All animal procedures in this study were approved by the Tianjin Medical University (TMU) Ethics Review Committee. This project was conducted according to Laboratory Animal Guidelines for Ethical Review of Animal Welfare (The National Standard of the People’s Republic of China, GB/T 35892-2018) and was approved by the Institutional Animal Care and Use Committee of TMU. In vitro cultures of Plasmodium falciparum were obtained using human erythrocytes as approved by the TMU Ethics Review Committee and the Tianjin Central Hospital of Gynecology Obstetrics Ethics Review Committee.

Male Wistar rats (80-100 g), 4 to 6-week-old female BALB/c mice, and 6 to 8-week-old female C57BL/6 mice were purchased from SPF Biotechnology Co., Ltd (Beijing, China). All animals were bred at the TMU Animal Care Facility. The blood-stage P. berghei ANKA strain and P. falciparum 3D7/Dd2/803 strains were stored as stabilates in liquid nitrogen. Parasitemia was counted by light microscope examination of Giemsa-stained thin blood smears.

BALB/c mice were infected intravenously with 1×10⁴-1×10⁶ P. berghei ANKA-iRBCs. For D-mannose (Sangon Biotech A600554, Shanghai, China) treatment, normal drinking water was exchanged for 20% D-mannose in drinking water (w/v) that was replaced once every 3 days. Mice received mannose by oral gavage (200 µL) every other day from the same stock of 20% mannose in drinking water. These mice were euthanized by cervical dislocation 13 days post-infection (p.i.). P. falciparum 3D7/Dd2/803 strains were cultured in human erythrocytes (obtained from Tianjin Central Hospital of Gynecology Obstetrics) in RPMI 1640 medium containing 25 mM HEPES supplemented with 0.5% AlbuMAX II, 100 μM hypoxanthine, and 12.5 μg/mL gentamicin (17). The initial parasitemia in each group was 1%. The cultured parasites were treated daily with 10 mM-50 mM D-mannose for 1 h, which was then replaced with fresh complete medium. The parasitemia of the infected mice and cultured P. falciparum was monitored every day.

Female C57BL/6 mice (6-8 weeks old) were infected intravenously with 1×10⁴ P. berghei-iRBCs to construct experimental cerebral malaria (ECM) models (18). Infected mice were treated with 200 μL 20% D-mannose or drinking water by oral gavage every other day starting from day 0, and the drinking water in the experimental groups was changed to 20% mannose. Whether the infected mice developed CM was assessed using the Rapid Murine Coma and Behavior Scale (RMCBS) every 24 hours from 6 days p.i. as described previously (19). Mice that developed CM died or were euthanized when found moribund. At the end point of this experiment, the remaining mice were euthanized by cervical dislocation. The integrity of the blood-brain barrier (BBB) was determined by Evans Blue staining, and Evans Blue extravasation of the brain was calculated as milligrams of Evans Blue dye per gram of brain tissue as described previously (20).

P. berghei-parasitized rat erythrocytes were purified by density gradient centrifugation using Histodenz (Sigma-Aldrich D2158, St. Louis, Missouri, US). The superoxide levels of iRBCs/RBCs were determined by detecting the mean fluorescence intensity of DHE (Beyotime, Shanghai, Beijing, China) and MitoSOX Red (Invitrogen, Carlsbad, CA, United States) by flow cytometry. Isolated iRBCs/RBCs were treated with 10 mM-30 mM D-mannose for 3 h, and then stained with 10 μM DHE for 30 min and 5 μM MitoSOX Red for 15 min. Menadione or H₂O₂-treated cells were used as positive controls. Stained cells were analyzed on a BD FACSCanto II Flow Cytometer (BD Biosciences, San Jose, CA, US) and data analyzed by FlowJo software (TreeStar, Ashland, OR, US).

Eryptosis of RBCs and iRBCs was detected by the Annexin V-FITC Apoptosis Detection Kit (Thermo Fisher Scientific, US) according to the manufacturer’s instructions. Rat RBCs and purified P. berghei-iRBCs were treated with 10-30 mM mannose for 3 h and stained with Annexin V-FITC before flow cytometry. Cells treated with 10 mM vitamin C for 1 h were used as positive controls. Cultured P. falciparum was treated with 10-30 mM mannose for 24 h and analyzed by the Annexin V apoptosis assay.

TUNEL assays were performed using the CoraLite^®488 TUNEL Fluorescence Assay Kit (Proteintech, US) in accordance with the manufacturer’s instructions. Parasite-infected erythrocytes treated or not treated with mannose and DNase I were fixed with 4% formaldehyde for 30 min and permeabilized with 0.2% Triton X-100 for 2 min. The cells were labeled with TUNEL reaction mix for 1 h at 37°C and analyzed by flow cytometry. The percentage of apoptotic cells was determined as the proportion of TUNEL-positive cells among total cells. DNase I-treated cells were used as the positive control.

To determine the macrophage population in mouse spleens, the organ was harvested on the indicated days and spleen single-cell suspensions prepared by homogenization through a 70-μM cell strainer (BD Falcon Bioswisstec, Switzerland). Erythrocyte lysis was performed to acquire splenocytes. Peripheral blood was collected by cardiac puncture and anticoagulated with heparin to measure CD4⁺ and CD8⁺ T cells. Peripheral blood mononuclear cells were isolated using the Mouse Peripheral Blood Mononuclear Cell Isolation Kit (Solarbio, China) according to the manufacturer’s protocol. To determine the infiltration of T cells into the brain, brains were dissected on the indicated days after perfusion with 1×PBS, and brain mononuclear cells were isolated following a previously reported procedure (21). The total cell concentrations of splenocytes or mononuclear cells from the brain and peripheral blood were calculated using a hemocytometer, and cell viability was evaluated using Trypan blue staining.

The following antibodies purchased from BioLegend, eBioscience, or BD were used: CD45-PerCP (30-F11), CD11b-APC (M1/70), CD11b-PerCP/Cy5.5 (M1/70), F4/80-FITC (BM8), F4/80-APC (BM8), Ly-6G-PE (1A8), MHC Class II-PE (M5/114.15.2), CD206-PE/Cyanine7 (C068C2), CD3e-FITC (145-2C11), CD4-PE (RM4-5), CD4-PerCP/Cy5.5 (RM4-5), CD8-APC (53-6.7), CXCR3-PE (CXCR3-173), CD69-PE/Cyanine7 (H1.2F3), and CD16/32 (93). All cell preparations were incubated with anti-mouse CD16/32 on ice for 15 min before staining, and then incubated with cocktails of mAbs in flow cytometry buffer. Stained cells were analyzed using a BD FACSCanto II Flow Cytometer and data analyzed using FlowJo software.

The phagocytosis assay was performed according to the protocol described by Yan et al. (22). Mouse spleen mononuclear cells were isolated and cultured in DMEM complete media with 10% FBS (D10) in a 6-well plate. Cells were incubated at 37°C for 1 hour to allow the macrophages to adhere, and non-adherent cells were subsequently removed by washing. IgG-opsonized GFP-expressing Escherichia coli (E. coli) was used for the phagocytosis of splenic macrophages. After washing the cultured cells, 2 mL fresh D10 containing E. coli (~100 bacteria per cell) was added to each well. Bacterial attachment was synchronized by spinning the 6-well plate at 500 rpm for 5 min at room temperature (RT). The cells were incubated at 37°C for 1 hour to allow bacterial uptake by macrophages. We added 200 μL 0.25% Trypsin-EDTA to the cells and incubated for 10 min at RT to remove the residual E. coli at the cell surface. The plate was washed three times with PBS and 0.5 mL 5 mM EDTA added to each well to detach the cells for 5 min, and then 2 mL D10 was used to stop digestion. The cells were harvested by spinning and the cells washed with PBS. The cells were then ready for flow cytometric analysis to evaluate the phagocytotic capacity.

Total RNA was extracted from tissues using TRIzol (Invitrogen, Carlsbad, CA, US) according to the manufacturer’s instructions, and cDNA was synthesized using a HiFiScript cDNA Synthesis kit (CoWin Biosciences, Beijing, China). The UltraSYBR Mixture kit (CoWin Biosciences, Beijing, China) was used for single-step real-time PCR reactions. For PCR, an initial cycle of 95°C for 10 min was followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min. A thermal denaturation protocol was then used to generate the dissociation curves to verify the amplification specificity (a single cycle of 95°C for 15 s, 60°C for 60 s, 95°C for 15 s, and 60°C for 10 s). The reactions were performed on a LightCycler ^® 96 System (Roche, Basel, Switzerland). Changes in mRNA levels were quantified using the 2^-ΔΔCT method with β-actin mRNA as the internal control. Primers for β-actin were 5’-GGCTGTATTCCCCTCCATCG-3’ (forward) and 5’-CCAGTTGGTAACAATGCCATGT-3’ (reverse). Primers for P. berghei 18S rRNA were 5’-AAGCATTAAATAAAGCGAATACATCCTTAC-3’ (forward) and 5’-GGAGATTGGTTTTGACGTTTATGTG-3’ (reverse). Primers for IFN-γ were 5’-CCCTCACACTCAGATCATCTTCT-3’ (forward) and 5’-TGCTACGACGTGGGCTACAG-3’ (reverse). Primers for TNF-α were 5’-ATGAACGCTACACACTGCATC-3’ (forward) and 5’-CCATCCTTTTGCCAGTTCCTC-3’ (reverse).

Infected and uninfected mice were treated with drinking water or 20% D-mannose and their body weight and food intake monitored daily. Average daily food consumption was expressed as grams of total food eaten per total weight of all animals in the cage. Blood samples were collected by retro-orbital bleeding after 20 days of treatment for normal BALB/c mice or on day 7 p.i. for infected C57BL/6 mice. Serum was harvested to detect biochemical indices using a Roche P800 Clinical Biochemistry Analyzer (F. Hoffmann-La Roche Ltd., Basel, Switzerland), including alanine aminotransferase (ALT), aspartate transaminase (AST), total bilirubin (TBIL), urea (BUN), and creatinine (CREA).

Data are presented as mean ± SD. The unpaired t test or Mann-Whitney U test was used to compare two different groups. One-way ANOVA (with Tukey’s multiple-comparison post-test) or Kruskal-Wallis ANOVA was used for comparisons between more than two groups. Two-way ANOVA with Tukey’s multiple comparisons test was performed to compare the effect of multiple levels of two factors. The cumulative survival rates of mice between two groups were plotted by the Kaplan-Meier method and compared using the log-rank test. P < 0.05 was considered significant. All statistical analyses were performed using GraphPad Prism 7.

To test whether mannose inhibits the growth of blood-stage parasites, we infected BALB/c mice with P. berghei ANKA, fed them drinking water containing 20% mannose daily, and treated them with 200 μL 20% mannose by oral gavage every other day (qod) from the day of infection. The same treatments have been used in mice for tumor inhibition (12) and a single oral gavage of 200 μL 20% mannose in mice elevates the serum concentration of mannose to ~3 mM (12). Infected mice not treated with mannose exhibited a typical parasite growth pattern over 13 days (23), whereas mice treated with mannose exhibited reduced parasitemia during both the initial peak and the latter proliferation (Figure 1A). Infected mice not treated with mannose exhibited a gradual decline in body weight and food intake, which correlated with sickness-related anorectic behavior (Figures 1B, C). In contrast, these aspects of infected and treated mice exhibited a slight decrease in the early phase of infection while remained relatively constant during the later course of infection (Figures 1B, C). These results suggest that mannose treatment delays the course of Plasmodium parasite infection even though complete elimination is not achieved.

Next, we tested whether mannose is still effective when infection has been established. When mannose treatment was started 5 days p.i., when parasitemia was evident, the inhibition of parasite growth was still visible in the later stages of infection but was less effective than treatment starting from 2 days p.i. (Supplementary Figure 1A). We also analyzed mannose efficacy in preventing Plasmodium infection by administering mannose 10 days prior to parasite infection. We observed no difference in the progression of infection compared to infected mice that were treated starting 1 day p.i. (Supplementary Figure 1B). These results suggest that mannose treatment is more efficient when administered at an early stage of infection but does not offer additional protection upon pre-treatment.

We also tested whether mannose treatment is sensitive to the genetic backgrounds of host mice. C57BL/6 mice are commonly used to generate the ECM model. We found growth inhibition of blood-stage parasite by mannose (Figure 1D). Similar patterns of food intake and body weight were seen in the treated and untreated groups compared to the BALB/c groups (Figures 1E, F). In addition, liver and kidney dysfunction observed during infection were reduced 7 days p.i. in mannose-treated mice (Figures 1G–J). These results indicate that mannose is effective in inhibiting Plasmodium growth and alleviating symptoms in C57BL/6 mice.

Finally, we tested whether mannose works on P. falciparum. Gonzalez et al. reported that 25 mM mannose impaired the growth of several kinds of cancer cells in vitro (12). Therefore, we added mannose to in vitro cultured P. falciparum for 1 h each day. Parasite growth was efficiently inhibited by the addition of high dose of mannose (≥ 20 mM) (Figure 1K) but was not effective at 10 mM (Figure 1L). We also tested the efficacy of mannose treatment against drug-resistant P. falciparum strains. Similar results were obtained when high dose of mannose (≥ 20 mM) was used in chloroquine-resistant P. falciparum Dd2 and artemisinin-resistant P. falciparum 803 strains (Supplementary Figures 1C, D). These results confirm that mannose treatment inhibits the growth of P. falciparum in vitro with high dose of mannose (≥ 20 mM) and is effective in previously known drug-resistant strains.

High-dose mannose causes oxidative stress in cells (14). To determine whether mannose treatment triggers oxidative stress in iRBCs, isolated P. berghei ANKA-infected rat RBCs and cultured P. falciparum 3D7 were incubated with 10-30 mM mannose for 3 h and subjected to superoxide measurements. Surprisingly, accumulation of reactive oxygen species (ROS), as monitored by DHE or MitoSOX, was not detected with increasing concentrations of mannose (Supplementary Figures 2A–C). The combination of mannose with chemotherapeutic drugs has been reported to trigger cancer cell death by potentiating the intrinsic pathway of apoptosis (12). However, in vitro treatment with mannose did not cause eryptosis in erythrocytes infected by P. berghei ANKA or P. falciparum 3D7 (Supplementary Figures 2D, E) or apoptosis in P. berghei parasites (Supplementary Figure 2F). In contrast, vitamin C treatment consistently triggered eryptosis in iRBCs (Supplementary Figure 2D) as reported previously (11). Taken together, these results suggest that mannose-induced Plasmodium inhibition is less likely attributed to increased oxidative stress or the induction of apoptosis.

The plasma concentrations of mannose in in vivo experiments are much lower than the dosage used in in vitro growth inhibition assay, suggesting different mechanisms of parasite-growth inhibition between in vivo and in vitro experiments. Hence, we investigated the effect of mannose on CM, a life-threatening form of malaria infection. Mannose treatment reduced the incidence of ECM in P. berghei-infected C57BL/6 mice based on the RMCBS (Figure 2A) and increased the survival rate to ~70% compared to the untreated group (Figure 2B). Consistently, BBB leakage was observed in infected mice, but the integrity of the BBB was protected by mannose treatment (Figure 2C). Parasite sequestration in the brains of infected mice was also diminished in the mannose-treated group when detected 7 days p.i. (Figure 2D). These results suggest that mannose treatment decreases the incidence of ECM.

Mannose treatment appears beneficial in eliminating CM, but the required dosage of administration is relatively high. Thus, we analyzed the general conditions of healthy mice treated with the same amount of mannose. Mannose treatment decreased the body weight of mice over a 7-day period (Supplementary Figure 3A). Similarly, food intake per cage was reduced upon mannose treatment (Supplementary Figure 3B). Loss of weight and food intake suggests that high-dose supplementation of mannose influences host metabolism. Therefore, we also measured biochemical indices in sera collected from the three groups that were treated for 20 days to assess their liver and kidney functions. ALT and AST, which represent the general liver condition, were not significantly changed among the groups (Supplementary Figure 3C). The same results were obtained when CREA and BUN, which indicate kidney function, were tested (Supplementary Figure 3D). Collectively, these results suggest that, although body weight and food intake were affected, the overall health condition of the mice was not compromised by treatment with 20% mannose.

Reduced parasitic burden in the peripheral blood and sequestration of iRBCs in the brain suggest enhanced parasite clearance in the mannose-treated group. Therefore, we analyzed the splenic macrophage population, which is mainly responsible for early parasite clearance (24), in ECM-susceptible C57BL/6 mice 4 and 7 days p.i. Detectable differences in parasitemia between untreated and treated groups of infected mice were observed from 5 days p.i. (Figure 3A). A swollen spleen, as measured by weight, was readily detected after infection. Comparing mannose-treated and untreated mice, spleen weights were lower 4 days p.i. but higher 7 days p.i. (Figure 3B). The percentage and absolute number of F4/80⁺ CD11b⁺ macrophages in the total population of CD45⁺ splenocytes were not remarkably different between the two infected groups 4 days p.i. (Figure 3C and Supplementary Figure 4A) but significantly increased in the mannose group 7 days p.i. (Figure 3D and Supplementary Figure 4B), consistent with the increased spleen weight (Figure 3B). The classification of splenic macrophages was subsequently analyzed and M1 macrophages (MHC II⁺ cells) found to comprise > 60% of total macrophages in the spleen. No significant differences in the macrophage population were observed between the mannose-treated and non-treated groups (Figure 3E). We also evaluated the phagocytotic capacities of the splenic macrophages, with or without mannose treatment, and found no obvious distinctions (Figure 3F). These results suggest that mannose promotes splenic macrophage accumulation which may participate in the removal of circulating parasites, but less likely influences their subtypes and phagocytosis.

Next, we tested the effects of mannose on splenic macrophages in uninfected C57BL/6 mice. As previously suggested, mannose treatment significantly decreased spleen weight and macrophage levels on day 7 (Figures 3G, H and Supplementary Figures 4C–F). Similar to infected mice, mannose had no effect on macrophage subtypes and their phagocytotic capacities in the uninfected mice (Figures 3I, J and Supplementary Figures 4G, H). These results suggest that mannose may play a distinct role in regulating splenic macrophage levels in P. berghei-infected C57BL/6 mice compared to uninfected, normal mice.

We also analyzed splenic macrophages in infected BALB/c mice. The spleen weight remained the same on 4 days p.i. and did not increase with mannose treatment at 7 days p.i. Instead, it slightly decreased on day 7, and the macrophage population showed no difference between the two infected groups 4 and 7 days p.i. (Supplementary Figure 5). These results suggest that mannose action is different in infected C57BL/6 mice than in BALB/c mice, likely due to differences in their genetic backgrounds and immune responses that are yet to be identified.

T-cell migration towards the brain and secretion of inflammatory factors have been shown to play important roles in CM pathogenesis (25). To investigate whether mannose treatment modulates the T-cell response during ECM development, we quantified the CD4⁺ and CD8⁺ T cell populations in peripheral blood and brain 7 days p.i., when CM-related neurological symptoms are most prominent. Circulating T cells were significantly increased upon infection compared to uninfected mice. In infected mice, CD8⁺ T cells slightly increased in the mannose-treated group, whereas CD4⁺ T cells did not change (Figures 4A, B and Supplementary Figures 6A, B). In contrast, the frequency and number of CD8⁺ T cells sequestered in the brain were significantly decreased after mannose treatment (Figures 4C, D and Supplementary Figures 6C, D). Because CXCR3 expression on T cells is associated with the migration of T cells to the brain (26), we tested CXCR3 expression on CD4⁺ and CD8⁺ T cells from the peripheral blood and brain. In peripheral blood, we noted a decrease in CXCR3 expression on both CD4⁺ and CD8⁺ T cells upon mannose treatment (Figure 4E). In the brain, we observed no detectable differences in CXCR3 expression on infiltrated T cells between the control and mannose-treated groups (Figure 4F). In addition, the expression of activation marker CD69 on T cells was significantly reduced in the peripheral blood and brain in the mannose-treated groups (Figures 4G, H).

The activation of T cells infiltrated in brain affects their secretion of cytokines, including IFN-γ and TNF-α, which play essential roles in ECM pathogenesis (27, 28). Therefore, we tested the mRNA expression levels of IFN-γ and TNF-α in brain tissues and found these two cytokines were significantly decreased in the infected mice treated with mannose (Supplementary Figures 6E, F). Taken together, these data suggest that mannose likely suppress ECM development by regulating T-cell activation and migration to the brain in mice infected with P. berghei ANKA.