The Deubiquitinating Enzyme Cylindromatosis Dampens CD8+ T Cell Responses and Is a Critical Factor for Experimental Cerebral Malaria and Blood–Brain Barrier Damage

Cerebral malaria is a severe complication of human malaria and may lead to death of Plasmodium falciparum-infected individuals. Cerebral malaria is associated with sequestration of parasitized red blood cells within the cerebral microvasculature resulting in damage of the blood–brain barrier and brain pathology. Although CD8+ T cells have been implicated in the development of murine experimental cerebral malaria (ECM), several other studies have shown that CD8+ T cells confer protection against blood-stage infections. Since the role of host deubiquitinating enzymes (DUBs) in malaria is yet unknown, we investigated how the DUB cylindromatosis (CYLD), an important inhibitor of several cellular signaling pathways, influences the outcome of ECM. Upon infection with Plasmodium berghei ANKA (PbA) sporozoites or PbA-infected red blood cells, at least 90% of Cyld−/− mice survived the infection, whereas all congenic C57BL/6 mice displayed signatures of ECM, impaired parasite control, and disruption of the blood–brain barrier integrity. Cyld deficiency prevented brain pathology, including hemorrhagic lesions, enhanced activation of astrocytes and microglia, infiltration of CD8+ T cells, and apoptosis of endothelial cells. Furthermore, PbA-specific CD8+ T cell responses were augmented in the blood of Cyld−/− mice with increased production of interferon-γ and granzyme B and elevated activation of protein kinase C-θ and nuclear factor “kappa light-chain enhancer” of activated B cells. Importantly, accumulation of CD8+ T cells in the brain of Cyld−/− mice was significantly reduced compared to C57BL/6 mice. Bone marrow chimera experiments showed that the absence of ECM signatures in infected Cyld−/− mice could be attributed to hematopoietic and radioresistant parenchymal cells, most likely endothelial cells that did not undergo apoptosis. Together, we were able to show that host deubiqutinating enzymes play an important role in ECM and that CYLD promotes ECM supporting it as a potential therapeutic target for adjunct therapy to prevent cerebral complications of severe malaria.

inTrODUcTiOn Proteolytic removal of ubiquitin and ubiquitin-like modifications on key target proteins in signaling pathways permits rapid responses to environmental cues (1). A central regulatory pathway in innate and adaptive immunity is controlled by the nuclear factor "kappa light-chain enhancer" of activated B cells (NF-κB), which is essential for inflammatory responses and pathogen defense (2). For termination of inflammation, NF-κB needs to be inactivated through the activity of deubiquitinating enzymes (DUBs) such as A20 and CYLD (3). Cylindromatosis (turban tumor syndrome), also called CYLD, was originally identified as tumor suppressor gene, a mutation of which causes tumors of the hair follicles (cylindromas) (4,5). CYLD is a DUB and negatively regulates activation of NF-κB, mitogen-activated protein kinases, and type I interferon (IFN) production by specifically cleaving lysine 63-linked polyubiquitin chains from several signaling molecules (6), including transforming growth factor beta-activated kinase 1, tumor necrosis factor (TNF) receptor-associated factors (TRAF) 2, TRAF6, B cell lymphoma 3, NF-κB essential modulator, retinoic acid-inducible gene 1, receptor-interacting serine/threonine kinase 2 (RIPK2), and signal transducer and activator of transcription 3 (STAT3) (7)(8)(9)(10)(11)(12), thereby regulating non-degradative functions like protein trafficking, protein-protein interaction, and signal transduction. Very few studies have addressed the role of CYLD in infectious diseases. Cyld −/− mice infected with Haemophilus influenzae and Escherichia coli develop hyperinflammation due to a strong activation of the NF-κB signaling pathway (13,14). On the contrary, CYLD deficiency protects mice from pneumolysin-induced acute lung injury and lethality (15). Previously, we have shown that CYLD inhibits IL-6 production by macrophages, which in turn induces STAT3-dependent protective fibrin production by hepatocytes in listeriosis (11). In addition, we could also show that CYLD impaired antilisterial immune responses in macrophages by deubiquitinating the kinase RIPK2 (12). However, how CYLD influences the course of parasitic infections remains entirely unknown.
Cerebral malaria is one of the most severe complications caused by infection with Plasmodium falciparum with fatality rates up to 25% (16). Brain pathology includes cerebral bleeding, brain edema, seizures, coma and, ultimately, death (17,18). Experimental cerebral malaria (ECM), the rodent disease model of human cerebral malaria, is a widely used surrogate model to study the pathogenesis of cerebral malaria (19)(20)(21). A hallmark of cerebral malaria is the sequestration of Plasmodium-infected red blood cells on the brain vascular endothelium (22)(23)(24). Other characteristic features include activation of endothelial cells (25), production of proinflammatory cytokines and chemokines (26,27), and disruption of the blood-brain barrier integrity (28). CD8 + T cells are known to play an important role in the pathology of ECM (29)(30)(31). CD8 + T cells mediate the disruption of the blood-brain barrier by fostering leukocyte accumulation (32) and by inducing apoptosis of endothelial cells through granzyme B and perforin-mediated cytotoxicity (29,30). However, recent reports have suggested a protective role of CD8 + T cells against blood-stage malaria. CD8 + T cells eliminate Plasmodium-infected red blood cells and also modulate the immune response through production of cytokines (33)(34)(35).
Currently, the functional relevance of DUBs in human and ECM is completely unknown. In this study, we describe that Cyld expression in the hematopoietic and parenchymal cells lethally aggravated ECM, whereas Cyld-deficient mice were protected. Furthermore, we uncovered a dual function of CYLD in the pathogenesis of ECM: (i) CYLD impaired activation of PKC-θ and NF-κB in CD8 + T cells and (ii) CYLD promoted apoptosis of endothelial cells, thereby augmenting brain pathology and preventing survival during ECM. resUlTs cYlD is essential for the Development of Plasmodium berghei anKa (Pba)-induced ecM To study the function of CYLD in ECM, WT and Cyld −/− mice were infected intravenously (i.v.) with 20,000 PbA sporozoites. WT mice developed signature ECM symptoms, such as ataxia, loss of grip strength, and progressive paralysis, within 7 days postinfection (p.i.) ( Figure 1A). In contrast, 90% of the infected Cyld −/− mice survived without showing any symptoms of ECM. Similarly, on intraperitoneal (i.p.) infection with 10 6 PbA-infected red blood cells, all Cyld −/− mice were protected from PbA blood stage-induced ECM ( Figure 1B). The brains of PbA-infected WT mice displayed extensive blood-brain barrier damage as shown by Evans blue distribution, while the brains of infected Cyld −/− mice showed no vascular leakage (Figures 1C,D), illustrating that CYLD is essential for the development ECM. A secondary lethal outcome of PbA infection in WT mice is anemia around day 20 after infection ( Figures 1A,B). In contrast to infected WT mice that did not survive beyond day 20, severe complications from parasite-induced anemia were not observed in infected Cyld −/− mice.
On infection with blood-stage PbA, parasitemia was detectable in WT and Cyld −/− mice at day 3 p.i., and parasitemia progressively increased in both mouse strains ( Figure 1E). However, the parasite burden was significantly reduced in Cyld −/− mice at day 7 p.i. Similarly, on sporozoite infection, parasitemia increased in both strains of mice from day 3 p.i. onward ( Figure 1F). This finding suggests that CYLD had no major impact on preerythrocytic parasite replication in the liver. As was seen in infections by blood transfusion, sporozoite-induced infections resulted in a significantly reduced number of parasites in Cyld −/− mice at day 7 p.i.

Presence of cYlD leads to ecM-associated Brain Pathology
To further study the effect of CYLD in ECM-associated brain pathology, a detailed histopathological analysis was performed (Figure 2). At day 7 p.i., WT mice showed severe brain hemorrhage, which was absent in Cyld −/− mice (Figure 2A). WT, but not and WT (n = 10 each) mice were intravenously (i.v.) infected with 20,000 PbA sporozoites, and survival rates were monitored daily up to day 28 postinfection (p.i.). Data represent one of two independent experiments. (B) C57BL/6 Cyld −/− and WT (n = 10 each) mice were intraperitoneally infected with 1 × 10 6 PbA-infected red blood cells, and the survival was monitored daily up to day 28 p.i. Data represent one of two independent experiments, p < 0.05 (log-rank test). (c,D) At day 7 p.i., Evans Blue was injected i.v. in mice infected with PbA sporozoites (c) and PbA-parasitized red blood cells (D). The mice were euthanized 1 h later and perfused with saline, and the isolated brains were photographed. Representative images from three independent experiments are shown. (e,F) The percentage of parasitized erythrocytes in the peripheral blood was enumerated daily from Giemsa-stained thin blood smears after infection with PbA sporozoites (e) and PbA-infected red blood cells (F), p < 0.05 (two-tailed Student's t-test). Cyld −/− , mice developed widespread neuroinflammation characterized by astrogliosis with increased expression of glial fibrillary acidic protein (GFAP) in activated astrocytes ( Figure 2B) and ionized calcium-binding adapter molecule 1 (Iba1) + -ramified microglia ( Figure 2C). Furthermore, MHC class I staining on microglial cells, some of which were tightly associated with cerebral microvessels, was strongly upregulated in the brains of infected WT mice ( Figure 2D). This was paralleled by focal intracerebral infiltration of CD8 + T cells in WT mice ( Figure 2E). Interestingly, CD31/ PECAM-1 + staining intensity on endothelial cells was considerably stronger in WT compared to Cyld −/− mice ( Figure 2F). In addition, CD31 + WT endothelial cells were frequently positive for cleaved caspase-3, an apoptosis-inducing effector caspase (Figures 2G,H).
Taken together, our data show that the presence of CYLD promotes brain pathology with increased neuroinflammation and hemorrhage as well as endothelial cell activation and apoptosis.   cYlD augments cytokine Production in the Brain of Pba-infected Mice Since cytokines and chemokines regulate the immunopathological processes leading to ECM, we next determined cytokine and chemokine levels in the brain and peripheral blood of PbAinfected mice. At day 7 p.i., quantitative reverse transcription polymerase chain reaction (RT-PCR) revealed that steady-state mRNA levels of IFN-γ, perforin, and lymphotoxin (LT)-α were significantly reduced in the brain of Cyld −/− mice, while levels of granzyme B, TNF, IL-6, CXCL-9, and CXCL-10 levels were comparable between the two mouse strains (Figure 3). This finding indicates that local expression of proinflammatory cytokines is significantly reduced in the absence of CYLD. This contrasts with systemic serum cytokine concentrations, since IFN-γ was increased in serum of Cyld −/− mice as compared to WT mice, whereas the levels of IL-2, IL-4, IL-6, IL-10, IL-17, and TNF levels did not differ significantly between the two mouse strains at day 7 p.i. ( Figure S1 in Supplementary Material).

cYlD reduces Parasite-specific cD8 + T cell responses in Peripheral Blood
Since the parasitemia in Cyld −/− mice was reduced and CD8 + T cells typically play an important role in the control of intracellular pathogens, we determined the influence of CYLD on the CD8 + T cell numbers in the blood during ECM. Before infection, WT and Cyld −/− mice harbored equally low numbers of CD8 + T cells ( Figure 4A). On infection with PbA-infected red blood cells, both the percentage and the total number of CD8 + T cells ( Figures 4A,B), as well as PbA-specific H2-Db SQLLNAKYL pentamer + CD44 + CD8 + T cells, were significantly increased in Cyld −/− mice at day 7 p.i. (Figures 4C,E).
cYlD augments cD8 + T cell response in the Brain Since pathogenic CD8 + T cells mediate end-stage immune pathology in ECM [reviewed in Ref. (39)], the intracerebral CD8 + T cell response of WT and Cyld −/− mice were analyzed in detail by flow cytometry (Figure 6). Uninfected WT and Cyld −/− mice harbored very low numbers of CD8 + T cells in the brain ( Figure 6A). Upon infection, both the percentage and the total number of CD8 + T cells (Figures 6A,B), as well as PbA-specific H2-Db SQLLNAKYL pentamer + CD44 + CD8 + T cells (Figure 6C), were significantly increased in the brains of WT mice at day 7 p.i. Numbers of PbA GAP-50-specific CD8 + T cells producing IFN-γ, TNF, and granzyme B were reduced in the brains of Cyld −/− mice ( Figure 6D). However, only differences for IFN-γ were statistically significant upon infection with blood-stage PbA (Figures 6D,E). We could independently confirm these data by sporozoite infection ( Figure S3 in Supplementary Material). Together, our findings corroborate the notion that antigenspecific CD8 + T cells might contribute to ECM pathology.  and survived the infection (Figure 7B), but harbored increased parasite loads compared to the WT and Cyld −/− mice receiving control antibody ( Figure 7C). However, the parasite load in the CD8 + T cell-depleted Cyld −/− mice was always lower compared to CD8 + T cell-depleted WT mice, indicating that in addition to the contribution of CD8 + T cells the enhanced parasite control in the Cyld −/− mice is also mediated by other cell types, likely Cyld −/− macrophages and granulocytes.
cYlD augments cytokine Production but impairs chemokine receptor expression by cD8 + T cells Although the parasite-specific CD8 + T cell response was enhanced in the peripheral blood of Cyld −/− mice, these mice harbored reduced numbers of pathogen-specific CD8 + T cells in the brain. Since chemokine receptors are essential for migration of CD8 + T cells, we checked for the cell numbers, cytokine production and the chemokine receptor expression of parasitespecific CD8 + T cells in the spleen. Upon infection with bloodstage PbA, the total number of parasite-specific CD8 + T cells were strongly increased in the spleen of Cyld −/− mice at day 7 p.i. (Figure 8A). In addition, absolute (Figures 8B,D) and relative (Figures 8C,E) numbers of PbA antigen-specific IFN-γ-and granzyme B-producing GAP-50-specific CD8 + T cells were significantly increased in the spleen of infected Cyld −/− mice. In marked contrast, expression of the chemokine receptor CXCR3 was significantly reduced in Cyld −/− CD8 + T cells (Figures 8F,G). A small, albeit non-significant, reduction in Cyld −/− CD8 + T cells Since CXCR3 plays an important role in migration of CD8 + T cells to the brain (40), the reduced accumulation of CD8 + T in the brain of Cyld −/− mice may be attributed to reduced levels of CXCR3.
cD4 + T cell responses are augmented in the Blood but reduced in the Brain of Cyld − / − Mice Since CD4 + T cells also play an important role in the control of blood-stage malaria, we determined the influence of CYLD on the CD4 + T cell numbers in the blood and brain during ECM. Before infection, WT and Cyld −/− mice harbored equal numbers of CD4 + T cells in the blood ( Figure 9A) and brain ( Figure 9B). Upon infection with PbA-infected red blood cells, the total number of CD4 + T cells increased in the blood and brain of both mouse strains. However, the numbers of CD4 + T cells were strongly increased in the blood ( Figure 9A), but markedly reduced in the brain of Cyld −/− mice ( Figure 9B). In addition, absolute ( Figure 9C) and relative ( Figure 9E) numbers of PbA antigen-specific IFN-γ-producing CD4 + T cells were significantly increased in the blood of Cyld −/− mice, but lower in the brain of Cyld −/− mice (Figures 9D,F). These findings indicate that similar to CD8 + T cells, the CD4 + T cell response to PbA is also regulated by CYLD.

absence of ecM in infected Cyld − / − Mice is Mediated by Both hematopoietic and Parenchymal cells
To validate whether the lethal course of cerebral malaria in the WT mice was caused by hematopoietic or parenchymal cells, reciprocal bone marrow chimera of WT and Cyld −/− mice were generated (Figure 10). Eight weeks after bone marrow transplantation, mice were infected with PbA-parasitized red blood cells. Seventy-five percent of WT mice reconstituted with Cyld −/− bone marrow did not develop clinical signs of ECM, whereas 50% of Cyld −/− mice reconstituted with WT bone marrow did show signature symptoms of ECM ( Figure 10A). In accordance with non-irradiated mice, all Cyld −/− mice reconstituted with Cyld −/− bone marrow remained healthy, whereas all WT mice reconstituted with WT bone marrow developed signature ECM symptoms. Thus, the clinical consequences of infection appear to depend on both Cyld-expressing hematopoietic and parenchymal cells from donor and recipient animals, respectively.
Furthermore, WT mice reconstituted with Cyld −/− bone marrow harbored low parasite numbers in the blood. Lower parasitemia was comparable with the Cyld −/− mice reconstituted with Cyld −/− bone marrow (Figure 10B). In contrast, Cyld −/− mice reconstituted with WT bone marrow harbored significantly higher parasite burden, which in turn was comparable to WT mice reconstituted with WT bone marrow ( Figure 10B). These To further analyze whether transplantation of Cyld −/− bone marrow cells into WT mice improved parasite-specific T cell responses in WT mice, the number of GAP-50-specific CD8 + T cells was determined. In blood, the total number of CD8 + T cells (Figure 10C) as well as IFN-γ-and granzyme B-producing GAP-50-specific CD8 + T cells (Figure 10D) was increased in WT | cD4 + T cell responses increase in the blood and decrease in the brain of Cyld −/− mice. (a,B) Total numbers of CD3 + CD4 + T cells in blood (a) and brain (B) of uninfected (day 0) and Plasmodium berghei ANKA (PbA)-infected (day 7) WT and Cyld −/− mice were determined by flow cytometry (n = 6 each), *p < 0.05 (two-tailed Student's t-test) (c-F) Absolute (c) and representative relative (e) numbers of interferon (IFN)-γ-producing CD4 + T cells in blood and absolute (D) and representative relative (F) numbers of IFN-γ-producing CD4 + T cells in the brain from uninfected (day 0) and PbA-infected mice (day 7) after ex vivo restimulation with anti-CD3/CD28 (n = 6 each), *p < 0.05 (two-tailed Student's t-test). Data from one of three independent experiments are shown. mice receiving Cyld −/− bone marrow, suggesting an important contribution of CYLD in the hematopoietic cell population from the donor mice in modulating systemic CD8 + T cell responses.
In the brains of bone marrow chimeric mice, significantly reduced numbers of the total CD8 + T cells (Figure 10E), as well as IFN-γ and granzyme B-producing GAP-50-specific CD8 + T cells (Figure 10F), were detected in WT mice receiving Cyld −/− bone marrow. Strikingly, the reverse transfer, i.e., WT bone marrow into Cyld −/− mice, resulted in similar low numbers of total and antigen-specific T cells. These results demonstrate that the observed lack of ECM symptoms in Cyld −/− mice can be attributed to both hematopoietic and parenchymal cells.

DiscUssiOn
This study identifies for the first time a DUB as a crucial host factor contributing to the lethal course of ECM. Our detailed comparative analysis of P. berghei infection, and the corresponding host immune responses in normal and Cyld −/− mice showed that the presence of CYLD impaired parasite control and augmented brain pathology by promoting hemorrhage, neuroinflammation with activation of astrocytes and microglia, increased accumulation of CD8 + T cells, and apoptosis of endothelial cells. During the pre-erythrocytic stage of the infection, the parasite liver stages replicate in host hepatocytes for 2-3 days before they enter the blood stream. Our study shows that upon both sporozoite and blood-stage infection, parasitemia was detectable in WT and Cyld −/− mice on day 3 p.i., indicating that CYLD deficiency did not prevent parasite replication in the liver. In this study, we addressed the role of CYLD in primary Plasmodium infections and can exclude a critical role in pre-erythrocytic parasite development and life cycle progression to blood infection, the only parasite stage that causes malaria. Future work is warranted to study a potential influence of CYLD on the hepatic immune response and acquisition of protective immunity after multiple sporozoite immunizations.
In marked contrast, the numbers of infected erythrocytes were significantly reduced in Cyld −/− mice upon both sporozoite or blood-stage PbA-induced infection. This is in contrast to our observation in listeriosis (11,12). Listeria monocytogenes (Lm) also replicates in the hepatocytes and additionally in the macrophages. We could show previously that CYLD inhibited protective hepatocytic and macrophage responses and impaired the control of Lm (11,12).
In both sporozoite and asexual blood stage infections, the systemic CD8 + T-cell response was significantly augmented when CYLD was absent. Previous studies have consistently shown that CD8 + T cells play no role in protection against blood-stage PbA infection (41)(42)(43)(44). More recent studies have challenged this view by showing a major role for parasite-specific CD8 + T cells in acute and chronic blood-stage infection (45). In this study, we demonstrated a strikingly enhanced CD8 + T cell response following acute blood-stage PbA infection in mice that lack the central regulator CYLD.
PKC-θ is a key component in the activation of T cells. Our previous studies have shown that mice deficient in PKC-θ exhibit impaired T cell activation and protection against infectious diseases due to reduced activation of NF-κB pathway (46,47). Thuille et al. (38) demonstrated that PKC-θ interacts with CYLD leading to inhibition of CYLD function (38). Interestingly, we observed increased activation of PKC-θ and NF-κB in CYLDdeficient CD8 + T cells, suggesting that the interaction between CYLD and PKC-θ might also lead to the inhibition of PKC-θmediated NF-κB activation.
Plasmodium berghei ANKA infection was associated with an increased expansion of pathogen-specific CD8 + T cells in Cyld −/− mice, which in turn was accompanied by augmented production of IFN-γ, TNF, granzyme B, and perforin. These cytokines are known to contribute to the control of PbA blood parasites. Although our CD8 + T cell depletion experiments show the importance of CD8 + T cells in the control of PbA blood parasites, CD8 + T cell depleted Cyld −/− mice harbored lower parasite loads compared to CD8 + T cell-depleted WT mice, indicating that the enhanced parasite control in the Cyld −/− mice is also mediated by other immune cells in addition to CD8 + T cells.
We have shown before that IFN-γ-stimulated Cyld −/− macrophages have an enhanced capacity to control intracellular pathogens (12). Furthermore, it is known that IFN-γ-stimulated macrophages exhibit enhanced phagocytosis of parasitized red blood cells (48). Our observation of reduced pathogen load in the blood of Cyld −/− mice at day 5 p.i., i.e., before the onset of clinical symptoms of ECM, indicates a possible role of macrophages in the enhanced clearance of parasitized red blood cells in the Cyld −/− mice.
Although CYLD is an important inhibitor of proinflammatory signaling pathways and reduces production of cytokines and chemokines, intracerebral cytokine and chemokine production of PbA-infected Cyld −/− mice was reduced compared to WT mice. Reduced accumulation of pathogen-specific cytokine-producing CD8 + T cells could be attributed to reduced expression of the chemokine receptor CXCR3 on the Cyld −/− CD8 + T cells, since CXCR3 expression on CD8 + T cells is required for T cell migration into the brain (40) and the development of ECM. Currently, it is unclear how CYLD regulates CXCR3 expression, and to clarify this aspect, further studies are required. One interpretation of this spatial reallocation of CD8 + T cells might be that an increased parasite load is a critical factor leading to intracellular cytokine and chemokine production and intracerebral accumulation of CD8 + T cells. In good agreement with the ameliorated course of ECM and the reduced production of intracellular proinflammatory cytokines, brain pathology including blood-brain barrier breakdown, perivascular bleeding, activation of astrocytes, and microglia were reduced in the Cyld −/− mice.
CD4 + T cells play an important role in controlling bloodstage malaria (49). Our observation of an enhanced CD4 + T cell response in blood of Cyld −/− mice indicates that CYLD also impaired CD4 + T cell-mediated protection against PbA infection. We also report that the expression of active caspase 3, the effector caspase-inducing apoptosis of endothelial cells, was increased in Cyld-expressing WT mice. Since Cyld −/− mice are protected from apoptosis due to increased NF-κB activation (7), this mechanism may also play an important protective role in PbA-infected Cyld −/− mice, and future studies on pathogen-induced apoptosis of endothelial cells are warranted.
In support of a role of CYLD in parenchymal, and, perhaps, endothelial, cells, our bone marrow chimera experiments demonstrate that Cyld expression in radioresistant parenchymal cells contributed to the development of lethal ECM. However, complete protection from death was dependent on Cyld deficiency in donor and recipient mice illustrating that CYLD inhibited protective host responses both in the immune system and in parenchymal cells.
Currently, inhibitors of CYLD and other DUBs are under clinical development, since DUBs are attractive candidate molecules in different diseases, including cancer (50). Our data indicate that CYLD inhibition might also be an attractive therapeutic option in severe malaria in combination with antiparasitic drugs.

MaTerials anD MeThODs ethics statement
All animal experiments were in compliance with the German Animal Welfare Act (TierSchG) in a protocol approved by the

Parasite infection
Plasmodium berghei strain ANKA was used for the experiments. For the hepatic stage infection, mice were infected i.v. with 20,000 live sporozoites obtained from salivary gland homogenate of day 21 PbA-infected female Anopheles mosquitoes. For bloodstage infection, parasites were passaged in C57BL/6J mice, and stabilates were harvested and stored in liquid nitrogen (100 µl of 1 × 10 6 infected red blood cells in 200 µl Alsever's solution and 10% glycerol). For infection, 1 × 10 6 PbA-infected red blood cells were injected i.p., with the dose adjusted for each stabilate batch such that neurological signs manifest 7 days later in the majority of mice. The infected mice were monitored for neurological symptoms, i.e., paralysis, ataxia, convulsions, and coma occurring between day 6 and 10 p.i. (51). Parasitemia was measured daily from day 3 p.i. onward by microscopic examination of Giemsastained thin blood smears.

evans Blue staining
Mice were i.v. injected with 100 µl of a 1% solution of Evans blue in phosphate-buffered saline (PBS) at day 7 p.i. One hour later, the mice were perfused intracardially with 0.9% NaCl after isoflurane anesthesia. Brains were removed and photographed.
histopathology For immunohistochemistry and immunofluorescence on frozen sections, mice were perfused intracardially with 0.9% NaCl after isoflurane anesthesia. For histology on paraffin sections, anesthetized mice were perfused with 4% paraformaldehyde in PBS and brain was removed and fixed with 4% paraformaldehyde for 24 h. Paraffin sections (4 µm) were used for hematoxylin and eosin staining and immunostaining with antibodies against GFAP and Iba1. A total of 7 µm cryostat sections were used for immunostains with active caspase-3 and CD31. Immunostaining for these two fluorescent markers was performed in staining chambers after fixation in acetone and chloroform for 10 and 8 min, respectively. The sections were then blocked with appropriate sera (1:10 in PBS) dependent on the source of the secondary antibody and incubated with the aforementioned primary antibodies at room temperature. After washing, the secondary antibody was added for 1 h. For this double immunostaining, the protocol was performed using the first primary antibody, and afterward, the same protocol was repeated with the second primary antibody and appropriate secondary antibodies. After a final washing step, the sections were aqueously mounted and stored at 4°C. Images were acquired with an Olympus microscope BX50 and the digital camera DP25 as well as cell Sens software (all by Olympus, Hamburg, Germany).

rT-Pcr
Isolation of mRNA from the brains of uninfected and PbAinfected mice was performed with an RNAeasy kit (Qiagen, Hilden, Germany). The SuperScript reverse transcriptase kit with oligo (dT) primers (Invitrogen) was used to transcribe mRNA into cDNA. Quantitative RT-PCR for IFN-γ, perforin, granzyme B, TNF, IL-6, CXCL-9, CXCL-10, LT-α, and hypoxanthine phosphoribosyltransferase (HPRT) was performed with cDNA from C57BL/6 WT and C57BL/6 Cyld −/− mice and the respective Taqman gene expression assay (Applied Biosystems, Darmstadt, Germany). Amplification was performed with a GeneAmp 5700 sequence detection system (Applied Biosystems). Quantitation was performed with the sequence detector software SDS (version 2.1; Applied Biosystems), according to the ΔΔCT threshold cycle method with HPRT as the housekeeping gene (52). Data are expressed as the increase in the level of mRNA expression in infected mice over that in uninfected controls of the respective mouse strain. All primers and probes were obtained from Applied Biosystems.

leukocyte isolation
Before isolation of cerebral leukocytes, the animals were anesthetized with isoflurane (Baxter, Deerfield, IL, USA) and intracardially perfused with 0.9% NaCl to remove contaminating intravascular leukocytes from the brain. Thereafter, brain tissue was minced through a 100-μm pore size cell strainer, and leukocytes were separated by Percoll (GE Healthcare, Freiburg, Germany) density gradient centrifugation. Peripheral blood was obtained by cardiac puncture. The erythrocytes were lysed with ammonium chloride, washed twice with ice-cold PBS at 300 × g for 10 min, and the leukocyte numbers were counted.

Flow cytometry
Leukocytes isolated from the blood and brain were analyzed by flow cytometry on a FACS Canto II with Cell Quest software (both from BD Biosciences, Heidelberg, Germany). Cells were stained with anti-CD45 in combination with anti-CD8 and anti-CD3 for CD8 + T cells. All antibodies were obtained from BD

generation of Bone Marrow chimera
Bone marrow was collected from WT and Cyld −/− mice. Femur and tibia were aseptically removed from mice, the bone ends were cut off, and the bone marrow cells were flushed out using PBS (4-8°C; Gibco, Darmstadt, Germany). For each chimera, 10 × 10 6 cells of bone marrow cells were transferred i.v. into lethally irradiated (one dose of 12 Gy) WT or Cyld −/− recipients. Recipient mice were allowed 8 weeks for reconstitution before infection with 1 × 10 6 PbA-infected red blood cells.

cytometric Bead array
The serum levels of IL-2, IL-4, IL-6, IL-10, IL-17, IFN-γ, and TNF of WT and Cyld −/− mice were analyzed by flow cytometry using the Cytometric Bead Array from BD Biosciences using the manufacturer's protocol. Peripheral blood obtained by cardiac puncture was centrifuged (490 × g, 5 min) to harvest the serum. Cytokine concentration in the serum was determined by adding 50 µl of cytokine-specific capture bead mixture to 50 µl of supernatant. Phycoerythrin detection agent (50 µl) was added to each sample and incubated in the dark at 4°C for 2 h. Thereafter, the samples were supplemented with 1 ml washing buffer and centrifuged at 200 × g for 5 min. The supernatant was discarded, and the cell pellet was resuspended in washing buffer (300 µl). The cytokine levels were measured using a FACS Canto II (BD Biosciences) and analyzed with FCAP Array™ software (version 3, BD Biosciences).
cD8 + T cell Depletion CD8 + T cells were depleted by the injection of rat anti-mouse CD8 β (clone 53.5.8) (BioXell, West lebanon, NH, USA). Depletion was performed by i.p. injection of 500 µg of purified anti-CD8 + T cell antibody starting 3 days before infection. Control mice were treated with i.p. injection of 500 µg of non-immune rat IgG. The depletion was assessed by flow cytometry.

statistics
Statistical differences for survival curves were analyzed by log-rank test. Statistical differences between groups were analyzed using the two-tailed Student's t-test. All experiments were performed at least twice. P values of <0.05 were considered significant.

acKnOWleDgMenTs
The technical assistance of Nadja Schlüter, Carolin Rauch, and Annette Sohnekind is gratefully acknowledged.

FUnDing
This study was supported by a grant from the Deutsche Forschungsgemeinschaft (SFB 854, TP5) to MN and DS (http:// www.sfb854.de/). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

sUPPleMenTarY MaTerial
The Supplementary Material for this article can be found online at http://journal.frontiersin.org/article/10.3389/fimmu. 2017.00027/full#supplementary-material.