Effects of Stress and MDMA on Hippocampal Gene Expression

MDMA (3,4-methylenedioxymethamphetamine) is a substituted amphetamine and popular drug of abuse. Its mood-enhancing short-term effects may prompt its consumption under stress. Clinical studies indicate that MDMA treatment may mitigate the symptoms of stress disorders such as posttraumatic stress syndrome (PTSD). On the other hand, repeated administration of MDMA results in persistent deficits in markers of serotonergic (5-HT) nerve terminals that have been viewed as indicative of 5-HT neurotoxicity. Exposure to chronic stress has been shown to augment MDMA-induced 5-HT neurotoxicity. Here, we examine the transcriptional responses in the hippocampus to MDMA treatment of control rats and rats exposed to chronic stress. MDMA altered the expression of genes that regulate unfolded protein binding, protein folding, calmodulin-dependent protein kinase activity, and neuropeptide signaling. In stressed rats, the gene expression profile in response to MDMA was altered to affect sensory processing and responses to tissue damage in nerve sheaths. Subsequent treatment with MDMA also markedly altered the genetic responses to stress such that the stress-induced downregulation of genes related to the circadian rhythm was reversed. The data support the view that MDMA-induced transcriptional responses accompany the persistent effects of this drug on neuronal structure/function. In addition, MDMA treatment alters the stress-induced transcriptional signature.


Introduction
3,4-Methylenedioxymethamphetamine (MDMA) is a ringsubstituted amphetamine analog. The primary effect of this drug is to alter perception, cognition, or mood. People who take it recreationally generally experience a feeling of elation [1]. Its short-term euphoria-inducing properties have made MDMA (ecstasy) a popular drug of abuse. These mood altering effects may also prompt MDMA intake by individuals who desire to overcome stressful experiences. In particular, MDMA treatment is under study to mitigate the clinical symptoms of PTSD [2]. However, it is not known whether or how the effects of MDMA may be altered in the context of chronic stress. It is conceivable that the pathophysiologic changes induced by stress could alter the efficacy or increase the toxicity of MDMA.
The hippocampus is a brain area that may be particularly susceptible to the effects of both MDMA and chronic stress. Chronic unpredictable stress suppresses neurogenesis within the hippocampus [3] and reduces dendritic complexity and spine density in this brain region [4,5]. Chronic stress also increases the vulnerability of hippocampal neurons to damage produced by neurotoxins and drugs [6,7]. The repeated administration of MDMA produces welldocumented, persistent deficits in biochemical markers of 5-HT axon terminals in multiple brain regions, including the hippocampus, in rodents and humans [8][9][10][11]. These findings have been viewed as evidence of MDMA-induced distal axotomy of 5-HT neurons [12]. Moreover, exposure to chronic unpredictable stress potentiates MDMA-induced 5-HT neurotoxicity in the hippocampus [13].
Beyond the persistent effects of MDMA on 5-HT axon terminals, there is evidence that MDMA produces neuronal degeneration within multiple brain regions, including the hippocampus [14][15][16][17]. Although the identity of the targeted neuronal populations has not been determined, Anneken et al. [18] and Perrine et al. [19] have reported that MDMA reduces biochemical markers of hippocampal GABAergic neurons.
The hippocampus is known to be critical for learning and memory function [20]. Deficits in spatial learning and memory [21][22][23][24] and impairments in developing a conditioned place preference to sex [25] have been reported in rats treated with MDMA. In human abusers of MDMA, the most consistent finding is that of impairments in short-term memory, particularly verbal recall [26,27], as well as deficits in tasks that assess executive functioning [28,29].
The purpose of the present study was to investigate the interaction between chronic stress and MDMA treatment within the hippocampus at the molecular level, as evaluated from transcriptional responses quantified by microarray analysis.

Animals and Stress
Exposure. Adult, male rats (250-300 g) of the Sprague-Dawley strain (Harlan Laboratories, Indianapolis, IN) were used in this study. Animals had free access to food and water in a temperature-and humiditycontrolled room. All procedures were performed in adherence to the NIH guidelines and were approved by the institutional animal care and use committee.

Drug Treatment.
Control and stress-exposed rats were injected with racemic 3,4-methylenedioxymethamphetamine (MDMA) (generously provided by the National Institute on Drug Abuse) or vehicle (0.15 N NaCl) 24 hr following exposure to the last stressor. The rats received MDMA (10 mg/kg, ip) or vehicle at 2 hr intervals for a total of 4 injections. The rats were euthanized for removal of the hippocampus 24 hr following the first injection of MDMA or vehicle.

RNA Extraction and
Hybridization. RNA was extracted from the hippocampi with 1.5 mL TriReagent according to the instructions by the manufacturer. The RNA samples were quantified by UV absorbance (260 and 280 nm) and their integrity was confirmed by Agilent Bioanalyzer 2100. Intact total RNA amounts of at least 50 ng were amplified using the NuGen Applause WT-Amp Plus ST system (http://www.nugeninc.com/nugen/index.cfm/products/apl/ applause-rna-amplification-systems/). The samples were then biotinylated with the Ambion Biotin-Enhanced Message Amp II kit. The GeneChip Rat Gene 1.0 ST Array was hybridized with 2.5 g of fragmented biotinylated aRNA. The hybridization, staining, and washing were carried out using the Affymetrix GeneChip Hybridization Wash and Stain Kit following the manufacturer's protocols. Briefly, the arrays were hybridized for 16 hr at 45 ∘ C, followed by protocol FS450 0007 for staining and washing the GeneChips in the Affymetrix Fluidics Station 450. The GeneChips were scanned with Affymetrix GeneChip Scanner 3000 7 G using GCOS and software and preset settings. The Rat Gene 1.0 ST chip provides comprehensive coverage of the transcribed rat genome by analyzing over 27,000 protein coding transcripts, over 24,000 Entrez genes, and alternatively spliced transcript variants with probes designed to maximize coverage of exons. For each treatment group, there were 4-5 hybridizations.

Microarray Analysis.
The data were analyzed to identify differentially expressed genes between the following conditions: (1) no stress saline versus stress saline, (2) no stress MDMA versus stress MDMA, (3) no stress saline versus no stress MDMA, and (4) stress saline versus stress MDMA. Four and five biological replicate arrays for Saline and MDMA samples, respectively, were performed.
Once the arrays had been scanned and the data files were generated, the results were analyzed by the UC Laboratory for Statistical Genomics and Systems Biology. Analysis was performed using statistical software R and the limma package of Bioconductor [31]. Data preprocessing, including background correction, normalization, and expression set summaries, was performed using RMA. Array quality was assessed using arrayQualityMetrics package of Bioconductor [32]. Estimated fold change for each comparison was calculated using ANOVA, and resulting -statistics from each comparison were modified using an intensity-based empirical Bayes method (IBMT) [33]. Genes with value < 0.05, absolute fold change > 1.2, and average intensity > 100 fluorescent units were considered significantly differentially expressed. These cutoffs are consistent with similar studies in the literature [34][35][36][37]. Significant genes were further analyzed by Fisher's exact tests to identify overrepresented gene ontology categories and KEGG pathways. For these categories and pathways, a false discovery rate (FDR) below 0.05 and an odds ratio (log OR) outside the interval 0.67 to 1.5 were considered significant. To eliminate GO categories that are too broad to be informative, we focused on categories of fewer than 100 genes.

PCR Array and Real-Time RT-PCR.
The microarray results for untreated and MDMA-treated rats were validated with the rat heat shock protein and chaperone PCR array (SA Biosciences), which profiles the expression of 84 Heat Shock Protein genes that regulate protein folding. The array contains a set of optimized real-time PCR primer assays on a 96-well plate for pathway focused genes as well as appropriate RNA quality controls. It performs gene expression analysis with real-time PCR sensitivity and the multigene profiling capability of a microarray. The assay was run and evaluated at the SA Biosciences PCR Array Service. For Per2, we measured the gene expression levels by real-time RT-PCR, using primers 5 -CTTCTGGTCTGG-ACTGCACA-3 (forward) and 5 -TGAGTCTGAGGTGGC-AGATG-3 (reverse), 1.5 mM magnesium, and 55 ∘ C annealing temperature for 40 cycles. Actin served as a loading control, and a reference cDNA validated each run.

Results
Exposure to chronic, unpredictable stress led to an overall moderate genetic response that comprised increases and decreases in gene expression levels. The consecutive administration of MDMA dramatically altered the profile of stressresponsive genes as follows.
(i) 118 genes were suppressed by stress alone. Most of these (i.e., 98) were distinct from the 99 genes suppressed by stress with ensuing exposure to MDMA (Figure 1(a)). Remarkably, there were 20 genes suppressed by stress in a fashion that was resistant to the consecutive exposure to MDMA (Table 1). (ii) 132 genes were upregulated by stress alone. Most (i.e., 128) of these genes differed from the only 34 genes upregulated after stress with ensuing administration to MDMA (see Figure 1(a)). The significance of the four genes (hypothetical protein LOC689399, hypothetical protein LOC680454, and hypothetical protein LOC686031, similar to 60S ribosomal protein L7a) induced by stress in a MDMA-resistant manner is unclear as all of them encode hypothetical or yet uncharacterized proteins.
(iii) Exposure to chronic stress resulted in altered gene expression in the KEGG pathway rno04710 (circadian rhythm). Four of twelve genes in this group were suppressed by stress, leading to a Fisher FDR of 3.23 × 10 −02 and a log odds ratio of 2.95. However, the stress-induced alteration in gene expression in the circadian rhythm pathway was not evident in stressed rats treated subsequently with MDMA (Table 2). For one of the affected genes, Per2, the stressdependent downregulation and MDMA-mediated reversal thereof was confirmed by real-time RT-PCR (not shown).
Exposure to MDMA led to the altered expression of 1225 genes in the hippocampus, with 588 genes being upregulated and 637 genes being suppressed (Figure 1(b)). Four gene ontology categories were affected (Table 3), comprising unfolded protein binding, protein folding, calmodulindependent protein kinase activity, and neuropeptide signaling pathway. The impact on protein folding seems to be of particular importance in the genetic response to MDMA as evidenced by the differential regulation of a large number of heat shock protein genes, for example, Hsp40, Hspd1, Hspa8, and Hsp90ab1 (Table 4(a)). The MDMA effect on heat shock proteins was validated by real-time RT-PCR array on samples from unstressed MDMA-treated versus unstressed untreated rats, which confirmed the MDMA-induced upregulation of several heat shock proteins and chaperones. In particular, Cct5, Dnajb1, Dnaja4, Dnajb11, Dnajb4, Hsp90ab1, Cct6a, SerpinH1, and Dnaja2 were increased under MDMA treatment in both microarray analysis and real-time RT-PCR (Table 4(b)). A moderate induction (1.2-to 1.4-fold) was seen  The average intensity of all genes shown is above threshold. Fold changes above or below threshold and significant values are shown in bold. On this basis, the italicized genes are judged to be suppressed by stress. in the PCR array in Cct3, Tcp1, and Cct4 (not shown), which were comparably moderately increased in the microarray. According to gene ontology analysis, the genetic response by unstressed rats to MDMA (described in the preceding paragraph) was different from the genetic response by stressed rats to MDMA, indicating that MDMA induces altered responses to tissue damage in nerve sheaths and altered sensory processing in stressed, but not unstressed individuals as follows.

Discussion
The present study examined the effects of chronic stress and subsequent MDMA treatment on gene transcription in the rat hippocampus. The key findings are the following: (1) MDMA alone increases the expression of genes in several gene ontology categories, including protein folding and unfolded protein binding, (2) MDMA upregulates the genes for several heat shock proteins, (3) MDMA negatively affects the expression of genes related to axon sheaths and tissue remodeling in stressed rats, (4) chronic stress alters gene expression within the circadian rhythm pathway, and this effect is absent in rats treated subsequently with MDMA, and (5) the number of stress-responsive genes is lessened by MDMA treatment.
The genetic responses to repeated stress included over two hundred genes that were either up-or downregulated. The reduction in the number of these genes by consecutive MDMA administration implies the potential for a druginduced amelioration of the stress response. Among the genetic adaptations to stress, transcripts in the KEGG pathway for circadian rhythm were downregulated in response to chronic, unpredictable stress. One prominent clinical characteristic of stress disorder, particularly PTSD, in humans is abnormalities in the circadian rhythms of several physiological responses such as cortisol, blood pressure, and sleep [38][39][40]. In the present study, the suppressive effect of stress on gene transcription within the KEGG pathway for circadian rhythm was not manifest in stressed rats treated subsequently with MDMA. There is evidence that MDMA may affect sleep cycles [41], implying that it could have the potential to reverse or reduce stress-induced sleep disturbance. MDMA has been shown to be beneficial in treating the symptoms of PTSD [2]. While it is not yet known whether this clinical response is related to a restoration of circadian rhythms, the anxiolytic drug prazosin can reduce stress-induced sleep disturbances [42], providing a rationale for non-antidepressant restoration of the circadian rhythm. In general, MDMA is under discussion for the treatment of anxiety disorders [43]. This drug significantly altered the pattern of stress-induced gene induction and suppression. It remains to be determined whether the ability of MDMA to alter the genetic response to stress is of significance in the clinical response of PTSD patients to MDMA. One component of the psychopharmacology of MDMA is that human abusers of the drug report enhanced sensory perceptions. In the present study, MDMA administration to chronically stressed rats exhibited an upregulation of related gene categories associated with sensory perception when compared to stressed rats not given MDMA. These findings suggest that MDMA may exert persistent effects on sensory perception that outlast the acute effects of the drug on neurotransmitter release.
Exposure to prior chronic stress impacted the genetic responses to MDMA. Notably, gene categories of tissue regeneration, axon ensheathment, myelin sheath, and cytokinecytokine receptor interaction were significantly downregulated following MDMA treatment to rats exposed to chronic stress when compared to unstressed rats given MDMA or to stressed rats not given MDMA. Hence, chronic stress appears to augment the deleterious effects of MDMA on the neuronal structure. Callahan et al. [44] have concluded that MDMA produces structural damage to axonal transport mechanisms in multiple brain regions. The susceptibility to that damage or the intensity of the damage may be enhanced consecutively to stressing experiences. Although the identity of neurons affected by the combination of stress and MDMA is unknown, it is of interest that chronic, unpredictable stress augments the persistent deficits in serotonergic neurons produced by MDMA [30].
The gene expression changes observed in this study are consistent with the existing literature. The MDMA-induced alterations in gene ontology categories are consistent with transcriptional events associated with drug-induced neuronal damage. Genes related to protein folding and unfolded protein binding were upregulated 24 hours following MDMA treatment. Gene transcripts for heat shock proteins were notably upregulated following MDMA administration. This finding is consistent with previous reports that MDMA increases heat shock protein mRNAs in the mouse striatum [16] and heat shock protein levels in several brain regions [45][46][47]. The effects of MDMA on heat shock protein gene expression are viewed as a response to cellular stress that may accompany MDMA-induced hyperthermia and/or oxidative stress.
Treatment with MDMA also increased gene expression in GO categories related to cell signaling (e.g., neuropeptide signaling pathway and calmodulin-dependent protein kinase activity). This finding is in accord with the reports of Thiriet et al. [48] and Salzmann et al. [16] in which MDMA was shown to increase transcription for several proteins belonging to signal transduction pathways. These changes can be viewed, most likely, as a consequence of 5-HT receptor activity subsequent to the MDMA-induced increase in extracellular 5-HT in the hippocampus [49].
The use of psychedelic drugs to treat emotional disorders has been controversial due to the intrinsic neurotoxicity of these agents [1]. The evaluation of such therapies largely depends, in the short term, on cognitive and psychological evaluations that may be particularly susceptible to placebo effects or interpretive differences. The supplementation of clinical studies with ex vivo molecular assessments may aid in providing mechanistic insights. The present investigation of hippocampal gene expression profiles has elucidated the potential of MDMA to ameliorate the genetic response to chronic unpredictable stress and to reverse stress-induced sleep disorders. However, it has also found evidence that the toxicity of MDMA on nerve sheaths may be increased in individuals previously exposed to stressful experiences.