Metabolic Engineering of Tissue Function

Abstract

Advances in the ability to grow, maintain, and manipulate cells and organs outside the body have opened new possibilities for building bio-artificial tissue devices to treat organ or tissue dysfunction. Moreover, engineered tissue constructs are ideal settings for detailed studies on cellular processes under highly controlled, yet physiologically relevant conditions. Our research is ultimately aimed at better understanding cellular metabolic regulation through the application of principles and techniques drawn from life sciences, physical sciences, and engineering. Of central importance is the notion that cellular activity is accomplished through the concerted actions of a network of biochemical reactions interconnected through shared intermediates, co-factors, and other regulatory features. Therefore, improved understanding of cellular metabolism can be achieved by considering the component reactions together, through comprehensive experimental observations guided by mathematical models of network regulation and process control.

One application of the above ideas is in the area of liver tissue engineering. Organ transplantation is the current standard for management of liver disease, but consistently increasing donor shortage has motivated the development of alternative treatment options. The objective of our liver research is to develop and implement a metabolic engineering strategy for improving the function of bio-artificial tissue constructs. The basic premise for this work is that the healthy liver is a multi-functional organ whose biochemical properties under normal and pathological conditions need to be understood at the systems level.

Recently, we have used the isolated perfused rat liver in conjunction with a non-lethal burn model to obtain liver-specific metabolic profiles at various times after an inflammatory stress. Projection of the metabolic profiling data onto a reduced-dimensional space by discriminant analysis showed a time-dependent evolution of distinct metabolic states. Using a stoichiometric network model, thermodynamic constraints, and optimization modeling, we found that the evolution of metabolic states is driven by changes to the distribution of metabolic fluxes and the corresponding biochemical objectives. The computed objective functions match the putative time course of liver stress response, which has been reported in the literature only in qualitative terms. In addition to the multi-objective optimization analysis, we have also used linear programming (simplex method) to predict hypothetical metabolic flux distributions within the liver in cases where the objective functions maximizes the flux through only one randomly chosen reaction in the metabolic network. The results of this naive optimization analysis were compared with data-derived flux distributions. In contrast to the multi-objective case, the simulated profiles from the single-objective cases correlated poorly with the data-derived cases, further corroborating the notion that liver simultaneously performs multiple metabolic objectives. On-going work is aimed at generalizing the above findings across multiple types of stresses and experimental model systems.