The accessibility of genome-scale stoichiometric models of cellular metabolism (Reed et al. , 2006) has facilitated the improvement of computational algorithms on behalf of the breakdown and drawing of composite metabolic networks. A fashionable approach is flux balance analysis (FBA) which involves a linear programming problem posed to determine the intracellular fluxes in an underdetermined stoichiometric model below the hypothesis which assume that the cell utilizes existing resources designed for development rate maximization (Stephanopoulos et al.
, 1998). FBA has been used expansively for the purpose of predicting cellular growth and product discharge patterns in microbial systems (Kauffman et al. , 2003; Sauer et al. , 1996; Segre et al. , 2002). Expansions of conventional FBA consent to the redesign of metabolic networks in favor of the overproduction of preferred metabolites in the course of gene deletions and insertions, which are put into operation by eliminating or accumulating intracellular responses to the network.
These computational techniques offer metabolic engineering objectives that can be tesed experimentally. In a current study through a Saccharomyces cerevisiae genome-scale network, the growth phenotypes of gene knockouts were envisaged with a 70 – 80 % success rate (Famili et al. , 2003). The FBA system assumes time-invariant extra-cellular circumstances and makes steady-state predictions consistent with continuous culture. Nevertheless, extensive fabrication of metabolic goods is frequently accomplished with batch and fed-batch culture.
An imperative benefit of fed-batch traditions is that substrate levels can be momentarily various to achieve a favorable tradeoff between cellular growth and product formation rates. Although batch culture experiments are often used to evaluate FBA predictions, the results are strictly valid only for the balanced growth phase. An alternative is to performcarry out metabolic network analysis and design using dynamic extensions of stoichiometric models.
Dynamic flux balance models are obtained by combining stoichiometric equations for intracellular metabolism with dynamic mass balances on key extracellular substrates and products under the assumption of fast intracellular dynamics (Gadkar et al. , 2004; Hjersted and Henson, 2006). The intracellular and extracellular descriptions are coupled through the cellular growth rate and substrate uptake kinetics, which can be formulated to include key regulatory effects such as product inhibition of growth.
Batch culture simulations with dynamic flux balance models have shown good agreement with experimental data (Sainz et al. , 2003; Varma and Palsson, 1994). Dynamic flux balance modeling offers important advantages over alternative dynamic modeling frameworks. Because simple unstructured models rely on phenomenological descriptions of cell growth and constant yield coefficients (Nielsen et al, 1994), they have no predictive capability for genetic alterations. Metabolic engineering applications of structured kinetic models (Steinmeyer and Shuler, 1989; Vaseghi et al.
, 1999), log-linear kinetic models (Hatzimanikatis et al. , 1998), and cybernetic models (Jones and Kompala, 1999; Varner and Ramkrishna, 1999) are often limited by the lack ofparameter values for in vivo enzyme kinetics. Dynamic flux balance modeling provides a practical alternative for incorporating intracellular structure. Given the availability of a steady-state flux balance model, only a small number of additional parameters are needed to account for the uptake of multiple substrates and the secretion of multiple products.
On the other hand, a well documented weakness of classical FBA is the difficulty associated with incorporating cellular regulation. This problem has been partially addressed by using gene expression data to constrain regulated fluxes within the metabolic network (Akesson et al. , 2004; Covert et al. , 2001). Dynamic flux balance analysis (DFBA) offers the additional possibility of formulating substrate uptake kinetics to account for known regulatory processes.
The production of ethanol from recombinant yeast strains has received considerable attention for renewable liquid fuel applications. Of particular interest is genetic engineering of xylose fermenting strains that can grow on media derived from agricultural products such as corn (Aristidou and Penttila, 2000; Jeffries and Jin, 2004; Kuyper et al. , 2005; Ostergaard et al. , 2000). A recent study by Bro et al. (2006) revealed novel metabolic engineering targets for improved ethanol production from glucose media based on classical FBA of a genome-scale.