Discovery of a novel T. gondii conoid-associated protein important for parasite resistance to reactive nitrogen intermediates

Abstract

T. gondii modifies its host cell to suppress its ability to become activated in response to IFN-γ and TNF-α and to develop intracellular antimicrobial effectors including nitric oxide. Mechanisms used by of T. gondii to modulate activation of its infected host cell likely underlies its ability to hijack monocytes and dendritic cells (DC) during infection to disseminate to the brain and CNS where it converts to bradyzoites contained in tissue cysts to establish persistent infection. To identify T. gondii genes important for resistance to the effects of host cell activation we developed an in vitro murine macrophage infection and activation model to identify parasite insertional mutants that have a fitness defect in infected macrophages following activation but normal invasion and replication in naïve macrophages. We identified fourteen independent T. gondii insertional mutants out of over 8000 screened that share a defect in their ability to survive macrophage activation due to macrophage production of reactive nitrogen intermediates (RNIs). These mutants have been designated counter-immune (CIM) mutants. We successfully used one of these mutants to identify a T. gondii cytoplasmic and conoid-associated protein important for parasite resistance to macrophage RNIs. Deletion of the entire gene or just the region encoding the protein in wild type parasites recapitulated the RNI-resistance defect in the CIM mutant confirming the role of the protein in resistance to macrophage RNI.

Introduction

Toxoplasma gondii is an obligate intracellular parasite that is the causative agent of toxoplasmosis. Human infection with T. gondii is largely due to the ingestion of food or water contaminated with oocysts shed from infected cats or bradyzoites in tissue cysts ingested from undercooked meat. Following ingestion of oocysts or cysts, sporozoites or bradyzoites respectively are released, penetrate the intestinal epithelium, convert to tachyzoites, invade cells and ultimately disseminate systemically. An IFN-γ-dependent cell mediated immune response eventually controls parasite replication and eliminates the majority of parasites but a subpopulation of parasites escape the host immune response and convert to a slower growing bradyzoite developmental stage of the parasite that is sequestered in tissue cysts primarily in the brain, CNS and muscle tissue where the parasites persist indefinitely. Bradyzoites in these cysts can convert back to tachyzoites if an individual becomes immune suppressed leading to serious disease.

During initial infection with T. gondii there is a race between the parasite and the host immune response pitting parasite replication and dissemination against full host cell activation sufficient to eliminate intracellular parasites. In this battle the macrophage acts as both a host cell for T. gondii during infection aiding dissemination to the brain and CNS (1) and also as an effector cell that when properly activated is capable of eliminating parasites. In terms of host immunity, IFN-γ is the key mediator of protection during both the acute (2) and chronic stage of T. gondii infection (2). It stimulates IFN-γ dependent antimicrobial effectors including production of reactive nitrogen intermediates (3) and IFN-γ inducible GTPases (4) (5) that aid in controlling parasites. Aiding parasite survival, T. gondii is able to suppress and activate diverse host cell pathways that impact the ability of the host cell to be activated by IFN-γ and TNF-α (6) and likely additional stimuli (7, 8). The parasite protects itself within infected cells in part by interfering with intracellular signaling downstream of activation stimuli including the action of transcription factors STAT1 (9), NFκB (10) and hypoxia inducible factor-1 (HIF) (11). Consequently, if T. gondii infects phagocytes prior to their activation by IFN-γ induced by the host immune response, the parasites can remain capable of replication even in the face of an ongoing immune response. T. gondii has been shown to hijack monocytes and dendritic cells (DCs) and use them as a relatively safe haven for dissemination even in the face of an ongoing immune response that activates adjacent non-infected cells (1). T. gondii even induces motility of infected DCs to aid parasite dissemination (12, 13). In contrast, if the parasite egresses from its safe intracellular niche and invades adjacent cells that are fully activated, the parasite is generally killed. This provides a balance during infection that allows the parasite time to replicate and disseminate to enable establishment of persistent infection but also ensures eventual parasite control by the host immune response to prevent the death of the host.

T. gondii can invade virtually any warm blooded vertebrate host cell by an active process dependent on the actin myosin motor of the parasite and independent of phagocytosis (14, 15). During invasion it forms a unique vacuole, the parasitophorous vacuole or PV, where the parasite replicates. Parasite invasion and PV formation requires the regulated secretion of products from secretory organelles at the apical end of the parasite that includes micronemes, rhoptries and dense granules involved in adhesion, invasion, PV formation and modification of the PV (16). The conoid is a truncated cone shaped structure present at the extreme apical end of the parasite. This is a retractable structure that seems to act like a probe during invasion and egress although its precise functions are not known. The PV is largely segregated from the endocytic and exocytic system of the host cell and this segregation is established during the invasion process and does not appear to require subsequent action of the parasite to maintain its segregation (15, 17, 18). In fact, the known parasite effects on host cell signaling are due to products secreted during the process of adhesion, invasion and PV formation rather than a consequence of constitutive secretion of products through the established PV into the host cell cytosol (19–21). Parasite intracellular survival requires active invasion as antibody-opsonized parasites are killed if taken up by phagocytosis (15, 18, 22).

Although T. gondii is known to dramatically modify the cell it infects, the parasite genes responsible for host cell modifications remain largely unknown. It is also likely that T. gondii, like other intracellular pathogens, has mechanisms that enable it to subvert or repair damage from intracellular antimicrobial mediators that are produced within infected cells. The present study combined random insertional mutagenesis of T. gondii along with a model of parasite invasion followed by macrophage activation to identify novel T. gondii genes that enable the parasite to survive the affects of macrophage activation. The study resulted in the isolation of a panel of parasite mutants that all share an increased susceptibility to macrophage reactive nitrogen intermediates (RNI) induced upon activation. The defect was associated with macrophage activation as none of the mutants were impaired for survival/replication in naïve macrophages. One of these mutants was used to identify a novel conoid-associated protein in T. gondii that contributes directly or indirectly to the parasites ability to withstand macrophage RNI.

Materials and Methods

Parasites and cell culture

The Prugniaud strain of T. gondii deleted in hypoxanthine-xanthine-guanosine phosphoribosyltransferase (ΔHPT) (Kind gift from Laura Knoll, University of Wisconsin-Madison) was the base strain used to generate mutants and was used for all of the studies except for endogenous tagging of proteins. Endogenous tagging of proteins was done in the RH ΔKu80ΔHPT strain (kind gift from Dr Vern Carruthers, University of Michigan). The deletion of the Ku80 gene involved in DNA repair via the NHEJ pathway, results in greater levels of homologous recombination allowing for incorporation of reporter proteins into the 3′ end of the endogenous genes (23, 24). Parasite strains were maintained in monolayers of human foreskin fibroblast cells (HFF) (ATCC) in DMEM supplemented with 10% FCS, 2mM L-glutamine, 100 U/ml penicillin and 100μg/ml stremptomycin (referred to as D10 media). Cells were cultured in a humidified 5% CO2 incubator at 37°C. All strains including T. gondii mutants were tested for mycoplasma contamination using MycoAlert Mycoplasma Detection Kit (Lonza Biologics Inc) and remained negative throughout the experiments. Bone marrow-derived macrophages were isolated from C57BL6 mice (Jackson Laboroatory, Bar Harbor, ME) and cultured as previously described (15). iNOS ^−/− and gp91 phox ^−/− mice and their control C57/BL6 background mice were obtained from Jackson Laboratories. Mice with a deletion of both Irgm1 and Irgm3 and their C57/BL6 controls were obtained from Dr. Gregory Taylor, Duke University (25). Macrophages were used for experiments between seven and ten days following isolation.

Insertional mutagenesis

Two independent T. gondii insertional mutant libraries were generated. One library of approximately 6000 mutants is published and was created as a signature tagged mutant library (26). The mutants 2B6, A40E4, 37B2, 14F9, 1BD11, 11BE9, 68C6 and the previously published mutant, 89B7, used for comparison were all identified from this library (26). An additional 2000 insertional mutants were generated in a similar manner but directly from Prugniaud ΔHPT to avoid insertion of a DNA signature tag in addition to the random insertional mutation. Mutants I02B6, I03D5, I01B3, I01B1, I01A5, A01B1 and I01A1 were all isolated from this second library.

Analysis of parasite replication in macrophages

Mutants were first screened in 96-well plates containing bone marrow-derived macrophages. Equal volumes of parasites were transferred from 96-well plates containing HFF cells to 96-well plates containing macrophages. Parasites were allowed to infect macrophages for 4 hours and then plates were rinsed with D10 media and 100 ng/ml LPS (E-coli, List Biologicals, Campbell, CA) and 100 U/ml IFN-γ (Peprotech, Rocky Hill, NJ) to remove extracellular parasites and then cultured for an additional 24 hours. 96 well plates were then fixed with 4% formaldehyde (EM Scientific) for 20 minutes, permeabilized with 0.02% triton x-100, blocked in 10% goat serum in PBS and then stained with rabbit polyclonal anti-Toxoplasma antibody (Fitzgerald Scientific) followed by Alexa 488 goat anti-rabbit antibody (Invitrogen). Mutants that predominantly failed to replicate beyond one parasite per vacuole were selected for further analysis. For subsequent screens macrophages were grown on glass chamber slides (BD Biosciences, San Jose, CA) overnight at a concentration of 2 × 10⁵ cells per ml in 250 μl of D10 media. Macrophages were challenged with 2 × 10⁴ T. gondii insertional mutants for four hours, rinsed with fresh medium to remove nonadherent parasites, and either left unstimulated or activated with 100 ng/ml lipopolysaccharide and 100 U/ml IFN-γ for an additional 24 h. Aminoguanidine (Acros Organics) was used at a concentration of 1mM to inhibit nitric oxide production concurrent with activation. In our studies we estimated nitric oxide production in our activated macrophage cultures by the Griess reaction that detects nitrite and nitrate, two stable downstream products of nitric oxide. Macrophage activation in the presence and absence of parasites resulted in approximately 50 μM of nitrite/nitrate in total after 24 hours. To evaluate the susceptibility of the CIM mutants to nitrosative stress in the absence of macrophage activation, the nitric oxide donors sodium nitroprusside (SNP) (100 μM) (Ricca Chemical Company) or DETA NONOate (100 μM) (ENZO Lab Sciences) was used in place of IFN-γ and LPS. DETA NONOate (100 μM) resulted in approximately 30 μM of total nitrite/nitrate in our cultures after 24 hours. Cells were stained as described above and analyzed by phase contrast and fluorescence microscopy. Samples were mounted using VectaShield mounting media with DAPI (4′,6-diamidino-2-phenylindole; Vector Laboratories, Burlingame, CA). Phase-contrast and fluorescence microscopy was used to examine parasite and PV morphology and to determine the number of parasites per vacuole. At least 100 parasite-containing vacuoles were examined twice and the mean ± standard deviation (SD) was derived from these. Mutants were selected if they were significantly impaired for replication/survival in three rounds of these screens. All immunofluorescence samples were visualized using a Zeiss inverted Axiovert 200 motorized microscope with a 100x objective (PlanApo 1.4-numerical-aperture oil PH3 objective); Zeiss filter sets 31, 34, 38, and 50; and Axiovision 4.3 software. Pictures were obtained using a Zeiss Axiocam MRm.

Immunofluorescence analysis

For immunofluorescence, cells were fixed and permeabilized as describe above prior to staining. To evaluate whether vacuoles fused with lysosomes, cells were stained 24 h after activation with a rat anti-mouse monoclonal antibody to lysosome-associated membrane protein (LAMP1) (ID4B; Developmental Studies Hybridoma Bank, University of Iowa) and a rabbit polyclonal antibody to T. gondii. In experiments analyzing expression of inducible nitric oxide synthase (iNOS) cells were costained with a rabbit polyclonal antibody to T. gondii and an antibody to murine iNOS (BD BioSciences). Expression of the endogenously YFP tagged TGME49_010810 protein in HFF and naïve macrophage cells was analyzed by immunoflourescence at distinct stages of the parasite’s lytic cycle including extracellular and 1, 24 and 36 hours post infection. Expression on recently egressed parasites was also examined at 36 hours post-infection. To confirm localization of the TGME49_010810-YFP tagged protein to the apical end of the parasite, cells were stained with the mAb 7E8 to ISP1 protein associated with the parasite inner membrane complex (complements of Peter Bradley UCLA, CA) (27).

iNOS quantification

For quantitative PCR and protein analysis of murine iNOS, macrophages were cultured in bacteriological petri dishes and challenged with WT or mutant parasites at a 1:1 ratio of parasites to macrophages for four hours. Macrophages were stimulated with 100 ng/ml LPS and 100 U/ml IFN-γ or left unstimulated. RNA was isolated from infected and noninfected macrophages for iNOS qRT-PCR 16 hours after activation. RNA was isolated using the Qiagen RNEasy plus protocol as per manufacturer’s instructions. Invitrogen Superscript was used to generate cDNA. For quantitative real time PCR cDNA was amplified using the Faststart Universal SYBYR Green Master mix (ROX) (Roche Diagnostics, IN) using an ABI-7900 PCR system. iNOS transcript was normalized against actin. For Western blot analysis of iNOS macrophages were isolated 24 hours after activation. Equal numbers of infected and noninfected macrophages were lysed in RIPA buffer and supernatant containing proteins analyzed by Western blot analysis using antibodies to iNOS (BD Laboratories) and actin (Santa Cruz Biotechnologies). Western blots were visualized using alkaline phosphatase-conjugated secondary antibodies and the Invitrogen Western Breeze detection system.

Analysis of genomic DNA flanking the plasmid insertion site

Genomic T. gondii DNA was prepared as previously described (28) or with the DNEasy Blood and Tissue Kit from Qiagen and digested with HindIII or NcoI restriction enzymes to cut once within the insertional plasmid and unknown sites in adjacent genomic DNA. After digestion, the DNA was ligated, transformed into GC10 Thunderbolt Electrocompetent bacteria (Gene Choice, Frederick, MD), and selected for kanamycin resistance. The DNA flanking the insertion site was sequenced and compared with the T. gondii genome database (http://ToxoDB.org) (29). The insertion site was confirmed by PCR using primers to the insertional plasmid and flanking genomic DNA.

TGME49_010810 mRNA analysis

Total RNA was isolated using RNEasy Plus Mini Kit (Qiagen) per manufacturers instructions. cDNA was synthesized from Prugniaud (ΔHPT) strain of T. gondii using SuperScript III First-Strand Synthesis Kit (Invitrogen). The primers used for amplifying cDNA were primer set 1: 5′-CAC GTT GAA AGT CTG CAG GAA AGC-3′ and 5′-AAC ATT TGGA GTC ACA GCG ACA GC-3′; primer set 2: 5′-GCT GTC GCT GTG ACT CCA AATGTT-3′ and 5′-CCA TGT TCC ACC CGT GTT CTT CTT-3′. The 5′ and 3′ ends of the cDNA were determined using First Choice RLM-RACE Kit (Ambion) and Accuprime Pfx DNA polymerase (Invitrogen) according to manufacturer’s instructions. The primers used for the 5′ RACE were: 5′-ACC CAG TCC GAT GTA CCC TTG AAA-3′ and 5′-TGT CTA CGC GAA AGA CTC CGT TCT-3′. The primers used for the 3′ RACE were 5′-TCC GAG TTG AGT GTG TGG TTG GTT-3′ and 5′-TCT CCT TCT TTG AGC AGG CCA TGT-3′. All PCR products were cloned into the pCR2.1-TOPO cloning vector and sequenced at Genewiz, NJ using the M13 primer sites present in the pCR® II-TOPO® plasmid. Conceptual translation of the identified cDNA was done using Translate tool at ExPASy (Expert Protein Analysis System) proteomics server to identify the encoded protein(s). Mass spectrometry peptide data available for TGME49_010810 encoded protein was used to confirm the identified protein (29, 30).

Gene deletion of TGME49_010810

Gene deletion was accomplished in two different ways. The first method recreated the original plasmid insertion event in the mutant 2B6 in wild type parasites to confirm that the identified plasmid insertion was responsible for the impaired phenotype of the mutant following macrophage activation. These are referred to as 2B6 recap clones. Restriction enzyme digestion (PvuII) was used to isolate a construct containing the intact insertional plasmid and the adjacent genomic flanking regions from the 2B6 mutant. Digested DNA was ligated using T4 DNA Ligase (NEB), ethanol precipitated, and electroporated into GC10 Thunderbolt E. coli (Gene Choice, Frederick, MD) according to the manufacturer’s instructions. 50μg of this construct linearized with Bsu361 was electroporated into Prugnaud ΔHPT Tg and selected for HPT using 50 ug/ml mycophenolic acid and 50 ug/ml xanthine (MPA/XAN). The population was cloned using limiting dilution to obtain single clones. PCR was performed on the individual clones using primers to generate a PCR product which spanned the insertion site: 5′-TTC GAC GCT CTT TGA CTC CAC CTT -3′ and 5′-TAC GAA TTG GCG TGC ATT TGG ACC-3′. Generation of the recap clones resulted in a functional deletion of the gene TGME49_010810 in the 2B6 recap clones as the entire cDNA was eliminated. Loss of cDNA was confirmed using primers 5′-GCT GTC GCT GTG ACT CCA AAT GTT-3′ and 5′-AAA TGT GCG CAT GCC ATC ACT GTC-3′. Amplification of tubulin beta chain cDNA was used as a control using primers 5′-TGC CTT CAG GGT TTC CAG ATC ACT-3′ and 5′-GCA GAA TGC GTA ACA AGG CGT TCA-3′

Deletion of just the protein coding region of the gene TGME49_010810 was also performed to ensure the protein itself and not its 5′ UTR was responsible for the phenotype. To generate the knockout construct, adjoining DNA flanking the gene TGME49_010810 were cloned into vector pMini-GFP.ht (Kind gift from Dr Gustavo Arrizabalaga, University of Idaho). Flank 1 and 2 of the construct was amplified from T. gondii using primers 5′-GGG AAA GGG AAA GCG GCC GCA TAC TTT CAC GCG AAT GCC TGC TC-3′ and 5′-GGA AGG TCT AGA AAT ATG CAG CGT TGG AAC TGA CGG-3′; 5′-TAC CGT CGA CCT CGA GGA AGC GCA AGA AAG TGT CAG GTT -3′ and 5′-ACA AAA GCT GGG TAC TCT CAG GCA TCC AAC ACA ATC CGA-3′ respectively. Flank 1 PCR product was digested with Not1 and Xba1 and cloned into pMini-GFP.ht digested with the same enzymes using T4 ligase (New England Biolabs). The resulting plasmid was digested with Acc651 and XhoI and Flank 2 was cloned in using In-Fusion® PCR cloning system (Clontech) according to the manufacturer’s instructions. 50 μg of the final targeting vector (p TGME49_010810KO) was linearized with NotI and electroporated into Pru ΔHXPRT T. gondii strain. Six to seven independent electroporations were done. Stable clones carrying the transfected plasmid were selected using MPA/XAN. Transformants that lost GFP expression due to a double crossover event leading to stable transformation were selected by fluorescence microscopy using an inverted Zeiss Axiovert 200 microscope (Carl Zeiss Inc, NY). Gene specific crossover as opposed to random integration of the plasmid was confirmed with PCR using primers 5′-GTG GCG ATT CTC ATC GAC TT-3′ and 5′-AAC GAC AGT TTG TAC TTG CGA GGC-3′ specific for HPT gene and for the T. gondii genome sequence adjacent to the 3′ region of the gene TGME49_010810 present in the targeting construct used to generate the gene deletion. Gene deletion in identified clones was confirmed by analyzing the TGME49_010810 transcript level using primers 5′-GCT GTC GCT GTG ACT CCA AAT GTT-3′ and 5′-AAA TGT GCG CAT GCC ATC ACT GTC -3′. Tubulin beta chain cDNA was amplified and used as the control.

Western blot analysis of TGME49_010810

Recombinant 6xHIS tagged TGME49_010810 protein was created to generate a polyclonal antibody. The translated DNA sequence of the TGME49_010810 gene was amplified using primers 5′-GGG CAG CAT ATG GTG AAG CTT CGG TGG AA-3′ and 5′-GAT CCG GAA TTC TTA TTC TTG TCC ATT GCT GCG-3′. The primers were designed to include restriction sites for enzymes NdeI in the forward primer and EcoRI in the reverse primer. The PCR products were digested with enzymes NdeI and EcoRI and ligated into pET28 (a) (Novagen) vector digested with the same enzymes using T4 DNA Ligase (New England Biolabs). This vector was electroporated into BL21(DE3) E. coli (Novagen). Expression of the recombinant protein which now encodes a His tag at the N′ terminus was induced using MagicMedia^™ (Invitrogen, Carlsbad, CA) according to manufacturer’s instructions. Transformed bacteria were lysed, bound to Ni-NTA Superflow Columns (Qiagen), and eluted under denaturing conditions. Purified proteins were inoculated into rabbits to produce polyclonal antisera (Cocalico Biologicals, Inc. PA). For western blots, lysates from 10⁷ parasites each was loaded per well and separated on a 8–16 % Tris Tricine gel (Biorad) and transferred to PVDF membrane (Invitrogen) using Tris-CAPS buffer. Rabbit anti-TGME49_010810 serum was used at a dilution of 1:200. Specificity was evaluated by comparing the antibody against WT and the mutant 2B6 and analysis of pre-serum. SAG1 expression was analyzed as a loading control by using mouse mAB DG52 (1:10 dilution). Secondary anti rabbit and mouse conjugated to alkaline phosphatase were used as provided in the Western Breeze kit (Invitrogen) and the blots were developed using the same kit as per manufacturer’s instructions.

Endogenous tagging of TGME49_010810 with YFP

Gene TGME49_010810 was endogenously tagged with the YFP gene to enable expression and trafficking studies using immunofluorescence microscopy as the peptide antibody did not detect the protein using a variety of IFA protocols. A 1.5 kb region at the 3′ end of the gene TGME49_010810 was amplified up to the gene’s stop codon using primers 5′-TAC TTC CAA TCC AAT TTA GCA CGG CAT GGT CGA TAA ACC GAG AA-3′ and 5′-TCC TCC ACT TCC AAT TTT AGC TTC TTG TCC ATT GCT GCG CAG AAC TTC C-3′ and inserted into Paf digested vector pyfp-lic-hpt (24)(kind gift from Vern Carruthers, University of Michigan) using In-Fusion® PCR cloning system in frame with the YFP gene. The resulting construct was linearized with EcoRV and electroporated into RH ΔHXPRT ΔKU80 (24) parasites followed by selection for HPT. The population was cloned using limiting dilution to obtain single clones. PCR was performed using primers 5′-CAC GTT GAA AGT CTG CAG GAA AGC-3′ and 5′-ACC ATG AAT TCC CGT CCT CCA CTT-3′ to identify clones where the YFP gene inserted before the stop codon of the endogenous TGME49_010810 gene by a homologous single cross over event with the targeting construct. Western blot analysis for the YFP protein was performed as described above using mouse anti-GFP monoclonal antibodies (clones 7.1 and 13.1) (Roche Applied Science).

Conoid extrusion

Parasites were harvested from freshly lysed HFF cells (naturally lysed by parasite egress) and passed through a 25 gauge needle, filtered through a 3 μm polycarbonate filter and collected by centrifugation at 1700 rpm, for 10 min at 4° C. Parasites were re-suspended in phosphate buffered saline (PBS) with 10% fetal bovine serum and 0.5 M ethanol for 30 seconds, fixed with 1.25 % glutaraldehyde (Ted Pella, Inc, CA) for 30 min, washed with PBS and smeared on glass slides (31) slides were air dried, mounted using Vectashield containing DAPI (Vector Laboratories, Burlingame, CA) and visualized using a Zeiss inverted Axiovert 200 motorized microscope with a 100X objective (PlanApo 1.4 na oil PH3 objective). Pictures were taken with the Zeiss Axiocam MRM cool CCD camera. At least 100 parasites were examined in duplicate.

Results

Isolation of T. gondii mutants with increased susceptibility to macrophage activation

We devised an in vitro screen in murine macrophages to identify T. gondii insertional mutants impaired in their ability to replicate in infected macrophages moderately activated by IFN-γ and LPS following parasite invasion. Parasites were added to bone marrow-derived macrophages four hours prior to initiating macrophage activation to allow parasites to invade, establish their PV and begin intracellular replication prior to activation. The concentration of IFN-γ and LPS and time course between parasite invasion and macrophage activation were designed to allow replication of wild type (WT) parasites while macrophages were in the process of becoming fully activated (24 hours). Thus we could use the process of macrophage activation to effectively titrate parasite fitness by comparing the extent of replication of mutant parasites compared to wild type parasites over time.

This model was used to screen two independent libraries of T. gondii insertional mutants to identify mutants that exhibit impaired replication following activation of infected macrophages with LPS and IFN-γ but normal survival/replication in naïve macrophages. Fourteen mutants were isolated, out of approximately 7000 screened, that were defective for replication following macrophage activation. Five of the insertional mutants that we have identified the genomic insertion site plus an additional mutant with the most pronounced defect (11BE9) are shown in the following figures. Whereas WT parasites replicated to two to four parasites per PV, mutants were selected that failed to replicate beyond one parasite per PV and/or appeared amorphous/degraded (Fig. 1B, 2). Replication of these mutants was also analyzed in naïve macrophages to confirm that the mutants replicated like wild type parasites in the absence of macrophage activation (Fig. 1A, 2) to confirm a defect specific for resistance to macrophage activation. These 14 mutants we designated counter-immune or CIM mutants as they have genes disrupted that counter the affects of macrophage activation. Representative pictures of WT parasites following macrophage activation and CIM mutants as represented by the 2B6 mutant are shown in figure 2.

Figure 1.

T. gondii CIM mutants display increased sensitivity compared to wild type parasites to the intracellular effects of macrophage activation. A. Parasite replication in naïve macrophages. B. Parasite replication or degradation following activation of infected macrophages with LPS and IFN-γ. C. Addition of aminoguanidine (1mM) to macrophages during activation to inhibit iNOS restores parasite replication. Quantification is the mean and standard deviation of two counts of 100 PVs each per conditions. Each experiments was performed a minimum of two times. Significance was determined by Student’s t-test. *Significance < 0.05; **Significance ** <0.01.

Figure 2.

Representative pictures of WT parasites that have replicated to two or four parasites per vacuole compared to the 2B6 mutant that either remained at one parasite per vacuole (static) or appeared amorphous/degraded (bottom panel) Parasites were stained with a polyclonal antibody against Toxplasma gondii.

The CIM mutants were compared to a previously published mutant, 89B7, with defective survival following macrophage activation as an independent evaluation of their degree of defect (32). All of the CIM mutants had a defect greater than or equal to that we observed for the 89B7 mutant (67% with 1 parasite per vacuole; 23% with 2 parasites/vacuole and 11% degraded parasites).

T. gondii mutants share an increased susceptibility to reactive nitrogen intermediates (RNIs) associated with macrophage activation

The 14 counter immune (CIM) mutants were examined to determine the macrophage activation pathway(s) responsible for their defect in replication/survival to which the mutation(s) increased their susceptibility. The experimental protocol for the following experiments was similar to that used for the primary screen. Parasites were added to macrophages at a 1:2 ratio of parasites to macrophages. Parasites were then allowed to invade macrophages for four hours, then monolayers were rinsed to remove extracellular parasites and macrophages activated or left unactivated for an additional 24 hours at which time monolayers were fixed, stained to detect parasites and parasite number per PV evaluated. First, the role of macrophage reactive nitrogen intermediates (RNIs) in the susceptibility of CIM mutants was examined by inhibiting inducible nitric oxide synthase (iNOS) with aminoguanidine during activation of infected macrophages. As shown in figure 1C inhibition of iNOS restored the replication of the CIM mutants to that of WT parasites. In fact, inhibition of iNOS enhanced replication of both WT parasites and CIM mutants to levels similar to that for naïve macrophages without activation (Fig 1A,B,C). The phenotype was confirmed using macrophages isolated from inducible nitric oxide synthase knock out mice (data not shown). Since reactive nitrogen and oxygen species often act synergistically, we evaluated the contribution of reactive oxygen species to the CIM mutants’ phenotypes using macrophages isolated from gp91-phox knockout mice. However, there was no difference in the phenotype of any of the CIM mutants in macrophage from wild type compared to gp-91-phox knockout mice (data not shown). IFN-γ-inducible GTPases are important for elimination of Toxoplasma gondii, particularly non-Type I geneotypes of the parasite, in murine macrophages activated with IFN-γ prior to parasite infection. In order to evaluate the role of p47 IFN-γ-inducible GTPases in the CIM mutants’ phenotype, we used macrophages from mice that had both Irgm1 and Irgm3 deleted. This double deletion has been shown to eliminate the anti-Toxoplasma activity of this family of GTPases during acute infection and to inhibit loading of IFN-γ-inducible GTPases to the parasitophorous vacuole (25). As shown in Figure 3, the absence of Irgm1/Irgm3 did not rescue any of the CIM mutants from the effects of macrophage activation (Figure 3B versus C). These studies support that the defect in each of the CIM mutants is dependent on production of reactive nitrogen intermediates associated with macrophage activation and largely independent of reactive oxygen species and IFN-γ-inducible p47 GTPases.

Figure 3.

The impaired replication/survival of T. gondii CIM mutants is not dependent on macrophage Irgm1/Irgm3. A. Parasite replication in naïve macrophages; B. activated macrophages; C. activated Irgm1/3 macrophages. Quantification is the mean and standard deviation of two counts of 100 PVs each per conditions. The experiment was performed once. Significance was determined by Student’s t-test. *Significance < 0.05; **Significance ** <0.01.

The CIM mutants could be directly susceptible to macrophage RNI production alone or to downstream determinants of RNI working in association with other macrophage mediators induced by activation. To determine if RNI was sufficient to suppress CIM mutants, sodium nitroprusside (SNP), a nitric oxide donor was added to infected naïve macrophages in the absence of activation. As shown in figure 4A, SNP in the absence of macrophage activation strongly suppressed replication of a subpopulation of CIM mutants including 2B6, 1BD11, 11BE9 and 68C6, suggesting these mutants were susceptible to RNIs alone and did not require the action of additional antimicrobial mediators induced by activation. In contrast, the A40E4 mutant replicated similar to wild type parasites in the presence of 100 μM sodium nitroprusside (SNP). These results were confirmed using 100 μM DETA NONOate as the nitric oxide donor instead of SNP to avoid potential complications related to cyanide liberation as a consequence of SNP treatment (Figure 4B). Parasite replication in naïve macrophages in the absence of nitric oxide donors is shown in figure 1A as the SNP and DETA NONOate data in figure 4 were performed as part of the experiment shown in figure 1. These result using nitric oxide donors in the absence of macrophage activation indicate that the defect in many of the CIM mutants is a consequence of an increased susceptibility to reactive nitrogen intermediates. Nitrite/nitrate was measured in the supernatants of our experimental culture conditions with activated macrophages and with DETA NONOate to roughly compare the amount of nitric oxide generated by both conditions. Macrophage activation with and without parasites resulted in approximately 50 μM/ml nitrite/nitrate after 24 hours while DETA NONOate resulted in approximately 30 μM ml nitrite/nitrate. Consequently, the amount of nitric oxide required to impair replication/survival of the sensitive CIM mutants is likely similar in the two conditions. Therefore, RNI in the absence of additional mediators of macrophage activation is sufficient to significantly impair replication/survival of many of the CIM mutants.

Figure 4.

Nitric oxide donors in the absence of macrophage activation are sufficient to impair replication/survival of many of the CIM mutants. A. Infected macrophages were cultured in the presence of 100μM/ml sodium nitroprusside (SNP) for 24 hours. B. Infected macrophages were cultured in the presence of 100μM/ml DETA NONOate for 24 hours. Quantification is the mean and standard deviation of two counts of 100 PVs each per conditions. Each experiments was performed a minimum of two times. Significance was determined by Student’s t-test. *Significance < 0.05; **Significance ** <0.01.

Pathogens like Salmonella and Mycobacteria have mechanisms to prevent recruitment of iNOS to their intracellular vacuole (33–35). Consequently, we wanted to determine if any of the mutations in the CIM mutants affected trafficking of macrophage cytosolic iNOS to the PV. Like PVs containing WT parasites, CIM mutant PVs did not accumulate iNOS in the PV or PV membrane (PVM) as fewer than 5% of WT or CIM mutant PVs stained positive for macrophage iNOS even though iNOS was expressed in the macrophage cytosol (Fig 2B and data not shown). To determine if the CIM mutants were altered in their ability to restrict lysosome fusion to the PV, macrophages were infected, activated and 24 hours later co-stained for parasites and lysosomal membrane associated protein -1 (LAMP1) (36). Both WT parasites and the CIM mutants resided in PVs that did not express LAMP1 as 99% of PVs for WT and CIM mutants were LAMP1 negative. Since autophagosomes generally fuse with lysosomal LAMP1-positive compartments (37), this also suggests that the defect in each of the CIM mutants is not associated with enhanced susceptibility to autophagosome mechanisms.

The CIM mutants appear similar to WT parasites in their effect on macrophage iNOS

The fact that nitric oxide donors were sufficient to impair replication/survival of many of the CIM mutants indicated that the defect in the majority of the mutants is due to an inability to withstand nitrosative stress. However, the CIM mutants were also evaluated by immunofluorescence microscopy to monitor their ability to inhibit iNOS induction in infected macrophages in response to activation. Figure 5A shows a representative picture of infected macrophages that have largely suppressed iNOS upregulation (no arrow head) compared to those unable to effectively suppress iNOS production (arrow heads). As shown in Figure 5B, WT parasites were only partially effective at inhibiting iNOS upregulation in infected macrophages following activation with LPS and IFN-γ. However, none of the CIM mutants had an obvious defect in their ability to suppress macrophage iNOS relative to WT parasites in this preliminary analysis. The level of iNOS transcript and protein was further quantified by comparing macrophages infected with WT parasites compared to the 2B6 CIM mutant. Macrophages were infected at a 1:1 ratio with parasites. Consequently, iNOS quantification includes that from infected macrophages as well as uninfected bystander macrophages. As shown in Figure 5C, naïve macrophages expressed low levels of iNOS transcript that was not altered by the presence of parasites. In contrast, activation of macrophages induced over a thousand fold increase in iNOS transcripts but levels were similar in the presence of both WT and 2B6 mutant parasites. Western blot analysis of iNOS protein also showed similar levels of iNOS protein in macrophages cultured with WT or 2B6 mutant parasites (Figure 5D).

Figure 5.

The defect in the CIM mutants is not associated with increased macrophage iNOS levels. A, B. Qualitative and quantitative analysis of iNOS levels in infected macrophages by fluorescent microscopy. Arrowheads point to infected macrophages that have increased iNOS expression. Parasites without arrowheads are in macrophages that have not increased iNOS expression following activation. C. Quantification of iNOS transcript by qRT-PCR. Fold change increase in iNOS transcript levels in macrophages with and without parasites following activation relative to naïve macrophages. The experiment was repeated twice. Normalization was performed with GAPDH. Lines indicate a significance value of <0.05 between naïve and activated macrophages but not between any of the activated samples. No bands were present in the absence of reverse transcriptase. D. Quantification of iNOS protein in macrophages with and without parasites relative to mouse actin. The experiment was repeated twice. 2B6 Recap and 2B6 deletion are deletions that recapitulate the original plasmid insertion (2B6 Recap) or have a targeted deletion of the 2B6 protein (2B6 deletion).

Identification of the genomic insertion site in the CIM mutants

In order to identify the T. gondii gene(s) contributing to parasite replication/survival following macrophage activation, we attempted to identify the site in the parasite genome where the mutagenesis plasmid inserted in each CIM mutant. We have identified the genomic insertion site for the plasmid in the 2B6, 40E4 68C6, I01A1 and 1BD11 mutants. The genomic insertion site for the A40E4 and 68C6 mutants were previously identified in our signature-tagged mutagenesis screen and is published (26). The plasmid genomic insertion sites are shown in Table 1. The genes disrupted by plasmid insertion all code for hypothetical proteins. Plasmid rearrangement occurred in some of the other CIM mutants along with tandem insertions and attempts to identify the insertion site in these mutants is ongoing.

Table 1.

Genomic site of plasmid insertion site each mutant.

Mutant	Insertion site	Prediction	Protein
2B6	Chr IV: 2015.888 K	TGME49_010810	Hypothetical
A40E4	Chr XII: 5802.190K	TGME49_078860	Hypothetical
68C6	Chr VIII: 5874.620K	TGME49_069350	Hypothetical
1BD11	Chr VIII: 3101.459K	TGME49_074010	Hypothetical
I01A1	Chr: III: 1647.491 K	ESTs (no prediction) or TGME49_054370	Adenylate and guanylate cyclase catalytic domain- containing protein

Characterization of the gene disrupted in the 2B6 mutant

The 2B6 mutant was one of the CIM mutants with the greatest sensitivity to addition of the nitric oxide donor SNP in macrophages in the absence of additional exogenous stimulators of macrophage activation. The insertion site in the 2B6 mutant is within the 5′UTR of the hypothetical gene TGME49_010810 (ToxoDB.org). The cDNA for TGME49_010810 was sequenced in entirety and two isoforms of the cDNA were identified. The larger transcript consists of five exons (168 amino acids) while the smaller transcript has the fourth exon spliced out. Conceptual translation of the cDNA revealed a single potential start methionine residue consistent with mass spectrometry peptide sequences for the proteins available at ToxoDB.org. The larger transcript encodes a 107 amino acid protein, 11.9 kDa, and the smaller transcript encodes a 78 amino acid, 8.7 kDa, protein with a truncated 3′end. The predicted protein isoforms do not contain conserved domains, transmembrane domains, signal peptides or homology to other known or hypothetical proteins. A schematic of both cDNA isoforms is shown in figure 6A. Based on the sequenced transcript, the plasmid inserted into the 2B6 mutant’s genome 52 bp downstream of the 5′UTR start site of the gene. qRT-PCR using primers downstream of the insertion site confirmed loss of the cDNA transcript for TGME49_010810 in the 2B6 mutant resulting in a functional gene deletion (Fig. 6B, C).

Figure 6.

TGME49_010810 gene encodes a transcript and results in an expressed protein A. Schematic of the transcript for the two isoforms of TGME49_010810 gene Untranslated (UTR) regions are darkly shaded, exons are represented by white boxes and introns as solid lines. The plasmid insertion site relative to the transcript is shown as an arrow. The region deleted in the gene deletion clone is shown. B. cDNA expression of the 2B6 transcript relative to actin in WT versus gene deleted parasites. The 2B6 mutant, Recap 1 (R1), Recap 2 (R2) and the 2B6 deletion of just the protein region (deletion) no longer express 2B6 cDNA. No bands were present in the absence of reverse transcriptase (data not shown). C. Western blot of wild type and 2B6 mutant parasites using a polyclonal antibody created to recombinant TGME49_010810 protein. T. gondii surface associated antigen-1 (SAG1) expression is shown as a loading control and was detected with the mAb DG52. The experiment was done twice. D. Western blot analysis of the 2B6 protein relative to SAG1 in WT parasites versus the 2B6 mutant, 2B6 Recap clone and 2B6 targeted protein deletion.

TGME49_010810 Protein expression

We expressed a recombinant form of the larger 168 amino acid protein in E. coli and used it to generate an antiserum to the TGME49_010810 protein. Western blot analysis using this antisera detected an approximately 16 kDa protein in WT parasites that was not present in the 2B6 mutant (Fig. 6C, D). The blot was probed with an antibody to T. gondii SAG1 protein to confirm equivalent loading of samples in each lane. The corresponding pre-serum showed no reactivity in this molecular weight range (data not shown).

The replication/survival defect in activated macrophages is recapitulated by re-creating the plasmid insertion in WT parasites

The methods used to generate random insertional mutants in T. gondii do not preclude creation of additional mutations due to sporadic point mutations elsewhere in the genome or insertions of plasmid fragments not readily identified by Southern blot analysis. To confirm the insertion of the plasmid into the identified genomic locus in the parasite was responsible for the defective replication/survival in activated macrophages, we re-created the plasmid insertion in the 2B6 mutant in WT parasites. Using the technique described in the materials and methods, we isolated two independent clones from different electroporations that we designated Recap 1 and Recap 2. Both clones replicated like WT parasites in naïve macrophages (Fig. 7A) but shared a similar phenotype to the original 2B6 mutant following activation of infected macrophages (Fig. 7B). Like the 2B6 mutant, the defective replication/survival following macrophage activation was reversed by the addition of aminoguanidine to inhibit inducible nitric oxide synthase (iNOS) during activation (data not shown). Similarly, addition of SNP or sodium nonanoate to infected naïve macrophages induced impaired replication of the Recap clones to levels comparable to that for the 2B6 mutant (data not shown).

Figure 7.

Replication of T. gondii ΔTGME49_010810 clones compared to WT parasites in A. naïve macrophages or B. following macrophage activation. Parasites with the entire TGME49_010810 cDNA disrupted (Recap 1 and 2) or parasites with just the protein coding region deleted (Deletion 1) recapitulate the phenotype of the 2B6 mutant. The mean and standard deviation of two counts of 100 PVs each is shown. Each experiment was performed at least twice.

Deletion of the protein coding region alone is sufficient to recapitulate the defect in replication/survival following macrophage activation

The plasmid insertion in the 2B6 mutant and the two independent Recap clones disrupts the cDNA of gene TGME49_010810 resulting in a functional gene deletion (Fig 6) However untranslated regions in eukaryotes can have effects independent of the transcribed protein. To confirm that the TGME49_010810 encoded protein itself was important for parasite replication/survival following macrophage activation, we created a directed gene deletion of just the protein coding region of TGME49_010810. The region deleted is shown in the schematic in figure 6A and the deletion of the cDNA is confirmed in Figure 6B,C,D. As shown in figure 7B, targeted deletion of just the protein coding region had a similar effect to the deletion of the entire cDNA, confirming a role for the protein itself in protection against nitrosative stress following macrophage activation.

Since in our model we let the parasites invade for four hours prior to activation, it seemed unlikely that the difference in replication/survival by WT versus the TGME49_010810 deletion clones was a consequence of differences in parasite invasion efficiency of macrophages. However, to ensure there was not a pronounced effect due to a defect in invasion efficiency in the gene deleted clones, we did a short 15 minute pulse of naïve macrophages with parasites and evaluated the number of internalized and extracellular parasites by differential permeabilization followed by staining for parasites. 57 +/− 2% of WT parasites were internalized relative to 62.5 +/− 8% and 49 +/− .7% of the gene deletion clones confirming gene deletion did not result in a pronounced defect in invasion.

T. gondii conoid associated protein-1 (TgCAP1) is expressed in the parasite conoid and cytosol

T. gondii has a number of organelles and structures specific for the Apicomplexa and each location often has a defined function(s). Consequently, where a protein localizes in the parasite or infected host cells, if and how it traffics during the parasites’ intracellular lytic cycle, and its time course for expression all provide clues to its potential function. This is particularly critical for a protein like TGME49_010810 that has no homology to other proteins or to functional domains. The antiserum we made to the recombinant protein worked for detection via Western blot analysis but was not effective for immunoflourescence detection even with a variety of fixation methods. As an alternative approach, we endogenously tagged TGME49_010810 isoform 1 with YFP. Expression of the fusion protein was verified by RT-PCR and western blot analysis (Fig. 8A). IFA analysis shows that the protein is localized to the parasite cytosol and also at the apical end, in both extracellular (data not shown) and intracellular parasites (Figure 8B,C). Additionally in parasites undergoing endodyogeny (replication) expression in the apical region of the daughter cells was also seen. Co-staining with an antibody to T. gondii ISP1 protein which stains the apical portion of the inner membrane complex localizes (27) the TGME49_010810 protein to the extreme apical end of the parasite beyond the apex of the inner membrane complex suggestive of conoid localization (Fig. 8B). The T. gondii conoid is a 380-nm diameter motile organelle in the extreme apical end of the parasite, consisting of a unique polymer of tublin fibers wound into a spiral like a compressed spring (38). The conoid can be retracted, enclosed within a shell formed by the parasite’s subpellicular microtubles or protruded beyond the apical end of the microtubules. Protrusion of the conoid can be induced by calcium ionophore or ethanol treatment in vitro (31, 39). In order to determine whether the tagged protein localized to the subpellicular microtubules or the conoid, we induced conoid protrusion in vitro using ethanol and then fixed the cells and examined the YFP-tagged protein by IFA. The tagged isoform 1 TGME49_010810 protein was observed in the extruded conoid. Consequently we have named the protein a T. gondii conoid associated protein-1 (Tg CAP1). Localization of the protein in the extended conoid but not at the extreme tip also suggests that it localizes to the conoid rather than the pre-conoidal or the polar rings. Closer analyses of the protein localization pattern revealed that in addition to the cytoplasmic and the conoid expression pattern, a strong peri-nuclear expression and sometimes a weak baso-lateral expression was also observed. This expression pattern was observed using both formaldehyde and glutaraldehyde as a fixative and with and without conoid extrusion ruling out this expression pattern as an artifact created by the use of these chemicals or dependent on signals that induce extrusion of the conoid.

Figure 8.

Immunolocalization of an endogenous YFP tag of the TGME49_010810 protein in intracellular and extracellular parasites. A. Expression of the YFP-tagged 2B6 protein in WT versus the 2B6-YFP parasite clone (left). Western blot analysis of YFP protein of WT parasites compared to the 2B6-YFP parasite clone (right). B. Localization of the endogenous TGME49_010810-YFP protein relative to the apical ISP-1 protein and the parasite nucleus (DAPI). The merge image is also stained with 594 ani-rabbit Toxoplasma antibody to mark the parasite (light red). C. Localization of TGME49_010810-YFP to the conoid following induction of conoid extrusion with ethanol. Lower panel shows a magnification of a single parasite with the conoid extruded and TGME49_010810-YFP expression. D. Quantification of the number of parasites that expressed the TGME49_010810-YFP in the conoid plus perenuclear, the conoid only, the perinuclear only, the conoid plus perinuclear plus basel end of the parasite or in the basel end of the parasite alone. TGME49_010810-YFP was expressed in the cytoplasm of the parasite in all of the parasites (B and C and data not shown). Each experiment was performed at least twice.

Discussion

Our aim in this study was to utilize the host innate immune response, in this case macrophages, to discover genes and mechanisms that T gondii has developed evolutionarily in order to counter host immunity. T. gondii has evolved to infect and cause natural disease in virtually any warm blooded vertebrate host species. Consequently, the mechanisms it uses to evade host immunity to successfully multiply, disseminate and convert to encysted bradyzoites must be effective across a broad range of species. Therefore it is not surprising if T. gondii has generated multiple mechanisms important for protecting itself from RNIs. T. gondii is known to modify its host cell by altering the action of host transcription factors. For example it suppresses host cell activation and cytokine production by inhibiting nuclear translocation of STAT1 (9) and NFkB (10) in infected cells and stimulating STAT3 (19, 40, 41) and Suppressor of Cytokine Signaling (SOCS) (7). It also prevents host cell chromatin remodeling to suppress TNF-α production (6). T. gondii is also likely to have an array of additional genes that provide protection or repair mechanisms against toxic intracellular mediators generated in response to host cell activation. For example, T. gondii has an active antioxidant system that is likely protective against oxidative stress during infection (42, 43) (44, 45). The mechanisms used by T. gondii to modify its host cell and protect itself from intracellular effectors in the midst of an ongoing immune response are likely critical for its ability to use monocytes and dendritic cells to disseminate to the brain and CNS to convert to bradyzoites contained in tissue cysts to establish persistent infection (1).

To identify T. gondii genes important for surviving activation of infected cells we developed a macrophage activation model that we used to screen parasite insertional mutants. In our macrophage activation model we allowed parasites to invade naïve macrophages and establish a PV four hours before activating macrophages with IFN-γ (100U/ml) and LPS (100ng/ml)). This time course was chosen to enrich for defects specific to the generation of macrophage intracellular mediators rather than parasite invasion efficiency. In this model WT parasite replication is slowed compared to naïve macrophages but replication still occurs for at least 24 hours. It is designed to act as a modified competition model between the time required for full macrophage activation versus the time in which WT parasites or mutant parasites can continue to replicate within their PV. We utilized this model in a forward genetic screen to isolate parasite insertional mutants that unlike WT parasites were unable to replicate or were degraded early following initiation of macrophage activation. Our studies indicate that all of the fourteen CIM mutants isolated are specifically impaired for resistance to pathways dependent on macrophage RNIs although their susceptibility in naïve macrophages to nitric oxide donors and their phenotype following macrophage activation differs. Inhibition of iNOS also restored replication of WT parasites following activation to levels similar to naïve macrophages. This is consistent with studies that indicate the importance of nitric oxide in suppression of T. gondii replication in murine macrophages following activation in vitro with IFN-γ and LPS (3, 46). Our observation that WT parasites did not substantially inhibit iNOS induction in macrophages may be a combination of a relatively low multiplicity of infection allowing iNOS production from uninfected macrophages as well as our use of a Type II genotype of Toxoplasma. The Type II genotype is associated with a lack of ROP16 important for downregulation of iNOS induced by IFN-γ stimulation (47, 48). The Type II genotype also is associated with the presence of GRA15 enhancing host cell signaling through NFκB (20).

Although the CIM mutants share an inability to withstand nitrosative stress, we observed no role for reactive oxygen species or p47 IFN-γ-inducible GTPases in the CIM mutants’ phenotype. p47 IFN-γ-inducible GTPases are critical in vivo in murine hosts for the control of Toxoplasma during infection particularly Type II and III parasite genotypes (49, 50) and important for the cidal activity against Toxoplasma by macrophages pre-activated with IFN-γ prior to parasite invasion (5). Type I genotypes of the parasite, in contrast, secrete a rhoptry protein, ROP18, that blocks recruitment of Igrb6 to the parasitophorous vacuole preventing killing of intracellular parasites (51, 52). The deletion of Irgm1 and Irgm3 in combination inhibits loading of p47 GTPases including Irga6 and Irgab6 to the parasitophorous vacuole (25). Therefore, this gene deletion combination is an effective way to evaluate the contribution of the p47 GTPases known to be important for cidal activity against Toxoplasma. Since p47 IFN-γ inducible GTPases require induction by IFN-γ, studies analyzing their activity against Toxoplasma pre-activate macrophages prior to parasite invasion and parasite killing occurs rapidly following invasion. Parasites that escape initial killing following invasion remain capable of replication suggesting that these GTPases are most effective during or immediately following parasite invasion instead of against replicating parasites in established vacuoles (5, 53). The fact that our model allows parasites to invade macrophages prior to activation means that the p47 family of GTPases are not fully induced by IFN-γ until nearly 24 hours after parasite invasion. It is possible that by the time IFN-γ-inducible GTPases are fully engaged the CIM mutants have already been killed by RNIs. However, it is also possible that the IFN-γ-inducible GTPases are less effective against Toxoplasma once the parasite-containing vacuole has been formed and parasite replication initiated.

In this study we focused primarily on characterizing the gene disrupted in the 2B6 mutant. This mutant was one of a subset of CIM mutants susceptible to nitric oxide donors in the absence of macrophage activation indicating a defect in the mutant’s ability to protect itself from RNIs. The insertional plasmid in the 2B6 mutant disrupted a novel hypothetical gene annotated as TGME49_010810 that our results show encodes two protein isoforms, of approximately 11.7 and 8.7 kDa each, that differ from each other based on alternative splicing and distinct 3′ ends. The proteins have no homology to known proteins or functional domains and no defining characteristics such as putative transmembrane domains or signal peptides. We used two approaches to confirm a role for the gene in parasite resistance to nitrosative stress. First, we created two independent recreations of the original plasmid insertion in the 2B6 mutant in WT parasites to confirm the identified locus was responsible for the phenotype of the 2B6 mutant. The plasmid insertion in the 5 UTR of the transcript resulted in a functional gene deletion of the entire transcript. We also created a deletion of just the protein coding region of the gene to confirm the role for the protein itself and not the 5′ UTR of the gene. Both disruption of the entire gene by insertion of the plasmid in the 5′ UTR and deletion of just the protein coding region confirmed a role for this gene in protecting T. gondii from nitrosative stress associated with macrophage activation.

Endogenous tagging of the dominant long form of the protein was used to evaluate its localization and trafficking. Our studies show that the protein is expressed in or associated with the parasite conoid as well as the parasite cytosol. It is unclear how Tg CAP1, a protein associated with the parasite conoid and cytosol protects the parasite from nitrosative stress although a role in the parasite cytosol for protection against RNI’s seems more plausible than the conoid localization. Non-structural proteins with primary conoid localization are limited to the calcium binding proteins Tg CAM1 and CAM2 (30). Calcium plays a key role in regulating conoid extrusion (31, 39) so a role for these proteins in sensing and responding to calcium seems likely. It may also be postulated that Tg CAP1 similarly functions in some way to sense the parasite’s intracellular environment and trigger an adaptive response. The conoid function itself remains ambiguous but it is hypothesized to have a role in invasion and egress of host cells (54). However, deletion of Tg CAP1 did not have an overt effect on parasite adhesion or invasion. We have currently only localized the 11.9 kDa protein using an endogenous tag. However, for our gene deletions both forms are deleted. Consequently, we can not rule out the possibility that the 8.7 kDa protein has a different localization pattern or we may lack sufficient sensitivity to detect Tg CAP1 protein at other locations. Future studies our aimed at dissecting the precise mechanism used by Tg CAP1 to protect the parasite from nitrosative stress.

In this paper we identify a novel T. gondii protein Tg CAP1 that contributes to the survival/replication of T. gondii following macrophage activation. We also describe the isolation of thirteen additional T. gondii CIM mutants that also act to either protect the parasite from RNIs produced during macrophage activation or affect the production of RNIs. Identification and characterization of the genes disrupted in these mutants will provide a foundation to begin dissecting parasite molecular pathways important for resistance to nitrosative stress associated with macrophage activation. The genes identified may also play a role in parasite regulation of endogenous nitrosative pathways as T. gondii also has putative genes involved in nitrosative systems including a putative nitric oxide synthase gene (55, 56). Since RNI can also trigger tachyzoite to bradyzoite conversion these parasite genes may also be important for cyst formation in response to certain stimuli. T. gondii has been in a relationship with immune systems in virtually every warm blooded vertebrate host type through its evolutionary history due to its wide host and host cell range. Understanding the defenses it has developed during its evolution with host immune systems may also lead to the identification of novel immune mechanisms it has had to develop defenses against in order to be successful.

Abbreviations used in this paper

HFFs

human foreskin fibroblasts

MOI

multiplicity of infection

IFA

immunofluorescence microscopy

D10

DMEM plus 10% FCS, L-glutamine and penicillin and streptomycin

DCs

dendritic cells

CIM mutants

counter-immune mutants

RNI

reactive nitrogen intermediates

iNOS

inducible nitric oxide synthase

SNP

sodium nitroprusside

LAMP1

lysosomal membrane associated-1 antigen

PV

parasitophorous vacuole

PVM

parasitophorous vacuole membrane

WT

wild type

HPT

hypoxanthinexanthineguanosine phosphoribosyltransferase

MPA/XAN

mycophenolic acid and xanthine

kDa

kilodalton

YFP

yellow fluorescent protein

ISP1

inner membrane complex (IMC) sub-compartment protein