C6-Modifications on chitosan

Accepted ManuscriptTitle:C6-Modifications on chitosan to develop chitosan-basedglycopolymers and their lectin-affinities with sigmoidalbinding profilesAuthor:Kazuhiro Koshiji Yuki Nonaka Maho IwamuraFumiko Dai Ryoji Matsuoka Teruaki HasegawaPII:S0144-8617(15)01053-XDOI:/doi:10.1016/j.carbpol.2015.10.073 Reference:CARP10490To appear in:Received date:3-3-2015Revised date:16-10-2015Accepted date:22-10-2015Please cite this article as:Koshiji,K.,Nonaka,Y.,Iwamura,M.,Dai,F.,Matsuoka, R.,and Hasegawa,T.,C6-Modifications on chitosan to develop chitosan-based glycopolymers and their lectin-affinities with sigmoidal binding profiles,Carbohydrate Polymers(2015),/10.1016/j.carbpol.2015.10.073This is a PDFfile of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting,typesetting,and review of the resulting proof before it is published in itsfinal form.Please note that during the production process errors may be discovered which could affect the content,and all legal disclaimers that apply to the journal pertain.A cc ep te dMa nu sc ri p tC6-Modifications on chitosan to develop chitosan-based glycopolymers and 1 their lectin-affinities with sigmoidal binding profiles2 Kazuhiro Koshiji,a, b Yuki Nonaka,a, b Maho Iwamura,c Fumiko Dai,c Ryoji Matsuoka,c and Teruaki3 Hasegawa b, c, ∗4 a Graduate School of Life Sciences, Toyo University, 1-1-1 Izumino, Itakura-machi, Ora-gun, Gumma 374-0193, Japan5 b Bio-Nano Electronics Research Centre, Toyo University, 2100 Kujirai, Kawagoe 350-8585, Japan6 cDepartment of Life Sciences, Toyo University, 1-1-1 Izumino, Itakura-machi, Ora-gun, Gumma 374-0193, Japan78 ABSTRACT910 Chitosan-based glycopolymers having multiple β-lactosides exclusively at their C6-positions were successfully synthesized from 11 partially deacetylated chitin through perfect N -deacetylation/phthaloylation and C6-selective bromination/azidation to afford 6-azide-12 6-deoxy-N -phthaloyl-chitosan and the subsequent Cu +-catalyzed Huisgen cycloadditions using alkyne-terminated β-lactoside and/or 13 quaternary ammonium modules followed by dephthaloylations. Lectin-affinities of the resultant chitosan-based glycopolymers were 14 assessed through fluorescence titration assays to show their unique sigmoidal binding profiles with amplified binding constants.1516 1. Introduction17 Carbohydrates ubiquitously exist as components of glycolipids and glycoproteins on cell surfaces and play essential roles in18 various bioprocesses (Amon, Reuven, Ben-Arye, & Padler-Karavani, 2014; Dwek, 1996; Yu, Tsai, Ariga, & Yanagisawa, 19 2011). Multivalent or clustered structures of carbohydrates are adopted in nature to enhance their affinities (Dennis & Brewer, 20 2013) towards carbohydrate-binding proteins, such as lectins. Linear polymers carrying multiple copies of pendent21 carbohydrates are called glycopolymers (GPs) and they also show amplified lectin affinities through the multivalent 22 carbohydrate-lectin interactions (Kohri et al. 2011; Miura, Koketsu, & Kobayashi, 2007; Nagatsuka, Uzawa, Ohsawa, Seto, & 23 Nishida, 2010; Narumi & Kakuchi, 2008). They are, therefore, now widely recognized as one of the most fascinating materials, 24 especially in bioorganic and medicinal chemistries (Nagatsuka et al., 2012; Sunasee & Narain, 2013). For example, 25 polyacrylamides carrying P K trisaccharides (Gal-α1,4-Gal-β1,4-Glc-β) (Gargano, Ngo, Kim, Acheson, & Lees, 2001; 26 Lundquist, Debenham, & Toone, 2000) or sialyllactoses (Sia-α2,3-Gal-β1,4-Glc-β) (Tsuchida et al. 1998) have been developed 27 to capture Shiga toxin and influenza virus, respectively. Some GPs having conductive and/or luminescent mainchains 28 (polythiophene, polyphenylene, polyphenyleneethynylene, polycarbazole, etc.) have also received increasing research efforts 29 because of their potential applications as sensory systems to detect the viruses and the toxins (Baek, Stevens, & Charych, 2000; 30 Xue, Jog, Murthy, & Liu, 2006; Kim, Wilson, & Bunz, 2005; Disney, Zheng, Swager, & Seeberger, 2004; Chen, Cheng, Zhao, 31 & Han, 2009). Other examples of GPs include polystyrenes having β-lactosides (Lac: Gal-β1,4-Glc β-) and they are now 32 commercially available as coating reagents for polystyrene culture dishes to cultivate hepatocytes. All these GPs can be 33 categorized in a same group from a viewpoint of their mainchain structures; that is, their mainchains are composed of synthetic 34 polymers those can be readily constructed through simple radical/ionic polymerizations and/or polycondensations of the 35 corresponding monomers. Such readily accessible natures of these GPs have strongly accelerated their research progress and 36 therefore, huge varieties of GPs in this group can be now found in the literature.37 On the contrary, GPs having naturally occurring polymers (polynucleotides, polypeptides, and polysaccharides) as their 38 polymeric scaffolds form the other group of GPs (Akasaka, Matsuura, & Kobayashi, 2001; Spinelli, Defrancq, & Morvan, 39 2013; Guo & Shao, 2005). These GPs are also attractive research targets in bioorganic/medicinal chemistry, green chemistry, 40A cc ep te dMa nu sc ri p t2and nano science, because of their biological functions, biodegradable natures, and unique superstructures. In spite of these1 potential applications, research progress on these GPs is relatively slow, possibly suffering from lack of simple synthetic2 strategy to access these GPs with desired structures. Recently, some research groups developed elaborate synthetic schemes to3 access polynucleotide-based GPs by applying solid phase synthetic techniques (Hunziker, 1999; Matsui & Ebara, 2012). In the4 case of polypeptide-based GPs, many research groups also reported successful syntheses of glycosylated Fmoc-amino acid5 monomers and their incorporations into multi glycosylated peptides through the solid phase peptide syntheses (Hasegawa &6 Sasaki, 2003; Maheshwari, Levenson, & Kiick, 2010; Ueki, Nakahara, Hojo, & Nakahara, 2007). On the contrary, no practical7 solid phase synthetic technique has been, however, established in carbohydrate chemistry. Direct modifications on native8 polysaccharides are sole alternative to access the polysaccharide-based GPs (Kurita, Shimada, Nishiyama, Shimojoh, &9 Nishimura, 1998). Such direct modifications are, however, still troublesome processes; that is, multiple hydroxy groups of the10 native polysaccharides have similar nucleophilicities and therefore, regioselective modifications are hardly achieved 11 (Matsuzaki, Sato, Enomoto, & Yamamoto, 1986). Such random modifications cause troublesome problems to hinder 12 developing the polysaccharide-based GPs. For example, when carbohydrate units are introduced onto the hydroxy groups of the 13 polysaccharides those participate in essential hydrogen bonding networks to construct their superstructures, the resultant 14 polysaccharide-based GPs would have disordered conformations and their potentials as chiral nano-materials would be strongly 15 spoiled.16We believe that C6-selective modifications of the polysaccharides should be the most appropriate strategy to avoid such 17 conformational disorders, since hydroxymethyl groups (6OHs) of most polysaccharides do not participate in their essential18 hydrogen bonding networks. In this decade, we have been launching our research efforts to establish general synthetic 19 procedures to access polysaccharide-based GPs carrying their pendent carbohydrates exclusively at their C6 positions. In a 20 series of our works, we established two-steps chemical modifications on native polysaccharides to access various21 polysaccharide-based GPs; that is (1) C6-selective bromination/azidation on native polysaccharides to afford the corresponding 22 6-azido-6-deoxy derivatives and (2) the subsequent Cu +-catalyzed Huisgen cycloadditions with alkyne-terminated carbohydrate 23 modules. These two-steps chemical modifications were firstly established in curdlan (Cur, linear β-1,3-glucan) chemistry, in 24 which we synthesized 6-azide-6-deoxycurdlan (Cur-N 3) through C6-selective bromination of native Cur by using 25 triphenylphosphine (PPh 3) and carbon tetrabromide (CBr 4) followed by S N 2 reaction with sodium azide (Hasegawa et al., 2006; 26 Hasegawa et al., 2007). The most significant feature of Cur-N 3 is its perfect structural homogeneity that is clearly proven by its 27 monosaccharide-like 13C NMR spectrum composed of 6 predominant peaks. The subsequent Cu +-catalyzed Huisgen couplings 28 with the alkyne-terminated carbohydrate modules proceeded with perfect chemoselectivity to afford Cur-based GPs (Cur-GPs). 29 We also applied this two-steps synthetic strategy in cellulose (Cel) chemistry to find that 6-azide-6-deoxycellulose (Cel-N 3) 30 prepared through a similar bromination/azidation process also acts as an excellent key substrate in the following Huisgen 31 cycloadditions (Yamashita, Okubo, Negishi, & Hasegawa, 2009; Negishi, Mashiko, Yamashita, Otsuka, & Hasegawa, 2011). 32 Through this strategy, we successfully developed varieties of Cel-GPs whose carbohydrate modules are attached exclusively 33 onto their C6-positions.34These two successful results in hand, we then shifted our research effort to chitosan (Chi) chemistry to develop Chi-GPs. 35 Not only biodegradability and biocompatibility but also unique immunopotentiative action of their Chi scaffolds would assure 36 great potentials of the Chi-GPs in fields of biosciences and pharmaceutical chemistry. In additions, Chi derivatives would be 37 also attractive research targets in nanoscience by taking advantages of 1D tape-like superstructures of their Chi mainchains 38 kept by the intrastranded hydrogen bonding networks (Fig. 1). It is, therefore, of quite interest to developing the Chi-GPs 39 whose carbohydrate-appendages are exclusively attached onto their C6 positions. We herein report preparation of such Chi-40 GPs and their properties including water-solubilities and lectin-affinities. 41 2. Results and Discussion42 2.1. Preparation of fully deacetylated chitosan43A cc ep te dMa nu sc ri p tIt is reasonably assumed that structural homogeneity of the starting material has critical impacts on those of the Chi-GPs. In 1 this respect, fully deacetylated Chi whose mainchains are exclusively composed of β-1,4-linked glucosaminides (GlcN) would 2 be the most appropriate starting material. Such Chi was, however, not available from commercial sources. Instead, we 3 purchased partially deacetylated chitin (PDA-chitin, Mw = ca. 10 kDa) whose mainchains are composed of random mixtures of 4 GlcN and N -acetyl-glucosaminides (GlcNAc) and then, converted it into Chi through complete deacetylation (Scheme 1). In 5 our first trial, we treated the PDA-chitin in aqueous NaOH and then, retrieved the resultant Chi through methanol 6 reprecipitation (Table 1, entry 1). We, however, found that only limited amount of Chi could be retrieved through this protocol. 7 We assumed that fragmentations of the Chi mainchains might occur during the deacetylation reaction. The similar deteriorated 8 conversion yields from PDA-chitin to Chi were also reported by Desbrieres et al. and they noted in their paper that additions of 9 NaBH 4 to the reaction mixtures are effective to improve the recovery yield (Tolaimate, Desbrieres, Rhazi, & Alagui, 2003). To10 verify this protective effect, we carried out similar deacetylation reactions in which varying amounts of NaBH 4 were added into 11 the reaction mixtures to find that NaBH 4 does have substantial effect to improve the recovery yields (entry 2-5), possibly by 12 preventing the mainchain fragmentations as mentioned by Desbrieres et al. Our trials to directly confirm the protective effects 13 of NaBH 4 towards the mainchain fragmentations, however, turned to be failure, since these Chi samples were hardly soluble in 14 DMF and therefore, we cannot carry out any gel permeation chromatographic (GPC) analyses to assess their molecular 15 weights.16We then assessed detailed chemical structures of Chi, together with that of the PDA-chitin, through 1H NMR spectral 17 analyses, in which D 2O acidified with small droplets of HCl aq was used as deuterated solvents. As shown in Fig. 2-a, the18 PDA-chitin shows a highly complicated and hardly assignable 1H NMR spectrum in which a broad singlet and a multiplet are 19 found at 2.18 and 3.30 ppm, respectively. According to the literature, the former and the latter are assignable to 2CH-20 NHCOC H 3 and 2C H -NH 3+, respectively, and degree of deacetylation (DDA) value of the PDA-chitin can be, thus, estimated to21 0.80 by comparison of these two peak areas (Hirai, Odani, & Nakajima, 1991). On the other hand, Chi showed a 1H NMR 22 spectrum in which the acetamide-derived peaks nearly disappears, indicating complete removal of the acetamide groups (Fig. 23 2-b). It should be noted that clear doublet (4.95 ppm) and triplet peaks (3.25 ppm) are indicative of excellent structural 24 homogeneity of Chi, again proving the complete removal. The complete removal of the acetamide groups was also proven by 25 their IR spectra in which an acetamide-derived peak at 1643 cm -1 observed for the PDA-chitin became negligible through this 26 deacetylation step (Fig. 3). We also carried out 13C NMR spectral analyses on Chi to assess their detailed chemical structures. 27 As shown in Fig. 4-a, the PDA-chitin showed a highly complicated 13C NMR spectrum composed of more than 21 peaks, 28 proving its highly heterogeneous structure. On the contrary, Chi showed a simple 13C NMR spectrum composed of 29 monosaccharide-like 6 predominant peaks (Fig. 4-b). All these 1H/13C NMR and IR spectral features indicate that Chi 30 mainchains are exclusively composed of the repeating GlcN units.312.2. Preparation of N-phthaloyl-chitosan32The amino groups of Chi were then protected by phthaloyl groups through a similar procedure reported in the literature; that 33 is, Chi was treated with phthalic anhydride in aqueous N ,N -dimethylformamide (DMF) to afford N -phthaloyl-chitosan (ChiPht) 34 (Satoh & Sakairi, 2003). The resultant ChiPht shows an IR spectrum in which two characteristic peaks (1715 and 1775 cm -1) 35 are indicative of the successful introduction of N -phthaloyl groups (Fig. 5-a). We tried to assess chemical structure of ChiPht 36 through 13C NMR spectral analysis, it was, however, hardly soluble in commonly used deuterated solvents, such as D 2O, 37 DMSO-d 6, DMF-d 7, and CDCl 3 and therefore, we could not directly assess its detailed chemical structure. Alternatively, we 38 treated a small aliquot of ChiPht with acetic anhydride in pyridine, and the resultant CDCl 3-soluble per -acetate (ChiPht-3,6Ac) 39 was subjected to 13C NMR spectral analysis. Highly homogeneous structure of ChiPht-3,6Ac was clearly proven by its simple40 13C NMR spectrum; that is, only 6 predominant peaks originated from the 6-menbered ring structure of the GlcN units were41 observed at 100-50 ppm region (Fig. 6-a). The other predominant peaks can be also assigned to N -phthaloyl and 3,6-O -acetyl 42 groups of ChiPht-3,6Ac. Especially, 4 predominant peaks derived from the N -phthaloyl groups clearly prove that the amino43A cc ep te dMa nu sc ri p t4groups are protected not as N -phthalamide groups but as N -phthalimide ones. All these spectral features of ChiPht-3,6Ac1 clearly indicate the excellent structural homogeneity of ChiPht from which ChiPht-3,6Ac was synthesized. It should be noted2 that some minor peaks, especially those at 100-50 ppm region in this spectrum, suggest deteriorated structural homogeneity of3 ChiPht-3,6Ac. We, however, assumed that not the N -phthaloylation step but the O -acetylation one should be responsible for4 this structural heterogeneity. Our assumption was clearly supported by a fact that other Chi derivatives prepared from ChiPht5 showed no such minor peak, as described below.6 2.3. Preparation of 6-bromo-6-deoxy-N-phthaloyl-chitosan7 We then tried to convert ChiPht to 6-bromo-6-deoxy-N -phthaloyl-chitosan (ChiPht-Br) whose mainchains are exclusively 8 composed of 6-bromo-6-deoxy-N -phthaloyl-β-D -glucosaminides (GlcPht-Br). We carried out the bromination reactions by9 using various reagent/solvent systems to confirm that the best result was achieved by using N -bromo-succinimide (NBS) in dry 10 N -methylpyrrolidone (NMP) (Ifuku, Wada, Morimoto, & Saimoto, 2011). Briefly, we dissolved ChiPht in dry NMP containing 11 triphenylphosphine (PPh 3) and NBS and the resultant mixture was stirred at 80 ºC for 2 h. The resultant ChiPht-Br was then12 retrieved from the reaction mixture through water precipitations. The resultant precipitate contained various water-insoluble 13 impurities, such as triphenylphosphine oxide. It was, therefore, essential to wash the precipitate with MeOH repeatedly to 14 obtain pure ChiPht-Br. As shown in Fig. 6-b, 13C NMR spectrum of ChiPht-Br is composed of 10 predominant peaks of which 15 6 and 4 peaks are attributable to their mainchain structures and N -phthaloyl protecting groups, respectively. No unassignable 16 peak can be found in this spectrum, indicating that all hydroxymethyl groups of ChiPht were quantitatively/selectively 17 converted into bromomethyl ones without any side reactions. 18 2.4. Preparation of 6-azide-6-deoxy-N-phthaloyl-chitosan19 We successfully converted ChiPht-Br into 6-azide-6-deoxy-N -phthaloyl-chitosan (ChiPht-N 3) through S N 2 type reactions 20 using NaN 3. The resultant ChiPht-N 3 shows an IR spectrum having a characteristic N 3-derived peak at 2108 cm -1 (Fig. 5-c). 21 Together with a fact that the IR peak stemmed from the N -phthaloyl groups remains intact, these spectral features are strongly 22 diagnostic of the successful synthesis of ChiPht-N 3. Structural homogeneity of ChiPht-N 3 is also proven by its 13C NMR23 spectrum in which the 13C peak derived from bromomethyl groups (–C H 2Br: 33.33 ppm) entirely disappears and that from 24 azidomethyl ones (–C H 2N 3: 50.12 ppm) newly appears (Fig. 6-c). It should be herein noted that the bulky N -phthaloyl groups 25 have substantial roles to obtain ChiPht-N 3 with excellent structural homogeneity. As shown in Fig. S1, Cel-N 3 synthesized26 from Cel showed a 13C NMR spectrum including some minor peaks, indicating its lower structural homogeneity than that of 27 ChiPht-N 3 (Miyazawa et al., 2014). These two polysaccharide derivatives both have β-1,4-linked glucan mainchains, structural28 difference between ChiPht-N 3 and Cel-N 3, therefore, entirely relies on structures of their C2 positions on which the former has 29 N -phthaloyl groups and the latter has hydroxy ones. We assume that the bulky N -phthaloyl groups of ChiPht should reduce 30 reactivity of their neighboring 3OHs on the bromination reaction, although low solubilities of Cel-Br in the commonly used31 deuterated solvents made it quite difficult to directly compare structural homogeneities of these two brominated polysaccharide 32 derivatives (ChiPht-Br and Cel-Br). Several research groups also noted the steric shielding effects of the N -phthaloyl groups in 33 the bromination reaction on ChiPht (Furuhata, Chang, Aoki, & Sakamoto, 1992). 34 2.5. Preparation of chitinosan-based GPs.35 We have two alternative routes from ChiPht-N 3 to access the Chi-GPs. One is composed of prior Cu +-catalyzed Huisgen 36 cycloadditions on ChiPht-N 3 followed by dephthaloylation. The other is prior dephthaloylation to produce 6-azide-6-deoxy-37 chitosan (Chi-N 3) followed by the Cu +-catalyzed Huisgen cycloadditions. These two alternative routes both involve their own 38 problems. The former, for example, cannot be applied to pendent oligosaccharides those are labile to dephthaloylation 39 conditions. In this respect, the latter has a significant advantage to certify its wide applications, since the Cu +-catalyzed 40 Huisgen cycloadditions are so chemoselective that co-existing functionalities do not undergo any side reactions. This route has, 41 however, one synthetic difficulty in its dephthaloylation step: that is, azidomethyl groups of ChiPht-N 3 would be 42A cc ep te dMa nu sc ri p tsimultaneously reduced by hydrazine in this dephthaloylation step. In fact, Nagasaki et al. treated ChiPht-N 3 with hydrazine in 1 an NMP/water mixed solvent system at 100 ºC to access 6-amino-6-deoxychitosan whose mainchain are exclusively composed 2 of 6-amino-6-deoxy-glucosamines (Satoh et al., 2007). Note that, under this reaction condition, both N -phthalimide and azide 3 groups of ChiPht-N 3 were completely/simultaneously converted into amino ones. This result clearly shows that it is quite 4 difficult to selectively remove N -phthaloyl groups without reducing the co-existing azido groups. In fact, Ifuku et al. adopted 5 the former route through which they synthesized Chi derivatives having triazole-tethered hydroxymethyl or phenyl groups at 6 their C6 positions (Ifuku, Wada, Morimoto, & Saimoto, 2011). At the first stage of this work, we tried to achieve selective N -7 dephthaloylation reactions under mild reaction conditions. Our trials turned to be failure and we obtained heterogeneous Chi 8 derivatives in which some azide groups were simultaneously reduced to amino ones even under mild reaction conditions (low 9 temperature, short reaction time, limited amounts of hydrazine, etc., data not shown). We, instead, adopted the former synthetic10 route to achieve the Chi-GPs with high structural homogeneity.11Introductions of multiple Lac modules onto the Chi scaffolds were achieved through Cu +-catalyzed Huisgen couplings using 12 ChiPht-N 3 and 2-propargyl-β-lactoside (Lac-yn) (Otsuka, Sakurai, & Hasegawa, 2009). Briefly, we dissolved ChiPht-N 3 in 13 DMSO and then, Lac-yn, CuBr 2, ascorbic acid, and propylamine were added into the reaction mixture to be kept at the ambient 14 temperature for 2 days. The resultant dialysis followed by lyophilization afforded N -phthaloyl-chitosan having triazole-tethered15 Lac modules (ChiPht-Lac). We also carried out a similar coupling procedure in which not only Lac-yn but also N ,N ,N -16 trimethyl-2-propargyl ammonium chloride (N +-yn) were added in the reaction mixture to develop a cationic Chi-GP having 17 both Lac and N ,N ,N -trimethyl ammonium (N +) modules along its mainchain (ChiPht-Lac/N +) (Kawagoe, Kasori, & Hasegawa, 18 2011). Successful syntheses of ChiPht-Lac and ChiPht-Lac/N + were confirmed by their IR spectra, in which no N 3-derived 19 peak was observed, clearly indicating that all azide groups in ChiPht-N 3 were successfully converted into the triazole-tethered 20 Lac or Lac/N + modules, respectively (Fig. 5-d and e). It should be noted that not only the aforementioned N 3-derived peak but 21 also Pht-derived ones also disappeared in the IR spectra of ChiPht-Lac and ChiPht-Lac/N +. In additions, new peaks 22 simultaneously appeared at around 1650 cm -1. These spectral changes imply that the N -phthaloyl groups were decomposed to 23 give N -phthalamide-type protecting groups during the Cu +-catalyzed Huisgen cycloadditions (for its chemical structure, see Fig 24 5 inset). We assume that propylamine as basic catalysts should be responsible for the decomposition of the N -phthaloyl groups. 25We carried out 13C NMR analyses to assess structural homogeneity of ChiPht-Lac. It was, however, quite difficult to obtain 26 its clear 13C NMR spectrum, suffering from its low solubility in the commonly used deuterated solvents. In fact, we cannot27 obtain any informative 13C NMR spectrum even after 3 days accumulation at 60 ºC. Not only the low solubilities, but also rigid 28 macromolecular nature of ChiPht-Lac also strongly hinders achieving its clear 13C NMR spectrum. On the other hand, in the29 case of 1H NMR, we found that ChiPht-Lac showed a broad singlet peak attributable to the 1,4-triazole-linkers at 8.32 ppm 30 (Fig. 7-a, top). Interestingly, ChiPht-Lac also shows broad peaks at 8.07 and 7.79 ppm, suggesting that the N -phthaloyl groups 31 were converted during the Huisgen cycloaddition into the N -phthalamide-type protecting groups carrying N -propylamido32 substituents (see Fig. 7, inset).331H NMR spectrum of ChiPht-Lac/N + also shows the similar broad peak at 7.78 ppm, again indicating that ChiPht-Lac/N +34 carries the N -phthalamide-type protecting groups (Fig. 7-b, top). In additions, Lac-N + ratios of ChiPht-Lac/N + were also 35 assessed based on this 1H NMR spectrum; that is, broad peaks at around 8.31 and 3.19 ppm can be attributed to triazolyl and 36 trimethyl protons, respectively, and the Lac-N + ratios was estimated to 0.84-0.16 by integrations of these peaks.37 We then treated ChiPht-Lac and ChiPht-Lac/N + in hydrazine monohydrate to remove their N -phthalamide-type protecting 38 groups. The resultant Chi-GPs (Chi-Lac and Chi-Lac/N +, respectively) showed 1H NMR spectra in which the broad peaks 39 derived from the N -phthalamide-type protecting groups entirely disappeared, confirming the successful syntheses of Chi-Lac 40 and Chi-Lac/N + with free amino groups (Fig. 7-a and b, bottom). 41 2.6. Water-solubilities of the Chi-GPs42A cc ep te dMa nu sc ri pt6Water-solubilities of the Chi-GPs, together with that of Chi, were assessed as described below. Excess amounts of these1 Chi derivatives and small amounts of water were added into small vessels and then, the resultant mixtures were incubated at the2 ambient temperature. After centrifugation to remove precipitates, concentrations of the resultant supernatants were determined3 through well-known phenol-sulfuric acid protocol. Although Chi was not soluble in water at all, Chi-Lac could be dissolved4 into water to some extent, although its water solubility was still low ([Chi-Lac] = 2.75 × 10-4 M). We assume that nonionic Lac 5 modules of Chi-Lac are not so effective to break interstranded hydrophobic interactions and/or hydrogen bonding networks6 among its Chi mainchains. On the other hand, the water-solubility of the Chi-GPs was much improved by the co-introductions7 of the cationic N + modules; that is, Chi-Lac/N + showed excellent water solubility and homogeneous solutions with high [Chi-8 Lac/N +] can be readily prepared ([Chi-La/N +] = 1.84 × 10-3 M or higher). In fact, the water solubility of Chi-Lac/N + was so9 high that we could not prepare its saturated aqueous solution. We, therefore, could not precisely assess its water solubility10 through our protocol.112.7. Lectin-affinity of the Chi-GPs12 We assessed lectin-affinities of the Chi-GPs through fluorescence titration assays using fluorescein isothiocyanate-labeled 13 lectins (FITC-lectins). Briefly, in these binding assays, we added small aliquots of Chi-GPs solutions into FITC-lectins in Tris-14 HCl buffer (20 mM, pH 7.2) every 5 min and measured fluorescence spectra of the resultant solutions just before the next 15 additions. As shown in Fig. 8-a, fluorescence intensity (I ) of FITC-RCA 120 (Ricinus communis agglutinin, βGal-specific) was 16 strongly weakened with increasing concentrations of Chi-Lac/N +, indicating their strong interaction. On the other hand, FITC-17 ConA (Concanavalin A, αMan/αGlc-specific) showed negligible decrement in its fluorescence intensity on additions of Chi-18 Lac/N +, showing their weak and/or negligible interactions. Monovalent Lac-yn also induced negligible changes in the 19 fluorescence intensities of FITC-lectins, indicating that the strong interaction between Chi-Lac/N + and FITC-RCA 120 arises 20 from the well-known carbohydrate cluster effects.21 Of our great interest, binding profiles of Chi-Lac/N + towards FITC-RCA 120 was sigmoidal rather than Langmuir-type one22 and accurately traced with the following Hill’s equation (eq. 1)23 I = I 0 + (I ∞ - I 0) K a n [Chi-GPs]n /(1+ K a n [Chi-GPs]n ) (eq. 1)24 where I 0, I ∞, K a , and n are relative fluorescence intensity of FITC-labeled lectins with no and excess amounts of the Chi-25 GPs, association constant, and Hill’s coefficient, respectively. Our software (Deltagraph ver 7.0.5) does not support 26 computational curve fittings with exponential variables, we, therefore, pre-substituted 1.0, 1.1, 1.2, and so forth for n and then,27 carried out the computational curve fittings (n = 1.0 ~ 4.0). As shortly summarized in Table 2, the best result was obtained by 28 the curve fitting in which 1.9 was pre-substituted for n (rms = 0.999, plane curve in Fig. 8-top), although a binding curve 29 obtained by using Langmuir-type equation (n = 1.0) poorly fit to the experimental data (rms = 0.973, dotted curve in Fig. 8-30 top).31 Note that, in this Hill’s equation, the n values indicate numbers of the Chi-GPs strands simultaneously binding to single 32 lectin molecule. These data, therefore, indicate that ca. two Chi-Lac/N + strands simultaneously bind to one FITC-RCA 120 33 molecule. Chi-Lac also showed the similar sigmoidal binding profile towards FITC-RCA 120 in which n value was estimated to 34 3.8, again indicating that ca. four Chi-Lac strands simultaneously bind to one FITC-RCA 120 molecule. Possible mechanisms to 35 reasonably explain such multiple bindings include aggregations of Chi-Lac/N + and Chi-Lac strands during the binding assays 36 (Fig. 9). Strong interstranded interactions among the Chi-GPs strands should be responsible for the aggregations; that is, the 37 resultant aggregates carrying larger numbers of Lac appendages than single Chi-GPs strand should show higher affinities than 38 the single Chi-GPs strand39 As summarized in Table 2, the K a values of Chi-Lac/N + and ChiN-Lac can be estimated to 7.56 × 104 and 8.75 × 105 M -1, 40 respectively. We then converted the K a into K a-Lac those are association constants based on [lactose-unit] to quantitatively assess41。