Human peripheral blood samples were donated by healthy volunteers and patients. The information of all individuals involved was anonymized. Written consent was obtained from all individuals for the publication of this study. All procedures performed on human samples were carried out in accordance with the tenets of the World Medical Association's Declaration of Helsinki. All procedures performed on the animals in this study were conducted according to the animal husbandry guidelines of the Chinese Academy of Medical Sciences. Studies on humans and animals were reviewed and approved by the Ethical Committee and the Experimental Animal Committee of the Chinese Academy of Medical Sciences.

Twenty-one patients suffering from falciparum malaria (FM) and 16 healthy individuals were recruited at Beijing Friendship Hospital at Capital Medical University from March 2015 to August 2016. Peripheral blood and plasma were obtained from all the subjects. The characteristics of the patients and healthy individuals are summarized in Supplementary Table 1. All FM patients had primary infections, documented by Giemsa-stained thin blood smears for parasite identification and confirmed by nest PCR that targets variant sequences in the small subunit ribosomal RNA genes (Kimura et al., 1997). All samples were obtained before treatment.

Six-week-old male C57BL/6 mice (special pathogen free) were purchased from Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). All mice were maintained in a pathogen-free facility and randomly divided into groups. PbANKA-infected mice were constructed as previously described (Hou et al., 2016), and they were infected intraperitoneally with 10⁶ parasitized red blood cells. For anti-Tim-3 treatment, the mice received one intraperitoneal injection of 100 μg of anti-Fc antibody (eBioscience, San Diego, USA) to reduce non-specific binding, followed by treatment with either 100 μg of purified anti-mouse Tim-3 antibody (catalog no. 14-5870, eBioscience) or purified rat IgG2α K isotype control (catalog no. 14-4321-85, eBioscience) once a day after infection. Parasitaemia was also monitored daily by examining tail blood smears stained with Giemsa stain (Sigma-Aldrich, St. Louis, MO, USA). The smears were observed using a digital camera and analyzed using the Image-Pro Plus 6.0 software.

Murine spleens were cut into pieces, minced, and pressed through 200-gauge stainless steel mesh. Then, the red blood cells (RBCs) were depleted with a red blood cell lysis solution as previously described (Hou et al., 2016). Murine splenic mononuclear cells were isolated by gradient centrifugation with HISTOPAQUE-1083 (Sigma-Aldrich) according to the manufacturer's instruction. Macrophages were isolated from murine splenic mononuclear cells using CD11b Microbeads (Miltenyi, Bergisch-Gladbach, Germany) according to the manufacturer's protocol. This procedure consistently yields a population of purified macrophages (>90% CD11b⁺ by flow cytometry) with more than 98% viability, as indicated by trypan blue exclusion. The mouse RBCs were enriched by gradient centrifugation using HISTOPAQUE-1083.

Flow cytometry was conducted as previously described (Hou et al., 2016). The antibodies used in this study were as follows: Anti-Human CD14 PerCP-Cy5.5, mouse IgG1κ isotype control PerCP-Cy5.5, anti-mouse CD11b FITC, rat IgG2a K isotype control FITC, anti-mouse TIM-3 PE, and rat IgG2α K isotype control PE (all obtained from eBioscience, San Diego, CA, USA), anti-human Tim-3-PE, and rat IgG_2A isotope control-PE (both obtained from R&D Systems). The cells were detected and analyzed using a FACS Canto II flow cytometer (BD Biosciences, San Jose, CA, USA), and the gates for positive cells were defined using the isotype controls.

Murine splenic mononuclear cells or splenic CD11b⁺ cells were cultured in RPMI 1640 medium (HyClone, Thermo, Beijing, China) supplemented with 10% fetal bovine serum (GIBCO, Grand Island, USA) and plated at 5 × 10⁵ cells per well in 96-well polystyrene plates. Three micrograms or milliliters of the anti-mouse TIM-3 purified antibody were simultaneously added and incubated for 30 min to block the Tim-3 signal pathway. Purified rat IgG2α K isotype was used as a control antibody. PbANKA-infected red blood cells with a parasitaemia >30% (Pb-iRBC) were then added into the indicated wells at a concentration of 1 × 10⁶ per well.

Phagocytosis was assessed by incubating murine mononuclear cells with or without Tim-3 antibodies in suspension for 4 h with Pb-iRBC prestained using the CellTrace™ Far Red Cell Proliferation Kit (Thermo Fisher Scientific, Carlsbad, CA, USA), according to the manufacturer's instruction. The frequencies of total monocytes/macrophages positive for Far Red were determined by flow cytometry.

Murine splenic CD11b⁺ cells treated with or without anti-Tim-3 antibodies were co-cultured with Pb-iRBC or RBCs from uninfected mice (uRBC) to detect parasiticidal mediator production. The cells were incubated for 24 h and then collected for subsequent experiments. The supernatants were harvested and maintained at −80°C. The levels of tumor necrosis factor (TNF-α) and nitric oxide (NO) in human plasma, mouse serum and cultured cell supernatants from murine CD11b⁺ cells were determined using the Human TNF-α ELISA Kit, the Mouse TNF-α ELISA Kit (all from R&D Systems) and the Total Nitric Oxide Assay Kit (Beyotime Biotechnology, Shanghai, China), respectively, according to the manufacturer's instructions.

Quantitative real-time PCR was performed to analyse the expression of genes in both human and mouse immune cells. The details of the QRT-PCR assay were previously described (Hou et al., 2015). The primers for the genes were derived from the Primer Bank (https://pga.mgh.harvard.edu/primerbank/) and are listed in Supplementary Table 2. The relative expression level of each gene was analyzed using SDS 1.4 software (Applied Biosystems, Carlsbad, USA).

CD11b⁺ cells were lysed with lysis buffer (Cytobuster, Novagen, San Diego, CA) for 30 min on ice, The mixture was centrifuged at 12,000 rpm for 30 min at 4°C and protein concentrations were quantified by the BCA protein Assay Kit (Pierce, Rockford, IL, USA). Fifty micrograms of protein of each sample was separated by SDS-PAGE and transferred onto polyvinylidene difluoride membranes (Millipore, Bedford, MA, USA). After blocking with 5% dried milk (wt/vol) in PBS for 1 h at room temperature, each membrane was incubated in the same buffer with CD36 monoclonal Antibody, CD54 (ICAM-1) Monoclonal Antibody (all from ThermoFisher Scientific, Rockford, IL, USA, diluted 1:1,000) or IgG control for overnight at 4°C, this was followed by washing and subsequent incubation with IRDye 800 CW Conjugated goat Anti-mouse IgG (H+L) antibody (Li-COR Biosciences, Lincoln, Nebraska, USA), and detection were carried out using Odyssey (Li-COR).

The data were analyzed using the GraphPad Prism 5.0 software and Microsoft Excel 2007. The results were analyzed using a 2-tailed paired t-test. The Wilcoxon test for paired samples was used for the data that did not fit a Gaussian distribution. Values of p < 0.05 were considered significant.

The quantity of peripheral monocytes and the expression of Tim-3 in monocytes of patients suffering from FM and healthy individuals were examined by flow cytometry. The proportion of CD14⁺ cells in peripheral leukocytes of FM patients was much higher than that of healthy individuals (Figure 1A), and the frequency of Tim-3⁺ cells in CD14⁺ cells was significantly decreased in FM patients (Figure 1B).

PbANKA is lethal in mice and has been widely used as experimental models for studies of human malaria diseases (Troye-Blomberg et al., 1994). Thus, we used PbANKA to study macrophages in murine malaria. The number of CD11b⁺ macrophages in the spleens of PbANKA-infected mice, counted manually, peaked at day 5 post-infection (Figure 2A). The proportion of splenic CD11b⁺ macrophages in the macrophage-like population reached a peak at day 3 post-infection (Figure 2C). However, the quantity of CD11b⁺ macrophages did not continuously increase. After the peak, both the number and the proportion started to decrease, but the number of splenic CD11b⁺ macrophages at day 9 was still much higher than that at day 0 (day 0, 0.94 ± 0.36% vs. day 9, 10.53 ± 1.94%, p = 0.0011, Figure 2A). Tim-3 expression in splenic CD11b⁺ macrophages of PbANKA-infected mice sharply decreased until day 3 post-infection and then gradually increased (Figure 2D). The Tim-3 expression in CD11b⁺ macrophages at day 9 post-infection was still lower than that at day 0 (day 0, 64.37 ± 1.05% vs. day 9, 47.27 ± 1.39%, p < 0.0001, Figure 2D).

Pb-iRBCs previously stained with CellTrace™ Far Red was used to quantify the level of phagocytosis of macrophages from PbANKA-infected mice. Splenic CD11b⁺ macrophages from mice at day 3 post-infection displayed significantly higher levels of phagocytosis of Pb-iRBCs compared to those from mice at day 0. However, the phagocytic ability of CD11b⁺ macrophages at day 5 post-infection sharply declined and was even lower than those at day 0 (Figures 3A,B). Tim-3 signal blockade with the anti-Tim-3 antibody effectively elevated the phagocytosis of Pb-iRBCs by macrophages, especially at day 3 post-infection (Figures 3A,B). Adhesion molecules, including cluster of differentiation 36 (CD36), intercellular adhesion molecule (ICAM)-1, vascular cell adhesion molecule (VCAM)-1 and platelet endothelial cell adhesion molecule (PECAM)-1, may favor the binding and uptake of parasite-infected RBCs by macrophages (Carvalho et al., 2010; Antonelli et al., 2014). The expression of the coding genes of splenic macrophages during the infectious periods was detected by QRT-PCR. The four molecules displayed different expression patterns. The level of CD36 sharply decreased as soon as PbANKA infection initiated. The levels of ICAM-1 and VCAM-1 increased to a peak and then gradually decreased. The level of PECAM-1 declined at day 1 post-infection and then gradually increased. The expression of ICAM-1 was much higher than that of the other genes during the first 5 days post-infection (p < 0.0001, Figure 3C). Similar result of CD36 expression in PbANKA-infected splenic CD11b⁺ macrophages were obtained by Western Blot, while the protein level of ICAM-1 decreased as soon as PbANKA infection initiated and then gradually increased to a peak at day 5 (Supplementary Figure 1A). To investigate the relationship of Tim-3 and these adhesion molecules, Tim-3 antibodies were applied to block the Tim-3 signaling pathway in mice models. PbANKA-infected mice received one intraperitoneal injection of 100 μg of an anti-mouse Tim-3 antibody or the IgG control every other day following infection. The splenic CD11b⁺ macrophages were obtained at days 0, 3, and 5 post-infection. The QRT-PCR results showed that the splenic CD11b⁺ macrophages from anti-Tim-3 antibody treated mice had higher expression levels of ICAM-1, VCAM-1, and PECAM-1 compared to the control groups (Figure 3D). Similar results of CD36 and ICAM-1 expression in splenic CD11b⁺ macrophages from PbANKA-infected mice were obtained by Western Blot (Supplementary Figure 1B). Thus, ICAM-1 was the most susceptible gene, while CD36 was not influenced by Tim-3 antibodies.

TNF-α is the key effector molecule released mainly by activated monocyte/macrophages during malaria (Giribaldi et al., 2010). The ELISA results showed that the level of TNF-α in the plasma of FM patients was much higher than that of healthy individuals (Figure 4A left panel) and the level of TNF-α in the sera of PbANKA-infected mice, detected along the infectious periods, quickly elevated to a peak at day 3 post-infection and then started to decline (Figure 4A right panel). The splenic CD11b⁺ macrophages from mice 0, 1, 3, 5, 7, and 9 days post-infection were obtained to detect the mRNA expression of TNF-α. The QRT-PCR results showed that variation in TNF-α mRNA expression in splenic CD11b⁺ macrophages was consistent with the variation in the TNF-α concentration in the sera of infected mice (Figure 4B). Our in vitro assay showed that splenic CD11b⁺ cells from PbANKA-infected mice at days 0, 3, and 5 post-infection co-cultured with Pb-iRBC had higher TNF-α production in the supernatant compared to that co-cultured with uRBC, and anti-Tim-3 antibody treatment further promoted TNF-α production (Figure 4C). In vivo assays were also carried out, and splenic CD11b⁺ cells from the anti-mouse Tim-3 antibody- or IgG control-treated PbANKA mice were obtained at days 0, 3, and 5 post-infection. The QRT-PCR results showed that the anti-Tim-3 treated groups had higher expression levels of TNF-α compared to that of the control groups (Figure 4D).

NO is also a key effector molecule made by activated monocyte/macrophages (Greve et al., 1999). The level of NO in the plasma was not significantly different between FM patients and healthy individuals (Figure 5A). The NO concentration in the plasma from PbANKA-infected mice sharply increased to a peak at day 1 post-infection and then quickly decline (Figure 5B). Serum samples were obtained from PbANKA-infected mice treated with anti-Tim-3 antibodies or IgG control, and the NO level was not different between the two groups (data not shown). Our in vitro assay showed that splenic CD11b⁺ cells from PbANKA-infected mice co-cultured with Pb-iRBC had higher NO production in the supernatant compared to those co-cultured with uninfected RBC, and anti-Tim-3 treatment further promoted the NO production (Figure 5C).

The splenic CD11b⁺ macrophages from mice at 0, 1, 3, 5, 7, and 9 days post-infection were obtained and used to detect the key markers of classically activated macrophages, inducible nitric oxide synthase (iNOS) and IL (interleukin)-12, and the key markers of alternatively activated macrophages, arginase-1 (Arg1) and IL-10 (Murray et al., 2014). The QRT-PCR results showed that the expression of iNOS and IL-12 peaked at day 3 post-infection and then declined, while the expression of Arg1 and IL-10 slowly increased in the first 3 days and rapidly increased in the following days (Figure 6A). To investigate the effect of Tim-3 on macrophage polarization, splenic CD11b⁺ cells from PbANKA-infected mice treated with the anti-Tim-3 antibodies or IgG control were obtained at days 0, 3, and 5 post-infection. Real-time PCR exhibited that the Tim-3 signal blockade by anti-Tim-3 antibodies promoted the expression of iNOS and IL-12 and inhibited the expression of Arg1 and IL-10 (Figure 6B).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.