Keywords

5.1 Introduction

Discorhabdins are structurally divergent marine alkaloids [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15] primarily found in sponges, Latrunculia. These compounds have demonstrated significant biological activities such as antiviral [16], antitumor [16, 17], antimalarial [18], and antimicrobial [18] activities. Discorhabdins exhibit high structural diversity, comprising approximately 50 congeners of different ring systems fused on a common pentacyclic skeleton containing a spirocyclic hexadienone fused with the pyrroloiminoquinone skeleton. The ring systems of structurally divergent discorhabdin can be classified into four classes based on structural complexity (Fig. 5.1). Class 1 congeners, such as discorhabdin C and E, have a common pentacyclic skeleton. Class 2 compounds, such as discorhabdin V and Z, have C2–N18 bonds in their E/F-rings. Class 3 congeners, such as discorhabdin A and B, have strained D/G rings containing an N, S-acetal moiety, which constitute the majority of congeners (34 compounds). Class 4 compounds, such as discorhabdin H and N, have a D/E/F/G ring system and are the most complicated congeners. In addition, a unique congener called aleutianamine, which was proposed to be biosynthetically derived from a Class 3 congener through skeletal rearrangement, was isolated from a deep-sea sponge in 2019. This compound has shown selective and potent activity against solid tumors, such as pancreatic cancer cell lines [19].

Fig. 5.1
20 bond line structures of the different classes of alkaloids. The compounds are discorhabdin C, E, U, V, Z, A, B, R, N, and H, epinardin D, 14 bromo 1 hydroxy discorhabdin V, epinardin A, 1 thiomethyl discorhabdin G slash I, 1 methoxy discorhabdin D, and aleutianamine.

Structurally divergent discorhabdin alkaloids and classification based on skeleton

Full size image

Discorhabdins have attracted considerable attention from synthetic chemists as attractive synthetic targets and drug candidates because of their promising biological activities and structurally intriguing complex molecular architectures [20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44]. Early-stage synthetic studies and total synthesis have mainly focused on simple Class 1 congeners such as discorhabdin C by Yamamura [21], Kita [22], Heathcock [23] and discorhabdin E by Heathcock [23]. Regarding Class 2 congeners, Heathcock’s and our groups have reported model studies on congeners categorized as Class 2 [23, 41]. More recently, the first total synthesis of the Class 2 congener (+)-discorhabdin V was reported by Burns and co-workers [42]. Regarding Class 3 congeners, Kita and co-workers disclosed the landmark total synthesis of (+)-discorhabdin A in 2003 [24, 25]. After two decades, our group reported the first total synthesis of (+)-discorhabdin B, as extensively discussed in this review article [43]. Based on the seminal synthetic studies on Class 4 congeners using natural discorhabdin B by Copp and co-workers, we completed the total synthesis of Class 4 congeners, (–)-discorhabdin H, in 2023 [43]. Since the isolation and structural determination of (–)-aleutianamine, this compound has attracted considerable attention as a synthetic target because of its highly fused heptacyclic structure and significant biological activities. In 2023, our group found a biomimetic synthetic route from a derivative of (+)-discorhabdin B to (–)-aleutianamine [43]. Shortly after our synthesis, Stoltz’s group reported the second example of its total synthesis [44]. As part of our ongoing project on pyrroloiminoquinoline marine natural products [45, 46], we conducted research to establish synthetic routes for discorhabdin congeners with various framework classes. As a result, we developed a divergent route to (+)-discorhabdin B, (–)-discorhabdin H, (+)-discorhabdin K, and (–)-aleutianamine via discorhabdin B, which are discussed in this review article, with the relevant background and related studies.

5.2 Biosynthetic Proposal and Previous Total Synthesis of N, S-Acetal-Containing Discorhabdins

The biosynthetic pathway of N, S-acetal-containing discorhabdins remains unclear. Scheme 5.1 shows Munro’s proposal for the biosynthesis of discorhabdin B [47]. According to this proposal, makaluvamine D was first synthesized from tyramine and tryptamine. Then, two routes were proposed for the conversion of makaluvamine D to discorhabdin B. The first route involves the formation of discorhabdin C from makaluvamine D via oxidative spirocyclization, followed by the introduction of a sulfur atom to convert discorhabdin C to B. The second route involves an early-stage introduction of sulfur atom to makaluvamine D, resulting in the formation of makaluvamine F, followed by oxidative spirocyclization to form discorhabdin B.

Scheme 5.1
A biosynthesis process. The tyramine reacts with tryptamine and undergoes oxidation and coupling to yield makaluvamine D, which further undergoes oxidative spiro cyclization to form discorhabdin D, followed by discorhabdin B. Makaluvamine D is reduced to makaluvamine F, resulting in discorhabdin B.

Proposed biosynthetic pathways for discorhabdin B

Full size image

Based on the two biosynthetic pathways proposed by Munro, Kita and co-workers conducted synthetic studies on N, S-acetal-containing discorhabdins, discorhabdins A and B [24, 25]. Scheme 5.2 shows the synthetic route to discorhabdin B from makaluvamine F by Kita and co-workers [48, 49]. The synthesis started with the preparation of the 2-aminodihydrobenzothiophene derivative 3. The direct introduction of nitrogen functionality at the C2 position of dihydrobenzothiophene derivative 1 was achieved through a Pummerer-type C2 azidation using a combination of iodosobenzene and TMSN3. After reducing the azide to an amine, the total synthesis of makaluvamine F was accomplished by a condensation of amine 3 with pyrroloiminoquinone 4 through an addition–elimination reaction [48, 49]. Finally, Kita and co-workers attempted the transformation of makaluvamine F into discorhabdin B under several oxidative conditions, such as PIFA or CuCl2 and H2O, as previously reported by Heathcook. However, the desired reactions did not provide discorhabdin B, which could be due to the high sensitivity of the N’S-acetal moiety toward oxidants.

Scheme 5.2
A synthesis route for discorhabdin B. Compound 1 forms compound 2 through an 8-step process, which further reacts in the presence of T M S N 3 P h I O and M e C N to form compound 3. Compound 3 reacts with M e O H and compound 4 to form makaluvamine F, which further reacts to form discorhabdin B.

Total synthesis of makaluvamine F by Kita and co-workers

Full size image

Scheme 5.3 depicts the total synthesis of discorhabdin A developed by Kita and co-workers, where the introduction of a sulfur atom was conducted at a later stage of the synthesis [24, 25]. The coupling reaction of amino alcohol 7, derived from L-tyrosine methyl ester (6), and pyrroloiminoquinone 5, followed by the PIFA-mediated diastereoselective oxidative cyclization of 8, provided spirodienones 9 in moderate yield and selectivity (49%, dr = 4.8:1). After cleavage of hydroxymethyl side chain in two steps, the introduction of sulfur on hemiaminal 10 was conducted by treatment with p-methoxybenzyl mercaptan under acidic conditions, which promoted N, O- to N, S-acetal exchange reaction to provide 11. Subsequently, the PMB group was removed using MeNH2 to generate free thiol 12. However, intramolecular thia-Michel reaction of 12 provided desired compound 13 in low yield; instead, a debrominated structural isomer 14 was obtained as the major product because of the low stereoselectivity of the N, S-acetal forming step. Finally, the first total synthesis of (+)-discorhabdin A was completed by deprotection of Ts group from 13.

Scheme 5.3
A synthesis route for discorhabdin A. Compound 8 forms compound 5 through a 5-step process, which further reacts in the presence of compound 5 to form compound 8. Compound 8 reacts with P I F A M K 10 to form compounds 10, 11, 12, 13, and 14. Compound 14 reacts with N a O M e to form discorhabdin A.

First total synthesis of discorhabdin A by Kita and co-workers

Full size image

5.3 Synthetic Plan

Inspired by Munro’s biosynthetic proposal, which suggested makaluvamine F as a precursor of discorhabdin B and Kita’s pioneering work (Scheme 5.2), we synthesized makaluvamine F and N-Ts-makaluvamine. We then examined oxidative spirocyclization by subjecting these compounds to various oxidation conditions. However, all attempted conditions failed to provide discorhabdin B or N-Ts-discorhabdin B and instead resulted in the formation of complex mixtures. Because of these setbacks, we abandoned the biomimetic synthetic route and investigated a nonbiomimetic synthetic strategy. Copp and co-workers reported the semi-synthesis of discorhabdin B from natural discorhabdin W by reductive cleavage of disulfide bond with dithiothreitol, followed by the final N, S-acetalization of free thiol 15 [6] (Scheme 5.4) to yield discorhabdin B. They observed that free thiol 15 was unstable and spontaneously underwent N, S-acetal formation at the C8 position, leading to the formation of discorhabdin B. Based on these observations, we speculated that N, S-acetal formation could occur if free thiol 15 could be generated from a precursor such as secondary amine 16 through a chemoselective oxidation of the secondary amine to imine, followed by the removal of a thiol-protecting group. We planned to construct spirodienone structure 16 through the PIFA-mediated oxidative spirocyclization of pyrroloiminoquinone 17, bearing a properly protected tyramine segment. Compound 17 can be easily prepared by condensation of pyrroloiminoquinone 5 with phenethylamine 18 using an addition/elimination reaction. Phenethylamine 18, containing a sulfur functionality, can be accessed through the reductive ring opening of 2-aminodihydrobenzothiophene 20, as described in Scheme 5.2, which can be readily prepared from the corresponding dihydrobenzothiophene.

Scheme 5.4
A mechanism of discorhabdin B. Discorhabdin W reacts with dithiothreitol to yield compound 15, followed by discorhabdin B. The compound 15 is produced from the compounds 20, 19, 18, 17, and 16. Makaluvamine F undergoes spiro cyclization to form discorhabdin B.

Synthetic strategy for discorhabdin B by Tokuyama and co-workers

Full size image

5.4 The First Racemic Total Synthesis of Discorhabdin B

Our synthesis of discorhabdin B started with the preparation of 2-aminobenzothiophene 25 using an alternative method of the protocol established in Kita’s total synthesis of makaluvamine F (Scheme 5.2) [49]. To achieve this, 2-methoxycarbonyl benzothiophene 23 [50] was synthesized from 2-fluorobenzaldehyde 21 and methyl 2-mercaptoacetate 22 via a SNAr reaction and intramolecular aldol condensation. Subsequently, it was reduced to dihydrobenzothiophene 24 using Mg metal under acidic conditions (Scheme 5.5) [51, 52]. After hydrolysis of the methyl ester in 24, the generated carboxylic acid was subjected to the Curtius rearrangement using DPPA and t-butanol to provide Boc-protected 2-aminodihydrobenzothiophene 25 with good yield and scalability (10 g) [53, 54]. Then, the chemoselective cleavage of the C1–S4 bond under the Birch reduction conditions yielded free thiol 26, [55,56,57] which was then converted to hydrochloride 27 via protection of the free thiol and removal of the Boc group in HCl in dioxane. Hydrochloride 27 and 5 were condensed to produce pyrroloiminoquinone 28 in good yield. Finally, the PIFA-mediated oxidative spirocyclization of 28 resulted in the formation of spirodienone 29 in excellent yield [24, 25].

Scheme 5.5
A mechanism of pyrroloiminoquinone 29. Compound 21 reacts with compound 22 and D M F to form compound 23, which further reacts with M g, A c O H, and M e O H to yield compound 24, followed by compounds 25, 26, 27, and 28. Compound 28 reacts with P I F A and T F E to yield compound 29.

Synthesis of pyrroloiminoquinone 29

Full size image

With spirodienone 29 in hand, we studied its oxidation to enamine 30 under several oxidation conditions using MnO2 and PdCl2 (Table 5.1). However, these conditions only yielded a trace amount of enamine 30 with recovery of a considerable amount of 29 (entries 1 and 2). Surprisingly, treatment of 29 with excess CuBr2 in MeCN at 80 °C did not yield the expected enamine 30; instead, it led to the formation of compound 31, which possessed an N, S-acetal structure and two bromo groups at the C4 and C2 positions (entry 3). This result indicated that CuBr2 promoted the chemoselective oxidation of secondary amine (29 to I), deacylation (I to II) [58, 59], formation of N, S-acetal moiety (II to 33), and sequential dibromination at the C4 and C2 positions (33 to 31) (Scheme 5.6). To mitigate excess bromination at the C4 position, we examined conditions with decreased amounts of CuBr2 (entry 4). However, only the C4 brominated product 32 was obtained as the major product, indicating that bromination at the C4 position proceeded faster than that at the C2 position. However, when THF was used as the solvent instead of MeCN, the undesired C4 bromination was suppressed, leading to the formation of product 33 (entry 5). Notably, a comparable yield of 33 was obtained using a catalytic amount of CuBr2 (30 mol%) in air (entry 6).

A catalytic mechanism. The compound 2 reacts in the presence of solvent to form compound 30, leading to the formation of compounds 31, 32, and 33. The compounds are composed of 4 cyclohexane rings and cyclopentane ring fused together to form a 19-carbon ring system.
Table 5.1 Oxidative N’S-acetal formation
Full size table
Scheme 5.6
A cascade mechanism of C u B r 2 promoted reaction. Compound 29 reacts with C u B r 2 to form compound 1, which further undergoes deacylation to yield compound 2. Compound 2 further undergoes N S acetal formation to yield compound 33, which is further processed with bromination to form compound 31.

Proposed reaction mechanism for CuBr2-mediated cascade reaction

Full size image

Although we successfully developed a CuBr2-catalyzed oxidative N, S-acetal formation cascade involving the oxidation of secondary amine to enamine, deacylation, and N, S-acetal formation, we failed to achieve regioselective bromination at the C2 position of compound 33. Instead, only decomposition was observed, possibly due to the sensitivity of the N, S-acetal structure to brominating agents (Scheme 5.7). Therefore, we abandoned the synthetic route from 33 to discorhabdin B and examined bromination before the oxidative N, S-acetalization cascade.

Scheme 5.7
A bromination reaction of compound 33. Compound 33 reacted under bromination conditions to yield compound 34, which further reacts under deprotection to from discorhabdin B.

Unsuccessful bromination of 33

Full size image

Table 5.2 summarizes the reaction of spirorenone 29 under various bromination conditions. Reaction using a combination of AIBN and NBS, DBDMH (1,3-Dibromo-5,5-dimethylhydantoin), NBS, and TBCO resulted in the decomposition of substrate 29 (entry 1–4). However, the use of PyHBr3 in a mixed solvent system (CHCl3/MeCN) effectively facilitated regioselective bromination at the C2 position, yielding the desired compound 35 in good yield (69%).

A reaction and 2 structures. Compound 29 forms compound 5. D B D M H has a 5-carbon ring. C1 and C3 are double bonded to O. C2 and C4 are replaced by N and bonded to B r. C5 bonded to 2 M e. T B C O structure has a 6-carbon ring. C1 is double bonded to O. C2 and C6 are bonded to Br. C4 is bonded to 2 Br.
Table 5.2 Exploration of the C2-selective bromination condition for 29
Full size table

After successfully establishing the C2-selective bromination reaction, we examined the CuBr2-catalyzed oxidative N, S-acetal formation cascade using brominated compound 35 (Scheme 5.8). However, the reaction of 35 under the established conditions resulted in a complex mixture. Alternatively, when THF was used as the solvent instead of CHCl3, the desired N, S-acetal 34 was obtained with a yield of 51%. Finally, the first racemic total synthesis of discorhabdin B was completed by removal of the Ts group [24, 25].

Scheme 5.8
A synthesis of discorhabdin B. Compound 35 reacts in the presence of C u B r 2 and C H C l 3 to yield compound 34, which further reacts with N a O M e and T H F slash M e O H to form discorhabdin B.

Endgame of the first total synthesis of (±)-discorhabdin B

Full size image

5.5 The First Asymmetric Total Synthesis of (+)-Discorhabdin B

Achieving the asymmetric total synthesis of (+)-discorhabdin B required the stereoselective construction of the C6 spirocenter. To accomplish this, we devised two strategies, as depicted in Scheme 5.9. The first strategy involved the reagent-controlled asymmetric oxidative spirocyclization of pyrroloiminoquinone 36 using a combination of chiral iodine catalyst and co-oxidants based on the protocol reported by Ishihara and co-workers [60, 61]. The second method relied on substrate-controlled PIFA-promoted diastereoselective oxidative spirocyclization of pyrroloiminoquinone 38 bearing chiral thioesters at the C5 position.

Scheme 5.9
2 reactions to yield C6 spirocenter. Top. Compound 36 reacts in the presence of compound 37 under reagent controlled enantioselective spiro cyclization to yield compound 29. Right. Compound 38 reacts with P I F A under substrate controlled diastereoselective spiro cyclization to form compound 39.

Two strategies for stereoselective construction of the C6 spirocenter

Full size image

Table 5.3 summarizes the results of the oxidative spirocyclization of pyrroloiminoquinone 36 using various chiral aryliodides. The reactions were conducted according to the protocol developed by Ishihara and co-workers using catalytic amount of chiral aryliodines 37a–c and mCPBA as the cooxidant [60]. Reactions using chiral aryliodides 37a–c resulted in the formation of compound 29 with low-to-moderate yields (12–43%) and poor enantioselectivities (entries 1–3). The reaction of phenol 36b was tested to facilitate oxidative spirocyclization. However, the desired compound was not obtained (entry 4).

A chemical reaction. Compounds 36 a and b react in the presence of catalyst m C P B A and T F E to yield compound 29. 3 bond line structures of compounds 37 a, 37 b, and 37 c are given below the reaction. The structures are composed of a benzene ring, O, N O H, and O E t groups.
Table 5.3 Attempts at reagent-controlled oxidative spirocyclization
Full size table

Having found that the reagent-controlled asymmetric oxidative spirocyclization approach was ineffective, we investigated substrate-controlled diastereoselective spirocyclization using pyrroloiminoquinone 38 bearing chiral thioesters (Table 5.4.). Although the chemical yields were relatively improved compared with the reagent-controlled enantioselective oxidative spirocyclization approach (Table 5.3) [24, 25], the diastereoselectivity remained unsatisfactory. Among the optically active thioesters prepared from N-protected amino acids (38a–38f), the tert-leucine substrate provided the corresponding spirocyclic compound with the highest chemical yield (92%) and diastereoselectivity (21% de). Subsequently, we examined a series of chiral thioesters 38 g–n derived from mandelic acid and its derivatives. The chemical yields and diastereoselectivity varied depending on the protective group on the hydroxyl group and the substituent on the benzene ring. Among the series of chiral thioesters derived from mandelic acid derivatives, we selected the TBS-protected chloromandelic acid derivative 38 h as the optimal substrate for oxidative spirocyclization. It provided 39 h almost quantitative yield with moderate diastereoselectivity (97%, 31% de). The MTPA (38o) derivative and the methoxynaphthalenyl propanoic acid derivative (38p) were not effective in the reaction. The absolute configuration of the spirocenter in product 39 h was not determined at this stage but was later determined via a few additional step transformations to discorhabdin B (Scheme 5.10).

Table 5.4 Results of diastereoselective oxidative spirocyclization using various chiral thioesters
Full size table
Scheme 5.10
A reaction of discorhabdin B. Compound 39 reacts in the presence of P y H B r, C H C l 3 slash M e C N at 35 degrees Celsius to yield compounds 40 a and 40 b, which further react with C u B r 2 and C H C l 3 to yield compound 34 that further reacts in the presence of N a O M e to form discorhabdin B.

First asymmetric total synthesis of (+)-discorhabdin B

Full size image

Although the diastereoselectivity of the oxidative spirocyclization in Table 5.4 could be improved, we transformed spirodienone 39 h to (+)-discorhabdin B (Scheme 5.10) according to the protocol established in our racemic total synthesis (Scheme 5.8). Thus, treatment of a diastereomeric mixture of spirodienones 39 ha and 39hb (65.5:34.5) with PyHBr3 yielded a mixture of 40a and 40b (67.5: 32:5) in moderate yield. After separation, the major diastereomer 40a was subjected to CuBr2-catalyzed oxidative spirocyclization, followed by deprotection of the Ts group using NaOMe, thereby completing the asymmetric total synthesis of (+)-discorhabdin B for the first time [24, 25].

5.6 The First Asymmetric Total Synthesis of Discorhabdins H and K

The construction of F-ring by the formation of C2–N18 bond presents a synthetic challenge for Class 4 discorhabdin congeners. Before describing our first total synthesis of discorhabdin H, synthetic studies on Class 4 discorhabdin congeners focusing on F-ring construction are briefly discussed (Scheme 5.11). Heathcock and co-workers constructed the F-ring in 43 via partial reduction of spirodienone 41 to provide spiroenone 42, followed by bromination at the C2 position and intramolecular N-alkylation (Scheme 5.10a) [23]. Our group conducted model studies to demonstrate the F-ring construction through a Pd-catalyzed intramolecular Heck cyclization of halogenated pyrroloiminoquinone 44 (Scheme 5.10b) [41]. Based on our Heck-cyclization strategy, Burns and co-workers accomplished the first asymmetric total synthesis of discorhabdin V (Scheme 5.10c) [42]. The biosynthetic pathway for the construction of the F-ring was supported by Copp’s seminal model experiment (Scheme 5.10d) [10]. Copp and co-workers treated natural (+)-discorhabdin B with N-Ac-L-cysteine in the presence of Et3N to facilitate a thia-Michael reaction at the C1 position, followed by formation of a C2–N18 bond from 50 to yield 51 in low yield. These transformations are widely accepted as a plausible biosynthetic pathway for Class 4 discorhabdin congeners.

Scheme 5.11
4 mechanisms. a. Compound 41 reacts with P h M e 3 N B r 3 to form compound 42, followed by compound 43. b. Compound 44 reacts to form compound 45, followed by compound 46. c. Compound 47 reacts to form compound 48, followed by compound 49. d. Discorhabdin B reacts to form compounds 50 and 51.

Synthetic approaches for the construction of the F-ring in class 4 congeners

Full size image

Inspired by the Copp’s model studies, we prepared L-ovothiol A according to reported procedures [61,62,63,64,65] and examined the thia-Michael reaction with N-Ts discorhabdin B (34). Initially, we attempted the thia-Michael reaction using L-ovothiol A under Copp’s conditions with triethylamine, but the desired reaction did not proceed (Scheme 5.12). After extensive studies using various combinations of bases and solvent systems, we eventually found that the thia-Michael reaction proceeded by simply mixing 34 with L-ovothiol A in DMSO/H2O (3:1) to furnish the desired N-Ts-discorhabdin H (52) along with N-Ts-discorhabdin K (53) in 65% combined yield. The generation of the major side product disulfide of L-ovothiol A was suppressed by Ar bubbling during the reaction. The first total syntheses of discorhabdin H and K were accomplished [7, 8] using high-performance liquid chromatography (HPLC) separation of 52 and 53, followed by deprotection of the Ts group [24, 25].

Scheme 5.12
A 3-step mechanism. Compound 34 reacts with L ovothiol A to yield compounds 52 and 53. Compound 52 further reacts with N a O M e and T H F slash M e O H to yield discorhabdin H. Compound 53 further reacts with N a O M e and T H F slash M e O H to form discorhabdin K.

First asymmetric total synthesis of (–)-discorhabdin H and (+)-discorhabdin K

Full size image

Scheme 5.13 outlines the proposed mechanism for the generation of 52 and 53 through the thia-Michael reaction of L-ovothiol A to 34, followed by the formation of the C2–N18 bond. The initial thia-Michel addition proceeded in a highly stereoselective manner from the less hindered side, avoiding the pyrroloiminoquinone skeleton to generate enol 54. The protonation of enol 54 is expected to proceed without steric repulsion of the ovothiol segment to furnish 55. Finally, intermediate 55 underwent an intramolecular SN2 reaction to form the C2–N18 bond, leading to the formation of N-Ts-discorhabdin H (52). Alternatively, the anti-elimination of HBr can proceed to produce N-Ts-discorhabdin K (53).

Scheme 5.13
A thia Michel mechanism. Compound 34 reacts to yield compounds 52 and 53. Compound 34 is formed from L-ovothiol, which further undergoes stereoselective 1 4 addition to yield compounds 54 and 55, which further undergo cyclization and antielimination to form compounds 53 and 53.

Proposed reaction mechanism of the thia-Michel addition of L-ovothiol A to 34 to furnish 52 and 53

Full size image

5.7 Total Synthesis of (–)-Aleutianamine

Aleutianamine is a new class of pyrroloiminoquinone alkaloids. This compound was isolated from a marine sponge, Latrunculia (Latrunculia) austini Samaai, Kelly & Gibbons, 2006 by Hamann and co-workers in 2019 [19]. Note that this compound exhibits potent and selective cytotoxicity against a pancreatic cancer cell line (PANC-1) at an IC50 of 25 nM (Scheme 5.14) [19]. In addition to its fascinating biological properties as a candidate for a new anticancer drug, aleutianamine possesses a highly complicated seven-membered ring system containing a tertiary sulfide moiety, which was elucidated through extensive spectroscopic analysis combined with computational studies [19]. Regarding the biosynthetic hypothesis of aleutianamine, Hamann and co-workers proposed two routes [19]. Route A involves the oxidation of the phenolic moiety makaluvamine F to quinone methide I, followed by the intramolecular Michael addition of the enamine moiety to quinone methide to form hexacyclic intermediate II. Then, C3–N18 bond formation gives intermediate III and finally reduction of N’O-acetal to provide aleutianamine. Route B involves the skeletal rearrangement of discorhabdin B to provide intermediate II, followed by the generation of N, O-acetal III and subsequent reduction to yield aleutianamine.

Scheme 5.14
2 biosynthetic routes. Route A. Makaluvamine F reacts to yield compound 1, which further reacts to form compound 2. Route B. Discorhabdin B reacts to yield the intermediate compound, which further undergoes skeletal rearrangement to form compound 2. Compound 2 reacts to form aleutianamine.

Proposed biosynthetic pathway for aleutianamine

Full size image

During synthetic studies on the transformation of N-Ts-discorhandin B (34) [43] to 3-dihydrodiscorhabdin [65], we found a similar skeletal rearrangement that involved in Route B, which allowed us to establish the first asymmetric total synthesis of aleutianamine. Thus, to establish the total synthesis of 3-dihydrodiscorhabdin via the reduction of the C3 ketone of 34, we examined various 1,2-reductions. Among them, we found that Luche reduction [66] promoted the chemoselective 1,2-reduction of the C3 ketone to form diallyl alcohol 56 (Scheme 5.15a). However, 56 was too unstable to isolate and was spontaneously converted into a structurally unidentified product during the purification process using reverse-phase HPLC (0.1% TFA in MeCN/H2O). Surprisingly, extensive NMR studies revealed that the structure of the unidentified product was N-Ts-aleutianamine (57). Finally, the Ts group of 56 was removed by treatment with NaOMe in THF/MeOH, completing the first total synthesis of (–)-aleutianamine [19, 43].

Scheme 5.15
3 chemical reactions of aleutianamine. a. Compound 34 reacts to form compound 36, followed by 3 dihydrodiscorhabdin, compound 57, and aleutianamine. b. Compound 56 reacts to form compound 2, which further forms compounds 2 and 4. c. Compound 58 reacts to form compounds 59, 60, and aleutianamine.

Total syntheses of aleutianamine by the Tokuyama and Stoltz groups

Full size image

A plausible reaction mechanism for the rearrangement of 56–57 is depicted in Scheme 5.15b [67]. Under the acidic conditions of HPLC purification, the C3 secondary alcohol of 56 was protonated to promote the dehydration reaction to generate diallyl cation species II, which then underwent skeletal rearrangement via an intercept of the allyl cation moiety with enamine to form fused cyclopropane III, followed by ring opening of the cyclopropane ring to generate a sulfur-bridged azepine derivative IV. Finally, a C3–N18 bond formed to furnish N-Ts-aleutianamine (57).

A few months after the disclosure of our asymmetric total synthesis of (–)-aleutianamine, Stoltz and co-workers also accomplished a racemic total synthesis of aleutianamine using a completely different synthetic approach involving an elegant intramolecular Heck-type reaction (Scheme 5.15c) [44]. They constructed densely fused indole scaffold 59 with a tertiary sulfide at the C5 position through a Pd-catalyzed dearomative intramolecular Heck-type reaction of bromopyrroloiminoquinone 58 with 2-aminotetrahydrobenzothiophene segment. The racemic total synthesis of aleutianamine was then achieved via the introduction of a Br group at the C2 position, oxidative pyrroloiminoquinone formation, and N, S-acetal formation.

5.8 Conclusion

We have accomplished total syntheses of a series of discorhabdin alkaloids including (+)-discorhabdin B, (–)-H, (+)-K, and (–)-aleutianamine based on the development of the substrate-controlled diastereoselective spirocyclization using chiral thioester and CuBr2-catalyzed late-stage oxidative N, S-acetalization as two key processes. Establishment of these syntheses would pave the way to divergent synthesis of all different classes of discorhabdins including hitherto not synthesized Class 2 and Class 4 congeners. Furthermore, successful syntheses should be helpful to understand biosynthesis of discorhabdin congeners including the unique skeletal rearranged congener, aleutianamine.