High prevalence of unusual genotypes of Toxoplasma gondii infection in pork meat samples from Erechim, Southern Brazil

Toxoplasmosis is the most common cause of infectious uveitis in Brazil, with a higher frequency in the South of the country. We have collected samples from porcine tongue and diaphragm obtained in both large and small abattoirs and used molecular biological technique to determine the prevalence of infection and RFLP analysis to type the parasites. Seventeen out of 50 (34%) samples from the diaphragm and 33 out of 50 (66%) samples from the tongue demonstrated a positive PCR reaction for T. gondii and restriction analysis of four of the positive samples revealed that all had a type I genotype at SAG2. However, when other unlinked loci were analyzed, these strains had a type III genotype at markers BTUB, SAG3, and GRA6. One of the strains (8T) had a type II allele at SAG3, indicating it has a combination of alleles normally seen in the clonal lineages. Our sampling indicates a high prevalence of infection and suggests that unusual genotypes of T. gondii are found in Brazil even among domesticated pigs.


INTRODUCTION
Toxoplasmosis is one of the most common causes of infectious uveitis -inflammation of the inner part of the eye -in humans and studies have shown a high frequency of toxoplasmic ocular lesions in the city of Erechim, Southern Brazil (Glasner et al. 1992, Silveira et al. 2001. The eye can be affected in both congenital and acquired infections (Holland 1996, Nussenblatt andBelfort 1994) and in this region of Brazil, the in-*Member Academia Brasileira de Ciências Correspondence to: Rubens Belfort Jr. E-mail: epm@belfort.med.br fection is usually transmitted by eating uncooked meat containing toxoplasma organisms. Pork is considered one of the most important sources of infection and it has been shown that porcine diaphragm and tongue frequently contain parasite cysts (Dubey et al. 1986).
Recently developed special techniques in molecular biology have allowed for the identifi cation of T. gondii DNA in different organs in humans and in various animals thought to harbor the parasite. Restriction fragment length polymorphism (RFLP) analysis of the organism has indicated that three major subgroups are predominant in North America and Europe, and it has been shown 112 RUBENS BELFORT-NETO et al. that human infection is mainly caused by types I and II (Ajzenberg et al. 2004, Howe andSibley 1995). Our goal was to determine the prevalence of T. gondii contamination in commercially available pork meat obtained from Erechim, RS, Brazil and the genotype of those parasites.

MATERIALS AND METHODS
Porcine tongue and diaphragm obtained in both large and small abattoirs in the Erechim area were studied.
The samples were kept in a saline solution and were later ground and frozen on site. Extraction of DNA was performed afterwards in São Paulo. Samples were prepared separately, using disposable material, with time intervals and cleaning after each sample to avoid crosscontamination. PCR was performed using two specifi c primers for T. gondii - -respectively from the 5 and 3 ends of a 533 bp fragment from the T. gondii genome. After optimizing the PCR reaction using tissue culture parasites, the PCR reaction using these samples was performed in a 50 ml reaction mixture containing 0.5 mM of each primer, 100 mM dNTP (Pharmacia Biotech), 60 mM Tris + HCl (pH 9.0), 15 mM (NH4)2SO4, 2 mM MgCl2, 0.5 U Taq PLATINUM (Applied Biosysthems). Amplifi cation was performed using a PerkinElmer/Applied Biosystems 9600 thermo cycler by 10 min incubation at 94 • C, followed by 38 cycles of 1.5 min at 94 • C, 1 min at 56 • C, 1 min at 72 • C and a fi nal 10 min incubation at 72 • C. The PCR products were analyzed using Southern blot methodology.
Six samples (four positive for T. gondii and two negative) were sent to Washington University for genotype analysis. DNAzol (Molecular Research Center, Inc.) treated tissue samples were extracted with equal volume of chloroform. Polyacryl carrier (Molecular Research Center, Inc.) was added (5 uL) to the aqueous phase and DNA was precipitated by addition of equal volume of ethanol and centrifugation at 5,000xg for 10 min. DNAs were subjected to nested PCR analysis for the unlinked markers SAG2, BTUB, SAG3, and GRA6 (Khan et al. 2005). Following amplifi cation, samples were digested with appropriate restriction enzymes, resolved on agarose gel and alleles scored by comparison to the type strains RH (type I), ME49 (type II) and CTG (type III).

RESULTS
Seventeen out of 50 (34%) samples from the diaphragm and 33 out of 50 (66%) samples from the tongue demonstrated a positive PCR reaction for T. gondii, while 28% had positive PCR reactions in both the tongue and diaphragm (data not shown).
Restriction analysis and DNA sequencing were performed on six random samples. Re-amplifi cation was negative on the two previously negative samples (negative controls). All four positive samples had a type I genotype at SAG2 (Table I). However, when other unlinked loci were analyzed, these strains had a type III genotype at markers BTUB, SAG3, and GRA6 (Table I).
One of the strains (8T) had a type II allele at SAG3, indicating that it has a mixed profi le (Table I). Although a high incidence of PCR positive samples were found, we were unable to infect interferon-gamma defi cient mice with either a macerated suspension, the meat or with serum from the animals (data not shown). This suggests that the parasite load in these samples is small.

DISCUSSION
These results suggest that a signifi cant proportion of meats commercially available in Erechim, Southern Brazil, are contaminated with T. gondii. This poses a real public health risk to those who live in that environment. Recently a study in the UK (Aspinall et al. 2002) assessed the prevalence of T. gondii by polymerase chain reaction and parasite typing by genotype in commercial meat products and demonstrated that 27 of the 71 samples harbored parasites. Restriction analysis and DNA sequencing showed that 21 of the contaminated meats contained parasites genotyped as type I at the SAG2 locus, whilst six of the samples contained parasites of both types I and II (Aspinall et al. 2002). A recent study conducted in Sao Paulo and Erechim has shown that in both sites only type I was identifi ed (Vallochi et al. 2005).
SAG2 has previously been shown to be highly sensitive means of genotyping strains: based on a combination of polymorphism in the 5' and 3' ends of the gene, it is possible to classify strains of the three con- ventional clonal lineages into one of the three genotypes (Howe et al. 1997). However, as shown in the present study, typing at SAG2 alone does not detect strains with recombinant genotypes. In the present case, all of the strains would have been misidentifi ed as type I, when in fact they contain alleles of different genotypes at different genetic loci. This pattern of apparently recombinant genotypes is highly unusual as has not been seen frequently in isolates studied previously. Typically, 95% of strains isolated in North America and Europe fall into one of three highly clonal genotypes (Howe and Sibley 1995). However, it has recently been proposed that a greater degree of genetic diversity is found among strains in some geographic areas and possibly among non-domesticated animals (Ajzenberg et al. 2004). Previous studies of strains isolated in Brazil has shown that they contain unusual, recombinant genotypes when analyzed using markers based on polymorphisms identifi ed in the North American and European clonal lineages (Ferreira et al. 2006, Fux et al. 2000. Our data indicate that unusual genotypes may be found in Brazil even among domesticated pigs. Consequently, the unusual clinical presentation of ocular toxoplasmosis in Brazil may be related to the presence of these unusual genotypes. Another recent published paper studied retinal scars from human eyes after enucleation and showed that parasites containing the type I genotype were found (Vallochi et al. 2005). This study used SAG2 region, amplifi ed by polymerase chain reaction and analyzed by restriction fragment length polymorphism, so they could be analyzing recombinant strains and further testing will be necessary to establish this.
The cause for such a high incidence and severity of toxoplasmic infection in the south of Brazil is still uncertain. One possibility is that it is due infection by different strains of T. gondii than in other parts of the world; another possibility is that a high contamination of food (mainly pork) can lead to chronic re-infection. Our results have supported such theories since there has been 66% of contamination on the pork samples.