In order to understand quantitatively how metabolic processes work, a mechanistic model is often formulated through biochemical mass balances and reaction kinetics. The modeling of a recognized, although not fully characterized yet, metabolic pathway is selected as an illustrative case study for this paper: the astrocyte-neuron lactate shuttle in mammalian brain metabolism. Extension to other biomedical and biotechnology applications is discussed.

The brain is the supervisory center of the nervous system in all mammals. Regulation of brain metabolism and cerebral blood flow involves complex control systems with many several interacting variables at both cellular and organ levels. Quantitative understanding of the spatially and temporally heterogeneous brain control mechanisms during internal and external stimuli requires the development and validation of a computational (mathematical) model of the metabolic processes in the brain tissue.

Unlike metabolic pathways in liver, heart and skeletal-muscle tissues, metabolism in mammalian brain tissue is widely believed to follow a compartmental model. Indeed, the conventional view that glucose oxidation fuels most activity-associated energy metabolism in neurons is challenged by the astrocyte-neuron lactate shuttle (ANLSH) hypothesis [1]. According to the ANLSH, neural activity increases the extra cellular concentration of glutamate, whose uptake by glia stimulates Na+ - K+ ATPase and glutamate synthetase activity. This, in turn, stimulates glial glycolysis leading to the lactate production in the cytoplasm of astrocytes. Lactate accumulates in the intercellular space and it is eventually transported into the neuronal cytoplasm. This hypothesis suggests that astrocyte-produced lactate is the main oxidative substrate for neurons, particularly during neuronal activation. The formulation of a mechanistic model of the astrocyte-neuron lactate shuttle is analyzed in this paper. The model structure consists of neurons, astrocytes, and a surrounding capillary network.

The resulting dynamic model is a highly nonlinear distributed parameter system with model equations that combine several chemical species, reactions and biochemical fluxes. Due to the difficulty in physically obtaining reaction rate constants for individual biochemical reactions, flux balances are often employed to reduce the number of unknown parameters [2]. Although flux balances can be valuable predictive tools, their use with the inclusion of proper thermodynamic constraints in pathway analysis can be used to more reliably estimate biochemical reaction rate parameters, and yield a robust dynamic model [3]. The integration of these principles within the formulation and verification of a metabolic model is demonstrated in this work.

References

[1] Magistretti, P.J., L.Pellerin, D.L. Rothman, and R.G. Shulman "Energy on demand," Science 283:496497 (1999).

[2] Patel, A.B., R.A. de Graaf, G.F. Mason, T. Kanamatsu, D.L. Rothman, R.G. Shulman, and K.L. Behar "Glutamatergic neurotransmission and neuronal glucose oxidation are coupled during intense neuronal activation," J Cereb Blood Flow Metab. 24: 972-985 (2004).

[3] Beard, D.A. "A biophysical model of the mitochondrial respiratory system and oxidative phosphorylation," PLoS Comput Biol 1(4): e36 (2005).