Fragmentation of Oligosaccharides from Sodium Adduct Molecules Depends on the Position of N-Acetyl Hexosamine Residue in Their Sequences in Mass Spectrometry

Abstract

Six different sequences of hexasaccharides, pyridylaminated malto-hexaoses containing one N-acetyl hexosamine (HexNAc) residue, were analyzed using matrix-assisted laser desorption/ionization (MALDI) tandem time-of-flight (TOF) mass spectrometry (MS). Based on the product ion spectra of sodium adducts [M+Na]⁺, the chemical species of the observed product ions contained a HexNAc residue and had high ion abundance, indicating that the HexNAc residue had a higher affinity to sodium atom than glucopyranose. The acetamide group coordinated easily to sodium atom. This general rule of product ion generation was useful to predict the structure of the oligosaccharides based on the MS/MS product ion spectra.

INTRODUCTION

Sugar chains are among the most important biological molecules, in addition to proteins and DNA. Until recently, the structural analyses of sugar chains had been more difficult than those of protein and DNA because sugar chains cannot be amplified and overexpressed, and because the amount of the sugar analytes contained in biological tissues is very small. Therefore, mass spectrometry (MS) and MS/MS techniques are among the most useful to analyze or identify the structures of sugars.^1–6) Sugar chain structures are complex due to their glycosyl bond variations and structural isomers. Relative ion abundance analysis of the fragment ions and spectral pattern matching based on the information obtained was a powerful strategy to distinguish the structures of isomeric sugar chains and to identify their fine structures.^6–11) MS/MS spectral patterns depend on many factors such as the sugar chain structure, sugar content, and ionic forms of sugars. Recently, we analyzed cyclodextrin derivatives with amino (–NH₂) and acetamide (–NHCOCH₃) groups using their tandem MS/MS product ion spectra.¹²⁾ The results suggested that sodium atom had a higher affinity to HexNAc residues than to glucopyranose residues, based on the isotope-like ion signal abundances of each product ion and their ion abundance of the cyclodextrin derivatives. In that study, the cyclic oligosaccharides had a unique cyclic structure and the number of analytes was small. The generality of the affinity of HexNAc to sodium atom should be verified for other linear oligosaccharides. In this study, we systematically prepare six malto-hexasaccharides containing one HexNAc residue, in which the sequential position of the HexNAc residue differed.¹³⁾ We reveal the influence of the HexNAc position on the sequential analysis of oligosaccharides.

MATERIALS AND METHODS

A matrix of 2,5-dihydroxybenzoic acid (DHBA) was purchased from Sigma-Aldrich (St. Louis, MO, USA). Acetonitrile was purchased from Nakarai Tesque Co. (Kyoto, Japan). A series of six pyridylaminated malto-hexasaccharides as shown in Fig. 1 containing one N-acetyl hexosamine (altrosamin) and five D-glucopyranoside (Glc) residues, in which the glycosyl bonds are of the α1–4 type linkage (Fig. 1), was prepared chemically from 3A-amino-3A-deoxy-(2AS,3AS)-α-cyclodextrin (Tokyo Chemical Industry Co., Ltd., Tokyo), as described in our previous paper.¹³⁾

Fig. 1. Structure of six 2-aminopyridine derivatized maltohexaoses with N-acetyl hexosamine 1–6 and their fragmentation species.

The ion peaks labeled by a red circle contained a PA reducing end with the HexNAc residue, and they were generated by sequential removal of the glucose (Glc) residues from the non-reducing terminal. The peaks labeled by a red inverted triangle contained a PA reducing end without the HexNAc residue, and they were generated by removal of the Glc and HexNAc residues from the non-reducing terminal. The peaks labeled by a blue square were generated by removal of the PA sugar reducing end.

MASS SPECTROMETRY

All MS/MS fragment spectra were acquired with an UltraflexIII MALDI-TOF/TOF MS instrument (Bruker daltonics, Bremen, Germany), in which the product ions were generated by laser-induced dissociation (LID). All MS/MS experiments were carried out under the same conditions. DHBA was used as a matrix at 10 mg/mL 80% acetonitrile aqueous solution. The analytes were dissolved in water and 1 μL analyte and matrix solution were mixed and dried on the target plate. In this study, a sodium dopant such as sodium TFA was not added because sodium adducts of the analytes were predominantly observed.

RESULTS

The LID-MS/MS product ion spectra of PA-malto-hexasaccharides containing one HexNAc residue 1 to 6 (Fig. 1) were measured from the sodium adducts [M+Na]⁺ (Fig. 2). All spectral patterns of the labeled peaks of 1 and 6 were different and depended on the sequential position of the N-acetyl hexosamine residue (HexNAc) in each oligosaccharide 1–6 (Fig. 1). These spectra also differed from the spectra produced from the protonated molecules [M+H]⁺, in which all the product ions were generated at the reducing-end side, with PA predominating in all compounds 1 to 6 because the PA amine group was strongly protonated as the charge center.¹³⁾

Fig. 2. LID MS/MS product ion spectra produced from sodium adducts [M+Na]⁺ of 1–6.

The ion peaks labeled by a red circle contained a PA reducing end with the HexNAc residue, and they were generated by sequential removal of the glucose (Glc) residues from the non-reducing terminal. The peaks labeled by a red inverted triangle contained a PA reducing end without the HexNAc residue, and they were generated by removal of the Glc and HexNAc residues from the non-reducing terminal. The peaks labeled by a blue square were generated by removal of the PA sugar reducing end.

In the product ion spectra of the sodium adducts [M+Na]⁺ (Fig. 2), the ion peaks labeled by a red circle contained a PA reducing end with the HexNAc residue, and they were generated by the sequential removal of the glucose (Glc) residues from the non-reducing terminal such as 5. The peaks labeled by a red inverted triangle had a PA reducing end without the HexNAc residue, and they were generated by the removal of the Glc and HexNAc residues from the non-reducing terminal (Figs. 1 and 2). The peaks labeled by a blue square were generated by the removal of PA-sugar reducing end such as 1–4 (Figs. 1 and 2).

Comparing the influence of the non-reducing end HexNAc and Glc residues of 1 and 2 on the fragment ion abundance (Fig. 1), the peak at m/z 970 for 2 had a higher abundance than that at m/z 929 for 1, normalized to the peak at m/z 874, which was produced commonly by the removal of the reducing end PA–Glc residue from both 1 and 2 (Fig. 2). This result indicated that the HexNAc residue has a high affinity to sodium atom. Namely, the peak at m/z 970 for 2, which contains HexNAc, has a higher abundance than that at m/z 929 for 1, which does not contain HexNAc residue (Figs. 1 and 2). The ion peaks labeled by a red circle corresponded to the number of the sequential Glc residues at the non-reducing end. The next HexNAc residue hindered the generation of these ions (Fig. 2, 2–6). When the HexNAc residue was removed from the ions, the abundance was greatly reduced. These results indicated that sodium atom attached to the HexNAc residue and that HexNAc easily became a charge center.

Since the sodium atom affinity to HexNAc is high and the charge centers on HexNAc differed in the structures of 1 to 6, the product ion spectra depended on their structure, i.e., the position of the HexNAc residue.

The ion peaks labeled by a blue square were observed in the product ion spectra of 1 to 4. These peaks were in part generated by multi-side cleavage of the glycosyl bonds both from the reducing and non-reducing ends. Their chemical species included the HexNAc residue in the middle of the structure.

For example, the ion peaks labeled by the blue sequence of 1 and 2 were generated by one-site cleavage from the reducing end. However, the ion peak at m/z 388 for 3 was generated by two-site cleavage from the reducing and non-reducing ends, and the ion species contained the middle HexNAc residue because HexNAc has a high affinity to sodium atom and the middle species with HexNAc easily became a charge center. The ion peaks at m/z 388 and 550 were also generated by two-side cleavage.

In the case of 5, the HexNAc residue is adjacent to the reducing end of PA–Glc and the charge center of the sodium ion located at the reducing end side. The ion peaks labeled by a red circle were observed predominately. The HexNAc is at the reducing end as PA–HexNAc in 6. The ion peaks labeled by a blue triangle were generated with the only glucose oligomers attached by a sodium atom. These results indicated that the acetamide groups were effective at coordinating to sodium ion and that the coordination of sodium atom to the oligosaccharides required multi-sites of oxygen atom. The structure of two sugar residues or more and glycosyl bonds were also important to coordinate to sodium ion.

DISCUSSION

The fragmentation of the oligosaccharides with HexNAc and PA differed between their sodium adducts [M+Na]⁺ and protonated forms [M+H]⁺. The fragmentation from [M+Na]⁺ was complex because the possible charge centers were not only at the PA unit but also at the HexNAc residues and others. In contrast, the product ion spectra of [M+H]⁺ were simple because the pyridyl amine unit was protonated strongly as a charge center, and the sequential fragmentation was generated only by removal of sugar residues from the non-reducing end side, including the HexNAc residue.

In the case of the fragmentation from [M+Na]⁺, many chemical species of the product ions contained the N-acetyl hexosamine residue. Namely, the HexNAc residue had a high affinity to sodium atom and it easily became a charge center, which is consistent with our previous report.¹²⁾ The stability of the product ions was important to analyze and predict the sequence and structure of oligosaccharides using the MS/MS product ion spectra. These rules in MS/MS fragmentation can be essential to automate prediction of oligosaccharide structures based on analysis of their product ion spectra in the future.

REFERENCES

• 1) D. J. Harvey. Analysis of carbohydrates and glycoconjugates by matrix-assisted laser desorption/ionization mass spectrometry: An update covering the period 1999–2000. Mass Spectrom. Rev. 25: 595–662, 2006.
• 2) Y. Mechref, M. V. Novotny, C. Krishnan. Structural characterization of oligosaccharides using MALDI-TOF/TOF tandem mass spectrometry. Anal. Chem. 75: 4895–4903, 2003.
• 3) D. J. Harvey. Fragmentation of negative ions from carbohydrates. Part 1. Use of nitrate and other anionic adducts for the production of negative ion electrospray spectra from N-linked carbohydrates. J. Am. Soc. Mass Spectrom. 16: 622–630, 2005.
• 4) V. N. Reinhold, B. B. Reinhold, C. E. Costello. Carbohydrate molecular weight profiling, sequence, linkage, and branching data: ES-MS and CID. Anal. Chem. 67: 1772–1784, 1995.
• 5) A. V. Everest-Dass, D. Kolarich, M. P. Campbell, N. H. Packer. Tandem mass spectra of glycan substructures enable the multistage mass spectrometric identification of determinants on oligosaccharides. Rapid Commun. Mass Spectrom. 27: 931–939, 2013.
• 6) T. Yamagaki, H. Nakanishi. Ion intensity analysis of post-source decay fragmentation in curved-field reflectron matrix-assisted laser desorption/ionization time-of-flight mass spectrometry of carbohydrates: For structural characterization of glycosylation in proteome analysis. Proteomics 1: 329–339, 2001.
• 7) Y. Takegawa, S. Ito, S. Yoshioka, K. Deguchi, H. Nakagawa, K. Monde, S. Nishimura. Structural assignment of isomeric 2-aminopyridine-derivatized oligosaccharides using MSⁿ spectral matching. Rapid Commun. Mass Spectrom. 18: 385–391, 2004.
• 8) A. Kameyama, N. Kikuchi, S. Nakaya, H. Ito, T. Sato, T. Shikanai, Y. Takahashi, K. Takahashi, H. Narimatsu. A strategy for identification of oligosaccharide structures using observational multistage mass spectral library. Anal. Chem. 77: 4719–4725, 2005.
• 9) A. Kameyama, S. Nakaya, H. Ito, N. Kikuchi, T. Angata, M. Nakamura, H. Ishida, H. Narimatsu. Strategy for simulation of CID spectra of N-linked oligosaccharides toward glycomics. J. Proteome Res. 5: 808–814, 2006.
• 10) T. Yamagaki, Y. Ishizuka, S. Kawabata, H. Nakanishi. Postsource decay fragment spectra of cyclomalto-octaose and branched cyclomaltohexaose by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Rapid Commun. Mass Spectrom. 10: 1887–1890, 1996.
• 11) T. Yamagaki, H. Nakanishi. Distinguishing of linkage isomers of lactotetra oligosaccharides by using the relative ion intensity analysis of post-source decay fragment ions in curved-field reflectron matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Anal. Sci. 17: 83–87, 2001.
• 12) T. Yamagaki, K. Sugahara, T. Watanabe. Amino and acetamide functional group effects on the ionization and fragmentation of sugar chains in positive-ion mass spectrometry. J. Am. Soc. Mass Spectrom. 25: 95–103, 2014.
• 13) M. Nakamura, Y. Makino, C. Takagi, T. Yamagaki, M. Sato. Probing the catalytic site of rabbit muscle glycogen phosphorylase using a series of specifically modified maltohexaose derivatives. Glycoconj. J. 34: 563–574, 2017.