|Home | About | Journals | Submit | Contact Us | Français|
The immunogenic properties of cysteine proteases obtained from excretory/secretory products (ES) of Haemonchus contortus were investigated with a fraction purified with a recombinant H. contortus cystatin affinity column. The enrichment of H. contortus ES for cysteine protease was confirmed with substrate SDS-PAGE gels since the cystatin-binding fraction activity was three times higher than total ES, despite representing only 3% of total ES. This activity was inhibited by a specific cysteine protease inhibitor (E64) and by recombinant cystatin. The one-dimensional profile of the cystatin-binding fraction displayed a single band with a molecular mass of 43 kDa. Mass spectrometry showed this to be AC-5, a cathepsin B-like cysteine protease which had not been identified in ES products of H. contortus before. The cystatin binding fraction was tested as an immunogen in lambs which were vaccinated three times (week 0, 2.5 and 5), challenged with 10 000 L3 H. contortus (week 6) before necropsy and compared to unvaccinated challenge controls and another group given total ES (n = 10 per group). The group vaccinated with cystatin-binding proteins showed 36% and 32% mean worm burden and eggs per gram of faeces (EPG) reductions, respectively, compared to the controls but total ES was almost without effect. After challenge the cystatin-binding proteins induced significantly higher local and systemic ES specific IgA and IgG responses.
Cysteine proteases have been identified in most helminth parasites as members of the papain-like clan, the largest subfamily among the cysteine protease class [13, 26]. Their presumed functions such as nutrition uptake, tissue penetration and evasion of host immune responses emphasize their importance as targets for helminth control [13, 23, 26, 33, 39].
Cysteine proteases of Haemonchus contortus, the most pathogenic gastrointestinal nematode parasite of small ruminants, have been the subject of extensive studies. H. contortus infects the abomasum of sheep, goats and other small ruminants and gains nourishment from the fact that it feeds on blood, often resulting in severe anaemia. Cysteine protease activity was detected in the microvillar intestinal tissue of the parasite gut and in its excreted/secreted products (ES) [10, 19, 22, 23]. The most predominant cysteine proteases are the cathepsin B-like proteases (CBL), which are encoded by a family of at least 22 genes . Their abundant expression is apparent from the analysis of all 21 975 expressed sequenced tag (EST) of H. contortus present in GenBank. Approximately 4% of these appear to be derived from CBL genes. However, only a few of these proteases have so far been characterized in more detail. The cysteine proteases detected in ES products of H. contortus ranged in size from 32 up to 51 kDa [10, 23] whereas the predicted molecular weights of cysteine proteases encoded by clustered EST range from 30 to 45 kDa. Despite the high abundance in EST and the presence of a signal peptide, the cysteine proteases were not identified during the proteomic mapping of the most abundant secreted proteins of H. contortus . The use of biotinylated inhibitors combined with a proteomic approach enabled the identification of nine different cysteine proteases present in ES (AC4, GCP7, HMCP1, HMCP1-like, HMCP2, HMCP-2-like, HMCP7, HMCP8, HMCP9) .
Specific cysteine protease inhibitors such as cystatin can provide an alternative method for enrichment of excreted cysteine proteases by affinity chromatography. Cystatins, members of the family 2 cysteine protease inhibitors, are natural, reversible, tight-binding cysteine protease inhibitors and represent important regulators of proteolytic processes [7, 30, 35].
In this report we enriched ES for cysteine proteases using recombinant H. contortus cystatin affinity chromatography, identified the proteins thus enriched and evaluated their protective effect against a challenge infection of H. contortus in a vaccination trial.
H. contortus cystatin (GenBank™ Accession No. AF035945) was expressed and purified as described by  with slight modifications. The soluble fraction was dialysed against 20 mM Tris-HCl, pH 7.4 and purified through sequential chromatography with ion exchange columns (Mono Q and Mono S, GE). The fractions were analysed by SDS-polyacrylamide gel electrophoresis (1D-SDS-PAGE).
Adult H. contortus (Moredun isolate) were harvested from the abomasum of donor sheep at 25 to 35 days post-infection. Total ES was obtained as described previously . Two affinity columns were prepared: one with recombinant H. contortus cystatin and one with an unrelated Escherichia coli recombinant protein (Cooperia punctata Cp-ASP-1a, Accession No. gi 13625909). Freeze-dried CNBr-activated Sepharose 4B Fast Flow (2 mg, GE) was swollen in 50 mL 1 mM HCl and washed with 200 mL 1 mM HCl using a sintered glass filter. After washing, 50 mL coupling buffer (0.1 M NaHCO3, 0.5 M NaCl pH 8.3) was added and removed under vacuum. The column material was gently mixed with 4 mL of either purified recombinant cystatin or Cp-ASP-1a recombinant protein (1 mg/mL) in coupling buffer for 4 h at room temperature (RT) and washed with 50 mL coupling buffer, using sintered glass filter and under vacuum. Remaining active groups were blocked with 0.1 M Tris-HCl pH 8.0 for 2 h at RT. The column material was rinsed (0.1 M Tris, 0.5 M NaCl pH 8.0 (50 mL) followed by 0.1 M acetate buffer, 0.5 M NaCl pH 4.0 (50 mL; this cycle was repeated three times), resuspended in PBS and packed into a column (BioRad, Hercules, CA, USA). The columns were washed in equilibration buffer (50 mL 50 mM acetate buffer, 0.15 M NaCl pH 5.0) at a flow rate of 1 mL/min using the Econo System (Controller-model ES-1, Pump-model EP-1, UV monitor-model EM-1, Biorad). Before loading on the Cp-ASP-1a Sepharose 4B column (flow rate 0.2 mL/min), 5 mg total ES was dialysed against equilibration buffer for 18 h at 4 °C (Slide-a-Lyzer, Pierce, Rockford, IL, USA). The unbound fraction (20 mL), obtained after washing the column with equilibration solution, was loaded on the cystatin Sepharose 4B column (flow rate 0.2 mL/min) and washed with 20 mL equilibration buffer. For elution 10 mM Tris-NaCl pH 7.4, 0.1% CHAPS, 6 M Urea (50 mL, flow rate 1 mL/min) was used. The fractions were dialysed against 10 mM Tris pH 7.4, 0.05% CHAPS for 18 h at 4 °C (Slide-a-Lyzer, Pierce), concentrated in 3 kDa filters (Centripep YM3, Millipore, Billerica, MA, USA) and the protein concentrations were measured.
The protein profiles of the cystatin-binding fractions were visualised by silver staining after SDS-PAGE. Protease activity was further characterized by electrophoresis under non-reducing conditions on gelatin substrate containing gels as described earlier . Protease activity was defined using class-indicative inhibitors (from Roche, Indianapolis, IN, USA) namely the cysteinyl (L-transepoxysuccinyl-L-leucylamido- (4-guanidino)-butane; E64; 40 μM), metallo (ethylenediaminotetraacetic acid; EDTA, 400 μM), aspartyl (Pepstatin, 1.4 μM) and serine (4-(2-Aminoethyl)-benzenesulfonyl-flouride hydrochloride; AEBSF, 2 mM) protease inhibitors as well as 0.12 μg/mL recombinant cystatin or 0.43 μg/mL B. bovis recombinant protein (as control for E. coli proteins) in 20 mM Tris/ 50 mM NaCl pH 5.0, supplemented with 2 mM DTT).
Thirty Zwart-Bles lambs, 6–6.5 months of age and kept indoors since birth to exclude helminth infections, were randomly divided into three groups with 10 animals each. The doses for the immunization were chosen proportionally to the purification (3% bound to the cystatin column). Animals from group 1 received 2 μg of the cystatin-binding fraction whereas group 2 was immunised with 75 μg of total ES and group 3 was the adjuvant control group (PBS only). The animals were vaccinated subcutaneously, at weeks 0, 2.5 and 5 from the start of the experiment. The antigens or PBS were dissolved in aluminium hydroxide gel (Al(OH)3; Allhydrogel, Superfos Biosector, Denmark) and each animal received 1.5 mg adjuvant/injection (1 mL/animal). At week 6 all animals were orally infected with 10 000 L3 H. contortus and killed at week 10. Faecal samples were collected weekly and after challenge three times a week and egg counting was performed according to the modified McMaster method. Worms were harvested and counted as described . All animal procedures were in accordance with the Ethical Committee from the Faculty of Veterinary Medicine from Utrecht University.
Animals were bled from the jugular vein in weeks 0, 6, 7 and 10 for isolation of lymphocytes  and LPA was done according to Schallig et al.  using 10 μg/mL ES and 5 μg/mL concanavalin A (conA) for lymphocyte stimulation. The results are presented as stimulation indices (SI) where SI = c.p.m. (experimental)/c.p.m. (medium control).
Individual abomasal tissues (~50 cm2) were collected at the time of slaughter for mucus isolation according to Kanobana et al. . All mucus samples were diluted to a concentration of 1 mg/mL for the performance of the enzyme linked immunosorbent assay (ELISA).
The ELISA were performed as described previously  with slight modifications. Briefly, ELISA plates coated with ES (2 μg/mL) were incubated with either serum or mucus, diluted at 1:20 for IgE and 1:100 for IgG in serum and 1:10 for all isotypes in mucus. The positive control serum consisted of a pool from 5 hyperimmune sheep which had been repeatedly infected with H. contortus. Each individual sample was tested in duplicate and the results are shown either as a percentage of the positive control serum that was present in duplicate on every plate or as OD values for the mucus samples.
The material that bound to the cystatin column (see section 2.2.) was trypsinised and analysed using liquid chromatography MS/MS as described previously . In summary, the bound fraction was delivered at 3 ul/min to a nano-LC system coupled to a Q-TOF (Micromass Ltd., Manchester, UK) and using a Famos autosampler (LCPackings, Amsterdam, The Netherlands) and trapped on an AquaTM C18RP column (Phenomenex, Torrance, CA; column dimensions 1 cm x 100 um inner diameter). After flow splitting down to 150–200 nL/min, peptides were transferred to the analytical column (PepMap; LC Packings, Amsterdam, The Netherlands; column dimensions 25 cm x 50 um inner diameter) in a gradient of acetonitrile (1% per min). Fragmentation of eluting peptides was performed in a data-dependent mode, and mass spectra were acquired in a full-scan mode. The MS/MS derived data were searched against GenBank Protein and EST databases using MASCOT software with the following parameters: oxidation of methionine as a variable modification, peptide and fragment mass tolerances of 0.3 and 0.2 Da respectively with a maximum of one missed cleavage.
Statistical analyses were carried out using the SPSS statistical package software (Chicago, IL, USA) and data were analysed with the non-parametric Kruskal-Wallis test. Subsequent group pairwise comparisons were analysed through the Post Hoc test as advised for Kruskal-Wallis and the confidence level was set at p < 0.05 (two-tailed). The Bonferroni correction was employed to avoid possible false positive associations generated by multiple comparisons. Correlations between the immunoglobulin levels and parasitological parameters were tested using the Spearman rank correlation coefficient and considered significant at p < 0.05.
Three percent of total ES protein bound to the recombinant cystatin column and the proteins were analysed by 1D-SDS-PAGE (Fig. 1) and gelatin-substrate gels (Fig. 2). Despite the bound protein fraction resolving as a single band around 43 kDa (Fig. 1, lane 3), this fraction displayed strong protease activity, over a wide molecular size range with a lower limit of 30 kDa (Fig. 2, lane 1). Further analyses were performed using different protease inhibitors. Recombinant cystatin and the specific cysteine protease inhibitor E64 (Fig. 2, lane 2 and 4, respectively) inhibited the activity of the cystatin binding fraction in contrast with AEBSF, a C. punctata recombinant protein or a mixture of EDTA, AEBSF and Pepstatin (Fig. 2, lane 3, 5 and 6).
Although a quantitative analysis of the specific activity of the partially purified H. contortus proteases was not feasible, a titration of ES and the cystatin-bound fraction by substrate-SDS-PAGE analysis indicated that protease activity was enriched at least four fold after purification (not shown).
The MS/MS spectra generated an identification of a sequence stretch composed of the 16 aminoacids, FFEYDGVVSGGPYLGK (See figure in Appendix), which was specifically identified to the cathepsin B-like protease AC-5 (AAA29176) and homologous EST (like BG734242) This sequence corresponds to aminoacids 170 to 185 and is located after the active site, in a region of high divergence between cathepsin B-like proteases allowing unambiguous identification of AC-5 (Fig. 3). The predicted molecular weight of AC-5 with the pro-region and without signal peptide is 36.7 kDa and without the pro-region is 29 kDa while the identified band was 43 kDa.
Lymphocytes from all groups proliferated after stimulation with concanavalin A were used as the positive control (Fig. 4a) although no significant differences were evident between groups. With ES as mitogen, the response in both vaccinated groups, 1 (cystatin) and 2 (ES), was greatly enhanced after the third immunisation (Fig. 4b, open bar) compared to the adjuvant group 3 (p < 0.05). Group 2, stimulated with total ES, displayed the highest ES-specific proliferation response for all time points compared to the other groups (p < 0.05 at week 4 after challenge). The local lymphocytes (Fig. 4b, striped bar) were more reactive to ES antigens than peripheral lymphocytes (Fig. 4b, white, grey and checkered bars). A negative correlation was found between cumulative eggs per gram of faeces (EPG) and LPA at the lymphocytes isolated from local lymph nodes at 4 weeks after challenge (p < 0.05, r = −0.046).
IgG anti-ES antibody concentrations increased markedly following immunisation with the cystatin-binding fraction (group 1) and total ES (group 2) with levels continuing to increase post challenge until the end of the experiment (Fig. 5a). Similarly, IgE levels (Fig. 5b) were elevated two to three-fold in these groups up to the time of infection where control group 3 showed a marked IgE response 2 weeks post challenge. IgA levels showed little response to immunisation but rose rapidly following infection with the response in the cystatin-binding group 1 being two-fold and significantly higher than that in the total ES group 2 and the adjuvant control group 3 (Fig. 5c).
Local immunoglobulins measured in the mucus showed that all immunised animals had higher local IgA and IgG responses compared to the controls (Fig. 5d). IgE was undetectable in all the groups. Group 1, vaccinated with the cystatin-binding fraction, exhibited the highest IgA- and IgG-levels (p < 0.05). Among all the correlations tested for systemic and local immunoglobulins with the parasitological parameters, only three correlations were found. Negative correlations between local IgA and local IgE with worm burden (r = −0.40 and r = −0.42, respectively, p < 0.05) and local IgG with cumulative EPG (r = −0.37, p < 0.05).
Mean EPG levels (Fig. 6) in all three groups increased until 25 days after challenge. From day 28 onwards a decrease in EPG was observed for group 1 (cystatin-bound fraction) in comparison to groups 2 (ES) and 3 (control) (Fig. 6). In general, animals of group 1 had lower mean EPG levels than the other groups (p < 0.05 at day 31).
Animals vaccinated with the cystatin-binding fraction (group 1) had means of 32% EPG and 36% worm burden reductions compared to the adjuvant control group (not significant). No effects were seen in Group 2, vaccinated with total ES (Tab. I) and changes in between groups in the sex ratio of the worms were not observed. There were positive correlations between EPG and worm burden (r = 0.64, p < 0.01), EPG and fecundity (r = 0.42, p < 0.05), worm burden and fecundity (r = 0.40, p < 0.05) and negative correlations between fecundity and protection (r = −0.59 for EPG and r = −0.53 for worm burden, p < 0.01).
Previous work showed that vaccination with adult ES antigens enriched for cysteine proteases using Thiol-Sepharose affinity purification, can induce a protective immune response in sheep against H. contortus . Protection was indicated by reductions in egg output and worm burden by 52 and 50%, respectively, compared to the adjuvant control group. However, these protective fractions contained a number of proteins including several metalloproteases, aminopeptidases and an apical gut protein . Here, we have purified the 43 kDa AC-5 cysteine protease from ES using a recombinant H. contortus cystatin affinity column. Lambs vaccinated with this AC-5 had mean reductions of 32% and 36% in cumulative egg output and worm burden, respectively, compared to the adjuvant control group (Tab. I). This level of protection is comparable to that of one of our previous studies  and with that observed in lambs vaccinated with H. contortus intestinal cysteine proteases purified from worm extracts using cystatin .
The AC-5 cysteine protease described here (Fig. 6) had not been identified in ES products of H. contortus before. Other secreted cysteine proteases such as AC-4 and GCP7 in ES  and HMCP-1, HMCP-1-like, HMCP-2 and HMCP-2-like, HMCP-7, HMCP-8 and HMCP-9  that had been identified before by Mass Spectrometry were not detected in the present cystatin bound fraction. 2D gel analysis of this fraction resolved only a single row of spots (data not shown). Since only a single protease was identified by MS, this suggests that the 43 kDa peptide is a single protein with different pIs and not a mixture of other cysteine proteases with different pIs but of similar molecular weight.
A characteristic of type-2 cystatins is the conservation of the pentapeptide sequence, recognised as a target enzyme-binding site [1, 4]. Two rice cysteine protease inhibitors (OCI-OCII), with small differences in their target enzyme-binding site showed different degrees of affinity for Meloidogyne hapla cysteine proteases [1, 15]. Thus, variation in the pentapeptide sequence may determine differences in affinity for different proteases and may indicate that the previously identified  seven secreted cysteine proteases are absent from the H. contortus cystatin-1-bound fraction because of a low affinity for this inhibitor. Alternatively it cannot be excluded that many secreted CBL have a cystatin bound to their active site thus prohibiting their binding to the cystatin column.
The enrichment of H. contortus ES for cysteine protease activity was confirmed using substrate SDS-PAGE gels (Fig. 2) with activity in the cystatin-binding fraction being judged to be at least 4-fold higher than in whole ES (not shown). This analysis also showed that the protease resolved as several zones of apparently differing molecular size, an observation which, at first sight, conflicts with the above discussion. This is likely to reflect post-translational modifications such as glycosylation, with two potential asparagines in the AC-5 sequence predicted to be potentially N-glycosylated by NetNGlyc1. Moreover, substrate gels are run under non-reducing conditions.
AC-5 is described to be, among five CBL tested in two H. contortus isolates, the only homozygote and monoallelic CBL gene whereas the others were extremely polymorphic . Therefore, AC-5 may be a particularly good vaccine target.
Animals vaccinated with the cystatin-binding fraction and subsequently challenged with H. contortus showed significantly higher mucosal ES-specific IgA levels (Fig. 5) and abomasal lymphocyte proliferative responses (Fig. 4) compared to the control animals. This may indicate the importance of the local antibody response against a challenge infection with H. contortus and was in agreement with higher mucosal Ostertagia specific IgA levels in protected calves after vaccination with thiol-binding proteins derived from ES . The cystatin-binding fraction also induced strong ES-specific systemic and local IgG responses. Previously, vaccination with thiol binding ES proteins of H. contortus resulted in significant higher systemic IgG1 responses .
The total ES fraction did not confer any reduction in EPG or worms, in agreement with a recent report . Variable results have been obtained with vaccination with ES in the past within our group [2, 27, 29, 34], due we suspect to variability between different ES batches (as observed by variable protein patterns observed on 1 and 2-dimensional protein gels) combined with the fact that some individual sheep animals fail to respond to vaccination.
There is now a consistent body of evidence suggesting that ES cysteine proteases may be appropriate targets for vaccine development against helminths of livestock. Calves vaccinated with a Thiol-Sepharose-enriched fraction of Ostertagia ostertagi ES had a 60% reduction in egg output . In a further experiment, this immunogen was subfractionated through Q-Sepharose anion exchange chromatography and a group injected with a resultant cysteine protease enriched fraction had a reduction of 80% in cumulative faecal egg output compared to controls . A similar thiol-enriched fraction was tested in goats, with 89% and 68% in egg and worm reduction, respectively . These experiments above used Freunds or QuilA as adjuvants, in contrast to the aluminium hydroxide employed in ours. In addition, cysteine proteases from the regurgitant of mature worms induced 80% protection in sheep against Fasciola hepatica, as judged by egg output,  and secreted cathepsin Ls are lead vaccine candidates for Fasciola and schistosomiasis (reviewed in ) and hookworms in man (e.g. ). The identification of AC-5, here, as an ES immunogen from H. contortus, adds to this body of evidence and expands the previously identified set of immunogens from H. contortus such as H11, Hc40, H-gal-GP [17, 21, 32].
This work was supported by the European Union (Project QRLT-PL-1999-00565). The H. contortus cystatin was kindly provided by Dr Diane Redmond from the Moredun Institute.