Systems Biology Approaches to Cancer Energy Metabolism - Systems Biology of Metabolic and Signaling Networks - page 228

9.8 Conclusion

Kinetic modeling is a Systems Biology bottom-up approach that enables us to
understand the controlling mechanisms underlying glycolysis in tumor cells. To
build a reliable and robust kinetic model of a metabolic pathway, it is essential
to use appropriate rate equations for each step or group of steps, in which the kinetic
parameters for the forward and reverse reactions are adequately determined. For
reactions with thermodynamic constraints (i.e., high negative G values Δ >

5 kcal/mol), the kinetic parameters of the reverse reaction are difficult to deter-
mine, and hence V max reverse can be replaced by the equilibrium constant K eq in the rate
equation. The use of rate equations that accurately reproduce the enzyme/trans-
porter behavior shields the kinetic modeling from arbitrarily introducing
“adjustments” to the parameters to forcing correct simulation of the in vivo path-
way behavior.
However, a drawback in constructing or extending kinetic models is the over-
whelming amount of experimental data required for validation. That is the main
reason why only few kinetic models have been developed (H¨bner et al. 2011 ) and,
in particular for cancer glycolysis (Mar´n-Hern´ndez et al. 2011 ). The latter model
has been used to construct a modular model of the most relevant metabolic
processes regulated by the PI3K/Akt signaling pathway in the human embryonic
kidney HEK293 cells (Mosca et al. 2012 ). Their model could reproduce the
experimentally determined metabolic fluxes and allowed to predict the metabolic
targets that may inhibit tumor cell growth under hyper activation of Akt kinase.
Regarding glycolysis, models that include other levels of regulation such as gene
expression and signal transduction are interesting aspects to be included in the
future for a full understanding of the metabolic circuitry in these pathological cells.
Furthermore, it has been claimed that kinetic parameters determined in cellular
extracts (i.e., in diluted cellular solutions) cannot reflect the kinetic properties of the
enzymes in vivo because the cellular aqueous phases are crowded with
macromolecules, altering the enzyme-substrate and enzyme-product interactions
and in consequence the rate reaction (Agrawal et al. 2008 ). However, Vopel and
Makhatadze ( 2012 ) have demonstrated that the presence of several synthetic
crowding agents (ficoll, dextran 40,000, or albumin) does not significantly perturb
the K m and catalytic turnover number ( k cat ) of several glycolytic enzymes such as
HK, GAPDH, PGK, and LDH. Then, it seems that macromolecular crowding may
only affect diffusion rates of metabolites (Agrawal et al. 2008 ). As these rates are
orders of magnitude higher than the enzyme/transporter rate constants, modeling
based on kinetic parameters determined in diluted solutions is expected not to be
significantly modified by macromolecular crowding. This last conclusion is
supported by the relatively elevated accuracy and robustness of the herein shown
glycolytic model developed for tumor cells (Mar´n-Hern´ndez et al. 2011 ).
Although our model of cancer glycolysis has only considered a few well-defined
steady states, it may serve as a validated platform for constructing large-scale
kinetic models, which can be applied to a wider variety of physiological conditions.