PT  - JOURNAL ARTICLE
AU  - Peng Xu
TI  - Branch point control at malonyl-CoA node: A computational framework to optimize the controller architecture toward ideal metabolic switches
AID  - 10.1101/847947
DP  - 2019 Jan 01
TA  - bioRxiv
PG  - 847947
4099  - http://biorxiv.org/content/early/2019/11/20/847947.short
4100  - http://biorxiv.org/content/early/2019/11/20/847947.full
AB  - Living organisms are an intelligent system consisting of sensors, transducers and actuators to perform precisely-controlled metabolic functions. With a better understanding of cellular regulation, metabolic engineers have been able to engineer both the chemistry modules (the mass flow) and the control modules (the information flow) inside the cell to design intelligent cell factories with improved performance. Biophysical models are important tools to understand genetic circuit dynamics, metabolic network constraints, and microbial consortia interactions. Based on a previously engineered malonyl-CoA switch (Xu et al, PNAS 2014), we have formulated nine differential equations to unravel the design principles underlying an ideal metabolic switch. By interrogating the physiologically accessible parameter space, we have determined the optimal control architecture to configure both the metabolic source pathway and metabolic sink pathway. We identified a number of biological parameters that strongly impact the system dynamics. We determined that low protein degradation rate, medium strength of metabolic inhibitory constant, high metabolic source pathway induction rate, strong TF-UAS (transcriptional factor-upstream activation sequence) binding affinity for the metabolic source pathway, weak TF-operator binding affinity for the metabolic sink pathway, and a strong cooperative repression of metabolic sink pathway by TF benefit the accumulation of the target molecule. The target molecule production is increased from 50% to 10-folds upon application of the metabolic switch. With strong metabolic inhibitory constant, the system displays hysteresis and multiplicity of steady states. Stable oscillation of metabolic intermediate is the driving force to allow the system deviate from its equilibrium state and permits alternating ON-OFF gene expression control of both the metabolic source and metabolic sink pathways. The computational framework may facilitate us to design and engineer predictable genetic-metabolic switches, quest for the optimal controller architecture of the metabolic source/sink pathways, as well as reshape metabolic function for diverse biotechnological and medical applications.Appendix: Symbols and variables used in this workμspecific growth rateμmaxmaximum specific growth rateα1cell growth-associated FapR production rate constant (constitutive expression)α2cell growth-associated FAS production rate constant (leaky expression)α3cell growth-associated ACC production rate constant (leaky expression)α4cell growth-associated PDH production rate constant (constitutive expression)β1non cell growth-associated FAS production rate (regulated expression)β2non cell growth-associated ACC production rate (regulated expression)K1Malonyl-CoA inhibitory (dissociation) constantK2Mal-CoA and FapR saturation constantK3dissociation rate constant of free FapR toward fapO in the FAS operon (to repress FAS transcription)K4dissociation rate constant of free FapR toward UAS in the ACC operon (to activate ACC transcription)K5acetyl-CoA saturation (Michaelis) constant toward ACCK6glucose saturation (Michaelis) constant toward glycolytic pathwayKsMonod constant for glucoseKmMalonyl-CoA saturation (Michaelis) constant toward FASk1FapR-inactivating rate constant due to the formation of MalCoA-FapR complexk2FA (fatty acids) production rate constant from Mal-CoA catalyzed by FASk3malonyl-CoA production rate constant from acetyl-CoA catalzyed by ACCk4acetyl-CoA production rate constant from glycolysis catalzyed by PDHSglucose concentrationS0glucose concentration in the feeding streamDdilution rate or degradation rateX0biomass concentration in the feeding streamYPS1malonyl-CoA to fatty acids conversion yieldYXSglucose to biomass conversion yieldYPS2glucose to acetyl-CoA conversion yieldmmalonyl-CoA-FapR (ligand-TF) Hill cooperativity coefficientnFapR-FapO nucleoprotein complex Hill cooperativity coefficientpFapR-UAS nucleoprotein complex Hill cooperativity coefficientqmalonyl-CoA-FAS (substrate-enzyme) Hill cooperativity coefficientracetyl-CoA-ACC (substrate-enzyme) Hill cooperativity coefficientuglucose-PDH (substrate-enzyme, artificial reaction) Hill cooperativity coefficient