Systems biology and metabolic engineering for obtaining E ...rombio.eu/vol19nr4/lucr 23 Zs. Bodor 2014_rec 6 febr...Keywords: systems biology, biotechnology, succinate, Escherichia

Romanian Biotechnological Letters Vol. 19, No. 4, 2014 Copyright © 2014 University of Bucharest Printed in Romania. All rights reserved

ORIGINAL PAPER

Romanian Biotechnological Letters, Vol. 19, No. 4, 2014 9625

Systems biology and metabolic engineering for obtaining E. coli mutants capable to produce succinate from renewable resources

Received for publication, December 20, 2013

Accepted, February 14, 2014

ZSOLT BODOR1*, ANDREA (IUHASZ) FAZAKAS1, ERIKA KOVÁCS 2, SZABOLCS LÁNYI2, BEÁTA ALBERT2 1 Department of Inorganic Substances Technology and Environment Protection, Polizu street No 1-7, 011061, “Politehnica” University of Bucharest, Bucharest, Romania 2Department of Bioengineering, Libertatii square, No. 1, 530104, Sapientia Hungarian University of Transylvania, Miercurea Ciuc, Romania *Address correspondence to: “Politehnica” University of Bucharest, Faculty of Applied Chemistry and Material Science, Analytical Chemistry and Environmental Engineering, Polizu street No 1-7, 011061, Bucharest, Romania, Tel: +40740791848; Fax: +40 266 372 099, Email:[email protected]

Abstract

The bio-based conversion of glycerol and glucose to valuable chemicals, such as succinate, by commonly used microorganisms, like Escherichia coli is of major interest. The petrochemically based succinate production can be replaced by a new sustainable and environmental friendly one and the glycerol resulted from biodiesel industry can be therefore reused. Genetically engineered strains could be used to provide a cost-effective, ecologically sustainable alternative to the current petrochemical production process. Systems biology and in silico analyses are necessary to study complex biological systems and successfully apply metabolic engineering. Here, a systems biology approach through model-driven evaluation of the cell metabolism and the production potential of an important biochemical compound, under different environmental conditions is presented. To investigate the genetic and environmental perturbations, as well as the relationship between biomass and succinate yield, in silico metabolic analysis was carried out using constraint-based metabolic flux simulations. Different methods were used to design strains with increased capabilities to produce succinate. The study provided specific metabolic interventions that can be experimentally implemented, characterized the metabolic network and outlined a strain design pipeline that can be used to study complex biological systems and processes.

Keywords: systems biology, biotechnology, succinate, Escherichia coli, metabolic engineering, modelling, glycerol.

1. Introduction The production of biodiesel generates large quantities of glycerol which is an inexpensive carbon source, like glucose, for many microorganisms (1-4). Glycerol can be converted biologically to succinate, which is an important precursor for many industrially manufactured chemical commodities and products (5-7). The level of succinate produced by native strains of E. coli in minimal medium is very low (2). To improve cellular capabilities and the production yield metabolic engineering should be carried out (8).

Metabolic engineering is successful in generating microbial strains with increased capabilities to produce different industrially important compounds (9, 10). The identification of genetic manipulations that led to mutant strains, able to produce a target compound is a promising and at the same time, a complex process. It is not an easy task to design and analyse a mutant strain, because several parameters should be considered simultaneously. Hence, the computational approach, especially systems biology is crucial to study complex

ZSOLT BODOR, ANDREA (IUHASZ) FAZAKAS, ERIKA KOVÁCS, SZABOLCS LÁNYI, BEÁTA ALBERT

9626 Romanian Biotechnological Letters, Vol. 19, No. 4, 2014

biological processes, like metabolic engineering, interactions inside the cell, relationships between genotype and phenotype, metabolic fluxes etc.

To analyse in silico large-scale biological networks, make predictions about cellular behaviours and to test the effect or perturbations, such as gene deletions, Flux Balance Analysis is an effective tool (11, 12). Systems biology is successful in predicting the outcomes or the growth rate utilizing the well known constraint-based reconstruction and analysis (COBRA) (13). The complexity of the intra-cellular metabolic network makes the identification of gene manipulations difficult, from here the role of in silico platforms being considered essential for the rational design of the network to improve the production rate of a target component (14). The biological models are constructed based on the known stoichiometry of the metabolic reactions, thermodynamic constraints and flux capacities (15, 16). One of the modelling platforms is COBRA Toolbox (17, 18) an open-source and modular platform (widely used in systems biology), incorporating strain optimization tasks, algorithms such as: FBA, dynamic FBA (dFBA), phenotypic phase plane analysis (PhPP), etc. The present large-scale computational study of cellular metabolism aimed to identify the biological way to produce succinate from renewable resources and how was the metabolic network influenced by different factors. Well-established metabolic modelling methods were used for metabolic engineering studies. The identified genes were either: a) selected by us or b) utilizing special optimizing algorithms such as OptKnock (19) and GDLS (20) under aerobic, microaerobic and anaerobic conditions, using glucose or glycerol as carbon sources. The methods outlined in this paper are fundamental approaches of applying systems biology to metabolic engineering for sustainable biotechnology and provides testable hypotheses. Strain optimization experiments were carried out, using λ-Red recombineering methods (21) and for the comparative analysis of metabolite profiles, we used gas chromatography coupled to mass spectrometry (GC-MS). Usually the results agreed well with the experimentally obtained ones (22, 23). This in silico approach of genetic engineering studies can reduce the time and cost of wet experiments. 2. Materials and Methods The metabolic reconstruction of E. coli K12 - named iJO1366 - (24) in SBML format was utilized in this study. The model is the most complex functionally available, tested and verified against experimental data to predict correctly the growth rates, metabolites excretion rates and growth phenotypes under different substrates and genetic conditions (25). The metabolic model is available in SBML (Systems Biology Markup Language) format at BioModels online database [http://www.ebi.ac.uk/biomodels-main/]. 2.1. Flux balance analysis (FBA) for wild-type and mutant strains: Steady-state metabolite flux assumption was performed for FBA calculations described in detail previously (13). OptKnock and GDLS bi-level optimization algorithms were implemented in the COBRA Toolbox as described in their original documentation. Simulations were run to completion for two, three and four maximum knockout simulations under different environmental conditions. Consumption rate for the main carbon substrate in each simulation was set to 10 mmol gDW-1h-1(millimoles per gram dry cell weight per hour). Simulations were carried out using minimal media (M9) containing only inorganic salts and for carbon source we used glucose or glycerol. Under aerobic growth the oxygen uptake was set to 1000 mmol gDW-1h-1 (unlimited oxygen uptake), for microaerobic we used 5 mmol gDW-1h-1 and 0 to create anaerobic conditions. All computations were performed in MATLAB (mathworks



Inc.; Natick, MA, USA) using COBRA Toolbox (version 2.0.5, http://opencobra.sourceforge.net/openCOBRA/Welcome.html) software packages with Gurobi optimization solver (Gurobi Optimizer version 5.1.0 Houston, Texas). The eliminated reactions selected by us (considering results from the literature) were: ΔpflB (pyruvate formate lyase), ΔldhA (lactate dehydrogenase), ΔadhE (alcohol dehydrogenase) and Δpts (phosphotransferase system). The maximum production potential for E. coli was determined by defining a minimum growth rate (μ) of 0.1 h-1 to simulate a minimum growth and maximizing the production rate of succinate with OptKnock and GDLS. Combining FBA with an iterative approach based on a quasi-steady-state assumption we are able to analyse dynamic processes, e.g. growth rates, metabolites production and consumption rates of the wild-type and the mutant strains. The initial substrate concentration was set to 10 mmol L-1, while the initial biomass concentration was set to 0.035 g L-1 (~0.1 optical density (OD)). In order to observe if diauxic growth is present, time step was set at a higher value to 25 min and the maximum number of steps to 150 to allow the consumption of metabolites. Phenotypic phase plane analysis was performed for wild-type and mutant strains to identify the robustness of the system under different conditions, if two parameters are varied simultaneously. Detailed description of this method can be found elsewhere (11). 2.2. Experimental procedure: The strain used in this study was Escherichia coli K12 MG1655 from “Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH” (DSMZ 18039). Fermentation was done at 37°C, in minimal medium under anaerobic conditions with glucose or glycerol at a concentration of 0.2% (v/v). Bacterial cells were grown in 5 mL of minimal medium to produce a starter culture and the seed culture was used to inoculate the fermentation medium (OD=0.1). Cells were grown with shaking at 150 rpm (Certomat BS-1 Sartorius) for 24 h in serum bottles (50 mL) with 20 mL M9 medium. Samples were taken every two hours for the analysis of cell growth. The optical density of the cell cultures was measured at 550 nm (OD550) to quantify cell growth, using a Cary 50 Conc UV-Visible spectrophotometer, as well as for dry cell weight determination. To estimate the cell mass we used the following simple assumption (1 OD550=0.36 gDW L-1). 1 mL of the culture supernatant was added in triplicate to pre-weighed Eppendorf tubes, centrifuged, washed with NaCl (0.9%) and dried until constant mass at 105°C. 2.3. Chromosomal gene deletion: λ-Red recombineering methods previously described (21) were used to create mutant strains. Plasmids (5 Strain Wanner Lambda Red Gene Disruption Kit) were obtained from the E. coli Genetic Stock Center (Yale University). 2.4. Analytical procedure: GC-MS (6890N/5975 Agilent) was used based on solid-phase microextraction (SPME) with on-fiber silylation to analyse metabolites. Silylation was carried out using N, O bis (trimethylsilyl) trifluoroacetamide (BSTFA), following the procedures described elsewhere (26). For chromatograms and spectral analyses we used the MassLab software (ThermoQuest, Manchester, UK). Compounds were identified by comparing the mass spectra obtained with commercially available MS libraries (Wiley, NIST and LIBTX). 3. Results and Discussions The present study aimed to redesign an E. coli strain to create different mutants being capable to produce succinate from renewable resources, such as glycerol and glucose in minimal media, under aerobic, microaerobic and anaerobic conditions and to understand the cellular responses to different factors. First we calculated the flux distribution of E. coli. As shown in Table 1 there are significant differences in specific growth rates (µ), in cells grown



under different conditions. Values are close to that observed experimentally, giving us confirmation that the genome-scale model reliability is good.

Table 1. The growth rates (µ) of E. coli K12 strain on different carbon sources (M9)

Glucose Growth rate (µ) Carbon-source Condition

E. coli 0.98 glucose aerobic

0.48 glucose microaerobic 0.24 glucose anaerobic

Glycerol

E. coli 0.55 glycerol aerobic 0.33 glycerol microaerobic 0.08 glycerol anaerobic

To investigate the consequences of alternative pathway eliminations on succinate yield

we decided to start with pflB, because the formate concentration was the highest under anaerobic condition (a yield of 1.73 mol mol-1 glucose and 1.81 mol mol-1 glycerol). Deletion of pflB had little effect on cell growth or even on succinate concentration in M9 mineral salts medium using glucose as sole carbon source because the carbon flow was diverted to lactate. The ΔpflB mutant showed slower growth than the wild-type, formate was not present, and the formation of acetate and ethanol was reduced to zero because the pfl is the primary route for pyruvate conversion to Ac-CoA. Eliminating the PFL:pyruvate formate lyase enzyme the pyruvate dissimilation was blocked and the, LDHA:lactate dehydrogenase was activated allosterically resulting in increased lactate synthesis. The succinate yield in this case was perturbed by the alcohol dehydrogenase because in case of the double mutant the concentration of ethanol was very high. Our next step was to delete the responsible pathway for ethanol production, adhE:alcohol dehydrogenase to improve the yield of succinate. To increase the yield of succinate it was necessary to eliminate all of the three pathways. The succinate molar yield was 0.9 mol mol-1 glucose after eliminating the above mentioned pathways. Mutations in the ΔpflB, ΔldhA, ΔadhE genes drove the metabolic fluxes toward succinate with only minor fermentation metabolites in microaerobic and anaerobic conditions. The biomass formation decreased more than 54% but the yield of succinate increased by 113 fold compared to the wild-type. Similar results were obtained under microaerobic conditions in the mutant strains, with 83% growth rate relative to wild-type. The main difference between microaerobic and anaerobic conditions is: acetate production (higher during microaerobic) and succinate (higher during anaerobic conditions) (table 2). The yield of succinate increased to 0.9 mol mol-1 with adh deletion, but growth was reduced with 40% compared to the double mutant when both of the alternative NADH oxidation pathways were inactivated. A significant increase of succinate yield was obtained only with ΔpflB, ΔldhA, ΔadhE, Δpts quadratic deletion, a molar yield of 1.25 mol mol-1 of glucose. The phosphoenolpyruvate-dependent phosphotransferase system is the primary mechanism for glucose uptake in E. coli. If we eliminate this mechanism the glucose uptake is replaced by GALP:galactose permease and GLK:glucokinase (27). Inactivation of the pts system results in an increase in the PEP pool, allowing the generation of more succinate. 3.1. Comparison of aerobic, microaerobic and anaerobic growth results for the designed mutants by using bi-level programming frameworks: To design strains of E. coli to produce the target product, i.e. succinate we used the available data from the literature and



the mentioned algorithms to identify the genetic modifications. Each simulation was allowed to run to completion, so the entire solution space was scanned. All of the designs indentified for glucose as carbon and energy source are presented in Table 2. First, we used OptKnock to design strains for each substrate and environmental conditions for a maximum of four reaction knock-outs, which allowed to find the global optimal sets of knockouts. After that GDLS simulations were conducted to identify if there are better solutions.

Table 2. Theoretical maximal production rate of succinate from glucose under different environmental conditions and genetic modifications

Designed by us OptKnock GDLS

Conditions Targets WTa 2KOb 3KOc 4KOd 2KOe 3KOf 4KOg 2KOh 3KOi 4KOj

aerobic Biomass 0.98 0.98 0.98 0.97 0.10 0.10 0.10 0.90 0.80 0.60

Succinate - - - - 0.95 1.40 5.96 0.94 1.38 6.08

microaerobic Biomass 0.49 0.45 0.41 0.38 0.10 0.10 0.10 0.49 0.41 0.40

Succinate - - 6.43 6.69 0.55 6.10 6.50 0.51 6.43 6.79

anaerobic Biomass 0.24 0.19 0.11 0.09 0.10 0.10 0.10 0.10 0.10 0.10

Succinate 0.08 0.06 9.10 12.53 1.68 12.03 12.39 1.41 12.35 12.46 a- wild-type; b- ΔpflB, ΔldhA; c- ΔpflB, ΔldhA, ΔadhE, d- ΔpflB, ΔldhA, ΔadhE, Δpts ; e- aerobic ΔfumA-fumarase, ΔpntB-NAD(P) transhydrogenase (periplasm); microaerobic Δmdh-malate dehydrogenase, ΔmaeB-malic enzyme (NADP); anaerobic ΔpflB, ΔtktB- transketolase; f- aerobic ΔfumA, ΔscpC- propanoyl-CoA:succinate CoA-transferase, ΔsucC-succinyl-CoA synthetase (ADP-forming); microaerobic ΔmhpF-acetaldehyde dehydrogenase, ΔldhA, ΔpflB; anaerobic ΔmhpF, ΔldhA, ΔpflB; g- aerobic ΔackA- acetate kinase, ΔfumA, Δgnd- phosphogluconate dehydrogenase, ΔserA- phosphoglycerate dehydrogenase; microaerobic ΔadhE, ΔldhA, ΔpflB, ΔtktA; anaerobic ΔadhE, ΔldhA, ΔpflB, Δpgi-glucose-6-phosphate isomerase; h- aerobic ΔfumA, ΔpntB; microaerobic Δmdh, ΔmaeB; anaerobic ΔpflB, ΔtktB; i- aerobic ΔfumA, ΔscpC, ΔsucC; microaerobic ΔmhpF, ΔldhA, ΔpflB; anaerobic ΔmhpF, ΔldhA, ΔpflB; j- aerobic ΔfumA, Δgnd, ΔserA, Δpta-phosphotransacetylase; microaerobic ΔadhE, ΔldhA, ΔpflB, ΔtktB; anaerobic ΔmhpF, ΔldhA, ΔpflB, ΔtktB. (Fluxes have units of mmol gDW-1h-1, except for biomass, which has units of h-1). The NADH generated during glycolysis is reoxidized in the process when the organic intermediates are reduced; the reducing equivalents are fully consumed. The cell tries to redress the reducing equivalents and with every genetic mutation different metabolites are produced. To maintain the redox balance acetyl-CoA is converted to lactate and ethanol and ATP is produced from the acetate pathway. The production rate of succinate being strongly related to the rate of biomass generation and, thu8s the biomass generation rate decreased with the increase in succinate yield. Utilizing optimization algorithms similar results were predicted under anaerobic conditions, only one deleted gene was different (pgi for OptKnock and tktA for GDLS). Similar tendency was predicted for microaerobic conditions, however during aerobic conditions the list of genes predicted to be eliminated from the system to improve succinate production are quite different (table 2). It is clear that under aerobic conditions different genes or even pathways are active and hence, different genes should be eliminated. The highest succinate yield was found to be under anaerobic conditions in each case paired with more complex metabolic interventions (4 KO). 3.2. Glycerol as carbon and energy source: Our next goal was to analyse the production potential if the carbon source is changed to glycerol. Glycerol is an abundant and inexpensive carbon source; it is generated in huge quantities as a by-product during biofuel production, so the value-added utilization in biotechnology is a promising future.

The growth of wild-type E. coli K12 MG1655 was very slowly on glycerol even in aerobic conditions. Succinate production from glycerol involves fixation of CO2 onto a 3–



carbon intermediate, which can be converted to succinate. Formate, ethanol and acetate were the major products, with smaller amounts of succinate during fermentation. To knock out formate production the PFL reaction must be deleted as we mentioned before in case of glucose, but after our simulations the ΔpflB mutant failed to grow anaerobically (table 3). The explication could be that acetyl-CoA is an essential metabolite for biosynthesis that is produced primarily by PFL during fermentative growth. As mentioned before different optimization algorithms were used to test if there is another optimal solution.

Table 3. Theoretical maximal production rate of succinate from glycerol under different environmental

conditions and genetic modifications Designed by

us OptKnock GDLS

Conditions Targets WT 2KO 3KO 2KOa 3KOb 4KOc 2KOd 3KOe 4KOf

aerobic Biomass 0.56 0.56 0.56 0.10 0.10 0.10 0.50 0.40 0.41

Succinate - - - 0.56 1.58 2.33 0.55 1.54 2.34

microaerobic Biomass 0.33 0.30 0.30 0.10 0.10 0.10 0.30 0.20 0.18

Succinate - - 0.32 2.21 5.45 5.78 2.33 6.00 6.27

anaerobic Biomass 0.08 - - 0.10 0.10 0.10 0.07 0.07 0.07

Succinate 0.03 - - 0.12 0.18 0.24 0.08 0.12 0.16 a- aerobic Δacs-acetyl-CoA synthetase, ΔfumA; microaerobic ΔmhpF, ΔpykF-pyruvate kinase; anaerobic Δadk-adenylate kinase (GTP), Δmdh; b- aerobic Δeno-enolase, ΔfumA, ΔtpiA-triose-phosphate isomerase; microaerobic ΔgldA-glycerol dehydrogenase, Δpta, ΔpykF; anaerobic Δmdh, ΔscpC, ΔsucC; c- aerobic ΔfumA, Δgnd, ΔserA, ΔpykF; microaerobic ΔgdhA- glutamate dehydrogenase (NADP), ΔgldA, Δpta, ΔpykF; anaerobic Δmdh, ΔcysH- phosphoadenylyl-sulfate reductase (thioredoxin), ΔscpC, ΔsucC; d- aerobic ΔfumA, Δacs; microaerobic ΔadhE, ΔpykF; anaerobic Δmdh, ΔaspA-L-aspartase; e- aerobic Δeno, ΔfumA, ΔtpiA; microaerobic ΔgldA, ΔackA, ΔpykF; anaerobic Δmdh, ΔscpC, ΔsucC; f- aerobic ΔfumA, Δgnd, ΔserA, ΔpykF; microaerobic ΔackA, ΔgdhA, ΔgldA, ΔpykF; anaerobic Δmdh, ΔcysH, ΔcodA-cytosine deaminase, ΔscpC. Utilizing the optimization algorithms we were able to find alternative solutions for aerobic, microaerobic and anaerobic conditions. The designs calculated during this analysis (table 3) contain the reactions that allow the diversion of flux in E. coli and on the other hand generates sufficient energy and biomass precursors. The predicted production potential was highest under microaerobic conditions and can be increased in higher amounts by increasing the number of knockouts (table 3). Growth coupled production was not possible because the biomass production is negatively affected by these modifications. The production of different metabolites, such as formate and ethanol was necessary for the biomass synthesis during the glycerol metabolism. The ATP is consumed under biomass synthesis process and reducing equivalents (NADH) are produced. The regeneration of both is resolved by the cell producing other by-products. With these genetic modifications the synthesis of succinate remained as the primary route of NAD+ regeneration. 3.3. Dynamic FBA of diauxic growth: Time-dependent processes can be analysed by using dFBA, as proposed before (18). The dFBA approaches were used to simulate batch growth of wild-type and mutant E. coli on glucose and glycerol. Dynamic FBA was performed to simulate batch growth in minimal media conditions for mutants with 4KO using glycerol or glucose as the input and biomass, acetate, formate, ethanol, lactate and succinate as the outputs (Figure 1).



Figure 1. Model predictions using dynamic FBA for the outcomes of mutant strains; aerobic conditions designed by us (A) and predicted with OptKnock/GDLS (D); microaerobic (B) and (E); anaerobic (C) and (F)

As we can observe diauxic growth was predicted especially for microaerobic conditions, the metabolites initially secreted being subsequently utilized after glucose exhaustion. The genes selected by us can be used under microaerobic and anaerobic conditions as expected. Changing the carbon source to glycerol in minimal medium the model predicted similar results but a much longer batch time due to the slower growth on this substrate. Few examples are presented in Figure 2.

Figure 2. Model predictions using dynamic FBA for the outcomes of mutant strains; designed by us, aerobic conditions (A), microaerobic (B) and predicted with OptKnock/GDLS (D); aerobic (C) and anaerobic (D)

Glycerol can be metabolized by E. coli but the growth rate is very low, however, cell mass yield is higher on glycerol. 3.4. Phenotypic phase plane analysis: To analyse the optimal utilization of the wild-type and mutant E. coli metabolic genotype, phenotypic phase plane analysis was carried out for cellular growth in silico on glucose and glycerol. We mapped the theoretical optimal metabolic characteristics for biomass production as a function of the environmental variables such as glucose or glycerol and carbon dioxide together with growth rates. Results for anaerobic conditions are presented below (Figure 3).



Figure 3. Phenotypic phase planes for growth of mutant strains: anaerobic wt (A); anaerobic designed by us and

OptKnock (B); and anaerobic GDLS (C)

Figure 4. Phenotypic phase planes for growth of mutant strains using glycerol: anaerobic wt (A); anaerobic

designed by us and OptKnock (B); and anaerobic GDLS (C)



It becomes clear from these plots that each surface has distinct regions, representing qualitatively distinct phenotypes. The identified phenotypic phase planes are: 4 distinct regions for mutants designed by OptKnock and GDLS using glucose (Figure 4 B, C) and 3 for the rest. Phase 1 (base plane) is characterized by 0 growth. Time-course fermentation experiments were carried out for wild-type and mutant strains to follow the metabolite changes. Metabolites were identified using GC-MS. Data analysis was carried out with MassLab via comparison with mass spectra obtained from different libraries. There was a good agreement between simulations and experiments taking into account the growth rates and succinate production rates. 4. Conclusions Systems biology, including the in silico analysis has the potential to analyse complex metabolic networks and the cellular behaviour. In silico analysis is a valuable tool to carry out cellular behaviour studies, to design industrially important strains by genetic engineering and to make predictions about the cellular behaviour. This work shows how systems biology can be used to analyse the succinate production potential of E. coli from glucose and glycerol. We have found that the modification of the redox pathways is a good option to improve succinate production during glucose fermentation, however it cannot be used if the substrate is glycerol (table 2). For glycerol microaerobic conditions were found to be optimal with four eliminated genes (table 3). Dynamic growth simulations and phenotype phase planes provided a deeper understanding of differences of the metabolic flux distributions between wild-type and genetically engineered strains, especially between genotype and phenotype. Growth rates were consistent during the experimental measurements, the cultures were in the different regions on the phase planes; a few primary phenotypes were identified. The λ-Red recombineering technology was successfully used for chromosomal modifications in E. coli. The workflow presented here can be utilized as a platform to perform similar analyses for different products and for different organisms. We know that the models can contain errors and are incomplete and therefore sometimes the in silico results include incorrect predictions. However, we can reduce the cost and time of wet experiments and can obtain valuable information. The in silico strains design presented here may serve as an important contribution to the implementation of biorefineries. The utilization of crude glycerol from biodiesel industry will improve the economic feasibility and a higher-value chemical (succinate) could be obtained reducing dependency on petrochemicals and at the same time, fixing CO2, a well known green house gas. ACKNOWLEDGEMENTS This work was supported by the Sectoral Operational Programme Human Resources Development 2007-2013 of the Romanian Ministry of Labour, Family and Social Protection through the Financial Agreement POSDRU/107/1.5/S/76903, by “BIOBUILD-Synthesis of some C4, C5 carboxylic acid building block chemicals from renewable biomass resources“ PN-II-PCCA-2011-3.2-1367 and by Collegium Talentum. GC-MS analysis was assisted by Center for Organic Chemistry “Costin D. Nenitescu”.

Romanian Biotechnological Letters Vol. 19, No. 4, 2014 Copyright © 2014 University of Bucharest Printed in Romania. All rights reserved

ORIGINAL PAPER


References

1. C. LI, K.L. LESNIK, H. LIU, Microbial Conversion of Waste Glycerol from Biodiesel Production into Value-Added Products, Energies, 6(9), 4739-4768 (2013).

2. X. ZHANG, K.T. SHANMUGAM, L.O. INGRAM, Fermentation of Glycerol to Succinate by Metabolically Engineered Strains of Escherichia coli, Appl. Environ. Microbiol., 76(8), 2397-2401 (2010).

3. X. FAN, R. BURTON, Y. ZHOU, Glycerol (Byproduct of Biodiesel Production) as a Source for Fuels and Chemicals – Mini Review, The Open Fuels & Energy Science Journal, 3, 17, 22 (2010).

4. G. PAULO DA SILVA, M. MACK, J. CONTIERO, Glycerol: A promising and abundant carbon source for industrial microbiology, Biotechnol. Adv., 27, 30-39 (2009).

5. H. YIM, R. HASELBECK, W. NIU, C. PUJOL-BAXLEY, A. BURGARD, J. BOLDT, J. KHANDURINA, J.D. TRAWICK, R.E. OSTERHOUT, R. STEPHEN, J. ESTADILLA, S. TEISAN, H.B. SCHREYER, S. ANDRAE, T.H. YANG, S.Y. LEE, M.J. BURK, S. VAN DIEN, Metabolic engineering of Escherichia coli for direct production of 1,4-butanediol, Nat. Chem. Biol., 7, 445-452 (2011).

6. B.J. MCKINLAY, C. VIEILLE, G.J. ZEIKUS, Prospects for a bio-based succinate industry, Appl. Microbiol. Biotechnol., 76, 727-740 (2007).

7. E. TAKIYAMA, E. FUJIMAKI, I. NIIKURA, Y. HATANO, Method for producing saturated polyester, US Patent 5, 306-787 (1994).

8. J. WANG, J. ZHU, N.G. BENNETT, Y.K. SAN, Succinate production from different carbon sources under anaerobic conditions by metabolic engineered Escherichia coli strains, Metab. Eng., 13, 328-335 (2011).

9. S. ATSUMI, J.C. LIAO, Metabolic engineering for advanced biofuels production from Escherichia coli, Curr. Opin. Biotechnol., 19(5), 414-419 (2008).

10. M. CHARTRAIN, P.M. SALMON, D.K. ROBINSON, B.C. BUCKLAND, Metabolic engineering and directed evolution for the production of pharmaceuticals, Curr. Opin. Biotechnol., 11(2), 209-214 (2000).

11. J.D. ORTH, I. THIELE, B.Ø. PALSSON, What is flux balance analysis?, Nat. Biotechnol., 28, 245-248 (2010).

12. K. RAMAN, N. CHANDRA, Flux balance analysis of biological systems: applications and challenges, Brief. Bioinforma., 10, 435-449 (2009).

13. N.D. PRICE, J.A. PAPIN, C.H. SCHILLING, B.Ø. PALSSON, Genome-scale microbial in silico models: the constraints-based approach, Trend Biotechnol., 21(4), 162-169 (2003).

14. S.J. LEE, D.Y. LEE, T.Y. KIM, B.H. KIM, J. LEE, S.Y. LEE, Metabolic engineering of Escherichia coli for enhanced production of succinate, based on genome comparison and in silico gene knockout simulation, Appl. Environ. Microbiol., 71, 7880-7887 (2005).

15. P. VILAÇA, I. ROCHA, M. ROCHA, A computational tool for the simulation and optimization of icrobial strains accounting integrated metabolic/regulatory information, Biosystems, 103, 435-441 (2011).

16. S. OHNO, C. FURUSAWA, H. SHIMIZU, In silico screening of triple reaction knockout Escherichia coli strains for overproduction of useful metabolites, J. Biosci. Bioeng., 115, 221-228 (2013).

17. S.A. BECKER, A.M. FEIST, M.L. MO, G. HANNUM, B.Ø. PALSSON, M.J. HERRGARD, Quantitative prediction of cellular metabolism with constraint-based models: the COBRA Toolbox, Nat. Protoc., 2, 727-738 (2007).

18. J. SCHELLENBERGER, R. QUE, R.M. FLEMING, I. THIELE, J.D. ORTH, A.M. FEIST, D.C. ZIELINSKI, A. BORDBAR, N.E. LEWIS, S. RAHMANIAN, J. KANG, D.R. HYDUKE, B.Ø. PALSSON, Quantitative prediction of cellular metabolism with constraint-based models: the COBRA Toolbox v2.0, Nat. Protoc., 4, 1290-1307 (2011).

19. A.P. BURGARD, P. PHARKYA, C.D. MARANAS, Optknock: a bilevel programming framework for identifying gene knockout strategies for microbial strain optimization, Biotechnol. Bioeng., 84(6), 647-657 (2003).



20. D.S. LUN, G. ROCKWELL, N.J. GUIDO, M. BAYM, J.A. KELNER, B. BERGER, J.E. GALAGAN, G.M. CHURCH, Large-scale identification of genetic design strategies using local search, Mol. Syst. Biol., 5, 296 (2009).

21. K.A. DATSENKO, B.L. WANNER, One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products, Proc. Natl. Acad. Sci. USA., 97, 6640-6645 (2000).

22. C. XU, L. LIU, Z. ZHANG, D. JIN, J. QIU, M. CHEN, Genome-scale metabolic model in guiding metabolic engineering of microbial improvement, Appl. Microbiol. Biotechnol., 97(2), 519-539 (2013).

23. C.S. THAKUR, M.E. BROWN, J.N. SAMA, M.E. JACKSON, T.K. DAYIE, Growth of wildtype and mutant E. coli strains in minimal media for optimal production of nucleic acids for preparing labelled nucleotides, Appl. Microbiol. Biotechnol., 88(3), 771-779 (2010).

24. J.D. ORTH, T.M. CONRAD, J. NA, J.A. LERMAN, H. NAM, A.M. FEIST, B.Ø. PALSSON, A comprehensive genome-scale reconstruction of Escherichia coli metabolism-2011, Mol. Syst. Bio., 7, 535 (2011).

25. H.J. KIM, B.K. HOU, S.G. LEE, J.S. KIM, D.W. LEE, S.J. LEE, Genome-wide analysis of redox reactions reveals metabolic engineering targets for D-lactate overproduction in Escherichia coli, Metab. Eng., 18, 44-52 (2013).

26. T.G. LUAN, K.S. YU, Y. ZHONG, H.W. ZHOU, C.Y. LAN, N.F. TAM, Study of metabolites from the degradation of polycyclic aromatic hydrocarbons (PAHs) by bacterial consortium enriched from mangrove sediments, Chemosphere, 65(11), 2289-2296 (2006).

27. Q. WANG, C. WU, T. CHEN, X. CHEN, X. ZHAO, Expression of galactose permease and pyruvate carboxylase in Escherichia coli ptsG mutant increases the growth rate and succinate yield under anaerobic conditions, Biotechnol Lett., 28(2), 89-93 (2006).

Documents

Systems biology and metabolic engineering for obtaining E ...rombio.eu/vol19nr4/lucr 23 Zs. Bodor 2014_rec 6 febr...Keywords: systems biology, biotechnology, succinate, Escherichia