Structural characterization of the N-glycans from Echinococcus granulosus hydatid cyst membrane and protoscoleces

Molecular and Biochemical Parasitology 86 (1997) 237 – 248

Structural characterization of the N-glycans from Echinococcus granulosus hydatid cyst membrane and protoscoleces Kay-Hooi Khoo 1,a, Alberto Nieto b, Howard R. Morris a, Anne Dell a,* b

a Department of Biochemistry, Imperial College, London SW7 2AY, UK Ca´tedra de Inmunologı´a, Facultad de Quimica, Uni6ersidad de la Republica, Monte6ideo, Uruguay

Received 13 December 1996; received in revised form 21 February 1997; accepted 5 March 1997

Abstract Infection by the tapeworm Echinococcus granulosus in the intermediate host results in the development of a hydatid cyst which contains the protoscoleces within a fluid-filled cavity enclosed by the bilayered cyst membrane. N-glycans were enzymatically released from crude extracts of homogenates of hydatid cyst membranes and protoscoleces and their structures were defined by high sensitivity fast atom bombardment mass spectrometry in conjunction with sequential exoglycosidase digestions.The major N-glycans from the cyst membrane were found to be non-charged structures having complex-type antennae and core fucosylation. The antennae are either truncated at the first N-acetylglucosamine or are extended with b-galactose to form N-acetyllactosamine (lacNAc). A significant proportion of the lacNAc backbones are capped by a-galactose. The resulting Gala-Galb-terminal structures may account for the earlier observation that antibodies against the blood group P1 epitope recognise components of hydatid cyst extracts. The complex-type N-glycans identified in the protoscoleces extracts were the same as the neutral structures found in the cyst membrane but a small proportion of high mannose structures and truncated di- and trimannosyl core structures were also identified. Sialylated N-glycans were identified as minor constituents of the cyst membrane preparation but were not observed in protoscoleces extracts. Whether the sialylated glycans are host derived or endogenously synthesized by the parasite remains to be established. This is the first reported structural analysis of N-glycans from cestodes and provides new insights into protein glycosylation in helminths. © 1997 Elsevier Science Ireland B.V. Keywords: Mass spectrometry; Echinococcus granulosus; Hydatid cyst; Protoscoleces; N-glycosylation; P1 epitope

Abbre6iations: FAB-MS, fast atom bombardment-mass spectrometry; FACE, fluorophore assisted carbohydrate electrophoresis; HCWA, hydatid cyst wall antigens; PSA, protoscolex antigens; NeuAc, N-acetylneuraminic acid; NeuXc, NeuAc or NeuGc; PNGase F, peptide N-glycosidase F. * Corresponding author. Tel.: + 44 171 5945219; fax: +44 171 2250458; e-mail: [email protected] 1 Present addrees: Institute of Biological Chemistry, Academia Sinica, Taipei 115, Taiwan 0166-6851/97/$17.00 © 1997 Elsevier Science Ireland B.V. All rights reserved. PII S 0 1 6 6 - 6 8 5 1 ( 9 7 ) 0 0 0 3 6 - 4

238

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

1. Introduction The dog tapeworm Echinococcus granulosus is the causative agent of unilocular hydatid disease, a zoonotic infection affecting humans, as well as sheep, cattle and pigs. In endemic areas, hydatidosis is a major public health problem which causes important economical losses [1,2].The pathology of the disease is mainly due to the physical pressure exerted on the intermediate host’s viscera by the developing hydatid cyst containing larval worms, or protoscoleces. The wall of the hydatid cyst consists of an outer laminate membrane and an inner germinal layer from which small packets of protoscoleces bud into the cyst lumen. Although polysaccharides and the blood group P1 carbohydrate antigen have long been recognized as constituents of the cyst fluid, membrane and protoscoleces [3– 6], virtually nothing is known about specific glycosylation structures. Immunological localization studies have suggested that the P1 antigen may be synthesized within the protoscolex tegument, stored in the cyst fluid, and then diffuses through the laminated cyst membrane [7]. Other cyst fluid antigens have also been localized to the protoscoleces and cyst membrane [8,9], supporting a dynamic relationship between the fluid antigens, the hydatid cyst wall antigens (HCWA) and the protoscolex antigens (PSA). March et al. [10] further demonstrated that one of the antigens common to the fluid, the germinal membrane and the protoscoleces possesses peptide N-glycosidase F (PNGase F) sensitive complex N-glycans. Glycosylated moieties have been shown to be immunodominant in both natural and experimental hydatid infections [11,12] although antibody responses to carbohydrate epitopes in a detergent extract from the surface of protoscoleces were observed not to correlate with protection [13]. In addition, a carbohydrate enriched fraction derived from endoglycosidase F treatment of intact protoscoleces was shown to be mitogenic in vitro and elicited hypergammaglobulinemia in vivo as well as a strong specific IgM and IgG3 low avidity antibody response [14]. Activation of complement by the hydatid cyst

fluid has also been attributed to periodate-sensitive carbohydrate epitopes which could be partially released by PNGase F [15]. It thus appears that although other types of glycosylation may be equally important, available evidence largely implicates the N-glycans as the immunodominant antigens. A firm structural foundation is a pre-requisite for understanding the glycobiology of parasite antigens. Previous structural studies of cestode glycoconjugates have focused on glycosphingolipids from a few representative genera [16– 19] and no data on protein glycosylation have been reported for any cestode. We have undertaken to rigorously characterise E. granulosus glycoconjugates and results from our studies on N-glycans from HCWA and PSA preparations are reported here. We show that the main Nglycans from both HCWA and PSA are complex type structures, a significant portion of which contain antennae which terminate in a-Gal attached to subterminal b-Gal resulting in a nonreducing disaccharide which probably constitutes the P1 epitope.

2. Materials and methods

2.1. Preparation of hydatid cyst wall antigens Fertile bovine hydatid cysts were obtained from Uruguayan abattoirs, their contents were asceptically aspirated according to Carol et al. [20] and cyst membranes excised from the adventitiae. The excised cyst membranes were extensively washed with phosphate buffer saline (PBS) and sliced. Approximately 500 mg of cyst membrane slices were suspended in 5 ml of PBS with 5 mM EDTA and disrupted by sonication in an ice bath. This disruption was carried out using 15 cycles of 15 s each, at maximum power in a Blackstone sonifier (Sheffield, USA). The resulting suspension was clarified by centrifugation (10 min, 8500× g), the supernatant stored and the pellet submitted six times to the same disruptive treatment. A pool was prepared with all the supernatants (HCWA) and stored at − 20°C.

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

2.2. Preparation of protoscolex antigens One ml of sedimented protoscoleces were disrupted by sonication at 4°C in 10 ml of PBS containing 20 mM phenylmethylsulfonyl fluoride (PMSF) and 5 mM EDTA, using 15 cycles of 10 s every minute at maximum power with an ultrasonic disruptor. The resulting suspension was clarified by centrifugation (6000 ×g, 10 min at room temperature) and the supernatant (PSA) stored at − 20°C.

2.3. Protein analysis The protein content of HCWA and PSA was determined by the bicynchoninic acid method (BCA Protein Assay Reagent, Pierce, IL) using bovine serum albumin (BSA) as standard, according to the manufacturer’s instructions.

2.4. Fluorophore assisted carbohydrate electrophoresis analysis Monosaccharide composition analysis and Nlinked oligosaccharide profiling were carried out according to Jackson [21] using the reagents and instructions supplied with the corresponding fluorophore assisted carbohydrate electrophoresis (FACE) kits (Glyko, Novato, CA). Photographic records were obtained using an ultra violet (UV) transilluminator (UVP Chromato Vue, Model TM20, 302 nm). The HCWA samples (4 mg ml − 1 protein) used were 45 ml for the profiling assay and 15 ml for the monosaccharide composition assay. The composition analysis of fluorophore labelled free oligosaccharides (using an aliquot from the solution used as the sample in the profiling of HCWA) was carried out according to the manufacturer’s instructions.

2.5. Preparation of glycans from hydatid cyst wall antigens and protoscolex antigens The crude extracts containing glycoproteins were first digested with non-TPCK treated trypsin (Sigma) (1:50 enzyme:protein ratio, w/w; 37°C for 5 h in 50 mM ammonium bicarbonate

239

buffer, pH 8.4) followed by C18 Sep-pak (Waters) purification. Glucan polymer and other hydrophilic contaminants were washed off with 5% acetic acid and the bound peptides/glycopeptides were eluted with a step gradient of 20, 40 and 60% propan-1-ol in 5% acetic acid. All the eluted fractions were pooled, dried down and then incubated with N-glycosidase F (1 unit, Boehringer–Mannheim) overnight at 37°C in 50 mM ammonium bicarbonate buffer, pH 8.4. Released N-glycans were separated from peptides/ glycopeptides using the same C18 Sep-pak procedure.

2.6. Sequential exo-glycosidase digestions Neuraminidase (from Vibrio cholerae, EC 3.2.1.18, Boehringer–Mannheim): 50 mU in 100 ml of ammonium acetate buffer at pH 5.5 for 24 h; N-Acetyl-b-D-hexosaminidase (from bovine kidney, EC 3.2.1.30, Boehringer–Mannheim): 0.2 U in 100 ml of 50 mM sodium-citrate-phosphate buffer at pH 4.6 for 18 h; b-galactosidase (from bovine testes, EC 3.2.1.23 Boehringer– Mannheim): 10 mU in 100 ml of 50 mM sodium-citrate-phosphate buffer at pH 4.6 for 24 h; a-galactosidase (from green coffee bean, EC 3.2.1.22, Boehringer–Mannheim): 0.5 U in 100 ml of 50 mM sodium-citrate-phosphate buffer at pH 6.0 for 24 h. All enzyme digestions were incubated at 37°C and terminated by boiling for 3 min before lyophilisation. An appropriate aliquot was taken after each digestion and permethylated for fast atom bombardment-mass (FAB-MS) analysis.

2.7. Chemical deri6atization and fast atom bombardment-mass analysis Samples were permethylated using the NaOH/ dimethyl sulfoxide slurry method as described by Dell et al. [22]. FAB-mass spectra were obtained using a VG Analytical ZAB-2SE 2FPD mass spectrometer fitted with a cesium ion gun operated at 25–30 kV. Data acquisition and processing were performed using the Opus® software. Monothioglycerol was used as matrix.

240

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

3. Results

3.1. Face analysis of hydatid cyst wall antigens HCWA samples were first investigated for sugar content and for N-glycan complexity using FACE [21]. N-glycans were released by peptide N-glycosidase F (PNGaseF) and subjected to FACE analysis giving the data in Fig. 1. Several bands were observed migrating in the region of the gel corresponding to 5.5 – 11 glucose units (Fig. 1, lane 2) which were not present in the non-PNGaseF-treated control (Fig. 1, lane 3). After hydrolysis the soluble fraction produced by PNGaseF treatment was shown by FACE analysis to contain GlcNAc, Gal, Glc, Fuc, Man and NeuAc (Fig. 2, Lane 3). The de-N-glycosylated pellet from PNGaseF-treated HCWA showed Gal, Man (very faint) and GalNAc (Fig. 2, lane 5). The absence of GlcNAc in the pellet suggested that the majority of N-glycans had been released by PNGaseF.

Fig. 2. FACE analysis of monosaccharides obtained by hydrolysis of HCWA and products of PNGaseF treatment. Lane 1, NeuAc standard. Lane 2, HCWA. Lane 3, N-linked oligosaccharides released from HCWA. Lanes 4 and 6, monosaccharide standard mixture. Lane 5, HCWA pellet after PNGaseF N-linked oligosaccharides had been released by PNGaseF.

3.2. Fast atom bombardment-mass spectometry strategy

Fig. 1. FACE analysis of oligosaccharides released from HCWA by PNGaseF. Lane 1, oligoglucose ladder standard. The low intensity band corresponds to the tetrasaccharide (four glucose units (GU)). Lane 2, fluorescently labelled oligosaccharides from the supernatant of PNGaseF digests of HCWA. The GU value corresponding to each of the major bands is shown on the right of the gel. Lane 3, control showing the oligosaccharide bands obtained following the same protocol used for preparing the sample run in lane 2, but without the PNGaseF.

Taking into account the results of the FACE profiling, which indicated that significant quantities of N-glycans could be released from HCWA preparations by PNGaseF digestion, we chose to focus our sophisticated mass spectrometric strategies on this class of glycans in the first phase of our Echinococcus structural study. A highly sensitive FAB-MS based strategy, which we have successfully applied in structural studies of other helminths, was employed to probe the structures of the N-glycans released by PNGase F from HCWA and PSA extracts. This strategy involves FAB-MS analyses of mixtures of permethylated glycans and exploits the unique strengths of FABMS in providing high quality molecular ion profiles as well as unambiguous fragment ion information in a single experiment. Because complex mixtures are amenable to FAB-MS, very few purification steps are required prior to FAB analysis thereby minimising sample losses. To facilitate glycan release in the absence of detergents (which are incompatible with subsequent MS

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

241

Fig. 3. The molecular ion region of the permethyl derivatives of HCWA. Major signals were assigned as in Table 1. Signals at m/z 1117, 1362, 1607, and 1770 are A-type fragment ions corresponding to Hex3HexNAc2+ , Hex3HexNAc3+ , Hex3HexNAc4+ , and Hex5HexNAc3+ respectively. These are derived from cleavage between the two GlcNAc of the chitobiose core of the major components and provided further evidence that the single fucose in most components are located at the reduing end GlcNAc residue. m/z value labelled are monoisotopic nominal mass corresponding to those calculated and tabulated in Table 1. The actual monoisotopic mass for signals above m/z 2000 and 2500 were 1 and 2 units higher than the nominal mass respectively.

analyses), HCWA and PSA homogenates were digested with trypsin prior to PNGase F treatment. Since the FACE analyses had indicated a high level of glucose (Fig. 2), indicative of glucose-containing polymers being possibly present in the homogenates, hydrophilic impurities were removed from the tryptic digests before the release of N-glycans. This was achieved by a simple Sep-pak purification of the peptide/glycopeptide products of tryptic digestion prior to PNGase F treatment. Once the glycans were released by PNGaseF they were separated from peptides and any remaining glycopeptides by a second Sep-pak step and were then permethylated prior to FAB-MS analyses.The structures of the glycans released by PNGase F treatment were tentatively assigned from molecular and fragment ion data and corroborated by sequential exo-glycosidase digestions monitored by FAB-MS. Data from these analyses are described below.

3.3. N-glycan profile of hydatid cyst wall antigens Preparations containing about 100 mg equivalent of protein/glycoprotein material, as estimated by protein assay, were analysed as described in the previous section. Five major A-type fragment ions were present in the low mass region of the FAB-mass spectrum of the permethyl derivative (data not shown) which defined the five major types of non-reducing terminal structures, namely HexNAc (m/z 260), Hex-HexNAc (m/z 464), HexHex-HexNAc (m/z 668), NeuAc-Hex-HexNAc (m/z 825), and NeuGc-Hex-HexNAc (m/z 855). Based on the molecular mass values afforded (Fig. 3), the range of non-reducing terminal structures present, and the sequential enzyme digestion data, the compositions and the probable structures of the major N-glycans in HCWA may be assigned as in Table 1. The absence of A-type fragment ions corresponding to9 Fuc1HexNAc2+ (m/z 505, 679) or

242

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

Table 1 The major N-glycans from the HCWA of E. granulosus

a The sample was treated with neuraminidase, b-galactosidase, and b-N-acetylhexosaminidase sequentially. After each digestion, an aliquot was removed, permethylated and analyzed by FAB-MS. ‘x’ denotes resistant to the particular enzyme used. When susceptible, the first number indicates how many NeuXc, Gal or HexNAc residues have been removed, followed by the new m/z values of the resulting digestion products, as found in Figs. 2A, 2B and 3A respectively. NeuXc denotes NeuAc and NeuGc which differ in mass by 30 u after permethylation. The neuraminidase used does not distinguish between these two sialic acids which are equally susceptible to digestion. b The persistence of the signal at m/z 2221 after b-galactosidase digestion indicated that it may be a mixture of two components as shown. c Probable structures consistent with sequential enzyme digestion and FAB-MS data. The distribution of peripheral and terminal structures on each arm of the trimannosyl core was not defined.

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

243

Fig. 4. The molecular ion region of the permethyl derivatives of desialylated HCWA. Most of the signals are as assigned in Table 1 and Fig. 3. A major new signal is present at m/z 2047, corresponding to the desialylated product from the major sialylated components (see Table 1). Another signal at m/z 2466 corresponds to Fuc1HexNAc5Hex5. Other signals at 22 units higher correspond to sodiated molecular ions.

Fuc1 Hex1 HexNAc1+ (m/z 638) indicates that antenna terminating with these structures are not present at detectable levels. Specifically, this implies that lacdiNac (GalNAcb1 “ 4GlcNAcb1“ ) and its fucosylated counterpart (GalNAcb1“ 4[Fuca1“3]GlcNAcb1 “), and the Lewisx epitope (Galb1“4[Fuca1 “ 3]GlcNAcb1 “) are either very minor or absent altogether. Trimming of the antenna with sequential exo-glycosidase digestions (Table 1, Figs. 3 – 6) largely abolished the glycosylation heterogeneity and yielded two major products of compositions 9 Fuc1Hex3HexNAc2 (m/z 1149 and 1323, Fig. 6), attributable to the trimannosyl N,N%-diacetyl-chitobiose core with and without a-6 fucosylation. In addition, two other products were identified (m/z 1976 and 2629, Fig. 6A) from which one and two a-Gal residues could be respectively removed by further digestion with a-galactosidase (Fig. 6B). Thus, the major N-glycans of HCWA are based on conventional complex type structures. The two most abundant components (m/z 1813 and 2058) have compositions consistent with two and three

HexNAc residues respectively being attached to the trimannosyl core (Table 1). The absence of a fragment ion for HexNAc2+ indicates that these HexNAc residues are separately attached to the core as shown in Table 1 and do not constitute a lacdiNAc moiety. However we cannot rule out minor amounts of the latter structure. Other larger and less abundant components have additional b-Gal and a portion of these structures may be capped by NeuAc/NeuGc or a-Gal. Interestingly, all NeuAc/NeuGc-capped glycans were found to be non-fucosylated whereas those without NeuAc/NeuGc carried a fucose on the core. The linkage of the core fucose was defined as a-6 by the specificity of PNGaseF which cannot release glycans carrying 3-linked fucose [23].

3.4. N-glycan profile of protoscolex antigens In comparison with HCWA, the yield of N-glycans from a 100 mg protein/glycoprotein equivalent of PSA homogenate appeared to be considerably lower. Thus, molecular ion signals in

244

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

Fig. 5. The molecular ion region of the permethyl derivatives of desialylated HCWA further digested with b-galactosidase. A major new signal is present at m/z 1639 (see Table 1).

the FAB-spectra of permethyl derivatives were weaker as indicated by the lower signal to noise ratio (Fig. 7, compare Fig. 3). Although some of the complex type N-glycans were common to both HCWA and PSA, notably those giving molecular ions at m/z 1568, 1813 and 2629 (see Table 1 for assignments), the latter did not appear to contain any of the sialylated components. In addition, a number of molecular ions were unique to PSA, the compositions of which corresponded to high mannose and truncated classes of N-glycans. Thus, signals at m/z 2373, 1965, 1557 and 1353 correspond to [M + H] + molecular ions of Hex9HexNAc2, Hex7HexNAc2, Hex5HexNAc2 and Hex4HexNAc2, respectively, whereas signals at m/z 945/1119 and 1149/1323 can be assigned to 9 Fuc(Hex2HexNAc2) and 9 Fuc (Hex3HexNAc2) respectively. The last pair were observed in HCWA experiments but only after sequential enzyme digestions (Fig. 4), and are likely to be the molecular ions for the trimannosyl N,N%-diacetyl-chitobiose core with and without a-6 fucosylation.

4. Discussion Recent structural studies on the glycans from parasitic helminths including those from the trematodes and nematodes have gradually furnished a much needed basic understanding of glycosylation in these lower animals [24–28]. Typically, the N-glycans from parasitic worms more closely resemble those found in plants and insects, rather than their better understood mammalian counterparts and may be summarized as having the following characteristics: (i) the trimannosyl N,N%-diacetyl-chitobiose core may carry a-3 fucosylation in place of or in addition to a-6 fucosylation; (ii) a range of truncated structures containing only two or three mannoses linked to the N,N%-diacetyl-chitobiose core with different combinations of core fucosylation are usually present and are sometimes the dominant structures; (iii) lacdiNac-based antennae, which are rare in mammals, are as common as the lacNActype which typify mammalian glycans; (iv) highly unusual or novel saccharide and non-saccharide substituents are often present which contribute to

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

245

Fig. 6. The molecular ion region of the permethyl derivatives of HCWA after sequential digestions with neuraminidase, b-galactosidase and b-hexosaminidase (A). The inset shows the two components (m/z 1976 and 2629) with one and two terminal a-galactoses respectively (see Table 1) which could be further released by a-galactosidase to give the spectrum in (B). Signals at 17 units higher than m/z 1149 and 1323 correspond to [M+ NH4] + molecular ions.

246

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

Fig. 7. The molecular ion region of the permethyl derivatives of PSA.

their antigenicity; and (v) there is a general lack of sialylation. Interestingly, the majority of the N-glycans observed in this study do not exhibit the above characteristics. For example the major complextype glycans carry only ‘conventional’ lacNAc antennae and cores are a6-fucosylated and not a3. The absence of terminal sequence ions which would correspond to lacdiNAc led us to conclude that the major glycans in HCWA are bi- and triantennary with non-extended terminal GlcNAcs on the mannoses although it should be pointed out that the compositions of the molecular ions are not incompatible with the presence of some lacdiNAc termini which failed to produce significant quantities of fragment ions due to their low abundance. The presence of a Gala “Galb “ GlcNAcb“ terminal structure is consistent with the long recognized presence of the blood group P1 epitope in the hydatid cyst. In addition to the cyst fluid, the blood group P1 antigen has previously been immuno-localized to the protoscolex tegument and the laminated layer of the cyst membrane [7].

Russi et al. [6] isolated a P1-containing glycoprotein antigen from the cyst membrane whilst Dennis et al. [17] identified the P1 epitope in the glycolipids isolated from the germinal layer of the cyst membrane. Thus the P1-epitope appears to be carried by both glycoproteins and glycolipids and is present on both HCWA and PSA. Although linkage data are required to rigorously confirm that the Gal-rich structures identified in this FAB study constitute the P1 epitope Gala“ 4Galb“ 4GlcNAcb“ [29,30] and not alternatives such as the common Gala“ 3Galb“ 4GlcNAcb “ terminal sequence, our data suggest that major carriers of the P1 epitope in both HCWA and PSA are likely to be biantennary complex type N-glycans. The major difficulty in structural analysis of glycosylated parasite antigens is the limited supply of sample material, often only at microgram levels, and usually only available as complex mixtures in crude extracts containing large amounts of impurities including polysaccharides. A mass spectrometric approach such as that adopted in this study allows high sensitivity definitive prob-

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

ing of the compositions and terminal structures of glycans present within complex mixtures of tissue extracts without resorting to extensive purification which is often not practicable. When coupled with specific exo-glycosidase digestions, additional information on the stereochemistry and sequence can be obtained although in most cases detailed linkage data requires experiments such as methylation analysis. In many cases, including the current study, linkage data are not very informative because of the complexity of the sample coupled with the presence of contaminating polsaccharides which were not observed in the FAB analyses because of their size. Studies are in progress to address questions of linkage. Another difficulty associated with the analysis of parasite samples not derived from in vitro culture is the possibility of contamination by hostderived molecules. This could either be a consequence of the extraction and isolation procedures, or possibly that the parasites actually actively adsorb and incorporate host molecules. Thus, although sialylated components were found in high abundance in the HCWA, it is possible that these were all host-derived. In fact, we have observed that the relative amount of sialylated components to non-sialylated ones fluctuated between batches and in some cases were not detected. The other glycans were detected at comparable levels in all batches consistent with parasite origin. Supporting this conclusion is the fact that the truncated glycans shown in the upper panels of Table 1 have not been observed in bovine glycoproteins. It is also noteworthy that whilst most of the non-sialylated components are core fucosylated, the sialylated ones are not, implying that they may be derived from two distinct sources. Furthermore, in recent unpublished work we have shown that the N-glycan profile of HCWA samples immunopurified by a monoclonal which reacts with HCWA, but not with several host tissues tested, is similar to the one we report in this paper minus the sialylated components. On the other hand, there are sporadic reports on the occurrence of sialylated glycans in the lower animals. Relevant to current studies, terminal sialic acid residues have been detected on the acidic glycolipid fractions of E. granulosus hydatid

247

cyst [17] and gangliosides from the related Echinococcus multilocularis metacestodes have been identified and characterized [19]. In the former case, the authors cautioned that the sialylated glycolipids may not be endogenous but acquired from the infected host tissue. Thus, although there is little doubt that sialylated glycans are present among the isolated HCWA material, their actual status in vivo remains to be established. It is of particular interest to determine if the prevalent view that helminth parasites, in common with plants and insects, do not possess endogenous sialylation machinery can be extended to the cestoda class.

Acknowledgements This work was supported by grants to H.R.M. and A.D. from the Medical Research Council, the Biotechnology and Biological Sciences Research Council and the Wellcome Trust (Grant 030826) and to A.N. from CSIC (Universidad de la Republica, Uruguay), The Royal Society, SAREC, EC (Grant TS3*CT91-0038) and CONICYT-BID (Grant 126).

References [1] McManus, D.P. and Smyth, J.D. (1986) Hydatid disease (hydatidosis): changing concepts in epidemiology and speciation. Parasitol. Today 2, 163 – 168. [2] Nieto, A., Ehrlich, R. and McManus, D. (1988) Regional workshop on hydatid disease. Parasitol. Today 4, 81. [3] Cameron, G.L. and Staveley, J.M. (1957) Blood group P substance in hydatid cyst fluids. Nature 179, 147 – 148. [4] Korc, I., Hierro, J., Lasalvia, E., Falco, M. and Calcagno, M. (1967) Chemical characterization of the polysaccharide of the hydatid membrane of Echinococcus granulosus. Exp. Parasitol. 20, 219 – 224. [5] Kilejian, A. and Schwabe, C.W. (1971) Studies on the polysaccharides of the Echinococcus granulosus cyst, with observations on a possible mechanism for laminated membrane formation. Comp. Biochem. Physiol. B 40, 25 – 36. [6] Russi, S., Siracusano, A. and Vicari, G. (1974) Isolation and characterization of a blood P1 active carbohydrate antigen of Echinococcus granulosus cyst membrane. J. Immunol. 112, 1061 – 1069.

K.-H. Khoo et al. / Molecular and Biochemical Parasitology 86 (1997) 237–248

248

[7] Makni, S., Ayed, K.H., Dalix, A.M. and Oriol, R. (1992) Immunological localization of blood P1 antigen in tissues of Echinococcus granulosus. Ann. Trop. Med. Parasitol. 86, 87 – 88. [8] Sanchez, F., Garcia, J., March, F., Cardenosa, N., Coll, P., Munoz, C., Auladell, C. and Prats, G. (1993) Ultrastructural localization of major hydatid fluid antigens in brood capsules and protoscoleces of Echinococcus granulosus of human origin. Parasite Immunol. 15, 441–447. [9] Sanchez, F., March, F., Mercader, M., Coll, P., Munoz, C. and Prats, G. (1991) Immunochemical localization of major hydatid fluid antigens in protoscoleces and cysts of Echinococcus granulosus from human origin. Parasite Immunol. 13, 583–592. [10] March, F., Enrich, C., Mercader, M., Sanchez, F., Munoz, C., Coll, P. and Prats, G. (1991) Echinococcus granulosus: antigen characterization by chemical treatment and enzymatic deglycosylation. Exp. Parasitol. 73, 433 – 439. [11] Ferragut, G. and Nieto, A. (1996) Antibody response of Echinococcus granulosus infected mice: recognition of glucidic and peptidic epitopes and lack of avidity maturation. Parasite Immunol. 18, 393–402. [12] Sterla, S., Ljungstro¨m, I. and Nieto, A. (1997) Modified ELISA for hydatid serodiagnosis: the potential of periodate treatment and phosphorylcholine inhibition. Serodiagnosis and immunotherapy in infectious diseases. Parasite Immunol. 8, 145–148. [13] Herna´ndez, A. and Nieto, A. (1994) Induction of protective immunity against murine secondary hydatidosis. Parasite Immunol. 16, 537–544. [14] Mı´guez, M., Baz, A. and Nieto, A. (1996) Carbohydrates on the surface of Echinococcus granulosus protoscoleces are immunodominant in mice. Parasite Immunol. 18, 559 – 569. [15] Ferreira, A.M., Wu¨rzner, R., Hobart, M.J. and Lachmann, P.J. (1995) Study of the in 6itro activation of the complement alternative pathway by Echinococcus granulosus hydatid cyst fluid. Parasite Immunol. 17, 245– 251. [16] Dennis, R.D. and Wiegandt, H. (1993) Glycosphingolipids of the invertebrata as exemplified by a cestode platyhelminth, Taenia crassiceps, and a dipteran insect, Calliphora 6icina. Adv. Lipid Res. 26, 321–351. [17] Dennis, R.D., Baumeister, S., Irmer, G., Gasser, R.B. and Geyer, E. (1993) Chromatographic and antigenic properties of Echinococcus granulosus hydatid cyst-derived glycolipids. Parasite Immunol. 15, 669–681. [18] Kawakami, Y., Nakamura, K., Kojima, H., Suzuki, M., Inagaki, F., Suzuki, A., Sonoki, S., Uchida, A., Murata, Y. and Tamai, Y. (1993) A novel fucosylated glycosphingolipid with a Galb1-4Glcb1-3Gal sequence in plerocercoids of the parasite, Spirometra erinacei. J. Biochem.

. .

(Tokyo) 114, 677 – 683. [19] Persat, F., Bouhours, J.-F., Petavy, A.-F. and Mojon, M. (1994) Gangliosides of Echinococcus multilocularis metacestodes. Biochim. Biophys. Acta Mol. Basis Dis. 1225, 297 – 303. [20] Carol, H., Hernandez, A., Baz, A. and Nieto, A. (1989) lack of interspecies barriers in anti-Id stimulated antibody production against Echinococcus granulosus antigens. Parasite Immunol. 11, 183 – 195. [21] Jackson-P (1994) The analysis of fluorophore-labeled glycans by high- resolution polyacrylamide gel electrophoresis. Anal. Biochem. 216, 243 – 252 [22] Dell, A., Reason, A.J., Khoo, K.-H., Panico, M., McDowell, R.A. and Morris, H.R. (1994) Mass spectrometry of carbohydrate-containing bipolymers. Methods Enzymol. 230, 108 – 132. [23] Tretter, V., Altmann, F., and Ma¨rz, L. (1991). PeptideN 4-(N-acetyl-b-glucosaminyl)asparagine amidase F cannot release glycans with fucose attached a1 – 3 to the asparagine-linked N-acetylglucosamine residue. Eur. J. Biochem. 199, 647 – 652. [24] Khoo, K.-H., Morris, H.R. and Dell, A. (1993) Structural characterisation of the major glycans of Toxocara canis ES antigens. In: Toxocara and Toxocariasis: Clinical, Epidemiological and Molecular Perspectives (Lewis, J.W. and Maizels, R.M., eds.), pp. 133 – 140. Institute of Biology. Birbeck & Sons Ltd., Birmingham. [25] Kang, S., Cummings, R.D. and McCall, J.W. (1993) Characterization of the N- linked oligosaccharides in glycoproteins synthesized by microfilariae of Dirofilaria immitis. J. Parasitol. 79, 815 – 828. [26] Reason, A.J., Ellis, L.A., Appleton, J.A., Wisnewski, N., Grieve, R.B., McNeil, M., Wassom, D.L., Morris, H.R. and Dell, A. (1994) Novel tyvelose-containing tri- and tetra-antennary N-glycans in the immunodominant antigens of the intracellular parasite Trichinella spiralis. Glycobiology 4, 593 – 604. [27] Cummings, R.D. and Nyame, A.K. (1996) Glycobiology of schistosomiasis. FASEB J. 10, 838 – 848. [28] Haslam, S.M., Coles, G.C., Munn, E.A., Smith, T.S., Smith, H.F., Morris, H.R. and Dell, A. (1996) Haemonchus contortus glycoproteins contain N-linked oligosaccharides with novel highly fucosylated core structures. J. Biol. Chem. 271, 30561 – 30571. [29] Cory, H.T., Yates, A.D., Donald, A.S., Watkins, W.M. and Morgan, W.T. (1974) The nature of the human blood group P1 determinant. Biochem. Biophys. Res. Commun. 61, 1289 – 1296. [30] Naiki, M., Fong, J., Ledeen, R. and Marcus, D.M. (1975) Structure of the human erythrocyte blood group P1 glycosphingolipid. Biochemistry 14, 4831 – 4837.