Quantitative Metabolite Profiling Utilizing Parallel Column Analysis for

(1-4) So far, in order to accomplish a large coverage of the primary carbon metabolome, two main strategies were pursued: ion pairing chromatography(5...
0 downloads 0 Views 806KB Size
Technical Note pubs.acs.org/ac

Quantitative Metabolite Profiling Utilizing Parallel Column Analysis for Simultaneous Reversed-Phase and Hydrophilic Interaction Liquid Chromatography Separations Combined with Tandem Mass Spectrometry Kristaps Klavins,†,‡ Hedda Drexler,† Stephan Hann,†,‡ and Gunda Koellensperger*,†,‡,§ †

Department of Chemistry, Division of Analytical Chemistry, University of Natural Resources and Life Sciences, BOKU-Vienna, Muthgasse 18, 1190 Vienna, Austria ‡ Austrian Centre of Industrial Biotechnology (ACIB), Muthgasse 18, 1190 Vienna, Austria § Institute of Analytical Chemistry, Faculty of Chemistry, University of Vienna, Währinger strasse 38, 1090 Vienna, Austria S Supporting Information *

ABSTRACT: In this work, a fully automated parallel LC column method was established in order to perform orthogonal hydrophilic interaction chromatography (HILIC) and reversed-phase (RPLC) chromatography within one analytical run for targeted quantitative mass spectrometric determination of metabolites from central carbon metabolism. In this way, the analytical throughput could be significantly improved compared to previously established dual separation work flows involving two separate analytical runs. Two sample aliquots were simultaneously injected onto a dual column setup columns using a ten-port valve, and parallel separations were carried out. Sub 2 μm particle size stationary phases were employed for both separation methods. HILIC and RPLC eluents were combined post column followed by ESI-MS/MS detection. The orthogonal separations were optimized, aiming at an overall separation with 2 retention time segments, while reversed-phase separation was accomplished within 5.5 min; metabolites on the HILIC phase were retained for a minimum time of 6 min. The overall run time was 15 min. The setup was applied to the quantification of 30 primary intercellular metabolites, including amino acids, organic acids, and nucleotides employing internal standardization by a fully 13C-labeled yeast extract. The comparison with HILIC−MS/MS and RPLC−MS/MS in separate analytical runs revealed that an excellent analytical performance was achieved by the parallel LC column method. The experimental repeatability (N = 5) was on average 5%). Moreover, limits of detection for the new approach ranging from 0.002−15 μM were in a good agreement with ones obtained in separate HILIC−MS/MS and RPLC−MS/MS runs (ranging from 0.01−44 μM).

M

ion pairing chromatography exclusively. Furthermore, ionpairing chromatography displayed limited separation power for isomeric compounds (e.g., citrate/isocitrate).5,6,9 Dual separation approaches implementing orthogonal HILIC and RPLC chromatographic separation approaches showed to be a valuable alternative providing excellent separation efficiency and metabolite coverage.10−15 As a drawback, conventionally, the combination of two separation methods implied two separated analytical runs for each sample, resulting in increased overall measurement time and a higher consumption of samples as well as isotopically labeled (internal) standards. Evidently, automated on-line two-dimensional (2D) strategies could have the potential of shorter analysis time and increased sample throughput, which is particularly important for routine analysis. However, successful online combination of HILIC and RPLC

etabolite profiling, one of the pillars of the youngest “omics” discipline metabolomics, addresses the comprehensive quantitative analysis of the primary metabolome. This analytical task inherently implies dealing with compounds of highly diverse chemical and physical properties, ideally using only as few as possible different analytical methods. In recent years, liquid chromatography combined with mass spectrometric detection evolved to one of the core techniques in the field, due to excellent figures of merit such as, for example, sensitivity and robustness and considering the challenges of global analysis, its versatility and high throughput capability.1−4 So far, in order to accomplish a large coverage of the primary carbon metabolome, two main strategies were pursued: ion pairing chromatography5−9 or the application of orthogonal separation modes, conventionally reversed-phase (RPLC) and hydrophilic interaction chromatography.10−15 Ion pairing chromatography displays several shortcomings compared to the latter approach. The used ion pairing reagent can lead to contamination of the instrumental system and, consequently, it is advisible to dedicate the MS system to the combination with © 2014 American Chemical Society

Received: January 24, 2014 Accepted: March 28, 2014 Published: March 28, 2014 4145

dx.doi.org/10.1021/ac5003454 | Anal. Chem. 2014, 86, 4145−4150

Analytical Chemistry

Technical Note

Table 1. Results of QC Sample Quantification Employing Separate Column and Parallel Column Approachesa separate columns

parallel columns

comp.

LC

Ctheor (μM)

Cmeasured (μM)

RSD (%)

accuracy (%)

LOD (μM)

Cmeasured (μM)

RSD (%)

accuracy (%)

LOD (μM)

PEP Thi Mali 5CMP IsoCit Val Met 5AMP Cit 5GMP Fum NADP Suc NAD 3AMP Ile Tyr Leu Phe Trp Ala Ser Gln Asn Pro Arg Lys Glu Asp His

RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC RPLC HILIC HILIC HILIC HILIC HILIC HILIC HILIC HILIC HILIC HILIC

5.0 0.5 10 2.5 2.5 200 5.0 2.5 10 2.5 5.0 5.0 5.0 5.0 2.5 25 100 25 5.0 5.0 100 100 50 200 100 50 100 80 200 50

5.1 0.49 9.9 2.7 2.6 218 4.9 2.6 11 2.8 5.2 5.1 5.1 4.8 2.5 25 96 25 5.3 5.1 99 100 40 197 101 50 97 79 180 50

5 10 4 12 1 3 8 1 2 1 4 2 9 5 5 3 2 2 2 3 8 1 2 8 1 1 2 1 2 2

102 98 99 109 105 109 98 102 105 114 104 102 101 96 100 101 96 99 107 103 99 100 80 98 101 100 97 99 90 100

0.09 0.02 0.24 0.01 0.19 0.19 0.15 0.02 0.09 0.05 3.6 0.46 0.5 0.09 0.01 0.10 0.18 0.08 0.01 0.003 25 1.2 1.7 2.6 1.0 0.28 3.4 2.3 44 2.0

4.9 0.5 9.7 2.6 2.5 198 5.1 2.5 10 2.5 4.9 5.0 5.2 5.0 2.4 26 97 25 5.2 4.8 101 101 41 203 105 50 104 79 198 50

4 3 2 6 2 2 2 2 2 3 5 3 5 1 5 1 2 1 2 4 5 2 1 2 3 2 1 6 1 2

98 100 97 103 101 99 101 101 100 101 97 100 105 100 97 103 97 100 104 95 101 101 81 102 105 100 104 99 99 99

0.06 0.01 0.1 0.007 0.03 0.49 0.01 0.004 0.04 0.002 1.1 0.01 0.6 0.004 0.01 0.04 0.01 0.14 0.04 0.04 5.1 3.2 0.93 2.2 6.7 0.12 0.16 15 2.1 0.02

a

Separation method used to assign each compound is indicated as well as theoretical compound concentration in the QC sample (Ctheor). For each approach average (N = 5) measured concentration (Cmeasured), relative standard deviation [RSD (N = 5)], accuracy, and limit of detection (LOD) are given.

equilibration and analysis times of the subsequent analytical runs. For example, while separation was carried out on one of the columns, others were undergoing equilibration. So far, the ability of the parallel column setup to enable parallel and simultaneous analysis of the same sample on two different columns has been overlooked. It especially holds true for targeted metabolite profiling.

separations is challenging due to the fact that weak eluting mobile phases in one separation are strong in the other. Technically, the developed automated 2D set-ups for untargeted analysis involved coupling HILIC and RPLC in series,16 using fast column switching,17,18 stop-flow comprehensive 2D liquid chromatography19 or off-line 2D chromatography.20 So far, in the field of metabolomics, online combination of RPLC and HILIC concerned nontargeted MS analysis increasing the number of detected and identified compounds within one chromatographic run.18 However, these setups required long run times (>45 min) and, therefore, fell short for the high-throughput analysis. In the presented work, we modified our recently introduced dual separation approach12,21 established for key components of the ubiquitous primary metabolome such as TCA cycle, PPP pathway, glycolysis, and amino acid synthesis by proposing an automated parallel dual column setup, enabling parallel and simultaneous analysis of the same sample on orthogonal HILIC and RPLC. This strategy is applied for the first time for targeted primary metabolite profiling. Analytical figures of merit for both dual separation methods will be discussed. Up to now, multiple parallel columns using the same separation mode were established when analysis aimed at high throughput (e.g., in the field of drug development22,23 and proteome analysis).24 The overall analysis time was reduced, by overlapping



EXPERIMENTAL SECTION Standards and Chemicals. The following standard substances were purchased from Sigma-Aldrich or Fluka (Vienna, Austria): L-alanine 99.5% (Ala), adenosine 3′monophosphoric acid 97% (3AMP), adenosine 5′-monophosphate sodium salt 99% (5AMP), L-arginine 98.5% (Arg), L-asparagine 98% (Asn), cytidine 5′-monophosphate sodium salt 99% (5CMP), fumaric acid 99.5% (Fum), guanosine 5′monophosphate disodium salt 99% (5GMP), DL-histidine 99% (His), DL-isocitric acid sodium salt 98% (IsoCit), L-isoleucine 98% (Ile), L-lysine 98% (Lys), β-nicotinamide adenine dinucleotide sodium salt 95% (NAD), β-nicotinamide adenine dinucleotide phosphate disodium salt 97% (NADP), phospho(enol)pyruvate 99% (PEP), L-proline 99.5% (Pro), DL-serine 99% (Ser), succinic acid 99.5% (Suc), L-tryptophan 99.5% (Trp), thiamine hydrochloride 99% (Thi), and L-tyrosine 99% (Tyr). The following substances were purchased from Merck 4146

dx.doi.org/10.1021/ac5003454 | Anal. Chem. 2014, 86, 4145−4150

Analytical Chemistry

Technical Note

guard (Macherey-Nagel, Düren, Germany) column, 50 × 3.0 mm, and particle diameter 1.8 μm. The LC analysis was carried out at 40 °C. The mobile phase A was 10 mM ammonium formate solution in LC−MS-grade water, with the pH adjusted to 3.25 using formic acid; mobile phase B was acetonitrile. The flow rate was set to 300 μL min−1. The mobile phase A gradient was applied as follows: mobile phase A was increased from 10% to 40% in 8 min and then increased to 90% in 0.1 min and was held for 1.9 min, followed by reconstitution of the starting conditions within 0.1 min and re-equilibration with 10% A for 4.9 min, resulting in a total analysis time of 15 min. Reversed-phase chromatography was conducted on the Accela 1250 HPLC system. Separation was performed using the Waters Acquity UPLC HSS T3 column 150 × 2 mm, 1.8 μm particle size column. The mobile phase A was 0.1% (v/v) formic acid solution in LC−MS-grade water, and the mobile phase B was methanol. The flow rate was set to 400 μL min−1 and the column temperature to 40 °C. The initial mobile phase B concentration was 0%. The mobile phase B gradient was applied as follows: 0% B was held for 0.7 min and then increased to 40% in 3 min and held for 0.7 min and then increased to 100% in 0.1 min and held for 0.6 min, followed by reconstitution of the starting conditions within 0.1 min and reequilibration with 0% B for 7.5 min. This resulted in a total analysis time of 12 min. The TSQ Vantage ESI-MS/MS (Thermo Scientific) featuring a heated ESI interface served for MS/MS analysis. Ion source parameters for both negative and positive modes were set as follows: vaporizer temperature, 350 °C; ion transfer tube temperature, 350 °C; auxiliary gas pressure, 15 arbitrary units; sheet gas pressure, 40 arbitrary units; ion sweep gas pressure, 0 arbitrary units; declustering voltage, 0 V; spray voltage, 3000 V; and the collision gas pressure for selected reaction monitoring (SRM) was set to 1.5 mTorr. Selected reaction monitoring (SRM) transitions of all compounds were obtained via the flow injection of a 10 μM single standard using a syringe pump coupled to a LC pump using a zero-volume Tpiece connector. XCalibur tune software was used to optimize SRM transitions for each compound. The scheduled SRM mode was employed for the detection. For each analyte, the specific SRM transition was monitored in its retention time (RT) window; for reversed-phase compounds it was RT ± 0.5 min, and for HILIC compounds it was RT ± 1 min. The cycle time for the SRM method was set to a fixed value of 0.2 s. More detailed information about SRM settings are provided in Table S-1 of the Supporting Information.

KGaA (Darmstadt, Germany): L-aspartic acid 99% (Asp), citric acid 99.5%−100.5% (Cit), L-glutamic acid 99% (Glu), Lglutamine 99% (Gln), L-leucine 99% (Leu), DL-malic acid 99.5% (Mali), L-methionine 99% (Met), L-phenylalanine 99% (Phe), and L-valine 99% (Val). For chromatography, LC−MSgrade water, LC−MS-grade acetonitrile and ammonium acetate 99.0% from Sigma-Aldrich, LC−MS-grade methanol from Fisher Scientific (Loughborough, U.K.), and formic acid 98%−100% Suprapur from Merck were used. In vivo fully 13 C-labeled yeast was used as the internal standard.21 Preparation of Samples and Internal Standard. The calibration solutions were prepared starting from single standard stock solutions (in case of amino acids diluted in in 0.1 M HCl, 3AMP in 0.1 M NaOH, and all other components in LC−MS grade water). Eight-hundred microliters of aqueous calibration solutions were spiked with the 200 μL of aqueous internal standard solution, resulting in a final biomass concentration of 4 mg mL−1. Independent quality control (QC) samples were prepared analogously, yielding metabolite concentration levels, as given in Table 1. As mentioned before, the fully 13C labeled yeast extract was used as the internal standard. A detailed description of the internal standard preparation procedures is available elsewhere.21 In brief, the Pichia pastoris cell culture grown in chemostat cultivation was sampled using a cold methanol quenching and fast filtration method. The extraction of the cell pellet corresponding to 4 mg cell dry weight utilizing a boiling ethanol protocol was carried out. The obtained extracts were evaporated using a speed vac system and reconstituted in 1 mL of water. Instrumental. The analysis was carried out on the LC− MS/MS system consisting of CTC PAL autosampler (CTC Analytics AG, Zwingen, Switzerland), Thermo Scientific Accela 1250 and Accela 600 pumps (Thermo Scientific, FL, USA), with a 10 port 15 kPsi switching valve (VICI Valco Instruments). The instrumental setup is shown in Figure 1. The HILIC separation was carried out using the Accela 600 pump. The separation was performed on a Nucleodur HILIC column (Macherey-Nagel, Düren, Germany), 100 × 3 mm with particle diameter 1.8 μm equipped with a Nucleodur HILIC



RESULTS AND DISCUSSION So far in our laboratory, quantitative profiling of the primary metabolome in yeast cell extracts has been carried out, employing dual LC separations involving orthogonal HILIC and RPLC. This strategy, which in the following will be referred to as the separate column approach, was the starting point of the presented work. A detailed description about the analytical tool set is available elsewhere.12,21,25 In brief, the RPLC separation was developed for the quantification of 24 metabolites, including apolar amino acids, nucleotides, and organic acids. For this purpose, a stationary phase (Water Atlantis T3, 150 × 4.6 mm, 3 μm particle size) compatible with 100% aqueous mobile phase was employed (run time of 20 min). Additionally, HILIC addressed the analysis of 10 hydrophilic amino acids. For SeQuant ZicHILIC 150 × 4.6 mm, 3.5 μm particle size columns, run times of 20 min could be

Figure 1. Instrument setup employed for parallel column analysis. 4147

dx.doi.org/10.1021/ac5003454 | Anal. Chem. 2014, 86, 4145−4150

Analytical Chemistry

Technical Note

Figure 2. Chromatogram of QC sample obtained with the developed parallel column setup.

that RPLC compounds eluted in a time segment from 1.1 to 5.2 min and HILIC compounds from 6.4 to 10.1 min allowed avoiding interferences caused by the possible peak overlapping from both separations. The ESI−MS/MS system was optimized for a flow rate of 700 μL min −1, which resulted from the combination of the final RPLC and HILIC flows. (3) A shorter overall analysis time would imply accommodating flow rates >700 μL min −1. However, in the past, drawbacks concerning robustness and sensitivity for ESI-MS analysis at flow rates of 1000 μL min −1 had been observed for the MS system used in this study. Significantly higher noise, hence, lower signal-tonoise ratio, was observed when the analyses at a flow rate of 1000 μL min −1 were carried out compared to a 500 μL min −1 flow rate, which was the optimal value for the given instrument setup. The combined flow rate of 700 μL min −1 provided the shortest possible run time with the minimal loss of sensitivity. The implementation of the 1.8 μm particle size compared to the earlier established 3 μm particle size resulted in a pressure increase from 120 to 600 bar for the RPLC separation. Hence, the UPLC pump with the maximum operating pressure of 1250 bar was employed (pump 2) for this separation. HILIC separation was carried out using a flow rate of 300 μL min −1, which displayed a moderate backpressure starting at 130 bar. Even though, judging by the backpressure, the flow rate could be increased, the set flow rate was ideal for the planned parallel column setup as it provided optimal separation and retention time windows distinct from RPLC separation for all assigned compounds. In a next step, the analytical performance of the new chromatographic design was evaluated in terms of limits of detection, repeatability, and accuracy. A panel of 30 metabolites was comparatively addressed by the parallel column setup and the conventional distinct RPLC−MS/MS and HILIC−MS/MS analysis.12,21 For both LC strategies, a six-point calibration curve with internal standardization was used for quantification. As in the earlier research, U13C-labeled yeast, namely Pichia pastoris, was used as an internal standard. Detailed information about calibration used for the parallel column analysis is given in Table S-2 of the Supporting Information. The analyzed sample, denoted as the quality control (QC) sample in the following, was prepared from commercially available standard substances and the U13C-labeled yeast cell extract. The concentration levels of each compound (Table 1) were adjusted to fit the range typically found in yeast cell extracts.12,21 Taking into account that the internal stand-

obtained. Recently, our HILIC method was improved in terms of repeatability and robustness as well as reduced analysis times utilizing a novel zwitterionic modified silica-based stationary phase with 1.8 μm particle size (Nucleodur HILIC, 100 × 3 mm). Moreover the selected stationary phase enabled injection of 100% aqueous samples. The method could be successfully applied for mass spectrometric quantification of 22 amino acids.25 The aim of this work was the development of an instrumental setup allowing the automated combination of the dual separation within one analytical run, increasing sample throughput. Moreover, the developed parallel column approach provided HILIC and RPLC chromatography on sub 2 μm particle stationary phases. As can be readily observed in Figure 1, a straightforward instrumental setup was designed. In the given setup the 10-port switching valve served as an injector valve for both chromatographic separations, enabling simultaneous and parallel RPLC and HILIC analysis. Hence, the fully automated procedure started with the sample loading process (see Figure 1). Both 5 μL sample loops were filled utilizing the autosampler connected to the 10-port valve. After 0.1 min, the 10-port valve switched to position “inject” (see Figure 1). Then the samples were transferred from the sample loops to the two analytical columns by two independent HPLC pumps. HILIC and RPLC separations were carried out independently applying gradient elution for each separation. After separation, eluents from both columns were combined using a zero-volume Tpiece and ESI-MS/MS detection was carried out. Evidently, the total analysis time of described setup was given by the longest separation method. On the RPLC column utilizing a 1.8 μm particle size stationary phase and a flow rate of 400 μL min −1 (resulting back pressure was 600 bar), 24 metabolites could be separated within 7 min. This was a considerable improvement compared with earlier work on 3 μm particle size columns, where the same number of metabolites were separated within 20 min. Hence, the HILIC separation, which was accomplished within 15 min, governed the total analysis time of the parallel column set up. This relatively long separation time, despite the fact that HILIC was conducted on sub-2 μm particulate phases was used because of three major reasons: (1) the long re-equilibration time necessary for robust HILIC separations and (2) the HILIC and the RPLC gradients were adjusted in the way that the two metabolite panels addressed by the orthogonal separation methods now had different retention time segments. The fact 4148

dx.doi.org/10.1021/ac5003454 | Anal. Chem. 2014, 86, 4145−4150

Analytical Chemistry

Technical Note

excellent retention time stability with RSDs not exceeding 1% (N = 5 measurements distributed over an investigational period of 10 h) for all compounds (Table S1 of the Supporting Information).

ardization was performed with a yeast cell extract (corresponding to a biomass concentration of 4 mg mL−1), the QC sample being a spiked cell extract could be considered as a close representation of a real sample matrix. The chromatogram of the QC sample obtained by the parallel column approach is shown in Figure 2. Excellent chromatographic separation of all 30 compounds was achieved. As a matter of fact, critical isomer pairs (leucine/isoleucine, citrate/isocitrate, and 3AMP/5AMP) as well as isobars (lysine/glutamine) were baseline separated. The first 20 eluting compounds, including amino acids, organic acids, and nucleotides were separated by RPLC separation and 10 compounds (hydrophilic amino acids) were determined, employing HILIC separation. It should be once again pointed out that two chromatographic peaks (one from RPLC, one from HILIC) for each investigated metabolite had to be expected as two separations were carried out simultaneously. However, due to the fact that the scheduled SRM measurement segments, adjusted to retention time windows were implemented, selectivity was provided. Two additional experiments were carried out in order to confirm this. First, the RPLC column was detached from the system and a standard mixture exclusively containing the compounds attributed to the RPLC separation was analyzed. In a next step, an analogous experiment was carried out for the metabolite panel separated by the HILIC. Accordingly, the HILIC column was detached from the system. In both cases, the signal intensity for all SRM segments corresponded to the method blank, hence confirming selectivity of the developed method. Analytical figures of merit for the parallel column approach and the separate column approach were compared and are summarized in Table 1. Repeatability was calculated from five repeated sample injections and expressed as relative standard deviation (RSD). The typically repeatability for LC−MS/MS measurements is around 1−5%, depending on signal height. As a matter of fact, the parallel column method displayed a comparable to even slightly improved repeatability compared to the two separate LC methods. The repeatability of the new design ranged from 1% to 6% and was >5% for only 2 compounds. For the conventional HILIC−MS/MS and RPLC−MS/MS, typically the repeatability ranged from 1% to 12% and was >5% for 6 compounds. Accuracy was calculated as the ratio between the measured metabolite concentrations and the theoretical compound concentration in the QC sample. Both measurement approaches provided good accuracy, ranging from 95% to 105% for the majority of the investigated compounds. The LODs listed in Table 1 were calculated considering the baseline noise using the 3σ criteria. It should be pointed out that, despite the post column dilution taking place in the parallel column approach, LOD values for both approaches were in the same range: from 0.002 to 15 μM for the parallel column approach and from 0.01 to 44 μM for the separate column approach. For compounds assessed by the RPLC separation, this could be explained by the fact that the sub 2 μm column was employed for the parallel column approach. This led to narrower peaks and hence increased S/N ratios compared to the RPLC−MS/MS determinations on 3 μm particle columns. For compounds assessed by the HILIC separation, the sensitivity loss due to the post column dilution was most likely compensated for by enhanced protonation efficiency of amino acids. The latter could be explained by the fact that the HILIC eluent with pH 3.25 was slightly acidified upon post column dilution with the RPLC mobile phase (pH 2.6). It is worth noting that parallel column analysis displayed



CONCLUSION For the first time, the potential of an instrumental LC/LC− MS/MS set up employing two parallel, orthogonal separation modes (i.e., reversed-phase and HILIC), for quantitative metabolite profiling was examined. We demonstrated that this strategy is a simple yet effective approach for coupling existing RPLC and HILIC methods, hence, decreasing analysis time and increasing throughput. The obtained total analysis time was 15 min. Moreover, the parallel column analysis displayed better performance and analytical figures of merit than previously established HILIC−MS/MS and RPLC−MS/MS.



ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: +43-476546087. Fax: +43-47654-6059. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We are deeply grateful for the help of Hannes Russmayer for the preparation of the internal standard and Halimat Ahmatova for LC−MS/MS data analysis. This work has been supported by the Federal Ministry of Science, Research, and Economy (BMWFW), the Federal Ministry of Traffic, Innovation and Technology (bmvit), the Styrian Business Promotion Agency SFG, the Standortagentur Tirol and ZIT-Technology Agency of the City of Vienna through the COMET-Funding Program managed by the Austrian Research Promotion Agency FFG. EQ BOKU VIBT GmbH is acknowledged for providing LC− MS/MS instrumentation.



REFERENCES

(1) Xiao, J. F.; Zhou, B.; Ressom, H. W. TrAC, Trends Anal. Chem. 2012, 32, 1−14. (2) Oldiges, M.; Lutz, S.; Pflug, S.; Schroer, K.; Stein, N.; Wiendahl, C. Appl. Microbiol. Biotechnol. 2007, 76, 495−511. (3) Theodoridis, G. A.; Gika, H. G.; Want, E. J.; Wilson, I. D. Anal. Chim. Acta 2012, 711, 7−16. (4) Gika, H. G.; Theodoridis, G. A.; Plumb, R. S.; Wilson, I. D. J. Pharm. Biomed. Anal. 2014, 87, 12−25. (5) Lu, W.; Clasquin, M. F.; Melamud, E.; Amador-Noguez, D.; Caudy, A. A.; Rabinowitz, J. D. Anal. Chem. 2010, 82, 3212−3221. (6) Luo, B.; Groenke, K.; Takors, R.; Wandrey, C.; Oldiges, M. J. Chromatogr., A 2007, 1147, 153−164. (7) Buescher, J. M.; Moco, S.; Sauer, U.; Zamboni, N. Anal. Chem. 2010, 82, 4403−4412. (8) Coulier, L.; Bas, R.; Jespersen, S.; Verheij, E.; van der Werf, M. J.; Hankemeier, T. Anal. Chem. 2006, 78, 6573−6582. (9) Kiefer, P.; Delmotte, N.; Vorholt, J. A. Anal. Chem. 2011, 83, 850−855. (10) Preinerstorfer, B.; Schiesel, S.; Lammerhofer, M.; Lindner, W. J. Chromatogr., A 2010, 1217, 312−328.

4149

dx.doi.org/10.1021/ac5003454 | Anal. Chem. 2014, 86, 4145−4150

Analytical Chemistry

Technical Note

(11) Yang, S.; Sadilek, M.; Synovec, R. E.; Lidstrom, M. E. J. Chromatogr., A 2009, 1216, 3280−3289. (12) Klavins, K.; Neubauer, S.; Al Chalabi, A.; Sonntag, D.; Haberhauer-Troyer, C.; Russmayer, H.; Sauer, M.; Mattanovich, D.; Hann, S.; Koellensperger, G. Anal. Bioanal. Chem. 2013, 405, 5159− 5169. (13) Lu, W.; Bennett, B. D.; Rabinowitz, J. D. J. Chromatogr., B 2008, 871, 236−242. (14) Honore, A. H.; Thorsen, M.; Skov, T. Anal. Bioanal. Chem. 2013, 405, 8151−8170. (15) Ivanisevic, J.; Zhu, Z.-J.; Plate, L.; Tautenhahn, R. Anal. Chem. 2013, 85, 6876−6884. (16) Greco, G.; Grosse, S.; Letzel, T. J. Sep. Sci. 2013, 36, 1379− 1388. (17) Lam, M. P.; Siu, S. O.; Lau, E.; Mao, X.; Sun, H. Z.; Chiu, P. C.; Yeung, W. S.; Cox, D. M.; Chu, I. K. Anal. Bioanal. Chem. 2010, 398, 791−804. (18) Wang, Y.; Lehmann, R.; Lu, X.; Zhao, X.; Xu, G. J. Chromatogr., A 2008, 1204, 28−34. (19) Dugo, P.; Fawzy, N.; Cichello, F.; Cacciola, F.; Donato, P.; Mondello, L. J. Chromatogr., A 2013, 1278, 46−53. (20) Schiesel, S.; Lammerhofer, M.; Lindner, W. J. Chromatogr., A 2012, 1259, 100−110. (21) Neubauer, S.; Haberhauer-Troyer, C.; Klavins, K.; Russmayer, H.; Steiger, M. G.; Gasser, B.; Sauer, M.; Mattanovich, D.; Hann, S.; Koellensperger, G. J. Sep. Sci. 2012, 35, 3091−3105. (22) Yang, L.; Wu, N.; Rudewicz, P. J. J. Chromatogr., A 2001, 926, 43−55. (23) Xia, Y. Q.; Hop, C. E.; Liu, D. Q.; Vincent, S. H.; Chiu, S. H. Rapid Commun. Mass Spectrom. 2001, 15, 2135−2144. (24) Shen, Y.; Tolić, N.; Rui Zhao, L. P.-T.; Li, L.; Berger, S. J.; Harkewicz, R.; Anderson, G. A.; Belov, M. E.; Smith, R. D. Anal. Chem. 2001, 73, 3011−3021. (25) Guerrasio, R.; Haberhauer-Troyer, C.; Mattanovich, D.; Koellensperger, G.; Hann, S. Anal. Bioanal. Chem. 2013, 496, 915− 922.

4150

dx.doi.org/10.1021/ac5003454 | Anal. Chem. 2014, 86, 4145−4150