Biochemical characterisation of the 56 and 82 kDa immunodominant gametocyte antigens from Eimeria maxima

International Journal for Parasitology 32 (2002) 805–816 www.parasitology-online.com

Biochemical characterisation of the 56 and 82 kDa immunodominant gametocyte antigens from Eimeria maxima Sabina I. Belli a,*, Michelle Lee a,1, Per Thebo b, Michael G. Wallach c, Boris Schwartsburd d, Nicholas C. Smith a a

Institute for the Biotechnology of Infectious Diseases, University of Technology Sydney, Gore Hill, Westbourne Street, Sydney, N.S.W. 2065, Australia b Swedish National Veterinary Institute, Uppsala, Sweden c ABIC Veterinary Products Ltd, Netanya, Israel d Department of Animal Science, The Hebrew University of Jerusalem, Rehovot, Israel Received 19 December 2001; received in revised form 11 January 2002; accepted 15 January 2002

Abstract Two immunodominant gametocyte antigens from Eimeria maxima with Mr 56 kDa and Mr 82 kDa have been identified previously as potential candidates for inclusion in a recombinant subunit vaccine against coccidiosis in poultry. Here, these proteins have been biochemically characterised, immunolocalised within the parasite, and sequences for their amino termini determined. These antigens co-purify by affinity chromatography suggesting an interaction with each other. However, separation of the proteins by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) in the absence of b-mercaptoethanol did not reveal the presence of inter-chain disulphide bonds. The true masses of the 56 and 82 kDa antigens are 52 450 and 62 450 Da, respectively, as determined by mass spectrometry. TX-114 separations suggested that they exist, in part, as soluble proteins within the parasite, and immunolocalisation studies indicated that they were found in the wall forming bodies of macrogametocytes. Separation of the proteins by 2D SDS-PAGE revealed that they are acidic in nature and heterogeneous in charge. Cleavage by neuraminidase and O-glycosidase indicated that the presence of O-linked glycans contributed to some of the charge microheterogeneity of both proteins. The absence of these O-glycans however, did not abolish antibody recognition, suggesting that the development of a recombinant subunit vaccine is possible. A more extensive investigation of the carbohydrate moieties of these proteins revealed that they also possess glucose, fucose, mannose and galactose. There was no evidence for the presence of N-linked glycans. The 56 and 82 kDa antigens were separated from a mixture of proteins in a crude gametocyte lysate by 2D SDS-PAGE, the proteins isolated, and the N-terminus amino acid sequence determined. They showed no homology to each other at the N-terminus, or to any other previously characterised protein. Characterisation of these novel proteins has provided further insights into the molecular mechanisms of gametocyte differentiation in E. maxima. q 2002 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved. Keywords: Eimeria; Gametocyte antigens; Localisation; 2D sodium dodecyl sulphate polyacrylamide gel electrophoresis; Glycosylation; N-terminus

1. Introduction Eimeria maxima is a parasitic protozoan belonging to the phylum Apicomplexa. This species of Eimeria, along with Eimeria acervulina, Eimeria tenella, Eimeria praecox, Eimeria necatrix, Eimeria brunetti, and Eimeria mitis parasitises the gut epithelium of poultry, giving rise to the debilitating disease known as coccidiosis (Raether, 1988). Two gametocyte antigens, with Mr 56 kDa and Mr 82 kDa, have been shown previously to be recognised by IgG in sera taken from chickens which had recovered from infection with E. * Corresponding author. Tel.: 161-2-9514-4043; fax: 161-2-9514-4026. E-mail address: [email protected] (S.I. Belli). 1 Present address: Children’s Cancer Institute Australia, Sydney Children’s Hospital, High Street, Randwick, 2031, Australia.

maxima (Wallach et al., 1989). In breeding hens, these antibodies are transferred to the developing embryo via the egg yolk, providing partial immunity to chicks upon hatching (Wallach et al., 1992). These observations suggest that it will be possible to develop a subunit vaccine based on maternal immunity which includes the 56 and 82 kDa antigens (Wallach et al., 1995). Here we describe the biochemical characterisation of the 56 and 82 kDa gametocyte antigens, and their location within the parasite. In particular, we determine their true sizes by mass spectrometry, describe their abberant migration in sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) gels, and demonstrate the presence of charged isoforms by 2D SDS-PAGE. Moreover, we analyse

0020-7519/02/$20.00 q 2002 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved. PII: S 0020-751 9(02)00011-5

806

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

the sugar moieties present on these proteins, and assess their importance in antibody recognition.

2. Materials and methods 2.1. Animals, parasites and proteins Chickens (10 week-old Australorps) were used for infections. They were housed at Gore Hill Research Laboratories at 218C with a 12 h light/dark cycle and free access to food and water. The Houghton strain of E. maxima was kindly provided by Dr Martin Shirley (Institute for Animal Health, Compton). The oocysts were passaged in SPF chickens at the Swedish National Veterinary Institute, Uppsala, and purified as described previously (Shirley, 1995). Gametocytes were isolated at the Swedish University of Agricultural Sciences, Uppsala, and purified from infected chicken intestines following techniques published previously (Wallach et al., 1989). They were stored as aliquots at 2808C until ready to use. The mAb 1E11-11 was used to affinity purify the 56 kDa gametocyte antigen from a crude gametocyte detergent lysate as described previously (Wallach et al., 1990). It has been noted that an additional protein of 82 kDa co-elutes with the 56 kDa protein in this affinity purification (Wallach et al., 1990). This protein preparation known as affinity purified gametocyte antigen was lyophilised and stored at 2208C. Just prior to use, the affinity purified gametocyte antigen was solubilised in PBS (145 mM NaCl, 7.5 mM Na2HPO4, 2.5 mM NaH2PO4 2H2O pH 7.1). 2.2. SDS-PAGE Gametocytes were lysed in 100 ml 0.5% sodium deoxycholate in PBS for 30 min on ice. Samples were then centrifuged (12 000 £ g, 5 min, 48C) to remove insoluble debris, and 4 £ Laemmli sample buffer (1 £ ¼ 62 mM Tris–HCl pH 6.8, 10% glycerol, 2.3% SDS, 0.5% bromophenol blue) added to the supernatant. Half of the sample was transferred to a new tube and b-mercaptoethanol added to a final concentration of 5%. For the affinity purified gametocyte antigen, 5 mg of protein was added to sample buffer, in the presence or absence of 5% b-mercaptoethanol, in a final volume of 10 ml. Gametocyte and the affinity purified gametocyte antigen samples were then heated for 2 min at 1008C, followed by centrifugation (12 000 £ g, 5 min) to remove insoluble and aggregated matter. Proteins were then separated on polyacrylamide gels according to protocols published previously (Laemmli, 1970) and visualised by staining in Coomassie Blue (0.2% Coomassie Brilliant Blue R (Sigma), 50% methanol, 10% glacial acetic acid) for 30 min followed by destaining (12% ethanol, 7% glacial acetic acid). Alternatively, proteins were transferred to polyvinylidene difluoride membrane (Pall Corporation) as described below.

2.3. Immunoblotting Proteins were transferred to polyvinylidene difluoride membrane for 2 h at 100 V according to protocols published previously (Harlow and Lane, 1988). Non-specific binding sites on membranes were blocked in 5% skim milk powder in PBS (blocking solution) for 16 h, at 48C. Membranes were then incubated in the presence of chicken anti-affinity purified gametocyte antigen antibody (1:1000; Wallach et al., 1990) or the mouse monoclonal anti-56 kDa antibody 1E11-11 (1:1000; (Wallach et al., 1990)) in blocking solution. As a control, serum taken from an uninfected chicken (#3246) was used. The membranes were left at room temperature with constant agitation for 60 min, prior to washing with three changes of 0.03% TWEEN-20/PBS over 30 min. The membranes were then probed with a rabbit anti-chicken IgG alkaline phosphatase conjugate (1:1000; Sigma) or a rabbit anti-mouse Immunoglobulin alkaline phosphatase conjugate (1:1000; Sigma). The membranes were washed and developed in the presence of 5-bromo-4chloro-3-indolyl-phosphate/nitro blue tetrazolium (SIGMA FAST BCIP/NBT, Sigma). 2.4. Mass determination Proteins were separated using Pharmacia’s Biotech SMART HPLC system interfaced with a Micromass Platform II single quadrapole electrospray mass spectrometer. Briefly, the affinity purified gametocyte antigen was resuspended in 0.1% trifluoroacetic acid in H2O before loading onto a Waters Symmetry C8 reverse phase column (2.1 £ 100 mm) which had been previously equilibrated in 0.1% trifluoroacetic acid. Proteins were eluted from the column using a gradient of 0–100% buffer B (80% acetonitrile, 0.08% trifluoroacetic acid) at 1%/min, sent into the electrospray source of the mass spectrometer and the mass determined as described previously (Greer and Morris, 1997). 2.5. TX-100 and TX-114 separations Gametocytes (10 5) were lysed in 0.5% TX-100 for 30 min on ice, followed by centrifugation (13 000 £ g, 10 min, 48C) to remove detergent insoluble material and nuclei. They were also lysed in the detergent TX-114 as described previously (Bordier, 1981). Briefly, gametocytes (10 5) were resuspended in 0.5% precondensed TX-114 in PBS containing 10 mM PMSF and 0.002% Bromophenol Blue (Sigma) for 30 min on ice. TX-114 insoluble material and nuclei were pelleted (13 000 £ g, 10 min, 48C). The cleared lysate was then loaded onto a sucrose gradient (6% sucrose, 0.06% TX-114 in PBS) and the tubes left at 378C for 3 min. The samples were then centrifuged (13 000 £ g, 3 min, room temperature). The top layer above the sucrose gradient was removed (aqueous phase), leaving an oily droplet at the bottom of the tube (detergent phase). The oily droplet at the

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

bottom of the tube was recovered, and the volume increased to its original volume with ice cold PBS. 2.6. Tissue section preparation and indirect immunofluorescence analysis Chickens were infected with 10 000 sporulated oocysts (E. maxima) orally. Chickens were killed by CO2 inhalation and cervical dislocation 138 h p.i. Intestines were removed, 5 cm either side of the diverticulum. They were washed with PBS and fixed in 3% paraformaldehyde and 0.05% gluteraldehyde in PBS. Fixed tissues were then sent to the Veterinary Laboratory, NSW Agriculture (Australia) for paraffin embedding and sectioning. Paraffin was removed from sections according to methods published elsewhere (Harlow and Lane, 1988) prior to blocking for 1 h in 0.1% BSA in PBS (BSA/PBS). Sections were then incubated in 1:500 mouse anti-affinity purified gametocyte antigen (#2113) in BSA/PBS for 1 h at room temperature in a humidified chamber. They were then washed for 15 min with three changes of PBS. Sections were then incubated in 1:50 goat antimouse Immunoglobulins F(ab 0 )2-FITC conjugate (Dako) in BSA/PBS, and the samples left for 1 h at room temperature. The slides were washed as described earlier, then incubated for 5 min in 0.003% Evans Blue Counterstain (Sigma). Sections were then rinsed in PBS. Tissues were mounted under coverslips with FluorSave Reagent (Calbiochem) prior to visualisation with an Olympus BX5 reflected fluorescence microscope with a PM-30 automatic photomicrographic system. Sections were photographed at £ 400 magnification using a narrow band 470–490 nm filter to detect the green fluorescence emitted by FITC, and a wide band 520–550 nm filter to detect the red fluorescence emitted by the Evans Blue Counterstain. 2.7. 2D SDS-PAGE Purified proteins (affinity purified gametocyte antigen) or proteins in crude gametocyte lysates were separated on the basis of charge in the first dimension and size in the second dimension as described previously (Walsh and Herbert, 1998). Briefly, gametocytes were washed twice by centrifugation (1100 £ g, 10 min, 48C) in 40 mM Tris pH 8 prior to solublisation. Washed gametocytes (0.5–1.0 £ 10 5) and the affinity purified gametocyte antigen (5 mg) were solublised in one of the following solutions (250–600 ml) depending on the experiment: (A) 0.5% CHAPS buffer consisting of 0.5% CHAPS, 8 M urea, 0.5% ampholytes (Ampholine pH 3.5– 10, Pharmacia), 40 mM Tris, 0.001% orange G; or (B) multiple surfactant buffer consisting of 2% CHAPS, 5 M urea, 2 M thiourea, 2% sulfobetaine 3–100, 0.5% ampholytes, 40 mM Tris, 0.001% orange G. Tributyl phosphine to 2 mM was added to each sample as well as 150 U endonuclease (Sigma), and the proteins left to solublise for 30 min at room temperature on a rotating wheel. Insoluble debris was pelleted by centrifugation at 13 000 £ g for 8 min. Immobilised pH gradient strips

807

(IPG; Amersham-Pharmacia) were rehydrated with the supernatant for ,4 h at room temperature. Isoelectric focusing was carried out using the Multiphor II system (Amersham-Pharmacia) under the following conditions: Phase 1– 300 V for 2.5 h; Phase 2–1000 V for 2.5 h; Phase 3–2500 V for 1 h; Phase 4–3500 V for 8 h, for a total of 33 750 V/h. Strips were then placed into equilibration buffer (6 M urea, 2% SDS, 375 mM Tris–HCl pH 8.8, 20% glycerol, 5 mM TBP, 2.5% acrylamide monomer) for 30 min. Equilibrated strips were then loaded onto precast 2D/Prep (Bio-Rad) polyacrylamide gels and the proteins separated by SDSPAGE as described earlier. The proteins were then visualised by silver staining, following the instructions outlined in the Amersham-Pharmacia Silver Staining Kit, or transferred to membrane for immunoblotting as described earlier. 2.8. O-glycosidase cleavage The affinity purified gametocyte antigen (20 mg) was treated with 5 mU neuraminidase (Sigma) in the presence of 0.1 M sodium acetate pH 5 for 16 h at 378C in a total volume of 100 ml. The neuraminidase treated protein sample was then neutralised with 40 ml 1 M NaHPO4 pH 7.2 (134 g Na2HPO4 7H2O and 4 ml 85% H3PO4 per liter) and heated to 1008C, 10 min. O-glycosidase (2.5 mU; Roche Molecular Biochemicals) was then added, and the samples left at 378C, 16 h. Multiple surfactant buffer (500 ml) was added, and the proteins solubilised prior to analysis by 2D SDS-PAGE as described earlier. Only half of the sample was analysed. 2.9. Carbohydrate analysis The affinity purified gametocyte antigen (100 mg) was separated on a 7.5% polyacrylamide gel, and transferred to polyvinylidene difluoride membrane as described earlier. The membrane was stained with Coomassie Blue (0.1% Commassie Blue R, 20% methanol, 0.5% glacial acetic acid) and destained (30% methanol). Bands corresponding to the 56 and 82 kDa proteins were excised and carbohydrate analysis performed using high-performance anion-exchange chromatography with pulsed amperometric detection (Townsend and Hardy, 1991) at the Australian Proteome Analysis Facility (Macquarie University, Sydney). 2.10. N-terminal sequencing Gametocytes (10 5) were lysed in multiple surfactant buffer and the proteins separated by 2D SDS-PAGE as described earlier. The proteins were separated in the first dimension on a pH gradient 4–7, and by size in the second dimension on a 7.5% polyacrylamide gel. Proteins were then transferred to polyvinylidene difluoride membrane and stained with Coomassie Blue as described above. The most intense spots corresponding to the isoforms of the 56 and 82 kDa proteins were identified and excised from the membranes. In order to obtain enough protein for N-term-

808

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

inal sequencing, eight separations were carried out, and the spots for each protein pooled. N-terminal sequences were determined by Edman degradation (Walker, 1997) at Biotech Australia Pty. Ltd. (Sydney).

3. Results 3.1. Characterisation of the 56 and 82 kDa gametocyte antigens by SDS-PAGE The 56 and 82 kDa gametocyte antigen enriched fraction (affinity purified gametocyte antigen), was analysed by SDS-PAGE in the presence (Fig. 1A, lane a) and absence (Fig. 1A, lane b) of the reducing agent b-mercaptoethanol, and stained with Coomassie Blue. Two predominant bands were seen with Mr 56 kDa and Mr 82 kDa (Fig. 1A, lanes a and b), consistent with the findings of Wallach et al. (1990). A number of smaller mol. wt. proteins were also seen. These proteins, along with the 56 and 82 kDa proteins were recognised by a chicken anti-affinity purified gametocyte antigen antibody by immunoblotting (Fig. 1B, lanes a and b). Control chicken serum did not recognise these bands (data not shown). In crude gametocyte extracts, the anti-affinity purified gametocyte antigen antibody recognised predominantly the two bands of 56 and 82 kDa (Fig. 1C, lanes a and c). Compared with the affinity purified gametocyte antigen samples (Fig. 1B, lanes a and b), fewer smaller mol. wt. proteins were recognised by the antibody in crude gametocyte extracts suggesting that these might represent breakdown products in the affinity purified gametocyte antigen samples. When the affinity purified gametocyte antigen was analysed in the absence of reducing agent, the presence of high mol. wt. disulphide bonded proteins was not detected by Coomassie Blue staining (Fig. 1A, lane b). However, the

proteins did migrate more slowly (Fig. 1A, lane b) under these conditions compared with those analysed under reducing conditions (Fig. 1A, lane a). When non-reduced affinity purified gametocyte antigen preparations were analysed by immunoblotting (Fig. 1B, lane b), a number of bands with Mr greater than 97 kDa were detected in these preparations, which were not observed in reduced samples (Fig 1B, lane a). In crude gametocyte preparations, the anti-affinity purified gametocyte antigen antiserum also recognised a band at ,160 kDa (Fig. 1C, lane c), not recognised by the control serum (Fig. 1C, lane d). A non-specific band was also observed at ,200 kDa (Fig. 1C, lanes c and d) in crude gametocyte lysates analysed under non-reducing conditions which was recognised by both the anti-affinity purified gametocyte antigen antiserum (Fig. 1C, lane c) and control chicken serum (Fig. 1C, lane d). 3.2. Physical properties of the 56 and 82 kDa gametocyte antigens The affinity purified gametocyte antigen preparation was separated by reverse phase-high pressure liquid chromatography, and the fractions containing the 56 and 82 kDa proteins used for mol. wt. determination by mass spectrometry. The true Mr for the oxidised forms of the proteins were determined as 52 450 Da for the 56 kDa protein and 64 236 Da for the 82 kDa protein. Eimeria maxima gametocytes were lysed in TX-114 to determine whether they were present as soluble proteins within the parasite or membrane bound. As a control for these experiments, gametocytes were also lysed in 0.5% TX-100. Gametocytes were lysed in the presence of 0.5% TX-100, and the detergent soluble fraction (Fig. 2, lane a) separated from the detergent insoluble/nuclear fraction (Fig. 2, lane b). Proteins were separated by SDS-PAGE and transferred to PVDF prior to probing with a chicken anti-affinity

Fig. 1. Analysis of the 56 and 82 kDa gametocyte antigens by SDS-PAGE. Affinity purified gametocyte antigen (5 mg; Panels A and B) or a crude E. maxima gametocyte lysate (4 £ 10 3 parasites; Panel C) were separated by SDS-PAGE on a 7.5% polyacrylamide gel in the presence (Panels A and B, lane a; Panel C, lanes a and b) or absence (Panels A and B, lane b; Panel C, lanes c and d) of 5% b-mercaptoethanol. The proteins were visualised by staining with Coomassie Brilliant Blue (Panels A). Alternatively, proteins were transferred to membrane and probed with chicken anti-affinity purified gametocyte antigen antiserum (Panel B, lanes a and b; Panel C, lanes a and c) or with control chicken serum (Panel C, lanes b and d). M ¼ molecular mass markers.

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

809

purified gametocyte antigen antibody. Both the 56 and 82 kDa proteins were present in the detergent soluble fraction (Fig. 2, lane a) consistent with the findings of Wallach et al. (1989). In addition, the proteins were also detected in the detergent insoluble fraction/nuclear pellet (Fig. 2, lane b). In the detergent TX-114, both the 56 and 82 kDa proteins separated to the TX-114 aqueous phase (Fig. 2, lane c) and to the detergent insoluble/nuclear pellet (Fig. 2, lane e). They were not detected in the TX-114 detergent phase (Fig. 2, lane d). 3.3. Immunolocalisation of the 56 and 82 kDa gametocyte antigens

Fig. 2. TX-100 and TX-114 fractionation of the 56 and 82 kDa E. maxima gametocyte antigens. Gametocytes were lysed either in 0.5% TX-100 (lanes a and b) or 0.5% TX-114 (lanes c–e) as described in Section 2. The TX-100 soluble phase (lane a), TX-100 insoluble proteins/nuclear pellet (lane b), TX-114 aqueous phase (lane c), TX-114 detergent phase (lane d) and TX114 nuclear pellet (lane e) were analysed by SDS-PAGE on a 7.5% polyacrylamide gel, followed by immunoblotting with chicken anti-affinity purified gametocyte antigen antibody.

A mouse anti-affinity purified gametocyte antigen antibody was used to localise the 56 and 82 kDa antigens in paraffin embedded sections of chicken intestines removed 138 h p.i. with E. maxima (Fig. 3). The antibody localised to the wall forming bodies in the macrogametocytes in infected tissue and not to a putative schizont (Panel A). The counterstain allowed the clear identification of macrogametocytes with their wall forming bodies, and schizonts (Panel B). No reactivity to the macrogametocytes was seen when a control mouse serum was used. When the two images were merged (Panels A and B), the co-localisation of the antibodies to the 56 and 82 kDa gametocyte antigens to the wall forming bodies was clear (Panel C).

Fig. 3. Localisation of the 56 and 82 kDa antigens in E. maxima gametocytes. Chickens were infected with E. maxima and sections of intestines prepared 138 h p.i. A mouse anti-affinity purified gametocyte antigen antibody followed by a secondary FITC-conjugated antibody was used to localise the immunofluorescence (Panel A). The sections were also stained with Evans Blue Counterstain (Panel B). Panel C represents the merged image of Panels A and B. Ggametocyte; and S-schizont.

810

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

3.4. Characterisation of the 56 and 82 kDa gametocyte antigens by 2D SDS-PAGE There are no published reports to date describing the solubilisation and analysis of crude gametocyte lysates or purified gametocyte antigens by 2D SDS-PAGE. Here, we have used two solubilisation procedures and 2D SDS-PAGE followed by silver staining or immunoblotting to gain further insights into the biophysical natures of the 56 and 82 kDa gametocyte antigens. Eimeria maxima gametocytes were lysed in the presence of 0.5% CHAPS and 8 M urea (solubilisation solution A, see Section 2) and analysed by 2D SDS-PAGE and immunoblotting (Fig. 4). A number of protein spots were recognised by the chicken anti-affinity purified gametocyte antigen antiserum when proteins were separated on a pH 3–11 gradient (Fig. 4A). However, the 56 and 82 kDa proteins (identified by the arrows) appeared to be the predominant proteins. They did not appear as discrete spots but as horizontal trails of spots, towards the acidic end of the pH gradient. The charge heterogeneity of these proteins was investigated further (Fig. 4B) by separating them in the first dimension on a narrower pH gradient (pH 4–7). Once again, the charge heterogeneity of these proteins was clearly seen for the 56 and 82 kDa proteins as horizontal trails of spots (indicated by arrows), and they appeared to be the predominant proteins in the gametocyte lysate recognised by the antiserum (Fig. 4B). Again, other protein spots were recognised by the antiserum on these membranes. Membranes probed with a control chicken serum did not recognise any spots (data not shown). A mAb to the 56 kDa protein was used to probe these blots in order to ascertain that the charged heterogeneous isoforms did indeed represent the same protein, and not other unrelated proteins. Once again, the charged heterogeneity of the 56 kDa protein was quite clear (Fig. 4C). In an attempt to exclude the possibility that the heterogeneity in charge observed for the 56 and 82 kDa gametocyte proteins was due to the poor solubilisation of proteins, a more rigorous extraction protocol was used. Gametocytes (Figs. 5A,B) or affinity purified gametocyte antigen (Figs. 5C,D) were solubilised in a multiple surfactant buffer containing 2% CHAPS, 5 M urea, 2 M thiourea and 2% sulfobetaine (solubilisation solution B, see Section 2), prior to separation by 2D SDS-PAGE and analysis by silver staining (Figs. 5A,C) or immunoblotting (Figs. 5B,D). In this buffer, the horizontal trail of spots representing charged isoforms was reduced in both gametocyte lysate preparations (Figs. 5A,B), and in affinity purified gametocyte antigen samples (Figs. 5C,D). However, under these solubilisation conditions, the 56 kDa gametocyte antigen appeared to be more prone to a small amount of aggregation, only detectable by immunoblotting (Figs. 5B,D). In these figures, a vertical line recognised by the anti-affinity purified gametocyte antigen antibody passes through a series of

spots at ,160 kDa which align with the pI of the spots representing the 56 kDa protein (Figs. 5B,D). Vertical background streaking was seen consistently when crude gametocyte lysates were separated by 2D SDS-PAGE and the proteins visualised by silver staining (Fig. 5A). This observation was not made when purified affinity purified gametocyte antigen preparations were analysed (Fig. 5C). The background streaking may reflect

Fig. 4. 2D SDS-PAGE analysis of E. maxima gametocytes. Eimeria maxima gametocytes (10 5) were separated by charge on a gradient of pH 3–11 (Panel A) or pH 4–7 (Panels B and C), then by size on a 10–20% polyacrylamide gel (Panel A) or a 7.5% polyacrylamide gel (Panels B and C). Proteins were then transferred to membrane and probed with chicken antiaffinity purified gametocyte antigen antiserum (Panels A and B) or mouse anti-56 kDa monoclonal antiserum (Panel C). Arrows point to the 82 kDa proteins (top horizontal smear) and the 56 kDa proteins (bottom horizontal smear).

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

811

Fig. 5. 2D SDS-PAGE analysis of charged isoforms of the 56 and 82 kDa proteins. Eimeria maxima gametocytes (5 £ 10 4; Panels A and B) and affinity purified gametocyte antigen (10 mg; Panels C and D) were solubilised in a multiple surfactant buffer prior to separation by charge on a gradient of pH 4–7 and by size on a 7.5% polyacrylamide gel. Proteins were either silver stained (Panels A and C), or transferred to membrane and probed with chicken anti-affinity purified gametocyte antigen antiserum (Panels B and D). Arrows point to the 82 kDa proteins (top horizontal smear) and the 56 kDa proteins (bottom horizontal smear).

the presence of salts and/or nucleic acids present in the crude gametocyte lysates (Rabilloud, 1996). Although the charged isoforms were clearly visualised by silver staining (Figs. 5A,C), the extent of the heterogeneity was more apparent on immunoblots (Figs. 5B,D), suggesting that they were present at low levels. 3.5. Analysis of the microheterogeneity of the 56 and 82 kDa gametocyte antigens Although the degree of charge heterogeneity for the 56 and 82 kDa gametocyte antigens was decreased when more rigorous extraction protocols were used, some isoforms were still present. This observation suggested that other post-translational modifications might be responsible for these isoforms (Dunbar, 1987). Since the proteins had been previously identified as glycoproteins (Wallach et al., 1995), the possibility that the sugar moieties contributed to the observed charge heterogeneity was explored. The affinity purified gametocyte antigen was treated with N-glycanase F, or with neuraminidase and O-glycosidase to assess for the presence of N- and O-linked glycans. There was no change in size observed for the 56 and 82 kDa proteins by SDS-PAGE after treatment with either enzyme (data not shown). When the effects of N-glycanase treatment of the proteins was assessed by 2D SDS-PAGE followed by immunoblotting with the anti-affinity purified gametocyte antigen antiserum, no difference in the degree of charge heterogeneity

was observed between treated and untreated samples (data not shown). When the proteins were cleaved with both neuraminidase and O-glycosidase, a clear reduction in the number of charged isoforms for the 56 and 82 kDa proteins was observed by 2D SDS-PAGE and immunoblotting (Fig. 6D) compared with untreated samples (Fig. 6A). Treatment of proteins with O-glycosidase alone (Fig. 6B) or neuraminidase alone (Fig. 6C) did not reduce the number of charged isoforms to a large extent. Treatment of the affinity purified gametocyte antigen with both neuraminidase alone (Fig. 6B), and the combination of both neuraminidase and Oglycosidase (Fig. 6D) did however, abolish the vertical streaking of the proteins, identified earlier as possible protein aggregates (see Figs. 5B,D). Treatment of the affinity purified gametocyte antigen with both neuramindase and O-glycosidase reduced the number of spots recognised by the anti-affinity purified gametocyte antigen antiserum (Fig. 6D). Despite this, the majority of the 56 and 82 kDa protein spots were still recognised by the anti-affinity purified gametocyte antigen antiserum. 3.6. Carbohydrate analysis of the 56 and 82 kDa proteins The carbohydrate moieties of the 56 and 82 gametocyte antigens were assessed through lectin binding studies and column chromatography. Immunoblotting with a number of lectins including Galanthus nivalis agglutinin, Sambucus nigra agglutinin, Maackia amurensis agglutinin, peanut agglutinin, Datura stramonium agglutinin and soy bean

812

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

Fig. 6. O-glycosidase treatment of the 56 and 82 kDa gametocyte antigens. Affinity purified gametocyte antigen (20 mg) was treated with 2.5 mU O-glycosidase (Panel B), or 5 mU neuraminidase (Panel C) or both neuraminidase and O-glycosidase (Panel D). Untreated samples are shown in Panel A. Proteins were separated by charge on a gradient of pH 4–7, then by size on a 7.5% polyacrylamide gel. They were then transferred to membrane and probed with chicken antiaffinity purified gametocyte antigen antiserum. Arrows point to the 82 kDa proteins (top horizontal smear) and the 56 kDa proteins (bottom horizontal smear).

agglutinin revealed that the proteins were only recognised by soy bean agglutinin (data not shown). Recognition of the 56 and 82 kDa gametocyte antigens by soy bean agglutinin was also observed previously by Wallach et al. (1995). When the proteins were characterised by column chromatography, glucose, fucose, galactose and mannose were detected (Table 1). Galactosamine was detected in both proteins indicative of the presence of O-linked glycans. Glucosamine was not detected.

3.7. N-terminus amino acid identification of the 56 and 82 kDa gametocyte antigens Two dimensional gel electrophoresis proved to be a suitable technique to separate the 56 and 82 kDa proteins from a gametocyte lysate for amino terminus identification. The most intensely stained spots corresponding to the 56 and 82 kDa proteins in the horizontal trail of proteins were excised from the membrane, and the proteins sequenced by Edman degradation. The N-termini were: Val-Pro-SerThr-Thr-Pro-Val-Glu-Asn-Gln-Val-His-Pro-Tyr-·-Glu-Met for the 56 kDa protein and, ·-Pro-Thr-Val-Leu-Asp-ThrThr-Thr-Gly-·-Gln-Val-Glu-Asp-Thr for the 82 kDa protein. The blanks (·) in the sequence indicate that there was some ambiguity in identifying the residue due to possible modifications.

4. Discussion The 56 and 82 kDa proteins were purified from gametocyte lysates by affinity chromatography as previously described (Wallach et al., 1990), and analysed by SDSPAGE. Two predominant proteins were clearly identified by Coomassie Blue staining after purification, as well as a number of smaller proteins. These smaller proteins are likely to represent break-down products of the 56 and 82 kDa antigens in the affinity purified gametocyte antigen preparation, since they were recognised by the anti-affinity purified gametocyte antigen antiserum, and were not Table 1 Carbohydrate analysis of the 56 and 82 kDa gametocyte antigens 56 kDa antigen

82 kDa antigen

pmol/mg

m/mg

pmol/mg

mg/mg

Neutral sugars Fucose Galactose Glucose Mannose

2.103 18.967 95.457 4.573

0.345 3.418 17.201 0.824

0.179 21.780 136.916 3.392

0.029 3.925 24.672 0.611

Amino sugars Galactosamine Glucosamine

43.197 0.000

9.313 0.000

32.528 0.000

7.013 0.000

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

observed when crude gametocyte lysates were probed with this antiserum. A mAb (1E11-11) to the 56 kDa protein bound to sepharose beads was used in the purification. The fact that both the 56 and 82 kDa proteins eluted from this antibody suggested that they either share the same epitope or interact with each other (Wallach et al., 1990). However, separation of gametocyte antigens by 1D SDSPAGE (data not shown) or by 2D SDS-PAGE followed by immunoblotting with the monoclonal 1E11-11 antibody (see Fig. 4, Panel C) only detected the 56 kDa antigen. These results suggested that the proteins do not share the same epitope but rather interact with each other. However, it cannot be ruled out that the presentation of epitopes on both antigens to the mAb on the affinity column is different to their presentation by SDS-PAGE. In previous reports, we also described the presence of a 230 kDa antigen in gametocytes (Wallach et al., 1989). In the present study, this high mol. wt. protein was not detected in gametocyte extracts. It has been reported that the 230 kDa gametocyte antigen is unstable and/or exists in small quantities in gametocytes (Fried et al., 1992), and this might explain why it remained undetectable by SDS-PAGE of gametocyte extracts. In addition, high mol. wt. proteins do not enter IPG strips easily, so it is not surprising that this protein was not detected here. The 56 and 82 kDa gametocyte antigens were analysed by electronspray ionisation mass spectrometry with maximum entropy processing (Green et al., 1996) to determine the masses of these charged glycoproteins. Both the 56 and 82 kDa proteins failed to migrate true to size by SDS-PAGE, in particular the 82 kDa protein whose true mass is 64 236 Da. This might be explained by the observation that proteins which possess an unusual amino acid composition for example a high proline content, or proteins that are glycosylated, can exhibit anomalously high Mr values by SDS-PAGE (Dunn, 1993). In the presence of the reducing agent b-mercaptoethanol, the proteins migrated further by SDS-PAGE compared with proteins analysed in the absence of reducing agent. In this case, the reducing agent had a greater effect on the migration of the 56 kDa protein than the 82 kDa protein on SDSPAGE gels. It might be expected that when disulphide bonds in a protein are broken by a reducing agent, that the protein adopts a more open configuration, therefore decreasing the distance migrated on SDS-PAGE gels. This was not the case for the 56 kDa protein, suggesting that the disruption of any intra-chain disulphide bonds affected the conformation of the protein in an alternative way such as to increase the distance travelled by SDS-PAGE. In the absence of reducing agent, no higher mol. wt., oligomeric forms of the 56 and 82 kDa proteins were observed by Commassie Blue staining indicating that these proteins did not form, within themselves, or each other, inter-chain disulphide bonds. However, immunoblotting of samples prepared in the absence of reducing agent, revealed the presence of a protein with Mr ,160 kDa, which

813

was also detected when gametocyte lysates were analysed (see Fig. 1). A protein of similar size was also seen by 2D SDS-PAGE only under reducing conditions (see Fig. 5). It is not certain whether the ,160 kDa protein observed by SDSPAGE and 2D SDS-PAGE are identical. However, it appears that the ,160 kDa protein observed by 2D SDSPAGE as a small number of spots and a vertical streak, represents an aggregated form, possibly of the 56 kDa protein, which is present in low abundance, and forms non-specifically under different solubilisation procedures. The ,200 kDa protein seen by SDS-PAGE (see Fig. 1, Panel C) was recognised by both the anti-affinity purified gametocyte antigen serum and control chicken serum under non-reducing conditions, indicating that it is a non-specific band which might represent an Fc receptor as reported previously (Pugatsch et al., 1989; Wallach et al., 1989). The localisation of the 56 and 82 kDa gametocyte antigens within the parasite was carried out biochemically and by immunofluorescence analysis. Gametocytes were lysed in TX-114 to determine whether the 56 and 82 kDa antigens were soluble proteins within the parasite, or membrane bound. Parasites were also lysed in TX-100 as a control for the experiment. In TX-100, some of the 56 and 82 kDa antigens were released into the detergent phase, although it appeared that the majority of the proteins were found in the detergent insoluble/nuclear pellet. This might indicate that more than one form of these proteins, with different degrees of detergent solubility, is present within the parasite. It cannot be excluded however, that the parasites have been incompletely lysed by this detergent. In TX114, both the 56 and 82 kDa gametocyte antigens partitioned to the TX-114 aqueous phase, as well as to the detergent insoluble pellet. The proteins were not detected in the detergent phase. Once again, these results supported the notion that some of the 56 and 82 kDa antigens exist as soluble proteins within the parasite, and that some might be detergent insoluble. Detergent lysis did not provide any evidence that the proteins were membrane bound. A mouse anti-affinity purified gametocyte antigen antibody was used to localise the 56 and 82 kDa antigens within the parasites. This antibody reacted exclusively with the wall forming bodies of macrogametocytes, suggesting a role for both proteins in oocyst wall formation. Two dimensional SDS-PAGE was used to further characterise the 56 and 82 kDa antigens. Both proteins were found to be acidic in nature, consisting of a number of charged isoforms. The number of charged isoforms were decreased when rigorous solubilisation procedures were used. When analysing the affinity purified gametocyte antigen by 2D SDS-PAGE, no background artefactual vertical streaking was observed. This streaking was more prominent when gametocyte lysates were separated by 2D SDS-PAGE, and is probably attributed to nucleic acids or salt within the parasite preparations (Rabilloud, 1996). Solubilisation conditions determined the degree of charged heterogeneity that was seen in the 56 and 82 kDa

814

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

gametocyte antigens. Although the number of isoforms was reduced under the rigorous solubilisation procedures used, some still remained. In addition, a number of these isoforms were present in small amounts, only detectable by immunoblotting. Both the 56 and 82 kDa gametocyte antigens had been previously identified as glycoproteins due to their reactivity with soy bean agglutinin (Wallach et al., 1995). Therefore an investigation was carried out to determine whether the charged isoforms of the 56 and 82 kDa proteins represented glycosylated modifications. The proteins were assessed for the presence of N- and Olinked glycans by enzymatic cleavage. N-linked glycans are bulky groups which, when removed, result in a reduction in mass of the core protein (Van den Steen et al., 1998). Enzymatic cleavage of the 56 and 82 kDa proteins did not reveal the presence of N-linked glycans. The proteins were then assessed for the presence of O-linked glycans. Unlike Nglycans, O-linked glycans are not bulky additions to proteins, therefore their removal will not alter the mass of the protein (Van den Steen et al., 1998). Thus the effects of removal of these carbohydrate moieties was assessed by 2D SDS-PAGE. In some cases, the removal of O-linked glycans by O-glycosidase (endo-a-N-acetyl galactosaminidase), an enzyme which specifically cleaves the unsubstituted disaccharide galactosea(1–3)N-acetylgalactosamine O-glycosidic linkage to serine or threonine (Van den Steen et al., 1998), is best achieved when terminal sialic acid residues are first removed with neuraminidase. Sialic acid residues interfere with O-glycosidase binding to the protein (Goldstein et al., 1978). Treatment of the proteins with neuraminidase alone, as well as a combination of neuraminidase and O-glycosidase reduced the number of aggregated forms of the 56 kDa proteins (see Fig. 6). However, neuraminidase alone did not appear to reduce the number of charged isoforms of the 56 and 82 kDa proteins as would have been expected following the removal of sialic acid residues from these proteins (see Fig. 6C). These results therefore might suggest that the number of sialic acid residues were too low to be detected by silver staining or immunoblotting, or that their contribution to the microheterogeneity of the proteins was minimal. In contrast, a marked change in the number of charged isoforms was seen after treatment with neuraminidase and O-glycosidase. These results were also unexpected given that cleavage of the uncharged disaccharide galactosea(1– 3)N-acetylgalactosamine should not affect the charge heterogeneity of the proteins. A possible explanation for these findings is that after the removal of terminal sialic acid residues from the proteins, O-glycosidase is free to specifically cleave the unsubstituted disaccharide galactosea(1–3)N-acetylgalactosamine moieties, giving rise to a conformational change within the proteins. The antibody species within the polyclonal anti-affinity purified gametocyte antigen serum used in these experiments, which recognise these conformational epitopes, no longer do so after the removal of the O-glycans. These results show that O-

glycans play a role in protein folding of the 56 and 82 kDa gametocyte antigens. However, their removal did not completely abolish antigen recognition. In light of these findings, the development of a recombinant protein, which maintains immunogenicity, may well be possible in a bacterial expression system. The remaining isoforms observed after O-glycosidase treatment might represent more complex O-linked carbohydrate moieties which remained uncleaved by this enzyme, or other post-translational modifications such as phosphorylation and sulphation (Dunbar, 1987). O-linked carbohydrates have been identified on both secreted and membrane bound proteins, and the most commonly studied is the mucin type, although other specific types also exist (reviewed by Van den Steen et al., 1998). They affect the secondary, tertiary and quaternary structures of proteins in a number of ways. In particular, they have been shown to confer stability to heat and proteases, as well as conformational rigidity to proteins. When O-linked glycans are removed from a protein, the protein converts from a linear to a globular state as seen in the case of the ovine submaxillary mucin (Rose et al., 1984). In the case of the 56 and 82 kDa antigens, it might be these moieties that influenced the conformation of the proteins after reduction in b-mercaptoethanol (see Fig. 1, and earlier discussion), or treatment with neuraminidase and O-glycosidase (see Fig. 6). It has also been shown that the ovine submaxillary mucin is prone to aggregation (Rose et al., 1984). Aggregation of the 56 kDa antigen in particular, was observed by 2D SDSPAGE (see Fig. 6), and was abolished when the proteins were treated with neuraminidase and O-glycosidase suggesting once again that O-glycans not only affect the tertiary structure of these gametocyte antigens, but also their quaternary structure. Cloning and sequencing of the genes encoding these proteins will provide insights into the O-glycosylation sites in these proteins, and their importance in determining function. Insights into the carbohydrate structures of proteins can be gained through lectin binding studies. Soy bean agglutinin for example, was the only lectin from the panel tested to bind to both the 56 and 82 kDa gametocyte antigens. Soy bean agglutinin has an affinity for the monosaccharide N-acetyl-d-galactosamine which is commonly O-linked to serine or threonine in proteins (Van den Steen et al., 1998). Galactosamine was detected by column chromatography in preparations of the 56 and 82 kDa gametocyte antigens, consistent with the observation of soy bean agglutinin binding to the proteins, and susceptibility to cleavage by O-glycosidase. Peanut agglutinin has an affinity for the disaccharide galactosea(1–3)N-acetylgalactosamine. This lectin did not bind to the proteins even after treatment with neuraminidase, suggesting that the disaccharide was either absent in these proteins, or the presence of more complex O-linked glycans containing fucose or galactose hindered the binding of peanut agglutinin to them (Gold-

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

stein et al., 1978). It is likely that the latter argument is true since the 56 and 82 kDa gametocyte antigens were susceptible to cleavage by O-glycosidase, and both fucose and galactose were detected by column chromatography. Datura stramonium agglutinin recognises disaccharides in complex N-glycan structures, and galactosea(1–4)N-acetylglucosamine in O-linkages. Since the presence of Nlinked glycans was not detected by N-glycanase F treatment of proteins, or by column chromatography, it is not surprising that Datura stramonium agglutinin did not recognise these proteins. In addition, these findings suggest that the particular complex carbohydrate structures associated with the O-glycosidic links recognised by Datura stramonium agglutinin are absent from the 56 and 82 kDa antigens. Galanthus nivalis has an affinity for N- or O-glycosidically linked mannose (Peumans and Van Damme, 1998). Although mannose was detected in both proteins, the levels may have been too low to enable Galanthus nivalis agglutinin binding, or the carbohydrate moieties may not have adopted the correct configuration for lectin recognition. Neuraminidase treatment of the 56 and 82 kDa proteins followed by 2D SDS-PAGE and immunoblotting suggested the presence of sialic acid residues, yet the lectins Sambucus nigra agglutinin and Maackia amurensis agglutinin, which recognise these moieties (Peumans and Van Damme, 1998) did not bind to the 56 and 82 kDa gametocyte antigens. Sambucus nigra agglutinin has an affinity for sialic acid terminally linked a(2–6) to galactose or N-acetylgalactosamine, and Maackia amurensis agglutinin has an affinity for sialic acid terminally linked a(2–3) to galactose. The fact that these two lectins did not bind to the proteins suggested that the level of sialic acid present in these proteins was too low to be detected by the assay. Glucose was the most abundant neutral sugar detected in both the 56 and 82 kDa antigens. This was an unusual observation given that glucose is not commonly found in glycoproteins. Glucose is often a contaminant in total monosaccharide analysis (Darbre, 1986), therefore the data presented here might represent an over estimation of the glucose content in these glycoproteins. Further investigation using an alternative strategy will be required to determine the true glucose content in these two proteins. Two dimensional SDS-PAGE enabled the successful purification of the 56 and 82 kDa antigens from a mixture of gametocyte antigens, which allowed for amino terminal sequence determination. The two proteins did not show any homology to each other at the N-terminus, or to any previously identified protein in the database. Once again, these observations suggest that the reason that both proteins elute from the same affinity column is that they interact with each other and not that they are derived from the same gene. A number of serine and threonine residues were observed on both proteins (one Ser and two Thr for the 56 kDa antigen; five Thr for the 82 kDa antigen), which might represent possible O-glycosylation sites.

815

The information presented here will now be used to isolate and clone the genes encoding the 56 and 82 kDa gametocyte antigens. The genes will then be expressed in a suitable system for the production of immunogenic recombinant forms of the proteins for incorporation in a maternal vaccine against coccidiosis. In addition, the function of both proteins in oocyst wall formation will be investigated. Acknowledgements This study was financed by the Commonwealth Government of Australia, Department of Education, Training and Youth Affairs, in partnership with ABIC Ltd, Israel, through an Australian Research Council SPIRT grant to Drs Nicholas Smith and Michael Wallach. We are grateful to David Witcombe (Molecular Parasitology Unit, Department of Cell and Molecular Biology, UTS) for assisting in the isolation and purification of gametocytes, and to Matthew Padula (Immunobiology Unit, Department of Cell and Molecular Biology, UTS) for his help with the mass spectrometry. References Bordier, C., 1981. Phase separation of integral membrane proteins in Triton X-114 solution. J. Biol. Chem. 256, 1604–7. Darbre, A., 1986. Analytical methods. In: Darbre, A. (Ed.). Practical Protein Chemistry-A Handbook, John Wiley and Sons Ltd, Chichester, pp. 230–335. Dunbar, B.S., 1987. Basic Principles of Posttranslational Modification of Proteins and their Analysis using High-Resolution Two-Dimensional Poylacrylamide Gel Electrophoresis, Two-Dimensional Electrophoresis and Immunological Techniques, Plenum Press, New York, pp. 77–102. Dunn, M.J., 1993. Gel electrophoresis: proteins. In: Graham, J.M., Billington, D. (Eds.). Bios Scientific Publishers Ltd, UK, pp. 57–58. Fried, M., Mencher, D., Sar-Shalom, O., Wallach, M., 1992. Developmental gene expression of a 230-kda macrogamete-specific protein of the avian coccidial parasite, Eimeria maxima. Mol. Biochem. Parasitol. 51, 251–62. Goldstein, I.J., Hayes, C.E., 1978. The lectins: carbohydrate-binding proteins of plants and animals. Adv. Carbohydr. Chem. Biochem. 35, 127–340. Green, A.N., Hutton, T., Vinogradov, S.N., 1996. Analysis of complex protein and glycoprotein mixtures by electrospray ionisation mass spectrometry with maximum entropy processing. In: Chapman, J.R. (Ed.). Protein and Peptide Analysis by Mass Spectrometry, Humana Press, New Jersey, pp. 279–94. Greer, F.M., Morris, H.R., 1997. Fast-atom bombardment and electrospray mass spectrometry of peptides and glycoproteins. In: Smith, B.J. (Ed.). Methods in Molecular Biology. Protein Sequencing Protocols, Humana Press, New Jersey, pp. 147–63. Harlow, E., Lane, D., 1988. Immunoblotting, Antibodies: a Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, pp. 490–2. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–5. Peumans, W.J., Van Damme, E.J., 1998. Plant lectins: specific tools for the identification, isolation, and characterisation of O-linked glycans. Crit. Rev. Biochem. Mol. Biol. 33, 209–58. Pugatsch, T., Mencher, D., Wallach, M., 1989. Eimeria maxima: isolation of gametocytes and their immunogenicity in mice, rabbits, and chickens. Exp. Parasitol. 68, 127–34.

816

S.I. Belli et al. / International Journal for Parasitology 32 (2002) 805–816

Rabilloud, T., 1996. Solubilisation of proteins for electrophoretic analyses. Electrophoresis 17, 813–29. Raether, W., 1988. Chemotherapy and control measures of parasitic diseases in domestic animals and man. In: Melhorn, H. (Ed.). Parasitology in Focus. Facts and Trends, Springer-Verlag, Heidelberg, pp. 740– 55. Rose, M.C., Voter, W.A., Sage, H., Brown, C.F., Kaufman, B., 1984. Effects of deglycosylation on the architecture of ovine submaxillary mucin glycoprotein. J. Biol. Chem. 259, 3167–72. Shirley, M.W., 1995. Eimeria and Isospora: Eimeria species and strains of chickens. In: Eckert, J., Braun, R., Shirley, M.W., Coudert, P. (Eds.). COST 89/820 Biotechnology. Guidelines on Techniques in Coccidiosis Research, 89/820. European Commission, Luxembourg, pp. 1–11. Townsend, R.R., Hardy, M.R., 1991. Analysis of glycoprotein oligosaccharides using high-pH anion exchange chromatography. Glycobiology 1, 139–47. Van den Steen, P., Rudd, P.M., Dwek, R.A., Opdenakker, G., 1998. Concepts and principles of O-linked glycosylation. Crit. Rev. Biochem. Mol. Biol. 33, 151–208.

Walker, J.M., 1997. The dansyl-Edman method for manual peptide sequencing. In: Smith, B.J. (Ed.). Methods in Molecular Biology, Humana Press Inc, New Jersey, pp. 183–7. Wallach, M.G., Mencher, D., Yarus, S., Pillemer, G., Halabi, A., Pugatsch, T., 1989. Eimeria maxima: identification of gametocyte protein antigens. Exp. Parasitol. 68, 49–56. Wallach, M., Pillemer, G., Yarus, S., Halabi, A., Pugatsch, T., Mencher, D., 1990. Passive immunisation of chickens against Eimeria maxima infection with a monoclonal antibody developed against a gametocyte antigen. Infect. Immun. 58, 557–62. Wallach, M., Halabi, A., Pillemer, G., Sar-Shalom, O., Mencher, D., Gilad, M., Bendheim, U., Danforth, H.D., Augustine, P.C., 1992. Maternal immunisation with gametocyte antigens as a means of providing protective immunity against Eimeria maxima in chickens. Infect. Immun. 60, 2036–9. Wallach, M., Smith, N.C., Petracca, M., Miller, C.M., Eckert, J., Braun, R., 1995. Eimeria maxima gametocyte antigens: potential use in a subunit maternal vaccine against coccidiosis in chickens. Vaccine 13, 347–54. Walsh, B.J., Herbert, B., 1998. Setting up two-dimensional gel electrophoresis for proteome projects. ABRFnews 9, 11–21.