SEARCH

SEARCH BY CITATION

Keywords:

  • alcohol;
  • astrocytes;
  • rafts;
  • toll-like receptors

Abstract

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

We have recently reported that ethanol-induced inflammatory processes in the brain and glial cells are mediated via the activation of interleukin-1 beta receptor type I (IL-1RI)/toll-like receptor type 4 (TLR4) signalling. The mechanism(s) by which ethanol activates these receptors in astroglial cells remains unknown. Recently, plasma membrane microdomains, lipid rafts, have been identified as platforms for receptor signalling and, in astrocytes, rafts/caveolae constitute an important integrators of signal events and trafficking. Here we show that stimulation of astrocytes with IL-1β, lipopolysaccharide or ethanol (10 and 50 mM), triggers the translocation of IL-1RI and/or TLR4 into lipid rafts caveolae-enriched fractions, promoting the recruitment of signalling molecules (phospho-IL-1R-associated kinase and phospho-extracellular regulated-kinase) into these microdomains. With confocal microscopy, we further demonstrate that IL-1RI is internalized by caveolar endocytosis via enlarged caveosomes organelles upon IL-1β or ethanol treatment, which sorted their IL-1RI cargo into the endoplasmic reticulum–Golgi compartment and into the nucleus of astrocytes. In short, our findings demonstrate that rafts/caveolae are critical for IL-1RI and TLR4 signalling in astrocytes, and reveal a novel mechanism by which ethanol, by interacting with lipid rafts caveolae, promotes IL-1RI and TLR4 receptors recruitment, triggering their endocytosis via caveosomes and downstream signalling stimulation. These results suggest that TLRs receptors are important targets of ethanol-induced inflammatory damage in the brain.

Abbreviations used
cav-1

caveolin-1

Ct-B

cholera toxin subunit B

ER

endoplasmic reticulum

EtOH

effect of ethanol

FBS

foetal bovine serum

GAPDH

glyceraldehyde-3-phosphate dehydrogenase

HRP

horseradish peroxidase

IL-1RI

interleukin-1 beta receptor type I

LAMP-1

lysosome-associated membrane protein 1

LPS

lipopolysaccharide

MyD88

myeloid differentiation primary-response protein 88

MβCD

methyl-β-cyclodextrin

PBS

phosphate-buffered saline

P-ERK

phospho-extracellular regulated-kinase

P-IRAK

phospho-IL-1R-associated kinase

SDS

sodium dodecyl sulphate

TLR4

toll-like receptor type 4

TRITC

tetramethylrhodamine isothiocyanate

Ethanol is known to affect the innate immune response in several organ systems (MacGregor 1986; Nagy 2003). Nevertheless, little is known about the potential action of ethanol on the CNS immune system, or how inflammation participates in alcohol-induced brain damage. Our recent studies demonstrate that astrocytes respond to ethanol by secreting cytokines and other inflammatory mediators (Blanco et al. 2004, 2005; Valles et al. 2004), and by contributing to an inflammatory environment in the brain of alcohol-fed animals (Valles et al. 2004). These effects seem to be mediated by the ethanol-induced activation of interleukin-1 beta receptor type I (IL-1RI), and by toll-like receptor type 4 (TLR4) signal-transduction pathways, as blocking these receptors with neutralizing antibodies abolishes most inflammatory signal events and prevents cell death (Blanco et al. 2005).

Interleukin-1 beta receptor type I and TLR4, the specific receptors of IL-1β and bacterial lipopolysaccharide (LPS), are members of a defined group of receptors which shares a cytoplasmic toll/IL-1 receptor domain, that participates in host responses to injury and infection (O’Neill 2000; Akira and Sato 2003). Activation of TLRs/IL-1Rs leads to the recruitment of specific adaptor molecules into the receptor complex, including IL-1R-associated kinase (IRAK), the adaptor molecule myeloid differentiation primary-response protein 88 (MyD88), and tumour-necrosis factor-receptor-associated factor 6 (Martin and Wesche 2002; Akira and Takeda 2004), which trigger the downstream stimulation of protein kinases (MAPK and stress-activated protein kinases/c-Jun N-terminal kinase), that ultimately lead to the activation of transcription factors, such as nuclear factor-κB and activator protein-1 (Martin and Wesche 2002; Akira and Takeda 2004).

Increasing evidence indicates the role of lipid rafts, cholesterol/sphingomyelin-enriched membrane microdomains, in immune system activation (Manes et al. 2003). The recruitment of TLR4 and other adaptor proteins, such as CD14, into lipid rafts has been observed upon LPS stimulation (Triantafilou et al. 2002). We have recently proposed that the effects of ethanol on the immune system and on TLRs receptors are, in part, mediated by the interaction of ethanol with lipid rafts (Blanco et al. 2005; Blanco and Guerri 2007). Indeed, high ethanol concentration seems to disrupt lipid rafts clustering, leading to the suppression of TLR4 signalling (Dai et al. 2005; Dolganiuc et al. 2006). However, low ethanol concentrations (10–50 mM) might facilitate protein–protein and protein–lipid interaction within the membrane microdomains to promote receptor recruitment into the lipid rafts, and to allow for IL-1RI/TLR4 activation and signalling.

Membrane rafts are specialized signalling platforms which are implicated in protein sorting, membrane trafficking and signal-transduction events (Pelkmans and Helenius 2002; Parton and Richards 2003; Pike 2003; Pelkmans et al. 2004). Different subtypes of lipid rafts have been described according to their protein and lipid composition. Caveolae are a major subclass of rafts enriched in the caveolin-1 (cav-1) protein, which are a specialized form of flask-shaped invaginations involved in endocytosis and transcytosis in many eukaryotic cell types (Nichols and Lippincott-Schwartz 2001; Pelkmans and Helenius 2002; Pelkmans et al. 2004). The endocytosed caveolar vesicles accumulate in a pre-existing population of cav-1-containing endosomes, termed caveosomes (Pelkmans et al. 2001). These structures serve as an intermediate station during the internalization of endocytosed ligands in the caveolar/raft endocytic pathway, and have some properties of a sorting compartment, delivering their cargo to the endoplasmic reticulum (ER)–Golgi apparatus, and also in the cellular nucleus (Pelkmans et al. 2001; Rajendran and Simons 2005). When compared with the clathrin-mediated pathway however, the non-clathrin endocytic pathways are poorly understood, and the emergence of caveosomes as a new organelle raises the question of the functional significance of uptake by a caveolar mechanism as opposed to clathrin-coated tips.

The present study was undertaken to test the hypothesis that ethanol-induced activation of IL-1RI/TLR4 in astroglial cells, is mediated by promoting receptors recruitment into lipid rafts caveolae, leading to receptor internalization and signalling. Here we show that astrocyte stimulation with ethanol (10 mM), IL-1β or LPS, induces a rapid translocation of IL-1RI and TLR4 into lipid rafts cav-rich fractions, as well as the recruitment and activation of signalling molecules [P-IRAK and phospho-extracellular regulated-kinase (P-ERK)] into these microdomains. Disruption of lipid rafts with nystatin, saponin or methyl-β-cyclodextrin (MβCD) abolishes the expression and activation of both receptors, suggesting the role of lipid rafts caveolae in IL-1RI and TLR4 signalling. Immunofluorescence studies and confocal microscopy further revealed that, in either ethanol or IL-1β-stimulated astrocytes, it is observed a rapid internalization of IL-1RI into enlarged cytoplasmic ring structures. These structures are positive for cholera toxin subunit B (Ct-B) (raft/caveolae marker) and cav-1 (caveosomes marker), and negative for the early/late endosomes and lysosomes, suggesting that ethanol- or IL-1β-induced internalization of IL-1RI in astrocytes occurs via the caveolar endocytic pathway.

Materials and methods

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

Antibodies and reagents

The antibodies used for western blotting were: anti-TLR4, anti-IL-1RI, anti-flotillin-1, anti-P-ERK1/2, anti-ERK2, anti-MyD88, anti-IRAK1 (Santa Cruz Biotechnology, Madrid, Spain), anti-cav-1-rabbit (Sigma-Aldrich, Madrid, Spain), anti-P-IRAK1 (Transduction Laboratories; Becton-Dickinson, Madrid, Spain), anti-GAPDH (Chemicon, Temecula, CA, USA), as well as horseradish peroxidase (HRP)-conjugated secondary antibodies against mouse (Santa Cruz Biotechnology), rabbit and goat (Sigma-Aldrich).

To visualize lipid rafts (GM-1 ganglioside), caveosomes, Golgi apparatus, ER and clathrin-dependent late endosomes, astrocytes were stained with Alexa-Fluor 488 conjugate of Ct-B (Molecular Probes, Eugene, OR, USA), anti-cav-1-mouse (Abcam, Ltd, Cambrigde, UK), BODIPY TR C5-ceramide, ER-Tracker Red dye (Molecular Probes) and anti-lysosome-associated membrane protein 1 (LAMP-1) (Abcam, Ltd) respectively. FITC-conjugated donkey anti-mouse antibodies (Jackson Immuno-Research, West Grove, PA, USA) were used as fluorescent secondary antibodies of cav-1 and LAMP-1. FITC and rhodamine [tetramethylrhodamine isothiocyanate (TRITC)]-conjugated goat anti-rabbit antibodies (Jackson Immuno-Research) were used to label IL-1RI.

Recombinant rat IL-1β, LPS from Escherichia Coli O26:B6 (LPS), nystatin dehydrate, saponin from Quillaja bark purified and MβCD, were obtained from Sigma-Aldrich.

Culture of astrocytes and experimental protocol

Primary cultures of rat cortical astrocytes were prepared from the cerebral cortex of 21-day-old rat foetuses, as previously described (Guerri et al. 1990). Cells were plated on 30 cm2 plates in Dulbecco’s modified Eagle medium containing 20% foetal bovine serum (FBS), supplemented with l-glutamine (1%), HEPES (10 mM), Fungizone (1%) and antibiotics (1%). Cultures were grown in a humidified atmosphere of 5% CO2/95% air at 37°C. After 1 week of culture, the FBS content was reduced to 10% and the medium was changed twice a week. Cells were grown to confluence and were used after 12 days in culture. The purity of astrocytes was assessed by immunofluorescence using: anti-glial fibrillary acidic protein (astrocyte marker; Sigma-Aldrich), anti-CD-11b (microglial marker, Serotec; Bionova, Madrid, Spain), anti-myelin basic protein (olygodendroglial marker; Sigma-Aldrich) and anti-microtubule-associated protein 2 (neuronal marker; Sigma-Aldrich). Astrocyte cultures were found to be at least 99% glial fibrillary acidic protein positive. No cells were found to express CD-11b, myelin basic protein or microtubule-associated protein 2.

In order to evaluate the effect of ethanol (EtOH, 10 and 50 mM), IL-1β treatment (10 ng/mL) or LPS (50 ng/mL) treatment, these compounds were added to Dulbecco’s modified Eagle medium culture medium containing 1 mg/mL bovine serum albumin, and cells were harvested at various exposure times (5, 10, 15, 30 and 60 min) for the different determinations. In some experiments, astrocytes were pre-incubated with a medium containing nystatin (50 μg/mL, for 15 min at 37°C) or saponin (0.5% for 30 min at 25°C) or MβCD (10 mM for 15 min at 37°C), before EtOH, IL-1β or LPS treatments.

Detergent-free purification of lipid rafts microdomains

Primary astrocytes treated with IL-1β, LPS or EtOH (10 and 50 mM) for 5, 10, 15, 30 and 60 min, were used to prepare lipid rafts/caveolae fractions as previously described (Song et al. 1996; Teixeira et al. 1999). Briefly, cells were washed in ice-cold phosphate-buffered saline (PBS) and scraped into 400 μL of sodium carbonate buffer, pH 11.0 (150 mM Na2CO3, 2 mM EDTA, 10 mM Na4P2O7, 100 mM NaF, 1 mM sodium orthovanadate, 2 μM leupeptine, 0.5 mM phenylmethylsulfonyl fluoride and 2 μM aprotinin). Homogenization was performed by passing the lysates through a 25-G × 5/8 inch needle 10 times, and then sonicating them (three times for 20 s). The homogenate was then adjusted to 40% sucrose by the addition of 400 μL of 80% sucrose solution prepared in 2-(N-morpholino)-ethanesulfonic acid-buffered saline (MBS) buffer [25 mM 2-(N-morpholino)-ethanesulfonic acid pH 6.5, 150 mM NaCl and 2 mM EDTA], and was placed at the bottom of an ultracentrifuge tube. A 5–35% discontinuous sucrose gradient formed above in MBS containing 150 mM sodium carbonate, and centrifuged at 39 000 rpm (200 000 g) for 16 h in an SW60Ti rotor (Beckman Instruments, Palo Alto, CA, USA). A light-scattering band confined to the 5–35% sucrose interface containing lipid rafts caveolae was observed. From the top of each gradient, nine fractions of 0.5 mL were harvested along with the pelleted material. Each fraction was mixed with sodium dodecyl sulphate (SDS) buffer (Laemmli 1970) and used for western blotting analysis.

Western blot analysis

Protein extracts from cultured astrocytes treated with nystatin, saponin or MβCD, before EtOH, IL-1β or LPS treatments (30 min), were mixed with equal volumes of lysis buffer [1% Nonidet P-40 (Sigma-Aldrich), 20 mM Tris–HCl pH 8, 130 mM NaCl, 10 mM NaF, 10 μg/mL aprotinin, 10 μg/mL leupeptin, 10 mM dithiothreitol, 1 mM Na3VO4 and 1 mM phenylmethylsulfonyl fluoride] and boiled for 5 min. Protein concentration in the cell lysate was determined by the Bradford assay (Bio-Rad, Hercules, CA, USA). Cell lysates and fractions from the sucrose gradient were separated by SDS–polyacrylamide gel electrophoresis gels and transferred to polyvinylidene difluoride membranes using standard techniques. Membranes were blocked with 5% dried milk in Tris-buffered saline containing 0.1% Tween-20, and then incubated overnight with the following primary antibodies: anti-flotillin-1 (1 : 100), anti-cav-1 (1 : 1000), anti-TLR4 (1 : 50), anti-IL-1RI (1 : 100), anti-MyD88 (1 : 500), anti-P-IRAK1 (1 : 200) and anti-P-ERK (1 : 250). After washing with Tris-buffered saline containing 0.1% Tween-20, blots were incubated with HRP-conjugated antibodies: anti-mouse IgG (1 : 1000) for P-IRAK1 and P-ERK, anti-rabbit (1 : 20 000) for flotillin-1, cav-1, IL-1RI, MyD88 and anti-goat (1 : 10 000) for TLR4. To analyse IRAK-1 and ERK-2, some membranes were stripped with SDS solution (2% SDS, 0.85% 2-mercaptoethanol, 65 mM Tris–HCl, pH 6.8) for 1 h at 60°C, and then incubated with anti-ERK-2 (1 : 250) or anti-IRAK-1 (1 : 200) for 2 h. For GAPDH detection, as loading protein control, after stripped, membranes were incubated with anti-GAPDH (1 : 5000). Blots were developed using the enhanced chemiluminescence system (ECL Plus; Amersham Pharmacia Biotech., Madrid, Spain), and the intensity of the bands was quantified with SigmaGel image analysis software version 1.0 (Jandel Scientific, Madrid, Spain).

Confocal scanning laser microscopy

Cultured astrocytes growing on 16-mm glass coverslips were used for immunofluorescence analysis to evaluate the internalization and endocytosis of IL-1RI. Cells treated either with or without IL-1β or EtOH (10 and 50 mM) for 5, 10 and 30 min, were fixed with formaldehyde (3.7%) in PBS for 15 min and blocked for 1 h with PBS containing 10% FBS, 5% milk powder and 0.5% bovine serum albumin. Cells were then co-incubated with anti-cav-1 (caveosomes marker, 1 : 50) or with anti-LAMP-1 (late endosomes marker, 1 : 100), along with anti-IL-1RI (1 : 50), followed by FITC-conjugated donkey anti-mouse (for cav-1 and LAMP-1) and TRITC-conjugated goat anti-rabbit (for IL-1RI). In some experiments, live astrocytes were treated with Ct-B labelled with the Alexa Fluor 488 (lipid rafts caveolae marker, 1 μg/mL, 10 min incubation at 4°C), with ER-Tracker Red (ER marker, 1 μM, 30 min of incubation at 37°C), or with BODIPY TR Red (Golgi marker, 5 μM, 30 min of incubation at 4°C, and then 30 min at 37°C). After these treatments, cells were stimulated with IL-1β or EtOH (10 and 50 mM), fixed and incubated with anti-IL-1RI (1 : 50), followed by FITC-conjugated goat anti-rabbit. Negative controls were obtained by omission of primary antibodies. Cell nuclei were detected by incubation with Hoechst 33342 (1 : 20 000; Molecular Probes). Immunofluorescence images were collected in a confocal microscope Leica TCS-SP2-AOBA (Leica Microsystems, Heidelberg GmbH, Mannheim, Germany). Quantitative analysis was performed by counting 500–1000 cells of 3–5 fields per coverslip. In each field, the proportion of cells (Hoechst positive) that were either IL-1RI-/Ct-B positive (IL-1RI-enriched caveosomes) or IL-1RI/BODIPY TR positive (IL-1RI in the Golgi–ER/nucleus region) was determined. An average of three coverslips per treatment from four different cultures was assessed.

Statistical analysis

Results are reported as the mean ± SD. Data were analysed using one- and two-way anova.

Results

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

Stimulation of astrocytes with either IL-1β or LPS results in a rapid translocation of IL-1RI and TLR4 into lipid rafts caveolae, and to the recruitment of activated signalling proteins (P-IRAK and P-ERK)

Rafts can be biochemically defined as detergent-resistant membranes that float at a light buoyant density on sucrose gradients. However, to avoid artefacts by the partitioning of detergent molecules into the lipid bilayer, detergent-free methods have also been successfully used in isolating membrane fractions with similar biochemical characteristics (Luria et al. 2002; Sot et al. 2006; Huang et al. 2007). Therefore, in order to investigate whether membrane rafts are involved in IL-1RI and TLR4 signalling in astrocytes, and whether these events occurs in rafts/caveolae microdomains, we used a sodium carbonate detergent-free method which allows the enrichment of rafts/caveolae on a sucrose gradient based on the buoyant-density properties of these structures (Song et al. 1996; Teixeira et al. 1999). Equal volumes of each of the nine collected fractions were submitted to immunoblotting analysis for flotillin-1, cav-1 and other proteins in both resting and stimulated cells.

Under our experimental conditions, in resting astrocytes, cav-1 was recovered in light density fractions (3–5), herein referred to as caveolae-enriched fractions, while flotillin-1, a protein expressed in lipid rafts (Langhorst et al. 2005) but present also in other cellular compartments (Kokubo et al. 2003; Lopez-Casas and del Mazo 2003; Reuter et al. 2004), was distributed in fractions 3–9 (Figs 1 and 2). Therefore, by using this established protocol we mainly separated the cav-enriched membrane fractions from the bulk of astrocytes membrane and cytosolic proteins (Song et al. 1996).

image

Figure 1.  Treatment with either interleukin-1 beta (IL-1β) or lipopolysaccharide (LPS) induces translocation of IL-1RI or toll-like receptor type 4 (TLR4) and the recruitment of downstream signalling proteins into rafts/caveolae. Astrocytes were treated with either IL-1β (a) or LPS (b) for 5, 10, 15, 30 and 60 min. After the different times, cells were washed and lysed with Na2CO3 buffer pH 11.0. Lysates were subjected to discontinuous sucrose gradient centrifugation (5–40%) as described in Materials and methods. Fractions (0.5 mL) were collected from top of the gradient and equal volumes (80 μL) were separated by 8–12% sodium dodecyl sulphate–polyacrylamide gel electrophoresis. After transferring onto polyvinylidene difluoride, and to assess the lipid rafts enrichment in the different fractions, membranes were probed for flotillin-1 and caveolin-1. Membranes were also probed for IL-1RI, TLR4, myeloid differentiation primary-response protein 88 (MyD88), phospho-IL-1R-associated kinase (P-IRAK) and phospho-extracellular regulated-kinase (P-ERK), using specific antibodies followed by horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence. Fraction numbers are indicated horizontally. Representative blots of three separate experiments are shown. (c) Quantitative analysis of IL-1RI, P-IRAK and P-ERK expression in the different sucrose fractions after 0, 5, 10, 15, 30 and 60 min of IL-1β treatment. Data were mean ± SD of densitometric values obtained of three different experiments. The caveolin-1-enriched fractions are marked in the graph to visualize the protein concentration in this area. (d) Quantitative analysis of TLR4, P-IRAK and P-ERK expression in sucrose fractions after 0, 5, 10, 15, 30 and 60 min of LPS treatment. Values were mean ± SD of three independent experiments. The caveolin-1-enriched fractions are also marked in the graph.

Download figure to PowerPoint

image

Figure 2.  Low concentrations of ethanol promote the translocation of interleukin-1 beta receptor type I (IL-1RI), toll-like receptor type 4 (TLR4), phospho-IL-1R-associated kinase (P-IRAK) and phospho-extracellular regulated-kinase (P-ERK) into lipid rafts caveolae, as IL-1β or lipopolysaccharide (LPS) treatment. Aliquots (100 μL) of detergent-free lysates of astrocytes obtained before and after stimulation with ethanol (EtOH) at 10 mM (a) or 50 mM (b) for different times (5, 10, 15, 30 and 60 min) were fractionated on discontinuous sucrose gradient (5–40%) centrifugation. Fractions from the gradient were subjected to 8–12% sodium dodecyl sulphate–polyacrylamide gel electrophoresis. After transfer onto polyvinylidene difluoride, membranes were probed for the presence of IL-1RI, TLR4, myeloid differentiation primary-response protein 88 (MyD88), P-IRAK and P-ERK. The lipid raft-enriched fractions were detected using anti-flotillin-1 and anti-caveolin-1. Similar results were obtained from four separate experiments. The quantitative analysis of TLR4, IL-1RI, P-IRAK and P-ERK expression in the different sucrose fractions was determined after 0, 5, 10, 15, 30 and 60 min upon ethanol at 10 mM (c) and 50 mM (d) treatment. Data were mean ± SD of four independent experiments. The caveolin-1-enriched fractions (fractions 3–5) are marked in the graphs.

Download figure to PowerPoint

The stimulation of astrocytes with either IL-1β (10 ng/mL) or LPS (50 ng/mL) induces a rapid translocation of IL-1RI and TLR4 from the bottom fractions in resting cells to the light buoyant density fractions or cav-enriched fractions. We observed a first recruitment of these receptors to rafts caveolae at 5 min and then the receptors returned to unstimulated conditions (10–15 min). Once more, a strong recruitment of IL-1RI and TLR4 into rafts/caveolae occurred at 30 min upon ligand stimulation. A gradual disappearance of these receptors in cav-enriched fractions was noted at 1 h (Fig. 1a and b).

Notably, IL-1β and LPS treatments also stimulate a translocation of signalling molecules associated with the IL-1RI/TLR4 response, such IRAK, P-IRAK, ERK and P-ERK, to the rafts/caveolae enriched fractions. Maximal translocation of these kinases also occurs at 5 and 30 min upon ligand treatment. Likewise, although the levels of IRAK-1 and ERK-2 in the different fractions grossly paralleled the levels of their phosphorylated forms (data not shown), the ratios of the signals for phosphorylated to total IRAK and ERK significantly increased (1.4 ± 0.11 for IRAK and 2.31 ± 0.12 for ERK; ratio of the signal given with an antibody specific for the phosphorylated forms vs. the signal given by an antibody against the total protein) at 5 and 30 min within the rafts/caveolae fractions (fractions 3–5), when compared with the non-caveolae-enriched fractions (0.94 ± 0.16 and 0.98 ± 0.12 for IRAK and ERK respectively). These results suggest that both IRAK-1 and ERK-2 are recruited into rafts/caveolae fractions upon stimulation, promoting the phosphorylation of ERK and IRAK, as well as of other signalling proteins (Cuschieri et al. 2004; Soong et al. 2004; Riteau et al. 2006; Kam et al. 2007). The adaptor protein MyD88 is also recruited into rafts/caveolae upon 30 min of stimulation with IL-1β (Fig. 1a), but not with LPS (Fig. 1b), indicating that the signal transduction induced by IL-1β occurs by MyD88-dependent pathway, as observed in other cell types (Koedel et al. 2004; Davis et al. 2006; Tabarean et al. 2006).

In order to visualize more clearly the data, we also assessed densitometrically the immunoreactive bands for IL-1RI, TLR4, P-IRAK and P-ERK, upon IL-1β and LPS treatment at 0, 5, 10, 15, 30 and 60 min. In Fig. 1(c), the presence of IL-1RI, P-IRAK and P-ERK in the raft/caveolae-enriched fractions (fractions 3–5) is clearly observed upon 5 and 30 min of IL-1β treatment. These proteins are not coupled to caveolae after 1 h, demonstrating the reversibility of their incorporation into these rafts. Similar results were observed after LPS treatment (Fig. 1d), in which the presence of TLR4, P-IRAK and P-ERK in caveolae-enriched fractions significantly increased at 5 and 30 min of cell stimulation.

Altogether, these results indicate that IL-1β or LPS stimulation triggers receptor (IL-1RI and TLR4) recruitment and signalling into the caveolae rafts in astrocytes.

Ethanol facilitates the translocation of IL-1RI and TLR4, and the recruitment of activated signalling molecules into rafts microdomains in astrocytes

The above findings clearly demonstrate that lipid rafts caveolae are involved in IL-1RI and TLR4 recruitment and signalling in astrocytes. Therefore, we next evaluate whether the activation of IL-1RI/TLR4 by ethanol is mediated by promoting receptors recruitment into the lipid rafts caveolae, leading to receptor signalling. For this purpose, we isolated rafs/caveolae fractions using a detergent-free method and discontinuous sucrose gradient (Song et al. 1996; Teixeira et al. 1999) from astrocytes treated with ethanol (10 and 50 mM) for different time periods (5, 10, 15, 30 and 60 min). An effective isolation of raft/caveolae was confirmed by the enrichment of cav-1 in fractions 3–5 from the sucrose gradient (Fig. 2a). In this figure we showed that stimulation with 10 mM ethanol induced a rapid translocation of IL-1RI and TLR4 to raft/caveolae-enriched fractions (3–5) at 5 min, followed by a redistribution of the receptors as the resting conditions at 10 min. However, while maximal translocation of IL-1RI and TLR4 occurred at 30 min upon IL-1β and LPS stimulation, ethanol treatment induced a strong and sustained recruitment of these receptors in caveolae rafts at 15 and 30 min, leading to gradual raft-receptor dissociation at 1 h. Similarly to IL-1β and LPS, 10 mM ethanol induces at 5 and 30 min: (i) the translocation and recruitment of IRAK-1 and ERK-2 (data not shown) and the activated kinases (P-IRAK and P-ERK) to rafts/caveolae enriched fractions (Fig. 2a and c), (ii) the phosphorylation of IRAK and ERK within the rafts/caveolae, as the proportion of phosphorylated form/total form of IRAK and ERK increased in rafts/caveolae fractions (1.52 ± 0.12 for IRAK and 2.40 ± 0.11 for ERK) when compared with the non-caveolae enriched fractions (0.88 ± 0.07 for IRAK and 1.05 ± 0.13 for ERK). These results suggest that ethanol is capable of facilitating the translocation of IL-1RI/TLR4 and of recruiting signalling molecules into rafts/caveolae. As with the LPS treatment, the adaptor protein MyD88 is not recruited into rafts/caveolae after ethanol stimulation (Fig. 2a).

Notably, when lipid rafts were isolated from astrocytes treated with 50 mM ethanol, we noted some differences when compared with cells stimulated with 10 mM ethanol. Thus, although treatment with 50 mM ethanol induces some translocation of the IL-1RI and TLR4 receptors, as well as activating signalling molecules (P-IRAK and P-ERK) to the rafts/caveolae, the diffuse immunoreactive bands corresponding to flotillin-1 and cav-1 from the different sucrose fraction suggests that a partial disruption of lipid rafts occurs with 50 mM ethanol treatment (Fig. 2b and d).

Collectively, our results suggest that low concentrations of ethanol (10 mM), like IL-1β and LPS, activate IL-1RI/TLR4 receptors by promoting the translocation of these receptors in the lipid rafts caveolae in astrocytes, triggering the recruitment of activated signalling molecules and downstream signalling events.

Lipid rafts integrity is essential to induce a successful activation of glial IL-1RI and TLR4, after IL-1β, LPS and ethanol treatments

To further analyse the role of rafts in the activation and protein expression of IL-1RI and TLR4 receptors, we pre-treated astrocytes with rafts-disrupting agents, and then cells were stimulated with IL-1β, LPS or ethanol (10 and 50 mM). Agents that sequester, chelate or prevent the synthesis of cholesterol are commonly used to disrupt lipid rafts in order to assess the role of these microdomains in receptor activation and signalling. Specifically, we used nystatin which intercalates into lipid moieties, disrupting lipid raft structures (cholesterol sequestration), saponin, a pore-forming agent (cholesterol sequestration) and MβCD, which causes cholesterol depletion. The results in Fig. 3(a and b) show that the stimulation of astrocytes with either IL-1β or LPS for 30 min increases the protein levels of IL-1RI and TLR4 respectively. Similarly, ethanol treatment (10 and 50 mM) also up-regulates the protein expression of both receptors. Pre-treatment of astrocytes with raft-depleted agents, such as nystatin, saponin or MβCD abolished the increase in protein expression of these receptors, suggesting the role of lipid rafts microdomains in the receptors activation. Notably, we also observed that immunoreactivity for the control protein GAPDH is not affected by these rafts-disrupting substances. These results suggest that these agents might promote the degradation of proteins associated with lipid rafts, such as TLR4 and IL-1RI, as occurs with other proteins in which raft microdomains are essential for their stability, function and signalling (Remacle-Bonnet et al. 2005; Arcaro et al. 2007; Jiang et al. 2007; Raghu et al. 2007).

image

Figure 3.  Treatment of astrocytes with lipid rafts disrupting agents abolishes the expression and activation of interleukin-1 beta receptor type I (IL-1RI) and toll-like receptor type 4 (TLR4) induced by IL-1β, lipopolysaccharide (LPS) or ethanol treatment. Cells were treated with or without IL-1β, LPS or ethanol (EtOH) (10 and 50 mM) for 30 min. In some experiments, astrocytes were pre-incubated with nystatin (50 μg/mL, 15 min at 37°C), saponin (0.5%, 30 min at 25°C) or methyl-β-cyclodextrin (MβCD) (10 mM, 15 min at 37°C), before IL-1β, LPS or EtOH induction. Lysates from resting and stimulated cells were analysed by western blotting for: (a) IL-1RI (80 μg/line) and (b) TLR4 (60 μg/line). Blots were redeveloped with anti-GAPDH to show equal protein loading. A representative immunoblot is shown. Values were mean ± SD for a total of four different experiments. *p < 0.05 versus control astrocytes, #p < 0.05 versus rafts-depleted cells and ##p < 0.001 versus rafts-depleted cells, two-way anova test.

Download figure to PowerPoint

Stimulation of astrocytes with either IL-1β or ethanol results in the internalization of IL-1RI into enlarged cytoplasmic ring structures or caveosomes

As receptor recruitment and signalling, as well as receptor trafficking, appear to be closely linked, we then decided to evaluate whether the activation of IL-1RI by either IL-1β or ethanol and the recruitment of the IL-1RI receptor into rafts-caveolae could lead to the internalization of IL-1RI via the cav-dependent pathway. For these experiments, astrocytes were incubated with fluorescein-Ct-B, a classical raft marker. Then cells were fixed, blocked and immunostained for IL-1RI. Cells were analysed by confocal microscopy. In resting cells, IL-1RI staining appeared on the plasma membrane (Fig. 4a), whereas Ct-B was localized in both the plasma membrane and intracellular compartment with a minimal co-localization with IL-1RI. After 5 min of IL-1β treatment, IL-1RI and Ct-B staining appear concentrated in large cytoplasmic ring-shaped structures. In fact, at that time point 62.3% ± 9.07% of the astrocytes displayed IL-1RI/Ct-B positive structures versus 2.6% ± 1.3% in non-stimulated cells (p < 0.001, one-way anova, Fig. 4b). The size of these structures varied, ranging between 5 and 8 μM in diameter. The redistribution of IL-1RI and Ct-B labelling was reversible after 10 min of IL-1β treatment, when the typical surface staining pattern was recovered, as in the resting conditions (7.3% ± 3.1% IL-1RI/Ct-B positive cells, Fig. 4c). However, a redistribution of IL-1RI and Ct-B into ring-shaped structures was observed once more upon 30 min of IL-1β treatment. At this time however, the number of cells containing the ring structures increased by 85.3% ± 10.26% (p < 0.001, vs. non-stimulated cells, one-way anova, Fig. 4d), and the intensity of IL-1RI/Ct-B labelling was higher than at 5 min of IL-1β treatment. To further characterize these ring structures, and to determine whether they are involved with caveolar endocytosis, astrocytes were labelled with cav-1, a caveosomes marker, and also with IL-1RI. As Fig. 4(e) depicts, a high-magnification picture indicated that IL-1RI accumulated within the cav-1 positive ring structures, suggesting that IL-1RI is endocytosed upon IL-1β stimulation via the caveolae-dependent pathway through enlarged caveosomes organelles.

image

Figure 4.  Interleukin-1 beta receptor type I (IL-1RI) internalize through caveolar endocytosis via enlarged caveosomes after IL-1β binding. Astrocytes growing on 16-mm glass coverslips were incubated with cholera toxin subunit B (Ct-B)-Alexa Fluor 488 (1 μg/mL, 10 min 4°C). Then cells were stimulated with IL-1β for 5, 10 and 30 min, fixed and incubated with IL-1RI, followed by secondary antibody tetramethylrhodamine isothiocyanate (TRITC). Finally, cells were incubated with Hoechst to visualize cellular nucleus and analysed by confocal microscopy. (a) In control cells, Ct-B and IL-1RI predominantly localized in the plasma membrane and in the cytoplasmic region. After 5 min of IL-1β treatment, both Ct-B and IL-1RI were located in enlarged cytosolic ring structures (b). At 10 min, Ct-B and IL-1RI were located on the cell surface, as resting conditions (c). At 30 min of stimulation, a new pronounced redistribution of Ct-B and IL-1RI into large ring-structures was observed. At this time, both the number and intensity of the IL-1RI labelling of these enlarged structures significantly increased (d, arrows). In some experiments, cells were co-incubated with anti-caveolin-1 (caveosomes marker) and anti-IL-1RI, conjugated with secondary antibodies FITC and TRITC respectively. The aberrant ring structures are also stained with caveolin-1 (e, arrows). Original magnification 63×, scale bar: 30 μm (a–d). Original magnification 100×, scale bar: 10 μm [left top rows showing magnification of boxed region in the second row and (e)].

Download figure to PowerPoint

We next tested whether ethanol treatment (10 and 50 mM) could also induce an internalization of IL-1RI via the caveolae/rafts-dependent endocytosis. As shown in Fig. 5(a), stimulation with 10 mM ethanol for 5 min also induced a redistribution of IL-1RI and Ct-B from the cell membrane to cytoplasmic ring-like structures (58.5% ± 14.54% IL-1RI/Ct-B positive cells vs. 2.6% ± 1.3% non-stimulated cells, p < 0.05, one-way anova, Fig. 5a). These structures disappeared after 10 min of ethanol stimulation (12.45% ± 5.86% IL-1RI/Ct-B positive cells, Fig. 5b). After 30 min of ethanol treatment however, the number and intensity of these structures significantly increased (82% ± 8.18% IL-1RI/Ct-B positive cells vs. non-stimulated cells, p < 0.001, one-way anova, Fig. 5c). As these structures are co-stained with IL-1RI, and with either cav-1 (Fig. 5d) or Ct-B, these results indicate that ethanol (10 mM), like IL-1β, is capable of activating IL-1RI, triggering its internalization via caveosomes or caveolae-dependent endocytosis.

image

Figure 5.  Treatment of astrocytes with ethanol triggers the internalization of interleukin-1 beta receptor type I (IL-1RI) into enlarged caveosomes. Cells were stimulated with ethanol (EtOH) at 10 or 50 mM for 5, 10 and 30 min. Confocal images of astrocytes with similar staining conditions used in Fig. 4 are shown. Both 10 and 50 mM of EtOH treatment induced the same redistribution of cholera toxin subunit B (Ct-B) and IL-1RI into cytoplasmic ring structures (at 5 and 30 min) (a, e and g, arrows), as observed with IL-1β stimulation in Fig. 4. However, some ring structures labelled with IL-1RI were observed after 10 min of 50 mM ethanol treatment (f, arrows). At this time, these structures do not appear with either IL-1β or 10 mM of ethanol treatment. Cells were also co-incubated with anti-caveolin-1 and anti-IL-1RI. Aberrant ring structures were stained with caveolin-1 in cells stimulated with either 10 or 50 mM ethanol (d, h, arrows). Original magnification 63×, scale bar: 30 μm (a–c and e–g). Original magnification 100×, scale bar: 10 μm [right top rows showing magnification of boxed region in the third row and (d and h)].

Download figure to PowerPoint

Similar results were obtained when cells were stimulated with ethanol at 50 mM. In this case, we also noted the localization of IL-1RI in cytosolic ring structures upon 5 min (65.2% ± 17.54% IL-1RI/Ct-B positive cells vs. 2.6% ± 1.3% control cells p < 0.001, one-way anova) and 30 min (84.3% ± 13.83% IL-1RI/Ct-B positive cells, p < 0.001, one-way anova) of ethanol stimulation, which were labelled with both cav-1 and Ct-B (Fig. 5e, g and h). After 10 min of ethanol stimulation however, the IL-1RI staining pattern did not return to the resting situation, as observed with IL-1β or ethanol at 10 mM, as some ring structures containing IL-1RI were still observed at this time (30.2% ± 13.83% IL-1RI/Ct-B positive cells, p < 0.05, one-way anova, Fig. 5f). All this suggests that 50 mM ethanol caused a maintained stimulation and endocytosis of this receptor.

Internalized IL-1RI accumulates in the ER–Golgi and in the nucleus of astrocytes through the endocytic caveosomes pathway

Previous studies suggested that both internalization and intracellular transport events associated with the IL-1RI response differ from other cytokines and hormones receptors. After ligand binding, most receptors are rapidly internalized in coated pits, passing through endosomes to be finally degraded in lysosomes. However, internalized IL-1RI is not degraded and is primarily localized in the cytoplasm and over the nucleus, which might suggest a novel mechanism by which IL-1RI could be internalized (Qwarnstrom et al. 1988; Curtis et al. 1990; Bavelloni et al. 1999). According to this early data, our results show that upon stimulation with either IL-1β or ethanol for 30 min, IL-1RI not only accumulated in ring-cytoplasmic structures (caveosomes), but a strong labelling was also observed in the nuclear region of astrocytes. Therefore, in order to characterize the intracellular transport of endocytosed IL-1RI, resting and stimulated astrocytes were incubated with either BODIPY TR Red™, a Golgi complex marker, or ER-Tracker Red™, an ER marker, together with IL-1RI conjugated with FITC secondary antibody. In resting astrocytes, IL-1RI labelling was mainly located on the cell surface, while BODIPY TR labelling appeared in the cytoplasm, and also in the perinuclear region as a punctate pattern (Fig. 6a); ER-Tracker labelling mostly appeared in the perinuclear region or the ER network (Fig. 6e). Upon stimulation with either IL-1β or ethanol (10 and 50 mM) for 30 min, the Golgi apparatus appears more compact and, in some cells, IL-1RI labelling overlapped with BODIPY TR (51.2% ± 2.6% IL-1RI/BODIPY TR positive cells in Fig. 6b; 55% ± 14.2% IL-1RI/BODIPY TR positive cells in Fig. 6c; 75% ± 4.52% IL-1RI/ BODIPY TR positive cells in Fig. 6d). Notably, we also observed that the IL-1RI staining located in the perinuclear region, and overlapping with the ER-Tracker staining, suggests that IL-1RI is internalized and is then delivered to the ER–Golgi compartment (Fig. 6f–h) and also to the nucleus. Indeed, the Hoechst staining was eliminated in Fig. 6 in order to facilitate the visualization of IL-1RI staining within the nucleus and the perinuclear region. Altogether, our results indicate that IL-1RI is internalized by caveosomes, delivered into the ER/Golgi compartment, and subsequently into the nucleus.

image

Figure 6.  Endocytosed interleukin-1 beta receptor type I (IL-1RI) is sorted in the endoplasmic reticulum (ER)–Golgi region and nucleus after IL-1β and ethanol stimulation. Astrocytes were treated with IL-1β or ethanol (10 or 50 mM) for 30 min, and analysed by confocal microscopy. Before fixation, cells were incubated with BODIPY TR Red™ (Golgi marker, 5 μM, 30 min at 4°C, and then 30 min at 37°C), or with ER-Tracker Red™ (ER marker, 1 μM, 30 min at 37°C). After these treatments, cells were stimulated with either IL-1β or ethanol for 30 min, fixed and incubated with anti-IL-1RI followed by FITC-conjugated secondary antibody. In resting cells, IL-1RI labelling was located on the cell surface, BODIPY TR labelling appeared in cytoplasm and also in the perinuclear region as a punctate pattern, and ER Tracker labelling appeared concentrated in the perinuclear region (a and e). After IL-1β (b and f) or ethanol treatment, at 10 mM (c and g) or at 50 mM (d and h), IL-1RI redistributed into caveosomes, as well as in both the Golgi–ER and nucleus region. We eliminated Hoechst to facilitate the visualization of IL-1RI in the nucleus. Original magnification 63×, scale bar: 30 μm. N, nucleus.

Download figure to PowerPoint

IL-1RI is exclusively internalized via caveolar endocytosis in astroglial cells

Recent studies demonstrate that raft markers could be taken up by multiple pathways, and not just to caveosomes, because certain raft-associated proteins have been shown to be endocytosed via clathrin-mediated endocytosis, although this pathway mostly excludes lipid rafts (Nichols 2003; Rajendran and Simons 2005). Therefore, to ascertain whether internalized IL-1RI exclusively proceeds via rafts/caveosomes pathway, or whether other pathways such as clathrin-coated pits are also involved, astrocytes stimulated with either IL-1β or ethanol were labelled with IL-1RI-TRITC, and LAMP-1, a late endosomes marker conjugated with FITC secondary antibody. Then, cells were analysed by confocal microscopy. In stimulated astrocytes, LAMP-1 was mainly located in the perinuclear region, and this staining did not overlap with IL-1RI labelling (Fig. 7). In addition, in these cells we also observed that IL-1RI-containing ring structures were not stained with LAMP-1 (Fig. 7, arrows), with an early endosome associated protein-1 marker (data not shown), or with Lysotracker™ (Molecular Probes), a lysosome marker (data not shown). These data suggest that in astroglial cells, IL-1RI is endocytosed via the clathrin-independent mechanism into caveosomes after IL-1β or ethanol stimulation.

image

Figure 7.  Interleukin-1 beta receptor type I (IL-1RI) did not accumulate in late-endosomes upon IL-1β or ethanol stimulation. Cells stimulated with IL-1β (a), ethanol 10 mM (b) or 50 mM (c) for 30 min were labelled with IL-1RI and lysosome-associated membrane protein 1 (LAMP-1) (a late endosomes marker), followed by tetramethylrhodamine isothiocyanate (TRITC)- and FITC-conjugated secondary antibodies respectively. In the stimulated astrocytes, LAMP-1 was located in the perinuclear region and this staining did not overlap with IL-1RI labelling. The ring-cytoplasmic structures enriched with IL-1RI were not stained with LAMP-1 (marked with arrows). Original magnification 63×, scale bar: 30 μm (a–c). Original magnification 100×, scale bar: 10 μm [right top rows showing magnification of boxed region in (b)]. N, nucleus.

Download figure to PowerPoint

Discussion

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

Clinical and experimental studies revealed that ethanol intake affects both the adaptive and innate immune system, resulting in specific alterations in the TLRs response (Medzhitov 2001; Nelson and Kolls 2002; Campbell 2004). We have recently demonstrated that ethanol is capable of activating astrocytes to release cytokines and inflammatory mediators, and that these effects are mediated by the activation of IL-1RI/TLR4 since blocking these receptors abolishes the production of inflammatory mediators and cell death induced by ethanol (Blanco et al. 2004, 2005; Valles et al. 2004). Although the mechanisms by which ethanol activates the IL-1RI/TLR4 response are presently unknown, we have suggested that low concentrations of ethanol could promote the aggregation and clustering of lipid rafts microdomains, leading to the recruitment and activation of IL-1RI and TLR4 into these rafts (Blanco and Guerri 2007). We herein demonstrate this hypothesis by showing that ethanol (10 mM), as well as IL-1β or LPS treatment of astrocytes, induce a rapid translocation of IL-1RI and/or TLR4 into lipid rafts caveolae to trigger a recruitment of activated signalling molecules (P-IRAK and P-ERK) into these microdomains. We further demonstrate that the recruitment and activation of IL-1RI upon IL-1β- or ethanol-stimulation is followed by its internalization and intracellular trafficking via caveolar-dependent endocytosis.

Increasing evidence indicates the role of lipid rafts as a coordinated signalling platform that regulates the dynamics and agonist-induced translocation of specific signalling proteins (Dykstra et al. 2001; Lai 2003; Pike 2003; Laude and Prior 2004). The existence of different classes of lipid rafts has significant implications for the function of these membrane domains in cell signalling (Pike 2003). In primary astrocytes, the lipid rafts caveolae constitute an important membrane compartment of signalization, necessary for the coordinated activation of the intracellular pathways (Cameron et al. 1997; Teixeira et al. 1999). Indeed, we demonstrate, for the first time, the importance of lipid rafts caveolae in the recruitment and signalling of IL-1RI and TLR4 upon IL-1β and LPS stimulation in primary astrocytes. Thus, after ligand binding we observe two peaks of induction in which IL-1RI, TLR4 and activated signalling molecules translocate into caveolae-rich fractions, one at 5 min and the other at 30 min. In the latter peak, we observe the maximal recruitment of receptors and signalling molecules into rafts/caveolae fractions. We also note a disassembly of receptors to rafts/caveolae, as a resting condition, at the intervals between the two peaks of activation (10, 15 and 60 min). It is known that the interactions that drive raft assembly are dynamic and reversible, and indeed, raft clusters can disassemble by the negative modulators implicated in the signalling response and/or by the removal of raft components (Simons and Toomre 2000).

Notably, the present results also demonstrate that ethanol treatment triggers a similar translocation and recruitment of IL-1RI/TLR4 and activated signalling molecules into rafts/caveolae as their specific ligands. These results suggest that low concentrations of ethanol, through its interaction with membrane lipids, might facilitate lipid–protein and protein–protein interactions, allowing receptor aggregation and signalling. Indeed, slight alterations in the structure of these membrane microdomains can facilitate the initiation of signal transduction pathways (Simons and Toomre 2000). Nevertheless, the interaction of ethanol with rafts seems to depend on ethanol concentration since high ethanol concentration (> 50 mM) can disrupt membrane lipid microdomains, interfering with lipid raft clustering and leading to the suppression of the IL-1RI/TLR4 response (Blanco et al. 2005; Dai et al. 2005; Dolganiuc et al. 2006; Fernandez-Lizarbe et al. 2008). Indeed, our results demonstrate that moderate levels of ethanol (50 mM) result in a partial disruption of lipid rafts caveolae, affecting the recruitment of IL-1RI and TLR4 into these rafts compared with 10 mM ethanol. Finally, the role of lipid rafts in the ethanol- or ligand-induced IL-1RI and TLR4 recruitment and signalling in astrocytes is further supported by the findings demonstrating that the disruption of rafts/caveolae with nystatin, saponin or MβCD abolishes the expression and activation of both receptors. These results indicate that lipid rafts caveolae are an essential platform in the IL-1RI and TLR4 signalling response in astrocytes, and that they also play a key role in the induction of inflammatory mediators induced by ethanol in these cells (Blanco et al. 2005).

Another important finding of this study is that ethanol- or ligand-mediated IL-1RI recruitment leads to the internalization and intracellular trafficking of this receptor via caveolar endocytosis. Internalization of membrane receptors occurs through both clathrin- and caveolar/rafts-mediated pathways (Nichols and Lippincott-Schwartz 2001; Parton and Richards 2003; Neel et al. 2005; Parton and Simons 2007). However, in comparison with the clathrin-mediated pathway, alternative non-clathrin endocytic pathways, such as caveolar endocytosis, are poorly understood (Lajoie and Nabi 2007; Parton and Simons 2007). In the latter pathway, it has been shown that endocytic caveolar vesicles accumulate in a discrete population of pre-existing cytoplasmic structures that are enriched in cav-1, called caveosomes (Pelkmans et al. 2001). These structures, which are devoid of markers of other endocytic and biosynthetic organelles, serve as an intermediate station during the internalization of endocytosed ligands in the caveolar/raft endocytic pathway, delivering their cargo into the ER–Golgi compartment, and in the nucleus (Pelkmans et al. 2001; Rajendran and Simons 2005). According to this pathway, after 5 min of ethanol or IL-1β stimulation, we observe that IL-1RI is located in cytoplasm cav-1-positive ring structures or caveosomes. These enlarged structures were negative to the early/late endosomal and lysosomal markers, as demonstrated by the lack of staining with early endosome associated protein-1, LAMP-1 and Lysotracker™. At longer times (30 min), IL-1RI labelling was located in the ER–Golgi compartment and in the nucleus of astrocytes. Therefore, these findings provide strong evidence that IL-1RI is exclusively internalized via caveolar endocytosis upon stimulation with either IL-1β or ethanol. In addition, we also observe a correlation between the translocation and signalling of IL-1RI in the caveolae-enriched sucrose fractions and the appearance of caveosomes containing IL-1RI at 5 and 30 min of stimulation.

An interesting aspect of our results is the enlarged size of caveosomes. Large caveosomes and invaginations-involved caveolins have also been observed in the caveolar endocytosis used by viruses to enter cells, such as SV-40 and polyoma virus (Anderson et al. 1996; Stang et al. 1997; Pelkmans et al. 2001). By passing through this pathway, viruses avoid their degradation in lysosomes to accumulate in enlarged caveosomes (Parton and Simons 2007). Interestingly, the cytoplasmic region of IL-1RI and both viruses contains similar consensus sequences which are known to mediate their nuclear transport (Heguy et al. 1991), and which might explain the similar characteristics of caveolae-dependent internalization. In fact, it is known that IL-1RI is transported to the nucleus and regulates the IL-1-induced gene transcription (Curtis et al. 1990). Alternatively, the enlarged caveosomes could be the result of the fusion of other caveosomes. This idea is supported by a recent study showing that both LPS and TLR4 are trafficked to early/sorting endosomes upon LPS stimulation, and that LPS-induced signalling increases the size of these endosomes, by endosome–endosome fusions (Husebye et al. 2006).

Finally, although evidence from our laboratory suggests that ethanol can interact with non-caveolae lipid rafts activating TLR4 signalling in RAW 264.7 macrophages (Fernandez-Lizarbe et al. 2008), the present study demonstrates for the first time that IL-1RI and TLR4 activation and signalling occurs exclusively via lipid rafts caveolae in astrocytes, and that ethanol is able to act as a ligand-mediated activation of these receptors. In addition, we provide new insights into the mechanisms of ethanol action, demonstrating that activation of IL-1RI by either IL-1β or ethanol triggers the internalization and intracellular trafficking of this receptor via caveolar-dependent endocytosis in astrocytes.

In conclusion, the present data provide a novel mechanism by which ethanol, through its interaction with rafts/caveolae in astrocytes, promotes IL-1RI and TLR4 recruitment into these microdomains, and triggers their internalization and downstream signalling pathways associated with inflammation. This mechanism could participate in ethanol-induced neuroinflammation and brain damage (Valles et al. 2004). The results might also contribute to understanding the mechanisms of the actions of ethanol.

Acknowledgements

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

This work was supported by Spanish governments Ministerio de Educación y Ciencia (SAF 2006-02178), Ministerio de Sanidad, Instituto de Salud Carlos III (RTA Network, G03/005), PNSD (G46923421), FEPAD and Fundación de Investigación Médica Mutua Madrileña. We would like to thank Alberto Hernandez Cano for his invaluable help with the confocal microscopy. We also thank Marisa March for her excellent technical assistance. The authors have not conflicting financial interests.

References

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