Astroglial CB1 Reveal Sex‐Specific Synaptic Effects of Amphetamine
INSERM, U1215 Neurocentre Magendie Université de Bordeaux Bordeaux France
Department of Biomedical and Biotechnological Sciences, Section of Pharmacology University of Catania Catania Italy
Department of Biomedicine, Neuroscience and Advanced Diagnostics (BIND) University of Palermo Palermo Italy
INRAE, Nutrition and Integrative Neurobiology, UMR 1286 Université de Bordeaux Bordeaux France
CINBIO University of Vigo Vigo Spain
Galicia Sur Health Research Institute (IISGS), Laboratory of Neuroscience Vigo Spain
* Correspondence:Luigi Bellocchio (luigi.bellocchio@inserm.fr)
Giovanni Marsicano (giovanni.marsicano@inserm.fr)
Ana Covelo (covelo.ana@gmail.com)
ABSTRACT
The Nucleus Accumbens (NAc) is a critical brain region for the effects of psychostimulant drugs. Type‐1 cannabinoid receptors (CB1), the main elements of the endocannabinoid system (ECS) in the brain, participate in these effects and modulate synaptic functions in the NAc. Besides their neuronal expression, CB1 receptors are also present in astrocytes, where they contribute to the regulation of synaptic plasticity and behavior. However, the impact of astroglial CB1 receptors on synaptic plasticity in the NAc and on psychostimulant‐induced synaptic and behavioral effects is currently unknown. This study shows that the psychostimulant amphetamine impairs a form of astroglial CB1 receptor‐dependent synaptic plasticity in the NAc of male, but not female mice. Consistently, locomotor effects of amphetamine require astroglial CB1 receptors in male, but not female mice. These results, by revealing unforeseen mechanisms underlying sex‐dependent effects of amphetamine, pave the way to a better understanding of the diverse impact of psychostimulants in women and men.
Main Points
- Astroglial CB1 mediate long‐term depression in the NAc through purine release.
- Amphetamine occludes LTD in male mice by inducing the saturation of A1 receptors.
- Astroglial CB1 mediate behavioral amphetamine effects in males, but not in females.
Boxed Text
Article notes
Footnote Group
1Introduction
Most research regarding drugs of abuse has been conducted in male animals, while research in females has been limited in the past and its systematic implementation has started only recently (Becker and Chartoff 2019; Becker and Koob 2016; Calipari et al. 2017; Torres 2022). As a consequence, very little is known concerning the potential mechanisms underlying sex‐dependent differences in the response to drugs of abuse in general, and of psychostimulants in particular (Becker and Hu 2008; Cornish and Prasad 2021; Torres 2022).
Being one of the main targets of dopamine circuits, the Nucleus Accumbens (NAc) bears a prominent role in reward processing and in the responses to drugs of abuse (Carlezon Jr. and Thomas 2009; Lammel et al. 2014; Sesack and Grace 2010), including psychostimulants (Carboni et al. 1989; Heinsbroek et al. 2017; Luscher and Malenka 2011). The activity of the NAc is highly regulated by glutamatergic projections arriving from the amygdala, the hippocampus, and the prefrontal cortex, which impinge onto GABAergic medium spiny neurons (MSNs) (Turner et al. 2018). These glutamatergic inputs express several forms of plasticity likely involved in adaptive and pathological behaviors (Heinsbroek et al. 2017; Jedynak et al. 2016; Quintero 2013; Turner et al. 2018). Additionally, changes in glutamatergic synaptic plasticity occur following exposure to drugs of abuse (Lee and Dong 2011; Li and Kauer 2004; Luscher and Malenka 2011; Mameli et al. 2009; Schmidt and Pierce 2010; Stuber et al. 2010). Importantly, psychostimulants have been suggested to induce higher levels of dopamine in the NAc of men as compared to women (Riccardi et al. 2006), but potential sex differences in drug‐related synaptic plasticity in this brain region are virtually unknown (Cornish and Prasad 2021; Torres 2022). Thus, understanding whether and how synaptic plasticity in the NAc is differentially regulated in male and female subjects is fundamental to disentangle the mechanisms underlying physiological as well as drug‐related behaviors.
The endocannabinoid system (ECS) and its main receptor in the brain, the type‐1 cannabinoid receptor (CB1), have emerged in the last decades as one of the most important sources of modulation in the central nervous system (CNS) (Kano et al. 2009). Functional CB1 receptors are present in different cell types throughout the brain, including neurons and glia, but also associated with different subcellular compartments such as the plasma membrane and the mitochondria (Bellocchio et al. 2010; Benard et al. 2012; Gomez‐Sotres et al. 2024; Gutierrez‐Rodriguez et al. 2018; Hebert‐Chatelain et al. 2016; Jimenez‐Blasco et al. 2020; Soria‐Gomez et al. 2021). Thus, CB1 receptors have been involved in many functions ranging from the regulation of cellular metabolic processes to the control of synaptic transmission and plasticity, up to the modulation of different behaviors (Busquets Garcia et al. 2016). The ECS is implicated in several forms of synaptic plasticity in the NAc (Bilbao et al. 2020; Covey and Yocky 2021; Grueter et al. 2010; Kasanetz et al. 2010; Parsons and Hurd 2015), and impaired endocannabinoid signaling in this brain region contributes to increased stress response, negative emotional states, and cravings that propel addiction (Sidhpura and Parsons 2011). However, the cellular and subcellular mechanisms involving the CB1 receptors in synaptic plasticity in the NAc are scantily known. Moreover, despite the existence of sex‐dependent differences in the activity of the ECS in specific brain regions (Liu et al. 2020; Rubino and Parolaro 2011; Wagner 2016), very little is known concerning the NAc.
By providing structural and metabolic support to other cells, astrocytes are key regulators of brain functions (Bonvento and Bolanos 2021; Chiareli et al. 2021). In addition, during the last decades, these cells emerged as direct controllers of synaptic functions, being active players of the so‐called Tripartite Synapse (Araque et al. 1999). Astrocytes express a wide variety of functional receptors that allow them to respond to neurotransmitters released by neurons. In turn, astrocytes release neuroactive molecules, named gliotransmitters (Volterra and Meldolesi 2005), that stimulate receptors on different brain cell types, including pre‐ or post‐synaptic neuronal terminals. Through the release of gliotransmitters, astrocytes can modulate neuronal activity and synaptic transmission in several brain areas, thereby directly impacting animal behavior (Araque et al. 2014; Covelo and Araque 2016; Kofuji and Araque 2021; Oliveira et al. 2015). Several regions in the CNS have been proven to contain functional astroglial CB1 receptors, and accumulating evidence points to a key role of these receptors in the astrocyte contribution to synaptic processes (Covelo et al. 2021; Gomez‐Gonzalo et al. 2015; Han et al. 2012; Martin‐Fernandez et al. 2017; Martin et al. 2015; Min and Nevian 2012; Navarrete and Araque 2010; Rasooli‐Nejad et al. 2014; Robin et al. 2018). Astrocytes in the NAc regulate neuronal excitability, synaptic activity, and behavior (Bull et al. 2014; Corkrum et al. 2020; D'Ascenzo et al. 2007; Fellin et al. 2007; Scofield et al. 2015; Serra et al. 2022), but whether and how this is controlled by CB1 receptors is currently unknown.
Here, we found that CB1 receptors mediate spike‐timing‐dependent synaptic plasticity (STDP) in the NAc of both male and female adult mice. Long‐term depression induced by pre‐post stimulations (t‐LTD) (Brzosko et al. 2019) is mediated by astroglial CB1 receptors and the release of purines that act at A1 receptors in both sexes. Interestingly, we observed that amphetamine application blocks t‐LTD in males but not in females. Consistently, astroglial CB1 receptors are involved in the behavioral effects of amphetamine in males, but not in female mice, thereby contributing to the sexually dimorphic responses to psychostimulants.
2Results
2.1Spike‐Timing‐Dependent Long‐Term Depression (t‐LTD) in the NAc Core Requires Astrocyte Activity
To study synaptic plasticity in the NAc, we focused on spike‐timing‐dependent plasticity (STDP), a form of long‐term synaptic plasticity in which the timing of pre‐ and post‐synaptic action potentials determines the direction and magnitude of synaptic changes (Brzosko et al. 2019). We performed patch‐clamp whole‐cell recordings from NAc core medium spiny neurons (MSNs) and monitored evoked excitatory post‐synaptic currents (EPSCs). STDP was induced by pairing a pre‐synaptic electrical stimulation with a post‐synaptic action potential (evoked by a short depolarization of 5–8 ms) within a time window of ≃10 ms (100 pairings at 1 Hz; Figure 1a). EPSC amplitude was then monitored for 45 min following STDP induction. This pre‐post pairing protocol induced a significant time‐dependent long‐term depression (t‐LTD) in both male and female mice (Figure 1b–e). As expected, t‐LTD was induced by the close sequential activation of the pre‐ and post‐synaptic compartments, as no changes in EPSC amplitude were observed when only the pre‐ or only the post‐synaptic stimulations were applied (Figure 1d). The analysis of the squared coefficient of variation (Figure 1f) indicated that t‐LTD in the NAc core is likely expressed pre‐synaptically (Faber and Korn 1991).
As expected (Bosch‐Bouju et al. 2016), the CB1 receptor antagonist Rimonabant (1 μM) blocked t‐LTD in slices from male mice, but also from females (Figure 1d,e), showing that CB1 receptors are required for this form of plasticity in both sexes. Astrocytes have also been shown to participate in t‐LTD in the cortex and hippocampus by releasing glutamate or D‐serine, respectively, to activate pre‐synaptic NMDA receptors (Andrade‐Talavera et al. 2016; Min and Nevian 2012). To explore astrocyte involvement in t‐LTD in the NAc core, we blocked astrocyte G Protein‐Coupled Receptor (GPCR) signaling by loading the astrocyte network with GDPβS (10 mM) (Corkrum et al. 2020; Covelo and Araque 2018). This treatment blocked t‐LTD in both male and female mice (Figure 1g,h), indicating that this form of synaptic plasticity involves the activation of astrocytes. To investigate the functional consequences of such activation, we next explored the identity of the potential gliotransmitter(s) involved. D‐AP5 (100 μM), a selective NMDA receptor antagonist, did not block the plasticity in either male or female mice (Figure 1g,h), excluding the involvement of NMDA receptors in NAc t‐LTD. Astrocytes in the NAc core have been reported to release purines (adenosine and/or ATP) that inhibit synaptic transmission by acting on inhibitory adenosine A1 receptors (Corkrum et al. 2020). Consistently, we found that t‐LTD was blocked by the adenosine A1 receptor antagonist CPT (2 μM) in slices from both male and female mice (Figure 1g,h), indicating that purines are involved in this form of NAc synaptic plasticity. Altogether, these data show that t‐LTD requires CB1 receptor signaling, astrocyte activation, and A1 adenosine receptor stimulation in both sexes.
2.2T‐LTD Is Mediated by CB1 Receptors Located in the Plasma Membrane of Astrocytes
Next, we investigated the cellular and subcellular location of the CB1 receptors involved in t‐LTD. At the subcellular level, CB1 receptors are largely localized at plasma membrane compartments (pmCB1). However, they can also be found intracellularly, for instance in association with mitochondrial membranes (mtCB1), where they can impact cellular, synaptic, and behavioral functions (Bellocchio et al. 2010; Benard et al. 2012; Gutierrez‐Rodriguez et al. 2018; Hebert‐Chatelain et al. 2016; Jimenez‐Blasco et al. 2020; Soria‐Gomez et al. 2021). We, thus, used a mouse line where CB1 receptors are excluded from mitochondrial membranes (DN22‐CB1‐KI) (Benard et al. 2012; Hebert‐Chatelain et al. 2016) to investigate whether mtCB1 receptors are involved in t‐LTD. This form of plasticity was still present in NAc slices derived from DN22‐CB1‐KI mice (Figure 2a) but it was blocked by Hemopressin (10 μM; Figure 2a), a CB1 receptor antagonist that cannot cross the cell membrane (Bellocchio et al. 2010; Benard et al. 2012). These data suggest that pmCB1 receptors mediate t‐LTD in the NAc core independently of mtCB1 receptors.
We next investigated the cellular location of the CB1 receptors involved in t‐LTD. At the synaptic level, the classical view of the ECS mode of action is that endocannabinoids are mobilized from the post‐synaptic compartment and travel retrogradely to the pre‐synaptic cell, where they activate neuronal CB1 receptors to reduce neurotransmitter release (Araque et al. 2017). Despite the involvement of astrocytes, it is still possible that similar neuronal mechanisms are involved in t‐LTD. Thus, we asked whether CB1 receptors expressed in glutamatergic axons impinging onto the NAc play a role in t‐LTD. To address this question, we used a mutant mouse line specifically lacking CB1 receptors in cortical glutamatergic neurons (Nex‐CB1 −/− mice, also known as Glu‐CB1 −/−) (Bellocchio et al. 2010; Monory et al. 2006). Notably, these mice exhibited normal t‐LTD (Figure 2b), indicating that CB1 receptors located at the pre‐synaptic terminals of cortical glutamatergic neurons are not involved in this form of NAc plasticity.
Astroglial CB1 receptors have been previously shown to participate in synaptic plasticity, including STDP in the neocortex (Min and Nevian 2012). To explore whether this subpopulation of CB1 receptors is involved in NAc t‐LTD, we used a transgenic mouse line that specifically lacks CB1 receptors in cells expressing the astrocyte marker glial fibrillary acidic protein (GFAP‐CB1 −/− mice) (Han et al. 2012). Strikingly, t‐LTD was not present in either male or female GFAP‐CB1 −/− mice (Figure 2c,d), indicating that t‐LTD requires the activation of astroglial CB1 receptors.
Finally, we investigated whether the lack of purine release might be the cause of the impairment of t‐LTD in the NAc of GFAP‐CB1 −/− mice. Notably, a puff application of adenosine during the STDP induction protocol rescued t‐LTD in both GFAP‐CB1 −/− male and female mice, without affecting the plasticity of WT littermates (Figure 2c,d). Overall, these data point to the idea that NAc t‐LTD is mediated by the activation of astrocyte pmCB1 receptors, eventually leading to purine release and stimulation of A1 adenosine receptors.
2.3Astrocyte Calcium Dynamics Associated With t‐LTD Induction
Changes in intracellular Ca2+ dynamics are currently considered the main mechanism underlying the cellular and synaptic functions of astrocytes (Covelo et al. 2022; Guerra‐Gomes et al. 2017). To investigate whether astrocytes undergo cytosolic Ca2+ changes during the t‐LTD induction protocol, we virally expressed the Ca2+ sensor GCaMP6f in astrocytes (Figure 3a). Analysis of live 2‐photon imaging of GCaMP6f‐expressing astrocytes in acute NAc slices from both male and female mice revealed that t‐LTD induction was associated with an increase in the frequency of astrocyte Ca2+ events in both sexes (Figures 3b–d and S1a), with no changes in amplitude, duration, or spreading (Figures 3e–g and S1b–d). The CB1 receptor blocker Rimonabant abolished the increase in Ca2+ event frequency associated with t‐LTD (Figures 3c,d and S1a), showing that CB1 receptors control astroglial Ca2+ dynamics upon t‐LTD induction.
Altogether, electrophysiological and imaging data indicate that STDP activates NAc astroglial pmCB1 receptors, which trigger intracellular Ca2+ signaling that eventually leads to the release of purines and to t‐LTD.
2.4Sex‐Dependent Involvement of Astroglial CB1 Receptors in Amphetamine‐Induced Hyperlocomotion
Synaptic plasticity in the NAc has been proposed to underlie the actions of most drugs of abuse, including psychostimulants (Heinsbroek et al. 2017; Jedynak et al. 2016; Quintero 2013; Turner et al. 2018). Therefore, we next investigated the potential role of NAc astroglial CB1 receptors in the psychomotor effects of the prototypical psychostimulant amphetamine. To this aim, we injected an adeno‐associated viral vector expressing Cre under the control of the GFAP promoter (AAV‐GFAP‐Cre) into the NAc of CB1 flox/flox mice to generate NAc‐GFAP‐CB1 −/− mice, that is, mice carrying an astrocyte‐specific deletion of the receptor in the NAc (Figure 4a). In NAc‐GFAP‐CB1 −/− mice, immunostaining showed that almost all Cre+ cells also expressed the astrocyte markers GFAP or S100β, while only very few cells contained the neuronal marker NeuN (Figure 4b,c). Importantly, the virus showed a high efficacy as most of the cells expressing GFAP or S100β also expressed Cre (Figure 4c). Additionally, when the AAV‐GFAP‐Cre was injected into the NAc of the Cre reporter mouse line Ai6 (Madisen et al. 2010) (Figure S2a) to check the actual recombination caused by the expression of the Cre protein, we found only marginal recombination in NeuN+ cells, whereas most of the recombination was present in astrocytes (Figure S2b). However, because some marginal virus expression and protein recombination were still present in NAc neurons, we performed fluorescent in situ hybridization (FISH) combined with immunohistochemistry to detect CB1 receptor mRNA in the NAc of NAc‐GFAP‐CB1 −/− mice. As already reported, CB1 receptor mRNA was sparse in the NAc core (Van Waes et al. 2012) and co‐localized with the NeuN protein (Figure S2c,d), indicating that, as in many brain regions, CB1 receptors are detectable by fluorescent microscopy in neurons only (Busquets‐Garcia et al. 2018), whereas astrocytes express low (although functionally important) levels of the receptor. There were no differences in the number of CB1 + cells found in NAc‐GFAP‐CB1 −/− mice and mice injected with the control virus (Figure S2e). Together, these data indicate that the AAV‐GFAP‐Cre was specifically expressed in astrocytes, where the recombination principally occurred, and that the injection of the viral vector did not decrease the expression of the neuronal CB1 receptor in CB1flox/flox mice. Importantly, while NAc astrocytes from mice injected with the control virus responded with calcium increases to the puff application of the CB1R agonist WIN 55,212–2 (10 μM), astrocytes from NAc‐GFAP‐CB1 −/− did not respond (Figure S3). However, they still showed robust calcium responses to ATP (2 mM), indicating that astrocytes from NAc‐GFAP‐CB1 −/− mice remain functional but selectively lack CB1R‐induced responses. Additionally, t‐LTD was absent in NAc‐GFAP‐CB1 −/− mice (Figure 4d), further supporting that local astroglial CB1 receptors are involved in this type of synaptic plasticity.
Amphetamine‐induced hyperlocomotion in male and female NAc‐GFAP‐CB1 −/− mice was studied in actimetry cages (Mariani et al. 2023). Locomotion was recorded during a habituation period of 1 h, after which mice were injected intraperitoneally (i.p.) with amphetamine (5 mg/kg) and their locomotion was monitored for another hour. No genotype differences were observed during habituation in total locomotor activity in either males or females (Figure 4e,f,h,i), indicating that the mutation did not alter basal locomotion. Amphetamine administration increased locomotor activity of all groups of mice (Figure 4e–j). However, the amplitude of amphetamine‐induced hyperlocomotion was reduced in male NAc‐GFAP‐CB1 −/− mice as compared to control littermates (Figure 4e–g). On the contrary, no differences in amphetamine‐induced hyperlocomotion were found in female NAc‐GFAP‐CB1 −/− mice as compared to controls (Figure 4h–j), indicating that astroglial CB1 receptors in the NAc are involved in amphetamine‐induced hyperlocomotion in males but not in females. Importantly, these effects were due to the deletion of astroglial CB1 receptors, since the expression of Cre in astrocytes from C57/Bl6 mice had no effect on amphetamine‐induced hyperlocomotion (Figure S4f,g). To exclude the possibility that the effects found in male mice are produced by the marginal expression of the virus in neurons, male CB1 flox/flox mice were injected with an AAV expressing Cre under the control of the Syn promoter (AAV‐Syn‐Cre) in the NAc (NAc‐Syn‐CB1 −/−), which drives the neuronal expression of Cre (Figure S4a,b). FISH to detect CB1 receptor mRNA showed that CB1 expression is virtually abolished in the NAc of NAc‐Syn‐CB1 −/− mice (Figure S4c,d). Importantly, these mice displayed no differences in amphetamine‐induced hyperlocomotion as compared to control mice (Figure S4e), indicating that neuronal CB1 receptors in the NAc do not contribute to this behavior. Altogether, these results show that astroglial CB1 receptors in the NAc are involved in the locomotor effects of amphetamine in male but not in female mice.
2.5Sex‐Dependent Impact of Amphetamine on t‐LTD
The data collected so far indicate that t‐LTD is equally expressed in male and female mice and it is fully dependent on astroglial CB1 receptors in both sexes. However, only male NAc‐GFAP‐CB1 −/− mice display a reduced hyperlocomotor response to the psychostimulant amphetamine. These results suggest that the interaction between amphetamine and CB1 receptors may be different in males and females. Thus, we next investigated whether the psychostimulant produces a differential effect on NAc t‐LTD in male as compared to female mice. To directly test this, we first treated in vivo male and female mice with amphetamine to then analyze t‐LTD in NAc slices (Figure 5a). To avoid confounding effects of the acute presence of the drug in the brain, the in vivo application of amphetamine was performed 24 h before the sacrifice and slice preparation. Strikingly, this treatment abolished t‐LTD in slices from male mice, but had no effect on slices from females (Figure 5b,c). Interestingly, this effect in male slices was accompanied by a significant reduction of the frequency, but not the amplitude, of spontaneous excitatory postsynaptic currents (sEPSCs) in NAc neurons (Figure 5d). However, no such effect was observed in slices from female mice (Figure 5e). This suggests that in vivo amphetamine treatment reduces synaptic inputs in male mice, thereby possibly occluding the induction of t‐LTD in the NAc. Importantly, this effect was fully absent in GFAP‐CB1 −/− mice (Figure 5f,g), showing that astroglial CB1 receptors play a necessary role in male regulation of synaptic plasticity in the NAc.
To better investigate the acute mechanisms of the sex‐dependent effect of amphetamine, we then turned to direct applications of the drug onto NAc slices. Amphetamine had a similar effect on synaptic transmission in slices obtained from male and female mice (Figure 6a). However, it fully reproduced the results obtained with the in vivo treatment, with amphetamine abolishing t‐LTD in NAc slices from male mice, but having no effect on slices from females (Figure 6b). Amphetamine effects on synaptic transmission and t‐LTD are mediated by a convergent cellular mechanism involving astrocyte activation and purine release that activates A1 receptors (Corkrum et al. 2020). Therefore, it is possible that amphetamine occludes t‐LTD in males by promoting the saturation of these receptors. To test this, we assessed the synaptic effects of adenosine before and after bath applications of amphetamine. In slices obtained from male mice, adenosine puffs induced a transient decrease of synaptic transmission under control conditions, but it had no effect in the presence of amphetamine (Figure 6c,d). On the contrary, adenosine could still decrease synaptic transmission in the presence of amphetamine in slices from female mice (Figure 6e,f). Importantly, adenosine effects were completely blocked by CPT (2 μM), an adenosine A1 receptor antagonist (Figure S5), indicating that the inhibitory effects of adenosine in the synapse in the NAc are fully mediated by this receptor. These data suggest that amphetamine induces the saturation of adenosine A1 receptors in males but not in females, possibly underlying the sex‐dependent differential effects of the psychostimulant.
Next, we investigated whether these sex‐dependent differences were due to a distinct adenosine receptor sensitivity. To test this, we assessed the synaptic effects of increasing concentrations of adenosine in slices obtained from male and female mice. Increasing doses of adenosine progressively reduced the amplitude of evoked excitatory synaptic transmission in NAc slices (Figure 7a,b). This effect was paralleled by an increase in paired‐pulse ratio (PPR, Figure 7c), indicating a presynaptic effect of the purine. We found that the lower doses of adenosine (1–10 μM) exerted a higher effect in males as compared to females (Figure 7a,b), while reaching similar effects at the highest doses (25–100 μM), suggesting that the differences found are due to a differential sensitivity of the receptor rather than a differential expression. Furthermore, we evaluated the levels of A1 receptor protein in the NAc using western blot and found no differences in receptor expression between male and female mice (Figure S6). Importantly, the A1 receptor antagonist CPT (2 μM) equally increased synaptic transmission in males and females (Figure 7d–f), indicating that the adenosine tone is similar in both sexes. These data suggest that NAc neurons from male mice are more sensitive to adenosine than female ones.
Altogether, these results show that male mice differ from females in their sensitivity to the occluding effects of amphetamine on t‐LTD, likely due to the fact that their NAc presynaptic adenosine receptors are more saturable than those in females.
3Discussion
The present results show that astroglial CB1 receptors are key mechanistic elements determining the sex specificity of amphetamine locomotor effects. By inducing purine release and activation of A1 receptors, astroglial pmCB1 receptors mediate t‐LTD in the NAc core. The intercellular mechanism proposed for this form of synaptic plasticity is analogous in male and female mice; however, the psychostimulant amphetamine abolishes t‐LTD in the NAc of male but not of female mice. Consistently, NAc astroglial CB1 receptors are necessary for the hyperlocomotor effect of amphetamine in male mice but not in females. This suggests that even though astroglial CB1 receptors in males and females mediate an equivalent form of t‐LTD in the NAc core, amphetamine exerts a sex‐specific impact on CB1‐dependent synaptic plasticity and behavior.
Time‐dependent sequential pre‐ and post‐synaptic stimulation is known to induce t‐LTP in the hippocampus and the cortex (Brzosko et al. 2019) and t‐LTD in the striatum (Fino et al. 2005) and the NAc (Bosch‐Bouju et al. 2016). Whereas t‐LTD in the NAc had been previously linked to the ECS (Bosch‐Bouju et al. 2016), the cellular and subcellular location of the CB1 receptors involved therein remained unknown. Our data demonstrate that CB1 receptors located in astrocytes, but not in neurons, are responsible for the induction of t‐LTD in the NAc core. Recently, an additional layer of complexity in the study of ECS‐regulated functions was added by the discovery of functional CB1 receptors associated with mitochondrial membranes (mtCB1) in both neurons and astrocytes (Benard et al. 2012; Gutierrez‐Rodriguez et al. 2018; Hebert‐Chatelain et al. 2016; Jimenez‐Blasco et al. 2020). Astroglial mtCB1 receptors play a key role in the regulation of cellular Ca2+ dynamics required to modulate synaptic integration in the hippocampus (Serrat et al. 2021) and in the regulation of glucose metabolism (Jimenez‐Blasco et al. 2020). However, here we found that t‐LTD is mediated by pmCB1 receptors, suggesting that the plasma and the mitochondrial membrane pools of CB1 receptors are involved in different regulatory mechanisms.
The timing of the pre‐ and post‐synaptic activity tightly controls the direction and magnitude of the synaptic changes observed after STDP induction (Brzosko et al. 2019). Our results demonstrate that astrocytes in the NAc core play a key role in this tight regulation, being a third element in the control of STDP by bi‐directionally communicating with the pre‐ and post‐synaptic neurons. Our results show that astrocytes respond to endocannabinoids released during t‐LTD induction with intracellular Ca2+ increases, followed by the release of purines acting at A1 receptors likely sitting at pre‐synaptic compartments. This mechanism of action differs from those found in the cortex and the hippocampus, which involve NMDA receptor signaling (Andrade‐Talavera et al. 2016; Min and Nevian 2012). Interestingly, a previous report found that astrocytes in the NAc core can release purines to induce a transient decrease in synaptic transmission (Corkrum et al. 2020). Here we show that the coincidence of this purinergic activity with neuronal stimulation leads to long‐term changes in synaptic transmission. Thus, our results assign a central role to astrocytes in the active regulation of long‐term synaptic plasticity in the NAc.
Amphetamine is known to regulate synaptic plasticity in brain regions related to reward processing, such as the NAc (Li and Kauer 2004), the VTA (Ahn et al. 2010; Jones et al. 2000; Medrano et al. 2021) and the cortex (Xu et al. 2010). However, these studies were conducted on male animals, in mixed cohorts, or omitting to state the sex of subjects (Ahn et al. 2010; Jones et al. 2000; Li and Kauer 2004; Xu et al. 2010), thereby missing the identification of potential sex‐dependent differences. Notably, here we found that t‐LTD in the NAc core is similar between sexes under physiological conditions, but it is blocked by amphetamine in male mice only. Amphetamine depresses synaptic transmission in the NAc core by increasing dopamine levels that over‐activate astrocytes, which then release purines that depress synaptic transmission by acting at A1 receptors (Corkrum et al. 2020). We found that this mechanism of action overlaps with the induction of t‐LTD and may underlie the differences found between males and females. Accordingly, our results show that this effect of amphetamine is sufficient to induce the saturation of A1 receptors and thereby occlude t‐LTD in males but not in females. Furthermore, these sex‐specific effects are due to the fact that adenosine receptors in males are more sensitive to the presynaptic depressive actions of their ligand. Studies using positron emission tomography (PET) found higher A1 receptor availability in the brain of women as compared to men (Hohoff et al. 2020; Pierling et al. 2021), suggesting that sex differences may exist in the expression levels of the A1 receptor. However, we found no differences in the expression of this receptor in the NAc of male and female mice, indicating that the effects observed are probably due to a higher sensitivity of the receptor in males than in females. Noteworthy, while the sex‐differential effects of adenosine were found in the presynapse, we have measured the expression levels of A1 receptors in the entire NAc. Thus, it is possible that potential differences in presynaptic levels of A1 receptors have been confounded by the methodology used.
Sex differences have been described for many effects of drugs of abuse (Becker and Chartoff 2019; Becker and Hu 2008; Becker and Koob 2016; Cornish and Prasad 2021; Torres 2022). Yet, most studies on the mechanisms of action of psychostimulants have been conducted in males, while research in females has been neglected. The ECS has been involved in the responses to psychostimulants and, in particular, to the acute effects of amphetamine, the expression of behavioral sensitization and relapse in males (Bonm et al. 2021; Corbille et al. 2007; Thiemann, Di Marzo et al. 2008; Thiemann, van der Stelt et al. 2008). Here, we found that astroglial CB1 receptors in the NAc are involved in amphetamine‐induced hyperlocomotion in male mice, but not in female mice. Constitutive CB1 −/− male mice express lower amplitudes of amphetamine‐induced hyperlocomotion (Corbille et al. 2007; Thiemann, van der Stelt et al. 2008), as we have observed in NAc‐GFAP‐CB1 −/− male mice. Interestingly, rescuing CB1 in MSNs of CB1 −/− mice is not sufficient to rescue amphetamine‐induced hyperlocomotion (Bonm et al. 2021), leaving the site of action of the ECS‐dependent control of amphetamine effects unknown (Bonm et al. 2021). In this context, it is important to note that our data show that global neuronal deletion of CB1 receptors in the NAc does not alter the effects of amphetamine, indicating that local MSNs or interneurons are not involved in this interaction. Accordingly, we show here that CB1 expression in astrocytes is necessary for the full development of amphetamine‐induced hyperlocomotion in male mice, but not in female mice. Additionally, astrocyte Ca2+ activity and astroglial D1 receptors are necessary for amphetamine‐induced hyperlocomotion in both males and females (Corkrum et al. 2020). Thus, while astrocyte activity is necessary for this behavioral effect of amphetamine in both sexes, the regulatory effects of ECS on psychostimulants are different depending on the sex. Interestingly, repeated pharmacological activation of CB1 receptors in females results in higher receptor desensitization and downregulation (Farquhar et al. 2019), and leads to an enhanced amphetamine sensitization (Lee et al. 2014). This suggests that the regulation of the ECS and its interaction with psychostimulants is fundamentally different between males and females. Importantly, it should be noted that sex hormones, like estradiol, regulate the behavioral responses to certain psychostimulants (Zovkic and McCormick 2019) and can interact with CB1 receptor signaling (Potier et al. 2023), suggesting that these interactions might play a role in the sex‐dependent effects of psychostimulants.
In summary, the present results show that astroglial CB1 receptors participate in synaptic plasticity in the NAc core by increasing cytosolic Ca2+ levels and inducing adenosine/ATP release. Even though this intercellular crosstalk is present in both males and females, its regulation by amphetamine occurs through sex‐specific mechanisms. Thus, amphetamine blocks t‐LTD in males but not in females, likely by promoting the saturation of adenosine A1 receptors. Consistently, astroglial CB1 receptors in the NAc participate in the behavioral responses to amphetamine in males but not in females. Understanding the regulation of synaptic plasticity in the NAc will shed light on the physiological processes that underlie the control of locomotion and the pathological effects of drugs of abuse. Our results show that astroglial CB1 receptors in the NAc play a critical role in this regulation of synaptic plasticity, as well as in behavior, in a sex‐specific manner, and demonstrate that the interaction between the ECS and amphetamine is fundamentally different in male and female animals.
4Materials and Methods
4.1Animals
All animal protocols were in accordance with the Guidelines for the Animal Care and Use and the European Communities Council Directive of September 22nd 2010 (2010/63/EU, 74) and approved by the French Ministry of Agriculture and Education.
Mice were housed in collective cages (maximum 8 mice per cage) under a 12–12 h light–dark cycle (light starting at 7:00 am) with ad libitum food and water and a controlled temperature of 21°C ± 2°C. All experiments were performed during the light period. Inbred constitutive and conditional CB1 receptor mutant male and female mice were used. Unless stated otherwise, experiments were performed in CB1 flox/flox mice (6–12 weeks old for electrophysiology and 10–13 weeks old for behavior) that carry a “floxed” version of the CB1 gene (Marsicano et al. 2003). The Nex‐CB1 −/− mouse line, carrying a conditional deletion of the CB1 gene in forebrain glutamatergic neurons under the control of a Nex‐Cre recombinase, was generated and bred as previously described (Bellocchio et al. 2010; Monory et al. 2006). For the generation of the GFAP‐CB1 −/− mouse line, CB1flox/flox mice were crossed with GFAP‐CreERT2 mice (Hirrlinger et al. 2006) using a three‐step backcrossing procedure to obtain CB1 flox/flox; GFAP‐CreERT2 and CB1 flox/flox littermates (Han et al. 2012), called GFAP‐CB1 −/− mice and WT littermates, respectively. GFAP‐CB1 −/− mice were treated with tamoxifen to induce the CreERT2‐dependent CB1 gene locus excision and achieve the deletion of astroglial CB1. GFAP‐CB1 −/− mice and WT littermates (> 6 weeks old) were injected for 8 consecutive days with tamoxifen (1 mg, i.p.) dissolved in 90% sesame oil and 10% ethanol to a final concentration of 10 mg/mL. The animals were used 4–8 weeks after tamoxifen treatment. The DN22‐CB1‐KI mouse line, a knock‐in mouse line replacing the wild‐type CB1 gene with a truncated form of the CB1 gene lacking the first 22 amino acids that reduces its mitochondrial‐associated expression (Benard et al. 2012; Hebert‐Chatelain et al. 2016) was generated and bred as previously described (Soria‐Gomez et al. 2021). The Ai6 mouse line, a reporter mouse line that expresses GFP when the Cre protein is present (Madisen et al. 2010) was used in combination with GFAP‐CRE AAV injection as described later. The respective wild‐type littermates of all lines were used as controls.
Mice were genotyped by PCR on tail DNA using the following primers: CB1‐flox locus (5′GCTGTCTCTGGTCCTCTTAAA and 5′GGTGTCACCTCTGAAAACAGA), Nex‐CRE locus (5′TCTTTTTCATGTGCTCTTGG and 5′CGCGCCTGAAGATATAGAAGA), GFAP‐CreERT2 locus (5′‐CCT GGA AAA TGC TTC TGT CCG and 5′‐CAG GGT GTT ATA AGC AAT CCC), DN22‐CB1‐KI locus (5′TGTGTGAATCGATAGTACTAAC and 5′CCTGGCACCTCTTTCTCAGT).
4.2Adeno‐Associated Viruses (AAV)
The viral vector AAV8‐hGFAP‐mCherry‐Cre (named here as AAV‐GFAP‐Cre) was purchased from the University of North Carolina (UNC) vector core, and the viral vectors ssAAV9/2‐hGFAP‐hHBbI/E‐GCaMP6f‐bGHp(A) (name here as AAV‐GFAP‐GCaMP6f), v146‐8 ssAAV‐hSyn1‐chI‐EGFP‐2A‐iCre‐WPRE‐SV40p(A) (named here as AAV‐Syn‐Cre) and v133‐8 ssAAV‐hSyn1‐chI‐mCherry‐WPRE‐SV40p(A) (control for AAV‐Syn‐Cre and named here as AAV‐Syn‐mCherry) were purchased from the University of Zurich. The viral vector AAV8‐hGFAP‐EGFP (named here as AAV‐GFAP‐EGFP and used as control for the viral vector AAV‐GFAP‐Cre) was subcloned using standard molecular cloning techniques previously used in the lab (Hebert‐Chatelain et al. 2016). The resulting vector was produced by PEI‐mediated transfection of HEK293 cells together with the AAV8‐serotype‐packaging plasmids (Hammond et al. 2017). The viruses were then purified by iodixanol density gradient and tittered as previously described (McClure et al. 2011). Virus titers were 1010 genomic copies per ml for AAV‐GFAP‐Cre and AAV‐GFAP‐EGFP and 1011 genomic copies per ml for AAV‐GFAP‐GCaMP6f, AAV‐Syn‐Cre, and AAV‐Syn‐mCherry.
4.3Stereotaxic Surgeries
Mice (> 6 weeks old) were injected subcutaneously with the analgesic buprenorphine (0.05 mg/kg, Buprecare) and anesthetized with isoflurane (5%). Anesthesia was then maintained with 2%–2.5% isoflurane. A heating pad was positioned underneath the animal to keep the body temperature at 37°C during all steps of the surgery. Eye dehydration was prevented by topical application of ophthalmic gel. The skin above the skull was shaved with a razor and disinfected with betadine. Mice were placed into a stereotaxic apparatus (Model 900, David Kopf Instruments) with mouse adaptor and lateral ear bars. Local analgesia was applied with lidocaine (0.1 mL at 0.5%, Lidor) before incision. The skin and the periosteum were carefully separated and kept hydrated with saline during the duration of the surgery. Two holes were drilled in the skull. Viral vectors (1 μL) were injected bilaterally using a Hamilton syringe attached to a 29‐gauge needle at a rate of 0.2 μL/min. The viral constructs were targeted to NAc astrocytes of CB1 flox/flox mice using the following coordinates (from bregma): anterior–posterior +1.50 mm; medial‐lateral +/− 0.75 mm; dorsal‐ventral −4.50 mm. After injections, the virus was allowed to diffuse for at least 5 min before the syringe was withdrawn. Mice were sutured, injected intraperitoneally with the nonsteroidal anti‐inflammatory meloxicam (5 mg/kg, Metacam) for analgesia, and left in an incubator at 37°C until they were fully awake. Mice were treated with meloxicam (5 mg/kg, Metacam) for 2 days after surgery, and the proper recovery and weight gain were monitored for 4–6 days after surgery. Mice were used for experiments ≥ 4 weeks after stereotaxic surgeries.
4.4Preparation of Acute Brain Slices
Animals for electrophysiology experiments were taken from the collective home cage. In specific experiments, they were placed for 1 h in a new cage, treated with amphetamine (5 mg/kg) or saline, and placed back in the collective home cage after another hour. Mice were sacrificed by decapitation, and the brain was rapidly extracted and placed in ice‐cold cutting solution containing (in mM): KCl 2.5, NaH2PO4 1.25, NaHCO3 26, CaCl2 0.2, MgCl2 12, sucrose 180, and glucose 11 (pH = 7.3–7.4). 350 μm thick coronal slices containing the NAc core were made with a vibratome (Leica VT 1200S) and put in artificial cerebral spinal fluid (ACSF) containing (in mM): NaCl 123, KCl 2.5, NaH2PO4 1.25, NaHCO3 26, CaCl2 2.5, MgCl2 1.3, and glucose 11 (pH = 7.3–7.4). Slices were incubated in ACSF at 34°C for 30 min and then at room temperature for > 30 min. Both cutting solution and ACSF were constantly oxygenated with carbogen (95% O2/5% CO2). Slices were placed in an immersion recording chamber and superfused (2 mL/min) with oxygenated ACSF and visualized with an upright Olympus microscope with infrared illumination for electrophysiology recordings or a Femtonics FEMTO3D‐RC 2‐photon microscope for Ca2+ imaging.
4.5Electrophysiology
Whole‐cell electrophysiological recordings from MSNs and astrocytes in the NAc core were made in voltage‐clamp configuration. For neuronal recordings, patch‐clamp electrodes had resistances of 4–6 MΩ when filled with the internal solution containing (in mM): K‐gluconate 135, KCl 10, HEPES 10, EGTA 1, MgCl2 2, CaCl2 0.3, ATP‐Mg+2 3, GTP‐Na+ 0.3, and phosphocreatine 7 (pH = 7.3). For astrocyte recordings, patch‐clamp electrodes had resistances of 6–10 MΩ when filled with the internal solution containing (in mM): KMeSO4 100, KCl 50, HEPES‐K 10, and ATP‐Na+2 4 (pH = 7.3). GDPßS (10 mM) was included in the astrocyte patch pipette to prevent G protein‐mediated intracellular signaling in astrocytes. Astrocytes were patched for at least 15 min to allow the dialysis of GDPßS throughout the gap‐junction connected astrocyte network. Then, an MSN within the same field of view was patch‐clamp recorded to monitor EPSCs (Corkrum et al. 2020).
Recordings were obtained with an amplifier Multiclamp 700B (Molecular devices). Membrane potential was held at −80 mV for both neurons and astrocytes. Series and input resistances were monitored throughout the experiment using a −5 mV pulse, and cells were discarded when they changed > 20%. Signals were fed to a PC through a DigiData 1440A interface board (Molecular devices). Signals were filtered at 1 KHz and acquired at a 10 KHz sampling rate. The pCLAMP 10.4 (Axon instruments) was used for stimulus generation, data display, acquisition, and storage.
4.6Synaptic Stimulation and STDP Induction
Synaptic currents were evoked using bipolar theta capillaries filled with ACSF placed in the NAc core. Paired pulses (2 ms duration with 50 ms interval) were continuously delivered at 0.33 Hz using an isolation unit DS2A‐Mk. II (Digitimer Ltd). Excitatory post‐synaptic currents (EPSCs) were isolated using picrotoxin (50 μM) and CGP5462 (1 μM) to block GABAA and GABAB receptors, respectively.
During STDP induction, neurons were recorded in current‐clamp. For STDP induction, a pre‐post protocol was used: a synaptic stimulation was applied followed by a short depolarization that elicited a single action potential (≃10 ms time window between the synaptic stimulation and the action potential). 100 pre‐post pairings were applied at 1 Hz. Baseline mean EPSC amplitude was obtained by averaging mean values recorded within the 10 min of baseline. Then, EPSC amplitudes were grouped in 1 min time bins and normalized to baseline. The presence of t‐LTD was established by comparing the normalized EPSCs recorded during baseline and 45 min after STDP induction.
For exogenous adenosine puff application (250 μM), a borosilicate glass pipette containing the drug was placed over the NAc core, and adenosine was applied with a pressure pulse (0.5 bar) during the STDP induction protocol (Figure 2) or for 5 s (Figure 6). For the adenosine dose–response (Figure 7), increasing concentrations of adenosine were bath applied.
For experiments assessing the effects of puff‐applied adenosine on synaptic transmission, EPSC amplitudes were grouped in 15 s time bins and EPSC obtained 2 min before and 30 s after amphetamine application were compared. For experiments investigating the effects of bath‐applied adenosine on synaptic transmission, EPSC amplitude 5 min before and 10 min after adenosine application were compared. For experiments studying the effects of amphetamine (10 μM) or CPT (2 μM) on synaptic transmission, EPSC amplitudes were grouped in 1 min time bins and EPSC obtained 5 min before and 25 min after amphetamine or CPT application were compared.
4.7Ca2+ Imaging
Changes in Ca2+ levels in astrocytes located in the NAc core were monitored by 2‐photon microscopy using the Ca2+ indicator GCaMP6f. Slices were visualized with an sCMOS camera (Hamamatsu Orca Flash 2.8), a stimulation electrode was put in the NAc core, and MSNs were patch‐clamped to induce STDP as described above. GCaMP6f 2‐photon acquisitions were performed using a Femtonics FEMTO3D‐RC microscope with a 20X water immersion objective (Olympus). The sample was illuminated using a femtosecond laser Coherent Chameleon Vision 2 and scanned with a galvanometric scanner. Images were acquired every 0.55–0.66 s with a 0.3–0.6 μm pixel size.
Image analysis was done using a custom ImageJ (NIH) macro (available on Zenodo: DOI: 10.5281/zenodo.11922155) (Serrat et al. 2021). Astrocytes were considered to undergo a Ca2+ event when fluorescence increased more than 4 times the standard deviation of background fluorescence. Ca2+ events were detected and analyzed in 3D (x, y, t). For each Ca2+ event, the amplitude, frequency, duration, and spreading coefficient (max x distance * max y distance) were calculated.
For experiments assessing astrocyte responses to WIN and ATP, astrocytes were imaged using an upright Olympus microscope illuminated with a CooLED, and images were acquired at 10 Hz. Image analysis was performed using a custom MATLAB App (available on Zenodo: 10.5281/zenodo.14999462 and at the GitHub page tommigara/Mandrake) that was used for the detection of active regions of interest (ROIs), extraction, correction, and normalization of fluorescence traces, and Ca2+ event detection. Active ROIs were manually inspected, and traces with no peaks detected were excluded. The time of occurrence of a Ca2+ event was determined at the onset of the Ca2+ event. Ca2+ events recorded from 30 to 100 active ROIs within the field of view were grouped into 10‐s bins. Each ROI was assigned a value of 0 for bins without a response and 1 for bins in which a Ca2+ event occurred. Finally, the Ca2+ event probability was calculated as the number of Ca2+ elevations in each time bin divided by the total number of active ROIs (Covelo and Araque 2018).
4.8Amphetamine‐Induced Hyperlocomotion
To assess the locomotor effects induced by amphetamine, we used Actimeter apparatus (Imetronic, Marcheprime) (Mariani et al. 2023) consisting of an aluminum ventilated rack covered with PVC plates. The rack contains 8 single boxes (actimetry cages), 30 × 16 × 11cm. Each box is equipped with infrared sensors to detect locomotor activity as beam breaks. The rack is connected to an electronic interface that receives information from the infrared sensors and allows communication with the computer.
Mice were habituated to the testing room for at least 45 min. After this period, mice were located in a single box and their basal locomotion was recorded for 1 h (habituation). Then, mice received an amphetamine injection (5 mg/kg, i.p.), and their locomotion was recorded for 1 h (test).
The software Imetronic stores data as beam‐breaks/5 min. The locomotor activity is expressed as total activity, corresponding to the total activity during 1 h of recording, or as Δactivity/min, calculated by subtracting the activity/min during the last 15 min of the habituation phase (range of time in which the locomotion induced by the novel environment exploration is stabilized) from the activity/min recorded during the total post‐injection period. Outliers (1 out of 19 male WT mice) and mice that were unresponsive to amphetamine (locomotion activity was higher during habituation than after amphetamine injection: 3 out of 19 male WT mice, 2 out of 12 female WT mice, 1 out of 22 male NAc‐ NAc‐GFAP‐CB1 −/− mice and 3 out of 13 female NAc‐GFAP‐CB1 −/− mice) were excluded from this study.
4.9Perfusion and Cryosection
Mice were deeply anesthetized with a fresh mix of pentobarbital (300 mg/kg) and lidocaine (30 mg/kg) and transcardially perfused with 20 mL of phosphate‐buffered solution (PBS 0.1 M, pH 7.4) followed by 50 mL of cold 4% paraformaldehyde (PFA, Sigma, HT501128‐4 L). After perfusion, the brains were extracted and post‐fixed in the same fixative solution overnight at 4°C. Brains were then embedded in sucrose 30% (Sigma, S0389) for 3 days, frozen in isopentane (Sigma, M32631) and stored at −80°C. Serial brain coronal cryosections (30 μm) of the NAc were cut in a cryostat (Leica, CM1950S) and collected in anti‐freeze solution (in 0.2 M phosphate buffer H2PO4 −/HPO4 −2: 20% v/v glycerol; 30% v/v Ethylene glycol). The samples were stored at −20°C until further use.
4.10Immunofluorescence
Two immunofluorescences were performed in order to spot Cre expression and neurons or astrocytes. Therefore, cryosections were permeabilized and blocked for 1 h at RT incubation with blocking solution (in PBS 1X: 10% donkey serum; 0.3% triton X‐100). Afterward, slices were incubated overnight at 4°C with a mix of primary antibodies. Mix 1: mouse anti‐NeuN (1:500; Millipore, MAB377) and rabbit anti‐DSred (1:1000; Takarabio, 632,496); or Mix 2: chicken anti‐GFAP (1:500, Abcam ab4674), mouse anti‐S100β (1:500; Sigma, S2532) and rabbit anti‐DSred (1:1000; Takarabio, 632,496). After some washes with PBS 1X, the tissue was incubated for 2 h at RT with a mix of secondary antibodies: Mix 1: donkey anti‐mouse AF 488 (1:500, Invitrogen) and goat anti‐rabbit AF 546 (1:500, Invitrogen); or Mix 2: goat anti‐chicken AF 647 (1:500, Invitrogen), donkey anti‐mouse AF 488 (1:500, Invitrogen) and goat anti‐rabbit AF 546 (1:500, Invitrogen). All the antibodies were prepared in blocking solution, and nuclei were stained with DAPI (1:20000 prepared in PBS). Sections were mounted, dried, coverslipped, and imaged with a Leica SP5 multi‐photon microscope.
4.11Free‐Floating In Situ Hybridization and Immunofluorescence
Endogenous peroxidases from the cryosections were inactivated with 3% H2O2 diluted in PBS‐DEPC for 30 min at RT. Then, slices were incubated for 20 min at RT in 0.2 mM HCl‐DEPC followed by an acetylation step of 10 min with 0.1 M triethanolamine‐HCl (pH 8) and 0.25% acetic anhydride. The riboprobe CB1‐DIG antisense (1:1000) (Marsicano and Lutz 1999) was used to detect mouse CB1 mRNA, and the riboprobe CB1‐DIG sense (1:1000) (Marsicano and Lutz 1999) was used as a negative control. Slices were hybridized overnight at 60°C with CB1‐DIG sense or antisense probes. Afterward, sections were incubated for 2 h at RT with sheep anti‐DIG POD antibody (1:1500, Roche 11,207,733,910) followed by a 10 min incubation with TSA Plus FITC (1:100, Akoya biosciences, NEL741001KT). Afterward, sections were permeabilized and blocked for 1 h at RT incubation in a blocking solution. Afterward, slices were incubated overnight at 4°C with a mix of primary antibodies: mouse anti‐NeuN (1:500, Millipore, MAB377) and rabbit anti‐GFP (1:1000, Chemicon, A11122). After some washes, sections were incubated for 2 h at RT with the secondary antibody donkey anti‐mouse AF 555 (1:500, Invitrogen A31570) and goat anti‐rabbit AF 488 (1:500, Invitrogen A11008). All the antibodies were prepared in blocking solution, and nuclei were stained with DAPI (1:20000 in PBS). Finally, sections were mounted, dried, coverslipped, and imaged with a Leica SP5 multi‐photon microscope. All DEPC solution has a concentration of 5% v/v DEPC.
4.12Western Blot
Naive male and female CB1 flox/flox mice were euthanized by cervical dislocation, the brain was extracted and the NAc was rapidly dissected using a coronal brain matrix, and processed as described before (Mariani et al. 2023; Vallee et al. 2014). Tissue samples were immediately frozen in dry ice and stored at −80°C. Samples were then homogenized using the Tissue Lyser (Quiagen, Hilden, Germany) in lysis buffer (0.05 M Tris–HCl pH 7.4, 0.15 M NaCl, 0.001 M EDTA, 10% Glycerol, 1% Triton X‐100) containing protease inhibitors (Roche, Basel, Switzerland). After 10 min of incubation at 4°C, samples were centrifuged (12,500 rpm, 10 min, 4°C). Bradford protein assay was performed on the solubilized fractions for measuring the protein content. Samples were then mixed with denaturing 4× Laemmli loading buffer (250 mM Tris–HCl, 40% Glycerol, 8% SDS, 5% β‐Mercaptoethanol, 0.2% Bromophenol blue) and heated at 37°C for 30 min. Samples were analyzed on 4%–20% precast polyacrylamide gels (Bio‐Rad, Hercules, California) and transferred onto PVDF membranes 0.45 μm (Merk Millipore, Burlington, MA). Membranes were blocked in a mixture of Tris‐buffered saline and polysorbate 20 (20 mM Tris–HCl pH 7.6, 150 mM NaCl, 0.05% Tween 20) containing 5% non‐fat dry milk for 1 h at RT. For the immunoblotting, the antibodies against Adora1AR (AAR‐006; 1:500, ON at 4°C, Alomone Labs, Israel) and tubulin (#3873; 1:5000, 1 h RT, Cell Signaling Technology, Danvers, MA) were used. The signal was detected with HRP‐linked antibodies (1:2000, Cell Signaling Technology, Danvers, MA) and visualized by enhanced chemiluminescence detection (Clarity Western ECL Substrate, Bio‐Rad, Hercules, California). Images were acquired on ChemiDoc Touch (Bio‐Rad, Hercules, California) and analyzed using the Image Lab software (Bio‐Rad, Hercules, California).
4.13Statistics
To assess the presence or absence of synaptic plasticity in each experimental condition, a paired student's t‐test was used comparing baseline with 45 min post‐STDP induction. As previously reported, few neurons underwent long‐term potentiation (3 out of 64 cells recorded from WT mice in control conditions) (Ji and Martin 2012) and were excluded when they were statistically outliers (4 out of 177). The presence of outliers was determined using a ROUT test (Q = 1%). The effects of adenosine on t‐LTD in GFAP‐CB1 −/− mice and behavioral effects of amphetamine were compared using two‐way ANOVA with Sidak's post hoc test. A full report of the exact p values found for each statistical test can be found in Table S1. Data are expressed as mean ± s.e.m. (standard error of the mean). Statistical thresholds used were *p < 0.05, **p < 0.01, and ***p < 0.001.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Acknowledgments
We would like to thank Ruby Racunica and all the personnel of the Animal Facilities of the Neurocentre Magendie for mouse care. We also thank the genotyping platform of the Neurocentre Magendie for the help in the experiments. We thank Sebastien Marais and all the personnel of the Bordeaux Imaging Center for their support with two‐photon microscopy. We thank all the members of Marsicano's lab for useful discussions and for their invaluable support. This study was funded by Inserm, EU–FP7 (PAINCAGE, HEALTH‐603191), European Research Council (Endofood, ERC‐2010‐StG‐260515; CannaPreg, ERC‐2014‐PoC‐640923; Micabra, ERC‐2017‐AdG‐786467), Fondation pour la Recherche Medicale (DRM20101220445), Region Aquitaine, Agence Nationale de la Recherche (LABEX BRAIN ANR‐10‐LABX‐43) and Human Frontiers Science Program to G.M.; and Human Frontiers Science Program, University of Bordeaux Investments for the Future program (IdEx), the Brain and Behavior Research Foundation (Narsad), and Ministerio de Ciencia, Innovación y Universidades (MICIU), Agencia Estatal de Investigación (AEI) and European Social Fund Plus ESF+ (RYC2022‐035546‐I) to A.C.
Data Availability Statement
The data obtained in this study are available upon reasonable request to the corresponding authors.