Exceptionally Potent Chiral Anandamide Analogs

Abstract

We recently reported an innovative design approach that allowed us to obtain potent endocannabinoids with enhanced metabolic stability. Our design is characterized by the incorporation of chiral centers within the endocannabinoid prototypes N-arachidonoylethanolamide and 2-arachidonoylglycerol. Work on N-arachidonoylethanolamide led to the identification of the first-generation lead analog (R)-N-(1-Methyl-2-hydroxyethyl)-13-(S)-methyl-arachidonamide (AMG315). Here, we synthesized a series of tail-modified AMG315 analogs to further optimize this novel chemotype for cannabinoid receptor binding affinity and potency. Our advanced molecule, namely, 20,20,20-trifluoro-(R)-N-(1-methyl-2-hydroxyethyl)-13-(S)-methyl-arachidonamide (AM12814, 12), is the first endocannabinoid analogue exhibiting unprecedented affinity for both the CB1 and CB2 receptors. In further in vitro functional characterization, 12 behaves as a potent, partial CB1 and CB2 agonist. Our SAR results are supported by docking studies of 12 on the crystal structures of cannabinoid receptors, while when tested in vivo, 12 behaves as a very potent and efficacious CB1 agonist. This analogue will serve as a unique endocannabinoid probe.

Graphical Abstract

INTRODUCTION

The legalization of cannabis and cannabis-related products in various US states has rekindled interest in the research and development of cannabinoid therapeutics in both academia and the private sector.¹ To support this growing interest in preclinical and clinical research, an in-depth understanding of the pharmacology of endocannabinoid system constituents, including receptors, endogenous ligands, and their synthesizing and degrading enzymes, is required. Despite the wide range of current cannabinoid system modulators and the recent structural studies on the two major cannabinoid receptors CB1 and CB2,^2,3 there is still need for the development of new cannabinoid ligands to serve as biological tools and as potential cannabinergic drug candidates with better side-effect profile. Particularly, our current arsenal of tool molecules and drug leads resembling the structure and pharmacological action of endogenous cannabinoid lipids is severely limited.

The main endogenous ligands for the cannabinoid system are N-arachidonoylethanolamide (AEA) and 2-arachidonoylglycerol (2-AG), two eicosanoid lipids consisting of an arachidonic acid chain coupled to ethanolamine and glycerol, respectively.⁴ 2-AG selectively binds cannabinoid receptors, being a high efficacy agonist for both CB1 and CB2, while AEA is a low-efficacy agonist for CB1 with modest affinity for CB2.^4,5 Physiologically, the endocannabinoid lipids are synthesized on-demand and exhibit short half-lives due to their rapid enzymatic degradation by various hydrolyzing enzymes with the major being monoacylglycerol lipase (MGL) and fatty acid amide hydrolase (FAAH).^6,7 Additionally, the endocannabinoids are susceptible to oxidation from oxygenases including cyclooxygenases (COX-2), cytochrome P450, and lipoxygenases (LOXs), which result in the formation of oxygenated eicosanoid-type products with distinct biological functions, mediated through the endocannabinoid or other biological systems.^8,9 The propensity of eCBs for enzymatic degradation, coupled with their generally unfavorable pharmacodynamic profile, underlines the need for developing novel lipid-like analogs with better stability and higher potency for both cannabinoid receptors to serve as tools for deconvoluting this complex biological system and as endogenous-like templates for drug discovery.^1,10 Critically, to our knowledge, there currently exists no endocannabinoid analog with high affinity for CB2 or combined CB1/CB2 affinity¹

Generally, endocannabinoid-related ligands possess 4 distinct pharmacophores (Figure 1), the ethanolamide/glyceride head, the propyl linker, the tetra-olefinic chain, and the n-pentyl tail.¹¹ In earlier work, we reported a novel design approach based on the incorporation of methyl groups in the S or R configuration at judiciously chosen positions within the tetra-olefinic chain.^11–14 Our initial SAR studies led to the identification of conformationally restricted AEA and 2-AG analogs, namely, AMG315 and AM8125, which exhibit enhanced stability toward enzymatic degradation and significantly improved potency and efficacy for both cannabinoid receptors. Two design findings stood out from this work: the incorporation of a (13S)-Me group, which protects the lipids from the major oxidative enzyme COX-2, and the incorporation of a (1′R)-Me group, which protects the AEA analog from all endocannabinoid hydrolyzing enzymes. These compounds currently serve as the gold standard for research involving endocannabinoid analogs and are used in a variety of in vitro, in vivo, and structural studies.^11,14–16 AMG315 in particular played a key role in determining the cryo-EM structure of the CB1 receptor bound to an endogenous-like ligand.¹⁶ Despite our initial success, we sought to further optimize our lead analogue AMG315, especially regarding its binding affinity and potency for the CB2 receptor.

Figure 1.

Summary of endocannabinoid lipid pharmacophores and current study design.

To achieve this, we shifted our attention to the relatively unexplored n-pentyl tail pharmacophore of the chiral endocannabinoid template. Particularly, in Liu et al.,¹¹ we observed that addition of the N₃ and NCS moieties on the terminal 20-position of the (13S)-Me-arachidonoyl amide scaffold leads to analogs with enhanced affinity for both the CB1 and CB2 receptors. This inspired further exploration of the SAR of the 20-position of our more successful first-generation chiral AEA analogue to produce novel ligands with higher affinity and potency for both cannabinoid receptors. Additionally, cryo-EM and crystal structure^2,3,16,17 data of cannabinoid receptors bound to exogenous and endogenous cannabinoids suggest that the n-pentyl tail of the eCBs and the side chain of tricyclic cannabinoids orient themselves into similar parts of the receptor binding pocket. Leveraging our laboratory’s extensive SAR knowledge of exogenous tricyclic cannabinoids, novel eCB ligands containing the most successful moieties were designed, synthesized, and pharmacologically evaluated. Namely, these moieties include terminal cyano, nitrate ester, hydroxy, bromo, fluoro, and trifluoromethyl groups, all of which are commonly used in our most potent tricyclic cannabinoids.^18–20 This SAR study led to the identification of the first AEA analogs with exceptionally high potency for both CB1 and CB2 receptors (Figure 1). Critically, these new compounds retain the chiral methyl functionalities at C1′ and C13 of the prototype AMG315 that provide robust stability against all endocannabinoid hydrolases, as well as the major endocannabinoid oxygenase COX-2.¹¹

CHEMISTRY

As reported in our earlier work with closely related systems,^11,13 we have used the Wittig approach for the construction of the C1–C20 chiral skeleton of analogs. Synthesis of the appropriate phosphonium salts 3 and 5 that serve as the tail synthetic fragments is depicted in Scheme 1. Briefly, commercially available 6-bromohexanol was first protected and then treated with triphenylphosphine in acetonitrile to afford intermediate phosphonium salt 3 in 99% yield. The synthesis of trifluoromethyl-containing phosphonium salt 5 was conducted in a similar manner.

Scheme 1. Synthesis of Intermediate Phosphonium Salts 3 and 5^a.

^aReagents and conditions. (a) 2,4-dihydropyran, PPTS, DCM, rt, 94%; (b) PPh₃, MeCN, 75 °C, 5 days, 3: 99%, 5: 98%.

To synthesize the 20-substituted anandamide analogs, we followed an approach similar to the one previously reported by our laboratory.¹³ The requisite aldehyde 8 (Scheme 2) was synthesized in 15 steps from commercially available enantiomerically pure (2S)-3-hydroxy-2-methylpropionic methyl ester in a manner similar to the one we reported earlier.¹³ Treatment of phosphonium salts 5 or 3 with KHMDS and coupling of the resulting ylides with aldehyde 8 at −110 °C produced the respective methyl esters 9a and 9b with the Z stereochemistry at the newly formed double bond as we published earlier.¹³ Subsequent hydrolysis of the resulting methyl esters and coupling of the respective carboxylic acids 10a,b with enantiomerically pure (R)-1-((tert-butyldiphenylsilyl)oxy)propan-2-amine (17)¹³ resulted in the formation of protected ethanolamides 11a and 11b in 86 and 80% yield, respectively. Silyl ethanolamide 11a was then deprotected with TBAF to afford the final anandamide analog 12. The tail and head protected amide 11b was instead treated with PPTS to cleave the tetrahydropyranyl protecting group and the resulting 20-hydroxy intermediate 13 was converted to the corresponding bromide using Appel reaction conditions in 83% from 11b. Deprotection of the silyl-protecting group followed by nucleophilic substitution of the 20-bromide with either NaCN or AgONO₂ resulted in the formation of final products 16a (66% yield) and 16b (70% yield), respectively.

Scheme 2. Synthesis of 20-Substituted Anandamide Analogs^a.

^aReagents and conditions. (a) Phosphonium salt 3 or 5, KHMDS, THF −78 to −60 °C, then 8, −110 to 0 °C, 9a: 75%, 9b: 85%; (b) LiOH, THF/H₂O, rt, 10a: 77%, 10b: 83%; (c) CDI, THF, rt, then addition of amine 17, rt, 11a: 86%, 11b: 80%; (d) TBAF, THF, 0 °C to rt, 85%.; (e) PPTS, EtOH, 55 °C, 91%; (f) CBr₄, Ph₃P, CH₂Cl₂, −25 to 0 °C, 92%; (g) TBAF, THF, 0 °C to rt, 88% (h) NaCN, DMSO rt, 66%; (i) AgONO₂, CH₃CN, 80 °C, 70%.

Lastly, the synthesis of analogs 18 and 21 proceeded from intermediate 13 (Scheme 3). Silyl deprotection of 13 with TBAF provided analog 18 directly, in 85% yield. The 20-fluoro analogue 21 was obtained via the mesylate intermediate 19, which was first deprotected with TBAF at 0 °C and then reacted with excess potassium fluoride in the presence of Kryptofix 222 to afford final fluoride 21 in 55% yield.

Scheme 3. Synthesis of 20-Hydroxy and 20-Fluoro Analogs^a.

^aReagents and conditions. (a) TBAF, THF, rt, 85%; (b) MsCl, Et₃N, CH₂Cl₂, 0 °C; (c) TBAF, THF, rt, 65% (over 2 steps); (d) KF, K₂CO₃, Kryptofix 222, MeCN 80 °C, 15 min, 55%.

BINDING AFFINITIES FOR CANNABINOID RECEPTORS

The initial pharmacological evaluation of the novel anandamide analogs was accomplished using standard competitive ligand-binding assay with the widely used [³H]CP-55,940 as the radioligand.¹¹ Employing this assay, we were able to determine their binding affinity in the form of Ki to rCB1 and hCB2 receptors. The results obtained from this pharmacological evaluation are summarized in Table 1, where data for AEA, our earlier lead compound AMG315, and CP-55,940 are also shown for comparison.

Table 1.

Affinities (K_i) of AEA analogs for CB1 and CB2 cannabinoid receptors (± 95% confidence limits).

Compound	Structure	rCB1 receptor	hCB2 receptor
		Ki, (nM)a (95%CI)	Ki (nM)a (95%CI)
AEA		74.5*11	16011
AMG315 22		7.8 (± 1.4)11	176 (± 28)11
AM11655 16a		1.4 (± 0.4)	2.2 (± 1.0)
AM11654 16b		1.7 (± 0.9)	10.9 (± 4.4)
AM11653 15		1.1 (± 1.5)	5.3 (± 3.1)
AM12814 12		0.70 (± 0.04)	3.4 (± 0.12)
AM12810 18		29.0 (± 4.2)	55.2 (± 9.4)
AM12817 21		4.2 (± 1.2)	22.2 (± 3.7)
CP-55,940		1.621	3.621

^aAffinities for CB1 and CB2 receptors were determined using membranes from rat brain (CB1) or HEK293 cells expressing human CB2 receptors and the widely used [³H]CP-55,940 as the radioligand following previously described procedures.¹¹ Data were analyzed using nonlinear regression analysis. K_i values were obtained from three independent experiments run in duplicate and are expressed as the mean of the three values.

^*This value was obtained in the presence of serine hydrolase inhibitor PMSF.

Assessment of the obtained results confirms our initial hypothesis, as substitution on the 20-position generally resulted in analogs with improved binding affinity for CB1 and substantially improved binding affinity for CB2. More specifically, analogs 15, 16a, 16b, and 21 show a 2–7-fold increase in binding affinity for CB1 when compared to our lead AMG315, while the 20-CF₃ analog 12, with a K_i in the subnanomolar range, exhibits 10-fold higher affinity for CB1 than that of the lead compound AMG315. Lastly, the introduction of a 20-hydroxy functionality on the eCB scaffold results in decreased binding affinity for the CB1 receptor. As predicted, substituents on the 20-position have an even more significant effect on the binding affinity of the produced analogs for the CB2 receptor. Our data indicates that all AEA analogs exhibit markedly higher binding affinities when compared to AMG315. Analogs 12 and 16a distinguished themselves as they exhibit 52–80-fold higher affinity for CB2 than the lead compound (K_is = 3.4 nM and 2.2 nM, respectively).

These results indicate that the cyano, fluoro, nitrate ester, and trifluoromethyl groups are very well tolerated in the 20-position of the chiral anandamide prototype, in both receptors. Specifically, introduction of the CF₃ group resulted in analog 12, which exhibits the highest binding affinity for CB1 ever reported in the literature as far as endocannabinoid analogs are concerned. Remarkably, this binding affinity eclipses even that of CP-55,940 which is commonly used as a model cannabinoid ligand and tritiated standard. This compound also exhibits excellent binding affinity for CB2 which is unusual for an anandamide analog. The exceptional binding affinity of compound 12 singles it out as the best analogue of this SAR.

FUNCTIONAL CHARACTERIZATION

We conducted further cell-based functional assays to examine the downstream signaling profile of the key analogue 12 in detail. The endogenous AEA and our earlier lead compound AMG315 are also assessed for comparison.¹¹

Specifically, we studied its downstream signaling via a forskolin-stimulated cAMP accumulation assay. We used CHO cells overexpressing human CB1 (hCB1) or human CB2 (hCB2), as these cells are used routinely in our laboratories and provide robust results.^18,19,22 Under physiological conditions, cannabinoid receptors couple predominantly to Gi proteins, which inhibit adenylyl cyclase. Thus, the measured % inhibition of forskolin-stimulated cAMP production is induced as part of the downstream signaling cascade initiated by the agonist-bound receptor and serves as a quantitative measure of function. Results from this assay along with the calculated functional potency (EC₅₀) and efficacy (E_max) are shown in Table 2 and Figure 2. Cannabinoid Receptors 1 and 2 have also been shown to signal via β-arrestin. For that reason, our most successful compound (12) was also assessed in the cellular β-arrestin 2 recruitment assays to further assess any potential biases between signaling pathways.¹⁹ The results of this assay are also shown in Table 2 and Figure 2.

Table 2.

EC₅₀ and E_max values (and the respective 95% confidence intervals) of AEA, AMG315, CP55,940, and 12 in forskolin stimulated cAMP accumulation and β-arrestin 2 recruitment assays at human CB1 and CB2 receptors.

Compd	Structure	hCB1 receptor		hCB2 receptor
		cAMP	β-Arrestin 2	cAMP	β-Arrestin 2
		EC50 (95% CI) Emax (95% CI)	EC50 (95% CI Emax (95% CI)	EC50 (95% CI) Emax (95% CI)	EC50 (95% CI Emax (95% CI)
AEA		5093 nM (2002–19111) 72.4 % @10 μM (48.4 – 96.4)	>10,000 nM 30.9 % @10 μM (21.5 – 40.4)	327 nM (206–507) 27 % (25 – 30)	91 nM (51–160) 36 % (32–40)
AMG315		64.0 nM (42.0 – 97.6) 80.5 % (74.9 – 86.2)	848 nM (502 – 1434) 38 % (34 – 44)	45 nM (23–84) 18 % (16 – 20)	314 nM (188–514) 36 % (32–41)
AM12814 12		8.75 nM (6.10 – 12.5) 67.5 % (63.5 – 71.5)	25 nM (12–52) 22 % (19 – 24)	22 nM (13–37) 29 % (26–32)	5.6 nM (2.0–14.2) 28 % (24–32)
CP55,940		3.23 nM (2.77 – 3.77) 100%	17.4 nM (15.4 – 19.7) 100%	3.2 nM (2.8 – 3.7) 100%	2.5 nM (2 – 3) 100%

EC₅₀ values were calculated following normalization holding baseline as zero and maximum CP-55,940 concentration as 100 and nonlinear regression analysis (see details in supporting information).

Figure 2.

Concentration-dependent curves of forskolin-stimulated cAMP accumulation and β-arrestin 2 recruitment assays for AEA, AMG315, CP-55,940, and 12 at hCB1 and hCB2.

Assessment of the cAMP assay data shows that compound 12 behaves as a potent agonist for both CB1 and CB2 receptors with EC₅₀ values comparable to that of the exogenous ligand CP-55,940 which is remarkable. Critically, in the cAMP assay, compound 12, despite its excellent potency, acts as a lower-efficacy agonist for CB1 and CB2 receptors (E_max = 67 and 29% respectively), which is consistent with the functional profile of the endogenous ligand, constituting compound 12 as an excellent tool for studying anandamide pharmacology. Compound 12 has a similar profile in the β-arrestin 2 recruitment assay, again acting as a potent, lower-efficacy agonist in both cannabinoid receptors. When compared to the standard CB1 and CB2 agonists CP-55,940, compound 12 behaves as a balanced agonist with comparable potency for the cyclase and arrestin signaling pathways. However, in CB1, the efficacy for the cyclase pathway is 3-fold higher than that of the β-arrestin 2 recruitment.

In summary, a careful assessment of the in vitro data for compound 12 shows that it acts as a substantially more potent version of the endogenous anandamide on human CB1 with a predicted resistance to metabolizing enzymes. Additionally, this compound is the first anandamide analog that acts as a potent lower-efficacy agonist on human CB2 receptor. The high in vitro potency along with the unique pharmacological profile of 12 makes it an excellent candidate for in vivo evaluation in various behavioral assays to evaluate the therapeutic potential of this unique endocannabinoid analog.

IN VIVO BEHAVIORAL CHARACTERIZATION

Results.

AM12814 Produces the Characteristic CB1-Mediated Behavioral Triad.

AM12814 elicited canonical CB1-mediated behaviors. In the ring test, AM12814 increased the immobility time in a dose- and time-dependent manner (Figure 3A; Drug Treatment: F(3, 140) = 75.07, p < 0.0001; Time: F(6, 140) = 13.47, p < 0.0001; Interaction: F(18, 140) = 2.683, p = 0.0006). The high (3 mg/kg, i.p.) and middle (1 mg/kg, i.p.) doses of AM12814 increased immobility time relative to either vehicle or the low dose (0.3 mg/kg, i.p.) overall (p < 0.0001 for each comparison). The high dose produced greater immobility than the middle dose, whereas the low dose did not differ reliably from vehicle overall (p = 0.0684). AM12814 (1 and 3 mg/kg, i.p.) increased immobility time across the entire 4-h observation interval (p ≤ 0.0107 for each time point), whereas the low dose transiently increased immobility time at 30 min postinjection only (p = 0.02). The CB1 antagonist AM251 (5 mg/kg, ip) blocked the catalytic effect of AM12814 (1 mg/kg ip) in the ring test (Figure 3B; Drug Treatment: F(3, 126) = 26.15, p < 0.0001; Time: F(6, 126) = 5.444, p < 0.0001; Interaction: F(18, 126) = 2.232, p < 0.0001). Immobility time was higher overall in groups receiving AM12814 (Veh-AM12814); “Veh-AM12814” refers to vehicle-treated animals that subsequently received AM12814 compared to AM251-pretreated mice that subsequently received either AM12814 (AM251-AM12814) or vehicle (AM251-Veh) (p < 0.0001 for each comparison). Immobility time was higher in the Veh-AM12814 (1 mg/kg, i.p.) group compared to all other groups at 0.5 and 1 h postinjection (p < 0.0001).

Figure 3.

AM12814 produces dose-dependent catalepsy, hypothermia, and antinociception through a CB1-dependent mechanism. (A) AM12814 (1 and 3 mg/kg, ip) induced catalepsy, as defined by an increase in immobility time in the ring test. (B) The cataleptic effects of AM12814 (3 mg/kg i.p.) were blocked by the CB1 antagonist AM251 (5 mg/kg, i.p.). (C) Compound AM12814 produced a dose-dependent reduction in body temperature. (D) Pretreatment with AM251 (5 mg/kg, i.p.) blocked the hypothermic effect of AM12814 (1 mg/kg, i.p.). (E) Compound AM12814 (0.3, 1, and 3 mg/kg, i.p.) produced a dose-dependent increase in tail-flick latency in the tail-immersion test. (F) AM251 (5 mg/kg, i.p.) blocked tail-flick antinociception induced by AM12814 (1 mg/kg, i.p.) in a dose- and time-dependent manner. Data are expressed as mean ± SEM (n = 5–6 per group). Brackets indicate the main treatment effects. * (A, C, E): ^0.3 mg/kg, *1 mg/kg, ⁺3 mg/kg AM12814 vs Vehicle. (B, D, F): ^Vehicle–Vehicle, *AM251-Vehicle, ^#AM251- AM12814 vs Veh-AM12814. Number of symbols denote significance level: ***p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05 determined by two-way repeated measures ANOVA followed by Tukey’s and Dunnett’s post hoc test.

AM12814 also produced dose-and time-dependent reductions in body temperature (Figure 3C; Drug Treatment: F(3, 140) = 47.89, p < 0.0001; Time: F(6, 140) = 7.586, p < 0.0001; Interaction: F(18, 140) = 2.141, p = 0.0072). The high (3 mg/kg, i.p.) and middle (1 mg/kg, i.p.) doses of AM12814 were equally effective (p = 0.0941) in reducing body temperature relative to vehicle (p < 0.0001) or the low (0.3 mg/kg, i.p.) dose (p < 0.0001) whereas effects of the low dose did not differ from vehicle (p = 0.8732). AM12814 (1 or 3 mg/kg) reduced body temperature as early as 10 min postinjection (p ≤ 0.0147), and this hypothermic effect was sustained from 0.5 to 4 h postinjection (p ≤ 0.0126). By contrast, the low (0.3 mg/kg, i.p.) dose of AM12814 did not alter body temperature at any postinjection time point. The hypothermic effect of AM12814 was also abolished by AM251 pretreatment (Figure 3D; Drug Treatment: F(3, 126) = 10.32, p < 0.0001; Time: F(6, 126) = 6.410, p < 0.0001; Interaction: F(18, 126) = 0.9651, p = 0.5037). Body temperature was lower overall in the Veh-AM12814 group compared to AM251–AM12814 (p < 0.0001), AM251-Veh (p = 0.0025), or the Veh–Veh (p = 0.0026) group. The CB1 antagonist reliably blocked the hypothermic effect of AM12814 from 0.5 to 2 h postinjection (p ≤ 0.0225), and body temperature in the Veh-12-treated group was also lower than either the Veh–Veh or AM251-Veh group at 60 min (p = 0.0055) postinjection.

Compound AM12814 increased tail-flick latency in a dose-dependent manner in the tail-immersion test, and tail-flick latencies also changed across the observation interval irrespective of drug treatment (Figure 3E; Drug Treatment: F(3, 140) = 17.05, p < 0.0001; Time: F(6, 140) = 1.417, p = 0.0123; Interaction: F(18, 140) = 1.073, p = 0.3854). The low (0.1 mg/kg i.p., p = 0.0173), middle (1 mg/kg i.p., p < 0.0001), and high (3 mg/kg i.p., p < 0.0001) doses of AM12814 all increased tail-flick latency relative to vehicle treatment overall. The high dose also produced a greater antinociceptive effect across the observation interval than the low dose (p = 0.0010). AM251 pretreatment reduced the overall antinociceptive efficacy of AM12814 (3 mg/kg i.p.) (p = 0.0224) (Figure 3F; Drug Treatment: F(3, 126) = 9.894, p < 0.0001; Time: F(6, 126) = 0.6569, p = 0.6845; Interaction: F(18, 126) = 0.7368, p = 0.7677), but the interaction over time did not reach statistical significance. Tail-flick latencies were higher overall in the Veh-AM12814 group compared to either the AM251–AM12814 (p = 0.0224), AM251-Veh (p < 0.0001), or Veh–Veh (p = 0.0002) treatment. Thus, AM12814 produced hallmark CB1-mediated behavioral effects—catalepsy, hypothermia, and tail-flick antinociception—that were dose-dependent, robust, and long-lasting.

AM12814 Reverses CFA-Induced Mechanical Hypersensitivity.

Complete Freund’s Adjuvant (CFA) administration reduced mechanical paw withdrawal thresholds in the ipsilateral (inflamed) paw, consistent with the development of inflammation-evoked mechanical hypersensitivity. AM12814 reversed CFA-induced mechanical hypersensitivity in a dose- and time-dependent manner (Figure 4A; Drug Treatment: F(4, 328) = 104.5, p < 0.0001; Time: F(7, 328) = 24.61, p < 0.0001; Interaction: F(28, 328) = 8.741, p < 0.0001). AM12814 at doses of 1 and 3 mg/kg increased paw withdrawal thresholds in the CFA-injected paw throughout the observation period relative to either vehicle or each of the lower doses (p < 0.0001), and with similar efficacies (p = 0.2621), whereas the lower doses (0.1 and 0.3 mg/kg, i.p.) were inactive (p = 0.6912 and p = 0.9097, respectively). Both the 1 mg/kg (p = 0.0083) and 3 mg/kg (p < 0.0001) doses of AM12814 elevated mechanical thresholds relative to vehicle with antiallodynic effects lasting at least 3 h relative to vehicle treatment (p ≤ 0.0039 for each comparison). AM12814 also produced modest but reliable changes in paw withdrawal thresholds in the noninflamed contralateral paw that were both dose- and time-dependent, and paw withdrawal thresholds also varied modestly across the observation interval irrespective of drug treatment (Figure 4B; Drug Treatment: F(4, 328) = 6.066, p = 0.0001; Time: F(7, 328) = 4.218, p = 0.0002; Interaction: F(28, 328) = 3.065, p < 0.0001). Low doses of 0.1 and 0.3 mg/kg (i.p.) were associated with modest reductions in contralateral paw withdrawal thresholds relative to vehicle (p ≤ 0.0283) or the 1 mg/kg i.p. dose (p ≤ 0.0461), whereas higher doses (1 mg/kg, p = 0.9988; 3 mg/kg, p = 0.4228) showed no effect on paw withdrawal latencies in the noninflamed paw overall. Contralateral paw withdrawal thresholds differed modestly between groups at a subset of time points (p ≤ 0.0372).

Figure 4.

AM12814 suppresses CFA-induced mechanical hypersensitivity through a CB1-dependent mechanism. (A) AM12814 (1 and 3 mg/kg, i.p.) reduced CFA-induced mechanical hypersensitivity in a dose- and time-dependent manner in the inflamed (ipsilateral) paw (B) while producing only minor effects on mechanical paw withdrawal thresholds in the contralateral (noninflamed) paw. (C) The antiallodynic effect of AM12814 (1 mg/kg, i.p.) on the CFA-injected paw was blocked by pretreatment with AM251 (5 mg/kg, i.p.), whereas (D) only minor differences in paw withdrawal thresholds were observed in the contralateral (noninflamed) paw. (D) Data are expressed as mean ± SEM (n = 6–15 per group). (A, B): ^#0.1 mg/kg, ^0.3 mg/kg, *1 mg/kg, ⁺3 mg/kg AM12814 vs Vehicle. (C, D): *Vehicle–Vehicle, ⁺AM251-Vehicle, ^#AM251–AM12814 vs Vehicle-AM12814; ^Vehicle–Vehicle, ^$AM251-Vehicle vs AM251–AM12814. Number of symbols denotes significance level: ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05, two-way repeated measures ANOVA followed by Tukey’s post hoc test.

AM251 blocked AM12814-induced elevations in paw withdrawal thresholds in the CFA-injected (ipsilateral) paw (Figure 4C; Drug Treatment: F(3, 207) = 26.69, p < 0.0001; Time: F(8, 207) = 5.498, p < 0.0001; Interaction: F(24, 207) = 5.317, p < 0.0001), consistent with mediation by CB1. AM12814 (1 mg/kg i.p.) increased paw withdrawal thresholds in the CFA-injected paw overall relative to groups receiving either AM251–AM12814, AM251-Veh, or Veh–Veh treatment overall (p ≤ 0.0007), and this effect was most prominent from 10 to 60 min postinjection (p ≤ 0.0020). Ipsilateral paw withdrawal thresholds were modestly elevated in the AM251-Veh group compared to Veh–Veh treatment (p = 0.0004) overall; this effect was reliable at a subset of postinjection time points only (10 min: p = 0.0363; 3 h: p = 0.0270). In the contralateral paw, modest differences in paw withdrawal thresholds were observed between groups overall (Figure 4D; Drug Treatment: F(3, 207) = 2.778, p = 0.0422; Time: F(8, 207) = 0.8867, p = 0.5286; Interaction: F(24, 207) = 1.565, p = 0.0512), but post hoc comparisons failed to reveal any difference between treatment groups (p > 0.0546).

Discussion.

Compound AM12814, administered systemically, produced characteristic CB1-mediated cannabimimetic effects in a triad of behavioral tests assessing catalepsy, hypothermia, and tail-flick antinociception. These effects were dose-dependent, long-lasting, and blocked by the CB1 antagonist AM251. These observations are consistent with the actions of AM12814 at central CB1 receptors. AM12814 was notably more potent in producing tail-flick antinociception than it was in inducing catalepsy or hypothermia. Moreover, AM12814 also reversed CFA-induced mechanical hypersensitivity through a CB1-dependent mechanism, as evidenced by complete blockade of the antiallodynic effects of AM12814 by AM251.^23–25

DOCKING ON CANNABINOID RECEPTORS

Docking of the three key compounds AEA, AMG315, and AM12814 into the orthosteric binding site of the active-state crystal structures of AMG315-bound CB1 receptor (PDBID: 8GHV)¹⁶ and AM12033-bound CB2 receptor (PDBID: 6KPF)³ showed reproducible docking poses and good docking scores (Table 3). The predicted docking poses of all three compounds are consistent with the ligand conformations in the corresponding structures of CB1 and CB2 receptors, with the acyl tail binding deep in the binding pocket and headgroup engaging in polar interactions with the residues at the top of the binding pocket (Figure 5). However, predicted docking scores are in the same range as scores obtained by redocking of compound CP-55,940 into the available CP-55,940-bound structures of CB1 and CB2 receptors.

Table 3.

Docking results for AEA, AMG315, AM12814, and CP-55,940 on the crystal structures of CB1 and CB2 receptors.

Compound	Structure	Binding affinity (Ki)		CB1 Receptor		CB2 Receptor
		CB1	CB2	Docking Score	RTCNN Score	Docking Score	RTCNN Score
				(PDBID: 8GHV)		(PDBID: 6KPF)
AM12814		0.7 nM	3.4 nM	−31.9	−23.4	−25.7	−34.4
AMG315		7.8 nM	176 nM	−31.7	−26.4	−24.0	−32.4
AEA		74.5 nM	160 nM	−32.5	−23.8	−27.1	−27.5
				(PDBID: 7V3Z)		(PDBID: 8GUR)
CP-55,940		1.6 nM	3.6 nM	−27.5	−37.4	−27.0	−36.1

Figure 5.

Predicted docking poses of AMG315 (orange) and AM12814 (yellow) in the (A) AMG315-bound CB1 receptor (PDBID: 8ghv) and (B) AM12033-bound CB2 receptor in complex with Gi (PDBID: 6kpf).

The lipid headgroup is predicted to form H-bonds with residues H178 (TM2) and backbone of residue I267 (ECL2) of CB1 receptor (Figure 5A), while in CB2 receptor, hydrophilic headgroups form hydrogen bonds with residue H95 (TM2) and backbone of residues F94 (TM2) and L182 (ECL2) (Figure 5B). The acyl tail of the lipid molecules binds in the hydrophobic part of the CB1 and CB2 orthosteric binding pockets. Predicted binding poses also show hydrophobic interaction between the (13S)-Me group of compounds AMG315 and AM12814 and residue F200 of the CB1 receptor and residue F117 of the CB2 receptor, a key part of F3.36 and W6.48 “toggle switch” that is involved in activation of cannabinoid receptors.²⁶ The substitution of the 20-methyl group by a trifluoromethyl group is a known step to improve the activity of a compound in medicinal chemistry. Indeed, introduction of a trifluoromethyl group in the AM12814 compound improves the affinity by 1 order of magnitude for the CB1 receptor and 50 times for the CB2 receptor. The CF₃ group is pointing at the end of the hydrophobic pocket formed by residues L193, T197, F268, I271, Y275, L276, W279, and M363 within 5 Å from the trifluoromethyl group in CB1 receptor and residues I110, T114, P168, F183, I186, Y190, L191, W194, and M265 within 5 Å from the trifluoromethyl group in CB2 receptor. The difference in binding energy created by the introduction of trifluoromethyl group is known to be driven by electrostatic and solvation energy.²⁷ The predicted positions of the methyl group in compound AMG315 and trifluoromethyl group in the AM12814 compound largely overlap. One of the likely reasons behind the large shift in affinity is due to the difference in interactions with high-energy water molecules left in the hydrophobic pocket. CF₃ group being more polar than CH₃ can stabilize high-energy water molecules by improving electrostatic interactions and has a higher likelihood to displace these waters from the pocket due to its slightly larger size, providing gain in solvation energy. Generation of low-energy water configurations around the trifluoromethyl group of the AM12814 compound and the methyl group of the AMG315 compound in CB1 and CB2 binding pockets showed that introduction of CF₃ group leads to the displacement of 2 out of 19 water molecules in the CB1 receptor pocket and 2 out of 12 water molecules in the CB2 receptor pocket (Figure 6).

Figure 6.

Predicted low-energy water configurations around the trifluoromethyl group of AM12814 (yellow) and methyl group AMG315 (orange) in CB1(A) and CB2 (B) binding pockets. Water molecules predicted for AM12814-protein complexes are shown as red spheres; water molecules predicted for AMG315-protein complexes are shown as blue spheres.

CONCLUSIONS

Earlier work on our endocannabinoid lipid development project led to the identification of the first-generation chiral lead anandamide analogue (R)-N-(1-Methyl-2-hydroxyethyl)-13-(S)-methyl-arachidonamide (AMG315) as a potent in vitro and in vivo CB1 agonist with stability to all hydrolytic endocannabinoid metabolizing enzymes as well as to major oxidative enzyme COX-2. We have now synthesized a series of chiral tail-modified AMG315 analogs in an effort to further optimize this novel endocannabinoid chemotype in terms of cannabinoid receptor binding affinity and in vitro and in vivo potency and efficacy. Our advanced molecule identified in the current study, namely, 20,20,20-trifluoro-(R)-N-(1-methyl-2-hydroxyethyl)-13-(S)-methyl-arachidonamide (12), showed picomolar affinity for the CB1 receptor (K_i = 0.7 nM), and it is the first reported endocannabinoid analog with low nano-molar affinity for the CB2 receptor (K_i = 3.4 nM). In further in vitro functional characterization, the compound behaves as a potent CB1 and CB2 agonist, and like anandamide, it exhibits partial efficacy for CB1 and CB2 receptors. Our SAR results are supported by docking studies of the key analogs on the crystal structures of cannabinoid receptors.

In vivo, AM12814 produces the cardinal signs of CB1 activation, while the CFA-induced inflammatory pain model behaves as a very potent and long-lasting analgesic. Currently, this chiral endocannabinoid analogue represents the state-of-the-art anandamide-like probe molecule.

EXPERIMENTAL SECTION

Experimental Procedures, Spectroscopic, Analytical, and Physical Data for Compounds.

Materials.

All reagents and solvents were purchased from Aldrich Chemical Co., unless otherwise specified, and used without further purification. All anhydrous reactions were performed under a static argon or nitrogen atmosphere in flame-dried glassware by using scrupulously anhydrous solvents. Flash column chromatography employed silica gel 60 (230–400 mesh). All compounds were demonstrated to be homogeneous by analytical TLC on precoated silica gel TLC plates (Merck, 60 F245 on glass, layer thickness 250 μm), and chromatograms were visualized by phosphomolybdic acid staining. NMR spectra were recorded in CDCl₃, unless otherwise stated, on Varian 500 (¹H at 500 MHz) and Bruker 400 (¹H at 400 MHz, ¹³C at 100 MHz) NMR spectra, and chemical shifts are reported in units of δ relative to internal TMS. Multiplicities are indicated as br (broadened), s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet) and coupling constants (J) are reported in hertz (Hz). Low and high-resolution mass spectra were performed in the School of Chemical Sciences, University of Illinois at Urbana–Champaign. Mass spectral data are reported in the form of m/z (intensity relative to base = 100). Purities of the tested compounds were determined by LC/MS analysis using a Waters MicroMass ZQ system (electrospray ionization, ESI) with Waters-2525 binary gradient module coupled to a photodiode array detector (Waters-2996) and ELS detector (Waters-2424) using a XTerra MS C18 (5 μm, 4.6 mm × 50 mm column and acetonitrile/water) and were >95%, or by HPLC analysis using an Agilent Infinity 1260 II system coupled to photodiode array detector (Agilent Infinity 1260 II Diode Array Detector HR) using a ZORBAX RX-SIL (4.6 mm × 150 mm, 5 μm column and isopropanol/water) and were >95%.

No unexpected or unusually high safety hazards were encountered.

2-((6-Bromohexyl)oxy)tetrahydro-2H-pyran (2).

To a stirred solution of commercially available 1 (1 g, 5.50 mmol) and dihydropyran (696 mg, 8.29 mmol) in anhydrous DCM (17 mL) was added PPTS (138 mg, 0,55 mmol) under an argon atmosphere, and the reaction mixture was stirred at ambient temperature for 4 h. The reaction mixture was then diluted with DCM and washed with water. The phases were separated, and the organic phase was washed with brine, dried over MgSO₄, and evaporated. Chromatography on silica gel (10–20% Et₂O-hexanes) afforded 1.37 g of the title compound as a colorless oil in 94% yield. The NMR spectra of this compound match the ones reported in the literature.²⁸

Triphenyl(6-((tetrahydro-2H-pyran-2-yl)oxy)hexyl)phosphonium Bromide (3).

To a stirred solution of 2 (1.37 g, 5.18 mmol) in anhydrous MeCN (22 mL) was added triphenylphosphine (2.69 g, 10.26 mmol), and the mixture was stirred at 75 °C for 7 days. The volatiles were evaporated under reduced pressure, and the resulting residue was chromatographed on silica gel (0–10% methanol-dichloromethane) to afford 2.71 g of pure title compound as a yellow gum in 99% yield. The NMR spectra of this compound match the ones reported in the literature.²⁸

Triphenyl(6,6,6-trifluorohexyl)phosphonium Bromide (5).

To a stirred solution of 4 (5 g, 22.83 mmol) in anhydrous MeCN (91 mL) was added triphenylphosphine (11.96 g, 45.66 mmol), and the mixture was stirred at 75 °C for 8 days. The volatiles were evaporated under reduced pressure, and the resulting residue was chromatographed on silica gel (0–10% methanol-dichloromethane) to afford 10.5 g of pure title compound as a yellow gum in 96% yield. The NMR spectra of this compound match the ones reported in the literature.²⁹

Methyl (S,5Z,8Z,11Z,14Z)-20,20,20-Trifluoro-13-methylicosa-5,8,11,14-tetraenoate (9a).

To a stirred solution of 5 (4.6 g, 9.60 mmol) in anhydrous THF (45 mL) at −78 °C under an argon atmosphere was added potassium bis(trimethylsilyl)amide (6.4 mL, 1 M solution in THF). The mixture was stirred for 30 min to ensure complete formation of the orange ylide, and then it was cooled to −115 °C. Subsequently, a solution of crude aldehyde 8 (850 mg, 3.20 mmol) in anhydrous THF (6 mL) was added dropwise. The reaction mixture was stirred for 25 min at −115 °C and then warmed to 0 °C over a 2-h period. The reaction mixture was then cooled to −115 °C and quenched with a saturated aqueous sodium bicarbonate solution. The mixture was warmed to room temperature, the phases were separated, the aqueous phase was extracted with Et₂O, and the combined organic extracts were washed with brine, dried over MgSO₄, and concentrated under reduced pressure. The residue was chromatographed on silica gel (0–10% Et₂O-hexanes) to afford 920 mg of the title compound in 75% yield as a yellow oil. ¹H NMR (500 MHz, CDCl₃) δ 5.43–5.32 (m, 4H, 5,6,8,9-H), 5.32–5.21 (m, 4H, 11,12,14,15-H), 3.66 (s, 3H, OMe), 3.48–3.43 (m, 1H, 13-H), 2.90–2.76 (m, 4H, 7,10-H), 2.32 (t, J = 7.5 Hz, 2H, 2-H), 2.16–2.01 (m, 6H, 4,16,19-H), 1.71 (p, J = 7.5 Hz, 2H, 3-H), 1.63–1.52 (m, 2H, 18-H), 1.43 (p, J = 7.4 Hz, 2H, 17-H), 1.02 (d, J = 6.7 Hz, 3H, 13-Me). ¹³C NMR (126 MHz, CDCl₃) δ 174.0, 135.1, 134.8, 129.0, 128.8, 128.2, 127.2 (q, J = 276.3 Hz), 127.1, 125.8, 51.5, 33.7 (q, J = 28.3 Hz), 33.4, 30.4, 28.8, 27.1, 26.5, 25.8, 25.6, 24.8, 21.9, 21.5 (d, J = 2.8 Hz), 15.3.

Methyl (5Z,8Z,11Z,13S,14Z)-13-Methyl-20-((tetrahydro-2H-pyran-2-yl)oxy)icosa-5,8,11,14-tetraenoate (9b).

To a stirred solution of 3 (2.38 g, 4.52 mmol) in anhydrous THF (23 mL) at −78 °C under an argon atmosphere was added potassium bis(trimethylsilyl)amide (3 mL of 1 M solution in THF). The mixture was stirred for 30 min to ensure complete formation of the orange ylide, and then, it was cooled to −115 °C. Subsequently, a solution of crude aldehyde 8 (389 mg, 1.51 mmol) in anhydrous THF (4.5 mL) was added dropwise. The reaction mixture was stirred for 25 min at −115 °C and then warmed to 0 °C over a 2-h period. The reaction mixture was then cooled to −115 °C and quenched with a saturated aqueous sodium bicarbonate solution. The mixture was warmed to room temperature, the phases were separated, the aqueous phase was extracted with Et₂O, and the combined organic extracts were washed with brine, dried over MgSO₄, and concentrated under reduced pressure. The residue was chromatographed on silica gel (0–10% Et₂O-hexanes) to afford 550 mg of the title compound in 85% yield as a yellow oil. ¹H NMR (400 MHz, CDCl₃) δ 5.46–5.32 (m, 4H, 5,6,8,9-H), 5.31–5.17 (m, 4H, 11,12,14,15-H), 4.62–4.57 (m, 1H, 2′-H), 3.92–3.83 (m, 1H, 6′-H), 3.73 (dt, J = 9.6, 7.6 Hz, 1H, 20-H), 3.65 (s, 3H, -OMe), 3.55–3.47 (m, 1H, 13-H), 3.43 (dt, J = 9.6, 7.6 Hz, 1H, 20-H), 3.42–3.33 (m, 1H, 6′-H), 2.89–2.75 (m, 4H, 7,10-H), 2.35 (t, J = 7.5 Hz, 2H, 2-H), 2.18–1.99 (m, 4H) [overlapping patterns i.e., 2.12 (dt, J = 7.5, 7.1 Hz, 2H, 4-H), 2.09–2.01 (m, 2H, 16-H)], 1.86–1.80 (m, 1H, 3′-H), 1.76–1.65 (m, 3H) [overlapping patterns i.e., 1.71 (p, J = 7.5 Hz, 2H, 3-H), 1.72–1.68 (m, 1H, 3′-H)], 1.64–1.56 (m, 4H, 19-H, 5′-H), 1.57–1.47 (m, 2H, 4′-H), 1.41–1.32 (m, 4H, 17,18-H), 1.01 (d, J = 6.7 Hz, 3H, 13-Me).

(S,5Z,8Z,11Z,14Z)-20,20,20-Trifluoro-13-methylicosa-5,8,11,14-tetraenoic Acid (10a).

To a stirred solution of 9a (840 mg, 2.18 mmol) in THF (43 mL) at room temperature under an argon atmosphere was added lithium hydroxide (11 mL, 1 M solution in water). The reaction mixture was stirred at the same temperature overnight. The reaction mixture was acidified with 5% HCl to pH 3 and extracted with Et₂O. The combined organic extracts were washed with brine, dried over MgSO₄, and evaporated under reduced pressure. The resulting residue was chromatographed on silica gel (30–50% EtOAc-hexanes) to afford 620 mg of the title carboxylic acid in 77% yield as a colorless oil. ¹H NMR (500 MHz, CDCl₃) δ 5.45–5.33 (m, 4H, 5,6,8,9-H), 5.32–5.21 (m, 4H, 11,12,14,15-H), 3.44–3.40 (m, 1H, 13-H), 2.89–2.77 (m, 4H, 7,10-H), 2.37 (t, J = 7.5 Hz, 2H, 2-H), 2.17–2.00 (m, 6H, 4,16,19-H), 1.72 (p, J = 7.5 Hz, 2H, 3-H), 1.62–1.52 (m, 2H, 18-H), 1.43 (p, J = 7.4 Hz, 2H, 17-H), 1.02 (d, J = 6.7 Hz, 3H, 13-Me). ¹³C NMR (126 MHz, CDCl₃) δ 174.0, 135.2, 134.6, 129.3, 128.7, 128.2, 127.2 (q, J = 276.3 Hz), 127.1, 125.8, 33.8 (q, J = 28.3 Hz), 33.4, 30.6, 28.9, 27.0, 26.5, 25.2, 25.5, 24.8, 21.8, 21.5 (d, J = 2.8 Hz), 15.1.

(5Z,8Z,11Z,13S,14Z)-13-Methyl-20-((tetrahydro-2H-pyran-2-yl)oxy)icosa-5,8,11,14-tetraenoic Acid (10b).

To a stirred solution 9b (500 mg, 1.16 mmol) in THF (23) at room temperature under an argon atmosphere was added lithium hydroxide (5.8 mL, 1 M solution in water). The reaction mixture was stirred at the same temperature overnight. The reaction mixture was acidified with 5% HCl to pH 3 and extracted with Et₂O. The combined organic extracts were washed with brine, dried over MgSO₄, and evaporated under reduced pressure. The resulting residue was chromatographed on silica gel (10–20% ethyl acetate–hexanes) to afford 400 mg of the title carboxylic acid in 83% yield as a colorless oil. ¹H NMR (400 MHz, CDCl₃) δ 5.46–5.32 (m, 4H, 5,6,8,9-H), 5.31–5.17 (m, 4H, 11,12,14,15-H), 4.61–4.57 (m, 1H, 2′-H), 3.92–3.83 (m, 1H, 6′-H), 3.73 (dt, J = 9.6, 6.9 Hz, 1H, 20-H), 3.55–3.47 (m, 1H, 13-H), 3.45–3.41 (m, 1H, 6′-H), 3.42 (dt, J = 9.6, 6.9, 20-H), 2.85–2.80 (m, 4H, 7,10-H), 2.35 (t, J = 7.5 Hz, 2H, 2-H), 2.18–1.99 (m, 4H, 4,16-H), 1.82–1.79 (m, 1H, 3′-H), 1.76–1.65 (m, 3H, 3-H, 3′-H), 1.64–1.56 (m, 6H, 19-H, 4′,5′-H), 1.41–1.32 (m, 4H, 17,18-H), 1.01 (d, J = 6.7 Hz, 3H, 13-Me).

(S,5Z,8Z,11Z,14Z)-N-((R)-1-((tert-Butyldiphenylsilyl)oxy)propan-2-yl)-20,20,20-trifluoro-13-methylicosa-5,8,11,14-tetraenamide (11a).

A solution of 10a (300 mg, 0.806 mmol) and dried carbonyldiimidazole (650 mg, 4.03 mmol) in anhydrous THF (16 mL) at room temperature under an argon atmosphere was stirred for 2 h, and then a solution of 17 (500 mg, 1.61 mmol) in THF was added. The reaction mixture was stirred for 2 h and then diluted with water and ethyl acetate. The organic phase was separated, and the aqueous phase was extracted with ethyl acetate. The combined organic layer was washed with brine, dried over MgSO₄, and concentrated under reduced pressure. The residue was chromatographed on silica gel (10–25% ethyl acetate–hexane) to afford 455 mg of the title amide in 86% yield as a light-yellow oil. ¹H NMR (500 MHz, CDCl₃) δ 7.63–7.55 (m, 4H, Ar–H), 7.46–7.35 (m, 6H, Ar–H), 5.58 (br d, J = 7.4 Hz, 1H, NH), 5.44–5.33 (m, 4H, 5,6,8,9-H), 5.33–5.20 (m, 4H, 11,12,14,15-H), 4.15–4.07 (m, 1H, 1′-H), 3.68 (dd, J = 10.1, 4.1 Hz, 1H, 2′-H), 3.60 (dd, J = 10.1, 6.1 Hz, 1H, 2′-H), 3.50–3.39 (m, 1H, 13-H), 2.89–2.76 (m, 4H, 7,10-H), 2.13–2.01 (m, 8H, 2,4,16,19-H), 1.72 (p, J = 7.5 Hz, 2H, 3-H), 1.59–1.52 (m, 2H, 18-H), 1.43 (p, J = 7.3 Hz, 2H, 17-H), 1.19 (d, J = 6.7 Hz, 3H, 1′-Me), 1.08 (s, 9H, t-bu), 1.02 (d, J = 6.8 Hz, 3H, 13-Me). ¹³C NMR (126 MHz, CDCl₃) δ 171.9, 135.5, 135.4, 135.0, 134.8, 133.3, 133.2, 129.8, 129.2, 128.6, 128.2, 128.2, 127.8, 127.2 (q, J = 276.3 Hz), 127.6, 125.7, 66.8, 60.4, 46.2, 36.3, 33.6 (q, J = 28.4 Hz), 30.5, 28.7, 27.0, 26.9, 26.7, 25.8, 25.6, 25.6, 21.9, 21.5 (q, J = 2.9 Hz), 21.6, 19.3, 17.6, 15.2, 14.1.

(5Z,8Z,11Z,13S,14Z)-N-((R)-1-((tert-Butyldiphenylsilyl)oxy)propan-2-yl)-13-methyl-20-((tetrahydro-2H-pyran-2-yl)oxy)icosa-5,8,11,14-tetraenamide (11b).

This compound was synthesized from 10b (400 mg, 0.56 mmol) in a manner similar to that for 11a. The crude residue was chromatographed on silica gel (10–25% ethyl acetate–hexane) to afford 450 mg of the title amide in 80% yield as a light-yellow oil. ¹H NMR (400 MHz, CDCl₃) δ 7.64–7.59 (m, 4H, Ar–H), 7.48–7.35 (m, 6H, Ar–H), 5.60 (br d, J = 7.3 Hz, 1H, NH), 5.44–5.33 (m, 4H, 5,6,8,9-H), 5.33–5.19 (m, 5H, 11,12,14,15-H, NH), 4.60–4.54 (m, 1H, 2″-H), 4.13–4.10 (m, 1H, 1′-H), 3.40–3.85 (m, J = 1H, 6″-H), 3.79–3.71 (m, 2H, 20-H), 3.68 (dd, J = 10.1, 4.2 Hz, 1H, 2′-H), 3.61 (dd, J = 10.1, 5.2 Hz, 1H, 2′-H), 3.54–3.45 (m, 1H, 13-H), 3.38–3.34 (m, 1H, 6″-H), 2.83–2.80 (m, 4H, 7,10-H), 2.18–2.06 (m, 6H, 2,4,16-H), 1.83–1.78 (m, 1H, 3″-H), 1.70–1.66 (m, 3H, 3″-H, 3-H), 1.65–1.48 (m, 4H, 19-H, 5″-H), 1.47–1.31 (m, 6H, 4″-H, 17,18-H), 1.19 (d, J = 6.7 Hz, 3H, 1′-Me), 1.08 (s, 9H, t-bu), 1.02 (d, J = 6.7 Hz, 3H, 13-Me).

(S,5Z,8Z,11Z,14Z)-20,20,20-Trifluoro-N-((R)-1-hydroxypropan-2-yl)-13-methylicosa-5,8,11,14-tetraenamide (12).

To a stirred solution of 11a (150 mg, 0.22 mmol) in anhydrous THF (3 mL) was added tetra-butylammonium fluoride (0.6 mL, 1 M solution in THF) dropwise at 0 °C under an argon atmosphere. The reaction mixture was stirred at the same temperature for 10 min and then warmed to ambient temperature and stirred for 1.5 h. The reaction was quenched via the addition of aq. sat. NH₄Cl solution, and the two phases were separated. The aqueous phase was then extracted with EtOAc, and the combined organic phases were washed with brine, dried over MgSO₄, and evaporated. The resulting residue was chromatographed on silica gel (70–100% EtOAc-hexanes) to afford 80 mg of the title compound in 85% yield as a light-yellow oil. ¹H NMR (500 MHz, CDCl₃) δ 5.60 (br s, 1H, NH), 5.45–5.32 (m, 4H, 5,6,8,9-H), 5.32–5.19 (m, 4H, 11,12,14,15), 4.07–4.03 (m, 1H, 1′-H), 3.66 (dd, J = 11.0, 3.8 Hz, 1H, 2′-H), 3.52 (dd, J = 11.0, 6.1,, 1H, 2′-H), 3.50–3.39 (m, 1H, 13-H), 2.89–2.76 (m, 4H, 7,10-H), 2.19 (t, J = 7.4 Hz, 2H, 2-H), 2.16–2.00 (m, 6H, 4,16,19-H), 1.72 (p, J = 7.4 Hz, 2H, 3-H), 1.56–1.50 (m, 2H, 18-H), 1.43 (p, J = 7.3 Hz, 2H, 17-H), 1.16 (d, J = 6.9 Hz, 3H, 1′-Me), 1.02 (d, J = 6.9 Hz, 3H, 13-Me). ¹³C NMR (126 MHz, CDCl₃) δ 173.7, 135.0, 134.8, 129.1, 128.7, 128.3, 128.2, 127.4 (q, J = 275.1 Hz), 127.0, 125.7, 67.4, 67.3, 47.8, 36.0, 33.6 (q, J = 28.3 Hz), 30.5, 28.7, 27.0, 26.6, 25.8, 25.6, 21.9, 21.5 (q, J = 3.0 Hz), 17.0. Mass spectrum (ESI) m/z (relative intensity): 430 (100), 431 (23). Exact mass (ESI) calculated for C₂₄H₃₉NO₂F₃ [M + H]⁺, 430.2933; found, 430.2939. HPLC analysis (Agilent 1260 Infinity II, ZORBAX RX-SIL) showed a purity of 98% and retention time of 4.8 min for the title compound.

(S,5Z,8Z,11Z,14Z)-N-((R)-1-((tert-Butyldiphenylsilyl)oxy)propan-2-yl)-20-hydroxy-13-methylicosa-5,8,11,14-tetraenamide (13).

To a stirred solution of 11b (400 mg, 0.56 mmol) in ethanol (4.5 mL) at room temperature and under an argon atmosphere was added pyridinium p-toluenesulfonate (15 mg, 0.056 mmol), and the mixture was heated to 55 °C and stirred at the same temperature for 4 h. The solvent was evaporated under reduced pressure, and the resulting residue was chromatographed on silica gel (10–30% ethyl acetate–hexanes) to afford 320 mg of the title compound in 91% yield as a light-yellow oil. ¹H NMR (500 MHz, CDCl₃) δ 7.67–7.61 (m,, 4H, Ar–H), 7.46–7.35 (m, 6H, Ar–H), 5.66 (br d, J = 7.3 Hz, 1H, NH), 5.44–5.33 (m, 4H, 5,6,8,9-H), 5.31–5.19 (m, 4H, 11,12,14,15-H), 4.13–4.09 (m, 1H, 1′-H), 3.68 (dd, J = 10.1, 4.2 Hz, 1H, 2′-H), 3.65–3.62 (m, 2H, 20-H), 3.61 (dd, J = 10.1, 5.8 Hz, 1H, 2′-H), 3.46–3.40 (m, 1H, 13-H), 2.83–2.76 (m, 4H, 7,10-H), 2.18–2.06 (m, 6H, 2,4,16-H), 1.73 (p, J = 7.5 Hz, 2H, 3-H), 1.65–1.48 (m, 2H, 19-H), 1.47–1.31 (m, 4H, 17,18-H), 1.19 (d, J = 6.7 Hz, 3H, 1′-Me), 1.08 (s, 9H, t-bu), 1.02 (d, J = 6.7 Hz, 3H, 13-Me).

(S,5Z,8Z,11Z,14Z)-20-Bromo-N-((R)-1-((tert-butyldiphenylsilyl)oxy)propan-2-yl)-13-methylicosa-5,8,11,14-tetraenamide (14).

A solution of 13 (280 mg, 0.44 mmol) in anhydrous dichloromethane (3 mL) was cooled to −25 °C under an argon atmosphere, and carbon tetrabromide (192 mg, 0.58 mmol) and triphenylphosphine (152 mg, 0.58 mmol) were added. The reaction mixture was stirred at−25 °C for 1 h and then at 0 °C for 30 min. The volatiles were removed under reduced pressure, and the resulting residue was chromatographed on silica gel (0–20% EtOAc-hexanes) to afford 280 mg of the title compound in 92% yield as a light-yellow oil. ¹H NMR (500 MHz, CDCl₃) δ 7.67–7.61 (m, 4H, Ar–H), 7.49–7.35 (m, 6H, Ar–H), 5.57 (br d, J = 7.9 Hz, 1H, NH), 5.45–5.31 (m, 4H, 5,6,8,9-H), 5.31–5.19 (m, 4H, 11,12,14,15-H), 4.15–4.09 (m, 1H, 1′-H), 3.68 (dd, J = 10.1, 4.1 Hz, 1H, 2′-H), 3.60 (dd, J = 10.1, 5.9 Hz, 1H, 2′-H), 3.46–3.38 (m, 1H, 13-H), 3.39 (t, J = 6.8 Hz, 2H, 20-H), 2.91–2.78 (m, 4H, 7,10-H), 2.18–2.06 (m, 6H, 2,4,16-H), 1.85 (p, J = 6.8 Hz, 2H, 19-H), 1.72 (p, J = 7.5 Hz, 2H, 3-H), 1.50–1.34 (m, 4H, 17,18-H), 1.19 (d, J = 6.7 Hz, 3H, 1′-Me), 1.08 (s, 9H, t-bu), 1.02 (d, J = 6.7 Hz, 3H, 13-Me).

(S,5Z,8Z,11Z,14Z)-20-Bromo-N-((R)-1-hydroxypropan-2-yl)-13-methylicosa-5,8,11,14-tetraenamide (15).

Compound 14 (260 mg, 0.376 mmol) was dissolved in anhydrous tetrahydrofuran (7.5 mL), and to the stirred mixture was added tetra-n-butylammonium fluoride (0.6 mL, 1 M solution in THF) at 0 °C under an argon atmosphere. Stirring was continued at 0 °C for 10 min and then for 1.5 h at ambient temperature. The reaction was kept at 0 °C for 10 min and then at room temperature for 1.5 h. The mixture was quenched by the addition of sat. aq. NH₄Cl solution, extracted with diethyl ether, and the combined organic extracts were washed with brine, dried over MgSO₄, and concentrated under reduced pressure. The residue was chromatographed on silica gel (30–60 EtOAc-hexanes) to afford 150 mg of the title bromide in 88% yield as a light-yellow oil. ¹H NMR (500 MHz, CDCl₃) 5.57 (br d, J = 8.3 Hz, 1H, NH), 5.46–5.32 (m, 4H, 5,6,8,9-H), 5.31–5.19 (m, 4H, 11,12,14,15-H), 4.08–4.04 (m, 1H, 1′-H), 3.67 (dd, J = 10.1, 4.2 Hz, 1H, 2′-H), 3.53 (dd, J = 10.1, 6.0 Hz, 1H, 2′-H), 3.45–3.39 (m, 1H, 13-H), 3.41 (t, J = 6.8 Hz, 2H, 20-H), 2.91–2.78 (m, 4H, 7,10-H), 2.20 (t, J = 7.5 Hz, 2H, 2-H), 2.16–2.06 (m, 4H, 4,16-H), 1.86 (p, J = 6.8 Hz, 2H, 19-H), 1.73 (p, J = 7.5 Hz, 2H, 3-H), 1.58 (bs, 1H, OH), 1.50–1.34 (m, 4H, 17,18-H), 1.17 (d, J = 6.7 Hz, 3H, 1′-Me), 1.02 (d, J = 6.7 Hz, 3H, 13-Me). Mass spectrum (ESI) m/z (relative intensity): 454 (100), 455 (23), 456 (100), 457 (25). Exact mass (ESI) calculated for C₂₄H₄₁NO₂Br [M + H]⁺, 454.2315; found, 454.2321. HPLC analysis (Agilent 1260 Infinity II, ZORBAX RX-SIL) showed 98% purity and 4.8 min retention time for the title compound.

(S,5Z,8Z,11Z,14Z)-20-Cyano-N-((R)-1-hydroxypropan-2-yl)-13-methylicosa-5,8,11,14-tetraenamide (16a).

To a stirred solution of 15 (70 mg, 0.154 mmol) in anhydrous DMSO (1.5 mL) was added NaCN (75 mg, 1.54 mmol) under an argon atmosphere, and the reaction mixture was stirred at ambient temperature overnight. The reaction mixture was quenched by the addition of ice-cold water. Et₂O was then added, and the heterogeneous mixture was stirred vigorously for 30 min. The two phases were then separated, and the aqueous phase was extracted with Et₂O. The combined organic phases were washed with brine, dried over MgSO₄, and evaporated. The resulting residue was chromatographed on silica gel (100% EtOAc) to afford 40 mg of the title compound in 66% as a colorless oil. ¹H NMR (500 MHz, CDCl₃) 5.65 (br d, J = 8.3 Hz, 1H, NH), 5.44–5.31 (m, 4H, 5,6,8,9-H), 5.31–5.20 (m, 4H, 11,12,14,15-H), 4.07–4.02 (m, 1H, 1′-H), 3.67 (dd, J = 10.1, 4.2 Hz, 1H, 2′-H), 3.52 (dd, J = 10.1, 6.1 Hz, 1H, 2′-H), 3.48–3.40 (m, 1H, 13-H), 2.87–2.76 (m, 4H, 7,10-H), 2.34 (t, J = 7.1 Hz, 2H, 20-H), 2.20 (t, J = 7.6 Hz, 2H, 2-H), 2.16–2.06 (m, 4H, 4,16-H), 1.73 (p, J = 7.6 Hz, 2H, 3-H), 1.67 (t, J = 7.1 Hz, 2H, 19-H) 1.56 (bs, 1H, OH), 1.51–1.40 (m, 4H, 17,18-H), 1.17 (d, J = 6.7 Hz, 3H, 1′-Me), 1.02 (d, J = 6.7 Hz, 3H, 13-Me). ¹³C NMR (126 MHz, CDCl₃) δ: 173.7, 134.9, 134.8, 129.1, 128.5, 128.3, 128.2, 127.3, 125.7, 119.8, 67.4, 47.8, 36.0, 30.5, 28.8, 28.3, 28.3, 27.1, 26.5, 25.8, 25.6, 25.5, 25.3, 22.0, 17.1. Mass spectrum (ESI) m/z (relative intensity) 403 (100), 404 (23). Exact mass (ESI) calcd for C₂₅H₄₁N₂O₂ [M + H]⁺, 401.3168; found, 401.3179. HPLC analysis (Agilent 1260 Infinity II, ZORBAX RX-SIL) showed 96% purity and 4.8 min retention time for the title compound.

(S,6Z,9Z,12Z,15Z)-20-(((R)-1-Hydroxypropan-2-yl)amino)-8-methyl-20-oxoicosa-6,9,12,15-tetraen-1-yl Nitrate (16b).

To a solution of 15 (40 mg, 0.091 mmol) in anhydrous CH₃CN (1.8 mL) was added AgNO₃ (231 mg, 1.363 mmol) under an argon atmosphere at room temperature. The reaction mixture was heated at 80 °C overnight. The mixture was then passed through a pad of Celite and washed with EtOAc. The organic phase was concentrated, and the resulting crude residue was chromatographed on silica gel (100% EtOAc) to afford 27 mg of pure title compound in 70% yield as a colorless oil. ¹H NMR (500 MHz, CDCl₃) 5.56 (br d, J = 8.3 Hz, 1H, NH), 5.45–5.32 (m, 4H, 5,6,8,9-H), 5.32–5.18 (m, 4H, 11,12,14,15-H), 4.44 (t, J = 6.7 Hz, 2H, 20-H), 4.08–4.03 (m, 1′-H), 3.67 (dd, J = 10.1, 3.8 Hz, 1H, 2′-H), 3.53 (dd, J = 10.1, 5.8 Hz, 1H, 2′-H), 3.45–3.38 (m, 1H, 13-H), 2.93–2.71 (m, 4H, 7,10-H), 2.20 (t, J = 7.4 Hz, 2H, 2-H), 2.15–1.97 (m, 4H, 4,16-H), 1.73 (p, J = 7.4 Hz, 2H, 3-H), 1.70 (p, J = 6.7 Hz, 2H, 19-H) 1.56 (bs, 1H, OH), 1.51–1.40 (m, 4H, 17,18-H), 1.17 (d, J = 6.7 Hz, 3H, 1′-Me), 1.02 (d, J = 6.7 Hz, 3H, 13-Me). ¹³C NMR (126 MHz, CDCl₃) δ: 173.7, 134.9, 134.8, 129.1, 128.7, 128.3, 128.2, 127.4, 125.7, 73.3, 67.5, 47.8, 36.0, 30.5, 29.2, 27.2, 27.2, 26.7, 26.6, 25.8, 25.6, 25.3, 22.0, 17.0. Mass spectrum (ESI) m/z (relative intensity) 438 (100), 439 (23). Exact mass (ESI) calcd for C₂₄H₄₁N₂O₅ [M + H]⁺, 437.3010; found 437.3002. HPLC analysis (Agilent 1260 Infinity II, ZORBAX RX-SIL) showed 96% purity and 5.4 min retention time for the title compound.

(S,5Z,8Z,11Z,14Z)-20-Hydroxy-N-((R)-1-hydroxypropan-2-yl)-13-methylicosa-5,8,11,14-tetraenamide (18).

To a stirred solution of 13 (200 mg, 0.32 mmol) in anhydrous THF (1 mL) was added TBAF (510 μL, 0.51 mmol, 1 M solution in anhydrous THF) at 0 °C under an argon atmosphere. The reaction mixture was then warmed to ambient temperature and stirred for 2 h. The reaction was quenched with sat. NH₄Cl solution, and the phases were separated. The aqueous phase was then extracted with EtOAc, and the combined organic phases were washed with brine, dried over MgSO₄, and evaporated. The resulting crude residue was chromatographed on silica gel (60–80% EtOAc-hexanes) to afford 100 mg of pure title compound in 84% yield. ¹H NMR (500 MHz, CDCl₃) 5.73 (br d, J = 7.9 Hz, 1H, NH), 5.45–5.33 (m, 4H, 5,6,8,9-H), 5.31–5.21 (m, 4H, 11,12,14,15-H), 4.10–4.06 (m, 1H, 1′-H), 3.67 (dd, J = 10.1, 3.9 Hz, 1H, 2′-H), 3.63 (t, J = 6.9 Hz, 2H), 3.51 (dd, J = 10.1, 5.1 Hz, 1H, 2′-H), 3.49–3.42 (m, 1H, 13-H), 2.90–2.79 (m, 4H, 7,10-H), 2.20 (t, J = 7.1 Hz, 2H, 2-H), 2.16–2.05 (m, 4H, 4,16-H), 1.73 (p, J = 7.1 Hz, 2H, 3-H), 1.57 (p, J = 6.9 Hz, 2H, 19-H) 1.56 (bs, 2H, 2xOH), 1.43–1.34 (m, 4H, 17,18-H), 1.16 (d, J = 6.7 Hz, 3H, 1′-Me), 1.01 (d, J = 6.7 Hz, 3H, 13-Me). ¹³C NMR (126 MHz, CDCl₃) δ: 173.6, 134.7, 135.1, 129.1, 128.7, 128.3, 128.2, 127.4, 125.7, 67.5, 61.4, 47.8, 36.0, 30.5, 29.2, 27.5, 27.3, 26.7, 26.3, 25.8, 25.4, 25.3, 22.1, 17.1. Exact mass (ESI) calculated for C₂₄H₄₃NO₃ [M + 2H]⁺, 393.3232; found 393.3219. HPLC analysis (Agilent 1260 Infinity II, ZORBAX RX-SIL) showed 97% purity and 5.5 min retention time for the title compound.

(S,6Z,9Z,12Z,15Z)-20-(((R)-1-((tert-Butyldiphenylsilyl)oxy)propan-2-yl)amino)-8-methyl-20-oxoicosa-6,9,12,15-tetraen-1-yl methanesulfonate (19).

To a stirred solution of 13 (100 mg, 0.158 mmol) in anhydrous DCM (1.5 mL) were added under an argon atmosphere at 0 °C methanesulfonyl chloride (22 mg, 15 μL, 0.190 mmol) and triethylamine (26 mg, 35 μL, 0.252 mmol). The reaction mixture was then stirred at the same temperature for 20 min. The reaction was quenched by the addition of sat. NH₄Cl solution, and the phases were separated. The aqueous phase was then extracted with DCM, and the combined organic phases were washed with brine, dried over MgSO₄, and evaporated. The resulting crude product was used for the next step without further purification.

(S,6Z,9Z,12Z,15Z)-20-(((R)-1-Hydroxypropan-2-yl)amino)-8-methyl-20-oxoicosa-6,9,12,15-tetraen-1-yl Methanesulfonate (20).

To a stirred solution of 19 (30 mg, 0.048 mmol) in anhydrous THF (4 mL) was added under an argon atmosphere at 0 °C TBAF (0.15 mL, 1 M solution in anhydrous THF), and the reaction mixture was stirred at ambient temperature for 3 h. The reaction was quenched via the addition of sat. NH₄Cl, and the resulting phases were separated. The aqueous phase was then extracted with EtOAc, and the combined organic phases were washed with brine, dried over MgSO₄, and evaporated. The resulting crude residue was chromatographed on silica gel (60–80% EtOAc-hexanes) to afford 16 mg of title compound as a colorless oil in 82% yield. ¹H NMR (500 MHz, CDCl₃) δ 5.74 (br d, J = 7.9 Hz, 1H, NH), 5.42–5.30 (m, 4H, 5,6,8,9-H), 5.30–5.19 (m, 4H, 11,12,14,15-H), 4.21 (t, J = 6.9 Hz, 2H, 20-H), 4.05–3.99 (m, 1H, 1′-H), 3.65 (dd, J = 11.0, 3.5 Hz, 1H, 2′-H), 3.51 (dd, J = 11.0, 6.3 Hz, 1H, 2′-H), 3.44–3.39 (m, 1H, 13-H), 3.00 (s, 3H, OMs), 2.89–2.74 (m, 4H, 7,10-H), 2.19 (t, J = 7.4 Hz, 2H, 2-H), 2.15–2.03 (m, 4H, 4,16-H), 1.81–1.66 (m, 4H, 3,19-H), 1.48–1.35 (m, 4H, 17,18-H), 1.16 (d, J = 6.9 Hz, 3H, 1′-Me), 1.01 (d, J = 6.8 Hz, 3H, 13-Me).

(S,5Z,8Z,11Z,14Z)-20-Fluoro-N-((R)-1-hydroxypropan-2-yl)-13-methylicosa-5,8,11,14-tetraenamide (21).

To a stirred solution of 20 (16 mg, 0.034 mmol) in anhydrous MeCN (0.7 mL) were added at ambient temperature: KF (40 mg, 0.68 mmol), K₂CO₃ (9 mg, 0.068 mmol), and 222 Kryptofix (3 mg). The reaction mixture was warmed to 80 °C and stirred at that temperature for 15 min. The reaction mixture was brought to ambient temperature, and the volatiles were evaporated under reduced pressure. The resulting crude residue was chromatographed on silica gel (80–100% EtOAc-hexanes) to afford 3 mg of the title compound as a colorless oil in 20% yield. ¹H NMR (500 MHz, CDCl₃) δ 5.58 (br d, J = 7.9 Hz, 1H, NH), 5.45–5.32 (m, 4H, 5,6,8,9-H), 5.32–5.19 (m, 4H, 11,12,14,15-H), 4.44 (dt, J = 25.3, 6.2 Hz, 2H, 20-H), 4.07–4.02 (m, 1H, 1′-H), 3.67 (dd, J = 11.0, 3.5 Hz, 1H, 2′-H), 3.53 (dd, J = 11.0, 6.2 Hz, 1H, 2′-H), 3.45–3.38 (m, 1H, 13-H), 2.91–2.78 (m, 4H, 7,10-H), 2.19 (t, J = 7.4 Hz, 2H, 2-H), 2.16–2.05 (m, 4H, 4,16-H), 1.73 (p, J = 7.4 Hz, 3H, 3-H), 1.71–1.61 (m, 2H, 19-H), 1.41–1.35 (m, 4H, 17,18-H), 1.17 (d, J = 6.9 Hz, 3H, 1′-Me), 1.02 (d, J = 6.8 Hz, 3H, 13-Me). ¹³C NMR (101 MHz, CDCl₃) δ 173.8, 135.1, 134.7, 129.2, 128.9, 128.5, 128.3, 127.8, 125.8, 84.32 (d, J = 164.1 Hz), 67.6, 48.0, 36.2, 30.6, 30.4 (d, J = 19.5 Hz), 29.5, 27.5, 26.7, 26.0, 25.7, 25.6, 25.0, 22.1, 17.2. Mass spectrum (ESI) m/z (relative intensity) 394 (100), 395 (23), 397 (22). Exact mass (ESI) calcd for C₂₄H₄₁FNO₂ [M + H]⁺, 394.3121; found, 394.3117. HPLC analysis (Agilent 1260 Infinity II, ZORBAX RX-SIL) showed 98% purity and 5.1 min retention time for the title compound.

rCB1 Radioligand Binding Assay.

The affinities (K_i) of the new compounds for rat CB1 were obtained by using membrane preparations from rat brain and [³H]CP-55,940 as the radioligand, as previously described.^18,19 Results from the competition assays were analyzed using nonlinear regression to determine the IC₅₀ values for the ligand. K_i values were calculated from the IC₅₀ (GraphPad Prism). Each experiment was performed in triplicate, and K_i values were determined from three independent experiments and are expressed as the mean of the three values.

Cell Membrane Isolation and hCB1 and hCB2 Radioligand Binding Assay.

HEK-293 cells stably expressing hCB1 or hCB2 receptors were disrupted by nitrogen cavitation, and the membrane fraction was extracted by ultracentrifugation.³⁰ Membrane protein concentrations were determined by using Bio-Rad Bradford protein assay kit.³¹ The binding assay was performed in a 96-well plate as modified by Lan et al.³² 25 μg membrane pellets were resuspended in 25 mM Tris base/5 mM MgCl₂/1 mM ethylenediaminetetraacetic acid containing 0.1% (W/V) BSA (TME-BSA) and added to each assay well. Radioligand ³H–CP-55,940 was diluted in TME-BSA buffer to yield the final assay concentration of 0.76 nM in a total reaction volume of 200 μL. Nonspecific binding was also performed in the presence of 5 μM unlabeled cold CP-55,940. The assay plate was incubated at 30 °C for 50 min with gentle agitation, and the samples were transferred to Unifilter GF/C filter plates and the unbound ligands were removed using Packard Filtermate 96 Cell Harvester (PerkinElmer Packard, Shelton, CT) by washing 3 times with ice-cold wash buffer (containing 50 mM Tris base, 5 mM MgCl₂, 0.5% BSA, pH 7.4). The bound radioactivity was quantified using a PerkinElmer TopCount Scintillation Counter. The IC₅₀ and K_i values were determined by nonlinear regression using GraphPad Prism 10.3.0 software, San Diego, CA. The K_i values were presented as means with 95% confidence intervals from at least 3 independent experiments performed in triplicate.

cAMP Accumulation and β-Arrestin 2 Recruitment Assays.

Compounds and Reagents.

CP-55,940 was purchased from Sigma-Aldrich (Product No: 83002–04–4, St. Louis, MO); all compounds were prepared as 10 mM stocks in DMSO. 4-(3-Butoxy-4-methoxybenzyl)imidazolidin-2-one (Ro-20–1724) (Product No: 29925–17–5), and forskolin (Product No: 66575–29–9) were purchased from Sigma-Aldrich. DMSO was purchased from Fisher Scientific (Product No: 67–68–5, Pittsburgh, PA). For cell culture, Opti-MEM, MEM, and DMEM/F12 were purchased from Thermo Fisher Scientific (Product Nos: 11058–021, 11095–080, and 11330–032, Waltham, MA), and heat-inactivated fetal bovine serum (HI-FBS) (Product No: F4135) was purchased from Sigma-Aldrich. The antibiotics puromycin, Geneticin, and hygromycin used in cell culture were purchased from Life Technologies (Product Nos: A11138–03, 10131–035, and 10687010, Carlsbad, CA).

Cell Lines and Cell Culture.

Chinese hamster ovary (CHO-K1, from ATCC) cells expressing hemeagglutinin (HA) tagged human cannabinoid 1 receptor or human cannabinoid 2 receptor (3xHA-hCB1-CHO, and 3xHA-hCB2-CHO) were maintained as previously described.³³ CHO cell line expressing modified β-arrestin 2 and hCB1 (hCB1-DRx-CHO), and U2OS cell line expressing modified β-arrestin 2 and hCB2 (hCB2-DRx-U2OS) were purchased from Eurofins DiscoveRx (Fremont, CA). hCB1-DRx-CHO cell line was maintained in DMEM/F12 media with 10% HI-FBS, 1% penicillin/streptomycin, 35 μL of Geneticin (5 μL/1 mL media), and 70 μL of Geneticin (10 μL/1 mL media). hCB2-DRx-CHO cell line was maintained in MEM media with 10% fetal bovine serum, 1% penicillin/streptomycin, 35 μL of Geneticin (5 μL/1 mL media), and 70 μL of Geneticin (10 μL/1 mL media). All cells were grown at 37°C (5% CO₂ and 95% relative humidity).

Inhibition of Forskolin-Stimulated cAMP Accumulation Assay.

Inhibition of forskolin (PLACE) stimulated the cAMP accumulation assay performed as previously described (CITE). Briefly, we used the Revvity HTRF cAMP Gs HiRange Detection Kit (Waltham, MA, USA). 3xHA-hCB1-CHO cells at 70% confluency were plated at 4,000 cells per well (5 μL) in a low-volume white 384-well plate in Opti-MEM (1% FBS, 1% pen. strep.) and were incubated at 37 °C for 3 h. Drugs were diluted in a vehicle composed of Opti-MEM (1% FBS, 1% pen. strep.), 1.8% DMSO, 40 μM forskolin, and 50 μM Ro-20–1724. The cells were treated with 5 μL of diluted drugs and were incubated at 37 °C for 30 min. Final concentrations per well: 1% DMSO, 20 μM forskolin, and 25 μM Ro-20–1724. Detection reagents were added (12 μL), and the plate was incubated at room temperature in the dark for 1 h. Fluorescence was read on Biotek Synergy Neo2 Hybrid Multimode reader (Agilent, Santa Clara, CA) at 665 and 620 nm. The FRET ratio was calculated from 665 nm/620 nm.

β-Arrestin 2 Recruitment Assay.

β-Arrestin 2 recruitment assay was performed using the DiscoveRx PathHunter β-arrestin assay Kit (DiscoveRx, Fremont, CA). hCB1-DRx-CHO cells were grown to 60–80% confluency and plated at 5,000 cells/well (20 μL) in Opti-MEM (1% FBS, 1% pen. strep.). Drugs were diluted in vehicle containing Opti-MEM (1% FBS, 1% pen. strep.), and 4.5% DMSO. Cells were treated with the diluted drugs (5 uL) and incubated at 37 °C for 90 min. Final concentration per well: 1% DMSO. Detection reagents were added (5 μL of each reagent), and the plate was incubated at room temperature for 1 h in the dark. Luminescence was measured using a BioTek Synergy Neo2 Hybrid Multimode reader (Agilent, Santa Clara, CA).

In Vivo Assays.

Animals.

Adult male CD1 mice (7–9 weeks old, 35–45 g; Charles River Laboratories, Wilmington, MA) were used in these experiments. Mice were housed in a temperature-controlled facility (73 ± 2 °F, 45% humidity) under a 12 h light/dark cycle with ad libitum access to food and water. All experimental procedures were approved by the Bloomington Institutional Animal Care and Use Committee (IACUC) at Indiana University and conducted in accordance with the guidelines of the International Association for the Study of Pain (IASP).

Drugs and Chemicals.

The CB1-selective antagonist/inverse agonist AM251 (1-(2,4-dichlorophenyl)-5-(4-iodophenyl)-4-methyl-N-1-piperidinyl-1H-pyrazole-3-carboxamide) was purchased from Cayman Chemical Company (Ann Arbor, MI). Sterile saline was purchased from Aqualite System (Hospira, Lake Forest, IL). Kolliphor was obtained from Sigma-Aldrich; ethanol from Thomas Scientific; and dimethyl sulfoxide (DMSO) from Sigma-Aldrich. AM12814 was prepared in a 10% DMSO vehicle with the same 1:1:18 ratio for all experimental procedures. AM251 was dissolved in a vehicle composed of 10% DMSO, Kolliphor, ethanol, and saline in a 1:1:18 ratio for the assessment of catalepsy (ring test), body temperature, and antinociception (Triad test). In the CFA studies, AM251 was prepared in a 20% DMSO vehicle with the same 1:1:18 ratio of DMSO, Kolliphor, ethanol, and saline. Compounds were administered intraperitoneally (ip) at a volume of 10 mL/kg.

Experimental Design.

We examined the effects of AM12814 (0.3, 1, or 3 mg/kg, ip) or vehicle on cardinal signs of CB1 activation in a triad of tests (assessing catalepsy, hypothermia, and tail-flick antinociception) with all three tests conducted sequentially using the same mice. To assess pharmacological specificity and ascertain the role of CB1 receptors in observed cannabimimetic effects, separate cohorts of mice received pretreatments with either AM251 (5 mg/kg, ip) or vehicle, followed 20 min later by a subsequent injection of AM12814 (3 mg/kg, ip) or vehicle. The order of testing in the triad was: ring immobility, rectal temperature, and tail-flick antinociception. Responses in the triad were assessed before and at 10 min, 0.5, 1, 2, 3, and 4 h postinjection (n = 5–6 per group).

To assess the antinociceptive effects of AM12814, we used a model of CFA-induced inflammatory hypersensitivity. To assess dose response, separate groups of mice received a single i.p. injection of AM12814 (0.3, 1, or 3 mg/kg, i.p.) or vehicle. In studies assessing pharmacological specificity, antinociceptive efficacy was assessed in mice receiving pretreatments with either AM251 (5 mg/kg, i.p.) or vehicle, followed 20 min later by a subsequent injection of AM12814 (1 mg/kg, i.p.) or vehicle. Mechanical paw withdrawal thresholds were measured before and at 20 min following the first injection (corresponding to AM251 or Vehicle pretreatment). Then, after the second ip injection of AM12814 or vehicle, paw withdrawal thresholds were measured again at 10 min, 0.5, 1, 2, 3, 4, and 24 h following the last injection (n = 6–15 per group).

Ring Test for Catalepsy.

Catalepsy was assessed using the ring test as described by Pertwee.³⁴ Mice were placed on a horizontal metal ring (5.7 cm in diameter) suspended 7.3 cm above the surface, and the duration of immobility (in seconds) was recorded over a 5 min observation period.

Measurement of Rectal Temperature.

Core body temperature was measured using a rectal thermometer (Physitemp Instruments, Inc., Clifton, NJ) equipped with a mouse rectal probe, following procedures described by Deng et al.³⁵

Tail-Flick Assay for Thermal Nociception.

Thermal nociception was evaluated using the hot water tail-immersion test (53–54 °C) as described by Slivicki et al.³⁶ The distal 2 cm of the tail was submerged in the water bath, and latency to tail withdrawal (“flick” response) was recorded. Baseline responses were measured three times at 15 min intervals prior to drug administration. A cutoff time of 15 s was applied to prevent tissue damage.

Assessment of Mechanical Hypersensitivity.

Paw withdrawal thresholds in response to mechanical stimulation were measured using an electronic von Frey anesthesiometer (IITC Model Alemo 2390–5, Woodland Hills, CA) as previously described.³⁵ Mice were placed on an elevated metal mesh platform and habituated under individual inverted plastic cages for at least 1 h before testing. A semiflexible tip connected to the anesthesiometer applied force to the midplantar region of the hind paw. Stimulation ceased upon paw withdrawal. Each paw was tested in duplicate, and responses were averaged for each animal. Mechanical paw withdrawal thresholds were measured before and after injection of CFA, and at various time points following pharmacological manipulations.

Complete Freund′s Adjuvant (CFA)-Induced Inflammatory Pain Model.

Inflammatory pain was induced via a single intraplantar (ipl) injection (20 μL) of complete Freund’s adjuvant (CFA; Sigma–Aldrich, St. Louis, MO) diluted 1:1 with saline into the right hind paw. Mechanical paw withdrawal thresholds were assessed before CFA injection (baseline-BL) and 72 h postinjection (predrug BL).³⁷ Responses to mechanical stimulation were also evaluated at multiple time points following pharmacological manipulations. AM12814 was administered systemically (intraperitoneally) ipsilaterally or contralaterally to the CFA injection.

Statistical Analysis.

Data were analyzed by Two-way repeated measures ANOVA followed by Tukey’s multiple comparison post hoc test (to ascertain group differences between all experimental groups), Dunnett’s multiple comparison post hoc test (to assess group differences vs control in the case of significant interactions), and Sidak’s post hoc test (in the case of pairwise comparisons across multiple time points). P < 0.05 was considered significant. All statistical analyses were performed using GraphPad Prism version 10.2.3 (GraphPad Software, Inc., 225 Franklin Street, Fl. Boston, MA 02110, USA).

Ethics Statement.

All animal experiments were approved by the Institutional Animal Care and Use Committee at Indiana University Bloomington (Protocol 22–029) and followed the guidelines of the International Association for the Study of Pain. Indiana University Bloomington is governed by the Office of Laboratory Animal Welfare (OLAW) Assurance Number D16–00587. The experimenter was blinded to the experimental conditions in all studies.

Molecular Docking.

ICM-Pro molecular modeling software (Molsoft LLC, version 3.9–4a) was used for structural model preparation, docking, and water modeling. Active-state structures of AMG315-bound CB1 receptor (PDBID: 8GHV), AM12033-bound CB2 receptor in complex with Gi protein (PDBID: 6KPF), CP-55,940-bound CB1 receptor (PDBID: 7V3Z), and CP-55,940-bound CB2 receptor (PDBID: 8GUR) were downloaded from the RCSB PDB server. The structures were converted to ICM objects before docking calculations. The ICM conversion algorithm includes addition of missing side chains, addition and optimization of hydrogen atoms, and optimization of side chain of polar residues (His, Pro, Asn, Gln, Cys) in the protein structure. Structures of the AMG315-bound CB1 receptor and AM12033-bound CB2 receptor were used for docking of compounds AEA, AMG315, and AM12814. Compound CP-55,940 was redocked into the available CP-55,940-bound structures of CB1 and CB2. The rigid docking maps were calculated in a rectangular box containing the orthosteric binding pocket. Three docking runs were performed with the thoroughness set to 5. Physics-based docking score and data-driven Radial and Topological Convolutional Neural Network Score (RTCNN Score) were calculated for the top 10 docking poses. The top-scoring predicted docking poses were used in the analysis. “Flood” procedure by Molsoft as implemented in ICM-Pro was used to generate a stack of low-energy water configurations inside of a 12 Å × 12 Å × 12 Å box around the predicted binding pose of the trifluoromethyl group of AM12814 compound and methyl group AMG315 compound in CB1 and CB2 binding pockets. 50 water configurations were generated for each protein–ligand pair, and the configuration with the lowest energy was used in the analysis.

Supplementary Material

Supporting Information

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jmedchem.5c02030.

Experimental procedures, spectroscopic, analytical, and physical data for all compounds; methods for radioligand binding, cAMP, and β-arrestin 2 assays, as well as cannabinoid triad and CFA-induced mechanical hypersensitivity assays and molecular modeling; reproductions of ¹H and ¹³C NMR spectra, and HPLC traces of key compounds; representative curves of competition binding assays; and additional references (PDF)

Molecular formula strings (CSV)

PDB coordinates (ZIP)

ABBREVIATIONS USED

2-AG

2-arachidonoylglycerol

AEA

anandamide

CB1

cannabinoid receptor 1

CB2

cannabinoid receptor 2

CNS

central nervous system

eCBs

endocannabinoids

FAAH

fatty acid amide hydrolase

HEK293

human embryonic kidney cell line

HPLC

High-Performance Liquid Chromatography

NMR

nuclear magnetic resonance

PMSF

phenylmethanesulfonylfluoride

SAR

structure–activity relationship