The factors involved in the resistance or susceptibility of mammalian hosts to infection with Trypanosoma cruzi remain obscure. A more detailed analysis of the interactions of the parasite and host cells would provide a better understanding of the pathogenesis of this infection. Several investigators have examined the mode of entry, intracellular development, and fate of T. cruzi in a variety of cells using both light and electron microscopic techniques, but the result of these studies are contradictory (1-11).We report here our studies on the mode of entry and fate of both epimastigote and trypomastigote forms ofT. cruzi at the light and electron microscope levels. The observations were mainly performed with mouse peritoneal macrophages, although parallel studies on several cultured cell types (HeLa, L cells, calf embryo fibroblasts) will also be described. Utilizing a technique to separate the cultured forms of epimastigotes and trypornastigotes (12), we were then able to study their different fates as well as the components of the host cell plasma membrane involved in the interiorization process. Some of these results have been previously published in abstract form (13).

As reported previously, mouse peritoneal macrophages could be activated to kill intracellular trypomastigotes of Trypanosoma cruzi, the agent of Chagas' disease, in either of two ways: by immunizing and boosting the mice (3), or by culturing resident or inflammatory macrophages in spleen cell factor(s) (SCF) in vitro (2). Macrophages activated in vivo became less trypanocidal with time in culture, and cells activated in vitro lost trypanocidal capacity when CSF was removed (2). In the present study, the ability of macrophages to release H2O2 in response to phorbol myristate acetate (PMA) could be induced in vivo and in vitro, and reversed in vitro, in a manner correlating closely with changes in trypanocidal activity. Macrophages could be activated in vitro with SCF in a time-dependent and dose-dependent fashion, so that they released as much H2O2 as macrophages activated in vivo. The sensitivity of epimastigotes and trypomastigotes to enzymatically generated H2O2 suggested that the generation of H2O2 by activated macrophages could be plausible explanation for their trypanocidal activity. Of the biochemical correlates of macrophage activation reported to date, increased ability to release H2O2 seems most closely allied to enhanced capacity to kill an intracellular pathogen.

OFIn previous papers we have shown that trypomastigotes of Trypanosoma cruzi, the etiological agent of Chagas' disease, survive quantitatively and replicate in the cytoplasm of both normal and inflammatory mouse peritoneal macrophages (1). Organisms enter the cell via phagocytosis, are initially enclosed within a phagocytic vacuole, and subsequently escape into the cytosol. In contrast, macrophages obtained from either T. cruzi or Bacille CalmetteGu~rin (BCG)l-infected mice, which had been elicited by a secondary challenge with specific antigen, were capable of destroying a majority of the intracellular parasites (2). Exposure of normal macrophages to lymphokines failed to modify their microbicidal activity, but promptly stimulated their secretion of plasminogen activator. The generation of the active lymphocyte supernatant fluid(s) required the presence of thymus-derived lymphocytes (3).In this study, we report the conditions under which we are now able to induce and maintain trypanocidat activity in both resident and inflammatory mouse peritoneal macrophages maintained in vitro. Materials and MethodsParasites. The Y strain of T. cruzi was obtained from Dr. Ruth Nussenzweig (New York University School of Medicine, New York). Blood form trypemastigotes were obtained from infected NCS mice and grown in liver-infusion tryptose (LIT) medium (4). Parasites were harvested from 30-day-old cultures in LIT, washed five times in ice-cold phosphate-buffered saline (PBS) (Dulbecco's; Grand Island Biological Co., Grand Island, N. Y.) at 750 g for 15 rain, resuspended in icecold PBS or Dulbecco's modified Eagle's medium (Grand Island Biological Co.) and counted in a hemocytometer with a × 40 objective. These cultures of Y strain in LIT contained >30% trypomastigotes. Trypomastigotes were purified by means of a modified procedure (5), utilizing a Metrizamide gradient (Nyegaard and Company A]S, Oslo, Norway). The gradients were prepared with 2 ml of 17.5% Metrizamide (d = 1.0924 g-cm -3) and 2 ml of 15% Metrizamide (d = 1.0787 g-cm -~) in Hepes buffer, pH 7.4, 260 mosmol (6). 107 parasites, after treatment with guinea-pig complement (Cordis Laboratories Inc., Miami, Fla.) were washed and resuspended in 1 ml of Hepes buffer, and overlayed on the Metrizamide gradient. The preparation was then centrifuged at 2,000 g for * Supported by grants 5-29055 from the Rockefeller Foundation, and AI 07012 from the U. S. Public Health Service.~Abbreviations used in this paper: BCG, Bacille Calmette-Gu~rin; Con A, concanavatin A; D2FBSME, Dulbecco's medium containing 2% fresh fetal bovine serum plus 5 × 10 -s M mercaptoethanol; D2HIFBS, Dulbecco's medium containing 2% heat-inactivated fetal bovine serum; FBS, fetal bovine serum; hkBCG, heat-killed BCG; HKT, heat-killed trypanosomes; LIT, liver-infusion tryptose; LPS, lipopolysaccharide; ME, mercaptoethanol; PBS, phosphate-buffered saline; PP, proteose peptone; PPD, purified protein derivative of tuberculin; SCF, spleen cell factor. 288J. ExP. MED.

The mechanism by which culture forms of Trypanosoma cruzi are lysed by normal mammalian sera was examined. Lysis is restricted to the epimastigote form of the organism and is not dependent on the presence of agglutinins. Lysis is a complement-dependent process, the activity being generated by the alternate pathway. The selective lysis by serum was exploited to purify viable trypomastigotes by means of centrifugation in an albumin column. Essentially pure trypomastigote populations are now being employed in cell culture experiments.

Soluble products from antigen stimulated Trypanosoma cruzi-immune spleen cells enhanced the expression of Ia antigens on proteose-peptone-elicited mouse peritoneal macrophages (M phi). Acquisition of Ia paralleled M phi activation, previously shown to be mediated by this same source of lymphokine (LK). Expression of Ia and four other plasma membrane antigens was monitored by quantitative binding and radioautographic studies with 125I-monoclonal antibodies. Immune LK selectively enhanced expression of Ia and, to a lesser extent, H-2D relative to control LK from antigen-stimulated noninfected spleen. The levels of three other non-major histocompatibility complex (MHC) antigens, including the trypsin-resistant Fc receptor, were similar in cells exposed to both sources of LK. As little as 1% immune LK induced one-half maximal expression of Ia. Kinetic studies revealed that much of the Ia on freshly explanted peritoneal M phi was lost during the 1st d of culture. In the continued presence of immune LK, Ia was re-expressed on virtually all M phi by the 2nd and 3rd d. Alternatively, > 95% Ia negative populations were obtained by culturing the cells 3 d; then, addition of LK induced Ia on most cells within 1 d. Once induced, Ia persisted on the M phi surface for at least 2 d. [35S]methionine radiolabeling indicated that immune LK selectively increased radiolabeling of M phi Ia, again with other non-MHC-linked plasma membrane polypeptides as controls. LK-induced Ia-bearing M phi were tested as primary mixed leukocyte reaction stimulators. 1 x 10(5)-2 x 10(5) M phi did not stimulate 4.5 x 10(6) responding T cells, whereas 10(4) dendritic cells induced strong responses, as previously described. Because Ia-positive M phi do not actively sensitize T cells in a model immune response, we propose that M phi MHC products serve primarily as recognition sites for previously sensitized T cells, thereby enhancing T cell-mediated M phi activation.