Toxoplasma gondii Co-opts the Unfolded Protein Response To Enhance Migration and Dissemination of Infected Host Cells

Cells that are infected with the parasite Toxoplasma gondii exhibit heightened migratory activity, which facilitates dissemination of the infection throughout the body. In this report, we identify a new mechanism used by Toxoplasma to hijack its host cell and increase its mobility. We further show that the ability of Toxoplasma to increase host cell migration involves not the enzymatic activity of IRE1 but rather IRE1 engagement with actin cytoskeletal remodeling. Depletion of IRE1 from infected host cells reduces their migration in vitro and significantly hinders dissemination of Toxoplasma in vivo. Our findings reveal a new mechanism underlying host-pathogen interactions, demonstrating how host cells are co-opted to spread a persistent infection around the body.

The ER is sensitive to the perturbations in protein homeostasis through a stresssensing pathway known as the unfolded protein response (UPR). Three ER transmembrane proteins, IRE1, ATF6, and PERK, operate as sensors that activate the UPR, leading to changes in gene expression that restore and expand the processing capacity of the organelle (6)(7)(8). IRE1 (ERN1) is a protein kinase and endoribonuclease that facilitates cytosolic splicing of XBP1 (XBP1s) mRNA, thereby enhancing expression of the XBP1s isoform, which induces transcription of genes involved in ER-associated protein degradation (ERAD), lipid synthesis, and protein folding (7,8). In response to ER stress, ATF6 transits from the ER to the Golgi apparatus, where it is cleaved, releasing an N-terminal cytosolic fragment (ATF6-N) that enters the nucleus and activates UPR target genes involved in protein folding and transport (6,9). PERK (EIF2AK3) is the third UPR sensor, which phosphorylates the ␣ subunit of eukaryotic initiation factor 2 (eIF2␣) to direct translational and transcriptional modes of gene expression that regulate ER processing of proteins, metabolism, and the oxidation status of cells (6,10). While the three ER stress sensory proteins function in parallel, there is cross-regulation that serves to coordinate the timing and magnitude of the UPR. For example, PERK was reported to induce expression of RPAP2, which serves to dephosphorylate and repress IRE1, thereby providing a means for the cell to abort failed ER stress adaptation and trigger apoptosis (11).
In addition to its role in the UPR, IRE1 was recently shown to modulate cytoskeletal remodeling and cell migration through direct interactions with the actin cross-linking factor filamin A (12). The role of IRE1 in cytoskeletal remodeling is enhanced by pharmacological induction of ER stress but occurs independently of IRE1 protein kinase and endoribonuclease activities (12); rather, IRE1 serves as a scaffolding protein for filamin A to orchestrate changes in cellular motility. This is noteworthy because Toxoplasma stimulates host cell migration, turning its host cell into a "Trojan horse" that can ferry parasites throughout the body (13). Given the recruitment of host ER to the PV, we postulated that migratory activities mediated by IRE1 function in parasite dissemination. In the present study, we uncovered a new mechanism by which Toxoplasma alters host ER homeostasis to produce hypermigratory activity in infected host cells. We show that during the course of Toxoplasma infection, the three UPR sensory proteins in the host cells are activated by a process involving calcium release from the ER, leading to IRE1 oligomerization, association with filamin A, and enhanced cell migration. Importantly, the IRE1-associated migration is a crucial determinant for successful dissemination of toxoplasmosis in a mouse model of infection.

RESULTS
Induction of the UPR in Toxoplasma-infected host cells. It is currently unclear why intracellular tachyzoites recruit host ER to the parasite PV. To address whether Toxoplasma perturbs host ER homeostasis, we infected mouse embryonic fibroblasts (MEFs) with RH strain parasites and measured three primary markers of the host UPR over a 36-h time course. Within 12 h of infection, Toxoplasma increased activation of PERK as measured by its self-phosphorylation (PERK-P), induced expression of ATF6 and formation of its cleavage product ATF6-N, and increased levels of the IRE1-derived spliced variant of XBP1 (XBP1s) (Fig. 1A). It is noteworthy that whereas expression of ATF6-N and XBP1s was transient, with increased amounts of the proteins appearing between 12 and 20 h postinfection (hpi), PERK-P increased throughout the 36 h of infection (Fig. 1A). These results indicate that Toxoplasma infection causes ER stress that activates each of the sensory proteins of the UPR, with some differences in the duration of their induction.
Activation of IRE1 involves oligomerization that can be visualized by a pattern of punctate spots by IFA (14)(15)(16). We expressed enhanced green fluorescent protein (EGFP)-tagged IRE1 in MEFs that had deletions of the endogenous IRE1 gene by CRISPR/Cas9 genome editing (see Fig. S1A to C in the supplemental material); upon Toxoplasma infection, we observed formation of IRE1 foci, which is consistent with reported IRE1 activation by oligomerization (Fig. 1B) (15,16). Furthermore, expres-sion of XBP1s mRNA, and its downstream target genes involved in ERAD and protein folding, was induced upon Toxoplasma infection of wild-type (WT) MEFs (Fig. 1C). These results indicate that IRE1 activation and signaling occur in response to Toxoplasma infection.
To further study the timing of UPR induction and potential cross-regulation between the UPR sensors in infected cells, we measured XBP1s mRNA levels by reverse transcription-quantitative PCR (RT-qPCR) in WT, IRE1 Ϫ/Ϫ , and PERK Ϫ/Ϫ MEFs. In WT cells, levels of XBP1s mRNA rose sharply until 18 hpi (Fig. 1D), consistent with the increase in XBP1s protein (Fig. 1A). As expected, XBP1s mRNA was not detected in infected IRE1 Ϫ/Ϫ cells (Fig. 1D). In comparison, there was induced XBP1s mRNA that was sustained during 36 h of Toxoplasma infection in PERK Ϫ/Ϫ cells, indicating that without PERK, IRE1 continues to facilitate XBP1s expression (Fig. 1D). These results are consistent with the idea that PERK governs IRE1 activity, as previously reported (11), and suggest that in infected WT cells, PERK operates to attenuate the response of IRE1 after 18 hpi. , and GAPDH were measured by immunoblot analyses. (B) IRE1 Ϫ/Ϫ MEFs were transfected with a plasmid encoding EGFP-IRE1 (green), followed by infection with Toxoplasma. DAPI (blue) was used to visualize host cell and parasite nuclei. Two boxed areas (1 and 2) of each condition are amplified to highlight the IRE1 distribution. Bar ϭ 5 m. Arrows indicate Toxoplasma infection as based on DAPI staining. (C) MEFs were infected with Toxoplasma for 18 h, or were mock infected, and the mRNA levels of the indicated ERAD/chaperone genes were measured by RT-qPCR. Levels of mRNA were normalized to those for mock-infected cells, represented as a value of 1 (ϮSD; n ϭ 3). **, P Ͻ 0.005; ***, P Ͻ 0.001; ****, P Ͻ 0.0001. (D) XBP1s mRNA levels were measured by RT-qPCR at the indicated times in WT, PERK Ϫ/Ϫ , and IRE1 Ϫ/Ϫ MEFs, as indicated. Levels of XBP1s mRNA were normalized to total XBP1 mRNA in mock-infected cells (value of 1) at each time point (ϮSD; n ϭ 3). ****, P Ͻ 0.0001.
Toxoplasma Infection Induces Host UPR and Migration ® IRE1 affects calcium release from ER in Toxoplasma-infected cells. Protein folding in the ER is highly sensitive to the concentration of calcium, which is released from the organelle by ryanodine receptors (RyR) and inositol 1,4,5-triphosphate (IP 3 ) receptors (IP 3 R) (17). The ER is a major reservoir of calcium; disruptions of calcium homeostasis can lead to unfolded proteins and initiation of the UPR (17). To determine whether calcium content is altered in the host ER during Toxoplasma infection, we first measured cytosolic calcium in infected MEFs. Over an 18-h period, Toxoplasma infection induced a steady increase in host cell cytosolic calcium levels ( Fig. 2A and Fig. S2A). Basal calcium levels in the cytosol of IRE1 Ϫ/Ϫ cells were lower than in WT cells, as previously reported (18), with some increase upon parasite infection ( Fig. 2A and Fig. S2A). To assess the mode of calcium release from the host ER during infection, we monitored calcium transport using fluorescent-4 acetoxymethyl ester (Fluo-4AM) in the presence of antagonists of RyR or IP 3 R. The cytosolic calcium levels were lower in infected cells treated with either antagonist, with the most robust calcium reduction occurring with inhibition of IP 3 R (Fig. 2B). We further addressed the contribution of RyR and IP 3 R to calcium release by adding increasing doses of caffeine and IP 3 , agonists of RyR and IP 3 R activity (19), respectively, to WT and IRE1 Ϫ/Ϫ cells infected with Toxoplasma for 18 h and incubated with Mag-Fluo-4. To compare calcium release between RyR and IP 3 R in WT and IRE1 Ϫ/Ϫ cells, the fluorescence values were represented as the percentage of calcium release and the respective start points were normalized to those for untreated WT and IRE1 Ϫ/Ϫ cells, respectively. The RyR agonist enhanced calcium to similar levels in infected WT and IRE1 Ϫ/Ϫ cells (Fig. 2C). In contrast, the agonist IP 3 induced appreciable calcium release only in infected WT cells (Fig. 2D), indicating a role for IRE1 in regulating IP 3 R activity, as previously reported (18). Surprisingly, the percentage of calcium release was higher when IP 3 R was stimulated than when RyR was stimulated, demonstrating that Toxoplasma infection differentially alters RyR and IP 3 R activities. Collectively, these results suggest that Toxoplasma infection induces significant calcium release from the host ER by processes involving both IP 3 R and RyR; this calcium release . **, P Ͻ 0.05; ***, P Ͻ 0.005. Infected cells were incubated with Mag-Fluo-4, followed by plasma membrane permeabilization with saponin and incubation with ATP to maintain the calcium in the ER. To estimate RyR (C) and IP 3 R (D) activities, WT and IRE1 Ϫ/Ϫ cells infected with Toxoplasma for 18 h were treated with caffeine or IP 3 at the indicated concentrations; calcium release was represented as fluorescence using Mag-Fluo-4 as described previously (19). Values are percent calcium release; the respective start points were normalized to those for untreated WT and IRE1 Ϫ/Ϫ cells, respectively (ϮSD; n ϭ 3). **, P Ͻ 0.05; ****, P Ͻ 0.001. is influenced by IRE1 and is a likely contributor to the activation of the host ER stress sensor proteins shown in Fig. 1A.

IRE1 activation induces cell migration in infected cells.
Toxoplasma triggers rapid morphological changes in host cells, including disappearance of podosome structures and appearance of lamellipodia (20). IRE1 has recently been shown to have noncanonical functions in actin cytoskeletal remodeling by directly binding to filamin A (12). To address whether activation of IRE1 by Toxoplasma infection enhances host cell migration, we quantified the number of lamellipodia per infected cell normalized to the value for uninfected cells ( Fig. 3A and Fig. S3A). At 18 hpi, Toxoplasma infection increased the number of lamellipodia in WT cells and these structures were significantly diminished in IRE1-deficient cells (Fig. 3A). There were greater numbers of lamellipodia in PERK-deficient cells or those treated with a PERK inhibitor, consistent with the idea that PERK is a negative regulator of IRE1 (Fig. 3A). Of interest, treatment with inhibitors of IRE1, namely, 48c, which interferes with endoribonuclease activity, and KIRA6, which blocks IRE1 protein kinase activity, did not change the number of lamellipodia compared with that in infected cells treated with a vehicle (Fig. 3A). These results indicate that IRE1 can control migration of Toxoplasma-infected cells independently of its known enzymatic activities.
The functions of filamin A in cytoskeleton dynamics and cell migration are dependent on phosphorylation of serine 2152 (S2152) (21). We detected a sharp increase in filamin A phosphorylation in MEFs infected with Toxoplasma for 18 h, which was not observed in host cells lacking IRE1 (Fig. 3B). To address whether there is an association between IRE1 and filamin A, we expressed Myc-tagged filamin A (Myc-FLNA) in WT MEFs. Following 18 h of Toxoplasma infection, we then performed an immunoprecipitation (IP) of the tagged filamin A, followed by immunoblot measurements of associated IRE1. There was enhanced association of IRE1 with filamin A in cells infected with the parasite compared to those left uninfected (Fig. 3C).
Next, we measured changes in host cell transmigration upon parasite infection and determined that there was a 2-fold increase in migration of WT MEFs upon infection with Toxoplasma ( Fig. 3D and Fig. S3B). Increased host cell migration was observed regardless of whether the cells were infected with type I RH or type II ME49 strain parasites (Fig. S3C). In contrast, parasite infection did not induce migration of cells lacking IRE1. Levels of migration in infected WT MEFs treated with one of the IRE1 enzymatic inhibitors, KIRA6 and 48c, or an ATF6 inhibitor (Ceapin-A7) were not significantly changed compared to those in untreated infected MEFs, nor were they altered in cells lacking the downstream IRE1 target XBP1 ( Fig. 3D and Fig. S3B). Notably, Toxoplasma-induced migration of PERK-deficient cells or WT cells treated with PERK inhibitor was increased Ͼ2.5-fold upon parasite infection compared to that in infected cells with functional PERK (Fig. 3D and Fig. S3B). These results suggest that IRE1 plays a critical role in inducing migration of Toxoplasma-infected cells and that this migration is independent of the protein kinase and endoribonuclease activities of IRE1 and its downstream target XBP1. Furthermore, PERK is suggested to dampen both IRE1 functions in XBP1 mRNA splicing and cell migration, which are induced upon Toxoplasma infection.
ER stress induces cell migration by mechanisms involving IRE1 oligomerization. To better understand the mechanisms by which IRE1 enhances host cell migration in response to Toxoplasma infection, we rescued the IRE1 Ϫ/Ϫ cells by expressing WT or defined mutant versions of IRE1 ( Fig. 3E and Fig. S4A). Equal amounts of the IRE1 proteins were expressed as judged by immunoblot and immunofluorescence analyses ( Fig. S4B and C). As expected, ire1-wt rescued the migration capacity of infected cells ( Fig. 3E and Fig. S3D). Expression of IRE1 defective in kinase (ire1-kD) or endoribonuclease (ire1-eD) activities still rescued the migration phenotype, further supporting the idea that these activities are dispensable for induced cell migration in response to parasite infection (Fig. 3E). In contrast, cells expressing IRE1 with mutations in the oligomerization domain (ire1-oD) were deficient in Toxoplasma-induced migration (Fig. 3E). Furthermore, a truncated version of IRE1 (ire1-⌬965) lacking the proline-rich carboxy-terminal segment of IRE1 that binds filamin A (12) was also impaired in cell migration following infection (Fig. 3E). As anticipated, only ire1-wt and ire1-⌬965 showed induction of spliced XBP1 mRNA upon pharmacological induction of ER stress (Fig. S4D). These results suggest that Toxoplasma infection induces IRE1 oligomerization (see also Fig. 1B) and that the hypermigratory behavior of infected host cells is reliant on both the oligomerization domain and the portion of IRE1 that interacts with filamin A.
Since we found that Toxoplasma induces calcium release from the host ER upon infection, we examined the importance of calcium release in host cell migration induced by Toxoplasma. We treated infected cells with RyR and IP 3 R blockers or activators during the migration assay: ryanodine and xestospongin-C (XeC) were used to inhibit RyR and IP 3 R, respectively; caffeine and IP 3 were used to activate RyR and IP 3 R, respectively. When infected cells were treated with the RyR and IP 3 R activators (releasing calcium from ER into the cytosol), the migration levels increased compared to those in cells not treated with these agents. In contrast, addition of the IP 3 R inhibitor significantly decreased migration of the infected cells ( Fig. 3F and Fig. S3E), consistent with IP 3 R being the major calcium release receptor involved in triggering the host UPR following Toxoplasma infection (Fig. 2B). These results support the idea that calcium release from the host ER contributes to IRE1 activation and its subsequent role in augmenting migration in response to infection.

IRE1 controls host cell migration in infected immune cells in vitro.
Toxoplasma makes use of immune cells as a "Trojan horse" to disseminate to distal organs and tissues throughout the body of the infected host (13). To address whether Toxoplasma targets IRE1 in immune cells to enhance their migration and facilitate dissemination, we infected bone marrow-derived dendritic cells (DCs) with Toxoplasma. Levels of XBP1s mRNA were sharply increased upon Toxoplasma infection in DCs (Fig. 4A), consistent with the idea that the parasite infection activated IRE1 in this cell type. Next, we used CRISPR/Cas9 and two distinct single guide RNAs (sgRNA1 and -2) to disrupt IRE1 in DCs (Fig. S5A). sgRNA1 and sgRNA2 decreased IRE1 mRNA levels by 66% and 93%, respectively (Fig. 4B). Each gRNA also led to a corresponding reduction in IRE1 protein in DCs, with sgRNA2 leading to no detectable IRE1 (Fig. 4C). IRE1-depleted DCs [ire1 (Ϫ)] did not exhibit decreased viability compared to that of WT cells, nor did they show any difference in infectivity with Toxoplasma ( Fig. 4D and Fig. S5B). However, loss of IRE1 in DCs significantly lowered the transmigratory capacity following infection with either type I or type II strains of Toxoplasma ( Fig. 4E and Fig. S5C and D). To determine whether IP 3 R plays a role in the migration of infected DCs, we carried out the migration assay in the presence of XeC. Inhibition of IP 3 R resulted in a loss of host cell migration following 18 h of Toxoplasma infection ( Fig. 4F and Fig. S5E), further supporting the importance of calcium homeostasis in the ER for migration of Toxoplasma-infected cells. To test whether IRE1 controls the migration of infected macrophages as well, we used CRISPR/Cas9 to deplete IRE1 in J774.1 macrophages (Fig. 4G and H). As observed for DCs, the loss of IRE1 significantly reduced the ability of infected macrophages to migrate ( Fig. 4I and Fig. S5F).

IRE1 facilitates dissemination of Toxoplasma in vivo.
To determine the importance of IRE1 in the migration of infected DCs in vivo, we inoculated C57BL/6 mice intraperitoneally (i.p.) with infected WT or infected ire1 (Ϫ) DCs and measured parasite burden in the spleen by PCR at various intervals over 3 days. Depletion of IRE1 in the DCs by CRISPR/Cas9 was confirmed by RT-qPCR and immunoblot analyses ( Fig. 5A and  B). Moreover, we ascertained that the depletion of IRE1 in DCs had no significant effect on parasite infectibility or replication rate ( Fig. 5C and Fig. S5G). Toxoplasma was first detected in the spleens of mice 12 h following i.p. inoculation with infected WT DCs, increasing at each time point over the 3-day period (Fig. 5D). In striking contrast, appreciable levels of Toxoplasma dissemination of infected IRE1-depleted DCs to the spleen were not detected until 3 days following inoculation of the mice (Fig. 5D). Even at the 3-day point, the loss of IRE1 from DCs produced lowered levels of parasitemia in the spleen that were similar to those measured at 12 h for infected WT DCs. We also measured Toxoplasma dissemination to the brain at 3 days, finding 200-fold fewer parasites when infected IRE1-depleted DCs were inoculated into the mice (Fig. 5E). Mice inoculated with infected DCs lacking IRE1 survived significantly longer than mice receiving infected WT DCs (Fig. 5F). These results demonstrate a novel role for host IRE1 in parasite pathogenesis, as IRE1 is crucial for the migration of immune cells being co-opted as Trojan horses for parasite dissemination.

DISCUSSION
Obligate intracellular pathogens create a niche inside their host cell that allows for parasite protection, nutrient acquisition, and the controlled release of pathogen effectors that promote infection and dissemination. Toxoplasma tachyzoites reside inside a nonfusogenic parasitophorous vacuole (PV) that forms intimate contacts with host organelles and vesicles, including the ER and mitochondria (4,5). It is suggested that the recruitment of host organelles to the PV allows Toxoplasma to control critical host cell operations, including antigen presentation, nutrient production, and suppression of apoptosis (5). In this study, we addressed the consequences of the Toxoplasma-ER engagement on parasite infection and dissemination in the host. As illustrated in a model presented in Fig. 5G, we showed that Toxoplasma infection activates each of the UPR sensor proteins, including IRE1, via ER stress that results at least in part from release of calcium from the organelle, primarily through IP 3 R. In addition to its role in the UPR, IRE1 has recently been shown to have noncanonical functions associated with the remodeling of the cytoskeleton through direct interactions with the actin cross-linking factor filamin A (12). We showed that Toxoplasma alters the morphology of its host cells through IRE1-filamin A interactions, which directs cytoskeletal remodeling that contributes to a hypermigratory phenotype that facilitated dissemination of the parasite into multiple organs of the infection host. The role of IRE1 in Toxoplasma-induced hypermigration is not reliant on its protein kinase or endoribonuclease activities that are central for classical UPR signaling; rather, it is IRE1 oligomerization and the carboxy- Toxoplasma Infection Induces Host UPR and Migration ® terminal residues required for filamin A association that prove to be important. Further mechanistic details into how ER recruitment to the PV elicits host UPR remain to be resolved.
After oral infection, Toxoplasma rapidly spreads from the lamina propria to distal organs using host immune cells as a vehicle for dissemination (22). Given our discovery that IRE1 is mobilized by Toxoplasma to enhance hypermigratory behavior in host cells, we tested whether IRE1 is crucial to in vivo dissemination in a mouse model of infection. We found that depletion of IRE1 in immune cells sharply decreases the number of parasites in the spleens or brains of infected mice (Fig. 5E). These results reveal a number of potential new targets for drug development aimed at thwarted spread of infection in the body.
It is noteworthy that other mechanisms have been suggested to contribute to the cell hypermotility upon Toxoplasma infection. For example, parasite effector protein TgWIP, TIMP-1, and GABAergic signaling was reported to signal hypermigration of certain infected host cells (23)(24)(25)(26). A key question for future studies is how these different mechanisms are coordinated to induce hypermigratory activity in Toxoplasmainfected host cells.

MATERIALS AND METHODS
Host cell and parasite culture. Mouse embryonic fibroblasts (MEFs) were cultured in Dulbecco's modification of Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Gibco/Invitrogen) and penicillin-streptomycin at 37°C with 5% CO 2 . PERK Ϫ/Ϫ cells were previously reported (27), and IRE1-deficient MEFs were engineered by CRISPR as described below. Host cells were seeded at a density of 2 ϫ 10 5 /well in a 6-well plate and cultured for 18 h. Infection was performed using a multiplicity of infection (MOI) of 3 with type I or II (RH or ME49) strain Toxoplasma parasites for 18 h. The infected cells were cultured in DMEM supplemented with 10% heat-inactivated FBS (Gibco/ Invitrogen) and penicillin-streptomycin at 37°C with 5% CO 2 . Cultures of DCs and J774.1 macrophages were cultivated in RPMI medium supplemented with 10% heat-inactivated FBS (Gibco/Invitrogen) and penicillin-streptomycin at 37°C with 5% CO 2 and infected as described above.
Generation of IRE1 knockout cells. Disruption of the IRE1 gene in MEFs was carried out using the CRISPR/Cas9 method (28). Two distinct sgRNAs designed using DESKGEN toll (g1-TGGACACGGAGCT GACT and g2-ACACGGAGCTGACTGGG) were examined individually. The sgRNAs were prepared using the EnGen sgRNA synthesis kit (New England BioLabs), along with an sg control (g-control-CATCCTCGGCA CCGTCACCC). The sgRNAs were then associated with EnGen Spy Cas9 NLS protein (New England BioLabs) at room temperature for 20 min. MEFs were then transfected with the bound sgRNA/Cas9 protein using the Lipofectamine CRISPRMAX Cas9 transfection reagent (Thermo Fisher Scientific). After culturing the transfected cells for 48 h, 100 cells were plated in 10-mm tissue-culture dishes for cloning. Cloned IRE1 Ϫ/Ϫ MEFs were validated by RT-qPCR using specific primers and by immunoblotting using IRE1specific antibody (Abcam; ab37073). IRE1 Ϫ/Ϫ cells were complemented with pcDNA3-derived vectors containing WT or mutant versions of IRE1. Briefly, the mouse cDNA sequence of IRE1 from MEFs was inserted into pcDNA3-EGFP plasmid (Addgene; number 13031), resulting in fusion proteins with EGFP fused to the carboxy terminus of IRE1. Mutations in IRE1 include changes inactivating the critical functions, including kD (kinase domain; K599A), eD (endoribonuclease domain; P830L), oD (oligomerization domain; D123P), and c-terminal (carboxy-terminally truncated Δ965), were carried out using specific primers (Table S1) and the Q5 site-directed mutagenesis kit (New England BioLabs). After sequence verification, the plasmids were transfected into IRE1 Ϫ/Ϫ cells using FuGENE 6 transfection reagent. Rescued WT and mutant IRE1 protein expression were confirmed by immunoblotting and immunofluorescence microscopy ( Fig. S4A to C).
To generate bone marrow-derived dendritic cells (DCs), 10 ϫ 10 6 bone marrow cells were isolated and cultured in a 6-well plate in 3 ml of complete medium (RPMI 1640 medium supplemented with 10% fetal bovine serum, penicillin, streptomycin, glutamine, 2-mercaptoethanol, 20 ng/ml of granulocytemacrophage colony-stimulating factor [GM-CSF], and 5 ng/ml of interleukin 4  [both from Peprotech]) for 7 days. Half of the medium was replaced every 2 days with medium supplemented with GM-CSF and IL-4, as previously described (29). DCs or J774.1 macrophages were transfected with IRE1-sgRNA1 or -2, associated with EnGen Spy Cas9 NLS protein (New England BioLabs), using the 4D-Nucleofector system (Lonza) in combination with the P3 Primary Cell 4D-Nucleofector X kit. After 48 h, the IRE1 mRNA and protein levels were measured by RT-qPCR and immunoblotting, respectively. Viability of DCs was examined by trypan blue staining.
Measurement of mRNA levels. Cells were first infected with Toxoplasma for 2 h, washed with phosphate-buffered saline (PBS), and then cultured in DMEM at the desired time points. RNA was isolated from cells using TRIzol LS reagent (Invitrogen), the cDNA was then generated using Omniscript (Qiagen), and RT-qPCR was performed using SYBR green real-time PCR master mixes (Invitrogen) and the StepOnePlus real-time system (Applied Biosystems). Oligonucleotide primers used to measure each target mRNA are listed in Table S1. Relative levels of transcripts were calculated with the threshold cycle (⌬⌬C T ) method using genes encoding glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and ␤-actin as internal controls, and then the relative levels of target mRNAs from the mock-infected samples were adjusted to 1 and served as the basal control value. Values of each time point were normalized to mock infection. Each experiment was performed three times, each with three technical replicates.
Immunoblot analyses. Cells were infected with Toxoplasma for 2 h, washed with PBS, and then cultured in DMEM for the desired times. The infected cells were harvested in radioimmunoprecipitation assay (RIPA) buffer solution supplemented with cOmplete, EDTA-free protease inhibitor cocktail (Roche). Protein quantification was performed using Bradford reagent (Sigma-Aldrich). Equal amounts of protein lysates were separated by SDS-PAGE, and proteins were transferred to nitrocellulose filters. Immunoblot analyses were done using primary antibodies against IRE1 (Abcam; ab37073), XBP1s (Cell Signaling; number D2C1F), ATF6 (30), GAPDH (Abcam; ab9485), and PERK (Cell Signaling; number 3192), followed by Amersham ECL horseradish peroxidase (HRP)-conjugated secondary antibody. These antibodies and additional reagents used in the study are listed in Table S2. Proteins in the immunoblots were visualized using FluorChem M (multiplex fluorescence; Protein Simple). Immunoblot analyses were carried out for three independent experiments.
Calcium measurement assay. Values derived from infected cells (⌬F) were divided by the resting intracellular calcium (Fo), ΔF/Fo, and the values of each time point were normalized to the value for mock-infected cells. In parallel, live infected cells were imaged by microscopy at the same exposure and a heat map was generated using ImageJ software. To determine the activity of RyR and IP 3 R, infected cells were incubated with 5 M Mag-Fluo-4, a low-affinity Ca 2ϩ indicator, and then permeabilized with 10 g/ml of saponin followed by incubation with 1.5 mM ATP to maintain Mag-Fluo-4 in the ER (19) (Fig. S2B). Infected cells were treated with the desired concentrations of caffeine (RyR, 0 to 200 mM) or IP 3 (IP 3 R, 0 to 3 M) (Sigma-Aldrich). A Synergy (BioTek) plate reader was used to monitor the Mag-Fluo-4 fluorescence at 490-nm excitation and 525-nm emission wavelengths (19). Values were normalized to those of mock-infected cells.
Immunofluorescence assay. Cultured cells were infected with Toxoplasma for 18 h, then fixed with 2.5% paraformaldehyde for 20 min, and blocked with PBS supplemented with 2% bovine serum albumin (BSA). Cells were permeabilized in blocking solution containing 0.01% Triton X-100 for 30 min and incubated with primary antibody (SAG1-p30; Invitrogen) for 1 h. Secondary goat anti-rabbit Alexa Fluor 488 (Invitrogen) was applied for 1 h in the presence of rhodamine phalloidin (Thermo Fisher Scientific) followed by Prolong gold antifade reagent (Invitrogen). 4=,6-Diamidino-2-phenylindole (DAPI) was used to visualize host cells and parasite nuclei (Vector Labs). Images were acquired with a Leica inverted DMI6000B microscope with a 63ϫ oil immersion objective and analyzed in ImageJ. Alternatively, IRE1 Ϫ/Ϫ cells were transfected with pcDNA3 encoding IRE1 fused with EGFP at the carboxy terminus and infected with Toxoplasma for 18 h (MOI, 3); the cells were then fixed and imaged as described above.
Immunoprecipitation assay. The mouse cDNA sequence of filamin A from MEFs was amplified and cloned into pcDNA3-myc plasmid (Addgene). The resulting plasmid, pcDNA3-myc-FLNA, was transiently transfected in MEFs, and then the transfected cells were infected with Toxoplasma for 18 h (MOI, 3). Cell lysates were prepared using IP-lysis solution (0.5% NP-40, 250 mM NaCl, 30 mM Tris, 0.5% glycerol [pH 7.4], 250 mM phenylmethylsulfonyl fluoride [PMSF] supplemented with cOmplete, EDTA-free protease inhibitor cocktail [Roche]). To immunoprecipitate Myc-tagged filamin A (Myc-filamin A), equal amounts of protein lysates were incubated with IgG magnetic beads (Pierce) for 2 h and then mixed with anti-Myc magnetic beads (Pierce) overnight at 4°C with rotation. Proteins bound to the beads were subsequently washed four times with IP-lysis solution at 4°C and then once with IP-lysis solution supplemented with 500 mM NaCl. Protein complexes were eluted at 95°C for 5 min in loading buffer solution and then separated by SDS-PAGE, followed by immunoblot analyses using specific antibodies to IRE1 (Abcam; ab37073) or Myc (Cell Signaling; no. 2276).
Cell migration assay. Cells were infected with Toxoplasma at an MOI of 3 for 18 h and then trypsinized and counted using a hemocytometer; 2 ϫ 10 4 cells were resuspended in serum-free medium and applied to the top of a membrane coated with collagen I (rat tail) (Gibco; A1048301). Transmigration assays were carried out using a Corning Transwell Costar apparatus (6.5-mm diameter and 8-m pore size) as described previously (32). After 18 h for MEFs and 6 h for DCs and macrophages, the medium was removed, and the cells were fixed with 2.5% paraformaldehyde for 20 min. To facilitate counting of migrated cells, cells that did not migrate and remained on the upper side of membrane (unmigrated cells) were removed with a swab. The membrane was incubated with Prolong gold antifade reagent with DAPI. Cells were counted using a Leica inverted DMI6000B microscope with a 63ϫ oil immersion objective. The transmigration was determined by numbers of migrated infected cells in 5 fields normalized to number of uninfected cells in each cell type and treatment. Each transmigration assay was carried out in technical triplicate (n ϭ 3).
In vivo migration assay. DCs were plated in 6-well plates (1 ϫ 10 6 cells/well) and allowed to adhere overnight. DCs were infected with Toxoplasma for 1 h (MOI, 3) and then washed with RPMI medium to remove extracellular parasites. After 18 h, infected DCs were incubated with CellTracker Orange CMTMR (Thermo Fisher) as described previously (33). Infected DCs were intraperitoneally inoculated into 6-weekold female C57BL/6J mice whose spleens and brains were subsequently harvested at the indicated time points. DC migration to the spleen was measured by fluorescence intensity using a Synergy (BioTek) plate reader at excitation and emission wavelengths of 541 and 565 nm (Fig. S5H). Also, DNA was isolated from the spleen and brain using TRIzol (Thermo Fisher), and the number of parasites was determined by using a PCR-based method measuring levels of the parasite-specific gene region B1 as previously described (34). After 3 days, the mice were observed twice a day and percent survival was recorded at each time point. The mouse experiments, including parasite measurement by B1, were performed blinded. The mice used in this study were housed in American Association for Accreditation of Laboratory Animal Care (AAALAC)-approved facilities at the Indiana University School of Medicine Laboratory Animal Research Center (LARC). The Institutional Animal Care and Use Committee (IACUC) at Indiana University School of Medicine approved the use of all animals and procedures (IACUC protocol number 11376).
Quantification and statistical analysis. Quantitative data are presented as means and standard deviations and were derived from three biological replicates. Statistical significance was determined using one-way analysis of variance (ANOVA) with Tukey's post hoc test and multiple two-tailed t test using Graph Prism software. The number of biological replicates and P values are indicated in figure legends. For immunoblot analyses, the reported images are representative of at least three independent experiments. The mouse survival curve was analyzed by Gehan-Breslow-Wilcoxon test for in vivo analysis.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.