MALDI-MS Procedures for

Biology and Chemistry,. The University of Georgia, 220 Riverbend Road, Athens, Georgia 30602-4712. Exoglycosidase digestion coupled with matrix-as...
0 downloads 0 Views 130KB Size
Technical Notes Anal. Chem. 1996, 68, 570-572

Simplifying the Exoglycosidase Digestion/ MALDI-MS Procedures for Sequencing N-Linked Carbohydrate Side Chains Yi Yang and Ron Orlando*

Complex Carbohydrate Research Center and Departments of Biochemistry and Molecular Biology and Chemistry, The University of Georgia, 220 Riverbend Road, Athens, Georgia 30602-4712

Exoglycosidase digestion coupled with matrix-assisted laser desorption/ionization mass spectrometry (MALDIMS) is an effective technique for sequencing the N-linked carbohydrate side chains of a glycoprotein. However, the buffers currently used in the enzymatic procedures are detrimental to MALDI-MS, and thus desalting is required before the digestion products can be analyzed. We demonstrate that a 25 mM ammonium acetate solution adjusted to the proper pH can replace the normal exoglycosidase digestion buffers. The use of these ammonium acetate solutions permits direct MALDI-MS analysis of the digestion mixture without desalting. More importantly, we show that many of the commonly used exoglycosidases retain both their activity and their specificity under these conditions. Exoglycosidase digestion coupled with matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) has been demonstrated to be an effective method for sequencing the N-linked carbohydrate side chains of a glycoprotein.1-3 The process is based on the use of exoglycosidases to selectively release monosaccharides from the nonreducing termini based on their stereochemistry, anomeric configuration, and linkage to the remainder of the carbohydrate side chain.4 After each exoglycosidase digestion, a portion of the sample is analyzed by MALDIMS to determine the shift in molecular weight (MW), which provides the number of monosaccharides released.1-3 By using these enzymes either in series or as an array, the stereochemical sequence, linkages, and anomeric centers of each monosaccharide can be established, which permits the primary structure of the carbohydrate side chain to be elucidated.1-3,5-7 (1) Sutton, C. W.; Poole, A. C.; Cottrell, J. S. In Techniques in Protein Chemistry IV; Angeletti, R. H., Ed.; Academic Press: San Diego, 1993; pp 109-116. (2) Sutton, C. W.; O’Neill, J. A.; Cottrell, J. S. Anal. Biochem. 1994, 218, 3446. (3) Harvey, D. J. Am. Lab. 1994, 26, 22-28. (4) Kobata, A. Anal. Biochem. 1979, 100, 1-14. (5) Schindler, P. A.; Settineri, C. A.; Collet, X.; Fielding, C. J.; Burlingame, A. L. Protein Sci. 1995, 4, 791-803. (6) Settineri, C. A.; Burlingame, A. L. In Techniques in Protein Chemistry V; Crabb, J. W., Ed.; Academic Press: San Diego, 1994; pp 97-104.

570

Analytical Chemistry, Vol. 68, No. 3, February 1, 1996

Much of the initial work on exoglycosidase carbohydrate sequencing was performed with size-exclusion chromatographic or gel electrophoretic analysis of the digestion products.4 These two techniques are immune to the buffers used for these enzymatic digestions, which contain either 100 mM sodium acetate or sodium citrate/phosphate. Since the activity and specificity of these enzymes have been established only in these buffers, they are currently used for exoglycosidase sequencing with MALDI-MS detection,1-3 even though these buffers are detrimental to MALDI-MS analysis. In particular, the high concentration of sodium in these buffers broadens the molecular ion peaks and reduces their intensities because of the numerous sodium-analyte adducts observed, which can totally conceal the sample signal when trace quantities of oligosaccharide are analyzed. Therefore, desalting is required before performing MALDI-MS analysis,1-3 which creates additional experimental procedures for the researcher and can result in sample losses. A better approach would be to eliminate the use of the sodiumcontaining exoglycosidase buffers and thus eliminate the need for desalting the sample prior to MALDI-MS analysis. In this paper, we demonstrate that a 25 mM ammonium acetate solution adjusted to the proper pH can replace the normal exoglycosidase digestion buffers. The use of these ammonium acetate solutions permits direct MALDI-MS analysis of the digestion mixture without desalting. More importantly, we show that many of the commonly used exoglycosidases retain both their activity and their specificity under these conditions. EXPERIMENTAL SECTION N-Linked glycopeptides with known carbohydrate side chains were obtained by HPLC purification of reduced/S-carboxymethylated and trypsin-digested bovine asialofetuin (Sigma Chemical Co., St. Louis, MO). Exoglycosidase digestions were performed sequentially on ∼50 pmol of glycopeptide, which was dissolved in 10 µL of buffer. All the exoglycosidases were purchased from Oxford GlycoSystems Ltd. (Abingdon, UK). Two sets of conditions were used for the exoglycosidase digestions. The first used the buffers suggested by Oxford GlycoSystems, while the second (7) Medzihradszky, K. F.; Maltby, D. A.; Hall, S. C.; Settineri, C. A.; Burlingame, A. L. J. Am. Soc. Mass Spectrom. 1994, 5, 350-358. 0003-2700/96/0368-0570$12.00/0

© 1996 American Chemical Society

Figure 1. Structures of the carbohydrate side chains attached to tryptic fragment 7 from bovine asialofetuin before and after release of the nonreducing terminal β-1-4-linked Gal residues by digestion with β-galactosidase from S. pneumoniae.

used a 25 mM ammonium acetate solution adjusted to the suggested pH for each enzyme. Aside from the difference in buffers, all other conditions were identical, i.e., performed overnight at 37 °C with the same enzyme-substrate ratio. After each digestion, 0.5 µL of the digestion mixture was removed from the reaction vial, mixed with the MALDI-MS matrix solution, and directly analyzed. The MALDI-MS experiments were performed on a HewlettPackard LDI 1700XP time-of-flight mass spectrometer. This instrument was operated at an accelerating voltage of 30 kV, an extractor voltage of 9 kV, and a pressure of ∼8 × 10-7 Torr. Samples were desorbed/ionized from the probe tip with a nitrogen laser (λ ) 337 nm) having a pulse width of 3 ns and delivering ∼10.5 µJ energy/laser pulse. A methanol-water solution (90: 10), saturated with sinapinic acid (3,5-dimethoxy-4-hydroxycinnamic acid), was used as the matrix. The instrument was calibrated with an external mixture of known peptides. RESULTS AND DISCUSSION Purified tryptic fragment 7 (expected fragments were numbered sequentially from the N-terminus of bovine asialofetuin) is used here to demonstrate the difference between these two buffers, although exoglycosidase digestions were performed on all three of the N-linked glycopeptides from this glycoprotein. The one N-linked glycosylation site in this glycopeptide is known to possess three different carbohydrate side chains, whose structures are shown in Figure 1.8,9 The bi- and triantennary glycoforms are easily identified by MALDI-MS analysis of this fragment (Figure 2A); however, the two triantennary glycomers cannot be distinguished, as both of these species have the same MW. The nonreducing terminal β-1-4-linked Gal residues were selectively released from these carbohydrate side chains by digestion with β-galactosidase from Streptococcus pneumoniae.4,10 As shown in Figure 1, this digestion releases two galactose residues (∆MW ) 324) from the biantennary carbohydrate side chain, three galactose residues (∆MW ) 486) from triantennary glycan I, and two galactose residues (∆MW ) 324) from triantennary glycan II. MALDI-MS analysis of the digestion (8) Green, E. D.; Gabriela, A.; Baenziger, J.; Wilson, S.; Van Halbeek, H. J. Biol. Chem. 1988, 263, 18253-18268. (9) Cumming, D. A.; Hellerqvist, C. G.; Harris-Brandts, M.; Michnick, S. W.; Carver, J. P.; Bendiak, B. Biochemistry 1989, 28, 6500-6512. (10) Distler, J. J.; Jourdian, G. W. J. Biol. Chem. 1973, 248, 6772-6780.

Figure 2. MALDI-MS spectra of purified tryptic fragment 7 from bovine asialofetuin (A) before and (B and C) after release of the nonreducing terminal β-1-4-linked Gal residues by digestion with β-galactosidase from S. pneumoniae. Spectrum B was obtained after digestion in the suggested buffer (100 mM sodium acetate at pH 6), and spectrum C was obtained after digestion in a 25 mM ammonium acetate solution adjusted to pH 6. Table 1. Exoglycosidases That Retain Their Activity and Specificity in 25 mM Ammonium Acetate Solutions

exoglycosidase

source

specificity

β-galactosidase β-galactosidase β-GlcNAcase β-HexNAcase R-mannosidase R-mannosidase

S. pneumoniae bovine testes S. pneumoniae chicken liver jack bean jack bean

Gal β 1-4 Gal β 1-3/1-4 GlcNAc β 1-2 HexNAc β 1-3/1-4 Man R 1-3d Man R 1-3,1-6d

suggested enzyme/ pH and substrate buffer ratiosa 6.0b 4.0c 6.0c 4.0c 4.5b 4.5b

40 800 2 200 20 1000

a Enzyme/substrate ratios are reported in microunits of exoglycosidase per picomole of glycopeptide. b 100 mM sodium acetate is the suggested buffer. c 100 mM sodium citrate/phosphate is the suggested buffer. d At low enzyme/substrate ratios, R-mannosidase is specific for Man R-1-3, while at high ratios this exoglycosidase will release both R-1-3- and R-1-6-linked Man residues.

mixture when the suggested buffer for this enzyme (100 mM sodium acetate at pH 6) was used (Figure 2B) demonstrates that the molecular ions of the digestion products are significantly broadened by the presence of multiple sodium adducts when compared with this sample before digestion (Figure 2A). This peak broadening decreases the MW accuracy of these measurements and is expected to cause problems when glycoforms with similar MWs are analyzed. By broadening the molecular ion envelope, the presence of excess sodium reduces the signal-tonoise ratio of the molecular ions. In fact, the new biantennary structure is observed just above the matrix background, and after subsequent exoglycosidase digestions with the suggested buffers, this glycopeptide could not be detected above the background. Alternatively, when β-galactosidase digestion is performed in a 25 mM ammonium acetate solution adjusted to pH 6, direct MALDI-MS analysis of this mixture (Figure 2C) demonstrates that the molecular ion peak widths and signal-to-noise ratios are comparable to those observed before this treatment (Figure 2A). The use of ammonium acetate solutions at this concentration eliminates the peak broadening problems discussed above because it is volatile, which permits the majority of this buffer to evaporate when the sample is placed into the MALDI-MS. Analytical Chemistry, Vol. 68, No. 3, February 1, 1996

571

However, at buffer concentrations over 25 mM, sufficient ammonium acetate is present in the sample, even under vacuum, to produce ammonium adducts in the MALDI-MS spectrum. The use of ammonium salts has also been reported to decrease the suppression effects encountered with MALDI-MS analysis of peptide mixtures.11 Both of these digestions yield the exact same products, demonstrating that this exoglycosidase has the same specificity in both solutions. However, the use of ammonium acetate has eliminated the problems associated with the sodium buffer. Numerous other exoglycosidase digestions were performed on the N-linked carbohydrate side chains of these glycopeptides to confirm their complete primary structures. These experiments demonstrated that the exoglycosidases listed in Table 1 retain (11) Woods, A. S.; Huang, A. Y. C.; Cotter, R. J.; Pasternack, G. R.; Pardoll, D. M.; Jaffee, E. M. Anal. Biochem. 1995, 226, 15-25.

572

Analytical Chemistry, Vol. 68, No. 3, February 1, 1996

both their activity and their specificity in a 25 mM ammonium acetate solution. The proteolytic activity reported for these exoglycosidases1,2,5-7 was not observed in ammonium acetate solutions. Furthermore, ammonium acetate solutions permit direct MALDI-MS analysis of the exoglycosidase digestion mixture without the need for desalting, thus saving time and minimizing sample loss. ACKNOWLEDGMENT This work was supported by a grant from the National Institutes of Health (NIH 2-P41-RR05351-06). Received September 18, 1995. Accepted November 7, 1995.X AC950932Z X

Abstract published in Advance ACS Abstracts, December 15, 1995.