Structure–Activity Relationship and Functional Evaluation of Cannabinoid Type-1 Receptor
1Pharmacology Laboratory, College of Pharmacy, Chonnam National University, Gwangju 61186, Republic of Korea
2Department of Life Science, Kyonggi University, Suwon 16227, Republic of Korea
3Pharmacology Laboratory, College of Pharmacy, Sungkyunkwan University, Suwon 16419, Republic of Korea
*Corresponding Author E-mail: kmkim@jnu.ac.kr, Tel: +82-62-530-2936, Fax: +82-62-530-2949Abstract
The type-1 cannabinoid receptor (CB1R) is a potential therapeutic target in several pathological conditions, including neuropsychological disorders and neurodegenerative diseases. Owing to their structural diversity, it is not easy to derive general structure–activity relationships (SARs) for CB1R ligands. In this study, CB1R ligands were classified into six structural families, and the corresponding SAR was determined for their affinities for CB1R. In addition, we determined their functional activities for the activation of extracellular signal-regulated kinases (ERKs). Among derivatives of indol-3-yl-methanone, the highest ligand affinity was observed when a pentyl and a naphthalenyl group were attached to the N1 position of the indole ring and the carbon site of the methanone moiety, respectively. In the case of adamantane indazole-3-carboxamide derivatives, the presence of fluorine in the pentyl group, the substituent at the N1 position of the indazole ring, strongly increased the affinity for CB1R. For (naphthalen-1-yl)methanone derivatives, the presence of 4-alkoxynaphthalene in the methanone moiety was more beneficial for the affinity to CB1R than that of a heterocyclic ring. The functional activities of the tested compounds, evaluated through ERK assay, were correlated with their affinity for CB1R, suggesting their agonistic nature. In conclusion, this study provides valuable insight for designing novel ligands for CB1R, which can be used to control psychiatric disorders and drug abuse.
INTRODUCTION
The biological effects of cannabinoids, the main constituents of the ancient medicinal plant Cannabis sativa (marijuana), are mediated by two members of the G protein-coupled receptor family, namely, the cannabinoid receptors 1 (CB1R) and 2 (CB2R). Along with the identification of CB1R (Gerard et al., 1990; Matsuda et al., 1990) and CB2R (Munro et al., 1993), two endogenous agonists, N-arachidonoyl ethanolamine (AEA; anandamide) (Devane et al., 1992) and 2-arachidonoylglycerol (2-AG), were discovered in previous studies (Mechoulam et al., 1995; Sugiura et al., 1995).
CB1R mediates a wide range of physiological functions of cannabinoids, possibly because of its widespread expression in the brain (Mackie, 2005), which includes the olfactory bulb, hippocampus, basal ganglia, and cerebellum (Matsuda et al., 1990; Galiegue et al., 1995). In fact, CB1R and the in vivo cannabinoid systems are largely involved in various aspects of neuropsychological disorders and neurodegenerative diseases. These include appetite disorders, anxiety, depression, schizophrenia, and addiction (Izzo et al., 2009; Hill et al., 2012, Patil et al., 2015).
Dopaminergic neurons in the ventral tegmental area (VTA) that project to the nucleus accumbens are presynaptically inhibited by GABAergic neurons. Stimulation of dopaminergic neurons results in increased production and secretion of endocannabinoids (Lupica and Riegel, 2005). Released endocannabinoids retrogradely act on CB1Rs located on the presynaptic GABAergic terminals in the VTA (Szabo et al., 2002), which reduces the GABA release, resulting in the activation of dopaminergic neurons (disinhibition) (Riegel and Lupica, 2004), and leading to an increased dopamine release into the nucleus accumbens. CB1R antagonists such as rimonabant block the CB1R, activating GABAergic neurons, and leading to decreased activity of the VTA dopaminergic neurons (Merrill et al., 2015).
In terms of chemical structure, established cannabinoid receptor agonists essentially fall into four main groups: classical, nonclassical, aminoalkylindoles, and eicosanoids (Howlett et al., 2002; Pertwee et al., 2010). The classical group consists of dibenzopyran derivatives. The nonclassical group contains bicyclic and tricyclic analogs of Δ9-THC that lack a pyran ring. Members of the aminoalkylindole group of CB1R/CB2R agonists have significantly different structures from both classical and nonclassical cannabinoids. Moreover, the structures of members of the eicosanoid group of CB1R/CB2R agonists are quite different from those of classical, nonclassical, or aminoalkylindole cannabinoids. Two prominent members of this group are the endocannabinoids anandamide and 2-arachidonoylglycerol (2-AG).
The CB1R, a member of the GPCR family, is coupled to the pertussis toxin (PTX)-sensitive Gi/o protein and suppresses the formation of cAMP upon receptor activation (Howlett et al., 1986). A rapid, transient, and PTX-sensitive release of Ca2+ from intracellular stores was also observed upon agonist binding to CB1Rs (Sugiura et al., 1996). At the same time, the CB1R interaction with Gs proteins has also been demonstrated in CHO cells (Maneuf and Brotchie, 1997; Calandra et al., 1999).
In this study, we evaluated the binding affinity for CB1R of a total of 30 compounds. For this purpose, the tested compounds were categorized into different subgroups to study their structure–activity relationship (SAR), and their binding affinities are presented in tables. In addition, the functional activities of CB1R ligands were assessed by measuring their efficacy to activate extracellular signal-regulated kinase (ERK).
MATERIALS AND METHODS
Reagents
Thirty putative ligands for CB1R were provided by the Korean Ministry of Food and Drug Safety (Cheongju, Korea). [3H]-SR141716A was purchased from PerkinElmer Life Sciences (Waltham, MA, USA). Antibodies against phospho-ERK1/2 and ERK2 were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA), and anti-mouse horseradish peroxidase (HRP)-conjugated secondary antibodies were obtained from Jackson ImmunoResearch (West Grove, PA, USA).
Cell culture
Human embryonic kidney 293 (HEK-293) cells were obtained from American Type Culture Collection (Manassas, VA, USA). Cells were cultured in a humid environment containing 5% CO2, using minimal essential medium containing 8% fetal bovine serum, 100 units/mL penicillin, and 100 μg/mL streptomycin. Cells were transfected using polyethylenimine (Polysciencies Inc., Warrington, PA, USA). Arrestin knockdown (KD) cells are described in previous studies (Min et al., 2023).
CB1R binding assay
For the affinity evaluation, cells stably expressing CB1R were dispensed into 24-well culture plates. The next day, [3H]-SR141716A was added along with increasing concentrations of test compounds, up to a final concentration of 3 nM, for 6 h at 4°C. Cells were washed five times with 200 μL of serum-free medium containing 0.1% bovine serum albumin (BSA), and 1% sodium dodecyl sulfate (SDS) was added to each well, followed by shaking for 2 h. Radioactivity measurements were performed using a liquid scintillation counter. JWH-018 was used as the positive control, whereas 10 μM rimonabant hydrochloride was employed to determine non-specific binding. Dose–response curves, reflecting binding curves, were obtained using the GraphPad Prism software to calculate the IC50 values; then, the Ki values were calculated as Ki=IC50/(1+[D]/Kd), where [D] represents the concentration of [3H]-SR141716A (3 nM) and Kd is the dissociation constant of [3H]-SR141716A (0.82 nM).
ERK measurements
Transfected cells were cultured in 6-well plates and starved overnight in a serum-free culture medium containing 0.1% BSA. Cells were treated with CB1R ligands dissolved in DMSO for a designated time period, and a SDS sample buffer was directly added to the culture wells. After incubating for 20 min at 65°C, samples were sonicated to shear genomic DNA. Proteins were separated by SDS–polyacrylamide gel electrophoresis (10% running gel, 5% stacking gel) and electroblotted onto polyvinylidene difluoride or nitrocellulose membranes. The membranes were incubated for 1 h at 22°C in TBS-Tween 20 (TBS-T) containing 5% nonfat dry milk or 4% BSA, followed by 1 h of incubation with antibodies against phospho-ERK (1:1000 dilution) and 1 h with alkaline phosphatase-conjugated secondary antibodies (1:5000 dilution) in 2% nonfat dry milk. Blots were visualized with a chemiluminescent Western blotting kit. The same samples were processed to detect ERK. Signals were quantified using the ChemiDoc MP imaging system (BioRad, Hercules, CA, USA).
Statistical analysis
Values are expressed as mean ± standard deviation. The statistical significance of the data was assessed by one-way analysis of variance with Tukey’s post-hoc test, using the GraphPad Prism 5 software (GraphPad, San Diego, CA, USA). A p-value <0.05 was considered significant.
RESULTS
Characterization of CB1R binding affinity of (1H-indol-3-yl)(aryl)methanone derivatives
The compounds presented in Table 1 have three substituents: R1 is located at the N1 position of the indole ring, whereas R2 is found at the C2 position of the indole ring, and R3 is attached to the ketone group along with the indole ring.
In the case of the R1 substituents, compound 6 with a longer carbon chain, displayed higher binding affinity than the 11 with shorter carbon units. However, an improved affinity was obtained when a methyl group was added to R2, even with a shorter alkyl group on R1 (12), suggesting that the presence of substituents at three possible sites (R1, R2, and R3) has positive effects on the binding affinity.
After examining the SAR for the R3 substituents, compounds containing an o-methyl (-OMe) (3, 4) or N-methylpiperazine (5) group showed lower affinity for CB1R than those containing a heteroaromatic (pyridine, 1) or 2-iodophenyl (2) ring. The affinity increased with the presence of a naphthalene ring (6). On the other hand, a naphthalene substituent with an ethyl group at C5 (7) or 8-methoxyquinoline (8) resulted in a reduced affinity. The binding affinity showed a considerable decrease when the 2-iodophenyl ring was replaced with an adamantane ring (9 vs. 10).
Binding affinity of N-(adamantan-1-yl)-1-alkyl-1H-indazole-3-carboxamide and N-(adamantan-1-yl)-1-alkyl-1H-indole-3-carboxamide for CB1 receptor
Differently from the first series of compounds, with three possible substitution positions (Table 1), the second series of compounds, shown in Table 2, have four possible substitution sites. The compounds included in Table 2 are divided into indazole and indole derivatives, depending on whether X is N or C. R4 represents an alkyl or fluoroalkyl group connected to the N1 position of the ring, while R5 and R6 represent groups linked to the N atom of the carboxamide moiety.
The presence of fluorine in the pentyl group, the substituent at the R4 position, resulted in a strongly increased affinity with indazole but not indole derivatives (13-16). No definitive conclusions could be obtained on the effects of substitution at R6, because of the limited number of corresponding compounds. It could be speculated that the bulkiness of the substituent at R5 and R6 had negative effects on the affinity for CB1R (17, 18).
Binding affinity of N-(1-amino-3-methyl-1-oxobutan-2-yl)-1-(alkyl)-3-(4-fluorophenyl)-1H-pyrazole-5-carboxamide for CB1 receptor
Table 3 shows two carboxamide derivatives connected to the fluorophenyl pyrazole ring. In this case, R7 represents the group connected to the N1 site of the pyrazole ring. The difference in binding affinity between 19 and 20 indicates that the nature of the substituent and the bond orientation might play a vital role in determining their affinity for CB1R.
Binding affinity of (naphthalen-1-yl)methanone derivatives for CB1 receptor
As shown in Table 4, (naphthalen-1-yl)methanone derivatives containing different rings, such as 4-alkoxynaphthalen-1-yl (21), 1H-benzo[d]imidazol-2-yl (22), 1H-pyrrol-3-yl (23), 9H-carbazol-3-yl (24), and 4-ethoxynaphthalen-1-yl) (21) had different binding affinities, increasing in the order 24<22< 23<21; this indicated that the attachment of 4-alkoxynaphthalene to the methanone moiety was more beneficial for the affinity for CB1R than that of the heterocyclic ring.
Binding affinity of N-(4-hydroxy-3-alkylbenzyl)-eicosatetraenamide for CB1 receptor
Arvanil, a structural hybrid between the endogenous CB1R ligand anandamide and capsaicin, is a potent CB1R agonist. This compound is also known to act as an inhibitor of anandamide transport and as a TRPV1 agonist at relatively high concentrations (Melck et al., 1999; Di Marzo et al., 2002). NADA, a potent endogenous cannabinoid and vanilloid receptor agonist, exhibits selectivity for CB1 over CB2 receptors. Additionally, it serves as a potent agonist at TRPV1 (VR1) receptors (Bisogno et al., 2000). As shown in Table 5, a compound with a hydroxy substituent at R9 exhibited slightly higher affinity for CB1R compared to the molecule with a methoxy substituent (25 vs. 26) (Di Marzo et al., 2002).
Binding affinity of miscellaneous group compounds for CB1 receptor
The remaining compounds 27-30 possess complex structures containing tetramethyl cyclopropane carboxamides (27), trans-tetrahydrocannabinol (28), trifluorobutane-1-sulfonates (29), and 1H-indole-2-carboxamides (30). Among these compounds, 28 showed the highest affinity for CB1R, demonstrating the importance of stereochemistry (bond orientation) and heterocyclic in the ring structure. Bay 59-3074 is a selective CB1R/CB2R partial agonist (De Vry et al., 2004).
Characterization of CB1R-mediated ERK activation
To complete the analysis of the SAR obtained by studying the receptor ligand binding, we selected compounds with various affinities and different structural characteristics and measured their effects on ERK activation. For this purpose, we first characterized the CB1R-mediated ERK activation for (–)-trans-Δ9-tetrahydrocannabinol (THC).
In a time-course experiment, the CB1R-mediated ERK activation reached its peak at 2 min and gradually decreased after that (Fig. 1A). In a dose–response experiment, the ERK activation peaked at 300 nM (Fig. 1B). Moreover, the ERK activation was inhibited by PTX treatment (Fig. 1C). These results show that ERK activation through CB1R occurs via Gi/o protein.
Correlation between ligand affinity and ERK activation
In the next stage of the study, we selected compounds with varying affinities for CB1R from seven different databases and conducted an ERK assay.
First, four compounds with different Ki values were selected from Table 1: JWH-018 (Ki=4.44 nM), JWH-210 (Ki=64.44 nM), JWH-073 (Ki=71.12 nM), and AM-1284 (Ki=103.78 nM). As shown in Fig. 2A, all compounds except AM-1284 significantly increased the ERK activation. JWH-018 and JWH-210 are known to be full agonists of both CB1R and CB2R (Huffman et al., 2005), whereas JWH-073 and AM1248 are known to be agonists of both cannabinoid receptors (Aung et al., 2000; Frost et al., 2010).
Next, we tested the ERK activation properties of JWH-018 (Ki=4.44 nM, Table 1), AM-2233 (Ki=32.67 nM, Table 1), STS-135 (Ki=44.59 nM, Table 2), and FUBIMINA (Ki=54.08 nM, Table 4). The level of ERK activation followed the order JWH-018>AM2233>STS-135≅FUBIMINA (Fig. 2B). AM-2233 and STS-135 act on both cannabinoid receptors (Jarbe et al., 2011; Banister et al., 2015). FUBIMINA has a more than 10 times higher selectivity to CB2R than CB1R and does not fully substitute for Δ9-THC in rat discrimination studies (Wiley et al., 2015).
After that, we tested JWH-018 (Ki=4.44 nM, Table 1), CB-13 (Ki=16.52 nM, Table 4), and EG-018 (Ki=170.6 nM, Table 4). As shown in Fig. 2C, JWH-018 and CB-13 (but not EG-018) significantly increased the ERK activation. CB-13 is a CB1R/CB2R dual agonist with higher affinity for CB1R (Dziadulewicz et al., 2007). EG-018 acts as a partial agonist of the CB1R and CB2R, with reasonably high binding affinity, but low efficacy in terms of induced signaling response (Gamage et al., 2020).
Finally, we tested JWH-018 (Ki=4.44 nM, Table 1), mepirapim (Ki=112.86 nM, Table 1), and A-836,339 (Ki=113.78 nM, Table 6). As shown in Fig. 2D, neither mepirapim nor A-836,339 caused a significant increase in ERK activation, in agreement with their Ki values for CB1R. Mepirapim, a structural analogue of JWH-018, has relatively low affinity for both CB receptors (>1,000 nM) (Schoeder et al., 2018). A-836,339 is a selective agonist for CB2R, with a Ki value higher than 250 nM for CB1R (McGaraughty et al., 2009).
DISCUSSION
In this study, the affinity (Ki value) of unlabeled ligands was determined through a competition assay using SR141716A (rimonabant) radiolabeled with tritium. SR141716A is a potent and selective antagonist of CB1R (Rinaldi-Carmona et al., 1994). The determination of the structure–activity relationship involved two stages. After elucidating the relationship between ligand structure and affinity, further experiments were carried out to investigate that between ligand affinity and ERK activation.
It is generally known that CB1R transmits signals by coupling with Gi/o proteins (Howlett et al., 1986; Sugiura et al., 1996). However, it has been reported that signal transmission of CB1R also takes place through Gs proteins. In mechanical terms, CB1R can be coupled with the Gs protein through cross-talk with different types of receptors (Glass and Felder, 1997; Breivogel and Childers, 2000). Alternatively, it has been hypothesized that distinct subpopulations of CB1R may exist, with one subpopulation coupling with Gi/o proteins and the other coupling with Gs proteins (Calandra et al., 1999). The results of our experiments showed that ERK activation induced by THC was inhibited by pretreatment with PTX, indicating that, under the present experimental conditions, CB1R transmits signals through Gi/o proteins for ERK activation.
Our results reveal a consistent relationship between structure of the ligand, affinity, and receptor function. These findings can provide valuable information for other research fields related to CB1R. For example, for medicinal chemists developing new derivatives for CB1R, our results can provide information on relevant SARs and highlight new directions for the design of ligands. In addition, the present results can support researchers conducting in vivo behavioral studies by providing the information necessary to determine the dose to be used in their experiments.
ACKNOWLEDGMENTS
This research was supported by the Ministry of Food and Drug Safety (19182MFDS403) and by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (RS-2023-00239943).
| Compd | Name | R1 | R2 | R3 | Ki (nM) |
|---|---|---|---|---|---|
| 1 | 5F3P | C5H10F | H | 8.73 | |
| 2 | AM-694 | C5H10F | H | 14.75 | |
| 3 | JWH-250 | C5H11 | H | 54.96 | |
| 4 | RCS-4 | C5H11 | H | 51.91 | |
| 5 | Mepirapim HCl | C5H11 | H | 112.86 | |
| 6 | JWH-018 | C5H11 | H | 4.44 | |
| 7 | JWH-210 | C5H11 | H | 64.44 | |
| 8 | QUPIC | C5H11 | H | 113.78 | |
| 9 | AM-2233 | H | 32.67 | ||
| 10 | AM-1248 | H | 103.78 | ||
| 11 | JWH-073 | C4H9 | H | 71.12 | |
| 12 | JWH-015 | C3H7 | CH3 | 11.14 |
| Compd | Name | R4 | X | R5 | R6 | Ki (nM) |
|---|---|---|---|---|---|---|
| 13 | APICA | C5H11 | C | - | 67.58 | |
| 14 | STS-135 | C5H10F | C | - | 44.59 | |
| 15 | AKB-48 | C5H11 | N | - | 59.83 | |
| 16 | 5F-AKB-48 | C5H10F | N | - | 3.1 | |
| 17 | AB-PINACA | C5H11 | N | 31.73 | ||
| 18 | BiPICANA | C5H11 | C | 145.31 |
| Compd | Name | R7 | Ki (nM) |
|---|---|---|---|
| 19 | 5F-AB-FUPPYCA | C5H10F | 318.59 |
| 20 | AB-CHFUPYCA | 48.71 |
| Compd | Name | R8 | Ki (nM) |
|---|---|---|---|
| 21 | CB-13 | 16.52 | |
| 22 | FUBIMINA | 54.08 | |
| 23 | JWH-307 | 37.42 | |
| 24 | EG-018 | 170.6 |
| Compd | Name | R9 | Ki (nM) |
|---|---|---|---|
| 25 | NADA | -OH | 27.37 |
| 26 | Arvanil | -OMe | 39.46 |
| Compd | Name | Structure | Ki (nM) |
|---|---|---|---|
| 27 | A-836,339 | 113.78 | |
| 28 | THC | 5.91 | |
| 29 | Bay 59-3074 | 18.12 | |
| 30 | ORG27569 | 399.74 |