INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Nuclear factor-B (NF-B) is a transcription factor involved in many cellular functions, including the innate immune response to pathogens (Kopp and Ghosh, 1995; Baldwin, 1996). NF-B family proteins normally exist as dimers, the most common form being the heterodimer p65/p50 (Baeuerle and Henkel, 1994). The dimers are retained in an inactive form in the cell cytoplasm by interaction with an inhibitory subunit, IB (Baeuerle and Baltimore, 1988). Activation occurs when phosphorylation-induced degradation of IB allows translocation of NF-B to the nucleus, where binding to specific DNA sequences induces gene expression (Henkel et al., 1993; Chen et al., 1995). Many microbial products, including viral proteins, bacterial lipopolysaccheride (Müller et al., 1993; Herrero et al., 1995), and glycophosphatidylinositols from various parasites (Tachado et al., 1996, 1997), can activate NF-B, thus inducing expression of proinflammatory cytokines such as tumor necrosis factor- (TNF-) and interleukin-1 (IL-1) (Collart et al., 1990; Goldfeld et al., 1990; Kopp and Ghosh, 1995). NF-B also regulates expression of inducible nitric oxide synthase (iNOS), which produces the antimicrobial radical NO (Xie et al., 1994). Activation of NF-B is therefore an essential step in the innate immune response to pathogens (Elewaut et al., 1999).

Trypanosoma cruzi is the causative agent of Chagas disease, which affects almost 20 million people in the Americas. The infection, once acquired, is lifelong with three distinct stages of disease, the acute stage, the indeterminate stage, and chronic stage. Acute infection is accompanied by mild to severe fever and is occasionally fatal in small children. Most chagasic patients in the indeterminate stage are asymptomatic, whereas those in the chronic stage may develop gross enlargement of the heart (cardiomegaly) and/or gastrointestinal organs (megaesophagus and megacolon). Although the pathology was initially attributed to autoimmune responses, it is now thought that persistence of parasite antigens is required for development of disease (Tarleton et al., 1997; Zhang and Tarleton, 1999).

T. cruzi is able to invade and multiply in many different cell types of many different species but shows a marked preference for myocytes. Thus, parasites are abundant and invade a variety of cells throughout the body in the acute stage of infection, but intracellular infection is limited to skeletal, smooth, and cardiac muscles in the indeterminate and chronic stages (Bice and Zeledon, 1970; Brener, 1973). Although parasite strain is believed to determine which particular organs are affected (Andrade, 1978; Melo and Brener, 1978), little is known of the host factors that lead to preferential infection of myocytes over other cell types.

Infection with T. cruzi causes increased expression of a number of proteins regulated by NF-B, including the cytokines TNF-, IL-1, and IL-6 and the adhesion molecules intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1) (Zhang and Tarleton, 1996; Huang et al., 1999). In addition, infection leads to up-regulation of iNOS (Huang et al., 1999). Some of these responses are essential for host control of the infection, because susceptibility to infection is enhanced in mice lacking iNOS or the TNF- receptor p55 (Castanos-Velez et al., 1998; Holscher et al., 1998). Although inflammatory cells play a major role in cytokine production during T. cruzi infection, other cells may also be involved. Infection of endothelial cells with T. cruzi causes direct induction of IL-1 and IL-6 (Tanowitz et al., 1992). In addition, a released surface protein of T. cruzi, trans-sialidase, can induce IL-6 production in isolated endothelial cells (Saavedra et al., 1999). These findings suggest that the response of nonimmune cells to the parasite may be a key regulatory step during infection.

The ability of T. cruzi to stimulate production of NF-B-regulated cytokines suggests that this transcription factor could be closely involved in the control of T. cruzi infection. Given that NF-B also controls expression of antiparasitic proteins such as iNOS, activation of NF-B could allow cells to limit intracellular infection. The aim of this work was to investigate the role of NF-B activation in T. cruzi infection of mammalian cells. We present evidence that T. cruzi trypomastigotes activate NF-B in a number of cells, which are relatively resistant to infection, including epithelial cells, endothelial cells, and fibroblasts. By contrast, the parasite fails to activate NF-B in myocytes, the cells that are most susceptible to invasion in vitro and in vivo. Furthermore, we demonstrate that inhibition of NF-B activation enhances infection of epithelial cells. These results suggest that NF-B does indeed regulate intracellular infection and provide a molecular explanation for the preferential infection of muscle cells by T. cruzi.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cells

Mink lung epithelial cells (Mv1lu) were cultured in minimum essential medium, whereas all other cells were maintained in Dulbecco's modified Eagle's medium (Life Technologies, Gaithersburg, MD). Media were supplemented with 10% FBS, 12.5 mM HEPES, 2 g/l sodium bicarbonate, penicillin, and streptomycin (all from Life Technologies). Myoblasts (L₆E₉ and H₉c₂) were transferred to low-serum medium before assay to induce myocyte differentiation (Nadal-Ginard, 1978).

Parasites

T. cruzi parasites of Silvio, Tulahuen, and MV13 strains (Prioli et al., 1990) were maintained in Vero cells in RPMI 1640 medium containing 2.5% Nuserum (Collaborative Laboratories, Bedford, MA), 12.5 mM HEPES, 2 g/l sodium bicarbonate, penicillin, and streptomycin (Life Technologies). Trypomastigotes were harvested by centrifugation at 500 x g for 5 min to remove host cells and 1200 x g for 10 min to recover parasites. For assays, trypomastigotes were resuspended in RPMI and 1% BSA. Conditioned medium was prepared by incubating trypomastigotes overnight at 5 × 10⁷ in RPMI and 1% BSA at 37°C. Infection assays were carried out as described (Ortega-Barria and Pereira, 1991).

Nuclear Translocation Assays

Cells were plated into 16-well Labtek chamber slides (Nalge Nunc International, Naperville, IL), incubated overnight, and then incubated for 1 h with samples in RPMI and 1%BSA, fixed with 4% paraformaldehyde in PBS for 20 min, permeabilized with 0.5% Triton-X-100 in PBS for 10 min, and blocked with 10% FBS in PBS. Cells were stained with goat anti-p65 polyclonal antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) at 2 µg/ml in PBS and 1%BSA followed by FITC-labeled anti-goat immunoglobulin G (Boehringer Mannheim, Indianapolis, IN) at a dilution of 1:50.

Electrophoretic Mobility Shift Assay

Mv1Lu cells (2 × 10⁷) in 10-cm dishes were incubated with parasites at 2 × 10⁷/ml for 2 h and then washed in ice-cold PBS. Cell harvesting, lysis, nuclear isolation, and extraction and binding to the NF-B binding site of the probe H2K were all carried out as described (Sen and Baltimore, 1986). Nuclear extract protein concentration was determined by Bradford assay, and binding reactions contained 3 µg of protein per assay.

NF-B-dependent Luciferase Activity

NF-B-dependent gene expression was studied in a transient transfection assay. Cells were plated at 1 × 10⁵ per well in six-well plates and incubated overnight. The cells were then cotransfected with 1 µg pBIIXluc, a plasmid containing luciferase under the control of two Ig-B sites (Kopp and Ghosh, 1994) and 1 µg of pSVbGal (Promega, Madison, WI) in the presence of 5 µl/well LipofectAMINE (Life Technologies) in serum-free RPMI. After 5 h of incubation the RPMI was replaced with minimum essential medium containing 10% FBS for Mv1Lu cells and Dulbecco's modified Eagle's medium containing 2.5% horse serum (growth medium) for L₆E₉ and H₉c₂ cells, and the cells were incubated overnight. Trypomastigotes in RPMI and 1%BSA were added for 2 h, and then the cells were washed with serum-free medium and incubated a further 24 h in growth medium. Cells were harvested and assayed for luciferase activity using the Promega luciferase detection system. Activity was normalized according to -galactosidase activity.

Generation of Stable Tetracycline-regulated IBaM/GFP Tranfectants

The plasmid pTR5-IBaM/GFP was constructed by excision of the murine mutant IBa gene IBaM from the plasmid pCMXIBaM (Van Antwerp et al., 1996) by restriction digestion with EcoRV and insertion into the PME1 site of the tetracycline-regulated dicistronic expression plasmid pTR5-DC/GFP (Mosser et al., 1997). To generate cells expressing the tetracycline-regulated transactivator protein tTA, which allows control of gene expression by tetracycline, Mv1Lu cells were transfected with PtTA-hygro and selected for hygromycin B resistance. Positive cells were isolated by fluorescence-activated cell sorting (FACS) after transient transfection with pTR-GFP, a plasmid expressing green fluorescent protein (GFP) under the control of the tet operator (Mosser et al., 1997). These cells were cotransfected with PTR5-IBaM/GFP and PCDNA3 and selected for neomycin resistance in the presence of 1 µg/ml Geneticin (Life Technologies). In addition, 10 ng/ml tetracycline was included in the medium to prevent expression of the IBaM gene during selection. Cells expressing IBaM and GFP were selected by removal of tetracycline for 24 h followed by FACS. Cells were maintained in the presence of 10 ng/ml tetracycline and transferred to tetracycline-free medium 24 h before assay.

Western Blotting

Cells were plated into 10-cm dishes at 5 × 10⁶ per plate and incubated overnight in the presence or absence of 5 ng/ml tetracycline, the minimum concentration required to block expression in these cells. Cells were washed with PBS, scraped from the plates, transferred to Eppendorf tubes, and centrifuged. The pellet was lysed in radioimmunoprecipitation assay buffer: 20 mM Tris, pH 7.5, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM dithiothreitol, 1 mM sodium orthovanadate, 1 mM sodium fluoride, and Complete protease inhibitor mixture (Boehringer Mannheim). Proteins were separated on a 12% SDS-PAGE gel and blotted onto nitrocellulose. IBa and IBaM were detected with rabbit anti-IBa antibodies (Santa Cruz). Blots were developed by enhanced chemiluminescence (Amersham, Arlington Heights, IL).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

To determine whether T. cruzi can activate NF-B, Mv1Lu cells were exposed to trypomastigotes, the infectious stage of T. cruzi, and stained with antibodies reactive to the p65 subunit of NF-B. In control Mv1Lu cells, p65 is limited to the cytoplasm (Figure 1A, first panel). As a positive control, cells were incubated with TNF-, which induces rapid translocation of p65 to the nucleus (Figure 1A, second panel). Exposure of Mv1Lu to trypomastigotes also causes nuclear translocation of p65 within 1 h of addition (Figure 1A, third panel). The percentage of positive nuclei increases in a dose-dependent manner with increasing parasite concentration (Figure 1B). Parasite invasion of the host cell is not required for stimulation, because similar activation can be induced by T. cruzi-free medium conditioned by overnight incubation with trypomastigotes (Figure 1A, fourth panel). Conditioned medium triggers nuclear translocation in 80% of cells, a level similar to that induced by TNF- at a concentration of 10 ng/ml. These results indicate that both intact T. cruzi and soluble material released by trypomastigotes can stimulate NF-B activation.

View larger version (35K): [in this window] [in a new window]   Figure 1.   Activation of epithelial cell NF-B by T. cruzi. (A) Mv1Lu cells incubated 1 h with RPMI and 1%BSA, 10 ng/ml recombinant human TNF- (R & D Systems, Minneapolis, MN), Silvio strain trypomastigotes (1 × 107/ml), or parasite conditioned medium and stained with anti-p65 antibodies. (B) Mv1Lu cells exposed to different concentrations of trypomastigotes and stained for p65 as above. At least 200 cells were counted per well. Each point represents the mean of triplicate assays ± SEM. (C) DNA binding in nuclear extracts from Mv1Lu incubated for 1 h with RPMI and 1% BSA with (+) or without () trypomastigotes. (D) Luciferase activity in Mv1Lu cells transiently transfected with pBIIXluc, exposed to different concentrations of trypomastigotes. Each point represents the mean of triplicate assays ± SEM.

To examine whether the translocated NF-B is active in DNA binding, nuclear extracts were prepared from Mv1Lu cells after exposure to trypomastigotes. NF-B-specific DNA binding was detected by gel shift assay in cells exposed to parasites but not in control cells (Figure 1C). Similar results were obtained in cells treated with conditioned medium (our unpublished results). To confirm that the parasite-induced translocation of p65 to the nucleus is sufficient to induce changes in gene expression, Mv1Lu cells were transiently transfected with the reporter plasmid pBIIXluc, in which luciferase gene expression is dependent on NF-B (Kopp and Ghosh, 1994). Trypomastigotes induce a dose-dependent increase in luciferase activity (Figure 1D), showing that the signal triggered by the parasites is sufficient to induce changes in NF-B-dependent gene expression.

The T. cruzi life cycle has multiple stages, each with distinct invasive properties, morphology, and antigenic makeup (Brener, 1973). Activation of NF-B is triggered by trypomastigotes, the life cycle stage of T. cruzi infectious to mammalian cells. Epimastigotes, the parasite stage that is infectious for insects but cannot infect mammalian cells, have no effect on NF-B activity in Mv1Lu cells, as determined by induction of luciferase expression (Figure 2A) and nuclear translocation assays (our unpublished results). The NF-B response is therefore specific for the stage of the parasite that invades mammalian cells during a natural infection.

View larger version (10K): [in this window] [in a new window]   Figure 2.   Specificity of NF-B activation. (A) Activation of NF-B-dependent luciferase transcription in Mv1Lu cells incubated 2 h with Silvio strain trypomastigotes or epimastigotes at a concentration of 1 × 107/ml in RPMI and 1% BSA. Cells were harvested 24 h after addition of parasites. Results represent the mean of triplicate transfections ± SEM. (B) Nuclear translocation of p65 in Mv1Lu cells exposed for 1 h to RPMI and 1% BSA and Silvio, Tulahuen, or MV13 strain parasites at 1 × 107/ml in RPMI and 1% BSA. Results represent the mean of duplicate assays ± range.

T. cruzi strain is regarded as an important component of tissue tropism (Andrade, 1978; Melo and Brener, 1978). To establish whether strains differ in their ability to induce activation of NF-B, Mv1Lu cells were exposed to three distinct strains of parasite, Silvio, MV13, and Tulahuen (Figure 2B). All three strains induce similar levels of p65 nuclear translocation in Mv1Lu cells, suggesting that activation of NF-B is independent of strain. A similar pattern was observed in activation of luciferase in pBXIIluc-transfected cells (our unpublished results). Stimulation of an NF-B response is therefore due to a component common to diverse T. cruzi strains.

To determine whether activation of NF-B has any direct impact on susceptibility to infection, we developed a system for inducible expression of the dominant negative mutant IBa gene IBaM (Van Antwerp et al., 1996). Mv1Lu cells were transfected with PtTA-hygro, a plasmid encoding a tetracycline-regulated transcriptional activator (tTA) (Mosser et al., 1997). Positively selected cells were transfected with PTR5-IBaM/GFP, a plasmid containing IBaM and GFP in a dicistronic cassette under the control of a promoter containing the tet operator sequence. Stable transfectants exhibit tetracycline-repressible IBaM and GFP expression. Expression of IBaM and its negative regulation by tetracycline were confirmed by Western blotting (Figure 3A).

View larger version (28K): [in this window] [in a new window]   Figure 3.   Expression of IBaM enhances infection in epithelial cells. (A) Tetracycline-regulatable expression of IBaM: Western blot of whole-cell lysates from tTA- or IBaM-expressing Mv1Lu cells incubated overnight with or without 5 ng/ml tetracycline. Blots were incubated with anti-IBa antibodies and visualized by ECL. Each lane contains 50 µg of protein. (B) Luciferase activity in pBXIIluc-transfected tTA- or IBaM-expressing Mv1Lu cells (Mv) 24 h after exposure to RPMI and 1%BSA, Silvio strain trypomastigotes (1 × 107/ml), tetracycline (+ Tet; 5 ng/ml), or tetracycline and trypomastigotes (+ Tet + Tryp; 1 × 107/ml). Each point represents the mean of triplicate assays ± SEM. (C) Infection level in wild-type Mv1Lu (WT Mv) and in stable transfectants expressing tTA (Mv-tTA), GFP alone (Mv-GFP), or GFP with IBaM (Mv-IBaM/GFP) in the absence ( Tet) or presence (+ Tet) of 5 ng/ml tetracycline. Cells were incubated overnight with or without tetracycline before infection. Infections were stopped at 48 h. Each point represents the mean of triplicate assays ± SEM. At least 300 cells were counted per well. (D) Intracellular amastigotes in cells transfected with PTR5-DC/GFP or PTR5-IBaM/GFP and infected with or without 5 ng/ml tetracycline as described above. Cells were fixed 48 h after infection, stained with human Chagasic sera, and visualized with TRITC-labeled anti-human immunoglobulin G (Sigma, St. Louis, MO). Similar numbers of host cells were present in each field.

The PTR5-IBaM/GFP Mv1Lu cells were transiently transfected with pBXIIluc and assayed for stimulation of luciferase expression by trypomastigotes. In the absence of tetracycline, these cells express IBaM and are no longer able to activate NF-B-dependent luciferase expression in response to trypomastigotes (Figure 3B). Addition of 5 ng/ml tetracycline blocks synthesis of IBaM and restores the response to control level. Cells expressing tTA alone respond to trypomastigotes with NF-B activation and show similar levels of activation in the presence or absence of tetracycline, showing that tetracycline itself has no effect on NF-B activation (Figure 3B). The failure of cells expressing IBaM to activate NF-B was confirmed in nuclear translocation assays (our unpublished results). Thus IBaM expression specifically blocks the activation of NF-B by T. cruzi trypomastigotes.

When infected with T. cruzi, Mv1Lu cells expressing IBaM and GFP show a significant enhancement in infection levels compared with wild type and tTA- or GFP-expressing Mv1Lu cells (p < 0.05; Figure 3C). In addition to increasing the percentage of infected cells, the number of parasites in cells expressing IBaM is higher than that in cells expressing GFP alone (Figure 3D). Addition of tetracycline, which inhibits IBaM expression (Figure 3, C and D), blocks this enhancement. A direct effect of tetracycline on T. cruzi infection can be ruled out, because the antibiotic has no effect on infection level in wild-type Mv1Lu, PtTA-hygro-transfected, or GFP-expressing control cells (Figure 3, C and D). These results show that NF-B activation by parasites limits infection levels in epithelial cells.

Given that T. cruzi-induced NF-B activation restricts infection level in an epithelial cell line, the possibility exists that parasite-induced activation will restrict infection in other cell types as well. Indeed, trypomastigotes activate NF-B nuclear translocation in the murine endothelial line SVEC4-10 and in primary cultures of human fibroblasts (Table 1), as well as in several other human epithelial and endothelial cell lines (our unpublished results). Interestingly, T. cruzi do not activate NF-B in any of the muscle-derived cells tested, namely, the rat skeletal muscle myoblast line L₆E₉, the rat cardiac myoblast line H₉C₂, and primary cultures of smooth muscle cells of human or bovine origin. None of these cells shows any stimulation of p65 translocation on exposure to trypomastigotes (Table 1). In addition, T. cruzi fail to trigger NF-B promoter activity in two other myoblast lines, L₆E₉ and H₉C₂ (Figure 4). However, muscle cells do activate NF-B in response to TNF- and become less permissive to T. cruzi invasion (our unpublished results), indicating that NF-B is present and functional in these cells. These results indicate that NF-B activation by T. cruzi is highly dependent on cell type and is absent in muscle cells, the prime target for infection.

View this table: [in this window] [in a new window]   Table 1.  Correlation of parasite-induced activation of NF-B in different cell types with resistance to infection

View larger version (21K): [in this window] [in a new window]   Figure 4.   TNF- activates NF-B-dependent gene expression in cells that do not respond to T. cruzi. Mv1Lu, L6E9, and H9c2 cells were transiently transfected with pBXIIluc and then exposed to RPMI and 1% BSA, Silvio strain trypomastigotes (1 × 107/ml), or recombinant human TNF- (10 ng/ml). Cells were incubated 24 h, harvested, and assayed for luciferase activity as described in MATERIALS AND METHODS.

As predicted by the results with IBaM-expressing cells, muscle cells, which do not respond to trypomastigotes with NF-B activation, are much more susceptible to T. cruzi infection than those cells that do activate NF-B (Table 1). The in vitro infection pattern reflects that observed in vivo in most natural and experimental infections, in which infection of epithelial and endothelial cells is particularly rare and most parasites are found in muscle cells. Thus a strong correlation exists between the ability of cells to respond to parasites with NF-B activation and resistance to infection. This suggests that the susceptibility of myocytes to infection may be at least in part due to their failure to activate an antiparasitic pathway on exposure to T. cruzi.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

T. cruzi are able to invade almost any mammalian cell type, but infection levels vary widely both in vitro and in vivo. This aspect of tissue tropism is poorly understood. Parasite strain may contribute to the variation, whereas on the host side, differences in metabolism and receptor expression may play a role. The importance of host cell signaling in parasite tissue tropism has not been previously addressed. Some host cell signals conferring susceptibility, such as transforming growth factor- receptor activation, have been identified (Ming et al., 1995). Calcium transients, tyrosine phosphorylation, and mitogen-activated protein kinase activation have also been implicated (Rodriguez et al., 1995; Villalta et al., 1998). However, none of these are sufficient to explain the specificity of the interaction of the parasite with host cells. This work shows that T. cruzi trypomastigotes induce NF-B signaling in a cell-specific manner and that NF-B activation increases resistance to infection.

In T. cruzi infection, NF-B activation is clearly beneficial to the host cell, in contrast to the reported role of NF-B in other parasitic and bacterial infections. In Theileria parva infection, for example, NF-B is activated in infected T lymphocytes and specifically stimulates proliferation of infected cells, thus enhancing survival of the parasite (Ivanov et al., 1989). In Rickettsia rickettsii infection, activation of NF-B prevents apoptosis of the host cells, thereby protecting the infecting organism (Clifton et al., 1998). NF-B activation by T. cruzi protects host cells against infection and limits infection to a specific cell type, which is unable to activate this pathway in response to the parasite.

This work is the first indication that nonimmune cells can regulate intracellular parasitic infection independently of exogenously added immune effectors. However, it is still possible that the regulation of infection is secondary to production of cytokines or antimicrobial proteins by the infected cell. NF-B-dependent induction of iNOS may be an important component in the regulation of intracellular infection. In macrophages NO generated by iNOS is responsible for cytokine-dependent killing of T. cruzi (Munoz- Fernandez et al., 1992), and this enzyme is induced in cardiac fibroblasts exposed to trypomastigotes (Rottenberg et al., 1996). In this case, however, the authors suggested a positive role of nitric oxide production on parasite survival. Nevertheless, iNOS knockout mice are highly susceptible to T. cruzi infection (Holscher et al., 1998), and in most systems NO appears to be an important regulator of intracellular infection.

The activation of NF-B in cultured nonmuscle cells may be significant in vivo. NF-B-dependent expression of cytokines, chemokines, and adhesion molecules is likely to stimulate localized innate immune responses against the parasite (Baeuerle and Henkel, 1994; Kopp and Ghosh, 1995; Elewaut et al., 1999). This may explain the rapid clearance of T. cruzi from most nonmuscle tissues. By contrast, the failure of myocytes to respond to trypomastigotes with NF-B activation, as well as enhancing susceptibility of these cells to infection, would limit the innate immune response in these tissues. Control of infection in muscle tissue is therefore more heavily dependent on the acquired immune response. A consequence of the inability of muscle cells to limit intracellular infection by direct activation of NF-B is the requirement for continuous presence of infiltrating inflammatory cells in infected tissue to maintain host control of the infection, leading to the immune-driven pathology of Chagas disease.

The demonstration of a role for NF-B activation in resistance to infection by T. cruzi highlights the dynamic nature of the host-parasite interaction. Parasites can activate host cell signaling pathways that promote or restrict intracellular growth. The innate resistance of most cells to T. cruzi infection is in part due to the ability of these cells to recognize and respond to the invading organism. The failure of myocytes to respond to T. cruzi with NF-B activation may be one of the factors that allow the parasite to establish infection in muscle cells in the acute stage of Chagas disease and maintain that infection in the face of the acquired immune response during the chronic phase. Understanding of the innate mechanisms of parasite control may help the development of effective therapy for Chagas disease.