1. Introduction

Nucleosides are fundamental building blocks of biological systems. The natural nucleosides and their analogs find numerous applications as antibiotic, fungicidal, anti-tumor and anti-viral agents. Therefore, the chemistry of these compounds has been studied extensively and today nucleoside chemistry represents an important area of research for modern drug discovery. Vital elements of the structure of nucleoside are: 1) hydroxymethyl functionality; 2) spacer [e.g., sugar component] on which the hydroxymethyl group is a part; 3) heteroaryl base or aryl group attached to the sugar. C-Nucleosides are a subtype of these compounds in which the spacer and heteroaryl(aryl) groups are linked via C-C bond (Figure 1). The attractive feature of C-Nucleosides arises from the presence of this C-C glycosidic bond, which gives a greater resistance towards chemical and enzymatic hydrolysis than N-Nucleosides.

Figure 1. Generalized structure of a C-nucleoside analogue

C-Nucleosides can be classified according to their sugar moiety into five different categories:

1- Cyclic C-Nucleoside

2- Acyclic C-Nucleoside

3- Homocyclic C-Nucleoside

4- Reverse C-Nucleoside

5- Carbocyclic C-Nucleoside

6- Heterocyclic C-nucleoside

These features have motivated organic chemists to develop efficient and practical synthetic methods, and numbers of strategies have been devised to design nucleoside analogues: 1) building up the heterocyclic unit on a suitably functionalized carbohydrate moiety; 2) building up a carbohydrate moiety on a suitably functionalized heterocyclic compound; 3) transformation of C-Nucleoside into another; 4) direct coupling between a suitably protected carbohydrate moiety with a preformed heterocyclic compound.

In this report we wish to view the latest development (2009-2011) with spot lighting on the tactics chemists deliberate based on their location(s).

2. Building up the Heterocyclic Unit on a Suitably Functionalized Carbohydrate Moiety

2.1. From Nitrile Terminal on C1 Atom

2.1.1. Cyclic C-Nucleosides

Functionalization of the carbohydrate moiety with nitrile group at the anomeric carbon gave an important precursor for synthesis of C-Nucleoside. Protected ribofuranosyl nitrile 5a-c was converted to the corresponding ribofuranosyl thioamide (6a-c), whose reaction with ethyl bromopyruvate gave the protected tiazofurin derivative 7a-c. Treatment of the latter with methanolic ammonia afforded The thiazole C-Nucleoside 8a-c (Scheme 1). [1, ].

Scheme 1. Construction of C-nucleosides 7 from nitrile precursors

2.1.2. Homocyclic C-Nucleosides

The homocyclic C-nucleoside analogue was also reported from the nitrile 9 using the same procedure already used for preparation of 8a-c to give the corresponding homo-C-tiazofurin analogue bearing 2,3-anhydro ribofuranosyl moiety 13 and 14, which represent the first biologically active tiazofurin analogue that demonstrate antiproliferative activity (Scheme 2). [2, ].

Scheme 2. Construction of homocyclic C-nucleoside 13 from nitrile precursor

2.2. From Alkynyl Terminal on C1 Atom

2.2.1. Cyclic C-Nucleosides

Another essential functional group for synthesis of hetrocyclic moiety upon carbohydrate is the ethynyl group at the anomeric carbon. Sonogashira reaction of 1-ethynyldeoxyriboside 16α or 16β with trifluoroacetanilide derivatives 15a-d furnished the corresponding alkynyldeoxyribose derivatives 17α or 17β, whose cyclization led to indolyl C-Nucleoside 18α or 18β (Scheme 3). [3, ].

Scheme 3. Synthetic steps to prepare C-nucleoside 18 from alkynyl precursor

Monosubstituted ribosyl acetylene 19a-b was also used for a 1,3- dipolar cycloaddition with benzylazide in presence of copper catalysis to direct the formation of only 1,4 disubstituted regioisomer of 1,2,3-triazole C-Nucleoside 20a-b which is considered as carbonylated analogue of ribavirin (Scheme 4). [4, ].

Scheme 4. Synthesis of 1,2,3-triazole C-nucleosides 20 from alkynyl precursor

Disubstituted ribosyl alkyne 22 obtained from riboside 21 using indium(0)-mediated alkynylation reaction was also involved into 1,3-dipolar cycloaddition reaction with benzylazide to afford 1,4,5-trisubstituted triazole C-Nucleosides 23 and 24 (Scheme 5). [4, ] The reaction was conducted under micellar catalysis to improve the regioselctivity.

Scheme 5. Synthetic approach to C-nucleosides 23 and 24 via 1,3-dipolar cyclization

2.2.2. Heterocyclic C-Nucleosides

The same methodolgy of heterocyclization of ethynyl group into 1,2,3-triazole was used to prepare the aza-C-Nucleoside analogue 28 (Scheme 6). [5, , ].

Scheme 6. Synthetic approach to aza-C-nucleosides 28 via 1,3-dipolar cyclization

2.3. From Carbonyl Functionality on C1 Atom

2.3.1. Cyclic and Homocyclic C-Nucleosides

Hantzsch reaction on the aldehydo of 29a with β-ketoester 30 and enamine ester 31, under L-proline catalysis gave dihydropyridine C-Nucleoside 32a in stereoselective manner with (de 95%). The use of C-glycosyl aldehyde with elongated chain 29b-f extends the synthesis to the homo-C-Nucleoside analogue 32b-f (Scheme 7). [6, , ].

Wittig type reaction of sugar functionalized with keto group at the 4- position 33 and stabilized phosphrane leads to 34 whose mild oxidation with pyridinium chlorochromate (PCC) afforded β-sugar- β-formy-α-β-unsaturated ester 35. Utilization of 35 in cyclocondensation reaction with hydrazine hydrate or its derivatives gave pyrazole, pyrazoline, pyridazionone pseudo C-Nucleoside (Scheme 8). [7, ].

2.3.2. Acyclic C-Nucleosides

Construction of the heterocyclic moiety on sugar can also be used for synthesis of acyclo C-Nucleoside analogs. Thus, the condensation of heterocyclic hydrazine derivatives with some monosaccharide gave the corresponding aldehydo sugar hydrazones, which on acetylation gave the o-acetylated sugar derivatives subsequent oxidative cyclization followed by de-o-acetylation afford the 1,2,4-triazole C-nucleoside. Scheme 9 [8-14, , ].

In most cases acetylation of the sugar hydrazones (40 a, b, c, f, g) gave directly the O- acetylated cyclic C-nucleoside, on the other hand all attempts dehydrogeative cyclization of sugar hydrazone (40d) or its acetylated derivative (41d) failed. Competition between two ortho ring nitrogen (NH, =N) in the oxidative cyclization process such as in case of the O-acetylated hydrazone (41c, g, h) gave product corresponded to cyclization between the azomethine carbon and the hydrogenated nitrogen of the ring.

Reaction of acid hydrazides with some aldehydo sugars gave the corresponding sugar aroylhydrazones. When those hydrazones were heated in acetic anhydride at 100 ^oC their corresponding 1,3,4-oxadiazoline acyclic C-nucleosides 46a-b were isolated (Scheme 10). [15-16, , ].

Cyclocondensation of aldehydo sugar Schiff base 48 with thioglycolic acid in dry dioxane afforded the corresponding thiazolidinone a cyclic C-nucleoside 49 (Scheme 11). [17, ].

Following the Hantzch approach for synthesis of pyridine, treatment of D-glyceraldehyde 50 with methyl acetoacetate and dimedone 52 in presence of bentonite clay as a support , ammonium nitrate as source of ammonia, and HNO₃ as oxidant was furnished, after chromatgraphic purification, C-acyclic pyridine nucleoside 53 (Scheme 12).. [18, , ]

The Kidawai^'s method for synthesis of imidazole derivative via condensation of 1,2- dicarbonyl compound such as benzil with aldehyde in excess of ammonium hydroxide under microwave irradiation (MWI) was used for the synthesis of imidazole C-acyclic nucleoside 59 by utilizing the aldehydo functional group of protected D-glyceraldehyde 54. While condensation of 54 with ethyl acetoacetate and urea using natural phosphate doped with ZnCl₂ under MWI afforded acyclic C-nucleoside having dihydropyrimidinone 55. An efficient synthesis of the C-acyclic nucleoside of indeno[1,2-b]pyridine derivative 62 was accomplished using MWI treatment of 54 with 1,3-indeno-1,3-dione in presence of ammonium acetate (Scheme 13). [18, , ]

Scheme 7. Construction of cyclic and homocyclic C-nucleosides 32 from carbonyl functionality on C1 of precursors

Scheme 8. Syntheses of C-nucleosides 36-38

Scheme 9. Synthetic steps for the preparations of acyclic C-nucleosides 43 from the reactions of heterocyclic hydrazines with aldoses

Scheme 10. Synthetic approach to acyclic C-nucleosides 46 from hydrazides

Scheme 11. Synthetic approach to acyclic C-nucleosides 49 from aniline derivatives and aldoses

Scheme 12. Synthesis of acyclic C-nucleoside 53 from aldoses via multicomponent reaction (MCR)

Scheme 13. Synthetic approachs to acyclic C-nucleosides 56, 59 and 62

3. Direct Coupling between a Suitably Protected Carbohydrate Moiety with a Preformed Heterocyclic Compound

The direct attachment of a preformed glycon unit to an appropriate carbohydrate is one of the most important method for synthesis of C-nucleoside. This strategy is based on ionic, free radical or metal mediated carbon-carbon bond formation.

Synthesis of nucleoside derivatives of quinuclidin-3-one 63 was achieved by the reaction of 63 with D-glucose in presence of catalytic amount of zinc chloride to give the cyclic C-nucleoside analogue 64a-d. This reaction involves nucleophilic dislpacemnt of the anomeric hydroxyl group which activated by Lewis acid catalysis. On the other hand condensation of 63 with D-glucose in sodium carbonate as basic catalyst instead of Lewis acid afforded the bis- quinclidine sugar 65. In addition the Mannich base analogue 66 was synthesized by treatment of 63 with glucose and morpholine under Mannich reaction condition( Scheme 14). [19, ].

Scheme 14. Construction of acyclic C-nucleosides 64-66

Coupling of aryl or hetroaryl lithium reactants to protected sugar lactone gave hemiketal intermediate, which subsequent steroselective reduction afforded the desired C-nucleoside with high β stereoselectivity. Thus, lithiating bromoheterocycle 67 using n-butyl lithium at -78 ^oC afforded the lithio species in situ which reacted with the protected lactone 68 to generate the hemiaketal 69. Subsequent anomeric reduction using triethylsilane and boron trifluoride etherate furnished 70. Metobo et. al. explain the steroselectivity to β -anomer on the basis that the chelation of silicon to either 2^\- or 3^\-benzyl ether oxygen electron pair results in delivery of hydride anion from a favored face to furnish only the β-anomer or almost exclusively (Scheme 15). [20, ].

Scheme 15. Reaction of heteroaryl lithium with sugar lactone to prepare 71

C- ribononucleoside that bear amino pyrimidine 78a or amino pyridine 78b have been synthesized using the same synthetic approach, but it was found that protection of the amino group with benzyl bromide are more suitable for coupling with the sugar component. Also it was found that the dehydroxylation of hemiketal intermediate 74a-b with Et₃SiH in presence of strong Lewis acid led to removal of the 2,3-O-isopropylidine group affording the corresponding C-nucleoside (Scheme 16). [21, ].

Scheme 16. Reaction of heteroaryl lithium with sugar lactone to prepare C-nucleosides 76

The synthesis of 5-bromopyridin-2-yl and 6-bromopyri -din-3-yl C-ribonucleosides was based on the regioselective lithiation of 2,5-dibromopyridine 79. In toluene the lithiation proceeded at position 2 leading to 5-bromo-2-lithiopyridine, whereas in Et₂O the lithiation took place at position 5 to furnish 2-bromo-5-lithiopyridine (Scheme 17).

Scheme 17. Synthetic approachs to C-nucleosides 84 and 87 from 2,4-dibromopyridine

4. Construction of the Spacer on the Base

A stereo-controlled synthesis via allylic substitution and ring closing metathesis sequence was reported to construct the sugar moiety of some aryl C-Nucleosde 108a-c. Where as the reaction of the enantiopure alcohol 88 with the enantiopure branched carbonated compounds 89a,b in presence of Ir(I) catalyst [23, , ] afforded the respective enantiopure products 90-93 in good yields (Scheme 18). Ring closing metathesis of the bisallyl ethers 90-93 in presence of a catalyst led to the formation of the desired 1,5-dihydrofuran derivatives 94a,b-97a,b which were deprotected by treatment with Et₃N.HF to afford the corresponding alcohols 98a,b-101a,b (Scheme 19). Also vicinal dihydroxylation of the isolated cis- dihydrofuran derivatives 94a,b and 97a,b using either osmium tetroxide or ruthenium tetroxide as catalysts were performed to obtain compounds 102a,b and 103a,b in excellent yield.

Scheme 18. Preparation of bisallyl ethers 90-93

Scheme 19. Synthetic approaches to C-nucleosides 102 and 103 via ring closing metathesis sequence of bisallyl ethers

Esterification of the tetrafluorinated alcohol 104 followed by cyclization and dehydroxylation was performed to yield the aryl C-nucleoside derivatives 108a-c (Scheme 20). [24, ] Thus, esterification of 104 using either acyl chlorides in CH₂Cl₂ in presence of Et₃N and 4-dimethylaminopyridine (DMAP), or carboxylic acids, which was performed in CH₂Cl₂ in presence of DCC and DMAP gave the desired esters 105a-c in good yields. The obtained esters are then cyclised using MeLi in THF to obtain the cyclised lactols 106a-c which underwent reductive dehydroxylation to give the protected derivatives 107a-c followed by debenzylation using either NaI in Me₃SiCl and MeCN or BBr₃ in CH₂Cl₂ to obtain aryl C-nucleoside derivatives 108a-c.

Scheme 20. Synthetic approach to fluorinated C-nucleosides 108

5. Transformation of the Existing C-nucleosides to Another Ones

Commonly transformation of the C-nucleosides to other ones is carried out to improve the biological activity of the existing C-nucleosides without breaking the C-C bond between both the sugar and the base moieties.

Modification of the existing C-nucleosides occurred either in the sugar or base parts of the existing natural or synthetic C-nucleosides.

5.1. Modifications on the Base Moiety

Modifications on the heteroaryl (aryl)-C-nucleosides using organometallic catalysis were reported. In 2009 Hocek et. al. described the aminocarbolylation of bromophenyl- C-ribonucleoside 109 with CO (1 atm) with different amines in presence of lead acetate, xantphos and potassium phosphate in toluene to give the corresponding amides 110a-c (Scheme 21, Table 1). [25, ] Deprotection of the silylated nucleosides 110a-c using Et₃N.3HF at 40°C for two days followed by treatment with K₂CO₃ led to the formation of the desired free C-ribonucleosides 111a-c. Benzamide-C-ribonucleosides were found to have a biological character as to be a strong cytostatic agent including apoptosis in cancer cells.

The unprotected 5-bromofuran-C-nucleoside 112 was used as a potential intermediate for aqueous phase cross coupling reactions. The Suzuki-Miyaura cross coupling reaction of this compound with different aryl boronic acids carried out in presence of lead acetate,tris(3-sulfophenyl)phosphine trisodium salt (TPPTS) ligand and Cs₂CO₃ as a base for 4h at 120°C yielded the biaryl β-C-nucleosides 113a-g (Scheme 22). [26, ] These nucleosides have high fluorescence properties that can be used as fluorescent labeling in biomolecules.

5.2. Modifications in the Spacer Part

The conversion of compounds 114 and 115 into the hitherto unknown 2’,3’-dideoxy- and2’,3’-dideoxy-2’,3’-didehydro- C-nucleosides 116, 117, 118 and 119 was performed (Scheme 23) [27, , , ]. These compounds were evaluated in a cell-based assay against Human immunodeficiency 1 (HIV-1). Protection of the 5’-hydroxyl group in 114 and 115 using tert-butyldimethylsilyl chloride and imidazole in anhydrous dimethylformamide or pyridine gave 116 and 117 respectively. The 4-thione product was converted to the 4- methylthio derivative 118 using methyl iodide in aqueous sodium hydroxide. Compounds 116 and 118 were then converted into 1,3-dioxalane-2-thione derivatives 119 and 120 followed by heating in presence of triethylphosphite to yield 121 and 122. Deprotection of these products in presence of tetrabutylammonium fluoride gave the 2’,3’-dideoxy-2’,3’-didehydro-C-nucleosides 123 and 124. Compound 124 reacts with ammonia in methanol at 120ºC under microwave to yield the amino derivative 125. Finally, hydrogenation of 123 and 125 using catalytic amount of Pd/C in ethanol yielded 2’,3’-dideoxy-C-nucleosides 126 and 127.

Scheme 21. Chemical reactions of modification to C-nucleosides 111

Table 1. Aminocarbonylation of 3-bromophenyl-C-ribofuranoside

Scheme 22. Suzuki-Miyaura cross coupling reaction to prepare C-nucleosides 113

Scheme 23. Synthetic modification of C-nucleosides 114-115 to 126-127

In order to incorporate s⁴ψiC into triplex forming oligonucleotides [28, ], S.-Q. Cao et al synthesized 3’-phosphoramidite derivative 131 starting from 128 (Scheme 24). There 128 was treated with 1.0 M tetrabutylammonium fluoride in THF followed by protecting the 5’-OH group using DMTr to give 129 in good yield. The 2’-OH protected derivative 130 was obtained by action of TBDMSCl on 129. Finally the reaction of 120 with NCCH₂CH₂OPClN(i-Pr)₂ in presence of 2,4,6-collidine and N-methylimidazole gave the phosphoramidite 131.

Scheme 24. Synthetic modification of C-nucleoside 128 to 131

5.3. Cyclization of the Acyclic Analogue

Transformation of the di-O-isopropylidine acyclic C-nucleoside compound 132 to its O-mesyl derivative 133 was performed [29, ] by addition of methansulfonyl chloride in pyridine at 0°C for 3 h, followed by deprotection and cyclization to yield the desired cyclic C-nucleoside 134 in good yield by refluxing compound 133 using 4% HCl in 1,2-dimethoxyethane (Scheme 25). The selective cyclization (C-4’-C’-1) and the configuration were confirmed using ¹³C-NMR and the nuclear overhauser effect (NOE) experiment. Similarly, the isopropylidine derivative 135 was transformed to the cyclic compound 137 under the same conditions through the formation of the tri-O-mesylated derivative 136 (Scheme 26).

Scheme 25. Ring closure steps of acyclic C-nucleoside 132 to the cyclic 134

Scheme 26. Ring closure steps of acyclic C-nucleoside 135 to the cyclic 137

The reaction of the acyclic C-nucleoside 138 with 2,4,6-tri-isopropylbenzenesulfonyl chloride (TIBSCl) in pyridine gave the homo-C-nucleoside derivative 4-(2,5-anhydro-D-manno-pentitol-1-yl)-2-pheny-2H-1,2,3-triazole (Scheme 27) 139 [30, ], where region-selectivity will occur due to the bulkiness of (TIBSCl). This dehydrative cyclization under basic conditions also yields a minor thermodynamically product 140 as a result of the 1,5-S_N² cyclization. During this reaction the intermediate 141 and the byproduct 142 were isolated from the reaction mixture.

Scheme 27. Ring closure steps of acyclic C-nucleoside 138 to the cyclic 139-140

Synthesis of 4-(5-deoxy-α- and β-L-arabinofuranosyl) -2-phenyl-2H-1,2,3-triazoles 144 and 145 were performed [31, ] and the anomeric configuration were determined by the CD and NMR spectroscopic measurements (Scheme 28). Acyclic 5-deoxy-L-manno-pentitol-1-yl nucleoside 143 was treated with methanolic sulfuric acid and subsequent reflux with copper sulphate yield the anomeric mixture of 144 and 145, which were separated by chromatography.

Scheme 28. Ring closure steps of acyclic C-nucleoside 143 to the cyclic 144-145

Similarly, the dehydrative cyclization of 4-(D-galacto-pentitol-1-yl)-2-phenyl-2H-1,2,3-triazole 146 catalyzed by acid gave the anomeric α,β-D-lyxopyranosyl and furanosyl derivatives 147-150 (Scheme29). [32, ]

Scheme 29. Dehydrative cyclization of acyclic C-nucleoside 146 to 147-150

6. Miscellaneous Synthesis

The homo-coupled product 152 was obtained from the reaction of THF-alkyne 151 with different catalytic systems in dry DMF in order to obtain the optimized reaction conditions such as i) CuI, CsCO₃ and NaI, ii) CuI and CsCO₃ and iii)CuI only, to give the desired product in good yields. It was found that both CsCO₃ and NaI have no effect in the reaction (Scheme 30). [33, ]

The reaction of butenonyl C-glycosides with cyanoacetamide via Micheal addition followed by dehydrative cyclization and oxidative aromatization gave the corresponding glycosyl pyridines. These compounds are evaluated for their antidiabetic potential in vitro. (E)-1-(β-D-glucopyranosyl)-4-(aryl)but-3-en-2-ones 153a-c reacted with cyanoacetamide in DMSO, t-BuOK, N₂ and O₂ at room temperature followed by acetylation to give the respective 3-cyano-4-(aryl)-6-[(2'',3'',4'',6''-tetra -O-acetyl - β-D-glucopyranosyl)methyl]pyridones 154a-c in good yields. Subsequently, deacetylation of these compounds with NaOMe/MeOH afforded the corresponding 3-cyano-4- (aryl)-6-[(β-D-glucopyranosyl) methyl]pyridones 155a-c (Scheme 31). [34, ]

Scheme 30. C-C coupling reaction of 151 to C-nucleoside 152

Scheme 31. Synthetic steps to build up the heterocyclic base on 153-155

Figure 2. Distribution of C-nucleosides synthetic tactics by countries

7. Conclusions

This work highlights a recent progress in developmental strategies to build up C-nucleosides and analogues. It does not attempt to be comprehensive in particular, it intended to spot locations of work and synthetic tactics (Figure 2), then correlate that to international vs. national transformation of methodologies. From the Middle East, choosing Egypt as a model, it is clear that the internal transformation of research perspectives there dominates. However, science diplomacy may provide a forum for international collaborative research, which leads to intriguing topics of potential applications for sustainable developments

Dedicated to the memory of Prof. Hans W. Zimmer¹

Notes

¹Prof. Hans Willi Zimmer (1921-2001) had served at University of Cincinnati (Ohio, USA) for 47 years and arranged for a student exchange program with the University of Stuttgart (Germany) in the early eighties. In keeping with the spirit of international exchange, which Hans initiated, Hans and Marlies Zimmer International Scholar Fund was endowed in his memory.