Recent Syntheses of Frog Alkaloid Epibatidine

Abstract

Many natives from Amazon use poison secreted by the skin of some colorful frogs (Dendrobatidae) on the tips of their arrows to hunt. This habit has generated interest in the isolation of these toxins. Among the over 500 isolated alkaloids, the most important is undoubtedly (-)-epibatidine. First isolated in 1992, by Daly from Epipedobates tricolor, this compound is highly toxic (LD₅₀ about 0.4 µg per mouse). Most remarkably, its non-opioid analgesic activity was found to be about 200 times stronger than morphine. Due to its scarcity, the limited availability of natural sources, and its intriguing biological activity, more than 100 synthetic routes have been developed since the epibatidine structure was assigned. This review presents the recent formal and total syntheses of epibatidine since the excellent review published in 2002 by Olivo et al.¹ Mainly, this review is summarized by the method used to obtain the azabicyclic core.

epibatidine; organic synthesis; azanorbornanes

1. Introduction

At an expedition to Western Ecuador in 1974, Daly and Myers isolated traces of an alkaloid with potential biological activity from the skin of the speciesEpipedobastes tricolor. Twenty years later, 750 frogs collected near a cocoa plantation, led to a total of 60 mg of a complex mixture of alkaloids. From this mixture, 500 µg of relatively pure (-)-epibatidine 1 was isolated. Tests with mice showed the analgesic activity of this compound to be 200 times stronger than that of morphine.²

The mechanism of action of epibatidine was only elucidated after the determination of the molecular structure by Daly in 1992.³Biological assays confirmed the analgesic activity by non-opioid mechanisms⁴ through the nicotinic receptor antagonist acetylcholine (nAChR)⁵ with analgesic activity at doses of 0.01 µmol kg^-1. Higher doses are considered highly toxic.

To date, many epibatidine analogs are synthesized and investigated as potential drugs for the treatment of diseases, such as Alzheimer's disease (AD).⁴ Epiboxidine (Figure 1), for example, was shown to be more selective to ganglionic nicotinic receptors and much less toxic than epibatidine.⁶

Figure 1
Structures of (-)-epibatidine and epiboxidine

The most recent review about epibatidine synthesis was published by Olivo et al.,¹ who synthesized epibatidine in a chemoenzymatic fashion.⁷ Since then, fewer reports about epibatidine synthesis have appeared (Figure 2). However, the number of articles is still impressive, and therefore, an update seems appropriate.

Figure 2
Number of articles about epibatidine found at Web of Science® (*until July, 2014).

This review attempts to cover epibatidine synthesis research (total and formal synthesis) since 2001. Despite the few available synthetic approaches for the construction of the azabicyclo ring (cycloaddition reactions, intramolecular nucleophilic substitution reactions, intramolecular SN, and rearrangement reaction) (Figure 3), this review presents new and modern strategies efficiently employed to obtain the cyclohexylamine core or to incorporate the 2-chloropyridyl moiety, including ring-closing metathesis, Suzuki/Negishi coupling reactions, and rearrangements. Within each of these categories, the content in this review will be presented in chronological order.

Figure 3
Recent synthetic approaches to form the azabicyclo ring during the synthesis of epibatidine

2. Cycloaddition Reactions

An efficient and short synthesis of optically pure (+)-N-Boc-azabicyclo[2.2.1]hept-2-one, (tert-butyloxycarbonyl, Boc), 10 was reported by Pandey (Scheme 1).⁸ The method involves the cycloaddition of ethynyl phenyl sulfone (3) with N-methoxycarbonyl pyrrole (2),⁹ followed by the introduction of the second phenyl sulfone group to the cycloadduct 4 by β-metalation. The desymmetrization of the bisphenylsulfonyl byciclic compound 5 was carried out by stirring with the disodium salt of meso-hydrobenzoin, giving 6 as a single diastereomer, indicated by nuclear magnetic resonance (NMR) and high performance liquid chromatography (HPLC) analysis. The reductive elimination of the phenyl sulfone group using Na-Hg amalgam gave the acetal 7. Simple acetal hydrolysis would lead to the corresponding N-acetylazabicyclo[2.2.1] hept-2-one, but the removal of the chiral acetal moiety was possible only by changing the acetyl protecting group to the t-butylacetyl group. Thus, the N-CO₂Me bond was cleaved with trimethylsilyl chloride (TMSCl), and the amine 8 was reprotected with di-tert-butyl dicarbonate (Boc₂O), giving the N-Boc compound 9. The ketone 10 was obtained by catalytic hydrogenation with Pd/C. The specific rotation ([a]^D) of ketone 10 corresponds to that reported by Fletcher and Trudell¹⁰ in their total synthesis of (-)-epibatidine.

Scheme 1
(a) 85-90 °C; (b) n-BuLi, -78 °C, tetrahydrofuran (THF), PhSO₂F, rt (43%, 2 steps); (c) meso-hydrobenzoin, NaH, THF, 0 °C → rt (85%); (d) 6% Na/Hg, NaH₂PO₄.H₂O, 0 °C (95%); (e) TMSCl, NaI, CH₃CN, rt; (f) Boc₂O, Et₃N, CH₂Cl₂, rt (95%, 2 steps); (g) Pd/C, H₂, 55 psi, EtOH/EtOAc, rt (90%) (adapted from reference 8).

In 1998, Pandey's group reported a [3+2]-cycloaddition strategy in the racemic synthesis of epibatidine (Scheme 2).¹¹ Years later, the same group reported an asymmetric approach using Oppolzer's sultam as a chiral auxiliary.¹² The key step of cycloaddition involves the reaction of the N-alkyl-bis(trimethylsilyl) cyclic amine 14 with the dipolarophile (-)-17 using AgF as a one-electron oxidant. The bis-silylated amine was prepared by double lithiation of pyrrolidine 11. Heck-coupling reaction between 2-chloro-5-iodopyridine 16 and chiral camphorsultam derivative15 gave the desired dipolarophile (-)-17. Surprisingly, the cycloaddition preferably gave the 2-chloropyridyl moiety in exo-stereochemistry, differently from the results obtained with model compounds. The authors did not mention any reasonable explanation for this favorable inversion of the exo/endo ratio. Careful separation of 18 followed by removal of the chiral auxiliary and methylation of the acid19 gave the compound 20. Starting from this compound, the synthesis of (±)-epibatidine was reported.¹¹

Scheme 2
(a) Boc-N₃, Et₃N, dioxane, rt (90%); (b) s-BuLi, TMSCl, TMEDA, -78 °C (90%); (c) s-BuLi, TMSCl, TMEDA, -50 → -30 °C (68%); (d) TFA, CH₂Cl₂, rt (quant.); (e) PhCH₂Cl, K₂CO₃, CH₃CN, rt (80%); (f) K₂CO₃, Pd(OAc)₂, PPh₃, CH₃CN, reflux (85%); (g) AgF, CH₂Cl₂ (58%); (h) LiOH.H₂O, THF:H₂O, 35 °C; (i) SOCl₂, MeOH, 0 °C → rt (90%, 2 steps) (adapted from reference 12).

The formal synthesis of (-)-epibatidine reported by Vogel¹³ uses the "naked aza-sugar" methodology for an efficient resolution of a racemic ketone by amination using (R, R)-1,2-diphenylethylenediamine (Scheme 3). Starting with the Diels-Alder adduct (±)-23,¹⁴ the tosyl group was removed with SmI₂, giving the ketone (±)-24. The resolution of (±)-24 gave a separable mixture of compounds (+)-25and (+)-26, which gives back the ketones (+)-24 and (-)-24 in optically active forms after acid hydrolysis. Reduction of the double bond led to (-)-10 and (+)-10 in quantitative yield. The total synthesis of (-)-1 and (+)-1 from compound 10 was reported by Fletcher and Trudell.¹⁰

Scheme 3
(a) SmI₂, THF/MeOH, -78 °C (75%); (b) (R,R)-1,2-diphenylethylenediamine, 4Å molecular sieves, CH₂Cl₂, room temperature (rt) (42% for (+)-25 and 43% for (+)-26); (c) H₃PO₄/THF, 20 °C (95% for (+)-24and 96% for (-)-24); (d) 10% Pd/C, H₂, MeOH, rt (> 99%) (adapted from reference 13).

Node and co-workers¹⁵ reported a second-generation route for the formal synthesis of (-)-epibatidine (Scheme 4). Diels-Alder reaction between the chiral dienophile di-(l)-menthyl (R)-allene-1,3-dicarboxylate 27 and the N-Boc-pyrrole 21 gives the endo-adduct 28 as the sole product. The authors mention the high endoexo selectivity could be attributed to the steric repulsion between the Boc group of 21 and the l0-menthyl group in the dienophile 27. Consecutive reductions of the bicyclic double bond and the menthyl diester give the diol 30. Ozonolysis of the remaining double bond generates the β-keto alcohol 31. The alcohol was oxidized by Jones oxidation to carboxylic acid 32, which was decarboxylated with reflux in toluene, giving the (-)-N-Boc-7-azabicyclo[2.2.1]heptan-2-one (10). Again, the total synthesis of 1 from compound 10 was already reported by Fletcher and Trudell.¹⁰

Scheme 4
(a) AlCl₃, CH₂Cl₂, -78 °C (88%); (b) Pd-C, H₂, EtOAc, rt (99%); (c) LiAlH₄, THF, rt (90%); (d) O₃, CH₂Cl₂, -78 °C, Me₂S, rt (93%); (e) Jones reagent, acetone, 0 °C; (f) toluene, reflux (54%, 2 steps) (adapted from reference 15).

3. Intramolecular SN-Type 1

An asymmetric exo-selective hetero Diels-Alder reaction mediated by Lewis acid and an unusual ring-opening fragmentation were the key steps for the asymmetric synthesis of (-)-epibatidine reported by Evans (Scheme 5).¹⁶ Dienophile 35 was formed by a Horner-Wadsworth-Emmons reaction of aldehyde 33 and phosphonate containing the chiral auxiliary 34. The heterodiene 36 was synthesized by treatment of glutarimide with triethylsilyl triflate under basic conditions. Evidenced by NMR, the exo selectivity of the cycloaddition reaction was rationalized based on steric interactions between the triethylsilyloxy substituents of 36 and the Lewis acid-acyl oxazolidinone complex. Removal of the chiral auxiliary by samarium(III) trifluoromethanesulfonate [Sm(OTf)₃] gave the methyl ester 38. Cleavage of the C1-N bond was accomplished by O-acylation with Boc₂O followed by treatment with tetrabutylammonium fluoride (TBAF) to produce the nitrile 40. Krapcho decarboxylation of 40 provided the ketone 41. The required transposition of the nitrogen atom was accomplished by the conversion of nitrile 41 to amide 42, followed by Hofmann rearrangement to afford the amine 43. Several reducing agents were tested for the stereoselective reduction of the ketone 43 to the desired axial alcohol for SN₂ displacement. However, all reactions gave a mixture of inseparable alcohols with equatorial alcohol as a major isomer. Therefore, to complete the synthesis, the alcohol 44 was converted into the corresponding mesylate 45, and subsequent SN₂ displacement with LiBr provided the bromide 46. Removal of the Boc group with trifluoroacetic acid (TFA) followed by reflux of amine 47 for 3 days afforded (-)-epibatidine in 13 steps and 13% overall yield.

Scheme 5
(a) LiCl, i-Pr₂EtN, CH₃CN, rt (81%); (b) Me₂AlCl, CH₂Cl₂, -78 °C (79%); (c) Sm(OTf)3, MeOH, reflux (84%); (d) Boc₂O, p-dimethylaminopyridine (DMAP), Et₃N, CH₂Cl₂, rt (98%); (e) n-Bu₄NF, THF/H₂O, rt (81%); (f) dimethyl sulfoxide (DMSO), H₂O, 130 °C (99%); (g) Me₃SiOK, toluene, 70 °C (72%); (h) Pd(OAc)₄, tert-butyl alcohol, 50 °C (70%); (i) NaBH₄, MeOH, -40 °C, (89%, 92% enantiomeric excess (ee)); (j) MsCl, Et₃N, CH₂Cl₂, rt (92%); (k) LiBr, THF, 50 °C (84%); (l) TFA, CH₂Cl₂, rt (91%); (m) CHCl₃, reflux (95%) (adapted from reference 16).

A practical 13-step process for the synthesis of (-)-epibatidine was reported by Loh (Scheme 6).¹⁷ The entire synthetic route is straightforward and convenient for gram-scale synthesis. The synthesis commenced from the reduction of commercially available chloro methylnicotinate 48 to the aldehyde derivative 50 in two steps, followed by Grignard addition with a vinyl group to provide the allylic alcohol 51. Bromination of 51 provided the desired terminal bromide 52 necessary for another allylic rearrangement under Barbier conditions. The "aza" Barbier reaction between 52 and the imine 55 (carrying a chiral auxiliary) provided only one single isomer, the diene 56. Using Grubbs 2^ndgeneration catalyst of this diene provided the desired product 57. For the intramolecular cyclization, the alkene 57 was brominated, and two isomers were obtained. Single X-ray structure confirmed that the minor isomer was the desired one, and the authors had to recycle by converting back the major isomer to the alkene 57. Deprotection of the chiral auxiliary group provided the amine 60, which underwent the intramolecular cyclization by heating, affording the azabicyclo compound 61. The completion of the synthesis involved radical debromination followed by epimerization of the endo-epibatidine.

Scheme 6
(a) NaBH₄, THF:MeOH (3:1), 0 °C (99%); (b) (COCl)₂, DMSO, Et₃N, CH₂Cl₂, -78 °C (99%); (c) vinyl magnesium bromide, THF, 0 °C (96%); (d) PBr₃, Et₂O, 0 °C (98%); (e) NaSO₄, CH₂Cl₂, 0 °C (99%); (f) Zn, THF, 0 °C (93%);. (g) 10% 62, CH₂Cl₂, 25 °C (94%); (h) Br₂, Et₄N+Br-, CH₂Cl₂, -78 °C (92%, dr 66:34); (i) diisobutylaluminium hydride (DIBAL-H), CH₂Cl₂, 0 °C (88%); (j) Pb(OAc)₄, CH₂Cl₂:MeOH (2:1), 0 °C (65%); (k) CH₃CN, 82 °C (85%); (l) Bu₃SnH, azobisisobutyronitrile (AIBN), benzene, reflux (99%); (m) t-BuOK, t-BuOH, reflux (58%) (adapted from reference 17).

Takemoto's research group developed a protocol to synthesize chiral 4-nitrocyclohexanones through an enantioselective tandem Michael addition between nitroalkenes and g,d-unsaturated b-ketoesters catalyzed by chiral thiourea 74.¹⁸ This protocol was applied to synthesize (-)-epibatidine in 10 steps at 19% overall yield (Scheme 7). The route commenced by treatment of enone 63with lithium bis(trimethylsilyl)amide (LHMDS), followed by the addition of allylcyanoformate to obtain 64. The Michael addition between 64 and 65 in the presence of chiral thiourea 74 gave 66 in lower enantioselectivity (75% ee) compared to the model compounds. The second Michael addition occurred with KOH in ethanol and provided the cyclic ketoester 67. Decarboxylation of 67 under Tsuji conditions gave 68 in quantitative yield. The remaining tasks were the stereoselective reduction of carbonyl group and the configuration inversion of the nitro group. For this purpose, the axial methoxy group assisted the stereoselective reduction of the ketone 68 by lithium trisec-butylborohydride (L-selectride), and the alcohol 69 was giving. After several attempts, the authors found that the best result to remove the OMe group was the treatment of 69 with NaOMe in t-butanol. At this stage, the ee was improved by recrystallization of compound 70. Finally, 1,4-hydride reduction of 70 with NaBH₃CN followed by mesylation of the remaining alcohol afforded the axial nitroalkane 72 as the major product. Successive treatment of 72 with zinc dust and refluxing in CHCl₃ gave (-)-epibatidine.

Scheme 7
(a) LiN(SiMe₃)₂, THF, -78 °C, vinyl 2-cyanoacetate, hexamethylphosphoramide, THF, rt (73%); (b) 10% (R,R)-74, toluene, 0 °C (90%, 75% ee); (c) KOH, EtOH, 0 °C (85%, 75% ee); (d) Pd(OAc)₂, Ph₃P, HCO₂H, Et₃N, THF, rt (99%); (e) L-selectride^®, THF, -78 °C → 0 °C (71%); (f) NaOMe, t-BuOH, rt (71%); (g) NaBH₃CN, CH₃CO₂H, MeOH, -20 °C (87%, 90% dr); (h) MsCl, Et₃N, DMAP, CH₂Cl₂, 0 °C (91%); (i) CH₃CO₂H, Zn, THF, rt; (j) CHCl₃, 60 °C (85%, 2 steps) (adapted from reference 18).

Stevenson and co-workers¹⁹ described an excellent route for the total synthesis of (-)-epibatidine through enzymatic cis-dihydroxylation of bromobenzene using the microorganism Pseudomonas putida UV (Scheme 8).^19,20 The cis-dihydrodiol (+)-76 was giving with > 98% ee, and its chemoselective hydrogenation produced the tetrahydrodiol (-)-77 with 90% yield. The hydroxyl group in C-2 prefers to occupy a pseudo-axial position to minimize allylic 1,2-strain with the substituent on the alkene, while the hydroxyl group in C-1 is equatorial. The silylation of (-)-77 was chemoselective with the hydroxyl group in the equatorial position with 86% yield. The alcohol (-)-78 was converted to urethane 79 using Ichikawa's conditions, with 90% yield. The urethane 80 was dehydrated, providing the intermediate cyanate 80, which underwent a rapid [3,3]-sigmatropic rearrangement under the reaction conditions to furnish the intermediate isocyanate 81. This intermediate was protected with N-protective group Boc by the Bruckner protocol with 76% yield for three steps in a one pot reaction.

Scheme 8
(a) Pseudomonas putida UV4 (> 98% ee); (b) H₂ (1,5 bar), 5% Rh-graphite catalyst, MeOH, rt (90%); (c) TBSCl, Py, rt (86%); (d) trichloroacetyl isocyanate, CH₂Cl₂, 0 °C, Na₂CO₃, Et₂O:MeOH:H₂O (4:4:1:), rt (90%); (e) Ph₃P, Et₃N, CBr₄, CH₂Cl₂, -10 °C; (f) [3,3] sigmatropic rearrangement; (g) 1% MoCl₂O₂, CH₂Cl₂, t-BuOH, rt (76%, 3 steps); (h) 6-chloropyridin-3-ylboronic acid, 2% 85, PhMe:EtOH:H₂O (1:1:1), Na₂CO₃, 100 °C (93%); (i) TBAF, THF (95%); (j) PtO₂, H₂ (1 bar), EtOH, rt (60%); (k) Et₃N, MsCl, CH₂Cl₂, 0 °C (91%); (l) LiBr, THF, 60 °C (78%); (m) TFA, CH₂Cl₂, rt (94%); (n) CHCl₃, 55 °C (95%) (adapted from reference 19).

The alkenyl bromide (+)-82 was subjected to Suzuki cross coupling with commercially available 6-chloropyridin-3-yl boronic acid, giving the compound (+)-83 with 93% yield. Deprotection of O-silyl was induced by fluoride, giving the alcohol (+)-84 with 95% yield. The next step involved a stereoselective catalytic reduction with the addition of the hydrogens in anti in relation to the hydroxyl and the amine, giving the compound (-)-44 with 60% yield. The alcohol (µ)-44 was made by Evan's protocol, giving the bromide (-)-46 with 91% yield for mesylation and 78% yield for bromination. The compound (-)-46 was deprotected with TFA, giving the amine (-)-47 with 94% yield, which was then cyclized (95% yield), giving the (-)-epibatidine with an overall yield of 18% and 14 steps.

Huang et al.²¹ described a fast and practical synthetic route involving seven steps for diastereoselective synthesis for (±)-endo-1, leaving the 6-choro-3-pyridinecarboxaldeide 86 (Scheme 9). Henry's reaction catalyzed by KF generates a β-nitro alcohol, which was treated with DMAP and acetic anhydride, giving the precursor 88 by means of desidratation with 76% yield for two steps. The next step was using the intramolecular Michael's reactions between the ketone and the pyridinylnitroolefin catalyzed with L-proline, giving the key intermediated 90 with 81% yield and dr ≥ 97%. The compound 90 was asymmetrically reduced by (4R,5R)-2-phenyl-1,3,2-dioxaborolane-4,5-dicarboxylic acid - 91 (TarB-H) and NaBH₄, building three chiral centers with 95% yield for compound 71. The alcohol giving was mesylated with 94% yield. The group nitro was reduced to amine with Zn/H⁺. The compound was refluxed with chloroform, giving (±)-endo-1 with 78% yield for two steps. The conversion of compound (±)-endo-1 to (±)-epibatidine was reported using t-BuOK/t-BuOH under reflux.^17,22

Scheme 9
(a) KF.H₂O, i-PrOH/THF, rt; (b) DMAP, Ac₂O, CH₂Cl₂, rt (76%, 2 steps); (c) L-proline, 2,5-dimethoxybenzoic acid, DMSO, rt (81%, dr > 97%); (d) 91, NaBH₄, THF, 0 °C (95%); (e) MsCl, Et₃N, CH₂Cl₂, 0 °C (94%); (f) Zn, CH₃CO₂H/THF, rt; (g) CHCl₃, reflux (78%, 2 steps) (adapted from reference 21).

Jorgensen and co-workers²³ developed a diastereoisomeric and enantioselective intramolecular Michael addition using a chiral organocatalyst for formal synthesis of (-)-epibatidine (Scheme 10). The Wittig reaction of ylide 92 and aldehyde 86 gave the γ-nitroenone 93 with 98% yield. The intramolecular cyclization by Michael addition using 10 mol% catalyst and hindered protic solvent gave the trans-4-nitrocyclohexanone 90 with 80% yield (96% ee, 14:1 dr). The compound 90 has been reduced with NaBH₄ to the 4-nitrocyclohexanone 71 with 87% yield (10:1 dr). The total synthesis of (-)-epibatidine from compound 71 was reported by Szántay,²⁴ involving over 4 steps with 22% yield.

Scheme 10
(a) 1,2-dicloroethane, 50 °C (98%); (b) 94(2×(S)-N-BocPhGly), t-BuOH, 75 °C (88%, 14:1 dr, 96% ee); (c) NaBH₄, MeOH, 0 °C (87%, 10:1 dr) (adapted from reference 23).

Blakemore and co-workers²⁵ investigated the functionalizing of α-chloroalkylllithiums, generating in situthe sulfoxide-lithium for stereospecific reagent-controlled homologation with available 6-chloropyridin-3-yl boronic acid for the formal synthesis of (-)-epibatidine,1 (Scheme 11). The reaction of p-thiocresol (95) with acrolein generates the aldehyde 96 with 95% yield, which was treated with ethylene glycol, generating the thioether 99. The two enantiomers of syn-chlorosulfoxide 99 were prepared by Ellamn-Bolm enantioselective oxidation using the appropriate enantiomer of Jackson's tert-leucinol derived ligand 109, giving 80% yield (> 98% ee) for the sulfoxidation and non-racemizung chlorination with 86% yield (95:5 dr).

Scheme 11
(a) Acrolein, Et₃N, CH₂Cl₂, 0 °C; (b) ethylene glycol, p-TsOH·H₂O, benzene, reflux (95%); (c) 109, [VO(acac)₂], aq. H₂O₂, CHCl₃, 0 °C (80%, > 98% ee); (d) NCS, K₂CO₃, CH₂Cl₂, 0 °C (86%, 95:5 dr); (e) NaHMDS, CD₃OD, THF, -78 °C (85%, 86% dr); (f) (i) 6-chloro-3-pyridineboronic acid pinacol ester, (S)-anti-D-100, PhLi, THF, -78 °C → rt; (ii) (R)-anti-D-100, PhLi, THF, -78 °C → rt, then KOOH (49%, 97% ee, 79% dr); (g) MsCl, Et₃N, CH₂Cl₂, 0 °C; (h) NaN₃, DMF, 80 °C; (i) TsOH, MeOH, reflux (52%, 3 steps); (j) aq. TFA, CHCl3, rt; (k) Ph3PCH3I, n-BuLi, THF, -10 °C → reflux; (l) (CF₃CO₂)₂O, Et₃N, rt (12%, 3 steps); (m) 10% 110, CH₂Cl₂, reflux (85%) (adapted from reference 25).

The epimerization of syn-chlorosulfoxide was conducted by sodium bis(trimethylsilyl)amide (NaHMDS) and [D₄]methanol, giving the anti-D-chlorosulfoxide 100 with 85% yield and 86% dr. There was a double one-pot reaction between the 6-chloropyridin-3-yl boronic acid and sulfoxide-lithium (generated from the anti-D-100in situ), which led to two reactional cycles, the first using the enantiomer (S)-anti-D-100 and the second using (R)-anti-D-100. The alcohol (R,S)-D,D-101 was obtained with 49% yield (97% ee and 79% dr).

The alcohol 101 was mesylated and substituted by azide group, and acetal was hydrolysed, generating the compound 104 with 52% yield for three steps. The compound 104 was converted to dialdehyde 105, which was used without purification and converted to dienyl amide 107 by exposure to methylenetriphenylphosphorane (generated in situ with n-BuLi and PPh₃CH₃I), giving 12% yield for two steps. The compound 107 was cyclized by Grubbs' first-generation catalyst, giving the compound 108 with 85% yield. The total synthesis of (-)-epibatidine from compound 108 was reported by Corey.²⁶

4. Intramolecular SN-Type 2

In 1999, Maycock and co-workers²⁷ reported a formal synthesis of (+)-epibatidine starting from (-)-quinic acid (Scheme 12).^27,28 Analogously, the same authors revealed years later that the same precursor (quinic acid) can be used for the formal synthesis of the corresponding levorotary epibatidine by the synthesis of the enantiopure R-4-hydroxy-2-cyclohexen-1-one (116).²⁹

Scheme 12
(a) (i) K-Slectride^®, THF, -78 °C; (ii) aq. 0.5 mol L_-1 NaOH, THF, 0 °C; (b) tert-butyldimethylsilyl chloride (TBDMSCl), (i-Pr)₂NEt, DMAP, CH₂Cl₂, 0 °C → rt (51%, 3 steps); (c) TFA, H₂O, CH₂Cl₂, reflux (85%); (d) TBDMSCl, (i-Pr)₂NEt, DMAP, CH₂Cl₂, 0 °C → rt (98%) (adapted from reference 29).

Avenoza et al.^30,31 described a synthetic route to 7-azabicyclo[2.2.1]heptane containing a carboxylic acid group at the bridgehead carbon atom (Scheme 13). The synthesis begins with preparation of the dienophile 120 from racemic serine through esterification followed by double benzoylation in the presence of base. Diels-Alder reaction between 120 and Danishefsky's diene gave a mixture of cycloadducts, which was treated under dilute HCl and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) to afford the enone 125. Hydrogenation of this product followed by reduction with L-selectride and mesylation gave the compound 127. Internal nucleophilic displacement with t-BuOK gave the desired compound 128 with the 7-azabicyclo[2.2.1]heptane core. Selective hydrolysis of the ester gave 129, which was used to prepare the Barton ester derivative 130. Simple decarboxylation of the Barton ester was achieved using tin hydride as a source of hydrogen radical, yielding the N-benzoyl-7-azabicyclo[2.2.1]heptane 131, the precursor for Olivo's synthesis of epibatidine.³²

Scheme 13
(a) HCl, MeOH, 0 °C → reflux (> 99%); (b) BzCl, Et₃N, CH₂Cl, rt, DBU, MeOH, rt (94%); (c) 2-trimethylsiloxy-1,3-butadiene, ZnI2, CH₂Cl₂, reflux; (d) aq. HCl, THF, rt (94%, 2 steps); (e) 132, toluene, reflux; (f) aq. HCl, THF, rt, DBU, MeOH, 5 °C (55%, 2 steps); (g) H₂, Pd/C, CH₂Cl₂, rt (95%); (h) L-selectride^®, THF, -78 °C; (i) MsCl, N,N-diisopropylethylamine (DIPEA), CH₂Cl₂, rt (68%, 2 steps); (j) t-BuOK, THF, rt (81%); (k) LiOH. H₂O, MeOH:H₂O, rt (99%); (l) (i) (COCl)₂, DMF, CH₃CH₂Cl₂; (ii) N-hydroxypyridine-2-thione, Et₃N, THF, 0 °C; (m) Bu₃SnH, THF, light, rt (61%, 3 steps) (adapted from reference 31).

Aggarwal and Olofsson³³ described a good methodology of direct asymmetric arylation for the cicloexanone 139, applying it for the total synthesis of (-)-epibatidine (Scheme 14). The electrophile pyridyl iodonium salt (135) was synthetized in two steps: obtaining the compound 134 from acetylene gas with 21% yield, which was reacted with 2-chloro-5-bromopyridine, generating the electrophile 135 with approximately 70% yield.

Scheme 14
(a) ICl₃, HCl, H₂O, 0 °C → rt (21%); (b) 2-chloro-5-bromopyridine, BuLi, Et₂O, -78 °C → rt (ca. 70%); (c) Boc₂O, DMAP, THF, reflux (> 99%); (d) (i) THF, -118 °C; (ii) DMF, -45 °C (41%, > 20:1 dr, 94% ee, 2 steps); (e) NaBH₄, MeOH, -98 °C (83%, 10:1 dr); (f) MsCl, Et₃N, CH₂Cl₂, rt; (g) TFA, CH₂Cl₂, rt; (h) CHCl₃, reflux (90%, 3 steps) (adapted from reference 33).

The carbamate 136 was protected with 100% yield. The ketone 137 was transformed in 2-pyridyl ketone 139 with 41% yield (> 20:1 dr). The base 138 was used to generate the enolate, favoring the coupling between 137 and 135. The 2-pyridyl ketone was reduced to alcohol 140 with 83% yield (10:1 dr). The alcohol was mesylated, giving the compound 141. This was deprotected, giving the amine 73, which was cyclized, giving the (-)-epibatidine with 90% yield for three steps.

Gómez-Sánchez and Marco-Contelles³⁴ reported a direct base-catalyzed heterocyclization of N-Boc protected 3,4-dibromocyclohex-1-yl amine (144) as a key-step for the synthesis of the azabicycloheptane core (Scheme 15). Starting from cyclohex-3-enecarboxylic acid 142, the N-Boc cyclohexene derivative 143 was obtained by Curtius rearrangement by diphenylphosphoryl azide (DPPA) under reflux of toluene. Bromination of the alkene gave the dibromo compound 144 in cis configuration as a major product. The authors found that NaH in dimethylformamide (DMF) is the best condition for the intramolecular cyclization of 144, affording the known compound 145, which gave the azabicycloheptene 146 after reaction with potassium t-butoxide. This azanorbornene is a known intermediate for the synthesis of epibatidine.³⁵

Scheme 15
(a) DPPA, Et3N, toluene, reflux, t-BuOH, reflux (85%); (b) Br₂, Et₄NBr, CH₂Cl₂, -78 °C (50%); (c) NaH, DMF, rt (52%); (d) t-BuOK, THF, rt (78%) (adapted from reference 34).

Lautens and Bexrud³⁶ developed a highly selective catalyst for the asymmetric hydroarylation of azabicycles and applied it to the synthesis of epibatidine (Scheme 16). Initially, the reaction of N-Boc-7-azanorbornene 146 with 2-chloropyridine-5-boronic acid 147 generated less than 12% of protected epibatidine 149. To circumvent the low reactivity of boronic acid, the authors used the potassium trifluoroborate 148 to produce the desired compound 149 in moderate yield, but with high enantiomeric excess. The steps a to d were reported by Gomez-Sanchez.³⁷The deprotection of N-Boc was reported by Armstrong.³⁸

Scheme 16
(a) KHF₂, MeOH, H₂O (71%); (b) Cs₂CO₃, 10% 150, THF/H₂O, 80 °C (45%) (adapted from reference 36).

5. Rearrangement

Armstrong et al.³⁸ accomplished the racemic synthesis of epibatidine from commercial 2-methoxy-3,4-dihydro-2H-pyran. The key step consisted of the construction of the 7-azabicyclo[2.2.1]heptane (154) through an elegant and exo-selective aza-Prins-pinacol rearrangement (Scheme 17). The bicyclic was obtained in only three steps from the pyran (151) (sulfonamidation using cheap chloramine-T),³⁹ the addition of vinyl Grignard reagent, and treatment with SnCl₄.⁴⁰ Attempts for a de novo synthesis of the chloropyridine group using the aldehyde moiety through a [4+2] cycloaddition approach failed, and the authors had to convert the aldehyde into a bromine group with aims at a cross-coupling approach. Thus, Jones oxidation of compound 154 followed by Barton bromodecarboxylation under sunlight gave the desired bromide 158. Changing the protecting group was necessary, with the authors commenting that the tosyl group interferes with the desired Suzuki coupling, and an unexpected intramolecular cyclization through a radical mechanism was observed. Deprotection of tosylate 158 with HBr and protection with Boc₂O led to the bromide 145. Suzuki coupling using bis(1,5-cyclooctadiene)nickel (Ni(COD)₂) in combination with bathophenanthroline yielded the N-Boc protected racemic (±)-epibatidine.

Scheme 17
(a) NBS, TsNClNa, CH₃CN, rt (79%); (b) vinyl magnesium bromide, Et₂O, -30 °C → rt (56%); (c) SnCl₄, CH₂Cl₂, 0 °C → rt (32%); (d) Jones reagent, acetone, 0 °C → rt (92%); (e) (COCl)₂, CH₂Cl₂, reflux, 2-mercaptopyridine-1-oxide, BrCCl3, 90 °C, sunlight (61%); (f) aq. HBr, phenol, reflux; (g) NaHCO₃, Boc₂O/THF, H₂O (67%, 2 steps); (h) Ni(COD)₂, bathophenanthroline, KO‘Bu, BuOH, rt, 6-chloropyridin-3-yl boronic acid, 60 °C (56%); (i) TFA, CH₂Cl₂ (89%) (adapted from reference 38).

6. Conclusions

The number of articles published from 2001 to 2014 about the synthesis of epibatidine is still significant. However, strategies for building azabicyclic systems are still limited due to its relatively low structural complexity. Therefore, the syntheses presented in this review focused on modern methods to obtain the cyclohexylamine ring or to incorporate the 4-chloropyridyl ring, including Pd catalyzed reactions and metathesis reactions.

Among the syntheses covered in this review, the synthesis reported by Jorgensen²³ seems to be the most attractive in terms of the number of steps. However, if the overall yield is evaluated, the Huang's route cannot be tolerated.

Concomitantly, advances in synthesis and biological activity studies of epibatidine analogs such as epiboxidine are ongoing.

• FAPESP has sponsored the publication of this article

7. Acknowledgments

The authors gratefully acknowledge financial support from the Fundação de Amparo à Pesquisa do Estado de São Paulo - FAPESP (2009/00617-2 and 2013/18700-9).

• ¹
  Olivo, H. F.; Hemenway, M. S.; Org. Prep. Proced. Int. 2002, 34, 1 and references therein.
• ²
  Daly, J. W.; Garraffo, H. M.; Spande, T. F.; Decker, M. W.; Sullivan, J. P.; Williams, M.; Nat. Prod. Rep. 2000, 17, 131.
• ³
  Spande, T. F.; Garraffo, H. M.; Edwards, M. W.; Yeh, H. J. C.; Lewis, P. U.; Daly, J. W.; J. Am. Chem. Soc. 1992, 114, 3415.
• ⁴
  Garraffo, H. M.; Spande, T. F.; Williams, M.; Heterocycles 2009, 79, 207.
• ⁵
  Falkenberg, M. B.; Pereira, P. A.; Lat. Am. J. Pharm. 2007, 26, 614.
• ⁶
  Badio, B.; Garraffo, H. M.; Plummer, C. V.; Padgett, W. L.; Daly, J. W.; Eur. J. Pharmacol. 1997, 321, 189.
• ⁷
  Olivo, H. F.; Colby, D. A.; Hemenway, M. S.; J. Org. Chem. 1999, 64, 4966.
• ⁸
  Pandey, G.; Tiwari, S. K.; Singh, R. S.; Mali, R. S.; Tetrahedron Lett. 2001, 42, 3947.
• ⁹
  Jones, C. D.; Simpkins, N. S.; Giblin, G. M. P.; Tetrahedron Lett. 1998, 39, 1021.
• ¹⁰
  Fletcher, S. R.; Baker, R.; Chambers, M. S.; Herbert, R. H.; Hobbs, S. C.; Thomas, S. R.; Verrier, H. M.; Watt, A. P.; Ball, R. G.; J. Org. Chem 1994, 59, 1771; Zhang, C.; Trudell, M. L.; J. Org. Chem. 1996, 61, 7189.
• ¹¹
  Pandey, G.; Bagul, T. D.; Sahoo, A. K.; J. Org. Chem 1998, 63, 760.
• ¹²
  Pandey, G.; Laha, J. K.; Lakshmaiah, G.; Tetrahedron 2002, 58, 3525.
• ¹³
  Moreno-Vargas, A. J.; Vogel, P.; Tetrahedron Asym. 2003, 14, 3173.
• ¹⁴
  Zhang, C.; Ballay II, C. J.; Trudell, M. L.; J. Chem. Soc., Perkin Trans. 1 1999, 1, 675.
• ¹⁵
  Kimura, H.; Fujiwara, T.; Katoh, T.; Nishide, K.; Kajimoto, T.; Node, M.; Chem. Pharm. Bull. 2006, 54, 399.
• ¹⁶
  Evans, D. A.; Scheidt, K. A.; Downey, W.; Org. Lett. 2001, 3, 3009.
• ¹⁷
  Lee, C-L. K.; Loh, T-P.; Org. Lett., 2005, 7, 2965.
• ¹⁸
  Hoashi, Y.; Yabuta, T.; Yuan, P.; Miyabe, H.; Takemoto, Y.; Tetrahedron 2006, 62, 365.
• ¹⁹
  Boyd, D. R.; Sharma, N. D.; Kaik, M.; McIntyre, P. B. A.; Stevenson, P. J.; Allen, C. C. R.; Org. Biomol. Chem. 2012, 10, 2774.
• ²⁰
  Boyd, D. R.; Dorrity, M. R. J.; Malone, J. F.; McMordie, R. A. S.; Sharma, N. D.; Dalton, H.; Williams, P.; J. Chem. Soc., Perkin Trans. 1 1990, 1, 489.
• ²¹
  Huang, X.; Shi, H.; Ren, J.; Liu, G.; Tang, Y.; Zeng, B.; Chin. J. Chem. 2012, 30, 1305.
• ²²
  Szántay, C.; Kardos-Balogh, Z.; Moldvai, I.; Szántay Jr., C.; Temesvári-Major, E.; Blaskó, G.; Tetrahedron Lett 1994, 35, 3171.
• ²³
  Jensen, K. L.; Weise, C. F.; Dickmeiss, G.; Morana, F.; Davis, R. L.; Jorgensen, K. A.; Chem. - Eur. J. 2012, 18, 11913.
• ²⁴
  Szántay, C.; Kardos-Balogh, Z.; Moldvai, I.; Szántay Jr., C.; Temesvári-Major, E.; Blaskó, G.; Tetrahedron 1996, 52, 11053.
• ²⁵
  Emerson, C. R.; Zakharov, L. N.; Blakemore, P. R.; Chem. - Eur. J. 2013, 19, 16342.
• ²⁶
  Corey, E. J.; Loh, T-P.; AchyuthaRao, S.; Daley, D. C.; Sarshar, S.; J. Org. Chem 1993, 58, 5600.
• ²⁷
  Barros, M. T.; Maycock, C. D.; Ventura, M. R.; Tetrahedron Lett 1999, 40, 557.
• ²⁸
  Barros, M. T.; Maycock, C. D.; Ventura, M. R.; J. Org. Chem. 1997, 62, 3984.
• ²⁹
  Barros, M. T.; Maycock, C. D.; Ventura, M. R.; J. Chem. Soc., Perkin Trans. 1 2001, 1, 166.
• ³⁰
  Avenoza, A.; Cativiela, C.; Busto, J. H.; Fernández-Recio, M. A.; Peregrina, J. M.; Rodríguez, F.; Tetrahedron 2001, 57, 545.
• ³¹
  Avenoza, A.; Busto, J. H.; Cativiela, C.; Peregrina, J. M.; Tetrahedron 2002, 58, 1193.
• ³²
  Olivo, H. F.; Hemenway, M. S.; J. Org. Chem 1999, 64, 8968.
• ³³
  Aggarwal, V. K.; Olofsson, B.; Angew. Chem., Int. Ed. 2005, 44, 5516.
• ³⁴
  Gómez-Sánches, E.; Marco-Contelles, J.; Lett. Org. Chem. 2006, 3, 827.
• ³⁵
  Che, C.; Petit, G.; Kotzyba-Hibert, F.; Bertrand, S.; Bertrand, D.; Grutter, T.; Goeldner, M.; Bioorg. Med. Chem. Lett. 2003, 13, 1001.
• ³⁶
  Bexrud, J.; Lautens, M.; Org. Lett. 2010, 12, 3160.
• ³⁷
  Gómez-Sánchez, E.; Soriano, E.; Marco-Contelles, J.; J. Org. Chem 2007, 72, 8656.
• ³⁸
  Armstrong, A.; Bhonoah, Y.; Shanahan, S. E.; J. Org. Chem 2007, 72, 8019.
• ³⁹
  Armstrong, A.; Cumming, G. R.; Pike, K.; Chem. Commun. 2004, 812.
• ⁴⁰
  Armstrong, A.; Shanahan, S. E.; Org. Lett 2005, 7, 1335.

Publication Dates

• Publication in this collection
  May 2015

History

• Received
  25 Nov 2014
• Accepted
  24 Feb 2015