Veterinary Parasitology, 36 (1990) 91-103 Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands
Detection of Taenia hydatigena Copro-Antigens by ELISA in Dogs P. DEPLAZES, B. GOTTSTEIN, Y. STINGELIN and J. ECKERT
Institute of Parasitology, University of ZFtrich, Winterthurerstrasse 266a, CH-8057 Zi~rich (Switzerland) (Accepted for publication 2 November 1989)
ABSTRACT Deplazes, P., Gottstein, B., Stingelin, Y. and Eckert, J., 1990. Detection of Taenia hydatigena copro-antigens by ELISA in dogs. Vet. Parasitol., 36: 91-103. A sandwich-ELISA was developed for the detection of soluble Taenia hydatigena antigens in fecal samples of dogs. Affinity-purifiedpolyclonal catching antibodies and alkaline phosphataseconjugated detecting antibodies were employed, which had been obtained from rabbits hyperimnmnized with excretory/secretory antigens derived from in vitro maintained adult Taenia hydatigena. The assay allowed the detection of 800 ng T. hydatigena antigen g- 1 of feces as a lower limit. Six helminth-free dogs were each infected with 10 T. hydatigena cysticerci isolated from Swiss sheep. After prepatent periods ranging from 57 to 71 days, the dogs started to excrete Taenia eggs and/or proglottids. The ELISA detected Taenia antigens in all six dogs during the prepatent period starting individuallybetween Day 18 and 45 post-infection (p.i.). Anthelmintic treatment of three dogs at Day 95 p.i. resulted in elimination of the cestodes and within the 5 following days in the disappearance of Taenia antigens from feces. The specificity of the assay was evaluated by testing crude antigens derived from helminths or bacteria. Four Taenia species showed cross-reactivity at concentrations of 5 zg protein m1-1. Conversely, no cross-reactions occurred with various antigen batches derived from Echinococcus granulosus, E. multilocularis, Dipylidium caninum, Mesocestoides corti, Diphyllobothrium sp., Tuxocara canis and bacterial antigens (Salmonella and Escherichia). Moreover, fecal samples from dogs naturally infected with 71. canis (n: 13), hookworms (n: 2), Trichuris vulpis (n: 13) and of 10 dogs with mixed infections with these three nematode groups were tested, and results confirmed the high degree of specificity. The Taenia antigens detectable by this ELISA remained immunologicallystable in native feces stored at + 25 °, + 4 ° or at - 2 0 °C for at least 5 days.
The parasitological diagnosis of intestinal cestode infections in dogs is often inaccurate as the parasites are not readily detectable during prepatency and because of irregular excretion of proglottids and/or eggs in feces during pat0304-4017/90/$03.50
© 1990 Elsevier Science Publishers B.V.
P. D E P L A Z E S E T AL.
ency. To improve the diagnostic sensitivity of fecal examinations for Echinococcus granulosus in control programs, arecolin purging of dogs with subsequent intensive fecal examination has been widely used (WHO, 1984). However, the accuracy of this technique is variable (WHO, 1984), and only about 50% of the infected dogs can be identified (Rickard and Arundel, 1985 ). Dogs artificially infected with Taenia hydatigena excreted highly variable numbers of proglottids per day (0-55) during patency; only 36% were eliminated with feces but 64% without defecation (Deplazes and Eckert, 1988a). In the same study of the detection of Taenia eggs, anal-skin swabs proved to be more reliable than examination of fecal samples using a flotation technique (Deplazes and Eckert, 1988a). These examples of the variable accuracy of parasitological techniques underline the need for diagnostic improvement by other means. Improvement may be achieved by the employment of immunodiagnostic methods, for the detection of specific antibodies in serum samples or of cestode antigen in feces (copro-antigen). Specific antibodies are detectable in sera of dogs experimentally or naturally infected with Taenia spp. or Echinococcus granulosus (Heath et al., 1985; Jenkins and Rickard, 1985, 1986a,b; Gasser et al., 1988). The application of this principle to field conditions in Australia has shown that 73% of 22 dogs infected with 300-302 600 E. granulosus could be identified by an enzymelinked immunosorbent assay (ELISA) using homologous protoscolex somatic antigen. Antibodies directed against antigens derived from E. granulosus oncospheres were demonstrable in 52% of 21 dogs. The sensitivity of this test was superior to the results of arecolin purging and the specificity was high enough to exclude cross-reaction with Taenia species and common nematodes on a reliable level (Gasser et al., 1988). Although further improvement of this technique can be expected in the future, some basic problems may limit its application; early cestode infections may be missed as seroconversion occurs about 2 weeks after infection or later. On the other hand antibodies may persist for long periods even after elimination of the parasites. It can be assumed that excretion of cestode antigens in feces of hosts may be directly correlated with the presence of intestinal stages of Taenia, Echinococcus ( --taeniid cestodes) or other cestode genera. On the other hand, parasite elimination should be directly followed by the cessation of antigen release. The diagnostic detection of virus antigen in stool samples of humans by ELISA was first established by Yolken et al. (1977). Similar assays have been employed in the diagnosis of Giardia lamblia and Entamoeba histolytica infections in humans (Grundy, 1982; Ungar et al., 1984, 1985; Baumann and Gottstein, 1987; Rosoff and O'Hanley, 1988). This paper describes an ELISA for the detection of cestode copro-antigen, using Taenia hydatigena in dogs as a model. The presented approach might also be applicable to infections with other Taenia spp., Echinococcus granulosus and E. multilocularis in carnivores as well as to Taenia saginata and T. solium infections in man.
DETECTION OF T. H Y D A T I G E N A COPR0-ANTIGENS
Experimental design A sandwich-ELISA was developed for the detection of Taenia hydatigena antigens in feces of dogs and was evaluated by the following steps. (a) Determination of the sensitivity of the assay in experiments using test solutions with known protein amounts of T. hydatigena metabolic antigens (excretory/secretory ( -- E / S ) antigens ), and similar determinations but with the inclusion of fecal material to determine its influence on the sensitivity. (b) The evaluation of specificity with somatic antigens derived from various cestode and a nematode species, from bacteria (Salmonella and Escherichia) and from other protein sources (see Fig. 2). (c) Identification of the threshold in the ELISA for discriminating between "positive" (cestode copro-antigen detectable) and "negative" (no cestode copro-antigen detectable). (d) Examination of the potential influence of different storage conditions of fecal samples prior to testing the sensitivity of the ELISA and determination of the reproducibility of the assay. (e) Definition of some test characteristics and parameters by using fecal samples from dogs experimentally infected with T. hydatigena.
Animals The following groups of dogs were used in our investigations. (1) Group 1: Sixteen female and four male dogs ("Niederlaufhunde", 6 months-6 years old) were kept under helminth-free conditions. Fecal samples from the dogs of this group were used for steps c-e. (2) Group 2: Four female and two male dogs ("Niederlaufhunde", all 6 months old) had been raised under helminth-free conditions. For the study they were artificially infected with metacestodes of T. hydatigena as previously described (Deplazes and Eckert, 1988a). Each dog received 10 metacestodes freshly isolated from slaughtered sheep. For these dogs the prepatent period was determined by daily examination for proglottid and egg excretion using the methods described by Deplazes and Eckert (1988a). (3) Group 3: Thirty-seven dogs of various ages and breeds. The dogs were examined for intestinal parasites by routine coprological procedures. Thirteen dogs had patent infections with Toxocara canis, two an infection with hookworms and 12 with Trichuris vulpis. Ten dogs had mixed infections with the three species mentioned above. The fecal samples from these dogs were used for step (e). The dogs of Groups 1 and 2 were housed according to the legal regulations of the "Swiss Animal Protection Codes" under conditions preventing access
P. DEPLAZES ET AI,.
to infective stages of helminths. The animals received commercial canned or dried dog food and water. The dogs of Group 3 were privately owned and maintained under usual local conditions.
Antigens Taenia E/S-antigens Living adult, gravid tapeworms were obtained from dogs experimentally infected with T. hydatigena and subsequently treated with Nemural ® according to Deplazes and Eckert (1988b). Taenia saginata was obtained from a human patient spontaneously excreting the tapeworm. The worms were abundantly washed with sterile PBS and individually maintained in 75-cm 2 tissue culture flasks (Corning Glass Works, No. 25110) each containing 150 ml Eagle's Minimal Essential Medium (EMEM) (Gibco Co., No. 072 1100) with D-glucose ( 12 mg m l - ' ), gentamycine (200 ]~g ml- 1) and Fungizone ® (250 ng m l - 1) at pH 7.2. The medium was replaced after 4, 18 and 28 h of incubation, then every 2 days until Day 21 (T. saginata until Day 6). The collected media (from Day 3 to 21) were stored at - 2 0 ° C until further processing. The viability of the worms was judged by their motility. The sterility of the cultures was tested on Days 3 and 13 of maintenance according to standard procedures. E/S-antigens were prepared by concentration of the collected culture media using an Amicon ultrafiltration unit and a YM-10 membrane, followed by dilution with PBS and reconcentration to a final concentration of 5 mg protein ml- ' solution. All protein concentrations were assessed by a Bio-Rad protein assay and with bovine albumin as a standard. The T. hydatigena E/S-antigen was used for the elaboration of the antigen-detecting ELISA (see rabbit antiT. hydatigena E / S hyperimmunoglobulins).
Helminth somatic antigens Somatic antigens prepared from adult and larval stages of various helminth species were used for the evaluation of the specificity of the test system. The individual helminth species and their hosts are listed in Fig. 2. All freshly obtained helminth material was abundantly washed with sterile PBS prior to processing. Somatic antigens (crude extracts) were prepared in the following way: worms were homogenized in a Polytron PCU-2 blender and frozen/thawed for three times using liquid nitrogen and a + 37 °C water-bath. Subsequently the material was ultrasonicated (60 s, 40 W, 80% pulse) and sedimented at 10 000 g for 30 min at + 4 ° C. Supernatants contained the somatic antigens.
Other antigens In addition, crude extracts processed in the same way as described above ("helminth somatic antigens") were obtained from Salmonella typhimurium
DETECTION OF 7: HYDATIGENA COPRO-ANTIGENS
strain LT2 MIC and from Escherichia coli strain Y 1089 (Gottstein et al., 1990). These antigens and the following proteins were used for background determinations in the ELISA: bovine albumin (Fluka AG No. 05480) and egg ovalbumin (Sigma Chem. Co., No.A-5503 ), dog milk was obtained from a bitch 5 days post-partum.
Fecal samples Fecal samples were collected from dogs of Group 2 prior to infection and in intervals of 7 days p.i. All samples were immediately frozen and stored at - 20 ° C until further processing. A separate amount of each sample was examined for helminth eggs by means of a sedimentation/flotation technique and microscopy (Boch and Supperer, 1983). All test specimens and all specimens obtained from dogs of Steps (c), (d) and (e) were prepared as follows. To the fecal samples, PBS containing 0.02% NAN3, 0.05% bovine hemoglobin (Fluka AG, No.51290) and 0.3% Tween 20 ( = " s a m p l e buffer") was added in a 1:4 ratio. This suspension was ultrasonicated for 30 s (40 W) and subsequently sedimented at 3000 g (at room temperature) for 15 min. The supernatant was used for testing by ELISA. Three of the six dogs in Group 2 were treated with a therapeutic dose of niclosamide 95 days after infection for removal of the cestode burden. The three other dogs were killed after 104 days.
Rabbit anti-T, hydatigena E/S hyperimmunoglobulins The production of rabbit hyperimmunoglobulins was carried out as described by Baumann and Gottstein (1987), but employing T. hydatigena E/Santigen for immunization. All other steps including the coupling of T. hydatigena E/S-antigen to CNBr-activated Sepharose 4B, the purification of antibodies by affinity-chromatography, the preparation of rabbit anti-T, hydatigena E/S-conjugates (alkaline phosphatase) and the preparation of normal rabbit IgG for control reactions were done according to Baumann and Gottstein (1987).
ELISA The polystyrene surface of alternate rows of wells on flat bottomed microELISA-plates (Nunc Immunoplates I ) were sensitized by incubation with 150 /~l each of a solution (0.1 M NaHCOffNa2CO3+0.02% NAN3, pH 9.6) containing either affinity-purified rabbit anti-T, hydatigena E/S-antigen-antibodies ("catching antibodies") or normal rabbit IgG ( control ) at concentrations of 10/~g m1-1 (overnight incubation at + 4 ° C ) . The further processing for testing fecal samples correspond to that described previously (Baumann and
P DEPLAZESET AI,.
Gottstein, 1987). The reading of the test was carried out by using a Dynatech MR710 Microplate Reader combined with a Macintosh II computer (software: Immunosoft ® from Dynatech). For the evaluation of potential cross-reactivity and nonspecific reactions in the ELISA, antigens of various helminths and bacteria (Fig. 2) were tested at a dilution of 5 ttg protein ml 1, bovine albumin and ovalbumin were tested at a concentration of 500 zg protein m l - ]. Dog milk was tested at a 1 : 1 dilution. RESULTS
The results are presented according to the scheme described under experimental design.
Sensitivity (Step a) Serial twofold dilutions of T. hydatigena E / S - a n t i g e n were assessed either in buffer or in feces by ELISA (Fig. 1). The threshold (cut-offpoint) for discriminating a negative from a positive reaction was determined by calculating the m e a n A405 .... value (2) of 20 fecal specimens originating from 20 helminthfree dogs plus 3 SD. According to this calculation the test system showed a sensitivity of detecting at least 200 ng T. hydatigena E/S-antigen ml-1 of a diluted ( 1 : 4 ) fecal sample, equivalent to 800 ng g - 1 feces. All control reactions with normal rabbit-IgG were very low ( A 4 o 5 n m < 0 . 0 2 ) ; therefore it could be E