Application of Isotope Labeling Experiments and 13C Flux Analysis
Total Page:16
File Type:pdf, Size:1020Kb
Available online at www.sciencedirect.com ScienceDirect 13 Application of isotope labeling experiments and C flux analysis to enable rational pathway engineering 1 1 1,2 Allison G McAtee , Lara J Jazmin and Jamey D Young 13 Isotope labeling experiments (ILEs) and C flux analysis industries to date, despite the fact that these approaches provide actionable information for metabolic engineers to can provide direct readouts on in vivo metabolic pathway identify knockout, overexpression, and/or media optimization activities. This may be partly due to the fact that many targets. ILEs have been used in both academic and industrial companies lack the combined experimental and compu- labs to increase product formation, discover novel metabolic tational expertise needed to effectively analyze ILEs, but functions in previously uncharacterized organisms, and perhaps even more important is the perception that these enhance the metabolic efficiency of host cell factories. This studies are intrinsically difficult and there have not been review highlights specific examples of how ILEs have been enough success stories to justify the requisite effort. used in conjunction with enzyme or metabolic engineering to Therefore, the purpose of this review is to present some elucidate host cell metabolism and improve product titer, rate, recent examples where ILEs and MFA have been suc- or yield in a directed manner. We discuss recent progress and cessfully applied to close the ‘design-build-test-learn’ 13 future opportunities involving the use of ILEs and C flux metabolic engineering cycle. analysis to characterize non-model host organisms and to identify and subsequently eliminate wasteful byproduct Applications of isotope labeling experiments pathways or metabolic bottlenecks. in metabolic engineering Addresses 1 ILEs allow strain and process engineers to peek inside the Department of Chemical and Biomolecular Engineering, Vanderbilt black box of host cell metabolism by tracing the progres- University, PMB 351604, Nashville, TN 37235-1604, USA 2 Department of Molecular Physiology and Biophysics, Vanderbilt sion and rearrangement of isotopically labeled substrate University, PMB 351604, Nashville, TN 37235-1604, USA atoms as they are metabolized through intermediary 13 biochemical pathways [5–7]. For example, C-labeled Corresponding author: Young, Jamey D ([email protected]) glucose tracers are commonly used to study central carbon 13 metabolism in heterotrophic microbes, while C-labeled Current Opinion in Biotechnology 2015, 36:50–56 amino acids and organic acids have been used extensively to assess citric acid cycle flux in mammalian cells. While This review comes from a themed issue on Pathway ILEs can provide useful information about metabolic Edited by William E Bentley and Michael J Betenbaugh rates even in the absence of quantitative flux analysis [8 ], several flux modeling approaches have been devel- oped to extract additional information from the raw http://dx.doi.org/10.1016/j.copbio.2015.08.004 isotope labeling data: MFA [9,10,11 ,12], isotopically 0958-1669/# Elsevier Ltd. All rights reserved. nonstationary MFA (INST-MFA) [13–15], metabolic flux ratio (METAFoR) analysis [16], and kinetic flux profiling (KFP) [17]. Both KFP and INST-MFA compute fluxes based on transient (rather than steady-state) isotope la- beling measurements, which reduces the duration of ILEs and provides increased flux resolution in most cases. Introduction METAFoR analysis and KFP rely on targeted isotope Mammalian, plant, yeast, and bacterial cells are used as labeling measurements to assess the local flux distribution industrial hosts for the production of commodity chemi- at specific network nodes. In contrast, MFA and INST- cals, specialty chemicals, small-molecule drugs, therapeu- MFA are able to integrate multiple independent mea- tic proteins, and other biomolecules of commercial surements into a comprehensive flux map while also interest [1 ]. Because these processes rely on living cells identifying inconsistent data that would not be readily as biocatalysts, they are often hindered by toxic byprod- detected through a localized analysis. The quantitative uct formation, low product yield, and slow production flux estimates produced by these modeling approaches rates. Genome-scale modeling, cell-wide ‘omics’ plat- have direct biological meaning and can be compared forms, and high-throughput screening approaches have across independent experiments. Flux information been developed to overcome these challenges by identi- obtained from ILEs has been effectively applied to (1) fying genes that can be engineered to improve host cell characterize new host organisms, (2) identify wasteful performance. However, isotope labeling experiments pathways that limit product yield, and (3) identify (ILEs) and metabolic flux analysis (MFA) [2–4] have metabolic bottlenecks that restrict production rate received limited attention in the biotech and biopharma (Figure 1). By quantifying fluxes at each major node of Current Opinion in Biotechnology 2015, 36:50–56 www.sciencedirect.com Isotope labeling experiments for pathway engineering McAtee, Jazmin and Young 51 Figure 1 Characterizing non-model hosts Isotope Tracer Identifying wasteful byproduct pathways Source Wasteful Desired Byproduct Product Actionable Pathway Engineering Results Identifying pathway bottlenecks Metabolic Network Source Product Current Opinion in Biotechnology Examples of how ILEs can lead to actionable results for metabolic pathway engineering. ILEs can elucidate genetic targets for improving cell metabolism by enabling researchers to (1) characterize the metabolism of non-model organisms, (2) identify wasteful processes that lead to yield losses, and (3) identify kinetic bottlenecks in metabolic networks that limit production rate. the metabolic network and determining how these fluxes or by NMR spectroscopy. By collecting a time series of become re-routed in response to targeted genetic or samples following tracer administration, it is possible to environmental perturbations, fundamental insights about obtain progressive snapshots of isotope labeling over network regulation can be obtained to guide further time, effectively mapping the progression of the tracer rounds of metabolic engineering. atoms through the unknown metabolic network. In this way, the intermediary pathways are elucidated, which Characterizing non-model host organisms may offer novel targets for improving host metabolism. Model host organisms such as Escherichia coli, yeast, and CHO cells have been thoroughly studied and are com- The scientific literature provides several illustrative monly used as industrial hosts [18,19]. However, non- examples where non-model hosts have been successfully 13 model host organisms often possess unique metabolic profiled by ILEs. For example, Swarup et al. [21] used C capabilities that make them better suited for production MFA to reconstruct the metabolic network of Thermus of specific compounds, especially biofuels and specialty thermophilus HB8, a potential host for biotech applications chemicals, although the genetic tools required for meta- due to its ability to thrive at high temperatures, and to 13 bolic engineering may still be underdeveloped. Another characterize its growth phenotype. Recently, C MFA potential limitation is that the biochemical pathways of was used to explore the largely unknown metabolic these non-model hosts are often incompletely under- phenotypes of 25 marine microbes that use glucose from stood. In this case, ILEs can be used to obtain missing seawater as their main carbon source [22 ]. Results from biochemical pathway information by tracing how labeled the ILEs showed that the majority of the marine bacteria substrates are broken down and rearranged as they tra- utilized a completely different glucose catabolic pathway verse the metabolic network. For example, non-targeted than their terrestrial counterparts, which was shown to tracer fate detection (NTFD) [20] can locate metabolite provide resistance against oxidative stress typically found peaks within a chromatographic trace that exhibit a mass in the marine environment. There has also been in- shift when cells are grown in the presence of an isotopic creased interest in the use of photosynthetic or methano- tracer. These peaks represent downstream intermediates trophic organisms as production hosts because of their or end-products derived from the tracer substrate, which advantageous ability to use CO2 or natural gas, respec- can be later identified by comparison to standard libraries tively, as their sole carbon source. Cyanobacteria and www.sciencedirect.com Current Opinion in Biotechnology 2015, 36:50–56 52 Pathway algae have been thoroughly probed and characterized E. coli strains that produced 39% of the theoretical through the application of ILEs under heterotrophic, maximum yield of C14–16 fatty acids, as opposed to the mixotrophic, and autotrophic conditions [23–29]. The wild-type species that produced only 11% of the theo- metabolic networks of methanol-consuming Corynebacte- retical yield. Similarly, when the newly discovered bacte- rium glutamicum [30] and Pichia pastoris [31–34] have also rial host Basfia succiniciproducens [48] was determined to been established through the utilization of ILEs. Com- have a natural affinity for bio-succinate production, Becker prehensively profiling the metabolic capabilities of these et al. [49] explored its intracellular metabolism using sys- 13 hosts has paved the way for rational