HETEROCYCLES

Review | Regular issue | Vol. 85, No. 6, 2012, pp. 1351-1376
Received, 14th February, 2012, Accepted, 6th April, 2012, Published online, 13th April, 2012.
DOI: 10.3987/REV-12-734

■ Chiral Synthesis of Iminosugars

Hiroki Takahata*

Faculty of Pharmaceutical Sciences, Tohoku Pharmaceutical University, 4-4-1 Komatsushima, Aoba-ku, Sendai 981-8558, Japan

Abstract

An chiral synthesis of iminosugars such as fagomine, 1-deoxynojirimycine, and isofagomine together with their stereoisomers are described.

INTRODUCTION
Since the discovery of nojirimycin as the first natural glucose mimic, over 200 naturally occurring iminosugars1 in which the oxygen of the sugar ring is replaced by nitrogen have been reported. The transition states for enzymatic glycosidase and transferase enzymes have considerable oxocarbenium-ion characteristics, in that the anomeric carbon acquires sp2 hybridization and a partial positive charge develops at the anomeric carbon and the endocyclic oxygen. Therefore, iminosugars as transition-state analogues are of particular interest in terms of inhibitor design. Iminosugars belong to the polyhydroxylated alkaloid family and display a broad range of interesting biological activities that are potentially useful in the treatment of ailments as varied as viral infections,2 including human immunodeficiency virus (HIV),2a-e human hepatitis C (HCV),2f,g or dengue virus,2i cancer,3 diabetes,4 tuberculosis,5 and lysosomal storage diseases.6 The tremendous therapeutic potential of this class of compounds has been attributed to their ability to interact with carbohydrate-processing enzymes, where they act as competitive inhibitors of glycosidases and/or glycosyltransferases, a phenomenon that has stimulated much research in this area of glycobiology.7 Because of the therapeutic importance of these compounds, many synthetic efforts have been directed toward their preparation. In practice, N-butyl-1-deoxynojirimycin (DNJ) 1 (Zavesca™) is used in the treatment of Gaucher disease. Another iminosugar, Miglitol (Glycet™) 2, which is commercially available in the USA and Canada, is used for the treatment of type II diabetes. In addition, galacto-DNJ (Migalastat) 3 has been shown to inhibit lysosomal α-galactosidase and is currently in phase III clinical trials for the treatment of Fabry’s disease (Figure 1). There are a number of reports and reviews on in vitro and in vivo glycosidase inhibition by DNJ, manno-DNJ, galacto-DNJ, and their derivatives. On the other hand, the biological properties of enantiomers of DNJ, manno-DNJ, and galacto-DNJ, and other diastereomers of DNJ have not been systematically studied. Therefore, our attention was focused on the syntheses of both enatiomers of several iminosugars such as fagomine, 1-deoxynojirimycine, and isofagomine, together with their congeners.

SYNTHESIS OF IMINOSUGARS
1. Preparation of Fagomine8
Three fagomine isomers 4-6 from Xanthocercis zambesiaca, which occurs in dry southern African forests have been isolated (Figure 2).9 Among these, fagomine 4 and 3-epi-fagomine 5 were found to have some activity against mammalian gut α-glucosidase and β-galactosidase.9 Recently, 4 was reported to have a potent antihyperglycemic effect in streptozocin-induced diabetic mice and the potentiation for the glucose-induced secretion of insulin.10 In addition, it was reported that the fagomine isomer 7 (not naturally occurring) is an inhibitor of lysosomal α-galactosidase activity in Fabry lymphoblasts.11 However, most synthetic efforts have been directed towards the synthesis of fagomine derivatives that have a naturally occurring D configuration. Therefore, we embarked on short and efficient syntheses of D-fagomine and its epimers via the preparation of a new common chiral building block, hydroxymethylpiperidene 8, which appears to be an ideal precursor for the synthesis of dihydroxypiperidinols. A new synthesis of all fagomine isomers 4-7 via the preparation of 8 in a straightforward and stereoselective manner with Garner aldehyde 9 and catalytic ring-closing metathesis (RCM) of the diene system 12 for the construction of the piperidine ring was achieved.12

Our synthesis of 8 began with the Wittig reaction of the D-serine-derived Garner aldehyde 9.13 Treatment of 9 with methyltriphenylphosphonium iodide in the presence of sodium bis(trimethylsilyl)amide gave olefin 10 in 63% yield. Hydrolysis of 10 with p-toluenesulfonic acid in MeOH followed by O-silylation afforded 11 in 72% yield. N-Alkylation of 11, using a three-step sequence [(1) deprotection of the N-Boc group; (2) alkylation; and (3) N-protection] provided the butenylated product 12 in 60% overall yield. RCM of 12 with the Grubbs' 1st generation catalyst, (benzylidine)bis(tricyclohexylphosphine)- ruthenium(IV) dichloride, under the usual conditions gave the desired intermediate 8 in 97% yield (Scheme 1).

With the promising educt 8 in hand, our interest was then directed to the synthesis of all isomers 4-7 of fagomine. We first introduced an epoxy-functionality into the double bond to produce both 4 and 6, which contain trans diols at the 3 and 4 positions. The dioxirane, generated in situ from Oxone® with 1,1,1-trifluoroacetone according to a recent procedure,14 was reacted with 8 to give a mixture of stereoisomeric epoxy compounds 13 (60%) and 14 (30%) which were separated by medium-pressure chromatography in a high yield of 90% although the diastereoselectivity was low (dr:  33%). The syn epoxide 14 was stereoselectively produced by an alternative method. Desilylation of 8 with TBAF was carried out to afford 15 (97% yield). Starting with 15, hydroxy-directed epoxidation with m-CPBA followed by silylation afforded 14 in 64% overall yield. Acid hydrolysis of the epoxy ring of 13 was accomplished using a mixture of H2SO4/1,4-dioxane/H2O in a volume ratio of 0.2/3/2,15 and further treatment with an ion-exchange resin (Amberlite® IRA-410) gave fagomine 4 exclusievly and in 75% yield. Although the explanation for this high selectivity remains unclear, we consider the following explanation:  an attack of H2O, as a nucleophile with backside displacement of the leaving oxygen in the epoxy-substituted ring, occurs at the more remote site (4 position) because nucleophilic attack at the 3 position has a syn orientation with respect to the adjacent siloxymethyl substiuent at the 5 position. In addition, the opening of cyclohexene oxide structures generally proceeds in such a manner that the diaxial reaction product is usually produced. In contrast, the basic cleavage of epoxide 14 by treatment with a mixture of KOH/1,4-dioxane/H2O gave 6 preferentially, in a ratio of 5:1 (6 to 4) in 99% combined yield.

The stereoselective dihydroxylation of the double bond was next examined. Under modified Upjohn conditions,16 the treatment of 8 with a catalytic amount of K2OsO4·2H2O (5 mol%) and 4-methylmorpholine N-oxide (1.5 equiv) as a cooxidant gave the diol 16 as a single diastereoisomer in a high yield of 92%. Deprotection of 16 by treatment with 10% hydrochloric acid in 1,4-dioxane followed by treatment with a basic ion-exchange resin (Dowex® 1×2 OH- form) afforded 3-epi-fagomine 5 in 91% yield. Surprisingly, both the AD-mix-α®- and β®-mediated dihydroxylation of 8 provided 16 with no diastereomer being detected, in 94% and 96% yields, respectively. Dihydroxylation exclusively occurred from the less hindered anti side of the siloxymethyl substituent, which adopts an axial position due to 1,3-allylic strain. The dihydroxylation of 15 under the above modified Upjohn conditions also took place from the anti side of the hydroxymethyl group, followed by the deprotection to give 5 as a single diastereomer in 87% combined yield. Donohoe reported17 that osmium tetroxide produces a bidendate and reactive complex with TMEDA, which can be utilized in the directed dihydroxylation of homoallyic alcohols. Therefore, the oxidation of 15 with a combination of OsO4 with TMEDA followed by deprotection gave 7 and 5 in low selectivity (2:1) in 56% and 30% yields, respectively. Since all of the D isomers were obtained, the preparation of L-forms, ent-4-7, using the L-serine-derived Garner aldehyde ent-9 was accomplished using nearly the same procedure as that reported for the preparations described in Schemes 1 -3.

2. Preparation of 1-Deoxynojirimycine and its stereoisomers18,19
Four naturally occurring 1-deoxyiminosugars have been isolated to date. 1-Deoxynojirimycin (DNJ) (17) and 1-deoxymannojirimycin (manno-DNJ) (19) have been isolated from many plants and microorganisms.20 1-Deoxyaltronojirimycin (altro-DNJ) (18) and 1-deoxygulonojirimycin (gulo-DNJ) were recently isolated from Scilla sibirica (Hyacinthaceae)21 and Angylocalyx pynaertii (Leguminosae).22 Many efforts have been devoted in recent years to develop methodologies for the asymmetric syntheses of iminosugars owing to their promising therapeutic potential in a wide range of diseases. However, many syntheses of iminosugars have been focused on derivatives with D-gluco, D-manno, or D-galacto configurations because their targeting enzymes are essential for survival and the existence of all living organisms.23,24 There are a number of reports and reviews on in vitro and in vivo glycosidase inhibition by DNJ, manno-DNJ, galacto-DNJ, and their derivatives.25-27 On the other hand, only a few systematic studies of the properties of the enantiomers of DNJ, manno-DNJ, and galacto-DNJ, and other diastereomers of DNJ have been reported. Herein, we report on the syntheses of both enantiomers of DNJ (17), manno-DNJ (19), allo-DNJ (20), galacto-DNJ (3), altro-DNJ (18), gulo-DNJ, and ido-DNJ and systematic studies of their inhibitory activities with respect to glycosidases.

Firstly, DNJ (17) and its three stereoisomers, i.e., altro-DNJ (18), manno-DNJ (19), and allo-DNJ (20) of trans-4,5-substituted 1-deoxyiminosugars (nojirimycin-type) were synthesized from a common chiral building block 21 (Figure 3). In a project focused on the asymmetric synthesis of glycosidase inhibitors, we envisioned dioxanylpiperidene 21 would serve as a new common chiral building block, which represents an ideal precursor for the synthesis of DNJ (17) and its congeners (Figure 3). Herein we describe a straightforward synthesis of DNJ (17) and its congeners 18-20 via 21 starting from the Garner aldehyde 9 using catalytic RCM to construct the piperidine ring (Scheme 4). The D-serine-derived Garner aldehyde 9 provided an attractive starting point for the synthesis, since it reacts with organometalic reagents with a high degree of diastereoselectivity and racemization is minimal.28 The diastereoselective addition of vinyl metals to 9 could furnish the anti-vinyl alcohol anti-22, depending on the reaction conditions used.29 The use of HMPA as a cosolvent gave 22 with high anti-selectivity (anti:syn = 5.2:1) in 91% yield. The diastereoselectivity is convincingly accounted for by considering the preferred transition state in each reaction. According to the well- known Felkin-Anh model, the nucleophile would preferentially attack the si-face of aldehyde 9 thereby leading to an anti-configuration. The chromatographic separation of a diastereomeric mixture of alcohols 22 was not successful. Accordingly, treatment of 22 (anti:syn = 5.2:1) with HCl in chloroform afforded the readily separable 1,3-acetonides 23 (1.8%), 24 (47%), and 22 (32%). The relative configuration at the C-4 and C-5 of 24 was confirmed from the H–H-coupling constant (C4H and C5H), which is 9.5 Hz for the trans configuration. The N-allylation of 24 with allyl iodide using NaH as a base gave the diolefin product 25 in 95% yield. Finally, 25 was subjected to RCM in the presence of Grubbs’ 1st generation catalyst in dichloromethane. Under these conditions, the desired piperidene 21 was obtained in 97% yield.

With the promising chiral building block 21 in hand, our interest was directed to the synthesis of 1-deoxynojirimycin compounds 17-20. We first converted the double bond to an epoxy-functionality to give both 17 and 18 containing a trans diol in the 2 and 3 positions. The dioxirane, generated in situ from Oxone® with 1,1,1-trifuluoroacetone was reacted with 21 to give the anti epoxide 26 and syn epoxide 27 in 45 and 44% yields, respectively. The use of m-CPBA also resulted in low stereoselectivity to provide 26 and 27 in a ratio of 47:27 in 74% yield.

Acid hydrolysis of the epoxy ring of 26 was achieved using a 0.2/3/2 mixture of H2SO4/dioxane/H2O, followed by treatment with an ion-exchange resin to give DNJ (17)30 and altro-DNJ (18)31 in a ratio of 1:1 in 89% yield. Base catalyzed cleavage of the epoxide 26 using a mixture of KOH/dioxane/H2O gave 18, preferentially, in a ratio of 1:1.5 (17 to 18) in 99% yield. In contrast, both acidic hydrolysis and basic cleavage of the syn epoxide 27 afforded only altro-DNJ (18) in 63% and 68% yields, respectively, after treatment with ion-exchange resin. Although the rationale for these stereoselectivities remains unclear, we propose the following explanation. It is known that the opening of cyclohexene oxides generally proceeds via a diaxial reaction.32 The epoxy substituents at the 3 position of 26 and the 2 position of 27 are in a quasi-equatorial orientation, because anti-dioxanyl ring has a diequtorial orientation. Accordingly, it would be expected that axial attack would occur at the 3 position of 26 and the 2 position of 27. In the case of 26, however, steric repulsion between a nucleophile and substituent at 4 position would exist. Hence, the nucleophile may attack at the more remote site (the 2 position) of 26 (Scheme 5).

The stereoselective dihydroxylation of the double bond was next examined. Deprotection of the acetonide with p-TsOH in methanol gave diol 28 (94%), which was treated under modified Upjohn conditions, with a catalytic amount of K2OsO4·2H2O (5 mol %) and NMO as a cooxidant to give an inseparable mixture of tetraols 29 in 87% yield. Fortunately, acetylation of the tetraol gave a mixture of separable tetraacetates 30 and 31, which were isolated in 45% and 49% yields. Finally, the exposure of 30 and 31 to 6N HCl in methanol followed by treatment with an ion-exchange resin provided manno-DNJ (19)33 and allo-DNJ (20)34 in 94 and 91% yields, respectively (Scheme 6).

Thus, the new promising chiral building block 21 for the synthesis of 1-deoxy-4,5-trans-oriented iminosugars was prepared in only four steps from the Garner aldehyde 9. In practice, the first synthesis of all four isomers 17-20 of trans-4,5-orientated 1-deoxyiminosugars using 21 as a common chiral building block was achieved.
Secondly, galacto-DNJ (3), ido-DNJ (32), and gulo-DNJ (33) of cis-4,5-substituted 1-deoxyiminosugars were synthesized from a new common chiral building block, dioxanylpiperidene 34, which represents an ideal precursor (Figure 4).

The common intermediate 34 can be prepared by the RCM of diolefin 37, produced by the stereoselective coupling of 9 with vinyl metals. The diastereoselective addition of vinyl metals to 9 may furnish the syn vinyl alcohol, depending on the reaction conditions.35 The reagent, formed from in situ prepared vinyllithium and anhydrous zinc dibromide in diethyl ether was found to provide the syn alcohol as a solid with a 5:1 diastereoselectivity (syn:anti) in 91% yield. The diastereoselectivity can be rationalized by considering the preferred transition state in each reaction. The vinyl zinc bromide complex coordinated with the carbamate carbonyl in the transition state is delivered to the re face of the aldehyde carbonyl to afford the vinyl alcohol 22 with syn selectivity. The chromatographic separation of the diastereomeric mixture of alcohols 22 was incomplete. However, the 67% de of the syn-preferred 22 was improved to 92% de (72%) by one recrystallization. When the recrystallized 22 was treated with HCl gas in chloroform, it was converted to the 1,3-acetonide 23 (69%) together with the recovery of 22 (24%). N-Allylation of 23 with allyl iodide using NaH as a base gave the diolefin product 35 in 76%. Finally, 35 was subjected to RCM in the presence of Grubbs' 1st catalyst, in dichloromethane to provide the desired piperidene 34 in excellent yield (Scheme 7).

With the promising educt 34 in hand, our interest was directed to the stereoselective synthesis of compounds 3, 32, and 33. We first introduced an epoxy functionality into the double bond to obtain both 3 and 32 containing a trans diol in the 3 and 4 positions. The dioxirane, generated in situ from Oxone with 1,1,1-trifuluoroacetone, was reacted with 34 to give the anti epoxide 36 as a single diastereomer in 99% yield. This indicates that the epoxidation occurred exclusively from the less hindered convex face, because the concave face is shielded by a methyl substituent. On the other hand, the syn epoxide was obtained by the hydroxy-directed epoxidation of diol 37, prepared by hydrolysis of the acetonide derivative of 34 with p-TsOH in methanol. The epoxidation of the diol with m-CPBA followed by acetonization afforded the syn epoxide 38 in 53% overall yield (Scheme 8).
Acid hydrolysis of the epoxy ring of the syn epoxide 41 was accomplished by means of a mixture of H2SO4/1,4-dioxane/H2O in a ratio of 0.2/3/2, and further treatment with an ion-exchange resin (DOWEX® 1x2, OH- form) gave only galacto-DNJ (3)36 in 83% yield. On the other hand, the basic cleavage of the epoxide 36, accomplished using a mixture of KOH/1,4-dioxane/H2O, followed by a sequence of deprotection and desalting gave ido-DNJ (32)37 exclusively in 87% combined yield.

The stereoselective dihydroxylation of the double bond was also examined. Under modified Upjohn conditions, treatment of 34 with a catalytic amount of K2OsO4·2H2O (5 mol %) and NMO as a cooxidant gave the diol 39 as a single diastereomer in 85% yield. This remarkably high diastereoselectivity of the dihydroxylation can be attributed to the same steric blocking of the concave face as explained for the epoxidation of 34. Deprotection of 39 with HCl in methanol followed by treatment with an ion-exchange (DOWEX® 50Wx8 H+ form) afforded 1-deoxygulonojirimycin (33) in 90% combined yield. Unfortunately, the dihydroxylation of 34 or 37 using Donohoe conditions,17 AD-mix-α®, and AD-mix-β® gave no syn-dihydroxyl products, which were converted to talo-DNJ (40).38 Thus, a straightforward synthesis of 3, 32, and 33 using 34 has been demonstrated in a highly stereocontrolled fashion (Scheme 9).

Since all DNJ isomers of the D-form, except for talo-DNJ (40), were obtained, the preparation of the L-forms ent-17-20 and 3, 32, and 33 using the L-serine-derived Garner aldehyde ent-9 was achieved following almost the same procedure as that reported as shown in Schemes 7-9.

3. Preparation of Isofagomine39-42
The search for anomer selective β-glycosidase inhibitors has led to a new class of sugar-mimics, 1-N-iminosugars with a nitrogen atom at the anomeric position. The representative 1-N-iminosugar isofagomine 41 was first designed by Bols et al.43 as an apparent transition state analog mimicking the carbocationic form of the oxycarbenium-like transition state in which the positive charge resides at the anomeric carbon. Isofagomine was found to be a selective and very strong inhibitor of β-glucosidase [Ki = 0.11 μM, sweet almonds]43,44 and its 2-hydroxy analog noeuromycin 42 functions as a β-glucosidase inhibitor in the nanomolar range.45 Isofagomine derivatives have recently received a great deal of attention because they are new candidates for the therapeutic treatment of Gaucher’s disease and are currently in Phase II of clinical development. Due to the pronounced and selective inhibition activities of isofagomine and its congeners, an increased interest has developed in the synthesis of such 1-N-iminosugars (Figure 5).46

We envisioned the use of N-Boc-5-hydroxy-3-piperidene (43) as a general representative chiral building block that might permit easy access to these classes of compounds. The synthesis of racemic 43 as the starting material began with the regioselective opening of butadiene monoxide with allylamine followed by protection of the secondary amine to afford the metathesis precursor 44 (66%). The Grubbs’ catalysts could be used directly on 44 to afford the ring-closing metathesis product 43 in 99% yield.　Ogasawara reported47 on the lipase-catalyzed transesterification of N-Cbz-5-hydroxy-3-piperidene with vinyl acetate. We applied this method to racemic 43 using lipase and vinyl acetate. Of the various lipases tested, resolution was best achieved with lipase PS (Pseudomonas cepacia), immobilized on ceramic particles, in tert-butyl methyl ether at 40 °C, which gave the acetate 45 in 49% yield, along with the unreacted alcohol (R)-43, in 48% (>99% ee) yield. In addition, the enzymatic hydrolysis of the acetate 45 with the same lipase in 0.1 M phosphate buffer afforded the enantiomeric alcohol (S)-43, in 98% (>99% ee) yield (Scheme 10). In addition, (R)- or (S)-43 was prepared by the palladium-catalyzed deracemization of the methyl carbonate derivative of racemic 43.48

Alcohol (R)-43 was initially converted into the TBDPS derivative 46, thus avoiding the possible assistance of the hydroxyl group in the favored approach of the oxidant. Treatment of (R)-43 with tert-butyldiphenylsilyl chloride under basic conditions gave the TBDPS derivative 46 in 99% yield. The dioxirane, generated in situ49 from Oxone with 1,1,1-trifluoroacetone, was subsequently reacted with 46 to give the anti epoxide 47 and the syn epoxide 48 in 72% and 17% yields, respectively (Scheme 11).

The nucleophilic opening of the anti epoxide 47 with “higher order” cuprate50 free halide ions in the presence of boron trifluoride etherate as the activating species was carried out as follows:  treatment of 47 with (CH2=CH)2CuCNLi2 in the presence of BF3·OEt2 at −78 °C for 2 h gave 49 as the sole isolable product in 74% yield. Analogously, the reaction of 47 with Me2CuCNLi2 afforded only 50 in 71% yield. An attempt to employ Grignard reagents in the presence of cuprous bromide51 resulted in no reaction. Although the rationale for this high selectivity remains unclear, the following mechanism is consistent with the results. The regiochemistry of the nucleophilic opening of the epoxide on a six-membered ring is mainly subject to trans diaxial opening (Fürst−Plattner rule).32 Consequently, a high regioselectivity would result if the opening proceeded through only one of the two possible half-chair conformations (A and B). Thus, the exclusive attack of the nucleophile at C-5 as conformer A would involve a trans diaxial opening. On the other hand, if a nucleophile attacked at C-4 of 47 through conformer B steric hindrance by the pseudoequatorial OTBDPS group at C-3 would be in play. Therefore, a trans diaxial opening through one of the half-chair conformers, namely conformer A, would occur (Scheme 12). A similar regioselectivity has been reported for ring-opening reactions of trans-3-(benzyloxy)-1,2-epoxycyclohexane derivatives.52

With the vinyl product 49 in hand, our interest was directed to its conversion to isofagomine (43) and homoisofagomine (51). Oxidative cleavage of the vinyl group of 49 with OsO4 and NaIO4 afforded the aldehyde, which without purification, and after reduction with NaBH4 followed by deprotection, afforded isofagomine (41) in 85% combined yield.53 Next, the hydroboration of the vinyl group of 49 with 9-BBN followed by treatment with hydrogen peroxide gave the corresponding primary alcohol. Deprotection of the alcohol with 10% HCl in 1,4-dioxane afforded 51 in 86% combined yield. Conversion of 50 to 5’-deoxyisofagomine (52) was accomplished by deprotection. The complete deprotection of 50 by treatment with 10% HCl in 1,4-dioxane afforded 52 in 92% yield. Thus, the first synthesis of 51 and 52 was achieved. Enantiomers of 41, 51, and 52 were also prepared starting from (-)-43, following the same procedure as above (Scheme 13).

Next, we report on the asymmetric synthesis of all stereoisomers of 1-iminosugars such as isofagomine (41), using the [2,3]-Wittig rearrangement54 as a key step starting from the chiral N-Boc-5-hydroxy-3-piperidene (43) as depicted in Scheme 14. We began with the synthesis of a precursor 53 for the [2,3]-Wittig rearrangement from 43. O-Alkylation of (S)-43 with tributyl(iodomethyl)stannane55 in the presence of KH and n-Bu4NI gave the stannane product (S)-53 in 98% yield. With (S)-53 in hand, the [2,3]-Wittig rearrangement of (S)-53 was examined by transmetallation using n-BuLi, producing the requisite hydroxymethyl substituent (R)-5456 with no racemization ([1,2]-Wittig rearrangement).57 The use of less polar solvents such as n-pentane and n-hexane resulted in yields of 65% and 53%, respectively.

Having the key intermediate (R)-54 in hand, the stereoselective dihydroxylation of the double bond was examined. Treatment of (R)-54 with a catalytic amount of OsO4 (5 mol%) and NMO as a cooxidant gave an inseparable mixture of diastereomeric diols, which, after deprotection with 10% HCl in dioxane, followed by silica gel column chromatography using an eluent comprised of a mixture of solvents (methanol:10% NH4OH), gave 3-epiisofagomine (55) (75%) and 4-epiisofagomine (56) (14%). Donohoe reported that osmium tetroxide produces a bidendate and reactive complex with TMEDA, that can be used in the directed dihydroxylation of cyclic homoallylic alcohols.58 Under these conditions, control via hydrogen bonding preferentially led to the formation of the syn isomer in almost every case. However, the oxidation of (R)-54 with a combination of OsO4 with TMEDA gave 55 and 56 in 61% and 37% yields, respectively. Although the yield of 56 was increased, reversal of the ratio was not observed in any case.

We next set out to synthesize 41 and 57 from the O-TBDPS protected intermediate (R)-58. The silylation of (R)-54 followed by epoxidation was attempted. Thus, the dioxirane, generated in situ59 from Oxone by treatment with 1,1,1-trifluoroacetone was reacted with (R)-58 to give the anti epoxide 59 and the syn epoxide 60 in 52% and 34% yields, respectively, which were tentatively assigned based on steric considerations between the allylic substitutent of the six membered cyclic alkene and a substituent of dioxrane.60 Subsequently, the basic cleavage of the epoxy ring of 59 was accomplished using a mixture of KOH/1,4-dioxane/H2O at reflux followed by a sequence of deprotection with 6 N HCl and desalting to give 41 and 57, 28% and 62% yields, respectively (Scheme 16). The regiochemistry of the nucleophilic opening of the epoxide on a six-membered ring is mainly subject to trans diaxial opening (Fürst-Plattner rule). Consequently, this regioselectivity would result, if the opening proceeded through the two possible half chair conformations (A and B). A substituent at C-3 would be preferentially located in a pseudoequatrial orientation compared with a pseudoaxial substituent. Thus, the somewhat preferential attack of the hydroxide anion at C-4 through conformer A would occur by trans diaxial opening. On the other hand, a similar reaction using 60 somewhat preferentially gave 41 (53%) together with 57 (37%). On the basis of the above reasoning, the existence of conformer D would be major species and conformer C, the minor species (Scheme 17).

In addition, four enantiomers of 41, 55, 56, and 57 were prepared from (R)-43 according to the above described procedure.

An interesting acceleration effect of an allylic hydroxy group on ring-closing enyne metathesis has been recently found by us.41 Ring-closing enyne metathesis of terminal alkynes with an allylic hydroxy group proceeded smoothly without ethylene atmosphere,

which is generally necessary to promote the reaction. Utilizing this efficient ring-closing enyne metathesis based on the acceleration effect of an allylic hydroxy group, the synthesis of (+)-isofagomine was achieved (Scheme 18). The enyne substrate with an allylic hydroxy group, 61, was synthesized by means of an addition reaction between propargyl amine and butadiene monoxide followed by protection of the imino group with a tert-butoxycarbonyl (Boc) group (71% for 2 steps). The allylic hydroxy group-accelerated ring-closing enyne metathesis (AHA-RCEM) of 61 efficiently gave the cyclic product 62 (>99%) in a short reaction time. The hydroxyl group of 62 was then protected with a tert-butyldiphenylsilyl (TBDPS) group (99%). The TBDPS-protected product was treated with AD-mix-α®. The highly regioselective dihydroxylation of the terminal olefin proceeded to provide diol 63 (78%). Oxidative cleavage of the diol 63 with NaIO4 followed by reduction with NaBH4 gave the allyl alcohol 64 (98% in 2 steps). Protection of the hydroxy group of 64 with a benzyl group (92%) and subsequent deprotection of the TBDPS group (98%) gave the benzylated product 65. The kinetic transesterification of 65 with vinyl acetate using lipase PS-C permitted the excellent resolution of the enantiomers, giving almost enantiomerically pure 67 (46%, 99% ee) together with acetate 66. Hydroboration of the internal olefin of 67 followed by oxidation with NaOH/H2O2 afforded diol 68 (67% for 2 steps). After both deprotection of the benzyl group (99%) and the Boc group, (+)-isofagomine (41) (78%) and (-)-3-epi-isofagomine (ent-55) (20%) were obtained. (+)-Isofagomine was synthesized in 11.5% total yield from commercially available 1,3-butadiene monoxide.

Our attention was next focused on the synthesis of 2-propylisofagomine 70 using allylic hydroxy group accelerated ring-closing enyne methasesis, since the synthesis of 2-alkyl isofagomines remains unexplored.61 Our strategy for the synthesis of 2-propylisofagomine is outlined in Scheme 19, which shows that the desired iminosugars A can be produced from the cyclic diene B by an operation similar to that described above. The piperidene core could be prepared by the AHA-RCEM of the terminal alkyne C as a key step. Hence, we began with a synthesis of the precursor C, which is available from the known chiral N-nosyl allylic amine D (Scheme 19).

In initial experiments, an asymmetric allylic amination between N-(o-nosyl)amine and carbonate provided N-hexenylnosylamide 71 in 82% yield with 94% ee using the procedure reported by Weihofen et al.62 The propargylation of 71 with propargyl bromide in the presence of potassium carbonate gave the acetylene product 72 in quantitative yield, which, on ozonolysis, afforded the aldehyde 73 in 88% yield. The vinylation of 73 with vinylmagnesium chloide in THF at –78 °C proceeded stereoselectively to give the allyl alcohol 74 as a single diastereomer in 66% yield. Although the stereochemistry of 74 remains unclear in this stage, we tentatively concluded that it is 3S,4R in the light of the Felkin-Anh model. Having the precursor 74, the AHA-RCEM of 74 was carried out using Grubbs’ 1st catalyst (10 mol %) as a catalyst at room temperature in a short reaction time to afford cyclic diene 75 in 85% yield. Treatment of 75 with AD-mix-β® as a bulky oxidant resulted in the highly regioselective dihydroxylation of the terminal olefin to provide diol 76 (81%). Oxidative cleavage of the diol 76 with NaIO4, followed by reduction with NaBH4, gave the allyl alcohol 77 (94% over two steps) (Scheme 20). At this stage, the stereochemistry of 77 was determined to be 2R,3S, by an X ray crystallographic analysis of the di-p-nitrobenzoate of 77.

With 77 in hand, we attempted to hydroxylate the 4 position of 77 by hydroboration followed by oxidation. The olefin 77 was treated with a BH3·THF complex (6 equiv.) at room temperature for 13 h, followed by oxidation with 3M NaOH and 30% H2O2 to give a separable mixture of triols 78 and 79 in 72% yield. Unfortunately, the ratio of the two diastereomeric triols was about 1:1 with no selectivity. We concluded that 78 was produced via transition state A with chelation between the hydroxy group at the 3 position and the borane reagent. Therefore, we hypothesized that the hydroboration of a protected silyl ether 81 could preferentially produce the desired isomer via transition state B due to steric repulsion between the boron reagent and the bulky O-silylated group (Figure 6). In practice, the hydroboration-oxidation of 81 produced the expected product 82 as a single isomer in 60% yield. Removal of the TBDPS groups by treatment with TBAF smoothly furnished the triol 83 in 96% yield. Finally, deprotection of the Ns group with benzenethiol in the presence of K2CO3 gave the desired 2-propylisofagomine 70 in 82% yield. Since a nuclear Overhauser effect (NOE) was observed between axial hydrogens at the 2 and 4 positions and also between axial hydrogens at the 3 and 5 positions, the stereochemistry of 70 was confirmed to be 2R,3R,4R,5R. In addition, 78 was transformed by denosylation into 80 in 90% yield. Thus, 2-propylisofagomine 70 was stereoselectively prepared from the carbonate in 13% overall yield using AHA-RCEM as the key step.

SUMMARY
In summary, we report herein on the synthesis of both enantiomers of iminosugars such as fagomine, 1-deoxynojirimycine, and isofagomine accompanied by their congeners. Although systematic studies of their glycosidase inhibitory activities are not described in this review, some quite interesting results have been observed.63 In addition, 1-C-alkyl-L-arabinoimonofuranoses, potential α-glucosidase inhibitors, were prepared and have potential for use in the treatment of type 2 diabetes.64

ACKNOWLEDGEMENT
This study was financially supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (JSPS) and in part by a grant of Strategic Research Foundation Grant-aided Project for Private Universities from Ministry of Education, Culture, Sport, Science, and Technology, Japan (MEXT).

References

1. (a) N. Asano, R. J. Nash, R. J. Molyneux, and G. W. J. Fleet, Tetrahedron: Asymmetry, 2000, 11, 1645; CrossRef (b) A. A. Watson, G. W. J. Fleet, N. Asano, R. J. Molyneux, and R. J. Nash, Phytochemistry, 2001, 56, 265. CrossRef
2. (a) L. Ratner, N. V. Heyden, and D. Dedera, Virology, 1991, 181, 180; CrossRef (b) P. M. Rudd, T. Elliott, P. Cresswell, I. A. Wilson, and R. A. Dwek, Science, 2001, 291, 2370; CrossRef (c) F. Chery, L. Cronin, J. L. O'Brien, and P. V. Murphy, Tetrahedron, 2004, 60, 6597; CrossRef (d) D.-S. Lee, K.-E. Jung, C.-H. Yoon, H. Lim, Y.-S. Bae, Antimicrob. Agents Chemother., 2005, 49, 4110; CrossRef (e) P. Greimel, J. Spreitz, A. E. Stütz, and T. M. Wrodnigg, Curr. Top. Med. Chem., 2003, 3, 513; CrossRef (f) J. Alper, Science, 2001, 291, 2338; CrossRef (g) D. Pavlovic, D. C. Neville, A. O. Argaud, B. Blumberg, R. A. Dwek, W. B. Fischer, and N. Zitzmann, Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 6104; CrossRef (h) D. Durantel, S. Carroueé-Durantel, N. Branza-Nichita, R. A. Dwek, and N. Zitzmann, Antimicrob. Agents Chemother., 2004, 48, 497; CrossRef (i) S.-F. Wu, C.-J. Lee, C.-L. Liao, R. A. Dwek, N. Zitzmann, and Y.-L. Lin, J. Virol., 2002, 76, 3596. CrossRef
3. (a) J.-Y. Sun, M.-Z. Zhu, S.-W. Wang, S. Miao, Y.-H. Xie, and J.-B. Wang, Phytomedicine, 2007, 14, 353; CrossRef (b) H. Paulsen and I. Brockhausen, Glycoconjugate J., 2001, 18, 867; CrossRef (c) P. E. Gross, M. A. Baker, J. P. Carver, and J. W. Dennis, Clin. Cancer Res., 1995, 1, 935.
4. (a) P. B. Anzeveno, L. J. Creemer, J. K. Daniel, C.-H. R. King, and P. S. Liu, J. Org. Chem., 1989, 54, 2539; CrossRef (b) J. A. Balfour and D. McTavish, Drugs, 1993, 46, 1025. CrossRef
5. (a) S. Cren, S. S. Gurcha, A. J. Blake, G. S. Besra, and N. R. Thomas, Org. Biomol. Chem., 2004, 2, 2418; CrossRef (b) T. M. Wrodnigg and F. K. Sprenger, Mini-Rev. Med. Chem., 2004, 4, 437.
6. (a) T. D. Butters, R. A. Dwek, and F. M. Platt, Chem. Rev., 2000, 100, 4683; CrossRef (b) T. M. Cox, J. M. F. G. Aerts, G. Andria, M. Beck, N. Belmatoug, B Bembi, R. Chertkoff, S. Vom Dahl, D. Elstein, R. Erickson, M. Giralt, R. Heitner, C. Hollak, M. Hrebicek, S. Lewis, A. Mehta, G. M. Pastores, A. Rolfs, M. C. Miranda, and A. Zimran, J. Inherit. Metab. Dis., 2003, 26, 513; CrossRef (c) J. Matsuda, O. Suzuki, A. Oshima, Y. Yamamoto, A. Noguchi, K. Takimoto, M. Itoh, Y. Matsuzaki, Y. Yasuda, S. Ogawa, Y. Sakata, E. Nanba, K. Higaki, Y. Ogawa, L. Tominaga, K. Ohno, H. Iwasaki, H. Watanabe, R O. Brady, and Y. Suzuki, Proc. Nat. Acad. Sci. U.S.A., 2003, 100, 15912. CrossRef
7. (a) P. Compain and O. R. Martin, ‘Iminosugars—From Synthesis to Therapeutic Applications,’John Wiley & Sons Ltd: West Sussex England, 2007; (b) E. Broges de Melo, A. da Silveira Gome, and I. Carvalho, Tetrahedron, 2006, 62, 10277; CrossRef (c) M. S. M. Pearson, M. Mathé-Allainmat, V. Fargeas, and J. Lebreton, Eur. J. Org. Chem., 2005, 2159; CrossRef (d) K. Afarinkia and A. Bahar, Tetrahedron: Asymmetry, 2005, 16, 1239; CrossRef (e) D. P. Germain, Clin. Genet., 2004, 65, 77; CrossRef (f) L. Cipolla, B. La Ferla, and F. Nicotra, Curr. Top. Med. Chem., 2003, 3, 1349; CrossRef (g) P. Compain and O. R. Martin, Curr. Top. Med. Chem., 2003, 3, 541; CrossRef (h) N. Asano, Curr. Top. Med. Chem., 2003, 3, 471; CrossRef (i) V. H. Lillelund, H. H. Jensen, X. Liang, and M. Bols, Chem. Rev., 2002, 102, 515; CrossRef (j) N. Asano, R. J. Nash, R. J. Molyneux, and G. W. J. Fleet, Tetrahedron: Asymmetry, 2000, 11, 1645; CrossRef (k) M. S. J. Simmonds, G. C. Kite, and E. A. Porter, Taxonomic Distribution of Iminosugars in Plants and Their Biological Activities. In Iminosugars as Glycosidase Inhibitors; A. Stütz, Ed.; Wiley-VCH: Weinheim, Germany 1999; p 8.
8. (a) Y. Banba, C. Abe, H. Nemoto, A. Kato, I. Adachi, and H. Takahata, Tetrahedron: Asymmetry, 2001, 12, 817; CrossRef (b) H. Takahata, Y. Banba, H. Ouchi, H. Nemoto, A. Kato, and I. Adachi, J. Org. Chem., 2003, 68, 3603. CrossRef
9. A. Kato, N. Asano, H. Kizu, K. Matsui, A. A. Watson, and R. J. Nash, J. Nat. Prod., 1997, 60, 312. CrossRef
10. (a) H. Nojima, I. Kimura, F.-J. Chen, Y. Sugiura, M. Haruno, A. Kato, and N. Asano, J. Nat. Prod., 1998, 61, 397; CrossRef (b) S. Taniguchi, N. Asano, F. Tomino, and I. Miwa, Horm. Metab. Res., 1998, 30, 679. CrossRef
11. J.-Q. Fan, S. Ishii, N. Asano, and Y. Suzuki, Nature Med., 1999, 5, 112. CrossRef
12. R. H. Grubbs and S. Chang, Tetrahedron, 1998, 54, 4413. CrossRef
13. A. Dondoni and D. Perrone, Org. Synth., 1999, 77, 64.
14. S. E. Denmark, D. C. Forbes, D. S. Hays, J. S. DePue, and R. G. Wilde, J. Org. Chem., 1995, 60, 1391. CrossRef
15. E. A. Severino and C. R. D. Correia, Org. Lett., 2000, 2, 3039. CrossRef
16. V. VanRheenen, R. C. Kelly, and D. Y. Cha, Tetrahedron Lett., 1976, 17, 1973. CrossRef
17. (a) T. J. Donohoe, L. Mitchell, M. J. Waring, M. Helliwell, A. Bell, and N. J. Newcombe, Tetrahedron Lett., 2001, 42, 8951; CrossRef (b) For a review, see: T. J. Donohoe, Synlett, 2002, 1223. CrossRef
18. H. Takahata, Y. Banba, M. Sasatani, H. Nemoto, A. Kato, and I. Adachi, Tetrahedron, 2004, 60, 8199. CrossRef
19. H. Takahata, Y. Banba, H. Ouchi, and H. Nemoto, Org. Lett., 2003, 5, 2527. CrossRef
20. A. A. Watson, G. W. J. Fleet, N. Asano, R. J. Molyneux, and R. J. Nash, Phytochemistry, 2001, 56, 265. CrossRef
21. T. Yamashita, K. Yasuda, H. Kizu, Y. Kameda, A. A. Watson, R. J. Nash, G. W. J. Fleet, and N. Asano, J. Nat. Prod., 2002, 65, 1875. CrossRef
22. N. Asano, K. Yasuda, H. Kizu, A. Kato, J.-Q. Fan, R. J. Nash, G. W. J. Fleet, and R. J. Molyneux, Eur. J. Biochem., 2001, 268, 35. CrossRef
23. B. Winchester and G. W. J. Fleet, J. Carbohydr. Chem., 2000, 19, 471. CrossRef
24. N. Asano and H. Hashimoto, Azaglycomimetics: synthesis and chemical biology. Glycoscience: Chemistry and Chemical Biology III; B. O. Fraser-Reid, K. Tatsuta, and J. Thiem, Eds.; Springer- 
Verlag: Berlin, Heidelberg, 2001; pp 2541-2594.
25. A. D. Elbein, Annu. Rev. Biochem., 1987, 56, 497. CrossRef
26. A. D. Elbein, FASEB J., 1991, 5, 3055.
27. B. Winchester and G. W. J. Fleet, Glycobiology, 1992, 2, 199. CrossRef
28. (a) P. Garner, J. M. Park, and E. Malecki, J. Org. Chem., 1988, 53, 4395; CrossRef (b) P. Herold, Helv. Chim. Acta, 1988, 71, 354. CrossRef
29. (a) L. Williams, Z. Zhang, F. Shao, P. J. Carroll, and M. M. Joullie, Tetrahedron, 1996, 52, 11673; CrossRef (b) R. S. Coleman and A. J. Carpenter, Tetrahedron Lett., 1992, 33, 1697. CrossRef
30. For recent synthesis of 1-deoxynojirimycin, see: (a) B. Zhou, H. Tang, H. Feng, and Y. Li, Synlett, 2011, 2709; CrossRef (b) S. K. Bagal, S. G. Davies, J. A. Lee, P. M. Roberts, P. M. Scott, and J. E. Thomson, J. Org. Chem., 2010, 75, 8133; CrossRef (c) G. Danoun, J. Ceccon, A. E. Greene, and J.-F. Poisson, Eur. J. Org. Chem., 2009, 4221; CrossRef (d) X. Song and R. I. Hollingsworth, Tetrahedron Lett., 2007, 48, 3115; CrossRef (e) A. Roy, B. Achari, and S. B. Mandal, Synthesis, 2006, 1035; CrossRef (f) P. Somfai, P. Marchand, S. Torsell, and U. M. Lindstrom, Tetrahedron, 2003, 59, 1293; CrossRef (g) D. L. Comins and A. B. Fulp, Tetrahedron Lett., 2001, 42, 6839. CrossRef
31. For recent syntheses of 1-deoxyaltronojirimycin, see: (a) O. K. Karjalainen and A. M. P. Koskinen, Org. Biomol. Chem., 2011, 9, 1231; CrossRef (b) D. D. Dhavale, S. D. Markad, N. S. Karanjule, and J. PrakashaReddy, J. Org. Chem., 2004, 69, 4760; CrossRef (c) O. V. Singh and H. Han, Tetrahedron Lett., 2003, 44, 2387; CrossRef (d) K. Asano, T. Hakogi, S. Iwama, and S. Katsumura, Chem. Commun., 1999, 41. CrossRef
32. E. L. Eliel, S. H. Wilen, and E. L Eliel, Ed., Stereochemistry of Organic Compounds. Wiley: New York; 1994. 730.
33. For recent syntheses of 1-deoxymannonojirimycin, see: (a) M. N. Gandy, M. J. Piggott, and K. A. Stubbs, Aust. J. Chem., 2010, 63, 1409; CrossRef (b) I. S. Kim, H. Y. Lee, and Y. H. Jung, Heterocycles, 2007, 71, 1787; CrossRef (c) J. G. Knight and K. Tchabanenko, Tetrahedron, 2003, 59, 281; CrossRef (d) J. L. O’Brien, M. Tosin, and P. V. Murphy, Org. Lett., 2001, 3, 3353; CrossRef (e) M. H. Haukaas and G. A. O’Doherty, Org. Lett., 2001, 3, 401; CrossRef (f) H. Yokoyama, K. Otaya, H. Kobayashi, M. Miyazawa, S. Yamaguchi, and Y. Hirai, Org. Lett., 2000, 2, 2427. CrossRef
34. For recent syntheses of 1-deoxyallonojirimycin, see: (a) R. Sridhar, B. Srinivas, and K. R. Rao, Tetrahedron, 2009, 65, 10701; CrossRef (b) F. Ferreira, C. Botuha, F. Chemla, and A. Perez-Luna, J. Org. Chem., 2009, 74, 2238; CrossRef (c) N. Ruiz, T. M. Ruanova, O. Blanco, F. Nunez, C. Pato, and V. Ojea, J. Org. Chem., 2008, 73, 2240. CrossRef
35. (a) L. Williams, Z. Zhang, F. Shao, P. J. Carroll, and M. M. Joullie, Tetrahedron, 1996, 52, 11673; CrossRef (b) R. S. Coleman and A. J. Carpenter, Tetrahedron Lett., 1992, 33, 1697. CrossRef
36. For recent syntheses of 1-deoxygalactonojirimycin, see: (a) M. S. M. Timmer, E. M. Dangerfield, J. M. H. Cheng, S. A. Gulab, and B. L. Stocker, Tetrahedron Lett., 2011, 52, 4803; CrossRef (b) T.-H. Chan, Y.-F. Chang, J.-J. Hsu, and W.-C. Cheng, Eur. J. Org. Chem., 2010, 5555; CrossRef (c) O. K. Karjalainen, M. Passiniemi, and A. M. P. Koskinen, Org. Lett., 2010, 12, 1145; CrossRef (d) C. Boucheron, P. Compain, and O. R. Martin, Tetrahedron Lett., 2006, 47, 3081. CrossRef
37. For recent syntheses of 1-deoxyidonojirimycin, see: (a) N. Palyam and M. Majewski, J. Org. Chem., 2009, 74, 4390; CrossRef (b) O. V. Singh and H. Han, Tetrahedron Lett., 2003, 44, 2387. CrossRef
38. For recent syntheses of 1-deoxygulonojirimycin, see: (a) S. Ghosh, J. Shashidhar, and S. Dutta, Tetrahedron Lett., 2006, 47, 6041; CrossRef (b) S.-J. Pyun, K.-Y. Lee, C.-Y. Oh, J.-E. Joo, S.-H. Cheon, and W.-H. Ham, Tetrahedron, 2005, 61, 1413; CrossRef (c) M. H. Haukaas and G. A. O’Doherty, Org. Lett. 2001, 3, 401. CrossRef
39. (a) H. Ouchi, Y. Mihara, and H. Takahata, J. Org. Chem., 2005, 70, 5207; CrossRef (b) H. Ouchi, Y. Mihara, H. Watanabe, and H. Takahata, Tetrahedron Lett., 2004, 45, 7053. CrossRef
40. Y. Mihara, H. Ojima, T. Imahori, Y. Yoshimura, H. Ouchi, and H. Takahata, Heterocycles, 2007, 72, 633. CrossRef
41. (a) T. Imahori, H. Ojima, Y. Yoshimura, and H. Takahata, Chem. Eur. J., 2008, 14, 10762; CrossRef (b) T. Imahori, H. Ojima, H. Tateyama, Y. Mihara, and H. Takahata, Tetrahedron Lett., 2008, 49, 265. CrossRef
42. T. Taguchi, T. Imahori, Y. Yoshimura, A. Kato, I. Adachi, M. Kawahata, K. Yamaguchi, and H. Takahata, Heterocycles, 2012, 84, 929. CrossRef
43. T. M. Jespersen, W. Dong, M. R. Sierks, T. Skrydstrup, I. Lundt, and M. Bols, Angew. Chem. Int. Ed. Engl., 1994, 33, 1778. CrossRef
44. A. Bülow, I. W. Plesner, and M. Bols, J. Am. Chem. Soc., 2000, 122, 8567. CrossRef
45. H. Liu, X. Liang, H. Søhoel, A. Bülow, and M. Bols, J. Am. Chem. Soc., 2001, 123, 5116. CrossRef
46. (a) Y. Ichikawa, Y. Igarashi, M. Ichikawa, and Y. Suhara, J. Am. Chem. Soc., 1998, 120, 3007; CrossRef b) M. M. Matin, T. Sharma, S. G. Sabharwal, and D. D. Dhavale, Org. Biomol. Chem., 2005, 3, 1702. CrossRef
47. H. Sakagami and K. Ogasawara, Synthesis, 2000, 521. CrossRef
48. H. Takahata, Y. Suto, E. Kato, Y. Yoshimura, and H. Ouchi, Adv. Syn. Catal., 2007, 349, 685. CrossRef
49. D. Yang, M.-K. Wong, and Y.-C. Yip, J. Org. Chem., 1995, 60, 3887. CrossRef
50. Review: B. H. Lipshutz, R. S. Wilhelm, and J. A. Kozlowski, Tetrahedron, 1984, 40, 5005. CrossRef
51. (a) M. A. Tius and A. H. Fauq, J. Org. Chem., 1983, 48, 4131; CrossRef (b) W. R. Roush, M. A. Adam, A. E. Walts, and D. J. Harris, J. Am. Chem. Soc., 1986, 108, 3422. CrossRef
52. (a) P. Crotti, G. Renzi, L. Favero, G. Roselli, V. D. Bussolo, and M. Caselli, Tetrahedron, 2003, 59, 1453; CrossRef (b) P. Crotti, V. D. Bussolo, L. Favero, F. Macchia, G. Renzi, and G. Roselli, Tetrahedron, 2002, 58, 7119; CrossRef (c) P. Crotti, V. D. Bussolo, L. Favero, M. Pineschi, F. Marianucci, G. Renzi, G. Amici, and G. Roselli, Tetrahedron, 2000, 56, 7513. CrossRef
53. For recent syntheses of isofagomine, see: (a) G. Malik, X. Guinchard, and D. Crich, Org. Lett., 2012, 14, 596; CrossRef (b) A. Rives, Y. Genisson, V. Faugeroux, N. Saffon, and M. Baltas, Synthesis, 2009, 3251; CrossRef (c) P. E. R. Espeel, K. Piens, N. Callewaert, and J. Van der Eycken, Synlett, 2008, 2321; CrossRef (d) G. Guanti and R. Riva, Tetrahedron Lett., 2003, 44, 357; CrossRef (e) G. Pandey, M. Kapur, M. I. Khan, and S. M. Gaikwad, Org. Biomol. Chem., 2003, 1, 3321; CrossRef (f) J. Andersch and M. Bols, Chem. Eur. J., 2001, 7, 3744; CrossRef (g) D. L. Comins and A. B. Fulp, Tetrahedron Lett., 2001, 42, 6839. CrossRef
54. For reviews on [2,3]-Wittig rearrangement, see: (a) T. Nakai and K. Mikami, Chem. Rev., 1986, 86, 885; CrossRef (b) J. A. Marshall, In Comprehensive Organic Synthesis; ed. by B. M. Trost and I. Fleming, Pergamon: Oxford, 1991; Vol. 3, 975; CrossRef (c) T. Nakai and K. Mikami, Synthesis, 1991, 594; CrossRef (d) T. Nakai and K. Mikami, Org. React., 1994, 46, 105; (e) T. Nakai and K. Tomooka, Pure Appl. Chem., 1997, 69, 595. CrossRef
55. J. Aahman and P. Somfai, Synth. Commun., 1994, 24, 1117. CrossRef
56. W. C. Still and A. Mitra, J. Am. Chem. Soc., 1978, 100, 1927. CrossRef
57. (a) L. A. Paquette and T. Sugimura, J. Am. Chem. Soc., 1986, 108, 3841; CrossRef (b) J. A. Marshall, E. D. Robinson, and A. Zapata, J. Org. Chem., 1989, 54, 5854. CrossRef
58. T. J. Donohoe, L. Mitchell, M. J. Waring, M. Helliwell, A. Bell, and N. J. Newcombe, Org. Biomol. Chem., 2003, 1, 2173. CrossRef
59. D. Yang, M.-K. Wong, and Y.-C. Yip, J. Org. Chem., 1995, 60, 3887. CrossRef
60. (a) R. W. Murray, M. Singh, B. L. Williams, and H. M. Moncrieff, J. Org. Chem., 1996, 61, 1830; CrossRef (b) D. Yang, G.-S. Jiao, Y.-C. Yip, and M.-K. Wong, J. Org. Chem., 1999, 64, 1635. CrossRef
61. X. Zhu, K. A. Sheth, S. Li, H-H. Chang, and J.-Q. Fan, Angew. Chem. Int. Ed., 2005, 44, 7450. CrossRef
62. R. Weihofen, O. Tverskoy, and G. Helmchen, Angew. Chem. Int. Ed., 2006, 45, 5546. CrossRef
63. (a) A. Kato, N. Kato, E. Kano, I. Adachi, K. Ikeda, L. Yu, T. Okamoto, Y. Banba, H. Ouchi, H. Takahata, and N. Asano, J. Med. Chem., 2005, 48, 2036; CrossRef (b) N. Asano, K. Ikeda, L. Yu, A. Kato, K. Takebayashi, I. Adachi, I. Kato, H. Ouchi, H. Takahata, and G. W. J. Feet, Tetrahedron: Asymmetry, 2005, 16, 223; CrossRef (c) C. Kuriyama, O. Kamiyama, K. Ikeda, F. Sanae, A. Kato, I. Adachi, T. Imahori, H. Takahata, T. Okamoto, and N. Asano, Bioorg. Med. Chem., 2008, 16, 7330; CrossRef (d) A. Kato, S. Miyauchi, N. Kato, R. J. Nash, Y. Yoshimura, I. Nakagome, S. Hirono, H. Takahata, and I. Adachi, Bioorg. Med. Chem., 2011, 19, 3558. CrossRef
64. Y. Natori, T. Imahori, K. Murakami, Y. Yoshimura, S. Nakagawa, A. Kato, I. Adachi, and H. Takahata, Bioorg. Med. Chem. Lett., 2011, 21, 738. CrossRef