Identification of metabolic engineering targets for the enhancement of 1,4-butanediol production in recombinant E. coli using large-scale kinetic models.

Rational metabolic engineering methods are increasingly employed in designing the commercially viable processes for the production of chemicals relevant to pharmaceutical, biotechnology, and food and beverage industries. With the growing availability of omics data and of methodologies capable to integrate the available data into models, mathematical modeling and computational analysis are becoming important in designing recombinant cellular organisms and optimizing cell performance with respect to desired criteria. In this contribution, we used the computational framework ORACLE (Optimization and Risk Analysis of Complex Living Entities) to analyze the physiology of recombinant Escherichia coli producing 1,4-butanediol (BDO) and to identify potential strategies for improved production of BDO. The framework allowed us to integrate data across multiple levels and to construct a population of large-scale kinetic models despite the lack of available information about kinetic properties of every enzyme in the metabolic pathways. We analyzed these models and we found that the enzymes that primarily control the fluxes leading to BDO production are part of central glycolysis, the lower branch of tricarboxylic acid (TCA) cycle and the novel BDO production route. Interestingly, among the enzymes between the glucose uptake and the BDO pathway, the enzymes belonging to the lower branch of TCA cycle have been identified as the most important for improving BDO production and yield. We also quantified the effects of changes of the target enzymes on other intracellular states like energy charge, cofactor levels, redox state, cellular growth, and byproduct formation. Independent earlier experiments on this strain confirmed that the computationally obtained conclusions are consistent with the experimentally tested designs, and the findings of the present studies can provide guidance for future work on strain improvement. Overall, these studies demonstrate the potential and effectiveness of ORACLE for the accelerated design of microbial cell factories.

An integrated biotechnology platform for developing sustainable chemical processes.

Genomatica has established an integrated computational/experimental metabolic engineering platform to design, create, and optimize novel high performance organisms and bioprocesses. Here we present our platform and its use to develop E. coli strains for production of the industrial chemical 1,4-butanediol (BDO) from sugars. A series of examples are given to demonstrate how a rational approach to strain engineering, including carefully designed diagnostic experiments, provided critical insights about pathway bottlenecks, byproducts, expression balancing, and commercial robustness, leading to a superior BDO production strain and process.

Dynamic analysis of CO2 labeling and cell respiration using membrane-inlet mass spectrometry.

Here, we introduce a mass spectrometry-based analytical method and relevant technical details for dynamic cell respiration and CO2 labeling analysis. Such measurements can be utilized as additional information and constraints for model-based (13)C metabolic flux analysis. Dissolved dynamics of oxygen consumption and CO2 mass isotopomer evolution from (13)C-labeled tracer substrates through different cellular processes can be precisely measured on-line using a miniaturized reactor system equipped with a membrane-inlet mass spectrometer. The corresponding specific rates of physiologically relevant gases and CO2 mass isotopomers can be quantified within a short-term range based on the liquid-phase dynamics of dissolved fermentation gases.

Overcoming the metabolic burden of protein secretion in Schizosaccharomyces pombe--a quantitative approach using 13C-based metabolic flux analysis.

Protein secretion in yeast is generally associated with a burden to cellular metabolism. To investigate this metabolic burden in Schizosaccharomyces pombe, we constructed a set of strains secreting the model protein maltase in different amounts. We quantified the influence of protein secretion on the metabolism applying (13)C-based metabolic flux analysis in chemostat cultures. Analysis of the macromolecular biomass composition revealed an increase in cellular lipid content at elevated levels of protein secretion and we observed altered metabolic fluxes in the pentose phosphate pathway, the TCA cycle, and around the pyruvate node including mitochondrial NADPH supply. Supplementing acetate to glucose or glycerol minimal media was found to improve protein secretion, accompanied by an increased cellular lipid content and carbon flux through the TCA cycle as well as increased mitochondrial NADPH production. Thus, systematic metabolic analyses can assist in identifying factors limiting protein secretion and in deriving strategies to overcome these limitations.

13C-based metabolic flux analysis: fundamentals and practice.

Isotope-based metabolic flux analysis is one of the emerging technologies applied to system level metabolic phenotype characterization in metabolic engineering. Among the developed approaches, (13)C-based metabolic flux analysis has been established as a standard tool and has been widely applied to quantitative pathway characterization of diverse biological systems. To implement (13)C-based metabolic flux analysis in practice, comprehending the underlying mathematical and computational modeling fundamentals is of importance along with carefully conducted experiments and analytical measurements. Such knowledge is also crucial when designing (13)C-labeling experiments and properly acquiring key data sets essential for in vivo flux analysis implementation. In this regard, the modeling fundamentals of (13)C-labeling systems and analytical data processing are the main topics we will deal with in this chapter. Along with this, the relevant numerical optimization techniques are addressed to help implementation of the entire computational procedures aiming at (13)C-based metabolic flux analysis in vivo.

Metabolic flux analysis gives an insight on verapamil induced changes in central metabolism of HL-1 cells.

Verapamil has been shown to inhibit glucose transport in several cell types. However, the consequences of this inhibition on central metabolism are not well known. In this study we focused on verapamil induced changes in metabolic fluxes in a murine atrial cell line (HL-1 cells). These cells were adapted to serum free conditions and incubated with 4 µM verapamil and [U-¹³C5] glutamine. Specific extracellular metabolite uptake/production rates together with mass isotopomer fractions in alanine and glutamate were implemented into a metabolic network model to calculate metabolic flux distributions in the central metabolism. Verapamil decreased specific glucose consumption rate and glycolytic activity by 60%. Although the HL-1 cells show Warburg effect with high lactate production, verapamil treated cells completely stopped lactate production after 24 h while maintaining growth comparable to the untreated cells. Calculated fluxes in TCA cycle reactions as well as NADH/FADH2 production rates were similar in both treated and untreated cells. This was confirmed by measurement of cell respiration. Reduction of lactate production seems to be the consequence of decreased glucose uptake due to verapamil. In case of tumors, this may have two fold effects; firstly depriving cancer cells of substrate for anaerobic glycolysis on which their growth is dependent; secondly changing pH of the tumor environment, as lactate secretion keeps the pH acidic and facilitates tumor growth. The results shown in this study may partly explain recent observations in which verapamil has been proposed to be a potential anticancer agent. Moreover, in biotechnological production using cell lines, verapamil may be used to reduce glucose uptake and lactate secretion thereby increasing protein production without introduction of genetic modifications and application of more complicated fed-batch processes.

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.

Metabolic response of Geobacter sulfurreducens towards electron donor/acceptor variation.

BACKGROUND: Geobacter sulfurreducens is capable of coupling the complete oxidation of organic compounds to iron reduction. The metabolic response of G. sulfurreducens towards variations in electron donors (acetate, hydrogen) and acceptors (Fe(III), fumarate) was investigated via (13)C-based metabolic flux analysis. We examined the (13)C-labeling patterns of proteinogenic amino acids obtained from G. sulfurreducens cultured with (13)C-acetate. RESULTS: Using (13)C-based metabolic flux analysis, we observed that donor and acceptor variations gave rise to differences in gluconeogenetic initiation, tricarboxylic acid cycle activity, and amino acid biosynthesis pathways. Culturing G. sulfurreducens cells with Fe(III) as the electron acceptor and acetate as the electron donor resulted in pyruvate as the primary carbon source for gluconeogenesis. When fumarate was provided as the electron acceptor and acetate as the electron donor, the flux analysis suggested that fumarate served as both an electron acceptor and, in conjunction with acetate, a carbon source. Growth on fumarate and acetate resulted in the initiation of gluconeogenesis by phosphoenolpyruvate carboxykinase and a slightly elevated flux through the oxidative tricarboxylic acid cycle as compared to growth with Fe(III) as the electron acceptor. In addition, the direction of net flux between acetyl-CoA and pyruvate was reversed during growth on fumarate relative to Fe(III), while growth in the presence of Fe(III) and acetate which provided hydrogen as an electron donor, resulted in decreased flux through the tricarboxylic acid cycle. CONCLUSIONS: We gained detailed insight into the metabolism of G. sulfurreducens cells under various electron donor/acceptor conditions using (13)C-based metabolic flux analysis. Our results can be used for the development of G. sulfurreducens as a chassis for a variety of applications including bioremediation and renewable biofuel production.

13C metabolic flux analysis for larger scale cultivation using gas chromatography-combustion-isotope ratio mass spectrometry.

(13)C-based metabolic flux analysis ((13)CMFA) is limited to smaller scale experiments due to very high costs of labeled substrates. We measured (13)C enrichment in proteinogenic amino acid hydrolyzates using gas chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS) from a series of parallel batch cultivations of Corynebacterium glutamicum utilizing mixtures of natural glucose and [1-(13)C] glucose, containing 0%, 0.5%, 1%, 2%, and 10% [1-(13)C] glucose. Decreasing the [1-(13)C] glucose content, kinetic isotope effects played an increasing role but could be corrected. From the corrected (13)C enrichments in vivo fluxes in the central metabolism were determined by numerical optimization. The obtained flux distribution was very similar to those obtained from parallel labeling experiments using conventional high labeling GC-MS method and to published results. The GC-C-IRMS-based method involving low labeling degree of expensive tracer substrate, e.g. 1%, is well suited for larger laboratory and industrial pilot scale fermentations.

Numerical bias estimation for mass spectrometric mass isotopomer analysis.

Mass spectrometric (MS) isotopomer analysis has become a standard tool for investigating biological systems using stable isotopes. In particular, metabolic flux analysis uses mass isotopomers of metabolic products typically formed from (13)C-labeled substrates to quantitate intracellular pathway fluxes. In the current work, we describe a model-driven method of numerical bias estimation regarding MS isotopomer analysis. Correct bias estimation is crucial for measuring statistical qualities of measurements and obtaining reliable fluxes. The model we developed for bias estimation corrects a priori unknown systematic errors unique for each individual mass isotopomer peak. For validation, we carried out both computational simulations and experimental measurements. From stochastic simulations, it was observed that carbon mass isotopomer distributions and measurement noise can be determined much more precisely only if signals are corrected for possible systematic errors. By removing the estimated background signals, the residuals resulting from experimental measurement and model expectation became consistent with normality, experimental variability was reduced, and data consistency was improved. The method is useful for obtaining systematic error-free data from (13)C tracer experiments and can also be extended to other stable isotopes. As a result, the reliability of metabolic fluxes that are typically computed from mass isotopomer measurements is increased.

Towards a metabolic and isotopic steady state in CHO batch cultures for reliable isotope-based metabolic profiling.

Attaining metabolic and isotopic balanced growth is one critical condition for physiological studies using isotope-labeled tracers, but is very difficult to obtain in batch culture due to the extensive metabolite exchange with the surrounding medium and related physiological changes. In the present study, we investigated metabolic and isotopic behavior of CHO cells in differently designed media. We observed that the assumption of balanced cell growth cannot be justified in batch culture of CHO cells directly using conventional, commercially available media. By systematically redesigning media composition and characterizing metabolic steady state based on mass balances and measurement of labeling dynamics, we achieved balanced cell growth for the main cellular substrates in CHO cells. This was done in a step-by-step analysis of growth and primary metabolism of CHO cells with the use of [U-13C]glucose feeding and adjusting concentrations of amino acids in the growth medium. The optimized media obtained at the end of the study provide balanced growth and isotopic steady state or at least asymptotic steady state. As a result, we established a platform to conduct isotope-based physiological studies of mammalian systems more reliably and therefore well suited for later use in metabolic profiling of mammalian systems such as 13C-labeled metabolic flux analysis.

Hybrid optimization for 13C metabolic flux analysis using systems parametrized by compactification.

BACKGROUND: The importance and power of isotope-based metabolic flux analysis and its contribution to understanding the metabolic network is increasingly recognized. Its application is, however, still limited partly due to computational inefficiency. 13C metabolic flux analysis aims to compute in vivo metabolic fluxes in terms of metabolite balancing extended by carbon isotopomer balances and involves a nonlinear least-squares problem. To solve the problem more efficiently, improved numerical optimization techniques are necessary. RESULTS: For flux computation, we developed a gradient-based hybrid optimization algorithm. Here, independent flux variables were compactified into [0, 1)-ranged variables using a single transformation rule. The compactified parameters could be discriminated between non-identifiable and identifiable variables after model linearization. The developed hybrid algorithm was applied to the central metabolism of Bacillus subtilis with only succinate and glutamate as carbon sources. This creates difficulties caused by symmetry of succinate leading to limited introduction of 13C labeling information into the system. The algorithm was found to be superior to its parent algorithms and to global optimization methods both in accuracy and speed. The hybrid optimization with tolerance adjustment quickly converged to the minimum with close to zero deviation and exactly re-estimated flux variables. In the metabolic network studied, some fluxes were found to be either non-identifiable or nonlinearly correlated. The non-identifiable fluxes could correctly be predicted a priori using the model identification method applied, whereas the nonlinear flux correlation was revealed only by identification runs using different starting values a posteriori. CONCLUSION: This fast, robust and accurate optimization method is useful for high-throughput metabolic flux analysis, a posteriori identification of possible parameter correlations, and also for Monte Carlo simulations to obtain statistical qualities for flux estimates. In this way, it contributes to future quantitative studies of central metabolic networks in the framework of systems biology.

Equilibria and dynamics of liquid-phase trinitrotoluene adsorption on granular activated carbon: effect of temperature and pH.

Environmental regulations for removal of trinitrotoluene (TNT) from wastewater have steadily become more stringent. This study focuses on the adsorption equilibrium, kinetics, and column dynamics of TNT on heterogeneous activated carbon. Adsorption equilibrium data obtained in terms of temperature (298.15, 313.15 and 323.15K) and pH (3, 8 and 10) were correlated by the Langmuir equation. In addition, the adsorption energy distribution functions which describe heterogeneous characteristics of porous solid sorbents were calculated by using the generalized nonlinear regularization method. Adsorption breakthrough curves were studied in activated column under various operating conditions such as temperature, pH, concentration, flow rate, and column length. We found that the effect of pH on adsorption breakthrough curves was considerably higher than other operating conditions. An adsorption model was formulated by employing the surface diffusion model inside the activated carbon particles. The model equation that was solved numerically by an orthogonal collocation method successfully simulated the adsorption breakthrough curves.

Respirometric 13C flux analysis, Part I: design, construction and validation of a novel multiple reactor system using on-line membrane inlet mass spectrometry.

A novel method for (13)C flux analysis based on on-line CO(2) labeling measurements is presented. This so-called respirometric (13)C flux analysis requires multiple parallel (13)C labeling experiments using differently labeled tracer substrates. In Part I of the work, a membrane-inlet mass spectrometry-based measurement system with 6 parallel reactors with each 12 ml liquid volume and associated experimental and computational methods for the respirometric (13)C data acquisition and evaluation are described. Signal dynamics after switching between membrane probes follow exactly first-order allowing extrapolation to steady state. Each measurement cycle involving 3 reactors takes about 2 min. After development of a dynamic calibration method, the suitability and reliability of the analysis was examined with a lysine-producing mutant of Corynebacterium glutamicum using [1-(13)C(1)], [6-(13)C(1)], [1,6-(13)C(2)] glucose. Specific rates of oxygen uptake and CO(2) production were estimated with an error less than +/-0.3 mmol g(-1) h(-1) and had +/-3% to +/-10% deviations between parallel reactors which is primarily caused by inaccuracies in initial biomass concentration. The respiratory quotient could be determined with an uncertainty less than +/-0.02 and varied only +/-3% between reactors. Fractional labeling of CO(2) was estimated with much higher precision of about +/-0.001 to +/-0.005. The detailed statistical analysis suggested that these data should be of sufficient quality to allow physiological interpretation and metabolic flux estimation. The obtained data were applied for the respirometric (13)C metabolic flux analysis in Part II.

Theoretical aspects of 13C metabolic flux analysis with sole quantification of carbon dioxide labeling.

The potential of using sole respirometric CO2 labeling measurement for 13C metabolic flux analysis was investigated by metabolic simulations. For this purpose a model was created, considering all CO2 forming and consuming reactions in the central catabolic and anabolic pathways. To facilitate the interpretation of the simulation results, the underlying metabolic network was parameterized by physiologically meaningful flux parameters such as flux partitioning ratios at metabolic branch points and reaction reversibilities. For real case flux scenarios of the industrial amino acid producer Corynebacterium glutamicum and different commercially available (13)C-labeled tracer substrates, observability and output sensitivity towards key flux parameters was investigated. Metabolic net fluxes in the central metabolism, involving, e.g. glycolysis, pentose phosphate pathway, tricarboxylic acid cycle, anaplerotic carboxylation, and glyoxylate pathway were found to be determinable by the respirometric approach using a combination of [1-13C] and [6-13C] glucose in two parallel studies. The reversibilities of bidirectional reactions influence the isotopic labeling of CO2 only to a negligible degree. On one hand, they therefore cannot be determined. On the other hand, their precise values are not required for the quantification of net fluxes. Computer-aided optimal experimental design was carried out to predict the quality of the information from the respirometric tracer experiments and identify suitable tracer substrates. A combination of [1-13C] and [6-13C] glucose in two parallel studies was found to yield a similar quality of information as compared to an approach with mass spectrometric labeling analysis of secreted products. The quality of information can be further increased by additional studies with [1,2-13C2] or [1,6-13C2] glucose. Respirometric tracer studies with sole labeling analysis of CO2 are therefore promising for 13C metabolic flux analysis.

Metabolic network simulation using logical loop algorithm and Jacobian matrix.

A novel method to accomplish efficient numerical simulation of metabolic networks for flux analysis was developed. The only inputs required are the set of stoichiometric balances and the atom mapping matrices of all components of the reaction network. The latter are used to automatically calculate isotopomer mapping matrices. Using the symbolic toolbox of MATLAB the analytical solution of the stoichiometric balance equation system, isotopomer balances and the analytical Jacobian matrix of the total set of stoichiometric and isotopomer balances are created automatically. The number of variables in the isotopomer distribution equation system is significantly reduced applying modified isotopomer mapping matrices. These allow lumping of several consecutive isotopomer reactions into a single one. The solution of the complete system of equations is improved by implementing an iterative logical loop algorithm and using the analytical Jacobian matrix. This new method provided quick and robust convergence to the root of such equation systems in all cases tested. The method was applied to a network of lysine producing Corynebacterium glutamicum. The resulting equation system with the dimension of 546 x 546 was directly derived from 12 isotopomer balance equations. The results obtained yielded identical labeling patterns for metabolites as compared to the relaxation method.

Dynamic calibration and dissolved gas analysis using membrane inlet mass spectrometry for the quantification of cell respiration.

A membrane inlet mass spectrometer connected to a miniaturized reactor was applied for dynamic dissolved gas analysis. Cell samples were taken from 7 mL shake flask cultures of Corynebacterium glutamicum ATCC 13032, and transferred to the 12 mL miniaturized reactor. There, oxygen uptake and carbon dioxide and its mass isotopomer production rates were determined using a new experimental procedure and applying nonlinear model equations. A novel dynamic method for the calibration of the membrane inlet mass spectrometer using first-order dynamics was developed. To derive total dissolved concentration of all carbon dioxide species (C(T)) from dissolved carbon dioxide concentration ([CO(2)](aq)), the ratio of C(T) to [CO(2)](aq) was determined by nonlinear parameter estimation, whereas the mass transfer coefficient of CO(2) was determined by the Wilke-Chang correlation. Subsequently, the suitability of the model equations for respiration measurements was examined using residual analysis and the Jarque-Bera hypothesis test. The resulting residuals were found to be random with normal distribution, which proved the adequacy of the application of the model for cell respiration analysis. Hence, dynamic changes in respiration activities could be accurately analyzed using membrane inlet mass spectrometry with the novel calibration method.