The iboga alkaloids
This academic book chapter (2017) offers the latest insight into the iboga alkaloids and related compounds (e.g. 18-MC) in order to update knowledge on the most recent advancements in the field.
Authors
- Lavaud, C.
- Massiot, G.
Published
Abstract
Iboga alkaloids are a particular class of indolomonoterpenes most often characterized by an isoquinuclidine nucleus. Their first occurrence was detected in the roots of Tabernanthe iboga, a sacred plant to the people of Gabon, which made it cult object. Ibogaine is the main representative of this class of alkaloids and its psychoactive properties are well documented. It has been proposed as a drug cessation treatment and has a wide range of activities in targeting opioids, cocaine, and alcohol. The purpose of this chapter is to provide a background on this molecule and related compounds and to update knowledge on the most recent advances made. Difficulties linked to the status of ibogaine as a drug in several countries have hampered its development, but 18-methoxycoronaridine is currently under evaluation for the same purposes and for the treatment of leishmaniasis. The chapter is divided into six parts: an introduction aiming at defining what is called an iboga alkaloid, and this is followed by current knowledge on their biosynthesis, which unfortunately remains a “black box” as far as the key construction step is concerned. Many of these alkaloids are still being discovered and the third and fourth parts of the chapter discuss the analytical tools in use for this purpose and give lists of new monomeric and dimeric alkaloids belonging to this class. When necessary, the structures are discussed especially with regard to absolute configuration determinations, which remain a point of weakness in their assignments. Part V gives an account of progress made in the synthesis, partial and total, which the authors believe is key to providing solid solutions to the industrial development of the most promising molecules. The last part of the chapter is devoted to the biological properties of iboga alkaloids, with particular emphasis on ibogaine and 18-methoxycoronaridine.
Research Summary of 'The iboga alkaloids'
Introduction
The paper defines "iboga alkaloids" at four overlapping levels: alkaloids isolated from Tabernanthe iboga; psychoactive alkaloids used in Central African ceremonial contexts; a chemical definition based on an indole nucleus fused to an isoquinuclidine system; and a biogenetic definition tied to specific rearrangements of secologanoside-derived carbons. Lotsof and colleagues adopt a biogenetic nomenclature for the review and restrict coverage to alkaloids built on the ibogan, isoplumeran and isoeburnan skeletons. The authors note that only a small subset of Apocynaceae genera produce these alkaloids and that, among roughly 100 known iboga-type compounds, catharanthine and ibogaine have driven the most chemical and biological interest, the latter for its putative anti‑addiction properties. This chapter aims to update chemical, biosynthetic, structural, synthetic and biological knowledge of iboga alkaloids from literature published roughly between 2000 and 2016. In doing so, the authors emphasise unresolved questions in biosynthesis, challenges in absolute configuration determination and the continued discovery of new monomeric and dimeric iboga derivatives, while signalling the medicinal interest in ibogaine and synthetic analogues such as 18‑OMe‑coronaridine (18‑MC).
Methods
The extracted text does not present a formal methods section because the work is a narrative, topical review rather than an original experimental study. The authors surveyed published literature spanning about 2000–2016, chose a biogenetic nomenclature for clarity, and organised the chapter into thematic sections covering biosynthesis, structural elucidation and reactivity, newly reported molecules (monomers and dimers), synthetic approaches, analytical methods, pharmacology and toxicology, and biological screening results. Where appropriate, the authors draw on structural determinations reported in the primary literature (NMR, NOE, coupling constants, ECD, Mosher derivatives, and X‑ray crystallography including anomalous scattering for absolute configuration). They discuss biosynthetic proposals based on classical tracer work and mechanistic hypotheses (for example, Diels–Alder‑like formations and possible enzymatic catalysis), and they integrate spectroscopic evidence and, when available, chemical correlations or total synthesis that have been used to confirm structures. For biological and pharmacological topics the review summarises receptor and activity profiling, metabolite identification, forensic/analytical methods (notably LC–MS with electrospray ionisation), and a range of in vitro screening assays that were reported for new isolates.
Results
Biosynthesis and chemistry: The authors report that detailed modern biosynthetic data for catharanthine and coronaridine remain limited; most mechanistic inferences still rest on tracer studies from the 1970s. Preakuammicine and stemmadenine remain central intermediates in proposed pathways, while the highly labile dehydrosecodine has not been detected. Formation of the azabicyclic (isoquinuclidine) ring is postulated to involve a Diels–Alder‑like step, and the possibility of an enzyme‑mediated Diels–Alderase is noted as an area of interest because such enzymes are rare. Structural elucidation: Advances in high‑field NMR, combined with NOE experiments, coupling constants, Mosher derivatisation, ECD (including ab initio calculations and the exciton chirality rule) and X‑ray crystallography, have improved assignments for many recent isolates. Nonetheless, absolute configuration remains unresolved for numerous compounds; the authors highlight four principal approaches (NMR/Mosher, circular dichroism/computational ECD, X‑ray with anomalous scattering, and total/partial synthesis) and note that extraction artefacts and mass spectrometric misbehaviour (e.g. [M‑2] fragments) complicate structural work. New natural products: The review documents continued discovery of new iboga‑type alkaloids, reporting about 100 compounds overall. Many new monomers are closely related to ibogamine or coronaridine, differing by oxidation at C‑19, C‑10/C‑11 of the indole, or in the isoquinuclidine fragment. Examples include several 19‑hydroxy and 19‑keto derivatives, hydroxy‑indolenines, and numerous 3‑substituted derivatives that may sometimes reflect extraction artefacts. Structural rearrangements (seco derivatives and ring‑contracted variants) and rare skeletons (isoplumeran, isoeburnan) are described, with several structures secured by X‑ray analysis. Dimers: The authors report 49 newly described dimeric (bisindole) alkaloids since the prior review, adding to roughly 40 earlier bisindoles. Most dimers combine a conserved vobasinyl residue with a diverse iboga moiety, and linkages commonly occur at C‑3'. Structural elucidation of dimers relies on HR‑MS and comprehensive 2D NMR, with NOE and HMBC correlations used to assign connectivity and relative stereochemistry; only a minority have had absolute configurations determined. Synthesis: Total synthesis remains of academic interest rather than routine practice for structure proof, since many target compounds remain available from natural sources. The review summarises several modern synthetic strategies: biosynthesis‑inspired routes stabilising dihydropyridine intermediates, Pd/Ag‑mediated cyclisations (and efforts to render these catalytic), Heck cyclisations, asymmetric organocatalytic construction of the isoquinuclidine system, and an example of a high‑yielding multistep total synthesis of voacangalactone. The synthesis of 18‑OMe‑coronaridine (127) is highlighted as short, convergent and enantioselective. Analytics and metabolism: For forensic and pharmacokinetic investigations, LC–MS with electrospray ionisation is the method of choice to discriminate iboga alkaloids. Ibogaine is rapidly metabolised to noribogaine (12‑hydroxy‑ibogamine) via CYP2D6; under daylight both ibogaine and noribogaine convert rapidly to other alkaloids (reported photolysis half‑lives of ~81.5 min for ibogaine and ~11 min for noribogaine in the cited study). Pharmacology and toxicology: Ibogaine exhibits a complex pharmacological profile, interacting with μ and κ opioid receptors, σ1/σ2 sites and showing affinity at serotonin and dopamine receptors. Functional studies described ibogaine as a weak μ‑opioid receptor antagonist (no agonist effect in rat thalamic membranes). Inhibition of acetylcholinesterase by ibogaine was found to be weak (IC50 ca. 520 ± 40 μM). Proteomic studies suggest ibogaine induces global metabolic adaptations and antioxidant responses in model systems. Clinically relevant safety concerns are emphasised: reports of fatalities following ibogaine administration exist, with suspected neuro‑ and cardiotoxicity among possible mechanisms; however, the forensic literature does not point to a single clear causal pattern. 18‑OMe‑coronaridine (18‑MC): 18‑MC, a synthetic coronaridine analogue, retains activity in preclinical models of drug self‑administration without tremorigenic side effects. It displays micromolar affinity for κ, μ and σ2 opioid receptors and 5‑HT3, no NMDA or serotonin transporter affinity, and notable activity at α3β4 nicotinic receptors. The compound has also shown activity in reward circuitry and antiparasitic activity against Leishmania amazonensis in preclinical work. A small company has pursued IND‑enabling studies and early human safety work in Brazil. Bioactivity screens: New isolates have been screened in a variety of in vitro assays. Several iboga alkaloids (voacangine, isovoacangine, coronaridine and a coronaridine hydroxyindolenine) inhibited Wnt signalling with IC50s in the ~10 μM range; coronaridine showed cytotoxicity near 2 μM in one assay. Some bisindoles and monomers displayed low‑micromolar cytotoxicity against cancer cell lines, with a few examples compared in potency to cisplatin in vitro. Additional activities reported include CB1 antagonism (submicromolar IC50s), acetylcholinesterase inhibition by select isolates, and antiplasmodial activity for one oxidised tabernaelegantine derivative (IC50 4.4 μM) with low cytotoxicity.
Discussion
The authors interpret their survey as evidence that iboga alkaloid chemistry and biology remain vibrant fields with unresolved fundamental questions and continued discovery. They emphasise that biosynthetic understanding is incomplete—modern omics and direct characterisation of proposed unstable intermediates are lacking—yet advances in spectroscopic methods and selective X‑ray studies have clarified many new structures. Lotsof and colleagues underline the recurrent problem of absolute configuration assignment for many isolates and note that extraction or chromatographic conditions can produce artefacts that complicate interpretation. In pharmacology, the review places particular weight on the complex, multisite pharmacology of ibogaine and its rapid metabolism to noribogaine, making the origin of both therapeutic and adverse effects difficult to disentangle. Safety concerns—reports of fatalities, possible cardiotoxicity, and neurotoxic signals in animal studies—are presented as a central barrier to clinical development of ibogaine itself. Against that backdrop, the authors highlight 18‑MC as a promising synthetic alternative that retains desirable anti‑addiction activity in preclinical models without some of the tremorigenic liabilities, and note active commercial efforts to advance it through IND‑enabling and early human safety studies. Key limitations acknowledged in the chapter include incomplete biosynthetic data, frequent reliance on spectroscopic rather than synthetic confirmation of absolute stereochemistry, potential artefacts during isolation and purification, limited mechanistic clarity for observed in vitro activities, and a predominance of screening data without extensive in vivo pharmacology or safety characterisation for many new compounds. The authors suggest that future work should combine modern biosynthetic tools, rigorous stereochemical assignment approaches, improved analytical characterisation of plant extracts and metabolites, and more systematic pharmacology and toxicology to assess therapeutic potential and risks.
Conclusion
Iboga‑type alkaloids are a chemically distinctive group of indolomonoterpene alkaloids notable for their diverse biological activities and, particularly, for the long‑standing interest in ibogaine as a potential agent for drug cessation. The authors conclude that, despite safety concerns and incomplete mechanistic understanding, ongoing discoveries of new natural products, advanced structural and synthetic methods, and the development of synthetic analogues such as 18‑MC justify continued research. They frame the work as contributing to the foundation for innovative studies that could, in time, translate into safer and clinically useful therapies inspired by Tabernanthe iboga.
View full paper sections
STRUCTURAL ELUCIDATION AND REACTIVITY
The widespread availability of high-field NMR spectrometers and of sophisticated pulse sequences now enables chemists to determine structures of moderate complexity, such as the iboga alkaloids and even the "dimers", leading to full ? ? Fig.The involvement of preakummicineand of precondylocarpinein the biosynthesis of iboga alkaloids unambiguous proton and carbon NMR assignments. Relative configurations in rings and ring junctions may be dealt with NOEs and coupling constant measurements. In these series, the only open chain fragment is the ethyl side chain, where chirality is present at C-19. The problem of its determination is best solved with one of the Mosher methods, an example of which is found in the structure elucidation of ervatamine F (12) (Fig.). In addition, X-ray crystallography allows the solving of structural problems whenever a single crystal is available and Sect. 4 lists several recent examples of such structural determinations. Catharanthinol, dihydrocatharanthinol, and N-hydroxymethylibogaine (15) have had their structure solved by X-ray crystallography, but unfortunately without absolute configuration determinations. The problem of the absolute configuration of these alkaloids remains a difficulty and there are currently four methods available for this purpose. NMR spectroscopy is one of these, provided one stereogenic center is linked to the others and determined using Mosher derivatives. This procedure has often been used to determine the configuration of C-19, which unfortunately cannot be easily linked with the remaining stereogenic centers. Circular dichroism is a general and sensitive method, which has been used to differentiate compounds in the (+)-catharanthine and (À)-ibogaine series. It is now used to determine the absolute configurations of the linkages in dimers by application of the exciton chirality rule, which was first proposed by Harada and Nakanishi, and then later used by others. Ab initio calculation of electronic CD spectra and comparison with the experimentally obtained values also allows the determination of absolute configurations, but so far most examples have dealt with monomers. X-ray crystallography using anomalous scattering is the most unambiguous method. It is more and more frequently used to solve structures. Finally, partial or total synthesis has often been used to link a new alkaloid to another of known absolute configuration. Chemical reactivity, and partial and total synthesis, were until recently integral parts of the structural elucidation process. Simple reactions were used to determine functionalities or to prepare crystalline derivatives. As a bonus, it often happened that new and unexpected reactions were discovered during the process. With the advent of highly sensitive spectroscopic techniques, this approach is almost no longer utilized, and only one example could be found in the more recent literature of a total synthesis, aimed at establishing the structure of voacangalactone (). The most intriguing aspects of the iboga alkaloids are their unusual oxidation reactions, which arise from the particular geometry of the isoquinuclidine ring system. The original experiments were described by Bartlett and Taylor, who transformed coronaridine (3) (Fig.) with iodine in aqueous methanol into the 3-hydroxy-and 3-oxo-coronaridine derivatives. These reactions went unnoticed for a long time, even though their products were later shown to be naturally occurring substances. Eglandine (17) (3-OH coronaridine) is one example, but its structure was assigned initially as containing a tetrahydro-oxazole ring and later had to be corrected. The present chapter authors long questioned the natural occurrence of these tetrahydro-oxazoles until definitive proof of their existence came from the X-ray structure of oxindole 57 (ervaoffine A). While working on this chapter, the authors corresponded with Professor Hiromitsu Takayama, whose group isolated the 3,6-oxido voacangine, and he provided good NMR spectroscopic evidence for the cyclic structure of oxidovoacangine. However, the probability is high that derivatives with substitutions at C-3 (OMe, OEt, or acetonyl) originate from the intermediate carbinolamines. Interestingly, these do not lend themselves to reduction or Pictet-Spengler-like reactions although they can be considered as masked aldehydes. One of the reasons why these particular structures are not easily identified is that they "misbehave" in mass spectrometers and often show [M-2] fragments (a result of a dismutation). As a consequence, there is a considerable tendency to make an extra bond as in ervaoffine Aor as in ervataine. Ervataineclearly violates Bredt's Fig.Oxidation of coronaridine at position 3 and "structure" of ervataineruleand its most probable structure is 3-OH ibogaine (C-3 showed a 13 C NMR resonance at 93.2 ppm). More classical oxidations arise at C-7, yielding the pseudo-indoxyl chromophore after rearrangement or at C-2 as an intermediate to oxindole. The addition of dioxygen on the double bond of indole and fragmentation produces N-acyl derivatives as observed in the degradation of tryptophan into kynurenine. All these chromophores are detected in natural products (vide infra) and these products, if they are not extraction artefacts, may be considered as end products in biosynthesis processes (Fig.). The most studied oxidation in the ibogan series is related to the N-oxides, which are easily prepared under acidic peroxide conditions. While in the ibogaine/coronaridine series the N-oxide is relatively stable, the N-oxide of catharanthine (20) (Fig.) lends itself to a Meisenheimer rearrangement, bringing about a N-4-C-3 bond cleavage. Such products have been observed but knowledge of their biosynthesis remains at the hypothetical stage. Fragmentation of the N-oxide of catharanthine is the key reaction in the synthesis of anhydrovinblastine and the related antitumor alkaloids vinblastine, vincristine, vinorelbine, and vinflunine. The subject has recently been reviewed and the interested reader will find an account of this chemistry involved in books by Langlois and Langloisand by Gue ´ritte and Le Roux. The formal addition of HCN across the C-16-C-21 bond of catharanthine (1) provides an intermediate aminonitrile, which allowed Beatty and Stephenson to bridge the gap between this series and the pseudo-tabersonine and -vincadifformine alkaloids. The initial reaction was brought about by visible light irradiation of 1 in the presence of the polyfluorinated catalyst Ir(dF(CF 3 ) ppy) 2 (dtbbpy)PF 6 and with trimethylsilyl cyanide as reactant (Fig.). Remarkably, this reaction was run in a flow reactor in order to better control reaction parameters and confine the hazards linked to HCN generation. Although the conversion to pseudotabersonineonly required iminium generation and its "through nitrogen" isomerization, it took carefully controlled acidic conditions to achieve the transformation. One of the unexpected difficulties was a partial racemization linked to C-14 equilibration, and treatment of 21 with TFA led to an almost complete conversion into 24. The optical rotation value and chiral HPLC parameters of the reaction product showed it to be a 2:1 mixture of enantiomers. While it was initially thought that these difficulties would be absent in the dihydro series, pseudovincadifforminewas not obtained but (+)-dihydrocatharanthine (3) occurred instead. Compound 23 was eventually reached in good yield through oxidative photoirradiation of 22, the over reduction product of compound 21. Among other things, this is to our knowledge, one of the rare articles daring to state that catharanthine (1) is a good starting material for partial synthesis due to its natural availability, a fact which is largely overlooked.
NEW MOLECULES
The list of compounds biogenetically linked to the iboga unit continues to be fed by new alkaloids but among the three skeletons of this group, the isoplumeran and isoeburnan variants remain scarce. This Section is divided into two parts: the monomers and the dimers, and, in each, molecules are grouped according to oxidation levels. The Apocynaceae family has continued to provide new occurrences, but at variance with other much more common indolomonoterpenes, the isolation of alkaloids of the iboga class is limited to only a few species in the genera Ervatamia, Tabernaemontana, Voacanga, and Catharanthus, with Ervatamia being considered as a synonym of Tabernaemontana by many authors (Tablesand). "The Plant List"was chosen as the botanical reference. The total number of species mentioned below does not exceed ten for a total of a little more than a hundred compounds. This may be the tip of the iceberg since the number of Tabernaemontana, Ervatamia, and Voacanga species is close to 700. The limited presence of iboga alkaloids in these particular genera is probably a consequence of their elaborate biosynthesis, pertaining to well-evolved plants. Alkaloids are considered here as new when not mentioned in previous reviews.
IBOGAMINE AND CORONARIDINE DERIVATIVES
Many of the newly isolated structures may be viewed as simple derivatives of ibogamine or coronaridine, with a limited range of oxidations. Typically, these variations concern oxidations at C-19 in the side chain, C-10 and C-11 of the indole nucleus, and more rarely C-15 and C-20 in the isoquinuclidine moiety. Molecules with substituents at C-11 remain scarce. Four new 19-hydroxy derivatives have been reported: (19S)-hydroxyibogamine (25) (Fig.), 19-epi-isovoacristine, and ervatamines Aand H (). The relative configuration of C-19 was determined by comparison of the 13 C NMR chemical shifts of C-15 and C-21, as proposed by Wenkert in 1976. The problem of absolute configuration was ignored in Ref., while ECD provided a solution with ab initio calculations for compound 28, and a comparison with ECD of coronaridine for 27. C-ring contraction is a very rare occurrence in the biosynthesis of indole alkaloids, for which the origin is a C-5-C-6 Polonosvski fragmentation. Ervatamine Aadds to this very short list and may be considered as a collateral pathway in the reaction of the corresponding ibogan Noxide. This is the second ibogan-type alkaloid isolated where ring C is contracted to form an unusual six-membered ring as in flabelliformidine. Ervatamine H (has a very unusual oxidation of C-15, which could also be envisioned as a hydration of a Δ 14 double bond. Observation of a NOE correlation between H-18 and H-15, and ECD comparison with coronaridine allowed the (14R,15R,16S,20S,21S)-configuration to be proposed. Five new alkaloids were found to contain an additional degree of oxidation and a ketone at C-19: isovoacryptine (29), conodusines A-C (30-32), and (À)ervatamine I (33), for which an isomer was named (+)-conodusine E (Fig.). The report on conodusine E was published subsequent to that on ervatamine I (33), and, given the similarities between the data for the two compounds, except for their optical rotations, great care was accorded to the determination of the relative and absolute configurations. First, a chemical correlation was carried out with (À)heyneanine (Dess-Martin oxidation), then the ECD was measured and compared with the calculated ECD values for the two enantiomers, and finally an X-ray analysis was performed. All these experiments converged to the structure depicted as 33. The Malaysian authors proposed that the so-called ervatamine I is an enantiomer of their own compound and therefore an alkaloid of the catharanthine series. However, the present authors feel that this hypothesis creates a precedent: the isolation of a catharanthine-like compound in a Tabernaemontana species. The sole rationale given for the absolute configuration of ervatamine I is the sign of the optical rotation (the same as coronaridine), which is an argument to be considered of little value at the present time. Conodusines A and B are isomeric at C-20, a position α to a carbonyl, and hence prone to isomerization. The possibility of conodusine B being an artefact was discussed but could not be established definitively. The same possibility holds for conodusine C (32), the N-oxide of conodusine B. Voacanga africana is one of the most highly investigated West African medicinal plants. From this species, a Japanese group isolated a new bioactive alkaloid, voacangalactone (16) (Fig.), for which the structure was determined by spectroscopic means and by total synthesis. All alkaloids with hydroxy groups at positions C-15, C-19, and C-20 may stem from precursors with a double bond between C-19 and C-20 or C-15 and C-20, isomeric with catharanthine (1), and until now never detected in this series. Fig.The 19-keto derivatives 29-32 Nucleophilic attack on C-11 is favored when C-10 is substituted by a methoxy group, which is the situation for ibogaine, and leads to a group of iboga alkaloids diversely substituted at C-11, all found in Tabernaemontana corymbosa from Malaysia. These were: cononusine (34) with a pyrrolidone unit, and the lignan conjugates conoliferine and isoconoliferine, conomicidines A and B (36), and isoconomicidines A and B (37) (Fig.). These included pairs of diastereomers that were not separated even though the authors discussed at length the relative configurations of the new chiral centers. The question of the absolute configuration of the iboga moiety was not a subject of debate, assuming that the configuration is prevalent in Tabernaemontana. In cononusine, the relative and absolute configurations of the chiral center in the pyrrolidone ring could not be determined. The lignan conjugates were assumed to arise from a nucleophilic attack of ibogaine on a quinone methide derived from one or two coniferyl units.
-ALKYL-OR 3-OXO-IBOGAMINE/-CORONARIDINE DERIVATIVES
These alkaloids are, for most of them, oxidized forms of known compounds. Alkaloids 38-40 (Fig.) are all 7-hydroxylated indolenines, further oxidized as lactams at position C-3. They derive, respectively, from ibogaine, coronaridine, and conodusine A (20-oxo-ibogamine), although no chemical correlations were attempted as structural proof. The reactions leading to these compounds are well established in the literature on indolomonoterpenes and their biosynthesis follows the usual oxidation pathways. The most well-documented structure is 3-oxo-(7R)-coronaridinehydroxyindolenine, which was determined by X-ray diffraction, with Flack parameter calculations, thus making the absolute configuration definite. The two other compounds have had their configuration determined by comparison of ECD spectra (38: λ max (Δε) 230 (À9.0), 258 (+18.6), 289 (À4.4) nm; 39: λ max (Δε) 221 (+26.3), 260 (À10.9), 294 (0) nm; 40: λ max (Δε) 221 (+11.6), 257 (À5.7), 289 (+1.7) nm). It would have been advisable to have the ECD also recorded for tabernaricatine F, for which the configuration of C-7 was proposed after observation of a NOE between OH and H-16. Following the same biosynthesis, three other hydroxy indolenines (41-43) (Fig.) have been encountered in the coronaridine/ibogaine series, displaying further oxidation at C-3. Compounds 41 and 42 share the same origin as 39, from which the hydroxylated derivative 41 is an obvious precursor. The cyano and acetonyl derivatives 42 and 43 arise from substitution of their respective 3-hydroxy precursors and while cyano derivatives are common in Nature, the origin of the acetone fragment is more dubious. It is not unreasonable to think that it comes from the solvent used for chromatography, with silica gel acting as an acid catalyst. Observation of the suitable Cotton effects for 41 and 42 allowed their configuration to be established as (7S). The configuration at C-3 was deduced from NOE correlations observed between H-3$H-6β$H-17β, placing the H-3 to the "left", towards the respective indole ring. Besides these cyano, hydroxy, and acetonyl derivatives, there are a large number of monomeric iboga alkaloids with miscellaneous alkyl or alkoxy substituents on C-3. The 3-alkoxy derivatives, (3R/S)-ethoxyheyneanineand (3R/S)-ethoxy-19-epi-heyneanine (45) (Fig.), were isolated as mixtures of diastereomers, and could be artefacts formed by addition of ethanol on an iminium intermediate during the extraction steps. Similar NOE effects were observed for four 3-alkyl derivatives: ervatamine G (46), tabernaricatine G (47), (3S,24S)-and (3S,24R)-hydroxyethyl-coronaridine (48 and 49), suggesting a (3S*) configuration. It is worth noting that these last two compounds were isolated from Ervatamia hainanensis roots in 1982 but their configuration could not be established at that time. An X-ray structure determination for the (24S) compound 48 secured the relative and absolute configurations for the two isomers. Their ECD spectrum exhibited similar Cotton effects as observed for coronaridine, confirming their common (3S,14R,16S,20S,21S)-configuration. This was confirmed by preparing Mosher esters of the C-24 alcohols. The Iboga Alkaloids
-AND/OR 6-OXO-IBOGAMINE/-CORONARIDINE DERIVATIVES
Positions α to the basic nitrogen atom are prone to oxidation and in the iboga alkaloids, this most often occurs at C-3. 19-epi-5-Oxovoacristine (50) from Ervatamia officinalis is an exception, and the location of the amide carbonyl is based on HMBC correlations and on the presence of an isolated AB quartet for CH 2 -6. The remainder of the asymmetric centers was determined by 13 C NMR chemical shifts (for C-19) and NOE correlations, and the absolute configuration was established by ECD (experimental vs. calculated). The location of the carbonyls in the 5,6-diones of ibogamine and ibogaine (51)and (52)was based on observation of HMBC correlations between H-3 and H-21 with O ¼ C-5 (lactam near 169 ppm). The ion at m/z 170 (m/z 156 in ibogamine) in their mass spectra also suggested the presence of an oxygen atom at C-6. The ECD spectrum of 52 was measured and compared to the calculated value, which confirmed the (14R,16R,20S,21S)-configuration. 5-Oxo-(6S)-hydroxy-and 5-oxo-(6S)methoxycoronaridine (53 and 54) are two other new products obtained from Ervatamia hainanensis. An X-ray structure was obtained for the alcohol 53, which allowed its full structural determination, and, as a consequence, establishment of the structure of 54. The simplest compound 55 (6-oxo-ibogaine) is unique and unexpected. There are two arguments in favor of oxidation at C-6: a downfield shift of H-9 due to carbonyl anisotropy and HMBC correlations of H-5 (two d, J ¼ 18.1 Hz) with C-3, C-7, and C-21. This compound has been an intermediate in a synthesis of ibogaine, but this is the first time that it was detected as a natural product(Fig.). In this group of alkaloids, the present authors have chosen to place those in which one bond of the skeleton is cleaved, thus giving four rearranged types of structure: the 2,7-seco-, 6,7-seco-, 2,16-seco-, and the N(4),21-seco-iboga derivatives. Ervaoffine D (56) (Fig.) is the only described iboga alkaloid with the 2,7-bond cleaved into a ketone and an amide. Its structure was proved unequivocally by X-ray crystallography inclusive of the absolute configuration. Among the 6,7-seco derivatives, the structures of ervaoffine A (57)and ervatamine F (12)were confirmed by X-ray crystallography and the absolute configurations were found to be (2S,3S,6S,14R,16R,20S,21S) and (2S,14R,16S,19S,20S,21S) (ibogaine series). The priority order of C-3 and C-6 is modified by the introduction of the hydroxy group on position C-6 and these two same (2S)-configured compounds indeed have opposite spatial configurations at C-2. Iboluteine, also produced by the same plant, and the deoxy analog of ervaoffine A, have also been shown by X-ray crystallography to have the opposite configuration, and so 58 is not the precursor of 57. In addition, measurements and calculations of ECD spectra for 57 were performed, allowing comparison to be made with alkaloids bearing the same chromophore. Mosher esters of 12 were prepared to confirm the C-19 stereochemistry. The third alkaloid, 59, was isolated by two different groups in 2014 and 2015, and although published in the same journal, it was given different trivial names: ervaoffine Band ervaluteine, when it could be simply have been named (6R)-hydroxy-(2S)-pseudo-indoxylibogaine. It is interesting to note that the absolute configuration was given in the first paper, since the ECD spectrum was similar of that of ervaoffine A (57)and a NOE was detected between OH-6 and H-3α and H-17β. In the second paper, the absolute configuration was deduced painfully from NOE correlations, comparison with (2R)-iboluteine, and examination of molecular models. Only one 2,16seco-ibogamine, ervaoffine C (60), was characterized, and, as for the other ervaoffines, its absolute configuration (3S,6R,7R,14R,16S,20S,21S) was established The Iboga Alkaloids from the similarity of Cotton effects with those obtained by calculated ECD for both enantiomers. Tabertinggineand voatinggine (62) (Fig.) are two exceptionally rearranged iboga alkaloids, for which their skeletons remain unique, and the X-ray crystallography used for both removed all doubt on their structures. A common biogenetic pathway was proposed for these compounds with initial cleavage of the C-21 to N-4 bond of conodusine A (30) and subsequent ring closure onto C-2 or C-3.
CATHARANTHINE AND PSEUDOEBURNAMONINE DERIVATIVES
All the monomers described until now in this Section are assumed to belong to the ibogaine/coronaridine series, whether demonstrated or not. There is but a single possible exception, alioline (63) (Fig.), isolated from Catharanthus roseus. This is a strange and unique molecule possessing the features of catharanthineto which is added a C 9 unit. However, if structure 63a is proposed in the original article, the "Dictionary of Natural Products" changed it to structure 63b under the same name and reference. Fragments I and II display the possibilities for the extra carbon atoms. The origin of the C 9 fragment is unclear: it could be a truncated terpene, although highly irregular, or an iridoid, but in this case, one of the methyl groups is missing. There is a single example of a new molecule with the isoeburnan skeleton, named tacamonidine, while just two reports of the isolation of pandine (64) (Fig.) are found in the isoplumeran categories. Tacamonidinebelongs to the rare isoplumeran series and differs from tacamonineby an OH group, for which the configuration was proposed by NMR spectroscopy.
MISCELLANEOUS REPRESENTATIVES AND ANOTHER ENIGMA
Four new alkaloids were isolated from Tabernaemontana corymbosa and named tabercarpamines G-J (65-68) (Fig.). They belong to the small family of the chippiines and are characterized by an N-1 to C-3 bond instead of the N-4 to C-3 bond. Their structures were established by NMR spectroscopy without any attempt at the determination of their absolute configurations. In tabercarpamine H, there is an OH at position C-19, the configuration of which was proposed on the basis of a ROE measurement. It would have been safer to make this determination on a partially synthetic derivative of 66, in which, for example, the free OH and the secondary amine would be engaged in a six-membered ring. The structures of the tabercarpamines are not exceptional but their mere existence is puzzling. It is not difficult to conceive a biosynthesis scheme starting from isovoacangine or coronaridine (3) and giving the tabercarpamine skeleton through the intermediate of a 3-hydroxy derivative and of an elusive aldehyde, 69 (Fig.). Whereas one would expect some stability of this aldehyde, it has never been identified and one hypothesis would be that it exists in such a high energy state that the seemingly more strained isoquinuclidine forms are predominant.
DIMERS
The count of new dimeric alkaloids containing an iboga moiety amounts to 49 (Table), which adds to the previous 40 such compounds described in an earlier review. This is therefore quite a well-represented class of alkaloids, and it appears that all but one contain the vobasinyl residue always substituted at position C-3 0 (vobasine numbering). In all examples also, the ibogan moiety is linked on the α face, which is deduced from the multiplicity of H-3 0 (dd, J ¼ 13 and 3 Hz), and from its NOE with H-15 0 and the indole NH. Nature utilizes a large variety of reactions to couple indole alkaloids, among which, coupling between an electronrich indole nucleus and a carbenium ion is a privileged route. The reason why the vobasinyl cation is so common in couplings remains to be explained. It could be that, being stabilized by N-4, it is an especially long-lived species lending itself to slow kinetics reactions. Most of the newly isolated bisindoles are derived from five vobasan-type monomers (Fig.): vobasinol (70) (syn. perivine), dregamine, vincadiffine (72) (3-oxo tautomeric form), pagicerine, and difforlemenine. For the iboga moiety, the chemical diversity is more extensive but it is possible that chemical modifications in this moiety occur after coupling. The ibogan skeleton is equally represented by the voacangine or ibogaine series (with or without a carbomethoxy group at C-16) and substitution preferentially occurs ortho to the aromatic group if present. Compounds without an OH-or OMe-directing group on the aromatic part of the ibogan molecule are scarce but over the years several have been isolated. Two alkaloids stand alone in the gallery of the iboga bisindoles (Fig.). One is biscarpamontine A (75), which is the result of a unique coupling between the iboga and aspidosperma units. A further particularity is the presence of a methylene bridging the two moieties and assigned to the iboga moiety, according to a biogenetic hypothesis from the original authors. As far as we know, there are no natural iboga alkaloids possessing N-1-methyl or formyl groups and it is our belief that this supernumerary carbon atom belongs to the aspidosperma moiety (as the one found in the vobtusines is also present in the same plant). The overall structure of 75 was established by NMR spectroscopy and only relative configurations are given. Vobatensine E (76) is the only dimeric iboga alkaloid to have a linkage with the C-9 of ibogaine and this was deduced from analysis of the 1 H NMR spectroscopic features of the aromatic part of the molecule. An attempt to prepare this (coronaridine-type) (dippinine-type) 69 ? Fig.A possible pathway between coronaridine-type and chippiine-type alkaloids compound by condensing vobasinol and ibogaine under acidic conditions was unsuccessful. This substitution pattern is so far unique. In the C-11-OMe series, however, condensations occur on both sides of the ArOMe (C-10 and C-12). Generally speaking, the structural elucidation strategy used for all bisindoles has been similar. The first step consists of recording high-resolution mass spectra (HR-ESI-TOF or HR-FAB-MS) and carbon NMR spectra to determine the nature of two moieties. Following this, a combination of COSY, HSQC, and HMBC NMR experiments allows the determination of the structures of the two monomers, which often are known, or by may be proposed by comparison with similar monomers and dimers. Finally, analysis of NOE and further HMBC correlations helps determine the linkage positions between the two monomeric moieties and the relative stereochemistry of some chiral centers. Very few bisindoles have had their absolute configuration determined (see below).
BISINDOLES WITH AN IBOGAMINE MOIETY
There are only four bisindoles that lack an alkoxy substituent on the aromatic ring of ibogamine: (19R)-and (19S)-hydroxytabernamine (77 and 78), and tabernamidines A and B (79 and 80) (Fig.). Comparison of the NMR spectra of 77 and 78 with those of the heyneanines was key to the C-19 configuration determinations. A compound named originally oxotabernamineand depicted as 79 was revised to 80 following a new isolation and comparison with the spectra of the corresponding monomers, conodusines A and B. Enolization of C-19 could explain the formation of the tabernamidines as seen in these last compounds. Vobatensine A () is the C-11-OMe analogue of (19R)hydroxytabernamine and its structure was proven definitively by a partial synthesis from vobasinol and 19-epi-iboxygaine. The same plant, Tabernaemontana corymbosa, yielded the dihydro derivative 82, which was named vobatensine B, and was also prepared from tabernaemontaninol and 19-epi-iboxygaine. The configuration at C-20 0 was deduced from the following NOE correlations: H-19 0 $H-16 0 , H-18 0 $H-21 0 , and H-20 0 $H-14 0 . Ervachinines B (83) and D (84) (Fig.) are two positional isomers, composed of vincadiffine () linked to 10-methoxy-ibogamine (ibogaine (2)) or to 11-methoxy-ibogamine (tabernanthine). Their absolute configurations were deduced from the similarity of their CD curves with that of ervachinine A (85) to which was applied, for the first time in bisindoles, the exciton chirality rule. Its indole chromophores were found to be oriented in a clockwise manner with Cotton effects at 223 (À) and 243 (+) nm. Ervachinine B (83) had been previously isolated from the stem bark of Tabernaemontana corymbosa (syn. Ervatamia chinensis), and the name "ervatensine A" given in the 2008 Ph.D. thesis of K.-H. Lim was maintained, even though the work was not published until 2015. Discrepancies between the 13 C NMR spectra of ervatensine A and of ervachinine B led the Malaysian authors to perform an X-ray structure determination to ascertain the structure of ervatensine A and to establish once and for all the absolute configuration. The series was completed with ervatensine B (86)and vobatensine F (87), which are the nor analogue and the N-oxide of decarbomethoxyvoacamine, respectively. Two other similar compounds, conodutarines A and B (88 and 89) both have a less typical linkage involving C-12 instead of C-10 [51] (Fig.).
BISINDOLES WITH A VOACANGINE (10-METHOXY-CORONARIDINE) MOIETY
Besides ervachinine A (85) (see above), there has been only one other bisindole reported with a voacangine unit, conodusarine, which is the 3-oxo derivative of voacamine (91). Its absolute configuration was not, strictly speaking, proven, but rather assumed, following the concomitant isolation of the parent compound (Fig.).
BISINDOLES WITH AN ISOVOACANGINE (11-METHOXY-CORONARIDINE) MOIETY
There is a wealth of alkaloids of this type, with vincadiffine (72) and derivatives in the iboga category, or vobasinoland derivatives in a non-iboga group. Variations of the iboga type are unsurprising and represent a similar diversity observed for the monomers. However, the authors of this chapter do have a concern in the The Iboga Alkaloids past lack of care in the choice of trivial names for new compounds, with little or no effort apparently having been made to correlate those names to literature references. Cononitarine B (92)and ervachinine C (93) (Fig.)are two bisindoles composed of a 11-hydroxy-or a 11-methoxy-coronaridine moiety linked to a vincadiffine unit. As for other ervachinines, the absolute configuration of 93 wasBisindoles with a voacangine moietyBisindoles with a isovoacangine moietydeduced from the similarity of its CD spectrum to that of ervachinine A (85). In the same series, 17-acetyl-tabernaecorymbosine A (94) was isolated and its absolute configuration was confirmed by the CD exciton chirality method. Conodiparine C () is the 19-oxo derivative of ervachinine C, and its isomer conodiparine Dsimply differs in its location of attachment on the aromatic ring (C-12 instead of C-10). Cononitarine A () is (19S)-hydroxycononitarine B. The other conodiparines (A (98), B (100), E (99), and F (101)) are pairs differing in the substitution pattern of N-4 in the affinisine moiety: NH vs. N-Me. A peculiarity of the four alkaloids, conodirinines A and B (102 and 103), and tabernaricatines C and D (104 and 105) (Fig.), is the presence of an extra tetrahydro-1,3-oxazine ring bridging C-16 and N-4. The configuration assignment of the hydroxyethyl side chain was based on the 13 C NMR chemical shifts. As discussed by the authors, the signal for the N-Me group was conspicuously missing and it was the first reason to propose that it was embedded in a ring. The original papersuggested that the methylene bridge is the result of a formaldehyde (or equivalent) condensation, but this also may be the result of the oxidation of the N-methyl into an iminium and subsequent trapping in a Polonovski fashion, as suggested by Zhang et al.. Like the set of preceding compounds, tabernaricatines A and B (106, 107), and tabercorine C (108) (Fig.), also possess a tetrahydro-1,3-oxazine ring, this time incorporating C-21 rather than a N-Me functionality. The structures of 106 and 107 were proposed mainly after comparison of their NMR data and observed NOE correlations with those of conodiparine B (100). No chiroptical data were reported and the structures were considered as flat although configurations were drawn for all centers including the epoxy ring. The absolute configuration of 108 was discussed after its CD data were measured and differences between the spectra of 106 and 108 led to the conclusion that they had an opposite epoxide configuration, but unfortunately taking the configuration of 106 for granted (which is not the case). The biosynthesis scheme proposed for 108is more appealing and has the merit of explaining the origin of the other compounds as well. The Iboga Alkaloids Muntafara sessifolia (syn. T. sessifolia) yielded 3 0 -oxo-tabernaelegantines A and B (109 and 110) (Fig.), which are regioisomers differing in the substitution pattern on the aromatic ring. Their structures, including the relative configurations within the two moieties, were determined mainly by NMR spectroscopy. These are the oxidation products of the well-known tabernaelegantines, although this has not been proven. A particular feature of the NMR spectra was a linebroadening phenomenon, said by the authors to disappear upon cooling at 273 K. It was even stated that "lowering the temperature favored one conformer", but this is not reasonable from a thermodynamic standpoint. Four 3-hydroxy-tabernaelegantines were characterized, with three of these having a C-12-C-3 0 bridge: (3R)-hydroxytabernaelegantine A (111), (3R)hydroxytabernaelegantine C (112), (3S)-hydroxytabernaelegantine A (113), and (3S)-hydroxytabernaelegantine C (114). (3R)-Hydroxytabernaelegantine C (112) was known previously as a semisynthetic bisindole obtained by acidic hydrolysis of tabernaelegantinine C (115). The configuration of C-3 was deduced from its 13 C NMR chemical shift value (i.e. near 86 ppm in the (S)-configuration and up to 93 ppm in the (R)-configuration). The interconversion of these carbinolamines under acidic conditions such as on the silica gel used for purification could explain the fact that hydroxytabernaelegantine A was isolated initially as a mixture of epimers, 111 and 113. The configuration of the ethyl chains in the vobasinyl units was determined from the chemical shifts of C-14 0 , C-16 0 , C-18 0 , C-19 0 , and C-20 0 , and a NOE correlation observed between H-3 0 $H-20 0 . (3R)-Hydroxytabernaelegantine D (116) was the sole carbinolamine bisindole containing a C-10-C-3 0 bond. The propensity of C-3 of the iboga alkaloids to capture nucleophiles was illustrated by the isolation of the three tabernaelegantinals A, B, and E (117-119) (Fig.). The (R)-configuration of C-3 was determined by the observation of a strong NOE interaction between the aldehyde proton and H-5b. It is assumed that the precursor, 3-formyl-isovoacangine, was formed by reduction of a 3-cyanoisovoacangine, itself produced from a 3-hydroxyisovoacangine via an iminium form in a Strecker-like reaction. The same iminium intermediate could be scavenged by nucleophiles such as the acetylacetate anion to form 3-alkyl-isovoacangine. Examples among the bisindoles are tabercorine A (120)and tabernaricatine E (121), which is 3-oxopropyl-ervachinine C. A NOE or ROE correlation was observed between H-3$H-17β, establishing that H-3 is β-("left")-oriented. The absolute The Iboga Alkaloids configuration of tabercorine A was established by means of the CD exciton chirality method with the two indoles oriented in a clockwise manner.
BISINDOLES WITH AN IBOGA-INDOLENINE OR REARRANGED MOIETY
Dimers in the iboga series with an oxidized indole nucleus are rather rare compared to the situation observed in the monomers. Only two 7-hydroxy-indolenines, tabercorine B (122)and vobatensine D (123) (Fig.), corresponding to tabernaricatine D (105) and ervatensine B (), have been isolated recently. The (7R)-configuration of 7-OH in 122 was deduced by comparison of the NMR data with those of the hydroxyindolenine of voacangine. No ROE correlation was observed for the OH, which is not surprising since the measurements were done in acetone-d 6 , but the absolute configuration was established by the exciton chirality CD rule. Vobatensine D (123) was claimed to have the inverted (7S)-configuration (α-OH) on the basis of a dubious argument: the co-occurrence of the (7S)hydroxyindolenine of ibogaine of known absolute configuration in the plant. A bisindole has been reported with a pseudo-indoxyl chromophore, namely, (2R)vobatensine C (124). The configuration of its spirocyclic C-2 was deduced from the chemical shifts of C-2, C-7, C-16, and C-21 as performed for iboluteine. A NOE correlation was observed between the indole NH and H-17β.
BISINDOLES WITH A CHIPPIINE MOIETY
Tabernaemontana corymbosa has yielded two dimeric alkaloids, tabercarpamines A (125) and B (126), possessing the usual vobasinol moiety and a chippiine portion(Fig.). Their structures were established by NMR spectroscopy and mass spectrometry and a proposal was made for their absolute configurations based on the application of the exciton chirality rule. While it is perfectly acceptable to determine the helicity around the junction between moieties, this is based on an assumption that the configurations of the monomers have been determined already. In this example, as in others described previously, these are simply assumed based on biogenetic hypotheses rather than on any rigorous experiments.
SYNTHESIS
As mentioned earlier in this chapter, total synthesis is almost no longer in use in the structural elucidation domain. Unless major changes occur in the marketplace, most industrially or pharmacologically important alkaloids (e.g. catharanthine (1) and ibogaine (2)) will continue to be available competitively on a large scale starting from natural materials. An exception is the 18-OMe derivative of coronaridine (127) (Chart 1), which is in the pre-development phase and, in the absence of a reliable source of its precursor, albiflorine (18-OH coronaridine), will remain a fully synthetic compound. Total synthesis in the area is justified by the preparation of unnatural analogues or derivatives and by the development of methods of a more general application. The Iboga Alkaloids Most of the first generation of iboga alkaloids syntheses were based on the assumed biosynthesis, and the main challenges were to build a molecule containing a reactive acrylate and an unstable dihydropyridine such as found in compound 130. A quite innovative and general route to 130 inspired by biosynthesis was recently One of the characteristics of this very clever approach is the stabilization of the dihydropyridine in the presence of the acrylate function in 130 allowing all possible Diels-Alder adducts to be obtained, thus leading to the iboga, aspidosperma, or even andranginine skeletons. As a bonus, chirality was also obtained with the introduction of a chiral oxazolidinone in 131, the precursor of the dihydropyridine (Chart 2). The synthesis follows a route similar to that used in the racemic synthesis and was illustrated by the preparation of (À)-catharanthine. Another approach follows the historical route developed by Trost, in which a bond is formed between C-2 of indole and the isoquinuclidine bearing a suitably placed double bond. The last cyclization stage (viz. 132 ! 133 in Chart 3) requires a full equivalent of palladium and another full equivalent of silver, which makes the end product more valuable than these two metals together. This is clearly a disadvantage and attempts have been made to circumvent this difficulty and make this reaction catalytic. After using the Pd/Ag sequence to prepare analogues of ibogamine such as 134 (Chart 3), Sinha introduced the Heck reaction to cyclize the azepinoindole system (Chart 4). Practically speaking, the precursor was a 2-iodinated indole either Chart 2 Asymmetric total synthesis of (À)-catharanthineprepared directly by iodination of the corresponding indole or in two steps via the triethylsilyl intermediate 135. As an alternative, the same authors synthesized a 2-iodo tryptophyl iodide, which they condensed with the appropriate isoquinuclidine. Almost simultaneously, and as part of a program aiming at alleviating the metal stoichiometry problem, Sames at Columbia University also made use of a Heck cyclization, this time with the 2-bromo intermediate 136 (Chart 4). However, the most innovative part in this article is the search for a true catalytic reaction to perform the desired cyclization. It indeed worked with a Ni catalyst, but on a benzofuran analogue 137 (Chart 4). So far, the genuine nucleus of the iboga skeleton did not surrender to a fully catalytic CH insertion process. The question of the crucial C-2 to C-16 ring closure is more obviously solved when C-16 is at the ketone oxidation level. It is illustrated in the synthesis of (19R)-ibogaminol (138) by H€ ock and Borschberg, where a single acid treatment suffices to close the seven-membered ring (step g in Chart 5). The key step of Asymmetric construction of the isoquinuclidine system was devised by Yang and Carter using organocatalysis based on the proline derivative 140 (Chart 6). The enantioselectivities obtained are excellent, but the presence of an aryl group and the location of the ketone on the isoquinuclidine will require several steps before reaching the natural products. There are not many articles associating medicinal chemistry and the iboga alkaloids. One of them by Sun et al. uses some of Kuehne's chemistry to build azepino-indoles like 141 adorned with functional groups, such as oxo-, thio-, and seleno-hydantoin units (Chart 6) Chart 4 Sinha's (top) and Sames's (middle) routes to ibogamine via Heck cyclization. Sames's fully catalytic route to the iboga alkaloid analog 137 (bottom). Unfortunately, no biological data were provided for the compounds synthesized. The synthesis of voacangalactone () by Harada et al. qualifies as a highyielding multiple-step total synthesis, for which the number of steps involved seems not to be an issue since every single step is high yielding (Chart 7). There are many innovations in this synthesis, and, for instance, the two-carbon atom chain of the tryptamine is installed in the last steps via an oxalyl chloride Friedel-Crafts reaction followed by a diborane reduction, a sequence used by Woodward to make methoxytryptamine in his approach to reserpine. To the best of our knowledge, this is the first time this reaction has been used to close the seven-membered ring of an azepino-indole, which is one of the difficulties in the synthesis of iboga alkaloids. Other salient features of this synthesis are the use of a Diels-Alder reaction to form the isoquinuclidine 142, with a chiral induction and a gold-mediated indole synthesis (143 ! 144). It took 14 steps for the synthesis of compound 142 and another 13 steps to make voacangalactone, for which 2 mg were added to the overall world resources of this compound! Until now, 18-OMe-coronaridine (127) has not been obtained as a natural product, but from a pharmacological point of view it is a very promising candidate for pharmaceutical development (vide infra). Its synthesis is what may be considered as state of the art: short, convergent, enantioselective, and high yielding (see Chart 8 for an example of the synthesis of 127). This starts with an azepino-indole, 145, made in two steps from tryptamine, and which reacts with a wide variety of aldehydes to give intermediates of the secodine type, to then undergo Diels-Alder type cyclization. When a chiral auxiliary is placed on the nitrogen atom, chirality induction is obtained, with the efficiency of the process depending on the match between the chiral azepino-indole and the aldehyde. In the synthesis of 18-OMe-coronaridine (127), the aldehyde contains provisos for the methoxyethyl side chain, and, after the reaction, a masked aldehyde tetracycle (146) is obtained. The auxiliary group was then "exchanged" to a "handier" benzyl group and this was followed by reductive opening to a nine-membered ring to provide 147. Upon debenzylation and removal of the aldehyde protection, a cyclization occurred uneventfully to produce the cleavamine derivative 148. Simple heating of enamine 148 under reflux in toluene produced a "miracle", with an exquisite CH shift and ring closure to give the title compound in overall good yield and less than ten steps.) is certainly included in any top list of alkaloids for its mystique and aura. It has its roots in the mythic Bwiti cult of Gabon and it has been for the past three decades the object of a fierce battle between its pros and cons when used as a drug abuse treatment. A full volume in the series "The Alkaloids" was dedicated to ibogaine, and, in the Introduction, G.A. Cordell wrote: "is ibogaine an alkaloid that can save the world from addiction? Probably not". That was in 2001, so why keep writing on this topic? Two of the reasons are that advances have been made in the understanding of the drug addiction phenomenon, and, also because, in the absence of an alternative treatment, 2 is still more or less openly in use. The story of ibogaine (2) has been written several times including its most recent developments as a treatment for drug abuse. It is less well-known, however, that iboga root extracts were at one time available on the market as an antifatigue and stimulant agent. Formulated tablets were sold by the Laboratoires Houde ´under the name of Lambarene®, after the city in Gabon where Albert Schweitzer had his dispensary. The drug was withdrawn from the market in 1970 but 2 reappeared on the scene when Howard Lotsof was granted a patent in 1985 for "a rapid method for interrupting the narcotic addiction syndrome". This was followed by a period of conflicts and experimental work that culminated in the first ibogaine conference held in New York in 1999. The story is still ongoing since a further ibogaine conference was convened in Mexico in 2016 [76].
SECURITY PROBLEMS AND FATALITIES
There is at least one major reason to explain the reluctance of health authorities to pursue ibogaine (2) as a potential therapeutic agent: security of use. Toxicity has always been a concern and it cannot be disclaimed that several "patients" have passed away some time after ingesting 2. One can find in the forensic literature articles investigating the reasons for these fatalities. However, there does not appear to be a general pattern linking such deaths to the intake of 2. Neurotoxicity and cardiotoxicity are the most often suspected causes of death, and also activity on Purkinje cells has been observed in female rats under ibogaine treatment. Not all of these effects have been ascertained during patient autopsies, and the known tremorigenic effects of 2 seem to be unrelated to such fatalities.
ANALYTICS
During forensic investigations, the levels of ibogaine (2) and related products in body fluids were determined. In the human body, the main metabolite of ibogaine (2) is noribogaine (12-hydroxy-ibogamine, 149) produced by demethylation of 2 with cytochrome CYP2D6. It is also a constituent of the crude alkaloid mixture of the root bark of Tabernanthe iboga. The major alkaloids from the plant (also involved in poisoning cases) are ibogaine, noribogaine, voacangine (150), iboluteine, ibochine (151), ibogaline (152), and ibogamine (153)(Fig.). In the same paper, the stability of the two main alkaloids ibogaine and noribogaine under daylight exposure was investigated and it was found that they rapidly transformed into iboluteine (58) and ibochine for 2, and into their desmethoxy counterparts of 149, with respective half-lives of 81.5 and 11 min. Liquid chromatography-MS with electrospray ionization is the method of choice to rapidly discriminate these alkaloids and many examples of the use of this technique are found in recent literature, as e.g., in Refs.and.
MECHANISM OF ACTION
The mechanism of action of ibogaine () is very complex and a glimpse at this complexity is given in an article where the affinity of psychedelic drugs towards a large series of receptors and transporters is documented. To make a long story short, ibogaine activates μ and κ opioid receptors as well as σ1 and σ2 receptors. Unsurprisingly, because of its structural similarities with serotonin, it also inhibits to a certain extent serotonin and dopamine receptors. The action of 2 on the μ opioid receptor (MOR) was investigated in detail by Alper et al. using hypothalamus cells overexpressing this receptor. It was eventually found that 2 is a weak MOR antagonist without any agonist effect in rat thalamic membranes. The inhibition of acetylcholinesterase (AChE) by 2 was invoked at one time to explain some properties of the molecule and in particular the potentiation of morphine analgesia. This was reinvestigated and an IC 50 of 520 AE 40 μM was found, a very high value suggesting that the physiological effect of 2 on AChE is negligible. These findings led to the conclusion that the mechanism of action of ibogaine is different from that of other molecules with effects on opioid tolerance and withdrawal, indicating a novel mechanism with targets still to be discovered. In a series of intriguing papers, Pas ˇkulin et al. raised the question of the duration of the effects of 2, which is much longer than its pharmacokinetic parameters would allow. In particular, the mood-elevating effects appear generally a day after ingestion, a lapse of time where most of the drug and its metabolites are almost no longer observable. Two proteomic studies were performed on whole rat brain and on the yeast Saccharomyces cerevisiae subjected or not to 2, and the overexpressed proteins were detected by 2D gel electrophoresis. These were involved with cell metabolism: glyceraldehyde-3-phosphate dehydrogenase, phosphoglycerate kinase, enolase, and alcohol dehydrogenase, all related to compensation of the ATP pool decrease. Ibogaine (2) induces a new metabolic equilibrium aiming at saving energyand at the same time increases the activity of antioxidative enzymes as an adaptation to oxidative stress. The complex pharmacological profile of ibogaine was highlighted recently in a zebrafish study which, inter alia, allowed observation of its effects on social behavior (suggesting a wider use of this model for hallucinogenic and drug abuse research). It must also be realized that due to the rapid transformation of 2 into Fig.Main alkaloids from the roots of Tabernanthe iboga (149-153) The Iboga Alkaloids 149, which possesses a different pharmacological profile, it is not easy to understand the origin of the effects (beneficial or deleterious) of the iboga alkaloids on human beings. This is probably the second major reason why it seems so difficult to bring one of these molecules to the market.
-METHOXY-CORONARIDINE (18-MC)
Faced with the difficulties associated with ibogaine (2), a search for an alternative molecule in the series was undertaken. Ibogamine (143) and coronaridine (3) were eventually considered as potential candidates since they retained the action on drug self-administration without any tremorigenic side effects. These properties were shared by many other similar compounds, among which 18-OMe-coronaridine (127) proved to be the most interesting. Compound 127 was fully synthetic and therefore patentable, which has been an advantage over ibogaine and other natural products, and it has its own particular mechanism of action. It was thus found to have micromolar activity on κ, μ, and σ2 opioid receptors and on 5-HT 3 receptor without affinity for the NMDA receptor and serotonin transporter. Particularly interesting was an affinity for the α3β4 nicotinic receptor, which was the object of pharmacological experimentation by intracerebral administration in rats in a studies related to drug self-administration. Compound 127 was also shown to act on the sucrose reward circuit and therefore may be considered as a possible antiobesity agent. Also exciting is the observation that coronaridine (3) and particularly 18-MC (127) show promise against Leishmania amazonensis, a causative agent of leishmaniasis. A small pharmaceutical company named Savant HWP (San Carlos, CA, USA) has commenced preclinical testing comprising IND-enabling studies and GMP scale-up work on 18-MC in the hope of gaining official approval as a drug cessation treatment. The company is also conducting human clinical safety studies on compound 127 in Brazil, so that it can be investigated further for the potential therapy of leishmaniasis.
CYTOTOXICITY AND ANTIPROLIFERATIVE ACTIVITY
As often is the case currently, the isolation of new natural products may be accompanied by investigation of their biological properties, using in-house available assays. In this respect, cytotoxicity screens against one or several cancer cell lines is a common asset of many laboratories. Due to compound quantity limitations, or the inavailability of follow-up in vivo assays, it may not always be possible to perform studies to determine antineoplastic activity using experimental tumor-bearing mouse models. In a general screening procedure for inhibitors of the Wnt pathway, T divaricata gave a positive response that led subsequently lead to four bioactive iboga alkaloids: voacangine, isovoacangine, coronaridine (3), and coronaridine hydroxyindolenine, with IC 50 values in the 10 μM range. The best compound was 3, shown to down-regulate mRNA and therefore decrease the βcatenin protein level and inhibit the Wnt signaling pathway, which controls, among other factors, cell proliferation. Compound 3 has been known for a long time to be cytotoxic [100] and this may be a possible mechanism of action. 19-Oxocoronaridine, also named ervatamine I and conodusine E, was found to be ca. 40 times less potent. A good level of activity was observed for tabercarpamine A (125) against the MCF-7, HepG2, and SMMC-7721 cancer cell lines. Tabercarpamine A (125) was found to induce apoptosis, a biological activity shared with ervatensines A (83) and B (86). Vobatensines A-Fwere assayed against a variety of cancer cell lines and showed cytotoxic effects in the 10 micromolar range. No indication was given on their possible mechanism of action. Tabercorine A (120) and 17-acetyltabernaecorymbosine A (94) also displayed cytotoxic activities in the low micromolar range and were shown to be as potent in this regard as cisplatin, which was used as a positive control. However, it is unlikely that they proceed through the same mechanism of action.
CENTRAL NERVOUS SYSTEM EFFECTS
Given the very potent and profound effects of ibogaine (2) on the central nervous system, it is not surprising to find reports of further new investigations of this type. One of the most promising arose from the total synthesis work of Sinha et al. and concerns a benzofuran analog of ibogaine, which possesses dual affinity for both the μ and κ opioid receptors (MOR and KOR). This compound was also nontremorigenic, and showed antinociception in mice in a standard hot-plate test of comparable potency to morphine. The identification of a new pharmacophore in this study may lead to the development of a new treatment for pain. The roots of Tabernaemontana divaricata yielded two known alkaloids with a promising level of inhibition of acetylcholinesterase (a demonstrated target for Alzheimer's disease). These compounds were 19,20-dihydrotabernamine [103] and 19,20dihydroervahanine A [104]. Voacamine, 3,6-oxidovoacangine, and 5-hydroxy-3,6oxidovoacangine from Voacanga africana were found to be potent antagonists of the cannabinoid receptor CB1, with IC 50 values in the submicromolar range. This target is involved in memory, pain, and appetite and inhibitory compounds may show promise in the treatment of obesity, metabolic syndrome, and related disorders inter alia. Coronaridine (3), alone among a series of 13 iboga alkaloids, was found to have protective effects in MPP+ (1-methyl-4phenylpyridinium)-injured primary cortical neurons. This may be of relevance in a treatment of Parkinson's disease but one should not underestimate the cytotoxicity of the molecule (ca. 2 μM ).
MISCELLANEOUS BIOLOGICAL ACTIVITIES
The so-called ervatamines A-I, among which iboga alkaloids, as well as coronaridine (3), heyneanine, their 2 0 -oxo derivatives and pandine, were subjected to an anti-inflammatory in vitro test, and inhibition of lipopolysaccharide-induced NO production in RAW 264.7 macrophages was demonstrated for 3 and 64. The tabernaeelegantinals (117-119) from Muntafara sessilifolia were assayed against the chloroquine-resistant strain FcB1 of Plasmodium falciparum and showed moderate to good activity. Selectivity indexes were determined after assessment of their cytotoxicity and it was concluded that their activity was the result of a general toxicity. However, when evaluated in vitro using the same strain of P. falciparum, 3 0 -oxo-tabernaelegantine A (109), also isolated from M. sessiloflora, was more promising, and showed antiplasmodial activity (IC 50 4.4 μM ) with non-significant cytotoxicity for the L-6 rat skeletal muscle cell line.
CONCLUSION AND PERSPECTIVES
Iboga-type alkaloids are one of the many types of indolomonoterpene alkaloids, but they are set apart owing to their unique biological properties. Notably, one of these is their possible use for drug cessation treatment. It must be kept in mind that cocaine addiction is so far without any reliable treatment and that the substitutes to opioids are not without any drawbacks, with one of them being their illicit diversion. Could Howard Lotsof's dream come true and a true drug cessation treatment be introduced to the market in the future, hopefully inspired by Tabernanthe iboga?
Full Text PDF
Study Details
- Study Typemeta
- Populationhumans
- Characteristicsliterature review
- Journal
- Compound