www.jss-journal.com

Page 1

Journal of Separation Science

Fast quantification of endogenous carbohydrates in plasma using hydrophilic interaction liquid chromatography coupled with tandem mass spectrometry Bangjie Zhu1, Feng Liu1, Xituo Li1, Yan Wang1, Xue Gu1, Jieyu Dai2, Guiming Wang1, Yu Cheng1*, Chao Yan1* 1

School of Pharmacy, Shanghai Jiao Tong University, Shanghai 200240, China;

2

DMPK Department, HD Biosciences Co., Ltd

*

Correspondence to: Professor Chao Yan, Email: [email protected];

or Dr. Yu Cheng, E-mail: [email protected]

Received: 19-Aug-2014; Revised: 08-Oct-2014; Accepted: 08-Oct-2014 This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1002/jssc.201400899. This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 2

Journal of Separation Science

Abstract Objective: Endogenous carbohydrates in biosamples are frequently highlighted as the most differential metabolites in many metabolomics studies. A simple, fast, simultaneous quantitative method for 16 endogenous carbohydrates in plasma has been developed using hydrophilic interaction liquid chromatography coupled with tandem mass spectrometry. Method: In order to quantify 16 endogenous carbohydrates in plasma, various conditions, including columns, chromatographic conditions, mass spectrometry conditions, and plasma preparation methods, were investigated. Different conditions in this quantified analysis were performed and optimized. The reproducibility, precision, recovery, and stability of the method were verified. Result: The results indicated that a methanol/acetonitrile (50:50, v/v) mixture could effectively and reproducibly precipitate rat plasma proteins. Cold organic solvents coupled with vortex for 1 min and incubated at –20°C for 20 min were the most optimal conditions for protein precipitation and extraction. The results, according to the linearity, recovery, precision, and stability, showed that the method was satisfactory in the quantification of endogenous carbohydrates in rat plasma. Conclusion: The quantified analysis of endogenous carbohydrates in rat plasma performed excellently in the sensitivity, high throughput and simple sample preparation, which met the requirement of quantification in specific expanded metabolomic studies after the global metabolic profiling research.

Keywords: endogenous carbohydrate, hydrophilic interaction chromatography, tandem mass spectrometry, plasma This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 3

Journal of Separation Science

Abbreviations: CE, collision energy; CV, coefficient of variation; DP, declustering potential; ELSD,

evaporative

light-scattering

detection;

HPAEC–PAD,

high-performance

anion-exchange chromatography coupled with pulsed amperometric detection; LLOQ, lower limit of quantification; ULOQ, upper limit of quantification; ACN, acetonitrile; MeOH, methanol

This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 4

Journal of Separation Science

1 Introduction Carbohydrates, the most abundant natural products, are generally regarded as food sources and structurally fundamental blocks in organisms. Recently, studies of carbohydrates have attracted more attention for their diverse bioactivities and biological functions. Carbohydrates have been described to be useful for improving vibriosis resistance and enhancing immune activity [1]. Yet, many of recent epidemiological studies, clinical trials, and animal studies continue to point to the contribution of excess dietary carbohydrate, especially fructose, to the risk factors for nonalcoholic fatty liver disease [2, 3]. The specific carbohydrate diet and other low-complex-carbohydrate diets may be possible therapeutic options for pediatric Crohn’s disease [4]. Several non-digestible oligosaccharides are also related to a lower risk of infections and diarrhea [5]. Moreover, during the last two decades, metabolomics has become a new approach to study different diseases and carbohydrates were frequently highlighted in the most differential metabolite list in the metabolic profiling studies of biosamples such as plasma, serum and urine [6–8]. Thus, a rapid and simple quantification method to analyze endogenous carbohydrates in the biosamples becomes necessary for expanded studies in the metabolomics approaches. The analysis of complex carbohydrate mixtures is still a challenge, not only because of the low concentration in the plasma or other biosamples, but also because many carbohydrates with similar properties and high hydrophilicity are isomeric, monosaccharides could have different structures (ring opened or closed, different ring size and conformations), structures of oligosaccharides are often branched and most carbohydrates lack chromophores This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 5

Journal of Separation Science

or fluorophores, a property that makes detection difficult [9]. Chromatographic techniques are commonly used for the analysis of carbohydrates. Complex mixtures of carbohydrates have been appropriately characterized with a relatively good resolution and sensitivity by GC [10] and GC–MS [11, 12]. However, carbohydrates are not suited to direct analysis by GC or GC–MS and require preceding derivatization which is not applicable for a high throughput measurement. Regarding carbohydrate detection, the lack of an effective chromophore or fluorophore group in saccharide structures prevents their direct detection from universal spectrophotometric detectors such as UV and fluorescence detection. This is the reason why the detection of carbohydrates by UV absorption and fluorescent detectors must be performed after a tedious derivatization [13]. High-performance anion-exchange chromatography coupled with pulsed amperometric detection (HPAEC–PAD) is a high-resolution technique for trace analysis of carbohydrates and a powerful alternative to traditional HPLC methods [14–16]. However, some sample pre-treatment may be required to remove interfering compounds that would affect detection, and degradation could become a problem at a high pH required in the analysis. Although evaporative light-scattering detection (ELSD) is used as a semi-universal detector for carbohydrates during the last decade [17–19], quantification by ELSD is less straightforward, since the response factor is generally not linear. The detection of compounds using MS allows the distinction between homologues of different masses, however, isobars such as glucose, fructose cannot be differentiated by mass alone. HPLC can provide a distinct separation of carbohydrate isobars. Reverse-phase chromatographic separation using C18 columns has been used to separate carbohydrates [20, 21], but a simultaneous use of derivatization reagent is necessary to This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 6

Journal of Separation Science

increase the hydrophobicity of the sugars. Recently, HILIC has bloomed to become an alternative to RPLC for the separation of the polar compounds and high compatibility with MS. In recent years, such HILIC approaches are increasingly being chosen for various polar compounds analysis [22–24]. However, the underivatized hybrid silica (BEH) ACQUITY UPLC HILIC column has been shown to be effective for the retention and separation of polar basic compounds, but limited for other polar compounds. To meet the analytical requirement, we investigated the use of HILIC–MS/MS with HILIC amide column using a new stationary phase, to achieve a simultaneous analysis of over ten carbohydrates. In this work, we aimed to develop a rapid and simple quantitative method to separate 16 carbohydrates and quantify these endogenous metabolites in rat plasma using HILIC–UHPLC–ESI-MS/MS.

2 Materials and methods 2.1 Chemicals Acetonitrile (LC–MS grade), methanol (HPLC grade), formic acid (99.5%, LC/MS), ammonium acetate and ammonium hydroxide solution (≥25% NH3 in H2O) were obtained from Sigma–Aldrich (St. Louis, MO, USA). Melibiose, sucrose, maltose, maltotriose, maltotetraose, glucose, fructose, xylose, ribose, raffinose, fucose, inosine, adenosine, mannitol, erythritol, arabitol and lamivudine were also purchased from Sigma–Aldrich. 13

L-Methionine-

C5 ,

15

N (99%) was from Cambridge Isotope Laboratories (Andover, MA,

USA). Pure water was prepared from a Milli-Q purification system (Millipore, Bedford, MA, USA). This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 7

Journal of Separation Science

2.2 Standard solutions and endogenous-metabolite-released matrix for calibration standards and QCs Endogenous-metabolite-released plasma for calibration standards and QCs were prepared from rat plasma, which had been stripped of endogenous carbohydrates using neutral decolorizing carbon (Sigma–Aldrich, MO, USA). Charcoal was added to rat plasma at the concentration of 4 g/100 mL. The suspension was stirred at room temperature for 30 min, vortexed for 30 min, and then centrifuged at 4°C, 13 000 rpm for 15 min. The supernatant was transferred to vials to be used as the endogenous-metabolite-released plasma [25, 26]. Stock solutions of the compounds were prepared by dissolving individual compound in pure water with the concentration of 10–100 mM and stored at –80°C until use. The compounds were divided into three groups: A, B and C, depending on their concentration in rat plasma, and standard working solutions were prepared by dilution with 70% acetonitrile at different concentration levels (13 points). Calibration curve were prepared by spiking aliquots of working solutions to plasma treated by a charcoal-stripping technique. The first calibration curve was in the range of 0.001–10 µM, the second within 0.01–100 µM and the third within 0.1–1000 µM. Details on the composition of mixtures A, B, C and the concentrations were given in the Supporting Information Table S1. The QC samples were prepared in charcoal-treated plasma spiked with known amounts of analytes and internal standard at three concentration levels (Table S1). The QC samples were also divided into three groups, including 0.5 µM (LQC), 2 µM (MQC), 10 µM (HQC) in the A group, 5 µM (LQC), 20 µM (MQC), 100 µM (HQC) in the B group, and 50 µM (LQC), 200 µM (MQC), 1000 µM (HQC) in the C group. Lamivudine and L-methionine-13C5, 15N were used as internal standards. This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 8

Journal of Separation Science

2.3 Sample preparation Before analysis, the plasma samples were thawed to room temperature, then, an aliquot of 100 µL plasma sample was precipitated with 500 µL cold ACN/MeOH (50:50, v/v) containing 2 μM lamivudine and 10 μM L-methionine-13C5,

15

N. The mixture was vortexed

for 1 min, placed at –20°C for 20 min, and then centrifuged at 4°C, 13000 rpm for 15 min. 300 µL supernatant was transferred to 96-well plate for LC–MS/MS analysis. 2.4 UHPLC–MS/MS Conditions The UHPLC–MS/MS system consisted of an ACQUITY ultra performance liquid chromatographic system (Waters, Milford, MA, USA) coupled with an API 4000 triple quadrupole mass spectrometer (AB Sciex, Foster City, CA, USA) equipped with a Turbo Ion Source (ESI) operating with negative mode. Chromatographic separation was conducted on an ACQUITY UPLC BEH amide column (2.1mm 100 mm, 1.7 µm particle size) equipped with an ACQUITY UPLC BEH Amide 1.7 µm Van-Guard Pre-column. The column temperature was maintained at 60°C. Mobile phase A consisted of 10 mM ammonium acetate in H2O/acetonitrile (95:5, v/v), and mobile phase B consisted of 10 mM ammonium acetate in acetonitrile/H2O (95:5, v/v). The separation of carbohydrates was achieved using the following gradient program at a flow rate of 250 µL/min over a course of 10 min: started with a linear increase from 2% A to 40% A in 6.0 min, then maintained at 40% A for 1.6 min. The mobile phase was allowed to return to the initial condition within 0.1 min, followed by column re-equilibration for 2.3 min. The injection volume was 10 µL. The compounds of interest were detected by MS/MS in negative ion multiple reaction monitoring (MRM) mode. In the RP-HPLC–MS/MS analysis, the precursor ions of This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 9

Journal of Separation Science

carbohydrates are usually detected as formate or acetate adducts, because of the derivatization process or additive conditions. Yet, in the HILIC–MS/MS analysis, the [M–H]– precursor ions were stronger in the MS response among the possible precursor ions in most analytes, as shown in Table S2 and Figure S1. Therefore, the [M–H]– precursor ions were used for the carbohydrates and their internal standards. Turbo V ion source parameters were common to all analytes: The capillary voltage was –4500 V, and the source temperature was at 550°C. The curtain gas (N2) and collision gas (N2) setting were 30 psi and 10 psi, respectively. The pressure for nebulization gas and vaporization gas setting were 60 psi. The entrance potential (EP) and collision cell exit (CXP) were –50 and –10 V, respectively. The declustering potential (DP) and collision energy were optimized for each analyte. The specific mass transition of each carbohydrate and IS are presented in Table S3. Analyst 1.5 software (AB Sciex, Foster City, CA, USA) was used for data acquisition and processing. 2.5 Method validation The method was validated for linearity, LOD, LOQ, precision, recovery, matrix effect, retention time reproducibility and stability. The analytical curves were examined by using internal standard spiked calibration solutions at 13 concentration levels, ranging from the lower limit of quantification (LLOQ) to the upper limit of quantification (ULOQ) with an injection volume of 10 µL. Integrated peak areas of the selected quantification MRM transitions were used to build the curves. Curves were fitted by a weighted (1/X) least squares regression analysis using the Analyst 1.5 software. LOD and LOQ were calculated based on the S/N of 3:1 and 10:1, respectively. The method precision was evaluated using three different concentration points (LQC, MQC, HQC) This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 10

Journal of Separation Science

that spiked the QC solutions to the treated matrix. Six replicates of each point of QCs were analyzed to determine the intra-day precision. This process was repeated three times over three days to determine the inter-days precision. Precision was expressed as the coefficient of variation (CV) of the determination of QCs. Inter-day precision were calculated similarly for the 18 replicates of each concentration point pooled from the three validation runs. The reproducibility of retention time for each analyte was also evaluated. Recovery was evaluated at low, medium and high QC concentration. The recovery was calculated by dividing the corrected mean peak area of six replicates of each analyte spiked before extraction by that of each analyte spiked after extraction, comparing the peak area ratio of the same concentration of analytes and IS in spiked samples [27]. Matrix effect was evaluated at high QC concentration in six replicates. As carbohydrates in the plasma were endogenous and the endogenous-metabolite-released surrogate matrix of charcoal-stripped plasma was used, the matrix effect was evaluated at two levels: the endogenous plasma sample compared to the endogenous-metabolite-released surrogate matrix, and the endogenous-metabolite-released surrogate matrix compared to the neat standard solution. In details, HQC work solution was spiked in the surrogate matrix, precipitated and got the peak area ratio of the analytes and IS (a); HQC work solution was spiked in the real plasma sample, precipitated and got the peak area ratio of the analytes and IS (b); HQC work solution was spiked in the neat standard solution (standard in 50% acetonitrile/water, v/v), precipitated and got the peak area ratio of the analytes and IS (c); The real plasma sample was precipitated directly, and got the peak area ratio of the analytes and IS, which was the endogenous value of each real plasma (d); The matrix effects were This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 11

Journal of Separation Science

calculated using a modified version of the equation described by Matuszewski et al. [28]: ME1 % = (b – d) / a × 100% (1) ME2 % = a / c × 100% (2) The stability was investigated in six replicates. HQC post-preparative samples were stored at 4°C in the dark for 0, 24, 48, and 72 h from the time of sample preparation. The peak area of each compounds and IS were determined and compared to the freshly prepared solutions (0 h).

3 RESULTS AND DISCUSSION 3.1 MS/MS condition optimization Sensitivity and specific mass spectrometric detection condition were evaluated and optimized. Most of the carbohydrates showed the MS response only in the negative ESI mode. The [M–H]– precursor ions were used for the carbohydrates and their internal standards, because the [M–H]– precursor ions were stronger in the MS response among the possible precursor ions in most analytes, as shown in Table S2 and Figure S1. The declustering potential (DP) and collision energy were important parameters for each compound and the optimized MS parameters are summarized in Table S2 and Table S3. Ion source values optimized to get proper response are described in Section 2. 3.2 Chromatographic condition optimization Carbohydrates might have different structures with the same precursor ion and daughter ion, even the DP and collision energy. It is difficult to differentiate them by mass alone, and UHPLC system coupled with MS was employed to achieve good separation. Columns, the This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 12

Journal of Separation Science

mobile phase composition and column temperature were evaluated and optimized. Several columns including a Phenomenex Gemini C18 column, ACQUITY UPLC HSS T3 column, and XBridge BEH C18 column were evaluated. These columns were found not suitable because of unsatisfactory peak shape, low response, poor separation or even no retention on the column. The newly released ACQUITY UPLC BEH Amide Column (2.1 100 mm, 1.7 µm particle size) is a HILIC column suitable for polar chemicals, the carbohydrates were well retained and separated from each other. Different additives of mobile phase, like formic acid, ammonium hydroxide solution, and ammonium acetate, were evaluated and optimized. Better peak shape and separation were obtained when the concentration of ammonium acetate in mobile phase increased from 0 to 10 mM. But higher concentration of ammonium acetate led to decrease MS response of most of analytes (Fig. S2a). The pH value of mobile phase was optimized by the addition of ammonium hydroxide solution, ammonium acetate or formic acid, respectively. It was observed that the reducing sugars were in double-peak in acidic mobile phase (pH=3.0), while the response of most carbohydrates was decreased in basic environment (pH=9.8; Fig. S2b). Therefore, the nearly neutral mobile phase (pH=6.8–7.0) was finally used. Column temperature is also an important parameter in chromatographic separation. Four temperatures of 30, 40, 50, 60°C were evaluated. The reducing sugars, e.g. glucose, fructose, ribose and fucose failed to present good peak shapes at the column temperature of 30, 40, and 50°C. Nevertheless, when the oven temperature increased to 60°C, the analytes performed good separation, high response, and good peak shapes. The reducing sugars exhibited single peaks, possibly due to alpha and beta anomers This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 13

Journal of Separation Science

formed by mutarotation [29, 30]. The reason for single peaks is the fast interconversion rates as compared to the chromatographic time scale. That means interconversion between alpha and beta anomers is by orders of magnitude faster than the chromatographic equilibria [19, 31]. However, a higher pH value of the eluent resulted in stronger degradation of these carbohydrates [16]. Considering the manufacturer recommendation of pH 2.0–11 and column temperature below 90°C in the use of this BEH Amide column, a column temperature of 60°C in combination with neutral mobile phase would be better for maintaining the column life time. The optimized chromatographic conditions are given in Section 2, with a high throughput of less than 10 min in a run. A typical MRM chromatogram of the analytes spiked in charcoal treated plasma is shown in Fig.1. 3.3 Sample preparation Sample preparation of plasma is a critical step in an accurate and reliable LC–MS/MS assay. The protein was precipitated by organic solvents. The composition of the organic solvents was evaluated and optimized. Cold solvent of ACN/MeOH (50:50, v/v) was used to precipitate the protein and placed at –20°C for 20 min after the vortex of the mixture of solvent and plasma, which was excellent for the preparation of plasma before HILIC–UHPLC–MS/MS. 3.4 Method validation As the analytes in the plasma are endogenous, the removal of the endogenous metabolites was critical for the matrix preparation in the linearity of standards in the quantitative method. A variety of matrices have been widely used for the quantification of endogenous compounds in biological samples, such as artificial matrix. In this assay, charcoal This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 14

Journal of Separation Science

stripped rat plasma was selected as the endogenous-metabolite-released matrix for the preparation of calibration standards [25]. 3.4.1 Linearity and sensitivity After the optimization of analytical conditions, the linearity was studied. The calibration curves showed a good linearity (mostly, correlation coefficient of r > 0.999), where y is the peak area ratio of the analyte to the IS and x is the plasma concentration of the analyte. All back-calculated standards met the criteria of ≤ ±15% deviation from nominal concentration. LODs and LOQs were calculated based on the S/N of 3:1 and 10:1, respectively. The linear range, coefficients of determination, LODs and LOQs of all analytes are summarized in Table 1. 3.4.2 Matrix effect, recovery and precision The matrix effect, extraction recovery, and intra- and inter-day precision of all analytes at three concentration levels were shown in Table 2. The two levels of matrix effect performed well,

ME1

%

(the

endogenous

plasma

sample

compared

to

the

endogenous-metabolite-released surrogate matrix) ranged from 96.0 to 118.0, and ME2 % (the endogenous-metabolite-released surrogate matrix compared to the neat standard solution) ranged from 73.0 to 120.0. The intra- and inter-day precision (CV) ranged from 0.53–7.50 and from 1.16–20.8%, respectively. The recovery values of three concentration levels of QC samples ranged from 67.2–110%. Precision and recovery were all within the acceptable range, indicating that the current method was reliable and reproducible. 3.4.3 Reproducibility of retention time and stability The reproducibility of retention time for the analytes, expressed as CV(%), was less than This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 15

Journal of Separation Science

0.25% in one day and below 1.62% over three days. The stability data fall within a 15% deviation range, suggesting that the stability of post-preparative samples was acceptable under routine laboratory analysis without incurring any significant loss of detection. The corresponding data is presented in Table 3. 3.4.4 Application in rat plasma samples The proposed method was applied in the determination of 18 Sprague–Dawley rat plasma samples. Using this analytical method, we were able to measure the concentration of 11 carbohydrates in all 18 plasma samples. Some of the carbohydrates were previously reported in the quantification of the rat plasma and our data was consistent with the results in the literatures. Some of the carbohydrates were first quantified in the rat plasma. The quantified results of carbohydrates are shown in Table 4. 4 Conclusions In this work, HILIC was coupled with negative ion mode ESI-MS/MS for the simultaneous quantitation of 16 polar metabolites, such as sugars, sugar alcohols and nucleosides in rat plasma. This quantification method performed excellently in the sensitivity, high throughput and simple sample preparation, which met the requirement of quantification in specific expanded metabolomic studies after the global metabolic profiling research.

Acknowledgements This work was financially supported by the National Natural Science Foundation of China (21175092 and 21105064), Shanghai Research Fund for the Post-Doctoral Program (13R21415000), Specially-funded program on the development of national key scientific This article is protected by copyright. All rights reserved.

www.jss-journal.com

instrument

and

Page 16

equipment

(2011YQ150072,

Journal of Separation Science

2011YQ15007204,

2011YQ15007207,

2011YQ15007210), Natural Science Foundation of Shanghai (12ZR1413600), National innovative program (12C26213202212), and Suzhou Specially-funded program on the development of Science and technology (SG201214). The work was also financially supported by Shanghai Ziyuan Pharmaceutical Co., Ltd.

Conflict of interest The authors have declared no conflict of interest.

References [1] Huang, X. X., Zhou, H. Q., Zhang, H., Fish Shellfish Immunol. 2006, 20, 750–757. [2] Neuschwander-Tetri, B. A., Curr. Opin. Clin. Nutr. Metab. Care 2013, 16, 446–452. [3] Ackerman, Z., Oron-Herman, M., Grozovski, M., Rosenthal, T., Pappo, O., Link, G., Sela, B. A., Hypertension 2005, 45, 1012–1018. [4] Suskind, D. L., Wahbeh, G., Gregory, N., Vendettuoli, H., Christie, D., J. Pediatr. Gastroenterol. Nutr. 2014, 58, 87–91. [5] Mussatto, S. I., Mancilha, I. M., Carbohydr. Polym. 2007, 68, 587–597. [6] Dai, Z. W., Leon, C., Feil, R., Lunn, J. E., Delrot, S., Gomes, E., J. Exp. Bot. 2013, 64, 1345–1355. [7] Liao, W., Wei, H., Wang, X., Qiu, Y., Gou, X., Zhang, X., Zhou, M., Wu, J., Wu, T., Kou, F., Zhang, Y., Bian, Z., Xie, G., Jia, W., J. Proteome Res. 2012, 11, 3436–3448. [8] Cheng, Y., Xie, G., Chen, T., Qiu, Y., Zou, X., Zheng, M., Tan, B., Feng, B., Dong, T., This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 17

Journal of Separation Science

He, P., Zhao, L., Zhao, A., Xu, L. X., Zhang, Y., Jia, W., J. Proteome Res. 2012, 11, 1354–1363. [9] Harvey, D. J., J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci. 2011, 879, 1196–1225. [10] Ruiz-Aceituno, L., Carrero-Carralero, C., Ramos, L., Martinez-Castro, I., Sanz, M. L., Anal. Chim. Acta. 2013, 787, 87–92. [11] Becker, M., Zweckmair, T., Forneck, A., Rosenau, T., Potthast, A., Liebner, F., J. Chromatogr. A. 2013, 1281, 115–126. [12] Ruiz-Matute, A. I., Sanz, M. L., Martinez-Castro, I., J. Chromatogr. A. 2007, 1157, 480–483. [13] Albalasmeh, A. A., Berhe, A. A., Ghezzehei, T. A., Carbohydr. Polym. 2013, 97, 253–261. [14] Panagiotopoulos, C., Sempere, R., Lafont, R., Kerherve, P., J. Chromatogr. A. 2001, 920, 13–22. [15] Xi, L., Wang, F., Zhu, Z., Huang, Z., Zhu, Y., Talanta 2014, 119, 440–446. [16] Cataldi, T. R. I., Campa, C., Angelotti, M., Bufo, S. A., J. Chromatogr. A. 1999, 855, 539–550. [17] Dvořáčková, E., Šnóblová, M., Hrdlička, P., J. Sep. Sci. 2014, 37, 323–337. [18] Terol, A., Paredes, E., Maestre, S. E., Prats, S., Todolí, J. L., J. Sep. Sci. 2012, 35, 929–936. [19] Karlsson, G., Winge, S., Sandberg, H., J. Chromatogr. A. 2005, 1092, 246–249. [20] McRae, G., Monreal, C. M., Anal. Bioanal. Chem. 2011, 400, 2205–2215. This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 18

Journal of Separation Science

[21] Wu, X., Jiang, W., Lu, J., Yu, Y., Wu, B., Food Chem. 2014, 145, 976–983. [22] Vilhena, R. D., Pontes, F. L., Marson, B. M., Ribeiro, R. P., Carvalho, K. A., Cardoso, M. A., Pontarolo, R., J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci. 2014, 967C, 41–49. [23] Trivedi, D. K., Iles, R. K., Biomed. Chromatogr. 2014. [24] Yao, X., Zhou, G., Tang, Y., Guo, S., Qian, D., Duan, J. A., Drug Test Anal. 2014. [25] Kinoshita, K., Jingu, S., Yamaguchi, J., Anal. Biochem. 2013, 432, 124–130. [26] Samtani, M. N., Jusko, W. J., Biomed. Chromatogr. 2007, 21, 585–597. [27] Wang, H., Chung-Davidson, Y. W., Li, K., Scott, A. M., Li, W., Talanta 2012, 89, 383–390. [28] Matuszewski, B. K., Constanzer, M. L., Chavez-Eng, C. M., Anal. Chem. 2003, 75, 3019–3030. [29] Brons, C., Olieman, C., J. Chromatogr. 1983, 259, 79–86. [30] Alpert, A. J., Shukla, M., Shukla, A. K., Zieske, L. R., Yuen, S. W., Ferguson, M. A., Mehlert, A., Pauly, M., Orlando, R., J. Chromatogr. A. 1994, 676, 191–122. [31] Hinterwirth, H., Lammerhofer, M., Preinerstorfer, B., Gargano, A., Reischl, R., Bicker, W., Trapp, O., Brecker, L., Lindner, W., J. Sep. Sci. 2010, 33, 3273–3282.

This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 19

Journal of Separation Science

Table 1 Analytical performance of the method Retention

Linearity

Correlation

LOD

Compound

LOQ (nM) Time (min)

range (μM)

Coefficient, r

(nM)

Ribose

3.06

1-1000

0.999

180

600

Adenosine

3.18

0.003-10

0.9994

0.9

3

Erythritol

3.18

0.1-600

0.9993

30

100

Fucose

3.48

0.3-1,000

0.9993

90

300

Inosine

3.8

0.003-3

0.9991

0.3

1

Xylose

3.83

0.6-1,000

0.9999

180

600

Arabitol

3.99

0.06-60

0.9976

18

60

Fructose

4.22

0.3-60

0.9995

90

300

Mannitol

4.66

0.6-100

0.998

120

400

Glucose

4.82

1-1,000

0.9991

300

1,000

Sucrose

5.6

0.6-100

0.999

180

600

Maltose

5.95

0.06-60

0.9991

18

60

Melibiose

6.3

0.1-100

0.9989

30

100

Raffinose

6.65

0.03-100

0.9996

3

10

Maltotriose

6.74

0.03-100

0.9991

6

20

Maltotetraose

7.31

0.06-100

0.9997

7.5

25

This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 20

Journal of Separation Science

Table 2 Matrix effect, recovery and precision of the method at low, medium and high concentrations

Compound

Matrix effect 1 (%)

Precision (CV%)

Precision (CV%)

Intra-day assay (n=6)

Inter-day assay (n=18)

Recovery

Matrix effect 2 (%) Low

Medium

High

Low

Medium

High

Low

Medium

High

Melibiose

107

99.7

92.9

97.3

91.4

0.758

2.36

1.92

3.63

11.1

4.50

Sucrose

117

106

100

92.6

92.2

1.76

2.04

2.87

4.05

5.99

3.15

Maltose

102

108

99.7

99.3

103

1.99

2.80

4.80

6.70

12.6

5.09

Maltotriose

115

99.3

102

98.2

91.3

1.83

2.09

0.877

3.42

3.29

2.35

Maltotetraose

98.1

78.2

72.4

67.7

67.2

0.680

1.48

2.73

2.67

5.98

3.90

Glucose

96.0

101

90.9

89.7

105

1.48

2.03

1.37

13.1

7.70

1.90

Fructose

118

98.9

95.5

75.5

110

2.71

2.07

2.06

8.53

10.7

3.55

Xylose

116

73.0

104

105

102

3.07

1.06

0.890

7.85

6.04

5.74

Ribose

100

81.4

95.8

97.4

102

1.37

2.38

2.35

7.76

9.14

4.70

This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 21

Journal of Separation Science

Raffinose

110

90.0

92.4

93.5

90.5

2.92

2.22

2.24

2.95

6.81

3.11

Fucose

107

106

93.9

90.1

89.7

1.96

1.28

1.29

8.76

9.60

6.96

Inosine

97.4

103

99.9

96.6

101

3.36

7.50

4.70

8.77

10.8

10.0

Adenosine

97.2

101

98.5

94.3

92.9

3.11

0.366

1.45

1.54

2.88

3.35

Mannitol

115

101

100

99.8

95.7

5.51

3.76

1.26

20.8

18.5

2.42

Erythritol

97.5

104

99.3

97.5

96.7

1.51

0.528

0.984

1.78

1.76

1.16

Arabitol

111

120

99.4

90.5

90.4

2.40

0.721

0.755

3.88

2.77

2.34

Matrix effect 1 %, the spiked endogenous plasma sample compared to the spiked endogenous-metabolite-released surrogate matrix. Matrix effect 2 %, the spiked endogenous-metabolite-released surrogate matrix compared to the neat standard solution.

This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 22

Journal of Separation Science

Table 3 The reproducibity of retention time and stability of the sixteen carbohydrates.

Compounds

0 h (%)

Retention time

Retention time

Stability

reproducibility

reproducibility

24 h (%) 48 h (%) 72 h (%)

(Intra-day, n=6, (Inter-day, n=18, CV%)

CV%)

Melibiose

100

102

98.9

100

0.000

0.125

Sucrose

100

101

100

96.9

0.000

0.0987

Maltose

100

104

101

98.2

0.000

0.0866

Maltotriose

100

106

103

104

0.000

0.0885

Maltotetraose

100

97.6

100

97.9

0.000

0.0569

Glucose

100

104

104

102

0.0863

0.208

Fructose

100

104

103

97.8

0.119

0.373

Xylose

100

95

96.7

97.4

0.113

1.62

Ribose

100

106

101

102

0.182

0.417

Raffinose

100

108

104

105

0.000

0.144

Fucose

100

105

103

101

0.250

1.05

Inosine

100

113

105

111

0.139

0.236

Adenosine

100

97.8

101

96.7

0.000

0.263

Mannitol

100

98.0

98.6

96.5

0.119

0.225

Erythritol

100

105

103

101

0.130

0.319

Arabitol

100

102

103

100

0.141

0.324

This article is protected by copyright. All rights reserved.

www.jss-journal.com

Page 23

Table 4 Carbohydrates concentration in SD rat plasma Compounds

Concentration (nM) (n=18)

Melibiose

169±30.8

Sucrose

1937±298

Maltose

1994±228

Maltotriose

BLQ

Maltotetraose

BLQ

Glucose

11,165,000±706,983

Fructose

6103±2065

Xylose

BLQ

Ribose

BLQ

Raffinose

BLQ

Fucose

1894±155

Inosine

60.9±34.8

Adenosine

5.01±3.45

Mannitol

6311±897

Erythritol

1372±92.6

Arabitol

1229±178

Note: BLQ- Below the lower limit of quantitation

This article is protected by copyright. All rights reserved.

Journal of Separation Science

www.jss-journal.com

Page 24

Journal of Separation Science

Figure 1 UHPLC–MS/MS chromatographic traces for 16 compounds and two internal standards on the HILIC Amide column. a. Chromatograms of standards spiked in the plasma. b. Chromatograms of endogenous compounds in the plasma sample. 1) ribose (RT: 3.06 min), 2) xylose (RT: 3.83 min), 3) fructose (RT: 4.22 min), 4) glucose (RT: 4.82 min), 5) maltotetraose (RT: 7.31 min), 6) lamivudine (RT: 2.61 min), 7) adenosine (RT: 3.18 min), 8) inosine (RT: 3.80 min), 9) sucrose (RT: 5.60 min), 10) raffinose (RT: 6.65 min), 11) fucose (RT: 3.48 min), 12) methionine (RT: 4.63 min), 13) melibiose (RT: 6.30 min), 14) maltotriose (RT: 6.74 min), 15) erythritol (RT: 3.18 min), 16) arabitol (RT: 3.99 min), 17) mannitol (RT: 4.66 min), 18) maltose (RT: 5.95 min).

This article is protected by copyright. All rights reserved.

Fast quantification of endogenous carbohydrates in plasma using hydrophilic interaction liquid chromatography coupled with tandem mass spectrometry.

Endogenous carbohydrates in biosamples are frequently highlighted as the most differential metabolites in many metabolomics studies. A simple, fast, s...
759KB Sizes 0 Downloads 13 Views

Recommend Documents