By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Oxazolines are five-membered cyclic compounds having an endo-imino ether (NCO) group. They belong to one of the classes of cyclic imino ethers, which contain cyclic endo- and exo-imino ethers of 5-, 6-, and 7-membered rings. An early stage of chemistry of oxazolines including their preparation, reactions, and applications was extensively reviewed previously.1, 2 Historically, oxazolines have been known since the end of 19th century. The first attempt of the preparation was made in 1884 via dehydrohalogenation of allylurea, resulting in detection of a new cyclic compound but failing to deduce its correct chemical structure.3 Five years later, the first successful synthesis of oxazolines was reported and then the chemistry of oxazolines started to be increasingly investigated.4
Oxazolines have three structural isomers; 2-oxazoline, 3-oxazoline, and 4-oxazolines. Among them, 2-oxazolines are by far the most well-known and extensively studied. The formation of an oxazoline ring was postulated in the cyclodehydration at peptide linkages constituted from the α-amino-β-hydroxy acids, serine, or threonine.5 2-Substituted-2-oxazolines were formed via the cyclization of an acyl derivative of α-amino acid with dehydration, the reaction of an amino alcohol with a carboxylic acid, the cyclization of a β-haloalkylamide via dehydrohalogenation, and so forth.1, 2 The preparation of unsubstituted 2-oxazoline was first reported in 1938.6 Chemistry of these hetero-cyclic compounds was widely studied by reactions with a variety of compounds. The chemistry has been extended to apply for the preparation of various functional materials such as coatings, surfactants, pharmaceuticals, gasoline additives, foam stabilizers, protective films, photographic materials, agricultural reagents, adhesives, binders, and plasticizers.
One of the big areas of the 2-oxazoline chemistry is the polymer formation, which started in the mid 1960s through the publication by four independent groups.7–13 With a cationic catalyst or initiator, 2-substituted-2-oxazolines (ROZOs) underwent the cationic ring-opening polymerization (CROP) to give poly(N-acylethylenimine)s (PROZOs) (Scheme 1). During the CROP, isomerization took place from the imino ether group of the monomer to a more thermally stable N-acyl group of the polymer. CROP of unsubstituted 2-oxazoline (OZO, R = H, Scheme 1) was first achieved in 1972, to afford poly(N-formylethylenimine).14 The product polymers are conveniently used as a starting material for the preparation of linear polyethylenimine (LPEI) via acid hydrolysis. This method can be compared with a CROP of unsubstituted ethylenimine, which produces a highly branched polyethylenimine (BPEI). The CROP of 2-oxazolines including other cyclic imino ether monomers has been extensively studied for these 4 decades and comprehensively reviewed from time to time.15–21
On the other hand, we have been developed new in vitro synthesis of natural and unnatural polysaccharides having complicated structures via enzymatic polymerization.22–31 One of the tough problems was the synthesis of chitin, one of 2-amino-2-deoxy sugar unit-containing polysaccharides; its first successful synthesis was achieved by the enzymatic ring-opening polyaddition (EROPA) of a chitobiose 2-oxazoline monomer (without protection) catalyzed by chitinase enzyme in one-step reaction (Scheme 2), in which the stereochemistry and regioselectivity were perfectly controlled.32–37 The monomer was found in a way like serendipity in 1995, to be a very reactive substrate for the enzymatic catalysis. At that time we speculated the reason of the high reactivity of the monomer and proposed a new concept of a “transition-state analogue substrate” (TSAS) monomer.34 The TSAS concept has been developed on the basis of the fundamentally important characteristics of enzymatic reactions, in which an enzyme stabilizes the transition-state to lower the activation energy,35, 36 accelerating the reaction in tremendous rate enhancement normally in 106–1012 fold or even 1020 fold.28, 30, 31, 37 And, the concept is well accepted presently.
The enzymatic chitin synthesis together with the TSAS concept opened a new door for the synthesis of amino-polysaccharides (mucopolysaccharides) having N-acetyl group at the 2-position of the sugar unit, which are widely found polysaccharides in nature. Scheme 3 indicates that the 2-methyl-2-oxazoline (MeOZO) is a latent N-acetyl group leading to the product chitin, as shown by a simple general reaction of MeOZO with an electrophile HX, which explains the formation of N-acetyl group in chitin.
Thus, both the above CROP and EROPA involve 2-oxazolines commonly in monomer structures. Chemistry of 2-oxazolines is involved in both polymerizations, and therefore, it is a crossing where CROP and EROPA meet in the polymer synthesis field. The present article describes recent developments of CROP of 2-oxazoline monomers for functional polyethylenimine derivatives and EROPA of sugar 2-oxazoline monomers for the synthesis of a family of various 2-amino-2-deoxy sugar unit-containing polysaccharides.
CATIONIC RING-OPENING POLYMERIZATION OF 2-OXAZOLINE MONOMERS
A general scheme of the CROP mechanism of ROZO is well established as given in Scheme 4, where an electrophile like alkylating agent R′X is used as initiator. Besides alkylating agents, protic acids and Lewis acids are widely used as catalysts or initiators for the CROP. In the initiation (a), an oxazolinium species is formed and another monomer ROZO nucleophilically attacks the oxygen atom of the oxazolinium to open the ring with involving an isomerization of the imino group to the N-acyl group. In the propagation (b), ROZO monomer attacks a propagating oxazolinium to open the ring involving the isomerization, giving a one-more-unit-elongated propagating species, and this type of reaction is repeated to produce a poly(N-acylethylenimine).15–21 The product amide polymers are often viewed as “pseudopeptides” due to the structural analogy as well as the nontoxic nature.18, 19, 21, 38, 39
Polymerization profiles depend upon the so-called reaction parameters such as nature of substituent R, position of the substituent, nature of catalyst or initiator, reaction solvents, reaction temperature, and so forth. Polymerization is often of living character, and then, block copolymers can be easily prepared. Functional group(s) can be introduced into the polymer chain end(s), at the initiating step by using a (macromolecular) initiator having the functional group(s) (initiator method) or at the termination step by using a (macromolecular) terminator having such group(s) (terminator method). Accordingly, functional poly(N-acylethleneimine)s like macromonomers, telechelics, block copolymers, and graft copolymers are obtained readily.
In some cases, depending upon the combination of monomer, initiator, temperature, and solvent, the polymerization becomes to involve a covalent propagation species, due to instability of the propagating 2-oxazolinium ion, particularly due to the highly nucleophilic nature of X− (Scheme 5). The polymerization rate then becomes much slower. The quantitative discussions on these points were previously made.15–19
Since CROP of 2-oxazoline monomers has been often reviewed, the present article will focus mainly on the recent developments particularly after the latest Highlight Article.21 As seen in the following, research in this area has been actively conducted continuingly, in particular, for these several years.40 The studies include fundamental works as well as new application works based on recently developed chemistry and techniques.
Polymerization Reactions and Methods
Reaction rate of CROP of ROZOs is relatively slow, normally taking hours to complete the reaction. To make the reaction faster, microwave irradiation was applied and found to accelerate the reaction very much, which has been extensively developed by Schubert et al. The first article by them in 2004 reported that compared with the conventional thermal heating, CROP of 2-ethyl-2-oxazoline (EtOZO) by methyl tosylate (MeOTs) catalyst in acetonitrile at 80–180 °C proceeded very fast under microwave irradiation in a living manner; rate-acceleration was a factor of 350 (reaction time 6 h→1 min) with Mn up to 9000 and a polydispersity index (PDI) of 1.1–1.2. A slightly yellow color of the product solution suggested side-reactions. It was also suggested that the rate enhancement is attributed solely to a temperature effect. The homogeneous microwave irradiation allowed conducting the CROP in bulk or with drastically reducing the solvent amount (green chemistry).41
Similarly, CROP of 2-phenyl-2-oxazoline (PhOZO) was much accelerated in rate with MeOTs catalyst in acetonitrile at 125 °C, showing a living nature to give PPhOZO almost quantitatively with Mn = 10,700 (PDI = 1.02) after 90 min. A strong microwave effect was suggested for a large rate enhancement. It is considered that microwave-assisted polymerization is a rapid, environmental friendly method.42 Whether or not the microwave effect is involved was investigated by CROP of also PhOZO initiated by MeOTs with carrying out both under the microwave irradiation and the conventional thermal heating. Then, it was concluded that the rate acceleration of CROP of PhOZO by microwave irradiation results from thermal effects only.43, 44
Microwave-assisted CROP of MeOZO, EtOZO, PhOZO, and 2-nonyl-2-oxazoline (NonOZO) was studied in detail with MeOTs initiator in acetonitrile at 80–200 °C. The polymerization proceeded with the first-order kinetics of the monomer consumption, exhibiting living nature of the reaction. Rate enhancement was up to 400 folds, yielding the polymers whose degree of polymerization (DP) higher than 300 with PDI <1.20. The fast, direct, and noncontact heating by the microwave irradiation permitted the CROP to be carried out in a highly concentrated solution or in bulk.45
By utilizing the rapid rate of reaction and the living nature of CROP of 2-oxazolines, diblock copolymers were prepared from the above four monomers under microwave irradiation in acetonitrile at 140 °C, which gave a library of four chain-extended homo-PROZOs and 12 diblock co-PROZOs. The reaction to yield a total number of 100 (50 + 50) monomer units was designed and thus-obtained 16 polymers showed a narrow PDI (<1.30). All polymers were stable up to 300 °C.46 Further, microwave-assisted polymerization was extended to synthesize amphiphilic triblock terpolymers to prepare a library of desired 30 terpolymers via CROP of the above four monomers, MeOZO, EtOZO, PhOZO, and NonOZO. The polymerization was performed in acetonitrile at 140 °C in a three step sequential procedure to complete the monomer consumption ranged from 13.2 to 61.6 min. The terpolymers exhibited narrow PDI (∼1.3) and showed minor deviations from the targeted monomer ratio of 33:33:33. The Tg values ranged from 50 to 100 °C depending on the incorporated monomers. Some amphiphilic terpolymers containing two hydrophilic and one hydrophobic blocks formed micelles, whose size ranged between 7 and 18 nm.47 Similarly, tetrablock ter- and quarterpolymers were prepared from the same four monomers to design the polymers having the monomer units as 25:25:25:25. Well-defined tetrablock polymers were obtained, having the structures of ABCA, ABCB, ABCD, and ABDC, where A (MeOZO), B (EtOZO), C (PhOZO), and D (NonOZO) blocks. PDI valued 1.20 or lower without NonOZO units and 1.38 or lower with NonOZO units. Aqueous micelles were formed and their size varied around 5–40 nm depending on the preparation procedure. Surface energy due to amphiphilic nature of the polymers was also measured.48
Under microwave irradiation, one-pot statistical copolymerization of PhOZO with MeOZO or EtOZO produced quasi-diblock copolymers.49 This was because of a big difference in the monomer reactivity ratio previously reported as rMeOZO = 10 ± 1 and rPhOZO = 0.02 ± 0.02. Cationic ring-opening copolymerization of EtOZO and 2-“soy alkyl”-2-oxazoline (SoyOZO) was performed with microwave irradiation with MeOTs initiator, where SoyOZO monomer is derived from soy-bean fatty acids (Scheme 6).50, 51 Copolymers obtained were of statistical structure, showing Mn values of 6–13 × 103 with PDI of 1.2–1.5. Both monomers possessed close monomer reactivity ratios; rEtOZO = 1.4 ± 0.3 and rSoyOZO = 1.7 ± 0.3. By utilizing the unsaturated group in the soy alkyl of the copolymer, UV-irradiation gave a crosslinked material.
Structure-property relationships were investigated in detail for the random and block copolymers from EtOZO and NonOZO monomers.52 Synthesis of the random copolymers were conducted with microwave irradiation in acetonitrile, and synthesis of the block copolymers with the conventional heating, both by MeOTs initiator. Under microwave irradiation conditions, both monomers showed close monomer reactivity ratio values; rEtOZO = 1.07 ± 0.07, rNonOZO = 1.05 ± 0.14. All the product copolymers were shown to have well-defined structure. The organization of the monomer units in the copolymers involves a significant effect on the surface energy, thermal transition, and mechanical properties of the copolymer material, showing that the structural control over monomer distributions is an excellent way to tune these properties.
Microwave-assisted one-step synthesis of EtOZO/2-(m-difluorophenyl)-2-oxazoline (F2PhOZO) and EtOZO/PhOZO copolymers was conducted via CROP with MeOTs initiator in acetonitrile at 140 °C.53 The one-step CROP reaction produced the well-defined statistical gradient copolymers, with DP unit ratios; EtOZO (20):F2PhOZO [or PhOZO] (40), and EtOZO (70):F2PhOZO [or PhOZO] (30). These gradient copolymers are good surfactants and formed micelles having a diameter of 9–13 nm determined by atomic force microscopy (AFM). Microwave-assisted CROP of PhOZO and F2PhOZO was possible in ionic liquids at 140 °C.54
Kinetic study of microwave-assisted CROP of two monomers, EtOZO and PhOZO, was performed at 160 °C in acetonitrile with four initiators, benzyl chloride (BnCl), BnBr, BnI, and BnOTs, to examine the effect on the reaction rate of the four different counter anions.55 Values of the rate-constant of propagation (kp, 10−3 L/mol/s) were for EtOZO: BnCl (2); BnBr (114); BnI (434); and BnOTs (350), and for PhOZO: BnCl (0); BnBr (2); BnI (7); and BnOTs (98). Percentages (%) of cationic propagating species were also determined by 1H NMR for EtOZO: BnCl (0); BnBr (96); BnI (100); and BnOTs (100), and for PhOZO: BnCl (-); BnBr (0); BnI (25); and BnOTs (100). Namely, the CROP of EtOZO with BzCl and of PhOZO with BzBr proceeded via covalent propagating species in 100%. The nucleophilicity of X− is in the order: Cl− > Br− > I− > TsO− (see Scheme 5). The polymer composition from copolymerization of EtOZO with PhOZO was not so much affected by different initiators.
Various monomers (ROZOs) were prepared with changing the structure of 2-substituents, and their CROP was carried out with MeOTs initiator under microwave irradiation to give the corresponding polymers.56 Scheme 7 shows structures of 12 substituents and results of their kinetic study by kp values (L/mol/s) in the parenthesis carried out at 140 °C in dichloromethane. The kp values are in the range from 1 × 10−3 to 241 × 10−3 depending on the structure/substitution of ROZOs. An electron-rich aromatic group (furan) shows a higher polymerizability close to an alkyl-substituent, whereas bulky groups decrease the kp value. Thermal properties and surface energy of these polymers were examined. For reference, the kp value of MeOZO (2050 × 10−4) can be interestingly compared with that (1.17 × 10−4) at 40 °C in acetonitrile.57
A pendant alkyne group-containing PROZO was prepared by CROP of 2-(4-pentynyl)-OZO (PynOZO) monomer, both via its homopolymerization and copolymerization with MeOZO or EtOZO. The pendant 4-pentynyl group of the water soluble copolymer unit, that is, the copolymer unit composition; PynOZO:MeOZO = 5:45, was reacted with an azide compound (R-N3) in aqueous solution via copper-catalyzed Huisgen 1,3-dipolar cycloaddition (click chemistry) to form the 1,2,3-triazole ring [Scheme 8(a)].58 It was suggested that this type of polymer modification reaction may lead to attach peptidic cell recognition sites for targeted drug delivery system. Another example of click chemistry is a “thio-click” modification of 3-butenyl (But) group of PButOZO unit with a mercaptan (RS-H) under UV light at room temperature [Scheme 8(b)].59 Further, to prepare a clickable PROZO, propargyl or 3-butynyl p-toluenesulfonate was used for initiating CROP of MeOZO, EtOZO, NonOZO, and PhOZO monomers under microwave irradiation at 140 °C in acetonitrile. The product PROZOs contain the alkyne end group and the click chemistry reaction of the group with an azide via a copper-catalyzed Huisgen 1,3-cycloaddition to form a 1,2,3-triazole. Reactions of these alkyne-end functionalized PROZOs and various azides will produce new materials for block and graft copolymers as well as other macromolecular architectures.60
Automated Reaction Equipments
With utilizing a commercially available automated synthesizer equipped with individually heatable reactors, the analysis of the temperature dependency of CROP of PhOZO was readily carried out. Thus, the synthesizer allowed to conduct kinetic studies of the CROP easily to result in an optimal reaction temperature and activation energy determination by MeOTs initiator.61 A system with an automated contact-angle measurement enabled to measure more than 100 contact angles per hour, calculating surface energy automatically. Hence, a library of PROZO homopolymers and copolymers obtained by microwave-assisted synthesis was prepared.62 CROP of EtOZO and its kinetics were performed by using a high-throughput workflow system with an automated gel permeation chromatography (GPC) as well as a gas chromatography (GC), which is powerful for direct monitoring of parallel polymerization reactions.63
It has been shown that bismuth compounds are effective catalysts for the CROP.64 Acidic Bi-salts like BiCl3, BiBr3, BiI3, and Bi-triflate catalyzed the CROP of MeOZO and EtOZO in bulk at 80–100 °C. PROZOs obtained had Mn value up to several thousands. Bi-triflate exhibited the highest catalytic activity due to the highest Lewis acidity. It was pointed out that using Bi salts as catalyst is of significance, when polymers are designed for pharmacological or biomedical applications, because Bi3+ is the least toxic heavy metal.
Functional Polymer Architectures and Applications
The first finding of amphiphilic copolymers from ROZOs for new surfactants was reported, using MeOZO and EtOZO as monomers for hydrophilic segments, and 2-propyl (PrOZO), 2-butyl (BuOZO), 2-octyl (OcOZO), 2-dodecyl (DoOZO), and PhOZO as monomers for hydrophobic segments. As a typical procedure, at first a living CROP of MeOZO was induced by MeOTs initiator in acetonitrile at 80 °C, and at the second stage BuOZO was polymerized from the MeOZO living end at 100 °C, to produce P[MeOZO-block-BuOZO] (Scheme 9).65 The amphiphilic (surfactant) nature was evaluated by the surface tension value (γ, dyn/cm) at 29 °C, for example, a block copolymer with R1 = Me (m = 18.8) and R2 = Bu (n = 5.0) exhibits γ = 27.5 (cf, γ of water = 71.3). In addition, surfactants of triblock type were prepared and the copolymer P[OcOZO5.7-block-MeOZO5.8-block-OcOZO5.1] showed the lowest γ value of 23.7 dyn/cm. The critical micelle concentration (CMC) of these copolymers was lower than 1.0 wt %. It is to be noted that PMeOZO segment is so far the most powerful as a nonionic hydrophilic polymer segment due to the polar amide structure; being more powerful than PEG [poly(ethylene glycol)] segment.
New amphiphilic polymer networks were prepared by using MeOZO and NonOZO as monomers, and 2,2′-tetramethylenebis(2-oxazoline) (BisOZO) as a crosslinking agent. The CROP of the mixture of these monomers was initiated with a copolymer of chloromethylstyrene (CMSt) and MMA or with a copolymer of CMSt and St (macroinitiator method) in benzonitrile at 110 °C to give a graft copolymer of the PROZO gels.66 The hydrophilic gels with higher content of MeOZO unit showed a maximal swelling degree of 13 mL of water, 20 mL of methanol, and 13 mL of chloroform per gram of the gel. The hydrophobic (lipophilic) gels only from NonOZO unit showed that of 8 mL of toluene, 14 mL of chloroform, and 2 mL of methanol, respectively. A recent article reported the synthesis of amphiphilic block and random copolymers from EtOZO and NonOZO and the behaviors of product polymers for solubility, dispersions, micellization, and the lower critical solution temperature (LCST).67 Both random and block copolymers with NonOZO unit higher than 60 mol % were solubilized in ethanol upon heating, but the block copolymer needed a higher temperature for the dissolution due to the higher crystallinity. The block copolymer with 10 mol % of NonOZO unit exhibited LCST at 68.7 °C in aqueous solution, whereas that of the random copolymer was only 20.8 °C, due to a hydrophobic microdomain formation in the block copolymer.
Five kinds of PROZO diblock, triblock, and gradient copolymers prepared from MeOZO and NonOZO monomers were examined for the micelle formation in water with using the fluorescence correlation spectroscopy (FCS). All these copolymers (Mn, roughly 4000–6000, PDI <1.4) formed micelles, whose sizes, the hydrodynamic radii r in nm, were, for examples, P[MeOZO40-block-NonOZO7]; 13 ± 2 and P[MeOZO30-block-NonOZO7-block-MeOZO26]; 5.6 ± 0.9, respectively. The radii of the triblock copolymer are lower than those of diblock copolymers, probably due to the space demands of the hydrophilic block as well as the stretching of the core block through the micellar core. The r value of unimers were around 1.2–1.4 nm and the CMC values around 10−5 mol/L.68 A fluoroalkyl-OZO (RfOZO) and NonOZO were used for hydrophobic segments and MeOZO for a hydrophilic segment in the preparation of amphiphilic block copolymers. These polymer amphiphiles formed core/shell micelles in aqueous solutions. Micelles from block copolymer P[MeOZO40-block-RfOZO6] exhibited radii lengths of core rc 3.7 nm and those of shell rm 5.2 nm.69 Micelle formation in solution was widely observed by using various amphiphilic copolymers prepared above, for example, via microwave irradiation.47, 49, 53
Micelle formation is normally observed in a selective solvent by dissolving an amphiphilic copolymer, but the formation is also possible by the deposition on a surface from a nonselective solvent. The micellization on surface was studied by using two quasi-block copolymers, P[PhOZO-block-MeOZO] and P[PhOZO-block-EtOZO]. Micelles were not formed in the initial ethanol solution of the copolymers, but formed during the spin-coating process by evaporation of the solvent. The morphology and size of the surface micelles were investigated in relation to the copolymer structure.70
Amphiphilic block copolymers were used as a polymer catalyst ligand studied by Nuyken et al., which is bound to the metal catalyst site and forms micelles in aqueous solution reaction. One example is a Ru-containing Hoveyda-Grubbs catalyst [Scheme 10(a)]71 and another is a Rh-containing catalyst, the polymer ligand of which is given in Scheme 10(c).72 Both copolymers utilize PMeOZO for hydrophilic segments and PNonOZO for hydrophobic segments. Ru-catalyst (a) was used for the ring-closing metathesis reaction of diethyl diallylmalonate to produce diethyl 3-cyclopentene-1,1′-dicarboxylate plus ethylene [Scheme 10(b)], in which the reaction proceeded in micelles effectively, showing a turn-over number reaching to 390 in water, the highest value ever reported. The polymer-bound catalyst (a) could be recycled. A Rh-catalyst complexed with polymer-ligand containing amino-group (c)73 was used as micellar catalyst for the 1-octene reaction of hydroformylation and hydroxyaminomethylation [Scheme 10(d)]. Best results were obtained by applying a dual Rh/Ir catalyst within the polymeric micelles, with respect to reaction temperature, product yield, and reaction selectivity.
Poly(2-isopropyl-2-OZO) (PiPrOZO) was first found to be a good thermosensitive polymer, similar to poly(N-isopropylacrylamide) (PNIPAM) (Scheme 11).74 PiPrOZO (Mn = 16,700, Mw/Mn = 1.13) showed a clouding point (CP) of 36.0 °C in water with 0.5 wt % of the polymer, hereby the polymer is soluble below CP. The hydrogel of PiPrOZO was prepared, whose swelling transition was around the CP near the temperature of human beings physiological conditions. This is expected for the application to thermosensitively functional biomaterials. For applying these results to biomedical materials, end-functionalized PiPrOZOs were synthesized, giving the polymers with structures of Me-PiPrOZO-OH, Me-PiPrOZO-NH2, and acetal-PiPrOZO-OH,39 and the temperature-induced phase-separation (similar to CP) was studied in detail with using microcalorimetry by Kataoka et al.75
Living copolymerization of iPrOZO as hydrophobic monomer with EtOZO as hydrophilic monomer was performed to produce statistical copolymer P[iPrOZO-co-EtOZO]. Monomer reactivity ratios, riPrOZO = 0.79 and rEtOZO = 1.78, were sufficiently different to lead to gradient copolymers (Mn = 8000–10,000) with varying the compositions and narrow PDI (≤1.02). Turbidity measurements revealed that LCST (corresponding to CP) of the copolymers could be precisely tuned over a broad range of 38.7–67.3 °C by varying the molar ratio of iPrOZO to EtOZO.76 A further study used three monomers, PrOZO, iPrOZO, and EtOZO, for developing more precise thermosensitive PROZOs.77 Copolymerization of PrOZO and iPrOZO (rPrOZO = 3.15 and riPrOZO = 0.57, respectively) formed gradient copolymers P[PrOZO-gradient-iPrOZO], whereas that of PrOZO and EtOZO (rPrOZO = 1.28 and rEtOZO = 1.04, respectively) gave random copolymers P[PrOZO-random-EtOZO). The LCST of both the gradient and random copolymers could be tuned over a wide range of temperatures by varying the molar ratio of the two monomers. Most of the random copolymers (Mn around 10,000–14,000, Mw/Mn < 1.1) showed a rapid and modulated response to the temperature change from 23.8 to 75.1 °C, being considered as an ideal system for tuning the LCST around the physiological conditions. It is notable that the homopolymer PPrOZO showed a sharp LCST-type phase transition around 23.8 °C.
Block and graft copolymers constituted from PNIPAM chain and PiPrOZO as well as PROZO chains were prepared and their thermosensitive behaviors were investigated.78, 79 For the graft polymers of Scheme 11, the turbidity temperature (same as CP) was tuned by the hydrophilic ROZO chain content from 28 to 42 °C and the increased content of ROZO chain made the CP value higher. A new P[iPrOZO-block-3-butenylOZO] was synthesized and a glucose moiety was introduced to the olefinic side chain via “click” addition of a glucose-thiol. Then, the CP of these copolymers was tuned from 10 to 85 °C over a wide range as a function of composition and hydrophilicity of the side chain.80
Molecular brushes are linear macromolecules with pendant polymer side chains at high grafting density. The side chain crowding induces a strong stretching of backbone and side chains to adopt a cylindrical molecular brush structure. Cylindrical molecular brushes of PROZOs from 2-isopropenyl-2-oxazoline (IPOZO) have been synthesized and characterized by Jordan et al.81 First, radical or anionic vinyl polymerization of IPOZO was conducted to form a backbone polymer (n = 88 with radical process and 220 with anionic process) (Scheme 12). Then, pendant 2-oxazoline groups were methylated by MeOTf to form oxazolinium ions, which serve as macroinitiators, and CROP of ROZO monomer was initiated from the macroinitiators (grafting-from approach), giving rise to cylindrical molecular brushes P[IPOZO-graft-ROZO] (oxazoline units, m = 18–23). As the substituent, Me, Et, and iPr were used. The LCST values determined by turbidity measurements of a 1.0 wt % aqueous solution were 63 °C for the brush (n = 88, EtOZO, m = 23), 30 °C for the brush (n = 88, iPrOZO, m = 18), and 27 °C for the brush (n = 220, iPrOZO, m = 18), respectively.
The method of the molecular brush synthesis was extended to synthesize brushes of bottle-brushes of PROZOs on polished glassy carbon (GC) substrates (Scheme 12).82 First, UV-induced vinyl polymerization of IPOZO monomer at room temperature was directly initiated from the GC surface (a) and gave PIPOZO grafted brush layers with thickness up to 160 nm (b). The pendant oxazoline ring of the PIPOZO was used followed by the oxazolinium ion formation (c) to perform a second CROP of ROZO monomers to result in bottle-brush brushes (d). It was possible to functionalize the bottle-brush brushes side chain end groups with a steric-demanding molecule such as rhodamine B, which was verified by fluorescence measurement of the dye molecule (e).
A one-pot, one-step preparation of binary mixed homopolymer brushes was reported, which combined simultaneous nitroxide-mediated radical polymerization of styrene and living CROP of PhOZO. The binary mixed homopolymer brush exhibited reversible self-adapting surface property changes when subjected to different solvents.83
Functionalized Monomers and Polymers
Macromonomers are known as providing a starting monomer to synthesize various graft polymers, when the macromonomer has a vinyl polymerizable group. Then, macromonomers having an oxazoline-polymerizable group were prepared via an initiator method, that is, by initiating anionic ROP of ethylene oxide and anionic vinyl polymerization of t-butyl methacrylate or MMA with MeOZO carbanion, and also by initiating anionic ROP of ε-caprolactone with MeOZO-containing alkoxide.84 Molecular weight (Mn) of these macromonomers was varied from 700 to 19,500 with high functionality.
Comb and graft polymers were synthesized based on macromonomer method and their LCST behavior was examined, whereby a hydrophilic oligoEtOZO was used as side chains and a hydrophobic methacrylate as backbone (Scheme 13).85 First, CROP of EtOZO was carried out under microwave irradiation using MeOTs initiator and the living end was terminated with ammonium methacrylate to form macromonomers (Mn = 900–3300, Mw/Mn < 1.1). Then, homopolymerization of the macromonomer was conducted radically via the reversible addition-fragmentation chain transfer (RAFT) polymerization method, giving rise to comb polymers (Mn = 7000–24,000, Mw/Mn ≤ 1.31). Copolymerization of the macromonomer with methyl methacrylate (MMA) was performed similarly, to produce graft copolymers (MMA, 40–80 mol %, Mn = 12,000–26,000, Mw/Mn < 1.3). The LCST of the graft polymers measured in aqueous solutions was tuned from 35 to 80 °C depending on the MMA content. LCST value of the comb polymer was significantly decreased when compared with that of linear PEtOZO, which was correlated with microdomain formation due to the crowded comb structure.67
Pluritriflate initiators, derived from mono-, di-, tri-, and tetra-functional alcohols, were used for CROP of MeOZO at 85 °C in acetonitrile for preparing hydrophilic star polymers. It was shown that the propagation of the each arm proceeded in a similar rate in the reaction of these initiators (kp value = ∼5 × 10−3 L/mol/s).86
Reactive functional groups were introduced into PROZO side chains for further utilization of PROZO chains. A 2-oxazoline monomer bearing a protected SH group was copolymerized with EtOZO, and the deprotection from the copolymer gave SH-containing PEtOZO (Mn around 2000–5000), which added to various activated olefinic compounds like a maleimide and amide compounds.87 An amino group-containing OZO monomer was prepared using Boc-protected amino function and the monomer was homopolymerized and copolymerized to lead to PROZOs with pendant amino-groups (Mn around 4000–8000, PDI < 1.4).88
Interestingly, glycosylated PROZO chains self-assembled into nanotubes in water (!) through intermolecular hydrogen bonding interactions (Scheme 14).89 The driving force for the nanotube formation is considered due to the unique polymer structure of tertiary polyamide backbone (hydrogen-accepting) and glucose side chains (hydrogen-donating), first forming and bending of a 2D hydrogen-bonded layer of interdigitated polymer chains and then closing to the tube. Rough values of outer and inner radii (ro and ri) and wall thickness (ro−ri) from three polymers are as follows (given in nm): 1; 2.0, 0.8, and 1.2. 2; 3.0, 2.0, and 1.0. 3; 4.5, 1.0, and 3.5.
Typical applications of ROZO polymers are biological and biomedical uses,90 because they are viewed as “pseudopeptides” and of nontoxic nature as mentioned in General Aspects. It was found that ammonium group-containing PEtOZOs and PMeOZOs showed efficient antimicrobial properties (Scheme 15).91 The antimicrobial potential of the polymers was tested by determining the minimal inhibitory concentration (MIC) of the polymer against the ubiquitous Gram-positive bacterium Staphylococcus aureus. MIC is the standard quantification for antimicrobial substances and is defined as the minimal required concentration to inhibit the growth of 99% of the bacterial cells in a suspension. For example, MIC value of the MeOZO polymer chain (R = Me, n = 53 in Scheme 15) was 1.0 mg/mL, which is nine times more effective than the corresponding PEG polymer chain. Also, the effect of end group structures on the antimicrobial activity was investigated.92
Linear polyethylenimine (PEI) is derived by the hydrolysis of PROZO chains. Novel triblock copolymers, PEI-PEG-PEI (2100-3400-2100 and 4000-3400-4000) [Scheme 16(a)] were shown to condense plasmid DNA effectively to give polymer/DNA complexes (polyplexes) of small sizes (<100 nm) and moderate ξ-potentials (∼+10 mV) at polymer/plasmid weight ratios ≥ 1.5/1. These polyplexes efficiently transfected COS-7 cells and primary bovine endothelial cells in vitro and are a novel class of nonviral gene delivery systems.93 The acetal-PEG-PEI block copolymer is considered to have a potential utility as targetable DNA carrier in the field of gene delivery.94 Lipopolymers were prepared as a potential candidate for constructing tailored model cell membranes, in which a lipid trifrate was used as initiator for CROP of hydrophilic monomers, MeOZO and EtOZO, to produce an amphiphilic polymers [Scheme 16(b)].95
Glycoproteins play important functions in mechanism like bioadhesion, cell–cell interactions, and recognition phenomena. A glycoprotein analog was newly prepared by combining a linear polysaccharide block and a PROZO block (psuedopeptide), for which hyaluronic acid (hyaluronan, HA) and PEtOZO were used, respectively. A low molecular weight HA (Mw = 2700) was obtained by hyaluronidase-catalyzed degradation of commercial HA, and then reacted with an amino group-containing PEtOZO to produce HA-block-PEtOZO (Mw = 10,200) (Scheme 17).96 Preliminary experiments showed that the anionic copolymer forms colloidally stable particles (Rh ∼ 130 nm) with the cationic drug diminazene.
In addition, the aforementioned various amphiphilic polymers were often discussed to involve possible applications for biomedical use as thermosensitive bioconjugates, drug delivery systems,76, 77 and for life sciences (biomedicine or cell sensing).80
A CROSSING WHERE CATIONIC AND ENZYMATIC PROCESSES OF 2-OXAZOLINE CHEMISTRY MEET
2-Oxazolines of sugar molecules have been known as a glycosylation donor in carbohydrate chemistry, and the use of the oxazolines for the polysaccharide synthesis was not efficient, resulting only in an oligomer formation without good control of stereochemistry.97 Then, the first example of a substituted polysaccharide synthesis was reported in 1996, in which ring-opening polyaddition (ROPA) of 3,6-di-O-benzylated chitin-oxazoline monomer was induced by 10-camphorsulfonic acid (CSA) catalyst to produce di-O-benzylated chitin [Scheme 18(a)].98 The ROPA was carried out at a reflux temperature in 1,2-dichloroethane and proceeded involving stereoregular glycosylation, giving rise to a dibenzylchitin product having β(1→4)-glucopyranan structure with Mn up to 4900 in around 50% yields. The acid-catalyzed ROPA was extended to the synthesis of the dibenzylchitin having β(1→6)-glucopyranan structure [Scheme 18(b)].99 The product is of unnatural-type structure, whose yield was at most 32% and Mn was up to 13,100. For both ROPA reactions, CSA was a very effective catalyst among other protic acids. The de-benzylation of the products via the catalytic hydrogenation was carried out; the reaction of product (a) was incomplete, whereas that of product (b) took place perfectly to produce the corresponding free aminopolysaccharide of unnatural type. The principle of these reactions was applied for the preparation of a hyperbranched polymer from an AB2 type monomer, in which 3- and 4-positions have free OH-groups [Scheme 18(c)].100 The product polymer possessed Mn up to 6600 (PDI < 2), whose de-tosylation gave a free hyperbranched aminopolysaccharide.
Furthermore, a sugar-oxazoline derived from N-acetyl-D-glucosamine was reacted with MeOTf to give a methyl oxazolinium triflate, which was used as an initiator for CROP of MeOZO and PhOZO to produce a sugar moiety-containing PROZO.101
These reactions can be taken as a platform to link the CROP of 2-oxazoline monomers and the EROPA of sugar oxazoline monomers (vide infra) from the viewpoint of reaction mode.
ENZYMATIC RING-OPENING POLYADDITION OF SUGAR OXAZOLINE MONOMERS
General Aspects: Role of 2-Oxazoline Structure
One of the most important and difficult reactions in sugar chemistry is glycosylation, connecting two sugar units with regioselective and stereocontrolled manner. Polysaccharide synthesis is only achieved when the repetition of highly selective glycosylation reactions is achieved multi-times. Normally, the glycosylation is carried out between a glycosyl donor and a glycosyl acceptor, in which the positions not to be reacted must be protected; the simple example to form a β(1→4) glycosidic linkage via condensation is given in Scheme 19. The anomeric C-1 carbon of the donor is to be activated by introducing X. One of possible ways of activation is an introduction of 2-oxazoline structure in Scheme 20, where a donor oxazoline is reacted with an acceptor of R′OH via ring-opening addition to form a β-glycosidic linkage. As mentioned in “Introduction,” it was found for the first time that a sugar oxazoline was recognized and activated by chitinase enzyme to lead to “synthetic chitin” in one-step (Scheme 2).34 The reaction of Scheme 2 is an enzymatic ring-opening polyaddition (EROPA) of a chitobiose 2-oxazoline monomer (without protection), in which the monomer acted as a donor as well as an acceptor. This EROPA method has been found very efficient for the synthesis of poly- and oligo-saccharides containing 2-amino-2-deoxy sugar units such as N-acetyl-D-glucosamine (GlcNAc), which are widely found in the living organisms and play crucial role in various biological functions. Since then, a variety of natural and unnatural amino-polysaccharides (mucopolysaccharides) have been synthesized on the basis of EROPA of sugar 2-oxazoline monomers.26, 28, 30, 31
Chitin is a natural homo-polysaccharide that consists of GlcNAc units connecting through β(1→4) glycosidic linkage. It is an amino-polysaccharide. It is one of the most abundant organic substances in the animal world, in contrast to that of cellulose on the Earth. A number of studies have been performed to synthesize chitin derivatives in vitro for the detailed investigation of their biological functions toward new functional materials architecture.102, 103 For example, chitin derivatives were synthesized via polycondensation of ethyl 6-O-acetyl-3-O-benzyl-2-deoxy-2-phthalimido-1-thio-β-D-glcopyranoside in the presence of methyl triflate catalyst. All glycosidic linkages were confirmed to be β by 1H NMR, and the total yield of oligomers up to dodecamer was 70% (Scheme 21).104 It was also shown that chitin derivatives were obtained by an acid-catalyzed ring-opening polyaddition of an oxazoline monomer [Scheme 18(a)].99 The key point for these polymerizations is the use of monomers having a bulky N-phthalimido group as C-2 protecting group or the 2-oxazoline ring so as to control stereochemistry of the anomer carbon. However, in both cases, there was a problem of deprotection process of the product polymers to the final free form; the deprotection was very hard and could not be completed.
As an extension of the first in vitro synthesis of cellulose via enzymatic polymerization,105, 106 we challenged and succeeded in the in vitro chitin synthesis in 1995, which utilized chitinase enzyme as catalyst and a chitobiose oxazoline as monomer, giving rise to “synthetic chitin” in quantitative yields as given in Scheme 2.32–34 The structure of the product chitin is perfectly controlled, whose DP value is 10–20 depending on the polymerization conditions.30 The oxazoline monomer was polymerized via enzymatic ring-opening polyaddition (EROPA) mode; the polymerization was induced through recognition of the monomer by the enzyme. Thus, the concept of TSAS monomer was proposed, which implies that the oxazoline monomer structure resembles to that of the transition-state of the hydrolysis reaction.34 On the other hand, a study to obtain mechanistic information on chitinase-catalyzed hydrolysis of chitin was performed by X-ray structural analysis; the article proposed a substrate-assisted mechanism involving an oxazoline structure as the intermediate for the hydrolysis and described that the fact of chitinase-catalyzed chitin synthesis from the oxazoline monomer also supports an oxazoline intermediate (or transition-state) of the chitinase-catalyzed hydrolysis.107, 108 In addition, lysozyme and chitinase (family 19) are known to involve an oxocarbenium intermediate, and their catalytic activity on EROPA of the oxazoline monomer was not observed.109 These results clearly indicate that taking an oxazolinium intermediate in the hydrolysis (via substrate-assisted mechanism) is essential for the catalysis of the EROPA.
Chitinase active for the EROPA is an exo-type enzyme, which is classified into a glycoside hydrolase (GH) family 18, having a cleft-like catalytic domain. At the catalytic domain, two conserved carboxylic residues are involved and considered to act as an acid–base catalyst. Further, allosamidine is a well-known inhibiting agent for chitinase.110, 111 On the basis of these observations, reaction mechanism of chitinase-catalysis is considered as illustrated in Figure 1.
In the case of hydrolysis, glycosidic oxygen atom of chitin chains in the catalytic domain is protonated by one carboxylic residue, and then carbonyl oxygen atom is nucleophilically attacked by the anomeric carbon (stage a). At the same time of C-1-O bond cleavage with inversion of C-1 anomeric stereochemistry, an oxazoline ring with α-C-1 structure is formed as an intermediate (or transition-state), which is stabilized by another carboxylic residue to form an oxazolinium ion (stage b). To the anomeric C-1 carbon of the ion, water molecule nucleophilically attacks from β-side, and the oxazolinium ring is opened with inversion of C-1 carbon, giving rise to the hydrolysate in a retaining manner due to two-times inversions (stage c).
On the other hand, in the polymerization the monomer has already an oxazoline with α-C-1 structure and then can be readily recognized at the donor side of chitinase (stage a′) to form an oxazolinium ion (stage b′), which is similar in structure to that of the intermediate (or transition-state) of stage b in the hydrolysis. The oxazolinium ion is attacked by 4-OH group of nonreducing end GlcNAc from β-side to form β(1→4) glycosidic linkage due to one-time inversion (stage c′). Repetitions of these steps of stereo- and region-selective glycosylation produce “synthetic chitin” (Scheme 2). These observations strongly support the concept of the transition-state analogue substrate (TSAS) monomer. An oxazolinium ion structure is commonly involved at both stages (b) and (b′). Thus, the monomer acted as a glycosyl donor as well as a glycosyl acceptor. The reaction mechanism of EROPA is an “activated monomer mechanism,” which is to be compared with a “activated chain-end mechanism” observed normally in vinyl polymerizations.
Further, optimal hydrolysis reaction condition by the chitinase was reported to be pH 8.0.112 However, the EROPA (reverse reaction to the hydrolysis) was still induced even under the weak alkaline conditions around pH 10–11, and therefore, hydrolysis of the product synthetic chitin was much suppressed.
In addition, the above depth-considerations for the chitinase-catalyzed polymerization involving an oxazolinium transition-state are also supported by the results that the chitinase-catalyzed hydrolysis occurs via oxazolinium transition-state (or intermediate) evidenced by several research groups in the late 1990s.107, 113
Chitin Derivatives and Unnatural Hybrid Polysaccharides
Substrate specificity of enzymatic reactions is normally considered to be extremely high. However, glycoside hydrolases including chitinase are somehow more dynamic than considered and can catalyze a specific reaction of an unnatural substrate with certain qualifications. C-2 Oxazoline structure at reducing end as a glycosyl donor and C-4′ hydroxy group at nonreducing end as an glycosyl acceptor are essential for the exclusive formation of the β(1→4) glycosidic linkage. Accordingly, 12 chitobiose oxazoline derivatives modified at C-3, C-6, C-2′, C-3′, and C-6′ were newly designed and synthesized as TSAS monomers (Scheme 22) and subjected to the chitinase-catalyzed polymerization (Table 1). In fact, chitinase catalyzed the formation of β(1→4) glycosidic linkage from the modified chitobiose oxazoline monomers as the followings.
Table 1. Chitinase-Catalyzed Polymerization of Various Chitobiose 2-Oxazoline Derivatives as TSAS Monomers
Chitobiose oxazoline derivatives were prepared by introducing methyl group at C-3 position of either or both sugar units (1-3, Scheme 22). These monomers showed less polymerizability, and oligomers up to octa- and hexa-saccharides were obtained from monomers 1 and 2, respectively.114 Polymerizability of monomers modified at C-6 depended on the bulkiness of the functional groups. For example, C-6 carboxylmethylated monomers, 4 and 5, having a bulky functional group were hardly polymerized, and only tetrasaccharide was obtained from monomer 5.115, 116 However, monomers substituted with fluorine atom (6–8), whose covalent radius (0.64 Å) is almost the same with that of oxygen (0.66 Å), showed good polymerizability.117 Despite the less polymerizability of monomers 1–5, conversion of the monomers was accelerated by the addition of chitinase to the reaction mixture, indicating that all these monomers were recognized at the donor site of the enzyme and hydrolyzed. Because of the steric hindrance and/or distortion of the monomer's three-dimensional structure, spatial location between the oxazoline anomeric carbon in donor molecule and the C-4 hydroxy group in acceptor molecule probably becomes less favored in the catalytic domain for the β(1→4) glycosidic linkage formation.
On the other hand, chitobiose oxazoline derivatives modified at C-2′ showed good polymerizability. N-Acetyl-chitobiose oxazoline monomer 9, which has an amino group at C-2′ position, was polymerized by chitinase catalysis, giving rise to a water-soluble polysaccharide with Mw 4280 in a 75% yield.118 The resulting polysaccharide has alternating structure of β(1→4)-linked D-glucosamine (GlcN) and N-acetyl-D-glucosamine (GlcNAc). β(1→4)-Linked GlcN polymer, which is normally prepared via deacetylation of chitin, is chitosan. Therefore, the polymer was named as a chitin–chitosan hybrid polysaccharide [Scheme 23(a)]. Further, monomer 10 having a bulky and ionic N-sulfonate group was also catalyzed by chitinase and gave corresponding chitin derivative of Mw 4990 in 62% yield.119 These results indicated chemical structure at C-2′ of chitobiose oxazoline is less influential for chitinase recognition.
In all cases, regio- and stereo-chemistry of the resulting polysaccharides is perfectly controlled because of the high catalysis specificity of the enzyme, and their chemical structure is consisted as that of the disaccharide repeating unit with corresponding substituted monomers connected through exclusive β(1→4) glycosidic linkages.
On the basis of the previous results, monomers 11 and 12 were newly designed and synthesized. Homopolymer of β(1→4) linked glucose (Glc) is cellulose, and that of xylose (Xyl) is xylan, which is one of major components of hemicellulose. Structural difference between Glc and GlcNAc is only C-2 substituent of a pyranoside unit: a hydroxy group on the cellulose molecule and an acetamido group on the chitin molecule. Xyl has C-6 methylol deleted form of Glc. Both monomers with chitinase catalysis gave polysaccharides having β(1→4) linked Glcβ(1→4)GlcNAc and Xylβ(1→4)GlcNAc structure, respectively [Scheme 23(b,c)].120, 121 The resulting polysaccharides are unnatural hybrid type polysaccharides of original two polysaccharide components, and named as a chitin-cellulose hybrid polysaccharide (b) and a chitin-xylan hybrid polysaccharide (c), respectively. These hybrid-type polysaccharides are expected to exhibit biological functions resulting from both natures.122
The catalytic domain of family 18 chitinase consists of a (β/α)8 barrel with a deep substrate-binding cleft. The point mutated chitinase from serratia marcescense was subjected to form a complex with a β(1→4)GlcNAc hexamer molecule, and the crystallographic studies of the complex showed that the hexamer occupies subsites −4 to + 2 of the cleft as shown in (a) and (b) (Fig. 2).123 At the catalytic domain, the aromatic residues that interact with the substrate are highlighted. Trp539, Trp167, and Tyr170 (green), which are located at the bottom of the cleft, are positioned to interact with the hydrophobic faces of the GlcNAc C-2 acetamido groups at sites -1, -3, and -5, respectively. However, C-2 acetamido groups located at subsites −2 and +1 face to the opposite sides of the cleft, and there seems to be a spatial margin. On the other hand, the location of C-3 and C-6 hydroxy groups looks relatively busy. These observations consistently explain the results of EROPA of monomers 1–12 from the viewpoint of substituted group effects.
Endo-A Catalyzed Polymerization
Endo-A from Arthrobacter protophormiae is an enzyme, which hydrolyzes the GlcNAcβ(1→4)GlcNAc glycosidic linkage located at the connecting region of N-linked glycoproteins. It is classified into GH family 85. Hydrolysis reaction mechanism of the enzyme is considered to proceed via a substrate-assisted mechanism involving an oxazoline intermediate in the same way with that of a family 18 chitinase. The enzyme has well-known activity for the transglycosylation of a Manβ(1→4)GlcNAc oxazoline derivative to a GlcNAc-Asn acceptor.124, 125
Recently, it has been found that under the reaction conditions with 25 times larger amounts of Endo-A than the previously reported transglycosylation reaction, the enzyme catalyzed the polymerization of Glcβ(1→4)GlcNAc oxazoline monomer 11 and gave the corresponding chitin-cellulose hybrid polysaccharide in 79% yields.126
Hyaluronic Acid and Chondroitin
Glycosaminoglycans (GAGs) are biomacromolecular hetero-polysaccharides, which are normally linked to various proteins to form proteoglycans. Together with collagens, fibronectins, and others, proteoglycans fill the interstitial space between living cells, and they are called extracellular matrices (ECMs). GAGs include hyaluronic acid (hyaluronan, HA), chondroitin sulfate (ChS), dermatan sulfate, heparin/heparan sulfate, and keratin sulfate (KS). All GAGs contain hexosamine unit such as N-acetyl-D-glucosamine (GlcNAc), N-acetyl-D-galactoamine (GalNAc), and their sulfated derivatives. HA is a linear polysaccharide having a repeating unit of β(1→3)-GlcNAc-β(1→4)-GlcA (= glucronic acid). Chondroitin (Ch) is non-sulfated ChS whose repeating unit structure is β(1→3)-GalNAc-β(1→4)-GlcA.30, 31
Hyaluronidase (HAase) belongs to GH Family 56 hyaluronidase which is known to hydrolyze β(1→4)-linked GlcNAc(or GalNAc) and GlcA in HA and Ch chains. On the basis of the consideration for the TSAS concept, N-acetylhyalobiuronate (GlcAβ(1→3)GlcNAc) oxazoline derivative and an N-acetylchondrosine oxazoline derivative are designed as TSAS monomers for the hyaluronidase-catalyzed polymerization.127, 128 Both monomers were effectively catalyzed by ovine and bovine testicular derived hyaluronidases (OTHase and BTHase, respectively), giving rise to the corresponding polysaccharides of HA and Ch (Scheme 24). The weight-averaged molecular weight (Mw) of the synthetic HA and synthetic Ch reached to 25,000 and 6800, respectively. The molecular weight value of Ch corresponds to that of natural Ch. These polymerizations are the first examples for the synthesis of hetero-polysaccharides via in vitro polymerization.
It is also known that at the catalytic domain of the HAase there is a conserved DXE(D) motif stabilizing an oxazolinium ion intermediate (or transition-state),129 and hence, the hydrolysis reaction mechanism is considered to be similar to that of family 18 chitinaseas as illustrated in Figure 1. These observations are accord with the concept of TSAS monomer, strongly suggesting that two EROPA reactions of Scheme 24 occur via an oxazolinium transition-state.
Derivatives of Hyaluronic Acid and Chondroitin
The HAase-catalyzed polymerization of modified monomers was carried out. 2-Ethyl, 2-n-propyl and 2-vinyl oxazoline monomers were newly prepared and polymerized with HAase catalysis. The reactions proceeded with total control of regioselectivity and stereochemistry, to afford the corresponding HA and Ch derivatives (unnatural polysaccharides) possessing N-propionyl, N-butyryl, and N-acryloyl group in every hexosamine unit (Scheme 25).128, 130 The resulting N-acryloyl HA and Ch are functional polymers having a reactive vinyl group.
Some of these monomers showed a close reactivity against HAase. Copolymerization of 2-methyl, 2-vinyl, 2-ethyl, and 2-n-propyl oxazoline monomers proceeded in a regio- and stereo-selective manner, to produce the corresponding copolymers. By varying the comonomer feed ratio, composition of the N-acyl groups could be easily controlled.131
HAase-catalyzed copolymerization of N-acetylhyalobiuronate oxazoline and N-acetylchondrosine oxazoline monomers also gave an unnatural intramolecularly hybridized HA-Ch copolymer, which can be hardly prepared via conventional methods (Scheme 26).132
Chondroitin (Ch) is widely distributed in the ECMs and on cell surfaces. Ch is normally sulfated and biological activities of chondroitin sulfate (ChS) depend on the sulfated pattern. According to the major sulfation pattern, natural ChS is classified into the following five types; ChS-A, ChS-C, ChS-D, ChS-E, and Chs-K. None of them is composed of a single repeating unit; all are of mixed type. For example, in a whale cartilage the so-called ChS-A contains ∼20% of ChS-C as a minor component besides the major ∼80% of ChS-A. Therefore, to prepare chondroitin sulfate (ChS) having a well-defined sulfation pattern is important for the fundamental investigation of the correlation between polysaccharide structure and the bioactivities. Then, N-acetylchondrosine oxazoline monomers having 4-, 6-, and 4,6-disulfate groups at GlcNAc unit were synthesized and their hyaluronidase-catalyzed polymerization was examined (Scheme 27).133 The monomer having sulfate group at C-4 (monomer a) was successfully polymerized by HAase, giving rise to ChS in high yields having sulfate groups only at C-4 position on all GlcNAc residues, which is a “pure” ChS-A with Mw values ranging from 5600 to 36,500 depending on the reaction conditions. Biological activity examination of the synthetic pure ChS-A will be a future study of interest. In case of monomers having sulfate group at C-6, ring-opening reaction was accelerated, but no polymeric products were obtained.
Reaction Mechanism of Hyaluronidase Catalysis
Hydrolysis reaction mechanism of testicular derived family 56 HAase has been considered to proceed via a substrate-assisted mechanism involving an oxazoline intermediate with homology to family 18 chitinase.107, 113 The above EROPA results that HA, ChS, and their derivatives can be synthesized from TSAS oxazoline monomers by the HAase catalysis also supported that the involvement of 2-oxazoline structure as an intermediate in the hydrolysis mechanism is a key-point. So far, X-ray structural analysis studies of HAase have been performed for human Hyal-1134, 135 and bee venom hyaluronidase (bvHyal);136 similar studies of BTHase and OTHase, that are active for the EROPA, have not been available, however. The sequence identity between BTHase and bvHyal is 32%, which is difficult to lead to a directly comparative model of the active site for BTHase. Nevertheless, construction of a comparative BTHase model using bvHyal crystalline structure was performed by theoretical calculation approach.137 Figure 3 shows a proposed binding mode using an HA fragment (tetrasaccharide) fitting into the binding site of the BTHase model. Asp147 and Glu149 indicate two catalytic acidic amino acid residues. From the model observation, C-4 GlcNAc hydroxy group at subsite -1 seems to be located just to the cleft opening. This spatial margin at the catalytic domain of BTHase probably allows to recognize N-acetylchondrosine oxazoline monomer having an axial C-4 hydroxy group and to induce its EROPA.
HAase showed high polymerizability for various kinds of TSAS monomers as discussed earlier. Therefore, HAase was called as a “supercatalyst” for the enzymatic polymerization (Scheme 28).138
Keratan sulfate (KS) is one of the classes of GAGs having a β(1→3) linked Gal-β(1→4)-GlcNAc (LacNAc) repeating structure, and it is sulfated at C-6 of either or both the Gal or GlcNAc moiety. Keratanase II is a glycoside hydrolase of KS, which catalyzes to cleave a GlcNAcβ(1→3)Gal linkage. Similar to the HAase mechanism, we postulated that during the hydrolysis the CO carbonyl oxygen atom in GlcNAc residue acts as catalytic nucleophile (base) and the hydrolysis proceeds via oxazolinium transition-state. Therefore, LacNAc oxazoline monomers having sulfate group at C-6 of GlcNAc and at both C-6 sugar units were designed as TSAS monomers for keratanase II, and their behaviors toward the catalyst were examined. Then, KS oligomers with perfectly controlled sulfation pattern were obtained (Scheme 29).139 Hydrolysis mechanism of keratanase II has not been well known, yet. These polymerization results, however, strongly suggests that the hydrolysis proceeds via a substrate-assisted mechanism involving an oxazolinium intermediate.
This article highlighted the recent developments of the polymer synthesis involving chemistry of 2-oxazolines. CROP of 2-oxazoline monomers has been known for more than 4 decades and it is still actively studied. EROPA of sugar 2-oxazoline monomers was found in 15 years ago and a new concept of TSAS monomer was created. CROP and EROPA look like involving a quite different class of reactions, yet interestingly they merged into the polymer synthesis field; they produce polymers with much different structures, poly(N-acylethylenimine)s and amino-polysaccharides, respectively. The key for the reactions to occur is an appropriate combination of a catalyst and a designed monomer. From the viewpoint of reaction mode, an acid-catalyzed ROPA is regarded as a crossing for the two classes of reaction, CROP and EROPA, to meet. Finally, it is to be noted that the EROPA reactions enabled for the first time to synthesize the above several hetero-polysaccharides, which have one of the most complicated structures of polymers ever synthesized in test-tube reactions.
Akira Makino studied Material Chemistry at Kyoto University and received his B.Sc. and M.Sc. degrees from Kyoto University in 2001 and 2003, respectively. In 2006, he obtained his Ph.D. in engineering from Kyoto University under the supervision of Prof. Shiro Kobayashi and Prof. Shunsaku Kimura. In 2006, he worked as a postdoctoral researcher at Shimadzu Corp., Japan. He has been an assistant professor at the Department of Material Chemistry, Kyoto University since 2007.
Shiro Kobayashi was graduated from Kyoto University and received his Ph.D. in 1969 there. After staying at Case Western Reserve University as a postdoc for 2 years, he joined Kyoto University in 1972, and started to investigate polymer synthesis. He then stayed at Mainz University as a Humboldt fellow in 1976. Following a lectureship in Kyoto University, he became a full professor of Tohoku University in 1986 and started to work on enzymatic polymerization. He moved back to Kyoto University in 1997 and officially retired in 2005 to become an emeritus professor. Since then he has been a distinguished professor at Kyoto Institute of Technology. His research interests include polymer synthesis, organic reactions, material science and, in particular, enzymatic polymerization, which enabled the first chemical synthesis of various natural and unnatural polysaccharides, functionalized polyesters and phenolic polymers. He received The Chemical Society of Japan Award for Young Chemists (1976), The Society of Polymer Science Japan Award (1986), The Distinguished Invention Award (1993), The Cellulose Society of Japan Award (1996), The Humboldt Research Award (1999), The Chemical Society of Japan Award (2001), The John Stauffer Distinguished Lecture Award (2002), The Society of Polymer Science Japan Award for Outstanding Achievement in Polymer Science and Technology (2004), and The Medal with Purple Ribbon (2007), and others. He is a foreign member of the Northrhine Westfalian Academy of Science since 1999. He currently serves as a member of (executive) advisory board and editorial (advisory) board for 14 international journals.