Tricyclic Pyrazole-Based Compounds as Useful Scaffolds for Cannabinoid CB1/CB2 Receptor Interaction

Cannabinoids comprise different classes of compounds, which aroused interest in recent years because of their several pharmacological properties. Such properties include analgesic activity, bodyweight reduction, the antiemetic effect, the reduction of intraocular pressure and many others, which appear correlated to the affinity of cannabinoids towards CB1 and/or CB2 receptors. Within the search aiming to identify novel chemical scaffolds for cannabinoid receptor interaction, the CB1 antagonist/inverse agonist pyrazole-based derivative rimonabant has been modified, giving rise to several tricyclic pyrazole-based compounds, most of which endowed of high affinity and selectivity for CB1 or CB2 receptors. The aim of this review is to present the synthesis and summarize the SAR study of such tricyclic pyrazole-based compounds, evidencing, for some derivatives, their potential in the treatment of neuropathic pain, obesity or in the management of glaucoma.


Introduction
Cannabinoids are a class of different chemical compounds (Figure 1), including the endocannabinoids (produced naturally in the body by humans and animals such as anandamide and 2-arachidonoylglycerol), the phytocannabinoids (derived from Cannabis, exemplified by ∆ 9 -tetrahydrocannabinol, the major psychoactive component of Cannabis sativa, commonly known as marijuana), and the synthetic cannabinoids (produced chemically by human, comprising a wide array of chemical entities, i.e., the pyrazole-based compounds SR141716A and SR144528 and the indole based compound WIN-55,212-2) [1][2][3]. The physiological and behavioral effects of cannabinoids appear directly correlated to their affinity towards two different classes of specific receptors: CB 1 receptors located predominantly in the central nervous system [4], and CB 2 receptors which are mostly found in peripheral tissues [5]. In the brain, the CB 1 receptors are abundantly expressed in the hippocampus, cerebellum and striatum [6,7]. Among the peripheral tissues wherein the CB 1 receptors have been found, the enteric nervous system [8], testis, urinary bladder, vas deferens can be mentioned [9]. CB 1 receptor mRNA and protein have been furthermore identified in the rat and human eye, both in the retina and in the iris, and in the ciliary body [10]. CB 2 receptors are located in the marginal zones of the spleen, tonsils, immune cells [9,11] and to a much lesser extent in CNS [12]. Furthermore, CB 2 receptor mRNA was expressed in the adult rat retina, including the somas of retinal ganglion cells [13].
During the last years, several ligands endowed with high affinities and subtype selectivity for both receptors were identified. Such ligands were proposed as potential therapeutic targets for the treatment of several diseases, including neuropathic pain [14], cancer [15,16] and osteopor4osis [17].
Non-selective CB 1 /CB 2 receptor agonists are the constituents of some approved medicines, i.e., Sativex ® , Cesamet ® , Marinol ® . Sativex contains approximately equal The only cannabinoid receptor antagonist approved as a medicine to-date is the CB 1 antagonist/inverse agonist SR141716A (rimonabant, Acomplia ® , Figure 1). This compound has been developed for the treatment of obesity and related metabolic risk factors [20]. However, it was soon withdrawn for its serious psychiatric disorders including anxiety, depression and suicidal tendency. Although most pharmaceutical companies were deterred from developing a drug that displayed rimonabant-like CB 1 receptor antago-nist/inverse agonist activity for the management of any disorders, this compound still remains an extremely valuable lead for the design of new ligands for CB receptors interaction [21,22]. Within this frame, the 4-alkyl-5-arylpyrazole skeleton of rimonabant has been modified, leading to benzocycloalkylpyrazole-based tricyclic systems of general formula I-IV ( Figure 2). Modifications carried out on such scaffolds, allowed to accede to further pyrazole-based tricyclic systems of general formula V-XVII ( Figure 2). Fine tuning of these tricyclic systems (vide infra) allowed to identify hundreds of molecules, most of which endowed with high affinity and selectivity for CB 1 or CB 2 receptors. In this review, pyrazole-based tricyclic compounds I-XVII are divided into four main groups according to the size of the central ring connected to the pyrazole one. If a five-membered ring is connected to the pyrazole, the compound belongs to (5,5)-condensed pyrazole derivatives, and if the connected ring is a six-, seven-or eight-membered ring, the compound belongs to (5,6)-, (5,7)-or (5,8)-condensed pyrazole derivatives. Here, we briefly present the synthesis and the bioactivities of pyrazole-based tricyclic compounds I-XVII investigated by us and other groups. Furthermore, this review surveys chemical and biological literature of some miscellaneous molecules featuring tricyclic pyrazole structure closely related to compounds I-XVII, investigated for cannabinoid receptor affinity. The first (5,5)-condensed pyrazole derivatives synthesized by us for cannabinoid receptors interaction are those of general formula I (Figure 2) featuring the 1,4-dihydroindeno(1,2c)pyrazole core. These compounds were designed as rigid analogs of CB 1 antagonist/inverse agonist SR141716A (rimonabant), endowed with a high affinity for CB 1 receptors and good selectivity over CB 2 receptors. It has been proposed that minimization of the flexibility of the lead compound through the introduction of structural constraints could have some impact on cannabinoid receptor interaction, maybe allowing the ligand to bind with high affinity and selectivity to its receptor. Accordingly, the first series of 1,4-dihydroindeno(1,2c)pyrazole derivatives have been synthesized and evaluated for their in vitro binding affinities for CB 1 and CB 2 receptors. Most of these compounds are N-piperidine-carboxamides that differ for R substituent (Cl, F, I, CH 3 , OCH 3 ), whereas G, quite often, maintained the 2,4-Cl-phenyl substitution pattern of rimonabant [23].
Compounds I were synthesized starting from the appropriate ketones 1 (Scheme 1) which were α-acylated by Claisen reaction, furnishing 1,3-diketoesters 2 as a tautomeric equilibrium shifted towards the alkenylidene structure (2 ). The cyclization with appropriate hydrazines in refluxing EtOH gave the tricyclic dihydroindeno(1,2-c)pyrazole esters 3, which were hydrolyzed with KOH, affording the corresponding acids 4. Target compounds I were synthesized by condensation of acids 4, previously activated to acyl chlorides with SOCl 2 , with the appropriate amines. Scheme 1. General synthetic route for the preparation of compounds I.
If not otherwise stated, the four-step general synthetic route for I compounds, with minor changes, was employed for the preparation of all compounds of general formula II-XVII, starting from the appropriate ketones and using the appropriate hydrazines and amines.
Our first SAR study revealed that several compounds incorporating the planar 1,4dihydroindeno(1,2-c)pyrazole scaffold displayed very high in vitro binding affinity for CB 2 receptors comparable to, or exceeding, that of SR144528 (Figure 1), identified as the first highly potent and selective ligand for the CB 2 receptors. Representative derivatives are IAaX, IAhX, IAfX, IBaX (Figure 3), compound IAhX emerging for its high affinity for CB 2 receptors and exceptional selectivity over CB 1 receptor (CB 1 /CB 2 selectivity ratio of 9811).
Our results prompted us to investigate new 1,4-dihydroindeno(1,2-c)pyrazoles (I), which were obtained modifying the N-carboxamide moiety (Q) and the aryl substitution of the lead compounds IAaX and IAhX [24]. In general, our SAR study showed that the CB 1 receptor affinities of all the investigated compounds were lower than their CB 2 receptor affinities, evidencing the suitability of 1,4-dihydroindeno(1,2-c)pyrazole architecture to generate ligands for CB 2 receptor interaction. Representative compounds of this series are IDhX, IEhX, IFhX, IGhX (Figure 4), all displaying high CB 2 affinity and reasonable selectivity over CB 1 receptors, even if lower than that elicited by compounds IAaX and IAhX. Affinities at CB 1 and CB 2 receptors were assessed by competition of ( 3 H)-CP 55,940 in mouse cerebellum membranes and in mouse spleen homogenates, respectively [23]. Permission has been obtained and there is no copyright issue. In vitro CB 2 intrinsic activity evaluation assay, based on the determination of P-ERK 1/2 increasing expression in human promyelocitic leukemia HL-60 cells exposed to the compounds to be assayed, highlighted agonist activity toward CB 2 receptors for derivatives IDhX, IEhX and for prototype IAaX, IAhX. In particular, the tested compounds significantly increased P-ERK 1/2 expression, reaching the maximum effect at the concentration of 10 nM with the following values: +61.3 ± 12.4% (IDhX), +125.0 ± 35.8% (IEhX), +65.0 ± 18.5% (IAaX), +83.0 ± 4.2% (IAhX) versus vehicle. Overall, the effect of tested compounds as CB 2 agonists was specific as it was blocked by the CB 2 receptor antagonist SR144528 [24].
By pursuing our interest in expanding SAR studies around 1,4-dihydroindeno(1,2c)pyrazole core, especially in the light of the potential of CB 2 ligands for the treatment of immune disorders or as anti-nociceptive agents [14,25], novel derivatives have been designed, making use of molecular hybridization based on scaffold hopping [26]. In particular, based upon the putative interacting sites and structural features of selective CB 2 antagonist SR144528 (Figure 1), i.e., N 1 -benzyl group, the C 3 carboxamide moiety, and the substitution of the C 5 phenyl ring [26,27], it was postulated that the introduction of such pharmacophoric elements in the tricyclic core of compound IAhX might provide new CB 2 ligands with potential therapeutic value. Briefly, different synthesized compounds were monoterpene derivatives incorporating bulky groups in the carboxamide moiety (i.e., fenchyl-, bornyl-, isopinocampheyl-, myrtanyl-, menthyl-), bearing a 6-CH 3 ,7-Cl or 6-Cl,7-CH 3 substitution pattern at the aryl ring of 1,4-dihydroindeno(1,2-c)pyrazole core. Some adamantane derivatives were also investigated, together with compounds incorporating simple cycloalkyl motifs at the carboxamide moiety. Compounds IAjX, IAkX, IHjX, IIjX, IKjX, IDjX, IDkX, IHkX ( Figure 5) are representatives of these third series of I compounds. As described above for representative compounds depicted in Figures 3 and 4, our SAR study carried out on a third series of molecules, evidenced for 1,4-dihydroindeno(1,2-c)pyrazole-based molecules preferential affinity for CB 2 receptors. However, the introduction of a chlorine atom, as well as its exchange with the methyl group in all-new hybrid compounds seemed to play a modest role in lowering the levels of CB 2 -affinity as compared to the reference compounds IAaX and IAhX. Figure 5. Structure of representative 1,4-dihydroindeno(1,2-c)pyrazoles (I) designed making use of molecular hybridization based on scaffold hopping. Affinities at CB 1 and CB 2 receptors were assessed by competition of ( 3 H)-CP 55,940 in mouse cerebellum membranes and in mouse spleen homogenates, respectively [26].
According to previous data obtained for compounds IAaX, IAhX, IDhX and IEhX, analogs IIjX, IDjX and IHkX, endowed with the highest affinity for CB 2 receptors, exhibited CB 2 agonism activity in in vitro model based on the determination of P-ERK 1/2 increasing expression in HL-60 cells, with maximum values reached at 125 nM for compounds IIjX (+54.5 ± 12.1%) and IDjX (+82.5 ± 19.1%), and at 50 nM for compound IHkX (+46.4 ± 15.6%) versus vehicle. Furthermore, the effect of such tested compounds on P-ERK 1 2 expression was counteracted by the reference CB 2 antagonist AM630, suggesting the correspondence between the detected effect and CB 2 modulation.
The obtained data confirmed that the flattening of 1,4-dihydroindeno(1,2-c)pyrazole core is important to assure agonist rather than antagonist activity, with respect to not condensed and more flexible analogs such as the antagonist SR144528.
Within this frame, a series of compounds incorporating the 1,4-dihydroindeno(1,2c)pyrazole scaffold (I), sharing the N 1 -benzyl group and bulky groups in the carboxamide moiety (i.e., IOkW, Figure 2) were claimed by Sanofi-Aventis, but no specific biological activity was presented [28]. However, the compounds of the invention were generally described as potent and selective CB 2 receptor antagonists with Ki values < 5 × 10 −7 M. Antagonistic properties of such compounds have been demonstrated by the results in the models for inhibition of adenylate cyclase induced by forskolin, although no specific examples or data were disclosed.
Furthermore, 1,4-dihydroindeno(1,2-c)pyrazoles featuring a cyclopropyl or cyclohexyl building block in C 6 position were investigated by us, in order to evaluate the effect of cycloalkyl moiety in place of methyl group both on cannabinoid receptor binding and activity [29]. The most interesting compounds coming from this fourth series are depicted in Figure 6. Whereas, these analogs provided further insight regarding the structural features for CB 2 affinity and selectivity, our data evidenced that the introduction of a cyclopropyl moiety in most of the synthesized compounds seemed to play a modest role in lowering the CB 2 receptor affinity, especially if compared to compound IAhX. Interestingly, intrinsic activity for compounds IAlX, IHlX, ILlX, IJlX, evaluated by GTPγS binding assay, showed antagonist/inverse agonist properties (IC 50 for compound IAlX = 294 nM, for IHlX = 80 nM, for ILlX = 27 nM and for IJlX = 51 nM). Independently, other groups investigated the biological properties of some 1,4-dihydroindeno(1,2-c])pyrazoles synthesized by our group. In particular, within the search aiming to characterize the molecular pharmacology of the most widely used CB 2 receptor ligands, it was reported for compound IAhX, namely Gp-1a, an affinity for the CB 2 receptors markedly different from that reported by our group (K i CB 1 = 426 ± 0.08 nM, K i CB 2 = 20.9 ± 0.23 nM; CB 1 /CB 2 selectivity ratio = 20), using ( 3 H)CP-55,940 and mouse brain and spleen as source for CB 1 and CB 2 receptors, respectively [30]. Furthermore, in contrast to our results, the same authors reported for Gp-1a inverse agonist properties on CB 2 receptors in GTPγS assay and resulted not active in hCB 2 receptor pERK assay [30]. Indeed, the first two series of 1,4-dihydroindeno(1,2-c)pyrazole-based compounds synthesized by our group [23,24], including Gp-1a, were described in a patent application providing methods and pharmaceutical compositions for reducing the serum level of immunoglobulin IgE in an animal or human subject [31]. In particular, it was reported the CB 2 agonist Gp-1a attenuated the serum levels of total IgE from BALB/e mice and the co-treatment with the CB 2 antagonist SR144528 reversed the Gp-1a effect. Therefore, the patent provided a method for modulating this type of antibody for the treatment of immune system-related conditions such as allergy, hay fever and the like. Biological literature concerning Gp-1a pointed out the relevance of such a compound as a useful pharmacological tool to ascertain in more detail the role of CB 2 receptors in physiopathological conditions [32][33][34][35][36]. Furthermore, the pharmacological relevance of 1,4-dihydroindeno(1,2-c)pyrazole-based I compounds emerged also from the significant anti-nociceptive activity of the potent and selective CB 2 agonist IDhX, namely NESS400, in Spared Nerve Injury (SNI) neuropathic mice, by alleviating both mechanical allodynia and thermal hyperalgesia [37].

Benzofuro(3,2-c)pyrazol-Based Derivatives (V)
Further condensed (5,5)-pyrazole derivatives were designed by us making use of bioisosteric replacement as drug design approach to obtain novel planar tricyclic scaffold for cannabinoid receptor interaction [38]. Following this approach, a new series of 1,4dihydroindeno(1,2-c)pyrazole analogs, namely benzofuro(3,2-c)pyrazoles (V, Figure 2), containing an oxygen atom at position 4 instead of a methylene unit, were designed. The structure of representative compounds is depicted in Figure 7. These compounds were previously described by Neuroscienze Pharmaness S.C.A.R.L in a patent application claiming specific composition of microemulsions of pharmaceutical compounds. Indeed, the patent covers a more extensive series of tricyclic pyrazoles belonging both to (5,5) as well as (5,6) and (5,7]) series [39]. In general, our SAR study revealed high CB 2 receptor affinity for the new compounds, with values qualitatively similar with those of compound IDhX, as well as those of related analogs incorporating the 1,4-dihydroindeno(1,2-c)pyrazole skeleton. In contrast, a comparison of K i CB 2 value of compound IAhX and that of the analog VAhX revealed two orders of magnitude decreased affinity for VAhX. Compounds VLhX and VMhX, featuring a bornyl and isopinocampheyl moiety at the carboxamide portion, exhibited the best CB 2 cannabinoid binding profiles among all synthesized derivatives. CB 2 functional assay carried out on HL-60 cells, based on the evaluation of P-ERK1/2 expression induced by cannabinoid ligands, evidenced agonism behavior for all synthesized compounds [38], with calculated maximum values, in most cases, exceeding that of the agonist WIN-55,212-2 (i.e., VAhX: dose 50 nM, 205 ± 12%; VDhX, dose 50 nM, 196 ± 20%; VLhX: dose 10 nM, 189 ± 11%; VMhX: dose 5 nM, 205 ± 6%; WIN-55,212-2: dose 75 nM, 190 ± 17% versus vehicle). Overall, the effect of tested compounds as CB 2 agonists was specific as it was blocked by the CB 2 receptor antagonist AM630 [38]. To further investigate the versatility of (5,5)-pyrazole condensed tricyclic derivatives in the development of CB 2 ligands as therapeutic agents, our attention was focused on the benzene ring of the tricyclic indenopyrazole scaffold (I), by replacing it with a thiophene ring, giving rise the novel dihydrothienocyclopentapyrazole architecture which fine-tuning furnished novel derivatives with general formula VII (Figure 2). It was postulated that bioisosteric replacement benzene/thiophene could be an efficient strategy to develop novel CB 2 selective ligands and maybe provide further insight concerning structural features for cannabinoid receptor interaction [40]. In our SAR study, we planned to investigate the effect of changing the size of carboxamide moiety (Q), of methyl shifting from C 6 to C 7 of the tricyclic platform, as well as the effect related to the replacement of the N 1 -dichlorophenyl group (X) with p-methylbenzyl moiety (Z) on cannabinoid receptor affinity. The structure of representative compounds is depicted in Figure 8. The major term, compound VIIAhX, displayed a high affinity for CB 2 receptors, even if 62-fold decreased with respect to the reference compound IAhX with selectivity ratio K i CB 1 /K i CB 2 = 191. A similar trend was observed in several thienocyclopentapyrazole compounds (Figure 8) with the exception of VIINhX, which exhibited a mixed binding profile, reaching K i values of 6.5 and 5.1 nM for CB 1 and CB 2 , respectively. All these compounds profiled as full agonists at CB 2 receptors in an assay based on the determination of P-ERK 1/2 increasing expression in HL-60 cells [40]. Within this framework, a series of 1,4-dihydrothieno(3 ,2 -4,5)cyclopenta(1,2-c)pyrazolebased derivatives with general formula X (Figure 2) were investigated. Representative compounds are depicted in Figure 9. Such compounds were described by Neuroscienze Pharmaness S.C.A.R.L in a patent application encompassing a more extensive series of condensed tricyclic pyrazoles [41]. According to binding data, shifting the sulfur atom of the 1,4-dihydrothieno(2 ,3 -4,5)cyclopenta(1,2-c)pyrazole structural template from position 5 to 7 induced a significant impact on cannabinoid receptor affinity. In general, most of the investigated compounds exhibited a preferential affinity for CB 2 receptors, with different compounds showing mixed and high CB 1 /CB 2 cannabinoid receptor affinities (XIhZ, XNhZ, XHhX, XIhY, XIhX, XIgX, Figure 9).
Amongst the compounds claimed in the patent application, derivatives XIhZ, XNhZ, XHhX were investigated to evaluate their intrinsic activity as agonist or antagonist on CB 1 receptors in an ex-vivo model based on the use of the vas deferens. According to the behavior of WIN-55,212-2 and other CB 1 ligands in mice vas deferens isolated organ assay [42], compounds XIhZ, XNhZ, XHhX showed agonist activity, with XIhZ the most effective in inhibiting the contractions induced by an electric stimulus compared to the basal value, as evidenced by the reported dose-response curve [41].
Compounds XIhZ, XNhZ, XHhX were also investigated for their ability to reduce intraocular pressure (IOP) which is considered a prominent risk factor for glaucoma development and progression [43]. The first study highlighting the relevance of CB 1 /CB 2 agonists for the treatment of glaucoma was conducted in 1971, reporting that smoking marijuana significantly lowered the IOP [44]. Very positive results were reported also for WIN-55,212-2, which is the prototype of the aminoalkylindole class of synthetic cannabinoids, activating both CB 1 and CB 2 receptors, albeit with a proclivity for the CB 2 receptors. In particular, in normotensive rabbits, a single dose of WIN-55,212-2, either topical or systemic, significantly reduced IOP without apparent ocular toxicity, most likely through effects on CB 1 receptors [45]. Subsequent studies showed a reduction of IOP in glaucomatous rats after local and chronic administration of WIN-55,212-2, without adverse effects [46]. These findings are consistent with another experiment showing that WIN-55,212-2 decreased IOP in human glaucoma resistant to conventional therapies [47]. Further studies have demonstrated increased aqueous outflow after exposure to the CB 2 agonist JWH015, suggesting a beneficial function derived from CB 2 receptor activation in the treatment of ocular diseases such as glaucoma [48]. Neuroprotective properties of cannabinoids have been demonstrated in CNS neurodegenerative diseases with different mechanisms [49].
Several studies have shown in the retina that CB 1 agonists (∆ 9 -tetrahydrocannabinol and cannabidiol) protected ganglion cells from glutamate-excitotoxicity and ischemia caused by increased IOP [50,51]. Although all these data are promising, an important issue for the clinical potential of cannabinoids as anti-glaucoma agents has been cardiovascular and psychotropic side effects mediated by systemic and brain cannabinoid receptor activation [52][53][54]. Additionally, short duration of action of cannabinoids on IOP reduction (i.e., the duration of action after smoking marijuana is only 3-4 h) is another issue that has to be overcome in the application of these compounds in treating glaucoma. Actually, standard therapeutic options in treating glaucoma include a few IOP-lowering drugs as prostaglandin analogs, β-adrenoreceptor antagonists, α 2 -adrenoreceptor agonists, carbonic anhydrase inhibitors, and cholinergic agonists [55]. When medical therapy failed to lower IOP, laser or surgical interventions are extremely considered in order to prevent disease progression toward blindness. According to unmet medical need, compounds XIhZ, XNhZ, XHhX, exhibiting high and mixed CB 1 /CB 2 receptor affinities, with a proclivity for CB 2 subtype, and endowed of CB 1 agonist properties, according to the CB receptor profile of the reference cannabinoidergic compound WIN-55,212-2, by using the animal model of old DBA/2J mice, were investigated for their ability to reduce IOP [41]. Compounds XIhZ, XNhZ, XHhX and WIN-55,212-2, which was used as reference compound, were dispersed in the commercial emulsion Tocrisolve™ and applied to the eye of old DBA/2J mice at a concentration of 100 µg or 50 µg. The obtained results, reported in Table 1, showed that commercial emulsion Tocrisolve™ (20, 40 µL) had no effect on the IOP. Furthermore, at the dose of 100 µg, all tested compounds were effective in reducing eye pressure as the reference compound WIN-55,212-2, while only XIhZ, at the dose of 50 µg was more effective in reducing IOP than the reference compound. What is claimed in the patent application was pharmaceutical compositions comprising emulsions or microemulsions may be useful to avoid systemic side effects of such cannabinoid compounds. The preliminary results reported in the patent application [41] for compound XIhZ could pave the way for the development of novel cannabinoidergic compounds as anti-glaucoma agents.

1,4-Dihydropyrazolo(3,4-a)pyrrolizine-Based Derivatives (XV)
Continuing with our interest in expanding SAR studies on cannabinoid receptors, and taking into consideration the binding profile of compounds IAaX and IAhX, a new tricyclic pyrazole scaffold, namely 1,4-dihydropyrazolo(3,4-a)pyrrolizine XV, was designed on the basis of bioisosteric replacement approach (benzene/pyrrole) [56]. Representative compounds are depicted in Figure 10. Surprisingly, none of the new compounds exhibited a high affinity for CB 2 receptors, with K i values above 314 nM. Negligible affinity was also determined for CB 1 receptors, with XVIhX reaching the K i value of 142 nM.

4,5-Dihydro-1H-benzo(g)indazole-Based Derivatives (II)
The first (5,6)-condensed pyrazole derivatives synthesized by us are those of general formula II (Figure 2), featuring the 4,5-dihydro-1H-benzo(g)indazole scaffold which is a homologue of 1,4-dihydroindeno(1,2-c)pyrazole one (compounds I). Representative compounds are depicted in Figure 11 [57]. Most of such compounds were described by Sanofi-Aventis in a patent application as cannabinoid CB 1 receptor antagonists, but no biological data were reported [58]. Interestingly, as evidenced by K i values of compounds IIAbX and IIAiX with respect to compounds IAaX and IAhX, increasing the carbocyclic central ring size by one methylene unit (from (5,5)-to (5,6)-condensed pyrazole derivatives) involved a marked loss of affinity for CB 2 receptors, a significant increase in CB 1 affinity, and a consequent loss of CB 2 selectivity. A similar trend was exhibited by other derivatives incorporating such 4,5dihydro-1H-benzo(g)indazole scaffold (i.e., IIAeX and IIBbX). The SAR study carried out on the two homologous series I and II, prompted us to suppose that to achieve high binding affinity to CB 1 receptors and CB 1 over CB 2 selectivity it is important for the tricyclic system to be non-planar. To evaluate the functional profile, several compounds were assayed for the capability to affect gastrointestinal transit in mice, making use of the upper gastrointestinal test which is based on the determination of the intestinal length traveled by a non-absorbable marker as a consequence of the active compound administration. From this test, compound IIAbX was able to induce a dose-dependent gastrointestinal motility increase, as well as compound IIAiX. This effect was markedly reversed by the in vivo administration of CP-55,940, suggesting for the series of 4,5-dihydro-1H-benzo(g)indazole-3-carboxamides antagonistic profile for CB receptors [57].

4,5-Dihydro-1H-Pyrazolo(4,3-g)indolizine-Based Derivatives (XVI)
To further extend SAR study on cannabinoid receptors, and driven by the negligible results on CB 1 /CB 2 receptor affinity of 1,4-dihydropyrazolo(3,4-a)pyrrolizines (XV), homologue 4,5-dihydro-1H-pyrazolo(4,3-g)indolizines (XVI, Figure 2) have been synthesized [56]. Representative compounds are depicted in Figure 13. According to binding data reported for compounds XVIAbX, XVICbX, XVIDbX with respect to compounds XVAaX, XVAhX, XVIaX, XVIhX, the expansion of the central ring size by a methylene unit led to conformational changes that promote the affinity for CB 1 receptors and improve CB 2 /CB 1 selectivity ratio, with XVICbX reaching reasonable affinity for CB 1 receptors (K i CB 1 = 81 nM) and the highest CB 2 /CB 1 selectivity ratio (>12). The intriguing SAR study carried out on ligands with the 1,4-dihydroindeno(1,2c)pyrazole core (i.e., IAaX, IAhX) endowed with very high binding affinity for CB 2 receptors in comparison to homologous ligands having 4,5-dihydro-1H-benzo(g)indazoles (i.e., IIAbX, IIAiX) exhibiting higher CB 1 binding affinity, prompted us to design and synthetize a new homologous series of general formula III, incorporating 1,4,5,6-tetrahydrobenzo(6,7)cyclohepta(1,2-c)pyrazole scaffold, for cannabinoid receptor interaction [61]. Representative compounds are depicted in Figure 14. Compound IIIAcX, namely NESS0327, was first claimed by Sanofi-Aventis, but no biological data was presented in the patent [58]. In general, all the investigated compounds exhibited preferential affinity for CB 1 receptors, with NESS0327, endowed of K i CB 1 = 4.2 nM, K i CB 2 = 55.7 nM and K i CB 2 /K i CB 1 selectivity ratio = 13.26. Slightly different values for CB 1 receptors, possibly due to different receptor matrix, were reported for IIIAcX by two other independent groups: Stoit et al. K i CB 1 = 126 nM [62] and Zhang et al. 18.4 nM [63] using ( 3 H)-CP 55,940 and hCB 1 receptor cloned in CHO cells or membranes isolated from a HEK-293 expression system, respectively. Compound IIIIcX, bearing the bulky myrtanil substituent, showed the best CB 1 receptor affinity which was equivalent to that exhibited by IIIAcX. Moving the chlorine atom from position 8 to 9 of the myrtanil-based derivative IIIIcX to give IIIIdX induced a decrease of CB 1 receptor affinity with a concurrent loss of selectivity. CB 1 receptor intrinsic activity of selected derivatives was evaluated through in vitro tests based on the determination of phosphorylated ERK 1/2 (P-ERK 1/2) expression in mouse neuroblastoma N1E-115 cell line. According to rimonabant, compounds IIIAcX and IIIBcX didn't affect P-ERK expression in N1E-115 cells in the concentration range 1 nM-10µM; P-ERK 1/2 (% of vehicle): vehicle, 100 ÷ 10; rimonabant (1 µM): 110 ÷ 8; IIIAcX (1 µM): 111 ÷ 12; IIIBcX (1 µM): 95 ÷ 9. Furthermore, these compounds inhibited the P-ERK 1/2 expression up-regulation induced by the reference cannabinoid agonist WIN-55,212-2; vehicle: 185 ÷ 12; rimonabant (1 µM): 108 ÷ 7; IIIAcX (1 µM): 115 ÷ 13; IIIBcX (1 µM): 103 ÷ 9. In contrast, the myrtanil-based derivatives IIIIcX and IIIIdX enhanced P-ERK 1/2 expression in N1E-115 cells, highlighting CB 1 receptor agonism profile for both compounds [61]. The pharmacological relevance of 1,4,5,6-tetrahydrobenzo(6,7)cyclohepta(1,2-c)pyrazole-based derivatives III emerged from the ability of antagonist NESS0327 to induce reduction of weight gain and food intake equivalent to that of rimonabant [64]. Furthermore, it was reported that the neutral antagonist profile exhibited by NESS0327 appeared important to avoid the side effects induced by rimonabant administration (i.e., suppression of the constitutive CB 1 receptor activity in the ventral tegmental area and basolateral amygdala causing anxiety and reduced motivation for reward). Several patents testify to the pharmaceutical interest of NESS0327 [65][66][67][68]. Within this frame, the University of Queensland claimed method and agents for reducing general anesthetic induced neuroexcitation, attributing such side effect especially to neurosteroid general anesthetic agents as alfaxalone, even if the patent covers all general anesthetics in use [69]. In vitro electrophysiological analysis showed alfaxalone significantly decreased the amplitude of inhibitory synaptic currents (IPSC) on hypoglossal motor neurons, at a concentration ranging between 100 nM to 10 µM, consistent with increased in vivo motor activity and muscle twitching induced by alfaxalone. Treatment of hypoglossal motor neurons with NESS0327 was demonstrated to attenuate the effect of alfaxalone both on evoked than spontaneous IPSC amplitude. In vivo analysis conducted on Wistar rats highlighted administration of NESS0327 as a premedication prior to alfaxalone anesthesia significantly reduced muscle twitching during the induction and recovery phases of anesthesia. More importantly, premedication with NESS0327 had no detrimental effect on arterial O 2 saturation or respiration rate during alfaxalone anesthesia [69].

4,5-Dihydro-1H-benzo(2,3)oxepino(4,5-c)pyrazole-Based Derivatives (VI)
Endocannabinoids are orexigenic factors promoting appetite via CB 1 receptor activation. This finding provided the rationale for the development of CB 1 antagonist/inverse agonist rimonabant for obesity treatment and its metabolic complications. However, rimonabant side effects responsible for its withdrawal were principally related to both activities on SNC and to inverse agonism profile [70,71]. Therefore, new strategies have been proposed for the development of more safe anti-obesity agents, such as the identification of peripherally restricted CB 1 receptor antagonists [71] as well as neutral CB 1 receptor antagonists or CB 1 allosteric modulators [72]. Within this frame, it was envisioned that bioisosteric modification of 1,4,5,6-tetrahydrobenzo(6,7)cyclohepta(1,2-c)pyrazole core (III) might offer new templates for CB 1 receptor interaction. Thus, a series of 4,5-dihydro-1H-benzo(2,3)oxepino(4,5-c)pyrazoles (VI, Figure 2), containing an oxygen atom at position 6 in place of a methylene unit, were designed and synthesized. Such bioisosteric methylene/oxygen replacement might give access to CB 1 ligands with increased polar surface area and decreased lipophilicity, which were considered critical parameters to influence the blood-brain permeability. Representative compounds are VIAcX, VIDcX, VIIcX (Figure 15), all exhibiting nanomolar/near nanomolar affinity for CB 1 receptors [73]. Within this frame, a series of compounds sharing a 4,5-dihydro-1H-benzo(2,3)oxepino(4,5c)pyrazole core, including VIAcX, were claimed by Cadila Healthcare Limited, but no binding data for cannabinoid receptors were presented [74]. However, the compounds of the invention were generally said, in the cAMP accumulation model, to antagonizes the WIN-55,212-2 inhibition of forskolin-induced cAMP accumulation in hCB 1 CHO cells. Furthermore, VIAcX, and other representative compounds have been shown to reduce, in the sucrose solution intake rat model, the sucrose solution consumption [74,75]. In our hands, compound VIAcX, namely NESS06SM, emerged for its nanomolar CB 1 receptor affinity and good selectivity with respect to CB 2 one (K i CB 1 = 10.25 nM, K i CB 2 > 5000 nM). Evaluation of intrinsic activity carried out on NESS06SM highlighted a neutral antagonist profile both in ( 35 S)GTPγS assay and in isolated organ assays (mouse vas deferens) [76]. In silico parameters (cLogP OW , tPSA, log BB), compared with IIIAcX (NESS0327) and other already known CB 1 antagonists (i.e., rimonabant), suggested that NESS06SM exhibited sparing BBB permeability. Moreover, chronic treatment with NESS06SM resulted in the reduction of body weight and cardiovascular risk factor improvement in C57BL/6N diet-induced obesity (DIO) mice fed with a fat diet. Furthermore, in contrast to rimonabant, the chronic treatment of NESS06SM did not change mRNA expression of both monoaminergic transporter and neurotrophins, highly related to anxiety and mood disorders [76]. Interestingly, the co-treatment of NESS06SM has been shown to reduce food intake and weight gain, mitigate the side effects induced by chronic administration of the atypical antipsychotic olanzapine, without altering the positive effects of olanzapine on behavior [77]. As expected, evaluation of the intrinsic activity of VIDcX, namely SM-11, evidenced CB 1 antagonist activity, both in in vitro test (P-ERK 1/2 expression in N1E-115 cells) and in isolated organs (mouse vas deferens). Behavioral studies highlighted that dosedependently SM-11 decreased food intake in rats by 15-20%. Moreover, the i.v. administration of SM-11 fully and readily antagonized the effect of the agonist WIN-55,212-2 on the activity of ventral tegmental area (VTA) dopamine neurons projecting to the nucleus accumbens cells, confirming its antagonist profile. Furthermore, this data supported that SM-11 can lessen the hedonic aspect of food thus promoting bodyweight reduction [78]. A series of derivatives incorporating a 1,4,5,6-tetrahydrothieno(2 ,3 -6,7)cyclohepta(1,2c)pyrazoleand 1,4,5,6-tetrahydrothieno(3 ,2 -6,7)cyclohepta(1,2-c)pyrazole-scaffolds IX and XII, respectively, (i.e., IXAcX or XIIAcX, Figure 2) were claimed by Neuroscienze Pharmaness S.C.A.R.L. for cannabinoid receptor interaction, but no biological activity was presented [59].  Figure 2) has been investigated for cannabinoid receptor interaction [56]. Representative compounds are depicted in Figure 16, all devoid of affinity for CB 2 receptors and endowed of negligible affinity for CB 1 receptors.  [63]. Within this frame, they synthesized both compound IIIAcX from our lab and IVAdX ( Figure 17) which is the only pyrazole-based tricyclic compound known belonging to (5,8)-condensed derivatives, featuring the 4,5,6,7tetrahydro-1H-benzo (7,8)cycloocta(1,2-c)pyrazole core. The compound, endowed of lower affinity for CB 2 than CB 1 receptors, exhibited a similar trend with respect to homologous compounds IIAbX and IIIAcX. However, increasing the carbocyclic central ring size by one methylene unit, i.e., from (5,7)-to (5,8)-condensed pyrazole derivatives, involved a marked loss of affinity for CB 1 receptors. Figure 17. Structure of 4,5,6,7-tetrahydro-1H-benzo (7,8])cycloocta(1,2-c)pyrazole derivative IVAdX. Affinities at CB 1 and CB 2 receptors were assessed by competition of ( 3 H)-CP 55,940 in HEK-293 and CHO-K1 cell membranes, respectively [63].
Compounds 5-7 exhibited affinity levels for CB 2 receptors in the near nM range (K i : 33-67 nM) with a good degree of selectivity for CB 2 receptors compared to CB 1 . According to in vitro assays based on the effects of forskolin-stimulated cAMP levels in human CB 2 CHO cells, compounds 5-7 exhibited antagonist/inverse agonist properties [79].

Conclusions
In the last decades, the CB 1 antagonist/inverse agonist rimonabant has been considered an extremely valuable lead for the design of new ligands for CB receptors interaction, with potential therapeutic value. In the context of this review, the introduction of structural constraints in rimonabant, the use of medicinal chemistry approaches as homologation or bioisosterism, gave access to several compounds most of which belong to (5,5), (5,6) and (5,7)-condensed tricyclic pyrazole derivatives. Here we have summarized the extensive SAR studies carried out on such compounds, allowing us to identify different ligands endowed with high affinity and selectivity for CB 1 or CB 2 receptors. In particular, the phar-macological relevance of 1,4-dihydroindeno(1,2-c)pyrazole-based I compounds, belonging to (5,5) tricyclic pyrazole derivatives, emerged from the significant anti-nociceptive activity of the potent and selective CB 2 agonist IDhX, by alleviating both mechanical allodynia and thermal hyperalgesia in Spared Nerve Injury (SNI) neuropathic mice. The interest of ligands belonging to (5,5) tricyclic pyrazole derivatives emerged also for their potential as anti-glaucoma agents, for the ability to reduce intraocular pressure (IOP). Compound XIhZ, featuring the 1,4-dihydrothieno(3 ,2 -4,5)cyclopenta(1,2-c)pyrazole core and exhibiting high and mixed CB 1 /CB 2 receptor affinities, effectively reduced intraocular pressure in the animal model of old DBA/2J mice, providing a potential alternative to the use of WIN-55,212-2 and smoking marijuana, known for their IOP lowering properties. SAR study pointed out the relevance of ligands belonging to (5,7) tricyclic pyrazole derivatives, for their anti-obesity potential. In this frame, the selective CB 1 ligand VIAcX, featuring the 4,5-dihydro-1H-benzo(2,3)oxepino(4,5-c)pyrazole core, exhibiting neutral antagonist profile, effectively reduced body weight with the improvement of cardiovascular risk factor in C57BL/6N diet-induced obesity (DIO) mice fed with fat diet. Preliminary data evidenced the potentiality of compound VIAcX in the treatment of obesity with a more safe profile with respect to the antagonist/inverse agonist rimonabant.