 |
 |
Science at the Edge Seminar Department of Engineering Friday, April 9 at 11:30am Room 1400 Biomedical and Physical Sciences Bldg. Refreshments at 11:15 Kyongbum Lee Department of Chemical and Biological Engineering, Tufts University, Medford, MA Metabolic Engineering of Cellular Energy (In)efficiency Our research seeks to better understand metabolic regulation at the cellular level, and ultimately develop strategies for rational manipulation of metabolic pathways to direct biological function. Central to our research is the idea that specialized cellular processes such as signaling, biosynthesis and biotransformation depend on and influence the metabolic state of the cell. New insights into manipulating cellular functions, whether to intervene in a disease process or to synthesize useful biomolecules may be obtained through systems-oriented studies examining the biochemical machinery of the cell as a whole. Our research encompasses tool development and hypothesis-driven investigation, with applications in biomedicine and biotechnology. One area of interest is obesity, which is one of the most pressing health concerns in the U.S. Obesity results from a chronic imbalance of nutrient intake and energy utilization at the whole body level leading to an excessive increase in body fat, or white adipose tissue. Currently available treatments are largely ineffective, including the limited number of pharmacological options whose modes of action are to influence whole body metabolism. Our work is aimed at generating new options targeting specific metabolic processes at the level of the fat cell, or adipocyte. Our approach is to establish energy inefficient, or "lean," phenotypes by systematically perturbing and characterizing the adipocyte metabolic network using both experimentation and computational analysis. Results to date, obtained from genetic modifications and chemical inhibitions, indicate that adipocyte lipid accumulation and ensuing growth can be limited by engineering inefficiencies in mitochondrial metabolism. Ongoing work builds on these findings to establish the lean phenotype in more advanced co-culture models of the adipose tissue. Future work will investigate the effects of multiple, dynamic interventions in adipose tissue metabolism using computational model-guided target selection. A second area of interest deals with developing the computational modeling tools to enable the aforementioned and other systems biological studies on cellular metabolism. Our work focuses on tools to identify functional units in the metabolic network, or modules. Modularization of a complex network provides a means to arrive at a dynamic model describing overall cellular behavior by tuning the kinetic resolution and the associated number parameters). Our partitioning algorithm exploits efficient, graph-based search algorithms and incorporates a novel metric for functional dependencies between metabolites, reactions and regulatory elements. The metric, termed Shortest Retroactive Distance, scores the reciprocal interactions between the components of a biochemical network. The algorithm modularizes the network so as to preserve reciprocal interactions. Results to date show that this algorithm produces partitions consistent with prior biological knowledge. Dynamic simulation of a partitioned metabolic network is illustrated with a model of CHO cell metabolism in fed-batch culture. Kinetic expressions are combined with a global regulator modeling the redox state to calculate cell densities and metabolite concentration profiles from a limited set of initial values and process parameters. Model simulations accurately predict the timing and magnitude of the metabolic shift, as well as changes in cell growth and recombinant protein production, in response to variations in process parameters such as temperature shift, seed density, and nutrient concentrations. This model is the first to date to account for the combined effects of temperature, redox, metabolite concentrations, and regulation.