Developments in the Chemical Synthesis of Heparin and Heparan Sulfate

Heparin and heparan sulfate represent key members of the glycosaminoglycan family of carbohydrates and underpin considerable repertoires of biological importance. As such, their efficiency of synthesis represents a key requirement, to further understand and exploit the H/HS structure‐to‐biological function axis. In this review we focus on chemical approaches to and methodology improvements for the synthesis of these essential sugars (from 2015 onwards). We first consider advances in accessing the heparin‐derived pentasaccharide anticoagulant fondaparinux. This is followed by heparan sulfate targets, including key building block synthesis, oligosaccharide construction and chemical sulfation techniques. We end with a consideration of technological improvements to traditional, solution‐phase synthesis approaches that are increasingly being utilised.


Introduction
Carbohydrates are indispensable to glycoconjugate biological function; this is typified by the glycosaminoglycans (GAGs). GAGs are present on most animal cell surfaces and in the surrounding extracellular matrix. They are extremely diverse, containing a linear and structurally heterogeneous anionic glycan chain and impart important biological functions by binding to different growth factors, enzymes, morphogens, cell adhesion molecules, and cytokines. One GAG in particular, heparan sulfate (HS) is involved in mediating mammalian cell function, exemplified by its interaction with growth factors such as fibroblast growth factors (FGFs), a protein family involved in cell proliferation, differentiation, and angiogenesis. [1] HS also mediates many pathological conditions including cancer, [2] Alzheimer's disease, [3] viral infections such as SARS-Cov-2, [4,5] HIV [6] and HSV, [7] alongside numerous bacterial infectivity events. [8] Structurally related to HS is the anticoagulant heparin (H), present within mast cells and considered a highly sulfated, tissue-specific variant of HS.
The chemical structures of H and HS ( Figure 1) consist of repeating disaccharide units, composed of glucosamine (d-GlcN) and a uronic acid (d-GlcA). Its microstructure is diverse: the amino sugar can be N-sulfated (d-GlcNS) or Nacetylated (d-GlcNAc), d-GlcA can be epimerised to l-IdoA and saccharide units are variably substituted with O-sulfate groups, at the C6 (and occasionally C3) of d-GlcN and at C2 of d-GlcA/l-IdoA. HS chains have an average molecular weight of 30 kDa and this, taken in context with the possibilities for functional group variation, presents a huge structural microheterogeneity and enormous scientific challenge in unravelling and understanding the HS structure-tofunction paradigm.
Due to their structural complexity and biological importance, the synthetic challenge surrounding access to (and application of) structurally defined H and HS fragments is significant and of continued interest. In this review we focus on chemical approaches to and methodology improvements for the synthesis of H/HS from 2015 onwards. Contextually, this should be considered alongside the impressive and complimentary advances in chemoenzymatic approaches to access H/HS that have been well reviewed elsewhere, [9,10] alongside GAG synthesis approaches in general. [11] We first consider advances towards the synthesis of the heparin pentasaccharide fondaparinux (and analogues), followed by HS targets, including building block synthesis, oligosaccharide construction and chemical sulfation methods. We end with a consideration of technological improvements to traditional, solution-phase synthesis approaches that are increasingly being utilised.

Developments in the Chemical Synthesis of Heparin-Related Pentasaccharide Sequences
Synthesis of the pentasaccharide anticoagulant drug fondaparinux 1 (Arixtra) by Sanofi and Organon in 2001 represented a landmark achievement in the construction of mimetic heparin oligosaccharide sequences. [12] However, the original 50-step route delivered the target molecule in an overall yield of only 0.1 %, presenting an opportunity and requirement for efficiency improvements. We first discuss recent efforts to develop more efficient synthetic routes to 1 using traditional modular synthetic routes and then explore one-pot, programmable methodologies. [13]

Improving Large-Scale Synthesis of 1
In 2017, Ding and co-workers' reported a practical and efficient large-scale synthesis of 1 using a combination of the Alchemia and Sanofi protocols (Scheme 1). [14] The group identified that the original Alchemia protocol involved simple and practical access to the monosaccharide building blocks 2-3 and, therefrom, trisaccharide donor 6 on a large scale. However, a drawback in the glycosidation of 6 with disaccharide acceptor 7 (to give pentasaccharide 8) was delivery of the desired material in very poor yield. This was suggested to arise from the use of acid labile PMB protecting groups within 6. Comparatively, in the case of the Sanofi protocol, glycosylation of acceptor 12 with trisaccharide 11 was successful and observed with excellent diastereoselectivity in delivering pentasaccharide 13. However, the use Cerny epoxide precursors 9 or 10 were not amenable to large scale application, due to stability issues.
Based on these observations, the group decided to adopt the Alchemia method to access trisaccharide donor 6 for subsequent glycosylation using the Sanofi method (Scheme 2).
The acid labile PMB groups within trisaccharide donor 6 were replaced with acetyl groups to provide trisaccharide 14.
Retaining the use of thioglycoside donors, trisaccharide 14 was successfully coupled with disaccharide acceptor 7 to deliver pentasaccharide 15 in 74 % yield and on > 100 g scale. Pentasaccharide 15 was converted through to 1 in 5 steps in 62 % overall yield and on an impressive > 30 g scale.

Iterative One-Pot Syntheses of 1
In 2018, Wong and co-workers' reported a highly efficient and programmable one-pot method for the synthesis of protected  . [15] The group designed eighteen key intermediates with improved relative reaction values (RRVs) for a one-pot synthesis to then access 6-O sulfation pattern variant pentasaccharides. The building blocks were designed to incorporate the following (a) highly stereoselective α-glycosylation outcomes, influenced by ring protecting groups (N 3 , TBDPS, and Ac) (b) use of TBDPS as a protecting group at the d-GlcN-O-6 position to increase the RRV of donor 16 [16] (c) stereoselective construction of β-1,4-glycosidic linkages using a neighbouring C2-O-benzoyl ester and (d) selective installation of a 6-Osulfate using orthogonal acetyl and levulinic protecting groups. To investigate if it was better to introduce the glucuronic acid as part of the disaccharide building block or through oxidation at the pentasaccharide level, 17 b (RRV = 38.2) was chosen as the thioglycoside donor to be used in the one-pot synthesis of pentasaccharide 19 e. This material was subsequently deprotected, oxidised and esterified to generate pentasaccharide 19 a in 42 % yield. In contrast, when glucuronic acid containing disaccharide acceptor 17 a (RRV = 18.2) was used, pentasaccharide 19 a was furnished in 54 % yield (Scheme 3). Thus, it was concluded here that the pre-glycosylation oxidation was more efficient than a post-glycosylation oxidation (54 % versus 42 % yields respectively).
Following the successful synthesis of pentasaccharides 19 a-d, 19 a and 19 b were further used to obtain regiodefined sulfate derivatives 22 and 26 (Scheme 4). Notably, when introducing O-sulfates, the group found it crucial to first replace the d-GlcN-O-6-TBDPS group in 19 a/b with a benzyl group (20 a/b), to prevent an unwanted insertion of SO 3 into the SiÀ O bond. It was also noted that co-solvent-promoted Obenzylation using Ag 2 O/BnBr in n Hex : CH 2 Cl 2 was found to be an ideal method to avoid acyl migration, ester hydrolysis and sugar decomposition. [17] Recently, Zhao and co-workers' also developed an optimised route towards 1, harnessing a preactivation-based, onepot glycosylation, concomitantly installing two new stereocentres with exclusive α-stereoselectivity (Scheme 5). [18] To synthesise pentasaccharide precursor 30, monosaccharide thioglycoside donor 27 and disaccharide acceptors 28 and 29 were glycosylated in one-pot using p-TolSCl and AgOTf as promotors, delivering 30 in 37 % yield. The target compound 1 was obtained following saponification, O-sulfation, hydrogenolysis and N-sulfation. The total number of synthetic steps was 34, from commercial materials, and the overall yield was 27 %; notably demonstrating a simplified route to 1 with improved synthetic efficiency.
Most recently, Wong and co-workers' reported a potentially scalable and programmable one-pot synthesis of 1 using a [1,2,2] strategy (Scheme 6). [19] This synthesis built on an earlier approach by the group accessing the related anticoagulant drug idraparinux. [20] The building blocks were designed with TBDPS and acetyl at the 6-O positions to support α-selective glycosylation. Initial glycosylations using disaccharide acceptor 32 a (R=Bz, Scheme 6) gave a poor yield of 26 %, however upon switching to acceptor 32 b (R=Ac, Scheme 6), an improved yield of 50 % was achieved. The total synthesis of 1 was achieved in 4.2 % yield over a total of 22 linear steps.

Synthetic Methodology Developments for Heparan Sulfate Synthesis
At a basic level, HS consists of a repeating disaccharide unit of either (-l-IdoA-α-1,4-d-GlcN-α-1,4-) or (-d-GlcA-β-1,4-d-GlcN-α-1,4-). Consequently, efforts in HS synthesis usually focus first on the provision of appropriate disaccharide building blocks, [21] for subsequent iterative oligosaccharide assembly. Additionally, de novo syntheses of mimetic fragments containing just one monosaccharide component (e. g., l-IdoA) have been completed. [22] With appropriately protected oligosaccharide sequences in hand, final functionalisations are made to incorporate essential O-and N-sulfate groups. Accordingly, we have divided recent accomplishments adopting this approach into three categories: i) synthesis of HS building blocks ii) oligosaccharide synthesis and iii) concepts to effect O-and N-sulfation.

HS Building Blocks from Polysaccharide Digestion
Hsieh-Wilson and co-workers' have recently described novel access to key HS disaccharide building blocks 35 a-d (Scheme 7). [23] Notably, two of the four core disaccharides generally required for HS assembly were obtained via digestion The authors were also able to access more complex, orthogonally protected disaccharide building blocks, exempli-fied by 37, in just nine steps from heparin (Scheme 8). Disaccharide 37 then underwent strategic protecting group manipulations to afford both donor 38 and acceptor 39 components. These were then combined to access tetrasaccharide 40 in a 60 % yield (Scheme 8). The incorporation of seven different protecting groups into 40 uniquely allowed each l-IdoA-2-O, d-GlcN-6-O and N-to be unmasked for regioselective sulfation.

R e c o r d R e v i e w T H E C H E M I C A L R E C O R D
(Scheme 9). [24] The optimised reaction conditions used Ph 3 PAuCl/AgOTf and a donor/acceptor ratio of 3 : 1. Given the known low reactivity of uronate d-GlcA/l-IdoA acceptors, high levels (3.0 equiv.) of donor were used to deliver high yields. A range of armed/disarmed donors and acceptors were screened (22 examples). Overall, the reaction yields were high (85-99 %), but donor/acceptor reactivity and donor O-6 protecting group identity had a significant effect on the stereoselectivity. Reactive donors/acceptors gave poorer αselectivity (43 a versus. 43 c) and an ester at O-6 improved αselectivity, suggested to be due to long range participation of such groups (43 a versus. 43 b). [2 + 2] glycosylations to access tetrasaccharides showed an l-IdoA acceptor outperform its d-GlcA counterpart in terms of yield and α-selectivity (43 d versus 43 e).

Programming for Regiodefined HS O/N Sulfation
In 2015, Huang and co-workers' assembled seven hexasaccharides with different sulfation patterns to demonstrate sequence diversity. These materials originated from a single hexasaccharide precursor 47 (Scheme 10). [25] To synthesise hexasaccharide precursor 47, disaccharide thioglycoside donor 44 and disaccharide acceptors 45 and 46 were glycosylated in one-pot using p-TolSCl and AgOTf as promotors, delivering 47 in 67 % yield. Following Levdeprotection at C6 of idose, 47 was oxidised to uronate ester 48. Next, the authors found it necessary to replace the nonreducing end O4-TBDMS group with Bn, prior to sulfation, to deliver 49. Hexosamine nitrogen was unmasked, following azide reduction with 1,3-propanedithiol and the remaining deprotections and O/N-sulfations were completed chemically and enzymatically (chemical steps shown in Scheme 10) to provide a panel of hexasaccharides 51 a-d.
Gardiner and co-workers' also reported a synthetic approach to a small matrix of protected heparin-type oligosaccharides containing orthogonal d-GlcN O-6 protecting groups. [26] Building on their earlier work, [27][28][29][30] this was completed to demonstrate capability in accessing programmability at specific sites, relevant to sulfation or other  In 2018, Boons and co-workers' reported an enzymatic modification of three chemically synthesised hexasaccharides, harnessing a d-GlcN-6-OMe blocking group, to effect regiodefined enzymatic modification and provide a library of 21 hexasaccharides. [31]

An S-Linked HS Disaccharide
In comparison to O-glycosides, S-glycosides can show improved stability towards hydrolytic enzymes. [32] Kovensky and co-workers' reported the synthesis of an S-glycoside analogue of the disaccharide unit of HS and prepared multivalent glycoclusters based around this unit (Scheme 11). [33] The authors opted to use d-Glc instead of d-GlcN, as previous studies showed that such a substitution made only a small difference to biological activity. [34] Initial attempts to synthesise 56 using donors 52 a-b and acceptor 53 gave only trace amounts of the desired product. However, switching the approach to synthesise 56 through nucleophilic substitution of 4-O-triflated galactoside 55 with glycosyl thiol 54 afforded the desired S-glycoside in 79 % yield (α/β, 12 : 82).

Iterative, Reducing End HS Oligosaccharide Synthesis
Gardiner and co-workers' described a novel approach towards the preparation of protected HS oligosaccharides bearing reducing end thioglycoside functionality (Scheme 12). [35] The authors employed a sequential chemoselective glycosylation at the reducing end, additive to their previously reported nonreducing terminus extension via lactone-terminated d-GlcN-l-IdoA intermediates. [36] This new approach introduced chemoselective activation of a reducing end l-IdoA-lactone to enable iterative HS oligosaccharide synthesis, without the need to interconvert the reducing end anomeric group.
Glycosylation of lactone thioglycoside acceptor 58 with GlcN donor 57 gave an intermediate disaccharide lactone which, upon subsequent opening and protection, afforded

Synthesis of Heparin-Related [4] n Oligosaccharides up to 40-Mer
In 2015 Gardiner and colleagues described the chemical synthesis of the longest heparin-related oligosaccharide to date (20-mer) using iterative synthesis and accessed protected oligosaccharides ranging from 16-mer through to the 40mer. [37] Tetrasaccharide donor 64 was utitised in a couplingdeprotection cycle of 2-step homologation reactions, yielding oligosaccharides of increasing [4] n chain length (Scheme 13). The synthesis initiated from O4-protected 12-mer 65 [38] and, following selective deprotection at O4, generated 12-mer acceptor 66 in excellent yield. The subsequent homologation to 16-mer and thence to 20-mer proceeded in very good overall yield (68 % and 79 % respectively) for each 2-step coupling/deprotection sequence. Glycosylations of increasingly long acceptors with donor 64 proved reliable throughout a series of [4] n iterations and afforded 24-mer, 28-mer, 32-mer, 36-mer and 40-mer materials. Oligosaccharide 69 a was deprotected and O-/N-sulfated to afford heparin-like 20-mer 70.

Exploring Functional Capability of Non-Reducing End D-GlcN Residues within HS Sequences
Gardiner's group also described installation of a non-reducing O4 handle in order to introduce conjugation-suitable, terminal functionality to biologically important HS oligosaccharides ( Figure 2). [39] Their approach introduced a d-GlcN O4aldehyde-level bearing unit (using an appropriate disaccharide donor) to remain latent throughout oligosaccharide synthesis. This latent aldehyde tag (LAT) was set to be unmasked at the final stage of synthesis, yielding fully sulfated, heparin-type   This compared to an earlier, reducing end LAT unit introduced by the group to enable radiolabelled HS dodecasaccharide synthesis. [38] The effect of moderating reducing end d-GlcN sulfation pattern was also investigated by the group. [40] Access to scalable, iterative synthesis delivered a site specifically O6 sulfated dodecasaccharide, which was used in conjunction with an earlier reported lower sulfated variant, [27] to then interrogate the effects of both level and site specificity of defined sulfation towards HS-dependant cytokines in vitro and in vivo . [40]

Modular Synthesis of HS Tetrasaccharide Libraries
To broadly explore HS-protein binding, Boons and coworkers' developed a library of 47 HS oligosaccharides for microarray immobilisation and screening. [41] The design conceived a limited number of tetrasaccharides which could be regioselectively O-and N-sulfated to rapidly diversify the final library. A total of 9 tetrasaccharides having 4 different backbone compositions with varying protecting group patterns were synthesised (Scheme 14A). With these in hand, regioselective O-sulfations were completed, giving access to materials of type 82 and 83 (Scheme 14B). Here the 6-OH at the distal d-GlcN 3 residue was found to be more reactive towards sulfation than the proximal d- The group also recently reported an integrated methodology to broadly determine the ligand requirements of HSbinding proteins. [42] This involved firstly completing partial degradation of natural HS, followed by fragment screening against an immobilised HS-binding protein. Compositional analysis using HILIC-HRMS identified moderately sulfated octasaccharide sequences that informed subsequent chemical synthesis of a structurally defined ligand for further structure activity studies. This workflow was demonstrated successfully to establish the ligand requirements of the human Roundabout receptor 1 (Robo 1).

Access to a Sulfonamide-Containing HS Fragment
Hung and co-workers' examined the effectiveness of substrates and inhibitors of human endo-O-sulfatase-1 (Scheme 15). [43] The team synthesised a range of HS oligosaccharides with various chain lengths and N-and O-sulfation patterns and studied their substrate and inhibitor specificity using a competitive fluorogenic substrate, 4-methylumbelliferyl sulfate (4-MUS).

R e c o r d R e v i e w T H E C H E M I C A L R E C O R D
To synthesise trisaccharides 113 a and 113 b, l-ido donor 106 and disaccharide acceptor 107 were glycosylated using NIS and TfOH to furnish α-linked trisaccharide 108. Sequential deacylation using Zemplén's conditions and TEM-PO oxidation converted 108 to lactone 109. Ring opening of 109 delivered methyl ester 110 which was subjected to Osulfation to yield a bis-O-sulfated derivative 111. Desilylation and treatment with either SO 3 · Et 3 N or ClSO 2 NHBn, converted 111 to the sulfate 112 a or sulfonamide 112 b. Following final deprotections, target trisaccharides 113 a and 113 b were accessed in 11 and 5 % overall yields respectively and in 11 steps.

Synthetic HS Sequences Containing GlcNAc and GlcNS
In 2020 Boons and colleagues presented a modular approach to access HS oligosaccharides containing both d-GlcNAc and d-GlcNS units. [44] A previous modular synthetic route from the group provided a diverse range of HS oligosaccharides from two common disaccharide building blocks, 114 and 115 (Scheme 16A), [45] but did not deliver sequences containing both d-GlcNS and d-GlcNAc.
To address this (ultimately, to mimic high and low Nsulfation HS domains) the group synthesised a new pair of disaccharide building blocks 116 a and 116 b (Scheme 16A). Nguyen and co-workers' had established that α-glycosides were selectively formed using a Ni(II) triflate promoted activation of glycosyl donors containing the C2-trifluoromethylphenylmethanimine moiety present in 116 a and 116 b. [46] This, alongside the possibility for its orthogonal removal under mild acidic conditions, presented the prospect to build d-GlcN-azide and d-GlcN-imine containing oligosaccharides.
Accordingly, optimal conditions for the preparation of tetrasaccharide 118 a from donor 116 a and acceptor 117 (Scheme 16B) were established. This afforded the α-anomer of 118 a in an acceptable 58 % yield. Scheme 16B also illustrates the subsequent conversion of 118 a into a final tetrasaccharide 119, containing both d-GlcNS and d-GlcNAc components.
Finally, the team were able to assemble hexasaccharides 120 a-d (Figure 3) in good yields from modular disaccharide donors and conventional acceptors via a series of sequential

Methodology Improvements to Effect O and N-Sulfation
Yu and co-workers' reported an effective method to simultaneously complete O-and N-sulfation under microwave conditions (Scheme 17). [47] Using a cocktail of SO 3 · Et 3 N, NEt 3 and pyridine, a simultaneous O,N-sulfation could be completed in a very short time (15 minutes). In the absence of NEt 3 only O-sulfation took place; hence, the presence of NEt 3 was deemed crucial to obtain simultaneous heteroatom sulfation. Interestingly, using the same cocktail with conventional heating gave only N-sulfation. A rationale for these observations was that as O-sulfonation occured, the system became acidic, leading to protonation of the amino group which was then not free to react with sulfur trioxide. However, the presence of NEt 3 deprotonates the hexosamine nitrogen, thus enabling N-sulfation. The reaction conditions worked well on a range of mono-, di-, tri-and tetrasaccharides (> 90 % yields) and sterically hindered hydroxyl groups (121 a) were sulfated without issue. In 2020, Niu and co-workers' reported an approach to early stage O-sulfation via the sulfur(VI) fluoride exchange (SuFEx) reaction between TMS-protected sugars and aryl fluorosulfates (Scheme 18). [48] Arylfluorosulfates bearing electron withdrawing substituents worked best for sulfate diester formation. The authors also demonstrated a one-pot, in-situ hydroxyl silylation with hexamethyldisilazane (HMDS) and subsequent SuFEx reaction. This one-pot protocol is useful as it circumvents any need to purify intermediate TMS silylethers, which are known to be unstable on silica. The sulfate diesters showed excellent robustness to subsequent acidic, basic, oxidising and reducing reaction conditions. Various glycosyl donors and acceptors decorated with sulfate diesters were successfully employed for glycosylation reactions. Sulfate diester deprotection was achieved using late-stage hydrogenolysis. This early-stage O-sulfation method provides a general, powerful tool for the synthesis of O-sulfated bioactive compounds.

Solid Phase Synthesis of a H/HS Hexasaccharide Precursor
The solid-phase synthesis (SPS) of carbohydrate targets is renowned for its challenges in comparison to solution-phase approaches. Unlike automated peptide and oligonucleotide assembly, oligosaccharide synthesis requires the development of procedures to account for multiple hydroxyl groups of similar reactivity and the requirement for high yielding and stereoselective glycosylation reactions. In 2015, Reichardt and co-workers' developed an effective strategy in this area for the solid-phase assembly of an HS precursor using monosaccharide building blocks (Scheme 19). [49] Previous work by the group established that monosaccharide 123 a had good reactivity and stability for the solid-phase synthesis of a HS trisaccharide precursor. [50] Therefrom, the sequential assembly of a hexasaccharide was performed using monosaccharides 123 a-d. Attachment of a pentyl spacer unit, protected with a 4-hydroxymethylbenzyl N-benzyl carbamate, to a carboxystyrene resin via an ester linkage provided a reliable solid-support (122). Successive capping, delevulination of idose-O4 and alternate glycosylation cycles with donors 123 a, 123 b and 123 c, followed by final glycosylation with d-GlcN donor 123 d furnished H/HS precursor 124 a, after cleavage from the resin. Acetylation of this material and HPLC purification gave the desired hexasaccharide 124 b in an 11 % yield over 14 steps (85 % average yield per step). This effective method delivered significant progress towards a routine and automatable solid-phase approach, notably completable within 2-3 weeks, as opposed to several months in solution phase.

Automated Glycan Assembly of H/HS
An alternative approach towards an HS hexasaccharide precursor was reported by Fascione and Schwörer using automated glycan assembly (AGA) (Scheme 20). [51] AGA eliminates the reliance on manual methods (chemoenzymatic, solution-phase and solid-phase) and provides a rapid approach towards the formation of biologically important carbohydrates. The team reported iterative, automated solid-phase synthesis of an HS hexasaccharide precursor using Glyconeer™, the automated oligosaccharide synthesiser.
Disaccharide 125 a was employed as the initial donor towards the synthesis of hexasaccharide 127. A labile Fmoc protecting group was used as the temporary protecting group  at O-4, to release acceptor capability for d-GlcN for chain elongation throughout the iterative glycosylations. Fmoc deprotection could be conveniently monitored by an inline UV detector, which measured the amount of dibenzofulvene adduct released in solution. A Merrifield resin was chosen as the solid-support and this was additionally functionalised with a photolabile nitrobenzyl ether-based linker (stable to acidic and basic reaction solutions). Excess hydrolysed donor 126 a from glycosylation steps was recycled back to 125 a in solution, at an efficiency of~60 %, highlighting the resourcefulness of AGA. Finally, cleavage from the resin was performed using an in house photoreactor, followed by filtration and evaporation to afford the crude hexasaccharide. Purification via silica gel flash column chromatography afforded 127 in 6 hours with a 30 % yield on 15 mg scale. Interestingly, this approach proposed to complete a latestage oxidation of d-Glc to d-GlcA (after oligosaccharide assembly) and a recent report from Schwörer has indeed highlighted the need for consideration of the complimentarity of this approach versus using d-GlcA disaccharide donors directly. [52]

Conclusion and Outlook
Developments in the field of chemical H/HS synthesis over the past five years have largely focused around and achieved the provision of ever longer and more diverse libraries of sulfation variant ligands. This has been expedited by improve-ments in access to requisite building blocks, notably introducing an upcycling of HS-related polysaccharides to deliver such materials. The matrix of structural possibilities for HS is enormous and often daunting, but the advent of integrated platforms to guide ligand design (and therefore synthesis) will certainly simplify future endeavours here.
Technologies to synthesise these essential glycan fragments have also undergone improvements. From the perspective of medically relevant heparin sequences, iterative, programmable, one-pot solution-phase syntheses to access defined fragment lengths surrounding Fondaparinux have emerged. Coupled with the recent establishments of automated glycan assembly and traditional solid phase synthesis, it is likely that there will be easier and more rapid access to ever more diverse and challenging sequences to support biological requirements in the near future. This must of course be considered in tandem and synergise with achievements made in the chemoenzymatic preparation of H/HS. An additional area that will surely continue to grow as a result of these accomplishments is around the fusion of HS synthesis with protein conjugation/synthesis, which will enable study of the larger and more challenging HS-proteoglycan axis. Such endeavours were recently exemplified through the synthesis and evaluation of a human syndecan-4 glycopeptide. [53,54]