Biomedical Engineering Reference
In-Depth Information
6 Conclusions
In a typical cell culture process there is a large number of environmental variables
that shape cellular physiology. One important implication is that the design space
for process development, namely culture medium optimization and process con-
trol, is potentially very large. Current process development methodologies in the
industry are essentially of empirical nature. Empirical methods are not well suited
to handle high-dimensional design spaces unless a substantial level of reduction-
ism is applied, and even then with potential reduction of performance.
With the advances in systems biology, accurate genome-scale metabolic net-
works are becoming available for several microorganisms used in industry. Such
metabolic networks contain the required information to enumerate all the opera-
tional modes of cells (i.e., elementary modes). With adequate systems biology
tools such as functional enviromics, one can learn how such operational modes are
controlled by the environment and/or how they modify the environment. This
paves the way for pathway-level process development strategies, which are much
more efficient than traditional empirical methods.
Here, we have laid out a process development methodology that can be sum-
marized in the following main steps
(i)
Formulation of an accurate (genome-scale) metabolic network
(ii)
Computation of the elementary modes and pre-reduction of the same
(iii)
Discrimination of elementary modes with high correlation with environ-
mental variables by functional enviromics
(iv)
Formulation of macroscopic material balances with explicit envirome-cor-
related elementary modes
(v)
Process
optimization
oriented
to
the
manipulation
of
elementary
mode
weighting factors
Such design tools can be used to optimize culture media and for advanced
process control. A main advantage is the significant reduction of the number of
experiments for very large design spaces. This is possible because the structure of
the metabolic network constrains the manipulation of the environment. Another
big benefit is the possibility to target intracellular control variables such as met-
abolic reactions or metabolic pathways directly linked with productivity and
product quality. All in all, such techniques have the potential to considerably
accelerate process development speed, to improve the mechanistic interpretability,
and to increase process performance.
References
1. Bastin G, Dochain D (1990) On-line estimation and adaptive control of bioreactors. Elsevier,
Amsterdam
2. Otero JM, Nielsen J (2010) Industrial systems biology. Biotechnol Bioeng 105:439-460
Search WWH ::
Custom Search