Grup de Recerca en Neurobiologia del Comportament (GReNeC), Departament de Ciències Experimentals i de la Salut, Universitat Pompeu Fabra, Parc de Recerca Biomèdica de Barcelona, Barcelona, Spain
Address correspondence and reprint requests to Olga Valverde, MD, PhD, Grup de Recerca en Neurobiologia del Comportament (GReNeC), Departament de Ciències Experimentals i de la Salut, Universitat Pompeu Fabra, C/ Dr. Aiguader, 88, 08003 Barcelona, Spain. E-mail: email@example.com
Ethanol and 3, 4-Methylenedioxymethamphetamine (MDMA) are popular recreational drugs widely abused by adolescents that may induce neurotoxic processes associated with behavioural alterations. Adolescent CD1 mice were subjected to ethanol intake using the drinking in the dark (DID) procedure, acute MDMA or a combination. Considering that both drugs of abuse cause oxidative stress in the brain, protein oxidative damage in different brain areas was analysed 72 h after treatment using a proteomic approach. Damage to specific proteins in treated animals was significant in the hippocampus but not in the prefrontal cortex. The damaged hippocampus proteins were then identified by mass spectrometry, revealing their involvement in energy metabolism, structural function, axonal outgrowth and stability, and neurotransmitter release. Mice treated with MDMA displayed higher oxidative damage than ethanol-treated mice. To determine whether this oxidative damage was affecting hippocampus activity, declarative memory was evaluated at 72 h after treatment using the object recognition assay and the radial arm maze. Although acquisition in the radial arm maze was not impaired by ethanol intake, MDMA treatment impaired long-term memory in both tests. Therefore, oxidative damage to specific proteins observed under MDMA treatment affects important cellular function on the hippocampus that may contribute to declarative memory deficits.
Adolescence is a critical developmental period in which the brain emerges from an immature state to adulthood (Spear 2000). Thus, the impact of drug abuse on the adolescent brain might have severe negative consequences. This is also a time when novel experiences involving drugs such as ethanol or psychostimulants may be sought. Many teens view risky behaviours as exciting and rewarding, ignoring any negative health consequences (Crews et al. 2007). Several reports indicate that consumption of ethanol and the psychostimulant 3, 4-Methylendioxymethampehtamine (MDMA or ecstasy) are common among adolescence and young adults (Barrett et al. 2006). It is well known that both ethanol and MDMA cause neuroinflammation and neurotoxicity in brain areas, including the prefrontal cortex, striatum and hippocampus (Izco et al. 2006; Rodríguez-Arias et al. 2011). Thus, both types of drug abuse, alone or in combination, induce glial activation (Vallés et al. 2004; Ros-Simó et al. 2012), increase reactive oxygen species and oxidative damage (Busceti et al. 2008; Alves et al. 2009; Rump et al. 2010), and proinflammatory cytokines release (Connor et al. 2001; Qin et al. 2008). Furthermore, MDMA induces neuronal terminal loss (Touriño et al. 2010). In those who consume alcohol and MDMA, these neurotoxic processes can lead to neurodegeneration in specific brain regions that may alter a variety of cognitive and performance tasks, including memory and learning processes (Fadda and Rossetti 1998; Parrott 2001). Cognitive dysfunctions are very prominent in several neuropsychiatric disorders and often diminish patients' quality of life (Millan et al. 2012). Several investigations have studied the effects of alcohol and MDMA on memory acquisition (Able et al. 2006; Kay et al. 2011) or short-term memory (Brooks et al. 2002; García-Moreno et al. 2002). However, little is known about the effects of these drugs on memory consolidation and long-term memory. The length of exposure to the drug and the pattern of consumption that will result in neurotoxicity and cognitive alterations remains a matter of debate.
Numerous studies have established that oxidative damage occurs in brain after ethanol and MDMA treatment in rodents and non-human primates (Busceti et al. 2008; Alves et al. 2009; Rump et al. 2010; Collins and Neafsey 2012), but little has been published about the specific oxidatively damaged proteins. Damage that negatively affects protein function can be measured by detecting carbonyl formation on amino acid side-chains (Levine et al. 1994; Tamarit et al. 2012). Therefore, identification of these proteins after treatment with ethanol, MDMA or both in brain areas involved in cognitive functions (Morris et al. 1982) can provide useful data to better understand how these impaired protein functions can affect these brain areas.
To test whether brain areas involved in cognitive functions become altered after ethanol and MDMA treatments, we carried out declarative memory tasks, which rely on the potentially affected areas. One of the tasks was the object recognition test, involving the hippocampus and the adjacent perirhinal cortex; which are strongly interconnected (Wixted and Squire 2011). The other task was the radial arm maze, which involves different aspects of spatial reference and working memory, involving the hippocampus and the prefrontal cortex respectively. In both tasks, memory consolidation was evaluated.
Materials and methods
All animal care and experimental procedures were conducted according to the guidelines of the European Communities Directive 86/609/EEC regulating animal research and were approved by the local ethics committee (CEEA-PRBB). The results of all studies are reported in accordance with the ARRIVE guidelines for reporting experiments involving animals (McGrath et al. 2010). All procedures were performed by an experimenter blind to the treatment.
A total of 120 adolescent male CD1 mice were used in this study. Mice (PND 21) were purchased from Charles River (Barcelona, Spain) and individually housed upon arrival. Animals remained on quarantine (7 days) until the beginning of the experiment (PND 28), weighing 18–20 g. Animal rooms were controlled for temperature (22 ± 1°C), humidity (55 ± 10%) and reversed light/dark photo-period (lights on between 20:00 h and 08:00 h). All the experiments took place during the dark phase. Food and water were available ad libitum except when water was substituted for ethanol for 2 or 4 h per day according to drinking in the dark (DID) procedure, as previously described by our team (Ros-Simó et al. 2012) with a few modifications from the original procedure (Rhodes et al. 2005). Briefly, DID is voluntary binge ethanol drinking that consists of replacing food and water bottles with a 20% (v/v) ethanol solution or water cylinders that remained in place for 2 h. After this 2-h period, food and water bottles were replaced again. This procedure was repeated on days 2 and 3 and fresh fluids were provided each day. On day 4, DID procedure lasted for 4 h (DID 1). During the following week, the DID procedure was repeated again (DID 2). In addition, on day 4, subjects received two injections of a neurotoxic MDMA dose (20 mg/kg, i.p.) or saline (i.p.), the first injection at the beginning of the 4-h DID procedure and the second two hours later, as previously described (Ros-Simó et al. 2012).
Racemic MDMA hydrochloride was purchased from Lipomed, A.G. (Arlesheim, Switzerland), dissolved in 0.9% physiological saline to obtain a dose of 20 mg/kg (2 mg/mL) expressed as the salt, and injected in a volume of 0.1 mL/10 g body weight by intraperitoneal (i.p.) route of administration. Ethyl alcohol was purchased from Merck Chemicals (Darmstadt, Germany) and diluted in tap water to obtain a 20% (v/v) ethanol solution.
Protein carbonylation analysis
To obtain crude extracts for protein oxidation analyses mice were killed 72 h after the last MDMA or saline injection; hippocampus and prefrontal cortex were immediately extracted from each mouse. The tissue was solubilized with 100 μL of 50 mM Tris-HCl (pH 7) with mammalian protease inhibitors (Roche, Roche Farma, Madrid, Spain) and 2% Sodium Dodecyl Sulphate (SDS). Samples containing 250 μg of protein were treated with the fluorescent probe Bodipy-FL-hydrazide to derivatize protein carbonyls, as described (Tamarit et al. 2012). Briefly, protein samples were mixed with an equal volume of 50 mM Bodipy-FL hydrazide in 0.1 M sodium acetate pH 5, 1 mM EDTA and 1% SDS and incubated at 25°C for 30 min. The reaction was stopped with 0.2 M NaCNBH4. The derivatized proteins were TCA precipitated and concentrated by centrifugation. The pellets were washed by repeating 3 cycles of ethanol/ethyl acetate (50% v/v), and then centrifuged 5 min at 16 870 g. For two-dimensional (2D)-gel electrophoresis, pellets were air dried and re-suspended in Protein Extraction Reagent type 4 (CO356, Sigma Chemical Co., Madrid, Spain), which consists of 40 mM Trizma Base, 7.0 M urea, 2.0 M thiourea and 1% C7BzO with low agitation for 2 h. A volume of 200 μL containing 100 μg of protein was used to rehydrate 11 cm IPG strips (3–10 NL, BioRad) for 12 h and then focused using a Protean IPG Cell (BioRad, Madrid, Spain) for a total of 31250 Vhr. After isoelectric focusing, strips were incubated for 15 min in 6 M urea, 2% SDS, 20% glycerol, 375 mM Tris pH 8.8 and 130 mM Dithiothreitol (DTT). For alkylation, strips were incubated with 135 mM iodoacetamide in the same buffer (without DTT) for an additional 15 min period. Two parallel 11-cm strips (samples from control and treated animals) were loaded side by side on the same 10% polyacrylamide gel using an Ettan Dalt-6 electrophoresis system. Bodipy-derivatized proteins were visualized by epifluorescence in a VersaDoc MP4000 imager (BioRad) using specific channel settings for Bodipy-FL. After obtaining these images, total protein staining was performed with Flamingo (Bio-Rad). Images were captured with a VersaDoc MP4000 and analysed with PDQuest software (Bio-Rad). For each spot, Bodipy signal was normalized to that of the protein signal. Spots displaying a Bodipy-fluorescence/protein signal ratio greater than 1.5 with respect to matched controls were selected for further identification. Protein identification was performed in Autoflex-Speed MALDI ToF/ToF mass spectrometer (Bruker Daltonics, Bruker Española, S.A., Madrid, Spain). The acquired spectra were processed with Flex Control version 3.0 software (BrukerDaltonics). MS spectra were externally calibrated using Peptide Calibration Standard II (BrukerDaltonics). Proteins were identified by peptide mass fingerprint, searching against the SwissProt database using MASCOT Server2.3 (www.matrixscience.com). Peptide mass fingerprinting assumes that peptides are monoisotopic, oxidized at methionine residues (variable modification), and carbamidomethylated at cysteine residues (fixed modification). One maximum missed cleavage was allowed with a maximum peptide mass tolerance of 100 ppm.
Hippocampus-related memory tasks
Object recognition test
The test was performed in a black rectangular open field (27 × 31 × 25 cm) in a moderate luminosity (100 lx). To evaluate the task, a discrimination index was calculated as the difference between the time spent exploring either the novel or familiar object divided by the total time exploring the two objects. A higher discrimination index is considered to reflect greater memory retention for the familiar object (Maccarrone et al. 2002). Behaviour was recorded and monitored with a video camera interfaced to a computer. Experimental procedure (Fig. 1a) consisted of habituation to the box, without objects, for 9 min on day one (D1 of the 2nd week). After that, mice were exposed to the corresponding 2 h DID procedure. Acquisition was performed from D2 to D4 for 9 min each day with subsequent exposure to the corresponding treatment (on D2 and D3, 2 h DID; on D4, MDMA 20 mg/kg × 2 or saline and 4 h DID, as detailed above). Acquisition was performed with two identical objects (familiar object) which were maintained all 3 days. After 72 h the testing phase took place, and one familiar object was replaced by a novel object.
Radial arm maze test
In a separate group of mice, radial arm maze was assessed. The radial arm maze (Panlab, Barcelona, Spain) made of black Plexiglas consisted of a central hub with eight-arms radiating from it, placed on a 100 cm high tripod. The maze arms were 35 cm in length, with outer arm walls 2.6 cm high, inner arm walls 15 cm high and 5.8 cm wide. The centre well of the maze was 16.7 cm in diameter. The food wells at the end of each arm consisted of a small piece of black Plexiglas, 1.2 cm high and 1.8 cm in diameter. The maze was situated in the middle of a dark room and surrounded by black curtains reaching from the ceiling to the base of the tripod, forming a circular enclosure 1.5 m in diameter and with moderate luminosity (100 lx). Three different extra-maze white and black visual cues (20 × 25 cm) to aid in spatial localization were affixed to the curtains. Behaviour was recorded and monitored with a video camera interfaced to a computer situated outside the curtains.
Animals batched for this experiment were food restricted to meet experimental requirements (15% of its initial weight). During the 1st and 2nd day of the habituation period animals were placed in the maze for 8 min each day (Fig. 1b). The 3rd day, one chocolate food pellet (Choco Krispis, Kellogg's) was left in the middle of each arm maze and on the 4th day the pellets were moved to the wells. Both days mice were placed in the maze until the mouse collected all chocolate food pellets or 8 min had passed, whichever came first. The following week, coinciding with the 1st week of the DID procedure; training sessions were carried out for 12 days (D1 to D12, being D12 the last day of the DID procedure) when acute MDMA or saline treatments were administered. During this acquisition period, animals were trained to find the chocolate pellets in three randomly selected arms. Animals were subjected to two trials of 5 min/day or until the mouse collected all the pellets. The day of the test (72 h after the last day of acquisition) animals were subjected to the same procedure. Working memory errors (i.e. entries into baited arms that had already been visited during the same trial) and spatial reference memory errors (i.e. entries into non-baited arms) were recorded. The average of the values from the two daily sessions was calculated for each measure.
Behavioural results are expressed as the mean ± SEM. Statistical analysis for treatment factor was determined by a one-way anova test and subsequent Tukey post hoc test when required. Two-way anova repeated measures (with treatment and day factors of variation) was used to analyse acquisition period in the radial arm maze test. In all experiments, differences were considered significant if the probability of error was less than 5%. SPSS (SPSS Inc., Chicago III, USA) statistical package was used.
Ethanol consumption was measured every day during the DID procedure. Measurements from the second week of the DID procedure are presented in Fig. 2. Although ethanol consumption does not differ from D1 to D3, animals receiving MDMA on D4 consumed less ethanol than those that did not receive the psychostimulant (p <0.05). This decreased intake could be attributed to MDMA-induced acute effects, such as hyperactivity, anxiety and excitability, which may prevent the animal's usual fluid intake.
Protein oxidative damage analysis
Proteins from hippocampus and prefrontal cortex were analysed by 2D gel electrophoresis. Although protein oxidative damage in the prefrontal cortex did not differ between treated animals and controls (data not shown), a significant difference was obtained in the hippocampus. Proteins were analysed using two pI ranges (5–8 and 7–10) to obtain a better resolution (Fig. 3a). Representative gels for carbonyl detection (BODIPY signal) are shown in Fig. 3b and c. When applying image gel analyses (described in 'Materials and methods'), 10 different spots displayed a significant increase in carbonylation, as shown in the enlarged gel sections (Fig. 4), and identified as listed in Table 1. As observed, cell functions mainly affected in the hippocampus are those involved in energy metabolism, cytoskeleton, axonal outgrowth and stability, and neurotransmitter.
Table 1. Carbonylated proteins identified in hippocampus of mice submitted to different treatments
Proteins with score greater than 50 are significant (Probability-based Mowse Score).
The oxidation fold is the mean value ± SEM of fold increase in carbonyl levels of matched pairs (treated vs. control water + saline). Data were obtained from hippocampus of three to four mice per group. Duplicates of each sample were run in three independent gels.
Heat shock cognate 71
2.44 ± 0.38
2.19 ± 0.29
5.52 ± 0.89
Heat shock cognate 71
2.42 ± 1.05
2.06 ± 0.19
2.40 ± 0.26
3.95 ± 2.05
1.69 ± 0.51
3.35 ± 1.27
1.55 ± 0.39
1.54 ± 0.22
2.32 ± 0.49
ATP synthase subunit β
3.19 ± 1.05
1.43 ± 0.79
1.88 ± 0.98
2.32 ± 0.41
1.97 ± 0.15
2.12 ± 0.21
4.02 ± 1.72
0.66 ± 0.10
1.60 ± 0.35
1.76 ± 0.30
1.36 ± 0.16
1.87 ± 0.34
1.96 ± 0.51
1.76 ± 0.11
2.55 ± 0.63
ATP synthase subunit α
1.93 ± 0.49
1.49 ± 0.32
1.81 ± 0.94
ATP synthase subunit α
2.17 ± 0.51
2.14 ± 0.64
2.10 ± 0.26
2.83 ± 0.96
2.15 ± 0.15
2.39 ± 0.20
1.85 ± 0.43
2.17 ± 0.73
2.41 ± 0.59
2.28 ± 0.33
1.64 ± 0.32
3.72 ± 2.62
2.56 ± 0.65
1.65 ± 0.46
2.06 ± 0.45
The three treated groups exhibited oxidatively modified enzymes in energy production pathways. Two of them were glycolytic enzymes (α-enolase and glyceralehyde-3P-dehydrogenase) and three were mitochondrial (aconitase, involved in Krebs cycle, and α and β subunits of the ATP synthase, the major ATP-producing enzyme inside the cell). Oxidation of these proteins suggests that hippocampus activity may be impaired because of the decreased ATP production.
Three additional targets were cytoskeletal proteins involved in axonal and dendritic outgrowth and stability. (i) dihyropyrimidinase-related protein 2 (CRMP-2), a tubulin binding protein that regulates the process of axonal outgrowth and branching; (ii) actin, a critical element in the neuron cytoskeleton involved in cell vesicle formation and movement, cell motility, junction and signalling, and (iii) α-internexin, involved in neuronal development, axonal outgrowth and regulation of axonal stability. CRMP-2 and actin were highly carbonylated in MDMA-treated mice but not in ethanol-only treated mice. α-internexin showed similar oxidation folds in all three treated groups compared to control (water + saline).
Synapsin-1, carbonylated in the three treated groups, is a vesicle-specific protein implicated in neurotransmitter release to the synaptic cleft and has a role in synapse formation and maturation.
Finally, heat shock cognate protein 71 (HSC71), a chaperone of the HSP-70 family, has a role in protection against apoptosis.
Hippocampus-related memory tasks
The data concerning protein oxidative damage indicate that important cell functions are compromised in hippocampus of treated animals. These data prompted us to evaluate whether the affected cellular functions can be related to impairment of tasks relying on hippocampus activity.
MDMA-treated mice show long-term memory deficits in the object recognition test
We first determined the effects of the drugs of abuse studied in the object recognition test, assessed 72 h after the last MDMA or saline injection (Fig. 1a). One-way anova revealed effects of the treatment [F(3, 40) = 4.51, p <0.01] (Fig. 5). Post hoc analysis showed a significant effect in MDMA-treated (p <0.05), and ethanol + MDMA-treated mice (p <0.01), compared to the water + saline control group. These results indicate that long-term memory is only affected after an acute and neurotoxic MDMA dose (20 mg/kg × 2).
Radial arm maze acquisition is not impaired by ethanol consumption
We next examined whether ethanol consumption could be affecting the acquisition process of the radial arm maze in which mice were trained for 12 days (Fig. 1b). Ethanol intake after daily training sessions in the radial arm maze did not impair memory acquisition, as indicated by two-way anova (treatment and day factors). Statistical analysis indicated working memory errors (Fig. 6a shows no treatment effect [F(1, 50) = 0.003, n.s.], effect depending on the day [F(11, 550) = 7.584, p <0.001] and no interaction between day and treatment factors [F(11, 550) = 0.491, n.s.]) and reference memory errors (Fig. 6b, again no treatment effect [F(1, 50) = 1.374, n.s.], but effect depending on the day [F(11, 550) = 8.583, p <0.001], no interaction between these factors [F(11, 550) = 0.480, n.s.]). Altogether, these results indicate that ethanol intake did not affect the acquisition task as they compared to controls.
MDMA impairs reference memory but not working memory in the radial arm maze
To assess the effects of acute MDMA, alone or added to ethanol intake, on the consolidation of the previous acquired task, we tested mice in the radial arm maze 72 h after the last MDMA or saline injection (Fig. 1b). Statistical analysis showed no treatment effects on working memory [F(3, 51) = 1.34, n.s.] (Fig. 6c) but effects on spatial reference memory [F(3, 51) = 4.89, p <0. 01] (Fig. 6d), as indicated by one-way anova. Post hoc analysis indicated that differences in reference memory errors (Fig. 5d) were observed in MDMA-treated (p <0.05) and ethanol + MDMA-treated mice (p < 0.05), compared to water + saline group. Our results indicate that MDMA treatment alone or in combination with ethanol impairs consolidation of spatial reference memory in mice.
Under our experimental conditions, both ethanol intake and acute MDMA-induced damage to specific hippocampus proteins in CD1 mice. MDMA-induced damage was more prominent than ethanol-induced effects, particularly to proteins related to structural functions and to outgrowth and stability. Indeed, an acute neurotoxic dose of MDMA affected consolidation of declarative memory in two different paradigms. Under this pattern of ethanol consumption, animals did not exhibit significant memory deficits compared to controls. The MDMA-induced memory impairments could be related to the specific oxidatively damaged hippocampus proteins.
Previous studies have revealed a clear involvement of oxidative stress in the neurotoxicity induced by MDMA and ethanol consumption (Alves et al. 2009; Rump et al. 2010). These studies led us to hypothesize that ethanol and MDMA could induce protein oxidation in discrete brain areas, and these changes could be associated with cognitive impairment. Therefore, our aim in this study was to identify oxidatively damaged proteins in different brain areas involved in learning and memory processes, specifically the hippocampus and prefrontal cortex, of mice treated with ethanol and/or acute MDMA. Oxidatively damaged proteins are known to be one of the most important causes of brain protein damage and dysfunction (Berlett and Stadtman 1997). Protein carbonyl formation is one of the most studied markers resulting from oxidative stress and can be easily detected and quantified (Levine et al. 1994; Tamarit et al. 2012). Under our experimental conditions, proteins involved in a variety of cellular functions were found oxidatively modified in hippocampus but not in prefrontal cortex (data not shown) by intake of ethanol, MDMA or its combined treatment. These included α-enolase, glyceralehyde-3P-dehydrogenase, aconitate hydratase, α and β subunit of the ATP synthase, CRMP-2, actin, α-internexin, synapsin-1 and HSC 71. Carbonylation, and thus, inactivation of proteins related to energy metabolism may contribute to an energy deficiency associated with drug use. This result could explain the observation that MDMA decreases brain ATP production, leading to membrane ionic dysregulation, calcium entry and additional free radical formation (Darvesh and Gudelsky 2005). In addition, carbonylation of proteins related to energy metabolism has been found in brains affected by Huntington's (Sorolla et al. 2010), Parkinson's (Malkus et al. 2009) or Alzheimer's (Castegna et al. 2002) neurodegenerative diseases. Furthermore, Castegna and colleagues (Castegna et al. 2002) report that in Alzheimer's disease CRMP-2 and HSC 71 proteins were oxidatively modified in the hippocampus. Moreover, synapsin-1 is involved in neurotransmitters release (Evergren et al. 2007) and hippocampal neuronal development (Fornasiero et al. 2009). Finally, actin and α-internexin (also known as neurofilament 66) are proteins involved in neuronal stability and axonal growth (Levavasseur et al. 1999). Considering the well-established MDMA-induced dopaminergic and serotonergic axon terminal loss (Sprague et al. 1998; Touriño et al. 2010), carbonylation of these proteins could well be involved in this mechanism of neurodegeneration. Our findings support protein oxidation as an additional important contributor to the mechanisms underlying the hippocampal neurotoxicity of ethanol and MDMA. Indeed, some researchers have reported attenuation of ethanol and MDMA-induced neurotoxicity by free radical scavengers and antioxidants (Colado and Green 1995; Crews et al. 2006), providing indirect evidence for the involvement of oxidative damage in the mechanism of ethanol and MDMA neurotoxicity.
To analyse whether this oxidative damage can be connected to a cognitive dysfunction, we evaluated the possible impairment in declarative memory tasks induced by the exposition to ethanol, MDMA and its combined treatment in mice. Under our experimental conditions, acute MDMA but not ethanol-induced memory impairments. No increased memory deficits were observed with the combination of both drugs of abuse indicating that such impairment is clearly related to MDMA administration. Previous studies have reported memory deficits with the administration of either drug alone without an enhancement of the deficits when ethanol and MDMA were administered together (Vidal-Infer et al. 2012).
Memory deficits induced by MDMA treatment were observed in the object recognition test 72 h. post-treatment even after three training days, which was performed to ensure a good acquisition of the task. In addition, in the radial arm maze test, animals treated with the psychostimulant performed significantly more spatial reference memory errors (Fig. 6d) than on the last day of acquisition, before MDMA administration (Fig. 6b, day 12). In contrast, those that did not receive MDMA performed the same number of spatial reference memory errors as the last training day. Several different learning and memory tasks following MDMA have been studied in rodents. For instance, non-spatial memory has been evaluated resulting in memory impairment in MDMA-treated rats (Camarasa et al. 2008). In agreement with our results, rats in the radial arm maze (Kay et al. 2011) or rats and mice in the Morris water maze (Camarasa et al. 2008) showed signs of spatial reference memory impairment without alterations in working memory, which depends more on prefrontal cortex. However, in all these studies animals received the MDMA treatment before the acquisition period; in our procedure animals had already acquired the task before the treatment. Thus, our results indicate that MDMA affects the process of consolidating previous learning. In line with these data, in previous murine studies MDMA impaired acquired tasks in behavioural paradigms, such as operant-delayed alternation task involving working memory but not spatial reference memory (Viñals et al. 2012) or active avoidance performance (Trigo et al. 2008), involving emotional memory.
There is evidence that human alcoholics show deficits in spatial memory tasks (Bowden and McCarter 1993), and similar results have been found in animal models (Kameda et al. 2007). However, some authors do not find such damage after ethanol consumption (Popović et al. 2004). In our study, ethanol consumption did not provoke deficits in the consolidation of any of the tasks evaluated, and no effect was observed in ethanol-treated adolescent animals during the acquisition period in the radial arm maze. These ethanol-induced neural changes and the potential for recovery seem to be dependent on length of ethanol exposure, volume of ethanol, degree of withdrawal signs or number of binge bouts, genetics and age (Crews and Nixon 2009). Thus, a pattern of binge exposure during adolescence impacts the developing brain and induces neural consequences, such as cognitive and behavioural dysfunctions (Guerri and Pascual 2010). However, our animals have low preference for ethanol as previously reported (Ros-Simó et al. 2012), which probably could not be considered as a pattern of binge drinking (Fig. 2). Thus, there is the possibility that the amount of ethanol ingested is not enough to produce alterations on memory consolidation, even though it can lead to long-term emotional-like behaviour alterations (Ros-Simó et al. 2012).
In agreement with this hypothesis, it has been reported that low doses of ethanol (0.5 g/kg) administered 30 min before the first training session on each day for 4 days did not impair spatial learning in rats (Acheson et al. 2001). Although acute ethanol induces memory impairments, after repeated treatment with low doses (0.6 g/kg) a tolerance develops to the amnesic effects (Kameda et al. 2007).
Interestingly, proteins carbonylated in MDMA-treated mice are involved in memory processes. In this context, CRMP-2 plays a critical role in axonal outgrowth and pathfinding through the transmission and modulation of extracellular signals (Fukata et al. 2002). Furthermore, it has been reported to be oxidatively modified in the brain in Alzheimer's disease and has been related to memory loss associated with decreased interneuronal connections and to shortened dendritic length (Coleman and Flood 1987). Other works have reported similarities between hippocampus damage related to memory deficits induced by MDMA abuse and Alzheimer's disease (Busceti et al. 2008). Actin, only carbonylated in MDMA-treated subjects, has a crucial role in cytoskeleton network integrity (Fletcher and Mullins 2010) and is concentrated in dendritic spines where it can produce changes in their shape that might be involved in memory function (Morgado-Bernal 2011). Thus, it could be suggested that both proteins may be involved in the MDMA-induced cognitive impairments observed.
We must take into account that protein carbonyl formation was also analysed in the prefrontal cortex with no significant results. Thus, our findings suggest that the impact of the MDMA-induced oxidative damage to specific proteins is distinctive for hippocampus, the brain area mainly involved in declarative memory (Morris et al. 1982) and where synaptic plasticity associated with the process of learning occurs (Malenka and Nicoll 1999). In addition, MDMA-induced energy dysfunction may enhance the loss of interneuronal connections (Butterfield et al. 2006). Ethanol-treated subjects also exhibited oxidatively damaged proteins, indicating that brain functions might be affected. However, ethanol-treated animals did not display memory deficits suggesting that other alterations may appear as a result of ethanol intake, as previously described (Ros-Simó et al. 2012).
In summary, MDMA but not ethanol intake affects consolidation of declarative memory in adolescent mice. Under this pattern of consumption, ethanol does not seem to affect learning acquisition in mice. The observed oxidative damage to specific proteins in hippocampus, especially those related to axonal and dendritic outgrowth and stability, may contribute to the cognitive deficits observed. However, we cannot discard other alterations, besides memory deficits, as a consequence of oxidative stress. Thus, studies of specific protein damage may provide new insights in understanding the MDMA and ethanol mechanisms of neurotoxicity and its behavioural consequences.
Authors thank to Alba Sorolla and Isabel Sánchez from “Servei de Proteòmica i Genòmica” UdL, for helping with peptide mass fingerprinting analyses. We also thank Elaine M. Lilly (“Writer's First Aid”) for the editorial assistance.
This study was supported by grants from Spanish Ministry of Science and Innovation (SAF2010-15793 and CSD2007-00020); the Spanish Ministry of Health (PNSD 2010); Red Temática de Investigación Cooperativa en salud (ISCIII) (RETIC-Trastornos Adictivos RTA 001/06/1001 and 12/0028/0024-FEDER) and Generalitat de Catalunya (2009SGR684). CR-S was funded by a FPI fellowship (BES-2008-007915).