2020-02-22 17:06:17
Chao Zheng and Shu-Li You
1.Introduction
The positive interplay between the studies on natural product synthesis and synthetic methodology development constitutes the major theme of fundamental organic chemistry . The fasci- nating structures of various natural products serve as the inspi- ration and driving force of endless emerging and evolving unprecedented synthetic methods. On the other hand, natural product synthesis is generally regarded as a premier stage, where a synthetic method can exhibit its practicality and robustness.
Indole-based polycyclic structures have constantly received intensive research interests due to their wide occurrence in numerous alkaloids that possess diverse biological activities.
Dearomatization of indole derivatives has long been viewed as an attractive strategy to readily access sophisticated spiro or fused polycyclic indoline scaffolds. Accordingly, Qin et al. developed an elegant cascade (or stepwise) reaction of cyclopropanation/ring-opening/iminium (CRI) cyclization of tryptamine derivatives, which enables the efficient assembly of a variety of indoline alkaloids.8 Meanwhile, Ma and co-workers carried out systematic studies on intramolecular dearomative oxidative coupling (IDOC) reactions of indoles and accom- plished the total synthesis of many indoline alkaloids.9
In recent years, catalytic asymmetric dearomatization (CADA) reactions have been recognized as powerful methods for the conversion of readily available planar aromatic feed- stocks to three-dimensional polycyclic molecules.10–18 Partic- ularly, the synthesis of many indole-based natural products becomes viable with the key ring system and the stereo- chemistry being established via a CADA reaction. A major advantage of these syntheses is that they do not rely on the chiral pool strategy. The absolute conftguration of the product is solely controlled by the chiral catalyst. Therefore, not only natural products themselves, but also their corresponding unnatural enantiomers, diastereoisomers, and even libraries of analogs can be easily synthesized, which is of great impor- tance in medicinal chemistry.
Herein, we summarize the most recent progress on the total synthesis of indole-based natural products enabled by an asymmetric dearomatization reaction as the key step (Fig. 1). Several ring-forming strategies for the construction of key polycyclic scaffolds will be ftrst highlighted. Then, the various applications of asymmetric dearomatization reactions in the total syntheses of these natural products will be discussed accordingly. Finally, we present a personal perspective of this dynamic research fteld.
2.General strategies for the formation of key ring systems
In general, a series of ring-formation strategies have been devel- oped in the total synthesis of indole-based natural products via a CADA reaction (Fig. 2). Most of these strategies take advantage of the inherent strong nucleophilicity of the C3 position of 3- substituted indoles and the electrophilicity of 3,3-disubstituted indolenine cations (types I to IV). By devising appropriately teth- ered nucleophiles and electrophiles, cascade cyclization reactions may be triggered. On the other hand, the recent enantioselective total synthesis of (+)-hinckdentine A involved a dearomative Heck- type cyclization, which can be regarded as the connection between the C2 position of indole with an electrophile (type V). Notably, the asymmetric dearomatization of indole derivatives via a visible- light-mediated sequential single-electron transfer process has also emerged, which allowed the formal installation of an external nucleophile at the C3 position of indole with concomitant cycli- zation (type VI). By employing these strategies, spiro, fused, and even bridged polycyclic scaffolds can be assembled by carefully manipulating the reaction sequence. In the following sections, detailed discussions will be presented on how these strategies have been applied in the construction of the core structures of indole-based natural products.
3.Type I cyclization
The most frequently employed strategy takes advantage of indole substrates possessing a pendant nucleophile on the C3 position, where tryptamine/tryptophol derivatives are among the commonly used substrates. A Friedel–Cras-type addition with an external electrophile generates a 3,3-disubstituted indolenine intermediate whose iminium moiety is readily trapped by the nucleophilic side chain, yielding the corre- sponding hexahydropyrrolo[2,3-b]indole (HPI) or tetrahydro- 2H-furo[2,3-b]indole core structures, which lays a solid foun- dation for further synthetic elaborations.
As early as in 2000, O¯mura, Smith and coworkers furnished the enantioselective synthesis of 3a-hydroxyfuroindoline from tryptophol by employing the Sharpless asymmetric epoxidation protocol (Scheme 1). This strategy was successfully applied in their study towards the total synthesis and the absolute conftg- uration assignment of (+)-madindoline A and (+)-madindoline B, which are potent inhibitors of interleukin 6 (IL-6).19 N-Alkylated tryptophol (+)-2 was prepared from oxazolidinone (+)-1 via 18 linear steps. The (+)-DET/Ti(OiPr)4-promoted Sharpless asym- metric epoxidation/cyclization, a process involving the asym- metric dearomatization of indole, served as the ftnal step to complete the synthesis of (+)-madindoline A (31% yield) and its diastereoisomer (—)-madindoline B (14% yield).
(—)-Flustramine B and (—)-debromoustramine B belong to a family of marine alkaloids isolated from the Bryozoa Flustra foliacea, which possess skeletal and smooth muscle relaxant activity and signiftcant butyrylcholinesterase inhibitory activity, respectively.20 Their embedded HPI core structure makes them popular targets in recent asymmetric total synthesis studies.21–23 In 2004, the MacMillan group reported the enantioselective synthesis of (—)-ustramine B and (—)-debromoustramine B (Scheme 2).24 In the presence of an imidazolidinone catalyst (S,S)-C1,25 6-bromo-substituted tryptamine derivative 3 reacted smoothly with acrolein in a cascade dearomative addition/ cyclization sequence. HPI compound 4 with a pendant primary alcohol group was achieved in 78% yield with 90% ee aer in situ reduction by NaBH4. The 3-hydroxy propyl side- chain was readily converted to a prenyl moiety in a three-step, two-pot sequence. The primary alcohol in 4 was ftrst mesylated and then transformed into terminal alkene by treat- ment with 2-nitrophenyl selenocyanate/H2O2. A cross- metathesis with 2-methyl-2-butene by Grubbs II catalyst incor- porated the prenyl moiety in 5. Finally, (—)-ustramine B and (—)-debromoustramine B were obtained via selective reduction of the N-Boc group and removal of the 6-bromo substituent in the latter case.
In 2012, Zhang and Antilla reported the asymmetric synthesis of (—)-debromoustramine B from tryptamine deriv- ative 6 and methyl vinyl ketone (MVK) using a chiral phosphoric acid (CPA) catalyst (Scheme 2).26 With TRIP-CPA (R)-C2 (ref. 27) as the optimal catalyst, the HPI core of 7 was assembled with two MVK segments incorporated at the N1 and C3a positions in good yield (90%) and enantioselectivity (93% ee). Subsequently, the two methyl ketone moieties were transformed into two prenyl groups of 8 via sequential Wittig reaction/oleftn isom- erization. (—)-Debromoustramine B was obtained in quanti- tative yield via Red-Al reduction of 8.
During their continuous efforts towards transition-metal- catalyzed asymmetric allylic dearomatization reactions,12 You's group developed a series of efficient asymmetric syntheses of the ustramine family of natural products (Scheme 3).28 With [Ir(cod) Cl]2 as the Ir-precursor, Carreira P/oleftn ligand L1 (ref. 29) as the optimal chiral ligand, and Fe(OTf)2 as the promoter, the dear- omatization reaction of tryptamine derivative 9 occurred with cinnamyl alcohol, leading to N1,C3a-biscinnamyl substituted HPI product 10 in good yield (87%) with high enantioselectivity (90% ee). The cinnamyl groups could be converted to prenyl groups through cross-metathesis reaction by the Grubbs II catalyst. The ftnal reduction of the N-CO2Me group furnished the total synthesis of (—)-debromoustramine B. Very recently, You and coworkers discovered an enabling Pd-catalyzed enantioselective dearomative prenylation of indole derivatives.30 By employing a Pd-complex derived from [Pd(prenyl)Cl]2 and a newly synthe- sized chiral phosphoramidite ligand allylphos (L2), prenyl carbonate could react with a large array of indole derivatives 12. In particular, the reaction of the 6-bromo-substituted tryptamine derivative 12a afforded the prenylated HPI product 13 in excellent yield (95%) and enantioselectivity (94% ee). The following N- prenylation reaction resulted in the formation of compound 14, which is a known precursor of (—)-debromoustramine B.24 Meanwhile, (—)-ustramine B could also be prepared from 14 aer deprotection of the N-Boc group and reductive amination. On the other hand, the reaction of 12b with a different prenyl precursor bearing a tethered silyl ether group provided the dearomatized product 16. The following reduction of the ester group and removal of the silyl group accomplished the synthesis of (—)-pseudophrynaminol. In addition, with the readily available tryptophan derivative Boc-L-Trp-OMe 12c as the starting material, the synthesis of prenylated HPI product 18 could be realized on the gram-scale with a catalyst loading as low as 0.5 mol%.
Subsequent protecting group manipulation and coupling with L- leucic acid furnished the total synthesis of mollenine A, which corrected the assignment of its absolute conftguration.31–33 Notably, this Pd-catalyzed enantioselective dearomative pre- nylation reaction tolerates highly functionalized indole deriva- tives as substrates. Under the standard conditions, the L-Trp-L- leucic acid 12d was involved in a cascade dearomative prenylation/cyclization/lactonization reaction, which allowed the one-step synthesis of mollenine A on the gram-scale, and also a library of analogues of mollenine A to be built from the corre- sponding tryptophan derivatives synthesized by reacting with various a-amino acids or chiral a-hydroxyl carboxylic acids.
(—)-Diazonamide A belongs to a structurally unique class of secondary metabolites isolated from the colonial marine ascidian Diazona angulata and is a potent antimitotic agent.34,35 In 2011, MacMillan and coworkers reported the asymmetric total synthesis of diazonamide A (Scheme 4).36 Starting from functionalized indole boron ester 19 and 2-iodophenol 20, Suzuki coupling furnished 2-(indol-3-yl)phenol 21, whose complexity was further enriched to 23 in two steps. With 23 and propynal as substrates, the imidazolidinone catalyst (R,R)-C1- enabled asymmetric dearomative addition/cyclization strategy was applied to assemble the key tetracyclic core with concomi- tant establishment of the quaternary chiral center at the C10 position, leading to the key intermediate 24 in good yield (86%) with excellent diastereoselectivity (>20 : 1 dr). Aer the exchange of the N-protecting group from PMB to TFA, ozonol- ysis furnished aldehyde 26. Subsequently, an intramolecular aldol reaction closed the 13-membered ring, affording 27 as a single diastereoisomer. The central oxazole ring (28) was then formed via a two-step oxidation/cyclodehydration sequence. In the presence of AgTFA, the thioester functionality of 28 was displaced by amine 29. The resulting intermediate 30 under- went a second cyclodehydration, furnishing bisoxazole 31. Aer debenzylation and treatment with PhNTf2, the bromo-bistriate intermediate 32 was obtained. The tandem Pd-catalyzed borylation/annulation of 32 occurred smoothly with B2pin2/KF under microwave heating, leading to the full installation of the carbon framework of (—)-diazonamide A. The ftnal selective chlorination was achieved in two steps, employing Br as a transient protecting group for the C5 position of the activated indoline ring.
(—)-Chimonanthine is a dimer of HPI units connected by a C3–C30 bond, which has long been known as a synthetic target.37,38 In 2013, Xie, Lai, Ma and coworkers reported the concise enantioselective total synthesis of (—)-chimonanthine based on the asymmetric dearomative bromocyclization of tryptamine 35 (Scheme 5).39 By utilizing DABCO-derived bromine salt 36 as the electrophilic bromination reagent and chiral phosphoric acid 8H-TRIP (R)-C3 as the catalyst, the C3- brominated HPI compound 37 was delivered on the gram- scale in excellent yield (96%) with high enantioselectivity (95% ee).40 Next, the Co-catalyzed homodimerization of 37 afforded 38 in 40% yield. The latter possesses the entire skel- eton of (—)-chimonanthine with correct stereochemistry.41,42 Simple removal of the protecting groups from 38 led to the target molecule. Subsequently, Lai, Xie and coworkers realized the asymmetric total synthesis of (—)-conolutinine, a terpenoid indole alkaloid isolated from Malaysian Tabernaemontana43 with a similar asymmetric dearomative bromocyclization reac- tion (39 / 40) as the key step (Scheme 5).44 Hydrolysis of the bromide and removal of the N-CO2Me protecting group of 40 provided 41. The enantiopurity of 41 was improved to 97.5% ee aer recrystallization. The desired eight-membered ring was constructed via double SN2 reactions of 41 with 1-bromo-2- (bromomethyl)but-2-ene. The ftnal ring in (—)-conolutinine was closed aer Co-mediated hydration of the exocyclic double bond followed by reduction of the amide by DIBAL-H.
3a-Amino-pyrroloindoline is another type of important core structure widely found in indole alkaloids and pharmacological compounds. In 2015, Deng, Liao and coworkers developed the asymmetric synthesis of 3a-amino-pyrroloindolines from the corresponding tryptamine derivatives and aryl diazonium salts catalyzed by a chiral phosphoric anion-paired catalyst (Scheme 6).45 With (S)-C2 as the chiral catalyst and Na2CO3 and 2,6-di-tBu-pyridine as the bases, the reaction between 43 and 44a produced C3-diazenated pyrroloindoline 45 in high enan- tiopurity (>96% ee).46 The N]N double bond could be cleaved with hydrazine hydrate in the presence of RANEY® Ni to afford the 3a-amino-pyrroloindoline compound 46, which underwent Buchwald–Hartwig amination with 1,2-dibromobenzene in the presence of a Pd-catalyst derived from Xantphos, leading to 47. Next, Pd-catalyzed cyclization of 47 with alkyne 48 afforded tryptamine dimer 49, from which routine protecting group manipulations delivered (—)-psychotriasine.47 Under slightly modifted conditions, the cyclic dipeptide 50 could couple with the phenyl diazonium salt 44b to give the dearomatized product 51 in 79% yield. Subsequently, the diazo group was reduced to primary amine 52 by hydrazine hydrate and Pd/C. Aer a series of screenings, the ortho-brominated hypervalent iodine reagent 53 was identifted as a suitable coupling partner for the Cu- catalyzed C–N bond formation, leading to 54 in reasonable yield (45%). The total synthesis of (+)-pestalazine B48,49 was completed aer ftnal cyclization of 54 with alkyne 55 (Scheme 6). Shortly aer, Tang, You and coworkers reported the Cu- catalyzed asymmetric dearomative amination reaction of trypt- amines (Scheme 6).50 In the presence of a Cu-catalyst ligated by the novel chiral BOX ligand L3, 3a-amino-pyrroloindoline 57 could be directly obtained in a reasonable yield (65%) and good enantioselectivity (90% ee) from tryptamine derivative 9 with O- (2,4-dinitrophenyl)hydroxylamine (DPH) 56 as the amination reagent. In this regard, (—)-psychotriasine was readily prepared from 57 through a known C–N bond formation/Larock cyclization/reduction sequence.
