Full Papers

Well-Defined Poly(N-glycosyl 1,2,3-triazole) Multivalent Ligands:

Design, Synthesis and Lectin Binding Studies
Jin Geng, Josefina Lindqvist, Giuseppe Mantovani, Gaojian Chen, Claire T. Sayers, Guy J. Clarkson and
David M. Haddleton*
Department of Chemistry, University of Warwick, Coventry CV4 7AL, UK, E-mail: [email protected]

Keywords: ATRP, Click, Glycopolymer, Lectin recognition, Multivalent ligands

Received: June 28, 2007; Accepted: October 01, 2007

DOI: 10.1002/qsar.200740089

Abstract
Glycopolymers have been synthesised by post-functionalisation of well-defined alkyne-
functional polymers with sugar azides to yield N-glycosyl 1,2,3-triazole functional poly-
mers. The Cu(I)-catalysed Huisgen cycloaddition was used to attach a-mannoside, b-
galactoside and b-lactoside derivatives via an azide functionality bound directly to the
sugar anomeric carbon. Three different catalytic systems were investigated for the click
reactions; [(PPh3)3Cu(I)Br], TBTA/Cu(I)Br and bathophenanthrolinedisulphonic acid
disodium salt/Cu(I)Br. The latter of these was found to be the most efficient for the
attachment of the larger/more sterically hindered disaccharide lactose moiety. The
interaction of the lactose- and galactose-bearing glycopolymers with Ricinus Communis
Agglutinin (RCA I) lectin was investigated by affinity HPLC analysis. The rate of the
interaction between mannose polymer and concanavalin A (Con A) lectin was assessed
by turbidimetry. The results from the lectin conjugation studies indicate that the
glycopolymers prepared in this work are able to function as multivalent ligands, further
suggesting that the attachment of the triazole directly to the sugar anomeric carbon has
no significant effect on the interaction of these glycopolymers with Con A and RCA I.

1 Introduction interact with lectins as multivalent ligands in a similar

manner to natural glycoproteins. Glycopolymers have
Carbohydrates are important for many biological process- been mainly synthesised using two different approaches;
es, including inflammation, cell – cell contacts, fertilisation polymerisation of a sugar-containing monomer, or post-
and signal transmission [1 – 5]. Mother nature has provided functionalisation of pre-formed polymers using sugar moi-
sugars with a great coding capacity, and the ability to store eties [18]. The development of controlled radical polymer-
biological information in oligosaccharide structures in gly- isation techniques has rendered it possible to produce
coproteins and glycolipids [6]. Lectins are carbohydrate- well-defined synthetic glycopolymers in a convenient man-
binding proteins that can selectively recognise and deci- ner. Transition Metal-Mediated Living Radical Polymeri-
pher the glycocode in oligosaccharides. Whereas individu- sation (TMM-LRP, often termed ATRP) has been widely
al protein – saccharide interactions are typically weak, the studied and proven useful for the synthesis of macromole-
multivalent interactions employed in biological systems cules with controlled molecular weight and narrow poly-
can be characterised by high affinity and high specificity dispersity [19, 20]. In particular, the polymerisation of sug-
[7]. The affinity of the lectin towards the multivalent li- ar-containing monomers by Living Radical Polymerisation
gand is even higher than would be expected only from the (LRP) to give well-defined synthetic glycopolymers has
increase in sugar concentration, the Dcluster glycoside ef- been reported [21 – 31].
fectE [6, 8 – 10]. Further understanding of the factors con- Using the approach of post-functionalisation with sugar-
trolling the protein – saccharide interactions is important containing compounds [32 – 34], it is possible to produce li-
for the development of new therapeutic strategies based braries of glycopolymers with the same macromolecular
on multivalent drugs [7]. features (degree of polymerisation, macromolecular archi-
Due to their biomimetic properties and many potential tecture, Mw/Mn) by attaching different sugar moieties to
applications, there is an increasing interest in synthetic pre-formed polymer scaffolds. This is of importance, for
polymers with pendent sugar moieties [11 – 18]. It has example, in studies focused on the effect of the density of
been shown that these synthetic glycopolymers are able to a particular sugar epitope on the biological behaviour of

1220 G 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim QSAR Comb. Sci. 26, 2007, No. 11-12, 1220 – 1228
Well-Defined Poly(N-glycosyl 1,2,3-triazole) Multivalent Ligands: Design, Synthesis and Lectin Binding Studies

glycopolymers, as carbohydrate-lectin recognition can be 2.2 Analysis

strongly dependent on the sugar polymer chain length
[35]. The post-functionalisation approach generally also All reactions were carried out using standard Schlenk
offers a simpler synthetic procedure, as some sugar-con- techniques under an inert atmosphere of oxygen-free ni-
taining monomers have the tendency to self-polymerise trogen, unless otherwise stated. Compounds were visual-
during purification procedures. For the attachment of sug- ised by use of UV light (254 nm) or a basic solution (10%
ars, click reactions, and in particular the Cu(I)-catalysed w/w K2CO3 in water) of KMnO4. Merck 60 (230 – 400
version of Huisgen 1,3-dipolar cycloaddition [36, 37], have mesh) silica gel was used for column chromatography.
been receiving increasing attention due to their attractive NMR spectra were obtained on a Bruker DPX300 and
features, which include high efficiency, stereoselectivity, Bruker DPX400 spectrometer. All chemical shifts are re-
mild reaction conditions and excellent functional group ported in ppm (d) relative to tetramethylsilane, referenced
compatibility [38]. to the chemical shifts of residual solvent resonances (1H
We have previously reported on the lectin recognition and 13C). The following abbreviations were used to explain
of synthetic glycopolymers obtained by Huisgen 1,3-dipo- the multiplicities: d ¼ doublet, m ¼ multiplet, t ¼ triplet.
lar cycloaddition of O-propenyl sugar azides with well-de- The molecular weight of the polymers Mn(NMR) were cal-
fined alkyne functional poly(methacrylates) prepared by culated by comparing the integrals of the benzyl chain-end
LRP [34]. In this study, we describe an alternative ap- signals and appropriate peaks related to the polymer back-
proach to glycopolymers via click reaction of unprotected bone.
sugar azides with alkyne functional scaffolds to produce Molar mass distributions of polymers (5a) and (5b) were
N-glycosyl 1,2,3-triazole polymers. Lectin-binding studies measured using Size Exclusion Chromatography (SEC),
were conducted using the model lectins Concanavalin A on a system equipped with two PL gel 5 mm mixed D col-
(Con A) and Ricinus Communis Agglutinin I (RCA I) to umns (300  7.5 mm) and one PL gel 5 mm guard column
investigate whether multivalent recognition is possible (50  7.5 mm; Polymer Laboratories, suitable for molecu-
also with the triazole attached directly to the sugar anome- lar weights between 200 and 400 000 g/mol) with differen-
ric carbon. tial refractive index detection using THF/triethylamine
95 : 5 v/v at 1.0 mL/min as the eluent. Poly(MMA) stand-
ards (3  105 – 200 g/mol) were used to calibrate the SEC.
Analyte samples contained (0.2% volume) toluene as the
2 Experimental Section
flow marker. Molar mass distributions of polymers (6 – 8)
were determined by SEC in a system equipped with a
2.1 Materials
guard column and two mixed C columns (2  106  200 g/
Copper(I) bromide (Aldrich, 98%) was purified according mol) with differential refractive index detection, and using
to the method of Keller and Wycoff [39]. [(PPh3)3CuBr] N,N-Dimethylformamide (DMF)/0.1m LiBr at 0.8 mL/min
[40] and Tris-(Benzyl-Triazolylmethyl)Amine (TBTA) [41] as eluent. Tubidimetry assay was performed using a Varian
were prepared following known synthetic protocols. 3-Tri- Cary 50 Bio UV – Vis spectrometer, using 2 mL volume
methylsilanyl-prop-2-yn-1-ol necessary for the preparation polycarbonate cuvettes (1 cm pathlength). RCA I recogni-
of the polymers (5a) and (5b) was either purchased from tion experiments were performed using an HPLC system
Sigma – Aldrich or synthesised following a known proce- equipped with a Shodex AFpak ARC-894 column, a Dio-
dure (for large-scale preparation) [42]. Polyalkyne poly- nex P680 HPLC pump and a Gilson UV – Vis-55 detector.
mers (5a) and (5b) were prepared as described earlier [34] Infrared absorption spectra were recorded on a Bruker
and stored at 0 8C. Peracetylated sugars were prepared by VECTOR-22 FTIR spectrometer using a Golden Gate di-
slight modification of a known general protocol [43]. O- amond attenuated total reflection cell. Mass spectra were
Ethenyl mannose azide (4) was prepared as previously de- recorded using a Micromass Autospec apparatus. Tur-
scribed [44], starting from 2-azidoethyl 2,3,4,6 tetra-O-ace- bidimetry was performed as described by Kiessling and
tyl-a-d-mannopyranoside intermediate prepared as de- coworkers [46]. The yields are not optimised.
scribed by Osborn and coworkers [45]. Triethylamine
(Fischer, 99%) was stored over sodium hydroxide pellets.
2.3 Crystal Structure Determination
All other reagents and solvents (including the bathophe-
nanthrolinedisulphonic acid disodium salt ligand) were ob- Crystallographic data for the complexes C30H30BrCuN10
tained at the highest purity available from Sigma – Aldrich and C30H30Br2CuN10 are summarised in Table 1. Suitable
Chemical Company and used without further purification crystals were quickly glued to quartz fibres, coated in dry
unless stated. Nujol and cooled in the cold nitrogen gas stream of the dif-
CAUTION!: Although we had never experienced any fractometer. The structure was solved by direct methods.
adverse event when working with these products, organic Anisotropic thermal parameters were used for all non-H
azides are potentially explosive compounds and they atoms whilst hydrogen atoms were inserted at calculated
should be handled with utmost care!! positions and fixed, with isotropic thermal parameters, rid-

QSAR Comb. Sci. 26, 2007, No. 11-12, 1220 – 1228 www.qcs.wiley-vch.de G 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1221
Full Papers Jin Geng et al.

Table 1. Selected bond lengths (R) and angles (8).

C30H30BrCuN10a C30H30Br2CuN10a
Cu(1)N(110) 2.183(3) Cu(2)N(210) 2.079(6)
Cu(1)N(114)b 2.503b Cu(2)N(214) 2.115(11)
Cu(1)Br(1) 2.503(7) Cu(2)Br(2) 2.398(2)
N(110)Cu(1)N(110) 111.93(18) N(210)Cu(1)N(210) 117.03(11)
N(110)Cu(1)Br(1) 106.9(2) N(210)Cu(1)N(214) 79.96(19)
N(210)Cu(1)Br(2) 100.04(19)
N(214)Cu(1)Br(2) 180.00(19)
a
Assignments were based on similarities with known crystal structures of copper complexes bearing ligands with analogous structure [53, 54].
b
The length of this bond is such that it is quite difficult to say for certain whether or not the nitrogen donor is effectively coordinated to the metal centre.

ing on the supporting atom. The structure solution was car- CH3C(O)O]. IR (neat): n~ ¼ 1739, 1639, 1435, 1367, 1208,
ried out using SHELXTL version 5.0 software on a Silicon 1174, 1143, 1040, 939, 899, 738 cm1. Anal. calcd. for C28
Graphics Indy workstation, refinements were carried out H38O19: C, 49.56; H, 5.64; O, 44.80; Found: C, 49.20; H,
using SHELXTL 96 software, minimising on the weighted 5.74; O, 44.56. MS(ESI): 641.2 (17) 701.2 (M þ Na) (100)
R factor, wR2. 702.0 (30) 703.1 (7) 717.1 (M þ K) (8).

2.4 Synthesis of the Sugar Azides 2.4.2 Acetylated N-Glycosyl Azides. General Procedure:
b-Azido-d-lactose Heptaacetate
2.4.1 Peracetylated Sugars. General Procedure: d-Lactose
Trimethylsilyl azide (1.61 mL, 12.2 mmol) was added un-
Octaacetate
der nitrogen to a solution of d-lactose octaacetate (4.14 g,
Acetic anhydride (48.5 mL 0.513 mol) was mixed with 6.11 mmol) in anhydrous CH2Cl2 (50 mL), followed by ti-
(18 g 0.052 mol) d-lactose and the resulting suspension n(IV) chloride (0.43 mL, 3.7 mmol). The mixture was
stirred for 5 min. Three drops of concentrate sulphuric stirred at ambient temperature for 12 h, then the solvent
acid were added and a substantial increase in the tempera- was removed under reduced pressure and the crude resi-
ture was detected. The solution was allowed to cool down due was purified by flash chromatography (CC, SiO2, ethyl
to ambient temperature, then CH2Cl2 (100 mL) was added acetate/petroleum ether 1 : 1 v/v). The relevant fractions
under stirring. After 10 min the solution was washed with were collected, combined and concentrated to dryness un-
saturated aqueous sodium bicarbonate solution (3  der reduced pressure. Obtained 3.67 g (91%) of b-azido-d-
100 mL) and finally with brine (2  200 mL). The organic lactose heptaacetate as a white solid. 1H NMR
layer was dried over sodium sulphate, filtered and the vol- (400.03 MHz, CDCl3, 298 K) d ¼ 1.88 (s, 3H, CH3), 1.89 (s,
atiles removed under reduced pressure. Obtained 32.0 g of 3H, CH3), 1.94 (s, 3H, CH3), 1.96 (s, 3H, CH3), 1.97 (s, 3H,
d-lactose octaacetate (89%). CH3); d ¼ 1.98 (s, 3H, CH3), 2.01 (s, 3H, CH3), 2.11 (s, 3H,
1
H NMR (400.03 MHz, CDCl3, 298 K) d ¼ 1.88 (s, 3H, CH3), 3.71 – 3.79 (m, 1H, CH), 3.81 – 3.89 (m, 1H, CH),
CH3), 1.89 (s, 3H, CH3), 1.94 (s, 3H, CH3), 1.96 (s, 3H, 3.91 – 4.06 (m, 1H, CH), 4.33 (s, 1H, CH), 4.34 – 4.40 (m,
CH3), 1.97(s, 3H, CH3); d ¼ 1.98 (s, 3H, CH3), 2.01 (s, 3H, 2H, CH2), 4.41 – 4.51 (m, 2H, CH2), 4.85 – 4.92 (m, 1H,
CH3), 2.11 (s, 3H, CH3), 3.71 – 3.79 (m, 1H, CH), 3.81 – CH), 4.86 – 4.99 (m, 1H, CH), 4.99 – 5.15 (m, 1H, CH), 5.25
3.89 (m, 1H, CH), 3.91 – 4.06 (m, 1H, CH), 4.33 (s, 1H, (t, J ¼ 9.0 Hz, 1H, CH), 5.39 (t, J ¼ 9.3 Hz, 1H, CH), 5.38
CH), 4.34 – 4.40 (m, 2H, CH2), 4.41 – 4.51 (m, 2H, CH2), (d, J ¼ 2.8 Hz, 1H, CHN3). 13C{1H} NMR (100.59 MHz,
4.85 – 4.92 (m, 1H, CH), 4.86 – 4.99 (m, 1H, CH), 4.99 – CDCl3, 298 K) d ¼ 20.60 (1C, CH3), 20.74 (1C, CH3), 21.78
5.15 (m, 1H, CH), 5.25 (t, J ¼ 9.0 Hz, 1H, CH), 5.39 (t, J ¼ (1C, CH3), 20.81 (1C, CH3), 20.82 (1C, CH3), 20.89 (1C,
9.3 Hz, 1H, CH), 5.61 (d, J ¼ 8.0 Hz, 1H, CH), 6.18 (d, J ¼ CH3), 21.02 (1C, CH3), 21.11 (1C, CH3), 60.89 (1C, CH2
3.8 Hz, CH). 13C{1H} NMR (100.59 MHz, CDCl3, 298 K) OAc), 60.57 (1C, CH2OAc), 66.70 (1C, CH), 69.23 (1C,
d ¼ 20.60 (1C, CH3), 20.74 (1C, CH3), 21.78 (1C, CH3), CH), 69.71 (1C, CH), 70.05 (1C, CH), 70.11 (1C, CH),
20.81 (1C, CH3), 20.82 (1C, CH3), 20.89 (1C, CH3), 21.02 70.23 (1C, CH), 70.71 (1C, CH), 75.21 (1C, CH), 87.83 (1C,
(1C, CH3), 21.11 (1C, CH3), 60.89 (1C, CH2OAc), 60.57 CHN3), 101.23 (1C, CH), 169.35 [1C, CH3C(O)O], 169.36
(1C, CH2OAc), 66.70 (1C, CH), 69.23 (1C, CH), 69.71 (1C, [1C, CH3C(O)O], 169.45 [1C, CH3C(O)O], 169.66 [1C,
CH), 70.05 (1C, CH), 70.11 (1C, CH), 70.23 (1C, CH), CH3C(O)O], 169.89 [1C, CH3C(O)O], 170.02 [1C,
70.71 (1C, CH), 75.21 (1C, CH), 91.63 (Canomeric), 101.31 CH3C(O)O], 170.23 [1C, CH3C(O)O], 170.51 [1C,
(1C, CH), 169.01 [1C, CH3C(O)O], 169.22 [1C, CH3C(O)O]. IR (neat): ~n ¼ 2118, 1708, 1378, 1227, 1167,
CH3C(O)O], 169.69 [1C, CH3C(O)O], 169.72 [1C, 1091, 1039, 960, 915, 775, 714, 611 cm1. Anal. calcd. for
CH3C(O)O], 169.79 [1C, CH3C(O)O], 169.82 [1C, C26H35N3O17: C, 47.20; H, 5.33; N, 6.35; O, 41.11; Found: C,
CH3C(O)O], 169.88 [1C, CH3C(O)O], 169.89 [1C, 47.11; H, 5.43; N, 6.00; O, 40.21. MS(ESI): 105 (86) 149

1222 G 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.qcs.wiley-vch.de QSAR Comb. Sci. 26, 2007, No. 11-12, 1220 – 1228
Well-Defined Poly(N-glycosyl 1,2,3-triazole) Multivalent Ligands: Design, Synthesis and Lectin Binding Studies

(47) 253 (100) 301 (47) 408 (49) 413 (61) 474 (51) 617 (53) fractions were collected, combined and volatiles removed
661 (M þ H) (52). under reduced pressure. Obtained 1.52 g (75%) of b-azi-
do-d-lactose (2) as a white solid. 1H NMR (400.03 MHz,
DMSO-d6, 298 K) d ¼ 3.02 – 3.05 (m, 1H, CH), 3.28 – 3.31
2.4.3 b-Azido-d-galactose Tetraacetate
(m, 1H, CH), 3.32 – 3.38 (m, 1H, CH), 3.38 – 3.42 (m, 1H,
Flash chromatography: (CC, SiO2, ethyl acetate/pet. ether CH), 3.43 – 3.45 (m, 1H, CH), 3.46 – 3.49 (m, 1H, CH),
1 : 2, v/v). 96% yield as a colourless solid. 1H NMR 3.51 – 3.55 (m, 2H, CH2), 3.52 – 3.56 (m, 1H, CH), 3.59 –
(400.03 MHz, CDCl3, 298 K) d ¼ 1.98 (s, 3H, CH3), 2.05 (s, 3.62 (m, 1H, CH), 3.85 (d, J ¼ 7.8 Hz, 1H, CH), 4.51 (d, J ¼
3H, CH3), 2.08 (s, 3H, CH3), 2.16 (s, 3H, CH3), 4.07 (m, 8.8 Hz, 1H, CHN3). 13C{1H} NMR (100.59 MHz, DMSO,
1H, CHCH2), 4.16 (m, 1H, CHCH2), 4.58 (d, J ¼ 8.8 Hz, 298 K) d ¼ 60.80 (1C, CH2), 61.69 (1C, CH), 69.28 (1C,
1H, CH), 5.03 (m, 1H, CH), 5.15 (m, 1H, CH), 5.41 (d, J ¼ CH), 73.76 (1C, CH), 73.81 (1C, CH), 75.49 (1C, CH),
2.3 Hz, 1H, CHN3). 13C{1H} NMR (100.59 MHz, CDCl3, 76.33 (1C, CH), 77.71 (1C, CH), 77.78 (1C, CH), 79.64 (1C,
298 K) d ¼ 20.78 (4C, 4  CH3), 61.38 (1C, CH2), 66.99 (1C, CH), 90.64 (1C, CHN3), 104.13 (1C, CH). IR (neat): ~ n¼
CH), 68.18 (1C, CH3), 70.86 (1C, CH3), 72.99 (1C, CH3), 3314, 2888, 2117, 1639, 1373, 1242, 1028, 890, 783,
88.41 (1C, CHN3), 169.50 [1C, CH3C(O)O], 170.15 [1C, 539 cm1. Anal. calcd. for C12H21N3O10: C, 39.24; H, 5.76;
CH3C(O)O], 170.27 [1C, CH3C(O)O], 170.54 [1C, N, 11.44; O, 43.56; Found: C, 39.01; H, 5.84; N, 11.22; O,
CH3C(O)O]. IR (neat): ~n ¼ 2125, 1746, 1435, 1274, 1082, 43.36. MS(ESI):105 (78) 129 (63) 185 (54) 230 (100) 239
1046, 951, 902, 842, 759, 718, 678 cm1. Anal. calcd. for C26 (53) (94) 301 (94) 309 (54) 391 (M þ Na) (45).
H35N3O17: C, 47.20; H, 5.33; N, 6.35; O, 41.11; Found: C,
47.11; H, 5.43; N, 6.00; O, 40.21. MS(ESI): 105 (65) 129
2.4.6 b-Azido-d-galactose (1)
(54) 185 (50) 229 (35) 230 (100) 253 (81) 301 (75) 309 (57)
413 (M þ K) (58). Flash chromatography: (CC, SiO2, methanol/CH2Cl2 1 : 5
v/v). 94 % yield as a white solid. 1H NMR (400.03 MHz,
D2O, 298 K) d ¼ 3.52 (m, 1H, CHCH2), 3.68 (m, 1H, CH),
2.4.4 a-Azido-d-mannose Tetraacetate
3.80 (m, 2H, CH2), 3.82 (m, 1H, CH), 3.97 (d, J ¼ 3.3 Hz,
Flash chromatography: (CC, SiO2, ethyl acetate/pet. ether 1H, CH), 4.68 (d, J ¼ 8.8 Hz, 1H, CH). 13C{1H} NMR
1 : 2 v/v). 91% yield as a colourless solid. 1H NMR (100.59 MHz, D2O, 298 K) d ¼ 60.95 (1C, CH2), 68.51 (1C,
(400.03 MHz, CDCl3, 298 K) d ¼ 2.0 (s, 3H, CH3), 2.05 (s, CH), 70.32 (1C, CH), 72.65 (1C, CH), 77.21 (1C, CHCH2),
3H, CH3), 2.11 (s, 3H, CH3), 2.17 (s, 3H, CH3), 4.14 (m, 90.55 (1C, CHN3). IR (neat): ~n ¼ 3315, 2127, 1738, 1498,
1H, CHCH2), 4.18 (m, 1H, CHCH2), 4.30 (m, 1H, CH), 1371, 1306, 1271, 1137, 1101, 982, 891, 858, 775, 706 cm1.
5.16 (m, 1H, CH), 5.26 (m, 1H, CH), 5.39 (d, J ¼ 1.8 Hz, Anal. calcd. for C12H21N3O10: C, 39.24; H, 5.76; N, 11.44; O,
1H, CHN3). 13C{1H} NMR (100.59 MHz, CDCl3, 298 K) 43.56; Found: C, 39.01; H, 5.84; N, 11.22; O, 43.36. MS(E-
d ¼ 20.75 (1C, CH3), 20.82 (1C, CH3), 20.86 (1C, CH3), SI):105 (61) 129 (73) 157 (42) 185 (48) 228 (M þ Na) (100).
20.96 (1C, CH3), 62.35 (1C, CH2), 66.80 (1C, CH), 68.40
(1C, CH3), 69.31 (1C, CH3), 70.74 (1C, CH3), 87.58 (1H,
2.4.7 a-Azido-d-mannose (3)
CHN3), 169.32 [1C, CH3C(O)O], 170.01 [1C, CH3C(O)O],
170.15 [1C, CH3C(O)O], 170.24 [1C, CH3C(O)O]. IR Flash chromatography (CC, SiO2, methanol/CH2Cl2 1 : 5
(neat): ~n ¼ 2120, 1780, 1369, 1212, 1046, 957, 684 cm1. v/v). 82 % yield as a colourless oil. 1H NMR (400.03 MHz,
Anal. calcd. for C26H35N3O17: C, 47.20; H, 5.33; N, 6.35; O, D2O, 298 K) d ¼ 3.66 (m, 1H, CHCH2), 3.72 (m, 1H, CH),
41.11; Found: C, 47.11; H, 5.43; N, 6.00; O, 40.21. MS(ESI): 3.76 (m, 2H, CH2), 3.87 (m, 1H, CH), 3.95 (m, 1H, CH),
105 (48) 129 (50) 185 (44) 229 (34) 230 (100) 239 (36) 301 5.46 (d, J ¼ 1.8 Hz, 1H, CH). 13C{1H} NMR (100.59 MHz,
(50) 309 (48) 413 (M þ K) (51). D2O, 298 K) d ¼ 61.46 (1C, CH2), 67.04 (1C, CH), 70.40
(1C, CH), 70.47 (1C, CH), 75.28 (1C, CHCH2), 90.38 (1C,
CHN3). IR (neat): ~n ¼ 3313, 2111, 1238, 1062, 936, 805, 668,
2.4.5 Deprotected N-Glycosyl Azides. General Procedure:
587 cm1. Anal. calcd. for C12H21N3O10: C, 39.24; H, 5.76;
b-Azido-d-Lactose (2)
N, 11.44; O, 43.56; Found: C, 39.01; H, 5.84; N, 11.22; O,
Sodium methoxide (25% w/w solution in methanol, 43.36. MS(ESI):105 (32) 129 (67) 157 (36) 185 (45) 228
21.9 mL, 96.0 mmol) was added to a solution of b-azido-d- (M þ Na) (100).
lactose heptaacetate (3.66 g, 5.53 mmol) anhydrous metha-
nol (150 mL) and the mixture was stirred at ambient tem-
2.5 Synthesis of Neoglycopolymers. General Procedure:
perature for 3 h. Amberlite IR-120 (PLUS) ion-exchange
Synthesis of Polymer (7)
resin was added and the reaction mixture was stirred for
further 30 min. The resin was removed by filtration and A solution of polymer (5a; 0.029 g, 0.23 mmol of DclickableE
the resulting solution concentrated under reduced pres- alkyne units), azido-lactose (2; 0.102 g, 0.278 mmol,
sure. The crude product was purified by flash chromatog- 1.2 equiv. per alkyne unit), triethylamine (0.016 mL,
raphy (CC, SiO2, methanol/CH2Cl2 1 : 2, v/v). The relevant 0.12 mmol) and TBTA ligand (0.040 g, 0.093 mmol) in

QSAR Comb. Sci. 26, 2007, No. 11-12, 1220 – 1228 www.qcs.wiley-vch.de G 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1223
Full Papers Jin Geng et al.

DMSO (10 mL), was deoxygenated by bubbling nitrogen 3 Results and Discussion
for 20 min. Cu(I)Br (0.017 g, 0.046 mmol) was then added
under nitrogen and the resulting solution was stirred at
3.1 Synthesis of Glycosyl Triazole Polymers
ambient temperature for 3 days. Ion exchange resin (Smo-
pex 112) was added and the suspension gently stirred at Glycosyl azides (1 – 3; Chart 1) employed in this work
ambient temperature for further 24 h. After filtration the were prepared in two steps starting from the peracetylated
solution was added dropwise to a CH2Cl2/methanol (2 : 1 sugars. The latter were treated with trimethylsilyl azide in
v/v) mixture and the polymer was isolated by centrifuga- the presence of SnCl4 [47 – 51] to give peracetylated glyco-
tion. The polymer was dissolved in water and dialysed sylazide intermediates that were then deprotected with so-
against water (dialysis tubing, MWCO 8.0 kDa) for more dium methoxide in anhydrous methanol. O-Ethenyl man-
than 20 h. The polymer aqueous solution was then freeze- nose azide (4, Chart 1) was used for comparative studies.
dried to give the glycopolymer (7) as a white light solid.

2.6 Crystal Structure Analysis

The crystal was coated in an inert oil and cooled to T ¼
100 K (Oxford Cryojet XL). Data were recorded on an
Oxford Diffraction Gemini R Ultra equipped with a Ruby
CCD area detector. Intensities were corrected semi-empir-
ically for absorption, based on symmetry-equivalent and
repeated reflections (CrysAlis PRO). Structures were solved
by direct methods (SHELXS) and refined on F2 Chart 1. Sugar azides employed in this study.
(SHELXTL). All H atoms were constrained with a riding
model with U(H) set at 1.2 times Ueq for the parent atom.
The alkyne functional and clickable well-defined polyal-
kyne scaffolds were prepared following a general synthetic
2.6.1 Crystal Data
strategy based on copper-catalysed LRP [34]. A benzyl
Compound {[TBTACu(I)Br] and [TBTACu(II)Br] þ Br  } polymerisation initiator was used for the synthesis of the
C60H60N20Cu2Br3 colourless block, 0.20  0.20  0.16 mm, two polymers (5a) and (5b) (Scheme 1), as the resulting
Cubic, P2(1)3 (No 198), alpha ¼ beta ¼ gamma ¼ 908, a ¼ benzyl moieties at the polymer chain-ends allowed for fac-
b ¼ c ¼ 18.107(5) R, U ¼ 5937(3) R3, Z 4, D(cal) ¼ 1.598 g/ ile determination of the Mn of the polyalkyne scaffolds by
cm3, 2qmax 58.04, l ¼ 0.71073 R, 25 262 reflections mea- 1
H NMR [34]. The click reactions were performed in
sured, 4920 unique [R(int) ¼ 0.1243], R1[for 2619 reflec- DMSO and triethylamine, in the presence of a slight ex-
tions with I > 2sigma(I)] ¼ 0.0836, wR2 ¼ 0.1969, Tmin 0.577, cess of a sugar azide [(1 – 4), 1.2 equiv. per alkyne mono-
Tmax 0.639, m(MoKa) ¼ 2.795 mm1. GooF ¼ 1.015, data/re- mer unit] and a copper(I) catalyst. This solvent/base sys-
straints/parameters 4920/2/257 refined against F2 tem was chosen because under these conditions the re-
(SHELXTL). Largest difference Fourier peak and hole agents, products and the Cu(I) catalyst are fully soluble.
0.340 and  0.301 e/R3. The crystal chosen was a racemic
twin and the BASF refined to 0.42(2).
3.2 Catalysts
The asymmetric unit contains both the Cu(I) and Cu(II)
complexes. The amine nitrogen, copper and chelated bro- For the clicking of monosaccharide azides (1), (3) and (4),
mine of each complex all lie on a three-fold axis. The bro- [(PPh3)3Cu(I)Br] in DMSO/triethylamine [34] was found
mide counter ion for the Cu(II) complex also lies on a to be a more than adequate catalytic system. Interestingly,
three-fold axis. in the presence of the lactose azide (2), under these experi-
CCDC 661131 contains the supplementary crystallo-
graphic data for this paper. These data can be obtained
free of charge at www.ccdc.cam.ac.uk or from the Cam-
bridge Crystallographic Data Centre, 12, Union Road,
Cambridge CB2 1EZ, UK; Fax: (internat.) þ 44-1223/336-
033; E-mail: [email protected].

Scheme 1. Synthetic approach to glycopolymers. Reaction con-

ditions (a) Sugar azide [(1), (2), (3) or (4)], triethylamine, cop-
per catalyst, DMSO.

1224 G 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.qcs.wiley-vch.de QSAR Comb. Sci. 26, 2007, No. 11-12, 1220 – 1228
Well-Defined Poly(N-glycosyl 1,2,3-triazole) Multivalent Ligands: Design, Synthesis and Lectin Binding Studies

mental conditions resulted in only 60% of the alkyne re-

peating units being converted into the desired 1,2,3-tria-
zole rings. This can be explained in terms of larger size of
the clicked lactose disaccharide molecules which made the
still unreacted polymer alkyne moieties less accessible for
further functionalisation. The use of TBTA ligand [41] did
not seem to improve the efficiency of this reaction, which
may be at least in part ascribed to preferential competitive
copper coordination to DMSO and/or triethylamine, used
in this process as the solvent and the base, respectively.
Bathophenanthrolinedisulphonic acid disodium salt, a
ligand that has been successfully employed in conjunction
with Cu(I) salts for the click decoration of virus capsids
[52] was subsequently tested. Online 1H NMR in DMSO-
d6 at 50 8C revealed that under these conditions the reac-
tion of (2) to the polyalkyne scaffold (5b) reached a maxi-
mum in conversion within 2 h (Figure 1). The 1H NMR
Figure 1. Online NMR reaction of lactose azide (2) to polyal-
data obtained do not allow us to prove whether or not the kyne scaffold (5b) in DMSO-d6: partial 1H NMR spectra show-
attachment of (2) was in 100% yield, yet from the analysis ing the formation of 1,2,3-triazole units.
of our results we can state that a loading of at least 85%
was reached.
Although this was beyond the scope of the present
work, a preliminary attempt to elucidate some click cata-
lyst structure – reactivity relationship was made. TBTA re-
acts with Cu(I)Br to give a 1 : 1 yellow complex that pre-
cipitates in THF and methanol. TBTA is known for stabil-
ising Cu(I) complexes towards oxidation to Cu(II) species;
therefore, a first attempt to obtain crystals suitable for X-
ray diffractometry was carried out in the presence of air.
This was also meant to be a quick assessment of the stabili-
ty of the TBTACu(I)Br complex under nonoxygen-free
conditions. Addition of one equivalent of Cu(I)Br to a so-
lution of TBTA in dichloromethane afforded a solution
that readily turned green in the presence of air.* X-ray
analysis of one of the crystals obtained following evapora-
tion of the solvent revealed the presence of two different
copper species, [TBTACu(I)Br] and [TBTACu(II)Br] þ in
the unit cell (one Br  counterion every two complex mole-
cules was detected), suggesting that some decomposition
of the Cu(I) catalyst occurred under these conditions. Sur- Figure 2. Crystal structure of complexes: a) [LCuBr] and b)
prisingly, the same Cu(I)/Cu(II)-TBTA crystals were sub- [LCuBr]Br, with L ¼TBTA.
sequently obtained also under oxygen-free conditions, by
slow diffusion of THF into a DMF solution of Cu(I)Br.
One of the two complexes had a rather regular trigonal
bipyramidal structure, with the Cu atom lying slightly above the equatorial plane formed by the three triazole ni-
trogen donors (Figure 2). The other species appeared to
have a pseudo-tetrahedral structure (see Table 1, first col-
* The green colour could also be due in some extent to reac- umn), with the copper centre coordinated to the three tria-
tion of TBTACu(I)Br with the solvent employed for this zole nitrogen atoms and to a bromine atom. In this case,
quick crystallisation attempt. We found that the latter com- the observed distance between the copper centre and basal
plex can also act as an LRP catalyst, although a poor control sp3 nitrogen was 2.5 R, a value that seemed to suggest very
over the polymer (PMMA) molecular weight distribution was weak or no interaction between the two atoms. Identifying
obtained. Cu(I) complexes bearing tris(2-pyridylmethyl)amine
unequivocally which one of the two observed species was
(TPA, often referred to as TMPA) have also been reported to
react with dichloromethane to give Cu(II) and radical deriva- Cu(I) and which was the Cu(II) derivative was at this stage
tives. R. R. Jacobson, Z. Tyeklar, K. D. Karlin, Inorg. Chim. rather difficult. From an analysis of the available litera-
Acta 1991, 181, 11 – 118. ture, we found that Cu(II) complexes bearing ligands with

QSAR Comb. Sci. 26, 2007, No. 11-12, 1220 – 1228 www.qcs.wiley-vch.de G 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1225
Full Papers Jin Geng et al.

and very plausible explanation for the poor catalytic per-

formance of TBTA/Cu(I)Br under the experimental con-
ditions employed in this work is that TBTA simply does
not coordinate efficiently enough to the copper centre in
the presence of the rather aggressive DMSO/triethylamine
solvent system. Important differences between the phe-
nanthroline-based ligand and subsequently employed
TBTA include superior ability of the former to form cop-
per complexes and that is phenanthrolines are bidentate li-
gands and one may speculate that the relative Cu(I) com-
plex may in principle temporarily lose one of the two coor-
dinated ligands during the catalytic cycle, further reducing
the size of the catalyst (assuming that the phenomena de-
scribed in this work are influenced by steric effect and not
just by electronic ones).

3.3 Lectin Conjugation Studies

The potential of these glycopolymers as multivalent li-
gands able to be recognised by, and bind to, appropriate
lectins was investigated. For the lactose and galactose
Figure 3. Affinity HPLC analysis of glycopolymers containing polymers, the lectin RCA I was chosen as it selectively in-
lactose (7; top), and galactose 6; bottom) at different concentra- teracts with b-d-galactose units. The sugar – lectin interac-
tions of galactose in the mobile phase. tion was studied by affinity chromatography analysed by
HPLC, injecting solutions of the glycopolymers (6) and (7)
into a column packed with immobilised RCA I. The glyco-
structure analogous to TBTA, such as tris(2-pyridylme- polymers are retained by the column as the sugar moieties
thyl)amine (TPA), have structures very similar to the are recognised by the lectin. A mobile phase containing
trigonal bipyramidal found in these crystals [53, 54]. Bas- galactose releases the polymer from the column which is
ing on these findings, and in addition to the fact that penta detected by UV absorbance. By increasing the concentra-
coordinated Cu(I) species are rare [55], we tentatively tion of galactose in the mobile phase, the amount of eluted
identify the trigonal bipyramidal structure as the Cu(II) glycopolymer increased (Figure 3). These results are anal-
derivative, and the pseudo-tetrahedral structure as the ogous to those we have previously reported for the lac-
Cu(I) complex. The latter represents a solid-state structure tose- and galactose-bearing glycopolymers based on O-
of the click pre-catalyst in the absence of both alkyne and propenyl sugars. This indicates that the lactose- and galac-
azide substrates and mechanistic considerations based tose-containing glycopolymers (6) and (7), respectively,
solely on these data should be drawn with utmost caution. can function as multivalent ligands for the recognition by
Yet, from an analysis of the Cu(I) structure a few observa- the galactose-selective lectin RCA I also when the 1,2,3-
tions can still be made. The Cu-Br unit lays on one-half of triazole moiety is bound directly to the anomeric sugar
the molecule, while the tripodal TBTA ligand essentially centre.
occupies the other half of the structure. This seems to indi- The rate of the ligand-lectin binding is a critical factor in
cate that the click reactants, first the 1-alkyne and subse- biological systems, where timescales can range from sec-
quently the organic azide, according to the reported mech- onds to hours. The rate of the binding of a lectin to the
anism [36, 56, 57], can have relatively easy access to the mannose-containing glycopolymer, (8), was investigated
metal centre, which may be in part responsible for the by turbidimetry. Concanavalin A (Con A) was used for
good efficiency of TBTA copper complexes as a catalyst these binding studies as it is a mannose-selective, well-
for the Huisgen [2 þ 3] cycloaddition. Our data could indi- studied lectin. By measuring absorbance changes at
cate that the TBTA ligand may provide significant equato- 420 nm over time in a solution of Con A and glycopolymer
rial steric hindrance which in part may explain the rela- (8) in HEPES buffer at pH 7.4, the rate of the ligand-lectin
tively low efficiency observed in the clicking of large lac- cluster formation could be obtained [34, 46]. As can be
tose azide (2) to the polyalkyne (5b). This may be due to seen in Figure 4, the absorbance reaches a plateau after
the fact that a relatively large complex may find difficult the initial increase, suggesting that nearly all the polymer
to first reach already sterically crowded unreacted alkyne was able to quickly interact with the lectin to form stable,
moieties pending from the polymer backbone, then allow precipitating clusters [46]. The initial aggregation rate ob-
the coordination of the large azide (2) to the metal centre tained for glycopolymer (8) is comparable (80%) to that
necessary for the click process to happen. An alternative of the O-ethenylmannose-functionalised glycopolymer,

1226 G 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.qcs.wiley-vch.de QSAR Comb. Sci. 26, 2007, No. 11-12, 1220 – 1228
Well-Defined Poly(N-glycosyl 1,2,3-triazole) Multivalent Ligands: Design, Synthesis and Lectin Binding Studies

lectin revealed that N-glycosyl 1,2,3-triazole polymers

were able to recognise this mannose binding protein, and
that its rate of clustering was comparable with that of an
analogous ligand prepared from the same polyalkyne pre-
cursor (and therefore with the same macromolecular fea-
tures, i.e. DP, PDI and macromolecular architecture), and
featuring an O-ethenyl linker between the sugar binding
units and the polymer backbone.
The synthetic strategy presented in this work appears to
be general and we believe it could be applied for the syn-
thesis of a number of multivalent ligands presenting multi-
ple copies of different carbohydrate binding units. The
synthesis of ligands with different relative content of two
or more sugar epitopes via coclicking approach [34] ap-
pears to be one of the possible evolution of the present
Figure 4. Turbidimetry assay results for glycopolymer (8) in work and will be the subject of future research.
the presence of mannose-binding lectin Con A.

prepared from O-ethenylmannose azide (4) and polyal-

Acknowledgements
kyne scaffold (5a). These results suggest that for the multi-
valent ligand/lectin receptor studied in this work, the pres-
The authors would like to thank Neil Brooks and Oxford
ence of a 1,2,3-triazole directly attached to the anomeric
Diffraction at Abingdon (UK) for recording the data for
sugar centre has no significant effect on the rate of the li-
the crystal structure of {[TBTACu(I)Br] and
gand – lectin interaction.
[TBTACu(II)Br] þ Br  }, the University of Warwick (J. G.
and G. C.) and Warwick Effect Polymers Ltd. and EPSRC
(WEP, C. T. S) for funding.
4 Conclusions

The design and synthesis of poly(N-glycosyl 1,2,3-triazole)

polymers via post-functionalisation of well-defined polyal- References
kyne scaffolds with appropriate sugar azide probes is re-
ported. Efficient post-functionalisation was obtained by [1] R. A. Dwek, Chem. Rev. 1996, 96, 683 – 720.
Cu(I)-catalysed Huisgen [2 þ 3] cycloaddition, using a- [2] A. Dove, Nat. Biotech. 2001, 19, 913 – 917.
[3] L. L. Kiessling, C. W. Cairo, Nat. Biotech. 2002, 20, 234 –
mannoside, b-galactoside and b-lactoside derivatives in 235.
which the azide unit was bound directly to the sugar [4] F. Fazio, M. C. Bryan, O. Blixt, J. C. Paulson, C.-H. Wong, J.
anomeric carbon. Three different catalytic systems, namely Am. Chem. Soc. 2002, 124, 14397 – 14402.
[(PPh3)3Cu(I)Br], TBTA/Cu(I)Br and bathophenanthroli- [5] D. A. Tirrell, Nature 2004, 430, 837.
nedisulphonic acid disodium salt/Cu(I)Br were employed, [6] M. Ambrosi, N. R. Cameron, B. G. Davis, Org. Biomol.
with the latter being the more efficient for the clicking of Chem. 2005, 3, 1593 – 1608.
[7] M. Mammen, S.-K. Chio, G. M. Whitesides, Angew. Chem.
large disaccharide lactose azide.
Int. Ed. 1998, 37, 2755 – 2794.
The ability of the N-glycosyl 1,2,3-triazole polymers to [8] Y. C. Lee, R. T. Lee, Acc. Chem. Res. 1995, 28, 321 – 327.
recognise specific protein receptors was then evaluated. [9] S. M. Dimick, S. C. Powell, S. A. McMahon, D. N. Moothoo,
Affinity chromatography using immobilised RCA I as the J. H. Naismith, E. J. Toone, J. Am. Chem. Soc. 1999, 121,
stationary phase showed that both the lactose- and galac- 10286 – 10296.
tose-containing glycopolymers prepared in this work were [10] J. J. Lundquist, E. J. Toone, Chem. Rev. 2002, 102, 555 – 578.
able to interact strongly with their immobilised protein [11] F. D. Tropper, A. Romanowska, R. Roy, Methods Enzymol.
1994, 242, 257 – 271.
counterpart and the use of a competitive monovalent li- [12] M. Okada, Prog. Polym. Sci. 2001, 26, 67 – 104.
gand, d-galactose in the mobile phase was necessary in or- [13] L. L. Kiessling, L. E. Strong, J. E. Gestwicki, Ann. Rep.
der to have elution of the neoglycopolymers. Med. Chem. 2000, 35, 321 – 330.
The versatility of the approach followed in this study al- [14] V. Ladmiral, E. Melia, D. M. Haddleton, Eur. Polym. J.
lowed not only to compare structurally identical ligands 2004, 40, 431 – 449.
that only differ for the nature of the sugar epitopes [34], [15] R. Roy, D. Page, S. F. Perez, V. V. Bencomo, Glycoconj. J.
1998, 15, 251 – 263.
but also to prepare glycopolymers that only differ from
[16] Q. Wang, J. S. Dordick, R. J. Linhardt, Chem. Mater. 2002,
each other for the length and the nature of the linker con- 14, 3232 – 3244.
necting the carbohydrate units to the macromolecular [17] L. L. Kiessling, J. E. Gestwicki, L. E. Strong, Angew. Chem.
backbone. Turbidrimetric assay in the presence of Con A Int. Ed. 2006, 45, 2348 – 2368.

QSAR Comb. Sci. 26, 2007, No. 11-12, 1220 – 1228 www.qcs.wiley-vch.de G 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1227
Full Papers Jin Geng et al.

[18] S. G. Spain, M. I. Gibson, N. R. Cameron, J. Polym. Sci., [38] S. Dedola, S. A. Nepogodiev, R. A. Field, Org. Biomol.
Part A: Polym. Chem. 2007, 45, 2059 – 2072. Chem. 2007, 5, 1006 – 1017.
[19] M. Kamigaito, T. Ando, M. Sawamoto, Chem. Rev. 2001, [39] R. N. Keller, H. D. Wycoff, Inorg. Synth. 1946, 1 – 4.
101, 3689 – 3745. [40] R. Gujadhur, D. Venkataraman, J. T. Kintigh, Tetrahedron
[20] K. Matyjaszewski, J. Xia, Chem. Rev. 2001, 101, 2921 – 2990. Lett. 2001, 42, 4791 – 4793.
[21] K. Ohno, Y. Tsujii, T. Fukuda, J. Polym. Sci., Part A: Polym. [41] R. T. Chan, R. Hilgraf, K. B. Sharpless, V. V. Fokin, Org.
Chem. 1998, 36, 2473 – 2481. Lett. 2004, 6, 2853 – 2855.
[22] M. Ejaz, K. Ohno, Y. Tsujii, T. Fukuda, Macromolecules [42] T. K. Jones, S. E. Denmark, Org. Synth. 1986, 64, 182 – 188.
2000, 33, 2870 – 2874. [43] G. Braun, Org. Synth. 1937, 17, 36 – 39.
[23] R. Narain, S. P. Armes, Biomacromolecules 2003, 4, 1746 – [44] M. Dowlut, D. G. Hall, O. Hindsgaul, J. Org. Chem. 2005,
1758. 70, 9809 – 9813.
[24] R. Narain, S. P. Armes, Macromolecules 2003, 36, 4675 – [45] W. Hayes, H. M. I. Osborn, S. D. Osborne, R. A. Rastall, B.
4678. Romagnoli, Tetrahedron 2003, 59, 7983 – 7996.
[25] S. Muthukrishnan, G. Jutz, X. Andre, H. Mori, A. H. E. Mu- [46] C. W. Cairo, J. E. Gestwicki, M. Kanai, L. L. Kiessling, J.
eller, Macromolecules 2005, 38, 9 – 18. Am. Chem. Soc. 2002, 124, 1615 – 1619.
[26] S. Muthukrishnan, M. Zhang, M. Burkhardt, M. Drechsler, [47] A. Stimac, J. Kobe, Carbohydr. Res. 2000, 324, 149 – 160.
H. Mori, A. H. E. Mueller, Macromolecules 2005, 38, 7926 – [48] V. Vicente, J. Martin, J. Jimenez-Barbero, J. L. Chiara, C.
7934. Vicent, Chem. Euro. J. 2004, 10, 4240 – 4251.
[27] V. Ladmiral, L. Monaghan, G. Mantovani, D. M. Haddleton, [49] S. Wu, C. Campbell, Y. Koda, J. T. Blanchfield, I. Toth,
Polymer 2005, 46, 8536 – 8545. Med. Chem. 2006, 2, 203 – 211.
[28] S. Sen Gupta, K. S. Raja, E. Kaltgrad, E. Strable, M. G. [50] A. Bianchi, A. Bernardi, J. Org. Chem. 2006, 71, 4565 –
Finn, Chem. Commun. 2005, 4315 – 4317. 4577.
[29] S. Muthukrishnan, D. P. Erhard, H. Mori, A. H. E. Mueller, [51] I.-H. Kim, K. Nishi, H.-J. Tsai, T. Bradford, Y. Koda, T. Wa-
Macromolecules 2006, 39, 2743 – 2750. tanabe, C. Morisseau, J. Blanchfield, I. Toth, B. D. Ham-
[30] V. Vazquez-Dorbatt, H. D. Maynard, Biomacromolecules mock, Bioorg. Med. Chem. 2007, 15, 312 – 323.
2006, 7, 2297 – 2302. [52] S. Sen Gupta, J. Kuzelka, P. Singh, W. G. Lewis, M. Man-
[31] C. Gao, S. Muthukrishnan, W. Li, J. Yuan, Y. Xu, A. H. E. chester, M. G. Finn, Bioconjug. Chem. 2005, 16, 1572 – 1579.
Mueller, Macromolecules 2007, 40, 1803 – 1815. [53] K. D. Karlin, J. C. Hayes, S. Juen, J. P. Hutchinson, J. Zubie-
[32] Z.-Q. Yang, E. B. Puffer, J. K. Pontrello, L. L. Kiessling, ta, Inorg. Chem. 1982, 21, 4106 – 4108.
Carbohydr. Res. 2002, 337, 1605 – 1613. [54] C. S. Allen, C.-L. Chuang, M. Cornebise, J. W. Canary, In-
[33] J. K. Pontrello, M. J. Allen, E. S. Underbakke, L. L. Kies- org. Chim. Acta 1995, 239, 29 – 37.
sling, J. Am. Chem. Soc. 2005, 127, 14536 – 14537. [55] J. A. Goodwin, D. M. Stanbury, L. J. Wilson, C. W. Eigen-
[34] V. Ladmiral, G. Mantovani, G. J. Clarkson, S. Cauet, J. L. Ir- brot, W. R. Scheidt, J. Am. Chem. Soc. 1987, 109, 2979 –
win, D. M. Haddleton, J. Am. Chem. Soc. 2006, 128, 4823 – 2991.
4830. [56] F. Himo, T. Lovell, R. Hilgraf, V. V. Rostovtsev, L. Noodle-
[35] M. Kanai, K. H. Mortell, L. L. Kiessling, J. Am. Chem. Soc. man, K. B. Sharpless, V. V. Fokin, J. Am. Chem. Soc. 2005,
1997, 119, 9931 – 9932. 127, 210 – 216.
[36] V. Rostovtsev, G. L. Green, V. V. Fokin, K. B. Sharpless, An- [57] C. Nolte, P. Mayer, B. F. Straub, Angew. Chem. Int. Ed.
gew. Chem. Int. Ed. 2002, 41, 2596 – 2599. 2007, 46, 2101 – 2103.
[37] C. W. Tornoe, C. Christensen, M. Meldal, J. Org. Chem.
2002, 67, 3057 – 3064.

1228 G 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.qcs.wiley-vch.de QSAR Comb. Sci. 26, 2007, No. 11-12, 1220 – 1228