Z-group ketone chain transfer agents for RAFT polymer nanoparticle modification via hydrazone conjugation

Article abstract of DOI:10.1021/ma301536f

A ketal-containing trithiocarbonyl compound has been synthesized and characterized as a chain transfer agent (CTA) in reversible addition- fragmentation transfer (RAFT) polymerization. The ketal functionality does not interfere with RAFT polymerization of acrylate monomers, which proceeds as previously reported to yield macro-CTA polymers and block copolymers. Postpolymerization ketal cleavage revealed ketone functionality at the polar terminus of an amphiphilic block copolymer. Hydrazone formation was facile in both organic solution and aqueous buffer where polymer nanoparticle assemblies were formed, indicating a conjugation/end-functionalization yield of 40-50%. Conjugation was verified with fluorescein, biotin, and Gd-DOTA derivatives, and though the trithiocarbonate linkage is hydrolytically labile, we observed stable conjugation for several days at pH 7.4. and 37 °C. As expected, streptavidin binding to biotinylated polymer micelles was observed, and size-change-based relaxivity increases were observed when Gd-DOTA hydrazide was conjugated to polymer micelles. Cell uptake of fluorescently labeled polymer micelles was also readily tracked by FACS and fluorescence microscopy. These polymer derivatives demonstrate a range of potential theranostic/ biotechnological applications for this conveniently accessible keto-CTA, which include ligand-based nanoparticle targeting and fluorescent/MR nanoparticle contrast agents.

Full text of DOI:10.1021/ma301536f

Article
pubs.acs.org/Macromolecules
Z‑Group Ketone Chain Transfer Agents for RAFT Polymer
Nanoparticle Modification via Hydrazone Conjugation
Saibal Bandyopadhyay,† Xin Xia,† Andrei Maiseiyeu,‡ Georgeta Mihai,‡ Sanjay Rajagopalan,‡
and Dennis Bong*,†
†Department of Chemistry, 100 W. 18th Avenue, The Ohio State University, Columbus, Ohio 43210, United States
‡The Davis Heart and Lung Institute, 473 W. 12th Avenue, The Ohio State University, Columbus, Ohio 43210, United States
S
* Supporting Information
ABSTRACT: A ketal-containing trithiocarbonyl compound has been
synthesized and characterized as a chain transfer agent (CTA) in reversible
addition−fragmentation transfer (RAFT) polymerization. The ketal
functionality does not interfere with RAFT polymerization of acrylate
monomers, which proceeds as previously reported to yield macro-CTA
polymers and block copolymers. Postpolymerization ketal cleavage
revealed ketone functionality at the polar terminus of an amphiphilic
block copolymer. Hydrazone formation was facile in both organic solution
and aqueous buffer where polymer nanoparticle assemblies were formed,
indicating a conjugation/end-functionalization yield of 40−50%. Conjugation was verified with fluorescein, biotin, and Gd-
DOTA derivatives, and though the trithiocarbonate linkage is hydrolytically labile, we observed stable conjugation for several
days at pH 7.4. and 37 °C. As expected, streptavidin binding to biotinylated polymer micelles was observed, and size-change-
based relaxivity increases were observed when Gd-DOTA hydrazide was conjugated to polymer micelles. Cell uptake of
fluorescently labeled polymer micelles was also readily tracked by FACS and fluorescence microscopy. These polymer derivatives
demonstrate a range of potential theranostic/biotechnological applications for this conveniently accessible keto-CTA, which
include ligand-based nanoparticle targeting and fluorescent/MR nanoparticle contrast agents.
ccessible strategies for controlled radical polymerization
of serum proteins, opsonization, and immunoclearance.7 These
surface protected assemblies have been shown to have longer in
vivo circulation times and altered tissue deposition relative to
their undecorated counterparts.8 The use of appropriately
functionalized CTAs has thus been used to synthesize polymer
micelles that surface-display poly(ethylene oxide) (PEO) in
order to take advantage of the immunoevasive properties
imparted by surface presentation of a hydrophilic polymer.5
Biotin-derivatized RAFT CTAs9 or ATRP initiators10 have
been used to enable protein conjugation to polymer scaffolds.
There have also been numerous reports of incorporation of
thiol groups as polymer-end nucleophiles to couple with
electrophilic disulfides or halides for postassembly functional-
ization or transformation of the thiol into other functionalitie-
s.1a,11 This is especially facile with RAFT polymers terminated
with a trithiocarbonate moiety that may be easily transformed
into a thiol-terminated polymer. Thus, through the use of
derivatized RAFT agents, carboxy, amino, thiol, azide, and more
complex metal ligands, fluorescent and biomolecular tags have
been incorporated selectively at the terminus of a RAFT
polymer.1a,12 While thiol chemistry provides a generally useful
platform for further polymer modification, the thiol itself is
prone to oxidation under ambient conditions; further, addi-
1
and ATRP2 have enabled a dramatic
Asuch as RAFT
expansion in the number of applications for which synthetic
polymers may play a role. The chain transfer agent (CTA) in
RAFT polymerization facilitates block copolymer synthesis,
controls polydispersity, and presents an opportunity for
polymer end-functionalization. Functionality pendant to the
CTA will be presented at a polymer terminus once polymer-
ization has been halted. This aspect of controlled radical
polymerization has been previously exploited to end-function-
alize polymers and, by extension, surface-functionalize polymer
assemblies.1c,2b,3 End-functionalization strategies of linear
polymers have minor impact on packing interactions along
the polymer strand4 relative to random side-chain functional-
ization and are thus less likely to affect polymer assembly. By
choosing the order of monomer use in an amphiphilic
copolymer RAFT synthesis, one may determine the function-
ality at each termini. Thus, modification at the terminus of the
polar block should result in presentation of that functionality
on the surface of an aqueous-phase assembly, while the same
functionality would be buried when assembly occurs in
nonpolar media.5
Surface derivatization of polymer assemblies such as micelles
is often desirable for biomedical applications, in order to
promote biomolecular interactions or to block them.3
Installation of poly(ethylene glycol) (PEG) on the surface of
nanoparticles and liposomes6 is well-known to reduce binding
Received: July 23, 2012
Revised: August 22, 2012
Published: August 30, 2012
© 2012 American Chemical Society
6766
dx.doi.org/10.1021/ma301536f | Macromolecules 2012, 45, 6766−6773

Macromolecules
Article
tional chemoselective functionality would broaden the utility of
this method. With this in mind, we designed and synthesized a
Z-group acetal-derivatized trithiocarbonate chain transfer
RAFT agent that would yield a ketone-terminated RAFT
polymer.
Z-group end-functionalization further avoids competitive
reinitation by AIBN-derived radicals.19 We demonstrate the
utility of this approach by surface-labeling keto polymer
assemblies with hydrazides bearing fluorophores, MRI contrast
agents, and biotin. As expected, we observe fluorescence
anisotropy, relaxivity increases, and streptavidin binding
consistent with modification of the polymer assemblies.
Furthermore, polymer nanoparticle uptake in A549 and
MCF-7 cell lines was tracked with an end-labeled fluorescent
polymer. Thus, this range of conjugation chemistry demon-
strates how this keto-RAFT agent could be applied in a wide
range of targeting and theranostic applications.8a
Ketone reaction with hydrazine and aminooxyethers to form
hydrazone and oxime linkages has considerable precedent as a
bioconjugation strategy.13 Like disulfides, these stabilized
imines are exchange-active covalent bonds that are hydrolyti-
cally labile under acidic conditions, whereas disulfides are labile
under reducing conditions. This property has been previously
exploited to produce hydrazone/oxime linked carrier systems14
are stable at normal physiological pH but will release cargo
once acidic pH is encountered, such as within an endosomal
compartment or the slightly acidic solid tumor microenviron-
ment. Hydrazone/oxime formation may be carried out in either
organic or aqueous conditions and is remarkably chemo-
selective, with no alternative reaction outcomes observed with
either the amine or keto components under normal
physiological conditions. Thus, polymer termination with
keto functionality may be derivatized by any synthetic
hydrazine or aminooxyether containing molecule, in aqueous
or organic milieu, with minimal concern for polymer self-
reaction (as with thiol oxidation). Though postpolymerization
ketone modification requires additional handling, prior
installation of hydrophilic groups on the CTA may lead to
poor chain extension with hydrophobic monomers due to
solubility issues.5,15 Recently, Maynard16 and Fulton17 have
demonstrated the polymer end-incorporation of aminooxyether
and aldehyde groups, using RAFT chain transfer agents
derivatized at the reinitiator site. These methods both utilized
S-alkyl-S′-(α,α-dimethyl-α″-acetic acid)trithiocarbonate18 chain
transfer agents with aminooxyether or aldehyde containing
groups amidated to the dimethyl acetic acid moiety, which
serves as the “R” group reinitiator on the RAFT CTA. We
report herein a complementary synthetic approach that places a
keto group on the “Z” group of the CTA, which does not
dissociate from the trithiocarbonate during radical polymer-
ization.
There are many appealing aspects of placing the chemo-
selective “handle” on the reinitating “R” group: the handle is
connected to the polymer via a hydrolytically stable carbon−
carbon bond, and the “Z” group may then be converted to the
chemically versatile thiol group or easily removed.1a,11a
Preparation of the dimethylacetic acid CTAs does require
multistep synthesis, albeit straightforward, with a range of yields
from high to low.18 We demonstrate herein that the “Z” group
functionalization remains chemically stable on polymer
assemblies over days at physiologically relevant conditions of
pH 7.4 and 37 °C, suggesting that the trithiocarbonate linkage
is sufficiently stable to be useful in bioapplications. Our
approach to the Z-keto CTA is a flexible, high yielding
approach (Scheme 1) that permits variation in the R group,
which may be convenient or necessary for some monomers.1a,d
EXPERIMENTAL SECTION
■
General. All commercially available materials were purchased from
Sigma-Aldrich. GdDOTA hydrazide was synthesized following
literature procedures.14a Monomers were purified using an alumina
column to remove inhibitor and stored at −20 °C. AIBN was
recrystallized from methanol. All polymerization solvents were
degassed prior to use under a nitrogen atmosphere. Concentration
of 1a and ratio of 1a:AIBN was held constant in all polymerizations.
Theoretical molecular weight (Mw,t) was calculated according to Mw,t
=
(nBAmw[nBA]C)/[1a] where nBAmw is the molecular weight of repeat
unit and C is conversion. NMR and mass spectra were acquired on a
400 MHz Bruker Avance instrument and a Bruker MicroTOF ESI,
respectively. Fluorescence anisotropy measurements were acquired at
25 °C on a PerkinElmer LS-50B fitted with polarizers. Transmission
electron microscopy samples were prepared on 200 mesh Formvar-
carbon film copper grids, stained with uranyl acetate and analyzed on a
FEI Tecnai G2 Bio TWIN TEM.
Relaxivity Measurements. Relaxivities (r1 longitudinal) were
determined as the slope of the curve obtained by plotting inverse
relaxation (1/T1) against the range of gadolinium concentration.
Measurements were conducted using a 1.5 T scanner at 25 °C
(Magnetom, Avanto MRI scanner, Siemens Medical Solutions,
Erlangen, Germany). Samples (100 μL) were prepared based on Gd
concentrations (2−0.03 mM) and analyzed in a sealed polypropylene
96-well plate in a styrene box filled with water. Relaxation curves for T1
were obtained by 2D imaging with an inversion−recovery turbo spin-
echo (IRTSE) pulse sequence. The following inversion times were
applied: 30, 60, 90, 120, 150, 250, 400, 600, 800, 1200, 1600, and 2000
ms. Relaxation times (r1) were calculated from each slope of the linear
relation between 1/T1 and Gd concentration. Gadolinium concen-
tration was determined by inductively coupled plasma optical emission
spectroscopy (ICP-OES).
Cell Uptake. Human breast adenocarcinoma (MCF-7) and human
lung carcinoma (A549) cell lines were maintained in GIBCO
Dulbecco’s Modified Eagle Medium: Nutrient Mixture F-12
(DMEM/F-12) media with 10% FBS and 1% penicillin−streptomycin.
All FACS samples were run on a LSR-II flow cytometry instrument
using FITC-A and PE-A lasers, and the results were analyzed using
FlowJo software. Cell microscopy images were acquired on an Zeiss
Axio Observer A1.
Synthesis of Benzyl 2-(2-Methyl-1,3-dioxolan-2-yl)-
ethyltrithiocarbonate (1a). S-Acetyl-4-mercapto-2-butanone (1.46
g, 0.01 mol) is commercially available but was readily prepared from
Michael addition of CH3COSK to methyl vinyl ketone20 and
converted to the acetal by reaction with ethylene glycol (2.48 g,
0.04 mol) and D-tartrate (0.05 g) in refluxing THF for 12 h using a
Dean−Stark apparatus. The reaction mixture was concentrated under
reduced pressure, dissolved in 80 mL of CH2Cl2, washed with
saturated aqueous NaHCO3, dried over anhydrous Na2SO4, and
concentrated under reduced pressure to give the desired acetal21 in
quantitative yield. The thiol ester (0.5 g, 2.63 mmol) was hydrolyzed
under nitrogen in 1:1 methanol/0.2 M NaOH solution (30 min), then
treated with sodium bicarbonate (0.22 g, 2.6 mmol), CS2 (0.27 mL,
4.5 mmol), and benzyl bromide (0.53 mL, 4.5 mmol), and then stirred
overnight. The reaction mixture was purified by silica gel
a
Scheme 1. Synthesis of 1a
a
Reagents and conditions: (a) D-tartrate, ethylene glycol, Dean−Stark
THF reflux; (b) NaOH(aq); (c) CS2, benzyl bromide.
6767
dx.doi.org/10.1021/ma301536f | Macromolecules 2012, 45, 6766−6773

Macromolecules
Article
chromatography (15% EtOAc/hexanes) following aqueous work-up
(CH2Cl2/H2O) to give pure 1a as a yellow liquid in 78% overall yield.
1H NMR (CDCl3) δ: 7.26 (5H, m), 4.55 (2H, s), 3.90 (4H, m), 3.39
(2H, m,), 2.01 (2H, m), 1.31 (3H, s). 13C NMR (CDCl3) δ: 223.9,
135.3, 129.4, 128.8, 127.9, 109.1, 65.0, 41.5, 37.3, 31.8, 24.2. HRMS
(ESI): m/z calcd for C14H18O2S3Na [M + Na]+ 337.0366. Found
337.0404.
Benzyl-3-oxobutyltrithiocarbonate (1b). 1a (0.04 g) was
stirred in methanol with 0.5% TFA at room temperature for 1 h.
Solvent was removed, and reaction mixture was redissolved in
dichloromethane and washed with NaHCO3(sat). After concentration
1b was obtained as yellow oil in quantitative yield. 1H NMR (CDCl3)
δ: 7.34 (5H, m), 4.62 (2H, s), 3.59 (2H, t, JHH = 6.4 Hz), 2.93 (2H, t,
JHH = 6.4 Hz), 2.18 (3H, s). 13C NMR (CDCl3) δ: 223.8, 206.2, 135.3,
129.5, 128.9, 128.0, 42.3, 41.7, 30.4, 30.1. HRMS (ESI): m/z calcd for
C12H14OS3Na [M + Na]+ 293.0104. Found 292.9182.
Purification of Hydrazone−Nanoparticle Suspensions. Crude
conjugation reaction mixtures in methanol (0.002 g) were dried to
polymer films as described above then hydrated with 1 mL of DPBS
buffer and sonicated to produce polymer nanoparticles, and pH
adjusted to 7.6 with NaHCO3 saturated solution as necessary. Polymer
nanoparticle suspensions were washed with DPBS (1 mL/10 min) in a
centricon size exclusion filter appropriate for the polymer length (e.g.,
10 and 30 kDa exclusion for conjugates with 6 kDa Pa(30,28) and 19.5
kDa Pa(100,89), respectively). Centricon filtration was performed at
4000 rpm for 2 h.
RESULTS AND DISCUSSION
■
The following studies focus on validation of the acetal-
derivatized trithiocarbonate as an effective chain transfer
agent for RAFT polymerization and quantification of accessible
ketone-terminated polymer. The general polymer architecture
is an amphiphilic block copolymer of n-butyl acrylate and
acrylic acid, PnBA-b-PAA, with free ketone terminating the
PAA block. Similar amphiphilic copolymers with poly(acrylic
acid) have been found to form “kinetically” frozen micelles of
∼200 monomer length;22 we have synthesized polymers with
∼50 mol % acrylic acid with total length of ∼200, 100, and 60
Hydrazone Formation (Small Molecule). Ketone 1b (0.010 g,
0.037 mmol) was dissolved in methanol with biotin hydrazide 3 (0.010
g, 0.047 mmol) followed by a catalytic amount (5 μL) of acetic acid.
The resulting reaction mixture was stirred at room temperature for 1 h
and concentrated under reduced pressure to give pale yellow solid
1
which was dissolved in d6-DMSO and analyzed by H NMR. HRMS
(ESI): m/z calcd for C22H30N4NaO2S4 [M + Na]+ 533.1149. Found
533.1145.
monomers, which we term Pb(100,89), Pb(49,48), and Pb(30,28)
,
General Procedure for Polymer−Hydrazone Conjugation.
PnBA-b-PAA (Pb(m,n), 0.012 mmol) was dissolved in 10 mL of
methanol and CH2Cl2 (3:1) mixture followed by the addition of
hydrazide (or semithiocarbazide) derivative (0.011 mmol) and 10 μL
of TFA. The resulting reaction mixture was stirred for 8 h, then
concentrated under reduced pressure, and dried under high vacuum to
yield a solid product that could be redissolved in dry methanol.
Polymerization Kinetics. n-Butyl acrylate (2.92 g, 22.8 mmol)
and 1a (0.03 g, 0.095 mmol) were dissolved in 3 mL of DMF and
heated to 70 °C in a sand bath under nitrogen, with DMSO (1.0 mL)
added as an internal standard. Polymerization was initiated with the
addition of AIBN (1.57 mg, 0.0095 mmol). Reaction aliquots were
collected every 10 min and quenched by cooling in dry ice acetone
bath under ambient atmosphere. Time point samples were analyzed
respectively (Figure 1 and Table 1). We further demonstrate
polymer−hydrazide conjugation and nanoparticle display of a
fluorescent tag, protein-targeting ligand, and MRI-contrast
agent.
1
using the combination of H NMR spectroscopy and gel permeation
chromatography (GPC/SEC).
General Polymerization Procedure (Pa(30,0)). As above, CTA 1a
(0.12 g, 0.38 mmol) and monomer n-butyl acrylate (2.92 g, 22.8
mmol) were dissolved in 25 mL of DMF with a DMSO (1.06 g)
internal standard and heated to 70 °C. Polymerization was initiated by
addition of AIBN (6.0 mg, 0.038 mmol) and quenched after 53%
conversion. Unreacted monomer and solvent were removed under
reduced pressure to obtain the macro-CTA, PnBA, as an oily liquid
that was analyzed by GPC to obtain the following parameters:
Mn(GPC) = 4210, Mw/Mn = 1.19 (theoretical molecular weight Mw,t
=
4390).
Figure 1. (top left) Chain transfer agents used with reinitiator “R” and
“Z” groups indicated. General polymer architectures and hydrazine
derivatives used in polymer and CTA conjugation are shown.
General Procedure for Block Copolymer Synthesis (Pa(30,28)).
In a 50 mL round-bottom flask tert-butyl acrylate (2.92 g, 22.8 mmol)
and macro-CTA (Pa(30,0), 1.0 g, 0.153 mmol) were dissolved in 23 mL
of DMF and heated to 70 °C after thorough degassing. After stirring
for 5 min, polymerization was initiated with the addition of AIBN
(0.006 g). Reaction was followed by H NMR spectroscopy and was
quenched after 53% monomer was consumed. Theoretical molecular
weight = 8455, Mn(GPC) = 7797, Mw/Mn = 1.21.
General Procedure for Polymer Acidolysis (Pb(m,n)). Block
copolymer (Pa(m,n)) was dissolved in 2 mL of methylene chloride with
8 mL of TFA and stirred for 3 h. Volatiles were removed under
reduced pressure, and the resulting solid was dissolved in CH2Cl2 and
washed with 1 N HCl to the organic layer concentrated to yield
Pb(m,n); the block copolymer as a pale yellow solid.
General Procedure for Polymer Nanoparticle Formation. A
methanolic solution of Pb(m,n) block copolymer (0.002 g) was placed
in a test tube and methanol was removed under nitrogen stream to
form polymer film which were subsequently dried under vacuum.
Polymer films were hydrated with 1 mL of DPBS buffer and sonicated
for 2 min using a benchtop sonicator.
Synthesis and Conjugation of Keto Chain Transfer
Agent. There have been numerous reports of polymer end-
functionalization that utilize trithiocarbonate RAFT agents that
are modified at the reinitiator site. Much of this exploration was
enabled by convenient synthesis of S-alkyl-S′-(α,α-dimethyl-α″-
acetic acid) trithiocarbonates,18 which are then used to acylate
an appropriately derivatized amine or alcohol. There have been
far fewer reports on modification of the “Z” group on
trithiocarbonate RAFT agents, possibly due to perceived lability
of the trithiocarbonate to nucleophilic cleavage. Detailed
studies on nucleophilic cleavage of dithiocarbonate RAFT
agents indicated strong dependence of cleavage rates on
conditions.23 The synthesis of ketone-derivatized CTAs is
straightforward and high-yielding (Scheme 1).
1
6768
dx.doi.org/10.1021/ma301536f | Macromolecules 2012, 45, 6766−6773

Macromolecules
Article
a
Table 1. Macro-CTA and Block Copolymer Properties
Scheme 2. Hydrazone Conjugation
a
b
c
polymer
Mn (kDa)
Mw/Mn
DP
30
ketone (%)
Pa(30,0)
4.2
7.8
1.19
1.21
n.d.
n.d.
Pa(30,28)
Pb(30,28)
Pa(49,0)
58
58
n.d.
6.1*
6.7
52, 61
n.d.
1.22
1.24
n.d.
49
a
Reagents and conditiions: (a) 1% TFA in methanol; (b) 1% acetic
Pa(49,48)
Pb(49,48)
Pa(100,0)
Pa(100,89)
Pb(100,89)
12.8
10.0*
13.1
24.6
19.5*
97
n.d.
acid in methanol and 3.
97
41, 45
n.d.
1.16
1.23
n.d.
100
189
189
n.d.
RAFT Polymerization. It was first established that the
acetal chain transfer agent 1a was an effective RAFT agent.
Indeed, the kinetic signatures of reversible chain transfer24 were
observed in the linear growth of ln(M0/Mt) over time (Figure
3) for n-butyl acrylate monomer (M) when a ratio of CTA:n-
BA 1:239 was used. Furthermore, polymer Mn (GPC)
dependence on monomer conversion (1H NMR) indicated
the linear trend with good fit to theoretical Mn while
polydispersity (PDI) remained relatively flat and ∼1.2 PDI,
as judged by GPC analysis. Thus, polymer growth exhibits
pseudo-first-order kinetics and low PDI, consistent with chain
transfer characteristics of known RAFT agents. Overall, narrow
polydispersity throughout the polymerization suggests effective
chain transfer even at high monomer conversion (>70%). We
employed CTA 1a in the preparation of 1:1 block copolymers
of tert-butyl acrylate and n-butyl acrylate, which on acidolysis
yielded PnBA-b-PAA-keto. Block copolymers were prepared
with total monomer units of ∼58, 97, and 189 that exhibited
PDIs of ∼1.2 (Table 1).
Quantification of Accessible Polymer-Displayed Ke-
tone. Each of these block copolymer products could be
functionalized with acyl hydrazides with 40−60% yield relative
to total polymer. Functional incorporation of the keto group for
all polymer lengths was conveniently quantified by conjugation
with fluorescein-5-thiosemicarbazide (FTSC). Conjugation was
conveniently carried out in methanol with catalytic acetic acid,
where the polymers are soluble, and could also be carried out in
aqueous buffer at pH 7, where PnBA-b-PAA-keto polymers are
assembled into low-polydispersity nanoparticles with diameters
of 24, 38, and 78 nm, correlating with increasing polymer
length (Figure 5). Negative-stain TEM indicated the presence
of clusters of uniform spherical particles (Figure 5, Supporting
46, 53
a
Mn and Mw/Mn determined by GPC, except for asterisks, which
b
indicate theoretical molecular weight. Degree of polymerization,
calculated using Mn. Ketone accessible in polymer determined by
FTSC conjugation. n.d. = not determined.
c
The acetal transfer agent 1a was used in all polymerizations
and deprotected under acidic conditions to yield the ketone-
terminated polymer product. It is well-known that nucleophiles
such as primary amines and hydrazine may cleave the
trithiocarbonate, which reasonably raises concerns that the
keto end group may be cleaved from the polymer terminus in
the process of conjugation with acylhydrazides. While the
trithiocarbonate clearly survives acylation conditions with
amine nucleophiles, hydrazides are more reactive as nucleo-
philes. The conjugation reaction was performed on the soluble
keto RAFT agent itself with biotin−hydrazide under acidic
conditions, which are known to be favorable for hydrazone
formation and suppress23 the hydrazidolysis of the trithiocar-
bonate.
Biotin−hydrazide Reaction with CTA. Gratifyingly, in
situ 1H NMR monitoring of the conjugation of 1b with 3 in d4-
methanol indicated clear formation of the hydrazone in 84%
yield with the remaining material unreacted ketone (Figure 2).
Conversion was measured by the relative integration of ketone
and hydrazone methyl resonances, which appear as a singlet in
the ketone (2.18 ppm) and a pair of singlets in the hydrazone
due to syn- and anti-hydrazone configurations (1.85, 1.82
ppm). Similar conditions were used for postpolymerization
hydrazone conjugation.
Figure 2. 1H NMR of 1b (left) and hydrazone 7 (right). Hydrazone methyl peaks shift upfield as two singlets representing syn and anti
stereoisomers relative to the ketone methyl singlet of 1b. The singlet at 2.15 ppm is acetic acid while the smaller singlet at 2.13 ppm in both spectra
is a nonreactive impurity.
6769
dx.doi.org/10.1021/ma301536f | Macromolecules 2012, 45, 6766−6773

Macromolecules
Article
Figure 3. Kinetic plots of n-butyl acrylate (nBA) polymerization in the presence of 1a indicating (A) steady state conditions and (B) consistently low
PDI as a function of monomer conversion with good correlation between experimental and theoretical Mn (Mw,t). Monomer:1a:AIBN stoichiometry
was 239:1:0.1.
Figure 4. Chain extension studies on macroCTAs to produce block copolymers: (A) Pa(30,0) to Pa(30,28); (B) Pa(49,0) to Pa(49,48); (C) Pa(100,0) to
Pa(100,89). Generation of macro-CTAs as intermediates in copolymer synthesis was confirmed by chain extension studies in these three size regimes.
Information). The FTSC-conjugated polymer was concentrated
into a film from methanol, hydrated in DPBS buffer to yield
polymer nanoparticles, and then washed with buffer using a
Centricon filter to remove unreacted FTSC. Particle size was
not detectably changed relative to the free ketone particles
upon hydrazone formation, as judged by DLS. Analysis of the
unreacted FTSC in the filtrate indicated ∼50% retention by the
nanoparticle suspension. In contrast, reaction of the fluorescein
semithiocarbazide with a control polymer in which the CTA
bears a carboxylic acid in place of a ketone led to complete loss
of fluorescein absorbance from the nanoparticle suspension
upon filtration. Conjugated FTSC was rapidly cleaved from the
nanoparticles by hydrazinolysis at pH 9.5 and quantified by
measurement of fluorescein UV absorbance, which confirmed
the ∼50% yield for block copolymers of 6, 10, and 19.5 kDa
molecular weight (Table 1). Though the FTSC linkage is not
stable to strongly basic treatment with hydrazine, labeled
nanoparticles exhibit less than 1% loss in fluorescence after 2
days incubation at pH 7.4, DPBS at 37 °C, indicating that the
conjugation is sufficiently stable in biologically relevant
conditions. Nanoparticle-FTSC coupling was also tracked by
monitoring fluorescence anisotropy change. As expected,
anisotropy increased significantly in the keto nanoparticle
reaction, while in the control nanoparticle reaction anisotropy
change was weaker (Figure 5). Though an anisotropy increase
was observed with the carboxylate-terminated nanoparticle
control, this likely represents minor nonspecific binding given
the relatively low fluorescence intensity of the sample.
Polymer Nanoparticle Biotinylation. Particle functional-
ization with protein ligands enables targeting to specific
biological targets directly or through antibody conjugation.
We demonstrate this functionality with biotin modification of
the polymer nanoparticle surface. Hydrazone conjugation of
biotin hydrazide with ketone-terminated polymer was per-
formed in a similar fashion to FTSC coupling. Biotin binding
was also assessed by fluorescence anisotropy change using
streptavidin−FITC. Indeed, treatment of fluorescently labeled
streptavidin with surface biotinylated polymer nanoparticles
resulted in the expected increase in FITC anisotropy, consistent
with the larger size of the putative nanoparticle-bound
streptavidin fluorophore relative to the protein. Incubation of
streptavidin−FITC with non-biotinylated polymer nanopar-
ticles did not significantly alter fluorescence anisotropy, nor did
incubation with free biotin (Figure 5); this is also expected
given the minor change in protein size upon small-molecule
binding. Biotin titration of streptavidin−FITC is accompanied
6770
dx.doi.org/10.1021/ma301536f | Macromolecules 2012, 45, 6766−6773

Macromolecules
Article
Figure 5. (A) Dynamic light scattering measurements (DPBS, pH 7.4) of polymer nanoparticles indicated. (B) Negative stain transmission electron
●
micrographs of Pb(49,48), with 100 nm scale bar shown. Fluorescence anisotropy measurements of ketone polymer Pb(100,89) ( ) and control
○
polymer Pc(100,89) ( ) during (C) nanoparticle conjugation with FTSC 2 and (D) incubation streptavidin−FITC following conjugation with biotin-
hydrazide 3.
by an increase in fluorescence intensity. Though this increase in
emission on biotin binding is not well-understood, it provides
an analytical measure of concentration.25 By this empirical
method, a much higher concentration of polymer-displayed
biotin is required to attain the same increase in streptavidin−
FITC emission relative to free biotin (Supporting Information).
The origin of this difference is unclear and may indicate
diminished accessibility26 of the biotinylated polymer terminus
or an altered FITC emission environment for the protein−
nanoparticle complex relative to the protein−biotin complex.
Conjugation of MRI Contrast Agent to Ketone
Nanoparticles. Nanoparticle accumulation in solid tumors
and atherosclerotic plaques27 through the enhanced permeation
and retention (EPR) effect has generated considerable interest
in their use as delivery and imaging agents. Significant
advantages in tumor/plaque imaging are presented by nano-
particle conjugation to traditional magnetic resonance imaging
(MRI) contrast agents such as gadolinium complexes: increased
retention in the tissue of interest and enhanced MRI sensitivity.
The gadolinium DOTA hydrazide complex 4 was incubated
with the keto-nanoparticle (Pb(100,89)) suspension as well as the
carboxylate nanoparticle control (Pc(100,89)) and purified by
dialysis. Conjugation of 4 to nanoparticle Pb(100,89) significantly
increased Gd relaxivity. Over the pH range 4−8 and field
strength of 1.5 T (60 MHz), conjugate relaxivity per Gd3+ was
∼50 mM−1 s−1, about 17 times higher than commercial
Gd(DOTA) (3.0 mM−1 s−1). Such an increase in relaxivity
could derive from more limited motion of the Gd complex
when bound to the polymer assembly. Protein, polymer, and
particle-bound gadolinium complexes have been previously
reported to have increased Gd relaxivities28 with r1 values as
high as 15.9 mM−1 s−1 for polymer-caged nanoparticles29 and
41 mM−1 s−1 for Gd-bound high-density lipoprotein.30
Similarly, our study shows that Gd(DOTA)−hydrazone
conjugation to polymer nanoparticles from Pb(100,89) elicits a
dramatic increase in r1 relaxivity per Gd ion. Contrast agents
possessing hypersensitivity warrant further detailed investiga-
tion, especially for in vivo imaging applications where high
sensitivity is critical. No gadolinium was detectable in the
control nanoparticle sample (Pc(100,89)) following dialysis.
Notably, similar relaxivity over a wide pH range of our
nanoparticle contrast agent (Figure 6) suggests utility in acidic
6771
dx.doi.org/10.1021/ma301536f | Macromolecules 2012, 45, 6766−6773

Macromolecules
Article
Figure 6. Dependence of gadolinium (A) T1 and (B) 1/T1 on concentration of Gd−DOTA hydrazide 4 incubated with nanoparticles from keto
polymer Pb(100,89) following dialysis purification. Relaxivities over the pH range 4−8 are 48.1, 35.2, 48.0, 50.2, and 54.7 mM−1 s−1, respectively (see
legend).
microenvironments as well as trithiocarbonate stability in basic
aqueous conditions. It is possible that decreased accessibility to
the trithiocarbonate linkage via polyacrylate neutralization
affords greater hydrazone stability in acidic pH regimes.
Tracking Cellular Uptake of Fluorescent Nanopar-
ticles. To showcase the potential utility of the keto-CTA to
track polymer biodistribution, we studied the cell uptake of
fluorescently labeled nanoparticles. Mixed films of fluorescently
labeled keto PnBA-b-PAA (19.5 kDa) and carboxylate
terminated polymer of the same length were prepared with
varying proportion of FTSC-tagged polymer. Buffer hydration
of these mixed films yielded fluorescent nanoparticle
populations that showed no loss of fluorescence or detectable
precipitation in serum at 37 °C. Cell uptake over a range of
polymer concentration from 0.1 to 250 μg/mL was studied in
two widely used cell lines, A549 and MCF-7. Full retention by
molecular weight cutoff filters greater than unimer size
established that concentrations of 0.1 μg/mL were safely
above polymer critical micelle concentration (cmc) for all three
block copolymer lengths; thus, conditions reflect nanoparticle
uptake rather than uptake of soluble polymer. Following a 6 h
incubation and wash, cells were trypsinized, fixed and analyzed
by FACS or fixed and examined by epifluorescence microscopy.
Interestingly, while both cell types showed increased
fluorescence, there was significantly higher uptake by MCF-7
for reasons we do not understand (Figure 7). Though this
polyacrylate architecture has been well-studied,22 this is the first
report of its cell uptake profile, enabled by polymer end-
labeling. Dose dependence of uptake was established for the
MCF-7 cell line, with increased fluorescence correlating with
higher degree of polymer labeling and total polymer
concentration.
Figure 7. Cellular uptake studies on FTSC labeled Pb(100,89). Overlaid
phase contrast and fluorescein epifluorescence microscopy images of
washed and fixed (A) A549 cells and (B) MCF-7 cells after incubation
with nanoparticles composed of 25% FTSC-labeled Pb(100,89) and 75%
Pc(100,89). Relative uptake profiles obtained by FACS are shown of (C)
MCF-7 vs A549 cells and (D) fluorescence in MCF-7 cells, as
proportion of labeled Pb(100,89) is increased as indicated.
derivatives of FITC, biotin, and Gd-DOTA yielded the
expected fluorescent, protein-binding, and relaxivity properties
in the assembled polymer nanoparticle. This strategy, in
combination with nanoparticle encapsulation strategies,28
could conveniently generate nanoparticle assemblies with
multiple functions. Discrete end-functionalization methodology
also enables necessary assembly characterization: fluorophore
installation allowed convenient determination of the upper limit
of polymer cmc and tracking of cellular uptake of polymer
nanoparticles without disruption of particle core packing
interactions. Overall, hydrazone conjugation to the Z-group
appears entirely stable to handling and extended incubation
under physiological pH and temperature, despite the known
lability of the trithiocarbonate linkage to nucleophilic attack.
Therefore, this method of polymer end-functionalization
CONCLUSIONS
■
Using a ketone-modified Z-group on a trithiocarbonate RAFT
agent, we have demonstrated effective and practical hydrazone
conjugation of amphiphilic block copolymers of the general
form PnB-b-PAA, using reporter tags commonly required for
biotechnological applications such as imaging, bioconjugation,
and targeting. Polymer−ketone conjugation with hydrazide
6772
dx.doi.org/10.1021/ma301536f | Macromolecules 2012, 45, 6766−6773

Macromolecules
Article
(15) Li, Y.; Lokitz, B. S.; McCormick, C. L. Macromolecules 2005, 39,
81−89.
provides a synthetically convenient and useful complement to
known methods of modification at the reinitiator “R” site of
RAFT agents. Synthetic possibilities in polymer architecture
may be broadened by a combination of the two approaches to
CTA derivatization.
́
(16) (a) Vazquez-Dorbatt, V.; Tolstyka, Z. P.; Maynard, H. D.
Macromolecules 2009, 42, 7650−7656. (b) Grover, G. N.; Lee, J.;
Matsumoto, N. M.; Maynard, H. D. Macromolecules 2012, 45, 4958−
4965.
(17) Jackson, A. W.; Fulton, D. A. Macromolecules 2009, 43, 1069−
ASSOCIATED CONTENT
* Supporting Information
Additional experimental details, compound characterization,
TEM. This material is available free of charge via the Internet at
http://pubs.acs.org.
1075.
■
S
(18) Lai, J. T.; Filla, D.; Shea, R. Macromolecules 2002, 35, 6754−
6756.
(19) Nakamura, Y.; Kitada, Y.; Kobayashi, Y.; Ray, B.; Yamago, S.
Macromolecules 2011, 44, 8388−8397.
(20) Coster, M. J.; De Voss, J. J. Org. Lett. 2002, 4, 3047−3050.
(21) Kim, J. Y.; Chae, H. K.; Choi, B. C.; Ha, H.-J.; Oh, S. J. Bull.
Korean Chem. Soc. 2004, 35, 1512−1514.
AUTHOR INFORMATION
Corresponding Author
*E-mail: bong@chem.osu.edu.
Notes
■
(22) (a) Lejeune, E.; Chassenieux, C.; Colombani, O. Prog. Colloid
Polym. Sci. 2011, 138, 7−16. (b) Lejeune, E.; Drechsler, M.; Jestin, J.;
Muller, A. H. E.; Chassenieux, C.; Colombani, O. Macromolecules
̈
2010, 43, 2667−2671. (c) Nicolai, T.; Colombani, O.; Chassenieux, C.
The authors declare no competing financial interest.
Soft Matter 2010, 6, 3111−3118.
(23) Thomas, D. B.; Convertine, A. J.; Hester, R. D.; Lowe, A. B.;
McCormick, C. L. Macromolecules 2004, 37, 1735−1741.
ACKNOWLEDGMENTS
This work was supported in part by NSF-0927778, NSF-
0747194 (D.B.), and NIH-HL106487 (S.R.).
■
(24) (a) Litvinenko, G.; Muller, A. H. E. Macromolecules 1997, 30,
̈
1253−1266. (b) Muller, A. H. E.; Zhuang, R.; Yan, D.; Litvinenko, G.
̈
Macromolecules 1995, 28, 4326−4333. (c) Chong, Y. K.; Krstina, J.; Le,
T. P. T.; Moad, G.; Postma, A.; Rizzardo, E.; Thang, S. H.
Macromolecules 2003, 36, 2256−2272.
REFERENCES
■
(1) (a) Keddie, D. J.; Moad, G.; Rizzardo, E.; Thang, S. H.
Macromolecules 2012, 45, 5321−5342. (b) Moad, G.; Rizzardo, E.;
Thang, S. H. Acc. Chem. Res. 2008, 41, 1133−1142. (c) Boyer, C.;
Stenzel, M. H.; Davis, T. P. J. Polym. Sci., Part A: Polym. Chem. 2011,
49, 551−595. (d) Stenzel, M. H. Chem. Commun. 2008, 2008, 3486−
3503.
(2) (a) Braunecker, W. A.; Matyjaszewski, K. Prog. Polym. Sci. 2007,
32, 93−146. (b) Zetterlund, P. B.; Kagawa, Y.; Okubo, M. Chem. Rev.
2008, 108, 3747−3794.
(3) Boyer, C.; Bulmus, V.; Davis, T. P.; Ladmiral, V.; Liu, J.; Perrier,
S. Chem. Rev. 2009, 109, 5402−5436.
(4) (a) Maibaum, L.; Dinner, A. R.; Chandler, D. J. Phys. Chem. B
2004, 108, 6778−6781. (b) Discher, D. E.; Eisenberg, A. Science 2002,
297, 967−973.
(5) Bartels, J. W.; Cauet, S. I.; Billings, P. L.; Lin, L. Y.; Zhu, J.; Fidge,
C.; Pochan, D. J.; Wooley, K. L. Macromolecules 2010, 43, 7128−7138.
(6) Immordino, M. L.; Dosio, F.; Cattel, L. Int. J. Nanomed. 2006, 1,
297−315.
(25) Smith-Palmer, T.; Barbarakis, M. S.; Cynkowski, T.; Bachas, L.
G. Anal. Chim. Acta 1993, 279, 287−292.
(26) Qi, K.; Ma, Q.; Remsen, E. E.; Clark; Wooley, K. L. J. Am. Chem.
Soc. 2004, 126, 6599−6607.
(27) (a) Perrault, S. D.; Walkey, C.; Jennings, T.; Fischer, H. C.;
Chan, W. C. W. Nano Lett. 2009, 9, 1909−1915. (b) Jain, R. K.;
Stylianopoulos, T. Nat. Rev. Clin. Oncol. 2010, 7, 653−664. (c) Phillips,
M. A.; Gran, M. L.; Peppas, N. A. Nano Today 2010, 5, 143−159.
(28) (a) Horacio, C.; Kazunori, K. Sci. Technol. Adv. Mater. 2010, 11,
014109. (b) Janib, S. M.; Moses, A. S.; MacKay, J. A. Adv. Drug
Delivery Rev. 2010, 62, 1052−1063.
(29) Lee, S. M.; Song, Y.; Hong, B. J.; MacRenaris, K. W.; Mastarone,
D. J.; O’Halloran, T. V.; Meade, T. J.; Nguyen, S. T. Angew. Chem., Int.
Ed. 2010, 49, 9960−9964.
(30) Briley-Saebo, K. C.; Geninatti-Crich, S.; Cormode, D. P.;
Barazza, A.; Mulder, W. J. M.; Chen, W.; Giovenzana, G. B.; Fisher, E.
A.; Aime, S.; Fayad, Z. A. J. Phys. Chem. B 2009, 113, 6283−6289.
(7) (a) Knop, K.; Hoogenboom, R.; Fischer, D.; Schubert, U. S.
Angew. Chem., Int. Ed. 2010, 49, 6288−6308. (b) Kang, J. S.; Deluca, P.
P.; Lee, K. C. Expert Opin. Emerging Drugs 2009, 14, 363−380.
(c) Veronese, F. M.; Mero, A.; Pasut, G. PEGylated Protein Drugs
2009, 11−31. (d) Fishburn, C. S. J. Pharm. Sci. 2008, 97, 4167−4183.
(8) (a) Elsabahy, M.; Wooley, K. L. Chem. Soc. Rev. 2012, 41, 2545−
2561. (b) Geng, Y.; Dalhaimer, P.; Cai, S.; Tsai, R.; Tewari, M.;
Minko, T.; Discher, D. E. Nat. Nanotechnol. 2007, 2, 249−255.
(9) Bathfield, M. l.; D’Agosto, F.; Spitz, R.; Charreyre, M.-T. r. s.;
Delair, T. J. Am. Chem. Soc. 2006, 128, 2546−2547.
(10) Bontempo, D.; Maynard, H. D. J. Am. Chem. Soc. 2005, 127,
6508−6509.
(11) (a) Moad, G.; Rizzardo, E.; Thang, S. H. Polym. Int. 2011, 60,
9−25. (b) Tasdelen, M. A.; Kahveci, M. U.; Yagci, Y. Prog. Polym. Sci.
2011, 36, 455−567.
(12) York, A. W.; Kirkland, S. E.; McCormick, C. L. Adv. Drug
Delivery Rev. 2008, 60, 1018−1036.
(13) (a) Kalia, J.; Raines, R. T. Angew. Chem., Int. Ed. 2008, 47,
7523−7526. (b) Kalia, J.; Raines, R. T. Curr. Org. Chem. 2010, 14,
138−147.
(14) (a) Frullano, L.; Tejerina, B.; Meade, T. J. Inorg. Chem. 2006,
45, 8489−8491. (b) Lee, C. C.; Gillies, E. R.; Fox, M. E.; Guillaudeu,
S. J.; Frechet, J. M.; Dy, E. E.; Szoka, F. C. Proc. Natl. Acad. Sci. U. S. A.
2006, 103, 16649−16654. (c) Kale, A. A.; Torchilin, V. P. Bioconjugate
Chem. 2007, 18, 363−370.
6773
dx.doi.org/10.1021/ma301536f | Macromolecules 2012, 45, 6766−6773