Quantitative O-Glycomics by Microwave-Assisted β-Elimination in the

Jul 1, 2015 - Shionogi Innovation Center for Drug Discovery, Shionogi & Co., Ltd., Kita-21 Nishi-11, Kita-ku, Sapporo 001-0021, Japan. § Graduate Sch...
0 downloads 0 Views 733KB Size
Technical Note pubs.acs.org/ac

Quantitative O‑Glycomics by Microwave-Assisted β‑Elimination in the Presence of Pyrazolone Analogues Jun-ichi Furukawa,*,† Jinhua Piao,† Yasunobu Yoshida,‡ Kazue Okada,† Ikuko Yokota,† Kenichi Higashino,‡ Nobuo Sakairi,§ and Yasuro Shinohara*,† †

Laboratory of Medical and Functional Glycomics, Graduate School of Advanced Life Science, Hokkaido University, Sapporo 001-0021, Japan ‡ Shionogi Innovation Center for Drug Discovery, Shionogi & Co., Ltd., Kita-21 Nishi-11, Kita-ku, Sapporo 001-0021, Japan § Graduate School of Environmental Science, Hokkaido University, Sapporo 060-0810, Japan S Supporting Information *

ABSTRACT: O-Linked glycosylation of serine/threonine residues is a posttranslational modification of proteins and is essential for protein recognition and lipid functions on cell surfaces and within cells. The characterization of differently structured O-linked glycans (O-glycans) is particularly challenging because there is no known endoglycosidase for such groups. Therefore, chemical digestion approaches have been widely used; however, it is sometimes difficult to suppress unwanted side reactions. Recently, we reported a novel O-glycomics procedure using β-elimination in the presence of pyrazolone analogues (BEP). In the present study, we describe a microwave (MW)-assisted BEP procedure for rapid and quantitative O-glycomic analysis. Following optimization of the reaction conditions, the MW-assisted BEP reaction substantially improved the recovery of total O-glycans from model glycoproteins (PSM) and the reaction time was reduced from 16 to 2 h. Combined with sequential solid-phase extractions, this MW-assisted BEP procedure enabled O-glycomic analyses of various biological samples.

G

modified proteins. These chemical digestion approaches are often accompanied by a significant loss of intact O-glycans due to glycan degradation (referred to as peeling reactions) caused by hydroxide anions. The traditional Carlson procedure (reductive β-elimination), which efficiently cleaves from glyco-proteins/peptides more than 95% without peeling reactions, is the most commonly used and successful method for O-glycomic analysis; however, it has two major limitations. The first limitation is that originally glycosylated proteins/ peptides are unstable, which leads to the loss of information regarding glycosylation sites. The second limitation is that in situ reduction converts the original hemiacetal group at the reducing end of the carbohydrate to resultant alditols, which precludes derivatization suitable for enrichment, chromatographic, and mass spectrometric analyses. Zauner et al., Wang et al., and ourselves have recently introduced novel βelimination approaches, in which β-elimination is performed in the presence of pyrazolone analogues (BEP). These methods use dimethylamine, ammonia, or sodium hydroxide as a base for β-elimination and allow simultaneous labeling of O-glycans

lycosylation is one of the most important posttranslational modifications of proteins and is essential for proteins to perform their intended functions.1 Glycosylation of proteins is classified as N- and O-linked glycosylation. All Nlinked glycans (N-glycans) share the common core structure of Man3GlcNAc2 linked to an asparagine residue and are classified into three general types (high mannose, complex, and hybrid). O-Glycans linked to serine/threonine residues represent an extremely diverse group of modifications that are often classified on the basis of the innermost monosaccharide (N-acetylgalactosamine (GalNAc), mannose (Man), fucose (Fuc), xylose (Xyl), N-acetylglucosamine (GlcNAc), galactose (Gal), and glucose (Glc)). Although O-GlcNAcylation involves a single O-GlcNAc residue, other modifications subsequently form more complicated oligosaccharide structures through the actions of various glycosyltransferases. N-glycans can be readily cleaved from glycoproteins by specific deglycosylation enzymes, namely, peptide N-glycosidase (PNGase) F and PNGase A2,3 for structural analysis. However, O-glycans attached to serine/ threonine residues have mostly been subjected to chemical digestion using reductive β-elimination,4 β-elimination with a mild base such as ammonia5 or alkylamine,6 and classical hydrazinolysis7 because there is no known endoglycosidase for the various O-glycan structures, with the exception of protein O-GlcNAcase, which cleaves a single GlcNAc residue from © XXXX American Chemical Society

Received: February 13, 2015 Accepted: July 1, 2015

A

DOI: 10.1021/acs.analchem.5b02155 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

Figure 1. Effect of MW irradiation on BEP reaction for O-glycomic analysis (a) The recovered amounts of glycans detected by the BEP reaction and the MW-assisted BEP reaction are indicated by black bars and gray bars, respectively. (b) The amounts of glycans recovered from PSM at various temperatures (85, 100, 110, 120, 130, and 140 °C) over 2 h. (c) Recovery of PMP-labeled glycans from PSM using the MW-assisted BEP reaction for various amounts of time. PMP-labeled glycans of PSM using the MW-assisted PMP procedure in 160 mM NaOH and 200 mM PMP at 100 °C for 0.5, 1, 2, 3, and 4 h. (d) MALDI-TOF MS spectra showing the O-glycan profiles of PSM. PSM was subjected to the MW-assisted BEP reaction at 85 °C (upper panel), 120 °C (middle panel), and 140 °C (lower panel) for 2 h. All experiments were performed independently three times and the reproducibility of these experiments is indicted by the error bars. The signal numbers correspond to those described in Table S-1 in the Supporting Information.



EXPERIMENTAL SECTION MW-Assisted β-Elimination of Mucin-Type Glycoproteins in the Presence of Pyrazolone Analogues. Glycoprotein (200 μL, 100 μg) was treated with 400 μL each of sodium hydroxide (0.4 M) and a 0.5 M methanolic PMP solution in a 2 mL Pyrex vial and then heated using a Monowave reactor (Monowave 300, Anton Paar) to various temperatures (85, 100, 110, 120, 130, and 140 °C) for 0.5−4 h. To quantify the reaction yield, bis-PMP-labeled GN4 was added as an external standard after the BEP reaction. The reaction mixture was neutralized with 1.0 M hydrochloric acid and then subjected to purification on an Iatrobeads silica gel column.13 Briefly, the reaction mixture was diluted with acetonitrile (MeCN) (final concentration, 95%) and applied to ∼25 mg of Iatrobeads packed in a disposable filter column pre-equilibrated with 1 M acetic acid and MeCN. After sample loading, the column was first washed with 95% MeCN containing 1% acetic acid prepared in water and then with MeCN containing 2% acetic acid. PMP labeled glycans were eluted from the silica gel with 50% aqueous MeCN. Detailed experimental conditions for O-glycomic analysis of various biological samples are provided in the Supporting Information. MALDI-TOF/TOF MS Analysis. Purified O-glycan solutions were mixed with 2,5-dihydrobenzoic acid solution (10 mg/mL prepared in 30% MeCN) and subjected to MALDI-TOF MS analysis as previously described.12

released from glycoproteins/glycopeptides with 1-phenyl-3methyl-5-pyrazolone (PMP).8−10 We further found that when the reaction is performed with sodium hydroxide, pyrazolone analogues function as Michael donors and the resultant deglycosylated site of proteins/peptides are concomitantly modified via the Michael addition.9 Although the BEP procedure has become a promising means of analyzing Oglycans, O-glycomic study using a minute amount of biological samples is still a challenging task. In addition it is timeconsuming like other β-elimination methods requires 16 h to maximize cleavage and labeling efficiencies. Similar to other βelimination methods, the rate-limiting step is considered to be the cleavage process. Recently, Maniatis et al. demonstrated rapid de-O-glycosylation by MW-assisted β-elimination using dimethylamine as a base.11 They demonstrated that MWassisted β-elimination not only accelerates the reaction time (down to ∼2 h) but also improves the cleavage efficiency. The aim of this study was to clarify the effect of MW on the BEP procedure, which comprises two distinct reactions: βelimination and labeling of resultant free glycans with a pyrazolone analogues. We herein demonstrated that the MWassisted BEP reaction is a powerful tool for rapid and quantitative O-glycomic analysis because it substantially improves the reaction time and reaction efficiencies. We applied the MW-assisted BEP procedure, in combination with sequential solid-phase extractions, to cellular O-glycomic analysis, as has already been performed using various human cells.12 In this study, we describe a detailed protocol not only for cellular O-glycomics but also for the O-glycomic analysis of other biological samples including serum, tissue, and formalinfixed paraffin-embedded (FFPE) samples.



RESULTS AND DISCUSSION MW-Assisted β-Elimination of Mucin-Type Glycoproteins in the Presence of Pyrazolone Derivatives. BEP9 comprises two distinct reactions: β-elimination and carbohydrate derivatization with pyrazolones, the latter of which was first reported by Honda et al.14 The usefulness of MW B

DOI: 10.1021/acs.analchem.5b02155 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry irradiation in β-elimination has been proven;12 therefore, we first investigated the effect of MW radiation on the derivatization efficiency of oligosaccharides by PMP using maltohexaose as a model. Relative yields were estimated by comparing the signal strength of an external standard by MALDI-TOF MS. MW irradiation accelerated the reaction and derivatization was completed in 10 min (Figure S-1 in the Supporting Information). By contrast, derivatization by conventional heating was slower and the reaction yield at 60 min was 80% of that observed when MW irradiation was applied. Thus, MW irradiation improved both the reaction time and the reaction yield. We next validated the effect of MW irradiation on the BEP reaction for O-glycomic analysis of glycoproteins using PSM as a model glycoprotein. BEP reactions were performed with or without MW irradiation under the same reaction conditions. The reaction time was fixed at 2 h, as this was previously demonstrated to be long enough to release O-glycans from a model glycopeptide by MW-assisted reductive β-elimination.11 Up to 14 mucin-type O-glycans labeled with PMP could be detected as sodiated cations (Figure S-2 in the Supporting Information). A quantitative summary of the major O-glycans in PSM is shown in Figure 1a. Although MW irradiation did not appear to significantly affect the relative quantitation, it increased the amount of O-glycans recovered by ∼3-fold. In our previous study, we observed almost quantitative release and labeling of O-glycans from model glycopeptides by BEP reaction with conventional heating;9 however, the introduction of MW irradiation to this reaction greatly improved the release and labeling of O-glycans from mucin-type glycoproteins. This may be explained by the increased efficiencies of glycoprotein cleavage and labeling with PMP. It is important to note that unwanted degraded O-glycan and peptide fragments were not observed, regardless of whether MW irradiation was applied. Optimization of the MW-Assisted BEP Reaction Conditions. We next sought to optimize the MW-assisted BEP reaction in order to maximize the reaction yield of PMPlabeled O-glycans and shorten the reaction time. First, we examined the yield of the MW-assisted BEP reaction under different temperature conditions in three independent experiments. The recovery of O-glycans labeled with PMP was highly temperature-dependent and was highest, with up to 43 mucintype O-glycans detected, at 120 °C (Figure 1b). Temperatures higher than 130 °C reduced the recovery due to degradation because many unidentifiable signals were clearly observed on MS spectra (Figure 1d, lower panel). O-Glycan signals at m/z 1027.8 (No. 7), 1597.2 (No. 21), and 2368.6 (No. 42) were only detected when the reaction was performed at 120 °C due to the improved recovery efficiency at this temperature. The recovery yield of PMP-labeled glycans was quantified at 0, 0.5, 1, 2, 3, and 4 h (Figure 1c). The yields of all O-glycans labeled with PMP increased up to 2 h, after which they slowly decreased. Therefore, the optimized reaction conditions with MW irradiation were determined to be 120 °C for 2 h. By contrast, when reaction was performed without MW irradiation, the recovery yield of PMP-labeled glycans did not reach a plateau even after 16 h, and the recovery was ∼50% compared with achieved using MW-assisted BEP reaction (Figure S-3 in the Supporting Information). To evaluate the mechanism by which MW irradiation improves the efficiency of the BEP reaction, we compared reactions performed in a 10 mL of Pyrex vial or a silicon carbide (SiC) vial. The latter is a strong MW-absorbing inert ceramic

material and effectively shields the reaction mixture from MW irradiation, thereby enabling MW thermal effects to be separated from nonthermal effects as reported by Obermayer et al.15 When a SiC vial was used, small peaks of degraded protein fragments were observed, while the yield of PMPlabeled O-glycans was almost identical to when a Pyrex vial was used (Figure S-4 in the Supporting Information). This suggests that temperature underlies the improvement in the efficiency of the BEP reaction by MW irradiation and that the electromagnetic field has no direct influence on the reaction pathway, as has been recently reported in a wide variety of chemical transformations including organic reactions, preparation of inorganic nanoparticles, and protein hydrolysis.16 To suppress the chemical noise of degraded proteins, Pyrex vials were used for all subsequent O-glycomic analyses. Effect of Structure on the Signal Strength of Oligosaccharides in MALDI-TOF MS. When the PMPlabeled O-glycans were analyzed by three independent MALDITOF MS analyses, we observed fairly good reproducibility (Figure S-5 in the Supporting Information), suggesting that the distribution of PMP labeled O-glycans in DHB matrix is fairly even regardless of differences in structure. Quantitative analysis of glycans by MALDI-TOF MS relies on the previous observations that oligosaccharides with masses of >∼1000 Da exhibit similar signal strengths, regardless of their structure17,18 and that methyl esterification of sialic acid renders sialylated oligosaccharides chemically equivalent to neutral oligosaccharides, permitting simultaneous analysis of neutral and sialylated oligosaccharides.19,20 Some of the O-glycans analyzed in this study have molecular masses of 130 °C) under MW irradiation. The optimized MW-assisted BEP procedure was applied to O-glycomic analyses of various biological samples. Cellular Oglycomic analyses have been performed using the reductive βelimination technique combined with permethylation to improve characterization by MALDI-TOF MS and hydrazinolysis followed by fluorescent labeling for high-performance liquid chromatography analysis. However, these approaches typically require 107 cells, which is more than one order higher than the number of cells routinely used for cellular N-glycomic analysis. The method presented in this study can be used for Oglycomic analysis of as few as 106 cells owing to the improved BEP reaction efficiency and the establishment of purification procedures using multiple solid-phase extractions. It is worth mentioning that that N-glycans are resistant to BEP is a great advantage, as cellular and serum glycoproteins often contain a large amount of N-glycans, which may interfere with the analysis of O-glycans. The protocol established in the current study was proven to be applicable to O-glycomic analysis of cell, serum, frozen tissue, and FFPE tissue, which are all important for the discovery of disease-related biomarkers. Thus, the MW-assisted BEP procedure will have many applications in drug discovery and medical care.



ASSOCIATED CONTENT

S Supporting Information *

Experimental Section, Supplementary Figures 1−7, and Supplementary Tables 1 and 2 as noted in text. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b02155.



AUTHOR INFORMATION

Corresponding Authors

*Phone: +81 11 706 9080. Fax: +81 11 706 9087. E-mail: [email protected]. *Phone: +81 11 706 9091. Fax: +81 11 706 9087. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported in part by the Special Coordination Funds for Promoting Science and Technology from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) of Japan and by JSPS KAKENHI (Grant Number 24550089). The authors thank Dr. Tomoya Tojima and Ms. Aiko Wada in Anton Paar Japan for the technical support on the MW reactor.



REFERENCES

(1) Spiro, R. G. Glycobiology 2002, 12, 43R−56R. E

DOI: 10.1021/acs.analchem.5b02155 Anal. Chem. XXXX, XXX, XXX−XXX