Intravenous morphine self-administration alters accumbal microRNA profiles in the mouse brain.

A significant amount of evidence indicates that microRNAs (miRNAs) play an important role in drug addiction. The nucleus accumbens (NAc) is a critical part of the brain's reward circuit and is involved in a variety of psychiatric disorders, including depression, anxiety, and drug addiction. However, few studies have examined the expression of miRNAs and their functional roles in the NAc under conditions of morphine addiction. In this study, mice were intravenously infused with morphine (0.01, 0.03, 0.3, 1 and 3 mg/kg/infusion) and showed inverted U-shaped response. After morphine self-administration, NAc was used to analyze the functional networks of altered miRNAs and their putative target mRNAs in the NAc following intravenous self-administration of morphine. We utilized several bioinformatics tools, including Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway mapping and CyTargetLinker. We found that 62 miRNAs were altered and exhibited differential expression patterns. The putative targets were related to diverse regulatory functions, such as neurogenesis, neurodegeneration, and synaptic plasticity, as well as the pharmacological effects of morphine (receptor internalization/endocytosis). The present findings provide novel insights into the regulatory mechanisms of accumbal molecules under conditions of morphine addiction and identify several novel biomarkers associated with morphine addiction.

Introduction

MicroRNAs (miRNAs) are short (~22 nucleotides) non-coding RNAs that play important roles in the regulation of almost all biological processes, including cell development, proliferation, and differentiation, as well as various neurological functions (Im and Kenny, 2012). miRNAs target their complementary binding sites at 3′-untranslated regions (3′-UTR) of mRNA by binding to argonaute (AGO) proteins in miRNA-induced silencing complexes (miRISCs). Subsequently, miRISCs interact with cytoplasmic deadenylase complexes (e.g., PAN2-PAN3 and CCR4-NOT) that catalyze the deadenylation of mRNA target sites. Additionally, miRISCs repress the translation of mRNA target proteins via inhibition of the eukaryotic initiation factor 4F (elF4F) complex (Jonas and Izaurralde, 2015). Previous studies have shown that miRNAs play an important role in the effects of addictive drugs, such as cocaine and methamphetamine (Hollander et al., 2010; Im et al., 2010), but few studies have investigated the regulation of miRNAs in morphine addiction.

Although morphine is considered to be the gold standard treatment for moderate-to-severe pain in patients with and without cancer (Dalal et al., 2012; Raphael et al., 2013), its clinical use is limited due to the rewarding effects and high risk of developing tolerance associated with this drug (Fitting et al., 2016). The primary pharmacological action of morphine is mediated by the activation of mu-opioid receptors (MOR), which are distributed throughout mesolimbic regions, such as the ventral tegmental area (VTA), striatum (STR), amygdala (Amy), and medial prefrontal cortex (mPFC) (Parker et al., 2014; Fields and Margolis, 2015; Kim et al., 2016). Furthermore, molecular and cellular changes in the STR are thought to be one of the most important characteristics involved in understanding the phenotypes of various psychiatric disorders, especially addiction (Nestler, 2004; Tan, 2008; Russo et al., 2010; Imperio et al., 2016).

The striatal complex is subdivided into the dorsal striatum (dST) and the nucleus accumbens (NAc), which comprises a reward circuit with the VTA (Beier et al., 2015). It is widely accepted that the VTA–NAc circuit plays diverse roles in the processes underlying mood disorders and psychostimulant-induced rewarding effects (Nestler, 2004; Russo and Nestler, 2013). Several lines of evidence suggest that there are two main regulatory mechanisms in the VTA–NAc circuit: 1) dopaminergic projections from the VTA to the NAc, and 2) gamma-aminobutyric acid (GABA)ergic projections from the NAc to the VTA. These two projections regulate the neuronal function of both regions in a complementary manner (Kalivas et al., 1993; Qi et al., 2016). Additionally, GABAergic neurons in the NAc receive glutamatergic projections from the hippocampus, Amy, and mPFC; these are known to modulate the rewarding and withdrawal effects induced by morphine (Russell et al., 2016).

Because MOR primarily mediates the pharmacological action of morphine, many studies have focused on the regulation of MOR when attempting to characterize the role that the NAc plays in morphine addiction (Dennis et al., 2016; Molaei et al., 2016). However, the molecular dynamics underlying the rewarding effects of morphine in the NAc remain poorly understood. Therefore, the present study utilized microarrays to evaluate the expression profiles of miRNAs in the NAc in mice that self-administered intravenous morphine. Furthermore, using various bioinformatics tools (i.e., mirWalk and Database for Annotation, Visualization, and Integrated Discovery [DAVID]), the transcriptional networks between the miRNAs and their putative target mRNAs in the NAc were analyzed. In this manner, the present study was able to investigate various molecular profiles and the putative molecular networks that may be implicated in functional changes in the NAc under conditions of morphine addiction.

Materials and methods

Animals

Eighteen 7-week-old male C57BL/6 mice, weighing 23.28 ± 1.56 g, purchased from Daehan Biolink (Chungbuk, Korea), were included in this study. All mice were individually housed under a 12-hour reverse light/dark cycle in a laboratory breeding room at the Korea Institute of Science and Technology (KIST) with ad libitum access to water. For the self-administration study, we randomly divided the mice into two groups, morphine and saline infusion group (8–9 mice per group). All mice were raised under mild food restriction throughout the study (approximately 85–90% of free-feeding body weight). During the morphine self-administration periods, the mice were placed in an operant chamber enclosed within a sound-attenuating cubicle (MED-307ACT-D1; Med Associates, Inc., St. Albans, VT, USA) in a dark room. The Institutional Animal Care and Use Committee of the KIST approved all protocols used in this study (approval no. 2016-081).

Intravenous self-administration (IVSA) of procedure

Apparatus

All procedures were performed using operant chambers (29.5 cm × 32.5 cm × 23.5 cm) with two retractable levers; the left lever was used as the ‘active lever’ that allowed the mice to earn a reward. A cue light was located above each lever and programmed to turn on only when the active lever was pressed. Syringe pumps (Med Associates, Fairfax, VT, USA) were connected to an intravenous catheter placed into the external jugular vein on the back of each mouse with metal spring-covered tubes infused the morphine. The behaviors of the mice during each session were recorded, and all actions were controlled using MED-PC software (Med Associates).

Food training

Prior to IVSA of morphine, all mice were trained to press the active lever rather than the inactive lever to gain a food reward using a fixed-ratio (FR) schedule (Weeks and Collins, 1978). The schedule gradually increased from FR1 to FR5 when mice received 25 food pellets for two consecutive days, and the sessions were completed when mice received 40 food pellets. After finishing the food training at FR5, intravenous catheters were implanted into the right jugular vein of each mouse for morphine infusions. Mice were allowed to recover for 3 days, and then two food-training sessions were performed to reinstate the memory of the active lever press.

IVSA of morphine

After completion of food training, we performed IVSA of morphine. During the session of 2-hour IVSA of morphine, mice received a solution of morphine hydrochloride (Myungmoon Pharm. Co., Ltd., Seoul, South Korea) for 3 seconds at a rate of 0.01 mL/s as a reward for every active lever press; a cue light above the active lever was illuminated for 20 seconds while morphine was being infused. Because preliminary data (not shown) revealed that unit dose of 0.3 mg/kg/infusion was a stable infusion, so it was employed as a ‘training dose’ between the changing of each dose.

Morphine dose and response tests were performed using unit doses of 0.03, 1, 0.01, and 3 mg/kg/infusion. Each dose was tested for 6 days and the average number of rewards earned during the final 3 days was calculated as the reward number (morphine infusion earned). A Latin square crossover design was performed for the tests to control for any order effects during the procedure. Following dose-response tests, a morphine unit dose of 0.3 mg/kg/infusion was presented for 5 days to achieve the optimal induction of morphine addiction and sacrificed immediately after the last IVSA session. A control group of mice underwent the catheter surgery and received yoked saline infusions (13.23 ± 3.27 lever presses). The overall experimental timeline for the morphine IVSA sessions is shown in .

Figure 1

Schematic diagram of intravenous self-administration procedure.

RNA extraction from the NAc and microarray experiment procedures

Mice in the control and morphine IVSA groups were sacrificed and decapitated on the day following the final morphine IVSA session. The extracted brains were frozen and bilateral 14-gauge NAc punch dissections were obtained using a cryostat (n = 2 per group, CM3050S; Leica Biosystems, Nussloch GmbH, Germany). Total RNA was isolated using the RNA STAT-60 kit (Amsbio, Abingdon, UK) according to the manufacturer's instructions. The concentration of the total RNA samples was determined by measuring optical density using the NanoDrop® ND-1000 system (Thermo Scientific, Waltham, MA, USA). miRNA expression profiling was conducted using Affymetrix GeneChip® miRNA 4.0 arrays (Affymetrix, Santa Clara, CA, USA) that contained 3,222 mature and stem-loop miRNAs. The RNA was labeled using the FlashTag™ Biotin HSR Labeling Kit (Genisphere, Hatfield, PA, USA), hybridized to the miRNA array, washed, and then scanned according to the standard Affymetrix array cassette staining protocol. Affymetrix® Expression Console Software version 1.2.1 (Affymetrix) was used for the last scanning step and analysis of the signals. Raw data (*.CEL files) were normalized at the transcript level using a robust multi-array average (RMA) method (Irizarry et al., 2003).

Microarray data analysis

The normalized miRNA array data were analyzed for each array. We employed fixed-fold-change (2-fold) and fixed-P-value (P < 0.05) cutoff to screen for miRNAs that exhibited significant expression patterns. Hierarchical clustering of the miRNAs was performed using the Cluster 3.0 software (http://bonsai.hgc.jp/~mdehoon/software/cluster/software.htm) and visualized with the MultiExperiment Viewer (http://www.tm4.org/).

Target prediction and pathway analysis

Responsive miRNAs were analyzed to predict putative targets using the following four databases: miRwalk 2.0 (zmf.umm.uni.heidelberg.de), miRanda (http://www.microrna.org/), RNA22 version 2.0 (http://cm.jefferson.edu/rna22/Interactive/), and Targetscan (http://www.targetscan.org/) in the miRwalk 2.0 platform (Dweep and Gretz, 2015). Next, the predicted targets that overlapped in at least three databases were selected for further analysis. To identify the functional pathways and reveal the miRNA/target gene regulatory network, Kyoto Encyclopedia of Genes and Genomes (KEGG; http://www.genome.ad.jp/kegg/) pathway analysis was applied using the predicted targets (Ogata et al., 1999). The list of target genes was used as the input for the Database for Annotation, Visualization, and Integrated Discovery (DAVID) Bioinformatics Resources 6.7 (https://david.ncifcrf.gov/) and the Mus musculus background was selected to cluster the KEGG pathways (Huang da et al., 2009).

CyTargetLinker

To show the network of miRNA and putative mRNA targets, we utilized CyTargetLinker (http://projects.bigcat.unimass.nl/cytargetlinker/), which is a plug-in for the Cytoscape software (http://www.cytoscape.org/). Representative images of the miRNA/mRNA networks were derived from the regulatory interaction networks (RegINs) provided by the MicroCosm (http://www.ebi.ac.uk/enright-srv/microcosm/htdocs/targets/v5/), mirTarBase (http://mirtarbase.mbc.nctu.edu.tw/), and TargetScan (http://www.targetscan.org/) databases.

Statistical analysis

All data were reported as the mean ± SEM and analyzed with one-way analysis of variance tests using Prism 6.01 software (Graphpad Software Inc, CA, USA). Fisher's Least Significant Difference post hoc tests for multiple comparisons were performed, and P values < 0.05 were considered to indicate statistical significance.

Results

Mice receiving IVSA of morphine exhibited general drug-seeking behaviors

Prior to morphine infusion sessions, all mice were trained to press the active lever rather than the inactive lever to obtain food pellets as a reward in a 5-day food-training protocol conducted in the operant IVSA chamber ( and ). Following catheter implantation surgery, mice received two additional food-training sessions ( and ) before the reward was changed from food pellets to morphine infusion (0.03 mg/kg/infusion). During the first three morphine sessions, mice earned 26.33 ± 1.39 rewards on average but their mean response level was adjusted to 17.58 ± 0.84 rewards over the following six consecutive 2-hour sessions (Figure 2B, ). The lever responses of the mice were also evaluated across different unit doses of morphine (0.03, 1, 0.01, and 3 mg/kg/infusion; Figure ). For all mice receiving IVSA of morphine, a dose-dependent relationship (0.01, 0.03, 0.3, 1, and 3 mg/kg/infusion) of the average infusions per lever press of the last three sessions revealed an inverted U-shaped curve (Figure ).

Figure 2

Mice that intravenously self-administered morphine exhibited general drug-seeking behavior.

Either food pellets or intravenous morphine infusions was earned as rewards during the consecutive self-administration sessions. (A) Over five consecutive food-training sessions, the number of active lever presses to acquire food pellets gradually increased during the 1-hour sessions. Additionally, the mice underwent two additional food-training sessions following the catheter implantation surgery to reinstate their memory of the procedure. After food training, we conducted 6–9 consecutive self-administration. (B) Number of morphine infusions earned and lever presses performed to receive a morphine infusion (0.3 mg/kg/infusion) during consecutive self-administration sessions. (C) Number of morphine infusions earned and lever presses performed to receive a morphine infusion (0.03 mg/kg/infusion) during consecutive self-administration sessions. (D) Number of morphine infusions earned and lever presses performed to receive a morphine infusion (1 mg/kg/infusion) during consecutive self-administration sessions. (E) Number of morphine infusions earned and lever presses performed to receive a morphine infusion (0.1 mg/kg/infusion) during consecutive self-administration sessions. (F) Number of morphine infusions earned and lever presses performed to receive a morphine infusion (3 mg/kg/infusion) during consecutive self-administration sessions. All data are reported as the mean ± SE (n = 8–9 mice per dose of morphine). FT: Food training; Re-FT: re-food training.

Lever presses in the operant chamber during morphine intravenous self-administration.

All data are reported as mean ± standard error (SE; n = 8–9 mice per dose of morphine). FT: food training; Re-FT: re-food training.

Click here for additional data file.

The number of morphine infusions (F(4, 10) = 114.6, P < 0.001; ) and lever presses (F(4, 10) = 62.10, P < 0.001; ) were significantly altered during the morphine self-administration session. Additionally, there were continuous increases in morphine intake across sessions for all unit doses of morphine infusion, and the volume of intake for each does was significantly changed during morphine self-administration session (F(4, 10) = 156.0, P < 0.001; ).

Figure 3

A typical inverted U-shaped curve in the dose-response test was generated from mice that intravenously self-administered morphine.

(A) Dose-response tests for morphine self-administration revealed relatively high numbers of infusions at unit doses of 0.03 and 0.3 mg/kg/infusions and low numbers of infusions at unit doses of 0.01, 1, and 3 mg/kg/infusions. The numbers of morphine infusions earned over the last 3 sessions were averaged for each dose. (B) The number of lever presses for morphine infusions over the last 3 sessions were averaged for each dose. (C) Total amount of morphine infused at each dose. All data are reported as mean ± SE (n = 8–9 mice per dose of morphine). All data are expressed as the mean ± standard error for each group, and analyzed with one-way analysis of variance.

miRNA profiling in the NAc revealed differential expression in mice receiving IVSA of morphine

There were different patterns of miRNA expression in the NAc regions of the morphine IVSA and control mice (). Of the 3,163 analyzed miRNAs in the NAc, there were significant alterations in 62 miRNAs after IVSA of morphine: 33 miRNAs were upregulated by more than 2-fold and 29 were downregulated by more than half ( and ).

Figure 4

Accumbal miRNAs clustered after intravenous self-administration of morphine and saline.

A heat map was created using significantly altered miRNAs according to the microarray data (n = 2); miRNA expression in the nucleus accumbens of the intravenous morphine self-administration and age-matched drug-naïve control groups were compared. The red color indicates an increased expression level and the green color indicates a decreased expression level.

Table 1

The 62 morphine responsive-miRNAs in the NAca

An integrated analysis of the miRNA putative target network using CyTargetLinker revealed interactions in 62 NAc miRNAs following morphine self-administration

To assess the relationships between the miRNAs and their putative mRNAs, 62 responsive miRNAs were analyzed using CyTargetLinker; the representative networks of the miRNAs and their putative targets were derived from RegINs and the interactions of each network were analyzed within the whole network. Among the upregulated miRNAs, mmu-miR-32-5p was associated with mmu-miR-451a to regulate their common target, Nsmal, and interacted with mmu-miR-202-5p to regulate B3galt2, while mmu-miR-32-5p and mmu-miR-201-5p were associated with the regulation of NM_008211.3 (). Of the downregulated miRNAs, mmu-miR-20b-5p and mmu-miR-193-3p shared the following four common targets: Arid4b, Vldlr, Fbxw11, and Ddhd1 ().

Figure 5

Representative miRNA/target interaction networks in the NAccreated using CyTargetLinker following morphine self-administration.

Based on the miRNA array data, we divided the responsive miRNAs into increased miRNAs and decreased miRNAs groups. After that, we used CyTargetLinker to show the interaction between responsive miRNAs and predicted target gene interaction after morphine self-administration. (A) Increased miRNAs-mRNA and (B) decreased miRNAs-mRNA showed diverse patterns of networks.

miRNAs altered by morphine self-administration had putative targets related to various functions of the NAc

In the present study, 62 responsive miRNAs were analyzed for putative target detection using miRwalk 2.0 web, miRanda, RNA22 version 2.0, and TargetScan. Of the putative targets derived using these four algorithms, 7,752 predicted mRNAs were potentially regulated by 62 miRNAs. A functional categorization of the target genes was performed using the DAVID functional annotation tool (https://david.ncifcrf.gov) prior to analysis using the KEGG Pathways (). A total of 43 different pathways corresponded to the altered miRNAs and several pathways, including Wnt signaling, mitogen-activated protein kinase (MAPK) signaling, transforming growth factor-β (TGF-β) signaling, longterm depression (LTD), calcium (Ca2+) signaling, and endocytosis pathways ( and Additional Figures ), are thought to be involved in neuronal function.

Figure 6

Classification of predicted target genes by morphine-responsive miRNAs according to the KEGG pathway analysis.

A total of 62 altered miRNAs were analyzed based on the target prediction algorithms of the bioinformatics database. KEGG pathways targeted by (A) upregulated and (B) downregulated miRNAs in the NAc. The vertical axis represents the pathway categories and the horizontal axis represents pathway enrichment calculated as –log [modified Fisher's exact P-value].

KEGG pathway analysis of the Wnt signaling pathway.

Red marks represent a putative target of a responsive microRNA (miRNA) in the nucleus accumbens (NAc) of mice that intravenously self-administered morphine.

Click here for additional data file.

KEGG pathway analysis of the MAPK signaling pathway.

Red marks represent a putative target of a responsive miRNA in the nucleus accumbens (NAc) of mice that intravenously self-administered morphine.

Click here for additional data file.

KEGG pathway analysis of the transforming growth factor-β (TGF-β) signaling pathway.

Red marks represent a putative target of a responsive miRNA in the nucleus accumbens (NAc) of mice that intravenously self-administered morphine.

Click here for additional data file.

KEGG pathway analysis of the long-term depression (LTD) pathway.

Red marks represent a putative target of a responsive miRNA in the nucleus accumbens (NAc) of mice that intravenously self-administered morphine.

Click here for additional data file.

KEGG pathway analysis of the Ca2+ signaling pathway.

Red marks represent a putative target of a responsive miRNA in the nucleus accumbens (NAc) of mice that intravenously self-administered morphine.

Click here for additional data file.

KEGG pathway analysis of the endocytosis pathway.

Red marks represent a putative target of a responsive miRNA in the nucleus accumbens (NAc) of mice that intravenously self-administered morphine.

Click here for additional data file.

Discussion

The present study is the first to examine the distinct expression patterns and functional roles of miRNAs in the NAc using a mouse model of morphine addiction. Two different expression patterns (upregulation and downregulation) of miRNA in mice receiving IVSA of morphine were compared to those in drug-naïve mice to determine the functional categorization of the miRNA targets associated with morphine addiction. Additionally, a bioinformatics analysis of the dynamic interactions within the miRNA target network was performed to characterize the roles of potential target mRNAs in the clinical symptoms of morphine addiction.

Current research indicates that the NAc plays a central role in the regulation of behavioral functions associated with depression, anxiety, and addiction due to its involvement in the brain reward circuit (Di Chiara and Imperato, 1988; Polter and Kauer, 2014). Approximately 95% of cells within the NAc are GABAergic neurons that possess dopamine D1- and D2-like receptors (Koo et al., 2014). NAc function is influenced by dopaminergic projections from the VTA (Fields and Margolis, 2015) and numerous studies have indicated that the high level of MOR distribution in the NAc might be related to the behavioral properties of opioid addiction, including reward and withdrawal symptoms (Dennis et al., 2016; Han et al., 2010). These behavioral consequences are reported to result from neuronal dysfunction induced by chronic exposure to opioids (Cunha-Oliveira et al., 2008; Ferrini et al., 2013). Therefore, identifying changes in NAc signaling pathways may be essential for a complete understanding of the mechanisms underlying the development of addictive behaviors.

According to the KEGG pathway analyses, 15 miRNAs and 45 of their putative targets appear to regulate the Wnt, MAPK, TGF-β, and neurotrophin signaling pathways, which are involved in diverse neuronal functions, such as neurogenesis, synaptic plasticity, and neuroinflammation (Fakira et al., 2014; Sanna et al., 2014; Zhang et al., 2016). Additionally, some of the miRNAs had multiple targets in each signaling pathway. In the Wnt signaling pathway, mmu-miR-202-5p regulated Fbxw11 and TBL1xr1; mmu-669d-2-3p regulated ROCK2, SMAD2, FZD7, and FZD10; and mmu-miR-32-5p regulated DKK2, MAPK8, and FZD10. In the MAPK signaling pathway, mmu-669d-2-3p regulated CHUK, RASA2, PTPRR, DUSP10, and CACNA1G; and mmu-miR-32-5p regulated ELK4, MAP2K4, and MAPK8. In the TGF-β signaling pathway, mmu-miR-669d-2-3p and mmu-miR-669-3p shared several target genes, including ROCK2, NODAL, and SMAD2, and mmu-miR-6416-3p regulated SMAD1 and LEFTY1. In the neurotrophin signaling pathway, mmu-miR-496b regulated IRS2 and MAP2K1 and mmu-miR-32-5p regulated PIK3CB and MAPK8.

Previous studies have suggested that there is a relationship between acute/chronic morphine treatment and synaptic plasticity in terms of Ca2+-dependent signaling in glutamatergic transmission (Chartoff and Connery, 2014) and that the NAc exhibits different types of intracellular signaling during the acute and chronic phases of morphine exposure (Nestler, 2004; Chartoff and Connery, 2014). During the acute phase of exposure, morphine inhibits presynaptic voltage-gated Ca2+ currents and upregulates protein kinase C (PKC)-dependent signaling (Martin et al., 1997), whereas morphine increases α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl-D-aspartate (NMDA)-mediated synaptic transmission during the chronic phase of exposure (Martin et al., 1997; Xu et al., 2012; Chartoff and Connery, 2014). In the present study, the KEGG analysis of putative targets for the altered miRNAs revealed that the regulatory pathway for Ca2+-dependent signaling was involved in the mouse model of morphine addiction.

It is known that individuals addicted to psychostimulants suffer from a variety of emotional states, such as depression or anxiety, which are caused by disruptions in LTD and long-term potentiation (LTP) (Nestler, 2001; Kasanetz et al., 2010). The KEGG pathway analysis in the present study identified a significant upregulation of mmu-miR-202-5p, which potentially regulates the LTD-associated genes PLC-β and KRAs. This is consistent with previous findings showing that morphine exposure modulates LTD in the NAc via mGLUR2/3 abolition during the withdrawal phase (Robbe et al., 2002) and that GABAergic synapses on VTA dopaminergic neurons are capable of inducing LTD while morphine stimulation significantly reduces LTD in the VTA (Dacher and Nugent, 2011).

Morphine is known for its rapid induction of tolerance, which limits its clinical use for anti-nociception (Gonzalez et al., 1997). Recent studies have suggested that the failure of morphine to promote MOR endocytosis can cause MOR desensitization, which, in turn, induces morphine tolerance and dependence (Kim et al., 2008; Berger and Whistler, 2010). The KEGG analysis of the putative targets for responsive miRNAs in the NAc after morphine self-administration in the present study identified 25 mRNAs that regulated endocytosis: FGFR2, EGFR, RET, CHMP5, TSG101, ASAP2, ARF6, EEA1, LDLRAP1, CHMP2B, ACVR1B, RAB11FIP5, DAB2, SH3GLB1, LOC547349, RAB22A, RAB11B, GRK6, RAB11A, HGS, PARD6G, EHD1, IQSEC1, EHD3, and RNF41.

To the best of our knowledge, this is the first study to perform a bioinformatics analysis of the networks between miRNAs and their putative targets in the NAc of mice intravenously self-administered morphine. Using this paradigm, a set of accumbal miRNAs that were altered by chronic morphine exposure were defined, and Wnt signaling, MAPK signaling, TGF-β signaling, LTD, calcium signaling, and endocytosis likely play important roles in morphine addiction. However, it is important to note that the networks among these miRNAs and their putative target genes were investigated using only a bioinformatics analysis; therefore, in vitro and in vivo experiments are needed to validate the links in these miRNA/mRNA networks and the functional effects of their changes on morphine addiction. Furthermore, it is likely necessary to determine the miRNA expression levels in specific brain regions and/or serum of humans/primates with morphine addiction.

Additional file:

Lever presses in the operant chamber during morphine intravenous self-administration.

KEGG pathway analysis of the Wnt signaling pathway.

KEGG pathway analysis of the MAPK signaling pathway.

KEGG pathway analysis of the TGF-β signaling pathway.

KEGG pathway analysis of the LTD pathway.

KEGG pathway analysis of the Ca

KEGG pathway analysis of the endocytosis pathway.