(626g) Optimizing Metabolic Dynamics for Increased Metabolite Production | AIChE

(626g) Optimizing Metabolic Dynamics for Increased Metabolite Production


Sowa, S. - Presenter, University of Texas at Austin
Baldea, M., The University of Texas at Austin
Contreras, L. M., University of Texas at Austin

Optimizing Metabolic Dynamics for Increased Metabolite Production

Steven Sowa1, M. Baldea2, L. Contreras2

1Microbiology Graduate Program, University of Texas at Austin, 100 E. 24th Street, STOP A6500
Austin, Texas 78712
2McKetta Department of Chemical Engineering, University of Texas at Austin, 200 E. Dean Keeton St., Stop C0400, Austin TX 78712

The development of in silico metabolic models has received extensive attention in the past decades. Static models, which are at present well understood and have been extensively validated for several organisms, have proven effective at screening and identifying genetic knockout strains that maximize production of a desired metabolite. More recent research efforts have focused on constructing and validating dynamic models, capable of predicting the transient behavior of specific pathways or entire organisms.
In the present work, we describe a new avenue for exploiting the newly acquired knowledge on dynamics of cellular metabolism. Specifically, we propose to tune the rates of metabolic reactions by varying relevant enzyme concentrations over time, with the objective of maximizing production of targeted metabolites. Equivalently, we seek time-varying enzyme concentration levels that can direct a wild-type cell towards optimal metabolic state on par with those that are engineered via traditional metabolic engineering approaches (e.g., gene knockouts). We formulate this concept as a dynamic optimization problem over a finite time horizon, using the total production of the metabolite of interest as the objective function, and the time-varying concentrations of the cellular enzymes as decision variables. We propose solving this problem using a feasible path approach, discretizing the decision variables using either piece-wise continuous functions or sinusoids with user-defined periods and amplitudes.
We illustrate this theoretical framework using an established kinetic model of Escherichia coli central carbon metabolism (Chassagnole 2002). We demonstrate that the production of phosenolpyruvate (PEP) can be significantly increased by optimizing specific sets of time-dependent enzyme levels. The optimal enzyme levels vary periodically in time, and we link this finding to classical results in chemical reaction engineering (Kevrekidis 1986, Ozgulsen 1993), which indicate that the time-integral of the quantity of a reaction product can be increased by operating the reactor in an unsteady regime. We conclude by discussing an experimental approach for tuning existing cellular regulatory networks and by validating these theoretical results.

Chassagnole, C., Noisommit-Rizzi, N., Schmid, J. W., Mauch, K. & Reuss, M. Dynamic modeling of the central carbon metabolism of Escherichia coli. Biotechnology and Bioengineering 79, 53-73, doi:10.1002/bit.10288 (2002).
I. G. Kevrekidis, L. D. Schmidt, and R. Aris, Some common features of periodically forced reacting systems. Chemical engineering science 41, 1263 (1986).
S. J. K. Ozgulsen, A. Cinar, Nonlinear predictive control of periodically forced chemical reactors. AIChE J. 39, 589 ( 1993).