|Home | About | Journals | Submit | Contact Us | Français|
A well-established body of work indicates a crucial role for corticotropin releasing factor (CRF) in neurobiological responses associated with excessive dependence-like ethanol drinking in ethanol vapor exposed rodents. Recent evidence demonstrates a role for CRF in the modulation of binge-like ethanol consumption by non-dependent mice, a behavior which can precede ethanol dependence. The CRF circuitry that is engaged by binge-like ethanol exposure, however, is unknown. Using converging approaches, we provide evidence that, similar to ethanol vapor-induced increases in ethanol intake, CRF signaling in the central nucleus of the amygdala (CeA) is engaged during binge-like ethanol consumption by C57BL/6J mice. Specifically, we found that binge-like consumption of an ethanol solution (20% ethanol v/v) was attenuated by pretreatment with the CRF1R antagonists antalarmin, (4-ethyl-[2,5,6-trimethyl-7-(2,4,6-trimethylphenyl)-7H-pyrrolo[2,3-d]pyrimidin-4-yl]amino-1-butanol (LWH-63), and NBI-27914 at doses (30 mg/kg, i.p.) that did not alter non-binge-like ethanol consumption. Binge-like ethanol consumption resulted in significant increases of CRF immunoreactivity in the CeA immediately following ethanol drinking and 18-24 h following ethanol removal and also blocked the ability of CRF to enhance GABAergic transmission in the CeA 18-24 h following ethanol removal. Pretreatment with bilateral injections of antalarmin (1 μg/ 0.5 μl per side) into the CeA, but not the adjacent basolateral amygdala (BLA), significantly attenuated binge-like ethanol consumption. These findings suggest that CRF signaling in the CeA is recruited during excessive ethanol intake, prior to the development of dependence. We hypothesize that plastic changes in CRF signaling develop with repeated binge-like drinking episodes, contributing to the transition to dependence.
Binge drinking defines an episode in which enough alcohol is consumed to generate blood alcohol concentrations (BECs) of 80 mg/dl or greater (Council, 2004). Binge drinking is associated with numerous adverse consequences (Read et al., 2008; Stahre et al., 2009), including an increased risk of developing alcohol dependence (Jennison, 2004; McCarty et al., 2004). As such, recruitment of mechanisms thought to underlie alcohol dependence may occur during bouts of high alcohol drinking, including alcohol binges.
Corticotropin releasing factor (CRF) plays a key role in excessive dependence-like ethanol drinking in rats exposed to high levels of ethanol via vapor exposure, through activation of the G-protein coupled CRF-1 receptors (CRF1R) in the central nucleus of the amygdala (CeA) (Roberto et al., 2010). The CRF system does not significantly modulate ethanol consumption in non-dependent rats (Roberts et al., 2000), thus leading to the hypothesis that neuroadaptations within the CRF system develop over an extended course of excessive ethanol exposure, culminating in ethanol dependence (Koob, 2003, 2009). A defining feature of this hypothesis is that CRF signaling is not crucial to maintain the initial stages of ethanol consumption that precede a dependent state (Heilig et al., 2010).
An extension of this model was revealed by several recent studies from our group which have established a role for CRF signaling in maintaining binge-like ethanol consumption by non-dependent animals (Sparta et al., 2008; Lowery et al., 2010). Specifically, systemic administration of a selective CRF1R antagonist, CP-154,526, attenuated binge-like ethanol consumption by male C57BL/6J mice that achieved BECs of at least 80 mg/dl under control conditions. Interestingly, CRF1R signaling appeared to be exclusively recruited by excessive levels of ethanol intake, as pretreatment with CP-154,526 did not alter ethanol consumption by mice that drank moderate levels of ethanol with associated BECs of 40 mg/dl or less (Sparta et al., 2008). While these findings indicate that the CRF system modulates excessive binge-like ethanol drinking, the central site of action and the overlap with the CRF pathways that modulate dependence-like drinking are unknown.
The current studies were designed to test the hypothesis that the CRF neurocircuitry that underlies binge-like ethanol consumption by non-dependent animals overlaps with the CRF neurocircuitry that has been implicated in dependence-like ethanol drinking. We show that three different selective CRF1R antagonists attenuate binge-like ethanol consumption, but do not affect low level, non-binge-like ethanol consumption. Furthermore, binge-like ethanol consumption alters CRF immunoreactivity (IR) in extrahypothalamic brain regions and alters the neuromodulatory actions of CRF in the CeA of non-dependent C57BL/6J mice. We also show that binge-like ethanol consumption is attenuated by intra-CeA microinjections of a selective CRF1R antagonist, while microinjections of the antagonist in to the basolateral amygdala (BLA) do not alter binge-like ethanol consumption. Together, these results provide support for the hypothesis that high levels of ethanol intake by non-dependent animals, such as those observed during binge-like drinking, recruit CRF signaling in the CeA, a region that is critical to the development of ethanol dependence.
Male C57BL/6J mice (Jackson Laboratories, Bar Harbor, ME) were 6-8 weeks of age and weighed 20-30 grams upon arrival. Mice were housed individually in plastic cages and allowed to habituate to the environment for at least one week before experimental procedures were initiated. The animal colony room was maintained at approximately 22 degrees C with a 12h/12h light/dark cycle (lights on at 0700 h or 1300 h). Mice had ad lib access to food throughout all experiments and free access to water except during ethanol access, as noted. All procedures used are in accordance with the National Institute of Health guidelines, and were approved by the University of North Carolina Institutional Animal Care and Use Committee.
Ethanol (20% v/v) solutions were prepared using tap water and 95% ethyl alcohol, and sucrose (10%, w/v) solutions were prepared using tap water and D-sucrose. The selective CRF1R non-peptidic antagonists NBI27914 (10-60 mg/kg; Tocris Biosciences, Ellisville, MO), antalarmin (30 mg/kg; Sigma-Aldrich, St. Louis, MO), and (4-ethyl-[2,5,6-trimethyl-7-(2,4,6-trimethylphenyl)-7H-pyrrolo[2,3-d]pyrimidin-4-yl]amino-1-butanol (LWH-63; 10-30 mg/kg; Sigma-Aldrich, St. Louis, MO) were dissolved in a 0.9% saline and emulphor (also known as Cremphor EL; Sigma-Aldrich, St. Louis, MO) solution (10% emulphor v/v) and delivered intraperitoneally (i.p.) in a 10 ml/kg injection volume approximately 30 min prior to ethanol access. Doses were extrapolated from previous studies (Robison et al., 2004; Sabino et al., 2006; Ji et al., 2007; Kim and Han, 2009). For the intra-CeA and intra-BLA microinjections, antalarmin (1.0 μg/ 0.5 μl per side) was dissolved in 0.9% saline and emulphor solution (5% emulphor, v/v) and delivered approximately 1 h prior to ethanol access. Previous studies were used to select the dose (Robison et al., 2004; Wang et al., 2006) and injection volume (Finn et al., 2007). CRF (Tocris Biosciences, Ellisville, MO) was dissolved in dH2O to a concentration of 0.1 mM, diluted to 200 nM in ACSF and bath applied.
All experiments used a 4-day drinking-in-the-dark (DID) procedure that has been used previously to generate high levels of ethanol intake that are associated with BECs in excess of 80 mg/dl (Rhodes et al., 2005; Rhodes et al., 2007; Sparta et al., 2008; Lowery et al., 2010). On days 1-3, beginning 3 h into the dark cycle, water bottles were removed from all cages and replaced with bottles containing a 20% (v/v) ethanol solution. Mice had 2 h of access to ethanol, after which ethanol bottles were removed from cages and water bottles were replaced. The same procedure was followed on day 4 except that ethanol access was extended to 4 h. Immediately following 4-h ethanol access on day 4, approximately 10 μl of blood were collected from the tail vein of each animal. Blood samples were centrifuged, and 5 μl of plasma were analyzed for BECs using the Analox Alcohol Analyzer (Analox Instruments USA, Lunenburg, MA). For all experiments involving pretreatment with drug on day 4, drug treatment groups were equated based on ethanol consumption on days 1-3. For all experiments involving multiple cycles of DID procedures, each cycle was separated by 3 days. To assess the specificity of experimental manipulations to binge-like ethanol consumption, sucrose control studies and non-binge-like ethanol consumption control studies were conducted. Briefly, sucrose control studies were conducted following the DID procedure similar to ethanol DID procedures except that 10% sucrose (w/v) solution was presented in place of 20% ethanol solution. Non-binge-like ethanol control studies were conducted using similar procedures to DID, except that access to 20% ethanol began 3 h into the light portion of the light/dark cycle.
Immunohistochemistry procedures were based on those routinely used in our laboratory (Thiele et al., 1996; Thiele et al., 1997; Thiele et al., 2000; Hayes et al., 2005). Mice were perfused transcardially with 0.1 mM phosphate-buffered saline (PBS; pH7.4) followed by 4% paraformaldehyde in phosphate buffer immediately following ethanol access on day 4 of the final DID cycle or 18-24 h following ethanol access on day 4 of the final DID cycle. The brains were collected and post-fixed in paraformaldehyde for 48 hours at 4°C, and then transferred to PBS. The brains were cut using a vibratome into 40 μm sections and stored in cryopreserve until the IHC assay. Sections were then transferred to PBS for 24 h before processing with CRF antibody. After rinsing in fresh PBS 4 times (10 minutes each), tissue sections were blocked in 10% goat serum and 0.1% triton-X-100 in PBS for 1 hour. Sections were then transferred to fresh PBS containing primary antibody for 72 h at 5°C. CRF expression was detected using primary rabbit anti-CRF (Peninsula Laboratories, LLC, San Carlos, CA; T-4037; 1:5000, 1:10,000, or 1:20,000 as determined by pilot experiments). As a control to determine if staining required the presence of the primary antibodies, some sections were run through the assay without primary antibody. After the 72 h of incubation, the sections were rinsed 4 times and then processed with Vectastain Elite kits (Vector Laboratories, Burlingame, CA) as per the manufacturer's instructions for standard ABC/HRP/diaminobenzidine-based immunohistochemistry. The sections processed for CRF were visualized by reacting the sections with a 3,3′-diamino-benzidine tetrahydrochloride (DAB; Polysciences, Inc., Bay Shore, NY) reaction solution containing 0.05% DAB, 0.005% cobalt, 0.007% nickel ammonium sulfate, and 0.006% hydrogen peroxide. All sections were mounted on glass slides, air-dried overnight, and cover slipped for viewing. Digital images of each candidate brain region were obtained on a Nikon E400 microscope equipped with a Nikon Digital Sight DS-U1 digital camera run with Nikon-provided software (Nikon Instruments Inc, Melville, NY). CRF-IR was quantified in each region at the level that showed the most robust staining and great care was taken to match sections within each brain region at the same level using anatomic landmarks with the aid of a mouse stereotaxic atlas (Paxinos and Franklin, 2004). Densitometric procedures were used to assess levels of CRF-IR in each brain region. Flat-field corrected digital pictures (8-bit grayscale) were taken using the Digital Sight DS-U1 camera. The density of staining was analyzed using Image J software (Image J, National Institutes of Health) by calculating the percent of the total area examined that showed signal (cell bodies and processes) relative to a subthreshold background. To account for differences in background between pictures within the same region, only the darkest area(s) of staining in each image was defined as CRF-IR. Specifically, the darkest area of staining was defined as the first complete cell body to be highlighted by Image J's threshold function. This cell body and all other areas of staining (including both cell bodies and processes) that were highlighted at this threshold were identified as CRF-IR. Anatomically matched pictures of the left and right sides of the brain were used to produce an average density for each brain region from each slice. In all cases, quantification of immunohistochemistry data was conducted by an experimenter that was blinded to group identity.
Mice were sacrificed 18-24 h following ethanol access on day 4 of the final DID cycle and their brains were placed in ice-cold sucrose artificial cerebro-spinal fluid (ACSF) containing: (in mM) 194 sucrose, 20 NaCl, 4.4 KCl, 2 CaCl2, 1 Mg Cl2, 1.2 NaH2PO4, 10.0 glucose, and 26.0 NaHCO3 saturated with 95% O2/ 5%CO2. The brains were cut using a vibratome into 300 μm sections and slices containing the CeA were identified using well characterized landmarks such as the BLA. Slices were then stored in a heated (28 °C), oxygenated (95% O2/ 5%CO2) holding chamber containing ACSF [(in mM) 124 NaCl, 4.4 KCl, 2 CaCl2, 1.2 MgSO4, 1 NaH2PO4, 10.0 glucose, and 26.0 NaHCO3 ] or transferred to a submerged recording chamber (Warner Instruments, Hamden, CT) where they were perfused with heated, oxygenated ACSF at a rate of 2 ml/min. Slices were allowed to equilibriate in ACSF for 1 h prior to the start of experiments. Whole-cell voltage clamp recordings were conducted from slices in the submerged recording chamber. Neurons of the CeA were visualized using infrared video microscopy (Olympus America Inc., Central Valley, PA). Recording electrodes (3-6 MΩ) were pulled on a micropipette puller (Sutter Instruments, Novato, CA) using thin-walled borosilicate glass capillaries. To analyze evoked inhibitory postsynaptic currents (eIPSCs), electrodes were filled with (in mM) K+-gluconate (70), KCl (80), HEPES (10), EGTA (1), ATP (4), GTP (0.4) pH 7.2, 290-295 mOsmol. Twisted nichrome wire stimulating electrodes were placed in the CeA, 100-500 μm from the recorded neuron. After entering a whole-cell configuration, cells were held at -70mV and GABA type A receptor (GABAAR)-mediated IPSCs were evoked at 0.2 Hz by local fiber stimulation with bipolar electrodes. GABAA-IPSCs were pharmacologically isolated by adding 10 μM NBQX. Signals were acquired via a Multiclamp 700B amplifier (Molecular Devices, Sunnyvale, CA), digitized and analyzed via pClamp 10.2 software (Molecular Devices, Sunnyvale, CA). Input resistance and access resistance were continuously monitored during experiments. Experiments in which changes in access resistance were greater than 20% were not included in the data analysis. eIPSC experiments were analyzed by measuring the peak amplitude of the synaptic response, which was normalized to the baseline period. The baseline period was defined as the 2 min period immediately preceding the application of the drug. CRF (200 nM) was bath applied at minutes 6-20.
Mice were surgically implanted with bilateral 26-gauge guide cannulae (Plastics One, Inc., Roanoke, VA) aimed at the CeA (0.94 mm posterior to bregma; 2.55 mm lateral to the midline; 2.60 mm ventral to the skull surface) or the BLA (1.22 mm posterior to bregma; 3.01 mm lateral to the midline; 4.74 ventral to the skull surface) using the Angle II stereotax (Leica Instruments Inc., Houston, TX) with the aid of a mouse stereotaxic atlas (Paxinos and Franklin, 2004). Following surgery, DID procedures were initiated. On day 4, animals were injected bilaterally with antalarmin (1 μg/ 0.5 μl per side) or vehicle (5% emulphor in saline; 0.5 μl per side) over 60 sec approximately 1 h prior to ethanol access using a 33-gauge injector designed to extend 2 mm beyond the cannula tip. A sucrose control study was conducted to evaluate the effects of intra-CeA antalarmin or vehicle on consumption of a 10% sucrose solution over a 4-h period beginning 3 h into the dark portion of the light/dark cycle. For this control study, mice were assigned to drug treatment groups based on their body weights. Cannulae placements were determined for each animal using injections of blue dye and thionin staining of mounted sections of brain tissue. Only the data from mice in which both cannulae were correctly aimed at the CeA or the BLA were included. All data from mice in which both cannulae (n= 1 for the CeA and n= 3 for the BLA) or one cannula (n= 8 for the CeA and n= 4 for the BLA) did not target the CeA or the BLA were excluded.
One-way ANOVAs were used to assess between-group differences in binge-like ethanol consumption and BECs, CRF-IR, and basal GABAergic signaling. Repeated measures ANOVAs were used to assess eIPSC peak amplitudes prior to CRF application (min 1-5) and while CRF was bath applied (min 6-20). A Student's t-test was used to assess the main effect of DID treatment on CRF's modulation of GABAergic signaling. As the aim of the immunohistochemistry experiments were to assess the specific effects of binge-like ethanol consumption on CRF-IR, data from animals that did not achieve BECs of 80 mg/dl or greater are not presented (n= 4). Because unquantifiable tissue varied by animal and region in these experiments, the degrees of freedom may differ for each statistical analysis. Tukey's post-hoc tests were used to assess specific between-group differences where appropriate. Significance was accepted at the p < 0.05 (two-tailed). All data are presented as mean ± SEM.
To test the hypothesis that high levels of ethanol intake require CRF signaling, the effects of several small molecule, non-peptide CRF1R antagonists on binge-like ethanol consumption were assessed. Average ethanol consumption and BEC data from day 4 are presented in Fig. 1. An ANOVA revealed a significant main effect of antalarmin on binge-like ethanol consumption (F(1, 25)= 6.548, p = 0.017), confirming that the 30 mg/kg dose of antalarmin significantly blunted binge-like ethanol drinking (Fig. 1 A). An ANOVA performed on associated BEC data was also significant (F(1, 25)= 4.296, p = 0.049), as antalarmin-induced reduction of ethanol drinking was associated with a significant reduction of BECs (Fig. 1 E). An ANOVA revealed a significant main effect of NBI27914 on binge-like ethanol consumption (F(3, 37)= 18.512, p < 0.001), and post-hoc analyses revealed that pretreatment with the 30 or 60, but not the 10, mg/kg doses of NBI27914 significantly attenuated binge-like ethanol consumption relative to vehicle (Fig. 1 B). Similar significant effects were observed on BEC data (F(3, 37)= 8.048, p < 0.001), and post-hoc analyses indicated that pretreatment with 60 mg/kg resulted in significantly reduced BECs relative to vehicle (Fig. 1 F). An ANOVA revealed a significant main effect of LWH-63 on binge-like ethanol consumption (F(3, 39)= 13.591, p < 0.001), and post-hoc analyses showed that pretreatment with both 30 and 60 mg/kg doses of LWH-63 significantly attenuated binge-like ethanol consumption relative to vehicle treatment (Fig. 1 C). Similarly, an ANOVA performed on associated BECs was statistically significant (F(3, 39)= 9.402, p < 0.001), and post-hoc analyses indicated that pretreatment with the 60 mg/kg dose significantly reduced BECs (see Fig. 1 G). Overall results showed that 30 mg/kg of antalarmin, 30 mg/kg of LWH-63, and 30 mg/kg of NBI27914 selectively altered binge-like ethanol consumption. A 30 mg/kg dose of antalarmin did not significantly alter sucrose intake (220.76 ± 20.49 ml/kg for vehicle-treated mice vs. 214.43 ± 39.67 ml/kg for antalarmin-treated mice). Further, pretreatment with 30 mg/kg doses of LWH-63 or NBI27914 did not significantly alter sucrose drinking (234.05 ± 29.30 ml/kg for vehicle-treated mice vs. 267.02 ± 17.07 ml/kg for LWH-63-treated mice and 173.48 ± 26.02 ml/kg for NBI27914-treated mice). These results indicate that CRF1R antagonists selectively modulate binge-like ethanol consumption.
To test the hypothesis that low or moderate levels of ethanol intake do not require CRF signaling, the effects of doses of each of the CRF1R antagonists that altered binge-like ethanol consumption were assessed on non-binge-like ethanol consumption. An ANOVA assessing the effects of CRF1R antagonists on non-binge-like ethanol consumption data was not statistically significant (F(3, 34)= 1.068, p = 0.377; see Fig. 1 D). An ANOVA performed on associated BECs was statistically significant (F(3, 34)= 3.190, p = 0.037), though post-hoc analyses did not indicate any significant differences in BECs between the vehicle group and any of the groups pretreated with a CRF1R antagonist (see Fig. 1 H). These results are in agreement with previous investigations showing that CRF1R antagonists have no effect on ethanol consumption that is not preceded by procedures to induce dependence-like drinking (Sabino et al., 2006; Chu et al., 2007; Gehlert et al., 2007) or to induce stress (Marinelli et al., 2007; Lowery et al., 2008), as well as a previous report comparing the effects of the CRF1R antagonist CP-154,526 on binge-like ethanol consumption and non-binge-like ethanol consumption (Sparta et al., 2008; Lowery et al., 2010). Because the Sparta et al. (2008) study also found that a CRF1R antagonist did not alter non-binge-like ethanol consumption during the dark portion of the light/dark cycle, the lack of effect observed in the current study is unlikely to be due to diurnal effects on CRF (Owens et al., 1990).
To verify the specificity of our antibody, brain tissue from a CRF wild-type (WT) mouse (see Fig. 2 A; CeA shown) and a CRF knockout (KO) mouse (see Fig. 2 B; CeA shown) was processed with CRF antibody. Importantly, CRF-IR was not observed in the PVN, the BNST, the VTA or the CeA of a CRF KO mouse, but was evident in each of these regions of a CRF WT mouse. The effects of binge-like ethanol consumption on CRF-IR immediately following 1 (1-cycle) or 6 (6-cycle) binge-like drinking exposures were investigated to assess the effects of a single episode and multiple episodes of high levels of ethanol intake on central CRF signaling. All comparisons were made with an ethanol-naive group of mice that consumed water for the duration of the experiment (Water). Consumption of ethanol during the 4-h access period for the duration of the experiment is shown in Table 1. Immediately following the final day of the final DID cycle, the 1-cycle group had mean BECs of 123.28 ± 12.69 mg/dl and the 6-cycle group had mean BECs of 161.71 ± 17.15 mg/dl. The results of a one-way ANOVA confirmed that ethanol consumption and BECs on the final day of the experiment did not differ between DID treatment groups, (F(1,15)= 0.430, p = 0.523 for ethanol consumption; F(1, 14)= 2.657, p = 0.127 for BECs). Fig. 2 shows CRF-IR in the CeA and ventral tegmental area (VTA) by DID treatment groups. Binge-like ethanol consumption significantly altered CRF-IR in the CeA (F(2,21)= 6.398, p = 0.008), and post-hoc analyses revealed that the 1-cycle and 6-cycle groups had significantly greater CRF-IR than the water group in this region (see Fig. 2 C-E). A significant effect of binge-like ethanol consumption on CRF-IR in the VTA was also observed (see Fig. 2; F(2,18)= 8.931, p = 0.002). Post-hoc analyses revealed that the 1-cycle group displayed significantly greater CRF-IR than the water group and the 6-cycle group (see Fig. 2 F-H). CRF-IR was not significantly altered following sucrose consumption in the CeA (0.10 ± 0.01 % area for the water group vs. 0.15 ± 0.01 % area for the 1-cycle group vs. 0.15 ± 0.02 % area for the 6-cycle group) or in the VTA (0.05 ± 0.01 % area for the water group vs. 0.06 ± 0.01 % area for the 1-cycle group vs. 0.04 ± 0.01 % area for the 6-cycle group). CRF-IR was not significantly altered in any other brain region assessed following binge-like ethanol consumption (please see Table 2) or following sucrose consumption (data not shown). These results indicate that binge-like ethanol consumption significantly alters CRF-IR of the CeA and the VTA. Because previous studies have primarily assessed the effects of ethanol on CRF expression after ethanol removal (Lowery and Thiele, 2009), we assessed CRF-IR in mice that had undergone 3 exposures to DID procedures, 18-24 h after their final ethanol exposure (3-cycle). Ethanol consumption over the course of this experiment is shown in Table 1. Animals of the 3-cycle group achieved mean BECs of 115.8 ± 20.72 mg/dl during the final 4-h ethanol access. A one-way ANOVA revealed that the 3-cycle group had significantly greater CRF-IR in the CeA than the water group (F(1,19)=7.586, p = 0.013; see Fig. 2 I-K). The resolution of the stain did not allow us to determine whether increases of CRF-IR were due to increases of CRF IR within cells or an increase in the number of cells exhibiting CRF IR (or both possibilities). Together, these findings extend those of previous reports showing upregulations of CRF markers in the CeA immediately following chronic ethanol ingestion (Lack et al., 2005) and during acute ethanol withdrawal from chronic intermittent ethanol vapor exposure (Sommer et al., 2008) by demonstrating that similar alterations in CRF IR are induced by voluntary ethanol consumption that is associated with intoxicating BECs. Additionally, a lack of effect of binge-like ethanol consumption on CRF expression in the PVN corroborate the results of a previous investigation from our group which suggested that binge-like ethanol consumption was modulated by CRF activity outside of the hypothalamic-pituitary-adrenal axis (Lowery et al., 2008).
Given that alterations of CRF-IR in the CeA were observed 18-24 h following ethanol removal in mice that had undergone 3 binge cycles, we hypothesized that multiple episodes of high ethanol intake would result in functional alterations in the CRF system. To maximize the effects of binge history on CRF function and minimize the likelihood of inducing dependence, we assessed the effects of 3 cycles of binge-like ethanol consumption on the neuromodulatory effects of CRF on GABAergic signaling in the CeA. As we were interested in the effects of a binge history on CRF function irrespective of the effects of acute ethanol, this experiment was conducted 18-24 h following the final binge episode to ensure that ethanol was not on board at the time of testing. Binge-like ethanol consumption on day 4 of each DID cycle is shown in Table 1. Mice of the 3-cycle group achieved mean BECs of 133.48 ± 10.5 mg/dl during the final 4-h ethanol access. Representative traces showing the effects of CRF on GABAergic transmission are shown in Fig. 3 A. The results of a repeated measures ANOVA revealed that a history of binge-like ethanol exposure lead to a significant reduction in the ability of CRF to enhance GABAergic transmission in the CeA (F(1, 9)= 9.163, p = 0.014; see Fig. 3 B). The results of a t-test revealed a significant main effect of ethanol history on the average modulation of eIPSC peak amplitude by CRF (t(9) = 2.445, p = 0.0370), as the average peak amplitude of IPSCs evoked in the presence of CRF was enhanced by 29 ± 11% in water drinking control mice but was not increased (-7 ± 9%) in mice with a history of binge-like ethanol consumption (see Fig. 3 C). In addition, to determine if there were any differences in the probability of basal GABA release, the paired-pulse ratio of the evoked IPSC in both water and alcohol exposed mice were assessed, and no significant differences were found (t(9) = 0.3392, p = 0.74, Fig. 3 D). Finally, we evaluated the ability of a CRFR1 antagonist, NBI27914 (1μM), to alter GABAergic transmission in both control and alcohol exposed mice. This experiment was done to determine if there was an endogenous CRF tone in the slice related to the ethanol exposure. We found that bath application of NBI27914 had no effect on the peak amplitude of eIPSC in either controls (93±8 % of baseline, n = 4) or alcohol exposed mice (108±4 % of baseline, n = 3), suggesting that there is no alteration in endogenous CRF tone in the CeA following binge-like ethanol drinking. Together, these results suggest that, despite concomitant upregulations of CRF IR in this region at the same timepoint, the modulatory effects of CRF on GABAergic transmission in the CeA are blunted in mice with a history of binge-like ethanol consumption.
Our IR results suggest that CRF signaling may be engaged in the CeA and motivate binge-like drinking. To test this hypothesis, the effects of microinjections of antalarmin into the CeA on binge-like ethanol consumption were assessed. Average ethanol consumption and BEC data from day 4 are presented in Fig. 4. The results of a one-way ANOVA revealed a significant effect of drug treatment (F(1, 14)= 8.921, p = 0.011), confirming that intra-CeA administration of antalarmin significantly blunted binge-like ethanol consumption (Fig. 4 A). Significant between-group differences in BECs were also observed (F(1, 14)= 4.847, p = 0.046), as mice pretreated with antalarmin achieved significantly lower BECs than mice that received vehicle (Fig. 4 B). Conversely, pretreatment with antalarmin into the CeA did not have effects on 4-h sucrose consumption (F(1, 16)= 1.458, p = 0.246; Fig. 4 C). Cannula placements are shown in Fig. 4 D. These results demonstrate that binge-like ethanol consumption, like dependence-like ethanol consumption (Funk et al., 2006), requires CRF signaling in the CeA and are the first to demonstrate a direct role for CRF1Rs in this region in ethanol consumption (Lowery and Thiele, 2009).
To determine whether the effects of microinjections of antalarmin into the CeA on binge-like ethanol consumption could have resulted from diffusion of the drug to other regions of the amygdala, we also assessed the effects of microinjections of antalarmin into the BLA, a region proximal to the CeA and dense in CRF1R expression (Hauger et al., 2006), on binge-like ethanol consumption. Average ethanol consumption and BEC data from day 4 are presented in Fig. 4. The results of a one-way ANOVA indicated no significant main effect of drug treatment on binge-like ethanol consumption (F(1, 16)= 1.124, p = 0.306 ; see Fig. 4 E) or BECs (F(1, 16)= 0.082, p = 0.778; see Fig. 4 F). Cannula placements are shown in Fig. 4 G. These results suggest that the attenuation of binge-like ethanol consumption by intra-CeA administration of antalarmin was not likely to be due to diffusion of the drug into the BLA.
The results of the current study provide novel evidence that endogenous CRF signaling is recruited in the CeA during bouts of binge-like ethanol consumption. First, three selective CRF1R antagonists reduced binge-like ethanol consumption at doses that did not alter non-binge-like ethanol intake. Second, a history of binge-like ethanol drinking was associated with a significant increase of CRF-IR in the CeA and in the VTA. Additionally, a history of binge-like ethanol drinking abolished the ability of exogenous CRF to increase GABAergic transmission in the CeA at a time point when endogenous CRF-IR was upregulated in the same region. Finally, antagonism of CRF1R in the CeA, but not in the BLA, attenuated binge-like ethanol consumption. These converging results show that binge-like ethanol consumption recruits CRF1R signaling in the CeA, which is necessary for maintaining excessive binge-like ethanol intake.
Previous investigations comparing the effects of CRF1R antagonists on binge-like ethanol consumption to their effects on non-binge-like ethanol consumption (Sparta et al., 2008; Lowery et al., 2010), have shown that CRF1R antagonists reduce high levels of ethanol consumption without effects on moderate or low ethanol consumption (see Lowery and Thiele, 2009 for review). Similarly, CRF1R antagonists attenuate elevated ethanol consumption stemming from procedures to induce dependence-like drinking (see Lowery and Thiele, 2009 for review), stress (Marinelli et al., 2007; Lowery et al., 2008; Sommer et al., 2008), and deprivation-induced drinking (Sparta et al., 2009). Additionally, CRF1R antagonists attenuate ethanol consumption by the selectively bred Marchigian Sardinian alcohol-preferring (msP) rats (Ciccocioppo et al., 2006; Gehlert et al., 2007). Thus, while these compounds are effective under conditions that elicit high levels of ethanol consumption, CRF1R antagonists do not alter moderate levels of ethanol consumption by non-stressed, non-selected animals in which dependence-like drinking has not been established (see Lowery and Thiele, 2009 for review). Together, these data may suggest that high levels of ethanol consumption that promote pharmacologically relevant BECs (i.e., ≥ 80 mg/dl) recruit CRF signaling, while moderate or low levels of ethanol consumption (associated with low BECs) do not. Interestingly, the role of CRF in cocaine self-administration also appears to be intake-dependent, as escalated cocaine self-administration by animals with a history of high cocaine intake was attenuated by CRF1R antagonists at doses that did not alter cocaine self-administration by animals with a history of low cocaine intake (Specio et al., 2008). Thus the hypothesis that the recruitment of CRF is dependent on the amount of ethanol consumed during a given period of access may also extend to other drugs of abuse.
Ethanol-induced alterations in CRF signaling in regions such as the CeA have previously been noted. Specifically, previous investigations have demonstrated significantly augmented CRF mRNA in the CeA of animals immediately following chronic ethanol ingestion (Lack et al., 2005) and in ethanol-dependent animals during acute withdrawal (Sommer et al., 2008). Additionally, increases in extracellular CRF were suggested in the CeA of ethanol-dependent animals during acute withdrawal, as decreases in total CRF-IR but no change in the total number of CRF immunoreactive cell bodies were observed relative to non-dependent animals (Funk et al., 2006). Our data extend the current literature by demonstrating that CRF-IR in the CeA is altered by high levels of ethanol consumption in animals with little (i.e., 1 binge-like drinking cycle) or moderate (i.e., 6 binge-like drinking cycles) ethanol exposure. This is an important observation because it suggests that recruitment of CRF signaling in the CeA during early binge-like drinking episodes, prior to dependence, may trigger initial neuroplastic changes in this system. Theoretically, these changes may become more robust and rigid with repeated binge-like drinking experiences, contributing to the transition to dependence. Though increases of CRF-IR may reflect increased CRF activity (via increased CRF synthesis and/or release) or decreased CRF activity (via decreased CRF release), our converging pharmacological and electrophysiological results strongly suggest that CRF activity is elevated during binge-like ethanol consumption.
We also observed changes in CRF-IR in the VTA following a single binge-like drinking cycle. Previous studies have implicated CRF signaling in the VTA in the modulation of stress-induced reinstatement to cocaine seeking (Wang et al., 2005; Wang et al., 2007), which may be due to CRF's neuromodulatory effects on dopaminergic and glutamatergic transmission within this region (Wise and Morales, 2010). Thus, administration of CRF or a CRF1R agonist directly in to the VTA significantly reinstated cocaine-seeking by animals that had undergone long access to cocaine self-administration (Blacktop et al., 2011). Notably, intra-VTA administration of CRF did not alter stress-induced reinstatement to cocaine-seeking by animals that had undergone short access to cocaine self-administration, suggesting that CRF of the VTA modulates drug-taking behavior only in animals with a history of high cocaine intake. As ethanol has known actions on dopaminergic and glutamatergic systems of the VTA (Stuber et al., 2008), in addition to the CeA, binge-like ethanol drinking may also be modulated by CRF in the VTA.
In the CeA, CRF exerts modulatory effects on GABAergic signaling via CRF1Rs in a way that mirrors ethanol's effects on GABAergic transmission (Nie et al., 2009; Roberto et al., 2010). The current data and previous work (Roberto et al., 2010) show that CRF enhances GABAergic transmission by neurons of the CeA in ethanol-naïve animals. Previous investigations have shown that relative to ethanol-naïve controls, this effect is augmented by chronic intermittent exposure to alcohol vapor (Roberto et al., 2010). However, we found that a history of binge-like ethanol consumption abolished the ability of CRF to enhance GABAergic transmission. Relative to water-drinking controls, we observed a blunted response to bath-applied CRF in brain slices of mice that experienced three binge-like drinking cycles at a time point when they also had increased CRF-IR in the CeA. This may reflect homeostatic regulation of the CRF-GABA system in response to high synaptic CRF concentrations (e.g., internalization of CRF receptors). Alternatively, upregulated tonic activity of CRF may have occluded the effect of subsequently applied CRF. However, this is unlikely given that there was no difference in basal paired-pulse ratio between water-drinking mice and mice that experienced three cycles of binge-like ethanol drinking (which would be expected if CRF was already active) and the lack of an effect of the CRFR1 antagonist on evoked GABAergic transmission.
The divergent results of the current study and previous studies of animals exposed to ethanol vapor suggest that binge-like ethanol consumption engenders different patterns of GABAergic transmission than does ethanol dependence, evidenced by differences in responses to exogenous CRF in the CeA between animals that experienced procedures to induce dependence-like drinking and non-dependent animals with a history of binge-like ethanol intake, and by the observation that a history of binge-like drinking did not alter baseline GABAergic transmission as has been described in vapor exposed rats (Roberto et al., 2010). It is important to consider factors that may have contributed to the differences between dependence-like and binge-like drinking models noted above. Few exposures to high levels of ethanol (as in the current study) may cause transient perturbations of the CRF system that return to a homeostatic set point, while many exposures to high levels of ethanol (as in studies of ethanol dependent animals) may cause long lasting adaptations in the CRF system that reflect the establishment of an allostatic set point. Interestingly, the disruption of CRF signaling found in the present report is similar to what was seen in the closely related structure, the bed nucleus of the stria terminalis (BNST), following chronic cocaine exposure (Nobis et al., 2011), as repeated cocaine exposure led to a transient disruption in CRFR1 modulation of excitatory transmission with no changes in basal synaptic properties. Also consistent with our data is a study demonstrating that stress leads to an internalization of CRFR1 receptor in the dorsal raphe (Waselus et al., 2009). It is important to note that while binge-like ethanol consumption and dependence-induced ethanol consumption appear to fundamentally differ in some aspects, behaviors require CRF signaling in the CeA.
The results of the current study suggest that CRF signaling in the CeA is critical for the expression of binge-like drinking behavior. CRF signaling in the CeA is recruited during excessive ethanol intake, prior to the development of dependence, and we speculate that plastic changes in CRF signaling within the CeA develop over the course of repeated binge-like drinking episodes, contributing to alterations in critical downstream targets like the BNST (Li et al., 2012), and ultimately to the transition to ethanol dependence. Future investigations will examine potential differences in the CRF systems of binge-like drinking relative to non-binge-like drinking mice and will elucidate the effects of binge- induced modifications of CRF activity in the CeA on downstream targets. Our observations suggest that compounds aimed at CRF1R may have therapeutic value for treating excessive binge drinking, perhaps protecting vulnerable individuals from progressing to dependence.
The authors thank Dr. A. Ryabinin for generously supplying CRF knockout and wildtype mice brains. This work was supported by NIH grants AA013573, AA015148, AA017818, AA017803, AA019839, AA017668, AA019454, an ABMRF Research Grant, and an INIA-Stress Pilot Project.
Conflicts of Interest: All authors declare that they have no financial, or any other, conflicts of interests related to the work presented in this manuscript.