Environmental Enrichment Rescues Endocannabinoid-Dependent Synaptic Plasticity Lost in Young Adult Male Mice after Ethanol Exposure during Adolescence

Binge drinking (BD) is a serious health concern in adolescents as high ethanol (EtOH) consumption can have cognitive sequelae later in life. Remarkably, an enriched environment (EE) in adulthood significantly recovers memory in mice after adolescent BD, and the endocannabinoid, 2-arachydonoyl-glycerol (2-AG), rescues synaptic plasticity and memory impaired in adult rodents upon adolescent EtOH intake. However, the mechanisms by which EE improves memory are unknown. We investigated this in adolescent male C57BL/6J mice exposed to a drinking in the dark (DID) procedure four days per week for a duration of 4 weeks. After DID, the mice were nurtured under an EE for 2 weeks and were subjected to the Barnes Maze Test performed the last 5 days of withdrawal. The EE rescued memory and restored the EtOH-disrupted endocannabinoid (eCB)-dependent excitatory long-term depression at the dentate medial perforant path synapses (MPP-LTD). This recovery was dependent on both the cannabinoid CB1 receptor and group I metabotropic glutamate receptors (mGluRs) and required 2-AG. Also, the EE had a positive effect on mice exposed to water through the transient receptor potential vanilloid 1 (TRPV1) and anandamide (AEA)-dependent MPP long-term potentiation (MPP-LTP). Taken together, EE positively impacts different forms of excitatory synaptic plasticity in water- and EtOH-exposed brains.


Introduction
Abusive EtOH consumption typically initiates during adolescence [1], having severe toxic effects on the maturing brain that leads to long-lasting neuropsychological alterations [2]. Recent studies have demonstrated that adolescent BD alters neurotransmitter systems [3], reduces neurogenesis [4], and affects motor coordination and balance [5]. In addition, EtOH intake causes changes in glial cells [6,7], activates immune responses [8], and induces a significant cognitive impairment through the modulation of synaptic transmission and plasticity, notably in the hippocampus [9].
Three-week-old C57BL/6J male mice (Janvier Labs, Le Genest-Saint-Isle, France) were randomly distributed in pairs of control and EtOH experimental groups. Animals were maintained at 22 • C in a temperature-controlled room with a 12 h light and dark cycle (red light on at 9:00 a.m.) to habituate to the new environment 1 week before experimental procedures were initiated. All procedures were approved by the Committee of Ethics for Animal Welfare of the University of the Basque Country (CEEA/M20/2016/073; CEIAB/2016/074) and were in accordance with the European Communities Council directive of 22 September 2010 (2010/63/EU) and Spanish regulations (Real Decreto 53/2013, BOE 08-02-2013). Ad libitum access to food and water was available along all experiments except during DID procedure.

DID Procedure and Timeline
Four-week-old adolescent male mice [postnatal day (PND) 32] were classified in either control or EtOH experimental groups. Mice were housed in pairs in standard 17 × 14.3 × 36.3 cm plexiglas cages. They were exposed to 4-day DID [32] over 4 weeks (PND 32-56). On days 1-4 of each week, mice were weighed at 8:00 a.m. 1 h before lights off. At noon (3 h into the dark cycle), they were individually placed into a cage provided with a single bottle of 10 mL tap water or a single bottle of 10 mL EtOH [20% EtOH (v/v) prepared from 96% EtOH (Alcoholes Aroca S.L., Madrid, Spain)]. Mice had free access to water (control group) or EtOH (EtOH group) for 2 h the first 3 days, and for 4 h the fourth day. They were kept resting in abstinence the last 3 days of each week with access to food and tap water ad libitum ( Figure 1A). Thirty minutes after the last 4 h exposure, blood samples were collected from the lateral tail vein, and the blood ethanol concentration (BEC) (mg/dL) was measured in each mouse using a commercial EtOH assay kit (Sigma-Aldrich, Madrid, Spain) ( Figure 1B). The effectiveness of the DID procedure was demonstrated by measuring EtOH intake during each day (grams of EtOH per kilogram per 2 or 4 h (g/kg/2 or 4 h)) ( Figure 1C) and total EtOH intake (grams of EtOH per kilogram per hour (g/kg/h)) ( Figure 1D). The correlation between total EtOH intake (g/kg/h) throughout DID and BEC (mg/dL) measured at the end of EtOH access was also calculated ( Figure 1E).
Biomedicines 2021, 9, x FOR PEER REVIEW 3 of 14 (mg/dL) was measured in each mouse using a commercial EtOH assay kit (Sigma-Aldrich, Madrid, Spain) ( Figure 1B). The effectiveness of the DID procedure was demonstrated by measuring EtOH intake during each day (grams of EtOH per kilogram per 2 or 4 h (g/kg/2 or 4 h)) ( Figure 1C) and total EtOH intake (grams of EtOH per kilogram per hour (g/kg/h)) ( Figure 1D). The correlation between total EtOH intake (g/kg/h) throughout DID and BEC (mg/dL) measured at the end of EtOH access was also calculated ( Figure 1E).

Figure 1.
Schematic representation of experimental timeline, BEC, daily voluntary oral EtOH intake, total EtOH intake, and correlation between total EtOH intake and BEC. (A) Male C57BL/6J mice were exposed to 4 weeks DID during adolescence (PND . Each week, mice were exposed to 2 and 4 h of free EtOH access (20% (v/v)). Several mice were under environmental enrichment over the 2 weeks of withdrawal (PND 60-73) and were subjected to the Barnes maze test in the last 5 days (PND 69-73). Then, they were sacrificed to perform physiological recordings at early adulthood (PND 74-78).

Enriched Environment
Mice were in withdrawal for 2 weeks after DID (PND 60-73). Half litters of EtOHexposed and control mice were placed into EE conditions (groups EE-EtOH and EE-H20), while the others remained in standard conditions (groups EtOH and control) ( Figure 1A). The EE consisted of a large two-level cage (50 cm length × 28 cm height × 54 cm width) equipped with nesting material, a little house, climbing ladders, a running wheel, tunnels, and toys of different colors, sizes, and material (plastic, wood, and metal). The toys were changed and rearranged twice a week in order to increase exploratory capacity and maintain novelty. Seven mice per cage were housed to ensure social interactions with food and water provided ad libitum.

Barnes Maze Test
Learning and memory were tested using the Barnes maze [33]. The escape hole randomly selected for the mouse was maintained for each daily trial. The escape box under the selected hole was not visible; therefore, the mice had to use different cues placed in the room to discover it. The task consisted of 4 trials/day over 5 days. In each trial, the Schematic representation of experimental timeline, BEC, daily voluntary oral EtOH intake, total EtOH intake, and correlation between total EtOH intake and BEC. (A) Male C57BL/6J mice were exposed to 4 weeks DID during adolescence (PND 32-56). Each week, mice were exposed to 2 and 4 h of free EtOH access (20% (v/v)). Several mice were under environmental enrichment over the 2 weeks of withdrawal (PND 60-73) and were subjected to the Barnes maze test in the last 5 days (PND 69-73). Then, they were sacrificed to perform physiological recordings at early adulthood (PND 74-78). (B) BEC obtained on the last day of EtOH exposure averaged 61.97 ± 2.84 mg/dL ((n = 16) Student's t-test; *** p = 0.0001 vs. control). (C) EtOH intake during each day (g EtOH/kg/2 or 4 h) was measured and (D) total EtOH intake averaged 2.078 ± 0.07 g/kg/h (n = 16), similar to previous studies [5,17,32]. (E) A direct correlation between total EtOH intake and BEC was detected. (F) Stimulating electrode (S) was placed in the middle one-third of the DGML to stimulate MPP. R recording electrode, GC granule cells, MCF mossy cell fibers, MPP medial perforant path, LPP lateral perforant path.

Enriched Environment
Mice were in withdrawal for 2 weeks after DID (PND 60-73). Half litters of EtOHexposed and control mice were placed into EE conditions (groups EE-EtOH and EE-H 2 0), while the others remained in standard conditions (groups EtOH and control) ( Figure 1A). The EE consisted of a large two-level cage (50 cm length × 28 cm height × 54 cm width) equipped with nesting material, a little house, climbing ladders, a running wheel, tunnels, and toys of different colors, sizes, and material (plastic, wood, and metal). The toys were changed and rearranged twice a week in order to increase exploratory capacity and maintain novelty. Seven mice per cage were housed to ensure social interactions with food and water provided ad libitum.

Barnes Maze Test
Learning and memory were tested using the Barnes maze [33]. The escape hole randomly selected for the mouse was maintained for each daily trial. The escape box under the selected hole was not visible; therefore, the mice had to use different cues placed in the room to discover it. The task consisted of 4 trials/day over 5 days. In each trial, the mouse was allowed to explore for a maximum of 240 s or until the escape box was found. If not found, the experimenter gently guided the animal into the box. The mouse rested in the box for 60 s after the trial. They were able to use different strategies to find the escape box ( Figure 2C-E): random, serial, and spatial. The latency to escape, errors (non-target holes visited before escaping), and the strategy used to escape were quantified.
Biomedicines 2021, 9, x FOR PEER REVIEW 4 of 14 mouse was allowed to explore for a maximum of 240 s or until the escape box was found.
If not found, the experimenter gently guided the animal into the box. The mouse rested in the box for 60 s after the trial. They were able to use different strategies to find the escape box ( Figure 2C-E): random, serial, and spatial. The latency to escape, errors (non-target holes visited before escaping), and the strategy used to escape were quantified.

Statistical Analysis
Before analysis, Shapiro-Wilk and Kolmogorov-Smirnov tests were used for examining normal distribution of the data. Homogeneity of variances was determined by Levene's test. Two-way ANOVA analysis was performed to evaluate DID and EE effects as well as the interaction between both conditions. To further explore the effect of DID and EE in different experimental conditions (sham, EtOH, EE-EtOH, and EE-H 2 0), one-way ANOVA with post hoc analysis (Bonferroni correction for equal variances or Tamhane's T2 correction for unequal variances) was used. Data obtained from the Barnes maze test were analyzed using two-way ANOVA's multiple comparison. BEC differences between the control and EtOH groups and statistical significance data (EE-EtOH vs. EE-EtOH + drugs and EE-H 2 0 vs. EE-H 2 0 + drugs) were analyzed using Student's t-test. p ≤ 0.05 was considered statistically significant. All statistical tests were performed with SPSS statistical software (version 22.0, IBM, Barcelona, Spain) and data were given as mean ± standard error of the mean (SEM) with p-values and sample size (n).

EE Recovers EtOH-Induced Memory Impairment
Young adult mice exposed to EtOH during adolescence showed significant higher latency to escape (Figure 2A, two-way ANOVA's multiple comparison (n = 12); day 1 ** p = 0.004, day 2 * p = 0.021, day 3 * p = 0.033, and day 5 * p = 0.034) and more errors committed in the Barnes maze than controls ( Figure 2B, two-way ANOVA's multiple comparison; day 1 * p = 0.027). These results suggest a direct negative effect of adolescent BD on memory in young adults. EE in early adulthood was able to rescue the memory impairment induced by EtOH. Thus, a lower latency to escape (Figure 2A, two-way ANOVA's multiple comparison (n = 12); day 1 + p = 0.014, day 5 + p = 0.046) and fewer errors ( Figure 2B, two-way ANOVA's multiple comparison (n = 12) day 1 + p = 0.032, day 2 + p = 0.025, day 3 + p = 0.043 and day 4 + p = 0.028) were observed after EE. Furthermore, the two-way ANOVA revealed the significant effect of DID (F (1. EtOH also had a negative impact on the process in which the mice solved the maze. Thus, EtOH-exposed mice showed mostly random strategies ( Figure 2C, one-way ANOVA; *** p = 0.000 vs. sham), with a reduction in serial, and, particularly, spatial strategies ( Figure 2D,E). However, control mice solved the maze mainly using a serial strategy ( Figure 2D, one-way ANOVA (n = 12); ** p = 0.0094 vs. EtOH). The EE was able to significantly revert the changes in random ( Figure 2C, one-way ANOVA (n = 12); +++ p = 0.001) and spatial strategies induced by EtOH ( Figure 2E, one-way ANOVA (n = 12); +++ p = 0.000). Altogether, the EE applied in early adulthood recovered the persistent memory impairment observed after BD.

Discussion
Although EE has been demonstrated to rescue memory impairment [5,34], the role of EE in LTD is not well understood [21]. We have shown in this study that young adult male mice exposed to EE recover the 9group I mGluR-dependent MPP-LTD that is lost after adolescent EtOH consumption [17,18].

Mechanisms of MPP-LTD Rescue by EE in Adult Mice after Adolescent EtOH Intake
The EE rescue of MPP-LTD involves several components of the canonical endocannabinoid signaling pathway: CB1R, group I mGluRs and 2-AG [18,35]. Noticeably, the mice had MPP-LTD deficits although the amount of EtOH intake was modest compared with the BEC criteria (≥0.08 g/dL) set for a pattern of adolescent BD [36]. Moderate EtOH intake during the critical adolescent period of brain development is sufficient to promote long lasting ECS-dependent changes in hippocampal plasticity [17] and cognition [5]. Both the ECS and EtOH are under mutual influence, as the former modulates EtOH-motivated behavior, and the latter profoundly dysregulates the ECS [37]. Thereby, the long-lasting harmful changes in hippocampal mGluR5 mRNA, CB1 receptor expression, localization, and function at MPP synapses caused by adolescent EtOH intake associate with recognition memory deficits that can be recovered with rising 2-AG [17,18,32].
The exposure to EE protects against neuronal and cognitive damage, reduces EtOHinduced reinforcing, and recovers different memory forms and motor coordination in adult mice exposed to adolescent EtOH intake [5,17,38]. EE stimulates hippocampal neurogenesis [39] and increases nerve growth factor and brain-derived neurotrophic factor (BDNF) [40,41]. These processes elicit neuroprotective responses and induce neural plasticity and synaptic structural brain changes [42] leading to an improvement of spatial navigation [43], learning, and spatial memory [5]. Furthermore, many synaptic proteins and ionotropic glutamate receptors involved in plasticity and memory can be changed by EE [34,44]. Among them, mGluR5, which participates in the EE-mediated BDNF increase [45], rises during EE conditions [46]. In addition, EE increases glutamate release [47] and decreases the neuronal glutamate transporter EAAC1 [48]. These combined events favor a greater extracellular glutamate milieu, eventually leading to group I mGluRs activation and mGluR-dependent synaptic plasticity. mGluR1 antagonism impairs LTD [49] and an mGluR5 blockade prevents LTD that is necessary for spatial learning and memory [50]. However, EtOH elicits a significant reduction in CB1R expression and receptor labeling, particularly in the middle one-third of the DGML targeted by MPP synaptic terminals [15,17]. CB1 receptor signaling is also affected by EtOH intake as a consequence of reduced CB1 receptor binding and decreased Gαi2 subunit expression, which is known to be linked to learning and social behavior [14,17]. Although 2-AG and DAGL expression does not seem to change, a drastic increase in arachidonic acid and MAGL was noticed in a similar EtOH intake model used in this study [17]. Therefore, 2-AG levels are likely to increase in our model [51], eventually balanced by more 2-AG degradation through MAGL increase [52]. DAGL inhibitors abolished the EE-induced MPP-LTD recovery, indicating that 2-AG is involved. Conversely, the enhancement of long-term synaptic plasticity and cognitive performance by MAGL suppression implies CB1Rs [53]. Our previous observations showed that MAGL inhibition rescues MPP-LTD and recognition memory in adult mice after EtOH treatment during adolescence [17]. These observations suggest that the 2-AG increase surmounts the CB1R reduction at MPP synapses, potentially due to the high coupling efficiency of CB1 receptors in glutamatergic terminals [54]. Thus, under EE conditions, the plausible increase in glutamate levels favored by EE after LFS will activate the available mGluR5 (and potentially mGluR1), leading to the rise in 2-AG synthesis and presynaptic CB1R activation. The net effect of EE on endocannabinoid levels and the expression of ECS components needs to be determined.

EE Prompts MPP-LTP in Water-Exposed Mice
Although CB1R activation reduced fEPSPs after EE in the mice who drank water to the same extent as the controls, they also exhibited MPP-LTP, which is triggered by the same LFS that elicits MPP-LTD [18]. This LTP involves TRPV1 and AEA, but not CB1Rs, group I mGluRs, or 2-AG. The potentiation of the MPP synaptic strength aligns with the transient potentiation of the perforant path synapses after EE [55]. In this sense, spatial learning associates with hippocampal LTP and EE improves both LTP and spatial learning [45]. EE also modifies the expression of presynaptic and postsynaptic proteins, including AMPA and NMDA receptor subunits, as well as LTP-related molecules such as Ca2+/calmodulindependent Kinase II and cAMP response element binding protein [34]. However, the observed MPP-LTP following LFS after EE did not seem to depend on NMDA receptors, as D-APV had no effect. Previous studies have shown that voluntary exercise promotes MPP-LTP [56] and, similar to EE, augments BDNF [57] that favors neurotransmitter release and LTP [34]. Group I mGluRs are also involved in hippocampal LTP [21]. In our study, neither mGluR1 nor mGluR5 seemed to participate in the EE-elicited MPP-LTP after LFS, as LTP was not changed by the antagonism of group I mGluRs. However, we cannot rule out that other MPP stimulation patterns may activate group I mGluRs, since the opposed effects of EE on LTP are described [58]. Furthermore, bidirectional LTD/LTP allows a switch at a given synapse, enabling adaptive changes in synaptic strength, and controlled by the level and timing of endocannabinoids [59]. The MPP-LTP observed upon EE in water conditions required AEA, as the FAAH inhibitor URB597 occluded the potentiation. This suggests that AEA regulation may be a limiting factor for MPP-LTP. AEA is an endogenous TRPV1 agonist and is restricted to certain brain cells and regions [60], facilitates glutamate release at excitatory synapses, and mediates LTD in the CA1 hippocampus [61]. TRPV1 in granule cell dendritic spines intervenes (together with mGluR5, postsynaptic calcium, and AEA) in LTD at MPP synapses [62]. Recent evidence indicates that the lack of TRPV1 alters the main 2-AG and AEA degrading enzymes, as well as CB1Rs localized to excitatory and inhibitory terminals in the outer 2/3 DGML. Thereby, a crosstalk between TRPV1 and CB1R has been proposed [63]. In this scenario, our findings support a switch from MPP-LTD to MPP-LTP when EE is applied. Our findings also suggest that EE modifies the intensity threshold of excitatory synapses. We observed that the switch in long-term plasticity was not dependent on CB1Rs, since the MPP-LTP was not blocked by AM251. Endocannabinoids can also trigger LTP in the hippocampus through stimulation of CB1Rs in astrocytes. However, LTP requires group I mGluRs activated by the release of astrocytic glutamate at distant synapses [64]. This mechanism seems to be unlikely in our model as MPP-LTP was not dependent on group I mGluRs.
In the case of excitatory synapses, postsynaptic calcium has been proposed as a principal mediator for bidirectional LTP/LTD [65]. In this sense, the MPP-LTP after EE may be triggered by an increase in glutamate release and intracellular calcium that can stimulate AEA production. Moreover, PLC activation by a rise in calcium generates 2-AG [66], which may yield to 1-AG due to acyl migration. This may eventually lead to TRPV1 activation [67]. In addition, PLC activity seems to be critical for TRPV1 since the receptor is sensitized by PLC-mediated hydrolysis of phosphoinositides (PI) and its activity is augmented by PI decrease [68]. Intracellular TRPV1 may regulate calcium release from the sarco/endoplasmic reticulum needed for excitatory synaptic plasticity [69]. The biosynthetic AEA enzyme, N-acyl phosphatidylethanolamine phospholipase D (NAPE-PLD), is highly expressed in dentate granule cells [70]. NAPE-PLD is found in the smooth endoplasmic reticulum of excitatory presynaptic terminals [71], as well as in postsynaptic dendrites and spines receiving excitatory synapses [72]. Hence, the generation of AEA or other N-acylethanolamines by the calcium-dependent catalytic activity of NAPE-PLD may cause postsynaptic TRPV1 activation.
Altogether, the cognitive improvement elicited by EE was similar in adult mice exposed to water and those exposed to EtOH during adolescence [8]. Thus, MPP-LTP elicited by EE may be implicated in this ceiling effect in which TRPV1 and AEA seem to be critical players.