|Home | About | Journals | Submit | Contact Us | Français|
Alcoholism is a chronically relapsing condition, indicative of long-term neuronal adaptations maintaining the disease even after prolonged abstinence. Previously we identified brain-derived neurotrophic factor (BDNF) in the dorsal striatum as the central mediator of a homeostatic mechanism which is activated by acute alcohol (ethanol) exposure and functions to decrease the sensitivity of rodents to ethanol-related behaviors. We hypothesized that extensive exposure to ethanol would result in dysregulation of this BDNF-mediated protective mechanism, accompanied by heightened ethanol intake. Here we demonstrate that while a single bout of ethanol intake increases BDNF mRNA expression in the dorsal striatum, this effect is no longer observed after 6 weeks of daily ethanol access. Additionally, 6 weeks of ethanol consumption decreases BDNF in the cortex, a main source of BDNF for the striatum. Importantly, these ethanol-induced changes in BDNF levels are not ameliorated by 2 weeks’ abstinence. Together, these data suggest that the BDNF pathway, which is activated following a single bout of ethanol drinking, breaks down by the end of 6 weeks of access and does not recover its protective function after a 2-week deprivation period. These results suggest that the persistence of altered BDNF signaling may contribute to the inflexibility of addictive behaviors.
BDNF is a neurotrophic factor expressed throughout the nervous system, with highest expression in the cortex and hippocampus (Hofer et al. 1990). BDNF is secreted by neurons in an activity-dependent fashion (Balkowiec & Katz 2000), thereby initiating signaling via binding to its receptor, TrkB. This results in the activation of three main downstream signaling cascades, namely the mitogen-activated protein kinase (MAPK), phosphatidylinositol-3-OH kinase (PI3K) and phospholipase C-γ (PLC-γ) pathways (Huang & Reichardt 2003). Activity-dependent BDNF signaling is thought to be involved in synaptic remodeling (Lu 2003), a process central to many of the known roles for BDNF, including synaptic plasticity (Bramham & Messaoudi 2005), learning and memory (Yamada et al. 2002, Tyler et al. 2002) and addiction (Russo et al. 2008). Interestingly, BDNF has been linked to pathways that negatively regulate the adverse actions of ethanol, and several lines of evidence from human studies suggest that genetic changes which reduce the function of the BDNF pathway are more prevalent in alcoholics than in the general population. For example, the region of chromosome 11 containing BDNF has been implicated in susceptibility to developing addiction to multiple drugs of abuse, including alcohol (Uhl et al. 2001). A similar association has been shown for the gene encoding TrkB in alcoholics with antisocial personality disorder (Xu et al. 2007). In addition, a polymorphism in the BDNF gene has been associated with an earlier onset of alcohol abuse (Matsushita et al. 2004), while decreased plasma BDNF levels have been observed in alcoholics, particularly those with a family history of alcoholism, than in nonalcoholic controls (Joe et al. 2007). Taken together, these data suggest a role for BDNF in preventing the development of alcoholism.
Accumulating evidence also demonstrates a role for BDNF in negatively regulating ethanol intake in rodents. Specifically, we and others have observed that transgenic mice with decreased BDNF levels show elevated ethanol intake (Hensler et al. 2003, Pandey et al. 2004, McGough et al. 2004). Similarly, inhibition of TrkB activation increases ethanol consumption (Jeanblanc et al. 2006), whereas global upregulation of BDNF levels reduces ethanol intake in mice (McGough et al. 2004). In addition, rat strains predisposed to drink high levels of ethanol show reduced BDNF expression as compared to their lower-drinking counterparts (Prakash et al. 2008, Yan et al. 2005), and modulation of BDNF levels in the amygdala by exogenous application of BDNF or by antisense inhibition of endogenous BDNF yield decreased or increased ethanol intake, respectively (Pandey et al. 2006). Together these data support a role for BDNF in controlling ethanol intake.
Previously we showed that BDNF in the dorsal striatum is a central mediator of a homeostatic pathway regulating ethanol intake (Jeanblanc et al. 2006, McGough et al. 2004, Logrip et al. 2008). Specifically, we found that acute ethanol treatment of striatal neurons leads to increased BDNF expression, resulting in the activation of the BDNF-mediated signaling pathway (McGough et al. 2004, Logrip et al. 2008). We further showed that moderate ethanol consumption by mice results in the upregulation of BDNF expression specifically in the dorsal striatum (McGough et al. 2004). Finally, we demonstrated that BDNF decreases ethanol intake via increasing the expression of downstream effectors such as the dopamine D3 receptor and dynorphin in the dorsal striatum (Jeanblanc et al. 2006, McGough et al. 2004, Logrip et al. 2008). Together these results suggest that BDNF in the dorsal striatum is part of a homeostatic protective pathway in which ethanol acutely increases BDNF expression and signaling to regulate subsequent ethanol intake.
Here we set out to investigate whether extensive ethanol experience results in the dysregulation of the dorsal striatal BDNF system. Specifically, we examined the regulation of BDNF expression in the dorsal striatum following voluntary ethanol intake as a function of the length of ethanol experience using a modified version of a limited access paradigm which has been shown to cause drinking to intoxication (Rhodes et al. 2007, Rhodes et al. 2005). We hypothesized that a single 4-h bout of ethanol drinking would result in elevated levels of BDNF in the dorsal striatum. Conversely, we hypothesized that extensive experience with ethanol would cause a breakdown in the homeostatic regulation of BDNF, a change which may persist even after significant abstinence from ethanol.
TRIzol and the Reverse Transcription System were purchased from Promega Corporation (Madison, WI, USA). DNase was obtained from Sigma Aldrich (St. Louis, MO, USA). All real-time PCR reagents, including TaqMan Gene Expression Assays, were obtained from Applied Biosystems, Inc. (Foster City, CA, USA).
Male C57BL/6J (C57) mice were obtained at 6–8 weeks of age from The Jackson Laboratory (Bar Harbor, ME). Animals were housed under a 12 h light/dark cycle, with lights off at 10:00 a.m. and lights on at 10:00 p.m., and were provided with continuous ad libitum access to food and water. Due to the reversal of normal light/dark cycle, animals were given 2 weeks to adjust to the new housing conditions prior to performance of any study procedures, as altered BDNF expression in various brain regions has recently been shown to last for up to a week after a single light phase shift (Katoh-Semba et al. 2008). All animal procedures were approved by the Gallo Center Institutional Animal Care and Use Committee and were conducted in agreement with the Guide for the Care and Use of Laboratory Animals, National Research Council (1996).
The paradigm used for limited access was adapted as a 2-bottle choice version of the limited access procedures described by Rhodes et al. (2005). Mice were singly housed in double grommet cages at least two weeks prior to the initiation of ethanol access. Fluid solutions were provided in cylindrical glass bottles equipped with standard stoppers. Mice were given two bottles – one containing water and the other containing either water, a solution of 10% ethanol in water (v/v) or a solution of 2% sucrose in water (w/v) – for 4 h each day, with access beginning 2 h into the dark cycle. A single water bottle was available the other 20 h. Bottles were weighed immediately before and after the 4-h access period, and the placement order of the bottles on the home cage was switched daily to ensure that consumption levels were not due to a side bias. Body weights were monitored every 6 days; the final weight was taken 36 h prior to sacrifice to ensure minimal disruption of consumption due to handling stress.
Mice were sacrificed immediately or 24 h after the final access session, and brain regions were isolated by microdissection and flash frozen to minimize RNA degradation. In order to maintain consistency in length of ethanol access, onset of access on the final day was staggered over a 1-h period. Order of sacrifice was counterbalanced by condition such that no systematic differences existed in the time of day at which brain tissue was taken for the water and ethanol groups.
The microdissection procedure was designed according to Paxinos and Franklin (2004), yielding 4 subdivisions of cortex, as well as dorsal and ventral subdivisions of the striatum. Prefrontal cortex consisted of the most anterior 2 mm of the cortex, dorsal to the olfactory bulbs, subdivided along the horizontal plane dorsal to orbitofrontal cortex. The tissue isolated for frontal cortex spanned the next 2 mm in the anterioposterior direction, dorsal to the corpus callosum, excluding the most ventral 2 mm. Posterior cortex samples were isolated from the remaining cortical regions posterior to frontal cortex, with the exclusion of the most ventral 1 mm. Dorsal striatum and nucleus accumbens were isolated via tissue punch of a coronal slice 2 mm wide in the anterior-posterior direction, beginning immediately posterior to the prefrontal cortex.
Brain sections were homogenized in TRIzol and mRNA isolated according to standard protocol. Samples were treated with DNase prior to reverse transcription using the Reverse Transcription System. The resulting cDNA samples were amplified by TaqMan quantitative PCR using commercially available primer/probe kits from Applied Biosystems for BDNF (Gene Expression Assay Mm00432069_m1) and GAPDH (Gene Expression Assay Mm99999915_g1), the internal control. The BDNF assay targets BDNF transcript 4, as designated by Liu et al. (2006), spanning the intron boundary between the 5’ UTR Exon IV (formerly Exon III) and the single protein coding Exon VIII (formerly Exon V).
Statistical analyses were performed using SigmaStat statistical software (Systat, San Jose, CA). Ethanol intake and preference data were analyzed by one-way ANOVA, with repeated measures when indicated. Gene expression data were analyzed by one-way ANOVA. All post-hoc analyses were performed using the Student-Newman-Keuls test.
We first assessed ethanol consumption using a 2-bottle choice limited access paradigm adapted from the dark cycle limited access procedure described by Rhodes et al. (2005). As demonstrated in Figure 1A, a single session of ethanol access under this paradigm results in significant ethanol consumption by C57 mice (left panel; mean g/kg +/− S.E.M.: 5.6 +/− 0.8). However, on this initial access day mice displayed only a slight preference for the ethanol-containing bottle, which was not statistically different from the equal sampling performed by mice provided with 2 water bottles, as analyzed by one-way ANOVA (Fig. 1A, right panel; no effect of Bottle, F(1,21) = 1.95, p = 0.18). After six weeks of daily limited access to ethanol, mice show a significant increase in both ethanol intake (Fig. 1B, left panel; main effect of Time, F(1,19) = 17.90, p < 0.001) and preference for the ethanol bottle over the water bottle (Fig. 1B, right panel; main effect of Time, F(1,19) = 17.33, p < 0.001). This level of ethanol consumption is comparable to the values obtained with limited access to forced ethanol, which yielded blood ethanol levels in the range of 100 mg/dl (Rhodes et al. 2005). These data suggest that voluntary ethanol intake under the limited access paradigm results in consumption of pharmacologically relevant quantities of ethanol in C57 mice. Importantly, under this paradigm ethanol intake is significant even upon first encounter and escalates over time.
To demonstrate specificity of effects for ethanol, separate groups of C57 mice were provided with 2-bottle choice limited access to 2% sucrose, an alternative orally administered reinforcer. Interestingly, unlike ethanol, high sucrose intake upon first exposure (Fig. 1C, left panel; mean ml/kg +/− S.E.M.: 123.4 +/− 6.8) was coupled with high preference for the sucrose bottle (Fig. 1C, right panel; main effect of Bottle, F(1,18) = 39.87, p < 0.001). After 6 weeks of daily limited access, we observed a significant increase in sucrose intake (Fig. 1D, left panel; main effect of Time, F(1, 9) = 10.70, p < 0.05), as observed for ethanol. However, the change in preference following 6 weeks’ access was not significant (Fig. 1D, right panel; no effect of Time, F(1,9) = 4.07, p = 0.08), and is likely to be due to the high preference observed on first presentation of sucrose. Together, these data demonstrate that 4-h daily access to ethanol or sucrose results in high levels of intake for both reinforcers during the initial access period, with intake increasing over time for both reinforcers, as observed after extended exposure.
Previously we have shown that both ethanol intake under a standard 24-h access 2-bottle choice paradigm and a single acute ethanol injection increase BDNF expression in the dorsal striatum (McGough et al. 2004). Therefore, we hypothesized that a single 4-h session of voluntary ethanol consumption would similarly increase BDNF expression in the dorsal striatum. To test this, C57 mice were given 4 h 2-bottle choice ethanol or water access, at the end of which brains were collected immediately. As shown in Figure 2A (left), we observed a significant increase in BDNF expression in the dorsal striatum following acute ethanol intake, as compared with water controls (main effect of Treatment, F(1,21) = 4.59, p < 0.05). This elevation in BDNF mRNA was specific to the dorsal striatum, as no change was observed in the nucleus accumbens (Fig. 2A, right panel; no effect of Treatment, F(1,21) = 0.49, p = 0.49) or in the overlying cortex (mean BDNF/GAPDH +/− S.E.M. intake Water Only: 1.70 +/− 0.07; Ethanol Available: 1.57 +/− 0.09; no effect of Treatment, F(1,21) = 1.35, p = 0.26). Importantly, this increase in dorsal striatal BDNF was specific for ethanol reinforcement, as 4 h access to 2% sucrose intake had no effect on BDNF levels in the dorsal striatum (Fig. 2B; no effect of Treatment, F(1,17) = 0.03, p = 0.87). Together these data demonstrate that a single bout of high-level ethanol drinking results in the elevation of BDNF expression in the dorsal striatum, and that this effect is specific to both the brain region and the reinforcer.
Given the ability of an acute bout of ethanol intake to increase dorsal striatal BDNF, and the proposed function of BDNF in the dorsal striatum as a homeostatic regulator of ethanol intake (Jeanblanc et al. 2006, McGough et al. 2004, Logrip et al. 2008), we hypothesized that the progression to ethanol dependence would result from the breakdown in BDNF homeostasis. Specifically, loss of ethanol-induced BDNF expression would be associated with escalating ethanol intake. As we observed increased ethanol intake after 6 weeks daily limited access to ethanol (Fig. 1B), we hypothesized that dorsal striatal BDNF levels would no longer be elevated following ethanol intake at that same time point. As shown in Figure 3A, in C57 mice given 6 weeks limited access to ethanol we found no change in BDNF expression in the dorsal striatum after the final ethanol intake session, as determined by one-way ANOVA (no effect of Treatment, F(1,17) = 0.40, p = 0.54). As expected, we also saw no difference in BDNF levels in the nucleus accumbens (no effect of Treatment, F(1,18) = 0.60, p = 0.45). Surprisingly, we observed a striking decrease in BDNF expression throughout the cortex, as demonstrated in Figure 3B. Comparison of water and ethanol groups showed main effects of Treatment in all cortical subregions except the dorsal prefrontal cortex, where the decrease almost attained significance (dorsal prefrontal cortex: F(1,18) = 3.60, p = 0.07; ventral prefrontal cortex: F(1,18) = 5.56, p < 0.05; frontal cortex: F(1,18) = 6.22, p < 0.05; posterior cortex: F(1,18) = 14.05, p < 0.005). In addition, this alteration in BDNF expression was dependent on proximal ethanol intake, rather than an alteration in basal BDNF levels, as 24 h withdrawal after the final ethanol access session resulted in BDNF levels indistinguishable from water controls (mean BDNF/GAPDH +/− S.E.M. 24 h after final ethanol access: dorsal prefrontal cortex: 1.80 +/− 0.08; ventral prefrontal cortex: 1.67 +/− 0.06; frontal cortex: 1.68 +/− 0.09; posterior cortex: 1.76 +/− 0.09; as compared to Water Only in Fig. 3B, no effect of Treatment, F’s < 0.41, p’s > 0.53 for all cortical subdivisions). As with a single day’s ethanol access, these changes in gene expression were specific to ethanol as a reinforcer, since 6 weeks of 2% sucrose consumption did not alter cortical BDNF expression (frontal cortex: mean BDNF/GAPDH +/− S.E.M. Water Only: 1.99 +/− 0.12; Sucrose Available: 2.26 +/− 0.10; F(1,18) = 2.98, p = 0.10). Taken together, these results demonstrate that 6 weeks of ethanol intake under a 4-h daily limited access schedule results not only in a breakdown in the homeostatic upregulation of dorsal striatal BDNF, but a concomitant decrease in BDNF levels throughout the cortex which is linked to the preceding ethanol drinking episode.
The chronically relapsing nature of alcoholism suggests long-lasting neuroadaptations which control ethanol intake, even after prolonged periods of sobriety. Thus, we next investigated whether corticostriatal BDNF expression would recover its acute response to ethanol after a 2-week period of deprivation. To test this, C57 mice were given 6 weeks daily limited access to 2-bottle choice ethanol, 2 weeks deprivation, then a single session of renewed access to ethanol or continued deprivation. We observed no recovery of corticostriatal ethanol-induced BDNF expression following a single session of renewed ethanol access after a 2-week deprivation. Specifically, ethanol failed to alter BDNF levels in the dorsal striatum (mean BDNF/GAPDH +/− S.E.M. Water Only: 0.18 +/− 0.08; Continued Deprivation: 0.10 +/− 0.04; Renewed Access: 0.26 +/− 0.07, no effect of Treatment, F(2,21) = 1.51, p = 0.24), but continued to depress BDNF expression throughout the cortex, as shown in Figure 4. Analysis by ANOVA determined that this decrease was statistically significant in the frontal and posterior cortical regions (main effects of Treatment, frontal cortex: F(2,20) = 7.23, p < 0.005; posterior cortex: F(2,21) = 5.41, p < 0.05), while the decrease in the prefrontal regions was marginally above the threshold for significance (no effect of Treatment, dorsal prefrontal cortex: F(2,21) = 2.88, p = 0.08; ventral prefrontal cortex: F(2,21) = 3.39, p = 0.053). Post-hoc analyses demonstrated significant differences between the ethanol-deprived mice and those given renewed access to ethanol in both the frontal (p < 0.005) and posterior (p < 0.01) cortex. Interestingly, cortical BDNF expression appeared to be elevated by ethanol deprivation, although this effect was only statistically significant in the frontal cortex (p < 0.05, water vs. continued deprivation). In light of this, the data were re-analyzed excluding the water controls to isolate the changes in BDNF expression induced by renewed access to ethanol, as compared to the deprivation baseline. In this analysis, we observed the same pattern of ethanol-induced changes in BDNF expression seen without deprivation – that is, a single session of renewed access significantly decreased BDNF expression in ventral prefrontal, frontal and posterior cortex, as compared to continued deprivation (main effects of Treatment, ventral prefrontal cortex: F(1,14) = 5.64, p < 0.05; frontal cortex: F(1,14) = 13.39, p < 0.005; posterior cortex: F(1,14) = 9.20, p < 0.01) with marginally insignificant effects in the dorsal prefrontal cortex (no effect of Treatment, F(1,14) = 3.31, p = 0.09). Taken together, these results demonstrate that 2 weeks deprivation from daily limited access to ethanol is insufficient to produce recovery of ethanol-induced corticostriatal BDNF expression.
In this study, we investigated ethanol-induced alteration of BDNF mRNA expression throughout the corticostriatal network as a function of the length of ethanol exposure. We found that ethanol consumption during a single 4-h access session resulted in an increase in BDNF expression in the dorsal striatum. However, following 6 weeks of 4-h daily access to ethanol, during which time ethanol intake escalates, we observed a breakdown in corticostriatal BDNF expression. Specifically, we show that ethanol intake ceased to increase dorsal striatal BDNF expression, while it resulted in the downregulation of BDNF expression throughout the cortex. We also demonstrate that these alterations in corticostriatal BDNF expression are specific for ethanol, as sucrose intake did not alter BDNF expression at either the 1-day or 6-week time points. Importantly, these changes in BDNF responsiveness to ethanol are long-lasting, as 2 weeks of ethanol deprivation did not alter the pattern of BDNF expression upon renewed access to ethanol.
Previously we demonstrated that ethanol exposure leads to the elevatation of BDNF expression and signaling in the dorsal striatum, resulting in increased levels of downstream effectors, such as the dopamine D3 receptor and dynorphin, whose activity is required for downregulation of ethanol consumption by BDNF (Jeanblanc et al. 2006, McGough et al. 2004, Logrip et al. 2008). We suggested that the BDNF signaling pathway functions to homeostatically regulate ethanol intake – that is, that ethanol intake elevates BDNF levels, resulting in the activation of downstream signaling cascades which serve as feedback to limit further ethanol consumption. However, the previous studies utilized 24 h continuous 2-bottle choice ethanol access, in which moderate levels of ethanol intake are consistently maintained and animals fail to develop ethanol dependence (Dole & Gentry 1984, Dole et al. 1985). Therefore, it was of interest to test the level of BDNF in the dorsal striatum in a model of excessive ethanol consumption. Several models have recently been developed to achieve high ethanol intake in mice in a limited period of time, resulting in intoxication (Finn et al. 2005, Rhodes et al. 2007, Szumlinski et al. 2007, Rhodes et al. 2005). Under such paradigms, mice experience repeated cycles of intoxication and withdrawal, a pattern of drinking more similar to human binge drinking than normally observed under 24-h access. Using a variation of this limited access paradigm, we provide the first demonstration that a single bout of ethanol consumption triggers BDNF expression in the dorsal striatum. This has significant implications for the function of BDNF as a putative regulator of social drinking in humans since these data demonstrate that dorsal striatal BDNF provides negative feedback following only one ethanol drinking bout. Since multiple drinking episodes are not required to elevate BDNF expression, this homeostatic pathway is poised to respond to ethanol drinking in the human social drinker. The importance of this regulatory pathway in the prevention of alcoholism is supported by data demonstrating a significant correlation between mutations in the BDNF gene and the onset of alcoholism (Matsushita et al. 2004) as well as evidence for reduced plasma BDNF levels in alcoholics (Joe et al. 2007). Together these data support the theory that BDNF functions as a mediator of ethanol intake in humans, perhaps by inhibiting the progression to addiction.
Although the dorsal striatum responds acutely to ethanol via increased BDNF expression, we demonstrate here that following 6 weeks of repeated daily limited ethanol intake at levels sufficient to induce intoxication (Rhodes et al. 2005), this protective pathway becomes unresponsive. Interestingly, in addition to the breakdown in the dorsal striatal BDNF response to ethanol, we observed a decrease in cortical BDNF mRNA subsequent to the final drinking bout. This effect was particularly surprising given the lack of change in cortical BDNF following the initial ethanol drinking bout. Nonetheless, as the cortex has been considered a major source of striatal BDNF (Kokaia et al. 1993, Baquet et al. 2004), this suggests that BDNF delivered by cortical afferents also decreases, resulting in an even greater depression of striatal BDNF-TrkB-mediated signaling. Although alterations in mRNA expression were limited to the period immediately following ethanol intake (as we observed no difference 24 h later), previous data have shown decreased cortical BDNF protein following 24 h withdrawal from chronic ethanol treatment (Pandey et al. 1999). This suggests that the reduction in cortical BDNF protein may be significantly prolonged as compared to the changes in gene expression, possibly due to differences in the temporal dynamics of BDNF mRNA and protein expression, as it has been suggested that the BDNF protein has a relatively long half-life (Nawa et al. 1995). In line with this possibility, Sanna and colleagues have demonstrated a prolonged decrease in the activation of the dorsal striatal MAPK extracellular signal-regulated kinase (ERK) during withdrawal from chronic intermittent ethanol exposure, lasting at least 7 h after the cessation of intermittent ethanol exposure (Sanna et al. 2002). As we have previously shown that striatal ERK phosphorylation is downstream of BDNF signaling (Logrip et al. 2008), together these data suggest that the prolonged decrease in ERK activity reported by Sanna et al. (2002) could be the functional consequence of decreased striatal BDNF signaling. It should also be noted that, in contrast to our results, Miller and colleagues have observed increased cortical BDNF levels using chronic ethanol treatments (Miller & Mooney 2004, Bruns & Miller 2007, Miller 2004). Therefore, the temporal dynamics of altered striatal signaling downstream of BDNF after extended limited access to ethanol require further investigation.
While acute ethanol intake triggers striatal BDNF expression, which is suggested to function as a negative regulator of further ethanol intake (McGough et al. 2004), we show here that this protective pathway breaks down over time, such that dorsal striatal BDNF becomes nonresponsive and cortical BDNF, which is unaltered by acute ethanol exposure, is reduced following ethanol intake. This differential regulation of BDNF in the striatum and cortex may result from a region-specific responsiveness to ethanol. Specifically, we have previously shown that acute ethanol triggers the nuclear translocation of the scaffolding protein RACK1 (Ron et al. 2000) via a mechanism that requires the activation of the cAMP/PKA pathway (He et al. 2002). We further found that activation of the cAMP/PKA pathway leads to the induction of BDNF expression (Yaka et al. 2003a). Importantly, we observed that ethanol-mediated nuclear translocation of RACK1 in the hippocampus and dorsal striatum is required for ethanol to increase BDNF expression (McGough et al. 2004). In addition, we showed that in the hippocampus and in the dorsal striatum under basal conditions RACK1 forms a trimolecular complex with Fyn and the NR2B subunit of the N-methyl-d-aspartate (NMDA) receptor, which is dissociated upon ethanol exposure leading to Fyn-mediated phosphorylation of NR2B via a mechanism that requires the activation of the cAMP/PKA pathway (Wang et al. 2007, Yaka et al. 2003b). Importantly, the NR2B/RACK1/Fyn complex cannot be detected in the cortex, and no NR2B phosphorylation is observed upon exposure of cortical neurons to ethanol (Yaka et al. 2003b). Together, these results raise the possibility that ethanol-induced dissociation of RACK1 from its complex with Fyn and the NMDA receptor may be required for RACK1’s translocation and subsequent induction of BDNF expression in the dorsal striatum but not in the cortex. In addition, the breakdown in responsiveness of striatal BDNF to ethanol may result from cessation of this nuclear translocation. We previously observed in an immortalized cell culture line that RACK1 is excluded from the nucleus following a 48-h (chronic) ethanol treatment and that a long withdrawal period of 24 h is required to restore ethanol-induced RACK1 nuclear translocation (Vagts et al. 2003). These data indicate that long-term ethanol exposure may cause a breakdown in BDNF homeostasis via the inhibition of RACK1-mediated BDNF expression, and that recovery of RACK1 function requires a considerable duration of ethanol withdrawal. Thus determining the responsiveness of RACK1 following prolonged ethanol intake in vivo presents a promising future line of study.
While cessation of ethanol-induced RACK1 translocation, and thus BDNF expression, presents a likely candidate for the breakdown in ethanol-induced dorsal striatal BDNF expression after extensive ethanol intake, this mechanism may not be responsible for the decreases in cortical BDNF observed after extensive ethanol experience, as acute ethanol does not alter cortical BDNF levels. A second putative mechanism that has recently gained much attention in addiction research involves chromatin remodeling, which can alter the responsiveness of genetic promoters. Epigenetic regulation has recently been suggested as a putative mechanism by which chronic drug exposure yields long-lasting changes in gene expression (Renthal & Nestler 2008). Elevation of striatal histone acetylation levels, resulting in increased promoter availability, has been demonstrated for BDNF during withdrawal from chronic cocaine exposure (Kumar et al. 2005), in line with increased BDNF levels observed during cocaine withdrawal (Grimm et al. 2003). Conversely, ethanol withdrawal yields reduced histone acetylation in the amygdala (Pandey et al. 2008), which may reduce BDNF expression. In light of the data presented here, the observed decrease in BDNF expression may result from decreased histone acetylation in the cortex, similar to that observed in the amygdala. Alternatively, BDNF expression may be downregulated by DNA methylation, a mechanism causing reduced promoter availability and thus decreased transcription (Miranda & Jones 2007). This presents an intriguing putative mechanism of epigenetic regulation since BDNF expression is regulated by the methyl CpG binding protein 2 (MeCP2), which normally represses BDNF transcription but is released in a phosphorylation-dependent manner following activating stimuli (Martinowich et al. 2003, Chen et al. 2003, Zhou et al. 2006). Prenatal ethanol exposure has been shown to increase the methylation of the BDNF gene and decrease BDNF expression in the postnatal olfactory bulb (Maier et al. 1999), but whether adult ethanol exposure yields similar changes is unknown. Another putative target is MeCP2 itself, as recent work has demonstrated regulation of MeCP2 expression, and thus BDNF expression, by microRNA (miR)-132 (Klein et al. 2007), although, surprisingly, the upregulation of MeCP2 levels by inhibition of miR-132 increased BDNF expression. As ethanol elevates the levels of another miR (Pietrzykowski et al. 2008), and several miRs have been shown to posttranscriptionally regulate prefrontal cortical BDNF (Mellios et al. 2008), further investigation into ethanol-induced changes in the epigenetic regulation of cortical BDNF presents an intriguing line of future investigation.
Together, our data support a biphasic function of BDNF homeostasis in the corticostriatal network. Initially, BDNF in the dorsal striatum acts to delay or prevent the onset of addiction via its upregulation subsequent to individual acute ethanol consumption bouts. However, escalation of ethanol intake may be due to a breakdown in ethanol-induced BDNF expression, suggestive of a loss of BDNF regulation of ethanol intake. Therefore, the progression to and persistence of addiction may be coupled to dysregulation of BDNF homeostasis. Thus defining the temporal dynamics of the observed breakdown in BDNF homeostasis, particularly its duration in abstinence, as well as determining its direct relationship to ethanol intake, present critical directions for future research. Finally, these data suggest that activation of the corticostriatal BDNF pathway may prove efficacious in the treatment of alcoholism.
This work was supported by NIAAA RO1 AA016848 (D.R.) and F31 AA015462 (M.L.L.), and the State of California for Medical Research on Alcohol and Substance Abuse through the University of California, San Francisco (P.H.J. and D.R.). This work was partially sponsored by NIAAA award U01 AA013489 through UCDHSC. This work was also supported by the Department of the Army, Grant # W81XWH-05-1-0212 (D.R.) for which the U.S. Army Medical Research Acquisition Activity, 820 Chandler Street, Fort Detrick, MD 21702-5014 is the awarding and administering acquisition office. The content of the information represented does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred. We thank Sebastien Carnicella, Jerome Jeanblanc and Quinn Yowell for their contributions.