Biology Reference
In-Depth Information
different objectives. For example, even for
E. coli
, at least five different objective
functions, among which maximal biomass, ATP, or CO
2
production, are consistent
with the observed flux distribution (Schuetz et al.
2012
). Future research on the
“real objective” of a bacterium subjected to a given growth condition will certainly
prove fruitful for fundamental research as well as for biotechnological applications.
Several extensions of the basic FBA have been proposed. As early as 2002, the
MOMA algorithm (minimization of metabolic adjustment) explicitly addressed the
question of the objective function (Segr` et al.
2002
). For wild type strains of
E. coli
it can reasonably be argued that evolution has optimized for growth (see above).
However, in the case of genetically modified bacteria, where certain genes have
been deleted, presumably not enough time has elapsed for metabolic network
rewiring. In this context, searching for the model that minimally perturbs the fluxes
of the wild type strain has proven more accurate than the pure biomass objective
function. More recent work suggests that bacterial metabolism has been evolution-
arily optimized for two competing goals: optimality under a given condition and
minimum adjustment between conditions (Schuetz et al.
2012
).
Several variants and extensions of MFA and FBA have been developed over the
years. Metabolic pathway analysis adopts a pathway centered view and exploits
thermodynamic and biochemical constraints to limit potential metabolic strategies
(Bar-Even et al.
2012
). The metabolic network is decomposed into individual
pathways by elementary mode analysis (Trinh et al.
2009
), thereby facilitating
the analysis and prediction of cellular phenotypes, and the construction of meta-
bolic networks. Following a related philosophy, resource balance analysis exploits
the modularity of metabolic pathways and obtains good predictions for different
growth phenotypes of
Bacillus subtilis
(Goelzer et al.
2011
). Many of these variants
are summarized in Kim et al. (
2012
).
MFA and FBA are descriptions of the metabolic system that predict the fluxes of
metabolites across the system. They do not explicitly address the
control
of these
fluxes. A powerful approach for analyzing the control structure of the metabolic
system is called “metabolic control analysis” (MCA). The key concept is the
“control coefficient” that describes the ratio of the relative change of a metabolic
variable or function (such as a steady state flux or a metabolite concentration) and
the relative change of a parameter of the system (Kremling et al.
2008
; Lewis
et al.
2012
). The control coefficient can thus quantify, for example, the influence of
enzyme concentrations on the flux of a metabolic pathway. More generally, MCA is
a mathematical framework capable of relating local properties, such as enzyme
activities, to global properties, such as the response of the entire system to an
external perturbation. MCA not only provides a tool for predicting system behavior
but also gives a deeper insight into system functioning by putting the accent on the
underlying control logic. One drawback of MCA is the requirement of extensive
experimental data for determining the system parameters, i.e., the control
coefficients. Recent extensions to MCA, such as optimization-based MCA
(OMCA) (Meadows et al.
2010
), reduce the experimental burden by introducing
simplifying assumptions. OMCA supposes that the metabolic network optimizes
homeostasis. The fluxes of the network are thus correlated with metabolite
Search WWH ::
Custom Search