Maximization of non-nitrogenous metabolite production in E. coli using population systems biology
The E. coli metabolome is an interconnected set of enzymes that has measurable kinetic parameters ascribed for the production of most of its metabolites. Flux Balance Analysis (FBA) or Ordinary Differential Equation (ODE) models are used to simulate and increase product yield augmented with rich carbon sources. However, such simulations either give a range (min-max) of the metabolite yield for FBA model or tries to enumerate the exact production in the ODE models. The transcriptome expression
... iptome expression diversity of the individual cells is not taken into account. We formulate the metabolic behaviour of individual cells by using a robust POpulation SYstem Biology ALgorithm (POSYBAL). This allowed for predicted multiple gene knockouts for increasing homologous metabolite like shikimate and heterologous metabolite like isobutanol. We demonstrate the performance one such triple knockout prediction viz. adhE, ackA and ldhA for isobutanol and aroK, aroA and aroL triple knockout for shikimate. The isobutanol yield increased by 40 times in the knockouts (>2000 ppm) compared to 50ppm produced in the wild-type, and the shikimate yield was increased to 42 times, i.e. from 58 to ~2100ppm. This formulation was also based on the additional concept of 'Nitrogen Swapping' where cells were grown in standard multi-component media during the growth phase and then swapped in low Nitrogen media in the production phase. This swap redistributed the flux distribution such that it flowed primarily through non-nitrogenous pathway such as to maximize the metabolites like shikimate and isobutanol that lack the element Nitrogen in its constituent. Further analysis of various feature of the prediction of the POSYBAL model indicates that even under normal glucose uptake the bacterial cell population diverges into rapidly growing and nearly non-growing cells thus increasing its diversity and hence robustness under any antibacterial attack. This feature is discussed from an evolutionary standpoint.