In vivo experiments were approved by the Animal Experimentation Ethics Committee of Zhongshan School of Medicine on Laboratory Animal Care at Sun Yat-sen University (No. 2016-081) and were carried out in strict accordance with the Institutional Guidelines for Care and Use of Laboratory Animals.

Female Kunming mice (outbred, 6-8 weeks old) were obtained from the Animal Center of Sun Yat-sen University, and PbANKA parasites were used throughout the study. Some mice were injected intraperitoneally (i.p.) with 300 mM of α-lactose solution in phosphate buffered saline (PBS) twice daily starting from day 1 post infection (p.i.) until the day mice were sacrificed. Animals were sacrificed using CO₂ asphyxiation and the appropriate organs were taken for further analysis 12 h after the last treatment (27). A total of 48 mice were used in the experiments: 16 mice were injected i.p. with 10⁶ PbANKA-pRBCs (e.g. malarial mice); 16 mice were injected i.p. with 10⁶ PbANKA-pRBCs and treated with α-lactose; 8 mice were injected i.p. with α-lactose alone, and 8 mice were injected with equal volume of PBS as negative controls. Mortality was monitored daily. Peripheral blood parasitemia was monitored daily by Giemsa-stained thin blood smears from the tail vein of mice. Erythrocyte counts were performed with a hematocytometer, and more than 1,000 RBCs were counted by light microscopy (×100) to determine the percentage of pRBCs. Eight PbANKA-infected mice were sacrificed for examination on days 5 and 7 p.i., respectively.

For histopathological analysis, the livers from PbANKA-infected mice, PbANKA-infected mice with α-lactose treatment, or control mice were harvested and immediately fixed in 10% buffered formaldehyde (Guangzhou Chemical Reagent Factory, China) for 48 h. Five-micrometer-thick sections of the organs from mice were stained with hematoxylin and eosin (H&E) (Sigma-Aldrich, China) and evaluated for histological changes. A semi-quantitative score described previously was used to score liver inflammation and damage. Changes of the histopathological features in liver tissues were classified based on the severity of four histological criteria: architecture loss, sequestration of pRBCs in microvessels, pigment deposition, and inflammation (4, 28). The histopathological scores were graded on a scale of 0 to 3 [nil (0), partial loss (1), moderate loss (2), and total loss (3)]. The highest possible total score was 12 (4 histological criteria × 3 as highest scale). Score 0 means no histopathological change and score 12 refers to the most severe histopathological change (28). Overall liver tissues were analyzed at a magnification of × 100 under light microscopy by counting 10 fields of each section from each mouse in each group. All the analyses were performed by two researchers.

Immunohistochemistry was carried out using the streptavidin–biotin–peroxidase complex (SABC) method. Tissue sections (5-μm) were deparaffinized and rehydrated in distilled water. Heat-induced antigen retrieval was carried out in an 800-W microwave oven for 30 min. Endogenous peroxidase activity was blocked by incubation with 3% hydrogen peroxide in methanol for 10 min at 37°C. Nonspecific binding was blocked by incubation in 5% bovine serum albumin (Sigma-Aldrich) in PBS (pH 7.4) for 10 min at room temperature. The sections were incubated with rabbit anti-CD68 (Wuhan Boster Biological Engineering Co., Ltd., Wuhan, China) (1:200 dilutions) overnight at 4°C, and then with secondary antibodies. Sections incubated with secondary antibodies only were used as isotype controls. Signals were detected with a SABC kit and developed in diaminobenzidine tetrahydrochloride (Zhongshan Golden Bridge Technology, Beijing, China). The sections were counterstained with hematoxylin and examined under a light microscope. CD68⁺ macrophages were identified by dark-brown staining and were quantified using images captured with a digital camera system and analyzed by using Image-Pro Plus (Image Z1 software, Media Cybernetics, MD, USA). The number of cells in each field was determined under high power field as well as the area of each field (0.015066 mm²). The density of positive cells was expressed as the number of cells per square millimeter.

For primary antibody incubation, liver tissues were incubated with mouse anti-mouse CD68 monoclonal antibody (IgG1; 1:200; Abcam, Cambridge, UK) and rabbit anti-Tim-3 polyclonal antibody (IgG1; 1:200 dilution; Abcam) or rabbit anti-Gal-9 monoclonal antibody (IgG1; 1:200 dilution; Bioss, Beijing, China) overnight at 4°C. Subsequently, PBS (pH 7.4) washed sections were incubated with anti-mouse IgG (H + L), F (ab’)₂ fragment (Alexa Fluor. 488 Conjugate) (2 mg/ml, 1:200 dilution; Cell Signaling Technology, MD, USA) or anti-rabbit IgG (H + L), F (ab’)₂ fragment (Alexa Fluor. 555) (2 mg/ml, 1:200 dilution; Cell Signaling Technology) for 1 h at room temperature in a dark chamber. After washing 3 times with PBS, the slides were mounted with polyvinylpyrrolidone antifade mounting medium (Beyotime, Haimen, China) with DAPI (1: 50,000; Sigma-Aldrich) in a dark chamber. For negative controls, a set of liver tissue slides was incubated under similar conditions without the primary antibodies. CD68⁺-macrophages were identified by their green fluorescence in liver, whereas Tim-3⁺ and Gal-9⁺ cells were identified by their red fluorescence, once CD68 and Gal-9 or CD68 and Tim-3 superimposed in one image that would appear as yellow fluorescence under a fluorescence microscope (Olympus BX63, Tokyo, Japan). Two researchers performed the observations.

Formalin-fixed paraffin embedded liver tissue sections (5-μm) were detected using TUNEL assay (Wuhan Boster Biological Engineering Co., Ltd.). Negative control was prepared by omission of terminal deoxynucleotidyl transferase. Apoptotic cells were determined under high power field as well as the area of each field (0.015066 mm²). The density of positive cells was expressed as the number of cells per square millimeter.

Female Kunming mice were used in this study. Naive mice, mice with α-lactose treatment, and PbANKA-infected mice with or without α-lactose treatment on day 5 p.i. were injected i.p. with 2 ml 3% thioglycollate broth (Sigma-Aldrich) solution in PBS once daily for 3 days. Animals were sacrificed and their peritoneal lavage fluid were spun at 800 g at 4°C for 5 min, and the pelleted peritoneal macrophages were re-suspended and seeded at 5 × 10⁵ cells/well in 12-well plates (Corning, NY, USA). After 4 h at 37°C in a 5% CO₂ atmosphere, cells were washed and isolated. Samples were stored at −80°C until subjected to further analysis.

The peritoneal macrophages from PbANKA-infected mice, PbANKA-infected mice with α-lactose treatment, or control mice were isolated by centrifugation at 1 000 g for 3 min. The precipitates were immediately fixed in 3% glutaraldehyde and 1% osmium tetroxide (both in 100 mM PBS, pH 7.2) overnight before being dehydrated through a series of graded ethanol solutions. The fixed tissues were then embedded in SPIPon 812 Embedding Kit (Structure Probe Inc., West Chester, PA, USA) following the manufacture’s instruction. Ultrathin sections (70 nm) were cut from the embedded tissues using the Leica EM UC6 ultramicrotome (Leica Microsystems, Wetzlar, Germany) and mounted on formvar-coated grids. The sections were then stained for 15 min in aqueous 1% uranyl acetate followed by 0.2% lead citrate, and were then analyzed under a JEM100CX-II transmission electron microscope (JEOL Ltd., Tokyo, Japan) at an accelerating voltage of 100 kV.

Total RNA was extracted from about 100 mg of mouse liver and spleen tissues or peritoneal macrophages of each group using a commercial RNA Extraction Kit (TaKaRa Bio Inc., Shiga, Japan) according to the manufacturer’s protocol. The quality of total RNA was analyzed by running 5 μl of each RNA sample on a 1.0% agarose gel stained with ethidium bromide. The quantity of total RNA was estimated by measuring the ratio of absorbance at 260 and 280 nm using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, DE, USA). First-strand cDNA was constructed from 1.0 μg of total RNA with oligo (dT) as primers using a PrimeScript™ II 1st Strand cDNA Synthesis Kit (TaKaRa Bio Inc.) following the manufacturer’s protocol, and cDNA was stored at −80°C until use. To determine the mRNA levels of IFNα, IFNβ, IFNγ, TREM-1, TREM-2, Gal-9, and Tim-3 in tissues or macrophages of different groups of mice, qRT-PCR was performed using SYBR Green QPCR Master Mix (TaKaRa Bio Inc.) according to the manufacturer’s instructions. For liver parasite burden, mRNA level of PbANKA 18S rRNA measured by qRT-PCR is shown as -∆∆Ct values. Primers for the qRT-PCR are listed in Table 1 . Briefly, a total of 10 μl reaction mixture contained 5.0 μl of SYBR^® Premix Ex TaqTM (2×), 0.5 μl of each primer (10 pM), 3.0 μl of dH₂O, and 1.0 μl of cDNA (0.2 μg/μl). Amplification was pre-denaturized for 30 s at 95°C followed by 43 cycles of 5 s at 95°C and 20 s at 60°C with a LightCycler^® 480 instrument (Roche Diagnostics, AL, USA). Specific mRNA expression levels were normalized to that of the housekeeping gene, β-actin, and the results are expressed as fold change compared with uninfected controls (with or without α-lactose treatment).

Results of experimental studies were reported as mean ± SD. Statistical analysis of the data was performed using Wilcoxon rank sum test, Student’s t-test, and one-way ANOVA followed by Bonferoni’s multiple comparison tests using SPSS software for windows (version 19.0; SPSS Inc., Chicago, IL, USA). All graphs were performed using GraphPad Prism software and a value of P < 0.05 was considered statistically significant.

Malarial mice died between 7 and 12 days p.i., while malarial mice treated with α-lactose died between 6 and 10 days p.i. ( Figure 1A ). Malarial mice treated with α-lactose developed a significantly higher peripheral blood parasitemia on days 5 (P < 0.01), 6 (P < 0.01), 7 (P < 0.001), and 8 (P < 0.01) p.i. compared with those of malarial mice ( Figure 1B ). In addition, measured by using qRT-PCR, there were significantly increased parasite burdens (PbANKA 18S rRNA) in the livers of malarial mice treated with α-lactose compared with those of malarial mice on days 5 (P < 0.001) and 7 (P < 0.05) p.i. ( Figure 1C ). Thus, blockage of galectin-receptor interactions increased peripheral blood parasitemia and tissue parasite burden, and shortened the survival of mice after PbANKA infection.

Histopathological analysis showed no obvious morphological changes in the liver tissues of uninfected mice and uninfected mice treated with α-lactose. However, on days 5 and 7 p.i., obvious inflammation and necrotic lesions (indicated with green arrow heads) were observed in the livers of both malarial mice and malarial mice treated with α-lactose. Furthermore, extensive malarial pigment deposition was observed in the livers of malarial mice on day 5 p.i. and in the livers of malarial mice treated with α-lactose on days 5 and 7 p.i. ( Figure 2A ). Semi-quantitative inflammation scores based on pathological changes of the liver tissues showed that compared with malarial mice, significantly increased histopathological scores in the livers of malarial mice with α-lactose treatment on days 5 (P < 0.05) and 7 (P < 0.01) p.i. Compared with day 5 p.i., there were significantly increased histopathological scores in the livers of both malarial mice and malarial mice with α-lactose treatment on day 7 p.i. (P <0.01) ( Figure 2B ). The data suggested that more severe liver pathology developed in malarial mice with α-lactose treatment than that in malarial mice.

Apoptosis in the livers was measured by TUNEL staining. There were only a few apoptotic cells observed in the liver tissues of uninfected controls. On days 5 and 7 p.i., a great number of apoptotic cells were observed in the livers of both malarial mice and malarial mice treated with α-lactose. However, more apoptotic cells, scattered throughout the liver tissue section, were observed in the livers of malarial mice treated with α-lactose in comparison of those of malarial mice on days 5 and 7 p.i. ( Figure 3A ). Quantitative analysis of TUNEL positive cells showed that compared with uninfected controls, the numbers of apoptotic cells were significantly increased in the livers of malarial mice with or without α-lactose treatment on days 5 and 7 p.i. (P < 0.001). Compared with malarial mice, the numbers of apoptotic cells were significantly increased in the livers of malarial mice treated with α-lactose on days 5 (P < 0.01) and 7 (P < 0.001) p.i. ( Figure 3B ). Phagocytosis of pRBCs causes induction of apoptosis in macrophages through release of cytosolic factors from P. falciparum 3D7-infected mouse macrophages (29). The data demonstrated that blockage of galectin-receptor interactions may increase cellular apoptosis in the liver during the erythrocytic stage of malaria.

A few CD68⁺ macrophages (brown color) were observed in the liver tissues of uninfected controls. However, a large number of CD68⁺ macrophages were dispersed in the liver sections of malarial mice or malarial mice treated with α-lactose on days 5 and 7 p.i. ( Figure 4A ). Quantitative analysis of CD68⁺ macrophages in the liver tissues showed that compared with uninfected controls, there were significantly higher numbers of CD68⁺ macrophages in the livers of malarial mice on days 5 and 7 p.i. (P < 0.001). Compared with malarial mice, there were significantly increased CD68⁺ macrophages in the livers of malarial mice with α-lactose treatment on days 5 and 7 p.i. (P < 0.001). Compared with malarial mice or malarial mice with α-lactose treatment on day 5 p.i., there were significantly increased numbers of CD68⁺ macrophages in the livers of malarial mice and malarial mice with α-lactose treatment on day 7 p.i. (P < 0.01 and P < 0.05, respectively) ( Figure 4B ). It has been reported that the number of Kupffer cells is significantly increased and their phagocytic activity is enhanced during the erythrocytic stage of malaria in Sprague-Dawley rats infected with PbANKA (30). In addition, Kupffer cells show marked hypertrophy and hyperplasia and are filled with malarial pigment, and total Kupffer cell phagocytic activity in the liver during the erythrocytic phase of malaria is also markedly increased in Wistar rats infected with PbANKA (31). Our data indicate that blockage of galectin-receptor interactions may increase the activity of liver CD68⁺ macrophages in the erythrocytic stage of malaria, which is in agreement with the increased parasitemia and tissue parasite burden in malarial mice with α-lactose treatment.

Immunofluorescence assay showed that both Gal-9 ( Figure 5A ) and Tim-3 ( Figure 5B ) were expressed on the CD68⁺ macrophages in the liver tissues of all the groups; more CD68⁺-Gal-9⁺ and CD68⁺-Tim-3⁺ macrophages were observed in malarial mice and malarial mice treated with α-lactose in comparison of uninfected controls. The data suggested that CD68⁺-Gal-9⁺ and CD68⁺-Tim-3⁺ macrophages play a role in the inflammatory response in liver during the erythrocytic stage of malaria.

Here, the mRNA expression levels of Gal-9, Tim-3, IFNα, IFNβ, IFNγ, TREM-1, and TREM-2 in the livers and spleens of PbANKA-infected mice with or without α-lactose treatment on day 7 p.i. were examined ( Figure 6 ). Compared with uninfected controls, there were significantly increased Gal-9 (P < 0.05 and P < 0.001, respectively) and Tim-3 (P < 0.01 and P < 0.001, respectively) in the livers, and significantly increased Gal-9 (P < 0.001) and Tim-3 (P < 0.01) in the spleens of malarial mice and malarial mice with α-lactose treatment on day 7 p.i. Compared with malarial mice, there were significantly increased mRNA levels of Gal-9 and Tim-3 in the livers of malarial mice with α-lactose treatment on day 7 p.i. (P < 0.01).

Compared with uninfected controls, there were significantly increased mRNA expression levels of IFNα (P < 0.01 and P < 0.001, respectively), IFNγ (P < 0.001), and TREM-1 (P < 0.05 and P < 0.001, respectively) in the livers, and significantly increased levels of IFNα (P < 0.01), IFNγ (P < 0.001), and TREM-1 (P < 0.001) in the spleens of malarial mice and malarial mice with α-lactose treatment on day 7 p.i. Compared with malarial mice, there were significantly increased levels of IFNα (P < 0.01), IFNγ (P < 0.001), and TREM-1 (P < 0.05) in the livers and significantly increased TREM-1 level (P < 0.05) in the spleens of malarial mice with α-lactose treatment on day 7 p.i.

The ultrastructures of peritoneal macrophages isolated from different groups of mice were observed by transmission electron microscopy, the peritoneal macrophages from PbANKA-infected mice with α-lactose treatment showed more pseudopodia in comparison of those from PbANKA-infected mice ( Figure 7 )

To test whether galectin-receptor interactions have an impact on functional activation of macrophages by PbANKA infection, we examined the mRNA expression levels of Gal-9, Tim-3, IFNα, IFNβ, IFNγ, TREM-1, and TREM-2 in the peritoneal macrophages ex vivo isolated from PbANKA-infected mice with or without α-lactose treatment on day 7 p.i. Compared with uninfected controls, there were significantly increased levels of Gal-9 (P < 0.05 and P < 0.001, respectively), Tim-3 (P < 0.05), IFNα (P < 0.01 and P < 0.001, respectively), IFNβ (P < 0.01 and P < 0.001, respectively), IFNγ (P < 0.01 and P < 0.001, respectively), and TREM-1 (P < 0.001) in the peritoneal macrophages isolated from both PbANKA-infected mice and PbANKA-infected mice treated with α-lactose. Compared with PbANKA-infected mice, there were significantly increased levels of Gal-9 (P < 0.05), IFNα (P < 0.05), IFNβ (P < 0.05), IFNγ (P < 0.01), and TREM-1 (P < 0.001) in those from PbANKA-infected mice treated with α-lactose ( Figure 8 ). Studies has revealed a substantial deleterious role for IFN-I-signaling in mediating severe and lethal disease during PbANKA infection (22). IFN-γ is a key contributor to the pathogenesis of cerebral malaria in the PbANKA mouse model (32). Using PbANKA-infected ICR mice to investigate the involvement of TREM-1 during malaria infection, and found that mice with low parasitemia levels have low concentrations of plasma TREM-1 and vice versa. Therefore, TREM-1 plays important pro-inflammatory roles during malaria infection and may be one of the key mediators of the disease’s severity (33). The data suggested that blockage of galectin-receptor interactions induces the expressions of IFN-I, IFNγ, and TREM-1 genes, which may exacerbate liver damage of malarial mice.

The correlations between the mRNA levels of Gal-9 and Tim-3/IFNα/IFNβ/IFNγ/TREM-1 in the livers, spleens, and peritoneal macrophages of PbANKA-infected mice or PbANKA-infected mice treated with α-lactose were evaluated, and only significant correlations were presented. There were significant correlations between the mRNA levels of Gal-9 and Tim-3 (r = 0.9966, P = 0.0034), Gal-9 and IFNγ (r = 0.9922, P = 0.0078), and Gal-9 and TREM-1 (r = 0.9847, P = 0.0153) in the livers of PbANKA-infected mice ( Figure 9A ), and there was significant correlation between the mRNA levels of Gal-9 and IFNγ (r = 0.9839, P = 0.0161) in the livers of PbANKA-infected mice treated with α-lactose ( Figure 9B ). There were significant correlations between the mRNA levels of Gal-9 and IFNγ (r = 0.9947, P = 0.0053), and Gal-9 and TREM-1 (r = 0.9918, P = 0.0882) in the spleens of PbANKA-infected mice ( Figure 9C ). There were significant correlations between the mRNA levels of Gal-9 and Tim-3 (r = 0.9642, P = 0.0396), Gal-9 and IFNγ (r = 0.9822, P = 0.0239), and Gal-9 and TREM-1 (r = 0.9998, P = 0.0069) in the peritoneal macrophages isolated from PbANKA-infected mice ( Figure 9D ), and there was significant correlation between the mRNA levels of Gal-9 and IFNα (r = 0.9225, P = 0.0451) in the peritoneal macrophages isolated from PbANKA-infected mice treated with α-lactose ( Figure 9E ). Taken together, significant positive correlations existed between the mRNA levels of Gal-9 and Tim-3/IFNγ/TREM-1 in both the livers and peritoneal macrophages, and between Gal-9 and Tim-3/TREM-1 in the spleens of PbANKA-infected mice; significant positive correlations existed between the mRNA levels of Gal-9 and IFNγ in the livers and between Gal-9 and IFNα in the peritoneal macrophages from PbANKA-infected mice treated with α-lactose. The data suggested that Gal-9/Tim-3 pathway may play an important role in the regulation of liver inflammatory response including upregulation of TREM-1, IFNγ, and IFNα expressions during the erythrocytic stage of PbANKA infection.