Chalmers Conferences, 9th European Conference on Mathematical and Theoretical Biology

Dynamic Flux Balance Analysis for Modeling Lactococcus lactis Mannitol Production
Armando Fernandes, Rafael Costa, Paula Gaspar, Ana Neves, Susana Vinga

Last modified: 2014-03-27


It is believed that mannitol has health protecting capabilities which make it an added-value compound. In fact, it has been reported that mannitol can be converted into fatty acids that are known for their ability to protect humans from colon cancer. Mannitol can be produced by Lactococcus lactis, which is a bacteria widely used in the production of dairy products such as cheese. This is very interesting because it allows including healthy components into food directly during fermentation. Consequently, there is a need to optimize the production of mannitol, which can be supported computationally through the creation of mathematical models that allow testing new hypotheses and obtain new predictions. Lactococcus lactis is adequate for this task due to its small genome, simple metabolism and availability of genome-scale metabolic models.

A report will be given on the results and difficulties of an ongoing work of creating a dynamic Flux Balance Analysis (dFBA) model capable of replicating experimental dynamic curves of mannitol production in Lactococcus lactis MG1363. The authors do not know of any model of dFBA for mannitol production. Dynamic computational models are frequently used to predict temporal variations of metabolite concentrations, but these models have the disadvantage of using a large number of chemical reaction parameters. Conventional FBA operates in steady-state meaning that the variation of metabolite concentration with time is zero, which makes it impossible to analyse transient problems. With dFBA, as described in Appl. Environ. Microbiol.,1994, 60(10):3724, it is possible to analyse the dynamic behavior of a system by solving a conventional FBA problem for a succession of time step with the objective of determining the flux of external metabolites in each of the steps. Once the fluxes are determined at a time step it is possible to update the concentration of the external metabolites so that new concentration values may be used in a new time step. In the dFBA developed the flux at each time step may be set, for a small number of external metabolites, according to experimental profiles measured. The experimental data available is extracted from Appl. Environ. Microbiol.,2004, 70(3):1466, where time curves for glucose, formate, ethanol, acetate, acetoin, 2,3-butanediol, lactate, and mannitol are provided. Two approaches will be followed in the dFBA simulations, one will employ the full genome-scale model of Lactococcus lactis and the other will try to use a smaller metabolic network, for which faster and simpler analysis can be conducted. Simulations are run in COnstraints Based Reconstruction and Analysis (COBRA) toolbox for Matlab, an optimization software that is widely used in the FBA community.

The present work develops and tests methods that will later on be used for the optimization of malonyl-coenzyme-A production, an anthocyanin precursor. Anthocyanins commercial value makes them the focus of the European project “BacHBerry- BACterial Hosts for production of Bioactivephenolics from bERRY fruits”.


Lactococcus Lactis; Mannitol; FBA; dynamic FBA;