Schisantherin A Attenuates Neuroinflammation in Activated Microglia: Role of Nrf2 Activation Through ERK Phosphorylation

Neuroinflammation is an inevitable and important pathological process involved in all types of damage and disorders in the brain [1]. Microglia, which serve as immune-surveillance cells, are well-established major components of neuroinflammatory responses in the central nervous system (CNS) [1, 2]. In response to stimuli, microglial cells can be polarised into activated inflammatory phenotypes, inducing overexpression of various pro-inflammatory cytokines and mediators [3]. These pro-inflammatory mediators can lead to lesions in neighbouring neurones, which induces more microglia activation as feedback [1-3]. Clinically, aberrant microglial activation and subsequent neuroinflammation contribute to several neurodegenerative diseases, like amyotrophic lateral sclerosis (ALS), Alzheimer’s disease (AD), and Parkinson’s disease (PD) [1-3]. Therefore, attenuating excessive microglial cells activation might be a potential treatment for neuroinflammation and neurodegeneration diseases.

The neuroinflammatory responses induced by activated microglia are characterised by excessive production of pro-inflammatory mediators and cytokines, e.g. nitric oxide (NO), prostaglandin E₂ (PGE₂), tumour necrosis factor (TNF)-α, IL(interleukin)-1β, and IL-6, et. al [4]. The activated microglia-mediated neuroinflammation is mainly regulated by transcription factor nuclear factor kappa B (NF-κB), a crucial modulator of various kinds of inflammation [5]. Normally, the inactive NF-κB proteins are bound to the inhibitor of kappa B (IκB) proteins and sequestered in the cytosol. To induce NF-κB activation, the IκB kinase (IKK) firstly phosphorylates and degrades IκB, which dissociates NF-κB from complex and induces NF-κB translocation into the nucleus, triggering the production of pro-inflammatory molecules [5]. Consequently, suppressing NF-κB activation is one of the actions of known anti-inflammatory agents [5]. Besides, mitogen-activated protein kinase (MAPKs) pathways and the phosphatidylinositol 3-kinase/Akt (PI3K/Akt) signal axis are also involved in NF-κB activation and pro-inflammatory responses [4, 5].

Accumulating evidence indicates that microglial activation also caused excessive oxidative stress, especially due to extravagant reactive oxygen species (ROS) accumulation [6]. Intracellular ROS act as important second messengers in microglial inflammation and subsequently induce neuronal cell death [7]. Interestingly, the level of endogenous ROS can be regulated by endogenous antioxidants and antioxidative enzymes via the nuclear factor erythroid 2-related factor 2 (Nrf2)/antioxidant response element (ARE) signal pathways [8]. In response to stimuli or stress, Nrf2 can be dissociated from Kelch-like ECH-associated protein 1 (Keap-1) in the cytoplasm, translocated into the nucleus. The nucleus Nrf2 then binds to the ARE sequence to activate the transcription of genes that encode for endogenous antioxidant and electrophile detoxification, like heme oxygenase-1 (HO-1), and NADPH quinone oxidoreductase-1 (NQO-1) [8]. Apart from protective actions against oxidative damage, current researches have also implicated Nrf2-related pathways in restoration of redox homoeostasis, attenuation of pro-inflammatory responses, and neuroprotective actions in microglia cells [9]. Notedly, activation of Nrf2 pathways can promote microglia polarisation toward an anti-inflammatory phenotype, thus inhibiting neuroinflammation [9-11]. Therefore, activation of Nrf2 signal pathways might be a valuable strategy for modulating neuroinflammation.

Schisantherin A (StA) is a dibenzo cyclooctadiene lignin, mainly isolated from the fruits of Schisandra sphenanthera [12]. StA was reported to have various kinds of biological properties, including antioxidation, detoxification, and liver protection [13, 14]. The anti-inflammatory and neuroprotective effects of StA were widely recognised in several different models [15-17]. Firstly, StA can remarkably inhibit pro-inflammatory responses in LPS-stimulated RAW264.7 cells [17], and significantly suppresses IL-1β-stimulated osteoarthritis in human chondrocytes. Also, StA was found to suppress tissue damage and improve the animal survival rate in an LPS-induced acute respiratory distress syndrome mouse model [18]. Moreover, our previous studies demonstrated that StA exhibited significant neuroprotective actions against neurotoxin-induced dopaminergic neuron loss and locomotion deficiency in zebrafish and mice models of PD [19, 20]. Results from our 6-hydroxydopamine (6-OHDA)-induced neurotoxic model in SH-SY5Y cells also demonstrated that StA could attenuate intracellular ROS accumulation, NO overproduction, and the expression of inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) via downregulating MAPKs and PI3K/ Akt pathways [19, 20].

Previous studies have indicated a potential relationship between StA and neuroinflammation. However, whether StA can inhibit neuroinflammation, and the underlying mechanisms, have not been investigated. In the current study, we aimed to investigate the anti-neuroinflammatory effects of StA in LPS-activated BV-2 microglial cells. Furthermore, the roles of StA in NF-κB- and Nrf2-related signaling pathways were investigated.

Schisantherin A (purity > 99%) was supplied by Chengdu Preferred Biological Technology (Chengdu, China). Zinc protoporphyrin-IX (ZnPP) and LPS (Escherichia coli serotype 055: B5) were purchased from Sigma-Aldrich (St. Louis, USA). Ro-31-8220, LY294002, SB203580, U0126, and PDTC were purchased from Selleck Chemicals (Shanghai, China). RPMI 1640, DMEM, OPTI medium, FBS, penicillin, streptomycin, RNA iMAX Kit, DAPI solution and BCA Kit were purchased from Thermo Fisher Scientific (Carlsbad, CA, USA). Nitric Oxide Colorimetric Assay Kit was purchased from BioVision (Milpitas, CA, USA). The ELISA Ready-SET-Go Kits purchased from eBiosciences (San Diego, CA, USA). High Pure RNA Isolation Kit, Transcriptor First Strand cDNA Synthesis Kit, and FastStart Universal SYBR Green Master Reagents were purchased from Roche Applied Science (Mannheim, Germany). p-AKT, AKT, p38, p-p p38, p44/42, p-p44/42, SAPK/JNK, p-SAPK/JNK, NQO-1, HO-1, β-actin, GAPDH, and HRP-linked secondary antibody were purchased from Cell Signaling Technology (Beverly, MA, USA); iNOS, COX-2 and p-Nrf2 were purchased from Invitrogen; Nrf2 purchased from Novus Biologicals (Littleton, CO, USA). The Alex Fluor 488 and Alex Fluor 647 secondary antibodies were obtained from Jackson ImmunoResearch Laboratories (West Grove, PA, USA).

BV-2 cells, a widely used immortalised murine microglial cell line [21], were obtained from the Kunming Cell Bank of Type Culture Collection, Kunming Institute of Zoology. SH-SY5Y cells were purchased from the American Type Culture Collection (Rockville, MD, USA). BV-2 cells and SH-SY5Y cells were cultured in RPMI 1640 and DMEM, respectively. The medium was supplemented with 10% FBS, 100 U/ml penicillin, and 100 µg/ml streptomycin. Cells were cultured in an atmosphere of 95% air and 5% CO₂ at 37 °C. In all experiments, cells were treated with the indicated concentrations of StA with or without LPS (500 ng/ml) in the FBS-free medium for indicated times.

Cell viability was measured by the MTT assay. Briefly, BV-2 cells were seeded in 24- or 96-well culture plates and received indicated treatments. After that, cells were incubated with MTT and finally the absorbance at 570 nm was measured using a FlexStation 3 Microplate Reader (Molecular Devices, CA, USA). All values were normalised to the control group.

Microglial production of NO in culture media was assessed using a Nitric Oxide Colorimetric Assay kit according to the manufacture’s protocol. The level of ROS was analysed using the fluorescent probe CM-H₂DCFDA. The ROS level was measured via determining the fluorescent intensity (excitation: 493 nm and emission: 522nm). Also, cells were visualised using an In Cell Analyzer 2000 system (GE Healthcare, Grandview Blvd, Waukesha, WI, USA).

The TNF-α, IL-6, IL-1β, and PGE₂ released in conditioned media were assessed by specific ELISA Ready-SET-Go kits. The levels were quantified following the manufacturers protocols.

Total RNA of BV-2 cells was extracted using the High Pure RNA Isolation Kit. Isolated RNA was then reverse-transcribed into cDNA using the Transcriptor First Strand cDNA Synthesis Kit. The qPCR assay was conducted using FastStart Universal SYBR Green Master reagentswith the Applied Biosystems 7900 HT Fast Real-Time PCR System (Applied Biosystems Inc., Foster City, CA, USA). The amplification parameters used here were 50 °C for 2 min then 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 30 s. Each sample was analysed in triplicate, and the relative expression of mRNA was calculated after normalisation to GAPDH. The primer sequences used are also listed: iNOS forward, 5’-CAAGAGTTTGACCAGAGGACC-3’; reverse, 5’-TGGAACCACTCGTACTTGGGA-3’; COX-2 forward, 5’-TTGAAGACCAGGAGTACAGC-3’; reverse, 5’-GGTACAGTTCCATGACATCG-3’; TNF-α forward, 5’-CCTATGTCTCAGCCTCTTCT-3’; reverse, 5’-CCTGGTATGAGATAGCAAAT-3’; IL-1β forward, 5’-GGCAACTGTTCCTGAACTCAACTG-3’; reverse, 5’-CCATTGAGGTGGAGAGCTTTCAGC-3’; IL-6 forward, 5’-CCACTTCACAAGTCGGAGGCTT-3’; reverse, 5’-CCAGCTTATCTGTTAGGAGA-3’; IL-10 forward, 5’-GCCAGTACAGCCGGGAAGACAATA-3’; reverse, 5’-GCCTTGTAGACACCTTGGTCTT-3’; HO-1 forward, 5’-TGTCACCCTGTGCTTGACCT-3’; reverse, 5’-ATACCCGCTACCTGGGTGAC-3’; NQO1 forward, 5’-AGAGGCTCTGAAGAAGAGAGG-3’; reverse, 5’-CACCCTGAAGAGAGTACATGG-3’; β-actin forward, 5’-AGCCATGTACGTAGCCATCC-3’; reverse, 5’-GCTGTGGTGGTGAAGCTGTA-3’; GAPDH forward, 5’-ATGTACGTAGCCATCCAGGC-3’; reverse, 5’-AGGAAGGAAGGCTGGAAGAG-3’.

Briefly, BV-2 cells were transfected using the pARE-luc and pNFκB-luc reporter plasmid provided by the Cignal Antioxidant Response Reporter (luc) Kit and Cignal NFκB Reporter (luc) Kit, respectively, using the Attractene Transfection Reagent according to the manufacturer’s protocol. Transfection efficiency was controlled by co-transfection of Renilla luciferase reporter plasmid. About 24 h after transfection, the transfected cells were subjected to further treatment. The samples were then analyzed using the Dual-Luciferase® Reporter Assay System, and luciferase activities were measured.

BV-2 cells with a confluence of 70 to 80% were transfected with Nrf2 siRNA (80 nM, Santa Cruz), ERK siRNA (100nM, Cell Signaling Technology), or scrambled-control siRNA (Santa Cruz), using RNA iMAX kit according to specified protocols. Further experiments were performed at 24 or 36 h after transfection.

For the whole cell protein extraction, BV-2 cells were incubated with RIPA lysis buffer (Beyotime) and cell lysates were then centrifuged and the supernatant was collected and stored. Subcellular fractionation was done using the Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime) according to the protocol of nuclear extraction kit manufacturer. The protein content was assayed using the BCA protein quantification kit (Pierce).

Aliquots of protein samples (∼30 µg) were resolved by SDS-PAGE (7.5% - 15%) and transferred to PVDF membranes. Membranes were blocked, followed by incubation at 4 °C overnight with diluted primary antibodies. Membranes were washed and incubated with HRP-linked secondary antibody for 1 h at room temperature. Finally, bands were visualised using an ECLplus Western Blotting Detection Reagents kit (GE Healthcare). The membranes were then scanned on a Bio-Rad ChemiDoc XRS Imaging System, and the intensity of the protein bands was analysed using Bio-Rad Quantity One Software (Bio-Rad, Hercules, CA, USA).

The DNA binding activity of NF-κB was analyzed using an ELISA-based TransAM NF-κB p65 EMSA kit (ActiveMotif, CA, USA) according to the manufacturer’s protocol. Absorbance at 450 nm was measured.

Cells were washed, fixed permeabilized, and blocked. After that, cells were incubated at 4 °C overnight with the primary antibodies. The next day, cells were incubated secondary antibodies and DAPI. Finally, the samples were imaged using the In Cell Analyzer 2000 system (GE Healthcare).

Statistical analyses were performed using GraphPad Prism software (ver. 6.0; GraphPad Software Inc., San Diego, CA, USA), and data are represented as means ± standard error of the mean (SEM). Statistical analysis of differences between two groups was done using the independent-samples t-test and one-way or two-way ANOVA with Bonferroni’s correction was used for multiple group comparisons. Pearson’s correlation coefficient was used for correlation analyses. P < 0.05 was considered as statistically significant in all analyses.

In order to estimate the range of effective concentrations, we firstly assessed the effect of StA on the cell viability of BV-2 cells. As shown in Fig. 1A and C, StA (0.1 to 50 µM) had no cytotoxic effect on BV-2 cells for 12 or 24 h incubations (all p > 0.01). LPS (500 ng/mL) treatment for 12 and 24 h slightly decreased the cell viability to 93.0 ± 4.3 %, and 90.8 ± 5.2 %, respectively, with no significant difference compared to the control group (Fig. 1B and D, p > 0.05). However, treatment with both StA (2.5 to 50 µM) and LPS also did not affect cell viability of BV-2 cells for 12 and 24 h (Fig. 1B and D, p > 0.05).

We initially evaluated the effects of StA on NO and PGE₂ production, and the expression of iNOS and COX-2, in LPS-stimulated BV-2 cells. As illustrated in Fig. 2A, pretreatment with StA for 1 h markedly decreased LPS-induced NO increases in a concentration-dependent manner (all p < 0.05, versus the LPS-treated group). Then, we found that LPS-induced expression of iNOS was also significantly inhibited by StA treatment in a concentration-dependent manner (Fig. 2B). The best inhibitory effect was found at the dose of 50 µM, at which the levels of NO and iNOS decreased to less than 50 % of that in the LPS-treated group (Fig. 2A and B). Our data further showed that StA (50 µM) did not confer a time-dependent inhibitory action on NO production or iNOS expression (Fig. 2C and D). There was no significant difference in inhibitory action between 1, 2, 4, and 8 h of pretreatment of StA (Fig. 2C and D, all P > 0.05). Correspondingly, pretreatment with StA (50 µM) for 1 h inhibited production of NO and PGE₂ (Fig. 2E), and increased expression of iNOS and COX-2 in mRNA and protein induced by LPS (Fig. 2F and G, all p < 0.01 versus the LPS-treated group). Hence, preincubation of 50 µM StA for 1 h was used in further experiments in the current study (Fig. 2E-G).

Concurrently, we investigated the effects of StA on pro-inflammatory cytokines in LPS-activated BV-2 microglial cells. As shown in Fig. 3A, B and C, the LPS-induced increases in the mRNA expression of TNF-α, IL-6, and IL-1β were significantly decreased following StA treatment. Moreover, StA also markedly suppressed the protein expression of TNF-α, IL-6, and IL-1β in LPS-activated BV-2 microglia (Fig. 3E, F and G, a ll p < 0.01 versus the LPS-treated group). Notedly, StA obviously increased the expression of IL-10 in both mRNA and protein levels (Fig. 3 D and H, all p < 0.01 versus the LPS-treated group).

To explore the role of StA in NF-κB activation and NF-κB-regulated gene transcription and signal transduction, the NF-κB luciferase reporter system was used in current study. The results indicated that StA could inhibit the NF-κB-driven gene transcription activity increased by LPS stimulation in a dose-dependent fashion (Fig. 4A, all p < 0.01 versus the LPS-treated group). Notably, treatment with 50 µM StA significantly decreased NF-κB transactivation to ∼ 42.67 % of that of the LPS-treated group (p < 0.01).

NF-κB is activated by phosphorylation, which triggers NF-κB nuclear translocation and promotes pro-inflammatory gene expression. Here, we then found that StA could significantly decrease the increase of NF-κB p65 phosphorylation in LPS-stimulated BV-2 cells (Fig. 4B, p < 0.01). Next, we examined StA’s effects on nuclear translocation of the NF-κB p65 subunit. As shown in Fig. 4C, treatment with 50 µM StA significantly blocked the NF-κB p65 LPS-activated nuclear translocation, to ∼ 43.2 % of that of the LPS-treated group (Fig. 4C and 4D, p < 0.01). Nuclear translocation was further revealed by immunofluorescence assay (Fig. 4E). As expected, StA significantly alleviated the LPS-stimulated accumulation of NF-κB p65 in the nucleus (Fig. 4E). Followed by nuclear translocation, NF-κB p65 binds to its target DNA and triggers the transcription of pro-inflammatory genes. The ELISA-based EMSA was conducted to assess the binding of NF-κB to DNA, and current results showed that StA significantly abolished the LPS-induced NF-κB-DNA binding activity (Fig. 4F, p < 0.01).

Next, we found that LPS obviously boosted the levels of phosphorylated IKKα/β and IκBα in BV-2 microglia (Fig. 4G and 4H, p < 0.01 versus the control). However, StA significantly decreased the increased phosphorylation of IKKα/β and IκBα (Fig. 4G and 4H, p < 0.01 versus the LPS-treated group). In addition, LPS significantly induced IκBα degradation in BV-2 cells (Fig. 4H, p < 0.001, versus the control group), which was reversed by StA treatment (Fig. 4H, p < 0.01 versus the LPS-treated group).

In order to further investigate StA’s effects on signaling pathways involving in NF-κB activation, we next examined the phosphorylation levels of MAPKs and PI3K/Akt. We found that StA could significantly suppress LPS-induced phosphorylation of p38 (Fig. 5A), JNK (Fig. 5C), and Akt (Fig. 5D), respectively (all p < 0.01 versus the LPS-treated group). Notably, StA did not alleviate the LPS-induced increase in phosphorylation of ERK (Fig. 5B, p > 0.05 versus the LPS-treated group), but StA significantly promoted the ERK phosphorylation in BV-2 cells without LPS stimulation (Fig. 5B, p < 0.01, versus the control group).

Intracellular ROS acts as an important second messenger in microglial inflammation. Hence, we next tested the role of StA in LPS-induced ROS production. Our results demonstrated that LPS stimulation obviously increased ROS production (Fig. 6A and 6B, p < 0.01, versus the control group), whereas treatment with StA significantly decreased the ROS (Fig. 6A, all p < 0.01 versus the LPS-treated group). We also found that StA significantly suppressed LPS-elevated level of phosphorylated p47phox (Fig. 6C, p < 0.01 versus the LPS-treated group).

we next examined the levels of HO-1 and NQO-1 in StA-treated BV-2 microglia. As shown in Fig. 6D and 6E, StA significantly enhanced the HO-1 and NQO-1 expression in LPS-activated BV-2 microglia in a dose-dependent fashion (all p < 0.01, versus LPS-treated group). We then used the HO-1-specific inhibitor, ZnPP to inhibit HO-1 activity and tested the role of StA-induced HO-1 expression in LPS-stimulated neuroinflammatory responses. As expected, pretreatment with ZnPP significantly abolished StA-mediated inhibition of NO, PGE₂, TNF-α, and IL-6, respectively (Fig. 6F and G, all p < 0.01, between the ZnPP-untreated group and ZnPP- treated group).

Encouraged by the elevated level of HO-1 and NQO-1 induced by StA, we further investigated the role of StA in Nrf2-related antioxidant pathways. Firstly, as shown in Fig. 7A, we found that treatment with StA could significantly induce the accumulation of nucleus Nrf2 in a dose-dependent fashion in BV-2 cells without LPS stimulation (all p < 0.01, versus the control group). Moreover, StA could significantly enhance the Nrf2 nucleus accumulation in LPS-activated BV-2 microglia when compared to the cells without LPS stimulation (Fig. 7A, p < 0.01). Furthermore, we found that StA-mediated enhancement of the accumulation of nucleus Nrf2 was time-dependent (Fig. 7B, all p < 0.01). As illustrated in Fig. 7B, the level of nuclear Nrf2 was significantly increased at 1 h after StA administration, peaked at 3 to 6 h, and then declined at about 12 h. As expected, we then found that incubation of 50 µM StA for 6 h could significantly enhance both total and phosphorylated Nrf2 levels in the nuclei of BV-2 cells (Fig. 7C, all p < 0.01). Then, results of immunofluorescence clearly showed that Nrf2 was enhanced in the nucleus (Fig. 7D). We next measured the activity of StA in antioxidant responsive elements by using a luciferase reporter containing the ARE consensus. The results of the ARE reporter assay showed that StA resulted in a dose-dependent increase of ARE-luciferase activity in both control and LPS-activated BV-2 microglial cells (Fig. 7E, all p < 0.01, versus the control group).

We next used siRNA to silence Nfr2 expression in BV-2 microglial cells. We found that StA could also induce a significant accumulation of nuclear Nrf2 in the control siRNA group (Fig. 8A, p < 0.01). However, StA was not able to induce Nrf2 nuclear translocation in Nrf2 siRNA-transfected BV-2 cells (Fig. 8A, p > 0.05, versus the control siRNA group). Moreover, StA-induced expression of HO-1 and NQO-1 was remarkably abolished by Nrf2 siRNA transfection (Fig. 8B, all p < 0.01 versus the control siRNA group). In addition, inhibitory actions of StA on ROS accumulation were totally abolished in Nrf2 siRNA-transfected BV-2 cells (Fig. 8C, all p < 0.01 versus the control siRNA group). Furthermore, we also found that Nrf2 knockdown significantly abrogated StA-mediated inhibition of the production of pro-inflammatory mediators and cytokines, including NO (Fig. 8D), TNF-α (Fig. 8E), and IL-6 (Fig. 8F) (all p < 0.01, versus the control-siRNA group). Moreover, to some extent, Nrf2 siRNA-transfection also suppressed StA-induced inhibitory effects on NF-κB reporter luciferases activity (Fig. 8G) and NF-κB-DNA binding activity (Fig. 8H) (p < 0.01, versus the control siRNA group).

To elucidate the upstream modulator that regulates the StA-mediated activation of Nrf2 in BV-2 microglial cells, various kinase inhibitors, including PKC inhibitor Ro-31-8220, PI3K/ Akt inhibitor LY294002, p38 MAPK inhibitor SB203580, and MEK 1/2 inhibitor U0126 were employed in the following investigation. As shown in Fig. 9, in the ARE reporter luciferase assay, we found that only pretreatment with MEK 1/2 inhibitor U0126 could significantly suppress StA-induced ARE reporter luciferase activity (Fig. 9A, p < 0.01), yet pretreatment of BV-2 microglia with the remaining inhibitors, like, LY294002 (Fig. 9B), SB203580 (Fig. 9C), or Ro-31-8220 (Fig. 9D) did not inhibit the ARE-luciferase activity (all p > 0.01). Moreover, we observed that only preincubation with U0126 significantly abrogated StA-induced activation of HO-1 in BV-2 cells (Fig. 9E, p < 0.01). Current results indicated that ERK kinase was involved in StA-induced Nfr2/ARE activation. This result was also consistent with the previously observed phenomenon (showed in Fig. 5B) of significant StA-induced ERK phosphorylation in BV-2 microglial cells.

Next, we further determined the role of StA in ERK phosphorylation. We found that increases in the phosphorylated forms of ERK were observed from 0.5h after exposure to StA, peaked at 1 to 2 h, and then declined (Fig. 10A, all p < 0.01). Preincubation of U0126 effectively suppressed ERK phosphorylation (Fig. 10B, p < 0.01 versus the control group) and also significantly abolished the StA-induced increase in ERK phosphorylation in BV-2 microglial cells (Fig. 10B, p < 0.01 versus the StA-treated group). Moreover, StA-mediated accumulation of nuclear Nrf2 was totally reversed by U0126 treatment (Fig. 10C, p < 0.01 versus the control group). Further, ERK siRNA was used to knockdown the expression of ERK. As shown in Fig. 10D, ERK siRNA transfection effectively silenced the ERK expression, evidenced by significantly decreased expression of total and phosphorylated ERK in BV-2 cells (both p < 0.01, versus the scramble-siRNA group). As expected, we found that StA-induced ERK phosphorylation and accumulation of nuclear Nrf2 was abolished in ERK siRNA transfected-BV-2 cells (Fig. 10E, p < 0.01, versus the scramble-siRNA group).

Neuroinflammatory responses are inevitable and important pathological processes in several types of damage and disorders in the CNS [1]. Microglial activation-mediated inflammatory responses are known to play a crucial role in brain neuroinflammation and subsequent neuronal injuries [1, 2]. Therefore, alleviating activation of the microglia-induced inflammatory process might be a valuable therapeutic approach for neuroinflammation-related diseases. Since StA was found to be anti-inflammatory and neuroprotective in several reports [22], this study aimed to investigate whether StA could inhibit microglia-mediated neuroinflammation in LPS-activated BV-2 microglia cells.

The BV-2 microglia cells are well known to be conferred pro-inflammatory properties following LPS stimulation [4]. Initially, we found that StA could suppress both NO/iNOS and PGE₂/COX-2 signaling pathways in LPS-activated BV-2 microglia. NO and PGE₂ are two crucial mediators of inflammation that are synthesised via iNOS and COX-2, respectively [4]. Moreover, iNOS and COX-2 are mainly expressed by activated glial cells, both of which are also involved in neuronal inflammation and consequently contribute to the pathogenesis of CNS diseases. Thus, compounds that could supress the level of iNOS and COX-2 might be able to prevent and treat neuroinflammation-related CNS disorders. Pro-inflammatory cytokines, including TNF-α, IL-6, and IL-1β can promote neuronal damage, and also induce more microglia activation as feedback [23]. Since inflammatory cytokines play a pivotal role in inflammation, modulating cytokine function is key to controlling inflammation-induced CNS disorders [23]. Based on the inhibitory effects of StA on NO/iNOS, PGE₂/COX-2, pro-inflammatory cytokines, we suggested that StA was able to abate aberrant microglial activation and activated microglial-mediated neuroinflammation.

The NF-κB signalling pathway is well known to be a central regulator of several kinds of inflammatory responses, including microglial-mediated neuroinflammation [5]. Since we previously found that StA could abate pro-inflammatory responses, investigation of the NF-κB pathway might help us to explain the underlying mechanisms further. Firstly, the inhibitory activity of StA against NF-κB transduction was confirmed in reporter gene assay. Next, StA-induced suppression of NF-κB p65 phosphorylation was also found in LPS-stimulated BV-2 microglia. Furthermore, StA inhibited nuclear translocation of NF-κB and attenuated NF-κB-DNA binding activity in LPS-treated BV-2 microglia. All these results demonstrated StA can suppress NF-κB activation in BV-2 cells, consistent with previous reports showing that StA-mediated inhibition of NF-κB activity induced anti-inflammatory actions on LPS-stimulated RAW264.7 cells and IL-1β-stimulated human chondrocytes [17]. Furthermore, we then tested the regulatory action of StA on IKK and IκB, two important upstream modulators of NF-κB signal pathways [24]. We found that StA could block not only the phosphorylation of IKKα/β and IκBα, but also the degradation of IκBα in LPS-activated microglia. Therefore, we clearly confirmed inhibitory effects on the activation of IKK and IκB by StA, which results in the inhibition of NF-κB signal pathways.

Several classical signaling pathways, like the MAPKs and PI3K/Akt pathways, are involved in NF-κB-mediated neuroinflammation [6, 7, 23]. Activation of MAPKs and Akt can regulate pro-inflammatory responses in microglia by modulating NF-κB signal pathways [13]. MAPKs and PI3K/Akt are also reported to be significantly phosphorylated in LPS-stimulated microglia [4, 5, 19]. Previous reports also showed that StA could inhibit the phosphorylation of JNK, p38, ERK, and Akt in response to LPS stimulation in RAW264.7 cells and human chondrocytes [17]. As expected, in the current study, we found that StA could inhibit the phosphorylation of JNK, p38, and Akt induced by LPS in BV-2 microglia. Therefore, we speculated that StA might also suppress these key signal pathways in LPS-activated BV-2 microglial cells, thereby inhibiting NF-κB activity, and subsequently suppressing neuroinflammatory responses.

Accumulating evidence indicates that activated microglia are the major producer of ROS in the CNS [7, 10]. Intracellular ROS acts as an important second messenger in microglial-mediated inflammation, which could induce neuronal cell death, subsequently exacerbating disorders of the brain [13, 18]. The elevated ROS level can activate a cascade of harmful events during inflammatory processes, for exmaple via MAPKs, PI3K/Akt, and NF-κB pathways [13]. Moreover, the activation of these ROS-sensitive signaling pathways can further enhance the generation of ROS as a feedback response [13]. Therefore, inhibiting microglial ROS release can protect microglial and neuronal cells against neuroinflammatory damage [7, 10]. In this study, we found that StA could significantly attenuate both the production of microglial ROS and expression of p47 phox, a key component of NADPH oxidase [7, 10]. As expected, these findings were consistent with our previous reports that StA could inhibit intracellular ROS accumulation in neurotoxin-induced SH-SY5Y cells [22]. Taken together, these results suggest that StA-mediated inhibition of ROS production might be a major mechanism of StA-induced anti-neuroinflammation in LPS-activated microglial cells.

Encouraged by the significant inhibition of ROS production by StA in LPS-induced BV-2 cells, we further investigated the antioxidant properties of StA in microglia. Accumulating evidence demonstrates that inducing the activation of antioxidant transcription factor Nrf2, to activate the transcription of genes encoding for endogenous antioxidants, is a promising strategy for decreasing endogenous ROS [7, 9]. In response to oxidative stress, Nrf2 can be dissociated from Keap-1 in the cytosol, and is then translocated into the nucleus, binding to the ARE sequence to activate transcription of antioxidant genes. In current study, we firstly observed StA-induced up-regulation of HO-1 and NQO-1. Importantly, apart from serving as markers of Nrf2 activation, the elevated expression of HO-1 and NQO-1 was also reported to enhance cellular resistance to inflammatory responses. Moreover, we unexpectedly found that HO-1 inhibition obviously decreased the anti-inflammatory action of StA. Therefore, we suggested that StA-induced HO-1 expression was a key mechanism for the potential therapeutic property of StA. Also, we also confirmed the roel of StA on enchancing Nrf2 nuclear translocation, increasing ARE-dependent transcription activity and finally up-regulating antioxidant protein expression in LPS-activated BV-2 microglia. Therefore, all of current results totally and fully demonstrated the actions of StA on Nr2 induction. Since that Keap1 was well associated with Nrf2 nuclear translocation, investigations are required to explain the role of Keap1 in StA-mediated Nrf2 activation in our further study.

Recent studies demonstrate that Nrf2 activation contributes to the inhibition of microglial hyperactivation and protection against microgliosis-induced neuronal damage in the brain [7-9]. Nrf2 activation not only confers anti-oxidative effects but also modulates redox homoeostasis and regulates neuroinflammatory conditions in activated microglial cells [7, 9, 10, 25, 26]. In response to various microenvironments, microglia can polarise into pro- and anti-inflammatory phenotypes [9, 10]. Moreover, more remarkably, activation of Nrf2 pathways has been shown to promote microglial polarisation toward the anti-inflammatory phenotype, thus inhibiting neuroinflammation [9-11]. It is worth noting that there are several reported agents, including ellagic acid, lutein, curcumin, sulforaphane and dimethyl fumarate that show considerable anti-inflammatory effects via activating Nrf2 pathways in microglia cells [27-29]. On the other hand, Nrf2-deficiency is reported to be more important in inflammatory models [10]. Therefore, induction of Nrf2 signaling pathways by StA might be a valuable strategy for modulating neuroinflammation. In current study, via using the Nrf2 siRNA, we further found that StA-mediated inhibitory actions on LPS-induced pro-inflammatory mediators and cytokines were dependent on Nrf2. Microglial neuroinflammation is well characterised by an elevated level of these pro-inflammatory mediators and cytokines. Therefore, we concluded that StA-mediated Nfr2 induction contributed to its effects on neuroinflammatory responses in LPS-activated BV-2 microglia.

The cross-talk between Nrf2 and NF-κB pathways was well reported in previous studies [25, 26, 30]. Firstly, both the Nrf2 and NF-κB pathways coordinate in order to maintain cellular redox homoeostasis [25, 26]. However, phytochemical-induced activation of Nrf2 signaling might have a modulatory effect on activation of NF-κB during neuroinflammation [31]. Moreover, several molecules, as we described above, inhibited NF-κB activation but also activated Nrf2 pathways in microglia [25, 26]. Conversely, accumulating evidence suggests NF-κB activation might directly suppress Nrf2/ARE pathways at transcription levels [25, 26]. In current study, we also observed that, Nrf2 silence also suppressed StA-induced inhibitory effects on NF-κB activity and NF-κB-DNA binding. These observations suggested that StA-mediated activation of Nrf2 pathways might contribute to inactivation of NF-κB pathways by StA in LPS-stimulated BV-2 microglia cells. However, the detailed mechanisms underlying this hypothesis require further experiments.

There are two main mechanisms involved in the activation of Nrf2: modification of Keap-1 conformation and phosphorylation of Nrf2 [32]. Some compounds can induce a change in Keap-1 conformation via covalent modification, resulting in the release of Nrf2 from the Nrf2-Keap1 complex [32]. On the other hand, Nrf2 phosphorylation by protein kinases can alter the interaction between Nrf2 and Keap-1 and induce Nrf2 activation [32]. Several protein kinases, including ERK, p38 MAPK, AKT, and PKCδ were reported to be strongly associated with Nrf2 pathways [33]. Interestingly, we found that MEK 1/2 inhibitor U0126 was able to block StA-induced Nrf2 nucleus transactivation, ARE-luciferase activity and HO-1 expression, whereas p38 MAPK, AKT, and PKCδ inhibitors did not abolish StA’s actions. Combined with our previous finding that StA induced ERK phosphorylation, we initially considered ERK to be involved in StA-induced Nrf2 activation. ERK signalling is well known to regulate the cell cycle, proliferation, and differentiation; however, it has also been reported to modulate Nrf2-dependent transcription of ARE-related antioxidant genes, in vivo and in vitro [30]. ERK phosphorylation-mediated Nrf2 activation was also found in some well-studied compounds, like lutein [22] and berberine [5, 34], which could subsequently result in an anti-neuroinflammatory action and neuroprotection in microglial cells. ERK siRNA transfection effectively silenced StA-induced ERK phosphorylation, which totally abolished Nrf2 activation in StA-treated BV-2 microglial cells. Therefore, based on the current results regarding pharmacological inhibiton and gene silencing, we suggest that Nrf2 activation in StA-treated BV-2 cells was mediated by increased ERK phosphorylation by StA.

In summary, this research yielded several principal findings, as follows: (1) StA could inhibit neuroinflammatory responses in LPS-activated BV-2 microglial cells; (2) StA inhibited activation of NF-κB signaling pathways in LPS-activated BV-2 microglial cells; (3) StA induced ERK-mediated activation of Nrf2 antioxidant signaling pathways in LPS-activated BV-2 microglial cells; and (4) StA-induced anti-neuroinflammatory effects were dependent on ERK-mediated Nrf2 activation in BV-2 microglial cells.

The authors have no conflicts of interest to declare.

This study was supported by grants from The Science and Technology Development Fund (FDCT) of Macao SAR (No. 134/2014/A3 and No. 017/2015/AMJ) and the Research Committee of the University of Macau (MYRG2016-00129-ICMS-QRCM, 2016-00133-ICMS-QRCM, 2015-00214-ICMSQRCM, and 2015-00182-ICMS-QRCM).