WIN55,212-2, a Dual Modulator of Cannabinoid Receptors and G Protein-Coupled Inward Rectifier Potassium Channels

The coupling of cannabinoid receptors, CB1 and CB2, to G protein-coupled inward rectifier potassium channels, GIRK1 and GIRK2, modulates neuronal excitability in the human brain. The present study established and validated the functional expression in a Xenopus laevis oocyte expression system of CB1 and CB2 receptors, interacting with heteromeric GIRK1/2 channels and a regulator of G protein signaling, RGS4. This ex vivo system enables the discovery of a wide range of ligands interacting orthosterically or allosterically with CB1 and/or CB2 receptors. WIN55,212-2, a non-selective agonist of CB1 and CB2, was used to explore the CB1- or CB2-GIRK1/2-RGS4 signaling cascade. We show that WIN55,212-2 activates CB1 and CB2 at low concentrations whereas at higher concentrations it exerts a direct block of GIRK1/2. This illustrates a dual modulatory function, a feature not described before, which helps to explain the adverse effects induced by WIN55,212-2 in vivo. When comparing the effects with other typical cannabinoids such as Δ9-THC, CBD, CP55,940, and rimonabant, only WIN55,212-2 can significantly block GIRK1/2. Interestingly, the inward rectifier potassium channel, IRK1, a non-G protein-coupled potassium channel important for setting the resting membrane voltage and highly similar to GIRK1 and GIRK2, is not sensitive to WIN55,212-2, Δ9-THC, CBD, CP55,940, or rimonabant. From this, it is concluded that WIN55,212-2 selectively blocks GIRK1/2.


Introduction
Cannabinoid receptors (CB) are G protein-coupled receptors (GPCRs) that have attracted broad attention since they are regarded as promising therapeutic targets for multiple pathologies, such as pain, epilepsy, anorexia, Parkinson's disease, and Alzheimer's disease [1]. Cannabinoid receptors are primarily classified as CB1 and CB2. CB1 is one of the most abundant GPCRs in the human brain [2], while CB2 was originally considered to be confined in the peripheral organs of humans [3,4]. Afterward, the expression of CB2 has also been reported in the mammalian healthy central nervous system (CNS) [5][6][7][8] and diseased brain cells from humans [9][10][11]. Ligands of CB1 and/or CB2 (also commonly known as cannabinoids), through their binding to cannabinoid receptors in the brain, produce a wide variety of effects. For instance, activating CB1 generates psychotropic effects including euphoria, enhancement of sensory perception, antinociception, appetite stimulation, and impairment of memory [1]. Activating CB2 produces antinociception, antiinflammation, and neuroprotection [1]. In the brain, signaling pathways of activating CB1 and/or CB2 have been revealed involving the phosphorylation of receptors by G protein receptor kinases and subsequently an association with β-arrestin1 or β-arrestin2 resulting in the desensitization and internalization of cannabinoid receptors [2,12]. Besides, activation of CB1 and/or CB2 elicits the dissociation of the βγ subunits (Gi/o(βγ)) of pertussis toxin-sensitive G proteins (Gi/o) from the α subunits (Gi/o(α)) [2,12,13]. Gi/o(α) expression Xenopus oocyte system. This assay provides a straightforward and efficient methodology for examining cannabinoid-stimulated effects. Moreover, the results may partly explain the contradictory effects produced by WIN55,212-2 in vivo.

Isolation of Xenopus Oocytes
All procedures for the use and handling of adult female Xenopus laevis frogs (CRB Xénopes, Rennes, France) were approved by the Animal Ethics Committee of the KU Leuven (Project No. P186/2019) following regulations of the European Union (EU) concerning the welfare of laboratory animals as declared in Directive 2010/63/EU. Oocytes were isolated from ovarian tissue surgically removed during hypothermia and 0.1% tricaine (Sigma-Aldrich Chemical, St. Louis, MO, USA) induced anesthesia. After recovery from anesthesia, frogs were returned to their tanks in the Aquatic Facility of KU Leuven and monitored daily.
The oocytes were enzymatically defolliculated by collagenase (3 mg/mL) (Sigma-Aldrich Chemical, St. Louis, MO, USA) digestion at 16°C on a rocker platform in a Ca 2+free ND96 solution. Isolated stage V-VI oocytes were then maintained in ND96 solution containing Theofylline and Gentamicin at 16 • C. The ND96 solution was composed of 96 mM NaCl, 2 mM MgCl 2 , 2 mM KCl, 5 mM HEPES, and 1.8 mM CaCl 2 , with a final pH of 7.5.

Electrophysiological Recordings
Macroscopic K + currents through GIRK1/2 were measured using a two-electrode voltage-clamp amplifier (GeneClamp 500B, Axon Instruments, San Jose, CA, USA). Electrodes were fabricated from borosilicate glass tubes (1.14 mm outside diameter, 0.7 mm inside diameter) by a programmable microelectrode puller (PUL-1, World Precision Instruments, Sarasota, FL, USA). Electrodes were filled with 3 M KCl and had tip resistances of 0.8 to 1.5 MΩ. Membrane currents from voltage-clamped oocytes were digitized using a Digidata 1550 low-noise data acquisition system (Axon Instruments, San Jose, CA, USA) and a Dell PC running pCLAMP 10.1 software (Axon Instruments, San Jose, CA, USA).
Oocytes were placed in a 0.2 mL recording chamber continuously perfused with ND96 solution. After electrode impalement and clamping the membrane potential to −90 mV, the perfusion solution was changed to a high K + (HK) solution composed of 96 mM KCl, 2 mM NaCl, 1 mM MgCl 2 , 1.8 mM CaCl 2 , 5 mM HEPES with a final pH of 7.5. The resulting increase in inward K + current represents a 'basal' K + current (I K,basal ) that is following primarily receptor-independent GIRK channel activity. Rapid application and washout of CB agonist WIN 55,212-2 (Sigma-Aldrich Chemical, St. Louis, MO, USA) in HK produced the reversible receptor-dependent K + current (I K,WIN ) and was performed with a controlled perfusion system that rapidly switched the perfusion of two syringe reservoirs (syringe A and syringe B) connected to syringe needles, two-way stopcocks, and tubing. The syringe reservoirs contained HK (syringe A) and HK plus WIN55,212-2 (syringe B). The perfusion system was located next to the oocyte. In addition to the abovementioned −90 mV voltage protocol, a 1-s voltage ramp protocol was also applied from −150 to +60 mV from a holding potential of −20 mV during perfusion.

Electrophysiological Data Analysis
The EC 50 values with corresponding 95% confidence intervals (CIs) were calculated using the Prism 8.0 program (GraphPad, San Diego, CA, USA). Statistical comparisons between the various experimental groups were performed by one-way ANOVA where p < 0.05 was considered significant. Experiments were replicated for three to six oocytes (n = 3-6).

Activation of CB-GIRK1/2-RGS4 Coupling in Xenopus Laevis Oocytes
First of all, we aim to assess the properties of receptor-dependent GIRK1/2 activation elicited by cannabinoid receptors CB1 and CB2. In addition to cannabinoid receptors and GIRK1/2 in this system, RGS4 was also transfected to the oocytes. RGS proteins have been previously reported to maintain the robustness of GPCR-GIRK coupling [29]. Moreover, co-expression with RGS4 was beneficial to the constructs of CB-G(α)-G(βγ) in heterogeneous cells, stimulating the basal activity of cannabinoid receptors [30]. In this study, oocytes co-injected with CB1 or CB2, GIRK1/2, and RGS4 cRNAs were voltageclamped at −90 mV. GIRK1/2 was initially inactivated by ND96, as shown in Figure 1A,B. After exchanging ND96 to HK, basal K + currents, which are known to depend on free G protein βγ-subunits present in the oocytes because of the inherent activity of G proteins [31], were observed and were independent on the level of GIRK1/2 (I K,basal ). In the presence of HK, K + current enhancement was immediately evoked on the application of 1 µM WIN55,212-2 (I K,WIN ) and was reversible via washing out WIN55,212-2 by HK. The receptordependent K + currents represent that GIRK1/2 were activated by CB1 or CB2 coupled specifically to GIRK1/2. As a control, K + currents carried by GIRK1/2 were not significantly enhanced upon the application of 1 µM WIN55,212-2 in oocytes injected with only GIRK1/2 (I K,WIN = 0.00997 ± 0.0157 µA, n = 3).
To determine the intrinsic current (I K,intrinsic ) that was responsible for I K,basal , noninjected oocytes were tested for sensitivity to HK. As a result, I K,intrinsic , which was 0.158±0.0595 µA (n = 5), constituted a negligible proportion of the total I K,basal . Exchanging HK to HK plus 1 µM WIN55,212-2 did not change I K,intrinsic (n = 5), suggesting no effect of WIN55,212-2 on intrinsic oocyte receptors or other endogenous ion channels.
Furthermore, to study whether WIN55,212-2 can activate inward rectification of GIRK1/2 through activating CB1 and CB2, a voltage ramp protocol was applied. Oocytes co-injected with CB1 or CB2, GIRK1/2, and RGS4 cRNAs were subjected to a 1-s voltage ramp protocol from −150 to +60 mV from a holding potential of −20 mV during perfusion. Figure 2 is a representative example of these experiments. Inward rectifying currents, activated under the conditions of HK or HK plus WIN55,212-2, had a reversal potential near 0 mV, as I K,in and I K,out were approximately equal. The 1 µM WIN55,212-2 did not shift the K + equilibrium potential of GIRK1/2, but simply increased the amplitude of inward K + currents induced by HK. Thus, the current enhancement was presumably a result of increased GIRK1/2 channel conductance. medicines 2021, 9, x FOR PEER REVIEW Figure 2 is a representative example of these experiments. Inward rectifying currents, activated under the conditions of HK or HK plus WIN55,212-2, had a reversal potential near 0 mV, as IK,in and IK,out were approximately equal. The 1 μM WIN55,212-2 did not shift the K + equilibrium potential of GIRK1/2, but simply increased the amplitude of inward K + currents induced by HK. Thus, the current enhancement was presumably a result of increased GIRK1/2 channel conductance.  Oocytes co-injected with CB1, GIRK1/2, and RGS4 cRNAs, and (B) oocytes co-injected with CB2, GIRK1/2, and RGS4 cRNAs were subjected to a 1-s voltage ramp protocol from −150 to +60 mV from a holding potential of −20 mV. Inward IK,basal was enhanced by exchanging HK to HK plus 1 μM WIN55,212-2. Typical IK,basal and IK,WIN were recorded from three to six different oocytes.
Biomedicines 2021, 9, x FOR PEER REVIEW 5 of 15 Figure 2 is a representative example of these experiments. Inward rectifying currents, activated under the conditions of HK or HK plus WIN55,212-2, had a reversal potential near 0 mV, as IK,in and IK,out were approximately equal. The 1 μM WIN55,212-2 did not shift the K + equilibrium potential of GIRK1/2, but simply increased the amplitude of inward K + currents induced by HK. Thus, the current enhancement was presumably a result of increased GIRK1/2 channel conductance.

Enhancement and Reduction of Inward K+ Currents by WIN55,212-2 in a CB-GIRK1/2-RGS4 Coupling Oocyte Expression System
We investigated the concentration-response relations of the effects of WIN55,212-2 on the CB-GIRK1/2-RGS4 coupling expressed in Xenopus oocytes. Oocytes co-injected with CB1 or CB2, GIRK1/2, and RGS4 cRNAs were subjected to the voltage ramp protocol abovementioned in "3.1". IK,basal was first allowed to reach a steady-state when perfusing HK. Afterward, HK plus increasing concentrations of WIN55,212-2 (0.01-≈100 μM) was applied to independent oocytes (three oocytes per concentration). Note the progressively Figure 2. Activation of inward rectification of GIRK1/2 in the CB-GIRK1/2-RGS4 coupling in Xenopus oocytes. (A) Oocytes co-injected with CB1, GIRK1/2, and RGS4 cRNAs, and (B) oocytes co-injected with CB2, GIRK1/2, and RGS4 cRNAs were subjected to a 1-s voltage ramp protocol from −150 to +60 mV from a holding potential of −20 mV. Inward I K,basal was enhanced by exchanging HK to HK plus 1 µM WIN55,212-2. Typical I K,basal and I K,WIN were recorded from three to six different oocytes.
Plotting the percentage of current enhancement and reduction (I K,WIN /I K,basal × 100) as a function of the WIN55,212-2 concentration reveals a concentration-dependent increase and subsequent decrease in the response of I K,WIN (Figure 3).
For the increase section of the concentration-response curve, the best-fit indicates an apparent EC 50 value of 0.374 µM (95% CI: 0.276 to 0.504) for activating CB1 (n = 3; Figure 5A), while the best-fit indicates an apparent EC 50 value of 0.260 µM (95% CI: 0.163 to 0.414) for activating CB2 (n = 3; Figure 5B). The EC 50 value for activating CB1 is consistent with previous studies where EC 50 values were shown to range from 0.03 µM to 3 µM depending on the CB1 expression systems [1].

Blockage of GIRK1/2 via High Concentrations of WIN55,212-2
WIN55,212-2 is a well-characterized agonist of CB1 and CB2; thus, it can exert activity by activating cannabinoid receptors. Moreover, low concentrations of WIN 55,212-2 (≤1 µM) can activate receptor-dependent inward K + current in CB1 or CB2 -GIRK1/2-RGS4 coupling oocyte expression system as described above, which is consistent with its agonistic activity. To investigate whether high concentrations of WIN55,212-2 interact with brain-type GIRK1/2, concentrations ranging from 1 to ≈100 µM WIN55,212-2 in HK were applied to oocytes injected with only GIRK1/2 cRNAs. The 1-s ramp protocol was also applied in this experiment. After reaching a steady-state of I K,basal when perfusing HK, the application of 1 µM WIN55,212-2 plus HK still did not cause a significant effect on the inward K + currents (n = 5, data not shown). Nonetheless, as shown in Figures 3 and 4C, the application of higher concentrations (10-≈100 µM) of WIN55,212-2 plus HK to GIRK1/2expressed oocytes caused a reduction of inward K + currents in a concentration-dependent manner, which indicates high concentrations of WIN55,212-2 blocked GIRK1/2. Therefore, the reduction of inward K + currents produced by WIN55,212-2 observed in the results of "3.2" was caused by the blockage of GIRK1/2 in the CB-GIRK1/2-RGS4 coupling expression system. In other words, these results suggest that high concentrations of WIN55,212-2 can block the effect of GIRK1/2 activated by CB1 and CB2, presumably following direct interaction with the GIRK channels.

Comparison of Effects of Other Typical Cannabinoid Members on GIRK1/2
We studied whether ∆9-THC (a partial agonist of CB1 and CB2), cannabidiol (CBD, an antagonist/inverse agonist or a negative allosteric modulator of CB1 and a partial agonist of CB2), CP55,940 (a non-selective agonist of CB1 and CB2), and rimonabant (a selective antagonist of CB1) also interact with GIRK1/2. Among these cannabinoids, ∆9-THC and CBD are derived from the plant Cannabis sativa, commonly known as marijuana. They are phytocannabinoids that have been extensively investigated and are classified as classical cannabinoids at the structural level [34] (Figure 6(B(a),C(a))). The classical group consists of dibenzopyran derivatives that are either cannabis-derived compounds (phytocannabinoids) or their synthetic analogs [34]. However, WIN55,212-2, CP55,940, and rimonabant are chemically synthesized, representing synthetic cannabinoids with three different chemical structures. The structure of WIN55,212-2 is the amino-alkylindole type and CP55,940 belongs to the non-classical type group which consists of bicyclic and tricyclic analogs of ∆9-THC that lack a pyran ring [34] (Figure 6(A(a),D(a))). Rimonabant belongs to the group of other types that have structures completely different from those of the abovementioned cannabinoids ( Figure 6(E(a))). In this experiment, large inward K + currents were also observed when the oocytes expressing GIRK1/2 were perfused with HK. Interestingly, the addition of 100 µM ∆9-THC to HK slightly enhanced the response of inward K + currents ( Figure 6(B(b))), whereas the addition of 100 µM CBD, CP55,940, or rimonabant to HK slightly reduced inward K + currents ( Figure 6(C(b),D(b),E(b))). Comparison of their effects with that of ≈100 µM WIN55,212-2 is shown in Figure 7. These results indicate that the ability of WIN55,212-2 to block GIRK1/2 is stronger than that of CBD, CP55,940, and rimonabant, while 100 µM ∆9-THC can modestly activate GIRK1/2.
The structure of all human Kir channels is very similar. They are all composed of four subunits that are each formed from two membrane-spanning helices connected by an extracellular domain and a pore helix, and cytosolic N-and C-termini [35]. Moreover, the homology among human Kir channels is relatively high with an identity that ranges between ∼28% (between Kir7.1 and Kir6.1 or Kir6.2) and ∼70% (between IRK1 and IRK2, and between GIRK2 and GIRK4) [35]. In particular, the identity between human IRK1 and GIRK1 is 42% [36], and the identity between human IRK1 and GIRK2 is ∼45% [35]. Additionally, IRK1 was found to be widely expressed in the human body [37] and is an important contributor to the resting membrane potential of the cells [38]. To explore the selectivity of the effect of WIN55,212-2, we tested whether the high concentration (≈100 µM) of WIN55, 212-2 interacts with IRK1, a constitutively active Kir channel. Intriguingly, in oocytes expressing IRK1, the application of ≈100 µM WIN55,212-2 in HK did not cause a significant effect on inward K + currents through the GIRK1/2 (Figure 6(c)). The other typical cannabinoids ∆9-THC, CBD, CP55,940, and rimonabant in the presence of HK also did not produce obvious effects on inward K + currents (Figure 6(c)).

Figure 3.
Concentration-response curve of WIN55,212-2 for CB-GIRK1/2-RGS4 coupling and GIRK1/2. Oocytes were co-injected with CB, GIRK1/2, and RGS4, or GIRK1/2 cRNAs. Inward K + current enhancement and reduction were produced on the application of a range of different concentrations of WIN55,212-2 in the presence of HK. The positive percentage represents enhancement and the negative percentage represents reduction. Each data point is the mean ± standard deviation (SD) of three determinations from two or three batches of oocytes. Oocytes co-injected with CB1, GIRK1/2, and RGS4 cRNAs, (B) oocytes co-injected with CB2, GIRK1/2, and RGS4 cRNAs, and (C) oocytes injected with GIRK1/2 cRNAs were subjected to a 1-s voltage ramp protocol from −150 to +60 mV from a holding potential of −20 mV. Inward I K,basal was strongly reduced by exchanging HK to HK plus ≈ 100 µM WIN55,212-2. Typical I K,basal and I K,WIN were recorded from three to six different oocytes.  Plotting the percentage of current enhancement and reduction (IK,WIN/IK,basal × 100) as a function of the WIN55,212-2 concentration reveals a concentration-dependent increase and subsequent decrease in the response of IK, WIN (Figure 3).
Or the increase section of the concentration-response curve, the best-fit indicates an apparent EC50 value of 0.374 μM (95% CI: 0.276 to 0.504) for activating CB1 (n = 3; Figure  5A), while the best-fit indicates an apparent EC50 value of 0.260 μM (95% CI: 0.163 to 0.414) for activating CB2 (n = 3; Figure 5B). The EC50 value for activating CB1 is consistent with previous studies where EC50 values were shown to range from 0.03 μM to 3 μM depending on the CB1 expression systems [1].

Discussion
Cannabinoid receptors have been expectantly viewed as hopeful targets for the treatment of diverse diseases. Therefore, many expression systems of CB1 and CB2 have emerged in response to promptly discovering drug candidates by screening potential ligands in vitro, e.g., mammalian cells transfected with human cannabinoid receptors, certain cultured cell lines that express cannabinoid receptors naturally, and cannabinoid receptor-containing membrane preparations obtained from tissues (such as brain and spleen) [1,39]. The CB-GIRK1/2-RGS4 oocyte expression system built in this research is also a robust and reliable system for rapidly analyzing functions of cannabinoids or potential cannabinoids. Using this system, it is possible to investigate the effects of cannabinoids or potential cannabinoids on CB1 and/or CB2 in combination with their accessory functional proteins, e.g., GIRK channels. As mentioned in "1. Introduction", GIRK channels are directly opened by Gi/o(βγ) subunits which are dissociated from Gi/o(α) subunits via activating cannabinoid receptors. Thus, the expression system of cannabinoid receptors used in this research nicely mimics the CB-GIRK signaling pathway. The other component, RGS4, in the expression system maintained the robustness of CB-GIRK1/2 coupling. RGS proteins are key components of GPCR complexes, interacting directly with G protein α-subunits to enhance their intrinsic GTPase activity [40]. Among RGS proteins composed of more than 30 genes, RGS4 expression has been found to markedly enhance Gi/o-coupled receptor-activated GIRK currents in Xenopus oocytes [29]. This study also reports that RGS4 augmented GIRK currents by increasing the availability of free Gi/o(βγ) rather than directly activating GIRK channels or potentiating Gi/o(βγ)-mediated gating of GIRK channels [29]. Furthermore, RGS4 dramatically improved the sensitivity of the CB-Gα-Gβγ-RGS4 construct system in Sf9 insect cells for screening potential ligands [30]. Thus, CB2-GIRK1/2-RGS4 was successfully constructed in this research, presumably because of the presence of RGS4 in a Xenopus oocyte expression system, compared with the previously built system [27]. Another advantage of using the heterologous Xenopus oocyte expression system is that different combinations of membrane-bound proteins can be , and rimonabant reduced inward K + currents, while 100 µM ∆9-THC slightly enhanced inward K + currents. The positive percentage represents enhancement and the negative percentage represents reduction. Statistical comparisons between the various experimental groups were performed by Dunnett's multiple comparisons test of one-way ANOVA where p < 0.05 was considered significant. ****, p < 0.0001 vs. percentage for WIN55,212-2. Each data point is the mean ± SD of three determinations from two or three batches of oocytes.

Discussion
Cannabinoid receptors have been expectantly viewed as hopeful targets for the treatment of diverse diseases. Therefore, many expression systems of CB1 and CB2 have emerged in response to promptly discovering drug candidates by screening potential ligands in vitro, e.g., mammalian cells transfected with human cannabinoid receptors, certain cultured cell lines that express cannabinoid receptors naturally, and cannabinoid receptorcontaining membrane preparations obtained from tissues (such as brain and spleen) [1,39]. The CB-GIRK1/2-RGS4 oocyte expression system built in this research is also a robust and reliable system for rapidly analyzing functions of cannabinoids or potential cannabinoids. Using this system, it is possible to investigate the effects of cannabinoids or potential cannabinoids on CB1 and/or CB2 in combination with their accessory functional proteins, e.g., GIRK channels. As mentioned in "1. Introduction", GIRK channels are directly opened by Gi/o(βγ) subunits which are dissociated from Gi/o(α) subunits via activating cannabinoid receptors. Thus, the expression system of cannabinoid receptors used in this research nicely mimics the CB-GIRK signaling pathway. The other component, RGS4, in the expression system maintained the robustness of CB-GIRK1/2 coupling. RGS proteins are key components of GPCR complexes, interacting directly with G protein α-subunits to enhance their intrinsic GTPase activity [40]. Among RGS proteins composed of more than 30 genes, RGS4 expression has been found to markedly enhance Gi/o-coupled receptor-activated GIRK currents in Xenopus oocytes [29]. This study also reports that RGS4 augmented GIRK currents by increasing the availability of free Gi/o(βγ) rather than directly activating GIRK channels or potentiating Gi/o(βγ)-mediated gating of GIRK channels [29]. Furthermore, RGS4 dramatically improved the sensitivity of the CB-Gα-Gβγ-RGS4 construct system in Sf9 insect cells for screening potential ligands [30]. Thus, CB2-GIRK1/2-RGS4 was successfully constructed in this research, presumably because of the presence of RGS4 in a Xenopus oocyte expression system, compared with the previously built system [27]. Another advantage of using the heterologous Xenopus oocyte expression system is that different combinations of membrane-bound proteins can be freely expressed: for example, co-inject mRNA encoding CB1 and TRPV1 (the vanilloid receptor, also the target of some cannabinoid ligands), or CB2 and µ-opioid receptors, to examine the possible promiscuity of ligands, or CB1 with different types of voltage-gated ion channels (e.g., Na, K, Ca, Cl). Besides, structure-function research, based on the expression of mutant CB1/CB2, can also be conducted using this system. As such, this functional bioassay can unravel hitherto unknown coupling(s) of signal-transmission pathways, e.g., the cannabinoid and opioid pathways, allowing us to better understand the properties of some cannabinoids [1]. In addition to orthosteric ligands, discovery and characterization of allosteric modulators can also be achieved using this system, thus, functionally and structurally exploring allosteric modulation of CB receptors.
Functionally, WIN55,212-2 under high concentrations significantly blocked GIRK1/2 activated by CB1 or CB2, and thus inhibited the CB-GIRK1/2 signaling. Previous studies have shown that after chronic administration of WIN55,212-2 to adult male rats, CB1 binding and mRNA levels were reduced in the substantia nigra and the striatum [42]. Moreover, chronic treatment with WIN55,212-2 impaired the recognition memory and brain network functional connectivity of adult male mice [26]. Also, chronic exposure to WIN55,212-2 during adolescence induced persistent emotional and cognitive dysfunctions that reflect some facets of schizophrenia disease, including increased anxiety and alterations in spatial and episodic-like memory [43,44]. It is notable that WIN55,212-2 disrupted sensorimotor gating and attentional filtering in adolescent animals and that these disruptions were selectively reversed by antipsychotics such as haloperidol and risperidone [45][46][47]. These behavioral impairments associated with chronic exposure to cannabinoids during adolescence were mainly related to alterations in dopaminergic neurotransmission in the prefrontal cortex, which is a prominent component of cognitive dysfunctions in schizophrenia disease [48,49]. Furthermore, the frequency and amplitude of transient dopamine release events increased as WIN55,212-2 was administered in a cumulative, ascending manner for a chronic vehicle in rats [50]. As described in "1. Introduction", the release of neurotransmitters including dopamine, i.e., neurotransmission, can be inhibited by activating GIRK channels. Additionally, mice lacking GIRK2 exhibited dopamine-dependent hyperactivity and elevated responses to drugs that stimulate dopamine neurotransmission [51]. Thus, the increased dopamine-releasing and altered neurotransmission in brain regions under chronic administration of WIN55,212-2 observed in the above studies might be because of the blockage of GIRK1/2 produced by accumulated high concentrations of WIN55,212-2, as shown in the present study.
Heavy or regular cannabis/cannabinoid abuse, generally defined as daily or almost daily use over a prolonged period, has been linked to cognitive dysfunction and increased risk of developing psychiatric symptoms, including schizophrenia spectrum disorders, acute psychosis, mania, and a motivational syndrome [26]. Nevertheless, no available data shows the concentrations of WIN55,212-2 after prolonged/chronic treatment in vivo. Similarly, the concentrations of ∆9-THC in vivo are, as far as we know, also unknown after chronic exposure of single ∆9-THC, though many studies have revealed the plasma concentration of ∆9-THC in humans after chronic treatment of cannabis. Meanwhile, cannabis/cannabinoid-based medicines are increasingly being used to treat several diseases such as epilepsy, chronic pain, multiple sclerosis, and neurodegenerative diseases [52][53][54][55], but the potential for negative side effects has not been well characterized. Understanding the effects of chronic cannabinoid exposure on brain and synaptic function will open a window into the development of therapeutic tools that could counteract the 'on target' side effects associated with chronic use of cannabis and cannabinoid-based medicines [56,57].
Finally, we compared the effects of ≈100 µM WIN55,212-2 with the effects of 100 µM of other typical cannabinoids ∆9-THC, CBD, CP55,940, and rimonabant on GIRK1/2. As shown in Figure 6((b),(c)) and Figure 7, only WIN55,212-2 significantly reduced inward K + currents under the high concentration. This result may be attributed to the aminoalkylindole-type structure of WIN55,212-2 which is distinct from the structures of other tested cannabinoids. On the other hand, human IRK1 was not sensitive to high concentrations of WIN55,212-2, ∆9-THC, CBD, CP55,940, and rimonabant. The core cytoplasmic domain elements of GIRK1 or GIRK2 channels are very similar to that of IRK1 [58]. The binding of G proteins triggers a conformational change of the cytoplasmic domain of GIRK channels [59]. The strong rectification of IRK1 has been attributed to three principal electronegative regions: Asp172 in the M2 domain, Glu224/Glu299 in the cytoplasmic domains, and di-aspartate cluster (Arg255/Arg259) in the cytoplasmic pore [60]. We hypothesize that WIN55,212-2 does not bind to these regions in IRK1, and thus cannot trigger inward rectification.

Conclusions
The present research built and validated a reliable and efficient CB1 and CB2 expression system coupled to GIRK1/2 and RGS4 in Xenopus oocytes. The CB2-GIRK1/2-RGS4 coupling system is the first successful heterologous expression system of CB2. This research firstly reveals the dual modulation of cannabinoid receptors and brain GIRK1/2 coupling by WIN55,212-2 in the oocyte expression system. Although WIN55,212-2 is an agonist of CB1 and CB2 and can activate CB-GIRK signaling in the CB-GIRK1/2-RGS4 coupling, it can also block GIRK1/2 activated by CB1 and CB2 under high concentrations. Moreover, WIN55,212-2 also can directly block GIRK1/2 in a concentration-dependent manner. Comparing the effects of WIN55,212-2 with the effects of ∆9-THC, CBD, CP55,940, and rimonabant which are also typical cannabinoids on GIRK1/2 shows that only WIN55,212-2 can significantly block GIRK1/2 at the high concentration. However, the high concentration of WIN55,212-2, ∆9-THC, CBD, CP55,940, and rimonabant did not affect IRK1, which is highly similar to GIRK channels, representing the selective blockage of WIN55,212-2 on GIRK1/2. The blockage of GIRK1/2 in CB-GIRK signaling provides new insights on the cannabinoid receptors-mediated side effects induced by the chronic accumulation of WIN55,212-2 in vivo. This research also provides a straightforward methodology to rapidly test the effects of cannabinoids/potential cannabinoids on CB1 and CB2 ex vivo.