Attenuated T Cell Responses Are Associated With the Blockade of Cerebral Malaria Development by YOP1-Deficient Plasmodium berghei ANKA

Reticulon and the REEP family of proteins stabilize the high curvature of endoplasmic reticulum tubules. The REEP5 homolog in Plasmodium, Plasmodium berghei YOP1 (PbYOP1), plays an important role in the erythrocytic cycle of the P. berghei ANKA and the pathogenesis of experimental cerebral malaria (ECM), but the mechanisms are largely unknown. Here, we show that protection from ECM in Pbyop1Δ-infected mice is associated with reduced intracerebral Th1 accumulation, decreased expression of pro-inflammatory cytokines and chemokines, and attenuated pathologies in the brainstem, though the total number of CD4+ and CD8+ T cells sequestered in the brain are not reduced. Expression of adhesive molecules on brain endothelial cells, including ICAM-1, VCAM-1, and CD36, are decreased, particularly in the brainstem, where fatal pathology is always induced during ECM. Subsequently, CD8+ T cell-mediated cell apoptosis in the brain is compromised. These findings suggest that Pbyop1Δ parasites can be a useful tool for mechanistic investigation of cerebral malaria pathogenesis.


INTRODUCTION
Malaria is caused by protozoan parasites of the genus Plasmodium and remains a leading cause of death and disease across many tropical and subtropical countries. An estimated 229 million cases of malaria and 409,000 deaths per year have been reported, mostly children under 5 years of age in sub-Saharan Africa (1). Cerebral malaria (CM) is the most severe complication of Plasmodium falciparum infection and a major cause of death in severe malaria. Mechanistic investigations of CM in humans are difficult for ethical reasons (2). Alternatively, an experimental cerebral malaria (ECM) model with Plasmodium berghei ANKA infection in C57BL/6 mice has been widely used (3).
The pathological features of ECM include increased proinflammatory cytokines, vascular pathology, disruption of the blood-brain barrier (BBB), and cerebral edema, in a similar fashion to that reported in human CM (3)(4)(5). Therefore, the ECM model is a valuable tool to elucidate the mechanisms of CM.
The endoplasmic reticulum (ER) likely plays an important role in Plasmodium infection. The organelle is involved in vital cellular processes, such as protein translation and secretion, lipid biosynthesis, and calcium homeostasis (6,7), and as such is directly linked to surface remodeling of infected red blood cells (iRBCs), which in turn regulate sequestration and host immune responses. In all eukaryotic cells, the ER forms a continuous membrane system of tubules and sheets, the shape of which is tightly associated with its physiological functions (8,9). An initial analysis of the plasmodial ER identified three homologs of proteins that stabilize membrane curvature and generate ER tubules in P. berghei ANKA, termed PbYOP1, PbYOP1L, and PbRTN1 (10). To investigate the importance of ER morphogenesis in Plasmodium, we generated YOP1-deficient P. berghei parasites (Pbyop1D) and found that the growth rate and virulence in ECM are severely attenuated during blood-stage infection (11). The decreased growth rate in Pbyop1D parasites is caused by the disordered digestive vacuole biogenesis associated with abnormal hemoglobin degradation. However, the mechanism of protection against ECM in Pbyop1D parasite-infected mice is unclear. In this study, we investigated the immune response and pathologies in the brain induced by Pbyop1D parasite infection during ECM induction. We found that T cells were efficiently trapped in the mouse brain, but Th1 cells were reduced compared to wild-type (WT) parasiteinfected mice and the secretion of pro-inflammatory cytokines and chemokines largely decreased. In addition, reduced expression of necessary adhesive molecules on the endothelial cell and decreased expression of perforin and granzyme B leads to insufficient killing of intracerebral cells by CD8 + T cells.

Ethics Statement
All animal work in this study was approved by the Institutional Animal Care and Use Committee (IACUC) of Tianjin Medical University (TMU), and was performed in accordance with ethical standards in the Laboratory Animal Guideline for Ethical Review of Animal Welfare (The National Standard of the People's Republic of China GB/T 35892-2018).

Animals and Parasites
Female C57BL/6 mice aged 6-8 weeks were purchased from SPF (Beijing) Biotechnology Co., Ltd (Beijing, China) and maintained at the Animal Care Facilities of Tianjin Medical University.
P. berghei ANKA lines (clone 15Cy1) were kindly gifted from Dr. Purnima Bhanot, Rutgers New Jersey Medical School, Newark, USA. Blood-staged P. berghei ANKA parasites were stored in liquid nitrogen and thawed for using in all experiments. Parasitemia was monitored by counting the number of iRBCs per 2000 total RBCs under light microscopy examination of Giemsastained thin smears of tail blood.

Experimental Cerebral Malaria Construction and Assessment
Cryopreserved P. berghei ANKA parasite was thawed and passaged once in vivo before being used to infect experimental animals. C57BL/6 mice were infected via intravenous injection of 1×10 4 WT parasites-infected RBCs or 1×10 4 or 1×10 6 Pbyop1D parasites-infected RBCs. The parasitemia of each mouse was recorded from 5 to 7 days post-infection (dpi). Mice were monitored daily for survival and neurological signs of ECM, such as ataxia, paralysis, and coma. According to Ana Villegas-Mendez et al. (12), signs of disease could be classified into five stages using the following clinical scale: 1 = no signs; 2 = ruffled fur and/or abnormal posture; 3 = lethargy; 4 = reduced responsiveness to stimulation and/or ataxia and/or respiratory distress/hyperventilation; and 5 = prostration and/or paralysis and/or convulsions. All animals were immediately euthanized when observed at stage 4 or 5.

Mononuclear Cell Isolation
To determine the migration of CD4 + and CD8 + T cells to the brain, brain mononuclear cells were isolated from the brain of mice 7 dpi following a previously reported procedure (13). Briefly, anesthetized mice received an intracardiac perfusion with 1×PBS to remove all blood leukocytes and other nonadhered cells. Brains were dissected and chopped into small pieces and incubated in 1×HBSS with collagenase/dispase (1 mg/ ml; Roche, Germany) for 30 min at 37°C. The suspension was filtered through a 70-mm cell strainer (Falcon, USA) and the volume of the cell suspension brought up to 7 ml with HBSS. We added 3 ml of 100% Percoll (GE Healthcare, Sweden) gradient to the cell suspension to achieve a final 30% gradient. The total 10 ml cell suspension was slowly overlaid on 2 ml of a 70% Percoll gradient and centrifuged at 500g for 30 min at 18°C with no brake. The 70%-30% interphase was gently removed to a clean tube containing 8 ml 1×HBSS, mixed a few times by interversion, and centrifuged at 500g for 7 min at 18°C. The pellet was collected and treated with ACK lysing buffer to remove RBCs, washed, and resuspended in flow cytometry buffer (1×PBS containing 1% FBS).
To determine the CD4 + and CD8 + T cell dynamic in peripheral blood, peripheral blood was collected by cardiac puncture and mixed with sodium heparin for anticoagulation. Peripheral blood mononuclear cells were isolated using the Mouse Peripheral Blood Mononuclear Cell Isolation Kit (Solarbio, China) according to the manufacturer's protocol. The absolute number of mononuclear cells from the brain and peripheral blood were determined using a hemocytometer, and live cells were distinguished from dead cells using trypan blue staining.

Real-Time PCR
Total RNA was extracted with TRIzol reagent (Invitrogen, CA, USA) and cDNA synthesis was performed using a reverse transcription kit (CWBIO, China) according to the manufacturer's instructions. The mRNA level of each gene was measured using the SYBR Green PCR Master Mix (CWBIO, China) and performed on a LightCycler ® 96 System (Roche, Basel, Switzerland). The relative mRNA expression levels were evaluated using the 2 −DDCt method and normalized to the housekeeping gene mouse b-actin. The level of parasites sequestered in the brain was determined as the expression of P. berghei ANKA 18S rRNA. The primer sequences are shown in Supplementary Table 1.

ELISA
Blood was collected by retro-orbital bleeding 7 dpi. Blood samples were allowed to clot for 30 min at room temperature, and spun at 2700g for 10 min at 4°C. The serum was collected and analyzed for IFN-g and TNF-a levels via ELISA (Lianke, China). The minimum limits of detection for IFN-g and TNF-a are 3.9 pg/ml and 1.63 pg/ml, respectively.

Immunohistochemistry
Mice were anesthetized with chloral hydrate and transcardially perfused with 20 ml of ice-cold 1×PBS. Brains were dissected and fixed in 4% paraformaldehyde for 24 h and processed for paraffin embedding. Sagittal brain sections (4 mm) were prepared for immunohistochemistry (IHC). After dewaxing and rehydration, brain sections were subjected to heat-mediated antigen retrieval in pre-heated sodium citrate buffer (pH 6.0) at 95°C for 30 minutes, and then allowed to cool at room temperature for at least 2 h. Slides were washed with 1× PBS, blocked with 5% goat serum in 1×PBST (with 0. Finally, the sections were counterstained with hematoxylin, washed, dehydrated, and sealed by coverslip with neutral balsam. Each field was chosen at random within one of four regions: olfactory bulb, cerebrum, brain stem, or cerebellum. The sections were visualized and the images acquired by a Nikon ECLIPSE 90i microscope using NIS-Elements BR (version 3.1) software. Positive vessels were counted in 6 (olfactory bulb), 20 (cerebrum), 15 (brain stem), or 5 fields (cerebellum) for each brain section per mouse at ×10 objective. Images are shown at ×40 objective.

Western Blotting
Brain tissues were homogenized and lysed in RIPA buffer supplemented with protease inhibitor cocktail (Roche, Germany). After extraction, the protein concentration was determined using the BCA Protein Assay Kit (ThermoFisher, USA). Approximately 50 mg of protein was used for SDS-PAGE and transferred to a PVDF membrane (Roche, Basel, Switzerland). Proteins were probed with anti-Caspase3 (1:1000; 9662S, Cell Signaling Technology, USA) or anti-b-tubulin antibody (1:1000; 2146S, Cell Signaling Technology, USA). Antibody binding was revealed using an HRP-conjugated goat anti-rabbit IgG (H+L) (1:3000; Sungene Biotech, China). Antibody complexes were detected using Immobilon Western HRP Substrate (Millipore, Germany) and exposed on a Tanon-5200 machine.

Apoptosis Detection In Situ
The sagittal brain sections were prepared as described above for IHC. Apoptotic cells were detected in situ by TUNEL staining according to the manufacturer's instructions (In Situ Cell Death Detection Kit, POD, Roche, Germany). TUNEL-positive cells were analyzed and counted using ImageJ software. Apoptotic cells were counted in 2 (olfactory bulb), 10 (cerebrum), 3 (brain stem), or 2 fields (cerebellum) for each brain section per mouse at ×4 objective. Images are shown at ×10 objective.

Statistical Analysis
Data are presented as means ± SD. The survival rates of the mice in different groups were analyzed using the Kaplan-Meier method and compared using the log-rank (Mantel-Cox) test. For comparisons between two groups, significance was analyzed by a t-test or Mann-Whitney U test depending on the normality of the data. For comparisons among three or more groups, significance was determined using a one-way ANOVA or Kruskal-Wallis ANOVA test depending on the normality of the data. All data were analyzed by GraphPad Prism software (version 6.01). P < 0.05 was considered significant.

PbYOP1 Deficiency Attenuates the Virulence of Parasites in ECM
To compare CM development by WT or Pbyop1D parasites and to eliminate the parasitemia-associated difference, C57BL/6 mice were intravenously inoculated with 1×10 4 or a high infectious dose of 1×10 6 Pbyop1D parasites, or with 1×10 4 of WT parasites.
As shown previously, all WT parasites-infected mice developed neurological symptoms classified to stage 4 or 5 (stage 4: reduced responsiveness to stimulation and/or ataxia and/or respiratory distress/hyperventilation; stage 5: prostration and/or paralysis and/or convulsions), thus counted as ECM-positive, and died within 6-8 dpi (11). In contrast, more than 90% of 1×10 6 Pbyop1D parasites-infected mice did not display any stage 4/5 signs, thus considered ECM-negative, during 6-14 dpi ( Figure 1A). Although the parasitemia was much lower in 1×10 4 Pbyop1D parasites-infected mice, the ECM incidence was similar to that of 1×10 6 Pbyop1D parasites-infected mice ( Figure 1B). Therefore, in this study, the samples were acquired from Pbyop1D parasites-infected mice that did not suffer from CM and WT parasites-infected mice that developed CM.
Sequestration of iRBCs in microvasculature is responsible for disease severity in malaria (14). Parasite sequestration in the brain was determined based on quantification of PbANKA parasite-specific 18S rRNA by real-time PCR (12). At 7 dpi, the levels of parasite 18S rRNA in 1×10 6 Pbyop1D-infected mice were at the same level as in WT-infected mice, both of which were significantly higher than in 1×10 4 Pbyop1D-infected mice, consistent with their corresponding parasitemia ( Figure 1C). These results indicate that PbYOP1 plays a role in CM development but is dispensable for iRBC sequestration.

PbYOP1 Deficiency Does Not Affect T Cell Migration and Sequestration in Brain
As ECM is an immunopathological disease, the attenuated virulence of Pbyop1D parasites may be related to changes in immune response during Plasmodium infection. Numerous studies have demonstrated that CD8 + and CD4 + T cells respond to blood-stage Plasmodium parasite infection and are the principal effector cells involved in the pathogenesis of ECM (15)(16)(17). To investigate whether PbYOP1 deficiency affects the T cell responses during ECM development, peripheral blood and brain mononuclear cells were isolated and quantified for the CD8 + and CD4 + T cell population by flow cytometry 7 dpi.
Circulating T cells are associated with the host systemic immune response to control the parasite burden and eradicate Plasmodium infection (18). As expected, CD8 + T cells in peripheral blood were significantly increased after infection, while no difference was detected among 1×10 4 WT, 1×10 4 Pbyop1D or 1×10 6 Pbyop1D-infected mice (Figures 2A, B). Similarly, both the frequency and cell number of CD4 + T cell in peripheral blood did not change among the three infection schemes (Figures 2A, C). These results suggest that Pbyop1D parasite primes similar number of activated T cells as the WT parasite does.
In the central nervous system, CD8 + and CD4 + T cells are sequestered in the microvasculature by adhering to the endothelial cells. Consistent with sequestration of iRBCs, CD8 + and CD4 + T cells were equivalently sequestered in the brain according to the levels of parasitemia in the three infection schemes. The proportion and number of CD8 + and CD4 + T cells were similar between WT and 1×10 6 Pbyop1D-infected mice, and were both significantly higher than in 1×10 4 Pbyop1D-infected mice ( Figure 3). These results suggest that T cell sequestration in the brain was not affected by deletion of PbYOP1.
CXCR3 is an important chemokine receptor associated with the migration of T cells into the brain and subsequent development of ECM (19,20). To determine whether migration of T cell to brain was insufficient in Pbyop1D-infected mice, CXCR3 expression on CD8 + and CD4 + T cells in the peripheral blood and brain 7 dpi was measured by flow cytometry, and analyzed by the mean fluorescence intensity (MFI) and the frequency of CXCR3positive T cells (Supplementary Figures 1A-F). In both the blood and brain, CXCR3 expression on CD8 + and CD4 + T cells was up-regulated upon infection, but no detectable difference was found between WT and Pbyop1D parasite-infected mice. Next, we examined the expression of CXCL9 and CXCL10 in brain, the CXCR3 ligands that facilitated peripheral CXCR3-positive T cells migrating up the chemokine gradient to the brain (20). Levels of CXCL9 and CXCL10 in brains were significantly reduced in 1×10 4 Pbyop1D-infected mice, but no difference was observed between WT and 1×10 6 Pbyop1D-infected mice (Supplementary Figure  1G). These results suggest that deletion of PbYOP1 does not affect T cell migration. inflammatory response and contribute to the development of ECM (21,22). Because the total CD4 + T cell number in blood and brain was not different between WT and 1×10 6 Pbyop1Dinfected mice, we examined the levels of Th1 cells, defined as CD4 + T-bet + , in the induction of ECM. At 7 dpi, the percentage of Th1 cell in CD4 + T lymphocytes and the number of Th1 cells in peripheral blood were increased compared to the uninfected group, while there were no remarkable differences among the three infected groups (Figures 4A, B). The frequency of Th1 cells in the CD4 + T cell population was also increased in brains upon infection. However, the Th1 cell proportion in CD4 + T cell population of brain was significantly lower in Pbyop1D-infected group than in WT-infected group, and no changes were detected between 1×10 4 and 1×10 6 Pbyop1D-infected mice ( Figures  4D, E). The absolute number of Th1 cells in brains was also lower in Pbyop1D-infected mice, particularly in the 1×10 4 infected group ( Figure 4E). Because T-bet is important for not only the differentiation of Th1 cells during induction of ECM, but also the generation of pathogenic CD8 + T cells (22), we calculated the frequencies and numbers of CD4 -T-bet + T cells, most of which were likely CD8 + T-bet + T cells, in peripheral blood and in brain ( Figures 4C, F ), and observed no significant difference between WT and 1×10 6 Pbyop1D-infected groups. We also tested regulatory T cell (Treg), another CD4 + T cell subset that may play a regulatory role in preventing the induction of ECM and in controlling fatal pathogenesis (23,24). No significant difference was detected between WT-and Pbyop1Dinfected mice (Supplementary Figure 2). Taken together, these data suggest that PbYOP1 deficiency in parasites downregulates Th1 cell sequestration in the brain while has no effect on Tregs.
PbYOP1 Deficiency Causes Reduced Secretion of IFN-g and TNF-a Th1 cell is responsible for the secretion of pro-inflammatory cytokines (25,26), which are important in activating other immune cells to respond to infection and in the pathogenesis of ECM (27). Since Th1 cells were reduced in the brains of Pbyop1D-infected mice during ECM construction, we examined whether deletion of Pbyop1 would influence the production of pro-inflammatory cytokines IFN-g and TNF-a, two crucial Th1type cytokines. IFN-g is required to activate brain endothelial cells and participate in brain endothelial cells cross-presentation of parasite antigen (28). Although TNF-a plays a dispensable role in ECM development, it exacerbates cerebral pathology (29,30). At 7 dpi, the mRNA expression of IFN-g and TNF-a were significantly decreased in the brains of Pbyop1D-infected mice compared to WT-infected mice ( Figure 5A). The levels of IFN-g and TNF-a in serum displayed similar results ( Figure 5B). These results suggest that reduction of brain-trapped Th1 cells and subsequent reduction of IFN-g and TNF-a may attribute to the blockage of CM development in Pbyop1D-infected mice.

PbYOP1 Deficiency Causes Reduced Cell Adhesion in the Brainstem
Under inflammatory conditions during Plasmodium infection, proinflammatory cytokines such as IFN-g induce brain endothelial activation and local inflammation (31). Activation of brain endothelial cells is associated with leukocyte adhesion, parasite sequestration, and function of antigen cross-presentation (28,32). A hallmark of endothelial activation is the upregulated expression of adhesion molecules, such as ICAM-1, VCAM-1, and CD36, on the endothelium of cerebral microvessels (33,34). At 7 dpi, the mRNA expressions of ICAM-1 and VCAM-1 were upregulated in the brain, but no prominent differences were detected among the three infected schemes. CD36 transcripts were more abundant in brains of 1×10 6 Pbyop1D-infected mice (Supplementary Figure 3). The elevation of CD36 is likely mostly contributed by the induced expression in innate immune cells in brain, because CD36 is also a scavenger receptor employed by phagocytes like monocytes, macrophages and microglia. Higher initial infection dose of parasites would trigger a more intensive innate immune response for phagocytic clearance of iRBCs (35). It is also reasonable to speculate that sustained CD36 expression in brains of 1×10 6 Pbyop1D-infected mice reflects a continuous demand on innate immunity due to attenuation in adaptive immune response. Fatal ECM is frequently linked to severe brainstem pathology (36). To test whether the extent of endothelial activation was distinct in specific brain regions of infected mice, we examined the protein levels of ICAM-1, VCAM-1, and CD36 by IHC staining and counting stain-positive vessels in multiple brain regions, including olfactory bulb, cerebrum, brainstem and cerebellum. At 7 dpi, the numbers of ICAM-1 + , VCAM-1 + and CD36 + vessels were significantly lower in the brainstems of Pbyop1D-infected mice compared to WT-infected mice ( Figure 6). Variation in ICAM-1 + , VCAM-1 + , and CD36 + vessels in other brain regions were also compared among the three infected schemes, but the differences were not as significant as detected in the brainstem. These results indicate insufficient endothelial activation in the brainstem in Pbyop1D-infected mice, which may influence the subsequent immune pathologies mediated by CD8 + T cells.

Cell Apoptosis Is Attenuated in the Brain of Pbyop1D-Infected Mice
During ECM pathogenesis, the activated endothelial cells adhere CD8 + T cells and present parasite-specific antigens to T cells. The effector CD8 + T cells secrete granzyme B and perforin to induce damage of intercellular tight junctions of the endothelium, trigger apoptosis of endothelial cells and neuronal cells, which disrupting the BBB and finally impairing central nervous system function (36)(37)(38). Thus, we measured the expression of granzyme B and perforin in the brain 7 dpi. The granzyme B and perforin mRNA levels were significantly decreased in the brains of Pbyop1D-infected mice compared to WT-infected mice (Supplementary Figure 4).
Caspase-3 is the main executioner of apoptosis and activated during ECM (39). Granzyme B can directly cleave pro-caspase-3, resulting in an active caspase-3. Activated caspase-3 induces DNA fragmentation and cell death, leading to the lethal pathogenesis of ECM (40,41). To further examine the brain cell damage, we tested the activation of caspase-3 and cell apoptosis in ECM. At 7 dpi, active caspase-3 was significantly decreased in the brains of Pbyop1D-infected mice compared to WT-infected mice. Consistent with ECM incidence, no significant difference was detected between 1×10 4 and 1×10 6 Pbyop1D-infected mice ( Figures 7A, B). In addition, the pro-caspase-3 protein levels did not change after infection ( Figures 7A, C).
Cell apoptosis was also detected in situ by TUNEL staining. Apoptotic cells were counted in the four brain regions. Pbyop1D parasite infection induced cell death was significantly reduced in the brainstem versus WT parasite infection (Figures 7D, E). Cell death was also decreased in the olfactory bulb, cerebrum, and cerebellum, where it was mild in Pbyop1D-infected mice. Cell apoptosis was consistent with endothelial activation in different brain regions. These results suggest that CD8 + T cell-mediated intracerebral cell apoptosis is attenuated in Pbyop1D parasite infection.

Inflammation Is Further Alleviated in the Brains of Pbyop1D Parasite-Infected Mice
Because the mice infected with Pbyop1D parasites died of severe anemia without distinct manifestations of ECM more than 3 weeks post-infection, we tested the expression of cytokines and cytolytic molecules after the time frame for the onset of ECM, 11 dpi (42). The expression of IFN-g, TNF-a, granzyme B, and perforin ( Figure 8A), and the activation of apoptosis molecule caspase-3 ( Figure 8B) were down-regulated in the brains of Pbyop1D parasite-infected mice 11 dpi compared to 7 dpi, particularly in 1×10 6 Pbyop1D-infected mice. However, the cell death in the brain neither expanded nor recovered for the irreversibility of apoptosis 11 dpi (data not shown). These data imply that Pbyop1D parasites not only give rise to mild proinflammatory responses and cytotoxic effects of brain infiltrating T cells that were not sufficient to lead to lethal pathogenesis of brain during ECM induction, but also that these immunopathological changes decreased gradually. integral membrane proteins that generate ER tubules by inducing high curvature in the membrane (8). PbYOP1-deleted P. berghei parasites were generated to explore the function of this important ER tubule-shaping protein in the Plasmodium parasite and malaria. We found that PbYOP1 had a profound effect on the parasite growth rate and the pathogenesis of ECM in blood-stage infection. As described recently, the slow growth rate of parasites in erythrocytic stage is associated with a dysfunction of hemoglobin degradation in the digestive vacuole and disordered parasite metabolism (11). However, the mechanism of susceptible mice infected with Pbyop1D parasites surviving from ECM induction is unclear. It is reported that the T cell response to malaria may contribute to ECM (43,44). In this study, we analyzed the effect of PbYOP1 on parasite virulence by detecting the T cell response associated with the pathogenesis of ECM. ECM is a complex neurological syndrome. In previous studies, iRBC sequestration in the brain microvasculature was associated with the development of ECM (45,46). Although the growth rate of Pbyop1D parasites is significantly decreased in the asexual phase, parasite sequestration in the brains of 10 6 Pbyop1D parasites-infected mice is comparable to that of 10 4 WT parasites-infected mice and fails to induce ECM. Sequestration of iRBCs only is inadequate to induce the brain injury leading to the fatal syndrome during infection.

DISCUSSION
Both CD4 + and CD8 + T cells have been shown to contribute to ECM development (47,48). The mechanisms by which CD4 + T cells

-bet + cells in T cells and cell number of CD4 -T-bet + T cells in
brain were quantified. Data are displayed as mean ± SD (n = 6/group) and are representative of three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant as determined by one-way ANOVA followed by Tukey's multiple comparison test.
mediate cerebral complications have not been fully elucidated, but it is thought to involve the production of Th1-type cytokines, such as IFN-g, that exacerbate the inflammatory cascade responsible for local and systemic inflammation in cerebral malaria (23). CD8 + T cell depletion or ablation of effective functions completely abrogates the development of ECM (27,49). Brain infiltrating CD8 + T cells induce opening of endothelium tight junction, endothelial cell apoptosis and other intracerebral cell apoptosis in a granzyme B and perforindependent manner (50,51). In addition, perforin secreted by CD8 + T cells is sufficient to cause cell death, disrupted BBB, and fatal brain edema in the specific regions of the brain, including brainstem and olfactory bulb, during ECM (52), indicating that the killing effect mediated by CD8 + T cells plays a vital and precise role in ECM pathology. Upon examination of sequestered T cells during infection, the PbYOP1-deficient parasites had attenuated virulence without influencing sequestration of the total CD4 + and CD8 + T cells in the brain, and the frequency and cell number were comparable between WT and 10 6 Pbyop1D-infected mice. Additionally, expression of CXCR3, which is associated with T cell migration, did not change in the three infection schemes including 10 4 Pbyop1D-infected mice.
Some studies have demonstrated that the pro-inflammatory Th1 response is involved in the pathogenesis of ECM (22,42). Consistent with the morbidity of ECM, the sequestration of Th1 cells in the brain was significantly decreased in Pbyop1D parasite-infected mice. Stimulation of T cell receptor and other extrinsic factors, particularly cytokines, which are associated with STAT activation, are crucial for the appropriate differentiation of CD4 + T cell subsets (53,54). Although the number of Th1 cells in peripheral blood did not change in Pbyop1D-infected mice compared to WT-infected mice, Th1 cells sequestered in brain decreased remarkably. This may due to the attenuated virulence of Pbyop1D parasites, which may influence parasites and leukocytes sequestrations in brain microvasculature, the process of parasite-derived antigen crosspresentation in endothelial cells, and the pro-inflammatory response inducing differentiation of CD4 + T cells (55).
>The cytokines associated with the pathogenesis of ECM were also detected in this study. The expression of pro-inflammatory cytokines IFN-g and TNF-a (56,57) and cytotoxic molecules granzyme B and perforin (51,58), and the activation of the main executioner of apoptosis caspase-3 (59) were significantly reduced A B FIGURE 5 | IFN-g and TNF-a expression are decreased in Pbyop1D-infected mice. (A) IFN-g and TNF-a mRNA expressions relative to b-actin in brain samples of uninfected and infected mice were evaluated by real-time PCR 7 dpi. (B) Serum IFN-g and TNF-a levels were quantified by ELISA 7 dpi. Data are presented as mean ± SD (n = 5/group) and are representative of three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; ns, not significant as determined by one-way ANOVA followed by Tukey's multiple comparison test. . ICAM-1, VCAM-1, or CD36-positive vessels (black arrows) were quantified for each sagittal brain section in 6 fields (olfactory bulb), 20 fields (cerebrum), 15 fields (brainstem), and 5 fields (cerebellum); one brain section per mouse. Data are presented as mean ± SD. Differences among the three groups were analyzed using Kruskal-Wallis ANOVA followed by Dunn's multiple comparisons test: *P < 0.05, **P < 0.01; ns, not significant.
in Pbyop1D parasites-infected mice; all of these reductions directly protect the Pbyop1D parasite-infected mice from ECM. Moreover, ECM occurs 6-10 dpi (42), and these pro-inflammatory factors were further decreased 11 dpi in Pbyop1D-infected mice; thus, the inflammation induced by the Pbyop1D parasite is not only attenuated, but also down-regulated gradually.
The brainstem regulates many vital functions, such as the cardiovascular and respiratory systems, and it is likely that mice succumb to ECM due to the widespread inflammation and neuron death observed in this brain region (36,60). The pathogenesis in the brainstem was significantly alleviated in Pbyop1D parasites infection. We detected that the adhesion were quantified for each sagittal brain section in 2 fields (olfactory bulb), 10 fields (cerebrum), 3 fields (brainstem), and 2 fields (cerebellum); one brain section per mouse. Data are presented as the mean ± SD. Differences among the three groups were analyzed using Kruskal-Wallis ANOVA followed by Dunn's multiple comparisons test: *P < 0.05, **P < 0.01; ns, not significant. molecules expressed on microvessels were decreased in the brainstem, as measured by IHC staining of ICAM-1, VCAM-I, and CD36. Pbyop1D parasite infection induced cell apoptosis detected by in situ TUNEL staining was also reduced in the four brain regions (olfactory bulb, cerebrum, brainstem, and cerebellum), particularly in the brainstem, consistent with the IHC results. While the connection of ICAM-1 and VCAM-1 to CM development is straightforward, the case for CD36 is complicated. CD36 plays a dual role in malaria: its expression in phagocytes induced at early stage of infection has an important effect on parasites clearance (61); whereas that in endothelial cells mediates parasite sequestration in microvasculature of organs (62). Notably, murine CD36mediated sequestration is not essential for CM pathology (62). However, it has been reported that ICAM-1 and CD36 synergize to mediate cytoadherence of Plasmodium falciparum-infected RBCs to human endothelial cells (63), suggesting that CD36 might contribute to CM pathogenesis in a collaborative manner. After invading the erythrocyte, hundreds of proteins are exported out of the parasite and beyond the parasitophorous vacuole membrane to numerous locations within the parasiteinfected erythrocyte (64). PbYOP1 deficiency may affect the ER function and tubule formation, resulting in disordered parasite metabolism and a defect in protein secretion. The pathologies in ECM are initially induced by the parasite antigen presented on the activated endothelial cell MHC I molecule and recognized by the specific T cell receptor on CD8 + T cells (16,65). Comparative analysis between the secretome of WT and PbYOP1-deleted parasites would reveal vital clues for understanding the development of CM.
The PbYOP1-deleted parasites offer a unique and important opportunity for further understanding of ECM. Specifically, previous studies showed that in non-ECM Pb NK65-infected C57BL/6 mice, iRBCs were not efficiently accumulated in brain microvessels (4). These differences between known non-ECM parasites and the PbYOP1-deleted parasites suggest that the Pbyop1D parasites could move at least one step further in ECM pathogenesis when compared to existing non-ECM Plasmodium strains, and thus become more useful materials for pinpointing the key elements during ECM development.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

ETHICS STATEMENT
The animal study was reviewed and approved by Institutional Animal Care and Use Committee (IACUC) of Tianjin Medical University.

AUTHOR CONTRIBUTIONS
QW and XS designed the study. LH and XS performed the experiments. LH analyzed the data. LH, XS and QW wrote the paper. All authors contributed to the article and approved the submitted version. (G) mRNA expressions of CXCL9 and CXCL10 in the brain were measured by realtime PCR. Data are presented as mean ± SD (n = 6/group) and are representative of three independent experiments. *P < 0.05, **P < 0.01; ns, not significant as determined by one-way ANOVA followed by Tukey's multiple comparison test.
Supplementary Figure 2 | Pbyop1D parasite has no effect on regulatory T cell accumulation in the blood and brain. (A, B) Representative flow cytometry dot plots showing regulatory T cells in the blood (A) and brain (B) of uninfected, WT parasites-infected (10 4 ), and Pbyop1D parasites-infected (10 4 or 10 6 ) mice 7 dpi gated on CD4 + T cells. (C, D) The frequency and cell number of Tregs in total CD4 + T cells in the blood (C) or brains (D). Data are presented as mean ± SD (n = 6/group) and are representative of three independent experiments. *P < 0.05, **P < 0.01; ns, not significant as determined by one-way ANOVA followed by Tukey's multiple comparison test.
Supplementary Figure 3 | mRNA expression of adhesion molecule ICAM-1, VCAM-1 and CD36 in the brain. mRNA expressions of ICAM-1, VCAM-1, and CD36 relative to b-actin in brain samples from uninfected and infected mice were evaluated by real-time PCR 7 dpi. Data are presented as mean ± SD (n = 5/group) and are representative of three independent experiments. ***P < 0.001; ns, not significant as determined by one-way ANOVA followed by Tukey's multiple comparison test.
Supplementary Figure 4 | Expression of granzyme B and perforin is decreased in the brains of Pbyop1D parasites-infected mice. Granzyme B and perforin mRNA expressions relative to b-actin in brain samples from uninfected and infected mice were evaluated by real-time PCR 7 dpi. Data are presented as mean ± SD (n = 5/ group) and are representative of three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant as determined by one-way ANOVA followed by Tukey's multiple comparison test.