Engineered Carbohydrate-Based Materials for Biomedical Applications: Polymers, Surfaces, Dendrimers, Nanoparticles, and Hydrogels

By Ravin Narain

This book addresses the need for a comprehensive book on the design, synthesis, and characterization of synthetic carbohydrate-based polymeric materials along with their biological applications. The first two chapters cover the synthesis and self-assembly of glycopolymers and different techniques for creating these synthetic polymers. Subsequent chapters account for the preparation of block copolymers, branched glycopolymers, glycosurfaces, glycodendrimers, cationic glycopolymers, bioconjugates, glyconanoparticles and hydrogels. While these chapters comprehensively review the synthetic and characterization methods of those carbohydrate-based materials, their biological applications are discussed in detail.

• Language: English
• Publisher: Wiley
• Release date: Mar 1, 2011
• ISBN: 9781118002247

Book preview

Chapter 1

SYNTHESIS OF GLYCOPOLYMERS

SAMUEL PEARSON, GAOJIAN CHEN, and MARTINA H. STENZEL

Centre for Advanced Macromolecular Design, University of New South Wales, Sydney, Australia

1.1 Introduction

1.2 Synthesis of Vinyl-Containing Glycomonomers

1.2.1 Monomers from Protected Carbohydrates

1.2.2 Monomers from Unprotected Sugars

1.3 Conventional Free Radical Polymerization

1.3.1 Acrylamide Monomers

1.3.2 (Meth)acrylate Monomers

1.3.3 Styrene-Based Monomers

1.3.4 Other Vinyl-Containing Glycomonomers

1.4 Controlled/Living Radical Polymerization

1.4.1 Stable Free Radical Polymerization

1.4.2 Atom Transfer Radical Polymerization

1.4.3 Reversible Addition–Fragmentation Chain Transfer Polymerization

1.5 Ring-Opening Polymerization

1.6 Ionic Chain Polymerization

1.6.1 Anionic Chain Polymerization

1.6.2 Cationic Chain Polymerization

1.7 Ring-Opening Metathesis Polymerization (ROMP)

1.8 Postfunctionalization of Preformed Polymers Using Sugar Moieties

1.8.1 Amide Linkage

1.8.2 Click Approach

1.8.3 Other Nonclick Approaches

1.9 Conclusions

References

1.1 INTRODUCTION

Glycopolymers—synthetic polymers with pendant carbohydrates—have received considerable attention in the fields of polymer chemistry, material science, and biomedicine due to their biocompatibility and their bioactivity. From humble beginnings where glycopolymers were synthesized from vinyl-functionalized sugars via free radical polymerization with little control over the resulting polymer characteristics, glycopolymer synthesis has now developed into a mature area where the control over molecular weight and polymer architecture is routinely sought and indeed achieved. Glycopolymer synthesis has now infiltrated most known techniques of polymer synthesis and is not restricted to controlled radical processes; ionic techniques also provide feasible means to polymerize glycomonomers in a controlled manner.

The aim of this chapter is to provide a comprehensive review of all the techniques that have been utilized for the synthesis of glycopolymers; however, greater emphasis will be placed on the techniques that supercede free radical polymerization. After highlighting the important achievements in the synthesis of glycopolymers via free radical polymerization, focus will turn toward glycopolymer synthesis via the controlled/living radical polymerization processes known as nitroxide-mediated polymerization (NMP), cyanoxyl-mediated polymerization, atom transfer radical polymerization (ATRP), and reversible addition–fragmentation chain transfer (RAFT) polymerization. Glycopolymer synthesis by ring-opening polymerization (ROP), anionic polymerization and cationic polymerization will detail the progress made in the area of ionic polymerization, and a discussion of the work carried out using ring-opening metathesis polymerization (ROMP) will conclude the section on the synthesis of glycopolymers by polymerizing sugar-containing monomers. The functionalization of reactive polymer scaffolds with carbohydrate species will then be discussed as an alternative strategy for synthesizing glycopolymers.

1.2 SYNTHESIS OF VINYL-CONTAINING GLYCOMONOMERS

1.2.1 Monomers from Protected Carbohydrates

The commercial availability of a range of carbohydrates provides access to a wide array of different glycomonomers, and significant efforts have been dedicated to the synthesis of polymerizable vinyl sugars. In an early feature article, Wulff et al. [1] highlighted possible avenues for generating glycomonomers in which an important distinction must be made between protected and unprotected carbohydrates. The choice of employing protected or unprotected sugars is dependent on the ease of stereospecific functionalization of the sugar, the solubility of the monomer and polymer, the potential incompleteness of the removal of the protective group, and the ease of purification. The most common synthetic approaches are outlined below, but they are discussed in more detail elsewhere [1].

1.2.1.1 Reactions Using Isopropylidene-Protected Sugars

Many sugars can easily be protected using acetone to form isopropylidene derivatives. This approach, which is suitable for a range of sugars including glucose, galactose, fructose, and sorbose, allows easy functionalization of the remaining hydroxyl functionality with acrylate [2], methacrylate [3], and 4-vinylbenzyl groups [4].

eps

1.2.1.2 Glycosides from Halogeno Sugars

Glycoside monomer synthesis via the reaction between halogeno sugars and hydroxyl groups of vinyl-containing species has been explored in detailed with varying degrees of success. The starting materials, typically acetylated 1-halogeno sugars, can be expensive or difficult to obtain, but the technique is especially useful for inserting longer spacers between the polymerizable moiety and the sugar. The cleavage of the acetyl protecting groups in alkaline media does not affect the glycoside bond [5].

eps

1.2.1.3 Grignard Reactions

The aldehyde functionality of a sugar molecule can be targeted by Grignard reagents [6]. Prior protection of the remaining hydroxyl groups is essential.

eps

1.2.2 Monomers from Unprotected Sugars

1.2.2.1 Enzymatic trans-esterification

Enzyme-catalysed trans-esterification reactions present a highly efficient and regioselective avenue for obtaining glycomonomers that would otherwise be inaccessible without utilizing protecting groups. The reaction of sugars with vinyl-containing esters in organic solvents is catalyzed by lipases such as Candida antarctica, usually yielding derivatives functionalized in 6-position [7], although efficient functionalization in the 1-position has also been reported [8].

eps

1.2.2.2 Fischer Glycoside Synthesis

Direct monosubstitution in the anomeric (C-1) position without the recourse to protective chemistry can be achieved by the reaction of an excess of hydroxyl groups, such as in hydroxyethyl acrylate, with the sugar in the presence of phosphomolybdic acid as catalyst [9].

eps

1.2.2.3 Synthesis via Barbituratic Acid

Barbituratic acid reacts readily with the C-1 position of the unprotected sugar to generate a reactive salt. Subsequent reaction with bromides such as 4-vinyl benzyl bromide leads to polymerizable monomers. Conversion of the barbituratic acid ring to a diamide further improves water solubility [10].

eps

1.2.2.4 Conversion of Aminosugars

A popular route to glycomonomers is the fast reaction between aminosugars and acyl halides or anhydrides. The high reactivity of the amine group ensures its preferential reaction even in the presence of unprotected hydroxyl groups. Reactions of acryloyl chloride and methacryloyl chloride [11] and also isocyanates [12] and epoxides with various aminosugars have been explored to confer the glycomonomers in high yields.

eps

1.2.2.5 Reaction between Oxidized Sugars and Amines

A range of amide-linked glycomonomers are accessible from sugars that have been oxidized to their corresponding lactones and can therefore be reacted with vinyl-functionalized amines [13].

eps

While these are the most common strategies used for glycomonomer synthesis, other pathways have emerged in recent years such as Cu(I) click chemistry [14]. Some of these approaches are highlighted in this chapter.

1.3 CONVENTIONAL FREE RADICAL POLYMERIZATION

Free radical polymerization is one of the most widely used techniques for making polymers. The polymerization reaction is initiated by free radical initiators and has been used to synthesize linear vinyl saccharide polymers since the 1960s. Despite its disadvantages, such as high polydispersities of the resulting polymers and difficulties in controlling terminal functionalities, the robustness of free radical polymerization has encouraged its widespread use. Indeed, a large number of reports have emerged on the synthesis of glycopolymers via free radical polymerization in both aqueous and nonaqueous media.

Glycopolymers, polymers with pendant sugar groups, were first reported in 1961 when Kimura et al. [15] and Whistler et al. [11a, 11c] reported the free radical homo- and copolymerization of glycomonomers. Significant activity in the field of free radical polymerization of glycomonomers emerged in the following years, which only declined in the late 1990s with the birth of living free radical polymerization techniques. The feature article by Wulff et al. [1] highlighted the body of work and the array of structures. Here, we only highlight some of the latest achievements in this area, mainly publications after 1990.

1.3.1 Acrylamide Monomers

Roy et al. copolymerized 4-acrylamidophenyl-β-lactoside (Table 1.1, entry 1) and acrylamide in water at 90°C in the presence of ammonium persulfate (APS) and tetramethylethylenediamine (TMEDA) [16]. The antigenicity of the resulting carbohydrate copolymer was then demonstrated by agar gel diffusion with peanut and castor bean lectins. Nishimura et al. outlined the synthesis of 3-(N-acryloylamino)propyl 2-acetamido-2-deoxy-β-D-glucopyranoside [17] (Table 1.1, entry 2) and 3-(N-acryloylamino)propyl O-(β-D-galactopyranosyl)-(l-4)-2-acetamido-2-deoxy-β-D-glucopyranoside [18] (Table 1.1, entry 3) and polymerized them in a similar manner to Roy et al. [16].

TABLE 1.1 Glycomonomers Synthesized via Free Radical Polymerization

Table 1-1Table 1-2Table 1-3Table 1-4Table 1-5Table 1-6Table 1-7Table 1-8Table 1-9Table 1-10Table 1-11

Methacrylamide-functionalized mannose monomers (Table 1.1, entries 4–5) were polymerized by Tagawa et al. using a lipophilic azo-initiator containing two long alkyl chains per initiating fragment [19]. Incorporation of the amphiphiles into liposomes generated structures that were able to recognize Concanavalin A (Con A) with little difference observed between the species with varying spacer lengths between the polymer backbone and the sugar residue. Also starting with protected sugars, Carpino et al. reported the synthesis of 2,3,4,6-tetra-O-acetyl-1-O-(4-methacryloylaminophenyl)-β-D-glucopyranoside (Table 1.1, entry 6) and 1-O-(4-methacryloylaminophenyl)-β-D-glucopyranoside (Table 1.1, entry 7), and homopolymerized them with 2,2′-azobisisobutyronitrile (AIBN) as initiator in dimethylformamide (DMF) to afford polymers that were then deprotected with sodium methoxide to give water-soluble glycopolymers [20].

1.3.2 (Meth)acrylate Monomers

Novel (meth)acrylic monomers (Table 1.1, entry 9) bearing a monosaccharide residue were developed by Kitazawa et al. by reacting methyl glycosides with 2-hydroxyethyl acrylate or methacrylate in the presence of heteropoly acid. The monomers were then polymerized in aqueous solution with potassium persulfate as initiator [9].

A galactose-based monomer containing a galactopyranose unit attached through an ester linkage to a vinyl group (Table 1.1, entry 10) was synthesized by Fortes and co-workers and was then copolymerized with ethyl acrylate in DMF under free radical conditions [22].

The protected monomers 2-(2′,3′,4′,6′-tetra-O-acetyl-β-D-glucosyloxy)ethyl methacrylate (Table 1.1, entry 11) and 2-(2′,3′,4′,6′-tetra-O-acetyl-β-D-galactosyloxy)ethyl methacrylate (Table 1.1, entry 12) were polymerized by Cameron et al. in chloroform, and the polymers deacetylated in a mixture of dichloromethane and methanol [23]. The alternative approach for obtaining deprotected polymers was also adopted; entries 11 and 12 were deprotected to give the corresponding monomers 2-(β-D-glucosyloxy)ethyl methacrylate (GlcEMA; Table 1.1, entry 13) and 2-(β-D-galactosyloxy)ethyl methacrylate (GalEMA; Table 1.1, entry 14), which were then polymerized in a water–methanol mixture. Poly(GalEMA) synthesized via the second approach was tested for the binding with peanut agglutinin (PNA) and the thermodynamic binding parameters were calculated [23].

In another report, Cuervo-Rodriguez and co-workers synthesized methacrylate derivatives bearing acetylated glucopyranoside (Table 1.1, entry 15) and galactopyranoside (Table 1.1, entry 16) residues. Glycopolymers were then obtained by homopolymerization (and copolymerization with methyl methacrylate) in chlorobenzene, and their binding to Ricinus communis agglutinin (RCA120) was investigated after deprotection using methoxide [24].

1.3.3 Styrene-Based Monomers

Wulff et al. invested significant effort into the synthesis and polymerization of styrenic glycomonomers (Table 1.1, entries 17–19). The oxidation of sugars to aldehydes and a subsequent Grignard reaction using 4-vinyl-phenylmagnesium chloride was the preferred method for generating the glycomonomers [6b]. Deprotection after the free radical polymerization in water produced polymers with high molecular weights [25]. Narain et al. polymerized the same monomer 4-vinylphenyl-D-gluco(D-manno)hexitol (Table 1.1, entry 17) using 2,2′-azobis-(2-amidinopropan)dihydrochloride (AAPD) initiator in water to obtain copolymers with acrylamide [26]. Thermal properties of the polymers were studied by differential scanning calorimetry (DSC). In the same report, the monomer was copolymerized with styrene in DMF using AIBN as initiator.

Kurth and co-workers prepared a new type of sugar monomer with an oxime linkage, D-lactose-O-(p-vinylbenzyl)oxime (Table 1.1, entry 20) and polymerized it using similar conditions as described above. High-molecular-weight glycopolymers were obtained that displayed narrow polydispersities, a feature attributed to the processing method: precipitation into methanol and two-stage thermal degradation [27]. Similar glycomonomers (Table 1.1, entry 21) containing a urea linkage were polymerized by the same researchers, with the resulting polymers displaying multimodal molecular weight distributions and high glass transition temperatures due to the presence of urea [28].

Kobayashi et al. prepared styrene derivatives with maltose, lactose, and maltotriose substituents on each benzene ring (Table 1.1, entries 22–24) by coupling the corresponding oligosaccharide lactones with p-vinylbenzylamine [13b]. The monomers were then polymerized in either dimethyl sulfoxide (DMSO) using AIBN or in water using potassium peroxydisulfate at 60°C. The maltose- and maltotriose-containing polymers interacted specifically with Con A.

Kobayashi et al. reported the synthesis of other types of p-vinylbenzamide glycoside derivatives (Table 1.1, entries 25–26). They were homo- and copolymerized with acrylamide using 2,2′-azobisisobutyronitrile (AIBN) as initiator in DMSO at 60°C [29]. They also investigated the interaction of the glycopolymers with lectins by means of a two-dimensional immunodiffusion test in agar and inhibition of the hemagglutinating activity. It was found that the specificity of lectins with these glycopolymers was similar to that reported for naturally occurring glycoconjugates and binding between wheat germ agglutinin lectin (WGA) and poly[(p-vinylbenzamido)-β-diacetylchitobiose] was increased by 10³ times compared with that of the oligosaccharide itself. Similar monomers (Table 1.1, entries 27–29) have also been synthesized by Akaike and co-workers and polymerized with AIBN in DMSO [30]. They investigated the specific binding of the glucose-derivatized polymers to the asialogylcoprotein receptor of mouse primary hepatocytes.

More recently, Nishida and co-workers reported the synthesis of novel vinyl monomers (Table 1.1, entries 30–31) bearing galacto-trehalose (GT), a novel class of 1,1′-linked nonreducing disaccharide possessing an α-galactoside epitope. The monomers were copolymerized with acrylamide, and the resulting glycopolymers showed specific binding to α-galactoside-specific proteins (BSI-B4 lectin and Shiga toxin-1) [31].

1.3.4 Other Vinyl-Containing Glycomonomers

Monomers derived from N-acetyl-D-glucosamine (Table 1.1, entry 32) and N-acetyl-D-lactosamine (Table 1.1, entry 33) and chitobiose (Table 1.1, entry 34) were synthesized by Nishimura and co-workers [32]. They were polymerized with acrylamide in deionized water in the presence of ammonium peroxodisulfate (APS) and tetramethylethylenediamine (TMEDA) at room temperature. The synthetic glycopolymers exhibited good solubility in water and specific adhesion to rat hepatocytes. This research group later reported the preparation of a trisaccharide monomer with a Lex structure, n-pentenyl-O-(β-D-galactopyranosyl)-(1-4)-[O-(α-L-fucopyranosyl)-(1-3)]-2-acetamido-2-deoxy-β-D-glucopyranoside (Table 1.1, entry 35) [33]. The monomer was then polymerized under the same conditions as described earlier.

A monomer based on vinyl ketone, 7,8-didesoxy-1,2:3,4-di-O-isopopylidene-α-D-galacto-oct-7-ene-1,5-pyranose-6-ulose (Table 1.1, entry 36) was copolymerized with a range of other monomers to generate hydrophilic surfaces. The reactivity ratios of different glycomonomers with methyl methacrylate (MMA), styrene, and acrylonitrile were determined [34].

Chemoenzymatically synthesized matitol- and lactitol-based glycomonomers (Table 1.1, entries 37–38) were polymerized to afford glucose-containing and galactose-containing polymers. The glycopolymers showed specific biological activities toward Con A and RCA120. Furthermore, positive adhesion to hepatocytes was also observed [35].

The solvent and oxygen effects on the free radical polymerization of 6-O-vinyladipoyl-D-glucopyranose (6-O-VAGlc; Table 1.1, entry 39) were recently investigated by Albertin and co-workers [36]. They polymerized 6-O-VAGlc in water and different alcohols at 60°C in the presence of 4,4′-azobis(cyanopentanoic acid) and found that in all cases long polymerization times (>24 h) were necessary to achieve reasonable conversions, and oxygen removal was critical for the success of the experiments.

1.4 CONTROLLED/LIVING RADICAL POLYMERIZATION

Controlled/living radical polymerization techniques have received widespread interest in recent years due to their ability to produce polymers with precise architectures, predefined compositions, and narrow molecular weight distributions, all of which are inaccessible via conventional radical polymerization. The term living polymerization implies that irreversible chain transfer and termination events are absent, a condition that is not strictly upheld in controlled/living radical polymerization since termination events are unavoidable; however, many of the features commonly associated with living polymerization are still attained [37]. Living characteristics include:

The linear evolution of molecular weight with monomer conversion

A constant concentration of active species, which is indicated by a linear plot of ln([M]0/[M]t) vs. time

Narrow molecular weight distributions, with the polydispersity index (PDI = Mw/Mn) remaining below 1.2; a conventional radical polymerization in which termination occurs exclusively by combination gives a theoretical minimum PDI of 1.5

The ability to polymerize until all monomer is consumed and continue chain growth with the addition of more monomer due to the retention of active end groups

Glycopolymers have been synthesized in a controlled/living fashion using the stable free radical polymerization techniques nitroxide-mediated polymerization (NMP) and cyanoxyl-mediated polymerization, the atom transfer radical polymerization (ATRP) technique, and the reversible addition–fragmentation chain transfer (RAFT) technique.

SCHEME 1.1 Mechanism of NMP.

ch01fig048.eps

1.4.1 Stable Free Radical Polymerization

1.4.1.1 Nitroxide-Mediated Polymerization

Nitroxide-mediated polymerization was developed as a controlled polymerization technique by Solomon et al. in 1985 [38], but it was not until the end of the 1990s that its potential for the synthesis of well-defined glycopolymers was finally realized. Nitroxide-mediated polymerization relies on the reversible capping of growing radical chains with nitroxide species, which are known as persistent radicals (2 in Scheme 1.1). The nitroxides themselves are not capable of initiating polymerization but instead serve to reduce the active radical concentration in the system and thereby limit bimolecular termination events.

The initiating radicals can originate from the fragmentation of an alkoxyamine (as in Scheme 1.1) or can be provided by another source such as a conventional free radical initiator; in the latter case, addition of a nitroxide rather than the corresponding alkoxyamine is the most suitable strategy for controlling the polymerization. Nitroxides are often commercially available as stable radical species. Additives known as accelerators are also commonly employed in NMP and serve the role of regulating the concentration of free nitroxide in the system, which would otherwise build up and retard the polymerization as propagating radicals are inevitably lost through termination. Sulfonic acids, conventional radical initiators, and unstable nitroxides are often used for this purpose [39].

Controlling agents used for the NMP of glycomonomers are shown in Figure 1.1. N1 (TEMPO, a very common controlling agent), N2, N6, and N10 are nitroxides, whereas the remaining species are alkoxyamines, which fragment to generate their corresponding nitroxides. Detailed discussion on the affect of nitroxide structure on NMP kinetics and monomer choice is provided elsewhere [40]. The range of monomers that can be polymerized using NMP is more restricted than other controlled polymerization techniques, and glycopolymer syntheses have largely involved styrenic and acrylic monomers. The majority are protected monomers. Table 1.2 lists the glycomonomers polymerized by NMP.

TABLE 1.1 Nitroxide and alkoxyamine species used for NMP of glycomonomers.

ch01fig009.eps

TABLE 1.2 Glycomonomers Polymerized Using NMP

Table 1-12Table 1-13Table 1-14Table 1-15

Ohno et al. were the first researchers to synthesize well-defined glycopolymers using NMP. N-(p-vinylbenzyl)-[O-β-D-galactopyranosyl-(1-4)]-D-gluconamide (VLA; Table 1.2, entry 1) and its acetylated equivalent (AcVLA; Table 1.2, entry 2) were polymerized in N,N-dimethylformamide (DMF) at 90–105°C using N3 and a radical initiator as accelerator [41]. Initial attempts to polymerize VLA using N1 were unsuccessful due to partial decomposition of the monomer at the temperatures required for sufficiently rapid dissociation of TEMPO (> 120°C). In contrast, N3 proved effective at lower temperatures (90–105°C) due to its higher dissociation constant. Polymerization of VLA using various concentrations of N3 was reasonably well controlled at high N3 concentrations (targeting lower molecular weights); however, high conversions could not be reached, PDIs increased with conversion, and no polymerization occurred when targeting higher molecular weights (target DPn > 100). The addition of the radical initiator dicumyl peroxide (DCP) accelerated the polymerization rate (as expected) without significantly increasing the number of polymer chains in the system; PDIs were unaffected, but the same limiting conversions were observed. Substitution of VLA with its protected analog AcVLA transformed the system into a highly living one, with PDIs consistently around 1.10 and 90% conversion attainable in short polymerization time. Again, the radical initiator greatly enhanced the polymerization rate without affecting the degree of control. Interestingly, the undesirable features of the VLA polymerization are attributed to side reactions, in particular chain transfer to the unprotected hydroxyl groups of the monomer; however, numerous unprotected glycomonomers have since been successfully polymerized using other controlled radical polymerization techniques without such reactions occurring. Despite this discrepancy, almost all subsequent glycopolymer syntheses by NMP have been restricted to protected monomers.

The same authors polymerized AcVLA using N4, an alkoxyamine with two long alkyl chains, to generate a lipophilic glycopolymer (Fig. 1.2) [42]. The polymerization was well controlled in 1,2-dichloroethane at 90°C, and after deprotection of the acetyl groups using hydrazine the resulting polymer formed a stable liposome when mixed with phospholipid. The liposomes specifically bound the galactose-specific lectin RCA120. A similar approach was adopted by Götz et al., who polymerized 3-O-acryloyl-1,2:5,6-di-O-isopropylidine-D-glucofuranose (AIpGlc; Table 1.2, entry 3) in DMF at 105°C using N5 [43]. PDIs remained between 1.06 and 1.20, and the molecular weight increased linearly with conversion. Copolymerization of the glycomonomer with a lipophilic acrylamide monomer was also an effective strategy to generate amphiphilic lipo-glycopolymers.

TABLE 1.2 Lipophilic glycopolymer synthesis by Ohno et al. [42].

ch01fig10.eps

Polymerizations of AIpGlc in p-xylene at 100°C using both N3- and an N2-capped poly(styrene) macroinitiator were deemed living by Ohno et al., although the level of control more closely resembled that of the unprotected monomer VLA than the protected AcVLA; even in the best case the PDI climbed beyond 1.6 at only 25% conversion, and high molecular weights were unattainable [44]. Despite the questionable livingness of the process, deprotection of the isopropylidine groups using formic acid generated amphiphilic diblock copolymers that exhibited microphase separation when cast as thin films.

Chen and Wulff produced diblock copolymers of styrene and 2,3;4,5-di-O-isopropylidine-1-(4-vinylphenyl)-D-gluco(D-manno)pentitol (Table 1.2, entry 4) and showed that either block could be polymerized first [45]. N7 (a TEMPO-derived alkoxyamine) was chosen as the mediating agent, and in contrast to the observations of Ohno et al. using VLA, no thermal decomposition of this monomer was noted even after polymerizing for 48 h at 130°C. The glycopolymer PDIs were around 1.33, and increased to 1.35–1.80 with the addition of different length poly(styrene) segments. After deprotection, films cast from the block copolymers showed surface properties that varied with the block lengths of the two components. A more extensive kinetic investigation into the N7-mediated polymerization of this and three other styrene-based glycomonomers 2,3-isopropylidine-1-(4-vinylphenyl)-D-threo(D-erythro)triol (Table 1.2, entry 5), 1,2;3,4-di-O-isopropylidine-1-(4-vinylphenyl)-D-glycero(L-glycero)-α-D-galactopyranose (Table 1.2, entry 6), and 2,3;4,5-di-O-isopropylidine-1-(4-vinylphenyl)-D-manno(D-gluco)-hexulo-2,6-pyranose (Table 1.2, entry 7), was performed by the same authors [46]. Polymerizations were conducted at 130°C in bulk without any monomer degradation; thermogravimetric analysis confirmed that the monomers are thermally stable up to 150°C. The polymerization of all four monomers saw a linear increase in Mn with conversion, and PDIs were low (1.24–1.28) when targeting low molecular weights, except in the case of entry 7 whose PDI was inexplicably large (> 1.50). High conversions were attainable for all but monomer 6, which was attributed to higher steric hindrance around its vinyl group. It is interesting to note that the presence of a free hydroxyl group in each monomer does not appear to have compromised the degree of control.

TABLE 1.3 Triblock copolymer synthesis by Narumi et al. [47].

ch01fig011.eps

Narumi et al. extended the NMP process to synthesize triblock copolymers of 4-vinylbenzyl glucoside peracetate (Table 1.2, entry 8) and styrene using N8, a difunctional TEMPO-based mediator [47a]. Polymerization of the glycomonomer was performed in chlorobenzene at 125°C for 5 h using (1S)-(+)-10-camphorsulfonic acid (CSA) as an accelerator to give a well-defined polymer (Mn 8.5 kDa, PDI 1.09) whose two terminal TEMPO groups were used to chain extend with styrene (Fig. 1.3). Small low-molecular-weight shoulders were observed in the size exclusion chromatography (SEC) traces when the polymerization time extended beyond 3 h, but restricting the polymerization time allowed generation of well-defined triblocks (Mn 26 kDa, PDI 1.17). The same monomer was used in the synthesis of star poly(styrene) with a glycopolymer core [47b]. Chains of well-defined poly(styrene) with TEMPO end groups were linked together by chain extending using divinylbenzene with either 4-vinylbenzyl glucoside peracetate or 4-vinylbenzyl maltohexaoside peracetate (Table 1.2, entry 9), thereby incorporating sugar groups into the core. The final star polymers had on average 18 arms, PDIs around 1.35, and after alkaline deprotection of the sugar groups were capable of encapsulating water-soluble molecules in chloroform solution.

Ting et al. published the most recent report on NMP of a sugar monomer, namely 2-(2′,3′,4′,6′-tetra-O-acetyl-β-D-galactosyloxy)ethyl methacrylate (AcGalEMA; Table 1.2, entry 10), which is the first report detailing the polymerization of a methacrylic sugar monomer by NMP [48]. Difficulties in polymerizing methacrylates by NMP (such as enhanced disproportionation between growing radicals and nitroxide) were overcome by copolymerizing with a small proportion of styrene, which reduced the average activation–deactivation constant and thereby suppressed the irreversible termination events. The commercially available alkoxyamine N9 (which contains an N10 nitroxide end group) was used to mediate the polymerization at 85°C in dioxane to generate the random copolymer poly(AcGalEMA-co-styrene), which exhibited linear evolution of molecular weight with conversion and a final PDI of 1.26. However, chain extension of this block with styrene at 115°C resulted in some low-molecular-weight tailing, which was attributed to loss of alkoxyamine end groups. Alternatively, chain extension of a well-defined poly(styrene) homopolymer with glycomonomer/styrene furnished the block copolymer with a PDI of 1.15. The polymerization was commenced at 120°C to encourage cleavage of the poly(styrene)-N10 macroinitiator before reducing the temperature to 85°C for the remainder of the polymerization, since the activation–deactivation constant for the methacrylate-based alkoxyamine is higher than that for styrene. Deprotection of the galactose acetyl groups afforded amphiphilic block copolymers capable of forming bioactive micelles and porous films.

1.4.1.2 Cyanoxyl-Mediated Polymerization

Cyanoxyl radicals, NCO•, were found by Druliner to act as persistent radicals capable of mediating the polymerization of acrylates, methacrylates, and methacrylamides [49]. The NCO• radicals are generated in situ by the one-electron reduction of cyanate anions using arenediazonium ions (e.g., p-chlorobenzenediazonium cations, Scheme 1.2). The aryl radicals produced simultaneously initiate polymerization, and their associated cyanoxyl radicals (which are not capable of initiating polymerization) act as reversible capping agents to form dormant species that can be reactivated by homolytic cleavage of the C–O bond.

SCHEME 1.2 Mechanism of cyanoxyl-mediated polymerization.

ch01fig049.eps

The C–O bonds in these systems are more easily cleaved than the equivalent bonds in alkoxyamines, which offers a distinct advantage over NMP in the polymerization of thermally sensitive glycomonomers; cyanoxyl-mediated polymerizations are typically conducted from 25 to 70°C. In addition, tolerance to a wide range of functional groups including hydroxyls, amines, and carboxylic acids broadens the scope of potential monomer–solvent combinations. It must be noted, however, that while cyanoxyl radicals introduce some degree of mediation, the inability to target predefined molecular weights by stopping the polymerization at a particular time precludes cyanoxyl-mediated polymerization systems from being considered truly living/controlled. Low initiator efficiency due to irreversible termination of aryl radicals in the initial stages of the polymerization is responsible for this lack of predictability in molecular weights [50].

Chaikof's work group pioneered the application of cyanoxyl-mediated polymerization for the synthesis of glycopolymers from unprotected monomers. Cyanoxyl radicals were generated in situ by the reduction of cyanate using p-chlorobenzenediazonium cations, simultaneously producing p-chlorobenzene radicals that initiated polymerization (Scheme 1.2). The glycoolymers were designed as heparin and heparin sulfate mimetics, with later publications focusing on their various biomimetic capabilities.

In their early work, nonsulfated and sulfated N-acetyl-D-glucosamine-based monomers with two different spacer arms between the vinyl group and the sugar (Table 1.3, entries 1–2) were copolymerized with acrylamide to generate statistical copolymers with different proportions of sugar groups [51]. The molecular weights increased with conversion and the final statistical copolymers displayed PDIs ranging from 1.10 to 1.47. A higher glycomonomer feed ratio did compromise the level of control, presumably due to the lower reactivity of the unactivated vinyl group. Indeed, further investigation revealed that homopolymerization of these alkene-derivatized monomers was not possible, prompting the development of acrylate alternatives (Table 1.3, entries 3–4) that were capable of homopolymerization [50]. It is interesting to note, however, that copolymers of these nonsulfated and sulfated acrylates with acrylamide generally displayed higher PDIs than those synthesized using the alkene glycomonomers, potentially due to the higher frequency of termination events in acrylate systems.

TABLE 1.3 Glycomonomers Polymerized by Cyanoxyl-Mediated Polymerization

Table 1-16Table 1-17Table 1-18

Extension of the same synthetic protocol to include alkene-derivatized nonsulfated and sulfated lactose monomers (Table 1.3, entries 5–6) broadened the library of copolymers to test for heparin-like abilities [52]. In vivo, heparin sulfate is known to bind to fibroblast growth factor 2 (FGF-2) and promote its binding to FGF receptor-1 (FGFR-1). When tested in this capacity, polymers containing either sulfated N-acetyl-D-glucosamine or sulfated lactose groups showed a pronounced enhancement in the binding of FGF-2 to FGFR-1, particularly for the lactose candidate, compared to their nonsulfated counterparts. Interestingly, the linker length had no influence. Further investigation into the chaperone ability of sulfated lactose/acrylamide copolymers found that a particular copolymer composition (Mn 9.3 kDa, 10 mol% glycomonomer) was as effective as heparin sulfate in protecting FGF-2 from enzymatic, acidic, and thermal degradation and for promoting binding to FGFR-1 [53].

Acrylamide-based lactose monomers (Table 1.3, entries 7–8) were copolymerized with acrylamide in an identical fashion to above, giving statistical copolymers with PDIs ranging from 1.19 to 1.47 [54]. Polymers containing sulfated lactose groups demonstrated a significant anticoagulation ability, whereas nonsulfated lactose polymers and neither sulfated nor nonsulfated N-acetyl-D-glucosamine-containing polymers showed any effect. Interestingly, sulfated lactose homopolymers were outperformed by their copolymers with acrylamide.

End-functionalization of glycopolymers synthesised by cyanoxyl-mediated polymerization was also reported by Chaikof's work group. Two biotin-derivatized arylamine initiators (each with different spacer lengths between the biotin and aryl groups) were used to generate biotin-terminated lactose/acrylamide copolymers that interacted strongly with streptavidin (irrespective of spacer length) [55]. Elaborating on this concept, a series of arylamine initiators derivatized with alkoxy, amino and Fmoc-protected amino, biotinyl, hydrazido, and carboxyl functionalities were effective in mediating the familiar glycomonomer/acrylamide copolymerizations (Fig. 1.4) [56]. The unprotected amino species and the hydrazine-functionalized species both gave lower yields, a feature that was attributed to quaternization of the amino group that suppressed radical formation. The series of end-functionalized glycopolymers were promising candidates for various bioconjugation reactions.

Figure 1.4 Synthesis of biotinylated statistical copolymers by Hou et al. [56].

ch01fig012.eps

1.4.2 Atom Transfer Radical Polymerization

Atom transfer radical polymerization (ATRP, Scheme 1.3) is a controlled radical polymerization technique that was developed independently by the research groups headed by Sawamoto [57] and Matyjaszewski [58] in 1995. An ATRP system contains a halogenated organic compound (initiator) R–X, a transition metal Mn, which can increase its oxidation number, a complexing ligand L to stabilize the metal, and monomer M. Initiation involves abstraction of the initiator's halogen atom by the metal complex, which simultaneously undergoes a single electron oxidation. This gives an organic radical R• and a new metal complex Mn+1X/L in which the metal's oxidation number has increased by 1. The radical can add monomer units, thereby initiating chain growth, before abstracting the hydrogen atom from the metal complex and restoring its dormant state Pn–X. This halogen-capped polymer chain assumes the same role that the initiator occupies in the initiation step; it remains in its dormant state until activated by the metal complex to reform the radical Pn•, which can then add a few more monomer units before being deactivated. These atom transfer equilibria lie heavily toward the dormant species, which reduces the effective radical concentration and thereby limits bimolecular termination. Termination still occurs but is greatly suppressed, imparting living characteristics on the polymerization process.

SCHEME 1.3 ATRP mechanism.

ch01fig050.eps

Atom transfer radical polymerization has been successfully performed using a variety of transition metals, but copper (Cu) complexes have proven to be the most efficient and versatile. The choice of ligand influences the relative rates of activation and deactivation and, therefore, the degree of control over the polymerization. Nitrogen-containing multidentate ligands are commonly employed for Cu-mediated ATRP, and those relevant to glycopolymer synthesis are shown in Figure 1.5. Figure 1.6 shows the initiators relevant to the present review. For simplicity, the bold code will be used in text to refer to each ligand and initiator, rather than including its full or abbreviated name. In cases where the initiating group was attached to another structure, such as a premade polymer or a surface, the structure will be stated followed by the code of the initiating group. For example, a poly(ɛ-caprolactone) (PCL) species end-functionalized with an ethyl 2-bromoisobutyryl group A1 will be denoted PCL-A1. The glycomonomers polymerized using ATRP are summarized in order of appearance in Table 1.4, including their polymerization conditions.

TABLE 1.5 Ligands employed Cu(I) catalyst systems for glycopolymer synthesis using ATRP.

ch01fig013.eps

TABLE 1.6 Initiators employed for glycopolymer synthesis using ATRP.

ch01fig014.eps

TABLE 1.4 Glycomonomers Polymerized Using ATRP

Table 1-19Table 1-20Table 1-21Table 1-22

1.4.2.1 Linear Polymers

Ohno et al. from Kyoto University reported the first ATRP of a sugar monomer by polymerizing 3-O-methacryloyl-1,2:5,6-di-O-isopropylidine-D-glucofuranose (MAIpGlc; Table 1.4, entry 1) using a CuBr/L3 catalyst in veritrole containing A1 as initiator at 80°C [59]. The first-order kinetic plots for the polymerizations of MAIpGlc were approximately linear, which means that the system obeys first-order kinetics with respect to monomer concentration. However, a twofold increase in the initiator concentration (holding all else constant) had a negligible effect on the rate of polymerization rp, whereas a doubling of the activator concentration increased rp by a factor of 3. Previous studies using styrene indicated that rp is first order with respect to both the dormant alkyl halide concentration and the activator concentration [60]. Despite these unexpected kinetic features, the polydispersity index (PDI) of the resulting glycopolymers was less than 1.30, indicating that the polymerizations were reasonably well controlled.

Liang and co-workers were one of the first research groups to generate well-defined glycopolymers using ATRP [61]. The polymerization of 2-(2′,3′,4′,6′-tetra-O-acetyl-β-D-glucopyranosyloxy) ethyl acrylate (AcGEA; Table 1.4, entry 2) was performed in chlorobenzene at 80°C using CuBr/L1 catalyst and A2 initiator. The polymerization obeyed first-order kinetics up to 70% conversion, after which the rising viscosity caused deviation from linearity. The measured molecular weights agreed with theoretical values and increased linearly with conversion, and the PDIs remained less than 1.4 throughout. Chain extension of a poly(styrene)-A2 macroinitiator under the same conditions generated well-defined block copolymers (PDI 1.31–1.37) as shown in Figure 1.7 [62]. Deprotection of the hydroxyl groups on the sugar units using sodium methoxide afforded amphiphilic block copolymers that self-assembled in water to form spheres, rods, vesicles, tubules, and finally large compound vesicles as the polymer concentration increased from 0.1 to 2.0 wt% [63]. The CuBr/L5-catalyzed ATRP of the same monomer using a poly(ethylene glycol)-A1 (PEG-A1) macroinitiator furnished a well-defined hydrophilic-hydrophilic block copolymer (after deprotection) that assembled into aggregates capable of binding concanavalin A (Con A) [64].

TABLE 1.7 Synthesis of amphiphilic block copolymers of poly(styrene)-b-poly(2-(2′,3′,4′,6′-tetra-O-acetyl-β-D-glucopyranosyloxy) ethyl acrylate) by Li et al. [63a].

ch01fig015.eps

Narain and Armes synthesized the two unprotected monomers 2-glucon-amidoethyl methacrylate (GAMA; Table 1.4, entry 3) and 2-lactobionamidoethyl methacrylate (LAMA; Table 1.4, entry 4) by reacting 2-aminoethyl methacrylate with D-gluconolactone [65]. Homopolymerizations were performed at 20°C using CuBr/L1 catalyst and a PEO-A1 macroinitiator in different methanol–water solvent combinations. A drastic increase in polymerization rate and a corresponding decline in control was observed when the polymerization medium was changed from pure methanol to 9:1 methanol–water, 3:2 methanol–water, and finally pure water (Fig. 1.8); full conversion was reached in 15 h using methanol (final PDI of 1.19) and only 30 min in water (final PDI of 1.82). [65a] The polymerization of LAMA proceeded with good control in 3:2 methanol–water (final PDI of 1.10) but not in pure water (final PDI of 1.78); its highly polar nature prohibited