SEARCH

SEARCH BY CITATION

Keywords:

  • alcohol;
  • TLR4;
  • MyD88;
  • IκBα;
  • loss of righting reflex;
  • sedation;
  • motor impairment;
  • mice studies

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusions
  7. Acknowledgements
  8. Conflicts of interest
  9. References

BACKGROUND AND PURPOSE Emerging evidence implicates a role for toll-like receptor 4 (TLR4) in the CNS effects of alcohol. The aim of the current study was to determine whether TLR4–MyD88-dependent signalling is involved in the acute behavioural actions of alcohol and if alcohol can activate TLR4-downstream MAPK and NF-κB pathways.

EXPERIMENTAL APPROACH The TLR4 pathway was evaluated using the TLR4 antagonist (+)-naloxone (µ-opioid receptor-inactive isomer) and mice with null mutations in the TLR4 and MyD88 genes. Sedation and motor impairment induced by a single dose of alcohol were assessed by loss of righting reflex (LORR) and rotarod tests, separately. The phosphorylation of JNK, ERK and p38, and levels of IκBα were measured to determine the effects of acute alcohol exposure on MAPK and NF-κB signalling.

KEY RESULTS After a single dose of alcohol, both pharmacological inhibition of TLR4 signalling with (+)-naloxone and genetic deficiency of TLR4 or MyD88 significantly (P < 0.0001) reduced the duration of LORR by 45–78% and significantly decreased motor impairment recovery time to 62–88% of controls. These behavioural actions were not due to changes in the peripheral or central alcohol pharmacokinetics. IκBα levels responded to alcohol by 30 min in mixed hippocampal cell samples, from wild-type mice, but not in cells from TLR4- or MyD88-deficient mice.

CONCLUSIONS AND IMPLICATIONS These data provide new evidence that TLR4–MyD88 signalling is involved in the acute behavioural actions of alcohol in mice.

LINKED ARTICLE This article is commented on by Pandey, pp. 1316–1318 of this issue. To view this commentary visit http://dx.doi.org/10.1111/j.1476-5381.2011.01695.x


Abbreviations
ADH

alcohol dehydrogenase

AUC

area under the alcohol concentration–time curves

BCA

bicinchoninic acid

CBA

cytometric bead array

LORR

loss of righting reflex

MyD88

myeloid differentiation primary response gene 88

NAD

nicotinamide adenine dinucleotide

TLR4

toll-like receptor 4

WT

wild-type

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusions
  7. Acknowledgements
  8. Conflicts of interest
  9. References

Alcohol is consumed annually by two billion people worldwide with its abuse posing a significant health and social problem, with over 76 million people diagnosed with an alcohol abuse disorder (WHO, 2004). Among acute alcohol-induced behavioural actions, sedation and motor inco-ordination are responsible for a significant number of traffic accident-related deaths (Lin and Kraus, 2009). The mechanisms causing impaired motor skills by alcohol were considered to be the enhanced GABA transmission to cerebellar granule cells (Carta et al., 2004) and Purkinje neurons (Hirono et al., 2009) in the cerebellum. Furthermore, mice with reduced affinity of the glycine binding site on NMDA receptor GluN1 subunit displayed an attenuated alcohol-induced motor dysfunction (Kiefer et al., 2003), implicating this system in alcohol action as well. Moreover, the GABA receptor (Linden et al., 2011), NMDA receptor (Boyce-Rustay and Holmes, 2005) and cAMP–protein kinase A (PKA) signalling (Wand et al., 2001) were demonstrated to be related to the sedative effects of alcohol. A variety of genes encoding second-messenger systems, neurotransmitters or opioid receptors, and alcohol metabolic enzymes has been demonstrated to be related to alcoholism (Schuckit et al., 2004). However, these purely neuronal and pharmacokinetic mechanisms of alcohol actions, which are still being elucidated, may not account for all of the behavioural effects induced by alcohol (Hyman et al., 2006), and a pro-inflammatory response induced by alcohol within the CNS may also play a role (He and Crews, 2008; Wu et al., 2011).

Glial cells and various immune modifying factors are activated following alcohol exposure in vitro (Alling et al., 1986; Hansson et al., 1987; Ronnback et al., 1988). Furthermore, rodents chronically treated with alcohol have increased levels of glial fibrillary acidic protein (GFAP, a pro-inflammatory astrocyte marker) in the ventral tegmental area (Ortiz et al., 1995), as well as CD11b (pro-inflammatory microglial marker) within the cerebral cortex (Alfonso-Loeches et al., 2010). In addition, several genes involved in the MAPK pathway are found to be up-regulated in the nucleus accumbens of a high alcohol-consuming rat line (Arlinde et al., 2004). Recently, toll-like receptor 4 (TLR4) has been demonstrated to be a key receptor in the activation of glial cells (microglia and astrocytes) following acute alcohol exposure in vitro and in chronic alcohol exposure ex vivo (Blanco et al., 2005; Fernandez-Lizarbe et al., 2009; Alfonso-Loeches et al., 2010). This is hypothesized to occur via the interaction between alcohol and the lipid rafts that trigger TLR4 signalling (Blanco et al., 2008), thus leading to an enhanced release of pro-inflammatory mediators following NF-κB up-regulation (Blanco et al., 2004; Valles et al., 2004). However, there is still a lack of direct evidence showing that acute alcohol administration triggers TLR4 signalling to modify its behavioural effects.

Emerging evidence indicates that the functions of certain neuroimmune molecules may contribute to the behavioural changes induced by alcohol exposure. At the cell signalling level, activation of the MAPK pathway reduced the motivation of rats to consume and seek alcohol (Carnicella et al., 2008). Moreover, null mutation of genes encoding chemokine (C–C motif) ligand 2 (CCL2, females), CCL3, or CCL receptor 2 (CCR2) resulted in a lower preference for alcohol in mice, and mice with a genetic deficiency of CCL2 or CCL3 showed a longer duration of alcohol-induced loss of righting reflex (LORR) than wild-type (WT) mice (Blednov et al., 2005). In addition, the systemic administration of LPS (a TLR4 ligand) in mice enhanced alcohol-induced motor impairment (Drugan et al., 2007) and alcohol consumption (Blednov et al., 2011). Furthermore, deletion of TLR4 protected mice against conditional learning and memory recognition dysfunctions elicited by chronic alcohol consumption (Pascual et al., 2011). However, the role of TLR4 signalling in the acute behavioural effects of alcohol has not been investigated.

In view of this new evidence for the role of TLR4 in the effects of alcohol within the brain, and the pivotal neuroinflammatory influence on the behavioural responses induced by alcohol, we hypothesized that inhibition of TLR4 signalling, by either genetic or pharmacological means, would reduce behavioural effects following acute alcohol administration in mice. Two behavioural tests, the LORR and rotarod test, were chosen to assess acute alcohol-induced sedation and motor inco-ordination, respectively. Our aim was to determine whether the TLR4-myeloid differentiation primary response gene 88 (MyD88)-dependent signalling cascade is involved in alcohol-induced sedation and motor impairment. Both genetic strategies (Tlr4 null mutant and Myd88 null mutant mice) and treatment with the TLR4 signalling inhibitor (+)-naloxone (the µ-opioid receptor-inactive isomer of naloxone) (Hutchinson et al., 2008; Hutchinson et al., 2010a) were used to assess the role of the TLR4 pathway. Furthermore, we examined whether any of the observed effects were related to changes in blood or brain pharmacokinetics of alcohol. Finally, we determined if the alcohol-induced activation of MAPK, JNK, ERK and p38, and IκBα (the main inhibitor protein of NF-κB), which are all involved in NF-κB signalling cascades, is TLR4-dependent.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusions
  7. Acknowledgements
  8. Conflicts of interest
  9. References

Animals

Pathogen-free male Balb/c WT mice, and mice with null mutations in the Tlr4 gene (Tlr4–/– mice) and Myd88 gene (Myd88–/– mice) (all 10–14 weeks old; n= 6–17 mice per group for behavioural studies, n= 4–5 mice per group for the pharmacokinetic study) were used in the experiments. Both Tlr4–/– and Myd88–/– mice, back-crossed onto Balb/c for more than 10 generations, were sourced from Prof Akira (Osaka University, Osaka, Japan) and purchased from Dr Simon Phipps (University of Queensland, Queensland, Australia) and Prof Paul Foster (University of Newcastle, New South Wales, Australia). Mice were housed in temperature (23 ± 3°C) and light/dark cycle (12/12 h) controlled rooms with standard rodent food and water available ad libitum. All animal care and experimental procedures complied with principles of the Australian code of practice for the care and use of animals for scientific purposes and were approved by the University of Adelaide Animal Ethics Committee.

Drugs, doses and solutions

Endotoxin-free (+)-naloxone was kindly provided by Dr Kenner Rice (Chemical Biology Research Branch, National Institute on Drug Abuse and National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Rockville, MD). Alcohol was obtained from Chem-Supply (99.5%, Gillman, South Australia, Australia). All other reagents and chemicals were of analytical grade quality. The receptor and channel nomenclature used in the paper follows Alexander et al. (2011).

For animal behavioural studies, (+)-naloxone was injected i.p. to the mice at 0.01 mL·g−1. The volume for injection of alcohol (20%, v/v, i.p.) varied and was based on animal weight and dose of alcohol. The weight of mice was 25 g on average and ranged from 22 to 30 g. Thus, the volume for injection of alcohol was 0.32 mL (range: 0.28–0.38 mL) at 2.0 g·kg−1, 0.40 mL (range: 0.35–0.48 mL) at 2.5 g·kg−1, 0.55 mL (range: 0.49–0.67 mL) at 3.5 g·kg−1 and 0.71 mL (range: 0.63–0.86 mL) at 4.5 g·kg−1 of alcohol. Endotoxin-free saline (0.9% sodium chloride) was used as the vehicle control.

For cell culture studies, (+)-naloxone and alcohol were diluted in endotoxin-free RPMI 1640 (Invitrogen, Carlsbad, CA), which was used as the vehicle control.

Alcohol-induced sedation and motor impairment

Effect of alcohol in WT, Tlr4–/– and Myd88–/– mice.  Following a dose of saline (−30 min), single alcohol doses of 2.5, 3.5 or 4.5 g·kg−1 were administered (0 min) once only to groups of WT, Tlr4–/– and Myd88–/– mice and alcohol-induced sedation, as measured by the duration of LORR, was recorded from 0 min. Subsequently, an ED50 value of alcohol was estimated (see Statistical analysis).

To assess alcohol-induced motor dysfunction with the rotarod test, saline (−30 min) was administered before a single dose of alcohol (2.0 g·kg−1) to these three groups of mice.

Effect of (+)-naloxone treatment.  To examine the effects of (+)-naloxone on alcohol-induced sedation, WT mice were treated with (+)-naloxone (10 or 60 mg·kg−1) or saline (each at −30 min), before a single 3.5 g·kg−1 alcohol dose (0 min) and duration of LORR recorded. To further assess any effect of (+)-naloxone on alcohol-induced sedation in null mutant mice, (+)-naloxone (60 mg·kg−1) or saline (each at −30 min) was administered to Tlr4/– or Myd88–/– mice before a single 4.5 g·kg−1 alcohol dose (0 min). These alcohol doses were chosen based on the ED50 in WT and null mutant mice, so that either reductions or enhancements in the effects of alcohol could be reliably quantified. Alcohol-induced sedation was subsequently evaluated via LORR test after alcohol administration.

To assess differences in motor co-ordination using the rotarod method, (+)-naloxone (60 mg·kg−1; −30 min) was administered before a single 2.0 g·kg−1 alcohol dose (0 min) in WT and the two null mutant groups of mice.

In vitro and ex vivo molecular studies

Effects of alcohol on IκBα protein levels in mixed hippocampal cell cultures.  Mixed hippocampal cells were isolated as previously described (Wu et al., 2011) from naïve drug-free WT, Tlr4–/– and Myd88–/– mice (n= 3 each). To analyse the effect of alcohol on IκBα levels, and the influence of (+)-naloxone, cells were treated with (+)-naloxone (153 µM, 50 µg·mL−1), or RPMI 1640 at 37°C, 5% CO2 for 30 min, before stimulation with alcohol (50 mM) or RPMI 1640 for a further 30 min, and relative IκBα protein levels were investigated by Western blotting.

Regulation of brain p38, JNK and ERK phosphorylation by alcohol.  To examine the effects of alcohol in vitro, hippocampal cells from each naïve mouse (WT or Tlr4–/–) were prepared as previously described (Wu et al., 2011). Cells were stimulated with 50 mM alcohol (based on previous studies; Alfonso-Loeches et al., 2010) or vehicle at 37°C, 5% CO2 for 10 min.

To evaluate these molecular effects of alcohol ex vivo, WT or Tlr4–/– mice were dosed with 3.5 g·kg−1 of alcohol or saline (0 min) and anaesthetized by an overdose of sodium pentobarbitone (300 mg·kg−1, 10 min). The hippocampus and cerebellum were isolated (15 min) with aseptic techniques after transcardial perfusion and immediately homogenized in 2 mL of Denaturation Buffer from BD CBA Cell Signaling Master Buffer Kit (BD Biosciences, San Diego, CA). Then, cellular enzymes within the samples were denatured by boiling at 100°C for 5 min with Denaturation Buffer, and samples were subsequently stored at −80°C until analysis. Phosphorylated ERK, JNK and p38 and total p38 (phosphorylated plus unphosphorylated) levels were quantified by a Cytometric Bead Array assay (see below for details). Protein concentrations of ex vivo samples were determined by the bicinchoninic acid (BCA) assay (Thermo Scientific, Waltham, MA) to normalize the data.

Alcohol pharmacokinetics

Effect of (+)-naloxone administration and genetic TLR4 or MyD88 deficiency on peripheral and brain alcohol concentrations.  To examine the influence of (+)-naloxone treatment and genetic TLR4 or MyD88 deficiency on blood and brain alcohol pharmacokinetics, (+)-naloxone (60 mg·kg−1) or saline (each at −30 min) was administered to WT or Tlr4–/– and Myd88–/– mice with a single dose of alcohol (3.5 g·kg−1, 0 min). Mice were anaesthetized by an overdose of sodium pentobarbitone 4 min before blood and tissue collection. Blood samples were taken via cardiac puncture at 15, 60, 120 or 180 min following the alcohol dose, or 4 min after the mice awoke. Following blood collection, mice were perfused transcardially with saline, and the brain then removed. Blood and brain samples were immediately placed on ice. The alcohol concentration in each sample was measured with a nicotinamide adenine dinucleotide (NAD)–alcohol dehydrogenase (ADH) assay (see below). For this experiment, blood and brain samples at each time point were collected from different animals.

Behavioural testing

LORR (sedation).  Mice were placed in separate cages with bedding, after being injected with alcohol. The duration of LORR was measured from the time of mice losing their righting reflex to the time of righting themselves three times in 30 s.

Rotarod (motor co-ordination).  The rotarod apparatus (Orchid Scientifics, Nashik, India) with a 3 cm-diameter dowel was set at a fixed speed of 4.3 × 10−3×g (16 rpm). The latency to fall was recorded as the duration that the mice remained on the rod, with a maximum cut-off latency of 180 s.

Each mouse underwent a training phase one day before the experimental testing. Training involved the mouse remaining on the rotarod for 180 s in three sequential trials. On the experimental testing day, mice underwent a baseline trial before any dosing to ensure they performed at the training standard time of 180 s; this was repeated if they fell off the rod before the 180 s cut-off. The mice were then dosed with (+)-naloxone or saline and returned to the cage for 30 min before alcohol administration (0 min). Another baseline test was conducted before alcohol administration. Mice were tested at 2, 5, 7, 13 and 20 min, and every 10 min thereafter, until they could remain on the rod for 180 s cut-off in two sequential trials. The duration of the mice remaining on the rod was recorded.

Molecular and chemical analyses

Western blotting.  The preparation of cellular lysates was performed as described previously (Lousberg et al., 2010). Briefly, mixed hippocampal cells were incubated on ice for 10 min, collected by centrifugation (2264 ×g, 4°C, 5 min) and washed with ice-cold Dulbecco's PBS (DPBS, Invitrogen, Carlsbad, CA). Cell pellets were resuspended in modified radioimmunoprecipitation assay (RIPA) buffer [10 mM Tris (pH 7.4), 100 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM NaF, 20 mM sodium orthovanadate, 20 mM Na4P2O7, 0.1% SDS, 0.5% sodium deoxycholate, 1% Triton X-100, 10% glycerol and complete EDTA-free protease inhibitor cocktail] for 15 min on ice. Following cell lysis, lysates were clarified via centrifugation (18 894× g, 4°C, 5 min) and stored until analysis at −80°C. Protein concentration was determined by BCA assay before Western blot.

For Western blotting, samples were heated in SDS loading buffer at 97°C for 5 min, fractionated by PAGE and transferred onto nitrocellulose membranes (GE Healthcare Biosciences, Pittsburgh, PA). The membranes were subsequently blocked with 5% ECL Blocking Agent (GE Healthcare Biosciences) in Tris-buffered saline (TBS) containing 0.1% Tween-20 for 1 h at room temperature and incubated with primary antibodies anti-IκBα (L35A5, 1:1000; Cell Signaling Technology, Danvers, MA) or anti-β-actin (3:5000; Rockland Immunochemicals, Gilbertsville, PA) overnight at 4°C. Blots were then washed and incubated with horseradish peroxidase (HRP)-conjugated anti-mouse (1:2000) or anti-rabbit (1:10 000) IgG antibodies respectively (GE Healthcare Biosciences) at room temperature for 1 h. The immunoreactive signal was visualized by a chemiluminescence method (ECL Western Blotting Detection Reagents, GE Healthcare Biosciences, Pittsburgh, PA, USA) followed by exposure to Hyperfilm ECL (GE Healthcare Biosciences). ImageJ software (http://rsb.info.nih.gov/ij/index.html) was used for quantifying the intensity of Western blot bands allowing comparison with the relevant β-actin controls.

Cytometric bead array.  Phosphorylated ERK 1/2 (T202/Y204), JNK 1/2 (T183/Y185) and p38 (T180/Y182) and total p38 were quantified in hippocampus and cerebellum tissue and in hippocampal cells with BD CBA Flex Set (BD Biosciences, San Diego, CA) and Cell Signaling Master Buffer Kit, according to the manufacturer's instructions. Data were acquired with a FACSCanto flow cytometer (BD Biosciences) and analysed with BD CBA Software according to the manufacturer's instructions.

NAD–ADH assay (alcohol quantification).  A NAD–ADH assay was used to quantify alcohol concentrations in blood and brain samples as previously described (Smolen and Smolen, 1989; Wu et al., 2011). The assay accuracy was expressed as the relative error (RE) according to the equation: RE (%) = 100% × (measured concentration − spiked concentration) / spiked concentration, and the precision evaluated by the coefficient of variation (CV). The intra-assay precision and accuracy were estimated by analysing five replicates at three different quality control levels (700, 500 and 200 mg 100 mL−1). Intra-assay precision and inaccuracy ranged from 2.8% to 9.8%, and −2.2% to 8.8%, respectively.

To test if (+)-naloxone or acetaldehyde would interfere with this assay, between 0.001 and 100 µM of each drug was added to serum (containing 500 mg alcohol 100 mL−1 serum), or an equal volume of brain was homogenized in solution (containing the equal concentration of alcohol) and assayed together in the absence of drugs. The results demonstrated that the presence of either of the drugs did not influence the results obtained from this assay.

Statistical analysis

GraphPad Prism 5.02 (GraphPad Software Inc., San Diego, CA) was used for all statistical analysis. One-way anova with Bonferroni's multiple comparison test, or two-way anova followed by Bonferroni's post hoc test were performed. Data are presented as mean ± SEM. P-values of 0.05 or less were considered significant.

ED50 was calculated from the dose–response curves, which were developed from the data of duration of alcohol-induced LORR. To fit these data, nonlinear regression (GraphPad Prism 5.02) was used, where minimum was set as 0, and maximum was the maximum duration of LORR. This resulted in estimate of ED50, slope and maximum.

  • image

The areas under the alcohol concentration–time curves (AUCs) from 0.25 to 3 h post-alcohol administration were calculated using the linear trapezoidal rule. Slopes of the alcohol concentration-time curves were calculated by linear regression, which followed zero-order kinetics.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusions
  7. Acknowledgements
  8. Conflicts of interest
  9. References

Mice deficient in TLR4 or MyD88 display decreased sedative and motor effects of alcohol

Tlr4–/– and Myd88–/– mice exhibit a shorter duration of alcohol-induced LORR.  After a single dose of alcohol, the sedative effect of alcohol was dose-dependent as an increase in dose led to an increase in the duration of LORR in WT, Tlr4–/– and Myd88–/– mice (P < 0.0001). Both Tlr4–/– and Myd88–/– mice displayed a shorter duration of alcohol-induced LORR than WT mice (P < 0.0001) at both 3.5 g·kg−1 (Tlr4–/–, 16 ± 4 min, P < 0.001, n= 9; Myd88–/–, 31 ± 8 min, P < 0.001, n= 10; WT, 73 ± 3 min, n= 13) and 4.5 g·kg−1 (Tlr4–/–, 56 ± 6 min, P < 0.001, n= 8; Myd88–/–, 78 ± 19 min, P < 0.01, n= 8; WT, 123 ± 3 min, n= 6) of alcohol (Figure 1A). The ED50 of alcohol was 3.4 ± 0.1 [95% confidence interval (95% CI), 3.2–3.6] and 4.2 ± 0.2 (95% CI, 3.8–4.6) g·kg−1 for WT and Myd88–/– mice, respectively, and more than 4.5 g·kg−1 (the highest dose used) for Tlr4–/– mice.

image

Figure 1. Mice deficient in TLR4 or MyD88 had an enhanced ability to recover from the behavioural effects associated with alcohol treatment. Tlr4–/– mice and Myd88–/– mice both had shorter duration of alcohol-induced LORR (sedation, A, n= 6–13), and recovered from alcohol-induced deficits in rotarod performance quicker (motor dysfunction, B, n= 6–9). Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Download figure to PowerPoint

Mice deficient in TLR4 or MyD88 recover more quickly from alcohol-induced motor impairment.  As shown in Figure 1B, the latency to fall-off the rotarod decreased from 180 s before treatment, to less than 3 s in all treatment groups after alcohol administration (2.0 g·kg−1), with a gradual improvement over the monitoring time.

Both Tlr4–/– and Myd88–/– mice displayed a shorter recovery time from alcohol-induced decreases in rotarod performance compared with WT mice (Tlr4–/–, P= 0.002, n= 9, two-way anova; P < 0.001 at 20 min, and P < 0.05 at 30, 40 and 50 min with Bonferroni's post hoc test; Myd88–/–, P= 0.030, n= 6, two-way anova; P < 0.001 at 20 min and P < 0.05 at 30, 40 and 50 min with Bonferroni's post hoc test; WT, n= 9; Figure 1B).

Alcohol-induced behavioural changes are reduced by (+)-naloxone treatment

Shorter duration of alcohol-induced LORR in (+)-naloxone-treated WT, but not in Tlr4–/– or Myd88–/– mice.  Administration of (+)-naloxone (10 or 60 mg·kg−1) in WT mice significantly reduced the duration of alcohol-induced (3.5 g·kg−1) LORR when compared with control untreated animals (P < 0.0001, one-way anova; 10 mg·kg−1, 40 ± 8 min, P < 0.01, n= 6, and 60 mg·kg−1, 32 ± 6 min, P < 0.001, n= 10, vs. saline, 73 ± 3 min, n= 13, with Bonferroni's multiple comparison test; Figure 2A). Conversely, however, (+)-naloxone treatment (60 mg·kg−1) in Tlr4/– mice (n= 8) or Myd88/– mice (n= 5) did not reduce alcohol-induced (4.5 g·kg−1) sedation (P= 0.72, two-way anova) as compared with untreated null mutant mice (n= 8), respectively (data not shown).

image

Figure 2. The treatment with (+)-naloxone significantly reduced the duration of LORR (sedation, A, n= 6–13) and deficits in rotarod performance (motor dysfunction, B, n= 9) in WT mice after a dose of alcohol. Data are presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Download figure to PowerPoint

Shorter recovery time from alcohol-induced motor impairment in (+)-naloxone-treated WT mice.  The alcohol-induced (2 g·kg−1) decrease in rotarod performance was reduced by (+)-naloxone treatment (60 mg·kg−1, n= 9) in WT mice when compared with controls (n= 9) that only received alcohol (P= 0.007, two-way anova; P < 0.001 at 20 and 30 min, P < 0.01 at 40 min and P < 0.05 at 50 min with Bonferroni's post hoc test; Figure 2B).

Cellular IκBα protein levels are differentially regulated by alcohol and (+)-naloxone in WT, but not in Tlr4–/– or Myd88–/–, mixed hippocampal cells in vitro

In WT cells (n= 3), IκBα protein levels were significantly increased by 30 min of alcohol exposure (P < 0.05, Bonferroni's post hoc test) or (+)-naloxone exposure (P < 0.01, Bonferroni's post hoc test), separately. A significant interaction between alcohol and (+)-naloxone treatments was also observed (P= 0.002). In the presence of both (+)-naloxone and alcohol, however, IκBα levels were decreased (P < 0.01; Figure 3A and B).

image

Figure 3. Cellular IκBα protein levels were differentially regulated by alcohol and (+)-naloxone in WT (A and B) but not in TLR4 deficient (C and D) or MyD88 (E and F) deficient mixed hippocampal cell samples in vitro. Band densities are presented graphically as a percentage of the density of the loading control β-actin. In WT groups (A and C), the relative IκBα levels were increased 30 min after the addition of alcohol, and in the presence of (+)-naloxone, IκBα levels were decreased 30 min after alcohol exposure, while (+)-naloxone alone increased IκBα levels (n= 3). Data are presented as mean ± SEM. *P < 0.05, **P < 0.01.

Download figure to PowerPoint

In cells from Tlr4/– mice and Myd88–/– mice (both n= 3), alcohol and/or (+)-naloxone exposure did not change IκBα levels when analysed by repeated two-way anova followed by Bonferroni's post hoc test (Tlr4–/–: alcohol, P= 0.57 (+)-naloxone, P= 0.63, interaction, P= 0.51; Myd88–/–: alcohol, P= 0.80 (+)-naloxone, P= 0.53, interaction, P= 0.091; P > 0.05 by Bonferroni's post hoc test; Figure 3C–F).

Acute alcohol stimulation does not change brain p38, JNK or ERK phosphorylation in vitro or ex vivo

On analysing mixed hippocampal cell samples treated with alcohol in vitro, we observed no significant difference in the phosphorylation of cell signalling proteins (JNK, ERK and p38) between control and alcohol-treated hippocampal cells (phosphorylated JNK, P= 0.33; phosphorylated ERK, P= 0.84; phosphorylated p38 was below the limit of detection of CBA), or between WT and Tlr4–/– cells (phosphorylated JNK, P= 0.10; phosphorylated ERK, P= 0.73). However, total p38 was significantly higher in Tlr4–/– than WT cells (genotype, P= 0.0043; alcohol treatment, P= 0.22; fluorescence intensity: Tlr4–/–, 1006 ± 44; WT, 569 ± 33). Data were tested by two-way anova.

In the ex vivo brain samples, lower phosphorylated JNK was observed in hippocampus of Tlr4–/– mice when compared with WT (genotype, P= 0.023; alcohol treatment, P= 0.62). In contrast, no significant difference was found in phosphorylated p38, phosphorylated ERK or total p38 in hippocampal samples (genotype, P= 0.23, 0.061, and 0.057, respectively; alcohol treatment, P= 0.60, 0.31, and 0.42, respectively). No significant effect of alcohol or of genotype was observed in phosphorylated p38, phosphorylated JNK, phosphorylated ERK or total p38 in cerebellum samples (alcohol treatment, P= 0.17, 0.78, 0.64, and 0.98, respectively; genotype, P= 0.86, 0.34, 0.45, and 0.88, respectively) (data not shown).

Genetic deficiency of TLR4 or MyD88, or (+)-naloxone treatment do not influence peripheral or brain alcohol pharmacokinetics

No significant differences in serum or brain alcohol concentrations (n= 4–5) were observed between (+)-naloxone-treated [(+)-naloxone/WT] and saline-treated WT mice (saline/WT) (serum, P= 0.35; brain, P= 0.24), or between WT mice and either Tlr4–/– (saline/Tlr4–/–) (serum, P= 0.053; brain, P= 0.099) or Myd88–/– mice (saline/Myd88–/–) (serum, P= 0.075; brain, P= 0.15) (Figure 4A and C).

image

Figure 4. Genetic deficits of TLR4 or MyD88 or (+)-naloxone treatment did not influence peripheral or brain alcohol pharmacokinetics. Serum and brain alcohol concentrations were assessed 0.25, 1, 2 and 3 h after alcohol exposure, and no significant differences were observed in either serum samples (A) or brain samples (C) collected from Tlr4–/– mice, Myd88–/– mice, (+)-naloxone-treated WT mice or saline-treated WT controls. At the time of the LORR test following the dose of alcohol, serum (B) and brain (D) alcohol concentrations were not significantly different across the groups (n= 4–5). Data are presented as mean ± SEM.

Download figure to PowerPoint

The serum alcohol AUC values of saline/WT (+)-naloxone/WT, saline/Tlr4–/– and saline/Myd88–/– groups were 9.8, 10.1, 9.2 and 9.3 mg·mL−1·h, respectively, and the slopes of the concentration-time curves were −0.47 ± 0.06, −0.46 ± 0.08, −0.45 ± 0.10 and −0.67 ± 0.06 h−1, respectively. The brain alcohol AUCs of saline/WT, (+)-naloxone/WT, saline/Tlr4–/– and saline/Myd88–/– groups were 345, 398, 327 and 387 mg·100 mg−1·h, respectively, and the slopes of the concentration-time curves were −0.18 ± 0.05, −0.25 ± 0.10, −0.14 ± 0.07 and −0.24 ± 0.08 h−1, respectively.

Serum and brain samples (n= 4–5) were also collected when the mice awoke after each treatment. Significant differences in the alcohol concentrations were found in serum samples (P= 0.019; 95% CI, 392–469, 430–490, 368–434 and 399–574 mg·100 mL−1 in saline/WT (+)-naloxone/WT, saline/Tlr4–/–, and saline/Myd88–/– groups, respectively; Figure 4B), but not in brain samples (P= 0.88; 95% CI, 106–206, 80–187, 62–234 and 59–257 mg·100 mg−1 in saline/WT (+)-naloxone/WT, saline/Tlr4–/–, and saline/Myd88–/– groups, respectively; Figure 4D) tested with one-way anova. However, no significant difference was observed between the null mutant groups or (+)-naloxone-treated WT group and saline-treated WT controls with Bonferroni's multiple comparison test.

Discussion and conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusions
  7. Acknowledgements
  8. Conflicts of interest
  9. References

The current study shows that inhibition of acute alcohol-induced pro-inflammation through the use of mice with a genetic deficiency in TLR4 or MyD88, or treatment with the TLR4 antagonist (+)-naloxone, was successful in attenuating acute alcohol-induced sedation and motor dysfunction in mice, as measured by duration of LORR and rotarod performance, respectively. These behavioural actions were unlikely to result from changes in the peripheral or central pharmacokinetics of alcohol. In addition, we demonstrated, at the cellular level, that IκBα protein levels are elevated in response to 30 min of alcohol exposure in mixed hippocampal cells from WT mice, but not in those from Tlr4–/– or Myd88–/– mice. However, acute alcohol exposure did not alter p38, JNK and ERK phosphorylation in vitro or ex vivo. These results provide a mechanistic hypothesis underlying the behavioural observations. Together, these findings suggest that alcohol is able to induce rapid modification of pro-inflammatory mediator signalling within the brain through the TLR4–MyD88 pathway and subsequently alter animal motor behaviour.

Acute alcohol exposure activates the TLR4–MyD88–NF-κB signalling pathway in the brain

Although brain TLR4 signalling, including MAPK and NF-κB pathways, has been demonstrated to be activated in vitro after acute alcohol exposure (Blanco et al., 2005; Fernandez-Lizarbe et al., 2009), as well as in vivo and ex vivo with chronic models (Valles et al., 2004; Alfonso-Loeches et al., 2010; Liu et al., 2011; Pascual et al., 2011), it is still not known whether this effect mechanistically contributes to the acute behavioural effects induced by alcohol. In this study, we have gone one step further by demonstrating that such signalling can occur after even one dose of alcohol. Importantly, our data indicated that the TLR4 signalling in vivo occurs rapidly, as the robust difference between the WT and null mutant groups started 20 min after alcohol administration in rotarod tests and after about 30 min in LORR tests.

To further explore the link between our behavioural findings and TLR4–MyD88 signalling, we analysed a number of cell signalling proteins that could be up-regulated by TLR4 signalling in the cerebellum and hippocampus. The cerebellum was chosen as it is generally considered to control motor activity (Valenzuela et al., 2010) in the brain regions influenced by alcohol (Vilpoux et al., 2009), and we assessed the modification of motor function by alcohol in this study. The hippocampus was investigated since hippocampal microglial activation was induced by adolescent binge alcohol exposure in rats (McClain et al., 2011). As attenuation of microglia, the prime component of the brain's immune system (Streit et al., 2004), inhibited acute alcohol-induced sedation in mice (Wu et al., 2011), the activation of TLR4–MyD88–NFκB signalling may occur in microglia.

Thus, due to the rapid activation of TLR4 signalling by alcohol suggested from the behavioural data, we assessed the phosphorylation of p38, JNK and ERK in MAPK pathway ex vivo in hippocampal or cerebellum tissue as well as in mixed hippocampal cells in vitro following alcohol exposure in an attempt to delineate the mechanism responsible. However, we found that acute alcohol exposure did not affect either p38, JNK or ERK phosphorylation, which differs from previous reports using chronic alcohol treatment ex vivo (Valles et al., 2004; Alfonso-Loeches et al., 2010) and fetal microglial or astrocyte cultures in vitro (Blanco et al., 2005; Fernandez-Lizarbe et al., 2008). This implies that non-MAPK signalling cascades, such as phosphoinositide 3 kinase (PI3K)/AKT pathways (Hua et al., 2007), may be involved in the acute alcohol-induced signalling downstream from TLR4. Recently, it was found that acute alcohol challenge induced a robust AKT phosphorylation in mouse striatum (Bjork et al., 2010), further highlighting the involvement of the non-MAPK pathways. It is possible that the disparity between our findings and those from previous studies may be related to different phenotypes between adult and neonatal glia (Beauvillain et al., 2008). Nonetheless, it is important to note that the concentration of alcohol (50 mM) used in all of the in vitro experiments is based on the maximum serum (85–100 mM) and brain (30–35 mM) alcohol concentrations observed in our pharmacokinetic study, which also show maximal activity in activating immune signalling in glial cells (Blanco et al., 2005; Fernandez-Lizarbe et al., 2008).

Furthermore, IκBα protein levels were determined in vitro in mixed hippocampal cells from WT, Tlr4–/– and Myd88–/– mice. Our previous study demonstrated that alcohol-induced cellular IκBα protein levels changed in a time-dependent manner with an increase at 15 and 30 min, and a decrease at 45 and 60 min following alcohol exposure in WT mouse mixed hippocampal cells (Wu et al., 2011). The time point of 30 min was chosen to match the behavioural response we observed, and we hypothesized that the increased IκBα protein levels following 30 min of alcohol exposure might be as a result of NF-κB activation leading to IκBα protein stabilization, free IκBα from nuclear NF-κB, or increased transcription of IκBα mRNA (Scott et al., 1993; Ferreiro and Komives, 2010). In this study, we have shown that the elevated cellular IκBα protein levels by alcohol in WT cells were not observed in cells from Tlr4–/– or Myd88–/– mice. As IκBα is the main inhibitory protein of NF-κB (Sun et al., 1993), these results imply that acute alcohol exposure induces a modification to the NF-κB cascade following activation of TLR4–MyD88 signalling. In addition, the elevated pro-inflammatory cytokine levels, such as TNF-α, IL-1β and IL-6, in the brains of WT mice seen after chronic alcohol treatment (Alfonso-Loeches et al., 2010), may also be due to alcohol-induced TLR4–NF-κB activation.

Collectively, the current results demonstrate that both a binge drinking dose (3.5 and 4.5 g·kg−1) and a lower moderate dose (2.0 g·kg−1) of alcohol rapidly activates pro-inflammatory signalling cascades within the brain, which appear to be critical to alcohol-induced sedation and motor impairment through activation of TLR4–MyD88-dependent signalling and NF-κB. The possible mechanisms between this immune activation and behavioural effects of alcohol are discussed below. It has been hypothesized that the acute activation of NF-κB leads to the release of pro-inflammatory cytokines, which in turn could modulate neuronal activity in the brain, although the mechanism by which this modulation occurs is only beginning to be understood (Ren and Dubner, 2008). Interestingly, IL-1β signalling, which was activated by acute alcohol administration in our previous study (Wu et al., 2011), drove excitotoxic motor neuron injury (Prow and Irani, 2008). Furthermore, chemokine (C-X-C motif) ligand 12 (CXCL12) may enhance GABA synaptic activity at 5-HT neurons in rats (Heinisch and Kirby, 2010). Therefore, cytokines and chemokines could alter neuronal receptor functions, and these actions raise the possibility that pro-inflammatory mediators could facilitate the activation of GABAA receptors by acute alcohol exposure (Ikonomidou et al., 2000; Mukherjee et al., 2008). Thus, apart from directly acting on neurons, alcohol could modify neuronal receptor signalling indirectly via immune signalling activation, and subsequently induce sedation and motor behaviours.

Alcohol-induced behavioural changes are protected by (+)-naloxone treatment

Signalling by TLR4 occurs in response to both clinically employed opioid antagonists [(–)-isomers] and their non-opioid receptor (+)-isomers (Hutchinson et al., 2010b). In this study, we showed firstly, that in contrast to WT mice, there is no effect of (+)-naloxone treatment in the LORR test when mice are deficient in TLR4 or MyD88. This is consistent with the specificity of (+)-naloxone for the TLR4–MyD88 signalling cascade. Secondly, (+)-naloxone induced an increase in IκBα protein levels 30 min following the initial (+)-naloxone exposure, indicating that the mechanism of (+)-naloxone action may be related to interference of IκBα protein synthesis or degradation. Thirdly, this alteration in IκBα protein levels by (+)-naloxone was TLR4–MyD88-dependent. To maintain physiological relevance, the (+)-naloxone concentration in our in vitro experiments was equivalent to the blood (–)-naloxone concentrations in a previous rodent pharmacokinetic study (Kleiman-Wexler et al., 1989), as there was a paucity of (+)-naloxone pharmacokinetic data available at the time of this study.

Behavioural changes are not the result of modified alcohol pharmacokinetic profiles in null mutant or (+)-naloxone-treated animals

To confirm that the behavioural changes induced by (+)-naloxone and genetic deficiency of either TLR4 or MyD88 were not simply a result of modifying the peripheral or central pharmacokinetics of alcohol, we measured alcohol concentrations following the dosing regimens used in the LORR tests (3.5 g·kg−1 of alcohol). Overall, neither (+)-naloxone treatment nor the absence of TLR4 or MyD88 altered alcohol concentrations in either serum or brain samples.

Because of the decreased alcohol pharmacodynamic responses and unchanged alcohol pharmacokinetics in TLR4 signalling attenuated groups compared with controls, we expected that mice which awoke earlier in the LORR test would have higher peripheral and brain alcohol concentrations following their awakening. However, there was no significant difference in serum or brain alcohol concentrations between groups at the time of waking from alcohol-induced sedation, which may be due to the shallow slopes of the alcohol concentration-time curves.

TLR4–MyD88 signalling plays a pivotal role in the acute behavioural actions of alcohol

Amongst the acute behavioural effects of alcohol, sedation and motor inco-ordination are probably responsible for traffic accident-related deaths in humans and accompany self-administration of alcohol in mice (Chuck et al., 2006). Thus, our results not only suggests that the initial effects of alcohol are related to TLR4 signalling but also may have important clinical applications in binge drinking-related brain conditions and alcohol dependence, which may culminate in preventing traffic accidents and decreasing the social burden of alcohol abuse.

In conclusion, the current study provides new evidence linking the contribution of TLR4–MyD88-dependent signalling to the behavioural response induced by acute alcohol administration. The consequences of blocking TLR4 signalling that support this theory include inhibition of the influence of alcohol on IκBα protein levels and a reduction in the sedative and motor effects of alcohol. Therefore, novel pharmacological strategies targeting TLR4 signalling, such as (+)-naloxone, may have an important and highly relevant clinical application. The use of TLR4 antagonists would potentially also reduce alcohol-induced peripheral TLR4 signalling in the liver and gut (Szabo and Bala, 2010).

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusions
  7. Acknowledgements
  8. Conflicts of interest
  9. References

YW is the recipient of China Scholarship Council (CSC)-University of Adelaide Joint Postgraduate Scholarship. MRH is a NHMRC CJ Martin Fellow (ID 465423; 2007–2010) and an Australian Research Council Research Fellow (DP110100297). JKC is a FTT Fricker Research Fellow (Medical Endowment Funds), Faculty of Health Sciences, University of Adelaide. EL was supported by the Cancer Council of South Australia through provision of a PhD stipend. The authors thank Prof Sarah Robertson for her technical assistance on flow cytometry and helpful comments on the manuscript and the reviewers for constructive suggestions. This work was partially supported by the NIH Intramural Research Programs of NIDA and NIAAA and NHMRC.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion and conclusions
  7. Acknowledgements
  8. Conflicts of interest
  9. References