Design of folic acid-conjugated nanoparticles for drug targeting

Article abstract of DOI:10.1002/1520-6017(200011)89:11<1452::AID-JPS8>3.0.CO;2-P

The new concept developed in this study is the design of poly(ethylene glycol) (PEG)-coated biodegradable nanoparticles coupled to folic acid to target the folate-binding protein; this molecule is the soluble form of the folate receptor that is overexpressed on the surface of many tumoral cells. For this purpose, a novel copolymer, the poly[aminopoly(ethylene glycol)cyanoacrylate-co-hexadecyl cyanoacrylate] [poly(H₂NPEGCA-co-HDCA)] was synthesized and characterized. Then nanoparticles were prepared by nanoprecipitation of the obtained copolymer, and their size, zeta potential, and surface hydrophobicity were investigated. Nanoparticles were then conjugated to the activated folic acid via PEG terminal amino groups and purified from unreacted products. Finally, the specific interaction between the conjugate folate-nanoparticles and the folate-binding protein was evaluated by surface plasmon resonance. This analysis confirmed a specific binding of the folate-nanoparticles to the folate-binding protein. This interaction did not occur with nonconjugated nanoparticles used as control. Thus, folate-linked nanoparticles represent a potential new drug carrier for tumor cell-selective targeting. (C) 2000 Wiley-Liss, Inc.

Full text of DOI:10.1002/1520-6017(200011)89:11<1452::AID-JPS8>3.0.CO;2-P

Design of Folic Acid-Conjugated Nanoparticles for
Drug Targeting
2
1
3
4
BARBARA STELLA,1,2 SILVIA ARPICCO, MARIA TERESA PERACCHIA, * DIDIER DESMAELE, JOHAN HOEBEKE,
¨
MICHEL RENOIR,1 JEAN D’ANGELO,3 LUIGI CATTEL,2 PATRICK COUVREUR1
1 Universite´ Paris-Sud XI, Physico-Chimie–Pharmacotechnie–Biopharmacie, UMR CNRS 8612-5, rue J. B. Cle´ment,
92296 Chaˆtenay-Malabry, France
2 Dipartimento di Scienza e Tecnologia del Farmaco, 10125 Torino, Italy
3 URA CNRS 1843, Chimie Organique, 92296 Chaˆtenay-Malabry, France
4 Universite´ de Strasbourg, UPR CNRS 9021, I.B.M.C., 67084 Strasbourg, France
Received 17 March 2000; accepted 1 August 2000
ABSTRACT: The new concept developed in this study is the design of poly(ethylene
glycol) (PEG)-coated biodegradable nanoparticles coupled to folic acid to target the
folate-binding protein; this molecule is the soluble form of the folate receptor that is
overexpressed on the surface of many tumoral cells. For this purpose, a novel copoly-
mer, the poly[aminopoly(ethylene glycol)cyanoacrylate-co-hexadecyl cyanoacrylate]
[poly(H2NPEGCA-co-HDCA)] was synthesized and characterized. Then nanoparticles
were prepared by nanoprecipitation of the obtained copolymer, and their size, zeta
potential, and surface hydrophobicity were investigated. Nanoparticles were then con-
jugated to the activated folic acid via PEG terminal amino groups and purified from
unreacted products. Finally, the specific interaction between the conjugate folate–
nanoparticles and the folate-binding protein was evaluated by surface plasmon reso-
nance. This analysis confirmed a specific binding of the folate–nanoparticles to the
folate-binding protein. This interaction did not occur with nonconjugated nanoparticles
used as control. Thus, folate-linked nanoparticles represent a potential new drug car-
rier for tumor cell-selective targeting. © 2000 Wiley-Liss, Inc. and the American Pharmaceu-
tical Association J Pharm Sci 89: 1452–1464, 2000
Keywords: poly(ethylene glycol); poly(cyanoacrylate); nanoparticles; folic acid; sur-
face plasmon resonance
proach to achieve controlled release, prevent drug
degradation, and avoid toxic effects. However, ef-
ficient drug delivery with these systems may be
compromised by two factors: a short blood half-
life (rapid elimination from the blood stream) and
nonspecific targeting.
The rapid elimination from the blood stream is
due to the recognition by the macrophages of the
mononuclear phagocyte system (MPS) as a conse-
quence of the adsorption of blood proteins (op-
sonins) onto the surface of colloidal carriers. This
fact causes the accumulation of particulate drug
carrier in the MPS organs, such as the liver and
INTRODUCTION
This paper proposes a new colloidal drug carrier
obtained by chemical coupling of folic acid at its
surface. Colloidal drug delivery systems for intra-
venous administration, such as liposomes and
nanoparticles, represent a very attractive ap-
Correspondence to: P. Couvreur (Telephone: 33-1-46-83-53-
96; Fax: 33-1-46-61-93-34; E-mail: patrick.couvreur@cep.u-
psud.fr)
Present address: Aventis, 94403 Vitry sur Seine, France
Journal of Pharmaceutical Sciences, Vol. 89, 1452–1464 (2000)
© 2000 Wiley-Liss, Inc. and the American Pharmaceutical Association
1452
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

FOLATE-CONJUGATED POLY(CYANOACRYLATE) NANOPARTICLES
1453
the spleen.1 To obtain long-circulating colloidal
drug carriers, it is possible to modify their surface
with hydrophilic, flexible, and nonionic polymers,
such as poly(ethylene glycol) (PEG).2–4
ing real-time analysis of the molecular associa-
tion.20
EXPERIMENTAL SECTION
Materials
To solve the problem of site-specific targeting,
some authors have attempted to increase the tis-
sue specificity of colloidal drug carriers by cou-
pling targeting agents, such as monoclonal anti-
bodies.5–8 However, this approach has at least
two drawbacks: the overall dimensions of the
antibodies, which cause particles to diffuse poorly
through biological barriers, and their immunoge-
nicity. A way to solve these problems is to use
small nonantigenic ligands. However, to our
knowledge, no attempts have been made in cou-
pling nanoparticles to small ligands for tumor tar-
geting. Among the possible low molecular weight
(MW) targeting agents, folic acid (MW ס 441 Da)
could be exploited to actively target cancer cells.
Indeed, folic acid is a vitamin whose receptor is
frequently overexpressed on the surface of human
cancer cells.9,10 Therefore, this receptor has been
identified as a tumor marker, especially in ovar-
ian carcinomas,11–13 although it is highly re-
stricted in most normal tissues. In addition, the
folate receptor is efficiently cell internalized after
binding with its ligand (folic acid).14,15 Thus,
folic acid presents advantages as a targeting
agent. First, it is stable, inexpensive, and non-
immunogenic compared with proteins such as
monoclonal antibodies, as already explained. Sec-
ond, folic acid has a very high affinity for its cell-
surface receptor (Kd 1 nM)14 and it moves into
the cell cytoplasm,16–19 which is an advantage for
more efficient intracellular delivery of anticancer
agents than using a cell membrane marker that is
not cell internalized.
Thus, the new concept proposed in this study is
to design PEG-coated biodegradable nanopar-
ticles conjugated to folic acid for the specific rec-
ognition of the soluble form of the folate receptor
expressed at the surface of cancer cells. Here we
describe the synthesis of the poly[aminopoly(eth-
ylene glycol)cyanoacrylate-co-hexadecyl cyanoac-
rylate] [poly(H2NPEGCA-co-HDCA)] copolymer,
the preparation of the nanoparticles using that
polymer, their characterization, the reaction of
conjugation to folic acid, and the purification of
the conjugate. Furthermore, to evaluate the abil-
ity of folic acid to address nanoparticles towards
the soluble form of the folate receptor (also
called folate-binding protein, FBP), we used sur-
face plasmon resonance (SPR) technology, allow-
t-Boc-HNPEG 3400 was purchased from Shear-
water Polymers, Inc. (Huntsville, AL). Cyanoace-
tic acid (purity, 99%) was obtained by Fluka
(Buchs, Switzerland). n-Hexadecanol, trifluoro-
acetic acid (TFA), folic acid, and FBP were pur-
chased from Sigma (St. Quentin, France). All
other reagents were of analytical grade.
Synthesis of the Poly(H2NPEGCA-co-HDCA)
Copolymer
The cyanoacetate esters of n-hexadecanol and of
N-(tert-butyloxycarbonyl)aminopoly(ethylene gly-
col) (t-Boc-HNPEG) 3400 were prepared accord-
ing to the method of Peracchia,3 with minor modi-
fications (Scheme 1). Briefly, for the synthesis of
the n-hexadecyl cyanoacetate (HDCA), n-hexa-
decanol (1.0 g, 4.1 mmol) was dissolved in 25 mL
of dichloromethane and mixed with cyanoacetic
acid (0.38 g, 4.5 mmol) in 2.5 mL of ethyl acetate.
Then, 1,4-(dimethylamino)pyridine (DMAP; cata-
lytic amount) dissolved in dichloromethane was
added. The reaction mixture was cooled at 4 °C,
and 1,3-dicyclohexylcarbodiimide (DCC; 0.93 g,
4.5 mmol) in dichloromethane was added in a
dropwise manner. After stirring at room tempera-
ture under nitrogen for 6 h, hexane was added
and the white solid that formed was filtered off.
The mixture was concentrated and purified by
flash chromatography (silica gel 60, 230–400
mesh, Merck, Darmstadt, D) eluting with hexan
acetate (90:10). The ester was obtained as a white
solid; yield: 1.2 g, 96% mp 51 °C; proton nuclear
magnetic resonance (1H NMR; CDCl3, ␦): 4.20 (t,
2H, COOCH2), 3.46 (s, 2H, CNCH2), 1.68 (m, 2H,
OCH2CH2), 1.26 (s, 26H, CH2), 0.88 (t, 3H, CH3);
electron-impact mass spectroscopy (EIMS; rela-
tive intensity) m/z: 309 (M+, 9), 294 (7), 280 (46),
266 (54), 111 (50), 97 (84), 84 (82), 69 (100), 55
(83), 43 (74).
For the synthesis of t-Boc-HNPEG 3400 cyano-
acetate, the same procedure of Peracchia3 was fol-
lowed, except that N-ethyl-NЈ-(dimethylamino-
propyl)-carbodiimide (EDC) was used instead of
DCC; yield: 89%.
The poly[N-(tert-butyloxycarbonyl)aminopoly-
(ethylene glycol)cyanoacrylate-co-hexadecyl cya-
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

1454
STELLA ET AL.
Scheme 1. Synthesis of the poly(H2NPEGCA-co-HCDA) copolymer.
noacrylate] [poly(t-Boc-HNPEGCA-co-HDCA)] co-
polymer was synthesized by condensation of t-
Boc-HNPEG 3400 cyanoacetate with n-hexadecyl
cyanoacetate (molar ratio, 1:5) in ethanol in the
presence of formalin and dimethylamine, as pre-
viously described for the poly[methoxypoly(ethyl-
ene glycol)cyanoacrylate-co-hexadecyl cyanoacry-
late] [poly(MePEGCA-co-HDCA)]3 and as shown
in Scheme 1; yield: 84%.
polymer (300 mg, 0.06 mmol) dissolved in dry di-
chloromethane (15 mL), and the reaction was car-
ried out for 1 h at room temperature under mag-
netic stirring. The reaction mixture was then
neutralized with 10% aqueous sodium bicarbon-
ate solution and extracted with dichloromethane.
The organic layer was dried over MgSO4 and
evaporated under reduced pressure to give a pale
yellow waxy material.
Then, the amino group of t-Boc-HNPEG was
deprotected by TFA in dichloromethane21
(Scheme 1): TFA (0.6 mL) was added to the co-
The poly(hexadecyl cyanoacrylate) (PHDCA)
and the poly(MePEGCA-co-HDCA) polymers
were also synthesized as described elsewhere.3
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

FOLATE-CONJUGATED POLY(CYANOACRYLATE) NANOPARTICLES
1455
Characterization of the Poly(H2NPEGCA-co-HDCA)
Copolymer
The optical density of the eluted samples was de-
termined at 350 nm with a Beckman spectropho-
tometer (Beckman Inst., Fullerton, CA). The con-
centration of the nanoparticles in the suspension
was based on dry weight analysis.
The poly(H2NPEGCA-co-HDCA) copolymer struc-
ture was confirmed by NMR. The 1H and 13C
NMR spectra were recorded in CDCl3 on a Bruker
AC 200P, 200 MHz spectrometer, with tetrameth-
ylsilane as internal standard.
The MW of the poly(H2NPEGCA-co-HDCA) co-
polymer was measured by gel permeation chro-
matography (GPC). The analysis was performed
in tetrahydrofuran (THF) with a Waters liquid
chromatograph equipped with a Waters 501
pump, 712 WISP, 745 data module, and R401 dif-
ferential refractometer. Polystyrene (Poly-
sciences, Inc., Warrington, PA) was used as the
standard.
Preparation and Purification of Folate–
Poly(H2NPEGCA-co-HDCA)
Nanoparticles Conjugate
Nanoparticles of poly(H2NPEGCA-co-HDCA)
were then conjugated to the folic acid, as shown in
Scheme 2. For this purpose, the N-hydroxysuc-
cinimide ester of folic acid (NHS-folate) was pre-
pared by esterification of folic acid with N-
hydroxysuccinimide (NHS) in dry dimethylsulfox-
ide (DMSO) in the presence of DCC and triethyl-
amine as catalyst, as reported elsewhere;17 the
structure of NHS-folate was confirmed by 1H and
13C NMR spectra.
Then, NHS-folate (23 mg) was dissolved in
DMSO (0.4 mL) and added to the nanoparticles
suspension (4 mL, 5 mg/mL, pH adjusted to 9.0
with 5 mM carbonate/bicarbonate buffer, pH
11.5). The reaction was carried out for 30 min at
room temperature.
To separate the folate-conjugated nanopar-
ticles from unreacted folic acid and other byprod-
ucts, the reaction mixture was purified with a 3 ×
14 cm Sepharose CL-4B column (Pharmacia Bio-
tech., St. Quentin, France) using 5 mM carborate/
bicarbonate buffer, pH 9.0, as eluent. The optical
density of the different collected fractions was
monitored at 365 nm. The folate–nanoparticles,
which eluted in the void volume, were then dia-
lyzed against bicarbonate buffer and MilliQ water
at 4 °C (Spectra/Por௡ 3500 MWCO dialysis mem-
brane, Spectrum., Houston, TX).
The size of folate–nanoparticles conjugate was
determined at 20 °C by QELS, with a nanosizer,
as already detailed, and the stability of the for-
mulation was evaluated by measuring the size of
the particles after 1 month of storage at 4 °C in
MilliQ water. Surface hydrophobicity of the fo-
late-conjugated nanoparticles was investigated
by HIC, with a propyl agarose column, as already
described for the poly(H2NPEGCA-co-HDCA)
nanoparticles before folic acid conjugation.
Preparation and Characterization of
the Nanoparticles
Nanoparticles were prepared by nanoprecipita-
tion.22 Practically, 20 mg of the polymer [PHDCA,
poly(MePEGCA-co-HDCA), or poly(H2NPEGCA-
co-HDCA)] was dissolved in 4 mL of warm ac-
etone, and this solution was added, with magnetic
stirring, to 8 mL of MilliQ water. Precipitation of
particles occurred spontaneously. After solvent
evaporation under reduced pressure, an aqueous
suspension of nanoparticles (2.5 mg/mL) was ob-
tained. The particles were then filtered through a
1.2-␮m filter (Millex௡ AP, Millipore, St. Quentin,
France) and stored at 4 °C.
The size of the nanoparticles was determined
at 20 °C by quasi-elastic light scattering (QELS)
with a nanosizer (Coulter௡ N4MD, Coulter Elec-
tronics, Inc., Hialeah, FL). The selected angle was
90°, and the measurement was made after dilu-
tion of the nanoparticles suspension in MilliQ wa-
ter. The surface charge of the nanoparticles was
evaluated by zeta potential measurements in wa-
ter and in 0.01 M phosphate buffer at pHs rang-
ing from 6.0 to 8.0 (Zetasizer 4, with a multi-8
correlator 7032, Malvern Inst., Malvern, UK).
Surface hydrophobicity of the nanoparticles was
also investigated by hydrophobic interaction chro-
matography (HIC). This analysis was performed
with 1 mL of a nanoparticles suspension, which
was injected into a column (bed volume, 10 mL)
filled with propyl agarose. Nanoparticles were
eluted with phosphate buffer saline (PBS, 53 mM
sodium phosphate, 82 mM NaCl). Particles that
interacted with the gel were removed by washing
with PBS containing Triton௡ X-100 (0.1% w/v).
Determination of Folate Content
The extent of folate conjugation on the folate-
conjugated poly(H2NPEGCA-co-HDCA) nanopar-
ticles was determined by quantitative ultraviolet
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

1456
STELLA ET AL.
Scheme 2. Preparation of the nanoparticles and conjugation with folic acid.
(UV) spectrophotometric analysis of the lyophi-
lized product. The analysis was performed in
DMSO/CH2Cl2 (4:1), and the amount of folic acid
in the conjugate was evaluated by measuring the
absorbance of the product at 358 nm (folic acid
␧ ס 15,760 M−1 cm−1).
the FBP, surface plasmon resonance measure-
ments were performed with the upgraded BI-
Acore 1000 (Pharmacia Biosensor AB, Uppsala,
Sweden). Nonconjugated poly(H2NPEGCA-co-
HDCA) nanoparticles were used as control.
For this purpose, FBP was immobilized on the
sensor surface of a carboxylated, activated, dex-
tran-coated gold film, according to the general
procedure for immobilization by amine coupling
recommended by the BIAcore constructor.23,24
Briefly, after equilibration of the instrument with
Surface Plasmon Resonance Measurements
To study the interaction of folic acid-conjugated
poly(H2NPEGCA-co-HDCA) nanoparticles with
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

FOLATE-CONJUGATED POLY(CYANOACRYLATE) NANOPARTICLES
1457
HBS buffer (10 mM Hepes, 150 mM NaCl, 0.05%
surfactant P20, pH 7.4), the following samples
were automatically and successively injected into
the BIAcore: (i) NHS/EDC in a mixed solution (1:
1, v/v) to activate the carboxylated dextran; (ii)
FBP dissolved at a concentration of 100 ␮g/mL in
100 mM potassium phosphate buffer, pH 7.0,
supplemented with 4 mM mercaptoethanol and
10% (v/v) glycerol; and (iii) 1 M ethanolamine in
water, pH 8.5, to deactivate residual NHS-esters
on the sensor chip. The immobilization protocol,
which was performed at a flow rate of 2 ␮L/min,
allowed the binding of 4 ng/mm2 of FBP per chan-
nel.
The binding of poly(H2NPEGCA-co-HDCA)
nanoparticles (both nonconjugated and folate-
conjugated ones) to the sensorchip was performed
at a flow rate of 5 ␮L/min with a concentration of
5 mg/mL. The nanoparticles were allowed to in-
teract with the FBP for 10 min. In view of the poor
regeneration of FBP in 3 M KSCN, every nano-
particles suspension was tested on newly immo-
bilized FBP. KSCN-denaturated FBP was used as
a control experiment where binding properties of
folate receptor are lost.
the composition and the MW of the copolymer
were the same before and after t-Boc group cleav-
age, as found by NMR and by GPC analysis. The
reaction with TFA, carried out under the condi-
tions described, led to the cleavage of 50% of t-Boc
groups, as found by NMR peaks integration.
Characterization of the
Poly(H2NPEGCA-co-HDCA) Copolymer
1
The analysis by H and 13C NMR (Figures 1 and
2) confirmed the structure of the expected
poly(H2NPEGCA-co-HDCA) copolymer. Figure 1
shows the 1H NMR spectrum of the poly-
(H2NPEGCA-co-HDCA) copolymer: the peak at
4.26 ppm was assigned to the methylene in the
␣-position to the ester groups. The signal at 3.64
ppm was attributed to the PEG backbone methy-
lene, whereas the broad peak at 2.90–2.20 ppm
was assigned to the methylene protons of poly-
(cyanoacrylate). The peak at 1.72 ppm corre-
sponds to the methylene in the ␤-position to the
In a second series of experiments, FBP was im-
mobilized at 6 ng/mm2 and 2.5 mg/mL of control
particles, and folate-conjugated nanoparticles
were allowed to interact at 5 ␮L/min to study the
kinetic parameters of the interaction. Under simi-
lar conditions, using a concentration of 1 mM fo-
late, the kinetic parameters of the free folic acid
were also determined.
RESULTS
Synthesis of the Poly(H2NPEGCA-co-HDCA)
Copolymer
The cyanoacetate esters were synthesized as de-
scribed elsewhere,3 with some modifications. To
protect the PEG amino group during the prepa-
ration of the copolymer, we used a t-Boc-protected
amino PEG, which allowed the synthesis of the
PEG cyanoacetate by reaction of the other distal
end, the hydroxyl group, with cyanoacetic acid, as
shown in Scheme 1. The same procedure was
followed for the synthesis of the hexadecyl cyano-
acetate. Then, the cyanoacetate esters were con-
densed in ethanol, in the presence of formalin
and dimethylamine, to form the poly(t-Boc-
HNPEGCA-co-HDCA) copolymer. The PEG
amino group was then selectively deprotected by
TFA, which does not cleave ester bonds;21 in fact,
Figure 1. 1H NMR spectrum of the poly(H2NPEGCA-
co-HDCA) copolymer.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

1458
STELLA ET AL.
Table 1. Composition and Molecular Weight of the
Poly(H2HPEGCA-co-HDCA)
Initial ratio
PEGCA/
HDCA
Weight-average MW
compositiona Theoreticalb Observedc
Polymer
1:5
1:4.9
5089
4922 400
a Calculated from 1H NMR spectra. bCalculated by consid-
ering the constitutional repeating units according to the ini-
c
tial ratio. Calculated from GPC chromatograms.
attributed to the quaternary carbon atoms of the
poly(cyanoacrylate), the hexadecyl methylene
groups appear between 32 and 22 ppm, and the
methyl group shows a signal at 13.7 ppm.
GPC analysis was performed to determine the
mean MW and polydispersity of the poly-
(H2NPEGCA-co-HDCA) copolymer (Table 1). The
average MW of the poly(H2NPEGCA-co-HDCA)
was 4922 400 Da, with a low index of polydis-
persity (0.14). The MW calculated according to
the initial PEGCA/HDCA ratio was 5089 Da.
Preparation and Characterization of
the Nanoparticles
Figure 2. 13C NMR spectrum of the poly(H2NPEGCA-co-
HDCA) copolymer.
The nanoparticles were prepared in a single step
by nanoprecipitation of PHDCA, poly(MePEGCA-
co-HDCA), or poly(H2NPEGCA-co-HDCA). The
organic solvent chosen for their preparation was
acetone; nevertheless, it was necessary to heat
this organic solvent because the polymers are not
soluble in acetone at room temperature.
ester group of the hexadecyl chain. The singlet at
1.44 ppm was attributed to the remaining t-Boc
groups. Finally, the peak at 1.25 ppm and the
triplet at 0.87 ppm were assigned to the methy-
lene and methyl protons of the hexadecyl chain,
respectively. Moreover, the integration of the
peaks corresponding to the PEG backbone and to
the methylene groups of the hexadecyl chain en-
abled the PEG/hexadecyl chain ratio to be esti-
mated.
These nanoparticles showed a unimodal size
distribution with a mean diameter of 96 9 nm
for poly(H2NPEGCA-co-HDCA), 78.9
12.2 nm
for the poly(MePEGCA-co-HDCA), and 89.2
26.2 nm for PHDCA nanoparticles. The size of the
three types of nanoparticles was unchanged after
4 weeks at 4 °C.
As shown in Table 1, starting from a PEGCA/
HDCA ratio of 1:5, the final polymer composition
was 1:4.9, showing a good correlation between
these two values.
The zeta potential value of poly(H2NPEGCA-
co-HDCA) nanoparticles was measured both in
water and in 0.01 M phosphate buffer to evaluate
the nanoparticles charge at different pHs and to
evidence the presence of amino groups at the sur-
face of the carrier (Table 2 and Figure 3). PHDCA
and poly(MePEGCA-co-HDCA) nanoparticles
were also tested as controls. It was observed that
in water, the zeta potential value of noncoated
PHDCA nanoparticles was lower than that of the
PEG-covered ones (MePEGCA-co-HDCA and
H2NPEGCA-co-HDCA; Table 2). Moreover, the
Figure 2 shows the 13C NMR spectrum of the
poly(H2NPEGCA-co-HDCA) copolymer. The sig-
nals at 164.2 and 116.2 ppm were attributed to
the ester carbonyl and to the cyano groups, re-
spectively. The sharp peak at 70.2 ppm was as-
signed to the PEG backbone, whereas the meth-
ylene in the ␣-position to the ester groups showed
a signal at 67.8 ppm. The peak at 43.0 ppm was
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

FOLATE-CONJUGATED POLY(CYANOACRYLATE) NANOPARTICLES
1459
Table 2. Zeta Potential of Nanoparticles Measured
in Water
Nanoparticle
Zeta potential (mV)
PHDCA
MePEG2000CA-co-HDCA
H2NPEG3400CA-co-HDCA
−48.06 0.20
−41.13 0.50
−32.40 2.10
longer PEG chains and the presence of terminal
amino groups on the poly(H2NPEGCA-co-HDCA)
particles confer a less negative zeta potential
value for these nanoparticles compared with the
ones bearing MePEG.
In phosphate buffer with poly(H2NPEGCA-co-
HDCA) nanoparticles, a dependence of the zeta
potential on the pH was observed, as shown in
Figure 3. At acidic pH, there was a significant
increase of the zeta potential value that corre-
sponded to the protonation of the amino groups,
whereas the zeta potential of PHDCA and
poly(MePEGCA-co-HDCA) nanoparticles did not
significantly change in the same range of pH. The
presence of amino groups on the nanoparticles
was also confirmed by the ninhydrin assay, which
was positive.
HIC chromatography data (Figure 4) showed
that PHDCA nanoparticles were strongly hydro-
phobic: they interacted with the gel phase and
were removed only by washing with Triton௡
X-100. On the contrary, poly(H2NPEGCA-co-
HDCA) nanoparticles exhibited a more hydro-
philic surface and were eluted easily.
Figure 4. HIC chromatograms of PHDCA and
poly(H2NPEGCA-co-HDCA) nanoparticles on a propyl
agarose column eluted with PBS buffer.
Preparation and Purification of
Folate–Poly(H2NPEGCA-co-HDCA)
Nanoparticles Conjugate
After poly(H2NPEGCA-co-HDCA) nanoparticles
preparation, they were covalently linked to folic
acid. To activate the folate carboxylic groups for
coupling with PEG terminal amino groups of the
poly(H2NPEGCA-co-HDCA) copolymer,25 the
NHS-ester of folic acid has been prepared, as
shown in Scheme 2. Then, the NHS-folate was
added to the poly(H2NPEGCA-co-HDCA) nano-
particles suspension. The reaction of NHS-folate
with PEG amino groups led to the formation of a
covalent amide bond. The folic acid–nanoparticles
conjugates were then purified from unreacted fo-
lic acid by Sepharose CL-4B gel chromatography,
using bicarbonate buffer as eluent (Figure 5). Two
distinct peaks were observed, the first one corre-
sponding to folate-nanoparticles, which were
eluted in the excluded volume, and the second one
corresponding to unreacted folate. As shown in
Figure 5, the separation was observed to be com-
plete. Then, conjugated nanoparticles were dia-
lyzed and their size was determined to verify the
absence of aggregates, which could result from
the coupling procedure. It was observed that the
size distribution was still unimodal and the diam-
eter near 100 nm, even if an increase in diameter
was found (136
47 nm instead of 96
9 nm
before conjugation). No change in nanoparticles
size was observed after 1 month of storage (4 °C).
To investigate the surface hydrophobicity of
the nanoparticles after folate conjugation, HIC
chromatography analysis was performed. The fo-
late-conjugated nanoparticles still showed a hy-
Figure 3. Zeta potential–pH profiles of PHDCA,
poly(MePEGCA-co-HDCA), and poly(H2NPEGCA-co-
HDCA) nanoparticles measured in 0.01 M phosphate
buffer.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

1460
STELLA ET AL.
FBP in the presence of free folic acid was just
slightly lower than the one bound in the absence
of free folate. Kinetic parameters for folate–nano-
particles and free folic acid are as follows: folate-
conjugated nanoparticles have an apparent disso-
ciation constant of 800 170 nM (calculated from
Figure 6C on the basis of 79 ␮M folate/2.5 mg
folate-nanoparticles) and free folic acid of 11
1
␮M (Figure 6D).
DISCUSSION
Figure 5. Gel filtration chromatogram of folate-
poly(H2NPEGCA-co-HDCA) nanoparticles on a Sepha-
rose CL-4B column eluted with 5 mM carbon-
ate/bicarbonate buffer pH 9.0. UV analysis of the frac-
tions indicated that the first peak corresponded to
purified conjugated nanoparticles and the second one to
the unbound folate.
Poly(H2NPEGCA-co-HDCA) copolymer was ana-
1
lyzed by H and 13C NMR, which confirmed the
expected structure. The GPC analysis showed a
good correlation between the initial proportion of
the monomers (PEGCA/HDCA, 1:5) and the mea-
sured average MW of the obtained copolymer
(4922 Da measured versus 5089 Da as calculated
MW theoretical value).
Nanoparticles with a size of 100 nm display-
ing a unimodal distribution were obtained after
nanoprecipitation of the poly(H2NPEGCA-co-
HDCA) copolymer in water. It is noteworthy that
using this methodology no surfactant was needed
for obtaining particles in a colloidal range. This is
an important point because most nanoparticle
preparation methods need one or more surface-
active agents to make the dispersion stable.
Clearly, it is a great advantage to avoid surfac-
tants because most of them are not allowed for
intravenous administration. Now the question
arises if sufficient PEG amino moieties were
available at the surface of the nanoparticles for
starting the folic acid coupling procedure. To an-
swer the question, zeta potential measurements
were performed in water and in phosphate buffer.
The more interesting observation was that in the
case of poly(H2NPEGCA-co-HDCA) nanopar-
ticles, the decrease of the pH led to an increase in
zeta potential value that corresponded to the pro-
tonation of the amino groups. This phenomenon
was not observed with poly(MePEGCA-co-HDCA)
or with non-PEGylated PHDCA nanoparticles.
Such a behavior suggests that amino groups of
poly(H2NPEGCA-co-HDCA) are well situated on
the surface of the nanoparticles, letting them
available for folic acid conjugation.
drophilic surface: they did not interact with the
gel phase and were eluted without the Triton௡
X-100; moreover, these particles were found to be
in the same collected fractions of poly(H2NPEGCA-
co-HDCA) nanoparticles before folic acid conjuga-
tion (data not shown).
Determination of Folate Content
To evaluate the extent of folate conjugation in fo-
late-conjugated poly(H2NPEGCA-co-HDCA)
nanoparticles, the quantitative UV analysis was
performed after lyophilization of the conjugate. It
was found that 14–16% of the total of PEG chains
were linked to folic acid molecules.
Surface Plasmon Resonance Measurements
To evaluate the ability of folate–poly(H2NPEGCA-
co-HDCA) nanoparticles to recognize the FBP, sur-
face plasmon resonance analysis was performed. As
shown in Figure 6A, folate-conjugated nanopar-
ticles were able to recognize the sensorchip-
immobilized FBP. On the contrary, the poly-
(H2NPEGCA-co-HDCA) nanoparticles, which were
not coupled to folic acid, failed to bind onto the sur-
face of the sensorchip in the same proportion.
To study the specificity of the interaction, bind-
ing experiments were performed in presence of 50
␮M of free folic acid. As shown in Figure 6B, the
amount of conjugated nanoparticles bound to the
The presence of PEG moieties at the surface of
the particles has also been evidenced by HIC
chromatography, showing that poly(H2NPEGCA-
co-HDCA) nanoparticles exhibited a more hydro-
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

FOLATE-CONJUGATED POLY(CYANOACRYLATE) NANOPARTICLES
1461
Figure 6. Surface plasmon resonance analysis of adsorption of folate-
poly(H2NPEGCA-co-HDCA) nanoparticles on immobilized FBP. Nonconjugated
poly(H2NPEGCA-co-HDCA) nanoparticles were used as control. These experiments
were conducted according to “Experimental Section” (flow rate: 5 ␮L/min). Every nano-
particle suspension was tested on a freshly prepared protein-covered sensor chip chan-
nel (RU ס resonance units). (A) Comparison of binding of folate-nanoparticles and
control nanoparticles (5 mg/mL) to FBP. (B) The same comparison in the presence of 50
␮M of free folic acid. (C) Kinetic analysis of folate-nanoparticles (2.5 mg/mL correspond-
ing to 79 ␮M folate) interaction with FBP. The heavy line shows the corrected sensor-
gram after subtraction of the sensorgram obtained with control nanoparticles. The
normal line shows the theoretical curve fitting according to a two-conformation state
model as calculated by BIA evaluation 3.0. The overall dissociation constant corre-
sponds to 800 170 nM. (D) Kinetic analysis of folate (1 mM) interaction with FBP. As
in C, the heavy line shows the sensorgram obtained for the binding of folate to FBP and
the normal line shows the theoretical curve. The apparent dissociation constant was
calculated as 11 1 ␮M.
philic surface than the nonPEGylated PHDCA
nanoparticles.
ticles preparation. First of all, the nanoparticles
can be prepared in optimal conditions before con-
jugation, considering for example copolymer solu-
bility in the organic solvents. Second, when the
conjugation is performed after nanoparticles
preparation, folic acid molecules may only be
linked to the surface functional groups of the
nanoparticles, thus being more available for fo-
late receptor recognition. In fact, if the coupling
procedure is made before nanoparticles prepara-
tion, some folate–PEG moieties can be included
into the particles core and so they become unuse-
The reaction of activated NHS-folate with PEG
terminal amino groups at the surface of
poly(H2NPEGCA-co-HDCA) nanoparticles
(Scheme 2) has then been performed to obtain a
stable covalent linkage between the folic acid and
the nanoparticle solid support.
This approach, which consists in the conjuga-
tion of folic acid on preformed nanoparticles,
presents some advantages in respect to the cou-
pling of folate to the copolymer before nanopar-
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

1462
STELLA ET AL.
ful for folate receptor targeting. Moreover, folate-
conjugated nanoparticles obtained by our proce-
dure were very stable after 1 month of storage
(4 °C).
Another step was the investigation of the sur-
face hydrophilicity of the nanoparticles after fo-
late conjugation. Results obtained after HIC chro-
matography analysis demonstrate that the conju-
gation of a part of the PEG moieties with folic acid
does not change the surface hydrophilicity of the
nanoparticles. This point is important to check
because surface hydrophilicity is known to reduce
the opsonization of the nanoparticles after intra-
venous administration, thus increasing their
blood circulation time.2,4
The number of PEG chains that have been con-
jugated to folic acid molecules has been calculated
to be 15% of total chains. To assess whether this
amount of folate-PEG derivatization is sufficient
to allow efficient recognition of the folate-
nanoparticles by the FBP, surface plasmon reso-
nance analysis was performed. As shown in Fig-
ure 6A, the folate-nanoparticles were able to ef-
fectively recognize the sensorchip-immobilized
FBP. More remarkable was the observation that
free folic acid (50 ␮M) was unable to displace fo-
late-conjugated nanoparticles from their interac-
tion with the folate receptor (Figure 6B). Analysis
of the kinetic parameters for folate-nanoparticles
and free folic acid (Figure 6C and 6D) explains the
low inhibitory capacity of folate as follows: folate-
conjugated nanoparticles have an apparent disso-
ciation constant of 800 170 nM, calculated on
the basis of 79 ␮M folate/2.5 mg folate-
nanoparticles. This affinity is due to two binding
steps: a fast one comparable to that of free folate
and a slow isomerization step (Figure 6C). In con-
trast, folic acid has an apparent dissociation con-
stant of only 11 1 ␮M (Figure 6D), necessitating
thus at least a 108-fold higher molar concentra-
tion over 79 ␮M to completely inhibit the interac-
tion of folate-nanoparticles with FBP. This con-
centration is over the solubility limit of folic acid
under the experimental conditions. The greater
binding avidity of folate-conjugated nanoparticles
towards FBP may be explained as follows: these
particles represent a multivalent form of the li-
gand folic acid, and folate receptors are often dis-
posed in clusters; thus, conjugated nanoparticles
could display a multivalent stronger interaction
with FBP. For this reason, the slow isomerization
step is supposed to correspond to the clustering of
FBP on the chip (Figure 7).
Figure 7. Schematic representation of the interac-
tion of free folic acid and folate-conjugated nanopar-
ticles with the immobilized FBP. (A) Free folic acid has
a single interaction with the receptor. (B) Folate-
conjugated nanoparticles display a multivalent inter-
action with the clusters of the FBP.
The relative low affinities measured in the BI-
Acore system compared with the known affinities
for folate of the FBP could be explained in two
ways. Immobilization of the binding protein could
alter its properties by linking the free lysine resi-
dues in an amide bond to the carboxylated dex-
tran. Alternatively, immobilization could inhibit
the conformational changes, which are leading to
high affinity binding. Whatever the mechanism,
the fact remains that folate bound to nanopar-
ticles has a 10-fold higher apparent affinity for
the FBP than free folate.
Thus, the prepared nanoparticles represent a
new potential delivery system for compounds
with anticancer activity because folate receptor is
frequently overexpressed onto the surface of ma-
lignant cells. Moreover, the encapsulation of an-
ticancer drugs into folate-conjugated nanopar-
ticles could not only increase the specific drug de-
livery towards target tumoral cells, but it may
also present the advantages to protect the drug
from in vivo degradation (case of antisense oligo-
nucleotides, for example) and to reduce drug side
effects and toxicity.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

FOLATE-CONJUGATED POLY(CYANOACRYLATE) NANOPARTICLES
1463
binding with long circulation times. Biochim Bio-
phys Acta 1149:180–184.
CONCLUSIONS
6. Kaplan MR, Calef E, Bercovici T, Gitler C. 1983.
The selective detection of cell surface determinants
by means of antibodies and acetylated avidin at-
tached to highly fluorescent polymer microspheres.
Biochim Biophys Acta 728:112–120.
7. Kumakura M, Suzuki M, Adachi S, Kaetsu I. 1983.
Polyacrolein microspheres as immunoreagents.
J Immunol Methods 63:115–122.
8. Bourel D, Rolland A, Le Verge R, Genetet B. 1988.
CD3 monoclonal antibody covalently coupled to
fluorescent polymethacrylic nanoparticles. J Im-
munol Methods 106:161–167.
9. Franklin WA, Waintrub M, Edwards D, Chris-
tensen K, Prendegrast P, Woods J, Bunn PA, Kol-
house JF. 1994. New anti-lung cancer antibody
cluster 12 reacts with human folate receptors pres-
ent on adenocarcinoma. Int J Cancer Suppl 8:
89–95.
In this study we have synthesized and character-
ized a novel amphiphilic PEG cyanoacrylate co-
polymer with terminal amino groups. This copoly-
mer allowed the preparation of stable and small-
sized nanoparticles, which were found to be an
excellent solid support for coupling activated folic
acid. The analysis by surface plasmon resonance
showed that the folate-nanoparticles conjugates
were able to bind specifically to the immobilized
FBP, whereas nanoparticles without folic acid
targeting moieties showed a much lower interac-
tion.
In conclusion, these PEGylated and targeted
nanoparticles represent potential carriers for tu-
mor cell-selective targeting of antitumoral drugs.
Further studies are in progress to test this new
drug delivery system in vivo.
10. Weitman SD, Lark RH, Coney LR, Fort DW, Fra-
sca V, Zurawski VR, Kamen BA. 1992. Distribution
of the folate receptor GP38 in normal and malignat
cell lines and tissues. Cancer Res 52:3396–3401.
11. Coney LR, Tomassetti A, Carayannopoulos L, Fra-
sca V, Kamen BA, Colnaghi MI, Zurawski WR.
1991. Cloning of a tumor-associated antigen:
Mov18 and Mov19 antibodies recognize a folate-
binding protein. Cancer Res 51:6125–6132.
12. Garin-Chesa P, Campbell I, Saigo PE, Lewis JL,
Old LJ, Rettig WJ. 1993. Trophoblast and ovarian
cancer antigen LK26. Sensitivity and specificity in
immunopathology and molecular identification as
a folate-binding protein. Am J Pathol 142:557–567.
13. Campbell IG, Jones TA, Foulkes WD, Trowsdale J.
1991. Folate-binding protein is a marker for ovar-
ian cancer. Cancer Res 51:5329–5338.
ACKNOWLEDGMENTS
This work was supported by MURST 60–40%
(Progetto Nazionale “Tecnologie Farmaceutiche”)
and by CNRS (programme “Physique et Chimie
du Vivant”) grants.
REFERENCES
1. Stolnik S, Illum L, Davis SS. 1995. Long circulat-
ing microparticulate drug carriers. Adv Drug Del
Rev 16:195–214.
2. Gref R, Minamitake Y, Peracchia MT, Langer R.
1996. PEG-coated biodegradable nanospheres for
intravenous drug administration. In Cohen S,
Bernstein H, editors. Microparticulate systems for
the delivery of proteins and vaccines, New York:
Marcel Dekker, pp 279–306.
3. Peracchia MT, Desmae¨le D, Couvreur P, d’Angelo
J. 1997. Synthesis of a novel poly(MePEG cyano-
acrylate-co-alkyl cyanoacrylate) amphiphilic co-
polymer for nanoparticle technology. Macromol-
ecules 30:846–851.
4. Peracchia MT, Fattal E, Desmae¨le D, Besnard M,
Noe¨l JP, Gomis JM, Appel M, d’Angelo J, Couvreur
P. 1999. Stealth௡ PEGylated polycyanoacrylate
nanoparticles for intravenous administration and
splenic targeting. J Controlled Release 60:121–
128.
14. Antony AC. 1992. The biological chemistry of folate
receptors. Blood 79:2807–2820.
15. Rothberg KG, Ying YS, Kolhouse JF, Kamen BA,
Anderson RG. 1990. The glycophospholipid-linked
folate receptor internalizes folate without entering
the clathrin-coated pit endocytic pathway. J Cell
Biol 110:637–649.
16. Wang S, Lee RJ, Mathias CJ, Green MA, Low PS.
1996. Synthesis, purification, and tumor cell up-
take of Gallium-67-deferoxamine-folate, a poten-
tial radiopharmaceutical for tumor imaging. Bio-
conjugate Chem 7:56–62.
17. Lee RJ, Low PS. 1994. Delivery of liposomes into
cultured KB cells via folate receptor-mediated en-
docytosis. J Biol Chem 269:3198–3204.
18. Wang S, Low PS. 1998. Folate-mediated targeting
of antineoplastic drugs, imaging agents, and
nucleic acids to cancer cells. J Controlled Release
53:39–48.
5. Blume G, Cevc G, Crommelin MDJA, Bakker-
Woundenberg IAJM, Kluft C, Storm G. 1993. Spe-
cific targeting with poly(ethylene glycol)-modified
liposomes: Coupling of homing devices to the ends
of the polymeric chains combines effective target
19. Gabizon A, Horowitz AT, Goren D, Tzemach D,
Mandelbaum-Shavit F, Qazen MM, Zalipsky S.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000

1464
STELLA ET AL.
1999. Targeting folate receptor with folate linked
to extremities of poly(ethylene glycol)-grafted lipo-
somes: in vitro studies. Bioconjugate Chem 10:
289–298.
terfacial polymer deposition following solvent dis-
placement. Int J Pharm 55:R1–R4.
23. Lo¨fås S, Johnsson B. 1990. A novel hydrogel matrix
on gold surfaces plasmon resonance sensors for fast
and efficient covalent immobilisation of ligands.
J Chem Soc Chem Commun 21:1526–1528.
20. Malmqvist M. 1993. Biospecific interaction analy-
sis using biosensor technology. Nature 361:186–
187.
21. Tarbell DS, Yamamoto Y, Pope BM. 1972. New
method to prepare N-t-butoxycarbonyl derivatives
and the corresponding sulfur analogs from di-t-
butyl dicarbonate or di-t-butyl dithiol dicarbonates
and amino acids. Proc Nat Acad Sci, USA 69:730–
732.
24. Johnsson B, Lo¨fås S, Lindquist G. 1991. Immobili-
zation of proteins to a carboxymethyldextran-
modified gold surface for biospecific interaction
analysis in surface plasmon resonance sensors.
Anal Biochem 198:268–277.
25. Mattson G, Conklin E, Desai S, Nielander G, Sav-
age MD, Morgensen S. 1993. A practical approach
to crosslinking. Mol Biol Rep 17:167–183.
22. Fessi H, Puisieux F, Devissaguet JPh, Ammoury
N, Benita S. 1989. Nanocapsule formation by in-
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 89, NO. 11, NOVEMBER 2000