MiR-218 targets MeCP2 and inhibits heroin seeking behavior

MicroRNAs (miRNAs) are a class of evolutionarily conserved, 18–25 nucleotide non-coding sequences that post-transcriptionally regulate gene expression. Recent studies implicated their roles in the regulation of neuronal functions, such as learning, cognition and memory formation. Here we report that miR-218 inhibits heroin-induced behavioral plasticity. First, network propagation-based method was used to predict candidate miRNAs that played potential key roles in regulating drug addiction-related genes. Microarray screening was also carried out to identify miRNAs responding to chronic heroin administration in the nucleus accumbens (NAc). Among the collapsed miRNAs, top-ranked miR-218 was decreased after chronic exposure to heroin. Lentiviral overexpression of miR-218 in NAc could inhibit heroin-induced reinforcement in both conditioned place preference (CPP) test and heroin self-administration experiments. Luciferase activity assay indicated that miR-218 could regulate 3′ untranslated regions (3′ UTR) of multiple neuroplasticity-related genes and directly target methyl CpG binding protein 2 (Mecp2). Consistently, Mecp2308/y mice exhibited reduced heroin seeking behavior in CPP test. These data reveal a functional role of miR-218 and its target, MeCP2, in the regulation of heroin-induced behavioral plasticity.


Results
Prediction of addiction-related miRNAs using network propagation-based strategy. First, we tried to screen for miRNAs involved in drug addiction-related regulatory networks. A miRNA could regulate gene expression either by directly targeting drug addiction-related genes, or by targeting regulatory elements (e.g. transcription factors) whose impact may propagate across the whole regulatory network. Thus we utilized a network propagation-based model (NP-method) 19 to predict miRNAs, whose target genes may contribute to major alterations in addiction-related pathways. We used miRNA-target regulation information (TargetScan 7.1) and the transcription regulatory database (HTRIdb) to model network effects of the miRNA perturbation, normalized reliability score from Knowledgebase for Addiction-Related Genes (KARG) to assess the potential involvement of addiction-related genes. The correlation between network effects of the miRNA perturbation and gene ranking was evaluated (Fig. 1a), which revealed 40 addiction-related miRNA families significantly enriched in regulation of addiction-related process (Table S1), among which, miR-132/212, miR-9, miR-181 etc. were topranked and previously reported 17,[20][21][22] , indicating that NP-method is efficient in identifying miRNAs involved in addiction-related process.
MiR-218 is down-regulated in response to chronic heroin administration. Microarray screening was carried out to identify miRNAs responding to chronic heroin administration. As shown in Fig. 1b, chronic treatment of heroin (CH, 1 mg/kg heroin, twice daily intraperitoneal injections for 7 days) resulted in differential expression (≥ 25%) of 111 miRNAs in the NAc, as compared with chronic saline group (CS, equivalent volume of saline, twice daily intraperitoneal injections for 7 days). The results obtained using NP-method and microarray screening were collapsed, resulting in 16 miRNAs from 14 families (Fig. 1d), which are likely important regulators of heroin-induced reinforcement. Ranked    Among these 16 candidate miRNAs, miR-181a, a miRNA previously implicated in cocaine-induced plasticity 17 , was top-ranked. MiR-218, whose role in drug addiction has not been reported, was second highest scored, and microarray analysis indicated that miR-218 was down regulated in response to chronic heroin exposure (Fig. 1d). Quantitative real-time PCR also confirmed that chronic, but not acute heroin administration (AH, saline twice daily intraperitoneal injections for 7 days, then one injection of heroin, 1 mg/kg), significantly reduced the expression level of miR-218 in NAc (Fig. 2a). Moreover, the decrease of miR-218 level induced by chronic heroin exposure was restricted to NAc, since no significant change of miR-218 level was observed in mPFC or hippocampus (Fig. 2b).

MiR-218 regulates heroin reinforcement.
To explore the role of miR-218 in heroin-induced behavioral plasticity, we delivered recombinant lentivirus that express miR-218 (LV-miR-218) or scramble control sequence (LV-control) into the NAc region (Fig. 3a). The expression of virus-derived reporter GFP and its co-localization with neuronal marker NeuN were observed (Fig. 3b). Quantitative PCR analysis revealed that significant increase of miR-218 level could be detected 3 days after lentivirus injection and a 2-fold increase was observed 7 days after LV-miR-218 injection in NAc (Fig. 3c). To determine if miR-218 played a role in heroin-induced reinforcement, the effect of miR-218 overexpression in NAc was examined on CPP. One week after NAc infusion of LV-miR-218 or LV-control, rats were given repeated heroin injections paired with a contextually-distinct chamber in CPP apparatus, and their performance in the subsequent CPP test was scored. We found a significant effect of treatment × test phase interaction (F (1,17) = 11.80, p = 0.003). Moreover, Bonferroni's post hoc analysis revealed that rats with miR-218 overexpression in the NAc exhibited reduced preference to the heroin-paired chamber, compared with LV-control injected rats (Fig. 4a).
Next, we carried out heroin self-administration tests to further validate the role of miR-218 in heroin seeking behavior. Rats were trained to self-administer heroin for 8 consecutive days, then subjected to LV-miR-218 or LV-control injection. Consistent with CPP results, miR-218 overexpression significantly reduced heroin acquisition compared with the control group (Fig. 4b).

MiR-218 regulates drug addiction network and directly targets MeCP2. To infer biological roles
and network communities of miR-218, miRSystem database 23 was used for characterizing enriched functions and pathways of miR-218 targets. MiR-218 targets, co-predicted by at least two canonical databases (PicTar, TargetScan, DIANA, miRanda), were enriched in KEGG pathways highly relevant to neuronal functions, such as gap junction and long-term depression (Fig. 5a), suggesting its potential involvement in regulation of neuronal plasticity. We selected a few predicted targets of miR-218, which are indicated by both KARG database and recent literature to contribute to drug-induced neuroplasticity, and tested the validity of the miRNA-gene regulatory relationship. Indeed, Gabrb3, and Mecp2, which play critical roles in drug induced synaptic plasticity 22,24-26 , Nrxn1, Gng3, and Ube3a, which are involved in axon guidance and dendritic morphogenesis [27][28][29] , were increased in NAc after chronic heroin administration (Fig. 5b). Using luciferase reporter system, we first cloned each predicted target site into the 3′ UTR of Renilla luciferase and co-transfected the constructs with miR-218 mimic or scramble control in HEK293T cells. Reduced luciferase activity indicated that these genes were the directly regulated by miR-218 (Fig. S1). Among the targets identified, we focused on MeCP2 for its crucial role in psychostimulus-induced plasticity 22 and broad transcription regulatory activity 30 . As shown in Fig. 6a, the predicted miR-218 target sequences of Mecp2 were highly conserved among mammals. Target sequence of Mecp2 3′ UTR and a seed region-mutated sequence were subjected to luciferase assay. We found that the expression of luciferase with mutant Mecp2 3′ UTR was not affected by miR-218 mimic transfection, in contrast to significantly attenuated activity of luciferase with wild type Mecp2 3′ UTR (Fig. 6b). Furthermore, overexpression of miR-218  Rats were subjected to LV-control or LV-miR-218 injection. Seven days later, they were subjected to pre-test, three conditioning sessions and CPP test. Heroin induced CPP is decreased in LV-miR-218 group (LV-control, n = 10, LV-miR-218, n = 9, **P < 0.01 vs. LV control during test session). (b) Lever presses of LV-control and LV-miR-218 rats in heroin self-administration. Rats were subjected to heroin selfadministration procedure for eight days and then randomly grouped to receive NAc LV-control or LV-miR-218 injection. Three days later, daily SA sessions were resumed. LV-miR-218 group showed decreased heroin consumption after lentivirus injection (LV-control, n = 6, LV-miR-218, n = 9, Two-way RM ANOVA. *P < 0.05 vs. LV-control on the same day). Data are shown as mean ± s.e.m. in C6 cells downregulated Mecp2 expression level (Fig. 6c), which further confirmed that Mecp2 is a target of miR-218. Coincident with miR-218 overexpression result, heroin CPP was significantly blocked in Mecp2 308/y mice (Fig. 6d). These data support the notion that miR-218 directly targets Mecp2 in a sequence-specific manner and thus inhibits heroin seeking behavior.

Discussion
MiRNA functions are executed by the multiprotein RNA-induced silencing complex containing Argonaute (Ago) 31 . The expression level of Ago2 is decreased in the VTA of morphine-dependent rats 32 . Ablation of Ago2 in dopamine 2 receptor-expressing neurons attenuates the reinforcing effects of cocaine and cocaine induced CPP, implicating miRNA may contribute to addictive behaviors 16 . Hollander et al. reported cocaine SA increase miR-212 in striatum and increasing miR-212 in this region decrease cocaine reward 18 . Chronic cocaine exposure also regulates miR-124 and miR-181a expression in brain, and overexpression of miR-124 in NAc reduced cocaine CPP while miR-181a has the opposite effect 33 . He et al. reported let-7 targets the 3′ UTR of MOR, and chronic morphine treatment increased let-7 expression in brain, while let-7 inhibition causes increased levels of MOR and partially attenuated opioid antinociceptive tolerance 34 . MiR-382 modulated the expression of Δ FosB which had been linked directly to several addiction-related behaviors, and overexpression of miR-382 in NAc decreased voluntary intake and preference for alcohol in rats 35 . Our research found that miR-218 in NAc decreased by heroin administration and overexpression of miR-218 inhibited heroin seeking behavior, providing a direct evidence of miR-218 involved in heroin induced behavioral plasticity.
In the present study, we found that miR-218 was downregulated by chronic heroin use in NAc. Consistently, bioinformatic analysis indicated the predicted targets of miR-218 was enriched in addiction-related genes and involved in neuroplasticity. MiR-218 is encoded by an intron of the Slit gene and inhibits the expression of Robo1 which play important roles in axonal growth 36,37 . Recent studies indicated that miR-218 plays a crucial role in motor neuron differentiation and loss of miR-218 cause systemic neuromuscular failure 38,39 . Recent large-scale miRNA expression profiling showed that miR-218 exhibits ubiquitous enrichment in hippocampus, cortex and cerebellum, but exhibits lower expression in olfactory bulb and brain stem 40 . Hippocampal cell-type-based analysis of miRNAs revealed that miR-218 is specifically enriched in neurons, compared with microglia, astrocytes and oligodendrocytes 41 . And more specifically, it is enriched in parvalbumin (PV)-positive, but not neuropeptide somatostatin expressing GABAergic interneurons. And its expression in CaMKIIα positive neurons exhibits three-fold higher expression than Gad2-expressing neurons 42 . Together with the predicted pathway of its target genes, we postulate that miR-218 is essential to the regulation of neuronal plasticity.
MeCP2 is a transcription factor binding to methylated cytosine residues in DNA and recruits histone deacetylases and other transcriptional repressors to silence target genes 43   a reduction in synaptic response whereas neurons with doubling of Mecp2 exhibit a two-fold enhancement in synapse number 44 . Mutations in MeCP2 results in nonsyndromic mental retardation, mild learning disability, and classic autism [45][46][47] . Recent studies reveal the role of MeCP2 in psychostimulant induced behavioral plasticity. Extended cocaine access increased Mecp2 expression in dorsal striatum and striatal Mecp2 knockdown decrease cocaine intake 22 . Coincidently, we found reduced heroin conditioned preference in Mecp2 308/y mice. However, Deng et al. found that Mecp2 308/y mutant mice showed enhanced locomotor response to acute amphetamine administration and impaired CPP, but lentivirus mediated knockdown or overexpression of MeCP2 in NAc enhanced or inhibited amphetamine-induced locomotion and CPP respectively 48 . Later, the same group found Mecp2 Ser421Ala knock-in mice display both a reduced threshold for amphetamine sensitization and enhanced behavioral sensitivity to the reinforcing properties in cocaine self-administration, indicating that MeCP2 in NAc limited behavioral plasticity upon repeated psychostimulant exposure 49 . Functionally, early studies found that MeCP2 interacts specifically with methylated cytosine in the CG context (mCG). However, recent evidence suggest that MeCP2 regulates expression of genes in a spatial-and temporal-distinct manner and binds to non-CG methylation 30,50,51 . These result implicated that the role of MeCP2 played in drug addiction seems much more complicated, and more detail understanding the molecular mechanism of MeCP2 in drug induced plasticity still need to be experimental verified. Taken together, our results reveal that miR-218 is involved in heroin induced behavioral plasticity possibly by regulating Mecp2. Overexpression of miR-218 in NAc decreased heroin conditioned preference and heroin self-administration. MiR-218 is likely to contribute to heroin related modifications, both transcriptionally and epigenetically. Our results provide a promising and novel approach to the treatment of heroin addiction. However, the role of miR-218 in other drug induced plasticity and the downstream signaling pathways regulated by miR-218 in addiction are remained to be elucidated.

Methods
Subjects. Male Sprague-Dawley rats were purchased from the Shanghai SLAC Laboratory Animal Co. Ltd. and housed with a 12 h reverse dark/light cycle at 23 °C. All subjects were allowed free access to food and water. Rats used for experiments were 8-10 weeks old. Mecp2 308/y mice and littermates were maintained by mating heterozygotes with C57BL/6 J. All animal treatments were in strict accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by Animal Care and Use Committee of Shanghai Medical College of Fudan University.
Lentivirus Construction and Packaging. The lentivirus construct and packaging were performed as we described previously 6 . Briefly, human U6 promoter and small hairpin RNA that overexpress miR-218 (LV-miR-218, TTGTGCTTGATCTAACCATGT) or control sequences (LV-control, GCAGTTATCACGTCTATGTTT) were cloned into FG12 plasmid and co-transfected into HEK293T cells with pMDLg/RRE, pRSV/rev and pHCMV-G for lentivirus packaging. Forty-eight hours after transfection, the culture medium was collected and concentrated by ultracentrifugation, titrated, aliquoted, and stored at − 80 °C, the final virus titer was estimated to be 1 × 10 8 IU/ml.

Intravenous heroin self-administration.
To facilitate subsequent heroin self-administration, rats were trained to press an active lever for food pellets until they self-administered 100 pellets for 3 consecutive days on fixed-ratio 1 program. Then the rats were anesthetized, and a silastic catheter (0.04 cm interior diameter) was inserted 3 cm into the jugular vein, and the other end of the catheter was attached to a stainless steel pedestal and mounted on the skull. Bilateral 26-gauge guide cannulae were implanted in NAc (1.7 mm A/P, ± 1.6 mm M/L, and − 5 mm D/V). After surgery, the rats were allowed to recover for 7 days and the catheters were flushed daily with saline supplemented with heparin (30 IU/ml) and gentamycin (5 mg/ml). Next the rats were allowed to self-administer heroin (0.05 mg/kg/injection, FR1 program) during a daily 4 h session. After the 8th session, the rats were anesthetized and subjected to lentivirus infusion with needle (Plastics One, Inc.) that extend 1.5 mm beyond the guide cannulae. The self-administration procedure was resumed 3 days after lentivirus injection and tested on FR1 schedule for six consecutive days.
Conditioned place preference. The apparatus was three-chambered (Med Associates), with two large side chambers with distinct floors and decorations and one small center chamber. In pretest, the rat was first confined in the center chamber for adaption for 5 min, and then allowed to freely explore the chambers for 15 min. Unbiased rats were subjected to lentiviral injection into NAc (1.7 mm A/P, ± 1.6 mm M/L, and − 6.5 mm D/V). The rats were allowed to recover for 7 days. Then the rats were alternatively confined in each large chamber for 30 min after injection of heroin (1 mg/kg, i.p., days 1, 3, 5) or saline (days 2, 4, 6). On the test session (day 8), the rats were adapted in center chamber for 5 min and allowed to explore all chambers unrestrictedly for 15 min. The time which the rat spent in each large chamber was recorded and CPP score was defined as the time spent in the heroin-paired chamber minus the time spent in the saline-paired chamber during the CPP test. The CPP pretest, conditioning and test procedure of Mecp2 308/y and Mecp2 +/y mice followed the same procedure.
MiRNA microarray and data analysis. Total RNA are harvested using TRIzol (Invitrogen) and RNeasy mini kit (QIAGEN) according to manufacturer's instructions. The RNA samples are labeled using the miR-CURY ™ Hy3 ™ /Hy5 ™ Power labeling kit and hybridized on the miRCURY ™ LNA Array (v.11.0). Scanning was performed with the Axon GenePix 4000B microarray scanner. GenePix pro V6.0 is used to read the raw intensity of the image. Median normalized signal of each miRNA was used to represent expression level. MiRNAs that exhibit 1.25-fold up-regulation or 0.75-fold down-regulation was chosen for further analysis. Network propagation based model was generated as described 19 . Briefly, miRNA families that are highly conserved (conservation score ≥ 2) were chosen from TargetScan Human 7.1, and their predicted targets were used.