Mass Spectrometry-based Workflow for Accurate Quantification of Escherichia coli Enzymes: How Proteomics Can Play a Key Role in Metabolic Engineering

Metabolic engineering aims to design high performance microbial strains producing compounds of interest. This requires systems-level understanding; genome-scale models have therefore been developed to predict metabolic fluxes. However, multi-omics data including genomics, transcriptomics, fluxomics, and proteomics may be required to model the metabolism of potential cell factories. Recent technological advances to quantitative proteomics have made mass spectrometry-based quantitative assays an interesting alternative to more traditional immuno-affinity based approaches. This has improved specificity and multiplexing capabilities. In this study, we developed a quantification workflow to analyze enzymes involved in central metabolism in Escherichia coli (E. coli). This workflow combined full-length isotopically labeled standards with selected reaction monitoring analysis. First, full-length 15N labeled standards were produced and calibrated to ensure accurate measurements. Liquid chromatography conditions were then optimized for reproducibility and multiplexing capabilities over a single 30-min liquid chromatography-MS analysis. This workflow was used to accurately quantify 22 enzymes involved in E. coli central metabolism in a wild-type reference strain and two derived strains, optimized for higher NADPH production. In combination with measurements of metabolic fluxes, proteomics data can be used to assess different levels of regulation, in particular enzyme abundance and catalytic rate. This provides information that can be used to design specific strains used in biotechnology. In addition, accurate measurement of absolute enzyme concentrations is key to the development of predictive kinetic models in the context of metabolic engineering. Metabolic engineering aims to design high performance microbial strains producing compounds of interest. This requires systems-level understanding; genome-scale models have therefore been developed to predict metabolic fluxes. However, multi-omics data including genomics, transcriptomics, fluxomics, and proteomics may be required to model the metabolism of potential cell factories. Recent technological advances to quantitative proteomics have made mass spectrometry-based quantitative assays an interesting alternative to more traditional immuno-affinity based approaches. This has improved specificity and multiplexing capabilities. In this study, we developed a quantification workflow to analyze enzymes involved in central metabolism in Escherichia coli (E. coli). This workflow combined full-length isotopically labeled standards with selected reaction monitoring analysis. First, full-length 15N labeled standards were produced and calibrated to ensure accurate measurements. Liquid chromatography conditions were then optimized for reproducibility and multiplexing capabilities over a single 30-min liquid chromatography-MS analysis. This workflow was used to accurately quantify 22 enzymes involved in E. coli central metabolism in a wild-type reference strain and two derived strains, optimized for higher NADPH production. In combination with measurements of metabolic fluxes, proteomics data can be used to assess different levels of regulation, in particular enzyme abundance and catalytic rate. This provides information that can be used to design specific strains used in biotechnology. In addition, accurate measurement of absolute enzyme concentrations is key to the development of predictive kinetic models in the context of metabolic engineering. The concept of producing economically useful compounds using a “cell factory” was introduced with the advent of recombinant DNA technologies, and became fully established in the 1980s when the Food and Drug Administration approved the use of recombinant insulin (1Johnson I.S. Human insulin from recombinant DNA technology.Science. 1983; 219: 632-637Crossref PubMed Scopus (250) Google Scholar). This concept is now largely used in the food, pharmaceutical and biotechnology industries for the production of various compounds such as polypeptides, l-threonine (2Lee J.H. Sung B.H. Kim M.S. Blattner F.R. Yoon B.H. Kim J.H. Kim S.C. Metabolic engineering of a reduced-genome strain of Escherichia coli for L-threonine production.Microb. Cell Fact. 2009; 8: 2Crossref PubMed Google Scholar), or artemisinin, an anti-malaria drug (3Ro D. Paradise E.M. Ouellet M. Fisher K.J. Newman K.L. Ndungu J.M. Ho K.A. Eachus R.A. Ham T.S. Kirby J. Chang M.C.Y. Withers S.T. Shiba Y. Sarpong R. Keasling J.D. Production of the antimalarial drug precursor artemisinic acid in engineered yeast.Nature. 2006; 440: 940-943Crossref PubMed Scopus (2154) Google Scholar) using adapted Escherichia coli (E. coli) 1The abbreviations used are:DNADeoxyribonucleic AcidE. coliEscherichia colimRNAmessenger Ribonucleic AcidRNARibonucleic AcidGEMsGenome-scale modelsCBMConstraint-Based ModelELISAEnzyme-Linked Immuno-Sorbent assayLC-MSLiquid Chromatography-Mass SpectrometrySRMSelected Reaction MonitoringSIDStable of and acid of of 1The abbreviations used are:DNADeoxyribonucleic AcidE. coliEscherichia colimRNAmessenger Ribonucleic AcidRNARibonucleic AcidGEMsGenome-scale modelsCBMConstraint-Based ModelELISAEnzyme-Linked Immuno-Sorbent assayLC-MSLiquid Chromatography-Mass SpectrometrySRMSelected Reaction MonitoringSIDStable of and acid of of a of and cell now used to from advances in metabolic engineering D. metabolic biotechnology and microbial cell Cell Fact. PubMed Scopus Google Scholar). Escherichia coli Ribonucleic Ribonucleic models Immuno-Sorbent Liquid Chromatography-Mass Reaction of and acid of of Escherichia coli Ribonucleic Ribonucleic models Immuno-Sorbent Liquid Chromatography-Mass Reaction of and acid of of Metabolic engineering in the and of the cell to of a Metabolic engineering requires to that can be to J. Keasling J.D. and metabolic PubMed Scopus Google Scholar). can the of a cell including of reaction metabolic of of or enzyme and metabolic the different levels of the cell be and to to this of metabolic can be to models metabolic to be a a PubMed Scopus Google Scholar) M. to the PubMed Scopus Google Scholar). and The of of genome-scale metabolic using Escherichia PubMed Scopus Google Scholar), the of such genome-scale models can be key of of a and of models for metabolic engineering. quantitative data with reference to PubMed Scopus Google Scholar) The and 8: Google Scholar) has been the of high and and the high of development The and The of the PubMed Scopus Google Scholar). M. mass in from to the PubMed Scopus Google Scholar), and advances in have made quantitative liquid a alternative for data using labeled can specific and accurate is for in with Reaction two levels of and specific R. reaction monitoring for quantitative a PubMed Scopus Google Scholar), to and quantify with high and specificity in a single with absolute quantification in proteomics in a of isotopically labeled or standards the and can be in of and mass In the and of the different in the information the of to be from the of the and the quantification of specific model with PubMed Scopus Google Scholar). standards have been such as labeled standards standards J. quantification of and from cell PubMed Scopus Google Scholar). standards of labeled that can be J.M. absolute quantification in proteomics using of PubMed Scopus Google Scholar). the of is and the of measurements be full-length isotopically labeled standards to in the as the M. D. M. J. absolute quantitative PubMed Scopus Google Scholar). J. for absolute quantitative 2009; PubMed Scopus Google D. M. M. M. J. quantification of in using and selected reaction PubMed Scopus Google Scholar) for quantitative in the context of when mass is combined with standards to the of more the requires the of full-length recombinant in a labeled and may specific to genome-scale models is This is from that an accurate quantitative proteomics can be used in combination with genome-scale models to in the design of cell in metabolic engineering. The Escherichia coli (E. is of the used cell and has been in and used to models The of of genome-scale metabolic using Escherichia PubMed Scopus Google Scholar). central metabolism is involved in and of the and the produced E. coli derived from precursor or and J. is metabolic PubMed Scopus Google Scholar). This an for metabolic engineering M. Recent advances in engineering the central metabolism of Cell Fact. PubMed Scopus Google Scholar). to use a to the concentrations of enzymes involved in E. coli with high in a we optimized a producing full-length be fully with were the E. coli and were using to J. R. R. of reaction monitoring of PubMed Scopus Google Scholar). were using The monitoring a single was used to accurately quantify 22 key enzymes involved in in E. coli and this workflow we a wild-type reference E. coli strain and two strains with an J. of Escherichia coli to in PubMed Scopus Google Scholar). measurements were combined with metabolic measurements for or catalytic of the enzymes Chang models of metabolism and and PubMed Scopus Google Scholar) R. Escherichia coli catalytic and of PubMed Scopus Google Scholar). The that is for metabolic engineering and is useful for a in is of a in the or to the catalytic of an data and this to be useful in kinetic models of in central involved in and of the 22 enzymes were produced in and in a quantification workflow to analyze enzymes involved in Escherichia coli central combined with an was developed using full-length 15N labeled standards to ensure ensure a were Reaction was used for and were from E. coli and E. coli were from and were from Fisher The and were from and were from was from was from and were from were from reference was from the of and and were from were for the of were from using and with and of is in were using and with the enzymes in were then a an and a of the the E. coli strains were and were using the of the was and a were then E. coli strains were in and wild-type E. coli strains were from J. of Escherichia coli to in PubMed Scopus Google Scholar), in were in were the based coli strains of the strains were in strain to an strain from the of to the of a for NADPH in and in a E. coli strains were was as labeled was used to labeled was with the strains of interest. in was with to an of was with and were to were was and were in the of were in for were and was were using to the The of in was adapted for to the and were were or in 15N recombinant was of 15N labeled as was in data a as with 15N were for selected 15N using The was and using a were calibrated using the to M. Y. J. M. a and for acid using and mass PubMed Scopus Google Scholar). the was with of was using the were in of The from an was used to of a a of the of an Fisher was for acid over The of acid in the and the were used to more M. Y. J. M. a and for acid using and mass PubMed Scopus Google Scholar). or E. coli were in for with and for in the with were with to a of was for an of were with to to for an of were and in a and were a mass Liquid chromatography was an to a were using a to over a of were a mass using a in in precursor was as a for the were with a of using a for and and were and is in with the of the were in with the and potential were to and was using a based the for the mass was in with a of and a of the was using was and were of was selected from involved in the central metabolic in E. coli in labeled and was and to the was a mass were using were to using a used as the a data from to with Escherichia coli was as the enzyme and was was as a and of was as a for and was and was and and was may be from data were used to the for of the 22 in to specific was were used to of for and of selected were using M. R. an for and proteomics PubMed Scopus Google Scholar). were and the was to be using the In with M. M. M. D. J. spectrometry-based absolute use of PubMed Scopus Google Scholar), were in the In were The for the selected were and were selected to a the This in of for in this selected were The higher in the were for a for were that the was for and may be from were using The for levels was and for the data were to ensure and accurate may be from The was for and were for the of a The for the the different from a were to the and were the and levels the and the quantification was from the the of the the of the The enzyme were to concentrations and of cell based data M. cell and of Escherichia coli to data for PubMed Scopus Google Scholar) J. E. D. Metabolic of Escherichia coli of and PubMed Scopus Google Scholar), as in In the a metabolic reaction can be as is the enzyme and is the or catalytic as in two Chang models of metabolism and and PubMed Scopus Google Scholar) R. Escherichia coli catalytic and of PubMed Scopus Google Scholar). was from measurements of and for enzyme in the strains as in that in the and strains to the reference can be the of the enzyme concentrations and the catalytic and enzyme were from the data for the and strains, and catalytic were The of this was to use a proteomics quantitative to ensure accurate data for the of enzyme concentrations and catalytic to strains and for predictive kinetic central metabolism in E. coli were including the the and the acid to key enzymes in involved in of the and from the full-length 15N labeled standards were produced in E. coli and ensure that labeled standards were produced in conditions were optimized to The was to ensure and the and in a single of a of PubMed Scopus Google Scholar). the of or 15N was to the rate. This is for accurate were using and to be for This with the of the used of standards were produced to was a the of standards and was as M. Y. J. M. a and for acid using and mass PubMed Scopus Google Scholar). In with and full-length 15N labeled standards can be using to were and to producing or and may be therefore to the of the standards and as the for was in The of for assays were and and of were therefore a to of labeled standards were to the for key enzymes involved in E. coli a of and to be be to ensure the specificity of M. R. an for and proteomics PubMed Scopus Google Scholar) was developed to this This requires two and for the were and of and and data were used to the were with is a producing were the selected were of in the E. coli and with M. D. J. R. R. of for quantitative PubMed Scopus Google Scholar). The for the of J. R. R. of reaction monitoring of PubMed Scopus Google Scholar) were were to the of for was to a the of to be and of the analysis. to analyze a with a This was in a of to be used with was to is the for quantification M. R. M. Recent in liquid and PubMed Scopus Google Scholar). conditions were optimized to ensure to a single 30-min 22 enzymes be and accurately in a single analysis. The of the quantification were using This involved of to an E. coli a of to The of was in from the to assess from to the potential as the Food and Drug workflow was including in conditions were were to the and the of in were the were as This to key of the of the is the of the of the is the to the for the of the The that is and that the is over the The of was and and in the was higher The for was with of a of can be for a were data that the be for was for the enzymes and for was in E. coli for the were used for for was were used of the was and were were based The were based the that be with a for and an and in the of from of from in a The is an of performance as the that can be accurately a of an of is used to the M. E. assays for abundance in mass and PubMed Scopus Google Scholar). However, with is and is to accurately therefore in with J. to Google Scholar) D. J. reaction absolute of in 2009; 8: PubMed Scopus Google Scholar), to the of the the that can be with a and an and in the is as the were for using data from for in The of was the in the or to the abundance of were for and and The for was therefore the of for The for of the assays in The that is and we therefore to use the labeled standards and the to enzyme concentrations in selected E. coli the of using labeled standards to E. coli enzyme we to the wild-type reference E. coli strain standards were E. coli The in was then were as were and the concentrations of were were for and was in quantification was for the 22 The was the was The for concentrations were and absolute quantification of the wild-type E. coli we a single The for the wild-type strain that this workflow is and has a high multiplexing and quantification for in E. coli The proteomics workflow was used to accurately quantify 22 key enzymes in E. coli central metabolism in with with standards were and were for concentrations in of The of based data for from of to the based to the based to the based to the based to the based to the based to the based to the based in a In of the was used to concentrations in two strains of engineered to high levels of NADPH and NADPH is involved in a when a strains producing high levels of NADPH interesting for metabolic engineering and in a recombinant Escherichia coli for PubMed Scopus Google Scholar). The for is in the and we the abundance of of the 22 enzymes in the wild-type The absolute from were as a of of cell and The in the enzymes in of the strains, and to in abundance The for measurements was the was to be cell in strain the was to be cell The of for for assays and of were for strain the cell for and were and with of and This the of the the from using and that the workflow developed is and can be used to accurate and data of the of metabolic engineering is to metabolic in as to the production of compounds of J. Metabolic and Google Scholar). Metabolic different including enzyme kinetic enzyme and in and the concentrations of and and of metabolic in PubMed Scopus Google Scholar). used a model of metabolic fluxes, the of the abundance and the catalytic of the The model the that metabolic to enzyme concentrations R. The of Google Scholar). metabolic enzyme and the catalytic rate. The catalytic the catalytic is concentrations of and as as the catalytic of the a is a enzyme enzyme and the catalytic is the catalytic The model can be used when in for the kinetic is the were for two of the in the for the of for the metabolic as J. of Escherichia coli to in PubMed Scopus Google Scholar), and the enzyme concentrations from the in this The catalytic be from the data to This can be used to a strains to concentrations and catalytic of the wild-type reference strain with strains and The metabolic in or to the of the enzyme concentrations and the catalytic a in metabolic in strain be of a in enzyme and a in catalytic the metabolic and enzyme and catalytic for of the enzymes in strains with and were as was data for the The for the metabolic of in enzyme for the of in the catalytic of the enzymes and in fluxes, enzyme and catalytic strains and Metabolic were enzyme concentrations were in this Metabolic and enzyme concentrations can be used to catalytic and The the in metabolic enzyme and catalytic for enzymes the reference and strains and for is from and strains, for the and data were in a J. of Escherichia coli to in PubMed Scopus Google Scholar) the strain with the wild-type strain and a in the reaction from as of the that this in from a in enzyme and a in catalytic This that the in catalytic the in metabolic have been to this based data can be made for the and for the in strain to the wild-type the enzyme concentrations for the catalytic and a in that this is the of an in enzyme and a in catalytic rate. The for in this of the central metabolism in Escherichia coli to and PubMed Scopus Google Scholar), that to of the with abundance catalytic This to a the with the In to the of enzyme concentrations and to the of catalytic may in in metabolic the catalytic is to the enzyme is from a a of the reaction an of the enzyme In the for the the is in with the wild-type strain and the catalytic of is this with the using the of E. Y. D. R. The and enzyme PubMed Scopus Google Scholar), the of and to be This a in the that be the that this the to with and the to absolute enzyme concentrations with high and in this the workflow and the labeled standards developed in this absolute enzyme concentrations can be for E. coli strain in and catalytic can be from the of the wild-type and strains, information the catalytic of an enzyme can be used to two levels of of metabolic enzyme concentrations and the enzymes The can be used to metabolic engineering for to is the of a for a specific This the use of an absolute and based quantification workflow for to the of E. coli central The is based combined with using Reaction have that be and accurately with high and specificity using R. of 2009; PubMed Scopus Google Scholar) or standards quantification of the in of a PubMed Scopus Google Scholar) to analysis. quantification of the in of a PubMed Scopus Google Scholar) the absolute quantification of enzymes of the using In the we an absolute quantification that the for enzyme concentrations measurements. in to accurate to be the of the and of the In the context of and standards used for absolute quantification a has been developed with the to absolute quantification for of R. R. absolute quantification of a in the of a PubMed Scopus Google Scholar). However, as is the or the such that the of is this is a to quantitative using to as to ensure accurate measurement of labeled standards were as the of standards and accurate to use full-length labeled standards to enzyme concentrations as accurately as model (E. to full-length isotopically labeled standards in E. coli to high production The was to a and production a to be to and production of selected a and were a single of a of PubMed Scopus Google Scholar). then the precursor mass with a standards were using an has the of of and a and has to M. Y. J. M. a and for acid using and mass PubMed Scopus Google Scholar). This workflow can be used for and production of of full-length isotopically labeled and is for This is to E. coli and can be for of a of PubMed Scopus Google Scholar) developed a 15N labeled production to quantify M. and quantification of DNA liquid using fully as PubMed Scopus Google Scholar) produced to accurately quantify DNA has that an can quantify over a and cell R. of 2009; PubMed Scopus Google Scholar). to Y. J. M. D. abundance of the Escherichia coli PubMed Scopus Google Scholar), the in E. coli strains is of The and capabilities this the for of the was to an workflow and accurate quantification the of for E. used a to with This to accurately quantify 22 monitoring a single 30-min analysis. and and the be used for accurate quantification of to The performance of workflow was using and a high and The to that with full-length be to and and be to accurately quantify in a single J. J. M. a for Escherichia PubMed Scopus Google Scholar) metabolism in E. coli to be This that accurately quantify of E. coli enzymes in a single 30-min analysis. the quantification workflow to strains with production of an involved in this the abundance of enzymes M. Recent advances in engineering the central metabolism of Cell Fact. PubMed Scopus Google Scholar). we the developed J. of Escherichia coli to in PubMed Scopus Google Scholar). In strain NADPH was the of of the and to for the of the two in two different to strain J. of Escherichia coli to in PubMed Scopus Google Scholar). The of in the a of the to NADPH The in the for the of a the of to the and the J. of Escherichia coli to in PubMed Scopus Google Scholar). in the of and were the the was to enzymes were of the NADPH production This in an of the to to the of M. Recent advances in engineering the central metabolism of Cell Fact. PubMed Scopus Google Scholar). can the accurate quantitative proteomics data with the in this be for the of the The absolute in interesting in as to enzymes the cell However, the data more when with of have combined the proteomics data with from the conditions as in this This to analyze the of the and the catalytic of enzymes to the the reaction in the was to be of a higher enzyme with a catalytic rate. was for In to be of a catalytic of the with of the This for the enzymes and as as to different of of the Escherichia coli PubMed Scopus Google Scholar) D. M. Escherichia coli and of with using and J. PubMed Scopus Google Scholar). that and were in the catalytic of with the of and as as the of In the from of the was higher in This to be of an catalytic as the enzyme in The the of for in catalytic in models from and enzymes catalytic were used for the the and the of the cell Chang models of metabolism and and PubMed Scopus Google R. Escherichia coli catalytic and of PubMed Scopus Google Scholar). the of catalytic of to in a In addition, this information may be of in the context of metabolic as enzymes to and may be in the in the the is the the of was to be to a for interesting as a be of the of this for of enzymes in and may therefore to a of the in R. in may in measurement of the in J. PubMed Scopus Google Scholar). more of metabolic is a and the of for is to the R. The of Google Scholar). be interesting to accurate quantitative proteomics data predictive models of the in metabolic and models including of enzyme with quantitative proteomics and as of E. quantitative proteomics and with a genome-scale metabolic PubMed Scopus Google Scholar), the of the metabolic quantitative information abundance may be useful when combined with and data in the of predictive kinetic the of the of the is the E. E. J. D. D. model of based a kinetic of PubMed Scopus Google Scholar), combined proteomics and data with enzyme kinetic assays to in for and when a kinetic to the of the of more over the is the data of in enzyme and concentrations as as and levels in of E. coli M. M. Ho Y. M. Y. Y. Y. M. the of E. coli to PubMed Scopus Google Scholar). data have been used to the in a model of the metabolic a of for model in genome-scale kinetic such as and the of M. D. E. of metabolic models from PubMed Scopus Google Scholar). of the is to the that were when catalytic a the of enzymes be as from the of However, in E. a is more a single the 22 the involved two and of was In to a of the in of enzyme abundance and catalytic to a in the of have to be In to the of in assays can be as E. E. J. D. D. model of based a kinetic of PubMed Scopus Google Scholar). In in this we the development of a proteomics workflow absolute quantification of enzymes involved in E. coli the data in an catalytic of the enzymes for wild-type and strains optimized for higher NADPH production. This provides the of enzyme concentrations and catalytic to different strains as as potential in the in the the kinetic models of metabolism that can be used as predictive in and in biotechnology. and for to to for with to for with to the for and when producing the to for with