Introduction

In addition to astrocyte activation [1–3], microglial activation is a histopathological hallmark of most if not all neuro-inflammatory and neuro-degenerative disorders such as AIDS dementia complex [4], multiple sclerosis [5], Parkinson’s and Alzheimer’s diseases [6, 7]. Upon activation, microglia cells release various pro-inflammatory cytokines and mediators such as nitric oxide (NO) and prostaglandins (PG) which are known to be implicated in neuroprotection [8]. On the other hand, activated microglial cells induce inflammatory responses that may cause neuronal damage [9]. Thus, it might be important to curtail excessive microglial activation [10]. Pro-inflammatory transcription factors such as nuclear factor κB (NF-κB) and activator protein-1 (AP-1) are activated by inflammatory cytokines such as tumor necrosis factor (TNF)-α, interleukin (IL)-1 or IL-6, which in turn positively influence their own expression by a positive feedback loop mechanism [11]. Cytokine-derived immune signalling is strictly regulated with respect to both magnitude and duration [12, 13].

Sulforaphane (SF) is a naturally occurring anticancer, antidiabetic, and antimicrobial compound mainly found in cruciferous vegetables [14–16]. The anticancer activity of SF is thought to be related to the induction of detoxifying or phase-II enzymes of xenobiotic transformation (such as quinone reductase and glutathione S-transferase), and enhancement of the transcription of tumor suppressor proteins, possibly via inhibitory effects on histone deacetylase. It has been shown that SF suppresses bacterial lipopolysaccharide (LPS)-mediated expression of inducible nitric oxide synthase (iNOS), COX-2, and IL-1β, TNF-α in RAW 264.7 and peritoneal macrophages [17–19] thus modulating inflammatory processes in the periphery. To what extend SF might also exert neuroprotective properties is far less investigated. Moreover, cellular interaction partners with SF have not been described so far.

Recent studies suggest that SF inhibits cytokine-mediated NFκB activation by blocking a signal leading to inhibitor protein κB (IκB) kinase activity in intestinal epithelial and mouse fibroblast cells. Furthermore, SF suppresses phorbol ester-induced c-Jun/AP-1 activation [20–22]. However, the mechanisms underlying interactions of SF with these signalling pathways are poorly understood. Although the molecular targets of this molecule are not completely characterized, the best known effect of SF is to induce nuclear factor erythroid 2-related factor 2 (Nrf2)-dependent gene expression [19]. Activation of Nrf2, a transcription factor implicated in neuroprotection, initiates or enhances the transcription of a battery of genes for anti-oxidative enzymes such as NAD(P)H:quinone oxidoreductase-1, thioredoxin reductase, glutathione peroxidase, and hemeoxygenase-1 [23–25].

In the present study, we investigated the influence of SF on the expression of the cytokines TNF-α, IL-1ß, and IL-6 in primary rat microglia cells after stimulation with LPS. Furthermore, the role of Nrf2 in this context was determined by Western blotting. We found that SF attenuates LPS-induced increase of IL-1β, IL-6, and TNF-α expression. In addition, SF significantly decreases the release of NO from LPS-stimulated microglia in a concentration-dependent manner. ERK1/2 and Jun-amino-terminal kinase (JNK) phosphorylation as well as NF-κB- and activator protein-1 (AP-1) activation seem to be implicated in this scenario. Moreover, we provide evidence that SF is able to activate the neuroprotective transcription factor Nrf2 in microglia cells.

Our results demonstrate that SF has antiphlogistic properties and interacts with brain resident inflammatory cells (microglia). These findings indicate that treatment of patients with SF or derivatives of SF might be considered as a novel therapeutic strategy to combat neurodegeneration and inflammation in the brain.

Materials and methods

Reagents

Sulforaphane (SF) and lipopolysaccharides (LPS) were obtained from Sigma-Aldrich (St. Louis, USA). WST-assay was obtained from Roche Diagnostics GmbH, Penzberg, Germany. Antibodies directed against pERK and ERK2 were obtained from Santa Cruz Biotechnology (CA, USA). Nrf2 antibodies were purchased from Abcam (UK), pJNK, NFκB p65 and Lamin A/C antibodies from Cell Signalling (Beverly, USA).

Cell culture

Microglia were collected as free-floating cells from primary cultures of neonatal astrocytes from Wistar rats and cultivated as described [26]. This culturing protocol has been demonstrated to yield microglia-enriched cell cultures with more than 98% OX42+-microglia and just few GFAP+-contaminating astrocytes. To test the cell purity, the cultures were stained with specific cell markers for astrocytes [glial fibrillary acidic protein (GFAP), an astrocyte marker; Sigma] and microglia (OX42; microglia/macrophage marker; Sera-Lab, Leicestershire, England) (data not shown). The murine microglia cell line BV2 was subcultivated in DMEM supplemented with 10% FCS and 1% penicillin/streptomycin (Carl Roth, Karlsruhe, Germany).

WST assay

For 4-[3-(4-Iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate (WST) assay, media were supplemented with 10 μl/well WST 2 hr before spectrophotometric evaluation. Conversion of WST to formazan was measured at 450 nm by microplate spectrophotometry (Model680; Bio-Rad, Hercules, CA, USA). This reaction reflects the reductive capacity of the cells, an indicator of cell viability. Results are expressed as percentage of control (±SEM) of the reductive capacity of untreated control cells.

RNA isolation and real time RT-PCR

Total RNA was isolated using the Trizol reagent (GIBCO BRL Life Technologies, Gaithersburg, USA) according to the manufacturer’s recommended protocol. RNA samples were reverse-transcribed by Moloney murine leukemia virus (MMLV) reverse transcriptase (Superscript RT; GIBCO BRL) and oligo-(dT)15 primers (Promega, USA). The cDNA products were used immediately for TaqMan real time RT-PCR using primers specific for IL-1β, IL-6 and TNF-α (Applied Biosystems, USA). Gene expression was monitored using the ABI Prism 7000 apparatus (Applied Biosystems, USA) applying a standardized protocol [27]. Relative quantification was performed using the ΔCt method which results in ratios between target genes and a housekeeping reference gene (18 s).

Western blotting

For Western blot analysis of MAP kinases phosphorylation, microglia cells were seeded in DMEM containing 10% FCS. Cells were harvested in a lysis buffer (50 mM Tris pH 7.5, 100 mM NaCl, 5 mM EDTA, 1% Triton, 2 mM sodium orthovanadate, 2.5 mM sodium pyrophosphate, 1 mM glycerol 2-phosphate, 1 mM phenylmethylsulfonylfluoride). Proteins (5 μg for pERK, pJNK, and ERK2) were resolved in SDS sample buffer and Western blotting procedure was performed as previously described in detail [28]. In brief, membranes were incubated with polyclonal primary antibodies against pERK1/2, pJNK, and ERK2 overnight at 4°C and subsequent detection was performed with peroxidase-labeled secondary antibodies. Antibody binding was detected via enhanced chemiluminescence (Amersham Pharmacia Biotech, Essex, UK). For Western blot analysis of NFκB p65 and Nrf2 translocation, the nuclear fraction of microglia cells was separated using the NE-PER kit purchased from Pierce, Rockford, IL (USA). Protein aliquots of lysate (15 μg for NFκB p65 and 20 μg for Nrf2 and reloaded with Lamin A/C as loading control) were detected using primary antibodies against NFκB p65, Lamin A/C and Nrf2.

Measurement of nitrite production

Microglia cells were treated with LPS (100 ng/ml) for 24 h. The generation of NO was determined by measuring the nitrite accumulation in the medium using a modified Griess reagent (1% sulphanilamide and 0.1% N-(naphthyl)-ethylenediamine dihydrochloride in 5% H3PO4; Sigma). The culture supernatant and Griess reagent were mixed and incubated for 5 min, and subsequently the absorption determined at 540 nm (SLT reader 340 ATTC). Sodium nitrite (NaNO2; Merck, Darmstadt, Germany) was used to generate a standard curve for quantification. Background nitrite was subtracted from the experimental value.

Luciferase assay

The vectors used in this study were constructed as follows: p(NFκB)3-Luc refers to a concatenated trimer of the wild-type sequence ATGTGGGATTTTCCCATG, while p(T5C)3-Luc corresponds to a three-time repeat with a T-for-C substitution at position 5 (underlined) of the IL-6 + NFκB motif ATGTGGGACTTTCCCATG (the core sequence is underlined).

5′-ATGTGGGATTTTCCCATGAGTGAGGGGACTTTCCCAGGCATGTGGGATTTTCCCATGG-3′ and 5′-CTAGCCATGGGAAAATCCCACATGCCTGGGAAAGTCCCCTCACTCATGGGAAAATCCCACATGGTAC-3′.

p5xAP-1-Luc refers to a concatenated pentamer of the AP-1 consensus sequence TGAGTCA.

All response elements were synthesized (Tib Molbiol, Berlin, Germany), annealed, and cloned at the Kpn1 and Nhe1 site of the pGL3-promoter (Promega, Madison, USA) to produce the reporter construct pNQO1-rARE. Then, 1.5 μg of the reporter plasmids containing the firefly luciferase reporter gene, and 0.5 μg of the pRL-TK plasmid, containing the Renilla reniformis luciferase gene under the control of the herpes simplex virus thymidine kinase promoter as an internal control, were cotransfected into cells in a 10-cm plate by the lipotransfection method (Lipofectamine 2000; Invitrogen, Carlsbad, USA) according to the manufacturer’s recommendation. Twenty-four hours after transfection, the cells were seeded to a 96-well plate. The activities of both firefly and R. reniformis luciferases were determined 48 h after transfection with the Dual-Luciferase Reporter Assay system (Promega, USA). The luciferase activities were normalized to the R. reniformis luciferase activity of the internal control.

Statistical analysis

All experiments were performed at least in triplicate and the values represent mean ± SEM. Significance of the difference between test and control groups was analyzed using ANOVA followed by the Bonferroni test. Data from the real time RT-PCR, densitometric quantification of Western blots, measurement of NO production and luciferase assay were analyzed using GraphPad Prism 3.0 software.

Results

Tolerability of Sulforaphane

In a first set of experiments, we exposed primary microglia cells to different concentrations of SF and LPS (applied alone or combined) to reveal information about their cell toxicity and to select a toxin concentration which would not result in cell death during the exposure period but might be high enough to cause an activation of microglia cells. After 24 h incubation with different SF concentrations (0.5–2 μM), the viability of cells was determined by a standardized WST assay. As shown in Fig. 1, the application of SF for 24 h in various concentrations did not affect the viability of microglia cells indicated by a stable metabolic activity. Furthermore, the application of LPS (100 ng/ml) alone or together with different SF concentrations did not affect cell viability. Based on these results, we chose SF at 1 μM in all further experiments.

Fig. 1
figure 1

Lack of effect of sulforaphane or LPS on cell viability in microglial cells. Sulforaphane (SF) and LPS together or alone were added to the cultures, which were then incubated for 24 h. Cell viability was determined using WST assay

Full size image

Sulforaphane inhibited pro-inflammatory cytokine mRNA expression in microglia cells

We next analyzed the influence of SF on pro-inflammatory cytokines’ expression of primary microglia cells after incubation with LPS with or without SF pre-treatment. Cells were treated with the endotoxin LPS to mimic bacterial inflammatory processes. Pre-treatment with SF was carried out for 30 min prior to LPS exposure. After preincubation with SF, the microglia cells were stimulated with LPS and SF simultaneously. Gene expression of IL-1β, IL-6, and TNF-α. was analyzed by the RT-PCR technique. As shown in Fig. 2, expression of TNF-α mRNA in non-stimulated microglia cultures was very low. LPS treatment induced strong proinflammatory cytokines IL-1β (a), IL-6 (b) and TNF-α (c) mRNA expression. The maximal increase of IL-1β and IL-6 transcripts was detected after 12 h of LPS treatment (279 ± 59- and 1,582 ± 90-fold expression increase, respectively). Peak TNF-α expression was measured after 6 h LPS treatment (139 ± 13-fold increase of expression). This demonstrates that the induction of TNF-α in microglia cells follows a different kinetic and regulation pattern. Pretreatment with SF (1 μM) 30 min prior to LPS application (6, 12, and 24 h) significantly (p < 0.001) attenuated TNF-α, IL-6, and IL-1β induction at the maximum points (IL-1β and IL-6 at 12 h, 8 ± 0.6- and 21 ± 2.8-fold, respectively; TNF-α at 6 h, 2.2 ± 0.5-fold) on the mRNA level in microglia cells.

Fig. 2
figure 2

Sulforaphane inhibited proinflammatory cytokine mRNA expression in primary rat microglia. For induction analysis of IL-1β (a), IL-6 (b) and TNF-α production (c), primary rat microglia were treated with 100 ng/ml LPS with and without SF and with SF alone for 0, 6, 12, and 24 h. mRNA expression was analyzed using TaqMan real time RT-PCR and results were compared to the untreated sample. 18 s (housekeeping gene) was used as an internal control. The data were assessed from three independent experiments in triplicate. An asterisks indicates a significant difference (*p < 0.05, **p < 0.001) between LPS and SF/LPS using ANOVA followed by the Bonferroni test

Full size image

LPS induced NO production was inhibited by sulforaphane in microglia

A hallmark of microglia activation is the production of Nitric oxide (NO) in response to pathogens. As a next step, we attempted to determine whether SF could modulate nitrite NO production in LPS activated microglial cells. As shown in Fig. 3, LPS induced a strong increase of NO concentration in the cell culture supernatants 24 h after stimulation (46.3 ± 3.2 μM nitrite) compared with unstimulated control cultures. The LPS-induced increase of nitrite was completely antagonized by SF (p < 0.001; 0.6 ± 0.1-fold). The single application of SF did not affect NO-production in microglial cells.

Fig. 3
figure 3

LPS induced NO production was inhibited by sulforaphane in microglia. Microglial cells were treated with 100 ng/ml LPS from Salmonella typhimurium with or without SF as well as with SF alone for 24 h at 37°C and nitric oxide (NO) concentration was determined. The values are the mean ± SEM from four separate measurements performed in duplicate. **Significant difference (p < 0.001) compared to the untreated control

Full size image

Sulforaphane inhibited LPS-induced ERK1/2 and JNK phosphorylation and NFκB signalling

To further investigate the mechanism of SF-mediated inhibition of LPS-induced microglia activation, we incubated microglia with LPS with and without SF pretreatment and determined the activation of different signalling pathways, namely ERK1/2 and JNK phosphorylation and translocation of NFκB p65 to the cell nucleus, by Western blotting. As shown in Fig. 4a, LPS treatment resulted in a pronounced ERK1/2 phosphorylation in microglia cells already after 5 min. After 30 min, ERK1/2 phosphorylation declined to basal control levels. Interestingly, this effect of LPS was almost completely inhibited by simultaneous intervention with SF. The single application of SF did not affect ERK1/2 phosphorylation. Further we examined the influence of SF treatment for LPS-induced JNK phosphorylation. As shown in Fig. 4b, LPS treatment resulted in a pronounced JNK phosphorylation in microglia cells after 5 min. SF blocked LPS-induced phosphorylation completely. SF alone did not affect JNK phosphorylation. After LPS stimulation the transcription factor NFκB p65 translocated into the nucleus in a time-dependent manner (Fig. 4c). The maximum translocation was detected after 1 h of treatment by Western blotting of the nuclear fraction. Treatment with SF completely blocked NFκB p65 translocation after 1 h of LPS stimulation. A single application of SF did not affect NFκB p65 translocation in microglia cells (Fig. 4c). The densities of the bands (pERK, pJNK and NFκB p65) and the loading control bands (ERK2 and Lamin A/C) were measured and the ratio calculated. Resultant ratios are presented as percentages of control in Fig. 4e, f.

Fig. 4
figure 4

Sulforaphane inhibited LPS-induced ERK1/2 and JNK phosphorylation and NFκB signalling. Microglial cells were treated with 100 ng/ml LPS from Salmonella typhimurium with and without SF as well as with SF alone for 5 and 30 min at 37°C. Phosphorylated ERK1/2 (a) or JNK (b) proteins were determined by immunoblotting. Microglial cells were treated with 100 ng/ml LPS with or without SF as well as with SF alone for 0, 30, 60, 120, and 240 min at 37°C. Translocation of NFκB p65 to the cell nucleus (c) was determined by immunoblotting. The mean ± SEM of the three independent experiments was evaluated by densitometric quantification (d–f). Asterisks indicate significant differences (*p < 0.05, **p < 0.001) between LPS and LPS + SF (one-way ANOVA followed by Bonferroni test)

Full size image

Inhibition of AP-1- and NFκB-mediated reporter activity by Sulforaphane

To compare validity of results obtained by NFκB Western blotting and using primary microglia cells, we additionally performed experiments using the murine microglia cell line BV2 to analyze activation of NFκB signalling pathways and additional AP-1 activation by dual luciferase assay. BV2 cells were transfected with a luciferase reporter construct harbouring NFκB (NFκB-Luc) or AP-1 (AP-1-Luc), respectively. As shown in Fig. 5a, b, LPS treatment induced a 3-fold increase of AP-1 activity (a) and a 5-fold increase in NFκB activity, respectively (b). Pre-treatment with SF (1 μM) resulted in a significant inhibition of LPS-induced (100 ng/ml) AP-1 and NFκB activity in BV2 cells, whereas SF (1 μM) alone showed no significant effect on AP-1 and NFκB activation (Fig. 5a, b).

Fig. 5
figure 5

Inhibition of LPS-induced AP-1- and NFκB-mediated reporter activity by sulforaphane. Cells of mice microglia line BV2 were transfected with (a) AP-1- or (b) NFκB-luciferase plasmid; luciferase activity was measured as described in “Materials and methods”. BV2 cells were treated with 100 ng/ml LPS from Salmonella typhimurium with and without SF as well as with SF alone (1 μM) for 6 h at 37°C and luciferase activity was determined. The values are the mean ±SEM from four separate measurements performed in duplicate. *Significant difference (p < 0.001) compared to the LPS (100 ng/ml) treated cells

Full size image

LPS- and Sulforaphane-induced Nrf2 protein expression

Encouraged by the observation that SF activates the antioxidant transcription factor Nrf2 by luciferase assay in BV2 cells [29], we analyzed SF-induced Nrf2 signalling by Western blotting. Primary rat microglia cells were exposed to SF to examine its effect on Nrf2 protein stability over time via Western blotting of the nuclear fraction. As shown in Fig. 6a, treatment with SF caused a significant time-dependent increase in Nrf2 protein content in nuclear extracts (max. at 3 h) which was still observed after 24 h. The densities of the Nrf2 band and the Lamin A/C band (loading control) were measured and the ratio calculated. Resultant ratios are presented as percentages of control in Fig. 6b.

Fig. 6
figure 6

Sulforaphane induced Nrf2 protein expression in microglia. Microglial cells were treated with SF for various times, and nuclear fractions were prepared as described in “Materials and methods”. a Nuclear proteins were separated on SDS–polyacrylamide gel electrophoresis, Western blotted, and probed with anti-Nrf2, and reprobed with anti-Lamin A/C (as loading control) antibodies. The positions of Nrf2 and Lamin A/C are indicated on the right. The positions of molecular mass markers are indicated on the left (in kDa). The mean ± SEM of the three independent experiments from (a) was evaluated using densitometric quantification (b). An asterisk indicates a significant difference (*p < 0.05, **p < 0.001) compared to controls using ANOVA followed by the Bonferroni test

Full size image

Discussion

SF, chemically defined as an isothiocyanate, is found in cruciferous vegetables, with particularly high levels detected in broccoli and brussels sprouts. Over a decade ago, SF was identified as a likely chemopreventive agent based on its ability to induce detoxification enzymes, as well as to inhibit enzymes involved in carcinogen activation [30]. Considerable attention has focused on the ability of SF to modulate the Nrf2/Keap1 pathway, but recent evidence suggests that SF acts by numerous other mechanisms. Moreover, pharmacological administration of SF may be a promising therapeutic approach for the treatment of inflammatory disorders, including those involving viral or bacterial-related pathologies.

To study the anti-inflammatory potential of SF in the brain, we used a well established model for neuroinflammation based on the stimulation of microglia with the endotoxin LPS to determine the influence of SF on LPS-induced microglia expression of certain cytokines, which, along with NO, are discussed in the context of neurodegeneration and neuro-inflammation. In our study we were able to show that SF attenuates LPS-induced increase of IL-1β, IL-6, and TNF-α expression in primary rat microglia. Furthermore, SF significantly decreases the release of NO from LPS-stimulated microglia in a concentration-dependent manner. Therefore, microglia might be an ideal interaction partner of SF in mediating the neuroprotective properties of SF.

Previous results have shown that SF attenuates LPS-mediated induction of iNOS, COX2 expression and TNF-α secretion in cultured raw macrophages [17]. In vivo administration of SF has resulted in decreased microglia cell activation and up-regulation of a set of inflammatory markers following endotoxin injection [29].

Heiss et al. [17] reported that SF possesses anti-inflammatory activity in raw macrophages due to inhibition on DNA binding of NFκB. Furthermore, SF had inhibitory effects on LPS-mediated retinal microglial activation processes, partly through a p38 MAPK-dependent mechanism. NFκB and MAPKs are two cellular pathways involved in LPS-mediated microglial activation [31]. Our findings suggest an involvement of NFκB in SF-mediated reduction of LPS-induced inflammation in primary rat microglia cells. Further yet, we were able to show that SF suppresses LPS activity via an inhibition of the ERK1/2 and JNK pathway. It was suggested that SF exerts its anti-inflammatory activity via activation of transcription factor nuclear factor erythroid 2-related factor 2 (Nrf2) [18]. SF can directly react with sulfhydryl groups, oxidize cysteine residues and deplete reduced cellular glutathione, thus mimicking an oxidative insult and releasing Nrf2 from Keap-1 binding [32]. It is generally accepted that Nrf2 plays a key role in the adaptive response to oxidative and electrophilic stress, maintaining the cellular self-defence. Nrf2 knockout mice were hypersensitive to the neuroinflammation induced by LPS compared with wild-type mice [29]. Disruption of Nrf2 resulted in a drastic increase in lethality during LPS-induced septic shock [33]. We saw a strong increase of Nrf2 translocation to the nucleus after SF treatment. Altogether, the observations supported that Nrf2 is involved in the anti-inflammatory activity of SF.

Consequently, SF down-regulates NFκB signalling pathways during inflammation. Besides NFκB, AP-1 is another LPS-induced pro-inflammatory transcription factor in peritoneal macrophages [34]. AP-1 is activated by MAPK [35]. Lin et al. [18] were able to show that SF displayed a moderate inhibition in LPS-stimulated Nrf2 knockout mice with peritoneal macrophages at high concentrations. They speculated on an impact of AP-1 activity which is not directly related to the Nrf2 pathway. In this context, our results show that SF inhibits AP-1 promoter activity in the murine microglia cell line BV2, using luciferase assay, and the inhibition resulted from the downregulation of ERK and JNK signal transduction.

This study was designed to discover nontoxic substances which exert anti-inflammatory as well as Nrf2-activating properties. The present data demonstrated that SF fulfils these requirements. Therefore, SF might be considered as an adjunct therapeutic strategy to combat neural demise in neurodegenerative diseases.