Study of the Micellization Behavior of Different Order Amino Block Copolymers with Heparin

Article abstract of DOI:10.1023/B:PHAM.0000012164.60867.c6

Purpose. The purpose of this study was to prepare and characterize copolymers presenting different pendant amino groups and to study their ability to form polyelectrolyte complexes with heparin. The responsiveness of the complexes to variations in pH and ionic strength was correlated to the nature of the copolymers. Methods. Copolymers composed of different aminoethyl methacrylate monomers were synthesized by atom transfer radical polymerization (ATRP) from a poly(ethylene glycol) macroinitiator. Copolymers were characterized by gel permeation chromatography and nuclear magnetic resonance spectroscopy. Micellization properties were assessed by atomic force microscopy, multiangle static light scattering, and dynamic light scattering on complexes formed from the addition of heparin to a solution of polymer. Results. Primary, tertiary, and quaternary amine-based diblock copolymers with molecular weights ranging from 4900 to 7400 and low polydispersity indexes were prepared. The synthesis of a copolymer bearing primary amines was achieved for the first time by ATRP. Micellization was found to be pH- and polymer-dependent. All polymers interacted with heparin at acidic pH to yield monodisperse assemblies of less than 30 nm. Complexes dissociated in response to increases in ionic strength. Conclusions. Electrostatic interactions between the amino copolymers and heparin triggered the formation of small, monodisperse, and stable complexes that present great potential as oral drug delivery systems.

Full text of DOI:10.1023/B:PHAM.0000012164.60867.c6

Pharmaceutical Research, Vol. 21, No. 1, January 2004 (© 2004)
Research Paper
in North America for preventing and treating deep vein
thrombosis and pulmonary embolisms. Unfortunately, its ad-
ministration is restricted to the parenteral route as its oral
bioavailability is impaired by a short half-life, large molecular
size, and overall negative charge. Current treatments require
careful patient monitoring and therefore point to the devel-
opment of new lower-cost and patient-friendly formulations.
Several strategies have been explored to increase the oral
bioavailability of heparin. These are generally aimed either at
the neutralization of its negative charges (to prevent electro-
static repulsions between the drug and gastrointestinal mu-
cosa) or the protection of the drug from pH and enzymatic
degradation. Examples include the use of absorption enhanc-
ers [such as ethylenediaminetetraacetic acid (1) or sulfonated
surfactants (2)], chemical modifications [deoxycholic acid–
heparin conjugates (3)], and particulate carrier systems (4).
However, these approaches have only had limited success.
Currently, the most effective formulation is the co-
Study of the Micellization Behavior of
Different Order Amino Block
Copolymers with Heparin
1
1,2
Marie-Hélène Dufresne and Jean-Christophe Leroux
Received September 2, 2003; accepted September 30, 2003
Purpose. The purpose of this study was to prepare and characterize
copolymers presenting different pendant amino groups and to study
their ability to form polyelectrolyte complexes with heparin. The
responsiveness of the complexes to variations in pH and ionic
strength was correlated to the nature of the copolymers.
Methods. Copolymers composed of different aminoethyl methacry-
late monomers were synthesized by atom transfer radical polymer-
ization (ATRP) from a poly(ethylene glycol) macroinitiator. Copoly- administration of heparin with N-[8-(2-hydroxybenzoyl)ami-
mers were characterized by gel permeation chromatography and no]caprylate (5), an absorption enhancer that improves the
nuclear magnetic resonance spectroscopy. Micellization properties oral absorption of heparin in humans. Though promising, this
were assessed by atomic force microscopy, multiangle static light scat-
formulation is not yet optimal in that the anticoagulant activ-
tering, and dynamic light scattering on complexes formed from the
ity only lasts 4 h at doses higher than those usually recom-
addition of heparin to a solution of polymer.
mended for humans. Recently, the use of polymeric carriers
Results. Primary, tertiary, and quaternary amine-based diblock co-
has attracted considerable attention as the technology can
polymers with molecular weights ranging from 4900 to 7400 and low
simultaneously protect a drug from the surroundings, provide
a controlled and sustained release, and potentially account for
increased absorption. In this light, heparin-loaded micro- and
polydispersity indexes were prepared. The synthesis of a copolymer
bearing primary amines was achieved for the first time by ATRP.
Micellization was found to be pH- and polymer-dependent. All poly-
mers interacted with heparin at acidic pH to yield monodisperse as- nanoparticles composed of blends of two biodegradable poly-
semblies of less than 30 nm. Complexes dissociated in response to mers [namely poly(∈-caprolactone) and poly(D,L-lactic-co-
increases in ionic strength.
glycolic acid)] and two nonbiodegradable cationic polymeth-
acrylates (Eudragit RS and RL) were proposed by Maincent
and co-workers (6,7). These systems presented anticoagulant
activity in a rabbit model for up to 7 h at doses close to those
commonly administered by the intravenous route. We now
Conclusions. Electrostatic interactions between the amino copoly-
mers and heparin triggered the formation of small, monodisperse,
and stable complexes that present great potential as oral drug deliv-
ery systems.
KEY WORDS: atom transfer radical polymerization; cationic co- intend to exploit the properties of polymeric carrier systems
polymers; heparin; polyion complex micelles.
by introducing a new formulation of heparin as polyion com-
plex micelles (PICM).
INTRODUCTION
PICM generally result from favorable interactions be-
Heparin is a macromolecular glycosaminoglycan with an-
tween a charged diblock copolymer and an oppositely
ticoagulant properties. It is one of the drugs most widely used
charged polyion (8). The formation of PICM is best described
as the sequential complexation of the polyions through elec-
1
Canada Research Chair in Drug Delivery, Faculty of Pharmacy,
trostatic interactions followed by the self-association of the
University of Montreal, P.O. Box 6128, Succ. Centre-Ville, Montre-
condensates into micelles. Neutralization of the charged seg-
ments indeed yields water-insoluble moieties that contribute
to a decrease in the entropy of the system. To offset this
entropy loss, the neutralized segments withdraw from the
al, Quebec, Canada, H3C 3J7.
To whom correspondence should be addressed. (e-mail: jean-
christophe.leroux@umontreal.ca)
2
ABBREVIATIONS: ␣, ionization degree; AEMA, 2-(N-amino)ethyl
methacrylate hydrochloride; AEMABoc, 2-[N-(tert-butoxycarbonyl)
aqueous phase and self-assemble into micelles presenting a
amino]ethyl methacrylate; AFM, atomic force microscopy; ATRP, atom core/shell architecture. The formation of PICM was shown to
transfer radical polymerization, CDCl , deuterated chloroform; be concentration-dependent, and the critical association con-
3
CD OD, deuterated methanol; D O, deuterated water; DCM, dichlo- centration of such systems has been evaluated by either moni-
3
2
romethane; DEAEMA, 2-(N,N-diethylamino)ethyl methacrylate;
DEMAEMA, 2-(N,N,N-diethylmethylamino)ethyl methacrylate;
DLS, dynamic light scattering; DMAEMA, 2-(N,N-dimethylamino)
toring changes in the intensity of pyrene fluorescence emis-
sion (9) or in the apparent molecular weight (10) with con-
centration. PICM have been designed for the delivery of
charged drugs such as oligonucleotides (11), DNA (12), and
enzymes (13). They present numerous advantages including
their straightforward synthesis, self-assembly in aqueous me-
dium, and narrow size distribution. Furthermore, entrapment
1
ethyl methacrylate; GPC, gel permeation chromatography; H NMR,
proton nuclear magnetic resonance; MASLS, multiangle static light
scattering; M , number-average molecular weight; M , weight-
n
w
average molecular weight; MW, molecular weight; PAMAM, poly-
amidoamide; PEG, poly(ethylene glycol); PEI, polyethyleneimine;
PI, polydispersity index; PICM, polyion complex micelles; PMDETA, of the drug in the core was shown to protect it from in vivo
N,N,N’,N’,N’’-pentamethyldiethylenetriamine; THF, tetrahydrofu- metabolism or enzymatic degradation (10). Polymers com-
ran; TMAEMA, 2-(N,N,N-trimethylamino)ethyl methacrylate.
monly used for the delivery of polyanionic drugs include poly-
0
724-8741/04/0100-0160/0 © 2004 Plenum Publishing Corporation
160

Complexation of Amino Copolymers with Heparin
161
ethylenimine (PEI) (14), poly(L-lysine) (11), polyamidoamide used to produce polymers bearing pendant primary amino
PAMAM) (15), and poly[2-(N,N-dimethylamino) ethyl groups.
(
methacrylate] (PDMAEMA) (16). All polymers, except for
PDMAEMA, present primary amines that are mostly proto- MATERIALS AND METHODS
nated at physiological pH. Furthermore, PEI and PAMAM
possess higher order amines that are responsible for their Materials
increased buffering capacity. Clearly, the order (i.e., degree of
Heparin sodium salt from porcine intestinal mucosa [mo-
substitution) of the amines confers certain characteristic
properties to the drug–polymer complexes. In this study,
block copolymers of poly(ethylene glycol) (PEG) and moi-
eties bearing different amino groups (Fig. 1) were synthesized
in order to evaluate the relationship between the different
order amines and their complexation behavior with heparin.
The polymerization of different alkylated derivatives of 2-(N-
amino)ethyl methacrylate provided a range of primary, ter-
tiary, and quaternary amine-based complexes.
lecular weight (MW) of ∼6000] was purchased from Sigma (St.
Louis, MO, USA). 2-(N-amino)ethyl methacrylate hydro-
chloride (AEMA) was obtained from Polysciences Inc. (War-
rington, PA, USA) and used as received. All other chemicals
were purchased from Aldrich Chemical Co. (Milwaukee, WI,
USA) and used as received, except for 2-(N,N-
dimethylamino)ethyl methacrylate (DMAEMA) and 2-(N,N-
diethylamino)ethyl methacrylate (DEAEMA), which were
vacuum-distilled prior to polymerization. Tetrahydrofuran
All copolymers were synthesized by a robust and versa-
tile controlled polymerization method known as atom trans-
fer radical polymerization (ATRP). ATRP allows for the con-
trol of molecular weight and polydispersity of polymer chains
as well as their composition (17), architecture (18), and end-
group functionality (19). The success of ATRP lies in the
establishment of a dynamic equilibrium between propagating
radicals and dormant polymer chains so as to keep the actual
concentration of active species much lower than the total con-
(
THF) was dried over sodium with benzophenone as indica-
tor, and dichloromethane (DCM) was dried over calcium
chloride.
Instrumentation
Nuclear Magnetic Resonance Analyses
1
All H NMR spectra were recorded on a Bruker NMR
centration of chains, thus reducing radical–radical termina- spectrometer operating at 300 MHz (Bruker, Milton, Ontario,
tion and enhancing control over the polymerization reaction Canada). Solutions were prepared in deuterated chloroform
(
20). Despite its capability of polymerizing a wide range of (CDCl ), methanol (CD OD), or water (D O).
3 3 2
monomers, the polymerization of methacrylates presenting
pendant amino groups remains difficult. This is because
amines can provoke the displacement of the ligand from the
metal complex (21) and, in the case of primary amines, react
with the halogen end group of the initiator or of the growing
polymer chain (19). Under appropriate reaction conditions,
we were able to exploit ATRP for the synthesis of copolymers
presenting different amino groups. To the best of our knowl-
edge, this is the first report in which ATRP was successfully
Molecular Weight Determination
Number- (M ) and weight-average (M ) MWs were de-
n
w
termined by gel permeation chromatography (GPC) using an
Alliance GPCV 2000 system (Waters, Milford, MA, USA)
equipped with a differential refractive index detector. GPC
analyses were performed in THF (HPLC grade) using three
Waters Styragel columns (HR1, HR3, and HR4) in series at a
flow rate of 1.0 ml/min and a temperature of 40°C. A calibra-
tion curve was obtained with near-monodisperse PEG stan-
dards.
Dynamic Light Scattering and Multiangle Static Light
Scattering Measurements
The mean hydrodynamic diameter, size distribution, and
scattering intensity of PICM were determined using a Mal-
vern Autosizer 4800 equipped with a 488-nm uniphase argon–
ion laser (Malvern Instruments, Worcestershire, UK). Mea-
surements were performed in triplicate at a scattering angle of
90° and a temperature of 25°C. The CONTIN program was
used to extract size distributions from the autocorrelation
functions. Alternatively, MASLS measurements were con-
ducted at 25°C at eight angles ranging from 50° to 120° for a
minimum of three concentrations. All samples were passed
through 0.45- and 0.22-␮m nylon filters prior to analysis. The
M s of the complexes were extracted from Eq. (1), which
w
describes the intensity of light scattered from a dilute solution
of macromolecules:
Kc
1
2
=
+ 2A c + 3A c
(1)
2
3
R0 M P͑␪͒
w
In this equation, c stands for the solution concentration
Fig. 1. Structure of the amino diblock copolymers studied.
in both polymer and heparin, R is the Rayleigh ratio of the
0

1
62
Dufresne and Leroux
1
solution, and A and A are the second and third virial coef-
H NMR (␦, ppm, CDCl ): 6.13 (1H, d), 5.60 (1H, t), 4.78
2
3
3
ficients. K is in turn defined by:
(1H, broad s), 4.21 (2H, t), 3.45 (2H, q), 1.95 (3H, s), 1.45 (9H,
s). Yield: 93%.
2
dn
2
2
4
␲ n ͩ ͪ
0
dc
K =
(2) ATRP Polymerization
␭4
N
Polymerization of monomers was performed in solution
using ␣-(2-bromoisobutyrylate)-␻-methylPEG as the ATRP
macroinitiator. The PEG ATRP macroinitiator (1 Eq) was
dissolved in distilled THF along with 16–25 Eq of the mono-
mer to a final monomer concentration of 0.5 or 0.8 M.
Concurrently, Cu(I)Br (1.1 Eq) and N,N,N’,N’,N’’-
pentamethyldiethylenetriamine (PMDETA, 1.1 Eq) were
mixed in a second round-bottom flask. The contents of both
reaction vessels were purged with argon for 15–20 min, after
which the PEG–monomer solution was transferred to the
metal complex activator using a two-head syringe. Polymer-
ization was typically carried out at 65°C for 16 h. The poly-
merization reaction was stopped by dilution with an ethanol–
THF solution (final ethanol concentration of 10%) and re-
moval of the copper complex by filtration on a silica gel
column using THF as eluent. Evaporation of the solvents
under reduced pressure provided crude polymers that were
purified by dialysis against water (Spectra/Por no.1, MW cut-
off 6000–8000) for at least 48 h and recovered by freeze-
drying (using a FreeZone 6 system, Labconco, Kansas City,
MO, USA).
where n is the refractive index of the solvent, ␭ the laser
0
wavelength in vacuo, and N Avogadro’s number. The specific
refractive index increments (dn/dc) were determined using a
differential Rudolph J157 automatic refractometer (Rudolph
Research Analytical, Flanders, NJ, USA), at a wavelength of
5
89.3 nm.
Atomic Force Microscopy Imaging
The topographic data of the complexes was collected us-
ing a Nanoscope III Dimension 3100 atomic force microscope
(Digital Instruments, Santa Barbara, CA, USA). Imaging was
performed in tapping mode with a silicon tip (tapping mode
etched Si probes – RTESP7) operating at a 300-kHz resonant
frequency and a 42 N/m constant force. Samples were pre-
pared by letting a drop of an acidic aqueous solution of PICM
(polymer concentration of 0.2 mg/ml, pH 5) dry on a freshly
cleaved mica surface at room temperature. PICM were
formed at a −/+ molar charge ratio of 1.5.
Synthesis of Poly(Ethylene Glycol) ATRP Macroinitiator
[
␣-(2-Bromoisobutyrylate)-␻-MethylPEG]
PEG-b-P(AEMABoc)
Fifteen grams of MeO–PEG–OH (M ס 2000, 7.5 mmol)
n
1
was dried with toluene by azeotropic distillation, dissolved in
0 ml anhydrous THF, and chilled for 20 min in an ice-water
H NMR (␦, ppm, CDCl ): 5.54 (16H, broad s), 4.01
3
7
(32H, broad s), 3.65 (185H, m), 3.39 (3H, s + 32H, broad s),
bath in presence of triethylamine (2.4 ml, 17.2 mmol). 2-Bro-
moisobutyryl bromide (6.5 ml, 33.5 mmol) was added drop-
wise; the reaction mixture was warmed to room temperature
and allowed to react for 48 h (22). The reaction was stopped
by adding DCM and washing with 10% aqueous HCl. The
organic phase was then extracted with 1 M NaOH, brine, and
dried over magnesium sulfate. Solvents were evaporated un-
der reduced pressure. The crude extract was dissolved in a
minimum of DCM and precipitated in diethyl ether. The
1
.84 (36H, m), 1.45 (160H, s), 1.26–0.89 (6H, s + 46H, m).
PEG-b-P(DMAEMA)
1
H NMR (␦, ppm, CDCl ): 4.08 (33H, broad s), 3.65
3
(176H, m), 3.38 (3H, s), 2.59 (33H, broad s), 2.30 (96H, s),
1.98–1.83 (31H, m), 1.14–0.90 (6H, s + 46H, m). Yield: 25%,
after purification.
product was recovered by simple filtration.
1
H NMR (␦, ppm, CDCl ): 4.33 (2H, t), 3.65 (180H, m),
3
PEG-b-P(DEAEMA)
3
.38 (3H, s), 1.94 (6H, s). Yield: 70%, after precipitation.
1
H NMR (␦, ppm, CDCl ): 3.99 (25H, broad s), 3.65
3
Synthesis of 2-[N-(tert-Butoxycarbonyl)Amino]Ethyl
Methacrylate (AEMABoc)
(
1
175H, m), 3.39 (3H, s), 2.70 (25H, broad s), 2.58 (50H, m),
.98–1.81 (24H, m), 1.26–0.90 (6H, s + 77H, t + 39H, m).
Yield: 31%, after purification.
A suspension of AEMA in anhydrous DCM (0.30 M)
was stirred for 10 min in an ice-water bath. Freshly distilled
triethylamine was then added (1.1 Eq) and the reaction mix- Deprotection of PEG-b-P(AEMABoc) to Its
ture stirred for 20 min. A solution of bis(tert-butyl)di- Hydrochloride Salt Derivative PEG-b-P(AEMA)
carbonate in DCM (1.1 Eq in sufficient DCM to bring AEMA
to a final concentration of 0.27 M) was added; the reaction
Crude polymer extracts were dissolved in 3 M HCl/
mixture was warmed to room temperature and allowed to EtOAc to a final amine concentration of 0.15 N. The reaction
1
react for 24 h. The reaction was stopped by extracting the was carried out for 2 ⁄
h after which solvents were removed
2
organic phase with water, 1 M HCl, saturated aqueous sodi- under reduced pressure. The polymer was dissolved in etha-
um bicarbonate, and brine. The organic extract was dried nol, dialyzed against water, and recovered by freeze-drying.
1
over sodium sulfate, and the solvent was removed in vacuo to
H NMR (␦, ppm, D O): 4.21 (34H, broad s), 3.66 (165H,
2
give a crude crystalline product that was used without further m), 3.34 (3H, s), 3.25 (34H, broad s), 1.99 (31H, m), 1.16–0.90
purification.
(6H, s + 50H, m). Yield: 28%, after purification.

Complexation of Amino Copolymers with Heparin
163
Transformation of PEG-b-P(DMAEMA) and
Effect of pH on Micelle Formation
PEG-b-P(DEAEMA) into PEG-b-Poly[2-(N,N,N-
Trimethylamino)Ethyl Methacrylate] [PEG-b-P(TMAEMA)]
and PEG-b-Poly[2-(N,N,N-Diethylmethylamino)Ethyl
Methacrylate] [PEG-b-P(DEMAEMA)], Respectively
PICM of heparin (−/+ molar charge ratio of 1.5) were
prepared to final polymer concentrations of 1 mg/ml. The pH
of the preparations was acidified with 1 N HCl to ∼4 and
sequentially increased to ∼11 using 1 N NaOH. Aliquots were
Alkylation of the tertiary amine groups of the methacry- sampled after each incremental addition of NaOH, passed
late residues was done in order to introduce permanent posi- through 0.45-␮m filters, and analyzed by DLS. Size data was
tive charges on the nitrogen atoms. An excess of methyliodide only recorded for samples with high enough scatter counts
(
2 Eq with respect to amino groups) was added to a solution (above the detection limit).
of polymer in water (30 mg/ml). Methylation was carried for
and 4 h for PEG-b-P(DMAEMA) and PEG-b- Effect of Salt Concentration on Micelle Stability
3
P(DEAEMA), respectively. The quaternized polymers were
purified by dialysis against water and recovered by freeze-
drying.
Stoichiometric PICM of heparin were prepared as acidi-
fied aqueous solutions (pH ∼4) with final polymer concentra-
tions of 5–7.5 mg/ml. Specific volumes of a 5 M NaCl solution
were added to yield solutions with salt concentrations ranging
from 0 to 1.2 M. Both the PICM and stock salt solutions were
passed through 0.45-␮m nylon filters. DLS analyses were per-
formed after each incremental addition of salt. Size data was
only recorded for samples with high enough scatter counts
PEG-b-P(TMAEMA)
1
H NMR (␦, ppm, CD OD): 4.56 (33H, broad s), 4.09
3
(32H, broad s), 3.64–3.36 (181 H, m + 150 H, broad s + 3H, s),
2
.07 (32H, broad s), 1.31–1.06 (6H, s + 51H, m). Yield: 76%,
(above the detection limit).
after purification.
Effect of Freeze-Drying on PICM Integrity
PEG-b-P(DEMAEMA)
1
Stoichiometric PICM of heparin were prepared in an ac-
etate buffer (67.8 mM sodium acetate, 32.2 mM acetic acid,
pH 5) to final polymer concentrations of 2.5 mg/ml. Compar-
atively, stoichiometric PICM of heparin were prepared as
acidified aqueous solutions (pH ∼5) with final polymer con-
centrations of 2.5 mg/ml. Solutions were passed through 0.45-
H NMR (␦, ppm, CD OD): 4.54 (24H, broad s), 3.96
3
(
3
24H, broad s), 3.64 (47H, broad s + 170H, m), 3.36 (3H, s),
.31 (CD OD + broad s), 2.06 (24H, broad s), 1.47 (72H,
broad s), 1.23–1.09 (6H, s + 37H, m). Yield: 59%, after puri-
fication.
3
␮
m nylon filters prior to initial DLS analysis. The samples
Potentiometric Titrations
were then frozen overnight at −80°C and freeze-dried for 48
h. Changes in micelle size were evaluated by DLS after re-
suspension of the dried cakes in deionized water.
Titration curves were generated by first titrating the
polymer solutions (1 mg/ml) to low pH using 0.01 N HCl to
ensure complete dissolution of the polymers. Changes in pH
values were then monitored with a portable Accumet AP61
pH-meter (Fisher Scientific, Montreal, Qc, Canada) following
RESULTS AND DISCUSSION
incremental additions of a standardized 0.01 N NaOH solu- Polymer Synthesis
tion. The apparent pK values were calculated from Eq. (3):
a
ATRP is a radical polymerization method that allows the
preparation of polymers with controlled and diverse compo-
sitions. Its success lies in its ability to combine fast initiation
rates with low concentrations of propagating radicals. This
1
− ␣
␣
pK = pH − logͩ ͪ
(3)
a
␣
is defined as the ratio CN+/C , where C stands for the critical equilibrium between growing radicals and dormant
m
N+
effective concentration of protonated amino groups and Cm is species is a function of the nature of the transition metal–
the monomeric concentration in methacrylates. Assuming ligand catalyst complex used. It then becomes of crucial im-
that all added hydroxide ions are able to deprotonate the portance that the species to be reacted do not interfere with
amine groups, a fair approximation of CN+ is calculated using the ATRP equilibrium. Accordingly, the polymerization of
the effective concentration of NaOH added (23).
methacrylates presenting pendant amino groups becomes
challenging for two reasons. First, displacement of the ligand
on the metal complex by the growing polymer chain could
occur. ATRP ligands often present amino groups used for
coordination to the metal catalyst; monomers bearing amines
Effect of Heparin/Polymer Molar Charge Ratio on
Micelle Formation
Specific amounts of a concentrated heparin solution (60 are direct competitors to the ligand and can disturb the sen-
mg/ml) were added to a solution of polymer (5 mg/ml) in an sitive metal–ligand equilibrium. This problem can be over-
acetate buffer (16.6 mM sodium acetate, 83.4 mM acetic acid, come by using polydentate ligands (21), given that their che-
pH 4) to yield complexes with −/+ molar charge ratios varying lating effect confers them an enhanced stability. PMDETA is
from 0 to 2.95. The density of charges on heparin was as- one such polydentate ligand and was selected as it presents
sumed to be of 5 mEq/g. DLS analyses were performed after both fast rates of activation and deactivation, thus rendering
each incremental addition of heparin. Both the stock heparin its metal complexes very active catalysts (24). Second, side
and polymer solutions were passed through 0.22-␮m nylon reactions between the free amines and the halogen end group
filters prior to experimentation.
of the macroinitiator and/or growing polymer chain are to be

1
64
Dufresne and Leroux
anticipated. Coessens et al. (19) have shown that bromide end
groups are available for nucleophilic substitution reactions to
yield polymers with amino end groups. These authors addi-
tionally revealed that under the experimental conditions used
during typical ATRP processes, interactions of the end bro-
mine group with tertiary amines were negligible whereas they
could take place with primary amines. The combination of
these adverse interactions advocates that great care must be
taken during the ATRP of amine-bearing monomers and calls
for the protection of the primary amine groups. In this light,
the tert-butoxycarbonyl (Boc) protective group was intro-
duced on AEMA prior to its polymerization. Boc was se-
lected as it reduces the accessibility of the lone amine electron
pair through both steric hindrance and its electron-
withdrawing activity. Furthermore, Boc can easily be cleaved
under mild acidic condition (3 M HCl/EtOAc), leaving the
ester bonds of the polymer intact. Under proper reaction con-
ditions, ATRP allowed for the preparation of well-defined
cationic copolymers originating from a PEG-based unit.
Characterization of Diblock Copolymers
The MW characteristics and polydispersity indexes (PIs)
of the polymers are reported in Table I. M s were evaluated
n
1
by H NMR from the integration of the two methylene groups
of the pendant aminoethyl segments with respect to that of
1
the methoxy group of the PEG macroinitiator. The H NMR
spectra of PEG-b-P(AEMABoc) and its derivative PEG-b-
P(AEMA) are shown as examples in Fig. 2. Figure 2B con-
firms the complete deprotection of the primary amine
through the disappearance of the peak at 1.45 ppm (assigned
1
Fig. 2. H NMR spectra of (A) diblock copolymer PEG-b-
P(AEMABoc) in CDCl and (B) of its deprotected equivalent PEG-
b-P(AEMA) in D O.
3
to the Boc tert-butyl group). Note that the –NH signal is lost
2
2
given that these protons are easily exchanged with the sur-
rounding deuterated solvent (D O). M s were also deter- THF] are suitable for the polymerization of both the AE-
2
n
mined by GPC relative to PEG standards. MWs calculated MABoc and DMAEMA derivatives but are inadequate for
from both techniques reasonably correlated, with the GPC DEAEMA; degrees of conversion of PEG-b-P(DEAEMA)
results being invariably underestimated. GPC analyses fur- are maximized at ∼80–88% of the feed.
ther allowed for the evaluation of PIs. PIs were relatively low
Potentiometric Titrations
(
Յ1.23), as expected for controlled polymerization processes.
Table I additionally reveals that the reaction conditions [i.e.,
PEG-b-P(TMAEMA) and PEG-b-P(DEMAEMA) co-
in presence of the Cu(I)Br/PMDETA complex at 65°C in polymers were not titrated as their amino groups are quater-
Table I. Characteristics of the Diblock Copolymers and Their Heparin-Based PICM
Conversion
degree
(Mn,NMR
PICM
a
b
c
Mn
Mn
Mn
/
diameter
PICM
MW
Aggregation
number
Rehydrated PICM
diameter (nm) [PI]
d
e
e
f
e
Description
(theo) (NMR) (GPC)
PI
Mn,theo
)
(nm) [PI]
g
PEG-b-P(AEMA)
5800
4700
5800
4900
3800
3900
1.17
1.23
1.00
1.04
24.7 [0.159] 340,000
30.8 [0.110] 446,000
41
53
24.2 [0.160]
44.3 (19.9%)
PEG-b-P(DMAEMA)
h
1
89.6 (83.1%)
PEG-b-P(TMAEMA)
PEG-b-P(DEAEMA)
PEG-b-P(DEMAEMA)
6900
5100
7400
7400
4500
5900
—
4700
—
—
1.16
—
1.07
0.88
0.80
30.6 [0.080] 496,000
32.4 [0.147] 692,000
31.6 [0.086] 380,000
45
80
40
35.4 [0.193]
342.1 [1.000]
40.2 [0.301]
PICM, polyion complex micelles; M , number-average molecular weight; PI, polydispersity index; MW, molecular weight; theo, theoretical.
n
a
Calculated from feed composition.
Evaluated by H NMR.
b
c
d
e
f
1
Evaluated by GPC relative to PEG standards.
Polydispersity index derived using MWs obtained by relative analysis.
Stoichiometric PICM were prepared in an acetate solution (pH 5.0).
The aggregation number refers to the number of polymer chains per micelle and was calculated by dividing the PICM MW by the average
MW of a single stoichiometric polymer–heparin condensate.
g
MW data is that of PEG-b-P(AEMABoc) whereas the PICM were synthesized using PEG-b-P(AEMA).
Numbers in parentheses represent the population distribution. PIs were not available.
h

Complexation of Amino Copolymers with Heparin
165
nized. Figure 3A illustrates the variations in pH expressed curves with drops of their apparent pK by unity as ␣ varies
a
against ␣ for the PEG-b-P(AEMA), PEG-b-P(DMAEMA), from 0 to 1. Such profiles correlate well to the classical coiled/
and PEG-b-P(DEAEMA) copolymers. The copolymers all extended transition brought about by the ionization of the
behave as weak bases and present buffering capacities, as amino groups. However, divergences were observed for the
evidenced by the presence of plateaus. pK values were di- PEG-b-P(DEAEMA) copolymer. The apparent pKa vs. ␣
a
rectly extracted from the titration curves and correspond to curve presented a plateau in the region of neutrality (i.e.,
the pH value at ␣ ס 0.5. pK s of 7.13 ± 0.01, 6.58 ± 0.07, and up to 18% ionization) indicating that no conformational
a
6
.75 ± 0.02 were obtained for the conjugate acids of PEG-b- changes took place. This can be rationalized by the fact that
P(AEMA), PEG-b-P(DMAEMA), and PEG-b- the DEAEMA segments are sufficiently hydrophobic to trig-
P(DEAEMA), respectively. pK values measured for PEG- ger the self-association of the block copolymer chains into
a
b-P(DMAEMA) and PEG-b-P(DEAEMA) are in agreement micelles (see section entitled “Effect of pH on Micelle For-
with those found in the literature (23,25). These results indi- mation,” below, for further details). Electrostatic repulsions en-
cate that PEG-b-P(AEMA) should form polyelectrolyte com- suing from ionization of the amines eventually led to dissocia-
plexes over a broader pH range as it is the most basic of all tion of the micelles and extension of the polymeric chains.
three copolymers.
In a past study, P rˇ ádný and Še␯ cˇ ík (25) monitored the Characterization of PICM
potentiometric behavior of PDMAEMA in water–ethanol so-
lutions and observed a dependence between the apparent pKa
and ␣ that was attributed to conformational transitions. The
authors explained that as ␣ varies from 0 to 1, the conforma-
Effect of Heparin/Polymer Molar Charge Ratio on
Micelle Formation
tion of the polymeric chains changes from statistical coils to
Variations in micelle size, distribution, and scattering in-
extended chains due to electrostatic repulsions between the tensity of heparin-based PICM were monitored by DLS as a
amino groups. Figure 3B plots the apparent pK values of function of the heparin/polymer molar charge ratio. Com-
a
PEG-b-P(AEMA), PEG-b-P(DMAEMA), and PEG-b- plexes were prepared in acidic buffered solutions (pH 4) to
P(DEAEMA) and illustrates their dependence on ␣. Both ensure that every amino group of the polymers was proto-
PEG-b-P(AEMA) and PEG-b-P(DMAEMA) exhibit similar nated. Heparin remains ionized in these conditions given that
its least acidic group has a pK of 3.13 (26). Figure 4A spe-
a
cifically illustrates the dependence of PEG-b-P(AEMA)
PICM on the heparin feed. This titration profile was roughly
subdivided into three main regions, the first of which corre-
sponds to an excess of polymer (domain I). The initial scat-
tering intensity (drug-free polymer solution) was very low,
indicating that the polymer was soluble in water and that its
chains existed as unimers. Addition of heparin slowly led to
increases in both particle size and scattering intensity. This
behavior is supportive of a mechanism where micelle forma-
tion is driven by electrostatic interactions. Careful examina-
tion of these first assemblies most often revealed the presence
of two populations and was associated to high PIs. It is as if
both free polymeric chains and PICM coexisted simulta-
neously.
Subsequent addition of heparin eventually resulted in a
domain of near stoichiometry (domain II). In this region, ev-
ery ionized polymer and heparin units electrostatically inter-
acted to yield neutral PICM with narrow size distributions.
PIs lower than 0.1 were observed and confirmed that the
PICM were evenly distributed and monodisperse. A typical
representation of the scattering measurements is presented in
Fig. 4B and illustrates the narrowness of the size distribution
of the complexes. An AFM image of the PEG-b-P(AEMA)–
heparin complexes is presented in Fig. 5. It can be seen that
the complexes most often have spherical shapes (∼20 nm) and
that their size distribution is narrow. The MW and aggrega-
tion number of similar stoichiometric assemblies were evalu-
ated by MASLS (Table I). It was found that PICM were
composed of 40–80 polymer chains and that the aggregation
number increased with increasing hydrophobicity of the mo-
nomeric core units.
Fig. 3. (A) pH and (B) apparent pK of (छ) PEG-b-P(AEMA), (ᮀ)
PEG-b-P(DMAEMA), and (᭝) PEG-b-P(DEAEMA) copolymers
a
The third region of the heparin titration profile (Fig. 4A)
in turn corresponded to the situation where a marked excess
as a function of ␣. Polymer solutions (1 mg/ml) were first acidified
with HCl and titrated with 0.01 N NaOH. Complete ionization occurs in heparin was met (domain III). Increases in the ratio of
at ␣ ס 1. heparin progressively induced a decrease in micelle size (data

1
66
Dufresne and Leroux
gesting that their chains existed as unimers. In these condi-
tions, the copolymers are completely unionized and are un-
able to interact electrostatically with heparin. Lowering the
pH progressively protonated the amines and induced electro-
static interactions at a critical pH value. Micellization was
translated as increases in both size and scattering intensity.
This threshold pH value varies between polymers and closely
correlates to the basicity of the amino groups (i.e., the pH at
which they get ionized). It can be extrapolated from Fig. 3A
that ionization of PEG-b-P(AEMA) and PEG-b-
P(DMAEMA) begins at pH 9.4 and 8.9, respectively; these
values intimately correspond to the pH at which their con-
densation to heparin was initiated (Fig. 6A, arrows 1 and 2).
PICM were observable from that point on because both the
copolymers and heparin existed as ionized species. Control
measurements performed on solutions of PEG-b-P(AEMA),
PEG-b-P(DMAEMA), or heparin demonstrated that the
components individually existed as unimers throughout the
whole pH range studied (data not shown).
PEG-b-P(DEAEMA) is peculiar in that pH has an effect
on the self-assembly of the polymer itself (Fig. 6A). In the
absence of heparin, the unionized DEAEMA units were suf-
ficiently hydrophobic to trigger their segregation into the core
of micelles (23) with diameters of ∼20 nm. PEG-b-
P(DEAEMA) chains retained their ability to form micelles
up to 18% ionization (pH 7.4), after which they broke apart.
This property of PEG-b-P(DEAEMA) greatly altered the
pH-responsiveness profile of the PEG-b-P(DEAEMA)–
heparin system. As opposed to the PEG-b-P(AEMA) and
PEG-b-P(DMAEMA) copolymers, the PEG-b-
P(DEAEMA) chains gathered, in basic conditions, into mi-
celles from which the polyanionic heparin was most likely
precluded. Acidifying the milieu gradually protonated the
amino groups, allowing for the incorporation of heparin mol-
ecules and the formation of larger polyelectrolyte assemblies.
Once again, the pH at which ionization of the polymer begins
(pH 9.0, Fig. 3A) compares well to the pH at which electro-
static interactions appear (arrow 3).
Fig. 4. (A) Effect of heparin/polymer molar charge ratio on (छ) the
scattering intensity and (ࡗ) polydispersity index of micelles prepared
in an acetate buffer of pH 4. The scattering intensity is expressed
relative to the maximum scattering count observed. The curve was
further subdivided in three regions corresponding respectively to do-
mains of (I) excess polymer, (II) near stoichiometry, and (III) excess
heparin. (B) A typical size distribution of stoichiometric PEG-b-
P(AEMA)–heparin PICM prepared in an acetate buffer of pH 4 is
also shown.
not shown), and, concomitantly, scattering intensity. These
micelles retained an extremely narrow size distribution and
exhibited a single population, unlike micelles formed under
an excess of polymer. Such observations suggest that micelle
formation proceeded in a nonstoichiometric way in that the
polymeric chains evenly distributed on the heparin molecules
to yield assemblies with excess negative charges. Similar pro-
files were obtained for the titration of PEG-b-P(DMAEMA),
PEG-b-P(TMAEMA), PEG-b-P(DEAEMA), and PEG-b-
P(DEMAEMA) with heparin (data not presented).
Effect of pH on Micelle Formation
It was evidenced from the potentiometric titrations (Fig.
3
) that the copolymers presented different basic properties.
Their capability to form polyelectrolyte complexes at a given
pH should therefore relate to their composition. The latter
was investigated by determining the influence of pH (and ␣)
on micelle formation. Figure 6A presents variations in scat-
tering intensity following changes in pH for PEG-b-
P(AEMA), PEG-b-P(DMAEMA), and PEG-b-
P(DEAEMA) while variations in particle size are shown in
Fig. 5. AFM image of PEG-b-P(AEMA)–heparin PICM. Polyelec-
trolyte complexes were prepared as an acidified aqueous solution
Fig. 6B. The scattering counts of both PEG-b-P(AEMA) and (pH 5) at a −/+ molar charge ratio of 1.5. Imaging was performed in
PEG-b-P(DMAEMA) were low in basic environment, sug- tapping mode in air.

Complexation of Amino Copolymers with Heparin
167
scattering intensity, thus indicating the dissociation of the
complexes. Of all polymers, PEG-b-P(AEMA) was the most
stable, exhibiting 94% of its initial scattering intensity under
physiological isotonic conditions (0.15 M NaCl).
Effect of Freeze-Drying on PICM Integrity
It is expected that aqueous solutions of heparin-based
PICM will have limited stability due to possible chemical and
physical degradation processes. Lyophilization is a method of
choice to grant acute stability and suitable shelf-life to labile
therapeutic systems. For example, lyophilization was success-
fully used to store oligonucleotide–PEG-polyamine com-
plexes without modifying the size of the complexes (29). Its
effect on the integrity of PICM prepared in an acetate buffer
(pH 5) was thus investigated and is summarized in Table I. It
can readily be seen that the PEG-b-P(AEMA), PEG-b-
P(TMAEMA), and PEG-b-P(DEMAEMA) complexes with-
stood both the freezing and drying procedures. The resus-
pended carriers indeed present unaltered or slightly increased
diameters. The greatest impact of lyophilization lied in in-
creased PIs for the rehydrated PICM. PEG-b-P(DMAEMA)
and PEG-b-P(DEAEMA) complexes, however, did not tol-
erate freeze-drying as satisfactorily. Resuspended PEG-b-
P(DMAEMA) PICM presented two populations while there
was a marked precipitation of the PEG-b-P(DEAEMA) com-
Fig. 6. Effect of pH on (A) scattering intensity and (B) particle di-
ameter of (छ) PEG-b-P(AEMA), (ᮀ) PEG-b-P(DMAEMA), and
(
᭝) PEG-b-P(DEAEMA) based PICM. The effect of pH on the
scattering intensity and particle diameter of (᭢) PEG-b-
P(DEAEMA) chains (without heparin) is also shown. Arrows indi-
cate the onset of ionization and complexation. PICM complexes were
prepared as aqueous solutions with −/+ molar charge ratios of 1.5.
The scattering intensity is expressed relative to the maximum scat-
tering count observed per polymer.
Effect of Ionic Strength on Micelle Stability
It is predicted that the stability of PICM will strongly be
dependent on the ionic strength of the surroundings. The
presence of salts can indeed compete with the polymer–
heparin electrostatic interactions, thereby destabilizing the
PICM (27). The influence of salt concentration on the stabil-
ity of micelles was therefore studied by measuring the scat-
tering intensity (Fig. 7A) and size (Fig. 7B) of micelles at
various NaCl concentrations. As a general trend, an increase
in ionic strength was associated with a slight increase in the
mean diameters of the micelles, up to NaCl concentrations of
about 0.1 M. The most probable explanation is that the com-
plexes presented a slight net charge that provided electro-
static repulsions between polymer chains and limited their
self-association. Addition of salt screened these charges, re-
duced electrostatic repulsions and allowed for more chains to
aggregate into the core of larger micelles (28). This effect was
particularly pronounced for the PICM of PEG-b-
P(DEAEMA) given that the shielded DEAEMA units are
Fig. 7. Effect of ionic strength on (A) scattering intensity and (B)
particle diameter of (छ) PEG-b-P(AEMA), (ᮀ) PEG-b-
P(DMAEMA), (᭿ PEG-b-P(TMAEMA), (᭝) PEG-b-
P(DEAEMA), and (᭡) PEG-b-P(DEMAEMA) based PICM. Poly-
electrolyte complexes were prepared as acidified aqueous solutions
the most hydrophobic. Further increases in NaCl concentra- (pH 4) with −/+ molar charge ratios of 1. The scattering intensity is
tions eventually led to decreases in both micelle size and light expressed relative to the initial scattering count.

1
68
Dufresne and Leroux
plexes. This phenomenon was attributed to the presence of
salts in the preparations. To illustrate this hypothesis, we pre-
pared complexes as acidic aqueous solutions and showed that
the lyophiles presented sizes and population distributions
comparable to that of the mother complexes (data not
shown). One possible explanation is that there is an increase
in the local concentration of polymer, heparin, and buffer
salts at the onset of freezing. It is known that as the prepa-
ration cools down, it reaches a temperature where pure ice
first begins to crystallize out of solution and induces the iso-
lation and concentration of the solution components (30).
Increased concentrations in buffer salts can contribute to par-
tial dehydration of the PEG corona and interpenetration of
micelle cores (31), thereby accounting for the formation of
larger micelles and even micellar aggregation. This mecha-
nism is in agreement with the observation that the aggrega-
tion of the lyophiles increased with increasing hydrophobicity
of the complexes.
Catella-Lawson, and G. A. FitzGerald. Oral delivery of antico-
agulant doses of heparin. A randomized, double-blind, controlled
study in humans. Circulation. 98:1610–1615 (1998).
. Y. Jiao, N. Ubrich, V. Hoffart, M. Marchand-Arvier, C. Vign-
eron, M. Hoffman, and P. Maincent. Anticoagulant activity of
heparin following oral administration of heparin-loaded micro-
particles in rabbits. J. Pharm. Sci. 91:760–768 (2002).
. Y. Jiao, N. Ubrich, M. Marchand-Arvier, C. Vigneron, M. Hoff-
man, T. Lecompte, and P. Maincent. In vitro and in vivo evalu-
ation of oral heparin-loaded polymeric nanoparticles in rabbits.
Circulation. 105:230–235 (2002).
8. K. Kataoka, A. Harada, and Y. Nagasaki. Block copolymer mi-
celles for drug delivery: design, characterization and biological
significance. Adv. Drug Deliv. Rev. 47:113–131 (2001).
. A. V. Kabanov, T. K. Bronich, V. A. Kabanov, K. Yu, and A.
Eisenberg. Soluble stoichiometric complexes from poly(N-ethyl-
4-vinylpyridinium) cations and poly(ethylene oxide)-block-
polymethacrylate anions. Macromolecules 29:6797–6802 (1996).
0. A. Harada, H. Togawa, and K. Kataoka. Physicochemical prop-
erties and nuclease resistance of antisense-oligonucleotides en-
trapped in the core of polyion complex micelles composed of
poly(ethylene glycol)-poly(L-lysine) block copolymers. Eur. J.
Pharm. Sci. 13:35–42 (2001).
6
7
9
1
CONCLUSIONS
11. K. Kataoka, H. Togawa, A. Harada, K. Yasugi, T. Matsumoto,
and S. Katayose. Spontaneous formation of polyion complexe
micelles with narrow distribution from antisense oligonucleotide
and cationic block copolymer in physiological saline. Macromol-
ecules 29:8556–8557 (1996).
ATRP was successfully used to prepare copolymers pre-
senting primary, tertiary, and quaternary amino groups. Elec-
trostatic interactions between the amino copolymers and hep-
arin triggered the formation of small and monodisperse poly-
electrolyte complexes with properties dependent on the
1
2. Y. Kakizawa and K. Kataoka. Block copolymer micelles for de-
livery of gene and related compounds. Adv. Drug Deliv. Rev.
54:203–222 (2002).
substitution of their constituents. For instance, copolymers 13. A. Harada and K. Kataoka. Novel polyion complex micelles en-
trapping enzyme molecules in the core: preparation of narrowly-
bearing primary amines were shown to form complexes under
the widest pH range and to yield assemblies stable to ionic
strength variations. Hydrophobicity was also shown to play a
distributed micelles from lysosyme and poly(ethylene glycol)-
poly(aspartic acid) block copolymer in aqueous medium. Macro-
molecules 31:288–294 (1998).
role, generally inducing the formation of complexes of larger 14. O. Boussif, F. Lezoualc’h, M. A. Zanta, M. D. Mergny, D.
Scherman, B. Demeneix, and J-P. Behr. A versatile vector for
diameters. Future studies will aim at relating the stability of
gene and oligonucleotide transfer into cells in culture and in vivo:
the PICM to the oral bioavailability of heparin. The combi-
polyethylenimine. Proc. Natl. Acad. Sci. U. S. A. 92:7297–7301
nation of conditions such as increases in ionic strength and
(
1995).
basification of the surroundings as the complexes transit from 15. J. F. Kukowska-Latallo, A. U. Bielinska, J. Johnson, R. Spindler,
the stomach to the jejunum are expected to induce the in vivo
dissociation of the complexes (along with heparin release).
Furthermore, the release profile should also be subject to
extreme dilution and possible electrostatic interactions be-
tween the cationic blocks and the negatively charged intesti-
nal mucosa and mucus.
D. A. Tomalia, and J. R. Baker, Jr. Efficient transfer of genetic
material into mammalian cells using starburst polyamidoamine
dendrimers. Proc. Natl. Acad. Sci. U. S. A. 93:4897–4902 (1996).
6. U. Rungsardthong, M. Deshpande, L. Bailey, M. Vamvakaki, S.
P. Armes, M. C. Garnett, and S. Stolnik. Copolymers of amine
methacrylate with poly(ethylene glycol) as vectors for gene
therapy. J. Controlled Release 73:359–380 (2001).
7. K. A. Davis and K. Matyjaszewski. ABC triblock copolymers
prepared using atom transfer radical polymerization techniques.
Macromolecules 34:2101–2107 (2001).
1
1
1
ACKOWLEDGMENTS
8. S. Angot, D. Taton, and Y. Gnanou. Amphiphilic stars and den-
drimer-like architectures based on poly(ethylene oxide) and
polystyrene. Macromolecules 33:5418–5426 (2000).
Financial support from the Natural Sciences and Engi-
neering Research Council of Canada and the Canada Re-
search Chair Program is acknowledged. The authors extend 19. V. Coessens and K. Matyjaszewski. Synthesis of polymers with
amino end groups by atom transfer radical polymerization. J.
Macromol. Sci. Pure Appl. Chem. A36:811–826 (1999).
0. K. Matyjaszewski. Mechanistic and synthetic aspects of atom
transfer radical polymerization. J. Macromol. Sci. Pure Appl.
Chem. A34:1785–1801 (1997).
their gratitude to Marc A. Gauthier and Marie-Christine
Jones for their critical reading of the manuscript and helpful
discussions.
2
REFERENCES
21. V. Coessens, T. Pintauer, and K. Matyjaszewski. Functional poly-
mers by atom transfer radical polymerization. Prog. Polym. Sci.
26:337–377 (2001).
22. M. Ranger, M-C. Jones, M-A. Yessine, and J-C. Leroux. From
well-defined diblock copolymers prepared by a versatile atom
transfer radical polymerization method to supramolecular assem-
blies. J. Polym. Sci. A: Polym. Chem. 39:3861–3874 (2001).
1
. E. Windsor and G. E. Cronheim. Gastro-intestinal absorption of
heparin and synthetic heparinoids. Nature 190:263–264 (1961).
. R. H. Engel and S. J. Riggi. Effect of sulfated and sulfonated
surfactants on the intestinal absorption of heparin. Proc. Soc.
Exp. Biol. Med. 130:879–884 (1969).
2
3
. Y-K. Lee, J. H. Nam, H-C. Shin, and Y. Byun. Coagulation of 23. A. S. Lee, A. P. Gast, V. Bütün, and S. P. Armes. Characterizing
low-molecular-weight heparin and deoxycholic acid for the de-
the structure of pH dependent polyelectrolyte block copolymer
micelles. Macromolecules 32:4302–4310 (1999).
24. K. Matyjaszewski, B. Göbelt, H-J. Paik, and C. P. Horwitz. Tri-
velopment of a new oral anticoagulant agent. Circulation. 104:
3116–3120 (2001).
4
. N. Sakuragawa, K. Takahashi, M. Ueno, and I. Horikoshi. Oral
administration of heparin. Acta Med. Biol. 29:33–37 (1981).
dendate nitrogen-based ligands in Cu-based ATRP: a structure-
activity study. Macromolecules 34:430–440 (2001).
a)
5
. R. A. Baughman, S. C. Kapoor, R. K. Agarwal, J. Kisicki, F. 25. M. P rˇ ádný and S. Šev cˇ ík. Precursors of hydrophilic polymers, 3

Complexation of Amino Copolymers with Heparin
169
The potentiometric behavior of isotactic and atactic poly(2-
dimethylaminoethyl methacrylate) in water/ethanol solutions.
Makromol. Chem. 186:111–121 (1985).
charge and ionic strength dependence. Macromolecules 35:8540–
8551 (2002).
2
3
9. S. V. Vinogradov, T. K. Bronich, and A. V. Kabanov. Self-
assembly of polyamine-poly(ethylene glycol) copolymers with
phosphorothioate oligonucleotides. Bioconjug. Chem. 9:805–812
2
2
6. H. M. Wang, D. Loganathan, and R. J. Linhardt. Determination
of the pK of gluronic acid and the carboxy groups of heparin by
a
1
3
C-nuclear-magnetic-resonance spectroscopy. Biochem. J. 278:
89–695 (1991).
(
1998).
6
0. T. W. Randolph. Phase separation of excipients during lyophili-
7. A. Harada and K. Kataoka. Formation of stable and monodis-
persive polyion complex micelles in aqueous medium from
poly(L-lysine) and poly(ethylene glycol)-poly(aspartic acid) block
zation: effects on protein stability. J. Pharm. Sci. 86:1198–1203
(
1997).
copolymer. J. Macromol. Sci. Pure Appl. Chem. A34:2119–2133 31. N. J. Jain, V. K. Aswal, P. S. Goyal, and P. Bahadur. Salt induced
(
1997).
micellization and micelle structures of PEO/PPO/PEO block co-
polymers in aqueous solution. Colloids Surf. A: Physicochem.
Eng. Asp. 173:85–94 (2000).
2
8. A. S. Lee, V. Bütün, M. Vamvakaki, S. P. Armes, J. A. Pople, and
A. P. Gast. Structure of pH-dependent block copolymer micelles: