Polyunsaturated fatty acids promote the expansion of myeloid-derived suppressor cells by activating the JAK/STAT3 pathway

Authors

  • Dehong Yan,

    1. Institute of Human Virology, Zhongshan School of Medicine, Sun Yat-Sen University, Guangzhou, China
    Search for more papers by this author
  • Quan Yang,

    1. Institute of Human Virology, Zhongshan School of Medicine, Sun Yat-Sen University, Guangzhou, China
    Search for more papers by this author
  • Maohua Shi,

    1. Institute of Human Virology, Zhongshan School of Medicine, Sun Yat-Sen University, Guangzhou, China
    Search for more papers by this author
  • Limei Zhong,

    1. Institute of Human Virology, Zhongshan School of Medicine, Sun Yat-Sen University, Guangzhou, China
    Search for more papers by this author
  • Changyou Wu,

    1. Institute of Immunology, Zhongshan School of Medicine, Sun Yat-Sen University, Guangzhou, China
    Search for more papers by this author
  • Tao Meng,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Huiyong Yin,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Jie Zhou

    Corresponding author
    1. Institute of Human Virology, Zhongshan School of Medicine, Sun Yat-Sen University, Guangzhou, China
    2. Key Laboratory of Tropical Disease Control, Sun Yat-sen University, Chinese Ministry of Education, Guangzhou, China
    • Full correspondence: Prof. Jie Zhou, Institute of Human Virology, Zhongshan School of Medicine, Sun Yat-Sen University, 74 Zhongshan 2nd Road, Guangzhou 510080, China

      Fax: +86-20-8733-2588

      e-mail: zhouj72@mail.sysu.edu.cn

    Search for more papers by this author

Abstract

Polyunsaturated fatty acids (PUFAs) exert immunosuppressive effects that could prove beneficial in clinical therapies for certain autoimmune and inflammatory disorders. However, the mechanism of PUFA-mediated immunosuppression is far from understood. Here, we provide evidence that PUFAs enhance the accumulation of myeloid-derived suppressor cells (MDSCs), a negative immune regulator. PUFA-induced MDSCs have a more potent suppressive effect on T-cell responses than do control MDSCs. These observations were found both in cultured mouse bone marrow cells in vitro and in vivo in mice fed diets enriched in PUFAs. The enhanced suppressive activity of MDSCs by PUFAs administration was coupled with a dramatic induction of nicotinamide adenine dinucleo- tide phosphate oxidase subunit p47phox and was dependent on reactive oxygen species (ROS) production. Mechanistic studies revealed that PUFAs mediate its effects through JAK-STAT3 signaling. Inhibition of STAT3 phosphorylation by JAK inhibitor JSI-124 almost completely abrogated the effects of PUFAs on MDSCs. Moreover, the effects of PUFAs on MDSCs and the underlying mechanisms were confirmed in tumor-bearing mice. In summary, this study sheds new light on the immune modulatory role of PUFAs, and demonstrates that MDSCs expansion may mediate the effects of PUFAs on the immune system.

Introduction

Polyunsaturated fatty acids (PUFAs) are fatty acids that have more than two carbon–carbon double bonds. PUFAs are classified as omega-3 (n-3) and omega-6 (n-6) fatty acids based on the location of the last carbon–carbon double bond from the omega end of the molecule [1]. They are consumed in the diet and can be metabolized by sequential reactions, to several biologically active mediators, such as eicosanoids [2]. Studies have demonstrated that PUFAs have diverse roles in lipid metabolism, inflammation, and immune responses, in addition to their essential functions in maintaining cell structure [3].

PUFAs have immune suppressive capabilities that provide clinical beneficial effects in some autoimmune and inflammatory diseases [4-6]. The mechanism of PUFA-mediated immune suppression has been studied extensively, but is not fully understood. Administration of n-3 PUFAs suppresses natural killer cell activity in healthy volunteers [7]. Dietary intake of n-3 PUFAs decreased cytokine production in murine peritoneal macrophages and human peripheral blood monocytes [8, 9]. PUFAs modulate T-cell signaling by altering plasma membrane micro-organization (lipid rafts) at the synapse formed between T cells and APCs [10, 11]. PUFAs also shift the differentiation of naive T cells from T-helper 1 (Th1) to T-helper 2 (Th2) [12]. Treatment of murine bone marrow-derived DCs with PUFAs attenuated LPS-induced DC maturation [13, 14]. These studies support the importance of PUFAs in the immune system.

Myeloid-derived suppressor cells (MDSCs) are an abnormal accumulation of different stages of myeloid progenitor cells and immature myeloid cells under some pathological conditions, including but not limited to cancer, inflammatory disorders, trauma, and infections [15, 16]. MDSCs are generated in bone marrow, and they migrate to the blood, peripheral lymphoid tissues, and other organs [16]. Studies have demonstrated that MDSCs possess broad and potent immune suppressive capacity by inhibiting other immune cells in a context-dependent manner [17-19]. MDSCs are identified as cells that coexpress the Gr1 and CD11b antigens in mice [20]. Human MDSCs are less well characterized because no uniform markers are available to date [15]. MDSCs consist of two major subsets: monocytic MDSCs (M-MDSCs) with a CD11b+Ly6GLy6Chigh phenotype, and granulocytic MDSCs (G-MDSCs) with a CD11b+Ly6G+Ly6C−/low phenotype [20, 21]. These two subsets differ in antigen specificity, tissue localization, and mechanisms for immune suppression [20, 22]. Studies about cancer and MDSCs have shown that M-MDSCs suppress T-cell responses in an antigen nonspecific manner, mostly via increased arginase activity, while G-MDSCs utilize reactive oxygen species (ROS) as immune mediators and suppress T cells in an antigen-specific manner [20, 23, 24]. Mechanistic studies have demonstrated that tumor-derived factors or inflammatory cytokines could activate some transcription factors, which promote the transcription of genes essential for MDSCs expansion or activation under pathological conditions [25, 26]. However, little is known about whether certain dietary components affect MDSCs biology.

Here, we provide evidence that the administration of PUFAs, a dietary component, strikingly enhances both the expansion and suppressive activity of mouse G-MDSCs in vitro and in vivo. The underlying mechanism of PUFA's effects involved the activation of STAT3 signaling and the transcription of its target genes. This study reveals a novel mechanism for immuno-suppression by PUFAs in cancer and certain inflammatory-related disorders.

Results

PUFAs enhance MDSCs expansion in vitro

Previous studies have demonstrated that culturing mouse bone marrow cells with different cytokines could generate MDSCs in vitro [27]. To evaluate the effect of PUFAs on MDSCs, mouse bone marrow (BM) cells were cultured in medium containing GM-CSF with or without IL-6 to generate MDSCs in vitro; α-linolenic acid (ALA) (n-3 PUFA) or LA (n-6 PUFA) was added at culture initiation, ethanol was used as a control. Cells were harvested for flow cytometric analysis after 6 days of culture. The data show that PUFAs dramatically increased the proportion of Gr-1+CD11b+ MDSCs, in the presence or absence of IL-6 (Fig. 1A and Supporting Information). The total number of cultured cells, however, did not change (data not shown). Based on the fact that MDSCs consist of granulocytic (G-MDSCs) and monocytic (M-MDSCs) populations, which may have different biological functions and utilize distinct mechanisms for immune suppression [21, 22], we therefore investigated which subtype was affected by exposure to PUFAs. We found that PUFAs consistently elevated the population of CD11b+Ly6G+Ly6C−/low G-MDSCs, but generated no noticeable change in CD11b+Ly6GLy6Chigh M-MDSCs (Fig. 1A). Together with the increased MDSCs proportion after PUFAs treatment, we observed a dramatic reduction in the percentage of CD11c+MHCII+ DCs (Fig. 1A), indicative of impaired myeloid cell differentiation.

Figure 1.

PUFAs enhanced the expansion and function of MDSCs in vitro. (A) Mouse BM cells were cultured in GM-CSF (20 ng/mL) for 6 days in the presence of 100 μM ALA or LA, ethanol was used as control. The proportions of MDSCs (Gr-1+CD11b+), MDSC subsets, and DCs (CD11c+MHCII+) were evaluated by flow cytometry analysis. Analysis from a single experiment (left), as well as mean +SEM of three independent experiments (right) are shown. (B and C) Allogeneic mixed lymphocytes reaction. Mouse (C57BL/6) BM cells were cultured in GM-CSF and IL-6 (20 ng/mL) for 6 days with PUFAs or ethanol control. MDSCs were purified by flow cytometric sorting. Allogeneic CD3+ T cells (from BALB/c mice) were stimulated with ConA, then cocultured with isolated MDSCs at different ratios for 3 days. T-cell proliferation was evaluated by CFSE dilution. Unstimulated T cells were used as a negative control. (B) The suppressive function of LA-derived and control MDSCs was measured. Representative flow cytometry data from single experiment (left) and mean +SEM from three independent experiments (right) are shown. (C) The comparison of suppressive activity on CD4+ (left) and CD8+ T cells between PUFA-derived MDSCs and control MDSCs are shown. Data are shown as mean +SEM of nine samples from three independent experiments. *p < 0.05, **p < 0.01 compared with the corresponding controls, unpaired t tests were used.

Our data suggest that PUFAs specifically causes the expansion of G-MDSCs. Considering that MDSCs suppress T-cell immune responses under some pathological conditions [17, 24], we next determined whether PUFA-induced MDSCs are functionally suppressive toward T cells. BM cells were cultured in GM-CSF alone or GM-CSF and IL-6 for 6 days, in the presence or absence of PUFAs. MDSCs were purified by flow cytometric sorting and cocultured with allogeneic T cells stimulated with concanavalin A (ConA) at different ratios, T-cell proliferation was evaluated by 5,6 carboxyfluo-rescein diacetate, succinimidyl ester (CFSE) dilution. Consistent with previous reports [27], the MDSCs generated by GM-CSF alone displayed much lower suppressive activity than the combination of GM-CSF and IL-6, while we found that PUFA-derived MDSCs suppressed T-cell responses in a dose-dependent manner under both culture conditions (Fig. 1B and Supporting Information). We further determined whether there was a difference in suppressive activity between PUFA-MDSCs and control-MDSCs. We found that the capability of PUFA-derived MDSCs was much stronger than control MDSCs in suppressing T-cell proliferation, when cultured with or without IL-6 (Fig. 1C and Supporting Information). Our data indicate that PUFAs also enhances the functionality of MDSCs.

PUFAs enhance MDSCs expansion in vivo

We next investigated the effects of PUFAs on MDSCs in vivo. C57BL/6 mice were fed a linseed oil diet (LOD) containing 45% ALA, or a sunflower oil diet (SOD) containing 45% linoleic acid (LA) for three consecutive months; a regular diet (RD) served as the control. After three months, the percentage of Gr-1+CD11b+ cells in BM, spleen, and peripheral blood was analyzed by flow cytometry. As expected, we observed consistently higher Gr-1+CD11b+ levels in these tissues from mice fed PUFA-enriched diets, than from mice fed control diet (Fig. 2A). Further analysis showed that CD11b+Ly6G+Ly6C−/low cells were significantly elevated in the BM and spleen of LOD- or SOD-fed mice, compared with that from RD-fed mice, while no differences in the frequency of CD11b+Ly6GLy6Chigh were found (Fig. 2B). These observations were consistent with the increased G-MDSCs (CD11b+Ly6G+Ly6C−/low) level by PUFAs administration in vitro. Since Gr-1+CD11b+ cells from naive mice do not possess suppressive function and are considered to be immature myeloid cells [28]. In order to determine whether the comparable proportions of Gr-1+CD11b+ cells from PUFA-fed mice were MDSCs, we did functional analysis and found that Gr-1+CD11b+ cells from the spleens of PUFA-fed mice were immune suppressive towards T cells, whereas MDSCs from RD-fed mice were nonsuppressive (Fig. 2C). These observations demonstrated that PUFAs enhance the suppressive function of MDSCs.

Figure 2.

PUFAs promote MDSCs accumulation in vivo. (A, B) C57BL/6 mice were fed with a RD, a diet enriched with ALA (LOD, linseed oil diet), or a diet enriched with LA (SOD, sunflower oil diet) for 3 months. Mice were sacrificed and different tissues were harvested. (A) The percentage of Gr-1+CD11b+ cells was evaluated by flow cytometry after staining with specific antibodies. Each group included 20 mice. (B) Proportions of MDSC subtypes in the BM and spleens were evaluated by flow cytometry. Each group included five mice. (A, B) Analysis from a single mouse (left) as well as the mean +SEM from all mice analyzed (right) are shown. (C) Allogeneic mixed lymphocytes reaction. CD3+ T cells (from BALB/c mice) were stimulated with ConA and cocultured with allogeneic MDSCs isolated from mice (C57BL/6) fed with different diets at different ratios for 3 days. T-cell proliferation was evaluated by CFSE dilution. Unstimulated T cells were used as a negative control. Three independent experiments with similar results were performed and mean +SEM of six samples pooled from three experiments is shown. *p < 0.05, **p < 0.01, compared with the controls, unpaired t tests were used.

PUFA-derived MDSCs suppressed T-cell responses in ROS-dependent manner

Based on the observation that MDSCs generated by PUFAs treatment suppressed T-cell responses, we further explored the underlying mechanisms controlling this suppression. It is well known that l-arginine and its metabolic products are essential for MDSC-mediated immune suppression under some pathological conditions, especially in cancer [24, 29]. We therefore compared the levels of the l-arginine metabolic products: arginase, nitric oxide (NO), and ROS, between PUFA-MDSCs and control-MDSCs. Mouse BM cells were cultured in GM-CSF to induce MDSCs generation in the presence or absence of PUFAs. Cells were harvested 6 days later and Gr-1+CD11b+ cells were purified by flow cytometric sorting. Biochemical assays found no considerable changes in the level of arginase activity (Fig. 3A) or NO content (Fig. 3B) in MDSCs between PUFAs and control groups, while ROS production yielded a three- to fourfold increase in PUFA-MDSCs, than in control-MDSCs (Fig. 3C). The elevation of ROS in PUFA-MDSCs was also observed in MDSCs from the BM and spleens of mice fed PUFA-enriched diets, when compared with that in RD-fed mice (Fig. 3D). We next examined the expression of enzymes responsible for l-arginine metabolism. Arginase-1 and iNOS, responsible for arginase activity and NO production, did not show any changes in mRNA and protein expression (Fig. 3E). We then evaluated the expression of all components of the nicotinamide adenine dinucleotide phosphate-oxidase complex that is responsible for ROS production [29]. Interestingly, only p47phox expression was increased in PUFA-derived MDSCs; other components of nicotinamide adenine dinucleo-tide phosphate oxidase complex, however, did not show any change between different groups (Fig. 3E). The elevation of p47phox protein was also observed in Gr-1+CD11b+ cells from the BM and spleens of mice fed PUFA-enriched diets, in comparison with that in RD-fed mice (Fig. 3F). In support of the essential role of ROS in PUFA-MDSC mediated immune suppression, addition of ROS inhibitor NAC into the MDSC-T-cell coculture system almost completely abrogated the suppressive effect of MDSCs on T-cell proliferation in CD4+ and CD8+ T cells, while the inhibitors for arginase or iNOS, or l-arginine supplementation had no effect (Fig. 3G). In order to eliminate the potential direct effect of these inhibitors on T cells, we performed experiments to test this possibility. Results showed that no considerable impacts were observed for T cells after administration of these inhibitors, supporting that the recovery of T-cell proliferation by ROS inhibitor was mediated by MDSCs (Supporting Information). We therefore conclude that PUFA-MDSCs suppress T cells in a ROS-dependent manner.

Figure 3.

PUFA-derived MDSCs suppress T-cell responses in a ROS-dependent manner. (A–C) Measurements of l-arginine metabolism in MDSCs. Mouse BM cells were cultured with GM-CSF (20 ng/mL) for 6 days in the presence of 100 μM ALA or LA; ethanol was used as a control. MDSCs isolated by flow cytometric sorting were subjected to biochemical assays, including (A) arginase activity and (B) NO content, as described in the Materials and methods. Mean +SEM of nine samples pooled from three independent experiments is shown. (C) ROS production in MDSCs was measured by flow cytometry. Gr-1+CD11b+ cells were gated and the percentage of CM-H2DCFDA+ cells are shown as mean +SEM of six samples pooled from three independent experiments. (D) Mice were fed diets enriched with PUFAs (LOD or SOD) or control diets (RD) for 3 months. Cells were harvested from BM and spleen and the ROS level was evaluated by flow cytometry. (C, D) Typical results from a single experiment are shown (left). Mean ± SEM of three independent experiments is shown (right). (E) MDSCs were isolated from 6-day cultured BM cells with or without PUFAs treatment, the expression of l-arginine metabolizing enzymes was measured by qRT-PCR (left) and Western blotting (right). (F) The expression of l-arginine metabolizing enzymes in MDSCs purified from mouse BM and spleen as described in (D) was measured by Western blotting and quantified and shown as mean +SEM of eight samples pooled from four experiments. (E, F) Data shown are representative of three independent experiments. β-Actin was used as a loading control. (G) Effect of different inhibitors on the function of PUFA-derived MDSCs was evaluated by allogeneic mixed lymphocytes reaction. CD3+ T cells from BALB/c mice were stimulated with ConA and cocultured with allogeneic MDSCs isolated from the cultured BM cells (C57BL/6 mice) as described in (A) at 2:1 ratio for 3 days with treatments as indicated. T-cell proliferation was evaluated by CFSE dilution. Unstimulated T cells were used as a negative control. nor-NOHA (100 μM): arginase inhibitor; l-arginine (1 mM); L-NMMA (100 μM): iNOS inhibitor; NAC (1 mM): ROS inhibitor. Means +SEM of six samples pooled from three independent experiments are shown. *p < 0.05, **p < 0.01, unpaired t test.

Activation of JAK-STAT3 signaling by PUFAs

STAT signaling is important for the expansion and activation of MDSCs under pathological conditions [25, 26]. To investigate whether STAT signaling mediates PUFA-induced MDSCs expansion, mouse BM cells were cultured in medium containing GM-CSF for 6 days in the presence or absence of PUFAs. The phosphorylation of STAT3 and STAT6 in purified MDSCs was determined by western blotting. Data showed that PUFAs administration specifically enhanced STAT3 phosphorylation, but had no effect on STAT6 signaling. The total STAT3 protein level did not change after PUFAs treatment (Fig. 4A). S100A8 and S100A9, important inflammatory molecules, are known transcriptional targets of STAT3 signaling and play a crucial role for MDSCs expansion [30]. We therefore examined the effect of PUFAs on the expression of S100A8/A9 in MDSCs. We show that PUFAs treatment dramatically induced the expression of S100A8/A9 at both the mRNA and protein levels (Fig. 4B). Consistent with the in vitro data, phosphorylation of STAT3 (Fig. 4C, Supporting Information) and the expression of S100A8/A9 (Fig. 4D, Supporting Information) were also clearly increased in Gr-1+CD11b+ cells isolated from BM and spleens of mice fed with PUFA-enriched diets. Cucurbitacin-I (JSI-124), a potent inhibitor of JAK-STAT3 signaling [31], was used to confirm that the upregulation of S100A8/A9 by PUFAs was mediated by STAT3 signaling. BM cells were cultured in GM-CSF and treated with LA, JSI-124, or both for 48 h, the expression of S100A8/A9 was evaluated by qRT-PCR. Results showed that JSI-124 remarkably overrode the effect of PUFAs on S100A8/A9 expression (Fig. 4E). An S100A8 reporter assay yielded the similar results (Fig. 4F). These observations suggest PUFAs upregulate the expression of S100A8/A9 by activating STAT3 signaling.

Figure 4.

Activation of JAK-STAT3 signaling by PUFAs. (A, B) Mouse BM cells were cultured in medium containing GM-CSF with ALA, LA, or ethanol control. Gr-1+CD11b+ cells were purified by flow cytometric sorting after 6 days. (A) The phosphorylation of STAT proteins was examined by Western blotting. (B) Expression of S100A8 and S100A9 was evaluated by qRT-PCR (left) and Western blotting (right). (C, D) Mice were fed diets enriched with LA-enriched diet (SOD) or control diets (RD) for 3 months. MDSCs were purified from BM and spleen. (C) Phosphorylation of STAT proteins was evaluated by Western blotting. (D) Expression of S100A8 and S100A9 was evaluated by qRT-PCR (left) and Western blotting (right). (E) Effect of STAT3 inhibitor JSI-124 on PUFAs induced S100A8/A9 expression. BM cells were cultured in GM-CSF with the indicated treatments for 24 h, mRNA expression of S100A8/A9 was determined by qRT-PCR. (F) HepG2 cells were transfected with S100A8 reporter plasmid, followed by culture in the presence of LA and/or JSI-124. Luciferase activity was measured at 48 h posttransfection. (A–F) Results represent three independent experiments, graphs show mean +SEM of nine pooled samples. *p < 0.05, **p < 0.01, unpaired t test was used.

The effects of PUFAs on MDSCs are dependent on JAK/STAT3 signaling

We next investigated whether the activation of STAT3 by PUFAs has any association with its effects on MDSCs. Mouse BM cells were cultured in medium containing GM-CSF with or without LA or JSI-124 treatment. The percentage of MDSCs was evaluated by flow cytometry after 6-day culture. As expected, LA treatment significantly increased the proportion of MDSCs, while cotreatment with JSI-124 abrogated the effect of LA (Fig. 5A). We next determined whether STAT3 influenced the suppressive activity of PUFA-MDSCs. Gr-1+CD11b+ cells were purified from 6-day cultured BM cells with or without LA or JSI-124 treatment, followed by coculture with allogeneic T cells for 3 days. T-cell proliferation was evaluated by CFSE dilution. The suppressive activity of MDSCs generated by PUFAs was clearly reduced by JSI-124 cotreatment, which was reflected by the higher T-cell proliferation in JSI-124 and LA cotreatment group, when compared with LA alone (Fig. 5B). We wanted to point out that the enhanced T-cell proliferation was not caused by the direct effect of JSI-124 on T cells [32], since we omitted JSI-124 from the MDSC/T cell coculture experiments. Consistent with the reduction in MDSCs functionality, the expression of p47phox, as well as ROS production, was also remarkably reduced by JSI-124 cotreatment, in comparison with LA alone (Fig. 5C, Fig. 5D). These observations indicate that STAT3 signaling could be essential for the suppressive activity of PUFA-induced MDSCs.

Figure 5.

PUFAs’ effect on MDSCs is mediated through JAK-STAT3 signaling. (A–D) Mouse BM cells were cultured in medium containing GM-CSF with the indicated treatments. Cells were harvested after 6-day culture. (A) The percentage of MDSCs was evaluated by flow cytometry. (B) Allogeneic mixed lymphocytes reaction. CD3+ T cells from mouse spleen were stimulated with ConA and cocultured with allogeneic MDSCs described in (A) at 2:1 for 3 days. T-cell proliferation was evaluated by CFSE dilution. Unstimulated T cells were used as a negative control. (C) Expression of p47phox in MDSCs isolated from cells in (A) was determined by qRT-PCR. (D) ROS levels in MDSCs from different groups were determined by CM-H2DCFDA labeling. (A, D) Both typical results from a single experiment (left), and cumulative results of six samples pooled from three independent experiments (right) are shown. (B, C) Data are shown as mean +SEM from three independent experiments. **p < 0.01, unpaired t test was used.

Pathological significance of PUFA-mediated MDSCs expansion in cancer

In order to know if the PUFA-mediated expansion of MDSCs has any pathological significance in cancer, we established tumor-bearing mice models in naive mice fed with diets enriched with PUFAs, or RDs. Mice were initially fed with different diets for three months, and then were subjected to subcutaneous injection of tumor cells, including CT26 and Lewis lung carcinoma (LLC) cell lines. Tumor size was monitored at different days and mice were sacrificed 24 days postinjection. Results showed that tumor growth in mice fed with PUFA-enriched diets (LOD or SOD) was much faster than mice fed with RD (Fig. 6A). Flow cytometry analysis showed that the percentage of MDSCs in different organs of PUFA-fed mice was significantly higher than in RD-fed mice (Fig. 6B). These observations were found in both tumor models established by CT26 and LLC cell lines, indicating a general effect of PUFAs on MDSCs expansion in tumor-bearing hosts. Further functional analysis confirmed that MDSCs generated by PUFAs administration displayed stronger capacity to suppress T-cell responses (Fig. 6C), as well as higher amount of ROS (Fig. 6D), when compared with control group. The phosphorylation of STAT3 and the transcription of its target genes were also induced in MDSCs from PUFA-fed mice than that from RD-fed mice (Fig. 6E and F). These results are consistent with the observations from naive mice. Our observations support that PUFA's effects on MDSCs have pathological significance in cancer.

Figure 6.

Pathological significance of PUFA-mediated MDSCs expansion in cancer. (A–F) Mice were initially fed a RD, a diet enriched with ALA (LOD), or a diet enriched with LA (SOD) for 3 months, then were subcutaneously injected with tumor cells, including CT26 (for BALB/c mice) or LLC (for C57B/L6 mice) cell lines; each group included six mice. (A) Tumor growth was monitored till 24 days postinjection, when the mice were sacrificed. (B) The percentages of MDSCs (Gr-1+CD11b+ cells) in different tissues were evaluated by flow cytometry analysis. Each symbol represents an individual mouse and the bar represents the mean. (C) Allogeneic mixed lymphocytes reaction. CD3+ T cells from mouse spleen were stimulated with ConA and cocultured with allogeneic MDSCs isolated from the spleens of tumor-bearing mice at 2:1 for 3 days. T-cell proliferation was evaluated by CFSE dilution. Unstimulated T cells were used as a negative control. (D) ROS levels in MDSCs from mouse BM and spleen were measured by flow cytometry. (E) Phosphorylation of STAT3 protein in purified MDSCs from mouse spleen was determined by Western blotting. (F) Gene expression at the mRNA level in MDSCs isolated from mouse spleen was determined by qRT-PCR. (C–F) Data are shown as the mean +SEM of six samples pooled from three independent experiments. (B–F) *p < 0.05, **p < 0.01; unpaired t test was used.

Discussion

Experimental and clinical studies have demonstrated that PUFAs have immune suppressive capabilities that contribute to their therapeutic effects in certain inflammatory and autoimmune diseases [3, 4, 6]. The mechanisms of PUFAs’ immunomodulation remain to be elucidated. Here, we provide evidence that PUFAs promote the accumulation and functional activity of MDSCs, one of the key brakes in the immune system. The enhanced generation of MDSCs may present a novel mechanism for the suppressive effects of PUFAs on immune responses.

MDSCs are generated by defective myeloid cell differentiation under pathological conditions such as cancer and inflammatory disorders [15, 19]. Studies have demonstrated that tumor-derived cytokines or certain inflammatory factors are contributors that drive MDSCs accumulation and activation [25, 26]. As far as we know, little is known about whether some dietary components could affect MDSCs. Our studies suggest that PUFAs administration could dramatically enhance MDSCs generation both in vitro and in vivo. The increased MDSCs generation by PUFAs was coupled with a clear reduction in DC differentiation, as evidenced by the lower proportion of immature DCs in the culture system. We therefore speculate that PUFAs impair myeloid cell differentiation, thus causing MDSCs accumulation. Although PUFAs are known to inhibit DC maturation [13, 14], little is known about the relationship between PUFAs and DC differentiation.

Previous studies have shown that several cytokines activate JAK-STAT3 signaling, which regulates the expression of genes related to MDSCs expansion, such as S100A8 and S100A9 [16, 29, 30]. Our studies showed that the effects of PUFAs on MDSCs were STAT3 signaling dependent and blockage of STAT3 phosphorylation nearly reversed the PUFAs’ effect on MDSCs. However, the mechanism upstream of STAT3 activation by PUFAs is still unknown. PUFAs are ligands for peroxisome proliferator-activated receptor-γ (PPARγ), which has anti-inflammatory properties [33, 34]. There is also cross talk between PPARγ and the STAT3 signaling pathway [35]. It is possible that PPARγ may mediate the effects of PUFAs on STAT3 signaling. However, our preliminary data show that PPARγ antagonist fails to counteract the effects of PUFAs on MDSCs, which argues against this possibility. Another possibility is that PUFAs may modulate the activity of enzymes responsible for generation of lipid second messengers from cell membrane phospholipids, which then activate protein kinases, such as PKC, and activate JAK-STAT3 signaling [36]. The mechanism of PUFAs’ effects on MDSCs accumulation or myeloid cell differentiation deserves further investigation.

In addition to its effect on MDSCs expansion, PUFAs supplementation also enhanced the suppressive activity of MDSCs, which was dependent on ROS production. The enhanced ROS production in PUFA-MDSCs is associated with the induction of NAPDH subunit p47phox, since other subunits did not differ between PUFAs-derived MDSCs and control MDSCs. In support of the importance of STAT3 in the effects of PUFAs on MDSCs, genes related to ROS production were also subjected to STAT3 regulation, and STAT3 inhibitor JSI-124 almost reversed the ROS level and p47phox expression in PUFA-MDSCs to that of control MDSCs. Therefore, we conclude that STAT3 signaling is essential for expansion and function of MDSCs generated by PUFAs.

Recent studies have demonstrated that COX-2 and one of its major synthetic products PGE2, play important role in MDSCs expansion under tumor-bearing conditions in mouse and human [37, 38]. Considering that PGE2 and PUFAs both are lipid mediators, we also investigated if there is potential relationship between them. Preliminary results showed that PUFAs administration did not impact the expression of COX2, COX2 inhibitor either failed to affect ROS production and the function of PUFA-derived MDSCs (Supporting Information), indicating there may be no direct relationship between PGE2 and PUFAs in MDSCs biology. However, further investigation is necessary before any conclusion could be drawn.

PUFAs have diverse effects on immune cells [14, 39-41]. PUFAs promote T-helper1 (Th1) to T-helper2 (Th2) effector T-cell differentiation, which favors inhibition of inflammation under pathological conditions [41]. PUFAs also impair LPS-induced maturation of DC and therefore influences the quality and quantity of adaptive immunity [13, 34, 40, 42]. Our studies suggest that the way PUFAs mediate expansion of suppressive MDSCs may be an alternative mechanism for PUFAs to exert their immune suppression. Considering that MDSCs have diverse effects in the immune system under a variety of pathological conditions, discovery of the effect of PUFAs on MDSCs may yield broader impacts on immune responses.

In addition to their effects on immune system, PUFAs can also influence tumor development. Previous reports have shown that dietary intake of PUFAs could impact tumor growth. Diets enriched with ALA or LA have been shown to promote tumor growth in mice [43], although the effects of some n-3 PUFAs on cancer are controversial [44]. Our studies support a tumor promoting effects of ALA or LA. The significant elevation of MDSCs levels in tumor-bearing mice in PUFA-fed groups indicates that MDSCs accumulation could represent one of the major mechanisms of PUFAs’ enhancive effects on cancer. Our study, therefore, may provide a novel explanation about the relationship between PUFAs and cancer.

In summary, our studies have uncovered a novel role for PUFAs in MDSCs biology, as well as its underlying molecular mechanism. Expansion of suppressive MDSCs may represent another mechanism for PUFA-mediated immune suppression in inflammatory diseases.

Materials and methods

Mice and cell lines

All mouse experiments were approved by the Sun Yat-Sen University Institutional Animal Care and Use Committee. C57BL/6 and BALB/c mice were obtained from the Animal Experimental Center of Sun Yat-Sen University. All mice were maintained under specific pathogen-free conditions and used at 6–8 weeks of age. For mouse feeding experiments, mice were fed with the RD, LOD, and SOD for 3 months. For tumor-bearing mice models, mice were subcutaneously injected with tumor cells (2 × 106 cells/mouse), including LLC for C57BL/6 mice, and CT26 for BALB/c mice. Tumor volumes were monitored and mice were sacrificed 24 days postinjection. HepG2, CT26, LLC cell lines were purchased from ATCC and cultured under conditions following the instructions from ATCC.

Reagents and antibodies

RPMI 1640, DMEM, Lipofectamine 2000, 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, IL6 were purchased from Peprotech (Oak Park, CA). The antibodies against arginase-1, iNOS, p47phox, S100A8, S100A9, STAT3, p-STAT3, STAT6, p-STAT6, β-actin, and HRP-conjugated secondary antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). NW-hydroxy-nor-Arginine (nor-NOHA) and L-NG-monomethyl-Arginine (L-NMMA) and Nimesulide (Nim) were obtained from Cayman Chemical (Ann Arbor, MI). N-Acetyl-L-cysteine (NAC), L-Arginine, ConA, α-linolenic acid (ALA, C18:3(n-3)), linoleic acid (LA, C18:2(n-6)), Cucurbitacin I hydrate (JSI-124), and dimethyl sulfoxide were purchased from Sigma-Aldrich (St. Louis, MO). The following fluorescein-conjugated anti-mouse antibodies: CD3e-PE-Cy7, CD4-PE, CD8a-PE-Cy5, Gr-1-PE-Cy5, CD11b-PE, CD11c-PE-Cy5, and their corresponding isotype controls were obtained from eBioscience (San Diego, CA). Gr-1-allophycocyanin, Ly6G-PE-Cy7, Ly6C-PE, and CD11b-FITC were obtained from BD Biosciences (San Jose, CA), and MHC Class II-PE was from Miltenyi Biotec (Auburn, CA).

Generation of mouse MDSCs in vitro

Mouse BM cells were obtained from the femurs and tibias. Two million BM cells were cultured in RPMI 1640 medium supplemented with 10% FBS, 20 ng/mL GM-CSF, 50 μM 2-ME, without or with 20 ng/mL IL-6. The cultures were maintained at 37°C in 5% CO2-humidified atmosphere in 24-well plates. Mediums were refreshed on day 3. Cells were analyzed by flow cytometry on day 6.

Quantative real-time PCR and Western blotting

These experiments were performed following the procedures described earlier [45].

Flow cytometry

Cells were stained with specific fluorescein-conjugated anti-mouse antibodies and the proportions of different cell populations were determined by a FACS Calibur flow cytometer (Beckman Coulter Fullerton, CA). All data were analyzed by CellQuest program (BD, Mountain View, CA). For purification of MDSCs, cultured BM cells or mouse splenocytes were stained with PE-CD11b and PE-Cy5-Gr-1 antibodies and Gr1+CD11b+ cells were isolated by cell sorting on a FACSAria cell sorter (BD, Mountain View, CA). For isolation of T cells, mouse splenocytes were stained with CD3e-PE-Cy7 antibody and CD3+ T cells were purified by flow cytometric sorting.

ROS production

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 and Gr-1-PE-Cy5 antibodies on ice. The ROS content in MDSCs was analyzed by flow cytometry.

Arginase activity and NO production

We followed the same procedures as described earlier [46, 47].

Allogeneic mixed lymphocytes reaction

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 ConA (5 μg/mL) and cultured alone or cocultured with allogeneic MDSCs (from C57BL/6 mice) at different ratios for 3 days. Cells were then stained with CD4-PE or CD8a-PE-Cy5 antibodies, and T-cell proliferation was analyzed by flow cytometry.

Plasmid constructs and transfection assays

The 5′-regulatory sequence of the mouse S100A8 gene was amplified by PCR using the primers listed in Supporting Information Table 1. Then the promoter fragments were cloned into pGL3-Basic vector (Promega, Madison, WI), and the recombinations were confirmed by DNA sequencing. Transient transfections of reporter plasmid were performed on HepG2 cells in 24-well culture plates using Lipofectamine 2000 following manufacturer's instructions. The luciferase activity was measured at 48 h posttransfection.

Statistics

Statistical analysis between different groups were done using unpaired t tests. GraphPad Prism version 5.0a and SPSS Statistics 17.0. softwares were used. p < 0.05 were considered significant.

Acknowledgments

This work was supported by the following grants to J.Z.: National Key Basic Research Program of China (973 Program, No. 2012CB524900), National Natural Science Foundation of China (No. 81072397; No.31270921), Guangdong Innovative Research Team Program (No. 2009010058), Key Research Projects of National 12th Five-year Plan for the Prevention and Treatment of Major Infectious Diseases (No.2012ZX10001003-003), Natural Science Foundation of Guangdong (No. S2011020006072), the Fundamental Research Funds for the Central Universities, the Provincial Talents Cultivated by “Thousand-Hundred-Ten” Program of Guangdong Province.

Conflict of interest

The authors declare no financial or commercial conflict of interest.

Abbreviations
ALA

α-linolenic acid

CFSE

5,6 carboxyfluorescein diacetate, succinimidyl ester

CM-H2DCFDA

5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate, acetyl ester

G-MDSC

granulocytic MDSC

LA

linoleic acid

LLC

Lewis lung carcinoma

LOD

linseed oil diet

MDSC

myeloid-derived suppressor cell

M-MDSC

monocytic MDSC

PPARmath formula

peroxisome proliferator-activated receptor-γ

PUFA

polyunsaturated fatty acid

RD

regular diet

SOD

sunflower oil diet

Ancillary