- Email: [email protected]
Evaluation of ELISA and Western Blot Analysis using three antigens to detect anti-Trichinella IgG in horses
Veterinary Parasitology 108 (2002) 163–178
Evaluation of ELISA and Western Blot Analysis using three antigens to detect anti-Trichinella IgG in horses Edoardo Pozio a,∗ , Ljiljana Sofronic-Milosavljevic b , Maria Angeles Gomez Morales a , Pascal Boireau b , Karsten Nöckler c a
Laboratory of Parasitology, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy b UMR 956 INRA-AFSSA-ENVA, Biologie Moléculaire et Immunologie Parasitaires et Fongiques, 22 rue Pierre Curie, 94703 Maisons Alfort, France c Federal Institute for Health Protection of Consumers and Veterinary Medicine, Diedersdorfer Weg 1, 12277 Berlin, Germany Received 18 January 2002; received in revised form 27 May 2002; accepted 10 June 2002
Abstract We assessed a serological method for detecting Trichinella infection in horses, specifically, an ELISA using three antigens to detect anti-Trichinella IgG (i.e. a synthetic tyvelose glycan–BSA (stg-BSA) antigen, an excretory/secretory (ES) antigen, and a crude worm extract (CWE) antigen). Serum samples were collected from 2502 horses (433 live horses from Romania and 2069 horses slaughtered in Italy and originating from Italy, Poland, Romania, and Serbia). Serum samples were also taken from horses experimentally infected with different doses of T. spiralis and T. murrelli larvae, as controls. The cut-off value of ELISA was determined on serum samples from 330 horses from Trichinella-free regions of Italy, which were also examined by artificial digestion of preferential-muscle samples. In the experimentally infected horses, the stg-BSA and ES antigens were less sensitive than the CWE antigen. Trichinella spiralis showed a higher immunogenicity than T. murrelli, and the IgG immunoresponse was dose-dependent. The kinetics of anti-Trichinella IgG were similar among all experimentally infected horses. No circulating antibodies were detected 4–5 months after experimental infection, although these horses still harbored infective larvae. Depending on the antigen used, for 4–7 of the 330 horses from Trichinella-free areas, the optical density (OD) of the serum sample was higher than the cut-off value, yet these samples were negative when subjected to Western Blot. Similar results were obtained for the 1739 horses slaughtered in Italy (originating from Italy, Poland, Romania, and Serbia) and the 433 live Romanian horses. Of the 4 horses with
∗ Corresponding author. Tel.: +39-06-4990-2304; fax: +39-06-4938-7065. E-mail address: [email protected] (E. Pozio).
0304-4017/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 0 1 7 ( 0 2 ) 0 0 1 8 5 - 1
164
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
muscle larvae, only one was positive by ELISA and Western Blot. Because the anti-Trichinella IgG remain circulating for only a short period of time, whereas the larvae remain infective for longer periods, serology cannot be used for either diagnosing Trichinella infection in horses or estimating the prevalence of infedction. Artificial digestion of at least 5 g of preferential-muscle tissue continues to be the method of choice at the slaughterhouse for preventing equine-borne trichinellosis in humans. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Trichinella spiralis; Trichinella murrelli; Horse; Europe; ELISA; Western Blot; Synthetic tyvelose glycan–BSA antigen; Excretory/secretory antigen
1. Introduction Trichinella infection in horses continues to represent a serious threat to animal and human health in France and Italy, the only two European countries where horse meat is frequently consumed raw (Pozio, 2001). Several diagnostic tests have been used to detect Trichinella infection in this animal, yet the artificial digestion of at least 5–10 g of muscle from the tip of the tongue seems to be the only suitable diagnostic method, although it is expensive and time consuming (Boireau et al., 2000). Since 1994, the year in which this diagnostic method was introduced as part of routine practice by EU legislation (Directive 77/96/EEC, amended by Directive 94/59/EEC), from one to two naturally infected horses have been detected in France and Italy per year, and all of these horses originated from eastern Europe. In the same period, human trichinellosis outbreaks occurred only when veterinary services failed to properly apply this method when controlling for the parasite at the slaughterhouse (Boireau et al., 2000; Pozio, 2001). To develop a serological method to detect this infection in live horses, many researches have studied the kinetics of the humoral immunoresponse against Trichinella in experimentally infected horses using excretory/secretory (ES) antigen or crude worm extract antigen in an ELISA (Smith and Snowdon, 1987; van Knapen et al., 1987; Polidori et al., 1989; Soulé et al., 1989; Gamble et al., 1996; Voigt et al., 1997). However, the results of these studies have been frustrating because the specific IgG become undetectable 4–5 months p.i. at the latest, despite the fact that the larvae remain infective in the muscle tissues, as reported for both experimentally infected (Polidori et al., 1989; Soulé et al., 1989) and naturally infected (Pozio et al., 1997) horses. Furthermore, a serosurvey carried out on horses from Romania, which is one of the European countries with the highest prevalence of domestic trichinellosis, has shown a high number of false-positive sera by ELISA, suggesting that the antigens used (excretory/secretory antigen) in the assays had a low specificity (Sofronic-Milosavljevic et al., 2001). The objective of the present study was to asses the diagnostic value of an ELISA using three antigens to detect anti-Trichinella IgG. To this end, we studied sera from both experimentally infected horses and from horses originating from several European countries with different prevalence rates of Trichinella infection and from Trichinella-free regions in Italy (Pozio, 1998).
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
165
2. Materials and methods 2.1. Serum samples We studied serum samples from eight horses bred and experimentally infected in France and eight horses bred and experimentally infected in Germany. The horses from France were infected with either T. spiralis (two animals infected with 5000 larvae) or T. murrelli (three animals infected with 5000 larvae and three animals with 20,000 larvae) (Soulé et al., 1989); the horses from Germany were infected with T. spiralis (with 5000 larvae each) (Voigt et al., 1997). Serum samples were collected from these horses beginning 1 week before experimental infection and up to 43 weeks p.i. Serum/blood samples were also collected from 2502 additional horses, which consisted of 433 live horses from Romania, which has an average prevalence of 1% of Trichinella infection in domestic pigs (Olteanu, 2001), and 2069 horses slaughtered in Italy. The horses slaughtered in Italy originated from various countries, specifically: 780 from Italy, which is free from domestic trichinellosis but which has a prevalence of vulpine infection (T. britovi) ranging from 0 to 35% (in some mountain areas) (Pozio et al., 2001); 627 from Poland, which has a 0.0004% prevalence of Trichinella infection in domestic pigs (Ramisz et al., 2001); 485 from Romania; and 177 from Serbia, which has a 0.1–0.3% prevalence of Trichinella infection in domestic pigs (Cuperlovic et al., 2001). Of the 780 Italian horses, 330 originated from Trichinella-free regions (Pozio, 1998); the other 450 horses originated from regions with different prevalence rates of sylvatic trichinellosis or from unspecified areas of Italy. After sample collection, blood was allowed to clot at room temperature. The serum was then collected and frozen at −80 ◦ C. We established the optimal serum dilution by testing serum samples from experimentally infected horses and from horses known to be negative for Trichinella at dilutions of 1:10, 1:20, 1:40, 1:50, 1:80, 1:100, 1:120, 1:150, and 1:200, using three different antigens by ELISA (see below). Using the paired Student t-test, the greatest difference in optical density (OD) between positive and negative serum samples was obtained at a dilution of 1:50 (data not shown), which was used for all subsequent ELISA. 2.2. Parasitological examinations To examine meat samples of horses slaughtered in Italy (n = 2069) for Trichinella larvae, 10 g of masseter or tongue were examined by artificial digestion (Gamble et al., 2000). To evaluate the specificity of serological analyses, the intestinal helminth fauna was studied. Fecal samples, collected directly from the rectum of 265 horses, were examined by flotation method. Positive samples were cultured at 26 ◦ C for 8 days to identify the third larval stage at the species level, in accordance with Soulsby (1965). 2.3. Antigens We used three antigens to test all serum samples: a crude worm extract antigen (CWE); an excretory/secretory antigen; and a 3,6-dideoxyhexose sugar (tyvelose), which is one of the major highly specific immunodominant epitopes of Trichinella, coupled with BSA
166
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
(i.e. synthetic tyvelose glycan–BSA (stg-BSA; Heska Corporation, Fort Collins, CO, USA) (Denkers et al., 1991; Wisnewski et al., 1993)). CWE was prepared using T. spiralis larvae collected from the muscles of experimentally infected Swiss mice by artificial digestion. After washing six times by settling in 0.1 M phosphate buffered saline (PBS), pH 7.2, larvae were stored at −80 ◦ C until use. After thawing, larvae were crushed in a glass Potter-homogenizer using a teflon pestle and further disintegrated by sonication. The larval suspension was maintained overnight at 4 ◦ C under magnetic stirring and centrifuged at 17, 200 × g at 4 ◦ C for 1 h. The protein concentration of the supernatant was determined by Bradford’s method (1976). To produce the ES antigen, the protocol of Gamble et al. (1983) was followed. Briefly, T. spiralis muscle larvae were washed three times in PBS, pH 7.2, with penicillin (500 units/ml) and streptomycin (500 g/ml). Larvae were then washed four times by settling in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with penicillin (500 units/ml) and streptomycin (500 g/ml). Five thousand worms per ml were resuspended in DMEM, supplemented with 1 M HEPES, 200 mM l-glutamine, 100 mM Na-pyruvate, and 5000 units of penicillin/streptomycin (GIBCO, Grand Island, NY), and incubated with 10% CO2 in a 75 cm2 culture flask (Costar) at 37 ◦ C for 18 h. Worms were then removed from the medium by settling in 50 ml conical tubes. The medium was filtered through a 0.2 m filter, and the supernatant was concentrated 100 times in an Amicon® (Amicon Inc., MA, USA) pressure concentrating chamber using a YM-5 filter. To determine the protein concentration and quality of the batch (i.e. without any bacterial or fungal contamination), OD was evaluated at a 280/260 nm ratio; antigens with a ratio higher than 1.2 were used. 2.4. ELISA using CWE or ES antigens An indirect ELISA method was used to detect Trichinella-specific IgG. Briefly, T. spiralis antigens (CWE or ES) were diluted in 0.1 M carbonate buffer, pH 9.6, and used at a concentration of 5.0 g/ml. Flat-bottomed polystyrene microtiter plates were coated with 100 l/well of CWE or ES in 0.1 M carbonate buffer, pH 9.6, at 37 ◦ C for 1 h. Coated plates were washed three times with 0.05% Tween 20 in H2 O. Plates were blocked with 200 l of 1.0% bovine serum albumin in PBS with 0.05% Tween 20 (PBST) at 37 ◦ C for 1 h. Plates were then washed and incubated with 100 l/well of five positive and five negative serum controls and horse serum samples at a 1:50 dilution at 37 ◦ C for 1 h. After washing as described above, 100 l/well of a goat anti-horse IgG conjugated with horse-radish peroxidase (Kirkegaard & Perry Laboratories Inc., Maryland) were added to the wells at a 1:10,000 dilution, and the plate was incubated at 37 ◦ C for 1 h. The plate was washed again, and 100 l/well of TMB Microwell Peroxidase Substrate System (Kirkegaard & Perry Laboratories, Maryland) was added. After 10 min, the reaction was stopped with 100 l/well HCl 1N, and the plate was read at 450 nm using an MRX Microplate Reader (Dynex technologies Inc., VA, USA). 2.5. ELISA using the stg-BSA antigen An indirect ELISA method was also used to detect Trichinella-specific IgG using the stg-BSA antigen. A Nunc MaxiSorpTM 96-well plate was coated with 100 l/well of the
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
167
stg-BSA antigen (1.25 g/ml) in 0.05 M Tris–EDTA–NaCl buffer, pH 7.3, at 5 ± 3 ◦ C for 16–18 h. The plate was then washed four times with PBST, and 100 l/well of PBST-diluted serum with 4% fetal bovine serum (FBS), and 0.5% Proclin® 300 (P 300) (Supelco Inc., PA, USA) was added. After 30 min incubation at 22±2 ◦ C, the plate was washed four times with PBST, and 100 l/well of a goat anti-horse IgG conjugated with horse-radish peroxidase was added at a 1:10,000 dilution in PBST, with 4% FBS and 0.5% P 300. The plate was incubated at 22 ± 2 ◦ C for 30 min. It was then washed four times in PBST and once with deionized water. Addition of the substrate and reading of the plate were done as described above. Since the synthetic antigen is conjugated to BSA, to avoid a possible reactivity of horse sera with BSA, serum samples were pre-incubated with nitrocellulose-BSA strips at room temperature for 1 h. 2.6. Cut-off values The cut-off value of ELISA was evaluated for the three antigens on the basis of the average OD plus three standard deviations (S.D.) of 330 serum samples of horses reared from birth in Trichinella-free areas of Italy, slaughtered in Italy, and without larvae of Trichinella in 10 g of masseter or tongue. 2.7. Western Blot Protein components of CWE, ES, and stg-BSA were electrophoretically separated on 8.0% sodium dodecyl sulfate polyacrylamide-gel (SDS-PAGE), according to Laemmli (1970), and transferred to nitrocellulose sheets, as performed by Towbin et al. (1979). Reference and ELISA-positive horse serum samples were diluted at 1:50, and goat anti-horse IgG with horse-radish peroxidase (Kirkegaard & Perry Laboratories Inc., Maryland), diluted at 1:3000, was used as conjugate. The immunoblot was evaluated using chemiluminescent principles (ECL Western Blotting detection reagents, Amersham Life Science, Little Chalfont, Buckinghamshire, UK) and visualized by exposing the membranes to autoradiography film for 30 s.
3. Results 3.1. Cut-off values The cut-off values detected by ELISA using the stg-BSA, ES, and CWE antigens on 330 serum samples from horses slaughtered in Italy and originating from Italian Trichinella-free areas, were 0.55, 0.92, and 1.02, respectively (Fig. 1). Four (1.2%), five (1.5%), and seven (2.1%) serum samples showed values higher than that of the cut-off values of the stg-BSA, ES, and CWE antigens, respectively (Fig. 1). By Western Blot Analysis, the ELISA-positive serum samples from horses slaughtered in Italy and originating from Italian Trichinella-free areas showed some unspecific bands when using the ES and CWE antigens, whereas they were negative when using the stg-BSA antigen (Fig. 2).
168
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Fig. 1. Optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of 330 serum samples from horses originating from Trichinella-free areas of Italy, slaughtered in Italy, and resulted negative by the artificial digestion of 10 g of muscles from the tongue and/or the masseter.
3.2. Experimentally infected horses In the ten horses infected with 5000 T. spiralis larvae, specific IgG were detectable between the third week and the eighth week p.i. when using stg-BSA and ES as antigen, and between the third week and the sixth week p.i. when using CWE. In some horses,
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
169
Fig. 2. Western Blot Analysis using three different antigens (a crude antigen, lines 1–4; an excretory/secretory antigen, lines 5–8; and a synthetic tyvelose glycan–BSA antigen, lines 9–12) of horse serum samples showing different patterns. Lines 1, 5, and 9: serum sample from an experimentally infected horse; lines 2, 6, and 10: serum sample from a Trichinella-negative horse; lines 3, 7, and 11: serum sample from the horse harboring Trichinella larvae with a thin collagen capsule imported to Italy from Poland; lines 4, 8, and 12: serum sample from a horse originating from a Trichinella-free area of Italy, slaughtered in Italy, and found to be negative by the artificial digestion of 10 g of muscles from the tongue, yet showing a positive ELISA with all three antigens tested. MW: molecular weight marker in Kd.
specific IgG were no longer detectable 9–20 weeks p.i. irrespective of the antigen used (Fig. 3). At the sixth and seventh week after infection, one horse, which showed a high positive OD value by ELISA using an ES antigen and a CWE antigen, showed negative OD values when using the stg-BSA antigen. In the three horses infected with 5000 T. murrelli larvae, when using the stg-BSA antigen, specific IgG were detectable in one horse 4 weeks p.i. whereas the OD of the other two horses remained below the cut-off value. When using the ES antigen, specific IgG were detectable 5–7 weeks p.i. for all three horses; for one horse, IgG remained detectable until 14 weeks p.i. When using CWE, specific IgG were detectable 3–5 weeks p.i. for all horses and remained detectable until 14 weeks p.i. in one horse (Fig. 4). In the three horses infected with 20,000 T. murrelli larvae, specific IgG were detectable between 4 and 5 weeks p.i. and 22 weeks p.i. (in only one horse) when using the stg-BSA antigen, and between 4 and 5 weeks p.i. and 33 weeks p.i. (in only one horse), when using ES; when using CWE, specific IgG were detectable between 3 and 5 weeks p.i. and 26 weeks p.i. (in only one horse) (Fig. 5). The kinetics of anti-Trichinella IgG were similar among all examined horses. The OD values peaked between weeks 5 and 8 p.i. (Figs. 3–5). The highest OD values were observed
170
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Fig. 3. Kinetic of optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of serum samples from ten horses experimentally infected with 5000 larvae of Trichinella spiralis. The cut-off values for the three antigens was that evaluated on the basis of the OD values of horses shown in Fig. 1.
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
171
Fig. 4. Kinetic of optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of serum samples from three horses experimentally infected with 5000 larvae of Trichinella murrelli. The cut-off values for the three antigens was that evaluated on the basis of the OD values of horses shown in Fig. 1.
172
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Fig. 5. Kinetic of optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of serum samples from three horses experimentally infected with 20,000 larvae of Trichinella murrelli. The cut-off values for the three antigens was that evaluated on the basis of the OD values of horses shown in Fig. 1.
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
173
Fig. 6. Optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of 1739 serum samples from horses originating from Italy (450), Poland (627), Romania (485), and Serbia (177), slaughtered in Italy, and found to be negative (1735) or positive (4) by the artificial digestion of 10 g of muscles from the tongue and/or the masseter. Serum samples from four horses imported to and slaughtered in Italy, with infective Trichinella larvae in their muscles (䉱).
174
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Fig. 7. Optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of 433 serum samples from live Romanian horses.
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
175
for the CWE antigen, followed by those obtained with the ES antigen and by the stg-BSA antigen. 3.3. Horses slaughtered in Italy Of the 1739 horses slaughtered in Italy, excluding the 330 horses from Trichinella-free Italian regions, four horses were found to be positive by artificial digestion: two horses from Romania, one from Serbia, and one from Poland, with 1.7, 11, and 615 larvae/g in the tongue and 256 larvae/g in the diaphragm (the horse from Poland), respectively. Only the sample from the Trichinella-positive horse imported from Poland showed an OD higher than that of the cut-off value for all three antigens (CWE, 1.29; ES, 1.13; and stg-BSA, 0.59), whereas the samples from the other three positive horses were negative for all three antigens. The OD values from Trichinella-negative horses ranged from 0.17 to 1.29 with the CWE antigen (average 0.55, S.D. 0.18), from 0.14 to 1.13 with the ES antigen (average 0.45, S.D. 0.18), and from 0.10 to 0.64 with the stg-BSA antigen (average 0.28, S.D. 0.11) (Fig. 6). The OD was higher than the cut-off value for three (0.17%), ten (0.57%), and seventeenth (0.98%) horses when using the stg-BSA antigen, the ES antigen, and the CWE antigen, respectively. The Western Blot Analysis of ELISA-positive serum samples showed that all but one sample (i.e. that belonging to the Trichinella-positive horse imported from Poland) showed some non-specific bands when using the ES and CWE antigens and no band at all when using the stg-BSA antigen (data not shown). The serum sample from the Trichinella-positive horse imported from Poland showed specific bands with all three antigens (Fig. 2). Of the 265 fecal samples examined, 233 (87.9%) showed strongylid larvae belonging to the families of Strongyloididae, Strongylidae, and Trichostrongylidae; 69 (26.0%) had Parascaris equorum eggs, 11 (4.1%) Anoplocephalidae eggs, and 16 (6.0%) were negative. The OD values of the serum samples from all 265 horses were lower than the cut-off values of the three antigens tested. 3.4. Live horses from Romania Of the 433 serum samples from live Romanian horses, three (0.7%), six (1.4%), and eleven (2.5%) samples showed OD values higher than the cut-off values for the stg-BSA antigen, the ES antigen, and the CWE antigen, respectively (Fig. 7). The Western Blot Analysis of ELISA-positive serum samples showed that all serum samples had some non-specific bands when using the ES and CWE antigens and no band at all when using the stg-BSA antigen (data not shown).
4. Discussion The results of this study confirm previous findings that anti-Trichinella IgG remain detectable in horse sera only up to 22–33 weeks p.i. depending on the specific antigen and
176
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Trichinella species, and that muscle larvae continue to be infective even after serology becomes negative (Polidori et al., 1989; Soulé et al., 1989; Pozio et al., 1997; Boireau et al., 2000). The highest OD values were observed between the fifth and the thirteenth weeks p.i. depending on the infecting dose and the Trichinella species, which influence the time of detection of the immune response by ELISA (Figs. 3–5). Specifically, T. spiralis induces higher levels of anti-Trichinella IgG than T. murrelli (Figs. 4 and 5), most probably because of its higher immunogenicity (Bruschi et al., 1999). The findings that only one of the serum samples with OD values higher than the cut-off values came from a horse positive for Trichinella larvae and that the Western Blot Analysis showed that the same horse was truly serologically-positive confirm that ELISA can provide false-positive results, even when using the stg-BSA antigen. These results are consistent with the observation that the serologically-positive horse was the only one of the four horses harboring Trichinella larvae that showed a very thin collagen capsule around muscle larvae, suggesting a recent infection, whereas the other three horses showed a very thick capsule around muscle larvae, suggesting an older infection (Pozio et al., 1999). For the live Romanian horses, all samples positive by ELISA were revealed as false-positives when subjected to Western Blot Analysis. One of the possible reasons for which ELISA may have provided false-positive results, despite the fact that a very specific antigen (stg-BSA) was used, could be that the method for calculating the cut-off value for OD may not be adequate for use on horses. The finding that all of the horses harboring an intestinal helminthic fauna showed ELISA OD values lower than the cut-off values for all three antigens indicates that the presence of these helminths does not play a role in the occurrence of cross-reactions, which result in false-positive ELISA results. However, the presence of some cross-reactive antibodies in horse sera cannot be excluded. An indirect support for this hypothesis could be the unusually high scattering of OD values obtained for individual sera (diluted 1:50) of Trichinella-free horses. A certain degree of similarity has been found between plant and animal-parasitic nematodes in terms of the chemical composition, origin, and biological role of the nematode surface coat (Spiegel and McClure, 1995). Monoclonal antibodies against ES products of plant-parasitic nematodes were shown to cross-react with ES products and the surface coats of the animal-parasites (T. spiralis and Haemonchus contortus) (Lopez de Mendoza et al., 1999); thus we cannot exclude that the observed cross-reactivity of serum samples from Trichinella-free horses is related to a previous sensitization of horses with plant-parasitic nematodes. It is not surprising that this problem occurs with serum samples from horses (i.e. a herbivorous animal), whereas a cross-reactivity of serum samples from Trichinella-free swine has not been reported (Gamble et al., 1997), although large numbers of pigs from organic farms have not been examined. The results of ELISA and Western Blot Analysis suggest that serology cannot be used to detect Trichinella infection in horses because the antibody response does not correlate with larval infection in the muscle tissue. Between 1975 and 1998, the prevalence of Trichinella infection in horses was estimated to be 3.6 per 1 million slaughtered (Pozio, 2001). In the past 2 years, the increased use of artificial digestion to detect Trichinella infection in horses in France and Italy has resulted in a greater number of infected horses being detected, resulting in a prevalence of 8.0 per 1 million slaughtered horses.
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
177
In conclusion, these results suggest that the detection of anti-Trichinella IgG in horse serum samples, irrespective of the specific technique or antigen used, cannot be considered as the method of choice and that the artificial digestion of at least 5 g of tissue from preferential-muscles, as suggested by the European Union regulation, continues to be the method of choice at the slaughterhouse for preventing equine-borne clinical trichinellosis in humans.
Acknowledgements We wish to thank the Heska Corporation, which provided us with the synthetic tyvelose glycan–BSA antigen. We also thank Marco Amati and Catherine Perret for their technical support. This study received financial support from the research project no. 9D/F entitled “Fast identification of parasitic agents in foods” of the Italian Ministry of Health, and from the research project no.1157/RI entitled “Studies of the immune response against zoonotic agents: Cryptosporidium and Trichinella” of the Istituto Superiore di Sanità, Rome, Italy. References Boireau, P., Vallée, I., Roman, T., Perret, C., Mingyuan, L., Gamble, H.R., Gajadhar, A., 2000. Trichinella in horses: a low frequency infection with high human risk. Vet. Parasitol. 93, 309–320. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of the protein-dye binding. Anal. Biochem. 72, 248–254. Bruschi, F., Pozio, E., Watanabe, N., Gomez Morales, M.A., Ito, M., Huang, Y., Binaghi, R., 1999. Anaphylactic response to parasite antigens: IgE and IgG1 independently induce death in Trichinella-infected mice. Int. Arch. Allergy Immunol. 119, 291–296. Cuperlovic, K., Djordjevic, M., Pavlovic, S., Sofronic-Milosavljevic, L., 2001. Present status of trichinellosis in Yugoslavia: Serbia. Parasite 8 (S2), S95–S97. Denkers, E.Y., Hayes, C.E., Wassom, D.L., 1991. Trichinella spiralis: influence of an immunodominant carbohydrate-associated determinant on the host antibody response repertoire. Exp. Parasitol. 72, 403–410. Gamble, H.R., Anderson, W.R., Graham, C.E., Murrell, K.D., 1983. Diagnosis of swine trichinosis by enzyme-linked immunosorbent assay (ELISA) using an excretory/secretory antigen. Vet. Parasitol. 13, 349– 361. Gamble, H.R., Bessonov, A.S., Cuperlovic, K., Gajadhar, A.A., Van Knapen, F., Nöckler, K., Schenone, H., Zhu, X., 2000. International commission on Trichinellosis: recommendations on methods for the control of Trichinella in domestic and wild animals intended for human consumption. Vet. Parasitol. 93, 393–408. Gamble, H.R., Gajadhar, A.A., Solomon, M.B., 1996. Methods for the detection of trichinellosis in horses. J. Food Prot. 59, 420–425. Gamble, H.R., Wisnewski, N., Wasson, D.L., 1997. Diagnosis of trichinellosis in swine by enzyme immunoassay, using a synthetic glycan antigen. Am. J. Vet. Res. 58, 1417–1421. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4 . Nature 227, 680–685. Lopez de Mendoza, M.E., Curtis, R.H.C., Gowen, S., 1999. Identification and characterization of excreted-secreted products and surface coat antigens of animal and plant-parasitic nematodes. Parasitology 118, 397–405. Olteanu, G., 2001. Trichinellosis in Romania: a short review over the past 20 years. Parasite 8 (2S), S98–S99. Polidori, G.A., Gramenzi, F., Piergili Fioretti, D., Ferri, N., Ranucci, S., Moretti, A., Scacchia, M., Bellelli, C., Baldelli, B., 1989. Experimental trichinellosis in horses. In: Tanner, C.E., Martinez-Fernandez, A.R., Bolas Fernandez, F. (Eds.), Trichinellosis, CSIC Press, Madrid, Spain, pp. 268–274. Pozio, E., 1998. Trichinellosis in the European union: epidemiology, ecology and economic impact. Parasitol. Today 14, 35–38.
178
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Pozio, E., 2001. New patterns of Trichinella infections. Vet. Parasitol. 98, 133–148. Pozio, E., La Rosa, G., Gomez Morales, M.A., 2001. Epidemiology of human and animal trichinellosis in Italy since its discovery in 1887. Parasite 8 (2S), S106–S108. Pozio, E., Paterlini, F., Pedarra, C., Sacchi, L., Bugarini, R., Goffredo, E., Boni, P., 1999. Predilection sites of Trichinella spiralis larvae in naturally infected horses. J. Helminthol. 73, 233–237. Pozio, E., Tamburrini, A., Sacchi, L., Gomez Morales, M.A., Corona, S., Goffredo, E., La Rosa, G., 1997. Detection of Trichinella spiralis in a horse during routine examination in Italy. Int. J. Parasitol. 27, 1613–1621. Ramisz, A., Szymborski, J., Balicka-Ramisz, A., 2001. Epidemiological studies on trichinellosis among swine, wild boars and humans in Poland. Parasite 8 (2S), S90–S91. Smith, H.J., Snowdon, K.E., 1987. Experimental Trichinella infections in Ponies. Can. J. Vet. Res. 51, 415–416. Sofronic-Milosavljevic, L., Pozio, E., Patrascu, I.V., Skerovic, N., Gomez Morales, M.A., Gamble, H.R., 2001. Immunodiagnosis of Trichinella infection in the horse. Parasite 8 (25), S260–S262. Soulé, C., Dupouy-Camet, J., Georges, P., Ancelle, T., Gillet, J.P., Vaissaire, J., Delvigne, A., Plateau, E., 1989. Experimental trichinellosis in horses: biological and parasitological evaluation. Vet. Parasitol. 31, 19–36. Soulsby, E.J.L., 1965. Textbook of Veterinary Clinical Parasitology: Helminths, Vol. 1. Blackwell Scientific Publication, Oxford. Spiegel, Y., McClure, M.A., 1995. The surface-coat of plant-parasitic nematodes—chemical composition, origin and biological role—a review. J. Nematology 27, 127–134. Towbin, H., Staehelin, T., Gordon, J., 1979. Electrophoretic transfer of protein from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Nat. Acad. Sci. U.S.A. 76, 4350–4354. van Knapen, F., Franchimont, J.M., Hendrikx, W.M., Eysker, M., 1987. Experimental Trichinella spiralis infection in two horses. Vet. Arch. 31, 123–132. Voigt, W.P., Nöckler, K., Freischem, B., Henriksen, S.A., van Knapen, F., Martinez Fernandez, A., Pfeiffer, G., Pozio, E., Reuter, G., Ring, Ch., Soulé, C., Weiss, H., 1997. Detection of low levels of Trichinella spiralis in experimentally infected horses. In: Ortega-Pierres, G., Gamble, H.R., van Knapen, F., Wakelin, D. (Eds.), Trichinellosis, Nonoalco Tlateloco, Mexico, pp. 629–634. Wisnewski, N., McNeil, M., Grieve, R.B., Wassom, D.L., 1993. Characterization of novel fucosil- and tyvelosyl-containing glycoconjugates from Trichinella spiralis muscle stage larvae. Mol. Biochem. Parasitol. 61, 25–36.
Evaluation of ELISA and Western Blot Analysis using three antigens to detect anti-Trichinella IgG in horses Edoardo Pozio a,∗ , Ljiljana Sofronic-Milosavljevic b , Maria Angeles Gomez Morales a , Pascal Boireau b , Karsten Nöckler c a
Laboratory of Parasitology, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy b UMR 956 INRA-AFSSA-ENVA, Biologie Moléculaire et Immunologie Parasitaires et Fongiques, 22 rue Pierre Curie, 94703 Maisons Alfort, France c Federal Institute for Health Protection of Consumers and Veterinary Medicine, Diedersdorfer Weg 1, 12277 Berlin, Germany Received 18 January 2002; received in revised form 27 May 2002; accepted 10 June 2002
Abstract We assessed a serological method for detecting Trichinella infection in horses, specifically, an ELISA using three antigens to detect anti-Trichinella IgG (i.e. a synthetic tyvelose glycan–BSA (stg-BSA) antigen, an excretory/secretory (ES) antigen, and a crude worm extract (CWE) antigen). Serum samples were collected from 2502 horses (433 live horses from Romania and 2069 horses slaughtered in Italy and originating from Italy, Poland, Romania, and Serbia). Serum samples were also taken from horses experimentally infected with different doses of T. spiralis and T. murrelli larvae, as controls. The cut-off value of ELISA was determined on serum samples from 330 horses from Trichinella-free regions of Italy, which were also examined by artificial digestion of preferential-muscle samples. In the experimentally infected horses, the stg-BSA and ES antigens were less sensitive than the CWE antigen. Trichinella spiralis showed a higher immunogenicity than T. murrelli, and the IgG immunoresponse was dose-dependent. The kinetics of anti-Trichinella IgG were similar among all experimentally infected horses. No circulating antibodies were detected 4–5 months after experimental infection, although these horses still harbored infective larvae. Depending on the antigen used, for 4–7 of the 330 horses from Trichinella-free areas, the optical density (OD) of the serum sample was higher than the cut-off value, yet these samples were negative when subjected to Western Blot. Similar results were obtained for the 1739 horses slaughtered in Italy (originating from Italy, Poland, Romania, and Serbia) and the 433 live Romanian horses. Of the 4 horses with
∗ Corresponding author. Tel.: +39-06-4990-2304; fax: +39-06-4938-7065. E-mail address: [email protected] (E. Pozio).
0304-4017/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 0 4 - 4 0 1 7 ( 0 2 ) 0 0 1 8 5 - 1
164
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
muscle larvae, only one was positive by ELISA and Western Blot. Because the anti-Trichinella IgG remain circulating for only a short period of time, whereas the larvae remain infective for longer periods, serology cannot be used for either diagnosing Trichinella infection in horses or estimating the prevalence of infedction. Artificial digestion of at least 5 g of preferential-muscle tissue continues to be the method of choice at the slaughterhouse for preventing equine-borne trichinellosis in humans. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Trichinella spiralis; Trichinella murrelli; Horse; Europe; ELISA; Western Blot; Synthetic tyvelose glycan–BSA antigen; Excretory/secretory antigen
1. Introduction Trichinella infection in horses continues to represent a serious threat to animal and human health in France and Italy, the only two European countries where horse meat is frequently consumed raw (Pozio, 2001). Several diagnostic tests have been used to detect Trichinella infection in this animal, yet the artificial digestion of at least 5–10 g of muscle from the tip of the tongue seems to be the only suitable diagnostic method, although it is expensive and time consuming (Boireau et al., 2000). Since 1994, the year in which this diagnostic method was introduced as part of routine practice by EU legislation (Directive 77/96/EEC, amended by Directive 94/59/EEC), from one to two naturally infected horses have been detected in France and Italy per year, and all of these horses originated from eastern Europe. In the same period, human trichinellosis outbreaks occurred only when veterinary services failed to properly apply this method when controlling for the parasite at the slaughterhouse (Boireau et al., 2000; Pozio, 2001). To develop a serological method to detect this infection in live horses, many researches have studied the kinetics of the humoral immunoresponse against Trichinella in experimentally infected horses using excretory/secretory (ES) antigen or crude worm extract antigen in an ELISA (Smith and Snowdon, 1987; van Knapen et al., 1987; Polidori et al., 1989; Soulé et al., 1989; Gamble et al., 1996; Voigt et al., 1997). However, the results of these studies have been frustrating because the specific IgG become undetectable 4–5 months p.i. at the latest, despite the fact that the larvae remain infective in the muscle tissues, as reported for both experimentally infected (Polidori et al., 1989; Soulé et al., 1989) and naturally infected (Pozio et al., 1997) horses. Furthermore, a serosurvey carried out on horses from Romania, which is one of the European countries with the highest prevalence of domestic trichinellosis, has shown a high number of false-positive sera by ELISA, suggesting that the antigens used (excretory/secretory antigen) in the assays had a low specificity (Sofronic-Milosavljevic et al., 2001). The objective of the present study was to asses the diagnostic value of an ELISA using three antigens to detect anti-Trichinella IgG. To this end, we studied sera from both experimentally infected horses and from horses originating from several European countries with different prevalence rates of Trichinella infection and from Trichinella-free regions in Italy (Pozio, 1998).
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
165
2. Materials and methods 2.1. Serum samples We studied serum samples from eight horses bred and experimentally infected in France and eight horses bred and experimentally infected in Germany. The horses from France were infected with either T. spiralis (two animals infected with 5000 larvae) or T. murrelli (three animals infected with 5000 larvae and three animals with 20,000 larvae) (Soulé et al., 1989); the horses from Germany were infected with T. spiralis (with 5000 larvae each) (Voigt et al., 1997). Serum samples were collected from these horses beginning 1 week before experimental infection and up to 43 weeks p.i. Serum/blood samples were also collected from 2502 additional horses, which consisted of 433 live horses from Romania, which has an average prevalence of 1% of Trichinella infection in domestic pigs (Olteanu, 2001), and 2069 horses slaughtered in Italy. The horses slaughtered in Italy originated from various countries, specifically: 780 from Italy, which is free from domestic trichinellosis but which has a prevalence of vulpine infection (T. britovi) ranging from 0 to 35% (in some mountain areas) (Pozio et al., 2001); 627 from Poland, which has a 0.0004% prevalence of Trichinella infection in domestic pigs (Ramisz et al., 2001); 485 from Romania; and 177 from Serbia, which has a 0.1–0.3% prevalence of Trichinella infection in domestic pigs (Cuperlovic et al., 2001). Of the 780 Italian horses, 330 originated from Trichinella-free regions (Pozio, 1998); the other 450 horses originated from regions with different prevalence rates of sylvatic trichinellosis or from unspecified areas of Italy. After sample collection, blood was allowed to clot at room temperature. The serum was then collected and frozen at −80 ◦ C. We established the optimal serum dilution by testing serum samples from experimentally infected horses and from horses known to be negative for Trichinella at dilutions of 1:10, 1:20, 1:40, 1:50, 1:80, 1:100, 1:120, 1:150, and 1:200, using three different antigens by ELISA (see below). Using the paired Student t-test, the greatest difference in optical density (OD) between positive and negative serum samples was obtained at a dilution of 1:50 (data not shown), which was used for all subsequent ELISA. 2.2. Parasitological examinations To examine meat samples of horses slaughtered in Italy (n = 2069) for Trichinella larvae, 10 g of masseter or tongue were examined by artificial digestion (Gamble et al., 2000). To evaluate the specificity of serological analyses, the intestinal helminth fauna was studied. Fecal samples, collected directly from the rectum of 265 horses, were examined by flotation method. Positive samples were cultured at 26 ◦ C for 8 days to identify the third larval stage at the species level, in accordance with Soulsby (1965). 2.3. Antigens We used three antigens to test all serum samples: a crude worm extract antigen (CWE); an excretory/secretory antigen; and a 3,6-dideoxyhexose sugar (tyvelose), which is one of the major highly specific immunodominant epitopes of Trichinella, coupled with BSA
166
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
(i.e. synthetic tyvelose glycan–BSA (stg-BSA; Heska Corporation, Fort Collins, CO, USA) (Denkers et al., 1991; Wisnewski et al., 1993)). CWE was prepared using T. spiralis larvae collected from the muscles of experimentally infected Swiss mice by artificial digestion. After washing six times by settling in 0.1 M phosphate buffered saline (PBS), pH 7.2, larvae were stored at −80 ◦ C until use. After thawing, larvae were crushed in a glass Potter-homogenizer using a teflon pestle and further disintegrated by sonication. The larval suspension was maintained overnight at 4 ◦ C under magnetic stirring and centrifuged at 17, 200 × g at 4 ◦ C for 1 h. The protein concentration of the supernatant was determined by Bradford’s method (1976). To produce the ES antigen, the protocol of Gamble et al. (1983) was followed. Briefly, T. spiralis muscle larvae were washed three times in PBS, pH 7.2, with penicillin (500 units/ml) and streptomycin (500 g/ml). Larvae were then washed four times by settling in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with penicillin (500 units/ml) and streptomycin (500 g/ml). Five thousand worms per ml were resuspended in DMEM, supplemented with 1 M HEPES, 200 mM l-glutamine, 100 mM Na-pyruvate, and 5000 units of penicillin/streptomycin (GIBCO, Grand Island, NY), and incubated with 10% CO2 in a 75 cm2 culture flask (Costar) at 37 ◦ C for 18 h. Worms were then removed from the medium by settling in 50 ml conical tubes. The medium was filtered through a 0.2 m filter, and the supernatant was concentrated 100 times in an Amicon® (Amicon Inc., MA, USA) pressure concentrating chamber using a YM-5 filter. To determine the protein concentration and quality of the batch (i.e. without any bacterial or fungal contamination), OD was evaluated at a 280/260 nm ratio; antigens with a ratio higher than 1.2 were used. 2.4. ELISA using CWE or ES antigens An indirect ELISA method was used to detect Trichinella-specific IgG. Briefly, T. spiralis antigens (CWE or ES) were diluted in 0.1 M carbonate buffer, pH 9.6, and used at a concentration of 5.0 g/ml. Flat-bottomed polystyrene microtiter plates were coated with 100 l/well of CWE or ES in 0.1 M carbonate buffer, pH 9.6, at 37 ◦ C for 1 h. Coated plates were washed three times with 0.05% Tween 20 in H2 O. Plates were blocked with 200 l of 1.0% bovine serum albumin in PBS with 0.05% Tween 20 (PBST) at 37 ◦ C for 1 h. Plates were then washed and incubated with 100 l/well of five positive and five negative serum controls and horse serum samples at a 1:50 dilution at 37 ◦ C for 1 h. After washing as described above, 100 l/well of a goat anti-horse IgG conjugated with horse-radish peroxidase (Kirkegaard & Perry Laboratories Inc., Maryland) were added to the wells at a 1:10,000 dilution, and the plate was incubated at 37 ◦ C for 1 h. The plate was washed again, and 100 l/well of TMB Microwell Peroxidase Substrate System (Kirkegaard & Perry Laboratories, Maryland) was added. After 10 min, the reaction was stopped with 100 l/well HCl 1N, and the plate was read at 450 nm using an MRX Microplate Reader (Dynex technologies Inc., VA, USA). 2.5. ELISA using the stg-BSA antigen An indirect ELISA method was also used to detect Trichinella-specific IgG using the stg-BSA antigen. A Nunc MaxiSorpTM 96-well plate was coated with 100 l/well of the
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
167
stg-BSA antigen (1.25 g/ml) in 0.05 M Tris–EDTA–NaCl buffer, pH 7.3, at 5 ± 3 ◦ C for 16–18 h. The plate was then washed four times with PBST, and 100 l/well of PBST-diluted serum with 4% fetal bovine serum (FBS), and 0.5% Proclin® 300 (P 300) (Supelco Inc., PA, USA) was added. After 30 min incubation at 22±2 ◦ C, the plate was washed four times with PBST, and 100 l/well of a goat anti-horse IgG conjugated with horse-radish peroxidase was added at a 1:10,000 dilution in PBST, with 4% FBS and 0.5% P 300. The plate was incubated at 22 ± 2 ◦ C for 30 min. It was then washed four times in PBST and once with deionized water. Addition of the substrate and reading of the plate were done as described above. Since the synthetic antigen is conjugated to BSA, to avoid a possible reactivity of horse sera with BSA, serum samples were pre-incubated with nitrocellulose-BSA strips at room temperature for 1 h. 2.6. Cut-off values The cut-off value of ELISA was evaluated for the three antigens on the basis of the average OD plus three standard deviations (S.D.) of 330 serum samples of horses reared from birth in Trichinella-free areas of Italy, slaughtered in Italy, and without larvae of Trichinella in 10 g of masseter or tongue. 2.7. Western Blot Protein components of CWE, ES, and stg-BSA were electrophoretically separated on 8.0% sodium dodecyl sulfate polyacrylamide-gel (SDS-PAGE), according to Laemmli (1970), and transferred to nitrocellulose sheets, as performed by Towbin et al. (1979). Reference and ELISA-positive horse serum samples were diluted at 1:50, and goat anti-horse IgG with horse-radish peroxidase (Kirkegaard & Perry Laboratories Inc., Maryland), diluted at 1:3000, was used as conjugate. The immunoblot was evaluated using chemiluminescent principles (ECL Western Blotting detection reagents, Amersham Life Science, Little Chalfont, Buckinghamshire, UK) and visualized by exposing the membranes to autoradiography film for 30 s.
3. Results 3.1. Cut-off values The cut-off values detected by ELISA using the stg-BSA, ES, and CWE antigens on 330 serum samples from horses slaughtered in Italy and originating from Italian Trichinella-free areas, were 0.55, 0.92, and 1.02, respectively (Fig. 1). Four (1.2%), five (1.5%), and seven (2.1%) serum samples showed values higher than that of the cut-off values of the stg-BSA, ES, and CWE antigens, respectively (Fig. 1). By Western Blot Analysis, the ELISA-positive serum samples from horses slaughtered in Italy and originating from Italian Trichinella-free areas showed some unspecific bands when using the ES and CWE antigens, whereas they were negative when using the stg-BSA antigen (Fig. 2).
168
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Fig. 1. Optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of 330 serum samples from horses originating from Trichinella-free areas of Italy, slaughtered in Italy, and resulted negative by the artificial digestion of 10 g of muscles from the tongue and/or the masseter.
3.2. Experimentally infected horses In the ten horses infected with 5000 T. spiralis larvae, specific IgG were detectable between the third week and the eighth week p.i. when using stg-BSA and ES as antigen, and between the third week and the sixth week p.i. when using CWE. In some horses,
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
169
Fig. 2. Western Blot Analysis using three different antigens (a crude antigen, lines 1–4; an excretory/secretory antigen, lines 5–8; and a synthetic tyvelose glycan–BSA antigen, lines 9–12) of horse serum samples showing different patterns. Lines 1, 5, and 9: serum sample from an experimentally infected horse; lines 2, 6, and 10: serum sample from a Trichinella-negative horse; lines 3, 7, and 11: serum sample from the horse harboring Trichinella larvae with a thin collagen capsule imported to Italy from Poland; lines 4, 8, and 12: serum sample from a horse originating from a Trichinella-free area of Italy, slaughtered in Italy, and found to be negative by the artificial digestion of 10 g of muscles from the tongue, yet showing a positive ELISA with all three antigens tested. MW: molecular weight marker in Kd.
specific IgG were no longer detectable 9–20 weeks p.i. irrespective of the antigen used (Fig. 3). At the sixth and seventh week after infection, one horse, which showed a high positive OD value by ELISA using an ES antigen and a CWE antigen, showed negative OD values when using the stg-BSA antigen. In the three horses infected with 5000 T. murrelli larvae, when using the stg-BSA antigen, specific IgG were detectable in one horse 4 weeks p.i. whereas the OD of the other two horses remained below the cut-off value. When using the ES antigen, specific IgG were detectable 5–7 weeks p.i. for all three horses; for one horse, IgG remained detectable until 14 weeks p.i. When using CWE, specific IgG were detectable 3–5 weeks p.i. for all horses and remained detectable until 14 weeks p.i. in one horse (Fig. 4). In the three horses infected with 20,000 T. murrelli larvae, specific IgG were detectable between 4 and 5 weeks p.i. and 22 weeks p.i. (in only one horse) when using the stg-BSA antigen, and between 4 and 5 weeks p.i. and 33 weeks p.i. (in only one horse), when using ES; when using CWE, specific IgG were detectable between 3 and 5 weeks p.i. and 26 weeks p.i. (in only one horse) (Fig. 5). The kinetics of anti-Trichinella IgG were similar among all examined horses. The OD values peaked between weeks 5 and 8 p.i. (Figs. 3–5). The highest OD values were observed
170
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Fig. 3. Kinetic of optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of serum samples from ten horses experimentally infected with 5000 larvae of Trichinella spiralis. The cut-off values for the three antigens was that evaluated on the basis of the OD values of horses shown in Fig. 1.
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
171
Fig. 4. Kinetic of optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of serum samples from three horses experimentally infected with 5000 larvae of Trichinella murrelli. The cut-off values for the three antigens was that evaluated on the basis of the OD values of horses shown in Fig. 1.
172
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Fig. 5. Kinetic of optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of serum samples from three horses experimentally infected with 20,000 larvae of Trichinella murrelli. The cut-off values for the three antigens was that evaluated on the basis of the OD values of horses shown in Fig. 1.
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
173
Fig. 6. Optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of 1739 serum samples from horses originating from Italy (450), Poland (627), Romania (485), and Serbia (177), slaughtered in Italy, and found to be negative (1735) or positive (4) by the artificial digestion of 10 g of muscles from the tongue and/or the masseter. Serum samples from four horses imported to and slaughtered in Italy, with infective Trichinella larvae in their muscles (䉱).
174
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Fig. 7. Optical density values in ELISA using three different antigens (a synthetic tyvelose glycan–BSA antigen, stg-BSA; an excretory/secretory antigen, ES; and a crude antigen) of 433 serum samples from live Romanian horses.
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
175
for the CWE antigen, followed by those obtained with the ES antigen and by the stg-BSA antigen. 3.3. Horses slaughtered in Italy Of the 1739 horses slaughtered in Italy, excluding the 330 horses from Trichinella-free Italian regions, four horses were found to be positive by artificial digestion: two horses from Romania, one from Serbia, and one from Poland, with 1.7, 11, and 615 larvae/g in the tongue and 256 larvae/g in the diaphragm (the horse from Poland), respectively. Only the sample from the Trichinella-positive horse imported from Poland showed an OD higher than that of the cut-off value for all three antigens (CWE, 1.29; ES, 1.13; and stg-BSA, 0.59), whereas the samples from the other three positive horses were negative for all three antigens. The OD values from Trichinella-negative horses ranged from 0.17 to 1.29 with the CWE antigen (average 0.55, S.D. 0.18), from 0.14 to 1.13 with the ES antigen (average 0.45, S.D. 0.18), and from 0.10 to 0.64 with the stg-BSA antigen (average 0.28, S.D. 0.11) (Fig. 6). The OD was higher than the cut-off value for three (0.17%), ten (0.57%), and seventeenth (0.98%) horses when using the stg-BSA antigen, the ES antigen, and the CWE antigen, respectively. The Western Blot Analysis of ELISA-positive serum samples showed that all but one sample (i.e. that belonging to the Trichinella-positive horse imported from Poland) showed some non-specific bands when using the ES and CWE antigens and no band at all when using the stg-BSA antigen (data not shown). The serum sample from the Trichinella-positive horse imported from Poland showed specific bands with all three antigens (Fig. 2). Of the 265 fecal samples examined, 233 (87.9%) showed strongylid larvae belonging to the families of Strongyloididae, Strongylidae, and Trichostrongylidae; 69 (26.0%) had Parascaris equorum eggs, 11 (4.1%) Anoplocephalidae eggs, and 16 (6.0%) were negative. The OD values of the serum samples from all 265 horses were lower than the cut-off values of the three antigens tested. 3.4. Live horses from Romania Of the 433 serum samples from live Romanian horses, three (0.7%), six (1.4%), and eleven (2.5%) samples showed OD values higher than the cut-off values for the stg-BSA antigen, the ES antigen, and the CWE antigen, respectively (Fig. 7). The Western Blot Analysis of ELISA-positive serum samples showed that all serum samples had some non-specific bands when using the ES and CWE antigens and no band at all when using the stg-BSA antigen (data not shown).
4. Discussion The results of this study confirm previous findings that anti-Trichinella IgG remain detectable in horse sera only up to 22–33 weeks p.i. depending on the specific antigen and
176
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Trichinella species, and that muscle larvae continue to be infective even after serology becomes negative (Polidori et al., 1989; Soulé et al., 1989; Pozio et al., 1997; Boireau et al., 2000). The highest OD values were observed between the fifth and the thirteenth weeks p.i. depending on the infecting dose and the Trichinella species, which influence the time of detection of the immune response by ELISA (Figs. 3–5). Specifically, T. spiralis induces higher levels of anti-Trichinella IgG than T. murrelli (Figs. 4 and 5), most probably because of its higher immunogenicity (Bruschi et al., 1999). The findings that only one of the serum samples with OD values higher than the cut-off values came from a horse positive for Trichinella larvae and that the Western Blot Analysis showed that the same horse was truly serologically-positive confirm that ELISA can provide false-positive results, even when using the stg-BSA antigen. These results are consistent with the observation that the serologically-positive horse was the only one of the four horses harboring Trichinella larvae that showed a very thin collagen capsule around muscle larvae, suggesting a recent infection, whereas the other three horses showed a very thick capsule around muscle larvae, suggesting an older infection (Pozio et al., 1999). For the live Romanian horses, all samples positive by ELISA were revealed as false-positives when subjected to Western Blot Analysis. One of the possible reasons for which ELISA may have provided false-positive results, despite the fact that a very specific antigen (stg-BSA) was used, could be that the method for calculating the cut-off value for OD may not be adequate for use on horses. The finding that all of the horses harboring an intestinal helminthic fauna showed ELISA OD values lower than the cut-off values for all three antigens indicates that the presence of these helminths does not play a role in the occurrence of cross-reactions, which result in false-positive ELISA results. However, the presence of some cross-reactive antibodies in horse sera cannot be excluded. An indirect support for this hypothesis could be the unusually high scattering of OD values obtained for individual sera (diluted 1:50) of Trichinella-free horses. A certain degree of similarity has been found between plant and animal-parasitic nematodes in terms of the chemical composition, origin, and biological role of the nematode surface coat (Spiegel and McClure, 1995). Monoclonal antibodies against ES products of plant-parasitic nematodes were shown to cross-react with ES products and the surface coats of the animal-parasites (T. spiralis and Haemonchus contortus) (Lopez de Mendoza et al., 1999); thus we cannot exclude that the observed cross-reactivity of serum samples from Trichinella-free horses is related to a previous sensitization of horses with plant-parasitic nematodes. It is not surprising that this problem occurs with serum samples from horses (i.e. a herbivorous animal), whereas a cross-reactivity of serum samples from Trichinella-free swine has not been reported (Gamble et al., 1997), although large numbers of pigs from organic farms have not been examined. The results of ELISA and Western Blot Analysis suggest that serology cannot be used to detect Trichinella infection in horses because the antibody response does not correlate with larval infection in the muscle tissue. Between 1975 and 1998, the prevalence of Trichinella infection in horses was estimated to be 3.6 per 1 million slaughtered (Pozio, 2001). In the past 2 years, the increased use of artificial digestion to detect Trichinella infection in horses in France and Italy has resulted in a greater number of infected horses being detected, resulting in a prevalence of 8.0 per 1 million slaughtered horses.
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
177
In conclusion, these results suggest that the detection of anti-Trichinella IgG in horse serum samples, irrespective of the specific technique or antigen used, cannot be considered as the method of choice and that the artificial digestion of at least 5 g of tissue from preferential-muscles, as suggested by the European Union regulation, continues to be the method of choice at the slaughterhouse for preventing equine-borne clinical trichinellosis in humans.
Acknowledgements We wish to thank the Heska Corporation, which provided us with the synthetic tyvelose glycan–BSA antigen. We also thank Marco Amati and Catherine Perret for their technical support. This study received financial support from the research project no. 9D/F entitled “Fast identification of parasitic agents in foods” of the Italian Ministry of Health, and from the research project no.1157/RI entitled “Studies of the immune response against zoonotic agents: Cryptosporidium and Trichinella” of the Istituto Superiore di Sanità, Rome, Italy. References Boireau, P., Vallée, I., Roman, T., Perret, C., Mingyuan, L., Gamble, H.R., Gajadhar, A., 2000. Trichinella in horses: a low frequency infection with high human risk. Vet. Parasitol. 93, 309–320. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of the protein-dye binding. Anal. Biochem. 72, 248–254. Bruschi, F., Pozio, E., Watanabe, N., Gomez Morales, M.A., Ito, M., Huang, Y., Binaghi, R., 1999. Anaphylactic response to parasite antigens: IgE and IgG1 independently induce death in Trichinella-infected mice. Int. Arch. Allergy Immunol. 119, 291–296. Cuperlovic, K., Djordjevic, M., Pavlovic, S., Sofronic-Milosavljevic, L., 2001. Present status of trichinellosis in Yugoslavia: Serbia. Parasite 8 (S2), S95–S97. Denkers, E.Y., Hayes, C.E., Wassom, D.L., 1991. Trichinella spiralis: influence of an immunodominant carbohydrate-associated determinant on the host antibody response repertoire. Exp. Parasitol. 72, 403–410. Gamble, H.R., Anderson, W.R., Graham, C.E., Murrell, K.D., 1983. Diagnosis of swine trichinosis by enzyme-linked immunosorbent assay (ELISA) using an excretory/secretory antigen. Vet. Parasitol. 13, 349– 361. Gamble, H.R., Bessonov, A.S., Cuperlovic, K., Gajadhar, A.A., Van Knapen, F., Nöckler, K., Schenone, H., Zhu, X., 2000. International commission on Trichinellosis: recommendations on methods for the control of Trichinella in domestic and wild animals intended for human consumption. Vet. Parasitol. 93, 393–408. Gamble, H.R., Gajadhar, A.A., Solomon, M.B., 1996. Methods for the detection of trichinellosis in horses. J. Food Prot. 59, 420–425. Gamble, H.R., Wisnewski, N., Wasson, D.L., 1997. Diagnosis of trichinellosis in swine by enzyme immunoassay, using a synthetic glycan antigen. Am. J. Vet. Res. 58, 1417–1421. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4 . Nature 227, 680–685. Lopez de Mendoza, M.E., Curtis, R.H.C., Gowen, S., 1999. Identification and characterization of excreted-secreted products and surface coat antigens of animal and plant-parasitic nematodes. Parasitology 118, 397–405. Olteanu, G., 2001. Trichinellosis in Romania: a short review over the past 20 years. Parasite 8 (2S), S98–S99. Polidori, G.A., Gramenzi, F., Piergili Fioretti, D., Ferri, N., Ranucci, S., Moretti, A., Scacchia, M., Bellelli, C., Baldelli, B., 1989. Experimental trichinellosis in horses. In: Tanner, C.E., Martinez-Fernandez, A.R., Bolas Fernandez, F. (Eds.), Trichinellosis, CSIC Press, Madrid, Spain, pp. 268–274. Pozio, E., 1998. Trichinellosis in the European union: epidemiology, ecology and economic impact. Parasitol. Today 14, 35–38.
178
E. Pozio et al. / Veterinary Parasitology 108 (2002) 163–178
Pozio, E., 2001. New patterns of Trichinella infections. Vet. Parasitol. 98, 133–148. Pozio, E., La Rosa, G., Gomez Morales, M.A., 2001. Epidemiology of human and animal trichinellosis in Italy since its discovery in 1887. Parasite 8 (2S), S106–S108. Pozio, E., Paterlini, F., Pedarra, C., Sacchi, L., Bugarini, R., Goffredo, E., Boni, P., 1999. Predilection sites of Trichinella spiralis larvae in naturally infected horses. J. Helminthol. 73, 233–237. Pozio, E., Tamburrini, A., Sacchi, L., Gomez Morales, M.A., Corona, S., Goffredo, E., La Rosa, G., 1997. Detection of Trichinella spiralis in a horse during routine examination in Italy. Int. J. Parasitol. 27, 1613–1621. Ramisz, A., Szymborski, J., Balicka-Ramisz, A., 2001. Epidemiological studies on trichinellosis among swine, wild boars and humans in Poland. Parasite 8 (2S), S90–S91. Smith, H.J., Snowdon, K.E., 1987. Experimental Trichinella infections in Ponies. Can. J. Vet. Res. 51, 415–416. Sofronic-Milosavljevic, L., Pozio, E., Patrascu, I.V., Skerovic, N., Gomez Morales, M.A., Gamble, H.R., 2001. Immunodiagnosis of Trichinella infection in the horse. Parasite 8 (25), S260–S262. Soulé, C., Dupouy-Camet, J., Georges, P., Ancelle, T., Gillet, J.P., Vaissaire, J., Delvigne, A., Plateau, E., 1989. Experimental trichinellosis in horses: biological and parasitological evaluation. Vet. Parasitol. 31, 19–36. Soulsby, E.J.L., 1965. Textbook of Veterinary Clinical Parasitology: Helminths, Vol. 1. Blackwell Scientific Publication, Oxford. Spiegel, Y., McClure, M.A., 1995. The surface-coat of plant-parasitic nematodes—chemical composition, origin and biological role—a review. J. Nematology 27, 127–134. Towbin, H., Staehelin, T., Gordon, J., 1979. Electrophoretic transfer of protein from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Nat. Acad. Sci. U.S.A. 76, 4350–4354. van Knapen, F., Franchimont, J.M., Hendrikx, W.M., Eysker, M., 1987. Experimental Trichinella spiralis infection in two horses. Vet. Arch. 31, 123–132. Voigt, W.P., Nöckler, K., Freischem, B., Henriksen, S.A., van Knapen, F., Martinez Fernandez, A., Pfeiffer, G., Pozio, E., Reuter, G., Ring, Ch., Soulé, C., Weiss, H., 1997. Detection of low levels of Trichinella spiralis in experimentally infected horses. In: Ortega-Pierres, G., Gamble, H.R., van Knapen, F., Wakelin, D. (Eds.), Trichinellosis, Nonoalco Tlateloco, Mexico, pp. 629–634. Wisnewski, N., McNeil, M., Grieve, R.B., Wassom, D.L., 1993. Characterization of novel fucosil- and tyvelosyl-containing glycoconjugates from Trichinella spiralis muscle stage larvae. Mol. Biochem. Parasitol. 61, 25–36.