SEARCH

SEARCH BY CITATION

Keywords:

  • glial activation;
  • in vitro;
  • neuroinflammation;
  • neuroprotection;
  • transcription factor

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

J. Neurochem. (2010) 115, 526–536.

Abstract

The control of neuroinflammation is a potential target to be considered in the treatment of neurodegenerative diseases. It is therefore important to find anti-inflammatory drugs and study new targets that inhibit neuroinflammation. We designed an experimental model of neuroinflammation in vitro to study the anti-inflammatory and neuroprotective effects of the flavonoid chrysin and the involvement of nuclear factor-κB p65 and CCAAT/enhancer binding proteins (C/EBPs) β and δ transcription factors in its mechanism of action. We used primary cultures of mouse embryonic cortical neurons and cultures of BV2 (murine microglial cell line) or mouse primary microglia. We induced neuronal death in neuronal-BV2/microglial co-cultures using lipopolysaccharide of Escherichia coli and interferon-γ. Chrysin pre-treatment inhibited nitric oxide and tumor necrosis factor-α production, as well as inducible nitric oxide synthase expression in lipopolysaccharide E. coli and interferon-γ-treated microglial cells, but did not affect cyclooxygenase-2 expression. Chrysin pre-treatment also protected neurons against the neurotoxicity induced by reactive microglial cells. These effects were associated to a decrease in C/EBPδ protein level, mRNA expression, and DNA-binding activity, with no effect on C/EBPβ and p65 nuclear protein levels or DNA-binding activity, pointing out C/EBPδ as a possible mediator of chrysin effects. Consequently, C/EBPδ is a possible target to act against neuroinflammation in neurodegenerative processes.

Abbreviations used:
ABTS

2,3′-azino-bisethylbenzothiazoline-6-sulphonic acid

BSA

bovine serum albumin

C/EBP

CCAAT/enhancer binding protein

COX

cyclooxygenase

CREB

cAMP response element binding protein

DIV

days in vitro

FBS

foetal bovine serum

IFN

interferon

iNOS

inducible nitric oxide synthase

LPS

lipopolysaccharide

MAP2

microtubule-associated protein 2

NF-κB

nuclear factor-kappa B

NO

nitric oxide

PBS

phosphate-buffered saline

STAT

signal transducer and activator of transcription

TBS

Tris-buffered saline

TNF

tumor necrosis factor

Neuroinflammation, which mainly involves reactive glial cells, has been repeatedly reported to play a key role in the development of most neurodegenerative diseases. The purpose of reactive glia is to compensate for changes in the normal function of the CNS that could result in neuronal damage. However, although glial activation is initially beneficial, an excessive and prolonged glial reactivity can result in an inflammatory reaction with negative effects on neuronal cells. The presence of inflammation and the involvement of reactive glial cells has been described in Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, amyotrophic lateral sclerosis and multiple sclerosis (Jack et al. 2005; Sargsyan et al. 2005; Lucas et al. 2006; Lobsiger and Cleveland 2007; Whitton 2007; Farfara et al. 2008).

Reactive glial cells produce a series of neurotrophic compounds, but they also produce potentially neurotoxic agents. Pro-inflammatory cytokines, nitric oxide (NO) and reactive oxygen species have been suggested as mediators of the neurotoxic effect of reactive glia (Bal-Price and Brown 2001; Lee et al. 2004; Choi et al. 2005; Qian et al. 2007). In addition, several authors show that the presence of reactive glial cells increases the neuronal sensitivity to a potentially noxious stimulus (Qin et al. 2002; Pérez-Capote et al. 2004; Mander et al. 2005; Zhang et al. 2009).

The expression of genes involved in the production of potentially neurotoxic molecules in reactive glial cells is regulated by several families of transcription factors. Among them, nuclear factor-kappa B (NF-κB) is believed to play a key role, because it controls the expression of many pro-inflammatory genes (O’Neill and Kaltschmidt 1997; Saha and Pahan 2006; Tsatsanis et al. 2006). In recent years, it has been suggested that the CCAAT/enhancer binding protein (C/EBPs) family of transcription factors also plays a role in neuroinflammation. Binding sites for the C/EBPs in the regulatory regions of a battery of genes that are involved in the inflammatory response have been described (Poli 1998; Ramji and Foka 2002; Saha and Pahan 2006). In addition, alterations in C/EBPs expression have been detected in the brain of Alzheimer’s disease patients (Colangelo et al. 2002; Li et al. 2004). We and others have described alterations in C/EBPs in response to glial activation in astroglial and/or microglial cells (Cardinaux et al. 2000; Pahan et al. 2002; Pérez-Capote et al. 2006; Ejarque-Ortiz et al. 2007a,b, 2010).

Several studies show that the neuroprotective properties of numerous compounds can be attributed to the inhibition of glial activation both in vitro (Lee et al. 2004; Qian et al. 2007; Jiang et al. 2008) and in vivo (Lee et al. 2003; Choi et al. 2005; Pereira et al. 2006). In addition, results of epidemiological studies show that long-term use of anti-inflammatory drugs reduces the risk of developing Alzheimer’s and Parkinson’s disease and delays their onset (Szekely et al. 2004; Chen et al. 2005). All these results suggest that the inhibition of the inflammatory response induced by reactive glial cells could be of therapeutic interest. Consequently, the use of anti-inflammatory agents in the treatment of neurodegenerative processes has generated considerable interest in recent years (Townsend and Praticò 2005).

Various in vitro studies have shown that plant flavonoids have anti-inflammatory properties, and the neuroprotective action of some of these compounds on the neurotoxicity induced by reactive glial cells has also been described (revised in Vafeiadeou et al. 2007; Vauzour et al. 2008). In the present study, we investigate the anti-inflammatory and neuroprotective effects of chrysin (5,7-dihydroxyflavone), a natural flavonoid contained in many plant extracts, and the involvement of NF-κB, C/EBPβ and C/EBPδ in the mechanism of action of this compound. We use an in vitro experimental model of neuroinflammation in which we induce neuronal death in neuron-microglia co-cultures by microglia activation with an inflammatory stimulus, lipopolysaccharide of Escherichia coli and interferon-γ (LPS/IFN-γ).

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Cell cultures

BV2 cells

The mouse microglial cell line BV2 (generated from primary mouse microglia transfected with a v-raf/v-myc oncogene, Blasi et al. 1990) was cultured in RPMI-1640 medium (Invitrogen, Carlsbad, CA, USA), supplemented with 0.1% penicillin-streptomycin (Invitrogen) and 10% heat-inactivated foetal bovine serum (FBS, Invitrogen). Cells were maintained at 37°C in a 5% CO2 humidified atmosphere. For immunocytochemistry, nitrite assay and tumor necrosis factor (TNF)-α production, cells were seeded at a density of 5 × 104 cells/mL (1.6 × 104 cells/cm2), and for protein and RNA extraction at 105 cells/mL (2.4 × 106 cells/cm2).

Primary cultures

Experiments were carried out in accordance with the Guidelines of the European Union Council (86/609/EU) and following the Spanish regulations (BOE 67/8509-12, 1988) for the use of laboratory animals, and were approved by the Ethics and Scientific Committees from the Hospital Clínic de Barcelona.

Microglial cultures

Primary microglia-enriched cultures were obtained from primary mixed glial cultures from 2- to 4-day-old C57BL/6 mice. To obtain mixed glial cultures, cerebral cortices were dissected, carefully stripped of their meninges, and digested with 0.25% trypsin-EDTA solution (Invitrogen) for 25 min at 37°C. Trypsinization was stopped by adding an equal volume of culture medium, to which 0.02% deoxyribonuclease I (Sigma-Aldrich, St. Louis, MO, USA) was added. The culture medium consisted of Dulbecco’s modified Eagle medium-F-12 nutrient mixture (Invitrogen) supplemented with 10% FBS, 0.1% penicillin-streptomycin (Invitrogen), and 0.5 μg/mL amphotericin B (Fungizone®, Invitrogen). Cells were pelleted (5 min, 200 g), resuspended in culture medium, and brought to a single cell suspension by repeated pipetting followed by passing through a 105 μm-pore mesh. Cells were seeded at a density of 3.5 × 105 cells/mL (1.2 × 105 cells/cm2) and cultured at 37°C in a 5% CO2 humidified atmosphere. Medium was replaced every 5–7 days. Microglial cultures were prepared by the mild trypsinization method previously described in our group (Saura et al. 2003). Briefly, after 19–21 days in vitro (DIV) mixed glial cultures were treated for 30 min with 0.06% trypsin in the presence of 0.25 mM EDTA and 0.5 mM Ca2+. This resulted in the detachment of an intact layer of cells containing virtually all the astrocytes, leaving a population of firmly attached cells identified as > 98% microglia. The microglial cultures were treated 24 h after isolation by this procedure.

Neuronal cultures Primary cortical neuronal cultures were prepared from C57BL/6 mice at embryonic day 16 according to the method described by Frandsen and Schousboe (1990). Cells were seeded at a density of 8 × 105 cells/mL (2.6 × 105 cells/cm2) in poly-d-lysine hydrobromide (Sigma-Aldrich) coated 48-well culture plates and cultured at 37°C in a 5% CO2 humidified atmosphere. These cultures contained 76% neurons (positive for microtubule-associated protein 2, MAP2, immunostaining), 17% astrocytes (positive for glial fibrillary acidic protein immunostaining), < 1% microglial cells (positive for CD11b immunostaining), and 6% other cell types. These neuronal-glial cultures are called ‘neuronal cultures’ here. Neuronal cultures were used at 5 DIV.

Co-cultures

Neuronal-BV2 co-cultures.  BV2 cells growing in T75–T150 culture flasks were gently scrapped in neuronal culture medium, and aliquots of the cell suspension (50 μL/well) were seeded on the top of 5 DIV primary neuronal cells at a final density of 1.5 × 105 cells/mL (5 × 104 cells/cm2).

Neuronal-primary microglia co-cultures.  Microglial cultures were obtained as described above. After the isolation, microglia-enriched cultures were incubated with 0.25% trypsin for 10 min at 37°C. Trypsinization was stopped by adding the same volume of culture medium with 10% FBS. Cells were gently scraped and centrifuged for 5 min at 200 g. Pellet was resuspended in neuronal culture medium and aliquots of the cell suspension (50 μL/well) were seeded on the top of 5 DIV primary neuronal culture at a final density of 4 × 105 cells/mL (1.3 × 105 cells/cm2).

Treatments

BV2 cells. 1 day after seeding, the culture medium was replaced by fresh RPMI medium. One day later, cells were treated with 100 ng/mL LPS (Sigma-Aldrich, E. coli serotype 026:B6) and 0.5 ng/mL recombinant mouse IFN-γ (Sigma-Aldrich).

Mouse primary microglia. 1 day after isolation, microglia-enriched cultures were treated with 100 ng/mL LPS and 30 ng/mL IFN-γ.

Neuronal-BV2 co-cultures. 100 ng/mL LPS and 0.5 ng/mL IFN-γ were added to the culture medium 2 h after seeding BV2 cells on the top of neuronal cultures.

Neuronal-primary microglia co-cultures. 100 ng/mL LPS and 30 ng/mL IFN-γ were added to the culture medium 1 day after seeding primary microglial cells on top of neuronal cultures.

A stock solution of chrysin (Sigma-Aldrich) was prepared in dimethylsulfoxide and stored at −20°C. It was added to the culture media 1 h prior to LPS/IFN-γ treatment at a final concentration of 20 μM and dimethylsulfoxide < 0.1%.

Nitrite assay

NO production was assessed by Griess reaction. Briefly, culture supernatants were collected 24 h after LPS/IFN-γ treatment, and incubated with equal volumes of Griess reagent for 10 min at 23–25°C. Optical density at 540 nm was determined using a microplate reader (Multiskan spectrum, Thermo Electron Corporation, Waltham, CA, USA). Nitrite concentration was determined from a sodium nitrite standard curve.

TNF-α assay

The amount of TNF-α released into the culture medium was determined using an ELISA kit specific for mouse TNF-α (Mouse TNF-α ELISA Ready-SET-Go kit, eBioscience, San Diego, CA, USA), following the instructions supplied by the manufacturer. Culture supernatants were collected 24 h after LPS/IFN-γ treatment and stored at −80°C until assayed for TNF-α content.

Isolation of nuclear and total proteins

Nuclear protein and total protein extraction were performed as previously described (Ejarque-Ortiz et al. 2010). p65, C/EBPβ and C/EBPδ levels were determined in nuclear protein extracts from BV2 cells 4 h after treatments, using a T75 confluent flask (75 cm2) for each experimental condition. Inducible nitric oxide synthase (iNOS; EC 1.14.13.39) and cyclooxygenase-2 (COX-2; EC 1.14.99.1) expression were determined in total protein extracts from BV2 and primary microglial cells 12 h after treatments, using two or three wells from 6-well plates for BV2 cells or mouse primary microglia respectively for each experimental condition. Protein amount was determined by Lowry assay (Total Protein kit micro-Lowry, Sigma-Aldrich).

Western blot

Around 20–40 μg of protein, of denatured (100°C for 5 min) total or nuclear extracts, was subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis on a 7% (iNOS and COX-2) or 10% (p65, C/EBPβ and C/EBPδ) polyacrylamide gel, together with a molecular weight marker (Fullrange Rainbow Molecular Weight Marker, Amersham, Buckinghamshire, UK), and transferred to a polyvinylidene difluoride membrane (Millipore, Bedford, MA, USA). After washing in Tris-buffered saline (TBS: 20 mM Tris, 0.15 M NaCl, pH 7.5) for 5 min, dipping in methanol for 10 s and air dry, the membranes were incubated with primary antibodies overnight at 4°C: polyclonal rabbit anti-C/EBPδ (1 : 300, Santa Cruz Biotechnology, Temecula, CA, USA), polyclonal rabbit anti-C/EBPβ (1 : 500, Santa Cruz Biotechnology), polyclonal rabbit anti-p65 (1 : 500, Santa Cruz Biotechnology), polyclonal rabbit anti-COX-2 (1 : 2000, Santa Cruz Biotechnology), polyclonal rabbit anti-NOSII (1 : 200, Chemicon, Temecula, CA, USA), monoclonal mouse anti-β actin (1 : 40 000, Sigma-Aldrich) or polyclonal goat anti-lamin B (1 : 5000, Santa Cruz Biotechnology) diluted in immunoblot buffer (TBS containing 0.05% Tween-20 and 5% non-fat dry milk). Then, the membranes were washed twice in 0.05% Tween-20 in TBS for 15 s and incubated in horseradish peroxidase (HRP)-labelled secondary antibodies for 1 h at 23–25°C: donkey anti-rabbit (1 : 5000, Amersham), goat anti-mouse (1 : 5000, Santa Cruz Biotechnology) or mouse anti-goat/sheep (1 : 2000, Sigma-Aldrich). After extensive washes in 0.05% Tween-20 in TBS, they were incubated in ECL-Plus (Amersham) for 5 min. Membranes were then exposed to the camera of a VersaDoc System (Bio-Rad Laboratories, Hercules, CA, USA), and pixel intensities of the immunoreactive bands were quantified using the % adjusted volume feature of Quantity One 5.4.1 software (Bio-Rad Laboratories). Data are expressed as the ratio between the intensity of the protein of interest band and the loading control protein band (lamin B or β-actin).

Quantitative real-time PCR

C/EBPδ mRNA expression was determined in BV2 cells 3 h after treatments. For isolation of total RNA from BV2 cells, two wells from 6-well plates were used per experimental condition. Total RNA was isolated with High Pure RNA Isolation Kit (Roche Diagnostics, Mannheim, Germany) and 1.5 μg of RNA for each condition was reverse transcribed with random primers using Transcriptor Reverse Transcriptase (Roche Diagnostics). cDNA was diluted 1/5 to perform real-time PCR. The primers (Roche Diagnostics) used to amplify mouse C/EBPδ mRNA were 5′-CTCCACGACTCCTGCCATGT-3′ (forward) and 5′-GAAGAGGTCGGCGAAGAGTTC-3′ (reverse). For normalization of cycle threshold (Ct) values to an endogenous control, the following mouse actin mRNA primers were used: 5′-CAACGAGCGGTTCCGATG-3′ (forward) and 5′-GCCACAGGATTCCATACCCA-3′ (reverse). Real-time PCR was carried out with IQ SYBR Green SuperMix (Bio-Rad Laboratories), and iCycler MyIQ equipment (Bio-Rad Laboratories). Primer efficiency was estimated from standard curves generated by dilution of a cDNA pool. Samples were run for 50 cycles (95°C for 15 s, 60°C for 30 s, 72°C for 15 s). Relative gene expression values were calculated with the comparative Ct or ΔΔCt method (Livak and Schmittgen 2001) using iQ5 2.0 software (Bio-Rad Laboratories). Ct values were corrected by the amplification efficiency of the respective primer pair which was estimated from standard curves generated by dilution of a cDNA pool.

DNA binding activity assay

NF-κB p65, C/EBPβ and C/EBPδ DNA binding activity were determined using transcription factor ELISA-based TransAMTM assay kits (Active Motif, Carlsbad, CA, USA). A rabbit polyclonal anti-C/EBPδ antibody (1 : 500, Santa Cruz Biotechnology) was used to perform the C/EBPδ DNA-binding assay as no antibody to recognize C/EBPδ was provided with the TransAM kit. Nuclear protein extraction and DNA binding assays were performed following manufacturer’s instructions. Nuclear proteins were obtained 4 h after LPS/IFN-γ treatment and stored at −80°C until used.

Immunocytochemistry

Cultured cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 20 min at 23–25°C. For immunocytochemistry using fluorescence labelling, cells were permeated with chilled methanol for 8 min and non-specific staining was blocked by incubating cells with 10% normal goat serum (Vector, Burlingame, CA, USA) in PBS containing 1% bovine serum albumin (BSA) for 20 min at 23–25°C. Cells were then incubated overnight at 4°C with polyclonal rabbit anti-C/EBPδ (1 : 2000, Rockland, Gilbertsville, PA, USA) and/or monoclonal rat anti-CD11b (1 : 300, Serotec, Oxford, UK) primary antibodies. After rinsing in PBS, cells were incubated for 1 h at 23–25°C with goat anti-rabbit ALEXA 546 (1 : 1000) and/or goat anti-rat ALEXA 488 (1 : 500) secondary antibodies (Molecular Probes, Eugene, OR, USA).

For immunocytochemistry using peroxidase labelling, cells were permeated and endogenous peroxidase activity was blocked by incubation with 0.3% H2O2 in methanol for 10 min. Non-specific staining was blocked by incubating the cells with 10% normal goat serum in PBS containing 1% BSA for 20 min at 23–25°C. The cells were then incubated with monoclonal mouse anti-MAP2 primary antibody (1 : 2000, Sigma-Aldrich) overnight at 4°C, and with biotinylated horse anti-mouse secondary antibody (1 : 200, Vector) for 1 h at 23–25°C. Following incubation with ExtrAvidin-HRP (1 : 500, Sigma-Aldrich) for 1 h at 23–25°C, colour was developed with diaminobenzidine (Sigma-Aldrich). The antibodies were diluted in PBS containing 1% BSA and 10% normal horse serum (Vector).

Microscopy images were obtained with an Olympus IX70 microscope (Olympus, Okoya, Japan) and a digital camera (CC-12, Soft Imaging System GmbH, Munich, Germany).

Assessment of neuronal viability

Neuronal viability was evaluated by MAP2 immunostaining using ABTS (2,3′-azino-bisethylbenzothiazoline-6-sulphonic acid) and absorbance reading according to Brooke et al. (1999) with some modifications. Briefly, MAP2 staining was performed using peroxidase labelling as described above, but colour was developed using the ABTS Peroxidase Substrate Kit (Vector) following the manufacturer’s instructions. Two wells per experimental condition were processed and each experimental condition was repeated at least three times. One well for each experimental condition was also processed without primary antibody and used as background control: its average absorbance was subtracted from samples absorbance to calculate neuronal viability. Neuronal viability was expressed as a percentage of control levels.

Data presentation and statistical analysis

Results are presented as the means ± SEM. Statistical analyses were performed using repeated measures one-way analysis of variance (anova) followed by Newman–Keuls post-test when three or more experimental groups were compared. Values of p < 0.05 were considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Inflammatory phenotype in activated primary microglial cell cultures: inhibitory effect of chrysin

Microglial activation was induced with LPS/IFN-γ (LPS 100 ng/mL + IFN-γ 30 ng/mL, a treatment that in preliminary experiments induced neurotoxicity in neuronal-microglial co-cultures) and the presence of an inflammatory response was evaluated through the measure of NO production, TNF-α release, and iNOS and COX-2 expression. Primary microglial cells showed an increase in NO production (Fig. 1a) and TNF-α release (Fig. 1b) 24 h after LPS/IFN-γ treatment. We also observed an increase in iNOS and COX-2 protein expression 12 h after LPS/IFN-γ treatment (Fig. 1c–e). Chrysin pre-treatment inhibited NO production and TNF-α release induced by LPS/IFN-γ treatment (Fig. 1a and b). Chrysin pre-treatment also inhibited the induction of iNOS expression observed after LPS/IFN-γ treatment, while no effect on the induction of COX-2 expression was detected (Fig. 1c–e).

image

Figure 1.  Anti-inflammatory effects of chrysin pre-treatment in primary microglial cell cultures treated with LPS/IFN-γ. (a) NO and (b) TNF-α production, and (c) iNOS and (d) COX-2 protein expression in control, LPS/IFN-γ, chrysin (Ch) and Ch + LPS/IFN-γ treated primary microglial cells. Measurements were performed 24 h (NO and TNF-α) or 12 h (iNOS and COX-2) after LPS/IFN-γ treatment. Protein expression was measured by western blot and data normalized with β-actin. Bars are means ± SEM of three to four independent experiments. *p < 0.05, **p < 0.01 and ***p < 0.001 vs. control; #p < 0.05, ##p < 0.01 and ###p < 0.001 vs. LPS/IFN-γ; one-way anova (repeated measures) and Newman–Keuls post-test. (e) Images show representative western blots.

Download figure to PowerPoint

Neurotoxicity induced by reactive primary microglial cells: anti-inflammatory and neuroprotective effect of chrysin

As an experimental model of neuroinflammation in vitro, we considered co-cultures of primary cortical neurons and primary microglial cells and treated them with a concentration of LPS/IFN-γ inducing neuronal death. In control conditions, the addition of primary microglial cells to primary neuronal cultures (neurons : microglia at a ratio of 2 : 1) did not modify neuronal viability per se, as determined by the MAP2-ABTS-ELISA assay (data not shown). Neuronal cultures treated with LPS/IFN-γ for 48 h did not show alterations in MAP2 immunostaining (images not shown). However, MAP2 immunoreactivity showed clear alterations of the neuronal network in LPS/IFN-γ treated neuronal-microglial co-cultures, effect that was inhibited by chrysin pre-treatment (Fig. 2a–d). The evaluation of neuronal viability with the MAP2-ABTS-ELISA assay showed a significant decrease (62%) in the co-cultures 48 h after LPS/IFN-γ treatment, while no significant alterations were observed in the neuronal cultures after the same treatment (Fig. 2e). Chrysin pre-treatment provided almost total protection from the neurotoxicity induced by LPS/IFN-γ in neuronal-microglial co-cultures (Fig. 2e). The neuroprotective effect of chrysin pre-treatment observed in neuronal-microglial co-cultures after LPS/IFN-γ treatment occurred in the presence of a reduction in NO production (Fig. 2f) and a inhibition of TNF-α release (Fig. 2g). Chrysin pre-treatment did not have any effect on the morphological changes induced by LPS/IFN-γ treatment in microglial cells (Figure S1).

image

Figure 2.  Neuroprotective action of chrysin pre-treatment in neuronal-microglial co-cultures treated with LPS/IFN-γ. MAP2 immunostaining in control (a) and LPS/IFN-γ (b), chrysin (Ch) (c) and Ch + LPS/IFN-γ (d) treated co-cultures. MAP2 immunostaining was performed 48 h after LPS/IFN-γ treatment. Arrowheads in (a) point out MAP2 immunostaining in neuronal processes, which are abundant in (a), (c) and (d) and dramatically decreased in (b), what is taken as an index of neuronal damage/death. Bar = 100 μm. (e) Evaluation of neuronal viability 48 h after treatment by MAP2-ABTS-ELISA assay. Results are presented as % of MAP2 immunostaining in control cultures. (f) NO and (g) TNF-α production in control, LPS/IFN-γ and Ch + LPS/IFN-γ treated co-cultures. Measurements were performed 24 h after LPS/IFN-γ treatment. Bars are means ± SEM of three independent experiments. *p < 0.05, **p < 0.01 and ***p < 0.001 vs. control; #p < 0.05 and ##p < 0.01 vs. LPS/IFN-γ; one-way anova (repeated measures) and Newman–Keuls post-test.

Download figure to PowerPoint

Involvement of p65, C/EBPβ and C/EBPδ in the mechanism of action of chrysin

We studied the involvement of NF-κB and C/EBPs in the anti-inflammatory effect of chrysin in order to investigate the mechanism of action of the flavonoid, because these families of transcription factors play a relevant role in the control of the expression of several genes involved in the inflammatory response, such as iNOS, TNF-α and COX-2.

We determined the expression of p65, C/EBPβ and C/EBPδ in nuclear protein extracts of microglial cells in response to LPS/IFN-γ and the effect of chrysin in these expressions. We used the murine microglial cell line BV2 to perform these studies, given the special difficulty of obtaining significant amounts of nuclear protein using murine primary microglial cells. In preliminary studies, we validated the use of BV2 cells to perform these experiments. We observed that BV2 cells treated with LPS/IFN-γ (LPS 100 ng/mL + IFN-γ 0.5 ng/mL, a treatment that induced neurotoxicity in neuronal-BV2 co-cultures) showed an increase in NO and TNF-α production, as well as an increase in iNOS and COX-2 expression, effects that were inhibited by chrysin pre-treatment, with the exception of COX-2 expression (Figure S2). We also observed that chrysin pre-treatment resulted in a decreased neurotoxicity in LPS/IFN-γ treated neuronal-BV2 co-cultures (Figure S3). Therefore, BV2 cells qualitatively responded in the same way than primary microglial cells as regards the different parameters studied.

The nuclear expression of p65 was increased in BV2 cells 4 h after LPS/IFN-γ treatment, effect that was not modified by chrysin pre-treatment (Fig. 3a). LPS/IFN-γ treatment also resulted in an increase in C/EBPβ and C/EBPδ nuclear expression in BV2 cells at 4 h (Fig. 3b and c). However, although chrysin pre-treatment did not modify the increase in C/EBPβ expression induced by LPS/IFN-γ (Fig. 3b), it strongly inhibited the expression of C/EBPδ induced by the treatment (Fig. 3c). To discriminate between an effect of chrysin at the level of protein or at the level of gene expression, we determined C/EBPδ mRNA levels. We observed an induction of C/EBPδ mRNA expression 3 h after LPS/IFN-γ treatment, effect that was significantly inhibited by chrysin pre-treatment (Fig. 3d).

image

Figure 3.  Chrysin action on p65, C/EBPβ, and C/EBPδ expression in BV2 cells. Effect of chrysin (Ch) pre-treatment on the nuclear protein expression of (a) p65, (b) C/EBPβ, and (c) C/EBPδ in control cells and in cells treated for 4 h with LPS/IFN-γ. Protein expression was evaluated by western blot and data normalized with lamin B expression. Images show representative western blots. (d) Effect of Ch pre-treatment on C/EBPδ mRNA expression 3 h after LPS/IFN-γ treatment. mRNA expression was measured by real-time PCR and data normalized with actin expression. Bars represent mean ± SEM of three to four independent experiments. *p < 0.05, **p < 0.01 and ***p < 0.001 vs. control; ##p < 0.01 and ###p < 0.001 vs. LPS/IFN-γ; one-way anova (repeated measures) and Newman–Keuls post-test.

Download figure to PowerPoint

To further discard the involvement of p65 and C/EBPβ on the mechanism of action of chrysin, we studied whether this flavonoid could modify the binding activity of these transcription factors to their respective DNA consensus sequences. LPS/IFN-γ treatment resulted in a significant increase in the DNA binding of p65 at 4 h, effect that was not inhibited by chrysin pre-treatment (Fig. 4a). C/EBPβ DNA-binding activity also showed a trend to increase 4 h after LPS/IFN-γ treatment, and chrysin pre-treatment did not inhibit it (Fig. 4b). As regards C/EBPδ DNA-binding activity, LPS/IFN-γ treatment resulted in a significant increase at 4 h, effect that was inhibited by chrysin pre-treatment (Fig. 4c). The specificity of p65, C/EBPβ and C/EBPδ DNA binding activities were proved in competition assays incubating nuclear protein extracts with mutated and wild-type competitor DNA consensus sequences (data not shown).

image

Figure 4.  Chrysin action on (a) p65, (b) C/EBPβ and (c) C/EBPδ activity. DNA binding was evaluated using a TransAM DNA binding determination kit. Bars represent mean ± SEM of three to five independent experiments. *p < 0.05 and **p < 0.01, ***p < 0.001 vs. control; #p < 0.05 vs. LPS/IFN-γ; one-way anova (repeated measures) and Newman–Keuls post-test.

Download figure to PowerPoint

Involvement of C/EBPδ in the neuroprotective action of chrysin

We finally studied the expression of C/EBPδ in neuronal-microglial co-cultures by immunocytochemistry, in order to confirm that the neuroprotective effect of chrysin pre-treatment in LPS/IFN-γ treated co-cultures occurred in the presence of an inhibition in C/EBPδ nuclear expression. We observed an increase in C/EBPδ immunostaining in the nuclei of microglial cells in LPS/IFN-γ treated co-cultures, effect that was inhibited by chrysin pre-treatment (Fig. 5). This effect occurred in the absence of any effect of chrysin pre-treatment on p65 and C/EBPβ immunoreactivity in the nuclei of microglial cells in LPS/IFN-γ treated co-cultures (data not shown).

image

Figure 5.  Effect of chrysin pre-treatment on microglial C/EBPδ expression in neuronal-primary microglia co-cultures. Double immunostaining for C/EBPδ (red) and CD11b (green) in control (a) and LPS/IFN-γ (b), chrysin (Ch) (c), and Ch + LPS/IFN-γ (d) treated co-cultures. Immunostaining was performed 4 h after LPS/IFN-γ treatment. Arrowheads point out C/EBPδ immunostaining in the nuclei of microglial cells, which are immunolabelled with the cell surface marker CD11b. Notice the increase in C/EBPδ immunostaining in LPS/IFN-γ treated microglial cells, effect that is inhibited by chrysin pre-treatment. Bar = 30 μm.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The present results show that the inflammatory pattern of reactive glial cells and its resulting neurotoxicity can be attenuated by chrysin. We suggest that these effects are mediated, at least in part, by the inhibitory effect of chrysin in microglial C/EBPδ expression.

We describe both an anti-inflammatory and a neuroprotective action of chrysin in our experimental model. Chrysin pre-treatment partially inhibited the inflammatory pattern induced by LPS/IFN-γ in microglial cells, inducing a decrease in NO production that could be because of the decreased expression of iNOS that we also observed in the presence of chrysin. In addition, we observed that chrysin pre-treatment inhibited TNF-α release induced by LPS/IFN-γ. However, chrysin inhibition of TNF-α release was transient, because it was not detected when TNF-α levels were determined 48 h after LPS/IFN-γ, while NO levels were still inhibited at 48 h of LPS/IFN-γ treatment in the presence of chrysin (data not shown). Both NO and TNF-α have been shown to mediate the neurotoxic effects of reactive glial cells in vitro, and the inhibition of their production has been shown to protect against the neurotoxicity induced by reactive glial cells (Bal-Price and Brown 2001; Lee et al. 2004). Both iNOS and TNF-α genes, as well as COX-2 gene, have binding sites for NF-κB and C/EBPs on their promoters (Ramji and Foka 2002; Saha and Pahan 2006; Tsatsanis et al. 2006). The decreased expression of C/EBPδ observed in LPS/IFN-γ treated microglial cells in the presence of chrysin pre-treatment could explain the inhibition of NO and TNF-α production, which in turn could account for chrysin’s neuroprotective effect. The absence of any effect of chrysin on COX-2 expression and its transient effect on TNF-α production in the presence of a maintained effect on NO production could be explained by the presence of different contributors in the final regulation of COX-2, TNF-α and iNOS genes under inflammatory stimuli. These effects occurred in the absence of alterations on p65 or C/EBPβ expression, stressing an important role for C/EBPδ in the regulation of the inflammatory response in microglial cells. According to our results, C/EBPδ may play a major role in iNOS regulation, and in the TNF-α initial response, but is not essential for COX-2 production in this model. Although several authors have studied the role of C/EBPβ and C/EBPδ in the regulation of COX-2, the different contribution of these transcription factors in this process is not fully understood. In macrophages, Wadleigh et al. (2000) suggest that C/EBPδ is necessary for both basal and LPS-induced expression of COX-2 mRNA and C/EBPβ for the expression of COX-2 in response to LPS induction. Caivano et al. (2001) suggest a role for C/EBPβ for the initiation and the maintenance COX-2 mRNA expression in macrophages in response to LPS and a role for C/EBPδ for the maintenance of the response. In a previous study, we described that C/EBPδ binds to COX-2 promoter in BV2 cells in the presence of LPS but not in basal conditions, and that C/EBPδ forms heterodimers with C/EBPβ [with both the inhibitory isoform liver-enriched inhibitory protein (LIP) and the activating isoform liver-enriched activating protein (LAP)] in response to LPS (Ejarque-Ortiz et al. 2010). In our experimental model, the lack of alterations in C/EBPβ expression induced by LPS/IFN-γ in the presence of chrysin may be responsible for the lack of effect of chrysin in COX-2 expression.

Several authors have previously reported an anti-inflammatory effect of chrysin or other flavonoids, using LPS-treated RAW 264.7 macrophages (Liang et al. 2001; Cho et al. 2004; Woo et al. 2005; Harris et al. 2006) or LPS or LPS/IFN-γ-treated BV2 or primary microglial cells (Suk et al. 2003; Chen et al. 2004). In addition, flavonoids such as silymarin, baicalein and fisetin have been shown to have neuroprotective effects on neurotoxicity induced by activated glial cells (Wang et al. 2002; Li et al. 2005; Zheng et al. 2008). Although different targets have been described for the anti-inflammatory action of several flavonoids, the primary mechanism of action of these compounds has not been established; some have been shown to induce different degrees of inhibition of the DNA binding for C/EBPs, NF-κB or activator protein-1 transcription factors (Suk et al. 2003; Chen et al. 2004; Woo et al. 2005, 2006). As for chrysin, Liang et al. (2001) showed that it activates peroxisome proliferator-activated receptor-γ in LPS-treated RAW 264.7 cells through an allosteric interaction. In addition, Woo et al. (2005) showed that chrysin inhibits C/EBPs DNA binding and increases NF-κB binding in activated RAW 264.7 cells, in the absence of any effect on cAMP response element binding protein (CREB) DNA binding. We found an action of chrysin on both C/EBPδ expression at the protein and mRNA level and C/EBPδ DNA-binding activity. C/EBPδ forms heterodimers with C/EBPβ, and it has been suggested that members of the C/EBP family of transcription factors constitute heterodimers with members of other families of transcription factors, such as NF-κB, CREB and activator protein-1 (revised in Ramji and Foka 2002). Consequently, the decreased expression of C/EBPδ we have observed under the action of chrysin could result in secondary effects on the DNA binding of the heterodimers on different DNA binding sites, in addition to its effect on the C/EBPs DNA binding site.

The relatively rapid inhibition of C/EBPδ mRNA expression induced by chrysin in LPS/IFN-γ treated microglial cells suggests a direct effect of the flavonoid in the C/EBPδ regulation pathway. The C/EBPδ promoter contains binding sites for CREB, NF-κB and signal transducer and activator of transcription (STAT) 3 for an early transcriptional response (Liu et al. 2007) as well as autoregulatory C/EBP binding sites for a late transcriptional response (Yamada et al. 1998). Sp1 transcription factor can also bind the C/EBPδ promoter, but its regulation is constitutive and it is more related to the basal expression of C/EBPδ protein (Zhang et al. 2007). As noted above, Woo et al. (2005) did not detect any effect of chrysin in CREB DNA binding in LPS-treated RAW 264.7 cells. Although these authors also showed that chrysin increases NF-κB DNA binding in LPS-treated RAW 264.7 cells, we did not detect any direct effect of chrysin on the nuclear expression or DNA-binding of the main NF-κB family member, p65. No effect of chrysin on STAT3 has been reported until now, but STAT3 could be a candidate to mediate the inhibitory effect of chrysin on C/EBPδ expression: an inhibition of the early C/EBPδ transcription could occur as a consequence of an inhibition of the Janus kinase/STAT pathway, and the resulting decrease in C/EBPδ expression could in turn inhibit a further increase in C/EBPδ expression because of the autoregulatory effect of the late transcriptional response. In addition, chrysin has been recently reported to interact with the aryl hydrocarbon receptor, a ligand-activated transcription factor suggested being involved in the inhibition of the inflammatory response (Sekine et al. 2009). Chrysin acts both as agonist and antagonist of this receptor, depending on the cell type and the experimental situation (Zhang et al. 2003; Van der Heiden et al. 2009). Although the mechanism of action of aryl hydrocarbon receptor in the control of the expression of inflammatory genes in microglial cells is not clear, C/EBPδ gene expression may be a potential candidate to be regulated by this transcription factor.

In summary, chrysin inhibits LPS/IFN-γ induced iNOS expression and NO and TNF-α production in microglial cells, as well as the neurotoxicity induced by LPS/IFN-γ activated microglial cells. These anti-inflammatory and neuroprotective effects are mediated, at least in part, by the inhibition of C/EBPδ expression at the level of both protein and mRNA expression, and C/EBPδ DNA-binding activity. Further studies are necessary to understand the mechanism of action of chrysin in C/EBPδ expression and its possible role at the level of other transcription factors involved in neuroinflammation. In addition, further studies are also necessary to understand the implication of C/EBPδ in neuroinflammation. In this context, C/EBPδ deficient mice will be a useful tool, and preliminary studies in our laboratory using these mice show that C/EBPδ deficient primary microglial cultures produce less NO and TNF-α in response to LPS/IFN-γ than wild-type cultures (Valente et al., unpublished observations), suggesting that the absence of C/EBPδ results in alterations in microglial activation. In general, few effective treatments are currently available to inhibit the progress of neurodegenerative diseases. The great complexity of these diseases suggests that a combination of different treatments may be needed to obtain better results. Neuroinflammation is a common factor in most neurodegenerative processes occurring in the human brain and a possible target in the treatment of these diseases. However, given the fact that neuroinflammation is also a complex process, different approaches can be considered to inhibit it. Our results show that the neurotoxicity induced by reactive microglial cells is prevented when C/EBPδ is inhibited in these cells. Thus, C/EBPδ is an additional target to be considered in the search for drugs inducing anti-inflammatory and neuroprotective effects to be used in therapy to control the progress of neurodegenerative processes occurring in the presence of neuroinflammation.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Núria Gresa-Arribas is the recipient of a JAE contract from CSIC. This study was supported by grants PI070455 and PI081396 from the Instituto Carlos III of the Ministerio de Ciencia e Innovación of Spain, and grant V-2006-Tv063031-O from Fundació La Marató de TV3. We acknowledge Marco Straccia for technical advice in real-time PCR technique. The authors declare no conflict of interest.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Figure S1. Morphology of microglial cells in neuron-microglia co-cultures. CD11b immunostaining in control (a) and LPS/IFN-γ (b), chrysin (Ch) (c) and Ch p;+ p;LPS/IFN-γ (d) treated co-cultures. CD11b immunostaining was performed 48 p;h after LPS/IFN-γ treatment. Insets in (a) and (b) show microglial cells at a higher magnification. Control microglia show a weak labelling in most of the cell surface, and a strong labelling at the end of some processes. Notice the morphological change and the increase in CD11b immunostaining in most of the microglial cell surface in LPS/IFN-γ-treated co-cultures, both in the absence and in the presence of chrysin pre-treatment. Bars p;= p;20 p;μm (insets) and 50 p;μm.

Figure S2. Anti-inflammatory effects of chrysin in BV2 cells. Effect of chrysin (Ch) pre-treatment on (a) NO and (b) TNF-α production, and (c) iNOS and (d) COX-2 protein expression in control and in LPS/IFN-γ, chrysin (Ch) and Ch p;+ p;LPS/IFN-γ treated cells. Measurements were performed after 24 p;h (NO and TNF-α production) or 12 p;h (iNOS and COX-2 expression) of LPS/IFN-γ treatment. Protein expression was determined by western blot and data normalized with ß-actin expression. Bars represent means p;± p;SEM of three to four independent experiments. *p p;< p;0.05, **p p;< p;0.01 and ***p p;< p;0.001 vs. control; #p p;< p;0.05; ##p p;< p;0.01 vs. LPS/IFN-γ; one-way anova (repeated measures) and Newman–Keuls post-test. (e) Images show representative western blots.

Figure S3. Neuroprotective action of chrysin pre-treatment in neuronal-BV2 co-cultures treated with LPS/IFN-γ. MAP2 immunostaining in control (a) and LPS/IFN-γ (b), chrysin (Ch) (c) and Ch p;+ p;LPS/IFN-γ (d) treated co-cultures. MAP2 immunostaining was performed 24 p;h after treatment. Arrowheads in (a) point out MAP2 immunostaining in neuronal processes, which are abundant in (a), (c) and (d) and dramatically decreased in (b), what is taken as an index of neuronal damage/death. Bar p;= p;100 p;μm. (e) Evaluation of neuronal viability 24 p;h after treatment by MAP2-ABTS-ELISA assay. Results are presented as % of MAP2 immunostaining in control cultures. (f) NO and (g) TNF-α production in control, LPS/IFN-γ, and Ch p;+ p;LPS/IFN-γ treated co-cultures. Bars represent means p;± p;SEM of four independent experiments. *p p;< p;0.05, **p p;< p;0.01 and ***p p;< p;0.001 vs. control; #p p;< p;0.05, ##p p;< p;0.01 and ###p p;< p;0.001 vs. LPS/IFN-γ; one-way anova (repeated measures) and Newman–Keuls post-test.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JNC_6952_sm_FigS1.tif2260KSupporting info item
JNC_6952_sm_FigS2.tif2008KSupporting info item
JNC_6952_sm_FigS3.tif8727KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.