Combined Use of Enzyme-Linked Immunosorbent Assay and Flow Cytometry To Detect Antibodies to Trypanosoma cruzi in Domestic Canines in Texas

CLINICAL AND DIAGNOSTIC LABORATORY IMMUNOLOGY, Mar. 2004, p. 313 319 Vol. 11, No. 2 1071-412X/04/$08.00 0 DOI: 10.1128/CDLI.11.2.313 319.2004 Copyright 2004, American Society for Microbiology. All Rights Reserved. Combined Use of Enzyme-Linked Immunosorbent Assay and Flow Cytometry To Detect Antibodies to Trypanosoma cruzi in Domestic Canines in Texas Sean V. Shadomy, Stephen C. Waring, and Cynthia L. Chappell* Center for Infectious Diseases, The University of Texas Health Science Center at Houston School of Public Health, Houston, Texas 77030 Received 10 September 2003/Returned for modification 21 October 2003/Accepted 25 November 2003 Canines may be sentinels and/or reservoirs for human Trypanosoma cruzi exposures. This study adapted a method originally designed for human diagnostics to detect serum immunoglobulin G to T. cruzi in canines. The method combined an enzyme-linked immunosorbent assay (ELISA) for screening and flow cytometry detection of anti-live trypomastigote antibodies (ALTA) for confirmation. The assays were optimized by using known positive and negative control canine sera, and cutoff values were established. The ELISA and ALTA assay easily distinguished between reactive (positive controls) and nonreactive (negative controls) sera and were used to test sera collected in a cross-sectional seroprevalence survey of 356 domestic canines from Harris County, Tex., and the surrounding area. Fifty-three (14.9%) of 356 asymptomatic canines in the survey were positive by ELISA, and 5 (1.4%) were confirmed positive with the ALTA assay, with an additional 4 (1.1%) canines classified as suspect positive. Thus, the overall prevalence of T. cruzi antibodies in this population was 2.6%. This is the first U.S. study to use the combination of ELISA and ALTA to detect serum antibodies to T. cruzi and the first report of the prevalence of T. cruzi infection in domestic canines in the Houston, Tex. (Harris County), region. Our results demonstrate that the combination of ELISA and ALTA has been successfully adapted for use in testing canines for serological evidence of T. cruzi infection. Seroprevalence survey results suggest that T. cruzi antibody-positive domestic canines in the peridomestic setting are present in the Houston, Tex., region and further suggest that T. cruzi is enzootic in the region. * Corresponding author. Present address: University of Texas School of Public Health, 1200 Herman Pressler Dr., Ste. 118A, Houston, TX 77030. Phone: (713) 500-9026. Fax: (713) 500-9020. E-mail: Cynthia.L.Chappell@uth.tmc.edu. American trypanosomiasis (Chagas disease) is a zoonotic disease caused by the protozoan parasite Trypanosoma cruzi, which is transmitted by hematophagous triatomine (reduviid) insects. The parasite is transmitted when vector feces containing metacyclic trypomastigotes is inoculated into the insect s bite site through scratching. In animals, transmission may also occur via ingestion of the infected reduviid. The parasite is endemic throughout Latin America and is a major cause of morbidity and mortality in the western hemisphere. According to the World Health Organization, an estimated 16 to 18 million persons are chronically infected with T. cruzi, and up to 50,000 deaths annually result from T. cruzi infection (26 28). The first naturally transmitted case in the United States was reported in Corpus Christi, Tex., in 1955 in a 10-month-old child residing in a triatomine insect-infested house (49, 51). There have since been four indigenous human cases of Chagas disease reported in the United States (two in Texas, one in California, and one in Tennessee) (23, 27, 40). Further, serological evidence of T. cruzi infection was found in 6 (2.5%) of 237 residents and in 6 (60%) of 10 dogs from the vicinity of the patient in California (36), and the case in Tennessee was associated with serological evidence of T. cruzi infection in a domestic canine owned by the patient s family, as well as the finding of an infected reduviid insect within the household (23). Dogs are considered important reservoirs for the transmission of T. cruzi in Latin American countries and in some areas of the United States (7, 20, 21, 35, 43) and serve as natural surveillance sentinels in areas of Latin America where vector control campaigns have been conducted (14, 22). Studies conducted in Texas have found T. cruzi infection in a number of reduviid insects. For example, T. cruzi infection was found in 286 (33.3%) of 859 reduviid insects sampled in southern Texas between 1941 and 1947 (41), and a 1978 survey in the lower Rio Grande Valley of Texas found that 22 (22.6%) of 84 reduviid insects were infected with T. cruzi (12). The persistence of T. cruzi in southern Texas triatomines was recently confirmed when approximately 80% (24 of 31) of the insects from a single residence were found to be infected (9). Seroprevalence studies of wild animals in the southern United States have detected T. cruzi infection in triatomines, coyotes, badgers, raccoons, armadillos, opossums, wood rats, and mice in Texas, California, Louisiana, Oklahoma, Alabama, Maryland, South Carolina, and Georgia (6, 12, 25, 48, 50). Seropositive dogs have been detected in both domestic and stray canine populations in Texas, Oklahoma, Louisiana, Virginia, and other southern states (7, 8, 10, 19, 44). Estimates of seroprevalence in stray-canine populations in the lower Rio Grande Valley of Texas range up to 8.8% (9, 12). T. cruzi-infected dogs from Texas have been found among canine cardiology patients referred to the Texas A&M University College of Veterinary Medicine. Nine fatal cases of canine cardiomyopathy due to American trypanosomiasis from six counties in central Texas were diagnosed between November 1972 and November 1975 (47). Between 1987 and 1996, the Texas Veterinary Medical Diagnostic Laboratory (TVMDL) reported a significant in- 313

314 SHADOMY ET AL. CLIN. DIAGN. LAB. IMMUNOL. crease (P 0.0001) in the proportion of seropositive domestic dogs among samples submitted for serological testing for T. cruzi. The proportion of positive cases increased from 1 (1.8%) of 55 samples in 1987 to 12 (17.1%) of 70 samples in 1996 (33). Between 1994 and 1998, samples from 351 domestic dogs from Texas suspected of having Chagas disease were submitted to the TVMDL for serological testing for antibodies to T. cruzi; 67 (19.1%) were seropositive (data kindly provided by C. N. Carter, TVMDL, personal communication, 8 May 1999). A number of approaches have been used in the diagnosis of Chagas disease, including xenodiagnosis, serological assays, and molecular detection methods. The use of serological assays, including radioimmunoprecipitation assay, direct and indirect hemagglutination, indirect immunofluorescent-antibody (IFA) testing, enzyme-linked immunosorbent assay (ELISA), and various combinations of these tests, has been described for the diagnosis of T. cruzi infection in canines (8, 10, 20, 21, 29, 30, 34, 52, 52). These serological tests have high sensitivity; however, their specificity may be low because of antigenic cross-reactivity with other parasitic species, such as Leishmania spp. Cross-reactivity is especially problematic when assays are based on epimastigote lysates. Intact, fixed T. cruzi epimastigotes are more specific but are still cross-reactive with T. rangeli (39). More recently, the PCR assay has been used to detect T. cruzi DNA in the blood and tissue of infected dogs (2, 10, 31). The PCR assay has shown a variable degree of efficiency in the detection of T. cruzi infection. Avila et al. found the sensitivity of the PCR assay to be 100% compared with serological results in a population of chronic chagasic and non-chagasic human patients (4). However, Junqueira et al. obtained positive PCR assay results in only 59.4% of seropositive human Chagas disease patients from a region of Brazil where the disease is endemic. Likewise, Britto et al. found that the PCR assay detected parasite DNA in only 44.7% of a population of human patients who were positive by three separate serological assays (indirect IFA testing, ELISA using the cytosolic epimastigote fraction, and ELISA using recombinant T. cruzi proteins) (11, 24). The use of flow cytometry detection of anti-live trypomastigote antibodies (ALTA) for the diagnosis of Chagas disease and detection of parasitologic cure in humans has been described (15, 32). Recently, Araujo et al. compared flow cytometry (ALTA assay) with PCR analysis and other, more conventional diagnostic methods in chronically, experimentally infected canines. In these dogs, at least two serial serum samples and up to nine repeated PCR assays per sample were required to obtain positive PCR results compared to the ALTA assay and ELISA, both of which yielded positive results with all of the samples tested. In contrast, xenodiagnosis and hemoculture were positive in only 11 and 22%, respectively, of the chronically infected canines (2). The purpose of this study was to optimize the (screening) ELISA by using intact, fixed epimastigotes and the flow cytometry (confirmatory) assay by using ALTA for testing canine sera. Secondly, the combination of a screening ELISA and the ALTA confirmatory test was used in a serological study of a domestic canine population in Harris County (Houston), Tex. (This research was conducted in partial fulfillment of the requirements for the degree of Master of Public Health [S. V. Shadomy].) MATERIALS AND METHODS Collection of canine sera. Fourteen ZIP code areas covering northwestern Harris County, Tex., and adjacent regions were chosen as the study area for the seroprevalence survey. Veterinary clinics in the study area were identified by using two separate Internet search engines and the regional business directory. Clinics were then randomized, contacted by investigators, and invited to participate. Serum samples of unknown T. cruzi antibody status were collected from 356 domestic canines visiting 11 clinics. Clients were asked by the attending veterinarian to participate in the study. Dogs were eligible for inclusion in the study if (i) the owner resided within the study region as verified by the ZIP code of the residence, (ii) the dog was determined to be healthy at the time of the clinic visit by the veterinarian, (iii) the dog was 6 months of age or older, and (iv) routine venipuncture was to be performed during the clinical visit, either for evaluation of canine heartworm (Dirofilaria immitis) infection status, for health screening for elective surgical procedures, or for other routine clinical purposes. Veterinary staff obtained a 2- to 3-ml blood sample by venipuncture from each dog enrolled in the study; blood was collected in serum separator tubes (Sherwood Medical Co., St. Louis, Mo.). Blood samples were centrifuged after clotting and stored at 4 C at the clinic until they were transported (weekly) to the Center for Infectious Disease, University of Texas Houston Health Science Center School of Public Health for analysis. This study was approved by the University of Texas Houston Health Science Center Animal Welfare Committee and Committee for the Protection of Human Subjects. Positive control sera (kindly provided by O. Martins-Filho, Centro de Pesquisas Rene Rachou, Oswaldo Cruz Foundation, Belo Horizonte, MG, Brazil) were collected from seven Brazilian dogs experimentally infected with the CL strain of T. cruzi. These sera were collected at time points between 28 and 55 days postinfection or from chronically infected dogs. A pool of positive control sera was made with equal volumes of serum from the seven experimentally infected dogs. A negative control serum was obtained from a local domestic canine with no known history of exposure to American trypanosomiasis and which tested negative for antibodies to T. cruzi at the TVMDL. Additional sera used to further evaluate the assays were obtained from three domestic canines that tested negative for antibodies to T. cruzi at the TVMDL (sera kindly provided by D. Jordan, Houston, Tex.; S. L. Hurwitz, Austin, Tex.; and D. W. Tinkey, Houston, Tex.). For these three domestic canines, serologic testing for T. cruzi infection was performed during clinical evaluation or because of a family history of cardiac disease; however, other etiologies were identified during the diagnostic evaluation and Chagas disease was ruled out by exclusion. Cultivation of parasites. T. cruzi epimastigotes were cultivated for use as the antigen for the screening ELISA as follows. Liver infusion tryptose (LIT) medium was prepared with 10% fetal bovine serum (FBS), 100 U of penicillin per ml, 100 g of streptomycin per ml, and 40 g of gentamicin per ml. T. cruzi CL strain trypomastigotes (kindly provided by O. Martins-Filho) were used to inoculate LIT medium (10 7 parasites in 10 ml of medium), and the culture was incubated at 28 C in 50-ml conical polycarbonate tubes (Sarstedt, Inc., Newton, N.C.). The tubes were gently agitated daily to aerate the medium. After 14 days, epimastigotes were observed in the medium and the culture was expanded to a 25-ml total volume. The epimastigotes were passed to fresh medium every 10 days and harvested during the exponential growth phase on day 8, 9, or 10 postpassage for use as the antigen in the ELISA. The culture was clarified of cellular debris by centrifugation at a relative centrifugal force of 134 for 10 min at room temperature, and the epimastigotes were washed three times in 25 ml of 0.15 M phosphate-buffered saline (PBS), ph 7.2, containing 2% bovine serum albumin and 0.02% sodium azide. Parasites were fixed by mixing with an equal volume of 4% paraformaldehyde in PBS and incubated overnight at 4 C. The fixed parasites were washed twice in PBS, resuspended in 10 ml of PBS with 0.02% sodium azide, and stored at 4 C. T. cruzi CL strain trypomastigotes were cultivated in an enterocyte cell line for use as the antigen in the flow cytometry confirmation assay as follows. HCT-8 human ileocecal adenocarcinoma tumor cells (American Type Culture Collection, Manassas, Va.) were incubated in RPMI 1640 medium (Sigma Chemical Co., St. Louis, Mo.) containing L-glutamine, 5% FBS and 8 g of gentamicin per ml at 37 C in 75-cm 2 polystyrene tissue culture flasks (Corning Glass Works, Corning, N.Y.). Once the enterocytes adhered to the flask surface and formed a 5 to 10% confluent monolayer, the medium was replaced with RPMI medium containing 10% FBS, and 10 7 trypomastigotes were added to the medium. The flasks were incubated at 37 C for 24 h before the medium was replaced with RPMI medium containing 5% FBS. Flasks were further incubated at 33 C, and trypomastigotes were harvested on day 5, 6, or 7 postinfection at the peak of the first-generation trypomastigote release from the infected monolayer. Harvested

VOL. 11, 2004 ELISA AND FLOW CYTOMETRY DETECTION OF IgG TO T. CRUZI 315 trypomastigotes were removed from culture supernatant by centrifugation at a relative centrifugal force of 1,010 for 10 min at 4 C and washed three times in 25 ml of PBS containing 10% FBS. Live trypomastigotes were then resuspended in PBS 10% FBS, and the trypomastigote concentration was determined with a hemacytometer. The suspension was then adjusted to 5 10 6 trypomastigotes per ml and used immediately as the antigen in the flow cytometry assay. Screening ELISA. Canine serum samples were tested for antibodies to T. cruzi by using a modification of protocols previously described (1, 8, 29). Fixed whole epimastigotes were washed and resuspended at a concentration of 500,000 epimastigotes per ml in 0.05 M carbonate buffer, ph 9.6. A volume of 100 l was added to each well of a 96-well polycarbonate microtiter plate (Nalge-Nunc NUNC-Immuno Plates with MaxiSorp Surface; Nalgene-Nunc, Rochester, N.Y.), and the plate was incubated at 37 C for 1 h and then incubated overnight at 4 C. On the following day, the plate was washed with ELISA wash buffer containing 0.15 M PBS (ph 7.2) with 0.1% polysorbate 20 (Tween 20) (ICI Americas, Inc., Wilmington, Del.) for three cycles in an automatic plate washer (TiterTek Smartwash 3; Titertek, Huntsville, Ala.). Blocking buffer (200 l) consisting of 5% nonfat powdered milk in ELISA wash buffer was added to the plate wells, and the plate was incubated at 37 C for 90 min. The plate was washed as described above, and 100 l of canine serum diluted in blocking buffer was added to wells. The pooled positive control sera and the negative control serum were tested in triplicate on each plate. Reagent controls and canine serum samples of unknown antibody status were each tested in duplicate. The plate containing canine sera was incubated at 37 C for 60 min and washed as described above. Horseradish peroxidase-labeled goat anti-dog immunoglobulin G (IgG; 100 l; ICN Pharmaceuticals, Aurora, Ohio) diluted 1:2,000 in blocking buffer was added to each well. The plate was incubated at 37 C for 60 min and then washed free of unbound antibody. Each well then received 50 l of2,2 -azinobis(3-ethylbenzthiazolinesulfonic acid) (ABTS) (6) activated with freshly added 0.05% hydrogen peroxide. Absorbances (414 nm) were monitored at 0, 3, 5, 10, 15, and 30 min postreaction with an automated plate reader (TiterTek MultiScan MCC/340). Results were expressed as the mean absorbance of duplicate (unknown sera) or triplicate (positive and negative control sera) wells. The screening ELISA was optimized by using a checkerboard dilution pattern of pooled sera from experimentally infected dogs or serum from a confirmed negative dog versus various concentrations of epimastigote antigen. Antigen concentrations ranged from 7,500 to 100,000 epimastigotes per well, and sera were serially diluted 1:20 to 1:640. Two repetitions of this experiment were conducted. For screening unknown canine serum samples, 50,000 epimastigotes per well and serum dilutions of 1:80 and 1:160 were used on all plates. The mean absorbance values for the positive (pooled) and negative control sera on each plate were compared to monitor plate-to-plate variation and reliability. To categorize unknown canine serum samples, a cutoff value for positivity was empirically established as a mean absorbance value 1.5 times the mean absorbance of the negative control on the same plate. The cutoff value was determined for each ELISA plate to standardize for possible variation. Flow cytometry (ALTA) assay. Flow cytometric analysis for the ALTA assay was performed as described by Martins-Filho et al. (32), except that canine sera were used in place of human sera and fluorescein isothiocyanate (FITC)-labeled mouse anti-dog IgG (1:400 dilution; Oswaldo Cruz Institute/Bio-Manguinhos, Rio de Janeiro, Brazil) was substituted for FITC-labeled anti-human IgG. Samples were analyzed with a Coulter/EPICS XL flow cytometer (Coulter Corporation, Miami, Fla.). For each assay, 10,000 events were analyzed and the trypomastigotes were identified and selectively gated on the basis of their specific forward and side light-scattering properties. The data were expressed as the relative percentage of positive fluorescent parasites (PPFP), and each serum sample was analyzed in duplicate to yield a mean PPFP. The WinMDI software package was used for both flow cytometry data storage and analysis. The optimum antigen and serum concentrations were determined by using 500,000, 250,000, and 125,000 live trypomastigotes per reaction versus twofold serial dilutions (1:32 to 1:2,048) of positive (pooled) or negative control sera. PPFP cutoff points were determined with 14 repetitions of the negative control. A positive result was defined as a PPFP that exceeded 3 standard deviations (SD) above the mean negative control value. Data analysis. The chi-square test or Fisher s exact test was used to compare categorical variables, and Student s t test (with the Welch correction, when warranted) was used for continuous variables. All calculations and statistical comparisons were done with the SAS, version 8, software (SAS Corporation, Cary, N.C.). FIG. 1. Serum titration curve comparing absorbance values for serial dilutions of positive control serum (pooled), negative control serum, and five serum samples from dogs experimentally infected with the CL strain of T. cruzi. Absorbance was measured at 414 nm 15 min after the substrate was added. Various serial dilutions (1:10 to 1:1,280) of serum are indicated. Fixed T. cruzi CL strain epimastigotes (50,000 per well) were used as the antigen. Experimentally infected dog sera were collected at 28 (dog 1), 29 (dog 2), 54 (dog 9), and 55 (dog 10) days postinfection or identified as being from a chronically infected dog (dog 13). The ratio of the absorbance of the positive control to that of negative control is shown above the graph for each serum dilution. RESULTS Optimization of the ELISA. The checkerboard titration pattern compared the degree of reactivity between various epimastigote antigen concentrations and serum antibody dilutions (data not shown). The highest signal-to-noise ratios (5.24, 5.43, and 5.88) between the positive and negative control sera occurred when 25,000 epimastigotes at serum dilutions of 1:80, 1:160, and 1:320, respectively, were tested. Tests of 356 canine sera of unknown T. cruzi antibody status were done with an antigen concentration of 50,000 epimastigotes per well and serum dilutions of 1:80 and 1:160. Including data from all of the microtiter plates (n 12) used in the serological tests, the net absorbance (i.e., minus that of the reagent control) for the positive and negative control sera yielded mean optical densities of 1.040 0.067 (standard error of the mean [SEM]) and 0.213 0.174 (SEM), respectively. These absorbances of positive and negative control sera were significantly different (P 0.0001). Positive (pooled) and negative control sera or sera from five acutely infected dogs or one chronically infected dog were tested with serum dilutions of 1:10 to 1:1,280 (Fig. 1). Acutely infected dogs yielded lower reactivity than the chronically infected dog and showed a general trend of increased reactivity as the duration of the infection increased. Optimization of the flow cytometry ALTA assay. The antigen titration curve (500,000, 250,000, and 125,000 live trypomastigotes per ALTA reaction) with positive control sera (pooled) yielded PPFPs of 38.1, 34.8, and 32.0%, respectively (data not shown). The highest signal-to-noise ratios were obtained at 250,000 and 500,000 trypomastigotes per reaction, and 500,000 trypomastigotes were used thereafter. The interassay variabilities of the positive and negative control sera were 87.1 3.32 (SEM) and 2.13 0.660 (SEM), respectively (P 0.0001). Serum titration curves with twofold serial dilutions (1:32 to

316 SHADOMY ET AL. CLIN. DIAGN. LAB. IMMUNOL. FIG. 2. Histogram spectra showing the FITC fluorescence intensities of the positive control sera (A) and the negative control serum (B). Each histogram shows the output for one of two FITC fluorescence readings taken of each sample on a repetition of the flow cytometry assay. The PPFP indicated above each histogram represents the percentage of parasite particle events exceeding the fluorescence cutoff (log 10 1 ). FIG. 3. Mean PPFPs for positive (Œ) and negative ( ) control sera, known negative domestic canine sera ( ), and seroprevalence study serum samples (F). Each circle represents the mean PPFP for duplicate tests of a single serum sample. T. cruzi CL strain trypomastigotes were used as the antigen in the assay. The cutoff PPFP (8.81%) shown on the y axis represents 3 SD above the mean negative control. 1:2,048) of positive and negative control sera yielded optimal results at a serum dilution of 1:256 (Fig. 2). The two event peaks are clearly distinguishable, with only 1.74% of the parasites incubated with negative serum fluorescing above the 10 1 fluorescence level, while 90.8% of the parasites incubated with positive serum occur above the same fluorescence intensity. Repetitions (n 14) of the ALTA assay performed with the negative control serum (at a 1:256 dilution) yielded a mean PPFP of 2.90% 1.97% (SD). The cutoff value for a negative result was set as the mean PPFP plus 3 SD, which equaled a PPFP of 8.81%. Further, sera obtained from the three local domestic canines that tested negative for antibodies to T. cruzi at the TVMDL also yielded negative results (PPFPs 1.0, 2.1, and 2.7%) in the ALTA assay (Fig. 3). Mean PPFPs for positive control sera (1:256) from six experimentally infected dogs ranged from 18.1 to 91.7% (Fig. 3). The lowest mean PPFP for any experimentally infected dog was 18.1%, a value associated with the most acutely infected dog. Because of the lack of clinical confirmation of serological results, PPFPs between 8.81 and 18.1% were considered suspect positive to avoid potential misclassification as true positives. Seroprevalence survey of domestic canines from the Harris County, Tex., region. Canine sera collected for the seroprevalence survey were tested at a 1:80 dilution with the ELISA. All survey samples with positive ELISA results and 30 randomly selected survey samples with negative ELISA results were retested at a dilution of 1:160. Of the 356 samples tested at a dilution of 1:80, 53 (14.89%) were considered positive in the screening ELISA. Thirty-four (64.2%) of the 53 samples positive in the ELISA at a 1:80 dilution were also positive at a 1:160 dilution. All 30 sera that were negative at the 1:80 dilution were also negative at the 1:160 dilution (Table 1). The 53 samples testing positive in the ELISA were tested by flow cytometry ALTA assay for confirmation of results (Fig. 3). Of these samples, 44 were considered to be negative, with an average PPFP of 1.83%. Nine sera yielded values above the cutoff point of 8.81%. Of these, five samples (1.4%) had a mean PPFP that fell within the range of PPFPs from dogs serving as positive controls. The remaining four sera (1.1%) fell between the 8.81% cutoff and the lowest value (18.1%) from a dog with a confirmed infection. These sera were considered to be suspect positives. Thus, the overall prevalence of T. cruzi antibodies in this population of 356 asymptomatic dogs, including the suspect positives, was 2.6%. DISCUSSION Previous studies have indicated that T. cruzi is endemic among wild animals in the United States. Local exposure to infected vectors and immigration of infected individuals from TABLE 1. Serological testing of 356 domestic canines from the Harris County, Tex., region a Test and serum dilution No. of samples tested/no. positive % Positive ELISA 1:80 356/53 14.9 1:160 83/34 9.55 ALTA flow cytometry, 1:256 53/5 1.4 b a Screening (ELISAs) were done on each serum sample, and positive sera were confirmed by ALTA assay. b An additional four samples (1.2%) were suspect positive, increasing the percent positive value to 2.6%.

VOL. 11, 2004 ELISA AND FLOW CYTOMETRY DETECTION OF IgG TO T. CRUZI 317 Central and South American countries ensure that human infections are present in the United States but may remain undetected. When human cases have occurred in the United States, the family dog has often been found to be infected (23). Thus, dogs may act not only as sources of parasite transmission in the environment but also as sentinels for infected vectors that have the potential of feeding on humans. A high percentage of dogs with Chagas disease suffer acute and/or chronic health problems; no effective drugs are currently available for treating T. cruzi infection in canines. Undiagnosed human infections can be problematic since the parasite can be acquired congenitally and can be transmitted to others via the blood supply. Also, persons with chagasic heart disease should be specifically managed. Because it is difficult to study the human population, surveillance of T. cruzi in domestic canines may prove useful in assessing potential threats of human exposure. To date, however, no standardized testing methods for canine testing have been developed and used in a seroprevalence study. It should be noted that proof of infection must rely on the direct visualization of parasites in tissues (often postmortem) or by finding the epimastigotes in laboratory-raised reduviid insects that have fed on the blood of animals or humans with suspected cases. Each of these methods has its own detection limits and associated problems. Serological assays are a convenient and practical method of detecting active infections with T. cruzi. In human infections, a positive diagnosis of Chagas disease relies on two positive tests done with different methods (ELISA, IFA testing, or indirect hemagglutination). Diagnosis and/or prevalence studies of veterinary populations face the same difficulties. The present study is the first to describe the use of flow cytometry detection (ALTA assay) in the testing of canines for American trypanosomiasis in the United States. This is also the first study to report a seroprevalence estimate of T. cruzi infection in a population of owned, asymptomatic domestic canines in Texas. The results indicate that the combination of the ELISA as a screening assay with confirmation testing by flow cytometry (ALTA assay), previously used for human studies, can be adapted for use in canine population studies. This combination of screening and confirmatory methods should be valuable in diagnosing T. cruzi infection in domestic canines and in conducting population surveillance studies. Previous studies of T. cruzi infection in canines in the United States have used a variety of test protocols, many with fixed epimastigote or trypomastigote antigens. Although sensitivity may be high with fixed parasite antigens, specificity is typically low because of cross-reactivity with other parasites. This is especially important in assays that use fixed epimastigote antigens (39). In contrast, the antigen that we used in the screening was intact epimastigotes, which limited antibody reactivity to surface proteins and decreased the amount of expected cross-reactivity. The flow cytometry ALTA assay was used to confirm the screening test results. This method has an increased specificity, an important factor in conducting seroprevalence studies. The greater specificity of this assay results from the use of intact live parasites (2, 15), which even further reduces the opportunity for nonspecific cross-reactivity. In this assay, surface proteins on live trypomastigotes are recognized by specific antibodies from the host. Because the trypomastigotes are amplified in cultured cells, they provide a convenient and consistent source of antigen for the flow cytometry assay. Although a number of serological assays with recombinant or purified parasite antigens have been described for use in diagnosing T. cruzi infection in humans (16 18, 37, 38, 45, 46), these assays have not been adapted or standardized for use with canine sera. Thus, we chose to combine the ELISA and ALTA assay systems in order to minimize the occurrence of false negatives with the screening assay (ELISA) and eliminate potential false positives with the confirmatory (ALTA) assay. In general, experimentally infected canines yielded absorbances and PPFPs that increased with the duration of the infection. In the ALTA, the experimentally infected dogs had PPFPs of 18.1 to 92.5%. The lowest value was consistent with the 20% PPFP cutoff value found in previous reports of human studies (2, 15, 32). It is of note that the canines that were the source of the positive control sera used in this study were experimentally infected with the same T. cruzi strain used as the antigen in this study. The use of a South American strain of T. cruzi for the serological assays of Texas dogs may have contributed to the results obtained with the four suspect positive sera. That is, antigens from a local T. cruzi strain may have yielded higher values in animals with a locally acquired infection than did the South American strain of the parasite. Indeed, genetic divergence and antigenic variation between T. cruzi strains from various regions are known to occur (3, 5, 13, 42). Therefore, the true prevalence may be closer to the 2.6% level that would include all of the positive sera, as well as the suspect positive sera. The estimated seroprevalence of 1.4 to 2.6% detected in this study is lower than those found in many previous studies of American trypanosomiasis in canines in the United States and in stray canine populations in Texas. The prevalence estimate reported here may be influenced by the fact that the survey population consisted of healthy domestic canines; canines with possible clinical illness due to infection or other comorbid conditions were excluded from the study. These restrictions were placed on subject selection in an effort to limit crossreactivity from other clinical illnesses. Even so, only 9 (17%) of 53 ELISA-positive samples were confirmed by flow cytometry (ALTA assay). Thus, the selection of healthy domestic dogs is likely to underestimate the prevalence of infection in stray dog populations or those animals that have symptoms consistent with T. cruzi infection. Another potential factor in underestimating the prevalence of infection may be related to the antigen source used in the serological tests. Parasites used in this study and in many others were from strains circulating in South America. Also, the positive control sera were from dogs that were infected with the homologous strain. In human populations, clinical manifestations of T. cruzi infection are known to vary with the strain of the parasite and geographic location. Likewise, serological differences may occur because of antigenic variation in epimastigote and trypomastigote surface proteins in the various parasite strains. Thus, use of a native North American strain of T. cruzi in the present study might have influenced the level of seropositivity detected in local canines. We are currently testing this hypothesis. Asymptomatic dogs from domestic environments were cho-

318 SHADOMY ET AL. CLIN. DIAGN. LAB. IMMUNOL. sen for this study, in part, as an indication of peridomestic exposure. Our study indicates that nine households had evidence of infected vectors in or near the household. Further, among the owners of dogs enrolled in the study who provided information, 38 (11.38%) of 334 indicated that they had seen reduviid insects around their premises (data not shown). This suggests that the cycle of T. cruzi transmission may be occurring in or around the household and may pose a potential health threat for the inhabitants. Further studies are needed to evaluate the prevalence of American trypanosomiasis in the domestic canine in the United States and to provide more information regarding exposure and risk factors for infection. Domestic dogs may serve as a sentinel for the presence of endemic T. cruzi in regions inhabited by reduviid vector species. Because of the potential public health risk suggested by these results, human serological studies are also needed to determine if T. cruzi infection exists in the human population in Texas, particularly along the Texas-Mexico border. Since Chagas disease in humans can be asymptomatic for long periods of time and can be transmitted congenitally and through the blood supply, serological tests specifically designed for the United States will enhance our ability to identify infected individuals, provide appropriate medical care, and prevent unnecessary spread of infection to others. ACKNOWLEDGMENTS We are grateful to Olindo Martins-Filho of the Centro de Pesquisas René Rachou, Oswald Cruz Foundation, Belo Horizonte, MG, Brazil, for scientific expertise regarding serological assay development and culture techniques and for providing the CL strain T. cruzi trypomastigote and epimastigote stocks and the positive control canine sera used in this study. REFERENCES 1. Antas, P. R., E. N. Azevedo, M. R. Luz, N. Medrano-Mercado, A. C. Chaves, P. G. Vidigal, A. C. Volpini, A. J. Romanha, and T. C. Araujo-Jorge. 2000. A reliable and specific enzyme-linked immunosorbent assay for the capture of IgM from human chagasic sera using fixed epimastigotes of Trypanosoma cruzi. Parasitol. Res. 86:813 820. 2. Araujo, F. M., M. T. Bahia, N. M. Magalhaes, O. A. Martins-Filho, V. M. Veloso, C. M. Carneiro, W. L. Tafuri, and M. Lana. 2002. Follow-up of experimental chronic Chagas disease in dogs: use of polymerase chain reaction (PCR) compared with parasitological and serological methods. Acta Trop. 81:21 31. 3. Avila, H., A. M. Goncalves, N. S. Nehme, C. M. Morel, and L. Simpson. 1990. Schizodeme analysis of Trypanosoma cruzi stocks from South and Central America by analysis of PCR-amplified minicircle variable region sequences. Mol. Biochem. Parasitol. 42(2):175 187. 4. Avila, H. A., J. B. Pereira, O. Thiemann, E. De Paiva, W. DeGrave, C. M. Morel, and L. Simpson. 1993. Detection of Trypanosoma cruzi in blood specimens of chronic chagasic patients by polymerase chain reaction amplification of kinetoplast minicircle DNA: comparison with serology and xenodiagnosis. J. Clin. Microbiol. 31:2421 2426. 5. Barnabe, C., R. Yaeger, O. Pung, and M. Tibayrenc. 2001. Trypanosoma cruzi: a considerable phylogenetic divergence indicates that the agent of Chagas disease is indigenous to the native fauna of the United States. Exp. Parasitol. 99:73 79. 6. Barr, S. C., C. C. Brown, V. A. Dennis, and T. R. Klei. 1991. The lesions and prevalence of Trypanosoma cruzi in opossums and armadillos from southern Louisiana. J. Parasitol. 77:624 627. 7. Barr, S. C., V. A. Dennis, and T. R. Klei. 1991. Serologic and blood culture survey of Trypanosoma cruzi infection in four canine populations of southern Louisiana. Am. J. Vet. Res. 52:570 573. 8. Barr, S. C., O. Van Beek, M. S. Carlisle-Nowak, J. W. Lopez, L. V. Kirchhoff, N. Allison, A. Zajac, A. de Lahunta, D. H. Schlafer, and W. T. Crandall. 1995. Trypanosoma cruzi infection in Walker hounds from Virginia. Am. J. Vet. Res. 56:1037 1044. 9. Beard, C. B., G. Pye, F. J. Steurer, R. Rodriguez, R. Campman, A. T. Peterson, J. Ramsey, R. A. Wirtz, and L. E. Robinson. 2003. Chagas disease in a domestic transmission cycle, southern Texas, USA. Emerg. Infect. Dis. 9(1):103 105. 10. Bradley, K. K., D. K. Bergman, J. P. Woods, J. M. Crutcher, and L. V. Kirchhoff. 2000. Prevalence of American trypanosomiasis (Chagas disease) among dogs in Oklahoma. J. Am. Vet. Med. Assoc. 217:1853 1857. 11. Britto, C., M. A. Cardoso, C. Ravel, A. Santoro, J. B. Pereira, J. R. Coura, C. M. Morel, and P. Wincker. 1995. Trypanosoma cruzi: parasite detection and strain discrimination in chronic chagasic patients from northeastern Brazil using PCR amplification of kinetoplast DNA and nonradioactive hybridization. Exp. Parasitol. 81(4):462 471. 12. Burkholder, J. E., T. C. Allison, and V. P. Kelly. 1980. Trypanosoma cruzi (Chagas) (Protozoa: Kinetoplastida) in invertebrate, reservoir, and human hosts of the lower Rio Grande valley of Texas. J. Parasitol. 66:305 311. 13. Burns, J. M., Jr., W. G.Shreffler, D. E. Rosman, P. R. Sleath, C. J. March, and S. G. Reed. 1992. Identification and synthesis of a major conserved antigenic epitope of Trypanosoma cruzi. Proc. Natl. Acad. Sci. USA 89:1239 1243. 14. Castanera, M. B., M. A. Lauricella, R. Chuit, and R. E. Gurtler. 2002. Evaluation of dogs as sentinels of the transmission of Trypanosoma cruzi in a rural area of north-western Argentina. Ann. Trop. Med. Parasitol. 92:671 683. 15. Cordeiro, F. D., O. A. Martins-Filho, M. O. Da Costa Rocha, S. J. Adad, R. Correa-Oliveira, and A. J. Romanha. 2001. Anti-Trypanosoma cruzi immunoglobulin G1 can be a useful tool for diagnosis and prognosis of human Chagas disease. Clin. Diagn. Lab. Immunol. 8:112 118. 16. da Silveira, J. F., E. S. Umezawa, and A. O. Luquetti. 2001. Chagas disease: recombinant Trypanosoma cruzi antigens for serological diagnosis. Trends Parasitol. 17:286 291. 17. Ferreira, A. W., Z. R. Belem, E. A. Lemos, S. G. Reed, and A. Campos-Neto. 2001. Enzyme-linked immunosorbent assay for serological diagnosis of Chagas disease employing a Trypanosoma cruzi recombinant antigen that consists of four different peptides. J. Clin. Microbiol. 39:4390 4395. 18. Gomes, Y. M., V. R. Pereira, M. Nakazawa, D. S. Rosa, M. D. Barros, A. G. Ferreira, E. D. Silva, S. F. Ogatta, M. A. Krieger, and S. Goldenberg. 2001. Serodiagnosis of chronic Chagas infection by using EIE-Recombinant-Chagas-Biomanguinhos kit. Mem. Inst. Oswaldo Cruz 96:497 501. 19. Grogl, M., R. E. Kuhn, D. S. Davis, and G. E. Green. 1984. Antibodies to Trypanosoma cruzi in coyotes in Texas. J. Parasitol. 70:189 191. 20. Gurtler, R. E., M. C. Cecere, D. N. Rubel, R. M. Petersen, N. J. Schweigmann, M. A. Lauricella, M. A. Bujas, E. L. Segura, and C. Wisnivesky-Colli. 1991. Chagas disease in north-west Argentina: infected dogs as a risk factor for the domestic transmission of Trypanosoma cruzi. Trans. R. Soc. Trop. Med. Hyg. 85:741 745. 21. Gurtler, R. E., J. E. Cohen, M. C. Cecere, M. A. Lauricella, R. Chuit, and E. L. Segura. 1998. Influence of humans and domestic animals on the household prevalence of Trypanosoma cruzi in Triatoma infestans populations in northwest Argentina. Am. J. Trop. Med. Hyg. 58:748 758. 22. Gurtler, R. E., F. O. Kravetz, R. M. Petersen, M. A. Lauricella, and C. Wisnivesky-Colli. 1990. The prevalence of Trypanosoma cruzi and the demography of dog populations after insecticidal spraying of houses: a predictive model. Ann. Trop. Med. Parasitol. 84:313 323. 23. Herwaldt, B. L., M. J. Grijalva, A. L. Newsome, C. R. McGhee, M. R. Powell, D. G. Nemec, F. J. Steurer, and M. L. Eberhard. 2000. Use of polymerase chain reaction to diagnose the fifth reported US case of autochthonous transmission of Trypanosoma cruzi, in Tennessee, 1998. J. Infect. Dis. 181: 395 399. 24. Junqueira, A. C., E. Chiari, and P. Wincker. 1996. Comparison of the polymerase chain reaction with two classical parasitological methods for the diagnosis of Chagas disease in an endemic region of north-eastern Brazil. Trans. R. Soc. Trop. Med. Hyg. 90(2):129 132. 25. Karsten, V., C. Davis, and R. Kuhn. 1992. Trypanosoma cruzi in wild raccoons and opossums in North Carolina. J. Parasitol. 78:547 549. 26. Kirchhoff, L. V. 1989. Is Trypanosoma cruzi a new threat to our blood supply? Ann. Intern. Med. 111:773 775. 27. Kirchhoff, L. V. 1993. American trypanosomiasis (Chagas disease) a tropical disease now in the United States. N. Engl. J. Med. 329:639 644. 28. Kirchhoff, L. V. 1993. Chagas disease. American trypanosomiasis. Infect. Dis. Clin. N. Am. 7:487 502. 29. Lauricella, M. A., M. B. Castanera, R. E. Gurtler, and E. L. Segura. 1998. Immunodiagnosis of Trypanosoma cruzi (Chagas disease) infection in naturally infected dogs. Mem. Inst. Oswaldo Cruz 93:501 507. 30. Lauricella, M. A., C. Wisnivesky-Colli, R. Gurtler, R. Petersen, M. Bujas, and E. L. Segura. 1993. Standardization of serological tests for detecting anti-trypanosoma cruzi antibodies in dogs. Mem. Inst. Oswaldo Cruz 88:413 417. 31. Machado, E. M., A. J. Fernandes, S. M. Murta, R. W.Vitor, D. J. Camilo, Jr., S. W. Pinheiro, E. R. Lopes, S. J. Adad, A. J. Romanha, and J. C. Pinto Dias. 2001. A study of experimental reinfection by Trypanosoma cruzi in dogs. Am. J. Trop. Med. Hyg. 65:958 965. 32. Martins-Filho, O. A., M. E. Pereira, J. F. Carvalho, J. R. Cancado, and Z. Brener. 1995. Flow cytometry, a new approach to detect anti-live trypomas-

VOL. 11, 2004 ELISA AND FLOW CYTOMETRY DETECTION OF IgG TO T. CRUZI 319 tigote antibodies and monitor the efficacy of specific treatment in human Chagas disease. Clin. Diagn. Lab. Immunol. 2:569 573. 33. Meurs, K. M., M. A. Anthony, M. Slater, and M. W. Miller. 1998. Chronic Trypanosoma cruzi infection in dogs: 11 cases (1987 1996). J. Am. Vet. Med. Assoc. 213:497 500. 34. Montenegro, V. M., M. Jimenez, J. C. Dias, and R. Zeledon. 2002. Chagas disease in dogs from endemic areas of Costa Rica. Mem. Inst. Oswaldo Cruz 97:491 494. 35. Mott, K. E., E. A. Mota, I. Sherlock, R. Hoff, T. M. Muniz, T. S. Oliveira, and C. C. Draper. 1978. Trypanosoma cruzi infection in dogs and cats and household seroreactivity to T. cruzi in a rural community in northeast Brazil. Am. J. Trop. Med. Hyg. 27:1123 1127. 36. Navin, T. R., R. R. Roberto, D. D. Juranek, K. Limpakarnjanarat, E. W. Mortenson, J. R. Clover, R. E. Yescott, C. Taclindo, F. Steurer, and D. Allain. 2002. Human and sylvatic Trypanosoma cruzi infection in California. Am. J. Public Health 75:366 369. 37. Oelemann, W. M., B. O. Vanderborght, G. C. Verissimo Da Costa, M. G. Teixeira, J. Borges-Pereira, J. A. De Castro, J. R. Coura, E. Stoops, F. Hulstaert, M. Zrein, and J. M. Peralta. 1999. A recombinant peptide antigen line immunoassay optimized for the confirmation of Chagas disease. Transfusion 39:711 717. 38. Partel, C. D., and C. L. Rossi. 1998. A rapid, quantitative enzyme-linked immunosorbent assay (ELISA) for the immunodiagnosis of Chagas disease. Immunol. Investig. 27:96. 39. Saldana, A., and O. E. Sousa. 1996. Trypanosoma rangeli: epimastigote immunogenicity and cross-reaction with Trypanosoma cruzi. J. Parasitol. 82: 363 366. 40. Schiffler, R. J., G. P. Mansur, T. R. Navin, and K. Limpakarnjanarat. 1984. Indigenous Chagas disease (American trypanosomiasis) in California. JAMA 251:2983 2984. 41. Sullivan, T. D., T. McGregor, R. B. Eads, and D. J. Davis. 1949. Incidence of Trypanosoma cruzi, Chagas, in Triatoma (Hemiptera, Reduviidae) in Texas. Am. J. Trop. Med. Hyg. 29:453 458. 42. Tibayrenc, M., P. Ward, A. Moya, F. J. and Ayala. 1986. Natural populations of Trypanosoma cruzi, the agent of Chagas disease, have a complex multiclonal structure. Proc. Natl. Acad. Sci. USA 83:115 119. 43. Tippit, T. S. 1978. Canine trypanosomiasis (Chagas disease). Southwest. Vet. 31:97 104. 44. Tomlinson, M. J., W. L. Chapman, Jr., W. L. Hanson, and H. S. Gosser. 1981. Occurrence of antibody to Trypanosoma cruzi in dogs in the southeastern United States. Am. J. Vet. Res. 42:1444 1446. 45. Umezawa, E. S., M. S. Nascimento, and A. M. Stolf. 2001. Enzyme-linked immunosorbent assay with Trypanosoma cruzi excreted-secreted antigens (TESA-ELISA) for serodiagnosis of acute and chronic Chagas disease. Diagn. Microbiol. Infect. Dis. 39:169 176. 46. Umezawa, E. S., and J. F. Silveira. 1999. Serological diagnosis of Chagas disease with purified and defined Trypanosoma cruzi antigens. Mem. Inst. Oswaldo Cruz 94:285 288. 47. Williams, G. D., L. G. Adams, R. G. Yaeger, R. K. McGrath, W. K. Read, and W. R. Bilderback. 1977. Naturally occurring trypanosomiasis (Chagas disease) in dogs. J. Am. Vet. Med. Assoc. 171:171 177. 48. Wood, S. F. 1975. Trypanosoma cruzi: new foci of enzootic Chagas disease in California. Exp. Parasitol. 38(2):153 160. 49. Woody, N. C., and H. B. Woody. 1955. American trypanosomiasis (Chagas disease). First indigenous case in the United States. JAMA 159:676 677. 50. Yabsley, M. J., and G. P. Noblet. 2002. Seroprevalence of Trypanosoma cruzi in raccoons from South Carolina and Georgia. J. Wildl. Dis. 38(1):75 83. 51. Yaeger, R. 1961. The present status of Chagas disease in the United States. Bull. Tulane Univ. Med. Fac. 21:9 13. 52. Zeledon, R., G. Solano, L. Burstin, and J. C. Swartzwelder. 1975. Epidemiological pattern of Chagas disease in an endemic area of Costa Rica. Am. J. Trop. Med. Hyg. 24:214 225.