Determination of intracellular metabolites concentrations in Escherichia coli under nutrition stress using liquid chromatography-tandem mass spectrometry

Talanta 189 (2018) 1–7

Contents lists available at ScienceDirect

Talanta journal homepage: www.elsevier.com/locate/talanta

Determination of intracellular metabolites concentrations in Escherichia coli under nutrition stress using liquid chromatography-tandem mass spectrometry Fenfen Jia,1, Yang Shena,1, Leihan Tangb, Zongwei Caia, a b

T

⁎

State Key Laboratory of Environmental and Biological Analysis, Department of Chemistry, Hong Kong Baptist University, Hong Kong SAR Department of Physics, Hong Kong Baptist University, Hong Kong SAR

A R T I C LE I N FO

A B S T R A C T

Keywords: Escherichia coli Nutrition stress Central carbon metabolism TCA intermediate

Cells show a timely and appropriate physiological adjustment on all levels of cellular activities in response to nutrient stress. However, the regulations for cells under different carbon/nitrogen influxes are poorly understood. To unveil a fully metabolic regulatory profile, we applied a mass spectrometry based “bottom-up” approach to investigate the metabolic response of Escherichia coli to nutrient stress. A novel cell sample preparation procedure was developed to decrease the variation and leakage of intracellular metabolites. Volatile ion-pair reagent tributylamine was used to improve the retention and selectivity of charged metabolites on a C18 reversed-phase column. The growth rate and intracellular concentrations of 12 central carbon metabolites were measured systematically under various carbon/nitrogen influxes by manipulating titratable promoters. Fructose1,6-biphosphate (FBP) concentration as a sensor of carbon influx was positively correlated with the growth rate, whereas α-ketoglutarate (αkg), served as a coordinator of carbon and nitrogen flux showed different dependence on growth rate between carbon limitation and nitrogen limitation. By integrating different behaviors of the metabolites with knowledge from previous reports, a scenario of feedback control under carbon and nitrogen limitations was proposed. Our findings revealed the key role of αkg in the coordination of carbon and nitrogen utilization under nutrition stress and highlighted the great potential of mass spectrometry based approach in deciphering the complex metabolic network.

1. Introduction Cells survive under different kinds of nutrient stress by adjusting their cellular metabolism. The growth of a cell would come to a halt under a variety of stress conditions, in particular, the depletion of a critical nutrient in the medium [1]. However, the timing of regulatory steps including shut down growth, adjust the influx and the way cells coordinate its protein composition to enter into a different growth mode or the stationary phase, are poorly understood up to now. Thus, characterization of the cellular response of microbial systems to environmental perturbations especially nutrient stress has attracted increasing interests in recent years. Many efforts have been made to investigate the correlation between the type of stress and the response in Escherichia coli (E. coli) at the gene expression level [2–4]. However, transcriptional regulation alone is not sufficient to completely determine the metabolite flow pattern on the complex metabolic network. It is known that metabolites and proteins can actually regulate

⁎

1

Corresponding author. E-mail address: [email protected] (Z. Cai). These authors contributed equally to this work.

https://doi.org/10.1016/j.talanta.2018.06.050 Received 6 April 2018; Received in revised form 7 June 2018; Accepted 13 June 2018 Available online 18 June 2018 0039-9140/ © 2018 Elsevier B.V. All rights reserved.

themselves to some extent and many downstream mechanisms can play important roles in cellular responses through self-adjustment. For example, allosteric regulation by small molecules can change the conformation of proteins so that rapid regulation can be achieved, with a much shorter time scale as compared to transcriptional regulation [5]. It is thus important to take into account the changes in metabolite concentrations in downstream in addition to gene and protein expression changes in order to obtain a more complete picture of cellular stress response. On the other hand, central carbon metabolism, as the main frame of metabolic network is considered crucial in the investigation on the regulation response to the environment in model organism E. coli [6]. The intracellular concentrations of the relevant metabolites carry information that can lead to quantitative understanding of in vivo enzyme kinetic parameters or control of metabolic flux with the aid of mathematical tools and theoretical models [7]. For the determination of central carbon metabolites in E. coli, liquid chromatogram-mass

Talanta 189 (2018) 1–7

F. Ji et al.

respectively. The glycerol minimal medium had 0.2% (w/v) glycerol added to the MOPS base medium. For the C-limitation growth, 1 mM IPTG and various concentrations (0–500 µM) of the inducer 3Methylbenzyl alcohol (3MBA) were added to the lactose minimal medium or glycerol minimal medium. For the N-limitation growth, concentrations of IPTG in the range of 30–100 µM and 25–75 µM were added to the glucose minimal medium and glycerol minimal medium, respectively. All batch culture growth was performed in a 37 °C water bath shaker shaking at 250 rpm. The culture volume was 10 ml in 25 mm × 150 mm test tubes. Each growth experiment was carried out in three steps: “seed culture” in LB broth, “pre-culture” and “experimental culture” in identical minimal medium. For seed culture, one colony from fresh LB agar plate was inoculated into liquid LB and cultured at 37 °C with shaking. After 4–5 h, cells were transferred into the minimal medium and cultured overnight (pre-culture), which allowed cells to grow for at least 3 doublings. Pre-culture cells was then diluted to OD600 = 0.005–0.025 in identical pre-warmed minimal medium, and cultured in 37 °C water bath shaker (experimental culture). 200 μL of cell culture was collected in a Starna Sub-Micro Cuvette (Starna Cells, Atascadero, CA) for OD600 measurement using a Thermo GENESYSTM 20 Spectrophotometer around every half doubling of growth. About 5–7 OD600 data points within the range of ~0.05 and ~0.5 (above OD600 =~0.6 the spectrophotometer was determined to be slightly nonlinear) were used for calculating growth rate.

spectrometry (LC-MS) based methods have been applied for most studies, with different sample preparation procedures [8–15]. Mass spectrometry (MS) is preferred for the aims of high throughput, resolution, and sensitivity to detect central carbon metabolites because most of them have no or weak UV absorption or florescence [16]. Most of the MS methods, however, focused on one step of sample preparation with miscellaneous results [17–19]. Therefore, evaluation and optimization of the comprehensive sample preparation procedures for central carbon metabolism is necessary to retain the original cell state. In this work, LC-MS based “bottom-up” approach was employed to investigate the cellular response of E. coli to nutrient stress. Firstly, different strains of E. coli were constructed for titrating carbon/nitrogen (C/N) sources. A novel and efficient sample preparation method with the subsequent LC-MS/MS analysis for quantifying 12 central carbon metabolites in E. coli was established. Then the growth rate and concentrations of the metabolites were monitored under various carbon or nitrogen influxes. Based on the metabolites changes, a scenario of feedback control coordinating the carbon and nitrogen utilization by αketoglutarate (αkg) was proposed. 2. Materials and methods 2.1. Strains construction for carbon/nitrogen source titration Construction of titratable lacY (NQ381) and titratable glpFK (NQ399) strains for C source titration: DNA fragment containing the Pu promoter (−1 bp to −178 bp relative to the transcriptional start site) was amplified by PCR from a Pu promoter containing plasmid pEZ9, then inserted into the SalI and BamHI sites of plasmid pKD13, producing plasmid pKDPu. Using this plasmid as a template, the region containing the km gene and Pu promoter was PCR amplified and integrated into the chromosome of E. coli strain NQ351 between the lacZ and lacY (from lacZ stop codon to lacY start codon), and in front of glpF (−1 bp to −252 bp relative to the translational start point of glpF) respectively, by using the λ Red system [20]. As the activation of Pu promoter needs the XylR protein, a strain NQ386 was constructed in which a synthetic lac promoter PLlac-O1 [21] (a promoter that is repressed by LacI but does not need Crp-cAMP for activation) driving xylR (xylR gene was cloned from pEZ6 [22]) was inserted at the attB site. The km-Pu-lacY and km-Pu-glpFK constructs in NQ351 were transferred into strain NQ386 containing PLlac-O1-xylR by P1 transduction, resulting in strains NQ381 and NQ399, respectively. Construction of titratable GOGAT strain (NQ393) for N source titration: Using the λ Red system [20], the promoter (+123 bp to −176 bp) of gltBDF operon was replaced by the synthetic lac promoter PLlac-O1 [21] (a promoter that is repressed by LacI but does not need Crp-cAMP for activation) together with selection maker Km gene. The resulting Km-PLlac-O1-gltBDF construct was transferred to strain NCM3722 by P1 transduction [23]. The Km gene was then eliminated by using plasmid PCP20 [24]. A sp-lacIQ-tetR cassette providing constitutive expression of lacI to tightly repress PLlac-O1 activity was inserted at the attB site by P1 transduction. Lactose permease encoded by lacY can concentrate intracellular Isopropyl β-D-1-thiogalactopyranoside (IPTG) and will narrow the titration range, we inactivated lacY by P1 transduction using strain JW0334-1from CGSC (E. coli Genetic Stock Center, Yale University) as lacY donor following by Km gene elimination. The gdhA gene was knocked out by P1 transduction using strain JW1750-2 from CGSC as gdhA donor following by Km gene elimination to obtain the final strain NQ393.

2.3. Sample preparation for quantification of intracellular metabolites The E. coli cell samples were collected during the exponential growth. The filter membrane (25 mm diameter with 0.45 µm pore size, Millipore) was pre-rinsed by pre-warmed Milli-Q water and culture medium, then 10 ml of culture with known OD was quickly filtrated under vacuum. 2 ml of wash medium was pipetted onto the membrane twice. Then the membrane was immediately immersed in 3 ml of acidic methanol/acetonitrile/water mixture solution (40:40:20 in volume, with 0.1 M formic acid) in a plastic tube. These two steps were finished within 30 s. After 30 s vortexing, the tube was frozen by liquid nitrogen for 1 min and vortexed for another 30 s after its thawing. After centrifugation, the supernatant was taken out and frozen by liquid nitrogen once again. The resulted solutions were then centrifuged through a Millipore Amicon Ultra Filter at 3000 g under 4 °C for 20 min for deproteinization. Finally 2 ml of supernatant was dried and re-dissolved in 200 μL of acidic acetonitrile/methanol/water for LC-MS/MS analysis. 2.4. LC-MS/MS analysis of intracellular metabolites Analysis was performed on a ultra-performance liquid chromatography (UPLC) coupled with a triple-quadrupole mass spectrometry equipped with an ESI source (Waters Corporation, Milford, MA, USA) with a reversed-phase C18 column (2.1 mm × 100 mm, 1.7 µm) running in ion-pair mode. The detailed method was developed based on a previous study with some modifications [25]. Briefly, the mobile phase consisted of two components: A (10 mM tributylamine aqueous solution adjusted pH to 4.95 with 15 mM acetic acid) and B (methanol). Linear gradient elution was as follows: 100% A at 0 mins, 80% A at 5 mins, 5% A at 20 mins, and 100% A at 22 mins. An injection volume of 10 μL was selected with a flow rate of 0.3 ml/min. MS analysis was operated in negative ion multiple reaction monitoring (MRM) mode. The ion transition, collision energy and cone voltage were optimized by relative standards from Sigma to increase the instrument response (Table S1). The optimized MS parameters were described as follows: the capillary voltage was 3000 V; the dwell time was 0.05 s; the extractor voltage was 2.5 V; the temperatures of the negative ESI source and desolvation gas were 118 and 500 °C, respectively; the cone gas and the desolvation gas flows were 40 and 650 L/h, respectively. Instrument operation and data acquisition were processed by using the Waters MassLynx V4.1

2.2. Mediums and growth measurements All growth media used in this study were based on M9 minimal medium and MOPS base medium with slight modifications. The lactose minimal medium and the glucose minimal medium had 0.2% (w/v) lactose and 0.2% (w/v) glucose added to the M9 minimal medium, 2

Talanta 189 (2018) 1–7

F. Ji et al.

3.3. Change of carbon flux revealed by fructose-1,6-biphosphate levels

SCN562 software package. The stock solution of standard mixture was prepared at a concentration of 100 mM in water. Calibration curves were obtained by analyzing calibration standards (Sigma) ranging from 10 µM to 10 mM (0.01, 0.05, 0.1, 0.5, 1, 5 and 10 mM). The concentration of metabolites in cell sample was quantified by external standard method.

FBP is an important intermediate of glycolysis, it acts as an inhibitor of the transcription factor Cra that regulates glycolytic and gluconeogenic genes in E. coli [30–33] (Scheme S1). Recently, Karl et al. suggested that the FBP–Cra interaction was part of a flux sensor, which senses the metabolic flux through glycolysis [34]. The glycolytic flux also reflects the carbon influx because it is the beginning of carbon utilization. FBP also activate the pyruvate kinase, which is enrolled to convert PEP to pyruvate, in order to supply sufficient precursor for TCA cycle or overflow metabolism. As shown in Figs. 2 and 3, a striking linear relationship was observed between the growth rate and intracellular FBP concentration in both carbon and nitrogen limitation conditions. Meanwhile, some differences were found between different carbon sources employed during carbon and nitrogen uptake titration. The concentrations of FBP from glycerol in carbon limitation were higher than that from lactose when they had the same or close growth rate. This may be attributed to the different utilization ways of lactose and glycerol. Glycerol needs several steps in gluconeogenesis pathway to be converted to the intermediate that can enter the glycolysis. When the growth rate increases, the demand of increasing the carbon influx will drive a higher level of metabolic flux of both glycolysis and gluconeogenesis pathway. Because FBP is a flux sensor for both pathways [34], more FBP is needed to tune the gene expression level for gluconeogenesis pathway at the same growth rate when the cell grown in glycerol. The similar phenomenon was also found in the nitrogen limitation. However, when the cells grew in glycerol under nitrogen limitation, the increase of FBP concentration appeared to become slower when the growth rate increased. Thus, the carbon and nitrogen limitation may utilize different regulation mechanisms due to that the behavior of cells grew in glycerol under nitrogen limitation was slightly different with carbon limitation. More data of other metabolites need to be investigated to confirm this hypothesis and to reveal the mechanisms of coordination of metabolism when cell met the carbon and nitrogen limitation.

3. Results and discussion 3.1. Optimization of cell sample preparation method A robust and efficient sample preparation method is crucial for studying intracellular central carbon metabolites due to their rapid and dynamic characteristics. In our study, metabolites we targeted were found to be present in significant quantities in the medium. The false high level caused by medium could reach up to 12.3% for Fructose-1,6biphosphate (FBP), 27.3% for phosphoenolpyruvate (PEP), 36.9% for succinate, 17% for malate and 30.5% for αkg (Table S2). Thus filtration step was necessary in our case to remove the medium interference. We further evaluated different sampling methods based on previous reports [26,27]. Reducing transfer loss and operation time turned out to be the key to preserve the original metabolic state. Thus we chose immediately connected filtration to extraction process, the two steps of which were completed within 30 s. Moreover, leakage of intracellular metabolites during quenching was reported [17]. In our case, we found the highest leakage (percentage of metabolites leak out to the quenching solution which was supposed to be discarded before extraction) can be up to 82.4% for αkg as shown in Table S3. Therefore, a simultaneous quenching and extraction with a cold extraction solution was employed in this study. Four different extraction solutions were compared: acidic acetonitrile/methanol/water (40:40:20 in volume, with 0.1 M formic acid) [28], pure methanol [29], 0.3 M KOH (dissolved in 25% ethanol) [25] and 0.8 M perchloric acid [26]. For KOH and HClO4 as extraction solution, 60 μL of glacial acetic acid or 1 ml of 2.5 M KOH were added respectively to neutralize the solution after frozen-thawing step. Results showed that acidic acetonitrile/methanol/water provides higher recovery for most central carbon metabolites (Fig. 1A). Deproteinization filter was further employed to remove residual proteins interference by cutting at the molecular weight of 3000 Da. The highest increase of metabolites level after applying the protein removal method was 19.6% (FBP) as shown in Fig. 1B. Finally, a fast filtration method coupled with simultaneous quenching and extraction by cold acidic acetonitrile/ methanol/water solution, followed by filter deproteinization was applied for the accurate quantification of the intracellular metabolites from different strains and conditions.

3.4. α-ketoglutarate coordination on the carbon and nitrogen flux Microbes could coordinate the uptake of carbon and nitrogen which were the primary substrates for biomass production, when they survived in numerous nutrition conditions. αkg, an intermediate located in both energy producing TCA cycle and the nitrogen assimilation, serves as a bridge of carbon and nitrogen metabolism [35]. In our study, opposite trends of αkg level were found between the different limitations (Figs. 2B and 3B, bottom): the αkg concentration increased when the growth rate increased under carbon limitation, but decreased under nitrogen limitation. In addition, linear correlations were observed in both two groups. In terms of the value of FBP/αkg, it showed the same trend with the growth rate changes under carbon limitation and opposite results under nitrogen limitation (the FBP levels increase with the growth rate, the αkg levels were linearly decreased with growth rate increased), which confirmed that the cells may use different strategy to carbon and nitrogen limitation. In other words, the carbon and nitrogen limitation may meet the different conditions in the regulations for cells to coordinate the nutrition availabilities. The relationship between growth rate and FBP/αkg from our titration study agreed with the mechanism reported by You et al. and provide more evidence for it that α-ketoacids were the key for catabolites by independent inhibition of adenylate cyclase, which was a crucial enzyme for the global regulator cyclic AMP (cAMP) synthesis (Scheme 1) [36]. Among several α-ketoacids, αkg is the most abundant one [37]. Therefore, it could be physiologically dominant. In our case, during carbon uptake titration, with the carbon influx increasing by titrating, the FBP level raised when sensing the carbon influx. The αkg level was

3.2. LC-MS/MS analysis of intracellular metabolites The separation and detection of central carbon metabolites were challenging because of their common properties such as high polarity and low molecular. In this study, volatile ion-pair reagent tributylamine was used to increase the retention of targeted metabolites on a C18 column without depositing at the ion source. Moreover, the second dimension separation from MS/MS analysis was also crucial for the determination of metabolites that could not be separated chromatographically. Fig. 1C showed the results from LC-ESI-MS/MS analysis of E.coli cell samples. Fig. 1C (1) and C (2) showed the separation of isomeric metabolites (G6P/F6P and GAP/DHAP) that had the same MS/ MS pattern from the ion-pair chromatography (IPLC). On the other hand, CIT/ISOCIT and metabolites that could not be separated chromatographically could be successfully distinguished from MS/MS analysis because they have different specific MRM transitions (Fig. 1C (4) and Table S1).

3

Talanta 189 (2018) 1–7

F. Ji et al.

Fig. 1. LC-MS/MS method development for detecting central carbon metabolites. (A) The effects of four different solvents on targeted metabolites extraction. Bars represent the averages with standard errors from three replicate samples taken from independent cultures and analyzed in duplicate (n = 6). (B) The results of applying protein removal process on decreasing the matrix effect. (C) MRM chromatograms of separated central carbon metabolites from IPLC-ESI-MS analysis: (1) separation of G6P and F6P which has the same MRM transition 259 > 97; (2) separation of GAP and DHAP which has the same MRM transition 169 > 97; (3) separation of ATP, ADP and AMP; and (4) separation of the metabolites that cannot be chromatographically separated with different MRM transitions.

3.5. Different behaviors of glycolytic and TCA intermediates between carbon and nitrogen limitations

also raised as a result of increased carbon influx and activated the nitrogen assimilation to enlarge the nitrogen influx, and to balance the carbon and nitrogen demand. On the other hand, when the cell met nitrogen limitation, with the nitrogen influx increased by titration, there were more demands for increasing the carbon influx to meet the requirement of carbon from more nitrogen. Thus the concentration of αkg was decreased to release the inhibition of cAMP in order to activate the carbon influx. Moreover, the tuning of carbon and nitrogen influx according to nutrition conditions resulted in different rates of amino acids and protein synthesis, which was the reason for different biomass synthesis and growth rates. This scenario revealed the mechanism of regulatory response to the nutrition conditions and origin of physiology changes.

Besides the two key regulators FBP and αkg, an interesting phenomenon was also observed in the other three important intermediates in glycolysis and TCA cycle that the relationship between growth rate and intracellular concentrations was different between carbon and nitrogen limitations (Fig. S1). Specifically, the concentrations of metabolites in the TCA cycle (succinate and malate) was increased linearly with increasing growth rate in carbon limitation, but showed no correlation with growth rate in nitrogen limitation. In contrast, the trend of PEP in glycolysis was opposite, and showed no correlation with growth rate in carbon limitation, however, a positive correlation with growth rate in nitrogen limitation. Thus, the results of the three intermediates 4

Talanta 189 (2018) 1–7

F. Ji et al.

A

B FBP concentration (mM)

Lactose

lacY 3MBA

xylR

Pu

lacY

αkg concentration (mM)

Glycerol

glpFK

3MBA

xylR

Pu

glpFK

Fig. 2. Carbon limitation by titrating the expression of transporter lacY and glpfk to control the lactose and glycerol uptake rate. (A) LacY or glpfk are the only transporter that allows E. coli to grow on lactose or glycerol as the sole carbon source. We inserted the titratable Pu promoter and the expression of Pu is activated by the regulator XylR upon induction by 3-methylbenzyl alcohol (3MBA) (0–500 µM). (B) The intracellular concentrations of FBP and αkg increased linearly with growth rate increasing under carbon limitations.

4. Conclusions

also confirmed that cell utilizes different strategies to coordinate the carbon and nitrogen flux, which also reflect the catabolic and anabolic metabolism, respectively. When carbon limitation was applied and slowed down the growth rate, the demand for carbon catabolic enzymes was raised and the demand for anabolic enzymes was reduced, because the biosynthesis needs to be promoted to save carbon as building block (Scheme 1). Thus, the concentration of precursors (αkg) needed to be decreased with the slowed growth, not only because lower concentration of precursors could tune down the nitrogen influx to match the carbon influx, but also due to more precursors were consumed for building blocks. Moreover, the intermediates in glycolysis (PEP) could keep homeostasis when slowed the growth. On the other hand, when nitrogen limitation was applied and slowed down the growth rate, less demand was required for carbon-catabolic enzymes and more demand was required for those involved in nitrogen assimilation [36]. Therefore, the concentration of αkg increased when growth rate decreased and most part of produced αkg were enrolled the nitrogen assimilation. In order to keep the homeostasis of other TCA intermediates, the concentration of glycolysis metabolite (PEP) decreased when growth rate slowed down. In summary, our results revealed the underlying mechanism that nutrition conditions altered the rate of catabolic and anabolic machinery. If carbon-catabolic machinery exceeded the anabolic machinery, accumulation of αkg inhibited the carbon-catabolic enzymes and tuned the carbon and nitrogen influx. In contrast, if anabolic machinery was in excess, the concentration of αkg fell down and carboncatabolic enzymes were activated. The results observed from carbon and nitrogen limitations and various carbons provided down-stream evidence for the scenario.

In this work, a novel sample preparation method was developed for the accurate measurement of intracellular concentrations of central carbon metabolites in E. coli by LC-MS/MS in ion-pair mode. Under various carbon/nitrogen uptake titrations, several key metabolites were determined to reveal correlations between their concentration and the regulation of metabolic pathways. For instance, the positive correlation between FBP concentration and the growth rate, as well as the different dependence on growth rate showed by αkg in terms of carbon and nitrogen limitations. These results provide supporting evidence for the mechanism of coordination of catabolism and anabolism under various nutrient conditions. To be specific, when sufficient carbon sources were served or nitrogen was limited, carbon influx exceeded anabolic capability and α-ketoacids (mainly αkg) accumulation inhibited cAMP, resulting in turning down of the carbon influx. Conversely, when carbon was limited, α-ketoacids levels fell down and cAMP increased to stimulate carbon-catabolic machinery. Overall, our studies highlight the capability of metabolic intermediates in interpreting the regulatory mechanism and provide evidence for the existence of flux sensing mechanism of E.coli with nutrition stress. However the information of metabolic changes in timedependent manner is lacking. Moreover, the integration of metabolic network knowledge with other omics data into large-scale systematic mathematical models is necessary for better understanding regulatory principles.

5

Talanta 189 (2018) 1–7

A

B

α-ketoglutarate NH3 GDH

α-ketoglutarate

αkg concentration (mM)

NH3

Glutamate GOGAT

FBP concentration (mM)

F. Ji et al.

GS

Glutamine

Fig. 3. Nitrogen limitation by titrating ammonia assimilation. (A) The constructed strain is deleted of gdhA and has the promoter of gltBD (GOGAT), which allows the titration of GOGAT expression under various IPTG levels (30–100 µM). (B) The intracellular concentrations of FBP and αkg showed opposite correlation with growth rate under nitrogen limitations.

[3]

[4] [5] [6]

Scheme 1. Proposed regulation of carbon and nitrogen flux by α-ketoacids and cAMP according to carbon and nitrogen availability.

[7]

[8]

Acknowledgement [9]

The authors would like to thank for the financial support from the Hong Kong Research Grants Council (RGC-CRF C2014-14E, RGC-GRF 12300114 & RGC-GRF 201910) and HKBU Interdisciplinary Research Matching Scheme (RC-IRMS/15-16/04).

[10]

[11]

Conflicts of interest There are no conflicts of interest to declare.

[12]

Appendix A. Supporting information [13]

Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.talanta.2018.06.050. [14]

References [15] [1] T. Ferenci, Adaptation to life at micromolar nutrient levels: the regulation of Escherichia coli glucose transport by endoinduction and cAMP, FEMS Microbiol. Rev. 18 (1996) 301–317, http://dx.doi.org/10.1016/0168-6445(96)00019-8. [2] L. López-Maury, S. Marguerat, J. Bähler, Tuning gene expression to changing

[16]

6

environments: from rapid responses to evolutionary adaptation (68–68), Nat. Rev. Genet. 10 (2009), http://dx.doi.org/10.1038/nrg2500. H. Weber, T. Polen, J. Heuveling, V.F. Wendisch, R. Hengge, Genome-wide analysis of the general stress response network in escherichia coli: σ s -dependent genes, promoters, and sigma factor selectivity, J. Bacteriol. 187 (2005) 1591–1603, http:// dx.doi.org/10.1128/JB.187.5.1591. J.D. Irr, Control of nucleotide metabolism and ribosomal ribonucleic acid synthesis during nitrogen starvation of Escherichia coli, J. Bacteriol. 110 (1972) 554–561. D. Kern, E.R.P. Zuiderweg, The role of dynamics in allosteric regulation, Curr. Opin. Struct. Biol. 13 (2003) 748–757, http://dx.doi.org/10.1016/j.sbi.2003.10.008. E. Fischer, U. Sauer, Metabolic flux profiling of Escherichia coli mutants in central carbon metabolism using GC-MS, Eur. J. Biochem. 270 (2003) 880–891, http://dx. doi.org/10.1046/j.1432-1033.2003.03448.x. R. Schuetz, L. Kuepfer, U. Sauer, Systematic evaluation of objective functions for predicting intracellular fluxes in Escherichia coli, Mol. Syst. Biol. 3 (2007), http:// dx.doi.org/10.1038/msb4100162. U. Schaefer, W. Boos, R. Takors, D. Weuster-Botz, Automated sampling device for monitoring intracellular metabolite dynamics, Anal. Biochem. 270 (1999) 88–96, http://dx.doi.org/10.1006/abio.1999.4048. C. Chassagnole, N. Noisommit-Rizzi, J.W. Schmid, K. Mauch, M. Reuss, Dynamic modeling of the central carbon metabolism of Escherichia coli, Biotechnol. Bioeng. 79 (2002) 53–73, http://dx.doi.org/10.1002/bit.10288. A. Buchholz, R. Takors, C. Wandrey, Quantification of intracellular metabolites in Escherichia coli K12 using liquid chromatographic-electrospray ionization tandem mass spectrometric techniques, Anal. Biochem. 295 (2001) 129–137, http://dx.doi. org/10.1006/abio.2001.5183. U. Theobald, W. Mailinger, M. Reuss, M. Rizzi, In vivo analysis of glucose-induced fast changes in yeast adenine nucleotide pool applying a rapid sampling technique, Anal. Biochem. 214 (1993) 31–37, http://dx.doi.org/10.1006/abio.1993.1452. H.C. Lange, M. Eman, G. Van Zuijlen, D. Visser, J.C. Van Dam, J. Frank, M.J. De. Teixeira Mattos, J.J. Heijnen, Improved rapid sampling for in vivo kinetics of intracellular metabolites in Saccharomyces cerevisiae, Biotechnol. Bioeng. 75 (2001) 406–415, http://dx.doi.org/10.1002/bit.10048. W. de Koning, K. van Dam, A method for the determination of changes of glycolytic metabolites in yeast on a subsecond time scale using extraction at neutral pH, Anal. Biochem. 204 (1992) 118–123, http://dx.doi.org/10.1016/0003-2697(92) 90149-2. C. Wittmann, J.O. Krömer, P. Kiefer, T. Binz, E. Heinzle, Impact of the cold shock phenomenon on quantification of intracellular metabolites in bacteria, Anal. Biochem 327 (2004) 135–139, http://dx.doi.org/10.1016/j.ab.2004.01.002. Y. Shen, T. Fatemeh, L. Tang, Z. Cai, Quantitative metabolic network profiling of Escherichia coli: an overview of analytical methods for measurement of intracellular metabolites, TrAC - Trends Anal. Chem. 75 (2016) 141–150, http://dx. doi.org/10.1016/j.trac.2015.07.006. E.J. Want, I.D. Wilson, H. Gika, G. Theodoridis, R.S. Plumb, J. Shockcor, E. Holmes,

Talanta 189 (2018) 1–7

F. Ji et al.

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

J.K. Nicholson, Global metabolic profiling procedures for urine using UPLC-MS, Nat. Protoc. 5 (2010) 1005–1018, http://dx.doi.org/10.1038/nprot.2010.50. C.L. Winder, W.B. Dunn, S. Schuler, D. Broadhurst, R. Jarvis, G.M. Stephens, R. Goodacre, Global metabolic profiling of Escherichia coli cultures: an evaluation of methods for quenching and extraction of intracellular metabolites, Anal. Chem. 80 (2008) 2939–2948, http://dx.doi.org/10.1021/ac7023409. A.B. Canelas, A. Ten Pierick, C. Ras, R.M. Seifar, J.C. Van Dam, W.M. Van Gulik, J.J. Heijnen, Quantitative evaluation of intracellular metabolite extraction techniques for yeast metabolomics, Anal. Chem. 81 (2009) 7379–7389, http://dx.doi.org/ 10.1021/ac900999t. M.J. va. der Werf, K.M. Overkamp, B. Muilwijk, L. Coulier, T. Hankemeier, Microbial metabolomics: toward a platform with full metabolome coverage, Anal. Biochem. 370 (2007) 17–25, http://dx.doi.org/10.1016/j.ab.2007.07.022. 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 (2000) 6640–6645, http://dx.doi.org/10.1073/pnas.120163297. R. Lutz, H. Bujard, Independent and tight regulation of transcriptional units in Escherichia coli via the LacR/O, the TetR/O and AraC/I1-I2 regulatory elements, Nucleic Acids Res. 25 (1997) 1203–1210, http://dx.doi.org/10.1093/nar/25.6. 1203. V. de Lorenzo, M. Herrero, M. Metzke, K.N. Timmis, An upstream XylR- and IHFinduced nucleoprotein complex regulates the sigma 54-dependent Pu promoter of TOL plasmid, EMBO J. 10 (1991) 1159–1167. L.C. Thomason, N. Costantino, D.L. Court, E. coli genome manipulation by P1 transduction, Curr. Protoc. Mol. Biol. (2007) 1.17.1–1.17.8, http://dx.doi.org/10. 1002/0471142727.mb0117s79. B. Doublet, G. Douard, H. Targant, D. Meunier, J.Y. Madec, A. Cloeckaert, Antibiotic marker modifications of λ Red and FLP helper plasmids, pKD46 and pCP20, for inactivation of chromosomal genes using PCR products in multidrugresistant strains, J. Microbiol. Methods 75 (2008) 359–361, http://dx.doi.org/10. 1016/j.mimet.2008.06.010. B. Luo, K. Groenke, R. Takors, C. Wandrey, M. Oldiges, Simultaneous determination of multiple intracellular metabolites in glycolysis, pentose phosphate pathway and tricarboxylic acid cycle by liquid chromatography-mass spectrometry, J. Chromatogr. A. 1147 (2007) 153–164, http://dx.doi.org/10.1016/j.chroma.2007. 02.034. D. Yan, P. Lenz, T. Hwa, Overcoming fluctuation and leakage problems in the quantification of intracellular 2-oxoglutarate levels in Escherichia coli, Appl.

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

7

Environ. Microbiol. 77 (2011) 6763–6771, http://dx.doi.org/10.1128/AEM. 05257-11. B.D. Bennett, J. Yuan, E.H. Kimball, J.D. Rabinowitz, Absolute quantitation of intracellular metabolite concentrations by an isotope ratio-based approach, Nat. Protoc. 3 (2008) 1299–1311, http://dx.doi.org/10.1038/nprot.2008.107. J.D. Rabinowitz, E. Kimball, Acidic acetonitrile for cellular metabolome extraction from Escherichia coli, Anal. Chem. 79 (2007) 6167–6173, http://dx.doi.org/10. 1021/ac070470c. R. Prasad Maharjan, T. Ferenci, Global metabolite analysis: the influence of extraction methodology on metabolome profiles of Escherichia coli, Anal. Biochem. 313 (2003) 145–154, http://dx.doi.org/10.1016/S0003-2697(02)00536-5. T.M. Ramseier, D. Nègre, J.C. Cortay, M. Scarabel, a.J. Cozzone, M.H. Saier, In vitro binding of the pleiotropic transcriptional regulatory protein, FruR, to the fru, pps, ace, pts and icd operons of Escherichia coli and Salmonella typhimurium, J. Mol. Biol. 234 (1993) 28–44, http://dx.doi.org/10.1006/jmbi.1993.1561. T.M. Ramseier, Cra and the control of carbon flux via metabolic pathways, Res. Microbiol. 147 (1996) 489–493, http://dx.doi.org/10.1016/0923-2508(96) 84003-4. T. Shimada, N. Fujita, M. Maeda, A. Ishihama, Systematic search for the Cra-binding promoters using genomic SELEX system, Genes Cells 10 (2005) 907–918, http://dx. doi.org/10.1111/j.1365-2443.2005.00888.x. T. Shimada, K. Yamamoto, A. Ishihama, Novel members of the Cra regulon involved in carbon metabolism in Escherichia coli, J. Bacteriol. 193 (2011) 649–659, http:// dx.doi.org/10.1128/JB.01214-10. K. Kochanowski, B. Volkmer, L. Gerosa, B.R. Haverkorn van Rijsewijk, A. Schmidt, M. Heinemann, Functioning of a metabolic flux sensor in Escherichia coli, Proc. Natl. Acad. Sci. 110 (2013) 1130–1135, http://dx.doi.org/10.1073/pnas. 1202582110. C.D. Doucette, D.J. Schwab, N.S. Wingreen, J.D. Rabinowitz, α-ketoglutarate coordinates carbon and nitrogen utilization via enzyme I inhibition, Nat. Chem. Biol. 7 (2011) 894–901, http://dx.doi.org/10.1038/nchembio.685. C. You, H. Okano, S. Hui, Z. Zhang, M. Kim, C.W. Gunderson, Y.-P. Wang, P. Lenz, D. Yan, T. Hwa, Coordination of bacterial proteome with metabolism by cyclic AMP signalling, Nature 500 (2013) 301–306, http://dx.doi.org/10.1038/nature12446. B.D. Bennett, E.H. Kimball, M. Gao, R. Osterhout, S.J. Van Dien, J.D. Rabinowitz, Absolute metabolite concentrations and implied enzyme active site occupancy in Escherichia coli, Nat. Chem. Biol. 5 (2009) 593–599, http://dx.doi.org/10.1038/ nchembio.186.