Predictive Dynamic Control Accurately Maps the Design Space for 2,3-Butanediol Production

Abstract 2,3-Butanediol is a valuable raw material for many industries. Compared to its classical production from petroleum, novel fermentation-based manufacturing is an ecologically superior alternative. To be also economically feasible, the production bioprocesses need to be well optimized. Here, we adapted and applied a novel process optimization algorithm, dynamic control flux-balance analysis (dcFBA), for 2,3-butanediol production in E. coli . First, we performed two-stage fed-batch process simulations with varying process lengths. There, we found that the solution space can be separated into a proportionality and a trade-off region. With the information of the simulations we were able to design close-to-optimal production processes for maximizing titer and productivity, respectively. Experimental validations resulted in a titer of 43.6±9.9 g L −1 and a productivity of 1.93 ± 0.08 g L −1 h −1 . Subsequently, we optimized a continuous two-reactor process setup for 2,3-butanediol productivity. We found that in this mode, it is possible to increase the productivity more than threefold with minor impact on the titer and yield. Biotechnological process optimization is cumbersome, therefore, many processes are run in suboptimal conditions. We are confident that the methods presented here, will help to make many biotechnological productions economically feasible in the future. Highlights Precise simulations are used to sample the process solution space. Our simulations uncover big productivity potential in the 2,3-butanediol production. Experiments validate the predictions and show a 2,3-butanediol productivity improvement of 104 %.