Systems Biology and Metabolic Engineering in Bacteria - Systems Biology of Metabolic and Signaling Networks - page 351

these nonlinearities are not taken into account in current models of the yeast
metabolism or in biotechnological applications aimed at maximizing product yield.
The collective behavior of cells has been exploited in synthetic biology to
stabilize engineered circuits in a noisy cellular environment. Contrary to natural
oscillators, which are extremely robust (Toya et al. 2011 ), synthetic circuits, such as
the repressilator (Elowitz and Leibler 1999 ), are very sensitive to fluctuations of
cellular components (noise). In the case of this oscillator, daughter cells may or may
not inherit the phase and frequency of the mother cell's oscillation. Theory and
modeling tell us that one way to remedy this shortcoming is to couple many
unstable oscillators. The collection of cells should produce stable and precise
oscillations. This prediction has recently been validated experimentally in E. coli
(Prindle et al. 2012 ). These authors have constructed bacteria containing an
oscillatory circuit responding to arsenite in the medium. Individual cells were
coupled by the transmission of two diffusible molecules: a quorum-sensing mole-
cule acyl-homoserine lactone (AHL) and the redox signaling molecule H 2 O 2 ,
produced by the periodic expression of NADH dehydrogenase. The bacteria were
placed in a microfluidics device where AHL provided the intra-channel, short-range
communication and H 2 O 2 , diffusing rapidly in the gas phase, was responsible for
the coordination between channels. The device produced a very stable oscillation,
coherent across a distance of 5 mm, the frequency of which revealed the concen-
tration of arsenite in the medium. An explicit, partial differential equations model
was used for the construction and optimization of the system. Explicit, dynamical
models, taking into account nonlinear interactions in time and space, are necessary
for understanding and engineering such systems. The increasing number of stan-
dard parts for synthetic biology along with the measurement of relevant parameters
of these parts will allow modeling and construction of large, complex, nonlinear
systems such as the one described above (Prindle et al. 2012 ).
13.5 Genetic Regulation of Metabolism
The integration of metabolism with gene regulation is important for metabolic
engineering because changing metabolite concentrations affect gene expression,
which, in turn, modulates enzyme activities (Keasling 2012 ). A complete metabolic
model should therefore include the dynamics of metabolism and the connections to
the genetic network and other regulatory mechanisms. This can be achieved in the
differential equation formalism, for example, the Heinemann model mentioned
above (Heinemann and Sauer 2010 ). A similar differential equation-based
modeling approach of central nitrogen metabolism in E. coli predicts complex
response patterns of the bacterium to diverse external and internal perturbations
(Yuan et al. 2009 ). Modeling is here successfully used as a discovery tool of
hitherto unknown regulatory mechanisms.
The improved FBA formalism of (Jamshidi and Palsson 2010 ) incorporates
regulation into the steady state model, resulting in mass action stoichiometric