Thiol-alkyne Chemistry for the Preparation of Micelles with Glycopolymer Corona: Dendritic Surfaces...
-
Upload
jatin-kumar -
Category
Documents
-
view
212 -
download
0
Transcript of Thiol-alkyne Chemistry for the Preparation of Micelles with Glycopolymer Corona: Dendritic Surfaces...
Communication
1620
Thiol-alkyne Chemistry for the Preparation ofMicelles with Glycopolymer Corona: DendriticSurfaces versus Linear Glycopolymer in TheirAbility to Bind to Lectinsa
Jatin Kumar, Antoine Bousquet, Martina H. Stenzel*
M. H. Stenzel, J. KumarCentre of Advanced Macromolecular Design (CAMD), School ofChemical Engineering, The University of New South Wales, NewSouth Wales 2052, AustraliaE-mail: [email protected]. KumarCooperative Research Centre for Polymers, 8 Redwood Drive,Notting Hill, Victoria 3168, AustraliaA. BousquetDepartment of Chemistry, Aarhus University, Langelandsgade140, DK-8000, Aarhus C., Denmark
a Supporting Information for this article is available at Wiley OnlineLibrary or from the author.
A poly(tert-butyl acrylate) (P(tBA)) with a glycodendric endfunctionality with eight glucosemoieties was synthesised in four steps via a combination of esterification, thiol-alkyneconjugation and hetero-Diels–Alder (HDA) cycloaddition. A linear glycopolymer of similarsize and composition was also synthesised in order to compare the protein binding charac-teristics of the polymer with glycodendritic endfunctionality to the linear glycol blockcopo-lymer. The two amphiphilic polymers were self-assembled in water into micelles. Theseparticles were then tested for their abilityto bind to Concanavalin A (Con A). In aturbidity assay, the polymer glycodendronexhibited a significantly faster clusteringrate to the lectin as compared to the linearglycopolymer. In a precipitation assay, it isfound that significantly less glucose resi-due is required for binding per Con A forthe polymer with the glycodendritic end-functionality.
Macromol. Rapid Commun. 2011, 32, 1620–1626
� 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonline
Introduction
Glycopolymers play a pivotal role in biological interac-
tions[1] anda lot of efforthasbeendevoted in recentyears to
optimise the synthesis of glycopolymers and glycodendri-
mers.[2] Dendrimers have been employed in a varied range
of applications,[3] however,most potential for them appear
to lie in drug delivery[2c,4] and biotechnological applica-
tions.[5] Glycodendrimers allow for multivalent interac-
tions with proteins, similar to the use of multivalent
glycopolymers, thereby expanding theuseof saccharides in
drug delivery mechanism.[1a] What sets glycodendrimers
apart from glycopolymers though, is that studies have
shown that glycodendrimers exhibit better protein inter-
action as compared to linear glycopolymers. Gillies and
library.com DOI: 10.1002/marc.201100331
Thiol-alkyne Chemistry for the Preparation of Micelles with Glycopolymer Corona: . . .
www.mrc-journal.de
colleagueshypothesise that this is due to the inherent steric
hindrance of the dendrimer that allows for the saccharides,
which are often terminally functionalised, to present
themselves on the surface of the molecular matrix as
opposed to the case of a glycopolymer where the
saccharides may be within the polymer matrix.[3e]
Previous studies of polymer-based dendrons have
generally involved repetitive multi-stage reactions to
achieve a higher number of branches. With an increase
in the number of steps, the most salient issue is ensuring
that yield is optimised through orthogonal and efficient
reactions.
‘Click’ chemistry techniques are by their very definition
orthogonal, efficient and versatile. Both copper catalysed
azide-alkyne click (CuAAC)[6] aswell as the recentlypopular
thiol-alkene[7] ‘click’ chemistry have been reportedly used
to synthesise dendrimers with high efficiency. There is
also a report of the synthesis of dendrimers by a
combination of CuAAC and traditional esterification
with high yields and efficiency.[8] However, dendrimers
synthesised via the CuAAC technique may ultimately
not be biocompatible due to the toxicity of copper.[9]
Chen et al. synthesised a dendrimer holding 48 terminal
hydroxyl functionalities within 3 steps via a combina-
tion of the thiol-alkyne technique as well as a simple
esterification reaction.[10] In comparison Antoni et al.
required eight steps to synthesise a dendrimer holding
the same number of functionality.[11]
Another polymer structure that has shown to have
promising lectin binding characteristics is a micelle.
Micelles were shown to have better lectin binding abilities
than their linear counterparts.[12] Their spherical shape
resembles that of dendrimers although the size is typically
much larger.
The central question is now if a dendritic structure is
advantageous over a traditional micelles structure or if the
introduction of a dendritic structure on the surface of a
micelle can increase binding with lectins. The aim of this
study is thus to synthesise polymers with glycodendritic
endfunctionality to assess their lectin binding character-
istics as compared to linear glycopolymers. Polymers with
dendritic endfunctionalities have been synthesised earlier
via reversible addition fragmentation chain transfer (RAFT)
polymerisation[13] but the aimof thiswork is to specifically
compare the performance of traditional glycopolymer
micelles with dendritic structures.
For this purpose two different glycopolymers are
envisaged: a poly(tert-butyl acrylate) (P(tBA)) was used
as the hydrophobic block while eight glucose units
represented the hydrophilic part either in linear or
dendritic arrangement. The dendritic amphiphilic block
copolymer was prepared by subsequent steps consisting
of RAFT polymerisation, RAFT-hetero-Diels–Alder (HDA)
reaction, thiol-yne click reaction and an esterification
www.MaterialsViews.com
Macromol. Rapid Commun.
� 2011 WILEY-VCH Verlag Gmb
step to yield polymers with four alkyne functionalities.
The linear block copolymer was prepared from poly(tert-
butyl acrylate)-block-poly(2-hydroxyethyl acrylate)
(P(tBA)-b-PHEA) synthesised via RAFT followed by ester-
ification to obtain a block copolymer with four alkyne
functionalities. Subsequent functionalisation with b-
thioglucose sodium salt leads to two different polymer
architectures with eight glucose functionalities each
(Scheme 1).
Experimental Section
See Supporting Information for detailed experimental procedure
and analysis.
Results and Discussion
The synthesis of the dendritic structure required an alkyne
endfunctional polymer. While direct synthesis using an
alkyne functional RAFT agent is possible, the postmodifica-
tion using a new bi-functional clicking agent was found to
be a convenient pathway to high endfunctionality. The
unique bi-functional clicking agent was synthesised in a 1-
step reactionwith2,4-trans,trans-hexadien-1-ol (HPYS) and
the 4-oxo-4-(prop-2-ynyloxy)butanoic anhydride (acety-
lene anhydride) as show in Scheme S1 of Supporting
Information. This bi-functional clicking agent can react
directly with the RAFT endfunctional terminal polymer via
HDAreaction converting theRAFTgroup inanefficientway
into an alkyne group.
The synthesis of the dienophile end-chain capable P(tBA)
was achieved by RAFT polymerisation using BPDF as the
RAFT agent and AIBN as a thermal initiator. The molecular
weightwasdeterminedvia 1HNMRandGPC.Twopolymers
of Mn (a) 1 200 g �mol�1 (PDI¼ 1.24/DPn ¼ 7) and (b)
5 600 g �mol�1 (PDI¼ 1.13/DPn ¼ 42) (SEC traces shown in
Figure S1 of Supporting Information) were obtained and
used in further experiments. The purpose of the polymer
with Mn of 1 200 g �mol�1 is only to follow the progress of
the reaction pathway via electrospray ionization mass
spectrometry (ESI-MS) so as to provide further structural
confirmation in addition to 1H NMR through the various
stages of functionalisation.
The two polymers were then reacted with HPYS via
an HDA reaction. TFA acts as a catalyst, and this reaction
was carried out at ambient temperature under vigorous
stirring in chloroform. As the reaction proceeded,
the solution lost its original pink colouration due to
the RAFT end-groups and became colourless due to the
formation of the 3,6-dihydro-2H-thiopyran ring. The
HPYS-P(tBA) conjugates were reacted with thioglycerol
via a thiol-alkyne conjugation reaction via a radical
pathway. The reactants were dissolved in DMF along
2011, 32, 1620–1626
H & Co. KGaA, Weinheim1621
S
SNn
S
SN
O O
Sn S
N
O
OO
OO
OO
O
AIBN, Toluene65 ºC
TFAChloroform25 ºC
Sn S
N
O
OO
O
SS
OHOH
HOOH
DMPA, DMF,N2 DegassedUV (365nm), 4 hr
HSOH
OH
SnS
N
O
OO
O
SS
OO
OO
O
O
O
O
OO
OO
O O
O O
DMAP, TEA,45 ºC, 18 hr,Chloroform
OO
OO
O
OO
(I)(II)
(III)(IV)
O O O O
OO
O O
OO
O
OHOH
OH
HO
SO
OHHO
OH
HO
S
OO
OS
OO
O OHOH
HO
HO
S
O
OHOH
OH
HO
SO
S OO
O OHOHHO
OH
S
OOHOH
HO
HO
SO
O
O
OOO
OHOH
HOOH
S
OOHHO
OH
S
O
O
OO
O
S
S N
O O
OH
HO
H
HO
H
HOHH SH
OH
DMPA, DMF,N2 DegassedUV (365nm), 4 hr
O O
OO
OH
S S
S
OH
O
OO O
OH
O
OO
OO
OO
O
O
OO
DMAP, TEAChloroform, 50 ºC, 24 hr
65 ºC, AIBN, Toluene
65 ºC, AIBN,DMAc
1)
2)
S S
S
OH
On m OO O
O
O
S S
S
OH
On m
O
OO
OO O
O
O
S S
S
OH
On m
OH
HO
HHO
H
HOHH SH
OH
DMPA, DMF,N2 DegassedUV (365nm), 4 hr
O
H OH
H
OHH
HOH
H
S
HO
O
H
HO HHO
H HOHH
SH
HO
Scheme 1. Reaction pathway to generate 4-alkyne functionalised P(tBA)-based polymer dendron and the alkyne functionalised linear blockcopolymer P(tBA)-b-PHEA and the subsequent modification to structures with eight glucose units.
1622
www.mrc-journal.de
J. Kumar, A. Bousquet, M. H. Stenzel
withasmall amountofDMPA, aphoto-initiatorand reacted
for 2h at 30 8C under UV light at a wavelength of 365nm.
Once this reaction was completed, the product was
precipitated in 50:50 vol% water/methanol to eliminate
all remnant thioglycerol and DMPA and then dried under
vacuum. Since thiols react with alkynes in a 2–1 ratio,
Macromol. Rapid Commun.
� 2011 WILEY-VCH Verlag Gmb
the HPYS-P(tBA) conjugates would have two thioglycerol
units attached terminally, thus affording four hydroxyl
groups at the chain ends.
Subsequent reaction 4-oxo-4-(prop-2-ynyloxy)butanoic
anhydride in the presence of DMAP and TEA over 18h at
45 8C afforded product (IV) (Scheme 1). ESI-MS of the lower
2011, 32, 1620–1626
H & Co. KGaA, Weinheim www.MaterialsViews.com
Figure 1. ESI-MS overlay of (1) both experimental (E), and theoretical (T) masses as well as the degree of polymerisation (DPn) are reported forNaþ charged species. Hþ and double charged (with Naþ) species are denoted by a and b respectively. E: experimental, T: theoretical.
Thiol-alkyne Chemistry for the Preparation of Micelles with Glycopolymer Corona: . . .
www.mrc-journal.de
Mn polymer was taken at various stages of functionalisa-
tion to serve as structural confirmation of the desired
product (Figure 1). The calculated mass was observed to be
close to the theoreticalmass. Other observedpopulations in
the ESI-MS spectra were due to double charged species.
Further structural confirmation via NMR is shown in
Figure 2 for all stages of functionalisation [structures (I)–
(IV) in Scheme 1].
As a control experiment, a linear version of the dendritic
polymer was synthesised. This is achieved by RAFT
polymerisation of tert-butyl acrylate using BPATT as the
RAFT agent and AIBN as a thermal initiator. The product
polymer was analysed for its molecular weight via 1H NMR
and GPC and is found to be 5900 g �mol�1 (PDI¼ 1.15/
DPn ¼ 46). P(tBA)-b-PHEA was then generated via a chain
extension of the homopolymer. The molecular weight
distribution of the P(tBA) homopolymer as well as the chain
extended copolymer is shown in Figure S2 of Supporting
Information. From 1H NMR, the polymer was found to have
www.MaterialsViews.com
Macromol. Rapid Commun.
� 2011 WILEY-VCH Verlag Gmb
anaverageof4.5HEAunitsperpolymer chainby integrating
the two protons (3.80ppm) next to the hydroxyl moieties
against the integraloftheprotonsofthe tert-butylgroups(2–
2.4ppm) as illustrated in Figure S3 of Supporting Informa-
tion. The polymer was then functionalised using a similar
procedure as outlines above resulting a block copolymer
with on average 4.5 alkyne functionalities per chain.
Glycosylation of the linear control as well as the
dendrimer bwas performed via a thiol-alkyne conjugation
process with a b-thioglucose sodium salt. The salt, HCl,
polymer and a small amount of DMPAphoto-initiatorwere
dissolved in DMF. HCl is required to protonate the glucose
salt for the purposes of this reaction. The mixture was
degassed viaN2 sparging and then reacted under UV light at a
wavelength of 365nm for a period of 4h. The product was
dialysed against 50:50 vol% water/methanol. The products,
P(tBA)44-dentritic-Glucose8andP(tBA)46-block-P(HEA-glucose)9,
were analysed via 1H NMR to verify the correct structure
(Figure3andFigure S5of Supporting Information). Only the
2011, 32, 1620–1626
H & Co. KGaA, Weinheim1623
8 7 6 5 4 3 2 1
D
O
OO
O
S
S N
O O
S
S
OO
O
O
OO
O
O
OO
O
O
OO
O
O
O
HOH
H
OH
H
HOH
H
HO
S OH
OH
HOH
H
HHO H
S
OH
OH
OH
H
OHH
HHO HS
HO
O
H
OHH OH
HHHO
H
S
OH
O
H
HOH OH
HHHO
H
S
OHO
HHO
H
OH
H
HHO
H
S
OHO
HHO
H
HO H
HOH
HS
OH
O
HHO
HOH
H
HHO
H
S
OH
AB
C
1 2 3 4 5 67
8E
F G
91011
* impurity from commercial thioglucose
*
Glucose
βDMF
A B G3,10
E
2,7,9
6F
DMFH2O
1,4,5,8,11
D
C
δ / ppm
Figure 3. 1H NMR (d-DMF) structural confirmation of P(tBA)-based dendron with eightglucose functionalities.
8 6 4 2 0
O
OO
O
S
S N
O O
S
S
OO
O
O
OO
O
O
OO
O
O
OO
O
O
F
F,K
3
J
E,I,Ha',12
a
B
32
1E
KJIHGF
GE,I,H
J,KK
JIHGFO
OO
O
S
S N
O O
S
S
OHHO
OH
OH
C AE,G,a'D,F
B
GFE
DCBA
a'a'
O
OO
O
S
S N
O O
n
CH-S ce
d S
N
O O
n
S
bbbba
a
bc
d
e
δ/ ppm
Figure 2. 1H NMR in CDCl3 of the dendritic structures based on P(tBA) with Mn ¼5 600 g �mol�1 providing structural confirmation of the products at the four stagesof dendron synthesis.
1624
www.mrc-journal.de
J. Kumar, A. Bousquet, M. H. Stenzel
anomeric glucose peak (b in Figure 3) is visible, the
remaining glucose signals are overridden by the water
signal. 1H{13C}HSQC NMR spectra (Figure S4 of Supporting
Information) reveal the missing glucose signals and
especially the H-6a and H-6b are clearly visible.
Macromol. Rapid Commun. 2011, 32, 1620–1626
� 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinhe
Micelles were prepared by dissolving
the polymers P(tBA)44-dentritic-Glucose8and P(tBA)46-block-P(HEA-glucose)9 in
acetone, followed by the slow addition
of deionised water and the subsequent
removal of acetone via dialysis. The
solutions were assessed for their particle
size by dynamic light scattering, and
their size and morphology via TEM. The
TEM images (Figure S6 of Supporting
Information and insert in Figure 4) and
the number size distribution of the
dynamic light scattering (Figure S7 of
Supporting Information) revealanumber
mean particle diameters via DLS of
133nm for the polymer dendron, and
124nm for the linear control polymer.
The low fraction of glycopolymer in the
block copolymer will result in crew-cut
aggregate, which have a brush-like gly-
copolymer configuration.
Lectin binding assays with Concana-
valin A (ConA) were then carried out to
fully assess and compare the ligation
characteristics of the two different poly-
mers. Two types of studies are performed
to assess two aspects of binding. The first
was a turbidimetry assay to evaluate the
rate at which the glycopolymer binds to
ConA and the extent of interchain cross-
linking due to binding. This analysis was
carried out by mixing a solution of ConA
in HEPES buffered saline into the micelle
solution of both the polymers. The UV
absorbance (at l¼ 420nm) of the solu-
tions was measured against time over
10min (Figure 4).
From the turbimetry assay, an initial
clustering rate constant, K can be
obtained. In the case of the polymer with
the glycodendritic end group, the value is
0.18 AU �min�1 while the value for the
linear glycopolymer is 0.05 AU �min�1. It
needs to be considered that there is an
error associated with this calculation
since significant turbidity can be
observed by the time the solutions are
mixed and placed in the UV/Vis spectro-
meter. The polymer with the dendrimer
end-group has a significantly higher K value despite the
similar overall glucose residue for both solutions.[14] It is
also evident that there is intermolecular cross-linking due
to the ConA binding from the increase in particle size as
obtained from the dynamic light scattering. The number
im www.MaterialsViews.com
Figure 4. Absorbance versus time turbidimetry assays of both the dendrimerb and linear polymermicelle solutionswhen a Con A solution isintroduced.
Figure 5. Correlation of ConA bound to glucose at various glucoseresidues for both the polymer with the dendritic end-group andlinear glycopolymer.
Thiol-alkyne Chemistry for the Preparation of Micelles with Glycopolymer Corona: . . .
www.mrc-journal.de
average particle diameters for the polymer with dendritic
corona and the linear polymer after binding are 269 and
573nm, respectively (Figure S7 of Supporting Information).
The bigger diameter is equivalent to more cross-linking
between the micelles, which is often observed when
binding is slower.[15] Turbidity assay determines the rate
of binding while the precipitation assay quantifies the
number of ConA bound per glucose residue. An excess of
ConA solution was added to solutions with different
polymer concentrations and allowed to bind at an
incubation temperature of 20 8C for 20h. The precipitate
was isolated and the amount of bound ConAwasmeasured
using UV/Vis spectroscopy (at l¼ 280nm). The correlation
bound ConA glucose concentration reveals that the
polymer dendron exhibits better binding characteristics
to ConA with more ConA bound to it as compared to the
linear polymer for similar glucose residues (Figure 5).
In summary, RAFT polymerisation in combination with
RAFT-HDA click chemistry and thiol-yne click chemistry
were used as efficient synthetic procedures to amphiphilic
blockcopolymerstructures, onewitheightglucosemoieties
in a dendritic arrangement and one with nine glucose
units in a linear architecture. Micellisation of both block
copolymers led to micelles with a dendritic glycopolymer
surface and a linear glycopolymer brush structure,
respectively. The turbidity assay and the precipitation
assay found more efficient binding of the dendritic
structure.
www.MaterialsViews.com
Macromol. Rapid Commun.
� 2011 WILEY-VCH Verlag Gmb
Acknowledgements: The authors like to acknowledge theAustralian Research Council for funding and like to thank thestaff from the UNSW NMR facility for their help.
Received: May 17, 2011; Revised: June 21, 2011; Published online:August 19, 2011; DOI: 10.1002/marc.201100331
2011, 32, 1620–1626
H & Co. KGaA, Weinheim1625
1626
www.mrc-journal.de
J. Kumar, A. Bousquet, M. H. Stenzel
Keywords: dendrimers; glycopolymer; lectin; micelles; thiol-yneclick
[1] a) S. R. S. Ting, G. Chen, M. H. Stenzel, Polym. Chem. 2010,1, 1392; b) S. G. Spain, N. R. Cameron, Polym. Chem. 2011, 2, 60;c) A. David, Israel J. Chem. 2010, 50, 204.
[2] a) B. Voit, D. Appelhans, Macromol. Chem. Phys. 2010, 211,727; b) S. G. Spain, M. I. Gibson, N. R. Cameron, J. Polym. Sci.,Part A: Polym. Chem. 2007, 45, 2059; c) Y. Chabre, R. Roy, Curr.Top. Med. Chem. 2008, 8, 1237.
[3] a) L. Tao, G. Chen, G. Mantovani, S. York, D. M. Haddleton,Chem. Commun. 2006, 4949; b) L. Tao, J. Geng, G. Chen, Y. Xu,V. Ladmiral, G. Mantovani, D. Haddleton, Chem. Commun.2007, 3441; c) L. M. Van Renterghem, M. Lammens, B.Dervaux, P. Viville, R. Lazzaroni, F. Du Prez, J. Am. Chem.Soc. 2008, 130, 10802; d) A. Agarwal, A. Asthana, U. Gupta,N. K. Jain, J. Pharm. Pharmacol. 2008, 60, 671; e) A. Martin,B. Li, E. Gillies, J. Am. Chem. Soc. 2009, 131, 734.
[4] M. J. Cloninger, Curr. Opin. Chem. Biol. 2002, 6, 742.[5] U. Boas, P. M. H. Heegaard, Chem. Soc. Rev. 2004, 33, 43.[6] a) W. Binder, R. Sachsenhofer, Macromol. Rapid Commun.
2008, 29, 952; b) C. N. Urbani, C. A. Bell, M. R. Whittaker,M. J. Monteiro, Macromolecules 2008, 41, 1057; c) C. Urbani,C. Bell, D. Lonsdale, M.Whittaker, M.Monteiro,Macromolecules2007, 40, 7056; d) M. Whittaker, C. Urbani, M. Monteiro, J. Am.
Macromol. Rapid Commun.
� 2011 WILEY-VCH Verlag Gmb
Chem. Soc. 2006, 128, 11360; e) J.Wiltshire, G. Qiao, J. Polym. Sci.,Part A: Polym. Chem. 2009, 47, 1485.
[7] a) A. B. Lowe, Polym. Chem. 2010, 1, 17; b) A. B. Lowe, M. A.Harvison, Aust. J. Chem. 2010, 63, 1251.
[8] P. Antoni, D. Nystrom, C. J. Hawker, A. Hult, M. Malkoch,Chem. Commun. 2007, 2249.
[9] R. A. Evans, Aust. J. Chem. 2007, 60, 384.[10] G. Chen, S. Amajjahe, M. H. Stenzel, Chem. Commun. 2009,
1198.[11] P. Antoni, Y. Hed, A. Nordberg, D. Nystrom, H. von Holst,
A. Hult, M. Malkoch, Angew. Chem., Int. Ed. 2009, 48, 2126.[12] a) S. R. S. Ting, E. H. Min, P. B. Zetterlund, M. H. Stenzel,
Macromolecules 2010, 43, 5211; b) M. Hetzer, G. J. Chen,C. Barner-Kowollik, M. H. Stenzel, Macromol. Biosci. 2010,10, 119.
[13] a) Z. S. Ge, S. Z. Luo, S. Y. Liu, J. Polym. Sci., Part A: Polym. Chem.2006, 44, 1357; b) Z. Ge, D. Chen, J. Zhang, J. Rao, J. Yin,D. Wang, X. Wan, W. Shi, S. Liu, J. Polym. Sci., Part A: Polym.Chem. 2007, 45, 1432; c) D. L. Patton, P. Taranekar, T. Fulghum,R. Advincula, Macromolecules (Washington, DC, US) 2008, 41,6703; d) R. Vestberg, A. M. Piekarski, E. D. Pressly, K. Y. VanBerkel, M. Malkoch, J. Gerbac, N. Ueno, C. J. Hawker, J. Polym.Sci., Part A: Polym. Chem. 2009, 47, 1237; e) J. Xu, C. Boyer,V. Bulmus, T. P. Davis, J. Polym. Sci., Part A: Polym. Chem. 2009,47, 4302.
[14] V. Ladmiral, G. Mantovani, G. J. Clarkson, S. Cauet, J. L. Irwin,D. M. Haddleton, J. Am. Chem. Soc. 2006, 128, 4823.
[15] C. W. Cairo, J. E. Gestwicki, M. Kanai, L. L. Kiessling, J. Am.Chem. Soc. 2002, 124, 1615.
2011, 32, 1620–1626
H & Co. KGaA, Weinheim www.MaterialsViews.com