Serotyping of naturally Toxoplasma gondii infected meat-producinganimals

Susana Sousa a,c,*, Nuno Canada b,c,d, Jose Manuel Correia da Costa a,c, Marie-Laure Darde e,f

a Centro de Imunologia e Biologia Parasitaria (CIBP), CSPGF, INSA Porto, Portugalb Instituto de Ciencias Biomedicas de Abel Salazar (ICBAS), Universidade do Porto, Porto, Portugalc Centro de Estudos de Ciencia Animal (CECA), Universidade do Porto, Vairao, Portugald Laboratorio Nacional de Investigacao Veterinaria (LNIV), Vairao, Vila do Conde, Portugale Laboratoire de Parasitologie-Mycologie, EA 3174-NETEC, Faculte de Medecine, Universite de Limoges, Limoges, 87025, Francef Centre National de Reference (CNR) Toxoplasmose/Toxoplasma Biological Resource Center (BRC), Limoges, France

Veterinary Parasitology 169 (2010) 24–28

A R T I C L E I N F O

Article history:

Received 17 July 2009

Received in revised form 10 December 2009

Accepted 16 December 2009

Keywords:

Toxoplasma gondii

Serotyping

Meat-producing-animals

A B S T R A C T

Serotyping was previously described as a promising method for typing strains of

Toxoplasma gondii. The majority of precedent studies utilized serum samples collected

from human patients with different T. gondii-associated pathologies. The aim of this work

was to study the applicability of the same procedure for serotyping naturally infected

meat-producing animals. An ELISA test based on GRA6 and GRA7 C-terminal polymorphic

peptides was used. Peptide GRA6II has polymorphisms specific for the archetypal strains

type II, GRA6I/III for strains type I and III, GRA7I for strains type I and GRA7III for strains

type III. As reference material, and to validate this approach, serum samples from eleven

free-range chickens and fifteen pigs used for Toxoplasma genotypes isolation were

selected. These strains integrate the Biological Resource Centre (BRC) ToxoBS Bank. Three

serum samples from chickens and two from pigs had serotyping results in agreement with

genotyping. Thirty-five serum samples from chickens, twenty-nine from pigs and fifty

from sheep, seropositive for T. gondii, from which no isolate was obtained, were also

serotyped. Serotype III appeared significantly more frequent among sheep. Our results

show that serotyping still need refinement, but may become a valuable tool for typing

Toxoplasma strains from animal origin.

� 2009 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Veterinary Parasitology

journa l homepage: www.e lsevier .com/ locate /vetpar

1. Introduction

Toxoplasmosis is the parasitic zoonosis with thehighest human incidence (EFSA, 2007). There is a wide-spread distribution of Toxoplasma infections in a variety oflivestock and wild animals. Ingestion of environmentaloocysts (by drinking water or eating raw vegetables) and

* Corresponding author at: Laboratorio de Parasitologia, Centro de

Saude Publica Dr Goncalves Ferreira, Rua Alexandre Herculano no. 321,

4000-055 Porto, Portugal. Tel.: +351 223401100; fax: +351 223401109.

E-mail address: sanrspt@yahoo.co.uk (S. Sousa).

0304-4017/$ – see front matter � 2009 Elsevier B.V. All rights reserved.

doi:10.1016/j.vetpar.2009.12.025

eating raw or undercooked meat containing tissue cystsstages are the main transmission routes in livestock and inhumans (Carme et al., 2002; Moura et al., 2006;Heukelbach et al., 2007). Knowledge about genotype ofstrains infecting meat producing animal is important.Relationship between Toxoplasma genotype and outcomeof human toxoplasmosis has been suggested (Darde, 2008),and severe cases of human toxoplasmosis due to atypicalgenotypes have been described after ingestion of meat(Carme et al., 2009; Elbez-Rubinstein et al., 2009).

Few studies regarding Toxoplasma infection in meat-producing animals were performed in Portugal. Antibodieswere found in 27.1% out of 225 free-range chickens (Dubey

S. Sousa et al. / Veterinary Parasitology 169 (2010) 24–28 25

et al., 2006), and in 15.6% out of 333 free-range pigs (Sousaet al., 2006) from different areas of Portugal. Prevalence oftoxoplasmosis in 1467 sheep randomly collected from 160farms from Northern Portugal, representing approximately10% of the ovine population was 17.1% (Sousa et al.,2009b). Also, T. gondii was isolated in a rare case ofcongenital toxoplasmosis in cattle (Canada et al., 2002).The three clonal lineages I, II and III were described inPortugal, although lineages II and III were found to bedominant (Dubey et al., 2006; Sousa et al., 2006). In fact,the distribution of these three clonal lineages differsbetween the European countries. The few studies con-ducted in other European countries, revealed a preponder-ance of genotype II in domestic animals (Dubey et al., 2005;Dumetre et al., 2006; Owen and Trees, 1999).

Genetic characterization is done subsequently to theisolation of the parasite by mouse inoculation, which is atime consuming and labour technique. This is, at ourknowledge, the first attempt to apply the serotypingtechnology to animal reference material. This method,based on the specific antibody–antigen recognition usingpolymorphic peptides was already described as a reliablemethod of typing Toxoplasma strains in humans (Konget al., 2003; Peyron et al., 2006; Morisset et al., 2008; Sousaet al., 2008).

Our objective was to test serotyping in serum samplesfrom animal origin, using peptides derived from GRA6 andGRA7 antigens.

2. Materials and methods

2.1. Animal serum samples

Serum samples from chickens and pigs from Portugal,from which T. gondii strains were previously isolated andgenotyped by a multilocus approach (SAG2 and micro-satellite analysis) (Ajzenberg et al., 2005), were selected inorder to validate serotyping as a typing method ofToxoplasma strains with animal origin. Eleven strains frompigs and eight strains from chickens were type II. Fourstrains from pigs and three from chickens were type III(Dubey et al., 2006; Sousa et al., 2006). Thirty-five serumsamples from chickens, twenty-nine from pigs and fiftyfrom sheep, seropositive for T. gondii with a modifiedagglutination test (Dubey et al., 2006; Sousa et al., 2006,2009b), from which no isolate was obtained, were alsostudied. Cut-off values for MAT were 1:5 for chickens, 1:10for pigs and 1:20 for sheep. Serum samples from 13Toxoplasma negative chickens, 10 Toxoplasma negative pigsand 16 Toxoplasma negative sheep were used to establishcut-off values.

2.2. Peptides

Four strain-specific peptides and a control peptide werestudied. These peptides were previously described by Sousaet al. (2008, 2009a). Peptide GRA6II has polymorphismsspecific for type II strain; peptide GRA6I/III has polymorph-isms specific for type I and III strains; peptide GRA7I haspolymorphisms specific for type I strain; peptide GRA7IIIhas polymorphisms specific for type III strain.

2.3. ELISA protocol

Immobilizer amino plates (Nunc, Denmark) werecoated with each peptide diluted to 10 mg/ml in 0.05 Mcarbonate/bicarbonate buffer pH 9.6 by incubation over-night at 4 8C. Wells were blocked with a solution of 3% BSAin PBS for 1 h at 37 8C in a moisture atmosphere and thenwashed 3 times with 0.3% Tween-20 in PBS. Serial dilutionsof sera and conjugate (data not show) were performed inorder to define the optimal work conditions. Chicken serawere diluted at 1/800 in a solution of 3% BSA in PBS with0.3% Tween-20, pig sera were diluted at 1/100 in a solutionof 3% BSA in PBS, sheep sera were diluted at 1/100 in asolution of 3% BSA in PBS with 0.3% Tween-20 andincubated for 2 h at 37 8C in a moist atmosphere. Wellswere washed 3 times with PBS/Tween 0.3%. Anti-chickenIgG peroxidase (KPL, USA) was diluted at 1/32,000 in asolution of 3% BSA in PBS with 0.3% Tween-20, anti-swineIgG peroxidase (KPL, USA) was diluted at 1/1000 in asolution of 3% BSA in PBS, anti-sheep/goat IgG peroxidaselabeled conjugate (The Binding Site, UK) was diluted at 1/10,000 in a solution of 3% BSA in PBS with 0.3% Tween-20and incubated for 1 h at 37 8C in a moist atmosphere. Wellswere washed 3 times with PBS/Tween 0.3% and developedwith o-phenylenediamine (OPD) for 15 min at 37 8C.Reaction was stopped with HCl 3 M and absorbance wasmeasured at 490 nm. Optical density (OD) index wascalculated by subtracting the OD of the peptide controlfrom the OD of each peptide. Cut off was set on the meanabsorbance readings of negative sera plus 2SD (standarddeviation).

2.4. Statistical analysis

Statistical analysis was performed using SPSS Version12.0 for Windows. The Chi-squared test was performed toassess the statistical significance of differences in theprevalence of serotypes for different hosts. p values of lessthan 0.05 were considered significant.

3. Results

Cut-off values were defined for each peptide accordinganimal species. Chickens were considered positive forGRA6II, GRA6I/III, GRA7I and GRA7III when OD index wasequal or higher than cut-off established (0.058, 0.051,0.180 and 0.244 respectively). Pigs were consideredpositive for GRA6II, GRA6I/III, GRA7I and GRA7III whenOD index was equal or higher than cut-off established(0.212, 0.249, 0.231 and 0.228 respectively). Sheep wereconsidered positive for GRA6II, GRA6I/III, GRA7I andGRA7III when OD index was equal or higher than cut-offestablished (0.452, 0.377, 0.038 and 0.076 respectively).

3.1. Serotyping vs genotyping

Serotyping results from 11 chickens and 15 pigs arereported in Tables 1 and 2. Three chickens (GA147, GA163and GA166) out of eleven had the same result forserotyping and genotyping (Table 1). Chicken GA43exclusively reacted with peptide GRA6I/III, suggesting a

Table 1

Serological reactivity of chicken sera with GRA6 and GRA7 specific peptides.

Case no. Strain GRA6 II GRA6 I/III GRA7 I GRA7 III Serotype Genotypea

GA19 TgCkPr1 0.023 �0.009 0.123 0.065 NR III

GA39 TgCkPr3 0.040 0.005 0.049 0.043 NR III

GA40 TgCkPr4 0.025 0.018 0.053 0.092 NR II

GA43 TgCkPr5 0.038 0.902 0.089 0.169 GRA6I/III III

GA147 TgCkPr6 0.104 0.004 0.012 0.051 II II

GA163 TgCkPr8 0.990 0.014 0.113 0.181 II II

GA164 TgCkPr9 0.170 0.003 0.279 0.230 II/I II

GA166 TgCkPr7 0.316 �0.005 0.070 0.073 II II

GA167 TgCkPr10 0.319 0.221 0.081 0.116 II/I/III II

GA170 TgCkPr11 0.026 0.008 0.069 0.097 NR II

GA176 TgCkPr12 0.053 �0.013 0.041 0.086 NR II

OD index above cut off values is in bold.

Cut-off value: GRA6II = 0.058, GRA6I/III = 0.051, GRA7I = 0.180 and GRA7III = 0.244.a Genotype defined by the study of SAG2 locus by PCR-RFLP.

Table 2

Serological reactivity of pig sera with GRA6 and GRA7 specific peptides.

Case no. Strain GRA6 II GRA6 I/III GRA7 I GRA7 III Serotype Genotypea

PV44 TgPiPr1 0.104 0.052 �0.017 0.162 NR II

PV116 TgPiPr2 0.008 �0.089 0.019 0.186 NR III

PV214 TgPiPr3 0.097 0.107 �0.023 0.054 NR II

PV220 TgPiPr4 �0.056 �0.099 0.011 0.028 NR III

PV227 TgPiPr5 0.824 �0.080 �0.018 0.134 II II

PV231 TgPiPr6 0.967 0.276 �0.090 0.720 II/III II

PV232 TgPiPr7 0.185 0.544 0.044 0.069 GRA6I/III II

PV238 TgPiPr8 0.615 0.020 �0.024 0.048 II II

PV266 TgPiPr9 �0.114 �0.125 �0.156 �0.059 NR II

PV272 TgPiPr10 0.742 0.477 0.032 0.033 II/I/III II

PV274 TgPiPr11 1.074 0.576 0.020 0.036 II/I/III II

PV282 TgPiPr12 0.044 �0.093 0.049 0.035 NR II

PV302 TgPiPr13 �0.080 �0.073 �0.019 0.042 NR II

PV311 TgPiPr14 �0.023 0.080 0.005 0.020 NR III

PV316 TgPiPr15 0.017 0.005 0.008 0.033 NR III

OD index above cut off values is in bold.

Cut-off value: GRA6II = 0.212, GRA6I/III = 0.249, GRA7I = 0.231 and GRA7III = 0.228.a Genotype defined by the study of five microsatelites markers (TUB2, TgM A, W35, B17, B18).

S. Sousa et al. / Veterinary Parasitology 169 (2010) 24–2826

possible infection with a strain type I or type III, which is inagreement with the genotyping results (type III). Twochickens reacted with more than one peptide, suggesting apossible mix infection (GA164 and GA167). From serumsamples from infected pigs, only two (PV227 and PV238)out of fifteen had the same results for genotyping andserotyping. One pig infected with a strain type II (PV232),reacted with the peptide GRA6I/III (Table 2). Some serumsamples had OD index below the established cut-off for allpeptides and were considered non-reactive. Five out ofeleven from chickens and nine out of fifteen from pigs werenon-reactive (Tables 1 and 2). No correlation was foundbetween MAT titers of the sera and response of thepeptides.

3.2. Prediction of T. gondii serotypes in naturally infected

animals from Portugal

Chickens, pigs and sheep positive for Toxoplasma fromwhich no isolate was obtained were serotyped. Differentserotype profiles were defined. Serotype I was defined bythe reactivity with peptides GRA7I or GRA7I and GRA6I/III.Serotype II was defined by the single reactivity with

peptide GRA6II. Serotype III was defined by the reactivitywith peptides GRA7III or GRA7III and GRA6I/III. SerotypeGRA6I/III was defined by the single reactivity with peptideGRA6I/III. Cross-reaction (CR) represents serum samplesthat have reacted with more than one peptide suggesting apossible mixed infection. Non-reactive (NR) represents theserum samples with OD index below cut-off values for allpeptides.

Five different reactivity profiles were found for chick-ens, three different profiles for pigs and five for sheep.Serotype III was significantly more frequently found insheep (p< 0.05) (Table 3). Serotype II was more frequent inchickens and pigs, while cross-reactivity (CR) was morefrequent in pigs and sheep. However, these differenceswere not significant (p> 0.05) (Table 3). The rate of non-reactive (NR) sera was higher than 50% for the threestudied animal species, with no significant difference(p> 0.05) (Table 3).

4. Discussion

Serotyping using the GRA6 and GRA7 derived peptideswere previously described for serotyping human serum

Table 3

Frequencies of serotype according to animal species.

Serotype Chickens,

n = 35

Pigs,

n = 29

Sheep,

n = 50

Significancea,

p

I 0 0 1 (2%) 0.524

II 9 (25.7%) 6 (20.7%) 4 (8%) 0.078

III 2 (5.7%) 0 10 (20%) 0.011GRA6I/III 2 (5.7%) 0 0 0.101

Cross-reaction 3 (8.6%) 6 (20.7%) 10 (20%) 0.303

Non-reactive 19 (54.3%) 17 (58.6%) 25 (50%) 0.756a Significant values are in bold.

S. Sousa et al. / Veterinary Parasitology 169 (2010) 24–28 27

samples (Sousa et al., 2008, 2009a). In this work, theapplicability of the same procedure for serotypingnaturally infected meat-producing animals was evaluated.Published data on meat-producing animals from Portugalrefers to chickens (Dubey et al., 2006), pigs (Sousa et al.,2006) and bovines (Canada et al., 2002). No data isavailable about genotypes isolated from sheep.

Match rate between genotyping and serotyping resultswas 36.4% for chickens and 13.3% for pigs. Peptide GRA6IIwas specifically recognized by type II strains from chickensand pigs, but with low sensitivity (52.6%). Peptide GRA6I/IIIalso has a low sensitivity (only one out of seven serumsamples from chickens and pigs infected with type IIIstrains were recognized). Peptide GRA7III does notrecognize any serum sample from animals infected withtype III strains. Some serum samples from animals infectedwith type II strains reacted with the peptide GRA6II butalso with the peptides specific for strains type I and III. Thiscross-reactivity may reflect peptides specificity problems,but it can also result from a natural mixed infection thatwas not detected by bioassay.

Serotyping was used in an attempt to determine thegenotypes from the infected animals for which no isolatewas obtained. For those animals, serotype II prevails inchickens and pigs, while for sheep serotype III seems to bemore prevalent. Results obtained for chickens and pigsare in agreement with the genotyping results ofToxoplasma isolates obtained from chickens and pigsfrom Portugal (Dubey et al., 2006; Sousa et al., 2006). Thehigher frequency of serotype III in sheep compared tochickens and pigs could suggest a possible strain selectionof type III by sheep, but this hypothesis is unlikely asother studies in Europe described the predominance ofgenotype II in sheep (Dumetre et al., 2006; Owen andTrees, 1999). This data may suggest that in Portugalgenotype III is more frequent than in France. Similarresults were obtained with serum samples from humanpatients, where serotype III was more frequent amongPortuguese patients compared to French patients (Sousaet al., 2008).

These data demonstrate that serotyping based on thesepeptides present some limitations for typing strains ofanimal origin (low sensitivity, cross-reactions). Otherpeptides from different antigens should be studied.

Since animals are a potential source of human infection,through tissue cysts, detection and genetic characteriza-tion of Toxoplasma infection in meat producing animals isimportant. A method that could characterize the Toxo-

plasma infection in an efficient and rapid way, could serve

to control infections with, for example, virulent atypicalstrains that are responsible for severe cases of humaninfections.

Acknowledgement

The authors would like to thank JP Dubey for kindlysupplying the anti-chicken and anti-swine conjugates.

References

Ajzenberg, D., Dumetre, A., Darde, M.L., 2005. Multiplex PCR for typingstrains of Toxoplasma gondii. J. Clin. Microbiol. 43, 1940–1943.

Canada, N., Meireles, C.S., Rocha, A., da Costa, J.M., Erickson, M.W., Dubey,J.P., 2002. Isolation of viable Toxoplasma gondii from naturally infectedaborted bovine fetuses. J. Parasitol. 88, 1247–1248.

Carme, B., Bissuel, F., Ajzenberg, D., Bouyne, R., Aznar, C., Demar, M.,Bichat, S., Louvel, D., Bourbigot, A.M., Peneau, C., Neron, P., Darde,M.L., 2002. Severe acquired toxoplasmosis in immunocompetentadult patients in French Guiana. J. Clin. Microbiol. 40, 4037–4044.

Carme, B., Demar, M., Ajzenberg, D., Darde, M.L., 2009. Severe acquiredtoxoplasmosis caused by wild cycle of Toxoplasma gondii, FrenchGuiana. Emerg. Infect. Dis. 15, 656–658.

Darde, M.L., 2008. Toxoplasma gondii, ‘‘new’’ genotypes and virulence.Parasite 15, 366–371.

Dubey, J.P., Edelhofer, R., Marcet, P., Vianna, M.C.B., Kwok, O.C.H., Leh-mann, T., 2005. Genetic and biologic characteristics of Toxoplasmagondii infections in free-range chickens from Austria. Vet. Parasitol.133, 299–306.

Dubey, J.P., Vianna, M.C.B., Sousa, S., Canada, N., Meireles, S., Correia daCosta, J.M., Marcet, P.L., Lehmann, T., Darde, M.L., Thulliez, P., 2006.Characterization of Toxoplasma gondii isolates in free-range chickensfrom Portugal. J. Parasitol. 92, 184–186.

Dumetre, A., Ajzenberg, D., Rozette, L., Mercier, A., Darde, M.L., 2006.Toxoplasma gondii infection in sheep from Haute-Vienne, France:seroprevalence and isolate genotyping by microsatellite analysis.Vet. Parasitol. 142, 376–379.

European Food Safety Authority, 2007. Scientific opinion of the panel onbiological hazards on a request from EFSA on Surveillance and mon-itoring of Toxoplasma in humans, foods and animals. EFSA J. 583, 1–64.

Elbez-Rubinstein, A., Ajzenberg, D., Darde, M.L., Cohen, R., Dumetre, A.,Yera, H., Gondon, E., Janaud, J.C., Thulliez, P., 2009. Congenital Tox-oplasmosis and reinfection during pregnancy: case report, straincharacterization, experimental model of reinfection, and review. J.Infect. Dis. 199, 280–285.

Heukelbach, J., Meyer-Cirkel, V., Moura, R.C.S., Gomide, M., Queiroz, J.A.N.,Saweljew, P., Liesenfeld, O., 2007. Waterborne toxoplasmosis, North-eastern Brazil. Emerg. Infect. Dis. 13, 287–289.

Kong, J.T., Grigg, M.E., Uyetake, L., Parmley, S., Boothroyd, J.C., 2003.Serotyping of Toxoplasma gondii infections in humans using syntheticpeptides. J. Infect. Dis. 187, 1484–1495.

Morisset, S., Peyron, F., Lobry, J.R., Garweg, J., Ferrandiz, J., Musset, K.,Gomez-Marin, J.E., de la Torre, A., Demar, M., Carme, B., Mercier, C.,Garin, J.F., Cesbron-Delauw, M.F., 2008. Serotyping of Toxoplasmagondii: striking homogeneous pattern between symptomatic andasymptomatic infections within Europe and South America. MicrobesInfect. 10, 742–747.

Moura, L., Bahia-Oliveira, L.M., Wada, M.Y., Jones, J.L., Tuboi, S.H.,Carmo, E.H., Ramalho, W.M., Camargo, N.J., Trevisan, R., Graca,R.M.T., Silva, A.J., Moura, I., Dubey, J.P., Garrett, D.O., 2006. Water-borne toxoplasmosis, Brazil, from field to gene. Emerg. Infect Dis.12, 326–329.

Owen, M.R., Trees, A.J., 1999. Genotyping of Toxoplasma gondii associatedwith abortion in sheep. J. Parasitol. 85, 382–384.

Peyron, F., Lobry, J.R., Musset, K., Ferrandiz, J., Gomez-Marin, J.E.,Petersen, E., Meroni, V., Rausher, B., Mercier, C., Picot, S., Ces-bron-Delauw, M.F., 2006. Serotyping of Toxoplasma gondii in chroni-cally infected pregnant women: predominance of type II in Europeand types I and III in Colombia (South America). Microbes Infect. 8,2333–2340.

Sousa, S., Ajzenberg, D., Canada, N., Freire, L., Correia da Costa, J.M., Darde,M.L., Thulliez, P., Dubey, J.P., 2006. Biologic and molecular character-ization of Toxoplasma gondii isolates from pigs from Portugal. Vet.Parasitol. 135, 133–136.

S. Sousa et al. / Veterinary Parasitology 169 (2010) 24–2828

Sousa, S., Ajzenberg, D., Vilanova, M., Costa, J., Darde, M.L., 2008.Use of GRA6-derived synthetic peptides in an immunoenzymaticassay to serotype Toxoplasma gondii in human serum samplescollected from three continents. Clin. Vaccine Immunol. 15,1380–1386.

Sousa, S., Ajzenberg, D., Marle, M., Aubert, D., Villena, I., Correia da Costa,J., Darde, M.L., 2009a. Selection of polymorphic peptides from GRA6

and GRA7 sequences of Toxoplasma gondii strains to be used inserotyping. Clin. Vaccine Immunol. 16, 1158–1169.

Sousa, S., Thompson, G., Silva, E., Freire, L., Lopes, D., Correia da Costa, J.M.,Castro, A., Carvalheira, J., Canada, N., 2009b. Determination of themore adequate modified agglutination test cut-off for serodiagnosisof Toxoplasma gondii infection in sheep. Zoonoses Public Health 56,252–256.