A Toxoplasma dense granule protein, GRA24, modulates the early immune response to infection by promoting a direct and sustained host p38 MAPK activation

The gene TGME49_030180 encoding GRA24 was originally identified in silico together with GRA16 (Bougdour et al., 2013) as strong parasite candidate genes with attributes to both target the parasite secretory pathway and to reach the nuclei of T. gondii tachyzoite-hosting cells. GRA24 encodes a 542-aa protein with a predicted N-terminal signal peptide, an internal putative bipartite nuclear localization sequence, and two repeats, R1 and R2, at the C-terminal end (Fig. 1 A). GRA24 protein is unique because it has no significant homology with any proteins, not even with the close relative Neospora caninum proteins. In Toxoplasma, GRA24 harbors significant polymorphism between the allele shared by types II and III clonal lineages and the allele typifying type I clonal lineage (Fig. 1 B). In fibroblasts hosting tachyzoites expressing a carboxy-terminal HAFlag-tagged version of GRA24, the protein is distinctly detected in the host cell nucleoplasm at 16 h after invasion and later, whereas there is no detectable protein left in the intracellular tachyzoites (Fig. 2 A). In contrast, in free tachyzoites, GRA24-HAFlag delineates punctuate structures throughout the parasite cytoplasm which occasionally overlap with the canonical DG proteins GRA1 and GRA7 (Fig. 2, D and E) while being excluded from the two typical apical secretory organelles using toxofilin and MIC2 as markers for rhoptry and microneme, respectively (Fig. 2, B and C). Interestingly, when overexpressed under the strong constitutive GRA1 promoter, a significant fraction of GRA24 is seen in the PV space similarly to GRA1, thus suggesting that GRA24 export across the PVM is a limiting step (Fig. 2, F and G). We concluded from these data that GRA24 is a novel DG-like protein that crosses the PVM of maturing vacuoles to reach the host cell nucleus, thereby belonging to the recently defined GRA exported protein subfamily typified by GRA16 (Bougdour et al., 2013).

Because GRA24 operates as a secreted DG that crosses the PVM and traffics to the host cell nucleus, we next sought partners of GRA24 by applying chromatography and ion-trap mass spectrometry to nuclear extracts that were prepared from inducible GRA24-HAFlag–expressing 293 cell lines. After induction, the protein localizes in the nucleus (Fig. 3 A), and regardless of the tag position (N- or C-terminal fusion), GRA24 is embedded in a remarkably robust molecular mass complex (400 kD, fractions 22–28; Fig. 3, B and E), which contains the CSBP2/SAPK2α isoform, also named p38α, as the only GRA24 relevant partner (Fig. 3 C). Western blotting confirmed the partnership between GRA24 and p38α (Fig. 3, B and E). The GRA24–p38α complex was also isolated from primary human foreskin fibroblasts (HFFs) and mouse J774 MØ line hosting GRA24-HAFlag–expressing tachyzoites, demonstrating that the complex forms not only in stromal cells but also in macrophages hosting cell-cycling tachyzoites (Fig. 3 D). Using the dual phosphorylation of Thr180/Tyr182 as a hallmark of p38 activation (Coulthard et al., 2009), we confirmed by immunoblotting that p38α in complex with GRA24 is activated (i.e., phospho Thr180/Tyr182 positive) when purified from inducible GRA24-HAFlag–expressing 293 cells (Fig. 3, B and E) or from infected murine and human cells (Fig. 3 D).

To investigate the role of GRA24 on host cell p38α activation during infection, WT, GRA24 knockout, or complemented tachyzoites were generated. The GRA24 gene was disrupted by homologous recombination (Δgra24) in both type I (RHku80) and type II (Pruku80) strains while the mutant was complemented with a single copy of a GRA24 functional allele driven by its endogenous promoter (Fig. S1 and Table S1). All strains behave similarly in vitro, indicating that GRA24 is not essential for either invasion or intracellular survival (unpublished data). Consistent with our previous data, no phospho-p38α population is ever detected in HFF and J774 MØ cells hosting GRA24 knockout tachyzoites in contrast to cells hosting GRA24 WT virulent type I or cystogenic type II parasite strains (Fig. 4, A and B). Importantly, reintroduction of the GRA24 gene in the mutant restores a parasite’s ability to induce p38α phosphorylation (Fig. 4, C and D), whereas p38α was never found to be activated by Neospora caninum regardless of the multiplicity of infection (MOI) or the duration of infection (Fig. 4, C and D). Remarkably, although p38α activation usually occurs transiently and within minutes in response to most stimuli (Ferreiro et al., 2010), GRA24-dependent p38α phosphorylation is, in contrast, impressively sustained as late as 24 h after the swift tachyzoite invasion step (Fig. 4), a trait which fits well with the unusual long-lasting presence of GRA24 into the host cell (Fig. 2 A) and suggests a direct control on the p38α kinase activity.

The prototypic p38α activation is achieved by upstream MAPK kinase (MKK3/6; Ashwell, 2006) or, alternatively, through TAB1-mediated p38α autophosphorylation (Ge et al., 2002). Another pathway, restricted so far to T cells upon T cell receptor activation, includes phosphorylation of Tyr323 by the ZAP-70 tyrosine kinase that leads to p38 autophosphorylation but solely on Thr180 (Salvador et al., 2005). The host p38α autophosphorylation, observed when tachyzoites are nested in their PV, was proposed to reflect a TAB1-dependent mechanism, thus a mechanism bypassing the MKK-dependent MAPK phosphorylation (Kim et al., 2005). In T. gondii–hosting J774 MØ cells exposed to SB203580, a competitive inhibitor of ATP binding blocking p38α autophosphorylation but not trans-phosphorylation by MKKs (Gum et al., 1998), there is a complete abrogation of GRA24-induced p38α phosphorylation, a phenotype which was not observed when cells were treated with PD98059 and SP600125, which have been shown to block the MEK1/2 and the JNK pathways, respectively (Fig. 5, A and B). In addition, expression of GRA24 is sufficient to trigger p38α autophosphorylation because SB203580 inhibits p38α activation in 293 cells induced for GRA24 synthesis, whereas MEK1/2 and JNK inhibitors do not (Fig. 5, C and D). GRA24 is the first molecular switch originating from a eukaryotic parasite that directly activates p38α in a sustained way independently of the prototypic kinase cascade in both human stromal cells and mouse macrophages.

To detail further how GRA24 binding to p38α promotes activation of the kinase, we mapped the GRA24 region interacting with the enzyme. To this end, HAFlag epitope-tagged carboxy-terminal truncated forms of GRA24 were stably expressed in RHku80 parasites (Fig. 6 A). The two chimeric proteins, namely GRA24 ΔR2 and GRA24 ΔR1/R2, migrate at their expected molecular weights (Fig. 6 C) and localize in the host cell nuclei after infection (Fig. 6, B and C), thereby indicating that the removed domains do not affect the export and the transit into the host cell of the chimeric proteins. However, although deleting only repeat R2 has no effect on p38α binding and activation, double deletion of R1 and R2 abolishes GRA24 binding property to p38α (Fig. 6 C), preventing any p38α phosphorylation (Fig. 6 B), thus pointing to the required contribution of GRA24 R1 domain to p38α activation. These observations prompted us to look for the molecular determinants within the repeat that drive the interaction with/activation of p38α. We identified a perfect replicate of the canonical KIM (also known as D motif, Fig. 6 D) defined as a key MAPK-interacting structural element that, in addition, provides specificity toward the MAP kinase partners (Reményi et al., 2005; Francis et al., 2011). The motif is typified by a cluster of basic residues at the N terminus and hydrophobic amino acids near the C terminus (Fig. 6 D). The GRA24-KIM clearly belongs to the tyrosine phosphatase PTP-KIM class (defined in Garai et al., 2012; Fig. 6 D) but displays an unusual insertion of a proline (Pro-451) between hydrophobic canonical residues ϕL and ϕA that, in tandem with Pro-452, confers local rigidity to the linear peptide embedded in the KIM binding site (Fig. 6 D) and, most likely, adds specificity toward p38α (Garai et al., 2012). Our docking models validate the proper binding of GRA24-KIMs to p38α (Fig. 7). More specifically, the two N-terminal arginine residues (Arg-444 and Arg-445) share salt bridges with CD-site Asp-316 and Glu-81 of p38α, whereas both Leu-450 (ϕL) and Leu-453 (ϕA) insert optimally into the shallow hydrophobic pockets (Fig. 7, A and B). Finally, Ile-455 (ϕB) reinforces the hydrophobic interactions between the two partners, as previously reported for the He-PTP Val-31 residue (ϕB; Francis et al., 2011). Collectively, peculiar GRA24-KIM structural features empower GRA24 ability to bind to p38α, thereby accounting for the remarkable resistance to the highly stringent conditions used during the GRA24–p38α complex purification (Fig. 3, B and E). Such binding properties are expected to potentiate p38α autophosphorylation, as KIM-mediated interaction with p38α has been shown to trigger local (Fig. 7 E) as well as long-range conformational changes, in particular within the activation loop that becomes disordered, therefore exposing the phosphorylatable Thr-180 and Tyr-182 residues (Chang et al., 2002; Zhou et al., 2006; Zhang et al., 2011). Of note, the KIM repeat itself, i.e., within R1 and R2 (Fig. 6 A), might favor p38α trans-autophosphorylation via the simultaneous binding of two p38α monomers (Diskin et al., 2007).

The unique features of the unraveled interaction between GRA24 and p38α and the consequences on p38α activation status prompted us to analyze whether the GRA24–p38α complex contributes to the typical changes in gene expression observed in either stromal or phagocytic cells infected with tachyzoites (Blader et al., 2001; Saeij et al., 2007; Jensen et al., 2011). To address this possibility, we performed a genome-wide expression profiling of mouse BM-derived macrophages (BMDMs) hosting parental versus Δgra24 parasites from both type I and type II strains (Fig. 8 A). We focused our analysis on genes that were altered with more than a twofold change when comparing each Δgra24 mutant with their respective parental strains. KEGG analysis highlights 10 pathways as significantly and selectively induced (P < 0.05) in a GRA24-dependent fashion, a majority of which regulates proinflammatory cytokines (e.g., IL-6, IL-12b, and TNF) and chemokines (e.g., MCP-1, CCL5, CXCL1, and CXCL10) known to operate in any sterile tissues colonized by invasive microbes, whether they are nonparasitic or parasitic (Fig. 8 C). Interestingly, the GRA24-dependent changes are exacerbated in cells infected with the type II clonal lineage, possibly as a consequence of allelic polymorphism between the strain types (Fig. 1 B), which leads to a concomitant up-regulation of 13 genes linked to M1-MØ activation and down-regulation of 17 genes associated with the alternative M2–MØ pathway (Fig. 8 B). We have also noticed that both IFN-γ– and iNOS-coding transcript populations were also up-regulated in a GRA24-dependent manner by 12- and 3-fold, respectively (Fig. 8, A and B). Of note, at least within the time frame (24 h) examined in these experiments, the BMDM supernatants display no difference in IFN-γ levels as assessed by ELISA (unpublished data).

As stated above, in laboratory mice, the T. gondii– and mouse-driven generation of immune type 1 effectors that prevent both uncontrolled tachyzoite population expansion and subsequent host death is characterized by the IL-12p40 production by tachyzoites-hosting macrophages (Scharton-Kersten et al., 1996; Sher et al., 2003; Yap et al., 2006). GRA24 is clearly playing a major role in regulating IL-12p40 synthesis, as its loss severely compromises both IL-12 mRNA and protein levels in mouse BMDM (Fig. 8 B and Fig. 9 A). Similarly, on day 4 after infection, mice infected with a Pruku80 Δgra24 strain had significantly less IL-12p40 in their i.p. cavities and serum than mice infected with the parental strain (Fig. 9 B). However, despite an efficient GRA24-mediated p38α activation (Fig. 4), type I parasites fail to trigger IL-12 secretion in BMDM (Fig. 8 A and not depicted), a phenotype which might reflect a dominant IL-12 suppressive activity driven by the ROP16 type I allele (Saeij et al., 2007; Yamamoto et al., 2009; Butcher et al., 2011). In addition, other effectors, such as GRA15, are involved in the control of IL-12 synthesis through the activation of the NF-κB pathway (Rosowski et al., 2011). Although GRA15 does not trigger p38α phosphorylation (Fig. 9 C), GRA24 does not affect NF-κB (not depicted), highlighting that several T. gondii strain-specific or -shared effectors display synergistic or antagonist effects on the regulation of IL-12p40–coding gene expression. In support of GRA24 key contribution to the IL-12 response, GRA24 is found to critically account for the typical increase in MCP-1, CXCL1, CCL5, and CXCL10 chemokine synthesis (Fig. 9 D) triggered by a type II tachyzoite lineage. Remarkably, MCP-1 was shown to play a substantial role in the recruitment, to T. gondii–loaded tissues, of Gr-1⁺ inflammatory monocytes that, in turn, produce nitric oxide and are essential for the local control of the tachyzoite population size (Robben et al., 2005; Dunay et al., 2008). The above results argue that the GRA24-dependent MCP-1 secretion might be involved in the recruitment of Gr-1⁺ monocytes in tissues transiently remodeled as cell-cycling tachyzoite-protective niches. To examine the functional consequences in vivo of GRA24-dependent MCP-1 secretion, we challenged C57BL/6 mice with 10³ tachyzoites of Pruku80 WT or Pruku80 Δgra24 strains and assessed parasite numbers in the peritoneal exudate as well as in the spleen. A larger number of parasites are detected beyond day 7 in the peritoneal contents of mice infected with Pruku80 Δgra24 than in mice infected with the parental strain (Fig. 9 E). Parasites are also found in the spleen, indicating that the infection becomes quickly systemic, but in this tissue no significant deviation in parasite loads was detected between the WT and the Δgra24 strains at any of the time points analyzed (Fig. 9 E). These results suggest that the deletion of GRA24, while suppressing MCP-1 production, most likely prevents Gr-1⁺ recruitment and contributes to impair the ability of the mice to suppress parasite numbers at the site of infection in the gut. We then assess if GRA24 is altering the overall host survival of C57BL/6 mice. A dose of 5 × 10² tachyzoites required to achieve a near LD₁₀₀ when using the Pruku80 strain was i.p. inoculated to C57BL/6 mice whose survival was monitored over the acute phase of infection. Mice infected by either the WT Pruku80 or Pruku80 Δgra24 parasites succumb to infection within 10 to 20 d (unpublished data; n = 6 mice per strain). Therefore, GRA24-deficient parasites did not fully recapitulate the phenotype of MCP-1^−/− and CCR2^−/− mice that were shown to be profoundly susceptible to acute toxoplasmosis (Robben et al., 2005; Dunay et al., 2008). This difference might be a result of the genetic background of the parasite strains, the type II Pruku80 strain (Fox et al., 2011) being more virulent and less cystogenic than the type II ME49 strain used in most Toxoplasma infection including those using MCP-1^−/− and CCR2^−/− mice (Robben et al., 2005).

Finally, as a link between the cytokine response and the direct GRA24-dependent p38α activation, we identified several transcription factors up-regulated in a GRA24-dependent fashion in MØ, among which those that either control the immediate early gene response or are immediate early prototype gene products like Egr-1 and c-Fos (Fig. 10 A). Of note, first, Egr-1 and c-Fos belong to a restricted set of 30 core genes whose expression is equally regulated by major p38α MAPK-activating stimuli (i.e., osmo-stress, the cytokine TNF, and the protein synthesis inhibitor anisomycin; Ferreiro et al., 2010), and second, the Egr-1 promoter activity is blocked by a pharmacologically inactivated p38α kinase (Tur et al., 2010). These observations strongly argue for a direct function of GRA24 on classical p38α MAPK-regulated genes and confirm previous studies on the induction of Egr-1 and its paralog Egr-2 during Toxoplasma infection (Phelps et al., 2008; Wiley et al., 2011). By Western blotting and immunofluorescence, we indeed confirmed at the protein level that Egr-1 is induced in the nuclei of human stromal cells and mouse macrophages hosting WT T. gondii, a phenotype which is not induced in cells hosting Δgra24 Toxoplasma or Neospora (Fig. 10, B and C; and not depicted). Reintroduction of the GRA24 gene restored the Toxoplasma ability to up-regulate Egr-1 (Fig. 10 B). Egr-1 induction was also validated in 293 cells conditionally expressing GRA24 (Fig. 10 D), and the level of expression was reduced in a dose-dependent fashion by the inhibitor SB203580, thereby establishing the link between Egr-1 and the autophosphorylation of p38α by GRA24 (Fig. 5 B). Finally, we identified the GRA24-KIM as the basis of p38α binding and autophosphorylation that further entails Egr-1 activation (Fig. 6 B). Besides Egr-1, c-fos is also regulated at the transcriptional level by GRA24 (Fig. 10 A). In particular, this transcription factor is phosphorylated on serine 32 in a GRA24-dependent fashion in both murine MØ and human fibroblasts (Fig. 10 E and not depicted). Unexpectedly, Toxoplasma elicits a prolonged induction of Egr-1 and c-Fos, which correlates with the nuclear persistence of the GRA24–p38α complex during infection but differs from the transient usual canonical activation of these genes. The induction of these transcription factors by GRA24 is most likely followed by a second wave of host cell transcriptome alteration because both Egr-1 and c-Fos then operate in a hierarchical manner by inducing the expression of their downstream genes (e.g., Egr-1-regulated genes, see GEO accession no. GSE38782).

Host cell invasion is a critical step by which Toxoplasma zoites rapidly establish themselves and further complete their developmental program as bradyzoite-containing cysts known to be located in the tissue of prey that are ingested by predators. Recent research has highlighted that early during the invasion process, the parasite secretes several rhoptry bulb resident proteins with a majority of kinases and phosphatases that co-opt host cells by interfacing with their signaling pathways (Melo et al., 2011). Once internalized, a Toxoplasma tachyzoite resides within a dynamic PV that is constantly reshaped through molecular exchanges with the host cell cytoplasm and that support parasite growth. Although nonenzymatic molecules secreted by DG were first thought of as parasite effectors restricted to the PV biogenesis and maturation, a recent study (Bougdour et al., 2013) and the present study highlight that DG-resident nonenzymatic proteins are secreted into the vacuole space, cross the PVM, and reach the host cell nucleus where they modulate host gene expression. Thus, GRA16 and GRA24 emerge as a subfamily of DG-scaffolding proteins exported beyond the tachyzoite-hosting vacuole to operate as effector molecules in infected cells.

In the present study, we have identified GRA24 as a strong interactor of host p38α, with the resulting complex acting as a regulator of gene expression in the host cell nucleus. Indeed, we establish that GRA24 interacts with the p38α enzyme, triggers its autophosphorylation, and consequently its activation. GRA24 bypasses the typical MKK-dependent MAPK phosphorylation cascade upstream of p38 and is the first parasitic microbe agonist so far described to promote p38α autophosphorylation through a dimeric stable complex. Interestingly, GRA24 specifically promotes Thr180 autophosphorylation of p38α because it neither activates ERK1,2 (Thr202/Tyr204 and Thr185/Tyr187, respectively) nor alters the JNK (Thr183/Tyr185) phosphorylation (unpublished data). In agreement with these results, none of the aforementioned kinases copurify with GRA24, reinforcing the notion of a high specificity of GRA24 function on p38α. Another unique feature of GRA24 function lies in its property to promote a sustained phosphorylation of p38α that correlates with the long-lasting secretion of GRA24. This kinetics of activation is remarkably distinct from the transient activation mediated by the prototypic MAPK pathway and has important consequences because it likely avoids conventional negative feedback by phosphatases. Consistent with this idea, we provide evidence that the carboxy-terminal repeat of GRA24 is essential for binding to p38α and its subsequent autophosphorylation (Fig. 6 B). Furthermore, molecular modeling reveals that GRA24 repeat contains a p38α docking KIM motif that is phylogenetically related to those of the PTP tyrosine phosphatase family (Fig. 6 D). Collectively, these data suggest that GRA24 has evolved a structurally optimized KIM domain that might be able to compete accurately with the phosphatases for binding to MAPK.

Toxoplasma-induced p38α activation was initially reported as TAK1- and Myd88-independent and proposed to occur through an association with the scaffold protein TAB1 (Kim et al., 2005). But—of note—first, TAB1 does not activate p38α after TNF exposure of fibroblasts derived from the MKK3-MKK6 double knockout mice (Brancho et al., 2003), and second, once bound to p38α, TAB1 prevents the p38α nuclear localization (Lu et al., 2006) disproving the validity of this proposal. In contrast, our many complementary data allow us to highlight a major role for GRA24 as the shuttling and scaffolding protein that allows not only the p38α nuclear localization but also its autophosphorylation, at least in two distinct cell lineages—fibroblasts and macrophages—that T. gondii tachyzoites do subvert as host cells. Moreover, a contribution of parasite-derived tyrosine kinase is also very unlikely because bypassing the recourse to tachyzoites, through the ectopic expression of GRA24 in mammalian cells, is clearly sufficient to promote p38α autophosphorylation. It is noteworthy that GRA24 has multiple alternative splice variants. Specifically, the presence of the exon E3b in the mRNA isoform 2 introduces a premature stop codon, leading to a protein variant devoid of the terminal repeats R1 and R2 (Fig. 1 A). In this fashion, the parasite might use alternative splicing to turn off GRA24-dependent p38α autophosphorylation under certain growth or developmental conditions, which remain to be discovered.

The ability of GRA24 to directly promote host p38α autophosphorylation places it as a potential regulator of the synthesis of growth factors and/or cytokines/chemokines and associated receptors, at least in macrophages once they are subverted as tachyzoite-hosting cells. Using microarray analysis, we showed that GRA24 controls, at the transcriptional level, an impressive network of genes encoding for cytokines/chemokines (Fig. 8), well beyond those associated previously with the secretion of GRA15 (Jensen et al., 2011; Rosowski et al., 2011). Chief among the cytokines is the proinflammatory cytokine IL-12p40, with a 20-fold transcript excess that typifies a robust M1 profile in MØ hosting type II tachyzoites. As reported previously, IL-12 synthesis induction by Toxoplasma-infected MØ occurs at lower levels and with distinct kinetics relative to endotoxin stimulation; in brief, although IL-12 levels peaks within 6 h of LPS stimulation (Kim et al., 2005), parasite-induced IL-12 is detected within a 24–40-h window after macrophage invasion (Fig. 9 A and Kim et al., 2005). This timing is consistent with the kinetics of GRA24 secretion and the requisite for de novo synthesis of p38-activated transcription factors driving IL-12 synthesis. Of note, once macrophages are hosting tachyzoites, both IL-1β and IL-18 signaling use MyD88 as an adaptor, leading to NF-κB and MAPK activation and subsequent IL-12 production (Adachi et al., 1998). It is also noteworthy that IL-18 is approximately sixfold induced by GRA24 in MØ infected with type II strain (Fig. 8 A). It is therefore possible that autocrine production of IL-18 results in GRA24-mediated IL-12 release. Anyhow, the massive production of IL-12 induced by type II strains when tachyzoites start to proliferate is most likely controlled by the synergistic action of GRA15 and GRA24, which keep parasite numbers low, thus facilitating long-term parasitism. This is consistent with the observation that GRA24 is one among the few polymorphic highly expressed genes that are under positive selection in Toxoplasma population (Minot et al., 2012). We can also speculate that GRA24 loss in Neospora caninum correlates with the inability of Neospora to properly modulate the host immune response, an argument which has been put forward earlier to explain why Neospora has a more limited host range than Toxoplasma (Reid et al., 2012).

Besides triggering a strong Th1 response, GRA24 also significantly contributes to suppress M2 markers at the transcriptional level in MØ hosting type II tachyzoites, a situation which powers the M1 response. The repertoire of M2 genes that are down-regulated in a GRA24-dependent manner again significantly exceeds that of GRA15, with Arg1 being distinctively ∼60-fold repressed by GRA24 (Jensen et al., 2011; Fig. 8 B). However, neither arginase-1 nor the mannose receptor type C (Mrc1/CD206) are induced at the protein level in Δgra24-infected BMDMs (unpublished data), suggesting that the suppression of the GRA24-dependent transcriptional repression of M2 genes in type II strain is not sufficient to induce the production of the corresponding protein product, consistent with previous studies (Butcher et al., 2011; Jensen et al., 2011). Additionally, GRA24 is found to critically account for the typical increase at the protein level of four chemokines including MCP-1/CCL2 (Fig. 8 A and Fig. 9 D). Remarkably, in mice lacking MCP-1 or its receptor CCR2, no recruitment, in T. gondii–hosting tissues, of inflammatory Gr-1⁺ monocytes, is detectable, a situation which likely accounts for the sustained parasite burden and early host mortality (Robben et al., 2005; Dunay et al., 2008). Consistent with these data, we show that mice infected with Δgra24 tachyzoites fail to control parasite numbers in the gut (Fig. 9 E), revealing a contributory role for GRA24 in resistance to acute toxoplasmosis, most likely through the regulation of cytokine expression and secretion including MCP-1.

GRA24 also significantly induces the secretion of CXCL10/IP-10 (Fig. 8 A and Fig. 9 D). This chemokine is required for preventing uncontrolled T. gondii tachyzoite burden (Khan et al., 2000), and its depletion prevents T cell recruitment and effector function in mouse tissues where the tachyzoite to bradyzoite developmental transition is known to occur. In this regard, it was recently shown that the CXCL10/IP-10 chemokine enhances the ability of CD8⁺ T cells to control tachyzoite reexpansion in cyst-hosting mouse brains (Harris et al., 2012). Collectively, these studies underscore the importance of GRA24 in orchestrating a cell response during Toxoplasma infections by promoting the induction of a selective set of chemokines, which contributes to the recruitment of inflammatory monocytes previously described as a first line of defense controlling intestinal pathogens, including T. gondii (Dunay et al., 2008).

Given that GRA24 was involved in nonprototypical direct p38α activation, we next sought to trace the factors that act downstream of this unusual signaling pathway. This led us to show that several transcription factors, including Egr-1 and c-Fos (Fig. 10 A), inhabit key downstream positions within this coordinated acute inflammatory process driven by tachyzoites nested within macrophage PV. Several factors are needed for full, regulated transactivation of the Egr1 (Tur et al., 2010) or c-Fos (O’Donnell et al., 2012) genes. As such, the mitogen- and stress-induced phosphorylation of CREB (CRE binding protein) at Ser133, especially by p38α, has been linked to the transcription of several immediate early genes, including c-fos, junB, and Egr-1 (Tur et al., 2010; O’Donnell et al., 2012). In this regard, the Toxoplasma tachyzoite, in a GRA24-dependent fashion, positively affects the transcription of two CREB orthologs, i.e., CREB2/ATF2 and CREBL2 (Fig. 10 A), as well as CREB-dependent downstream genes (GEO accession no. GSE38782). We first thought that GRA24 might activate p38α, which in turn directly phosphorylates CREB-related factors in macrophages hosting tachyzoites. But this is apparently more complicated because Toxoplasma induces the phosphorylation of both CREB-Ser133 and ATF2-Thr-71 independently of GRA24 (Fig. 4 D). Therefore, GRA24 might act upstream in the pathway. It was previously shown that the phosphorylation of Ser 133 facilitates the ability of CREB to interact with CBP (CREB-binding protein) and its paralog p300, therefore leading to the transcription of CREB-responsive genes (Johannessen et al., 2004). CBP and p300 possess an intrinsic acetylation activity and are thought to acetylate specific histone tails, thereby assuring relaxation of the chromatin environment (Johannessen et al., 2004). As such, the sustained presence of GRA24-p38α in the nucleus of the tachyzoites-hosting macrophages might reflect the need for the complex to be embedded in the chromatin to operate in the vicinity of p38α MAPK-responsive genes. A GRA24-dependent response might alternatively be mediated through Egr-1 that physically and functionally interacts with CBP/p300 to modulate gene transcription (Silverman et al., 1998). Although interesting, this hypothesis deserves further investigation.

In conclusion, GRA24 extends the scope of the function of T. gondii DG proteins beyond their initial and dedicated role in PV formation and maturation. Moreover, the specific attributes of the GRA24-dependent p38α activation unveil a novel regulatory path for p38α-controlled transcriptional changes. This path not only shortcuts the canonical MAPK signaling cascade turned on in response to various stress-related stimuli and intracellular pathogens (Ferreiro et al., 2010; Krachler et al., 2011) but it also differs from the alternate p38 activation pathways described so far involving the scaffolding protein TAB1 or the T cell–restricted kinase ZAP-70 (Brancho et al., 2003; Salvador et al., 2005).

HFF primary cells, 293-TRex, and J774 cell lines were cultured in DMEM (Invitrogen) supplemented with 10% heat inactivated FBS (Invitrogen), 10 mM Hepes buffer, pH 7.2, 2 mM l-glutamine, and 50 µg/ml penicillin and streptomycin (Invitrogen). Cells were incubated at 37°C in 5% CO_2. BMDMs were obtained from female C57BL/6 mice. Bone marrow was isolated by flushing hind tibias and femurs using a 25-gauge needle followed by passages through an 18-gauge needle to disperse cell clamps. Cells were suspended in DMEM supplemented with 10% heat-inactivated FBS and 20 ng/ml recombinant M-CSF (Invitrogen), and incubated in a tissue culture-treated flask for 8–12 h. Then, nonadherent cells were harvested and transferred in 55 cm² non–tissue culture–treated plates (Corning) at 4–6 × 10⁶ cells per plate and further incubated at 37°C, with 5% CO₂ in humidified air. After 6 d, cells were washed with PBS to remove nonadherent cells, harvested by dislodging with a cell scraper in ice-cold PBS, and replated for the assay. This method yielded a highly pure population of F4/80⁺ macrophages by IFA. The Toxoplasma strains used in this study are listed in Table S1. All T. gondii strains were maintained in vitro by serial passage on monolayers of HFFs.

Antibodies against HA (3F10; Roche), MIC2 (provided by D. Sibley, Washington University School of Medicine, St. Louis, MO), Toxofilin (provided by I. Tardieux, INSERM, U1016, Institut Cochin, Paris, France), GRA1 and GRA7 (provided by J.-F. Dubremetz, UMR 5235 Centre National de la Recherche Scientifique, Montpellier, France), p38 MAPK (Cell Signaling Technology), Phospho-p38 MAPK (Cell Signaling Technology), Phospho-c-Fos (Ser32; Cell Signaling Technology), H4 Acetylated (EMD Millipore), TgHDAC3 (Bougdour et al., 2009), EGR1 (Cell Signaling Technology), and anti-TBP (Abcam) were used in the immunofluorescence assay and/or in Western blotting. Immunofluorescence secondary antibodies were coupled with Alexa Fluor 488 or Alexa Fluor 594 (Invitrogen). Secondary antibodies used in Western blotting were conjugated to alkaline phosphatase (Promega). The inhibitors SB203580, SP600125, and PD98059 were purchased from InvivoGen.

Immunofluorescence assays were performed as described in Bougdour et al. (2009). Cells grown on coverslips were washed in PBS and fixed for 20 min at room temperature with PBS containing 3% (vol/vol) formaldehyde. Fixed cells were permeabilized with PBS-0.1% Triton X-100 (vol/vol) for 10 min and blocked in PBS-3% BSA for 1 h at room temperature. Samples were incubated in PBS containing 3% BSA with the primary antibodies indicated in the figures, followed by the secondary antibodies coupled with Alexa Fluor 488 or Alexa Fluor 568 (Invitrogen) at a 1:1,000 dilution each in PBS–3% BSA. Nuclei of both host cells and parasites were stained for 10 min at room temperature with Hoechst 33258 at 2 µg/ml in PBS. After four washes in PBS, coverslips were mounted on a glass slide with Mowiol mounting medium. Images were acquired with a fluorescence microscope (Axioplan 2; Carl Zeiss).

To construct the vector pLIC-GRA24-HF, the coding sequence of GRA24 was amplified using primers LICF-117710_F and LICR-117710_R using RH genomic DNA as template. The resulting PCR product was cloned into the pLIC-HF-dhfr vector (Bougdour et al., 2013) using the LIC cloning method previously reported (Huynh and Carruthers, 2009). The C-terminal truncated versions of GRA24 were amplified and cloned in the aforementioned pLIC-HF-dhfr vector to generate pLIC-GRA24ΔR2-HF and pLIC-GRA24ΔR1/R2-HF. The pLIC-P_GRA24-GRA24-HF was obtained by amplifying the promoter and coding region of GRA24 (strain GT1, Chromosome VIII, position 664,928–671,201) with LICF–5′ region–117710_F2 and LICR-117710_R2 primers. The resulting PCR product was cloned in the aforementioned pLIC-HF-dhfr vector. To construct the mammalian expression vector pcDNA-GRA24-HF, the coding sequence of GRA24 (residues 25–542) was amplified from a poly(A)⁺ enriched RH cDNA library using pcDNA4-LIC-117710-HAFlag_F and pcDNA4-LIC-117710-HAFlag_R primers. The resulting PCR product was cloned by LIC into the pcDNA-LIC-HF plasmid (Bougdour et al., 2013). Likewise, pcDNA4-FlagHA-LIC-117710_F and pcDNA4-FlagHA-LIC-117710_R primers were used to generate the plasmid pcDNA-HF-GRA24 harboring an N-terminal HAFlag tag. The vector pDEST14 KOGRA24 was generated to construct the deletion/insertion mutation of GRA24 in Toxoplasma strains RHku80 (type I, Huynh and Carruthers, 2009) and Pruku80 (type II, Fox et al., 2011). The Multisite Gateway Pro 3-fragment Recombination system was used to clone the dhfr cassette flanked by the 5′ and 3′ surrounding regions of GRA24 coding sequence of RH genomic DNA. The 5′ flanking region of GRA24 of RH strain was amplified using primers attB1-117710_F and attB4-117710_R, and was cloned into the plasmid pDONR221 P1-P4 (Invitrogen). The 3′ flanking region of GRA24 was amplified using primers attB3-117710_F and attB2-117710_R, and was cloned into the plasmid pDONR221 P3-P2 (Invitrogen). The resulting vectors, pDONR221/5′GRA24 and pDONR221/3′GRA24, respectively, were then recombined with pDONR221/DHFR (Bougdour et al., 2013) into the destination vector pDEST14, yielding the pDEST14 KOGRA24. The vector pDEST14 KOGRA15 was generated to construct the deletion/insertion mutation of gra15 in T. gondii Pruku80 strain. The 5′ flanking region of GRA15 of type II Pru strain was amplified using primers attB1-275470_F and attB4-275470_R, and was cloned into the plasmid pDONR221 P1-P4 (Invitrogen). The 3′ flanking region of gra15 was amplified using primers attB3-275470_F and attB2-275470_R, and was cloned into the plasmid pDONR221 P3-P2 (Invitrogen). The resulting vectors, pDONR221/5′gra15 and pDONR221/3′gra15, respectively, were then recombined with the pDONR221/DHFR (Bougdour et al., 2013) into the destination vector pDEST14, yielding the pDEST14 KOGRA15.

Vectors were transfected into RHku80 and Pruku80 parasites by electroporation. Electroporation was done in a 2-mm cuvette in a BTX ECM 630 (Harvard Apparatus) at 1,100 V, 25 Ω, and 25 μF. Stable integrants were selected in media with 1 µM pyrimethamine and cloned by limiting dilution.

We used the RHku80 strain (Huynh and Carruthers, 2009) to endogenously tag GRA24 in Toxoplasma. In brief, ∼20 µg of each DNA construct was linearized within the region of homology. Linearized vectors pLIC-GRA24-HF (BsiWI), pLIC-GRA24ΔR2-HF (BlpI), and pLIC-GRA24ΔR1/R2-HF (BlpI) were phenol-chloroform extracted and ethanol precipitated. Constructions were resuspended in TlowE buffer (10 mM Tris-HCl, pH 8.0, and 0.1 mM EDTA) for transfection. Stable recombinants were selected with pyrimethamine, single-cloned by limiting dilution, and verified by IFA. The resulting strains are RHku80 GRA24-HF, RHku80 GRA24ΔR2-HF, and RHku80 GRA24ΔR1/R2-HF, respectively (Table S1).

To construct the deletion/insertion mutation of GRA24 in type I (RHku80) and type II (Pruku80) strains of Toxoplasma, the plasmid pDEST14 KOGRA24 was amplified by PCR using primers attB1-117710_F and attB2-117710_R. After sodium acetate/ethanol precipitation, DNA was resuspended in TlowE buffer (10 mM Tris-HCl, pH 8.0, and 0.1 mM EDTA) and ∼20 µg of PCR product was used for transfection. Recombinants were single-cell cloned by limiting dilution and verified by PCR analysis as follows. PCR with either a forward primer (5′ region–117710_F2) upstream of the 5′ flanking region used to generate the knockout construction and a reverse primer within the dhfr cassette (dhfr-KO-ol1_R) or a forward primer within the dhfr cassette (dhfr-KO-ol1_F) and a reverse primer (3′ region–117710_R1) downstream of the 3′ flanking region of GRA24 confirmed a disruption in the GRA24 locus. Further PCRs were performed to confirm the inability to amplify GRA24 coding sequence using primers (pcDNA4-FlagHA-LIC-117710_F and pcDNA4-FlagHA-LIC-117710_R) in both type I and type II strains. The deletion/insertion mutation of GRA15 in Pruku80 was constructed as described above with the plasmid pDEST14 KOGRA15 amplified by PCR using primers attB1-275470_F and attB2-275470_R. Disruption in the GRA15 locus was confirmed by PCR analysis with either a forward primer (5′ region–275470_F) upstream of the 5′ flanking region used to generate the knockout construction and a reverse primer within the dhfr cassette (dhfr-KO-ol1_R) or a forward primer within the dhfr cassette (dhfr-KO-ol1_F) and a reverse primer (3′ region–275470_R) downstream of the 3′ flanking region of gra15. To generate a complemented RHku80Δgra24 strain, vector pLIC-P_GRA24-GRA24-HF was co-electroporated with the pMiniHX at 1:10 ratio, followed by selection with mycophenolic acid and xanthine, and cloned by limiting dilution. The resulting strain is named RHku80 Δgra24, GRA24⁺ (Table S1).

24 h before transfection, the cells were plated (80% confluency) in 6-well tissue culture dishes. 1 µg Flag–fusion protein–expressing plasmids and 0.1 µg puromycin selection plasmid were cotransfected into 293-TRex cells with Lipofectamine reagent (Invitrogen) according to the manufacturer’s instructions. 72 h later, cells were diluted in the presence of 5 µg/ml puromycin (Sigma-Aldrich) for selection. Individual drug-resistant clones were expanded and tested for tetracycline-inducible gene expression.

Nuclear extracts from 293-TRex cells stably expressing Flag-tagged protein were incubated with anti-FLAG M2 affinity gel (Sigma-Aldrich) for 1 h at 4°C. Beads were washed with 10-column volumes of BC500 buffer (20 mM Tris-HCl, pH 8.0, 500 mM KCl, 10% glycerol, 1 mM EDTA, 1 mM DTT, 0.1% NP-40, 0.5 mM PMSF, and 1 µg ml⁻¹ each of aprotinin, leupeptin, and pepstatin). Bound peptides were eluted stepwise with 250 µg ml⁻¹ FLAG peptide (Sigma-Aldrich) diluted in BC100 buffer. For size-exclusion chromatography, protein eluates were loaded onto a Superose 6 HR 10/30 column equilibrated with BC500. Flow rate was fixed at 0.35 ml/min, and 0.5-ml fractions were collected.

Protein bands were excised from colloidal blue–stained gels (Invitrogen), treated with DTT and iodoacetamide to alkylate the cysteines, and then immediately subjected to in-gel tryptic digestion. Peptides were extracted with 5% vol/vol formic acid solution and acetonitrile, and injected into an Ultimate 3000 (Dionex) nanoLC system that was directly coupled to a LTQ-Orbitrap mass spectrometer (Thermo Fisher Scientific). MS and MS/MS data were acquired using Xcalibur (Thermo Fischer Scientific) and processed automatically using Mascot Daemon software (Matrix Science). Tandem mass spectra were searched against a compiled T. gondii database using the MASCOT program (Matrix Sciences).

For cytosolic and nuclear fractionation analysis, cells were washed in ice-cold PBS and harvested using a cell scraper. After centrifugation, cells were resuspended in buffer D (10 mM Hepes, pH 7.9, 10 mM KCl, 1.5 mM MgCl₂, 0.34 M Sucrose, 10% Glycerol, 1 mM DTT, and 1× Roche protease inhibitor cocktail) and lysed for 8 min with Triton X-100 at a final concentration of 0.1%. Cell lysates were centrifuged at 1,300 g for 5 min and the supernatant containing the cytosolic fraction was collected. The pellet, containing the nuclear fraction, was washed once in buffer D and resuspended in cell extraction buffer (Invitrogen) for protein extraction. The supernatant and pellet fractions were clarified by centrifugation at 20,000 g for 10 min at 4°C. Supernatants were mixed with protein sample buffer (Invitrogen) for Western blot analysis.

Proteins were separated by SDS-PAGE, transferred to a polyvinylidene fluoride membrane (Immobilon-P; Millipore) by liquid transfer and Western blots were probed using appropriate primary antibodies followed by phosphatase-conjugated goat secondary antibodies (Promega). Signals were detected using NBT-BCIP (Amresco).

C57BL/6 BMDMs were seeded (4 × 10⁵ per well) in 96-well plates and left to adhere overnight at 37°C in 5% CO₂. Cells were infected with freshly lysed T. gondii tachyzoites at MOI = 3, and supernatants (200 µl) were collected 24 h after infection and stored at −20°C if necessary. IL-12p40 levels were determined using a commercially available ELISA kit (Mouse IL-12 p40 ELISA kit; Thermo Fisher Scientific) according to the manufacturer’s instructions.

For qualitative determination of the kinds of chemokines secreted by infected BMDM, chemokine Ab arrays were performed with culture supernatant. Macrophages were left uninfected or infected with Pruku80 or Pruku80 Δgra24 strain parasites at an MOI of 3. Culture supernatant was harvested after 24 h and used in mouse chemokine Ab arrays according to the manufacturer’s instructions (R&D Systems). In brief, supernatant was incubated with an array membrane on which 25 different mouse chemokine Abs were bound. The membrane was then incubated with a mixture of biotinylated anti-chemokine Abs followed by HRP-conjugated streptavidin. Detection was by chemiluminescence reagents provided by the manufacturer.

Total RNA was isolated from HFFs or BMDM infected with Pru or RH strains using TRIzol reagent (Invitrogen). cDNA was synthesized with random hexamers by using the High Capacity RNA-to-cDNA kit (Applied Biosystems). Samples were analyzed by real-time quantitative PCR for IL-12b, IFN-γ, CD70, and CD200 using TaqMan Gene Expression Master Mix (Applied Biosystems) according to the manufacturer’s instructions. β2-microglobulin was used as an internal control gene.

Based on sequence similarities, we have generated a GRA24-KIM 16-aa (GLLERRGVSELPPLYI) model using the HePTP phosphatase KIM domain (RLQERRGSNVALMLDV) from structure 2GPH (Zhou et al., 2006) as a template, with appropriate residue mutations. The conformation of the side-chains and orientation of the main-chain for N- and C-terminal regions of peptides were modeled based on the bound peptide conformations in crystal structures of ERK2-pepHePTPm (2GPH) and p38-pepMEF2 (1LEW) peptide complexes, respectively. Similar work has been done with GRA24-KIM2(519–534), i.e., with an R instead of Y, with no significant differences. A p38–GRA24-KIM binary complex was then prepared using Coot modeling software (Emsley et al., 2010). The structure of p38, isolated from the p38–pepMEF2 complex (PDB accession no. 1LEW; Chang et al., 2002) was used as template for p38–GRA24-KIM docking and energy minimization calculations. The p38–GRA24-KIM complex was then subjected to energy minimization cycles under the MacroModel software (Schrödinger, LLC) using the OPLS-2005 (Optimized Potential for Liquid Simulations) force field and water as a solvent with constant dielectric value of 1. All the atoms in protein as well as peptide models were allowed to move freely. Energy minimization of structures were completed via the PRCG (Polak-Ribiere Conjugate Gradient) energy minimization procedure at 10,000 iterations and performed to a convergence threshold of 0.02 kJ mol⁻¹-Å).

BMDMs were plated at 8–10 × 10⁶ cells per 55 cm² plate, infected (MOI = 3) for 18 h with RHku80, RHku80 Δgra24, Pruku80, and Pruku80 Δgra24. RNA was isolated using TRIzol reagent (Invitrogen), followed by phenol/chloroform/isoamyl extraction. Transcripts were obtained from three biological replicates each. RNA Labeling and Array Hybridization of BMDM/RHku80 and BMDM/Pruku80 experiments were done by ArrayStar as follows: sample labeling and array hybridization were performed according to the One-Color Microarray-Based Gene Expression Analysis protocol (Agilent Technology). In brief, total RNA from each sample was linearly amplified and labeled with Cy3-UTP. The labeled cRNAs were purified by the RNeasy Mini kit (QIAGEN). The concentration and specific activity of the labeled cRNAs (pmol Cy3/µg cRNA) were measured by NanoDrop ND-1000. 1 µg of each labeled cRNA was fragmented by adding 11 µl of 10× blocking agent and 2.2 µl of 25× fragmentation buffer, the mixture was heated at 60°C for 30 min, and finally, 55 µl of 2× GE Hybridization buffer was added to dilute the labeled cRNA. 100 µl of hybridization solution was dispensed into the gasket slide and assembled to the gene expression microarray slide. The Mouse 4× 44K Gene Expression Array was manufactured by Agilent Technologies. The slides were incubated for 17 h at 65°C in a hybridization oven (Agilent Technologies). The hybridized arrays were washed, fixed, and scanned using the DNA Microarray Scanner (part number G2505B; Agilent Technologies). Microarray data analysis of acquired array images was performed using the Feature Extraction software (version 10.7.3.1; Agilent Technologies). Quantile normalization and subsequent data processing were performed using the GeneSpring GX software package (version 11.5.1; Agilent Technologies). After quantile normalization of the raw data, genes in which at least six out of nine samples have flags in Detected or Not Detected (All Targets Value) were chosen for further data analysis. Differentially expressed genes with statistical significance were identified through Volcano Plot filtering between two groups (WT versus GRA24 mutant) of compared arrays. The thresholds are fold change ≥ 2.0, p-value ≤ 0.05. GO analysis and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analysis were performed using the DAVID Bioinformatics Resources. Microarray data has been uploaded to GEO Datasets under accession no. GSE38782.

Fig. S1 shows a schematic representation of the strategy and analysis of GRA24 knockout. Table S1 provides a list of the vectors and the oligonucleotides used in this study.

We thank Dr. Vern B. Carruthers (University of Michigan Medical School) and Dr. David J. Bzik (Dartmouth Medical School) for sharing the RHku80 and Pruku80 strains, respectively. We thank Dr. Geneviève Milon for her review and fruitful discussion on the manuscript.

This work was supported by the ANR grants ANR Blanc 2010 APImiR (ANR 2010 BLAN 1315 01), ANR Jeune Chercheur 2012 ToxoEffect (ANR-12-JSV3-0004-01), and the Laboratoire d’Excellence (LabEx) ParaFrap (ANR-11-LABX-0024), as well as by the FINOVI foundation.

The authors have no competing financial interests.