A Schistosoma japonicum Infection Promotes the Expansion of Myeloid-Derived Suppressor Cells by Activating the JAK/STAT3 Pathway

Schistosomiasis is a worldwide, chronic, parasitic disease that is caused by blood flukes and leads to significant morbidity and mortality (1). After infection, the schistosomula and its eggs migrate through a variety of tissues, such as the liver, intestines, and vesical mucosa (2, 3). Soluble egg Ag (SEA) and schistosome worm Ag (SWA) are the main Ags related to Schistosoma japonicum infection (4), which induce multiple cytokines, including IFN-γ, IL-4, IL-5, IL-10, IL-17, IL-22, and IL-33, and mediate the immune response (4–7). A liver granulomatous reaction is the major pathological change and is tightly correlated with typical clinical symptoms of schistosomiasis, such as portal hypertension and jaundice (8).

Myeloid-derived suppressor cells (MDSCs) are a heterogeneous population of immature myeloid cells that promote tumor progression by inhibiting the function of T cells, especially CD8⁺ T cells (9). Recent studies showed that MDSCs could be induced in some virus infectious diseases, such as HIV and hepatitis C virus (10, 11), as well as in parasite infection (12). MDSCs can be divided into two groups, granulocytic (G-MDSCs) and monocytic (M-MDSCs) populations, according to the expression of the Ly6G and Ly6C molecules. Each group has a different biological function and different mechanism involved in its mediation function (13, 14).

MDSCs are inflammatory cells that secrete many types of cytokines, including GM-CSF, IL-1α, IL-6, and IL-10, among others. However, the primary suppression function-related cytokines of MDSCs are TGF-β and IL-10 (11). Additionally, the expression of programmed cell death 1 ligand (PD-L)1/2, which causes T cell apoptosis, was induced in myeloid cells, including MDSCs, during some pathologic conditions, such as tumors and pathogen infection (15, 16).

It has been reported that MDSCs inhibit the T cell response mainly by altering l-arginine pathways during cancer or other pathologic conditions (17, 18). The mechanism might proceed in three different ways, including an increase in the expression of ARG1, induction of the activity of inducible NO synthase (iNOS), and production of reactive oxygen species (ROS). Although many types of mechanisms are involved in the regulation of ROS production, NADPH oxidase (NOX)2 is the key factor in MDSCs. NOX2 is constructed of two membrane proteins, gp91^phox and p22^phox, and four cytoplasmic components, p67^phox, p40^phox, p47^phox, and rac1 (a type of small G protein) (19).

S100A8/A9 are important inflammation factors that play a key role in MDSC expansion (20). An increase in MDSC accumulation and tumor growth could be stopped in mice with a S100A9 gene knockout (21). Recent research found that S100A8/A9 were important target genes in the STAT3 signaling pathway. STAT3 enhances MDSC proliferation by inducing S100A8 and S100A8 expression (20). The JAK signal transducer and activator of transcription 3 signaling pathway inhibitor 124 (JSI-124) is special inhibitor of the JAK/STAT3 signaling pathway, which blocks STAT3 phosphorylation and activation by inhibiting the activation of JAK kinase (22).

In this study, C57BL/6 mice were infected by S. japonicum, the phenotype and function of infection-induced MDSCs were investigated, and the mechanism was explored.

Female C57BL/6 and BALB/c mice were purchased from the Animal Experimental Center of Sun Yat-Sen University (Guangzhou, China). All of the mice were maintained under pathogen-free conditions and used at 6–8 wk of age. S. japonicum cercariae were shed from naturally infected Oncomelania hupensis snails, which were purchased from the Jiangsu Institute of Parasitic Disease (Wuxi, China). C57BL/6 mice were infected percutaneously with 40 ± 5 cercariae and sacrificed at 5–6 wk postinfection. Animal experiments were performed in strict accordance with the regulations for the Administration of Affairs Concerning Experimental Animals, and all efforts were made to minimize suffering.

RPMI 1640, FBS, 2-ME, penicillin, 5-(and-6)-chloromethyl-2′,7′- dichlorodihydrofluorescein diacetate, acetyl ester (CM-H2DCFDA), CFSE, and streptomycin were obtained from Invitrogen (Grand Island, NY). Recombinant murine GM-CSF and IL-6 were purchased from PeproTech (Oak Park, CA). Abs against p47^phox, gp91^phox, S100A8, S100A9, STAT3, p-STAT3, STAT6, p-STAT6, and β-actin as well as HRP-conjugated secondary Abs were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). N^ω-hydroxy-nor-l-arginine (nor-NOHA) and N^G-monomethyl-l-arginine (L-NMMA), and nimesulide were obtained from Cayman Chemical (Ann Arbor, MI). N-acetyl-l-cysteine (NAC), l-arginine, PMA, brefeldin A, ionomycin, Con A, cucurbitacin I hydrate (JSI-124), and DMSO were purchased from Sigma-Aldrich (St. Louis, MO). Fluorescein-conjugated anti-mouse Abs (CD3e-PE-Cy7, CD4-PE, CD8a-PE-Cy5, Gr-1–PE-Cy7, Gr-1–PE, Gr-1–FITC, CD11b-FITC, CD11b-PE-Cy7, CD11b-allophycocyanin-Cy7, CD11c-PE-Cy5, PD-L1–allophycocyanin, PD-L2–Brilliant Violet 421, Ly6C-PerCP-Cy5.5, CD49d-PE, CD117-PE-Cy7, CD115-allophycocyanin-Cy7, CD135, F4/80-PE-Cy7, GM-CSF-PerCP-Cy5.5, IL-1α–PE, IL-10–allophycocyanin, and IL-6–allophycocyanin) and their corresponding isotype controls were obtained from eBioscience (San Diego, CA). Ly6G-PE, Ly6G-allophycocyanin, and Ly6C-allophycocyanin were obtained from BD Biosciences (San Jose, CA), and MHC class II–PE was from Miltenyi Biotec (Auburn, CA).

S. japonicum cercariae, SEA, and SWA were obtained from the Jiangsu Institute of Parasitic Diseases. SEA and SWA were sterile filtered and the endotoxin was removed with polymyxin B agarose beads (Sigma-Aldrich). A Limulus amebocyte lysate assay kit (Lonza, Basel, Switzerland) was used to confirm the removal of the endotoxin from SEA and SWA as previously described (23).

Mice were narcotized, the precava was cut, and sterile normal saline was injected to remove blood from the liver through the ventriculus sinister. Then, the liver, spleen, and mesenteric lymph nodes were removed, pressed through a 100-μm cell strainer (BD Falcon), and suspended in HBSS. Lymphocytes in liver were isolated using Ficoll-Hypaque (Dakewe Biotech, Shenzhen, China) by density gradient centrifugation. Lymphocytes in spleen were isolated using blood cell lysis buffer (Dakewe Biotech). Isolated cells were washed twice with HBSS and resuspended at 2 × 10⁶ cells/ml in complete RPMI 1640 medium supplemented with 10% heat-inactivated FCS, 100 U/ml penicillin, 100 μg/ml streptomycin, 2 mM l-glutamine, and 50 μM 2-ME.

Mouse bone marrow (BM) cells were obtained from the femur and tibia. One million BM cells were cultured in RPMI 1640 medium supplemented with 10% FBS, 20 ng/ml GM-CSF, 50 μM 2-ME, and 20 ng/ml IL-6. The cultures were maintained at 37°C in a 5% CO₂-humidified atmosphere in 48-well plates. The medium was refreshed on day 3. The cells were analyzed by flow cytometry on day 6.

These experiments were performed following previously described procedures (24). In brief, RNA was extracted with an RNase mini kit and cDNA was synthesized with a SuperScript III reverse transcriptase kit (Qiagen, Valencia, CA). Real-time PCR amplification was carried out in the presence of 2.5 μl of cDNA template, 12.5 μl of SYBR master mixture (Applied Biosystems, Foster City, CA), and target gene-specific primers (Table I) in a CFX Connect Real-Time PCR Detection System (Bio-Rad, Hercules, CA). Amplification of β-actin was used as an internal control.

Cultured or purified cells were collected and lysed. The protein concentration was measured by a bicinchoninic acid protein assay kit (Beyotime). Protein sample was separated in 10% SDS-denatured polyacrylamide gel and transferred onto a polyvinylidene difluoride membrane. The polyvinylidene difluoride membranes were blocked with 5% skim milk in TBST at room temperature for 2 h. The targeted molecules were probed using specific primary Abs and HRP-conjugated secondary Abs and detected with an ECL HRP chemiluminescent substrate reagent kit (Invitrogen, Carlsbad, CA).

Cells were washed twice in PBS and blocked in PBS buffer containing 1% BSA for 30 min. Then, the cells were stained with conjugated Abs that were specific for cell surface Ags for 30 min at 4°C in the dark. These Ags included CD11b, Gr-1, Ly6G, Ly6C, PD-L1, PD-L2, F4/80, CD49d, CD115, CD117 (c-Kit), and CD135 (Flt-3). The stained lymphocytes were analyzed using flow cytometry (Beckman Coulter, Fullerton, CA), and the results were analyzed with the program FlowJo version 6.0 (Tree Star, Ashland, OR). Isotype-matched cytokine controls were included in each staining protocol.

For purification of MDSCs, cultured BM cells or mouse splenocytes were stained with PE-CD11b and PE–Gr-1 Abs. Gr-1⁺CD11b⁺ cells were isolated by cell sorting on a FACSAria cell sorter (BD Biosciences, Mountain View, CA). For isolation of T cells, mouse splenocytes were stained with the CD3e-PE-Cy7 Ab, and CD3⁺ T cells were purified by flow cytometric sorting.

Single-cell suspensions from the spleens of control mice and mice infected with S. japonicum were stimulated with 20 ng/ml PMA plus 1 μg/ml ionomycin for 5 h, at 37°C, under a 5% CO₂ atmosphere. Brefeldin A (10 g/ml; Sigma-Aldrich, Shanghai, China) was added during the last 4 h of incubation. Cells were washed twice in PBS, fixed with 4% paraformaldehyde, and permeabilized overnight at 4°C in PBS buffer containing 0.1% saponin (Sigma-Aldrich), 0.1% BSA, and 0.05% NaN₃. Cells were then stained for 30 min at 4°C in the dark with conjugated Abs specific for the cell surface Ags CD11b and Gr-1 as well as the cytokines GM-CSF, IL-1α, IL-6, and IL-10. The expression phenotypes of the Ab-labeled lymphocytes were analyzed by flow cytometry (BD FACSCalibur and FACSAria II) and the results were analyzed with the software FlowJo version 6.0 (Tree Star). Isotype-matched cytokine controls were included in each staining protocol.

The oxidation-sensitive dye CM-H2DCFDA (Invitrogen) was used to measure ROS production. Cells were incubated at 37°C in RPMI 1640 in the presence of 1 μM CM-H2DCFDA for 30 min and then labeled with CD11b-PE-Cy7 and Gr-1-PE Abs on ice. The ROS content in MDSCs was analyzed by flow cytometry.

Previously described procedures were employed (25). Briefly, 0.5 million cells were lysed with 100 μl of buffer containing 0.1% Triton X-100 and 10 mM Tris-HCl for 30 min. Arginine hydrolysis was performed by incubating the lysate with 0.5 M l-arginine at 37°C for 120 min. Urea concentration was measured at 540 nm after the addition of α-isonitrosopropiophenone (dissolved in 100% ethanol), followed by heating at 95°C for 30 min. A 1 mM urea standard and dH₂O were used as controls. The arginase activity was calculated following the manufacturer’s instructions (BioAssay Systems, Hayward, CA). NO content in plasma was measured following the manufacturer’s protocol (BioVision, Milpitas, CA). An equal volume of plasma (100 μl) was mixed with the Griess reagent and incubated for 10 min at room temperature. The absorbance at 550 nm was measured using microplate plate reader (Bio-Rad). Nitrite concentrations were determined by comparing the absorbance values of test samples to a standard curve generated by serial dilution of 0.25 mM sodium nitrite.

T cell proliferation was determined by CFSE dilution. Purified CD3⁺ T cells from BALB/c mice were labeled with CFSE (1 μM) (Invitrogen), stimulated with Con A (5 μg/ml), and cultured alone or cocultured with allogeneic MDSCs (from C57BL/6 mice) at different ratios for 3 d. Cells were then stained with CD4-PE or CD8a-PE-Cy7 Abs, and T cell proliferation was analyzed by flow cytometry.

Statistical analysis was conducted using unpaired t tests (GraphPad Prism version 5.0a and SPSS Statistics 17.0). A p value <0.05 was considered statistically significant.

Mice were sacrificed 5–6 wk postinfection, the livers were harvested, and sections were made as described in 2Materials and Methods. H&E staining showed clear lesions and granuloma in infected mouse livers (Fig. 1A, 1B). To explore the existence and distribution of MDSCs in S. japonicum–infected mice at 5–6 wk postinfection, mononuclear cells were isolated from mouse BM, spleen, and mesenteric lymph nodes, and the percentage of CD11b and Gr-1 coexpressed MDSCs was detected by FACS. The results demonstrated that the percentage and absolute numbers of CD11b⁺Gr-1⁺ MDSCs in all of these organs in infected mice were higher than that in normal mice (p < 0.05, Fig. 1C, 1D). Furthermore, different fluorescence-labeled mAbs to ly6G, Ly6C, F4/80, CD49d, CD115, CD117 (c-Kit), and CD135 (Flt-3) were used to define the CD11b⁺Gr-1⁺ MDSCs, and Wright staining was performed to show the Gr-1⁺CD11b⁺ cell morphology. Results showed that the expression of Ly6G, CD115, CD117, and CD135 on MDSCs was increased significantly after infection (p < 0.05, Fig. 1E, 1F). Also, much smaller mononuclear leukocyte and multinuclear cell were observed in isolated MDSCs from infected mice (Fig. 1G). Moreover, the subtype of S. japonicum infection–induced MDSCs was also investigated by FACS. As shown in Fig. 1H, the percentage of CD11b⁺Ly6G⁺Ly6C^−/low G-MDSCs increased significantly (p < 0.05). However, the change of CD11b ⁺Ly6G⁻Ly6C^high M-MDSCs was not significant (p > 0.05).

To further explore the function of S. japonicum infection–induced MDSCs, lymphocytes were isolated from S. japonicum–infected mouse spleen. CD11b⁺Gr-1⁺ MDSCs were sorted by FACS and cocultured with Con A–prestimulated, CFSE-stained T cells from BALB/c mice at different ratios. Three days later, the status of T cell proliferation was analyzed by FACS. The results indicated that S. japonicum infection–induced MDSCs could inhibit the proliferation of both CD4⁺ and CD8⁺ T cells in a concentration-dependent manner. When the ratio of MDSCs/T cells was 1:2, MDSCs could inhibit CD4⁺ and CD8⁺ T cell proliferation significantly. When the ratio of MDSCs/T cells changed to 1:4, the suppressive role of MDSCs decreased, and the suppressive role of MDSCs disappeared at ratio of 1:8 (MDSCs/T cells) (Fig. 2A). No suppressive function was found in MDSCs from normal mice.

Furthermore, inflammatory cytokine production was detected in S. japonicum infection–induced MDSCs by intracellular cytokine staining. As shown in Fig. 2B, CD11b⁺Gr-1⁺ MDSCs were gated first, and the percentages of IL-1a–, IL-6–, IL-10–, and GM-CSF–secreting cells were investigated. The results indicated that the percentages of IL-1a, IL-6, IL-10 and GM-CSF cells in MDSCs from infected mouse spleen were higher than those from control mice (p < 0.05). Among these cytokines, the increase of IL-1a⁺ MDSCs was most significant (p < 0.01). However, the change in the percentage of GM-CSF–secreting cells between normal and infected mice was not significant (p > 0.05). Moreover, PD-L1/2 expression was detected in MDSCs by FACS as well. The results (Fig. 2B) showed that the difference in the percentage of PD-L1– or PD-L2–expressed MDSCs between infected and control mouse spleen was not obvious (p > 0.05).

To further explore the roles of SEA and SWA in S. japonicum infection–induced MDSCs and myeloid cell differentiation, BM cells from normal C57BL/6 mice were isolated and stimulated with GM-CSF and IL-6 for 6 d. SEA and SWA were added to the cells alone or together, and a negative control was used as described in 2Materials and Methods. Six days later, the percentages of immature DCs (CD11c⁺MHC class II⁺) and MDSCs (Gr-1⁺CD11b⁺) were detected by FACS. The results indicated that both SEA and SWA could significantly inhibit immature CD11c⁺ DCs and CD11c⁺MHC class II⁺ DC differentiation (p < 0.05), and the effects of SEA and SWA could be overlaid (Fig. 3A, 3B). The subtype of SEA- and SWA-induced MDSCs was further investigated by FACS. The results indicated that SEA and SWA could dramatically increase the percentage of CD11b⁺Ly6G⁺Ly6C^−/low G-MDSCs (p < 0.05), but not CD11⁺Ly6G⁻Ly6C^high M-MDSCs (p > 0.05, Fig. 3C).

Additionally, the function of SEA- and SWA-induced MDSCs was investigated. SEA-induced MDSCs were sorted by FACS and cocultured with Con A–prestimulated and CFSE-labeled allogeneic T cells (from BALB/c mice) at various ratios (1:8, 1:4, or 1:2). Three days later, T cell proliferation was assessed by FACS. The results indicated that SEA-induced MDSCs could inhibit the proliferation of both CD4⁺ and CD8⁺ T cells in a concentration-dependent manner (Fig. 3D).

Differences between SEA- or SWA-induced MDSCs and GM-CSF–induced MDSCs in the inhibition of T cell proliferation were explored as well. BM cells were isolated from normal C57BL/6 mice and stimulated with GM-CSF, SEA, or SWA. Six days later, the MDSCs were sorted from each treatment separately and cocultured with Con A–prestimulated, CFSE-labeled allogeneic T cells at different ratios. Three days later, a stronger suppressive effect on both CD4⁺ and CD8⁺ T cell proliferation was observed in SEA- and SWA-induced MDSCs by FACS (p < 0.05, Fig. 3D).

To investigate the mechanism of S. japonicum infection–induced MDSCs on the T cell response, MDSCs were sorted from the spleen of normal and S. japonicum–infected mice by FACS. The levels of ARG1, NO, and ROS in the MDSCs were detected as described in 2Materials and Methods. No obvious increase was found in the expression and activity of ARG1 in MDSCs from infected mice (Fig. 4A, 4B). Similar results were observed in the production of NO and expression of the NOS2 gene (p > 0.05, Fig. 4C, 4D). Interestingly, the expression levels of ROS in BM and spleen from infected mouse MDSCs increased dramatically (p < 0.05, Fig. 4E). Additionally, an in vitro experiment indicated that the expression of ROS in SEA- or SWA-induced MDSCs also increased significantly (p < 0.05, Fig. 4F).

The expression level of NOX2 in MDSCs from normal and infected mice was detected by quantitative RT-PCR (qRT-PCR). The results indicated that the RNA expression of two subunit types of NOX2 (gp91^phox and p47^phox) were increased in MDSCs from infected mice (p < 0.05). No significant differences were observed in the expression of other genes subunits (p > 0.05, Fig. 4G, Table I). The Western blotting results showed greater protein expression of gp91^phox and p47^phox in MDSCs from infected mice (Fig. 4H), as well as MDSCs induced by SEA or SWA in vitro (Fig. 4I). Furthermore, the effects of ROS-specific inhibitor NAC, arginase inhibitor nor-NOHA, l-arginine, and the NOS-specific inhibitor L-NMMA on MDSCs were investigated. Inhibitors were added to the cocultured mouse splenic MDSCs and CFSE-labeled allogeneic T cells as previously described at a ratio of 1:2, and an unstimulated T cell was used as control. Three days later, the FACS results showed that the ROS inhibitor NAC could obviously mitigate the suppressive effect of the MDSCs on both CD4⁺ and CD8⁺ T cell proliferation (p < 0.05). However, the suppressive effects of the MDSCs were not affected by nor-NOHA, l-arginine, or L-NMMA (p > 0.05, Fig. 4J).

To further explore the mechanism of the increased MDSCs in S. japonicum–infected mouse immune organs, apoptosis of S. japonicum infection–induced MDSCs in the spleen were investigated by FACS. As shown in Fig. 5A, no significant decrease was detected on the expression of annexin V in MDSCs from infected mouse spleen compared with MDSCs from normal mice (p > 0.05). Furthermore, the expressions of MDSC essential genes S100A8/9, cyclin D1, c-Myc, CDX2, Bcl-x_L, and BAX were detected by quantitative RT-PCR (qRT-PCR), and the results (Fig. 5B) showed that the expression of the S100A8/9 gene was dramatically increased in MDSCs from infected mice compared with those from normal mice (p < 0.05). No significant differences were observed in the expression of other MDSC essential genes (p > 0.05). The Western blotting confirmed an increased expression of S100A8/A9 in protein level as well (Fig. 5C).

To check the possible role of the STAT3 pathway in the accumulation of S. japonicum infection–induced MDSCs, the phosphorylation level of STAT3 and STAT6 (the control) in sorted MDSCs from both infected and normal mouse spleen was analyzed. The results indicated a higher level of phosphorylation of STAT3 but not STAT6 in MDSCs from infected mouse spleen (p < 0.05, Fig. 5D). Additionally, overexpression of the S100A8/A9 gene and a higher level of S100A8/A9 protein were detected in SEA- or SWA-induced MDSCs in vitro (p < 0.05, Fig. 5F), which is consistent with the in vivo experiment.

To further explore the role of the STAT3 signaling pathway on the increase of the S100A8/A9 gene expression in infection-induced MDSCs, SEA was added with or without JSI-124 to a solution of isolated normal mouse BM cells. Forty-eight hours later, the expression levels of S100A8/A9 genes in sorted MDSCs were detected by RT-PCR. The results (Fig. 5G) indicated that the level of the S100A8/A9 gene was significantly increased in SEA stimulation–inducted MDSCs (p < 0.05). JSI-124 costimulation significantly suppressed the level of the S100A8/A9 gene in SEA-induced MDSCs.

The correlation between SEA-activated STAT3 signals and the effect of MDSCs was explored. SEA or JSI-124 was added to isolated normal mouse BM cell solutions alone or together. Six days later, the percentage of Gr-1⁺CD11b⁺ MDSCs was detected by FACS. The results (Fig. 6A) showed that SEA stimulation could increase the percentage of MDSCs in cultured BM cells (p < 0.05), and JSI-124 was an agonist for the effect of SEA in MDSC induction (p < 0.05).

The role of the STAT3 signaling pathway was further explored for the suppressive effects of SEA-induced MDSCs. In vitro–cultured MDSCs were isolated from SEA- or JSI-124–induced BM cells and cocultured with SCFE-labeled allogeneic T cells as previously described. Three days later, T cell proliferation was assessed by FACS. As shown in Fig. 6B, the proliferation of SEA-induced MDSCs was significantly inhibited by JSI-124 (p < 0.05). Consistent with this result, the expression of the gp91^phox and p47^phox genes was increased in the SEA-induced MDCSs (p < 0.05, Fig. 6C, 6D).

MDSCs are a new type of suppressor cell that can be induced in different types of parasite infections (12). In this study, the percentage and the absolute numbers of CD11b⁺Gr-1⁺ MDSCs in infected mouse organs was significantly increased (p < 0.05, Fig. 1C, 1D), which suggested that S. japonicum infection could induce MDSC accumulation in BM, spleen, and mesenteric lymph nodes of mice. Consistent with our results, accumulation of myeloid-derived suppressor cells was reported in the spleen and peripheral blood of S. japonicum–infected mice (26). To further define this population of cells, mature granulocyte and myeloid population–associated molecules (9, 27), including Ly6G, Ly6C, F4/80, CD49d, CD115, CD117 (c-Kit), and CD135 (Flt-3), were detected in CD11b⁺Gr-1⁺ MDSCs by FACS (Fig. 1E). As shown in Fig. 1F, the expression of Ly6G, CD115, CD117, and CD135 increased significantly (p < 0.05). It implied that these induced MDSCs are mainly immature myeloid cells. Furthermore, the morphology of MDSCs was examined using Wright staining. As showed in Fig. 1G, many small mononuclear leukocyte and multinuclear cell were observed, indicating that MDSCs were mainly immature monocytes and neutrophils. Further experimental results indicated that the percentage of CD11b⁺Ly6G⁺Ly6C^−/low G-MDSCs, but not CD11⁺Ly6G⁻Ly6C^high M-MDSCs, was significantly increased (Fig. 1H, p < 0.05).

It is well known that MDSCs are a population of heterogeneous immature myeloid cells, and one of their principal functions is to inhibit T cell functions (28). Clearly, our study showed that S. japonicum infection–induced MDSCs inhibited the proliferation of both CD4⁺ T cells and CD8⁺ T cells in a concentration-dependent manner (Fig. 2A). Moreover, no suppressive function was found for MDSCs from normal mice. This result is consistent with a report by Corzo et al. (29) and suggested that Gr-1⁺CD11b⁺ cells from naive mice do not have a suppressive function and are considered to be immature myeloid cells.

It has been reported that MDSCs could secrete the inhibitory cytokines TGF-β and IL-10 (11). To check the inflammatory cytokine production in S. japonicum infection–induced MDSCs, we compared the percentage of IL-1a, IL-6, IL-10, and GM-CSF cells in MDSCs from uninfected and infected mouse spleen. Significant increases of these cytokines were detected in infected mice (p < 0.05) (Fig. 2A). Among these cytokines, the increase of IL-1a⁺ MDSCs was the most significant (p < 0.01), which suggested that S. japonicum infection–induced MDSCs might have an inhibitory function via the secretion of IL-10.

PD-L1 or PD-L2 was expressed in myeloid cells, including MDSCs, which induce T cell apoptosis (15, 16). In this study, we showed no differences in the percentage of PD-L1– or PD-L2–expressed MDSCs between infected and control mouse spleen (p > 0.05) (Fig. 2B), which indicated that the overexpression of PD-L1 or PD-L2 was not involved in the function of S. japonicum infection–induced MDSCs.

SEA and SWA were the primary soluble proteins correlated with the S. japonicum infection–induced adaptive immune response. Recently, it has been reported that SEA and SWA could alter macrophage polarization during a S. japonicum infection in mice (30). We, therefore, explored the roles of SEA and SWA in S. japonicum infection–induced MDSCs. As shown in Fig. 3A and 3B, SEA and SWA could increase MDSC accumulation by inhibiting myeloid cell differentiation in vitro. The subtypes of the SEA- and SWA-induced MDSCs were CD11b⁺Ly6G⁺Ly6C^−/low G-MDSCs (p < 0.05), but not CD11⁺Ly6G⁻Ly6C^high M-MDSCs (Fig. 3C). Additionally, these SEA- and SWA-induced MDSCs could inhibit the proliferation of both CD4⁺ and CD8⁺ T cells in a concentration-dependent manner (Fig. 3D). It suggested that SEA and SEA could promote the production of MDSCs and also improve their immunosuppressive function.

It has been reported that MDSCs inhibited the function of the T cells response mainly via an alteration in the metabolic pathway of l-arginine when cancer or other pathologic conditions were present. The underlying mechanism might involve the increase in the expression of ARG1, induction of the activity of iNOS, as well as the production of ROS (17, 18). It has been reported that SEA could upregulate the expression of ARG1 in isolated CD11b⁺ myeloid cells from S. japonicum–infected mouse spleen in vitro (31). In our study, the level of l-arginine metabolic products (ARG1, NO, and ROS) in spleen MDSCs was compared between normal and S. japonicum–infected mice. No significant increase of ARG1 expression was detected in CD11b⁺Gr-1⁺MDSCs (p > 0.05, Fig. 4B), but increased ARG1 expression in CD11b⁺ myeloid cells and CD11b⁺Gr-1⁻ cells was detected (Supplemental Fig. 1). The expression levels of ROS in MDSCs from infected mouse BM and spleen were dramatically increased (p < 0.05, Fig. 4E). Additionally, an in vitro experiment indicated that the expression of ROS in SEA- or SWA-induced MDSCs was also significantly higher than that in the control (p < 0.05, Fig. 4F), which indicated that the SEA- or SWA-induced higher ROS expression might be related to S. japonicum infection–induced MDSC differentiation. In the S. mansoni model, Pearce and colleagues (32) have shown the induction of ROS using the CM-H2DCFDA reagent, which is consistent with our work.

As reported, ROS production in MDSCs was mainly regulated by NADPH oxidase (NOX2) (19), which is constructed of two membrane proteins (gp91^phox and p22^phox) and four cytoplasmic components (p67^phox, p40^phox, p47^phox, and rac1, a type of small G protein). We therefore performed qRT-PCR to determine the expression levels of NOX2 in MDSCs from normal and infected mice. Our results (Fig. 4G) indicated that the expression of gp91^phox and p47^phox, but not other subunits, was significantly higher in RNA from infected mouse MDSCs (p < 0.05). The Western blotting results confirmed the greater protein expression of 91^phox and p47^phox in MDSCs from infected mice (Fig. 4H) and SEA- or SWA-induced MDSCs (Fig. 4I). Furthermore, the ROS-specific inhibitor NAC, arginase inhibitor nor-NOHA, l-arginine, and the NOS-specific inhibitor L-NMMA were added to the cocultured infected mouse splenic MDSCs and allogeneic T cells as previously described. The FACS results (Fig. 4F) showed that the ROS inhibitor NAC could obviously mitigate the suppressive effect of MDSCs on both CD4⁺ and CD8⁺ T cell proliferation (p < 0.05). In contrast, the suppressive effects of MDSCs were unchanged by nor-NOHA, l-arginine, or L-NMMA (p > 0.05). These results indicated that an S. japonicum infection could induce the expression of the NOX2 subunits gp91^phox and p47^phox, which caused an increase of ROS in MDSCs and mediated the suppressive effects of MDSCs.

Previous study showed that the survival conditions of MDSCs, or the differentiation of myeloid cells, could affect the accumulation of MDSCs as well (33). To evaluate possible effects of these factors on MDSCs, apoptosis of S. japonicum infection–induced MDSCs was investigated. As shown in Fig. 5A, no significant decrease was detected on the expression of annexin V in MDSCs from infected mouse spleen compared with MDSCs from normal mouse (p < 0.05), which suggested that MDSC apoptosis was not responsible for the increase of MDSCs in S. japonicum–infected mouse spleen.

S100A8/9 is an important inflammation factor that plays a key role in MDSC expansion (20). Higher MDSC accumulation and tumor growth could be stopped in S100A9 gene knockout mice (21). Deregulation of the apoptotic factors Bcl-x_L and Bax confers apoptotic resistance to MDSCs and contributes to their persistence in cancer (34). Cox2, an intestine-specific transcription factor, is expressed in Barrett’s esophagus and is involved in this process as well (35). In this study, MDSC-related genes S100A8/9, cyclin D1, c-Myc, COX2, Bcl-x_L, and BAX were detected by qRT-PCR. The results showed that only the expression of the S100A8/9 gene increased dramatically in MDSCs from infected mice (Fig. 5B). The Western blotting results confirmed the increasing S100A8/A9 protein expression as well (Fig. 5C). Taken together, these results suggested that S. japonicum infection could induce MDSC accumulation through overexpression of intracellular S100A8/A9.

Recent research has found that S100A8 and S100A9 are important target genes in the STAT3 signaling pathway (20). STAT3 could enhance MDSC proliferation through induction of S100A8 and S100A9 expression. In this study, we found a higher phosphorylation level of STAT3 but not of STAT6 in MDSCs from infected mouse spleen (Fig. 5D). Additionally, overexpression of the S100A8/A9 gene and a higher level of S100A8/A9 protein were also detected in SEA- or SWA-induced MDSCs in vitro (p < 0.05, Fig. 5F). These results help to clarify that the STAT3 pathway is involved in S. japonicum infection–induced expansion of MDSCs.

Cucurbitacin I hydrate (JSI-124) is an inhibitor of the JAK/STAT3 signaling pathway, which could block STAT3 phosphorylation and activation and inhibit the activation of JAK kinase (22). To further explore the role of the STAT3 signaling pathway and S100A8/A9 gene expression in infection-induced MDSCs, SEA was added to an isolated normal mouse BM cell solution with or without JSI-124. The RT-PCR results (Fig. 5G) indicated that JSI-124 costimulation could significantly suppress the level of the S100A8/A9 gene in SEA-induced MDSCs, which suggested that SEA could increase the expression of the S100A8/A9 gene via the STAT3 signaling pathway during MDSCs induction.

Next, the correlation between SEA-activated STAT3 signals and the effect of MDSCs was explored. Our results indicated that JSI-124 was an agonist for the effect of SEA in inducing MDSCs (Fig. 6A, p < 0.05) as well as the function of MDSCs (Fig. 6B). Consistently, the expression of the gp91^phox and p47^phox genes was higher in SEA-induced MDCSs (p < 0.05, Fig. 6C, 6D). These results suggested that the STAT3 signal was involved in the effect of SEA on MDSCs.

In summary, SEA and SWA were confirmed to enhance the accumulation of MDSCs in S. japonicum–infected mouse spleen by inducing the NOX subunits gp91^phox and p47^phox, and MDSC induction was dependent on the ROS production via JAK/STAT3 signaling.

The authors have no financial conflicts of interest.