Characterization of protein glycosylation by mass spectrometry

Characterization of protein glycosylation by mass spectrometry

Characterization of protein glycosylation by mass spectrometry Alma L Burlingame Electrospray ionization, a natural interface with microbore and capi...

617KB Sizes 0 Downloads 186 Views

Characterization of protein glycosylation by mass spectrometry Alma L Burlingame

Electrospray ionization, a natural interface with microbore and capillary high-pressure liquid chromatography, has become the method of choice for the reliable structural characterization of protein glycosylation by mass spectrometry at the picomole level. Its advantages include inherent sensitivity in the femtomole range, compatibility with collisional activation methods that both permit the detection and monitoring of structurally specific ions and enable the induction of glycopeptide fragmentation that facilitates determination of glycoform sequence and branching. Developments in high-performance electrospray mass spectrometry include sample introduction at nanoliter flow rates, tandem magnetic sector/orthogonal time-of-flight instruments, Fourier transform instruments, and new ion optical strategies, including ion traps. Although a sensitive and important complementary technique, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry suffers from matrix-dependent deposition of excess internal energies, which produce extensive metastable fragmentation and (photo)adduct formation. These metastable fragments may be focused into a mass spectrum by employing an ion mirror (reflectron) in time-of-flight instrumentation. In favorable cases, structural information may be obtained. Address Department of Pharmaceutical Chemistry and the Liver Center, University of California, San Francisco, California 94143-0446, USA Current Opinion in Biotechnology 1996, 7:4-10 © Current Biology Ltd ISSN 0958-1669

Abbreviations ES electrospray GPI glycosylphosphoinositol HPLC high-pressureliquid chromatography LC liquid chromatography LSI liquid secondary ion MALDI matrix-assistedlaser desorption/ionization m/z mass-to-charge

Introduction An inherent advantage of mass spectrometric-based strategies for use in the structural characterization of protein glycosylation is their ability to detect unanticipated carbohydrate components that may not be observed by standard biochemical methods of analysis in common use. Since the advent of ionization techniques suitable for biopolymers, the discovery of new moieties or unanticipated structures has become commonplace. For example, (1---~6)-linkage branching of the (1---)3)-arm of the high-mannose Man9 structure in yeast mannan could not be detected using two-dimensional N M R or other methods of analytical

scrutiny prior to its detection by mass spectrometric investigation [1]. T h e classes of protein glycosylation are diverse, and each class poses its own distinct challenges to analysis [2]. T h e three most common of these protein modifications arise from co-translational or post-translational processes forming asparagine-linked glycopeptides [3], serine-linked or threonine-linked glycopeptides [4] and protein carboxyterminal elaboration with a glycosylphosphoinositol (GPI) lipid membrane anchor [51. Other less common types of glycosylation involve moieties such as 0-1inked N-acerylD-glucosamine [6], 0-fucose [7], 4-sulfo galactosyl [8] and xylose [91. In this field, the elucidation of rigorous structure/function correlations has been compromised by the difficulties of obtaining sufficient site-specific structural detail, a particularly problematic issue when only small quantities of natural glycoproteins are available. This is well illustrated by the challenge of elucidating the structural nature of the multivalent carbohydrates present on selectins and their ligands, which modulate inflammatory processes such as leukocyte adhesion to the endothelium and platelets [10]. T h e choice of strategy for structural characterization of protein glycosylation is intimately intertwined with the amount of glycoprotein readily available for analysis. In practical terms, circulating and recombinant glycoproteins are usually not sample-limited, unlike many other natural glycoproteins of major importance such as cell-surface receptors. If the sample size available is in the micromole range, then methods of N M R coupled with mass spectrometry will provide the most detailed and revealing information [11,12",13",14-18]. If the quantities are limited to the nanomole to subpicomole range, however, strategies involving mass spectrometric methods, together with various combinations of separation techniques and selective enzymic/chemical degradation and derivatization methods, are the most effective [2,19-22,23°]. More than one way can always be found to employ established techniques to solve a given problem, but the mass spectrometric techniques, in particular, continue to develop at a rapid pace, as considered below. T h e choices to be made and manner in which they are employed depend, to a large extent, on the exact nature of the problem, the amount of biopolymer readily at hand, and the particular techniques available to the investigator. This review is concerned with the use of enzymic or chemical digestions and the techniques of mass spectrometry for the structural characterization of asparagine-linked

Characterization of protein glycosylationby mass spectrometry Burlingame

glycosylation, serine/threonine-linked glycosylation and GPI-modified peptides, with highlights from the current literature.

Oligosaccharides Until recently, mass spectrometric studies of intact glycoproteins were impossible, and even studies at the glycopeptide level were severely compromised by the relatively low ionization efficiency (sensitivity) and limited effective mass range of the only 'soft' ionization technique available, liquid secondary ion (LSI) mass spectrometry (or fast-atom bombardment). Of necessity, for almost a decade, attention in most laboratories was focused on methods that involved quantitative liberation of the oligosaccharides from the glycoprotein (or glycopeptide), either enzymatically or chemically [20,21], followed by separation of these mixtures and isomers by anionic exchange chromatography [22]. Because mass spectrometric techniques require that the analyte be surface-active in a viscous matrix, the extreme hydrophilicity of the free oligosaccharides obtained from these methods necessitated the use of derivatization to enhance hydrophobicity. This was accomplished either by derivatization of all labile hydrogens (e.g. permethylation or peracetylation) [21] or by reductive amination with a hydrophobic chromophore [24]. These methods are still in common use in several laboratories [25,26,27",28-30].

Glycoproteins T h e more recent developments in matrix-assisted laser desorption/ionization (MALDI) [31°-34",35-39] and electrospray (ES) ionization [39-43] permit direct measurement of the molecular weight of glycoproteins in favorable cases. Spectra derived from both methods have been obtained for the same glycoprotein in two recent studies [39,40]. Treatment of such glycoproteins with glycolytic enzymes and subsequent re-measurement of their molecular weight profiles usually result in sharper and more-accurate measurements. This effect results from selective elimination of non-reducing terminal heterogeneity (truncation of the antennae) [44] or, as a final step in the case of asparagine (N)-linked glycosylation (discussed below), conversion of the glycoprotein to the corresponding aspartyl protein analog [45]. Mass centroid values from MALDI and stable isotope profiles from ES ionization (which provide higher mass resolutions and thus greater mass-measurement accuracies) can reveal global glycosylation content and/or glycoform heterogeneity directly [39] (assuming no protein heterogeneity [42]). Unlike ES ionization, which forms the lowest internal energy molecular ionic species (i.e. it is the 'softest' ionization technique), MALDI results in excess internal energy deposition and presents certain problems

5

in making such measurements with similar reliability [31",32"].

Glycopeptides In a remarkably short time since the advent of ES ionization, liquid chromatography (LC) mass spectrometry has become the methodology of choice for characterization of protein digests and may be carried out routinely at high sensitivity [23°,46,47]. These developments have revolutionized protein/glycoprotein mass mapping, site-specific assessment of global glycosylation (in studies ranging from fundamental investigations to detection of covalent modifications), and quality control for recombinant proteins destined for therapeutic usage [23"]. One of the most important practical benefits recently arising from this methodology involves the ease of direct selective detection of glycopeptides, among other components in a protein digest, at high sensitivity during an LC mass spectrometry run. It has been established that collisional deposition of sufficient excess internal energy to the pseudomolecular ions induces the formation of oxonium type ions from all glycopeptide components present [48,49,50",51",52]. This is achieved by continuously monitoring the specific mass values of such ions during the high-pressure liquid chromatography (HPLC) run. qpecific signals are recorded at these preselected mass values that 'flag' when glycopeptide components elute from the HPLC separation. For example, oxonium ions, such as HexNAc ÷ (mass-to-charge/m/z] ratio =204), HexHexNAc ÷ (m/z ratio=366), or NeuNAc-Hex-HexNAc ÷ (m/z ratio=657) (where Hex is hexose, HexNAc is N-acetylhexosamine and NeuNAc is sialic acid), may be used to locate the presence of particular suspected structural types, bearing in mind that isobaric isomeric moieties will not be distinguished de novo from such data. Because usually only -5% of a digest is consumed during a LC/ES mass spectrometry run, the remainder may be collected in fractions for further experiments. Simultaneously, successive mass spectra can be recorded using the higher mass portion of each full scan of the mass range, detecting pseudomolecular ions corresponding to each emerging eluent. Usually, when glycopeptides elute, the mass spectra reveal the presence of more than one oligosaccharide species (viz. glycoform) at any given glycosylation site (a phenomenon termed microheterogeneity). Under reverse-phase conditions, the retention times of glycopeptides are governed to a first approximation by the hydrophilicity:hydrophobicity ratio of the amino acids in the peptide portion in the commonly used water/acetonitrile/trifluoroacetic acid solvent system. As a result, the mixture of oligosaccharide species attached to a given site (i.e. glycoforms) will co-elute in a chromatographic peak broadened to some extent compared with eluting peptides. It should be noted that the separation of glycoforms attached to a single site may be achieved using an alternative solvent

6

Analytical biotechnology

system (ethanoi/propanol/water/formic acid) that is also fully compatible with ES mass spectrometry [53]. Further insight into the structural nature of the glycoforms attached at a given site may then be obtained by employing a series of glycosidases for sequential digestion of particular non-reducing terminal residues (see [54°]; Table 1) and again employing mass analysis after each digestion to ascertain the oligosaccharide heterogeneity and composition remaining on each glycopeptide [50"',51°°,54 .] In addition, tandem mass spectrometry may be employed to obtain complementary information [55°], as discussed further in connection with O-linked glycopeptides below. Analogous glycopeptide mass-mapping strategies have been explored using MALDI mass spectrometry, as well [56",57,58]. The sensitivity, resolution and mass-measurement accuracy is significantly poorer than that achieved routinely by ES mass spectrometry [50"',51"']. In addition, there is evidence that negatively charged glycopeptides (containing sialic acid or sulfate) do not behave as well using either polarity in MALDI. This can result, for example, in an unpredictable extent of loss of sialic acid [32°,58].

N-linked glycosylation In the case of N-linked glycosylation, the consensus sequence that provides the location of putative sites of occupation is known (Asn-XXX-Ser/Thr; three-letter amino acid code, where X is any residue). Three types of sugar residue composition and oligosaccharide branching pattern are common among mammalian proteins, namely, high mannose, complex and hybrid [3]. The type of oligosaccharide, the antennicity, and microheterogeneity present at each site of occupancy are most readily determined by use of information obtained from LC mass spectrometry experiments using ES ionization [50",51°°,53,59,60o°1. In addition, peptide-N-glycosidase F may be employed to liberate all of the three glycosylation types, thus yielding a mixture of free reducing terminal oligosaccharides intact [24,61-63], together with the aspartyl analog of the peptide in question [45]. Of course, the structural possibilities among non-mammalian N-linked glycoproteins may present considerably greater complexity that requires significant further structural investigation [27°].

O-linked glycosylation Glycoproteins bearing serine-linked and threonine-linked glycosylation pose special problems and challenges compared with those with solely N-linked glycosylation.

Table 1 Glycosldases and their linkage Sl)edficltles. Linkage Enzymes

Enzyme source

specificity*

Sialidase

Newcastle disease v i r u s

NeuNAca2-~3,8R NeuGc(x2-~4,8R

Arthrobacter ureafaciens

NeuNAc(t2-~6(>3,8)R NeuGco.2-~6(>3,8) R

~-galactosidase

Streptococcuspneumoniae Gall 134---~GInNAc

Concentration (U m1-1ormU ld-1)

pH Digestion optimum buffer

Digestion time (h)

0.02-0.05

5.0-6.0

30 mm sodium acetate, pH 5.0

16-22

0.8-1.0

5.0-5.5

ibid

ibid

0.03-0.07

5.5-6.5

ibid

ibid

Gall 134-->GalNAc 13-N-acetylChicken liver glucosaminidase

GIcNAc~I -->3,4R

1.0

4.0

30 mM sodium acetate, pH 5.0 or 100 mm sodium citrate phosphate, pH 4.0 and 250 mg m1-1 BSA

ibid

(~-Mannosidase

Jackbean

Man(t1 -->2,3,6Man

3.5

4.0-4.5

30 mM sodium acetate, pH 5.0 25 mM zinc chloride

ibid

~-Mannosidase

He~ixpomatia

Man~l -~4GIcNAc

2.0-4.0

4.0-4.5

ibid

(~-Fucosidase

Bovine epididymis

Fucotl--)6(>2,3,4)R

0.2

6.0-6.5

1O0 mM sodium citrate phosphate, pH 6.0 and 250 mg m1-1 BSA

15-22 16

*Parentheses around linkage sites, for example (>3,8), indicate that enzymes have lower rates of reactivity towards them. R is the attachment moiety on the non-reducing part of the glycopeptide or oligosaccharide. BSA, bovine serum albumin; Fuc, fucose; Gal, galactose; GIcNAc, N-acetylglucosamine; GalNAc, N-acetylgalactosamine;Man, mannose; NeuGc, N-glycolyl-neuraminicacid; NeuNAc, sialic acid.

Characterization of protein glycosyletion by mass spectrometry Burlingame

T h e s e problems include the lack of a reliable amino acid consensus sequence and availability of a general enzyme that will cleave all O-linked structures. This necessitates the use of less desirable experimental protocols, such as use of base-catalyzed elimination of oligosaccharides [64], with the loss of site-linkage specificity and concomitant destruction of the peptide itself, or, at this juncture, the adoption of tandem mass spectrometric strategies for analysis of such glycopeptides intact. These methods of tandem mass spectrometry are emerging as partial, but very important, solutions to these problems [65,66•,67,68•]. Even so, further complexities exist because of the frequency of semicontiguous hydroxyamino acids that are potential sites of occupancy as well as the common occurrence of having more than one of such sites occupied in particular peptide sequences. Often, this situation is complicated even more by a lack of sites in such peptide sequences permitting specific enzymic cleavage. Thus, one may be forced in these cases to resort to the use of non-specific proteases or chemical digests in conjunction with tandem mass spectrometric methods. In addition, some combination of Edman procedures and mass spectrometry may be required [69].

Glycosylphosphoinositol lipid membrane anchors Comparatively recently, a new type of cell surface protein membrane anchor has been discovered that consists of a GPI lipid [5]. Early uses of mass spectrometry [70-72] recognized heterogeneity of GPI glycans. In addition, use of ES ionization was important in detecting the presence of sialic acid on the scrapie glycan [71]. Structural elucidation of other GPI anchors has also been reported [72-74].

sacrificing mass resolution and sensitivity, permits the detection of low charge-state higher mass components, such as glycopeptides, as well as the direct determination of their charge state from stable isotope spacing.

Acknowledgements I wish to acknowledge the expertise and editorial assistance of Candy Stoner in library work and in production of the final copy for this manuscript. I am grateful for financial support from NIH National Center for Research Resources Grant 01614.

References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as: • of special interest • • of outstanding interest 1.

Hernandez LM, Ballou L, Alvarado E, Gillaos-Castro BL, Burlingame AL, Ballou CE: A new Saccharomyces cerevisiee mnn mutant N-linked oligosaccharlde structure. J Biol Chem 1989, 269:11849-11856.

2.

Settineri CA, Budingame AL: Mass spectrometry of carbohydrates and glycoconjugates. In Modem Chromatographic and Electrophoretic Methods in Carbohydrate Analysis. Edited by El Razzi Z. Amsterdam: Elsevier; 1995:447-514.

3.

Cummings RD: Synthesis of esperaglne-Ilnked ollgosaccharldes: pathways, geneUcs, and metabolic regulation. In G/¥coconjugates: Composition, Structure, and Function. Edited by Allen HI, Kisailus EC. New York: Marcel Dekker; 1992:333-360.

4.

Shachter H, Brockhausen I: The biosynthesis of sedne (threonlne)-N-acetylgalactosamlne-Iinked carbohydrate moieties, In G/¥coconjugates: Composition, Structure, and Function. Edited by Allen HI, Kisailus EC. New York: Marcel Dekker; 1992:263-332.

5.

Udenfriend S, Kodukula K: How glycosyl-phosphatldyllnositolanchored membrane proteins are made. Annu Rev Biochem 1995, 64:553-591.

6.

Hart GW, Haltiwanger RS, Holt GD, Kelly WG: Glycosylatlon In the nucleus and cytoplasm. Annu Rev Biochem 1989, 58:841-874.

7.

Harris R.I, Molony MS, Kwong MY, Ling VT: Identifying unexpected protein modifications, In Mass Spectrometry in the Biological Sciences. Edited by Burlingame AL, Carr SA. Totowa, New Jersey: Humana Press; 1996:333-350.

8.

Yamada S, Oyama M, Kinugasa H, Nakagawa T, Kawasaki T, Nakasawa S, Khoo KH, Dell A, Sugahara K: The sulfated carbohydrate-protein linkage region isolated from chondroltin 4-sulfate chains of inter-(~-bypsln Inhibitor In human plasma. Glycobiology 1995, 5:335-341.

g.

Nishimura H, Kawabata SI, Klaiel W, Hsse S, Ikenaka T, Takao 1", Shimonishi Y, Iwanaga S: IdentificaUon of a dlsaocharide (XleOGI¢) and a trlsaccharlde (Xyl2-GIc) O-glycosldlcally linked to a serlne residue in the first epidermal growth factor-like domain of human factors VII and IX and protein Z and bovine protein Z. J Biol Chem 1989, 264:20320-20325.

10.

Rosen SD, Bertozzi CR: The selectlns and their llgands, Curt Opin Cell Bio11994, 6:663-673.

11.

Lo-Guidica JM, Wieruszeski JM, Lemoine J, Verbert A, Roussel P, Lamblin G: Sialylatlon and sultation of the carbohydrate chains In respiratory muclns from a patient with cystic fibrosis. J Biol Chem 1994, 269:18'/94-18813.

Conclusions Throughout the past decade, mass spectrometry has emerged as an indispensable approach for revealing unprecedented detail in the delineation of the structural aspects of protein glycosylation. This information is essential for understanding the structural basis of glycoprotein activity in biological assays, where only a limited amount of sample is available. At this juncture, it is fair to point out the competitive disadvantage involved in tackling structural investigations of protein glycosylation without employing the current advanced methods of mass spectrometry. In addition, many technical advances are emerging that promise to provide further analytical power to deal with the myriad of post-translational challenges of proteins. T h e quality and preciseness of LC/ES mass spectrometry is improving in several ways; sensitivity, mass range and mass resolution are all progressing. This is a result of on-going developments in low flow rate ES sources [75], magnetic sector instruments possessing multichannnel array detection [76] and Fourier transform mass spectrometry systems [77,78]. T h e extended mass range of current mass spectrometry systems, without

7

12. •

Van Dam G J, Bergwerff AA, Thomss-Oatss JE, Rotmans JP, Kamerling JP, Vliegenthart JFG, Deelder AM: The immunologlcally reactive O-linked polysaccharlde chains derived from clrculaUng cathodic antigen Isolated from the human blood fluke Schistosoma mansoni have Lewis x as repeating unit. Eur J Biochem 1994, 225:467-482. These authors employ base-catalyzed elimination of O-glycans, together with tandem mass spectrometry of the permethylated derivative and NMR analysis, to characterize the glycosylation of circulating cathodic antigen (CCA) from Schistosoma mansonL Proton NMR was used to establish the atruc-

8

Analytical biotechnology

tures of the major oligoseccharides; mass spectrometry demonstrated that these alditols can be fucosytated and enabled determination of the structure and branching of an unusual minor component. This work establishes that the antigenicity results from the presence of the poly Lewisx determinant on CCA. 13. •

Teguchi T, Seko A, Kitajima K, Muto Y, Inoue S, Khco KH, Morris HR, Dell A, Inoue Y: Structural studies of a novel type of pentaantennary large glycan unit in the ferUllzatlon-associated carbohydrste-rlch glycopeptlde Isolated from the fertilized eggs of Oryzias latlpes. J Biol Chem 1994, 269:8752-8771. In this paper, extensive use is made of chemical degradation/derivatization methods, as well as LSI mass spectrometry using a 20-25 kV primary cesium ion beam and NMR techniques, to characterize the N-glycan structures in L-hyosophorin isolated from the fertilized eggs of Oryzias latipes and Oryzias me/astigma. This study represents the first detailed investigation establishing the highly branched nature of these N-linked pentantennary glycans. Expression of such bulky multiantennery structures in hyoeophorin molecules suggests that they may play a role in regulating cell-cell interactions of blastomeres at certain developing stages of embryos. 14.

Taguchi T, Kitajima K, Muto Y, Inoue S, Khoo KH, Morris HR, Dell A, Wallace RA, Selman K, Inoue Y: A precise structural analysis of a fertilization-associated carbohydrate-rich glycopeptlde Isolated from the fertilized eggs of euryhallne killl fish (Fundulus heteroclitus). Novel penta-antennary Nglycan chains with a bisecting N-ecetylglucosamlnyl residue. Glycobio/ogy 1995, 5:611-624.

15.

Nimtz M, Wray V, R0digar A, Conradt HS: Identification and structural characterization of a mannose-6-phosphate containing otlgomannosldlc N-glycan from human erythropoletin secreted by recombinant BHK-21 cells. FEB8 Lett 1995, 365:203-208.

15.

Hofsteenge J, MOiler DR, De Beer T, L6ffler A, Richter WJ, Vllegenthart JFG: New type of linkage between a carbohydrate and a protein: C-glycosylstlon of a specific tryptophan residue In human RNase Us. Biochemistry 1994, 33:13524-13530.

17.

Book K, Schuster-Kolbe J, Altman E, AIImsier G, Stahl B, Christian R, Sleytr UB, Messner P: Primary structure of the O-glycesidlcally linked glycan chain of the crystalline surface layer glycoproteln of Thermoaneerobacter thermohydrosulfuHcus L111-69. J Biol Chem 1994, 269:7137-7144.

18.

Zlegler FD, Gsmmill TR, Trimble RB: Glycoproteln synthesis in yeast. J BioI Chem 1994, 269:12527-12535.

19.

Budingame AL, Boyd RK, Gaskell SJ: Mass spectrometry. Anal Chem 1994, 66:634R-683R.

20.

Peter-Katalinic J: Analysis of glycoconjugates by fast atom bombardment mass spectrometry and related MS techniques. Mass Spectrom Rays 1994, 13:77-98.

21.

Dell A, Reason AJ, Khoc KH, Panico M, McDowell RA, Morris HR: Mess spectrometry of carbohydrste-contalnlng blopolymers. Methods Enzymol 1994, 230:108-132.

22.

Hardy MR, Townsend RR: High pH anion exchange chromatography of glycoproteln dedved carbohydrates. Methods Enzymol 1994, 230:208-225.

23. •

Hemling ME, Mentzer MA, Capiau C, Carr SA: A mulflfaceted strategy for the characterization of recombinant gD-2, a potential herpes vaccine. In Mass Spectrometry in the Biological Sciences. Edited by Burlingame AL, Carr SA. Totowa, New Jersey: Humans Press; 1996:30'/-331. This chapter illustrates the effective use of all of the newest techniques of mass spectrometry in the characterization of the glycosylation of the potential herpes vaccine recombinant gD-2, expressed in Chinese hamster ovary cells. It also includes a discussion of the tolerance and intolerance of these techniques to common buffers, salts, and detergents used in biochemical laboratories. 24.

Poulter L, Budingame AL: Desorption mass spectrometry of oligosecchaddes coupled with hydrophobl¢ ¢hromophores. Methods Enzymol 1990, 193:661-688.

25.

HoffmannA, Nimtz M, Wurster U, Conradt HS: Carbohydrate structures of B-trace protein from human cerebrosplnal fluid: evidence for 'breln-type' N-glycosylstlon. J Neurochem 1994, 63:2185-2196.

26.

Reason A J, Ellis LA, Appleton JA, Wianewski N, Grieve RB, McNeil M, Wassom DL, Morris HR, Dell A: Novel tyvelosecontaining td- and tetra-antennary N-glycans In the immunodominent antigens of the intrecellular parasite Trichinelle spire/is. Glycobio/ogy 1994, 4:593-603.

27. •

Khoo KH, Sarda S, Xu X, Caulfield JP, McNeil MR, Homans SW, Morris HR, Dell A: A unique multlfucosyIsted - 3GalNAc~I --)4GIcNAc~I -->3Galc~l - motif constitutes the repeating unit of the complex O-glycans dedved from the cercadal glycocalyx of Schistosoma mansoni. J Biol Chem 1995, 270:17114-17123. This paper is an excellent illustration of how the well established techniques of selective degradation and derivatization, coupled with mass spectrometry, can be used to reveal and define unanticipated structures in protein glycosylation. 28.

PfeifferG, Strube KH, Schmidt M, Geyer R: Glycosylstion of two recombinant human utedne tissue plasminogen activator variants carrying an additional N-glycosylation site in the epidermal-growth-factor-like domain. Eur J Biochem 1994, 219:331-348.

29.

Hogeland KE Jr, Deinzer ML: Mass spectrometric studies on the N-linked ollgosaccharldes of baculovlrus-expressed mouse Intedeukln-3. Biol Mass Spectrom 1994, 23:218-224.

30.

Stadie TRE, Chai W, Lawson AM, Byfield PGH, Hanisch FG: Studies on the order and site specificity of GalNA¢ transfer to MUC1 tandem repeats by UDP-GalNAc:polypeptlde Nacatylgalectosamlnyltransferase from milk or mammary carcinoma cells. Eur J Biochem 1995, 229:140-147.

31. •

Tserbopouios A, Karas M, Strupat K, Pramanik BN, NagabhushanTL, Hillenkamp F: Comparative mapping of recombinant proteins and glycoprotelns by plasma desorption end matrix-assisted laser desorption/Ionlzation mass spectrometry. Anal Chem 1994, 66:2062-2070. Deposition of excess intemal energy in analyte molecular ions during the MALDI process causes metastable fragmentation, the extent of which is matrix dependent. In practical terms, the analysis of glycopeptides may be compromised by the lability (partial loss) of sialic acid residues as well as other small neutral species. These authors show that this problem may be ameliorated to some extent using a 9:1 mixture of 2,5-dihydroxybenzoic acid (DHB) and 2-hydroxy-5-methoxybenzoic acid instead of neat DHB. 32. •

KarasM, Bahr U, Strupat K, Hillenkamp F, Tserbopouios A, PramanikBN: Matrix dependence of metastable fragmentation of glycoprotelns In MALOI TOF mass spec~ometry. Anal Chem 1995, 67:675-679. This contains a discussion of the relative stabilities or lifetimes of glycoproteins during the MALDI process using different matrices, the mass measurement accuracy problems associated with excess internal energy deposition causing metastable fragmentation (causing tailing on the low mass side of the spectrum) and problems associated with matrix adduct formation (causing tailing on the high mass side of the spectrum). Use of the least acidic matrix appears to minimize fragmentation, but causes more photoadduction. 33. •

KieliszewskiMJ, O'Neill M, Leykarn J, Orlando R: Tandem mass spectrometry and structural elucidation of glycopeptldes from a hydroxyprollne-rlch plant cell wall glycoproteln Indicate that contiguous hydroxyproline residues ere the major sites of hydroxyprollne O-areblnosylation. J BiG/Chem 1995, 270:2541-2549. Tandem mass spectrometry on an underivatized O-linked plant cell wall glycopeptide is used to determine which hydroxyproline residue is glycosylated. Precise hydroxyproline-arabinosylation motifs were established, suggesting s sequenceodependent, rather than conformation-dependent, enzymatic mechanism. 34. •

NakanishiT, Okamoto N, Tanaka K, Shimizu A: Laser desorptlon time-of-flight mass spectrometric analysis of transferrln precipitated with antiserum: a unique simple method to identify molecular weight variants. Biol Mass Spectrom 1994, 23:230-233. Worth reading as a cautionary example; these authors appear unaware of the problems of MALDI-induced partial and irreproducible losses of sialic acid (see [31",32",58]). This could compromise their assertion that MALDI is a reliable technique for screening carbohydrate-deficisnt glycoprotein syndrome and analogous disorders on the basis of solely the measurement of a glycoprotein's molecular weight profile. 35.

Lochnit G, Geyer R: Carbohydrate structure analysis of betroxobin, • thrombln-Ilke serlne protease from Bothrops moojeni venom. Eur J Biochem 1995, 228:805-816.

36.

Verges Romero C, Neudorfer I, Mann K, Sch~fer W: Pudflcatlon and amino acid sequence of amlnopeptldase P from pig kidney. Eur J Biochem 1995, 229:252-269.

37.

Gowda, DC, Jackson CM, Hensley P, Davidson EA: Factor Xactivating glycoproteln of Russell's viper venom. J Biol Chem 1994, 269:10644-I 0650.

38.

Mohr MD, B6rnsen KO, Widmer HM: Mstrlx-esslsted laser desorptlonllonlzatlon mass spectrometry: Improved matrix

Charectedzetion of protein glycosyletion by mass spectrometry Burlingame

for oligosaccharldes. Rapid Commun Mass Spectrom 1995, 9:809-814. 39.

MOller D, Domon B, Karas M, Van Ooetrum J, Richter WJ: Charectedzation and direct glycoform profiling of a hybrid plasminogen activator by matrix-assisted laser desorption and eleotrospray mass spectrometry:, correlation with highperformance liquid chromatographic and nuclear magnetic resonance analyses of the released glycans. Biol Mass Spectrom 1994, 23:330-338.

40.

Ashton DS, Bed(Jell CR, Cooper DJ, Craig SJ, Lines AC, Oliver RWA, Smith MA: Mass spectrometry of the humanized monodonal antibody CAMPATH 1H. Anal Chem 1995, 67:835-842.

41.

Petillot Y, Thibault P, Thielens NM, Rosei V, Lacroix M, Coddeville B, Spik G, Schumaker VN, Gagnon J, Arlaud G J: Analysis of the N-Ilnkod ollgosacchaddes of human C l s using eleotrospray Ionisetion mass spectrometry. FEBS Lett 1995, 358:323-358.

42.

Young NM, Watson DC, Yaguchi M, Adar R, Arango R, RodriguezArango E, Sharon N, Blay PKS, Thibault P: C-terrninal posttranslational proteolysls of plant leotlns and their recombinant forms expressed in Escherichie coil J Bio/ Chem 1995, 270:2563-2570.

43.

44.

45.

Lewis DA, Guzzetta AW, Hancock WS, Costello M: Characterization of humanized antl-TAC, an antibody directed against the Intadeukln 2 receptor, using electrospray Ionization mass spectrometry by direct infusion, LC/MS, and MS/MS. Anal Chem 1994, 66:585-595. Stahl B, Klabunde T, Witzel H, Krebs B, St•up M, Karas M, Hillenkamp F: The ollgosacchaddes of the Fe(lll)-Zn(ll) purple add phosphates• of the red kidney bean. Eur J Biechem 1994, 220:321-330. Carr SA, Barr JR, Roberts GD, Anumula KR, Taylor PB: Identification of attachment sitas and structural classes of asparagine-Ilnked carbohydrates in glycoprotalns. Methods Enzymol 1990, 193:501-518.

46.

Hemling ME, Roberts GD, Johnson W, Cart SA, Covey TR: Analysis of proteins and glycoprotelns at the plcomole level by on-line coupling of mlcrobore high-performance liquid chromatography with flow fast atom bombardment and electrospray mass spectromet~, a comparative evaluation. Biomed Environ Mass Spectrom 1990, 19:677-691.

47.

Ling V, Guzzetta AW, Canova-Davis E, Stults JT, Hancock WS, Covey TR, Shushan Bh Characterization of the tryptlc map of recombinant DNA dedved tissue plasmlnogen activator by high performance liquid chromatography-electrospray ionization mess spectrometry. Anal Chem 1991, 63:2909-2915.

48.

Huddleston MJ, Bean MF, Carr SA: Collislonal fragmentation of glycopeptides by electrospray Ionization LC/MS and LC/MS/MS: methods for selective detection of glycopeptides In protein digests. Anal Chern 1993, 65:877-884.

49.

50. ee

Carr SA, Huddleston MJ, Bean MF: Selective identification and dlfferentietion of N- and O-linked ollgoseccharldes in glycoprotelns by liquid chromatography-mass spectrometry. Protein Sci 1993, 2:183-196.

Schindler PA, Settineri CA, Collet X, Fielding C J, Budingame AL: Site-specific detection and structural characterization of the glycosylatlon of human plasma proteins lecithin:cholesterol acyltrensferase and apollpoprotein D using HPLC/electrospray mass speotromeby and sequential glycosides• digestion. Protein Sci 1995, 4:791-801. By raising the nozzle-skimmer voltage in the ES source, collision-induced dissociation is employed to detect the elution times of glycopeptide components in the protein digest and to monitor for the appearance of carbohydrate-specific oxonium ions. This technique was developed by Carr and co-workers [48,49] and is further illustrated in their comprehensive work on the glycosyletion of monocional antibodies [51"]. The approach is of general importance and is particularly effective and well suited for detecting unsuspected glycopeptides in protein digests, as exemplified in this study by the detection of O-linked glycosyletion on lecithin:choleeterol acetyl transfersse and the presence of Apo D in the preparation. This paper describes the use of sequential digestion of glycopeptides by glycosidasos of appropriate specifiolties with concomitant analysis of the remaining glycoforms by ES mass spectrometry. A note of caution must be raised concerning the need to establish the epecifioity and purity of commercial glycosides• enzyme preparations used in such studies. 51. Roberts GD, Johnson WP, Burman S, Anumula KR, Cart SA: An o• integrated strategy for structural characterization of the protein

9

and carbohydmta components of monoclonal antibodies: application to anti-respiratory syncytial virus mAb. Anal Chem 1995, 67:3613-3625. This paper illustrates the power of LC/ES mass spectrometry for the detection of post-transletional modifications and their detailed characterization using a recombinant reshaped human monocional antibody. The authors demonstrate how this information fits into an integrated strategy comprising most current methods for structural characterization of a recombinant glycoprotein. Use of analogous LC/ES mass spectrometry protocols is widespread in the biopharmaceutical industry for characterization of recombinant glycoproteins destined for therapeutic use as well as for subsequent quality control requirements. It should be noted, however, that the potential problems involving analysis of non-reducing terminal sialic acid residues by MALDI techniques were not at issue in this study. 52.

Bean MF, Annan RS, Hemling ME, Mentzer M, Huddleston MJ, Can" SA: LC-MS methods for selective detection of posttranslational modifications In proteins: glycosylaflon, phosphorylaflon, suItstion, and acylation. In Techniques in Protein Chemistry VI. Edited by Crabb JW. San Diego: Academic Press; 1995:107-116.

53.

Medzihradszky KF, Maltby DA, Hall SC, Settineri CA, Burlingeme AL: Characterization of protein N-glycosylation by reversed-phase mlcrobore liquid chromatogrephy/eleotrospray mass spectrometry, complementary mobile phases, and sequential exoglycosldase digestion. J Am Soc Mass Spectrom 1994, 5:350-358.

54. •

Settineri CA, Buifingame AL: Strstegles for the characterization of carbohydrates from glycoprotains by mass spectrometry. In Techniques in Protein Chemistry V. Edited by Crabb JW. San Diego: Academic Press; 1994:9'7-113. This chapter shows the use of sequential glycosidsse digestions monitored by mass spectrometry to investigate recombinant extracellular domains of platelet-derived growth factor and nerve growth factor receptors expressed in Chinese hamster ovary cells. 55. •

Peter-Ketalinic J, Williger K, Egge H, Green B, Haniech FG, Schindler D: The application of electrospray mass spectromeby for structural studies on a tatracacchedde monopeptlde from the urine or a patient wlth a N-ecetylhexosamlnldase deficiency. J Carbohydr Chem 1994, 13:447-456. Details the use of both negative-ion and positive-ion ES mass spectrometry for the characterization of glycopeptides containing sialic acid. Enhanced information derived from collisionaJ activation in the ES source is described in the negative-ion mode. 56. •

Sutton CW, O'Neill J, Cottrell JS: Slta-speclfic characterlzatlon of glycoproteln carbohydrates by exoglycosldese digestion and laser desorption mess spectrometry. Anal Biochem 1994, 218:34-46. This paper illustrates that MALDI, similar to ES mass spectrometry [50",51 "], may be employed to characterize the gtycoforms present in glycopeptide fractions using combinations of glycoaldase digestions and concomitant mass analysis of the truncation products. 57.

Derby PL, Wypych J, Rush RS, Clogston CL, Rohde MF: Site-specific glycosylation of recombinant stem cell factor. In Techniques in Protein Chemistry V. Edited by Crabb JW. San Diego: Academic Press; 1994:89-96.

58.

Huberty MC, Vath JE, Yu W, Martin SA: Site-specific carbohydrate Identification in recombinant proteins using MALD-TOF MS. Anal Chem 1993, 65:2791-2800.

59.

Linsley KB, Chan SY, Chan S, Reinhold BB, Lisi PJ, Reinhold VN: Applications of electrospray mass spectrometry to erythropoietin N- and O-linked glycans. Anal Biochem 1994, 219:207-217,

60. oo

Rush RS, Derby PL, Smith DM, Merry C, Rogers G, Rohde MF, Katta V: Mlcrohetarogeneity of erythropoletln carbohydrate structure. Anal Chem 1995, 67:1442-1452. This work illustrates the advantages of LC/ES mass spectrometry in providing a global assessment of intact glycosylation microheterogeneity with minimal sample handling and alteration of the original structural integrity o! glycosylation. An evaluation of O-acetylation of sialic acid residues is presented. 61.

Ip CCY, Miller WJ, Silberklang M, Mark GE, Ellis RW, Huang L, Glushka J, Van Halbeek H, Zhu J, Alhadeff JA: Structural characterization of the N-glycans of a humanized anti-CD18 murlne immunoglobulln G. Arch Biochem Biophys 1994, 308:387-399.

62.

Oxley D, Becic A: Mlcroheterogenelty of N-glycosylatlon on a styler self-Incompatibility glycoprotetn of Nicotiane a/eta. Glycobio/ogy 1995, 5:517-523.

10

Analytical biotechnology

63.

Hoffmann A, Nimtz M, Getzlaff R, Conradt H: 'Braln type' Nglycosylatlon of aslalo-trensferrln from human cerebrosplnal fluid. FEBS Lett 1995, 359:154-I 68.

64.

Carlson DM: Structures and Immunochemlcel propertles of ollgosaccharldes isolated from pig submaxlllary muclns. J Biol Chem 1968, 243:616-626.

65.

Medzihradszky KF, Gillece-Castro BL, Settineri CA, Townsend RR, Masiarz FR, Budingame AL: Structure determlnatlon of O-llnked glycopeptldes by tandem mass spectrometry. Biomed Environ Mass Spectrom 1990, 19:777-781.

Agarwala KL, Kawabata SI, Takao T, Murata H, Shimonishi Y, Nishimura H, Iwanaga S: Activation peptlde of human factor IX has oligosaccharides O-glycosldlcelly linked to threonlne residues at 159 and 169. Biochemistry 1994, 33:5167-5171. LSI mass spectrometry and high-energy collision-induced dissociation with multichannel array detection are employed to aid in establishing the structure of the O-linked glycopeptide of the activation peptide of human factor IX. 67. Rummer TH, Tarentino AL, Hauer CR: Novel, specific O-glycosylatlon of secreted Revobacterium meningosepticum proteins. J Biol Chem 1995, 270:13192-13196.

71.

Stahl N, Baldwin MA, Tapiow DB, Hood L, Gibson BW, Burlingarne AL, Prusiner SB: Structural studles of the scraple prlon proteln using mass spectrometry and amino acid sequenclng. Biochemistry 1993, 32:1991-2002.

72.

Taguchi R, Hamakawa N, Harada-Nishida M, Fukui T, Nojima K, Ikezawa H: Mlcrohetarogenelty In glycosylphosphatldyllnositol anchor structures of bovlne llver 5'-nucleotldese. Biochemistry 1994, 33:1017-I 022.

73.

Redman CA, Thomss-Oates JE, Ogata S, Ikehara Y, Ferguson MAJ: Structure of the glycosylphosphatidyllnosltol membrane anchor of human placental alksllne phosphatase. Biochem J 1994, 302:861-865.

74.

Treumann A, Lifely MR, Schneider P, Ferguson MAJ: Primary structure of CD52. J BIOl Chem 1995, 270:6088-6099.

75.

Wilm M, Houthaeve T, Talbo G, Kellner R, Mortanssn P, Mann M: Approaches to the practical use of MS/MS In a proteln sequencing fadllty. In Mass Spectrometry in the Biological Sciences. Edited by Burlingame AL, Carr SA. Totowa, NJ: Humana Press; 1996:245-265.

76.

Burlingame AL, Medzihradszky KF, Claussr KR, Hall SC, Maltby DA, Walls FC: From protein primary sequence to the gamut of covalent modifications using mess spectrometry. In Biological and Biotechnological Applications of ESI.MS. Edited by Snyder P. Washington, DC: ACS Books; 1996:472-511.

77.

O'Connor PB, Speir JR Senko MW, Liffie DP, McLafferty FW: Tandem mass spectrometry of carbonic snhydrase (29 kDa). J Mass Spectrom 1995, 30:88-93.

66. •

68. •

Reinhold BB, Hauer CR, Plummer TH, Reinhold VN: Detailed structural analysis of a novel, specific O-linked glycan from the prokaryote Revobecterium meningosepticum. J BIOl Chem 1995, 270:13197-13203. This paper and [67] report the characterization of an unusual @linked heptasaccharide isolated from hydrolases of Flavobacterium. The authors employ ES mass spectrometry and ES collison-induced dissociation mass spectrometry of the intact glycopeptide and an array of chemical degradations and derivatives. 69.

Bulet P, Hegy G, Lambert J, Van Dorsselaer A, Hoffmann JA, Hetru C: Insect Immunity. The Inducible antibacterial peptlde dlptercin carries two O-91ycens necessary for biological activity. Biochemistry 1995, 34:7394-7400.

70.

Baldwin MA, Burlingame AL, Prusiner SB: Mass spectrometric analysis of a GPI-anchored protein: the screple prlon protein. Trends Anal Chem 1993, 12:239-248.

78.

Smith RD, Bruce JE, Wu Q, Chang X, Hofstadler SA, Anderson GA, Chen R, Bakhtiar R, Van Orden SO, Gale DC et el.: The role of Fouder transform Ion cyclotron resonance mass spectrometry In biological research-new developments and applications. In Mass Spectrometry in the Biological Sciences. Edited by Burlingame AL, Carr SA. Totowa, New Jersey: Humans Press; 1996:25-68.