Veterinary hnmunology and Immunopathology, 29 ( 1991 ) 267-283

267

Elsevier Science Publishers B.V., Amsterdam

Antibody and lyrnphoblastogenic responses of dogs experimentally infected with Trypanosoma cruzi isolates from North American mammals Stephen C. Barr a'', Vida A. Dennis b'2, Thomas R. Klei a'b and Neil L. Norcross c aDepartment of Veterinary Microbiology and Parasitology and bDepartment of Veterinary Science, Louisiana State University, Baton Rouge, LA 70803, U.S.A. CDepartment of Clinical Sciences, College of Veterinary Medicine, New York State College of Veterinary Medicine, Corneli University, Ithaca, NY 148.~.i, LT.4 (Accepted 6 November 1990) ABSTRACT Ban', S.C., Dennis, V.A., Klei, T.R. and Norcross, N.L., 1991. Antibody and lymphoblastogenic responses of dogs experimentally infected with Trypanosoma cruzi isolates from North American mammals. Vet. Immunol. Immunopathol., 29: 267-283. The humoral and cellular immune responses of dogs infected with either a non-pathogenic Trypanosoma cruzi isolate from a North American dog (Tc-D) or a pathogenic T. cruzi isolate from an opossum (Tc-O) were studied over a 240 day period. Antibody to T. cruzi epimastigote antigens prepared from Tc-O or Tc-D isolates were first detected by ELISA by Day 26 post infection (PI), peaked by day 175 PI and remained elevated throughout the experimental period in both Tc-O and Tc-D infected dogs. Differences in antibody levels between infected groups were not detected. Western blot analyses were performed using Tc-O and Tc-D epimastigote antigens probed with pooled sera and sera from individual Tc-O and Tc-D infected dogs prior to infection (Day 0), and during the acule ( Day 16-35 PI ), indeterminate (Day 50- ! 35 P! ) and chronic (Day 235 P! ) stages of infection. Generally, the patterns, number of protein bands, and temporal appearance of the protein bands identified by pooled sera and sera from individual dogs within each antigen preparation were similar. However, similarities and differences were present in antibody responses between sera from Tc-O and Tc-D infected dogs. Blastogenic responses of peripheral blood mer:onuclear cells (PBMC) from TcO and Tc-D infected dogs to mitogens (concanavalin A, phytohemagglutinin and pokeweed) were not significantly different from controls at any time during the experimental period. The PBMC from both groups of dogs were unresponsive to epimastigote antigens during the acute stage of infection. Statistically significant differences (P< 0.05) in PBMC responsiveness from controls were observed on Days 70 and 175 PI. Responses decreased to pre-infection levels by Day 240 PI. These studies demonstrate that although two North American T. cruzi isolates have markedly different virulence for dogs, some aspects of their cellular and humoral immune responses are similar while other responses, such as antibody recognition of specific T. cruzi antigens, vary. ABBREVIATIONS aEU, adjusted Eu values; Con A, concanavalin A; FCS, fetal calf serum; LIT lives infusion tryptose; MEM, eagle's minimum essential medium; PBMC, peripheral blood mononuclear cells; PBS, phos~Present address: Department of Clinical Sciences, College of Veterinary Medicine, New York State College of Veterinary Medicine, Cornell University, lthaca, NY 14853, USA. 2present address: Delta Regional Primate Research Center, Tulane University, Covington, LA 70433, USA. 0165-2427/91/$03,50 © 1991 Elsevier Science Publishers B.V. All rights reserved.

268

S.C. BARRET AL.

phate buffered saline; PHA, phytohemagglutinin; PI, post infection; PWM, pokeweed mitogen; RT, room temperature; SI, situation index.

INTRODUCTION

Trypanosoma cruzi, the etiological agent of Chagas' disease, is endemic in many parts of South America. In some infected individuals, infection is initially characterized by an acute myocarditis, followed by a prolonged latent period. Some patients then develop a chronic and often fatal cardiomyopathy (Hudson, 1981 ). The pathophysiology of the cardiomyopathy is essentially unknown although there is evidence that cellular and humoral immune responses play an important role in its development (Teixeira, 1987). A major stumbling block in the research efforts to elucidate the pathogenesis of chronic Chagas' disease is the lack of a suitable animal model (Hudson, 1981 ). Inbred mice, monkeys, rabbits and dogs have all been used as models for human chronic Chagas' disease, but no one system has proved sufficiently reproducible to gain general acceptance (Hudson, 1981 ). The canine model has been shown to mimic some of the cardiac, pathologic and electrocardiographic changes seen in human Chagasic patients (Anselmi et al., 1967, 1971; Andrade et al., 1981, 1984). In contrast to other animal models, little is known of the immune response of dogs to T. cruzi infection. Anti-T. cruzi antibodies have been demonstrated in experimentally infected and naturally infected dogs but serial studies on humoral or cellular responses have not been reported (Goble, I""": : . . . . .i~ / i; Barr e~ al., i986 ). ~a,~; wmmIns e[ al., The clinical, cardiographic and pathologic changes seen in dogs infected with T. cruzi isolates from an opossum (Tc-O) and a dog (Tc-D) are described elsewhere (Barr, 1989; Barr et al., 1991c). In this paper we describe some serial humoral and cellular immune responses over an 8 month period in dogs infected with the two isolates of T. cruzi. Using Western blot analysis, we demonstrate the epimastigote antigens that are identified by sera from dogs infected with the two isolates at different periods during infection. MATERIALS AND METHODS

Animals and experimental infections All dogs were vaccinated against Bordetella bronchiseptica, parainfluenza, parvovirus, distemper virus, hepatitis virus, rabies and leptospirosis using standard vaccination protocols. Puppies were weaned at 4 weeks of age. They

ANTIBODY RESPONSES OF TR YPANOSOMA CRUZ! INFECTED DOGS

269

were dewormed regularly, had three consecutively negative fecal examinations, and tested negative to T. cruzi antibodies prior to entering the study. Animals were housed indoors in individual cages in an AAALAC accredited facility for a 4 week acclimatization period and during the study. They were fed a commercial dry dog chow. Prior to initiation of the study, the experimental protocol was reviewed and approved by the University Committee on the Use of Humans and Animals as Research Subjects. Two groups each of six pure bred beagle dogs were used in the study. One group consisted of specific pathogen-free females (Hazleton Research Animals Inc., Cumberland, VA), which were 47 weeks old and weighed on average 6.75 kg (range 6-8 kg ) when infected. The second group (Louisiana State University Beagle Colony, Baton Rouge, LA) were from the same litter (five males, one female), 13 weeks old and weighed on average 3.28 kg (range 33.6 kg) when infected. Two dogs from each group were inoculated subcutaneously between the shoulder blades with either Tc-O, Tc-D, or 1.0 ml eagle minimum essential medium (MEM; Gibco Labs, Grand Island, New York) plus antibiotics (uninfected controls). A dose of 5 × 103 trypomastigotes per gram body weight up to a maximum dose of 2 × 107 organisms was given. Organisms of each isolate used in the dog inoculations were inoculated into Vero cell culture as an added check of viability. Parasites Trypanosoma cruzi isolates were obtained from an opossum (Tc-O) and a dog (Tc-D) from Louisiana as previously described (Barr et al. 1986, 1991a). Briefly, T. cruzi were initially isolated by the culture of blood from the hosts in liver infusion tryptose (LIT) media (Logan and Hanson, 1974), and maintained in African green monkey kidney (Vero) cell culture. The organisms were identified as T. cruzi based on morphology, in vitro characteristics (Barr et al., 1990b), and infectivity studies in inbred mice (Barr et al., 1990a). The media overlying Vero cell cultures and containing trypomastigotes released from the cells was harvested, washed three times in MEM supplemented with 10% heat inactivated ( 56 ° C, 30 min) fetal calf serum (FCS, Sigma Chemical Co., St. Louis, MO), 100 IU penicillin, and 100/lg streptomycin/ml of media (antibiotics). Trypomastigotes were counted and a viability check performed in a modified Neubauer hemocytometer and re-suspended to a concentration of 107/ml in MEM plus antibiotics. Antigens and mitogens Soluble parasite antigen was produced from T. cruzi epimastigotes of each isolate (Tc-O and Tc-D) grown in LIT media supplemented with 10% FCS as described previously (Logan and Hanson, 1974). Parasites were harvested during the log phase of growth, washed three times in 0.01 M phosphate buff-

270

S.C. BARR ET AL.

ered saline (PBS, 0.01M, pH 7.2), freeze/thawed four times (dry ice/acetone bath ) and sonicated with a Bronson Sonic Power Sonifier W 350 (Plainview, NY) set at 50% of maximum. Parasite debris was removed by centrifugation ( 100 000 × g for I h at 4 ° C). Following protein content determination (Bradford, 1976), the supernatant was filtered through a 0.2 gm filter (Acrodisc, Gelman Sciences, Ann Arbor, MI), aliquoted in 0.5 ml volumes, and stored at - 8 0 ° C until used. Concanavalin A (Con A), phytohemagglutinin (PHA), and pokeweed mitogens (PWM, Sigma Chemical Co., St. Louis, MO) were reconstituted with PBS, sterilized by filtration, aliquoted and stored at - 8 0 ° C until used. Antigens and mitogens were made up to optimum final culture concentration in MEM supplemented with 10% FCS and antibiotics (complete media) as described previously (Barr, 1989). ELISA assay

The assay was conducted as described previously (Barr et al., 1991b). Briefly, wells of flat-bottomed microtiter plates (Immulon 1R, Dynatech Laboratories, Alexandria, VI ) were coated with 2.5 ~g/ml of protein antigen (from Tc-O or Tc-D isolates) in 50/~l of carbonate-bicarbonate buffer (0.01 M, pH 9.6), and incubated overnight at 4 °C. For use, the plates were washed three times (3 rain per wash) in PBS, containing 0.05% Tween 20 (Sigma). After washing, 50/~1 of serum diluted to 1:64 in PBS-Tween 20 was added to each well and incubated for 2 h at room temperature (RT). All samples were tested in duplicate. After further washing, 50/~l of a 1" 1000 dilution of affinity purified, peroxidase conjugated goat anti-dog IgG (heavy and light chain specific; Kirkegaard and Perry Laboratories Inc., Gathersburg, MD) was added to each well and incubated for 2 h at RT. The plates were washed and reacted with 100/~1 of peroxidase substrate per well (nine parts of 0.8% 5-amino salicylic acid, pH 6, in one part 0.05% H202). After 30 rain the reaction was stopped with 25 ~1 of I M NaOH. Color development was measured with an MR 700 Microplate Reader (Dynatech) at a wavelength of 630 ~tm. Results are expressed as arbitrary ELISA units (EU) defined on the basis of a reference pool of sera from dogs experimentally infected with T. cruzi to which all other samples were compared. The same positive and negative pool control sera were used in each plate so that comparisons could be made between plates. ELISA unit values were considered positive if they were greater than three times the mean EU reading of six negative control samples (duplicate readings of sera from three uninfected control dogs). Using this cut off EU level, 95% of experimentally infected dogs become positive using this assay by Day 26 post infection (PI) (Barr et al., 1991b). To standardize readings between plates, all EU values for a plate were divided by three times the mean EU of negative control samples for that plate so that any EU value greater than one was positive. These values were called adjusted EU values (aEU). The opti-

ANTIBODY RESPONSES OF TRYPANOSOMA CRUZi INFECTED DOGS

271

mum antigen, serum and conjugate concentrations were determined by checkerboard titrations as described previously (Barr et al., 1991 b; VoUer et al., 1976).

SDS-PAGE SDS-PAGE of Tc-O and Tc-D antigens was performed using 7.5-20% gradient gels in a discontinuous buffer system as described previously (Laemmli, 1970). The stacking gel was 4% and the ratio of acrylamide to his acrylamide was 30:1. High and low molecular weight protein standards (Bio Rad Laboratories, Richmond, CA) run in each experiment were used for molecular weight determinations. Prior to electrophoresis antigen samples (350 /zg each) were diluted l : l with sample buffer (0.2M Tris-HCl pH 6.8, 4% SDS, 10% 2-mercaptoethanol, 10% glycerol and 0.1% bromophenol blue) and boiled at 95 °C for 4 min. All gels ( 1.0 mm) were run at a constant current of 16 mA per gel for I h for stacking gels, and 24 mA per gel for 4 h for resolving gels in a Protean II slab cell (BioRad Laboratories, Richmond, CA). All reagents were electrophoresis grade (BioRad).

Immunoblot Proteins separated by SDS-PAGE were electro-transfered (Transblot system, BioRad) for 18 h at 65 V to nitrocellulose sheets as described previously (Towbin et al., 1979). Following transfer, the nitrocellulose membrane was washed in distilled H20 and blocked in PBS containing 0.3% Tween-20 and 5% Carnation non-fat dry milk (blocking buffer) to saturate all unused protein binding sites. Strips (4 mm wide) were cut from the nitrocellulose memu,,,u~ a,,u ,~t~ut,atcu ~o~ l h at ~ °C with pooled sera obtained from either Tc-O or Tc-D infected dogs at Day 35, 135 and 235 PI, or individual animals at Days 0, 5, 16, 35, 50, 84, 102, 135 and 235 PI. Sera were diluted to 1-50 in PBS containing 1% non-fat dry milk. The strips were then washed twice for l 0 min each in PBS Tween-20, then once in PBS for 10 min, and incubated for 1 h at 37 ° C with a l- 1000 dilution of affinity purified, peroxidase conjugated goat anti-dog IgG (Kirkegaard and Perry Laboratories ) in PBS containing 1% non-fat dry milk. The strips were again washed and reacted at RT with 0.3% 4-chloro-l-naphthol (BioRad) in 20 ml of ice cold methanol and 100 ml of PBS containing 0.018% H202. The reaction was stopped after it reached the desired visual end point by rinsing the strips in distilled H20 and air drying.

Lymphocyte blastogenesis Prior to and at selected times throughout the experiment, 3 ml of blood was drawn from the jugular vein of dogs into heparinized tubes and centrifuged (450 × g for 15 rain) to obtain buffy coat cells which were resuspended in MEM. Peripheral blood mononuclear cells (PBMC) were separated from other blood components and trypomastigotes using a discontinuous ficoll

__7.

S.C. BARR ET AL.

(Sigma) gradient as described previously (DeTitto et al., 1982 ). PBMC were washed three times by centrifugation (450 × g for 10 min at 4 °C) in MEM, and counted in a modified Neubauer hemacytometer. Viability of cells (usually greater than 95%) was determined by trypan blue exclusion. Cells were diluted to 2 × 106 viable cells/ml in complete media, and 100/~1 added to the wells of fiat-bottomed tissue culture plates (Costar R, Data Packaging Co., Cambridge, MA) with 100/~l of mitogen, or round-bottomed tissue culture plates (Linbro R, Flow Laboratories Inc., McLean, VA) with 100 ~1 of Tc-O antigen. All samples were tested in triplicate. Optimal concentrations of antigen (Tc-O, 1 ~g/ml) and mitogens (Con A, 5/~g/ml; PHA, 5 gg/ml; PWM, 8/~g/ml) that produced optimal stimulation of PBMC from all dogs were determined prior to infection. Cultures were incubated at 37°C in a humidified atmosphere of 5% CO2, 95% air for 3 days for mitogens and 5 days for antigens, then pulsed for 18 h with 1 #Ci of 3H-thymidine (New England Nuclear Corp., Boston MA). Cells were harvested and counts per minute (c.p.m.) measured by liquid scintillation spectroscopy. The results are expressed as stimulation index, (SI, c.p.m, of stimulated cultures/c.p.m, of unstimulated cultures) or net c.p.m.s (c.p.m. of stimulated culture-c.p.m, of un-stimulated culture).

Statistics Student's unpaired 2-tailed t-test was used to compare the mean differences in antibody titers and SI between normal and infected dogs. Findings were considered significant at P< 0.05. Bars on all graphs represent the standard error of the mean. A SI greater than 2.0 was considered significant based on the 95th oercentile of pre-infectian v a h m ~ n f 19 d n o ~ RESULTS

Clinical outcome of infection A marked difference in the clinical outcome between Tc-O and Tc-D infected dogs was observed and has been described in detail elsewhere (Barr et al., 1991c). Briefly, during the 237 day experimental period, clinical, clinical pathologic, cardiac is.enzymes, electrocardiographic and echocardiographic data were studied and found to be normal in all Tc-D infected dogs. Parasitemias, although subpatent, remained present in these dogs throughout the experimental period. This was contrary to data collected from the Tc-O infected dogs. Dogs in this group, irrespective of their age at infection, developed acute myocarditis 2-3 weeks PI, which was fatal in one dog 28 days PI. Parasitemias were detectable in all dogs 9 days PI, peaked 2-3 weeks PI when acute myocarditis developed, then became subpatent or negative by 36 days PI. Survivors of acute disease remained clinically normal for several months PI then developed electrocardiographic abnormalities and dilational my.car-

ANTIBODY RESPONSES OF TRYPANOSOM.,~ CRUZI INFECTED DOGS

273

ditis. At necropsy, Tc-D infected dogs had minor cardiac changes in comparison to the severe biventricular dilation, cardiac wall thinning, multifocal lymphohistiocytic cellular infiltrate and extensive fibrosis throughout the cardiac tissue in Tc-O infected dogs.

ELISA analysis Both Tc-O and Tc-D infected dogs had detectable antibody levels to T. cruzi as detected by ELISA by Day 26 PI (Fig. 1 ). Antibody levels rose gradually until Day 152 PI when there was a two fold increase. Antibody levels then showed a continued increase throughout the experimental period. Statistically significant differences were not observed in antibody levels to T. cruzi between Tc-D and Tc-O infected dogs, nor between dogs of different ages infected with the same T. cruzi isolate at any time during infection Western blot analysis Pooled sera from Tc-O (Fig. 2A) and Tc-D (Fig. 2B) infected dogs were selected for Western blot analysis at Days 35, 135 and 235 PI because these times tended not only to reflect the main phases of infection (acute, indeterminate and chronic) but also when changes in antibody titers occurred (Fig. 1 ). To assess possible variation in antigen recognition between sera from individual dogs that made up the pooled sera, sera from each individual dog was also reacted with antigen preparations derived from the isolate the dogs were infected with. Representative reactions of sera obtained on Days, 0, 5, 16, 35, 50, 84, 102, 135 and 235 PI from individual dogs infected with either Tc-O or Tc-D are shown in Figs. 3 and 4, respectively. Generally, all the reactions described for the pooled sera were consistently observed in the sera from individual dogs. However, because sera from more time points were analyzed throughout infection, trends in, and the time of recognition of, certain antigens could be better appreciated in blots from individual dogs. How120

T

100

.o

e~e

Te-O

0--0

Tc-D

/ 0

I/T

orU

"o

0

25

50

75

100

125

,

,

,

,

1,50 175 200 225

DAYS POSTINFECTION

Fig. 1. Mean + SEM serum antibody levels Io 7". cruzi versus days PI for Tc-O infected dogs and Tc-D infected dogs expressed as aEU (see Materials and methods).

274

S.C. BARR ET AL.

123

!

I ~3ol~

2 3

4

~3ol~

i

lo2l~

~o2~

8o,!,i

so~

s3 l~

30 ~' .............~,

18~I,

A

181,

B

Fig. 2. T~.panosoma cruzi epimastigote antigens of each isolate (2A, Lanes 1-4, Tc-O; 2B, Lanes !-4, Tc-D) were electrophoretically separated in 7.5-20% SDS-polyacrylamide gels, transferred to nitrocellulose paper, and probed with pooled sea from Tc-O infected dogs (2A, Lanes 1-3 ), Tc-D infected dogs (2B, Lanes 1-3 ), or uninfected control dogs (Lanes 4). The relative migration of molecular weight standards ( × 1000) for the antigen preparations are indicated to the left.

ever, some antigens recognized in blots of pooled sera were not observed in blots of individual dog sera. A total of 25 different protein bands with apparent molecular weights of 115 to 25 kDa were identified in Tc-O antigen preparations by pooled sera and sera from individual dogs infected with Tc-O. Most protein bands were present by Day 35 PI (acute stage on infection), and increased in intensity by Day 135 PI (indeterminate stage of infection) and Day 235 PI (chronic stage of infection) in blots of pooled sera (Fig. 2A). Specific protein bands

ANTIBODYRESPONSESOF TR YPANOSOMA CRUZi INFECTEDDOGS 1

2

3

4

5

~

7

8

~

275

10

v

130 I~

i

102~ !

L

1

so~

531~

.!

ii

I i i

a0~

18t~

I

Fig. 3. Tr.vpanosoma cruzi epimastigote antigens of Tc-O were electrophoretically separated in 7.5-20% SDS-polyacrylamide gels, transfered to nitrocellulose paper, and probed with sera collected preinfection (Lane 1 ), and on PI Days 5 (Lane 2), 16 (lane 3), 35 (Lane 4), 50 (Lane 5), 84 (Lane 6), 102 (Lane 7), 135 (Lane 8) and 235 (Lane 9) from an i n d i v i d u a l dog representative of those infected with Tc-O, and on Day 235 PI (Lane 10) from an uninfected control dog. The relative migration o f molecular weight standards ( × 1000) for the antigen preparations are indicated to the left.

of interest were: ( 1 ) a 80 kDa antigen recognized by all pooled sera from TcO infected dogs (Fig. 2A) and first recognized by sera from individual Tc-O infected dogs at Day 35 PI (Fig. 3), and (2) a 26 kDa antigen recognized very strongly by individual dog sera during the acute and indeterminate stages of infection but not during the chronic stage of infection (Fig. 3 ). A total of l0 different protein bands with apparent molecular weights of 125 to 25 kDa were identified in Tc-D antigen preparations by pooled sera

276

S.C. BARR ET AL.

1

2

3

4

5

~

T

8

?

10

z3o )'

z02~

soF

53~,

30~

Fig. 4. T~.panosoma cruzi epimastigote antigens of Tc-D were electrophoretically separated in 7.5-20% SDS-polyacrylamide gels, transfered to nitrocellulose paper, and probed with sera collected preinfection (Lane 1 ), and on PI Days 5 (Lane 2), 16 (Lane 3), 35 (Lane 4), 50 (Lane 5), 84 (Lane 6), 102 (Lane 7), 135 (Lane 8) and 235 (Lane 9) from an individual dog representative of those infected with Tc-D, and on Day 235 PI (Lane 10) from an uninfected control dog. The relative migration of molecular weight standards ( × 1000) for the antigen preparations are indicated to the left.

and sera from individual dogs infected with Tc-D (Figs. 2B and 4, respectively). Five immunodominant antigens (90, 88, 83, 80 and 26 kDa) were recognized by all individual dog sera during the acute, indeterminate and chronic stages of infection (Fig. 4). Also of interest were: ( l ) a 50 kDa antigen recognized very st, ongly by individual dog sera during the acute stage of infection but less so daring the indeterminate and chronic stages of infection; and (2) a 124 and 115 kDa antigen recognized first on Days 135 and 85 PI

ANTIBODY RESPONSES OF TR YPANOSOMA CRUZi INFECTED DOGS

2 77

(indeterminate state) by the pooled sera and individual dog sera, respectively. The intensity of staining of these protein bands by both pooled and individual dog sera increased as the infection became more chronic. Some specific similarities and differences were detected in Western blot analysis between sera from Tc-O and Tc-D infected dogs. Overall, sera from Tc-D infected dogs recognized antigens more intensely than sera from Tc-O infected dogs. The 80 and 26 kDa antigens were recognized in both antigen preparations by both pooled and individual dog sera. However, by comparing antigen recognition by sera from individual dogs, Tc-D infected dogs consistently recognized the 80 kDa antigen on Day 16 PI (Fig. 4), while Tc-O infected dogs recognized it on Day 35 PI (Fig. 3 ). Although both Tc-O and TcD infected dogs recognized the 26 kDa antigen by Day 35 PI, Tc-D infected dogs recognized it more strongly throughout the course of infection. Differences were also observed in Tc-D and Tc-O infected dog sera recognition of 90, 88, 83 and 50 kDa antigens. These antigens were intensely stained in the Tc-D antigen preparation and only faintly stained in the Tc-O antigen preparation. Also, these antigens were observed earlier in infection in the Tc-D antigen preparation. Antigens with apparent molecular weights of I 15 and 125 kDa were observed in the Tc-D preparation but not in the Tc-O antigen preparation. Differences also occurred between antigen recognition by pooled sera and individual dog sera. A 7 kDa antigen was only recognized in the TcO antigen preparations by pooled sera from Tc-O infected dogs on Days 135 and 235 PI (data not shown), but not by sera from individual Tc-O infected dogs. PB MC biastogenic responses to mitogens and antigens The mean c.p.m. +_SD for unstimulated PBMC for all dogs throughout the experiment was 458.52+_698.87 with a range of 4626.93 to 57.67 c.p.m. PBMC from Tc-O and Tc-D infected dogs showed similar in vitro responses to mitogens. At no time were these responses significantly different from control values during the infection. Responses to Con A were higher than those to PHA which were higher than responses to PWM (Table 1 ). Although responses to mitogens as measured by net c.p.m, showed fluctuations from one sample day to another, there was not significant difference in responses between groups of dogs within each mitogen (data not shown). At no time did infected dogs show a suppression of responses to mitogens. PBMC proliferative responses to T. cruzi antigens were not detected in controis or infected dogs prior to infection. PBMC proliferative responses in infected dogs to T. cruzi antigens were significantly different from uninfected control dogs (P 2) at Days 70 and 175 PI (Fig. 5). At only one time during the experimental period did two control dogs show responses to T. cruzi antigens similar to infected dogs (SI > 2 ), this being on Day 91 PI.

278

S.C. BARR ET AL.

TABLE 1 Mean net c.p.m. _+SD as a measure of in vitro proliferative response of PBMCs from Tc-O infected, Tc-D infected, and control dogs stimulated with Con A, PHA and PWM throughout the experimental period Mitogen

Source of PBMCs

Con A PHA PWM

i!

Tc-O

Tc-D

Comrois

45.909 + 39.037 37.916 + 40.004 14.207 + 11.526

56.260 + 53.236 42.075 + 38.330 19.424 + 19.86 !

69.053 + 75.695 54.530 + 57.808 20.329 + 24.310

O - - O Tc-O e--e Tc-O A - - A Cont.~l

/if

;;

~!

O %1"I ' :"-¢'~ I I I -25 0 25 50 75 100 125 150 17~ 2C5 z25 DAYSPOST1NFECTION Fig. 5. Mean_+SEM PBMC blastogenic responses to 7". cruzi epimastigote antigen (1 p g / m l ) versus days PI for Tc-O infected, Tc-D infected and control dogs. *P< 0.05 from control dog values° DISCUSSION

This is the first study that details the longitudinal cellular and ht~mora! immune responses of dogs infected with pathogenic and non-pathogenic isolates of T cruzi to this parasite. Dogs infected with either non-pathogenic or pathogenic T cruzi isolates produced elevated antibody levels against T. cruzi by 4 weeks PI. This is a period of decreasing parasitemias (Barr et al., 1991c). Antibody levels persisted for many months and in one dog infected with the non-pathogenic T cruzi isolate (Tc-D), high antibody levels persisted for at least 490 days. This is consistent with findings in mice in which antibody titers persist for long periods (Araujo et al., 1984). Persistent antibody titers have also been reported in man and monkeys (Hanson, 1976; Szarfman et al., 1981 ). Dogs infected with either the pathogenic or non-pathogenic isolate had similar antibody levels throughout the experiment. These observations differ from those in human patiems. In one such study, patients infected with T.

ANTIBODY RESPONSES OF TRYPANOSOMA CP, UZI INFECTED DOGS

279

cruzi developed anti-T, cruzi antibody titers, but those patients who had car-

diomyopathy had significantly higher titers than those with no evidence of heart disease (Mosca et al., 1985 ). There is a considerable body of evidence which favors the role of antibody in the pathogenesis of Chagasic cardiomyopathy (Teixeira, 1987 ). Our data show that dogs with a T. cruzi induced chronic dilational cardiomyopathy develop antibody levels similar to those in T. cruzi infected dogs without heart disease. This suggests that more specific studies are required to elucidate the role of antibody, if any, in the generation of heart disease. The dog model discussed could provide considerable advantages in such studies. As with previously described immunoblots of T. cruzi antigens using sera from infected animals, the general pattern shown in our data is complex (Nogueira et al., 1981; Zingales et al., 1984; Martins et al., 1985; Grogl and Kuhn, 1985 ). Previous studies have shown that both human Chagasic, rabbit anti-trypomastigote (Y strain) and mouse anti-trypomastigote (Y and Brazil strains) sera consistently recognize a 90 to 95 kDa and 80 to 75 kDa protein bands common to epimastigote forms (Nogueira et al., 1981; Zingales et al., 1982, 1984, Grogl et al.; 1985; Martins et al., 1985 ). In our study, sera from dogs during the acute, indeterminate and chronic stages of the disease (Tc-O infected dogs) also recognize these two protein bands. The 90 kDa antigen may correspond to the 90 kDa glycoprotein described by Snary (1980) and the 95 kDa protein recognized by human Chagasic sera (Snary, 1980; Zingales et al., 1984). This protein band is common to epimastigotes derived from many different clones or strains from different geographic areas (Snary, 1980; Zingales et al., 1984). The 80 kDa protein identified by all sera in our study may correspond to the 80 to 75 kDa giycoprotein fourld in a number of clones and stocks (Zingales et al., 1984). Zingales (1984) found that this antigen was strongly recognized by all Chagasic sera tested, whereas the 95 kDa glycoprotein was less noticeably recognized by sera of lower indirect immunofluorescence titers. This may explain why some authors are unable to identify the 95 kDa antigen in their studies of the epimastigote state (Nogueira et al., 1981 ). Certainly, in our studies, a 90 kDa protein was difficult to detect in Tc-O antigen preparations blotted with either pooled sera or sera from individual dogs, and represents one of the main differences in patterns of protein band recognition between sera from non-pathogenic infections (Tc-D infected dogs) and pathogenic infection (Tc-O infected dogs). These studies suggest that T. cruzi isolates from an opossum and naturally infected dog from North America contain the major antigens found in pathogenic South American isolates. Furthermore, although a cause and effect relationship remains to be demonstrated, it is possible that the differences shown in antigen recognition between sera from the Tc-O and Tc-D infected dogs may be important in dictating the outcome of infection. Also, the temporal relationship

280

S.C. BARR ET AL.

between the stages of disease and the appearance and disappearance of certain protein bands (e.g. 26 kDa) in blots of Tc-O antigen preparations throughout the infection may also indicate a cause and effect relationship. The in vitro proliferative capacities of PBMC from human patients suffering from cardiomyopathy, megaesophagus and megacolon caused by T. cruzi do not differ from the responses of PBMC from uninfected individuals in their response to mitogens (Tschudi et ai., i 972; Montufar et al., 1977; Gusmao et al., 1984). Our results show that the aoility of PBMC from both groups of T. cntzi infected dogs to respond to mitogens were also not significantly different from uninfected control dogs. The inability of PBMC from T. cruzi infected dogs to respond to epimastigote antigens during the first 40 days of infection (acute stage) coincided with the period when parasitemias were highest. At approximately Day 40 PI, parasitemias had dropped to non-quantifiable levels and PBMC responses to epimastigote antigens began to increase. This pattern of unresponsiveness to epimastigote antigens during the acute stage of T. cruzi infection has also been demonstrated in mice and is thought to be caused, in part, by a number of factors including suppressor macrophages, deficient T helper-cell activity, and reduction of interleukin production (Rowland and Kuhn, 1978; Ramos et al., 1979; Cunningham and Kuhn, 1980; Hayes and Kierszenbaum, 198 l, Reed et al., 1984; Tarleton and Kuhn, 1984; Choromanski and Kuhn, 1987). Mice recover their ability to respond to specific antigens soon after the disappearance of the parasitemia (Hays and Kierszenbaum, 1981 ). In man, interleukin-2 receptor expression has been shown to be inhibited by T. cruzi in vitro (Laemmli, 1970), and it has been shown that acute Chagasic patients may exhibit reduced cutaneous reactivity to T. cruzi antigens (Teixeira et a!., 1978). PBMC from human patients with Chagas' disease proliferate in response to I: cruzi amigel~s to the same degree irrespective of whether the stage of disease is indeterminate or chronic (Gusmao et al., 1984; Mosca et al., 1985 ), or whether the antigens used are from an isolate or clone (Morato et al., 1986). In this study, PBMC from infected dogs responded significantly above control values during the indeterminate and chronic periods. Additionally, as in human patients, there was no significant difference between infected dogs showing no disease or dogs with heart disease. These results support the hypothesis that PBMC blastogenic responses to crude 7'. cruzi antigens indicate exposure to the organism but are not a reliable indicator of a particular Chagasic disease state. There is a considerable body of evidence in studies in rabbits, mice and man, that cell-mediated immune responses play a major role in Chagasic cardiomyopathy (Teixeira, 1987). Although much of this evidence indicates the presence of an antigenic determinant common to both T. cruzi and host cardiac tissues which is also recognized by immune T-cells, further studies are required to clarify the role of cellular immunity in Cha-

ANTIBODY RESPONSES OF TRYPANOSOMA CRUZI INFECTED DOGS

281

gasic cardiomyopathy (Teixeira, 1987). This study illustrates the feasibility of using the Tc-O infected dog as a model for Chagasic cardiomyopathy with Tc-D infected dogs acting as positive controls for future cellular immunological and biochemical studies. The results presented here document the longitudinal humoral and cellular immune responses to pathogenic and non-pathogenic isolates of T. cruzi in the dog. Results suggest there are marked similarities between immune responses in the dog and those in infected humans. These data further expand the potential role of the canine model for human Chagas' disease.

REFERENCES Andrade, A.Z., Andrade, S.G. and Sadigursky, M., 198 !. Experimental Chagas' disease in dogs. Arch. Pathol. Lab. Med., 105: 460-464. Andrade, Z.A., Andrade, S.G. and Sadigursky, M., 1984. Damage and healing in the conducting tissue of the heart (an experimental study in dogs infected with Trypanosoma cruzi). J. Parasitol., 143: 93-101. Anselml, A., Gurdiel. O., Suarez, J.A. and Anselmi, G., 1967. Disturbances in the A-V conduction system in Chagas' myocarditis in the dog. Circ. Res., 20: 56-64. Anselmi, A., Moleiro, F., Suarez, R., Suarez, J.A and Ruesta, V., 1971. Ventricular aneuD'sms in acute experimental Chagas' myocardiopathy. Chest, 59" 654-658. Araujo, F.G., Heilman, B. and Tighe, L., 1984. Antigens of Trypanosoma cru=i detected by different classes and subclasses of antibodies. Trans. R. Soc. Trop. Med. Hyg., 7g: 672-677. Barr, S.C., 1989. Studies on Trypanosoma cruzi isolates from Louisiana mammals with particular reference to infectivity Io dogs. PhD dissertation, Louisiana State University, Baton Rouge, LA. Barr, S.C., Baker, D. and Markovits, J., 1986. Trypanos~,,aiasis and laryngeal paralysis in a dog. J. Am. Vet. Med. Assoc., 188: 1307-1309. Barr, S.C., Brown, C.C., Dennis, V.A. and Klei, T.R., 1990a. Infections of inbred mice with three Trypanosoma cruzi isolates from Louisiana mammals. J. Parasitol., 76:9 i 8-921. Barr, S.C., Dennis, V.A. and Klei, T.R., 1990b. Growth parameters in axenic and cell cultures, protein profiles, and zymodeme typing of three Trypanosoma cruzi isolates from Louisiana mammals. J. Parasitol., 76:631-638. Barr, S.C., Brown, C.C., Dennis, V.A. and Klei, T.R., 1991a. The lesions and prevalence of Trvpanosoma cruzi in opossums and armadillos from southern Louisiana. J. Parasitol., in press. Barr, S.C., Dennis, V.A. and Klei, T.R., 1991b. Serological and blood culture survey of four dog populations for Trypanosoma cruzi from south Louisiana. Am. J. Vet. Res.. 52: 570-573. Barr, S.C., Gossett, K.A. and Klei, T.R., !991c. Clinical, clinical pathological, and parasitological observations of trypanosomiasis in dogs infected with North American Trypanosoma cruzi isolates. Am. J. Vet. Res., 52: 954-960. Bradford, M.M., ! 976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem., 72: 248-254. Choromanski, L. and Kuhn, R.E., 1987. Use of parasite antigens and intedeukin-2 to enhance suppressed immune responses during Trypanosoma cruzi infection mice. Infect. lmmun., 55: 403-408. Cunningham, D.S. and Kuhn, R.E., 1980. Trypanosoma cruzi-Jnduced suppression of the pri-

282

S.C. BARR ET AL.

mar) immune responses in murine cell cultures to T-cell-dependent and -independent antigens. J. Parasitol., 66:16-27. DeTitto, E., Braun, M. and Segura, E.L., 1982. Density gradient purification of human lymphocytes from contaminating trypomastigotes of T~.panosoma cruzi. J. Immunol. Met., 50:281287. Goble, EC., 1952. Observations on experimental Chagas' disease in dogs. Am. J. Trop. Med. Hyg., l" 189-204. Grogi, M and Kuhn, R.E., 1985. Identification of antigens of Trypanosoma cruzi which induce antibodies during experimental Chagas' disease. J. Parasitol., 71: 183-19 I. Gusmao, R.D ~Rassi, A., Rezende, J.M. and Neva, F.A., Specific and non-specific lymphocyte blastogenic responses m individuals infected with Trypanosoma cruzi. Am. J. Trop. Med. Hyg., 33: 827-834. Hanson, W.L., 1976. Immunology of American Trypanosomiasis. In: S. Cohen and E.H. Sadum (Editors), Immunology of Parasitic Infections. Blackwell Scientific Publications, Oxford, pp. 222-234. Hayes, M.M. and Kierszenbaum, F., 198 i. Experimental Chagas' disease: Kinetics of lymphocyte responses and immunological control of the transition from acute to chronic Trypanosoma cru=i infection. Infect. Immun., 31: 1117-1124. Hudson, L., 1981. Immunobiology of Trypanosoma cruzi infection and Chagas' disease. Trans. R. Soc. Trop. Med. Hyg., 75: 493-498. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head bacteriophage T4. Nature, 277: 680-685. Logan, L. and Hanson, W.L., 1974. Trypanosoma cruzi: Morphogenesis in diffusion chambers in the mouse peritoneal cavity and attempted stimulation of host immunity. Exp. Parasitol., 36: 439-454. Martins, M.S., Hudson, L., Krettli, A.U., Cancado, J.R. and Brener, Z., 1985. Human and mouse sera recognize the same polypeptide associated with immunological resistance to Trypanosoma cruzi infection. Clin. Exp. Immunol., 61: 343-350. Montufar, O.M.B., Musatti, C.C.,. Mendes, E. and Mendes, N.F., 1977. Cellular immunity in chronic Chagas' disease. J. Clin. Microbiol., 5:401-404. Morato, M.J.F., Breaer, Z., Cancado, J.R., Nunes, R.M.Bo~ E~er, C. and Gazzine!!i, G., 1986. Cellular immune responses of Chagasic patients to antigens derived from different Trypanosoma cru:i strains and clones. Am. J. Trop. Med. Hyg., 35:505-511. Mosca, W., Plaja, J., Hubsch, R. and Cedillos, R., 1985. Longitudinal study of immune response in human Chagas' disease. J. Clin. Microbiol., 22: 438-441. Nogueira, N., Chaplan, S., Tydings, J.D., Unkeless, J. and Cohn, Z., 1981. Trypanosoma cruzi: Surface antigens of blood and culture forms. J. Exp. Med., 153: 629-639. Ramos, C.I., Schadtler, S. and Ortiz-Ortiz, L., 1979. Suppressor cell present in the spleen of Trvpanosoma cruzi-infected mice. J. Immunol., 122:1243-1247. Reed, S.G., lnverso, J.A. and Roters, S.B., 1984. Heterologous antibody responses in mice with chronic Trypanosoma cruzi infection: Depressed T helper function restored with supernarants containing interleukin-2. J. Immuno!., 133:1558-1563. Rowland, E.C and Kuhn, R.E., 1978. Suppression of cellular responses in mice during Trypap!osoma cruzi infection. Infect. Immun., 20: 393-397. Snary, D., 1980. Trypanosoma cruzi: antigen invariance of cell surface glycoprotein. Exp. Parasitol., 49: 68-77. Szarfman, A., Gerecht, D., Draper, C.C. and Marsden, P.D., 1981. Tissue-reacting immunoglobulins in rhesus monkeys infected with Trypanosoma cruzi: A follow-up study. Trans. R. Soc. Trop. Mcd. Hyg~ 75:114-116. Tarleton, R.L. and Kuhn, R.E., 1984. Restoration cf in vitro immune responses of spleen cells

ANTIBODY RESPONSES OF TR YP,4NOSOMA CRUZI INFECTED DOGS

283

from mice infected with Trypanosoma cruzi by supernatants containing interleukin-2. J. Immunol., 133: 1570-1575. Teixeira, A.R.L., 1987. The stercorarian trypanosomes. In: E.J.L. Soulsby (Editor), |nmmnc Responses in Parasitic Infection: Immunology, Immunopathology, and Immunoprophylaxis. CRC Press, Boca Raton, FL, pp. 25-! ! 8. Teixeira, A.R.L., Teixeira, G., Macedo, V. and Prata, A., 1978. Acquired cell-mediated immunodepression in acute Chagas' disease. J. Clin. Invest., 62:1132-1141. Towbin, H., Staehelin, T. and Gordon, J., 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets. Proc. Natl., Acad. Sci., U.S.A., 76: 4350-4354. Tschudi, E.I., Anziano, D.F. and Dalmasso, A.P., 1972. Lymphocyte transformation in Chagas" disease. Infect. Immun., 6: 905-908. Voller, A., Bidwell, D. and Bartlett, A., 1976. Microplate enzyme immunoassays for the immunodiagnosis of viral infections. In: N.R. Rose and H. Friedman (Editors), Manual of Clinical Immunology. American Society of Microbiology, Washington, pp. 506-512. Williams, G.C., Adams, L.G., Yaeger, R.G., McGrath. R.K., R:~ad, W.K. and Biiderback, W.R., 1977. Naturally occurring Trypuno~oma cruzi (Chagas' dLease) in dogz. !. Am. Vet. Med. Assoc., 171: 171-177. Zingales, B., Andrews, N.W., Kuwajima, V.Y. and Colli, W., 1982. Cell surface antigens to T~,panosoma cruzi; possible correlation with interiorization in mammalian cells. Mol. Biochem. Parasitol., 6: I I l-124. Zingales, B., Aubin, G., Romanha, A.J., Chiari, E. and Colli, W., 1984. Surface antigens of stocks and clones of Trypanosoma cruzi isolated from humans. Acta Trop., 41: 5-16.