Optimal knockout strategies in genome-scale metabolic networks using particle swarm optimization

Dynamic Solution Space Division-Based Methods for Calculating Reaction Deletion Strategies for Constraint-Based Metabolic Networks for Substance Production: DynCubeProd

Flux balance analysis (FBA) is a crucial method to analyze large-scale constraint-based metabolic networks and computing design strategies for strain production in metabolic engineering. However, as it is often non-straightforward to obtain such design strategies to produce valuable metabolites, many tools have been proposed based on FBA. Among them, GridProd, which divides the solution space into small squares by focusing on the cell growth rate and the target metabolite production rate to efficiently find the reaction deletion strategies, was extended to CubeProd, which divides the solution space into small cubes. However, as GridProd and CubeProd naively divide the solution space into equal sizes, even places where solutions are unlikely to exist are examined. To address this issue, we introduce dynamic solution space division methods based on CubeProd for faster computing by avoiding searching in places where the solutions do not exist. We applied the proposed method DynCubeProd to iJO1366, which is a genome-scale constraint-based model of Escherichia coli. Compared with CubeProd, DynCubeProd significantly accelerated the calculation of the reaction deletion strategy for each target metabolite production. In addition, under the anaerobic condition of iJO1366, DynCubeProd could obtain the reaction deletion strategies for almost 40% of the target metabolites that the elementary flux vector-based method, which is one of the most effective methods in existence, could not. The developed software is available on https://github.com/Ma-Yier/DynCubeProd.

Comparison of Optimization-Modelling Methods for Metabolites Production in Escherichia coli

The metabolic network is the reconstruction of the metabolic pathway of an organism that is used to represent the interaction between enzymes and metabolites in genome level. Meanwhile, metabolic engineering is a process that modifies the metabolic network of a cell to increase the production of metabolites. However, the metabolic networks are too complex that cause problem in identifying near-optimal knockout genes/reactions for maximizing the metabolite’s production. Therefore, through constraint-based modelling, various metaheuristic algorithms have been improvised to optimize the desired phenotypes. In this paper, PSOMOMA was compared with CSMOMA and ABCMOMA for maximizing the production of succinic acid in E. coli. Furthermore, the results obtained from PSOMOMA were validated with results from the wet lab experiment.

Determination of growth-coupling strategies and their underlying principles

BACKGROUND

Metabolic coupling of product synthesis and microbial growth is a prominent approach for maximizing production performance. Growth-coupling (GC) also helps stabilizing target production and allows the selection of superior production strains by adaptive laboratory evolution. To support the implementation of growth-coupling strain designs, we seek to identify biologically relevant, metabolic principles that enforce strong growth-coupling on the basis of reaction knockouts.

RESULTS

We adapted an established bilevel programming framework to maximize the minimally guaranteed production rate at a fixed, medium growth rate. Using this revised formulation, we identified various GC intervention strategies for metabolites of the central carbon metabolism, which were examined for GC generating principles under diverse conditions. Curtailing the metabolism to render product formation an essential carbon drain was identified as one major strategy generating strong coupling of metabolic activity and target synthesis. Impeding the balancing of cofactors and protons in the absence of target production was the underlying principle of all other strategies and further increased the GC strength of the aforementioned strategies.

CONCLUSION

Maximizing the minimally guaranteed production rate at a medium growth rate is an attractive principle for the identification of strain designs that couple growth to target metabolite production. Moreover, it allows for controlling the inevitable compromise between growth coupling strength and the retaining of microbial viability. With regard to the corresponding metabolic principles, generating a dependency between the supply of global metabolic cofactors and product synthesis appears to be advantageous in enforcing strong GC for any metabolite. Deriving such strategies manually, is a hard task, due to which we suggest incorporating computational metabolic network analyses in metabolic engineering projects seeking to determine GC strain designs.

A non-dominated sorting Differential Search Algorithm Flux Balance Analysis (ndsDSAFBA) for in silico multiobjective optimization in identifying reactions knockout

Metabolic engineering is defined as improving the cellular activities of an organism by manipulating the metabolic, signal or regulatory network. In silico reaction knockout simulation is one of the techniques applied to analyse the effects of genetic perturbations on metabolite production. Many methods consider growth coupling as the objective function, whereby it searches for mutants that maximise the growth and production rate. However, the final goal is to increase the production rate. Furthermore, they produce one single solution, though in reality, cells do not focus on one objective and they need to consider various different competing objectives. In this work, a method, termed ndsDSAFBA (non-dominated sorting Differential Search Algorithm and Flux Balance Analysis), has been developed to find the reaction knockouts involved in maximising the production rate and growth rate of the mutant, by incorporating Pareto dominance concepts. The proposed ndsDSAFBA method was validated using three genome-scale metabolic models. We obtained a set of non-dominated solutions, with each solution representing a different mutant strain. The results obtained were compared with the single objective optimisation (SOO) and multi-objective optimisation (MOO) methods. The results demonstrate that ndsDSAFBA is better than the other methods in terms of production rate and growth rate.

Improving designer glycan production in Escherichia coli through model-guided metabolic engineering

Asparagine-linked (N-linked) glycosylation is the most common protein modification in eukaryotes, affecting over two-thirds of the proteome. Glycosylation is also critical to the pharmacokinetic activity and immunogenicity of many therapeutic proteins currently produced in complex eukaryotic hosts. The discovery of a protein glycosylation pathway in the pathogen Campylobacter jejuni and its subsequent transfer into laboratory strains of Escherichia coli has spurred great interest in glycoprotein production in prokaryotes. However, prokaryotic glycoprotein production has several drawbacks, including insufficient availability of non-native glycan precursors. To address this limitation, we used a constraint-based model of E. coli metabolism in combination with heuristic optimization to design gene knockout strains that overproduced glycan precursors. First, we incorporated reactions associated with C. jejuni glycan assembly into a genome-scale model of E. coli metabolism. We then identified gene knockout strains that coupled optimal growth to glycan synthesis. Simulations suggested that these growth-coupled glycan overproducing strains had metabolic imbalances that rerouted flux toward glycan precursor synthesis. We then validated the model-identified knockout strains experimentally by measuring glycan expression using a flow cytometric-based assay involving fluorescent labeling of cell surface-displayed glycans. Overall, this study demonstrates the promising role that metabolic modeling can play in optimizing the performance of a next-generation microbial glycosylation platform.

Genetic Optimization Algorithm for Metabolic Engineering Revisited

To date, several independent methods and algorithms exist for exploiting constraint-based stoichiometric models to find metabolic engineering strategies that optimize microbial production performance. Optimization procedures based on metaheuristics facilitate a straightforward adaption and expansion of engineering objectives, as well as fitness functions, while being particularly suited for solving problems of high complexity. With the increasing interest in multi-scale models and a need for solving advanced engineering problems, we strive to advance genetic algorithms, which stand out due to their intuitive optimization principles and the proven usefulness in this field of research. A drawback of genetic algorithms is that premature convergence to sub-optimal solutions easily occurs if the optimization parameters are not adapted to the specific problem. Here, we conducted comprehensive parameter sensitivity analyses to study their impact on finding optimal strain designs. We further demonstrate the capability of genetic algorithms to simultaneously handle (i) multiple, non-linear engineering objectives; (ii) the identification of gene target-sets according to logical gene-protein-reaction associations; (iii) minimization of the number of network perturbations; and (iv) the insertion of non-native reactions, while employing genome-scale metabolic models. This framework adds a level of sophistication in terms of strain design robustness, which is exemplarily tested on succinate overproduction in Escherichia coli.