In situ immunocytochemical detection of cells containing antibodies specific for Eimeria tenella antigens

Journal of Immunological Melhodv. 151 (1992) 191-199

191

g'~ 1992ElsevierScience PublishersB.V. All rightsreserved0022-1759/92/$05.00

JIM 06312

In situ irnmunocytochemical detection of cells containing antibodies specific for Eimeria tenella antigens L. Vervelde a, A.N. Vermeulen h and S.H.M. Jeurissen ~ a Central Veterinary hzstitlae, Department of Viroh~gy. Lelystad. Netherlands, and h Interret International B. V., Boxmeer, Netherlands

(Received I I November1991,revisedreceived 17 December 1991,accepted 7 February 1992) A three-step immunocytochemical method for the in situ detection of antibodies specific for Eimeria tenella has been developed. The method is based on the binding of E. tenella antigens to antibodies in cryostat sections of chicken tissues and the recognition of these antigens by rabbit antiserum specific for E. tenella or mouse monoclonal antibodies specific for E. tenella, l'hc rabbit antiserum and mouse monoclonal antibodies were revealed by the immunoperoxidase technique. Suspensions of sonicated sporulated oocysts, incubated with or without various concentrations of the non-ionic detergents Triton X-100 (TX-100) or Nonidet P-40 (NP-40), were used as antigen. Ceils containing antibodies specific for E. tenella were detected only when detergent extracts of sonicated sporulated oocysts were used. After chickens were intravenously immunized with a suspension of sonicated sporulated oocyst antigen, cells containing antibodies specific for E. tenella antigens were detected in the red pulp of tile spleen. After simultaneous immunoenzyme staining for isotype and antigen specificity, the E. tenella-specqfie antibody-containing cells were of the IgM isotype after the primary immunization and of the igM and lgG isotype after the booster immunization. Immune complexes specific for E. tenella on the surfaces of follicular dendritic cells in the germinal centers were also stained. Chickens were also orally infected with sporulated oocysts. In these experiments, cells containing antibodies specific for E. tenella were detected in the lamina propria of the ceca and in the red pulp of the spleen. Specific immune complexes were also detected in the germinal centers of the cecal tonsils. When detergent extracts of sonicated sporulated ooeysts were characterized by immunoblotting, rabbit antiserum specific for E. tenella reacted with proteins ranging in size from 16 kDa to 200 kDa, with major bands of 20 kDa, 24 kDa, 45 kDa, and 100 kDa. Monoclonal antibodies specific for E. tenella recognized only proteins of low molecular weight (20 kDa and 24 kDa) or high molecular weight (80-100 kDa). Immune chicken serum reacted with proteins of low and high molecular weight but especially with proteins of 100 kDa and 113 kDa. This method is the first by which immune complexes and cells containing antibodies specific for parasitic antigens can be detected in situ and may be of value for studies of the local humeral immune response to E. teneUa in the mucosa of chickens. Key words: lmmunocytochemica[staining;Antibody-containingcell: In situ detection; Eimena tenella; (Chicken)

Correspondence to: L. Vervelde, Central Veterina~ Institute, Department of Virology. P.O. Box 36J. 8200 AI Lelyslad, Netherlands Tel.: 31-32011-76.61l; Fax:31-3200-42,804. Abh,eriations: E. tenella, Eimeria tenella; TX-10fl, Triton X-10ft; NP-40, Nonidet P-40; BSA. bovineserum albumin: PBS, phosphate-buffered saline.

192 Introduction

Various approaches have been described for the detection of specific antlbody-containing cells that recognize the antigens responsible for their development (reviewed in Van Rooijen and Claassen, 1986). Such ceils are usually detected by two. or three-step immunoenzyme or immunofluorescence methods and are used to study humoral immune responses to model antigens, e.g., sheep red blood cells (Steven, 1980), human serum albumin (White et al., 1970), horseradish peroxidase (Strans, 1968; Sminia et al., 1983), and trinitrophenyl (Claassen and Van Rooijen, 1984). The indirect immunocnzyme method for detecting specific antibody-containing cells in tissue sections has also been used to study responses to bacterial antigens (Jeurissen et al., 1987). However, this method has never been used to detect cells containing antigen-specific antibodies in parasitic infections, although protozoan and helminthie infections of man and domestic animals have a tremendous impact upon health and socio-economic development. The development of such a test has undoubtedly been problematic, because the life cycles of parasites are complex, and parasites express different antigens at successive stages of their development. Coccidiosis is an economically important disease in chickens and is caused by protozoan parasites of the genus Eimeria. Eimeria tenella specifieaUy parasitizes the ccea of chickens. Once infected with E. tenella, chickens become immune to challenge beeanse of cellular and humoral immune responses and because of non-immune defense mechanisms (Davis, 1981). The induction of humoral immune responses in naive chickens is generally studied in serum and intestinal fluid, but not in the lymphoid tissues of the intestine. Although earlier studies on the humoral immune response in chickens infected with Eimeria were based on the staining of intracellular immunoglobulins, the antigen specificity of intraeellular immunoglobulins could not be determined (Davis et al., 1978; Nash and Speer, 1988; Jeurissen et al., 1989). Indirect fluorescence tests, however, were able to detect antigen-specific responses, such as IgA antibodies directed against sporozoites of E. tenella in bile (Rose and lies-

keth, 1987). Davis et al. (1978) used immunodiffusion to test sera and cecal contents for precipitating antibodies against preparations of E. tenella antigens, but were unable to locate or quantify antigen-specific plasma cells. Although many investigators have studied immune responses to eoccidia, it is still not dear which stages of the life cycle of the parasite elicit a protective immune response in the ccca of the host. We studied the humoral immune response in the ceca in order to detect cells containing antibodies specific for E. tenella antigens. The detergent extracts of sonicated sporulated oocysts used as antigen suspensions in the detection method were characterized by immunoblotting.

Materials and methods

Chickens White Leghorn chickens were maintained under ~pecific pathogen-free conditions with free access to food and water in accordance with institutional guidelines. Parasites E. tenella (Weybridge strain) was passaged with regular intervals through coeeidia-free chickens. Oocysts were purified, sporulated, and stored according to the procedures of Long et al. (1976). Antigen preparation Sporulated oocysts of E. tenella (Weybridge strain) were washed free of potassium bichromate and suspended in 25 mM Tris-HCl (pH 8.0) with 1 i~M phenylmethylsulfonyl fluoride (PMSF, Sigma Chemicals Co., St. Louis, MO, USA, l0 T oocysts/ml). The oooysts were broken by shaking them with glass beads on a Vibrofix (Janke and Kunkel, Germany) set at maximum for 5 min. The suspension was microscopically examined in order to verify that most oocysts and sporocysts were ruptured. Antigen was obtained by sonication of this suspension on ice. To obtain detergent extracts of sonicated sporulated oocysts, the suspension was incubated with 0.1%, 0.5%, 1.0%, or 2.0% TX-100 (Sigma) or NP-40 (BDH Chemicals, Poole, UK) for 1 h on ice and clarified by centrifugation at 3500 × g (Baetifuge, Heraeus,

193

Germany) for 25 min. The protein content of the supernatant was determined according to the method of Lowry et al. (1951) using bovine serum albumin (BSA) as the standard. The extracts of sonicated sporulated oocysts were used as antigen in the immunoenzyme staining technique, while the suspension of sonieated sporulated ooeysts was used to inoculate the chickens. Both suspensions were stored at -20°C before use.

Experimental design Five adult chickens were intravenously inoculated with 0.5 ml sonicated sporulated oocyst antigen suspended in 0.9% NaCI. They were given a booster inoculation with the same suspension after 16 days. All chickens were exsangninated 6 days after inoculation. Nine chickens (4 weeks of age) were orally inoculated with 5000 sporulated oocysts of E. tenella (Weybridge strain). The chickens were exsanguinated 7, 10, and 14 days after inoculation. Three uninfected control chickens were killed at day 10. Seven chickens (2 weeks of age) were orally inoculated with 2500 sporulated oocysts and were inoculated again 2 weeks later with 7500 oocysts. They were then challenged 2 weeks later with 15,060 oocysts and were exsanguinated 4 days after challenge. Control animals were not infected or challenged. After exsangnination, the spleen, ceca, duodenum, and bursa of Fabricius were remeved snap frozen in liquid nitrogen, and stored at -20°C.

lmmunohistochemistry Cryostat sections (-20°C, 8 ~ m thick) of chicken tissue were picked up on glass slides and stored over silica gel. Slides were fixed for 10 rain in pure: acetone, and air-dried. Although 0.01% H 2 0 2 can be used during fixation to inhibit endogenous peroxidase activity, we prefered to use the endogenous peroxidase activity of polymorphonoclear cells as a marker for the red pulp of the spleen. Slides were then incubated with antigen for 45 min at 4°C (1/5 diluted in TrisHCI/PMSF). Slides were rinsed in phosphatebuffered saline (PBS, 0.01 M, pH 7.4) and covered for 45 min with rabbit antiserum, which had been raised against sporozoites but recognized all

developmental stages of E. ten.ella (Jeurissen et al., 1989). Also, mouse monoclonal antibody E.TEN IIG-2, which recognizes sporozoites and gametocytes of E. tenella, and monoclonal antibody E.TEN 11P-2, which recognizes mature first and second generation schizonts, were used to detect specific antibody-containing cells (Jeurissen et al., 1989). After washing the slides in PBS, they were covered with peroxidase--conjugated goat anti-rabbit Ig (Diagnostics Pasteur, MarnesIn-Coquette, France) or peroxidase-conjngated rabbit anti-mouse lg (Dakopatts, Glostrup, Denmark) for 45 rain. The antiserum, the monodonal antibodies, and the conjugates were diluted in PBS containing 0.6% BSA and used at appropriate concentrations. To develop peroxidase activity, slides were treated with 0.5 mg 3,.3'-dlaminobenzidine-tetrahydrochloride (DAB, Sigma) per ml Tris-tlCI buffer (0.05 M, pH 7.6) containing 0.01% H 2 0 2. The slides were briefly counterstained with hematox~lin, rinsed in tap water, dehydrated, and mounted in DePeX mountant (BDH). Control slides were incubated as described above, except that the antigen was omitted, and then examined for non-specific staining. To simultaneously detect specific antibodycontaining cells as well as the isotype of the antibodies, we covered the slides with antigen suspension for 45 rain at 4°C (1/5 diluted in Tris-HC1/PMSF) and then incubated them with a mixture of rabbit polyclonal antiserum and mouse monoclonal antibodies in an appropriate concentration for 45 rain. Mouse monoclonal antibodies specific for chicken lgM (HIS-C12), chicken lgG (CVI-ChIgG-47.3), and chicken IgA (CV1-ChIgA.46.5) were used (Jeurissen et al., 1988). After rinsing in PBS, the slides were covered with alkaline phosphatase-conjugated goat anti-rabbit lg (Calbiochem Behring Diagnostics, La Jolla, CA) for 45 rain. They were then rinsed and covered with peroxidase-conjugated rabbit anti-mouse Ig for 45 min. The slides were rinsed again and alkaline phosphatase activity was developed with 0.12 mg naphthol AS-MX phosphate (Sigma) and 0.25 mg Fast Blue BB Salt (BDH) per ml Tris-HCl buffer (0.2 M, pH 8.5, 37°C). The slides were treated for peroxidase activity with 0.2 mg 3-amino-9-etbylcarbazole (AEC, Sigma) per ml sodium acetate buffer (0.05

M, pH 5) containing 0.01% H202 and finally mounted in Aquamount (BDH). Control slides were incubated as described above. Charactetlzation o f antigen

The detergent extracts of sonicated sporulated oocysts were analyzed on a 10% non-reducing radium dodecyl sulfate-polyaerylamide gel (SDSPAGE) according to the method of Laemmli (1970). Stained protein molecular weight standards (206 kDa, 111 kDa, 71 kDa, 44 kDa, 29 kDa, 18 kDa, 15 kDa, Bethesda Research Laboratories, Bethesda, MD) were used to determine approximate size. Electrophoresis was performed at 40 mA. The slab gels were used for silver staining to detect proteins and for immunoblot analysis. After SDS-PAGE, the slab gel was equilibrated for ! h in transfer buffer made of 25 mM Tris, 190 mM glycine, 20% (v/v) methanol, and 0.1% SDS (pH 8.4). Proteins in the gel were transferred eleetrophoretica[ly to nitrocellulose paper (BAg5, 0.45 /zm, Schleicher & Schuell, Dassel, Germany) in a Trans-Biot cell (Bio-Rad Laboratories, Richmond, USA). EIectrophoresis was performed with transfer buffer at room temperature for 18 h at 40 mA. After transfer of the proteins, excess binding sites on the nitrocellulose paper were blocked by washing the paper in PBS

(pH 7.0) containing 0.05% Tween 80, and 0.87 M NaCI, and 4% (v/v) equine serum (ES, CVI, Lelystad, Netherlands) for 30 rain, The blots were then treated with rabbit antiserum, E.TEN 110-2, and E.TEN I1P-2 for 1 h at room temperature. The blots were rinsed three times in PBS for 5 min each time, and then exposed to peroxidaseconjugated goat anti-rabbit Ig or peroxidase-con. jugated rabbit anti-mouse Ig in appropriate concentrations. The rabbit antiserum, mouse monoelonal antibodies, and peroxidase conjugates were diluted in PBS (pH 7.0) containing 0.05% "l'~veen 80, and 0.87 M NaCI, and 4% (v/v) ES. The blots were rinsed three times in PBS and treated for peroxidase activity with 3-amino-9-ethylcarbazole (AEC).

Results To develop an in situ immunocytochemical staining technique to detect cells containing antibodies specific for E. teneUa antigens, we intravenously inoculated chickens with a suspension of sonieated sporulated oocysts and examined cryostat sections of spleen tissue. No cells containing antibodies specific for E. tenella were detected in tissue sections after the suspension of sonieated

Fig. 1. Chickenspleenafter intravenousimmunizationwith a suspensionof sonicated sporulatedoocyst antigen,a: in additionto the $ranuloc,/lescontainingendogenousperoxidase, cellscontainingantibodiesspecificfor E. tenella antigens(arrows) are visible after incubationwith TX-100(x 250). Insert shows magnificationof a cell containingantibodiesspecific for E. tenelta antigens t x 400). b: immunecomplexesspecificfor E. t¢oella on the surface of folliculardendriticcellsin a germinalcenter after incubation with NP-40 ( ~<250).

195 sporulated oocysts was used. In contrast, such cells were detected after detergent extracts of such oecysts were used (Fig. 1). We used the detergents NP-40 or TX-100 at concentrations of 0.1%, 0.5%, 1.0% and 2.0% and found that cell detection was optimal when the concentration of the detergent was 0.1%. Diluting the detergent extracts resulted in a weaker staining pattern; this 'edieated that the staining was not non-specific out antigen-dependent. The optimal dilution for staining was 1/5. The detergent extract of sonicated sporulated oocysts contained 0.2 mg protein/ml. Specific antibody-containing cells were stained with rabbit antiserum as well as with the monoelonal antibodies E.TEN 11G-2 and E.TEN 11P2, although fewer cells were stained by the monoelonal antibodies than by the rabbit antiserum. Cells containing antibodies specific for E. tenella were detected in the red pulp of the spleen, often near large arteries. Ceils with strong endogenous peroxidase activity and polymorphie nuclei were considered to represent granulocytes, and they were easy to distinguish from ceils containing antibodies specific for E. tenella. Specific E. tenella antibodies were also detected in germinal centers. After complete deletion of the antigen, neither cells containing antibodies specific for E. tenella nor germinal center staining were detected. lsotype determination Antibodies specific for E. tenella were double stained with HIS-C12 (anti-lgM), CVI-ChlgG47.3 (anti-lgG), and CVI-ChIgA-46.5 (anti-lgA) in order to determine the isotype of the antibodies. Blue staining indicated antibodies specific for E. tenella, while red staining indicated IgM, IgG or lgA isotype. Double staining, which appeared violet, indicated E. tenella.specific antibodies of the particular isotype. E. tenella-speeific antibodies detected in the chickens inoculated once only were of the IgM isotype. E. tenella-specifie antibodies in chickens which had been inoculated twice were of the lgM or lgG isotype and were detected in about equal numbers. Double-stained cells were also detected in germinal centers of the spleen. Double staining the spleen sections with CVI-IgA-46.5 resulted in red and blue ceils,

hut no violet cells, demonstrating that antibodies specific for E. tenella in the spleen were not of the IgA isotype. Oral infection To investigate whether the present detection technique could also be used in chickens infected via the natural route, we orally infected chickens with E. tenella oocysts. In the ceca of chickens inoculated once only, which were exsanguinatcd after 7, 10 and 14 days, many remnants of E. tenclla oocystg were recognized by the rabbit antiserum. In addition, few antibody-containing cells were detected. Therefore, in a second experiment chickens were inoculated three times to reduce these remnants and to increase the number of cells containing antibodies specific for E. tenella. Such ceils were detected in the lamina propria of the cecal tonsils. Specific E. tenella antibodies were also detected in the germinal centers of the cecal tonsils. Surprisingly, cells containing antibodies specific for E. tenella were also detected in the red pulp of the spleen. These cells resem-

MW

Fig. 2. SDS-PAGEanalysisof detergent extracts of sonicated sporulated oocysts stained with silver. The detergents used were 0.1% TX-IflOand 0.1% NP-~O. Numbers on the left indicate majorband sizes.

iq6

IdW

r

"-!"

12

34

Fig. 3. lmmunoblotanalysisof detergent extractsof sonicaled sporulated ooCysls.The sonicated sporulated oocyst SUSpellsion was incubatedwith 0.1% NP-40. Proteinswere resolved by SDS-PAGE,transferred to a nitrocellulosesheet and incubated with variousantibodies, peroxldase conjugate and the color development subslrate. The antibodies used for the immunohlot were: lane 1, rabbit antiserum; lane 2, E.TEN llG-2; lane 3, E.TEN lIP-2; lane 4, immunechicken serum. Numberson the left indicatemajorhand sizes.

bled plasmablasts and contained only a small amount of cytoplasmic antibody. No specific E. tenella antibodies were detected in the duodenum or in the bursa of Fabricius.

Characterization o f antigen The two detergent extracts used in the study were analyzed by SDS-PAGE and visualized by silver stain. Both detergents extracted from the sonicated sporulated oocyst suspension proteins of virtually the same weight (Fig. 2). Likewise, batches of antigen suspension did not differ (data not shown). Two major protein bands with molecular weights of 45 kDa and 100 kDa were detected and also various minor bands with molecular weights ranging between 16 kDa and 200 kDa. Using the immunoblot technique, we tested two different mouse monoclonal antibodies and a rabbit antiserum (Fig. 3). The rabbit antiserum

reacted with protein bands of molecular weights ranging between 16 kDa and 200 kDa. The reaction was strong with proteins of 20 kDa, 24 kDa, 45 kDa, and 100 kDa. Monoelonal antibody E.TEN 11G-2 reacted with two protein bands of approximately 20 kDa and 24 kDa, while monoclonal antibody E.TEN 11P-2 reacted with proteins ranging between 80 kDa and 100 kDa, To detect which antigens were also recognized by antibodies specific for E. tenella in the serum of infected chickens, we tested immune serum collected from chickens infected at 6 weeks of age. The chickens were infected six times at 3 day intervals with 10,000 oocysts of E. tenella and killed 7 days after the last immunization. The antibodies reacted strongly with prote!ns of high molecular weight (100 kDa and 113 kDa) but also with proteins of low molecular weight. No proteins were recognized by antibodies in normal serum collected from coeeidla-free chickens reared and maintained in isolators.

Discussion We have developed a three-step immunocytochemical method for the in situ detection of cells containing antibodies specific for E. tenella antigens. The method is based on the binding of antigens to antibodies specific for E. tenella present in cryostat sections of chicken tissue and the recognition of the bound antigens by rabbit antiserum or mouse monoelonal antibodies specific for E. tenella. The chickens were immunized with a suspension of sonicated sporulated oocysts, and a detergent extract of such oocysts was used as a source of antigen in the immunoenzyme staining technique. The choice of the detergents was based on their ionic properties. Non-ionic detergents are less denaturing than ionic detergents and are used to remove proteins from a membrane without disrupting protein-protein interactions. Nonionic detergents of the polyox3'ethylene type, such as NP-40 and TX-100, preserve the antigenicity as well as the charge properties of the native protein (Johnstone and Thorpe, 1987). The proteins can then be identified with specific monoclonal antibodies and separated by conventional charge-based techniques (Labota ct al., 1988).

197

Cells stained most vividly when the suspension of sonicatcd sporulated oo_~sts was incubated with 0.1% detergent. Dilution of these detergent extracts resulted in a weaker staining pattern, indicating that the technique was specific. Furthermore, after removal of the antigen no antibodycontaining cells were detected. In general, immune responses towards systemic antigens develop identically in chickens and mammals (Tempells, 1988). Because chickens lack lymph nodes the spleen is the primary organ where immune responses towards systemically administered antigens develop. After chickens were intravenously inoculated with a suspension of sonicated sporulated oocysts, cells containing antibodies specific for E. tenella antigens were detected after staining with rabbit antiserum, and also with mouse monoclonal antibodies specific for different parasitic proteins. These cells were detected outside the peri-cllipsoid lymphocyte sheaths (PELS) in the red pulp of the spleen. In a previous study using a TNP-specific detection technique, after a single intravenous injection of KLH-TNP, all plasma cells containing antibedics against TNP were located outside the PELS areas in the red pulp (Jeurissen, 1991). Some antiTNP-specific lymphoblasts were detected in the outer border of the PELS. These results agree with the results of others who have found that B lymphocytes of the PELS form the basis of the plasma cellular reaction in the red pulp (Ogata et al., 1981). In germinal centers, E. tenella.specific antibodies stained in a cobweb pattern. These antibodies represented immune complexes on the surface of follicular dendritic cells, as previously shown by Kroese et al. (1982). Follicular dendritic cells trap immune complexes in germinal centers, where they are preserved long after the initiation of an immune response. The most important function of antigen preservation is the generation of memory B cells (Klaus et al., 1980). In our study, the antigens in the immune complexes could not be demonstrated after incubation with the rabbit antiserum only. As this result contradicts that of Van den Dobbelstcen et al. (1991), we assume that the E. tenella antigens had lost several secondary and tertiary structural features and could no longer be recognized by the rabbit antiserum.

Much staining was seen in cecal tissue sections obtained from chickeas which were inoculated orally once only and used after 7, 10 and 14 days due to remnants of E. tenella in the moco~a. The distinction between this specific staining of remaining parasites and specific antibody-containing cells is possible, but difficult. In addition, few cells contained antibodies specific for E. tenella. Therefore, we used immune chickens which contained virtually no remnants of E. tenella. When these chickens were infected, cells containing antibedies specific for E. tenella were detected in the cecal tonsils and spleen. In the spleen, these cells resembled plasma blasts and contained only a small amount of cytoplasmic antibody, probably because the chickens were used 4 days after challenge. The spleen may have contained such cells because the parasites can travel extra-intestinally and cause a systemic response (Fernando et al., 1987; Rose and Hesketh, 1991), or because antigen-specific plasma blasts, which arc induced in the ccca, migrated to the spleen before becoming plasma cells (Jeurissen et al., 1985). Oral inoculation of chickens with sporulated oocysts may cause antigen to be degraded by enzymes in the digestive tract. The immune response may or may not have been directed towards epitopes that were also expressed in the detergent extract of sonicated sporulated oocysts, which was used as antigen in the staining technique. After passing through the digestive tract, however, at least some epitopes remain similar to epitopes in the detergent extract of sonicated sporulated oocysts. Intravenous inoculation of chickens with a suspension of sonicated spornlated oocysts induced numerous cells containing antibodies specific for E. tenella in the spleen, suggesting that the immune response was directed towards the same epitopes expressed in the inoculated antigen suspension. In order to determine which polypeptides were present in the detergent extract of sonieated sporolated oocysts, we characterized the antigens by immunoblotting. The finding that the monoclonal antibodies recognized only a few of the proteins recognized by the rabbit antiserum correlated with the finding that fewer cells containing antibodies specific for E. tendla were detected in the tissue sections after using mono-

198 clonal antibodies. The target antigens of the monoclonal antibodies E . T E N 1 ]G-2 and E . T E N 1 IP-2 arc a sporozoite surface protein of respectively 25 kDa (Vermeulen, anpublished results) and a high-molecular-welght protein of between 90 kDa and 100 kDa known to be present in micronemes of sporozoites and merozoites of E. tenella (Tomley el al., 1991 in press). The use of non-ionic detergents for isolating coccidial antigens has already been shown to be successful Karkhanis et al. (1991) measured the ability of parasitic extracts to confer protection and purified the protective antigens. T h e y prepared antigens from sonicated sporulated oocysts and from sonicated sporozoites of E. teriella. T h e extract of sporulated oocysts contained a single predominant polypeptide with a molecular mass of 26 k D a as well as smaller polypeptides. T h e Zwittergent extract of sporozoites contained a 22 k D a polypeptide that was protective. Files ¢t al, (1987) also described the isolation of a 23 kDa surface protein of E. tenella that provided partial protection against oocyst challenge. T h e immunocytochemical method for the in situ detection of cells containing antibodies and immune complexes specific for E. tenella can be used to simultaneously detect these cells and the isotype of the antibodies. Moreover, using this technique, we can detect cells in situ that contain antibodies specific for antigens of sonicated sporulated oocysts, which are also important for the humeral response in naturally infected chickens. T h e technique offers the possibility of detecting humeral responses to single parasitic polypeptides in situ, T h e developmental stage of the parasite used to prepare antigen can be varied, in order to study the importance of these stages and the antigen can then be characterized with immunobiotting techniques. T h e technique will be of use for further studies of the local humeral immune response to E. tenella in the chicken mucosa.

Acknowledgements W e thank Sjoerd V e l d k a m p and M a r g a Janse for their expert technical assistance.

References Cluassen~ E. and Van Rooijen, N, (1984) TNP-enzyme conjusates For the detection of anti-TNP antibody containing cells in vivo. J. lmmunol. Methods "/5, ISl. Davis, P.J. (1981). Immunity 1o coccidia. In: M,E. Rose. L.N.

Payne and B.M. Freeman (Eds.), Avian Immunology. British Poultry Science Ltd., Edinburgh, p. 361. Davis, P.J., parry, S.H. and Porter, P. (1978) The role of secretory IgA in anti coecidial immunity in the chicken. Immunology 34, 879. Fernando, M.A., Rose, M.E. and MiHard, B.J. (1987) Eimeria spp. of domestic fowl: the migration of sporozoites intraand exlra-emerically. J. ParasitoL 73, 561. Files, J.G., Paul, L.S. and Gabe, LD. (1987) Identification and characterization of the gene for a major surface antigen of Eimeria tenella. ]n: N. Agahian, M. Goodman and N. Nogueira (Eds.), Molecular Strategies of Parasite Invasion. Alan R. Liss, New York, p. 713. Jeurissen. S.H.M. (1991) Structure and function of the chicken spleen. Res~ lmmunol, 142, 352. Jeurissen, S.H.M., Claassen, E., Van Rooijen, N. and Kraal, G. (1985) lnlra-intestinal priming leads to antigcn-spacific ]gA memory cells in peripheral lymphoid organs. Immunology 56, 417. Jeurissen, S.H.M., Sminia, T. and Beuvery, E.C. (1987) Induction of a mucosal IgA immune response with preparations of porin proteins of Ne£vserla gonorrRoea¢. Infect. Immun. 55, 253. Jeurissen, S.H.M., Jansc, E.M., Ekino~ S., Nieuwenhuis, P., Koch, G. and De Boer, G.F. (1988) Monoclonal antibodies as probes for defining cellular subsets in the bone marrow, thymus, bursa of Fabricius, and spleen of the chicken. Vet. Immunol. Patbol. 19, 225. Jeurissen, S.H.M., Janse, E.M. and Vermeulen, A.N. (1989) Lymphoid and non-lymphoid cells in ;he caecal mucosa of naive and immune chickens related to the development of Eimeria tenella. In: P. Ivore (Ed.), Coccidia and Intestinal Coecidiomorphs. INRA, Paris, is. 551. Johnstone, A.P. and Thorpe, R. (1987) In: Immunochemistry in practice. Blackwell Scientific Publications, Oxford, p. 105.

Karkhanis, Y.D., Nollstadt, K.A., Bhogal, B.S., Ravine, O., Peliegrino, R., Crane, M.S., Ivlurray, P.K. and Turner. M.J. (199l) Purification and char,qcterization of a protective antigen from Eimeria zenella. Infect. lmmun. 59, 983. Klaus. G.G., Humphrey, J.H., Kunkl, A. and Dongwort, D.W. (1980) The follicular dendritic cell: its role in antigen presentation in the generation of immunological memory. ImmunoL Rev. 53, 3. Kroese, F.G.M., Eikelenboom, P. and Van Rooijen, N. (1982) Electronmieroscopical and enzymehistochemical characterization of immuge complex trapping cells in the spleen of different vertebrate species. Dev. Comp. ImmunoL 2 (suppl.), 69. Labeta, M.O., Fernandez, N. and Festenstein, H. (1988)Solubilisation effect of Nonidet P-40. Triton X-100 and CHAPS

199 in the detection of MHC-like glycoproteins. J, ImmunoL Methods 112, 133. Laemmli. U.K. (1970) Cleavage of structural proteins during Ihe assembly of Ihe head of bacteriophage T4. Nature 27% 680. [xmg. P., Millard, B,J., Joynt:r. L.P. and Nortt~n. C.C. (197fi) A guide 1o lahoraloty techniques used in the study and diagnosis of avian coccidiosis, Folia Vet. Lal. 6, 2(}1. Lowl'J, O.H., Rosebrough. N.J., Farr. A.L. and Randall. R J . (1951) Protein measurement with the fidin phenol reagent. J. Biol. Chem. 193. 969. Nash, P.V. and Speer. C.A. (Iqgg) B-Lymphocyte responses in the large intestine and mesenteric !ymph nodes of mice infected with Eimeria [alcifonnis (Apieomplexa). J, ParasitoL 74. 144. Ogata. K,. Kitawa, H. and Kudo, N. (1981) Formation of two types of germinal centers during immune response in chicken spleen. Jpn. J. Vet. Sci. 43. 645. Rose, M.E. and Heskcth, P. (1987) Eimeriu re,ella: Effects of immunity on sporozoites within the lumen of the small intestine. Exp. Parasito]. b3, 337. Rose, M.E. and Hesketh, P. {1991} Eimeria tern,ira: localization of the sporozoites in the cecum of the domestic fowl. Parasitology 102, 3i7. Sminia, T.. D~lemarre, F. and Janse, E,M. (1983) Histological observations on Ihc intestinal immune respunse towards horse radish peroxidase in rats. Immunology 511.53.

Steven, W.M. (19801 Effects of antigen dosage on early localization of specific antibodies in rat splenic germinal centers. Anat. Rec. 198, 503. Straus, W. (19681 Cytochemica| demonstration of sites of antibody to horseradish peroxidas¢ in spleen and lymph nodes, J. Histcchem. Cyzochem. lf~. 237. Tempelis. C.H. 09881 Ontogeny and regulation of the immune response in the chicken, Prop,, Vet. MicrobioL Immua. 4, I. Tomlcy, F.M.. Clark, L.E. Ka,,~azoe. U., Dijkcma, R. and Kok. J.J. {1991| Sequence of the gene encoding an immunodominan! microneme protein of Eimeri~ tenella. MoL Biochem. Parasi:ol. 49, 277. Van den Dobbelslcen. G.PJ.M., Van Rooi ien, N.. Sminia, T, and Van Rues, E.P. (199l) The immune response in rat to $trL,plVCOCCn.~ pne,moniae type 3 and type 4 capsular polysaccharide. Detection by double immunocytochemical staining of antibody-containing cells in silu and ELISA, J. Immunol. Methods, 145.93. Van Rooijen, N. and Claassen. E. (19~6t Recent advances in the detection and characterization of specific antilyody-forruing cells in tissue section. Histochem. J. 18. 465. While. R.G.. French. V.I. and Stark. J.M. (lq70) A study of the localization of a protein antigen in the chicken spleen and its relation t,) the formation cf g~rmina[ centers. J. Mud. Microhiol. 3, 65,