- Email: [email protected]
A bead-based suspension array for the serological detection of Trichinella in pigs
The Veterinary Journal 196 (2013) 439–444
Contents lists available at SciVerse ScienceDirect
The Veterinary Journal journal homepage: www.elsevier.com/locate/tvjl
A bead-based suspension array for the serological detection of Trichinella in pigs F.J. van der Wal ⇑, R.P. Achterberg, A. Kant, C.B.M. Maassen 1 Central Veterinary Institute, Wageningen UR, Lelystad, The Netherlands
a r t i c l e
i n f o
Article history: Accepted 19 October 2012
Keywords: Trichinella spiralis Serology Pig Suspension array Human Zoonosis
a b s t r a c t The feasibility of using bead-based suspension arrays to detect serological evidence of Trichinella in pigs was assessed. Trichinella spiralis excretory–secretory antigen was covalently coupled to paramagnetic beads and used to bind serum antibodies, which were subsequently detected using anti-swine antibody. The assay was evaluated by testing pig sera from farms where trichinellosis was endemic and comparing the results with those obtained using two commercially available ELISAs. With cut-offs established by receiver operating characteristic (ROC) analysis, digestion-negative sera from a Trichinella-free population of pigs were deemed seronegative. When anti-swine antibody was replaced with protein A/G, higher test sensitivity (94% vs. 88%) at similar test specificity (95%), was achieved. The potential use of this assay in species other than swine was also demonstrated by testing human sera. Ó 2012 Elsevier Ltd. All rights reserved.
Introduction Trichinellosis is a parasitic zoonosis reported to affect as many as 11 million people worldwide (Dupouy-Camet, 2000). The disease is caused by nematodes of the genus Trichinella of which Trichinella spiralis is the species most commonly implicated in human infections (Pozio and Murrell, 2006). Infection occurs through the ingestion of raw or undercooked meat, and the clinical sequelae for humans can be severe (Capo and Despommier, 1996). While Trichinella spp. can be found in many different hosts (Pozio, 2005), the major reservoirs for human infection are domestic pigs, wild boar, and horses (Pozio and Murrell, 2006). While in North America and most of Western Europe infections due to Trichinella spp. are rare in domestic animals, in other parts of the world infection of the local animal populations is endemic (Gottstein et al., 2009). In order to control trichinellosis, the meat from >167 million pigs is subjected to mandatory inspection within the EU annually (Alban et al., 2011). This involves the artificial digestion and microscopic examination of pooled meat samples from 100 pigs to look for larvae. When such a ‘pool’ is deemed positive, individual samples are investigated by artificial digestion (Nöckler et al., 2000). Serological screening, using the excretory/secretory (E/S) antigen of T. spiralis larvae in an ELISA (Gamble et al., 1988; Nöckler et al., 1995), has been used and can be more sensitive than artificial
⇑ Corresponding author. Tel.: +31 320 238395. E-mail address: fi[email protected] (F.J. van der Wal). Present address: National Institute for Public Health (RIVM), Centre for Infectious Disease Control, Laboratory for Zoonoses and Environmental Microbiology, Bilthoven, The Netherlands. 1
1090-0233/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tvjl.2012.10.029
digestion (Gajadhar et al., 2009). However, as there is an inevitable lag between serological positivity and infection, the larval burden may not always be reflected in the test result (Nöckler et al., 1995). Nevertheless, serology can be used to monitor trichinellosis-free herds or seronegative populations in low-risk regions (Commission Regulation EC 2075/2005). In this context the use of serology to contemporaneously screen for this and other zoonotic pathogens such as Salmonella spp. (Wegener et al., 2003) in a multiplex format is an appealing goal. Although there are many examples illustrating bead-based assays can be used for multiplex serology in humans (Dias et al., 2005; van Gageldonk et al., 2008; Casabonne et al., 2009; Antonsson et al., 2010), the use of bead-based serology for veterinary applications is limited (Clavijo et al., 2006; Perkins et al., 2006; Go et al., 2008; Watson et al., 2009; Anderson et al., 2011; van der Wal et al., 2012). In this study, we first investigated whether a serological test for Trichinella infection in pigs could be established in a format suitable for multiplexing. The Luminex platform was selected as it allows the use of multiple antigens in bead-based suspension arrays (Perkins et al., 2006; Krishhan et al., 2009). Our second objective was to investigate if the use of bead-based serology was possible with an immunoglobulin binding protein such as protein A/G (Inoshima et al., 1999; Zhang et al., 2010) in order to create an assay that could potentially be used in multiple species, as previously described using protein A (Gamble et al., 1983). We compared the results of our candidate assay with those of an ELISA using sera from experimentally infected pigs, as well as from pigs from Trichinella-endemic and -free populations. In addition, the use of protein A/G instead of an anti-swine antibody was evaluated with both porcine and human sera to investigate if the assay could be used in more than one species.
440
F.J. van der Wal et al. / The Veterinary Journal 196 (2013) 439–444
Fig. 1. Graphs A and B illustrating the performance of bead-based Trichinella assay (BBA, Luminex anti-swine) in pigs relative to two commercially available ELISAs (Pourquier and Safepath) using longitudinal serum samples from two pigs (graphs A and B), experimentally infected with 250 T. spiralis larvae. Signal strength of the beadbased assay (expressed as median fluorescence intensity [MFI], left Y-axis) is plotted against days post-infection (DPI). The results of the two ELISAs are detailed on the right Y-axis: ranging from 0% to 200% for the Pourquier, and at a range of optical densities between 0 and 2 for the Safepath ELISA, respectively. Both sets of ELISA results are presented ranging from 0 to 2.
Materials and methods Selection of serum samples Sera of pigs infected with 100, 200, 1000 and 10,000 T. spiralis larvae (days 83, 84, 92, and 92 post-infection [p.i.], respectively) were kindly donated by Dr. M. Swanenburg and used to establish the assay. A longitudinal series of sera from two pigs infected with 250 larvae were kindly provided by Dr. R. Gamble. The average number of larvae/g (LPG) in samples of diaphragm at necropsy were 2.67 (day 120 p.i.) and 0.07 (day 365 p.i.), respectively (R. Gamble, personal communication). Sera from pigs experimentally infected with different amounts of T. spiralis larvae were obtained from Dr. K. Nöckler (Nöckler et al., 1995). For examination of cross reactivity, sera from pigs infected with Trichuris suis (n = 6), Toxoplasma gondii (n = 5), and Ascaris suum (n = 12), were kindly provided by Drs. H. Kringel, J. Cornelissen, F. Borgsteede, and sera from pigs infected with Sarcoptes scabiei (n = 3), Dermatophagoides pteronysinnus (n = 2), and Acarus siro (n = 2), were kindly provided by Dr. H. van der Heijden. The work was approved by the institute’s Animal Experiments Committee (DEC) in accordance with Dutch regulations (DEC reference numbers: T. spiralis, 2006148.a and 2007079.a; Toxoplasma gondii, 2006105.b; Ascaris suum, 2006134.a and 2007154.a). To evaluate the assay, a set of 244 swine sera collected in endemic regions of Argentina (Döpfer et al., 2006; Teunis et al., 2009), and a set of 120 sera from digestion-negative pigs (seronegative by ELISA [Maassen et al., 2007]), were used. To test the bead-based assay using protein A/G in species other than swine, 30 Trichinella seronegative and 31 Trichinella seropositive human sera were kindly provided by Drs. T. Garate and E. Rodriguez. The serological status of these samples had been established using indirect immunofluorescence (Sulzer, 1965) and ELISA (Escalante et al., 2004). The human samples had already been collected as part of public health
diagnostic activities and had been submitted to both 15/1999 (Spanish Law) and 1720/2007 (Spanish Royal Order on Data Protection). Control (negative) samples were obtained from volunteers after they had provided written informed consent. Coupling of E/S antigens to carboxylated beads Excretory–secretory antigen, produced as previously described (Franssen et al., 2011), was obtained from Dr. J. van der Giessen; 13.75 lg was covalently coupled to 2.5 106 carboxylated paramagnetic beads (MagPlex microspheres, Luminex). Coupling was achieved through a generic two-step carbodiimide coupling with sulfo-N-hydroxysulfosuccinimide (NHS) and 1-ethyl-3-[3-dimethylaminopropyl]carbodiimide hydrochloride (EDC) (Pierce) according to the manufacturer’s instructions. Bead-based serology Assays were performed with a Luminex 200 system (Luminex) using 1000 beads/ 50 lL PBS-T (0.05% Tween 20). Aliquots of 50 lL bead mix and 50 lL diluted serum (1:200 in PBS-T as established in pilot experiments) were mixed and incubated in a 96-well plate for 30 min in the dark at room temperature on a plate shaker (at approximately 600 rpm). After incubation, the plate was placed on a magnet (Dynal MPC-96-S; Invitrogen) for 1 min after which the beads were washed with 100 lL of PBS-T. Next, 100 lL of biotinylated secondary antibody (goat anti-swine IgG, catalogue number. 114-065-003, Jackson ImmunoResearch) in PBS-T were added (1:5000). Alternatively, biotinylated protein A/G (1:5000 of 430 lg/mL) was used: 500 lg of recombinant protein A/G (Cat. No. 21186, Pierce) was biotinylated using EZ-Link Sulfo-NHS-Biotin (Pierce) in a 1:1 ratio, according to the manufacturer’s instructions.
F.J. van der Wal et al. / The Veterinary Journal 196 (2013) 439–444 Table 1 Results of bead-based serology and ELISA on sera (n = 39) from pigs experimentally infected with various doses of Trichinella spiralis. Infecting dose, numbers of larvae/g of tissue (LPG), titre, and ELISA data were taken from Nöckler et al. (1995). Bead-based assay results (expressed as MFI) on ELISA-negative sera are highlighted in bold. ID, serum identification number; dose, number of infecting larvae; week, sampling timepoint (weeks post-infection). ID
Dose
Week
LPG
Titre
ELISA
MFI
74 120 123 127 129 441 442 223 4a 427 469 97 103 107 124 233 6a 5a 226 417 472 126 100 108 231 7a 243 224 443 451 98 116 119 234 225 484 6 4 40
50 50 50 50 50 100 100 150 150 150 150 150 150 150 150 150 150 150 500 500 500 500 500 500 500 500 500 1500 1500 1500 1500 1500 1500 1500 1500 5000 40,000 40,000 40,000
8 11 11 11 11 6 6 4 4 6 6 11 11 11 11 20 20 20 4 6 6 11 11 20 20 20 60 4 6 6 11 11 11 20 20 130 21 36 60
1.74 0.06 0.23 0.00 0.21 0.91 0.57 0.50 2.51 2.01 0.19 0.21 4.80 3.65 0.03 0.00 1.04 2.21 4.80 21.14 0.53 18.05 7.84 10.67 1.40 8.24 42.79 9.15 33.87 61.09 19.84 33.15 37.44 22.20 16.40 93.75 427.92 362 378
>1:80 1:20 >1:40 1:160 1:160 1:40 >1:40 – – 1:160 1:40 >1:40 1:160 1:160 1:640 1:320 >1:40 1:320 1:160 1:320 1:80 1:40 1:80 1:160 1:640 >1:320 >1:80 – 1:1280 1:640 1:160 1:640 1:640 >1:160 1:1280 1:80 1:640 1:320 1:320
+ + + + + + +
6938 2659 3082 6938 8340 3390 3762 295 572 3875 3962 5834 3180 3306 5103 6555 7811 6059 6320 7305 5495 3207 7176 6025 7181 7279 5465 815 4530 7520 7720 7634 7657 6058 8428 3774 3937 6452 8783
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + +
The beads were incubated for 30 min with anti-swine or protein A/G as described above, washed once, followed by 30 min incubation with 100 lL of phycoerythrin-conjugated streptavidin (Invitrogen, cat. No. SNN1007, 0.25 mg/mL) diluted 1:2000 in PBS-T. Following re-suspension in 80 lL PBS-T, the beads were analysed by measuring the fluorescence of 100 beads/sample at default settings, using xPONENT 3.1 software (Luminex). The results were expressed as median fluorescence intensity (MFI). Assay evaluation Data previously obtained with two commercial ELISAs (T. spiralis antibody test kit [Safepath Laboratories], and Trichinella serum screening [Institut Pourquier, now IDEXX]) (Maassen et al., 2007) were used to evaluate the bead-based assay. Recommended cut-offs were applied and where the result was considered ‘doubtful’ by the Pourquier ELISA (according to the manufacturer between 30% and 40%) a serum was deemed negative. For the sera collected in endemic regions, results of bead-based serology were compared to the ELISA results, using receiver operating characteristic (ROC) analysis (GraphPad Prism 5.04, GraphPad Software). Lists were generated with cut-offs and the corresponding sensitivities and specificities, and at selected specificities, the sensitivities and cut-offs were extracted and used to calculate the efficiencies and Cohen’s kappa (Mackinnon, 2000). To compare serological assays, the following definitions were used. Sensitivity is defined as true seropositives (i.e. samples that are positive on the criterion [a commercial ELISA] and the studied test [e.g. the bead-based assay]) divided by the sum of true positives and false negatives (i.e. samples positive on the criterion, but negative on the studied test). Specificity is defined as true negatives (i.e. samples that are negative on the criterion and the studied test), divided by the sum of true negatives and false positives (samples negative on the criterion, but positive in the studied test). Efficiency reflects overall agreement, and is defined as the sum
441
of true positive and true negative samples divided by the total number of samples. Cohen’s kappa is a measure of the level of agreement between assays (Mackinnon, 2000), where 0.6–0.8 indicates substantial, and 0.8–1 almost perfect agreement.
Results The test was established using experimentally derived sera. Then the ‘longitudinal sera’ from the two pigs infected with a low dose of 250 larvae were tested and the results were compared with those of two commercial ELISAs (Fig. 1). With the bead-based assay an antibody response was detected that changed over time, and was similar to the results obtained with the ‘Safepath’ and ‘Pourquier’ ELISAs. Sera from experimentally infected pigs were tested where the ELISA results and the infecting larval burdens were known (Nöckler et al., 1995). With the bead-based assay, and using an anti-swine antibody for detection, two groups were identified with signals <1000 MFI (median fluorescence intensity) in the first group (n = 3), and signals ranging from 2659 to 8783 MFI in the second (n = 36) (Table 1). These groups were seronegative and seropositive in the ELISA as described by Nöckler et al. (1995). There was no correlation between the larval burden and the results of the ELISA or the bead-based assay. To investigate cross-reactivity of the bead-based assay with antibodies against other parasites, 31 sera from pigs infected with the following parasites were tested: Trichuris suis, Toxoplasma gondii, Ascaris suum and siro, Sarcoptes scabiei, and Dermatophagoides pteronysinnus. None of the tested sera registered high signals in the bead-based assay (<480 MFI) (data not shown). The Trichinella serological bead-based assay was evaluated with 244 porcine sera from endemic regions, and the results compared with the commercial ELISAs using ROC analysis. For a range of specificities, the corresponding sensitivities, efficiencies and Cohen’s kappa are presented in Table 2. The highest sensitivity of the bead-based assay with an almost perfect test agreement (Cohen’s kappa >0.8), was 87% sensitivity at 95% specificity in comparison with the Pourquier ELISA, and 98% sensitivity at 95% specificity relative to the Safepath ELISA. This finding suggested the bead-based assay correlated better with the Safepath rather than the Pourquier ELISA. The sensitivity of the Safepath ELISA relative to the Pourquier ELISA was 79% with 100% specificity (95% efficiency, Cohen’s kappa 0.85), indicating that a substantial number of samples (12/56) positive in the Pourquier ELISA were negative in the Safepath ELISA. Thus, given that the Pourquier ELISA was more sensitive than the Safepath ELISA, the Pourquier ELISA was used for further comparisons in order to prevent the use of potential false negatives when comparing the results of the bead-based serological assay and the ELISA. To investigate the possibilities of bead-based serology using a generic Ig-binding protein instead of a species-specific antibody, the potential use of using protein A/G was explored. Firstly, protein A/G reacted with Ig from a broad panel of mammalian species in an ELISA (data not shown). When protein A/G was tested with the sera detailed in Table 1 (Nöckler et al., 1995), the same two groups were identified: those with MFI values <150 (n = 3), and those with MFI values ranging from 999 to 8962 (n = 36) (data not shown). When protein A/G was tested with the panel of 31 sera compiled to investigate cross-reactivity, none of the sera registered high signals (all <260 MFI). A subset (n = 88) of the field sera collected in endemic regions of Argentina was tested with the bead-based assay, using anti-swine antibody and protein A/G for detection and the results compared with the Pourquier ELISA by ROC analysis. For a range of specificities, the corresponding sensitivities, efficiencies and Cohen’s kappa are presented in Table 3. In comparison to the Pourquier ELISA, the highest sensitivities of the bead-based assay with a Cohen’s kappa of >0.8, were 91% sensitivity for anti-swine antibody at 91% speci-
442
F.J. van der Wal et al. / The Veterinary Journal 196 (2013) 439–444
Table 2 Performance of bead-based Trichinella assay (BBA) in pigs relative to two commercially available ELISAs (Pourquier and Safepath). Sera (n = 244) from a Trichinella-endemic region were tested with the BBA using anti-swine antibodies. At a range of specificities (85%, 90%, 95%, 99%, and 100%), the sensitivities, efficiencies, Cohen’s kappa, and corresponding cut-offs (MFI), are detailed as determined by ROC analysis. BBA vs. Pourquier ELISA Specificity (%) Sensitivity (%) Efficiency (%) Cohen’s kappa Cut-off (MFI)
85 96 88 0.7 961
90 91 90 0.74 1077
BBA vs. Safepath ELISA 95 87 93 0.81 1242
99 84 95 0.86 1432
100 78 95 0.85 1634
85 98 87 0.65 1053
90 98 91 0.75 1151
95 98 95 0.86 1350
99 95 98 0.94 1634
100 53 92 0.65 3481
Table 3 Performance of bead-based Trichinella assay (BBA) in pigs relative to a commercially available ELISA (Pourquier). Sera (n = 88) from a Trichinella-endemic region were tested with the BBA using anti-swine antibody (ab) and protein A/G, respectively. At a range of specificities (86%, 91%, 95%, 98%, and 100%), the sensitivities, efficiencies, Cohen’s kappa, and corresponding cut-offs (MFI), are detailed. BBA (anti-swine ab) vs. Pourquier Specificity (%) Sensitivity (%) Efficiency (%) Cohen’s kappa Cut-off (MFI)
86 97 90 0.79 855
91 91 91 0.81 1071
95 88 92 0.83 1137
BBA (protein A/G) vs. Pourquier 98 78 91 0.80 1392
ficity (cut-off 1071 MFI), and 94% sensitivity for protein A/G at 91% specificity (cut-off 589 MFI). When using protein A/G at 94% sensitivity the specificity of the bead-based assay increased to 95% at a cut-off of 692 MFI – the equivalent sensitivity for the anti-swine antibody variant of the assay was only 88%. These results demonstrate that higher sensitivities were achieved using protein A/G than using anti-swine antibody in the bead-based assay. For the subsequent analysis of sera from a negative population, cut-offs of 1071 MFI for anti-swine antibody and 692 for protein A/G were used, respectively. The bead-based assay was further evaluated in a Trichinellanegative population of 120 slaughter pigs. Both anti-swine antibody and protein A/G were used for detection. For anti-swine antibodies, signals varied from 49 to 683 MFI, and for protein A/ G from 18 to 79 (Fig. 2), all below the selected cut-offs of 1071 MFI for anti-swine antibody and 692 MFI for protein A/G, respectively. This result demonstrated that all the tested sera were negative in the bead-based assay, regardless of the detection molecule used (i.e. anti-swine antibody or protein A/G). To demonstrate its potential use in species other than pigs, 30 Trichinella seronegative and 31 seropositive human sera were tested using the bead-based assay incorporating protein A/G. The seronegative sera gave signals ranging from 4 to 219 MFI, whereas the seropositive sera registered values from 7 to 6118 MFI (Fig. 3). At a cut-off of 30 MFI, and with a sensitivity of 90%, the specificity was 93% (efficiency 92%, Cohen’s kappa 0.84), demonstrating that human IgG can be detected using protein A/G.
100 78 92 0.82 1433
86 97 90 0.79 319
91 94 92 0.83 589
95 94 94 0.88 692
98 78 91 0.80 910
100 69 89 0.74 1515
Fig. 2. Graph illustrating that sera (n = 120) from a Trichinella-negative population of pigs were seronegative using the bead-based assay, using anti-swine antibodies or protein A/G for detection. Signal strength of the assay (expressed as MFI), is plotted against (arbitrary) serum numbers.
Discussion The first objective of this study was to investigate whether a bead-based assay could be used to carry out serology in pigs using Trichinella as the test pathogen. Evaluation of previously wellcategorised sera (Nöckler et al., 1995) demonstrated that all seropositive animals registered a strong positive result using the bead-based assay, but there was no obvious correlation between the dose of larvae used to infect the pigs/larval burden and this result is as described with a conventional ELISA (Nöckler et al., 1995). The bead-based assay was further evaluated using ‘field’ sera from a region where trichinellosis was endemic (Döpfer et al., 2006; Teunis et al., 2009), and a Trichinella-free pig population. Using the porcine sera from the endemic region, the results of the bead-based assay and commercial ELISAs were compared using
Fig. 3. Graph illustrating the results of serology on human sera (30 seronegative and 31 seropositive, respectively) using the bead-based assay with protein A/G for detection. Signal strength of the assay (expressed as MFI), is plotted against (arbitrary) serum numbers.
ROC analyses and cut-offs were established. With these cut-offs, all the pigs from the Trichinella-free region were found to be seronegative, as expected.
F.J. van der Wal et al. / The Veterinary Journal 196 (2013) 439–444
The second objective of this research was to investigate if protein A/G could be used in the bead-based test format. Relative to the Pourquier ELISA, use of this protein in the bead-based assay resulted in a higher sensitivity (at a specificity of 95%) than when an anti-swine antibody was used. Furthermore, we demonstrated that protein A/G could be used to differentiate seronegative from seropositive human serum. Although several ELISAs are available to serologically screen for Trichinella, this approach is not commonly practiced. In the present work, when the effect of the cut-off on test sensitivity and specificity was evaluated for our bead-based assay, a decreased cut-off resulted in an increased sensitivity (less false negatives), and decreased specificity (more false positives), respectively. Given that this type of serological screening of pigs is used to monitor Trichinella-free/seronegative populations, test sensitivity may be prioritised over specificity in order to minimise the number of false negatives even though this will lead to an increase in the number of false positives. This seems a pragmatic approach since false positives can be verified with an appropriate confirmatory test whereas the false negatives may remain undetected. Since Trichinella occurs in a wide range of mammals (Pozio, 2005; Pozio and Murrell, 2006), we evaluated the capacity of our bead-based assay to serologically screen other species by using a more ubiquitous detection system focussed on protein A/G (Inoshima et al., 1999; Zhang et al., 2010). Our results demonstrated that this assay can also be performed using protein A/G which serendipitously increases the sensitivity relative to using anti-swine antibody. This increased sensitivity was largely due to a reduction in background binding (i.e. the signal of the ELISAnegative samples) – the MFI values of the positive samples were actually lower when protein A/G was used. Thus, used in this way, protein A/G lowers the assay cut-off point, increasing the sensitivity of the test when compared to an ELISA at a given specificity. Protein A/G could also be successfully incorporated into our beadbased assay when human serum was tested. Preliminary observations (data not shown) suggest that some sera, positive by ELISA and/or IFF, contain other Ig such as A and M, only weakly bound by protein A/G. The suspected presence of such immunoglobulins is supported by previous observations (Reiterova et al., 2007), and suggests that for the serological diagnosis of trichinellosis in humans the use of Ig-specific secondary antibodies may be preferred. Nevertheless, the ‘proof of principle’ of using protein A/G in the assay for use in species other than pigs was demonstrated, and the feasibility of such an approach in other serological assays and species remains to be determined.
Conclusions Although there are many examples illustrating how the Luminex platform can be used for multiplex serology in humans, the previous use of bead-based serology in veterinary applications is limited. Our results highlight the feasibility of using a bead-based assay to serologically screen pigs for evidence of Trichinella. This assay can serve as a ‘scaffold’ test system that can potentially be expanded to include other serological assays by adding other antigens. Such an approach could be used as a low cost method of screening pigs at slaughter for a range of pathogens including Salmonella spp., swine vesicular disease virus, and Aujeszky’s disease virus, or to differentiate between vaccinated and non-vaccinated animals. Recently, a duplex bead-based assay has been developed that can simultaneously detect antibodies against two proteins of the Rift valley fever virus, facilitating the differentiation of vaccinated from non-vaccinated animals. Bead-based methodology has considerable potential, and is likely to find a range of useful and novel veterinary applications.
443
Conflict of interest statement None of the authors of this paper has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.
Acknowledgments The authors would like to thank: J. van der Giessen for providing E/S antigen and for critical reading of the manuscript; R. Gamble, K. Nöckler, H. Kringel, and H. van der Heijden for providing us with porcine sera, D. Döpfer for Argentine field sera, M. Swanenburg, J. Cornelissen, and F. Borgsteede for sera from animal experiments performed at CVI, T. Garate and E. Rodriguez for human serum; and J. Bergervoet and J. Peters for helpful discussions in relation to the Luminex platform. This research was partly funded by the Dutch Ministry of Economic Affairs, Agriculture and Innovation (Strategic Research Program Food Safety, Monitoring and Detection KB-06-005), and by the European Union Seventh Framework Programme (FP7/2007-2013) under Grant Agreement Number 222633 (WildTech).
References Alban, L., Pozio, E., Boes, J., Boireau, P., Boue, F., Claes, M., Cook, A.J., Dorny, P., Enemark, H.L., van der Giessen, J., Hunt, K.R., Howell, M., Kirjusina, M., Nöckler, K., Rossi, P., Smith, G.C., Snow, L., Taylor, M.A., Theodoropoulos, G., Vallee, I., Viera-Pinto, M.M., Zimmer, I.A., 2011. Towards a standardised surveillance for Trichinella in the European Union. Preventive Veterinary Medicine 99, 148–160. Anderson, S., Wakeley, P., Wibberley, G., Webster, K., Sawyer, J., 2011. Development and evaluation of a Luminex multiplex serology assay to detect antibodies to bovine herpes virus, parainfluenza virus, bovine viral diarrhoea virus, and bovine respiratory syncytial virus, with comparison to existing ELISA detection methods. Journal of Immunological Methods 366, 79–88. Antonsson, A., Green, A.C., Mallitt, K.A., O’Rourke, P.K., Pawlita, M., Waterboer, T., Neale, R.E., 2010. Prevalence and stability of antibodies to the BK and JC polyomaviruses: A long-term longitudinal study of Australians. The Journal of General Virology 91, 1849–1853. Capo, V., Despommier, D.D., 1996. Clinical aspects of infection with Trichinella spp. Clinical Microbiology Reviews 9, 47–54. Casabonne, D., Waterboer, T., Michael, K.M., Pawlita, M., Mitchell, L., Newton, R., Harwood, C., Proby, C., 2009. The seroprevalence of human papillomavirus by immune status and by ethnicity in London. Infectious Agents and Cancer 4, 14. Clavijo, A., Hole, K., Li, M., Collignon, B., 2006. Simultaneous detection of antibodies to foot-and-mouth disease non-structural proteins 3ABC, 3D, 3A and 3B by a multiplexed Luminex assay to differentiate infected from vaccinated cattle. Vaccine 24, 1693–1704. Dias, D., Van Doren, J., Schlottmann, S., Kelly, S., Puchalski, D., Ruiz, W., Boerckel, P., Kessler, J., Antonello, J.M., Green, T., Brown, M., Smith, J., Chirmule, N., Barr, E., Jansen, K.U., Esser, M.T., 2005. Optimization and validation of a multiplexed Luminex assay to quantify antibodies to neutralizing epitopes on human papillomaviruses 6, 11, 16, and 18. Clinical and Diagnostic Laboratory Immunology 12, 959–969. Döpfer, D., Maassen, K., Achterberg, R.P., van der Giessen, J., Teunis, P., Buist, W., Molina, V., Guarnera, E., Gonzales, M., Krivokapich, S., Rodriguez, M., Peralta, J.L., Trabattoni, H., Larrieu, E., Engel, B., 2006. Diagnostic test evaluation for Trichinella infections in pigs – A renewed interaction between diagnostic test data and their statistical analysis. In: Proceedings of the 11th Symposium of the International Society for Veterinary Epidemiology and Economics, vol. 11, pp. 649–651. Dupouy-Camet, J., 2000. Trichinellosis: A worldwide zoonosis. Veterinary Parasitology 93, 191–200. Escalante, M., Romaris, F., Rodriguez, M., Rodriguez, E., Leiro, J., Garate, M.T., Ubeira, F.M., 2004. Evaluation of Trichinella spiralis larva group 1 antigens for serodiagnosis of human trichinellosis. Journal of Clinical Microbiology 42, 4060–4066. Franssen, F.F., Fonville, M., Takumi, K., Vallee, I., Grasset, A., Koedam, M.A., Wester, P.W., Boireau, P., van der Giessen, J.W., 2011. Antibody response against Trichinella spiralis in experimentally infected rats is dose dependent. Veterinary Research 42, 113. Gajadhar, A.A., Pozio, E., Gamble, H.R., Nöckler, K., Maddox-Hyttel, C., Forbes, L.B., Vallee, I., Rossi, P., Marinculic, A., Boireau, P., 2009. Trichinella diagnostics and control: Mandatory and best practices for ensuring food safety. Veterinary Parasitology 159, 197–205. 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. Veterinary Parasitology 13, 349–361.
444
F.J. van der Wal et al. / The Veterinary Journal 196 (2013) 439–444
Gamble, H.R., Rapic, D., Marinculic, A., Murrell, K.D., 1988. Evaluation of excretory– secretory antigens for the serodiagnosis of swine trichinellosis. Veterinary Parasitology 30, 131–137. Go, Y.Y., Wong, S.J., Branscum, A.J., Demarest, V.L., Shuck, K.M., Vickers, M.L., Zhang, J., McCollum, W.H., Timoney, P.J., Balasuriya, U.B., 2008. Development of a fluorescent-microsphere immunoassay for detection of antibodies specific to equine arteritis virus and comparison with the virus neutralization test. Clinical and Vaccine Immunology 15, 76–87. Gottstein, B., Pozio, E., Nöckler, K., 2009. Epidemiology, diagnosis, treatment, and control of trichinellosis. Clinical Microbiology Reviews 22, 127–145. Inoshima, Y., Shimizu, S., Minamoto, N., Hirai, K., Sentsui, H., 1999. Use of protein AG in an enzyme-linked immunosorbent assay for screening for antibodies against parapoxvirus in wild animals in Japan. Clinical Diagnostic Laboratory Immunology 6, 388–391. Krishhan, V.V., Khan, I.H., Luciw, P.A., 2009. Multiplexed microbead immunoassays by flow cytometry for molecular profiling: Basic concepts and proteomics applications. Critical Reviews in Biotechnology 29, 29–43. Maassen, C.B.M., Achterberg, R.P., Engel, B., van der Giessen, J., Teunis, P., van der Heijden, H.M.J.F., Haughey, S., Molina, V., Peralta, J.L., Döpfer, D., 2007. Evaluation of serological test for Trichinella infections in pigs. In: Proceedings of 7th International Safepork Symposium, Padova, Italy, pp. 211–213. Mackinnon, A., 2000. A spreadsheet for the calculation of comprehensive statistics for the assessment of diagnostic tests and inter-rater agreement. Computers in Biology and Medicine 30, 127–134. Nöckler, K., Voigt, W.P., Protz, D., Miko, A., Ziedler, K., 1995. Diagnosis of trichinellosis in living pigs using indirect ELISA. Berliner und Munchener Tierärztliche Wochenschrift 108, 167–174. Nöckler, K., Pozio, E., Voigt, W.P., Heidrich, J., 2000. Detection of Trichinella infection in food animals. Veterinary Parasitology 93, 335–350. Perkins, J., Clavijo, A., Hindson, B.J., Lenhoff, R.J., McBride, M.T., 2006. Multiplexed detection of antibodies to nonstructural proteins of foot-and-mouth disease virus. Analytical Chemistry 78, 5462–5468.
Pozio, E., 2005. The broad spectrum of Trichinella hosts: From cold- to warmblooded animals. Veterinary Parasitology 132, 3–11. Pozio, E., Murrell, K.D., 2006. Systematics and epidemiology of Trichinella. Advances in Parasitology 63, 367–439. Reiterova, K., Kincekova, J., Snabel, V., Marucci, G., Pozio, E., Dubinsky, P., 2007. Trichinella spiralis-outbreak in the Slovak Republic. Infection 35, 89–93. Sulzer, A.J., 1965. Indirect fluorescent antibody tests for parasitic diseases. I. Preparation of a stable antigen from larvae of Trichinella spiralis. Journal of Parasitology 51, 717–721. Teunis, P.F., Fonville, M.T., Dopfer, D.D., Eijck, I.A., Molina, V., Guarnera, E., van der Giessen, J.W., 2009. Usefulness of sero-surveillance for Trichinella infections in animal populations. Veterinary Parasitology 159, 345–349. van der Wal, F.J., Achterberg, R.P., Boer, S.M., Boshra, H., Brun, A., Maassen, C.B., Kortekaas, J., 2012. Bead-based suspension array for simultaneous detection of antibodies against the Rift Valley fever virus nucleocapsid and Gn glycoprotein. Journal of Virological Methods 183, 99–105. van Gageldonk, P.G., van Schaijk, F.G., van der Klis, F.R., Berbers, G.A., 2008. Development and validation of a multiplex immunoassay for the simultaneous determination of serum antibodies to Bordetella pertussis, diphtheria and tetanus. Journal of Immunological Methods 335, 79–89. Watson, D.S., Reddy, S.M., Brahmakshatriya, V., Lupiani, B., 2009. A multiplexed immunoassay for detection of antibodies against avian influenza virus. Journal of Immunological Methods 340, 123–131. Wegener, H.C., Hald, T., LoFoWong, D., Madsen, M., Korsgaard, H., Bager, F., GernerSmidt, P., Molbak, K., 2003. Salmonella control programs in Denmark. Emerging Infectious Diseases 9, 774–780. Zhang, D., Wang, Z., Fang, R., Nie, H., Feng, H., Zhou, Y., Zhao, J., 2010. Use of protein AG in an enzyme-linked immunosorbent assay for serodiagnosis of Toxoplasma gondii infection in four species of animals. Clinical Vaccine Immunology 17, 485–486.
Contents lists available at SciVerse ScienceDirect
The Veterinary Journal journal homepage: www.elsevier.com/locate/tvjl
A bead-based suspension array for the serological detection of Trichinella in pigs F.J. van der Wal ⇑, R.P. Achterberg, A. Kant, C.B.M. Maassen 1 Central Veterinary Institute, Wageningen UR, Lelystad, The Netherlands
a r t i c l e
i n f o
Article history: Accepted 19 October 2012
Keywords: Trichinella spiralis Serology Pig Suspension array Human Zoonosis
a b s t r a c t The feasibility of using bead-based suspension arrays to detect serological evidence of Trichinella in pigs was assessed. Trichinella spiralis excretory–secretory antigen was covalently coupled to paramagnetic beads and used to bind serum antibodies, which were subsequently detected using anti-swine antibody. The assay was evaluated by testing pig sera from farms where trichinellosis was endemic and comparing the results with those obtained using two commercially available ELISAs. With cut-offs established by receiver operating characteristic (ROC) analysis, digestion-negative sera from a Trichinella-free population of pigs were deemed seronegative. When anti-swine antibody was replaced with protein A/G, higher test sensitivity (94% vs. 88%) at similar test specificity (95%), was achieved. The potential use of this assay in species other than swine was also demonstrated by testing human sera. Ó 2012 Elsevier Ltd. All rights reserved.
Introduction Trichinellosis is a parasitic zoonosis reported to affect as many as 11 million people worldwide (Dupouy-Camet, 2000). The disease is caused by nematodes of the genus Trichinella of which Trichinella spiralis is the species most commonly implicated in human infections (Pozio and Murrell, 2006). Infection occurs through the ingestion of raw or undercooked meat, and the clinical sequelae for humans can be severe (Capo and Despommier, 1996). While Trichinella spp. can be found in many different hosts (Pozio, 2005), the major reservoirs for human infection are domestic pigs, wild boar, and horses (Pozio and Murrell, 2006). While in North America and most of Western Europe infections due to Trichinella spp. are rare in domestic animals, in other parts of the world infection of the local animal populations is endemic (Gottstein et al., 2009). In order to control trichinellosis, the meat from >167 million pigs is subjected to mandatory inspection within the EU annually (Alban et al., 2011). This involves the artificial digestion and microscopic examination of pooled meat samples from 100 pigs to look for larvae. When such a ‘pool’ is deemed positive, individual samples are investigated by artificial digestion (Nöckler et al., 2000). Serological screening, using the excretory/secretory (E/S) antigen of T. spiralis larvae in an ELISA (Gamble et al., 1988; Nöckler et al., 1995), has been used and can be more sensitive than artificial
⇑ Corresponding author. Tel.: +31 320 238395. E-mail address: fi[email protected] (F.J. van der Wal). Present address: National Institute for Public Health (RIVM), Centre for Infectious Disease Control, Laboratory for Zoonoses and Environmental Microbiology, Bilthoven, The Netherlands. 1
1090-0233/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tvjl.2012.10.029
digestion (Gajadhar et al., 2009). However, as there is an inevitable lag between serological positivity and infection, the larval burden may not always be reflected in the test result (Nöckler et al., 1995). Nevertheless, serology can be used to monitor trichinellosis-free herds or seronegative populations in low-risk regions (Commission Regulation EC 2075/2005). In this context the use of serology to contemporaneously screen for this and other zoonotic pathogens such as Salmonella spp. (Wegener et al., 2003) in a multiplex format is an appealing goal. Although there are many examples illustrating bead-based assays can be used for multiplex serology in humans (Dias et al., 2005; van Gageldonk et al., 2008; Casabonne et al., 2009; Antonsson et al., 2010), the use of bead-based serology for veterinary applications is limited (Clavijo et al., 2006; Perkins et al., 2006; Go et al., 2008; Watson et al., 2009; Anderson et al., 2011; van der Wal et al., 2012). In this study, we first investigated whether a serological test for Trichinella infection in pigs could be established in a format suitable for multiplexing. The Luminex platform was selected as it allows the use of multiple antigens in bead-based suspension arrays (Perkins et al., 2006; Krishhan et al., 2009). Our second objective was to investigate if the use of bead-based serology was possible with an immunoglobulin binding protein such as protein A/G (Inoshima et al., 1999; Zhang et al., 2010) in order to create an assay that could potentially be used in multiple species, as previously described using protein A (Gamble et al., 1983). We compared the results of our candidate assay with those of an ELISA using sera from experimentally infected pigs, as well as from pigs from Trichinella-endemic and -free populations. In addition, the use of protein A/G instead of an anti-swine antibody was evaluated with both porcine and human sera to investigate if the assay could be used in more than one species.
440
F.J. van der Wal et al. / The Veterinary Journal 196 (2013) 439–444
Fig. 1. Graphs A and B illustrating the performance of bead-based Trichinella assay (BBA, Luminex anti-swine) in pigs relative to two commercially available ELISAs (Pourquier and Safepath) using longitudinal serum samples from two pigs (graphs A and B), experimentally infected with 250 T. spiralis larvae. Signal strength of the beadbased assay (expressed as median fluorescence intensity [MFI], left Y-axis) is plotted against days post-infection (DPI). The results of the two ELISAs are detailed on the right Y-axis: ranging from 0% to 200% for the Pourquier, and at a range of optical densities between 0 and 2 for the Safepath ELISA, respectively. Both sets of ELISA results are presented ranging from 0 to 2.
Materials and methods Selection of serum samples Sera of pigs infected with 100, 200, 1000 and 10,000 T. spiralis larvae (days 83, 84, 92, and 92 post-infection [p.i.], respectively) were kindly donated by Dr. M. Swanenburg and used to establish the assay. A longitudinal series of sera from two pigs infected with 250 larvae were kindly provided by Dr. R. Gamble. The average number of larvae/g (LPG) in samples of diaphragm at necropsy were 2.67 (day 120 p.i.) and 0.07 (day 365 p.i.), respectively (R. Gamble, personal communication). Sera from pigs experimentally infected with different amounts of T. spiralis larvae were obtained from Dr. K. Nöckler (Nöckler et al., 1995). For examination of cross reactivity, sera from pigs infected with Trichuris suis (n = 6), Toxoplasma gondii (n = 5), and Ascaris suum (n = 12), were kindly provided by Drs. H. Kringel, J. Cornelissen, F. Borgsteede, and sera from pigs infected with Sarcoptes scabiei (n = 3), Dermatophagoides pteronysinnus (n = 2), and Acarus siro (n = 2), were kindly provided by Dr. H. van der Heijden. The work was approved by the institute’s Animal Experiments Committee (DEC) in accordance with Dutch regulations (DEC reference numbers: T. spiralis, 2006148.a and 2007079.a; Toxoplasma gondii, 2006105.b; Ascaris suum, 2006134.a and 2007154.a). To evaluate the assay, a set of 244 swine sera collected in endemic regions of Argentina (Döpfer et al., 2006; Teunis et al., 2009), and a set of 120 sera from digestion-negative pigs (seronegative by ELISA [Maassen et al., 2007]), were used. To test the bead-based assay using protein A/G in species other than swine, 30 Trichinella seronegative and 31 Trichinella seropositive human sera were kindly provided by Drs. T. Garate and E. Rodriguez. The serological status of these samples had been established using indirect immunofluorescence (Sulzer, 1965) and ELISA (Escalante et al., 2004). The human samples had already been collected as part of public health
diagnostic activities and had been submitted to both 15/1999 (Spanish Law) and 1720/2007 (Spanish Royal Order on Data Protection). Control (negative) samples were obtained from volunteers after they had provided written informed consent. Coupling of E/S antigens to carboxylated beads Excretory–secretory antigen, produced as previously described (Franssen et al., 2011), was obtained from Dr. J. van der Giessen; 13.75 lg was covalently coupled to 2.5 106 carboxylated paramagnetic beads (MagPlex microspheres, Luminex). Coupling was achieved through a generic two-step carbodiimide coupling with sulfo-N-hydroxysulfosuccinimide (NHS) and 1-ethyl-3-[3-dimethylaminopropyl]carbodiimide hydrochloride (EDC) (Pierce) according to the manufacturer’s instructions. Bead-based serology Assays were performed with a Luminex 200 system (Luminex) using 1000 beads/ 50 lL PBS-T (0.05% Tween 20). Aliquots of 50 lL bead mix and 50 lL diluted serum (1:200 in PBS-T as established in pilot experiments) were mixed and incubated in a 96-well plate for 30 min in the dark at room temperature on a plate shaker (at approximately 600 rpm). After incubation, the plate was placed on a magnet (Dynal MPC-96-S; Invitrogen) for 1 min after which the beads were washed with 100 lL of PBS-T. Next, 100 lL of biotinylated secondary antibody (goat anti-swine IgG, catalogue number. 114-065-003, Jackson ImmunoResearch) in PBS-T were added (1:5000). Alternatively, biotinylated protein A/G (1:5000 of 430 lg/mL) was used: 500 lg of recombinant protein A/G (Cat. No. 21186, Pierce) was biotinylated using EZ-Link Sulfo-NHS-Biotin (Pierce) in a 1:1 ratio, according to the manufacturer’s instructions.
F.J. van der Wal et al. / The Veterinary Journal 196 (2013) 439–444 Table 1 Results of bead-based serology and ELISA on sera (n = 39) from pigs experimentally infected with various doses of Trichinella spiralis. Infecting dose, numbers of larvae/g of tissue (LPG), titre, and ELISA data were taken from Nöckler et al. (1995). Bead-based assay results (expressed as MFI) on ELISA-negative sera are highlighted in bold. ID, serum identification number; dose, number of infecting larvae; week, sampling timepoint (weeks post-infection). ID
Dose
Week
LPG
Titre
ELISA
MFI
74 120 123 127 129 441 442 223 4a 427 469 97 103 107 124 233 6a 5a 226 417 472 126 100 108 231 7a 243 224 443 451 98 116 119 234 225 484 6 4 40
50 50 50 50 50 100 100 150 150 150 150 150 150 150 150 150 150 150 500 500 500 500 500 500 500 500 500 1500 1500 1500 1500 1500 1500 1500 1500 5000 40,000 40,000 40,000
8 11 11 11 11 6 6 4 4 6 6 11 11 11 11 20 20 20 4 6 6 11 11 20 20 20 60 4 6 6 11 11 11 20 20 130 21 36 60
1.74 0.06 0.23 0.00 0.21 0.91 0.57 0.50 2.51 2.01 0.19 0.21 4.80 3.65 0.03 0.00 1.04 2.21 4.80 21.14 0.53 18.05 7.84 10.67 1.40 8.24 42.79 9.15 33.87 61.09 19.84 33.15 37.44 22.20 16.40 93.75 427.92 362 378
>1:80 1:20 >1:40 1:160 1:160 1:40 >1:40 – – 1:160 1:40 >1:40 1:160 1:160 1:640 1:320 >1:40 1:320 1:160 1:320 1:80 1:40 1:80 1:160 1:640 >1:320 >1:80 – 1:1280 1:640 1:160 1:640 1:640 >1:160 1:1280 1:80 1:640 1:320 1:320
+ + + + + + +
6938 2659 3082 6938 8340 3390 3762 295 572 3875 3962 5834 3180 3306 5103 6555 7811 6059 6320 7305 5495 3207 7176 6025 7181 7279 5465 815 4530 7520 7720 7634 7657 6058 8428 3774 3937 6452 8783
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + +
The beads were incubated for 30 min with anti-swine or protein A/G as described above, washed once, followed by 30 min incubation with 100 lL of phycoerythrin-conjugated streptavidin (Invitrogen, cat. No. SNN1007, 0.25 mg/mL) diluted 1:2000 in PBS-T. Following re-suspension in 80 lL PBS-T, the beads were analysed by measuring the fluorescence of 100 beads/sample at default settings, using xPONENT 3.1 software (Luminex). The results were expressed as median fluorescence intensity (MFI). Assay evaluation Data previously obtained with two commercial ELISAs (T. spiralis antibody test kit [Safepath Laboratories], and Trichinella serum screening [Institut Pourquier, now IDEXX]) (Maassen et al., 2007) were used to evaluate the bead-based assay. Recommended cut-offs were applied and where the result was considered ‘doubtful’ by the Pourquier ELISA (according to the manufacturer between 30% and 40%) a serum was deemed negative. For the sera collected in endemic regions, results of bead-based serology were compared to the ELISA results, using receiver operating characteristic (ROC) analysis (GraphPad Prism 5.04, GraphPad Software). Lists were generated with cut-offs and the corresponding sensitivities and specificities, and at selected specificities, the sensitivities and cut-offs were extracted and used to calculate the efficiencies and Cohen’s kappa (Mackinnon, 2000). To compare serological assays, the following definitions were used. Sensitivity is defined as true seropositives (i.e. samples that are positive on the criterion [a commercial ELISA] and the studied test [e.g. the bead-based assay]) divided by the sum of true positives and false negatives (i.e. samples positive on the criterion, but negative on the studied test). Specificity is defined as true negatives (i.e. samples that are negative on the criterion and the studied test), divided by the sum of true negatives and false positives (samples negative on the criterion, but positive in the studied test). Efficiency reflects overall agreement, and is defined as the sum
441
of true positive and true negative samples divided by the total number of samples. Cohen’s kappa is a measure of the level of agreement between assays (Mackinnon, 2000), where 0.6–0.8 indicates substantial, and 0.8–1 almost perfect agreement.
Results The test was established using experimentally derived sera. Then the ‘longitudinal sera’ from the two pigs infected with a low dose of 250 larvae were tested and the results were compared with those of two commercial ELISAs (Fig. 1). With the bead-based assay an antibody response was detected that changed over time, and was similar to the results obtained with the ‘Safepath’ and ‘Pourquier’ ELISAs. Sera from experimentally infected pigs were tested where the ELISA results and the infecting larval burdens were known (Nöckler et al., 1995). With the bead-based assay, and using an anti-swine antibody for detection, two groups were identified with signals <1000 MFI (median fluorescence intensity) in the first group (n = 3), and signals ranging from 2659 to 8783 MFI in the second (n = 36) (Table 1). These groups were seronegative and seropositive in the ELISA as described by Nöckler et al. (1995). There was no correlation between the larval burden and the results of the ELISA or the bead-based assay. To investigate cross-reactivity of the bead-based assay with antibodies against other parasites, 31 sera from pigs infected with the following parasites were tested: Trichuris suis, Toxoplasma gondii, Ascaris suum and siro, Sarcoptes scabiei, and Dermatophagoides pteronysinnus. None of the tested sera registered high signals in the bead-based assay (<480 MFI) (data not shown). The Trichinella serological bead-based assay was evaluated with 244 porcine sera from endemic regions, and the results compared with the commercial ELISAs using ROC analysis. For a range of specificities, the corresponding sensitivities, efficiencies and Cohen’s kappa are presented in Table 2. The highest sensitivity of the bead-based assay with an almost perfect test agreement (Cohen’s kappa >0.8), was 87% sensitivity at 95% specificity in comparison with the Pourquier ELISA, and 98% sensitivity at 95% specificity relative to the Safepath ELISA. This finding suggested the bead-based assay correlated better with the Safepath rather than the Pourquier ELISA. The sensitivity of the Safepath ELISA relative to the Pourquier ELISA was 79% with 100% specificity (95% efficiency, Cohen’s kappa 0.85), indicating that a substantial number of samples (12/56) positive in the Pourquier ELISA were negative in the Safepath ELISA. Thus, given that the Pourquier ELISA was more sensitive than the Safepath ELISA, the Pourquier ELISA was used for further comparisons in order to prevent the use of potential false negatives when comparing the results of the bead-based serological assay and the ELISA. To investigate the possibilities of bead-based serology using a generic Ig-binding protein instead of a species-specific antibody, the potential use of using protein A/G was explored. Firstly, protein A/G reacted with Ig from a broad panel of mammalian species in an ELISA (data not shown). When protein A/G was tested with the sera detailed in Table 1 (Nöckler et al., 1995), the same two groups were identified: those with MFI values <150 (n = 3), and those with MFI values ranging from 999 to 8962 (n = 36) (data not shown). When protein A/G was tested with the panel of 31 sera compiled to investigate cross-reactivity, none of the sera registered high signals (all <260 MFI). A subset (n = 88) of the field sera collected in endemic regions of Argentina was tested with the bead-based assay, using anti-swine antibody and protein A/G for detection and the results compared with the Pourquier ELISA by ROC analysis. For a range of specificities, the corresponding sensitivities, efficiencies and Cohen’s kappa are presented in Table 3. In comparison to the Pourquier ELISA, the highest sensitivities of the bead-based assay with a Cohen’s kappa of >0.8, were 91% sensitivity for anti-swine antibody at 91% speci-
442
F.J. van der Wal et al. / The Veterinary Journal 196 (2013) 439–444
Table 2 Performance of bead-based Trichinella assay (BBA) in pigs relative to two commercially available ELISAs (Pourquier and Safepath). Sera (n = 244) from a Trichinella-endemic region were tested with the BBA using anti-swine antibodies. At a range of specificities (85%, 90%, 95%, 99%, and 100%), the sensitivities, efficiencies, Cohen’s kappa, and corresponding cut-offs (MFI), are detailed as determined by ROC analysis. BBA vs. Pourquier ELISA Specificity (%) Sensitivity (%) Efficiency (%) Cohen’s kappa Cut-off (MFI)
85 96 88 0.7 961
90 91 90 0.74 1077
BBA vs. Safepath ELISA 95 87 93 0.81 1242
99 84 95 0.86 1432
100 78 95 0.85 1634
85 98 87 0.65 1053
90 98 91 0.75 1151
95 98 95 0.86 1350
99 95 98 0.94 1634
100 53 92 0.65 3481
Table 3 Performance of bead-based Trichinella assay (BBA) in pigs relative to a commercially available ELISA (Pourquier). Sera (n = 88) from a Trichinella-endemic region were tested with the BBA using anti-swine antibody (ab) and protein A/G, respectively. At a range of specificities (86%, 91%, 95%, 98%, and 100%), the sensitivities, efficiencies, Cohen’s kappa, and corresponding cut-offs (MFI), are detailed. BBA (anti-swine ab) vs. Pourquier Specificity (%) Sensitivity (%) Efficiency (%) Cohen’s kappa Cut-off (MFI)
86 97 90 0.79 855
91 91 91 0.81 1071
95 88 92 0.83 1137
BBA (protein A/G) vs. Pourquier 98 78 91 0.80 1392
ficity (cut-off 1071 MFI), and 94% sensitivity for protein A/G at 91% specificity (cut-off 589 MFI). When using protein A/G at 94% sensitivity the specificity of the bead-based assay increased to 95% at a cut-off of 692 MFI – the equivalent sensitivity for the anti-swine antibody variant of the assay was only 88%. These results demonstrate that higher sensitivities were achieved using protein A/G than using anti-swine antibody in the bead-based assay. For the subsequent analysis of sera from a negative population, cut-offs of 1071 MFI for anti-swine antibody and 692 for protein A/G were used, respectively. The bead-based assay was further evaluated in a Trichinellanegative population of 120 slaughter pigs. Both anti-swine antibody and protein A/G were used for detection. For anti-swine antibodies, signals varied from 49 to 683 MFI, and for protein A/ G from 18 to 79 (Fig. 2), all below the selected cut-offs of 1071 MFI for anti-swine antibody and 692 MFI for protein A/G, respectively. This result demonstrated that all the tested sera were negative in the bead-based assay, regardless of the detection molecule used (i.e. anti-swine antibody or protein A/G). To demonstrate its potential use in species other than pigs, 30 Trichinella seronegative and 31 seropositive human sera were tested using the bead-based assay incorporating protein A/G. The seronegative sera gave signals ranging from 4 to 219 MFI, whereas the seropositive sera registered values from 7 to 6118 MFI (Fig. 3). At a cut-off of 30 MFI, and with a sensitivity of 90%, the specificity was 93% (efficiency 92%, Cohen’s kappa 0.84), demonstrating that human IgG can be detected using protein A/G.
100 78 92 0.82 1433
86 97 90 0.79 319
91 94 92 0.83 589
95 94 94 0.88 692
98 78 91 0.80 910
100 69 89 0.74 1515
Fig. 2. Graph illustrating that sera (n = 120) from a Trichinella-negative population of pigs were seronegative using the bead-based assay, using anti-swine antibodies or protein A/G for detection. Signal strength of the assay (expressed as MFI), is plotted against (arbitrary) serum numbers.
Discussion The first objective of this study was to investigate whether a bead-based assay could be used to carry out serology in pigs using Trichinella as the test pathogen. Evaluation of previously wellcategorised sera (Nöckler et al., 1995) demonstrated that all seropositive animals registered a strong positive result using the bead-based assay, but there was no obvious correlation between the dose of larvae used to infect the pigs/larval burden and this result is as described with a conventional ELISA (Nöckler et al., 1995). The bead-based assay was further evaluated using ‘field’ sera from a region where trichinellosis was endemic (Döpfer et al., 2006; Teunis et al., 2009), and a Trichinella-free pig population. Using the porcine sera from the endemic region, the results of the bead-based assay and commercial ELISAs were compared using
Fig. 3. Graph illustrating the results of serology on human sera (30 seronegative and 31 seropositive, respectively) using the bead-based assay with protein A/G for detection. Signal strength of the assay (expressed as MFI), is plotted against (arbitrary) serum numbers.
ROC analyses and cut-offs were established. With these cut-offs, all the pigs from the Trichinella-free region were found to be seronegative, as expected.
F.J. van der Wal et al. / The Veterinary Journal 196 (2013) 439–444
The second objective of this research was to investigate if protein A/G could be used in the bead-based test format. Relative to the Pourquier ELISA, use of this protein in the bead-based assay resulted in a higher sensitivity (at a specificity of 95%) than when an anti-swine antibody was used. Furthermore, we demonstrated that protein A/G could be used to differentiate seronegative from seropositive human serum. Although several ELISAs are available to serologically screen for Trichinella, this approach is not commonly practiced. In the present work, when the effect of the cut-off on test sensitivity and specificity was evaluated for our bead-based assay, a decreased cut-off resulted in an increased sensitivity (less false negatives), and decreased specificity (more false positives), respectively. Given that this type of serological screening of pigs is used to monitor Trichinella-free/seronegative populations, test sensitivity may be prioritised over specificity in order to minimise the number of false negatives even though this will lead to an increase in the number of false positives. This seems a pragmatic approach since false positives can be verified with an appropriate confirmatory test whereas the false negatives may remain undetected. Since Trichinella occurs in a wide range of mammals (Pozio, 2005; Pozio and Murrell, 2006), we evaluated the capacity of our bead-based assay to serologically screen other species by using a more ubiquitous detection system focussed on protein A/G (Inoshima et al., 1999; Zhang et al., 2010). Our results demonstrated that this assay can also be performed using protein A/G which serendipitously increases the sensitivity relative to using anti-swine antibody. This increased sensitivity was largely due to a reduction in background binding (i.e. the signal of the ELISAnegative samples) – the MFI values of the positive samples were actually lower when protein A/G was used. Thus, used in this way, protein A/G lowers the assay cut-off point, increasing the sensitivity of the test when compared to an ELISA at a given specificity. Protein A/G could also be successfully incorporated into our beadbased assay when human serum was tested. Preliminary observations (data not shown) suggest that some sera, positive by ELISA and/or IFF, contain other Ig such as A and M, only weakly bound by protein A/G. The suspected presence of such immunoglobulins is supported by previous observations (Reiterova et al., 2007), and suggests that for the serological diagnosis of trichinellosis in humans the use of Ig-specific secondary antibodies may be preferred. Nevertheless, the ‘proof of principle’ of using protein A/G in the assay for use in species other than pigs was demonstrated, and the feasibility of such an approach in other serological assays and species remains to be determined.
Conclusions Although there are many examples illustrating how the Luminex platform can be used for multiplex serology in humans, the previous use of bead-based serology in veterinary applications is limited. Our results highlight the feasibility of using a bead-based assay to serologically screen pigs for evidence of Trichinella. This assay can serve as a ‘scaffold’ test system that can potentially be expanded to include other serological assays by adding other antigens. Such an approach could be used as a low cost method of screening pigs at slaughter for a range of pathogens including Salmonella spp., swine vesicular disease virus, and Aujeszky’s disease virus, or to differentiate between vaccinated and non-vaccinated animals. Recently, a duplex bead-based assay has been developed that can simultaneously detect antibodies against two proteins of the Rift valley fever virus, facilitating the differentiation of vaccinated from non-vaccinated animals. Bead-based methodology has considerable potential, and is likely to find a range of useful and novel veterinary applications.
443
Conflict of interest statement None of the authors of this paper has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.
Acknowledgments The authors would like to thank: J. van der Giessen for providing E/S antigen and for critical reading of the manuscript; R. Gamble, K. Nöckler, H. Kringel, and H. van der Heijden for providing us with porcine sera, D. Döpfer for Argentine field sera, M. Swanenburg, J. Cornelissen, and F. Borgsteede for sera from animal experiments performed at CVI, T. Garate and E. Rodriguez for human serum; and J. Bergervoet and J. Peters for helpful discussions in relation to the Luminex platform. This research was partly funded by the Dutch Ministry of Economic Affairs, Agriculture and Innovation (Strategic Research Program Food Safety, Monitoring and Detection KB-06-005), and by the European Union Seventh Framework Programme (FP7/2007-2013) under Grant Agreement Number 222633 (WildTech).
References Alban, L., Pozio, E., Boes, J., Boireau, P., Boue, F., Claes, M., Cook, A.J., Dorny, P., Enemark, H.L., van der Giessen, J., Hunt, K.R., Howell, M., Kirjusina, M., Nöckler, K., Rossi, P., Smith, G.C., Snow, L., Taylor, M.A., Theodoropoulos, G., Vallee, I., Viera-Pinto, M.M., Zimmer, I.A., 2011. Towards a standardised surveillance for Trichinella in the European Union. Preventive Veterinary Medicine 99, 148–160. Anderson, S., Wakeley, P., Wibberley, G., Webster, K., Sawyer, J., 2011. Development and evaluation of a Luminex multiplex serology assay to detect antibodies to bovine herpes virus, parainfluenza virus, bovine viral diarrhoea virus, and bovine respiratory syncytial virus, with comparison to existing ELISA detection methods. Journal of Immunological Methods 366, 79–88. Antonsson, A., Green, A.C., Mallitt, K.A., O’Rourke, P.K., Pawlita, M., Waterboer, T., Neale, R.E., 2010. Prevalence and stability of antibodies to the BK and JC polyomaviruses: A long-term longitudinal study of Australians. The Journal of General Virology 91, 1849–1853. Capo, V., Despommier, D.D., 1996. Clinical aspects of infection with Trichinella spp. Clinical Microbiology Reviews 9, 47–54. Casabonne, D., Waterboer, T., Michael, K.M., Pawlita, M., Mitchell, L., Newton, R., Harwood, C., Proby, C., 2009. The seroprevalence of human papillomavirus by immune status and by ethnicity in London. Infectious Agents and Cancer 4, 14. Clavijo, A., Hole, K., Li, M., Collignon, B., 2006. Simultaneous detection of antibodies to foot-and-mouth disease non-structural proteins 3ABC, 3D, 3A and 3B by a multiplexed Luminex assay to differentiate infected from vaccinated cattle. Vaccine 24, 1693–1704. Dias, D., Van Doren, J., Schlottmann, S., Kelly, S., Puchalski, D., Ruiz, W., Boerckel, P., Kessler, J., Antonello, J.M., Green, T., Brown, M., Smith, J., Chirmule, N., Barr, E., Jansen, K.U., Esser, M.T., 2005. Optimization and validation of a multiplexed Luminex assay to quantify antibodies to neutralizing epitopes on human papillomaviruses 6, 11, 16, and 18. Clinical and Diagnostic Laboratory Immunology 12, 959–969. Döpfer, D., Maassen, K., Achterberg, R.P., van der Giessen, J., Teunis, P., Buist, W., Molina, V., Guarnera, E., Gonzales, M., Krivokapich, S., Rodriguez, M., Peralta, J.L., Trabattoni, H., Larrieu, E., Engel, B., 2006. Diagnostic test evaluation for Trichinella infections in pigs – A renewed interaction between diagnostic test data and their statistical analysis. In: Proceedings of the 11th Symposium of the International Society for Veterinary Epidemiology and Economics, vol. 11, pp. 649–651. Dupouy-Camet, J., 2000. Trichinellosis: A worldwide zoonosis. Veterinary Parasitology 93, 191–200. Escalante, M., Romaris, F., Rodriguez, M., Rodriguez, E., Leiro, J., Garate, M.T., Ubeira, F.M., 2004. Evaluation of Trichinella spiralis larva group 1 antigens for serodiagnosis of human trichinellosis. Journal of Clinical Microbiology 42, 4060–4066. Franssen, F.F., Fonville, M., Takumi, K., Vallee, I., Grasset, A., Koedam, M.A., Wester, P.W., Boireau, P., van der Giessen, J.W., 2011. Antibody response against Trichinella spiralis in experimentally infected rats is dose dependent. Veterinary Research 42, 113. Gajadhar, A.A., Pozio, E., Gamble, H.R., Nöckler, K., Maddox-Hyttel, C., Forbes, L.B., Vallee, I., Rossi, P., Marinculic, A., Boireau, P., 2009. Trichinella diagnostics and control: Mandatory and best practices for ensuring food safety. Veterinary Parasitology 159, 197–205. 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. Veterinary Parasitology 13, 349–361.
444
F.J. van der Wal et al. / The Veterinary Journal 196 (2013) 439–444
Gamble, H.R., Rapic, D., Marinculic, A., Murrell, K.D., 1988. Evaluation of excretory– secretory antigens for the serodiagnosis of swine trichinellosis. Veterinary Parasitology 30, 131–137. Go, Y.Y., Wong, S.J., Branscum, A.J., Demarest, V.L., Shuck, K.M., Vickers, M.L., Zhang, J., McCollum, W.H., Timoney, P.J., Balasuriya, U.B., 2008. Development of a fluorescent-microsphere immunoassay for detection of antibodies specific to equine arteritis virus and comparison with the virus neutralization test. Clinical and Vaccine Immunology 15, 76–87. Gottstein, B., Pozio, E., Nöckler, K., 2009. Epidemiology, diagnosis, treatment, and control of trichinellosis. Clinical Microbiology Reviews 22, 127–145. Inoshima, Y., Shimizu, S., Minamoto, N., Hirai, K., Sentsui, H., 1999. Use of protein AG in an enzyme-linked immunosorbent assay for screening for antibodies against parapoxvirus in wild animals in Japan. Clinical Diagnostic Laboratory Immunology 6, 388–391. Krishhan, V.V., Khan, I.H., Luciw, P.A., 2009. Multiplexed microbead immunoassays by flow cytometry for molecular profiling: Basic concepts and proteomics applications. Critical Reviews in Biotechnology 29, 29–43. Maassen, C.B.M., Achterberg, R.P., Engel, B., van der Giessen, J., Teunis, P., van der Heijden, H.M.J.F., Haughey, S., Molina, V., Peralta, J.L., Döpfer, D., 2007. Evaluation of serological test for Trichinella infections in pigs. In: Proceedings of 7th International Safepork Symposium, Padova, Italy, pp. 211–213. Mackinnon, A., 2000. A spreadsheet for the calculation of comprehensive statistics for the assessment of diagnostic tests and inter-rater agreement. Computers in Biology and Medicine 30, 127–134. Nöckler, K., Voigt, W.P., Protz, D., Miko, A., Ziedler, K., 1995. Diagnosis of trichinellosis in living pigs using indirect ELISA. Berliner und Munchener Tierärztliche Wochenschrift 108, 167–174. Nöckler, K., Pozio, E., Voigt, W.P., Heidrich, J., 2000. Detection of Trichinella infection in food animals. Veterinary Parasitology 93, 335–350. Perkins, J., Clavijo, A., Hindson, B.J., Lenhoff, R.J., McBride, M.T., 2006. Multiplexed detection of antibodies to nonstructural proteins of foot-and-mouth disease virus. Analytical Chemistry 78, 5462–5468.
Pozio, E., 2005. The broad spectrum of Trichinella hosts: From cold- to warmblooded animals. Veterinary Parasitology 132, 3–11. Pozio, E., Murrell, K.D., 2006. Systematics and epidemiology of Trichinella. Advances in Parasitology 63, 367–439. Reiterova, K., Kincekova, J., Snabel, V., Marucci, G., Pozio, E., Dubinsky, P., 2007. Trichinella spiralis-outbreak in the Slovak Republic. Infection 35, 89–93. Sulzer, A.J., 1965. Indirect fluorescent antibody tests for parasitic diseases. I. Preparation of a stable antigen from larvae of Trichinella spiralis. Journal of Parasitology 51, 717–721. Teunis, P.F., Fonville, M.T., Dopfer, D.D., Eijck, I.A., Molina, V., Guarnera, E., van der Giessen, J.W., 2009. Usefulness of sero-surveillance for Trichinella infections in animal populations. Veterinary Parasitology 159, 345–349. van der Wal, F.J., Achterberg, R.P., Boer, S.M., Boshra, H., Brun, A., Maassen, C.B., Kortekaas, J., 2012. Bead-based suspension array for simultaneous detection of antibodies against the Rift Valley fever virus nucleocapsid and Gn glycoprotein. Journal of Virological Methods 183, 99–105. van Gageldonk, P.G., van Schaijk, F.G., van der Klis, F.R., Berbers, G.A., 2008. Development and validation of a multiplex immunoassay for the simultaneous determination of serum antibodies to Bordetella pertussis, diphtheria and tetanus. Journal of Immunological Methods 335, 79–89. Watson, D.S., Reddy, S.M., Brahmakshatriya, V., Lupiani, B., 2009. A multiplexed immunoassay for detection of antibodies against avian influenza virus. Journal of Immunological Methods 340, 123–131. Wegener, H.C., Hald, T., LoFoWong, D., Madsen, M., Korsgaard, H., Bager, F., GernerSmidt, P., Molbak, K., 2003. Salmonella control programs in Denmark. Emerging Infectious Diseases 9, 774–780. Zhang, D., Wang, Z., Fang, R., Nie, H., Feng, H., Zhou, Y., Zhao, J., 2010. Use of protein AG in an enzyme-linked immunosorbent assay for serodiagnosis of Toxoplasma gondii infection in four species of animals. Clinical Vaccine Immunology 17, 485–486.