The Echinococcus granulosus antigen EgA31: localization during development and immunogenic properties

© 2004 Blackwell Publishing Ltd

489

Parasite Immunology

,

2003,

25

, 489

–

501

Blackwell Publishing, Ltd.

ORIGINAL ARTICLE

Echinococcus granulosus

antigen EgA31

The


antigen EgA31: localization during

development and immunogenic properties

DIDIER SABOULARD,

1

SAMIA LAHMAR,

3

ANNE-FRANÇOISE PETAVY

2

& GEORGES BOSQUET

1

1

C.G.M.C., UMR CNRS 5534, Université Lyon I, 69622 Villeurbanne Cedex, France,

2

Laboratoire de Parasitologie, Faculté de Pharmacie, Université Lyon I, 69373 Lyon Cedex 08, France and

3

Ecole Nationale de Médecine Vétérinaire, Service de Parasitologie, 2020 Sidi Thabet,

Tunisia

SUMMARY

EgA31 is a fibrillar protein from


thatbehaves as a potent antigen during infestation of dogs. Thelocalization of this antigen during development of the parasitewas investigated by immunohistochemistry in optical andelectron microscopy. The protein is mostly abundant in themicrotriches and subtegumental cells of the adult, whereas it isabsent from protoscolex microtriches. Eggs, the periphery ofcalcareous corpuscles, and the germinal layer were other sitesof accumulation. Immunogenicity of different domains of theprotein was assessed during experimental infection of dogs. Itwas shown that the polypeptide encoded by the

Pst I–Hind III

fragment of the complete cDNA is the most antigenic duringthe infection. The uses of such polypeptide for infectiondiagnosis and as a candidate vaccine protein are discussed.

INTRODUCTION

The cestode


is the causativeagent of hydatidosis, one of the world’s major zoonoses. Thisparasite requires two successive hosts for its complete lifecycle. A wide range of mammals, including cattle and man,can become intermediate hosts, where the larvae form hydatidcysts, mainly in the liver and lungs. The adult worm developsin the small intestine of the definitive host, generally Canidaespecies (1).

Numerous reports have shown the importance of economiclosses and health problems linked to this zoonosis, and attemptshave been made to break the parasite cycle by preventionof metacestode transmission from dogs to human or toherbivore animals. Serological diagnostic tests (2–6) andvaccines (7–11), using purified parasite products and recom-binant proteins, have been developed. In order to overcomethe problems of cross-reactivity linked to crude or partiallypurified preparations derived from

E. granulosus

, severalrecombinant antigens have been expressed and evaluatedas immunological markers or potential vaccine candidates.Among these are the EG95 host-protective antigen from

E.granulosus

oncospheres (12,13), the actin filament fragment-ing protein EgAFFP (14), the recombinant protein 14-3-3(15), P-29 (16) which has recently been shown to be distinctfrom Ag5 (17), and EgDf1, a fatty acid-binding protein (18,19).None of them proved to be totally satisfactory either fortheir efficacy, or with regard to the difficulties encounteredin application on a large scale. We thus decided to increaseour knowledge of EgA31, a fibrillar protein that is one of themajor antigens of

E. granulosus

identified by Fu

et al

. (20).In the present paper, we report on the localization of this

protein during the parasite life cycle. This localization wasdetermined by immunohistochemistry using a new anti-body. We also show that four recombinant polypeptidesderived from EgA31 reveal various immunogenic propertiesof different domains of the native protein in dogs experi-mentally infected with

E. granulosus

.

Correspondence

: Pr Georges Bosquet, C.G.M.C., UMR CNRS 5534, Université Lyon I, 43 boulevard du 11 Novembre 1918, 69622 Villeurbanne Cedex, France (e-mail: [email protected]) and Pr Anne-Françoise Petavy, Laboratoire de Parasitologie, Faculté de Pharmacie, Université Lyon I, 8 avenue Rockefeller, 69373 Lyon Cedex 08, France (e-mail: [email protected]).

Received

: 8 September 2003

Accepted for publication

: 18 November 2003

490

© 2004 Blackwell Publishing Ltd,

Parasite Immunology

,

25

, 489–501

D. Saboulard

et al.

Parasite Immunology

MATERIALS AND METHODS

Sera and animals

For all immunoassays, samples of sera from 25 animals wereused. Most of these samples were collected in regions ofTunisia endemic for human and canine echinococcosis; theyincluded sera from clinically ill and from healthy animalskept at the National Medical Veterinary School of SidiThabet (Tunisia). Sera of dogs devoid of any parasite (naivebeagles, Charles River Laboratories, L’Arbresle, France) andcommercial normal dog sera

(

TEBU, France

)

were also usedas negative controls. All serum samples were stored at

−

20

°

Cuntil used. Young adult

E. granulosus

worms were collectedfrom dogs of various breeds at the National Medical VeterinarySchool of Sidi Thabet (Tunisia). The dogs were anaesthet-ized and sacrificed 4 weeks after infection. The intestinaltract was opened along its entire length to search for immatureadult cestodes (IA), which were found both in the contentsof the intestine and in the scrapings of the mucosa. Proto-scoleces of

E. granulosus

were isolated from sheep liver obtainedfrom various slaughterhouses in Tunisia.

Brood capsules, protoscoleces, and young adults freshfrom the dog intestine (28 days post-infection), were fixedfor 2 h in phosphate-buffered 4% paraformaldehyde.Fixation was followed by a buffer wash. Samples weredehydrated through a graded ethanol series (50%, 60%, 70%)and stored at

−

20

°

C. Soluble extracts from IA and proto-scoleces were prepared by homogenization in 0·1

PBSbuffer (pH 7·4) followed by centrifugation at 10 000

g

at+4

°

C for 30 min. Protein concentration in the supernatantwas determined by Bradford assay (Bio-Rad, France).

Immunohistochemistry

Native EgA31 protein was immunolocalized in broodcapsules, protoscoleces and 28-day-old

E. granulosus

adultsections. For this, fixed samples were dehydrated in ethanol,embedded in paraffin, serially sectioned at 5-

µ

m, deparaffi-nized, and passed through graded alcohols. Sections wereincubated in 5% H

2

O

2

in ethanol to block endogenousperoxidase activity. After brief PBS-Tween-20 washes, theywere blocked for 1 h in a 3% bovine serum albumin solution(Sigma) to avoid non-specific cross reactions. Tissue sectionswere then incubated overnight at +4

°

C with the primary anti-body, a guinea pig polyclonal anti-EgA31 [Pst-Dra] antibody(see below) diluted 1 : 1000 in PBS-Tween-20. After threewashes in PBS-Tween-20, sections were incubated for1 h at room temperature with a secondary antibody, anHRP-conjugated rabbit anti-guinea pig immunoglobulindiluted 1 : 1000 in PBS (Sigma). The sections were thenwashed in PBS-Tween-20 buffer three times and the

secondary antibody fixation visualized by treatment with 3,3

′

-diaminobenzidine (Sigma) and H

2

O

2

. After brief washesin PBS-Tween-20, sections were mounted in Gel/Mount(Microm, France), a permanent aqueous mounting medium.All immunohistochemical experiments included negativecontrols carried out with pre-immune guinea pig serum.Counterstaining was carried out with haematoxylin andshandon blueing reagent (CML, France).

Electron microscopy

The material was dehydrated in an ascending series ofethanols, embedded in LR-White resin, and sectioned withan ultramicrotome. Sections were mounted on nickel grids,rinsed in PBS-Tween-20 buffer

,

pH 7·4, and incubated for2 h in the presence of the anti-EgA31 [Pst-Dra] guinea pigserum diluted at 1/2500 in PBS-Tween 20 buffer

,

pH 7·4, atroom temperature.

Grids were then washed with PBS-Tween-20 buffer andincubated with an anti-guinea pig IgG coupled to 10-nmgold particles (GAG10, Sigma) diluted at 1 : 50 in PBS,pH 8·2, for 1 h, at room temperature. After the reaction,samples were washed five times in PBS-Tween-20 pH 8·2buffer, to remove non-reactive colloidal gold particles.Further, parasites were fixed for 5 min in 3% glutaraldehydeand were counterstained with 3% uranyl acetate (Fluka) for45 min. Sections were observed and photographed on anS800 Hitachi electron microscope.

Cloning into pQE80L expression vector

EgA31 cDNA (1836 base pairs; ref. AF067807) was origin-ally cloned in the

Sal I–Not I

cloning site of the plasmidpBluescript II SK

+

(Stratagene Europe, France). Four restric-tion fragments of this cDNA,

Sal I–Pst I

(506 bp),

Sal I–Hind III

(1132 bp),

Pst I–Hind III

(626 bp) and

Pst I–Dra I

(1243 bp) were subcloned in the polylinker of the expressionvector pQE80L (Qiagen, Courtaboeuf, France) designed toproduce proteins with a His

6

tag at the N-terminus with con-servation of the original reading frame. The recombinantplasmids were used to transform

E. coli

BL21-Gold (DE3)pLysS competent cells (Stratagene Europe, France). Trans-formants were selected on LB-broth agar plates containing100

µ

g /mL ampicillin after 24 h at +37

°

C. Clones werecultured overnight in LB-broth containing 100

µ

g /mL am-picillin and 100

µ

g /mL chloramphenicol at +37

°

C; plasmidDNA was extracted using DNA-binding columns as out-lined in the QIAprep Spin Miniprep Kit protocol (Qiagen).The sequencing reaction of the clones was performedusing pQE80L Sequencing-Primer Set (Qiagen), and sequenceanalysis was obtained with a MegaBACE 4000 automaticDNA sequencer (Amersham Biosciences, Orsay, France). The


Parasite Immunology

,

25

, 489–501

491

Volume 25, Number 10, October 2003


antigen EgA31

various recombinant His-Tag polypeptides are hereafterreferred to as EgA31 [Sal-Pst], EgA31 [Sal-Hind], EgA31[Pst-Hind] and EgA31 [Pst-Dra]. Determination of theirexpression was performed by SDS-PAGE.

Briefly, the cells were diluted 100 times in LB-broth con-taining 100

µ

g /mL ampicillin and 100

µ

g /mL chloramphen-icol and expression of the recombinant polypeptide wasinduced by the addition of isopropyl-

β

-D-thiogalactoside(IPTG, Sigma, France) to a final concentration of 1 m

,when the culture absorbance at 600 nm was at an OD of 0·5.Three hours later, the cells were centrifuged for 10 min at15 000

g

and the cell pellet was resuspended in Laemmlibuffer. The suspension was heated for 15 min at +95

°

C, and50

µ

g of each culture suspension was analysed on 15% SDS-polyacrylamide gels (Bio-Rad, France). Gels were stainedwith the standard Coomassie Brilliant Blue G-250 (Sigma,France).

Mass production and purification of recombinant His-Tag EgA31 proteins

The four His-Tag EgA31 clones were inoculated into250 mL of LB-broth containing 100

µ

g /mL ampicillin and100

µ

g /mL chloramphenicol and incubated with shakingat +37

°

C until an A

600

of 0·5 was reached. Isopropyl-

β

-D-thiogalactoside (IPTG, Sigma) was added to a final concen-tration of 1 m

and the culture allowed to incubate for anadditional 3 h. To purify recombinant His-Tag EgA31proteins, the cell pellets were lysed at room temperature bystirring the pellets in a buffered solution containing 8

urea,pH 8·0. Once the solution became translucent, the cellulardebris was removed by centrifugation. The clarified super-natants were loaded onto nickel-nitrilotriacetic acid-agarosecolumns (Ni-NTA column, Qiagen) to capture the His-Tagproteins. The columns were washed with several volumes ofbuffered 8

urea, pH 6·3. Elution of recombinant His-TagEgA31 was carried out using buffered urea 8

, pH 5·9 andpH 4·5. Eluted proteins were concentrated with CentriconPlus-20 Ultracel-PL (Millipore, St-Quentin en Yvelines,France) and stored at

−

20

°

C until used. They were checkedby SDS-PAGE and Coomassie Brilliant Blue R-250 stainingfor size and purity, and concentrations were determinedusing the Bradford assay (Bio-Rad).

Production of polyclonal antibodies against EgA31 [Pst-Dra] His-Tag protein

Polyclonal antibodies were generated in guinea pigs againstthe purified His-Tag EgA31 [Pst-Dra] subunit (50·6 KD)of EgA31 (Ref. AF067807) by Eurogentec s.a. (Seraing,Belgium). Prior to initial immunization, a pre-immune bloodsample (PB) was drawn from the two guinea pigs, and the

sera were used as negative controls in subsequent Westernblots. His-Tag EgA31 [Pst-Dra] protein was emulsified inComplete Freund’s Adjuvant, and injected into guinea pigs(45

µ

g of protein). Three booster injections (45

µ

g of pro-tein) were made at days 14, 28 and 56 following the initialinjection. Blood was collected at days 38, 66 and 87 (finalbleed) after the initial injection.

Immunoblot analysis

Identification of parasite antigens by the guinea pig poly-clonal anti-EgA31 [Pst-Dra] antibody was carried out fol-lowing the separation of total cestode antigens by sodiumdodecylsulphate polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting (21,22). Identification of thefour His-Tag proteins was made by Western blotting anddot-blotting assays. 150

µ

g of parasite proteins from proto-scolex or immature adults or 20

µ

g of purified His-Tag pro-teins, EgA31 [Sal-Pst], EgA31 [Sal-Hind], EgA31 [Pst-Hind]and EgA31 [Pst-Dra], were run on 15% polyacrylamide gelsprepared in Tris-HCl buffer. Following SDS-PAGE, theseparated proteins were transferred electrophoretically ontonitrocellulose sheets (Bio-Rad, France). The membraneswere washed twice for 5 min with TBS (200 m

Tris, 5

NaCl, pH 7·5) and blocked for 1 h in TBS-5% milk powder.All washing steps with TBS and TBS-T (200 m

Tris, 5

NaCl, pH 7·5 containing 0·05% Tween 20) were performedat room temperature. After blocking, the membranes werewashed for 20 min with TBS-T, then incubated overnight at+4

°

C with the guinea pig polyclonal anti-EgA31 [Pst-Dra]antibody diluted 1 : 1000 in TBS-1% milk powder, or withthe sera from dogs diluted 1 : 1000 in TBS-1% milk powder.

After three washes for 20 min in TBS-T, the membraneswere incubated for 1 h at room temperature with HRP-conjugated rabbit anti-guinea pig immunoglobulin IgG(diluted 1 : 1000 in TBS-1% milk; Sigma, France) or HRP-conjugated rabbit anti-dog immunoglobulin IgG (diluted1 : 3000 in TBS-1% milk; Sigma, France). Excess of secondantibody was removed by washing three times for 20 min withTBS-T and bound antibodies were detected with ready-to-use AEC substrate (Dako, France). Dot-blotting assayswere performed as follows. The four purified His-Tag pro-teins were spotted onto nitrocellulose membrane (Bio-Rad,France) at 100 ng and 50 ng per spot and then allowed todry. The membranes were washed twice for 5 min with TBS(200 m

Tris, 5

NaCl, pH 7·5) and blocked for 1 h inTBS-5% milk powder. Then, the membranes were washedfor 20 min with TBS-T, and treated overnight at +4

°

C withdog sera (1 : 1000 in TBS-1% milk powder). After threewashes for 20 min in TBS-T, the membranes were incubatedfor 1 h at room temperature with HRP-conjugated rabbitanti-dog immunoglobulin IgG (1 : 3000 in TBS-1% milk;

492


Parasite Immunology

,

25

, 489–501

D. Saboulard

et al.

Parasite Immunology

Sigma, France). After three washes for 20 min in TBS-T,bound antibodies were detected with ready-to-use AEC sub-strate (Dako, France).

Enzyme-linked immunosorbent assay (ELISA) procedure – antigenicity of EgA31

The presence of anti-EgA31 antibodies in dog sera wasevaluated by enzyme-linked immunosorbent assay (ELISA)in 22 samples; these included 13 dogs with confirmed

E.granulosus

infection and nine dogs free of parasites (a con-firmation was carried out by necropsy). In the first assay,96-well polystyrene microtitre plates (Ni-NTA HisSorb,Qiagen, France) were coated with the four His-Tag antigensEgA31 [Sal-Pst], EgA31 [Sal-Hind], EgA31 [Pst-Hind] andEgA31 [Pst-Dra], diluted to 1 : 10 in carbonate–bicarbonatebuffer (pH 9·6). The antigen concentrations in this assaywere 1

µ

g, 0·2

µ

g, 40 ng, and 8 ng per well.In the second assay, polystyrene microtitre plates were

coated with EgA31 [Pst-Hind] only. The antigen concentra-tions ranged from 0·25

µ

g /mL to 2

µ

g /mL per well. After anovernight incubation at +4

°

C, wells were washed fivetimes with washing buffer (PBS containing 0·1% Tween 20,Sigma), then blocked with 3% bovine serum albumin inwashing buffer for 2 h. Plates were incubated for 1 h at+37

°

C with dog serum samples (diluted 1/1000) in 0·3%bovine albumin in PBS. For negative control, two rows ofthe microtitre plate were incubated with TBS-0·1% BSAinstead of sera. After five washes, wells were incubatedfor 1 h at +37

°

C with horseradish peroxidase conjugatedanti-dog IgG raised in rabbit (Sigma) diluted 1 : 3000. Theplates were then washed several times, and freshly preparedo-phenylenediamine dihydrochloride (OPD tablets, Dako,France) was added. Reaction was stopped after 30 min with0·5

sulphuric acid and OD was measured at 492 nm usinga microplate reader (Dynatech MR5000 spectrophotometer;Chantilly, VA). All sera were tested in triplicate and averageabsorbance was taken as the final OD. Tests were repeatedtwice to check for reproducibility of results.

RESULTS

Immunohistochemistry

Immunological studies were performed with a polyclonalantiserum, which was generated against EgA31 [Pst-Dra] ina guinea pig. This serum was first checked against proteinextracts of protoscoleces or adult worms to assess the pres-ence of specific antibodies against EgA31. A unique proteinwas detected at the expected size on Western blots (notshown) and the serum was further used for immunodetec-tion after depletion against putative cross reacting proteins.

Protoscoleces

In contrast to the absence of reaction of the pre-immuneserum (Figure 1b), the 2-month antiserum reacted stronglywith the native protein EgA31 on protoscoleces sections(Figure 1a). DAB staining was light on the subtegumentallayer (Figure 1d), and totally absent in the tegument or inhooks (Figure 1a,c,d). The outer layers of the calcareouscorpuscles (Cc) reacted strongly (Figure 1d) and were themost intensely stained structure in the protoscolex. Anapparently uniform labelling was also detected on the broodcapsule, but the EgA31 antigen was obviously no longerpresent in cells forming the stalk connecting the protoscolexto the capsule (Figure 1d). In the cyst wall, the labelling waslimited to the germinal membrane (GM), which producesthe brood capsule (Figure 1e), whereas no EgA31 proteinwas detected in the laminated layer (LL).

These observations were confirmed by immunoelectronmicroscopy. The background labelling was almost negligiblewhen the first incubation was performed with the pre-immuneguinea pig serum (Figure 3c).

Utilization of the immune serum showed that the antigenwas abundant in the subtegumental layer (ST) containinglongitudinal muscles (LM), but no labelling was detectedin either the microtriches or the tegument (Figure 3a). The hookregion was also devoid of fixed gold particles (not shown).Conversely, an intense accumulation was observed in externaland surrounding layers of the calcareous corpuscles (Figure 3e).

Immature adults

Figure 2a shows that on

E. granulosus

pre-adults, the dis-tribution of the EgA31 antigen is slightly different from thaton the larvae: the most intense immunolabelling is locatedin the outer layers of the parasites. Note that the distribu-tion of this labelling is uniform along the anteroposterioraxis, none of the segments exhibiting a more intense stainingthan the adjacent ones; this was confirmed by the observa-tion of numerous animals. On enlargements (Figure 2d), theouter surface of the parasite, the tegument and underlyingparenchymal cells (UP) are the most labelled structures.The muscle region is not intensely stained, contrary tothe calcareous corpuscles (Figure 2e), as previously men-tioned for the protoscoleces. Immature eggs (EG) are veryintensively labelled (Figure 2c). No labelling was observedin controls performed with the corresponding pre-immuneserum (Figure 2b).

The cytological locations of the labelling were examinedby electron microscopy. As previously mentioned for pro-toscolex sections, the non-specific background labelling wasvery low or absent on adult

E. granulosus

sections: almostno gold particles could be detected when the polyclonal


Parasite Immunology

,

25

, 489–501

493

Volume 25, Number 10, October 2003


antigen EgA31

Figure 1 Immunolocalization of EgA31 in protoscoleces sections. DAB immunohistochemistry of a longitudinal section of a protoscolex (100×) (a). Longitudinal section of E. granulosus protoscolex. Negative control. (100×) (b). High magnification of the anterior part of a protoscolex. No staining is observed in the microtriches (MT) (100×) (c). Section of brood capsule wall near the point of attachment of the posterior part of the protoscolex (100×). Note the presence of DAB staining in the brood capsule (BC) and in the calcareous corpuscle (Cc) (d). High magnification of a cyst wall reveals intense DAB staining in the germinal membrane (GM) (100×). Note the absence of DAB staining in the amorphous laminated layer (LL) (e).

494 © 2004 Blackwell Publishing Ltd, Parasite Immunology, 25, 489–501

D. Saboulard et al. Parasite Immunology

Figure 2 EgA31 immunolocalization in 28-day-old E. granulosus adult parasite sections. The longitudinal section shows that the DAB staining is principally bound to the outer layer of the worm’s tegument (10×) (a). Negative control (10×) (b). Portion of the cestode body showing the uterus (UT) with eggs (EG) (40×) (c). Positive labelling on the surface of the parasite as well as in the underlying parenchymal cells (UP) and in the calcareous corpuscle (Cc). Longitudinal muscles (LM). (40×) (d). Height magnification (100×) of a DAB-labelled calcareous corpuscle (Cc) (100×) (e).

© 2004 Blackwell Publishing Ltd, Parasite Immunology, 25, 489–501 495

Volume 25, Number 10, October 2003 Echinococcus granulosus antigen EgA31

antibody was omitted (not shown) or when the pre-immuneserum was used (Figure 3c). The highest density of immu-nogold particles was observed in the tegument (TG) and inthe microtriches (MT), up to their distal part (Figure 3b).

Cytoplasmic extensions of the subtegumental cells(Figure 3f) and glycogen cells (Figure 3g) were also intens-ively labelled. Gold particles could be seen associated with

the calcareous corpuscles (Figure 3e), mainly on the externallayer. Although EgA31 shares some similarities with para-myosins, it was not found specifically associated with musclecells (not shown). It was not detected either in the lumen ofthe protonephridial duct or in bordering cells (not shown),suggesting that the protein was not excreted. It is worthmentioning that whatever the location of the immunolabelling

Figure 3 Electron micrographs illustrating immunogold staining of protoscoleces and 28-day-old E. granulosus adult parasite sections. Proto-scolex: positive labelling in the subtegumental layer containing longitudinal muscles (LM). No labelling is observed either in the microtriches (MT) or the tegument (TG) (a). Immature adult: high magnification reveals gold particles in the microtriches (MT) and in the tegument (TG) (b). Longitudinal section through the microtriches (MT) and the tegument (TG). Negative control (c). Enlargement of syncytial layer (tegument). Note the doubling of the gold particles marked with arrows (d). Protoscolex: high magnification of the calcareous corpuscle (Cc) with gold particles marked with arrows in the external and surrounding layers (e). Cytoplasmic extension of a subtegumental cell with intense labelling (f). Abundant gold particles in the cytoplasm of a glycogen cell. Note that the nuclear structure never retains the anti-EgA31 antibody (g).



Figure 3 Continued.



in the organism, the gold particles were not randomlydistributed, but could clearly be seen gathered in pairs(Figure 3d). This distribution was observed in the micro-triches, the tegument, the layer containing the subtegumentalcells, and to a less extent in parenchymal cells and calcareouscorpuscles of the protoscoleces.

Dot blot and Western blot

In order to assess the presence of antibodies against each ofthe four recombinant antigens, sera from infected and con-trol dogs were used to probe a membrane on which 50 and100 ng of EgA31 [Sal-Pst], EgA31 [Sal-Hind], EgA31 [Pst-Hind] and EgA31 [Pst-Dra] were spotted. Neither the com-mercial dog serum (Figure 4, lane 1) nor the pre-infected dogserum (Figure 4, lane 4) reacted with any of the antigens.After the experimental infection, the serum of the same dogreacted unambiguously with the four recombinant proteins(Figure 4, lane 3). Moreover, as a proof, this serum wasfirst incubated overnight at +4°C with a total protein extractfrom E. granulosus before the reaction on the blots: no signalbound to the membrane. These results were confirmedon Western blots: lanes 5–8 on Figure 5 show that the fourrecombinant antigens were strongly recognized by the infecteddog serum, whereas the serum of the same dog collected beforeinfection failed to react (Figure 5, lanes 9–12).

ELISA

We wanted to determine if the various domains of the EgA31protein were able to trigger immunogenic reactions at com-parable levels. For this, similar amounts of each of the fourrecombinant polypeptides were tested in ELISA experimentsagainst the sera from three dogs collected 28 days after experi-mental infection. Sera of healthy, non-infected dogs, were usedas controls. Concentrations of 8 ng/mL to 1 µ /mL of EgA31[Sal-Pst], EgA31 [Sal-Hind], EgA31 [Pst-Hind] and EgA31[Pst-Dra] recombinant antigens were checked against eachof the sera: Figure 6(a–c), shows a strong reaction with serafrom infected dogs for all these antigens, suggesting that thepresence of E. granulosus in the digestive tract has triggeredthe production of IgG against each of the domains of EgA31,whereas none of the control sera contained anti-EgA31IgG (Figure 6d). However, it is worth mentioning thatthe levels of antibodies against each recombinant antigenwere significantly different, the response against EgA31[Pst-Hind] being the most intense. This difference wasobserved with all the infected-dog sera tested (numerousrepetitions were carried out to ensure the reproducibility ofthe result).

We thus decided to focus on the use of this recombinantantigen to compare the levels of antibodies between sera fromcontrol uninfected dogs, sera from dogs before an experimentalinfection and 28 days after this infection. The EgA31 [Pst-Hind]antigen concentrations range from 0·25 µg/mL to 2 µg/mL.Figure. 7(e,f ) represents the mean OD levels of three ELISAassays for each of the 19 dog sera at 0·5 µg/mL and 2 µg/mL,respectively. Sera corresponding to the control commercialdog and to the pre-infected dogs gave absorbencies less thanOD = 0·2, except for dog number 15. As expected, all dog seracollected 28 days after the infection unambiguously containedantibodies against EgA31 [Pst-Hind], and their responses wererather comparable.

However the dog number 15 remained again as an excep-tion, as the ODs were similar before and after experimental

Figure 4 Dot immunobinding assay using the four recombinant proteins EgA31 [Sal-Pst], EgA31 [Sal-Hind], EgA31 [Pst-Hind] and EgA31 [Pst-Dra]. 50 ng (left) and 100 ng (right) of each recombinant protein were spotted. Lane 1: non-infected dog serum. Lane 2: PBS as negative control. Lane 3: post-infected dog serum. Lane 4: pre-infected serum of the same dog before infection. Lane 5: post-infected serum of the same dog incubated overnight with total protein extracts from E. granulosus immature adults.

Figure 5 SDS-PAGE and Western blot analysis of the four EgA31 recombinant antigens. Lanes 1, 5, 9: EgA31 [Pst-Dra]; lanes 2, 6, 10: EgA31 [Pst-Hind]; lanes 3, 7, 11: EgA31 [Sal-Hind]; lanes 4, 8, 12: EgA31 [Sal-Pst]. Lanes 1–4: Coomassie-stained proteins after SDS-PAGE in 12% gels under reducing conditions. Lanes 5–8: reactivity of the post-infected dog serum against respective proteins, by immunoblot. Lanes 9–12: control reactivity of the pre-infected serum of the same dog against respective proteins, by immunoblot.



infection. A necropsy carried on this 3·5-year-old dogshowed that its intestine was free of any E. granulosus para-site 28 days after the experimental infection. The presenceof the high level of anti-EgA31 [Pst-Hind] antibodies beforethe experimental infection was thus attributed to a likelycontamination of this dog before its recruitment in the pro-tected zone of the veterinary school. It is worth mentioningthat this high anti-EgA31 [Pst-Hind] antibody level beforethe oral infection correlates with the absence of living para-sites in the intestine 28 days later.

Figure 7 Detection of anti-EgA31 [Pst-Hind] antibodies in dog sera by ELISA assay. 0·5 µg and 2 µg of EgA31 [Pst-Hind] were tested for immunoreactivity as shown in histograms a and b, respectively. Results are displayed for non-infected and experimentally infected dogs. Sample 1: control (no serum); sample 2: commercial dog serum (naive beagles, Charles River Laboratories, L’Arbresle, France); samples 3–9: non-infected dog sera; samples 10–20: pre-infected dog sera (solid bars) and post-infected dog sera (open bars).

Figure 6 ELISA reactions of the four EgA31 recombinant proteins with sera from dogs experimentally infected with E. granulosus. Equivalent amounts of the four antigens were tested for immunoreactivity with three post-infected dog sera (a, b, c) and a non-infected dog serum as negative control (d). The EgA31 [Pst-Hind] antigen showed a better reactivity with parasite antibodies. P/D: EgA31 [Pst-Dra]; P/H: EgA31 [Pst-Hind]; S/H: EgA31 [Sal-Hind]; S/P: EgA31 [Sal-Pst].



DISCUSSION

EgA31 is a 66-kDa fibrillar polypeptide identified by immuno-screening of an E. granulosus cDNA library with sera frominfected dogs (20). Although some homology could be estab-lished with the paramyosin family, no totally convincingidentification of this protein could be made from com-parison with sequences in data bases. The presence of theC-terminal domain suggested the possibility of association ofEgA31 in polymers or with other proteins. Preliminary stud-ies revealed that this polypeptide elicits immune reactions inthe dog soon after infection both at the humoral and cellu-lar levels (23), and prompted us to gain more insight into theproperties of this protein.

EgA31 localization during development

During previous studies, the recombinant polypeptide cor-responding to the Sal I–Hind III cDNA restriction fragmentof EgA31 was used to generate an antibody. In our studies,very high levels of EgA31 specific polyclonal antibody wereelicited following guinea pig immunization by the Pst I–Dra I derived fragment. The first result that is worth empha-sizing is the predominant presence of EgA31 at the outersurface of the pre-adult worm: the syncytial tegument,and more strikingly, the microtriches. This observation wasmade both in optical and electron microscopy. Thompsonet al. (24) reported the existence of a microtrichial polymor-phism in the adult parasite, and suggested that the filamen-tous microtriches could allow a closer contact betweenE. granulosus and its host. Careful examination of numer-ous pre-adults led us to conclude that EgA31 distributionwas not affected by this polymorphism, the observed differ-ences along the animal resulting mainly from the variationof orientation of the section with respect to the surface.

This predominant presence of EgA31 in the microtrichesof the pre-adult is in striking contrast with its total absencefrom the microtriches of the still invaginated protoscolex,even in regions where they are well developed (see Figure 1c).The structure of the syncytial tegument of the protoscolexhas been described in detail by Bui et al. (25). The absenceof EgA31 in the outer tegument of the larva and its abund-ance in the adult worm suggests that either the change ofstage is associated with the colonization by a new type ofcell, or more likely that the EgA31 gene is turned on at thisdevelopmental step. This later suggestion is strengthened bythe observation of brood capsules containing protoscoleces(see Figure 1d,e): the germinal layer is strongly labelled(contrary to the amorphous laminated layer), but as thecells enter the vesicle to form the protoscoleces (26), theylose EgA31 expression. Thus, the developmental changescan be characterized by an early expression in the germinal

cells, an arrest of expression as protoscoleces differentiate,and a re-expression in the early and pre-adult worms.

The anti-EgA31 [Pst-Dra] antibody also labelled otherparts of the pre-adult parasite, namely the underlying paren-chyma, the suckers, and the calcareous corpuscles. Intrigu-ingly, a common feature between protoscolex and pre-adultstages is the presence of EgA31 in the calcareous corpuscles.These structures consist of an organic matrix, organizedinto concentric rings, and an inorganic component. Divers-ity in the morphology of calcareous corpuscles has beenreported for E. granulosus (27). Examination of the EgA31antigen localization clearly demonstrated that the nativeform of EgA31 was heavily enriched in the external layer.It has been shown for other species that, aside from thevarious functions ascribed to calcareous corpuscles, e.g.osmoregulation and protection from abnormal calcification(28), these structures have the capacity of binding proteins(29,30). Whether they behave as a reservoir for EgA31 can-not be decided with the present observations.

Two additional observations are worth mentioning. Onthe one hand, no EgA31 was found in the lumen of protone-phridial ducts. This could indicate that, although present atthe surface of the adult, it is not a component of proteinssecreted and excreted by the worm, as for example the 14-3-3 protein produced by the rostellar glands (15). On the otherhand, in all of its adult localization, EgA31 is revealed bypairs of immunogold particles: the distance between thepoints is too great to correspond to the fixation of two par-ticles on a single polypeptide. This arrangement could thussuggest that EgA31 is part of a larger structure involving atleast two EgA31 units and other polypeptides: the predic-tion, from the putative EgA31 amino acid sequence of C-terminal domains which generally participate in polypeptideassociations, supports this hypothesis.

The immunogenicity of the native EgA31 domains

Truncated forms of the native EgA31 were prepared in orderto analyse the production of antibodies by the immunesystem of the dog during an infection by the parasite. Fouroverlapping domains derived from the initial cDNA andcorresponding approximately to domains of the predictedprotein limited by putative changes in the secondary struc-ture (Antheprot program, IBCP, Lyon, France) were pro-duced in pQE80L systems. They were assessed in Dot blot,Western blot and ELISA against sera of uninfected, pre-infected and post-infected dogs. As expected, a uniquepolypeptide of 66 kDa was revealed on Western blot of totalprotein extracts, by the sera of infected dogs (data non shown).These sera were analysed against each of the overlappingdomains, both in Dot blot and Western blot: all the serareacted with each of the domains designed. Specificity of



this reaction was assessed by the absence of signal when thesera were depleted with total protein extracts of E. granulosus,previously to the incubation with the Dot blot membrane.This thus demonstrates that all domains of the EgA31 par-ticipate in the immune reaction evoked by the parasite.

Interestingly, after necropsy several of the pre-infecteddogs were found to be infected by other parasites (e.g. Anky-lostoma caninum, Mesocestoides sp.). Nevertheless, no reac-tion with any of the recombinant antigens was observed,suggesting, at least for these samples, that our recombinantEgA31 proteins could be species- or genus-specific. Thisspecificity remains to be explored before further utilizationof this antigen in the detection of infected dogs.

ELISA assays have been carried out in an attempt tocompare the importance of each EgA31 domain in the stim-ulation of antibody production during infection. Equivalentamounts of recombinant proteins were used in these ELISAassays. All the repetitions of the assays with all the dogsera indicated that the most active epitope is located in therecombinant 26·8-kDa polypeptide corresponding to thePst I–Hind III subfragment of EgA31 cDNA, althoughthe differences from other recombinant polypeptides werenot considerable.

The recombinant EgA31 [Pst-Hind] antigen was thusused to compare the responses of a batch of dogs to aninfection with the parasite: the EgA31 [Pst-Hind] domain isclearly a good marker of the infection in ELISA assays, witha tendency for a correlation between the highest levels ofantibodies and the highest score of worms in the intestine.Moreover, the distribution of absorbency values revealed aclear difference between negative and positive sera. The onlyexception was dog 15, which exhibited rather similar andhigh values of IgG responses prior to and after protoscolexingestion, and which was the only dog without any pre-adult parasites in the intestine 28 days after this ingestion.It must be remembered that this dog was 3·5 years old, anage at which most the dogs have been in contact, possiblyseveral times, with the parasite in this endemic region (31).We thus suggest that prior to our measurements, this farmdog had been infected one or several times before beingdewormed. These previous infestations could have protectedthe dog in our experiments, as suggested by Gemmel andLawson (32), Deplazes et al. (33), Herd (34), and Vuitton (35).Whether the presence of anti-EgA31 [Pst-Hind] antibodieshas participated to the protection mechanism remains to beanalysed.

In summary EgA31 is a developmentally regulatedantigen associated with the germinal membrane of the broodcapsule, calcareous corpuscles and parenchymal cells of theprotoscolex, and with the microtriches, subtegumental cellsand eggs of the adult and immature adult. These locationsexplain the rapid immune response against this protein after

an infestation. The domain encoded by the Pst I–Hind IIIfragment of the cDNA is the most antigenic during theinfection. Work is now in progress to investigate the potentialuse of the recombinant protein EgA31 [Pst-Hind] in develop-ing a new diagnostic tool and a candidate vaccine for E.granulosus infection in dogs.

ACKNOWLEDGEMENTS

This work was supported by European Commission Con-tract ICA4-CT-2000-30030 (Echinostop). The authors aregrateful to Pr. G. Morel Laboratory (Université Lyon I,Villeurbanne, France) for helpful contributions in electronmicroscopy. The authors would like to thank Ms C. Jullien-Moutelon for excellent technical assistance.

REFERENCES

1 Thompson RCA. In Echinococcus and Hydatid Disease, ed.Thompson, RCA. Wallingford: Thompson-Lymbery; 1995: 1.

2 Kittelberger R, Reichel MP, Jenner J et al. Evaluation of threeenzyme-linked immunosorbent assays (ELISAs) for the detec-tion of serum antibodies in sheep infected with Echinococcusgranulosus. Vet Parasitol 2002; 110: 57–76.

3 Ortona E, Rigano R, Buttari B et al. An update on immuno-diagnosis of cystic echinococcosis. Acta Trop 2003; 85: 165–171.

4 Sbihi Y, Rmiqui A, Rodriguez-Cabezas MN et al. Comparativesensitivity of six serological tests and diagnostic value of ELISAusing purified antigen in hydatidosis. J Clin Lab Anal 2001; 15:14–18.

5 Gonzalez-Sapienza G, Lorenzo C & Nieto A. Improved immuno-diagnosis of cystic hydatid disease by using a synthetic peptidewith higher diagnostic value than that of its parent protein,Echinococcus granulosus antigen. J Clin Microbiol 2000; 38:3979–3983.

6 Virginio VG, Hernandez A, Rott MB et al. A set of recom-binant antigens from Echinococcus granulosus with potential foruse in the immunodiagnosis of human cystic hydatid disease.Clin Exp Immunol 2003; 132: 309–315.

7 Lightowlers MW & Gauci CG. Vaccines against cysticercosisand hydatidosis. Vet Parasitol 2001; 101: 337–352.

8 Heath DD, Jensen O & Lightowlers MW. Progress in controlof hydatidosis using vaccination: a review of formulation anddelivery of the vaccine and recommendations for practical use incontrol programmes. Acta Trop 2003; 85: 133–143.

9 Mastroeni P, Chabalgoity JA, Dunstan SJ et al. Salmonella:immune responses and vaccines. Vet J 2001; 161: 104–106.

10 Dalton JP & Mulcahy G. Parasite vaccines: a reality? Vet Para-sitol 2001; 98: 149–167.

11 Dalton JP, Brindley PJ, Knox DP et al. Helminth vaccines:from mining genomic information for vaccine targets to systemsused for protein expression. Int J Parasitol 2003; 33: 621–640.

12 Chow C, Gauci CG, Cowman AF & Lightowlers MW. A genefamily expressing a host-protective antigen of Echinococcusgranulosus. Mol Biochem Parasitol 2001; 118: 83–88.

13 Woollard DJ, Gauci CG, Heath DD & Lightowlers MW. Pro-tection against hydatid disease induced with the EG95 vaccine



is associated with conformational epitopes. Vaccine 2000; 19:498–507.

14 Cortez-Herrera E, Yamamoto RR, Rodrigues JJ et al. Echinoc-occus granulosus: Cloning and functional in vitro characteriza-tion of an actin filament fragmenting protein. Exp Parasitol2001; 97: 215–225.

15 Siles-Lucas M, Nunes CP, Zaha A & Breijo M. The 14-3-3 pro-tein is secreted by the adult worm of Echinococcus granulosus.Parasite Immunol 2000; 22: 521–528.

16 Gonzalez G, Spinelli P, Lorenzo C et al. Molecular characte-rization of P-29, a metacestode-specific component of Echinococcusgranulosus which is immunologically related to, but distinct from,antigen 5. Mol Biochem Parasitol 2000; 105: 177–184.

17 Lorenzo C, Salinas G, Brugnini A et al. Echinococcus granulo-sus antigen 5 is closely related to proteases of the trypsin family.Biochem J 2003; 369: 191–198.

18 Chabalgoity JA, Harrison JA, Esteves A et al. Expression andimmunogenicity of an Echinococcus granulosus fatty acid-binding protein in live attenuated Salmonella vaccine strains.Infect Immun 1997; 65: 2402–2412.

19 Chabalgoity JA, Moreno M, Carol H et al. Salmonella typhimu-rium as a basis for a live oral Echinococcus granulosus vaccine.Vaccine 2000; 19: 460–469.

20 Fu Y, Martinez C, Chalar C et al. A new potent antigen fromEchinococcus granulosus associated with muscles and tegument.Mol Biochem Parasitol 1999; 102: 43–52.

21 Laemmli UK. Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature 1970; 227:680–685.

22 Towbin H, Staehelin T & Gordon J. Electrophoretic transfer ofproteins from polyacrylamide gels to nitrocellulose sheets: pro-cedure and some applications. Biotechnology 1992; 24: 145–149.

23 Fu Y, Saint-Andre Marchal I, Marchal T et al. Cellular immuneresponse of lymph nodes from dogs following the intradermalinjection of a recombinant antigen corresponding to a 66 kDaprotein of Echinococcus granulosus. Vet Immunol Immunopathol2000; 74: 195–208.

24 Thompson RCA, Houghton A & Zaman V. A study of themicrotriches of adult Echinococcus granulosus by scanning elec-tron microscopy. Int J Parasitol 1982; 12: 579–583.

25 Bui LT, Stenzel DJ & Jones MK. Reappraisal of vesicular typesin the syncytial tegument of the Echinococcus granulosus proto-scolex. Parasitol Res 1999; 85: 35–40.

26 Holcman B & Heath DD. The early stages of Echinococcusgranulosus development. Acta Trop 1997; 64: 5–17.

27 Pawlowski ID, Yap KW & Thompson RC. Observations on thepossible origin, formation and structure of calcareous corpus-cles in taeniid cestodes. Parasitol Res 1988; 74: 293–296.

28 Vargas-Parada L & Laclette JP. Role of the calcareous corpus-cles in cestode physiology: a review. Rev Latinoam Microbiol1999; 41: 303–307.

29 Yang HJ. Separation of calcareous corpuscles from plerocer-coids of Spirometra mansoni and their binding proteins. Para-sitol Res 2000; 86: 781–782.

30 Chung YB, Kong Y, Cho SY & Yang HJ. Purification andlocalization of a 10 kDa calcareous corpuscle binding protein ofSpirometra mansoni plerocercoid. Parasitol Res 2003; 89: 235–237.

31 Lahmar S, Kilani M & Torgerson PR. Frequency distributionsof Echinococcus granulosus and other helminths in stray dogs inTunisia. Ann Trop Med Parasitol 2001; 95: 69–76.

32 Gemmell MA & Lawson JR. Epidemiology and control ofhydatid disease. In The Biology of Echinococcus and HydatidDisease, ed. Thompson RCA. London: Allen & Unwin; 1986:189–216.

33 Deplazes P, Thompson RC, Constantine CC & Penhale WJ.Primary infection of dogs with Echinococcus granulosus:systemic and local (Peyer’s patches) immune responses. VetImmunol Immunopathol 1994; 40: 171–184.

34 Herd RP. Resistance of dogs to Echinococcus granulosus. Int JParasitol 1977; 7: 135–138.

35 Vuitton DA. The ambiguous role of immunity in echinococco-sis: protection of the host or of the parasite? Acta Trop 2003; 85:119–132.

The Echinococcus granulosus antigen EgA31: localization during development and immunogenic properties

Documents