Synthetic Studies toward the Myrioneuron Alkaloids

Abstract

The Myrioneuron alkaloids are a relatively small family of plant-derived alkaloids that present an intriguing array of structural intricacy and biological properties. As such, these natural products have drawn interest from the synthetic community, resulting in creative total syntheses of several family members. This review showcases recent synthetic efforts towards these polycyclic alkaloids.

Graphical Abstract:

1. Introduction

Polycyclic alkaloids have a rich history both as vital therapeutics in human health and in the development of the art of modern organic synthesis. The Myrioneuron alkaloids are a relatively recent addition to this broader class of natural products, with its first members isolated by Bodo et al. in 2002 from Myrioneuron nutans, a small tree native to Vietnam.^1,2a In subsequent years, the class has grown to over 35 members isolated by the Bodo and Hao groups from various Myrioneuron species (Figure 1).^1,2 These plants form part of the large Rubiaceae family of flowering plants, comprised of hundreds of genera, from which many alkaloids like caffeine, psychotridine, yohimbine, emetine, and quinine with important agricultural and medicinal applications have been isolated.³ Although far less studied, the Myrioneuron alkaloids present a promising array of biological activity, which, in combination with their complex polycyclic structures, has attracted the attention of synthetic chemists.

Figure 1.

The Myrioneuron alkaloids.

Structurally, the Myrioneuron alkaloids contain a core decahydroquinoline framework (see inset, Figure 1) that is fused to additional oxazine, diazine, or cyclohexane rings. The resulting polycyclic structures typically present as interlocked chair-like conformers, as illustrated for myrioneurinol (3)^2c and myrobotinol (4).^2d Further diversity is achieved through oxidation state variation, connectivity, and the number of monomer units, resulting in tricyclic to decacyclic congeners. Despite this structural diversity, all members are believed to derive biosynthetically from lysine.^1,2c,d,e,k

It is worth noting that a few Myrioneuron alkaloids had been isolated from other plant species prior to Bodo’s initial description of this family. For example, schoberine (11) was previously isolated from several Nitraria species⁴ before its later discovery in M. nutans,^2e being grouped among the ‘tripiperidine alkaloids’ that include other structurally distinct alkaloids like matrine.⁵ We have chosen to mainly focus on a structural definition for the class according to the key substructure shown in Figure 1.

A prior review by Poupon and Gravel discusses Nitraria and Myrioneuron alkaloids, insights into their common biosynthetic origins, and early synthetic efforts through 2009.¹ However, since then, the Myrioneuron family has seen significant growth through the isolation of new natural products and several synthetic studies and successful total syntheses. This review, therefore, intends to showcase the current state of the Myrioneuron alkaloid family, with a particular focus on recent synthetic efforts toward this class of natural products.

1.1. Biological Activity

As part of the larger Rubiaceae family of plants, a rich source of bioactive alkaloids, it is unsurprising that alkaloids isolated from Myrioneuron species display some interesting biological properties. These activities include antimalarial and antibacterial properties, cytotoxicity against cancer cell lines, and inhibition of hepatitis C virus (HCV) replication.

Several Myrioneuron alkaloids display antimalarial effects. Myrioneurinol (3) has shown activity against Plasmodium falciparum (IC₅₀ = 11 μg/mL), which was suggested to not derive from cytotoxicity given its weak inhibition of KB cell proliferation (IC₅₀ = 26 μg/mL).^2c More potent antimalarial activity is found for myrionidine (24) (and its unnatural enantiomer) and schoberine (11) (IC₅₀ = 0.3 and 4 μg/mL, respectively, against P. falciparum), which again does not appear to be solely attributable to cytotoxicity (IC₅₀ = 6 and 20 μg/mL against KB cells, respectively).^2e

Significant cytotoxic effects are less common among the class, but myrifabine (14), a decacyclic heterodimeric Myrioneuron alkaloid, has shown cytotoxicity towards multiple human tumor cell lines (IC₅₀ = 16.4–21.2 μM), as well as activity against Gram-positive bacteria including Staphylococcus aureus and methicillin-resistant S. aureus (MIC = 6.3–25.1 μg/mL).^2h Additionally, myrionamide (8) exhibited weak cytotoxicity towards HeLa, A549, and COLO-205 cell lines (IC₅₀ = 35.2–38.7 μM).^2m

Perhaps the most prevalent property exhibited by this alkaloid family is antiviral activity against HCV. Schoberine B (5), containing a unique hydroxyalkyl chain only found within one other family member (20), exhibits inhibitory activity (EC₅₀ = 2.76 μM) towards HCV with a selectivity index (SI) higher than 36.2 (CC₅₀ >100 μM).^2k A number of other Myrioneuron alkaloids also display anti-HCV activity, including myriberines A (12) and B (13, EC₅₀ = 5.1–8.5 μg/mL, SI >12),^2g,k myrifabrals A (18) and B (19, EC₅₀ = 2.2–4.7 μM, SI >25),²ⁱ myritonines A (29) and B (30, EC₅₀ = 16–17 μM, SI >12),^2l 12-oxomyriberine (10) and isomyrionine (22, SI >15.2 and 36.2, respectively),²ⁿ and myrifamines A (15) and B (16, EC₅₀ = 0.92–3.3 μM, SI >30.6).^2j

This encouraging biochemical profile notwithstanding, it is likely that other interesting biochemical properties are inherent to the family, with the aforementioned studies limited in several cases with respect to natural material supply and assay availability. Further biological studies with synthetic material should enable the full therapeutic potential of this family to be explored.

1.2. Proposed Biosynthesis

The Myrioneuron alkaloids and the structurally related Nitraria alkaloids (both metabolites of Rubiaceae family) are proposed to derive biosynthetically from the common amino acid building block L-lysine (36). Their distinct structures are believed to be the result of divergent bond-forming processes from a common aldehyde intermediate (42, Scheme 1A). In this pathway, L-lysine first undergoes decarboxylation to form the intermediate cadaverine (37), which then is subjected to oxidative deamination to lead to 5-aminopentanal (38). Intermediate 38 can cyclize under dehydrating conditions to yield tetrahydropyridine (39) in either its enamine or imine form. Enamine/imine 39 can then dimerize through a Mannich-type reaction to form tetrahydroanabasine (40), which can undergo a retro-aza-Michael reaction and oxidative deamination to form the key intermediate 42.^1,5

Scheme 1.

(a) Biosynthesis of key intermediate 42; (b) Elaboration of 42 to the Myrioneuron alkaloids.

From imine 42, the characteristic decahydroquinoline framework of Myrioneuron alkaloids can be constructed via conjugate reduction and a subsequent Mannich cyclization (42 → 44, Scheme 1B). Simple tricyclic members such as myrioxazine A (1) and myrioxazine B (2) are formed via aldehyde reduction, with or without prior epimerization, followed by condensation of the resulting amino alcohol with a formaldehyde equivalent to form the natural products. The incorporation of other piperidine units (e.g., 44 → 47) yields family members bearing additional azine rings like myrionine (6) and schoberine (11). Similar pathways with additional lysine-derived units might explain the biogenesis of other more elaborate congeners, including (hetero)dimers and higher oligomers like myrobotinol (4), myrifabine (14), and myrifamine C (17).

The biogenesis of myrioneurinol (3), which differs from most other members due to the presence of a quaternary center at C-5, is also envisaged to proceed via the key intermediate 42 (Scheme 2).^2c This compound undergoes conjugate addition of enamine 39 to its unsaturated imine to form intermediate 48. An enamine-aldol closure, followed by deoxygenation, delivers the characteristic cyclohexane ring in the form of compound 49. Imine hydrolysis and similar oxidative deamination as seen for the simpler members (Scheme 1A) gives imine-dialdehyde 50, which can perform an intramolecular Mannich reaction to form 51. A final reduction of the dialdehyde units and oxazine ring formation provides myrioneurinol (3).

Scheme 2.

Proposed biosynthesis of myrioneurinol (3) from key intermediate 42.

Although the pathways outlined above can account for the majority of Myrioneuron structures, several non-canonical family members like schoberine B (5),^2k containing a pentanol chain (presumably lysine-derived) with unique connectivity, and myrifabrals A–D (18–21) necessitate the intervention of alternative biogeneses, some aspects of which require further clarification.^2i,o

2. Synthetic Studies towards the Myrioneuron Alkaloids

Although the Myrioneuron alkaloid family has grown to contain over 35 members, to date only 8 have been prepared by total synthesis. With the exception of myrioneurinol (3) and myrifabrals A and B (18–19), the majority of these completed targets have been simpler members containing bicyclic or tricyclic scaffolds. In fact, many of the more structurally complex Myrioneuron alkaloids, such as myrobotinol (4) or myrifabine (14), pose a formidable challenge for modern organic synthesis and have yet to be addressed synthetically. In the following section, we outline the synthetic work conducted toward this intriguing alkaloid family to date.

2.1. Total Synthesis of Myrioxazines A and B

The initial synthetic work toward this class of natural products was conducted by Bodo and coworkers in 2002, coinciding with their isolation of the first family members myrioxazines A (1) and B (2).^2a In this report, these authors were able to confirm their proposed structures through asymmetric synthesis of (+)-myrioxazine A and (+)-myrioxazine B, as well as their C-5,10 epimers (Scheme 3).

Scheme 3.

Bodo’s total synthesis of (+)-myrioxazine A (1) and (+)-myrioxazine B (2).

In the Bodo group’s synthesis, the tetrahydroquinoline core was constructed from readily accessible cyclohexanone (52) via Michael addition of its piperidine enamine to acrolein, followed by cyclization of the resulting 1,5-dicarbonyl compound with hydroxylamine hydrochloride to yield 5,6,7,8-tetrahydroquinoline (53; also commercial). Reacting 53 with paraformaldehyde provides the racemic primary alcohol (±)-54. The attachment of a (−)-camphanoyl group allowed for separation of the resulting ester diastereomers, which after hydrolysis led to enantiopure (−)-54 and (+)-54. Moving forward, PtO₂-catalyzed hydrogenation of the pyridine ring of (−)-54 gave 55 and 56 (dr = 1:4) with hydrogen delivered preferentially to the face opposite the hydroxymethyl unit; in a similar fashion, (+)-54 provided ent-56 as the major diastereomer. Amino alcohols 55 and ent-56 each underwent cyclization with aqueous formaldehyde to provide (+)-myrioxazine A (1) and (+)-myrioxazine B (2), respectively. Overall, the syntheses are quite concise, proceeding in 8 steps with overall yields of 1.5% and 6.0% for (+)-myrioxazine A (1) and (+)-myrioxazine B (2), respectively. However, an unselective hydrogenation step (especially for 1) and the classical resolution strategy led to lower overall efficiency. Nonetheless, these studies provided the first inroads – both in isolation and synthetic terms – into the Myrioneuron alkaloid family, and the developed strategies would be utilized by Bodo and coworkers in future syntheses within the class.

One further racemic synthesis of myrioxazine A (1) was reported by the Coldham group in 2009 (Scheme 4), taking advantage of a key nitrone formation/dipolar cycloaddition cascade to stereoselectively construct the substituted decahydroquinoline core.^6b Their synthesis starts with the alkylation of 6-heptenenitrile (57) with bromide 58 followed by an acidic workup to furnish primary alcohol 59. Chlorination of 59 followed by DIBAL-H reduction yielded aldehyde 60, the precursor for the key step of the synthesis.

Scheme 4.

Coldham’s total synthesis of (±)-myrioxazine A (1).

This planned cascade involves a tandem oxime formation/N-alkylation/intramolecular nitrone [3+2] cycloaddition reaction to forge the decahydroquinoline core. To effect this process, aldehyde 60 is treated with hydroxylamine hydrochloride and i-Pr₂NEt in refluxing toluene to initially form oxime 61, which is alkylated by the pendant primary alkyl chloride to form cyclic nitrone 62. 62 then undergoes a diastereoselective intramolecular [3+2] cycloaddition with the pendant olefin to provide tricycle 63 in 77% yield bearing the correct relative configuration for myrioxazine A (1). Zinc-mediated cleavage of the N─O bond in 63 gave an amino alcohol 55 that upon heating with formaldehyde and catalytic p-TsOH yielded (±)-myrioxazine A (1). Overall, the synthesis is short, similar to the Bodo approach, but far higher yielding, proceeding in 42% yield over the six steps. This is principally due to the elegant tandem formation of two of the three rings in a highly diastereoselective fashion. The preparation of amino alcohol 55 also constitutes a formal synthesis of myrionine (6), myrionidine (24) and schoberine (11) based on the work of Bodo (vide infra).^2e,7

Coldham and coworkers additionally reported another very similar route to (±)-myrioxazine A (1) in 2012 (Scheme 5A).^6c Here, they utilize α-sulfonyl aldehyde 64, which is prepared in a similar fashion to their previous route, in the key dipolar cycloaddition cascade. Using their prior conditions, 64 is converted to tricyclic product 66 in 93% yield (via intermediate nitrone 65). To complete the synthesis, sodium amalgam-mediated N–O cleavage and desulfonylation and subsequent treatment with paraformaldehyde and p-TsOH furnishes myrioxazine (1) in high yield.

Scheme 5.

(a) Coldham’s alternative route to (±)-myrioxazine A (1); (b) Model study to access myrioneurinol core 69.

A final study by the same group details their efforts to extend this approach toward the framework of myrioneurinol (3), which contains a C-5 quaternary center (Scheme 5B).^6a Starting from model aldehyde precursor 67, with an ethyl-containing quaternary center in place, the same intramolecular cycloaddition cascade furnishes the tricyclic amine 68 as a single diastereomer. Similar processing as before affords tricyclic oxazine 69. Structurally, 69 contains three of four rings of the natural product, and demonstrates that their developed tandem reaction has the potential to furnish the core framework of myrioneurinol (3) with its key quaternary stereocenter. No further studies toward 3 have been reported by the Coldham group to date.

2.2. Total Synthesis of Myrionine, Myrionidine, and Schoberine

In further investigations on Myrioneuron nutans, in 2007/8 the Bodo group isolated and performed the asymmetric synthesis of myrionine (6), myrionidine (24), and schoberine (11, Scheme 6).^2b,e Their synthetic strategy made use of the same intermediate 55, prepared in seven steps from cyclohexanone (Scheme 3), that they had utilized in their prior syntheses of myrioxazines A (1) and B (2). Amino alcohol 55 is N-benzyl protected to afford primary alcohol 70, which is converted to a mesylate and reacted with the anion of 2-piperidinone to afford lactam 71. Benzyl deprotection of 71 (H₂, Pd/C, AcOH) delivers (−)-myrionine (6). Alternatively, a dehydrative cyclization of 6 via POCl₃ and base treatment gives (−)-myrionidine (24). LiAlH₄ reduction of the amidinium of 24 yielded (−)-schoberine (11). Overall, (−)-myrionine (6), (−)-myrionidine (24), and (−)-schoberine (11) were synthesized in 4–6 steps from compound 55 (11–13 steps from cyclohexanone).

Scheme 6.

Bodo’s total syntheses of (−)-myrionine (6), (−)-myrionidine (24), and (−)-schoberine (11).

2.3. Total Synthesis of Myrifabrals A and B

Myrifabrals A (18) and B (19) are structurally atypical Myrioneuron alkaloids containing a bridged octahydroquinolizine skeleton, which were isolated in racemic form as pairs of acetal epimers by Hao et al. in 2014.²ⁱ Shortly thereafter, the She group reported in 2016 the first racemic synthesis of these compounds via an impressively short sequence (Scheme 7).⁸ Their synthesis begins with the alkylation of β-ketoester 72 with iodide 73, which can be prepared from acrolein in one step,⁹ providing the alkylated product 74. Engaging 74 in a key tandem Mannich/amidation reaction with imine 39 delivered 75 in 46% yield in an optimized one-step process. This reaction constructs the central bicyclic piperidine ring of the natural product, and notably this reaction can be performed on scales up to 5 g with only slightly decreased yield. Upon treatment with LiAlH₄, compound 75 was reduced to an amino alcohol (dr = 6.75:1), with the major diastereomer 76 isolated in 81% yield. Upon treatment with aqueous HCl, 76 underwent acetal hydrolysis and hemiacetalization to provide (±)-myrifabral A (18a/b) in 92% yield, completing an impressive 4-step total synthesis from known 73 (28% overall yield). A similar treatment of 76 with HCl, followed by subsequent addition of hemiaminal ether 77 yielded (±)-myrifabral B (19a/b, 42%) as well as (±)-myrifabral A (18a/b, 21%). Presumably, 19a/b – which Hao et al. suggest may be an isolation artifact²ⁱ – is formed via α-alkylation of the aldehyde tautomer of 18a/b with an iminium ion derived from 77.

Scheme 7.

The She group’s total synthesis of (±)-myrifabrals A (18) and B (19).

An additional enantioselective synthesis of myrifabrals A (18) and B (19) was recently reported by the Stoltz group (Scheme 8).¹⁰ Starting from β-ketoester 78, Mannich reaction with sulfonylmethyl carbamate 79 provided protected α-(aminomethyl)ketone 80. After optimization, the Boc-protected amine was directly converted to glutarimide 82 using glutaric acid and a catalytic amount of arylboronic acid 81 in refluxing xylenes. From 82, the authors utilized their previously developed palladium-catalyzed decarboxylative asymmetric allylic alkylation conditions¹¹ to provide enantioenriched 83 in 94% yield and 88% ee. Ketone 83 was transformed to its ethyl enol ether 84, the substrate for the key Mannich cyclization step in the synthesis. Initial monoreduction of the glutarimide to hemiaminal 85 followed by BF₃•OEt₂-mediated N-acyl iminium ion cyclization provided tricyclic lactam 86 as a single diastereomer in 89% yield. 86 was converted to amino alcohol 87 through an efficient one-pot reduction procedure where initial L-Selectride treatment diastereoselectively reduced the ketone (>19:1 dr) and subsequent LiAlH₄ reduction at reflux converted the lactam to the corresponding tertiary amine. Olefin cross-metathesis with 88 using the Hoveyda–Grubbs II catalyst (HG-II) provided vinylboronic acid ester 89 that was oxidized to a transient aldehyde which underwent hemiacetalization to provide (−)-myrifabral A (18a/b) in 85% yield. Using an adaptation of She’s previously reported conditions, (−)-myrifabral A (18) was converted to (−)-myrifabral B (19) in 44% yield. Overall, while longer than the She approach, the Stoltz synthesis is still relatively concise (7–8 steps) and importantly provides the first catalyst-controlled asymmetric synthesis of these targets – and indeed any Myrioneuron alkaloid – to date.

Scheme 8.

The Stoltz group’s enantioselective synthesis of myrifabrals A (18) and B (19).

2.4. Total Synthesis of Myrioneurinol

Myrioneurinol (3) was isolated by Bodo and coworkers in 2007^2c but was not prepared by total synthesis until 2014, when its racemic synthesis was reported by the Weinreb group (Scheme 9).¹² Myrioneurinol (3), with its unique C-5 quaternary center, is arguably the most structurally complex Myrioneuron alkaloid synthesized to date.

Scheme 9.

Weinreb’s synthesis of spirocycle 106.

The Weinreb synthesis begins with a C-alkylation of δ-valerolactam (90) with 6-bromo-1-hexene (91) to furnish lactam 92. Cbz-protection followed by ozonolysis yielded aldehyde 93, which underwent a Horner–Wadsworth–Emmons olefination to give the (E)-α,β-unsaturated ester 94. In the first key step of the synthesis, an intramolecular Michael addition promoted by titanium tetrachloride and triethylamine provides the spirocyclization product 96 in 86% yield as a single diastereomer, installing the C-5 quaternary stereocenter via a proposed chelated Ti-intermediate 95. A protecting group switch gave the N-benzyl lactam 98, which was then converted to aldehyde 100 via Weinreb amide 99 in good overall yield. After conversion to the pyrrolidine enamine 101, NCS treatment and subsequent hydrolysis provided the α-chloroaldehyde 102 as an inconsequential mixture of diastereomers (82% overall). Condensation with O-TBS-hydroxylamine gave O-TBS-α-chlorooxime 103, the precursor for the second key step in the authors’ approach. Using their previously developed conditions,¹³ the enolate of dimethyl malonate was added in a conjugate fashion to nitrosoalkene 104 (formed in situ with TBAF from the chlorooxime 103) yielding oximes 105a/b as a ~5:1 E/Z mixture but with perfect control of the tertiary C-7 stereocenter. Reductive hydrolysis of the oxime then delivers aldehyde 106.

106 underwent further reduction/lactonization and Krapcho decarboxylation to deliver γ-lactone 107 (Scheme 10). Lactone was then opened to a Weinreb amide-alcohol, which was MOM-protected (108) and reduced to yield aldehyde 109. A Wittig reaction with the Seyferth β-trimethylsilylethyl ylide derived from 110 converted 109 to the allyl silane 111 as a 2.5:1 E/Z mixture in 74% yield (2 steps). Debenzylation under Birch conditions (Na/NH₃) gave an N-H lactam that was tosyl-protected to provide 112. In preparation for the final C─C bonding forming step, DIBAL-H reduction of 112 furnished hemiaminal intermediate 113 in situ. Addition of anhydrous ferric chloride generated N-sulfonyliminium ion intermediate 114, which engaged the pendant allylsilane in an aza-Sakurai reaction to give tricyclic compound 115 (78%) as a single diastereomer. 115 bears all stereocenters of the target 3 in the correct configuration. To complete the synthesis, ozonolysis of 115 converted its vinyl group to an aldehyde, which was reduced and MOM-protected to give diether 116. Birch conditions (Li/NH₃) removed the N-Ts group, providing secondary amine 117. Treatment with 6 N HCl led to the hydrolytic cleavage of the distal MOM group and N-cyclization of the other to form the final 1,3-oxazine ring, furnishing (±)-myrioneurinol (3) in 27 steps from commercial δ-valerolactam (90). The Weinreb synthesis features several highly diastereoselective C─C bond-forming transformations but was lengthened by multiple protecting group switches and redox manipulations. Importantly, though, it constitutes the first synthesis of one of the more complex natural products in the Myrioneuron alkaloid family.

Scheme 10.

Completion of the Weinreb synthesis of (±)-myrioneurinol (3).

In 2022, the Ma group reported a more concise racemic synthesis of myrioneurinol (3) centering upon a key intramolecular [2+2] cycloaddition and a retro-Mannich/Mannich transformation (Scheme 11).¹⁴ Similar to the Weinreb approach, their synthesis began with alkylation of δ-valerolactam (90), this time with (6-iodohex-1-ynyl)trimethylsilane (118), and subsequent Boc-protection efficiently provided lactam 119. DIBAL-H reduction and elimination affords the corresponding enamine-yne 120 that can be further elaborated to 122 through acetylide acylation with Weinreb amide 121. To establish the tricyclic core, addition of silver hexafluoroantimonate and t-BuCl to 122 promotes an intramolecular [2+2] cycloaddition (presumably in a stepwise fashion) furnishing cyclobutene 123 in 47% yield. Under acidic conditions a key retro-Mannich fragmentation followed by reclosure by Mannich reaction remodels 123 into the core tricycle 124 with the net expansion from a 4- to 6-membered ring. α-Methylenation of 124 with sodium methoxide and paraformaldehyde provides dienone 125. To form the final 1,3-oxazine ring of the natural product, treatment of 125 with TFA and formalin cleaves the Boc group, performs N-hydroxymethylation (to intermediate 126), and promotes subsequent intramolecular oxa-Michael addition to provide tetracycle 127 (47%). Stereoselective hydrogenation followed by α-ethoxycarbonylation delivers β-keto ester 128 containing the tetracyclic core and all chiral centers of the natural product. Reduction of the ketone and conversion of the resulting alcohol to its xanthate 129 set the stage for Barton–McCombie deoxygenation to remove the superfluous keto unit. This step could be effected under standard conditions (Bu₃SnH, AIBN) and subsequent lithium aluminum hydride reduction in the same pot completes the synthesis of myrioneurinol (3) in 14 steps from 118. Overall, the Ma synthesis of (±)-myrioneurinol (3) is built upon a creative intramolecular [2+2] cycloaddition and retro-Mannich/Mannich sequence which enables construction of the central tricyclic core of the natural product in a highly concise fashion.

Scheme 11.

The Ma group’s synthesis of (±)-myrioneurinol (3).

The most recent synthesis of myrioneurinol (3) was reported in 2022 by our own group (Scheme 12).¹⁵ Our approach is distinct from these earlier syntheses, exploiting hidden symmetry within the target to construct the polycyclic piperidine core via a desymmetrizing double reductive amination. Our work also provides the first asymmetric entry to this natural product.

Scheme 12.

The Smith group’s synthesis of (±)-myrioneurinol (3).

Starting with commercially available pentachlorocyclopropane (130), decagram quantities of bicyclic chlorodiketone 132 were able to be prepared via a known sequence of elimination to the cyclopropene with KOH, Diels–Alder cycloaddition with cyclopentadiene, ring-opening to bicycle 131, and subsequent basic hydrolysis in one pot.¹⁶ O-Alloc protection provides 133, which under palladium-catalyzed Tsuji–Trost decarboxylative allylation conditions can furnish mono-allylated diketone 134 after in situ dechlorination with Zn/AcOH. A Michael addition to acrolein under mild aqueous conditions provided key tricarbonyl precursor 135/5-epi-135 in 75% yield, albeit in moderate dr (1.4:1) favoring the desired diastereomer (135). From this epimeric mixture of diketo aldehydes 135/5-epi-135, the key desymmetrizing double reductive amination was performed utilizing N-benzylamine and NaBH₃CN at high temperature to yield tricyclic amine (±)-136 as a single diastereomer (62% based on desired epimer of 135). Next, chemoselective dihydroxylation of the strained bicyclic alkene and acetonide protection provides 137. A diastereoselective allylation followed by ring-closing metathesis with Hoveyda–Grubbs second generation catalyst (HG-II) delivered pentacyclic alkene 138 in 67% over four steps from 136. Subjection of 138 to standard hydrogenation conditions led to saturation of the alkene and concomitant hydrogenolysis of the N-Bn group, which was protected as N-tosylamide 139. Elimination of the C-10 tertiary alcohol in 139 provided trisubstituted alkene 140 in 92% yield.

Reduction of the alkene at this stage was found to uniformly yield the incorrect epimer at C-10. For these reasons, acetonide deprotection using a mild CeCl₃-based system¹⁷ delivered diol 141 (65%, 80% brsm), which was subjected to oxidative cleavage with PhI(OAc)₂¹⁸ followed by in situ reduction of the corresponding dialdehyde to yield the bridge-cleaved primary diol 142. Advancement to bis-MOM ether 143 (86%, two steps from 141) allowed for diastereoselective alkene reduction under hydrogen atom transfer conditions (Fe(acac)₃/PhSiH₃),¹⁹ providing the desired C-10 epimer 116 as the major product (dr = 12:1) in 64% yield. Since tricycle 116 is a known intermediate in the prior Weinreb synthesis,¹² submission to a similar sequence (Li/NH₃ followed by treatment with 6 N HCl) gave racemic myrioneurinol (3) in 35% yield (2 steps). Overall, our synthesis proceeds in 18 steps (~1% yield) from commercial materials.²⁰

An advantage of the developed desymmetrization-based approach is its adaptability to an asymmetric synthesis of myrioneurinol (3). Thus, substituting benzylamine for inexpensive (R)-α-methylbenzylamine (144) in the key double reductive amination of 135 led to a diastereoselective desymmetrization with reasonable selectivity (dr = 4:1) for one of the two diastereotopic ketones, allowing for the isolation of pure major isomer 145 in 34% yield (Scheme 13). This transformation sets the absolute configuration of four of the five stereocenters of myrioneurionol (3), including the quaternary center, in a single step. 145 can be advanced through the same sequence of reactions as the racemic series to deliver (−)-139 (>99% ee by HPLC), for which the absolute configuration was determined by single-crystal X-ray analysis. The synthesis of (−)-139 constitutes a formal asymmetric synthesis of (−)-ent-myrioneurinol [(−)-3]; given that (S)-α-methylbenzylamine is also readily available, access to the natural (+)-enantiomer via such a process should be possible.

Scheme 13.

The Smith group’s formal asymmetric synthesis of (−)-myrioneurinol (3).

3. Conclusions and Outlook

As a relatively recent addition to the array of secondary metabolites isolated from nature, the Myrioneuron alkaloids present an intriguing combination of structural complexity and promising preliminary biological activity. Increased attention from both isolation and synthetic chemists over the past two decades has resulted in steady growth of the family in tandem with the development of strategies to construct their scaffolds. Though such synthetic studies logically began with the simpler family members, recent efforts have begun to tackle more complex targets such as myrifabrals A and B (18–19) and myrioneurinol (3), some in an asymmetric fashion. Cumulatively, the total synthesis of 8 distinct family members has been achieved by 7 research groups. However, still more complex targets such as myrobotinol (4), myrifabine (14), and myritonines A–C (29–31) containing extended fused ring systems and differing functionality and connectivity remain unconquered. We believe that these more challenging members, and any new congeners added to the class, will serve to inspire chemists to develop novel tools and tactics to address their intricate structures, studies which in turn will empower their biological exploration.

Biographies

Jake Aquilina is from Long Island, New York. He earned his BSc in Forensic Science from CUNY John Jay College of Criminal Justice in 2019 and is currently a PhD student at UT Southwestern Medical Center with Prof. Myles Smith. His current research interests focus on the total synthesis of complex alkaloids.

Myles Smith is a native of Cape Town, South Africa. He obtained his BSc (Hons) degree at the University of Cape Town in 2006, and subsequently undertook graduate studies at Columbia University under the direction of Scott Snyder. After obtaining his PhD in 2015, Myles conducted postdoctoral training at The Scripps Research Institute and then at Stanford University with Phil Baran and Noah Burns, respectively. He began his independent career at UT Southwestern in 2019, where his lab seeks to streamline the synthesis of complex molecules of medicinal value and develop novel synthetic tools.