'13TSB_SynBio' Genome-scale metabolic modelling to optimise high value microbial manufacturing
Lead Research Organisation:
University of Cambridge
Department Name: Biochemistry
Abstract
We are beginning to see a shift in focus from of large-scale biotechnology away from biofuels (such as bioethanol and biobutanol) towards fine and bulk chemicals (e.g. succinic acid and propanediol) production. Methacrylic acid is the most widely used starting material to produce methylmethacrylate (MMA) for the manufacture of Perspex (also known as Lucite). MMA is a primary building block for the methacrylates industry and supports an extensive supply chain globally. All of this material is currently produced from petrochemical feedstocks. As a major player in the production of MMA (24% of total global production) one of the participating companies in this project, Lucite, is committed to developing a route based on renewable feedstocks. This project aims to engineer an industrial microbe to produce commercially viable amounts of a methacrylate precursor. The diversion of a large proportion of a microbe's resources to the generation of a product of no benefit to itself generally involves extensive realignment of its metabolic network. Scientists at the University of Cambridge will construct a computer model of the microbe's metabolism and use this in simulations to guide the engineering of the microbe by the third partner in this grouping, Ingenza. Together the three partners will optimise and scale-up the process.
Both the general methodologies followed and specific results obtained will be of interest to biotechnologists, since they will be applicable to develop routes to many other bulk and speciality chemicals using renewable feedstocks. Success in a bulk chemical-based project of this type will help to promote the UK's aim to be at the leading edge of a Bio-based economy. Success in this project will also give impetus, where needed, to other chemical producers to engage in industrial biotechnology and the development of Bio-based processes. Furthermore, there will be extensive benefits throughout the supply chain associated with Lucite's activities, thus providing more general benefits across the UK economy.
Both the general methodologies followed and specific results obtained will be of interest to biotechnologists, since they will be applicable to develop routes to many other bulk and speciality chemicals using renewable feedstocks. Success in a bulk chemical-based project of this type will help to promote the UK's aim to be at the leading edge of a Bio-based economy. Success in this project will also give impetus, where needed, to other chemical producers to engage in industrial biotechnology and the development of Bio-based processes. Furthermore, there will be extensive benefits throughout the supply chain associated with Lucite's activities, thus providing more general benefits across the UK economy.
Technical Summary
Methacrylic acid is the most widely used starting material to produce methylmethacrylate (MMA) for the manufacture of Perspex (also known as Lucite). MMA is a primary building block for the methacrylates industry and supports an extensive supply chain globally. All of this material is currently produced from petrochemical feedstocks. As a major player in the production of MMA (24% of total global production) one of the participating companies in this project, Lucite, is committed to developing a route based on renewable feedstocks. This project aims to engineer an industrial microbe to produce commercially viable amounts of a methacrylate precursor. The diversion of a large proportion of a microbe's resources to the generation of a product of no benefit to itself generally involves extensive realignment of its metabolic network. Scientists at the University of Cambridge will construct a computer model of the microbe's metabolism and use this in simulations to guide the engineering of the microbe by the third partner in this grouping, Ingenza. Together the three partners will optimise and scale-up the process.
Planned Impact
Methacrylic acid is the most widely used starting material to produce methylmethacrylate (MMA). MMA is a primary building block for the methacrylates industry and supports an extensive supply chain globally. All of this material is currently produced from petrochemical feedstocks. As a major player in the production of MMA (24% of total global production) Lucite is committed to developing a route based on renewable feedstocks. The success of this project will be of huge significance in helping Lucite to advance toward this goal. Furthermore, success in a bulk chemical-based project of this type will help to promote the UK's aim to be at the leading edge of a Bio-based economy. We are beginning to see a shift in focus from biofuels (bioethanol and biobutanol) towards fine and bulk chemicals (e.g. succinic acid and propanediol) production. Success in this project will give impetus, where needed, to other chemical producers to engage in industrial biotechnology and the development of Bio-based processes. Furthermore, there will be extensive benefits throughout the supply chain associated with Lucite's activities, thus providing more general benefits across the UK economy.
Ingenza will exploit their proprietary approaches to engineering microbes and Lucite will lead the economic evaluation and process development work following on from this project. The collaboration will help both Ingenza and Lucite to build their capabilities in fermentation technology, molecular biology and metabolic modelling, and future scale up will be developed with NIBF (at an adjacent site at Wilton). It is envisaged that graduate and postgraduate employment opportunities will arise as this technology develops.
We also expect UK academia to benefit significantly from the work carried out in this project. It will also inform and refine the industrial relevance of the metabolic modelling and simulations at which the Cambridge Systems Biology Centre (CSBC) is expert and which the Oliver lab has deployed in support of local Biotech companies. This will be CSBC's first opportunity to interact directly with a major chemicals manufacturer and this will provide valuable insights into the problems of scale-up and downstream processing.
The wider academic community will also benefit from our dissemination activities. Clearly IP issues will need to be addressed first and this is vital for the competitiveness of Ingenza and Lucite, and for the establishment of revenue streams to the University. However, once these aspects have been addressed, we expect publication of results in respected peer-reviewed journals. Both the general methodologies followed and specific results obtained will be of interest to academia both in the UK and internationally, since they will be applicable to develop routes to many other bulk and speciality chemicals using renewable feedstocks.
Ingenza will exploit their proprietary approaches to engineering microbes and Lucite will lead the economic evaluation and process development work following on from this project. The collaboration will help both Ingenza and Lucite to build their capabilities in fermentation technology, molecular biology and metabolic modelling, and future scale up will be developed with NIBF (at an adjacent site at Wilton). It is envisaged that graduate and postgraduate employment opportunities will arise as this technology develops.
We also expect UK academia to benefit significantly from the work carried out in this project. It will also inform and refine the industrial relevance of the metabolic modelling and simulations at which the Cambridge Systems Biology Centre (CSBC) is expert and which the Oliver lab has deployed in support of local Biotech companies. This will be CSBC's first opportunity to interact directly with a major chemicals manufacturer and this will provide valuable insights into the problems of scale-up and downstream processing.
The wider academic community will also benefit from our dissemination activities. Clearly IP issues will need to be addressed first and this is vital for the competitiveness of Ingenza and Lucite, and for the establishment of revenue streams to the University. However, once these aspects have been addressed, we expect publication of results in respected peer-reviewed journals. Both the general methodologies followed and specific results obtained will be of interest to academia both in the UK and internationally, since they will be applicable to develop routes to many other bulk and speciality chemicals using renewable feedstocks.
People |
ORCID iD |
| Stephen Oliver (Principal Investigator) |
Publications
Cankorur-Cetinkaya A
(2017)
CamOptimus: a tool for exploiting complex adaptive evolution to optimize experiments and processes in biotechnology.
in Microbiology (Reading, England)
Dikicioglu D
(2019)
Extension of the yeast metabolic model to include iron metabolism and its use to estimate global levels of iron-recruiting enzyme abundance from cofactor requirements.
in Biotechnology and bioengineering
Kell DB
(2016)
The metabolome 18 years on: a concept comes of age.
in Metabolomics : Official journal of the Metabolomic Society
Kell DB
(2015)
Membrane transporter engineering in industrial biotechnology and whole cell biocatalysis.
in Trends in biotechnology
Mendes P
(2015)
Fitting Transporter Activities to Cellular Drug Concentrations and Fluxes: Why the Bumblebee Can Fly.
in Trends in pharmacological sciences
| Description | The overall objective of this project was to optimize the biological production of citramalate under anaerobic or low levels of oxygen using E. coli and C. glutamicum as expression systems and a genome-scale metabolic modeling approach. Natural occurrence of this metabolite is not observed in these two industrial host organisms. Therefore, the main challenge is to identify the optimal combination of genes insertions/deletions to engineer the organisms to produce citramalate without compromising cell viability. The experimental validation of the potential combinations becomes costly and time-consuming due to the large number of the candidate combinations for genetic manipulations. The use of genome-scale modeling to perform these modifications in silico is a faster and very much cheaper way to develop this process. The computational study involved in this project included the following steps: - Identification of an appropriate genome-scale model for each expression system. - Data preprocessing using an automated program for fitting linear and non-linear functions to the metabolite concentrations (e.g. glucose, organic acids and biomass) to estimate their consumption/production rates. - Development of a suitable constraint-based genome-scale model (e.g. a stoichiometric model for Flux Balance Analysis). This step required the insertion/deletion of genes or reactions to/from the model to reproduce the initial strains, fine-tuning of the upper and lower bounds of the metabolic reactions and identification of the cellular objective functions of the organism (e.g. maximization or minimization of one or two metabolic reactions). The multi-objective Flux Balance Analysis approach was developed and implemented for the first time with a biological process in this project. Preferably, the model should be able to predict the synthesis of the main product, including biomass formation, using as input the estimated uptake flux of the main substrate (e.g. glucose). - Simulation of the metabolic flux distribution under different gene insertions/deletions and substrates (e.g. glucose, oxygen and nitrate) uptake rates. - Development of an optimization program based on genetic algorithms to find the optimal combination of gene insertions/deletions that maximizes the process objective function (e.g. maximization of biomass and/or the target product). This program was suitable for both the design of both the cell culture medium and the strain. - Ideally, iterative experimental validation of the optimized combination of gene insertions/deletions and cell culture media composition to fine-tune the model. - Implementation of a graphical user interface (GUI) to facilitate the use by experimentalists of the computational tools developed in this project. The GUI includes two different executable programs: one to create and adapt genome-scale models and one to perform Flux Balance Analysis using single and double cellular objective functions, to visualize the results and to save the metabolic flux distribution. Moreover, the program allows the user to analyze the effect of an additional constraint to a metabolic reaction on the results. A manual with the instructions for using both programs was provided. A brief description of the inputs used for the computational study and the specific outcomes for each expression will be presented below. C. glutamicum Metabolic model The C. glutamicum model proposed by Kjeldsen and Nielsen [1] and later modified by Woo et al. [2] and Ooyen et al. [3] was considered in this project. The model was further modified to integrate gene deletions present in BOL-3 strain [4] and gene insertions involved in the transport and production of citramalate. Ferredoxin regeneration reactions were also included in the model. Experimental data Data from 25 biotransformation (production phase under anaerobic conditions) and 8 (fed)-batch (growth + succinate production phase) experiments to validate the model were provided. Some experiments were excluded from the analysis because they did not contain enough data to calculate the fluxes or the main substrate was not consumed. Additionally, the (fed)-batch experiments were not considered during the project due to issues with the concentration of glucose in the feed and the unpredictable composition of oxygen in the inlet and outlet gasses. For the biotransformation experiments data, second-order functions were fitted to the measured metabolite concentrations and their consumption/production rates were estimated from the derivative of the function, converted from g/l/h to mmol/l/h and normalized to the biomass concentration. Single-objective Flux balance analysis The metabolic flux distribution over time for most of the biotransformation experiments was determined using the C. glutamicum model and the consumption rate of glucose as input. The model performance was evaluated using the succinate production rates. The best results for the prediction of succinate were achieved using the minimization of NAD+ utilization as cellular objective function. FBA model simulations The model was used to simulate different scenarios of gene deletions (discussed with Ingenza) and nitrate (discussed with Lucite) and glucose levels to evaluate the conditions that promote glutamate production in the presence of low levels of oxygen or in its complete absence. Nitrate reduction was considered as an alternative pathway to oxygen reduction in the regeneration of NAD(P)+ species. Glutamate production was not observed under any of the simulated scenarios without further genetic modifications to the strains. Strain engineering The model-based optimization program for strain engineering was implemented using the FBA model and the maximization of glutamate production as process objective functions. The optimization results suggested the deletion of the gene encoding glucose-6-phosphate 1-dehydrogenase (EC 1.1.1.49) subunit or 6-phosphogluconate dehydrogenase (EC 1.1.1.43) subunit, both part of a heteromultimeric complex [5], and the deletion of the gene encoding cystathionine gamma-synthase (EC 2.5.1.48) or cystathionine beta-lyase (EC 4.4.1.8), both from cysteine and methionine metabolism. This double deletion enables glutamate production under strictly anaerobic conditions. Alternatively, the model suggests the double deletion of the genes encoding malate synthase (EC 2.3.3.9) and 3-methyl-2-oxobutanoate hydroxymethyltransferase (EC 2.1.2.11) for glutamate production under both anaerobic and aerobic conditions. These results need to be validated experimentally. Escherichia coli Metabolic network The E. coli K-12 MG1655 proposed by Feist et al. [6] was adapted to incorporate the deletion of the genes araD, araB, rhaD, rhaB, ldhA, pflB and the insertion of genes involved on the synthesis of citramalate (EC: 2.3.1.182) and citraconate (EC: 4.2.1.35) and their respective production and exchange reactions. Data preprocessing Data from 12 biotransformation (organic acids production phase under anaerobic conditions) and 8 fed-batch (growth + citramalate production phase) experiments to validate the metabolic model were provided. Most of the data provided from biotransformation experiments (10 out of 12) were excluded because glucose was apparently not consumed. Only data from fed-batch experiments were considered for modelling purposes. Uptake and production rates were estimated from the slope of linear functions fitted to the experimental data during the growth phase and (growth +) citramalate production phase. The slopes were converted from g/l/h to mmol/l/h and normalized to the biomass concentration. Four out of the 8 fed-batch experiments were also excluded from the modeling study. Two experiments presented high specific growth rate relative to their specific glucose uptake rate and two very high observed glucose uptake rates. Multiobjective Flux balance analysis The FBA model for E. coli was developed using the data from the four remaining experiments and using the consumption rate of glucose as input and the growth and citramalate production rates as output. The cellular objective functions that better predict both growth and citramalate rates were the maximization of growth followed by the maximization of citramalate. A significant improvement in the results, when compared to the single-objective optimization was observed, especially in the experiments with simultaneous growth and citramalate production. The results revealed that two of the experiments were performed under oxygen limitation. This result was later confirmed by Ingenza. This fact may explain the extension of the growth phase over the citramalate production phase in these two experiments. Metabolic flux distribution The results from the average metabolic flux distribution obtained for the four experiments suggest that, under oxygen limitation, part of acetyl-CoA required for citramalate and citraconate production is produced from PEP via aspartate, threonine and serine pathway. Additionally, under such conditions, the reactions FBA (EC 4.1.2.13) and PFK (EC 2.7.1.11) are activated. The reason that could explain the use of both these pathways is its efficiency in energetic terms. An alternative explanation is that these pathways are activated by growth. For both situations, and for greater energetic efficiency, it would be interesting to promote simultaneous growth and citramalate production. Under excess oxygen, the energy required to produce acetyl-CoA from pyruvate is recovered through the TCA cycle. Cell culture medium optimization An opportunity for cell culture medium optimization was explored after analyzing the results obtained in the two experiments performed under oxygen limitation, when the growth phase was extended to the citramalate production phase. Model simulations were performed to identify the nutrients required for different growth rates at a fixed supply of glucose. According to the model, it was recommended to increase the amount of Ca+2, Co+2, Cu+2, Mn+2, Zn+2 and NH4+ to prevent any growth limitations under excess oxygen conditions and at different growth rate levels. These results need to be validated experimentally. |
| Exploitation Route | Will be pursued by industrial partners Lucite & Ingenza. The work in this project will also contribute to a IB Catalyst Programme. |
| Sectors | Chemicals |
| Description | Tool for construction and manipulation of genome-scale metabolic model to industrial collaborators. |
| First Year Of Impact | 2016 |
| Sector | Chemicals,Manufacturing, including Industrial Biotechology |
| Impact Types | Economic |
| Description | IB Catalyst |
| Amount | £525,090 (GBP) |
| Funding ID | BB/N02348X/1 |
| Organisation | Biotechnology and Biological Sciences Research Council (BBSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 09/2016 |
| End | 08/2021 |
| Title | CamOptimus |
| Description | CamOptimus is a multi parameter optimisation tool. It exploits a Genetic Algorithm to discover the subspace containing the optimal combination of parameters, and Symbolic Regression to construct a model to evaluate the sensitivity of the experiment to each parameter under investigation. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2017 |
| Provided To Others? | Yes |
| Impact | A number of researchers have used CamOptimus to optimise both microbial growth media for production of proteins or organelles, and for the optimisation of growth protocols to generate quantum dots. |
| URL | https://github.com/DuyguD/CamOptimus.git |