Biohybrid glycopolymer capable of ionotropic gelation

Article abstract of DOI:10.1021/bm300925j

Ionotropic gelation is particularly appealing for the formation of hydrogels because it takes place under mild conditions, is not thermoreversible, and does not involve toxic chemicals. A well-known example is the gelation of alginate in the presence of calcium ions, which is at the base of numerous applications involving this polymer. In this study, alginate-derived oligosaccharides were converted into acrylamide- and methacrylamide-type macromonomers in two steps without resorting to protective group chemistry. They were then copolymerized with 2-hydroxyethylmethacrylamide in aqueous solution to yield high molar mass biohybrid glycopolymers containing between 25 and 52% by mass of oligosaccharide graft chains. A comparative kinetic study showed that both acrylamide- and methacrylamide-type macromonomers reacted since the early stages of the copolymerization, but that the mole fraction in the polymer was smaller than in the feed up to 50-60% conversion and increased markedly afterward. This effect was slighter for the methacrylamide-type macromonomer though. Copolymers carrying oligosaccharide chains with 16-20 repeating units were synthesized and used for a gelation experiment: When dialyzed against CaCl₂ 0.5 mol L^-1, the polymer carrying (1→4)-α-l-guluronan residues led to a soft isotropic self-standing transparent hydrogel, while the polymer carrying (1→4)-β-d-mannuronan residues gave a loose opaque gel. This study demonstrates that alginate-extracted oligosaccharides and aqueous radical polymerization can be combined for the flexible design of biohybrid glycopolymers capable of ionotropic gelation under very mild conditions.

Full text of DOI:10.1021/bm300925j

Subscriber access provided by University of Guelph Library
Article
Biohybrid glycopolymer capable of ionotropic gelation
Ali Ghadban, Luca Albertin, Marguerite Rinaudo, and Alain Heyraud
Biomacromolecules, Just Accepted Manuscript • DOI: 10.1021/bm300925j • Publication Date (Web): 28 Aug 2012
Downloaded from http://pubs.acs.org on September 3, 2012
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Biomacromolecules is published by the American Chemical Society. 1155 Sixteenth
Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 40
Biomacromolecules
1
2
3
4
5
6
Biohybrid glycopolymer capable of ionotropic gelation
7
8
9
Ali Ghadban 1, Luca Albertin 1,*, Marguerite Rinaudo 2, Alain Heyraud 1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1 Centre de Recherches sur les Macromolécules Végétales (CERMAVꢀCNRS)†, BP53, 38041 Grenoble,
France. 2 European Synchrotron Radiation Facility, BP 220, 6 Rue Jules Horowitz, 38043, Grenoble,
France.
luca.albertin@cermav.cnrs.fr
RECEIVED DATE (to be automatically inserted after your manuscript is accepted if required
according to the journal that you are submitting your paper to)
ABSTRACT
Ionotropic gelation is particularly appealing for the formation of hydrogels since it takes place under
mild conditions, is not thermoreversible and does not involve toxic chemicals. A wellꢀknown example is
the gelation of alginate in the presence of calcium ions, which is at the base of numerous applications
involving this polymer. In this study, alginateꢀderived oligosaccharides were converted into acrylamideꢀ
and methacrylamideꢀtype macromonomers in two steps without resorting to protective group chemistry.
They were then copolymerized with 2ꢀhydroxyethylmethacrylamide in aqueous solution to yield high
molar mass biohybrid glycopolymers containing between 25% and 52% by mass of oligosaccharide graft
* Corresponding author, luca.albertin@cermav.cnrs.fr. Tel. +33 (0)4 76 03 76 60. Fax +33 (0)4 76 54 72
03.
†Affiliated with Université Joseph Fourier, and member of the Institut de Chimie Moléculaire de
1
Grenoble.
ACS Paragon Plus Environment

Biomacromolecules
Page 2 of 40
chains. A comparative kinetic study showed that both acrylamideꢀ and methacrylamideꢀtype
macromonomers reacted since the early stages of the copolymerization, but that the mole fraction in the
polymer was smaller than in the feed up to 50ꢀ60% conversion and increased markedly afterwards. This
effect was slighter for the methacrylamideꢀtype macromonomer though. Copolymers carrying
oligosaccharide chains with 16ꢀ20 repeating units were synthesized and used for a gelation experiment:
When dialyzed against CaCl2 0.5 mol Lꢀ1, the polymer carrying (1→4)ꢀαꢀLꢀguluronan residues led to a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
soft isotropic selfꢀstanding transparent hydrogel while the polymer carrying (1→4)ꢀβꢀDꢀmannuronan
residues gave a loose opaque gel. This study demonstrates that alginateꢀextracted oligosaccharides and
aqueous radical polymerization can be combined for the flexible design of biohybrid glycopolymers
capable of ionotropic gelation under very mild conditions.
INTRODUCTION
Hydrogels are hydrophilic polymer networks swollen with water. Thanks to their high water content
(usually a multiple of their dry mass) and soft rubbery consistence they closely resemble soft living
1ꢀ10
tissues and are extensively studied for pharmaceutical and medical applications.
The attractive
features of hydrogels include their low interfacial tension (which results in a minimal tendency to adsorb
proteins from body fluids and in good biocompatibility), the fact that an aqueous environment can
protect cells and fragile drugs incorporated into the gel, and the good transport of nutrients to cells and
of products from cells throughout the gel pores. 9
11
Among the various mechanisms leading the formation of hydrogels, the complexation of ions with
the creation of extended junction zones (i.e. ionotropic gelation) is particularly appealing 12 since it takes
place under mild conditions, is not thermoreversible (as long as an excess of ions is present and the
junction zone is sufficiently long) 13, 14 and does not involve toxic chemicals. A wellꢀknown example is
the gelation of alginate in the presence of calcium ions, which is at the base of numerous applications
involving this polymer. Alginate is a heteropolysaccharide extracted from brown algae such as
Laminaria hyperborea and Lessonia. At a molecular level it is a linear unbranched copolymer of
2
ACS Paragon Plus Environment

Page 3 of 40
Biomacromolecules
(1→4)ꢀlinked βꢀDꢀmannuronic acid (M) and αꢀLꢀguluronic acid (G) residues distributed in long
1
2
3
4
5
6
7
8
9
homopolymeric blocks (e.g. MMMM or GGGG) and in shorter alternating copolymer blocks (e.g.
MGMG). 15, 16 In aqueous solution, alginate forms transparent hydrogels upon addition of some divalent
cations (e.g. Ca2+, Ba2+, Sr2+ and Zn2+) mostly due to complexation of said ions by guluronan on adjacent
polymer chains and to the formation of junction zones. 17 Depending on the author,18, 19 a minimum of 8
to 18 repeating units in a guluronan block is reported to be necessary to the formation of a stable
complex with calcium ions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A major difficulty in the use of alginate for the preparation of hydrogels with wellꢀdefined physicoꢀ
chemical properties is the intrinsic variability of the molar mass and composition of the polysaccharide,
which depend on the source species, the conditions of growth, the period of harvest and the extraction
15
process.
Besides, the type of chemical functionality present on any given polysaccharide is
intrinsically limited (only hydroxyl, carboxyl and hemiacetal groups in the case of alginate) and any
other residue (e.g. an adhesion peptide or a biodegradable linker) has to be introduced via polymerꢀ
modification reactions. The latter are often cumbersome, 20ꢀ22 rarely efficient 23, 24 and generally limited
in scope. 25, 26
An attractive feature of alginate is the possibility to extract its constituting blocks by controlled
27ꢀ29
hydrolysis and selective precipitation.
Since at pH ≅ 2ꢀ3 the rate of hydrolysis of GꢀM glycosidic
linkages is 3ꢀ4 times that of MꢀM, GꢀG or MꢀG linkages, alternating blocks are preferentially degraded
and fairly pure homopolymeric sequences are recovered (> 85% in G or M). 30ꢀ32 In this way, (1→4)ꢀαꢀ
Lꢀguluronan and (1→4)ꢀβꢀDꢀmannuronan samples with Xn ≅ 20ꢀ30 and low mass dispersity can be
obtained in good yield on a multiꢀgram scale (Xn is the number average degree of polymerization).
In our laboratory, we are investigating the use of alginateꢀderived oligosaccharides as building blocks
for the synthesis of biohybrid polymers with original gelation properties. The rationale behind this work
is that (1→4)ꢀαꢀLꢀguluronan residues with Xn ≥ 18 are ionotropic gelators that can be incorporated into
a water soluble synthetic polymer to bestow it with original gelation properties. As a result, the
3
ACS Paragon Plus Environment

Biomacromolecules
Page 4 of 40
versatility and precision of modern polymer chemistry could be combined with the gelation properties of
guluronan to afford well defined glycopolymers (in terms of composition, molar mass, dispersity and
functionality) capable of gelling under mild conditions.
1
2
3
4
5
6
7
8
In this communication we describe the synthesis of (meth)acrylamideꢀfunctionalized (1→4)ꢀβꢀDꢀ
9
mannuronan and (1→4)ꢀαꢀLꢀguluronan macromonomers (Scheme 1), their copolymerization with 2ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
hydroxyethylmethacrylamide (HEMAm) in water and the gelation properties of two of the resulting graft
copolymers.
Scheme 1. Synthesis of (1→4)ꢀβꢀDꢀmannuronanꢀ and (1→4)ꢀαꢀLꢀguluronanꢀfunctionalized
macromonomers.
Conditions: i) H2O, 30 °C, 5ꢀ18 days. ii) H2O / CH3OH (or DMSO) ≥ 9:1, 0 °C→Tamb, 6ꢀ7 ½ hours.
EXPERIMENTAL
Materials
The following chemicals were reagent grade and were used as received: 4,4´ꢀ Azobis(cyanopentanoic
acid) (ACPA, 98%, Aldrich), CaCl
NaHCO (≥ 99%, SDS), DMSOꢀd6 (99.8%, Eurisotop), ethanol amine (≥ 99%, Fluka), 2,6ꢀdiꢀtertꢀbutylꢀ
4ꢀmethyl phenol (BHT, ≥ 99%, Fluka), methyl αꢀ ꢀglucoside (≥ 99%, Fluka),
2
(97%, Prolabo), D
2
O (99.8%, Eurisotop), HCl (37%, Carlo Erba),
3
D
4
ACS Paragon Plus Environment

Page 5 of 40
Biomacromolecules
ethylenediaminetetraacetic acid tetrasodium salt EDTA (99%, Acros), methacryloyl chloride (≥97.0%
Fluka), NaCl (≥ 99%, Aldrich), NH4OAc (98 %, Aldrich), NaBH3CN (95 %, Aldrich), NaNO (≥ 99%,
Aldrich), NaN (≥ 99%, Merck), sodium hydroxide solutions (pure, Acros). Deionized water was
produced in house with a MilliQ apparatus (Millipore) and used for all experiments. Cellulose nitrate
1
2
3
4
5
6
7
8
3
3
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
membranes were supplied by Sartorius. (1→4)ꢀβꢀ
D
ꢀMannuronan and (1→4)ꢀαꢀLꢀguluronan oligomers
1
were obtained from Elicityl SA (Crolles, France) and characterized by H NMR and SECꢀMALLS for
their molar mass and, following total acid hydrolysis, by High Performance Anion Exchange
Chromatography (HPAEC) for composition.
General methods
Accurate volumes (10ꢀ1000 ꢁL) were measured with calibrated automatic pipettes (Eppendorf
Research). Accurate pH and conductivity values were measured with a pHꢀmeter (Cyberscan PC510);
alternatively, a special pH indicator paper was used (MachereyꢀNagel, ± 0.5 pH units). The density of
solutions and liquids was measured with precision hydrometers (Alla France). Dialysis purifications
were performed against deꢀionized water (100ꢀ150 times the volume of the sample) at room temperature
over a 48 hours period with SlideꢀAꢀLyzer Dialysis Cassettes (Pierce Biotechnology). During the
process the water was changed thrice, typically after 2, 16, and 26 hours. Diafiltrations were carried out
using ultrafiltration cells equipped with a cellulose acetate membrane (∅ 63.5 mm, Millipore) and
connected to an auxiliary reservoir filled with deꢀionized water (p = 2ꢀ3 bars; stirring rate ~300 rpm).
Purifications were stopped once the conductivity of the eluate had fallen below 5 ꢀS cmꢀ1. TLC analyses
were performed on aluminiumꢀbacked silica gel plates (60 Å, 15 ꢀm, SDS); following solvent
evaporation, the developed plates were exposed to a UV lamp (λ=254 nm) for spots detection. Flash
chromatography was carried with a glass column using silica gel from Merck (60 Å, 40ꢀ60 ꢂm).
Synthesis of N-(2-hydroxyethyl)methacrylamide (HEMAm)
In a round bottom flask, ethanol amine (10 mL, 0.160 mol) was dissolved in MeOH (100 mL) and
Et3N (21 mL, 0.150 mol) was added. The mixture was cooled in an ice bath and methacryloyl chloride
5
ACS Paragon Plus Environment

Biomacromolecules
Page 6 of 40
(14.4 mL, 0.150 mol) was introduced drop wise under stirring using a gas tight syringe. After 2 hours
the ice bath was removed and the mixture was left reacting for a further 2 hours at ambient temperature.
The reaction was stopped by adding HCl 2 N (~8 mL), the volatiles were eliminated at reduced pressure,
and 10 g of the resulting solid were reꢀsolubilized in EtOH and adsorbed on silica gel (70 g). The
resulting white powder was added at the top a column (∅: 7cm) preꢀpacked with silica gel (h ≅ 24 cm)
and the product was eluted using a gradient of Petroleum Ether/EtOAc/EtOH (from 6:3:1 to 6:2:2). The
fractions containing the product were pooled, stabilized with BHT, concentrated at the rotary evaporator
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
and further dried under mechanical vacuum (p ≈ 10ꢀ1 mbar) overnight. Viscous oil, Rf 0.45 (Petroleum
1
Ether/EtOAc/EtOH 6:2:2). HꢀNMR (400 MHz, D2O, 298 K)
δ (ppm): 1.93 (m, H6, 3H), 3.40 (t, H4,
2H), 3.68 (t, H5, 2H), 5.45 (m, H3b, 1H), 5.71 (m, H3a, 1H). 13CꢀNMR (100 MHz, D2O, 10 °C)
δ
(ppm): 20.41 (C6), 44.34 (C4), 62.60 (C5), 123.95 (C3), 141.61 (C2), 174.83 (C1). ESIꢀMS: m/z
calculated 130.09, found: 130.0 [M.H+]. Note: In the absence of a radical inhibitor the monomer
homopolymerized during drying under mechanical vacuum.
Synthesis of macromonomers
In a typical experiment, (1→4)ꢀβꢀDꢀmannuronan (ManA10, Xn = 10, 98 %, 4.50 g, 2.79 mmol) was
weighed in a 200 mL Erlenmeyer flask, solubilized in H2O (50 mL) and mixed with an aqueous solution
of ammonium acetate (9.08g, 115 mmol in 35 mL H2O) and sodium cyanoborohydride (3.21 g, 48.0
mmol in 15 mL H2O). The final pH was checked with pH paper (pH ≅ 7), a magnetic bar was added and
the flask was sealed with a rubber septum and immersed in a water bath preheated at 30 ºC. After
stirring at 200 rpm for 7 days, the mixture was transferred to a centrifugation tube and the
oligosaccharide precipitated by the addition of ethanol under vigorous stirring (final proportion 80 %
v/v). After centrifugation (10 000 rpm for 10 min), the supernatant was discarded and the precipitate
was reꢀdissolved in H2O (~80 mL) and diafiltered (500 Da cut off membrane). The desalted solution was
transferred to a round bottom flask, frozen in liquid nitrogen and freeze dried overnight to isolate the 1ꢀ
aminoꢀ1ꢀdeoxyalditol (ManA10-NH2). Conversion of starting carbohydrate > 90 %. Functionalization
6
ACS Paragon Plus Environment

Page 7 of 40
Biomacromolecules
rate: 42 %. 1HꢀNMR (400 MHz, D2O, 323 K, see Scheme 2 for position numbering) δ (ppm): 3.00 (dd,
H1a, 1H, J1a,1b 12.9 Hz, J1a,2 9.4 Hz), 3.47 (dd, H1b, 1H, J1b,1a 13.1 Hz, J1b,2 3.0 Hz), 3.61ꢀ4.30 (H2ꢀH5,
H2´ꢀH5´), 4.62ꢀ4.81 (H1´), 5.02 (H1´, residual G unit). Mn (SEC) 1900 Da with dn/dc = 0.165, Ð 1.08.
Part of the ManA10ꢀNH2 (1.50 g, 3.57 × 10ꢀ4 mol) was dissolved in 34 mL of a 9:1 mixture of sodium
carbonate – sodium bicarbonate buffer solution (0.6 mol Lꢀ1, pH 9.5) and MeOH. The resulting solution
was cooled on an ice bath for 10 minutes and methacryloyl chloride (493 ꢁL, 50.9 × 10ꢀ4 mol) was
added dropꢀwise under vigorous stirring (250ꢀ300 rpm). From time to time the pH was adjusted to ~9.5
with solid Na2CO3. After 2 hours on ice and 4.5 hours at Tamb the reaction was stopped by precipitating
the oligosaccharide with ethanol (final proportion 80 % v/v). The mixture was centrifuged (10,000 rpm,
15 °C, 10 min), the precipitate was reꢀsolubilized in water, diafiltered (500 Da cut off membrane) and
freeze dried (ManA10-MAm). Conversion of starting 1ꢀaminoꢀ1ꢀdeoxyalditol: 100 %. Functionalization
rate: 42 %. MS calcd for C40H56NO37 (X=6), 1142.26; found 1142.3 ([MꢀH]ꢀ); calcd for C46H64NO43
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(X=7) 1318.29; found 1318.4 ([MꢀH]ꢀ). 1H NMR (400 MHz, D2O, 323 K)
δ (ppm): 1.94 (m, H10, 3H),
3.38 (dd, 1H, H1a, J1a,1b 14.0 Hz, J1a,2 7.6 Hz), 3.59ꢀ4.34 (H2ꢀH5, H2´ꢀH5), 4.64ꢀ4.80 (H1´), 5.00 (H1´,
residual G unit), 5.44 (m, 1H, H7 cis to C10), 5.70 (m, 1H, H7 trans to C10).
7
ACS Paragon Plus Environment

Biomacromolecules
Page 8 of 40
Scheme 2. Position numbering used for NMR assignment.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Homopolymerization of 2-hydroxyethyl methacrylamide (HEMAm)
(Run no. 1, Table 1) 4,4´Azobis(cyanopentanoic acid) (3.14 × 10ꢀ2 g, 1.12 × 10ꢀ4 mol) was dissolved
in DMSOꢀd6 (1 mL), cooled to ≅ 8 °C and diluted with an equal volume of D2O. HEMAm (0.350 g,
2.71 × 10ꢀ3 mol) was dissolved in D2O (2.4 mL) and filtered through a syringe filter (0.22 ꢁm, cellulose
acetate) to remove the suspended inhibitor (BHT). Part of the latter solution (2.0 mL, 1.13 mol Lꢀ1), was
mixed with a calculated amount of ACPA solution (200 ꢁL, 5.60 × 10ꢀ2 mol Lꢀ1, 1.12 × 10ꢀ5 mol) and
transferred to a Schlenk tube. The tube was then sealed with a rubber septum, degassed by 3 freezeꢀ
evacuateꢀthaw cycles and transferred to a water bath preheated at 60 °C. After 3.5 hours the
polymerization was stopped by plunging the tube in icy water, a sample was withdrawn for 1H NMR and
SEC analysis and the remaining polymer was precipitated twice in an excess of acetone. The fiberꢀlike
precipitate was dried under mechanical vacuum (10ꢀ15 torr) at room temperature (~21°C) for 65 hours.
1HꢀNMR (400 MHz, D2O, 298 K)
δ (ppm): 0.97 and 1.12 (3H, CH3), 1.75 and 1.86 (2H, CH2
backbone), 3.28 (2H, CH2ꢀNH), 3.66 (2H, CH2ꢀOH).
8
ACS Paragon Plus Environment

Page 9 of 40
Biomacromolecules
Table 1. Summary of polymerization experiments.
1
2
3
4
5
6
run HEMAm macromonomer
t (min) p (%)
dn/dc Mn (Da)
Ð
f
F
Fm
[η]
no.
(mol L-1)
(%) (%) (%)
(mL g-1)
7
8
9
type
c (mmol L-1)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
1.03
0.52
0.58
ꢀ
ꢀ
210
120
270
270
300
300
300
300
270
255
80
68
76
76
0.208 395 000
0.197 208 000
0.186 520 000
0.187 480 000
93
1.83
ꢀ
ꢀ
ꢀ
ManA5MAm 26.9
ManA10MAm 29.6
ManA10MAm 29.6
ManA10MAm 30.0
104
199
193
205
211
218
192
205
205
2.05 4.9
2.32 4.9
2.41 4.9
2.05 5.0
2.59 4.9
2.52 4.9
2.40 5.0
2.12 4.0
1.95 2.7
4.1
25
6.6a 50a
6.6a 49a
4 b 0.58
5
6
0.58
0.58
86 (76) 0.191 743 000
85 (51) 0.188 566 000
87 (55) 0.188 611 000
84 (68) 0.192 678 000
98 (100) 0.185 482 000
82 (76) 0.189 612 000
4.3
3.0
3.1
4.1
4.2
2.5
39
30
31
38
52
44
ManA10Am
ManA10Am
ManA10Am
29.9
29.6
31.0
7 b 0.58
8
9
0.58
0.50
ManA16MAm 21.0
GulA20MAm 11.0
10 0.39
General conditions: D2O / DMSOꢀd6 (< 2%), pD ≅ 6ꢀ7, 60 °C, 4,4´ꢀazobis(cyanopentanoic acid),
c0
/ mM = 5.1 (run 1), 3.1 (run 2), 1.0 mM (run 3ꢀ8) or 1.1 mM (run 9ꢀ10). ManAX and GulAX
ACPA
represent (1→4)ꢀβꢀDꢀmannuronan and (1→4)ꢀαꢀLꢀguluronan, respectively, with number average
repeating units “X”. MAm stands for methacrylamide, Am stands for acrylamide. Total conversions (p)
for run 1ꢀ4 were calculated by 1H NMR. Individual monomer conversions were calculated by 1H NMR
for run 6ꢀ8 and by SEC for run 5ꢀ10 and are reported as pHEMAm (pmacromonomer). Ð is the molar mass
dispersity; f and F are the molar fraction of macromonomer in the feed and in the polymer, respectively;
1
Fm is the weight fraction in the polymer; F values were calculated from the H NMR spectra of the
purified polymer (run 2ꢀ4) or from individual monomer conversions. These values are overestimated
a
due to the presence of residual macromonomer and starting oligosaccharide in the dialyzed / diafiltered
polymer sample. b Polymerization carried out in NaCl 0.2 mol Lꢀ1.
Copolymerization of ManAX-(M)Am with HEMAm (X = 5, 10)
(Run no. 2ꢀ4 and 6ꢀ7, Table 1). In a typical experiment (run no. 2) 4,4´ꢀazobis(cyanopentanoic acid)
(ACPA, 1.47 × 10ꢀ2 g, 5.24 × 10ꢀ5 mol) was dissolved in DMSOꢀd6 (0.50 mL), cooled to ~ 8 °C and
diluted with an equal volume of D2O. HEMAm (0.0880 g, 6.82 × 10ꢀ4 mol) was dissolved in D2O (0.40
mL) and filtered through two syringe filters connected in series (1.22 ꢁm glass fiber and 0.22 ꢁm nylon
filter) to remove the suspended inhibitor (BHT). ManA5ꢀMAm (0.0940 g, 47% purity, 4.77 × 10ꢀ6 mol)
was dissolved in D2O (0.62 mL), added to part of the HEMAm solution (0.3 mL, 1.70 mol Lꢀ1) and
9
ACS Paragon Plus Environment

Biomacromolecules
Page 10 of 40
mixed with a calculated amount of initiator (57 ꢁL, 5.24 × 10ꢀ2 mol Lꢀ1, 2.99 × 10ꢀ6 mol). The resulting
1
2
3
4
5
6
7
8
solution (pD ≅ 6ꢀ7) was transferred to a NMR tube equipped with a Young valve, a 1H NMR spectrum
at t = 0 was recorded and the tube was sealed, degassed by 4 freezeꢀevacuateꢀthaw cycles and transferred
to a water bath preheated at 60 °C. After 2 hours the reaction was stopped by plunging the tube in iceꢀ
9
1
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
water and a H NMR spectrum was acquired for global conversion calculation (p). To this end, the Hꢀ
NMR spectra of the starting and final polymerization mixture were normalized to the internal anomeric
proton of the macromonomer (4.64 ppm) and the ethylenic protons were integrated (5.44 and 5.70 ppm
for HEMAm and ManXMAm). The following formula was then applied:
A




ethylenic 

A


ref t
p =1−
(1)
A



ethylenic 

A
ref

0
where Aethylenic and Aref are the area of peak ethylenic and reference peaks, respectively. The polymer
1
was purified by dialysis (20,000 Da cutꢀoff membrane) for 50 hours respectively and freeze dried. Hꢀ
NMR (400 MHz, D2O, 323 K)
δ (ppm): 0.96 (H7, 2H), 1.11 and 1.74 (H10, 3H), 3.27 (H11, 2H), 3.64
(H12, 2H), 3.6ꢀ4.4 (H2ꢀH5, H2´ꢀH5), 4.67 (H1´). The composition of the purified glycopolymer was
1
calculated by normalizing to 2.00 the integral the HꢀNMR CH2ꢀNH signal of HEMAm (3.27 ppm) and
by dividing the integral of the internal anomeric protons of the macromonomer (4.67 ppm for
mannuronan, 5.04 ppm for guluronan) by (XnNMRꢀ1), where XnNMR is the number of repeating units in the
1
oligosaccharide estimated by H NMR. For polymerizations at high ionic strength D2O was replaced
with NaCl 0.2 mol Lꢀ1 in D2O. When the macromonomer was Man10Am (δethylenic 5.76, 6.18 and 6.32
ppm) individual monomer conversions were calculated from 1HꢀNMR.
Kinetic study of the copolymerization of ManA10-MAm and ManA10Am with HEMAm
The above described procedure was modified as follows: The macromonomer was dissolved in D2O
together with a calculated amount of methyl αꢀDꢀglucoside as internal standard; a t0 sample (~ 120 ꢀL)
was drawn from the polymerization mixture, frozen in liquid nitrogen and stored in a freezer until
10
ACS Paragon Plus Environment

Page 11 of 40
Biomacromolecules
needed. The polymerization mixture (pD ≅ 6ꢀ7) was then transferred to a Schlenk tube, sealed with a
rubber septum, degassed with 4 freezeꢀevacuateꢀthaw cycles and transferred to a water bath preheated at
60 °C. At preꢀset intervals, samples (~ 6 mg mLꢀ1) were drawn from the reaction mixture, frozen in
liquid nitrogen and stored in the freezer until needed. After 5 hours (end of the kinetic monitoring) more
initiator was added and the reaction was left running for another 11.5 hours (final conversion ≅ 100%).
Intermediate samples were analyzed by SEC with a column set allowing baseline resolution between
monomer, polymer and internal standard peaks (Shodex OH pak SBꢀ(Guard + 802 + 803) HQ columns).
Individual monomer conversions (p) were then calculated according to the formula:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60

 

A
A
i
i


 


−
⋅
(1− P
)
 
A
A
ISTD
ISTD

t 

0
pi =1−
(2)

A
i



⋅P

A

ISTD 0
where Ai and AISTD are the area of the monomer and internal standard peak respectively, P is the mass
1
purity of the monomer (assumed to be 1 for HEMAm and determined by HꢀNMR for the monomer).
This formula takes into account the fact that macromonomer samples contained an amount (1ꢀP) of the
oligosaccharide from which they were derived (or the corresponding alditol) and that the corresponding
SEC peaks are perfectly superimposed.
Copolymerization of GulA20-MAm and ManA16-MAm with HEMAm
The procedure used for kinetic experiments was modified as follows: The polymerization mixture was
degassed with 3 freezeꢀevacuateꢀthaw cycles; no intermediate samples were drawn; after 4.25 hours the
reaction was stopped by plunging the tube in cold water, a tend sample (~ 120 ꢀL) was drawn for SEC
analysis and the polymer was recovered by diafiltration (30,000 Da cut off membrane) followed by
freeze drying. GulA20-Mam / HEMAm (run no. 10, Table 1), final conversion (SEC): HEMAm 82%,
GulA20ꢀMAm 76%. 1HꢀNMR (400 MHz, D2O, 328 K)
δ (ppm): 0.98 and 1.12 (3H, H10), 1.75 and 1.86
(2H, H7), 3.28 (2H, H11), 3.66 (2H, H12), 3.88 (H2, H2´), 4.00 (H4, H4´), 4.11 (H3, H3´), 4.44 (H5,
H5´), 5.05 (H1´). ManA16-MAm / HEMAm (run no. 9, Table 1), final conversion (SEC): HEMAm
11
ACS Paragon Plus Environment

Biomacromolecules
Page 12 of 40
98%, ManA16ꢀMAm ~100%. 1H NMR (400 MHz, D2O, 328 K)
δ (ppm): 0.98 and 1.12 (3H, H10), 1.75
1
2
3
and 1.86 (2H, H7), 3.28 (H11, 2H), 3.6ꢀ4.4 (H2ꢀH5, H2´ꢀH5´), 4.64 (H1´).
4
5
6
Analytical techniques
7
8
Nuclear Magnetic Resonance. Spectra were acquired on a Bruker DPX400 spectrometer equipped
9
1
with a Variable Temperature (VT) module (resonance frequency of 400.13 and 100.62 MHz for H and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13C nuclei, respectively). Two different 5 mm detection probes were used: QNP (direct) and BBIZ
(inverse). Unless otherwise specified, for 1H experiments 90° pulses and pulse sequence recycle times of
3 s were used. The probe temperature was calibrated in the range 303ꢀ363 K using 80% ethylene glycol
33
1
in DMSOꢀd6. 1D H spectra were obtained with 32ꢀ128 scans and 32 K data points, and were reꢀ
processed using MestReNova software (v5.1). Sodium 3ꢀ(trimethylsilyl)propanoate (TSP) or sodium 3ꢀ
(trimethylsilyl)propaneꢀ1ꢀsulfonate (DSS) were used as an internal reference for samples dissolved in
D2O, whereas tetramethylsilane (TMS) was used in all other cases. Chemical shifts (in ppm) were
referenced to δTSP = ꢀ0.017 ppm (1H) and ꢀ0.149 ppm (13C), or to δDSS and δTMS = 0.000 ppm (1H and
13C).
Mass spectrometry. ESIꢀMS analyses were performed with a Waters ZQ (Altrincham, GB) single
quadrupole atmospheric pressure ionization mass spectrometer fitted with a Z electrospray interface
(ESI). The instrument was calibrated with mass spectra generated by ion spray ionization of a 0.1 mol.Lꢀ
1
solution of sodium iodide in aqueous acetonitrile (50%, v/v) in the mass range of 23ꢀ1972 amu.
Nitrogen was used as the drying and nebulizing gas. Samples (~1 mg mLꢀ1) were dissolved in deionised
water or water/methanol mixtures and infused to the ESI interface at constant flow rate (50 ꢁL minꢀ1).
MALDI –Tof analyses were performed on a Voyager DEꢀPro (AB Sciex) equipped with a nitrogen laser
(λ=337 nm). The spectrometer was operated in negative ion reflector or linear mode and mass spectra
were obtained by accumulating 300 laser shots for each analysis. Samples were solubilized in water (20
g Lꢀ1) and mixed with an equal volume of 2ꢀ(4ꢀhydroxyphenylazo)benzoic acid / 1,1,3,3ꢀ
34
tetramethylguanidine (1:2 molar ratio) in methanol as described by Przybylski et al. The resulting
12
ACS Paragon Plus Environment

Page 13 of 40
Biomacromolecules
solution (1 ꢂL) was then deposited target and allowed to dry at room temperature and atmospheric
pressure.
1
2
3
4
5
6
7
8
Size Exclusion Chromatography (SEC-DV-MALLS). Molar mass, molar mass distributions and
intrinsic viscosities were measured with a Size Exclusion Chromatography (SEC) system consisting of
an Alliance GPCV 2000 chromatograph (Waters) equipped with a differential refractometer (λ = 880
nm) and a 3 capillary differential viscometer (DV), and interfaced with a multiꢀangle laser light
scattering (MALLS) detector (DAWN HELEOS II, Wyatt Technology Corp., Santa Barbara, California;
λ = 658 nm). The system was equipped with a 50×6 mm guard column and two 300×8 mm linear
columns (Shodex SBꢀ800 HQ series). An aqueous buffer (NaNO3 0.1 M, NaN3 0.03% w/v, NaꢀEDTA
0.01 M) was used as eluant at a flow rate of 0.5 mL minꢀ1 while temperature of the columns, DRI and
viscometer was maintained at 30 °C. Samples were prepared by diluting the polymerization mixtures or
by dissolving pure polymer samples directly in the eluant at concentrations of ~1 g Lꢀ1. The resulting
solutions were filtered through 0.22 ꢂm sterile syringe filters (Millex GS, Millipore) and injected in 100
ꢀL volumes. Differential refractive index increments dn/dc used for molecular weight calculations were
measured at 30 °C with an Optilab rEX differential refractometer (Wyatt Technology Corp., λ = 633
nm). Results were analyzed with ASTRA 5.3 software (Wyatt Technology Corp.).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For each slide of the chromatogram, the molar mass was obtained from the intensity of the scattered
light at different angles as the reciprocal of the intercept at
θ
= 0 of a plot: 35
Kc
Rθ
θ
2
 
 
vs.sin2
(3)
 
where
θ is the scattering angle, K is an optical constant containing the specific refractive index
increment of the polymer (dn/dc), Rθ is the Rayleigh ratio and c is mass concentration. To this end, the
term 2 A2c in the classic Zimm plot was considered to be negligible. Radii of gyration Rg were obtained
from the slope of the same plot and their zꢀaverage was computed for the whole sample. Intrinsic
viscosities were obtained as mass averages by using the approximation:
13
ACS Paragon Plus Environment

Biomacromolecules
Page 14 of 40
ηsp
1
2
(4)
[η] ≅
c
3
4
5
where ηsp is the specific viscosity. Eq. 4 is generally valid for dilute solutions of low molar mass
6
7
8
9
36
polymers of small intrinsic viscosity.
chromatograms and [
In our case c was less than 0.3g Lꢀ1 at the peak of the
η
] ≤ 220 mL gꢀ1. Intrinsic viscosities were used to calculate approximated
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
hydrodynamic radii (Rh) for individual polymer chains via the EinsteinꢀSimha equation for inelastic
spheres: 35, 36

1/3
3M[η]
10πNA
(5)




Rh ≅


where M is the molar mass of the polymer and NA is Avogadro’s number.
Differential refractive index increments. The value of dn/dc for poly(HEMAm) (run no. 2) and
poly(HEMAmꢀcoꢀManA5MAm) (run no. 3) was measured at 30 °C with an offline Optilab® rEX
differential refractometer (λ = 633 nm; Wyatt Technology Corp.) operated with ASTRA 5.3 software.
To this end, poly(HEMAm) was isolated by reꢀprecipitating it in acetone twice followed by extensive
drying under mechanical vacuum, whereas poly(HEMAmꢀcoꢀManA5MAm) was purified by diafiltration
(30,000 Da molecular weight cutꢀoff) and freezeꢀdried. In both cases, the residual water content was
estimated by TGA (vide infra). Polymer solutions of known concentration were then prepared
gravimetrically by dissolving the polymers into the eluant used for SEC analysis (d30= 1.0035 g mLꢀ1)
and injected at a flow rate of 0.25 mL minꢀ1. The pure eluant was also injected. The instrument measures
the refractive index of (n) for each concentration (c), subtracts the refractive index of the eluant (n0) and
calculates dn/dc from the slope of the plot ꢁn vs. c according to the following equation:
dn
(6)
ꢁn =
× c
dc
where ꢁn = n ꢀ n0. The refractive index increments of the copolymers were estimated from the mass
fraction (Fm) of each monomer and the dn/dc of the corresponding homopolymers according to the
14
ACS Paragon Plus Environment

Page 15 of 40
Biomacromolecules
formula: 37
1
2
3
4
5
6
7
8
dn/dc = Fm,1(dn/dc)1 + Fm,2(dn/dc)2
To this end, dn/dc values of 0.165 mL gꢀ1 and = 0.208 mL gꢀ1 were used for alginate
poly(HEMAm) (this study), respectively.
(7)
17
and
9
Rheometry. The rheological properties of polymer gels were characterized with an AR2000
rheometer (TA instruments) at 25 °C. To this end, polymers were isolated by diafiltration (30,000 Da
molecular weight cutꢀoff), lyophilized, and their residual solvent content was determined by Thermo
Gravimetric Analysis. For solution measurements, polymers were dissolved in NaCl 0.1 mol Lꢀ1 at twice
the critical overlapping concentration (c∗) as estimated from the intrinsic viscosity in NaNO3 0.1 mol Lꢀ1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
obtained by SECꢀIVꢀMALLS analysis (c∗ ≈ 2.5 / [
η]). The resulting solutions were subjected to steadyꢀ
state and dynamic (oscillatory) experiments using a coneꢀplate rheometer (diameter = 4 cm, angle =
3°59) with an interꢀconeꢀplate gap of 113 ꢀm. Gels of poly(HEMAmꢀcoꢀManA16MAm) and
poly(HEMAmꢀcoꢀGulA20MAm) were prepared by dialyzing a glycopolymer solution in deionized water
(c0 ≅ 1.5 c* = 18 mg mLꢀ1, pH ≅ 6ꢀ7) against a solution of CaCl2 (0.5 mol Lꢀ1) for 28ꢀ48 hours. To this
end, small dialysis cassettes were used (SlideꢀAꢀLyzer, 0.5ꢀ3.0 mL, MWCO 2000 Da, Pierce), from
which the gel was recovered by cutting off the dialysis membrane with a scalpel. The resulting material
was either punched into a disk (poly(HEMAmꢀcoꢀGulA20MAm), ∅ = 2.0 cm, h ≅ 0.32 cm) with the
same diameter of the rheometer plate, or transferred onto the same plate as a film. In order to prevent
slippage effects, a sanded plate was used for gel characterization. For each type of hydrogel, the linear
viscoelastic regime was determined by an amplitude sweep to check that the measured moduli were
independent of the deformation applied. The rheological properties were then investigated on fresh
samples in oscillatory and compression mode using a parallelꢀplate system. The Young modulus of the
hydrogel of poly(HEMAmꢀcoꢀGulA20MAm) was measured by placing a disk of area S = 3.14 × 10ꢀ4 m2
between the two parallel plates of the rheometer and by squeezing it while recording the backward force
15
ACS Paragon Plus Environment

Biomacromolecules
Page 16 of 40
exerted. The elastic modulus E was calculated by dividing the initial slope of the force vs. deformation
curve by the area of the disk.
1
2
3
4
5
6
RESULTS AND DISCUSSION
7
8
Synthesis of macromonomers
9
Glycuronans are polysaccharides containing only uronic acid residues, i.e. monocarboxylic acids
formally derived from aldoses by replacement of the hydroxymethyl group ꢀCH2OH with a carboxy
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
38
group. In order to incorporate a glycuronan into a biohybrid polymer (or other glycoconjugage) it is
advantageous to selectively functionalize its reducing end without resorting to protective group
39
chemistry, the latter being exceedingly time consuming. A common strategy for the derivatization of
monoꢀ and oligosaccharides is their transformation into the corresponding 1ꢀaminoꢀ1ꢀdeoxyalditols via
40ꢀ44
reductive amination.
In this reaction, the masked aldehyde group at the reducing end of the
carbohydrate condenses with an amine (or ammonia) molecule to give an amino alcohol, which is then
protonated at the oxygen atom and eliminates a water molecule to give an iminium ion. In the presence
45, 46
of a suitable reducing agent, the latter is rapidly reduced to a stable amine.
At a laboratory scale,
sodium or lithium cyanoborohydride is the reducing agent of choice for reductive aminations in protic
solvent, thanks to its remarkable stability in aqueous solution (at pH > 2.5) and to its pHꢀdependent
chemoselectivity. 47 In particular, if the substrate is a ketone or aldehyde, at pH 6ꢀ8 the reduction of the
iminium ion is sufficiently faster than that of carbonyl groups to enable its formation and reduction in
situ. The reductive amination of oligoalginates with a bifunctional amine in water has already been
48, 49
reported in the literature
but the characterization of the resulting materials was basic and
functionalization yields were not determined. Concerning other glycuronans, Ridley et al. reacted
analytical quantities of (1→4)ꢀαꢀDꢀgalacturonan in methanol with a biotinꢀhydrazide derivative and
40
reduced the resulting hydrazone with NaBH3CN in water. The procedure involved repeated solvent
evaporation steps and is mostly applicable to microscale reactions. Also, the yield of oligogalacturonan–
16
ACS Paragon Plus Environment

Page 17 of 40
Biomacromolecules
hydrazide from a single reaction decreased with the size of the oligosaccharide, from 88% for X = 8 to
39% for X = 16 (where X is the degree of polymerization).
1
2
3
4
5
6
In this study (Scheme 1), (1→4)ꢀβꢀDꢀmannuronan (ManAX, X = 5, 10 and 16) and (1→4)ꢀαꢀLꢀ
7
8
9
guluronan oligosaccharides (GulA20) were reacted for several days with an excess of ammonium acetate
(>40 eq.) in aqueous solution and in the presence of NaBH3CN (~50 eq.). 50 Although conversion of the
starting carbohydrate as measured by the disappearance of the reducing end anomeric proton was >90%
in all cases, the mole fraction of 1ꢀaminoꢀ1ꢀalditol was in the range 37ꢀ46 %, thus suggesting the
concomitant formation of byꢀproducts (most probably alditol). The reaction rate was generally slow and
under identical conditions, the reaction of GulA20 was ~3 times slower than that of ManA16, thus
indicating a slower formation of the intermediate iminium ion for guluronan in the rate determining
step.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Optimization of this reaction was beyond the scope of this work and, after elimination of salts by
diafiltration, the obtained products were directly used for the following step. The rationale behind this
choice was that residual nonꢀaminated oligosaccharides would be inert towards the remaining synthetic
steps (acylation and radical polymerization) and could be separated from the final polymer by simple
diafiltration or dialysis (the hydrodynamic diameter of the two species differing by two orders of
magnitude). In particular, in basic aqueous solution (Schotten–Baumann conditions) an acid chloride
will rapidly react with the primary amine and any excess reagent will be hydrolyzed by water (c = 55
mol Lꢀ1) before acylating the hydroxyl groups of the carbohydrate (c ≤ 0.6 mol Lꢀ1), the two reactions
51
having similar rate constants. Likewise, carbohydrates readily react with oxygenꢀcentered radicals by
52‡
hydrogenꢀatom abstraction at the ring CꢀH bonds
but are unaffected by most of the secondary and
tertiary carbonꢀcentered radicals encountered in radical polymerization processes (alkyl ester radicals
being a notable exception). 53
‡ For the same reason peroxy initiators should be avoided in the synthesis of glycopolymers.
17
ACS Paragon Plus Environment

Biomacromolecules
Page 18 of 40
1
Hence, the purity of the 1ꢀaminoꢀ1ꢀalditols obtained by reductive amination was quantified by H
NMR and the same were directly reacted with (meth)acryloyl chloride (~15 eq.) in aqueous carbonate
buffer at pH 9.5. This pH value was found to be the best compromise between the quest for fast amide
bond formation (which demands a high pH) and the need to avoid degradation of the glycuronans under
1
2
3
4
5
6
7
8
9
54
basic conditions. When an excess of base was used (> 2 eq.), the acylation step was quantitative and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
the limiting step for endꢀofꢀchain functionalization was reductive amination. Figure 1 shows the H
NMR spectra of (1→4)ꢀβꢀDꢀmannuronan (Xn = 10; ManA10), of the corresponding 1ꢀaminoꢀ1ꢀalditol
(ManA10ꢀNH2) and of the final macromonomer ManA10ꢀMAm. Note the almost complete disappearance
of the reducing end anomeric protons H1α/β at 4.84ꢀ4.89 ppm and 4.20 ppm and the appearance of two
double doublets at 3.00 and 3.46 ppm when going from ManA10 to ManA10ꢀNH2. Also, note the
deshielding of H1a (from 3.00 ppm to 3.38 ppm) and H1b (no longer visible after acylation) and the
appearance of two ethylenic protons at 5.45 ppm (H7 cis to C10) and 5.70 ppm (H7 trans to C10) when
going from ManA10ꢀNH2 to ManA10MAm. The fact that the two H1 protons in the latter compounds are
not magnetically equivalent is due to the presence of an asymmetric center at C2. The structure of the
macromonomer was further confirmed by MALDIꢀToF analysis (Figure 2, reflectron mode), in which
peaks corresponding to [MꢀH]ꢀ and spaced of 176 Da are visible which correspond to oligosaccharides
containing between 5 and 10 repeating units and bearing an endꢀofꢀchain methacrylamide group. Higher
molar mass species were not visible in the reflectron spectrum but could be seen in spectra acquired in
linear mode, albeit with a lower resolution.
18
ACS Paragon Plus Environment

Page 19 of 40
Biomacromolecules
1
2
3
H1
H7 trans
H1a
4
H7 cis
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
H1
H1b
H1a
H1(G)
H1
H1α
H1β
H1(G)
1
Figure 1. H NMR spectra of (a) (1→4)ꢀβꢀDꢀmannuronan (Xn = 10), (b) the corresponding 1ꢀaminoꢀ1ꢀ
alditol (ManA10ꢀNH2) and (c) the final macromonomer ManA10ꢀMAm. Note that the internal anomeric
proton of residual guluronic acid units H1(G) is visible at 5.04 ppm. Conditions: 400.13 MHz, D2O, T =
328 K for (a) and 323 K for (b) and (c).
19
ACS Paragon Plus Environment

Biomacromolecules
Page 20 of 40
1
2
3
4
7
ManA10MAm
5
6
negative ion mode
7
8
9
8
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6
1494
1496
1498
X = 5
9
176
10
1000
1200
1400
m/z
1600
1800
Figure 2. MALDIꢀToF spectrum of ManA10ꢀMAm. Reflectron mode, negative ions. For each peak, the
degree of polymerization (X) of the corresponding oligosaccharide is also reported.
Polymerization experiments and synthesis of graft copolymers
The obtained macromonomers were copolymerized with 2ꢀhydroxyethylmethacrylamide (HEMAm) in
aqueous solution at 60 °C in the presence of 4,4´ꢀazobis(cyanopentanoic acid) (ACPA; ~1 mmol Lꢀ1).
55
HEMAm was chosen as a comonomer since it leads to highly hydrophilic copolymers, has already
56
been studied for the preparation of soft contact lenses
and because it closely resembles 2ꢀ
hydroxypropylmethacrylamide (HPMAm) (the only difference being a methyl group on the side chain),
whose homopolymer is nonꢀtoxic and nonꢀimmunogenic and whose copolymers have already been used
for the preparation of bioconjugates for intravenous administration, some of which are now under
20
ACS Paragon Plus Environment

Page 21 of 40
Biomacromolecules
57, 58
clinical trial.
Obviously, any other hydrophilic methacrylamide derivative could be used as a
1
2
3
4
5
comonomer, but we expect poly(HEMAm) to be as biocompatible as poly(HPMAm) while being even
more hydrophilic.
6
The total monomer concentration was 0.5ꢀ0.6 mol Lꢀ1 in all runs except run no. 10 (Table 1), in which
case the lower solubility of GulA20MAm required a slight dilution of the reaction mixture. Nonꢀuniform
(dispersity Ð ≅ 2.0ꢀ2.6), high molar mass copolymers were obtained in run 3ꢀ10 (Mn ≅ 480 000 – 680
000 Da), whereas a somewhat lower molar mass copolymer was obtained with ManA5MAm due to the
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
§
higher concentration of initiator and, possibly, to the inferior purity of the macromonomer batch. After
purification by diafiltration, the copolymers composition was determined by 1H NMR and it was found
that the final materials contained between 2.5 mole % and 4.3 mole % of macromonomer, which
correspond to 25% and 52 % by mass of oligosaccharide graft chains. In the case of ManA10MAm and
ManA10Am the rate of incorporation of the macromonomer was investigated by adding an internal
standard (ISTD, methyl αꢀDꢀglucoside) to the polymerization mixture and by taking samples at
intermediate reaction times (run 5 and 8 in Table 1). Since the ethylenic protons of HEMAm and
1
ManA10MAm have identical chemical shifts in H NMR, individual monomer conversions were
calculated by SEC as described in the experimental section. The two experiments were carried out under
identical conditions and the resulting conversion vs. time and firstꢀorder plots are reported in Figure 3.
The rate of polymerization of HEMAm was identical in the two cases and followed pseudo first order
kinetics. Both macromonomers were incorporated into the polymer since the early stages of the process
but their rate of conversion was slower than that of HEMAm, the difference being more pronounced in
the case of ManA10Am. Also, the first order plot (Figure 3b) shows that for the two macromonomers the
rate of conversion was slower at the beginning and slightly accelerated with time. The absence of an
inhibition period at the beginning of the polymerization confirms that simple filtration of the aqueous
§ ManA5MAm was the only macromonomer not to purified by diafiltration due to its small size.
21
ACS Paragon Plus Environment

Biomacromolecules
Page 22 of 40
monomer stock solution through a 0.22 ꢀm nylon filter is an effective way to remove 2,6ꢀdiꢀtertꢀbutylꢀ4ꢀ
1
2
3
methylphenol (an hydrophobic radical inhibitor). 59
4
5
6
7
8
9
2.0
(b)
1.6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1.2
0.8
0.4
0.0
0.8
(a)
0.6
0.4
run 5 ManA10MAm
HEMAm
run 8 ManA10Am
0.2
HEMAm
0.0
0
40 80 120 160 200 240 280 320
t / min
Figure 3. Conversion vs. time (a) and firstꢀorder plot (b) for the radical copolymerization of
ManA10MAm (full symbols) and ManA10Am (empty symbols) with 2ꢀhydroxyethylmethacrylamide
(D2O, 60 °C; run 5 and 8 in Table 1). The two experiments were carried out under identical conditions
and monitored by offꢀline Size Exclusion Chromatography (p is fractional monomer conversion).
Dashed and solid lines indicate linear regression of experimental data.
22
ACS Paragon Plus Environment

Page 23 of 40
Biomacromolecules
From the conversion vs. time data for individual comonomers (Figure 3a), the instantaneous feed (f)
and copolymer composition (F) was calculated as:
1
2
3
4
5
6
7
fm0acro
⋅
(
1− pmacro
)
fmacro
=
(8)
fm0acro
⋅
(
1− pmacro
)
+ fH0EMAm
⋅
(
1− pHEMAm
)
8
9
dp




fm0acro
⋅
d[macro]
macro 
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
dt

dt
Fmacro
=
=
(9)
d[macro] d[HEMAm]
dp
dp


macro 


fm0acro
⋅
+ f 0
⋅



HEMAm 
dt

+
HEMAm
dt
dt
dt


where [A], f 0A and pA are the concentration, initial feed ratio and conversion, respectively, of species
A.
This analysis shows that up to 50ꢀ60% conversion the mole fraction of macromonomer in the newly
formed polymer was smaller than in the feed (Figure 4). ManA10MAm was better incorporated than
ManA10Am, with F / f = 0.74 at 10% conversion versus 0.60 for the analogous acrylamide derivative.
Also, above 40% conversion F(ManA10Am) started to drift rapidly, by 70% conversion it overtook f and
ended up being a multiple of it by the end of the process (F / f ≅ 3 at 83 % conversion). A qualitatively
similar behavior was observed for ManA10MAm, but in this case the highest F / f ratio was 1.4 at 76%
conversion. Overall, for the methacrylamideꢀderived macromonomer F increased from 3.9% at p = 0.10
to 10% at p = 0.76 (i.e. a 2.6ꢀfold increase), whereas for the acrylamide analogue F was 3.2% at p =
0.10 but 32% at p = 0.84 (i.e. a 10ꢀfold increase). These findings highlight the importance of the nature
of the polymerizable moiety for the copolymerization of glycuronanꢀderived macromonomers with
HEMAm and show that even when both monomers are methacrylamide derivatives (i.e. HEMAm and
ManA10MAm), the glycuronanꢀderived one reacts more slowly. The higher reactivity of ManA10MAm
towards poly(2ꢀhydroxyethylmethacrylamid)yl radicals is consistent with the findings of Dainton and
Sisley 60 for the copolymerization of methacrylamide and acrylamide, for which the values of r1 = 1.10 ±
0.20 and r2 = 0.74 ± 0.10 were estimated. In our case though, the copolymerization cannot be analysed
in the frame of a simple terminal model because of the bulkiness of the macromonomer and the multiple
23
ACS Paragon Plus Environment

Biomacromolecules
Page 24 of 40
negative charges that it bears. For instance, under the conditions used for copolymerizations neither
ManA10MAm nor ManA10Am homopolymerized.
1
2
3
4
5
6
7
8
To take into account electrostatic repulsion as a cause for the slower incorporation of the
macromonomers, parallel experiments were carried out in which either ManA10MAm or ManA10Am
were copolymerized with HEMAm in deionized water (run 3 and 6 in Table 1) or in NaCl 0.2 mol Lꢀ1
(run 4 and 7). Here, the role of the salt was to screen the electrostatic repulsion and facilitate the
incorporation of the macromonomer. In the case of ManA10MAm, the copolymer composition was not
determined since the purification step did not eliminate all unreacted macromonomer and since the
individual monomer conversion was not available (the ethylenic protons of the two monomers have
identical chemical shift in 1H NMR). Nevertheless, for an identical total conversion (p = 0.76) polymers
with similar molar masses were obtained in deionized water and in NaCl 0.2 mol Lꢀ1, suggesting that
electrostatic screening did not significantly increase the mole fraction of incorporated macromonomer.
This conclusion is supported by the results obtained with ManA10Am: The addition of salt resulted in an
increase from 51% to 55% of the final macromonomer conversion, from 570 000 Da to 610 000 Da of
the molar mass of the copolymer and from 3.0% to 3.1% of the mole fraction of macromonomer in the
polymer; i.e. differences comparable with the error inherent to the experimental procedures used.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
24
ACS Paragon Plus Environment

Page 25 of 40
Biomacromolecules
1
2
3
4
5
6
fmacro
Fmacro
F cum
F cum
macro
fHEMAm
FHEMAm
HEMAm
7
8
9
0.90
0.80
0.70
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
(b) ManA10Am / HEMAm
0.08
0.04
0.00
0.30
0.20
0.10
0.1 0.2 0.3 0.4 0.5 0.6 0.7
0.95
0.90
(a) ManA10MAm / HEMAm
0.10
0.08
0.06
0.04
0.02
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
p
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Variations in feed (f) and copolymer composition (F) with conversion for the copolymerization
of ManA10MAm (a) and ManA10Am (b) with HEMAm (run no. 5 and 8 in Table 1). The cumulative
mole fraction in the copolymer (F cum) is also shown.
As explained in the introduction, between 8 and 18 repeating units in a guluronan block are thought to
be necessary to the formation of a stable complex with calcium ions. 18, 19 Based on the results reported
above, two macromonomers (ManA16MAm and GulA20MAm) possessing a suitably long glycuronan
25
ACS Paragon Plus Environment

Biomacromolecules
Page 26 of 40
chain were copolymerized with HEMAm and the resulting graft copolymers were purified by
diafiltration and freeze drying (run 9 and 10 in Table 1). In the two cases, high molar mass polymers
were obtained (Mn ≅ 550 000 Da, Ð = 2.0ꢀ2.1) whose structure was investigated by 1H NMR (Figure 5;
see Scheme 2 for position numbering): The absence of ethylenic proton signals at 5.45 and 5.70 ppm
confirms the elimination of all unreacted monomer. The broad peaks in the 0.90 – 2.00 ppm region are
due to the polymethacrylamide main chain (H7 and H10) whereas the side chain methylene protons of
HEMAm are visible at 3.28 ppm (H11) and 3.66 ppm (H12). The internal anomeric proton peak H1 is
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
found at 4.64 ppm for (1→4)ꢀβꢀDꢀmannuronan and at 5.05 ppm for (1→4)ꢀαꢀLꢀguluronan, whereas all
other methine protons from the carbohydrates resonate in the region 3.7ꢀ4.3 ppm. From the integration
of H11 and H1 the mole faction of oligosaccharide in each polymer was calculated and the
corresponding mass fraction was deduced by taking into account the relative molar mass of the
comonomers. It was thus found that poly(HEMAmꢀgꢀManA16) and poly(HEMAmꢀgꢀGulA20) contained
52% and 44% by mass of graft chains, respectively.
26
ACS Paragon Plus Environment

Page 27 of 40
Biomacromolecules
1
2
3
4
H12
5
6
H11
H7
7
8
9
H2ꢀH4
H1
H10
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
H12
H2ꢀH5
H1
H11
H7
H10
1
Figure 5. H NMR spectra of (a) poly(HEMAmꢀgꢀManA16) and (b) poly(HEMAmꢀgꢀGulA20). See
Scheme 2 for peak assignment. The integral of CH2ꢀNHR (H11) was normalized to 2 in both spectra.
Conditions: 400.13 MHz, D2O, 328 K.
Ionotropic gelation
Poly(HEMAmꢀgꢀGulA20) is a high molar mass biohybrid copolymer carrying (1→4)ꢀαꢀLꢀguluronan
residues that are long enough to form stable complexes with suitable divalent cations. Such complexes
were expected to act as cross links between polymer chains and to induce ionotropic gelation of the
polymer. In this respect, poly(HEMAmꢀgꢀManA16) was designed as a “negative control”, i.e. an
alginateꢀderived biohybrid polymer incapable of ionotropic gelation. The two polymers had similar
molecular size and architecture and the main difference was the degree of branching: Fmacromonomer = 4.0
% for poly(HEMAmꢀgꢀManA16) vs. 2.7 % for poly(HEMAmꢀgꢀGulA20).
27
ACS Paragon Plus Environment

Biomacromolecules
Page 28 of 40
Initially, the presence of loose interꢀchain interactions in the absence of gelling ions was checked by
preparing semiꢀdilute solutions of both polymers in NaCl 0.1 mol Lꢀ1. To this aim, samples were
dissolved at twice the critical overlapping concentration c∗ (estimated by the intrinsic viscosity measured
in NaNO3 0.1 mol Lꢀ1) and their rheological behaviour was investigated. Figure 6 shows the variation of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
&
the steady state shear viscosity (η) with shear rate ( ) for the two polymers: Clearly, in the narrow shear
γ
−1
≤10 s−1 ) both solutions behaved as Newtonian fluids (
η
≈ constant).
&
γ
rate window tested (0.5 s ≤
Dynamic (oscillatory) experiments confirmed that in the frequency range 0.5 – 10 Hz the value of the
loss modulus G´´ was the same for poly(HEMAmꢀgꢀManA16) and poly(HEMAmꢀgꢀGulA20) (Figure 7)
and that G´´ was systematically higher than the corresponding storage modulus G´. Indeed, in the case of
poly(HEMAmꢀgꢀGulA20) G´ was too small to be accurately measured.
10-1
9x10-2
8x10-2
10-1
9x10-2
8x10-2
7x10-2
6x10-2
7x10-2
6x10-2
5x10-2
poly(HEMAm-g-GulA20)
poly(HEMAm-g-ManA16)
5x10-2
4x10-2
4x10-2
3x10-2
2x10-2
3x10-2
2x10-2
η
10-2
10-2
1
10
.
γ
/ s-1
&
η as a function of the shear rate for poly(HEMAmꢀgꢀManA ) (blue triangles) and
γ
16
Figure 6. Viscosity
poly(HEMAmꢀgꢀGulA20) solutions (red squares) in NaCl 0.1 mol Lꢀ1 at 25 °C.
28
ACS Paragon Plus Environment

Page 29 of 40
Biomacromolecules
1
2
3
4
5
gel
solution
,
poly(HEMAm-g-ManA16)
,
poly(HEMAm-g-GulA20)
,
6
7
8
9
103
102
101
100
10-1
10-2
103
102
101
100
10-1
10-2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
1
10
Frequency (Hz)
Figure 7. Variation of G´ and G´´ with frequency for the gels obtained from poly(HEMAmꢀgꢀManA16)
and poly(HEMAmꢀcoꢀGulA20) at 25 °C in the linear domain. Note that in the case of poly(HEMAmꢀgꢀ
GulA20) the value of G´ is missing because it was too small to be accurately measured.
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A gelation experiment was then realized starting from polymer solutions in deionized water (c0 ≅ 1.5
c∗). The latter were transferred to small dialysis cassettes possessing two parallel membranes and
dialyzed against a CaCl2 0.5 mol Lꢀ1 for 28ꢀ48 hours. The resulting gels were recovered by cutting off
the dialysis membrane with a scalpel: As shown in Figure 8, poly(HEMAmꢀgꢀManA16) formed a loose
opaque gel with the consistency of an after shave balm that was peeled off the dialysis membrane;
whereas poly(HEMAmꢀcoꢀGulA20) gave a soft selfꢀstanding transparent hydrogel from which a disk was
punched out for rheological characterization (vide infra).
29
ACS Paragon Plus Environment