Large-scale bi-level strain design approaches and mixed-integer programming solution techniques

The use of computational models in metabolic engineering has been increasing as more genome-scale metabolic models and computational approaches become available. Various computational approaches have been developed to predict how genetic perturbations affect metabolic behavior at a systems level, and have been successfully used to engineer microbial strains with improved primary or secondary metabolite production. However, identification of metabolic engineering strategies involving a large number of perturbations is currently limited by computational resources due to the size of genome-scale models and the combinatorial nature of the problem. In this study, we present (i) two new bi-level strain design approaches using mixed-integer programming (MIP), and (ii) general solution techniques that improve the performance of MIP-based bi-level approaches. The first approach (SimOptStrain) simultaneously considers gene deletion and non-native reaction addition, while the second approach (BiMOMA) uses minimization of metabolic adjustment to predict knockout behavior in a MIP-based bi-level problem for the first time. Our general MIP solution techniques significantly reduced the CPU times needed to find optimal strategies when applied to an existing strain design approach (OptORF) (e.g., from ～10 days to ～5 minutes for metabolic engineering strategies with 4 gene deletions), and identified strategies for producing compounds where previous studies could not (e.g., malate and serine). Additionally, we found novel strategies using SimOptStrain with higher predicted production levels (for succinate and glycerol) than could have been found using an existing approach that considers network additions and deletions in sequential steps rather than simultaneously. Finally, using BiMOMA we found novel strategies involving large numbers of modifications (for pyruvate and glutamate), which sequential search and genetic algorithms were unable to find. The approaches and solution techniques developed here will facilitate the strain design process and extend the scope of its application to metabolic engineering.     

A Genetic Algorithm for the Bi-Level Topological Design of Local Area Networks

Local access networks (LAN) are commonly used as communication infrastructures which meet the demand of a set of users in the local environment. Usually these networks consist of several LAN segments connected by bridges. The topological LAN design bi-level problem consists on assigning users to clusters and the union of clusters by bridges in order to obtain a minimum response time network with minimum connection cost. Therefore, the decision of optimally assigning users to clusters will be made by the leader and the follower will make the decision of connecting all the clusters while forming a spanning tree. In this paper, we propose a genetic algorithm for solving the bi-level topological design of a Local Access Network. Our solution method considers the Stackelberg equilibrium to solve the bi-level problem. The Stackelberg-Genetic algorithm procedure deals with the fact that the follower’s problem cannot be optimally solved in a straightforward manner. The computational results obtained from two different sets of instances show that the performance of the developed algorithm is efficient and that it is more suitable for solving the bi-level problem than a previous Nash-Genetic approach.     

Robust mutant strain design by pessimistic optimization

Background

Flux Balance Analysis (FBA) based mathematical modeling enables in silico prediction of systems behavior for genome-scale metabolic networks. Computational methods have been derived in the FBA framework to solve bi-level optimization for deriving “optimal” mutant microbial strains with targeted biochemical overproduction. The common inherent assumption of these methods is that the surviving mutants will always cooperate with the engineering objective by overproducing the maximum desired biochemicals. However, it has been shown that this optimistic assumption may not be valid in practice.

Methods

We study the validity and robustness of existing bi-level methods for strain optimization under uncertainty and non-cooperative environment. More importantly, we propose new pessimistic optimization formulations: P-ROOM and P-OptKnock, aiming to derive robust mutants with the desired overproduction under two different mutant cell survival models: (1) ROOM assuming mutants have the minimum changes in reaction fluxes from wild-type flux values, and (2) the one considered by OptKnock maximizing the biomass production yield. When optimizing for desired overproduction, our pessimistic formulations derive more robust mutant strains by considering the uncertainty of the cell survival models at the inner level and the cooperation between the outer- and inner-level decision makers. For both P-ROOM and P-OptKnock, by converting multi-level formulations into single-level Mixed Integer Programming (MIP) problems based on the strong duality theorem, we can derive exact optimal solutions that are highly scalable with large networks.

Results

Our robust formulations P-ROOM and P-OptKnock are tested with a small E. coli core metabolic network and a large-scale E. coli iAF1260 network. We demonstrate that the original bi-level formulations (ROOM and OptKnock) derive mutants that may not achieve the predicted overproduction under uncertainty and non-cooperative environment. The knockouts obtained by the proposed pessimistic formulations yield higher chemical production rates than those by the optimistic formulations. Moreover, with higher uncertainty levels, both cellular models under pessimistic approaches produce the same mutant strains.

Conclusions

In this paper, we propose a new pessimistic optimization framework for mutant strain design. Our pessimistic strain optimization methods produce more robust solutions regardless of the inner-level mutant survival models, which is desired as the models for cell survival are often approximate to real-world systems. Such robust and reliable knockout strategies obtained by the pessimistic formulations would provide confidence for in-vivo experimental design of microbial mutants of interest.

     

Metabolic networks are NP-hard to reconstruct

High-throughput data from various omics and sequencing techniques have rendered the automated metabolic network reconstruction a highly relevant problem. Our approach reflects the inherent probabilistic nature of the steps involved in metabolic network reconstruction. Here, the goal is to arrive at networks which combine probabilistic information with the possibility to obtain a small number of disconnected network constituents by reduction of a given preliminary probabilistic metabolic network. We define automated metabolic network reconstruction as an optimization problem on four-partite graph (nodes representing genes, enzymes, reactions, and metabolites) which integrates: (1) probabilistic information obtained from the existing process for metabolic reconstruction from a given genome, (2) connectedness of the raw metabolic network, and (3) clustering of components in the reconstructed metabolic network. The practical implications of our theoretical analysis refer to the quality of reconstructed metabolic networks and shed light on the problem of finding more efficient and effective methods for automated reconstruction. Our main contributions include: a completeness result for the defined problem, polynomial-time approximation algorithm, and an optimal polynomial-time algorithm for trees. Moreover, we exemplify our approach by the reconstruction of the sucrose biosynthesis pathway in Chlamydomonas reinhardtii.     

Generation of an atlas for commodity chemical production in Escherichia coli and a novel pathway prediction algorithm,GEM-Path

The production of 75% of the current drug molecules and 35% of all chemicals could be achieved through bioprocessing (Arundel and Sawaya, 2009). To accelerate the transition from a petroleum-based chemical industry to a sustainable bio-based industry, systems metabolic engineering has emerged to computationally design metabolic pathways for chemical production. Although algorithms able to provide specific metabolic interventions and heterologous production pathways are available, a systematic analysis for all possible production routes to commodity chemicals in Escherichia coli is lacking. Furthermore, a pathway prediction algorithm that combines direct integration of genome-scale models at each step of the search to reduce the search space does not exist. Previous work (Feist et al., 2010) performed a model-driven evaluation of the growth-coupled production potential for E. coli to produce multiple native compounds from different feedstocks. In this study, we extended this analysis for non-native compounds by using an integrated approach through heterologous pathway integration and growth-coupled metabolite production design. In addition to integration with genome-scale model integration, the GEM-Path algorithm developed in this work also contains a novel approach to address reaction promiscuity. In total, 245 unique synthetic pathways for 20 large volume compounds were predicted. Host metabolism with these synthetic pathways was then analyzed for feasible growth-coupled production and designs could be identified for 1271 of the 6615 conditions evaluated. This study characterizes the potential for E. coli to produce commodity chemicals, and outlines a generic strain design workflow to design production strains.     

Divergent polo box domains underpin the unique kinetoplastid kinetochore

Kinetochores are macromolecular machines that drive eukaryotic chromosome segregation by interacting with centromeric DNA and spindle microtubules. While most eukaryotes possess conventional kinetochore proteins, evolutionarily distant kinetoplastid species have unconventional kinetochore proteins, composed of at least 19 proteins (KKT1–19). Polo-like kinase (PLK) is not a structural kinetochore component in either system. Here, we report the identification of an additional kinetochore protein, KKT20, in Trypanosoma brucei. KKT20 has sequence similarity with KKT2 and KKT3 in the Cys-rich region, and all three proteins have weak but significant similarity to the polo box domain (PBD) of PLK. These divergent PBDs of KKT2 and KKT20 are sufficient for kinetochore localization in vivo. We propose that the ancestral PLK acquired a Cys-rich region and then underwent gene duplication events to give rise to three structural kinetochore proteins in kinetoplastids.     

EasyCloneMulti: A Set of Vectors for Simultaneous and Multiple Genomic Integrations in Saccharomyces cerevisiae

Saccharomyces cerevisiae is widely used in the biotechnology industry for production of ethanol, recombinant proteins, food ingredients and other chemicals. In order to generate highly producing and stable strains, genome integration of genes encoding metabolic pathway enzymes is the preferred option. However, integration of pathway genes in single or few copies, especially those encoding rate-controlling steps, is often not sufficient to sustain high metabolic fluxes. By exploiting the sequence diversity in the long terminal repeats (LTR) of Ty retrotransposons, we developed a new set of integrative vectors, EasyCloneMulti, that enables multiple and simultaneous integration of genes in S. cerevisiae. By creating vector backbones that combine consensus sequences that aim at targeting subsets of Ty sequences and a quickly degrading selective marker, integrations at multiple genomic loci and a range of expression levels were obtained, as assessed with the green fluorescent protein (GFP) reporter system. The EasyCloneMulti vector set was applied to balance the expression of the rate-controlling step in the β-alanine pathway for biosynthesis of 3-hydroxypropionic acid (3HP). The best 3HP producing clone, with 5.45 g.L^-1 of 3HP, produced 11 times more 3HP than the lowest producing clone, which demonstrates the capability of EasyCloneMulti vectors to impact metabolic pathway enzyme activity.     

Metabolic engineering of Escherichia coli for direct production of 1,4-butanediol

1,4-Butanediol (BDO) is an important commodity chemical used to manufacture over 2.5 million tons annually of valuable polymers, and it is currently produced exclusively through feedstocks derived from oil and natural gas. Herein we report what are to our knowledge the first direct biocatalytic routes to BDO from renewable carbohydrate feedstocks, leading to a strain of Escherichia coli capable of producing 18 g l(-1) of this highly reduced, non-natural chemical. A pathway-identification algorithm elucidated multiple pathways for the biosynthesis of BDO from common metabolic intermediates. Guided by a genome-scale metabolic model, we engineered the E. coli host to enhance anaerobic operation of the oxidative tricarboxylic acid cycle, thereby generating reducing power to drive the BDO pathway. The organism produced BDO from glucose, xylose, sucrose and biomass-derived mixed sugar streams. This work demonstrates a systems-based metabolic engineering approach to strain design and development that can enable new bioprocesses for commodity chemicals that are not naturally produced by living cells.     

Optimal In Silico Target Gene Deletion through Nonlinear Programming for Genetic Engineering

Background

Optimal selection of multiple regulatory genes, known as targets, for deletion to enhance or suppress the activities of downstream genes or metabolites is an important problem in genetic engineering. Such problems become more feasible to address in silico due to the availability of more realistic dynamical system models of gene regulatory and metabolic networks. The goal of the computational problem is to search for a subset of genes to knock out so that the activity of a downstream gene or a metabolite is optimized.

Methodology/Principal Findings

Based on discrete dynamical system modeling of gene regulatory networks, an integer programming problem is formulated for the optimal in silico target gene deletion problem. In the first result, the integer programming problem is proved to be NP-hard and equivalent to a nonlinear programming problem. In the second result, a heuristic algorithm, called GKONP, is designed to approximate the optimal solution, involving an approach to prune insignificant terms in the objective function, and the parallel differential evolution algorithm. In the third result, the effectiveness of the GKONP algorithm is demonstrated by applying it to a discrete dynamical system model of the yeast pheromone pathways. The empirical accuracy and time efficiency are assessed in comparison to an optimal, but exhaustive search strategy.

Significance

Although the in silico target gene deletion problem has enormous potential applications in genetic engineering, one must overcome the computational challenge due to its NP-hardness. The presented solution, which has been demonstrated to approximate the optimal solution in a practical amount of time, is among the few that address the computational challenge. In the experiment on the yeast pheromone pathways, the identified best subset of genes for deletion showed advantage over genes that were selected empirically. Once validated in vivo, the optimal target genes are expected to achieve higher genetic engineering effectiveness than a trial-and-error procedure.     

Combinatorial network of transcriptional regulation and microRNA regulation in human cancer

Background

In order to reduce time and efforts to develop microbial strains with better capability of producing desired bioproducts, genome-scale metabolic simulations have proven useful in identifying gene knockout and amplification targets. Constraints-based flux analysis has successfully been employed for such simulation, but is limited in its ability to properly describe the complex nature of biological systems. Gene knockout simulations are relatively straightforward to implement, simply by constraining the flux values of the target reaction to zero, but the identification of reliable gene amplification targets is rather difficult. Here, we report a new algorithm which incorporates physiological data into a model to improve the model??s prediction capabilities and to capitalize on the relationships between genes and metabolic fluxes.

Results

We developed an algorithm, flux variability scanning based on enforced objective flux (FVSEOF) with grouping reaction (GR) constraints, in an effort to identify gene amplification targets by considering reactions that co-carry flux values based on physiological omics data via ??GR constraints??. This method scans changes in the variabilities of metabolic fluxes in response to an artificially enforced objective flux of product formation. The gene amplification targets predicted using this method were validated by comparing the predicted effects with the previous experimental results obtained for the production of shikimic acid and putrescine in Escherichia coli. Moreover, new gene amplification targets for further enhancing putrescine production were validated through experiments involving the overexpression of each identified targeted gene under condition-controlled batch cultivation.

Conclusions

FVSEOF with GR constraints allows identification of gene amplification targets for metabolic engineering of microbial strains in order to enhance the production of desired bioproducts. The algorithm was validated through the experiments on the enhanced production of putrescine in E. coli, in addition to the comparison with the previously reported experimental data. The FVSEOF strategy with GR constraints will be generally useful for developing industrially important microbial strains having enhanced capabilities of producing chemicals of interest.     

G6P-capturing molecules in the periplasm of Escherichia coli accelerate the shikimate pathway

Escherichia coli, the most studied prokaryote, is an excellent host for producing valuable chemicals from renewable resources as it is easy to manipulate genetically. Since the periplasmic environment can be easily controlled externally, elucidating how the localization of specific proteins or small molecules in the periplasm affects metabolism may lead to bioproduction development using E. coli. We investigated metabolic changes and its mechanisms occurring when specific proteins are localized to the E. coli periplasm. We found that the periplasmic localization of β-glucosidase promoted the shikimate pathway involved in the synthesis of aromatic chemicals. The periplasmic localization of other proteins with an affinity for glucose-6-phosphate (G6P), such as inactivated mutants of Pgi, Zwf, and PhoA, similarly accelerated the shikimate pathway. Our results indicate that G6P is transported from the cytoplasm to the periplasm by the glucose transporter protein EIICB^Glc, and then captured by β-glucosidase.     

Improvement of ClosTron for successive gene disruption in Clostridium cellulolyticum using a pyrF-based screening system

Clostridium includes a number of species, such as thermophilic Clostridium thermocellum and mesophilic Clostridium cellulolyticum, producing biofuels and chemicals from lignocellulose, while genetic engineering is necessary to improve wild-type strains to fulfill the requirement of industrialization. ClosTron system is widely used in the gene targeting of Clostridium because of its high efficiency and operability. However, the targetron plasmid present in cell hinders the successive gene disruption. To solve this problem, a pyrF-based screening system was developed and implemented in C. cellulolyticum strain H10 in this study for efficient targetron plasmid curing. The screening system was composed of a pyrF-deleted cell chassis (H10ΔpyrF) constructed via homologous recombination and a PyrF expression cassette located in a targetron plasmid containing an erythromycin resistance gene. With the screening system, the gene targeting could be achieved following a two-step procedure, including the first step of gene disruption through targetron transformation and erythromycin selection and the second step of plasmid curing by screening with 5-fluoroorotic acid. To test the developed screening system, successive inactivation of the major cellulosomal exocellulase Cel48F and the scaffoldin protein CipC was achieved in C. cellulolyticum, and the efficient plasmid curing was confirmed. With the assistance of the pyrF-based screening system, the targetron plasmid-cured colonies can be rapidly selected by one-plate screening instead of traditional days' unguaranteed screening, and the successive gene disruption becomes accomplishable with ClosTron system with improved stability and efficiency, which may promote the metabolic engineering of Clostridium species aiming at enhanced production of biofuels and chemicals.     

Developing Bacillus spp. as a cell factory for production of microbial enzymes and industrially important biochemicals in the context of systems and synthetic biology

Increasing concerns over limited petroleum resources and associated environmental problems are motivating the development of efficient cell factories to produce chemicals, fuels, and materials from renewable resources in an environmentally sustainable economical manner. Bacillus spp., the best characterized Gram-positive bacteria, possesses unique advantages as a host for producing microbial enzymes and industrially important biochemicals. With appropriate modifications to heterologous protein expression and metabolic engineering, Bacillus species are favorable industrial candidates for efficiently converting renewable resources to microbial enzymes, fine chemicals, bulk chemicals, and fuels. Here, we summarize the recent advances in developing Bacillus spp. as a cell factory. We review the available genetic tools, engineering strategies, genome sequence, genome-scale structure models, proteome, and secretion pathways, and we list successful examples of enzymes and industrially important biochemicals produced by Bacillus spp. Furthermore, we highlight the limitations and challenges in developing Bacillus spp. as a robust and efficient production host, and we discuss in the context of systems and synthetic biology the emerging opportunities and future research prospects in developing Bacillus spp. as a microbial cell factory.     

Engineered biosynthesis of medium-chain esters in Escherichia coli

Medium-chain esters such as isobutyl acetate (IBAc) and isoamyl acetate (IAAc) are high-volume solvents, flavors and fragrances. In this work, we engineered synthetic metabolic pathways in Escherichia coli for the total biosynthesis of IBAc and IAAc directly from glucose. Our pathways harnessed the power of natural amino acid biosynthesis. In particular, the native valine and leucine pathways in E. coli were utilized to supply the precursors. Then alcohol acyltransferases from various organisms were investigated on their capability to catalyze esterification reactions. It was discovered that ATF1 from Saccharomyces cerevisiae was the best enzyme for the formation of both IBAc and IAAc in E. coli. In vitro biochemical characterization of ATF1 confirmed the fermentation results and provided rational guidance for future enzyme engineering. We also performed strain improvement by removing byproduct pathways (Δldh, ΔpoxB, Δpta) and increased the production of both target chemicals. Then the best IBAc producing strain was used for scale-up fermentation in a 1.3-L benchtop bioreactor. 36 g/L of IBAc was produced after 72 h fermentation. This work demonstrates the feasibility of total biosynthesis of medium-chain esters as renewable chemicals.     

Systematic engineering of TCA cycle for optimal production of a four-carbon platform chemical 4-hydroxybutyric acid in Escherichia coli

To address climate change and environmental problems, it is becoming increasingly important to establish biorefineries for the production of chemicals from renewable non-food biomass. Here we report the development of Escherichia coli strains capable of overproducing a four-carbon platform chemical 4-hybroxybutyric acid (4-HB). Because 4-HB production is significantly affected by aeration level, genome-scale metabolic model-based engineering strategies were designed under aerobic and microaerobic conditions with emphasis on oxidative/reductive TCA branches and glyoxylate shunt. Several different metabolic engineering strategies were employed to develop strains suitable for fermentation both under aerobic and microaerobic conditions. It was found that microaerobic condition was more efficient than aerobic condition in achieving higher titer and productivity of 4-HB. The final engineered strain produced 103.4 g/L of 4-HB by microaerobic fed-batch fermentation using glycerol. The aeration-dependent optimization strategy of TCA cycle will be useful for developing microbial strains producing other reduced derivative chemicals of TCA cycle intermediates.     

Mutation tests in Neurospora crassa: A report of the U.S. environmental protection agency gene-tox program

Many mutation tests have been developed in Neurospora crassa during the almost 40 years of its use in mutation research. These tests detect two major classes of mutation: gene mutation and meiotic nondisjunction. Within the first class, forward- and reverse-mutation tests have been used. The forward-mutation tests include those that detect mutations at many loci and at specific loci. Both kinds of forward-mutation tests have been done in homokaryons (n) and heterokaryons (n + n′). From the publications that were not rejected by our pre-established criteria, data were extracted for 166 chemicals that had been tested for mutagenicity. Only 6 of the 166 chemicals have been tested in one or more gene mutation test and the meiotic nondisjunction test; these 6 chemicals were positive in the first and negative in the second. Of the 102 chemicals tested in one or more gene mutation tests, 94 were positive and 8 were negative. Of the 70 chemicals tested in the meiotic nondisjunction test, 7 were positive and 63 were negative.Two tests, the ad-3) forward-mutation test and the meiotic nondisjunction test, have been used most frequently. These two tests are especially important for hazard evaluation, because each detects a class of mutations that is likely to be deleterious or lethal in the F₁ - disomics by the meiotic nondisjunction test and multilocus deletions by the ad-3 forward-mutation test in heterokaryons. Generally, direct-acting chemicals are mutagenic in the gene mutation tests, but few chemicals that required metabolic activation have been tested. Only 31 of the 166 chemicals tested in N. crassa have been tested for carcinogenicity. Among these chemicals, there is a good association between mutagenicity in gene mutation tests and carcinogenicity but a poorer association between meiotic nondisjunction and carcinogenicity; however, only a small number of chemicals has been tested in the meiotic nondisjunction test. Further use and development of certain mutation tests in N. crassa are desirable.     

Kinetoplastid kinetochore proteins KKT2 and KKT3 have unique centromere localization domains

The kinetochore is the macromolecular protein complex that assembles onto centromeric DNA and binds spindle microtubules. Evolutionarily divergent kinetoplastids have an unconventional set of kinetochore proteins. It remains unknown how kinetochores assemble at centromeres in these organisms. Here, we characterize KKT2 and KKT3 in the kinetoplastid parasite Trypanosoma brucei. In addition to the N-terminal kinase domain and C-terminal divergent polo boxes, these proteins have a central domain of unknown function. We show that KKT2 and KKT3 are important for the localization of several kinetochore proteins and that their central domains are sufficient for centromere localization. Crystal structures of the KKT2 central domain from two divergent kinetoplastids reveal a unique zinc-binding domain (termed the CL domain for centromere localization), which promotes its kinetochore localization in T. brucei. Mutations in the equivalent domain in KKT3 abolish its kinetochore localization and function. Our work shows that the unique central domains play a critical role in mediating the centromere localization of KKT2 and KKT3.