Divergent Total Syntheses of (−)-Pseudocopsinine and (−)-Minovincinine

Abstract

Herein the first total syntheses of (−)-pseudocopsinine (1) and (−)-minovincine (3) from a common intermediate 8 are detailed, enlisting late-stage, hydrogen atom transfer (HAT)-mediated free radical bond formations (C20–C2 and C20–OH, respectively) that are unique to their core or structure. The approach to 1 features an Fe-mediated HAT reaction of the intermediate olefin 2, effecting a transannular C20–C2 free radical cyclization of a challenging substrate with formation of a strained [2.2.1] ring system and reaction of poor acceptor tetrasubstituted alkene with a hindered secondary free radical to form a bond and quaternary center adjacent to another quaternary center. Central to the assemblage of their underlying Aspidosperma skeleton is a powerful [4+2]/[3+2] cycloaddition cascade of 1,3,4-oxadiazole 9, which affords the stereochemically-rich and highly functionalized pentacyclic intermediate 8 as a single diastereomer in one step. The work extends the divergent total synthesis of four to now six different natural product alkaloid classes by distinguishing late stage key strategic bond formations within the underlying Aspidosperma core from the common intermediate 8. Together, the work represents use of strategic bond analysis combined with the strategy of divergent synthesis to access six different natural product classes from a single intermediate.

Graphical Abstract

INTRODUCTION

(−)-Pseudocopsinine (1, aka pseudokopsinine) was first isolated in 1967 by Yunusov from Vinca erecta Regel et. Schmalh.¹ Later its structure and absolute configuration were established by X-ray crystallography by Adrianov in 1974.² Related to the Aspidosperma alkaloids but containing an additional C20–C2 bond, the unique hexacyclic system of 1 features an unusual central strained bicyclo[2.2.1]heptane core and six contiguous stereocenters of which three are quaternary and two are adjacent to one another. Prior synthetic efforts towards 1 are limited to the model studies of Penkett and Parsons,^3,4 and semi-syntheses of the related natural products tuboxenin⁵ and vindolinine (Δ^6,7-1) by Levy.⁶ The latter employed a sodium-mediated radical cyclization of 20-iodotabersonine to effect C20–C2 bond formation without control of the relative stereochemistry, resulting in a mixture of all four possible C20/C3 diastereomers in a low combined yield of ca. 8% of which Δ^6,7-1 was a minor component. Notably, a more traditional AIBN-initiated Bu₃SnH-mediated free radical reaction of the secondary radical onto a hindered tetrasubstituted and poor acceptor alkene failed to promote cyclization with bond formation adjacent to a quaternary carbon, resulting simply in iodide reduction.⁶ Herein, we disclose the first total synthesis of 1 through late-stage formation of the strategic C20–C2 bond effected by an iron-mediated hydrogen atom transfer (HAT)^7,8 transannular radical cyclization of olefin 2 (Figure 1). Our focus on formation of the C20–C2 bond in spite of the largely failed precedent arises from the fact that it is a strategic bond, providing the greatest simplification of target structure in a retrosynthetic analysis of 1.⁹ A metal-mediated HAT reaction, of which our early studies⁷ served as a catalyst for its present day renaissance, was explored and successfully implemented based on insights detailed herein. To date, a metal-based HAT-mediated free radical cyclization has been implemented in less than 10 complex natural products total syntheses.^10,11 Only two constitute a more challenging and typically slower transannular cyclization,¹⁰ of which one is reported to be inferior to a more traditional alkylbromide/Bu₃SnH-mediated cyclization.^10b The underlying Aspidosperma skeleton in 2 was accessed in a single step through the scalable tandem [4+2]/[3+2] cycloaddition cascade of acyclic precursor 9, furnishing 8 in good yield and perfect diastereoselectivity.^12,13

Figure 1.

Late-stage strategic bond formation enables divergent syntheses of six natural product classes from key intermediate 8, the product of a [4+2]/[3+2] cycloaddition cascade of 9.

A bonus of this strategy is that it also permitted the first total synthesis of (−)-minovincinine (3) from the same intermediate 2. Oxidation (O₂)^7,14 versus transannular cyclization of the HAT radical intermediate derived from 2 was used to form the C20-OH bond and access 3. This Aspidosperma alkaloid was first isolated from Vinca erecta L. in 1962 by Janot,¹⁵ and its C20 relative stereochemistry was established initially through Horeau’s method (preferential acylation reactivity)¹⁶ and recently supported by 2D-NMR analysis.¹⁷ Minovincinine (3) has been the subject of biosynthetic studies^18–20 and one semisynthesis,²¹ but no total synthesis of 3 has been reported to date. The efforts detailed herein represent the first total syntheses of 1 and 3, both accessed from 8, complementing our recent total syntheses of (−)-kopsinine (4),^22,23 (+)-fendleridine (6),²⁴ (−)-kopsifoline D (5) and (−)-deoxoapodine (7)²⁵ from the same common intermediate 8 (Figure 1). The prior studies enlisted the masked C21 alcohol in 8 as a synthetic handle for formation of the strategic C21–C2, C21-O-C19, C21–C3, and C21-O-C6 bonds, respectively. The total syntheses of 1 and 3 described herein shifts this functionalization site from C21 to C20 to access two new classes of alkaloids. Combined with earlier studies, these efforts now provide the divergent total syntheses²⁶ of six different alkaloid classes from the common intermediate (8), notably with late-stage strategic bond formation. Combined, the work represents use of strategic bond analysis⁹ and the strategy of divergent synthesis²⁶ to access six different natural product classes from a common intermediate.

RESULTS AND DISCUSSION

Central to the assemblage of 8 was the [4+2]/[3+2] cycloaddition cascade of 9 (Figure 1), which was prepared in four steps from N-benzyltryptamine and 4-(2-tert-butyldimethylsilyloxy)pent-4-enoic acid, requiring only two purifications.²⁴ As detailed earlier,²⁴ warming a solution of 9 in o-dichlorobenzene (o-DCB) at 180 °C initiates an intramolecular [4+2] cycloaddition between the tethered alkene and 1,3,4-oxadiazole, which is followed by a loss of N₂ to generate a uniquely stabilized 1,3-dipole. This intermediate undergoes a subsequent [3+2] cycloaddition with the tethered indole with complete endo selectivity directed to the face opposite the newly formed lactam to give 8 as a single diastereomer (74–84%, gram-scale),^24b a stereochemical outcome dictated by the linking tether. Diastereoselective reductive oxido bridge ring opening of 8 (NaCNBH₃, AcOH/iPrOH, 73%) with reduction of the intermediate N-acyliminium ion exclusively from the less hindered convex face provided alcohol 10 (Scheme 1).^22–25

Scheme 1.

Synthesis of (−)-14 from 8

Compound 10 could be advanced to intermediate 14 by our previous reported sequence. In detail, debenzylation of 10 (Raney Ni, H₂, 80%) provided 11 that was converted to carbamate 12 (CbzCl, K₂CO₃, 92%). The enantiomers of 12 can be easily resolved chromatographically to give natural (+)-12, displaying a chromatographic a value of 1.8 (Chiralcel OD, 20% i-PrOH/hexanes) and providing an intermediate for which the absolute configuration was unambiguously assigned by X-ray.²² In addition to providing a remarkably simple chromatographic resolution of 12, the protecting group exchange also improved the regioselectivity of subsequent Chugaev elimination of methyl xanthate (+)-13, which was prepared from (+)-12 (NaH, CS₂; MeI, 90%). Heating a solution of (+)-13 (toluene, 150 °C) afforded the tetrasubstituted olefin (−)-14 (58%).²³

(−)-Pseudocopsinine.

With newly prepared (−)-14 in hand, the studies commenced targeting olefin (−)-2. The amide of (−)-14 was reduced selectively to the corresponding tertiary amine and subsequent acidic work-up with in situ silyl ether deprotection afforded (−)-15 (BH₃; 1 M aq. HCl, 83%). After surveying a range of methods for alcohol dehydration, a selenoxide elimination²⁷ (Scheme 2) reported by Mukai and co-workers²⁸ was found to provide the olefin (−)-2 from (−)-15 in high yield. Thus, alcohol (−)-15 was converted to tosylate (−)-16 (TsCl, Et₃N, 95%), which was subsequently displaced by in situ generated PhSeNa to afford selenide (−)-17²⁹ (NaBH₄, (PhSe)₂, 85%). Treatment of (−)-17 with N-chlorosuccinimide³⁰ (NCS) provided the corresponding selenoxide that underwent syn elimination at room temperature to afford (−)-2 in 86% yield (69% overall from 15), setting the stage for the HAT-initiated transannular free radical cyclization to the pseudocopsinine skeleton.

Scheme 2.

Synthesis of (−)-2 by selenoxide elimination

Initial efforts conducted with racemic material, which were inspired by the success of the dehydration of 12, examined the Chugaev elimination (Scheme 3) for dehydration of (±)-15. Use of typical conditions for formation of the methyl xanthate (CS₂, NaH, ImH; MeI) afforded (±)-18 in poor and capricious yields (10–35%). However, stoichiometric deprotonation of (±)-15 (KHMDS, 1 equiv) followed by sequential reaction with CS₂ and MeI afforded (±)-18 reliably in good yield (82%). The methyl xanthate (±)-18 underwent elimination in xylenes to afford olefin (±)-2 in 44% yield, accompanied by a near-equal amount of (±)-19 (34%), bearing the kopsifoline core (6,7-dihydro-kopsifoline D)²⁵ and arising through an intramolecular enamide displacement of the xanthate with Cbz cleavage. In efforts to increase selectivity for formation of (±)-2, the corresponding phenyl and benzyl xanthates of (±)-18 were also examined and afforded similar results. Although the rates of elimination for both were twice as fast as the methyl xanthate (±)-18 (see Supporting Information), the relative competitive formation of (±)-2 and (±)-19 proved essentially independent of the xanthate structure.

Scheme 3.

Synthesis of (±)-2 by Chugaev elimination

Gratifyingly, treatment of alkene (−)-2 with phenylsilane (PhSiH₃) in the presence of Fe(acac)₃⁸ (Scheme 4) provided the desired transannular cyclization reaction in excellent yield (80%) and good diastereoselectivity (3:1), favoring formation of the desired diastereomer (−)-22 (60%), as well as a minor amount of the alkene reduction product (−)-20 (7%) and a small amount of recovered (−)-2 (10%) (Scheme 4).

Scheme 4.

HAT-initiated transannular free radical cyclization of (−)-2

Notably, the reaction provided a 3:1 diasteroselectivity for formation of the C20 center and exclusive diastereoselectivity for formation of the C3 center. The yield of cyclization increased with increasing Fe(acac)₃ (1.5 vs 0.5 equiv) and time (16 vs 2 h) and decreased relative to reduction with use of a stronger reductant (PhSiH2(OiPr)³¹ vs PhSiH₃), see Supporting Information. In addition, the cyclization diastereoselectivity for generation of the desired diastereomer (−)-22 versus (−)-21 increased with decreasing reaction temperature (25 vs 50 °C).

It is a remarkable result given the ring strain introduced with the cyclization, the poor acceptor alkene and its tetrasubstitution, the hindered nature of the reacting secondary radical located adjacent to a quaternary center, and bond formation to generate a quaternary center adjacent to yet another quaternary center. Key to the success of the HAT-mediated transannular cyclization relative to more traditional methods (eg, Bu₃SnH) can be attributed to slow, low level Fe(III)-H generation (25 °C) with use of a weak reductant (PhSiH₃), both of which minimize intermediate radical reduction. It is also plausible that the released Fe(II) reacts with and stabilizes the initially formed C20 radical through reversible formation of a C20-Fe(III) bond^14,32 that slows or prevents its reduction before desired cyclization and where the adjacent C5 quaternary center prevents olefin isomerization^8b through counterproductive Fe(III)-H elimination in competition with the slow cyclization. It is even plausible, and our studies do not rule out, that an intermediate C20 Fe-species coordinates the acceptor alkene and undergoes direct carbometallation.

The diastereoselectivity for preferential formation of 22 can be rationalized based on the mechanistic studies of Holland and co-workers³² (Scheme 5). Markovnikov hydrogen atom transfer to 2 first generates radical A by way of an in situ generated Fe(III) hydride. The free radical intermediate undergoes an irreversible and kinetically controlled conjugate addition through two competing transition states (TS-21.1 and TS-22.1). In the more favorable transition state TS-22.1, the C20 methyl group avoids the 1,3-steric repulsion of the C19 and C6 hydrogen atoms and that of the C12 aryl substituent that disfavors the more sterically congested TS-21.1, preferentially providing the desired C20 stereochemistry. Both TS-21.1 and TS-22.1 yield the corresponding C3 radicals that are either quenched by free radical hydrogen atom abstraction or further reduced by Fe(II) to provide the intermediate enolates. Either enolate protonation or the hydrogen atom reduction occurs exclusively from the sterically less congested convex face for both c20 diastereomers to give (−)-21 and (−)-22 in 20% and 60%, respectively.

Scheme 5.

Rationale for diastereoselectivity of the transannular free radical cyclization of (−)-2

Without optimization, Cbz removal from (−)-22 (Pd/C, H₂, 64%) and (−)-21 (Pd/C, H₂, 58%) completed the total synthesis of (−)-1 (${\left[\alpha \right]}_{\text{D}}^{25}\text{-}23$ (c 0.075, MeOH), reported [α]_D-30 (c 1.5, MeOH)³³) and 20-epi-1 ${[\alpha ]}_{\text{D}}^{25}\text{-}7.5\pm 0.2$ (c 0.33, MeOH), respectively (Scheme 6). The ¹³C NMR chemical shifts of synthetic 1 matched closely those reported by Yagudaev³³ and Janot³⁴ for 1 (Supporting Information Table S1). Unambiguous confirmation of the synthesis of pseudocopsinine (1) was established by X-ray crystallography conducted on the hydrochloride salt of 1 with determination of the relative stereochemistry of synthetic 1.³⁵ This latter result, matching the X-ray of the natural product, resolves a disparity raised by the subtle differences in the ¹³C NMRs recorded by Yagudaev³³ and Janot,³⁴ which match ours well, and earlier spectra (See SI Table S1). It is not the structure assignment that is in question, rather simply the result of an understandable less accurate recording of the spectra in the early days of ¹³C NMR.³³

Scheme 6.

Completion of the total synthesis of 1 and 20-epi-1

(−)-Minovincinine.

An additional attribute of the approach is that this same intermediate (−)-2 allowed for the divergent synthesis of (−)-minovincinine. In order to introduce the C20 hydroxyl group (Scheme 7), (−)-2 was subjected to a series of HAT-initiated oxidation conditions for alcohol introduction¹⁴ and representative results are summarized in Scheme 7. Optimized conditions (Scheme 7, Entry 1) afforded a near 1:1 mixture (82%) of the desired diastereomer (−)-24 and its isomer (−)-23. Interestingly, the classical Mukaiyama hydration conditions (Scheme 7, Entry 3) gave 25 as a single diastereomer exclusively (See SI for characterization), which upon reduction with NaCNBH3 provided exclusively 23. Replacement of Co(acac)₂ with Co complex A³⁶ suppressed formation of 25 and provided a modest 34% yield of olefin hydration (Scheme 7, Entry 2) with a 2.5:1 preference for the desired diastereomer (−)-24. Although not investigated in detail, it is likely that C19 oxidation prior to olefin oxidation, perhaps to an iminium radical cation, followed by oxygen or peroxide attack^37,38 to form a Co(III) peroxide intermediate such as I-25^39,40 is responsible for formation of 25, accounting for the single diastereomer formed.⁴¹ The use of FePc and Fe₂(ox)₃⁷ (Scheme 7, Entries 4–5) Mn(acac)₃³¹ (Scheme 7, Entry 6) as well as Fe(NO₃)₃, Fe₂(SO₄)₃, Fe(acac)₃, Fe(dpm)₃ and Mn(dpm)₃ (not shown) gave low yields of alkene hydration.

Scheme 7.

HAT-mediated oxidation of (−)-2

Removal of the Cbz group from (−)-23 (PtO₂, H₂, 89%) and (−)-24 (PtO₂, H₂, 99%) afforded 20-epi-3 (${[\alpha ]}_{\text{D}}^{25}\text{-}429$ (c 0.07, MeOH) and (−)-3 (${[\alpha ]}_{\text{D}}^{25}\text{-}497$ (c 0.07, MeOH), reported ${[\alpha ]}_{\text{D}}^{20}\text{-}514$ (c 0.07, EtOH)¹⁷), respectively, (Scheme 8) and the latter displayed an ¹H NMR identical with that reported for the natural isolate¹⁴ (Supporting Information).

Scheme 8.

Completion of total syntheses of 3 and 20-epi-3

CONCLUSIONS

The first total syntheses of (−)-pseudocopsinine and (−)-minovincinine detailed herein were achieved in a divergent synthesis from a common intermediate 8 with late stage strategic bond formation or functionalization unique to their cores or structure. The approach features two metal-mediated HAT reactions of the intermediate olefin 2, effecting a transannular C20–C2 free radical cyclization of a challenging substrate with formation of a strained [2.2.1] ring system employing a poor tetrasubstituted acceptor alkene and hindered secondary free radical (for 1), or a Markovnikov free radical oxidation reaction (for 3). These efforts complement our previous total syntheses of (−)-kopsinine (4), (+)-fendleridine (6), (−)-kopsifoline D (5) and (−)-deoxoapodine (7), culminating in a total of six syntheses of natural products belonging to different alkaloid classes from the same common intermediate 8. Central to assemblage of the common intermediate 8 is the powerful intramolecular [4+2]/[3+2] cascade of a 1,3,4-oxadiazole that provided the underlying densely functionalized pentacyclic core as a single diastereomer in one step from a simple linear precursor. The preprogrammed oxidation at C21 was subsequently and efficiently relayed to C20, permitting access to both (−)-pseudocopsinine and (−)-minovincinine.

EXPERIMENTAL

General Methods.

Unless otherwise stated, all reactions were performed in oven-dried or flame-dried glass round-bottom flask with a rubber septum, or Fisherbrand® borosilicate glass reaction tubes with a black phenolic screw cap, under an atmosphere of dry nitrogen or argon. All reactions that require heating used oil bath as heat source. All reagents and solvents were purchased from either Fischer Chemical or Millipore-Aldrich and used directly without further purification. Preparative TLC (PTLC) and column chromatography were conducted using Millipore SiO2 60 F254 PTLC (0.5 mm) and Zeochem ZEOprep 60 ECO SiO2 (40–63 μm), respectively. Volatile solvents were removed under reduced pressure using a rotary evaporator. IR spectra were obtained using a Thermo Nicolet 380 FT-IR with a SmartOrbit Diamond ATR accessory. High resolution mass spectrometry (HRMS) analysis was performed by direct sample injection on an Agilent G1969A ESI-TOF mass spectrometer. The single crystal X-ray diffraction studies were carried out on a Bruker Kappa APEX-II CCD diffractometer equipped with Mo Kα radiation (λ = 0.71073) at UCSD Crystallography Facility. Structural assignments were made with additional information from gCOSY, gHSQC, gNOESY and gHMBC experiments.

(−)-15.

A stirred solution of (−)-14²² (351 mg, 0.569 mmol) in anhydrous THF (50 mL) under Ar at 0 °C was treated with BH₃•THF (1 M in THF, 8.5 mL, 8.5 mmol), and was stirred at 0 °C until full conversion (1 h). The reaction mixture was quenched with addition of distilled water until gas evolution ceased (1 mL added), followed by addition of 40 mL of aqueous 1 M HCl. The resulting solution was removed from cold bath and stirred for 16 h at 25 °C. The mixture was basified to pH 8 with addition of aqueous 1 M NaOH (22 mL) and extracted with EtOAc (3 × 50 mL). The combined organic layers were dried over MgSO₄ and concentrated under reduced pressure. Flash chromatography (SiO₂, 40% EtOAc/hexane) afforded (−)-15 (227 mg, 83%) as white foam. ¹H NMR (600 MHz, CDCl₃) δ 7.77 – 7.71 (m, 1H), 7.39 – 7.29 (m, 5H), 7.21 (td, J = 7.7, 1.3 Hz, 1H), 7.16 (dd, J = 7.4, 1.3 Hz, 1H), 7.03 (td, J = 7.5, 1.1 Hz, 1H), 5.36 (d, J = 12.4 Hz, 1H), 5.13 (d, J = 12.4 Hz, 1H), 3.53 (s, 3H), 3.52 – 3.43 (m, 2H), 3.12 (ddt, J = 10.8, 4.2, 1.9 Hz, 1H), 3.07 (d, J = 15.2 Hz, 1H), 2.93 – 2.86 (m, 1H), 2.41 – 2.33 (m, 2H), 2.27 (td, J = 11.2, 2.9 Hz, 1H), 2.17 (td, J = 11.7, 7.1 Hz, 1H), 2.08 (dd, J = 15.3, 1.9 Hz, 1H), 1.91 – 1.80 (m, 2H), 1.64 (dd, J = 12.0, 5.5 Hz, 1H), 1.58 (ddd, J = 13.4, 6.2, 2.9 Hz, 1H), 1.44 – 1.31 (m, 2H), 0.94 (ddd, J = 13.5, 7.5, 5.8 Hz, 2H); ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 168.1, 152.5, 149.6, 140.3, 138.1, 135.9, 128.7, 128.3, 128.2, 127.7, 124.1, 120.6, 115.9, 112.6, 71.0, 68.0, 58.6, 53.4, 51.8, 51.8, 51.6, 42.8, 39.3, 37.3, 33.9, 28.8, 22.3; IR (film) ν_max 3433, 2933, 2778, 1717, 1473, 1391, 1356, 1328, 1303, 1282, 1253, 1215, 1180, 1126, 1068, 1045, 905, 851, 752, 697 cm⁻¹; HRMS (ESI-TOF) m/z 489.2376 [(M + H)⁺, C₂₉H₃₃N₂O₅⁺ requires 489.2389]; ${[\alpha ]}_{\text{D}}^{23}\text{-}85$ (c 0.61, CHCl₃).

(−)-16.

A stirred solution of (−)-15 (213 mg, 0.436 mmol) in CH₂Cl₂ (10 mL) was cooled to 0 °C and treated with Et₃N (3.30 mL, 23.7 mmol) followed by TsCl (997 mg, 5.23 mmol), and the reaction mixture was warmed to 25 °C and stirred for 6 h. The resulting solution was poured into H₂O (15 mL) and the organic layer was separated. The aqueous layer was extracted with CH₂Cl₂ (3 × 10 mL). The combined organic layers were dried over MgSO₄ and concentrated under reduced pressure. Flash column chromatography (SiO₂, 30% EtOAc/hexane) afforded (−)-16 (267 mg, 95%) as white foam. ¹H NMR (600 MHz CDCl₃) δ 7.75 (d, J = 8.0 Hz, 1H), 7.64 (d, J = 8.1 Hz, 2H), 7.41 – 7.29 (m, 5H), 7.27 (d, J = 8.1 Hz, 2H), 7.21 (td, J = 7.8, 1.3 Hz, 1H), 7.11 – 7.07 (m, 1H), 7.01 (td, J = 7.5, 1.0 Hz, 1H), 5.35 (d, J = 12.3 Hz, 1H), 5.13 (d, J = 12.3 Hz, 1H), 3.83 (td, J = 6.6, 2.8 Hz, 2H), 3.52 (s, 3H), 3.08 (m, 2H), 2.88 (t, J = 7.8 Hz, 1H), 2.41 (s, 3H), 2.35 (ddd, J = 11.6, 8.6, 5.6 Hz, 1H), 2.31 (s, 1H), 2.27 – 2.20 (m, 1H), 2.15 (td, J = 11.6, 7.1 Hz, 1H), 2.00 (dd, J = 15.5, 1.9 Hz, 1H), 1.8 – 1.78 (m, 1H), 1.74 (d, J = 14.0 Hz, 1H), 1.62 (dd, J = 12.0, 5.5 Hz, 1H), 1.55 (m, 1H), 1.47 (dt, J = 14.2, 6.9 Hz, 1H), 1.29 (td, J = 13.6, 5.0 Hz, 1H), 1.01 (dt, J = 13.8, 6.3 Hz, 1H); ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 167.5, 152.4, 149.8, 144.7, 140.3, 137.6, 135.8, 132.9, 129.9, 128.7, 128.4, 128.2, 128.0, 127.7, 124.1, 120.6, 115.9, 112.1, 70.7, 68.0, 66.8, 53.4, 51.7, 51.6, 51.6, 42.7, 37.3, 34.9, 33.5, 28.5, 22.2, 21.8; IR (film) ν_max 2932, 2779, 1721, 1601, 1473, 1460, 1391, 1358, 1330, 1303, 1282, 1232, 1177, 0097, 1047, 1019, 952, 908, 874, 814, 753, 698, 664 cm⁻¹; HRMS (ESI-TOF) m/z 643.2482 [(M + H)⁺, C₃₆H₃₉N₂O₇S⁺, required 643.2478]; ${\left[\alpha \right]}_D^{\mathbf{22}}\text{-}51$ (c 0.39, CHCl₃).

(−)-17.

A solution of (−)-16 (267 mg, 0.416 mmol) in EtOH (15 mL) was cooled to 0 °C and treated with diphenyl diselenide (649 mg, 2.08 mmol). A solution of NaBH₄ (157 mg, 4.16 mmol) in EtOH (10 mL) was added to the reaction mixture dropwise, which then was warmed slowly to 25 °C over 30 min and subsequently warmed at 60 °C for 40 min. The reaction mixture was cooled to 25 °C and concentrated under reduced pressure. The crude residue was purified by flash column chromatography (SiO₂, 20% EtOAc/hexane) to afford (−)-17 (223 mg, 85%) as white foam. ¹H NMR (600 MHz, CDCl₃) δ 7.73 (d, J = 8.0 Hz, 1H), 7.42 – 7.29 (m, 6H), 7.20 (t, J = 7.8 Hz, 1H), 7.15 – 7.08 (m, 5H), 7.01 (t, J = 7.5 Hz, 1H), 5.36 (d, J = 12.3 Hz, 1H), 5.13 (d, J = 12.3 Hz, 1H), 3.49 (s, 3H), 3.10 (d, J = 10.9 Hz, 1H), 3.02 (d, J = 15.5 Hz, 1H), 2.88 (t, J = 7.8 Hz, 1H), 2.64 (td, J = 12.1, 4.5 Hz, 1H), 2.52 (td, J = 12.0, 5.2 Hz, 1H), 2.40 – 2.31 (m, 2H), 2.28 – 2.21 (m, 1H), 2.14 (td, J = 11.6, 7.1 Hz, 1H), 2.04 (d, J = 15.1 Hz, 1H), 1.82 (dd, J = 28.9, 13.6 Hz, 2H), 1.65 – 1.60 (m, 2H), 1.47 (td, J = 13.4, 13.0, 5.1 Hz, 1H), 1.32 (td, J = 13.4, 13.0, 4.6 Hz, 1H), 1.12 (td, J = 13.4, 12.9, 4.4 Hz, 1H); ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 167.6, 152.5, 150.5, 140.3, 137.7, 135.9, 132.4, 129.8, 129.0, 128.7, 128.4, 128.2, 127.6, 126.8, 124.1, 120.5, 115.9, 111.9, 70.4, 68.0, 53.4, 51.8, 51.7, 51.5, 42.7, 39.0, 37.6, 33.5, 28.3, 22.3, 21.4; IR (film) ν_max 2931, 2787, 1721, 1474, 1436, 1391, 1356, 1328, 1303, 1281, 1250, 1231, 1200, 1181, 1157, 1116, 1068, 1047, 1019, 735, 694 cm⁻¹; HRMS (ESI-TOF) m/z 623.1966 [(M + H)⁺, C₃₅H₃₇N₂O₄⁷⁴Se⁺, required 623.1978] and m/z 629.1919 [(M +H)⁺, C₃₅H₃₇N₂O₄⁸⁰Se⁺, required 629.1919]; ${\left[\alpha \right]}_D^{22}\text{-}79$ (c 0.54, CHCL₃).

(−)-2.

A solution of (−)-17 (223 mg, 0.355 mmol) in MeOH (10 mL) and CH₂Cl₂ (10 mL) was cool to 0 °C and treated with a solution of N-chlorosuccinimide (NCS, 47.4 mg, 0.355 mmol) in MeOH (1 mL) and CH₂Cl₂ (1 mL) dropwise. The reaction mixture was stirred at 0 °C for 40 min before being quenched with the addition of saturated aqueous NaHCO₃ (7 mL). The biphasic mixture was separated, and the aqueous layer was extracted with CH₂Cl₂ (3 × 40 mL). The combined organic layers were dried over MgSO₄, concentrated under reduced pressure, and the resulting residue was dissolved in benzene and left at 25 °C for 16 h. The resulting light-yellow solution was concentrated under reduced pressure, and the crude residue was subjected to flash column chromatography (SiO₂, 30% EtOAc/hexane) and afforded (−)-2 (144 mg, 86%) as a white foam. ¹H NMR (600 MHz, CDCl₃) δ 7.72 (dt, J = 8.0, 0.7 Hz, 1H), 7.38 – 7.29 (m, 5H), 7.22 – 7.15 (m, 2H), 7.03 (td, J = 7.5, 1.1 Hz, 1H), 5.41 (dd, J = 17.9, 10.9 Hz, 1H), 5.35 (d, J = 12.3 Hz, 1H), 5.12 (d, J = 12.3 Hz, 1H), 4.73 – 4.68 (m, 2H), 3.54 (s, 3H), 3.19 (d, J = 15.1 Hz, 1H), 3.16 – 3.11 (m, 1H), 2.93 (dd, J = 8.6, 7.1 Hz, 1H), 2.63 (d, J = 1.8 Hz, 1H), 2.43 (ddd, J = 11.4, 8.6, 5.6 Hz, 1H), 2.35 – 2.27 (m, 1H), 2.23 – 2.11 (m, 2H), 1.88 (tdt, J = 13.4, 12.0, 4.6 Hz, 1H), 1.71 – 1.64 (m, 2H), 1.61 (ddd, J = 13.3, 5.1, 2.5 Hz, 1H), 1.46 (td, J = 13.7, 4.9 Hz, 1H); ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 167.9, 152.5, 150.2, 143.6, 140.4, 137.9, 135.9, 128.7, 128.3, 128.2, 127.5, 124.0, 120.6, 115.8, 113.9, 113.3, 69.9, 66.8, 53.7, 51.8, 51.6, 51.5, 42.9, 41.7, 35.2, 30.2, 22.1.; IR (film) ν_max 2935, 2776, 1722, 1603, 1473, 1461, 1435, 1390, 1356, 1326, 1302, 1281, 1251, 1231, 1204, 1182, 1120, 1050, 914, 751, 697 cm⁻¹; HRMS (ESI-TOF) m/z 471.2283 [(M + H)⁺, C₂₉H₃₁N₂O₄⁺, required 471.2284]; ${\left[\alpha \right]}_D^{21}\text{-}114$ (c 0.48, CHCl₃).

(−)-22.

A solution of (−)-2 (20 mg, 0.042 mmol) and Fe(acac)₃ (22.5 mg, 0.0637 mmol) in EtOH (12.5 mL) was subjected to 3 cycles of freeze-pump-thaw and placed under Ar, before being treated with a solution of phenylsilane in dichloroethane (0.15 M, 1.4 mL, 0.21 mmol, subjected to 3 cycles of freeze-pump-thaw) at 25 °C under Ar. The reaction mixture was stirred 25 °C for 16 h before being quenched by the addition of a mixture of saturated aqueous NaHCO₃, saturated aqueous Rochelle’s salt, and saturated aqueous Na₂S₂O₄ (1:1:1, v/v, 10 mL). The resulting biphasic mixture was stirred vigorously for 30 min at 25 °C under open air, which was then extracted with EtOAc (3 × 10 mL). The combined organic layers were dried over MgSO₄ then concentrated under reduced pressure, and the resulting crude residue was purified by PTLC (SiO₂, 50% acetone/hexane, developed 3 times) to afford recovered (−)-2 (2.1 mg, 10%, R_f: 0.9), (−)-20 (1.4 mg, 7%, R_f: 0.8), the desired diastereomer (−)-22 (11.8 mg, 60%, R_f: 0.5) and undesired diastereomer (−)-21 (4.0 mg, 20%, R_f: 0.45) as white powders. For (−)-20: ¹H NMR (600 MHz, CDCl₃) δ 7.75 (d, J = 8.0 Hz, 1H), 7.40 – 7.29 (m, 5H), 7.20 (td, J = 7.8, 1.3 Hz, 1H), 7.16 (d, J = 7.3 Hz, 1H), 7.03 (t, J = 7.5 Hz, 1H), 5.36 (d, J = 12.3 Hz, 1H), 5.13 (d, J = 12.3 Hz, 1H), 3.54 (s, 3H), 3.12 (d, J = 11.0 Hz, 1H), 2.98 (d, J = 15.3 Hz, 1H), 2.93 – 2.86 (m, 1H), 2.39 – 2.33 (m, 2H), 2.25 (td, J = 11.3, 2.8 Hz, 1H), 2.21 – 2.14 (m, 1H), 2.02 (dd, J = 15.4, 1.8 Hz, 1H), 1.90 – 1.81 (m, 1H), 1.76 (d, J = 13.8 Hz, 1H), 1.63 (dd, J = 12.0, 5.6 Hz, 1H), 1.59 (m, 1H), 1.23 (dd, J = 13.5, 4.9 Hz, 1H), 1.15 – 1.06 (m, 1H), 0.70 (dq, J = 14.3, 7.3 Hz, 1H), 0.56 (t, J = 7.5 Hz, 3H); ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 167.9, 152.5, 150.0, 140.4, 138.4, 135.9, 128.7, 128.3, 128.2, 127.5, 123.9, 120.6, 115.8, 112.5, 71.1, 67.9, 53.4, 51.9, 51.9, 51.5, 42.7, 38.0, 32.9, 29.0, 28.0, 22.4, 7.3; IR (film) ν_max 2931, 2776, 1722, 1473, 1460, 1391, 1356, 1328, 1303, 1280, 1250, 1231, 1183, 1114, 1049, 1019, 752, 698, 623 cm⁻¹; HRMS (ESI-TOF) m/z 473.2443 [(M + H)⁺, C₂₉H₃₃N₂O₄⁺, required 473.2440); ${[\alpha ]}_D^{20}\text{-}44$ (c 0.14, CHCl₃). For (−)-21: ¹H NMR (500 MHz, DMSO-d₆, 55 °C) δ 7.69 (s, 1H), 7.49 – 7.33 (m, 5H), 7.22 – 7.12 (m, 2H), 7.01 (t, J = 7.6 Hz, 1H), 5.32 (s, 2H), 3.99 (s, 1H), 3.51 (s, 3H), 3.25 (s, 2H), 3.12 (s, 1H), 3.02 (d, J = 12.1 Hz, 1H), 2.92 (d, J = 11.0 Hz, 1H), 2.79 (dd, J = 13.9, 6.4 Hz, 1H), 2.33 (s, 1H), 2.23 – 2.07 (m, 1H), 1.69 (d, J = 12.9 Hz, 2H), 1.62 – 1.46 (m, 3H), 1.42 (s, 1H), 0.36 (d, J = 7.2 Hz, 3H); ¹³C{¹H} NMR (126 MHz, DMSO-d₆, 55 °C) δ 173.2, 152.4, 138.3, 136.2, 128.2, 127.8, 127.7, 127.2, 123.3, 121.4, 113.8, 79.8, 75.7, 66.3, 57.8, 55.0, 53.5, 51.2, 47.0, 42.7, 32.5, 28.2, 20.5, 9.1; IR (film) ν_max 2938, 1705, 1599, 1480, 1459, 1398, 1350, 1325, 1255, 1219, 1204, 1155, 1130, 1102, 1040, 1023, 753, 699, 618 cm⁻¹; HRMS (ESI-TOF) m/z 473.2441 [(M + H)⁺, C₂₉H₃₃N₂O₄⁺, required 473.2440]. ${[\alpha ]}_D^{21}\text{-}50$ (c 0.40, CHCl₃). For (−)-22: ¹H NMR (500 MHz, DMSO-d₆, 70 °C) δ 7.66 (s, 1H), 7.46 – 7.30 (m, 5H), 7.23 (dd, J = 7.5, 1.4 Hz, 1H), 7.19 – 7.13 (m, 1H), 7.02 (td, J = 7.4, 1.1 Hz, 1H), 5.24 (s, 2H), 4.20 (s, 1H), 3.48 (s, 3H), 3.18 (dt, J = 8.4, 6.5 Hz, 1H), 3.15 – 3.00 (m, 2H), 2.91 – 2.80 (m, 2H), 2.74 (ddd, J = 14.4, 6.8, 1.8 Hz, 1H), 2.12 – 2.04 (m, 1H), 1.81 – 1.70 (m, 2H), 1.69 – 1.49 (m, 3H), 1.48 – 1.33 (m, 2H), 0.90 (d, J = 6.8 Hz, 3H); ¹³C{¹H} NMR (126 MHz, DMSO-d₆, 70 °C) δ 173.6, 152.6, 141.1, 138.7, 135.9, 128.0, 127.9, 127.7, 126.8, 123.2, 122.5, 115.2, 80.1, 79.2, 66.3, 58.9, 53.8, 52.0, 50.9, 47.5, 43.5, 37.9, 29.4, 28.2, 20.1, 13.3, 8.1; IR (film) ν_max 2935, 1705, 1480, 1461, 1399, 1349, 1324, 1292, 1219, 1153, 1106, 1047, 752, 698, 643 cm⁻¹; HRMS (ESI-TOF) m/z 473.2445 [(M + H)⁺, C₂₉H₃₃N₂O₄⁺, required 473.2440]; ${[\alpha ]}_D^{21}\text{-}53$ (c 1.18, CHCl₃).

(−)-Pseudocopsinine ((−)-1).

Pd/C (10% Pd w/w, 4.5 mg, 0.0042 mmol) was suspended in a solution of (−)-22 (4.0 mg, 0.0085 mmol) in MeOH (2 mL). The suspension was sparged with H₂ gas for 20 min while vigorously stirring at 25 °C before being stirred for 16 h at 25 °C under a H₂ atmosphere. The reaction mixture was filtered through Celite, and the filtrate was concentrated under reduced pressure. The residue was purified by PTLC (SiO₂, 5% MeOH/CH₂Cl₂) to afford (−)-pseudocopsinine (1, 1.9 mg, 64%). ¹H NMR (600 MHz, CDCl₃) δ 7.27 (s, 1H), 7.06 (td, J = 7.6, 1.3 Hz, 1H), 6.85 (td, J = 7.4, 1.0 Hz, 1H), 6.80 – 6.76 (m, 1H), 3.70 (s, 3H), 3.33 (td, J = 8.9, 6.8 Hz, 1H), 3.19 (m, 2H), 3.14 – 3.08 (m, 1H), 3.05 (dq, J = 11.8, 6.2, 5.6 Hz, 2H), 2.87 (ddd, J = 14.3, 6.3, 2.1 Hz, 1H), 2.12 (ddd, J = 14.7, 6.7, 2.8 Hz, 1H), 1.84 – 1.68 (m, 4H), 1.57 (qd, J = 9.1, 4.5 Hz, 2H), 1.52 – 1.45 (m, 1H), 0.85 (d, J = 6.7 Hz, 3H); ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 175.2, 149.6, 140.2, 127.4, 123.8, 121.4, 112.9, 80.6, 78.7, 60.4, 55.1, 52.1, 51.1, 48.1, 44.6, 40.3, 37.3, 31.3, 29.0, 20.6, 7.6; IR (film) ν_max 3369, 2931, 1725, 1607, 1464, 1435, 1375, 1347, 1322, 1203, 1132, 1102, 1074, 950, 889, 746, 699, 669, 651 cm⁻¹; HRMS (ESI-TOF) m/z 339.2077 [(M + H)⁺, C₂₁H₂₇N₂O₂⁺, required 339.2073]; ${[\alpha ]}_D^{25}\text{-}5.5$ (c 1.5, CHCl₃) and ${[\alpha ]}_D^{25}\text{-}23.4$ (c 0.075, MeOH), Yagudaev³³ reported [α]_D −30.4 (c 1.51, MeOH).

(±)-Pseudocopsinine Hydrochloride ((±)-1•HCl).

A solution of (±)-pseudocopsinine (4.6 mg, 0.014 mmol, prepared by the route above with racemic material) in EtOH (0.2 mL) at 0 °C was treated with ethanolic HCl (0.42 M, 0.16 mL) and warmed slowly to 25 °C. The reaction solvent was removed under a stream of N₂ and the resulting residue was washed with MeCN (2 × 0.5 mL) and dried under a stream of N₂ to yield (±)-pseudocopsinine hydrochloride as a white solid. ¹H NMR (600 MHz, CD₃OD) δ 7.25 – 7.20 (m, 1H), 7.07 (td, J = 7.7, 1.2 Hz, 1H), 6.78 (td, J = 7.5, 1.0 Hz, 1H), 6.69 (dt, J = 7.8, 0.8 Hz, 1H), 4.59 (s, 1H), 3.96 (s, 1H), 3.91 – 3.83 (m, 2H), 3.74 (s, 3H), 3.70 (dt, J = 13.2, 4.7 Hz, 1H), 3.48 (ddd, J = 13.2, 10.4, 4.5 Hz, 1H), 3.36 – 3.32 (m, 1H), 2.75 (ddd, J = 15.3, 5.8, 2.4 Hz, 1H), 2.48 – 2.40 (m, 1H), 2.10 (ddd, J = 15.5, 12.3, 1.4 Hz, 1H), 2.06 – 1.90 (m, 4H), 1.83 (ddd, J = 12.6, 7.6, 4.5 Hz, 1H), 1.66 (dt, J = 13.5, 7.9 Hz, 1H), 1.00 (d, J = 6.8 Hz, 3H); ¹³C{¹H} NMR (151 MHz, CD₃OD) δ 175.5, 150.9, 135.0, 129.7, 124.5, 120.9, 111.9, 81.7, 74.6, 60.8, 57.7, 52.7, 52.1, 44.1, 41.3, 35.8, 34.2, 30.7, 25.8, 20.3, 6.8. The structure and the relative configuration of (±)-pseudocopsinine hydrochloride were unambiguously established with a single crystal X-ray structure determination conducted on a light yellow prism obtained from vapor diffusion between MeOH and Et₂O at −20 °C (CCDC 2015375), see Scheme 6.³⁵

(−)-20-epi-Pseudocopsinine ((−)-20-epi-1).

Pd/C (10% Pd w/w, 4.5 mg, 0.0042 mmol) was suspended in a solution of (−)-21 (4.0 mg, 0.0085 mmol) in MeOH (2 mL). The suspension was sparged with H₂ gas for 20 min while being vigorously stirred at 25 °C, then stirred for 16 h at the same temperature under a H₂ atmosphere. The reaction mixture was filtered through Celite, and the filtrate was concentrated under reduced pressure. The residue was purified by PTLC (SiO₂, 5% MeOH/CH₂Cl₂) to afford (−)-20-epi-pseudocopsinine (1.7 mg, 58%). ¹H NMR (600 MHz, CDCl₃) δ 7.27 (m, 1H), 7.06 (td, J = 7.6, 1.1 Hz, 1H), 6.86 (t, J = 7.4 Hz, 1H), 6.78 (d, J = 7.8 Hz, 1H), 3.71 (s, 3H), 3.65 (d, J = 8.5 Hz, 1H), 3.31 (dd, J = 22.0, 16.2 Hz, 3H), 3.09 – 3.00 (m, 2H), 2.90 (dd, J = 14.1, 6.2 Hz, 1H), 2.17 (dt, J = 14.4, 4.6 Hz, 1H), 1.90 (q, J = 7.3 Hz, 1H), 1.86 – 1.78 (m, 2H), 1.75 (t, J = 13.1 Hz, 1H), 1.68 – 1.62 (m, 2H), 1.57 (q, J = 7.3 Hz, 1H), 0.53 (d, J = 7.2 Hz, 3H); ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 174.0, 149.8, 127.7, 122.8, 121.9, 112.6, 105.6, 80.0, 73.6, 59.3, 59.1, 54.7, 52.2, 47.8, 43.54, 43.46, 38.0, 35.8, 27.7, 20.3, 9.5; IR (film) ν_max 3355, 2930, 1723, 1906, 1482, 1465, 1350, 1322, 1220, 1203, 1152, 1110, 1063, 1018, 924, 828, 747, 699, 653, 621 cm⁻¹; HRMS (ESI-TOF) m/z 339.2064 [(M + H)⁺, C₂₁H₂₇N₂O₂⁺, required 339.2073]; ${[\alpha ]}_D^{25}\text{-}7.5\pm 0.2$ (c 0.33, MeOH).

(−)-24.

A stirred solution of (−)-2 (20.0 mg, 0.0425 mmol) and complex 1 (17.4 mg, 0.0425 mmol) in EtOH (10 mL) was sparged with O₂ for 15 min at 25 °C before being treated with 30 μL of PhSiH₃ at 3 h intervals until completion (180 μL PhSiH₃ added in total, 1.5 mmol). The resulting solution was quenched with the addition of saturated aqueous Na₂S₂O₄ (10 mL) and extracted with EtOAc (3 × 5 mL). The combined organic layers were dried over MgSO₄ then concentrated under reduced pressure. The resulting dark brown liquid was loaded onto a silica gel plug which was flushed with hexanes to remove PhSiH₃ byproducts, then eluted with 40% EtOAc/hexanes to provide the hydration products as mixture of two diastereomers. The fractions containing the hydration (oxidation) products were concentrated under reduced pressure, and the resulting white foam was purified by PTLC (SiO₂, 60% Et₂O/hexane, developed 3 times) to afford desired diastereomer (−)-24 (8.3 mg, 38%, R_f: 0.45) and undesired diastereomer (−)-23 (9.2 mg, 44%, R_f: 0.5) as white powders. For (−)-23: ¹H NMR (600 MHz, CDCl₃) δ 7.76 (dt, J = 7.9, 0.8 Hz, 1H), 7.38 – 7.28 (m, 5H), 7.23 (qd, J = 7.5, 1.3 Hz, 2H), 7.05 (td, J = 7.5, 1.1 Hz, 1H), 5.35 (d, J = 12.3 Hz, 1H), 5.14 (d, J = 12.4 Hz, 1H), 3.53 (s, 3H), 3.50 – 3.44 (m, 1H), 3.17 (d, J = 15.8 Hz, 1H), 3.12 – 3.05 (m, 1H), 2.96 – 2.90 (m, 2H), 2.37 (ddd, J = 11.5, 8.6, 5.8 Hz, 1H), 2.27 (ddd, J = 12.0, 10.7, 2.9 Hz, 1H), 2.19 (td, J = 11.7, 7.2 Hz, 1H), 2.01 (dd, J = 15.7, 1.9 Hz, 1H), 1.85 (dtt, J = 17.3, 8.6, 4.6 Hz, 1H), 1.76 – 1.69 (m, 1H), 1.69 – 1.62 (m, 2H), 1.49 (ddd, J = 13.2, 5.3, 1.8 Hz, 1H), 0.86 (d, J = 6.4 Hz, 3H); ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 167.5, 152.5, 150.9, 140.5, 137.9, 135.9, 128.7, 128.4, 128.2, 127.8, 124.2, 120.5, 115.9, 111.4, 68.0, 66.6, 65.8, 53.3, 52.0, 51.8, 51.6, 42.5, 42.2, 28.2, 24.7, 21.9, 17.4; IR (film) ν_max 2934, 2781, 1721, 1473, 1460, 1392, 1357, 1327, 1305, 1282, 1250, 1234, 1204, 1181, 1111, 1083, 1048, 752, 698, 616 cm⁻¹; HRMS (ESI-TOF) m/z 489.2391 [(M + H)⁺, C₂₉H₃₃N₂O₅⁺, required 489.2389]; ${[\alpha ]}_D^{22}\text{-}103$ (c 0.92, CHCl₃). For (−)-24: ¹H NMR (600 MHz, CDCl₃) δ 7.74 (dt, J = 8.1, 0.7 Hz, 1H), 7.40 – 7.29 (m, 5H), 7.20 (td, J = 7.8, 1.3 Hz, 1H), 7.13 (dd, J = 7.3, 1.3 Hz, 1H), 7.03 (td, J = 7.5, 1.0 Hz, 1H), 5.37 (d, J = 12.3 Hz, 1H), 5.14 (d, J = 12.4 Hz, 1H), 3.55 (s, 3H), 3.50 (q, J = 6.4 Hz, 1H), 3.14 – 3.10 (m, 1H), 3.09 (d, J = 15.2 Hz, 1H), 2.92 (dd, J = 8.6, 7.1 Hz, 1H), 2.45 (d, J = 1.8 Hz, 1H), 2.40 (dd, J = 15.2, 1.9 Hz, 1H), 2.38 – 2.33 (m, 1H), 2.25 (td, J = 11.3, 3.1 Hz, 1H), 2.21 – 2.13 (m, 1H), 1.91 – 1.79 (m, 1H), 1.68 – 1.60 (m, 3H), 1.52 – 1.44 (m, 1H), 0.93 (d, J = 6.4 Hz, 3H); ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 168.0, 152.5, 149.6, 140.5, 137.4, 135.9, 128.7, 128.3, 128.3, 127.7, 124.0, 120.6, 115.9, 112.4, 68.0, 67.0, 67.0, 53.4, 52.0, 51.9, 51.7, 42.8, 42.7, 27.5, 25.7, 21.8, 17.6; IR (film) ν_max 1940, 2780, 1721, 1473, 1460, 1392, 1356, 1304, 1283, 1253, 1233, 1204, 1181, 1110, 1070, 1049, 754, 732, 636 cm⁻¹; HRMS (ESI-TOF) m/z 489.2387 [(M + H)⁺, C₂₉H₃₃N₂O₅⁺, required 489.2389]; ${[\alpha ]}_D^{22}\text{-}105$ (c 0.83, CHCl₃).

For 25:

¹H NMR (600 MHz, CDCl₃) δ 7.69 (d, J = 8.1 Hz, 1H), 7.51 (dd, J = 7.6, 1.3 Hz, 1H), 7.44 – 7.30 (m, 5H), 7.22 (td, J = 7.9, 1.4 Hz, 1H), 7.10 – 7.06 (m, 1H), 5.37 (d, J = 12.3 Hz, 1H), 5.18 (d, J = 12.3 Hz, 1H), 4.08 (q, J = 6.4 Hz, 1H), 3.56 (s, 3H), 3.41 (td, J = 8.8, 5.9 Hz, 1H), 3.29 (dt, J = 12.9, 6.6 Hz, 1H), 3.00 (td, J = 8.6, 2.7 Hz, 1H), 2.90 (dt, J = 11.5, 5.3 Hz, 1H), 2.70 (d, J = 16.3 Hz, 1H), 2.35 – 2.27 (m, 2H), 2.03 – 1.96 (m, 1H), 1.82 (ddd, J = 12.1, 6.0, 2.6 Hz, 1H), 1.78 – 1.71 (m, 2H), 1.41 (dt, J = 13.7, 5.4 Hz, 1H), 1.01 (d, J = 6.4 Hz, 3H); ¹³C{¹H} NMR (151 MHz, CDCl₃) δ 167.8, 152.5, 148.6, 140.2, 135.6, 134.3, 128.8, 128.6, 128.5, 127.9, 125.2, 124.4, 115.5, 109.6, 104.6, 81.2, 68.4, 59.4, 51.7, 49.4, 49.1, 43.7, 39.0, 30.3, 27.4, 18.3, 11.6; IR (film) ν_max 2945, 1727, 1601, 1462, 1385, 1385, 1350, 1283, 1223, 1134, 1114, 1043, 758, 699 cm⁻¹; HRMS (ESI-TOF) m/z 503.2184 [(M + H)⁺, C₂₉H₃₁N₂O₆⁺, required 503.2182].

(−)-Minovincinine (3).

PtO₂ (6.8 mg, 0.025 mmol) was suspended in a solution of (−)-24 (4.0 mg, 0.0082 mmol) in MeOH (2 mL). The suspension was vigorously stirred and sparged with H₂ for 20 min at 25 °C, then was stirred for 16 h at 25 °C under a H₂ atmosphere. The reaction mixture was filtered through Celite and the filtrate was concentrated under reduced pressure. The crude residue was purified by PTLC (SiO₂, 60% EtOAc/hexane) to afford (−)-minovincinine (3, 2.9 mg, 99%). Synthetic (−)- minovincinine is spectroscopically identical to that reported for naturally isolated material as reported by Williams et al.^{17 1}H NMR (600 MHz, acetone-d₆) δ 9.16 (s, 1H), 7.26 (d, J = 7.4 Hz, 1H), 7.10 (td, J = 7.7, 1.2 Hz, 1H), 7.00 (dt, J = 7.8, 0.8 Hz, 1H), 6.83 (td, J = 7.4, 1.0 Hz, 1H), 3.66 (s, 3H), 3.42 – 3.35 (m, 1H), 3.14 – 3.07 (m, 1H), 2.98 (d, J = 5.3 Hz, 1H), 2.89 (dd, J = 8.3, 6.5 Hz, 1H), 2.75 (d, J = 14.6 Hz, 1H), 2.68 (dd, J = 14.7, 1.9 Hz, 1H), 2.65 (d, J = 1.8 Hz, 1H), 2.54 (ddd, J = 11.6, 8.3, 4.7 Hz, 1H), 2.40 (td, J = 10.6, 3.6 Hz, 1H), 1.99 (td, J = 11.5, 6.5 Hz, 1H), 1.87 – 1.77 (m, 1H), 1.61 (m, 3H), 1.52 (ddd, J = 13.4, 12.0, 4.7 Hz, 1H), 0.86 (d, J = 6.5 Hz, 3H); ¹³C{¹H} NMR (151 MHz, acetone-d₆) δ 169.3, 168.0, 144.9, 138.1, 128.3, 121.6, 121.1, 110.5, 93.2, 68.8, 67.0, 56.2, 52.2, 51.4, 50.9, 46.6, 43.8, 26.9, 26.5, 22.7, 18.3; IR (film) ν_max 3365, 2933, 1776, 1668, 1605, 1463, 1438, 1371, 1296, 1279, 1253, 1235, 1154, 1129, 1104, 1044, 902, 796, 745 cm⁻¹; HRMS (ESI-TOF) m/z 355.2023 [(M + H)⁺, C₂₁H₂₇N₂O₃⁺, required 355.2022]; ${[\alpha ]}_D^{25}\text{-}497\pm 1$ (c 0.07, MeOH), Janot and co-workers¹⁵ reported ${[\alpha ]}_D^{20}\text{-}504\pm 5$ (c 0.07, EtOH), De Luca and co-workers¹⁷ reported ${[\alpha ]}_D^{20}\text{-}514$ (c 0.07, MeOH).

(−)-20-epi-Minovincinine ((−)-20-epi-3).

PtO₂ (6.8 mg, 0.025 mmol) was suspended in a solution of (−)-23 (4.0 mg, 0.0082 mmol) in MeOH (2 mL). The suspension was vigorously stirred and sparged with H₂ for 20 min at 25 °C, then was stirred for 16 h at 25 °C under a H₂ atmosphere. The reaction mixture was filtered through Celite and the filtrate was concentrated under reduced pressure. The crude residue was purified by PTLC (SiO₂, 60% EtOAc/hexane) to afford (−)-20-epi-minovincinine (2.6 mg, 89%). ¹H NMR (600 MHz, acetone-d₆) δ 9.20 (s, 1H), 7.25 (d, J = 7.3 Hz, 1H), 7.10 (td, J = 7.6, 1.2 Hz, 1H), 7.03 – 6.95 (m, 1H), 6.86 – 6.79 (m, 1H), 3.69 (s, 3H), 3.26 (p, J = 6.2 Hz, 1H), 3.21 (s, 1H), 3.09 (m, 2H), 2.94 – 2.85 (m, 2H), 2.60 (s, 1H), 2.44 (t, J = 10.3 Hz, 1H), 2.26 (d, J = 15.2 Hz, 1H), 2.01 – 1.85 (m, 2H), 1.80 (d, J = 13.4 Hz, 1H), 1.62 (m, 2H), 1.42 (d, J = 13.4 Hz, 1H), 0.84 (d, J = 6.4 Hz, 3H); ¹³C{¹H} NMR (151 MHz, acetone-d₆) δ 168.6, 168.3, 144.4, 138.8, 128.1, 122.1, 121.1, 110.3, 91.8, 67.6, 67.1, 56.4, 52.1, 50.9, 50.8, 46.6, 43.2, 27.3, 25.5, 22.7, 17.8; IR (film) ν_max 3373, 2936, 2777, 1671, 1605, 1463, 1437, 1377, 1296, 1278, 1253, 1237, 1201, 1154, 1130, 1105, 1045, 897, 795, 746, 659 cm⁻¹; HRMS (ESI-TOF) m/z 355.2026 [(M + H)⁺, C₂₁H₂₇N₂O₃⁺, required 355.2022]; ${[\alpha ]}_D^{25}\text{-}429$ (c 0.07, MeOH).