Arabinogalactan-folic acid-drug conjugate for targeted delivery and target-activated release of anticancer drugs to folate receptor-overexpressing cells

Article abstract of DOI:10.1021/bm900853z

Folic acid (FA) is a high affinity ligand (K_d = 0.1-1 nM) of folate receptors (FRs) responsible for cellular uptake of folates via receptor-mediated endocytosis. FRs are frequently overexpressed in malignant epithelial cells including ovary, brain, kidney, breast, colon, and lung. FR has emerged as a target for the differential-delivery of anticancer chemotherapeutics with several FA-linked therapeutic agents currently undergoing clinical trials. Here we show that by tethering both FA and the anticancer drug methotrexate (MTX) to arabinogalactan (AG), a highly branched natural polysaccharide with unusual water solubility, a targeted biomacromolecular nanovehicle is formed, which can differentially deliver a cytotoxic cargo into FR-overexpressing cells. Moreover, by linking MTX via an endosomally cleavable peptide (GFLG), we demonstrate a target-activated release mechanism. This FA-AG-GFLG-MTX drug conjugate displayed 6.3-fold increased cytotoxic activity to FR-overexpressing cells compared to their FR-lacking counterparts. These findings establish a novel FA-tethered polymeric nanoconjugate for the targeted delivery of antitumor agents into cancer cells overexpressing FR.

Full text of DOI:10.1021/bm900853z

294
Biomacromolecules 2010, 11, 294–303
Arabinogalactan-Folic Acid-Drug Conjugate for Targeted
Delivery and Target-Activated Release of Anticancer Drugs to
Folate Receptor-Overexpressing Cells
Roy I. Pinhassi,#,‡ Yehuda G. Assaraf,§ Shimon Farber,| Michal Stark,§ Diana Ickowicz,|
Stavit Drori,§ Abraham J. Domb,| and Yoav D. Livney*,#,‡
Laboratory of Biopolymers and Food Nanotechnology, Department of Biotechnology and Food
Engineering, The Russell Berrie Nanotechnology Institute, and The Fred Wyszkowski Cancer Research
Laboratory, Department of Biology, The Technion, Israel Institute of Technology, Israel, and Institute of
Drug Research, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Israel
Received October 19, 2009; Revised Manuscript Received November 19, 2009
Folic acid (FA) is a high affinity ligand (Kd ) 0.1-1 nM) of folate receptors (FRs) responsible for cellular uptake
of folates via receptor-mediated endocytosis. FRs are frequently overexpressed in malignant epithelial cells including
ovary, brain, kidney, breast, colon, and lung. FR has emerged as a target for the differential-delivery of anticancer
chemotherapeutics with several FA-linked therapeutic agents currently undergoing clinical trials. Here we show
that by tethering both FA and the anticancer drug methotrexate (MTX) to arabinogalactan (AG), a highly branched
natural polysaccharide with unusual water solubility, a targeted biomacromolecular nanovehicle is formed, which
can differentially deliver a cytotoxic cargo into FR-overexpressing cells. Moreover, by linking MTX via an
endosomally cleavable peptide (GFLG), we demonstrate a target-activated release mechanism. This FA-AG-
GFLG-MTX drug conjugate displayed 6.3-fold increased cytotoxic activity to FR-overexpressing cells compared
to their FR-lacking counterparts. These findings establish a novel FA-tethered polymeric nanoconjugate for the
targeted delivery of antitumor agents into cancer cells overexpressing FR.
elimination from the body, hence, avoiding accumulation in
vivo, low polydispersity to ensure an acceptable homogeneity
of the final conjugates, and longer body residence time either
to prolong the conjugate action or to allow distribution and
accumulation in the target tumors. In this respect, arabinoga-
lactan (AG), a highly branched natural polysaccharide with
unusual water solubility (70% w/w in water) satisfies these strict
requirements. AG is extracted from the Larix tree and is
available at 99.9% purity with reproducible molecular weight
(MW) and physicochemical properties.16 The high water
solubility, biocompatibility, biodegradability, and ease of drug
conjugation in an aqueous solution render AG attractive as a
potential biomacromolecular nanovehicle for folic acid-based
targeted anticancer drug delivery, which has not yet been
proposed for this task. While most nanovehicles are passive,
active targeting modalities based on molecular recognition are
rapidly advancing, using antibodies or ligands, for example, folic
acid (FA), for which certain malignant tumors overexpress
specific receptors, known as folate receptors (FRs).17-19 An
example is FA-drug conjugates, which have been shown to
undergo receptor-mediated endocytosis into acidic endosomes
in cancer cells, thereby offering a highly effective route for the
delivery of anticancer drugs.20,21 The FR constitutes a useful
target for tumor-specific drug delivery primarily because (i) it
is up-regulated in many human carcinomas, including malignan-
cies of the ovary, brain, kidney, breast, colon, myeloid cells,
and lung, and (ii) FR density appears to increase as the stage/
grade of the cancer is more advanced, including in metastatic
disease.22 Thus, solid tumors that are currently most difficult
to treat by classical therapeutic modalities may be readily
targeted with FA-linked therapeutics.23 To exploit these unique
characteristics of differential FR overexpression in various
1. Introduction
Chemotherapy of neoplastic diseases is often restricted by
adverse systemic toxicity, which consequently limits the dose
of antitumor agents that can be administered, or by the frequent
emergence of anticancer drug resistance. The insufficient
selectivity of antitumor agents is only one obstacle currently
hindering drug efficacy. Other impediments include inacces-
sibility of the target, premature drug metabolism and multi-
drug resistance phenomena.1 Hence, there is a growing demand
for innovative targeted drug delivery systems that can selectively
deliver anticancer drugs and overcome drug resistance.2-8 The
most promising approaches in the delivery of anticancer
chemotherapeutics include the use of long-circulating nanodrug
carriers, including polymeric drug carriers,9-11 nanoparticles,8,12
nanogels,13 or liposomes coated with hydrophilic polymers that
can evade elimination by the immune system.14,15 Polymer-
conjugated drugs11 generally exhibit prolonged half-life, higher
stability, water solubility, lower immunogenicity, and antige-
nicity and possibly a specific targeting to certain tissues or cells.
This approach, explored for the first time in the 1950s, is recently
gaining growing attention due to the introduction of new
polymersandtoadvancedchemicalstrategiesofdrugcoupling.14,15
An ideal polymer for drug delivery should be characterized by
biodegradability or adequate molecular weight that allows
favorable pharmacokinetics and pharmacodynamics facilitating
* To whom correspondence should be addressed. Phone: 972-4-8294225.
Fax: 972-4-8293399. E-mail: livney@technion.ac.il.
# Department of Biotechnology and Food Engineering, Israel Institute
of Technology.
‡ The Russell Berrie Nanotechnology Institute, Israel Institute of
Technology.
§ Department of Biology, Israel Institute of Technology.
| School of Pharmacy, The Hebrew University of Jerusalem.
10.1021/bm900853z
 2010 American Chemical Society
Published on Web 12/16/2009

Target-Activated Release of Anticancer Drugs
Biomacromolecules, Vol. 11, No. 1, 2010 295
at an average yield of 75% ( 5% (w/w). FT-IR (KBr) )1724 cm-1
(CdO). The aldehyde content was determined by the hydroxylamine
hydrochloride method.31
carcinomas, FA has been tethered to both low molecular weight
drugs as well as macromolecular complexes as a means of
targeting the attached chemotherapeutic cargo molecules to
malignant cells. Conjugation of FA to macromolecules has been
shown to enhance their delivery to FR-expressing cancer
cells.24-29 To further maximize the efficacy of antitumor agents
and minimize untoward side effects, nanoscopic drug delivery
systems have been developed by introducing target-activated
triggering mechanisms for drug release. These should increase
the feasibility of obtaining local high-dose therapy in malignant
tissues and at intracellular compartments, while minimizing
premature drug release in the blood. Such an approach requires
sophisticated strategies and systems that perform the desired
chemical/physical functions depending on the triggering signals.
For example, polymer-drug conjugates linked by a peptide bond
such as the peptide linker Gly-Phe-Leu-Gly (GFLG), which is
known to be stable in serum yet cleavable by endolysosomal
peptidases,10,30 bears promise for an efficient target-activated
drug release mechanism inside tumor cells.
Synthesis of NHS Esters of FA or of MTX (General Method).
Lyophilized FA (1.0 g, 2.26 mmol) dissolved in 20 mL of distilled
dimethyl sulfoxide (DMSO) was activated with 3 equiv of 1,3-
dicyclohexylcarbodiimide (DCC) coupling agent (1.4 g, 6.78 mmol)
for 2 h at room temperature. Then 0.78 g of N-hydroxysuccinimide
(NHS; 3 equiv, 6.78 mmol) and 0.14 g of 4-dimethylaminopyridine
(DMAP; 0.5 equiv, 1.14 mmol) were added to the flask. The mixture
was stirred in a light-protected container overnight at 40 °C under a
nitrogen atmosphere. The resulting precipitate (dicyclohexylurea (DCU))
was discarded by filtration and the filtrate was concentrated by partial
removal of the solvent under reduced pressure at 40 °C. Precipitation
of the product was achieved by dropwise addition of the concentrated
solution into a cold solution of acetone/diethyl ether (30:70 v/v ratio)
followed by filtration. The precipitate was washed with 100 mL of
acetone/diethyl ether (30:70 v/v ratio) solution and diethyl ether (3 ×
100 mL) and vacuum-dried over P2O5 to yield 60% (mol/mol) of the
product.
1H NMR (FA-NHS, DMSO, ppm): 2.069 (m, 2H, -CH2-, FA
hydrogens), 2.391 (m, 2H, -CH2-, FA hydrogens), 2.791 (m, 4H,
-CH2-, NHS hydrogens), 4.467 (d, 1H, CH, FA hydrogen), 4.605 (s,
1H, CH, FA hydrogen), 4.938 (m, 2H, NH2, FA hydrogens), 5.012 (m,
2H, NH2, FA hydrogens), 6.643 (t, 1H, CH, FA benzyl hydrogen), 6.962
(t, 1H, NH, FA hydrogen), 7.655 (t, 1H, CH, FA benzyl hydrogen),
8.149 (d, 1H, NH, FA hydrogen), 8.657 (d, 1H, CH, FA benzyl
hydrogen), 11.511 (broad, 1H, COOH, FA carboxylic hydrogen).
1H NMR (MTX-NHS, DMSO, ppm): 1.244 (m, 2H, -CH2-, MTX
hydrogens), 1.683 (m, 2H, -CH2-, MTX hydrogens), 2.790 (t, 4H,
-CH2-, NHS hydrogens), 3.291 (s, 3H, CH3, MTX hydrogens), 4.467
(d, 1H, CH, MTX hydrogen), 4.777 (s, 1H, CH, MTX hydrogen), 4.846
(m, 2H, NH2, MTX hydrogens), 6.589 (s, 1H, CH, MTX benzyl
hydrogen), 6.821 (m, 1H, NH, MTX hydrogen), 7.518 (m, 1H, CH,
MTX benzyl hydrogen), 7.736 (d, 1H, NH, MTX hydrogen), 8.603 (d,
1H, CH, MTX benzyl hydrogen), 12.324 (broad, 1H, COOH, MTX
carboxylic hydrogen).
Synthesis of Folic Acid-Propaneamine DeriVatiVe. Volumes of 0.62
mL (7.4 mmol) of propane-1,3-diamine and 0.2 mL of DIEA were
dissolved in 1 mL of dry DMSO. Then 400 mg FA-NHS (0.74 mmol)
dissolved in 6 mL of dry DMSO was added dropwise to the clear
solution. The mixture was stirred in a light-protected container overnight
at room temperature under a nitrogen atmosphere. The clear solution
was then concentrated by partial removal of the solvent under reduced
pressure at 40 °C. The solution was poured into a cold solution of
acetone/diethyl ether (30:70% v/v ratio) resulting in precipitation of
the product. The obtained precipitate was washed with 50 mL of
acetone/diethyl ether solution (30:70% v/v ratio) and diethyl ether (3
× 50 mL) to remove traces of unreacted reagents and diisopropylethy-
lammonium-NHS salt. The product was dried in vacuum over P2O5
overnight and the yield was 60% (mol/mol).
Herein we devised a new targeted delivery system for
anticancer drugs based upon the naturally occurring polymer
AG and characterized its ability to differentially recognize,
penetrate, and eliminate tumor cells overexpressing FRR. The
mode of synthesis employed here may be easily adapted to
prepare therapeutic conjugates with several different cytotoxic
drugs which may act synergistically to exploit the complete
potential of the polymeric nanocarrier. Using an in vitro model
cell line overexpressing FRR, we show here for the first time
that this novel antitumor drug conjugate can differentially target
and eliminate cells overexpressing FRR.
2. Experimental Section
Materials. All reactants employed in this study were of analytical
grade: ethylenediamine (EDA), propane-1,3-diamine (PDA; Biolab,
Jerusalem, Israel), methotrexate (MTX; Iffect Chemphar Co., Ltd.,
Shenzhen, China), dimethyl sulfoxide (DMSO), and 4-(dimethylamino)
pyridine (DMAP; Merck and Co., Inc.). 1,3-Dicyclohexylcarbodiimide
(DCC), N-hydroxysuccinimide (NHS), diisopropylethylamine (DIEA),
triethylamine, folic acid (FA), and fluorescein isothiocyanate (FITC)
were purchased from Sigma-Aldrich Inc., Rehovot, Israel. The peptide
Gly-Phe-Leu-Gly (GFLG) was tailor-synthesized by BioSight Ltd.,
Israel. Arabinogalactan (AG) was purchased from Lonza Inc. (Allendale,
NJ). Growth media for cell culture were obtained from Gibco, U.S.A.,
and Beth-Haemek, Israel.
Methods. Determination of Molecular Weight of Polymers and
Conjugates. Molecular weights of starting polymers and synthetic
conjugates were estimated using a size exclusion chromatography (SEC)
system (Spectra Physics instrument, Darmstadt, Germany) consisting
of an isocratic pump (HP-1050, Waldbronn, Germany), methyl-
methacrylate-gel size exclusion column (Shodex KB-803 HQ, Phe-
nomenex, Japan), refractive index (RI) detector, and Breeze 3.20
software (Waters Corp., Milford, U.S.A.). The eluent used was 0.05
M NaNO3. Average molecular weights were determined according to
pullulan standards (PSS, Mainz, Germany) with molecular weights
between 5800 and 212000. A sage-metering pump model 365 (Orion,
NJ) was used for slow and reproducible addition of reactants.
Oxidation of Arabinogalactan. AG (10 g, 62.5 mmol of monomer
saccharide units) was dissolved in 100 mL of doubly deionized water
(DDW). To this solution potassium periodate was separately added at
1:1 mol ratio (IO4-/AG units), and the mixture was stirred in the dark
at room temperature for 2 h. The resulting polyaldehyde was purified
from iodate (IO3-) and unreacted periodate (IO4-) by Dowex-1 (acetate
form) anion exchange chromatography, followed by extensive dialysis
against DDW (5 L × 4; 12000 Da cutoff cellulose tubing) at 4 °C for
3 days. Purified polyaldehyde was freeze-dried to obtain a white powder
Methotrexate-propaneamine derivative was synthesized by the same
method resulting in 58% (mol/mol) yield.
1H NMR (FA-propaneamine, DMSO, ppm): 1.902 (m, 2H, -CH2-,
PDA hydrogens), 2.087 (m, 2H, -CH2-, FA hydrogens), 2.405 (m,
2H, -CH2-, FA hydrogens), 2.734 (m, 2H, -CH2-NH2, PDA
hydrogens), 3.333 (m, 2H, -CH2-NH, PDA hydrogens), 4.475 (d, 1H,
CH, FA hydrogen), 4.622 (s, 1H, CH, FA hydrogen), 5.002 (m, 2H,
NH2, FA hydrogen), 6.611 (t, 1H, CH, FA benzyl hydrogen), 7.636 (t,
1H, CH, FA benzyl hydrogen), 8.188 (d, 1H, NH, FA hydrogen), 8.671
(d, 1H, CH, FA benzyl hydrogen).
Synthesis of MTXsGFLG Conjugate. MTX-NHS active ester (MTX-
NHS) (100 mg, 0.18 mmol) was dissolved in 3.5 mL of dry DMSO.
GFLG (trifluoroacetate form; 1.2 equiv, 110 mg, 0.216 mmol) dissolved
in 3.5 mL of dry DMSO and 0.15 mL of triethylamine (TEA) was
added to the flask. The mixture was stirred in a light-protected container
overnight at room temperature under a nitrogen atmosphere. The
resulting solution was then concentrated by partial removal of the

296 Biomacromolecules, Vol. 11, No. 1, 2010
Pinhassi et al.
solvent under reduced pressure at 40 °C. Precipitation of the product
was achieved by dropwise addition of the concentrated solution into
cold solution of acetone/diethyl ether (30:70 v/v ratio) followed by
filtration. The precipitate was washed with 100 mL of acetone/diethyl
ether (30:70 v/v ratio) solution and diethyl ether (3 × 100 mL) and
vacuum-dried over P2O5 to yield 78% (mol/mol) of the product.
1H NMR (MTX-GFLG, DMSO, ppm): 0.821 (dd, 6H, 2 CH3, Leu
hydrogens), 1.452 (m, 2H, -CH2-, Leu hydrogens), 1.890 (s, 1H, CH,
Leu hydrogen), 2.246 (m, 2H, -CH2-, MTX hydrogens), 2.476 (m,
2H, -CH2-, MTX hydrogens), 2.714 (d, 1H, -CH2-, Phe hydrogen),
3.193 (d, 1H, -CH2-, Phe hydrogen), 3.402 (s, 3H, CH3, MTX
hydrogens), 4.471 (d, 1H, CH, MTX hydrogen), 4.228 (d, 2H, -CH2-,
Gly hydrogens), 4.301 (d, 2H, -CH2-, Gly hydrogens), 4.507 (m, 1H,
CH, Leu hydrogen), 4.773 (s, 1H, CH, MTX hydrogen), 4.862 (s, 1H,
CH, Phe hydrogen), 6.561 (m, 1H, CH, MTX benzyl hydrogen), 6.811
(m, 1H, CH, MTX benzyl hydrogen), 7.141 (m, 1H, CH, MTX benzyl
hydrogen), 7.180 (m, 5H, CH, Phe benzyl hydrogens), 7.486 (m, 1H,
CH, MTX benzyl hydrogen), 7.716 (m, 2H, CH, MTX benzyl
hydrogens), 8.013 (m, 1H, CH, MTX benzyl hydrogen), 8.215, 8.561
(m, 2H, NH, GFLG hydrogens).
Synthesis of MTX-GFLG-NHS Ester. Previously prepared MTX-
GFLG conjugate (100 mg, 0.12 mmol) dissolved in 5 mL of dry DMSO
was activated with 3 equiv of dicyclohexylcarbodiimide (DCC; 75 mg,
0.36 mmol) for 2 h at room temperature. Then 41.6 mg (NHS; 3 equiv,
0.36 mmol) and 7.3 mg of DMAP (0.5 equiv, 0.06 mmol) were added
to the flask. The mixture was stirred in a light-protected container
overnight at 40 °C under a nitrogen atmosphere. The resulting
precipitate dicyclohexylurea (DCU) was discarded by filtration and the
filtrate was concentrated by partial removal of the solvent under reduced
pressure at 40 °C. Precipitation of the product was achieved by dropwise
addition of the concentrated solution into a cold acetone/diethyl ether
solution (30:70 v/v ratio) followed by filtration. The precipitate was
washed with 100 mL of acetone/diethyl ether solution (30:70 v/v ratio)
followed by diethyl ether (3 × 100 mL) and vacuum-dried over P2O5
to yield 60% (mol/mol) product.
1H NMR (MTX-GFLG-NHS, DMSO, ppm): 0.811 (m, 6H, 2 CH3,
Leu hydrogens), 1.457 (m, 2H, -CH2-, Leu hydrogens), 1.891 (s, 1H,
CH, Leu hydrogen), 2.192 (m, 2H, -CH2-, MTX hydrogens), 2.481
(m, 2H, -CH2-, MTX hydrogens), 2.711 (d, 1H, -CH2-, Phe
hydrogen), 2.791 (m, 4H, -CH2-, NHS hydrogens), 3.193 (d, 1H,
-CH2-, Phe hydrogen), 3.337 (s, 3H, CH3, MTX hydrogens), 3.601
(d, 1H, CH, MTX hydrogen), 4.229 (d, 2H, -CH2-, Gly hydrogens),
4.295 (d, 2H, -CH2-, Gly hydrogens), 4.520 (m, 1H, CH, Leu
hydrogen), 4.771 (s, 1H, CH, MTX hydrogen), 4.842 (s, 1H, CH, Phe
hydrogen), 6.577 (m, 1H, CH, MTX benzyl hydrogen), 6.810 (m, 1H,
CH, MTX benzyl hydrogen), 7.158 (m, 1H, CH, MTX benzyl
hydrogen), 7.187 (m, 5H, CH, Phe benzyl hydrogens), 7.592 (m, 1H,
CH, MTX benzyl hydrogen), 7.720 (m, 2H, CH, MTX benzyl
hydrogens), 8.016 (m, 1H, CH, MTX benzyl hydrogen), 8.287, 8.440,
8.573 (m, 3H, NH, GFLG hydrogens).
Synthesis of MTX-GFLG-Ethyleneamine DeriVatiVe. Ethylenediamine
(41 µL, 10 equiv, 0.54 mmol) and 100 µL of triethylamine, added to
facilitate the reaction, were dissolved in 1 mL of dry DMSO. MTX-
GFLG-NHS (50 mg, 0.054 mmol) dissolved in 2 mL of dry DMSO
was added dropwise. The mixture was stirred in a light-protected
container at room temperature for 20 h under a nitrogen atmosphere.
The solution was concentrated by partial removal of the solvent under
reduced pressure at 40 °C. Precipitation of the product was achieved
by dropwise addition of the concentrated solution into a cold solution
of acetone/diethyl ether (30:70 v/v ratio) followed by filtration. The
precipitate was washed with 50 mL of acetone/diethyl ether (30:70
v/v ratio) solution and diethyl ether (3 × 50 mL) and vacuum-dried
over P2O5 to yield 93% (mol/mol) product.
hydrogens), 2.708 (d, 1H, -CH2-, Phe hydrogen), 3.071 (d, 1H,
-CH2-, Phe hydrogen), 3.182 (s, 3H, CH3, MTX hydrogens), 3.457
(m, 2H, -CH2-, EDA hydrogens), 3.629 (d, 1H, CH, MTX hydrogen),
4.068 (d, 2H, -CH2-, Gly hydrogens), 4.254 (d, 2H, -CH2-, Gly
hydrogens), 4.763 (s, 1H, CH, MTX hydrogen), 4.856 (s, 1H, CH, Phe
hydrogen), 6.556 (m, 1H, CH, MTX benzyl hydrogen), 6.788 (m, 1H,
CH, MTX benzyl hydrogen), 7.191 (m, 5H, CH, Phe benzyl hydrogens),
7.592 (m, 1H, CH, MTX benzyl hydrogen), 7.695 (m, 2H, CH, MTX
benzyl hydrogens), 8.016 (m, 1H, CH, MTX benzyl hydrogen), 8.287,
8.440, 8.541 (m, 3H, NH, GFLG hydrogens).
Synthesis of FITC-EDA. An amount of 3.5 mg (0.009 mmol)
fluorescein isothiocyanate (FITC) dissolved in 0.3 mL of dry DMSO
was added dropwise to the solution of EDA (0.5 µL 0.0075 mmol) in
0.3 mL DMSO under continuous stirring. The mixture was then stirred
in a light-protected container at room temperature for 20 h. The solution
was added to the desired products for labeling.
Synthesis of FA-PDA-AG-EDA-MTX. A total of 400 mg (2.5 mmol)
oxidized AG (7.75 mmol/g aldehyde groups) was dissolved in 40 mL
of borate buffer (0.1 N, pH ) 11). A total of 25 mg crude
FA-propaneamine and 75 mg crude methotrexate-ethyleneamine were
added to the reaction mixture, resulting in a turbid solution. The mixture
was gently stirred in a light-protected container at 37 °C for 72 h. The
solution was centrifuged to remove the insoluble reactants, receiving a
clear yellow solution. The amine-based conjugate (reduced) was
obtained after reducing 20 mL of the imine conjugate solution with
excess of 95 mg (2.5 mmol) of NaBH4 for 2 h at 4 °C to obtain a more
stable amine form. The resulting light-yellow solution was purified by
dialysis against DDW (5 L × 4), applying a 3500 Da MW cutoff
cellulose dialysis tubing at 4 °C for 48 h followed by lyophilization
resulting in 50% (mol/mol) overall yield.
A total of 50 mg of the conjugate (imine form) was dissolved in 5
mL of the borate buffer solution (0.1 N, pH ) 11). FITC-EDA (5 mg)
was added dropwise to the solution. The mixture was gently stirred in
a light-protected container at 37 °C for 18 h. The reduction reaction
was conducted by the addition of 23.6 mg (0.62 mmol) NaBH4 for 2 h
at room temperature. The labeled product was purified by dialysis
through 3500 MWCO dialysis tubing, against DDW (5 L × 4) for 48 h
at 4 °C, followed by size exclusion column purification with Sephadex
G-25, and lyophilization.
Synthesis of FA-AG-GFLG-MTX. Oxidized AG (30 mg, 0.1875
mmol; 7.75 mmol/g aldehyde groups) was dissolved in 3.5 mL of borate
buffer (0.1 N, pH 11). Folate-propaneamine (4.5 mg, 0.009 mmol) and
20.7 mg EDA-GFLG-MTX (0.023 mmol) dissolved in 0.2 mL of
DMSO were added to the polysaccharide solution. The mixture was
gently stirred in a light-protected container at 37 °C for 72 h.
Nonconjugated insoluble folate-propaneamine and EDA-GFLG-MTX
were discarded by centrifugation before the reduction step. The amine-
based conjugate (reduced) was obtained after reducing 1.8 mL of the
imines conjugate solution with an excess of NaBH4 (4.3 mg, 0.1125
mmol) at 4 °C and dark conditions for 2 h under continuous stirring.
The resulting light-yellow solution was purified by dialysis against
DDW (5 L × 4), applying a 3500 Da MW cutoff cellulose dialysis
tubing at 4 °C for 48 h, followed by lyophilization, achieving 65%
(mol/mol) overall yields.
Synthesis of AG-GFLG-MTX. The synthesis of AG-GFLG-MTX was
done by the same method as the FA containing FA-AG-GFLG-MTX
conjugate excluding FA addition. Yield 55% (mol/mol).
EValuation of the Degree of Substitution with FA and MTX. The
degree of substitution with FA and MTX was estimated by UV
spectroscopy at 350 and 400 nm wavelengths, using Kontron Instru-
ments Uvicon model 930 (Ms-scientific, Berlin, Germany). The samples
were prepared as follows: 1 mg of the appropriate conjugate was
dissolved in 100 µL of DDW and diluted with DMSO to a final
1H NMR (MTX-GFLG-ethyleneamine, DMSO, ppm): 0.786 (m, 6H,
2 CH3, Leu hydrogens), 1.486 (m, 2H, -CH2-, Leu hydrogens), 1.883
(s, 1H, CH, Leu hydrogen), 2.250 (m, 2H, -CH2-, MTX hydrogens),
2.477 (m, 2H, -CH2-, MTX hydrogens), 2.588 (m, 2H, -CH2-, EDA

Target-Activated Release of Anticancer Drugs
Biomacromolecules, Vol. 11, No. 1, 2010 297
Scheme 1. Synthesis of FA-AG-GFLG-MTX Conjugate
concentration of 0.1 mg/mL. Spectral data were analyzed according to
eq 1:
then calculated from the set of equations obtained for two wavelengths,
350 and 400 nm, in which the differences between the absorbance
spectra of the pure compounds are relatively large, enabling to
quantitatively distinguish the two compounds.
Tissue Culture and Cell Lines. Parental Chinese hamster ovary
(CHO) AA8 cells and their reduced folate carrier (RFC)-null CHO
subline (termed C5) were grown under monolayer culture conditions,
as previously described.32 C5 cells were stably transfected with a
pcDNA3.1 expression vector harboring human folate receptor R, as
detailed previously.33 The latter cells, which highly overexpress folate
AFA(λi)(pure)
A(λi) ) CFA(conjugate)
+
CFA(pure)
AMTX(λi)(pure)
CMTX(pure)
CMTX(conjugate)
(1)
where A(λi) is the absorbance at wavelength λi, and CFA and CMTX are
the concentrations of FA and MTX, respectively (in pure solutions or
in the conjugate solution). The concentrations of FA and MTX were

298 Biomacromolecules, Vol. 11, No. 1, 2010
Pinhassi et al.
Table 1. Chemical Characteristics of the Oxidized AG and AG-MTX Conjugates
FAb %w (441 Da)
(mol/mol AG)
MTXb %w (454 Da)
(mol/mol AG)
a
conjugate
oxidized AG
Mw (Da)
Da
15280
16070
1.62
1.58
FA-PDA-AG-EDA-MTX
4%
7.50%
(1.5)
(0.0)
(2.7)
11.80%
(4.0)
2.40%
(0.8)
AG-EDA-GFLG-MTX
15530
15750
1.49
1.66
FA-PDA- AG-EDA-GFLG-MTX
4.80%
(1.7)
a Average molecular weight (Mw) and polydispersity (D ) Mw/Mn) were determined by SEC. b FA and MTX content (% w/w) were determined by
spectrophotometric measurement of the conjugates (UV, λ1 ) 350, λ2 ) 400) using calibration curves of FA and MTX and eq 1 (determination accuracy
( 0.5%).
receptor R, were termed C5-FRR, and were routinely grown in a
medium containing 0.2 nM FA.
anticancer drug MTX, as well as FA, as the targeting molecule,
onto AG, a natural polysaccharide, and investigated its efficacy
as a potential antitumor agent against a model cell line
overexpressing FRR. MTX was conjugated to the polymer
backbone either only via a short linker ethylenediamine (EDA)
or also via a tetra-peptide linker, GFLG, which can be efficiently
cleaved by endosomal peptidases, but is stable in the circula-
tion.34 The steps of synthesis performed to generate this targeted
biomacromolecular vehicle are depicted in Scheme 1.
MTX-GFLG was prepared via several steps using an NHS
activating group. NHS was added to the in situ activated MTX
with DCC to obtain the MTX-NHS ester derivative, followed
by a reaction with an excess of the tetra-peptide GFLG, thereby
resulting in a MTX-GFLG derivative. The MTX-GFLG was
then treated with NHS followed by addition of EDA, hence,
yielding the MTX-GFLG-ethylamine derivative. FA propy-
lamine was similarly prepared using an NHS activating group
followed by addition of propanediamine.
Flow Cytometric Analysis. C5-FRR cells were routinely grown in
FA-deficient RPMI-1640 medium (GIBCO) supplemented with 0.2 nM
FA, 1 mM sodium pyruvate, and 10% dialyzed fetal calf serum (dFCS).
Parental C5 cells were grown in RPMI-1640 medium containing 2.3
µM FA, 1 mM pyruvate and 10% FCS. All cells were grown under
monolayer conditions in disposable T-75 tissue culture flasks (NUNC,
Roskilde Denmark) in a humidified atmosphere of 5% CO2 at 37 °C.
A total of 24 h prior to the experiment, the growth medium was removed
and cells were washed once with excess PBS to eliminate residual FA.
After removing the PBS, warm FA-free medium (37 °C) was added.
Cells were then incubated at 37 °C until the next day. Then cells were
detached by trypsinization and resuspended in FA-free medium
containing 5% dialyzed FCS, supplemented with 20 mM HEPES buffer
pH ) 7, setting cell concentration to 105 cells/mL. Portions of 1 mL
of cell suspension (105 cells) were dispensed into 1.5 mL disposable
Eppendorf tubes, followed by the addition of the tested conjugate at
the predetermined concentration. Test tubes were then incubated
protected from light for 90 min at 37 °C and shaken every 20 min.
Then the test tubes were centrifuged (370 × g at 4 °C for 3 min) and
the cell pellets were washed three times with cold (4 °C) PBS. Cells
were then resuspended in cold PBS and kept on ice protected from
light until flow cytometric analysis.
Analysis Using Two-Photon Confocal Scanning Laser Microscope.
A total of 48 h prior to the experiment, cells were seeded in 10 mL of
FA-free growth medium in 10 cm Petri dishes (Nunc, Roskilde,
Denmark). On the day of the experiment, the growth medium was
replaced with 5 mL of fresh FA-free medium. Cells were then subjected
to various treatments and the culture plates were covered with aluminum
foil to avoid photobleaching of the fluorescent samples. Draq-5 DNA
dye was added 15 min prior to microscopic analysis, as recommended
by the manufacturer (Alexis Biochemicals, Montreal, Quebec, Canada),
and cells were incubated at 37 °C for 10 min. Cells were then washed
with fresh medium and examined in a two-photon confocal laser
scanning microscope (LSM510, Carl Zeiss, Germany).
Prior to the conjugation, arabinogalactan was oxidized by
reacting the polysaccharide in aqueous medium with potassium
periodate at 1:1 mol ratio (IO4-/AG unit) for 2 h in a light-
protected container. This oxidation process results in saccharide
ring-opening,16 as shown in Scheme 1. A typical aldehyde-
carbonyl peak appeared in the FTIR spectrum at 1724 cm-1
.
The resulting polyaldehyde was purified from iodate and
unreacted periodate ions by DOWEX-1 anion exchange chro-
matography (acetate form, pH 7).16
To prepare the polymeric targeted nanovehicle, oxidized AG
was reacted with the FA-propylamine and the MTX-GFLG-
ethylamine to form an azomethene bond (Schiff base). The drug-
AG Schiff bases were then reduced to the amine bond using
sodium borohydride. The unreacted aldehyde groups on the
oxidized AG were reduced to the corresponding alcohols to
prevent further binding of the FA-AG-GFLG-MTX conjugate
to proteins and body components in vivo.
To prepare the polymeric targeted nanovehicle, oxidized AG
was reacted with the FA-propylamine and the MTX-GFLG-
ethylamine followed by reduction with sodium borohydride to
obtain the amine-based conjugate.
Cytotoxicity Assay and Analysis of Folate Requirement for
Growth. Each well of a 96-well microplate (Nunc, Roskilde, Denmark)
was seeded with 2000 cells in 0.1 mL and incubated in growth medium
for 24 h prior to analysis. Various treatments (depending on the specific
experiment performed as detailed below) were introduced to these plates
which were then incubated for 96 h in a humidified atmosphere of 5%
CO2 at 37 °C. Cell growth was followed by the XTT reagent assay
(Biological Industries, Beth-Haemek, Israel), with a formazan color
development incubation time of 4 h. The number of viable cells in
each well was determined using a calibration curve that was prepared
for this purpose based on color development obtained at that time.
Absorbance at a wavelength of 490 nm was determined with a Bio-
Tek plate reader spectrophotometer (Synergy HT, Bio-Tek, U.S.A.).
A portion of this conjugate was also linked to the fluorescent
molecule, FITC, and was subsequently used to visualize the
membrane binding, endocytosis, and intracellular localization
of the drug carrier in FRR-overexpressing cells.
Grafting of FA and MTX on the conjugates had no significant
effect on the molecular weights and polydispersity index,
evaluated by SEC analysis, as they were found very close to
that of the oxidized AG used (Table 1). Average molecular
weight of the FA-AG-GFLG-MTX conjugate was 15.75 kDa.
1H NMR Analysis of the Conjugates. The conjugations were
confirmed by 1H NMR spectroscopy (as detailed in the Methods
3. Results and Discussion
Chemistry. In the current study we have prepared a novel
biomacromolecular drug delivery system by grafting a model
1
section). Figure 1 depicts the H NMR spectra of FA, MTX,

Target-Activated Release of Anticancer Drugs
Biomacromolecules, Vol. 11, No. 1, 2010 299
Figure 3. Flow cytometric analysis of FRR expression in parental C5
cells (A) and in FRR-overexpressing C5-FRR cells (B). Autofluores-
cence of both cell lines (C). FRR-expressing cells were detected by
viable immunofluorescence using the anti-FRR MOv18 monoclonal
antibody. This assay revealed an 85-fold FRR overexpression of the
transfectant C5-FRR cell line relative to parental C5 cells.
Figure 1. 1H NMR spectra of FA, MTX, MTX-GFLG, oxidized AG,
and the complete construct, FA-(PDA)-AG-(EDA)-GFLG-MTX. Peaks
common to the components, and the final construct are marked with
vertical frames. Spectra were obtained at 25 °C using a 300 Hz 1H
NMR.
Figure 2. UV spectrograms of free FA and MTX as well as the FA-
AG-GFLG-MTX conjugate.
Figure 4. FA growth requirement of parental C5 cells (circles) and
their C5-FRR transfectant cells (triangles). Error bars represent
standard error. Cells were incubated for 7 days, and cell growth was
evaluated using a colorimetric assay. Whereas the concentration of
FA that supports 50% maximal growth (EC50) of parental C5 cells
was ∼100 nM, that of the stable C5-FRR transfectant cells was 100-
fold lower (i.e., ∼1 nM).
GFLG-MTX, ungrafted AG, and grafted AG with FA and
GFLG-MTX. In contrast to unconjugated AG, AG linked with
FA and GFLG-MTX demonstrated common signal peaks with
that of unbound drug and of FA. In a detail, the peaks which
appeared at 0.8, 1.4, 1.6, and 7.2 ppm were common to both
MTX-GFLG and GFLG. Further resonance recorded at 4.2, 6.9,
7.6, and 8.6 ppm is attributed to FA. Peaks obtained from MTX
and FA appeared in the same region due to similar chemical
structure. Glucose hydrogens of the drug conjugate were
detected at 3.5-4.5 ppm.
To corroborate the functionality of FRR overexpression, we
determined the FA growth requirement of both parental C5 and
C5-FRR transfectant cells (Figure 4); whereas the concentration
of FA that supports 50% maximal growth (EC50) of parental
C5 cells was ∼100 nM, that of the stable C5-FRR transfectant
cells was ∼1 nM, that is, 100-fold lower. Hence, the 100-fold
decrease in the FA growth requirement in transfectant cells is
in good agreement with the extent of surface FRR overexpres-
sion and further corroborates the functionality of the overex-
pressed FRR.
We further investigated our paired cell line model to
determine the exact concentrations where FRR saturation occurs;
this was achieved by measuring the uptake of a FA-FITC
conjugate in the presence of increasing concentrations of the
latter (Figure 5A). C5-FRR transfectant cells displayed an EC50
of ∼2 nM of FA-FITC and saturation was reached at ∼10 nM.
In contrast, parental C5 cells showed no detectable FA-FITC
uptake in the concentration range tested, which emphasizes the
sensitivity of this in vitro paired cell line model. This experiment
also provides crucial information for future experiments regard-
ing the effective concentration range of FA-AG-FITC conjugates
UV Spectroscopy Results. The degrees of substitution with
FA and MTX were evaluated by UV spectroscopy (Figure 2,
eq 1) and the results are presented in Table 1.
Characterization of FRr-Overexpressing Cells. We have
established a cell line system comprising (1) parental CHO-
AA8-C5 cells (designated “C5”) that are devoid of reduced
folate carrier (RFC/SLC19A1) transport activity,32 which es-
sentially lack FRR expression, and (2) their stable transfectant
cell line CHO-AA8-C5-FRR highly overexpressing FRR (des-
ignated “C5-FRR”). Flow cytometric analysis of plasma mem-
brane expression of FRR in viable C5-FRR transfectant cells
stably overexpressing FRR was performed using the anti-FRR
monoclonal antibody MOv18 which recognizes an external
epitope of FRR on the surface of viable cells; this analysis
revealed 85-fold FRR overexpression in the stable C5-FRR
transfectants cells when compared to their parental C5 coun-
terpart (Figure 3).

300 Biomacromolecules, Vol. 11, No. 1, 2010
Pinhassi et al.
Figure 6. Dose-dependence of FA-AG-MTX-FITC conjugate ac-
cumulation in parental C5 cells (circles) and their FRR overexpressing
subline, C5-FRR (triangles). Following a 24 h incubation in FA-free
medium, cells were detached and resuspended in FA-free medium
at pH 7.4. To 1 mL aliquots of 105 cells, FA-AG-MTX-FITC conjugate
was added, incubated for 1 h (37 °C), and analyzed with a flow
cytometer. The polymeric AG-conjugate showed an EC50 of ∼80 nM
and saturation was reached at ∼300 nM. Parental C5 cells showed
no conjugate uptake in the concentration range tested.
Figure 5. (A) FA-FITC conjugate titration curve for C5-FRR cells
(triangle) and parental C5 cells (circles). Following a 24 h incubation
in FA-free medium, cells were detached and resuspended in FA-free
medium at pH 7.4 (105 cells/mL). To 1 mL aliquots of 105 cells, FA-
FITC conjugate was added at each predetermined concentration,
incubated for 1 h (37 °C), and analyzed by a flow cytometer (AU )
arbitrary units). C5-FRR transfectant cells displayed an EC50 of ∼2
nM of FA-FITC and saturation was reached at ∼10 nM, whereas
parental C5 cells showed no detectable FA-FITC uptake in the
concentration range tested. (B) Time-course of cellular FA-FITC
accumulation in C5-FRR transfectant cells and their parental C5 cells,
at a constant FA-FITC conjugate concentration of 200 nM. Following
the addition of the FA-FITC conjugate, at a fixed concentration of
200 nM, cells were incubated for predetermined times, washed with
cold PBS, and immediately analyzed with a flow cytometer.
Figure 7. Relative FRR affinity assay: Competition between FA-AG-
MTX-FITC conjugate and free FA for binding to C5-FRR and to parental
C5 cells. Cells were coincubated for 2 h in increasing concentrations of
free FA at a constant concentration (200 nM) of the FA-AG-MTX-FITC
conjugate. Then fluorescence per cell was determined by flow cytometry.
A 50% competition was achieved with 100 nM free FA: C5-FRR cells
(triangles) and parental C5 cells (circles).
(0.1-1 µM, assuming that their affinity does not change
drastically relatively to that of FA-FITC). It is noteworthy that
this concentration range is exactly the one previously reported.36
Moreover, our data (Figure 5B) suggest that membrane
binding occurs within a few minutes. This finding is in accord
with a previous study by Turek et al.35 who demonstrated that
FA-protein particles bind membrane FR in KB carcinoma cells
in less than 15 min. According to their study, intracellular
endosomes were noticeable after an hour, and conjugate release
to the cytoplasm was detectable after ∼6 h. We hence conclude
that the present in vitro paired cell line model is a highly
sensitive and reliable system for the examination of FA-
conjugate based targeted delivery systems.
Exploiting this paired cell line model, we screened the AG-
based conjugates for specificity and selectivity.37 Conjugates
underwent three stages of evaluation. (a) Examination of the
selectivity and specificity of binding and cell association to
membrane FRR: (i) One negative control for binding specificity
tested the conjugate’s binding to parental C5 cells lacking FRR
expression. (ii) A second negative control was coincubation of
FRR-overexpressing cells with the tested conjugate along with
a large molar excess of free FA. (b) Subcellular localization of
fluorescent conjugates in various intracellular compartments,
including endosomes by confocal microscopy, to examine
internalization of the conjugate via endocytosis into FRR-
overexpressing transfectant cells. (c) Evaluation of the cytotoxic
activity of the conjugate: conjugates harboring drug cargo should
selectively eliminate target FRR-overexpressing cells due to the
intracellular release of the cytotoxic cargo.
Cell-Association and Selectivity of Cellular Uptake of a
FA-AG-MTX Conjugate. Following these three complemen-
tary readouts, we first estimated the affinity of FITC-labeled
FA-(PDA)-AG-(EDA)-MTX conjugate for FRR using increasing
conjugate concentrations in the presence or absence of free FA.
As apparent from Figure 6, the polymeric AG-conjugate showed
an EC50 of ∼80 nM and saturation was reached at ∼300 nM.
Notably, parental C5 cells showed no conjugate uptake in the
concentration range tested, thereby underlining the selective
mode in which the AG-conjugate associates with FRR.
Moreover, cell-association was competitively eliminated when
excess FA was coincubated with the FA-AG-MTX-FITC
conjugate. Figure 7 shows an assay estimating the relative
affinity (RA) for FRR; this technique measures the ability of

Target-Activated Release of Anticancer Drugs
Biomacromolecules, Vol. 11, No. 1, 2010 301
Figure 8. Confocal laser micrographs of the subcellular localization of the FA-AG-MTX-FITC conjugate in parental C5 (A) and C5-FRR ( cells
(B). Cells were seeded in FA-free growth medium (10 mL/10 cm Petri dish) and incubated for 48 h prior to the experiment. Then, medium was
replaced with fresh FA-free medium and cells were subjected to 0.5 µM FA-AG-MTX-FITC conjugate for 2 h in FA-free medium in the dark.
Draq 5 DNA dye was added at 5 µM, and equilibrated for 10 min at 37 °C, 15 min prior to confocal microscopy analysis. Cells were then washed
and examined with confocal laser microscope. Green fluorescent endosomes are denoted by red arrows, whereas the blue fluorescent structures
are nuclei. Plasma membrane staining of FRR ( overexpressing cells is also apparent. Scale bar is 20 µm.
Figure 9. Cytotoxicity test of FA-AG-(EDA)-MTX conjugate to C5-
FRR cells (triangles) and parental C5 cells (circles). Each of the 96
wells was seeded with 2000 cells and incubated in growth medium
Figure 10. Cytotoxicity of FA-AG-EDA-GFLG-MTX conjugate to C5-
for 24 h prior to analysis. Cell growth was followed by the colorimetric
FRR cells (triangles) and to parental C5 cells (circles). Each of the
XTT assay. Evidently, no significant growth inhibition was observed
96 wells was seeded with 2000 cells and incubated in growth medium
for both cell lines with this conjugate.
for 24 h prior to analysis. FA-AG-EDA-GFLG-MTX conjugate was
added and plates were incubated for 96 h at 37 °C. Cell growth was
followed by the colorimetric XTT assay. Absorbance was determined
at 490 nm. The mean IC50 of C5-FRR and for parental C5 cells were
320 ( 5 and 2000 ( 126, respectively. Tests were conducted in
tetraplicates (error bars represent standard error).
the conjugate to directly compete with free FA for binding to
cell surface FRR;38 and when 500 nM of the AG-conjugate (in
terms of the whole polymer) were coincubated with increasing
concentrations of free FA, we observed that 100 nM of free
FA were required to achieve 50% competition. Thus, this
polymeric carrier had successfully complied with the first two
criteria in the screening process, thereby demonstrating selective
cell association and uptake by FRR-overexpressing cells.
Using confocal laser microscopy (Figure 8), we then con-
firmed that selective endocytosis of FA-AG-MTX-FITC con-
jugate by C5-FRR transfectant cells occurred and not by C5,
wild-type cells.
Next, we tested the FA-AG-(EDA)-MTX conjugate for selective
cytotoxicity. Figure 9 shows the results of the comparison between
growth of FRR-overexpressing cells and parental C5 cells in the
presence of FA-AG-(EDA)-MTX. Evidently, no significant cyto-
toxic effect was observed on both cell lines.
The fact that no cytotoxicity was observed albeit the effective
selective uptake, suggested the drug is not released from its
polymeric carrier after endocytosis. Therefore, we added a
cleavable peptide linker (GFLG), which has been reported10,30
to be stable in the circulation but cleavable by endolysosomal
proteases. Thus, the addition of such a linker constitutes a target-
activated release mechanism, which may improve the safety and
the targeting selectivity of drug delivery. The FA-AG-(EDA)-
GFLG-MTX conjugate (that is devoid of FITC labeling) was
tested for its cytotoxic activity on cells lacking or overexpressing
FRR. We first compared the cytotoxicity of this construct to
parental C5 cells versus that observed with C5-FRR cells. We
hypothesized that the preferential accumulation of the MTX-
bearing FA-conjugate to cells overexpressing FRR should result
in enhanced cytotoxicity when compared to FRR-lacking
parental cells and that the cleavability of the peptide linker would
enable drug efficacy. Indeed, the cytotoxicity data support our
hypotheses; we observed a 6.3-fold higher cytotoxicity for FRR
overexpressing C5-FRR cells when compared to parental C5
cells lacking FRR expression (Figure 10). This indicates a
moderate but significant selectivity. The mean IC50 values
obtained for C5-FRR and parental C5 cells were 320 ( 5 and

302 Biomacromolecules, Vol. 11, No. 1, 2010
Pinhassi et al.
Figure 11. Cytotoxicity of FA-AG-EDA-GFLG-MTX conjugate to C5-
FRR cells incubated in FA lacking medium (empty triangles) or in 50
µM FA-containing medium (solid triangles). The experiment was
conducted in pentaplicates. Error bars represent standard error.
Cytotoxicity test was performed as described in Figure 10 legend. A
6.3-fold higher cytotoxicity was obtained for cells incubated in FA-
free medium relative to cells incubated with excess FA. The average
IC50 values measured for FRR cells in the absence or presence of
50 µM of FA were 400 ( 13 and 2500 ( 65, respectively.
Figure 12. Evaluating the necessity for FA as a targeting moiety on
the AG-GFLG-MTX conjugate. FRR cells exposed to FA-AG-EDA-
GFLG-MTX (triangles) compared to the same cells exposed to AG-
EDA-GFLG-MTX (solid squares). Cytotoxicity tests were performed
as described in Figure 10 legend.
selective delivery and target-activated release of chemothera-
peutic agents. An important potential advantage of such a
polymeric vehicle is that, by loading it with several synergisti-
cally active antitumor agents at optimal stoichiometric ratios,
effective treatment may be achieved toward malignant cells as
well as multidrug resistant tumor cells.1,39 Moreover, such a
vehicle may be tailored for a particular disease or a particular
patient as an initial step toward personalized medicine.
2000 ( 126 nM, respectively. An increased ratio of the IC50
values observed for the above paired cell lines might be
achievable by increasing the accessibility of the polymer-bound
FA to FRR by conjugating more FA residues per polymer
molecule or by using a longer linker arm, thereby improving
the effectiveness and selectivity of this polymeric nanovehicle.
Moreover, the cytotoxicity may be improved by loading the
polymer with more MTX, and/or other anticancer drugs.
To further corroborate the FA-mediated uptake of this
cytotoxic construct, an additional experiment was undertaken
(Figure 11), where C5-FRR cells were incubated in the presence
of excess free FA (50 µM). In this case, a 6.3-fold higher
cytotoxicity was also obtained for C5-FRR cells incubated in
FA-free medium relative to cells incubated with excess FA. The
average IC50 values obtained for C5-FRR cells in FA-lacking
medium and in medium supplemented with 50 µM FA were
400 ( 13 and 2500 ( 65, respectively.
The targeting function of FA is obvious when the vehicle is
carrying drugs that have no interaction with cell membrane
receptors. However, because in the current example, MTX, a
FA analogue was used as the test anticancer drug, it was
important to evaluate the necessity of FA, as MTX has some
affinity to FRR, which may suffice for targeting, thereby serving
a dual role. To test this hypothesis, an AG conjugate lacking
FA was synthesized (AG-EDA-GFLG-MTX), and its cytotox-
icity was tested on FRR cells in comparison with the complete
conjugate (FA-AG-EDA-GFLG-MTX). These results are de-
picted in Figure 12. Considering also the relative compositions
of these two vehicles (Table 1), the advantage of including FA
as a targeting moiety is clearly demonstrated: AG-EDA-GFLG-
MTX carried 4 moles of MTX per mole of AG, while FA-AG-
EDA-GFLG-MTX carried only 0.8 moles MTX per mole AG,
and 1.7 mole FA, and yet, the IC50 of the latter was 40% when
compared to that of the former. This is due to the significantly
higher affinity of FRR for FA, compared to its affinity for MTX.
Acknowledgment. The financial assitance of the Eliyahu
Penn Fund and The Technion Interdisciplinary Collaboration
fund are gratefully acknowledged. The generous contribution
of the Russell Berrie Nanotechnology Institute for the acquisition
of the spectrofluorometer used in this study is thankfully
accredited. This study was also supported by the Fred Wysz-
kowski Cancer Research Fund. We gratefully acknowledge Prof.
Philip S. Low for the generous gift of the FA-FITC conjugate
and Dr. Silvana Canevari for the Mov18 monoclonal antibody.
References and Notes
(1) Assaraf, Y. Cancer Metastasis ReV. 2007, 26, 153–181.
(2) Sinha, R.; Kim, G. J.; Nie, S. M.; Shin, D. M. Mol. Cancer Ther.
2006, 5, 1909–1917.
(3) Brannon-Peppas, L.; Blanchette, J. O. AdV. Drug DeliVery ReV. 2004,
56, 1649–1659.
(4) Dyba, M.; Tarasova, N. I.; Michejda, C. J. Curr. Pharm. Des. 2004,
10, 2311–2334.
(5) Kumar, A.; Srivastava, A.; Galaev, I. Y.; Mattiasson, B. Prog. Polym.
Sci. 2007, 32, 1205–1237.
(6) Duncan, R. Nat. ReV. Cancer 2006, 6, 688–701.
(7) Haag, R.; Kratz, F. Angew. Chem., Int. Ed. 2006, 45, 1198–1215.
(8) Zhang, J.; Lan, C. Q.; Post, M.; Simard, B.; Deslandes, Y.; Hsieh,
T. H. Cancer Genomics Proteomics 2006, 3, 147–157.
(9) Freiberg, S.; Zhu, X. X Int. J. Pharm. 2004, 282, 1–18.
(10) Satchi-Fainaro, R.; Duncan, R.; Barnes, C. M. AdV. Polym. Sci. 2006,
193, 1–65.
(11) Vicent, M. J.; Duncan, R. Trends Biotechnol. 2006, 24, 39–47.
(12) Watanabe, J.; Iwamoto, S.; Ichikawa, S. Colloids Surf., B 2005, 42,
141–146.
(13) Zhang, H.; Mardyani, S.; Chan, W. C. W.; Kumacheva, E. Biomac-
romolecules 2006, 7, 1568–1572.
(14) Emanuel, N.; Kedar, E.; Bolotin, E. M.; Smorodinsky, N. I.; Barenholz,
Y. Pharm. Res. 1996, 13, 861–868.
(15) Gabizon, A.; Catane, R.; Uziely, B.; Kaufman, B.; Safra, T.; Cohen,
R.; Martin, F.; Huang, A.; Barenholz, Y. Cancer Res. 1994, 54, 987–
992.
(16) Falk, R.; Domb, A. J.; Polacheck, I. Antimicrob. Agents Chemother.
1999, 43, 1975–1981.
(17) Duncan, R. Nat. ReV. Drug DiscoVery 2003, 2, 347–360.
(18) Leamon, C. P. Curr. Opin. InVest. Drugs 2008, 9, 1277–1286.
4. Conclusions
Taken collectively, our data establish that the novel AG
anticancer biomacromolecular drug carrier devised here pos-
sesses a promising potential as a targeted nanovehicle for the

Target-Activated Release of Anticancer Drugs
Biomacromolecules, Vol. 11, No. 1, 2010 303
(19) Leamon, C. P.; Jackman, A. L. Vitam. Horm. 2008, 79, 203–233.
(20) Wang, S.; Low, P. S. J. Controlled Release 1998, 53, 39–48.
(21) Leamon, C. P.; Low, P. S. Biochem. J. 1993, 291 (Pt 3), 855–860.
(22) Shia, J.; Klimstra, D. S.; Nitzkorski, J. R.; Low, P. S.; Gonen, M.;
Landmann, R.; Weiser, M. R.; Franklin, W. A.; Prendergast, F. G.;
Murphy, L.; Tang, L. H.; Temple, L.; Guillem, J. G.; Wong, W. D.;
Paty, P. B. Hum. Pathol. 2008, 39, 498–505.
(30) Rejmanova, P.; Kopecek, J.; Duncan, R.; Lloyd, J. B. Biomaterials
1985, 6, 45–48.
(31) Zhao, H.; Heindel, N. D. Pharm. Res. 1991, 8 (3), 400–402.
(32) Rothem, L.; Berman, B.; Stark, M.; Jansen, G.; Assaraf, Y. G. Mol.
Pharmacol. 2005, 68, 616–624.
(33) Roberts, S. J.; Petropavlovskaja, M.; Chung, K. N.; Knight, C. B.;
Elwood, P. C. Arch. Biochem. Biophys. 1998, 351, 227–235.
(34) Duncan, R. Nat. ReV. Drug DiscoVery 2003, 2, 347–360.
(35) Gabizon, A.; Horowitz, A. T.; Goren, D.; Tzemach, D.; Mandelbaum-
Shavit, F.; Qazen, M. M.; Zalipsky, S. Bioconjugate Chem. 1999, 10,
289–298.
(23) Leamon, C. P.; Reddy, J. A. AdV. Drug Deliery ReV. 2004, 56, 1127–
1141.
(24) Lee, R. J.; Low, P. S. Biochim. Biophys. Acta, Biomembr. 1995, 1233,
134–144.
(25) Wang, S.; Low, P. S. J. Controlled Release 1998, 53, 39–48.
(26) Goren, D.; Horowitz, A. T.; Tzemach, D.; Tarshish, M.; Zalipsky, S.;
Gabizon, A. Clin. Cancer Res. 2000, 6, 1949–1957.
(27) Leamon, C. P.; Reddy, J. A. AdV. Drug DeliVery ReV. 2004, 56, 1127–
1141.
(28) Salazar, M.; Ratnam, M. Cancer Metastasis ReV. 2007, 26, 141–152.
(29) Russell-Jones, G.; McTavish, K.; McEwan, J.; Rice, J.; Nowotnik, D.
J. Inorg. Biochem. 2004, 98, 1625–1633.
(36) Turek, J. J.; Leamon, C. P.; Low, P. S. J. Cell Sci. 1993, 106, 423–
430.
(37) Leamon, C. P.; Reddy, J. A. AdV. Drug DeliVery ReV. 2004, 56, 1127–
1141.
(38) Lu, Y.; Low, P. S. AdV. Drug DeliVery ReV. 2002, 54, 675–693.
(39) Assaraf, Y. G. Drug Resist. Updates 2006, 9, 227–246.
BM900853Z