Introduction to Glycosylation: new methodologies and applications

Antony J. Fairbanks *a, Sabine L. Flitsch *b and M. Carmen Galan *c
aDepartment of Chemistry, University of Canterbury, Private Bag 4800, Christchurch 8140, New Zealand. E-mail: antony.fairbanks@canterbury.ac.nz
bUniversity of Manchester, Manchester Institute of Biotechnology, 131 Princess Street, Manchester M1 7DN, UK. E-mail: sabine.flitsch@manchester.ac.uk
cSchool of Chemistry, Cantock's Close, University of Bristol, Bristol BS8 1TS, UK. E-mail: m.c.galan@bristol.ac.uk

Carbohydrate chemistry is a core enabling aspect of glycoscience, a scientific discipline that centres on both unravelling the myriad roles that complex carbohydrates play throughout biology, and also exploiting them for therapeutic purposes. Glycosylation is the key reaction by which the complex carbohydrate structures (glycans) that are found in glycoproteins, glycolipids and other biologically important glycoconjugate structures, are assembled.

One of the key considerations of glycosylation, and most difficult aspect to control, is the anomeric stereochemistry of the product. Carbohydrate chemists have been grappling with this problem for many years, and as yet no ‘universal solution’ applicable to every situation exists. Thus, the topic continues to be one of serious interest, and several of the papers in this themed collection concern the development of new methods for control of the stereochemical outcome of glycosylation.

The 1,2-cis-β-mannose linkage is commonly regarded as one of the most difficult to synthesise. However, β-rhamnose, which is of course of identical configuration but lacks the 6-hydroxyl group, is an even more difficult problem, since approaches like Crich's β-mannosylation that require cyclic 4,6-protection for high levels of stereocontrol cannot be used effectively. In their excellent review (Org. Biomol. Chem., 2020, 18, 3216–3228), Kulkarni and Rai have collated and summarised the ingenious advances in β-rhamnosylation.

In their paper, Iadonisi and co-workers report (Org. Biomol. Chem., 2020, 18, 5157–5163) a method for the synthesis of some other 1,2-cis glycosidic linkages, e.g. α-glucosides and α-galactosides, taking inspiration from Lemeux's in situ anomerisation method. In this study they use glycosyl chlorides as the donors, and activate them by heating in the presence of a combination of Bu4NBr, P(OEt)3 and excess Hünig's base as a neat liquid amine in the absence of any solvent. A one-pot method involving in situ formation of the glycosyl chloride donors from the corresponding reducing sugars was also developed.

Traditionally neighbouring group participation (NGP) of simple ester protecting groups at the 2-position of the glycosyl donor has been used to access 1,2-trans glycosidic linkages. In their paper (Org. Biomol. Chem., 2020, 18, 1165–1184), Boltje and co-workers report an expansion of this type of NGP during glycosylation, using a variety of functionalised ethers at position-2 of the glycosyl donor which contain phosphine oxides or tertiary amides.

Ragains and co-workers report (Org. Biomol. Chem., 2020, 18, 2405–2409) investigations into a method for stereoselective 1,2-cis glycosylation (namely α-glucosylation) which exploits the combined use of donors bearing benzyl ethers decorated with electron withdrawing groups and Lewis basic solvents/additives.

Stereochemical control of glycosylation of ketoses can be extremely challenging. Moreover, stereocontrolled glycosylation of furanoses is considerably more difficult than that of pyranoses since the anomeric effect cannot be taken advantage of, and usually no well-defined conformational preference, such as a chair, exists for the donor. Lowary and Huang report (Org. Biomol. Chem., 2020, 18, 2264–2273) elegant β-selective glycosylation using xylulose donors in the furanose form, that they achieve by use of a conformational lock (3,4-O-xylelene protection), and apply it to the synthesis of the repeating unit of the lipopolysaccharide O-antigen from Y. enterocolitica O:5/IO:5,27.

The logistical advantages of catalytic synthetic processes over stoichiometric ones are plain to see, and glycosylation is no different. In this vein, Liu and co-workers report (Org. Biomol. Chem., 2020, 18, 2242–2251) the α-selective Ferrier-type C-glycosylation of a range of glycals using several palladium(II) complexes and aryl iodonium salts.

Turning the tables somewhat, and using the carbohydrate as part of the catalyst, Galan and co-workers report (Org. Biomol. Chem., 2020, 18, 3012–3016) the production of some rhodium complexes of N-heterocyclic carbenes (NHCs) that comprise two carbohydrates as chiral ligands, and their application to the enantioselective hydrosilylation of acetophenone. Interestingly the two chiral NHC complexes, in which the sugars were linked either through C-2 or C-3 to the N atoms of the NHC, behave as pseudo enantiomers.

Glycosyl trichloroacetimidates are one of the most useful and widely applied classes of glycosyl donor, but they typically require activation by a reactive and sensitive Lewis acid catalyst, such as BF3·etherate or TMSOTf. In their paper (Org. Biomol. Chem., 2020, 18, 851–855), Toshima and co-workers report the photo-induced activation and glycosylation of trichloroacetimidates using diaryldisulfides as organo-Lewis photoacid catalysts.

The basic reactivity and effectiveness of trichloroacetimidates as donors has been modulated by changing features of the anomeric leaving group, for example by replacement of the three chlorines by fluorines and/or replacement of the H atom attached to N with an aryl group. In their paper (Org. Biomol. Chem., 2020, 18, 1918–1925), Pedersen and Kowalska take this idea further, and report the synthesis and activation of a series of glycosyl trifluoroacetimidates bearing ortho-methoxyaryl groups on the N atom. Donor synthesis was, except in the case of mannose, completely β-selective, and their activation and glycosylation, the preferred conditions using either Fe(II) or Bi(III) triflates in ether, was highly α-selective.

Glycosyl acetamides have so far found significantly less application than their glycosyl imidate counterparts. In their paper (Org. Biomol. Chem., 2020, 18, 3043–3046), Ye and co-workers expand applications of glycosyl acetamides as useful glycosyl donors by demonstrating that their Lewis acid catalysed activation with SnBr4 in the presence of allyl trimethylsilane and a range of enol ethers and phenols and derivatives, leads to the synthesis of a range of C-glycosides. Interestingly when phenols are used as the acceptors the reaction is highly β-selective and proceeds via an O to C rearrangement.

Besides the difficulties associated with controlling stereochemistry, glycosylation chemistry is also logistically arduous. As such there has been significant effort over the past ∼20 years to facilitate glycosylation chemistry in a more general sense. For example, controlling the outcome of glycosylation also typically means the protracted production of selectively protected glycosyl donors and acceptors, involving multiple protecting group manipulations.

In an attempt to address this type of logistical issue Pohl and Bennett and their co-workers report (Org. Biomol. Chem., 2020, 18, 3254–3257) the development of an efficient modular flow synthesis of orthogonally protected 6-deoxy glucose glycals; monosaccharide building blocks that have wide application in oligosaccharide synthesis. Significant advantages that continuous flow routes such as this have are those of scale and required reaction time. For example, in this paper glycals were produced in only 20–40 min of flow time that would typically require a week to make via a traditional batch approach in the lab.

There is of course still room for improvement in increasing the efficiency by which traditional batch chemistry can be applied to protecting group chemistry of monosaccharides. In their paper Mong and co-workers report (Org. Biomol. Chem., 2020, 18, 3135–3141) a useful one-pot reductive etherification procedure of OH groups, via intermediate TMS ethers, involving subsequent reaction with aldehydes and reduction by polymethylhydrosilane; the process can be used to install benzyl and Nap ethers on substrates with base labile protecting groups, and can also be applied as part of one-pot protecting group manipulations of completely deprotected monosaccharides.

Finally with donors and acceptors in hand, then there is the time and effort required to perform multiple glycosylation reactions. Solid phase automated synthesis clearly has particular advantages for this type of reaction. In their paper (Org. Biomol. Chem., 2020, 18, 1349–1353), Delbianco and co-workers demonstrate the ever-widening applicability of automated glycosylation by reporting the automated solid-supported synthesis of charge bearing oligosaccharides (from hexa- to dodeca-) comprising monosaccharides bearing either carboxylic acid and/or amino functionality. The low reactivity of uronic acid donors and their limited applicability to solid-supported glycosylation was circumvented by the attachment of carboxylic acid functionality onto the monosaccharide donors as ethers of glycolic acid.

The main objective of the methodological investigations outlined above is of course to facilitate the actual production of oligosaccharide structures of biological interest for further study or application.

The synthesis of glycosaminoglycans (GAGs) is particularly challenging and many approaches have been developed over the years to overcome the issues associated with these complex synthetic targets. One of the main hurdles, as alluded to above, is the presence of uronic acid building blocks in the target which are difficult glycosyl donors to activate. To overcome this, the corresponding C-6 protected hexoses are often employed, but the method then requires selective C-6 deprotection and late oxidation. Schwörer et al. (Org. Biomol. Chem., 2020, 18, 4728–4733) have now evaluated the chemical synthesis of a range of uronic acid- and hexose-based disaccharide imidate and thioglycoside donors in the [2 + 2] synthesis of heparan sulfate fragments. They find that, using the same protecting group pattern, hexose imidate donors perform better than the corresponding uronate-counterparts with high selectivity and very good yields; while generally lower yields are obtained in the case of thioglycoside donors with no difference between the respective hexose and glucuronic acid versions.

Another important glycosaminoglycan is hyaluronic acid (HA) which is found in the extracellular matrix and believed to be involved in cell migration through interactions with the transmembrane glycoprotein CD44. Li, Hung et al. (Org. Biomol. Chem., 2020, 18, 5370–5387) have developed an efficient synthetic route to access HA di-, tetra- and hexasaccharides containing the key GlcNAc–GlcA repeating unit. This group also measured the binding interaction between these sugars and human CD44 by isothermal titration calorimetry (ITC) and found that a tetrasaccharide of the pattern (GlcNAc–GlcA)2 is the shortest structure that can interact effectively with CD44.

In other work on GAGs, Lopin-Bon et al. (Org. Biomol. Chem., 2020, 18, 4831–4842) describe an efficient, regioselective and stereocontrolled synthesis of various sulfoforms of the GAG disaccharide GlcA-1,3-β-D-Gal and the trisaccharides GlcNAc-1,4-α-D-GlcA-1,3-β-D-Gal and GalNAc-1,4-β-D-GlcA-1,3-β-D-Gal. The compounds were evaluated as substrates for chondroitin sulfate N-acetylgalactosaminyltransferase-1 (CSGalNAcT-1) and preliminary results showed that non-sulfated and 6-sulfated disaccharides are substrates for the enzyme, while the 4-sulfated analogue was not, which highlights the importance of the sulfation pattern for biological recognition.

Glycoconjugates, such as resin glycosides are also an important class of glycolipids with interesting bioactivities including cytotoxicity against human cancer cell lines. Wan et al. (Org. Biomol. Chem., 2020, 18, 3818–3822) describe two parallel approaches to access tricolorin A, a bioactive glycolipid. A stepwise [2 + 2] synthesis using the interrupted Pummerer reaction-mediated (IPRm) glycosylation to efficiently construct the glycosidic bonds is initially showcased, and they find that the O-2-[(propan-2-yl)sulfinyl]benzyl (OPSB) glycosyl donor could achieve better stereoselectivity than the S-2-[(propan-2-yl)sulfinyl]benzyl (SPSB) version. On the other hand, a one-pot relay glycosylation strategy was also employed where two different glycosidic bonds are sequentially assembled with only one equivalent of external activator to access the tetrasaccharide with significantly improved overall efficiency.

In other developments, van der Marel and Codée et al. (Org. Biomol. Chem., 2020, 18, 2834–2837) report the synthesis of an orthogonally protected bacillosamine (2,4-diamino-2,4,6-trideoxyglucose) synthon that allows for the differentiation of the two amino functionalities. They also demonstrate its versatility in the preparation of an N. meningitidis PilE disaccharide with a C-3-α-D-galactose appendage and a C-4-N-glyceroyl, which are post-translational modifications of N. meningitidis PilE proteins and make up the bacterial pili. These novel probes have therapeutic and diagnostic potential to generate antibodies directed at N. meningitidis PilE.

The same group (Org. Biomol. Chem., 2020, 18, 2038–2050) also reports on the synthesis of a D-alanine kojibiose functionalized glycerol phosphate teichoic acid fragment, which can be found in the cell wall of the opportunistic Gram-positive bacterium, Enterococcus faecalis. To assemble the challenging α-1,2-glucosyl linkages, an additive controlled glycosylation strategy is explored whereby per-benzylated glucosyl imidate donors with stoichiometric TfOH in the presence of an excess DMF is used. This work further demonstrates how nucleophilic additives can be used to aid the construction of difficult to access 1,2-cis-glycosidic linkages.

Bp and Bm lipopolysaccharides (LPS) have been identified as attractive vaccine candidates for the development of prophylactic measures against melioidosis and glanders, which are caused by Gram-negative bacteria Burkholderia pseudomallei (Bp) and Burkholderia mallei (Bm). Gauthier et al. (Org. Biomol. Chem., 2019, 17, 8878–8901) report a glycosylation strategy that relies on the late-stage epimerization of the inner rhamnose into a 6-deoxy-L-talose residue to access two tetrasaccharide fragments which mimic the main substitution epitopes of the Bp and Bm LPS O-antigens. Moreover, the team demonstrate that these probes have potential as candidates for the development of vaccines and/or diagnostic tools against melioidosis and glanders since they exhibit antigenicity similar to the native Bp O-antigen.

Human milk oligosaccharides (HMO) have been linked to many beneficial health effects. However, their exact biological roles and functions remain unknown beyond prebiotic effects due to the lack of well-defined HMOs in sufficient amounts for evaluation. Demchenko et al. (Org. Biomol. Chem., 2020, 18, 1747–1753) describe the first chemical synthesis of lacto-N-neohexaose and lacto-N-neotetraose via a convergent [2 + 2 + 2] strategy, which employs orthogonally protected lactose and lactosamine building blocks to access the targets in high yield.

In addition to new chemical glycosylation strategies, enzymatic and chemoenzymatic methods are gaining popularity. Whilst the application of biocatalysis has in the past been mainly used for the synthesis of natural carbohydrates, it is increasingly appreciated that enzymes can be promiscuous, and also accept non-natural substrates. The review by Council et al. (Org. Biomol. Chem., 2020, 18, 3423–3451) illustrates that a wide range of fluorinated carbohydrates have been accessible through chemoenzymatic strategies. Related to this topic, Huang et al. (Org. Biomol. Chem., 2020, 18, 3142–3148) describe the characterization of a new recombinant bacterial α1,4 galactosyltransferase from N. weaveri, which led to the synthesis of Gb3, P1, and their amino- and deoxyfluoro-analogues.

Thioglycosides are of great interest as biostable analogues of glycosidic linkages. Romanò et al. (Org. Biomol. Chem., 2020, 18, 2696–2701) report on the introduction of S-linkages into linear and branched arabinoxylan oligosaccharides (AXs) in order to obtain a small collection of synthetic tools as inert substrates or inhibitors for the study of AX-degrading enzymes. Thioglycosides are also accessible through enzymatic synthesis, as shown by Kurdziel et al. (Org. Biomol. Chem., 2020, 18, 5582–5585), who reported engineering of a β-D-glucuronidase DtGlcA from Dictyoglomus thermophilum, to generate an active thioglycoligase that is able to catalyse the formation of numerous S-glucuronides starting with natural sugar donors.

The introduction of biorthogonal and metabolically stable azido-groups into carbohydrates continues to be a widely used tool for understanding disease. In elegant studies, Parker et al. (Org. Biomol. Chem., 2020, 18, 3607–3612) describe the chemo-enzymatic synthesis of azido-functionalised asymmetric trehalose analogues that were subsequently used as probes for trehalose processing pathways in mycobacteria, the causative agents of tuberculosis. Jaiswal et al. (Org. Biomol. Chem., 2020, 18, 2938–2948) describe the synthesis of Nα,Nε-diacetyl-L-lysine-inositol conjugates containing azides, which were shown to be incorporated into the GPI biosynthesis pathways of cancer cells, but not into normal cells, thus allowing for selective metabolic labelling as a novel strategy.

A significant advantage of enzymatic strategies is the possibility of telescoping several reactions in one experiment. McArthur et al. (Org. Biomol. Chem., 2020, 18, 738–744) report the efficient use of one-pot multienzyme (OPME) glycosylation systems for synthesizing 5,7-di-N-acetyllegionaminic acid glycosides and analogues from the corresponding N-acyl mannosamines.

Carbohydrate chemistry provides access to valuable synthetic molecular tools that can interrogate biological processes and help with development of diagnostics and therapeutics. Haksar et al. (Org. Biomol. Chem., 2020, 18, 52–55) report the synthesis of a polymeric hybrid containing both the known ligand GM1 and also L-fucose. These materials can mimic receptors for cholera toxin, an AB5 toxin secreted by Vibrio cholera, in the intestine. Robinson et al. (Org. Biomol. Chem., 2020, 18, 2739–2746) used the Crich methodology for β-mannosylation to generate a number of β-mannosylceramide analogues that were evaluated for their ability to activate D32.D3 NKT cells and induce anti-tumour activity.

Dimitriou et al. (Org. Biomol. Chem., 2019, 17, 9321–9335) develop synthetic tools towards alginate analogues containing carboxyl bioisosteric hydroxamate functionalities at the C-6 position. These studies shed light on the contribution of both acceptor and donor reactivity underpinning uronate glycosylations.

Nuclear magnetic resonance studies are very powerful tools for understanding carbohydrate function, but require selective labelling strategies that have to be introduced by de novo synthesis. Long et al. (Org. Biomol. Chem., 2020, 18, 4452–4458) report the synthesis of L-[U-13C6]-Fuc labelled type I Lewis b (Leb) structures for use in NMR binding studies with the Blood-group Antigen Binding Adhesin (BabA), a membrane-bound protein from the bacterium Helicobacter pylori. Takahashi et al. (Org. Biomol. Chem., 2020, 18, 2902–2913) describe a rational and efficient synthesis of sialoglycolipids via direct sialylation of a glycolipid at a late-stage, based on their novel sialidation method. The synthetic method enabled the development of GM3 ganglioside analogs with various C5-modifications of the sialosyl moiety. Furthermore, the synthesized analogs were subjected to solid-state 19F NMR analysis on the model membranes revealing the influence of cholesterol on glycan dynamics.

These highlights further demonstrate how the development of improved synthetic methods and strategies to access structurally defined glycans can help us better understand how these complex molecules modulate key biological interactions. They also illustrate how furthering our understanding of the role and function of glycosylation in nature can be used for the development of novel and improved therapeutic and diagnostic tools.


This journal is © The Royal Society of Chemistry 2020