- Email: [email protected]
Detection of glycation sites in proteins by high-resolution mass spectrometry combined with isotopic labeling
Analytical Biochemistry 400 (2010) 237–243
Contents lists available at ScienceDirect
Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio
Detection of glycation sites in proteins by high-resolution mass spectrometry combined with isotopic labeling Piotr Stefanowicz *, Monika Kijewska, Alicja Kluczyk, Zbigniew Szewczuk Faculty of Chemistry, University of Wrocław, 50-137 Wrocław, Poland
a r t i c l e
i n f o
Article history: Received 3 December 2009 Received in revised form 1 February 2010 Accepted 10 February 2010 Available online 13 February 2010 Keywords: Glycation Mass spectrometry Isotopic labeling
a b s t r a c t The products of nonenzymatic glycation of proteins are formed in a chemical reaction between reducing sugars and the free amino group located either at the N terminus of the polypeptide chain or in the lysine side chain. Glycated proteins and their fragments could be used as markers of the aging process as well as diabetes mellitus and Alzheimer’s disease, making them an object of interest in clinical chemistry. In this article, we propose a new method for the identification of peptide-derived Amadori products in the mixtures obtained by enzymatic hydrolysis of glycated proteins. Two proteins, ubiquitin and human serum albumin (HSA), were modified with an equimolar mixture of glucose and [13C6]glucose and were subjected to enzymatic hydrolysis. The obtained enzymatic digests were analyzed by high-resolution mass spectrometry (HRMS), and the peptide-derived Amadori products were identified on the basis of specific isotopic patterns resulting from 13C substitution. The number of glycated peptides in the digest of HSA detected by our procedure was in agreement with the data recently reported in the literature. Ó 2010 Elsevier Inc. All rights reserved.
Reducing sugars, mainly glucose, interact under physiological conditions, with the amino groups of proteins forming products of Amadori rearrangement [1]. The concentration of these compounds increases in diabetes [2]; therefore, glycated proteins and their fragments are useful markers of high blood concentration of glucose. Currently, glycated hemoglobin and, to some extent, glycated serum albumin are used as markers of diabetes mellitus [3,4]. Systematic studies on the glycation of blood proteins are vital for understanding the mechanisms of diabetic complications. Fragments of glycated proteins could also be used in new diagnostic procedures. Enzymatic hydrolysis of glycated proteins produces a complex mixture of peptides. Peptide-derived Amadori products are minor components of such mixtures, making their detection difficult. Recently, three strategies allowing detection of peptide-derived Amadori products in the mixtures obtained by enzymatic hydrolysis of proteins from more complex systems, such as serum and plasma, have been published. The first approach is based on preconcentration of glycated peptides using affinity chromatography on borate columns [5]. The fractions enriched in Amadori products are consequently analyzed by liquid chromatography–mass spec-
* Corresponding author. E-mail address: [email protected] (P. Stefanowicz). 0003-2697/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2010.02.011
trometry (LC–MS).1 This method is suitable for the analysis of complex multiprotein systems, and its usefulness was confirmed in proteomic studies [6]. The second approach uses the characteristic neutral losses of Amadori products in tandem mass spectrometry (MS/MS) analysis. At relatively low collision energy, glycated peptides eliminate a series of water molecules and formaldehyde, forming stable furylium (neutral loss of 84 Da) and pyrylium (neutral loss of 54 Da) ions [7–9]. These fragments are produced from [MH]+ ions as well as [MH2]2+ and [MH3]3+ ions. This feature of Amadori products allows their selective detection by neutral loss scanning on a triple quadrupole instrument [9] or by using equivalent procedures on quadrupole time-of-flight (Q-TOF) instruments [10,11]. The boronic acid affinity enrichment of the sample can be combined with MS/MS neutral loss-based procedures. The third method is based on direct application of borate buffer to MS analysis of peptide-derived Amadori products. Complexation of the glycated peptide by the borate anion may be used to distinguish glycated and nonglycated peptides in the digest; moreover, the complex stabilizes the hexose moiety attached to the lysine
1 Abbreviations used: LC–MS, liquid chromatography–mass spectrometry; MS/MS, tandem mass spectrometry; Q-TOF, quadrupole time-of-flight; CID, collision-induced dissociation; ESI/MALDI, electrospray ionization/matrix-assisted laser desorption/ ionization; HSA, human serum albumin; TFA, trifluoroacetic acid; HPLC, highperformance liquid chromatography; UV, ultraviolet; FT–ICR, Fourier transform ion cyclotron resonance; SP, solid phase; HRMS, high-resolution mass spectrometry; AGE, advanced glycation end product.
238
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
side chain, reducing the neutral losses and significantly simplifying the collision-induced dissociation (CID) spectrum [12]. In this work, we propose an alternative method of identification of glycated peptides in multicomponent mixtures. A sample of protein is glycated using an equimolar mixture of glucose and [13C6]glucose and then is hydrolyzed by the selected proteolytic enzyme. This procedure produces a mixture of nonglycated and glycated peptides. Amadori products can be recognized on the basis of their unique isotopic distribution resulting from 13C substitution. The search for the glycated peptides can be performed automatically using, for example, a simple Excel spreadsheet recognizing the pairs of peaks with equal abundance and defined distance (e.g., 6 m/z units for +1 ion, 3 m/z units for +2 ion). A similar method for identification of the products of protein modification was proposed several years ago by the McLafferty group [13]. Such an approach, based on isotopic labeling for the identification of cross-linked protein fragments, was further developed by the Muller group [14] and the Sinz group [15,16].
Materials and methods Reagents All reagents, solvents, and ubiquitin (isolated from bovine blood cells) were purchased from Sigma–Aldrich. The reagents were used without further purification. Mass spectrometry All MS experiments were performed on an Apex-Qe Ultra 7T instrument (Bruker Daltonics) equipped with a dual electrospray ionization/matrix-assisted laser desorption/ionization (ESI/MALDI) source. The instrument was operated in the positive ion mode and calibrated with the Tunemix mixture (Bruker Daltonics). The mass accuracy was better than 5 ppm. The obtained mass spectra were analyzed using Biotools software (Bruker Daltonics). The instrumental parameters were as follows: scan range, 300–2500 m/z; drying gas, nitrogen; temperature of drying gas, 200 °C; potential between spray needle and orifice, set at 4.5 kV; source accumulation time, 0.5 s; and ion accumulation time, 0.5 s. In MS/MS mode, the precursor ions were selected in the quadrupole collision cell and subsequently fragmented in the hexapole collision cell. Argon was used as a collision gas. The obtained fragment ions were directed to the ICR mass analyzer and registered as an MS/MS spectrum. The collision energy in the hexapole collision cell was set at 20 V. The samples for MS and MS/MS experiments (0.05 mg) were dissolved in 1 ml of an acetonitrile/water mixture containing formic acid (50:50:0.1, v/v/v). Glycation of proteins Samples of proteins were glycated according to the Boratyn´ski method [17,18] using an equimolar mixture of glucose with natural isotopic composition and [13C6]glucose. Briefly, the selected protein and the mixture of glucose and [13C6]glucose were dissolved in pure water. The solution did not contain any buffering additives, but the buffering capacity of protein itself was sufficient to maintain pH in the range 6.0–7.0. The sample was lyophilized. The dry lyophilizate was then placed in an air oven at 80 °C for 25 min (according to the method described by us previously [19]). The molar ratio of protein to glucose was 1:10 for ubiquitin and 1:370 for human serum albumin (HSA).
Enzymatic hydrolysis The enzymatic hydrolysis of glycated ubiquitin was performed according the procedure described previously [19]. Glycated ubiquitin (1 mg) was dissolved in 5% aqueous formic acid (100 ll), and a 10-ll aliquot of stock solution of pepsin (1 mg/ml in water) was added. The enzyme/substrate mass ratio was 1:100. The reaction mixture was incubated for 36 h at 22 °C. The resulting digest was lyophilized and used for MS experiments. The enzymatic hydrolysis of glycated HSA was performed according to the modified procedure described by Lapolla and coworkers [20]. Briefly, the glycated HSA (3 mg) was dissolved in NH4HCO3 buffer solution (3 ml, 10 mM, pH 8.0). After the addition of dithiothreitol (45 ll and 45 mM in water), the mixture was incubated at 50 °C for 15 min. The final concentration of dithiothreitol in the mixture was approximately 0.67 mM. Then the 100-ll aliquot of trypsin stock solution (1.5 mg and 500 ll in water) was added to obtain the 1:10 enzyme/substrate mass ratio. The reaction mixture was incubated for 24 h at 37 °C. The reaction was quenched by the addition of 10% aqueous trifluoroacetic acid (TFA, 240 ll). The resulting digest was lyophilized and used for MS experiments. High-performance liquid chromatographyAnalytical high-performance liquid chromatography (HPLC) was carried out on a Thermo Separation HPLC system using a Vydac RP C18 column (4.6 250 mm) with ultraviolet (UV) detection at 220 nm. The flow rate was 1 ml/min, and the gradient was 0–80% B in A in 40 min, where A is water containing 0.1% TFA and B is acetonitrile containing 0.1% TFA. For the partial separation of tryptic digest of HSA, the time-dependent fraction collection was applied. Data analysis The mass list generated by the Data Analysis program (Bruker Daltonics) was analyzed using an Excel spreadsheet to find the pairs of ions for which the mass difference equals 6.0201 ± 10 ppm and the intensity is the same (±10%). The analyses were performed separately for ions of different charges. Results and discussion The experiments were performed on two proteins: ubiquitin and HSA. Ubiquitin was chosen because of its relatively simple structure and low molecular mass so as to simplify the characterization of the glycated material before the enzymatic hydrolysis. As described in Materials and methods, the glycated protein was dissolved in an acetonitrile/water mixture containing formic acid and was subsequently analyzed on an Apex Ultra Fourier transform ion cyclotron resonance (FT–ICR) mass spectrometer equipped with an ESI source. The deconvoluted spectrum is presented in Fig. 1. The results indicate the glycation level, which corresponds to approximately 2 hexose moieties per protein molecule. In addition, the isotopic patterns presented in the Fig. 1 insets confirm that the protein was glycated by an equimolar mixture of glucose and [13C6]glucose. The enzymatic hydrolysis in the presence of pepsin was carried out in acidic solution (5% aqueous formic acid). In our experience, peptide-derived Amadori products are relatively stable at acidic pH. We previously reported [21,22] the solid phase (SP) method for synthesis of Amadori products, which were removed from the SP resin using a mixture of TFA and water (95:5, v/v). Even in such harsh conditions, the reaction products were stable for at least 24 h. The glycated ubiquitin was hydrolyzed in the presence of pepsin, and the mixture of peptic fragments was directly analyzed on an FT–MS spectrometer without chromatographic separation
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
239
Fig. 1. ESI–MS deconvoluted spectrum of glycated ubiquitin.
Fig. 2. ESI–MS spectrum of the hydrolysate of glycated ubiquitin. The symbols *, ** and *** mark fragments of spectrum which are presented in expanded form as inserts.
240
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
Table 1 Glycated peptides identified by HRMS analyses of unseparated products of peptic hydrolysis of glycated ubiquitin.
a b
Glycated peptide
Charge state
Mass (calculated/found)a
Modified amino acid
[59–67] [5–15] [4–15] [5–15]b [59–69] [46–58] [46–59] [44–58] [25–40] [25–41] [25–40]b [25–43] [4–23] or [3–22] [25–45] [25–45]b [22–45]
1+, 1+, 1+, 2+ 1+, 2+ 2+ 2+, 2+, 2+ 2+ 2+, 2+ 2+, 2+, 3+
1256.6136/1256.6184 1335.7861/1335.7914 1482.8545/1482.8603 1497.8389/1497.8458 1506.7566/1506.7631 1550.7424/1550.7500 1713.8057/1713.8142 1810.8949/1810.9072 1941.0055/1941.0154 2069.0641/2069.0722 2103.0583/2103.0694 2338.2492/2338.2596 2353.2516/2353.2478 2598.4017/2598.4116 2760.4545/2760.4628 2941.5761/2941.5864
K63 K6 or K11 K6 or K11 K6 or K11 K63 K48 K48 K48 K27, K29, or K27, K29, or K27, K29, or K27, K29, or K6 or K11 K27, K29, or K27, K29, or K27, K29, or
2+ 2+ 2+ 2+
3+ 3+
3+, 4+ 3+, 4+ 3+, 4+
K33 K33 K33 K33 K33 K33 K33
Monoisotopic mass. Fragments containing two hexose moieties.
(see Fig. 2). The peaks corresponding to the glycated peptides were identified by their specific isotopic distribution. The signals of Amadori products can be found by either visual inspection of the spectrum or automatic data analysis (see Materials and methods). The identified glycated fragments of ubiquitin are presented in Table 1. Our procedure detected 16 glycated fragments. In the obtained ESI–MS spectrum, the following ubiquitin monoglycated fragments were found: [59–67], [5–15], [4–15], [59–69], [46–58], [46–59], [44–58], [25–40], [25–41], [25–43], [4–23], [3–22], [25–45], and [22–45]. The glycated peptides were accom-
panied by their doubly glycated analogues in three cases: [5–15], [25–45], and [25–40]. The large number of identified glycated fragments suggests that the thermal glycation of ubiquitin is relatively unspecific and that the majority of lysine side chains are glycated to some extent even when the total glycation level of ubiquitin is moderate. This result contradicts the prediction obtained using the NetGlycate 1.0 server [23], which suggested that only Lys11 and Lys29 are susceptible to nonenzymatic glycation. One possible explanation for this discrepancy is that the data used as a training set for the sequence-based
Fig. 3. ESI–MS spectrum of glycated HSA hydrolysate (fraction 4). The symbols * and ** mark fragments of spectrum which are presented in expanded form as inserts.
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
predictor included both in vivo and in vitro data, and the in vitro glycation was performed at conditions preserving the native structure of proteins [23]. The high-temperature glycation used by us may alter the protein structure and, consequently, may result in the lower specificity of glycation. A similar protocol was applied to the HSA sample. Glycated protein was treated with dithiothreitol to reduce the disulfide bridges and was subjected to tryptic hydrolysis. For the enzymatic digestion, we selected trypsin, which cleaves the protein chain after Lys residues that are also the subject of glycation. This may be considered as a disadvantage because it impedes attempts at comparing concentrations of modified and unmodified peptides to get an idea of the degree of glycation at a given site. On the other hand, a recent article compared the applications of various enzymes in the proteomic analysis of glycation sites in plasma proteins. Three enzymes were taken into consideration: trypsin, Arg-C, and Lys-C. The highest number of glycated peptides was identified in the case of trypsin [24]. One of the main reasons for the selection of trypsin was the possibility of a direct comparison of our results with those published previously for HSA by Lapolla and coworkers [20] and Zhang and coworkers [24].
241
The unseparated products of enzymatic hydrolysis of HSA were directly analyzed by an ESI–FT–ICR mass spectrometer, but the number of detected glycated products was low (only 11 peptides); therefore, we decided to partially fractionate the enzymatic digest by HPLC (according to the procedure given in Materials and methods), and further experiments were performed on the fractions collected from the whole tryptic digest of HSA. The obtained spectra were analyzed automatically using an Excel spreadsheet. A representative example of the digest fraction spectrum is presented in Fig. 3. This spectrum, recorded for fraction 4 (retention time of 15–16 min), shows several signals, with a distinct isotopic pattern, corresponding to glycated peptides. The combined results of the analysis of HPLC fractions 1–15 are presented in Table 2. In this experiment, 45 glycated peptides were identified. This number is similar to that recently reported for HSA [19,25], confirming the usefulness of our method. The HSA sequence coverage is 75% according to the list of modified peptides. The glycated Lys residues are shown in Fig. 4. We also noticed some additional pairs of peaks with the characteristic shift of 6.0201 Da and comparable intensity; however, we were not able to assign these signals to the tryptic fragments of HSA. These peaks (e.g., the pair of peaks at m/z
Table 2 Glycated peptides identified by HRMS analyses of fractionated products of tryptic hydrolysis of glycated HSA.
a b
Glycated peptide
Retention time (min)
Charge state
Mass (calculated/found)a
Modified amino acid
[1–10] [433–444] [429–436] [198–205] [467–475] [182–190] [314–323] [535–545] [182–195] [561–574] [182–197] [182–197]b [200–209] [11–20] [349–359] [223–233] [198–209] [198–209]b [263–276] [219–233] [94–114] [137–144] [525–534] [187–197] [522–534] [414–428] [411–428] [187–197]b [175–190] [485–500] [373–389] [318–337] [501–521] [65–81] [115–137] [226–240] [542–557] [287–317] [223–240]b [219–240]b [373–402] [445–472] [223–240] [446–472] [21–51]
11–12 11–12 13–14 13–14 13–14 13–14 13–14 13–14 13–14 13–14 13–14 13–14 15–16 15–16 13–14 15–16 15–16 15–16 15–16 15–16 15–16 17–18 17–18 17–18 17–18 17–18 17–18 17–18 19 19 20 20 20 21 21 22 22 22 23 23 24 24 24 25 25
2+ 2+ 1+ 1+ 2+ 2+ 1+, 2+ 2+, 2+ 2+, 2+, 1+, 1+, 2+ 2+ 2+ 2+, 2+, 3+ 3+, 2+ 2+ 2+ 3+ 2+ 3+ 2+ 3+ 3+ 2+ 3+ 3+ 2+ 3+, 2+, 3+ 3+, 3+ 3+, 3+, 4+ 3+ 4+ 4+
1310.6215/1310.6272 1447.6799/1447.6872 963.5237/963.5296 1051.5583/1051.5656 1163.6034/1163.6092 1235.5881/1235.5932 1301.5809/1301.5872 1468.8249/1468.8352 1679.8214/1679.8288 1731.7509/1731.7582 1963.9811/1963.9892 2126.0339/2126.0452 1299.6129/1299.6192 1387.6507/1387.6572 1457.7501/1457.7572 1413.7028/1413.7092 1540.7919/1540.8012 1702.8448/1702.8528 1788.8451/1788.8538 1897.9785/1897.9878 2740.2511/2740.2664 1216.6340/1216.6392 1289.7442/1289.7502 1337.6423/1337.6512 1658.9818/1658.9898 1800.9833/1800.9932 2193.1893/2193.1957 1499.6951/1499.7015 1931.9874/1931.9958 2014.9558/2014.9608 2206.1409/2206.1492 2617.2840/2617.2888 2649.1905/2649.2058 2036.0612/2036.0732 2882.3830/2882.3930 1811.9404/1811.9478 2001.9605/2001.9708 3523.5681/3523.5768 2346.2094/2346.2170 2830.4852/2830.4958 3787.8551/3787.8678 3376.6465/3376.6594 2184.1566/2184.1638 3220.5454/3220.5584 3724.9071/3724.9184
K4 K436 or K439 K432 K199 R472 R186 K317 K536, K538, or K541 K190 K564 or K573 K190 or K195 K190 and K195 K205 K12 K351 K225 K199 or K205 K199 and K205 K274 K225 K106 K137 K525 K190 or K195 K524 or K525 K414 K413 or K414 K190 and K195 K181 R485 K378 K322 K519 K73 K136 K233 K545 K313 K225 and K233 K225 and K233 K378 or K389 K466 K225 or K233 K466 K41
Monoisotopic mass. Fragments containing two hexose moieties.
2+ 3+ 3+ 3+ 2+ 2+
3+ 3+ 4+
4+, 5+ 3+ 4+ 4+ 4+
242
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
Fig. 4. Scheme of HSA sequence showing the glycated Lys residues.
756.86 and 759.87 labelled with symbol [X] in Fig. 3) were subjected to the CID analysis (Fig. 5). The observed fragment ions demonstrate the pattern of neutral losses characteristic for Amadori products. The ion representing the isotopically labeled glycated
peptide eliminates the formaldehyde containing 13C. The presence of Amadori products that do not correspond to the predicted tryptic fragments of HSA may be explained by the incomplete reduction of disulfide bridges, additional posttranslational
Fig. 5. (A) ESI–MS spectrum of monoglycated unidentified peptide. (B and C) CID MS/MS fragmentation of nonlabeled (B) and labeled (C) ions of glycated peptide.
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
modifications of albumin [26], and/or partial chymotrypsin-like specificity of trypsin used for the digestion. We also compared the data obtained using our procedure with those reported previously for both in vivo and in vitro HSA glycation by Garlick and Mazer [27]. The authors indicated several specific lysine and arginine modification sites in glycated albumin; however, the most predominant site of nonenzymatic glycation in HSA in vivo is Lys525. Later results showed that Lys199, Lys281, Lys439, Lys525, Lys233, Lys317, Lys351, Lys12, and Lys534 are also susceptible to nonenzymatic glycation. In another report [28], four lysine residues (Lys199, Lys281, Lys439, and Lys525) were identified as the main glycation sites, although approximately 33% of the overall glycation occurs at Lys525. Recently, the additional modifications at Lys51, Lys159, Lys205, Lys286, and Lys538, as well as at Arg160, Arg222, and Arg472, were reported on the basis of MALDI–TOF analysis of in vitro minimally glycated HSA [29]. The number of lysine moieties that underwent the Amadori rearrangement in our experiment seems to be higher when compared with that of those experimentally detected in previous reports or calculated using the NetGlycate 1.0 server [23]. The fragments detected in our studies, covering nearly the whole sequence of HSA, demonstrate the low specificity of the thermal glycation. The presented data clearly prove that high-resolution mass spectrometry (HRMS) combined with isotopic labeling can be applied directly to the analysis of the glycation of proteins containing approximately 102 amino acid residues, whereas the more complex systems may require an LC–MS procedure. Although our experiments were performed on the instrument equipped with an ESI source, we believe that the MALDI source can also be used. The proposed method was applied in searching for products of Amadori rearrangement, the early stage of the glycation process. The existing methods, based on characteristic neutral loss patterns, could be successfully used for detection of peptide-derived Amadori products even in complex mixtures. However, the approach based on stable isotope labeling described here is not limited to products of Amadori rearrangement. The early glycation products undergo further chemical modifications, forming numerous advanced glycation end products (AGEs). The diversity of chemical structures of AGEs makes their detection difficult, and this is also complicated by the lack of a general method of preconcentration of AGEs. Theoretically, the method developed by us could also detect the unknown products of reaction of glucose (or glucose degradation products) with proteins. In general, the isotopic pattern characteristic of peaks corresponding to the product of reaction with glucose/[13C]glucose should be accompanied by specific neutral losses in an MS/MS experiment. Otherwise, the identified ion may represent an AGE; therefore, our procedure allows the detection of potential AGEs without prior information on their structure. The described method is suitable for model studies in vitro. The experiment in vivo would be more difficult and expensive because of the problems with the introduction of [13C]glucose at the concentration required for the in vivo glycation.
Conclusion We have presented a convenient and straightforward HRMS procedure for selective detection of protein glycation sites and peptide-derived Amadori products based on isotopic labeling during glycation with an equimolar mixture of glucose and [13C6]glucose. This new technique may also be useful for detection of AGEs in complex mixtures (e.g., enzymatic digests).
243
References [1] M. Amadori, Products of condensation between glucose and p-phenetidine I, Atti Accad. Nazl. Lincei 2 (1925) 337–345. [2] S.R. Thorp, J.W. Baynes, Role of the Maillard reaction in diabetes mellitus and diseases of aging, Drugs Aging 9 (1996) 69–77. [3] C. Rohlfing, M.H. Wiedmeyer, R. Little, Biological variation of glycohemoglobin, Clin. Chem. 48 (2002) 1116–1118. [4] D.M. Nathan, H. Turgeon, S. Regan, Relationship between glycated hemoglobin levels and mean glucose levels over time, Diabetologia 50 (2007) 2239–2244. [5] A. Frolov, R. Hoffmann, Analysis of Amadori peptides enriched by boronic acid affinity chromatography, Ann. NY Acad. Sci. 1126 (2008) 253–256. [6] Q. Zhang, N. Tang, J.W.C. Brock, H.M. Mottaz, J.M. Ames, J.W. Baynes, R.D. Smith, T.O. Metz, Affinity chromatography coupled with electron transfer dissociation mass spectrometry, J. Proteome Res. 6 (2007) 2323–2330. [7] I. Jeric, C. Versluis, S. Horvat, A.J.R. Heck, Tracing glycoprotein structures: electron ionization tandem mass spectrometric analysis of sugar–peptide adducts, J. Mass Spectrom. 37 (2002) 803–811. [8] P. Stefanowicz, J. Boratyn´ski, U. Kan´ska, I. Petry, Z. Szewczuk, Evaluation of high temperature glycation of proteins and peptides by electrospray ionization mass spectrometry, Acta Biochim. Pol. 48 (2001) 1137–1141. [9] A. Frolov, P. Hoffmann, R. Hoffmann, Fragmentation behavior of glycated peptides derived from D-glucose, D-fructose, and D-ribose in tandem mass spectrometry, J. Mass Spectrom. 41 (2006) 1459–1469. [10] M. Sniatynski, J.C. Rogalski, M.D. Hoffman, J. Kast, Correlation and convolution analysis of peptide mass spectra, Anal. Chem. 78 (2006) 2600–2607. [11] M.D. Hoffman, M.J. Sniatyn´ski, J.C. Rogalski, J.C. Le Blanc, J. Kast, Multiple neutral loss monitoring (MNM): a multiplexed method for post-translational modification screening, J. Am. Soc. Mass Spectrom. 17 (2006) 307–317. [12] M. Kijewska, A. Kluczyk, P. Stefanowicz, Z. Szewczuk, Electrospray ionization mass spectrometric analysis of complexes between peptide-derived Amadori products and borate ions, Rapid Commun. Mass Spectrom. 23 (2009) 4038– 4046. [13] N.L. Kelleher, R.B. Nicewonger, T.P. Begley, F.W. McLafferty, Identification of modification sites in large biomolecules by stable isotope labeling and tandem high resolution mass spectrometry: the active site nucleophile of thiaminase I, J. Biol. Chem. 27 (1997) 32215–32220. [14] D.R. Muller, P. Schindler, H. Towbin, U. Wirth, H. Voshol, S. Hoving, M.O. Steinmetz, Isotope-tagged cross-linking reagents: a new tool in mass spectrometric protein interaction analysis, Anal. Chem. 73 (2001) 1927–1934. [15] C. Ihling, A. Schmidt, S. Kalkhof, D.M. Schulz, C. Stingl, K. Mechtler, M. Haack, A.G. Beck-Sickinger, D.M. Cooper, A. Sinz, Isotope-labeled cross-linkers and Fourier transform ion cyclotron resonance mass spectrometry for structural analysis of a protein/peptide complex, J. Am. Soc. Mass Spectrom. 17 (2006) 1100–1113. [16] A. Sinz, Isotope-labeled photoaffinity reagents and mass spectrometry to identify protein–ligand interactions, Angew. Chem. Intl. Ed. 46 (2007) 660– 662. [17] J. Boratyn´ski, Dry reaction of proteins with carbohydrates at 120 °C yields neoglycoconjugates, J. Biotechnol. Tech. 12 (1998) 707–710. [18] J. Boratyn´ski, R. Roy, High temperature conjugation of proteins with carbohydrates, Glycoconj. J. 15 (1998) 131–138. [19] P. Stefanowicz, M. Kijewska, Z. Szewczuk, Sequencing of peptide-derived Amadori products by the electron capture dissociation method, J. Mass Spectrom. 44 (2009) 1047–1052. [20] A. Lapolla, D. Pedele, R. Reitano, N.C. Arico, Enzymatic digestion and mass spectrometry in the study of advanced glycation end products/peptides, J. Am. Soc. Mass Spectrom. 15 (2004) 496–509. [21] P. Stefanowicz, K. Kapczyn´ska, A. Kluczyk, Z. Szewczuk, A new procedure for the synthesis of peptide-derived Amadori products on a solid support, Tetrahedron Lett. 48 (2007) 967–969. [22] P. Stefanowicz, M. Kijewska, K. Kapczyn´ska, Z. Szewczuk, Methods of the siteselective solid phase synthesis of peptide-derived Amadori products, Amino Acids (in press), doi: 10.1007/s00726-009-0294-z. [23] M.B. Johansen, L. Kiemer, S. Brunak, Analysis and prediction of mammalian protein glycation, Glycobiology 16 (2006) 844–853. [24] Q. Zhang, A.A. Schepmoes, J.W.C. Brock, S. Wu, R.J. Moore, S.O. Purvine, J.W. Baynes, R.D. Smith, T.O. Metz, Improved methods for the enrichment and analysis of glycated peptides, Anal. Chem. 80 (2008) 9822–9829. [25] F.L. Brancia, J.Z. Bereszczak, A. Lapolla, D. Fedele, L. Baccarin, R. Seraglia, P. Traldi, Comprehensive analysis of glycated human serum albumin tryptic peptides by off-line liquid chromatography followed by MALDI analysis on a time-of-flight/curved field reflectron tandem mass spectrometer, J. Mass Spectrom. 41 (2006) 1179–1185. [26] D. Bar-Or, R. Bar-Or, L.T. Rael, D.K. Gardner, D.S. Slone, M.L. Craun, Heterogeneity and oxidation status of commercial human albumin preparations in clinical use, Crit. Care Med. 33 (2005) 1638–1641. [27] R.L. Garlick, J.S. Mazer, The principal site of non-enzymatic glycosylation of human serum albumin in vivo, J. Biol. Chem. 258 (1983) 6142–6146. [28] N. Iberg, R. Fluckiger, Nonezymatic glycosylation of albumin in vivo: identification of multiple glycosylated sites, J. Biol. Chem. 261 (1986) 13542–13545. [29] C. Wa, R.L. Cerny, W.A. Clarke, D.S. Hage, Characterization of glycation adducts on human serum albumin by matrix-assisted laser desorption/ionization timeof-flight mass spectrometry, Clin. Chim. Acta 385 (2007) 48–60.
Contents lists available at ScienceDirect
Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio
Detection of glycation sites in proteins by high-resolution mass spectrometry combined with isotopic labeling Piotr Stefanowicz *, Monika Kijewska, Alicja Kluczyk, Zbigniew Szewczuk Faculty of Chemistry, University of Wrocław, 50-137 Wrocław, Poland
a r t i c l e
i n f o
Article history: Received 3 December 2009 Received in revised form 1 February 2010 Accepted 10 February 2010 Available online 13 February 2010 Keywords: Glycation Mass spectrometry Isotopic labeling
a b s t r a c t The products of nonenzymatic glycation of proteins are formed in a chemical reaction between reducing sugars and the free amino group located either at the N terminus of the polypeptide chain or in the lysine side chain. Glycated proteins and their fragments could be used as markers of the aging process as well as diabetes mellitus and Alzheimer’s disease, making them an object of interest in clinical chemistry. In this article, we propose a new method for the identification of peptide-derived Amadori products in the mixtures obtained by enzymatic hydrolysis of glycated proteins. Two proteins, ubiquitin and human serum albumin (HSA), were modified with an equimolar mixture of glucose and [13C6]glucose and were subjected to enzymatic hydrolysis. The obtained enzymatic digests were analyzed by high-resolution mass spectrometry (HRMS), and the peptide-derived Amadori products were identified on the basis of specific isotopic patterns resulting from 13C substitution. The number of glycated peptides in the digest of HSA detected by our procedure was in agreement with the data recently reported in the literature. Ó 2010 Elsevier Inc. All rights reserved.
Reducing sugars, mainly glucose, interact under physiological conditions, with the amino groups of proteins forming products of Amadori rearrangement [1]. The concentration of these compounds increases in diabetes [2]; therefore, glycated proteins and their fragments are useful markers of high blood concentration of glucose. Currently, glycated hemoglobin and, to some extent, glycated serum albumin are used as markers of diabetes mellitus [3,4]. Systematic studies on the glycation of blood proteins are vital for understanding the mechanisms of diabetic complications. Fragments of glycated proteins could also be used in new diagnostic procedures. Enzymatic hydrolysis of glycated proteins produces a complex mixture of peptides. Peptide-derived Amadori products are minor components of such mixtures, making their detection difficult. Recently, three strategies allowing detection of peptide-derived Amadori products in the mixtures obtained by enzymatic hydrolysis of proteins from more complex systems, such as serum and plasma, have been published. The first approach is based on preconcentration of glycated peptides using affinity chromatography on borate columns [5]. The fractions enriched in Amadori products are consequently analyzed by liquid chromatography–mass spec-
* Corresponding author. E-mail address: [email protected] (P. Stefanowicz). 0003-2697/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2010.02.011
trometry (LC–MS).1 This method is suitable for the analysis of complex multiprotein systems, and its usefulness was confirmed in proteomic studies [6]. The second approach uses the characteristic neutral losses of Amadori products in tandem mass spectrometry (MS/MS) analysis. At relatively low collision energy, glycated peptides eliminate a series of water molecules and formaldehyde, forming stable furylium (neutral loss of 84 Da) and pyrylium (neutral loss of 54 Da) ions [7–9]. These fragments are produced from [MH]+ ions as well as [MH2]2+ and [MH3]3+ ions. This feature of Amadori products allows their selective detection by neutral loss scanning on a triple quadrupole instrument [9] or by using equivalent procedures on quadrupole time-of-flight (Q-TOF) instruments [10,11]. The boronic acid affinity enrichment of the sample can be combined with MS/MS neutral loss-based procedures. The third method is based on direct application of borate buffer to MS analysis of peptide-derived Amadori products. Complexation of the glycated peptide by the borate anion may be used to distinguish glycated and nonglycated peptides in the digest; moreover, the complex stabilizes the hexose moiety attached to the lysine
1 Abbreviations used: LC–MS, liquid chromatography–mass spectrometry; MS/MS, tandem mass spectrometry; Q-TOF, quadrupole time-of-flight; CID, collision-induced dissociation; ESI/MALDI, electrospray ionization/matrix-assisted laser desorption/ ionization; HSA, human serum albumin; TFA, trifluoroacetic acid; HPLC, highperformance liquid chromatography; UV, ultraviolet; FT–ICR, Fourier transform ion cyclotron resonance; SP, solid phase; HRMS, high-resolution mass spectrometry; AGE, advanced glycation end product.
238
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
side chain, reducing the neutral losses and significantly simplifying the collision-induced dissociation (CID) spectrum [12]. In this work, we propose an alternative method of identification of glycated peptides in multicomponent mixtures. A sample of protein is glycated using an equimolar mixture of glucose and [13C6]glucose and then is hydrolyzed by the selected proteolytic enzyme. This procedure produces a mixture of nonglycated and glycated peptides. Amadori products can be recognized on the basis of their unique isotopic distribution resulting from 13C substitution. The search for the glycated peptides can be performed automatically using, for example, a simple Excel spreadsheet recognizing the pairs of peaks with equal abundance and defined distance (e.g., 6 m/z units for +1 ion, 3 m/z units for +2 ion). A similar method for identification of the products of protein modification was proposed several years ago by the McLafferty group [13]. Such an approach, based on isotopic labeling for the identification of cross-linked protein fragments, was further developed by the Muller group [14] and the Sinz group [15,16].
Materials and methods Reagents All reagents, solvents, and ubiquitin (isolated from bovine blood cells) were purchased from Sigma–Aldrich. The reagents were used without further purification. Mass spectrometry All MS experiments were performed on an Apex-Qe Ultra 7T instrument (Bruker Daltonics) equipped with a dual electrospray ionization/matrix-assisted laser desorption/ionization (ESI/MALDI) source. The instrument was operated in the positive ion mode and calibrated with the Tunemix mixture (Bruker Daltonics). The mass accuracy was better than 5 ppm. The obtained mass spectra were analyzed using Biotools software (Bruker Daltonics). The instrumental parameters were as follows: scan range, 300–2500 m/z; drying gas, nitrogen; temperature of drying gas, 200 °C; potential between spray needle and orifice, set at 4.5 kV; source accumulation time, 0.5 s; and ion accumulation time, 0.5 s. In MS/MS mode, the precursor ions were selected in the quadrupole collision cell and subsequently fragmented in the hexapole collision cell. Argon was used as a collision gas. The obtained fragment ions were directed to the ICR mass analyzer and registered as an MS/MS spectrum. The collision energy in the hexapole collision cell was set at 20 V. The samples for MS and MS/MS experiments (0.05 mg) were dissolved in 1 ml of an acetonitrile/water mixture containing formic acid (50:50:0.1, v/v/v). Glycation of proteins Samples of proteins were glycated according to the Boratyn´ski method [17,18] using an equimolar mixture of glucose with natural isotopic composition and [13C6]glucose. Briefly, the selected protein and the mixture of glucose and [13C6]glucose were dissolved in pure water. The solution did not contain any buffering additives, but the buffering capacity of protein itself was sufficient to maintain pH in the range 6.0–7.0. The sample was lyophilized. The dry lyophilizate was then placed in an air oven at 80 °C for 25 min (according to the method described by us previously [19]). The molar ratio of protein to glucose was 1:10 for ubiquitin and 1:370 for human serum albumin (HSA).
Enzymatic hydrolysis The enzymatic hydrolysis of glycated ubiquitin was performed according the procedure described previously [19]. Glycated ubiquitin (1 mg) was dissolved in 5% aqueous formic acid (100 ll), and a 10-ll aliquot of stock solution of pepsin (1 mg/ml in water) was added. The enzyme/substrate mass ratio was 1:100. The reaction mixture was incubated for 36 h at 22 °C. The resulting digest was lyophilized and used for MS experiments. The enzymatic hydrolysis of glycated HSA was performed according to the modified procedure described by Lapolla and coworkers [20]. Briefly, the glycated HSA (3 mg) was dissolved in NH4HCO3 buffer solution (3 ml, 10 mM, pH 8.0). After the addition of dithiothreitol (45 ll and 45 mM in water), the mixture was incubated at 50 °C for 15 min. The final concentration of dithiothreitol in the mixture was approximately 0.67 mM. Then the 100-ll aliquot of trypsin stock solution (1.5 mg and 500 ll in water) was added to obtain the 1:10 enzyme/substrate mass ratio. The reaction mixture was incubated for 24 h at 37 °C. The reaction was quenched by the addition of 10% aqueous trifluoroacetic acid (TFA, 240 ll). The resulting digest was lyophilized and used for MS experiments. High-performance liquid chromatographyAnalytical high-performance liquid chromatography (HPLC) was carried out on a Thermo Separation HPLC system using a Vydac RP C18 column (4.6 250 mm) with ultraviolet (UV) detection at 220 nm. The flow rate was 1 ml/min, and the gradient was 0–80% B in A in 40 min, where A is water containing 0.1% TFA and B is acetonitrile containing 0.1% TFA. For the partial separation of tryptic digest of HSA, the time-dependent fraction collection was applied. Data analysis The mass list generated by the Data Analysis program (Bruker Daltonics) was analyzed using an Excel spreadsheet to find the pairs of ions for which the mass difference equals 6.0201 ± 10 ppm and the intensity is the same (±10%). The analyses were performed separately for ions of different charges. Results and discussion The experiments were performed on two proteins: ubiquitin and HSA. Ubiquitin was chosen because of its relatively simple structure and low molecular mass so as to simplify the characterization of the glycated material before the enzymatic hydrolysis. As described in Materials and methods, the glycated protein was dissolved in an acetonitrile/water mixture containing formic acid and was subsequently analyzed on an Apex Ultra Fourier transform ion cyclotron resonance (FT–ICR) mass spectrometer equipped with an ESI source. The deconvoluted spectrum is presented in Fig. 1. The results indicate the glycation level, which corresponds to approximately 2 hexose moieties per protein molecule. In addition, the isotopic patterns presented in the Fig. 1 insets confirm that the protein was glycated by an equimolar mixture of glucose and [13C6]glucose. The enzymatic hydrolysis in the presence of pepsin was carried out in acidic solution (5% aqueous formic acid). In our experience, peptide-derived Amadori products are relatively stable at acidic pH. We previously reported [21,22] the solid phase (SP) method for synthesis of Amadori products, which were removed from the SP resin using a mixture of TFA and water (95:5, v/v). Even in such harsh conditions, the reaction products were stable for at least 24 h. The glycated ubiquitin was hydrolyzed in the presence of pepsin, and the mixture of peptic fragments was directly analyzed on an FT–MS spectrometer without chromatographic separation
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
239
Fig. 1. ESI–MS deconvoluted spectrum of glycated ubiquitin.
Fig. 2. ESI–MS spectrum of the hydrolysate of glycated ubiquitin. The symbols *, ** and *** mark fragments of spectrum which are presented in expanded form as inserts.
240
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
Table 1 Glycated peptides identified by HRMS analyses of unseparated products of peptic hydrolysis of glycated ubiquitin.
a b
Glycated peptide
Charge state
Mass (calculated/found)a
Modified amino acid
[59–67] [5–15] [4–15] [5–15]b [59–69] [46–58] [46–59] [44–58] [25–40] [25–41] [25–40]b [25–43] [4–23] or [3–22] [25–45] [25–45]b [22–45]
1+, 1+, 1+, 2+ 1+, 2+ 2+ 2+, 2+, 2+ 2+ 2+, 2+ 2+, 2+, 3+
1256.6136/1256.6184 1335.7861/1335.7914 1482.8545/1482.8603 1497.8389/1497.8458 1506.7566/1506.7631 1550.7424/1550.7500 1713.8057/1713.8142 1810.8949/1810.9072 1941.0055/1941.0154 2069.0641/2069.0722 2103.0583/2103.0694 2338.2492/2338.2596 2353.2516/2353.2478 2598.4017/2598.4116 2760.4545/2760.4628 2941.5761/2941.5864
K63 K6 or K11 K6 or K11 K6 or K11 K63 K48 K48 K48 K27, K29, or K27, K29, or K27, K29, or K27, K29, or K6 or K11 K27, K29, or K27, K29, or K27, K29, or
2+ 2+ 2+ 2+
3+ 3+
3+, 4+ 3+, 4+ 3+, 4+
K33 K33 K33 K33 K33 K33 K33
Monoisotopic mass. Fragments containing two hexose moieties.
(see Fig. 2). The peaks corresponding to the glycated peptides were identified by their specific isotopic distribution. The signals of Amadori products can be found by either visual inspection of the spectrum or automatic data analysis (see Materials and methods). The identified glycated fragments of ubiquitin are presented in Table 1. Our procedure detected 16 glycated fragments. In the obtained ESI–MS spectrum, the following ubiquitin monoglycated fragments were found: [59–67], [5–15], [4–15], [59–69], [46–58], [46–59], [44–58], [25–40], [25–41], [25–43], [4–23], [3–22], [25–45], and [22–45]. The glycated peptides were accom-
panied by their doubly glycated analogues in three cases: [5–15], [25–45], and [25–40]. The large number of identified glycated fragments suggests that the thermal glycation of ubiquitin is relatively unspecific and that the majority of lysine side chains are glycated to some extent even when the total glycation level of ubiquitin is moderate. This result contradicts the prediction obtained using the NetGlycate 1.0 server [23], which suggested that only Lys11 and Lys29 are susceptible to nonenzymatic glycation. One possible explanation for this discrepancy is that the data used as a training set for the sequence-based
Fig. 3. ESI–MS spectrum of glycated HSA hydrolysate (fraction 4). The symbols * and ** mark fragments of spectrum which are presented in expanded form as inserts.
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
predictor included both in vivo and in vitro data, and the in vitro glycation was performed at conditions preserving the native structure of proteins [23]. The high-temperature glycation used by us may alter the protein structure and, consequently, may result in the lower specificity of glycation. A similar protocol was applied to the HSA sample. Glycated protein was treated with dithiothreitol to reduce the disulfide bridges and was subjected to tryptic hydrolysis. For the enzymatic digestion, we selected trypsin, which cleaves the protein chain after Lys residues that are also the subject of glycation. This may be considered as a disadvantage because it impedes attempts at comparing concentrations of modified and unmodified peptides to get an idea of the degree of glycation at a given site. On the other hand, a recent article compared the applications of various enzymes in the proteomic analysis of glycation sites in plasma proteins. Three enzymes were taken into consideration: trypsin, Arg-C, and Lys-C. The highest number of glycated peptides was identified in the case of trypsin [24]. One of the main reasons for the selection of trypsin was the possibility of a direct comparison of our results with those published previously for HSA by Lapolla and coworkers [20] and Zhang and coworkers [24].
241
The unseparated products of enzymatic hydrolysis of HSA were directly analyzed by an ESI–FT–ICR mass spectrometer, but the number of detected glycated products was low (only 11 peptides); therefore, we decided to partially fractionate the enzymatic digest by HPLC (according to the procedure given in Materials and methods), and further experiments were performed on the fractions collected from the whole tryptic digest of HSA. The obtained spectra were analyzed automatically using an Excel spreadsheet. A representative example of the digest fraction spectrum is presented in Fig. 3. This spectrum, recorded for fraction 4 (retention time of 15–16 min), shows several signals, with a distinct isotopic pattern, corresponding to glycated peptides. The combined results of the analysis of HPLC fractions 1–15 are presented in Table 2. In this experiment, 45 glycated peptides were identified. This number is similar to that recently reported for HSA [19,25], confirming the usefulness of our method. The HSA sequence coverage is 75% according to the list of modified peptides. The glycated Lys residues are shown in Fig. 4. We also noticed some additional pairs of peaks with the characteristic shift of 6.0201 Da and comparable intensity; however, we were not able to assign these signals to the tryptic fragments of HSA. These peaks (e.g., the pair of peaks at m/z
Table 2 Glycated peptides identified by HRMS analyses of fractionated products of tryptic hydrolysis of glycated HSA.
a b
Glycated peptide
Retention time (min)
Charge state
Mass (calculated/found)a
Modified amino acid
[1–10] [433–444] [429–436] [198–205] [467–475] [182–190] [314–323] [535–545] [182–195] [561–574] [182–197] [182–197]b [200–209] [11–20] [349–359] [223–233] [198–209] [198–209]b [263–276] [219–233] [94–114] [137–144] [525–534] [187–197] [522–534] [414–428] [411–428] [187–197]b [175–190] [485–500] [373–389] [318–337] [501–521] [65–81] [115–137] [226–240] [542–557] [287–317] [223–240]b [219–240]b [373–402] [445–472] [223–240] [446–472] [21–51]
11–12 11–12 13–14 13–14 13–14 13–14 13–14 13–14 13–14 13–14 13–14 13–14 15–16 15–16 13–14 15–16 15–16 15–16 15–16 15–16 15–16 17–18 17–18 17–18 17–18 17–18 17–18 17–18 19 19 20 20 20 21 21 22 22 22 23 23 24 24 24 25 25
2+ 2+ 1+ 1+ 2+ 2+ 1+, 2+ 2+, 2+ 2+, 2+, 1+, 1+, 2+ 2+ 2+ 2+, 2+, 3+ 3+, 2+ 2+ 2+ 3+ 2+ 3+ 2+ 3+ 3+ 2+ 3+ 3+ 2+ 3+, 2+, 3+ 3+, 3+ 3+, 3+, 4+ 3+ 4+ 4+
1310.6215/1310.6272 1447.6799/1447.6872 963.5237/963.5296 1051.5583/1051.5656 1163.6034/1163.6092 1235.5881/1235.5932 1301.5809/1301.5872 1468.8249/1468.8352 1679.8214/1679.8288 1731.7509/1731.7582 1963.9811/1963.9892 2126.0339/2126.0452 1299.6129/1299.6192 1387.6507/1387.6572 1457.7501/1457.7572 1413.7028/1413.7092 1540.7919/1540.8012 1702.8448/1702.8528 1788.8451/1788.8538 1897.9785/1897.9878 2740.2511/2740.2664 1216.6340/1216.6392 1289.7442/1289.7502 1337.6423/1337.6512 1658.9818/1658.9898 1800.9833/1800.9932 2193.1893/2193.1957 1499.6951/1499.7015 1931.9874/1931.9958 2014.9558/2014.9608 2206.1409/2206.1492 2617.2840/2617.2888 2649.1905/2649.2058 2036.0612/2036.0732 2882.3830/2882.3930 1811.9404/1811.9478 2001.9605/2001.9708 3523.5681/3523.5768 2346.2094/2346.2170 2830.4852/2830.4958 3787.8551/3787.8678 3376.6465/3376.6594 2184.1566/2184.1638 3220.5454/3220.5584 3724.9071/3724.9184
K4 K436 or K439 K432 K199 R472 R186 K317 K536, K538, or K541 K190 K564 or K573 K190 or K195 K190 and K195 K205 K12 K351 K225 K199 or K205 K199 and K205 K274 K225 K106 K137 K525 K190 or K195 K524 or K525 K414 K413 or K414 K190 and K195 K181 R485 K378 K322 K519 K73 K136 K233 K545 K313 K225 and K233 K225 and K233 K378 or K389 K466 K225 or K233 K466 K41
Monoisotopic mass. Fragments containing two hexose moieties.
2+ 3+ 3+ 3+ 2+ 2+
3+ 3+ 4+
4+, 5+ 3+ 4+ 4+ 4+
242
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
Fig. 4. Scheme of HSA sequence showing the glycated Lys residues.
756.86 and 759.87 labelled with symbol [X] in Fig. 3) were subjected to the CID analysis (Fig. 5). The observed fragment ions demonstrate the pattern of neutral losses characteristic for Amadori products. The ion representing the isotopically labeled glycated
peptide eliminates the formaldehyde containing 13C. The presence of Amadori products that do not correspond to the predicted tryptic fragments of HSA may be explained by the incomplete reduction of disulfide bridges, additional posttranslational
Fig. 5. (A) ESI–MS spectrum of monoglycated unidentified peptide. (B and C) CID MS/MS fragmentation of nonlabeled (B) and labeled (C) ions of glycated peptide.
Detection of glycation sites in proteins / P. Stefanowicz et al. / Anal. Biochem. 400 (2010) 237–243
modifications of albumin [26], and/or partial chymotrypsin-like specificity of trypsin used for the digestion. We also compared the data obtained using our procedure with those reported previously for both in vivo and in vitro HSA glycation by Garlick and Mazer [27]. The authors indicated several specific lysine and arginine modification sites in glycated albumin; however, the most predominant site of nonenzymatic glycation in HSA in vivo is Lys525. Later results showed that Lys199, Lys281, Lys439, Lys525, Lys233, Lys317, Lys351, Lys12, and Lys534 are also susceptible to nonenzymatic glycation. In another report [28], four lysine residues (Lys199, Lys281, Lys439, and Lys525) were identified as the main glycation sites, although approximately 33% of the overall glycation occurs at Lys525. Recently, the additional modifications at Lys51, Lys159, Lys205, Lys286, and Lys538, as well as at Arg160, Arg222, and Arg472, were reported on the basis of MALDI–TOF analysis of in vitro minimally glycated HSA [29]. The number of lysine moieties that underwent the Amadori rearrangement in our experiment seems to be higher when compared with that of those experimentally detected in previous reports or calculated using the NetGlycate 1.0 server [23]. The fragments detected in our studies, covering nearly the whole sequence of HSA, demonstrate the low specificity of the thermal glycation. The presented data clearly prove that high-resolution mass spectrometry (HRMS) combined with isotopic labeling can be applied directly to the analysis of the glycation of proteins containing approximately 102 amino acid residues, whereas the more complex systems may require an LC–MS procedure. Although our experiments were performed on the instrument equipped with an ESI source, we believe that the MALDI source can also be used. The proposed method was applied in searching for products of Amadori rearrangement, the early stage of the glycation process. The existing methods, based on characteristic neutral loss patterns, could be successfully used for detection of peptide-derived Amadori products even in complex mixtures. However, the approach based on stable isotope labeling described here is not limited to products of Amadori rearrangement. The early glycation products undergo further chemical modifications, forming numerous advanced glycation end products (AGEs). The diversity of chemical structures of AGEs makes their detection difficult, and this is also complicated by the lack of a general method of preconcentration of AGEs. Theoretically, the method developed by us could also detect the unknown products of reaction of glucose (or glucose degradation products) with proteins. In general, the isotopic pattern characteristic of peaks corresponding to the product of reaction with glucose/[13C]glucose should be accompanied by specific neutral losses in an MS/MS experiment. Otherwise, the identified ion may represent an AGE; therefore, our procedure allows the detection of potential AGEs without prior information on their structure. The described method is suitable for model studies in vitro. The experiment in vivo would be more difficult and expensive because of the problems with the introduction of [13C]glucose at the concentration required for the in vivo glycation.
Conclusion We have presented a convenient and straightforward HRMS procedure for selective detection of protein glycation sites and peptide-derived Amadori products based on isotopic labeling during glycation with an equimolar mixture of glucose and [13C6]glucose. This new technique may also be useful for detection of AGEs in complex mixtures (e.g., enzymatic digests).
243
References [1] M. Amadori, Products of condensation between glucose and p-phenetidine I, Atti Accad. Nazl. Lincei 2 (1925) 337–345. [2] S.R. Thorp, J.W. Baynes, Role of the Maillard reaction in diabetes mellitus and diseases of aging, Drugs Aging 9 (1996) 69–77. [3] C. Rohlfing, M.H. Wiedmeyer, R. Little, Biological variation of glycohemoglobin, Clin. Chem. 48 (2002) 1116–1118. [4] D.M. Nathan, H. Turgeon, S. Regan, Relationship between glycated hemoglobin levels and mean glucose levels over time, Diabetologia 50 (2007) 2239–2244. [5] A. Frolov, R. Hoffmann, Analysis of Amadori peptides enriched by boronic acid affinity chromatography, Ann. NY Acad. Sci. 1126 (2008) 253–256. [6] Q. Zhang, N. Tang, J.W.C. Brock, H.M. Mottaz, J.M. Ames, J.W. Baynes, R.D. Smith, T.O. Metz, Affinity chromatography coupled with electron transfer dissociation mass spectrometry, J. Proteome Res. 6 (2007) 2323–2330. [7] I. Jeric, C. Versluis, S. Horvat, A.J.R. Heck, Tracing glycoprotein structures: electron ionization tandem mass spectrometric analysis of sugar–peptide adducts, J. Mass Spectrom. 37 (2002) 803–811. [8] P. Stefanowicz, J. Boratyn´ski, U. Kan´ska, I. Petry, Z. Szewczuk, Evaluation of high temperature glycation of proteins and peptides by electrospray ionization mass spectrometry, Acta Biochim. Pol. 48 (2001) 1137–1141. [9] A. Frolov, P. Hoffmann, R. Hoffmann, Fragmentation behavior of glycated peptides derived from D-glucose, D-fructose, and D-ribose in tandem mass spectrometry, J. Mass Spectrom. 41 (2006) 1459–1469. [10] M. Sniatynski, J.C. Rogalski, M.D. Hoffman, J. Kast, Correlation and convolution analysis of peptide mass spectra, Anal. Chem. 78 (2006) 2600–2607. [11] M.D. Hoffman, M.J. Sniatyn´ski, J.C. Rogalski, J.C. Le Blanc, J. Kast, Multiple neutral loss monitoring (MNM): a multiplexed method for post-translational modification screening, J. Am. Soc. Mass Spectrom. 17 (2006) 307–317. [12] M. Kijewska, A. Kluczyk, P. Stefanowicz, Z. Szewczuk, Electrospray ionization mass spectrometric analysis of complexes between peptide-derived Amadori products and borate ions, Rapid Commun. Mass Spectrom. 23 (2009) 4038– 4046. [13] N.L. Kelleher, R.B. Nicewonger, T.P. Begley, F.W. McLafferty, Identification of modification sites in large biomolecules by stable isotope labeling and tandem high resolution mass spectrometry: the active site nucleophile of thiaminase I, J. Biol. Chem. 27 (1997) 32215–32220. [14] D.R. Muller, P. Schindler, H. Towbin, U. Wirth, H. Voshol, S. Hoving, M.O. Steinmetz, Isotope-tagged cross-linking reagents: a new tool in mass spectrometric protein interaction analysis, Anal. Chem. 73 (2001) 1927–1934. [15] C. Ihling, A. Schmidt, S. Kalkhof, D.M. Schulz, C. Stingl, K. Mechtler, M. Haack, A.G. Beck-Sickinger, D.M. Cooper, A. Sinz, Isotope-labeled cross-linkers and Fourier transform ion cyclotron resonance mass spectrometry for structural analysis of a protein/peptide complex, J. Am. Soc. Mass Spectrom. 17 (2006) 1100–1113. [16] A. Sinz, Isotope-labeled photoaffinity reagents and mass spectrometry to identify protein–ligand interactions, Angew. Chem. Intl. Ed. 46 (2007) 660– 662. [17] J. Boratyn´ski, Dry reaction of proteins with carbohydrates at 120 °C yields neoglycoconjugates, J. Biotechnol. Tech. 12 (1998) 707–710. [18] J. Boratyn´ski, R. Roy, High temperature conjugation of proteins with carbohydrates, Glycoconj. J. 15 (1998) 131–138. [19] P. Stefanowicz, M. Kijewska, Z. Szewczuk, Sequencing of peptide-derived Amadori products by the electron capture dissociation method, J. Mass Spectrom. 44 (2009) 1047–1052. [20] A. Lapolla, D. Pedele, R. Reitano, N.C. Arico, Enzymatic digestion and mass spectrometry in the study of advanced glycation end products/peptides, J. Am. Soc. Mass Spectrom. 15 (2004) 496–509. [21] P. Stefanowicz, K. Kapczyn´ska, A. Kluczyk, Z. Szewczuk, A new procedure for the synthesis of peptide-derived Amadori products on a solid support, Tetrahedron Lett. 48 (2007) 967–969. [22] P. Stefanowicz, M. Kijewska, K. Kapczyn´ska, Z. Szewczuk, Methods of the siteselective solid phase synthesis of peptide-derived Amadori products, Amino Acids (in press), doi: 10.1007/s00726-009-0294-z. [23] M.B. Johansen, L. Kiemer, S. Brunak, Analysis and prediction of mammalian protein glycation, Glycobiology 16 (2006) 844–853. [24] Q. Zhang, A.A. Schepmoes, J.W.C. Brock, S. Wu, R.J. Moore, S.O. Purvine, J.W. Baynes, R.D. Smith, T.O. Metz, Improved methods for the enrichment and analysis of glycated peptides, Anal. Chem. 80 (2008) 9822–9829. [25] F.L. Brancia, J.Z. Bereszczak, A. Lapolla, D. Fedele, L. Baccarin, R. Seraglia, P. Traldi, Comprehensive analysis of glycated human serum albumin tryptic peptides by off-line liquid chromatography followed by MALDI analysis on a time-of-flight/curved field reflectron tandem mass spectrometer, J. Mass Spectrom. 41 (2006) 1179–1185. [26] D. Bar-Or, R. Bar-Or, L.T. Rael, D.K. Gardner, D.S. Slone, M.L. Craun, Heterogeneity and oxidation status of commercial human albumin preparations in clinical use, Crit. Care Med. 33 (2005) 1638–1641. [27] R.L. Garlick, J.S. Mazer, The principal site of non-enzymatic glycosylation of human serum albumin in vivo, J. Biol. Chem. 258 (1983) 6142–6146. [28] N. Iberg, R. Fluckiger, Nonezymatic glycosylation of albumin in vivo: identification of multiple glycosylated sites, J. Biol. Chem. 261 (1986) 13542–13545. [29] C. Wa, R.L. Cerny, W.A. Clarke, D.S. Hage, Characterization of glycation adducts on human serum albumin by matrix-assisted laser desorption/ionization timeof-flight mass spectrometry, Clin. Chim. Acta 385 (2007) 48–60.