“Photo-Rimonabant”: Synthesis and Biological Evaluation of Novel Photoswitchable Molecules Derived from Rimonabant Lead to a Highly Selective and Nanomolar “Cis-On” CB₁R Antagonist

Abstract

Human cannabinoid receptor type 1 (hCB₁R) plays important roles in the regulation of appetite and development of addictive behaviors. Herein, we describe the design, synthesis, photocharacterization, molecular docking, and in vitro characterization of “photo-rimonabant”, i.e., azo-derivatives of the selective hCB₁R antagonist SR1411716A (rimonabant). By applying azo-extension strategies, we yielded compound 16a, which shows marked affinity for CB₁R (K_{i (cis form}) = 29 nM), whose potency increases by illumination with ultraviolet light (CB₁R K_itrans/cis ratio = 15.3). Through radioligand binding, calcium mobilization, and cell luminescence assays, we established that 16a is highly selective for hCB₁R over hCB₂R. These selective antagonists can be valuable molecular tools for optical modulation of CBRs and better understanding of disorders associated with the endocannabinoid system.

Graphical Abstract

INTRODUCTION

The isolation and elucidation of Δ⁹-tetrahydrocannabinol (Δ⁹-THC) as one of the main active components of marijuana¹ gave way to the subsequent discovery of the receptors involved in the endocannabinoid system. To date, two canonical receptors have been described as part of this system.^2,3 These G-protein-coupled receptors (GPCRs) are named as human cannabinoid receptor type 1 (hCB₁R) and type 2 (hCB₂R). The main differences between them lie in the characteristics of their transduction system, given by differences between their primary structures.² The distribution of both receptors is diverse. The density of hCB₂R is predominant in lung and testicles, as well as peripheral tissues with immune function, such as tonsils, spleen, thymus, mast cells, leukocytes, and macrophages.^2,4,5 Expression of hCB₁R is noted in the central and peripheral nervous system and in peripheral tissues including liver, gastrointestinal tract, and some tissues of the musculoskeletal, cardiovascular, and reproductive systems.^3,6,7 This wide distribution of hCB₁R has generated several studies to understand the relationship between the receptor and multiple pathologies in the tissues in which the receptor is present.^3,4,8–11 By study of the association between hCBs and appetite increase, hCB₁R was found to be an important target for treatment of metabolic disorders such as obesity, anorexia, and cachexia. By activation of CB₁R in specific areas of the brain, such as nucleus accumbens and hippocampus, an increase in the desire to eat palatable food is produced, which is enhanced by a possible relationship between the cannabinoid system and the modulation of metabolic hormones in the blood that stimulate lipogenesis and fat accumulation.^12–14

Through the discovery and development of selective ligands for hCB₁R, a novel orally active and selective hCB₁R antagonist was reported in 1994.¹⁵ This compound, known as SR141716A or rimonabant, was marketed as an anorectic medicine for the treatment of obesity and metabolic disorders. However, despite its pronounced therapeutical effects during the clinical research phases and application, it was withdrawn from the market in 2008 due to adverse events related to increased anxiety and depression.¹⁶ Even being withdrawn from the market, the rimonabant scaffold remains a highly interesting pharmacological structure and entity. Recently published studies suggest that rimonabant-derived molecules are effective in in vitro assays against tuberculosis-causing pathogens.¹⁷ Similarly, the role of cannabinoids in light-induced retinal degeneration has been studied, where hCB₁R antagonists such as rimonabant are shown to be effective in protecting against photoreceptor cell death in murine models.¹⁸ Additionally, hCB₁R antagonists were found to have interesting pharmacological properties for the treatment of addiction to drugs of abuse. Studies suggest that rimonabant acts dually to generate anorectic effects, increasing satiety and decreasing neural response to positive stimuli in areas of the brain involved with reward, thereby inhibiting motivation for eating palatable food.^19,20 This mechanism is complemented by the relationship between hCB₁R and memory-related plasticity, suggesting that hCB₁R antagonists have a positive effect on relapse prevention in patients with addiction to abused drugs by modifying the impact on reward-related memories.^21,22

These pharmacological advances highlight the need to further investigate the specific roles of the hCB₁R in different brain regions in more detail by novel selective molecular probes. Molecular tool compounds are necessary in which an optimal response in the specific tissue can be obtained with the possibility to control this action also in time (spatiotemporal control). In the pharmaceutical field, numerous approaches have been followed to control drug activation in the body precisely at its site of action. These applications are based on the use of an external stimulus such as heat, electricity, magnetic field, ultrasound, among others, to release the active drug in a specific tissue.²³ Although these advances allow the release in precise doses and specific times, nevertheless most of these are irreversible, preventing the subsequent inactivation of the molecule by nonmetabolic routes once it has been released.

A strategy used to generate reversibility in the activation of a molecule is “photopharmacology”, which is based on the design and synthesis of molecules whose activation can be modulated by the application of light irradiation.²⁴ When these molecules are irradiated, their molecular structure changes, which generates a change in their pharmacological activity.^23,24 The use of photoswitchable molecules such as stilbenes, azobenzenes, or diarylethenes, and their coupling to specific molecular ligands, has allowed creation of systems that can be modified with high spatiotemporal precision.²⁴ This allows the study of the type of interaction between ligands and their receptors, as well as interaction between ligands and other biological processes after the activation/inactivation of the target molecules. All these possibilities generate useful tools for the study of complex pathologies where different physiological processes are simultaneously involved.²⁵ Applying the principles of photopharmacology for GPCRs, several research groups have successfully synthesized new photoligands, which act irreversibly or reversibly over their target protein (opioid receptors, adrenoreceptors, glutamate receptors, acetylcholine receptors, among others), thus expanding the toolbox for understanding the mechanisms of action and physiological responses at different levels in the organism.^26–29

The use of photopharmacology in the design of molecules involved in the function of hCBR is of great interest because of the multiple biological processes related to this receptor. The synthesis of molecules based on the Δ⁹-THC pharmacophore and the coupling of an azobenzene moiety have been reported, with which it was possible to yield two photoswitchable hCB₁R agonists, achieving control over hCB₁R and its associated signaling pathways through isomerization of compounds.³⁰ Although this most interesting type of molecule allows activation of the receptor with high spatiotemporal resolution, the selectivity over other targets in the endocannabinoid system is an important element to consider. Δ⁹-THC binds with similar affinity to both hCB₁ and hCB₂ in radioligand displacement assays.³¹ Cannabinoids such as Δ⁹-THC have been found to bind to a large extent to several targets, such as potassium channels and glycine receptors, among others.³² This lack of selectivity leads, in vivo, to greater unwanted side effects, which is why ligands with a high degree of specificity for their biological targets are preferred. In a previous work by our research group, it was possible to obtain hCB₂R partial agonist photochromic compounds with high selectivity over hCB₁R.³³ These compounds were based on abenzimidazole structure and showed a higher affinity for its cis photoisomer and with variations in its biological parameters depending on the substitution pattern introduced in the molecule. Very recently, the synthesis of hybrid compounds by combining the structural characteristics of the cannabinoid agonists HU-308 and AM841 was described and ligands based on HU-308 proved to be highly selective CB₂R ligands with the potential to be chemically coupled to fluorescent and photoswitchable units.³⁴ Subsequent work by the Carreira and Frank groups led to a set of highly remarkable CB₂R ligands that enabled real-time control of calcium release allowing monitoring of CB₂R signaling, which shows the remarkable potential of selective photoswitchable ligands as molecular tool compounds.^34,35

Herein, we describe the design, synthesis, photochemical characterization, and affinity and efficacy studies of “photorimonabant” as new photoswitchable molecules based on the structure of the hCB₁R antagonist. Using molecular docking and in vitro biological assays, we demonstrate that compound 16a has a very high affinity in the nanomolar range, which is even higher in its cis isoform. Additionally, this molecule is highly selective for hCB₁R over hCB₂R, becoming the first molecule with these characteristics reported to date.

RESULTS AND DISCUSSION

Chemistry.

Compounds were developed based on published rimonabant’s structure–activity relationship (SAR) studies.^36–38 We define that the most reasonable positions to incorporate the photoswitch are atoms 3 and 5 of the central pyrazole ring (Chart 1). For target compounds 9a–e and 14a,b, the azologization strategy was applied, by which the photoswitch is introduced inside the scaffold of the pharmacophoric structure, in such a way that the azoarene replaces in shape and size analogs coming from the original molecule.^39,40 For this case, the pyrazole ring of rimonabant was used, to which a diazene unit (N=N) was attached in positions 3 or 5 for the introduction of the second aromatic ring. For target compounds 16a–d, the azo-extension strategy was applied, whereby a phenylazo group is added to the pharmacophore in a position expected to generate changes in ligand–receptor interaction when photoisomerization is carried out.^39,41 For this project, we considered position 3 as optimal, extending the molecule through introduction of azobenzene in the position of the piperidin-1-ylcarbamoyl moiety.

Chart 1.

Design of Rimonabant Photoderivatives “Photo-Rimonabant” by Azologization (Blue) and Azo-Extension (Orange) Strategies

For the choice of the photoswitchable moiety to introduce, we considered that incorporation of 1,2-dithienylethenes leads to photoswitchable ligands with high thermal stability and quantum yields, e.g., enabling their application in kinetic studies.^25,42 For the present study, however, we decided to take azoarenes as the photoswitches of choice. The main advantage of azoarenes in this case is their comparatively small molecular size. For the introduction of the photoswitch in the different positions of the pyrazole ring, it is necessary to get similar proportions to the benzene ring (position 5) or to the piperidine ring (position 3) of the parent molecule rimonabant. With the introduction of an azo group attached to an aromatic ring, we could generate similar-sized structures. Another consideration is the thermal stability of the cis-isomer. By introducing different substituents on the aromatic rings, it is possible to obtain a wide variety of molecules with thermal half-lives from seconds to several days.⁴³ Finally, the ease of synthesis was also considered, since azoarenes are highly stable molecules under acidic or basic conditions, so the possibilities for chemical synthesis are broad.^44,45

For the introduction of the azoarene into position 5 of the pyrazole ring (Scheme 1), the synthesis started from the reaction between diethyl oxalate (1) and propionitrile in the presence of potassium tert-butoxide and 18-crown-6, yielding compound 2. Subsequently, reflux of 2 in EtOH was carried out together with 2,4-dichlorophenyl hydrazine, generating the N-substituted pyrazole 3. Hydrolysis of 3 with lithium hydroxide at room temperature yielded the carboxylic acid 4.

Scheme 1. Synthesis of Analogs of Rimonabant with the Introduction of Photoswitch in Position 5^a.

^aReagents and conditions: (a) CH₃CH₂CN, 18-crown-6, t-BuOK, THF, 60 °C, 30 min, 81%; (b) 2,4-dichlorophenylhydrazine hydrochloride, EtOH, reflux, 2 h, 25%; (c) LiOH, THF, H₂O, rt, overnight, quant; (d) MPS, DCM/H₂O, rt, overnight; (e) 40% aqueous NaOH/pyridine, 80 °C, 2 h, 20–25%; (f) (COCl)₂, DMF, DCM, rt, 1 h; (g) N-aminopiperidine, HBTU, TEA, DCM, rt, 4 h, 45–61%.

In a first approach, compound 4 was reacted with (COCl)₂ and subsequent reaction with N-aminopiperidine in the presence of HBTU and TEA, yielding the respective carboxamide 5a. However, formation of byproduct 5b with physicochemical characteristics like 5a did not allow the continuation of the synthetic route due to problems related to complex purification. As an alternative, it was decided to carry out the incorporation of the azoarene prior to formation of the carboxamide. For this purpose, the respective substituted anilines 6a–e were reacted with MPS for partial oxidation yielding nitroso compounds 7a–e. Subsequently, compound 4 was coupled in position 5 of the central ring by refluxing with 7a–e in 40% aqueous solution of NaOH and pyridine, obtaining substituted azoarenes 8a–e. Finally, after activation with (COCl)₂ and subsequent amide formation with N-aminopiperidine, compounds 9a–e were obtained.

For generation of compounds with azoarene moieties in position 3, azologization (compounds 14a,b) and azoextension (compound 16a) strategies were used in order to enlarge the structural diversity for photoswitchable CB₁R ligands. Synthesis started from commercially available rimonabant carboxylic acid 10 (Scheme 2). As a first step, the activation with (COCl)₂ and subsequent reaction with ammonia at room temperature were performed, yielding primary amide 11. Subsequently, a Hofmann rearrangement was carried out with NBS and metallic sodium in MeOH to generate the secondary amide 12. Refluxing with NaOH and MeOH was carried out, yielding primary amine 13 by carbamate hydrolysis. For the azo-coupling in position 3 of the pyrazole ring, reflux with 40% of aqueous NaOH and pyridine solution in the presence of nitroso compounds 7a,b was performed obtaining target compounds 14a,b. In parallel, compound 10 and 4-aminobenzylamine were treated with TEA and ECF to generate amide 15. For the formation of the respective azobenzene, coupling was carried out using nitrosobenzene through a Baeyer–Mills reaction under acidic conditions, yielding 16a.

Scheme 2. Synthesis of Analogs of Rimonabant with the Introduction of Photoswitch in Position 3^a.

^aReagents and conditions: (a) (COCl)₂, DMF, DCM, rt, 1 h; (b) 25% NH₄OH, DCM, rt, 30 min, quant; (c) Na, NBS, MeOH, reflux, 2 h, 51%; (d) NaOH, MeOH, 70 °C, 12 h, quant; (e) 40% aqueous NaOH/pyridine, 80 °C, 2 h, 30–34%; (f) ECF, TEA, 4-aminobenzylamine, DMF, rt, 0 °C, 3 h, rt, overnight, 88%; (g) nitrosobenzene, acetic acid/TFA/toluol, rt, overnight, 41%.

By studying the crystal structure of hCB₁R, it has been established that residues Phe170, Phe174, and Met384 are important binding points for the activity of antagonist ligands, especially for hydrophobic interactions with the piperidin-1-ylcarbamoyl moiety.⁴⁶ This is confirmed by SAR studies that suggest that introduction of an additional aromatic ring at position 5 generates an increase in the activity of the antagonists, possibly due to a greater interaction between this second ring with the hydrophobic residues of the receptor.³⁸ In the present work, an additional synthetic route was designed to obtain analogs of compound 16a (Scheme 3). The aim was to evaluate the influence of aromatic substitution of azobenzene on the biological activity of the photoligand, in a similar way as previously published for hCB₂.³³ Our hypothesis was that structural changes between the differently substituted azobenzene isomers could generate a different interaction behavior with the hydrophobic pocket of the receptor. Additionally, through the synthesis of analog 16d, we evaluated whether an oxygen atom incorporated in the distal position of azobenzene leads to significant changes in pharmacological activity.

Scheme 3. Synthesis of Analogs of 16a^a.

^aReagents and conditions: (a) TEA, Boc anhydride, DCM, 0 °C, 2 h, quant; (b) nitrosobenzene, acetic acid, rt, overnight, 58–68%; (c) MPS, DCM/H₂O, rt, overnight; (d) 1,3-dihydroisobenzofuran-5-amine, acetic acid, rt, overnight, 62%; (e) TFA, DCM, rt, overnight, 70–96%; (f) rimonabant carboxylic acid, ECF, TEA, DMF, rt, 0 °C, 3 h, rt, overnight, 88–95%.

The synthesis started using amino-substituted aminobenzylamines in the meta (17b), ortho (17c), and para (17d) positions, which were selectively protected by coupling with Boc on the aliphatic amine, obtaining compounds 18b–d. Subsequently, the aromatic amine 18d was partially oxidized using MPS, yielding nitroso-compound 19d. Baeyer–Mills reaction in acidic medium was conducted for generation of azobenzenes, carrying out the reaction of the amines 18b,c with nitrosobenzene, or 1,3-dihydroisobenzofuran-5-amine with nitroso-compound 19d, yielding photoligands 20b–d. Subsequent deprotection of the aliphatic amines in acidic medium generated compounds 21b–d. Finally, the coupling of the respective amines with 10 was carried out in the presence of ECF and TEA, yielding final compounds 16b–d. All synthesized target compounds showed stability in aqueous media and organic solvents during synthesis, purification, and biological assays.

Photophysicochemical Properties.

To confirm photo-isomeration and to rule out possible photofatigue (loss of photochromic behavior after multiple isomerization cycles) of the new analogs, the photostationary states were evaluated by UV/vis spectroscopy. Through the use of light at different wavelengths, it was established that all the tested compounds show their highest photoconversion to their cis forms by UV irradiation (λ = 366 nm) and to their trans photoisomers by blue light (λ = 454 nm) (Figure 1). These differences are explained by π → π* and n → π* transitions, typical of this type of photoswitchable compounds.⁴⁷ Compounds 9a–e have maximum absorption value of ~350 nm for the trans photoisomer and ~450 nm for the cis form. Compounds 14a,b and 16a–d show slightly lower values, being close to 325 and 430 nm for the trans and cis photoisomers, respectively. By alternately irradiating with UV and blue light, it was shown that all compounds can easily be photoconverted between their cis and trans states (t < 3 min). This process can be repeated for many cycles without noticeable photofatigue (Figure 2).

Figure 1.

Representation of the photoisomerization of 16a and its UV absorbance spectra by irradiation with λ = 454 nm blue light (blue, solid line) and λ = 366 nm UV light (purple, dotted line).

Figure 2.

Photochemical characterization of 16a: (a) HPLC chromatogram and the changes in their respective stationary states induced by light irradiation; (b) switching cycles without noticeable photofatigue; (c) thermal stability in a period of >3 h (buffer pH 7.4; 37 °C).

Most of the compounds show long thermal stability (>2 h, 37 °C); the only exception is compound 9c, which presented a reversal in the photoconversion of 45% after being stored in the dark for 2 h. This instability can be explained by the push–pull system effect observed for substituted photocompounds.⁴⁸ This effect is due to the introduction of electron-withdrawing groups in the para position of the azoarene, giving a considerable decrease in the thermal cis → trans relaxation time.⁴⁹ Furthermore, the photostationary distribution of the compounds was quantified by HPLC. All the compounds are distributed mainly toward their trans isomer (70–84%), which is thermodynamically more stable. By use of UV light at λ = 366 nm, the ratio shifts in favor of its cis isomer. For chlorinated compounds 9b, 9c, and 14b, their cis forms are given with a maximum of approximately 50%, while for nonchlorinated compounds their cis photoisomer can be obtained almost quantitatively.

Radioligand Binding Studies.

As all the tested compounds showed different extents of photoisomerization upon irradiation in physicochemical characterization, i.e., they are photoswitchable, it was decided to test their affinity at CB₁ and CB₂ to evaluate whether these differences indicate a change in affinity or selectivity for the receptors. The radioligand binding assay was used to assess whether the cis photoisomers had a higher or lower affinity for the receptor (“switching-on”/”switching-off”). In contrast to optical in vitro assays, radioligand binding studies can be performed under light exclusion and therefore effectively avoid false-negative (or false-low) differences between the in vitro activity of the different photoisomers due to back-isomerization upon measurement.^27,28

All the tested compounds have a notable affinity for CB₁, showing values in the 2 or 3 digits nanomolar range (Table 1). The 9a–e family of compounds, despite having a good affinity for CB₁, showed no change in their binding values (K_i value) when irradiated with λ = 366 nm UV light or λ= 454 nm blue light. Compounds with photoswitch incorporation at position 3 of the pyrazole ring showed differences in the affinity of their respective photoisomers. Compounds 14a and 14b, obtained by the azologization strategy, showed a slight change in affinity when irradiated with λ= 366 nm UV light (trans/cis ratio = 2.16 for 14a; 2.74 for 14b). In contrast, 16a, obtained using the azo-extension strategy, showed the highest difference regarding this key pharmacological parameter. The affinity (ratio trans/cis) is more than 15-fold higher for CB₁ in its cis photoisomer (K_i(CB₁) = 29 nM) compared to the trans isoform (K_i(CB₁) = 444 nM) (Figure 3a). Additionally, its cis form showed even stronger affinity for CB₁ than the reference compound rimonabant (K_i(CB₁) = 45 nM).

Table 1.

Affinity Values at rCB₁ and hCB₂ Determined by Radioligand Binding Studies

photoisomer		CB1 (Ki ± SD) [nM]a	CB1 trans/cis ratio	CB2 (Ki ± SD) or [3H]CP55950 displ at 10 μMb
rimonabant		45 ± 18		<10%
MN-I-79				38 ± 7.8 nM
9a	cis	461 ± 337	1.0	14%
				51%
	trans	458 ± 239
9b	cis	716 ± 26	0.83	61%
	trans	602 ± 85
				81%
9c	cis	358 ± 62	1.11	55%
				62%
	trans	401 ± 33
9d	cis	384 ± 40	0.76	45%
	trans	294 ± 66
				68%
9e	cis	679 ± 160	0.60	46%
				65%
	trans	410 ± 103
14a	cis	249 ± 155	2.16	21%
	trans	538 ± 66
				<10%
14b	cis	195 ± 63	2.74	60%
				31%
	trans	521 ± 200
16a	cis	29 ± 2.8	15.31	<10%
	trans	444 ± 116
				<10%
16b	cis	311 ± 196	1.23	<10%
				14%
	trans	377 ± 243
16c	cis	142 ± 28	1.58	14%
	trans	224 ± 148
				46%
16d	cis	221 ± 109	2.72	<10%
				32%
	trans	602 ± 416

^aPerformed on CB₁ membranes prepared from rat brain homogenate. Values are mean values from at least two independent experiments, each performed in triplicate.

^bPerformed on CB₂ membranes harvested from stably transfected hCB₂-HEK cells.

Figure 3.

Pharmacological evaluation of compound 16a. (a) Radioligand binding data of cis (purple, dotted line) and trans (blue, solid line) photoisomers. Symbols represent mean values ± SEM (n = 9). (b) Cell luminescence assay, hCB₁R receptor antagonism: results of HEK-293-CB₁-CRELuc cells pretreated with 16a-cis and stimulated with the agonist CP55940 (1 μM). Results are shown as the mean ± SD (n = 6), p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. versus negative control.

Regarding the effect of aromatic substitution of the photoswitch, we analyzed the binding of the azo-extended compounds with para (16a), meta (16b), and ortho (16c) substitution. As a result, the substitution in para position had the highest affinity toward CB₁. The substitution in ortho position showed moderate activity on its cis form (K_i(CB₁) = 142 nM), and the substitution in meta showed weaker affinity. Concerning the selectivity parameter, photocompounds 16a and 14a showed strong selectivity for CB₁ over CB₂ in the radioligand displacement assay at 10 μM.

Cell Luminescence Assay.

To further study the possible activation mechanism by which the tested compounds bind to hCB₁R, we studied the effect over HEK293-CB₁ and HEK293-CB₂ cells, transfected with the reporter plasmid pCRE-luc. In the luciferase assay with HEK-293-CB₂-CRELuc cells, no activity was observed for the tested molecules and hCB₂R. None of the compounds could inhibit forskolin-induced luciferase activity (agonist mode assay) (Supporting Information, Figures S14 and S15). Similarly, no compound was able to recover the activity in cells previously treated with the agonist CP55940 (antagonist mode assay).

Regarding the luciferase assay with HEK-293-CB₁-CRELuc cells, the compounds showed antagonistic behavior at hCB₁R. Compounds tested at doses of >1 μM antagonized the activity of the agonist CP55940 (1 μM) (Figure 3b). These results support the selectivity of compounds such as 16a, which for the radioligand binding assay showed high affinity at hCB₁R but no affinity at hCB₂R.

hCB₁ Calcium Mobilization Assay for Antagonism.

Antagonism of 16a was evaluated by measuring hCB₁-activated Gα_q16-coupled calcium mobilization in CHO-K₁ cells over-expressing hCB₁, using a fluorescent dye as described in previous studies.^50,51 Compound 16a (tested as a mixture of isomers due to interference between the dyes and the external light sources used for isomerization)²⁷ was found to be capable of moving the concentration response curve of the CB₁ reference agonist CP55940 rightwards, thus confirming the activity 16a as a human CB₁ antagonist (see section “Methods” below). The apparent antagonist dissociation equilibrium constant (K_e) of 16a was determined (~85.3 nM), corroborating the data obtained by radioligand binding assay and cell luminescence assay.

Molecular Docking.

To gain insights into the molecular framework of 16a binding to the CB₁R, we optimized a docking protocol using the coordinates of the AM6538-CB₁ complex (PDB code 5TGZ). By testing of all scoring functions available within GOLD,^52,53 the ASP⁵⁴ scoring function was found to consistently reproduce the experimental pose of AM6538. Rescoring with the DSX_CSD⁵⁵ scoring function yielded a top-ranked pose that differed only by 0.45 Å from the experimentally observed binding mode of AM6538 (Figure 4). The 4-methylpyrazole core of the molecules forms hydrophobic interactions with the side chains of Phe170 and Phe379, while the 2,4-dichlorophenyl arm shows an edge-face aryl–aryl interaction with Phe170 and a hydrophobic interaction with Val196. The 4-substituted chlorophenyl ring of 16a is predicted to occupy the same side pocket as the 4-substituted phenyl ring of AM6538, thereby addressing Phe102, Phe268, and Trp356 through aryl–aryl interactions, as well as Leu193, Val196, Leu359 through hydrophobic interactions.

Figure 4.

(a) Pose reproduction of AM6538 (0.45 Å RMSD). (b) Top scored docking pose for 16a-cis (1.30 Å partial RMSD) and (c) 16a-trans (0.59 Å partial RMSD). (d) Detail of the interaction network of the photoswitchable group of 16a-cis within CB₁, as predicted by docking. Hydrogen bonds are denoted by black dashed lines. (e) Comparison of the predicted binding poses of the two photoisomers of 16a. The photoswitchable group in the trans conformation is pushed toward the solvent.

Moreover, the proximal phenyl ring of the photoswitchable group of 16a occupies a similar space as the piperidin-1-ylcarbamoyl moiety of AM6538, forming interactions with the hydrophobic residues Ile105, Phe170, and Phe174.

The shared structural features between 16a and AM6538, a rimonabant derivative, lead to similar interaction networks between the ligands and the receptor, as suggested by our docking experiments. This may account for the experimentally observed CB₁R antagonistic activity of both 16a photoisomers. Finally, our docking experiments suggest that the distal phenyl ring of the photoswitchable group of 16a occupies a hydrophobic pocket defined by Ile105, Ile119, and Phe381 in the cis conformation, while the nitrogen atoms on the double bond address Ser123 through hydrogen bonds (Figure 4d). In contrast, the distal phenyl ring of the photoswitchable group is pushed toward the solvent in the trans conformation, exiting the globular core of the receptor (Figure 4e). We hypothesize that the energy penalty of the binding mode resulting from this conformational change on the ligand may contribute to the ~15-fold affinity decrease of the trans conformation of 16a, as compared to the cis conformation.

To explain the potential reasons for selectivity of 16a for CB₁ over CB₂, the superimposed structures of antagonist-bound CB₁ (AM6538) and CB₂ (AM10257) reveal that, in contrast to CB₂, CB₁ antagonist pushes α-helices I and II outward at the extracellular part of the receptor (Supporting Information Figure S13a, denoted by arrows). These helices adopt a more compact conformation in the antagonist-bound (AM10257) CB₂ structure. The compact binding pocket of CB₂ would cause the CB₁-bound superimposed pose of 16a to clash with V36, F87, and F91 located at α-helices I and II in CB₂ (Supporting Information Figure S13b). Additionally, S123 (which is predicted to form hydrogen bonds with 16a in CB₁) is replaced by C40 at the equivalent position in CB₂, which may affect hydrogen bond formation and promote CB₁ selectivity. These observations are congruent with the incompatibility of the CB₂ binding pocket geometry with the extended conformation of rimonabant, as it has been suggested in the literature.⁵⁶

CONCLUSIONS

Through the application of the azo-extension and azologization approaches, we were able to successfully synthesize molecules derived from rimonabant, which have a photoswitchable moiety in position 3 or 5 of the central pyrazole ring. These compounds can be easily switched between their cis and trans photoforms by short-time irradiation with light. This isomerization can be repeated for many cycles without showing photofatigue.

By pharmacological characterization, it was found that photorimonabant molecules have a high affinity for CB₁ in the nanomolar range. The incorporation of the photoswitch at position 3 of the pyrazole generated molecules with higher affinity differences between their photoisomers (14a,b; 16a–d) with respect to their modified homologues in position 5 (9a–e). In turn, by use of the azo-extension approach, the affinity difference between the cis and trans forms was further increased. By synthesis of compound 16a (“photo-rimonabant”), a >15-fold higher affinity was evidenced in its cis form (K_i(CB₁) = 29 nM), compared to the trans isomer (K_i(CB₁) = 444 nM). Likewise, it was found that 16a conserves its activity as an antagonist (K_e ~ 85.3 nM), with high selectivity for CB₁ over CB₂. These excellent pharmacological properties are explained by the interaction between 16a and different aromatic residues of CB₁. Our predictions indicate that residues Ile105, Ile119, and Phe381 are of special importance for the marked affinity of 16a-cis, mainly due to hydrophobic interactions. However, when switched to the trans photoisomer, conformational changes in the distal azobenzene ring generate increased exposure toward the solvent that leads to a decrease in affinity for the receptor.

Since photo-rimonabant shows “cis-on” activation and holds all in vitro characteristics of a GPCR molecular tool compound in terms of affinity and selectivity in various assays, we currently use photorimonabant in several advanced pharmacological applications.

METHODS

Chemistry.

General Methods.

Both solvents and commonly used reagents were purchased directly from commercial suppliers and used without further purification. Tetrahydrofuran (THF) was distilled from sodium/benzophenone under an argon atmosphere. The monitoring of the reactions was performed with thin layer chromatography (TLC) on silica gel 60 on alumina foils with fluorescent indicator, and spots were detected with UV light (254 nm). Melting points were determined with a Stuart melting point SMP3 apparatus (Bibby Sterilin Ltd., Staffordshire, U.K.). Preparative TLC purification (prep-TLC) was done by preparing glass surface chromatographic plates (20 cm × 20 cm) with silica gel 60GF254 (Merck). For column chromatography, silica gel 60, 230–400 mesh (Merck) was used. Nuclear magnetic resonance spectra were recorded with a Bruker AV-400 NMR instrument (Bruker, Karlsruhe, Germany) in deuterated solvents, and chemical shifts were expressed in ppm (DMSO, ¹H, 2.50 ppm, ¹³C, 39.52 ppm; CDCl₃, ¹H, 7.26 ppm, ¹³C, 77.16 ppm; MeOD, ¹H, 4.87 ppm, ¹³C, 49.0 ppm; acetone-d₆ ¹H, 2.05 ppm, ¹³C, 203.6 ppm). To monitor the purity of the products by HPLC, a Shimadzu kit, equipped with a DGU-20A3R degassing unit, a LC20AB liquid chromatograph, and an SPD-20A UV/vis detector, was used. By use of an LCMS 2020, mass spectra were obtained. The stationary phase was a Synergi 4 μm fusion-RP (150 mm × 4.6 mm) column, and a MeOH/water gradient with 0.1% formic acid was used as the mobile phase (parameters, A, water; B, MeOH, V(B)/(V(A) + V(B)) = from 5% to 90% over 10 min, V(B)/(V(A) + V(B)) = 90% for 5 min, V(B)/(V(A) + V(B)) = from 90% to 5% over 3 min. The method was performed with a flow rate of 1.0 mL/min and scan range of 60–1000 m/z. The new compounds were accepted with a purity of ≥95%. UV/vis spectra and experiments were made on a Varian Cary 50 Bio UV/vis spectrophotometer using Hellma (type 100-QS) cuvettes (10 mm light path). Target compounds used for biological evaluation were tested only if the purity was ≥95%.

General Procedure I for the Partial Oxidation of Anilines (7b–e, 19d).

The respective aniline (1 equiv) was dissolved in DCM. Then, an aqueous solution of potassium peroxymonosulfate (MPS) (2 equiv) was added to it (water/DCM, 1:1). The reaction mixture was vigorously stirred at room temperature overnight. After reaction, the mixture was diluted with DCM and washed with water. The combined organic phases were dried over Na₂SO₄, and the solvent was partially removed in vacuo; the brown-green crude resulting solution was used immediately for the next reaction step without purification.

General Procedure II for Azo-Coupling under Alkaline Conditions (8a–e).

Crude DCM solution of nitroso compound (4 equiv) was added to the respective amine (1 equiv) dissolved in pyridine and 40% aq NaOH (1:1). The mixture was stirred at 80 °C for 4 h. The reaction mixture was cooled, quenched with water, acidified to pH 2–5, and extracted with ethyl acetate and water (3×). The combined organic layers were dried over Na₂SO₄, and the solvent was removed under reduced pressure.

General Procedure III for Amidation in Position 3 of Pyrazole (9a–e).

To a stirred solution of the respective carboxylic acid (1 equiv) in DCM were added 3 drops of DMF. The reaction mixture was cooled to 0 °C and then, (COCl)₂ (1.7 equiv) was added dropwise. The reaction mixture was stirred for 1 h to room temperature. Excess of (COCl)₂ was removed under reduced pressure. In another flask were introduced N-aminopiperidine (1.5 equiv) and TEA (1 equiv) in DCM, and the mixture was cooled to 0 °C. Subsequently, a solution of the intermediate acyl chloride in DCM was added dropwise at the same temperature. The reaction mixture was stirred at 0 °C for 15 min and for 15 min more at room temperature. The mixture was diluted with DCM, washed with NaHCO₃ and brine, dried over Na₂SO₄, and concentrated under reduced pressure.

General Procedure IV for Selective Boc-Protection of Aminometylanilines (18b–d).

A solution of the respective diamine (1 equiv) and TEA (2 equiv) in DCM was cooled to 0 °C. Then, a solution of di-tert-butyl decarbonate (1 equiv) in DCM was added dropwise to the mixture. The reaction was stirred at 0 °C for a further 2 h. After completing the reaction, the mixture was washed with sat. NaHCO₃ (aq). The organic layers were dried over Na₂SO₄, and the solvent was removed under reduced pressure.

General Procedure V for Azo-Coupling under Acidic Conditions (20b–d).

The respective amine (1 equiv) and the nitroso compound (1.1 equiv) were dissolved in acetic acid. The mixture was stirred at room temperature overnight. After reaction, the solvent was removed in vacuo. The raw material was dissolved in DCM and washed with sat. NaHCO₃ (aq). The organic layers were dried over Na₂SO₄, and the solvent was removed under reduced pressure.

General Procedure VI for Boc-Deprotection of Aminometylanilines (21b–d).

The Boc-protected amine was dissolved in a mixture DCM/TFA (9:1) and stirred at room temperature overnight. After reaction, the mixture was neutralized with NaHCO₃ (aq) and washed with brine.

General Procedure VII for Incorporation of Azobenzene by Azo-Extension Strategy (16b–d).

Rimonabant carboxylic acid (1 equiv) and TEA (2.1 equiv) were dissolved in anhydrous DMF. Then, ECF (2.05 equiv) was added dropwise and the reaction was maintained under stirring at 0 °C for 2 h. To the reaction mixture, the respective azobenzene (2.4 equiv) was added, and the reaction was stirred at 0 °C for 1 h and subsequently at room temperature overnight. After the reaction was completed, the reaction was poured into ice–water. The precipitate was filtered and washed with water and petroleum ether.

Potassium (Z)-3-Cyano-1-ethoxy-1-oxobut-2-en-2-olate (2).

Diethyl oxalate (1.5 mL, 10.94 mmol, 1 equiv) was added to a solution of 18-crown-6 (280 mg, 1.1 mmol, 0.1 equiv) and sodium tert-butoxide (1.22 g, 10.94 mmol, 1 equiv) in anhydrous THF at 0 °C. The mixture was heated at 60 °C while anhydrous propionitrile (0.82 mL, 10.94 mmol, 1 equiv) was added dropwise. The reaction mixture was stirred at 60 °C for 1 h. After cooling, the solid was filtered and washed with anhydrous diethyl ether to yield the title compound (2) as yellow powder (1.71 g, 81%). Mp 114 °C (dec). ¹H NMR (400 MHz, DMSO): δ = 4.07−3.92 (m, 3H), 1.51 (s, 2H), 1.23−1.13 (m, 3H) ppm. ¹³C NMR (101 MHz, DMSO): δ = 167.9 (1C), 162.7 (1C), 128.5 (1C), 63.6 (1C), 58.9 (1C), 14.5 (1C), 12.5 (1C) ppm. LC: t_R = 6.52 min, purity = 99.15%. MS: [M − K]⁻ calcd for [C₇H₈KNO₃ − K]⁻ = 154.05, found 154.05.

Ethyl-5-amino-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxylate (3).

2,4-Dichlorophenylhydrazine hydrochloride (718 mg, 3.36 mmol, 1.3 equiv) was added to a suspension of 2 (500 mg, 2.58 mmol, 1 equiv) in EtOH. The reaction mixture was refluxed for 2 h. After cooling, water was added and the precipitate was collected, washed with water, and dried. Then, the compound was dissolved in ethyl acetate and the organic layer was washed with water, dried over Na₂SO₄ and the solvent was concentrated under reduced pressure. The residue was purified with column chromatography (petroleum ether/ethyl acetate = 3/1) to yield to yield the title compound (3), as a yellow solid (197 mg, 25%). Mp 180–183 °C. ¹H NMR (400 MHz, DMSO): δ = 7.86 (d, J = 2.2 Hz, 1H), 7.58 (dd, J = 8.5, 2.3 Hz, 1H), 7.50 (d, J = 8.4 Hz, 1H), 5.20 (s, 2H), 4.21 (q, J = 7.1 Hz, 2H), 2.05 (s, 3H), 1.25 (t, J = 7.1 Hz, 3H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 162.9 (1C), 144.0 (1C), 142.3 (1C), 136.2 (1C), 134.2 (1C), 132.8 (1C), 130.9 (1C), 130.0 (1C), 128.1 (1C), 102.6 (1C), 60.6 (1C), 14.3 (1C), 8.1 (1C) ppm. LC: t_R = 9.57 and 11.92 min, purity = 84.42%. MS: [M + H]⁺ calcd for [C₁₃H₁₃Cl₂N₃O₂]⁺ = 314.05, found 314.05.

5-Amino-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxylic Acid (4).

Lithium hydroxide monohydrate (66 mg, 1.54 mmol, 2 equiv) was added to a solution of 3 (242 mg, 0.77 mmol, 1 equiv) in THF/H₂O (2:1). The reaction was stirred at room temperature for 72 h. The mixture was acidified with HCl to pH 2–5 and extracted with ethyl acetate. The combined organic layers were washed with water, dried over Na₂SO₄ and the solvent was concentrated under reduced pressure to yield the title compound (4) as yellow solid (216 mg, quant). Mp 198 °C (dec). ¹H NMR (400 MHz, DMSO): δ = 7.86 (d, J = 2.2 Hz, 1H), 7.58 (dd, J = 8.5, 2.3 Hz, 1H), 7.50 (d, J = 8.5 Hz, 1H), 5.14 (s, 2H), 2.05 (s, 3H) ppm. ¹³C NMR (101 MHz, acetone-d₆): δ = 163.9 (1C), 149.1 (1C), 136.7 (1C), 135.9 (1C), 133.5 (1C), 131.7 (1C), 130.5 (1C), 129.0 (1C), 128.8 (1C), 106.3 (1C), 9.2 (1C) ppm. LC: t_R = 8.47 min, purity = 98.53%. MS: [M + H]⁺ calcd for [C₁₁H₉Cl₂N₃O₂]⁺ = 286.01, found 286.00.

2,3-Dichloro-1-nitrosobenzene (7b).

The reaction was carried out according to general procedure I, using 2,3-dichloroaniline (227 mg, 1.4 mmol) and MPS (0.86 g, 2.8 mmol). The crude product was immediately used without further purification for the next reaction step.

2,4-Dichloro-1-nitrosobenzene (7c).

The reaction was carried out according to general procedure I, using 2,4-dicholoroaniline (243 mg, 1.5 mmol) and MPS (0.92 g, 3 mmol). The crude product was immediately used without further purification for the next reaction step.

1-Methyl-2-nitrosobenzene (7d).

The reaction was carried out according to general procedure I, using o-toluidine (255 μL, 2.37 mmol) and MPS (1.46 g, 4.74 mmol). The crude product was immediately used without further purification for the next reaction step.

1-Methyl-3-nitrosobenzene (7e).

The reaction was carried out according to general procedure I, using m-toluidine (150 μL, 1.42 mmol) and MPS (0.87 g, 2.84 mmol). The crude product was immediately used without further purification for the next reaction step.

(E)-1-(2,4-Dichlorophenyl)-4-methyl-5-(phenyldiazenyl)-1H-pyrazole-3-carboxylic Acid (8a).

The reaction was carried out according to general procedure II, using nitrosobenzene (117 mg, 1.09 mmol) and 4 (155 mg, 0.55 mmol). The residue was purified by column chromatography (dichloromethane/methanol/formic acid = 98/2/0.05) to yield the title compound (8a) as an orange solid. (38.5 mg, 25%). Mp 248 °C. ¹H NMR (400 MHz, DMSO) δ 7.95 (d, J = 2.2 Hz, 1H), 7.77 (d, J = 8.5 Hz, 1H), 7.67 (dd, J = 8.5, 2.3 Hz, 1H), 7.61−7.51 (m, 5H), 2.61 (s, 3H) ppm. ¹³C NMR (101 MHz, DMSO): δ = 163.2 (1C), 152.2 (1C), 148.9 (1C), 135.6 (1C), 135.5 (1C), 134.8 (1C), 132.1 (1C), 131.7 (1C), 130.8 (1C), 129.3 (2C), 129.1 (1C), 128.0 (1C), 121.9 (2C), 113.4 (1C), 9.4 (1C) ppm. LC: t_R (min) = 10.69 (cis) + 11.11 (trans), purity = 76.8%. MS: [M + H]⁺ calcd for [C₁₇H₁₂Cl₂N₄O₂]⁺ = 375.04, found 375.05.

(E)-1-(2,4-Dichlorophenyl)-5-((2,3-dichlorophenyl)-diazenyl)-4-methyl-1H-pyrazole-3-carboxylic Acid (8b).

The reaction was carried out according to general procedure II, using 7b solution (1.4 mmol, 4 equiv) and 4 (100 mg, 0.35 mmol, 1 equiv) The residue was purified by column chromatography (petroleum ether/ethyl acetate/formic acid = 1/1/0.05) to achieve the title compound (8b) as an orange solid (31.1 mg, 20%). Mp 253 °C (dec). ¹H NMR (400 MHz, DMSO): δ = 13.35 (s, 1H), 7.97 (d, J = 2.2 Hz, 1H), 7.83 (dd, J = 7.9, 1.5 Hz, 1H), 7.78 (d, J = 8.5 Hz, 1H), 7.69 (dd, J = 8.5, 2.3 Hz, 1H), 7.48 (t, J = 8.0 Hz, 1H), 7.41 (dd, J = 8.2, 1.5 Hz, 1H), 2.66 (s, J = 8.1 Hz, 3H) ppm. ¹³C NMR (101 MHz, DMSO): δ = 163.0 (1C), 149.7 (1C), 149.3 (1C), 143.4 (1C), 135.9 (1C), 135.13 (1C), 133.3 (1C), 133.2 (1C), 132.7 (1C), 131.8 (1C), 130.8 (1C), 129.7 (1C), 128.8 (1C), 128.4 (1C), 116.8 (1C), 115.1 (1C), 9.7 (1C) ppm. LC: t_R (min) = 10.80 (cis) +11.79 (trans), purity = 99.0%. MS: [M + H]⁺ calcd for [C₁₇H₁₀Cl₄N₄O₂]⁺ = 442.96, 444.96, found 442.95, 444.95.

(E)-1-(2,4-Dichlorophenyl)-5-((2,4-dichlorophenyl)-diazenyl)-4-methyl-1H-pyrazole-3-carboxylic Acid (8c).

The reaction was carried out according to general procedure II, using 7c solution (1.4 mmol, 4 equiv) and 4 (100 mg, 0.35 mmol, 1 equiv). The residue was purified by column chromatography (petroleum ether/ethyl acetate/formic acid = 2/1/0.05) to achieve the title compound (8c) as an orange solid. (37.5 mg, 24%). Mp 242 °C. ¹H NMR (400 MHz, DMSO): δ = 7.93 (t, J = 2.6 Hz, 1H), 7.86 (d, J = 2.2 Hz, 1H), 7.75 (dd, J = 8.5, 2.4 Hz, 1H), 7.68−7.63 (m, 1H), 7.53 (dd, J = 8.8, 2.2 Hz, 1H), 7.43 (d, J = 8.8 Hz, 1H), 2.62 (s, 3H) ppm. ¹³C NMR (101 MHz, DMSO): δ = 163.0 (1C), 149.4 (1C), 146.9 (1C), 143.4 (1C), 137.5 (1C), 136.0 (1C), 135.8 (1C), 135.1 (1C), 131.8 (1C), 130.9 (1C), 130.4 (1C), 129.7 (1C), 128.7 (1C), 128.4 (1C), 117.6 (1C), 116.5 (1C), 9.8 (1C) ppm. LC: t_R (min) = 9.77 (cis) + 10.86 (trans), purity = 80.5%. MS: [M + H]⁺ calcd for [C₁₇H₁₀Cl₄N₄O₂]⁺ = 442.96, 444.96, found 442.95, 444.95.

(E)-1-(2,4-Dichlorophenyl)-4-methyl-5-(o-tolyldiazenyl)-1H-pyrazole-3-carboxylic Acid (8d).

The reaction was carried out according to general procedure II, using 7d solution (2.4 mmol, 6 equiv) and 4 (113 mg, 0.4 mmol, 1 equiv). The residue was purified by column chromatography (petroleum ether/ethyl acetate/formic acid = 2/1/0.1) to achieve the title compound (8d) as an orange solid. (38.9 mg, 25%). Mp 225 °C. ¹H NMR (400 MHz, DMSO): δ = 7.94 (d, J = 2.3 Hz, 1H), 7.74 (d, J = 8.5 Hz, 1H), 7.66 (dd, J = 8.5, 2.3 Hz, 1H), 7.46−7.31 (m, 3H), 7.26 (ddd J = 8.3, 6.7, 1.7 Hz, 1H) 2.61 (s, 3H), 2.29 (s, 3H) ppm. ¹³C NMR (101 MHz, DMSO): δ = 163.6 (1C), 150.8 (1C), 149.9 (1C), 143.6 (1C), 139.9 (1C), 136.8 (1C), 135.4 (1C), 132.8 (1C), 132.3 (1C), 132.0 (1C), 131.2 (1C), 129.9 (1C), 128.8 (1C), 127.2 (1C), 115.9 (1C), 114.5 (1C), 17.1 (1C), 10.1 (1C) ppm. LC: t_R (min) = 10.26 (cis) + 11.19 (trans), purity = 74.1%. MS: [M + H]⁺ calcd for [C₁₈H₁₄Cl₂N₄O₂]⁺ = 389.06, found 389.05.

(E)-1-(2,4-Dichlorophenyl)-4-methyl-5-(m-tolyldiazenyl)-1H-pyrazole-3-carboxylic Acid (8e).

The reaction was carried out according to general procedure II, using 7e solution (1.27 mmol, 6 equiv) and 4 (60 mg, 0.21 mmol, 1 equiv). The residue was purified by column chromatography (petroleum ether/ethyl acetate/formic acid = 2/1/0.1) to achieve the title compound (8e) as an orange solid. (16.0 mg, 20%). Mp 237 °C (dec). ¹H NMR (400 MHz, DMSO): δ = 7.85 (d, J = 2.2 Hz, 1H), 7.66 (d, J = 8.5 Hz, 1H), 7.59−7.55 (m, 1H), 7.34 (d J = 7.2 Hz, 1H), 7.30 (d J = 7.6 Hz, 1H), 7.28−7.19 (m, 2H), 2.50 (s, 3H), 2.26 (s, 3H) ppm. ¹³C NMR (101 MHz, DMSO): δ = 163.5 (1C), 152.5 (1C), 149.7 (1C), 144.6 (1C), 139.4 (1C), 136.2 (1C), 135.3 (1C), 133.2 (1C), 132.3 (1C), 131.2 (1C), 129.7 (1C), 128.5 (1C), 123.7 (1C), 118.8 (1C), 114.0 (1C), 111.5 (1C), 21.0 (1C), 9.9 (1C) ppm. LC: t_R (min) = 10.77 (cis) + 11.75 (trans), purity = 82.5%. MS: [M + H]⁺ calcd for [C₁₈H₁₄Cl₂N₄O₂]⁺ = 389.06, found 389.05.

1-(2,4-Dichlorophenyl)-4-methyl-5-[(E)-2-phenyldiazen-1-yl]-N-(piperidin-1-yl)-1H-pyrazole-3-carboxamide (9a).

The reaction was carried out according to general procedure III, using 8a (250 mg, 0.55 mmol, 1 equiv), (COCl)₂ (0.08 mL, 0.94 mmol, 1.7 equiv), N-aminopiperidine (0.1 mL, 0.83 mmol, 1.5 equiv), and TEA (0.08 mL, 0.55 mmol, 1 equiv). The residue was purified by prep-TLC (DCM/MeOH = 99/1) to achieve the title compound (9a) as an orange solid (155 mg, 61%). Mp 141 °C. ¹H NMR (400 MHz, CDCl₃) δ 7.68−7.61 (m, 3H), 7.59−7.55 (m, 1H), 7.46−7.39 (m, 5H), 2.86 (dd, J = 15.7, 10.6 Hz, 4H), 2.73 (s, 3H), 1.80−1.71 (m, 4H), 1.48−1.38 (m, 2H). ¹³C NMR (101 MHz, CDCl₃): δ = 159.7 (1C), 152.9 (1C), 150.1 (1C), 144.6 (1C), 136.3 (1C), 135.9 (1C), 133.2 (1C), 131.9 (1C), 130.3 (1C), 130.0 (1C), 129.3 (2C), 127.7 (1C), 122.8 (2C), 114.0 (1C), 57.23 (2C), 25.6 (2C), 23.5 (1C), 9.8 (1C). LC: t_R (min) = 10.73 (cis) + 11.62 (trans), purity = 99.1%. MS: m/z [M + H]⁺ calcd for (C₂₂H₂₂Cl₂N₆O)⁺ = 457.13, found 457.10.

1-(2,4-Dichlorophenyl)-4-methyl-5-[(E)-2-(2,3-dichlorophenyl)diazen-1-yl]-N-(piperidin-1-yl)-1H-pyrazole-3-carboxamide (9b).

The reaction was carried out according to general procedure III, using 8b (38 mg, 0.09 mmol, 1 equiv), (COCl)₂ (12.5 μL, 0.15 mmol, 1.7 equiv), N-aminopiperidine (16.3 μL, 0.14 mmol, 1.5 equiv), and TEA (12.5 μL, 0.09 mmol, 1 equiv). The residue was purified by column chromatography (petroleum ether/ethyl acetate = 2/1) to achieve the title compound as an orange solid (27.5 mg, 0.052 mmol, 61%). Mp 164 °C. ¹H NMR (400 MHz, CDCl₃): δ = 7.65 (s, 1H), 7.56 (dt, J = 3.4, 1.7 Hz, 1H), 7.54−7.48 (m, 1H), 7.45−7.39 (m, 2H), 7.37−7.30 (m, 1H), 7.19 (t, J = 8.0 Hz, 1H), 2.94−2.80 (m, 4H), 2.76 (s, 3H), 1.79−1.72 (m, 4H), 1.49−1.36 (m, 2H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 159.1 (1C), 150.2 (1C), 150.0 (1C), 144.4 (1C), 135.9 (1C), 135.8 (1C), 134.4 (1C), 134.2 (1C), 132.7 (1C), 132.5 (1C), 129.9 (1C), 129.8 (1C), 127.6 (1C), 127.0 (1C), 116.4 (1C), 114.5 (1C), 56.9 (2C), 25.2 (2C), 23.1 (1C), 9.6 (1C) ppm. LC: t_R (min) = 11.12 (cis) + 12.39 (trans), purity = 99.2%. MS: m/z [M + H]⁺ calcd for (C₂₂H₂₀Cl₄N₆O)⁺ = 525.05, 527.05, found 525.00, 527.00.

1-(2,4-Dichlorophenyl)-5-[(E)-2-(2,4-dichlorophenyl)diazen-1-yl]-4-methyl-N-(piperidin-1-yl)-1H-pyrazole-3-carboxamide (9c).

The reaction was carried out according to general procedure III, using 8c (40 mg, 0.09 mmol, 1 equiv), (COCl)₂ (12.5 μL, 0.15 mmol, 1.7 equiv), N-aminopiperidine (16.3 μL, 0.14 mmol, 1.5 equiv), and TEA (12.5 μL, 0.09 mmol, 1 equiv). The residue was purified by column chromatography (petroleum ether/ethyl acetate = 4/1) to achieve the target compound as an orange solid (25 mg, 52%). Mp 144–147 °C. ¹H NMR (400 MHz, CDCl₃): δ = 7.64 (s, 1H), 7.58−7.55 (m, 1H), 7.52 (d, J = 2.1 Hz, 1H), 7.44−7.37 (m, 3H), 7.23 (dd, J = 8.8, 2.2 Hz, 1H), 2.91−2.81 (m, 4H), 2.75 (s, 3H), 1.81−1.72 (m, 4H), 1.49−1.39 (m, 2H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 159.8 (1C), 150.8 (1C), 147.9 (1C), 145.0 (1C), 138.6 (1C), 137.6 (1C), 136.6 (1C), 136.4 (1C), 133.3 (1C), 131.0 (1C), 130.6 (1C), 130.5 (1C), 128.3 (1C), 128.2 (1C), 117.7 (1C), 116.6 (1C), 57.6 (2C), 25.9 (2C), 23.8 (1C), 10.3 (1C) ppm. LC: t_R (min) = 10.94 (cis) + 12.56 (trans), purity = 96.2%. MS: m/z [M + H]⁺ calcd for (C₂₂H₂₀Cl₄N₆O)⁺ = 525.05, 527.05, found 525.05, 527.05.

1-(2,4-Dichlorophenyl)-4-methyl-5-[(1E)-2-(2-methylphenyl)diazen-1-yl]-N-(piperidin-1-yl)pyrazole-3-carboxamide (9d).

The reaction was carried out according to general procedure III, using 8d (44 mg, 0.11 mmol, 1 equiv), (COCl)₂ (16 μL, 0.19 mmol, 1.7 equiv), N-aminopiperidine (18 μL, 0.17 mmol, 1.5 equiv), and TEA (15 μL, 0.11 mmol, 1 equiv). The residue was purified by column chromatography (petroleum ether/ethyl acetate = 2/1) to achieve the target compound as an orange solid (23 mg, 45%). Mp 172−174 °C. ¹H NMR (400 MHz, DMSO): δ = 9.26 (s, 1H), 7.95 (d, J = 2.2 Hz, 1H), 7.78 (d, J = 8.5 Hz, 1H), 7.68 (dd J = 8.5, 2.3 Hz, 1H), 7.42 (dt J = 14.1, 6.6 Hz, 2H), 7.35 (d, J = 7.2 Hz, 1H), 7.28 (t, J = 7.4 Hz, 1H), 2.84−2.77 (m, 4H), 2.60 (s, 3H), 2.32 (s, 3H), 1.59 (dd J = 10.6, 5.5 Hz, 4H), 1.25 (s, 2H). ¹³C NMR (101 MHz, CDCl₃): δ = 160.5 (1C), 159.8 (1C), 151.2 (1C), 144.6 (1C), 139.5 (1C), 135.9 (1C), 133.2 (1C), 132.1 (1C), 131.6 (1C), 130.3 (1C), 130.2 (1C), 128.9 (1C), 127.9 (1C), 126.6 (1C), 114.7 (1C), 114.6 (1C), 57.3 (2C), 25.6 (2C), 23.5 (1C), 17.5 (1C), 9.9 (1C). LC: t_R (min)= 10.47 (cis) + 11.25 (trans), purity = 98.0%. MS: m/z [M + H]⁺ calcd for (C₂₃H₂₄Cl₂N₆O)⁺ = 471.15, found 471.15.

1-(2,4-Dichlorophenyl)-4-methyl-5-[(1E)-2-(3-methylphenyl)diazen-1-yl]-N-(piperidin-1-yl)-1H-pyrazole-3-carboxamide (9e).

The reaction was carried out according to general procedure III, using 8e (24 mg, 0.062 mmol, 1 equiv), (COCl)₂ (8 μL, 0.086 mmol, 1.4 equiv), N-aminopiperidine (11 μL, 0.092 mmol, 1.5 equiv), and TEA (9 μL, 0.062 mmol, 1 equiv). The residue was purified by prep-TLC (petroleum ether/ethyl acetate = 4/1) to achieve the final compound as an orange solid (16 mg, 55%). Mp 134 °C. ¹H NMR (400 MHz, CDCl₃): δ = 7.67 (s, 1H), 7.58 (d, J = 2.0 Hz, 1H), 7.48 (d, J = 8.4 Hz, 1H), 7.45−7.38 (m, 3H), 7.32 (t, J = 7.6 Hz, 1H), 7.29−7.27 (m, J = 6.4 Hz, 1H), 2.95−2.83 (m, 4H), 2.73 (s, 3H), 2.41 (s, 3H), 1.85−1.71 (m, 4H), 1.51−1.40 (m, 2H). ¹³C NMR (101 MHz, CDCl₃): δ = 159.8 (1C), 153.1 (1C), 150.2 (1C), 144.6 (1C), 139.2 (1C), 136.4 (1C), 135.9 (1C), 133.2 (1C), 132.8 (1C), 130.4 (1C), 130.0 (1C), 129.1 (1C), 127.8 (1C), 124.0 (1C), 119.4 (1C), 114.1 (1C), 57.2 (2C), 25.6 (2C), 23.5 (1C), 21.4 (1C), 9.8 (1C). LC: t_R (min) = 10.74 (cis) + 11.73 (trans), purity = 98.4%. MS: m/z [M + H]⁺ calcd for (C₂₃H₂₄Cl₂N₆O)⁺ = 471.15, found 471.10.

5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide (11).

Four drops of DMF were added to a stirred solution of rimonabant carboxylic acid (100 mg, 0.262 mmol, 1 equiv) in DCM (4 mL). The reaction mixture was cooled to 0 °C, and then, (COCl)₂ (40 μL, 0.445 mmol, 1.7 equiv) was added dropwise. The reaction mixture was stirred for 1 h at room temperature. Excess of (COCl)₂ was removed under reduced pressure. Then, the intermediate acyl chloride was diluted in DCM, cooled to 0 °C, and introduced ammonia solution 25% dropwise (500 μL). The reaction mixture was stirred at 0 °C for 15 min and for 15 min more at room temperature. The mixture was quenched with NH₄Cl saturated solution and washed with water and DCM. The organic layer was dried over Na₂SO₄ and concentrated under reduced pressure to yield the title compound (11) as a white solid without further purification (109.3 mg, quant.). Mp 185–187 °C. ¹H NMR (400 MHz, DMSO): δ = 7.75 (dd, J = 13.7, 5.4 Hz, 2H), 7.57 (dd, J = 8.3, 2.2 Hz, 2H), 7.45 (d, J = 8.4 Hz, 2H), 7.28 (s, 1H), 7.23 (d, J = 8.4 Hz, 2H), 2.24 (s, 3H) ppm. ¹³C NMR (101 MHz, DMSO): δ = 164.0 (1C), 144.8 (1C), 142.5 (1C), 135.8 (1C), 135.0 (1C), 133.7 (1C), 132.1 (1C), 131.9 (1C), 131.2 (2C), 129.6 (1C), 128.7 (2C), 128.3 (1C), 127.3 (1C), 116.5 (1C), 9.20 (1C) ppm. LC: t_R = 10.98; purity = 97.5%. MS: [M + H]⁺ calcd for [C₁₇H₁₂Cl₃N₃O]⁺ = 380.01, found 380.00.

Methyl N-[5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazol-3-yl]carbamate (12).

Metallic sodium (3.25 g, 140.4 mmol, 30 equiv) was dissolved in methanol. Then, compound 11 (1.8 g, 4.68 mmol, 1 equiv) and NBS (1.26 g, 7.02 mmol, 1.5 equiv) were added to the previously solution. The reaction was carried out under reflux for 1.5 h, and then the solvent was concentrated under reduced pressure. The crude was dissolved in ethyl acetate, washed with water and brine, and dried over Na₂SO₄. The residue was purified by flash chromatography (petroleum ether/ethyl acetate = 3:1) to achieve the title compound (12), as a white solid (0.989 g, 51%). Mp 176–178 °C. 1H NMR (400 MHz, DMSO): δ = 9.42 (s, 1H), 7.73 (d, J = 2.1 Hz, 1H), 7.60−7.48 (m, 2H), 7.44 (d, J = 8.4 Hz, 2H), 7.20 (d, J = 8.4 Hz, 2H), 3.65 (s, 3H), 1.94 (s, 3H) ppm.¹³C NMR (101 MHz, DMSO): δ = 154.8 (1C), 146.7 (1C), 141.0 (1C), 136.1 (1C), 134.3 (1C), 133.4 (1C), 132.0 (1C), 131.8 (1C), 130.8 (2C), 129.6 (1C), 128.7 (2C), 128.3 (1C), 127.9 (1C), 110.0 (1C), 51.9 (1C), 8.31 (1C) ppm. LC: t_R = 10.72 min, purity = 90.8%. MS: [M + 2]⁺ calcd for [C₁₈H₁₄Cl₃N₃O₂]⁺ = 410.02, 412.02, found 410.00, 412.00.

5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazol-3-amine (13).

Compound 12 (30 mg, 0.073 mmol, 1 equiv) was dissolved in methanol. Then to the mixture was added NaOH (30 mg, 6.5 mmol, 90 equiv). The reaction was heated to 70 °C for 12 h. After completion of the reaction, the solvent was concentrated under reduced pressure. The crude was dissolved in ethyl acetate, washed with water and brine, and dried with Na₂SO₄ to yield the title compound (13) as a white solid without further purification (26 mg, quant). Mp 140–142 °C. ¹H NMR (400 MHz, DMSO): δ = 7.66 (d, J = 2.3 Hz, 1H), 7.49−7.35 (m, 4H), 7.17−7.09 (m, 2H), 4.90 (s, 2H), 1.91 (s, 3H) ppm.¹³C NMR (101 MHz, DMSO): δ = 155.8 (1C), 140.4 (1C), 137.1 (1C), 133.1 (1C), 132.7 (1C), 132.1 (1C), 131.7 (1C), 130.5 (2C), 129.5 (1C), 129.0 (1C), 128.5 (2C), 128.0 (1C), 102.3 (1C), 7.66 (1C) ppm. LC: t_R = 10.86 min, purity = 91.3%. MS: [M + H]⁺ calcd for [C₁₆H₁₂Cl₃N₃]⁺ = 352.02, found 352.00.

5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-3-[(E)-2-phenyldiazen-1-yl]-1H-pyrazole (14a).

Nitrosobenzene (24 mg, 0.11 mmol, 2 equiv) was added to a solution of 13 (40 mg, 0.11 mmol, 1 equiv) in pyridine/NaOH 40% (1:1). The mixture was stirred at 80 °C for 2 h. Then, the reaction was cooled, quenched with water, acidified to pH 2–5, and extracted with ethyl acetate and water. The organic layer was dried over Na₂SO₄ and the solvent removed under reduced pressure. The residue was purified by prep-TLC (petroleum ether/ethyl acetate = 20/1) to achieve the target compound as an orange solid (15 mg, 30%). Mp 160 °C. ¹H NMR (400 MHz, CDCl₃): δ = 8.03−7.97 (m, 2H), 7.55−7.44 (m, 3H), 7.42 (dd, J = 5.3, 3.1 Hz, 2H), 7.36−7.28 (m, 3H), 7.18−7.11 (m, 2H), 2.39 (d, J = 5.5 Hz, 3H). ¹³C NMR (101 MHz, CDCl₃): δ = 160.9 (1C), 153.2 (1C), 143.3 (1C), 136.2 (1C), 136.0 (1C), 135.1 (1C), 133.2 (1C), 131.4 (1C), 131.0 (2C), 130.9 (1C), 130.3 (1C), 129.2 (2C), 129.1 (2C), 128.0 (1C), 127.4 (1C), 123.1 (2C), 110.9 (1C), 9.9 (1C). LC: t_R (min) = 11.06 (cis) + 12.07 (trans), purity = 98.5%. MS: m/z [M + H]⁺ calcd for (C₂₂H₁₅Cl₃N₄)⁺ = 441.04, found 440.95.

5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-3-[(E)-2-(2,3-dichlorophenyl)diazen-1-yl]-4-methyl-1H-pyrazole (14b).

This compound was synthesized analogously as 14a, using 7b (4 equiv) and 13 (100 mg, 0.26 mmol, 1 equiv). The residue was purified by prep-TLC (petroleum ether/ethyl acetate = 15/1) to achieve the final compounds as an orange solid (46 mg, 34%). Mp 154 °C. ¹H NMR (400 MHz, CDCl₃): δ= 8.06 (d, J = 8.1, 1.0 Hz, 1H), 7.88 (d, J = 7.9, 1.1 Hz, 1H), 7.75 (d, J = 2.2 Hz, 1H), 7.70 (d, J = 8.5 Hz, 1H), 7.63 (td, J = 5.2, 2.8 Hz, 3H), 7.59−7.53 (m, 1H), 7.43 (t, J = 7.5 Hz, 2H), 2.68 (d, J = 16.7 Hz, 3H). ¹³C NMR (101 MHz, CDCl₃): δ = 161.5 (1C), 150.8 (1C), 143.6 (1C), 136.2 (1C), 136.1 (1C), 135.3 (1C), 134.4 (1C), 133.0 (1C), 132.4 (1C), 131.1 (2C), 130.7 (1C), 130.4 (1C), 129.2 (1C), 129.1 (2C), 128.0 (1C), 127.5 (1C), 127.1 (1C), 115.4 (1C), 109.0 (1C), 10.8 (1C). LC: t_R (min) = 11.86 (cis) + 14.12 (trans), purity = 97.3%. MS: m/z [M + H]⁺ calcd for (C₂₂H₁₃Cl₅N₄)⁺ = 508.97, 510.96, found 508.90, 510.95.

N-(4-Aminobenzyl)-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-1H-pyrazole-3-carboxamide (15).

Rimonabant carboxylic acid (200 mg, 0.524 mmol, 1 equiv) and TEA (0.15 mL, 1.10 mmol, 2.1 equiv) were dissolved in anhydrous DMF. Then, ECF (0.11 mL, 1.07 mmol, 2.05 equiv) was added dropwise and the reaction was maintained under stirring at 0 °C for 2 h. To the reaction mixture, 4-aminobenzylamine (0.14 mL, 1.26 mmol, 2.4 equiv) was added, and the reaction was stirred at 0 °C for an hour and subsequently at room temperature overnight. After the reaction was completed, the reaction was poured into ice–water. The precipitate was filtered and washed with water and petroleum ether to afford the title compound (15) as a white solid (225 mg, 88%). Mp 102 °C. ¹H NMR (400 MHz, CDCl₃): δ = 7.44−7.36 (m, 1H), 7.31−7.24 (m, 4H), 7.15 (d, J = 8.3 Hz, 3H), 7.06 (t, J = 8.2 Hz, 2H), 6.64 (d, J = 8.4 Hz, 2H), 4.49 (d, J = 5.8 Hz, 2H), 4.11−3.08 (m, 2H), 2.41 (d, J = 9.7 Hz, 3H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 162.5 (1C), 145.9 (1C), 145.1 (1C), 143.1 (1C), 136.1 (1C), 136.0 (1C), 135.0 (1C), 133.1 (1C), 130.9 (2C), 130.6 (1C), 130.4 (1C), 129.5 (2C), 129.0 (2C), 128.4 (1C), 128.0 (1C), 127.4 (1C), 117.9 (1C), 115.3 (2C), 42.8 (1C), 9.6 (1C) ppm. LC: t_R = 10.00 min, purity = 93.8%. MS: [M + H]⁺ calcd for [C₂₄H₁₉Cl₃N₄O]⁺ = 485.07, found 485.05.

5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-N-((4-[(1E)-2-phenyldiazen-1-yl]phenyl)methyl)-1H-pyrazole-3-carboxamide (16a).

Nitrosobenzene (48 mg, 0.45 mmol, 2 equiv) was added to a solution of 15 (110 mg, 0.226 mmol, 1 equiv) in a mixture of acetic acid/toluene/TFA (6:6:1). Then, the reaction was stirred at room temperature overnight. After reaction, the mixture was diluted with ethyl acetate and washed with water. The organic layers were separated, dried over Na₂SO₄, and concentrated under reduced pressure. The residue was purified by prep-TLC (petroleum ether/ethyl acetate = 4/1) to achieve the title compound (53 mg, 41%) as an orange solid. Mp 93 °C. ¹H NMR (400 MHz, CDCl₃): δ = 7.70 (ddd, J = 8.3, 4.1, 1.7 Hz, 3H), 7.34−7.27 (m, 4H), 7.21 (d, J = 1.8 Hz, 1H), 7.16 (t, J = 6.1 Hz, 1H), 7.12−7.02 (m, 5H), 6.90−6.84 (m, 2H), 6.63 (ddd, J = 6.5, 4.3, 1.6 Hz, 1H), 4.43 (t, J = 55.8, 6.2 Hz, 2H), 2.20 (s, 3H). ¹³C NMR (101 MHz, CDCl₃): δ = 163.1 (1C), 153.1 (1C), 152.5 (1C), 145.2 (1C), 143.6 (1C), 142.0 (1C), 136.5 (1C), 136.4 (1C), 135.4 (1C), 133.4 (1C), 131.4 (1C), 131.3 (2C), 130.9 (1C), 130.8 (1C), 129.5 (2C), 129.4 (2C), 129.0 (2C), 128.4 (1C), 127.6 (1C), 123.6 (2C), 123.3 (2C), 118.4 (1C), 43.1 (1C), 9.9 (1C). LC: t_R (min) = 11.12 (cis) + 12.23 (trans), purity = 99.9%. MS: m/z [M + H]⁺ calcd for (C₃₀H₂₂Cl₃N₅O)⁺ = 574.09, found 574.05.

tert-Butyl (3-Aminobenzyl)carbamate (18b).

The reaction was carried out according to general procedure IV using 3-aminobenzylamine (1 g, 8.19 mmol), TEA (2.27 mL, 16.4 mmol), and di-tert-butyl decarbonate (1.79, 8.19 mmol). The product was used for the next step without further purification. The title compound (18b) was obtained as a white solid (1.80 g, quant). Mp 50–54 °C. ¹H NMR (400 MHz, CDCl₃): δ = 7.11 (t, J = 7.7 Hz, 1H), 6.63 (d, J = 7.5 Hz, 1H), 6.56 (dd, J = 11.0, 3.1 Hz, 2H), 4.83 (s, 1H), 4.20 (d, J = 5.6 Hz, 2H), 3.64 (s, 2H), 1.45 (s, 9H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 156.2 (1C), 146.6 (1C), 140.3 (1C), 129.7 (1C), 117.7 (1C), 114.1 (1C), 79.3 (1C), 44.9 (1C), 28.5 (1C), 28.5 (3C) ppm.

tert-Butyl (2-Aminobenzyl)carbamate (18c).

The reaction was carried out according to general procedure IV using 3-aminobenzylamine (1 g, 8.1gmmol), TEA (2.27 mL, 16.4 mmol), and di-tert-butyl decarbonate (1.79, 8.19 mmol). The product was used for the next step without further purification. The title compound (18c) was obtained as a white solid (1.74 g, 96%). Mp 85–88 °C. ¹H NMR (400 MHz, CDCl₃): δ = 7.12 (td, J = 7.7, 1.5 Hz, 1H), 7.13 (d, J = 6.9 Hz, 1H), 6.89 (t, J = 7.4 Hz, 2H), 4.91 (s, 1H), 4.36−4.16 (m, 4H), 1.45 (s, 9H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 157.7 (1C), 146.2(1C), 132.6 (1C), 129.0 (1C), 124.2 (1C), 119.0 (1C), 116.7 (1C), 79.7 (1C), 42.3 (1C), 26.7 (3C) ppm.

tert-Butyl (4-Aminobenzyl)carbamate (18d).

The reaction was carried out according to general procedure IV using 4-aminobenzylamine (1 g, 8.19 mmol), TEA (2.27 mL, 16.4 mmol), and di-tert-butyl decarbonate (1.79, 8.19 mmol). The product was used for the next step without further purification. The title (18d) compound was obtained as a white solid (1.86 g, quant). Mp 107–110 °C. ¹H NMR (400 MHz, CDCl₃): δ = 7.07 (d, J = 8.2 Hz, 2H), 6.64 (d, J = 8.3 Hz, 2H), 4.71 (s, 1H), 4.18 (d, J = 5.1 Hz, 2H), 3.62 (s, J = 55.5 Hz, 2H), 1.45 (s, 9H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 156.0 (1C), 145.8 (1C), 138.7 (1C), 129.0 (2C), 115.3 (2C), 79.38 (1C), 44.51 (1C), 28.57 (3C) ppm.

tert-Butyl (4-Nitrosobenzyl)carbamate (19d).

The reaction was carried out according to general procedure I, using 18d (200 mg, 0.89 mmol) and MPS (0.55 g, 1.8 mmol). The crude product was immediately used without further purification for the next reaction step.

tert-Butyl (E)-(3-(Phenyldiazenyl)benzyl)carbamate (20b).

The reaction was carried out according to general procedure V, using 18b (100 mg, 0.45 mmol) and nitrosobenzene (53 mg, 0,49 mmol). The crude product was purified by column chromatography using DCM to afford the title product (20b) as an orange oil. (94 mg, 68%). ¹H NMR (400 MHz, CDCl₃): δ = 8.00−7.95 (m, 2H), 7.90−7.87 (m, 2H), 7.68−7.70 (m, 4H), 7.64−7.38 (m, 1H), 5.29−5.13 (m, 1H), 4.63 (d, J = 5.1 Hz, 2H), 1.74 (s, 9H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 159.3 (1C), 154.0 (1C), 155.7 (1C), 143.1 (1C), 134.4 (2C), 132.6 (1C), 131.5 (1C), 131.3 (1C), 130.6 (2C), 125.0 (1C), 123.2 (1C), 81.5 (1C), 47.3 (1C), 30.3 (3C) ppm.

tert-Butyl (E)-(2-(Phenyldiazenyl)benzyl)carbamate (20c).

The reaction was carried out according to general procedure V, using 18c (100 mg, 0.45 mmol) and nitrosobenzene (53 mg, 0,49 mmol). The crude product was purified by column chromatography using DCM to afford the title product (20c) as an orange oil. (80 mg, 58%). ¹H NMR (400 MHz, CDCl₃): δ = 7.87 (d, J = 7.2 Hz, 2H), 7.69 (d, J = 7.9 Hz, 1H), 7.57−7.50 (m, 4H), 7.46−7.41 (m, 1H), 7.40−7.36 (m, 1H), 5.08 (s, 1H), 4.85 (d, J = 5.0 Hz, 2H), 1.47 (s, 9H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 156.0 (1C), 152.9 (1C), 150.1 (1C), 138.11 (2C), 131.5 (1C), 131.4 (1C), 129.8 (2C), 129.3 (1C), 128.3 (1C), 123.2 (1C), 116.0 (1C), 79.5 (1C), 41.2 (1C), 28.6 (3C)ppm.

tert-Butyl (E)-(4-((1,3-Dihydroisobenzofuran-5-yl)diazenyl)-benzyl)carbamate (20d).

The reaction was carried out according to general procedure V, using (1,3-dihydroisobenzofuran-5-yl)-methanamine (110 mg, 0.81 mmol) and crude 19d (0,89 mmol). The crude product was purified by column chromatography using DCM to afford the title product (20d) as an orange oil. (177 mg, 62%). ¹H NMR (400 MHz, CDCl₃): δ = 7.85 (t, J = 6.1 Hz, 3H), 7.75 (s, 1H), 7.40 (d, J = 8.2 Hz, 2H), 7.34 (d, J = 8.0 Hz, 1H), 5.14 (s, 4H), 5.08 (s, 1H), 4.37 (d, J = 5.2 Hz, 2H), 1.47 (s, 9H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 155.7 (1C), 152.4 (1C), 151.6 (1C), 141.9 (1C), 140.1 (1C), 127.8 (1C), 123.6 (2C), 122.9 (2C), 121.2 (2C), 113.9 (1C), 79.4 (1C), 73.1 (2C), 44.1 (1C), 28.2 (3C) ppm. LC: t_R (min) = 9.10 (cis) + 10.38 (trans), purity = 87.7%. MS: [M + H]⁺ calcd for [C₂₀H₂₃N₃O₃]⁺ = 354.18, found 354.13.

(E)-(3-(Phenyldiazenyl)phenyl)methanamine (21b).

The reaction was carried out according to general procedure VI, using 20b (80 mg, 0.26 mmol) dissolved in TFA (500 μL) and DCM (4500 μL) to afford the title product (21b) as an orange oil. No further purification was performed (57 mg, quant.). ¹H NMR (400 MHz, CDCl₃): δ = 7.85−7.80 (m, 2H), 7.68 (s, 1H), 7.64 (d, J = 7.7 Hz, 1H), 7.43−7.39 (m, 5H), 3.74 (s, 2H), 1.41 (s, 2H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 154.0 (1C), 153.6 (1C), 142.1 (1C), 130.0 (1C), 129.2 (1C), 129.0 (2C), 128.8 (1C), 121.7 (2C), 121.0 (1C), 118.9 (1C), 45.3 (1C) ppm. LC: t_R (min) = 5.24 (cis) + 7.26 (trans), purity = 81.6%. MS: [M + H]⁺ calcd for [C₁₃H₁₃N₃]⁺ = 212.12, found 212.20.

(E)-(2-(Phenyldiazenyl)phenyl)methanamine (21c).

The reaction was carried out according to general procedure VI, using 20c (80 mg, 0.26 mmol) dissolved in TFA (500 μL) and DCM (4500 μL) to afford the title product (21c) as an orange oil. No further purification was performed (55 mg, quant). ¹H NMR (400 MHz, CDCl₃): δ = 7.99 (d, J = 7.3 Hz, 2H), 7.80 (d, J = 7.9 Hz, 1H), 7.60−7.49 (m, 5H), 7.42−7.39 (m, 1H), 3.72 (s, 2H), 1.52 (s, 2H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 152.8 (1C), 150.0 (1C), 142.6 (1C), 131.5 (1C), 131.2 (2C), 129.2 (1C), 129.1 (2C), 127.7 (1C), 123.0 (1C), 115.9 (1C), 43.5 (1C) ppm. LC: t_R (min) = 5.86 (cis) + 7.21 (trans), purity = 84.3%. MS: [M + H]⁺ calcd for [C₁₃H₁₃N₃]⁺ = 212.12, found 212.20.

(E)-(4-((1,3-Dihydroisobenzofuran-5-yl)diazenyl)phenyl)-methanamine (21d).

The reaction was carried out according to general procedure VI, using 20d (170 mg, 0.67 mmol) dissolved in TFA (500 μL) and DCM (4500 μL). The crude material was purified by column chromatography (DCM/MeOH/NH₃ = 10/1/0.1) to afford the title product (21d) as an orange oil. (119 mg, 70%). ¹H NMR (400 MHz, DMSO): δ = 7.97−7.90 (m, J = 12.6, 5.0 Hz, 3H), 7.88 (s, 1H), 7.69 (d, J = 8.4 Hz, 2H), 7.61 (d, J = 8.0 Hz, 1H), 5.16 (s, 4H), 4.02 (s, 2H) ppm. ¹³C NMR (101 MHz, DMSO): δ = 153.8 (1C), 153.0 (1C), 146.1 (1C), 144.6 (1C), 142.7 (1C), 130.7 (2C), 125.5 (1C), 124.5 (2C), 124.0 (1C), 116.1 (1C), 74.4 (1C), 74.3 (1C), 46.0 (1C) ppm. LC: t_R (min) = 5.49 (cis) + 7.18 (trans), purity = 91.3%. MS: [M + H]⁺ calcd for [C₁₅H₁₅N₃O]⁺ = 254.13, found 254.05.

5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-N-((3-[(1E)-2-phenyldiazen-1-yl]phenyl)methyl)pyrazole-3-carboxamide (16b).

The reaction was carried out according to general procedure VII, using rimonabant carboxylic acid (58 mg, 0.15 mmol, 1 equiv), TEA (43 μL, 0.31 mmol, 2.1 equiv), ECF (29 μL, 0.3 mmol, 2.05 equiv), and 21b (74 mg, 0.35 mmol, 2.4 equiv). The product was purified by prep-TLC (petroleum ether/ethyl acetate = 4/1) to afford the target compound as an orange oil (82 mg, 95%). ¹H NMR (400 MHz, DMSO): δ = 7.87 (dd, J = 8.0, 1.6 Hz, 3H), 7.84−7.72 (m, 3H), 7.64−7.51 (m, 6H), 7.50−7.41 (m, 2H), 7.28−7.04 (m, 3H), 4.54 (d, J = 6.2 Hz, 2H), 2.25 (s, 3H). ¹³C NMR (101 MHz, DMSO): δ = 162.2(1C), 151.9(1C), 144.6(1C), 142.6(1C), 141.5(1C), 135.8(1C), 135.1(1C), 133.8(1C), 132.2(1C), 132.0 (1C), 131.5(1C), 131.3(2C), 130.7(1C), 129.6(1C), 129.5(2C), 129.3(1C), 128.8(1C), 128.7(1C), 128.3(1C), 127.2(1C), 122.5(2C), 121.4(1C), 121.2(1C), 119.7(1C), 116.5(1C), 41.7(1C), 9.1 (1C). LC: t_R (min) = 10.99 (cis) + 11.92 (trans), purity = 98.6%. MS: m/z [M + H]⁺ calcd for (C₃₀H₂₂Cl₃N₅O₂)⁺ = 574.09, found 574.10.

5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-N-((2-[(1E)-2-phenyldiazen-1-yl]phenyl)methyl)-1H-pyrazole-3-carboxamide (16c).

The reaction was carried out according to general procedure VII, using rimonabant carboxylic acid (50 mg, 0.13 mmol, 1 equiv), TEA (36 μL, 0.26 mmol, 2.1 equiv), ECF (25 μL, 0.26 mmol, 2.05 equiv), and 21c (63 mg, 0.3 mmol, 2.4 equiv). The crude product was purified by column chromatography (petroleum ether/ethyl acetate = 4/1) to afford the target compound as an orange oil (69 mg, 92%). Mp 73 °C. ¹H NMR (400 MHz, DMSO): δ = 7.96 (q, J = 6.5, 3.3 Hz, 2H), 7.82−7.70 (m, 2H), 7.68−7.50 (m, 6H), 7.50−7.34 (m, 3H), 7.34−7.11 (m, 3H), 7.05−6.94 (m, 1H), 4.90 (d, J = 101.7, 6.0 Hz, 2H), 2.27 (s, 3H). ¹³C NMR (101 MHz, DMSO): δ = 162.0(1C), 152.2(1C), 149.0(1C), 144.6(1C), 142.6(1C), 138.5(1C), 135.7(1C), 135.0(1C), 133.8(1C), 132.1(1C), 131.9(1C), 131.5(1C), 131.2(2C), 129.6(1C), 129.4(2C), 128.8(2C), 128.7(1C), 128.3(1C), 127.6(1C), 127.2(1C), 122.8(2C), 119.9(1C), 116.4(1C), 115.0(1C), 38.1(1C), 9.1(1C). LC: t_R (min) = 11.09 (cis) + 12.52 (trans), purity = 98.9%. MS: m/z [M + H]⁺ calcd for (C₃₀H₂₂Cl₃N₅O₂)⁺ = 574.09, found 574.10.

5-(4-Chlorophenyl)-1-(2,4-dichlorophenyl)-N-((4-[(1E)-2-(1,3-dihydro-2-benzofuran-5-yl)diazen-1-yl]phenyl)methyl)-4-methyl-1H-pyrazole-3-carboxamide (16d).

The reaction was carried out according to general procedure VII, using rimonabant carboxylic acid (20 mg, 0.05 mmol, 1 equiv), TEA (15 μL, 0.10 mmol, 2.1 equiv), ECF (10 μL, 0.10 mmol, 2.05 equiv), and 21d (33 mg, 0.05 mmol, 1 equiv). The crude product was purified by prep-TLC (petroleum ether/ethyl acetate = 4/1) to afford the target compound as an orange oil (27 mg, 88%). Mp 102 °C. ¹H NMR (400 MHz, CDCl₃): δ = 7.77 (dd, J = 12.9, 4.9 Hz, 3H), 7.67 (s, J = 11.8 Hz, 1H), 7.42 (d, J = 8.3 Hz, 2H), 7.36−7.25 (m, 3H), 7.23−7.13 (m, 4H), 7.03−6.93 (m, 2H), 5.13−4.85 (m, 4H), 4.61 (d, J = 6.1 Hz, 2H), 2.31 (d, J = 10.9 Hz, 3H). ¹³C NMR (101 MHz, CDCl₃): δ = 162.8 (1C), 152.8 (1C), 152.0 (1C), 144.9 (1C), 143.3 (1C), 142.3 (1C), 141.79 (1C), 140.5 (1C), 136.15 (1C), 136.0 (1C), 135.1 (1C), 133.1 (1C), 130.9 (2C), 130.6 (1C), 130.5 (1C), 129.1 (2C), 128.7 (2C), 128.0 (1C), 127.3 (1C), 124.0 (1C), 123.3 (2C), 121.6 (1C), 118.1 (1C), 114.3 (1C), 73.5 (2C), 42.8 (1C), 9.6 (1C). LC: t_R (min) = 10.80 (cis) + 11.56 (trans), purity = 99.1%. MS: m/z [M + H]⁺ calcd for (C₃₂H₂₄Cl₃N₅O₂)⁺ = 616.10, found 616.15.

Rimonabant Hydrochloride.

Five drops of DMF were added to a stirred solution of rimonabant carboxylic acid (1 g, 2.62 mmol, 1 equiv) in DCM (10 mL). The reaction mixture was cooled to 0 °C, and then (COCl)₂ (0.37 mL, 4.45 mmol, 1.7 equiv) was added dropwise. The reaction mixture was stirred for 1 h at room temperature. Excess of (COCl)₂ was removed under reduced pressure. In another flask were introduced N-aminopiperidine (0.43 mL, 3.93 mmol, 1.5 equiv) and TEA (0.36 mL, 2.62 mmol, 1 equiv) in DCM, and the mixture was cooled to 0 °C. Subsequently, to the mixture was added dropwise a solution of the intermediate acyl chloride in DCM at the same temperature. The reaction mixture was stirred at 0 °C for 15 min and for 15 min more at room temperature. The mixture was diluted with DCM, washed with NaHCO₃ and brine, dried over Na₂SO₄, and concentrated under reduced pressure, yielding rimonabant without further purification as a pale-yellow solid (1.2 g, quant). The product was treated with a solution of HCl in 2-propanol. The precipitate was collected and washed with anhydrous diethyl ether to obtain the respective rimonabant hydrochloride salt as a beige powder. Mp 225–227 °C (dec). ¹H NMR (400 MHz, CDCl₃): δ = 7.62 (s, 1H), 7.42 (d, J = 1.9 Hz, 1H), 7.32−7.27 (m, 4H), 7.05 (d, J = 8.3 Hz, 2H), 2.94−2.80 (m, 4H), 2.36 (s, 3H), 1.75 (dt, J = 11.1, 5.6 Hz, 4H), 1.41 (t, J = 14.5 Hz, 2H) ppm. ¹³C NMR (101 MHz, CDCl₃): δ = 160.4 (1C), 144.9 (1C), 143.4 (1C), 136.4 (1C), 135.4 (1C), 133.5 (1C), 132.9 (1C), 131.3 (2C), 131.0 (1C), 130.8 (1C), 129.4 (2C), 128.4 (1C), 127.7 (1C), 118.7 (1C), 57.6 (2C), 25.9 (2C), 23.8 (1C), 9.8 (1C) ppm. LC: t_R = 10.68 min, purity = 99.9%. MS: [M]⁺ calcd for [C₂₂H₂₁Cl₃N₄O] ⁺ = 463.09, found 463.05.

Molecular Docking.

The 2D structures of the compounds were drawn with ChemDraw (version 18.0, PerkinElmer) and Marvin-Sketch for Windows (version 19.10.0, ChemAxon Ltd.). Subsequently, the input files for the molecular docking studies were generated. The crystal structure of the CB₁R (PDB code 5TGZ) was prepared for docking using MOE (version 2016.08).⁵⁷ Removal of ligands and water molecules was followed by automatic assignment of tautomer and protonation states, using MOE’s protein structure preparation pipeline^58–63 and Protonate3D⁶⁴ at pH 7.5. Pose reproduction performance was assessed by RMSD between docked poses and the experimental binding pose of AM6538 (PDB code 5TGZ). Geometry optimization of AM6538 was performed using the MMFF94x^65,66 force field (gradient convergence criterion: RMS 0.001·kcal·mol⁻¹·Å⁻¹). The atoms on the nitrate group of AM6538 were not considered as they were not visible in the electron density. The GOLD program^52,53 (version 5.4.1) was used to dock the ligand under default settings. The knowledge-based scoring function ASP⁵⁴ was found to robustly model experimentally observed intermolecular interactions and to yield the best crystal pose reproduction among all of GOLD’s scoring functions. The 50 best ASP poses were rescored using the knowledge-based scoring function DSX⁵⁵ (version 0.89), with CSD and PDB potentials (version 05/11). The cis and trans isomers of compound 16a were manually prepared in MOE’s Builder tool and subsequently prepared and evaluated under the same conditions as AM6538. All figures representing structures and superimpositions were prepared using PyMOL (The PyMOL Molecular Graphics System, version 1.8.0.6, Schrödinger, LLC).

Biological Assays.

Cell Luminescence Assay.

The assay was established in a similar way as previously described.⁶⁷ HEK-293 cells stably expressing hCB₁R or hCB₂R were cultured in DMEM supplemented with FBS 10%, l-glutamine 2 mM, and penicillin/streptomycin 1%. All cells were maintained at 37 °C and 5% CO₂ in a humidified atmosphere. The CRE luciferase reporter contains the firefly luciferase gene under the control of multimerized cAMP response element (CRE) located upstream of a minimal promoter. The cells were treated with the compounds (previously isomerized to their cis or trans states with light of 366 or 454 nm, respectively) for 5h.

For hCB₁R agonistic activity, HEK-293-CB₁-CRELuc cells were stimulated with CP55940 1 μM as a positive control. For hCB₁R antagonistic activity, HEK-293-CB₁-CRELuc cells were pretreated with the test compounds for 30 min and then stimulated with CP55940 1 μM for 5 h. As a positive control, the cells were stimulated with rimonabant 1 μM. For hCB₂R agonistic assay, HEK-293-CB₂-CRELuc cells were pretreated with the test compounds for 30 min and then stimulated with forskolin 10 μM for 5 h. As a positive control the cells were stimulated with CP55940 1 μM. For hCB₂R antagonistic assay, HEK-293-CB₂-CRELuc cells were pretreated with test compounds for 30 min and then stimulated with forskolin 10 μM + CP55940 1 μM for 5 h. As positive controls, the cells were stimulated with SR144528 10 μM and AM630 10 μM.

After stimulation and incubation, the cells were dissolved in a solution containing tris-phosphate 25 μmol/L (pH = 7.8); MgCl₂, 8 μmol/L; dithiothreitol, 1 μmol/L; Triton X-100, 1%; and glycerol, 7%. The luciferase activity was measured on lysed cells with the luciferase assay kit (Promega) according to the manufacturer’s protocol in a Tristar² LB942 multimode reader machine (Berthold Technologies). To avoid isomerization reversion, all the experiments were carried out in the dark. Statistical analyses were performed using one-way ANOVA followed by Dunnett’s multiple comparison post-tests using GraphPad Prism 6 for Windows (version 6.01, September 21, 2012).

Radioligand Binding Assay.

For the experiments, hCB₂-HEK cells were a kind gift from AbbVie Laboratories (Chicago, U.S.). Cells were grown in Dulbecco’s modified Eagle’s medium containing high glucose supplemented with 8% fetal calve serum and 25 μg/mL zeocin in a 37 °C incubator in the presence of 5% CO₂. The respective hCB₂R membranes were prepared as described in the literature.³³ The rCB₁R membrane homogenate was prepared from brains of adult female rats.^15,68 The tissues were obtained from an alternate project of the Institute of Pharmacology and Toxicology at the University of Würzburg and frozen at −80 °C. The respective rCB₁R membranes were freshly prepared according to the protocol described by Catani and Gasperi for preparation of membrane homogenates.⁶⁹

Saturation and competition assays were done according to the protocol previously established in M. Decker’s research group.³³ For the determination of K_D value of the membrane samples, saturation assays were done with 8 concentrations of [³H]CP55940 (Hartmann Analytic GmbH) ranging from 0.088 nM to 4.4 nM. Reactions were started by adding membrane (25 μg/well for rCB₁ or 8 μg/well for hCB₂) of a 96 well Multiscreen filter plate (Millipore) containing the radioligand in assay buffer (50 mM Tris-HCl, pH = 7.4; 5 mM MgCl₂·6H₂O; 2.5 mM EDTA; 2 mg/mL BSA). After 3 h incubation at room temperature, the reaction was stopped by vacuum filtration and each well was washed 4 times with cold binding buffer (50 mM Tris-HCl, pH = 7.4; 5 mM MgCl₂·6H₂O; 2.5 mM EDTA). The filter plate was dried at 45 °C. Lately, 20 μL of IRGA Safe plus-scintillation cocktail (PerkinElmer) was added to each well. The activity was counted in a Micro Beta Trilux counter (Wallac). Competition assays were done with 5–10 concentrations (0.1 nM to 0.4 mM) of target compound (previously isomerized to their cis or trans states with light of 366 or 454 nm, respectively) and 0.6 nM [³H]CP55940. The positive controls for the assays over CB₁ and CB₂ were respectively the selective ligands rimonabant and MN-I-79 (both synthesized in house).

The stock solutions of all tested compounds (4 mM) were prepared by dissolving in DMSO. The dilution series of all stock solutions were prepared diluting with binding buffer.

Statistical analyses and sigmoidal dose–response curve fittings were performed with GraphPad Prism 6 for Windows (version 6.01, September 21, 2012).

K_i values were determined according to the Cheng–Prusoff equation:

$$K_i=\frac{{\text{EC}}_{50}}{1+\frac{\left[\text{L}*\right]}{K_{\text{D}}}}$$

with [L*] as radioligand concentration. The K_i value was calculated from at least two individual experiments. K_D values were measured in at least two individual experiments for each new batch of prepared membrane.

hCB₁ Calcium Mobilization Assay.

By use of a functional fluorescent hCB₁ activated Gα_q16-coupled intracellular calcium mobilization assay in CHO-K1 cells, the antagonist effect of the target compound was characterized pharmacologically as described previously in the literature and apparent affinity (K_e) values were determined.^50,51 Calcium flux was monitored in a 96-well format using the fluorescent calcium-5 dye in an automated plate reader (FLIPR Tetra, Molecular Devices). The antagonist activity of 16a was measured by its ability to shift the concentration–response curve of the reference CB₁ agonist CP55940 rightward using the equation

$$K_{\text{e}}=\left[\text{ligand}\right]/\left[\text{DR}-1\right]$$

where DR is the EC₅₀ ratio of CP55940 in the presence or absence of a test agent. Statistical analysis was performed using GraphPad Prism 6 for Windows (version 6.01, September 21, 2012).

ABBREVIATIONS

hCB₁R

human cannabinoid receptor 1

hCB₂R

human cannabinoid receptor 2

rCB₁R

rat cannabinoid receptor 1

DCM

dichloromethane

DMF

dimethylformamide

ECF

Ethyl chloroformate

MPS

potassium peroxymonosulfate

NBS

N-bromosuccinimide

TEA

triethylamine

TFA

tri-fluoroacetic acid

THF

tetrahydrofuran HBTU (2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate

TLC

thin layer chromatography