Abstract
The phosphotransferase system (PTS) is the sugar transportation machinery that is widely distributed in prokaryotes and is critical for enhanced production of useful metabolites. To increase the glucose uptake rate, we propose a rational strategy for designing the molecular architecture of the Escherichia coli glucose PTS by using a computer‐aided design (CAD) system and verified the simulated results with biological experiments. CAD supports construction of a biochemical map, mathematical modeling, simulation, and system analysis. Assuming that the PTS aims at controlling the glucose uptake rate, the PTS was decomposed into hierarchical modules, functional and flux modules, and the effect of changes in gene expression on the glucose uptake rate was simulated to make a rational strategy of how the gene regulatory network is engineered. Such design and analysis predicted that the mlc knockout mutant with ptsI gene overexpression would greatly increase the specific glucose uptake rate. By using biological experiments, we validated the prediction and the presented strategy, thereby enhancing the specific glucose uptake rate. ### Synopsis The goals of systems biology are to understand the mechanisms of how biochemical networks generate particular cellular functions in response to environmental stresses or genetic changes, and to rationally design these molecular processes to meet an engineering purpose. To design biological systems at the molecular interaction level, it is essential to identify a biochemical network map, to build a dynamic model of the system, and to perform system analysis. Perturbation analysis is useful for identifying critical parameters that affect the system's performance. CAD is now a key technology to simulate or design the molecular architecture of a genetically engineered cell ([Kurata et al , 2005][1]). Furthermore, module‐based analysis has been presented that decomposes a biochemical network map into hierarchical modules: functional and flux modules, in a manner analogous to control engineering architectures ([Kurata et al , 2006][2]). If an engineering purpose is provided for biochemical networks, functional modules are readily determined and an engineering function is assigned to each module. This greatly facilitates rational design of a genetic manipulation responsible for achieving a purpose and an intuitive understanding of how the biochemical network of interest can be improved or modified. In the field of industrial production of useful substances by microorganisms, Glc is one of the most important substrates necessary for substance production or cell growth. E. coli is a commonly used microbe in the industrial field, including the manufacture of amino acids ([Ikeda, 2003][3]) and organic acids ([Wendisch et al , 2006][4]). To enhance the ability to utilize substrates is of critical importance to increase the productivity of useful substances from the viewpoint of economic impacts. In this study, to increase the glucose uptake rate we rationally designed the molecular architecture of the E. coli glucose PTS by using a CAD and verified the simulated results with biological experiments, as shown in [Figure 1][5]. The presented strategy consisted of construction of a biochemical network, dynamic modeling, module‐based and perturbation analyses to find critical genes responsible for enhancing glucose‐uptake rate, and experimental validation. We present a biochemical map of the E. coli glucose PTS ([Figure 2][6]) and developed its dynamic model by using CADLIVE ([Kurata et al , 2005][1]). To assign a particular function to each gene in terms of an enhanced glucose uptake rate, the biochemical map of the glucose PTS was decomposed into functional modules ([Figure 2A][6]) and flux modules ([Figure 2B][6]) analogous to engineering control‐systems block diagrams ([Kurata et al , 2006][2]). Specifically, if we define the phosphotransfer cascade of IICBGlc ( ptsG ), IIAGlc ( crr ), HPr ( ptsH ), and EI ( ptsI ) as the plant to be controlled, then the quantity that drives this plant is the signal from the accelerator actuator. The phosphorylation signal of IIAGlc‐P is sensed by the FB sensor module and sent to the computer module that consists of CYA, cAMP, and CRP. The complex of IIAGlc‐P:CYA, which is formed in the FB sensor module, produces the cAMP signal in the computer module. This signal molecule modulates the accelerator and brake actuator modules. The accelerator actuator module comprises the synthesis process of the IICBGlc, HPr, and EI proteins, supplying these PTS proteins to the plant module; the brake actuator module of the Mlc protein suppresses the synthesis of the same PTS proteins. Superimposed on these functional modules, two major flux modules have been identified: the accelerator flux module and the brake flux module. In the accelerator flux module the signal of the IIAGlc‐P increase goes through the FB sensor, computer, accelerator actuator, and plant modules to induce the synthesis of IICBGlc, HPr, and EI ([Figure 2B][6]). In the brake flux module the signal of IIAGlc‐P increase goes through the FB sensor, computer, brake actuator, accelerator actuator, and plant modules, which suppresses the accelerator actuator. The functional and flux module decomposition readily provides a rational strategy for enhanced glucose uptake. It is critical for enhanced glucose uptake to increase the quantity of PTS proteins in the plant module (the ptsG , ptsH , ptsI , and crr genes), to delete the brake actuator module (the mlc gene), and to enhance the accelerator actuator module through the computer module (the crp gene). To further explore target genes critically responsible for enhancing glucose uptake, we performed perturbation analysis of crp , mlc , ptsG , ptsH , ptsI , and crr gene expression. The specific glucose uptake rates were simulated with respect to a 10‐fold increase in the...

This publication has 54 references indexed in Scilit: