Twenty-four male Wistar rats (Charles River Laboratories, Barcelona, Spain) weighing 100–125 g on postnatal day (pnd) 24 were housed individually in a humidity and temperature-controlled vivarium on a 12 h light/dark cycle (lights off at 19:00 h). Rats were allowed to acclimatize to the new environment for 7 days before any experimental procedure was performed (pnd 31). Water and rat chow pellets were available ad libitum.

This study was carried out in accordance with the European Directive 2010/63/EU for the protection of animals used for scientific purposes and Spanish regulations (Real Decreto 53/2013 and 178/2004, Ley 32/2007 and 9/2003 and Decreto 320/2010) for the care and use of laboratory animals. The protocol was approved by the Ethic and research Committee at the Universidad de Malaga. All efforts were made to minimize animal suffering and social isolation, as well as to reduce the number of animals used.

As shown in Figure 1, adolescent rats (pnd 31) were randomly assigned to the experimental alcohol group following 4 weeks of intermittent alcohol exposure (n = 12) or control group (n = 12).

Rats were tail-bled 1 h after the last alcohol exposure. Blood was collected into a microtube containing anticoagulant (4 μl heparin; 1000 USP units/ml) and centrifuged at 2000 × g for 10 min. Serum was extracted and assayed for ethanol concentration (BEC) using the alcohol oxidase method with an AM1 Alcohol Analyzer (Analox Instrument, London, UK).

Behavioral tests were conducted by trained observers who were unaware of the experimental conditions.

Two weeks after the last alcohol exposure (pnd 70), rats were anesthetized with sodium pentobarbital (50 mg/kg, i.p.) and brain samples were collected. The brains were quickly removed, immediately frozen on dry ice and stored at −80°C, until determination of the gene expression of proteins using quantitative real-time reverse transcription polymerase chain reaction (RT-qPCR).

The frozen brains were placed in acrylic rat brain matrices, and 2-mm thick slices were obtained using brain matrix razor blades. The medial prefrontal cortex (mPFC), striatum (accumbens, caudate, and putamen nucleus), entire amygdala and dorsal and ventral hippocampus were dissected out bilaterally and collected using a scalpel (mPFC and hippocampus) and a sample corer (striatum and amygdala). The localization of the brain regions was performed using a rat brain atlas and considering the bregma point as zero coordinate in the rostral-caudal coordinates: mPFC, +3.0 mm to +5.0 mm from bregma; striatum, +1.0 mm to −1.0 mm from bregma; amygdala, −1.0 mm to −3.0 mm from bregma; and hippocampus, −3.0 mm to −5.0 mm from bregma (Paxinos and Watson, 1998).

Total RNA was extracted from brain samples using Trizol Reagent (Gibco BRL Life Technologies, Baltimore, MD, USA) and the concentrations were quantified using a spectrophotometer to ensure ratios of absorbance at 260 to 280 nm of 1.8–2.0. The reverse transcription was performed using the Transcriptor Reverse Transcriptase kit and random hexamer primers (Transcriptor RT; Roche Diagnostic, Mannheim, Germany). The RT-qPCR was performed using an ABI PRISM® 7300 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) and the FAM dye label format for the TaqMan Gene Expression Assays (Applied Biosystems). The absolute values from each sample were normalized relative to the housekeeping β-actin gene (Actb). The relative quantification was calculated using the ΔΔCt method and normalized to the control group. Primers for the RT-qPCR were obtained based on the Applied Biosystems genome database of rat mRNA references (http://bioinfo.appliedbiosystems.com/genome-database/gene-expression.html) (Table 1).

All the data in the graphs are expressed as the means ± SEM. The statistical analysis was performed using GraphPad Prism version 5.04 (GraphPad Software, San Diego, CA, USA) and the normal distribution of data was evaluated by D'Agostino-Pearson omnibus test. Student's t-test and Welch's t-test for unequal variances were conducted to compare continuous variables between the two experimental groups (Control and Alcohol). Two-way analysis of variance (ANOVA) was conducted in the Novel Object Recognition Test using alcohol exposure (control and alcohol) and type of object (familiar and novel) as factors and using Bonferroni as post hoc test. Benjamini–Hochberg false discovery rate (FDR) approach was used for multiple comparisons (Benjamini et al., 2001) of the mRNA expression analysis. The statistics (t-statistic and F-statistic) and degrees of freedom were indicated in the description of results. The significance level was established at 0.05 and a p-value of less than 0.05 was considered statistically significant. Additionally, FDR adjusted significance levels were indicated for the multiple comparisons in each case.

As shown in Figure 2A, during the period of alcohol exposure, there was a significant decrease in body weight gain in the alcohol group compared with the control group [t₍₂₂₎ = 4.00, p < 0.001]. Similarly, the averages of food intake during these 4 weeks were also significantly different [t₍₂₂₎ = 3.68, p = 0.001] (Figure 2B).

After the last alcohol exposure, the BEC in the alcohol group reached an average of 201 ± 6 mg/dl.

We next explored whether alcohol exposure during adolescence induced locomotor or emotional alterations.

One week after the last adolescent alcohol exposure, we measured spontaneous locomotor and exploratory activity using the open field test in rats. We observed no significant differences in the number of crossings between both groups (Figure 3A). Additionally, we measured anxiety-like behavior by time spent in the center area of the field (Figure 3B) and we detected a significant decrease in the time spent in the center area in the alcohol group compared with the control group [t₍₁₄₎ = 4.19, p < 0.001].

The rats were also tested on the elevated plus-maze under novelty conditions 2 weeks after alcohol exposure. Regarding the locomotor activity in the elevated plus-maze, we found no differences in the number of entries in the closed arms between groups (Figure 3C). In this maze, we measured anxiety-like behavior by the time spent in the open arms and significant differences were found. As shown in Figure 3D, there was a significant decrease in the time of open arm exploration in the alcohol group compared with the control group [t₍₁₄₎ = 4.65, p < 0.001].

One week after adolescent alcohol exposure, the recognition memory was measured using the novel object recognition task. During the familiarization phase, none of the groups displayed any preference for the objects but the statistical analysis revealed significant differences on the total exploration time (Figure 4A). Thus, the alcohol group spent less time to explore both objects than the control group [t₍₂₁₎ = 4.73, p < 0.001].

One hour later, the test session was repeated using a novel object. A two-way ANOVA revealed a main effect of alcohol exposure on the exploration time [F_{(1, 42)} = 12.76, p < 0.001] but also an interaction between alcohol exposure and type of object [F_{(1, 42)} = 14.40, p < 0.001]. As shown in Figure 4B, the post hoc comparison indicated that the control group had a higher preference for the novel object relative to the familiar object (p < 0.05), whereas the alcohol group displayed a lower preference for the novel object (p < 0.05). Additionally, there was a significant decrease in the preference of the novel object in the alcohol group compared with the control group (p < 0.001).

Consequently, the alcohol group exhibited a negative discrimination ratio that indicated an impaired recognition memory [t₍₂₁₎ = 5.83, p < 0.001] (Figure 4C).

We examined the effect of adolescent alcohol exposure on the gene expression encoding relevant enzymes and receptors involved in endocannabinoid signaling [receptors: CB₁ (Cnr1), CB₂ (Cnr2) and peroxisome proliferator-activated receptor-α (PPAR-α) (Ppara); synthesis enzymes: NAPE-PLD (Napepld), DAGL-α (Dagla) and DAGL-β (Daglb); degradation enzymes: FAAH (Faah) and MAGL (Mgll)] following 2 weeks after the last exposure.

We also evaluated the effect of alcohol exposure during adolescence on the gene expression of neuropeptides linked to anxiety/stress responses and their receptors [neuropeptides: corticotropin-releasing factor (CRF) (Crh) and neuropeptides Y (NPY) (Npy); and receptors: CRF1R (Crhr1), CRF2R (Crhr2), NPY1R (Npy1r), NPY2R (Npy2r) and NPY5R (Npy5r)]. Determinations were conducted in the amygdala and hippocampus, two main regions belonging to the circuit that modulates emotional responses.

Finally, we examined the effect of alcohol exposure during adolescence on the gene expression of relevant inflammatory-related markers [pro-inflammatory mediators: tumor necrosis factor-α (TNF-α) (Tnf), TLR4 (Tlr4), COX-2 (Ptgs2) and NF-κβ (Rela); and glia-activating factors: glial fibrillary acidic protein (GFAP) (Gfap) and allograft inflammatory factor-1 or microglia response factor (MRF-1) (Aif1)] following 2 weeks after the last alcohol exposure.

As expected, the alcohol group showed elevated ethanol levels in the blood. These BECs were about 200 mg/dl and they are comparable to the concentrations reported in other studies conducted in adolescent Wistar rats using a model of binge-like drinking with i.p. injections of 3 g/kg, which reported 150–180 mg/dl 1 h postinjection (Przybycien-Szymanska et al., 2010, 2012). Although our procedure raised plasma alcohol levels in a binge-like manner [binge drinking is indeed defined as heavy episodic drinking within a BEC of 80 mg/dl or higher (NIAAA, 2004)], the alcohol administration was injected and it is possible that our model would yield higher alcohol doses than animals would consume voluntarily as it is expected in a binge drinking pattern (Spanagel, 2000). However, both patterns of excessive alcohol consumption are characterized by high BECs, and this can be critical in the effects of adolescent alcohol on behavior and expression of signaling systems in the brain.

During the alcohol exposure, we monitored the feeding behavior of adolescent rats by measuring body weight and food intake. Thus, the alcohol group displayed a lower body weight gain accompanied by a reduction in food intake during the alcohol exposure. These observations agree with previous studies reporting a significant decrease in body weight after i.p. administration of ethanol (Luz et al., 1996; Iwaniec and Turner, 2013). In fact, Lauing and colleagues reported decreased weight gain after modeling binge drinking using i.p. alcohol administration delivered on 2 consecutive days followed by 5 days of abstinence (Lauing et al., 2008). Several factors such as local inflammation and pain related to repeated injections, repeated acute ethanol intoxications with subsequent physical discomfort…could participate in this negative energy balance. However, there are studies reporting no differences in the final body weight after an intermittent binge-drinking exposure (Przybycien-Szymanska et al., 2010) although the total duration of this protocol was only 8 days.

After the alcohol exposure, locomotor activity, anxiety-like behavior and cognitive responses were assessed using a behavioral test battery consisting of the open field, elevated plus maze and novel object recognition task.

One week after the last alcohol session, we observed no differences in locomotor activity in the open field but there was an anxiogenic-like response to the test in the alcohol group compared with the control group, which could be related to the effects of alcohol deprivation on emotional behavior. These results are consistent with previous studies in rodents exposed to repeated binge-drinking during adolescence, which reported long-term anxiety disorders (Gilpin et al., 2012; Montesinos et al., 2016). One week later, the same rats were assessed in the elevated plus-maze and the anxiogenic-like behavior was also detected in the alcohol group. Ethanol withdrawal in rodents is associated with a negative affective state, including an enhanced anxiety-like behavior. Following a period of abstinence, rats spent significantly less time in the open arms of the elevated-plus maze than control animals (File et al., 1993; Moy et al., 1997; Wilson et al., 1998; Pandey et al., 2003). This behavioral profile can persist for as long as 28 days (Rasmussen et al., 2001). In our animals, the elevated anxiety in the alcohol group persisted unchanged for at least 2 weeks following the last alcohol exposure, suggesting enduring changes on emotionality.

A novel object recognition task was used to detect the existence of cognitive impairments because this paradigm relies on the innate preference that rodents display for exploring novel rather than familiar objects (Ennaceur and Delacour, 1988). To perform the task, we used a 1 h delay interval between the familiarization and novelty phases to consolidate a short-term memory that depends on the integrity of the hippocampal formation (García-Moreno et al., 2002). The results showed that rats exposed to alcohol displayed a decreased preference for the novel object over the familiar object but similar exploration time of the familiar object. This attracted our attention, because this less exploration time of the novel object was not associated with a locomotor impairment in this task, which was in agreement with the open field observations. In fact, the alcohol group displayed less exploration time during the familiarization phase when both identical objects were considered novel. Thus, the reduced preference for the novel object exhibited by the rats exposed to alcohol might be associated with the development of neophobia-induced anxiety and cognitive deficit, which is thought to prevent exploratory behavior (Myhrer, 1988).

The brain is undergoing extensive maturation during adolescence. Similarly, the ECS is also undergoing maturational changes and some alteration of these processes of maturation can produces long-term alterations, including deficits in emotional behavior and cognition.

Several lines of evidence indicate that alcohol leads to neuroadaptations in endocannabinoid signaling mechanisms (Basavarajappa and Hungund, 1999; Basavarajappa et al., 2003). The present results showed that rats exposed to adolescent alcohol displayed alterations in the expression of endocannabinoid enzymes and receptors that appear to be reliant on the brain region, which is consistent with previous studies reporting regionally distinct effects on brain endocannabinoid levels after alcohol exposure (González et al., 2002, 2004; Rubio et al., 2007).

The mPFC is a key region for multiple cognitive functions, including executive function or anxiety. In addition, the ECS is involved in the regulation of emotional behaviors in this brain region (Rubino et al., 2008; McLaughlin et al., 2014; Morena et al., 2016). Thus, the negative impact of the alcohol to the adolescent mPFC may be associated to the observed alterations in cognition and behavior. We found that rats exposed to adolescent alcohol had a strong increase in the mRNA expression of Napepld and Dagl in this region. These observations suggest that the ECS is activated in response to the negative affective state associated to alcohol withdrawal. Additionally, the alcohol group displayed an increase in the mRNA expression of Faah and the receptor Ppara. It has been described that localized infusion of the FAAH inhibitor URB597 into the mPFC increases ethanol consumption by rats (Hansson et al., 2007). Also, a recent study has reported that pretreatment with an inhibitor of FAAH prevents against oxidative stress caused by binge ethanol consumption in the mPFC of adolescent rats (Pelição et al., 2016). Furthermore, PPAR-α is involved in the actions of other N-acylethanolamines with no endocannabinoid activity (e.g., oleoylethanolamide and palmitoylethanolamide) that have been reported to exhibit neuroprotective effects (Fu et al., 2003; Scuderi et al., 2012). Collectively these findings suggest a strengthening of the ECS that may reflect a homeostatic mechanism to prevent the neurotoxic effects induced by alcohol with a relevant role of other non-cannabinoid congeners in the alcohol exposure.

The ECS is an important modulator of neuroplasticity in the striatum. Overall, we observed that alcohol-exposed rats displayed a decreased mRNA expression of Cnr1 and Ppara, as well as a decreasing trend in Cnr2. These results agree with preclinical studies that indicate a decrease in Cnr1 mRNA levels after alcohol exposure in mice (Vinod et al., 2006) and rats (Adermark et al., 2011). Regarding the Cnr2 mRNA levels, it is known that alcohol consumption alters the Cnr2 expression in the brain and that an increased ethanol preference is associated with a reduced Cnr2 expression in the striatum (Onaivi et al., 2008).

The amygdala is a crucial subcortical area that integrates reward, emotions, and conditioned learning. Previous studies have demonstrated drug-induced alterations in amygdalar endocannabinoid function (Orio et al., 2009; Kamprath et al., 2011; Schmidt et al., 2011). In the present study, we observed a decreasing trend in the receptors and synthesis enzymes but not in the degradative enzymes. Consistently, we have reported that chronic alcohol treatment alters the expression of different cannabinoid signaling-related genes in the amygdala (Serrano et al., 2012).

The hippocampus is a brain region involved in learning which cooperates with the amygdala to modulate emotion and belongs to the reward circuit. Numerous studies have shown that alcohol exposure alters hippocampal function (for review see Kutlu and Gould, 2016). Regarding the expression of the endocannabinoid signaling in the hippocampus, we found a dramatic decrease in the expression of Cnr2. In the hippocampus, such cannabinoid receptor is involved in the modulation of excitatory synapses (García-Gutiérrez et al., 2013; Kim and Li, 2015). A recent study has demonstrated a role of the hippocampal CB₂ to modulate cognitive functions in mice (Li and Kim, 2016). Thus, the effects on Cnr2 mRNA expression in our alcohol-exposed rats may be associated to the adverse effects of alcohol on synaptic plasticity (basically long-term potentiation and long-term depression processes) and the consequent defects in learning and memory (Lovinger and Roberto, 2013).

As mentioned above, the presence of anxiety is one of the most consistent features of alcohol withdrawal. Since the amygdala and hippocampus are key brain regions in the modulation of emotional behaviors and the ECS is also involved in the regulation of emotionality (Hill and Gorzalka, 2009), the present findings suggest that the differential changes observed in the ECS after adolescent alcohol exposure may be associated with different responses to anxiety. In this regard, to get a better understanding of the potential mechanisms underlying the anxiety effects that were observed, we analyzed the CRF and NPY signals in these regions. Furthermore, both neuropeptides are clearly involved in alcohol-related behaviors and binge alcohol drinking (Pleil et al., 2015).

Interestingly, we found an increase in the hippocampal expression of Crh mRNA in the alcohol group with no changes in the CRF receptors. In line with this, previous studies have associated CRF with binge drinking and alcohol dependence (Lowery et al., 2010; Gilpin et al., 2012; Pleil et al., 2015).

Regarding the NPY signal, in both brain areas the alcohol group displayed a down-regulation of this system with a strong decrease in the mRNA levels of Npy2r that may be associated with the anxiogenic-like behavior observed in this group. These findings are in agreement with a recent study showing that a decrease in the expression of receptors for NPY are associated with an increase in the anxiety-like behavior in adolescent rats exposed to repeated binge-like alcohol drinking (McClintick et al., 2016).

Since alcohol increases neuroinflammation through its ability of activating natural immunity, we evaluated the gene expression of factors involved in pro-inflammatory signaling pathways but also factors associated with plasticity and neurodevelopment.

The data indicate a region-specific susceptibility to the alcohol regulation of the mRNA expression of these factors, being these alterations more prominent in the mPFC. In line with this, binge drinking has been reported to produce alcohol-induced inflammatory PFC damage and this can be accompanied by reduced executive functions and compulsive behavior (Crews et al., 2000). Thus, we observed that adult rats exposed to adolescent alcohol displayed an overall increase in the mRNA expression of the neuroinflammation-related factors Tlr4, Tnf and Ptgs2 in the mPFC. Several evidence indicates that alcohol induces the TLR4 signaling in the PFC (Vetreno and Crews, 2012; Pascual et al., 2014), triggering the induction of a cascade of pro-inflammatory mediators that affect, among others, cognitive impairments and anxiety-like behaviors associated with the alcohol abuse (Montesinos et al., 2015; Pascual et al., 2015).

TLR4 receptors are mediators of alcohol-induced inflammatory damage in the adolescent and adult brain (Alfonso-Loeches et al., 2010). In fact, Crews and colleagues have reported an increased Tlr4 expression in the brain of alcoholics as well as in mice treated chronically with ethanol (Crews et al., 2013). In addition to TLR4, there was an up-regulation of Tnf and Ptgs2 mRNA levels in alcohol-exposed rats, which is consistent with previous studies reporting that ethanol binge drinking is associated with increased cortical levels of TNF-α and COX-2 (Knapp and Crews, 1999; Antón et al., 2016). By contrast, the gene expression of various pro-inflammatory mediators were down-regulated (e.g., COX-2 mRNA levels were decreased in the striatum and hippocampus and NF-κβ mRNA levels were decreased in the mPFC and striatum), which appears to be incongruent with literature since alcohol exposure activates COX-2 and NF-κβ by triggering cytokine and chemokine release and neuroinflammation. In the mPFC, the decreased gene expression of Rela may be associated with the up-regulation of Ppara receptor observed in this brain region because it has been described that the activation of this nuclear receptor inhibits Rela (Stahel et al., 2008). Nevertheless, we have to keep in mind that changes in gene expression are often not concordant with protein expression (Vogel and Marcotte, 2012).

In addition to these factors and receptors, the inflammatory response is associated with astrocytes activation (Ransohoff and Brown, 2012). In fact, previous studies have demonstrated that Gfap levels are affected by alcohol exposure (Miguel-Hidalgo et al., 2006; Udomuksorn et al., 2011). Consistently, the gene expression of GFAP and MRF-1 were found up-regulated in the mPFC and hippocampus. Similar to COX-2 and NF-κβ, the gene expression of both astrocyte marker Gfap and microglial marker Aif1 were also down-regulated in other brain regions (hippocampus and amygdala, respectively) and this was only reported in human alcoholic brain in conjunction with astrocytic loss (Lewohl et al., 2005).

In summary, adult rats exposed to intermittent alcohol exposure during adolescence displayed anxiety-like behaviors and cognitive deficits related to recognition memory. Furthermore, the mRNA expression of some components of signaling systems involved in behavior, neuroinflammation and plasticity were found altered in these animals, with brain region-dependent changes.

Although adult rats exposed to adolescent alcohol had no alterations in locomotor activity, they exhibited a strong anxiogenic-like behavior. In addition to the anxiety-like response, we observed a clear cognitive deficit with a reduced preference for the novelty. These behavioral and cognitive effects in adulthood were accompanied by molecular changes in the brain.

Alterations in the mRNA expression of genes of the ECS appear to be reliant on the brain region. However, we observed an overall decreasing trend in the mRNA expression of the cannabinoid receptors in the alcohol group. This down-regulation can be associated to the adverse effects of alcohol on synaptic plasticity with cognitive consequences in adulthood such as deficits in learning and memory.

Whereas, the ECS is also associated with emotionality, neuropeptides such as CRF and NPY are clearly involved in alcohol-related behaviors and binge alcohol drinking. Thus, we observed changes in both signals that were linked to anxiogenic-like behaviors, in particular a decrease in the NPY signaling expression.

Regarding neuroinflammatory-related factors, our data indicated a brain region-specific susceptibility in the adult rats exposed to intermittent alcohol. These alterations were more prominent in the mPFC with an up-regulation of the mRNA levels of Tlr4, Tnf, Ptgs2, and Gfap and a decrease of Rela. The activation of PPAR-α inhibits NF-κβ and the mRNA expression of Ppara was increased in the mPFC, a key region for multiple cognitive functions. Interestingly, Ppara is involved in the anti-inflammatory and neuroprotective actions of endocannabinoid congeners. In contrast to the mPFC, the mRNA expression of Ptgs2 and Gfap were down-regulated in the rest of brain regions.

We are aware of the limitations of our study, mainly because of the exploratory nature of the present investigation and the necessity of performing future studies that include female subjects and alcohol exposure during adulthood. Furthermore, the protein expression and function of signaling systems and factors involved in behavior, neuroinflammation and neuroplasticity will open new lines of research using pharmacological and genetic approaches to characterize the association and regulation of these systems with emotional behavior and cognition in the context of adolescent alcohol and its long-term consequences.

The existence of sex differences in the response to alcohol and other drugs has been extensively reported (Becker and Koob, 2016) but we have conducted our experiments only in male rats for clarity. Studies using female rats will reveal whether the estrous cycle and gonadal hormones influence the behavioral and molecular alterations that we have observed here.

Although our data indicate changes in the mRNA expression of signaling systems in the brain and alterations in both emotional and cognitive responses of adult rats exposed to adolescent alcohol, we cannot link these molecular and behavioral effects exclusively to an alcohol exposure during adolescence because we have no rats exposed to alcohol during adulthood for comparison. Consequently, the specificity of the effects associated with intermittent alcohol exposure on these biological substrates will have to be elucidated.

Finally, additional studies will have to resolve the lack of the protein expression of these enzymes, factors and receptors because changes in gene expression level are frequently not reflected at the protein level. In fact, post-transcriptional, translational and degradation regulation must also be taken into account in the determination of protein concentrations because contribute at least as much as transcriptional itself (Vogel and Marcotte, 2012). Also, functional assays will be required to elucidate the role of these signaling systems in the appearance of emotional behaviors and addictive behaviors related to an adolescent alcohol exposure.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.