The Function of Gamma Interferon-Inducible GTP-Binding Protein IGTP in Host Resistance to Toxoplasma gondii Is Stat1 Dependent and Requires Expression in Both Hematopoietic and Nonhematopoietic Cellular Compartments

Abstract

IGTP is a member of the 47-kDa family of gamma interferon (IFN-γ)-induced GTPases. We have previously shown that IGTP is critical for host resistance to Toxoplasma gondii infection. In the present study, we demonstrate that T. gondii-induced IGTP expression in vivo and IFN-γ-driven synthesis of the protein in vitro are dependent on Stat1. Consistent with this observation, Stat1-deficient animals succumbed to T. gondii infection with the same rapid kinetics as IGTP^−/− mice. To ascertain the cellular levels at which IGTP functions in host control of acute infection, we constructed reciprocal bone marrow chimeras between IGTP-deficient and wild-type mice. Resistance to infection was observed only when IGTP was present in both hematopoietic and nonhematopoietic compartments. To assess the possible contribution of IGTP to the maintenance of parasite latency, partial chemotherapy was used to allow the establishment of chronic infection in IGTP-deficient animals. Upon cessation of drug treatment, these animals showed delayed mortality compared with similarly infected and treated IFN-γ-deficient or inducible nitric oxide synthase-deficient mice, which succumbed rapidly. Parallel experiments performed with drug-treated bone marrow chimeras supported a role for the hematopoietic compartment in this NO-dependent, IGTP-independent control of chronic infection. Taken together, our findings demonstrate that host resistance mediated by IGTP is a Stat1-induced function which in the case of T. gondii acts predominantly to restrict acute as opposed to chronic infection. This effector mechanism requires expression of IGTP in cells of both hematopoietic and nonhematopoietic origin. In contrast, in latent infection, hematopoietically derived cells mediate resistance by means of a largely NO-dependent pathway.

Toxoplasma gondii is an apicomplexan intracellular parasite that can promiscuously invade many mammalian cell types, leading to an acute phase characterized by a rapidly dividing tachyzoite stage. With the onset of the host immune response, the growth of tachyzoites in tissues is shut down, resulting in the emergence of encysted latent forms known as bradyzoites that maintain chronic infection primarily within the central nervous system (9, 10, 19).

Cell-mediated immune mechanisms play a major role in the control of T. gondii infection (8, 38). Gamma interferon (IFN-γ) is essential for acute as well as chronic resistance, and its expression in both stages is dependent on a vigorous interleukin 12 (IL-12) response stimulated by the parasite (11, 12, 17, 20, 30, 33, 37). Studies employing bone marrow chimeric mice have revealed that IFN-γ-mediated control of T. gondii growth requires cells of both hematopoietic and nonhematopoietic origins (39).

Although the critical involvement of the IFN-γ-IL-12 axis is well established in vivo, the nature of the IFN-γ-dependent effector mechanisms that limit the intracellular replication of the parasite is poorly understood, particularly during the acute phase of infection. While nitric oxide (NO) appears to play a role in controlling early parasite growth in perorally infected mice, it was also found to be detrimental to the host, leading to acute death of the animals (21). In contrast, in another model employing intraperitoneal inoculation with a different parasite strain, mice deficient in either inducible nitric oxide synthase (NOS2) or p47-phox survived acute infection, suggesting that neither NO nor reactive oxygen intermediates are needed at this initial stage (31). Nevertheless, in both of the experimental models studied, NO was found to play a critical role in the control of chronic infection.

A major clue concerning the intracellular pathway that leads to acute resistance to T. gondii has been obtained from recent studies with gene-targeted mice deficient for members of a 47-kDa family of IFN-γ-inducible GTP-binding proteins. This family includes at least six molecules, IGTP (35), LRG-47 (32), GTPI (1), IRG-47 (14), TGTP/Mg21 (3, 22), and IIGP (1), several of which have been demonstrated to have GTPase activity. Importantly, expression of these proteins is strongly upregulated by IFN-γ in both professional phagocytes and nonhematopoietic cells (1, 35, 36). It has previously been shown that mice lacking IGTP or LRG-47 are unable to control T. gondii infection, succumbing within 8 to 10 days after parasite inoculation despite the induction of a robust IL-12-IFN-γ response (5, 34). Therefore, IGTP and LRG-47 appear to play nonredundant roles in host defense against T. gondii and function downstream of IFN-γ in the signaling pathway of this cytokine. Interestingly, while IGTP-deficient mice are resistant to Listeria monocytogenes infection, LRG-47 knockout (KO) animals inoculated with the same infectious agent are highly susceptible, dying within 5 days after bacterial challenge. In contrast, mice lacking either IGTP or LRG-47 are resistant to murine cytomegalovirus infection. Thus, IGTP, LRG-47, and perhaps other members of this family may have pathogen-specific functions in IFN-γ-dependent host defense.

In the present study, we further analyzed the role of IGTP in host resistance to T. gondii infection by focusing on three important aspects of how this gene product regulates control of parasite growth. First, we explored the position of IGTP in the IFN-γ signaling cascade by examining the requirement for Stat1 in the expression of IGTP both in vivo and in vitro. Stat1 is a transcription factor activated as a consequence of IFN-γ receptor ligation (18). Although Stat1 has been shown to play a critical role in many IFN-γ-dependent host defense functions, recent work has indicated the presence of an IFN-γ-induced signaling pathway that is Stat1 independent (13, 28, 29). For this reason, it was of interest to determine whether the induction of IGTP by T. gondii requires Stat1 signaling and, in turn, whether Stat1 itself is essential for host resistance to the parasite. Second, we used reciprocal bone marrow chimeras constructed between IGTP-deficient and wild-type (WT) mice to determine whether, in common with IFN-γ receptor expression (39), IGTP must be present in cells derived from both hematopoietic and nonhematopoietic sources to achieve normal control of T. gondii infection. While IGTP is clearly important for host resistance during the acute stage, its role in the IFN-γ-dependent control of chronic infection is not clear. Third, we used partial chemotherapy to generate chronically infected IGTP-deficient mice and then tested their survival following drug removal. Our findings indicate that for normal resistance to T. gondii infection, IGTP expression must be induced in both hematopoietic and nonhematopoietic cells via a Stat1-dependent mechanism. In addition, our results suggest that while essential for acute resistance, IGTP plays only a partial role in the control of latent Toxoplasma infection.

MATERIALS AND METHODS

Experimental animals.

Stat1-deficient mice (25) and their WT counterparts with a 129/Sv/Ev background were purchased from Taconic Farms, Inc. (Germantown, N.Y.). IGTP-deficient (34) as well as WT control animals had a C57BL/6 × 129Sv genetic background. The mutant mouse strain with a disruption in the gene coding for the α-chain of IFN-γ receptor has been described previously (16) and had a 129 background. These mice were bred in our animal facility, housed in pathogen-free conditions, and sex and age matched for each experiment. IFN-γ (7) and NOS2 (24) KO animals with a C57BL/6 background were obtained from the National Institute of Allergy and Infectious Diseases Contract Facility at Taconic Farms, Inc. All mouse manipulations were done in accordance with procedures outlined in the Guide for the Care and Use of Laboratory Animals (26) under an animal study proposal approved by the National Institute of Allergy and Infectious Diseases Animal Care and Use Committee. In our model of intraperitoneal infection, B6/129 F1, 129, and C57BL/6 mice were equally resistant to the ME49 strain of T. gondii, surviving for at least 60 days following parasite inoculation (5, 31, 36, 39)

Preparation of bone marrow chimeric mice.

Reciprocal chimeras between IGTP KO and WT mice were generated as described previously (39). Briefly, recipient animals were lethally irradiated (900 rads) and reconstituted with 25 × 10⁶ bone marrow cells administered via the tail vein. Donor and recipient mice were sex matched. Chimeric mice were treated with Bactrim (sulfamethoxazole [150 mg/ml] and N-trimethoprim [30 mg/ml]), administered in their drinking water, for 5 weeks. The extent of hematopoietic cell reconstitution was assessed by flow cytometry using markers for macrophages and NK, T, and B cells. Animals were infected 8 to 12 weeks following the initial time of transplantation and monitored daily.

T. gondii infections.

Cysts of the avirulent T. gondii ME49 strain were used for infections and were prepared from brains of chronically infected C57BL/6 mice. Experimental animals received 20 cysts in 0.5 ml of phosphate-buffered saline via the intraperitoneal route and were monitored daily (30).

To assess the replication of the parasite during the acute phase, cytocentrifuge smears were prepared as described previously (30). Briefly, cells were harvested from mice at 5 days postinfection by injection of cold RPMI medium into the peritoneal cavity, and samples were prepared by concentrating 10⁵ cells with a Cytospin (Shandon Lipshaw, Pittsburgh, Pa.) for 5 min at 900 rpm. Slides were processed by using Diff-Quik fixative and staining solutions according to the instructions of the manufacturer (Dade Behring Inc., Newark, Del.). Approximately 500 cells were examined microscopically, using oil immersion (100× objective) to determine the percentage of infected cells.

Serum preparation and cytokine analysis.

Blood was collected from mice at the time of sacrifice in microcentrifuge tubes for serum separation according to the instructions of the manufacturer (Sarstedt, Newton, N.C.). Samples were allowed to clot for at least 30 min prior to centrifugation at 6,000 × g for 5 min. IFN-γ production was measured in serum samples by using sandwich enzyme-linked immunosorbent assay protocols as previously described (6).

Establishment of chronic infection by partial chemotherapy.

To allow the survival of IFN-γ-, IFN-γ receptor-, and IGTP-deficient mice through the acute phase of the infection, animals were given Bactrim in their drinking water beginning at 3 or 5 days after parasite inoculation (39). Based on previous experience, the day 3 time point was utilized for immunodeficient mice susceptible to acute infection, while day 5 was used for animals resistant to acute infection. Drug treatment was stopped at day 20 postinfection, and assessment of persistent infection was performed by examination of brain tissues from the mice for the presence of cysts and measurement of the ability of the cysts to induce mortality in IFN-γ KO animals. In some experiments, 2.5% aminoguanidine hemisulfate (AG) (Sigma, St. Louis, Mo.), an inhibitor of NO production, was added to sterilized drinking water at the time of drug removal (2, 23). The treatment was continued every day until the experiment was concluded. Mice were observed daily and mortality was scored.

Measurement of IGTP expression.

Total protein lysates were prepared from organs of infected and uninfected experimental animals as follows. Tissues were homogenized in ice-cold lysis buffer (1% NP-40, 0.15 M NaCl, 0.05 M Tris [pH 7.5], and protease inhibitors) by using an Omni-TH tissue homogenizer (Thomas Scientific, Swedesboro, N.J.) at medium speed. Protein concentration was determined with a bicinchoninic acid protein assay system (Pierce, Rockford, Ill.). Samples (25 μg of protein) were loaded onto 4 to 20% Tris-glycine gels (Invitrogen, Carlsbad, Calif.) and transferred to nitrocellulose (Schleicher & Schuell, Dassel, Germany). Proteins were detected with an anti-IGTP antibody as described previously (35) by using enhanced chemiluminescence (NEN Life Science Products, Inc., Boston, Mass.).

To measure IGTP levels in inflammatory macrophages, peritoneal cells (PC) were harvested from animals injected 5 days earlier with 3 ml of 3% thioglycolate (Sigma) and were plated in a 24-well plate at 2 × 10⁶ cells per well for 2 h in the presence or absence of 100 U of recombinant murine IFN-γ/ml (Genentech, Inc., San Francisco, Calif.). At this time, lipopolysaccharide (LPS) (Sigma) was added at a concentration of 10 ng/ml to the samples that received recombinant murine IFN-γ. After overnight incubation, cells were harvested and centrifuged and the pellet was resuspended in ice-cold lysis buffer. Protein lysates were processed as described above and analyzed by immunoblotting with an anti-IGTP antibody (35).

RESULTS

Stat1 is required for T. gondii- as well as IFN-γ-induced IGTP expression.

To address the role of Stat1 signaling in the induction of IGTP, Stat1 KO mice were inoculated with T. gondii cysts and IGTP levels in tissue homogenates were measured by Western blotting on day 5 postinfection. This time point was chosen because it is within the range of peak expression of IGTP during infection in WT animals (34). As shown in Fig. 1A, IGTP was readily detected in the spleens, livers, and lungs of WT animals. In contrast, IGTP expression was greatly reduced in Stat1-deficient animals. This reduction in IGTP was not due to an absence of IFN-γ production, since the sera of these infected animals contained high levels of the cytokine (1.19 ± 0.69 ng/ml [mean ± standard deviation] in Stat1 KO mice versus 2.32 ± 0.12 ng/ml in WT mice).

FIG. 1.

Requirement for Stat1 in IFN-γ-dependent IGTP induction and in vivo control of T. gondii infection. (A) Production of IGTP in tissues from T. gondii-infected Stat1-deficient animals and WT control animals. Protein lysates were prepared from the spleen, liver, and lungs of individual uninfected (−) and infected (for 5 days) (+) mice and analyzed for the presence of IGTP by Western blotting. The bands shown had an approximate molecular mass of 47 kDa as determined by reference to molecular mass markers and were not seen in control samples from infected IGTP-deficient animals (data not shown). The background bands observed in uninfected WT as well as Stat1 KO mice are likely to represent basal expression of IGTP, since in a previous study, no reactivity was observed when comparable Western blots were probed with the preimmune serum for this antibody (35). (B) Expression of IGTP in IFN-γ-LPS-stimulated inflammatory macrophages from Stat1-deficient and WT control animals. Thioglycolate-elicited PC from individual uninfected WT or Stat1-deficient animals were cultured for 18 h in the presence (+) or absence (−) of IFN-γ plus LPS. Protein lysates were then prepared and subjected to immunoblotting with anti-IGTP antibody as described above. (C) Survival of WT (129/Sv/Ev), IGTP-deficient, and Stat1-deficient mice (n = 5 animals per group) following infection with 20 cysts of T. gondii strain ME49. The insert shows the percentage (mean and standard deviation) of tachyzoite-infected PC harvested from WT (n = 3) or Stat1 KO (n = 5) mice on day 5 after T. gondii exposure. The experiment shown is representative of two performed with nearly identical results.

To confirm that Stat1 is a critical transcription factor for the induction of IGTP at the cellular level, we assessed IGTP expression in IFN-γ-LPS-stimulated inflammatory macrophages obtained from uninfected Stat1 KO and WT mice by immunoblotting. As shown in Fig. 1B, cells from the animals lacking Stat1 showed a near complete reduction in IGTP levels compared with their WT counterparts, indicating that Stat1 plays a central role in IFN-γ-dependent IGTP induction.

Consistent with their defective induction of IGTP following T. gondii infection, Stat1-deficient mice were extremely susceptible, succumbing 5 to 8 days after parasite challenge. The kinetics of mortality was similar to that of T. gondii-infected, IGTP-deficient mice (Fig. 1C) (34). To confirm that the death of the animals was associated with loss of control of the infection, we determined the percentage of PC infected with the parasite on day 5 following T. gondii exposure. In contrast to cells recovered from WT animals, PC from Stat1 KO mice were heavily infected with tachyzoites (Fig. 1C, insert).

Acute resistance to T. gondii infection requires IGTP expression in both hematopoietic and nonhematopoietic cellular compartments.

A previous study has demonstrated a crucial requirement for IFN-γ receptor expression in both hematopoietic and somatic cells for complete control of T. gondii infection (39). To assess the role of IGTP in each cellular compartment, we constructed reciprocal bone marrow chimeric mice by using IGTP KO and WT animals with a C57BL/6 × 129Sv background. The resulting animals were challenged with T. gondii. As shown in Fig. 2, while many of the WT mice sham reconstituted with WT bone marrow survived for over 30 days after T. gondii exposure, both WT bone marrow-reconstituted KO mice and KO bone marrow-reconstituted WT mice succumbed to acute infection with the same kinetics as that of KO-to-KO controls or nonchimeric KO animals. These observations establish that IGTP function is required in both hematopoietic and nonhematopoietic cells for control of T. gondii infection.

FIG. 2.

Survival of IGTP KO and WT reciprocal bone marrow chimeric mice following infection with T. gondii. Chimeric animals were constructed by reconstituting lethally irradiated WT or IGTP KO mice with bone marrow from either of the two strains. The animals were infected intraperitoneally with 20 cysts of T. gondii strain ME49 8 to 12 weeks after reconstitution (WT to KO, n = 6; KO to WT, n = 7; KO to KO, n = 6; WT to WT, n = 6). The results shown were reproduced in a second experiment.

Expression of IGTP is maintained during the chronic phase of T. gondii infection.

Although our study indicates that the presence of IGTP is critical for resistance during the acute phase of T. gondii infection, the role of this GTPase in maintaining control of the parasite during chronic infection is less clear. It has previously been shown that levels of IGTP increase in the livers, spleens, and PC of mice infected with T. gondii for 5 days (34). In contrast, IGTP induction was not observed in brain tissue examined at the same acute-phase time point (34). To determine whether IGTP is produced during the chronic stage, tissue homogenates from brains of WT mice were analyzed by Western blotting at days 14, 20, and 29 following parasite exposure. As shown in Fig. 3, IGTP was found to be strongly induced at all of the time points assayed, clearly establishing its expression during chronic infection.

FIG. 3.

IGTP is expressed in brain tissue during chronic T. gondii infection. Protein extracts were prepared from brain homogenates of WT mice before (d0) and on days 14 (d14), 20 (d20), and 29 (d29) following T. gondii infection. The samples were then analyzed by immunoblotting with an anti-IGTP antibody for the presence of IGTP. Each lane represents an individual animal.

IGTP plays a subordinate role relative to inducible nitric oxide synthase in controlling chronic T. gondii infection.

To determine whether IGTP is required for resistance in chronically infected mice, partial chemotherapy was used to allow the animals to survive through the acute phase and establish a persistent infection. As shown in Fig. 4, in contrast to animals lacking expression of the IFN-γ receptor, which died early after drug removal (after 6 to 15 days), IGTP KO mice survived for an extended period of time, dying after 17 to 59 days. This observation suggests that host resistance to chronic infection is partially dependent on IGTP.

FIG. 4.

Limited role of IGTP in control of latent T. gondii infection. WT (C57BL/6 × 129Sv), IFN-γ receptor-deficient, and IGTP-deficient mice were infected intraperitoneally with 20 cysts of T. gondii strain ME49. Bactrim-containing water was administered daily beginning on day 3 (IFN-γ receptor-deficient and IGTP KO mice) or day 5 (WT animals) (Rx start) and withdrawn on day 20 (Rx stop) postinfection. Survival of the animals (WT, n = 5; IFN-γ receptor [R] KO, n = 6; IGTP KO, n = 12) was then monitored. The results shown are representative of four experiments performed.

Role for NO in mediating IGTP-independent control of latent T. gondii infection.

In mice infected by intraperitoneal inoculation, NO has been proposed to play a pivotal role during the chronic but not the acute phase of T. gondii infection (31). Indeed, NOS2-deficient animals succumb 3 to 4 weeks postinfection, and their mortality is associated with increased cyst burdens (31). To determine whether NOS2-dependent effector mechanisms are responsible for the residual survival observed in the drug-treated IGTP-deficient mice, AG was administered to the drinking water when the Bactrim therapy was stopped. As shown in Fig. 5, treatment with AG led to decreased survival of IGTP KO mice, with the animals dying 10 to 31 days after antibiotic withdrawal. The kinetics of mortality of AG-treated IGTP KO animals resembled that of the NOS2-deficient mice that had received Bactrim, most of which died after 10 to 35 days. In contrast, as observed previously, animals lacking IFN-γ receptor (or IFN-γ) died soon after drug removal (exhibiting 100% mortality by day 12). These observations suggest that NO is at least partially responsible for the prolonged survival of antibiotic-treated mice lacking IGTP. However, the finding that both NOS2-deficient animals and AG-treated IGTP KO mice survive longer than their IFN-γ receptor-deficient (or IFN-γ-deficient) counterparts suggests that other effector mechanisms in addition to those dependent on IGTP and NO play an important role in the control of established infections. Unexpectedly, AG-treated WT mice did not succumb to the parasite at the same rate as did animals lacking NOS2, perhaps due to the failure of the compound to induce total depletion of inducible nitric oxide synthase.

FIG. 5.

The IGTP-independent functions responsible for control of chronic T. gondii infection are in part NOS2-dependent effector mechanisms. WT (C57BL/6 × 129Sv) and IGTP KO, IFN-γ KO, and NOS2 KO mice were infected intraperitoneally with 20 cysts of T. gondii strain ME49. Bactrim-containing water was supplied continuously beginning on day 3 (IFN-γ KO and IGTP KO mice) or day 5 (WT and NOS2 KO animals) (Rx start) and withdrawn on day 20 (Rx stop) postinfection. Beginning at day 20, the WT and IGTP KO animals received AG daily in their drinking water. Survival of the animals (WT, n = 12; WT plus AG, n = 11; IGTP KO, n = 14; IGTP KO plus AG, n = 10; IFN-γ KO, n = 8; NOS2 KO, n = 14) was then monitored. Data shown are representative of three experiments performed.

A previous study has demonstrated that the presence of NOS2 in the hematopoietic compartment is sufficient to confer control of chronic T. gondii infection (39). To address whether hematopoietically derived cells acting in an NOS2-dependent fashion are responsible for the extended survival of Bactrim-treated IGTP KO mice, we examined the role of NOS2 in the resistance of similarly treated chimeric mice constructed by lethally irradiating WT animals and reconstituting them with IGTP-deficient bone marrow. As shown in Fig. 6, the majority (83%) of these KO-to-WT chimeric animals survived for over 30 days following antibiotic withdrawal. Importantly, when the same chimeras were treated with AG, they displayed enhanced susceptibility, dying 19 to 22 days after cessation of antibiotic treatment. The mortality curve closely resembled that of Bactrim-treated NOS2-deficient mice. These results support the hypothesis that the NOS2-dependent, IGTP-independent effector mechanism controlling chronic infection is a function of the hematopoietic cellular compartment.

FIG. 6.

Role of cells of hematopoietic origin in the NO-dependent, IGTP-independent control of latent T. gondii infection. Bone marrow chimeras were constructed by reconstituting lethally irradiated WT animals with bone marrow from either WT or IGTP KO mice. The resulting animals were injected intraperitoneally with 20 cysts of T. gondii strain ME49 8 to 12 weeks after reconstitution. Bactrim-containing water was given daily starting on day 3 (IGTP KO-to-WT and IFN-γ KO mice) or day 5 (WT-to-WT and NOS2 KO animals) (Rx start), and antibiotic treatment was stopped at day 20 (Rx stop) postinfection. Beginning at day 20, AG was administered daily to the drinking water of the IGTP KO-to-WT chimeric mice. Cumulative mortality of the mice (WT to WT, n = 7; IGTP KO to WT plus AG, n = 7; IGTP KO to WT, n = 6; IFN-γ KO, n = 8; NOS2 KO, n = 8) was recorded daily.

DISCUSSION

Since T. gondii invades a wide variety of nucleated cell types, the IFN-γ-dependent mechanism by which the growth of the parasite is controlled must operate in other cells in addition to professional phagocytes. This idea has been underscored by a previous study employing reciprocal bone marrow chimeras that established that acute resistance to T. gondii occurs only when IFN-γ receptors are expressed in cells of both hematopoietic and nonhematopoietic origins (39). In the present study, we utilized a similar approach to determine whether IGTP must also be present in both lineages for resistance to be generated. The answer to this question was not obvious, since reciprocal chimera experiments with NOS2-deficient mice indicated a requirement for only the hematopoietic compartment in the control of chronic infection (39). Our finding that acute resistance depends on IGTP expression in both hematopoietic and nonhematopoietic lineages is consistent with the known IFN-γ-dependent production of IGTP by phagocytic as well as nonphagocytic cells (36) and, more importantly, supports a direct role for this GTPase in the IFN-γ receptor-triggered mechanism that leads to intracellular control of the parasite in multiple cell types and tissues.

Because IGTP-deficient mice succumb during the acute phase, IGTP is clearly important for control of T. gondii during this early stage following intraperitoneal infection (34). To evaluate the role of IGTP in resistance to chronic infection, it was necessary to treat infected IGTP KO mice with drugs (Bactrim) to ensure their survival during the acute phase and then assess their ability to control the parasite following removal of the antibiotic. The results of this analysis, which was performed as a comparison study with control groups consisting of IFN-γ-deficient, IFN-γ receptor-deficient, NOS2-deficient, and AG-treated IGTP-deficient animals, revealed that IGTP plays a subordinate role in controlling chronic T. gondii infection and support the results of previous studies indicating a primary contribution for NOS2-dependent effector functions at this stage (21, 31). Nevertheless, since NOS2-deficient animals were significantly more resistant than IFN-γ KO mice when subjected to the same drug treatment and withdrawal protocol, it is probable that other as-yet-unidentified IFN-γ-dependent mechanisms participate with NOS2 in maintaining resistance to chronic infection.

While it has been difficult to demonstrate defective IFN-γ-dependent control of T. gondii in IGTP-deficient fibroblast and macrophage cultures, a recent study has indicated that astrocytes from IGTP KO mice, in contrast to those obtained from WT animals, fail to limit parasite growth following IFN-γ treatment (15). As a cell associated with the central nervous system, the astrocyte should be infected by T. gondii primarily during the chronic stage. Thus, the above-mentioned observation would seem to be at odds with our conclusion that IGTP plays a limited role in IFN-γ-dependent control of chronic infection. This apparent discrepancy may reflect the differential effects of IGTP in vivo versus in vitro or, alternatively, point to the astrocyte as the key effector cell responsible for limiting infection beyond the period when IGTP KO mice succumb.

Stat1 is a critical component in the signaling pathway triggered by IFN-γ receptor ligation and has been shown to be essential for host resistance to a number of different viral and bacterial pathogens (18). Nevertheless, recent studies have documented a pathway of IFN-γ receptor signaling that does not depend on this transcription factor (13, 28, 29). For this reason, we examined the role of Stat1 in both IGTP induction and the IFN-γ-dependent control of T. gondii infection. Our results clearly establish that the upregulation of IGTP expression in tissues which is triggered by T. gondii is Stat1 dependent and, consistent with this observation, that Stat1-deficient animals show the same loss of resistance to infection that is displayed by IGTP KO mice. To the best of our knowledge, the latter finding is the first in vivo demonstration of the function of Stat1 in host resistance to a protozoan pathogen, although a role for this transcription factor in cell-mediated immunity to T. gondii was previously hypothesized on the basis of in vitro studies in which human fibroblast lines with chemically induced mutations in Stat1 were found to be defective in IFN-γ-dependent control of intracellular infection (4).

A generalized requirement for Stat1 in the IFN-γ-dependent IGTP induction was suggested by the failure of thioglycolate-elicited macrophages from Stat1-deficient mice to upregulate IGTP protein expression upon stimulation with IFN-γ-LPS. Such a requirement is also supported by a recent gene array study performed on IFN-γ-treated bone marrow-derived macrophages from Stat1 KO mice in which defective induction of IGTP gene expression was apparent but not discussed (13). Interestingly, in the same study, Stat1-deficient macrophages also displayed impaired upregulation of the genes encoding the IGTP-related proteins LRG-47 and IRG-47. The latter observation suggests that expression of all of the members of the 47-kDa family of GTPases may be Stat1 regulated.

Although Stat1 can now be included as a critical component in the IGTP-dependent pathway of host resistance to T. gondii, the final mechanism by which this GTPase restricts parasite growth remains unknown. As noted previously, IGTP deficiency does not affect IFN-γ-dependent induction of NO, which argues against a role for the latter effector molecule in the mechanism of host resistance (34). A second IFN-γ-dependent pathway previously implicated in the restriction of T. gondii growth is tryptophan starvation induced by activation of indoleamine 2,3-dioxygenase (IDO) (27). Nevertheless, in recent experiments, we observed normal induction of IDO in tissues of IGTP KO mice following T. gondii infection (unpublished observations). Thus, IGTP would seem to function by a novel mechanism that is yet to be elucidated. It has previously been proposed that IGTP, as an endoplasmic reticulum-associated GTPase (36), may be a critical regulator of trafficking to parasite-containing vacuoles, which in turn could affect pathogen survival. Interestingly, the recent observation that LRG-47-deficient mice are also highly susceptible to T. gondii infection (5) indicates that this GTP-binding protein plays an equally important role in the IFN-γ-dependent mechanism that controls parasite growth and therefore must act in concert with IGTP. We are currently determining the localization of these two critical 47-kDa GTPases in IFN-γ-stimulated, T. gondii-infected cells in an attempt to gain new information about their function in relation to the intracellular milieu of this pathogen.

In conclusion, the findings reported here strengthen the proposed association between IGTP gene function and the mechanism of IFN-γ-dependent resistance to T. gondii infection by linking both processes to the Stat1 signaling pathway and to the dual requirement for hematopoietic and nonhematopoietic cells in the restriction of parasite growth. In addition, our results reveal a dichotomy in the respective roles of IGTP and NO during acute versus chronic infection with this protozoan pathogen.