Inhibition of ATF6β-dependent host adaptive immune response by a Toxoplasma virulence factor ROP18

Abstract

Toxoplasma gondii (T. gondii) secretes various effector molecules, which co-opt host cells and enable parasite proliferation. Of these, the rhoptry protein, ROP18, is a parasite-derived factor that determines acute virulence. ROP18 is injected into the host cytoplasm during infection and, eventually, localizes to parasitophorous vacuole (PV) membranes. ROP18 is predicted to be a serine/threonine kinase; however, the molecular mechanism by which ROP18 mediates its pathological effects remains unclear. At the end of 2010, two groups reported that ROP18 targets and phosphorylates interferon-inducible p47 small GTPases (IRGs), demonstrating the parasite’s strategy for disarming the innate defense system. Recently, we described a mechanism by which ROP18 mediates degradation of the host endoplasmic reticulum-localizing transcription factor, ATF6β, to downregulate CD8 T cell-mediated type I adaptive immune responses. Taken together, these results suggest that T. gondii inactivates host innate and adaptive immune responses by targeting different host immunity-related molecules: IRGs and ATF6β.

Keywords: :

• ATF6β
• IRGs
• effector molecules
• virulence

This article refers to:

Toxoplasma gondii (T. gondii) is a eukaryotic protozoan pathogen, which causes life-threatening encephalitis, called toxoplasmosis, in humans and animals.¹T. gondii belongs to the phylum Apicomplexa, which includes Plasmodium spp, the causative agent of the deadly parasite-mediated disease malaria. Apicomplexan parasites are defined by the presence of apical secretory organelle complexes, such as rhoptries and micronemes.² During T. gondii infection, the contents of the apical complex are secreted into the host cytosol, where they act as effector molecules. These molecules have recently been shown to directly affect gene expression, thus co-opting the host cell machinery for the survival of the parasite.³

T. gondii is divided into three main clonal lines (type I, II and III) in addition to other exotic strains.⁴ Parasite-induced innate immune responses are differentially induced depending on genotype differences. For instance, infection by type I or III parasites suppresses the production of proinflammatory cytokines such as IL-6 and IL-12, both of which are essential for the induction of effective host adaptive type I immunity, which is required for eradication of the parasite.^5,6 However, infection by type II parasites fails to reduce the innate immune response, confirming the genotype-dependent phenotype in T. gondii. Infection by type I or III, but not type II, parasites strongly activates host transcription factors such as Stat3 and Stat6.^7,8 Stat3 has an anti-inflammatory effect on innate immune cells by suppressing expression of inflammatory genes, particularly in response to IL-10.⁹ Stat6 is required for the development of alternatively activated M2-macrophages, which are, if anything, anti-inflammatory in comparison to inflammatory M1-type macrophages.¹⁰ A forward genetic study, in which Stat3/6 activation-proficient type III and activation-deficient type II parasites were intercrossed identified a highly polymorphic rhoptry-localizing protein called ROP16 as the determinant of genotype-dependent Stat3/6 activation.¹¹ Reverse genetic analysis showed that ROP16 directly associates with Stat3/6, and that a single amino acid polymorphism within the kinase domain determines the genotype-dependence.^12-14 Thus, the rhoptry protein, ROP16, determines the genotype-dependent T. gondii phenotype.

The T. gondii genotypes were originally characterized by comparing their virulence in mice. Type I parasites are the most virulent, with an LD50 in mice of 10⁰. The LD50 of type II and III parasites is 10³ and 10⁵, respectively.^15,16 Similar to genotype-dependent Stat activation, type II (virulent) and type III (avirulent) parasites were intercrossed to identify the candidate genes responsible for genotype-dependent virulence. Forward genetic studies identified another rhoptry-localized kinase called ROP18, which is also secreted into the host cytoplasm during T. gondii infection and, finally, localizes to the membranes of parasitophorous vacuoles (PV).^11,17,18 Expression of ROP18 is severely reduced in avirulent type III parasites, compared with that in virulent strains such as type I or II parasites. Furthermore, overexpression of ROP18 in type III parasites restores virulence in mice, indicating that ROP18 may determine the genotype-dependent virulence of T. gondii. The substrates of ROP18 have been analyzed by several groups. Biochemical and parasitological analyses show that interferon-γ (IFN-γ)-inducible p47 small GTPases (IRGs) are phosphorylated by ROP18.^19,20 IRGs play important roles in innate immunity against T. gondii infection, particularly in macrophages and dendritic cells.²¹ Loss of IRGs such as LRG-47, IRG47, TGTP and IGTP results in profound defects in host defense against T. gondii infection.²¹ IRGs localize to the membranes of PV (PVM) in response to parasitic infection, where they are predicted to destroy the integrity of the membrane.²² Phosphorylation of threonine residues within IRGs such as TGTP impairs localization to PVM.^19,20 Furthermore, type III or ROP18-deficient (rop18-ko) type I parasites, but not wild-type type I or ROP18-sufficient type III parasites, allow IRG localization at the PVM, indicating that innate immunity-related IRGs are the targets of ROP18. The mechanism underlying IRG phosphorylation and the subsequent loss of defensive function remains unclear.

Our own group independently attempted to identify the substrate of ROP18. Previously, we showed that ROP16 associates with Stat3 by the N-terminus. Therefore, we first analyzed the significance of the N-terminal domain of ROP18 through genetic complementation of rop18-ko type I parasites with a ROP18 mutant lacking a portion of the N-terminal domain. The results showed that the N-terminal portion of ROP18 is required for full virulence in type I parasites.²³ Next we hypothesized that ROP18 also interacts with its host-derived substrates via its N-terminus. We performed yeast-two-hybrid screening using the N-terminal portion as the bait, resulting in the identification of ATF6β as the candidate interacting protein. ATF6β is a host endoplasmic reticulum (ER)-localizing transcription factor, which is activated by a stress response called ER stress.²⁴ This transcription factor is cleaved in response to ER stress, liberating its N-terminus (containing the transactivation domain), which then localizes in the nucleus. We detected the interaction between full length ROP18 and ATF6β only in the presence of a proteasome inhibitor, suggesting that ROP18 induces degradation of ATF6β in a proteasome-dependent fashion. Furthermore, the kinase activity of ROP18 is essential for ATF6β degradation, and kinase-inactive ROP18 associates with ATF6β. Why does this parasite kinase degrade ATF6β? In other words, does ATF6β have a potential defensive function against rop18-ko parasites? To test this possibility, we infected wild-type, or ATF6β-deficient, mice with rop18-ko parasites and compared the survival rates. We found that ATF6β-deficient mice were highly susceptible to rop18-ko parasites because of a defect in CD8 T cell-mediated IFN-γ production. Moreover, impairment of CD8 T cells may be caused by the reduced expression of ATF6β in the dendritic cells required for live parasite-induced cross-presentation.²⁵ Thus, our results suggest that ROP18 may target ATF6β in dendritic cells to compromise adaptive CD8 T cell activation.

Together with another model, we would like to propose that ROP18 functions as a virulence factor during two different phases of T. gondii infection. During the early phase of infection, ROP18 inactivates IRGs via phosphorylation, leading to the failure of IRG localization at the PVM, and allowing efficient parasite proliferation within the PV. During the later stages, ROP18 degrades ATF6β and impairs CD8 T cell-mediated adaptive immune responses in the vicinity of the PVM and the host ER, both which are directly or indirectly involved in the transfer of parasite-derived antigens from the PV to the host ER for cross presentation. Thus, T. gondii may disarm host innate and adaptive defense systems at both the early and late stages of infection by targeting different host factors. Our study suggests that ER stress or ER stress-related molecules are implicated in host defense. Recently, an ATF6β homolog in plants was shown to be involved in anti-bacterial innate immune responses.²⁶ Also, another ER stress sensor, XBP-1, participates in host defense against bacteria in C. elegans.²⁷ Myeloid-specific ablation of XBP-1 in mice results in reduced expression of proinflammatory cytokines, such as IL-6 and IL-12, and type I interferon.²⁸ The role of ATF6β in the ER stress response remains unclear. However, another family protein, ATF6α, critically regulates the expression of a proportion of ER stress-related genes. Although our result demonstrates the decreased resistance to rop18-ko parasites in ATF6β-deficient mice, the gene(s) regulated by ATF6β in this cell type, and that determine the phenotype, are unknown.

The ROP18 deletion mutant (ΔN2-ROP18) lacking a portion of the N-terminal arginine-rich amphipathic helix called RAH domain, which has been reported to be involved in the localization in PVM,²⁹ does not associate with ATF6β. When rop18-ko parasites complemented with ΔN2-ROP18, the mutant parasites showed “modest” virulence and were able to still interact with and phosphorylate an IRG in the infected macrophages. Given that ROP18 inactivates IRGs at PVM,^19,20 ΔN2-ROP18 might be still recruited to PVM and inactivate IRGs by other PVM-localizing molecules such as ROP2 family members including the closely related protein ROP5.²⁹

Both ROP16 and ROP18 belong to the ROP2 family of proteins, which comprise more than 30 members, including 17 active kinases.³⁰ Although they are active kinases, a pseudokinase, ROP5, and its homologs are critically involved in acute toxoplasmosis in type I parasites, suggesting that not only active kinases, but also other kinase-inactive members, might play a role in virulence.^31,32 Moreover, another class of T. gondii-derived molecules, called GRA proteins, secreted from a parasite-specific organelle-dense granule, may also function as effector proteins. In particular, the GRA protein, GRA15, is involved in NFκB activation and the expression of inflammatory genes in macrophages.³³ Further studies focusing on T. gondii-secreting effector molecules will help us to gain new insights into host-T. gondii interactions in the future.

Acknowledgments

We thank C. Hidaka and M. Yasuda for excellent secretarial assistance; Y. Magota for technical assistance; and members of Takeda’s lab for discussions. This work was supported by grants from the Ministry of Education, Culture, Sports, Science and Technology, the Strategic International Cooperative Program (Research Exchange Type), the Japan Science and Technology Agency (JST), and the Takeda Science Foundation; KANAE FOUNDATION FOR THE PROMOTION OF MEDICAL SCIENCE; The Cell Science Research Foundation; THE ICHIRO KANEHARA FOUNDATION; Kato Memorial Bioscience Foundation; THE UEHARA MEMORIAL FOUNDATION; and Mochida Memorial Foundation for Medical and Pharmaceutical Research.