Total Synthesis of Tabernanthine and Ibogaline: Rapid Access to Nosyl Tryptamines

Abstract

We describe the first total syntheses of tabernanthine and ibogaline. Entry to these iboga alkaloid natural products is enabled by a thermal coupling of indoles and aziridines to furnish the requisite nosyl tryptamine starting materials. This route features a Friedel-Crafts type alkylation to form the key indole-isoquinuclidine C-C bond. Finally, a regio- and diastereoselective hydroboration-oxidation enables C-N bond formation to close the isoquinuclidine ring system and deliver tabernanthine and ibogaline in 10 and 14% yield respectively. Both syntheses were completed in eight steps.

ToC Graphic

The total syntheses of tabernanthine and ibogaline were enabled by application of a thermal coupling of nosyl aziridine and indoles in eight steps each, 14% and 10% yield respectively. A Friedel-Crafts type macrocyclization facilitated key C-C bond formation. Regio- and diastereoselective hydroboration-oxidation of the resultant macrocyclic alkene enabled C-N bond formation and furnished the completed isoquinuclidine ring system.

Graphical Abstract

Introduction

The iboga alkaloids are a well-known class of natural products that have received continued interest for their unique antiaddictive properties. The first anecdotal reports of their antiaddictive qualities came as early as the 1960s.^{1, 2} However, the U.S. Food and Drug Administration (FDA) classified ibogaine (the predominant alkaloid in the root of Tabernanthe iboga) as likely addictive and it received a schedule one classification. While many of the other alkaloids found in Tabernanthe iboga were discovered and characterized in the late 1950s and 1960s, their syntheses have yet to be reported. Our group is interested in uniquely oxidized iboga alkaloids and how they might be used to treat substance use disorder (SUD).³

Previously, we disclosed syntheses of the iboga alkaloids ervaoffine J and K.⁴ That synthetic campaign required 5-methoxy tryptamine (which is commercially available) as a starting material. In the present study, we focused our efforts on accessing the substituted indole iboga alkaloids tabernanthine (2) and ibogaline (3, Figure 1). In contrast to the ervaoffines, our strategy required 6-methoxy and 5,6-dimethoxy tryptamines, which are not commercially available and would require a robust synthetic approach. These two iboga alkaloids were first discovered in the 1950s and, to the best of our knowledge, have not fallen to total synthesis.^{5, 6}

Figure 1:

Representative iboga alkaloids

Results and Discussion

One of the most used methods of synthesizing tryptamines occurs via conjugate addition of indole to nitroalkenes.⁷ However, in our hands, alkylation of 6-methoxy indole (9) with N,N-dimethyl-2-nitroethen-1-amine (10) followed by reduction of the nitroalkene with lithium aluminum hydride was low yielding (Scheme 1). The poor throughput of this sequence is well documented in the literature, as both the alkylation and the subsequent reduction steps generate undesired byproducts.⁸

Scheme 1:

Synthesis of nosyl tryptamine via 3-nitrovinyl indole

As a more direct approach was desired, we turned our attention to using an aziridine as the indole coupling partner to produce substituted tryptophans and tryptamines. It is well documented that indoles undergo reaction at C3 with electrophiles when using strong Lewis or Bronsted acid catalysts. Unfortunately, strong acids facilitate the formation of indole dimers, such as indigo and indirubin, as byproducts. To mitigate this issue, the coupling of indole with aziridine has traditionally required specialized conditions. For example, indole has been coupled with aziridines incorporating electron withdrawing protecting groups on nitrogen.⁹ This scenario is showcased in the synthesis of α-aryl tryptamines where a variety of mild Lewis acids catalyze the reaction with very high regioselectivity (Figure 2a).^10–15 A second common approach is to employ lanthanide Lewis acids to catalyze the reaction of indoles with aziridines bearing electron withdrawing groups such as esters on one of the two carbon atoms (Figure 2b).^16–19 Mechanistically, placement of ester functionality on the aziridine is used to increase the yield of the alkylation and maintain regiocontrol. The least studied classes of reactions feature aziridines that are alkyl substituted. Traditionally, these reactions require strong Lewis acid catalyst for the alkylation event (Figure 2c). Few strategies exist for alkylations involving unsubstituted aziridines (Figure 2d).^{20, 21}

Figure 2:

Coupling classifications of indoles and aziridines

Due to our desired coupling of 6-methoxy indole (9) and nosyl aziridine (12) belonging to the most challenging and least precedented type of coupling, we embarked on a reaction development campaign (Table 1). Nosyl tryptamine (12) can be accessed in 73% yield over three steps through a modification of Hadjichristidis’s procedure.²² Initial attempts to catalyze the indole alkylation with standard Lewis acids such as BF₃•OEt₂ (entry 1) and AlCl₃ (entry 2) resulted in decomposition and the coupling product was not detected. The use of Grieco’s ionic liquid conditions (5M lithium perchlorate in diethyl ether) afforded trace adduct detection.²³ After initial failures in employing Lewis acid catalysis, we were encouraged to find the Hanamoto group’s report on the synthesis of β-trifluoromethyl tryptamines via thermal ring opening of trifluoromethyl nosyl aziridine.²⁴ Heating 1.5 equivalents of indole 9 with one equivalent of nosyl aziridine 12 to 110 °C overnight in toluene afforded trace formation of 11a. Increasing the temperature to 140 °C in o-xylenes provided nosyl tryptamine 11a in 40% yield. Increasing the temperature to 165 °C, using mesitylene as a solvent, did not enhance the reaction. Analysis of the crude reaction mixture provided evidence for bisalkylation. To combat this undesired reactivity, we adjusted the stoichiometry such that 6-methoxy indole was in two-fold excess. This change improved the yield to 67%.

Table 1:

Aziridine ring opening optimization

Entry	Conditions	Yield (%)
1[a]	BF3•OEt2, CH2Cl2, -78 to 0 °C	not detected
2[a]	AlCl3, CH2Cl2, -78 to 0 °C	not detected
3[a]	5M LiClO4 in Et2O, 23 °C	trace
4[a]	PhCH3, 110 °C	trace
5[a]	o-xylene, 140 °C	40
6[a]	mesitylene, 140 °C	26
7[a]	PhCl, 140 °C	40
8[b]	o-xylene, 140 °C	67

^[a]1.5 equiv. 6-methoxy indole.

^[b]3.0 equiv. 6-methoxy indole

With these conditions in hand, we explored a targeted scope of the reaction – paying specific attention to how the electron density of the indole coupling partner influenced reactivity (Scheme 2). Electron rich indoles, 5,6-dimethoxy and 5-methoxy indole, were competent in this reaction providing nosyl tryptamines 11b and 11c in 62 and 52% yields respectively. Indole itself provided nosyl tryptamine 11d in 65% yield. When examining more electron poor indoles, such as 6-chloro indole, we observed a decrease in yield to 29%. Predictably, 6-nitro indole did not provide any adduct.

Scheme 2:

Scope of nosyl tryptamine coupling

We turned our attention to the synthesis of tabernanthine (2) and ibogaline (3) (Scheme 3). Starting from 6-methoxy and 5,6-dimethoxy nosyl tryptamines (11a and 11b respectively), Fukuyama-Mitsunobu coupling with primary alcohol 14 afforded adducts 15a and 15b.²⁵ Luche reduction, followed by acetylation of the resultant allylic alcohol, gave allylic acetates 16a and 16b.²⁶ This sequence set the stage for applying our recently disclosed magnesium(II) perchlorate mediated Friedel-Crafts alkylation.²⁷ Hydroboration-oxidation of the macrocyclic alkenes gave a single regioisomer and diastereomer in both cases. The resultant secondary alcohol was mesylated and the nosyl group removed under nucleophilic conditions. Concomitant C-N bond formation closed the isoquinuclidine ring system to give tabernanthine (2) and ibogaline (3).

Scheme 3:

Total synthesis of tabernanthine (2) and ibogaline (3)

Conclusion

In summary, we report the first total syntheses of tabernanthine and ibogaline in 8 steps each with 14% and 10% overall yields respectively. Our route features an efficient nosyl tryptamine synthesis via indole-aziridine coupling under thermal conditions. This strategy of coupling indoles with nosyl aziridine is significant to our program as it eliminates two steps from our previous general synthetic strategy. Future directions of this project include the synthesis of differentially oxidized iboga alkaloids and their use as probes to further understand substance use disorder. Results in this regard will be reported in due course.