Identification of an Atypical Strain of Toxoplasma gondii as the Cause of a Waterborne Outbreak of Toxoplasmosis in Santa Isabel do Ivai, Brazil Jean D. Vaudaux1,a, Cristina Muccioli6, Erick R. James7, Claudio Silveira6, Spencer L. Magargal3, Calvin Jung7, J. P. Dubey4, Jeffrey L. Jones5, Mehmet Z. Doymaz2, David A. Bruckner2, Rubens Belfort Jr6, Gary N. Holland1, and Michael E. Grigg3 1Ocular Inflammatory Disease Center, Jules Stein Eye Institute, and the Department of Ophthalmology, David Geffen School of Medicine at UCLA, Los Angeles, California 2Department of Pathology and Laboratory Medicine, David Geffen School of Medicine at UCLA, Los Angeles, California 3Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Diseases, Bethesda 4Animal Parasitic Diseases Laboratory, Animal and Natural Resources Institute, Agricultural Research Service, United States Department of Agriculture, Beltsville, Maryland 5Division of Parasitic Diseases, National Center for Zoonotic Vector-borne and Enteric Infections, United States Centers for Disease Control and Prevention, Atlanta, Georgia 6Department of Ophthalmology, Federal University of São Paulo, Brazil 7Department of Medicine, University of British Columbia, Vancouver, Canada Abstract Multilocus DNA sequencing has identified a nonarchetypal strain of Toxoplasma gondii as the causal agent of a waterborne outbreak in Brazil in 2001. The strain, isolated from a water supply epidemiologically linked to the outbreak, was virulent to mice, and it has previously been identified as BrI. Using a serologic assay that detects strain-specific antibodies, we found that 13 (65%) of 20 individuals who were immunoglobulin (Ig) M positive during the outbreak possessed the same serotype as mice infected with the purported epidemic strain. The remaining 7 individuals, plus additional IgM-negative, IgG-positive individuals, possessed 1 of 4 novel serotypes, the most common of which matched the serotype of mice infected with strains isolated from chickens foraging near the outbreak site. The latter strains likely reflect the genetic diversity of T. gondii circulating in highly endemic regions of Brazil. The serotyping assay proved a useful tool for identification of specific individuals infected with the outbreak agent. Reprints or correspondence: Michael E. Grigg, Laboratory of Parasitic Diseases, NIAID, 4 Center Dr, Rm B1-06, Bethesda, MD 20892-0425 (griggm@niaid.nih.gov); or Gary N. Holland, Jules Stein Eye Institute, 100 Stein Plaza, UCLA, Los Angeles, CA 90095-7003 (uveitis@jsei.ucla.edu). aPresent affiliation: Jules Gonin Eye Hospital and the Department of Ophthalmology, University of Lausanne, Lausanne, Switzerland. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the US Department of Health and Human Services, the National Institutes of Health, or the Centers for Disease Control and Prevention. Potential conflicts of interest: none reported. HHS Public Access Author manuscript J Infect Dis. Author manuscript; available in PMC 2017 December 07. Published in final edited form as: J Infect Dis. 2010 October 15; 202(8): 1226–1233. doi:10.1086/656397. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript Between October and December 2001, an outbreak of human toxoplasmosis occurred in the town of Santa Isabel do Ivai, Brazil [1]. The outbreak was epidemiologically linked to a cistern that served as a municipal water supply, and it was characterized by a high prevalence of symptomatic, systemic disease [1] and associated eye involvement (C.S., C.M., J.L.J., R.B., and G.N.H., unpublished data). IgM antibody evidence of infection was found in 426 people. It has been hypothesized that the genotype of the parasite may contribute to variations in the prevalence and severity of Toxoplasma gondii infections in different populations [2, 3]. Preliminary study of the Santa Isabel outbreak identified a “type I” lineage allele at the SAG2 genotyping locus for parasites isolated from the local water supply [1]; however, most strains in Brazil are nonarchetypal and genetically distinct from the clonal type I, II, and III strains common to North America and Europe [4]. Consequently, the genotype analysis performed using the single genetic marker most likely failed to capture the true genetic “type” of the agent thought to be associated with the epidemic. The single marker was also insufficient to resolve whether 3 isolates linked to the epidemic point-source possessed the same genotype. Furthermore, no test was performed on symptomatic people infected during the outbreak to establish whether these studied strains were the true source of infection in humans. Polymerase chain reaction (PCR) analyses have been used successfully in the past to identify the genotypes of strains infecting people, but this technique requires isolation of parasite DNA from host body fluids or tissues. The morbidity associated with specimen collection usually precludes such determinations. In the absence of parasite DNA from individuals infected during the Santa Isabel outbreak, serotyping was performed to test whether isolates epidemiologically linked to the outbreak were the causal agents. The serologic assay identifies strain-specific antibodies against polymorphic epitopes in the T. gondii dense granule proteins GRA6 and GRA7, without requiring isolation of parasite DNA [5]. Specifically, we compared serologic profiles obtained from individuals infected during the outbreak with serologic profiles from mice experimentally infected with epidemic strains isolated from the contaminated water supply, to confirm that isolates epidemiologically linked to the outbreak were responsible for the infections in humans [1]. METHODS Genotyping analysis A T. gondii isolate was recovered from a cat living near the cistern implicated in the outbreak [1, 6, 7]; this isolate (TgCatBr85) previously was mistakenly designated by Dubey et al. [7] as TgCatBr38. Two isolates were obtained from pigs fed filters through which water from the cistern had been passed. Isolates were also obtained from 11 free-ranging chickens foraging in the vicinity of the cistern shortly after the outbreak. Strains isolated from chickens are believed to reflect the diversity of genotypes naturally circulating in an area where the disease is endemic [6]. Multilocus PCR DNA sequencing was performed on genomic DNA from each isolate at the B1, GRA6, and GRA7 loci.Primer sequences and PCR reaction conditions for B1, GRA6, Vaudaux et al. Page 2 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript and GRA7 have been described elsewhere [8–10]. PCR products were incubated for 15 min with ExoSAP-IT (USB) before DNA sequencing. DNA sequencing was performed by the RML Genomics Unit. Mouse infection studies A total of 5 T. gondii isolates recovered from chickens foraging near the outbreak site were tested against the epidemic strain by infecting mice; serum samples obtained from infected mice at 17 days after infection were tested using the serotyping assay to determine serologic profiles for each mouse. Serologic analysis In July 2003, serum samples (labeled B21-B40) were collected from 20 individuals in Santa Isabel who had anti–T. gondii immunoglobulin (Ig) M antibodies and symptomatic toxoplasmosis during the outbreak [1]. Specimens were anonymized before serologic testing and were provided without corresponding clinical data about disease manifestations, in compliance with regulations associated with the public health investigation. Also in 2003, potential control samples (samples B1–B20) were collected from 20 additional individuals who were seronegative for anti–T. gondii antibodies during the outbreak. Two samples (B5 and B10) possessed anti–T. gondii IgG antibodies when they were retested using a commercial immunofluorescence assay (IFA; General Biometrics) and 2 in-house assays, utilizing either recombinant SAG1 (rSAG1) or Toxoplasma-solubilized tachyzoite antigen preparation to detect anti–T. gondii IgG antibodies. These 2 samples presumably represent individuals infected with T. gondii during the 18 months after the outbreak. Thus, there were 22 seropositive (20 outbreak-associated samples and 2 other types) and 18 seronegative samples. This study was conducted in accordance with the tenets of the Declaration of Helsinki. Because specimens were collected from volunteers during a public health emergency, the study was not subject to institutional review board approval by the Federal University of São Paulo. Testing of anonymized serum samples was considered to be exempt from institutional review board approval at the University of California, Los Angeles; the US Centers for Disease Control and Prevention; and the National Institutes of Health. Peptide-based enzyme-linked immunosorbent assay The GRA6 and GRA7 strain-specific polymorphic peptides were coupled to keyhole limpet hemocyanin (KLH; Biosource), as described elsewhere [5]. Peptide names were abbreviated accordingly, with “6” denoting peptides from GRA6; “7” indicating peptides from GRA7; “I/III” or “II” denoting the peptide epitope allele; and “d” indicating a truncated version of the diagnostic peptide. Coupled peptides were diluted to 2 micrograms per milliliter in 0.1 mol/L carbonate buffer, pH 8.5. The assay was performed as described elsewhere [5] and was summarized in Table 1, which appears only in the electronic version of the Journal. Enzyme-linked immunosorbent assay (ELISA) data were presented as an optical density (OD) index by dividing the OD value obtained at 405 nm for each of the 5 serotyping peptides (6-I/III, d6-I/III, 6-II, d6-II, and 7-II) by the mean of OD readings for 2 control Vaudaux et al. Page 3 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript peptides with results expressed as arbitrary units. Serum samples obtained from unrelated individuals infected with T. gondii for whom the genotype of the infecting parasite was known were included in each experiment as positive controls. Determination of threshold values Threshold values (above which normalized assay values were considered to denote positive results) were determined by averaging the normalized OD ratio from 16 of the 18 immunofluorescence assay–negative samples and adding 2 standard deviations (SDs). The other 2 samples (B14 and B15) showed strongly false-positive reactions against single diagnostic peptides: B14 with peptide 6-I/III and B15 with peptide d6-I/III. Neither sample had a positive result, according to our 3 screening assays (immunofluorescence assay, rSAG1, and solubilized tachyzoite antigen) for anti–T. gondii IgG antibodies. False reactivity occurred in <2% of samples for each of the 5 diagnostic peptides, as determined for a cohort of 115 seronegative individuals throughout North and South America (E.R.J. and M.E.G., unpublished data). RESULTS Strain characterization In Brazil, the prevalence of T. gondii infection is high, and the population genetic structure for this parasite is highly diverse. To determine whether the waterborne outbreak was caused by a single, epidemic clone, parasite DNA was obtained from 3 isolates (designated outbreak 1, 2, and TgCatBr85) linked to the outbreak source. To interrogate the diversity of strains endemic to Santa Isabel, parasite DNA isolated from 11 chickens (TgCkBr93–TgCkBr103) foraging nearby the outbreak was tested. DNA sequences obtained at the highly polymorphic B1, GRA6, and GRA7 genotyping loci were aligned against archetypal type I, II, and III strain alleles (Figure 1). At B1, 2 alleles were identified that were equally distributed among the 14 specimens analyzed. An archetypal type I and a unique allele were detected, but the latter was closely related to a type I allele and consequently was given the designation UI (for type I–like). At GRA7, 3 alleles were identified. Two were archetypal (I or III), and the other was unique, differing from the type III allele at 2 polymorphic sites (nucleotides 168 and 436); it was designated UIII. At GRA6, 4 alleles were present. Three novel alleles were detected that diverged substantially from archetypal alleles; as such, they were designated u-1, u-2, and u-3, respectively. Intriguingly, a canonical type II allele was identified at GRA6 in 3 isolates, suggesting that type II strains have introgressed with indigenous strains circulating in this region of Brazil. The 3 isolates linked to the outbreak (outbreak 1, outbreak 2, and TgCatBr85) were genotypically identical according to multilocus DNA sequencing; all possessed the same alleles across 3 highly polymorphic typing loci, indicating a common, clonal origin for these epidemic strains (Table 2). Previous PCR–restriction fragment–length polymorphism (RFLP) genetic analyses performed on chicken isolates from Santa Isabel [11] identified 4 isolates (TgCkBr98, TgCkBr101, TgCkBr102, and TgCkBr103) that possessed the same alleles as the outbreak strains at B1, GRA6, and GRA7, and they were identical genotypically (ie, they shared the same inheritance pattern of alleles) across an additional 9 Vaudaux et al. Page 4 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript loci. This finding supports the conclusion that the outbreak genotype was nonarchetypal, clonal, and the dominant type recovered among animals foraging near the reservoir. This Brazilian clonal type has previously been designated BrI [11]. Furthermore, for the remaining 7 chicken isolates, 6 additional atypical genotypes were identified, confirming that a vast array of genetically heterogeneous strains indigenous to this endemic niche was circulating at the time of the outbreak (Table 2). Experimental infection and serotyping analysis To determine whether the purported outbreak isolate was responsible for the waterborne epidemic, Swiss-Webster mice were experimentally infected with the outbreak clone, and then the resulting mouse serotype was compared with the serotypes for people infected during the outbreak. The serologic assay identifies strain-specific IgG antibodies and previously has been successfully applied against both mice and human samples [5, 12]. As controls, mice were infected with archetypal type I and II strains; a chicken isolate (TgCkBr98) that possessed the outbreak clone BrI genotype; and 4 other chicken isolates that each possessed novel, nonarchetypal genotypes indigenous to the Santa Isabel area. All parasite strains recovered from the Santa Isabel region were virulent; mice typically died within 20 days of inoculation with as few as 50 tachyzoites (data not shown). Serum collected at day 17 from mice infected with the outbreak clone (outbreak 1), as well as with the chicken isolate that possessed the identical multilocus sequence type (TgCkBr98), reacted only with the 6-I/III epitope. IgG reactivity was not dependent on the allele present at GRA6 (u-1) but, rather, on the epitope sequence. The outbreak clone possessed the GRA6 epitope identical to that expressed by type I strains, and both yielded a type I/III serotype in the assay (Table 3). The GRA6 epitope in strain TgCkBr97 had one polymorphism (an arginine-to-histidine substitution); however, serum antibodies to infection still reacted specifically with the 6-I/III peptide. For the other non-archetypal strains TgCkBr94 and TgCkBr100 (which possessed type II alleles at GRA6), antibodies in the infection serum reacted specifically with the 6-II and 6 d-II peptides. These 2 chicken isolates did not react with the 7-II epitope, presumably because these 2 strains did not possess a type II allele sequence at GRA7 (shown by DNA sequencing). Intriguingly, strain TgCkBr95, which possessed a novel GRA6 epitope sequence LHPGSVNVFDY (containing neither the type I/III “ER” nor the type II “E” diagnostic polymorphisms), produced antibodies that reacted to both the type I/III and type II diagnostic peptides. The most parsimonious explanation for this result is that the GRA6 epitope is immunodominant and that antibodies induced by TgCkBr95 infection reacted against the nonpolymorphic residues shared in the 6-I/III and 6- II epitopes. Hence, the epitope expressed by the infecting strain ultimately determined the reactivity against the diagnostic peptides, not the genotype of the infecting strain. To test our hypothesis, serum samples from the 20 IgM-positive individuals believed to be infected during the epidemic, the two other individuals believed to have subsequent endemic disease, and the 16 seronegative control individuals were tested using the strain-typing assay, to compare serologic profiles obtained from outbreak-infected individuals with those from mice infected experimentally with (1) the epidemic clone isolated from the contaminated water supply, or (2) indigenous strains circulating in Santa Isabel. Hence, 22 T. gondii– Vaudaux et al. Page 5 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript positive samples were tested against the polymorphic peptides, and the serotype(s) of the strain(s) associated with the outbreak in humans was determined. A type II–positive control sample established that the assay was reproducible and was working within the detection and cutoff limits, as shown in Table 4 [5]. This outbreak was also particularly remarkable for the high antibody titers achieved, compared with previously tested samples from North America and Europe [5, 12]. Assay values of ≥7.0 at 6-I/III were achieved by 11 samples, which is unprecedented. Table 4 shows serotyping assay results for individuals believed to be infected in the outbreak; all values at or above the cutoff values are denoted by references to footnote c in the table. The serologic profile of mice infected with the outbreak clone was identical to the dominant serotype of individuals infected during the epidemic. Strong reactivity only at the 6-I/III peptide was possessed by 13 (65%) of 20 people. No type II serologic profiles were identified. Analysis of the remaining seropositive samples revealed a unique collection of serologic signatures (Figure 2). In 3 (17%; B21, B23, and B33) of 20 samples, the infecting strain type could not be predicted on the basis of assay results, because only weak or no reactivity against the serotyping peptides was detected. Four additional samples (B22, B30, B31, and B36) showed no allele specificity, with high reactivity to both type-specific GRA6 serotyping peptides but no reactivity to the GRA7-derived 7-II peptide. Because mixed infections are highly unusual, this latter serotype suggests infection with a nonarchetypal strain equivalent to that found in mice infected with strain TgCkBr95, which was indigenous to the Santa Isabel environs. One sample (B26) showed consistently high reactivity across the entire panel of serotyping peptides, which may also suggest infection with a nonarchetypal strain. One of the postoutbreak seropositive samples (B10) strongly reacted to both 6-I/III and 7-II, indicating yet another unique serologic signature indigenous to the area. In all, 5 serotyping patterns emerged for infected individuals. For 6 (27%) of the total 22 samples, the different combinations of reactivity with the type II and type I/III diagnostic peptides suggested either (1) the possibility of mixed infections, or (2) infections with atypical strains that possessed cross-reacting alleles different from the 3 archetypal genotypes, consistent with the diversity of endemic strains identified to be circulating in animals in Santa Isabel at the time of the outbreak. DISCUSSION This study was undertaken to determine whether T. gondii strains recovered from the water supply implicated as the source of the Santa Isabel epidemic were identical genetically and were responsible for the human disease outbreak. Because no parasites were recovered from infected individuals, no direct determination of the responsible strain type(s) could be performed. We therefore used a serologic strain-typing assay to determine “serotypes” for a group of such individuals. There was a dominant serotype in the majority (65%) of individuals, which was indistinguishable from the serotype found in serum from mice that had been infected with isolates from the implicated water supply. Furthermore, the outbreak strains were genotyped at multiple polymorphic loci and found to be clonotypic and nonarchetypal. Hence, our data support prior evidence that the municipal cistern was the point source of a clonal outbreak and that the causal agent possessed a novel genotype. Vaudaux et al. Page 6 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript It is believed that the cistern was contaminated by feral cats [1]. The water was chlorinated before consumption but not filtered. Drinking unfiltered water has been identified as a risk factor for T. gondii infection in Brazil [13, 14], but the reason for the severity of this outbreak remains enigmatic. Current evidence suggests that strain type may influence the severity of toxoplasmosis in immunocompetent hosts [2, 15, 16]. Type I or recombinant strains that possess different combinations of type I and type III alleles (like those found in Brazil) have been associated with severe ocular toxoplasmosis in people [16]. Although it was not our purpose to identify the clinical manifestations of infection with the outbreak strain, it would be useful to perform a follow-up study that investigates the relationship between disease severity and strain type. Previous studies reported the outbreak strain to be type I by PCR-RFLP analysis performed using a single genetic marker [1, 7]; however, strains in Brazil do not fit the clonal pattern prevalent in Europe and North America, and extensive testing at multiple genetic loci, followed by DNA sequencing, is necessary to resolve a true genotype. Multilocus PCR- RFLP analyses have shown that type I strains are rare in Brazil. Among 149 chicken isolates examined by this technique, only one type I strain was identified, in a chicken from southern Brazil [11]. No type I isolates have been recovered from cats and chickens in the environs of Santa Isabel; instead, there is a rich diversity of nonarchetypal strains [11, 17, 18]. Our data show that the nonarchetypal outbreak clone was also the dominant genotype infecting local chickens (4 [36%] of 11 isolates) at the time of the outbreak (Table 2). More recent multilocus PCR-RFLP analyses have shown a decreasing prevalence of the outbreak clone. In 2003, it was isolated from only 5 (18%) of 28 cats; in 2006, it was isolated from only 1 (7%) of 15 chickens [7, 17]. In both studies, new genotypes had replaced the outbreak clone as the dominant genotype. In this situation, single-locus determinations can fail to provide an accurate description of the population genetic structure. A greater level of genetic resolution is required to derive the informative associations pertaining to the evolution, expansion, and dynamics of potential epidemic clones that cause disease in high transmission zones, such as Brazil, where there is an abundance of genetically diverse, nonarchetypal strains. Nonarchetypal strains found in Brazil share alleles with type I, II, and III strains, despite the fact that these archetypal lineages are rare in Brazil [11, 19]. This fact prompted us to test whether a serotyping assay, which captures antibodies restricted to key polymorphic amino acids present in diagnostic GRA6 and GRA7 epitopes, might be able to identify accurately those individuals infected with the outbreak strain. DNA sequencing of the outbreak isolates identified a divergent GRA6 allele; however, the 12-amino-acid sequence containing the GRA6 diagnostic epitope was identical to that of type I and III strains, which was sufficient to produce a type I or III strain serotype when mice were infected with the outbreak strain. This serotype was also identified in the majority of people infected during the outbreak, clearly demonstrating that it is the epitope, not the underlying allele or genetic type, that determines the specificity of the serotyping assay. Taken together, the data provide compelling evidence that the isolates recovered from the water supply were the agents responsible for causing the outbreak. Vaudaux et al. Page 7 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript We also identified 4 additional, nonarchetypal serologic profiles among the other 7 human samples associated with the outbreak (Figure 2). We interpret these unique serologic “signature” profiles as being diagnostic in their own right, the result of infection with genetically distinct T. gondii genotypes, a variety of which were identified in chickens and cats from Santa Isabel. Atypical strains possessing different epitopes at GRA6 and GRA7 generally induce antibodies that either do not react or cross-react with nonpolymorphic amino acids present in the diagnostic epitopes [20]. These signatures suggest that the 7 individuals may have been infected before the outbreak (with systemic symptoms caused by some other, unrelated disease process) or at the time of the outbreak, but from a different source. Another possible, but unlikely, explanation for the unusual serologic results is sequential, mixed infections with 2 genetically distinct strains (assay results would reflect the presence of antibodies against both strains). Such mixed infections have been reported in cases of human toxoplasmosis [21, 22], but they remain rare. The mouse infection experiment utilizing TgCkBr95 supports a cross-reaction against nonpolymorphic amino acids as the likely cause for the nonarchetypal serotypes. It is also possible that the cistern was contaminated with more than one T. gondii strain. Without isolation and genetic characterization of the strains giving rise to specific serotypes from different individuals, further clarification is impossible. Peptides restricted to diagnostic polymorphisms contained in epitopes from indigenous strains will be promising reagents for future studies attempting to track the epidemiological dynamics of nonarchetypal strains that emerge in areas like Brazil. In summary, our results support epidemiologic evidence that a contaminated cistern was the source of the Santa Isabel epidemic, and that a nonarchetypal parasite, designated BrI, was the causal agent. The identification of a dominant serologic profile in the context of 4 additional serologic profiles indicates that a heterogeneous group of unique or divergent T. gondii genotypes was circulating in this endemic niche. The 4 additional serotypes reflect infection, unrelated to the outbreak, with other indigenous strains circulating in animals at the time of the outbreak. Our findings highlight the potential difficulty of studying an outbreak superimposed on a disease that is otherwise endemic. The strain-typing assay identified those specific individuals infected with the outbreak agent, and it may be a useful epidemiologic tool for future outbreaks that occur in areas having a large genetic diversity of parasites. Acknowledgments Financial support: This work was supported by the Intramural Research Program of the National Institutes of Health and the National Institute of Allergy and Infectious Diseases (grant AI001018 to M.E.G.), the US Department of Agriculture (Agricultural Research Service, Current Research Information System [CRIS] Project 1265-32000-076 to J.P.D.), the US Centers for Disease Control and Prevention (contract to J.D.V., G.N.H., and J.L.J.), Fonds National Suisse pour la Recherche Scientifique (PBLA33102339 to J.D.V.), Research to Prevent Blindness (RPB; support to G.N.H.), Conselho Nacional de Desenvolvimento Científíco e Tecnológico (CNPq I-A Investigator Grant to C.M., C.S., and R.B.), the Skirball Foundation, (support to G.N.H.), and the Jack H. Skirball Endowed Professorship (to G.N.H.). G.N.H. is the recipient of an RPB Physician-Scientist Award. M.E.G. is a scholar of the Michael Smith Foundation for Health Research and the Canadian Institute for Advanced Research Program for Integrated Microbial Biodiversity. Vaudaux et al. Page 8 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript References 1. deMoura L, Bahia-Oliveira LM, Wada MY, et al. Waterborne toxoplasmosis, Brazil, from field to gene. Emerg Infect Dis. 2006; 12:326–9. [PubMed: 16494765] 2. Boothroyd JC, Grigg ME. Population biology of Toxoplasma gondii and its relevance to human infection: do different strains cause different disease? Curr Opin Microbiol. 2002; 5:438–42. [PubMed: 12160866] 3. Holland GN. Ocular toxoplasmosis: a global reassessment. Part II: disease manifestations and management. Am J Ophthalmol. 2004; 137:1–17. [PubMed: 14700638] 4. Howe DK, Sibley LD. Toxoplasma gondii comprises three clonal lineages: correlation of parasite genotype with human disease. J Infect Dis. 1995; 172:1561–6. [PubMed: 7594717] 5. Kong JT, Grigg ME, Uyetake L, Parmley S, Boothroyd JC. Serotyping of Toxoplasma gondii infections in humans using synthetic peptides. J Infect Dis. 2003; 187:1484–95. [PubMed: 12717631] 6. Dubey JP, Navarro IT, Graham DH, et al. Characterization of Toxoplasma gondii isolates from free range chickens from Parana, Brazil. Vet Parasitol. 2003; 117:229–34. [PubMed: 14630431] 7. Dubey JP, Navarro IT, Sreekumar C, et al. Toxoplasma gondii infections in cats from Paranaá, Brazil: seroprevalence, tissue distribution, and biologic and genetic characterization of isolates. J Parasitol. 2004; 90:721–6. [PubMed: 15359466] 8. Bottos J, Miller RH, Belfort RN, et al. Bilateral retinochoroiditis caused by an atypical strain of Toxoplasma gondii. Br J Ophthalmol. 2009; 93:1546–50. [PubMed: 19666926] 9. Grigg ME, Boothroyd JC. Rapid identification of virulent type I strains of the protozoan pathóogen Toxoplasma gondii by PCR-restriction fragment length polymorphism analysis at the B1 gene. J Clin Microbiol. 2001; 39:398–400. [PubMed: 11136812] 10. Miller MA, Grigg ME, Kreuder C, et al. An unusual genotype of Toxoplasma gondii is common in California sea otters (Enhydra lutris nereis) and is a cause of mortality. Int J Parasitol. 2004; 34:275–84. [PubMed: 15003489] 11. Dubey JP, Velmurugan GV, Chockalingam A, et al. Genetic diversity of Toxoplasma gondii isolates from chickens from Brazil. Vet Parasitol. 2008; 157:299–305. [PubMed: 18804329] 12. Nowakowska D, Colón I, Remington JS, et al. Genotyping of Toxoplasma gondii by multiplex PCR and peptide-based serological testing of samples from infants in Poland diagnosed with congenital toxoplasmosis. J Clin Microbiol. 2006; 44:1382–9. [PubMed: 16597866] 13. Bahia-Oliveira LMG, Jones JL, Azevedo-Silva J, Alves CC, Oréfice F, Addiss D. Highly endemic, waterborne toxoplasmosis in north Rio de Janeiro state, Brazil. Emerg Infect Dis. 2003; 9:55–62. [PubMed: 12533282] 14. Heukelbach J, Meyer-Cirkel V, Moura RC, et al. Waterborne toxoplasmosis, northeastern Brazil. Emerg Infect Dis. 2007; 13:287–9. [PubMed: 17479893] 15. Fuentes I, Rubio JM, Ramirez C, Alvar J. Genotypic characterization of Toxoplasma gondii strains associated with human toxoplasmosis in Spain: direct analysis from clinical samples. J Clin Microbiol. 2001; 39:1566–70. [PubMed: 11283088] 16. Grigg ME, Ganatra J, Boothroyd JC, Margolis TP. Unusual abundance of atypical strains associated with human ocular toxoplasmosis. J Infect Dis. 2001; 184:633–9. [PubMed: 11474426] 17. Dubey JP, Sundar N, Gennari SM, et al. Biologic and genetic comparison of Toxoplasma gondii isolates in free-range chickens from the northern Pará state and the southern state Rio Grande do Sul, Brazil revealed highly diverse and distinct parasite populations. Veterinary Parasitology. 2007; 143:182–8. [PubMed: 16982151] 18. Su C, Zhang X, Dubey JP. Genotyping of Toxoplasma gondii by multilocus PCR-RFLP markers: a high resolution and simple method for identification of parasites. Int J Parasitol. 2006; 36:841–8. [PubMed: 16643922] 19. Khan A, Jordan C, Muccioli C, et al. Genetic divergence of Toxoplasma gondii strains associated with ocular toxoplasmosis, Brazil. Emerg Infect Dis. 2006; 12:942–9. [PubMed: 16707050] 20. Grigg, ME. The biology of Toxoplasma gondii. In: Soldata, D., Ajioka, JW., editors. Population genetics, sex and the emergence of clonal lines of Toxoplasma gondii.. Horizon Bioscience; Norfolk, United Kingdom: 2007. Vaudaux et al. Page 9 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript 21. Aspinall TV, Guy EC, Roberts KE, Joynson DH, Hyde JE, Sims PF. Molecular evidence for multiple Toxoplasma gondii infections in individual patients in England and Wales: public health implications. Int J Parasitol. 2003; 33:97–103. [PubMed: 12547351] 22. Elbez-Rubinstein A, Ajzenberg D, Darde ML, et al. Congenital toxoplasmosis and reinfection during pregnancy: case report, strain characterization, experimental model of reinfection, and review. J Infect Dis. 2009; 199:280–5. [PubMed: 19032062] Vaudaux et al. Page 10 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript Figure 1. Polymorphic sites at the B1 (A), GRA7 (B), and GRA6 (C) loci among archetypal lineages were compared with sequences obtained from the 3 outbreak isolates and 11 strains recovered from chickens grazing in the vicinity of the outbreak in Parana State, Brazil. Consensus sequences were determined by the nucleotide common in at least 2 of the 3 archetypal lineages. Periods (.) denote agreement with the consensus sequence. Dashes (−) denote insertions and deletions (INDELS) in the nucleotide sequence. Nucleotide positions were determined based on the published GenBank sequences for Toxoplasma gondii at the B1 gene (AF179871), GRA7 gene (Y13863.1), and GRA6 gene (AF239283). The first 3 rows contain sequences for archetypal strains I (RH), II (76K), and III (CEP), obtained from laboratory stock; TgCkBr (rows 4–12) designates isolates from chickens in Brazil; outbreak strains (last 3 rows) were obtained from filtered water from a cistern that was epidemiologically linked to the outbreak (strains 1 and 2) and a cat living near the cistern (strain 3). UI and UIII denote novel alleles that are closely related to types I and III, respectively; u-1, u-2, and u-3 denote novel alleles that are related to, but substantially divergent from, types I, II, and III, respectively. Vaudaux et al. Page 11 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript Figure 2. An enzyme-linked immunosorbent assay plate depicting the 5 distinct serotype patterns identified. The majority of samples exhibited a reactivity pattern consistent with infection by a non–type II strain (labeled “type I/III”), which supported epidemiological evidence. Three additional patterns (labeled “atypical”) were suggestive of infection by nonarchetypal strains, as evidenced by high reactivity to more than 1 allele. In 3 cases, no peptide reactivity was detected, although infection with Toxoplasma gondii was confirmed by SAG1 reactivity (labeled “no reactivity”). Reaction with a type II strain–infected control serum (bottom row) confirmed the specificity of the assay. Vaudaux et al. Page 12 J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript Vaudaux et al. Page 13 Ta b le 1 C ut -o ff V al ue s fo r Fi ve G en ot yp e- sp ec if ic P ep tid es U se d in a n Ig G S er ot yp in g A ss ay a nd f or r SA G 1 B as ed o n Se ra f ro m U ni nf ec te d In di vi du al s. ID a G en ot yp e- sp ec if ic P ep ti de sb rS A G 1b 6 I/ II I 6 d- I/ II I 6 II 6 d- II 7 II B 0 1 1. 42 1. 48 1. 63 1. 27 1. 36 1. 01 B 0 2 1. 69 1. 19 1. 32 1. 31 1. 79 1. 03 B 0 3 1. 42 1. 29 1. 33 1. 05 1. 01 0. 87 B 0 4 2. 07 1. 85 2. 28 1. 48 1. 25 0. 93 B 0 6 1. 94 2. 14 1. 52 1. 60 1. 23 1. 02 B 0 7 2. 13 2. 11 1. 86 1. 51 1. 35 0. 96 B 0 8 1. 46 1. 78 1. 73 1. 50 1. 21 1. 01 B 0 9 1. 46 1. 90 1. 64 1. 42 1. 02 1. 13 B 1 1 0. 90 1. 00 1. 00 0. 90 0. 90 1. 14 B 1 2 1. 69 2. 46 2. 44 2. 06 1. 13 0. 83 B 1 3 1. 88 1. 94 1. 42 1. 36 1. 21 1. 07 B 1 6 1. 79 2. 12 1. 91 2. 07 1. 35 0. 96 B 1 7 1. 21 1. 31 1. 35 1. 33 0. 80 1. 09 B 1 8 1. 53 2. 14 2. 00 2. 08 1. 31 1. 02 B 1 9 1. 30 2. 81 2. 08 2. 17 0. 93 1. 08 B 2 0 1. 56 1. 65 1. 31 1. 34 1. 24 0. 95 M ea nc 1. 64 1. 88 1. 72 1. 57 1. 21 1. 01 SD c 0. 28 0. 45 0. 36 0. 35 0. 23 0. 09 C ut -o ff v al ue sd 2. 2 2. 8 2. 4 2. 3 1. 7 1. 2 SD = st an da rd d ev ia tio n. a A no ny m iz ed s er um s pe ci m en n um be r. Se ru m w as f ro m in di vi du al s w ho w er e liv in g in S an ta I sa be l d o Iv ai a t t he ti m e of th e ou tb re ak u nd er in ve st ig at io n. A ll w er e co nf ir m ed to b e se ro ne ga tiv e fo r T. go nd ii Ig G a nt ib od ie s us in g a co m m er ci al a ss ay . b Fo r se ru m s am pl es , v al ue s re pr es en t o pt ic al d en si ty ( O D ; o bt ai ne d at A 40 5 nm ) va lu es f or th e di ag no st ic a nt ig en s (f iv e se ro ty pi ng p ep tid es o r rS A G 1) d iv id ed b y th e m ea n of th e O D v al ue s fo r tw o co nt ro l pe pt id es ( ra nd om iz ed s eq ue nc e of th e G R A 6 pe pt id e; m ix o f pe pt id es d er iv ed f ro m h um an a nd L ei sh m an ia s p. E F1 α p ro te in s) ; t he se n or m al iz ed v al ue s re pr es en t b ac kg ro un d re ac tiv ity a m on g th e un in fe ct ed in di vi du al s, e xp re ss ed a s ar bi tr ar y un its . V al ue s fr om o ne r ep re se nt at iv e ex pe ri m en t a re s ho w n. J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript Vaudaux et al. Page 14 c T he m ea n va lu e of th e no rm al iz ed d at a fr om f ou r in de pe nd en t e xp er im en ts . d D ef in ed a s th e m ea n va lu e of th e no rm al iz ed d at a fo r al l f ou r in de pe nd en t e xp er im en ts + 2 SD . C ut -o ff v al ue s w er e ro un de d to th e ne ar es t d ec im al p la ce . R es ul ts a bo ve th es e cu t- of f va lu es w er e co ns id er ed po si tiv e fo r th e se ro ty pi ng a ss ay a nd f or d et er m in in g rS A G 1- sp ec if ic r ea ct iv ity in th is s tu dy . J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript Vaudaux et al. Page 15 Ta b le 2 G en ot yp es o f To xo pl as m a go nd ii Is ol at es A ss oc ia te d w ith a n O ut br ea k of T ox op la sm os is in S an ta I sa be l d o Iv ai , B ra zi l Is ol at e D N A s eq ue nc e P C R -R F L P a C lo na l t yp e B 1 G R A 6 G R A 7 SA G 1 SA G 2 SA G 3 B T U B G R A 6 c2 2– 8 c2 9– 2 L 35 8 P K 1 A pi co O ut br ea k 1b I u- 1 I N A N A N A N A N A N A N A N A N A N A O ut br ea k O ut br ea k 2b I u- 1 I N A N A N A N A N A N A N A N A N A N A O ut br ea k T gC at B r8 5c I u- 1 I N A N A N A N A N A N A N A N A N A N A O ut br ea k T gC kB r9 8 I u- 1 I I I II I I II u- 1 I I I I O ut br ea k T gC kB r1 01 I u- 1 I I I II I I II u- 1 I I I I O ut br ea k T gC kB r1 02 I u- 1 I I I II I I II u- 1 I I I I O ut br ea k T gC kB r1 03 I u- 1 I I I II I I II u- 1 I I I I O ut br ea k T gC kB r9 3 U I II … I II I II I II I II I II I I II I Sa nt a Is ab el 1 T gC kB r9 4 U I II … I II I II I II I II I II I I II I Sa nt a Is ab el 1 T gC kB r9 5 U I u- 2 U II I I II I II I II I II I I I I II I II I Sa nt a Is ab el 2 T gC kB r9 6 U I u- 2 II I u- 1 II II I II I II I II I I II I I Sa nt a Is ab el 3 T gC kB r9 7 U I u- 3 U II I I II II I II I II I I II I I II II I Sa nt a Is ab el 4 T gC kB r9 9 U I u- 1 U II I I II I II I II I II u- 1 I I II I Sa nt a Is ab el 5 T gC kB r1 00 U I II U II I I II I II I II I II u- 1 I I II I Sa nt a Is ab el 6 N O T E . I , I I, I II , T ox op la sm a go nd ii ar ch et yp al c lo na l s tr ai n ty pe s; N A , n ot a va ila bl e; P C R , p ol ym er as e ch ai n re ac tio n; R FL P, r es tr ic tio n fr ag m en t– le ng th p ol ym or ph is m ; T gC K B r, T. g on di i i so la te s fr om ch ic ke ns in B ra zi l; U I- II I, u ni qu e al le le s th at a re c lo se ly r el at ed to ty pe s I, I I, a nd I II , r es pe ct iv el y; u -1 , - 2, a nd - 3, u ni qu e al le le s th at a re r el at ed to , b ut s ub st an tia lly d iv er ge nt f ro m , t yp es I , I I, a nd I II , re sp ec tiv el y. a PC R -R FL P da ta a re f ro m D ub ey e t a l. [6 , 1 1] . b Is ol at es r ec ov er ed f ro m w at er ta ke n fr om th e ci st er n th at w as e pi de m io lo gi ca lly im pl ic at ed a s th e so ur ce o f th e ou tb re ak . J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript Vaudaux et al. Page 16 c Is ol at e re co ve re d fr om a c at li vi ng in th e ar ea o f th e ci st er n. J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript Vaudaux et al. Page 17 Ta b le 3 Se ro ty pe D et er m in at io n B as ed o n th e Im m un og lo bu lin G R ea ct io n to 5 G en ot yp e- Sp ec if ic P ep tid es in M ic e In fe ct ed w ith I so la te s A ss oc ia te d w ith a n O ut br ea k of T ox op la sm os is in S an ta I sa be l d o Iv ai , B ra zi l Is ol at e G en ot yp e- sp ec if ic p ep ti de s SA G 1 Se ro ty pe G R A 6 ep it op e 6 I/ II I 6 d- I/ II I 6 II 6 d- II 7 II O ut br ea ka 4. 5b 0. 8 1. 2 1. 0 0. 9 6. 9b I/ II I L H PE R V N V FD Y T gC kB r 98 3. 5b 1. 4 0. 8 0. 7 1. 4 4. 8b I/ II I L H PE R V N V FD Y T gC kB r 97 6. 2b 1. 1 1. 3 1. 0 0. 7 7. 3b I/ II I L H PE H V N V FD Y T gC kB r 95 3. 0b 1. 5 3. 8b 1. 9 1. 2 2. 7b A ty pi ca l L H PG SV N V FD Y T gC kB r 94 1. 3 1. 0 4. 7b 3. 9b 1. 6 5. 4b II L H PG SV N E FD F T gC kB r 10 0 1. 0 1. 0 6. 8b 4. 1b 1. 0 9. 2b II L H PG SV N E FD F I 14 .5 b 1. 2 1. 3 0. 9 1. 0 19 .3 b I/ II I L H PE R V N V FD Y II 1. 1 1. 1 3. 4b 2. 3b 3. 7b 21 .2 b II L H PG SV N E FD F N il 1. 4 1. 5 1. 4 1. 4 0. 9 1. 0 N eg at iv e … N O T E . E nz ym e- lin ke d im m un oa ss ay d at a ar e pr es en te d as a n op tic al d en si ty ( O D ) in de x as f ol lo w s: th e O D v al ue o bt ai ne d at 4 05 n m f or e ac h of th e 5 se ro ty pi ng p ep tid es ( 6- I/ II I, d 6- I/ II I, 6 -I I, d 6- II , a nd 7- II ) w as d iv id ed b y th e m ea n of O D r ea di ng s fo r 2 co nt ro l p ep tid es , a nd th e re su lt is e xp re ss ed a s an a rb itr ar y un it. I , I I, I II , T ox op la sm a go nd ii ar ch et yp al c lo na l s tr ai n ty pe s; T gC kB r, T. g on di i i so la te s fr om c hi ck en s in B ra zi l. a A n is ol at e re co ve re d fr om w at er in th e ci st er n ep id em io lo gi ca lly li nk ed to th e ou tb re ak . b V al ue s th at a re g re at er th an c ut of f th re sh ol ds a nd th at r ep re se nt p os iti ve r es ul ts . C ut of f va lu es w er e de te rm in ed b y av er ag in g th e no rm al iz ed O D r at io f ro m 1 6 im m un of lu or es ce nc e as sa y– ne ga tiv e sa m pl es an d ad di ng 2 s ta nd ar d de vi at io ns ( SD s) , a s de sc ri be d in T ab le 1 , w hi ch a pp ea rs o nl y in th e el ec tr on ic v er si on o f th e Jo ur na l, us in g m et ho ds p re vi ou sl y pu bl is he d el se w he re [ 5] . J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript Vaudaux et al. Page 18 Ta b le 4 Se ro ty pe D et er m in at io n B as ed o n th e Im m un og lo bu lin G R ea ct io n to 5 G en ot yp e- Sp ec if ic P ep tid es a nd th e R ea ct io n to S A G 1 am on g In di vi du al s In fe ct ed d ur in g an O ut br ea k of T ox op la sm os is in S an ta I sa be l d o Iv ai , B ra zi l Sp ec im en a G en ot yp e- sp ec if ic p ep ti de sb rS A G 1b Se ro ty pe 6 I/ II I 6 d- I/ II I 6 II 6 d- II 7 II B 21 1. 2 0. 9 1. 0 0. 9 0. 9 8. 7c N R B 23 1. 4 1. 0 1. 0 1. 1 1. 5 18 .6 c N R B 33 1. 2 0. 9 1. 0 1. 0 1. 0 4. 4c N R B 05 3. 7c 1. 0 1. 2 1. 0 1. 7 12 .6 c I/ II I B 24 10 .8 c 1. 0 1. 7 1. 0 1. 4 15 .5 c I/ II I B 25 13 .1 c 0. 8 2. 1 0. 9 0. 9 16 .7 c I/ II I B 27 8. 1c 0. 9 1. 0 1. 1 1. 5 22 .4 c I/ II I B 28 9. 5c 0. 9 2. 1 0. 9 1. 0 17 .6 c I/ II I B 29 2. 3c 0. 8 1. 0 1. 0 1. 4 15 .9 c I/ II I B 32 4. 3c 0. 9 1. 3 1. 1 0. 9 5. 8c I/ II I B 34 4. 9c 0. 9 1. 4 1. 0 1. 0 8. 5c I/ II I B 35 3. 7c 0. 7 1. 0 0. 7 1. 2 12 .8 c I/ II I B 37 3. 2c 0. 9 1. 0 0. 9 1. 1 11 .3 c I/ II I B 38 5. 7c 0. 9 1. 1 1. 0 1. 6 10 .5 c I/ II I B 39 3. 1c 1. 0 1. 7 1. 0 1. 1 12 .2 c I/ II I B 40 9. 1c 1. 1 1. 9 1. 0 1. 1 12 .0 c I/ II I B 10 7. 8c 0. 9 1. 0 0. 9 2. 3c 4. 3c A ty pi ca l B 36 9. 3c 2. 7 3. 0c 1. 0 1. 6 12 .1 c A ty pi ca l B 22 17 .3 c 8. 7c 9. 2c 1. 0 1. 7 15 .3 c A ty pi ca l B 30 14 .6 c 3. 7c 9. 5c 1. 4 1. 2 19 .1 c A ty pi ca l B 31 10 .0 c 3. 3c 4. 9c 1. 0 0. 9 16 .3 c A ty pi ca l J Infect Dis. Author manuscript; available in PMC 2017 December 07. A uthor M anuscript A uthor M anuscript A uthor M anuscript A uthor M anuscript Vaudaux et al. Page 19 Sp ec im en a G en ot yp e- sp ec if ic p ep ti de sb rS A G 1b Se ro ty pe 6 I/ II I 6 d- I/ II I 6 II 6 d- II 7 II B 26 8. 0c 6. 1c 5. 5c 3. 8c 2. 4c 4. 0c A ty pi ca l Ty pe I Id 2. 2 1. 0 4. 2c 3. 6c 5. 5c 14 .1 c … N O T E . N R , n on re ac tiv e. C ut of f va lu es w er e as f ol lo w s: f or 6 I /I II , 2 .2 ; f or 6 d -I /I II , 2 .8 ; f or 6 I I, 2 .4 ; f or 6 d -I I, 2 .3 ; f or 7 I I, 1 .7 ; a nd f or r SA G 1, 1 .2 . T he c ut of f va lu e w as d ef in ed a s th e m ea n va lu e ± 2 st an da rd d ev ia tio ns f or a ss ay r es ul ts u si ng s er um f ro m a g ro up o f un in fe ct ed in di vi du al s (f ro m T ab le 1 , w hi ch a pp ea rs o nl y in th e el ec tr on ic v er si on o f th e Jo ur na l) . a A no ny m iz ed s er um s pe ci m en n um be r. b Fo r se ru m s am pl es , v al ue s ar e ex pr es se d as a rb itr ar y un its , o n th e ba si s of m et ho ds p re vi ou sl y de sc ri be d el se w he re [ 5, 1 2] , a nd a re s um m ar iz ed in T ab le 1 , w hi ch a pp ea rs o nl y in th e el ec tr on ic v er si on o f th e Jo ur na l. c V al ue s th at a re g re at er th an c ut of f th re sh ol ds a nd th at d en ot e po si tiv e re su lts . d Se ru m f ro m a c on tr ol in di vi du al k no w n to b e in fe ct ed w ith a ty pe I I st ra in . J Infect Dis. Author manuscript; available in PMC 2017 December 07. Abstract METHODS Genotyping analysis Mouse infection studies Serologic analysis Peptide-based enzyme-linked immunosorbent assay Determination of threshold values RESULTS Strain characterization Experimental infection and serotyping analysis DISCUSSION References Figure 1 Figure 2 Table 1 Table 2 Table 3 Table 4