Total Synthesis of Shearinines D and G: A Convergent Approach to Indole Diterpenoids.

The first total syntheses of the indole diterpenoids (+)-shearinine G and D are disclosed. The successful routes rely on late-stage coupling of two complex fragments. Formation of the challenging trans-hydrindane motif was accomplished by diastereoselective, intramolecular cyclopropanation. A one-pot sequence consisting of Sharpless dihydroxylation/Achmatowicz reaction was developed to install the dioxabicyclo[3.2.1]octane motif. The indenone subunit was accessed by Prins cyclization. Tuning the electronic nature of the substituents on the parent arylcarboxaldehyde allowed access to divergent products that were further transformed into shearinines G and D. Riley-type oxidation of a bicyclic enone yielded a surprising stereochemical outcome.

Introduction

Shearinines G and D are complex indole diterpenoids from the Janthitrem class of natural products (Scheme 1). They were initially isolated from the marine fungi Eupenicillium spp. and Penicillium janthinellum and later from Escovopsis weberi, a fungal pathogen interfering with the symbiosis of Acromyrmex leaf‐cutter ants and the garden fungus Leucoagaricus gongylophorus. Other well‐known, closely related indole diterpenes include paspalicine and tremorgenic paspalinine, which have been the subject of several synthetic studies (Figure 1). These have culminated in the first total syntheses by Smith of the bioactive indole terpenoids as well as penitrem and nodulisporic acids (Figure 1), resulting in a variety of creative approaches to the synthetically challenging motifs. Recent syntheses of paspalicine, paspaline, nodulisporic acid C, emindole SB and emindole PB highlight the continued interest of the synthetic community in complex indole diterpenes.

Scheme 1

Shearinines G and D; approach to the trans‐hydrindane.

Figure 1

Related indole diterpenes.

Shearinines A–C were isolated for the first time in 1995 and have since been accompanied by an increasing number of related family members. They show broad variety of bioactivities, such as blocking of high‐conductance calcium‐activated potassium channels, cytotoxicity towards human leukemia HL‐60 cells, and anti‐insectan activity.[ , ] The group of Cichewicz reported the ability of shearinines D and E to inhibit Candida albicans biofilm formation. The size and structural complexity of these secondary metabolites render them attractive targets for total synthesis studies. Herein, we present the first total syntheses of shearinines G (1) and D (2) made possible through the development of modular routes. Salient features of the approaches include late‐stage, convergent coupling of two advanced fragments, intramolecular cyclopropanation that provides access to the trans‐hydrindane, and unexpected observations in connection with γ‐hydroxylation of a bicyclic enone in the presence of SeO2 to install the C13 hydroxy group found in the natural product.

Background and Retrosynthetic Analysis

A major challenge associated with the synthesis of shearinines was identified as the installation of the trans‐hydrindane decorated with two vicinal quaternary stereocenters. The thermodynamic preferences for trans‐ versus cis‐hydrindanes is complicated as it depends on the substitution pattern. For example, despite the inherent preference for the trans isomer in the parent hydrindane itself, methyl substitution at C8 can lead to overwhelming preference for the cis‐fused system.

A time‐honored approach to overcome the inherent substrate bias towards the undesired cis‐hydrindane largely involves recourse to hydroxy‐directed transformations, such as Simmons–Smith cyclopropanation or directed hydrogenation (Scheme 2).[ , ] In a complementary manner, in other highly diastereoselective approaches to the trans‐hydrindane core, the stereochemical relationships are set prior to installation of the 5/6 ring system. These tend to rely on diastereoselective functionalization of cyclohexenones. Most notably, Smith has pioneered the preparation of thermodynamically favored trans‐decalins, which are then subject to oxidative cleavage and subsequent cyclocondensation to furnish the derived trans‐hydrindanes.[ , ] In a complementary approach, conjugate addition to cyclopentenone derivatives followed by alkylation of the resulting enolate set the necessary vicinal anti relationship, which sets the stage for installation of the fused 6‐ring.[ , ] Another strategy to these relies on a sequence of 1,4‐addition and α‐alkylation reactions to generate 1,6‐dienes that are then subject to ring‐closing metathesis reactions.[ , , ] More recently, cationic cyclization has provided entry to the trans‐hydrindane, albeit as the minor product.

Scheme 2

Previous work and retrosynthesis.

Our retrosynthetic analysis led to disconnection of the nonacyclic system at the centrally located heterocycle, which results in two fragments of similar size and complexity, namely 3 and 4. In contrast to earlier work, however, we aimed to reduce manipulations of the fully assembled natural product skeleton to a minimum. In crafting our approach, we wondered whether intramolecular cyclopropanation by a diazo ketone of a methylene cyclohexane would provide proper control of the ring‐fusion configuration (cf. 5) and thus give access to trans‐hydrindane. Reductive opening of the cyclopropane and concomitant regioselective vinyl triflate formation was anticipated to complete the synthesis of the hydrindane fragment. The preparation of indene 3 was designed to be modular and inspired by early work by Magnus and Mansley.

Results and Discussion

Our focus for the preparation of the shearinine core was related to the stereoselective synthesis of the requisite trans‐hydrindane at C3 and C16. In our first‐generation route we prioritized control of the challenging 3R quaternary center, which necessitated attendant control of C16. Ample precedent in the literature suggested strong preference for cis‐hydrindane formation through the implementation of intramolecular cyclopropanation reactions. Thus, the initial plan was to generate ax‐9 and subsequently invert the configuration at C16 through the intermediacy of enone 12 and its stereoselective reduction (11→12→13). In this respect, oxidation of the cyclopentanone to the cyclopentenone would allow corrective action to be taken, wherein conjugate reduction from the olefin face opposite the axial methyl group at C3 would produce the trans‐hydrindane. The feasibility of this approach was evaluated with 6 (Scheme 3).

Scheme 3

First‐generation route to cyclopropyl ketone 13. Reagents and conditions: a) 1 M aq. HCl–THF (2:1), RT, 94 %; b) LiN(SiMe3)2, Me3SiCl, THF, then Comins reagent, −78 °C→RT, 94 %; c) Pd(PPh3)4 (10 mol %), Et3N, CO (1 atm), MeOH, RT→65 °C, 71 %; d) i‐Bu2AlH, CH2Cl2, −78 °C→RT, 88 %; e) Ac2O, pyridine, DMAP (cat.), 0 °C, 94 %; f) LiN(SiMe3)2, (Me2N)3PO, THF, then t‐BuMe2SiCl, −78 °C→RT, then 40 °C, 72 %, d.r. 5:1; g) MeLi (2.6 equiv), Et2O, 0 °C, 71 % (2 cycles); h) LiN(SiMe3)2, THF, −78 to −40 °C, then F3CCH2O2CCF3; i) H2O (1 equiv), Et3N, p‐ABSA, MeCN, RT, 68 % (over 2 steps); j) Rh2(cap)4 (2 mol %), CH2Cl2, 40 °C, 79 %, d.r. 10:1; k) LiN(SiMe3)2, −78→−40 °C, then Me3SiCl, −78→−40 °C; l) NBS, THF, −78 °C, 74 % (two steps); m) Li2CO3, LiBr, DMF, 120 °C, 68 % (78 % brsm); n) Li, NH3, THF, NH4Cl, −78 °C, 51 %. DMAP=N,N‐dimethyl‐4‐aminopyridine, p‐ABSA=4‐acetamidobenzenesulfonyl azide, cap=caprolactamate, NBS=N‐bromosuccinimide.

Sequential dioxolane cleavage in 6 and treatment with LiN(SiMe3)2 followed by the Comins reagent at −78 °C furnished a vinyl triflate, which was subjected to palladium‐catalyzed carbonylation (1 atm CO, Et3N, MeOH) to give enoate 7. Reduction to the allylic alcohol and acetylation afforded allylic acetate 8. Formation of the silyl ketene acetal at low temperatures followed by slow warming of the reaction mixture to 40 °C led to carboxylic acid 9 with a d.r. of 5:1, favoring the C16α isomer (ax‐9) in 70 % yield in which the side chain is axially positioned. As the diastereomers were inseparable, we proceeded to move forward with the diastereomeric mixture.

Treatment of carboxylic acid 9 with excess MeLi led to its conversion into the corresponding methyl ketone in 71 % yield. Following Danheiser's protocol, it was sequentially treated with LiN(SiMe3)2 in THF at −78 °C and trifluoroethyl trifluoroacetate (TFEA). The mixture was warmed to −40 °C and the unpurified 1,3‐diketone was subjected to Et3N (1.5 equiv) in MeCN in presence of water (1.0 equiv). 4‐Acetamidobenzenesulfonyl azide (p‐ABSA) was added dropwise as a solution in MeCN at ambient temperature, to yield diazoketone 10 in 68 % yield. Intramolecular cyclopropanation of 10 mediated by Rh2(cap)4 in CH2Cl2 afforded cyclopropyl ketones 11 and 13 in 79 % yield and d.r. 10:1, favoring cis‐hydrindane 11 as determined by NOE studies on a closely related substrate (see the Supporting Information). Formation of the targeted trans‐hydrindane necessitated inversion of configuration at C16. This feat was accomplished through a desaturation/reduction sequence. After extensive optimization (see the Supporting Information), an efficient three‐step sequence was developed. Silyl enol ether formation with LiN(SiMe3)2 and Me3SiCl was followed by sequential α‐bromination using NBS and elimination in the presence of LiBr and Li2CO3 to yield enone 12. Finally, treatment with Li (2.3 equiv) afforded cyclopropyl ketone 13. The relative configuration of 13 as a trans‐hydrindane was confirmed by X‐ray crystallography (CCDC 1979001).

In the previous sequence, Claisen rearrangement 8→9 produced an inseparable mixture of diastereomers (eq‐9 and ax‐9 in Scheme 3), which were taken without separation through the steps described. As shown in Scheme 4, analysis of the mixture of products from the cyclopropanation reaction led to the observation that the 16S diastereomer of 10 from eq‐9 yielded desired trans‐hydrindane 13. The stereoselective formation of this seemingly unexpected product finds support in a single literature precedent by Corey and co‐workers in which intramolecular cyclopropanation of a steroid ABC ring precursor led to closure of the trans‐fused steroid D ring. Based on these findings, we redesigned our route to provide more direct access to trans‐hydrindane 13 (Scheme 5).

Scheme 4

Analysis of the Claisen diastereomer.

Scheme 5

Optimized synthesis of vinyl triflate 26. Reagents and conditions: a) Li, NH3, THF, −78 °C, then isoprene, −78 °C→RT, then Me3SiCl, NEt3, THF, 0 °C, 75 %; b) MeLi, (Me2N)3PO, 15, THF, 0 °C, 73 %; c) aq. HCl–THF, 50 °C; d) AuCl (cat.), AgOTf (cat.), p‐TsOH (cat.), PhMe, 50 °C, 71 % (2 steps); e) LiN(SiMe3)2, (Me2N)3PO, THF −78 °C→RT, then MeOH, DBU (cat.), 71 % (d.r.>20:1); f) KOt‐Bu, MePPh3Br, PhMe, 50 °C, 70 %; g) aq. LiOH, EtOH, 40 °C; h) MeLi, Et2O, −10 °C, 65 % (2 steps); i) LiN(SiMe3)2, F3CCH2O2CCF3, −78→−40 °C; j) NEt3, H2O, p‐ABSA, MeCN, 63 %; k) Rh2(OAc)4, CH2Cl2, 40 °C, 74 %; l) NaNp., t‐BuOH, then isoprene, (Me2N)3PO, Comins reagent, THF, −78 °C→RT, 57 %; m) OsO4 (10 mol %), (DHQ)2PHAL (15 mol %), K3Fe(CN)6, K2CO3, MeSO2NH2, t‐BuOH–H2O–THF (10:10:1), then K3Fe(CN)6, K2CO3, t‐BuOH–H2O, 57 %, d.r.>15:1; n) p‐TsOH (cat.), CuSO4, PhH, RT; o) SeO2, 1,4‐dioxane, 90 °C, 30 % (2 steps). DBU=1,8‐diazabicyclo[5.4.0]undec‐7‐ene, p‐ABSA=4‐acetamidobenzenesulfonyl azide.

In this respect, cyclopropanation precursor 20 with the 16S configuration was anticipated to be accessible through α‐alkylation and olefin methenylation. Controlling the configuration at C16 would not be of concern as the desired configuration would be expected to be thermodynamically preferred because the acetic acid sidechain is positioned equatorially as shown in Scheme 4. Furthermore, our second‐generation route involved gold‐catalyzed propargyl ketone cycloisomerization of 16 inspired by reports from the groups of Krause and Wipf.

Wieland–Miescher ketone derivative 14 was subjected to Birch reduction conditions, and the lithium enolate generated in situ was isolated as the corresponding Me3Si‐enol ether. Birch reduction on a gram to multigram scale proceeded reliably using large excess of Me3SiCl/NEt3 (4 equiv). In situ lithium enolate generation was followed by alkylation with readily accessible propargyl iodide 15 (see the Supporting Information) to give 16. Dioxolane hydrolysis and subjecting the unpurified ketone to Krause's conditions proved to be crucial for high yields of furan 17. Notably, the reaction also proceeded with p‐TsOH in toluene at elevated temperatures but resulted in a lower yield of 17.

Enolization of 17 and alkylation with ethyl bromoacetate proceeded in 64 % yield to give a mixture of axial and equatorial substituted adducts in varying diastereomeric ratios, favoring desired eq‐18. Treatment of this mixture with MeOH and catalytic amounts of DBU gave desired C16 epimer eq‐18 with d.r. >20:1 (CCDC 1979002). Wittig olefination afforded methylene cyclohexane 19 in 70 % yield. Ester 19 was hydrolyzed in aqueous ethanol in presence of LiOH at elevated temperature, and the carboxylic acid was then converted into diazoketone 20. Subjecting 20 to Rh2(OAc)4 in CH2Cl2 at 0 °C gave a single product, which was taken forward in the synthesis. The relative configuration of 23 was confirmed by X‐ray crystallographic analysis (CCDC 1979000).

The stereochemical outcome of the cyclopropanation was surprising. While Corey had noted a similar outcome in the context of a steroid synthesis, this observation has not, to the best of our knowledge, been exploited for the synthesis of other trans‐hydrindanes. We believe that the reaction proceeds through transition state 22, which avoids unfavorable interactions present in 21. Cyclopropyl ketone in 23 was reductively opened with sodium naphthalenide in the presence of freshly distilled t‐BuOH, and the resulting enolate was trapped with the Comins reagent to yield vinyl triflate 24.

The dioxabicyclo[3.2.1]oct‐3‐en‐2‐one motif in shearinines G (1) and D (2) has been the subject of a number of studies. While early computations suggested the Achmatowicz reaction provided the undesired configuration of the intermediate alcohol at C7 thereby disfavoring acetal closure, Saxton later showed in synthetic studies towards paspalinine that the dioxabicyclo[3.2.1]oct‐3‐en‐2‐one was accessible by an Achmatowicz reaction.[ , ] The convergent nature of our approach required that the dihydroxylation be effected in a diastereoselective manner. This is in contrast to the implementation of dihydroxylation in Saxton's synthetic studies.

Sharpless dihydroxylation using commercially available AD‐mix α gave low conversion of 24. Inspired by a report by Nicolaou and others, a so‐called “super” AD‐mix was employed, which consists of a higher loading of osmium(VIII), ligand and stoichiometric oxidants. Under these conditions, the enantioenriched diol was isolated along with hemiketal 25. Further optimization of the osmium‐catalyzed transformation allowed a one‐pot synthesis of hemiketal 25 by adding a second portion of K3[Fe(CN)6] and K2CO3 after overnight reaction. The synthesis of vinyl triflate 26 was completed by ketal formation in the presence of catalytic p‐TsOH and Riley oxidation in 30 % yield over two steps. In the initial prospecting experiments this reaction was conducted on small scale and produced product as a single isomer. On the basis of related oxidations in similar systems, we decided to proceed with the synthesis route.

Synthesis of Shearinine G

The recent synthesis of nodulisporic acid C by Pronin and co‐workers featured an elegant, clever cycloisomerization approach to the indenol. However, the substrate requirements reported to secure high diastereocontrol preclude its use for the asymmetric synthesis of the shearinines. We reasoned that the use of a Prins cyclization would provide access to a versatile intermediate that might be amenable to diversification at the indenopyran subunit and enable access to other closely related natural products.

Accordingly, the synthesis of aryl stannane 3 commenced with Suzuki–Miyaura coupling of bromoarene 27 with vinyl boronic acid pinacol ester 28 (Scheme 6). Treatment of benzaldehyde 29 with Me3SiOTf (1 equiv) in CH2Cl2 at −20 °C triggered Prins cyclization to afford a 1:3 mixture of homoallylic alcohol 31 and desired allylic alcohol 30. However, when 4 equivalents of the Lewis acid were used the ratio improved to roughly 1:8. Further experimentation revealed that the best ratio (1:10) and yield (69 % combined) of undesired to desired indenol 30:31 were obtained at around −15 °C, albeit with traces of unreacted starting material. Sequential oxidation to the enone using Dess–Martin periodinane and nitro group reduction with SnCl2 in ethanol at 70 °C was followed by N‐Boc protection in the presence of 1 equivalent of guanidinium chloride in a 10:1 mixture of EtOH and di‐tert‐butyl dicarbonate. Electrophilic bromination with NBS in the presence of AcOH occurred with high regioselectivity. Subsequent Stille coupling afforded stannane 3 in 69 % yield.

Scheme 6

Assembly of aryl stannane 3. Reagents and conditions: a) 27, Pd(PPh3)4 (10 mol %), K2CO3, DME, 80 °C, 63 %; b) Me3SiOTf (4.0 equiv), CH2Cl2, −15 °C, 69 %; c) DMP, CH2Cl2, 89 %; d) SnCl2⋅6 H2O, EtOH, 70 °C, 86 %; e) guanidinium chloride, EtOH– (t‐BuOCO)2O (10:1), RT, 89 %; f) NBS, AcOH‐1,4‐dioxane (1:1), 87 %; g) Pd(PPh3)4 (10 mol %), Me6Sn2, 1,4‐dioxane, 90 °C, 69 %. DME=1,2‐dimethoxyethane, DMP=Dess–Martin periodinane.

Aryl stannane 3 and vinyl triflate 26 were coupled under Corey's CuCl‐accelerated Stille reaction conditions (Scheme 7). Oxidative indole formation mediated by Pd(OCOCF3)2 yielded 32 in 62 % over two steps. Pyrolytic N‐Boc cleavage was accomplished by adsorption of 32 on neutral silica gel and heating under high vacuum to 90 °C. The NMR data of indole 33 thus obtained featured significant discrepancies compared to the literature data of (+)‐shearinine G (1). Since the largest deviation was observed for hydrogen and carbon atoms in the vicinity of the enone, we hypothesized that the Riley oxidation had unexpectedly afforded the undesired cis‐decalin. To investigate this further, the oxidation of 34 was scaled up, which allowed isolation of small amounts of a minor product [Eq. 1]. Analysis by X‐ray crystallography revealed that the minor product was indeed desired trans‐decalin 4 (CCDC 1977991).

Scheme 7

Completion of the synthesis of shearinine G (1). Reagents and conditions: a) Pd(PPh3)4 (30 mol %), LiCl, CuCl, DMSO–CH2Cl2 (3.5:1), 50 °C, 90 % C13‐β; b) Pd(OCOCF3)2 (3 equiv), NaOAc (6 equiv), DMSO, 60 °C, 62 % 32 over two steps, 77 % 36; c) neutral SiO2, 2–3 Pa, 90 °C, 57 % 33, 67 % 1. DMSO=dimethyl sulfoxide.

Based on this unexpected finding, a study was initiated of the allylic oxidation reaction (Table 1). Conducting the reaction under microwave irradiation allowed us to maintain short reaction times (3×20 min) at 80 °C, whereupon a 1.7:1 ratio of diastereomers was obtained. The addition of pyridine did not affect the diastereomeric ratio, but the addition of formic acid (pK a=3.75) as a cosolvent proved beneficial with desired 4 to undesired 26 being formed in a 1:1.2 ratio. The use of approximately 15 equivalents of selenous acid (pK a=2.64) as a reagent led to a 1:1 ratio of the two epimers. The experiments involving acidic conditions suggest the possibility of an incipient trend, with stronger acids leading to a greater proportion of 36. We were, however, unable to examine this further, because under stronger acidic conditions we observed instability of the bicyclic acetal.

Table 1

Studies of the Riley oxidation.

Reagent	T [° C]	Additive	26:4
SeO2 (2 equiv)	80–90	–	3:1
SeO2 (excess)[a]	80	–	1.7:1
SeO2 (excess)[a]	80	pyridine (10 equiv)	1.7:1
SeO2 (excess)[a]	80	HCO2H (pK a=3.75; 90 equiv)	1.3:1
SeO2 (excess)[a]	80	HCO2H (pK a=3.75; 900 equiv)	1.2:1
H2SeO3 (15 equiv)	80	(selenous acid: pK a=2.64)	1:1

[a] Reactions were carried out on an approximately 0.5 mg scale with 10–15‐fold excess of reagent.

–

–

Under optimized conditions, 4 was isolated in 33 % yield along with 34 % of 26. Stille coupling of 3 and 4, followed by Pd(OCOCF3)2‐mediated indole formation in the presence of NaOAc afforded 35 in 69 % yield. Pyrolytic removal of the N‐Boc protecting group completed the synthesis of (+)‐shearinine G (1) in 67 % yield. The spectroscopic data (1H NMR,13C NMR, IR, [α]D) and high resolution mass collected for 1 were in agreement with that reported for the natural product.

The experimental outcome of the Riley oxidation under standard conditions came as a surprise as it implicated an unexpected isomerization process. In addition, it was reminiscent of computational studies by Smith, who examined the thermodynamic preferences of two diastereomers of paspalinine that differ in the relative configuration of the tertiary alcohol and the bicyclic ketal analogous to 4 and 26. The formation of stereoisomeric products in our study is depicted in Scheme 8. Following generation of the dienol derived from 34, selenium dioxide may react from convex or concave faces to give I or II, respectively. Irrespective of the diastereoface from which initial selenylation takes place, we hypothesize that in the presence of acid, equilibration of the C11 stereocenter occurs, thus leading to the formation of the desired epimer after [2,3]‐sigmatropic rearrangement.

Scheme 8

Possible mechanistic pathways for the Riley oxidation.

Synthesis of Shearinine D

Shearinine D is arguably the most complex shearinine reported to date. Its indenopyrane subunit features a 22S and 23S configuration. We envisioned to use the same overall strategy as for the indenone in shearinine G but go through styrene intermediate 37 to access the C22 alcohol enantioselectively (Scheme 9). The synthesis commenced with Suzuki–Miyaura coupling of known benzaldehyde 36 with vinyl boronic pinacol ester 28. Treatment with TMSOTf (1 equiv) in CH2Cl2, followed by Buchwald–Hartwig coupling with H2NBoc yielded 37 in 20 % yield over three steps. None of our attempts at catalytic asymmetric hydroboration of 37 were fruitful, confirming that trisubstituted olefins are still challenging substrates for these transformations. For example, Hartwig's CuI−H catalyzed hydroboration/oxidation sequence that had been showcased on a single trisubstituted olefin substrate did not lead to reaction of 37. Accordingly, we resorted to Brown's traditional chiral reagent approach. With readily available (+)‐monoisopinocampheylborane the corresponding benzylic alcohol was obtained in 58 % yield with e.r. 3.7:1. Steglich esterification with (R)‐(−)‐α‐methoxyphenylacetic acid (38) afforded 39 in 57 % yield with a d.r.>20:1. Ester hydrolysis, followed by TBS protection and directed metalation with Me3SnCl afforded arylstannane 40 in 56 % yield over three steps.

Scheme 9

Synthesis of enantiomerically pure aryl stannane 40. Reagents and conditions: a) 36 (1.05 equiv), Pd(PPh3)4 (10 mol %), aq. K2CO3–DME (1:1.9), 80 °C, 65 %; b) Me3SiOTf (1.0 equiv), CH2Cl2, 0 °C→RT, 52 %; c) H2NBoc (1.2 equiv), Pd(OAc)2 (10 mol %), XPhos (30 mol %), Cs2CO3 (1.4 equiv), 1,4‐dioxane, 100 °C, 57 %; d) (+)‐IpcBH2 (4.8 equiv), THF, 0 °C, then NaBO3, THF–H2O (1:1), RT, 58 %, e.r. 3.7:1; e) 38 (1.7 equiv), DCC (1.7 equiv), DMAP (10 mol %), CH2Cl2, 57 %, d.r.>20:1; f) LiOH⋅H2O (10 equiv), THF–H2O (1:1), 95 %; g) t‐BuMe2SiCl (6.0 equiv), imidazole (8.0 equiv), DMF, RT, 86 %; h) t‐BuLi (3.9 equiv), Et2O, −40→−15 °C, then Me3SnCl (2.0 equiv), −78 °C→RT, 69 %. Ipc=isopinocampheyl, DCC=N,N‐dicyclohexylcarbodiimide.

CuCl‐accelerated Stille coupling of arylstannane 40 and vinyl triflate 4 was followed by oxidative indole formation to yield 41 (Scheme 10). Subsequent desilylation with HF⋅pyridine and N‐Boc deprotection by adsorption on neutral silica gel and heating on high vacuum completed the synthesis of shearinine D (2). Notably, the reaction was stopped before full conversion to avoid elimination of the benzylic alcohol in shearinine D (2). The spectroscopic data (1H NMR,13C NMR, IR, [α]D) and high‐resolution mass spectral data collected for 2 were in agreement with those reported for the natural product.[ , ]

Scheme 10

Completion of the synthesis of shearinine D (2). Reagents and conditions: a) Pd(PPh3)4 (30 mol %), LiCl, CuCl, DMSO–CH2Cl2 (3.5:1), 50 °C, 86 %; b) Pd(OCOCF3)2 (9 equiv), NaOAc (19 equiv), DMSO, 60 °C, 93 %; c) HF⋅py, pyridine, MeCN, 0 °C to RT, 87 %; d) neutral SiO2, 2–3 Pa, 90 °C, 56 %; DMSO=dimethyl sulfoxide, py=pyridine.

Conclusion

In summary, we have accomplished the first total syntheses of (+)‐shearinines G (1) and D (2) through convergent and efficient routes. Highlights are a gold(I)‐catalyzed cycloisomerization to access 2‐isobutenyl furans, intramolecular rhodium(II)‐catalyzed cyclopropanation to form the trans‐hydrindane motif with two quaternary stereocenters, and one‐pot Sharpless dihydroxylation/Achmatowicz reaction en route to the dioxabicyclo[3.2.1]octane. Furthermore, the unexpected preference of the late‐stage Riley oxidation for the cis‐C13‐hydroxydecalin from the trans‐decalin precursor was investigated and rationalized. The modular route towards the indenopyran subunit and our convergent strategy allow access to related natural products and congeners.

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Click here for additional data file.