6.2 Computational
Modeling of
Fermentation
Processes
Biomass conversion is a critical step in biorefineries. In the
biological conversion pathway, once the set of feedstocks possibly
available for fermentation is identified, biorefinery designers need
to select the organisms for the fermentation process. Natural feedstock biomass is composed of various different molecules, such as
monosugars (glucose, galactose, rhamnose, xylose, etc.), amino
acids (valine, histidine, lysine, etc.), fatty acids (myristic, oleic,
palmitic, etc.), fibers (cellulose, hemicellulose, lignin, ulvan, etc.),
and others. Thus, selection of the fermenting organism is not a
trivial issue, since most of the organisms cannot metabolize part of
the existing biomass components, leading to significant amount of
residual media and low fermentation efficiency. For example, wildtype Saccharomyces cerevisiae, which is a first choice organism for
bioethanol production, poorly utilizes carbohydrates such as xylose,
rhamnose, and galactose. One approach to overcoming this deficiency and thereby to improving the bioethanol yields is to genetically modify S. cerevisiae to improve sugar uptake mechanisms.
Studies in this direction are undertaken for several years but successful implementation remains an open challenge [62]. Another
approach is to induce or to increase the required functionality in the
organism, leading to broader digestion ability. For example, BondWatts et al. [63] proposed different plasmid inserts into Escherichia
coli to introduce butanol-producing pathways. However, broad
digestion ability of some single organisms reduces the total yields
of desired products. For example, E. coli is less efficient in production of ethanol from glucose than S. cerevisiae [64, 65].
Fermentation by bacterial communities is a natural alternative
to genetic modifications of selected organisms. Community members can be selected to naturally digest the broader range of existing
biomass compounds and further convert them to desired products.
However, fermentation by communities has some serious drawbacks, such as the need for expertise in growing of several organisms, and the understanding of inter-organism interactions and of
competition for resources. Mathematical modelling of the
community-based fermentation process is also further complicated,
since the natural inter-organism interactions are still poorly understood [66] (relatively to intra-organism metabolism) and since
mapping of metabolites between models representing individual
organisms is not trivial [67, 68]. There are several approaches for
mathematical modeling of the community behavior. For example,
OptCom [66] is a methodology that proposes a computational
framework to describe different intraspecies interactions grown
together. This framework aims to describe trade-offs between individual vs. community level fitness criteria. Similarly, cFBA [69] is a
method that integrates interspecies interactions to achieve maximal
growth rate of the entire bacterial community, while SUMEX [70]
performs the same task by maximizing the total molar output
exchange minus input exchange of metabolites. Another approach
26
Alexander Golberg et al.
Modeling of
Fermentation
Processes
Biomass conversion is a critical step in biorefineries. In the
biological conversion pathway, once the set of feedstocks possibly
available for fermentation is identified, biorefinery designers need
to select the organisms for the fermentation process. Natural feedstock biomass is composed of various different molecules, such as
monosugars (glucose, galactose, rhamnose, xylose, etc.), amino
acids (valine, histidine, lysine, etc.), fatty acids (myristic, oleic,
palmitic, etc.), fibers (cellulose, hemicellulose, lignin, ulvan, etc.),
and others. Thus, selection of the fermenting organism is not a
trivial issue, since most of the organisms cannot metabolize part of
the existing biomass components, leading to significant amount of
residual media and low fermentation efficiency. For example, wildtype Saccharomyces cerevisiae, which is a first choice organism for
bioethanol production, poorly utilizes carbohydrates such as xylose,
rhamnose, and galactose. One approach to overcoming this deficiency and thereby to improving the bioethanol yields is to genetically modify S. cerevisiae to improve sugar uptake mechanisms.
Studies in this direction are undertaken for several years but successful implementation remains an open challenge [62]. Another
approach is to induce or to increase the required functionality in the
organism, leading to broader digestion ability. For example, BondWatts et al. [63] proposed different plasmid inserts into Escherichia
coli to introduce butanol-producing pathways. However, broad
digestion ability of some single organisms reduces the total yields
of desired products. For example, E. coli is less efficient in production of ethanol from glucose than S. cerevisiae [64, 65].
Fermentation by bacterial communities is a natural alternative
to genetic modifications of selected organisms. Community members can be selected to naturally digest the broader range of existing
biomass compounds and further convert them to desired products.
However, fermentation by communities has some serious drawbacks, such as the need for expertise in growing of several organisms, and the understanding of inter-organism interactions and of
competition for resources. Mathematical modelling of the
community-based fermentation process is also further complicated,
since the natural inter-organism interactions are still poorly understood [66] (relatively to intra-organism metabolism) and since
mapping of metabolites between models representing individual
organisms is not trivial [67, 68]. There are several approaches for
mathematical modeling of the community behavior. For example,
OptCom [66] is a methodology that proposes a computational
framework to describe different intraspecies interactions grown
together. This framework aims to describe trade-offs between individual vs. community level fitness criteria. Similarly, cFBA [69] is a
method that integrates interspecies interactions to achieve maximal
growth rate of the entire bacterial community, while SUMEX [70]
performs the same task by maximizing the total molar output
exchange minus input exchange of metabolites. Another approach
26
Alexander Golberg et al.