Astragalus polysaccharides exerts immunomodulatory effects via TLR4-mediated MyD88-dependent signaling pathway in vitro and in vivo

Astragalus polysaccharides (APS), which is widely used as a remedy to promote immunity of breast cancer patients, can enhance immune responses and exert anti-tumor effects. In this study, we investigated the effects and mechanisms of APS on macrophage RAW 264.7 and EAC tumor-bearing mice. Griess reaction and ELISA assays revealed that the concentrations of nitric oxide, TNF-α, IL-1β and IL-6 were increased by APS. However, this effect was diminished in the presence of TAK-242 (TLR4 inhibitor) or ST-2825(MyD88 inhibitor). In C57BL/10J (TLR4+/+wild-type) and C57BL/6J (MyD88+/+wild-type) tumor-bearing mice, the tumor apoptosis rate, immune organ indexes and the levels of TNF-α, IL-1β and IL-6 in blood increased and the tumor weight decreased by oral administration of APS for 25 days. APS had no obvious effects on IL-12p70. However, these effects were not significant in C57BL/10ScNJ (TLR4-deficient) and C57BL/B6.129P2(SJL)-Myd88m1.1Defr/J (MyD88-deficient) tumor-bearing mice. qRT-PCR and Western blot indicated that APS stimulated the key nodes in the TLR4-MyD88 dependent signaling pathway, including TLR4, MyD88, TRAF-6, NF-κB and AP-1, both in vitro and in vivo. However, TRAM was an exception. Moreover, TRAF-6 and NF-κB were not triggered by APS in gene-deficient tumor-bearing mice. Therefore, APS may modulate immunity of host organism through activation of TLR4-mediated MyD88-dependent signaling pathway.

Griess reaction. RAW 264.7 cells were seeded (5 × 10 4 cells/ml) in 24-well culture plates and treated with or without APS (400 μ g/ml) or LPS (100 ng/ml) for different periods of time (4 h, 8 h, 16 h, 24 h, 32 h, 48 h and 72 h) in vitro. The volume of cell culture medium was 0.5 ml. The culture supernatants in different time groups were collected at each time point for the analysis. NO 2 − accumulation was used as an indicator of nitric oxide (NO) production. And, the nitrite content in the culture supernatant was determined by Griess reaction as previously described 7 . Measurement of apoptosis by flow cytometry. Tumor tissues was minced into small pieces and then digested by collagenase type IV with a concentration of 10 mg/ml for 2 h at room temperature. After digestion, the cells were washed twice in DMEM and then washed in PBS. Tumor cell apoptosis was detected by flow cytometry using Annexin-V-FITC apoptosis detection kit according to the instructions.
Quantitative real-time PCR assay (qRT-PCR). The total RNA of spleen homogenates was extracted and reverse transcribed into cDNA. qRT-PCR was performed on Bio-Rad CFX-96 (Bio-Rad, Foster City, CA, USA). The condition of qRT-PCR amplification was as follows: initial denaturation for 30 s at 95 °C, followed by 40 cycles of denaturation at 95 °C for 5 s, annealing and extension at 56 °C for 30 s. Reduced glyceraldehyde-phosphate dehydrogenase (GAPDH) and β -actin served as internal controls. The relative mRNA expressions of TLR4, MyD88, TRAM, TRAF-6, NF-κ B and AP-1 were calculated by Vandesompele Method 26 . The sequences of the primers were listed in Table 1. Triplicate reactions were run per sample and each experiment was repeated three times.
Western blot. The proteins were extracted from the spleen tissues and the treated RAW 264.7 macrophage cells. Then, proteins were separated by 10% sodium dodecyl sulfate-polyacrylamide gel and electro-blotted onto polyvinylidene difluoride membranes (Immunobilon TM-P; Millipore, USA). After blocking with 5% BSA in TBST buffer (Tris 10 mM, NaCl 150 mM, pH 7.6, 0.1% Tween 20), the membranes were probed with primary and secondary antibodies. Protein bands were visualized by enhanced chemiluminescence (Millipore) and analyzed with ChemiDoc Imaging system (Bio-Red, USA). Statistical analysis. SPSS 17.0 software (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. Data were shown as mean ± standard deviation (SD). Differences between two groups were assessed by unpaired two-tailed Student's t-test. Data sets that involved more than two groups were assessed by one-way analysis of variance (ANOVA). Differences were considered statistically significant at P < 0.05.

In vitro Experiments. APS increases the secretion of immunomodulatory factors by RAW 264.7 mac-
rophages. To investigate the immunomodulatory effects of APS on macrophages, the levels of NO and cytokines in the culture supernatants of RAW 264.7 cells were detected by Griess reaction and ELISA. A preliminary test determined that the optimal dose and optimal incubation time of APS were 400 μ g/ml and 24 h for RAW 264.7 cells In vitro. As shown in Fig. 1, the NO production was significantly increased by APS and LPS, remained relative high level from 8 h to 72 h (P < 0.05). As shown in Table 2, the treatment of APS significantly increased the IL-1β , IL-6 and TNF-α secretion compared with control group (P < 0.05). However, APS had no significant effect on IL-12p70 (P > 0.05). The results suggest that APS can stimulate RAW 264.7 macrophages to secrete NO, IL-1β , IL-6 and TNF-α in vitro.
APS has effects on the expressions of mRNAs and proteins of TLR4 signaling pathway in RAW 264.7 macrophages. To explore the immunoregulatory mechanism of APS, the expressions of mRNAs and proteins of the key nodes (TLR4, MyD88, TRAM, TRAF-6, NF-κ B and AP-1) in TLR4 signaling pathway were detected using qRT-PCR ( Fig. 2A) and Western blot ( Fig. 2B and C). After incubation with APS or LPS for 24 h, the mRNA and protein expression levels of TLR4, MyD88, TRAF-6, NF-κ B and AP-1 were significantly elevated compared with those in the control group (all P < 0.05). In contrast, the mRNA and protein expression levels of TRAM in the APS group were not significantly different from those in the control group (P > 0.05). These results suggest that  APS activates TLR4 signaling pathway but selectively up-regulates the mRNA and protein expressions of some key nodes. These results also imply that the effects of immunoregulation by APS are probably mediated through TLR4 signaling pathway.
APS promotes immunomodulatory effects of RAW 264.7 macrophages via TLR4 and MyD88. To further analyze whether TLR4 and MyD88 are involved in APS-induced macrophage activation, cells were pre-treated with TAK-242 and ST2825, which are inhibitors of TLR4 and MyD88, respectively. Then, ELISA was performed to detect cytokine levels of TNF-α and IL-6. As shown in Fig. 2D and E, in cells without inhibitors, the production of TNF-α and IL-6 was significantly increased by APS and LPS, compared with those in control group (P < 0.05). However, with the presence of inhibitors, the TNF-α and IL-6 production induced by APS were suppressed and were significantly lower than those without inhibitors (P < 0.05). The results confirm that the effect of APS on immunoregulation in macrophages is probably acted through TLR4 and MyD88.

In vivo Experiments. APS inhibits tumor weight and facilitates immune organ indexes and cytokines
secretion in EAC-bearing mice. To further investigate the immunoregulatory effect of APS in vivo, EAC tumor-bearing mice were used. EAC cells were diluted and inoculated into the right armpit of each mouse. The solid-tumor-bearing mice model was established and treated as described in the Materials and Methods section. Adriamycin (ADM), commonly used in the treatment of cancers 27 , was used as the positive control for the in vivo experiments. Tumor weight was analyzed to evaluate tumor inhibition by APS. As shown in Table 3, the tumor weight in APS group and ADM group was significantly declined, compared with those in the NS group (P < 0.05). Moreover, the tumor cells apoptosis of EAC tumor-bearing mice were detected by flow cytometry. As shown in Fig. 3A,B, cell apoptosis rate in NS group, ADM group, APS group and LPS group was 20.65 ± 12.56, 72.04 ± 29.03, 54.22 ± 23.93 and 29.21 ± 6.07, respectively. Compared with NS group, the apoptosis rate in APS group and ADM group were significantly increased (P < 0.05), while those in LPS group had no obvious changes (P > 0.05). For immunomodulation, the weight of thymus and spleen was measured and organ index was calculated ( Table 2). The thymus and spleen index of APS group were significantly higher than those in NS group (P < 0.05). However, the thymus and spleen index had no significant differences between in ADM group and NS group (P > 0.05). Furthermore, cytokines in serum of EAC tumor-bearing mice were analyzed by ELISA. As shown in Fig. 3C, the levels of IL-1β , IL-6 and TNF-α were significantly elevated by APS and LPS, compared with those in NS group (P < 0.05). In LPS group, IL-12p70 level was also significantly increased than NS group. However, APS had no significant influence on the level of IL-12p70 (P > 0.05). ADM has no significant effects on the levels of all detected cytokines in mice peripheral blood (P > 0.05). The results indicate that APS has significant immunoregulatory and anti-tumor effects on tumor-bearing mice. Macrophages were cultured with APS (400 μ g/ml) or LPS (100 ng/ml) for 4-72 h. The supernatants were assayed for NO production by Griess reaction. *P < 0.05, **P < 0.01 vs. Control group.
As shown in Fig. 4, APS and LPS were found to significantly induce the expressions of mRNAs and proteins of TLR4, TRAF-6, NF-κ B and AP-1 in the TLR4 +/+ tumor-bearing mice, compared with those in the NS group (P < 0.05). Inversely, there was no remarkable differences in the expression of mRNA and proteins of TLR4, TRAF-6, NF-κ B and AP-1 among groups in the TLR4 −/− tumor-bearing mice (P > 0.05).
In TLR4 +/+ tumor-bearing mice, the levels of TNF-α and IL-6 ( Fig. 5A,B) in APS and LPS groups were significantly higher than those in NS and ADM TLR4 +/+ groups (P < 0.05). In TLR4 −/− tumor-bearing mice, the concentrations of TNF-α and IL-6 had no significant differences among groups (P > 0.05). Additionally, the secretion of cytokines in the TLR4 −/− mice of the APS and LPS groups were significantly lower than those in the TLR4 +/+ mice (P < 0.05).
In summary, APS induces the expressions of mRNAs and proteins of TLR4, TRAF-6, NF-κ B and AP-1, as well as increasing the levels of TNF-α and IL-6 in the serum of the TLR4 +/+ tumor-bearing mice, but not TLR4 −/− tumor-bearing mice. This indicates that TLR4 signaling pathway is probably involved in the anti-tumor and immunomodulation effects induced by APS.
Anti-tumor immunomodulatory effects of APS are mediated by TLR4-MyD88-dependent signaling pathway. To clarify the detailed mechanism of anti-tumor immunomodulatory effects of APS and to verify whether APS activates TLR4 signaling pathway through MyD88-dependent pathway, we used C57BL/6J (MyD88 +/+ ) and MyD88-deficient (MyD88 −/− ) tumor-bearing mice in this study.
The mRNA and protein expressions of TLR4, MyD88, TRAF-6, NF-κ B and AP-1 were detected by qRT-PCR and Western blot (Figs 6 and 7). Results showed that the mRNA and protein levels were obviously provoked by APS and LPS in the MyD88 +/+ tumor-bearing mice, compared with those in the NS group (P < 0.05). However, there was no significant difference in the mRNA and protein expressions of TRAF-6 and NF-κ B among groups in the MyD88 −/− tumor-bearing mice (P > 0.05). TLR4 is in the up-stream of the pathway. Therefore, the deficiency of MyD88 slightly affected the induction of TLR4 expression by APS and LPS (Figs 6 and 7). On the  Table 3. The comparison of tumor weight and immune organ indexes of EAC tumor-bearing mice (mean ± SD, n ≥ 4). Note: *P < 0.05 vs NS group. **P < 0.01 vs NS group.  Moreover, both TNF-α and IL-6 in the serum of the C57BL/6J (MyD88 +/+ ) tumor-bearing mice stimulated by APS and LPS were obviously higher than those in NS and ADM groups (P< 0.05, Fig. 8A and B). The concentrations of TNF-α and IL-6 in the APS group had no obvious difference from those in the NS group of the MyD88 −/− tumor-bearing mice (P > 0.05). In addition, the secretion of cytokines after treatment with APS in the MyD88 −/− mice were significantly reduced, compared with those in the APS groups of the MyD88 +/+ tumor-bearing mice (P < 0.05). The positive control LPS also significantly increased TNF-α and IL-6 levels in both MyD88+ /+ and MyD88− /− mice (P < 0.05). However, the levels of TNF-α and IL-6 in LPS groups had no significant differences between in MyD88+ /+ and MyD88− /− mice (P > 0.05). All these results suggest that APS can selectively activate the down-stream key nodes and the terminal effect factors of TLR4-MyD88-dependent pathway in vivo.

Discussion
APS, have been shown to have multiple immunomodulatory functions, such as inhibiting the proliferation of CD4 + CD25 + regulatory T cells 10 , promoting the maturation of dendritic cells (Shao et al., 2016), regulating the imbalance of Thl/Th2 subgroups, regulating the differentiation of the erythroid lineage 28,29 , and enhancing the cytostatic activity of macrophages 29 . Macrophages are specific antigen-presenting cells that play important roles in anti-infection and anti-tumor immunity through engulfing and eradicating pathogens 7 . There are two approaches for activated macrophages to immunomodulate and kill tumor cells or pathogens, including direct contact or releasing cytotoxic molecules such as NO and cytokines 30 .
NO, synthesized by macrophage-induced nitric oxide synthase, has been identified as a major effect molecule involved in many biological processes 31 , including pathogen elimination 32 and destruction of tumor cells by activated macrophages 33 . IL-1β , IL-6, IL-12p70 and TNF-α are typical multifunctional cytokines involved immune responses, hematopoiesis and inflammation. IL-1β has a wide range of immunomodulaory effects and may mediated inflammation or be directly involved in the inflammatory process 34 . IL-6 is a cytokines with multiple immunomodulatory functions, and it can stimulate B cells, T cells and stem cell proliferation, promote the B cell production of immnoglobulin and promote cytotoxic lymphocyte and stem cell differentiation 35 IL-12p70, the active heterodime, is known as a T-cell stimulating factor which can stimulate the growth and function of T cells 36 . TNF-α generated by macrophages is implicated in cytotoxic function in certain tumors 5 , and in the development and procession of immunoregulatroy effects of APS 37,18 . Our result showed that APS could directly increase the NO, IL-1β , IL-6 and TNF-α production by macrophages in vitro, but not IL-12p70. Similarly, the level of IL-1β , IL-6 and TNF-α was also increased by APS in EAC tumor-bearing mice in vivo. APS was reported to improve the spleen/thymus indexes of H22 tumor bearing mice 38 . In this study, the immune organ indexes of tumor-bearing mice were also higher than those in NS group. These data indicate that APS improves the secretion of immunomodulatory factors in vitro and in vivo.
As mentioned before, APS is wildly accepted as a complementary and alternative therapy for cancer patients 9,10 . APS is also reported to inhibit the growth of breast cancer cell line MDA-MB-468 17 and enhance the therapeutic effect of cisplatin 39 . It is found that the anti-tumor effect of APS on H22 tumor-bearing mice might be related to its ability to enhance the expression of IL-1, IL-2, IL-6 and TNF-α and decrease of IL-10 40 . In the present work, in tumor bearing mice, the tumor weight was inhibited by APS and the tumor cell apoptosis rate was increased by APS. Together, these results suggest that APS has anti-tumor activity in tumor-bearing mice. And, this effect may be acted via regulating the production of cytokines. Subsequently, we explored the underlying mechanism of APS. TLR4 signaling pathways have two branches: MyD88-dependent and MyD88-independent signaling pathways 19 MyD88, containing a death domain in cytoplasm 11 , is bound to the structural domain of TLR by MyD88 adaptor-like (MAL) protein, which is an essential adapter protein binding to MyD88 and activating the downstream molecule TRAF-6 39 . TRAM, with a similar role as MAL, is able to combine with TRIF and TLR 41 . and is specifically involved in the MyD88-independent signaling pathway 19 . It delivers TRIF to the endosomes via a specific region of the plasma membrane 41 . TRAF-6, is also recruited to the complex of TRAM/TRIF to regulate the expression of correlated cytokines and interferon type I/II 42 . Therefore, TRAF-6 can be considered as the intersection of the two types of TLR4 signaling pathway, which can activate NF-κ B and MAPK 7 to regulate the expression of the downstream key nodes 43 . In addition, AP-1 is the heterodimer of c-Fos and c-Jun 44 , whose binding sites are regulated by TNF-α through MAPKs-mediated pathway 45 . In view of these, we detected the mRNA and protein levels of TLR4, TRAF-6, NF-κ B and AP-1 in RAW 264.7 cells. Results showed that the expressions of the key nodes were significantly induced by APS, compared with the control group in vitro. Besides, the mRNA and protein expressions of TLR4, TRAF-6, NF-κ B and AP-1 in the splenocytes of the TLR4 +/+ EAC tumor-bearing mice were also observed increased significantly. Furthermore, the APS-induced TLR4, TRAF-6, NF-κ B and AP-1 expression were decreased in the splenocytes of the TLR4 −/− EAC tumor-bearing mice. TAK-242 (Resatorvid), a small-molecule specific inhibitor of TLR4, binds selectively to TLR4 and interferes with the interactions between TLR4 and its adaptor molecules 46,47 . The APS had slight influence on the secretions of TNF-α and IL-6 in RAW 264.7 macrophages treated with TLR4 inhibitors (TAK-242). The data indicates that TLR4 involves in APS-mediated immune activation. This result partly supports the opinion of the study implying directly interaction between APS and TLR4 on cell surface 25 . Moreover, APS-induced immunoregulation may via TLR4 signaling pathway.
To give a good insight into the deep mechanism of APS immunoregulation and to deeply verified which branch of TLR4 signaling pathway is involved in APS-mediated immune activation, the key nodes expression in RAW 264.7 cells, MyD88 +/+ and MyD88 −/− tumor-bearing mice were detected. TRAM, specifically involved in the MyD88-independent signaling pathway, had no obvious changes with treatment of APS both in RAW 264.7 cells and in MyD88 +/+ tumor-bearing mice. The mRNA and protein expressions of TRAF-6, NF-κ B and AP-1 in the splenocytes of MyD88 −/− tumor-bearing mice were not provoked by APS. And the improvement of APS on the secretion of IL-6 and TNF-α was also suppressed by ST-2825. ST-2825, a MyD88 pharmacologic inhibitor, is widely used to impede the dimerization of MyD88 48 . The activation of TLR4 signaling pathway by APS were almost lost when TLR4 and MyD88 were deficient or inhibited. Taken together, we suppose that APS may activate the TLR4-MyD88 dependent pathway through TLR4 (Fig. 9). Then, the key nodes in the TLR4-MyD88    The TLR4-mediated MyD88-dependent signaling pathway is probably one of mechanisms underlying the immunoregulation and anti-tumor effects of APS. dependent pathway, including TRAF-6, NF-κ B and AP-1, are activated. Finally, the production of effector cytokines such as IL-6 and TNF-α is enhanced to mediate the immunomodulation effects of APS.
However, exceptions still exist. It is confusing that the mRNA and protein expressions of AP-1 were increased by APS and LPS in both MyD88 +/+ mice and MyD88 −/− mice. AP-1, which is another eukaryotic transcription factor targeted by MAPK signaling pathways 49 , is an important regulatory protein involved in various biological activities, and also contributes to inflammatory and immune responses 50 . Thus, we assumed that, when MyD88 was deficient, APS activated AP-1 through another signaling pathway other than the TLR4-MyD88-denpendent signaling pathway. On account of the complexities of signal transduction with various intersections, we speculated that APS might provoke other signaling pathways through TRAM or might have potential relationship with AP-1 when MyD88 was deficient. However, no relative reports are found. The detailed mechanism of this phenomenon needs to be thoroughly investigated in future studies.
In conclusion, the TLR4-mediated MyD88-dependent signaling pathway is probably one of the APS-induced signal pathways underlying the immunoregulation and anti-tumor effects of APS both in vitro and in vivo.