Design, synthesis, and biological evaluation of novel cRGD-paclitaxel conjugates for integrin-assisted drug delivery

Article abstract of DOI:10.1021/bc300164t

The efficacy of taxane-based antitumor therapy is limited by several drawbacks which result in a poor therapeutic index. Thus, the development of approaches that favor selective delivery of taxane drugs (e.g., paclitaxel, PTX) to the disease area represen

Full text of DOI:10.1021/bc300164t

Article
pubs.acs.org/bc
Design, Synthesis, and Biological Evaluation of Novel cRGD−
Paclitaxel Conjugates for Integrin-Assisted Drug Delivery
Michael Pilkington-Miksa,† Daniela Arosio,‡ Lucia Battistini,§ Laura Belvisi,# Marilenia De Matteo,†
Francesca Vasile,† Paola Burreddu,∥ Paola Carta,▲ Gloria Rassu,∥ Paola Perego,⊥ Nives Carenini,⊥
Franco Zunino,⊥ Michelandrea De Cesare,⊥ Vittoria Castiglioni,□,¶ Eugenio Scanziani,□,¶
Carlo Scolastico,† Giovanni Casiraghi,§ Franca Zanardi,*,§ and Leonardo Manzoni*,‡
†Centro Interdipartimentale Studi Biomolecolari e Applicazioni Industriali, Universita
I-20138 Milano, Italy
̀
degli Studi di Milano, Via Fantoli 16/15,
‡Istituto di Scienze e Tecnologie Molecolari, Consiglio Nazionale delle Ricerche, Via Golgi 19, I-20133 Milano, Italy
§Dipartimento Farmaceutico, Universita
#Dipartimento di Chimica Organica e Industriale, Universita
∥Istituto di Chimica Biomolecolare, Consiglio Nazionale delle Ricerche, Traversa La Crucca 3, I-07100 Li Punti, Sassari, Italy
▲Porto Conte Ricerche Srl, I-07041 Tramariglio Alghero, Sassari, Italy
⊥Dipartimento di Oncologia Sperimentale e Medicina Molecolare, Fondazione IRCCS Istituto Nazionale Tumori, Via Amadeo 42,
I-20133 Milano, Italy
̀
degli Studi di Parma, Parco Area delle Scienze 27A, I-43124 Parma, Italy
degli Studi di Milano, Via Venezian 21, I-20133 Milano, Italy
̀
□Dipartimento di Patologia Animale, Igiene e Sanita
Studi di Milano, Via Celoria 10, I-20133 Milano, Italy
̀ ̀ ̀
Pubblica Veterinaria (DIPAV), Facolta di Medicina Veterinaria, Universita degli
¶Mouse and Animal Pathology Laboratory, Fondazione Filarete, Viale Ortles 22/4, I-20139 Milano, Italy
S
* Supporting Information
ABSTRACT: The efficacy of taxane-based antitumor therapy
is limited by several drawbacks which result in a poor
therapeutic index. Thus, the development of approaches that
favor selective delivery of taxane drugs (e.g., paclitaxel, PTX)
to the disease area represents a truly challenging goal. On the
basis of the strategic role of integrins in tumor cell survival and
tumor progression, as well as on integrin expression in tumors,
novel molecular conjugates were prepared where PTX is
covalently attached to either cyclic AbaRGD (Azabicycloal-
kane-RGD) or AmproRGD (Aminoproline-RGD) integrin-
recognizing matrices via structurally diverse connections.
Receptor-binding assays indicated satisfactory-to-excellent
αVβ3 binding capabilities for most conjugates, while in vitro
growth inhibition assays on a panel of human tumor cell lines revealed outstanding cell sensitivity values. Among the nine
conjugate ensemble, derivative 21, bearing a robust triazole ring connected to ethylene glycol units by an amide function and
showing excellent cell sensitivity properties, was selected for in vivo studies in an ovarian carcinoma model xenografted in
immunodeficient mice. Remarkable antitumor activity was attained, superior to that of PTX itself, which was associated with a
marked induction of aberrant mitoses, consistent with the mechanism of action of spindle poisons. Overall, the novel cRGD-PTX
conjugates disclosed here represent promising candidates for further advancement in the domain of targeted antitumor therapy.
use of available drugs, in particular, cytotoxic agents. Impressive
advances in the molecular and biological characterization of
human tumors have led to the identification of specific
alterations exploitable for tumor targeting. In this context,
strong interest for integrins has recently emerged.2 Various
INTRODUCTION
■
The curative potential of clinically available antitumor agents is
limited by a number of factors which may include tumor-related
factors (e.g., activation of drug resistance mechanisms by tumor
cells), drug-related factors (e.g., inadequate intratumor
concentration of the drug), tumor microenvironment inter-
actions (e.g., hypoxia, microvesicle release), and individual
features of the patients (i.e., genetic polymorphisms).1 Thus, an
impressive effort is still required in an attempt to optimize the
Received: April 5, 2012
Revised: May 21, 2012
Published: July 9, 2012
© 2012 American Chemical Society
1610
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
studies have indicated that integrins promote cell attachment
and migration on the surrounding extracellular matrix and
mediate cell interactions during several pathological processes.2
Indeed, the available evidence supports a role for integrins in
proliferation, migration, and survival of tumor cells, as well as in
additional processes that favor tumor progression such as
angiogenesis and inflammation.2 In particular, specific integrin
receptors, i.e., the αVβ3 and αVβ5 families, have been elected as
compelling molecular indicators of tumor angiogenesis and
tumor progression, invasion, and metastasis, since they are
overexpressed on proliferating endothelial cells as well as on
various tumor cells including breast,3,4 glioma,5 melanoma,6,7
prostate,8 and ovarian carcinoma.9
A large number of specific, highly potent αVβ3- and αVβ5-
addressed small molecule ligands have been designed and
implemented so far, which contain or mimic the key receptor-
recognizing RGD (Arg-Gly-Asp) tripeptide motif;10,11 these
include the landmark cyclic pentapeptide c(RGDf[N-Me]V)
(Cilengitide, I)12 (Chart 1), the azabicycloalkane-based cyclic
RGD semipeptide II (AbaRGD),13 and the 4-aminoproline-
based cyclic RGD semipeptide III (AmproRGD).14
therapeutics where the RGD-homing function is coupled to the
effector moiety under the shape of either conventional covalent
conjugates10,20−29 or nanosized molecular/supramolecular
entities.10,20,21,30−34 Among the chemotherapeutics which
were incorporated in these RGD-conjugates, the microtubule-
interfering agent paclitaxel (PTX, 1, Scheme 1) has often been
selected, in an attempt to overcome the inherent drawbacks
such as low aqueous solubility, short half-life, poor bioavail-
ability, and systemic toxicityall features resulting in a poor
therapeutic index. The use of paclitaxel is of specific interest in
this approach, because taxanes are known to have anti-
angiogenic effects.35 Quite surprisingly, however, the broad
spectrum of tumor-addressing integrin ligands to be exploited
in conjugated entities was limited to linear RGD-embedded
peptides or no more than a couple of cyclic peptide candidates,
namely, c(RGDfK), c(RGDyK), or E-[c(RGDfK)2], which
represent the conjugable versions of Cilengitide.10,27−29
In the structural design of novel, potentially bioactive
conjugates, the judicious calibration of both the targeting
device and the appropriate linker is crucial for the successful
delivery of a given cytotoxic drug. Along this line, we present
here nine novel molecular conjugates where the common PTX
cytotoxic cargo is covalently attached to either cyclic AbaRGD
or AmproRGD integrin-recognizing matrices via structurally
diverse connections. Preeminent goals of the study were as
follows: (1) to assay the feasibility of the chemistry involved by
means of simple and efficient conjugation reactions under
gentle and faithful conditions; (2) to assess the structural
integrity/purity of the targets; (3) to evaluate by biological tests
in vitro the mutual influence of the Aba/AmproRGD, PTX, and
linker modules on the αVβ3/αVβ5 targeting capabilities,
solubility, and tumor cell sensitivity; and (4) to assay via in
vivo biological studies the antitumor activity of a selected
candidate in order to gain insight into possible structural
amendments and future development of the proposed
conjugates.
a
Chart 1High-Affinity αVβ3-Targeting Cyclic RGD Ligands
EXPERIMENTAL PROCEDURES
■
Chemistry. General. All chemicals and solvents were of
reagent grade and were used without further purification.
Solvents were dried by standard procedures and reactions
requiring anhydrous conditions were performed under nitrogen
1
or argon atmosphere. H and 13C NMR spectra were recorded
at 300 K on Bruker AVANCE-400 or Bruker AVANCE-600
MHz spectrometers. Chemical shifts δ are expressed in ppm
relative to internal Me4Si as the standard reference. Mass
spectra were obtained with an ESI apparatus Bruker Esquire
3000 plus. Thin-layer chromatography (TLC) was carried out
with precoated Merck F254 silica-gel plates. Flash chromatog-
raphy (FC) was carried out with Macherey-Nagel silica gel 60
(230−400 mesh). Reverse-phase column chromatography was
carried out with the Biotage SP1 or SP4 systems using the
Biotage 25+M C18 cartridges. Unless otherwise stated,
semipreparative HPLC was carried out on a Waters
SymmetryPrep C18-7 μm 7.8 × 300 mm column or a Waters
Atlantis C18 OBD 5 μm 19 mm × 10 cm or a Supelco Ascentis
RPAmide 5 μm 21.1 mm × 15 cm; gradient from 100% H2O (+
0.1% TFA) to 75% H2O (+ 0.1% TFA)/25% MeCN (+ 0.1%
TFA) over 30 min.
a
b
c
aRef 12. Ref 13. Ref 14.
Apart from the preeminent use of small-molecule integrin
antagonists as antiangiogenic and antitumor agents per
se,10,11,15−18 such motifs may be exploited with even more
success as active targeting tools, provided that appropriate
cytotoxic and/or imaging cargos are consigned to them.10,19−21
Since the pioneering work by Arap and Pasqualini,22 where
effective delivery of the antitumor agent doxorubicin to human
breast cancer xenografts in nude mice was given in charge of a
small RGD-doxo conjugate, several reports have appeared
concerning the synthesis and biology of actively targeted
Representative Procedure A. Preparation of C2′ Paclitaxel
Ester 7. To Paclitaxel 1 (PTX, 50 mg, 0.059 mmol, 1.0 equiv),
4-pentynoic acid 2 (6.32 mg, 0.0644 mmol, 1.1 equiv), and a
catalytic quantity of N,N-dimethylaminopyridine (DMAP) in
1611
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
Scheme 1. Synthesis of Activated Taxane Modules 7−11
dry dichloromethane (1.0 mL), diisopropylcarbodiimide (DIC,
13.6 μL, 11.08 mg, 0.088 mmol, 1.5 equiv) was added. The
resulting solution was left to stir under argon at room
temperature for 20 h. The reaction mixture was then
concentrated in vacuo to give a yellow solid. This was purified
by reverse-phase chromatography eluting with water and
acetonitrile to give a white solid in 99% yield. For character-
ization data of compound 7, see Supporting Information.
Compounds 8 and 9 were obtained following the above
procedure as white solid in 99% and 94% yield, respectively.
Paclitaxel Ester 8. The title compound was obtained from
Paclitaxel 1 and undec-10-ynoic acid 3 following representative
procedure A. After reverse-phase chromatography, compound 8
was recovered as a white solid in 99% yield. For character-
ization data of compound 8, see Supporting Information.
Paclitaxel Ester 9. The title compound was obtained from
Paclitaxel 1 and alkyne linker 4 (preparation reported in the
Supporting Information) following representative procedure A.
After reverse-phase chromatography, compound 9 was
recovered as a white solid in 94% yield. For characterization
data of compound 9, see Supporting Information.
Representative Procedure B. Preparation of C2′ Paclitaxel
Ester 10. To PTX 1 (50 mg, 0.059 mmol, 1.0 equiv) and
diglycolic anhydride 5 (8.6 mg, 0.074 mmol, 1.25 equiv) in dry
dichloromethane (1.0 mL), pyridine (0.474 mL, 463 mg, 5.855
mmol, 100 equiv) and a catalytic quantity of DMAP were
added. The resulting solution was left to stir under argon for 3
h at 0 °C and then left to warm to room temperature. After 20
h, the reaction mixture was concentrated in vacuo; the resulting
solid was dissolved in water/acetonitrile/acetic acid (3 mL,
1:1:1) and purified by reverse-phase chromatography eluting
with water and acetonitrile to give a white solid in 99% yield.
For characterization data of compound 10, see Supporting
Information.
11 was recovered as a white solid in 95% yield. For
characterization data of compound 11, see Supporting
Information.
Preparation of Cyclic AbaRGD (Compounds 12 and 13)
and AmproRGD (Compounds 14−16 and 18) Modules.
Cyclic AbaRGD compounds 12 and 13 were prepared
according to a reported procedure.13 Cyclic AmproRGD
compounds 14−16 were prepared according to a reported
procedure.14
Compound 18. Azidopeptide 14 (280 mg, 0.55 mmol, 2.2
equiv) and dialkyne 1736 (93 mg, 0.25 mmol, 1.0 equiv),
dissolved in 20 mL of tert-butanol/water (1:1), were
sequentially treated with 0.3 M solution of copper(II) acetate
(333 μL, 0.1 mmol, 0.4 equiv) and 0.9 M solution of sodium L-
ascorbate (222 μL, 0.2 mmol, 0.8 equiv). The resulting
heterogeneous mixture was sonicated for 1 h and then stirred at
room temperature for additional 4 h. After reaction completion,
the mixture was lyophilized and the residue was purified by
flash chromatography (EtOAc/MeOH, 85:15 to 75:25)
affording a protected dimeric peptide (250 mg, 72%) as a
glassy solid. This protected intermediate (250 mg, 0.18 mmol,
1.0 equiv) was treated with a solution of TFA/TIS/water
95:2.5:2.5 (16 mL) at room temperature. After 5 h, the solvent
was evaporated under vacuum and the residue dissolved in
water (15 mL) and thoroughly washed with Et2O (4×). The
aqueous phase was concentrated under vacuum furnishing 240
mg (95%) of dimeric amine 18 as a trifluoroacetate salt, which
was used as such in the following coupling step. For
characterization data of compound 18, see Supporting
Information.
Representative Procedure C. Preparation of Triazolyl-
Linked Conjugate 19. To azido-cRGD compound 12 (33.8
mg, 0.059 mmol, 1.0 equiv) and PTX ester 7 (54.7 mg, 0.059
mmol, 1.0 equiv) in tert-butanol/water (6:4, 3 mL) was added
0.9 M sodium L-ascorbate (28.39 μL, 0.026 mmol, 0.4 equiv).
To this solution was then added 0.3 M copper(II) acetate (42.6
μL, 0.013 mmol, 0.2 equiv) and the resulting mixture was left to
Paclitaxel Ester 11. The title compound was obtained from
PTX 1 and succinic anhydride 6 following representative
procedure B. After reverse-phase chromatography, compound
1612
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
stir at room temperature and monitored by LC-MS. When the
reaction reached completion (2−4 h), the solution was frozen
in liquid nitrogen and then freeze−dried. The solid recovered
from freeze−drying was purified by RP-HPLC eluting with
water + 0.1% formic acid and acetonitrile + 0.1% formic acid.
The purified product was then freeze−dried to give a white
solid in 75% yield. For characterization data of compound 19,
see Supporting Information.
Triazolyl-Linked Conjugate 20. The title compound was
obtained from azido-cRGD 12 and PTX ester 8 following
representative procedure C. After RP-HPLC purification,
compound 20 was recovered as a white solid in 59% yield.
For characterization data of compound 20, see Supporting
Information.
Triazolyl-Linked Conjugate 21. The title compound was
obtained from azido-cRGD 12 and PTX ester 9 following
representative procedure C. After RP-HPLC purification,
compound 21 was recovered as a white solid in 46% yield.
For characterization data of compound 21, see Supporting
Information.
Triazolyl-Linked Conjugate 22. The title compound was
obtained from azido-cRGD 14 and PTX ester 7 following
representative procedure C. After RP-HPLC purification,
compound 22 was recovered as a white solid in 57% yield.
For characterization data of compound 22, see Supporting
Information.
compound 26 was recovered as a white solid in 40% yield.
For characterization data of compound 26, see Supporting
Information.
Amide-Linked Conjugate 27. The title compound was
obtained from amino-cRGD 13 and PTX ester 11 following
representative procedure D. After RP-HPLC purification,
compound 27 was recovered as a white solid in 62% yield.
For characterization data of compound 27, see Supporting
Information.
Biology. Solid-Phase Receptor Binding Assay. Purified
αvβ3 and αvβ5 receptors (Chemicon International, Inc.,
Temecula, CA, USA) were diluted to 0.5 μg/mL in coating
buffer containing 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1
mM MnCl2, 2 mM CaCl2, and 1 mM MgCl2. An aliquot of
diluted receptors (100 μL/well) was added to 96-well
microtiter plates (NUNC MW 96F Medisorp Straight) and
incubated overnight at 4 °C. The plates were then incubated
with blocking solution (coating buffer plus 1% bovine serum
albumin) for an additional 2 h at room temperature to block
nonspecific binding, followed by 3 h incubation at room
temperature with various concentrations (10−5−10−12 M) of
test compounds in the presence of biotinylated vitronectin (1
μg/mL). Biotinylation was performed using an EZ-Link Sulfo-
NHS-Biotinylation kit (Pierce, Rockford, IL, USA). After
washing, the plates were incubated for 1 h at room temperature
with biotinylated streptavidin−peroxidase complex (Amersham
Biosciences, Uppsala, Sweden) followed by 30 min incubation
with 100 μL/well Substrate Reagent Solution (R&D Systems,
Minneapolis, MN) before stopping the reaction with the
addition of 50 μL/well 2 N H2SO4. Absorbance at 415 nm was
read in a SynergyTM HT Multi-Detection Microplate Reader
(BioTek Instruments, Inc.). Each data point represents the
average of triplicate wells; data analysis was carried out by
nonlinear regression analysis with GraphPad Prism 5.0 software.
Each experiment was repeated in triplicate.
Drugs. For in vitro studies, PTX and compounds under
investigation were dissolved in dimethylsulfoxide (DMSO) and
then added to culture medium. DMSO concentration in
medium never exceeded 0.5%. For in vivo studies, PTX was
dissolved in a mixture of ethanol and cremophor ELP (50 +
50%) and kept at 4 °C. At treatment, the drug was diluted in
90% of cold saline after magnetic stirring and administered i.v.
keeping the vial in ice. Compound 21 was dissolved and
administered like PTX at room temperature.
Cell Lines and Growth Conditions. The human ovarian
carcinoma IGROV-1 cell line,37 the cisplatin-resistant IGROV-
1/Pt1 subline38 and the human large cell lung H460 carcinoma
cell line (ATCC, HTB-177) were cultured in RPMI-1640
medium; the human osteosarcoma U2-OS cell line (ATCC,
HTB-96) was grown in McCoy’s 5A medium. Medium was
supplemented with 10% fetal calf serum.
Triazolyl-Linked Conjugate 23. The title compound was
obtained from azido-cRGD 15 and PTX ester 7 following
representative procedure C. After RP-HPLC purification,
compound 23 was recovered as a white solid in 48% yield.
For characterization data of compound 23, see Supporting
Information.
Representative Procedure D. Preparation of Amide-Linked
Conjugate 24. To PTX-ester 10 (122 mg, 0.126 mmol, 1.0
equiv) and N-hydroxysulfosuccinimide sodium salt (34.1 mg,
0.157 mmol, 1.25 equiv) in dry N,N-dimethylformamide
(DMF, 4.0 mL) was added diisopropylcarbodiimide (DIC,
29.2 μL, 23.8 mg, 0.188 mmol, 1.5 equiv). The resulting
mixture was left to stir under argon at room temperature for 24
h. The reaction was then concentrated in vacuo to give an off-
white solid. To this solid was added amino-cRGD 13 (108 mg,
0.126 mmol, 1.0 equiv) in acetonitrile (2.0 mL) and pH 7.0
phosphate buffer (PBS, 0.5 M, 1.0 mL), and the resulting
mixture was rapidly cooled to 0 °C and then sonicated to favor
dissolution. The reaction mixture was then left to stir at 0 °C
for 10 h after which time it was allowed to warm to room
temperature. After 18 h stirring, dioxane/water (1:1, 10 mL)
was added and the resulting solution was freeze−dried. The
solid recovered from freeze−drying was purified by RP-HPLC
eluting with water + 0.1% formic acid and acetonitrile + 0.1%
formic acid. The purified product was then freeze−dried to give
a white solid in 43% yield.
Cell Sensitivity to Drugs. The cell sensitivity to antitumor
agents was measured by using the growth-inhibition assay based
on cell counting. Cells were seeded in duplicates into 6-well
plates and exposed to drug 24 h later. The compounds were
dissolved in DMSO at 3.3 mM. Before cell treatment, the
compounds were further diluted in DMSO. Five microliters of
200× solutions were used for treating cells in a volume of 1 mL
of culture medium. The final DMSO concentration was not
toxic. After 72 h of drug incubation, cells were harvested for
counting with a cell counter. IC50 is defined as the drug
concentration producing 50% decrease of cell growth. At least
three independent experiments were performed.
Compounds 25−27 were obtained following the above
procedure as white solid in 45%, 40%, and 62%, respectively.
Amide-Linked Conjugate 25. The title compound was
obtained from amino-cRGD 16 and PTX ester 10 following
representative procedure D. After RP-HPLC purification,
compound 25 was recovered as a white solid in 45% yield.
For characterization data of compound 25, see Supporting
Information.
Amide-Linked Conjugate 26. The title compound was
obtained from amino-cRGD 18 and PTX ester 10 following
representative procedure D. After RP-HPLC purification,
1613
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
Scheme 2. Structures of Monomeric AbaRGD and AmproRGD Modules 12−16 and Synthesis of Dimeric AmproRGD Module
18
Analysis of Integrin Levels. The expression of integrins was
measured by flow cytometry, following optimization of
antibody concentration. Exponentially growing cells were
harvested and incubated for 30 min at 4 °C with anti-human
αvβ3 or αvβ5 antibodies or isotypic controls (Millipore,
Temecula, CA; Chemicon International). Cells were then
washed and samples were immediately used for flow cytometric
analysis (FACScan, Becton-Dickinson). Integrin levels were
expressed as the ratio between the mean fluorescence intensity
obtained in cells incubated with anti-integrin antibodies and
that of cells incubated with isotypic control (Table 2).
Antitumor Activity Studies. All experiments of antitumor
activity were carried out using female athymic Swiss nude mice,
8−10 weeks old (Charles River, Calco, Italy). Mice were
maintained in laminar flow rooms keeping temperature and
humidity constant. Mice had free access to food and water.
Experiments were approved by the Ethics Committee for
Animal Experimentation of the Istituto Nazionale Tumori of
Milan according to institutional guidelines. The IGROV-1/Pt1
human tumor xenograft derived from cultures of the ovarian
carcinoma cell line38 was used. Exponentially growing cells were
s.c. injected into the right flank of athymic nude mice, and the
tumor line was achieved by serial s.c. passages of fragments of
regrowing tumors into healthy mice. Groups of five mice
bearing bilateral s.c. tumors were employed. Tumor fragments
were implanted on day 0 and tumor growth was followed by
biweekly measurements of tumor diameters with a Vernier
caliper. Tumor volume (TV) was calculated according to the
formula: TV (mm3) = d2 × D/2 where d and D are the shortest
and the longest diameter, respectively. Drugs were delivered i.v.
into the lateral tail vein using a 26G needle equipped syringe.
Drugs were administered in a volume of 10 mL/kg of body
weight every 4 days for 4 times (q4d × 4). Treatment started
three days after tumor implant, when tumors were just palpable.
The efficacy of the drug treatment was assessed as tumor
volume inhibition percentage (TVI%) in treated versus control
mice, calculated as follows: TVI% = 100 − (mean TV treated/
mean TV control × 100) and log cell kill, evaluated by the
formula: LCK = (T − C)/(3.32 × DT), where T and C are the
times (days) to reach a tumor volume in treated and control
mice, respectively, and DT is the doubling time of control
tumors. Tumor DT (days) was obtained from semilog best-fit
curves of mean tumor volumes in untreated control mice
plotted against time. The toxicity of the drug treatment was
determined as body weight loss and lethal toxicity. Student's t
test (two-tailed) was used for statistical comparison of tumor
volumes in mice.
Immunohistochemistry. Tumor xenografts and adjacent
tissues were excised and formalin fixed and paraffin embedded.
Four micrometer sections from each tumor xenograft were
routinely stained with Hematoxylin-Eosin (HE) and evaluated
under a light microscope. Mitoses were evaluated in 3 randomly
selected 400× fields within the bulk of the xenograft, avoiding
areas of necrosis and hemorrhage. The total number of mitoses
and the mean value for each sample were evaluated.
Furthermore, mitoses were differentiated as “normal” and
“aberrant”, the latter class characterized by both small,
condensed hyperchromatic nuclei and large cells composed of
nuclear envelope around individual clusters of mis-segregated
chromosomes (mitotic catastrophe), and the ratio between
these two classes was evaluated. The analysis of mitoses was
performed in a blind fashion. Statistical analysis of the obtained
data was carried out with the Kruskal−Wallis test followed by
Dunn’s multiple comparison test using GraphPad Prism
(GraphPad Software, Inc.).
RESULTS
■
Chemistry. In engineering versatile, covalent connections of
the building modules (PTX, RGD-semipeptide, and linker),
mainly two requisites had to be satisfied. First, these linkages
had to be easy to make and robust enough to guarantee the
structural integrity of the whole construct until contacting and
entering the cell compartment in such a way to fully exploit the
specific targeting capabilities and the target-mediated cell
internalization elicited by the RGD moiety. Second, connection
to PTX, chosen at its C2′−OH position, has to be scissile
1614
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
Scheme 3. Synthesis of Triazolyl-Linked Paclitaxel-AbaRGD and Paclitaxel-AmproRGD Conjugates 19−23
(hopefully inside the cell) to allow complete exercise of the
cytotoxic function.39 Accordingly, either a triazole-forming
II and III (Chart 1), these RGD peptides have the notable
presence of exocyclic free azide or amine functional groups
prone to conjugation with external partners.
Huisgen 1,3-dipolar cycloaddition reaction (“click reaction”)40
̈
or popular amide conjugation was judged a viable solution,
considering their intrinsic robustness, easy execution, and
compatibility of the experimental conditions with the exposed
functional groups within the moieties.
Compounds 12−16 were synthesized from the respective
amino acid building blocks in 30−44% yields, as reported.13,14
Another point of interest was the multipresentation of the
ligands, since it is known that in many cases close correlation
exists between the “RGD-valency” of the construct and the
targeting capability enhancement.41−43 To this end, compound
18 was built, where a rigid phenolic core is twofold linked to
the AmproRGD unit through robust triazolyl connections,
while leaving a free amino function ready for further
conjugation. Thus, ad hoc-prepared bis-alkyne 17, in turn
obtained from 3,5-dihydroxy methyl benzoate and 1,2-diamino-
ethane according to a reported procedure,36 was reacted with
Synthesis of Activated Taxane Modules. The first step in
the synthesis of the conjugates involved preparation of various
C2′-esters of PTX 7−11 (Scheme 1) bearing either alkyne or
carboxylic acid terminating moieties for subsequent triazolyl/
amide conjugation. For 7−9, two commercially available,
diversely sized alkynoic acids 2 and 3, along with an ad hoc
prepared short PEG acid 4 (for preparation, see Supporting
Information), were regioselectively coupled to PTX at the free
C2′-position. Standard DIC/DMAP chemistry in dichloro-
methane was effective for preparing the corresponding esters
7−9 in 94−99% isolated yields. Notably, no parasite
esterification at the unprotected C7-hydroxyl of the taxane
was observed. For carboxylic acids 10 and 11, two commercially
available cyclic acid anhydrides, namely, glycolic anhydride 5
and succinic anhydride 6, were used and coupled to the
cytotoxic drug 1 without any byproducts under conventional
esterification conditions.
Synthesis of Cyclic AbaRGD and AmproRGD Modules. As
disclosed above, a key point of innovation of the present work
with respect to the reported studies in the field is the selection
of the αVβ3-integrin targeting unit. A first class of integrin
ligands (compounds 12 and 13 in Scheme 2) features a fifteen-
membered cyclopentapeptide where a 5,6-fused azabicycloal-
kane (Aba) amino acid moiety is tethered to the amino/acid
terminals of the Arg-Gly-Asp tripeptide sequence. A second,
different category of integrin ligands is represented by
compounds 14, 15, 16, and 18 (Scheme 2), all characterized
by a fourteen-membered cyclotetrapeptide where a 4-amino-
proline scaffold (Ampro) is grafted onto the Arg-Gly-Asp
moiety. Structurally related to the known, high-affinity ligands
azide 14 (2 mol equiv) under copper-catalyzed Huisgen 1,3-
̈
dipolar cycloaddition conditions40 to furnish, after acidic
removal of the Boc protection, dimeric amine 18 in 68%
yield for the two steps.
Of note, RGD-cyclopeptides 12−16 and 18 were recovered
as fully deprotected, >95% pure, and homogeneous materials as
trifluoroacetate salts after semipreparative HPLC and proved
completely water-soluble (25 °C, pH 7).
Synthesis of Triazolyl-Linked Paclitaxel-AbaRGD and
Paclitaxel-AmproRGD Conjugates. With alkyne-ending taxane
esters 7−9 and azide-terminating cycloRGD-peptides 12, 14,
and 15 in hand, the next task was to couple these modules to
create the projected triazolyl-linked conjugates. The regiose-
lective copper-catalyzed Huisgen 1,3-dipolar cycloaddition
̈
between an azide and a terminal alkyne, often referred to as
“click reaction”,40 has been reported to proceed in very good
yields in t-butanol/water, a solvent mixture that appeared ideal
for dissolution of both lipophilic PTX and water-soluble
cyclopeptides. Thus, the “click” reactions of completely
deprotected azides 12, 14, or 15 with taxane alkynes 7, 8, or
9 proceeded fairly rapidly giving the corresponding triazolyl
conjugates 19−23 in 46−75% yields (Scheme 3).
1615
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
Scheme 4. Synthesis of Amide-Linked Paclitaxel-AbaRGD and Paclitaxel-AmproRGD Conjugates 24−27
Structurally speaking, these conjugates differed from each
other by the nature of the integrin-ligand moiety (e.g., AbaRGD
19−21 vs AmproRGD 22−23), by the distance of the taxane
unit from the triazole (e.g., 19 vs 20 vs 21) and by the distance
of the RGD unit from the triazole (e.g., 22 vs 23).
common AbaRGD unit linked to the taxane by different
diglycolic vs succinic acid spacers.
Biology. Receptor Binding Assay. The ability of the nine
PTX-AbaRGD and PTX-AmproRGD conjugates 19−27 to
bind to human, isolated αVβ3 and αVβ5 integrin receptors in
vitro was evaluated by competitive displacement assays using
biotinylated vitronectin (Table 1). To better estimate the
impact of taxane-linker conjugation on the binding capability of
the RGD portions, the data were compared to those of
unconjugated parent RGD counterparts 12−16 and to known
high-affinity binders c(RGDfV), Aba-based compound II, and
Ampro-based derivative III (see Chart 1).
Triazolyl conjugates 19−23 showed variable results. In
particular, AbaRGD derivatives exhibited some decrease (8-fold
for 19 and 7-fold for 21) or complete abrogation (for 20) of
αVβ3-binding affinity with respect to the parent azide
counterpart 12, while AmproRGD compounds 22 and 23
showed substantially unaltered, if not improved, one-digit
nanomolar affinity and αVβ3/αVβ5 selectivity (as compared to
azides 14 and 15).
For diglycolic amide-linked conjugates 24 and 25, a
nanomolar binding capability was observed in both cases: the
former displayed a 30-fold IC50 decrease with untouched αVβ3/
αVβ5 selectivity as compared to the parent compound 13, while
the latter exhibited a comparable IC50 value and 4-fold
increased αVβ3/αVβ5 selectivity as compared to the parent
compound 16. Succinic ester-linked Aba-conjugate 27 exhibited
dual affinity for both receptors similar to triazolyl derivative 19;
compound 26, bearing a 2-fold AmproRGD repeat, showed
notable 0.23 nM αVβ3 binding capability, in line with the
behavior observed for similar multimeric presentations.14,41−43
In this case, again, remarkable selectivity amplification was
observed.
Synthesis of Amide-Linked Paclitaxel-AbaRGD and
Paclitaxel-AmproRGD Conjugates. For the amide coupling
between primary amine RGD ligands of types 13, 16, and 18
and carboxylic acid taxanes 10 and 11, we had to face and solve
some inherent, exquisitely chemical problems such as (1) the
use of solvent mixtures compatible with both reacting partners;
(2) the minimization of undesired side reactions between the
free functional groups of the cyclopeptide and the taxane
moiety, which is normally susceptible to attacks from
nucleophiles; (3) the annihilation of the possibly competitive
acylation at the RGD guanidinium group; and (4) acceleration
of the amidation reaction as compared to unwanted
competitive hydrolytic paths.
Following indications given by Staros44 and Anjaneyulu,45
optimum conditions were found using the carbodiimide DIC
and N-hydroxysulfosuccinimide (NHSS) activating couple in
organic−aqueous media at nearly neutral pH at low temper-
ature. In practice, as outlined in Scheme 4, DMF solutions of
carboxylic acids 10 or 11 were reacted with DIC and NHSS to
furnish the corresponding hydrophilic N-hydroxysulfosuccini-
midyl ester intermediates. Subsequent addition of the amine
component (13, 16, or 18) in a pH 7.0 phosphate buffered
water/acetonitrile (1:1) solvent mixture at 0 °C ensured
preparation of the respective amide conjugates 24−27 in 40−
62% yields.
From a structural viewpoint, amides 24 and 25 share a
common diglycolic ester link but differ from each other in their
targeting RGD unit, while compound 26 is the dimeric version
of its counterpart 25. Finally, compounds 24 and 27 exhibit a
1616
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
a
Table 1. Solid-Phase Receptor Binding Assay
IC50 (nM) SD
compound
αVβ3
αVβ5
19
266 75
>10−5
330 54
>10−5
20
21
220.0 13.2
4.3 1.6
805 309
187 65
22
23
2.3 2.0
82 16
24
35.6 9.7
8.9 0.9
344 196
1909 351
725.0 234.4
219 51
25
26
0.23 0.012
237 89
27
12
32.6 15.7
1077 624
6.0 1.5
98.3 51.4
11300 1480
1208 373
38 20
13
14
15
6.4 2.4
16
6.1 1.6
315 137
7.5 4.8
c(RGDfV)
3.2 1.3
II
53.7 17.3
2.4 1.0
205 33.5
38 11
III
vitronectin
47.1 10.0
13.7 6.2
a
IC50 values were calculated as the concentration of compound required for 50% inhibition of biotinylated vitronectin binding to human, isolated
αVβ3 and αVβ5 integrin receptors. High-affinity integrin binders c(RGDfV), II and III, were used as positive controls.
Overall, the presence of a common taxane unit and
chemically different linkers covalently attached to the
integrin-recognizing RGD units led to appreciable results,
with satisfactory to excellent αVβ3 binding capabilities (in the
266−0.23 nM range) which were not generally compromized
by the added cargo. It appears that interposition of a flexible
chain (e.g., ethylene) between the RGD moiety and the linker
is favorable for affinity (e.g., 22, 23, 25, and 26); also, the
presence of the amide-attached diglycolic unit seems to be
beneficial to the overall binding capability (e.g., 24, 25, and 26).
As a sole exception, compound 20 exhibited a completely
annihilated binding affinity, possibly due to the presence of a
long, flexible hydrocarbon linker chain, which would allow
folding of the cytotoxic unit onto the RGD portion, thereby
suppressing its integrin recognizing capability.
In Vitro Cell Sensitivity Assays. The ability of PTX
conjugates 19−27 to inhibit growth of tumor cells was assessed
using a growth inhibition assay (72 h exposure) in which the
effect of free PTX 1 was also examined. Cell sensitivity to the
different conjugates was evaluated using cell lines characterized
1617
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
by a variable expression of αvβ3 and αvβ5 integrins. In U2-OS
cells, the basal expression of αvβ5 was more marked than that of
all other cell systems, whereas the cell line displaying the
highest level of expression of αvβ3 was the cisplatin-resistant
IGROV-1/Pt1 subline (Table 2). For this reason, the IGROV-
22 and 23, displaying very good in vitro αVβ3 affinity, the cell
sensitivity of the different cell lines was variable, reaching in
some cases interesting results. Conjugates 24 and 25, bearing
diglycolic amide linkers, show nonflattened one-to-two-digit
nanomolar IC50 values for all cell lines; interestingly enough,
treatment with compound 24 inhibited proliferation of ovarian
carcinoma IGROV-1 cells at a lower IC50 as compared to free
PTX (1.04 nM vs 23.4 nM) and this could be due to an
additive cytotoxic effect deriving from both PTX and the RGD
unit itself. Compound 26, bearing two RGDs per PTX,
represents a dimeric version of 25 and displays superior in vitro
αvβ3-targeting capabilities. Accordingly, 26 shows a 2-fold
cytotoxic activity as compared to 25, supporting the hypothesis
of active directioning and αVβ3-mediated endocytosis. Uncon-
jugated integrin ligands 13 and 16 do not show any sign of
cytotoxicity in the present assay.
Table 2. Cytofluorimetric Analysis of αVβ3 and αVβ5 Levels
in Cell Lines of Different Tumor Types
a
relative mean fluorescence intensity
cell line
H460
αVβ3
αVβ5
1.1 0.5
2.43 0.3
2.70 0.6
13.27 4.9
3.36 1.1
27.3 1.4
4.33 1.7
2.8 0.3
U2-OS
IGROV-1
IGROV-1/Pt1
a
The value is the ratio between mean fluorescence intensity of cells
In Vivo Antitumor Activity. Due to the good growth
inhibition capability (slightly improved in comparison to PTX)
displayed in vitro by compound 21, it was selected to be further
examined for in vivo studies. When mice were xenografted s.c.
with IGROV-1/Pt1, a significant inhibition of tumor growth
was observed following 21 administration. The compound was
administered i.v. (36 mg/kg) every 4 days for 4 times. Under
these conditions, the TVI (tumor volume inhibition) was 98%
(Figure 1, Table 4, P < 0.05 vs PTX-treated mice). PTX used at
the same treatment conditions displayed a lower antitumor
effect. An improved antitumor activity of 21 was also indicated
by the LCK (log cell kill) value.
Histological analysis highlighted an exceedingly higher
number of mitoses in the group treated with compound 21,
compared to other groups (Figure 2). The majority of the
mitoses observed in the group treated with compound 21 and
in the group treated with PTX were aberrant, an observation
consistent with the mechanism of action of the drug.46−48
Examples are provided in Figure 3.
incubated with primary antibody and that of cells incubated with
isotypic control.
1/Pt1 cell model was employed for subsequent in vivo studies.
As detailed in Table 3, all conjugate compounds 19−27 showed
an antiproliferative effect which varies from low nanomolar to
micromolar IC50 values, indirectly supporting a possible
cleavage and release of the PTX drug from the conjugate in
the 72 h exposure time.39 Triazolyl-linked compound 20, which
does not possess appreciable binding affinity toward the αVβ3/
αVβ5 integrins, may be regarded as a nondirecting control; the
IC50 values of 20 are the highest in the series along all tested
tumor cell lines supporting the belief of a close correlation
between overall antiproliferative activity and αVβ3/αVβ5
recognition capability. Residual, poor activity could be ascribed
to aspecific delivery of the drug. Compound 19, a shortened
homologue of 20, is endowed with improved targeting ability,
and accordingly, the IC50 values appear appreciably improved,
though still far from action of the free drug. Derivative 21,
which presents a robust triazole ring connected to ethylene
glycol units by an amide function, shows flattened, low-
nanomolar IC50 values that are comparable, if not superior, to
those of free PTX; and this could be imputable either to
excellent active targeting and αVβ3-mediated internalization or,
simply, to premature and complete release of the cytotoxic
cargo during the 72 h exposure. For robust triazolyl compounds
DISCUSSION
■
In the present study, we report the synthesis and biological
evaluation of nine PTX conjugates characterized by structural
novelty due to Aba-based or Ampro-based RGD moieties which
are joined as C2′ PTX esters through variable triazolyl- or
amide-connecting moieties. A number of molecules for which
a
Table 3. Cell Sensitivity of Cells with Variable Expression of Integrins to cRGD Ligand Conjugates
IC50 (nM) SD
b
c
d
e
compound
U2-OS
IGROV-1
IGROV-1/Pt1
H460
19
20
21
22
23
24
25
26
27
PTX
13
16
107 27
244 20
2.0 1.1
nd
541 334
2870 822
1.6 0.9
140 53
219 63
1.6 1.1
11.7 6.2
73.3 6.7
10.1 4.7
3.4 2.4
1.6 0.5
5.8 0.5
2.2 0.8
58 27
236 79
9.5 2.2
nd
277 180.1
333.4 200
1.04 0.47
43.1 6.8
28.0 2.2
81.0 6.8
23.4 8.2
22 300
28.0 6.7
10.1 3.3
nd
nd
13.1 2.7
nd
nd
nd
6.36 0.18
3.4 0.5
6.9 1.4
5.4 1.3
>4000
>50 200
>30 000
>16 700
>4000
nd
>16 700
a
Cell sensitivity was evaluated by growth inhibition assays based on cell counting after 72 h exposure. IC50 is defined as the drug concentration
b
c
d
inhibiting cell growth by 50%. Human osteosarcoma. Human ovarian adenocarcinoma. Cisplatin resistant human ovarian adenocarcinoma.
e
Human large cell lung carcinoma.
1618
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
Figure 1. Antitumor activity studies of compound 21. Efficacy of i.v.
●
▲
21 ( ) and paclitaxel ( ) administered at 36 mg/kg every fourth day
four times on the ovarian carcinoma IGROV-1/Pt1 xenografted s.c. in
○
athymic nude mice. Solvent was injected for the control group ( ).
Each point represents the mean tumor volume from 9 to 11 tumors.
Bars represent SD.
Table 4. Efficacy of i.v. 21 and PTX (36 mg/kg q4d×4) on
the Ovarian Carcinoma IGROV-1/Pt1 Xenografted s.c. in
Athymic Nude Mice
a
b
c
d
compound
TVI%
LCK
BWL%
Tox
21
98
81
2
7
5
0/5
0/5
PTX
0.7
a
Tumor volume inhibition % in treated over control mice assessed 7
b
days after last treatment. Gross log10 cell kill to reach 700 mm3 of
tumor volume. Body weight loss % induced by treatment; the highest
change is reported. Dead/treated mice.
c
d
■
□
Figure 2. Comparison between normal ( ) and aberrant ( ) mitoses
in different groups of treatment. Mitoses were evaluated in 3 randomly
selected 400× fields using quadruplicate samples. The reported
numbers correspond to the total number of normal/aberrant mitoses
in analyzed group 1, control group; 2, group treated with ligand 13; 3,
group treated with compound 21; 4, group treated with PTX. Group
treated with compound 21 has the highest number of mitoses and in
the meanwhile the highest number of aberrant mitoses (p < 0.05).
Figure 3. Randomly selected fields of (A) G1 sample, characterized by
up to two “normal” mitoses (hematoxylin and eosin; Bar, 50 μm); (B)
sample from mice treated with compound 21. Note the central
aberrant mitosis, characterized by mis-segregated chromosomes within
a nuclear envelope (mitotic catastrophe), surrounded by a large
number of small condensed hyperchromatin nuclei, both examples of
“atypical” mitoses (hematoxylin and eosin; Bar, 50 μm); (C) G4
sample. In the center, there is an aberrant mitosis, characterized by
mis-segregated chromosomes surrounded by a nuclear envelope
(mitotic catastrophe) (hematoxylin and eosin; Bar, 50 μm).
conjugation did not negatively influence binding to integrins
was obtained. Such molecules were also endowed with marked
cytotoxic activity in the range of that observed for PTX. In fact,
cell sensitivity assays toward αVβ3/αVβ5-overexpressing human
tumor cell lines support that PTX preserves or improves its
cytotoxic activity when delivered through the covalent
conjugates 19−27, thus indirectly supporting its release as a
free drug.39 Whether it is prematurely released outside the cell
1619
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
and subsequently internalized through cell membrane diffusion
or it is actively targeted and released inside the cell following
integrin-mediated endocytosis, it is hard to definitely conclude
here. Observation of nonflattened and nonhomogeneous data
in the tested assay (with exception of compound 21), which are
closely related to the corresponding affinity toward the
receptors, favors the second hypothesis.
moieties which are joined as C2′ PTX esters through variable
triazolyl- or amide-connecting moieties. The main goals of the
work have been accomplished. The synthesis proved viable and
faithful, efficiently furnishing the nine conjugates in 38−74%
yields from the taxane and respective RGD starting modules.
Purification and complete structural characterization of the final
products guaranteed reliability during subsequent biological
tests. Preservation of high in vitro αVβ3-receptor affinity
observed for almost all conjugate candidates in the series as
compared to free RGD-counterparts demonstrated that the
cytotoxic cargo and/or the linker unit may comfortably coexist
with the RGD moiety without damaging the exquisite αVβ3-
selective targeting capability. In vitro growth inhibition assays
on a panel of αVβ3/αVβ5-overexpressing human tumor cell lines
showed remarkable cytotoxic activity. Finally, in vivo evaluation
of conjugate 21 confirmed the effectiveness of this compound
to inhibit tumor growth, displaying an enhanced activity in
comparison to that of free Paclitaxel. This result is further
strengthened by the fact that the effective quantity of Paclitaxel
administered to the animals treated with compound 21 is
almost half the given doses.
Though the details of the release process remain to be
defined, it would seem likely that the PTX-AbaRGD and PTX-
AmproRGD junction is a well-done marriage where the
partners, more than simply tolerant, cooperate with each other.
Last, during in vivo assays performed on compound 21,
outstanding control of tumor growth (TVI 98% for 21 vs 81%
for PTX at 25 day) in a model overexpressing αVβ3 was
achieved at well-tolerated doses of the conjugate (36 mg/kg).
This result is even more impressive if we consider that the
effective amount of PTX administered to the animals treated
with compound 21 is 17 mg/kg (the MW being almost
double). Since tumors from mice treated with compound 21
had the highest number of mitoses and the major part of them
were atypical, it can be speculated that neoplastic cells treated
with compound 21 enter mitosis but fail to replicate and incur
in mitotic arrest, as already described in the current literature
for “spindle poisons” such as PTX. As mentioned in the
Introduction, a few studies exist where cyclopeptide RGD
moietiesnamely, dimeric E-[c(RGDfK)2] or E-[c-
(RGDyK)2]were covalently conjugated to PTX in its 2′-
OH position through a succinate spacer.27−29 In a first
contribution, Chen and Neamati28 tested their PTX-RGD
construct against metastatic breast cancer cells (MDA-MB-438)
showing a cell proliferation inhibition comparable to that
observed for paclitaxel and a slightly decreased integrin binding
affinity with respect to the unconjugated counterpart. In vivo
and ex vivo data of the same conjugate were reported shortly
afterward29 and its antitumor therapeutic effect compared to
simple PTX/RGD combinations. In this case, although the
tritiated conjugate showed higher tumor uptake and longer
retention of PTX in the MDA-MB-438 tumor than PTX alone,
the absolute tumor uptake value was still rather low. The E-
[c(RGDfK)2]-PTX conjugate was also evaluated by Satchi-
Fainaro and Kratz et al.27 demonstrating that in a standard 72 h
assay it inhibited the proliferation of HUVECs in a similar
manner as free PTX (IC50 0.4 nM) suggesting complete
hydrolysis of the succinate ester under the tested conditions.
Furthermore, in vivo studies in a OVCAR-3 xenograft model
demonstrated no antitumor efficacy.
Favorable indications of the present work incite further
investigations (e.g., evaluation of metabolic stability, toxicity,
and actual drug release localization) to be performed on
selected candidates in order to gain insight into optimal
structural compromise between selective targeting ability, in-
cell drug release capability, and overall solubility/permeability
in human tissues.
ASSOCIATED CONTENT
* Supporting Information
Characterization data, HPLC traces, and copies of H NMR
■
S
1
spectra of RGD conjugates. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
■
*Leonardo Manzoni: Tel. (+39) 02 50314064, Fax (+39) 02
50314072, e-mail: leonardo.manzoni@istm.cnr.it. Franca Za-
nardi: Tel. (+39) 0521 905067, Fax (+39) 0521 905006, e-mail:
franca.zanardi@unipr.it.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by Associazione Italiana per la Ricerca
sul Cancro (Milano, Italy), Ministero dell’Istruzione, dell’Uni-
■
Here, the connection of robust semipeptidic and highly αVβ3-
selective RGD ligands to PTX through chemically diverse
tethers gave rise to constitutionally and geometrically diverse
conjugate products whose varied physical and chemical
properties were anticipated to significantly impact the biological
response. Indeed, the diversified, yet interesting in vitro
cytotoxic activity as well as the outstanding in vivo antitumor
response obtained in this work seem unique in the field; and
this strengthens the notion that the Aba- and Ampro-based
RGD semipeptides may be conveniently added in the list of the
integrin-targeting moieties for the construction of effective and
selective antitumor and anti-angiogenesis agents.
versita
Banco di Sardegna. Thanks are due to the Centro
Interdipartimentale Misure “G. Casnati” (Universita degli
̀
e della Ricerca (MIUR, PRIN 2008), and Fondazione
̀
Studi di Parma) for instrumental facilities, and to “Regione
Autonoma della Sardegna” for research fellowships to P.B. and
P.C. (PO Sardegna FSE 2007-2013, L.R. 7/2007 “Promozione
della ricerca scientifica e dell’innovazione tecnologica in
Sardegna”).
REFERENCES
■
(1) Cossa, G., Gatti, L., Zunino, F., and Perego, P. (2009) Strategies
to improve the efficacy of platinum compounds. Curr. Med. Chem. 16,
2355−2365.
CONCLUSION
■
The study focused on the synthesis and biological evaluation of
nine PTX-conjugates compounds 19−27, whose structural
novelty is represented by Aba-based or Ampro-based RGD
(2) Desgrosellier, J. S., and Cheresh, D. A. (2010) Integrins in cancer:
biological implications and therapeutic opportunities. Nat. Rev. Cancer
10, 9−22.
1620
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
(3) Harms, J. F., Welch, D. R., Samant, R. S., Shevde, L. A., Miele, M.
E., Babu, G. R., Goldberg, S. F., Gilman, V. R., Sosnowski, D. M.,
Campo, D. A., Gay, C. V., Budgeon, L. R., Mercer, R., Jewell, J.,
Mastro, A. M., Donahue, H. J., Erin, N., Debies, M. T., Meehan, W. J.,
Jones, A. L., Mbalaviele, G., Nickols, A., Christensen, N. D., Melly, R.,
Beck, L. N., Kent, J., Rader, R. K., Kotyk, J. J., Pagel, M. D., Westlin, W.
F., and Griggs, D. W. (2004) A small molecule antagonist of the
alpha(v)beta3 integrin suppresses MDA-MB-435 skeletal metastasis.
Clin. Exp. Metastasis 21, 119−128.
(4) Vellon, L., Menendez, J. A., and Lupu, R. (2006) A bidirectional
″alpha(v)beta(3) integrin-ERK1/ERK2MAPK″ connection regulates
the proliferation of breast cancer cells. Mol. Carcinog. 45, 795−804.
(5) Gingras, M. C., Roussel, E., Bruner, J. M., Branch, C. D., and
Moser, R. P. (1995) Comparison of cell adhesion molecule expression
between glioblastoma multiforme and autologous normal brain tissue.
J. Neuroimmunol. 57, 143−153.
(6) Albelda, S. M., Mette, S. A., Elder, D. E., Stewart, R.,
Damjanovich, L., Herlyn, M., and Buck, C. A. (1990) Integrin
distribution in malignant melanoma: association of the beta3 subunit
with tumor progression. Cancer Res. 50, 6757−6764.
(7) Dang, D., Bamburg, J. R., and Ramos, D. M. (2006) Alphavbeta3
integrin and cofilin modulate K1735 melanoma cell invasion. Exp. Cell
Res. 312, 468−477.
(8) Cooper, C. R., Chay, C. H., and Pienta, K. J. (2002) The role of
alpha(v)beta(3) in prostate cancer progression. Neoplasia 4, 191−194.
(9) Hapke, S., Kessler, H., Luber, B., Benge, A., Hutzler, P., Hofler,
H., Schmitt, M., and Reuning, U. (2003) Ovarian cancer cell
proliferation and motility is induced by engagement of integrin
alpha(v)beta3/vitronectin interaction. Biol. Chem. 384, 1073−1083.
(10) Auzzas, L., Zanardi, F., Battistini, L., Burreddu, P., Carta, P.,
Rassu, G., Curti, C., and Casiraghi, G. (2010) Targeting αVβ3 integrin:
design and applications of mono- and multifunctional RGD-based
peptides and semipeptides. Curr. Med. Chem. 17, 1255−1299.
(11) Paolillo, M., Russo, M. A., Serra, M., Colombo, L., and Schinelli,
S. (2009) Small molecule integrin antagonists in cancer therapy. Mini-
Rev. Med. Chem. 9, 1439−1446.
(12) Dechantsreiter, M. A., Planker, E., Matha, B., Lohof, E.,
Holzemann, G., Jonczyk, A., Goodmann, S. L., and Kessler, H. (1999)
N-Methylated cyclic RGD peptides as highly active and selective αVβ3
integrin antagonist. J. Med. Chem. 42, 3033−3040.
(13) Manzoni, L., Belvisi, L., Arosio, D., Civera, M., Pilkington-Miksa,
M., Potenza, D., Caprini, A., Araldi, E. M. V., Monferini, E., Mancino,
(19) Schottelius, M., Laufer, B., Kessler, H., and Wester, H.-J. (2009)
Ligands for mapping αVβ3-integrin expression in vivo. Acc. Chem. Res.
42, 969−980.
(20) Garanger, E., Boturyn, D., and Dumy, P. (2007) Tumor
targeting with RGD peptide ligands − Design of new molecular
conjugates for imaging and therapy of cancers. Anti Cancer Agents Med.
Chem. 7, 552−558.
(21) Temming, K., Schiffelers, R. M., Molema, G., and Kok, R. J.
(2005) RGD-Based strategies for selective delivery of therapeutics and
imaging agents to the tumor vasculature. Drug Resistance Updates 8,
381−402.
(22) Arap, W., Pasqualini, R., and Ruoslahti, E. (1998) Cancer
treatment by targeted drug delivery to tumor vasculature in a mouse
model. Science 279, 377−380.
(23) Liu, Y., Bajjuri, K. M., Liu, C., and Sinha, S. C. (2012) Targeting
cell surface alpha(v)beta(3,5) integrins increases therapeutic efficacies
of a legumain protease-activated auristatin prodrug. Mol. Pharm. 9,
168−175.
(24) Dufort, S., Sancey, L., Hurbin, A., Foillard, S., Boturyn, D.,
Dumy, P., and Coll, J.-L. (2011) Targeted delivery of a proapoptotic
peptide to tumors in vivo. J. Drug Targeting 19, 582−588.
(25) Dal Pozzo, A., Esposito, E., Ni, M., Muzi, L., Pisano, C., Bucci,
F., Vesci, L., Castorina, M., and Penco, S. (2010) Conjugates of a novel
7-substituted camptothecin with RGD-peptides as αVβ3 integrin
ligands: an approach to tumor-targeted therapy. Bioconjugate Chem.
21, 1956−1967.
(26) Braun, K., Wiessler, M., Pipkorn, R., Ehemann, V., Bauerle, T.,
̈
Fleischhacker, H., Muller, G., Lorenz, P., and Waldeck, W. (2010) A
̈
cyclic-RGD-BioShuttle functionalized with TMZ by DARinv “Click
Chemistry” targeted to αVβ3 integrin for therapy. Int. J. Med. Sci. 7,
326−339.
(27) Ryppa, C., Mann-Steinberg, H., Biniossek, M. L., Satchi-Fainaro,
R., and Kratz, F. (2009) In vitro and in vivo evaluation of a paclitaxel
conjugate with the divalent peptide E-[c(RGDfK)2] that targets
integrin αVβ3. Int. J. Pharm. 368, 89−97.
(28) Chen, X., Plasencia, C., Hou, Y., and Neamati, N. (2005)
Synthesis and biological evaluation of dimeric RGD peptide-paclitaxel
conjugate as a model for integrin-targeted drug delivery. J. Med. Chem.
48, 1098−1106. (corrigendum (2005) J. Med. Chem. 48, 5874).
(29) Cao, Q., Li, Z.-B., Chen, K., Wu, Z., He, L., Neamati, N., and
Chen, X. (2008) Evaluation of biodistribution and anti-tumor effect of
a dimeric RGD peptide-paclitaxel conjugate in mice with breast cancer.
Eur. J. Nucl. Med. Mol. Imaging 35, 1489−1498.
(30) Arosio, D., Casagrande, C., and Manzoni, L. (2012) Integrin-
mediated drug delivery in cancer and cardiovascular diseases with
peptide-functionalized nanoparticles. Curr. Med. Chem. 19, 3128−
3151.
(31) Zhan, C., Meng, Q., Li, Q., Feng, L., Zhu, J., and Lu, W. (2012)
Cyclic RGD−Polyethylene Glycol−Polyethylenimine for Intracranial
Glioblastoma-Targeted Gene Delivery. Chem. Asian J. 7, 91−96.
(32) Yao, H.-J., Ju, R.-J., Wang, X.-X., Zhang, Y., Li, R.-J., Yu, Y.,
Zhang, L., and Lu, W.-L. (2011) The antitumor efficacy of functional
paclitaxel nanomicelles in treating resistant breast cancers by oral
delivery. Biomaterials 32, 3285−3302.
̀
M., Podesta, F., and Scolastico, C. (2009) Cyclic RGD-containing
functionalized azabicycloalkane peptides as potent integrin antagonists
for tumor targeting. ChemMedChem 4, 615−632.
(14) Battistini, L., Burreddu, P., Carta, P., Rassu, G., Auzzas, L., Curti,
C., Zanardi, F., Manzoni, L., Araldi, E. M. V., Scolastico, C., and
Casiraghi, G. (2009) 4-Aminoproline-based arginine-glycine-aspartate
integrin binders with exposed ligation points: practical in-solution
synthesis, conjugation and binding affinity evaluation. Org. Biomol.
Chem. 7, 4924−4935.
(15) Liu, Z., Wang, F., and Chen, X. (2008) Integrin αVβ3-targeted
cancer therapy. Drug Dev. Res. 69, 329−339.
(16) Stupp, R., and Ruegg, C. (2007) Integrin inhibitors reaching the
clinic. J. Clin. Oncol. 25, 1637−1638.
(17) Reardon, D. A., Fink, K. L., Mikkelsen, T., Cloughesy, T. F.,
O’Neill, A., Plotkin, S., Glantz, M., Ravin, P., Raizer, J. J., Rich, K. M.,
Schiff, D., Shapiro, W. R., Burdette-Radoux, S., Dropcho, E. J.,
Wittemer, S. M., Nippgen, J., Picard, M., and Nabors, B. L. (2008)
Randomized phase II study of cilengitide, an integrin-targeting
arginine-glycine-aspartic acid peptide, in recurrent glioblastoma
multiforme. J. Clin. Oncol. 26, 5610−5617.
(18) MacDonald, T. J., Stewart, C. F., Kocak, M., Goldman, S.,
Ellenbogen, R. G., Phillips, P., Lafond, D., Young Poussaint, T., Kieran,
M. W., Boyett, J. M., and Kun, L. E. (2008) Phase I clinical trial of
Cilengitide in children with refractory brain tumors: pedriatric brain
tumor consortium study PBTC-012. J. Clin. Oncol. 26, 919−924.
(33) Eldar-Boock, A., Miller, K., Sanchis, J., Lupu, R., Vicent, M. J.,
and Satchi-Fainaro, R. (2011) Integrin-assisted drug delivery of nano-
scaled polymer therapeutics bearing paclitaxel. Biomaterials 32, 3862−
3874.
(34) Austin, L. A., Kang, B., Yen, C.-W., and El-Sayed, M. A. (2011)
Nuclear targeted silver nanospheres perturb the cancer cell cycle
differently than those of nanogold. Bioconjugate Chem. 22, 2324−2331.
(35) Ubezio, P., Garofalo, A., Borsotti, P., Cassis, L., Riccardi, E.,
Scanziani, E., Eccles, S. A., Bani, M. R., and Giavazzi, R. (2006) The
vascular targeting property of paclitaxel is enhanced by SU6668, a
receptor tyrosine kinase inhibitor, causing apoptosis of endothelial
cells and inhibition of angiogenesis. Clin. Cancer Res. 12, 1839−1849.
(36) Dijkgraaf, I., Rijnders, A. Y., Soede, A., Dechesne, A. C., van
Esse, G. W., Brouwer, A. J., Corstens, F. H. M., Boerman, O. C.,
Rijkers, D. T. S., and Liskamp, R. M. (2007) Synthesis of DOTA-
1621
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622

Bioconjugate Chemistry
Article
conjugated multivalent cyclic-RGD peptide dendrimers via 1,3-dipolar
cycloaddition and their biological evaluation: implications for tumor
targeting and tumor imaging purposes. Org. Biomol. Chem. 5, 935−944.
(37) Benard, J., Da Silva, J., De Blois, M. C., Boyer, P., Duvillard, P.,
Chiric, E., and Riou, G. (1985) Characterization of a human ovarian
adenocarcinoma line, IGROV1, in tissue culture and in nude mice.
Cancer Res. 45, 4970−4979.
(38) Perego, P., Giarola, M., Righetti, S. C., Supino, R., Caserini, C.,
Delia, D., Pierotti, M. A., Miyashita, T., Reed, J. C., and Zunino, F.
(1996) Association between cisplatin resistance and mutation of p53
gene and reduced bax expression in ovarian carcinoma cell systems.
Cancer Res. 56, 556−562.
(39) Kingston, D. G. (2000) Recent advances in the chemistry of
taxol. J. Nat. Prod. 63, 726−734.
(40) Himo, F., Lovell, T., Hilgraf, R., Rostovtsev, V. V., Noodleman,
L., Sharpless, K. B., and Fokin, V. V. (2005) Copper(I)-catalyzed
synthesis of azoles. DFT study predicts unprecedented reactivity and
intermediates. J. Am. Chem. Soc. 127, 210−216.
(41) Wester, H. J., and Kessler, H. (2005) Molecular targeting with
peptides or peptide-polymer conjugates: just a question of size? J. Nucl.
Med. 46, 1940−1945.
(42) Sancey, L., Garanger, E., Foillard, S., Schoehn, G., Hurbin, A.,
Albiges-Rizo, C., Boturyn, D., Souchier, C., Grichine, A., Dumy, P., and
Coll, J.-L. (2009) Clustering and internalization of integrin αVβ3 with a
tetrameric RGD-synthetic peptide. Mol. Ther. 17, 837−843.
(43) Liu, S., Hsieh, W.-Y., Jiang, Y., Kim, Y.-S., Sreerama, S. G., Chen,
X., Jia, B., and Wang, F. (2007) Evaluation of a 99mTc-labeled cyclic
RGD tetramer for non invasive imaging integrin αVβ3-positive breast
cancer. Bioconjugate Chem. 18, 438−446.
(44) Staros, J. V. (1982) N-Hydroxysulfosuccinimide active esters:
bis(N-hydroxysulfosuccinimide) esters of two dicarboxylic acids are
hydrophilic, membrane-impermeant, protein cross-linkers. Biochemistry
21, 3950−3955.
(45) Anjaneyulu, P. S. R., and Staros, J. V. (1987) Reactions of N-
hydroxysulfosuccinimide active esters. Int. J. Pept. Protein Res. 30, 117−
124.
(46) Matson, D. R., and Stukenberg, P. T. (2011) Spindle poisons
and cell fate: a tale of two pathways. Mol. Interv. 11, 141−150.
(47) Portugal, J., Mandilla, S., and Bataller, M. (2010) Mechanisms of
drug-induced mitotic catastrophe in cancer cells. Curr. Pharm. Des. 16,
69−78.
(48) Roninson, I. B., Broude, E. V., and Chang, B. D. (2001) If not
apoptosis, then what? Treatment-induced senescence and mitotic
catastrophe in tumor cells. Drug Res. Updates 4, 303−313.
1622
dx.doi.org/10.1021/bc300164t | Bioconjugate Chem. 2012, 23, 1610−1622