Heligmosomoides polygyrus elicits a dominant non-protective antibody response directed against restricted glycan and peptide epitopes

Summary

Heligmosomoides polygyrus is a widely-used gastrointestinal helminth model of long-term chronic infection in mice, which has not been well-characterized at the antigenic level. We now identify the major targets of the murine primary antibody response as a subset of the secreted products in H. polygyrus Excretory-Secretory (HES) antigen. An immunodominant epitope is an O-linked glycan (named Glycan A) carried on 3 highly-expressed HES glycoproteins (VAL-1, -2 and -5), which stimulates only IgM antibodies, is exposed on the adult worm surface, and is poorly represented in somatic parasite extracts. A second carbohydrate epitope (Glycan B), present on both a non-protein high molecular weight component (HM) and a 65-kDa molecule, is widely distributed in adult somatic tissues; while HM-65 molecules bear phosphorylcholine (PC), the Glycan B epitope itself is not PC. Class-switched IgG1 antibodies are found to Glycan B, but the dominant primary IgG1 response is to the polypeptides of VAL proteins, including also VAL-3 and VAL-4. Secondary antibody responses are similar while recognizing in addition VAL-7. Whilst vaccination with HES conferred complete protection against challenge H. polygyrus infection, monoclonal antibodies raised against each of the glycan epitopes, and against VAL-1, -2 and -4 proteins were unable to do so, even though these specificities (with the exception of VAL-2) are also secreted by tissue-phase L4 larvae. The primary immune response in susceptible mice is, therefore, dominated by nonprotective antibodies against a small subset of antigenic epitopes, raising the possibility that these act as decoy specificities that generate ineffective humoral immunity.

Introduction

Heligmosomoides polygyrus is a widely-used experimental mouse model for the highly prevalent human and animal gastrointestinal helminth infections (1, 2). This system has provided major new findings in parasite immunology (3, 4), immune regulation (5, 6), nutrition (7) and ecology (8), and yet little information is available on the specific parasite antigens to which the host immune system is exposed. In this study, we set out to identify the molecular targets of murine humoral antibodies, to define individual H. polygyrus antigens, and to investigate the role of major antibody specificities in the host-parasite relationship.

Among the interesting facets of H. polygyrus is its ability to establish a chronic infection in most strains of laboratory mice, with the genetic background influencing the rate of expulsion rather than susceptibility per se (9-11). Genetically resistant mice mount a more rapid serum antibody response measured against adult worm somatic extract (12) or excretory-secretory (ES) antigens from cultured adult parasites (13, 14), and immunity to reinfection is compromised in B cell-deficient mice (4, 15-17). Early investigations had reported that passive transfer of serum from infected mice can confer a degree of immunity to H. polygyrus both in terms of worm number and fecundity (18); this effect was associated with IgG1 isotype antibodies (19, 20). More recently, IgG1 serum antibodies have been demonstrated to reduce the fecundity and viability of adult worms, and shown to require affinity maturation to confer any resistant effect (15).

As has been recently pointed out (21), in current nematode model systems, few serologically important antigens have yet been identified. Previous studies have relied either on crude whole-worm homogenates, or collected secreted products as a more restricted but nevertheless complex antigenic set. We therefore decided to analyze the humoral antibody response to H. polygyrus in terms of specific immunoglobulins, to define the molecular targets of parasite-specific antibodies, and to test whether these played any protective role against the infection in vivo. We tested antibody reactivities both to crude parasite antigenic extracts, and also to preparations collected from in vitro culture of adult worms, termed “excretory-secretory” (ES) antigens, which are strongly implicated in immunomodulation of the host (6, 22). We report here that several major constituents are homologues of Venom allergen-Ancylostoma secreted protein-Like (VAL) antigens related to the vaccine candidates of human and canine hookworms (23, 24). However, the response to infection is dominated by anti-glycan specificities, and the murine antibody profile is highly restricted with respect to the range of antigens recognized.

Materials and Methods

Parasites, antigens and mice

The original stock of H. polygyrus bakeri used in these studies was kindly supplied to us by Professor J M Behnke, University of Nottingham, UK. Parasites, H. polygyrus Excretory-Secretory (HES) antigen and adult worm somatic extract (HEx) were produced as previously described (6, 25, 26). Day 5 fourth-stage larvae were collected from the intestinal wall of infected mice and ES collected over a 3-day culture period in the same manner as adult HES. Female C57BL/6 and BALB/c mice (6-10 weeks old) were bred in-house, and animal studies were performed under UK Home Office Licence. Mice were infected with 200 H. polygyrus L3 by oral gavage, and fecal egg counts and adult worm burdens determined by standard procedures (2). For secondary infection, mice were treated orally with pyrantel embonate (27) in the form of 2.5 mg Strongid P paste in 0.2 ml water on days 28 and 29 post-primary infection. Drug-treated mice were rechallenged with 200 L3 by gavage two weeks later. Where indicated, HES was heat denatured by incubating at 95°C for 20 minutes (6).

1D and 2D gel electrophoresis and Western blotting

HES and HEx (1-10 μg) were separated, silver stained or blotted as previously described (28). Blots were blocked in 2% BSA-TBS with 0.05% Tween 20 (TBST) for 2 hours at room temperature, before being probed with sera (1/500 dilution) or monoclonal antibodies (2 μg/ml) at 4°C overnight. Following extensive washing in TBST, blots were incubated with HRP-conjugated secondary antibodies (anti-mouse Ig 1/2000, Dako P0460; anti-mouse IgM 1/1000, Southern Biotech 1020-05; anti-mouse IgG1, 1/2000 Southern Biotech 1070-05) for 1 hour at 37°C, washed in TBST, then developed as previously described (28). Alternatively, IgA blots were incubated with biotinylated anti-mouse IgA (1/500, Invitrogen M31115) followed by HRP-streptavidin (1/2000, Sigma), and developed as above. Mouse IgM monoclonal antibody Bp-1 (29) was used for anti-PC blots at 1/1000 dilution and detected with anti-mouse Ig as above.

ELISA

HES and HEx (1 μg/ml) were coated on Immunoplates (Nunc) in 0.06 M carbonate buffer overnight (4°C), blocked with block solution (2% BSA-TBST) for 2 hours (37°C), then incubated with doubling dilutions of sera (in block solution) for 2 hours (37°C). For comparison of HES and L4 ES, each were used to coat plates at a range of dilutions. Worm-specific antibody titers were detected using the secondary reagents described above and developed with ABTS (Insight Biotech). Titer was determined as the reciprocal dilution at which the sample dropped below background levels. For anti-PC ELISA, plates were coated with 1 μg/ml PC-conjugated BSA (30), serum used at a 1/500 dilution, and mAb at 5 μg/ml. Anti-H. polygyrus monoclonal Abs were used at 5 μg/ml for all ELISAs, unless stated. Goat anti-rat Ig (1/2000, Dako P0450) was used as a secondary for experiments with rat sera.

Monoclonal Ab and polyclonal Ab production

For mAbs, spleens and mesenteric lymph nodes were recovered from C57BL/6 mice at day 28 post-infection, and fused with SP2 cells. Fused cells were cultured for 12-14 days in HAT selection media (RPMI 1640 supplemented with 20% FCS (Hyclone), 100 U/ml penicillin, 100 μg/ml streptomycin, 2 mM L-glutamine (all Gibco), HAT (100 μM hypoxanthine, 0.4 μM aminopterin, 16 μM thymidine) and OPI (1 mM oxaloacetate, 0.45 mM pyruvate, 0.2 U/ml insulin; all Sigma)). Plates were ELISA screened for the production of anti-HES antibodies as above, and positive wells cloned by 2-3 rounds of limiting dilution. Cells were then adapted into standard complete RPMI media (RPMI 1640 supplemented with 10% FCS and the above concentrations of penicillin, streptomycin and L-glutamine), and grown in bulk cultures in Vectra Cell Bioreactors (Bio-Vectra) to produce mAb. Antibody isotypes were determined either with a mouse antibody isotype kit (Isostrip, Roche) or the anti-mouse Ig secondary Ab described above. Antibodies were purified using an AKTA prime FPLC with a HiTrap protein G HP (IgG1 mAb) or HiTrap IgM purification (IgM and IgA) column, according to the manufacturer’s instructions, and then dialyzed extensively into PBS. For the rat polyclonal antibodies, rats were immunized with 25 μg HES or recombinant H. polygyrus calreticulin (AM296015, (31) produced in E. coli) in alum adjuvant i.p., then boosted with 10 μg antigen on days 28 and 35, before serum collection on day 42. The mouse IgG1 myeloma MOPC 31C from ATCC was used as an isotype control. A hybridoma producing mouse anti-DNP IgM control was produced as above from mice immunized with 50 μg DNP-conjugated KLH in alum adjuvant i.p., then boosted on days 12 and 13 with 1 μg DNP-KLH in PBS i.v. before spleen harvest on day 14. Screening was performed using DNP-OVA and anti-mouse IgM secondary antibody as above. Hybridomas were also produced from mice immunized with 25 μg HES in incomplete Freund’s adjuvant (IFA, Sigma) i.p. then boosted on days 48 and 49 with 1 μg HES in PBS i.v. before spleen harvest and fusion on day 50.

Immunoprecipitation

HES was labeled with biotin (approx 40 μg biotin reagent per 100 μg HES) using EZ-link Sulfo-NHS Biotinylation kit (Pierce) for 2 hours on ice, then dialyzed overnight into PBS. Biotinylated HES was then pre-cleared with Protein G agarose beads (Millipore, 16-266) in the presence of MOPC 31C IgG1 isotype control for 30 minutes at room temperature. Unbound HES (2 μg) was then incubated with 2 μg of various anti-HES IgG1 mAb, MOPC IgG1 control, or 5 μl C57BL/6 day 28 primary or d14 secondary infection sera, in non-denaturing IP buffer (20 mM Tris pH 8, 150 mM NaCl, 1 mM EDTA, 10% glycerol, 1% Triton-X100) for 2 hours, then with Protein G agarose beads overnight, at 4°C with rotation. Beads were then washed 5 × 5 minutes in IP buffer, and bound proteins eluted with 0.1 M glycine pH 2.7. Eluted proteins were buffer exchanged into PBS (with MicroBio-Spin 6 chromatography columns, Bio-Rad), run on 1-D and 2-D gels and Western blotted as described above, then probed with 1/2000 streptavidin HPO (Sigma) before developing to allow visualization of biotinylated proteins.

Deglycosylation

For PNGase F treatment, HES or ribonuclease B (Sigma) was denatured by heating to 95°C for 5 minutes in the presence of 0.1% SDS and 100 mM 2-mercaptoethanol, before the addition of 10 U PNGase F (Sigma) and 1% Triton-X100. Samples were incubated at 37°C for 3 hours. For PNGase A treatment, HES was diluted in 100 mM ammonium bicarbonate pH 8.5, heat-denatured, and digested overnight with 1 μg of proteomics grade trypsin at 37°C (Sigma), after which trypsin was inactivated by boiling. Complete digestion of HES was confirmed by SDS-PAGE analysis (data not shown). Tryptic peptides were then lyophilized and resuspended in 100 mM sodium acetate pH 5.0 and treated with 0.2 mU PNGase A (Roche) for 24 hours at 37°C. For subsequent ELISA analysis, peptides were bound to plates in carbonate buffer at 1 μg/ml and then tested for antibody binding as above. For PNGase F and A digestions, control proteins were treated in an identical manner but in the absence of enzyme. For chemical deglycosylation of HES with TFMS (trifluromethanesulfonic acid), HES or ribonuclease B were dialyzed into 0.1 M ammonium acetate then lyophilized until completely dry. The pellet was resuspended in 10% anisole in TFMS (both Sigma) at 4°C, and the reaction was allowed to proceed for 2 hours, before neutralization with a 60% pyridine solution in a methanol-dry ice bath. The soluble fraction of HES was dialyzed into PBS, and the residual precipitate solubilized in 1% SDS.

Affinity purification of VAL proteins

Twenty mgs each of anti-VAL-1 (4-M15), 2 (4-S4) and 4 (2-11) mAbs were dialyzed into coupling buffer (0.1 M sodium bicarbonate, 0.5 M NaCl pH 8.4) then reacted with swollen cyanogen bromide-activated sepharose beads (Sigma) overnight at 4°C with rotation. Unreacted groups were blocked with 0.2 M glycine pH 8.0 for 2 hours at room temperature, after which the beads were washed in 5 cycles of coupling buffer followed by 0.1 M sodium acetate, 0.5 M NaCl pH 4.0. For affinity purification, 10 μg of HES was treated as before for IP, eluted proteins run on 1-D SDS-PAGE and bands of interest excised for mass spectrometry analysis.

Sequence database and mass spectrometry

In studies to be published elsewhere, a database compiled of ~466,000 Roche 454 sequence reads from normalized and non-normalized adult H polygyrus cDNA was assembled into ~20,000 predicted gene products³. This database was used to match peptides identified by mass spectrometry of SDS-PAGE and 2D gel purified proteins. The individual protein genes described here (VAL-1, -2, -3, -4 and -5) were each amplified by PCR from H. polygyrus mRNA using gene-specific primers, and multiple independent clones were sequenced to verify the sequences predicted by the assembly algorithm. The full sequences for VAL-1 to -5 have been deposited with NCBI (http://www.ncbi.nlm.nih.gov/) under accession numbers JF914902, JF914906, JF914909, JF91410 and JF914911 respectively. Immunoprecipitated HES proteins were prepared for mass spectrometry analysis as previously described (28). Positive-ion MALDI mass spectra were obtained using a Bruker ultraflex III in reflectron mode, equipped with a Nd:YAG smart beam laser. MS spectra were acquired over a mass range of m/z 800-4000, and monoisotopic masses were obtained using a SNAP averaging algorithm. The ten strongest peaks of interest, with a S/N greater than 30, were selected for MS/MS fragmentation in LIFT mode. Bruker flexAnalysis software (version 3.3) was used to perform the spectral processing and peak list generation for both the MS and MS/MS spectra. Identification of mAb 2-11 target VAL-4 required LC-ESI-MS/MS using an Ultimate nano-LC system (Dionex) equipped with a PepMap C₁₈ trap (300 μm × 0.5 cm, Dionex) and an Onyx C₁₈ monolithic silica capillary column (100 μm × 15 cm, Phenomenex). Peptides were eluted over with a acetonitrile gradient (Solvent A = 2% (v/v) acetonitrile, 0.1% (v/v) formic acid in H₂O; Solvent B = acetonitrile, 0.1% (v/v) formic acid; Gradient conditions consisted of 3 min solvent A, and then a linear 0-50% gradient of Solvent B over 20 min was applied, followed by a 5 min wash at 95% Solvent B. The flow rate was 1.2 μL/min and the column temperature was 60°C. The nano-LC was interfaced with an HCT ultra ETD II ion-trap LC-MS/MS system (Bruker Daltonics) with an online nanoESI source. Positive ESI-MS & MS/MS spectra were acquired using AutoMSn mode, over the 300 – 1800 m/z range. Instrument control, data acquisition and processing were performed using Compass 1.3 SP1 software (Esquire control, Hystar and DataAnalysis, Bruker Daltonics). Tandem mass spectra data were submitted to database searching against an in-house database containing the VAL sequences (276 sequences; 118058 residues), using the Mascot program (Matrix Science Ltd., version 2.1), through the Bruker ProteinScape interface (version 2.1). Database searching was run with a peptide tolerance of 250 ppm and MS/MS tolerance of 0.5 Da. Peptide matches with expect values less than 0.05 at a Mowse significance threshold of p<0.05 were considered significant

Immunofluorescence

For sections, adult H. polygyrus worms were snap-frozen on dry ice in Cryo-M-Bed mountant (Bright Instruments). Cryostat sections (5 μm; Leica) were cut onto Polysine™ slides (VWR), dried and then fixed in 100% acetone for 10 min. Sections were washed twice with PBS for 10 min, and then incubated with the various mAb (50 μg/ml in PBS containing 1% FCS) for 2 hours at room temperature, washed twice in PBS as before, and then incubated with secondary anti-mouse Ig TRITC (1/100 in PBS) for 1 hour at room temperature. Following extensive washing, sections were mounted in anti-fade Vectashield mountant (Vector Labs). Staining was visualised with a Olympus fluorescent microscope. Non-fixed intact worms were stained on ice in round bottom 96 well plates, and then treated as above.

Radiolabeling of adult worm surface

Adult H. polygyrus were surface radiolabeled essentially as described in earlier publications (32, 33). Eppendorf tubes (1.5ml) were coated with 200 μl of a 1 mg/ml solution of Iodination reagent (Pierce) in chloroform. Once dried, the tubes were washed several times with PBS, before transfer of approximately 500 adult worms and 500 μCi ¹²⁵Iodine (Perkin Elmer) on ice. The sample was incubated with frequent agitation for 10 minutes, and then quenched by the addition of a saturated solution of L-tyrosine (Sigma). Radio-labeled parasite surface material was then produced as described above as with HEx, except parasites were homogenized in 1.5 % nOG detergent and 1% protease inhibitor cocktail (Sigma P8340). Immunoprecipitates were performed as above (anti-HM-65 mAb 9.1.3 or rat anti-HES pAb) or with anti-mouse IgM agarose (Sigma A4540). Autoradiographs were carried out on dried gels as previously described (32, 33).

Vaccination and Passive immunization

C57BL/6 females were immunized with 25 μg of HES in alum adjuvant i.p., then boosted on days 28 and 35 with 5 μg of HES-alum i.p. Mice were challenged with 200 H. polygyrus L3 larvae, and fecal egg counts determined at days 14 and 28 post-infection, and adult worms counted at day 28. For passive immunization, C57BL/6 females were treated on day -1 and then every 2-3 days post-infection (with 200 H. polygyrus L3 larvae) with either 0.2 or 1.0 mg of mAb i.p. (for IgG1 mAbs) or i.v. (for IgM mAbs) as detailed in the Figure Legend. Eggs and worm numbers were determined as above.

Results

Antibody responses are predominantly directed at secreted, rather than somatic, parasite antigens

As a first step in defining immunogenic products of adult H. polygyrus, we compared a conventional antigenic preparation comprising a soluble whole parasite extract (HEx), with products released by live parasites maintained in serum-free tissue culture medium (H. polygyrus excretory-secretory antigen, HES). The overall protein compositions of HES and HEx are very different, as shown in Fig. 1 A and B respectively, and confirmed by recent proteomic analysis (34).

Figure 1. The predominant serological antigens of H. polygyrus are in HES rather than somatic extract.

A, B 2-D silver stained gels of 10 μg of H. polygyrus excretory-secretory products (HES, A) and somatic extract (HEx, B). Molecular weight markers (in kDa) are indicated on the left.

C, D Anti-HES (open symbols) and HEx (solid symbols) IgG1 serum antibody titres from BALB/c (C) and C57BL/6 (D) mice at various time points following primary infection.

E, F Anti-HES and HEx IgM serum antibody titres in the same BALB/c (E) and C57BL/6 (F) mice.

Each point represents the mean and SEM of data from 5 individual mice separately assayed. Data are representative of two independent experiments

As IgG1 antibodies predominate, and have a protective role in both primary and secondary H. polygyrus infection (15, 19, 20), we then compared IgG1 responses to HES and HEx following a primary infection. Antigen-specific IgG1 responses to HES were detected by ELISA in both BALB/c and C57BL/6 mice by day 14 post-infection, and by day 28 reached a high titer which was maintained for at least 63 (BALB/c, Fig. 1 C) and 100 days (C57BL/6, Fig. 1 D). Moreover, IgG1 titers to HES were up to 20-fold higher than those to HEx, and showed less variation between individual animals. We therefore focussed in the majority of subsequent antibody investigations on HES.

The anti-parasite IgM response differed from the IgG1 response in several regards (Fig. 1 E and F). Significant background levels of both anti-HES and anti-HEx IgM antibodies were noted in naïve mice, and following infection, specific IgM titers reached a relatively early plateau (day 14). Furthermore, reactivity to HES and HEx was equivalent at all time points. We were unable to detect the target of (presumably natural) IgM antibodies present in naïve mice using both Western blot and monoclonal antibody techniques, perhaps due to their relatively weak affinity (data not shown).

Antigen specificity of polyclonal antibody responses

To identify individual antigenic targets of the antibody responses, we adopted both polyclonal and monoclonal strategies. First, we used 2D SDS-PAGE separation of HES antigens, by which ~100 distinct protein spots are observed (Fig. 1 A and 2 A); the identity of most major proteins has been determined by mass spectrometry (34). Despite the abundance of potential antigens, however, the 2D Western blot profile of polyclonal sera from C57BL/6 mice with 28-day primary infection is much more restricted (Fig. 2 B). These were identified by proteomics as members of the Venom allergen/Ancylostoma secreted protein-Like (VAL) family of proteins (35, 36), specifically VAL-1, VAL-2 and VAL-5 (Fig. 2 A and (34)). Additional reactivity was noted to both a high molecular weight component and smaller 65-kDa antigen which do not appreciably silver stain and did not give measureable peptides for proteomic analysis; we term this antigen combination HM-65 (see below).

Figure 2. Polyclonal antibodies recognise a restricted set of H. polygyrus antigens by Western Blot.

A. Silver-stained 2-D gel of HES indicating major components identified by mass spectrometry (34). HMAg and Ag65 indicate the position of antigens (subsequently termed HM-65, see Figure 3 D) that do not stain with silver.

B, C. 2-D Western blot of HES with sera from C57BL/6 mice taken at days 28 (B) and 100 (C) following primary infection with H. polygyrus, developed with a polyvalent anti-Ig conjugate.

D, E. As above, but with HEx.

F. 2-D Western blot of anti-HES IgM in sera from 28-day infected C57BL/6 mice.

G. As above, for anti-HES IgG1

H. As above, for anti-HES IgA

I. 2-D Western blot of anti-HES IgG1 in sera C57BL/6 mice collected 14 days following a secondary challenge infection.

For each assay, sera were pooled from 5 mice. Molecular weight markers are indicated in kDa, and results are representative of two or more experiments. Naïve mouse sera showed no positive binding under the same conditions (data not shown).

As reactivity increases over a long time frame, and as C57BL/6 mouse is considered to be a slow responder to infection (37, 38), we also examined the serological response at 100 days post-infection (Fig. 2 C); at this time point, responses to all the antigens previously noted were substantially stronger, but there remained a restricted repertoire of target antigens.

Consistent with ELISA results indicating that anti-HEx somatic extract responses were relatively weak, the same sera showed only slight reactivity by Western blot (Fig. 2 D, E). The primary anti-VAL antibody response to HES was predominantly of the IgM isotype, although a smaller amount of anti-VAL-2 IgA was also noted (Fig. 2 F and H). In contrast, anti-HES IgG1 response was detected against HM-65 (Fig. 2 G). Polyclonal IgE did not provide a measurable signal by Western blot, likely due to the low levels of this isotype following infection (data not shown).

We also examined the secondary response mounted by genetically susceptible mice cleared of infection by chemotherapy; in these mice, protective immunity is stimulated against challenge infection (9, 39). Importantly, immunity to challenge infection is ablated in B-cell or antibody-deficient mice (4, 15, 16, 40). By Western blot analysis, secondary IgG1 antibodies showed a similar profile to samples from primary infection, albeit with a 10-30-fold higher titer (data not shown) and correspondingly stronger binding patterns (Fig 2 I). In contrast, secondary IgM responses were similar to primary antibodies with respect both the titer and specificity profile (data not shown).

Monoclonal antibody specificities

We next generated a panel of monoclonal antibodies to dissect the antigenic response in fine detail, using spleens and draining MLN from infected mice at day 28. Despite taking cells at this relatively late time point in the primary response, a substantial proportion of monoclonals represented antibodies which had not undergone class switch from the IgM isotype, particularly when splenocytes rather than MLNC were used (Fig. 3 A). The great majority (12/14) of these IgM mAbs displayed a similar Western blot profile against HES (Fig. 3 C) and bound to proteins previously identified as VAL-1, VAL-2 and VAL-5 by proteomic analysis. In fact, this indicates that the observed specificity of polyclonal serum (Fig. 2 B, C) can be largely replicated by a single (monoclonal) antibody specificity, and that an immunodominant epitope is shared by at least three members of the VAL family.

Figure 3. HES-specific monoclonal antibodies from mice infected with H. polygyrus.

A, B. Frequency of HES-specific hybridoma cells producing IgM (A) or IgG1 (B) from the spleens and mesenteric lymph node (MLNs) of infected C57BL/6 mice at day 28 post-infection, as determined by ELISA. Each dot represents a single well of a 96 well plate.

C. Representative 2-D Western blots of 3 different IgM anti-HES mAbs derived from infected mice which bind to Glycan A.

D. Representative 2-D Western blots of anti-HES IgM, IgG1 and IgA mAbs recognizing Glycan B on the HM-65 antigens.

E. Anti-phosphorylcholine (PC) antibodies measured by ELISA against PC-BSA in naïve and 28-day infected sera from BALB/c and C57BL/6 mice.

F. 2-D Western blot of HES probed with anti-PC mAb Bp-1. The positions of the HM-65 antigens are circled.

G. Reactivity to PC measured by ELISA on BSA-PC-coated plates; data from a panel of anti-HES mAbs are shown, together with Bp-1 positive control.

Molecular weight markers are indicated in kDa.

A different pattern was shared by a smaller number of antibodies, including IgM, IgG1 and IgA isotypes, which bound to HM-65 (Fig. 3 D). Notably, these components do not stain with silver and do not give measurable peptides for proteomic analysis, mimicking the pattern seen with polyclonal IgG1 from primary infections (Fig. 2 G).

Because of the prominence of phosphorylcholine (PC) in many helminth products, including high molecular weight species with non-proteinaceous composition (30, 41), we tested for PC reactivity in infection sera and for PC epitopes in HES. Serum antibodies from H. polygyrus-infected mice showed a modest degree of anti-PC binding (Fig. 3 E), as has previously been reported (42). When 2D Western blots were probed with monoclonal anti-PC antibody, the major positive reaction was with HM-65 (Fig. 3 F). Despite this, none of the anti-HES monoclonals bound directly to PC (Fig. 3 G) indicating that they target a non-PC specificity.

IgG antibodies recognise heat-labile epitopes of secreted VAL antigens

The majority of IgG1 mAbs raised from infected mice (12/14) failed to react with HES by Western blot, despite their strong reactivity to native HES by ELISA (Fig. 4 A and data not shown). We interpreted this to mean that the IgG1 response is predominantly directed against conformational protein epitopes, which are destroyed following detergent denaturation during SDS-PAGE. In support of this, while IgM monoclonals are equally reactive to native or heat-denatured HES (Fig. 4 B), heat denaturation of HES ablated all IgG1 mAb ELISA reactivity, with the exception of the two anti-HM-65 IgG1 antibodies (Fig. 4 A). Furthermore, while most IgG1 mAbs (Fig. 4 C) and IgM mAbs (Fig. 4 D) show little reactivity to somatic extract (HEx), all antibodies specific for HM-65 show equally strong binding to HES and HEx.

Figure 4. Most HES-specific IgG1 antibodies recognise heat-labile epitopes absent from parasite extract.

A. ELISA reactivity of anti-HES IgG1 mAbs to native (white bars) and boiled (black bars) HES. Polyclonal sera from 28-day infected C57BL/6 mice were included as a positive control.

B. As above, for IgM mAbs.

C. ELISA reactivity of anti-HES IgG1 mAbs to HES (white bars) and HEx (black bars). MOPC 31C myeloma IgG1 was included as a negative control.

D. ELISA reactivity of anti-HES IgM and IgA mAbs to HES (white bars) and HEx (black bars). Anti-DNP IgM mAb was included as a negative control.

To identify the heat-labile determinant(s) recognised by the IgG1 mAbs, an immunoprecipitation strategy was employed using biotin-labeled non-denatured HES and protein G beads. The majority of these IgG1 monoclonals (10/14) were specific for a band that migrates in the position of VAL-1, as represented by 4-M15 (Fig. 5 A). Two additional mAbs immunoprecipitated antigens comigrating with VAL-2 (4-S4) and VAL-4 (2-11).

Figure 5. Conformation-dependent IgG1 antibodies primarily target VAL antigens.

A-E. Immunoprecipitation of biotin-labeled HES by the indicated mAb or polyclonal infection sera, separated by 2-D SDS-PAGE, and visualised by Western blotting with streptavidin-HRP.

A. Anti-VAL-1 mAb 4-M15.

B. Anti-VAL-2 mAb 4-S4.

C. Anti-VAL-4 mAb 2-11.

D. Primary day 28 C57BL/6 polyclonal infection serum. VAL-1, 2 and VAL-3 are indicated.

E. Secondary day 14 C57BL/6 polyclonal infection serum. Postions of VAL-4 and VAL-7 are indicated, as well as VAL-1, -2 and -3 as above.

F. 1-D SDS-PAGE of immunoprecipitated (unlabeled) HES proteins using bead-conjugated mAb as indicated; boxes indicate segments eluted for mass spectrometric analysis. Additional bands likely reflect heavy and light chains of mAb leaching from beads. Molecular weight markers are indicated in kDa.

G. Peptides of VAL-1, 2 and 4 matched by mass spectrometry indicated by boxes. Mascot scores are also shown.

Two-dimensional analysis of antigens immunoprecipated by day 28 polyclonal infection serum confirmed that the dominant target of primary IgG antibody in C57BL/6 mice was VAL-1 (Fig. 5 D), as well as a spot that co-migrates with another VAL protein abundant in HES, VAL-3 (Fig. 2 A). Moroever, secondary IgG immunoprecipitated two further homologues, VAL-4 and VAL-7 (Fig. 5 E). Comparison of profiles from Western blot (Fig. 2 G, I) and immunoprecipitation (Fig. 5 D, E) indicates that Western blotting gives incomplete picture of IgG1 antibody specificity, as a result of the predominantly conformationally-dependent VAL epitopes, while immuno-precipitation of biotinylated antigens omits the HM-65 group of antigens with low protein content.

To formally identify the immunoprecipitated proteins, samples bound by each antibody were eluted from 1-D gel bands (Fig. 5 F) and subjected to mass spectrometry, matching the respective protein sequences (Fig. 5 G). These assays also indicate that VAL-1, 2 and 4 do not interact with other proteins present in HES, given the absence of any co-immunoprecipitated components.

IgM antibodies recognise a common O-linked glycan epitope on VAL glycoproteins

Despite being bound by several IgM monoclonals (Fig. 3 C), the target antigens VAL-1, -2 and -5 show only limited amino acid homology (14.9% identity, 23.5% similarity, Supplementary Fig. 1), and reactivity is heat stable (Fig. 4 B). We therefore evaluated the possibility of a carbohydrate nature of the target epitope(s). We also noted that the antigens’ predicted molecular weights, based on primary amino acid sequences, were 10-39 kDa lower than their observed gel migration, yet each contained only a single potential N-linked glycosylation site (Fig. 6 A). However, they encoded abundant Ser and Thr residues in a central domain, which were predicted to be O-glycosylated by the NetOGlyc 3.1 program (43), Fig. 6 A, Supplementary Figure 2). This contrasts with another secreted VAL protein of similar abundance, VAL-3 (Fig. 2 A), which does not appear to be recognised by IgM antibodies, lacks predicted O-glycosylation sites (Supplementary Figure 2), and migrates on 2-D gels in a manner consistent with its predicted molecular weight (Fig. 3 A and 6 A).

Figure 6. The immunodominant IgM target is not an N-linked glycan.

A. Molecular weight discrepancies between the predicted MW of indicated VAL proteins based on amino acid sequence and observed migration on 2-D gels. Mature MW is that predicted without the signal sequence, predicted N-glycosylation sites are defined as N(X)S/T, O-glycosylation sites were predicted using NetOGlyc version 3.1 (43). The proportion of predicted sites utilized is not known.

B. Western blot of anti-VAL-1 mAb (4-M15) -precipitated antigen separated on 2D SDS-PAGE and probed with anti-Glycan A mAb (13.1).

C. Silver stained 1-D SDS-PAGE of PNGase F treated (+) or mock treated (−) ribonuclease B control glycoprotein and HES.

D. 1-D Western blot of HES with or without PNGase F treatment using day 28 C57BL/6 polyclonal sera, and anti-Glycan B (HM-65) mAb 14.3. Band shifts associated with enzymatic removal of N-glycans are arrowed.

E. 1-D Western blots of HES with or without PNGase F treatment using anti-Glycan A mAbs. IgM denotes an anti-DNP IgM mAb negative control.

Molecular weight markers are indicated in kDa.

To first confirm that Glycan A is carried on VAL-1 and VAL-2, we showed that the anti-Glycan A mAb 13.1 bound to VAL-1 and VAL-2 on Western blots following their affinity purification using IgG1 mAbs to the conformational epitopes of the VAL antigens (Fig. 6 B and data not shown). Comparison with the profile of immunoprecipitated VAL-1 antigen (Fig. 5 A) indicates that only 3 of the 6 VAL-1 spots react with the anti-Glycan A mAb, as is also evident from Fig. 3 B.

To then investigate the potential role of antigenic carbohydrates, we employed both enzymatic and chemical deglycosylation strategies. When HES was pre-treated with PNGase F to remove N-linked carbohydrates, small mobility shifts were evident in silver stained gels (Fig. 6 C) and Western blots (Fig. 6 D), indicating the removal of N-glycans, yet polyclonal sera and IgM mAb reactivity remained intact (Fig 6 D and E). Similarly, the dominant antigen is unlikely to be an PNGase F-resistant N-glycan with a core α1,3-fucose, since PNGase A treatment of tryptic HES peptides failed to ablate either polyclonal or monoclonal antibody binding (Supplementary Figure 3). Furthermore, mass spectrometric analysis identified a number of peptides containing unconjugated N-glycosylation sites (NxS/T), calling into question whether these are utilized in the VAL proteins of this species (data not shown).

To assess the role of O-glycans, HES was treated with trifluoromethane sulfonic acid (TFMS), which removes both N- and O-glycans. Such chemical deglycosylation resulted in mobility shifts of silver stained HES bands, indicative of glycan removal, comparable to that seen with a control glycoprotein, ribonuclease B (Fig. 7 A). Importantly, TFMS treatment ablated both anti-VAL IgM mAb and polyclonal infection sera recognition of HES (Fig. 7 B and C), whereas TFMS-treated HES retained reactivity with an anti-protein antibody, raised against recombinant H. polygyrus calreticulin, as well as polyclonal serum from rats immunized with HES (Fig. 7 D). These data therefore indicate that the dominant and persistent IgM antibody response is to an O-linked glycan shared by several polypeptide secreted antigens of H. polygyrus. The status of the HM-65 antigen is less clear; whilst PNGase F treatment ablates binding of anti-Glycan B 14.3 mAb to the 65-kDa component but not the high molecular weight complex, the reverse is true for TFMS (Figs 6 D and 7 B). Because the 65-kDa antigen is less abundant, the overall reactivity of 14.3 to TFMS-treated HES is reduced approximately tenfold, but not abolished (Fig. 7 E). Thus a similar or identical epitope is conjugated through different linkages to different carrier macromolecules. It was also noted that TFMS does not remove the PC moiety from HM-65 (data not shown), indicating that a different linkage is used in H. polygrus from that described through N-linked glycans for Acanthocheilonema viteae (44).

Figure 7. Chemical deglycosylation ablates anti-Glycan A IgM antibody binding and reduces anti-Glycan B binding to HM-65.

A. Silver stained 1-D SDS-PAGE of TFMS-treated (+) or mock treated (−) HES and ribonuclease B control glycoprotein. TFMS-treated HES was separated into PBS and SDS-soluble fractions.

B. 1-D Western blots of mock-treated and TFMS-treated HES probed with anti-HES mAbs against Glycans A and B, and day 28 C57BL/6 polyclonal antiserum.

C. ELISA of TFMS-treated (open symbols) and mock-treated (solid symbols) HES, used to measure binding of anti-Glycan A mAb 13.1 (squares), 2-13 (circles) and 3-42 (triangles).

D. ELISA of TFMS-treated (open symbols) and mock-treated (solid symbols) HES probed with polyclonal rat anti-HES serum (circles) or rat antibody to recombinant H. polygyrus calreticulin (Hp-CRT; squares).

E. ELISA of TFMS-treated (open symbols) and mock-treated (solid symbols) HES probed with the anti-Glycan B (HM-65) mAb 14.3. The PBS-soluble fraction of TFMS was used for ELISA in C-E.

Molecular weight markers are indicated in kDa.

Glycans A and B, VAL-1 and VAL-4, but not VAL-2 are expressed by tissue-phase larvae

Upon infection of mice, H. polygyrus larvae first invade the submucosal tissue of the intestinal tract, molt twice (from L3, to L4 and then adult), emerging 8 days later into the lumen of the small intestine (1). As immunity to challenge infection is directed, in part at least, at the tissue-phase larvae (3), we investigated whether the antigens defined from adult parasites are also expressed by fourth-stage larvae (L4) recovered from the submucosa at day 5 post infection. Larvae were cultivated to yield ES antigens, which were probed with each of the specific antibody reagents. Polyclonal anti-HES antiserum reacted strongly with L4 ES (Fig. 8 A); positive reactions to L4 ES were also seen with mAbs to Glycan A (Fig. 8 B) and Glycan B (Fig. 8 C), although in both cases levels were lower than with HES. Interestingly, among the VAL antigens it was found that while VAL-1 (Fig. 8 D) and VAL-4 (Fig. 8 F) were present at similar levels in HES and L4 ES, VAL-2 was absent from the larval stage (Fig. 8 E).

Figure 8. Monoclonal binding to ES from the tissue-phase L4 larvae show expression of Glycans A and B, VAL-1 and VAL-4, but not VAL-2.

A. ELISA reactivity of polyclonal rat anti-HES binding to plates coated with range of concentrations of adult HES (solid symbols) or L4 ES (open symbols)

B. As above, with anti-Glycan A mAb 13.1

C. As above, with anti-Glycan B mAb 9.1.3

D. As above, with anti-VAL-1 mAb 4-M15

E. As above, with anti-VAL-2 mAb 4-S4

F. As above, with anti-VAL-4 mAb 2-11

Glycan A is strongly represented on the adult cuticle, while Glycan B is a somatic antigen

To localise Glycan A and B epitopes within the adult parasite, we probed intact worms and sections by immunofluorescent microscopy. In sections, anti-Glycan A mAb bound the worm surface, highlighting the longitudinal ridges of the cuticle (Fig. 9 A). Anti-Glycan A antibody also bound to the surface of intact adults, in a pattern which similarly emphasised the ridges (Fig. 9 B). At higher magnification, anti-Glycan A antibody was seen to stain an ordered array of epitopes organized longitudinally along the cuticular furrows (Fig. 9 C). In contrast, anti-Glycan B antibodies failed to bind to intact worms (Fig. 9 D) while reacting strongly to somatic constituents in cross sections (Fig. 9 E); in particular, no cuticular binding was observed with anti-Glycan B antibody.

Figure 9. The immunodominant Glycan A is present on the surface of adult worms.

A. Binding of anti-Glycan A mAb 13.1 to transverse sections of adult H. polygyrus. The corrugated cuticular ridges are visible.

B. Binding of anti-Glycan A mAb 13.1 to the surface of intact adult H. polygyrus

C. Binding of anti-Glycan A mAb 13.1 ordered structures within cuticular furrows on the surface of H. polygyrus.

D. Failure of anti-Glycan B mAb 14.3 to bind to intact worms

E. Binding of anti-Glycan B mAb 14.3 to muscle layer and other structures in transverse sections of adult worms.

F. SDS-PAGE of ¹²⁵I-surface-labeled adult proteins immunoprecipitated with anti-Glycan A mAb 13.1, revealed by autoradiography; similar analysis with anti-Glycan B mAb 9.1.3 is also shown. Control IgM and IgG1 (MOPC31C) proteins are shown, as well as polyclonal rat anti-HES and normal rat serum (NRS).

Scale bars in A-E represent 100 μm. Molecular weight markers are indicated in kDa.

To determine the macromolecules to which cuticular Glycan A is conjugated, we surface radiolabeled adult worms and used mAbs to immunoprecipitate Glycan A-bearing antigens. As shown in Fig. 9 F, this procedure indicates that Glycan A is expressed on at least four different molecular weight species, with a 180 kDa band predominating (Fig. 9 F); only a small proportion of ~55-kDa VAL proteins are similarly immunoprecipitated. The conclusion that most surface Glycan A is not borne on VAL proteins is supported by studies reported elsewhere that anti-VAL-1 and -2 mAbs bind only to localized areas of the cuticle (34). In contrast to glycan A, the Glycan B-specific antibodies were unable to immunoprecipitate radiolabelled surface components (Fig. 9 F). A remarkable degree of cross-reactivity between ES and surface proteins was indicated as a polyclonal sera raised against HES was able to precipitate essentially all radiolabelled surface components (Fig. 9 F).

Vaccination with HES confers protection against challenge, whereas passive immunisation with anti-HES mAb does not

Given the protective ability of VAL family members such as Ancylostoma secreted protein in related helminth infections (45), we determined whether passive immunization with the anti-VAL protein and glycan mAbs were able to confer protection against challenge infection. We first wished to verify that adult-worm derived HES, in which VAL proteins are among the major antigens, could effectively vaccinate against challenge with larval parasites, as negative results have been reported in the literature (46). We found, however, that HES vaccination with alum adjuvant induced a potent humoral response, with titers of anti-HES IgG1 comparable to those seen following secondary infection (data not shown). Moreover, vaccinated animals had greatly reduced egg counts at day 14 than animals immunized with PBS-adjuvant alone (Fig. 10 A) while by day 28, HES-immunized animals showed no fecal eggs and had expelled all adult worms (Fig. 10 A and B). Thus, immune responses against adult secretions conferred highly effective and significant protection against larval challenge.

Figure 10. Vaccination with adult HES confers sterile immunity to challenge, but passive immunization with mAbs from infected mice does not lead to worm expulsion.

A. Fecal egg counts at 14 and 28 days post-infection in HES-alum vaccinated mice compared to PBS-alum alone. Each symbol represents a single mouse, and data shown are combined from two independent experiments. *** P<0.0001.

B. Adult worm burdens at 28 days post-infection in HES-alum vaccinated mice compared to PBS-alum alone. Each symbol represents a single mouse, horizontal bars denote mean values, and data shown are combined from two independent experiments. *** P<0.0001.

C, D. Fecal egg counts at 14 (C) and 28 (D) days post-infection in mice receiving passive immunizations of anti-HES IgG1 Mabs, or of MOPC31C control IgG1 antibody. Each symbol represents a single mouse, horizontal bars denote mean values, and data shown are combined from two independent experiments using 0.2 mg (open symbols) or 1 mg (closed symbol) mAb for each dose.

E. Adult worm burdens at day 28 post-infection in the experiments shown in C and D.

F. Adult worm burdens at day 28 post-infection in mice passively immunized with PBS alone, control anti-DNP IgM, or IgM anti-Glycan A mAb 13.1. i.v. Each symbol represents a single mouse, and horizontal bars denote mean values; similar results were obtained in an independent experiment following i.p delivery of mAb.

Mice receiving i.p. injections of IgG1 antibodies were assayed 24 hr following the first transfer, and were found to have specific anti-HES serum antibodies equivalent to between 217-633 μg/ml of monoclonal antibody (data not shown).

We next performed passive immunization experiments with each of the defined specificity mAbs given throughout the infection period. Using either 0.2 mg or 1 mg doses of mAb every 2-3 days, mice given anti-VAL-1, -2 or -4 IgG1 antibodies showed no diminution in egg counts at d14 of infection (Fig. 10C) or d28 (Fig. 10 D), and indeed anti-VAL-4 recipients showed elevated egg numbers in two independent experiments (p<0.05 compared to recipients of MOPC31C). Furthermore, none of the anti-VAL mAbs induced worm expulsion as measured at day 28 (Fig. 10 E).

In the same experiments, we also tested IgG1 anti-Glycan B mAb for ability to passively protect recipient mice; however, as shown in Fig. 10 C-E, this antibody also failed to reduce egg numbers or elicit expulsion of adult worms. Finally, IgM antibody against Glycan A was tested, as no class-switched antibodies of this specificity were observed. As with the other mAbs, anti-Glycan A did not protect against egg production (data not shown) or worm persistence (Fig. 10 F).

Discussion

The model system of H. polygyrus captures many essential characteristics of the gastrointestinal nematode infections that are highly prevalent in human and animal populations (47, 48). The parasite establishes a chronic infection, driving regulatory T and B cell subsets within a Th2-dominated environment (5, 25, 49-51), and altering innate populations such as dendritic cells (52, 53) and macrophages (54, 55). Immunity to H. polygyrus is slow to develop, particularly in the genetically most susceptible hosts (14), but both B cells and the antibodies they produce are important constituents of the protective immune response (4, 15, 16). In this study, we aimed primarily to define the antigens of adult parasites recognised by host serum antibodies, and secondarily to address whether those antibody specificities serve a protective function in the host-parasite relationship.

Previous antigenic analyses of H. polygyrus have involved a mixture of approaches and relatively simple characterization such as 1-dimensional SDS-PAGE or column fractionation (32, 33, 56), or have investigated individual gene products that are postulated to play a role in immune recognition (26, 31, 57, 58). We have adopted here a more global approach to identify the major antigens recognized during primary infection, which we show are well represented in H. polygyrus adult ES (HES) although not in somatic extracts, presumably reflecting the fact that the immune system is exposed in a more continuous fashion to the secreted products of a luminal-dwelling live parasite. Indeed, we have detected Glycan A in the serum of 7-day infected mice, demonstrating that products of a gastrointestinal parasite can disseminate to distant sites (JPH, unpublished data). On-going work has also revealed that T cell antigens, driving the secretion of Th2 cytokines, are enriched in HES compared to somatic extract in a similar manner to serological antigens (JPH and KJF, unpublished observations).

Murine antibody responses are known to be predominantly IgG1 with primary reactivity to secreted antigens in the 50-70 kDa range (14, 59). These antigens correspond to those we have now defined as VAL-1, VAL-2 and VAL-5, and appear likely to represent the products isolated from HES by Monroy and co-workers to achieve a 40% reduction in egg production following vaccination (57). Interestingly, these glycoproteins bear a conserved antigenic carbohydrate we have termed Glycan A, the structural analysis of which is now under way. Glycan A is also strongly associated with the cuticular surface of the adult worm, and is partly responsible for the extensive degree of antigenic sharing between the secretions and surface of adult H. polygyrus, as reported by earlier investigators (33). This may indicate that HES, and the VAL components in particular, are shed from the surface of the worm, or alternatively that molecules secreted in HES remain associated in some form with the parasite cuticle.

The VAL glycoproteins represent the immunodominant target of both IgM (against glycan A) and the class-switched IgG1 response (against a conformational epitope, presumably the protein backbone). Most VAL proteins from other species have been reported to be glycosylated based on discrepancies between predicted and observed mol.wt. (eg Ancylostoma caninum VALs (60) although only N-glycans have so far been identified (on a VAL protein from the cattle nematode Ostertagia ostertagi, (61)). Here we show that a subset of H. polygyrus secreted VAL proteins are decorated with highly antigenic O-glycans, which are most likely concentrated in an serine/threonine-rich tract linking two conserved sperm coating protein (SCP; cd05380) domains. Similar stretches of predicted O-glycosylation can be observed in other nematode VAL proteins, including Haemonchus contortus Hc40 (accession number AAC03562 http://www.ncbi.nlm.nih.gov/), Cooperia punctata ASPs (AAK35199 and AAK35187) and A. caninum ASP-4 and 6 (AAO63576 and AAO63578). It is also important to note that the VAL proteins themselves, rather than only their associated glycans, are likely to play a key role in host-parasite interactions, as H. polygyrus secretes a number of abundant non-glycosylated VAL proteins, such as VAL-3 (34).

The nematode cuticle is an extracellular matrix assembled from specialized collagens and cuticlins (62), with sugar components that may vary greatly between species. For example, Trichinella spiralis conjugates a unique immunodominant glycan with a terminal tyvelose sugar onto multiple peptide backbones (63), providing a target for protective antibodies (64, 65), while larvae of Toxocara canis (which invade the intestinal tract and migrate in tissues), release two unusually methylated and highly antigenic O-linked trisaccharides (66-68). Larvae of Trichostrongylus colubriformis which, like H. polygyrus is a member of the Trichostrongylid family, express a protease-resistant carbohydrate antigen which may act as the target of protective immunity (69). Another antigenic moiety closely related with helminth parasites is phosphorylcholine, which may be associated with high molecular weight proteoglycan-like molecules (30), or individual proteins that resolve on SDS-PAGE (41). Interestingly, both appear to be the case for HES, in which PC is present on a high molecular weight product that does not stain for protein, as well as a 65-kDa component. Most intriguingly, mAbs to this combination (termed HM-65) do not react to PC itself, indicating that a distinct but as yet uncharacterized structure (provisionally named Glycan B) is expressed. It remains possible that “Glycan B” is not a true carbohydrate, but a small haptenic group similar to diethylaminoethanol that is found in ES antigens of the filarial nematode Litomosoides sigmodontis (70). Hence, although we do not yet know the structural nature of Glycans A and B, it is clear that H. polygyrus is not unusual in presenting extensive and immunodominant non-protein specificities to the mammalian host.

Antibodies may exert a protective effect by several pathways (17); in particular they may impede growth and migration during the histotropic larval phase (20, 71), possibly by neutralizing key ES products (6), and may target the exposed epitopes on the surface of adult worms during the luminal phase. A major objective in studying antibody-antigen interactions in H. polygyrus is therefore to address whether particular specificities can confer immunological protection against infection. It is known that antibodies play an important role in protective immunity to this parasite, as B-cell deficient animals suffer impaired immunity against secondary infection (4, 15, 16) as a result of absence of antibody as well as B cell participation in the cellular response. In tests of polyclonal serum antibodies, the passive transfer of secondary infection sera can protect against challenge infection (15, 72), and this was associated with the IgG1 fraction (19). Whilst attempts to transfer immunity with primary infection sera have been less successful (15), purified IgG1 can reduce worm burdens and lead to stunting of adult parasites (20). This may imply that the efficacy of secondary serum antibodies over primary sera reflects elevated titers of anti-worm antibodies, rather than any difference in specificity induced by repeat exposure.

We therefore tested each of the mAb types generated in this study, for their ability to confer protection by passive transfer. However, none exerted any effect on immunity. As we transferred considerable quantities of each, and these mAbs represent the major specificities present in primary infection sera, our data instead argue that the failure of primary sera to protect does not reflect a quantitative insufficiency in terms of concentration, but either an inadequate affinity as a result of limited affinity maturation or, most interestingly, the absence of key new specificities that may arise only after multiple infections. The failure of mAb-mediated passive immunization contrasts with the sterile immunity generated following vaccination with HES which generates circulating anti-HES titers comparable to those seen following secondary infection. We also therefore examined the antibody profile following secondary infection, which is broadly similar to the primary but contains some additional specificities (such as the single-domain VAL-7 antigen). Future work will examine whether these are more effective targets of protective immunity. A further possibility to be tested is that combinations of antibodies, for example against each of the related VAL proteins, are required to neutralize a common function, and that the full range of anti-VAL antibodies are only generated through secondary infection.

An intriguing finding of this study is that the dominant anti-Glycan A response, which is rapidly and extensively stimulated by H. polygyrus, shows little protective capacity. This epitope may thus represent an example of a decoy antigen that is elaborated by the parasite to distract immune responses without risk of it inducing a lethal attack on the worm. It is also notable that despite strongly binding the adult worm surface, anti-Glycan A antibodies are not protective in vivo; this could reflect their restriction to IgM, suggesting that in the absence of class switching, antibodies may not gain access to the intestinal sites of infection. The lack of other isotypes even after 28 days of infection suggests a deficiency in T cell help, potentially at the level of the TFH. As it is known that there are abundant IL-4 producing TFH in the draining mesenteric lymph nodes of H. polygyrus mice early in primary infection (73), it is surprising that no IgG response to Glycan A is mounted, suggesting that this specificity is recognized and/or processed in an unusual manner.

The longevity of parasites can be attributed to their ability to evade or divert host immunity (74), and hence the molecular basis of immune attack is of paramount interest. Our understanding of immune interactions with gastrointestinal helminths is primarily at the level of effector cell populations (39, 75) rather than identification of target molecules. At this time, the definition of parasite molecules in this important model of chronic gastrointestinal infection will provide an essential platform both to analyze the recognition of antigens, and to identify the parasite products (proteins and sugars) that can modulate host immunity and facilitate protection (22). In this report, we have accordingly moved our definition forward from using crude parasite extracts, to the more antigenic and less complex HES, and finally to individual antigenic species, so that future work can study defined glycans and recombinant proteins. We have also established that most of the prominent antigens are secreted not only by the luminal-dwelling mature adult worms, but also by the histotropic larval stage which is considered to be a major target of protective immunity. We can now begin to investigate how and where antibodies act (21), the relative importance of functional neutralization and the recruitment of host effector cells, and the lethal mechanisms that achieve sterilizing immunity against intestinal helminths.

Table 1. Monoclonal antibodies to HES Antigens.

Antigen, Specificity	Clone	Source	Isotype
Glycan A (VAL-1-2-5)	13.1, 2-2, 2-12, 2-13, 2-62, 3-8, 3-11, 3-28, 3-29, 3-40, 3-42, 3-55	d28 SPL	IgM
Glycan B (HM-65)	14.3	d28 SPL	IgA
	3-31	d28 SPL	IgM
	4-M9	d28 MLN	IgM
	4-M7, 4-M17	d28 MLN	IgG1
	9.1.3	HES/IFA	IgG1
VAL-1	2-6, 3-6, 3-10, 3-38, 3-39	d28 SPL	IgG1
	4-M4, 4-M15, 4-M20, 4-M23, 4-M25	d28 MLN,	IgG1,
VAL-2	4-S4	d28 SPL	IgG1
VAL-4	2-11	d28 SPL	IgG1

Abbreviations

ASP

Ancylostoma secreted protein

ES

excretory-secretory products

HES

H. polygyrus adult excretory-secretory products

HEx

H. polygyrus adult somatic extract

HM-65

high mol.wt.-65 kDa antigenic complex

PC

phosphorylcholine

TFH

T follicular helper cell

TFMS

trifluoromethane sulfonic acid

VAL

Venom allergen-Ancylostoma secreted protein-Like