A MOR Antagonist with High Potency and Antagonist Efficacy among Diastereomeric C9-Alkyl-Substituted N-Phenethyl-5-(3-hydroxy)phenylmorphans

The 5-(3-hydroxy)phenylmorphan structural class of compounds are unlike the classical morphinans, 4,5-epoxymorphinans, and 6,7-benzomorphans, in that they have an equatorially oriented aromatic ring rather than the axial orientation of that ring found in the classical opioids. This modified and simplified opioid-like structure has been shown to retain antinociceptive activity, depending on its stereochemistry and substituents, and some of them have been found to be much more potent than morphine. A simple C9-hydroxy-5-(3-hydroxy)phenylmorphan enantiomer was found to be about 500 times more potent than morphine in vivo. We have previously examined C9-alkenyl and hydroxyalkyl substituents in the N-phenethyl-5-(3-hydroxy)phenylmorphan class of compounds. Comparable C9-alkyl (methyl through butyl) substituents, with their sets of diastereomers, have not been explored. All these compounds have now been synthesized to determine the effect chain-length and stereochemistry at the C9 position in the molecule might have on their interaction with opioid receptors. We now report the synthesis and in vitro activity of 16 compounds, the C9-methyl, ethyl, propyl, and butyl diastereomers, using the inhibition of forskolin-induced cAMP accumulation assay. Several potent (sub-nanomolar and nanomolar) MOR compounds were found to be selective agonists with varying efficacy. Of greatest interest, a selective MOR antagonist was discovered; it did not display any DOR or KOR agonist activity in vitro, was three times more potent than naltrexone, and was found to antagonize the EC90 of fentanyl at MOR to a greater extent than naltrexone.


Introduction
The Center for Disease Control and Prevention (CDC) reported that more than 932,000 people have died since 1999 from a drug overdose. In 2020, opioids were involved in 68,630 overdose deaths (URL (accessed on 7 June 2023) https://www.cdc.gov/ drugoverdose/deaths/index.html). According to the United States Department of Justice, the direct and indirect monetary cost of illicit drug use in the U.S. totaled around $193 billion in 2007. The cost in lives destroyed by the illicit use of drugs is immeasurable and, unfortunately, with increasingly potent analgesics reaching the market, those totals will only increase in future years. For these reasons, researchers have sought new types of analgesics as well as new and more selective and potent antagonists. Although no theory has succeeded in predicting the properties of a molecule that could act as a selective potent opioid antagonist, several theories have been developed that relate the properties of potential analgesics to their undesired side effects [1][2][3][4][5][6]. Some new types of analgesics (e.g., PZM21, SR17018, tramadol, tianeptine, oliceridine, Figure 1) have structures that appear to be quite different from the opioids used clinically (e.g., morphine, oxymorphone, and oxycodone (Figure 1)). Of these new compounds, only oliceridine has been approved for human use as an analgesic. It was shown to have somewhat fewer side effects than those associated with the opioids (the major opioid side effects are respiratory depression, gastrointestinal effects, tolerance, dependence, and OUD (opioid use disorder)).
tures that appear to be quite different from the opioids used clinically (e.g., morphine, oxymorphone, and oxycodone (Figure 1)). Of these new compounds, only oliceridine has been approved for human use as an analgesic. It was shown to have somewhat fewer side effects than those associated with the opioids (the major opioid side effects are respiratory depression, gastrointestinal effects, tolerance, dependence, and OUD (opioid use disorder)).
The 5-(3-hydroxy)phenylmorphan molecule ( Figure 1) was designed [7] as a simplified form of morphine, but that simplification introduced major changes. Morphine, and all the classical 4,5-epoxymorphinan opioids, have a rigid aromatic phenyl ring axially oriented towards the piperidine ring, unlike the 5-phenylmorphans where the phenyl ring is detached from morphine's B-ring and equatorially oriented towards the piperidine ring. The phenyl ring in the 5-phenylmorphans can rotate around a C-C bond, unlike the rigid phenyl ring in morphine. We have employed an N-phenethyl substituent, rather than Nmethyl substituent, since it has been found to increase the potency of the molecule in the chiral C9-alkyl series; although in a racemic 5-phenylmorphan, it was initially found to decrease antinociceptive activity in mice [8]. Many other types of N-substituents have been noted in the literature [9,10]. We have used the N-phenethyl substituent for all of our C9-diastereomers, and we maintain that, The 5-(3-hydroxy)phenylmorphan molecule (Figure 1) was designed [7] as a simplified form of morphine, but that simplification introduced major changes. Morphine, and all the classical 4,5-epoxymorphinan opioids, have a rigid aromatic phenyl ring axially oriented towards the piperidine ring, unlike the 5-phenylmorphans where the phenyl ring is detached from morphine's B-ring and equatorially oriented towards the piperidine ring. The phenyl ring in the 5-phenylmorphans can rotate around a C-C bond, unlike the rigid phenyl ring in morphine. We have employed an N-phenethyl substituent, rather than N-methyl substituent, since it has been found to increase the potency of the molecule in the chiral C9-alkyl series; although in a racemic 5-phenylmorphan, it was initially found to decrease antinociceptive activity in mice [8].
Many other types of N-substituents have been noted in the literature [9,10]. We have used the N-phenethyl substituent for all of our C9-diastereomers, and we maintain that, initially, as a constant for comparison of their activity in vitro. In earlier work with the 5-phenylmorphans, we determined the effect of the enantiomers of a C9-hydroxyl group and a C9-methyl group in a few chiral N-phenethyl-substituted 5-phenylmorphans [11]. The 1R,5R,9S-OH compound and the 1R,5S,9R-methyl compound were found to have sub-nanomolar MOR affinity in a receptor binding assay. A later study with a 1S,5R,9R-propyl group at C9 in the phenylmorphans was also intriguing [12]. In an inhibition of forskolin-induced cAMP accumulation assay (cAMP assay), it acted as a partial agonist (EC 50 (potency) = 1.42 nM; %E max (efficacy) = 45%). Further testing in vivo has indicated that it was unlike morphine in its locomotor effects [13]. The C9-propyl compound appeared to have some antinociceptive activity, depending on the animal species used and the pain stimulus [13]. With our previous discovery that a chiral C9-methyl [11] and the C9-propyl [12] compound had interesting activity, we thought we should design compounds that would explore the complete 3-dimensional space around the C9 area of the 5-phenylmorphan, while maintaining the N-phenethyl substituent as a constant. A C9substituent introduced a third chiral atom in the molecule. However, with one chiral atom (at C5) fixed, we would only need four diastereomers for every C9-substituent. To examine C9-alkyl substituents from methyl through butyl, we synthesized a total of 16 compounds. Three of these (two C9-methyl compounds and the C9-propyl) had previously been prepared [11,12], and they were reexamined in vitro for comparison with the novel compounds. We have, in this same manner, examined C9-alkenyl compounds [14] and hydroxyalkyl compounds and found interesting MOR partial agonists and potent MOR antagonists among them. The agonists and antagonists were distinguishable using molecular modeling [14]. The antagonists were found to interact with the inactive (4DKL) MOR crystal structures and agonists were found to interact with the active (6DDF) MOR crystal structures [14]. The determination that several C9-alkenyl compounds were of interest [15] gave us hope that we would find interesting alkyl compounds.

Synthesis of Diastereomeric C9-Alkyl-5-phenylmorphans
The synthesis of the C9 ethyl though butyl series (n = 1-3, Figure 1) of phenylmorphans was achieved through the route shown in Scheme 1. The racemic N-methyl compound 1 (Scheme 1) was synthesized from commercially available compounds following literature procedures [7,16]. Using the procedure of Hiebel et al. [11], optical resolution to separate the enantiomers of ±1 was achieved using tartaric acid, and the enantiomers were purified to >99% ee (as determined using 1 H NMR). The tartrate salt of (1S,5S)-9-ketophenylmorpohan (1S, 5S-1) is shown in Scheme 1. For each enantiomer, the N-methyl substituent was replaced with an N-phenethyl moiety. Treatment of N-methyl compound 1 was with cyanogen bromide and subsequent hydrolysis with 3 N HCl yielded secondary amine 2 in high yield and purity. Further purification was unnecessary before alkylation. Treatment of secondary amine 2 with phenethyl bromide gave the alkylation product 3. Formation of alkylated product 3 gave a >60% yield over two steps. This procedure was scalable, and reactions were run on 1 g to 26 g scale without loss of yield.
Functionalization of the ketone at the C9 position via the formation of enol ether 4 (Scheme 1) was achieved by a Wittig reaction of 3 with methoxymethyl triphenylphosphonium chloride and LiHMDS. Installation of the C9 stereocenter was accomplished by the hydrolysis of enol ether 4 with aqueous HCl. The concentration of HCl, reaction time, and quenching conditions affect the C9 R:S ratio, as noted previously [14]. Aldehyde 5 was used immediately upon formation without purification because it was unstable to silica gel column chromatography and not stable at room temperature for longer than a few hours. From the common intermediate 5, Wittig reactions using varying phosphonium salts (Scheme 2) were performed to obtain alkenes at C9 [14]. The Wittig reactions resulted in a mixture of epimers at the C9 position, which were not separable by silica gel column chromatography. The mixture of epimers was subject to O-demethylation conditions using BBr 3 to form phenols 9-14 in 60-80% yields (Scheme 2). The phenol compounds 9-14 could be separated into their C9S and C9R diastereomers by column chromatography. Functionalization of the ketone at the C9 position via the formation of enol ether 4 (Scheme 1) was achieved by a Wittig reaction of 3 with methoxymethyl triphenylphosphonium chloride and LiHMDS. Installation of the C9 stereocenter was accomplished by the hydrolysis of enol ether 4 with aqueous HCl. The concentration of HCl, reaction time, and quenching conditions affect the C9 R:S ratio, as noted previously [14]. Aldehyde 5 was used immediately upon formation without purification because it was unstable to silica gel column chromatography and not stable at room temperature for longer than a few hours. From the common intermediate 5, Wittig reactions using varying phosphonium salts (Scheme 2) were performed to obtain alkenes at C9 [14]. The Wittig reactions resulted in a mixture of epimers at the C9 position, which were not separable by silica gel column chromatography. The mixture of epimers was subject to O-demethylation conditions using BBr3 to form phenols 9-14 in 60-80% yields (Scheme 2). The phenol compounds 9-14 could be separated into their C9S and C9R diastereomers by column chromatography. Alkenes 9-14 were hydrogenated in the presence of Pd/C (0.1 equiv) to give the desired alkyl compounds 15-20 (Scheme 2) in the 1S,5R series. The use of HCl salts of 9-14, or the treatment of the free base with one equivalent of HCl, for the hydrogenation reactions was necessary for high yielding reductions. When HCl was not employed, reaction yields dropped significantly. Alkyl compounds 15-20 were purified by column chromatography and isolated as oils. All the free base oils were treated with HCl in isopropanol to form the HCl salts as white solids, except for the C9 butyl compound 20, which was isolated as the tartrate salt.
The enantiomers (21-26) of compounds 15-20 were synthesized from the tartrate salt of (1R,5S)-5-(3-methoxyphenyl)-2-methyl-2-azabicyclo[3.3.1]nonan-9-one (1R,5S-21, Scheme 3) by the same synthetic steps as in Scheme 1. Alkenes 9-14 were hydrogenated in the presence of Pd/C (0.1 equiv) to give the desired alkyl compounds 15-20 (Scheme 2) in the 1S,5R series. The use of HCl salts of 9-14, or the treatment of the free base with one equivalent of HCl, for the hydrogenation reactions was necessary for high yielding reductions. When HCl was not employed, reaction yields dropped significantly. Alkyl compounds 15-20 were purified by column chromatography and isolated as oils. All the free base oils were treated with HCl in isopropanol to form the HCl salts as white solids, except for the C9 butyl compound 20, which was isolated as the tartrate salt.
sired alkyl compounds 15-20 (Scheme 2) in the 1S,5R series. The use of HCl salts of 9-14, or the treatment of the free base with one equivalent of HCl, for the hydrogenation reactions was necessary for high yielding reductions. When HCl was not employed, reaction yields dropped significantly. Alkyl compounds 15-20 were purified by column chromatography and isolated as oils. All the free base oils were treated with HCl in isopropanol to form the HCl salts as white solids, except for the C9 butyl compound 20, which was isolated as the tartrate salt.
The C9R-methyl diastereomer (48, Scheme 6) had not been previously made as the literature method [11] selectively formed only the C9S-methyl epimer 43. To obtain the C9R-epimer, N-methyl phenylmorphan 1 (Scheme 6) was subjected to N-demethylation conditions followed by a Boc protection resulting in 44. Using a Tebbe's olefination reaction to form the C9 methylene compound 45, followed by a hydrogenation reaction, resulted in a mixture of the desired C9R-stereochemistry in 46 and the C9S-epimer. Isolation and subsequent deprotection of the C9R isomer and the introduction of the phenethyl moiety resulted in compound 47. Finally, O-demethylation using standard conditions yielded the desired C9S-methyl compound 48.
Using the same procedures described in Scheme 2, the Wittig reactions with aldehyde 25 (Scheme 4) yielded compounds 26-28 as a mixture of epimers at the C9 position that were not easily separable. O-Demethylation of 26-28 with BBr3 gave phenols 29-34, respectively. The phenolic compounds were separable by silica gel column chromatography and provided C9R and C9S compounds. Hydrogenation using Pd/C gave the final products 35-40. The C9S-methyl compound (43) in the 1S,5R-series were synthesized (Scheme 5) from compound 3 using the literature procedure [11]. The alkylation product 3 olefination using Tebbe's reagent gave alkene 41 in a 71% yield (Scheme 5). Reduction of 41 gave stereoselectively the C9S-methyl 42 with only a small impurity of the C9R-methyl diastereomer. The final O-demethylated product 43 [11] was achieved by treating 42 with BBr3.  The C9S-methyl compound (43) in the 1S,5R-series were synthesized (Scheme 5) from compound 3 using the literature procedure [11]. The alkylation product 3 olefination using Tebbe's reagent gave alkene 41 in a 71% yield (Scheme 5). Reduction of 41 gave stereoselectively the C9S-methyl 42 with only a small impurity of the C9R-methyl diastereomer. The final O-demethylated product 43 [11] was achieved by treating 42 with BBr3. The selectivity of hydrogenation could be shifted based on the substituent on the oxygen and nitrogen (Table 1). With a phenethyl group on nitrogen and a m-methoxy substituted aromatic ring, the C9S epimer is formed exclusively. After O-demethylation the same hydrogenation conditions result in a 5:1 ratio still favoring the C9S epimer. Alternatively, when a Boc group is exchanged for the phenethyl group on the nitrogen and a meta-methoxoyphenyl moiety is present the selectivity shifts to favor the C9R epimer in a 3:1 ratio. This observation was exploited to obtain all four diastereomers of the C9-methyl phenylmorphan (e.g., Scheme 5 for the C9R-methyl in the 1S,5R series using the starting material in reaction #3 in Table 1; the C9S epimer was obtained using the starting material in reaction 1 in Table 1).
conditions followed by a Boc protection resulting in 44. Using a Tebbe's olefination reaction to form the C9 methylene compound 45, followed by a hydrogenation reaction, resulted in a mixture of the desired C9R-stereochemistry in 46 and the C9S-epimer. Isolation and subsequent deprotection of the C9R isomer and the introduction of the phenethyl moiety resulted in compound 47. Finally, O-demethylation using standard conditions yielded the desired C9S-methyl compound 48. The selectivity of hydrogenation could be shifted based on the substituent on the oxygen and nitrogen (Table 1). With a phenethyl group on nitrogen and a m-methoxy substituted aromatic ring, the C9S epimer is formed exclusively. After O-demethylation the same hydrogenation conditions result in a 5:1 ratio still favoring the C9S epimer. Alternatively, when a Boc group is exchanged for the phenethyl group on the nitrogen and a metamethoxoyphenyl moiety is present the selectivity shifts to favor the C9R epimer in a 3:1 ratio. This observation was exploited to obtain all four diastereomers of the C9-methyl phenylmorphan (e.g., Scheme 5 for the C9R-methyl in the 1S,5R series using the starting material in reaction #3 in Table 1; the C9S epimer was obtained using the starting material in reaction 1 in Table 1). The selectivity of hydrogenation could be shifted based on the substituent on the oxygen and nitrogen (Table 1). With a phenethyl group on nitrogen and a m-methoxy substituted aromatic ring, the C9S epimer is formed exclusively. After O-demethylation the same hydrogenation conditions result in a 5:1 ratio still favoring the C9S epimer. Alternatively, when a Boc group is exchanged for the phenethyl group on the nitrogen and a metamethoxoyphenyl moiety is present the selectivity shifts to favor the C9R epimer in a 3:1 ratio. This observation was exploited to obtain all four diastereomers of the C9-methyl phenylmorphan (e.g., Scheme 5 for the C9R-methyl in the 1S,5R series using the starting material in reaction #3 in Table 1; the C9S epimer was obtained using the starting material in reaction 1 in Table 1). Using the same procedures described in Schemes 5 and 6, the two C9-methyl diastereomers were obtained for the 1R,5S series (Schemes 7 and 8). The C9S and C9R-methyl compounds were differentiated based on the previously determined X-ray diffraction analyses of 51 [11]. Using the same procedures described in Schemes 5 and 6, the two C9-methyl diastereomers were obtained for the 1R,5S series (Schemes 7 and 8). The C9S and C9R-methyl compounds were differentiated based on the previously determined X-ray diffraction analyses of 51 [11]. Using the same procedures described in Schemes 5 and 6, the two C9-methyl diastereomers were obtained for the 1R,5S series (Schemes 7 and 8). The C9S and C9R-methyl compounds were differentiated based on the previously determined X-ray diffraction analyses of 51 [11]. Using the same procedures described in Schemes 5 and 6, the two C9-methyl diastereomers were obtained for the 1R,5S series (Schemes 7 and 8). The C9S and C9R-methyl compounds were differentiated based on the previously determined X-ray diffraction analyses of 51 [11]. 3 1:3 Using the same procedures described in Schemes 5 and 6, the two C9-methyl diastereomers were obtained for the 1R,5S series (Schemes 7 and 8). The C9S and C9R-methyl compounds were differentiated based on the previously determined X-ray diffraction analyses of 51 [11]. Using the same procedures described in Schemes 5 and 6, the two C9-methyl d stereomers were obtained for the 1R,5S series (Schemes 7 and 8). The C9S and C9R-met compounds were differentiated based on the previously determined X-ray diffract analyses of 51 [11].

Forskolin-Induced cAMP Accumulation Assay for In Vitro Determination of the Potency and Efficacy of the Diastereomers
Compounds were examined for their functional activity using the inhibition of forskolininduced cAMP accumulation assay ( Table 2). Four diastereomers for each of the C9-alkyl compounds, C9-methyl, ethyl, propyl, and butyl were evaluated. All except three of these (compounds 43 [11], 51 [11] and 16 [12]) in Table 2 are novel, and the three formerly known compounds have been included and re-evaluated so that the data on the complete series could be compared.
The C9-methyl series were of most interest in that one of them, 48, was a selective potent MOR antagonist (IC 50 = 3.9 nM) with a high %I max = 153%. It did not have DOR or KOR agonist activity, unlike naltrexone, which acts as a potent partial kappa agonist (KOR EC 50 = 0.6 nM). The C9-methyl antagonist 48 was also found to antagonize the EC90 of fentanyl at MOR to a greater extent than naltrexone (%I max = 153% vs 104% for naltrexone).      3.3 ± 0.9 (69 ± 7%) >10,000 75 ± 30 (157 ± 11%) >10,000 11 ± 6 (112 ± 9%)   1.1 ± 0.2 (107 ± 5%) a Inhibition of forskolin-induced cAMP accumulation; cAMP Hunter TM Chinese hamster ovary cells (CHO-K1) that express human μ-opioid receptor (OPRM1), human κ-opioid receptor (OPRK1), and human δ-opioid receptor (OPRD1) were used for the forskolin-induced cAMP accumulation assay in order to determine potency and efficacy of the compounds following the previously established methods; [17] to determine % efficacy in forskolin-induced cAMP assays, data were blank subtracted with the vehicle control, followed by normalization to the forskolin control. Data were then analyzed in GraphPad Prism 8 (GraphPad, LaJolla, CA, USA) using nonlinear regression; values are expressed as the mean ± SEM of at least three independent experiments; N/D = not determined. b MOR Antagonist potency (IC50) determined versus EC90 of fentanyl; Degree of antagonism (Imax) normalized to naltrexone. c DOR Antagonist potency (IC50) determined versus EC50 of SNC80; degree of antagonism (Imax) normalized to naltrexone. d KOR Antagonist potency (IC50) determined versus EC90 of U50488H; degree of antagonism (Imax) normalized to nor-BNI. e Structure previously published [11]. Data from re-assay. 55  1.1 ± 0.2 (107 ± 5%) a Inhibition of forskolin-induced cAMP accumulation; cAMP Hunter TM Chinese hamster ovary cells (CHO-K1) that express human µ-opioid receptor (OPRM1), human κ-opioid receptor (OPRK1), and human δ-opioid receptor (OPRD1) were used for the forskolin-induced cAMP accumulation assay in order to determine potency and efficacy of the compounds following the previously established methods; [17] to determine % efficacy in forskolin-induced cAMP assays, data were blank subtracted with the vehicle control, followed by normalization to the forskolin control. Data were then analyzed in GraphPad Prism 8 (GraphPad, LaJolla, CA, USA) using nonlinear regression; values are expressed as the mean ± SEM of at least three independent experiments; N/D = not determined. b MOR Antagonist potency (IC 50 ) determined versus EC90 of fentanyl; Degree of antagonism (I max ) normalized to naltrexone. c DOR Antagonist potency (IC 50 ) determined versus EC50 of SNC80; degree of antagonism (I max ) normalized to naltrexone. d KOR Antagonist potency (IC 50 ) determined versus EC90 of U50488H; degree of antagonism (Imax) normalized to nor-BNI. e Structure previously published [11]. Data from re-assay.
The enantiomer of 48, compound 56, was found to be a potent sub-nanomolar partial agonist (EC 50 = 0.8 nM, %E max = 39%), and the diastereomer of 56, compound 51, was a potent full agonist (EC 50 = 1.89 nM, %E max = 95.7%). As agonists, compounds 56 and 51 were found to be highly selective at MOR (EC 50 > 10,000 nM for DOR and KOR). Whereas the reference MOR agonist morphine was less potent (EC 50 = 5.8 nM) than the three compounds 48, 56, and 51, and was a partial agonist at DOR (EC 50 = 525 nM, E max = 75%) and KOR (EC 50 = 346 nM, E max = 86%). The remaining, previously described diastereomer 43 was the least potent partial C9-methyl agonist (EC 50 = 16.1 nM, %E max = 78%). These data exemplified our findings with diastereomers at C9 in the 5-phenylmorphan series and could have very different properties in vitro, and their activity was not predictable in silico. We previously determined that agonists and antagonist C9-substituted phenylmorphans could be distinguished using molecular modeling [14], but that partial and full agonists could not be differentiated.

General Information
Melting points were determined on a Mettler Toledo MP70 and are uncorrected. Proton and carbon nuclear magnetic resonance ( 1 H and 13 C NMR) spectra (Figures S1-S12 in Supplementary Material) were recorded on a Varian Gemini-400 spectrometer in CDCl 3 (unless otherwise noted) with the values given in ppm (TMS as internal standard) and J (Hz) assignments of 1 H resonance coupling. The analyses were performed on the free base, unless otherwise noted. Mass spectra (HRMS) were recorded on a Waters (Mitford, MA, USA) Xevo-G X5 QTof. The optical rotation data were obtained on a PerkinElmer polarimeter model 341. Thin layer chromatography (TLC) analyses were carried out on Analtech silica gel GHLF 0.25 mm plates using various gradients of CHCl 3 /MeOH containing 1% NH 4 OH or gradients of EtOAc/n-hexane. Visualization was accomplished under UV light or by staining in an iodine chamber. Flash column chromatography was performed with Fluka silica gel 60 (mesh 220-400). Flash column chromatography was performed using RediSep Rf normal phase silica gel cartridges. Robertson Microlit Laboratories, Ledgewood, N.J., performed elemental analyses, and the results were within ±0.4% of the theoretical values. Experimental procedures for intermediates, and their characterization, can be found in a previous publication [14]. Compounds 16, 43, and 51 in Table 2 were discussed in previous publications [11,12].

Synthesis
General procedure for hydrogenation of C9-alkene to C9-alkane. The HCl salt of the alkene was dissolved in MeOH (10 mL) and was transferred to a Parr shaker degassed with argon. The solution was treated with palladium on carbon (Escat 103), and the Parr shaker was pressurized to 20 psi with H 2 and was shaken at 23 • C for 16 h. Upon completion, the reaction mixture was filtered through Celite to remove the palladium and was washed several times with methanol. The filtrate was washed with saturated NaHCO 3 , several times with CHCl 3 , and the filtrate was concentrated in vacuo. The resulting oil was taken up in MeOH (3.0 mL) and treated with 3 M HCl (29 mL). The reaction mixture was stirred at reflux for 16 h and subsequently quenched with 7 N NH 4 OH in MeOH, extracted with CHCl 3 and concentrated in vacuo. To the crude solution was added dry dichloromethane (12 mL) at 0 • C, di-tert-butyl dicarbonate (2.9 g, 1.4 equiv, 13 mmol), triethylamine (1.9 mL, 1.4 equiv, 13 mmol), and 4-dimethylaminopyridine (0.12 g, 0.1 equiv, 0.96 mmol) dropwise. The solution was allowed to stir under argon for 30 min at 0 • C then warmed to room temperature. After 1 h, TLC showed consumption of starting material. Saturated ammonium chloride was added, and the mixture was extracted with CH 2 Cl 2 , washed with brine, and dried over sodium sulfate. The crude mixture was loaded onto silica and purified via flash chromatography eluting with 0-30% ethyl acetate in hexane.
tert-Butyl (1S,5S)-5-(3-methoxyphenyl)-9-methylene-2-azabicyclo[3.3.1]nonane-2-ca rboxylate (45). To a vial containing ketone 44 (1.85 g, 1 equiv, 5.36 mmol) in THF (25.0 mL) at 0 • C was added Tebbe's Reagent (10.7 mL, 1 equiv, 5.36 mmol). The resulting mixture was stirred at 0 • C for 1 h and then slowly warmed up to room temperature for an additional 4 h. Upon completion by TLC, the reaction was cooled to 0 • C and 50 mL of Et 2 O was added. The reaction was quenched carefully with 1.8 N NaOH. A very vigorous gas evolution took place and a thick red/orange precipitate formed. Magnesium sulfate was added, and the mixture was allowed to stir an additional 5 min. The solids were filtered and washed with EtOAc. Flash column chromatography using 1-5% CMA in CHCl 3 yielded 45 as an orange foam (0.81 g, 44% yield).
(1S,5R,9R)-5-(3-Methoxyphenyl)-9-methyl-2-phenethyl-2-azabicyclo[3.3.1]nonane (47). To a solution of 46 (1 g, 1 equiv, 3 mmol) in dichloromethane (30 mL) at 0 • C was added trifluoroacetic acid (2 mL, 10 equiv, 0.03 mol), dropwise. After 15 min, the reaction was allowed to warm to room temperature and allowed to stir for 1 h. Upon completion as determined using TLC, the reaction mixture was cooled to 0 • C and quenched with NaHCO 3 and extracted with dichloromethane. The crude oil was taken up in acetonitrile (32 mL), was treated with K 2 CO 3 , and the mixture was purged with argon. (2-Bromoethyl)benzene (0.8 g, 1.5 equiv, 4 mmol) was added, and the reaction was refluxed under argon for 16 h. Upon completion, the reaction mixture was concentrated in vacuo then extracted with CHCl 3 . The organic extracts were washed with water and brine, dried with sodium sulfate and concentrated in vacuo. Purification by flash column chromatography on silica using 0-100% EtOAc: Hexanes to isolate 47 as a yellow oil (830 mg, 80% yield). The resulting oil was a mixture of epimers and was used without further purification.
3-((1S,5R,9R)-9-Methyl-2-phenethyl-2-azabicyclo[3.3.1]nonan-5-yl)phenol (48). In an oven-dried round-bottom flask, 47 (600 mg, 1 equiv, 2.9 mmol) was suspended in dichlo romethane (30 mL) and the mixture was cooled to −78 • C. Tribromoborane (0.3 mL, 2 equiv, 5.8 mmol) was added dropwise and the reaction was stirred at −78 • C. The reaction mixture was allowed to warm to room temperature and stirred 2 h. Upon completion, the reaction mixture was cooled to 0 • C and quenched with 15 mL MeOH dropwise and stirred for 30 min. A total of 20 mL 1N HCl was added, and the reaction mixture was refluxed at 100 • C for 1 h. The reaction mixture was then cooled to 0 • C and made basic (pH > 10.5) with NH 4 OH and extracted with 9:1 CHCl 3 :MeOH. The combined organic layers were washed with water and brine, dried with sodium sulfate, and concentrated in vacuo. Purification by silica gel flash chromatography using 10-100% EtOAc Hexanes resulted in a tan foam (420 mg, 73%). The HCl salt of 48 was formed in iPrOH with 37% HCl (0.1 mL) and recrystallized from ethanol to give a white solid: mp 264-269 • C. 1