Development of large scale dynamic metabolic model of Penicillium chrysogenum using linlog kinetics
Integration of life sciences & engineering
Integration of Life Sciences & Engineering - Poster (T5-P)
Keywords: large scale kinetic modelling, model reduction, linlog kinetics, Penicillium chrysogenum
Biological systems present a complexity beyond intuitive comprehension and to obtain a better understanding of the behavior of the living organisms, we now use large scale dynamic mathematical models. From a system biology perspective, these models should not only describe the kinetic behavior of metabolic reaction networks that feature metabolite-enzyme interactions (allosteric feedback or feed forward), intercompartmental transport, and cofactor coupling, but, they should also ultimately allow combining several pathways (horizontal modeling) and/or “omic” levels (vertical modeling) in the cell. However, currently by far most of the available models are limited to only one pathway and one “omic” level.
In this work, we present a large scale kinetic metabolic model of Penicillium chrysogenum which is intensively studied in our laboratory. Although still focusing on one level (i.e. metabolome), such a model aims encompassing all the major pathways present in the organism. In constructing the kinetic model, we first considered the stoichiometric network presented in [1], which consists of 188 metabolites and 167 reactions located in 3 compartments (cytosol, mitochondria and peroxisome) and postulated a kinetic expression for each of the reactions. We used approximative linlog kinetics for the rate expressions, which allowed us to represent the enzyme-metabolite kinetic interactions by an elasticity matrix. Information on the presence and absence of mass action and allosteric enzyme kinetic information was obtained from literature survey and database search. The final values of the elasticities needed to be estimated by fitting the model to the available short term kinetic response data.
We encountered two major limitations on measurements of metabolites: (1) measurement of a metabolite at all and (2) compartmentation, i.e. measurements of metabolites/reactions that are present in multiple compartments [2]. To deal with the compartmentation problem, we reduced the system size to 96 metabolites and 82 reactions within one compartment, by lumping using insights gained both from biochemical knowledge and from data recently published by our group on short term kinetic responses of primary metabolism of Penicillium chrysogenum [2]. The limited number of available measurements was dealt by data-driven model reduction while applying the parameter estimation scheme to the large model. To estimate the kinetic parameters, we followed the methodology presented in [3] in which the theory was applied to a small example system.
We present the results and considerations in reducing the model from 3 to one compartment and from 167 to about 50 reactions. As such, the model describes central metabolism (glycolysis, TCA cycle, pentose phosphate pathway, storage material pathways) amino acid production and nucleotide pathways and product pathways (the biosynthesis of PenG), and represents the time dependent dynamic perturbation data presented in [2]. From the results, it is concluded that the linlog modeling framework facilitated model reduction and parameter identification of this complex system.
References
1.Van Gulik et al., 2000, Biotechnology and Bioengineering, 68(6), pp 602-618.
2.Nasution U et al., 2006, Metabolic Engineering, 8(5), 395-405.
3.Nikerel IE et al., 2006, BMC Bioinformatics, 7:540.
Presented Wednesday 19, 13:30 to 15:00, in session Integration of Life Sciences & Engineering - Poster (T5-P).
