GC-MS Method for the Quantitation of ... - ACS Publications

Article pubs.acs.org/JAFC

GC-MS Method for the Quantitation of Carbohydrate Intermediates in Glycation Systems Sanja Milkovska-Stamenova,†,‡ Rico Schmidt,†,‡ Andrej Frolov,*,†,‡ and Claudia Birkemeyer# †

Institute of Bioanalytical Chemistry, Faculty of Chemistry and Mineralogy, ‡Center for Biotechnology and Biomedicine (BBZ), and Institute of Analytical Chemistry, Faculty of Chemistry and Mineralogy, Universität Leipzig, 04103 Leipzig, Germany

#

S Supporting Information *

ABSTRACT: Glycation is a ubiquitous nonenzymatic reaction of carbonyl compounds with amino groups of peptides and proteins, resulting in the formation of advanced glycation end-products (AGEs) and thereby affecting the properties and quality of thermally processed foods. In this context, mechanisms of the Maillard reaction of proteins need to be understood; that is, glycation products and intermediates (α-dicarbonyls and sugars) need to be characterized. Although the chemical analysis of proteins, peptides, and α-dicarbonyls is well established, sensitive and precise determination of multiple sugars in glycation mixtures is still challenging. This paper presents a gas chromatography−mass spectrometry (GC-MS) method for absolute quantitation of 22 carbohydrates in the model of phosphate-buffered glycation systems. The approach relied on the removal of the phosphate component by polymer-based ion exchange solid phase extraction (SPE) followed by derivatization of carbohydrates and subsequent GC-MS analysis. Thereby, baseline separation for most of the analytes and detection limits of up to 10 fmol were achieved. The method was successfully applied to the analysis of in vitro glycation reactions. Thereby, at least seven sugar-related Maillard reaction intermediates could be identified and quantified. The most abundant reaction product was D-fructose, reaching 2.70 ± 0.12 and 2.38 ± 0.66 mmol/L after 120 min of incubation in the absence and presence of the model peptide, respectively. KEYWORDS: carbohydrates, gas chromatography−mass spectrometry, glycation, phosphate removal, solid phase extraction

■

INTRODUCTION Glycation is a ubiquitous nonenzymatic reaction of carbonyl compounds (reducing sugars and α-dicarbonyls) with amino groups of amino acids, peptides, and proteins.1 Thus, glycation of amino acids is a traditional way to synthesize flavor compounds and is, therefore, often desired and not always accompanied by the formation of toxic derivatives.2 When peptides or proteins are involved in the Maillard reaction, the early glycation products (Amadori and Heyns compounds) undergo degradation, yielding a heterogeneous group of advanced glycation endproducts (AGEs).3−5 Formation of AGEs was observed during thermal processing of foods, even at mild “boiling” conditions.6 In mammals, these AGEs exert a pro-inflammatory effect mediated by macrophage and endothelial receptors of the immunoglobulin family (e.g., receptor for AGEs, RAGE).7 As AGEs are readily absorbed upon oral consumption,8 their presence in food proteins needs to be minimized. For this, the exact mechanisms of AGE formation during heat treatment need to be understood. Taking into account the chemistry of early and advanced glycation and the major pathways of sugar degradation,9 it is obvious that the identification of AGE formation pathways must rely on the analysis of not only protein and peptide modifications6 and reactive α-dicarbonyls as their key intermediates10 but also carbohydrates. Indeed, even simple in vitro glycation systems contained erythrose,11 arabinose,12 and glyceraldehyde13 upon incubation with glucose at 37 °C. Moreover, formation of a pentose-derived AGE pentosidine in incubations with hexoses−glucose, fructose, and ascorbic acid14 clearly indicates the existence of a pentose intermediate. Therefore, the analysis of © XXXX American Chemical Society

carbohydrate patterns in glycation systems (both in vitro and in vivo) is desired. As various sugars represented by multiple isomers are expected to be present in such mixtures,9,12 their chromatographic separation is necessary for the accurate identification and quantitation of individual representatives. It can be accomplished by highperformance anion exchange chromatography (HPAEC)15 and hydrophilic interaction liquid chromatography (HILIC).16 However, as the detection relies on refractive index or evaporative light scattering (ELS), these techniques lack sensitivity. In contrast, fluorescent detection after precolumn derivatization with anthranilic acid,17 as well as electrochemical detection (e.g., pulsed amperometric detection, PAD), ensures superior sensitivity.15 Precolumn derivatization of reducing sugars with 1-naphthylamine using fluorescent and mass spectrometric (MS) detection resulted in further sensitivity improvement.18 In general, GC-MS is the preferred technique for carbohydrate analyses as GC ensures high resolution and sensitivity, whereas electron ionization (EI) fragmentation patterns provide an additional level of analyte annotation.19 Low thermal stability and insufficient volatility of carbohydrates require derivatization of their polar functional groups prior to GC-MS analysis. Derivatization is most often accomplished by trimethylsilylation with hexamethyldisilazane (HMDS) and trimethylchlorosilane (TMCS), which was first introduced by Bentley et al.20 Received: November 27, 2014 Revised: May 12, 2015 Accepted: June 5, 2015

A

DOI: 10.1021/jf505757m J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

Journal of Agricultural and Food Chemistry

buffer (pH 7.4) containing 18 μmol/L iron(II) sulfate in the presence or absence of 25 mmol/L D-glucose and incubated for 0, 30, 60, and 120 min at 95 °C under continuous shaking at 550 rpm. The incubated samples were cooled on ice, allowing the gas phase to condense completely and centrifuged (2 min at 2500g), and the reaction was quenched by the addition of 20 μL of a 6 mmol/L diethylenetriaminepentaacetic acid (DTPA) solution. Afterward, the samples were dried under vacuum and stored at −80 °C prior to GC-EI-MS analysis. The incubations performed in the absence of peptide served as controls. Solid Phase Extraction. Ion exchange cartridges (mixed mode anion/cation exchanger CHROMABOND PSMix, 1 mL/100 mg, Macherey-Nagel), strong anion (CHROMABOND PSOH−, 1 mL/100 mg, Macherey-Nagel), and weak anion exchanger (CHROMABOND HR-XAW, 45 μm, 1 mL/30 mg, Macherey-Nage) were conditioned with 3 mL of methanol and equilibrated with 3 mL of water. After the sample application, elution was performed in five steps: 1 mL of 5% (v/v) aqueous ammonia (E1), 1 mL of 5% (v/v) aqeuous acetic acid (E2), 1 mL of 50% (v/v) aqeuous methanol containing 10% (v/v) ammonia (E3), 1 mL of 50% aqeuous methanol containing 10% acetic acid (E4), and 1 mL of 20% methanol in acetone (E5). The individual fractions were dried under vacuum, and the carbohydrates and phosphate were analyzed by GC-MS after methoximation/trimethylsilylation. For analysis of glycation mixtures, polymer-based mixed mode ion exchange SPE cartridges (CHROMABOND PSMix, 1 mL/100 mg, Macherey-Nagel) were conditioned and equilibrated with 3 mL (3 × 1 mL) of methanol and water, respectively. A 50 μL sample (n = 3) was diluted 20-fold with water and loaded, and carbohydrates were eluted with 10 mL of water. The eluates were lyophilized and reconstituted with 500 μL of 30% aqueous methanol. Eighty-five microliters of each diluted fraction was transferred to a 0.5 mL polypropylene tube, dried under vacuum, and stored at −20 °C prior to derivatization and GC-EI-MS analysis. Method Standardization and Validation. Standardization and validation of the method was performed by a serial dilution (2−2.5-fold increment) of the authentic standard mixture (100 pmol/L−1 mmol/L each, 20 calibration levels, n = 3) in 30% aqueous methanol. The calibration mixtures were dried and derivatized directly or subjected to SPE (when spiked to the glycation matrix) before derivatization as described above. The experimentally determined limits of detection (LODs) were defined as 3.3 SD/m, where m is the slope of the calibration curve and SD is its standard deviation (calculated from the residual standard deviation of the relevant points in LDR).24 The limits of quantitation (LOQs) were defined as the lower limit of the linear dynamic range, but not smaller that S/N = 10/1. The linear dynamic ranges (LDRs) were determined with 12 and 20 calibration levels (n = 3, increment 2.0−2.5) in the presence and absence of matrix, respectively. The accuracies were assessed using three calibration levels within the LDRs considering the SPE procedure. The uncertainty of the analysis was determined at the LOD for each analyte. Derivatization Procedure and GC-MS. The samples were derivatized by employing a modification of the protocol described by Fiehn and co-workers.22 In detail, 30 μL of freshly prepared 20 mg/mL MOA in pyridine solution was added to the dried sample and shaken at 600 rpm for 90 min at 30 °C. Then, 55 μL of MSTFA was added, and the sample was incubated for 30 min at 37 °C under continuous shaking at 600 rpm. The derivatized sample was transferred to a GC autosampler glass vial with micro inserts (NeoLab, Heidelberg, Germany) and subjected to GC-EI-MS analysis. One microliter of sample was injected by an A200S autosampler (CTC Analytics, Zwingen, Switzerland) into a Trace GC Ultra chromatograph equipped with a standard split/splitless injector (Thermo-Fisher Scientific, Bremen, Germany) as described in the Supporting Information (protocol S-1).

N-Trimethylsilylimidazole (TMSI), N-methyl-N-trimethylsilylacetamide (MSA), N-trimethylsilyldiethylamine (TMSDEA), N-trimethylsilyldimethylamine (TMSDMA), N-methyl-N-trimethylsilyltrifluoroacetamide (MSTFA), N,O-bis(trimethylsilyl)acetamide (BSA), and N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) have also been used as silylation agents.21 MSTFA was found to be superior in comparison to other reagents, giving better results for a broader range of analytes and producing fewer byproducts, especially when preceded with alkoxyamine treatment to prevent the formation of cyclic anomers.22 This two-step procedure was further extended to other polar compounds present in multicomponent extracts, being now a gold standard in GC-MS-based metabolomics.23 However, as in vitro glycation mixtures often contain 50−200 mmol/L phosphate, that is, concentrations dramatically exceeding its dynamic range,6,11,12 their analyses by GC-MS can be challenging. Additionally, peptides or proteins present in the mixture might also affect both method and instrument performance. Here we describe a new approach for the analysis of carbohydrates in heated model phosphate-buffered glycation systems containing D-glucose and peptide. The method relied on the removal of the phosphate component by SPE on polymerbased ion exchange cartridges followed by two-step derivatization with methoxyamine hydrochloride (MOA) and MSTFA and subsequent GC-MS analysis.

■

MATERIALS AND METHODS

Chemicals. Fmoc-rinkamide AM resin and Fmoc-L-Arg(Pbf)-OH (peptide synthesis grade) were bought from Iris Biotech GmbH (Marktredwitz, Germany). All other Fmoc-protected L-amino acids (peptide synthesis grade) were from ORPEGEN Pharma (Heidelberg, Germany). N,N′-Dimethylformamide (DMF, ≥99.8%), piperidine (≥99.5%), and dichloromethane (DCM, ≥99.9%) were purchased from Biosolve (Valkenswaard, The Netherlands). Trifluoroacetic acid (≥99.9%), α-D-glucose monohydrate (≥99.5%), D-(−)-fructose (≥99.5%), methanol (HPLC grade, ≥99.9%), 2-propanol (UV/IR grade, ≥99.9%), and D-(+)-sucrose (≥99.5%) were from Carl Roth GmbH & Co (Karlsruhe, Germany). L-(+)-Arabinose (≥99%), D-(+)-maltose monohydrate (≥98%), and D-(+)-mannose (≥98%) were from Merck KGaA (Darmstadt, Germany). MSTFA (≥95%) was bought from Macherey-Nagel (Düren, Germany). D-(+)-Galactose, D-(−)-ribose, D-(−)-mannitol, myo-inositol, diethyl ether (99%), and acetonitrile (≥99.9%, HPLC grade) were bought from VWR International GmbH (Dresden, Germany). 3-Deoxyglucosone (95%; 3-DG) was from Santa Cruz Biotechnology (Heidelberg, Germany). Ammonia solution (∼25% p.a.) was bought from KMF Laborchemie Handles GmbH (Lohmar, Germany). All other chemicals were purchased in required purities from Sigma-Aldrich Chemie GmbH (Taufkirchen, Germany). Water was purified in-house (resistance = 18 mΩ) on a PureLab Ultra Analytic System (ELGA Lab Water, Celle, Germany). Peptide Synthesis. The sequence Ac-AFGSARASGA-NH2 was synthesized on a Syro2000 multiple-peptide synthesizer (MultiSynTech GmbH, Witten, Germany) by 9-fluorenylmethoxycarbonyl/tert-butyl (Fmoc/tBu)-chemistry using 8 equiv (eq) of Fmoc-amino acid derivatives activated with 1,3-diisopropylcarbodiimide/hydroxybenzotriazol (DIC/HOBt) in DMF. N-Terminal acetylation was implemented in the automated synthesis by using 8 eq of acetic acid. After cleavage from the resin and removal of all protecting groups, the peptide was purified by RP-HPLC using 0.1% (v/v) aqueous trifluoroacetic acid (TFA) as ion pair reagent added to water (eluent A) and 60% (v/v) aqueous acetonitrile (eluent B) in 1% eluent B/min gradient (22−42% eluent B in 20 min). The purified peptide was reconstituted in 20% aqueous acetonitrile and split into aliquots of 50 nmol in 0.5 mL polypropylene tubes (Eppendorf AG, Hamburg, Germany). The solvent was removed under reduced pressure, and residues were stored at −20 °C. In Vitro Glycation of the Model Peptide. The dried peptide aliquots were reconstituted in 100 μL of 100 mmol/L sodium phosphate

■

RESULTS GC-EI-MS of Carbohydrate Derivatives. GC-MS analysis of the standard mixture containing 22 TMS-methoxime carbohydrate derivatives resulted in rich signal patterns in the corresponding total ion current (TIC) chromatograms (Figure 1). Individual analytes were typically represented by two well-resolved

B


Article


Figure 1. Total ion current chromatogram of TMS derivatives of carbohydrate methoximes identified in a 25 μmol/L carbohydrate standard mixture.

chromatographic peaks corresponding to syn- and anti-isomers of the TMS-methoximes. All carbohydrates were annotated by tR and their characteristic MS fragmentation patterns. The numbers and intensities of the signals observed in the EI spectra (and, hence, obtained matches in spectra similarity search) were sufficient for unambiguous identification of analytes by spectra similarity (NIST database) or manual interpretation with a reversed match >900. Moreover, the annotation was confirmed by experiments performed with individual authentic standards. Extracted ion chromatograms (XICs) of the most intense and characteristic m/z values (m/z ± 0.5) were integrated at specific retention times for relative and absolute quantitation. When the temperature gradient from 40 to 350 °C was applied, baseline separation could be achieved for the majority of the sugars (Figure 1). However, some sugars, namely, D-ribose, D-ribulose, and D-xylulose (a diastereomer of D-ribulose), showed a strong coelution under these conditions (Figure S-1A). Moreover, the EI spectra of these analytes demonstrated a high similarity and absence of characteristic signals that made distinguishing these sugars impossible (Figure S-1B−D). Unfortunately, further optimization of the GC temperature gradient did not improve the separation of these three analytes. Therefore, in sugar mixtures, these unresolved signals were annotated as “ribose equivalents”. To assess the potential of our method for carbohydrate quantitation, we determined LODs and LOQs as well as LDRs for all analytes (Table S-1). For most of the analytes LODs were between 10 and 500 fmol (only L-fucose showed LOD of 1 pmol). The LOQs were in the range of 50−500 fmol, being higher only for L-rhamnose and L-fucose (1 and 2.5 pmol, respectively). The method showed a good linearity: the LDRs typically covered 3 orders of magnitude, whereas two analytes (D-sucrose and ribitol) produced linear GC-MS response over only 2 orders of magnitude. Only for two sugars, namely, L-rhamnose and L-fucose, the LDRs were 74%, although for three of them, namely, D-mannitol, D-glucose, and D-fructose, they were 63.8, 66.2, and 69.0%, respectively. For the determination of intraday and interday precision, glycation mixtures (n = 5) were incubated for 60 min as described under Materials and Methods on three nonconsecutive days and analyzed by the workflow presented in Figure 5. The method demonstrated a high precision: for concentration determination neither intraday nor interday relative standard deviation (RSD%) exceeded 10%, whereas for retention times these values did not exceed 1% (Table S-2). Quantitation of Sugar Intermediates in Model Glycation Mixtures. The established method was applied for the absolute quantitation of carbohydrate intermediates of the Maillard reaction. For this, D-glucose (25 mmol/L) was incubated in the absence and presence of synthetic peptide as described under Materials and Methods. GC-MS analysis of the glycation mixtures, performed after phosphate removal, revealed the formation of at least eight pentose and hexose monosaccharides present at least at one time point of incubation at concentrations exceeding LOQm, identified by characteristic retention times and similarity of EI mass spectra (Table 2). Peak areas were derived by integration of corresponding extracted ion chromatograms at retention times specific for each sugar. Quantitation relied on external calibration in authentic matrix. For this, the standard mixture was serially diluted

■

DISCUSSION Model reactions are known to be a convenient tool for the study of the Maillard reaction in protein systems. Typically, to E


Article


Figure 5. Experimental workflow for carbohydrate analysis.

challenging task. Indeed, the differences in the structures of isomeric sugars are obviously too small to affect the chromatographic behavior of the analytes, and in this study, that was the case for D-ribose, D-ribulose, and D-xylulose (Figure S-1A). These three sugars could not be separated under any chromatographic conditions tested, without loss of selectivity in other sections of the chromatogram. Although Medeiros and Simoneit29 recently separated ribose and xylulose TMS derivatives obtained without a methoximation step on a 5% phenyl-methylpolysiloxane column, their method suffered from coelution of fucose and xylose, galactose, and altrose, as well as arabitol and xylitol. Application of liquid chromatography (LC)-based techniques (LC-MS) resulted in compromised selectivity30 and sensitivity,31 although it can be successfully applied for the analysis of specific compound groups (e.g., dicarbonyls).32 Remarkably, the LODs observed in this study for some monosaccharides (e.g., 1, 3, 7, 11, 12, 13, 15, and 16) and for all oligosaccharides (18−22) were lower than those observed with LC-MS,33 GC-flame ionization detection (FID),34 GC-MS,23 and GC-MS/MS35 methods. The values obtained here were also advantageous in comparison to those obtained with soft ionization techniques: ESI-MS,36 MELDI-MS,37 and MALDI-TOF-MS.38 The lower method LODs and LOQs in comparison to the corresponding instrument values can be explained by different drying procedures: freeze-drying versus Speedvac. Additionally, reconstitution of dried SPE eluates in a low volume of derivatization mixture provided further increase of sensitivity due to the concentration effect (Tables 1 and S-1). Thus, in terms of both analyte coverage and selectivity, our method is, to the best of our knowledge, superior in comparison to those reported before. The sensitivity improvement can be also attributed to the double-focusing sector field mass analyzer used here, which is known for its high sensitivity. Another challenge of multicomponent analysis is the interference between analytes and sample matrix (i.e., undefined compounds behind the scope of analysis) directly affecting

Table 2. GC-MS Parameters of D-Glucose and the Main Products Formed during Its Thermal Degradationa tR (min) analyte D-glucose D-mannose D-fructose 3-DG myo-inositol arabinose D-xylose a D-ribose

peak 1

peak 2

25.2 24.9 24.7 22.9 28.2 21.1 21.0 21.4

25.4 24.9 23.6

MS data m/z (rel intensity, %) 160 (48), 205 (97), 217 (42), 319 (100) 160 (47), 205 (99), 217 (42), 319 (100) 103 (100), 205 (12), 217 (55), 307 (57) 205 (10), 217 (7), 231 (100), 315 (4) 191 (40), 217 (97), 305 (100), 318 (45) 103 (100), 205 (13), 217 (50), 307 (47) 103 (100), 205 (14), 217 (49), 307 (44) 103 (100), 205 (31), 217 (45), 263 (7), 307 (42)

a D-Ribose, D-ribulose, and D-xylulose were quantified as D-ribose equivalents. For all analytes, the boldfaced selective m/z and peak 1 were used for quantitation; 3-DG, 3-deoxyglucosone.

characterize the patterns of glycoxidative protein modifications, biologically relevant polypeptides, such as collagen,25 bovine or human serum albumins (BSA and HSA, respectively),26 were used. However, due to the large number of potential glycation sites, such mixtures are often too complex for efficient analysis of reaction pathways and mechanisms. Therefore, in the past decades, model peptide-based systems for the glycation of amino groups with selected glycation agents were established.27 Recently we have described an advanced glycation model system based on arginine-containing peptides.6 This approach provided adequate information about the peptide modifications and the kinetics of their formation. However, to reveal the underlying reaction mechanisms, this information needs to be complemented with knowledge about α-dicarbonyl and sugar profiles. As methods for comprehensive α-dicarbonyl analysis are well-established and reported elsewhere,28 here we focused on carbohydrate profiling. In the whole workflow of carbohydrate analysis, complete separation of all monosaccharides seems to be the most F


Article

Journal of Agricultural and Food Chemistry Table 3. Concentrations of D-Glucose and Main Sugar Glycation Intermediates at Each Incubation Time Point concentrations at each reaction time pointa (mmol/L) 0 min analyte D-glucose D-mannose D-fructose d

3-DG myo-inositol arabinose D-xylose e D-ribose

control b

25 blq blq blq blq blq blq blq

30 min

Glc +Pepta

60 min

120 min

control

Glc+Pepta

control

Glc+Pepta

control

Glc+Pepta

blq 0.64 ± 0.05 (2.5)c blq blq 0.47 ± 0.04 (1.9)c blq blq



0.17 ± 0.04 (0.7)c 1.22 ± 0.71 (4.9)c blq blq 0.46 ± 0.29 (1.8)c blq blq

0.50 ± 0.04 (2.0)c 2.70 ± 0.12 (10.8)c blq blq 0.76 ± 0.09 (3.1)c blq 0.22 ± 0.03 (0.9)c

0.30 ± 0.07 (1.2)c 2.38 ± 0.66 (9.5)c blq blq 0.84 ± 0.06 (3.4)c blq blq

b

25 blq blq blq blq blq blq blq

D-Glucose was heated in 100 mmol/L phosphate buffer (pH 7.4) containing 18 μmol/L FeSO4 in the absence and presence of synthetic peptide Ac-AFGSARASGA-NH2 at 95 °C during 2 h. The incubated samples were subjected to mixed mode anion/cation exchange (PSMix) SPE, derivatized with MOA/MSTFA, and analyzed by GC-EI-MS. bInitial D-glucose concentration. cPercentage of the initial glucose concentration. Data are expressed as mean ± SD; blq, below the limit of quantitation. d3-DG, 3-deoxyglucosone. eD-Ribose, D-ribulose, and D-xylulose were quantified as D-ribose equivalents. a

High analyte recoveries and method precision allowed the application of our approach to in vitro glycation reactions performed with arginine-containing peptide.6 As early glycation was not described for arginine side chains yet, this peptide can be subjected exclusively to advanced glycation, that is, interaction with α-dicarbonyls, first of all glyoxal and methylglyoxal.41 According to our results, the detected intermediates of dicarbonyl formation were mostly hexoses and pentoses (Table 2). In most cases, their origin can be explained by known pathways of the protein Maillard reaction or simple interconversion of monosaccharides (e.g., Lobry de Bruyn− Alberda van Ekenstein rearrangement driven by proton abstraction).42 Thus, mannose and 3-deoxyglucosone can be generated from glucose via 1,2-endiol intermediate.42 As the presence of the peptide did not influence fructose concentrations in the reaction mixture (Table 3), it can be concluded that this monosaccharide was generated solely from free glucose. Most probably, mannose was rather reactive toward the model peptide or its degradation was catalyzed by the peptide: its abundance was much lower when incubations were performed in the presence of peptide (Table 3). The formation of myo-inositol (even in trace amounts) apparently identified (Figure S-4) under the conditions of incubation seems to be confusing. In animals, this polyol is synthesized from glucose-6-phosphate in two enzymatic steps.43 However, its nonenzymatic generation, to the best of our knowledge, was not described so far. Generally, pentoses are well-known intermediates of protein Maillard reaction and typically generated in glucose-based glycation mixtures via retroaldolization.44 All of them, especially aldoses, are potent glycation agents and readily react with free amino groups.45 It can be assumed that, although ribose is much less abundant in the reaction mixtures (Table 3), it has an essentially higher reactivity in comparison to arabinose (represented here, most probably, with a D-isomer (Figure S-3)). Indeed, the concentration of ribose is significantly decreased in the presence of peptide, which might indicate its enhanced autoxidation. This enhancement, in turn, can be explained by the shift of chemical equilibrium due to the reaction of sugar autoxidation products with arginine residues. The same was observed for mannose, which is known to have a relatively low glycation potential. In conclusion, our method allowed simultaneous quantitation of carbohydrates present in model glycation systems. Gas chromatographic analyses of sugars after two-step derivatization

analyte signal intensities. These events, generally denoted as matrix effects, result in signal enhancement or suppression and represent thereby an essential challenge in instrumental analyses.39 Thus, this phenomenon was described for sugars: for example, glucose was reported to enhance the signals of lysine and fumaric and citric acids in plant samples.19 Here, we observed that sugar-derived signals themselves could be suppressed in the presence of high (100 mmol/L) phosphate concentrations (Figure 2). Indeed, as this inorganic anion can be readily derivatized with MSTFA and converted into its volatile 3-TMS ether, it is involved in retention and separation, and, when present in concentrations exceeding column capacity, might interfere with analytes. Commonly, suppression effects within the GC-MS system can be explained by the formation of active sites in the analytical system.40 Typically, the injector is the preferred part where these active sites can emerge. This might occur when a highly abundant analyte does not undergo quantitative derivatization and, due to compromised volatility, deposits in the liner. In such cases, subsequent injections with MSTFA as the solvent can furthermore provoke an in situ derivatization of these deposited remainders yielding irreproducible memory peaks in the corresponding chromatograms. Indeed, we observed such signals in MSTFA blank runs analyzed after the samples overloaded with phosphate (data not shown). This clearly indicated that formation of active sites could be responsible for the observed suppression effect (Figure 2). Obviously, removal of phosphate from the samples before derivatization was the only way to escape this problem. Typically, SPE is the method of choice for this. As both analytes (i.e., sugars) and contaminating phosphate anions are polar substances, our strategy relied on the use of ion exchange cartridges. Surprisingly, a comparison of the three tested materials, that is, mixed mode anion/cation exchanger (PS-Mix), strong anion (PS-OH−), and weak anion (HR-XAW) exchangers, applied to four carbohydrates dissolved in 100 mmol/L sodium phosphate buffer, revealed a consistently better recovery from the PS-Mix material in comparison to the other two phases (Figure 3). This improvement related to additional PS-H+ functionality of the PSMix cartridges most probably can be attributed to the binding of iron and sodium cations, removing them from the mobile phase, where they otherwise could induce dipoles in the carbohydrate molecules that may lead to a more anionic character of the carbohydrates increasing the interaction with the PSOH− column material. G


Article

Journal of Agricultural and Food Chemistry DCM Fmoc/tBu TFA DIC/HOBt DTPA LODs S/N LOQs TMS SD RSD TIC XICs LDR BSA HSA LC FID FT SC

with MOA and MSTFA showed very good performance and can be applied to the analysis of carbohydrate patterns in Maillard reaction mixtures. The method provided high sensitivity and precision for more than 10 relevant carbohydrates. After removal of >99% phosphate with mixed-mode ion exchange SPE, at least seven sugars (Maillard reaction intermediates) were identified by GC-MS and reproducibly quantified (RSD < 10%) using the peak areas of the corresponding extracted ion chromatograms. The method for the analysis of carbohydrates in model glycation mixtures established here can be used to get better insights into the mechanisms of protein Maillard reaction and pathways of AGE formation.

■

ASSOCIATED CONTENT

S Supporting Information *

Additional protocols, figures, tables, and references. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jf505757m.

■

AUTHOR INFORMATION

■

Corresponding Author

*(A.F.) Mail: Institut für Bioanalytische Chemie, BiotechnologischBiomedizinisches Zentrum, Deutscher Platz 5, 04103 Leipzig, Germany. Phone: +49 (0) 341 9731332. Fax: +49 (0) 341 9731339. E-mail: [email protected].

dichloromethane 9-fluorenylmethoxycarbonyl/tert-butyl trifluoroacetic acid 3-diisopropylcarbodiimide/hydroxybenzotriazol diethylenetriaminepentaacetic acid limits of detection signal/noise limits of quantitation trimethylsilyl standard deviation relative standard deviation total ion current extracted ion chromatograms linear dynamic ranges bovine serum albumin human serum albumin liquid chromatography flame ionization detection flow through sugars-only control

REFERENCES

(1) Ulrich, P.; Cerami, A. Protein glycation, diabetes, and aging. Recent Prog. Horm. Res. 2001, 56, 1−21. (2) van Boekel, M. A. J. S. Formation of flavour compounds in the Maillard reaction. Biotechnol. Adv. 2006, 24, 230−233. (3) Hodge, J. E. The Amadori rearrangement. Adv. Carbohydr. Chem. 1955, 10, 169−205. (4) Heyns, K.; Noack, H. Die Umsetzung von D-Fructose mit L-Lysin und L-Arginin und deren Beziehung zu nichtenzymatischen Bräunungsreaktionen. Chem. Ber. 1962, 95, 720−727. (5) Grillo, M. A.; Colombatto, S. Advanced glycation end-products (AGEs): involvement in aging and in neurodegenerative diseases. Amino Acids 2008, 35, 29−36. (6) Frolov, A.; Schmidt, R.; Spiller, S.; Greifenhagen, U.; Hoffmann, R. Arginine-derived advanced glycation end products generated in peptideglucose mixtures during boiling. J. Agric. Food Chem. 2014, 62, 3626− 3635. (7) Huttunen, H. J.; Fages, C.; Rauvala, H. Receptor for advanced glycation end products (RAGE)-mediated neurite outgrowth and activation of NF-κB require the cytoplasmic domain of the receptor but different downstream signaling pathways. J. Biol. Chem. 1999, 274, 19919−19924. (8) Koschinsky, T.; He, C. J.; Mitsuhashi, T.; Bucala, R.; Liu, C.; Buenting, C.; Heitmann, K.; Vlassara, H. Orally absorbed reactive glycation products (glycotoxins): an environmental risk factor in diabetic nephropathy. Proc. Natl. Acad. Sci. U. S. A. 1997, 94, 6474− 6479. (9) Smuda, M.; Glomb, M. A. Fragmentation pathways during Maillard-induced carbohydrate degradation. J. Agric. Food Chem. 2013, 61, 10198−10208. (10) Gobert, J.; Glomb, M. A. Degradation of glucose: reinvestigation of reactive α-dicarbonyl compounds. J. Agric. Food Chem. 2009, 57, 8591−8597. (11) Thornalley, P. J.; Langborg, A.; Minhas, H. S. Formation of glyoxal, methylglyoxal and 3-deoxyglucosone in the glycation of proteins by glucose. Biochem. J. 1999, 344 (Part 1), 109−116. (12) Wells-Knecht, K. J.; Zyzak, D. V.; Litchfield, J. E.; Thorpe, S. R.; Baynes, J. W. Mechanism of autoxidative glycosylation: identification of glyoxal and arabinose as intermediates in the autoxidative modification of proteins by glucose. Biochemistry 1995, 34, 3702−3709. (13) Usui, T.; Yoshino, M.; Watanabe, H.; Hayase, F. Determination of glyceraldehyde formed in glucose degradation and glycation. Biosci., Biotechnol., Biochem. 2007, 71, 2162−2168. (14) Grandhee, S. K.; Monnier, V. M. Mechanism of formation of the Maillard protein cross-link pentosidine. Glucose, fructose, and ascorbate as pentosidine precursors. J. Biol. Chem. 1991, 266, 11649−11653.

Funding

We thank Prof. Dr. Ralf Hoffmann for financial support. We gratefully acknowledge financial support from the Deutsche Forschungsgemeinschaft (DFG, FR-3117/2-1 and HO-2222/7-1), the European Fund for Regional Structure Development (EFRE, European Union and Free State Saxony, and the Bundesministerium für Bildung and Forschung (BMBF) to R.H. and C.B. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Prof. Dr. Ralf Hoffmann for a critical reading of the manuscript and helpful discussions, as well as Antje Hutschenreuther for initial experiments and Divya Varadharajan for assistance in sample preparation.

■

ABBREVIATIONS USED AGEs advanced glycation end-products GC-MS gas chromatography−mass spectrometry SPE solid phase extraction HPAEC high-performance anion exchange chromatography HILIC hydrophilic interaction liquid chromatography ELS evaporative light scattering PAD pulsed amperometric detection MS mass spectrometry EI electron ionization HMDS hexamethyldisilazane TMCS trimethylchlorosilane TMSI N-trimethylsilylimidazole MSA N-methyl-N-trimethylsilylacetamide TMSDEA N-trimethylsilyldiethylamine TMSDMA N-trimethylsilyldimethylamine MSTFA N-methyl-N-trimethylsilyltrifluoroacetamide BSA N,O-bis(trimethylsilyl)acetamide BSTFA N,O-bis(trimethylsilyl) trifluoroacetamide MOA methoxyamine hydrochloride DMF dimethylformamide H


Article


degradation of sugars. Biosci., Biotechnol., Biochem. 2007, 71, 2465− 2472. (33) Henning, C.; Liehr, K.; Girndt, M.; Ulrich, C.; Glomb, M. A. Extending the spectrum of α-dicarbonyl compounds in vivo. J. Biol. Chem. 2014, 289, 28676−28688. (34) Ciucanu, I.; Swallow, K. C.; Căpriţa,̆ R. Micro-solid phase extraction with helical-solid-sorbent in the presence of organic solvent for gas chromatography−mass spectrometry analysis of per-Omethylated mono- and disaccharides. Anal. Chim. Acta 2004, 519, 93−101. (35) Gómez-González, S.; Ruiz-Jiménez, J.; Priego-Capote, F.; Luque de Castro, M. a. D. Qualitative and quantitative sugar profiling in olive fruits, leaves, and stems by gas chromatography−tandem mass spectrometry (GC-MS/MS) after ultrasound-assisted leaching. J. Agric. Food Chem. 2010, 58, 12292−12299. (36) Wunschel, D. S.; Fox, K. F.; Fox, A.; Nagpal, M. L.; Kim, K.; Stewart, G. C.; Shahgholi, M. Quantitative analysis of neutral and acidic sugars in whole bacterial cell hydrolysates using high-performance anion-exchange liquid chromatography−electrospray ionization tandem mass spectrometry. J. Chromatogr., A 1997, 776, 205−219. (37) Qureshi, M. N.; Stecher, G.; Sultana, T.; Abel, G.; Popp, M.; Bonn, G. K. Determination of carbohydrates in medicinal plants − comparison between TLC, mf-MELDI-MS and GC-MS. Phytochem. Anal. 2011, 22, 296−302. (38) Grant, G. A.; Frison, S. L.; Yeung, J.; Vasanthan, T.; Sporns, P. Comparison of MALDI-TOF mass spectrometric to enzyme colorimetric quantification of glucose from enzyme-hydrolyzed starch. J. Agric. Food Chem. 2003, 51, 6137−6144. (39) Hutschenreuther, A.; Kiontke, A.; Birkenmeier, G.; Birkemeyer, C. Comparison of extraction conditions and normalization approaches for cellular metabolomics of adherent growing cells with GC-MS. Anal. Methods 2012, 4, 1953−1963. (40) Poole, C. F. Matrix-induced response enhancement in pesticide residue analysis by gas chromatography. J. Chromatogr., A 2007, 1158, 241−250. (41) Wolff, S. P.; Dean, R. T. Aldehydes and dicarbonyls in nonenzymic glycosylation of proteins. Biochem. J. 1988, 249, 618−619. (42) Martins, S. I.; Marcelis, A. T.; van Boekel, M. A. Kinetic modelling of Amadori N-(1-deoxy-D-fructos-1-yl)-glycine degradation pathways. Part I − reaction mechanism. Carbohydr. Res. 2003, 338, 1651−1663. (43) Holub, B. J. Metabolism and function of myo-inositol and inositol phospholipids. Annu. Rev. Nutr. 1986, 6, 563−597. (44) Kerler, J.; Winkel, C.; Davidek, T.; Blank, I. Basic chemistry and process conditions for reaction flavours with particular focus on Maillard-type reactions. In Food Flavour Technology; Wiley-Blackwell: Chichester, UK, 2010; pp 51−88. (45) Syrovy, I. Glycation of albumin: reaction with glucose, fructose, galactose, ribose or glyceraldehyde measured using four methods. J. Biochem. Biophys. Methods 1994, 28, 115−121.

(15) Cataldi, T. R. I.; Campa, C.; De Benedetto, G. E. Carbohydrate analysis by high-performance anion-exchange chromatography with pulsed amperometric detection: the potential is still growing. Fresenius' J. Anal. Chem. 2000, 368, 739−758. (16) Buszewski, B.; Noga, S. Hydrophilic interaction liquid chromatography (HILIC) − a powerful separation technique. Anal. Bioanal. Chem. 2012, 402, 231−247. (17) Anumula, K. R.; Dhume, S. T. High resolution and high sensitivity methods for oligosaccharide mapping and characterization by normal phase high performance liquid chromatography following derivatization with highly fluorescent anthranilic acid. Glycobiology 1998, 8, 685−694. (18) Rakete, S.; Glomb, M. A. A novel approach for the quantitation of carbohydrates in mash, wort, and beer with RP-HPLC using 1naphthylamine for precolumn derivatization. J. Agric. Food Chem. 2013, 61, 3828−3833. (19) Koek, M. M.; Jellema, R. H.; van der Greef, J.; Tas, A. C.; Hankemeier, T. Quantitative metabolomics based on gas chromatography mass spectrometry: status and perspectives. Metabolomics 2011, 7, 307−328. (20) Bentley, R.; Sweeley, C. C.; Makita, M.; Wells, W. W. Gas chromatography of sugars and other polyhydroxy compounds. Biochem. Biophys. Res. Commun. 1963, 11, 14−18. (21) Ruiz-Matute, A. I.; Hernandez-Hernandez, O.; RodriguezSanchez, S.; Sanz, M. L.; Martinez-Castro, I. Derivatization of carbohydrates for GC and GC-MS analyses. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci. 2011, 879, 1226−1240. (22) Fiehn, O.; Kopka, J.; Trethewey, R. N.; Willmitzer, L. Identification of uncommon plant metabolites based on calculation of elemental compositions using gas chromatography and quadrupole mass spectrometry. Anal. Chem. 2000, 72, 3573−3580. (23) Koek, M. M.; Muilwijk, B.; van der Werf, M. J.; Hankemeier, T. Microbial metabolomics with gas chromatography/mass spectrometry. Anal. Chem. 2006, 78, 1272−1281. (24) International Conference on the Harmonization of Technical Requirements for the Registration of Pharmaceuticals for Human Use; ICH harmonized tripartite guideline validation of analytical procedures: text and methodology Q2 (R1), 2005. Current Step 4 version, Parent Guideline dated 27 October 1994 (Complementary Guideline on Methodology dated 6 November 1996 incorporated in November 2005). (25) Fu, M. X.; Wells-Knecht, K. J.; Blackledge, J. A.; Lyons, T. J.; Thorpe, S. R.; Baynes, J. W. Glycation, glycoxidation, and cross-linking of collagen by glucose. Kinetics, mechanisms, and inhibition of late stages of the Maillard reaction. Diabetes 1994, 43, 676−683. (26) Frolov, A.; Hoffmann, R. Analysis of amadori peptides enriched by boronic acid affinity chromatography. Ann. N. Y. Acad. Sci. 2008, 1126, 253−256. (27) Ehrlich, H.; Hanke, T.; Frolov, A.; Langrock, T.; Hoffmann, R.; Fischer, C.; Schwarzenbolz, U.; Henle, T.; Born, R.; Worch, H. Modification of collagen in vitro with respect to formation of Nεcarboxymethyllysine. Int. J. Biol. Macromol. 2009, 44, 51−56. (28) Gensberger, S.; Glomb, M. A.; Pischetsrieder, M. Analysis of sugar degradation products with α-dicarbonyl structure in carbonated soft drinks by UHPLC-DAD-MS/MS. J. Agric. Food Chem. 2013, 61, 10238−10245. (29) Medeiros, P. M.; Simoneit, B. R. Analysis of sugars in environmental samples by gas chromatography-mass spectrometry. J. Chromatogr., A 2007, 1141, 271−278. (30) Sevcik, R. S.; Mowery, R. A.; Becker, C.; Chambliss, C. K. Rapid analysis of carbohydrates in aqueous extracts and hydrolysates of biomass using a carbonate-modified anion-exchange column. J. Chromatogr., A 2011, 1218, 1236−1243. (31) Fox, A. Chapter 23: A current perspective on analysis of sugar monomers using GC-MS and GC-MS/MS. In Journal of Chromatography Library; Ziad El, R., Ed.; Elsevier: Amsterdam, The Netherlands, 2002; Vol. 66, pp 829−843. (32) Usui, T.; Yanagisawa, S.; Ohguchi, M.; Yoshino, M.; Kawabata, R.; Kishimoto, J.; Arai, Y.; Aida, K.; Watanabe, H.; Hayase, F. Identification and determination of alpha-dicarbonyl compounds formed in the I


GC-MS Method for the Quantitation of ... - ACS Publications

Recommend Documents