the journal of vol. 2, 14, pp. by for in u. s. a. o-glycosylation … · 2001-06-25 ·...

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 269, No. 2, Issue of January 14, pp. 910-920,1994 Printed in U. S. A.

O-Glycosylation Mimics N-Glycosylation in the 16-kDa Fragment of Bovine Pro-opiomelanocortin THE MAJOR O-GLYCAN ATTACHED TO THR-45 CARRIES S04-4GalNAcpl-4GlcNAcpl-, WHICH IS THE ARCHETYPAL NON-REDUCING EPITOPE IN THE N-GLYCANS OF PITUITARY GLYCOHORMONES*

(Received for publication, July 21, 1993, and in revised form, September 1, 1993)

Rosa A. Siciliano, Howard R. Morris$, Hugh P. J. BennettGll, and Anne Dell$ From the Department of Biochemistry, Imperial College of Science Technology and Medicine, London S W7 2AZ, United Kingdom, and §Royal Victoria Hospital, Montreal, Quebec H3A IAl . Canada

The NHz-terminal domain of pro-opiomelanocortin, designated as the 16-kDa fragment, is highly con- served throughout the vertebrate family and is likely therefore to have an important functional role. Bovine 16-kDa fragment is a 77- residue glycopeptide, which has been found to be glycosylated at threonine 45 and asparagine 65. Available evidence suggests that glycoforms lacking glycans at the O-linked site are processed in the intermediate pituitary at -Arg4’-Lyseo- to give the residue 1-49 amino-terminal peptide and a carboxyl-terminal glycopeptide referred to as Lys1y3- melanotropin. Glycoforms carrying O-glycans remain unprocessed in the intermediate pituitary. Thus 0- glycosylation is likely to play an important role in controlling the fate of the NH2-terminal portion of pro- opiomelanocortin, thereby affecting the biological events that are influenced by peptides and glycopeptides derived from this domain. In a recent study (Si- ciliano, R. A., Morris, H. R., McDowell, R. A., Azadi, P., Rogers, M. E., Bennett, H. P. J., and Dell, A. (1993) Glycobiology 3, 225-239), we sequenced the N-glycans attached to Asn-65 of bovine 16-kDa fragment and demonstrated that the acidic components contain, in addition to neutral antennae, a single sod- 4GalNAc/31-4GlcNAc/31- antenna, which is characteristic of the pituitary glycohormone N-glycans (Baen- ziger, J. U., and Green, E. D. (1988) Biochim. Biophys. Acta 947, 287-306). We now report the structural characterization of the O-linked oligosaccharides found in bovine 16-kDa fragment. The major component, which constitutes about 80% of the O-glycan population, is a novel sulfated tetrasaccharide, which carries the same sulfated epitope as the N-glycans. This is the first time that the SOr-4GalNAc/31-4GlcNAc/31- moiety has been observed in O-glycans, and it raises the interesting possibility that the B-N-acetylgalacto- saminyltransferase responsible for the addition of N- acetylgalactosamine to the pituitary glycohormones (Smith, P. L., and Baenziger, J. U. (1988) Science, 242,930-933) might be capable of glycosylating both N- and O-linked acceptors. ~~

* The Imperial College work was supported by the Medical Re- search Council (program grant to H. R. M. and A. D.) and the Wellcome Trust (Grant 030826/1.27/LA/GR to H. R. M. and A. D.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

8263; Fax: 44-71-225-0458. $ To whom correspondence should be addressed. Tel. 44-71-225-

ll Supported by Operating Grant MT-6733 from the Medical Re- search Council of Canada.

Pro-opiomelanocortin (POMC),l the common precursor for corticotropin (ACTH), endorphin, and related peptides secreted by the pituitary, has been known since its discovery to be a glycoprotein (Eipper and Mains, 1980). It was Seidah and co-workers (Seidah et al., 1981) who first discovered that the amino-terminal biosynthetic fragment of POMC bore both 0- and N-linked glycans at threonine 45 and asparagine 65, respectively (Fig. 1). In addition, a second N-linked glycosylation site was localized to the carboxyl-terminal region of rat and mouse ACTH (Mains and Eipper, 1976; Eipper and Mains, 1978). Glycosylation of ACTH is confined to these species since the structural cue for N-glycosylation ( i e . -Am- X-Ser/Thr) is missing from the ACTHs of other mammalian species (Kawauchi, 1983).

Within the anterior pituitary, the 16-kDa or amino-terminal fragment is a major product of POMC processing and is not subject to further cleavage (Seidah et al., 1981; Keutmann et al., 1981). Structural analysis of human 16-kDa fragment derived from the anterior pituitary revealed that it is glycosylated a t both 0- and N-sites with high efficiency (Bennett et al., 1986). In the bovine intermediate pituitary, the 16-kDa fragment consists of 77 residues and has been found to be partially cleaved at an -Arg4’-Lys5’- processing site to yield an amino-terminal49-residue peptide and a carboxyl-terminal peptide consisting of residues 50-77 (James and Bennett, 1985). This latter peptide, known as Lysly,-melanotropin (Lys1y3MSH), has been shown to be glycosylated at asparagine 65 (Bennett, 1986a). In contrast, structural analysis of the bovine residue 1-49 amino-terminal peptide yielded no evidence for O-glycosylation at threonine 45 (Bennett, 1984). The homologous peptides isolated from rat and mouse intermediate pituitaries have also been found to lack O-glycans (Bennett, 1986b; Seger and Bennett, 1986). Characterization of the 16-kDa fragment, which remains unprocessed in the bovine intermediate pituitary, indicated that this glycopeptide is fully O-glycosylated at threonine 45 (James and Bennett, 1985). These observations prompted the suggestion that cleavage of the processing site at -Arg4’-Lys5’- is dependent upon the absence of O-glycosylation at the adjacent threonine 45. Since cleavage at this site does not occur in the anterior pituitary, it was suggested that this putative regulatory mech- anism might account at least in part for the observed tissue specific processing of 16-kDa fragment (Seger and Bennett,

’ The abbreviations used are: POMC, pro-opiomelanocortin; ACTH, adrenocorticotropin; FAB, fast atom bombardment; GC, gas chromatography; Hex, hexose; HexNAc, N-acetylhexosamine; HPLC, high performance liquid chromatography; 16-kDa fragment, the amino-terminal fragment of pro-opiomelanocortin; MSH, melanotropin stimulating hormone; MS, mass spectrometry; NeuGc, N-glyco- lylneuraminic acid; PCE, pro-hormone converting enzyme.

910

0-Glycosylation of Bovine POMC 911

1 77 80 103 106 144147 265 16 K FRAGMENT

FIG. 1. Diagram showing the major domains of POMC together with the sequence of the 16-kDa domain of the bovine polypeptide. The 0- and N-glycosylation sites at positions 45 and 65, respectively, are bored. Residues 48- 50 are believed to constitute the tripeptide motif, which is recognized by the transferase that cdds GalNAc to the N- glycans a t Asn-65.

21

E Q P L m E N P R K Y V M G H F R W D R F G R R H G S S S S G V G G A A Q 61 65 n

1986). However, the use of site-directed mutagenesis technology has indicated that 0-glycosylation is most likely not a determinant in the regulation of processing of the 16-kDa fragment in the anterior pituitary corticotroph. Neither wild- type POMC nor POMC bearing an alanine for threonine substitution at position 45 was cleaved at -Arg4'-Lys5'- by mouse AtT-20 tumor cells, which are known to process POMC in a manner similar to that of the anterior pituitary (Noel et al., 1991). Therefore, it is more likely that 0-glycosylation plays a critical role in the regulation of 16-kDa processing within the intermediate pituitary rather than its anterior counterpart. Indeed, a putative pro-hormone converting enzyme (PCE), an aspartyl protease isolated from bovine pituitary tissue, has been shown in vitro to have the appropriate specificity with respect to cleavage of the bovine 16-kDa fragment. The peptide lacking the 0-glycan structure was found to be a substrate for PCE, whereas the fully glycosylated peptide was unprocessed (Birch et al., 1991). PCE is thought to be the mammalian homologue of the YAP-3 aspartyl protease found in yeast, which has been shown to process POMC at pairs of basic residues in vitro (Azaryan et al., 1993). Whether PCE is involved in POMC processing in vivo remains an open question, especially since recent studies of the properties of pro-hormone convertase 2, a member of the mammalian family of subtilisin-like proteases, have produced evidence that pro-hormone convertase 2 could be responsible for all the POMC cleavages occurring specifically within the intermediate lobe including the processing of the 16-kDa fragment (Zhou et al., 1993).

0-Glycosylation of the 16-kDa fragment may also play a role in determining its fate following secretion. It has been proposed that cleavage of the 16-kDa fragment by an adrenal cortical protease generates a peptide corresponding to the amino-terminal residue 1-49 sequence. This peptide in turn has been proposed to act as a locally acting mitogen for cells of the adrenal cortex (Estivariz et al., 1982; Lowry et al., 1983). If the adrenal cortical protease is inhibited in an analogous manner to its pituitary counterpart by glycosylation at threonine 45, it is probable that the presence of 0-glycans will be critical in the expression of the mitogenic activity of the 16- kDa fragment.

The acquisition of structural data on the 0-glycans is an essential prerequisite to studies aimed at clarifying the role 0-glycosylation plays in 16-kDa processing and function. We have recently used strategies based on fast atom bombardment mass spectrometry (FAB-MS) to characterize the N- glycans present in bovine 16-kDa fragment (Siciliano et ul., 1993). The major acidic N-glycans are biantennary structures having S04-4GalNAcpl-4GlcNAcp1- on the 3-arm and a neutral 6-arm that is fucosylated in some glycoforms. The sulfated epitope is characteristic of pituitary glycohormone N-glycans (Baenziger and Green, 1988). We now present results from a similar mass spectrometric study of the 0-linked oligosaccharides of bovine 16-kDa fragment and provide rigorous evidence for the most abundant 0-glycan being a tetrasaccharide

bearing the same S04-4GalNAc~l-4GlcNAc~l- epitope. This structure has not been previously detected in the 0-glycans of any glycoprotein. We discuss the possibility that the same enzymes could be responsible for the synthesis of the sulfated P-GalNAc moiety in both the N- and 0-glycans of the 16-kDa fragment.

EXPERIMENTAL PROCEDURES

Preparation of 16-kDa Fragment-To determine the 0-glycan structure of the 16-kDa or amino-terminal fragment of bovine POMC, two preparations of glycopeptide were used. The first was a portion of that used to determine the N-glycan structure linked to asparagine 65 (Siciliano et al., 1993). For this preparation, 600 bovine posterior pituitaries were extracted in acid and the solubilized peptides subjected to both reversed-phase and anion-exchange batch fractionation. Final purification of the 16-kDa fragment was achieved using preparative and analytical HPLC (Siciliano et al., 1993). To complete the 0-glycan structure, a second preparation of 16-kDa fragment was used. 100 lyophilized bovine posterior pituitaries (PelFreez Biolog- icals, Rogers, AR) were extracted, and the solubilized glycopeptide was purified as described previously (James and Bennett, 1985; Ben- nett, 1986a) with the exception that the batch anion-exchange step was eliminated.

Reduction and Carbozymethylation-Reduction of the 16-kDa fragment was carried out in Tris buffer (0.6 M pH 8.4) for 1 h at 37 "C using a 4 M excess (over cysteine residues) of dithiothreitol. Carbox- ymethylation was carried out in the same buffer using a &fold molar excess over total thiol for 30 min a t room temperature. Desalting was achieved using a C,, Sep-Pak (Waters) previously primed with methanol, propan-1-01, and 5% acetic acid and eluted with 40% propan-l- 01 in 5% acetic acid.

V8 Protease Digestion-Reduced and carboxymethylated 16-kDa fragment was digested with V8 protease (1:25 enzyme:substrate) in 100 p1 of ammoniun bicarbonate buffer (50 mM, pH 8.4) for 6 h a t 40 "C. The reaction was stopped by freeze-drying, and the enzyme was inactivated by boiling the sample for 5 min.

Peptide N-Glycosidase F Digestion-Peptide N-glycosidase F digestion was carried out in ammonium bicarbonate buffer (50 mM, pH 8.4) for 16 h a t 37 "C using 2.5 p1 of enzyme solution. Lyophilized sample was dissolved in 5% acetic acid and loaded onto a Cla Sep- Pak and eluted with 5% acetic acid followed by 40% propan-1-01 in 5% acetic acid. N-Linked oligosaccharides were eluted in the 5% acetic acid fraction, and peptides and 0-linked glycopeptides were eluted in the 40% propan-1-01 fraction.

Reductiue Elimination-The 40% propan-1-01 fraction was treated with 400 p1 of 1 M sodium borohydride (NaBHJ, 0.05 M NaOH a t 45 "C overnight. The sample was then neutralized with glacial acetic acid and desalted on a Dowex (500W X 8(H)) column, and the borates were removed by co-evaporation under N, with 10% acetic acid in methanol.

Permethylation-Oligosaccharides were permethylated using the sodium hydroxide procedure as previously described (Dell, 1990). Briefly 0.5-1.0 ml of dimethyl sulfoxide/NaOH slurry (prepared by grinding 5 pellets of sodium hydroxide with approximately 3 ml of dry M e 8 0 1 was added followed by about 0.5 ml of methyl iodide. The mixture was shaken for 10 min at room temperature and then quenched with 1 ml of water, and the products were extracted into chloroform, which was washed several times with water. The chloroform layer was dried under a stream of nitrogen, and the residue was dissolved in methanol for FAB-MS analysis.

Perdeuteroacetylation-Base-catalyzed peracetylation was per- formed using 1-methyl imidazole/deuteroacetic anhydride (15 (v/v) ratio, 30 p1 total volume) a t room temperature for 1 h. The reagents

912 0-Glycosylation of Bovine POMC were dried down under nitrogen. Sulfated oligosaccharides were eluted in the 30% fraction from a CIS Sep-Pak, and neutral and sialylated oligosaccharides were eluted in the 50% acetonitrile fraction.

Periodate Oxidation-The reduced 0-linked glycans were treated with 20 1.r1 of 10 mM sodium metaperiodate (NaI04) in acetate buffer (0.1 M, pH 5.5) at room temperature for 4 h in the dark. The reaction was quenched with 2 pl of ethylene glycol and left to stand for 30 min. The glycans were reduced with 10 mg/ml NaBH4 in 2 M NH40H at room temperature for 2 h before being subjected to perdeuteroacetylation and desalting using a Cl8 Sep-Pak.

Removal of Sulfate Groups-Removal of sulfate groups was carried out in 0.5 M HCl in methanol at room temperature for 3 h. The reagents were dried down under nitrogen.

Eroglycosydase Digestions-Digestion with exo-P-D-galactosidase (EC 3.2.1.23, 0.01 unit) from bovine testes was carried out in 200 pl of 5 mM sodium citrate-phosphate buffer, pH 4.6, containing 10 pl of toluene at 37 "C for 48 h. Digestion with exo-0-N-acetyl-D-glucosaminidase (EC 3.2.1.30,O.l unit) from bovine kidney was carried out in the same buffer at 37 "C for 48 h. After each digestion, aliquots (10%) were taken. The remaining oligosaccharides produced by the enzymatic reactions were deuteroacetylated and desalted by Cl8 Sep- Pak. In sequential digests, enzymes were deactivated by boiling the sample for 5 min prior to addition of the next enzyme.

HPLC Fractionation of the Neutral Glycans-The neutral glycans eluted in the 50% fraction were fractionated by reverse phase HPLC using a Shisheido C l S column attached to a Kontron HPLC system (model 450 data system, model 420 HPLC pump, model 430 detector, and model M800 mixer). The sample was loaded in 10% aqueous acetonitrile (v/v) and eluted with a solvent system of MilliQ water (A) and acetonitrile (B) (Rathburn) at a flow rate of 1 ml/min, isocratically with 10% B for 5 min, then with linear gradients of 10- 30% B in 5 min, 30-80% B in 60 min, and 80-100% B in 5 min. The elution was monitored at 214 and 280 nm, and 1-ml fractions were collected, dried, and redissolved in methanol for FAB-MS screening. The tetra- and pentasaccharides eluted in fractions 23-29 with major molecular ions for HexlHexNAc3 (m/z 1332), FuclHexlHexNAcs (m/ z 1568), and HexzHexNAc2 (m/z 1336) being observed in fractions 23, 27, and 29, respectively.

Chromium Trioxide Oxidation-The deuteroacetylated sample was dissolved in 100 pl of glacial acetic acid, followed by the addition of 10 mg of chromium trioxide. The resulting mixture was shaken at 50 "C for 1 h, then the reaction was quenced with 2 ml of water. The products were extracted into chloroform and washed several times with water, and the chloroform was dried down under a stream of nitrogen.

FABIMS-Low mass positive ion mode FAB-MS spectra were acquired using a VG analytical high field ZAB-HF mass spectrometer fitted with an "Scan gun operated at 10 kV. Spectra were recorded on oscillographic chart paper and counted manually, giving nomina1 mass assignments. The negative ion mode FAB spectra and the positive ion mode high mass region spectra were acquired using a ZAB-2SE FPD mass spectrometer fitted with a cesium gun. Spectra from this instrument were computer-processed, giving accurate mass assignments, which are rounded to the nearest whole number in most of the spectra that are reproduced in this paper. The matrix was thioglycerol, and the derivatized oligosaccharides were dissolved in methanol prior to loading on the target.

Deriuatization for Linkage Analysis-Partially methylated alditol acetates for GC-MS analysis were prepared from permethylated oligosaccharides according to the procedure described by Albersheim (Albersheim et al., 1967). Briefly, the permethylated samples were hydrolyzed with 2 M trifluoroacetic acid for 2 h at 121 "C, reduced with 10 mg/ml sodium borodeuteride (NaBDa) and 2 M NH,OH at room temperature for 2 h, and then acetylated with acetic anhydride at 100 "C for 1 h.

GC-MS Analyses-GC-MS were carried out using a Hewlett Pack- ard model 5890 gas cromatograph connected to a VG Trio 1 quadru- pole mass spectrometer. The partially methylated alditol acetates were dissolved in dichloromethane prior to injection on a DB-5 (25 m X 0.32 mm, internal diameter, J&W Scientific) column at room temperature. The temperature was then increased to 100 "C over 1 min and held for 1 min before increasing to 290 "C at 8 "C/min.

RESULTS

Isolation of 0-Glycans-Reduced carboxymethylated 16- kDa fragment was digested sequentially with V8 protease and peptide N-glycosidase F, and the released N-glycans were

separated from the peptides and 0-linked glycopeptide using a Sep-Pak. The included fraction, which contained the 0- glycopeptide, was reductively eliminated, and the released 0-glycans were purified by Dowex chromatography and converted to their deuteroacetyl derivatives using N- methylimidazole/deuteroacetic anhydride (Khoo et al., 1993). Upon Sep-Pak purification, derivatized sulfated oligosaccharides were recovered in the 30% acetonitrile fraction, while the remaining components eluted in the 50% acetonitrile fraction. Sugar analysis of the 30% and 50% fractions indicated that at least 80% of the oligosaccharides were recovered in the 30% fraction (data not shown).

Characterization of the Sulfated Fraction by FAB-MS and Linkage Analysis-The positive and negative FAB mass spectra of the 30% acetonitrile fraction are reproduced in Fig. 2. The negative spectrum (Fig. 2a) is dominated by an (M - H)- cluster centered at m/z 1365 corresponding to SIHexlHex- NAczHexNAcitol. The Gaussian-like appearance of the molecular ion cluster is caused by a portion of the deuteroace- tylating reagent being incompletely deuterated, resulting in the signals on the low mass side of the m/z 1365 cluster. Those on the high mass side are predominantly due to natural abundance 13C, although it is possible that there is a minor amount of sulfated HexzHexNAclHexNAcitol (m/z 1369 (M - H)") in this preparation because the isotopic spread is wider than expected. In the low mass region, fragment ions corresponding to sulfated HexNAc are present at m/z 325, 327, 372,388, and 491. The origin of these ions has been reported previously (Dell et al., 1991). Interestingly, the presence of the signal at m/z 491, which arises from ring cleavage of the penultimate sugar residue with retention of carbons 4-6 on the fragment ion, is indicative of SIHexNAc being linked to position 4 of the next sugar. In the positive FAB spectrum (Fig. 2b), the major peaks at m/z 1411 and 1309 are assigned to SIHexlHexNAczHexNAcitol ((M + 2Na - H)') and (OH)lHexlHexNAczHexNAcitol (loss of sodium sulfite from m/z 1411). In accord with the negative ion data, it appears that HexzHexNAclHexNAcitol is present as a minor component (m/z 1415 and 1313). No peaks due to neutral glycans were detected, suggesting that a complete separation of the sulfated glycans from the neutral ones had been achieved. A- type cleavage gave signals at m/z 343 (Hex+), 374/294 (SIHexNAc'/(OH)lHexNAc+), and 667/587 (S1HexNAc2+/ (OH)lHexNAcz+), thus defining two non-reducing sequences Hex+ and SIHexNAcHexNAc+ for the major component. The @-cleavage ions at m/z 723 and 1016, attributed to sodium adducts of (OH)lHexlHexNAcitol and (OH)lHexlHexNAclHexNAcitol provide further evidence for the attachment of the sulfate group to a terminal HexNAc. The signal at m/z 683 cor responds to elimination of the sulfated HexNAcz moiety to give "dehydrated" HexlHexNAcitol, which is observed as a protonated species. Signals at m / z 671 and 591 are assigned to SIHexHexNAc+ and (0H)HexHexNAc' and are probably derived from the minor sulfated component SIHexZHexNAclHexNAcitol.

Linkage analysis experiments were carried out on the Sep- Pak-purified perdeuteroacetylated oligosaccharides. Aliquots of the 30% and 50% acetonitrile fractions were permethylated using mild Hakomori and sodium hydroxide permethylation conditions, respectively; the products were hydrolyzed, reduced, and acetylated, and the resulting partially methylated alditol acetates were analyzed by GC-MS. In order to define the linkage position of the sulfate group, a second aliquot of the 30% acetonitrile fraction was desulfated with methanol- HCl prior to permethylation and linkage analysis. The desulfated permethylated sample was analyzed by FAB-MS before

O-Glycosylation of Bovine POMC

x5.00

so

913

1 0 0 ? y x 5 . 0 0 m 13,65 r

1369

1009 3 f 3 - x5.00,

FIG. 2. a, negative FAB spectrum of the deuteroacetylated O-linked glycans that eluted from Sep-Pak in the 30% acetonitrile fraction. Signals between m/z 500 and 1150 are magnified five times compared with the molecular ion region, while signals below m/z 500 are magnified three times. The very minor signals between m/z 600 and 700 are a HexNAc increment above those corresponding to sulfated HexNAc (assigned in the text). The signal at m/z 1320 corresponds to (M - H)- of glycans lacking one deuteroacetyl group. The signal at m/z 1051 is derived from ring cleavage of non-reducing Hex residue. The minor cluster at m/z 944 is not assigned. b, positive FAB spectrum of the deuteroacetylated O-linked glycans that eluted from Sep-Pak in the 30% acetonitrile fraction. Signals between m/z 500 and 1100 are magnified five times compared with the rest of the spectrum. Carbohydrate-derived signals are assigned in the text.

conversion to the partially methylated alditol acetates. The FAB data (not shown) confirmed that no significant hydrol- ysis of glycosidic linkages had occurred under the conditions used for the desulfation. Results from linkage analysis of the sulfated fraction are reported in Tables I (before desulfation) and I1 (after desulfation). Data from the 50% fraction are reported in Table IV. The following conclusions can be drawn from the data in Tables I and 11. (i) The presence of 3,6-

linked GalNAcitol, but not 3- or 6-linked GalNAcitol, is indicative of a branched core. (ii) Both 4-linked GlcNAc and 4-linked GalNAc are detected prior to desulfation, but 4- linked GalNAc is absent in the desulfated fraction, suggesting that the sulfate is attached at the 4-position of terminal GalNAc. (iii) Terminal GalNAc is present in the desulfated sample but not in the sulfated sample, thus confirming the conclusion in (ii). (iv) All the detected galactose is terminal.

914 0-Glycosylation of Bovine POMC

Taken together, the GC-MS and FAB-MS data are consistent with the major component comprising a Type 2 core substituted at the 4-position of the 6-linked GlcNAc by a 4-0- sulfated GalNAc residue. However, the data do not rule out an alternative novel core in which the attachment sites of the two branches to the GalNAcitol are reversed. This ambiguity was resolved by periodate oxidation.

Characterization of the Core of the Major Sulfated Glycan by Periodate Oxidation and FAB-MS-The mixture of glycans obtained by reductive elimination of the 0-linked glycopeptide was subjected to mild periodate oxidation, and the products were reduced, deuteroacetylated, and analyzed by negative FAB-MS after Sep-Pak purification. The negative FAB spectrum of the 30% acetonitrile fraction (Fig. 3) gave a strong signal at m/z 772, which is the correct mass for SlHexNAc2 attached to a two-carbon fragment derived from GalNAcitol.

TABLE I GC-MS analysis of partially methylated alditol acetates derived from

sulfated 0-linked glycans (30% fraction)

Assignment Characteristic fragment ions time min

14.76 t-Gal 102,118, 129,145, 161,162,205 18.91 3,6-GalNAc-itol 130, 246, 318

19.55 4-GalNAc 117, 159, 233 19.40 4-GlcNAc 117, 159, 233

TABLE I1 GC-MS analysis of partially methylated alditol acetates derived from

desulfated 0-linked glycans (30% fraction)

Assignment Characteristic fragment ions time

min 14.77 t-Gal 102, 118,129, 145, 161, 162,205 18.91 3,6-GalNAc-it01 130, 246, 318 18.92 t-GalNAc 117, 159,203,205 19.40 4-GlcNAc 117,159,233

This is the result expected when periodate cleaves between carbons 4 and 5 of 3,6-linked GalNAcitol carrying the sulfated branch on position 6, i.e. the common Type 2 core structure of 0-glycans.

Determination of Anomeric Configurations in the Major Sulfated Glycan-Anomeric configurations were determined by a combination of hexosaminidase digestion, chromium trioxide oxidation, and FAB-MS. After deuteroacetylation, the sulfated glycan was separated from non-sulfated components by Sep-Pak purification, and sulfate and deuteroacetyl groups were removed by mild methanolysis. A portion of the product was reacetylated with non-deuterated acetic anhydride and analyzed by FAB-MS to check for completeness of 0-deacetylation and to establish whether any glycosidic cleavages had occurred. The FAB spectrum (not shown) contained signals at m/z 1296, 1299, 1302, and 1305 (approximate ratio 1:2:3:2) consistent with intact HexlHexNAc2HexNAcito1 re- taining zero, one, two, and three deuteroacetyl groups, respectively. We considered that this level of 0-acylation would not seriously prejudice the hexosaminidase experiment, and the remainder of the methanolyzed sample was therefore incu- bated with bovine kidney exo-@-N-acetyl-D-glucosaminidase for 48 h. The products were fully deuteroacetylated and analyzed by FAB-MS (Fig. 4). Molecular ion clusters at m/z 746, 1039, and 1332 correspond to (M + H)' for HexlHexNAcitol, HexlHexNAclHexNAcitol, and He~~HexNAc~HexNAcitol, respectively, showing that the @-hexosaminidase is able to sequentially remove GalNAc and GlcNAc from the GalNAcl- 4GlcNAc branch.

The hexosaminidase experiment was corroborated by chromium trioxide oxidation, which rapidly oxidizes @-linkages while a-linkages remain largely unaffected under the conditions used (Angyal and James, 1970). The degree of oxidation can be readily monitored by FAB-MS because each oxidized sugar residue is shifted in mass by 14 mass units (Khoo and Dell, 1990). The mixture of 0-glycans was desulfated by mild methanolysis, and the products were deuteroacetylated and

FIG. 3. Negative FAB spectrum of the products of mild periodate cleavage. The sample was reduced, deuteroacetylated, and purified by Sep-Pak prior to FAB analysis. The data shown were obtained from the 30% acetonitrile fraction.


XI0 1039 r

1332

FIG. 4. Positive FAB spectrum of the deuteroacetylated products of j3-hexosaminidase digestion of the desulfated glycan. Digestion was not complete because a portion of the sample had retained acyl groups incorporated during purification (see “Results”).

TABLE I11 FAB-MS data on desulfated, perdeuteroacetylated 0-glycans before

and after chromium trioxide oxidation m/z value

Before AM Assignment

After

1336 1378 42 1332

3 @-linkages 1374 42

1039 1067 28 3 @-linkages

746 760 14 2 @-linkages 1 @-linkage

S O ~ - ~ G ~ I N A C ~ ~ - ~ G ~ C N A C ~ ~ ~ 3Ga1NAc

GalpV

nine 45 in bovine 16-kDa glycopeptide.

purified using a Sep-Pak. Fractions were screened by positive FAB-MS, which showed that the HexlHexNAczHexNAcitol (m/z 1332) component was recovered in the 50% acetonitrile fraction, together with some Hex2HexNAclHexNAcito1 (m/z 1336), HexlHexNAcHexNAcitol (m/z 1039), and HexlHex- NAcitol (m/z 746). The FAB data reproduced in Table I11 were obtained after chromium trioxide oxidation of this fraction. These data show that HexlHexNAc~HexNAcitol, Hex2- HexNAclHexNAcitol, HexlHexNAclHexNAcitol, and HexlHexNAcitol have three, three, two, and one @-linkages, respectively, i.e. all linkages in these oligosaccharides are @.

Structure of the Major Sulfated 0-Glycan-Taken together, the above data show that the major sulfated 0-glycan attached to the bovine 16-kDa fragment has the structure shown in Fig. 5. This is an oligosaccharide with a Type 2 core structure, which is elongated on the 6-branch by the sulfated epitope previously observed on N-linked oligosaccharides of pituitary glycoproteins including POMC (Baenziger and Green, 1988; Siciliano et al., 1993).

Analysis of Non-sulfated 0-Glycans-Non-sulfated glycans comprised less than 20% of the 0-glycan population, and

FIG. 5. Structure of the major 0-glycan attached to threo-

complete characterization was difficult because of heteroge- neity and because there were significant differences among the minor oligosaccharides isolated from two preparations of 16-kDa fragment. The two samples of 16-kDa were prepared using similar, but not identical, protocols and were derived from different batches of pituitaries. 0-Glycan variability appears to be confined to the non-sulfated components, since the glycan shown in Fig. 5 was the major component in both preparations.

After deuteroacetylation and Sep-Pak purification of 0- glycans from the first 16-kDa sample, the non-sulfated glycans, eluting in the 50% acetonitrile fraction, gave the positive FAB data reproduced in Fig. 6. The spectrum is dominated by the signal at m/z 1336, which corresponds to (M + H)+ for Hex2HexNAclHexNAcito1. Less abundant molecular ions are present at m/z 746 (HexlHexNAcitol) and 1039 (HexlHexNAclHexNAcitol) and major A-type fragment ions at m/z 343 and 636 are assigned to Hex+ and HexHexNAc’, respectively. The chloride ion adducts of each of the above species were observed in the negative ion mode and the absence of any (M - H)- signals attributable to sulfated components indicated that the Sep-Pak procedure had effi- ciently separated the sulfated glycans from their non-sulfated counterparts (data not shown). In order to search for very minor components, which are more readily observed as their permethyl derivatives, the remainder of the perdeuteroacetylated sample was permethylated and 10% of the products were analyzed in the positive ion mode. The major signals in the FAB spectra were fully consistent with the results from the deuteroacetyl derivative but there were additional very minor signals corresponding to (M + H)+ for NeuAcHexlHex- NAcitol, NeuGcHexlHexNAcitol, NeuAcHex2HexNAclHex- NAcitol, and NeuGcHexzHexNAclHexNAcitol (data not shown). Linkage analysis data acquired on the remainder of the permethylated sample are reported in Table IV. Terminal Gal, 4-linked GlcNAc, 3-linked GalNAcitol, and 3,6-linked GalNAcitol were observed as prominent constituents, to-


I

40

FIG. 6. Positive FAB spectrum of the deuteroacetylated 0-linked glycans that eluted from Sep-Pak in the 50% acetonitrile fraction. The (M + H)+ signals of fully deuteroacetylated components are assigned in the text. Signals are also present for (M + Na)+ species 22 mass units above each (M + H)+ and for components lacking one or more deuteroacetyl groups that occur at 45 mass unit intervals below fully derivatized species.

TABLE IV GC-MS analysis of partially methylated alditol acetates derived from

neutral 0-linked glycans (50% fraction) Major components are indicated by an asterisk.

Assignment Characteristic fragment ions time min

12.90 t-Fuc 118, 131, 162,175 14.75* t-Gal 102, 118, 129, 145, 161, 162, 205 15.96 3-Gal 118,129,161,234 16.50 6-Gal 118,129,162, 189,233 16.71 3,4-Gal 118,129 16.83 2,3-Gal 129,161,262 17.02' 3-GalNAc-it01 130, 246, 290 17.28 2,6-Gal 129,130,189,190 18.83* 3,6-GalNAc-it01 130, 246, 318

18.92 t-GalNAc 117,159,203,205 18.43 t-GlcNAc 117, 159, 203, 205

19.35; 4-GlcNAc 117,159,233 20.23 3,4-GlcNAc 117,159,346 20.68 4,g-GlcNAc 117,159, 261

gether with a considerable number of minor components. FAB-MS monitoring of exoglycosidase digests of the total 0- glycan population revealed that the major neutral oligosaccharides were fully degraded by sequential digestion with exo- @-galactosidase and exo-@-N-acetyl-D-glucosaminidase (data not shown). Taken together, the FAB-MS, GC-MS, and enzyme experiments suggest that two of the major neutral glycans in this preparation are Galj31-3GalNAcitol and Gal@l-4GlcNAc@1-6(Gal@l-3)GalNAcitol.

The non-sulfated deuteroacetylated 0-glycans from the second 16-kDa preparation gave data (Fig. 7a) that were signif- icantly different from those described above (Fig. 6). Although HexlHexNAcitol (m/z 746), HexlHexNAclHexNAcitol ( n / z 1039), and HexzHexNAclHexNAcitol (m/z 1336) are still present, the last is no longer the major component. Instead, the most abundant high mass molecular ion corresponds to a

fucosylated glycan of composition FuclHexlHexNAc2Hex- NAcitol (m/z 1568), while a less intense, but nevertheless quite prominent, signal is observed at m/z 1332 (HexlHex- NAc2HexNAcitol) corresponding to its non-fucosylated counterpart. The A-type fragment ion at m/z 876 is indicative of a HexNAc(Fuc)HexNAc' branch in the fucosylated glycan. The mixture of deuteroacetylated glycans was purified by HPLC, and fractions were monitored by off-line FAB-MS. Linkage analysis was carried out on the major fucosylated component, which eluted in fraction 27. Partially methylated alditol acetates were observed for terminal Fuc, terminal Gal, terminal GalNAc, 3,4-linked GlcNAc, and 3,6-linked Gal- NAcitol together with additional minor components (Table V). These data suggest that the fucosylated glycan has the same structure as the major sulfated glycan (Fig. 5), except that the sulfate is absent and the GlcNAc is substituted at position 3 by fucose. However, we cannot rule out the alternative structure in which the positions of GalNAc and Fuc are exchanged. We attempted to resolve this ambiguity by permethylating the mixture of 0-glycans and searching for linkage specific fragment ions in the FAB spectra. Unfortu- nately, fragment ions from the fucosylated component were insufficiently abundant for a firm conclusion to be made. In accord with the data obtained from the first 16-kDa sample, permethylation enabled the facile identification of sialylated components (Fig. 7b), which gave molecular ions at n / z 873 (NeuAclHexlHexNAcitol), 903 (NeuGclHexlHexNAcitol), 1234 (NeuAc~HexlHexNAcitol), 1265 (NeuAclNeuGclHexl- HexNAcitol), 1295 (NeuGc2HexlHexNAcito1), 1323 (NeuAcl- Hex2HexNAclHexNAcito1), and 1353 (NeuGclHex2Hex- NAclHexNAcitol). The disialylated oligosaccharides were not observed in the first sample, but this may not be a significant observation because higher quality data were obtained from the second batch.

DISCUSSION

Using a FAB-MS/derivatization strategy, complemented by chemical and enzymatic degradation, we have shown that

O-Glycosylation of Bovine POMC

’r I

917

1039

1336 1332 I

1 5 6 8

FIG. 7. a, positive FAB spectrum of the deuteroacetylated O-linked glycans from the second preparation of 16-kDa fragment (see “Experimental Procedures”), which eluted from Sep-Pak in the 50% acetonitrile fraction. Signals are assigned in the text. b, positive FAB spectrum of the permethylated O-linked glycans from the second preparation of 16-kDa fragment.

the major O-glycan attached to threonine 45 in bovine 16- kDa fragment is a sulfated tetrasaccharide (Fig. 5). This structure is novel because of the sulfated P-linked GalNAc residue attached to the 6-arm of the Type 2 core giving the SO,-4GalNAc~l-4GlcNAc~l- moiety, which is normally found in the N-glycans of pituitary glycohormones (Baenziger and Green, 1988). Although recent studies have suggested that this epitope is more widespread in N-glycans than previously thought (Smith et al., 1992; Dell and Khoo, 1993), there is no published evidence for its presence in O-glycans.

The GalNAc transferase responsible for the addition of p- linked GalNAc to the N-glycans in the pituitary glycohormones recognizes a tripeptide motif, Pro-X-Arg/Lys (where X is usually a hydrophobic residue), located 6-9 residues NH2- terminal of the glycosylation site (Smith and Baenziger, 1992). In the 16-kDa fragment the proteolytic processing site, Pro-Arg-Lys, which is 15 residues upstream from the N - glycosylation site (Fig. l ) , is believed to be the recognition motif for GalNAc addition to the N-glycans (Skelton et al., 1992; Siciliano et al., 1993). Thus, some flexibility in both the

918 0-Glycosylation of Bovine POMC TABLE V

GC-MS linkage amlyses of HPLC fraction 27, which gave a major molecular ion corresponding to FuclHexlHexNAca in the FAB

spectrum Major components are indicated by an asterisk.

E'ution Assignment Characteristic fragment ions time

min 12.87* t-Fuc 118, 131, 162, 175 14.73* t-Gal 102, 118, 129, 145, 161, 162, 205 15.96 3-Gal 118, 129, 161,234

18.82* 3,6-GalNAc-itol 130, 246, 318 18.92* t-GalNAc 117,159,203,205

18.43 t-GlcNAc 117, 159, 203,205

19.35 4-GlcNAc 117,159,233 20.23, 3,4-GlcNAc 117,159,346

location of the tripeptide motif and the nature of X can be accommodated by the P-GalNAc transferase, although its activity is probably compromised, as exemplified by the fact that none of the N-glycans in the 16-kDa fragment contains more than one sulfated P-GalNAc epitope (Siciliano et al., 1993), in contrast to lutropin and thyrotropin, which have major glycoforms containing two sulfated B-GalNAc antennae. We consider it likely that the Pro-Arg-Lys motif regu- lates @-GalNAc addition at both the N- and 0-glycosylation sites in bovine 16-kDa fragment, with a single P-GalNAc transferase being responsible for synthesis a t both sites. Prec- edent exists for a capping p-GalNAc transferase being capable of accepting both N- and 0-glycans as substrates. Conzelmann and Kornfeld (1984b) have isolated a p1,4-GalNAc transferase from a cytotoxic T lymphocyte cell line, which is similar, if not identical, to the enzyme that is responsible for synthesis of the Sd" and Cad blood group determinants, both of which contain P-linked GalNAc as a capping group attached to a p- Gal residue (Blanchard et al., 1983, 1985). Using glycophorin and Tamm-Horsfall glycoprotein as 0- and N-glycan acceptors, respectively, Conzelmann and Kornfeld (1984b) showed that the purified enzyme was able to transfer GalNAc to Gal acceptors on both 0- and N-glycans. A similar broad specificity might be exhibited by the pituitary p-GalNAc transferase, provided the putative 0-glycan acceptor is appropriately located with respect to the tripeptide recognition site. In the 16-kDa fragment, the 0-glycosylation site is very close to the tripeptide motif that is believed to be recognized by the GalNAc transferase (Fig. 1).

A number of other protein hormones have been shown to be 0-glycosylated. These include the a-subunit of lutropin (Parsons and Pierce, 1984), the &subunit of human chorionic gonadotropin (Kessler et al., 1979), placental lactogen (Shi- momura and Bremel, 1988), human erythropoietin (Lai et al., 1986; Sasaki et al., 1987), catfish somatostatin (Andrews et al., 1984), rat pro-glucagon (Patzelt and Weber, 1986) and chromogranin A (Rosa et at., 1985; Gorr et al., 1991). With the exception of chromogranin A, the 0-glycan structures linked to these proteins do not appear to be sulfated. Chro- mogranin A is a member of the chromogranin/secretogranin family of acidic proteins found in many endocrine and neu- roendocrine cells (Winkler and Fischer-Colbrie, 1992). They are segregated into the granule fraction with processed hormone and co-secreted upon stimulation. These ubiquitous proteins are thought to promote granule formation by aggre- gating within the trans Golgi network (Scammell, 1993). It will be interesting to determine whether the novel sulfated 0- glycan found in bovine POMC is prototypic for sulfated 0- linked oligosaccharides in chromogranin, the structures of

which await elucidation, or whether it is uniquely found in the 16-kDa fragment.

Terminal P-linked GalNAc residues are relatively rare in 0-glycans, although they are increasingly being discovered in a wide variety of N-glycans (reviewed by Dell and Khoo (1993)). Recently characterized N-linked glycoproteins containing this moiety include Bowes melanoma tissue plasmin- ogen activator (Chan et al., 1991), human Tamm-Horsfall glycoprotein ( H k d et al., 1992), bovine lactotransferrin (Coddeville et al., 1992), human urokinase (Bergwerff et al., 19921, and proteases from snake venom (Pfeiffer et al., 1992; Tanaka et al., 1992). In the 0-glycan family, GalNAcpl-4Gal is a constituent of the very rare blood group determinant called Cad (discussed earlier) and has been observed on non- reducing termini of oligosaccharides present in a cloned mu- rine cytotoxic T lymphocyte cell line (Conzelmann and Korn- feld, 1984a). /3-Linked GalNAc has also been found in cervical glycoproteins from the bonnet monkey (Nasir-ud-Din et al., 1990), in fish egg glycoproteins (Shimamura et al., 1983; Iwasaki et al., 1984), and as a minor constituent of porcine zona pellucida glycoproteins (Hirano et al., 1993). In none of these instances is it known to be sulfated. Sulfate is, nevertheless, a relatively common substituent on 0-glycans, especially the heavily glycosylated mucins, where it usually occurs attached in various linkages to Gal and GlcNAc (see, for example, Capon et al. (1989) and Mawhinney et al. (1992)). The biological significance of sulfation is poorly understood. Sulfated epitopes might be important for anti-adhesion by masking binding epitopes and providing a negatively charged surface on the glycoprotein. On the other hand, they appear to constitute an important part of the binding epitope in the L-selectin ligand (Lasky et al., 1992) and are an effective alternative to sialic acid in the E-selectin epitope (Yuen et al., 1992). The sulfated antennae of the pituitary glycohormone N-glycans are recognized by a specific hepatic reticuloendo- thelial cell receptor, which is believed to be responsible for their rapid clearance from the circulation (Fiete et al., 1991). However, studies of the metabolic clearance rate and half- time disappearance rate of human 16-kDa fragment have suggested that the hepatic receptor is unlikely to play a critical role in determining the metabolic fate of this glycopeptide (Lu et al., 1983; Siciliano et al., 1993). Carbohydrate composition analysis of the 16-kDa fragment and its biosynthetic derivatives isolated from pituitary tissue has indicated that processing at the -Arg49-Lys50- site is dependent upon the absence of 0-glycosylation at threonine 45 (Browne et al., 1981; Bennett, 1984; James and Bennett, 1985; Seger and Bennett, 1986). This is consistent with the known predilection of 0-glycans to inhibit the action of proteases (Jentoft, 1990). Whether the novel sulfated structure identified in the present study is critical for masking the cleavage site within the 16- kDa fragment has yet to be determined. The non-sulfated structures that are present as minor constituents might be equally capable of inhibiting processing. Alternatively, non- sulfated glycoforms could be processed to yield glycosylated forms of the amino-terminal residue 1-49 fragment that have remained undetected because of their low abundance.

The present study shows that less than 20% of the 0- glycans in the 16-kDa fragment are non-sulfated and that they constitute a highly heterogeneous population of neutral and sialylated glycans, the latter containing both N-acetyl and N-glycolyl neuraminic acid. In two separate preparations of bovine pituitaries, considerable differences were detected in the neutral 0-glycans despite minimal differences being observed in the acidic glycans. Although the neutral 0-glycan variability could be due to differences in the purification

0-Glycosylation of Bovine POMC 919

protocols for the two preparations (see “Experimental Pro- cedures”), which could conceivably have separated some glycoforms into fractions that were not characterized, we believe this to be unlikely because the same N-glycans (neutral as well as acidic) were observed in both preparations (data not shown). Furthermore, the sizes of the neutral 0-glycans observed in each preparation are similar and they are expected to have comparable chromatographic behavior. In future work we intend to explore the possibility that the structures of the non-sulfated 0-glycans are affected by the metabolic status of the animals from which the pituitaries are obtained.

The expression of the various pressor, steroidogenic, and mitogenic actions of derivatives of the 16-kDa fragment is dependent upon correct processing of POMC at the -Arg4’- Lys6’- cleavage site, which in turn is dependent upon the 0- glycosylation status. Several hormonal functions have been proposed for the 16-kDa fragment and the products resulting from the action of pro-hormone convertases (Seger and Ben- nett, 1986; Estivariz et al., 1989). The amino-terminal fragment encompassing the first 49 residues has been shown to be a mitogen for adrenal cortical cells (Estivariz et al., 1982; Lowry et al., 1983). The carboxyl-terminal fragment comprising the y M S H sequence has been shown to potentiate the steroidogenic actions of ACTH on adrenal cells both in vivo and in uitro (Pederson and Brownie, 1980; Al-Dujaili et al., 1981). The secretion of both glucocorticoids and mineralocor- ticoids by the adrenal cortex is enhanced by peptides bearing the y M S H sequence (Pederson and Brownie, 1980; Pederson et al., 1980; Al-Dujaili et al., 1981; Seidah et al., 1981). The latter class of steroid is critical in maintenance of electrolyte homeostasis and blood pressure. In keeping with these observations, yMSHs have also been shown to have pressor and cardioaccelerator activity following intravenous administra- tion to rats (Callahan et al., 1984; Klein et al., 1985; Sun et al., 1992; DeWildt et al., 1993). These hemodynamic effects have been shown to be mediated primarily through the sym- pathetic nervous system (Callahan et al., 1984; DeWildt et al., 1993). Some of the effects of 7-MSH related peptides may also be mediated centrally through the autonomic nervous system (Gruber and Callahan, 1989). POMC is synthesized in the brain primarily in the arcuate nucleus of the hypothalamus. The biosynthetic processing of POMC in the rat hypothalamus is reminiscent of that observed in the intermediate lobe of the pituitary including extensive cleavage of the -Arg4’-LysS0- processing site within the 16-kDa fragment (Emeson and Eipper, 1986). The 0-glycosylation status of the 16-kDa fragment is likely to be important for processing in the hypothalamus as well as the pituitary. It will therefore be interesting to establish whether the novel sulfated 0-glycan is also present in POMC biosynthesized in the hypothalamus.

The structural studies reported in this paper provide a firm foundation for future research directed toward establishing the role of 0-glycosylation in the control of POMC processing and for addressing other structure/function issues. For example, it will be interesting to examine whether truncated 0-glycans, prepared by desulfation and controlled exoglycosidase digestion of the 16-kDa fragment, are cleaved at the -Arg4’-LysSo- processing site. Whether or not the previously identified pituitary GalNAc transferase (Smith and Baen- ziger, 1988) is responsible for addition of the capping GalNAc in the 16-kDa 0-glycan can conceivably be resolved by as- sessing whether appropriately truncated glycoforms are substrates for this enzyme.

Acknowledgment-H. P. J. B. is grateful for the skilled technical assistance of Susan James.

REFERENCES Albersheim, P., Nevins, D. J., English, P. D., and Karr, A. (1967) Carbohydr.

Al-Dujaili, E. A. S., Hope, J., Estivariz, F. E., Lowry, P. J., and Edwards, C. R.

Andrews, P. C., Pubols, M. H., Hermodson, M. A,, Sheares, B. T., and Dixon,

Angyal, S. J., and James, K. (1970) A u t . J. Chem. 23,1209-1221 Azaryan, A. V., Wong, M., Friedman, T. C., Cawley, N. X., Estivariz, F. E.,

Cben, H.-C., and Loh, Y. P. (1993) J. Biol. Chem. 268,11968-11975 Baenziger, J. U., and Green, E. D. (1988) Biochim. Biophys. Acta 947,287-306 Bennett, H. P. J. (1984) Biochem. Biophys. Res. Commun. 125,229-236 Bennett, H. P. J. (1986a) J. Chromatogr. 359,383-390 Bennett, H. P. J. (1986b) Peptides 7,615-622 Bennett, H. P. J., Seidah, N. G., Benjannet, S., Solomon, S., and Chrhtien, M.

Bergwerff, A. A,, Thomas-Oates, J. E., Van Oostrum, J., Kamerling, J. P., and

Birch, N. P., Estivariz, F. E., Bennett, H. P. J., and Loh, Y. P. (1991) FEBS

Blanchard, D., Cartron, J. P., Fournet, B., Montreuil, J., Van Halbeek, H., and

Blanchard, D., Capon, C., Leroy, Y. , Cartron, J. P., and Fournet, B. (1985)

Blanchard, D., Piller, F., Gillard, B., Marcus, D., and Cartron, J. P. (1985) J.

Res. 5,340-345

H. (1981) Nature 291,156-159

J. E. (1984) J. Biol. Chem. 2 5 9 , 13267-13272

(1986) Int. J. Pept. Protein Res. 27,306-313

Vliegenthart, J. F. G. (1992) FEBS Lett. 314,389-394

Lett. 290,191-194

Vliegenthart, J. F. G. (1983) J. Biol. Chem. 2 5 8 , 7691-7695

Biochern. J. 232,813-818

Biol. Chem. 260. 7813-7816 ~ ~ ~~~ ~~~

Browne, C. A,, Bennett, H. P. J., and Solomon, S. (1981) Biochem. Biophys. Res. Commun. 100,336-343

Callahan, M. F., Kirby, R. F., Lymangrover, J. R., Johnson, A. K., and Gruber, K. A. (1984) Clin. Enp. Hypertern. A6, 1727

Capon, C., Leroy, Y., Wieruszeski, J.-M., Ricart, G., Strecker, G., Montreuil, J., and Fournet, B. (1989) Eur. J. Biochem. 182,139-152

Chan, A. L., Morris, H. R., Panico, M., Etienne, A. T., Rogers, M. E., Gaffney, P., Creighton-Kempsford, L., and Dell, A. (1991) Glycobiology 2 , 173-185

Coddeville, B., Strecker, G., Wieruszeski, J. M., Vliegenthart, J. F. G., Van

Res. 236,145-164 Halbeek, H., Peter-Katalinic, J., Egge, H., and Spik, G. (1992) Carbohydr.

. ~~~ ~~~

Conzelmann, A,, and Kornfeld, S. (1984a) J. Biol. Chem. 259,12536-12542 Conzelmann, A., and Kornfeld, S. (1984b) J. Biol. Chem. 2 5 9 , 12528-12535 Dell, A. (1990) Methods Enzymol. 193,647-660 Dell, A,, and Khoo, K.-H. (1993) Curr. Opinion Struct. Biol. 3,687-693 Dell, A., Morris, H. R., Greer, F., Redfern, J. M., Rogers, M. E., Weisshaar, G.,

DeWildt, D. J., Krugers, H., Kashergen, C. M., DeLong, H., and Versteeg, D.

Eipper, B. A., and Mains, R. E. (1978) J. Supramol. Struct. 8,247-262 Eipper, B. A,, and Mains, R. E. (1980) Endocr. Reu. 1 , 1-27 Emeson, R., and Eipper, B. A. (1986) J. Neurosci. 6,837-849 Estivariz, F. E., Iturriza, F., McLean, C., Hope, J., and Lowry, P. J. (1982)

Estivariz, F. E., Birch, N. P., and Loh, Y. P. (1989) J. Biol. Chem. 2 6 4 , 17796-

Fiete, D., Srivastava, V., Hindsgaul, O., and Baenziger, J. U. (1991) Cell 6 7 ,

Gorr, S.-U., Hamilton, J. W., and Cohn, D. V. (1991) J. Bid. Chem. 266,5780-

Gruber, K. A., and Callahan, M. F. (1989) Am. J. Physiol. 2 5 7 , R681 H k d , K., Van Zadelhoff, G., Moonen, P., Kamerling, J. P., and Vliegenthart,

Hiyama, J., and Renwick, A. G. C. (1991) Carbohydr. Res. 209,33-50

H. G. (1993) Eur. J. Biochem. 233,157-164

Nature 297,419-422

1780 1

1103-1110

5784

J. F. G. (1992) Eur. J. Biochem. 209.895-915 Hirano, T.; Takasaki, S., Hedrick, J. L.,’Wardrip, N. J., Amano, J., and Kobata,

Iwasakl, M., Inoue, S., Kitajima, K., Nomoto, H., and Inoue, Y. (1984) Biochem-

James, S., and Bennett, H. P. J. (1985) J. Chromatogr. 326,329-338 Jentoft. N. (1990) Trends Biochem. Sci. 15.291-294

A. (1993) Eur. J. Biochem. 2 1 4 , 763-769

istry 23,305-310

Kawauchi, H . (1983) Arch. Biochem. Biophys. 227,343-350 Kessler, M. J., Mise, T., Ghai, R. D., and Bahl, 0. P. (1979) J. Biol. Chem.

Keutmann, H. T., Lampman, G. W., Mains, R. E., and Eipper, B. A. (1981)

Khoo, K.-H., and Dell, A. (1990) Glycobiology 1,83-91 Khoo, K.-H., Morris, H. R., McDowell, R. A., Dell, A., Maccarana, M., and

Lindabl, U. (1993) Carbohydr. Res. 244,205-223 Klein, M. C., Hutchins, P. M., Lymangrover, J. R., and Gruber, K. A. (1985)

Lije Sci. 3 6 , 769-775 Lai, P.-H., Everett, R., Wang, F.-F., Arakawa, T., and Goldwasser, E. (1986) J.

Biol. Chem. 261,3116-3121 Lasky, L. A., Singer, M. S., Dowbenko, D., Imai, Y., Henzel, W. J., Grimley,

C., Fennie, C., Gillett, N., Watson, S. R., and Rosen, S. D. (1992) Cell 6 9 ,

Lowry, P. J., Silas, L., McLean, C., Linton, E. A., and Estivariz, F. E. (1983) 927-938

Lu, C.-L., Chan, J. S. D., de Lean, A., Cben, A., Seidah, N. G., and Chretien, Nature 306.70-73

M. (1983) Lije Sci. 3 3 , 2599-2608 Mains, R. E., and Eipper, B. A. (1976) J. Bid. Chem. 251,4115-4120 Mawhinney, T. P., Adelstein, E., Gayer, D. A., Landrum, D. C., and Barbero,

Nasir-ud-Din, Hussain, S. A., Jeanloz, R. W., and Walker-Nasir, E. (1990)

Noel, G., Keutmann, H. T., and Mains, R. E. (1991) Mol. Endocrinol. 5 , 404-

2 5 4 , 7909-7914

Biochemistry 20,4148-4155

G. J. (1992) Carbohydr. Res. 2 2 3 , 187-207

Carbohydr. Res. 205,444-452

dl ?I

Pa;;&, T. F., and Pierce, J. G. (1984) J. Biol. Chem. 259,2662-2666 Patzelt, C., and Weber, B. (1986) EMBO J. 5,2103-2108 Pederson, R. C., and Brownie, A. C. (1980) Proc. Natl. Acad. Sci. U. S. A. 7 7 ,

2239-2243 Pederson, R . C., Ling, N., and Brownie, A. C. (1980) Science 208 , 1044-1046 Pfeiffer, G., Dabrowski, U., Dabrowski, J., Stirm, S., Strube, K.-H., and Geyer,

Rosa, P., Hille, A., Lee, R. W. H., Zanini, A,, DeCamilli, P., and Huttner, W. R. (1992) Eur. J. Biochem. 205,961-978

920 O-Glycosylation of Bovine POMC

Sasaki, H., Bothner, B., Dell, A., and Fukuda, M. (1987) J. Bwl. Chem. 282,

Seger, M. A., and Bennett, H. P. J. (1986) J. Steroid Biochem. 25, 703-710 Scammell, J. G. (1993) Trends EndocrinoL Metab. 4, 14-18

Seidah, N. G., Rochemont, J., Hamelin, J., Lis, M., and Ch&tien, M. (1981) J.

Shimamurn, M., Endo, T., Inoue, Y., and Inoue, S. (1983) Biochemistry 22,

Shimomura, K., and Bremel, R. D. (1988) Mol. Endocrinol. 2,845-853 Siciliano, R. A., Morris, H. R., McDowell, R. A,, Azadi, P., Rogers, M. E.,

Skelton, T. P., Kumar, S., Smith, P. L., Beranek, M. C., and Baenziger, J. U.

B. (1885) J. CeU Biol. 101,1977-2011

12059-12076

BwL Chem. 256, 1977-7984

959-963

Bennett, H. P. J., and Dell, A. (1993) Glycobiol~gy 3,225-239

(1992) J. Biol. Chem. 287,12998-13006 Smith, P. L., and Baenziger, J. U. (1988) Science 242,930-933

Smith, p. L., and Baenziger, J. U. (1992) Proc. Natl. Acad. Sci. U. s. A. 89, 329-333

Smith p. L., Skelton T. P., Fiete D., Dharmesh S. M. Beranek M. c M c h i l L Broze.'G. J., Jr., a i d Baenziger, J.'U. (1992) J, Bioi C h i : 287,19i40:19146

Sun, x. y., Feng, P. Gong, Q. L., Edvinsson, L., and Hedner, T. (1992) Am. J. Physwl. 262%7?

Tanaka, N., Nakada, H., Itoh, N., Mizuno Y. Takanishi M. Kawasaki T Tab , S. I., Inagaki, F., and Yamashina, i. (1b92) J. Bidhei. (Tokyo) ili: 68-74

~" ."

Winkler, H., and Fisher-Colbrie, R. (1992) Neuroscience 49,497-528 Yuen, C. T., Lawson, A. M., Chai, W., Larkin, M Stoll, M. S., Stuart A. C

9131 Sullivan, F. X., Ahern, T. J., and Feizi, T. (1942) Biochemistry 31,'912<

Zhou, A., Bloomquist, B. T., and Mains, R. E. (1993) J. Bwl. Chem. 268,1763- 1769

the journal of vol. 2, 14, pp. by for in u. s. a. o-glycosylation … · 2001-06-25 ·...

Documents