Metabolic engineering from a cybernetic perspective. A conceptual framework

Jeffrey David Varner, Purdue University

Abstract

The cybernetic framework developed by Ramkrishna and coworkers is extended to metabolic engineering. A conceptual methodology is formulated that affords both the description of the regulation of arbitrarily complex metabolic networks as well as local genetic alterations to network structure. Additionally, a metabolic design framework is conceived in which Metabolic Control Analysis is applied to cybernetic models to yield sensitivity measurements that are directly cognizant of network regulation. This understanding is employed in the analysis of two realistic pathway systems, namely, the penicillin V biosynthetic network and the aspartate family of amino acids. In both cases cybernetic models describing the time evolution of the metabolic network are constructed and analyzed to determine likely routes genetic manipulation that yield performance enhancements. Secondly, continuous biocatalysis of heterologous products is explored. Continuous operation of recombinant microorganisms is often impossible because of genetic instability. The outward manifestation of this phenomena is the loss of insert bearing population as a function of time. From a productivity standpoint continuous reactor operation would be the preferred operational route in the absence of such instabilities. Consequently, a conceptual cybernetic model describing the biocatalysis of heterologous products is formulated and analyzed to determine qualitative strategies that can be employed to eliminate genetic instability. These strategies take the form of the rational design of plasmid characteristics and promoter configuration as well as the formulation of nonlinear reactor control systems designed to control the level of insert bearing biomass in the reactor. Lastly, we focus upon the more conceptual exercise of formulating guidelines that afford rational construction of cybernetic models. One of the conceptual roadblocks hindering the application of cybernetic models to nonlinear bioreactor control has been the issue of model formulation. The formulation of appropriately abstracted cybernetic models of microbial processes has proven difficult, in the past, using an entirely first principles approach. Accordingly, we present a set of postulates that have been formulated using tools from bifurcation theory that allow the rational formulation of cybernetic models, such that the model framework is guaranteed to possess desired behavioral characteristics. Three example models are formulated and analyzed using the abstraction procedure.

Degree

Ph.D.

Advisors

Ramkrishna, Purdue University.

Subject Area

Chemical engineering|Pharmacology|Anatomy & physiology|Animals

COinS