6,5-Fused Ring, C2-Salvinorin Ester, Dual Kappa and Mu Opioid Receptor Agonists as Analgesics Devoid of Anxiogenic Effects

Abstract

Current common analgesics are mediated through the mu or kappa opioid receptor agonism. Unfortunately, selective mu or kappa receptor agonists often cause harmful side effects. However, ligands exhibiting dual agonism to the opioid receptors, such as to mu and kappa, or to mu and delta, have been suggested to temper undesirable adverse effects while retaining analgesic activity. Herein we report an introduction of various 6,5-fused rings to C2 of the salvinorin scaffold via an ester linker. In vitro studies showed that many of these compounds have dual agonism on kappa and mu opioid receptors. In vivo studies on the lead dual kappa and mu opioid receptor agonist demonstrated supraspinal thermal analgesic activity while avoiding anxiogenic effects in male mice, thus providing further strong evidence in support of the therapeutic advantages of dual opioid receptor agonists over selective opioid receptor agonists.

Graphical Abstract

Salvinorin-based Analgesics: Compounds with dual agonism to opioid receptor subtypes have been suggested to reduce adverse effects while retaining analgesic activity. Herein we report the introduction of various 6,5-fused rings to C2 of the salvinorin scaffold via an ester linker. The lead dual kappa and mu opioid receptor agonist demonstrated supraspinal thermal analgesic activity while avoiding anxiogenic effects in male mice.

Introduction

The Centers for Disease Control (CDC) has estimated that one in five Americans suffers from chronic pain, and one in six American adults suffers from a mental health disorder.^[1,2] Studies have indicated that these disorders share a neural network and thus have a bidirectional relationship.^[1] Therefore, discovery and development of new analgesics that exert efficacy for pain and mental health disorders are vital. Historically, selective agonists of classical opioid receptors, including kappa (κ, KOR), mu (μ, MOR), and delta opioid receptor (δ, DOR), are commonly used and clinically efficacious against a number of anxiety, depression, and chronic pain states.^[1] Unfortunately, these compounds often result in many serious and harmful adverse effects, such as addiction, dysphoria, reduced bowel motility, tolerance, convulsions, and overdose.^[3,4] The opioid epidemic has thus sparked an urgent need to develop effective pain management therapeutics devoid of such adverse effects.^[3] Importantly, many studies have suggested that compounds exhibiting dual agonism to the opioid receptors, such as to MOR and KOR,^[5–10] or to MOR and DOR,^[11] demonstrate the ability to temper undesirable adverse effects while retaining analgesic activity.

Salvinorin A (Figure 1) is the main active ingredient in the hallucinogenic plant Salvia divinorum. It is one of the most potent, naturally occurring opioid agonists with high selectivity and affinity for KOR.^[13] It has potential therapeutic benefits for various central nervous system (CNS) disorders. Salvinorin A induces analgesic and anti-inflammatory effects, but uniquely does not present affinity to other receptors associated with perceptual alterations, such as dopamine, serotonin, or glutamate receptors.^[13] Due to its potent hallucinatory effects, salvinorin A has never advanced to clinical trials,^[14] but has been used as an important prototype for the development of related drug candidates, especially salvinorin analogs that exert therapeutic effects while being devoid of the hallucinatory side effects of KOR agonists.^[15–17]

Figure 1.

Structures of salvinorin A, herkinorin, compound 3k,^[12] PR-38, and salvindolin.

Previous studies have shown salvinorin analogs with aromatic or heteroaromatic moieties at C2 to exhibit significant changes in their pharmacological profiles, associated with a change in the affinity from KOR to MOR.^[5,6,8,9] Four representative molecules of this change are herkinorin, compound 3k, PR-38, and salvindolin (Figure 1), all of which display dual affinity to KOR and MOR.^{[5,6,8,12,18]} Herkinorin, discovered by the Prisinzano group, was the very first salvinorin compound that displays dual agonism on KOR and MOR.^[8] Compared to salvinorin A, herkinorin has a 47-fold lower affinity for KOR (K_i=90 nM vs. K_i=1.9 nM), and at least a 83-fold higher affinity for MOR (K_i=12 nM vs. K_i>1000 nM). Herkinorin also exerts micro-molar binding affinity to DOR that salvinorin A does not (K_i=1170 nM).^[8] Herkinorin was initially suggested to not promote the recruitment of β-arrestin-2 to the intracellular domain of MOR nor induce receptor internalization,^[18] but arrestin recruitment was later evidenced utilizing a sensitive BRET assay.^[19] The Prisinzano group later discovered other analogues of herkinorin with differential β-arrestin-2 interactions.^[12] Compound 3k displays dual agonism on KOR and MOR with a decreased affinity at KOR greater than 50-fold compared to herkinorin (K_i=5490 nM vs. K_i=90 nM) and a reduced affinity at MOR approximately 10-fold compared to herkinorin (K_i=180 nM vs. K_i=12 nM). No studies were done on compound 3k’s ability to promote the recruitment of β-arrestin-2 nor induce receptor internalization.^[12] PR-38 and salvindolin were discovered by the Zjawiony group and displayed dual agonism on KOR and MOR.^[5,6] PR-38 had roughly a five-fold preference towards KOR over MOR, while salvindolin displayed a 100-fold preference towards MOR over KOR.

To study the mechanism for the switch to dual affinity and the effects of novel moieties at C2, we introduced a series of 6,5-fused rings to C2 via an ester linker (Figure 2), including the indene (which is non-nitrogenous, compounds 1 and 2), the indane (which is non-aromatic, compound 3), and many bicyclic heterocycles (other compounds). We planned to use these synthetic salvinorin analogs to study the structure-activity relationships between variations on the 6,5-fused rings and the resulting affinities/efficacies at KOR, MOR, and DOR.

Figure 2.

Introduction of various 6,5-fused rings to C2 of the salvinorin scaffold via an ester linker: structures of compounds 1–14.

Results and Discussion

Synthesis

Compounds 1–10 were synthesized via an ester coupling reaction between commercially available salvinorin B and the corresponding carboxylic acids, which were also commercially available, using EDC and DMAP,^[20] in various yields (Scheme 1A).

Scheme 1.

Synthesis of compounds 1–10 (A), 11–12 (B), and 13–14 (C).

The corresponding carboxylic acids for compounds 11–14 were not commercially available, so we synthesized them in our lab. A base hydrolysis of the ethyl ester in commercially available 15 using lithium hydroxide provided carboxylic acid 16 in a 57% yield (Scheme 1B). An ester coupling reaction between salvinorin B and 16 gave compound 11 in 71% yield. A literature 3-step procedure^[21] to synthesize carboxylic acid 20 from commercially available 17 was followed, and the overall yield of the three steps was 22% (Scheme 1B). An ester coupling reaction between salvinorin B and 20 gave compound 12 in 49% yield.

A direct ester coupling reaction between salvinorin B and commercially available carboxylic acid 21 did not yield 13. Therefore, the free indoline amino group of 21 was then first protected using the fluorenylmethyloxycarbonyl (Fmoc) group to provide 22 in a 90% yield (Scheme 1C). An ester coupling reaction between salvinorin B and 22 gave the Fmoc-protected ester 23 in 89% yield. A subsequent Fmoc deprotection using piperidine in DMF^[22] gave compound 13 in 84% yield. In a similar fashion, an ester coupling reaction between salvinorin B and commercially available Fmoc-protected carboxylic acid 24, followed by Fmoc deprotection, gave compound 14 in a 53% yield over two steps (Scheme 1C).

In vitro evaluation

Compounds 1–14 were subjected to in vitro evaluation to determine their affinity, efficacy, and functionality on classical opioid receptors. A competitive binding assay^[23] was utilized to screen receptor binding and determine the K_D values of each compound for KOR, MOR, and DOR. These K_D values are shown in Table 1. A GTPγS assay was also completed to determine the functionality and potency of selected compounds of interest. All of these compounds were shown to be agonists of the opioid receptors, and their EC₅₀ values are shown in Table 1. Several compounds were found to be partial agonists, and their respective E_max values are shown in Table S1 in the Supporting Information.

Table 1.

K_D and GTPγS EC₅₀ values of compounds 1–14 towards KOR, MOR, and DOR.

Compound	C2 substituent	KD [nM]	EC50 [nM]	KD [nM]	EC50 [nM]	KD [nM]	EC50 [nM]
		(KOR)	(KOR)	(MOR)	(MOR)	(DOR)	(DOR)
Naloxone	N/A	9.98 (6.89–14.5)	NT	4.39 (3.14–6.11)	NT	60.0 (47.9–75.1)	NT
U69,593	N/A	190 (105–345)	64.0 (39.0–105)[c]	NT	NT	NT	NT
Cebranopadol	N/A	448 (246–815)	NT	3.77 (2.54–5.60)	NT	35.3 (24.9–50.0)	NT
Nalbuphine	N/A	14.8 (11.0–20.0)	NT	4.02 (2.82–5.72)	NT	353 (233–535)	NT
Herkinorin	Phenyl	90 ± 2[a]	NT[a]	12 ± 1[a]	NT[a]	1170 ± 60[a]	NT[a]
Salvindolin	1H-Indole-2-	1010 ± 170[a]	ND[a]	10.5 ± 1.8[a]	188 ± 52[a]	ND[a]	ND[a]
1	1H-Indene-3-	62.3 (39.6–98.0)	1570 (548–4500)[c]	531(184–1540)	> 10,000[b]	ND	ND
2	1H-Indene-2-	49.7 (23.5–105)	853 (456–1600)[d]	538 (335–864)	49.5 (38.8–63.0)[c]	714 (385–1330)	2350 (964–5710)[c]
3	1H-Indane-2-	1270 (785–2040)	3000 (409–22000)[d]	3020 (931–9760)	> 10,000[b]	ND	ND
4	Benzo[d]oxazole-2-	ND	ND	ND	ND	ND	ND
5	1H-indazole-3-	ND	ND	208 (91.5–472)	> 10,000[b]	ND	ND
6	1H-benzimidazole-2-	ND	ND	83.2 (57.9–119)	> 10,000[b]	ND	ND
7	4-Fluoro-1H-indole-2-	57.3 (30.8–107)	877 (564–1360)[d]	13.1 (8.1–21.3)	211 (74.5–598)[d]	ND	ND
8	5-Fluoro-1H-indole-2-	97.8 (74.0–129)	1110 (441–2790)[c]	ND	ND	ND	ND
9	6-Fluoro-1H-indole-2-	17.0 (11.4–25.5)	419 (240–729)[c]	39.7 (15.9–99.0)	973 (317–2990)[d]	ND	ND
10	7-Fluoro-1H-indole-2-	53.8 (36.0–80.6)	36.5 (13.6–98.2)[d]	58.4 (26.2–130)	132 (46.3–378)[c]	ND	ND
11	Benzo[d]thiazole-2-	91.7 (61.9–136)	294 (179–482)[d]	ND	ND	ND	ND
12	Indolizine-2-	18.4 (14.6–23.2)	1560 (213–11500)[c]	1390 (785–2470)	200 (124–323)[c]	480 (209–1100)	742 (215–2560)[c]
13	(R)-1H-indoline-2-	130 (66.0–255)	2260 (1360–3760)[c]	9.04 (6.37–12.8)	840 (521–1350)[d]	ND	ND
14	(S)-1H-indoline-2-	32.1 (16.8–61.3)	361 (216–602)[d]	77.9 (24.9–244)	> 10,000[b]	ND	ND

N/A: Not applicable. NT: Not tested. ND: Not determined; initial screening resulted in an out-of-range value and therefore was not tested further. Italics: 95% confidence intervals.

^[a]Literature values (herkinorin,^[8] salvindolin)^[6],

^[b]EC₅₀ value assays were carried out and the values were above the upper limit of the range (10.0 μM).

^[c]Full agonist.

^[d]Partial agonist.

It has been hypothesized from the studies of herkinorin, compound 3k, PR-38, and salvindolin that: (1) the introduction of an aromatic moiety to C2 of the salvinorin scaffold substantially changed the pharmacological profile of the compounds from mono affinity (KOR) to dual affinity (KOR and MOR),^{[5,6,8,12,18]} and (2) the nitrogen on the indole ring of salvindolin was responsible for the affinity change to MOR by interacting with nearby residues (coming from a molecular modeling study).^[6] However, our results showed that (1) compounds 13 and 14 do not have an aromatic moiety to C2 of the salvinorin scaffold and yet exhibited dual activity on KOR and MOR, and (2) compounds 1 and 2 lacking the nitrogen in the 6,5-fused ring, but still exhibited dual affinity to KOR and MOR. Our results stand in contrast to the two aforementioned hypotheses and rather suggest that the mechanism for the switch from mono affinity (KOR) to dual affinity (KOR and MOR) is more complicated than previously thought.

In fact, it is not easy to make a general trend of structure-activity relationships that goes through all the variations of the 6,5-fused rings for each of the opioid receptors. We think it is more suitable and appropriate to put them into groups and make the trends of structure-activity relationships within each group (Figure 3). With non-nitrogenous 6,5-fused rings, breaking the trigonal planar geometry of carbon 2 on the 6,5-fused ring, such as in the indane 3, diminished the affinity towards KOR (Figure 3A). The indene-2- 2 exhibited triple affinity towards KOR, MOR, and DOR, while the indene-3- 1 exhibited dual affinity towards KOR and MOR.

Figure 3.

Structure-activity relationship of 6,5-fused rings on the binding affinity towards opioid receptors: (A) non-nitrogenous 6,5-fused rings; (B) 6,5-fused rings that have an additional heteroatom besides nitrogen; (C) fluorine substitutions on the indole ring; (D) breaking the trigonal planar geometry of carbon 2 on the indole ring.

6,5-Fused rings that have an additional heteroatom besides nitrogen, such as 4, 6, and 11, displayed significant difference in opioid receptor affinity and selectivity from one another (Figure 3B). While the benzoxazole 4 exhibited no affinity to any of the opioid receptors, the benzimidazole 6 exhibited selective affinity towards MOR and the benzothiazole 11 exhibited selective affinity towards KOR. Meanwhile, the 6,5-fused ring with a nitrogen at the bridge location, indolizine 12, exhibited triple affinity towards KOR, MOR, and DOR.

A fluorine substitution at the positions 4, 6, and 7 on the indole, such as in the cases of compounds 7, 9, and 10, significantly enhanced the binding affinity to both KOR and MOR (Figure 3C). Meanwhile, a fluorine substitution at the position 5 on the indole, such as in the case of compound 8, led to a loss of binding affinity to MOR.

The geometry of carbon 2 on the indole had a significant impact on the binding affinity of the compounds to KOR and MOR. With the trigonal planar geometry of carbon 2 on the indole in the case of salvindolin, MOR affinity was 100-fold more favorable. Meanwhile, breaking the trigonal planar geometry of carbon 2 on the indole, in the case of compounds 13 and 14, brought the KOR and MOR affinities to a more even ground (Figure 3D). The (R) indoline isomer 13 had a 14-fold preference towards MOR, while the (S) indoline isomer 14 had ~2-fold preference towards KOR.

In vivo evaluation

As mentioned earlier, compounds exhibiting dual agonism to the opioid receptors, such as to MOR and KOR,^[5,6] or to MOR and DOR,^[11] have been suggested to temper undesirable adverse effects while retaining analgesic activity; therefore, we decided to focus our in vivo studies on the compounds that possessed dual agonism on the opioid receptors. Some of these compounds had high binding affinity, despite poor potency, such as compound 9 at KOR and MOR, while others had exerted high binding affinity and commensurately high potency, such as compound 10 at KOR and MOR. A higher binding affinity to a receptor does not always translate to a better potency as potency also depends on the intrinsic activity of the drug-receptor complex to produce a functional response. Therefore, we selected our lead compounds (2 and 10) for in vivo studies based on both the activities and the potencies of the compounds at the opioid receptors. Both 2 and 10 demonstrated dual agonism on KOR and MOR. Interestingly, compound 2 has an 11-fold preference towards KOR in binding affinity, but exerted 17-fold higher in potency at MOR. Compound 2 also demonstrated some weak agonism on DOR. Meanwhile, compound 10 has equal binding affinity towards KOR and MOR, but exerted ~4-fold higher in potency at KOR. In addition to selecting compounds 2 and 10 for in vivo studies, we used salvinorin A and herkinorin (synthesized by following the previously published procedure)^[8] as respective positive controls for selective KOR and dual KOR/MOR agonists.

The anti-nociceptive effects of the two lead compounds, 2 and 10, against acute pain were assessed in male C57BL/6NHsd mice (Figure 4). In the hot plate assay, which examines supraspinally-mediated nociception, pretreatment with the current compounds significantly altered paw lick/flutter latencies [F(5,42)=34.34, p<0.05] (Figure 4A). Mice treated with morphine (p<0.0001) or compound 2 (p=0.01) showed a significant increase in the latency to paw response compared to vehicle-treated mice. Neither salivinorin A, herkinorin, nor compound 10 were observed to significantly influence the latency to paw response compared to vehicle-treated mice.

Figure 4. Effects of compounds 2 and 10 in the hot plate and tail-flick thermal anti-nociception assays.

Average latency to hind paw lick or flutter in the hot plate assay (A) or latency to withdraw tail in the tail-flick assay (B) following dosing with vehicle (n=10 for both assays), morphine (5 mg/kg, n=5 for hot plate, n=10 for tail-flick), salvinorin A (2.5 mg/kg, n=5 for both assays), herkinorin (5 mg/kg, n=10 for both assays), compound 2 (2 mg/kg, n=8 for hot plate, n=7 for tail-flick), or compound 10 (2.5 mg/kg, n=10 for both assays). All data were analyzed via one-way ANOVA, * indicates significant difference from vehicle, p<0.05.

In the spinally-mediated thermal tail-flick nociception assay, only pretreatment with morphine significantly increased the latency to tail withdrawal [F(5,46)=30.84, p<0.0001] (Figure 4B). Neither salvinorin A, herkinorin, nor compounds 2 or 10 demonstrated significant antinociception.

When anxiety-like behavior was assessed in an elevated plus maze, two compounds significantly increased anti-anxiety-like behavior [F(4,45)=2.71, p<0.05] (Figure 5). Compared to vehicle treatment, both salvinorin A (p=0.02) and compound 2 (p=0.04) significantly increased the amount of time spent on the open arms of the maze, indicating anxiolysis (Figure 5A). Neither herkinorin nor compound 10 significantly influenced anxiety-like behavior. Albeit, herkinorin produced a notable increase in open arm time, but this was not statistically significant. Changes in the anxiety-like response were not due to gross motor deficits given that neither salvinorin A, nor compound 2, altered the distance (Figure 5B) or velocity (Figure 5C) traveled in the maze. Interestingly, compound 10 exerted a decrease in motor behavior as indicated by a reduction in the total distance traveled [F(4,45)=3.89, p=0.01].

Figure 5. Effects of compounds 2and 10in the elevated plus maze anxiety-like behavior assay.

The average time spent in the open arms (A), distance traveled (B), and velocity (C) in the elevated plus maze following dosing with vehicle (n=10), salvinorin A (2.5 mg/kg, n=10), herkinorin (5 mg/kg, n=10), compound 2 (5 mg/kg, n=10), or compound 10 (5 mg/kg, n=10). All data were analyzed via one-way ANOVA, * indicates significant difference from vehicle, p<0.05.

Together, these results indicate compound 2 to exert significant analgesia in the hot plate test but not antinociception in the tail-flick test, suggesting that this effect was mediated supraspinally. Importantly, both salvinorin A and compound 2 exerted anxiolytic effects, but only compound 2 exerted concomitant analgesia in the hotplate test. While the analgesic efficacy of 2 did not match the levels produced by morphine, the dose of morphine was notably higher (5 mg/kg) than that of 2 (2 mg/kg). As such, these data provide an intriguing indication for the potential applications of compound 2 in the treatment of supraspinally-mediated pain and anxiety.

One limitation of our in vivo studies is the exclusion of assessment in female mice. Sex-specific differences in nociceptive assays have been reported, and it will be important to include sex as a biological variable in follow-up studies related to salvinorin-based compounds. Another consideration is the potential for bias towards arrestin recruitment which was not assessed in the current report, but could explain the differences between K_D and GTPγS EC₅₀ results. Lastly, future work should rule out the potential psychotomimetic effects that are known to occur in selective KOR agonists.

Conclusion

In conclusion, we have reported an introduction of various 6,5-fused rings to C2 of the salvinorin scaffold via an ester linker. The compounds were subjected to in vitro evaluation to determine their affinity, efficacy, and functionality at the classical opioid receptors. The results showed that many of these compounds have dual agonism on KOR and MOR, and a couple of these compounds have some weak agonism on DOR. The compounds were categorized into groups, and the trends of structure-activity relationships within each group were identified. The trends stand in contrast to the two current hypotheses on the structure-activity relationships of salvinorin-based, dual KOR and MOR agonists, and rather suggest that the mechanism for the switch from mono affinity (KOR) to dual affinity (KOR and MOR) is more complicated than previously thought. As selective mu or kappa receptor agonists often cause harmful side effects such as addiction, dysphoria, reduced bowel motility, tolerance, convulsions, and overdose, dual agonists to the opioid receptors, such as to MOR and KOR, or to MOR and DOR, have been suggested in previous studies to temper undesirable adverse effects while retaining analgesic activity. To this end, the dual agonism on KOR and MOR of compound 2 may contribute to its unique supraspinal analgesic and anxiolytic activities. Together, these findings provide further strong evidence in support of the therapeutic advantages of dual opioid receptor agonists over selective opioid receptor agonists.

Experimental Section

Synthesis of compounds 1–14

Chemicals and Instruments

The starting material carboxylic acids for the syntheses of compounds 1–10 were commercially available and purchased from Sigma Aldrich (St. Louis, MO), Combi-Blocks (San Diego, CA), and Fisher Scientific (Hampton, NH). The carboxylic acid starting materials for the syntheses of compounds 11–14 were synthesized in our lab. Salvinorin B was purchased from Apple Pharms Ingredients Inc. (Bakersville, NC). All other chemicals were purchased from Sigma-Aldrich or Fisher Scientific and used as received unless specified. All syntheses were conducted with anhydrous conditions under an atmosphere of argon, using flame-dried glassware and employing standard techniques for handling air-sensitive materials unless otherwise noted. All solvents were distilled and stored under an argon or nitrogen atmosphere before use. ¹H NMR and ¹³C NMR spectra were recorded on a Bruker-400 and or a Bruker-500 spectrometer using CDCl₃, DMSO-d₆, acetone-d₆, or CD₃CN as the solvent. Chemical shifts (δ) were recorded in parts per million and referenced to CDCl₃ (7.24 ppm for ¹H NMR and 77.23 ppm for ¹³C NMR), DMSO-d₆ (2.50 ppm for ¹H NMR and 39.51 ppm for ¹³C NMR), acetone-d₆ (2.05 ppm for ¹H NMR and 29.92 and 206.68 ppm for ¹³C NMR), CD₃CN (1.94 for ¹H NMR and 1.39 and 118.69 for ¹³C NMR). ¹⁹F NMR spectra were recorded on a Bruker-400 spectrometer. Coupling constants (J) are in Hz. The following abbreviations were used to designate the multiplicities: s=singlet, d=doublet, t=triplet, q=quartet, quint=quintet, m= multiplet, br=broad. Melting points were measured using an OptiMelt automated melting point system. LC–MS were measured using an ACQUITY-Waters micromass (ESCi) system. High-resolution mass spectra (HRMS) were measured using a Waters Synapt XS HRMS. Compounds 1–14 were purified via column chromatography (1:2 ethyl acetate:hexanes) and further by HPLC if necessary (7.8× 30 mm, 7 μm, C18, gradient water in acetonitrile, flow rate 2 mL/min) until their purities were higher than 95% before being evaluated in in vitro and in vivo assays; purities were measured using a Waters 2695 analytical HPLC system.

General ester coupling procedure

To a solution of salvinorin B (1 eq, 0.0768 mmol, 30 mg), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC, 4 eq, 0.307 mmol, 47 mg), and dimethylaminopyridine (DMAP. 2 eq, 0.154 mmol 19 mg) in CH₂Cl₂ (4 mL) were added the corresponding starting material carboxylic acid (2 eq, 0.154 mmol) and stirred under argon atmosphere for 8–16 hours at 40°C in a single-necked round bottom flask. After completion of the reaction, indicated by thin layer chromatography (TLC), the mixture was cooled to room temperature and washed with water (3×15 mL), 1 N HCl (15 mL), and brine (15 mL). The ester product was then purified via column chromatography (1:2 ethyl acetate:hexanes).

General Fmoc deprotection procedure^[22]

The Fmoc-protected compound (0.036 mmol) was dissolved in dry dimethylformamide (DMF, 0.3 mL), followed by the addition of piperidine (0.007 mmol). The mixture was stirred under inert atmosphere (N₂) for 30 minutes at room temperature, and upon completion of the reaction, indicated by thin layer chromatography (TLC), was added 5 mL of water and extracted 3 times with ethyl acetate. The combined organic layers were washed with brine and concentrated in vacuo. The product was then purified via column chromatography (1:2, ethyl acetate:hexanes).

Methyl-(2S,4aR,6aR,7R,9S,10aS,10bR)-9-((1H-indene-3-carbonyl) oxy)-2-(furan-3-yl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromene-7-carboxylate, 1

General ester coupling procedure was followed to afford 1. Yield: 49%. Mp 208–210°C. ¹H NMR (500 MHz, CDCl₃) δ 8.03 (d, J=7.6 Hz, 1H), 7.62 (s, 1H), 7.51 (d, J=7.4 Hz, 1H), 7.46−7.36 (m, 3H), 7.31 (d, J=7.4 Hz, 1H), 6.42 (s, 1H), 5.55 (dd, J=11.6, 5.0 Hz, 1H), 5.43 (t, J=9.9 Hz, 1H), 3.77 (s, 3H), 3.59 (s, 2H), 2.92−2.82 (m, 1H), 2.63−2.44 (m, 3H), 2.28 (s, 1H), 2.17 (dd, J=42.2, 13.2 Hz, 2H), 1.85 (d, J=12.0 Hz, 1H), 1.65 (td, J=21.3, 19.6, 10.2 Hz, 4H), 1.20 (s, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 201.83, 171.64, 171.15, 162.71, 146.03, 143.72, 143.28, 140.37, 139.47, 135.27, 126.74, 125.76, 125.23, 123.84, 122.42, 108.43, 75.03, 72.08, 64.16, 53.67, 52.03, 51.44, 43.44, 42.22, 38.65, 38.23, 35.52, 30.99, 18.19, 16.52, 15.24. HRMS m/z calcd for C₃₁H₃₁O₈ [M−H]⁻ 531.2019, found: 531.1992. Purity after HPLC: 95.30%.

Methyl-(2S,4aR,6aR,7R,9S,10aS,10bR)-9-((1H-indene-2-carbonyl) oxy)-2-(furan-3-yl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromene-7-carboxylate, 2

General ester coupling procedure was followed to afford 2. Yield: 66%. Mp 209–210°C. ¹H NMR (500 MHz, CDCl₃) δ 7.85 (s, 1H), 7.53 (td, J=9.2, 7.6, 4.5 Hz, 2H), 7.43−7.33 (m, 4H), 6.40−6.36 (m, 1H), 5.52 (dd, J=11.7, 5.1 Hz, 1H), 5.33 (d, J=9.6 Hz, 1H), 3.74 (s, 3H), 3.73 (s, 2H), 2.86−2.79 (m, 1H), 2.55 (dd, J=13.5, 5.2 Hz, 1H), 2.48−2.41 (m, 2H), 2.25 (s, 1H), 2.20−2.16 (m, 1H), 2.10 (dd, J=11.5, 2.7 Hz, 1H), 1.85−1.80 (m, 1H), 1.68−1.57 (m, 4H), 1.47 (s, 3H), 1.17 (s, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 202.28, 171.77, 171.29, 163.81, 145.09, 143.85, 142.94, 142.62, 139.59, 135.86, 128.09, 127.12, 125.35, 124.45, 123.78, 108.55, 75.12, 72.21, 64.26, 53.81, 52.16, 51.58, 43.57, 42.34, 38.44, 38.37, 35.64, 31.15, 18.32, 16.64, 15.38. HRMS m/z calcd for C₃₁H₃₂O₈Cs [M+Cs]⁺ 665.1152, Found 665.1160. Purity after HPLC: 95.17%.

Methyl-(2S,4aR,6aR,7R,9S,10aS,10bR)-9-((2,3-dihydro-1H-indene-2-carbonyl)oxy)-2-(furan-3-yl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromene-7-carboxylate, 3

General ester coupling procedure was followed to afford 3. Yield: 85%. Mp 217–219°C. ¹H NMR (500 MHz, Acetone-d₆) δ 7.60 (s, 1H), 7.53 (s, 1H), 7.21 (dd, J=11.1, 4.3 Hz, 2H), 7.17−7.12 (m, 2H), 6.53 (s, 1H), 5.58 (dd, J=11.7, 5.2 Hz, 1H), 5.32 (dd, J=12.5, 7.4 Hz, 1H), 3.70 (s, 3H), 3.50−3.19 (m, 6H), 3.05 (dd, J=13.3, 3.2 Hz, 1H), 2.41−2.19 (m, 4H), 1.82−1.57 (m, 4H), 1.42 (s, 3H), 1.11 (s, 3H). ¹³C NMR (126 MHz, Acetone-d₆) δ 202.68, 173.66, 171.75, 170.49, 143.75, 141.57, 141.53, 139.98, 126.47, 126.47, 126.23, 124.21, 124.20, 108.68, 75.22, 71.24, 62.50, 52.75, 51.07, 50.38, 43.09, 42.64, 41.78, 37.76, 36.13, 35.67, 35.28, 30.78, 18.19, 15.77, 14.54. HRMS m/z calcd for C₃₁H₃₄O₈Cs [M+Cs]⁺ 667.1308, found: 667.1295. Purity after HPLC: 98.73%.

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl benzo[d]oxazole-2-carboxylate, 4

General ester coupling procedure was followed to afford 4. Yield: 25%. Mp 217–219°C. ¹H NMR (400 MHz, DMSO-d₆) δ 7.96 (dd, J=22.4, 8.1 Hz, 2H), 7.76−7.63 (m, 3H), 7.56 (t, J=7.7 Hz, 1H), 6.61 (s, 1H), 5.69 (dd, J=12.4, 7.4 Hz, 1H), 5.59 (dd, J=11.7, 5.2 Hz, 1H), 3.69 (s, 3H), 3.09 (d, J=10.6 Hz, 1H), 2.90 (s, 1H), 2.46−2.12 (m, 4H), 1.94 (d, J=11.0 Hz, 1H), 1.84−1.48 (m, 5H), 1.32 (s, 3H), 1.03 (s, 3H). ¹³C NMR (126 MHz, DMSO-d₆) δ 202.51, 172.07, 171.52, 155.02, 152.48, 150.69, 144.53, 140.71, 140.42, 129.11, 126.56, 126.10, 122.27, 112.53, 109.40, 77.67, 71.45, 61.80, 52.21, 52.04, 50.00, 42.78, 42.03, 37.69, 35.36, 30.68, 18.33, 16.53, 15.24. HRMS m/z calcd for C₂₉H₂₉NO₉Cs [M+Cs]⁺ 668.0897, found 668.0891. Purity after HPLC: 95.86%.

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl 1H-indazole-3-carboxylate, 5

General ester coupling procedure was followed to afford 5. Yield: 25%. Mp 263–266°C. ¹H NMR (400 MHz, DMSO-d₆) δ 13.99 (s, 1H), 8.07 (d, J=8.2 Hz, 1H), 7.72 (s, 1H), 7.70−7.66 (m, 2H), 7.46 (t, J=7.2 Hz, 1H), 7.33 (t, J=7.3 Hz, 1H), 6.62−6.60 (m, 1H), 5.65−5.56 (m, 2H), 3.67 (s, 3H), 3.09 (dd, J=13.2, 3.2 Hz, 1H), 2.88 (s, 1H), 2.48−2.40 (m, 2H), 2.34−2.16 (m, 2H), 1.95 (d, J=24.0 Hz, 1H), 1.80 (d, J=12.7 Hz, 1H), 1.68−1.44 (m, 4H), 1.32 (s, 3H), 1.04 (s, 3H). ¹³C NMR(101 MHz, DMSO-d₆) δ 203.55, 172.26, 171.58, 161.37, 144.50, 141.39, 140.73, 135.03, 127.21, 126.12, 123.49, 122.64, 121.58, 111.59, 109.42, 75.81, 71.49, 61.87, 52.37, 52.14, 50.04, 42.83, 42.08, 37.75, 35.37, 31.08, 18.38, 16.56, 15.29. HRMS m/z calcd for C₂₉H₃₀N₂O₈Cs [M+Cs]⁺ 667.1057, found: 667.1067. Purity after HPLC: 95.02%.

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl 1H-benzo[d]imidazole-2-carboxylate, 6

General ester coupling procedure was followed to afford 6. Yield: 42%. Mp 205–207°C. ¹H NMR (400 MHz, DMSO-d₆) δ 13.49 (s, 1H), 7.79 (d, J=8.1 Hz, 1H), 7.70 (d, J=16.2 Hz, 2H), 7.59 (d, J=8.1 Hz, 1H), 7.36 (dt, J=32.7, 7.2 Hz, 2H), 6.61 (s, 1H), 5.60 (td, J=16.0, 14.2, 6.3 Hz, 2H), 3.68 (s, 3H), 3.08 (d, J=10.3 Hz, 1H), 2.89 (s, 1H), 2.43 (d, J=11.8 Hz, 1H), 2.36−2.27 (m, 1H), 2.19 (dd, J=13.1, 5.0 Hz, 1H), 2.09 (s, 1H), 1.94 (d, J=11.2 Hz, 1H), 1.78 (t, J=12.4 Hz, 1H), 1.64 (s, 2H), 1.31 (s, 3H), 1.03 (s, 3H). ¹³C NMR (101 MHz, DMSO-d₆) δ 202.52, 171.70, 171.04, 157.77, 144.02, 142.91, 140.97, 140.18, 134.24, 125.64, 125.32, 123.05, 120.79, 112.69, 108.88, 76.28, 70.98, 61.31, 51.67, 49.52, 42.32, 41.56, 37.24, 34.88, 30.65, 30.32, 17.87, 16.04, 14.77. HRMS m/z calcd for C₂₉H₂₉N₂O₈ [M−H]⁻ 533.1924, found: 533.1894. Purity after HPLC: 100%.

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl 4-fluoro-1H-indole-2-carboxylate, 7

General ester coupling procedure was followed to afford 7. Yield: 42%. Mp 248–249°C. ¹H NMR (400 MHz, DMSO-d₆) δ 12.24 (t, J=2.3 Hz, 1H), 7.73 (t, J=1.1 Hz, 1H), 7.68 (t, J=1.7 Hz, 1H), 7.33−7.23 (m, 3H), 6.88 (ddd, J=10.6, 7.3, 1.1 Hz, 1H), 6.61 (dd, J=1.9, 0.9 Hz, 1H), 5.62−5.52 (m, 2H), 3.68 (s, 3H), 3.07 (dd, J=13.3, 3.4 Hz, 1H), 2.88 (s, 1H), 2.44 (tt, J=11.3, 3.4 Hz, 2H), 2.30 (q, J=12.9 Hz, 1H), 2.20 (dd, J=13.4, 5.3 Hz, 1H), 1.94 (dd, J=13.5, 3.4 Hz, 1H), 1.78 (t, J=12.5 Hz, 1H), 1.70−1.46 (m, 3H), 1.32 (s, 3H), 1.03 (s, 3H). ¹³C NMR (101 MHz, DMSO-d₆) δ 203.48, 172.26, 171.57, 159.95, 144.51, 140.68, 140.27, 127.62, 126.15, 125.99, 116.72, 109.61, 109.39, 105.02, 104.85, 104.28, 76.03, 71.50, 61.79, 52.32, 52.13, 50.04, 42.83, 42.07, 37.79, 35.36, 30.94, 18.38, 16.55, 15.27. ¹⁹F NMR (377 MHz, CD₃CN) δ −122.14. HRMS m/z calcd for C₃₀H₃₀FNO₈Cs [M+Cs]⁺ 684.1010, found: 684.1010. Purity after HPLC: 95.03%.

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl 5-fluoro-1H-indole-2-carboxylate, 8

General ester coupling procedure was followed to afford 8. Yield: 43%. Mp 247–249°C. ¹H NMR (400 MHz, Acetone-d₆) δ 11.12 (s, 1H), 7.64 (s, 1H), 7.59−7.53 (m, 2H), 7.43 (dd, J=9.6, 2.6 Hz, 1H), 7.29 (d, J=2.2 Hz, 1H), 7.14 (td, J=9.2, 2.6 Hz, 1H), 6.58 (d, J=1.9 Hz, 1H), 5.58 (ddd, J=19.9, 12.1, 6.4 Hz, 2H), 3.74 (s, 3H), 3.14 (dd, J=13.3, 3.6 Hz, 1H), 2.52 (ddd, J=13.1, 7.5, 3.6 Hz, 1H), 2.45−2.37 (m, 3H), 1.91−1.75 (m, 3H), 1.75−1.59 (m, 2H), 1.44 (s, 3H), 1.16 (s, 3H). ¹³C NMR (101 MHz, Acetone-d₆) δ 202.31, 171.76, 170.48, 159.71, 143.76, 139.94, 134.36, 128.78, 126.30, 114.03, 113.73, 113.64, 108.67, 108.55 (d, J=5.1 Hz), 106.24, 106.00, 75.61, 71.28, 62.55, 52.79, 51.10, 50.41, 43.17, 41.87, 37.82, 35.34, 30.86, 18.22, 15.83, 14.56. ¹⁹F NMR (377 MHz, DMSO-d₆) δ −123.97. HRMS m/z calcd for C₃₀H₃₁FNO₈ [M+H]⁺ 552.19, Found: 552.22. Purity after HPLC: 100%.

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl 6-fluoro-1H-indole-2-carboxylate, 9

General ester coupling procedure was followed to afford 9. Yield: 49%. Mp 246–250°C. ¹H NMR (400 MHz, DMSO-d₆) δ 11.97 (d, J=2.3 Hz, 1H), 7.75−7.66 (m, 3H), 7.26 (dd, J=2.2, 0.9 Hz, 1H), 7.18 (dd, J=9.8, 2.4 Hz, 1H), 6.98 (ddd, J=9.7, 8.8, 2.4 Hz, 1H), 6.61 (dd, J=1.9, 0.9 Hz, 1H), 5.63−5.47 (m, 2H), 3.68 (s, 3H), 3.07 (dd, J=13.2, 3.4 Hz, 1H), 2.44 (tt, J=11.9, 3.4 Hz, 2H), 2.32−2.15 (m, 2H), 1.94 (dd, J=13.6, 3.3 Hz, 1H), 1.78 (t, J=12.5 Hz, 1H), 1.65 (t, J=5.4 Hz, 2H), 1.59−1.46 (m, 1H), 1.32 (s, 3H), 1.03 (s, 3H). ¹³C NMR (101 MHz, DMSO-d₆) δ 204.08, 173.89, 170.99, 162.63, 160.07, 147.36, 142.52, 138.31, 127.88, 126.15, 124.29, 123.73, 108.81, 98.55, 98.29, 76.20, 73.09, 63.35, 52.92, 52.14, 50.03, 45.04, 42.05, 37.03, 35.36, 31.14, 31.00, 19.62, 17.20, 15.41. ¹⁹F NMR (377 MHz, DMSO-d₆) δ −123.97 (td, J=9.6, 4.7 Hz). HRMS m/z calcd for C₃₀H₃₀FNO₈Cs [M+Cs]⁺ 684.1010, found 684.1010. Purity after HPLC: 95.32%.

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl 7-fluoro-1H-indole-2-carboxylate, 10

General ester coupling procedure was followed to afford 10. Yield: 41%. Mp 249–251°C. ¹H NMR (400 MHz, DMSO-d₆) δ 12.42−12.29 (m, 1H), 7.73−7.66 (m, 2H), 7.51 (d, J=7.8 Hz, 1H), 7.32 (t, J=2.6 Hz, 1H), 7.17−7.04 (m, 2H), 6.63−6.59 (m, 1H), 5.55 (ddd, J=22.6, 11.9, 6.4 Hz, 2H), 3.68 (s, 3H), 3.07 (dd, J=13.2, 3.6 Hz, 1H), 2.87 (s, 1H), 2.46−2.40 (m, 2H), 2.38−2.28 (m, 1H), 2.19 (dd, J=13.4, 5.3 Hz, 1H), 1.94 (d, J=10.7 Hz, 1H), 1.76 (d, J=12.5 Hz, 1H), 1.65 (d, J=8.5 Hz, 2H), 1.60−1.52 (m, 1H), 1.32 (s, 3H), 1.03 (s, 3H). ¹³C NMR (101 MHz, DMSO-d₆) δ 204.48, 172.32, 171.58, 159.96, 151.04, 144.52, 140.70, 130.90 (d, J=5.1 Hz), 128.63, 126.34 (d, J=14 Hz), 126.14, 121.01, 118.78, 109.73 (d, J=16 Hz), 109.39, 76.00, 71.49, 61.78, 52.35, 52.13, 50.03, 42.83, 42.07, 37.79, 35.36, 31.14, 30.89, 18.38, 16.56, 15.28. ¹⁹F NMR (377 MHz, DMSO-d₆) δ −130.85. HRMS m/z calcd for C₃₀H₃₀FNO₈Cs [M+Cs]⁺ 684.1010, found: 684.1010. Purity after HPLC: 95.06%.

Benzo[d]thiazole-2-carboxylic acid, 16^[24]

To a solution of commercially available 15 (207mg, 1 mmol) in tetrahydrofuran (THF, 0.5 mL) at 10°C was added a solution of LiOH (42mg, 1mmol) in water (2 mL), and stirred for 30 min. This was followed by dropwise acidification with 1 N HCl until precipitation occurred. The precipitate was filtered and dried to afford 16 that was used immediately for the next reaction. Yield: 57%. Mp 66–72°C. ¹H NMR: (400 MHz, DMSO-d₆) δ 14.42 (s, 1H), 8.26−8.13 (m, 2H), 7.63 (tt, J=7.4, 5.5 Hz, 2H); matching with literature values.^[24,25]

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl benzo[d]thiazole-2-carboxylate, 11

General ester coupling procedure was followed to afford 11. Yield: 71%. Mp 278–280°C. ¹H NMR (500 MHz, CDCl₃) δ 8.28 (d, J=8.0 Hz, 1H), 7.99 (d, J=7.7 Hz, 1H), 7.58 (p, J=7.1 Hz, 2H), 7.41 (d, J=9.5 Hz, 2H), 6.38 (s, 1H), 5.55−5.45 (m, 2H), 3.75 (s, 3H), 2.84 (d, J=13.1 Hz, 1H), 2.60 (dt, J=36.1, 10.9 Hz, 3H), 2.27 (s, 1H), 2.19 (d, J=13.4 Hz, 1H), 2.11 (d, J=11.0 Hz, 1H), 1.84 (d, J=10.9 Hz, 1H), 1.66 (d, J=22.2 Hz, 3H), 1.46 (s, 3H), 1.18 (s, 3H). ¹³C NMR (126 MHz, CDCl₃) δ 200.34, 171.30, 171.05, 159.59, 156.83, 153.26, 143.75, 139.41, 136.89, 127.88, 127.27, 125.76, 125.21, 122.11, 108.38, 72.07, 64.18, 53.62, 52.10, 51.40, 43.36, 42.27, 38.21, 35.55, 30.72, 18.17, 16.58, 15.22. HRMS m/z calcd for C₂₉H₂₉NO₈SCs [M+Cs]⁺ 684.0668, found: 684.0657. Purity after HPLC: 97.90%.

Methyl 2-(hydroxy(pyridin-2-yl)methyl)acrylate, 18^[21,24,25]

A mixture of commercially available 17 (536 mg, 5.0 mmol), methyl acrylate (0.54 mL, 6.0 mmol) and 1,4-diazabicyclo[2.2.2]octane (DABCO, 34 mg, 0.3 mmol) in dioxane/H₂O (3 mL/1 mL per mmol) was stirred for 3 hours at room temperature. Then the reaction mixture was concentrated in vacuo, and the residue was purified via column chromatography (1:3 ethyl acetate:hexanes) to afford 0.43 g of 18 as a dark yellow oil. Yield: 45%. Mp 50–54°C. ¹H NMR (400 MHz, CDCl₃) δ 8.55 (dt, J=4.8, 1.7 Hz, 1H), 7.68 (td, J=7.7, 1.7 Hz, 1H), 7.43 (dq, J=7.9, 1.0 Hz, 1H), 7.22 (ddd, J=7.5, 4.9, 1.3 Hz, 1H), 6.37 (d, J=1.1 Hz, 1H), 5.98 (t, J=1.1 Hz, 1H), 5.64 (s, 1H), 4.89 (s, 1H), 3.74 (d, J=1.2 Hz, 3H). ¹³C NMR (101 MHz, CDCl₃) δ 166.58, 159.49, 148.28, 141.78, 136.84, 126.86, 122.68, 121.28, 72.08, 51.89.

Methyl indolizine-2-carboxylate, 19^[21]

In a round-bottomed flask, a solution of 18 (0.43 g, 2.24 mmol) and acetic anhydride (10 mL) was stirred at 100°C for 1 hour under N₂ atmosphere. After the complete disappearance of 18, indicated by thin layer chromatography, the reaction solution was then heated at reflux (160°C) for additional 2 hours. The solution was cooled to room temperature, then poured into a mixture of ice and saturated aqueous sodium bicarbonate, and stirred for 1 hour. After stirring, the solution was extracted with ethyl acetate (3x). The combined organic layers were dried with anhydrous sodium sulfate, filtered, and concentrated in vacuo. The crude product was purified via column chromatography (1:3 ethyl acetate:hexanes) to afford 0.18 g of 19 as an off-white solid. Yield: 50%. Mp 96–99°C. ¹H NMR (400 MHz, DMSO-d₆) δ 8.28 (dq, J=7.2, 1.1 Hz, 1H), 8.09 (d, J=1.7 Hz, 1H), 7.46 (d, J=9.1 Hz, 1H), 6.79−6.72 (m, 2H), 6.65 (td, J=6.8, 1.4 Hz, 1H), 3.80 (d, J=1.4 Hz, 3H). ¹³C NMR (101 MHz, DMSO-d₆) δ 165.14, 132.71, 126.69, 120.24, 119.15, 116.91, 112.65, 100.14, 51.72.

Indolizine-2-carboxylic acid, 20^[21]

To a solution of 19 (0.193 g, 1 mmol) in MeOH/THF (4 mL:4 mL) was added 1 mL solution of 20% aqueous sodium hydroxide and refluxed at 80°C for 12 hours. The organic solvents were then removed under reduced pressure. The remaining solution was cooled to 0–5°C and adjusted to pH 3–4 with 1 M HCl. The suspension was then stirred for 30 min and filtered. The filter cake was dried to dryness to afford 0.16 g of 20 which was used directly for the subsequent reaction. Yield: 95%. Mp 240–242°C. ¹H NMR (400 MHz, DMSO-d₆) δ 12.27(bs), 8.27 (dq, J=7.1, 1.1 Hz, 1H), 8.01 (d, J=1.6 Hz, 1H), 7.44 (d, J=9.1 Hz, 1H), 6.77−6.68 (m, 2H), 6.62 (td, J=6.7, 1.3 Hz, 1H); matching literature values.^[21]

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl indolizine-2-carboxylate, 12

General ester coupling procedure was followed to afford 12. Yield: 49%. Mp 217–219°C. ¹H NMR (500 MHz, Acetone-d₆) δ 8.28−8.22 (m, 1H), 8.09−8.03 (m, 1H), 7.64 (s, 1H), 7.56 (t, J=1.6 Hz, 1H), 7.47 (d, J=9.1 Hz, 1H), 6.83 (s, 1H), 6.78 (dd, J=9.1, 6.5 Hz, 1H), 6.69−6.62 (m, 1H), 6.60−6.56 (m, 1H), 5.62 (dt, J=11.8, 5.9 Hz, 1H), 5.49 (dd, J=12.5, 7.4 Hz, 1H), 3.74 (s, 3H), 3.12 (dd, J=13.3, 3.5 Hz, 1H), 2.81 (s, 2H), 2.50−2.27 (m, 4H), 1.89−1.58 (m, 5H), 1.44 (s, 3H), 1.15 (s, 3H). 13C NMR (126 MHz, Acetone-d₆) δ 202.72, 171.84, 170.55, 163.02, 143.77, 139.96, 132.82, 126.28, 126.00, 119.92, 118.98, 118.46, 116.63, 112.25, 108.70, 100.20, 74.90, 71.31, 62.52, 52.86, 51.08, 50.41, 43.15, 41.82, 37.81, 35.33, 31.00, 18.22, 15.80, 14.56. HRMS m/z calcd for C₃₀H₃₁NO₈Cs [M+Cs]⁺ 666.1104, found: 666.1085. Purity after HPLC: 98.89%.

(R)-1-(((9H-fluoren-9-yl)methoxy)carbonyl)indoline-2-carboxylic acid, 22^[26]

A solution of Fmoc-Cl (1 equiv) in dioxane (2.6 mL/mmol) was added to a round-bottom flask containing a suspension of commercially available 21 in dioxane (1.3 mL/mmol and 10% aqueous Na₂CO₃ 2.6 mL/mmol) at 0°C. The mixture was stirred for 1 hour at 0°C and then for 1 hour at room temperature. The reaction mixture was poured into water and washed with diethyl ether. The aqueous phase was then acidified with concentrated HCl, and the precipitated product was isolated by filtration and dried in vacuo and used immediately for the next reaction. Yield: 90%.

1-((9H-fluoren-9-yl)methyl)-2-((2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl) (R)-indoline-1,2-dicarboxylate, 23

General ester coupling procedure was followed to afford a mixture of 23 and 13, which was used immediately for the next reaction. Yield: 89%. HRMS m/z calcd for C₄₅H₄₃NO₁₀ [M⁺] 757.29, found: 757.59.

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl (R)-indoline-2-carboxylate, 13

General Fmoc deprotection procedure was followed to afford 13. Yield: 84%. Mp 204–205°C. ¹H NMR (500 MHz, DMSO-d₆) δ 11.91 (s, 1H), 7.77−7.61 (m, 3H), 7.47 (d, J=8.3 Hz, 1H), 7.32−7.20 (m, 2H), 7.10 (t, J=7.5 Hz, 1H), 6.62 (s, 1H), 5.56 (ddd, J=27.5, 12.0, 6.3 Hz, 2H), 3.68 (s, 3H), 3.07 (dd, J=13.2, 2.9 Hz, 1H), 2.87 (s, 1H), 2.44 (t, J=10.1 Hz, 2H), 2.33−2.14 (m, 2H), 2.07−1.85 (m, 1H), 1.78 (t, J=12.6 Hz, 1H), 1.65−1.48 (m, 3H), 1.31 (s, 4H), 1.02 (s, 3H). ¹³C NMR (126 MHz, DMSO-d₆) δ 203.67, 172.31, 171.63, 160.36, 144.54, 140.71, 137.91, 127.15, 127.01, 126.08, 125.37, 122.63, 120.73, 113.03, 109.40, 108.99, 75.78, 71.48, 61.72, 52.29, 52.17, 49.99, 42.78, 42.05, 37.76, 35.33, 30.96, 18.36, 16.54, 15.25. HRMS m/z calcd for C₃₀H₃₄NO₈ [M+H]⁺ 536.2284, found: 535.2384. Purity after HPLC: 95.30%.

1-((9H-fluoren-9-yl)methyl)-2-((2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl) (S)-indoline-1,2-dicarboxylate, 25

General ester coupling procedure was followed to afford 25. Yield 75%. ¹H NMR (500 MHz, DMSO-d₆) δ 7.94 (dd, J=22.4, 7.2 Hz, 2H), 7.75−7.60 (m, 5H), 7.48−7.26 (m, 5H), 7.24−7.12 (m, 1H), 7.01 (t, J=7.3 Hz, 1H), 6.79 (dt, J=46.4, 7.4 Hz, 1H), 6.53 (s, 1H), 5.50 (dd, J=11.6, 5.4 Hz, 1H), 5.35−5.22 (m, 1H), 5.06−4.99 (m, 1H), 4.91 (d, J=10.6 Hz, 1H), 4.83 (s, 1H), 4.53−4.39 (m, 2H), 4.25 (d, J=5.9 Hz, 1H), 3.65 (s, 3H), 3.00 (dd, J=13.1, 2.8 Hz, 1H), 2.76 (d, J=19.5 Hz, 1H), 2.36 (d, J=9.7 Hz, 1H), 2.31−2.00 (m, 3H), 1.89 (d, J=12.8 Hz, 1H), 1.68−1.44 (m, 5H), 1.30−1.22 (m, 4H), 0.95 (d, J=10.6 Hz, 3H). ¹³C NMR (126 MHz, DMSO-d₆) δ 203.65, 172.10, 171.52, 170.82, 144.44, 144.25, 142.31, 141.84, 140.64, 128.12, 127.61, 125.95, 125.39, 120.66, 120.60, 114.26, 109.33, 75.78, 71.40, 67.40, 61.66, 60.24, 59.45, 52.18, 51.92, 49.91, 42.69, 41.98, 37.66, 35.20, 30.84, 18.29, 16.46, 15.19. LC–MS m/z calcd for C₄₅H₄₃NO₁₀Cs [M+Cs]: 890.2, found: 890.2. Purity: 96.06%.

(2S,4aR,6aR,7R,9S,10aS,10bR)-2-(furan-3-yl)-7-(methoxycarbonyl)-6a,10b-dimethyl-4,10-dioxododecahydro-2H-benzo[f]isochromen-9-yl (S)-indoline-2-carboxylate, 14

General Fmoc deprotection procedure was followed to afford 14. Yield: 84%. Mp 201–204°C. ¹H NMR (400 MHz, CDCl₃) δ 8.97 (s, 1H), 7.72 (d, J=8.1 Hz, 1H), 7.50−7.33 (m, 5H), 7.18 (t, J=7.5 Hz, 1H), 6.39 (d, J=9.0 Hz, 1H), 5.54 (dd, J=11.7, 5.1 Hz, 1H), 5.43 (t, J=10.0 Hz, 1H), 3.76 (d, J=9.7 Hz, 3H), 2.89−2.82 (m, 1H), 2.53 (ddd, J=20.4, 15.5, 7.3 Hz, 3H), 2.32−2.06 (m, 4H), 1.85 (d, J=12.7 Hz, 1H), 1.76−1.58 (m, 5H), 1.48 (s, 3H), 1.20 (s, 3H). ¹³C NMR (101 MHz, CDCl₃) δ 201.92, 171.67, 171.26, 160.74, 143.86, 139.58, 137.29, 127.56, 126.11, 126.03, 125.37, 122.93, 121.17, 112.05, 110.38, 108.54, 75.52, 72.19, 64.30, 53.77, 52.19, 51.55, 43.55, 42.36, 38.36, 35.66, 31.14, 18.31, 16.66, 15.37. HRMS m/z calcd for C₃₀H₃₂ NO₈ [M−H]⁻ 534.2128, found: 534.2095. Purity after HPLC: 100%.

Receptor membrane isolation

Chinese hamster ovary (CHO) cells expressing hKOR, hMOR, or hDOR (Valiscreen cells, Perkin Elmer) were cultured in T-175 plates with Ham’s F12 media supplemented with fetal bovine serum (FBS), penicillin, and streptomycin (0.5 mg/mL), as well as selection antibiotics, per manufacturers’ recommendations. After reaching confluency, cells were washed with ice cold phosphate buffered saline (PBS) and scraped from the cell culture plate in cold 50 mM Tris-HCl buffer (pH 7.4). The cells were pelleted at 1,000×g for 10 minutes at 4°C. The supernatant was discarded, and the pellet was again washed with Tris-HCl buffer, sonicated, and centrifuged at 5,200×g at 4°C for 10 minutes to remove debris and contaminated proteins. Supernatants were pooled and centrifuged at 40,000×g for 40 minutes at 4°C. The membrane-rich pellet was re-suspended in Tris-HCl buffer, sonicated, and aliquoted for storage at −80°C until use. A Pierce^™ bicinchoninic acid (BCA) Protein Assay was utilized to determine the protein concentration in the membrane, according to the manufacturer’s instructions. Membranes were evaluated in saturation experiments to determine the optimal concentration for radioligand binding and functional assays.

Determination of K_D values

A competitive binding assay was utilized to screen for receptor binding and determine the K_D values of each compound against MOR, KOR, and DOR, similar to the previous reports.^[23] For this, 15–25 micrograms of the cell membrane were diluted in 50 mM Tris-HCl, pH 7.4 and pipetted into a 96-well plate. Test compounds were reconstituted in DMSO and added to the reaction plate. Radio-ligands specific to each receptor were then added at a final concentration of 2–4 nM: KOR [³H]-U-69,593; MOR [³H]-DAMGO; DOR [³H]-Deltorphin II (all radioligands were purchased from Perkin Elmer). The reaction was incubated for 1 hour at room temperature prior to transfer onto GF/B filter plates (Perkin Elmer). The plates were washed 10× with cold reaction buffer and dried for 15 minutes at 50°C. MicroScint20 was then added to each well, the plates were sealed, and the counts per minute were quantified on the TopCount NXT or Microbeta Microplate Scintillation counter (Perkin Elmer). The percent displacement of the radiolabeled compound was calculated with the following equation: 100*(cpm of test compound-cpm of non-specific binding)/(cpm max binding-cpm of non-specific binding). Unlabeled U-69,593, DAMGO, and DPDPE were used to determine non-specific control counts, while vehicle (DMSO) was added to determine the total radioligand binding on each plate. A series of three-fold dilutions were used for each test compound starting at 10.0 μM. A control curve of naloxone was included on each plate for internal control.

GTPγS functional assay

Twenty-five micrograms of the CHO cell MOR, KOR, or DOR cell membrane were diluted in reaction buffer [50 mM Tris-HCl, 150 mM NaCl, 9 mM MgCl, 0.2 mM (ethylene glycol-bis(β-aminoethyl ether)-N,N,N’,N’-tetraacetic acid), 0.14% Bovine serum albumin (BSA), 10 μM guanosine 5’-diphosphate (GDP), pH 7.4] and pipetted into a 96-well plate. Test compounds were reconstituted in DMSO and then added to the reaction, followed by radiolabeled GTPγ³⁵S (0.05 nM final concentration). The reaction was incubated for 1 hour at room temperature and subsequently transferred to a pre-soaked GF/B plate (soaked in 0.3% BSA). The plate was washed 10x in 50 mM Tris-HCl, pH 7.4 to remove unbound GTPγ³⁵S. The plate was then dried for 15 minutes at 50°C. MicroScint20 was added to each well; the plate was then sealed; and the radioactivity was quantified in 3 replicates on the TopCount. For control purposes, six wells were incubated with (D-Ala(2)-mephe(4)-gly-ol-(5))enkephalin (DAMGO), n-methyl-2-phenyl-n-[(5r,7s,8s)-7-(pyrroli-din-1-yl)-1-oxaspiro[4.5]dec-8-yl]acetamide (U69,593), or (D-Pen2,D-Pen5)-Enkephalin (DPDPE) (10 mM) to calculate E_max, vehicle (DMSO) to calculate basal activity, and 40 μM unlabeled GTPγ³⁵S to determine non-specific binding. Data were quantified as (cpm of test compound-cpm of basal)/(cpm of basal-cpm of non-specific binding) and presented as percent GTPγ³⁵S stimulation.

In vivo assays

Mice

All experimental procedures were approved by the University of Mississippi Institutional Animal Care and Use Committee (protocol #19-009). This study was carried out in male wild-type C57BL/6NHsd mice (N=150; 6—8 weeks old, 19—30 grams) purchased from Envigo (Indianapolis, IN, USA). Mice were group-housed (n=4–5/cage) in a polycarbonate homecage with soft bedding in a temperature- (22°±2) and humidity-controlled vivarium under a 12:12 h light/dark cycle (lights on at 07:00 h) with food and water ad libitum. Mice were acclimated for at least 5 days prior to experimentation. Behavioral testing was performed during the light period (between 08:00–18:00) after a 30-minute acclimation to the testing room. Equipment was sanitized with 70% ethanol between mice in all assays. Compounds 2, 10, salvinorin A, and herkinorin were administered via intraperitoneal injection in a vehicle solution of 4:4:2 DMSO:PEG200:PBS at 2 mg/mL. Morphine was administered via intraperitoneal injection in a solution of saline at 0.5 mg/mL. Compound treatments were coded to allow for double-blinded observations, and block randomization was utilized to assign equally sized treatment groups. All compounds were administered 30 min prior to testing.

Hot plate test

Mice were placed in an open top observation chamber (12.7×15.24 cm) on a 52°C heated plate (IITC Life Science, CA). A digital timer was initiated by the depression of a foot switch that measured the latency (seconds) of withdrawal responses (hind paw flutter or lick), which indicated supraspinally-mediated thermal nociception. A 45-second cut-off was used to prevent tissue damage.

Tail-flick test

Thermal nociception was also assessed in the tail-flick assay. Mice were confined in a restrainer with their tails positioned above a radiant heat source (IITC #33; IITC Life Science, Inc., Woodland Hills, CA). The latency to tail withdrawal from the light stimulus was used as an index of spinally-mediated thermal nociception. The intensity of the light was adjusted to yield a baseline latency around 2 seconds in control animals, and a 20 second cutoff was used to prevent tissue damage.

Elevated plus maze test

Mice were placed on the elevated plus maze (San Diego Instruments) facing an outer arm and were allowed to freely explore the maze for 5 min. Behavior was digitally-encoded using Noldus Ethovision (ver. 14) software. Total locomotion, velocity, and amount of time spent in the open arms were recorded as indices of motor and anxiety-like behavior, respectively. If mice fell off the maze, they were immediately placed in the center of the maze to complete the 5-minute trial, and a notation was made. No mice were excluded from analyses due to repeated falls.

Statistical analyses

Median effective concentrations (EC₅₀; reported with 95% confidence intervals), maximal effective concentrations (E_max; reported with SEM), and inhibition constants (K_D; reported with 95% confidence intervals) were determined via non-linear regression (sigmoidal curvilinear modeling with variable slope) using a least-squares fit for each compound via GraphPad Prism software (ver. 7.03; GraphPad Software, Inc., San Diego, CA, USA). Dependent behavioral measures were assessed via one-way ANOVA with drug condition as the between-subjects factors. Fisher’s Protected Least Significant Difference post-hoc tests determined group differences following main effects. All analyses were considered significant when p≤0.05.

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.