Elsevier

Metabolic Engineering

Volume 25, September 2014, Pages 8-19
Metabolic Engineering

A novel platform for automated high-throughput fluxome profiling of metabolic variants

https://doi.org/10.1016/j.ymben.2014.06.001Get rights and content

Highlights

  • Integrated solution for fluxome profiling of large sets of biological systems and conditions.

  • Suite of high-throughput compatible methods from cell cultivation to flux calculation.

  • Discrimination of metabolic variants on the basis of their metabolic fluxes.

  • In depth understanding of biological processes.

  • Exceeding existing methods in terms of throughput, robustness, release of resources and screening capacity.

Abstract

Advances in metabolic engineering are enabling the creation of a large number of cell factories. However, high-throughput platforms do not yet exist for rapidly analyzing the metabolic network of the engineered cells. To fill the gap, we developed an integrated solution for fluxome profiling of large sets of biological systems and conditions. This platform combines a robotic system for 13C-labelling experiments and sampling of labelled material with NMR-based isotopic fingerprinting and automated data interpretation. As a proof-of-concept, this workflow was applied to discriminate between Escherichia coli mutants with gradual expression of the glucose-6-phosphate dehydrogenase. Metabolic variants were clearly discriminated while pathways that support metabolic flexibility towards modulation of a single enzyme were elucidating. By directly connecting the data flow between cell cultivation and flux quantification, considerable advances in throughput, robustness, release of resources and screening capacity were achieved. This will undoubtedly facilitate the development of efficient cell factories.

Introduction

In the field of biotechnology, metabolic engineering and synthetic biology, fluxomics has been identified as a key analytical technology (Ellis and Goodacre, 2012, Feng et al., 2010, Sanford et al., 2002, Stephanopoulos, 1999) not only for the rational design of cells but also for comprehensive understanding of the link between genotype and phenotype. Fluxomics, i.e. the cell-wide quantification of metabolic fluxes, reveals the actual operation of metabolic networks in given environmental conditions, resulting from the integrated flow of interactions between all molecular components - genes, mRNAs, proteins and metabolites (Sauer, 2004, 2006). Hence, metabolic flux analysis is the most comprehensive description of the metabolic phenotype at the cellular level (Wittmann and Portais, 2013). In practice, 13C-fluxomics is still a tedious and rather time consuming process. The method combines both 13C-labelling experiments and mathematical modelling of biochemical networks. Cells are gown on 13C-labelled substrates to metabolic and isotopic steady state and the labelling patterns of metabolites are monitored by mass spectrometry (MS), nuclear magnetic resonance (NMR), or both. The labelling information can be collected on metabolic end-products such as protein-bound amino acids, which accumulate to larger extents than true metabolic intermediates (Sauer, 2004, Szyperski, 1995, Wiechert, 2001). Combined with quantitative physiological data and a detailed metabolic model of metabolism, the labelling patterns give access to the in vivo reaction rates (i.e. fluxes) associated with the cellular network of an organism. When this information is not available, multivariate statistics can be applied to the isotopic data to provide detailed phenotyping of biological systems without any prior knowledge (Raghevendran et al., 2004, Zamboni and Sauer, 2004).

The rapid expansion of fluxomics and its areas of application is driving the need for high-throughput approaches to enable comprehensive metabolic investigations of a growing number of organisms, engineered mutants and physiological conditions (Ellis and Goodacre, 2012). Significant advances have been made in the field in the last decade. Miniaturized cell cultivation systems have been used to perform 13C-labelling experiments with less effort and cost in labelled substrate (Balcarcel and Clark, 2003, Betts and Baganz, 2006, Ge et al., 2006, Girard et al., 2001, Huber et al., 2009, Isett et al., 2007, Kocincova et al., 2008, Tang et al., 2009). The sensitivity, speed, and robustness of both NMR-based and MS-based isotopic analysis have been improved (Boisseau et al., 2013, Cahoreau et al., 2012, Fan and Lane, 2008, Fischer et al., 2004, Giraudeau et al., 2012, Giraudeau et al., 2011, Massou et al., 2007a, Peng, 2012). Tools have been developed for large scale isotopic data processing (Millard et al., 2012, Poskar et al., 2012) as along with improved algorithms and software for flux calculation and statistical analysis of isotopic data (Antoniewicz et al., 2007, Quek et al., 2009, Raghevendran et al., 2004, Sokol et al., 2012, Weitzel et al., 2012, Zamboni et al., 2005, Zamboni and Sauer, 2004). These developments have allowed large-scale flux analysis to be applied to microorganisms (Amador-Noguez et al., 2010, Blank et al., 2005, Fischer and Sauer, 2005, Fischer et al., 2004, Haverkorn van Rijsewijk et al., 2011, Wittmann et al., 2004) and mammalian cells (Munger et al., 2008).

However, despite all the above improvements, HT fluxomics is still in early development. In particular, there is no existing HT fluxomics platform which combines all experimental and in silico steps of the complete workflow in a single and fully integrated manner. The design and development of such a platform is hindered by several technical challenges that need to be addressed in parallel. This includes (1) the tight control of cultivation parameters to ensure metabolic and isotopic steady state; (2) the automated monitoring of growth parameters to collect labelled material once steady state is achieved (3) parallel cultivation to increase throughput; (4) miniaturization of working volumes to reduce the cost of labelled substrates and to facilitate the parallel processes; (5) automated, rapid, parallel sampling of labeled material to avoid degradation of the metabolites; (6) rapid, sensitive measurement of isotopic profiles to cope with both the small amounts of biological material and the large number of samples; (7) automated extraction of labeling information from raw analytical data, which today is still mainly done manually, thereby saving time and effort and improving data robustness (8), data interpretation tools that provide valuable metabolic information in a high throughput manner, i.e. with reduced user supervision.

For the first time, these challenges were met with the development of a fully integrated solution for fluxome analysis that combines a robotic cultivation and sampling workstation for 13C-labelling experiments with NMR-based isotopic profiling, and tools for processing and interpreting isotopic data. The automation, parallelization, optimization and integration of all steps in the workflow are described in detail. As a proof of concept, the new platform was applied to a set of Escherichia coli mutants with varying levels of a single enzyme and grown on two different 13C-labelled carbon sources. The power of the overall approach to provide discriminating metabolic information for a large number of mutants and conditions as well as its value in generating valuable metabolic knowledge was demonstrated.

Section snippets

Materials and methods

Fig. 1 is a schematic diagram of the complete workflow developed in this study.

Development of a robotic cultivation and sampling system

A workstation enabling (i) the automated parallel cultivation of bacteria with 13C-labelled substrates and (ii) the automated sampling of labelled metabolites under steady state conditions was designed and constructed. Cell cultivation was automated, parallelized and miniaturized by incorporating a block of 48 micro-scale bioreactors (Kusterer et al., 2008, Puskeiler et al., 2005) in the automatic workstation. This system allows fully automated and reproducible monitoring of growth, pH and

Conclusion

The integrated fluxome platform presented here is a first but critical step towards robotic screening of quantitative metabolic phenotypes. Due to the parallelization and automation of the workflow, significant improvements in throughput, robustness and release of resources were achieved. The 80 flux data collected on E. coli G6PDH-modulation mutants were generated in four days, excluding preparation time but including all automated and manual steps from cultivation to calculated fluxes. This

Acknowledgments

MetaToul (Metabolomics & Fluxomics Facitilies, Toulouse, France, www.metatoul.fr) and its staff members Edern Cahoreau and Lindsay Peyriga are gratefully acknowledged for technical support and access to NMR. MetaToul is part of the national infrastructure MetaboHUB (The French National infrastructure for metabolomics and fluxomics, www.metabohub.fr, MetaboHUB-ANR-11-INBS-0010). MetaToul is supported by the Région Midi-Pyrénées, the European Regional Development Fund, the SICOVAL, the

References (74)

  • S. Massou et al.

    Application of 2D-TOCSY NMR to the measurement of specific 13C-enrichments in complex mixtures of 13C-labeled metabolites

    Metab. Eng.

    (2007)
  • S. Massou et al.

    NMR-based fluxomics: quantitative 2D NMR methods for isotopomers analysis

    Phytochemistry

    (2007)
  • D. Mueller et al.

    Stable isotope-assisted metabolomics to detect metabolic flux changes in mammalian cell cultures

    Curr. Opin. Biotechnol.

    (2013)
  • C. Nicolas et al.

    Response of the central metabolism of Escherichia coli to modified expression of the gene encoding the glucose-6-phosphate dehydrogenase

    FEBS Lett.

    (2007)
  • K. Sanford et al.

    Genomics to fluxomics and physiomics—pathway engineering

    Curr. Opin. Microbiol.

    (2002)
  • U. Sauer

    High-throughput phenomics: experimental methods for mapping fluxomes

    Curr. Opin. Biotechnol.

    (2004)
  • U. Sauer et al.

    The soluble and membrane-bound transhydrogenases UdhA and PntAB have divergent functions in NADPH metabolism of Escherichia coli

    J. Biol. Chem.

    (2004)
  • G. Stephanopoulos

    Metabolic fluxes and metabolic engineering

    Metab. Eng.

    (1999)
  • W. Wiechert

    13C metabolic flux analysis

    Metab. Eng.

    (2001)
  • J. Zhao et al.

    Effect of zwf gene knockout on the metabolism of Escherichia coli grown on glucose or acetate

    Metab. Eng.

    (2004)
  • D. Amador-Noguez et al.

    Systems-level metabolic flux profiling elucidates a complete, bifurcated tricarboxylic acid cycle in Clostridium acetobutylicum

    J. Bacteriol.

    (2010)
  • R.R. Balcarcel et al.

    Metabolic screening of mammalian cell cultures using well-plates

    Biotechnol. Progr.

    (2003)
  • O. Beckonert et al.

    Metabolic profiling, metabolomic and metabonomic procedures for NMR spectroscopy of urine, plasma, serum and tissue extracts

    Nat. Protoc

    (2007)
  • J.I. Betts et al.

    Miniature bioreactors: current practices and future opportunities

    Microb. Cell Fact.

    (2006)
  • L.M. Blank et al.

    Large-scale 13C-flux analysis reveals mechanistic principles of metabolic network robustness to null mutations in yeast

    Genome Biol.

    (2005)
  • R. Boisseau et al.

    Fast spatially encoded 3D NMR strategies for C-based metabolic flux analysis

    Anal. Chem

    (2013)
  • X. Feng et al.

    Bridging the gap between fluxomics and industrial biotechnology

    J. Biomed. Biotechnol.

    (2010)
  • E. Fischer et al.

    Metabolic flux profiling of Escherichia coli mutants in central carbon metabolism using GC–MS

    Eur. J. Biochem.

    (2003)
  • E. Fischer et al.

    Large-scale in vivo flux analysis shows rigidity and suboptimal performance of Bacillus subtilis metabolism

    Nat. Genet.

    (2005)
  • P. Giraudeau et al.

    UFJCOSY: a fast 3D NMR method for measuring isotopic enrichments in complex samples

    ChemPhysChema European journal of chemical physics and physical chemistry

    (2012)
  • P. Giraudeau et al.

    Ultrafast quantitative 2D NMR: an efficient tool for the measurement of specific isotopic enrichments in complex biological mixtures

    Anal. Chem.

    (2011)
  • B.A. Hanson

    ChemoSpec: Exploratory Chemometrics for Spectroscopy

    R Package Version

    (2012)
  • B.R. Haverkorn van Rijsewijk et al.

    Large-scale 13C-flux analysis reveals distinct transcriptional control of respiratory and fermentative metabolism in Escherichia coli

    Mol. Syst. Biol.

    (2011)
  • S. Heux et al.

    High-throughput workflow for monitoring and mining bioprocess data and its application to inferring the physiological response of Escherichia coli to perturbations

    Appl. Environ. Microbiol.

    (2011)
  • R. Hortsch et al.

    Process performance of parallel bioreactors for batch cultivation of Streptomyces tendae

    Bioprocess. Biosyst. Eng.

    (2011)
  • Q. Hua et al.

    Responses of the central metabolism in Escherichia coli to phosphoglucose isomerase and glucose-6-phosphate dehydrogenase knockouts

    J. Bacteriol.

    (2003)
  • R. Huber et al.

    Robo-Lector—a novel platform for automated high-throughput cultivations in microtiter plates with high information content

    Microb. Cell. Fact.

    (2009)
  • Cited by (30)

    • Metabolomics and modelling approaches for systems metabolic engineering

      2022, Metabolic Engineering Communications
      Citation Excerpt :

      The use of liquid handling technologies also allows for good reproducibility of samples extracted and is therefore vital for high throughput sample preparation (Liu et al., 2019a). Cultivation devices such as microtiter plates or micro-bioreactors have also been integrated to robotic cultivation platforms, enabling growth parameters for cell cultures to be fully automated and controlled (Long et al., 2014; Faust et al., 2014; Unthan et al., 2015; Heux et al., 2014; Bergès et al., 2021). Microfluidic platforms allow for enhanced analysis speed, diminished sample volumes, and improved multiplexing capabilities (Wang et al., 2015).

    • Classification and biomarker identification of prostate tissue from TRAMP mice with hyperpolarized <sup>13</sup>C-SIRA

      2021, Talanta
      Citation Excerpt :

      Biological interpretations are supported by the isotopic patterns contributing to group discrimination [6]. In particular, isotopic profiling studies take advantage of unsupervised, multivariate statistical classification (such as principal component analyses) to discriminate strains or conditions based on 13C-labeling data, without a priori knowledge of the metabolic system [7,8]. Metabolic activity can be quantitatively studied with nuclear magnetic resonance (NMR).

    • Recent advances in high-throughput <sup>13</sup>C-fluxomics

      2017, Current Opinion in Biotechnology
      Citation Excerpt :

      Significant progress in the development of HT fluxomics made in the last few years has contributed to deepening our understanding of cellular metabolism. As illustrative examples, HT fluxomics has allowed to identify transcription factors that control metabolic fluxes in yeast under a given condition [17•], to reveal the conservation of the Entner–Doudoroff pathway as a common glycolytic strategy among glucose-using marine bacteria [8], or to elucidate the control exerted by key enzymes on metabolic fluxes [9••]. Being a very complex process requiring many heterogeneous tasks and sub-tasks to be performed and iterated, improvements still remain to be made to automatize the complete workflow, to increase the number of investigations, and to improve the quality of flux measurements.

    View all citing articles on Scopus
    View full text