| Glycobiology | Pages |
Partially glucose-capped oligosaccharides are found on the hemoglobins of the deep-sea tube worm Riftia pachyptila
Introduction
Results
Discussion
Materials and methods
Acknowledgments
Abbreviations
References
Partially glucose-capped oligosaccharides are found on the hemoglobins of the deep-sea tube worm Riftia pachyptila
Introduction
The giant tube worm Riftia pachyptila is a vestimentiferan which lives around deep-sea hydrothermal vents along the East Pacific Rise. Having no mouth or gut, it relies entirely on endosymbiotic chemolithoautotrophic bacteria for metabolism and growth, and must, therefore, provide these symbionts with the inorganic metabolites they need: oxygen, sulfide, and carbon dioxide. The former two are concentrated and transported from the external environment to the bacteria-containing organ by hemoglobins (reviewed in Childress and Fisher, 1992).
Three extracellular hemoglobins (EHb) are found in Riftia body fluids, and they have recently been wholly identified and characterized (Zal et al. 1996a,b). Two of these EHbs are dissolved in the vascular blood (V1 and V2) and one in the coelomic fluid (C1). The EHb V1, with a molecular weight of 3503 ± 13 kDa (Zal et al., 1996b), consists of an arrangement of both globin and linker chains (Zal et al., 1996a) and a hexagonal bilayer architecture (Terwilliger et al., 1980; De Haas et al., 1996; Zal et al., 1996b) which is typical of EHbs and chlorocruorins found in annelids (Lamy et al., 1996) and vestimentifera (Suzuki et al., 1988). The other two EHbs have lower molecular masses (433 ± 8 kDa for V2 and 380 ± 4 kDa for C1), lack linker chains, and are only found in other vestimentifera and related pogonophora (Suzuki et al., 1988; Terwilliger, 1992; Yuasa et al., 1996; Zal et al., 1996a,b).
Electrospray mass spectrometric analysis showed that, in contrast with EHb V1, EHbs V2 and C1 contained glycosylated monomeric globin chains, a1-a3,which account for ~10% of the total hemoglobin content(Zal et al., 1996a). The difference in mass between the enzymatically deglycosylated chain a and the chains a1-3 was, within experimental error, equal to the mass of glycans with the composition (HexNAc)2(Hex)8-10. The presence of two other putative glycoforms, a4 and a5, could not be established conclusively because the differences between the calculated and measured masses for these species were significantly larger than the estimated experimental error. From ESI-MS data alone, it was not possible to determine the exact nature of these glycans since all hexose residues have identical masses (i.e., 162.1 Da). Glycosylated chains in annelid EHb are not common although several other cases have been reported (e.g., Lumbricus terrestris EHb globin chain a and linker chain L1; Martin et al., 1996). Some authors have even postulated a role for carbohydrates in the supramolecular structure of annelid EHbs as for Perinereis aibuhitensis (Ebina et al., 1995).
The aim of the present work was to determine the exact nature of the glycan residues found in Riftia EHbs to refine further the model of the quaternary structure of Riftia EHbs developed in earlier studies (Zal et al., 1996a). In this report we have characterized the N-linked oligosaccharides of Riftia EHbs V2 and C1 by a combination of mass spectrometry (MALDI and ESI-MS/MS), NP-HPLC, and exoglycosidase digestion.
Results
Electrospray tandem mass spectral data
The electrospray fragment ion spectra of the [M + 11H]11+ ions from the two major glycoforms from Riftia EHbs (Figure
Figure 1. Electrospray fragmentation spectra of the [M + 11H]11+ ions from the (HexNAc)2(Hex)8 (upper spectrum) and (HexNAc)2(Hex)9 (lower spectrum) glycoforms of EHb V2 chain a. The spectra were obtained with methane at 3 × 10-2 mbar as the collision gas and collision cell potential of 45 V. Ionic compositions are indicated using the symbols H for hexose and N for N-acetylhexosamine. The relative abundance of the fragments from the (HexNAc)2(Hex)9 reflects the branching pattern (Oliver et al. 1996) which, again, indicates that these two carbohydrates belong to the normal high-mannose series. Thus, major fragments at m/z 1748.6, 1732.3, and 1700.0 reflect losses at the branch points of two (the hexoses from either branch of the 6-arm), three (the hexoses of the 3-arm) and five (the entire 6-arm) hexoses, respectively. Evidence for this antenna specificity was obtained from MALDI-PSD fragmentation and is described below. Major losses of two, three, and four hexoses from the glycoprotein containing the (HexNAc)2(Hex)8 structure indicates that the compound has one hexose less on the 6-arm. The same pattern is evident in the ions of the 11+ charge state. Further structural analysis was performed on the released sugars by use of MALDI-MS, NP-HPLC and exoglycosidase digestion.
Oligosaccharide profiling
Oligosaccharides linked to the side chain of asparagine were first released from the two EHb samples by PNGase F and then subjected to MALDI MS analysis. Figure
Table I.
Figure 2. (A) MALDI mass spectrum of oligosaccharides released from EHb V2 by PNGase F treatment. Molecular weights correspond to the monoisotopic masses of the [M+Na]+ ions. Satellite peaks corresponding to [M+K]+ ions are observed 16 mass units above the [M+Na]+ ions. Signals below m/z 900 correspond to matrix-related ions. (B) MALDI mass spectrum of oligosaccharides from EHb V2 after treatment with jack bean [alpha]-mannosidase. Molecular weights correspond to the average masses of the [M+Na]+ ions. The MALDI-MS profile of carbohydrates released from the coelomic EHb C1, was very similar to the one obtained for V2 (data not shown). However, the small amounts of (HexNAc)2(Hex)4-6 were not detected in this sample.
Exoglycosidase digests
Although the monosaccharide composition of an oligosaccharide can be deduced from the masses measured by mass spectrometry, no information can be obtained as to which of the isobaric hexoses (mannose, glucose, galactose) or N-acetylhexosamines (N-acetylglucosamine (GlcNAc), N-acetylgalactosamine (GalNAc)) are present. Because of the low relative abundance of the glycosylated globin chains (~10%) within the hemoglobin complex, classical composition analysis could not be performed. However, some features of the sugar chains can be deduced from the fact that PNGase F cleaves N-linked oligosaccharides only and, thus, requires the presence of the di-N-acetyl chitobiose core for catalysis to occur (Chu, 1986). Although there is no formal proof for the nature of the HexNAc residues deduced from the mass measurement of released glycans in this study, it would seem unlikely that HexNAc isomers other than GlcNAc would be present. Therefore, all sugars in the present study share the same pentasaccharide core, which is almost certainly (GlcNAc)2(Man)3. For a more detailed account of the structures present, exoglycosidase digestions were performed since these enzymes are specific for the type of monosaccharide as well as the linkage and stereochemistry of the glycosidic bond hydrolyzed.
For V2, Figure
To determine the nature of the additional hexose(s), aliquots of the oligosaccharides from both EHbs were fluorescently labeled with 2-AB and analyzed by normal-phase HPLC. The HPLC profiles obtained (Figure
Figure 3A. Partial normal-phase HPLC chromatograms of 2-AB labeled oligosaccharides from EHb V2. Peaks are annotated according to their elution position relative to a 2-AB labeled partial dextran hydrolysate (glucose units, GU) and monosaccharide compositions determined by MALDI. (A) glycan pool (B) isolated H9N2 (the shaded area corresponds to the material actually collected) after treatment with rat liver glucosidase II (C) glycan pool (D) isolated H10N2 after treatment with rat liver glucosidase II. Vascular EHb V2, Figure Further evidence for the branching pattern of these high mannose glycans was obtained from the MALDI PSD fragmentation pattern (Figure Confirmation of this structure was provided by the presence of two cross-ring cleavage ions produced from the branching mannose. In the spectrum of (HexNAc)2(Man)9 (Figure The PSD spectra of the other glycans displayed similar ions (Table II). Thus, compound 5 (HexNAc)2(Man)8 displayed Y-type losses of (Hex)1 (Hex)2, (Hex)3, and (Hex)4 but not (Hex)5, consistent with the structure of compound 5 (Table I) where the missing mannose was one attached to the 6-antenna (the specific mannose could not be identified). The weak internal cleavage ion at m/z 995.8 [(Hex)6] and the strong one corresponding to the (Hex)5 from the 6-antenna at m/z 833.7, together with the 0,4A4, and 3,5A4 cross-ring cleavage ions, confirmed this structure. The PSD spectrum of (HexNAc)2(Man)10 from EHb V2 was weak and consisted of a mixture of compounds 1 and 2. However, the comparable relative abundances of the ions at m/z 995.8 and 833.7 (Table II) is consistent with the presence of compounds 1 and 2 (Table I). Table II. Figure 4A. (A)MALDI-PSD spectrum of (HexNAc)2(Man)9 from EHb C1. Ions are identified in (C), and masses are listed in Table II, (B) MALDI-PSD spectrum of (HexNAc)2(Man)8 from EHb CI. Ions are listed in Table II. (C) Formation of the fragment ions in the MALDI-PSD mass spectrum of (HexNAc)2(Man)9 shown in (A). Ions labeled with a prime ([prime]) or double prime ([prime][prime]) are formed by cleavage of one of the branches of the 6-antenna. However, the spectra do not allow these branches to be differentiated and they are labeled as in (A) for illustrative purposes by analogy with the structures of similar compounds that have been fully characterized in other species. The linkage between the HexNAc residues is similarly shown as between GlcNAc. Structural symbols are as in Table I.
Fig. 4B. See caption under Figure 4A.
Fig. 4C. See caption under Figure 4A.
Coelomic EHb C1, Figure Figure 5. Partial normal-phase HPLC chromatograms of 2-AB labeled oligosaccharides from C1. Peak annotations are as in Figure 3. (A) Glycan pool (B) isolated H9N2 after treatment with rat liver glucosidase II; (C) glycan pool (D) isolated H10N2 after treatment with rat liver glucosidase II. Compound 1 ((HexNAc)2(Man)9(Glc)1) from EHb C1 gave a weak PSD spectrum but clearly showed that the (Hex)6 B-type ion (m/z 995.8) from the 6-antenna was much stronger than the (Hex)7 ion confirming the substitution of the glucose on the 3-antenna as shown in Table I. A reference PSD spectrum of (GlcNAc)2(Man)9(Glc)3 also showed an enhanced relative abundance for this ion confirming that it did not contain the glucose substituents. In a reference spectrum of (GlcNAc)2(Man)8(Glc)3 this ion appeared at m/z 834 as it contained one less mannose residue than in the spectrum of (GlcNAc)2(Man)9(Glc)3. Fragment ions in the PSD spectra of (HexNAc)2(Man)8 and (HexNAc)2(Man)8, shown in Figure Table I summarizes the data obtained for EHbs V2 and C1. Quantification of the individual species was based on integration of the respective fluorescent NP-HPLC signal.
Discussion
Hitherto, the glycosylation of annelid EHbs HBL has not been studied in great detail, but some cases have been reported in which glycosylation was indicated by diagnostic mass differences in ESI-MS spectra (Lumbricus terrestris; Martin et al., 1996) or staining of 2-D protein gels with lectins (Perinereis aibuhitensis; Ebina et al., 1995). However, glycosylation does not appear to be a general feature of certain constituent polypeptide chains of the EHbs. In both cases mentioned above, carbohydrates were associated with linker chains as well as globin chains. For Riftia it was found that only V2 and C1 which lack linker chains were glycosylated via the attachment of a single oligosaccharide molecule per globin chain a. In contrast, V1 which contains linker chains appeared to be unglycosylated (Zal et al., 1996a). Nonetheless, glycosylation is not believed to be random because no other constituent of the EHbs, apart from globin chain a, is glycosylated. Furthermore, this chain is fully glycosylated since no signal for unglycosylated protein could be detected in the ESI-MS spectrum (Zal et al., 1996a).
The function of the carbohydrates on globin chain a of Riftia remains unknown. However, for another marine worm (Perinereis aibuhitensis), it has recently been demonstrated that removal of N-linked carbohydrates resulted in the irreversible dissociation of the EHb HBL complex (Ebina et al., 1995). Moreover, the amino acid sequence of the linker chains showed some similarity to lectins. The authors concluded that assembly into the quaternary structure of EHbs was possibly mediated by lectin-like protein-carbohydrate interactions, a mechanism which was termed 'carbohydrate gluing." EHbs V2 and C1 from Riftia do not contain linker chains. Whether glycosylation of the globin chain a could serve a similar function needs to be investigated.
In none of the previously mentioned studies were the carbohydrates actually isolated and their structures determined. This study has shown that the N-linked carbohydrates found on Riftia EHbs V2 and C1 are mainly of the high-mannose type with some of the carbohydrate structures capped with one or two terminal glucose residues. The observation of partial glucose capping is interesting because secreted glycoproteins do not usually contain these structures. In the well established pathway for protein glycosylation in eukaryotes (Kornfeld and Kornfeld, 1985), a carbohydrate precursor structure (GlcNAc)2(Man)9(Glc)3 is cotranslationally attached to the growing polypeptide chain. The terminal glucose residue is subsequently trimmed back to (GlcNAc)2(Man)9(Glc)2 by glucosidase I and then further processed to (GlcNAc)2(Man)9 by glucosidase II inside the Golgi compartment of cells before either further processing of the glycans or secretion of the protein occurs.
Since the EHbs of Riftia are extracellular proteins, but still display some glucose capping, they currently represent the only example of immature oligosaccharide structures on an otherwise normal and functional secreted glycoprotein of a metazoan organism. In fact, there appear to be only two reports in the literature in which an unusual (GlcNAc)2(Man)6(Glc)1 structure from a membrane-bound cell surface glycoprotein of the unicellular parasite Leishmania mexicana amazonensis was described (Olafson et al., 1990; Funk et al., 1997). These authors explain the presence of the glucosylated structure by a lack of glucosidase II activity or a low level of activity of this enzyme. For Riftia it appears that, although glucosidase II is present (mono- and di-glucosylated structures are observed), the enzyme does not act very efficiently, resulting in some 9-13% of all oligosaccharides on EHbs V2 and C1, respectively, being poorly processed.
Materials and methods
Animal sampling
Specimens of Riftia used in this study were collected from 2600 m depth at 13°N site (12°46[prime]N-103°56[prime]W and 12°50[prime]N-103°57[prime]W) on the East Pacific Rise during the HERO'92 expedition in April 1992 (F.Z.). Animals were plucked from the substrate by the manipulators of the American submersible 'Alvin" and stored at deep-sea water temperature (2-4°C) in a thermally insulated basket during the trip to the surface (2-3 h). On board, specimen which had not been damaged during collection had their tubes removed and were dissected dorsally on ice. The blood contained in the closed vascular system and in the coelomic cavity was collected separately as carefully as possible (Zal et al., 1996b). Each sample was centrifuged at low speed to remove cellular material (5 min, 4°C) and the supernatant was rapidly frozen and stored in liquid nitrogen until use.
Protein purification
The EHbs contained in the vascular blood and coelomic fluid were purified by gel filtration on a 1×30 cm Superose 6-C column (Pharmacia LKB Biotechnology, Inc.) using a low-pressure FPLC system (Pharmacia). The column was equilibrated with a saline buffer as described previously (Zal et al., 1996b). The flow rate was typically 0.5 ml/min, and protein elution was monitored at 280 nm and 414 nm. Generally, two purification steps were necessary to obtain homogeneous proteins. Peaks corresponding to the EHbs were collected and concentrated with a 10 kDa cut-off microconcentrator, Centricon-10 (Amicon).
Release of N-linked oligosaccharides
2.7 mg (6.1 nmol) Hb of the gel filtration fraction V2 from vascular blood and 3 mg (6.8 nmol) Hb of the gel filtration fraction C1 from coelomic fluid were dissolved in 10 mM ammonium bicarbonate pH 8.6, containing 1% SDS. The mixture was heated to 100°C for 5 min and then diluted 10-fold with 1% NP40 in 10 mM ammonium bicarbonate pH 8.6 (Tarentino et al., 1985). PNGase F (Boehringer Mannheim) was added to the sample to give a final enzyme concentration of 100 U/ml. The reaction mixture was incubated at 37°C for 20 h. The deglycosylated protein was precipitated by adding 4 volumes of ice-cold ethanol and incubated for 1 h on ice. After centrifugation at 10,000 × g the supernatant (which contains the glycans) was removed. The pellet was then washed with 200 µl of cold ethanol, and the combined supernatants were dried. SDS and NP40 were removed by disposable C-18 reversed-phase cartridges (Analytichem International) and mixed ion exchange chromatography (AG3/AG50, Bio-Rad) prior to MALDI MS analysis.
Electrospray ionization mass spectrometry
Purified Hb samples were extensively dialyzed against distilled water and lyophilized prior to ESI-MS. Electrospray spectra were acquired on a Quattro II (Micromass UK Ltd.) triple quadrupole mass spectrometer. Samples were infused into the ion source at a flow rate of approximately 5 µl/min and at a concentration of 0.5 µg/µl in 50/50 water/acetonitrile containing 0.2% formic acid. Fragment ion spectra were obtained by collision induced dissociation using a collision energy of 45 volts and methane (3 × 10-2 mbar) as the collision gas.
Matrix-assisted laser desorption/ionization (MALDI) mass spectrometry
Prior to MALDI-MS, samples were incubated for 10 min on a piece of Nafion membrane (Aldrich) floating on water (Börnsen et al., 1995). On the MS target the sample was then mixed with 1 µl of a saturated 2,5-dihydroxybenzoic acid (2,5-DHB) (Aldrich) solution in acetonitrile. The air-dried sample was subsequently recrystallized from ethanol (Harvey, 1993). Positive ion MALDI mass spectra were acquired on two different instruments. (1) In reflectron mode on a Perseptive Biosystems Voyager Elite mass spectrometer fitted with a pulsed nitrogen laser (337 nm) and a delayed extraction ion source. The molecular weights obtained using this instrument represent the monoisotopic masses of the sodiated ([M+Na]+) compounds. (2) In linear mode on a Thermo Bioanalysis Lasermat 2000 (Cottrell, 1992) fitted with a pulsed nitrogen laser (337 nm). Molecular weights obtained using this instrument represent the average masses of the sodiated compounds.
MALDI Post-source decay (PSD) fragmentation spectra were acquired with the Perseptive Biosystems Voyager Elite mass spectrometer with sample preparation as described above.
Normal-phase high performance liquid chromatography (NP-HPLC)
Prior to NP-HPLC, oligosaccharides were fluorescently labeled by reductive amination with 2-aminobenzamide (2-AB) (Oxford GlycoSciences) according to the manufacturer's directions. 2-AB labeled oligosaccharides were separated on a 200 × 4.6 mm PolyLC hydroxyethyl A normal phase HPLC column (Hichrom) using two Waters 510 pumps. The solvent system consisted of 50 mM formic acid (A) and 80% acetonitrile, 20% 50 mM formic acid (B). The gradient used was 0-120 min 0-40% A; 120-123 min 40-100% A; 123-128 min 100% A at a flow rate of 0.4 ml/min (Guile et al., 1996). Elution of 2-AB-labeled oligosaccharides was monitored by fluorescence detection and fractions were collected manually. The elution positions of individual oligosaccharides were measured in glucose units (GU) by calibrating the column with a partial hydrolysate of 2-AB-labeled dextran.
Exoglycosidase digestions
Samples were digested with jack bean [alpha]-mannosidase (Oxford GlycoSciences) at 100 U/ml in 20 mM sodium acetate pH 5.0 containing 2mM zinc acetate for 16h at 37°C or with rat liver [alpha]1,3 glucosidase II (purified from rat liver microsomes) using the same conditions as described by Olafson et al., 1990.
Acknowledgments
We are indebted to the members of the 'Alvin" groups, the captains and crews of R/V 'Vickers" and R/V 'Atlantis II," and the chief scientists of the research cruises, Jim Childress and Horst Felbeck, who allowed us to conduct this work. F.Z. and F.H.L. are very grateful to André Toulmond for his encouragements and constant support during all stages of this work. This study was supported by research grants from CNRS (UPR 9042), INSU, UPMC, IFREMER (URM N° 7) (F.Z. and F.H.L.). F.Z. thanks the Ministère des Affaires Étrangères ('Lavoisier" program) and the Conseil Régional de Bretagne for the grants that enabled him to work at the University of California of Santa Barbara as a postdoctoral fellow and to write this publication. B.K. and D.J.H. thank Professor Raymond A. Dwek for his support in this project, and B.K. thanks his studentship sponsor, the Biotechnology and Biological Sciences Research Council and Deutscher Akademischer Austauschdienst Grant D/94/14920. In addition, we want to mention here that B.K. and F.Z. contributed equally to this work.
Abbreviations
2-AB, 2-aminobenzamide; DHB, 2,5-di-hydroxybenzoic acid; ESI-MS, electrospray ionization mass spectrometry; Glc, glucose; GlcNAc, N-acetylglucosamine; GU, glucose unit; Hb, hemoglobin; HBL, hexagonal bilayer; EHb, extracellular Hb; Hex, hexose; HexNAc, N-acetylhexosamine; MALDI-MS, matrix-assisted laser desorption/ionization mass spectrometry; Man, mannose; NP40, Nonidet P-40, octylphenoxypolyethoxy-ethanol; NP-HPLC, normal-phase high performance liquid chromatography; PNGase F, peptide-N4-(acetyl-[beta]-glucosaminyl)asparagine amidase, EC 3.2.218; PSD, post source decay; SDS, sodium dodecyl sulfate.
References
4To whom correspondence should be addressed
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 10 Jun 1998
Copyright©Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Handelsman Metagenomics: Application of Genomics to Uncultured Microorganisms Microbiol. Mol. Biol. Rev., December 1, 2004; 68(4): 669 - 685. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Khalaila, J. Peter-Katalinic, C. Tsang, C. M. Radcliffe, E. D. Aflalo, D. J. Harvey, R. A. Dwek, P. M. Rudd, and A. Sagi Structural characterization of the N-glycan moiety and site of glycosylation in vitellogenin from the decapod crustacean Cherax quadricarinatus Glycobiology, September 1, 2004; 14(9): 767 - 774. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||