MicroRNA let-7d is a target of cannabinoid CB1 receptor and controls cannabinoid signaling

Neuropharmacology 108 (2016) 345e352

Contents lists available at ScienceDirect

Neuropharmacology journal homepage: www.elsevier.com/locate/neuropharm

MicroRNA let-7d is a target of cannabinoid CB1 receptor and controls cannabinoid signaling € rner c, 1, Laura Martín-Go  mez b, 1, Ada Jime nez-Gonza lez d, Anna Chiarlone a, b, Christine Bo zquez a, b, n García-Concejo d, María L. García-Bermejo b, Mar Lorente a, Cristina Bla Adria a, b a c c € llt , , Amador de Haro , Elisa Martella , Volker Ho Elena García-Taboada n a, b, * Raquel Rodríguez d, Ismael Galve-Roperh a, b, Jürgen Kraus c, Manuel Guzma a n Biom n Neuroquímica Centro de Investigacio edica en Red sobre Enfermedades Neurodegenerativas (CIBERNED) and Instituto Universitario de Investigacio (IUIN), Department of Biochemistry and Molecular Biology I, Complutense University, 28040 Madrid, Spain b n y Cajal de Investigacio n Sanitaria (IRYCIS), 28034 Madrid, Spain Instituto Ramo c Department of Pharmacology and Toxicology, University of Magdeburg, 39106 Magdeburg, Germany d Instituto de Investigaciones Biom edicas de Salamanca (IBSAL), 37007 Salamanca, Spain

a r t i c l e i n f o

a b s t r a c t

Article history: Received 6 October 2015 Received in revised form 19 April 2016 Accepted 10 May 2016 Available online 11 May 2016

Cannabinoid CB1 receptor, the molecular target of endocannabinoids and cannabis active components, is one of the most abundant metabotropic receptors in the brain. Cannabis is widely used for both recreational and medicinal purposes. Despite the ever-growing fundamental roles of microRNAs in the brain, the possible molecular connections between the CB1 receptor and microRNAs are surprisingly unknown. Here, by using reporter gene constructs that express interaction sequences for microRNAs in human SH-SY5Y neuroblastoma cells, we show that CB1 receptor activation enhances the expression of several microRNAs, including let-7d. This was confirmed by measuring hsa-let-7d expression levels. Accordingly, knocking-down CB1 receptor in zebrafish reduced dre-let-7d levels, and knocking-out CB1 receptor in mice decreased mmu-let-7d levels in the cortex, striatum and hippocampus. Conversely, knocking-down let-7d increased CB1 receptor mRNA expression in zebrafish, SH-SY5Y cells and primary striatal neurons. Likewise, in primary striatal neurons chronically exposed to a cannabinoid or opioid agonist, a let-7d-inhibiting sequence facilitated not only cannabinoid or opioid signaling but also cannabinoid/opioid cross-signaling. Taken together, these findings provide the first evidence for a bidirectional link between the CB1 receptor and a microRNA, namely let-7d, and thus unveil a new player in the complex process of cannabinoid action. © 2016 Elsevier Ltd. All rights reserved.

Keywords: Cannabinoid CB1 receptor microRNA Cell signaling Synaptic transmission

1. Introduction Cannabinoid CB1 receptor is one of the most abundant metabotropic receptors in the central nervous system. It is particularly expressed in discrete brain areas involved in the control of learning and memory (cortex, hippocampus), motor behavior (basal ganglia, cerebellum), emotions (amygdala), and autonomic and endocrine functions (hypothalamus, pons, medulla), therefore participating in

* Corresponding author. Department of Biochemistry and Molecular Biology I,  Antonio Novais 12, 28040 Madrid, School of Biology, Complutense University, c/Jose Spain. n). E-mail address: [email protected] (M. Guzma 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.neuropharm.2016.05.007 0028-3908/© 2016 Elsevier Ltd. All rights reserved.

the control of a large plethora of neurobiological processes (Piomelli, 2003; Katona and Freund, 2008). The CB1 receptor is engaged by a family of neuronal retrograde messengers, the endocannabinoids, and is also targeted by D9-tetrahydrocannabinol (THC), the major psychoactive component of Cannabis sativa (marijuana). Endocannabinoid signaling serves as a major feedback mechanism to prevent excessive neuronal activity, thereby tuning the functionality and plasticity of many synapses (Piomelli, 2003; Katona and Freund, 2008). Cannabis, together with its selectively cultivated subspecies (e.g., skunk) and its synthetic derivatives (“designer drugs”), are the most common illicit drugs of abuse currently used for recreational purposes (Alp ar et al., 2016; World Drug Report, 2015). Cannabis preparations have also been used in medicine for centuries, and

346

A. Chiarlone et al. / Neuropharmacology 108 (2016) 345e352

nowadays there is a renaissance in the study and application of their therapeutic effects. In this context, THC and other cannabinoids are already approved as anti-emetic, anti-cachexic, analgesic and anti-spastic compounds (Pertwee, 2012; Syed et al., 2014; Notcutt, 2015). However, tolerance to the effects of cannabinoids develops rapidly, which progressively decreases the efficacy of cannabinoid-based medicines, especially upon use over long time periods. Cannabinoid tolerance is mainly attributed to pharmacodynamic changes, such as a decrease in the number of total and functionally-coupled CB1 receptors on the cell surface, with a minor pharmacokinetic component caused by enhanced cannabinoid biotransformation and excretion (Adams and Martin, 1996; Grotenhermen, 2003; Panlilio et al., 2015). The CB1 receptor signals via Gi/o proteins, and, in fact, alterations in both the cAMP/ protein kinase A (PKA) pathway (Rubino et al., 2000; Lee et al., 2003) and the extracellular signal-regulated kinase (ERK) pathway (Rubino et al., 2005; Tonini et al., 2006), two canonical Gi/o protein-coupled pathways, have been implicated in cannabinoid tolerance. However, the lack of knowledge on the precise downstream effectors of these signals -as well as on the possible role of other molecular cues- in CB1 receptor desensitization hampers the biological understanding and potential therapeutic management of cannabinoid tolerance. Since their discovery, a very large number of microRNAs (miRNAs) and other small non-coding RNAs have been identified from humans to viruses, thus allowing a new level of transcriptional and post-transcriptional regulation of gene expression. The brain is highly enriched in miRNAs, and these molecules play ever-growing roles in fundamental neurobiological processes such as neuronal development, plasticity and survival (Smalheiser, 2014; Sun and Shi, 2015). Moreover, alterations in the expression of various miRNAs have been shown to contribute importantly to different neurological diseases (Sun and Shi, 2015; Issler and Chen, 2015). However, in spite of the ample expression and key physiopathological functions of the CB1 receptor in the central nervous system, it is surprising that the possible bidirectional connection between the CB1 receptor and miRNAs has not been studied so far. In particular, the miRNA let-7d has a widespread action in the brain during health and disease, including the control of drug addiction processes (He et al., 2010; Barbierato et al., 2015; Rehfeld et al., 2015). Hence, this background prompted us to test (i) whether the CB1 receptor controls let-7d expression, and (ii) whether let-7d controls the CB1 receptor. 2. Material and methods 2.1. Mice Mice knockout for the cannabinoid CB1 receptor gene (Cnr1
/
) and their wild-type littermates (Marsicano et al., 2002) were zquez et al., 2011). Only male animals maintained as described (Bla were used for the assays. Animal handling procedures were approved by Complutense University Animal Research Committee (Madrid, Spain) in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC). All efforts were made to minimize animal suffering, to reduce the number of animals used, and to utilize alternatives to in vivo techniques. 2.2. Zebrafish Experiments were performed using the wild-type AB zebrafish line, which was maintained according to the standard protocols of dicas de Salamanca (Spain) the Instituto de Investigaciones Biome in accordance with the directives of the European Commission. Animals were maintained at a constant temperature of 28  C in a

14-h light cycle and fed 3 times a day. Embryos were obtained by natural mating, cultured in E3 medium, and staged according to hpf (Kimmel et al., 1995). The morpholinos targeting dre-cnr1 (50 GAACAGCATGGTCAGAGATGCTCTA-30 ), dre-let-7d (50 -AAACCATACAACCAACTACCTCAGC-30 ) and dre-miR-124 (50 -TGGCATT0 CACCGCGTGCCTTAA-3 ) were purchased from GeneTools. 2.3. Cell culture, cAMP determination and Western blot SH-SY5Y and HEK293 cells were cultured at 37  C under 5% CO2 in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum, 1 mM sodium pyruvate, 2 mM L-glutamine and 1 mM non-essential amino acids. Primary rat striatal neurons were €rner et al., 2007). For miRNA obtained and cultured as described (Bo expression experiments, cells were transferred to 1%-FBS medium and, after 24 h, they were pre-incubated (when indicated) with 0.25 mM SR141716, (rimonabant), 2 mM SR144528, 2 mM U0126, 25 mM forskolin or vehicle (0.1% dimethylsulfoxide). Thirty min later, 0.5 mM methanandamide (Sigma), 1 mM morphine (Sigma) or vehicle (1% ethanol) was added to the cultures for the times indicated. For cAMP determinations, cells were incubated for 15 min with 25 mM forskolin (alone or in combination with 0.5 mM methanandamide or 1 mM morphine), and then lysed with 50 mM HCl for 30 min on ice. A competitive cAMP ELISA was used (Horton € rner et al., 2008). Western blots were conducted et al., 1992; Bo with antibodies against phosphorylated ERK (1:1000; Cell Signaling) and b-actin (1:4000, Santa Cruz Biotechnology) following standard procedures. 2.4. Cell transfection and reporter assays The ptkCAT3 construct was generated by cloning the thymidine kinase promoter of pBLCAT2 upstream the chloramphenicol acetyltransferase (CAT) reporter gene of pCAT3 (Promega). The recognition sequences for various miRNAs were subsequently inserted into the single XbaI site downstream the CAT gene of ptkCAT3. The inserted sequences were: let-7a, 50 -CTAGAACTATACAACCTACTACCTCA-30 ; let-7d, 50 -CTAGGACTATGCAACCTAC0 0 TACCTCT-3 ; miR-23b, 5 -CTAGGGTAATCCCTGGCAATGTGAT-30 , miR-98, 50 -CTAGAACAATACAACTTACTACCTCA-30 , miR-124, 50 CTAGAGGCATTCACCGCGTGCCTTA-30 ; scrambled, 50 -CTAGA0 GACGATTCACCCGACCACATCG-3 . The constructs were transfected €rner et al., 2008) and CAT into SH-SY5Y cells as described (Bo expression was measured by ELISA (Roche). HEK293 cells were transfected using Lipofectamine 2000 (Invitrogen) with the reporter plasmid pmirGlo (Promega) and a let-7d mimic (Life Technologies). Luciferase was assayed with the Dual-Glo Luciferase Assay System (Promega). The let-7d inhibitor (a chemically synthesized, single-stranded, modified RNA sequence that specifically hybridizes to let-7d and so blocks its function) was purchased from Active Motif. 2.5. Quantitative PCR zquez et al., 2011) and retroRNA was isolated as described (Bla transcribed using miRCURY LNA™ Universal RT microRNA PCR, polyadenylation and cDNA synthesis kit (Exiqon). cDNA was used as template for quantitative PCR reactions with SYBR Green (11066420, SYBR Green I Master, Roche Diagnostics). Specific LNA PCR primer set UniRT (Exiqon) was used for each miRNA. 5S rRNA was used for data normalization. Quantification crossing points (Cq) were determined by Light Cycler 480 Software 1.5 (Roche) and miRNAs expression was calculated using the 2
DDCq formula (second derivative method) (Aguado-Fraile et al., 2012).

A. Chiarlone et al. / Neuropharmacology 108 (2016) 345e352

347

2.6. Statistical analysis All variables were first tested for normality (Kolmogorov-Smirnov’s test) and homoscedasticity (Levene’s test). When implied variables satisfied these conditions, two-way ANOVA was used to test the significance of genetic manipulations/incubation times/ pharmacological treatments as well as their possible interactions. One-way ANOVA with post-hoc Fisher’s LSD test was then used to assess differences between groups. Significance level was set at 0.05 in all cases. Data are shown as mean ± SEM. All analyses were carried out with GraphPad Prism 6 (GraphPad Software). 3. Results 3.1. Up-regulation of let-7d expression by CB1 receptor activation We first tested whether CB1 receptor engagement affects let-7d expression in human SH-SY5Y neuroblastoma cells. These cells constitute a widely used experimental model to study neuronal cell function and degeneration. Specifically, its endogenous cannabinoid system -including the CB1 receptor- has been characterized in detail (Pasquariello et al., 2009). The interaction sequence for hsalet-7d and other hsa-miRNAs was cloned downstream the CAT gene, so reporter activity would be expected to decrease upon binding of the corresponding endogenous miRNA to that sequence. We exposed the transfected cells to methanandamide, a CB1 receptorselective synthetic agonist, and used in parallel morphine as a control compound because (i) it signals via the m-opioid receptor, which, like the CB1 receptor, is a Gi/o protein-coupled receptor, and (ii) it has been previously shown to enhance let-7d expression in SH-SY5Y cells (He et al., 2010). Two-way ANOVA of CAT activity showed a significant overall effect of both genetic manipulations (F(6,96) ¼ 207.7, P < 0.0001; Fig. 1A) and pharmacological treatments (F(2,96) ¼ 559.4, P < 0.0001; Fig. 1A), as well as a significant interaction between these two main factors (F(12,96) ¼ 51.01, P < 0.0001; Fig. 1A). Methanandamide strongly reduced CAT activity of the hsa-let-7d-interacting construct when compared to untreated transfectants (P < 0.0001; Fig. 1A). This effect was not observed in cells transfected with the empty construct or in cells transfected with the construct containing a scrambled sequence, to which no endogenous miRNA is predicted to bind (Fig. 1A). Using a similar strategy, we found that methanandamide also heightened the expression of other miRNAs that are functionally relevant in the brain such as miR-124, let-7a, miR-23b and miR-98 (P < 0.0001 in all cases; Fig. 1A). Morphine produced similar effects than methanandamide (P < 0.0001 in all cases; Fig. 1A). To provide further support to the connection between the CB1 receptor and let-7d we directly measured the expression levels of hsa-let-7d by quantitative PCR, and found that they were increased upon methanandamide or morphine challenge in a timedependent manner (F(4,46) ¼ 4.081, P ¼ 0.0065; Fig. 1B). hsamiR-124 was used as a positive control (F(4,38) ¼ 4.070, P ¼ 0.0250; Fig. 1B). In addition, we used miR-93 and miR-126 as negative controls because, according to previous studies, these two miRNAs are not as plastic as let-7d and miR-124 in the brain (Barbierato Fig. 1. CB1 receptor activation up-regulates let-7d expression in SH-SY5Y cells. (A) SH-SY5Y cells were transfected with the reporter plasmid ptkCAT3 or with ptkCAT3 additionally containing binding sites for several miRNAs (or a scrambled sequence as control), then incubated for 72 h with vehicle, 0.5 mM methanandamide (MAEA) or 1 mM morphine, and CAT activity was measured (n ¼ 8 empty, 7 scrambled, 8 let-7d, 6 miR-124, 4 miR-let-7a, 4 miR-23b, 4 miR-98). (B) SH-SY5Y cells were incubated with vehicle, 0.5 mM MAEA or 1 mM morphine, and the expression levels of hsa-let-7d, hsamiR124, hsa-miR-93 and hsa-miR-126 were measured by quantitative PCR. 5S rRNA was used for data normalization (n ¼ 5 for each time point). (C) SH-SY5Y cells were preincubated for 30 min with vehicle, 0.25 mM SR141716 (SR1) or 2 mM SR144528 (SR2),

then incubated for 2 h with vehicle or 0.5 mM MAEA, and the expression levels of hsalet-7d and hsa-miR124 were measured by quantitative PCR. 5S rRNA was used for data normalization (n ¼ 3e4 for each condition). (D) SH-SY5Y cells were pre-incubated for 30 min with vehicle, 2 mM U0126 or 25 mM forskolin (FSK), then incubated for 2 h with vehicle, 0.5 mM MAEA or 1 mM morphine, and the expression levels of hsa-let-7d and hsa-miR124 were measured by quantitative PCR. 5S rRNA was used for data normalization (n ¼ 6 vehicle, 4 U0126, 4 FSK). Data were analyzed by ANOVA with post-hoc Fisher’s LSD test. *P < 0.05, **P < 0.01 from the corresponding vehicle-treated cells; ## P < 0.01 from the corresponding methanandamide/vehicle-treated cells (a.u. ¼ arbitrary units).

A. Chiarlone et al. / Neuropharmacology 108 (2016) 345e352

3.2. Down-regulation of let-7d expression by CB1 receptor inactivation in vivo We next evaluated whether the CB1 receptor controls let-7d expression in vivo. First, we used zebrafish (Danio rerio) as a model organism as its endogenous cannabinoid system, including the CB1 receptor, resembles quite closely that expressed by mammals (Lam et al., 2006; McPartland et al., 2007; Rodríguez-Martín et al., 2007). Thus, zebrafish embryos at 24 h post fertilization (hpf), in which the central nervous system is completely developed, were microinjected with a dre-cnr1-targeted or a scrambled morpholino. CB1 receptor knockdown reduced the levels of dre-let-7d, with a similar effect being exerted on dre-miR-124 (F(3,8) ¼ 23.42, P ¼ 0.0003; Fig. 2A). Likewise, we examined the brains of CB1 receptor knockout mice and found a lower expression of mmu-let-7d (F(7,26) ¼ 3.247, P ¼ 0.0130; Fig. 2B) and mmu-miR-124 (F(7,26) ¼ 2.910, P ¼ 0.0247; Fig. 2B) in the cortex, striatum and hippocampus, but not in the cerebellum, compared to wild-type littermates. 3.3. Up-regulation of CB1 receptor mRNA expression by let-7d inactivation We next asked whether let-7d regulates CB1 receptor mRNA

A

Zebrafish Scrambled

dre-miRNA (a.u.)

1.5

dre-cnr1 morphants

1.0

****

** **

dre-let-7d

dre-miR-124

0.5

0

B

Mouse Cnr1 +/+ 1.5

mmu-let-7d (a.u.)

et al., 2015; Brites and Fernandes, 2015; Qu et al., 2015 Wan et al., 2015) but are rather associated to the physiopathology of other tissues such as kidney and vascular endothelium (Aguado-Fraile et al., 2015; Lorenzen, 2015; Rudnicki et al., 2015). Thus, neither methanandamide nor morphine exposure changed the expression levels of hsa-miR-93 (F(8,97) ¼ 0.1882, P ¼ 0.9920) or hsa-miR-126 (F(8,95) ¼ 0.1639, P ¼ 0.9950) in SH-SY5Y cells (Fig. 1B), thus providing support to a specific up-regulation of let-7d and miR-124. We next aimed at further substantiating the involvement of the CB1 receptor in methanandamide action. First, by using quantitative PCR, we found that our SH-SY5Y cells expressed CB1 receptor mRNA (Cq ¼ 29). Additionally, and in line with a previous study (Pasquariello et al., 2009), these cells also expressed CB2 receptor mRNA, although at a lower extent (Cq ¼ 32). The m-opioid receptor mRNA was also amplified as a control (Cq ¼ 32). Next, we used cannabinoid receptor type-selective antagonists to assess the participation of the two receptors in the observed effects. Thus, the CB1 receptor-selective antagonist SR141716 (rimonabant) abrogated the methanandamide-induced elevation of hsa-let-7d expression (F(3,71) ¼ 9.142, P < 0.0001) and hsa-miR-124 expression (F(3,58) ¼ 11.63, P < 0.0001) (Fig. 1C). A preventive effect ethough milder than that of SR141716- of the CB2 receptorselective antagonist SR144528 was also evident (F(3,64) ¼ 9.362, P < 0.0001 for hsa-let-7d expression; F(3,61) ¼ 9.133, P < 0.0001 for hsa-miR-124 expression) (Fig. 1C). These data support that the CB1 receptor is functional in mediating cannabinoid actions on miRNA expression in SH-SY5Y cells, and that the closely-related CB2 receptor also seems to be active in this process. As the Gi/o protein-coupled CB1 receptor is well known to inhibit the cAMP/PKA pathway and activate the ERK pathway (Pertwee et al., 2010), we also tested the possible involvement of these two canonical routes in the up-regulation of let-7d. Incubation of cells with forskolin (that activates adenylyl cyclase, and so raises cAMP) or U0126 (that inhibits MEK, and so blocks ERK activation) prevented the methanandamide-induced elevation of hsa-let-7d levels (F(4,40) ¼ 5.363, P ¼ 0.0086; Fig. 1D). A similar mechanism was evident for the up-regulation of hsa-miR-124 expression upon CB1 receptor engagement (F(4,49) ¼ 2.633, P ¼ 0.0453; Fig. 1D). Likewise, the induction of these two miRNAs by morphine was blocked by forskolin and U0126 (Fig. 1D).

Cnr1 -/-

1.0

* *

0.5

*

0

Cx

Str

Hip

Cbl

1.5

mmu-miR-124 (a.u.)

348

1.0

*

* *

0.5

0

Cx

Str

Hip

Cbl

Fig. 2. CB1 receptor inactivation lowers let-7d expression in vivo. (A) Zebrafish embryos at 24 hpf were microinjected with a dre-cnr1-targeted morpholino (or a scrambled morpholino as control) and the expression levels of dre-let-7d and dremiR124 were measured (n ¼ 3 pools of 50 embryos for each condition). (B) The expression levels of hsa-let-7d and hsa-miR124 were measured in the cortex (Cx), striatum (Str), hippocampus (Hip) and cerebellum (Cbl) of 10 week-old male Cnr1
/
mice and wild-type littermates (n ¼ 5 for each genotype and each brain region). Data were analyzed by ANOVA with post-hoc Fisher’s LSD test. *P < 0.05, **P < 0.01 from the corresponding control animals (a.u. ¼ arbitrary units).

expression in vivo. We knocked-down dre-let-7d in zebrafish embryos at 24 hpf, and found an increased expression of dre-cnr1 mRNA (F(2,6) ¼ 8.651, P ¼ 0.0139; Fig. 3A). A similar effect was evident when dre-miR-124 was silenced (F(2,6) ¼ 8.651, P ¼ 0.0095; Fig. 3A). To assess the regulation of CB1 receptor mRNA expression by let7d in mammalian cells we tested the effect of a let-7d-inhibiting RNA sequence, which specifically hybridizes to let-7d and so blocks its function. Exposure of cultured SH-SY5Y cells to the let-7d inhibitor for 6 days increased CB1 receptor mRNA expression by 50 ± 5% from control untransfected cells (P < 0.0001). We analyzed in parallel CB2 receptor mRNA expression and found that it was not significantly affected by let-7d inhibition (13 ± 26% decrease from

A. Chiarlone et al. / Neuropharmacology 108 (2016) 345e352

349

control untransfected cells; P ¼ 0.8439). When similar experiments were conducted in primary cultures of rat striatal neurons we observed a similar effect of a 6-day exposure to the let-7d inhibitor on CB1 receptor mRNA expression (42 ± 8% increase from control untransfected cells; P ¼ 0.0076), with no significant changes in CB2 receptor mRNA expression (8 ± 18% increase from control untransfected cells; P ¼ 0.8890). To evaluate whether let-7d can directly target CB1 receptor mRNA we first conducted bioinformatic analyses, using Ensembl (http://www.ensembl.org/index.html), MicroCosm Targets (http:// www.ebi.ac.uk/enright-srv/microcosm/htdocs/targets/v5) and mFold (http://mfold.rna.albany.edu/?q¼DINAMelt/Two-statemelting), in search of putative let-7d binding sites. As a result, we carried out luciferase assays on the dre-cnr1 exon 2 sequence 50 CTAGCTAGGTACCTAGGCCATAGGAACCAGCACCGCGAAG-30 because in this underlined region, but not in the UTR regions, a predicted let-7d binding site was found. This sequence (either wild-type or with a mutation in the seed region) was cloned into the reporter plasmid pmirGlo, and the resulting constructs were transfected into HEK293 cells. In parallel, a let-7d mimic was transfected to assess the putative binding site. Reporter assays showed that let-7d overexpression did not reduce luciferase activity on the wild-type plasmid (F(4,20) ¼ 1.017, P ¼ 0.4223; Fig. 3B). Likewise, we subsequently cloned in the reporter plasmid the equivalent sequences located on exon 2 of human CB1 receptor mRNA (hsa-cnr1; 50 CTAGCTAGGTACCTAGACTGTGCAGTTGCTGTTTACTTAAG-30 ) and mouse CB1 receptor mRNA (mmu-cnr1; 50 -CTAGCTAGGTACCTAGAACTGTGTTAGGCCTGCATTTTCAAG-30 ), as well as the corresponding mismatch versions. These sequences were transfected into HEK293 cells and challenged to the let-7d mimic. In both hsacnr1 and mmu-cnr1 reporter assays, luciferase activity on the wildtype plasmids did not change upon let-7d overexpression (F(4,20) ¼ 1.022, P ¼ 0.4201 for hsa-cnr1 plasmid; F(4,20) ¼ 0.590, P ¼ 0.6735 for the mmu-cnr1 plasmid; Fig. 3B). Together, these data indicate that, at least in this reporter-construct experimental setting, the only predicted binding site of let-7d on CB1 receptor mRNA does not seem to be functional. 3.4. Inhibition of cannabinoid signaling and cannabinoid-opioid cross-signaling by let-7d We next asked whether let-7d exerts a feedback control on CB1 receptor function. Overexpression of let-7d, miR-124 or miR-23b in SH-SY5Y cells prevented the typical Gi protein-mediated cAMPlowering action of the CB1 receptor upon forskolin challenge (F(5,36) ¼ 9.643, P < 0.0001; Fig. 4A). A similar effect was found for let-7d, miR-124 and let-7a in the case of the Gi protein-coupled mopioid receptor, which was again used as control (F(5,26) ¼ 7.327, P < 0.0001; Fig. 4B). Overexpression of let-7d also impaired the typical agonist-induced ERK phosphorylating/activating action of the CB1 and m-opioid receptor (Fig. 4B). As let-7d impairs cannabinoid and opioid biochemical actions, and functional interactions between cannabinoid and opioids have been reported in both animal models and humans (Manzanares et al., 1999; Robledo et al., 2008; Befort, 2015), we asked whether let-7d could affect cannabinoid signaling and cannabinoid-opioid cross-signaling. For this purpose we used primary striatal neurons as a physiologically-relevant cellular model. These cells Fig. 3. let-7d inactivation enhances CB1 receptor mRNA expression in vivo. (A) Zebrafish embryos at 24 hpf were microinjected with a dre-let-7d-targeted or a dremiR124-targeted morpholino (or a scrambled morpholino as control) and the expression levels of dre-cnr1 mRNA were measured (n ¼ 3 pools of 50 embryos for each condition). (B) HEK293 cells were transfected with the reporter plasmid pmirGlo or with pmirGlo additionally containing a putative let-7d binding site in the dre-cnr1

exon 2, hsa-cnr1 exon 2 or mmu-cnr1 exon 2 [either wild-type (wt) or with a mutation in the seed region (mismatch, mis)]. In parallel, cells were transfected or not with a let7d mimic. Finally, luciferase activity was measured after 48 h (n ¼ 5 for each condition). Data were analyzed by ANOVA with post-hoc Fisher’s LSD test. **P < 0.01 from the corresponding control animals (a.u. ¼ arbitrary units).

350

A. Chiarlone et al. / Neuropharmacology 108 (2016) 345e352

A

SH-SY5Y cells Vehicle

MAEA

Vehicle

1.5

1.5

Morphine

cAMP levels (a.u.)

## ##

##

##

##

1.0

##

1.0

*

* ##

*

0.5

0.5

**

**

**

**

** 0

0 Scr

Scr

let-7d miR-124 let-7a miR-23b miR-98

let-7d miR-124 let-7a miR-23b miR-98

B SH-SY5Y cells Scrambled Vehicle

Scrambled

let-7d

MAEA

Vehicle

MAEA

Vehicle

Morphine

let-7d Vehicle

Morphine

pERK (42-44 kD) -actin

C

Primary striatal neurons Acute vehicle

cAMP levels (a.u.)

1.5

Acute MAEA

##

1.0

##

##

** **

**

**

0.5

6d morphine

##

**

**

**

** ** 0

i-let-7d

Acute morphine

6d MAEA

6d vehicle

A

B

C

D

E

F

+

G

H

+

I

J

K

+

L

M

+

Fig. 4. let-7d inhibits cannabinoid signaling and cannabinoid-opioid cross-signaling. (A) SH-SY5Y cells were transfected with plasmids encoding different miRNAs [or a scrambled (Scr) sequence as control]. Seventy two h later, cells were incubated for 15 min with vehicle or 25 mM forskolin, alone or in combination with vehicle, 0.5 mM methanandamide (MAEA) or 1 mM morphine, and cAMP was assayed (n ¼ 2 for each condition). (B) In the scrambled-RNA and let-7d-expressing cells, Western blot analysis of phosphorylated ERK was conducted after incubation for 15 min with vehicle, 0.5 mM MAEA or 1 mM morphine (n ¼ 2 for each condition). Representative blots are shown. Images from different parts of the same gels were grouped. (C) Primary rat striatal neurons, either “naïve” (vehicle-treated) or exposed to 0.5 mM MAEA or 1 mM morphine for 6 days, were used to measure the acute effect of 0.5 mM MAEA or 1 mM morphine on cAMP levels. In parallel, cells were transfected or not with a let-7d-inhibitor (i-let-7d) 4 days prior to cAMP determinations (n ¼ 2 for each condition). Data were analyzed by ANOVA with post-hoc Fisher’s LSD test. *P < 0.05, **P < 0.01 from the corresponding vehicle-treated cells. ## P < 0.01 from the corresponding scrambled RNA-expressing cells (panel A) or i-let-7d non-expressing cells (panel C) (a.u. ¼ arbitrary units).

expressed considerable levels of CB1 receptor mRNA (Cq ¼ 32) and m-opioid receptor mRNA (Cq ¼ 33), but only negligible amounts of CB2 receptor mRNA (Cq ¼ 39). In this experimental setting we first controlled that exposure to methanandamide for 6 days reduced CB1 receptor mRNA expression (32 ± 7% decrease from vehicletreated cells; P ¼ 0.0450) and m-opioid receptor mRNA expression (55 ± 10% decrease from vehicle-treated cells; P ¼ 0.0377), without significantly affecting CB2 receptor mRNA expression (13 ± 20% decrease from vehicle-treated cells; P ¼ 0.8489). Likewise, we found that exposure to morphine for 6 days reduced CB1 receptor

mRNA expression (44 ± 14% decrease from vehicle-treated cells; P ¼ 0.0215) and m-opioid receptor mRNA expression (51 ± 6% decrease from vehicle-treated cells; P ¼ 0.0406), without significantly affecting CB2 receptor mRNA expression (7 ± 18% decrease from vehicle-treated cells; P ¼ 0.9004). Subsequently, “naïve” neurons and neurons that had been exposed to methanandamide or morphine for 6 days were used to evaluate the acute cAMPlowering effect of methanandamide or morphine. Signaling inhibition, expressed as the inability to acutely lower cAMP levels down to those of “naïve” cells (P < 0.0001, bars B and C; Fig. 4C), was

A. Chiarlone et al. / Neuropharmacology 108 (2016) 345e352

observed for methanandamide (P < 0.0001, bars E vs. B; Fig. 4C) and morphine (P < 0.0001, bars L vs. C; Fig. 4C). Inhibition of crosssignaling between the two drugs was also evident (P < 0.0001, bars G vs. C and bars J vs. B; Fig. 4C). Of note, let-7d inhibition restored the acute cAMP-lowering action of the cannabinoid and the opioid, as well as of the cannabinoid-opioid and opioidcannabinoid cross-treatments (acute treatment x chronic treatment: F(8,39) ¼ 26.59, P < 0.0001; Fig. 4C). 4. Discussion Despite the wide expression and key function of the CB1 receptor in the central nervous system during health and disease, the possible bidirectional connection between the receptor and miRNAs had not been studied so far in neural tissue. Some previous studies had focused on the effect of cannabinoid ligands on miRNA expression in cells of non-neural origin (Hegde et al., 2013; Sreevalsan and Safe, 2013; Jackson et al., 2014). In these studies let-7d was not among the miRNAs whose expression was reported to change upon cannabinoid challenge. However, this is not surprising owing to the well known dependency of CB1 receptor action on contextual factors such as the characteristics of the cell under n, 2003) and the pattern of drug administration study (Guzma (Mechoulam and Parker, 2013). In fact, for example, cannabinoids decreased miR-27a expression in one of those studies (Sreevalsan and Safe, 2013) but increased it in another (Hegde et al., 2013). Here, by using various in vitro and in vivo systems, we unequivocally show that, in the brain, the CB1 receptor up-regulates let-7d, which, in turn, impairs CB1 receptor signaling. Our findings thus unveil a new player in the complex process of cannabinoid action. We are nonetheless aware that other molecular determinants can participate as well in the control of cannabinoid signaling. For example, as miRNAs act at a post-transcriptional level, let-7d is not expected to control gene transcription. It is also conceivable that, according to the present report, other miRNAs, especially miR-124, may contribute to tuning cannabinoid signaling, as it has been shown for opioid signaling (Barbierato et al., 2015). However, it is still unknown whether there are mechanisms of cross-talk that control the expression and/or function of let-7d and miR-124 coordinately, or whether these two miRNAs are produced and/or act independently. Additionally, as supported by our data, let-7d most likely targets the mRNA(s) encoding protein(s) that control CB1 receptor expression and/or function, rather than targeting CB1 receptor mRNA directly. In fact, the notion of a single mechanism underlying cannabinoid tolerance appears too simplistic and, instead, multiple mechanisms are most likely involved. Our data suggest that let-7dinduced impairment of CB1 receptor signaling might be one of those mechanisms. Synergism between opioids and cannabinoids has been demonstrated in a number of animal models, especially in the field of pain. Thus, the antinociceptive effects of opioids can be enhanced by cannabinoids, an effect that may rely, for example, on the common Gi/o protein-dependent signal transduction mechanisms shared by m-opioid and CB1 cannabinoid receptors and the ability of cannabinoids to enhance the production of opioid peptides (Manzanares et al., 1999; Nadal et al., 2013). In humans, a pilot study has been conducted to investigate the effect of concomitant cannabinoids on the pharmacokinetics of opioid analgesics (Abrams et al., 2011). Specifically, patients with chronic pain on a stable dose of sustained-release opioid (morphine or oxycodone) were monitored before and after 4 days of exposure to a cannabis preparation. No adverse side-effects of combining cannabinoids and opioids were observed, and there were no significant alterations in the area under the curves for the opioids after the administration of cannabinoids (Abrams et al., 2011). Moreover, a

351

potential synergistic relief of pain was appreciated between cannabinoids and opioids, in agreement with a previous phase I/II trial conducted on chronic-pain patients (Narang et al., 2008). If cannabinoids and opioids were shown to be synergistic in larger clinical studies, it is conceivable that lower doses of opioids would be effective for longer periods of time with fewer side-effects, clearly benefiting patients with pain -as it has been shown in some animal models (Manzanares et al., 1999; Nadal et al., 2013). In fact, a randomized, double-blind, placebo-controlled, phase III trial has shown that cannabinoids may be useful add-on analgesic drugs for patients with opioid-refractory cancer pain (Portenoy et al., 2012). Although more exhaustive follow-on controlled human studies are indeed necessary, this evidence strongly supports that cannabinoids might be combined with opioids for benefiting patients with pain -and perhaps other diseases (Bisaga et al., 2015; Hurd et al., 2015). Conflict of interest The authors declare no conflict of interest. Acknowledgements This work was supported by Ministerio de Economía y Competitividad (grants SAF2012-35759 and SAF2015-64945-R to MG), Comunidad de Madrid (grant S2010/BMD-2308 to MG), DoktorRobert-Pfleger-Stiftung (to JK), and Instituto de Salud Carlos III (grant FIS 12/00094 to MLGB). AC was supported by Ministerio de Economía y Competitividad (FPI Program). We are grateful to Eva Resel for expert technical assistance. References Abrams, D.I., Couey, P., Shade, S.B., Kelly, M.E., Benowitz, N.L., 2011. Cannabinoidopioid interaction in chronic pain. Clin. Pharmacol. Ther. 90, 844e851. Adams, I.B., Martin, B.R., 1996. Cannabis: pharmacology and toxicology in animals and humans. Addiction 91, 1585e1614. Aguado-Fraile, E., Ramos, E., Saenz-Morales, D., Conde, E., Blanco-S anchez, I., Stamatakis, K., et al., 2012. miR-127 induced during renal ischemia/reperfusion via HIF-1a protects proximal tubule cells against ischemic injury: KI3FB as miR127 target. PLoS One 7, e44305.  mez, L., Lietor, A., Aguado-Fraile, E., Ramos, E., Conde, E., Rodríguez, M., Martín-Go et al., 2015. A pilot study identifying a set of microRNAs as precise diagnostic biomarkers of acute kidney injury. PLoS One 10, e0127175. r, A., Di Marzo, V., Harkany, T., 2016. At the tip of an iceberg: prenatal marijuana Alpa and its possible relation to neuropsychiatric outcome in the offspring. Biol. Psychiatry 79, e33ee45. Barbierato, M., Zusso, M., Skaper, S.D., Giusti, P., 2015. MicroRNAs: emerging role in the endogenous m opioid system. CNS Neurol. Disord. Drug Targets 14, 239e250. Befort, K., 2015. Interactions of the opioid and cannabinoid systems in reward: insights from knockout studies. Front. Pharmacol. 6, 6. Bisaga, A., Sullivan, M.A., Glass, A., Mishlen, K., Pavlicova, M., Haney, M., et al., 2015. The effects of dronabinol during detoxification and the initiation of treatment with extended release naltrexone. Drug Alcohol Depend. 154, 38e45. zquez, C., Chiarlone, A., Sagredo, O., Aguado, T., Pazos, M.R., Resel, E., et al., 2011. Bla Loss of striatal type 1 cannabinoid receptors is a key pathogenic factor in Huntington’s disease. Brain 134, 119e136. €rner, C., Stumm, R., Ho €llt, V., Kraus, J., 2007. Comparative analysis of mu-opioid Bo receptor expression in immune and neuronal cells. J. Neuroimmunol. 188, 56e63. €rner, C., Bedini, A., Hollt, V., Kraus, J., 2008. Analysis of promoter regions reguBo lating basal and interleukin-4-inducible expression of the human CB1 receptor gene in T lymphocytes. Mol. Pharmacol. 73, 1013e1019. Brites, D., Fernandes, A., 2015. Neuroinflammation and depression: microglia activation, extracellular microvesicles and microRNA dysregulation. Front. Cell Neurosci. 9, 476. Grotenhermen, F., 2003. Pharmacokinetics and pharmacodynamics of cannabinoids. Clin. Pharmacokinet. 42, 327e360. n, M., 2003. Cannabinoids: potential anticancer agents. Nat. Rev. Cancer 3, Guzma 745e755. He, Y., Yang, C., Kirkmire, C.M., Wang, Z.J., 2010. Regulation of opioid tolerance by let-7 family microRNA targeting the mu opioid receptor. J. Neurosci. 30, 10251e10258. Hegde, V.L., Tomar, S., Jackson, A., Rao, R., Yang, X., Singh, U.P., et al., 2013. Distinct

352

A. Chiarlone et al. / Neuropharmacology 108 (2016) 345e352

microRNA expression profile and targeted biological pathways in functional myeloid-derived suppressor cells induced by D9-tetrahydrocannabinol in vivo: regulation of CCAAT/enhancer-binding protein a by microRNA-690. J. Biol. Chem. 288, 36810e368126. Horton, J.K., Martin, R.C., Kalinka, S., Cushing, A., Kitcher, J.P., O’Sullivan, M.J., et al., 1992. Enzyme immunoassays for the estimation of adenosine 30 ,50 cyclic monophosphate and guanosine 30 ,50 cyclic monophosphate in biological fluids. J. Immunol. Methods 155, 31e40. Hurd, Y.L., Yoon, M., Manini, A.F., Hernandez, S., Olmedo, R., Ostman, M., et al., 2015. Early phase in the development of cannabidiol as a treatment for addiction: opioid relapse takes initial center stage. Neurotherapeutics 12, 807e815. Issler, O., Chen, A., 2015. Determining the role of microRNAs in psychiatric disorders. Nat. Rev. Neurosci. 16, 201e212. Jackson, A.R., Nagarkatti, P., Nagarkatti, M., 2014. Anandamide attenuates Th-17 cellmediated delayed-type hypersensitivity response by triggering IL-10 production and consequent microRNA induction. PLoS One 9, e93954. Katona, I., Freund, T.F., 2008. Endocannabinoid signaling as a synaptic circuit breaker in neurological disease. Nat. Med. 14, 923e930. Kimmel, C.B., Ballard, W.W., Kimmel, S.R., Ullmann, B., Schilling, T.F., 1995. Stages of embryonic development of the zebrafish. Dev. Dyn. 203, 253e310. Lam, C.S., Rastegar, S., Str€ ahle, U., 2006. Distribution of cannabinoid receptor 1 in the CNS of zebrafish. Neuroscience 138, 83e95. Lee, M.C., Smith, F.L., Stevens, D.L., Welch, S.P., 2003. The role of several kinases in mice tolerant to delta-9-tetrahydrocannabinol. J. Pharmacol. Exp. Ther. 305, 593e5999. Lorenzen, J.M., 2015. Vascular and circulating microRNAs in renal ischaemiareperfusion injury. J. Physiol. 593, 1777e1784. Manzanares, J., Corchero, J., Romero, J., Fernandez-Ruiz, J.J., Ramos, J.A., Fuentes, J.A., 1999. Pharmacological and biochemical interactions between opioids and cannabinoids. Trends Pharmacol. Sci. 20, 287e294. Marsicano, G., Wotjak, C.T., Azad, S.C., Bisogno, T., Rammes, G., Cascio, M.G., et al., 2002. The endogenous cannabinoid system controls extinction of aversive memories. Nature 418, 530e534. McPartland, J.M., Glass, M., Matias, I., Norris, R.W., Kilpatrick, C.W., 2007. A shifted repertoire of endocannabinoid genes in the zebrafish (Danio rerio). Mol. Genet. Genomics 277, 555e570. Mechoulam, R., Parker, L.A., 2013. The endocannabinoid system and the brain. Annu. Rev. Psychol. 64, 21e47. Nadal, X., La Porta, C., Andreea Bura, S., Maldonado, R., 2013. Involvement of the opioid and cannabinoid systems in pain control: new insights from knockout studies. Eur. J. Pharmacol. 716, 142e157. Narang, S., Gibson, D., Wasan, A.D., Ross, E.L., Michna, E., Nedeljkovic, S.S., et al., 2008. Efficacy of dronabinol as an adjuvant treatment for chronic pain patients on opioid therapy. J. Pain 9, 254e264. Notcutt, W.G., 2015. Clinical use of cannabinoids for symptom control in multiple sclerosis. Neurotherapeutics 12, 769e777. Panlilio, L.V., Goldberg, S.R., Justinova, Z., 2015. Cannabinoid abuse and addiction: clinical and preclinical findings. Clin. Pharmacol. Ther. 97, 616e627. Pasquariello, N., Catanzaro, G., Marzano, V., Amadio, D., Barcaroli, D., Oddi, S., et al., 2009. Characterization of the endocannabinoid system in human neuronal cells and proteomic analysis of anandamide-induced apoptosis. J. Biol. Chem. 284, 29413e29426.

Pertwee, R.G., 2012. Targeting the endocannabinoid system with cannabinoid receptor agonists: pharmacological strategies and therapeutic possibilities. Philos. Trans. R. Soc. Lond. B Biol. Sci. 367, 3353e3363. Pertwee, R.G., Howlett, A.C., Abood, M.E., Alexander, S.P., Di Marzo, V., Elphick, M.R., et al., 2010. International union of basic and clinical pharmacology. LXXIX. Cannabinoid receptors and their ligands: beyond CB1 and CB2. Pharmacol. Rev. 62, 588e631. Piomelli, D., 2003. The molecular logic of endocannabinoid signalling. Nat. Rev. Neurosci. 4, 873e884. Portenoy, R.K., Ganae-Motan, E.D., Allende, S., Yanagihara, R., Shaiova, L., Weinstein, S., et al., 2012. Nabiximols for opioid-treated cancer patients with poorly-controlled chronic pain: a randomized, placebo-controlled, graded-dose trial. J. Pain 13, 438e449. Qu, S., Guan, J., Liu, Y., 2015. Identification of microRNAs as novel biomarkers for glioma detection: a meta-analysis based on 11 articles. J. Neurol. Sci. 348, 181e187. Rehfeld, F., Rohde, A.M., Nguyen, D.T., Wulczyn, F.G., 2015. Lin28 and let-7: ancient milestones on the road from pluripotency to neurogenesis. Cell Tissue Res. 359, 145e160. Robledo, P., Berrendero, F., Ozaita, A., Maldonado, R., 2008. Advances in the field of cannabinoideopioid cross-talk. Addict. Biol. 13, 213e224. n Ferna ndez de Velasco, E., Rodríguez, R.E., 2007. Rodríguez-Martín, I., Marro Characterization of cannabinoid-binding sites in zebrafish brain. Neurosci. Lett. 413, 249e254. Rubino, T., Vigano, D., Massi, P., Spinello, M., Zagato, E., Giagnoni, G., et al., 2000. Chronic delta-9-tetrahydrocannabinol treatment increases cAMP levels and cAMP-dependent protein kinase activity in some rat brain regions. Neuropharmacology 39, 1331e1336. , D., Zippel, R., Parolaro, D., 2005. Ras/ERK signalling in Rubino, T., Forlani, G., Vigano cannabinoid tolerance: from behaviour to cellular aspects. J. Neurochem. 93, 984e991. Rudnicki, M., Beckers, A., Neuwirt, H., Vandesompele, J., 2015. RNA expression signatures and posttranscriptional regulation in diabetic nephropathy. Nephrol. Dial. Transpl. (Suppl. 4), iv35eiv42. Smalheiser, N.R., 2014. The RNA-centred view of the synapse: non-coding RNAs and synaptic plasticity. Philos. Trans. R. Soc. Lond. B Biol. Sci. 369, 1652. Sreevalsan, S., Safe, S., 2013. The cannabinoid WIN 55,212-2 decreases specificity protein transcription factors and the oncogenic cap protein eIF4E in colon cancer cells. Mol. Cancer Ther. 12, 2483e2493. Sun, E., Shi, Y., 2015. MicroRNAs: small molecules with big roles in neurodevelopment and diseases. Exp. Neurol. 268, 46e53. Syed, Y.Y., McKeage, K., Scott, L.J., 2014. Delta-9-tetrahydrocannabinol/cannabidiol (Sativex®): a review of its use in patients with moderate to severe spasticity due to multiple sclerosis. Drugs 74, 563e578. Tonini, R., Ciardo, S., Cerovic, M., Rubino, T., Parolaro, D., Mazzanti, M., et al., 2006. ERK-dependent modulation of cerebellar synaptic plasticity after chronic D9tetrahydrocannabinol exposure. J. Neurosci. 26, 5810e5818. Wan, Y., Liu, Y., Wang, X., Wu, J., Liu, K., Zhou, J., et al., 2015. Identification of differential microRNAs in cerebrospinal fluid and serum of patients with major depressive disorder. PLoS One 10, e0121975. World Drug Report, 2015. United Nations Office on Drugs and Crime. New York, USA.