Systems biology of cancer: moving toward the integrative study of the metabolic alterations in cancer cells

Abstract

One of the main objectives in systems biology is to understand the biological mechanisms that give rise to the phenotype of a microorganism by using high-throughput technologies (HTs) and genome-scale mathematical modeling. The computational modeling of genome-scale metabolic reconstructions is one systemic and quantitative strategy for characterizing the metabolic phenotype associated with human diseases and potentially for designing drugs with optimal clinical effects. The purpose of this short review is to describe how computational modeling, including the specific case of constraint-based modeling, can be used to explore, characterize, and predict the metabolic capacities that distinguish the metabolic phenotype of cancer cell lines. As we show herein, this computational framework is far from a pure theoretical description, and to ensure proper biological interpretation, it is necessary to integrate high-throughput data and generate predictions for later experimental assessment. Hence, genome-scale modeling serves as a platform for the following: (1) the integration of data from HTs, (2) the assessment of how metabolic activity is related to phenotype in cancer cell lines, and (3) the design of new experiments to evaluate the outcomes of the in silico analysis. By combining the functions described above, we show that computational modeling is a useful methodology to construct an integrative, systemic, and quantitative scheme for understanding the metabolic profiles of cancer cell lines, a first step to determine the metabolic mechanism by which cancer cells maintain and support their malignant phenotype in human tissues.

Keywords: cancer metabolic phenotype; computational modeling of metabolism; constraint-based modeling; genome scale metabolic reconstruction; high throughput biology.

Figure 1

Scheme of Flux Balance Analysis (FBA). First is the definition of the network, HT data is used to complete different levels of information (either in the network, concentration, or fluxes), the network get represented in an stoichiometric matrix. Given the fluxes, the concentration of each metabolite and the stoichiometric matrix the steady state is found.

Figure 2

Central metabolism in cancer cell lines. LAC represents lactate, G6P represents glucose-6phosphate, F6P represents fructose 6 phosphate, FDP represents fructose 1,6 biphospate, G3P represents glyceraldehydes 3 phosphate, 13DPG represents 1,3 biphosoglycerate, 3PG represent 3 phosphoglycerate, 2PG represents 2-phopho glycerate, PEP represents phosphenol pyruvate, PYR represents pyruvate, 6PGCL represents 6-phosphoglucono-δ-lactone, 6PGC represents 6-phosphogluconate, RU5Prepresents ribulose 5-phosphate, R5P represents Ribose 5 phosphate, X5P represents Xylulose 5 phosphate, S7P represents sedoheptulose 7-phosphate, E4P represents erythrose 4-phosphate, OAA represents oxaloacetate, SUC-COA represents succinyl-CoA, ACCOA represents acetyl CoA.

Figure 3

Analysis with constraint-based modeling. (A) Having reconstructed the metabolic network, in silico gene deletion allows to identify the phenotype behavior in cancer cell lines. In this case, phenotype is defined in terms of biomass production. Three effects can occur in this situation: (1) genes whose activity is dispensable, (2) genes whose expression reduce the biomass, and (3) genes whose activity is essential to biomass production. Based on in silico analysis, we conclude that lactate dehydrogenase (LDH) has a pivotal metabolic control on cancer cell growth (Resendis-Antonio et al., 2010). Supporting this finding, panel (B) shows the effects that variations on the enzymatic activity of LDH have on some enzymes participating in glycolysis, TCA cycle and pentose phosphate. As the figure shows according the metabolic activity of LDH decrease, we note a reduced activity on all the enzymes integrating these pathways. Regions in red and blue represent a higher (H) and lower (L) metabolic flux activity, respectively. (C) Phenotype phase plane considering the activity of pyruvate dehydrogenase (PDH) and glucose uptake rate.