone, respectively. We integrated our biomass composition results into a flux balance analysis model, placing upper bounds on mitochondrial NADH oxidation to simulate metformin treatment; these simulations reproduced several empirically observed metabolic phenotypes, including increased reductive isocitrate dehydrogenase flux.

Conclusions: Our analysis clarifies the differential needs for central carbon metabolism precursors, glutamine-derived nitrogen, and cofactors such as ATP, NADPH, and NAD(+), while also providing justification for various extracellular nutrient uptake behaviors observed in tumors. Collectively, these results demonstrate how stoichiometric considerations alone can successfully predict empirically observed phenotypes and provide insight into biochemical dynamics that underlie responses to metabolic perturbations.

Keywords: Anabolism; Biosynthesis; Cancer metabolism; Metabolic modeling; Proliferation; Stoichiometric analysis.

Fig. 1

Simplified schematic of central carbon metabolism. Rectangular boxes contain branchpoint metabolite intermediates, and rounded rectangular boxes contain amino acid and fatty acid products that can be incorporated into biomass macromolecules. Arrows indicate carbon flux. Additional metabolic intermediates are not shown; instead, they are implicitly lumped into pools with displayed metabolites (e.g., fructose 6-phosphate with G6P)

Fig. 2

Fates of major biomass precursors and cofactor equivalents consumed in synthesis of macromolecules. Fates of biomass precursors (3-phosphoglycerate and oxaloacetate), nitrogen/amine groups, and cofactors (NAD⁺, NADPH, and ATP) are classified by their requirements for major classes of macromolecules (proteins, nucleotides, lipids, and polysaccharides). Demands for each macromolecule include both costs of polymerization and de novo synthesis of monomers

Fig. 3

Schematic of the major routes of one-carbon unit production. Serine is catabolized through serine hydroxymethyltransferase (SHMT), and glycine is catabolized through the glycine cleavage system (GCS). Intracellular compartmentalization is not shown

Fig. 4

Schematic of the major routes of glutamine contribution to carbon and nitrogen biomass. Deamidation of glutamine to glutamate occurs either via glutaminase (GLS) or various enzymes in nucleotide biosynthesis pathways. Glutamate subsequently can donate its remaining α carbon amine group (NH₄ ⁺ _α-C) to α-keto acids via aminotransferases (ATs) to form amino acids, resulting in conversion of the glutamate carbon skeleton to αKG. GLS also produces free ammonium (NH₄ ⁺ _amide), which can subsequently be incorporated into αKG to regenerate glutamate by glutamate dehydrogenase (GDH)

Fig. 5

Metabolic flux alterations predicted to occur in response to inhibition of NADH oxidation in the ETC in simulations of metformin treatment. (a) Schematic indicating absolute and relative flux values in central carbon metabolism for 0, 50, and 100 % inhibition cases; cellular compartmentalization not shown for simplicity. (b) Oxygen consumption, (c) glucose consumption, (d) lactate production, and (e) net IDH fluxes plotted as functions of percent NADH oxidation inhibition