EXPERIMENTAL
74, 46-56 (1992)
PARASITOLOGY
Trypanosoma cruzi: Response
Predominance of IgG2a in Nonspecific during Experimental Chagas’ Disease
SYLVIANE SPINELLA,PASCALE LIEGEARD,AND Unit6 d’lmmunoparasitologie,
Humoral
MIREILLE HONTEBEYRIE-JOSKOWICZ'
Institut Pasteur, 25, Rue du Dr Roux, 75724 Paris Cedex 15, France
SPINELLA, S., LIEGEARD, P., AND HONTEBEYRIE-JOSKOWICZ, M. 1992. Trypanosoma cruzi: Predominance of IgG2a in nonspecific humoral response during experimental Chagas’ disease. Experimental Parasitology, 74, 4656. The kinetic and isotypic pattern of hypergammaglobulinemia has been investigated in C3H/HeJ infected with Trypanosoma cruzi. Hypergammaglobulinemia appeared 14 days postinfection, increased until Day 28 postinfection, and persisted throughout the chronic phase (>60 days of infection). The main isotype secreted was IgGZa, reaching IO-fold the control level. High titers of autoantibodies were found of IgM and IgG subclasses. Isotypic characterization of antibodies against myosin, myelin, and keratin, was performed and determined to be IgG2a subclass in the chronic stage of infection. Specific responses against T. cruzi took place 2 weeks postinfection when the parasitemia was high. Interestingly, parasite-specific response was maximal after 4 weeks of infection and plateaued during the chronic phase when parasites were rare. In contrast to the humoral polyclonal response in the chronic stage, showing a preferential IgC2a pattern, the anti-T. cruzi response consisted of all the different isotypes: IgM, IgGl, IgG3, IgGZa, and IgGZb, throughout the infection. Identical patterns of parasite antigens were recognized by IgG2a and IgG2b antibodies. Few different antigens were identified by the IgG3. Some antigens were recognized by several isotypes, others by only one isotype. With regard to the existence of antigenic cross-reactivities between host and parasite, we designed absorption experiments on parasite-specific immunoadsorbent showing that specific antibodies eluted from the column failed to recognize the natural antigens. These studies suggest that nonspecific and antiparasite-specific responses may be maintained by different regulatory pathways. o 1992 Academic PITHS, I”~. INDEX DESCRIPTORS AND ABBREVIATIONS: Trypanosoma cruzi, protozoan, experimental Chagas’ disease; Specific and nonspecific humoral response; Autoantibodies; Isotypic pattern; Immunoglobulin (Ig); Keratin (Ker); Myoglobin (Myo): Myosin (Myos); Phosphatebuffered saline (PBS); peroxydase (PO); thyroglobulin (Thyr); Trinitrophenyl (TNP); Tubulin (Tub).
that, at least in part, the immunosuppression observed during the acute phase and the autoimmunity associated with T. cruzi chronic infection (Plata 1985; Reed 1984; Hudson 1985) could be due to the intense polyclonal activation of the early stage of infection (Minoprio 1989). Indeed, an association between polyclonal B cell activation and autoimmune syndrome has been described in systemic lupus erythematous experimental models (Klinman and Steinberg 1987). In turn, the existence of immunological and molecular cross-reactivities between the parasite and its mammalian host-the so-called molecular mimicry-
Mouse infection with the protozoan parasite Trypanosoma cruzi (T. cruzi) sets off polyclonal B cell activation which appears early in the infection and persists during the life-long chronic phase (Ortiz-Ortiz et al. 1984; D’Imperio Lima et al. 1985, 1986). In the early stage of infection, the number of Ig secreting cells in the spleen and peripheral lymph nodes is very high and the majority of activated B cells secrete nonparasite-specific antibodies (Minoprio et al. 1988). The hypothesis has been proposed ’ To whom correspondence
should be addressed. 46
0014-4894/92 $3.00 Copyright0
1992 by Academic Press.Inc.
All rights of reproduction
in any form reserved.
ISOTYPE
DISTRIBUTION
IN EXPERIMENTAL
may account for some autoantibodies present in the sera of infected patients or animals and, possibly, relevant to the pathogenesis of chronic Chagas’ disease (Van Voorhis and Eisen 1989; Kahn et al. 1990; Levin et al. 1990; Mesri et al. 1990; Kerner et al. 1991). Whether the production of autoantibodies is the result of nonspecific polyclonal B cell activation or parasite-antigen-driven B cell stimulation remains to be elucidated. Polyclonal B cell responses in T. cruziinfected mice have been clearly demonstrated in lymphoid organs. In contrast, the isotype profiles of nonspecific humoral responses have been poorly investigated in experimental Chagas’ disease. Nevertheless, the isotype profile of T. cruzi-specific humoral responses has been well documented in mouse model and in humans (Krettli and Brener 1976, 1982; Scott and Goss-Sampson 1984; Araujo et al. 1984; Takehara 1989). We were interested in the kinetic of nonspecific immunoglobulins in the sera of infected animals and its comparison to that of the total parasite-specific responses. In the present report, we show (a) that hypergammaglobulinemia appeared 2 weeks after infection; (b) that the predominant isotype secreted during the chronic stage of infection was the IgG2a, including the natural autoantibodies, as evidenced for those reacting against myelin, keratin, and myosin; (c) that IgGl, IgG2a, and IgG2b were represented in the specific responses and recognized different parasite antigens; and (d) that the parasite-specific antibodies did not bind to natural antigens. Those data suggest that specific and nonspecific humoral responses may be triggered by different mechanisms during experimental Chagas’ disease. MATERIALS
AND
METHODS
Mice and infection. C3H/HeJ mice were bred at the Pasteur Institute. Eight- to lo-week-old mice were infected by ip injection of lo4 freshly collected blood-
CHAGAS’
DISEASE
47
stream trypomastigotes of the CL strain of T. cruzi. Mice were sacrificed on Days 14, 21, 28, 35, 60, 90, 120, and 180 postinfection. Sex- and age-matched uninfected mice were used as normal controls. Parasitemia was determined at the tail vein (Brener 1962). Antigens and antibodies. T. cruzi extract was obtained as previously described (HontebeyrieJoskowicz et a/. 1987). Briefly, Vero cells grown in Dulbecco’s medium with 5% calf serum (Flow laboratories, Asnieres, France) were infected with T. cruzi trypomastigotes. Trypanosomes released in culture medium were extensively washed and adjusted to 10’ parasites/ml in serum-free RPMI, frozen and thawed three times, and then passed through a 0.45urn Millipore filter. Native double-stranded (ds)-DNA, (porcine type II) thyroglobulin and (whale skeletal muscle type II) myoglobin, myosin of rabbit muscle, and keratin from human epidermis were purchased from Sigma Co. (St. Louis, MO); actin (Spudish and Watt 1971)and tubulin from rabbit muscle (Shelansky et al. 1973) were prepared as previously described; horseradish peroxidase (PO) was purchased from Boehringer (Manheim, Germany); TNP-BSA (TNP) was prepared according to Little and Eisen, 1966;myelin from mouse central nervous system was prepared according to Norton and Poduslo 1973, and was a generous gift of Dr. Ben You&-Chennouli (La Salpetriere, Paris). Peroxidase labeled-goat anti-mouse (IgM, IgGl, IgG3, IgG2b, or IgG2a) and unlabeled goat anti-mouse Ig were purchased from Southern Biotechnology Associates, Inc., as well as unlabeled isotypes (IgM, IgGl , IgG3, IgG2b, and IgGZa) used for standard. Enzyme-linked immunoassay (ELISA) for T. cruzi and natural antigens. Flat-bottomed plates (CML,
France) were coated with different antigens overnight at 4°C: 50 PI/well of T. cruzi antigen (25 ug/ml in KH,PO,/K,HPO, buffer, 50 n&f pH 8) or 100 uhwell of different antigens diluted in carbonate buffer pH 8.5; actin (10 p.g/ml), TNP-BSA (1 &ml), thyroglohulin (10 l&ml), PO (10 ug/ml), myoglobin (10 p&ml), tubulin (10 t.q$ml), myelin (1 l&ml), myosin (5 l&ml), except ds-DNA which was diluted in citrate buffer (0.5 mg/ml) and keratin (10 kg/ml) which was diluted in Tris buffer 0.1 M pH 7.4, added with 8 M urea and 1% mercaptoethanol. Coated plates were washed thoroughly with phosphate-buffered saline (PBS) containing 0.1% Tween 20 (Merck, France) (PBS-T) and incubated for saturation with PBS 1% gelatin (200 p,l/well) for 1 hr at room temperature (RT). Then, they were incubated in presence of serial dilutions of each sera (initial dilution l/10) for 2 hr at RT. After extensive washing with PBS-T, the plates were incubated for 1 hr at RT with peroxidase-labeled goat anti-mouse isotypes (labeled anti-IgM, anti-IgG3, and anti-Ig were used at l/2000, anti-IgGl l/4000, anti-IgG2b and IgG2a l/8000) and revealed with appropriate peroxidase sub-
48
SPINELLA,
LIEGEARD,
AND
skate. Results were expressed as the percentage of absorbance of infected sera on absorbance of a pool of normal sera, when natural autoantibodies were tested. For T. cruzi-specific response, the titer was expressed as the reciprocal dilution of the 50% of the maximal OD after titration at serial twofold dilutions. ELISA for total immunoglobulins in the serum. Flatbottomed plates were coated with unlabeled goat antimouse Ig (5 ug/ml) in carbonate buffer, pH 8.5, overnight at 4°C. The same protocol as described above was used, and the initial dilution used was l/500. Unlabeled isotypes were used as standards, at serial dilutions beginning with 180 &ml. Western blot analysis. Analysis of T. cruzi polypeptides recognized by the different Ig isotypes of sera from chronically infected mice was performed by Western blotting (Coding and Handman 1984).Briefly, crude lysates of purified trypomastigotes (2 x lO’/ml) from culture of Vero cells were loaded on SDSpolyacrylamide gels. After migration, the polypeptides were transfered onto nitrocellulose according to Towbin et al. (1979). The nitrocellulose strips were then saturated and incubated with normal or chronic sera diluted l/100. The reaction was developed with peroxidase-labeled goat anti-mouse Ig isotypes antibodies and 4-chloro-a-naphtol as substrate (Bio-Rad, France). Affinity chromatography. T. cruzi antigens were coupled to CNBr-activated Sepharose 4B (Pharmacia, France) following the manufacturer instructions and used to affinity purify the parasite-specific antibodies in sera from infected mice. Ten milligrams of parasite extract were incubated with 1 g of CNBr-Sepharose. Pools of sera from Day 28 of infection (acute serum) or of chronic sera were incubated in phosphate-buffered saline at pH 7.2 and then eluted with 0.2 Mglycine (pH 2.5) immediately neutralized with 2 M Tris. Effluent and eluate fractions were then concentrated and assayed in ELISA against T. cruzi antigen and the panel of natural antigens as described above. RESULTS
In order to determine the time-course and magnitude of total immunoglobulin responses in the serum of T. cruzi C3H/HeJinfected mice, we measured by ELISA the levels of total Ig on different days postinfection (Table I). Titers of total Ig were unchanged 1 week after infection (not shown), increased on Day 14 postinfection when the parasitemia reached a plateau, and remained high throughout the life of the animals. Nevertheless, maximal titers were obtained after 1 month of infection repre-
HONTEBEYRIE-JOSKOWICZ
TABLE I High Levels of Total Immunoglobulin in the Serum of Trypanosoma cruzi-Infected C3H/HeJ Mice Days of infection
Parasite/ml xIO-~ k SD
0” 14 21 28 35
4*1 321 421
74, 46-56 (1992)
PARASITOLOGY
Trypanosoma cruzi: Response
Predominance of IgG2a in Nonspecific during Experimental Chagas’ Disease
SYLVIANE SPINELLA,PASCALE LIEGEARD,AND Unit6 d’lmmunoparasitologie,
Humoral
MIREILLE HONTEBEYRIE-JOSKOWICZ'
Institut Pasteur, 25, Rue du Dr Roux, 75724 Paris Cedex 15, France
SPINELLA, S., LIEGEARD, P., AND HONTEBEYRIE-JOSKOWICZ, M. 1992. Trypanosoma cruzi: Predominance of IgG2a in nonspecific humoral response during experimental Chagas’ disease. Experimental Parasitology, 74, 4656. The kinetic and isotypic pattern of hypergammaglobulinemia has been investigated in C3H/HeJ infected with Trypanosoma cruzi. Hypergammaglobulinemia appeared 14 days postinfection, increased until Day 28 postinfection, and persisted throughout the chronic phase (>60 days of infection). The main isotype secreted was IgGZa, reaching IO-fold the control level. High titers of autoantibodies were found of IgM and IgG subclasses. Isotypic characterization of antibodies against myosin, myelin, and keratin, was performed and determined to be IgG2a subclass in the chronic stage of infection. Specific responses against T. cruzi took place 2 weeks postinfection when the parasitemia was high. Interestingly, parasite-specific response was maximal after 4 weeks of infection and plateaued during the chronic phase when parasites were rare. In contrast to the humoral polyclonal response in the chronic stage, showing a preferential IgC2a pattern, the anti-T. cruzi response consisted of all the different isotypes: IgM, IgGl, IgG3, IgGZa, and IgGZb, throughout the infection. Identical patterns of parasite antigens were recognized by IgG2a and IgG2b antibodies. Few different antigens were identified by the IgG3. Some antigens were recognized by several isotypes, others by only one isotype. With regard to the existence of antigenic cross-reactivities between host and parasite, we designed absorption experiments on parasite-specific immunoadsorbent showing that specific antibodies eluted from the column failed to recognize the natural antigens. These studies suggest that nonspecific and antiparasite-specific responses may be maintained by different regulatory pathways. o 1992 Academic PITHS, I”~. INDEX DESCRIPTORS AND ABBREVIATIONS: Trypanosoma cruzi, protozoan, experimental Chagas’ disease; Specific and nonspecific humoral response; Autoantibodies; Isotypic pattern; Immunoglobulin (Ig); Keratin (Ker); Myoglobin (Myo): Myosin (Myos); Phosphatebuffered saline (PBS); peroxydase (PO); thyroglobulin (Thyr); Trinitrophenyl (TNP); Tubulin (Tub).
that, at least in part, the immunosuppression observed during the acute phase and the autoimmunity associated with T. cruzi chronic infection (Plata 1985; Reed 1984; Hudson 1985) could be due to the intense polyclonal activation of the early stage of infection (Minoprio 1989). Indeed, an association between polyclonal B cell activation and autoimmune syndrome has been described in systemic lupus erythematous experimental models (Klinman and Steinberg 1987). In turn, the existence of immunological and molecular cross-reactivities between the parasite and its mammalian host-the so-called molecular mimicry-
Mouse infection with the protozoan parasite Trypanosoma cruzi (T. cruzi) sets off polyclonal B cell activation which appears early in the infection and persists during the life-long chronic phase (Ortiz-Ortiz et al. 1984; D’Imperio Lima et al. 1985, 1986). In the early stage of infection, the number of Ig secreting cells in the spleen and peripheral lymph nodes is very high and the majority of activated B cells secrete nonparasite-specific antibodies (Minoprio et al. 1988). The hypothesis has been proposed ’ To whom correspondence
should be addressed. 46
0014-4894/92 $3.00 Copyright0
1992 by Academic Press.Inc.
All rights of reproduction
in any form reserved.
ISOTYPE
DISTRIBUTION
IN EXPERIMENTAL
may account for some autoantibodies present in the sera of infected patients or animals and, possibly, relevant to the pathogenesis of chronic Chagas’ disease (Van Voorhis and Eisen 1989; Kahn et al. 1990; Levin et al. 1990; Mesri et al. 1990; Kerner et al. 1991). Whether the production of autoantibodies is the result of nonspecific polyclonal B cell activation or parasite-antigen-driven B cell stimulation remains to be elucidated. Polyclonal B cell responses in T. cruziinfected mice have been clearly demonstrated in lymphoid organs. In contrast, the isotype profiles of nonspecific humoral responses have been poorly investigated in experimental Chagas’ disease. Nevertheless, the isotype profile of T. cruzi-specific humoral responses has been well documented in mouse model and in humans (Krettli and Brener 1976, 1982; Scott and Goss-Sampson 1984; Araujo et al. 1984; Takehara 1989). We were interested in the kinetic of nonspecific immunoglobulins in the sera of infected animals and its comparison to that of the total parasite-specific responses. In the present report, we show (a) that hypergammaglobulinemia appeared 2 weeks after infection; (b) that the predominant isotype secreted during the chronic stage of infection was the IgG2a, including the natural autoantibodies, as evidenced for those reacting against myelin, keratin, and myosin; (c) that IgGl, IgG2a, and IgG2b were represented in the specific responses and recognized different parasite antigens; and (d) that the parasite-specific antibodies did not bind to natural antigens. Those data suggest that specific and nonspecific humoral responses may be triggered by different mechanisms during experimental Chagas’ disease. MATERIALS
AND
METHODS
Mice and infection. C3H/HeJ mice were bred at the Pasteur Institute. Eight- to lo-week-old mice were infected by ip injection of lo4 freshly collected blood-
CHAGAS’
DISEASE
47
stream trypomastigotes of the CL strain of T. cruzi. Mice were sacrificed on Days 14, 21, 28, 35, 60, 90, 120, and 180 postinfection. Sex- and age-matched uninfected mice were used as normal controls. Parasitemia was determined at the tail vein (Brener 1962). Antigens and antibodies. T. cruzi extract was obtained as previously described (HontebeyrieJoskowicz et a/. 1987). Briefly, Vero cells grown in Dulbecco’s medium with 5% calf serum (Flow laboratories, Asnieres, France) were infected with T. cruzi trypomastigotes. Trypanosomes released in culture medium were extensively washed and adjusted to 10’ parasites/ml in serum-free RPMI, frozen and thawed three times, and then passed through a 0.45urn Millipore filter. Native double-stranded (ds)-DNA, (porcine type II) thyroglobulin and (whale skeletal muscle type II) myoglobin, myosin of rabbit muscle, and keratin from human epidermis were purchased from Sigma Co. (St. Louis, MO); actin (Spudish and Watt 1971)and tubulin from rabbit muscle (Shelansky et al. 1973) were prepared as previously described; horseradish peroxidase (PO) was purchased from Boehringer (Manheim, Germany); TNP-BSA (TNP) was prepared according to Little and Eisen, 1966;myelin from mouse central nervous system was prepared according to Norton and Poduslo 1973, and was a generous gift of Dr. Ben You&-Chennouli (La Salpetriere, Paris). Peroxidase labeled-goat anti-mouse (IgM, IgGl, IgG3, IgG2b, or IgG2a) and unlabeled goat anti-mouse Ig were purchased from Southern Biotechnology Associates, Inc., as well as unlabeled isotypes (IgM, IgGl , IgG3, IgG2b, and IgGZa) used for standard. Enzyme-linked immunoassay (ELISA) for T. cruzi and natural antigens. Flat-bottomed plates (CML,
France) were coated with different antigens overnight at 4°C: 50 PI/well of T. cruzi antigen (25 ug/ml in KH,PO,/K,HPO, buffer, 50 n&f pH 8) or 100 uhwell of different antigens diluted in carbonate buffer pH 8.5; actin (10 p.g/ml), TNP-BSA (1 &ml), thyroglohulin (10 l&ml), PO (10 ug/ml), myoglobin (10 p&ml), tubulin (10 t.q$ml), myelin (1 l&ml), myosin (5 l&ml), except ds-DNA which was diluted in citrate buffer (0.5 mg/ml) and keratin (10 kg/ml) which was diluted in Tris buffer 0.1 M pH 7.4, added with 8 M urea and 1% mercaptoethanol. Coated plates were washed thoroughly with phosphate-buffered saline (PBS) containing 0.1% Tween 20 (Merck, France) (PBS-T) and incubated for saturation with PBS 1% gelatin (200 p,l/well) for 1 hr at room temperature (RT). Then, they were incubated in presence of serial dilutions of each sera (initial dilution l/10) for 2 hr at RT. After extensive washing with PBS-T, the plates were incubated for 1 hr at RT with peroxidase-labeled goat anti-mouse isotypes (labeled anti-IgM, anti-IgG3, and anti-Ig were used at l/2000, anti-IgGl l/4000, anti-IgG2b and IgG2a l/8000) and revealed with appropriate peroxidase sub-
48
SPINELLA,
LIEGEARD,
AND
skate. Results were expressed as the percentage of absorbance of infected sera on absorbance of a pool of normal sera, when natural autoantibodies were tested. For T. cruzi-specific response, the titer was expressed as the reciprocal dilution of the 50% of the maximal OD after titration at serial twofold dilutions. ELISA for total immunoglobulins in the serum. Flatbottomed plates were coated with unlabeled goat antimouse Ig (5 ug/ml) in carbonate buffer, pH 8.5, overnight at 4°C. The same protocol as described above was used, and the initial dilution used was l/500. Unlabeled isotypes were used as standards, at serial dilutions beginning with 180 &ml. Western blot analysis. Analysis of T. cruzi polypeptides recognized by the different Ig isotypes of sera from chronically infected mice was performed by Western blotting (Coding and Handman 1984).Briefly, crude lysates of purified trypomastigotes (2 x lO’/ml) from culture of Vero cells were loaded on SDSpolyacrylamide gels. After migration, the polypeptides were transfered onto nitrocellulose according to Towbin et al. (1979). The nitrocellulose strips were then saturated and incubated with normal or chronic sera diluted l/100. The reaction was developed with peroxidase-labeled goat anti-mouse Ig isotypes antibodies and 4-chloro-a-naphtol as substrate (Bio-Rad, France). Affinity chromatography. T. cruzi antigens were coupled to CNBr-activated Sepharose 4B (Pharmacia, France) following the manufacturer instructions and used to affinity purify the parasite-specific antibodies in sera from infected mice. Ten milligrams of parasite extract were incubated with 1 g of CNBr-Sepharose. Pools of sera from Day 28 of infection (acute serum) or of chronic sera were incubated in phosphate-buffered saline at pH 7.2 and then eluted with 0.2 Mglycine (pH 2.5) immediately neutralized with 2 M Tris. Effluent and eluate fractions were then concentrated and assayed in ELISA against T. cruzi antigen and the panel of natural antigens as described above. RESULTS
In order to determine the time-course and magnitude of total immunoglobulin responses in the serum of T. cruzi C3H/HeJinfected mice, we measured by ELISA the levels of total Ig on different days postinfection (Table I). Titers of total Ig were unchanged 1 week after infection (not shown), increased on Day 14 postinfection when the parasitemia reached a plateau, and remained high throughout the life of the animals. Nevertheless, maximal titers were obtained after 1 month of infection repre-
HONTEBEYRIE-JOSKOWICZ
TABLE I High Levels of Total Immunoglobulin in the Serum of Trypanosoma cruzi-Infected C3H/HeJ Mice Days of infection
Parasite/ml xIO-~ k SD
0” 14 21 28 35
4*1 321 421
