Functional groups containing a nitrogen-carbon double bond (C = N), where nitrogen forms a bond with aryl or alkyl groups, are known as Schiff bases. The Schiff base reaction (Scheme 1), first described by Schiff (1864), can proceed by a condensation reaction of an aldehyde or ketone with a primary amine (Schiff, 1864; da Silva et al., 2011; Ciaccia and Di Stefano, 2015). In order to favor the reaction, the solution must be sufficiently acid, resulting in protonation of the carbonyl compound and making the carbon more susceptible to nucleophilic attack (da Silva et al., 2011; Ciaccia and Di Stefano, 2015).

Substituent groups can be inserted into chitosan backbones by reductive amination (Scheme 2), which is a variation of Schiff base formation (Baxter and Reitz, 2002; Goszczynska et al., 2015). However, in this case, chitosan reacts with an aldehyde or ketone, resulting in an imine intermediate that is subsequently converted to an N-alkyl or N-aryl derivative by reduction employing sodium borohydride (NaBH₄) or sodium cyanoborohydride (NaCNBH₃) (Baxter and Reitz, 2002). The amines generated by these reactions are stable toward hydrolysis, and other advantages of this route are selective functionalization of the amino groups of chitosan and a homogeneous distribution of substituent groups in the chitosan backbone (Sajomsang, 2010; Badawy and Rabea, 2013).

Du and Hsieh (2007) performed PEGylation of chitosan by reductive amination and acylation, resulting in PEG-N-chitosan and PEG-N,O-chitosan, respectively. Chitosans with different molecular weights (137, 190, and 400 kDa) were used, together with different molecular weight PEGs (550, 2,000, and 5,000 Da). Both reductive amination and acylation resulted in the conversion of hydroxyl-terminal groups of PEG to more reactive groups. It was found that in both reactions, shorter chain lengths of either chitosan or PEG resulted in increased PEGylation. The degree of substitution of PEG-N-CS was between 0.12 and 0.44, while for PEG-N,O-CS, values of 1.50 and 0.60 were obtained for chitosan functionalized with 550 and 2,000 Da PEGs, respectively. All the PEGylated materials became water-soluble, due to the high content of PEG. The authors reported that DNA condensation was much more effective when mediated by pyrene or multi-walled carbon nanotubes.

Liu et al. (2008) prepared chitosan modified by β-cyclodextrin using a Schiff base reaction between 6-O-(4-formylphenyl)-β-cyclodextrin and chitosan. The modified chitosan was used to construct different supramolecular aggregations with pyrene, multi-walled carbon nanotubes, and a combination of pyrene and multi-walled carbon nanotubes. The effect of these supramolecular aggregates on DNA condensation was also evaluated.

Amide bonds are usually formed from pre-activated carboxylic acid derivatives such as acid chlorides and anhydrides, or by using diimide derivatives (Montalbetti and Falque, 2005; Pattabiraman and Bode, 2011). This reaction (Scheme 3) can proceed in different ways: (i) formation of an intermediate acylating agent, followed by aminolysis; (ii) formation of a reactive acylating agent from the acid, followed by treatment with amine; and (iii) addition of an activating or coupling agent, resulting in the in situ generation of an acylating agent from the acid, in the presence of imine (Montalbetti and Falque, 2005; Pattabiraman and Bode, 2011). Carbodiimides have shown great potential for the attachment of molecules onto the chitosan backbone, due to their allene functional groups (Mourya and Inamdar, 2008). The compound 1-ethyl-3-(3-(dimethylamino)propyl) carbodiimide (EDC) is the most widely used carbodiimide, because of its aqueous solubility (Hermanson, 2013). However, there are other carbodiimides that can be used, such as dicyclohexylcarbodiimide (DCC) and N,N′-diisopropylcarbodiimide (DIC), which require organic solvents since they are water-insoluble. The transformation of EDC into a non-toxic urea derivative results in low toxicity associated with the reactions performed with this compound (Hermanson, 2013).

Lin and Hsu (2015) prepared galactosylated chitosan grafted with methoxy poly(ethylene glycol) (mPEG) or short chain poly(ethylene glycol) diacid (PEGd), under aqueous conditions, by amide formation employing EDC, involving the amino groups of chitosan and the carboxylic groups of PEG. The degree of substitution of PEG on the chitosan was 3.7%. Evaluation was also made of the formation of a polyplex with DNA. Chitosan grafted with mPEG formed the most stable polyplex with DNA, followed by galactosylated chitosan grafted with PEGd. Compared to naked DNA, the polyplex enhanced DNA cellular transfection in HepG2 cells.

Tan et al. (2012) prepared and characterized chitosan grafted with carboxymethyl-β-cyclodextrin (CM-β-CD) for subsequent application in anticancer drug delivery. The grafting employed carbodiimide (EDC), which promoted amide formation between the carboxylic groups of CM-β-CD and the amine groups of chitosan. The degree of substitution was measured by a colorimetric method and showed that there was about one grafted CM-β-CD on every fifteenth monosaccharide unit of chitosan (223.05 μmol g⁻¹). The carrier was able to release the anticancer drug at pH 5.0, but not at pH 7.4. This carrier could be used for anticancer drug release because it was pH-sensitive.

Although traditional organic syntheses are important tools for the production of new materials, they are inevitably time consuming and expensive (Nwe and Brechbiel, 2009). In order to overcome these limitations, Kolb et al. (2001) introduced the concept of click chemistry, establishing a set of criteria to encourage simplicity and robustness in chemical syntheses. They described a set of reactions that fulfilled these criteria, including chemoselective reactions with high yield, broad scope, removal of the product without use of chromatography, simplicity, and the ability to operate under different conditions. Any reaction fulfilling these prerequisites could be considered a click reaction. The term “click chemistry” refers to reactions where two functional groups react exclusively with each other and which can be performed at room and physiological temperatures. These reactions are divided into four major groups: (i) cycloadditions of unsaturated species; (ii) nucleophilic ring-openings; (iii) non-aldol type carbonyl chemistry; and (iv) additions to carbon-carbon multiple bonds (Hein et al., 2008). Among all the click reactions (Scheme 4), copper-catalyzed Huisgen 1,3-dipolar cycloadditions (CuAAC) of alkynes and azides are the most used (Wallyn et al., 2011; Kim and Kim, 2014).

Kulbokaite et al. (2009) studied the preparation and characterization of methoxy PEGylated chitosan (chitosan-N-MPEG) copolymers using click reactions. The polymer was synthesized by the reaction of N-azidated (azide groups) chitosan with acetylene-terminated mPEG (alkyne groups), using 1,3-dipolar cycloaddition. The reaction was successful in a medium consisting of water and methylene chloride (1:1 v/v), but failed in a solution of 5% LiCl in N-methyl-2-pyrrolidone. The failure was attributed to the high concentration of Li^I, which resulted in coordination competition with Cu^I and decreased its catalytic activity. It was shown that an equimolar ratio of CS and mPEG, or even low concentrations of mPEG, resulted in unreactivity of the alkyne groups of mPEG and consequently low degrees of substitution. Polymers synthesized with excess amounts of alkyne groups from mPEG showed degrees of substitution up to 40%, solubility in water, and residual amounts of Cu.

Chen et al. (2013) synthesized chitosan 6-OH immobilized cyclodextrin by click reaction. Firstly, the amino group of chitosan was protected by Schiff base formation with benzaldehyde, followed by the generation of C₆-OH p-toluenesulfonate. Subsequently, nucleophilic substitution of sulfonate with NaN₃ resulted in Schiff base protected chitosan (BCTS-6-N₃). Separately, alkynylic β-cyclodextrin (CD-OPg) was obtained. After the production of these substrates, the Cu (I)-catalyzed click reaction between BCTS-6-N₃ and CD-OPg resulted in the 2-benzadehylde Schiff base chitosan 6-OH immobilized cyclodextrin derivative (BCTS-6-CD). The removal of the protecting group then resulted in CTS-6-CD. The loading of immobilized cyclodextrin in the final product was 223.17 μmol g⁻¹.

In addition to the two hydroxyl groups (primary and secondary), chitosan has a third reactive site that is attractive for chemical modification: the primary amine group (Kyzas and Bikiaris, 2015). The non-bonding electron pair on the primary unit acts as a proton acceptor, making chitosan a potent nucleophile (Scheme 5) (Buschmann et al., 2013). Tosyl groups are well known in organic chemistry because they are good electrophiles and leaving groups. Tosylation of the 6-OH group of chitosan has been employed in order to produce tosyl-chitosan as an intermediate, prior to further chitosan modification. Chen et al. (2012) prepared chitin with tosylated 6-OH groups, which were displaced by a monoamino β-cyclodextrin derivative via nucleophilic substitution. Subsequently, the cyclodextrin-grafted chitin was deacetylated, producing the modified chitosan. This method was effective for the production of 6-OH substituted cyclodextrin derivatives of chitosan with high levels of substitution, where the final polymer showed a substitution capacity of around 128.68 mol g⁻¹.

Monochlorotriazinyl groups are known in organic chemistry for their capacity to form covalent bonds with nucleophilic groups. Martel et al. (2001) prepared a chitosan derivative by attaching a monochlorotriazinyl derivative of β-CD onto the chitosan backbone. This reaction occurred by nucleophilic substitution of the chloride atoms of the β-CD by the amino groups of chitosan. The high degree of substitution (2.8) resulted in an insoluble product.

Photo-initiated grafting is a simple process based on the formation of radicals by the application of electromagnetic radiation (ultraviolet or microwave). The technique can follow two different routes, either with or without a sensitizer such as methyl methacrylate (Muftuoglu et al., 2004). When the polymer is non-photolabile, the use of a sensitizer is required, with the molecule absorbing light and attaining an excited/high energy state. The sensitizer then transfers this energy to the polymer, causing dissociation of the molecule and formation of the free radicals responsible for the grafting process (Deng et al., 2009). The main photosensitizers include benzoin ethyl ether, acrylated azo dyes, aromatic ketones (benzophenone and xanthone), and metal ions. However, there is still a need for photosensitizers that are more efficient (Deng et al., 2009; Wang et al., 2010).

Sharma and Rajesh (2017) reported a fast way to prepare β-cyclodextrin-grafted chitosan by a microwave assisted method. The capability of the resultant polymer to adsorb palladium (II) was evaluated. The method employing the microwave reactor required only 5 min at 140°C to complete the reaction, whereas the conventional method necessitated approximately 6 h of stirring at 70°C. The polymers showed high selectivity toward palladium (II), with around 96% being recovered from a contaminated solution.

Graft polymerization of chitosan can also be initiated by gamma radiation and enzymes (Wang et al., 2007; Karaki et al., 2016; Li et al., 2016; Yang et al., 2016). It can therefore be seen that this is a useful technique for improving the properties of chitosan and broadening the range of its possible applications. Selection of the graft copolymerization method to be used can be based on the intended use of the hybrid polymer, and each of the different methods has variables that can influence the grafting process (Sun et al., 2003; Thakur et al., 2014).

Asamoah-Asare et al. (2014) studied the preparation and characterization of nanoparticles composed of carboxymethyl chitosan (CMC) and carboxymethyl β-cyclodextrin (CM-β-CD), in order to produce a carrier system capable of carrying and releasing both hydrophilic and hydrophobic drugs. Investigation of the feasibility of producing the nanoparticles employed different concentrations of CMC (0.05–1.50%, w/v), while maintaining a constant CM-β-CD concentration (0.8 mg/mL). The results showed that use of a low concentration of CMC produced large nanoparticles, while a high concentration of CMC produced small nanoparticles. In addition, it was observed that use of the lowest and highest CMC concentrations did not result in production of sufficient nanoparticles to be characterized. All the nanoparticles showed negative surface charge, and the zeta potential of the nanoparticles increased as the CMC concentration decreased. The nanoparticles produced were very promising as carriers for both hydrophilic and hydrophobic drugs.

In another example involving drug delivery, Daimon et al. (2014) studied the preparation and physicochemical properties of β-CD grafted onto chitosan in order to develop a possible carrier system for drugs, employing insulin as a model compound. Investigation was made of the supramolecular aggregate formation mechanism and the morphological effects of the action of ionic species. The results revealed different morphologies of the macromolecular aggregates, depending on the media conditions employed. The macromolecular structure, nanoparticles, and large aggregates were formed in acetate, citrate, and phosphate buffers, respectively. In order to simulate the environment after oral administration, the network structure formed in buffer solution was diluted with phosphate buffer and it was observed that the network was maintained. Multivalent interactions consisting of host-guest and electrostatic interactions between β-CD-grafted chitosan and insulin enabled strong binding over a wide pH range.

In another study, Kono and Teshirogi (2015) reported the preparation of smart hydrogels for use as carrier systems for drugs, consisting of β-CD-grafted CMC hydrogels with different ratios of CMC and CD (CD-g-CMCs), produced using 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) and N-hydroxysuccinimide (NHS) as the coupling method. One hydrogel was prepared containing only carboxymethyl chitosan, without addition of the CD, and CD-g-CMC gels were prepared with different CD concentrations. Evaluation was made of the swelling at pH 4, 7, and 10, as well as drug adsorption/absorption and release. The drug model employed was acetylsalicylic acid (aspirin). The hydrogels showed low swelling at pH 4 and similar absorptions at pH 7 and 10, indicating that charged groups in the hydrogels participated in the swelling mechanism. The uptake of acetylsalicylic acid by the CD-g-CMC hydrogels was strongly dependent on the amount of CD grafted. After 24 h, the CD-g-CMC hydrogels adsorbed 23.9 μmol.g⁻¹, while the CMC hydrogel only adsorbed 2.4 μmol.g⁻¹ during the same period. The CD-g-CMC hydrogels showed fast release of acetylsalicylic acid in the first 2 h, after which the release became slower. The CMC hydrogels showed more rapid release, with the majority of the drug (86%) released within 2 h.

Other potential biomedical applications of systems based on cyclodextrins grafted onto chitosan that have been reported in the literature are summarized in Table 3.

It can be seen from Table 2 that there are various potential applications of cyclodextrin-grafted chitosan systems. However, further advances are required in order to make it viable to manufacture these products at the commercial scale. Some of the synthesis procedures will need to be optimized in order to produce these grafted compounds for use in commercial products and industrial processes.

This section describes examples of the use of PEG grafted onto chitosan, considering the potential advantages of this strategy.

In a recent study by Anraku et al. (2014), chitosan was modified with PEG to increase the biocompatibility and water solubility of the polysaccharide. The formaldehyde linking method was used to modify chitosan, using different molecular weights of chitosan (22, 38, and 52 kDa) and a derivative of polyethylene glycol. Nanoparticle aggregates of PEG-CS were also prepared, and the functional and structural properties of the materials were investigated at neutral pH. The copolymer obtained possessed a cationic main chain composed of chitosan, and a non-ionic hydrophilic chain containing PEG. The degrees of substitution of the different copolymers synthesized were 25.6, 28.9, and 27.2 mol% for the mPEG-CS prepared with 22, 38, and 53 kDa chitosan, respectively. The mPEG-CS22, mPEG-CS38, and mPEG-CS53 particle sizes were 259, 413, and 452 nm, respectively, and all the copolymers exhibited almost identical zeta potentials (around +31 mV). The scavenging activities for 2,2-diphenyl-1-picrylhydrazyl (DPPH) revealed that the smallest nanoparticles had the greatest antioxidant activity.

Fu et al. (2014) reported the preparation and characterization of methoxypolyethylene glycol (mPEG) grafted onto chitosan (mPEG-g-CS) and the subsequent preparation of self-assembled polymeric micelles by the ultrasonic method. The integrity of the micelles was evaluated under different pH conditions, using 5-fluorouracil (5-FU) as a model drug. The solution pH affected the solubility of the copolymer, which increased slightly between pH 4 and 5. Micelles were not formed at solution pH greater than 7 or less than 4, and a higher pH favored larger particles. The micelles presented spherical morphology, without aggregation, and particle sizes ranged between 150 and 200 nm. The release profile showed a burst effect during the initial 0.5 h, followed by delayed release.

In another example, Hassani Najafabadi et al. (2014) prepared a PEGylated chitosan by conjugating methoxypolyethylene glycol to the hydroxyl group of chitosan. For this, the most reactive groups (NH₂) were first protected with a surfactant. The resulting copolymer was used to prepare nanoparticles loaded with ibuprofen. Dynamic light scattering measurements showed that the loaded nanoparticles presented a mean size distribution of around 80.6 nm, while AFM, TEM, and SEM analyses revealed that the nanoparticles were spherical. The encapsulation efficiency was dependent on the amount of chitosan-PEG, with an increase in the chitosan-PEG amount from 0.1 to 1% (w/v) increasing the ibuprofen encapsulation efficiency from 41 to 99%. When only chitosan was employed in the drug release experiments, the drug was completely released in 12 h. When the copolymer (chitosan-PEG) was employed, release of ibuprofen from the nanoparticles remained incomplete after 48 h. Both chitosan and the copolymer showed near-first order drug release.

Other examples of applications of PEG grafted onto chitosan are listed in Table 4.

The findings reported in the literature indicate that the strategy of grafting PEG onto chitosan is a good option for extending the applications of chitosan in different biomedical areas. However, synthesis procedures need to be developed at the industrial scale in order to enable the use of these materials in biotechnological and pharmaceutical applications, including commercial products and industrial processes.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.