Trisubstituted Sulfonamides: A New Chemotype for Development of Potent and Selective CB₂ Receptor Inverse Agonists

Abstract

An extensive exploration of the structure–activity relationship of a trisubstituted sulfonamide series led to the identification of 39, which is a potent and selective CB₂ receptor inverse agonist [K_i(CB₂) = 5.4 nM, and K_i(CB₁) = 500 nM]. The functional properties measured by cAMP assays indicated that the selected compounds were CB₂ inverse agonists with high potency values (for 34, EC₅₀ = 8.2 nM, and for 39, EC₅₀ = 2.5 nM). Furthermore, an osteoclastogenesis bioassay indicated that trisubstituted sulfonamide compounds showed great inhibition of osteoclast formation.

Cannabinoid receptors 1 and 2 (CB₁ and CB₂, respectively) were identified in the early 1990s as members of the G protein-coupled receptor (GPCR) superfamily.^1,55 While CB₁ receptors are primarily found in the central nervous system (CNS), CB₂ receptors are predominantly located in tissues and cells of the immune system, such as tonsils, spleen, macrophages, and lymphocytes.² Also, there is some evidence of the presence of the CB₂ receptor in the CNS.^3,66

Recently, numerous agonists and antagonists of cannabinoid receptors have been explored because of the important role of the endocannabinoid system in various diseases and disorders.⁴ Among these, CB₁ receptor ligands have been developed for pain, appetite stimulation, nausea, neurodegeneration, hypermotility, and inflammation.⁵ However, the CB₁ receptor ligands are known to cause side effects in the CNS such as cognitive dysfunction, motor incoordination, and sedation.⁶ Because of differences in receptor distribution and signal transduction mechanisms, CB₂-selective ligands are considered as medications without CNS side effects,⁷ and such ligands are being actively investigated for use in a multitude of diseases and pathological conditions,⁸ such as atherosclerosis,⁹ myocardial infarction,¹⁰ stroke,¹¹ gastrointestinal inflammatory,¹² autoimmune,¹³ and neurodegenerative¹⁴ disorders, bone disorders,¹⁵⁻⁸⁸ and cancer.¹⁶

The development of CB₂ receptor-selective ligands has attracted significant attention because of the therapeutic potential of CB₂ receptor modulation.¹⁷⁻³³ The first CB₂ inverse agonist is SR144528, which is extensively used as the standard to measure the specificity of various cannabinoid inverse agonists for CB₂ in animal models.¹⁸ Other notable examples of CB₂ receptor agonists and antagonists include AM630,¹⁹ JTE-907,²⁰ Sch225336,²¹ and JWH-133.²² Recently, on the basis of the research in three-dimensional CB₂ receptor structure model^23,56 and pharmacophore database searches, our group also reported the discovery of novel bis-amide derivatives [1 (Figure 1)] as CB₂ receptor inverse agonists and osteoclast inhibitors.²⁴ However, the optimization of bis-amide derivatives is limited by the synthesis method and symmetrical scaffold.

Figure 1.

Structures of CB₂ receptor inverse agonists and new scaffold discovery.

On the basis of continuing virtual screening and QSAR results,²⁵ we designed and synthesized 2, with a trisubstituted sulfonamide scaffold, as a novel chemotype with CB₂ binding activity [K_i(CB₂) = 750 nM]. Compared with the structure of 1 and considering the QSAR results, we believed that a longer chain in zone A was important for the CB₂ inverse agonist (Figure 1). Compound 3 with a diethylamino group was synthesized and confirmed to have a better CB₂ binding affinity [K_i(CB₂) = 53 nM] and a good CB₂ selectivity index [SI = 43, calculated as the K_i(CB₁)/K_i(CB₂) ratio]. Given this promising result, 3 was chosen as a prototype for further structure–activity relationship (SAR) studies. Herein, we reported the design and synthesis of novel trisubstituted sulfonamide derivatives as CB₂ inverse agonists. Binding activities were investigated to define their SAR and ligand functionality. After we modified the groups at zones A–C, some derivatives, such as 34 [K_i(CB₂) = 5.5 nM, and SI = 15], 39 [K_i(CB₂) = 5.4 nM, and SI = 92], and 45 [K_i(CB₂) = 4.0 nM, and SI = 120], were identified as CB₂-selective ligands with improved CB₂ binding affinity and high selectivity. These compounds were selected for the functional property investigation by a cAMP assay, which showed their high potency (for 34, EC₅₀ = 8.2 nM, and for 39, EC₅₀ = 2.5 nM) as CB₂ inverse agonists. Moreover, these compounds also showed great inhibition activity with osteoclast cells.

The trisubstituted sulfonamide derivatives were synthesized by two general strategies (Scheme 1). In the first strategy, the different imine intermediates, synthesized from reductive amination reactions of substituted aryl aldehydes and various amines, were reacted with different sulfonyl chlorides to prepare compounds 3–12, 14–25, 31–36, and 39–48, as well as with benzyl bromide to obtain 13. Compounds 26–30, 37, and 38 were obtained by the reaction of intermediate 25 with various bromides. The structures of all the compounds were characterized by ¹H NMR and ESI-HRMS spectra; purity was confirmed by HPLC.

Scheme 1. Synthesis of Trisubstituted Sulfonamides.

Reagents and conditions: (a) CH₃OH, reflux; (b) CH₃OH, NaBH₄; (c) Et₃N, CH₂Cl₂; (d) K₂CO₃, CH₃COCH₃, reflux.

Biological data of compounds with different substituents of zone A are listed in Table 1. Some alkyl chains, whose substituents had been shown in our previous study to be favorable for CB₂ receptor affinity,²⁴ were introduced at position 4 of phenyl, such as the bialkyl chain (dimethylamino, isopropoxyl, and isopropyl), monoalkyl chain substituents (ethoxyl, propoxyl, and butyl), and cycloalkanes (1-piperidinyl). Among them, 3 (R₁ = diethylamino) and 6 (R₁ = 1-piperidinyl) showed higher affinities for the CB₂ receptor (53 and 44 nM, respectively) with a good CB₂ selective index (SI = 43 and 28, respectively). These results indicated that longer chains and double chains are necessary for improving binding affinity in this scaffold.

Table 1. CB₁ and CB₂ Receptor Affinities of Compounds 3–12 with a Modified Zone A.

		Ki (nM)
compd	R1	CB2a,b	CB1a,b	SIc
3	-N(CH2CH3)2	53 ± 6	2300 ± 200	43
4	-N(CH3)2	680 ± 10	NT
5	-(CH2CH2Cl2)2	166 ± 4	NT
6	1-piperidinyl	44 ± 8	1300 ± 200	28
7	-OCH(CH3)2	174 ± 7	NT
8	-OCH2CH3	230 ± 60	NT
9	-CH2CH2CH3	130 ± 20	NT
10	-CH2CH=CH2	130 ± 40	NT
11	-n-butyl	280 ± 50	NT
12	-CH(CH3)2	230 ± 40	NT
SR144528d,e		2.1 ± 0.4	NT
SR141716d,f		NT	11 ± 1

^aBinding affinities of compounds for CB₁ and CB₂ receptors were evaluated using the [³H]CP-55,940 radioligand competition binding assay. Data are means ± the standard error of the mean of at least three experiments performed in duplicate.

^bNT, not tested.

^cSI, selectivity index for CB₂, calculated as the K_i(CB₁)/K_i(CB₂) ratio.

^dThe binding affinities of reference compounds were evaluated in parallel with compounds 3–12 under the same conditions.

^eCB₂ reference compound.

^fCB₁ reference compound.

The modification of zone B was based on maintenance of the diethylamino group in zone A, as shown in Table 2. Compound 13, in which the sulfo group was replaced with a CH₂ group, showed a slightly lower affinity. A similar result was found when a CH₂ group was added between the sulfo and the phenyl group (14). Many substituents were introduced at position 4 of the sulfophenyl, such as Cl, F, H, acetylamide, isopropyl, methoxyl, isopropoxyl, and trifluoromethoxyl. Other than 20 (R₂ = 4-methoxyphenyl), which had a binding affinity for the CB₂ receptor (K_i = 46 nM) similar to that of 3, 13–22 did not show obvious increased binding affinity. Large groups, like naphthyl and 1,1′-biphenyl, were also used to replace the 4-methylphenyl. This replacement resulted in a lower binding affinity. These results indicated that 4-methylphenyl and 4-methoxylphenyl are the best groups among our modifications of zone B. Moreover, the CB₂ selective indexes of 15, 16, 19, and 20 were quite low, ranging from 6 to 15.

Table 2. CB₁ and CB₂ Receptor Affinities of Compounds 13–24 with a Modified Zone B.

^a–gSame as the footnotes of Table 1.

After investigation of the SAR of R₁ and R₂ in zones A and B, the binding affinities were not obviously improved. The following modification of group R₃ in zone C was based on maintenance of the R₁ and R₂ groups in 3 (Table 3). As expected, group R₃ is a key substituent for enhancing the CB₂ affinity. First, another CH₂ group was introduced at the N atom and the phenyl ring; compound 26 showed a lower CB₂ binding affinity (K_i = 120 nM). The same reduced trends were found when diethylamino and halogen (Cl and F) were added to the phenyl ring in zone C. The binding affinities for the CB₂ receptor of 27–30 were 540, 210, 390, and 120 nM, respectively. After substitution of the phenyl in zone C with cyclohexyl, a small increase in affinity was achieved [for 31, K_i(CB₂) = 34 nM], while replacement by five-membered heterocyclic moieties (in 32 and 33) enhanced affinity dramatically, yielding a K_i values of 9.5 and 14 nM at CB₂ receptors and 570 and 610 nM at CB₁ receptors, respectively. It seemed that a smaller group in zone C is preferable for CB₂ binding activity, so some compounds with smaller group were synthesized. Compound 34 with the reduced CH₂ between the N atom in the core and phenyl group showed a higher CB₂ binding affinity (K_i = 5.5 nM, and SI = 15). However, replacement with other smaller groups, such as cyclohexyl (35), 5-methylthiazolyl (36), cyclopentyl (37), and allyl (38), significantly reduced the binding activity [K_i(CB₂) = 18, 14, 36, and 66 nM, respectively]. Especially, 25, without any substituted group except with an H atom at R₃, showed quite low CB₂ binding activity [K_i(CB₂) = 3600 nM]. This result suggests that the space of the phenyl group is suitable for zone C and thus addition of a larger or smaller group results in lower CB₂ binding activity.

Table 3. CB₁ and CB₂ Receptor Affinities of Compounds 25–38 with a Modified Zone C.

^a–gSame as the footnotes of Table 1.

On the basis of the SAR research of zones A–C, we found that the introduction of some groups at R₁–R₃ could improve CB₂ binding activity. Further SRA research was based on the compound library of combinations of these groups: R₁ = diethylamino and 1-piperidinyl, R₃ = phenyl, cyclohexyl, furan-2-ylmethyl, and 5-methylthiazolyl, and R₄ = methoxyl, Cl, isopropyl, and methyl. The results are listed in Table 4 with the CB₂ binding affinity ranging from 4 to 310 nM and the selective index ranging from 9 to 120. Among them, two compounds were identified with better CB₂ binding activity and selectivity, 39 [K_i(CB₂) = 5.4 nM, and SI = 92] and 45 [K_i(CB₂) = 4.0 nM, and SI = 120].

Table 4. CB₁ and CB₂ Receptor Affinities of Compounds 39–48.

				Ki (nM)
compd	R1	R3	R4	CB2a,b	CB1a,b	SIc
39	-N(CH2CH3)2	phenyl	-OCH3	5.4 ± 0.5	500 ± 60	92
40	-N(CH2CH3)2	phenyl	-CH(CH3)2	30 ± 4	2000 ± 200	66
41	-N(CH2CH3)2	furan-2-ylmethyl	-OCH3	20 ± 2	840 ± 80	42
42	-N(CH2CH3)2	furan-2-ylmethyl	-CH(CH3)2	31 ± 4	1300 ± 100	42
43	-N(CH2CH3)2	cyclohexyl	-OCH3	75 ± 6	660 ± 60	9
44	-N(CH2CH3)2	cyclohexyl	-CH(CH3)2	55 ± 2	1400 ± 100	25
45	-N(CH2CH3)2	5-methylthiazolyl	-OCH3	4.0 ± 0.6	600 ± 10	120
46	1-piperidinyl	phenyl	-CH3	310 ± 20	NT
47	1-piperidinyl	phenyl	-OCH3	32 ± 7	590 ± 50	18
48	1-piperidinyl	phenyl	-Cl	140 ± 10	NT
SR144528d,e				2.1 ± 0.4	NT
SR141716d,f				NT	11 ± 1

^a–fSame as the footnotes of Table 1.

Functional properties were investigated in cAMP assays by using cell-based LANCE cAMP assays as our published protocol²⁶ to measure the agonistic or antagonistic functional activities of the CB₂-selective compounds. Because CB₂ is a G_αi-coupled receptor, an agonist inhibits the forskolin-induced cAMP production, resulting in an increase in the magnitude of the LANCE signal, while an antagonist or inverse agonist decreases the magnitude of the LANCE signal. In addition to 34, 39, 45, and 47, with respective modifications in zones A–C for good binding activity, were selected for cAMP assays. As shown in Figure 2, reduction of the magnitude of the LANCE signal occurred with increasing concentrations of 34, 39, 45, 47, and SR144528 with EC₅₀ values of 8.2 ± 3.1, 2.5 ± 1.4, 73 ± 2, 49 ± 2, and 14 ± 3 nM, respectively. Such a contrary phenomenon was observed with agonists CP55,940 and HU308,²⁷ which showed cAMP production with EC₅₀ values of 11 ± 2 and 85 ± 5 nM, respectively. On the basis of the LANCE signal change and the high EC₅₀ value closely correlated with the high affinity value, it suggests that 34, 39, 45, and 47 behaved as CB₂ receptor inverse agonists.

Figure 2.

Comparisons of the LANCE signal of different CB₂ receptor ligands in stably transfected CHO cells expressing human CB₂ receptors in a concentration-dependent fashion. EC₅₀ values of compounds 34, 39, 45, 47, and SR144528 are 8.2 ± 3.1, 2.5 ± 1.4, 73 ± 2, 49 ± 2, and 14 ± 3 nM, respectively. EC₅₀ values for CP55,940 and HU308 are 11 ± 2 and 85 ± 5 nM, respectively. Data are means ± the standard error of the mean of one representative experiment of two or more performed in duplicate or triplicate.

Modulating osteoclast function is a well-known activity of CB₂ receptor agonists²⁸ and inverse agonists.^15c Three compounds, 34, 39, and 45, were selected on the basis of the results of binding affinity and selectivity, as well as their functionality as candidate inhibitors of osteoclast (OCL) formation. As shown in Figure 3, we tested the effects of these most promising CB₂ ligands on osteoclast (OCL) formation using mouse bone marrow mononuclear cells treated by the mouse receptor activator of NF-κB ligand (RANKL) plus macrophage-colony-stimulating factor (M-CSF) (see the Supporting Information). These three compounds exhibited strong inhibition of osteoclastogenesis. Among them, 34 showed the most favorable activity. At 0.1, 1, and 5 μM, it suppressed osteoclast formation by 55, 83, and 99.7%, respectively. To investigate their cell toxicity, 34 was tested in a cytotoxicity assay without showing any cytotoxic effects at a concentration of 5 μM. These results indicate that our compounds possess favorable therapeutic indexes and the inhibition of osteoclastogenesis is not a result of their cytotoxicity.

Figure 3.

Inhibition of osteoclastogenesis by CB₂ ligands 34, 39, and 45. All experiments were performed in triplicate. Results are means ± the standard deviation. Note that the control is vehicle control.

In summary, we have discovered the trisubstituted sulfonamide chemotype as a novel series possessing significant cannabinoid CB₂ receptors affinity. Some compounds with high binding affinities and selective indexes of CB₂ receptors were identified by optimization of zones A–C. The potencies of the novel compounds were measured in functional assays, with high potency values (represented by EC₅₀) that are closely correlated with the high affinities (expressed as K_i), revealing that the novel series behaves as CB₂ receptor inverse agonists. The promising inhibition activity to osteoclast cells of this novel series of compounds offers an attractive starting point for further optimization.

Supporting Information Available

Synthetic experimental details, analytical data, and biological assay protocols. This material is available free of charge via the Internet at http://pubs.acs.org.

We authors from the University of Pittsburgh gratefully acknowledge the financial support of our laboratory by National Institutes of Health Grants R01DA025612 and HL109654 (X.-Q.X.). We also acknowledge the collaboration support from the National Natural Science Foundation of China (NSFC81090410, NSFC90913018, and NSFC21202201).

The authors declare no competing financial interest.

Funding Statement

National Institutes of Health, United States