Cell surface antigens of Trypanosoma cruzi: Possible correlation with the interiorization process in mammalian cells

Molecular and Biochemical Parasitology , 6 (1982) 111-124 Elsevier Biomedical Press

111

CELL SURFACE ANTIGENS OF TR YPANO$OMA CR UZI: POSSIBLE CORRELATION WITH THE INTERIORIZATION PROCESS IN MAMMALIAN CELLS

BIANCA ZINGALES, NORMA W. ANDREWS,VERA Y. KUWAJIMAand WALTER COLLI

Departamento de Bioquimtca, Instituto de Quimica, BIoco 10 T, Universidade de 5~o Paulo, Caixa Postal, 20. 780, S~o Paulo, Brazil (Received 15 December 1981; accepted 7 April 1982)

Differences were observed in the pattern of the cell surface proteins of evolutive stages of Trypanosoma cruzi after radioiodination of the cells and subsequent analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Affinity chromatography revealed that surface glycoproteins also vary in the epimastigote and trypomastigote forms of the parasite. The cell surface antigens of the two forms were identified after radioiodination or biosYnthetic labeling with [35S]methionine and incubation with different antisera. Both epimastigotes and ttypomastigotes share two main antigens, possibly gly6aproteins, of apparent molecular weight 95 000 and 80 000, respectively, which are recognized by all antisera tested. On the other hand, the trypomastigote form possesses somewhat more cell surface antigens recognized both by rabbit anti-aTpomastigote and by human Chagasicsera. Sequential immunoprecipitations with heterologous and homologous antisera established that an 85 000 and some higher molecular weight antigens are specific to the trypomastigote form. Preincubation of the trypomastigotes with sera from Chagasic patients elicits a 60% inhibition of the parasite interiorization into mammalian cells in culture when compared with normal human serum or anti-epimastigote serum. This result suggests that the antigens specific to the trypomastigote form are involved in the interiorization process. Key words: Trypanosoma cruzi, Trypomastigotes, Endocytosis inhibition, Glycoproteins, Chagas' disease.

INTRODUCTION

Trypanosoma cruzi, the causative agent of Chagas' disease, belongs to the group of parasites which necessarily have to invade host cells to complete their biological cycle. Despite the fact that some preferential tissue tropism can be ascribed to particular strains [1], infections by 7". cruzi have been reported in a wide range of mammalian hosts where

Abbreviations: DME, Dulbecco's modified Eagle medium; FCS, fetal calf serum; LIT, liver infusion, tryptose; PBS, phosphate-buffered saline; PMSF, phenylmethylsulfonylfluoride; SDS, sodium dodecyl sulfate; TLCK, N-a-tosyl-L-lysylchloromethyl ketone. 0166-6851/82/0000-0000/$02.75 © 1982 Elsevier BiomedicalPress

112 several tissues are parasitized. The nonspecificity of tissue invasion is well documented in vitro where no cell line resistant to T. cruzi invasion has been described. Capacity to penetrate cells and subsequent resistance to intracellular lytic mechanisms may be regarded as properties acquired by the invasive form to protect it against the otherwise effective humoral and cellular immune attacks seen in in vivo infections [2]. It has been observed that in contrast to T. cruzi trypomastigotes, epimastigotes are not interiorized by vertebrate cells, with the exception of professional phagocytes. Since the primary event in the parasite interiorization by the host cells is the contact between both plasma membranes, it is very likely that specific surface components of the trypomastigore form are involved in the mechanism of interiorization. Using iodo-gen [3] as a catalyst of cell surface radioiodination, we have compared the distribution of proteins in both the epimastigote and trypomastigote forms of T. cruzL Immunoprecipitation studies demonstrated several antigenic determinants on the cell surface of the trypomastigote stage which could not be found in the epimastigote stage. These results were confirmed by endogenous labeling of the surface proteins with [as S]methionine. It has been possible to further identify the stage specific antigens by sequential immunoprecipitation with different antisera. Experiments with mammalian cells cultured in,vitro as a system for the study of T. cruzi infection [4] suggested that some of these antigens might be involved in host-parasite interaction. MATERIALS AND METHODS Parasites. Epimastigotes from the Y strain [5] of T. cruzi were grown in LIT medium [6]

in a rotatory shaker at 28°C. The parasites were obtained from 3 days cultures (98% epimastigotes), recovered by centdfugation (800 × g, 10 min), and washed three times with phosphate-buffered saline (PBS) before use. Trypomastigotes from the Y strain were maintained by weekly transfers in male A/Snell mice. Bloodstream forms were purified from citrate-treated blood of 7-d infected mice or irradiated mice (irradiated 24 h before infection with 400 rad from a cesium source). Parasites were collected and freed from blood cell contaminants by two centrifugations in Metrizamide (Nyegaard and Co., Oslo, Norway) gradients [7]. The parasites were further washed with PBS containing 1% glucose. Trypomastigotes were also obtained from cultures of LLC-MK2 cells (rhesus monkey kidney epithelial cells). The cells were maintained in Dulbecco's modified Eagle (DME)/10% fetal calf serum (FCS) plus 100 units/ml penicillin and 100/zg/ ml streptomycin at 37°C in a humidified atmosphere containing 5% CO2, and were propagated every three days after trypsin digestion. Blood from infected mice was aseptically harvested in the presence of 3.8% sodium citrate, centrifuged (800 × g, 10 min), and incubated for 60 rain at 37°C. The trypomastigotes in the supematant were added to flasks containing monolayers of LLC-MK2 cells in the DME medium (without FCS). After incubation for 72 h at 34°C the medium was changed to DME/2% FCS. The parasite burst occurred after three additional days of incubation at 34°C. Trypomastigotes were freed from cell debris and transition forms by centrifugation of the culture medium

113 and incubation at 37°C for 60 min. The trypomastigotes in the supematant were used in the experiments after thre 0 washings in medium 199. The parasites were propagated in ceil monolayers by 10-15 weekly transfers, without loosing infectivity in mice [4]. In a particular experiment, trypomastigotes were obtained in ceil cultures developed in absence of serum, in the medium described by Taub and Sato [8]. The parasites were collected after four consecutive transfers in this medium.

Antisera. Antiserum against trypomastigotes was obtained from a rabbit infected with three doses of 10a live trypomastigotes at four-week intervals. Anti-vesicles serum was prepared by immunizing rabbits with 100/ag of epimastigote plasma membrane vesicles [9], in 1 ml of PBS emulsified in an equal volume of Fretmd's complete adjuvant. Four weeks later the animals received a second dosis of 100/ag vesicles in 1 ml of PBS and 1 ml of Freund's incomplete adjuvant, and were bled after one month. Rabbit antiserum against the epimastigote form was kindly supplied by Mariza Morgado (Fundaqa'o Oswaldo Cruz, Rio de Janeiro, Brazil). Anti-epimastigote glycoprotein serum was kindly provided by Dr. D. Snary (Wellcome Research Laboratories, Beckenham, U.K.). Human Chagasic sera were obtained from chronic patients from different regions in Brazil. IgG was isolated from these sera by sodium sulfate precipitation and DEAEcellulose chromatography. Protein determinations were performed by the Lowry method [10]. Sera were characterized by indirect immunofluorescent antibody (IFA) and indirect hemagglutination (IHA) tests and were stored at -70°C. lodination of cell surface proteins. Purified parasites were washed three times by centrifugation as described above and radioiodinated by iodo-gen (1,3,4,6-tetrachloro-3a,6t~,diphenylglycoluril, Pierce Chemical Co.) [3]. Parasites (10 s cells)were incubated in 1 ml of PBS containing 500/aCi of NalalI (IPEN, S~o Paulo, Brazil), at 4°C for 10 min in tubes precoated with 20/ag ofiodo-gen. All of the parasites were found viable and motile after the labeling procedure. Furthermore, the trypomastigote form could still infect LLC-MK: monolayers. Ceils were washed twice with large volumes of Hanks' solution, and immediately lysed with electrophoresis sample buffer [11] containing sodium dodecyl sulfate and fl-mercaptoethanol in the presence of N-a.tosyl-L-lysylchloromethyl ketone (TLCK) (1 mM) and phenylmethylsulfonyl fluoride (PMSF) (1 mM). Samples were heated for 3 min and submitted to electrophoresis in SDS-polyacrylamide gels. Isolation of cell surface glycoproteins. The radioiodinated parasites were lysed in 2% (v/v) Nonidet P-40, 0.15 M NaC1, 10 mM Tris-HC1 (pH 7.5) containing 1 mM TLCK, 1 mM PMSF, 10 mM e-aminocaproic acid, 2.8 units/ml aprotinin, 25/~g/ml antipain and 6 mM p-aminobenzamidine by incubation at 37°C for 10 min. After centrifugation at 10 000 X g for 30 rain, the supematant was applied to Lens culinaris lectin Sepharose (Pharmacia) columns. The glycoproteins bound to the column were ehited with 0.1 M ot-methylmannoside, as described previously [12], and analysed by SDS-polyacrylamide gels. The six protease inhibitors were maintained during the whole procedure.

114

[aSS/Methionine labeling o f parasites. Parasites (5 × 107 ceUs/ml) were cultured for 4 h in methionine-free medium containing 50/aCi/ml of [2 S]methionine (1093 Ci/mmol, New England Nuclear). Epimastigotes were cultured in Ben6 and Parent's medium [13]. Trypomastigotes from tissue culture were incubated in DME medium containing 5%, amino acid-depleted FCS. After the labeling period the parasites were washed and used for immunoprecipitation studies. Immunoprecipitation analyses. Two different protocols were followed: (a) with the total lysate; (b) with intact living parasites. In the first case, the radioiodinated parasites were resuspended in a lysis buffer composed of 1% (v/v) Nonidet P-40, 10 mM Tris-HC1 (pH 7.5), 1 mM TLCK, 1 mM PMSF and 2.8 units/ml of aprotinin and subsequently incubated for 10 min at 37°C. The lysates were centrifuged at 10 000 X g for 30 min at 4°C and the supematants were stored at -70°C. Optimal immunoprecipitations were obtained with amounts of parasites around 3 X 107 cells, per experimental point. Lysate aliquots were incubated with appropriate dilutions of the immune sera for 2 h at 4°C, with occasional stirring. Antigen-antibody complexes were incubated for 30 min at room temperature with 70/al of 10% (w/v) suspension of heat killed and formalin fixed Staphylococcus aureus of the Cowan 1 strain [14]. The immunocomplexes formed were washed twice [14] and once more with a buffer containing 10 mM Tris-HC1 (pH 8.7), 0.3 M NaC1, 0.1% (w/v) SDS and 0.05% (v/v) Nonidet P-40, to reduce nonspecific precipitations. The precipitates were resuspended in 70/j1 of electrophoresis sample buffer [11 ], heated for 10 min at 40°C and for 3 rain at 100°C, centrifuged, and the supernatant was analysed by SDS-polyacrylamide gel electrophoresis. Identical results were obtained when controls were performed by precipitation of the cell lysates with suspensions ofS. aureus prior to incubation with the antibodies. For identification of the cell surface components with intact living parasites, 5 X 107 labeled cells were incubated in 1 ml of PBS containing either 700 #g of IgG or total serum (1 : 20 dilution) from different sources for 1 h at 4°C, with occasional stirring. After this period the parasites were washed with PBS, lysed with 250/al of the same lysis buffer as above, and centrifuged at 10 000 × g for 30 min. Immunocomplexes were precipitated with 100/al of a 10% (w/v) suspension ofS. aureus, washed and prepared for electrophoresis as described above. Sequential immunoprecipitations were performed by incubating the cell lysate with the first antiserum under the conditions specified before. The immunocomplexes were precipitated with S. aureus and the supernatant reincubated with the second antiserum. Antigen-antibody complexes were isolated with S. aureus again. Both precipitates were washed as above. Polyacrylamide gel electrophoresis. Electrophoresis was performed according to Laemli [11] in linear 7-14% gradients of polyacrylamide containing SDS. Molecular weight markers previously radioiodinated by the iodo-gen technique [3] were run on each slab gel in amounts less than 1/ag protein. Sharp bands were always observed. Gels containing 3s S-labeled proteins were processed for fluorography [ 15 ].

115 Gels were dried and exposed to Kodak X-RP 5 film. Short and long exposures were taken of each gel. The molecular weight adopted for the markers used were: fl-galactosdase = 130 kDa, phosphorylase a = 95 kDa; bovine serum albumin = 68 kDa; heavy chain of IgG = 55 kDa; light chain of IgG = 25 kDa. Interiorization o f T. cmzi in cultured mammalian cells. The system employed in these studies has been characterized before [4, 16]. Briefly, confluent monolayers of HeLa cells were obtained by plating 4 × 104 cells/cm2 in DME/5% FCS on coverslips 48 h before the experiment. The parasites were obtained from tissue culture as described above and preincubated at a density of 107 trypomastigotes/ml in DME containing different inactivated sera at a dilution of 1 : 24, for 60 rain at 4°C. After this period, cell monolayers were infected with these suspensions at a multiplicity of 25 parasites/ ceU, for 3 h at 34°C, in a humidified atmosphere containing 5% CO2. The antisera were maintained during the infection. The coversllps were then washed three times with PBS, pulsed with water for 2 min (in order to lyse adhered trypomastigotes [4, 16]) and washed again three times with PBS. The interi0rized parasites were observed under a 100 × oil immersion CF plan apochromat objective (Nikon HFM optiphot Microscope) after staining with dilute Giemsa [4]. The total number of parasites associated with the cells was determined and expressed as number of parasites/100 cells. Results are the mean of 4 coverslips where at least 200 cells were scored. The experiment was repeated 4 times.

RESULTS The cell surface proteins of the different stages of T. cruzi have been analysed after radioiodination of the parasites and SDS-polyacrylamide gel electrophoresis [12, 17, 18]. Several bands are common to the three forms, but the overall electrophoretic profiles are distinct and stage specific [12, 17]. It has been previously demonstrated [16, 19] that the radioiodination of the surface of trypanosomes through the lactoperoxidase catalysed reaction presents serious problems, since the enzyme (and its many contaminants) becomes iodinated during the incubation period and is strongly adsorbed on the cell surface of both epimastigote and trypomastigote forms [16]. Furthermore, iodo-gen is superior to chloramine-T since it provides a reproducible and gentle radioiodination of the surface proteins with no direct exposure to the oxidizing agent. Total surface patterns using iodo-gen as the iodinating agent are seen in Figs. 2a and 4a, confirming previous observations [12, 17]. These results are at variance with the extremely simple patterns presented by Nogueira et al. [18]. The cell surface glycoproteins from epimastigotes and tissue culture-derived trypomastigotes were analysed after in vivo radioiodination followed by parasite lysis and affinity chromatography. The material which was eluted with 0.1 M ~-methyhnannoside was electrophoresed in gels (Fig. 1). Epimastigotes show a strongly labeled glycoprotein (app. M r 95 kDa), as already described [12]. Additional bands were always detected with

116

Fig. 1. Autoradiogzaphs of SDS-polyacrylamide gel electrophoresis patterns of the cell surface glyco proteins of T. cruzi: (a) epimastigotes; (b) trypomastigotes.

Fig. 2. (Legend opposite)

117 the following app.Mr: 62 kDa, 80 kDa, and one higher than 130 kDa (probably 150kDa). A very faint band is seen at 70 kDa. Trypomastigotes show more than one glycoprotein on their cell surface. In the region between 80 kDa and 105 kDa a diffuse band is observed making it difficult to assess the true number of components by this method. This point was clarified by the immunoprecipitation studies (see below). Three different antisera were obtained against epimastigote antigens: anti-total epimastigotes, anti-purified plasma membrane vesicles, and anti-total glycoproteins. Immunoprecipitation studies showed that all three antisera recognized the same antigens of app. M r of 95,80 and 7 0 - 7 2 kDa (Fig. 2 b - d ) . When anti-trypomastigote serum was added to the epimastigote lysate, it recognized the 95 and 80 kDa antigens (Fig. 2e). Three sequential absorptions of epimastigote antigens with anti-trypomastigote serum practically exhausted the common antigens. Subsequent addition of anti~pirnastigote serum precipitated the 7 0 - 7 2 kDa protein (Fig. 20. This result suggests that either the latter antigen is specific to the epimastigote form or is little represented in the trypomastigote form. The presence of non-iodinated surface proteins was analysed by biosynthetic labeling of the epimastigote form with [as S]methionine. After 4 h incubation, the protein pattern (Fig. 3a) is much more complex than that obtained by radioiodination (Fig. 2a). The immunoprecipitation studies were performed by adding several antisera to viable 3s Slabeled cells, removing excess antibody by centrifugation, and by lysing the cells with detergent. The 95 and 80 kDa cell surface proteins were recognized by both rabbit anti-epimastigote and anti-trypomastigote sera, and by an IgG fraction from a human Chagasic patient, whereas the 7 0 - 7 2 kDa antigen was not precipitated by any of the above sera (Fig. 3c--e). Several other antigens are expressed on the trypomastigote cell surface as shown by immunoprecipitation with hyperimmune rabbit anti-trypomastigote and human Chagasic sera (Fig. 4b, c). A cluster of bands between 80 and 105 kDa, and well defined high molecular weight antigens up to 180 kDa were observed. Antisera of epimastigote origin precipitated only three antigens (95, 82, and 80 kDa proteins, Fig. 4d, e). These antigens are probably common to both forms and have a glycoprotein nature since they are precipitated by an antiserum against dpimastigote glycoproteins (Fig. 4e). When trypomastigote cell lysates were previously immunoprecipitated with excess anti-epimastigote serum and the supematant further precipitated with human Chagasic sera, specific trypomastigote antigens were visualized (Fig. 4f). In addition to the wide range of high molecular weight proteins already described, a conspicuous antigen of 85 kDa can be recognized (Fig. 4f).

Fig. 2. Labeled cell surface polypeptides (~alI) of the epimastigote form isolated by immunoprecipitation: (a) whole cell lysate; antigens precipitated with sera raised against: (b) epirnastigotes;(c) plasma membrane vesicles; (d) epimastigote glycoproteins; (e) trypomastigotes; (f) epimastigotes after three previous immunoprecipitations of the lysate with anti-trypomastigote serum.The irnmunoprecipitation with rabbit normal serum is not shown since no bands are detected even at long term exposures.

118

Fig. 3. Labeled polypeptides ([as S]methionme) of epimastigotes isolated by immunoprecipitation: (a) control, whole cell lysate; surface antigens precipitated with: (b) normal rabbit serum; (c) rabbit anti-epimastigote serum; (d) human Chagasic igG fraction; (e) rabbit anti-trypomastigote serum. The antigens observed on the trypomastigote cell surface are not derived from unspecific adsorption of LLC-MK2 components. No antigen was precipitated when controls were made in which the total proteins from the mammalian ceils were radioiodinated and incubated with anti-trypomastigote or human Chagasic sera. The endogenous nature of these specific antigens has been further demonstrated by labeling the parasites with [as S]methionine. Surface antigens were identified by incubating the living parasites either with anti-trypomastigote serum (Fig. 5a) or with lgG from human Chagasic sera (Fig. 5b). These antigens also migrate in the gel region between 80 and 150 kDa. In another set of experiments, trypomastigotes were obtained from LLCMK2 ceils in the absence o f fetal calf serum [8]. These parasites were labeled with Na 131 I and immunoprecipitated in vivo as above (Fig. 5c, d). The results were qualitatively the same as with 3s S-labeling. As a control, iodinated living trypomastigotes were precipitated by anti-epirnastigote plasma membrane vesicle serum and the results were similar to the patterns shown in Fig. 4d or e. In order to test whether anti-trypomastigote sera contained antibodies specifically dir ~ed a~ainst parasite interiorization, we used a system previously characterized in our labora ~ " "or th~ ~+udy of in vitro infection [4, 16]. HeLa cells were exposed to

119

Fig. 4. Labeled cell surface polypeptides (131I) of tissue culture trypomastigotes isolated by immunoprecipitation: (a) whole cell lysate; antigens precipitated with (b) human Chagasic serum; (c) rabbit anti-trypomastigote serum; (d) rabbit anti-epimastigote serum; (e) rabbit anti-epirnastigote glycoproteins serum; (f) human Chagasie serum after previous immunoprecipitation of the lysate with antiepimastigote serum. The immunoprecipitations with control sera are not shown since no bands were detected even at long term exposures.

trypomastigotes previously incubated with human serum from two Chagasic patients (Ch3 and Ch5) and serum against epirnastigote plasma membrane vesicles. None of these sera agglutinated the trypomastigotes at a dilution of 1 : 24 in DME medium. In contrast to the serum raised against purified epimastigote plasma membrane vesicles, both human sera inhibited interiorization (Ch3 = 37% and Ch5 = 63%) of trypomastigotes in HeLa cells (Fig. 6). These data suggest that the antigens shared by both differentiation stages do not play a role in the interiorization process. DISCUSSION The cell surface labeling of T. cruzi shows a rather complex pattern for both the epimastigote and trypomastigote forms. These results are at variance with those reported by Nogueira et al. [18] who refer to the existence of a single polypeptide both on the surface of the epimastigote (75 kDa) and trypomastigote (90 kDa) cells. The observed discrepancies are probably due to the different radioiodinating catalyst (iodogen) and to the more penetrating isotope (131 I) used in our studies. In fact, our autoradiographies

120

o

b

c

d

130-

25-

Fig. 5. Comparison between cell surface antigens of trypomastigotes labeled with [3SS]methionine (a, b) or with 1311 (c, d) and immunoprecipitation by incubating living parasites with rabbit antitrypomastigote serum (a, d)or with an IgG fraction from human Chagasic serum (b, c). Trypomastigores used for radioiodination were obtained from tissue cultures developed in the absence of fetal calf s e r u m .

Fig. 6. Effect of antisera in the interiorization of trypomastigotes in cultured mammalian HeLa cells. Parasites were preincubated (107 cells/ml, 60 min, 4°C) with normal human serum (NH); human Chagasic sera (Ch3, Ch 5); normal rabbit serum (NR); and rabbit anti-vesicle serum (V). lnteriorization was determined after an exposure of 3 h at 34°C. IFA titers in Ch3, Ch5 and V sera were 1 : 320, 1 : 320, and 1 : 640, respectively. IHA titers in Ch3, Ch 5 and V sera were 1 : 160, 1 : 320, and 1 : 320, respectively. of trypanosome cell surfaces labeled with Na12SI always showed less defined profiles when compared to those obtained with 131 I. The following evidence strongly indicates that we are n o t labeling internal proteins by the iodo-gen technique: (a) tubulin (app M r 56 kDa), the major Coomassie blue-stained protein of T. cruzi, is n o t labeled with Na13~I unless the parasites are treated with proteasesj (b) the same antigens are revealed by immunoprecipitation whether antibodies are added to radioiodinated samples of live parasites (Fig. 5) or cell lysates (Fig. 4). Using the isolation procedure described in refs. 12 and 20 we have shown that epimastigotes and trypomastigotes possess more than one glycoprotein on their surfaces. It has been suggested that the observation of high molecular weight glycoproteins is due to aggregation of lower molecular weight compounds [21]. Although the possibility

121 cannot be entirely ruled out by the methodology used we fred it improbable, since the high molecular weight glycoproteins are consistently observed even when the concentrations of SDS and ~-mercaptoethanol are doubled in the electrophoresis sample buffer. Furthermore, the 150 kDa glycoprotein from epimastigotes is not immunoprecipitated by the anfisera which precipitate the 95 and 80 kDa glycoproteins. On the other hand, the addition of six protease inhibitors during the isolation procedure excludes the possibility that the lower molecular weight bands are derived from proteolytic cleavage of higher molecular weight compounds. Controls using iodinated exogenous or endogenous proteins from T. cruzi demonstrated the efficiency of such inhibitors preventing proteolysis. Immunoprecipitation studies showed that epimastigotes have two main antigenic proteins, with apparent molecular weights of 80 and 95 kDa, which are recognized by rabbit anti-epimastigote and anti-trypomastigote sera, and by human Chagasic sera. Epimastigotes also possess another antigen with an apparent molecular weight of 70-72 kDa which appears in immunoprecipitations with homologous sera when the parasites are labeled with Na TM I. This antigen, possibly specific to the epimastigote form, is not recognized by anti-epimastigote serum when the parasites are labeled with [as S]methionine. We suggest that the lack of as S-labeling of this protein could be due to the low level of this antigen present in epimastigotes or to its low methionine content. Using a monoclonal antibody, Snary et al. [22] described an epimastigote-specific, cell surface glycoprotein with a 2 : 1 tyrosine: methionine ratio, and with an app.M r of 72 kDa. Thus, it is likely that the 70-72 kDa antigen detected in our immunoprecipitation studies corresponds to the glycoprotein described by these authors. The molecular weights of the two main antigens of the epimastigote surface coincide with those of two main glycoproteins isolated from epimastigotes by affinity chromatography (Fig. 1). In addition, the anti-epimastigote glycoprotein serum also recognizes the same proteins. These results suggest that the two antigens above are glycoproteins. The crossed immunoprecipitation studies with heterologous sera are indicative that these antigens are common to both differentiation stages of T. cruzi. The fact that anti-epimastigote serum precipitates an 82 kDa antigen in the trypomastigote surface, but does not recognize this protein in epimastigotes, suggests that this protein might share common antigenic regions with the 95 or the 80 kDa antigens. Using a different experimental approach a 90 kDa glycoprotein was found [12]. It was concluded that this protein is common to trypomastigotes (derived from blood of infected mice) and to epimastigotes obtained from different geographical areas or from clones [20]. It is highly likely that the 95 and 80 kDa glycoproteins found in our studies correspond, respectively, to the 90 kDa [12, 20] and 75 kDa [18] glycoproteinspreviously described. The difference in molecular weights could be explained by the higher resolution of slab gradient gels used by us and by the slightly different molecular weights assigned to protein markers in the various papers. The 75 kDa (80 kDa) glycoprotein has been found both in the epimastigote and trypomastigote (metacyclic) cell membrane [18]. We also fend a remarkable similarity ha the cell surface composition of these two

122 differentiation forms derived from axeuic liquid cultures (not shown). This means that metacyclic trypomastigotes seem to differ from blood or tissue culture trypomastigotes regarding their membrane protein patterns. Several other antigens, specific to trypomastigote surface, are immunoprecipitated by both rabbit anti-trypomastigote and human Chagasic sera. We assayed 20 sera from patients with chronic Chagas' disease, collected at random in different regions of Brazil. Most probably, all these patients were not infected by the same T. cruzi strain. It is remarkable that the immunoprecipitation patterns were basically the same, strongly suggesting that the antigenic characteristics of the plasma membrane of T. cruzi trypomastigotes are essentially constant. This hypothesis will be tested (work in progress) by correlating surface immunoprecipitation patterns with different 'zymodemes' [23, 24] and 'schizodemes' [25] of the parasite. The endogenous labeling of the same surface proteins with [aSS]methionlne, as revealed by in vivo immunoprecipitations, unequivocally demonstrates that these antigens are specific to the trypomastigote stage. Different experimental approaches have established that humoral immunity plays an important role in controlling infection by T. cruzi (cf. [1] ). However, the mechanism of antibody action is as yet unclear. Specific antigens could be involved in a recognition mechanism relevant for interiorization of the trypomastigote into the host cell. Antibodies directed against such antigens could disturb or inhibit this recognition. It is well known that trypomastigotes are interiorized by mammalian cells in culture, while epimastigotes are not. We found that, while human Chagasic serum does inhibit trypomastigote interiorization in LLC-MK2 and HeLa cells, anti-epimastigote serum does not. It must be concluded that the surface antigens common to both stages, which are recognized by the latter serum, do not play any role in host-cell recognition and/ or penetration. Chagasic or anti-trypomastigote sera recognize antigens specific to the cell surface of trypomastigotes. Evidence is not available as yet for a specific role of these antigens in host-parasite recognition and interiorization. Sera from Chagasic patients are effective in inhibiting in vitro interiorization, suggesting that specific antibodies against penetration are present. The 85 kDa or any of the proteins of molecular weights above 95 kDa, found in our work to be present only in trypomastigotes, seem to be good candidates for cell surface proteins involved in the mechanism of parasite interiorization. ACKNOWLEDGEMENTS The authors are indebted to Dr. S. Schreier for the helpful suggestions during the preparation of the manuscript. This investigation was supported by grants from UNDP/ World Bank/WHO Special Programme for Research and Training in Tropical Diseases, Conselho Nacional de Desenvolvimento Cientffico e Tecnol6gico (Projeto CNPq/FINEP), and PNUD/UNESCO RLA 78/024 to W. Colli and Funda~o de Amparo h Pesquisa do Estado de Sffo Paulo (FAPESP) to B. Zingales.

123

REFERENCES 1 2 3

4 5 6 7 8 9 10 11 12 13 14

15 16

17 18 19 20 21

22

Brener, Z. (1980) Immunity to Trypanosoma cruzi. Adv. Parasitol. 1 8 , 2 4 7 - 2 9 2 . Colli, W. (1979) The membrane pathobiology of tropical diseases. In: Tropical diseases research series: 2, (Wallach, D.F.H. ed.,), pp. 131-153. Schwabe & Co., Basel. Fraker, F.J. and Speck, J.C. (1978) Protein and ceil membrane iodinations with a sparingly soluble chloroamide 1,3,4,6-tetrachloro-3~,6a-diphenylglycoluril. Biochem. Biophys. Res. Commun. 80, 8 4 9 - 8 5 7 . Andrews, N.W. and CoUi, W. (1982) Adhesion and interiorization of Trypanosoma cruzi in mammalian ceils. J. Protozool. 29, 2 6 4 - 2 6 9 . Silva, L.H.P. and Nussenzweig, V. (1953) Sobre uma cepa de Trypanosoma cruzi altamente virulenta para o camundongo branco. Folia Clin. Biol. 20, 191-207. Casteilani, O., Ribiero, L.V. and Fernandes, J.F. (1967) Differentiation of Trypanosoma cruzi in culture. J. Protozool. 1 4 , 4 4 7 - 4 5 1 . Pereira, M.E.A., Loures, M.A., Villalta, F. and Andrade, A.F.B. (1980) Lectin receptors as markers for Trypanosoma cruzi. J. Exp. Med. 152, 1375-1392. Taub, M. and Sato, G. (1979) Growth of kidney epithelial cells in hormone-supplemented, serumfree medium. J. Supramol. Struct. 1 1 , 2 0 7 - 2 1 6 . Franco da Silveira, J., Abrahamsohn, P.A. and Colli, W. (1979) Plasma membrane vesicles isolated from epimastigote forms of Trypanosoma cruzi. Biochim. Biophys. Acta 5 5 0 , 2 2 2 - 2 3 2 . Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 1 9 3 , 2 6 5 - 2 7 5 . Laemli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T 4 . Nature 2 2 7 , 6 8 0 - 6 8 5 . Snary, D. and Hudson, L. (1979) Trypanosorna cruzi cell surface proteins: identification of one major glycoprotein. FEBS Lett. 1 0 0 , 1 6 6 - 1 7 0 . Bon6, G.J. and Parent, G. (1963) Stearic acid an essential growth factor for Trypanosoma cruzi. J. Gen. Microbiol. 3 1 , 2 6 1 - 2 6 6 . Kessler, S.W. (1975) Rapid isolation of antigens from cells with a Staphylococcal protein A-antibody adsorbent: parameters of the interaction of antibody-antigen complexes with protein A. J. lmmunol. 115, 1617-1624. Chamberlain, J.P. (1979) Fluorographic detection of radioactivity in polyacrylamide gels with water soluble fluor, sodium salicylate. Anal. Biochem. 9 8 , 1 3 2 - 1 3 5 . Colli, W., Andrews, N.W. and Zingales, B. (1981) Surface determinants in American trypanosomiasis. In: International Cell Biology 1980-1981, (Schweiger, H.C. ed.), pp. 4 0 1 - 4 1 0 . SpringerVerlag Berlin, Heidelberg. Araujo, F.G. and Remington, J.S. (1981) Characterization of stages and strains of Trypanosoma cruzi by analysis of cell membrane components. J. Immunol. 1 2 7 , 8 5 5 - 8 5 9 . 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. 1 5 3 , 6 2 9 - 6 3 9 . Camargo, E.P., Barbieri, C.L. and Jankevicius, J.V. (1982) Possible artifact in the radioiodination of surface proteins of Trypanosomatids. J. Immunol. Methods (in press). Snary, D. (1980) Trypanosoma cruzi: antigenic invariance of the cell surface glycoprotein. Exp. Parasitol. 49, 6 8 - 7 7 . Snary, D. and Scott, M. (1980) A cell surface glycoprotein from Trypanosoma cruzi capable of inducing protective immunity. In: The host-invader interplay (Van den Bossche, H. ed.), pp. 2 5 3 - 2 5 6 . Elsevier/North-Holland Biomedical Press, Amsterdam. Snary, D., Ferguson, M.A.J., Scott, M.T. and Allen, A.K. (1981) Cell surface antigens of Trypanosoma cruzi: use of monoclonal antibodies to identify and isolate an epimastigote specific glycoprotein. Mol. Biochem. Parasitol. 3 , 3 4 3 - 3 5 6 .

124

23

24

25

Miles, M.A. (1979) Transmission cycles and the heterogeneity of Trypanosoma cruzi. In: Biology of the Kinetoplastida (Lumsden, W.H.R. and Evans, D.A., eds.), vol. 2, pp. 117-196. Academic Press, London. Romanha, A.J., Pereira, A.A.S., Chiari, E. and Kilgour, V. (1979) Isoenzyme patterns of cultured Trypanosoma cruzi. Changes after prolonged subculture. Comp. Biochem. Physiol. B 62, 139142. Morel, C., Chiari, E., Camargo, E.P., Mattei, D.M., Romanha, A.J. and Simpson, L. (1980) Strains and clones of Trypanosoma cruzi can be characterized by pattern of restriction endonuclease products of kinetoplast DNA minicircles. Proc. Natl. Acad. Sci. U.S.A. 77, 68106814.