SYNTHETIC ANALOGS OF NATURAL GLYCOSIDES …structural information storage, related to biological function, looks incomparably more complicated, that previously defined fields of functional

Acta Poloniae Pharmaceutica ñ Drug Research, Vol. 65 No. 6 pp. 655ñ676, 2008 ISSN 0001-6837Polish Pharmaceutical Society

Sugars are extensively engaged in nearly allaspects of life processes, from primary energetics,through post-translational modification of function-al proteins, to production of structural elements ofthe cell and generating signalling molecules.Glycosylation is quite common biochemical trans-formation, performed by more than 80 families ofglycosyl transferases (GT). Considering estimatedgigatonns of UDPG per month used by terrestrialplants in order to build up cellulose, starch and otherpolysaccharides globally, glucose transfer alonestands out as the largest biotech operation run on theplanet. Plants synthesize well over 100 000 naturalproducts, and considerable proportion of them areglycosylated. Glycobiology describes the role ofsugar oligomers (glycans) in variety of intra- andintermolecular events, which involve molecularrecognition of carbohydrates (1, 2). Unlikeaminoacids and nucleotides ñ the building blocs,

which tend to produce linear biopolimers upon con-densation, monosaccharides offer much more roomfor structural diversity, leading to branched struc-tures with great variety of shapes, depending ontypes of interglycosidic linkages and monomericconstituents involved. Therefore, from the presentpoint of view, glycomics perceived as a system ofstructural information storage, related to biologicalfunction, looks incomparably more complicated,that previously defined fields of functional biology:genomics and proteomics. Although bioinformaticshas not reached the level, at which rational miningof sugar structural diversity becomes viable, evensuperficial knowledge of natural products can offeruseful hints concerning distribution and biologicalactivity of small molecular weight compounds ofglycosidic structure isolated from bacteria, fungiand higher plants, which found application as drugsin traditional medicine. More often than not, these

SYNTHETIC ANALOGS OF NATURAL GLYCOSIDES IN DRUG

DISCOVERY AND DEVELOPMENT

GRZEGORZ GRYNKIEWICZ1*, WIES£AW SZEJA2, and JERZY BORYSKI3

1Pharmaceutical Research Institute, 8 Rydygiera St., 01-793 Warsaw, Poland,2Chemistry Department, Silesian Technical University, 4 Krzywoustego St. 4, 44-100 Gliwice, Poland,

3Institute of Bioorganic Chemistry, Polish Academy of Science, 12/14 Z. Noskowskiego St., 61-704 PoznaÒ, Poland

Abstract: Secondary metabolites, which have vital environmental and allelopathic functions for a host, andlong tradition of ethnopharmacological applications preceding modern medicinal use, often occur in their nativestate as glycosides. The role of sugar moiety looks completely different from plant physiology point of viewand from drug discovery and development perspective. Based on a short survey of cases, in which structuralmodification of natural glycone (saccharide part of a low molecular weight secondary metabolite) resulted inadvantageous pharmacological changes, we postulate that glycosides of natural origin can be quite promisingas drug leads, based on general rules of drug design. In particular, polyfunctional sugar moieties offer ampleopportunities for almost continuous changes in shape, electron density and polarity. By the same token, glyco-sylation of other biologically active natural products, which are not natively glycosylated, can be viewed as atool for tune up of their activity in direction of higher efficacy and better selectivity. Despite of considerableadvances towards turning enzymatic glycosylations into biotechnological processes, chemical transformationsstill remain more practical, particularly for synthesis of modified glycosides, both: in research laboratory andin industry.

Keywords: natural products, secondary metabolites, naturally occurring glycosides, glycosidic drugs, chemicalglycosylation, modification of sugar moietiesAbbreviations: BA ñ bioavailability, CNS ñ central nervous system, CYP ñ cytochrome P450, DNA ñdeoxyribonucleic acid, GT ñ glycosyl transferases, RNA ñ ribonucleic acid, SAR ñ structure-activity relation-ship, SM(s) ñ secondary metabolite(s), UDPG ñ uridine diphosphoglucose, LG ñ leaving group, IDCP ñ iodo-nium dicollidine perchlorate.

655

Correspondence: e-mail: [email protected]

656 GRZEGORZ GRYNKIEWICZ et al.

natural glycosides rooted in ethnopharmacologicaltradition, pawed the way to the next generations ofsemisynthetic medicines, used in modern therapy.Nature provides unsurpassed structural versatility inmany categories of secondary metabolites, whichbecame valuable lead compounds for drug discoveryand development, often inaccessible by convention-al chemical synthesis. It is estimated that more thana half of modern drug substances originated in thenatural products pool (3-5). It is likely, that most ofthese valuable compounds will be eventually obtain-able by biotechnological processes as a result ofadvances in functional plant genomics. Meanwhile,the window of opportunity for chemical synthesis,applied for rationally designed structural modifica-tion of the natural lead compounds, seems to remainwide open. In particular, chemical glycosylation,although intrinsically difficult when applied to com-plex and multifunctional natural products, can beperceived as type of derivatization, which allows fora fine tune up of various molecular properties (suchas overall size, polarity, local electron density,lipophilicity, membrane permeability, affinity toactive transport systems, biodegradability, etc.),within a single glycosylation protocol. This shortsurvey, focused on selected categories of plant sec-ondary metabolites, which are known to occur inglycosylated state, is intended for illustration of thethesis that a switch from the natural glycoside to asemisynthetic construct, comprising a structurallymodified glycone, can be very useful for therapeuticpurposes (resulting for example, in lower toxicityand/or better selectivity of the modified conjugate).In our opinion, chemical glycosylation can serve asa technical platform for rational structural modifica-tion and as an useful tool in medicinal chemistry,particularly for more effective exploitation of chem-ical structure diversity offered by Nature.

Biological significance of secondary metabolism

Plant genome is surprisingly large in compari-son with that of animals. The model organism,Arabidopsis thaliana (for which reasonable bio-chemical pathway database exists) has ca. 130 kbp,while wheat or onion are known to contain almosttwo orders of magnitude more base pairs. This canbe interpreted in terms of need for stress adaptationreaction in form of chemical response, in absence ofthe flight option. Plants have encoded during theirevolution a great variety of gene clusters responsiblefor synthesis of compounds which are not essentialfor survival but offer some environmental advantageto the host (6). These compounds, collectively clas-sified as secondary metabolites (SMs), often func-

tion as deterrents to herbivory and although biocom-patible in general, sometimes exhibit significant tox-icity against animals, including humans.Nevertheless, many of these SMs, initially foreign tohuman physiology, have been adopted over milleniaby ethnopharmacology for treatment of variousalignments, often because selective toxicity exhibit-ed by corresponding plant herbal preparations,helped fighting infective pathogens or parasites. Ingeneral, these non-nutritional constituents of plants(or microbial organisms) are processed metabolical-ly after human ingestion as typical xenobiotics,undergoing standard phase I (e.g. oxidative activa-tion by CYP) and phase II (conjugation leading toexcretion) processes, which make them sub-optimalas drug leads or drug candidates, despite of pro-nounced biological activity in model systems.

Secondary metabolites have evolved biogeneti-cally on various enzymatic pathways, utilizing suchprincipal metabolic constituents as fatty acids,aminoacids, shikimate or mevalonate units, iso-prenoids, etc., to produce a great variety of struc-tures, traditionally categorized as: alkaloids, antibi-otics, flavonoids, terpenoids, etc. (7-9). Many ofthem undergo glycosylation, by action of glycosyl-transferases, in order to facilitate their cellular trans-port and compartmentalization, but stability of gly-cosides in biological material is often quite limited,sometime leading to a false impression that agly-cones are primary constituents of a biological matrix.While glycosylated SMs are mobilized as a local orsystemic chemical defense system of a host organ-ism, their rationally designed synthetic analogs can,in principle, serve a variety of functions, in humanmedicine. Sugar moiety of such glycosides is anideal target for structural manipulations, because itsmultifunctionality offers wide opportunities forchanges in both: topology and charge/electron densi-ty distribution. In some cases intact natural glycosidestructure constitute a proper substrate for desiredchemical derivatization. However, quite often de-conjugation of such SMs (either enzymatic or chem-ical) leads to inactive aglycone and inactive sugar.Such cases (anthracycline antibiotics, digitalis gly-cosides, antiviral nucleosides, etc.) are particularlyinteresting for study of the sugar moiety role andfunction (10-12). Often, disconnection and reassem-bly of the glycosidic bond is the simplest way to alab scale derivatization. However, such cases canpose a serious operational challenge, particularly forscale up and process development. Therefore, acomment on glycosylation as a limiting step, sup-plemented by examples from contemporary litera-ture, is enclosed (13).

Synthesis analogs of natural glycosides in drug... 657

Scope of chemical glycosylations as applied to

natural products

There are great many synthetic protocols forconjugating sugars with alcohols and phenols bychemical means (14-19), but unlike enzymatic cou-plings, which are high yielding, stereo- andregiospecific one step transformations, they requireprotection of sugar donor hydroxyl groups, carefulchoice of anomeric leaving group and design of acompatible activation system. The glycosidic bondis most commonly formed by a nucleophilic dis-placement of the leaving group (LG) of the glycosyldonor with the hydroxyl or thiol moiety of the nucle-ophile (glycosyl acceptor). These reactions are usu-ally performed in the presence of activator, promot-er or catalyst. The glycosidic linkage adds a chiralring element and the two diastereofacial isomers, α-and β-glycosides, can be formed. The stereochemi-cal outcome of the reaction is influenced by a rangeof factors, including steric effect, character of pro-tecting group, promoter, leaving group, and reactionconditions. For many years in the past, anomericbromides, chlorides (20) hemiacetals and acetates(21) were used as glycosyl donors for the synthesisof oligosaccharides and complex glycosides. Duringlast 30 years of the past century, new classes ofdonors were developed (19). Nowadays,

trichloroacetimidates (16), thioglycosides (17), fluo-rides (20), glycosyl xanthates and thioimidates (22)n-pentenyl glycosides (23) and glycals (24, 25) havealso become common glycosyl donors.

In most glycosidation reactions, the resultinganomeric stereochemistry is controlled by the natureof the C-2 substituent in the sugar moiety. Thus 1,2-trans glycosidic linkage can be stereoselectivelyformed with the use of participating group at C-2such as O-acetyl (Ac), O-benzoyl ( Bz), N-phthal-imido (Phth) (14).

There are a number of papers on influence ofprotecting and leaving group on the 1,2-cis stereose-lectivity (19). When the C-2 oxygen is protected withan alkyl or benzyl group, the anomeric effect domi-nates and the α-anomer is preferentially formed (19).Milder reaction conditions are beneficial for 1,2-cisglycosylation. Thus, halide ion catalyzed reaction ofglycosyl halides is superior for the synthesis of 1,2-cis glucosides, galactosides, and fucosides (26).Thioglycosides are favorably activated with iodo-nium dicollidine perchlorate (IDCP) (27). When gly-cosylation proceeds via bimolecular SN2 mechanism,glycosyl donors with 1,2-trans orientation form 1,2-cis glycosides. For example, α-mannosyl chlorideswith insoluble catalyst (28), or α-mannosyl triflates(29, 30) are used for β-mannosylation. Since in most

Scheme 1. Synthesis of quercetin 3-O-β-D-glucuronide.


Scheme 2. Synthesis of glucoglycerolipid.

cases the reaction proceeds according to SN1 mecha-nism (intermediate carbooxonic cation is formed) theorientation of the leaving group is of lesser impor-tance (31, 32). Although variation of reaction condi-tions and structure of glycosyl donor or glycosylacceptor may occasionally lead to high 1,2-cis stere-oselectivity, no efficient general method for 1,2-cisglycosylation has been elaborated yet.

Abundance of natural glycosides inspired con-siderable synthetic efforts towards complex glyco-sides and their analogs (33, 34). Some selectedexamples are presented below, to illustrate typicalapproaches by which various glyconjugates can beobtained, even from complex, multifunctional agly-cones. A number of natural products have been pre-pared by the use of the O-glycosylation reactionwith glycosyl halides. The earliest known glycosy-lation method is that of Koenigs and Knorr. Thisreaction involves the coupling of a glycosyl bromideor chloride with a hydroxyl component upon activa-tion of the former with a heavy metal ion, typicallysilver or mercury (35). Extensions of theKoenigsñKnorr conditions include the use of Lewisacids and phase-transfer catalysis (26) to activate theanomeric halides. Several authors have reported theglycosylation of flavonoid derivatives, sinceflavonoid glycosides play a variety of essential rolesin the growth and development of plants (36). A gly-cosyl derivative of the secondary plant metabolite ñ

quercetin, a very efficient antioxidant, was preparedby Rolando (37) via standard glycosylation, fol-lowed by selective catalytic oxidation, providing thedesired glucuronide, as depicted in Scheme 1.

In order to investigate alterations of the glyco-lipid composition due to HTLV-I virus (human T-lymphotropic virus type I) infections, a new gluco-syl glycerolipid with a 6-O-phosphocholine groupwas isolated from the culture of HTLV-I-infectedhuman helper T-cells, and synthesized by coupling aglucopyranosyl bromide with (S)-glycidol (Scheme2) (38). The mild glycosylation was carried outaccording to Lemieux halide methodology, asshown below (26).

Trichloroimidate-mediated glycosylation wasdesigned and developed by Schmidt (16) as an alter-native useful method to Koenigs-Knorr glycosyla-tion. Some advantages of this method include mildreaction conditions, thermal stability and ease ofpreparation of the trichloroacetimidate, usually goodchemical yield, good stereocontrol and lack of effecton other glycosidic bonds. Therefore, synthesis ofnumerous disaccharides, key subunits of biological-ly potent oligosaccharides such as antigens, antibi-otics, glycoproteins, and glycolipids has been per-formed using glycosyl imidates, as exemplified bypreparation of biologically potent α-1,2-linked di-saccharide derivatives via regioselective one-potprotectionñglycosylation (39) .


Scheme 3. Synthesis of docetaxel 7-α-D-glucopyranoside.

Paclitaxel (taxol) is considered an exceptional-ly effctive anticancer drug of natural origin. One ofthe major drawbacks of this drug is its extremelylow aqueous solubility. In order to prepare taxaneglycoconjugates, similar to those found in nature,which would be more soluble in water than pacli-taxel, Zamir et al. (40) performed the semisynthesisof an O-glycosylated docetaxel analogue by reactionof the natural taxane with protected imidate esterderivative according to Scheme 3.

Thioglycosides are amongst the most widelyused glycosyl donors (17). Their popularity is partlydue to their facile synthesis, stability and their easyconversion into glycosides. A number of glycosyla-tions with thioglycosides have constituted key stepsin the syntheses of various glycolipids. The thiogly-coside methodology was applied by van Boom (41)to the synthesis of naturally occurring biologicallyinteresting disaccharide rhamnolipids as shown inScheme 4.

1,2-Unsaturated sugars (glycals) are versatilesynthetic intermediates, often used for preparationof 2-deoxyglycosides (42). The most direct route to2-deoxyglycosides is by acid-promoted addition ofalcohols to a glycal, although care must be taken toprevent acid-catalyzed Ferrier-rearrangement. Thestereochemical outcome of the reaction is influencedby a range of factors including the conformation of

glycal, steric factors in both the glycal and nucle-ophile, the nature of electrophilic species, the sol-vent and the temperature.

Small molecule drugs and drug candidates result-

ing from structural modification of a natural gly-

cone

Alkaloids

Class of alkaloids, as compounds distinctly dif-ferent from other plant constituents, because of theiralkaline character, gained recognition at the begin-ning of XIX century with discovery of morphine,and have grown to ca. 20 thousands member grouptoday. They are formed predominantly in higherplants, by multistep transformations involvingamino acids and biogenic amines, and are consid-ered to function as phytoalexins. This category ofSM is not commonly glycosylated, perhaps becausein order to render them soluble it is enough, inmajority of cases, to combine them with ubiquitousorganic acids, instead of forming energeticallyexpensive covalent glycosidic bond. Althoughreports on glycosylated quinolizidines, ergot deriva-tives, betalaines and indoles of plant origin can befound in journals devoted to plant physiology, thebest known group of alkaloids natively combinedwith sugars are cholesterol metabolites ofSolanaceae, like solanine 1 and tomatine 2 (Fig. 1),


Scheme 4. Synthesis of disaccharide rhamnolipid.

similar in origin and character to saponins. Theytypically contain branched oligosaccharide com-posed of D-glucose, D-galactose, D-xylose, and L-rhamnose units, linked to the alicyclic secondaryalcohol. These compounds are toxic to mammalsand exhibit pronounced antifungal activity. So far,they did not found medicinal applications, despite oftheir occurence in cultivated plants of Lycopersicumand Solanum families (9).

Staurosporine 3 and rebeccamycin 4 (Fig. 2)are good examples of bis-indole N-glycosidic natu-ral products (isolated from Streptomyces, hence clas-sified also as antibiotics), which evoked more inter-est in terms of strive for semisynthesis of analogs.Staurosporine was found to be a potent inhibitor ofprotein kinases and therefore, was considered as alead compound for anticancer application. Itsanalogs were synthesized, among others, fromrebeccamycin, itself a topoisomerase inhibitor (43).

Colchicine can be considered a special case ofan alkaloid without basic nitrogen atom, besides itcontains rather unusual tropolone ring system, but itshould be mentioned here because of interestingantimitotic action, based on its ability to bind totubulin and interfere with microtubule assembly.While colchicine itself is appliedd in experimentalplant genetics and as veterinary antineoplastic, itssemisynthetic glycoside: thiocolchicoside 5 (Fig. 2),is used for gout treatment and as a muscle relaxant(9). In this case, glycosidation is clearly chosen as amean for toxicity remediation. In connection withthis motive, it should be pointed out that metabolismof alkaloids can also provide useful hints for explo-ration of SM structural diversity. Morphine, forexample, is metabolized to 3-O-glucuronide and/or6-O-glucuronide in mammals, but only to the sec-ond of these metabolites in man, which could inspireresearch on long acting prodrug conjugates. It


seems, however, that alkaloids have been relativelyseldom a subject of chemical glycosylation with aimto study biological activity of resulting conjugate. Inour opinion, it can be expected, that many interest-ing medicinal chemistry hits could be obtained onsuch route by intensified research in the field.

Antibiotics

This category of SM, which bears clear medic-inal connotation, is often of microbial origin, andcomprises enormous structural variety. Amongantibiotics, which constitutively contain sugar moi-eties, the following groups can be listed: aver-mectins, lincomycin derivatives, polyenes (nystatin,amphotericin), aminoglycosides (streptomycin, gen-tamycin), glycopeptides (vancomycin, bleomycin),enediynes (calicheamycin, neocarzinostatin) andmacrolides (erythromycin, azithromycin). In each ofthese groups, much effort has been devoted to chem-ical derivatization and semisynthesis, in order toobtain more active and more selectively actingantimicrobial and antifungal agents. In many casesswitch, or structural modification of a glycone,resulted in a new pharmaceutical active substance(44, 45). In order to discuss some conceptual groundfor such modification, we will bring up examples

from the field of anthracycline antitumor antibiotics,in which we share some personal experience.Daunorubicin 6 and doxorubicin 7 (Fig. 3), toxicpigments from Streptomyces, are in clinical use fornearly four decades and the second from the twonatural drugs became one of the principal con-stituents of the multidrug chemotherapy schemes,typical for contemporary clinical oncology (46)Nevertheless, the quest for ìbetter doxorubicinîcontinues all the time, and thousands of new syn-thetic derivatives have been obtained and tested,throughout the years (47-49).

Although DNA intercalation was indicatedfrom the beginning as the principal mechanism ofcytotoxicity, alternative modes of action, in particu-lar inhibition of topoisomerase II and generation ofreactive oxygen species through free radical mecha-nism, were also considered. However, in retrospec-tive examination of the extensive synthetic effort inpreparation of new dauno/doxo analogs, the line ofdefinite molecular targeting can hardly be traced.Since molecules of natural anthracyclines are veryreactive and susceptible to destruction by variety ofchemical reagents (electrophilic, nucleophilic, oxi-dating, reductive, etc.), search for new derivativeswas severely limited to ìdoable chemistryî per-

Figure 1. α-Solanine (1) and α-tomatine (2)


formed on prototype antibiotics, at least during ini-tial period. Meanwhile, the central dogma of anthra-cycline-DNA interaction, stating that 3-aminosugar,L-daunosamine, is necessary component of the bina-ry complex, because it stabilizes minor grove bind-ing through electrostatic interaction, was contestedby W. Priebe from UT MD Anderson Cancer Centerin Houston, TX, USA (50, 51). Priebe with co-work-ers executed radical structural changes in the gly-cone part of anthracyclines, using glycoside discon-nection approach, which relied on steady aglyconesupply, in house synthesis of sugar moieties, andmastering chemical glycosylation processes.Meanwhile, interactions of partial structures withappropriate DNA fragments were measured andmodeled, resulting in energy estimates with partitionto incremental contributions from molecular frag-

ments (52). This seminal work paved way to a seriesof highly innovative new drug candidates, which arenow in clinical trials. Annamycin 8 (Fig. 3) is per-haps the most spectacular example of such innova-tion, worth discussing in some detail. The followingradical changes have been introduced (all in the gly-cone part), in comparison to the daunosamine moi-ety in prototype antibiotic doxorubicin: a) 2í-iodosubstituent replaced 2-deoxy group; b) 3í-hydroxyfunction replaced amino group, with retention ofconfiguration; c) configuration of the 4í-OH groupwas inverted. Commentaries to these changes are asfollows:

Original 2-deoxyglycosidic bond in antibioticsis relatively easy to break, both: enzymatically andchemically. Introduction of iodine atom at C-2 addslipophilicity and strengthens the glycosidic bond, by

Figure 2. Staurosporine (3), rebeccamycin (4) and thiocolchico-

side (5)

Figure 3. Daunomycin (6), doxorubicin (7), annamycin (8) and

amrubicin (9)


estimated factor 103 (both changes consideredadvantageous from ìdruglikenessî point of view).Position of the halogen atom is highly relevant forbiological activity ñ only compounds with 2í-axialsubstituents demonstrate efficacy in cytotoxicitytests. Removal of the amino group from sugar C-3represents a revolutionary change, from the point ofview of early SAR of anthracycline antibiotics,which stressed the need for 3-deoxy-3-aminopyra-nose with equatorial NH2 group as a DNA minorgrove anchoring element. This structural changeappeared in the research program as a result ofreducing basicity campaign, following observationthat antibiotics with more basic sugar moiety arebetter substrates of Pgp efflux pumps, reducingeffective intracellular drug concentration (53). Thelast change (C-4í configuration) results from anearly observation that L-lyxo- to L-arabino- config-urational switch does not have detrimental effect onantitumor activity, as clearly demonstrated on theepirubicin example. Another synthetic anthracyclinewithout natural L-daunosamine moiety, which madeit into the drug level, is amrubicin 9, which providesadditional proof for SAR ideology formulated by W.Priebe.

Antibiotic monosaccharides, both: natural andsynthetic, are subjects to extensive critical reviews

(54, 55). Although antibiotic sugars are repeatedlysubjects of total syntheses, treated as an academicexercise and also as a proof of certain synthetic con-cepts, intermediates in such multistep preparationscan often be treated as valuable intermediates fordisconnection ñ glycosylation approach to newanalogs synthesis. For example, there is a consider-able renewal of interest in precursors of aminosug-ars bearing azido function, because they turned outto be compatible with efficient glycosylation proce-dures (56, 57). Azido analogs of aminosugars arealso suitable precursors for novel glycones bearingN-heterocyclic substituents in place of originalamino group, which are easily obtainable by 1,3-dipolar cycloadditions (58).

Simple phenolics and polyphenols

It is well known that the most popular drugever, aspirin, and its precursor ñ salicylic acid, orig-inate from salicin 10 (Fig. 4), simple phenolic gly-coside, which made Salix plants bark famous aspreparation for common cold and rheumatoid fever.Corresponding salicylic acid β-D-glucoside 11, hasbeen postulated as plant systemic defense agent fordecades, but turned out to be elusive as isolablechemical entity, until synthesized in the laboratory(59). Salicylic acid glucoside 11 could be a viabledrug candidate, except for economic reasons andtechnical difficulties involved in its preparation. Thesame concerns D-glucose acetylsalicylic acidanomeric esters, which require ingenious procedureto prepare (60). Further examples of simple, singlearomatic ring phenol glucosides include: arbutin,coniferin, glucogallin, glucovanillin, phloridzin,populin and syringin. To our knowledge no attemptshave been made to study derivatized or glycone-modified glycoside activity in this group of com-pounds.

Polyphenols are ubiquitous as plant SM andthey apparently are good substrates for glycosyl-transferases, since they occur in glycosylated formas a rule (9). It is quite likely that glycosylationserves not only for intracellular solubilization butalso stabilize this type of aglycons, which are sus-ceptible to various destructive transformations invitro, particularly in solution, apparently initiated bythe presence of oxygen, light, metal ions, etc.Therefore, there is a distinct difference betweenfacility of biogenetic glycosylation of polyphenolicsubstrates, and difficulties of their chemical glyco-sylation, which tends to be low yielding and nonse-lective in most cases. There are well over thousandnatural polyphenol glycosides known, and theflavonoid class is the largest and perhaps the most

Figure 4. Salicin (11), salicylic acid β-D-glucopyranoside (11),and 1-O-β-D-glucopyranose salicylate (12)


interesting group from the point of view of rationalglycone modifications for prospective medicinaluses. Flavonoid glycosylation is an old evolutionarytrait, as evidenced from large accumulation of thesecompounds in Ginkgo biloba, considered a livingfossil among plants. Standardized extract of driedleaves of Ginkgo biloba (used as remedies for vari-ous symptoms since deep antiquity) contains ca.24% of flavonol glycosides! This well defined mate-rial (EGb 761) is a subject of research as therapeuticagent for cognitive decline in senile dementia,including Alzheimerís disease (61).

Flavonoids are significant, non nutritional con-stituents of the human diet for millenia and their rolein maintaining health and well being is a subject ofnumerous monographs (62-64). Flavonol glyco-sides, particularly quercetin derivatives (rutin, hes-peridin) are/can be ingested in gram quantities, sincetheir content in common vegetables (onion, pepper,citrus fruits) is high, but chemistry of this group ofSM is not well developed, despite of continuousinterest started in 1930-ties by A. Szent-Gyˆrgistudies on vitamin C bioavailability, which elevatedcitrus flavonoids to the vitamin level. A notableexception is polyhydroxyethylated rutin(TroxerutinÆ), obtained by action of ethylene oxide(or 2-chloroethanol) on quercetine rutoside, whichbecame a drug in 1950-ties, and is still used as amixture of derivatives, for vein and capillary bloodvessel protection. An example of flavonoid glyco-sides used for pharmaceutical purposes is sylibin, aflavonolignan obtained from seeds of milk thistle(Silibum marianum). Compound 13 (Fig. 5) is apotent detoxicant and hepatoprotectant, whichdespite of its five hydroxyl groups suffers from lowwater solubility, which limits its efficacy. A groupof glycosides bearing monosaccharide or disaccha-ride moieties have been obtained from sylibin bybiotechnological and chemical methods and all ofthem exhibited advantageous features (apart fromexpected rise in solubility) in biological tests oninhibition of lipooxidation and hepatoprotection.Glycosides, containing β-lactosyl, β-maltosyl, β-glucosyl and β-galactosyl residues at the primaryhydroxyl were highly preferred as products of bio-transformation and chemical synthesis alike (65).

Plant isoflavones, which are generated by thesame phenylpropanoid biogenetic pathway, evoke alot of interest and some controversies, because ofepidemiological consequences of dramatic differ-ences in their dietary intake (66-68). Althoughoccurence of isoflavones is quite limited, their over-all consumption is very high, because they are pres-ent in soybean, one of the principal crops of modern

industrial agriculture (producing globally ca. 200million metric tons of beans annually, with an aver-age content of isoflavones 0.15% by weight). Themain isoflavone constituent of soybean are genistin16 (Fig. 6) and its 6-O- esters of acetic and malonicacid, D-glucosides very easily converted during var-ious food processing operations into its aglycone ñgenistein 15. An interest in genistein as biologicalactive substance started in 1957 with discovery of itsinhibitory action on protein kinase C, and reachedafterwards overwhelming proportion, manifested inover 500 papers annually devoted to studies of itsproperties. A list of genistein activities include:antioxidant, antitumor, cardiovascular, estrogenic,and inhibitory towards several enzymes, includingprotein kinases. These findings offer support for thethesis that dietary isoflavones may be the causativeepidemiological factor responsible for significantlylower tumor mortality and morbidity consistentlyobserved within Asian societies using traditionaldiet rich in soy food (69). Consequently, genisteinbecame a drug lead, investigated in number of clin-ical trials with indication in tumor chemopreventionand chemoprotection as well as treatment of hor-mone dependent tumors (in connection with itsestrogenic activity), and osteoporosis (70, 71).

It should be noted that genistein itself exhibitsvery poor characteristics as a drug candidate: thecompound is extremely polar (log P 1.74), practical-ly insoluble in water, is poorly bioavailable, andundergoes extensive conjugative metabolism whichfacilitates its excretion. A question arises, if genistin16, its natural glycoside, is any better, in terms ofhuman pharmacokinetics? To our knowledge, thereare no data available, from such side by side com-parison. Experiments performed on animals suggestthat there is no significant difference betweenbiavailability of genistein and its glycoside, alleged-ly because of efficient deconjugation of the latter byintestinal bacteria. In contemporary nutraceuticalmarket preparations containing genistein prevail butsome OTC market products based on genistin (e.g.Soyfem), gain evidence of clinical efficacy (72).

Figure 5. Sylibin (13) and its glycosides (14)


Obviously, synthetic glycosides of genistein (e.g.17), bearing unnatural and protected glycone moi-eties are also of considerable interest in respect totheir biological activity. In the authors opinion, theycan offer much better scope and perspectives of drugdevelopment, that further examination of genisteinand genistin. In support of this opinion we can pres-ent an evidence that some synthetic glycosides ofgenistein are not just a source of the active aglycone(like genistin), but in some cases show significantlyhigher activity and exhibit individual profiles of bio-logical action. In general, lipophilicity factor wassingled out as promoting cytotoxic and antiprolifer-ative activity. Moreover, in specific cases (17), newmolecular mechanism of action never beforeobserved in flavonoids, was uncovered (73, 74).

Unfortunately, wide screening of numerousglycosylation procedures with genistein as the glyco-syl acceptor failed to afford good yield of expectedproducts and until now only limited cases of suc-cessful sugar conjugation of the isoflavone werereported (75-77). Similar case of advantageous influ-ence of structurally modified glycosides on polyphe-nolic pharmacophores is available from lignan cate-gory. Podophylotoxin 18 (Fig. 7), was a promisingantitumor drug candidate which was disqualified indevelopment on toxicity ground. After two structur-al changes: 4í-de-O-methylation and inversion ofconfiguration at C-4, new intermediate was obtained.Fortunately, its synthetic glycosides 19 and 20, madeit to the market (as etoposide and teniposide anti-cancer drugs). An additional structural element ñacetylidene ring in hexopyranose moiety proved

essential in drug developent phase (78). Like in thepreviously discussed case, glycosylation radicallyaffected the mechanism of action characteristic forthe prototype drug ñ podophylotoxin (79).

Obviously, other groups of natural polypheno-lic glycosides, which have not yet found medicinalchemistry connection, also deserve attention.Human dietary intake of flavonol glycosides, in par-ticular quercetin glycosides can easily reach gramlevel daily, because their significant content in veg-etables, particularly in onion, and also in fruits andbeverages. However, all quercetin glycosides (3-O-,4í-O- and 3,4í-di-O- glucosides) are rather poorlybioavailable, in comparison with the aglycone (80).

Figure 6. Genistein (15), genistin (16) and synthetic derivativeIFG-21 (17)

Figure 7. Podophylotoxin (18), etoposide (19), and teniposide(20)


Supposedly, even partial acylation could changesuch BA characteristics, which is relevant to recentfindings that flavanone, flavone and flavonol glyco-sides exert, (despite of their poor BA) some specif-ic, depressant action, on CNS (81), an activity large-ly overlooked in earlier research. Less common,hydroxylated stilbenes, with their much popularizedrepresentative ñ resveratrol, also occur in plants asD-glucosides and seem to be a relatively easy target,for exploratory glycodiversification. Anthrone gly-cosides are well known folk medicine used as laxa-tives. Anthocyanidins ñ widespread plant pigments,are interestig for at least two reasons: oligosaccha-ride moieties of the peonidine aglycone are natural-ly modified by acylation with caffeyl acid residues;some simple glycosides from this group are potentnatural insecticides (82).

Lipids

Glycolipids are carbohydrate-attached lipids,widely distributed in animals, plants, algae and bac-teria. It is well establish that they are involved in thebiosynthesis of glycoproteins and also serve as bio-logical markers and as ligands for toxins, lectins,bacteria and viruses (83). At the surface of cells,complex carbohydrates are encountered as organ-ized systems, such as glycolipids anchored into thebilayer via their lipid moiety.

Two main classes of glycolipids can be distin-guished: glycosphingolipids and glycosyl glycerides(Fig. 8). In typical glycosyl glycerides two hydroxylgroups of C3 unit are esterified by fatty acids. Theglycosphingolipids cerebrosides (ceramides) andgangliosides are a diverse family of molecules com-posed of one or more sugar residues linked via aglycosidic bond to the sphingoid base.Galactocerebrosides are typically found in neural

tissue, while glucocerebrosides are found in othertissues.

Glycosyl gangliosides are the compounds com-posed of a glycosphingolipid with one or more sial-ic acids linked on the sugar chain. They are compo-nents of the cell plasma membrane which modulatecell signaling transduction events.

It is presumed that compounds with glycolipidstructure have a little chance to become a drug can-didate, because with their physicochemical charac-teristics they will inevitably end up as componentsof lipid bilayer forming membranes. Nevertheless,an idea of glycosidic prodrugs for lipophilic activecompounds still persists, as evidenced by attemptsto glycosylate carotenoids, tocopherols and vitaminD group members (84). It is also possible, that evenshort living glyceroglycosides, can exert some sig-nalling action, since various acylglycerols are activeas second messengers. An example of simple galac-tolipid: GOPO (1-O-β-D-galactosyl glycerol doublyesterified with unsaturated fatty acid), which is clin-ically tested as an antiinflamatory agent, provides an

Figure 8. Structure of natural glycolipids.

Figure 9. Galactosyl cerebroside ( KRN7000).


argument for considering also this class as possiblemediators of intracellular communication (85).

Glycolipids with alpha glycosidic bond haveshown potent anti-tumor and anti-viral activities aswell as potential for the treatment of certain autoim-mune disorders. Detailed mechanistic studies sug-gest this biological activity occurs via Natural KillerT (NKT) cell activation. Since the discovery of O-galactosyl ceramides from marine sponge (86) sev-eral studies of KRN7000 and its derivatives havebeen reported. These investigations have revealedthat tumor-associated cell-surface glycoproteinssuch as CD1d present exogenous lipids to NKT cellscausing the release of chemokines that regulateimmune response to cancer. (87, 89). The potentimmunostimulatory activities of lipid antigens suchas KRN7000 (Fig. 9) have led to the development ofanticancer chemotherapeutics that are currently inclinical trials (90).

Some glycolipids have also found their use intreating neurological diseases. Some attempts todevelop carbohydrate drugs for Parkinsonís diseaseand other neurological indications have been report-ed (91). The development initially focused on themodification of certain glycolipid compounds thathave previously demonstrated clinical promise.Since there is no known prevention or cure for thisdisease and current treatments focus only on con-trolling the symptoms of the disease and do notretard its progression, this would be a significantimprovement.

Gangliosides, a heterogeneous family of gly-cosphingolipids abundant in the brain, have beenshown to affect neuronal plasticity during develop-ment, adulthood and aging. Natural and semisyn-

thetic gangliosides are considered possible therapeu-tics for neurodegenerative disorders. (92).

An interesting example of simple glycolipid,which exhibits pronounced selective biologicalactivity is flocculosin (Fig. 10), natural antifungal ofmicrobial origin (93). The structure of flocculosin,which is an acylated drivative of simple maltoside,strongly supports our thesis that sugar scaffold canplay decisive role in rendering biological activity toa glycoside, even through such simple operation asselective multiple acylation.

Several medical applications of glycolipids arein close relationship to molecular recognition. Thelipo-polysaccharide-anchored liposomes have beenstudied as stable and targetable drug carriers adapt-able in effective chemotherapy, particularly forintroducing chemotherapeutics into target cells ortumor cell lines (94).

Adriamycin bearing liposomes anchored withimmuno-polysaccharide derivative demonstrated invivo targetability against human lung cancer (PC-9grafted) in experimental athymic mice (95).Targeted chemotherapy of brain tumor using poly-saccharide-anchored liposomes loaded with antitu-mor drug cisplatin has been attempted by Ochi et al.(96). Survival of 9L-glioma implanted rats withCHP based liposomes loaded with cis-platinumdiamino-dichloride (cisplatin) was significantlyhigher as compared to average survival recorded foruntreated groups.

Nucleosides

Nucleosides and their synthetic analogs consti-tute the major class of glycosylated compoundswhich to date exhibit pronounced in vitro and in vivo

Figure 10. Structure of flocculosin.


antiviral activity (97-99). The naturally occurringnucleosides are constituents of nucleic acids. Theyare composed of an aglycone part (a purine orpyrimidine base) and sugar moiety (β-D-ribofura-nose in RNA or 2í-deoxy-β-D-ribofuranose inDNA). In order to obtain effective therapeutic activ-ities, the natural nucleosides (Fig. 11) have to bechemically modified. From a structural viewpoint,all biologically active nucleoside analogs can bedivided into two categories: compounds modified inthe aglycone fragment, and compounds of modifiedsugar portion.

In general, the mechanism of their antiviralactivity consists in interference of the viral replica-tion cycle, and most of the known nucleosideanalogs act as so-called prodrugs. To be effective asantivirals, therapeutic compounds have to cross theplasma membrane of the cell, and this condition isusually met in the case of nucleosidic compoundswhich have been developed or are currently underdevelopment. After entering into the host cell, mostof the nucleosides are phosphorylated by viral orcellular kinases to their 5í-phosphates which, inturn, undergo further phosphorylation to the biolog-ically active 5í-triphosphates. Due to their negative-ly charged phosphate groups, the nucleoside deriva-tives in their active form cannot leave the cell, andmay stay there long enough to efficiently inhibitviral replication. The nucleoside analog triphos-phates mimic the buiding blocks of genetic material,and thus a majority of antiviral agents are targeted atone or another step of nucleic acid biosynthesis,which is vital for virus reproduction. To avoid orminimize the cytotoxic effects, a nucleoside 5í-triphosphate should not be incorporated into cellularnucleic acids, but should be a good substrate forviral enzymes responsible for virus proliferation, i.e.virus-encoded specific DNA or RNA polymerases.Therefore, high selectivity of nucleosidic agents canbe rationalized by less stringent substrate require-ments of viral enzymes than those of the correspon-ding cellular ones. And one of the structural require-ment for biological activity within this group ofchemical compounds is the presence of sugar moi-ety, which is necessary for their enzymatic phos-phorylation.

In some cases, however, the sugar part can bereduced to a chain, as it has been shown for acy-

Scheme 5. Enzymatic phosphorylation of acyclovir to its 5í-triphosphates.

Figure 11. Natural nucleosides, constituents of RNA and DNA:uridine (25), cytidine (26), adenosine (27), guanosine (28), thymi-dine (29), 2í-deoxycytidine (30), 2í-deoxyadenosine (31), 2í-deoxyguanosine (32).


clovir (Scheme 5), an acyclic analog of guanosine,which has proved to be active against herpes virus-es. The first synthesis of acyclovir from 2,6-dichloropurine has been reported by Schaeffer at al.(100), and the mechanism of its action has beenreviewed by Ellion (101). Acyclovir is phosporylat-ed to its 5í-monophosphate by a virus-encodedthymidine kinase (HSV TK) and therefore, it is acti-vated only in the infected cells. This explains highselectivity of the compound for herpes viruses. Inturn, the monophosphate is converted to acyclovirdi- and triphospate, as mentioned above. The lattercompound is then recognized by viral DNA-poly-merase, and acts as a DNA-chain terminator due tothe lack of the 3í-hydroxyl group.

The structure of acyclovir has been modified indifferent manners to get new acyclonucleosides ofpronounced therapeutic activities (102). Perhaps themost successful result of this study is ganciclovir(33; Fig. 12), which inhibits HSV-1 and HSV-2 pro-liferation at concentrations as low as 10-8 mol/L.This compound, however, is less selective than acy-clovir because, to some extent, it is phosphorylatedin uninfected cells by cellular enzymes. In this wayganciclovir is active against viruses which do notcode their own TK, like human cytomegalovirus(HCMV), but thereby it is more cytotoxic than acy-clovir. Further examples of sugar-modified nucleo-side analogs are presented in Figure 12. A similarmode of action exhibits a carba-analog of ganci-clovir, penciclovir (34). In turn, the structure ofcompound named HPMPA (35) represents a familyof phosphonylmetoxyalkyl purines. The acyclonu-

cleosides of this type are broad-spectrum anti-DNAvirus agents and can be considered as phosphonatecongeners of acyclovir 5í-monophoshate. Thus,virus-specified phosphorylation is not required fortheir activity (103). Some of the related compoundsact as inhibitors of S-adenosylhomocysteine hydro-lase, suppressing methyltransfers which are neces-sary for the maturation of viral mRNAs.

Other sugar-modified nucleoside analogs ofthe structure of 2í,3í-dideoxy-β-D-ribofuranosides,represented here by 3í-azido-3í-deoxythymidine(36; AZT, retrovir) and 2í,3í-dideoxyadenosine (37;ddA), are targeted at the reverse transcriptase ofRNA viruses, including retroviruses like the humanimmunodeficiency virus (HIV). Despite numerousside-effects, retrovir (104) is still the most effectivedrug in treatment of AIDS. Even more importantly,some nucleoside analogs are promising drug candi-dates in inhibition of hepatitis C virus RNA synthe-sis. Recently it has been found that some 2í-branched nucleosides (e.g. 2í-C-methylcytidine; 38)act as a chain terminator of HCV RNA-polymerase.

The base-modified nucleosidic drugs alsodeserve attention, and some typical structures arepresented in Figure 13. In this case, the pentofura-nosyl sugar fragment of natural precursors remainsuntouched, while the aglycone portion has beenchemically modified in order to trigger the biologi-cal activity. For instance, 5-iodo-2í-deoxyuridine(39; idoxuridine) was the first antiviral nucleosideever synthesized (105). Idoxuridine was shown to beactive against various DNA viruses, and shortlyafter that it was approved for human use as the first

Figure 12. Examples of sugar-modified antiviral nucleoside analogs (33 ñ 38).


chemotherapeutic antiviral agent. It became a proto-type compound in the search for new antiviral drugs.Indeed, quite a number of related 5-substitutedderivatives of 2í-deoxyuridine exhibit antiviralactivity, and the most active of them is 2-bro-movinyl analog (BVDU; 40).

Ribavirin (41), a ribofuranoside of 1,2,4-tria-zolecarboxamide, has been synthesized byWitkowski et al. (106). The compound represents astructure, in which the aglycone hardly resemblesany of the naturally occurring nucleosides. Ribavirinis rapidly phosporylated by cellular adenosinekinase and, therefore, exhibits a broad spectrum ofantiviral activity, including DNA and RNA viruses.Another type of base-modified nucleosides is repre-sented by the general structure 42. Deoxynucleo-sides of furo- and pyrrolo[2,3-d]pyrimidines, one ofthe newest class of biologically active nucleosidiccompounds, show remarkably potent and specificactivity against varicella-zoster virus (VZV) (107).

Apart from their aniviral activity, the nucleo-side analogs have found therapeutic application inthe treatment of cancers, and several of them hasbeen licensed for clinical use (Fig. 14). For instance,capecitabine (43), a 5í-deoxy-5-fluorouridine nucle-oside, is applied in the treatment of metastatic breastand colorectal cancers (108). Acting as a prodrug,capecitabine undergoes enzymatic conversion to 5-

fluorouracil which is an inhibitor of DNA synthesisin tumor cells. In turn, cladribine (2-CDA; 44), asimple 2-chloroanalog of 2í-deoxyadenosine (7), isan inhibitor of adenosine deaminase and interfereswith the DNA processing. This synthetic antineo-plastic drug has found application in the treatment ofhairy cell leukemia. Immucilin-H (45) can be con-sidered as an inosine analog which mimics thegeometry of transition state. The compound inhibitspurine nucleoside phosphorylase (PNP) (109).Cytarabine, an arabinofuranosyl isomer of cytosine(2), is a drug commonly used to treat hematologicalmalignancies such as leukemia and lymphoma.Gemcitabine, 2í,2í-difluoro-2í-deoxyguanosine, isapplied in various carcinoma, like breast, lung andpancreatic cancers.

Although the present mini-review in devoted tothe synthetic drug glycosides, it would be also ofinterest to mention some naturally occurring nucle-osidic antibiotics. A number of nucleoside analogsof this class have been isolated from bacteria andfungi (110, 111). From a structural viewpoint, someof them are simple analogs on natural nucleosideswhile the others are more complex di- or trisaccha-rides. For example, tubercidin and formycin can beconsidered as 7-deaza and 9-deaza-8-aza analogs ofadenosine (3), respectively. Tubercidin is incorpo-rated into DNA and inhibits DNA polymerases. This

Figure 13. Examples of base-modified antiviral nucleoside analogs (39 ñ 42).

Figure 14. Examples of anticancer nucleoside analogs (43 ñ 45).


compound exhibits antifungal, antiviral and antineo-plastic activities. In turn, tunicamycin, produced byStreptomyces lysosuperficus, (46; Fig. 14) is a com-plex trisacharide derived from uridine (25) and hasfound some applications in experimental biology.Tunicamycin acts as an inhibitor of N-linked glyco-sylation and the formation of N-glycosidic protein-carbohydrate linkages which causes cell cycle arrestin G1 phase.

In conclusion of this part we may notice thatsynthetic nucleoside analogs, resulting either fromnitrogen glycosylation reactions of heterocylic bases(convergent approach), or from chemical modifica-tion of intact nucleosides (divergent approach),proved to be particularly active against a variety ofviruses and, to some extent, in treatment of cancer.Moreover, there are some examples of applicationof nucleosidic compounds in therapy of human dis-eases caused by bacteria, mycobacteria or parasites.However, the use of nucleosides for chemotherapyof other than viral or cancer diseases is of marginalsignificance in contemporary clinical practice.

Sterols, Saponins, Terpenoids

Complex plant glycosides have been studiedextensively long before modern chemistry providedsophisticated spectral tools, which allow to performcomplete structural analysis within hours.Remarkably, until late 1950-ties, more than a half ofnew structures established for digitalis and sterolglycosides came from a single laboratory: that of T.Reichstein, in Basel, Switzerland (112). The reasonfor this intensive exploration of plant materials wasReichsteinís strong belief that plant sterols (andtheir natural glycosides) can serve as rich and effi-cient source of raw materials for chemical synthesisof mammalian hormones. In fact, his forsight has

later materialized in large scale industrial applica-tion of diosgenin and progesterone of plant origin.Saponins constitute a large and continuously grow-ing group of SM, generally split along the aglyconestructure line into steroidal and triterpenoidsaponins. Both are well covered in phytochemicalliterature (113, 114) and the first group is of consid-erable pharmaceutical importance, because of closerelation to steroidal hormones. Until recently, dios-genin, obtained from Mexican yam saponins consti-tuted the sole source of steroidal contraceptivepreparation. Hecogenin (obtained from Sisal speciessaponins) is still an effective source of pharmaceuti-cal corticosteroids. However, this kind of naturalglycosides application, which relies on exploitationof aglycon chemistry (the name: genin is adopted forsaponin aglycones) is beyond the scope of thisreview, which focuses on glycoside activity changesevoked by structural modifications of the sugar moi-ety (glycone). Simple example of such change canbe found in cardiac glycosides category, to whichScilla and Digitalis cardenolides and bufadienolidesbelong. Herbal preparations of foxglove plant(Digitalis purpurea, in documented clinical usesince W. Withering account published in 1785) relyon the presence of digoxin and lanatosides, whichare oligosaccharide glycosides of cardenolide agly-cones. Enzymatic de-glycosylation is important,since, as a rule, monoglycosides (particularly of 6-deoxypyranoses) are more active than disaccharideand trisccharide glycosides, in this group of SMs.Digoxin, trisaccharide glycoside containing 1-4linked digitoxose moieties, is a pharmaceuticalactive substance in its own right, but its 4ííí-O-methylated derivative ñ medigoxin ñ have been alsointroduced as drug because of improved bioavail-ability. Similarly, proscillaridine 47 (Fig. 16), whichis a contemporary drug with monosacharide bufa-dienolide structure, excerting positive inotropic andtonotropic action has been supplemented by semi-synthetic congener ñ 4í-O-methyl proscillaridine 48

used for the same therapeutic indications (115).These examples clearly indicate that even marginalstructural changes in a glycone portion can bringabout desirable changes in biological activity.

Pentacyclic triterpenoid saponins (which haveα-amyrin, β-amyrin, lupeol and the correspondingtriterpenoid acids as genins) are well known con-stituents of about 80 families of dicotyledone plants,but in most cases are rather poorly exploited formedicinal chemistry purposes, perhaps because ofhemolytic properties ñ the feature which they col-lectively share. Separation and structural analysis ofthese compounds require quite sophisticated equip-

Figure 15. Tunicamycin (46), a nucleoside antibiotic.


ment and individual compounds are not easily avail-able, which hampers exploratory chemistry of theclass. Aescines isolated from Aesculus hippocas-tanum (horse chesnut seed, contains ca. 20%saponins counted on dried mass) are employed, incontinuation of long lasting ethnopharmacologicaltradition, for treatment of peripheral vascular disor-ders. Oligosaccharides linked to aescigenins arecomposed of glucose, xylose, galactose and glu-curonic acid. To our knowledge no attempt has beenmade to modify glycone parts of individual con-stituents of aescine complex.

Triterpenoid carboxylic acids are known tooccur as glycosides, as well as glycosyl esters. Bothgroups (and their combination) are of potential inter-est in medicinal chemistry. Glycyrrhizinic acid is adiglucuronic acid glycoside; its salts are responsiblefor sweet taste of liquorice (Glycyrrhiza spp. rootextracts), used in pharmacy as sweetening and fla-voring agent, also employed in confectionery andtobacco industry. Steviosides, kaurenoic acidanomeric esters and glycosides from Stevia rebaudi-ana are ca. 300 times sweeter than sucrose and foundapplication in food industry, mainly as soft drinksweeteners. In this case a number of syntheticanalogs were obtained and their sensory evaluationwas made (116). Asiaticoside ñ β-anomeric ester oftrihydroxyursenoic acid and trisaccharide containingtwo D-glucose units and terminal L-rhamnose foundin Centella asiatica has interesting wound healingproperties. Betulinic acid, for which antiviral andantitumor properties have been reported is anotherexample of potential drug lead for which there is anample room for design glycorandomization.

Perhaps the most famous saponin containingplant is Asian ginseng (Panax ginseng C.A. Meyer),very difficult to grow beyond its natural habitat andrequiring at least five years to mature for harvest.The root of this plant, used in Asia as general tonicand vitality enhancing agent since time immemorial,contains close to forty ginsenosides, with five majorones (Rb, Rb2 , Rc, Re and Rg) constituting morethan 80% of total saponins. Structures of these com-pounds, which are derivatives of triterpenoiddammarane are presently well known and their bio-logical activity is being studied, since ginsenosideshave attracted much attention of western medicine,at first as herbal product and later as a collection oflead compounds for drug candidates for CNS and/oroncological therapies. A number of structural modi-fications of natural ginsenosides have been per-formed, including enzymatic degradation as well aschemical acylation, in order to modify their pharma-codynamic properties (117).

Miscellaneous

Plant glycosides with typically deterrent func-tion, like cyanogenic glycosides or glucosinolates,which easily release volatile toxins, are not likely tofind medical applications, but can serve as an inspi-ration, how to construct easily biodegradable carbonñ toxin bond (118).

Taxoids, of which acclaimed antitumor drug ñpaclitaxel ñ is the best known example, belong tohigher plant diterpenoids, which can occur also inglycosylated state. In search for better soluble pacli-taxel derivatives, many analogs of natural 7-O-D-xylopyranosyl paclitaxel have been obtained. Sincethe tetracyclic diterpenoid aglycone is not an easysubstrate for chemical glycosylation, another type ofsugar conjugates were designed. The glycosylationsite: 7-O- secondary function was esterified withprotected glycosyloxyacetic acids (119). This typeof ìthrough spacerî glycoconjugation of complexand multifunctional natural products, is becomingmore popular recently.

Phenylethanoid glycosides are also an inter-esting group of SMs, exhibiting various, potenta-ially useful biological activities. Verbascoside 49

(also called acteoside; Fig. 17) is known as aninhibitor of important enzymes: protein kinase Cand aldose reductase. It is also claimed to possessantitumor and immunomodulatory properties(120). This glycoside contains four phenolic andfive hydroxyl groups located in disaccharide moi-ety, which offer ample opportunity for structuremodifications inquest for selectivity of biologicalaction.

Figure 16. Proscillaridin (47) and meproscillaridin (49)


Obviously, examples of natural glycosideswith potential for medicinal chemistry evaluationcould be largely multiplied, based on such sourcesas Natural Products Reports, but we intend to keepthis collection of examples concise, rather than com-prehensive.

CONCLUSIONS

Catastrophic attrition rates observed recently inbig pharma pipelines, call for reconsidering of drugdiscovery and development programs on everylevel. Since natural products remain to be our bestsource for the drug lead compounds, the questionhow to intensify the lead generation process,becomes pressing one. Plant and microbial second-ary metabolites, which offer various environmentaladvantages to the host, occasionally happen to exertspecific biological action in higher organisms, aphenomenon well recognized by medicinal chem-istry and exploited in pharmacy and clinical medi-cine. Various hybrid structures, combining differentbiogenetic pathways proved particularly useful aslead compounds for drug discovery and develop-ment. For example, SMs are often glycosylated,regardless of their biogenetic origin, and the glyconesometimes is essential for a given biological action.In traditional medicinal chemistry, prodrugs of gly-cosidic structure are considered primary as site spe-cific drug delivery systems undergoing enzymaticsplit, liberating an aglycone in the colon (121).Current ideas concerning this issue are less straight-forward (84). Although from glycobiology perspec-tive, an oligosaccharide is the smallest unit securingmolecular recognition on the biopolymer level,numerous examples clearly indicate that small mol-ecules of glycosidic structure can exert a variety ofselective biological action. Moreover, in such caseseven marginal structural changes in glycone or agly-cone part, can be used to tune up efficacy and selec-tivity of activity. It seems evident that chemical gly-cosylation can provide a key technology for both:structural modifications of natural glycosides and

glycosidation of SMs, which are capable of formingO-, N-, S- or C-glycosidic bond. Among all classesof glycosyl donors, glycals deserve special attentionas very versatile precursors of structurally modifiedglycones (122). Practically all classes of naturalproducts can be converted to glycosylated deriva-tives by chemical methods discussed above.Furthermore, newly installed glycone part offersample opportunity for fine tune up of lipophilicity,which is one of the fundamental features determin-ing ìdruglikenessî (123). Such chemocontrolledstructural variation of the SM pool could, in ouropinion, provide many new and useful lead com-pounds for drug development programs, at least dur-ing the period before biotechnological methods,offering parallel transformations for glycodiversifi-cation, reach maturity (124).

Acknowledgments

Financial support from the Scientific Network:Synthesis, Structure and Therapeutic Properties ofOrganic Compounds, coordinated by Institute ofOrganic Chemistry, Polish Academy of Sciences inWarsaw (G.G. and J. B.) and the Polish StateCommittee for Scientific Research (Grant No. 3T09B 11229; W. S.) is gratefully acknowledged.

REFERENCES

1. Taylor M.E., Drickamer K.: Introduction toGlycobiology, Oxford University Press, Oxford2003.

2. Comprehensive Glycoscience; From Chemistryto Systems Biology. Kamerling J.P. Ed.,Elsevier, Amsterdam 2007.

3. Samuelsson G.: Drugs of natural origin, 4th ed.,Apotekarsocieten, Stockholm 1999.

4. Plotkin M.J.: Medicine Quest: In search ofNatureís Healing Secrets, Viking, New York2000.

5. Ko≥odziejczyk A.: Natural organic compounds(in Polish), PWN, Warszawa 2004.

Figure 17. Verbascoside (acteoside 49).


6. Secondary Metabolism in Model Systems;Recent Advances in Phytochemistry, vol. 38,Romeo J.T. Ed., Elsevier, Amsterdam 2004.

7. Harborne J.: Ecological biochemistry (Polishtranslation), PWN, Warszawa 1997.

8. Oleszek W., Glowniak K., Leszczynski B. Eds.:Biochemical environmental interactions (inPolish), AM, Lublin 2001.

9. Evans W.C.: Trease and Evans Pharmacognosy,15th ed., W.B. Saunders, Edinburgh, 2002.

10. Weymouth-Wilson A.C.: Nat. Prod. Rep. 14, 99(1997).

11. K¯en V., Martinkova L.: Curr. Med. Chem. 8,1313 (2001).

12. Thorson J.S., Vogt T.: in Carbohydrate-basedDrug Discovery, Wong C-H. Ed., vol 1, p. 685,Wiley-VCH, Weinheim 2003.

13. Cutler S.J, Cutler H.G. Eds., Biologically activenatural products, CRC Press, Boca Raton 2000.

14. Goodman L,: Adv. Carbohydr. Chem. 22, 109(1967).

15. Lemieux R.U.: Adv. Carbohydr. Chem. 9, 1(1954).

16. Schmidt R.R.: Angew. Chem. Int. Ed. Engl. 25,212 (1986).

17. Garegg P.J.: Adv. Carbohydr. Chem. Biochem.52, 174 (1997).

18. Mukaiyama T.: Angew. Chem. Int. Ed. Engl.43, 5590 (2004).

19. Handbook of Chemical Glycosylation, Dem-chenko A.V. Ed., Wiley-VCH, Weinheim, NewYork 2008.

20. Shoda S.I.: in Handbook of Chemical Glyco-sylation, Demchenko A.V. Ed., p. 29, Wiley-VCH, Weinheim, New York 2008.

21. Ryan D.A., Gin D.Y.: in Handbook of ChemicalGlycosylation, Demchenko A.V. Ed., p. 95,Wiley-VCH, Weinheim, New York 2008.

22. Szeja W., Grynkiewicz G.: in Handbook ofChemical Glycosylation, Demchenko A.V. Ed.,p. 329, Wiley-VCH, Weinheim, New York 2008.

23. Kim K.S., Jeon H.B.: in Handbook of ChemicalGlycosylation, Demchenko A.V. Ed., p. 185,Wiley-VCH, Weinheim, New York 2008.

24. Friesen R.W., Danishefsky S.J.: J. Am. Chem.Soc. 111, 6656 (1989).

25. Halcomb R.L,: Danishefsky S.J.: J. Am. Chem.Soc. 111, 6661 (1989).

26. Lemieux R.U., Hendriks K.B., Stick R.V.,James K.: J. Am. Chem. Soc. 97, 4056 (1975).

27. Veeneman G.H., van Leeuwen S.H., van BoomJ.H.: Tetrahedron Lett. 31, 1334 (1990).

28. Paulsen H.: Angew. Chem. Int. Ed. Engl. 21,155 (1982).

29. Crich D., Sun S.: J. Org. Chem. 61, 4506(1996).

30. Crich D., Smith M.: Org. Lett. 2, 4067 (2000).31. Anderson F., Fugedi P., Garegg P.J., Nashed

M:. Tetrahedron Lett. 27, 3019 (1986).32. Garegg P.J., Oscarson S., Szonyi H.:

Carbohydr. Res. 205, 125 (1990).33. Toshima K., Tatsuta K.: Chem. Rev. 93, 1503

(1993).34. Pellissier M.: Tetrahedron 61, 2947 (2005).35. Paulsen H.: Angew. Chem., Int. Ed. Engl. 21,

184 (1982).36. Needs P. W., Williamson G.: Carbohydr. Res.

330, 511 (2001).37. Bouktaib M., Atmani A., Rolando C.:

Tetrahedron Lett. 43, 6263 (2002).38. Nishida Y., Ohrui H., Meguro H., Ishizawa M.,

Matsuda K., Taki T., Handa S., Yamamoto N.:Tetrahedron Lett. 35, 5465 (1994).

39. Wang C.C., Lee J.C., Luo S.Y., Fan H.F., PaiC.L., Yang W.C., Lu L.D., Hung S.C.: Angew.Chem., Int. Ed. Engl. 41, 2360 (2002).

40. Nikolakakis A., Haidara K., Sauriol F., MamerO., Zamir L.O.: Biorg. Med. Chem. 11, 1551(2003).

41. Kanie O., Ito Y., Ogawa T.: Tetrahedron Lett.37, 4551 (1996).

42. Kopper S., Thiem J.: Carbohydr. Res. 260, 219(1994).

43. Messaoudi S., Anizon F., Pfeiffer B., GolsteynR., Prudhomme M.: Tetrahedron Lett. 45, 4643(2005).

44. Grabley S., Thiericke R. Eds.: Drug discoveryfrom nature, Springer, Berlin 2000.

45. Butler M.S.: Nat. Prod. Rep. 22, 162 (2005).46. Booser D.J., Hortobagyi G.N.: Drugs 47, 223

(1994).47. Monneret C.: Eur J. Med. Chem. 36, 483

(2001).48. Minotti G., Menna P., Salvatorelli E., Cairo G.,

Gianni L.: Pharmacol Rev. 56, 185 (2004).49. Anthracycline Chemistry and Biology, Krohn

K. Ed., Topics in Current Chemistry, Vol. 282and 283, Springer, Berlin 2008.

50. Priebe W., Perez-Soler R.: Pharmacol. Ther. 60,215 (1993).

51. Priebe W.: Curr. Drug Design 1, 73 (1995).52. Chaires J.B., Satyanarayana S., Suh D., Fokt I.,

Przewloka T., Priebe W.: Biochemistry 35,2047 (1996).

53. Priebe W., Van N., Burke T.G., Perez-Soler R.:Anticancer Drugs 4, 37 (1993).

54. Otsomaa L.A., Koskinen A.M.P.: Fortschr.Chem. Org. Naturst. 74, 197 (1998).


55. Grynkiewicz G., Szeja W.: Top. Curr. Chem.282, 249 (2008).

56. Grynkiewicz G., Achmatowicz O., Fokt I.,Priebe W., Ramza J., Szechner B., Szeja W.:Wiad. Chem. 56, 536 (2002).

57. Grynkiewicz G., Fokt I., Skibicki P., PrzewlokaT., Szeja W., Priebe W.: Polish J. Chem. 79,335 (2005).

58. Fang L., Zhang G., Li C., et al.: J. Med. Chem.49, 932 (2006).

59. Grynkiewicz G., Achmatowicz O., Hennig J.,Indulski J., Klessig D.F.: Polish J. Chem. 67,965 (1993).

60. Hanessian S., Mascitti V., Lu P.P., Ishida H.:Synthesis 1959 (2002).

61. Clostre F.: Ann. Pharm. Fr. 57 (Suppl. 1), 1S8(1999).

62. Middelton E., Kandaswami C., TheoharidesT.C.: Pharm. Rev. 52, 673 (2000).

63. Flavonoids in Health and Disease, Rice-EvansC.A., Packer L., Eds., Marcel Dekker, NewYork 2003.

64. Flavonoids: Chemistry, Biochemistry andApplications, Andersen O.M., Markham K.R.Ed., CRC Press, Boca Raton 2006.

65. K¯en V., Kubisch J., Sedmera P., et al.: J.Chem. Soc., Perkin Trans. 1, 2467 (1997).

66. Ross J.A., Kasum C.M.: Ann. Rev. Nutr. 22, 19(2002).

67. Wu A.H., Wan P., Hankin J., Tseng C.C., Yu M.C., Pike M.C.: Carcinogenesis 23, 1491 (2002).

68. Meritt R.J., Jenks B.H.: J. Nutr. 134, 1220S(2004).

69. Harborne J.B., Williams C.A.: Phytochemistry55, 481 (2000).

70. Birt D.F., Hendrich S., Wang W.: Pharmacol.Ther. 90, 157 (2001).

71. Dixon R.A., Ferreira D.: Phytochemistry 60,205 (2002).

72. Stanosz S., Puk E., Grobelny W., Stanosz M., Ka-zikowska A.: Przegl. Menopauz. 3, 182 (2006).

73. Polkowski K., Popiolkiewicz J., KrzeczynskiP., et al.: Cancer Lett. 203, 59 (2004).

74. Rusin A., Gogler A., Bochenek D., Glowala-Kosinska M., Grynkiewicz G., Krawczyk Z.:Cancer Mol. Pharmacol. submitted.

75. Lewis P.T., Kaltia S., W‰h‰l‰ K.: J. Chem.Soc., Perkin Trans. 1. 2481 (1998).

76. Lewis P.T., W‰h‰l‰ K.: Tetrahedron Lett. 39,9559 (1998).

77. Boryski J., Grynkiewicz G.: Synthesis 2170(2001).

78. Allevi P., Anastasia M., Ciuffreda P., Bigatti E.,Macdonald P.: J. Org. Chem. 58, 4175 (1993).

79. Bender R.P., Jablonsky M.J., Shadid M., et al.:Biochemistry 47, 4501 (2008).

80. Wiczkowski W., Romaszko J., Bucinski A.,Szawara-Nowak D., Honke J., Zielinski H.,Piskula M.K.: J. Nutr. 138, 885 (2008).

81. Fernandez S.P., Wasowski C., Loscalzo L.M.,Granger R.E., Johnston G.A.R., Paladini A.C.,Marder M.: Eur. J. Pharmacol. 539, 168 (2006).

82. Harborne J.B., Williams C.A.: Nat. Prod. Rep.15, 631 (1998).

83. Feizi T.: Curr. Opin. Struct. Biol. 3, 701(1993).

84. K¯en V.: in Glycoscience; Chemistry andChemical Biology, B. Fraser-Reid, K.Tatsuka, J. Thiem, Eds., Vol 1, Chapter 9.3, p.2471, Springer, Berlin 2001.

85. Larsen E., Kharazmi A., Christensen L.P.,Christensen S.B.: J. Nat. Prod. 66, 994 (2003).

86. Natori N.; Koezuka Y., Higa T.: TetrahedronLett. 34, 5591 (1993).

87. Brutkiewicz R. R., Sriram V.: Crit. Rev.Oncol./Hematol. 41, 287 (2002).

88. Brutkiewicz R. R., Lin Y., Cho S., Hwang Y.K., Sriram V., Roberts T. J.: Crit. Rev.Immunol. 23, 403 (2003).

89. Zhou D., Cantu C., Sagiv Y., et al.: Science303, 523 (2004).

90. Crul M., Mathor R.A.A., Giaccone G., et al.:Cancer Chemother. Pharmacol. 49, 287(2002).

91. Wrotnowski C.: Genet. Eng. News 21, 1(2001).

92. Mocchetti I.: Cell Mol. Life Sci. 62 , 2283(2005).

93. Mimee B., Labbe C., Pelletier R., BelangerR.R.: Antimicrob. Agents Chemother. 49,1597 (2005).

94. Sihorkar V., Vyas S.P.: J. Pharm. Pharmaceut.Sci. 4, 138 (2001).

95. Sato T., Sunamoto J.: Prog. Lipid Res. 31, 345(1992).

96. Ochi A., Shibata S., Mori K., Sato T.,Sunamoto J.: Drug Deliv. Syst. 5, 261 (1990).

97. Antiviral Drug Development. A Multidiscipli-nary Approach, De Clercq E., Walker R.T.Eds., Plenum Press, New York & London 1988.

98. Antiviral chemotherapy, Jefferies D.J., DeClercq E. Ed., Wiley, Chichester 1995.

99. Boryski J.: in: Antiviral activity of nucleosideanalogs and their application in therapy.Barciszewski J., £astowski K., Koroniak H.Eds., New trends in molecular biology, genet-ic engineering and medicine (in Polish), Sorus,PoznaÒ 1996.


100. Schaeffer H.J., Beuchamp L., de Miranda P.,Elion G.B, Bauer D.J., Collins P.: Nature 272,583 (1978).

101. Elion G.B.: Am. J. Med., 73, 7 (1982). 102. Agrofolio L.A., Challand S.R.: in Acyclic,

Carbocyclic and L-Nucleosides, Kluwer,Dordrecht, Boston, London 1998.

103. De Clercq E., Holy A., Rosenberg I., SakumaT., Balzarini J., Maudgal P.C.: Nature 323,464 (1986).

104. De Clercq E.: J. Med. Chem. 29, 1561 (1986).105. Prusoff W.H.: Biochim. Biophys. Acta 32, 295

(1959).106. Witkowski J.T., Robins R.K., Sidwell R.W.,

Simon L.N.: J. Med. Chem. 15, 1150 (1972).107. Janeba Z., Balzarini J., Andrei G., Snoeck R.,

De Clercq E., Robins M.J.: J. Med. Chem. 48,4690 (2005).

108. Malet-Martino M., Martino R.: Oncologist 7,288 (2002).

109. Miles R.W., Tyler P.C., Furneaux R.H., Bag-dassarian C.K., Schramm V.L.: Biochemistry37, 8615 (1998).

110. Knapp S.: Chem. Rev. 95, 1859 (1995).111. Garner P.: in Synthetic Approaches to

Complex Nucleoside Antibiotics. Studies inNatural Product Chemistry, Atta-ur-RahmanEd., Elsevier, Amsterdam 1988.

112. Reichstein T., Weiss E.: Angew. Chem. Int.Ed. 1, 572 (1962).

113. Oleszek W., G≥owniak K., LeszczyÒski B.Eds., Biochemical environmental interactions(in Polish), p. 213, AM, Lublin 2001.

114. Evans W.C.: Trease and Evans Pharmaco-gnosy 15th ed., p. 289, WB Saunders, Edin-burgh 2002.

115. Kubinyi H.: Arzneim. Forsch. 28, 493 (1978).116. Geuns J.M.: Phytochemistry 64, 913 (2003).117. Lei J., Li X., Gong X., Zheng Y.: Molecules

12, 2140 (2007).118. Fahey J.W., Zalcmann A.T., Talalay P.:

Phytochemistry 56, 5 (2001).119. Mikuni T., Nakanishi K., Hara K., et al.: Biol.

Pharm. Bull. 31, 1155 (2008).120. Jimenez C., Riguera R.: Nat. Prod. Rep. 11,

591 (1994).121. Stella V. J.: J. Med. Chem. 23, 1275 (1980).122. Priebe W., Fokt I., Grynkiewicz G. in:

Glycoscience Chemistry and ChemicalBiology, Fraser-Reid B.O., Tatsuta K., ThiemJ. Eds., Vol. 1, p. 699, Springer Verlag, Berlin2008.

123. Lipinski C.A., Lombard F., Dominy B.W.,Feeney P.J: Adv. Drug Deliv. Rev. 23, 3(1997).

124. Thibodeaux C.J., Melancon C.E., Liu H.:Nature 446, 1008 (2007).

Documents

SYNTHETIC ANALOGS OF NATURAL GLYCOSIDES …structural information storage, related to biological function, looks incomparably more complicated, that previously defined fields of functional