Catalyzed Glycosylation with Glycosyl oâAlkynylbenzoates as Donors

Article Cite This: Acc. Chem. Res. XXXX, XXX, XXX−XXX

pubs.acs.org/accounts

Gold(I)-Catalyzed Glycosylation with Glycosyl o‑Alkynylbenzoates as Donors Biao Yu* State Key Laboratory of Bioorganic and Natural Products Chemistry, Center for Excellence in Molecular Synthesis, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, and University of Chinese Academy of Sciences, Shanghai 200032, China CONSPECTUS: Naturally occurring glycans and glycoconjugates have extremely diverse structures and biological functions. Syntheses of these molecules and their artificial mimics, which have attracted the interest of those developing new therapeutic agents, rely on glycosylation methodologies to construct the various glycosidic linkages. In this regard, a wide array of glycosylation methods have been developed, and they mainly involve the substitution of a leaving group on the anomeric carbon of a glycosyl donor with an acceptor (a nucleophile) under the action of a particular promoter (usually a stoichiometric electrophile). However, glycosylations involving inherently unstable or unreactive donors/acceptors are still problematic. In those systems, reactions involving nucleophilic, electrophilic, or acidic species present on the leaving group and the promoter could become competitive and detrimental to the glycosylation. To address this problem, we applied the recently developed chemistry of alkynophilic gold(I) catalysts to the development of new glycosylation reactions that would avoid the use of the conventional leaving groups and promoters. Gratifyingly, glycosyl o-alkynylbenzoates (namely, glycosyl o-hexynyl- and o-cyclopropylethynylbenzoates) turned out to be privileged donors under gold(I) catalysis with Ph3PAuNTf2 and Ph3PAuOTf. The merits of this new glycosylation protocol include the following: (1) the donors are easily prepared and are generally shelf-stable; (2) the promotion is catalytic; (3) the substrate scope is extremely wide; (4) relatively few side reactions are observed; (5) the glycosylation conditions are orthogonal to those of conventional methods; and (6) the method is operationally simple. Indeed, this method has been successfully applied in the synthesis of a wide variety of complex glycans and glycoconjugates, including complex glycosides of epoxides, nucleobases, flavonoids, lignans, steroids, triterpenes, and peptides. The direct glycosylation of some sensitive aglycones, such as dammarane C20-ol and sugar oximes, and the glycosylation-initiated polymerization of tetrahydrofuran were achieved for the first time. The gold(I) catalytic cycle of the present glycosylation protocol has been fully elucidated. In particular, key intermediates, such as the 1-glycosyloxyisochromenylium-4-gold(I) and isochromen-4-ylgold(I) complexes, have been unambiguously characterized. Exploiting the former glycosyloxypyrylium intermediate, SN2-type glycosylations were realized in specific cases, such as β-mannosylation/rhamnosylation. The protodeauration of the latter vinylgold(I) intermediate has been reported to be critically important for the gold(I) catalytic cycle. Thus, the addition of a strong acid as a cocatalyst can dramatically reduce the required loading of the gold(I) catalyst (down to 0.001 equiv). C-Glycosylation with silyl nucleophiles can proceed catalytically when moisture, which is sequestered by molecular sieves, can serve as the H+ donor for the required protodeauration step. Indeed, the unique mechanism explains the merits and broad applicability of the present glycosylation method and provides a foundation for future developments in glycosylation methodologies that mainly involve improving the diastereoselectivity and catalytic efficiency of glycosylations.

■

INTRODUCTION In hindsight, I feel extremely fortunate that I started learning about carbohydrate synthesis from the two seminal review articles that were written by Paulsen in 1982 and Schmidt in 1986.1,2 These reports discuss the overall state of the field at the time and details of the glycosylation reactions, and they have guided me throughout my journey toward the successful synthesis of a large variety of biologically significant glycoconjugates.3,4 Paulsen noted that “each oligosaccharide synthesis remains an independent problem” in that the diastereoselective formation of each glycosidic linkage is critically dependent on the coupling partners and especially the glycosyl donors, including their configuration and pattern of protecting groups. This was a major factor when he wrote the review since at the time most glycosidic bonds were prepared via Koenigs−Knorr reactions © XXXX American Chemical Society

(Figure 1). Schmidt, on the other hand, noted the limitations of this classical reaction, including the fact that most of the glycosyl bromides/chlorides are unstable and are thus difficult to prepare and use and that more than 1 equiv of a Hg2+/Ag+ salt is required for promotion, which makes the reaction conditions both harsh and toxic. In this regard, a number of promising new glycosylation protocols were introduced; one notable method is the Schmidt reaction, which uses easily prepared glycosyl trichloroacetimidates as donors (Figure 1). Schmidt glycosylation can often proceed under such mild conditions (e.g., in the presence of a catalytic amount of TMSOTf or BF3·OEt2 at −78 °C) that side reactions (especially those caused by the promoter) can be Received: November 14, 2017

A

DOI: 10.1021/acs.accounts.7b00573 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

Scheme 1. Attempts at the Glycosylation of Dammarane Derivative 1

Figure 1. A typical glycosylation reaction (L = leaving group, P = protecting groups, E+X− = promoter; electrophiles are highlighted in pink and nucleophiles are highlighted in blue).

largely avoided and the diastereoselectivity can be finely tuned.5 Indeed, this method has since enabled the synthesis of virtually any type of glycosidic linkage, and it has set a formidably high standard for newly developed glycosylation methodologies. The Schmidt reaction is catalytic because the promoter (e.g., E+X− = TMSOTf) can be regenerated after glycoside formation [TMSNH(CO)CCl3 (E+L−) + HOTf (H+X−) ⇄ TMSOTf (E+X−) + NH2(CO)CCl3 (HL)] and the leaving group, NH2(CO)CCl3 (HL), remains inert (Figure 1). Other glycosylation reactions require stoichiometric amounts of the promoter (usually a strong electrophile) to sequester the nucleophilic leaving species (e.g., Br−/Cl− in the Koenigs−Knorr reaction), and they require an additive (usually a hindered base) to intercept the in situ-generated protonic acid H+X− when the acid would be detrimental to the reactants or products. On the basis of this foundation, for over 20 years my lab has generally been able to select the appropriate glycosylation methods and conditions for the construction of various glycosidic linkages and, more importantly, to understand failed glycosylation experiments.4,6 In fact, glycosylation with unusual acceptors has continuously pushed the limits of the well-established glycosylation methods:7 (1) When the acceptor is poorly nucleophilic or unreactive, other nucleophilic species, including those present on the donor (such as protecting groups and leaving groups), the acceptor (functional groups), and the promoter, can react with the oxocarbenium intermediate. In the case of the Schmidt reaction, reaction with the leaving group, NH2(C O)CCl3, can become competitive. (2) When the acceptor is vulnerable to electrophilic reactions and a stoichiometric amount of the promotor is used, the promoter can interfere in the glycosylation. (3) When the acceptor or nascent glycoside is extremely acidsensitive, even the catalytic amount of the Lewis acid present in a Schmidt reaction will be detrimental. As an example,8 dammarane acceptor 1, which contains a γ-hydroxyolefin motif, could easily undergo cyclization to form the corresponding tetrahydrofuran derivative in the presence of an electrophile (Scheme 1). Therefore, glycosylation methods using stoichiometric promoters (such as the Koenigs−Knorr method and the thioglycoside method, which use halophilic and thiophilic reagents, respectively, as promoters) are likely to be unsuccessful. Indeed, even with glucosyl sulfoxide 2 as the donor and Tf2O (1.1 equiv) as the promoter in the presence of a hindered base (DTBMP = 2,6-di-tert-butyl-4-methylpyridine), furan derivative 5 was prepared in 78% yield via activation of the

olefin presumably by the in situ-generated TolSOTf.9 A Schmidttype glycosylation also failed to provide the desired glycoside; the hindered C20-OH moiety remained intact when treated with glucosyl trifluoroacetimidate 3 and TMSOTf (0.1 equiv) at −78 °C, but it underwent dehydration to provide olefin 6 (74%) when the temperature was increased to ∼18 °C. This challenging glycosylation step was finally achieved with o-alkynylbenzoate donor 4 under Ph3PAuNTf2 catalysis (0.1 equiv), which furnished dammarane 20-O-glycoside 7 in excellent yield (80%).8,10 Indeed, the gold(I)-catalyzed glycosylation method has essentially fulfilled my long-standing dream of a new glycosylation reaction: it uses stable donors and a catalytic amount of promoter and proceeds under neutral conditions.10 Thus, the scope of this glycosylation method can be expanded. Herein, I present a sequential account of the evolution, mechanistic studies, application, and future perspective of this new glycosylation methodology.

■

EVOLUTION OF THE METHOD As of 2005, “gold catalysis has become a hot spot in organometallic chemistry”,11 and impressively, an easily prepared, airstable gold(I) catalyst (i.e., Ph3PAuNTf2) was reported.12 As a nonspecialist, the most obvious benefit of this gold catalysis was the “alkynophilicity” of the gold(I) species (LAu+), which could lead to the selective activation of a C−C triple bond toward nucleophilic addition.13 In a simple analogy to the classic Fraser− Reid glycosylation with pent-4-enyl glycosides as donors,14 a gold(I) catalyst may be able to glycosylate pent-4-yne glycosides and similar compounds (Figure 2). Importantly, the stoichiometric amount of the highly electrophilic halonium promoter required in the former method is replaced by a catalytic amount of the π-acidic gold(I) species in this method. Recently, two inspiring communications were published. The group of Imagawa and Nishizawa reported a Hg(OTf)2-catalyzed glycosylation method using alkynoates as the leaving group (A, Figure 3).15 Hg2+ is isolobal to LAu+,13 but it is highly toxic and electrophilic. Hotha and Kashyap16 reported a gold(III)-catalyzed glycosylation protocol with propargyl glycosides (B) as donors. However, this reaction was limited to only the coupling of active donors and acceptors even under relatively harsh conditions (CH3CN, 60 °C). B


Article


with alcohols or aromatic compounds under Ph3PAuCl/AgOTf catalysis to provide the corresponding ethers or Friedel−Crafts products. Thus, Yao Li was very disappointed that the attempted glycosylation reaction with peracetylglucopyranosyl o-hexynylbenzoate as the donor in the presence of Ph3PAuOTf or Ph3PAuNTf2 did not proceed well but resulted in complex mixtures. However, the second donor he tested, perbenzoylglucopyranosyl o-hexynylbenzoate, underwent glycosylation perfectly. In fact, the coupling of this donor with a wide variety of acceptors consistently led to the desired glycosides in nearly quantitative yields at rt.10,17 The peracetyl glucopyranosyl donor was one of the few exceptions in the exploration of this method.20 Meanwhile, Yao Li developed an simple procedure for the large-scale preparation of o-alkynylbenzoic acids that involved Sonogashira coupling of inexpensive methyl 2-iodobenzoate with 1-alkynes followed by saponification.17,21 Since 1-hexyne and cyclopropylethyne were the least expensive terminal alkynes, they became the preferred compounds. The condensation of various 1-OH sugar derivatives with o-hexynyl- or o-cyclopropylethynylbenzoic acid could be achieved under various esterification conditions to give the donors in satisfactory yields and various α/β ratios.17 Notably, modifications of the o-alkynylbenzoate leaving group, such as replacing the alkane residue (R in E) with a phenyl residue or installing a substituent on the benzoate ring (R′ in E), rarely improved the glycosylation reactions22 because the effect of the steric and electronic variations in the leaving group could easily be mitigated by variations in the reaction conditions (solvent, temperature, etc.). Potential donors bearing other types of alkyne aglycones were also examined in the gold(I)-catalyzed glycosylation. In fact, donors with alkyne-containing carbonate (I)17 and phenyl (J)23 leaving groups, as well as the alkyne glycosides B and F24−H and allenoate (K),25 remained largely intact in the presence of Ph3PAuOTf/Ph3PAuNTf2. o-(Methyltosylaminoethynyl)benzyl

Figure 2. (left) Fraser−Reid glycosylation protocol and (right) the proposed gold(I)-catalyzed glycosylation reaction (the counteranions are not shown).

Yao Li examined the proposed gold(I)-catalyzed glycosylation reaction with pent-4-yne and hex-5-yne glycosides (G and H); however, no glycosylated products were observed in the presence of Ph3PAuOTf or Ph3PAuNTf2 under various conditions.17 The hydrated terminal alkynes were observed, which indicated that the C−C triple bond was being activated, but the activation was not strong enough to allow nucleophilic attack from the exoanomeric oxygen. We reasoned that improving the stability of the leaving group (i.e., the vinylgold(I) species), such as by the introduction of aromaticity, could enhance its leaving group ability. Accordingly, inspired by the disclosure of a facile transformation of o-alkynylbenzoyl acetal C into the corresponding isochromene-1-one derivative under PtCl2 catalysis by Fürstner and Davies,18 we first turned our attention to o-alkynylbenzoates. Around this time, Asao et al.19 reported o-alkynylbenzoic acid alkyl esters (e.g., D) as effective alkylation agents that could react

Figure 3. Relevant precedents (A−D), optimal donors (E), failed donors (F−K), less successful donors (L and M), and representative developments from other research groups (N and O). C


Article


Figure 4. Mechanism of a typical gold(I)-catalyzed glycosylation reaction with a glycosyl o-alkynylbenzoate as the donor.

Scheme 2. First Reaction Leading to the Isolation of Isochromen-4-ylgold(I) Complex 8

It is possible that the imine residue intercepted the H+ generated in situ (and was then hydrolyzed), which is required for protodeauration of isochromen-4-ylgold(I) complex 8 to regenerate the active Ph3PAu+ species. Indeed, he managed to isolate vinylgold(I) complex 8 in high yield (91% based on the starting Ph3PAuOTf), and he found that 8 is stable in the presence of alcohols and is incapable of promoting the glycosylation reaction. Therefore, with the addition of a strong acid (e.g., HOTf) as a cocatalyst, the loading of the gold(I) catalyst could be significantly reduced (down to 0.001 equiv) in gold(I)catalyzed glycosylation reactions.29 Considering the critical role of the protodeauration step in the gold(I)-catalyzed glycosylation, we wondered whether the C-glycosylation with allyltrimethylsilane or silyl enol ethers as acceptors would be catalytic given that no H+ would be generated during the C-glycosylation. Interestingly, Xiaoping Chen found that the C-glycosylation under normal conditions could proceed to completion in the presence of a catalytic amount of Ph3PAuNTf2, with isocoumarin 10 being recovered as the side product (Scheme 3).30 After careful analysis of a series of control experiments, he ultimately confirmed that the H+ was supplied by the moisture sequestered in the molecular sieves (MS). Thus, when he dried the CH2Cl2 solutions of the donor and acceptor and the solution of Ph3PAuNTf2 (10 mol %) with 4 Å MS and took only the dry solutions for the reaction, the C-glycosylation stopped at ∼13% conversion. Early in the method development process, Yao Li observed that when an excess of a pure β anomer was used as the donor, the donor present after workup of the gold(I)-catalyzed glycosylation reaction was predominantly the α derivative. This anomerization of the glycosyl o-alkynylbenzoate was surprising because both the endo- and exocyclic cleavage pathways observed in conventional

glycosides (L) were glycosylated more effectively in the presence of TMSOTf than in the presence of a gold(I) complex.26 o-Alkynylphenyl thioglycosides (M) underwent gold(I)-catalyzed glycosylation with a narrow substrate scope;27 a broader scope was found with gem-dimethyl S-but-3-ynyl thioglycosides (N) as donors.24 Recently, Hotha and co-workers disclosed that ethynylcyclohexyl carbonates (O) were well tolerated as donors in the gold(I)-catalyzed glycosylation.28

■

INSIGHTS INTO THE MECHANISM The mechanism of the gold(I)-catalyzed glycosylation was proposed prior to the methodological development (Figure 2). Nevertheless, it took years to collect sufficient experimental evidence to determine the details of the mechanism. The key findings are summarized in Figure 4 and include the characterizations of (i) the isochromen-4-ylgold(I) intermediate 829 and (ii) its protodeaurated form, which was prepared in a key step in the catalytic cycle (8 → 10),29,30 (iii) the 1-glycosyloxyisochromenylium-4-gold(I) intermediate II,31,32 (iv) the reversibility of the conversion from the donor to the glycosyl oxocarbenium intermediate (E + LAu+ ⇄ I ⇄ II ⇄ III + 8),31 and (v) the isochromen-4-yl gem-gold(I) complex 9 as an off-cycle gold(I) resting intermediate.31 When performing a routine glycosylation experiment with 2-p-methoxybenzylideneamino-β-D-glucopyranosyl o-hexynylbenzoate (11) as the donor and n-pentenol as the acceptor, Yugen Zhu was surprised to find that the reaction could reach completion only when the amount of Ph3PAuOTf was increased to ∼0.5 equiv, and the resulting products were β-glycoside 12 (37%) and, unexpectedly, α-glycoside 13 (47%), in which the N-p-methoxybenzylidene group was cleaved (Scheme 2).29 D


Article


than −OTf), the desired glycosyloxypyrylium complex 17 was cleanly formed and was fully characterized by 2D NMR spectroscopy at −35 °C (Scheme 5). Similar intermediates were not detected when Ph3PAuNTf2 was used, indicating the important role of the counteranion in the glycosylation reactions.35 In fact, highly β-selective mannosylation could be achieved with α-D-mannopyranosyl o-alkynylbenzoates under Ph3PAuBAr4F catalysis.32

Scheme 3. C-Glycosylation with Silyl C-Nucleophiles, in Which the Moisture Sequestered by the Molecular Sieves (MS) Plays a Key Role in the Catalytic Cycle

■

BROAD APPLICATIONS The present gold(I)-catalyzed glycosylation method has sufficiently broad applicability.17 This is evidenced by the successful synthesis of naturally occurring glycans and glycoconjugates that contain a wide variety of glycosidic linkages. Representative examples from our research group are as listed in Figure 5,36−48 while those from other groups are presented in Figure 6.49−54 Of the many proven advantages of this method, its application in the synthesis of complex glycoconjugates bearing vulnerable scaffolds or functional groups that are not stable under other glycosylation conditions is the most significant (Schemes 1 and 6).

anomerizations33 of glycosides seemed unlikely under gold(I) catalysis; the former involves the activation of the ring oxygen (instead of the alkyne) by the gold(I) cation, and the latter involves nucleophilic addition of the isochromen-4-ylgold(I) species (i.e., III + 8 → II) and elimination of the vinylgold(I) complex to give the alkyne (II → I). Jiakun Li and Yu Tang carefully examined the gold(I)-catalyzed anomerization of glycosyl o-alkynylbenzoates in the presence of a stoichiometric, exogenous vinylgold(I) complex (Scheme 4), and in all cases they isolated the crossover glycosyl o-alkynylbenzoates (e.g., 15).31 These results supported the exocyclic cleavage pathway, which allowed them to prove the occurrence of the relevant 1-glycosyloxyisochromenylium-4-gold(I) intermediate (II). Although efforts to isolate this glycosyloxypyrylium complex were unsuccessful, it was successfully trapped through a cycloaddition reaction with vinyl ether by employing carefully selected substrates and reaction conditions.31 On the other hand, Yu Tang observed that the gold(I)catalyzed anomerization became much slower in the presence of the exogenous vinylgold(I) complex. By monitoring the reactions using NMR spectroscopy, he detected the presence of a new gold(I) species, which was then prepared from vinylgold(I) complex 8 and unambiguously determined to be the isochromen4-yl gem-gold(I) complex 9.31 Indeed, this gem-gold(I) species, which did not show any catalytic activity, can also be detected in the normal gold(I)-catalyzed glycosylation reactions. Continuing their efforts to characterize the glycosyloxypyrylium intermediate (II) under normal glycosylation conditions, Yugen Zhu tried the treatment of 4,6-O-benzylidene-2,3-di-Obenzyl-α-D-mannopyranosyl o-hexynylbenzoate (16α) with gold(I) complexes bearing a variety of counterions to prepare a presumably stable complex.32 The corresponding mannosyl α-triflate, a key intermediate in the Crich-type β-mannosylation reaction, is known to be relatively stable because of the steric and electronic properties of the sugar residue.34 Indeed, mannosyl α-triflate was detected by NMR spectroscopy upon treatment of 16α (or its β isomer) with Ph3PAuOTf; however, upon treatment with Ph3PAuBAr4F (BAr4F = tetrakis[3,5-bis(trifluoromethyl)phenyl]borate, which is known to be a significantly poorer nucleophile

Glycosylation with Vulnerable Acceptors

For the first time, the C20-OH moiety of dammarane was glycosylated (Scheme 1), which enabled us to synthesize a number of complex ginsenosides (e.g., 20) with diverse biological activities.8,55 Direct glycosylations of other vulnerable acceptors, including sugar oximes,56 flavonoids,57 lignans,51 and complex epoxides,49,58 were also achieved. Another example of a glycosylation using this method is shown in Scheme 6. The glycosylation of a complex nucleoside derivative (i.e., 36), which contains an enol ether motif, was achieved using o-alkynylbenzoate 37 as the donor along with Ph3PAuOTf. The reaction provided the desired α-rhamnoside 38, which was then converted into the target antibiotic, A201A (25).43 A stoichiometric amount of Ph3PAuOTf was used because of the reaction of the in situ-generated H+, which is required for the regeneration of the gold(I) catalyst, with the basic nitrogen atoms in 36. In contrast, attempts at the glycosylation of 36 with the relevant imidate donors were not successful because the enol ether decomposed in the presence of either TMSOTf or BF3·OEt2. Glycosylation with Nucleobases

The present method has been reliably used in the glycosylation of nucleobases.43,59,60 Most notably, Boc-protected purine derivatives, which are unstable under the classical Vorbrüggen reaction conditions, could be effectively glycosylated with various glycosyl o-alkynylbenzoates under Ph3PAuNTf2 catalysis at room temperature. In the total synthesis of tunicamycins, a late-stage glycosylation of uracil (or a purine) using a complex trisaccharide o-alkynylbenzoate (i.e., 39) successfully provided 40 (or its purine analogue) in satisfactory yield (Scheme 7).44,61

Scheme 4. Representative Crossover Experiment To Prove the Exocyclic Cleavage Mechanism of the Gold(I)-Catalyzed Anomerization of Glycosyl o-Alkynylbenzoates and the Occurrence of the Isochromen-4-yl gem-Gold(I) Complex 9

E


Article


Scheme 5. Characterization of a Glycosyloxypyrylium Intermediate (i.e., 17) in the β-Selective Mannosylation Reaction (Numbers in Blue Are the Chemical Shifts (in ppm) of the Carbon Atoms)

Figure 5. Representative glycans and glycoconjugates synthesized by Yu and co-workers. The bonds highlighted in red were made by the present glycosylation method.

Glycosylation with Poor Nucleophiles and the Glycosylation-Initiated Polymerization of THF

(i.e., 3,4,6-tri-O-acetyl-2-azido-2-deoxy- D -glucopyranosyl o-hexynylbenzoate) was mixed with PPh3AuOTf (0.3 equiv) in tetrahydrofuran (THF) at room temperature with stirring, Yao Li observed that prior to the addition of an acceptor, the initially clear solution became viscous and gelled substantially after standing overnight. Workup of the reaction afforded a white, plastic solid, which was determined to be a glycosyl polytetrahydrofuran.62 This highly efficient

With no competitive nucleophilic species that could be derived from the conventional leaving group of the donors and the promoters being present, the developed gold(I)catalyzed method is the most efficient reaction for the glycosylation of poorly nucleophilic (or unreactive) acceptors.8,50,57 As shown in Scheme 8, when a donor F


Article


Figure 6. Representative glycoconjugates synthesized by other research groups. The bonds highlighted in red were made by the present glycosylation method.

Scheme 6. Glycosylation of an Acid-Labile Enol Ether Acceptor (i.e., 36) en Route toward the Synthesis of A201A (25)

similar results. Removal of the acetyl group of 43β led to gordonoside F (23), which was found to be responsible for the weight-reducing effect of the popular Hoodia plants.41 The glycosylation of steroid disaccharide 44, which contains a unique seven-membered formyl acetal-bridged orthoester (FABO) motif, with deoxytetrasaccharide o-alkynylbenzoate 45 under Ph3PAuNTf2 catalysis led to the coupled hexasaccharide, albeit with the α anomer as the major product (α/β ≈ 3:1).42 With Ph3PAuOTf as the catalyst, the glycosylation resulted in a β product, likely involving a glycosyl α-triflate intermediate with concomitant cleavage of the FABO linkage. To intercept the incipient HOTf, a hindered base (TTBP = 2,4,6-tri-tert-butylpyrimidine) was introduced. The sequestration of HOTf also retarded the gold(I) catalytic cycle, and thus, 0.8 equiv of the gold(I) complex was required to drive the reaction to completion. Finally, coupled hexasaccharide 46 was obtained in a satisfactory yield (64% yield, 87% b.r.s.m.) with a β/α ratio of 2:1. Removal of the

glycosylation-initiated cationic ring-opening polymerization of THF was unprecedented and highlights the unique properties of the gold(I)-catalyzed glycosylation reaction. Glycosylation with Deoxy Sugars

While the imidate donors of deoxy sugars, especially those of the di- and trideoxy sugars, are usually unstable and difficult to purify, the corresponding o-alkynylbenzoates are stable and convenient to prepare.63−65 In addition, the mild conditions of the gold(I) catalysis can prevent the anomerization and hydrolysis of the resulting glycosidic linkages of the deoxy sugars. Thus, glycosylation of hoodigogenin A (41) with deoxytetrasaccharide o-alkynylbenzoate 42 under the normal gold(I)catalyzed conditions provided coupled steroid tetrasaccharide 43 (α/β = 1:1) in nearly quantitative yield (Scheme 9).41 Even though the β selectivity has not yet been optimized, the thermodynamically favored α isomer is typically the predominant product with other types of donor. Remarkably, a gram-scale glycosylation gave G


Article


in the presence of nucleophilic, electrophilic, or acidic species, and on the basis of our mechanistic understanding of alkynophilic gold(I) catalysis, we have developed a new glycosylation method that uses glycosyl o-alkynylbenzoates as donors and a gold(I) complex as the catalyst. This method has been successfully applied in the synthesis of virtually all types of glycosidic linkages. The merits of this method include the following: • The donors, including those of the di- and trideoxy sugars, are easily prepared and shelf-stable. • Promotion of the reaction requires a catalytic amount of a gold(I) complex (the most convenient of which are Ph3PAuOTf and Ph3PAuNTf2), which possesses little oxophilic character or Lewis acidity. • The substrate scope is extremely wide, and side reactions are minimal. This phenomenon is due to the fact that nucleophilic, electrophilic, and acidic species cannot be derived from the leaving group and the promoter as they can in other glycosylation methods, which makes these conditions more mild. • The glycosylation conditions are orthogonal to conventional glycosylation conditions; thus, new, one-pot, sequential glycosylation methods can be developed in combination with other glycosylation reactions.10,17 • The method is operationally quite convenient in that the reactions are often performed at room temperature and the reactants do not need to be quenched during workup. The mechanism of glycosyl o-alkynylbenzoate activation under gold(I) catalysis has been fully elucidated, and the key intermediates (i.e., isochromen-4-ylgold(I) complex 8 and glycosyloxypyrylium intermediate II) have been characterized. This unique mechanism66 explains the advantages of the present glycosylation reaction and provides a foundation for the development of further glycosylation methodologies. In the interim, two problems have yet to be solved: • The diastereoselectivity is still largely dependent on the steric and electronic properties of the coupling partners.67 Adjusting the steric and electronic properties of the leaving group and the gold(I) catalyst led to satisfactory stereoselectivity in only a limited number of cases (i.e., the β-mannosylation/rhamnosylation and the glycosylation

Scheme 7. Late-Stage Glycosylation of a Nucleobase with a Complex o-Alkynylbenzoate Donor (i.e., 39) en Route toward the Synthesis of Tunicamycins (e.g., 26)

Scheme 8. Glycosylation-Initiated Polymerization of THF Due to the Lack of Competitive Nucleophiles in the Present Gold(I)-Catalyzed Glycosylation Reaction

terminal chloroacetyl (CA) group and tert-butyldimethylsilyl (TBS) group in 46β furnished periploside A (24), which is a potent immunosuppressive compound isolated from the Chinese medicinal plant Periploca sepium.

■

SUMMARY AND PERSPECTIVE With the aim of synthesizing complex glycosides that involve glycosidic couplings of donors and acceptors that are not stable

Scheme 9. Glycosylation with Deoxy Sugar o-Alkynylbenzoate Donors (42 and 45) en Route toward the Synthesis of Gordonoside F (23) and Periploside A (24)

H


Article


(8) Yu, J.; Sun, J.; Niu, Y.; Li, R.; Liao, J.; Zhang, F.; Yu, B. Synthetic access toward the diverse ginsenosides. Chem. Sci. 2013, 4, 3899−3905. (9) Crich, D.; Sun, S. Are glycosyl triflates intermediates in the sulfoxide glycosylation method? A chemical and 1H, 13C, and 19F NMR spectroscopic investigation. J. Am. Chem. Soc. 1997, 119, 11217−11223. (10) Li, Y.; Yang, Y.; Yu, B. An efficient glycosylation protocol with glycosyl ortho-alkynylbenzoates as donors under the catalysis of Ph3PAuOTf. Tetrahedron Lett. 2008, 49, 3608. (11) Hashmi, A. S. K. The catalysis gold rush: new claims. Angew. Chem., Int. Ed. 2005, 44, 6990−6993. (12) Mézailles, N.; Ricard, L.; Gagosz, F. Phosphine gold(I) bis(trifluoromethanesulfonyl)imidate complexes as new highly efficient and air-stable catalysts for the cycloisomerization of enynes. Org. Lett. 2005, 7, 4133−4136. (13) Fürstner, A.; Davies, P. W. Catalytic carbophilic activation: catalysis by platinum and gold π acids. Angew. Chem., Int. Ed. 2007, 46, 3410−3449. (14) Fraser-Reid, B.; Konradsson, P.; Mootoo, D. R.; Udodong, U. Direct elaboration of pent-4-enyl glycosides into disaccharides. J. Chem. Soc., Chem. Commun. 1988, 823−825. (15) Imagawa, H.; Kinoshita, A.; Fukuyama, T.; Yamamoto, H.; Nishizawa, M. Hg(OTf)2-catalyzed glycosylation using alkynoate as the leaving group. Tetrahedron Lett. 2006, 47, 4729−4731. (16) Hotha, S.; Kashyap, S. Propargyl glycosides as stable glycosyl donors: anomeric activation and glycoside syntheses. J. Am. Chem. Soc. 2006, 128, 9620−9621. (17) Li, Y.; Yang, X.; Liu, Y.; Zhu, C.; Yang, Y.; Yu, B. Gold(I)catalyzed glycosylation with glycosyl ortho-alkynylbenzoates as donors, general scope and application to the synthesis of a cyclic triterpene saponin. Chem. - Eur. J. 2010, 16, 1871−1882. (18) Fürstner, A.; Davies, P. W. Heterocycles by PtCl2-catalyzed intramolecular carboalkoxylation or carboamination of alkynes. J. Am. Chem. Soc. 2005, 127, 15024−15025. (19) Asao, N.; Aikawa, H.; Tago, S.; Umetsu, K. Gold-catalyzed etherification and Friedel-Crafts alkylation using ortho-alkynylbenzoic acid alkyl ester as an efficient alkylating agent. Org. Lett. 2007, 9, 4299− 4302. (20) Ma, Y.; Lian, G.; Li, Y.; Yu, B. Identification of 3,6-di-O-acetyl1,2,4-O-orthoacetyl-α-D-glucopyranose as a direct evidence for the 4-Oacyl group participation in glycosylation. Chem. Commun. 2011, 47, 7515−7517. (21) Yu, B.; Cao, X. 2-(1-Hexyn-1-yl)benzoic acid. In Electronic Encyclopedia of Reagents for Organic Synthesis (e-EROS); John Wiley & Sons: Chichester, U.K., 2014; RN01722. (22) Zhu, Y.; Shen, Z.; Li, W.; Yu, B. Stereoselective synthesis of βrhamnopyranosides via gold(I)-catalyzed glycosylation with 2-alkynyl4-nitro-benzoate donors. Org. Biomol. Chem. 2016, 14, 1536−1539. (23) Li, Y. Ph.D. Thesis, University of Chinese Academy of Sciences, Beijing, 2010. (24) Adhikari, S.; Li, X.; Zhu, J. Studies of S-but-3-ynyl and gemdimethyl S-but-3-ynyl thioglycoside donors in gold-catalyzed glycosylations. J. Carbohydr. Chem. 2013, 32, 336−359. (25) Zhang, Y.; Wang, P.; Song, N.; Li, M. Bromodimethylsulfonium bromide/silver triflate-promoted glycosylations using glycosyl allenoates as donors. Carbohydr. Res. 2013, 381, 101−111. (26) Chen, X.; Shen, D.; Wang, Q.; Yang, Y.; Yu, B. ortho(Methyltosylaminoethynyl)benzyl glycosides as new glycosyl donors for latent-active glycosylation. Chem. Commun. 2015, 51, 13957−13960. (27) Yang, F.; Wang, Q.; Yu, B. ortho-Alkynylphenyl thioglycosides as a new type of glycosylation donors under the catalysis of Au(I) complexes. Tetrahedron Lett. 2012, 53, 5231−5234. (28) Mishra, B.; Neralkar, M.; Hotha, S. Stable alkynyl glycosyl carbonates: catalytic anomeric activation and synthesis of a tridecasaccharide reminiscent of mycobacterium tuberculosis cell wall lipoarabinomannan. Angew. Chem., Int. Ed. 2016, 55, 7786−7791. (29) Zhu, Y.; Yu, B. Characterization of the isochromen-4-yl-gold(I) intermediate in the gold(I)-catalyzed glycosidation of glycosyl oalkynylbenzoates and enhancement of the catalytic efficiency thereof. Angew. Chem., Int. Ed. 2011, 50, 8329−8332.

with a large excess of the reactive acceptors via SN2-type substitution of the glycosyloxypyrylium intermediates).22,31,32 • Effective glycosylation typically requires a 5−10 mol % loading of the gold(I) catalyst. Moreover, when coupling basic substrates (which can inhibit the gold(I) catalytic cycle)43 or unreactive substrates,68 as much as 1 equiv of the gold(I) catalyst is required for the reaction to be practical. The development of a more effective gold(I) catalyst or a cocatalyst that will facilitate the large-scale synthesis of glycans and glycoconjugates is still underway.68−70 With these successes and pending challenges, glycosylation methodologies are still being studied, although they are being approached from a new starting point.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Biao Yu: 0000-0002-3607-578X Notes

The author declares no competing financial interest. Biography Biao Yu received his B.Sc. in Radiochemistry from Peking University in 1989 and his Ph.D. from Shanghai Institute of Organic Chemistry (SIOC), Chinese Academy of Sciences (CAS), in 1995. After a one-year postdoctoral stay at New York University, he returned to SIOC as an assistant professor and became a professor in 1999. His laboratory is dedicated to the total synthesis, synthetic methodology, and chemical biology of glycans and glycoconjugates.

■

ACKNOWLEDGMENTS B.Y. is grateful to his co-workers whose names appear in the text and references for their invaluable contributions to this project and for the financial support from the National Natural Science Foundation of China (21432012 and 21621002), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB20020000), the K. C. Wong Education Foundation, and the E-Institutes of Shanghai Municipal Education Commission (E09013).

■

REFERENCES

(1) Paulsen, H. Advances in selective chemical syntheses of complex oligosaccharides. Angew. Chem., Int. Ed. Engl. 1982, 21, 155−173. (2) Schmidt, R. R. New methods for the synthesis of glycosides and oligosaccharides - Are there alternatives to the Koenigs-Knorr method? Angew. Chem., Int. Ed. Engl. 1986, 25, 212−235. (3) Yu, B.; Zhang, Y.; Tang, P. Carbohydrate chemistry in the total synthesis of saponins. Eur. J. Org. Chem. 2007, 5145−5161. (4) Yu, B.; Sun, J.; Yang, X. Assembly of naturally occurring glycosides, evolved tactics and glycosylation methods. Acc. Chem. Res. 2012, 45, 1227−1236. (5) Peng, P.; Schmidt, R. R. Acid−base catalysis in glycosidations: a nature derived alternative to the generally employed methodology. Acc. Chem. Res. 2017, 50, 1171−1183. (6) Yu, B.; Sun, J. Glycosylation with glycosyl N-phenyltrifluoroacetimidates (PTFAI) and a perspective of the future development of new glycosylation methods. Chem. Commun. 2010, 46, 4668−4679. (7) Yang, Y.; Zhang, X.; Yu, B. O-Glycosylation methods in the total synthesis of complex natural glycosides. Nat. Prod. Rep. 2015, 32, 1331− 1355. I


Article

Accounts of Chemical Research (30) Chen, X.; Wang, Q.; Yu, B. Gold(I)-catalyzed C-glycosylation of glycosyl ortho-alkynylbenzoates, a role of the moisture sequestered by molecular sieves. Chem. Commun. 2016, 52, 12183−12186. (31) Tang, Y.; Li, J.; Zhu, Y.; Li, Y.; Yu, B. Mechanistic insights into the gold(I)-catalyzed activation of glycosyl ortho-alkynylbenzoates for glycosidation. J. Am. Chem. Soc. 2013, 135, 18396−18405. (32) Zhu, Y.; Yu, B. Highly stereoselective β-mannopyranosylation via the 1-α-glycosyloxy-isochromenylium-4-gold(I) intermediates. Chem. Eur. J. 2015, 21, 8771−8780. (33) Capon, B. Mechanism in carbohydrate chemistry. Chem. Rev. 1969, 69, 407−498. (34) Aubry, S.; Sasaki, K.; Sharma, I.; Crich, D. Influence of protecting groups on the reactivity and selectivity of glycosylation: chemistry of the 4,6-O-benzylidene protected mannopyranosyl donors and related species. Top. Curr. Chem. 2011, 301, 141−188. (35) Bohe, L.; Crich, D. A propos of glycosyl cations and the mechanism of chemical glycosylation; the current state of the art. Carbohydr. Res. 2015, 403, 48−59. (36) Yang, Y.; Li, Y.; Yu, B. Total synthesis and structural revision of TMG-chitotriomycin, a specific inhibitor of insect and fungal β-Nacetylglucosaminidases. J. Am. Chem. Soc. 2009, 131, 12076−12077. (37) Shen, Z.; Mobarak, H.; Li, W.; Widmalm, G.; Yu, B. Synthesis of β(1→2)-linked 6-deoxy-L-altropyranose oligosaccharides via gold(I)catalyzed glycosylation of a ortho-hexynylbenzoate donor. J. Org. Chem. 2017, 82, 3062−3071. (38) Li, Y.; Sun, J.; Yu, B. Efficient synthesis of lupane-type saponins via gold(I)-catalyzed glycosylation with glycosyl ortho-alkynylbenzoates as donors. Org. Lett. 2011, 13, 5508−5511. (39) Li, Y.; Yang, W.; Ma, Y.; Sun, J.; Shan, L.; Zhang, W.-D.; Yu, B. Synthesis of kaempferol 3-O-(2″,3″- and 2″,4″-di-O-E-p-coumaroyl)-αL-rhamnopyranosides. Synlett 2011, 915−918. (40) Yang, W.; Sun, J.; Lu, W.; Li, Y.; Shan, L.; Han, W.; Zhang, W.-D.; Yu, B. Synthesis of kaempferol 3-O-(3″,6″-di-O-E-p-coumaroyl)-β-Dglucopyranoside, efficient glycosylation of flavonol 3-OH with glycosyl ortho-alkynylbenzoates as donors. J. Org. Chem. 2010, 75, 6879−6888. (41) Zhang, S.; Ma, Y.; Li, J.; Ma, J.; Yu, B.; Xie, X. Molecular matchmaking between the popular ‘weight loss’ herb Hoodia gordonii and GPR119, a potential drug target for metabolic disorder. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 14571−14576. (42) Zhang, X.; Zhou, Y.; Zuo, J.; Yu, B. Total synthesis of periploside A, a unique pregnane hexasaccharide with potent immunosuppressive effects. Nat. Commun. 2015, 6, 5879. (43) Nie, S.; Li, W.; Yu, B. Total synthesis of nucleoside antibiotic A201A. J. Am. Chem. Soc. 2014, 136, 4157−4160. (44) Li, J.; Yu, B. A modular approach to the total synthesis of tunicamycins. Angew. Chem., Int. Ed. 2015, 54, 6618−6621. (45) Zhu, D.; Yu, B. Total synthesis of linckosides A and B, the representative starfish polyhydroxysteroid glycosides with neuritogenic activities. J. Am. Chem. Soc. 2015, 137, 15098−15101. (46) Dai, Y.; Yu, B. Total synthesis of astrosterioside A, an antiinflammatory asterosaponin. Chem. Commun. 2015, 51, 13826−13829. (47) Xiao, G.; Yu, B. Total synthesis of starfish saponin goniopectenoside B. Chem. - Eur. J. 2013, 19, 7708−7712. (48) Chen, X.; Shao, X.; Li, W.; Zhang, X.; Yu, B. Total synthesis of echinoside A, a representative triterpene glycoside of sea cucumbers. Angew. Chem., Int. Ed. 2017, 56, 7648−7652. (49) Nicolaou, K. C.; Cai, Q.; Sun, H.; Qin, B.; Zhu, S. Total synthesis of trioxacarcins DC-45-A1, A, D, C, and C7″-epi-C and full structural assignment of trioxacarcin C. J. Am. Chem. Soc. 2016, 138, 3118−3124. (50) Wang, B.; Liu, Y.; Jiao, R.; Feng, Y.; Li, Q.; Chen, C.; Liu, L.; He, G.; Chen, G. Total synthesis of mannopeptimycins α and β. J. Am. Chem. Soc. 2016, 138, 3926−3932. (51) Liu, H.; Liao, J.-X.; Hu, Y.; Tu, Y.-H.; Sun, J.-S. A highly efficient approach to construct (epi)-podophyllotoxin-4-O-glycosidic linkages as well as its application in concise syntheses of etoposide and teniposide. Org. Lett. 2016, 18, 1294−1297. (52) Bai, Y.; Shen, X.; Li, Y.; Dai, M. Total synthesis of (−)-spinosyn A via carbonylative macrolactonization. J. Am. Chem. Soc. 2016, 138, 10838−10841.

(53) Sun, P.; Wang, P.; Zhang, Y.; Zhang, X.; Wang, C.; Liu, S.; Lu, J.; Li, M. Construction of β-mannosidic bonds via gold(I)-catalyzed glycosylations with mannopyranosyl ortho-hexynylbenzoates and its application in synthesis of acremomannolipin A. J. Org. Chem. 2015, 80, 4164−4175. (54) Wang, Q.; Kuramoto, Y.; Okazaki, Y.; Ota, E.; Morita, M.; Hirai, G.; Saito, K.; Sodeoka, M. Synthesis of polyunsaturated fatty acidcontaining glucuronosyldiacylglycerol through direct glycosylation. Tetrahedron Lett. 2017, 58, 2915−2918. (55) Shen, R.; Cao, X.; Laval, S.; Sun, J.; Yu, B. Synthesis of ocotilloltype ginsenosides. J. Org. Chem. 2016, 81, 10279−10294. (56) Yu, J.; Sun, J.; Yu, B. Construction of interglycosidic N−O linkage via direct glycosylation of sugar oximes. Org. Lett. 2012, 14, 4022−5. (57) Liao, J.-X.; Fan, N.-L.; Liu, H.; Tu, Y.-H.; Sun, J.-S. Highly efficient synthesis of flavonol 5-O-glycosides with glycosyl ortho-alkynylbenzoates as donors. Org. Biomol. Chem. 2016, 14, 1221−1225. (58) He, Q.-L.; Minn, I.; Wang, Q.; Xu, P.; Head, S. A.; Datan, E.; Yu, B.; Pomper, M. G.; Liu, J. O. Targeted delivery and sustained antitumor activity of triptolide through glucose conjugation. Angew. Chem., Int. Ed. 2016, 55, 12035−12039. (59) Zhang, Q.; Sun, J.; Zhu, Y.; Zhang, F.; Yu, B. An efficient approach to the synthesis of nucleosides, gold(I)-catalyzed N-glycosylation of pyrimidines and purines with glycosyl ortho-alkynylbenzoates. Angew. Chem., Int. Ed. 2011, 50, 4933−4936. (60) Yang, F.; Zhu, Y.; Yu, B. A dramatic concentration effect on the stereoselectivity of N-glycosylation for the synthesis of 2′-deoxy-βribonucleosides. Chem. Commun. 2012, 48, 7097−7099. (61) Price, N. P. J.; Hartman, T. M.; Li, J.; Velpula, K. K.; Naumann, T. A.; Guda, M. R.; Yu, B.; Bischoff, K. M. Modified tunicamycins with reduced eukaryotic toxicity that enhance the antibacterial activity of βlactams. J. Antibiot. 2017, 70, 1070−1077. (62) Li, Y.; Yu, B. Glycosylation initiated cationic ring-opening polymerization of tetrahydrofuran to prepare neo-glycopolymers. Chem. Commun. 2010, 46, 6060−6062. (63) Ma, Y.; Li, Z.; Shi, H.; Zhang, J.; Yu, B. Assembly of digitoxin by gold(I)-catalyzed glycosidation of glycosyl ortho-alkynylbenzoates. J. Org. Chem. 2011, 76, 9748−9756. (64) Zeng, J.; Sun, G.; Yao, W.; Zhu, Y.; Wang, R.; Cai, L.; Liu, K.; Zhang, Q.; Liu, X.-W.; Wan, Q. 3-Aminodeoxypyranoses in glycosylation: diversity-oriented synthesis and assembly in oligosaccharides. Angew. Chem., Int. Ed. 2017, 56, 5227−5231. (65) Zeng, J.; Sun, G.; Wang, R.; Zhang, S.; Teng, S.; Liao, Z.; Meng, L.; Wan, Q. Gold-catalyzed diversified synthesis of 3-aminosugar analogues of digitoxin and digoxin. Org. Chem. Front. 2017, 4, 2450−2454. (66) Ranade, S. C.; Demchenko, A. V. Mechanism of chemical glycosylation: focus on the mode of activation and departure of anomeric leaving groups. J. Carbohydr. Chem. 2013, 32, 1−43. (67) Nigudkar, S. S.; Demchenko, A. V. Stereocontrolled 1,2-cis glycosylation as the driving force of progress in synthetic carbohydrate chemistry. Chem. Sci. 2015, 6, 2687−2704. (68) Li, J.; Dai, Y.; Li, W.; Laval, S.; Xu, P.; Yu, B. Effective synthesis of α-D-GlcN-(1→4)-D-GlcA/L-IdoA glycosidic linkage under gold(I)catalysis. Asian J. Org. Chem. 2015, 4, 756−762. (69) Zhu, D.; Cao, X.; Yu, B. Au(I) π-bis(tert-butyldimethylsilyl)acetylene triphenylphosphine complex, an effective pre-catalyst for Au(I)-catalyzed reactions. Org. Chem. Front. 2015, 2, 360−365. (70) Zhu, Y.; Laval, S.; Tang, Y.; Lian, G.; Yu, B. A polystyrene-bound triphenylphosphine gold(I) catalyst for the glycosylation of glycosyl ortho-hexynylbenzoates. Asian J. Org. Chem. 2015, 4, 1034−1039.

J


Catalyzed Glycosylation with Glycosyl oâAlkynylbenzoates as Donors

Recommend Documents

Catalyzed Glycosylation with Glycosyl oâAlkynylbenzoates as Donors