Conjugates of a novel 7-substituted camptothecin with RGD-peptides as αvβ3 integrin ligands: An approach to tumor-targeted therapy

Article abstract of DOI:10.1021/bc100097r

Eight conjugates of a novel camptothecin derivative (Namitecan, NMT) with RGD peptides have been synthesized and biologically evaluated. This study focused on factors that optimize the drug linkage to the transport vector. The different linkages investigated consist of heterofunctional glycol fragments and a lysosomally cleavable peptide. The linkage length and conformation were systematically modified with the purpose to understand their effect on receptor affinity, systemic stability, cytotoxicity, and solubility of the corresponding conjugates. Among the new conjugates prepared, C6 and C7 showed high receptor affinity and tumor cell adhesion, acceptable stability in murine blood, and high cytotoxic activity (IC₅₀ = 8 nM). The rationale, synthetic strategy, and preliminary biological results will be presented.

Full text of DOI:10.1021/bc100097r

1956
Bioconjugate Chem. 2010, 21, 1956–1967
Conjugates of a Novel 7-Substituted Camptothecin with RGD-Peptides as
rvꢀ3 Integrin Ligands: An Approach to Tumor-Targeted Therapy
Alma Dal Pozzo,*,† Emiliano Esposito,† Minghong Ni,† Laura Muzi,† Claudio Pisano,‡ Federica Bucci,‡
Loredana Vesci,‡ Massimo Castorina,‡ and Sergio Penco‡
Istituto di Ricerche Chimiche e Biochimiche G. Ronzoni, via G. Colombo, 81, 20133, Milano, Italy, and Area of Oncology
Research and Development, Sigma-Tau, Pomezia, Roma, Italy. Received February 24, 2010; Revised Manuscript Received
September 30, 2010
Eight conjugates of a novel camptothecin derivative (Namitecan, NMT) with RGD peptides have been synthesized
and biologically evaluated. This study focused on factors that optimize the drug linkage to the transport vector.
The different linkages investigated consist of heterofunctional glycol fragments and a lysosomally cleavable peptide.
The linkage length and conformation were systematically modified with the purpose to understand their effect on
receptor affinity, systemic stability, cytotoxicity, and solubility of the corresponding conjugates. Among the new
conjugates prepared, C6 and C7 showed high receptor affinity and tumor cell adhesion, acceptable stability in
murine blood, and high cytotoxic activity (IC50 ) 8 nM). The rationale, synthetic strategy, and preliminary biological
results will be presented.
surface receptor. This integrin has a predominant role in tumor-
induced angiogenesis and is overexpressed on the activated
endothelial cells in several tumor forms, but expressed at a low
level on normal and mature endothelial cells. Inhibition of
angiogenesis has been shown to prevent tumor growth and even
cause tumor regression in various experimental models (9).
However, antiangiogenic therapy usually is not sufficient to
eradicate tumors in the late stage. Thus, combination of
antiangiogenic agents with either chemo or radiation therapy
has demonstrated enhanced activity in multiple tumor systems
(10). Ruoslathi et al. first raised the idea of conjugating a
cytotoxic agent with RGD-containing peptides (11-13). Later
on, many RGD peptides have been evaluated for their potential
as antiangiogenic agents or integrin-targeted radiotracers and/
or therapeutics (14-19).
We were interested in studying the effect of RGD cyclopep-
tides in conjugation with a new potent antitumoral drug
belonging to the camptothecin family, Namitecan, NMT (20).
A number of camptothecin prodrugs and delivery systems have
been developed to improve the pharmacokinetic profile of this
anticancer drug (21-23). We have previously reported on the
synthesis and biological evaluation of a number of RGD-NMT
conjugates (24), where the targeting RGD peptide was attached
to the drug by either amide or hydrazone bond; in theory, both
linkages should be potentially stable at physiological pH,
whereas the main difference between them lies in the acid
stability. During that study, we demonstrated that the amide
bond-containing conjugates maintained good affinity to integrin
receptors and accumulated in tumor cells after internalization,
but showed no appreciable activity both in vitro and in vivo,
indicating that the amide bond is too stable to promptly release
the drug into the tumor cells. On the contrary, the hydrazone
bond-containing conjugates exhibited high in vitro cytotoxicity,
but their stability at pH 7.4 was much lower than expected; as
a consequence, their activity has been mostly attributed to the
NMT prematurely liberated within the cell culture medium.
Thus, it was concluded that both types of linkage are not suitable
for these low-molecular-weight conjugates. Moreover, all
conjugates prepared were not soluble enough for a complete in
vivo study, likely because the proximity of ligand and drug
INTRODUCTION
The antitumor efficacy of clinically used anticancer drugs is
limited by their nonspecific toxicity to normal cells, resulting
in a low therapeutic index. Therefore, great efforts are being
devoted to the aim of targeting therapeutics preferentially to
tumors, sparing physiological tissues and decreasing side effects.
In late years, several different approaches to this problem have
been studied, which mainly focused on conjugating cytotoxic
agents with macromolecules such as monoclonal antibodies
(mAbs) or polymeric materials (1-5). First successes have been
achieved with mAbs immunoconjugates, and some of them are
currently in clinical trials. However, there are factors restricting
the scope of this approach: the antigen-dependent drug delivery
to target tumor cell population can be limited by the cell surface
density of expressed antigens and by saturation of the receptors
with immunoconjugates that are endocytosed at a given rate.
The new generation immunoconjugates have incorporated drugs
that are considerably more potent than the standard anticancer
agents. Among them, the anti-CD33-calicheamicin (Mylotarg)
has been approved by FDA for treatment of acute myeloid
leukemia (6).
During the past two decades, the understanding of many
receptors overexpressed by tumor cells during proliferation has
given impetus to the development of low-molecular-weight
selective ligands. Therefore, following a trend toward replacing
the antibody scaffolds with small molecules, several peptides
have been proposed as tumor-targeting tools, to be used as
receptor imaging or as therapeutics when conjugated with
antitumor agents; for recent reviews on this topic, see ref 7, 8.
Small receptor-binding peptides offer the advantage of being
easily synthesized, modified, and characterized in order to
optimize their affinity to specific receptors.
Cyclic peptides containing the RGD (Arg-Gly-Asp) sequence
are well-known as specific ligands of the Rvꢀ3 integrin cell
* To whom correspondence should be addressed. Dr. Alma Dal
Pozzo, Ist. Di Ricerche Chimiche e Biochimiche G. Ronzoni, via
Colombo 81, 20133 Milano, Italy. Tel.: +39-02-70600223; Fax: +39-
02-70641634, e-mail: dalpozzo@ronzoni.it.
† Istituto di Ricerche Chimiche e Biochimiche G. Ronzoni.
‡ Area of Oncology Research and Development.
10.1021/bc100097r  2010 American Chemical Society
Published on Web 10/15/2010

RGD Peptide-Camptothecin Conjugates
Bioconjugate Chem., Vol. 21, No. 11, 2010 1957
Chart 1. Structure of NMT, Cyclopeptides P1 and P2, and Conjugates C1-C7

1958 Bioconjugate Chem., Vol. 21, No. 11, 2010
Dal Pozzo et al.
Chart 2. Structure of Fragments F1-F6, Containing Cyclopeptides
forms efficient intramolecular stacking interactions. Inspired by
these results, we sought to verify another interesting approach,
proposed in the literature and based on the introduction of
selective protease-sensitive peptides into tumor-targeted con-
jugate spacers. These peptides have been claimed to be specific
substrates of enzymes overexpressed within the lysosomal
compartment of tumor cells, but rather stable to proteases in
the systemic circulation (25-28).
Because of our interest in the potential of this approach, we
have undertaken the development of a series of tumor-targeted
conjugates containing an Rvꢀ3 integrin recognition moiety
connected to NMT by different molecular bridges as spacers.
The structure of these spacers include a peptide, that is
enzymatically cleavable, but stable enough in the systemic
circulation, and a number of glycol residues, whose length and
conformation were systematically modified, in order to enhance
solubility, without disturbing the binding of the RGD ligand to
the receptor.The objective of the work described in this article
was to study the effect of the spacer modifications on receptor
affinity, stability in murine blood, cytotoxicity, and solubility
of the new conjugates.
Among a series of RGD cyclopeptide analogues of Cilengitide
(29), designed and synthesized in our laboratories for the
purpose of this study, we identified P1 and P2 to be used as
targeting ligands, owing to their optimal stability and affinity
to Rv integrins (30). Peptide ligands, camptothecin derivative
NMT, and conjugate structures are illustrated in Chart 1. The
structure of the fragments used for assembling the conjugates
is illustrated in Charts 2 and 3. Each pair of fragments, which
were coupled together to obtain the desired conjugate, is
described in Table 1.
EXPERIMENTAL PROCEDURES
Materials and Instruments. All reagents and solvents were
reagent grade and were distilled prior to use. All natural amino

RGD Peptide-Camptothecin Conjugates
Bioconjugate Chem., Vol. 21, No. 11, 2010 1959
Chart 3. Structure of Fragments F7-F11, Containing the Cytotoxic Drug NMT
acids and resin were purchased from Bachem (Switzerland).
Fmoc-4-aminomethyl-phenylalanine (Amp) was purchased from
PepTech Corporation, Burlington, MA, USA; the corresponding
building blocks, Fmoc-Amp-glycol-R (where R can be amino,
hydrazido, or azido group), were easily synthesized by standard
methods. PEG derivatives were purchased from Iris Biotech
GMBH (Germany). Namitecan (ST1968) was synthesized in
Sigma-Tau laboratories (20). Microwave-assisted reactions were
performed on a CEM MW instrument. Compounds were purified
and characterized by a RP-HPLC 600 Waters instrument,
equipped with a semipreparative column Alltima, Alltech, RP18,
10 µ, 22 × 250 mm, or an analytical column Gemini,
Phenomenex, C18, 5 µ, 4.6 × 250 mm, at λ ) 220 nm, using
acetonitrile/water with 0.1% trifluoroacetic acid as mobile phase.
1H NMR spectra were recorded on an AC300 Bruker instrument.
Mass spectra were recorded on Bruker Autoflex Maldi-Tof or
Micro-Tof Q instruments. Flash chromatography was carried
out on silica gel Merck 230-400 mesh.
after repeated precipitations with cool diethyl ether followed
by lyophilization, with 97% purity.
Synthesis of Conjugate C2. Coupling between fragments
F1 and F8 was performed following the procedure described
for conjugate C1. After purification by preparative HPLC and
lyophilization, the compound was obtained with 97.6% purity.
Synthesis of Conjugates C3-C7. (Method B) General
procedure. Two fragments, containing an alkyne and an alky-
lazido group, respectively, were assembled by Huisgen 1,3-
dipolar cycloaddition. Equimolar amounts of the two desired
fragments were dissolved in DMF. Aqueous solutions of sodium
ascorbate (1 equiv) and CuSO4 ·5H2O (0.1 equiv) were added,
pH was adjusted to 6 by addition of NaOH, and the suspension
was stirred at room temperature overnight. When necessary to
speed up the reaction, the mixture was submitted to microwave
irradiation, 90 W for 2 min (Tmax obs 120 °C) once or twice
until completion of the reaction, as monitored by HPLC. Final
conjugates were obtained with >98% purity, after preparative
HPLC and lyophilization.
Synthesis of Conjugate C1. (Method A) As described in
Scheme 1, fragment F1 (350 mg, 0.319 mmol) and HOAT (259
mg, 1.91 mmol) were dissolved with 7 mL DMF; then, t-butyl
nitrite (45 µL, 0.383 mmol) was added and the mixture left
under stirring at room temperature. After 30 min, 1 equiv of
DIPEA was added, and the reaction mixture was poured into a
solution containing fragment F7 (337 mg, 0.319 mmol) in 2
mL DMF; pH was adjusted to 7.5 with DIPEA and the mixture
left under stirring at room temperature, protected from light.
After 2 h, an additional equivalent of the activated ester,
prepared in situ as described above, was added, pH adjusted at
7.5, and left overnight. The solvent was removed at reduced
pressure, and the residue was purified by preparative HPLC
(40% CH3CN in H2O + 0.1% TFA). To remove the N-Alloc
group, the compound was dissolved in 1 mL DMF, under argon,
together with 2.5 mg of (Ph3P)4Pd, AcOH (12 µL), and Bu3SnH
(25 µL), and stirred for 1 h. The final conjugate C1 was obtained
MW-Assisted Solid-Phase Synthesis of Cyclopeptides.
General Procedure. Solid-phase synthesis followed the standard
Fmoc-protocol, starting from Fmoc-Gly-SASRIN. The amino
acids (2 equiv) were added in the following order: Fmoc-
Arg(Pmc)-OH, Fmoc-Amp(building block)-OH, Fmoc-D-Phe-
OH (or Fmoc-D-Tyr(tBu)-OH), and Fmoc-Asp(OtBu)-OH. At
each step, Fmoc deprotection was carried out with 20%
piperidine. Energy applied: 30 W for 5 min (tmax obs. 70 °C)
for couplings and 25 W for 3 min (tmax obs. 40 °C) for
deprotections. For cleavage of the peptides, the resin was treated
with 1% TFA in DCM during 15 min for 5 times and the filtrates
immediately neutralized with pyridine. The collected filtrates
were taken to dryness under vacuum, and the linear peptides
were dissolved with acetonitrile containing 1% DIPEA to obtain
a 1.5 × 10-3 M solution, then added with 3 equiv TBTU/HOBT
and stirred for 1 h, until complete cyclization, as monitored by
HPLC. After evaporation of the solvent, the cyclopeptides were

1960 Bioconjugate Chem., Vol. 21, No. 11, 2010
Dal Pozzo et al.
Scheme 1. Synthesis of the Conjugate C1
Scheme 2. Synthesis of Fragment F5
dissolved with DCM and washed (water, 0.1 N HCl, water).
The organic phase, after drying with Na2SO4, was concentrated
and the residue purified by flash chromatography (mobile phase,
95:5 f 85:15 DCM/MeOH). Removal of the protecting groups
was performed with 95:5 TFA/H2O within 1 h. After concentra-
tion to small volume, the products were further purified by
repeated precipitations with dry and cold diethyl ether. Total
yields were around 50%, calculated on the amount of starting
resin.The cyclopeptides prepared with this method differ from
each other for the presence of D-Phe (P1) or D-Tyr (P2) at
position 4 and for the Amp-building blocks at position 5.
Synthesis of Fragment F5. As described in the Scheme 2,
DCC (84 mg, 0.405 mmol) was added to a cold solution of
azidoacetic acid 1 (41 mg, 0.405 mmol), glutamic acid di-tert-
butyl ester. HCl 2 (100 mg, 0.338 mmol), HOAT (0.405 mmol),
and DIPEA (127 mL, 0.744 mmol) in 5 mL DCM. The reaction
was stirred at room temperature for 2.5 h and after filtration
was diluted up to 30 mL and washed with water, 0.1 N HCl,
5% NaHCO3, and water. After removing the solvent, the residue
was treated with 3 mL of TFA for 1 h to afford 2-(2-
azidoacetylamino)-pentanedioic acid 3, which was dissolved
with 45 mL of 8:1 DCM/DMF and coupled with tert-butyl 12-
amino-4,7,10-trioxadodecanoate (281 mg, 1.01 mmol), HOAT
(137 mg, 1.014 mmol), DIPEA (174 mL), and DCC (209 mg,
1.014 mmol) by the standard method to give a crude product
that was purified by flash chromatography (95:5 DCM/MeOH)
to obtain 175 mg of a solid white powder. Yield 68% (calc. for
3 steps). After deprotection with TFA, the bis-carboxylic
intermediate 4 (0.234 mmol) was dissolved with 5 mL of DMF
and treated with N-hydroxysuccinimide (108 mg, 0.936 mmol)
and DCC (145 mg, 0.700 mmol) by stirring overnight. Solvent
was removed and the succinimide 5, dissolved with 8 mL DCM,
was added to a solution of the partially protected cyclopeptide
c[Arg(Pmc)-Gly-Asp(OtBu)-D-Tyr(tBu)-Amp], 728 mg, 0.66
mmol and DIPEA (113 µL, 0.66 mmol) in 14 mL of DMF.
After 1.5 h, solvent was removed and the residue purified by
preparative HPLC (69% CH3CN). Yield 48%. Total deprotection
was performed with 1:1 TFA/DCM containing 220 equiv of

RGD Peptide-Camptothecin Conjugates
Bioconjugate Chem., Vol. 21, No. 11, 2010 1961
Scheme 3. Synthesis of Fragment F6
thioanisole. The pure fragment F5 was obtained as a white solid
after evaporation of the solvents and subsequent precipitations
from cold diethyl ether. Yield 79%.
Conjugates Binding to rvꢀ3 Integrin, Tumor Cell Adhe-
sion, and in Vitro Cytotoxicity. These assays were performed
following the experimental procedures already published (24),
and the results are reported in the Tables 5 and 7.
Synthesis of Fragment F6. As reported in Scheme 3, the
totally protected cyclopeptide c{Arg(Pmc)-Gly-Asp(OtBu)-D-
Tyr(OtBu)-Amp[CO-(CH2)2-(O-CH2-CH2)-O-(CH2)2-N3]} 6 (572
mg, 0.448 mmol) and 1,3-bis-(prop-2-ynylcarbamoyl-propyl)-
carbamic acid benzyl ester 7 (72.5 mg, 0.206 mmol) were
dissolved with 12 mL of 7:5 DMF/H2O. Then, 2.5 M sodium
ascorbate (90 µL) and 0.5 M CuSO4 ·5H2O (45 µL) were added,
and the mixture submitted to microwave irradiation (90 W, Tmax
obs. 121 °C). The same operation was repeated three times until
complete disappearance of the starting material, which was
monitored by HPLC (35% CH3CN). After removing the solvent
under reduced pressure, the crude residue was purified by flash
chromatography (93:7 f 90:10 f 80:20 DCM/MeOH). Yield
71%. ESI mass: 1453.6 (m/z 2+), 969.4 (m/z 3+). 406 mg of
this product was dissolved with 3:5 DMF/MeOH and orthogo-
nally deprotected by means of ammonium formate (44 mg, 0.7
mmol) and Pd/C (200 mg). After 3 h stirring, the reaction was
filtered, and the filtrate concentrated to give the intermediate 8,
which was dissolved with 3 mL DMF and used for the next
step. N-(Methyl-PEG11-carboxy)-propargylglycine (102 mg,
0.149 mmol in 4.5 mL DCM) followed by HCTU (62 mg, 0.149
mmol) and DIPEA (51 µL, 0.298 mmol) were added, and the
resulting solution stirred for 2 h. After evaporation, the residue
was redissolved with 300 mL of DCM and washed with water
(2 × 50 mL).The organic phase was evaporated, and the
fragment fully deprotected by means of a mixture of TFA/DCM/
thioanisole (1:1:0.3). After taking to dryness, the residue was
purified by repeated precipitations from TFA with cold diethyl
ether, to afford 245 mg of fragment F6, with 41% yield
(calculated from 3 steps).
Conjugate Stability in Murine Blood. The blood of untreated
healthy mice was collected in tubes containing 2% EDTA.The
sample molecules were added to obtain a concentration of 5
µM.The tubes were incubated at 37 °C under rotation for different
sampling times during an interval of 3 h. At each time point, the
samples were centrifuged at 1000g for 10 min, and the supernatants
transferred in new tubes and stored at -20 °C. Plasma (100 µL)
was processed by adding 700 µL of a cold mixture of 0.1% acetic
acid and methanol (1:5 v/v). After vortexing, samples were kept
at 20 °C for 10 min and then centrifuged at 14 000g for 10 min at
4 °C. The supernatants were analyzed by RP-HPLC-FL Beckman;
column, Discovery HS-F5, 100 × 4.6 mm, 5 µ, Supelco; mobile
phase, A ) 0.1 M AcOH + 0.1% TEA, B ) CH3CN; flow rate,
1 mL/min; spectrofluorometric detector, RF-10AXL Shimadzu;
wavelengths, λex 370 nm and λem 510 nm. A gradient elution was
employed ranging from 25% to 33% B within 30 min. Data were
acquired by computer system and processed by 32Karat Software,
Beckman. Calibration curves were performed in plasma, and peak
areas were used to calculate each calibration curve intercept and
the slope of each calibration curve equation by weighted least-
squares method, using l/y as weighting factor. All peaks were well-
resolved from impurities of plasma and methods were linear in
the tested range (r2 > 0.985), with results suitable for quantification.
Results are reported in Table 6 as a percentage of residual amounts
found at each sampling time versus time zero (calculated using
mean values of triplicate samples).
RESULTS
Synthesis. Two groups of fragments have been prepared: the
fragments F1-F6 of the first group (Chart 2) contain one or

1962 Bioconjugate Chem., Vol. 21, No. 11, 2010
Dal Pozzo et al.
Scheme 4. Synthesis of Fragment F7a
a Reagents and conditions: (i) LiOH · H2O; (ii) 4-aminobenzylalcohol, HOAT, DCC; (iii) 4-nitrophenylchloroformate, py; (iv) 7-(2-
aminoethoxyimine)-methyl-camptothecin.HCl, TEA, DMF; (v) TFA/DCM 1:1. Total yield 46.8%.
Scheme 5. Synthesis of Fragments F8-F11a
a Reagents and conditions: (i) LiOH ·H2O; (ii) 4-aminobenzylalcohol, HOAT, DCC; (iii) 4-nitrophenylchloroformate, py; (iv) NMT, TEA; (v)
0.5% TFA in trifluoroethanol, 30 min, total yield 56.2%; (vi) propiolic acid, HOAT, DCC; yield 72.6%; (vii) HCC-CH2-NH-CO-CH2-O-(CH2-
CH2O)2-CH2-COOH, DCC, yield 63.5%; (viii) N3-(CH2)2-O-(CH2-CH2-O)2-(CH2)2-COOH, DCC, yield 62.8%.
two RGD cyclopeptides, which differ from each other for the
amino acid in position 5 bearing glycol chains of different length
and conformation. These fragments were built with good yields
by MW-assisted SPPS, following cyclization in solution. In the
case of the dimeric fragments F5 and F6, two cyclopeptides
were connected in solution through a dendron based on glutamic
acid. The experimental procedures are given in detail and
illustrated in Schemes 2 and 3. Fragments F7-F11 of the second
group (Chart 3) contain the camptothecin derivative and the
lysosomally cleavable dipeptide sequence.The dipeptide is
connected to the N-terminal of the drug through 4-aminoben-
zylalcohol (PABA). PABA is a known self-immolative entity
(31) that accomplishes a double function, to increase the distance
between peptide and cytotoxic drug, allowing a better recogni-
tion by the endopeptidases, and to guarantee prompt release of
the parent drug through a 1,6-elimination mechanism, only after
the enzymatic hydrolysis. The fragments belonging to the latter
group were all prepared with similar synthetic procedures, except
for minor modifications. In fact, during the synthesis of the first
fragment F7 (Scheme 4), deprotection of the terminal N-Boc
caused a loss of about 10% of NMT, even reducing the TFA
treatment to a few minutes. Therefore, N-Bpoc protection was
used for fragments F8-F11 (Scheme 5), because it can be
removed with 0.5% TFA in trifluoroethanol without any
cleavage of the carbamate bond.
In each fragment of both groups, a unique amino, hydrazido,
azido, or alkyne function was introduced for further reaction
with the corresponding counterpart to give the desired conjugates
C1-C7 (Chart 1). Each pair of fragments used for the synthesis
of each conjugate is indicated in Table 1. The formation of

RGD Peptide-Camptothecin Conjugates
Bioconjugate Chem., Vol. 21, No. 11, 2010 1963
Table 1. Reacting Fragments, Coupling Methods Used in the
Synthesis of Conjugates C1-C7 and Yields
procedure (TAEC) introduced by Ramage et al. (32), as
described in Scheme 1 (Method 1). Later on, for conjugates
C3-C7, we found more convenient the use of Huisgen 1,3-
dipolar reaction (33), (Method B). In fact, this cycloaddition
affords cleaner products, because of its mild nature; more-
over, grafting of the residual alkyne with the azide-terminated
molecule enables the use of equivalent amounts of the two
fragments, whereas the acylazide reaction requires two
equivalents of the RGD counterpart. In the case of the bulkier
conjugates, MW irradiation was applied to accelerate the
reaction. Physico-chemical characterization of all fragments
and conjugates is reported in the Tables 2, 3, and 4.
fragments
conjugates
yields %
F1 + F7
F1 + F8
F2 + F9
F3a + F9
F3b + F9
F4 + F9
F5 + F10
F6 + F11
C1a
C2a
C3b
C4ab
C4bb
C5b
C6b
C7b
46.6
55.0
44.0
42.5
45.0
39.0
41.0
42.0
a Method A. b Method B.
Effect of Spacer on the Conjugate Receptor Affinity,
Tumor Cell Adhesion, and Stability in Murine Blood. The
effects of spacer modifications on the behavior of the conjugates
RGD-NMT are illustrated in Tables 5, 6, and 7. All spacers
contain the dipeptide alanine-citrulline as a lysosomally cleav-
able sequence, except for C1, which contains phenylalanine-
lysine. Within the series C2-C5, the spacers differ from each
conjugates needs a chemoselective synthesis. In fact, the reacting
fragments must be used in a totally deprotected form, because
the carbamate bond between NMT and the rest of the
molecule is very sensitive to all late deprotection conditions.
Two different methods were used: the conjugates C1 and
C2 were synthesized via the acylazide method from the initial
hydrazide, using the transfer active ester condensation
Table 2. Physico-Chemical Characterization of Conjugates C1-C7
entry
HPLCa, rt, minb
Maldi mass [M+H]+
1H NMR
C1
9.0, 12.0 (35%)
1696.6
(DMSO-d6 + D2O) δ 9.29 (s, 1H), 8.60 (d, 1H; J ) 9.4), 8.22 (d, 1H, J ) 9.4), 7.87 (t,
1H, J ) 8.2), 7.73 (t, 1H, J ) 8.2), 7.40 (s, 1H), 7.25 (d, 2H, J ) 8.2), 7.20-7.10 (m,
14H), 7.03 (d, 2H; J ) 8.2), 5.42-5.30 (m, 4H), 5.02 (s, 2H), 4.71-4.58 (m, 2H),
4.48-4.32 (m, 4H), 4.30-4.23 (m, 2H), 4.20-4.00 (m, 4H), 3.93 (s, 2H), 3.89-3.72 (m,
4H), 3.63-3.11 (m, 10H), 2.92-2.63 (m, 9H), 2.40-2.21 (m, 1H), 1.95-1.80 (m, 2H),
1.65-1.30 (m, 10H), 0.88 (t, 3H, J ) 7.4).
C2
C3
7.96, 10.4 (34%)
7.7, 9.9 (34%)
1650.71
1745.7
(DMSO-d6 + D2O): δ 9.29 (s, 1H), 8.60 (d, 1H, J ) 9.4), 8.22 (d, 1H, J ) 9.4), 7.87 (t,
1H, J ) 8.2), 7.73 (t, 1H, J ) 8.2), 7.36 (s, 1H), 7.24 (d, 2H, J ) 8.5), 7.15-7.11 (m,
9H), 7.03 (d, 2H, J ) 8.5), 5.42 (s, 2H), 5.36 (s, 2H), 4.95 (s, 2H), 4.64-4.60 (m, 1H),
4.45-4.27 (m, 8H), 4.07-4.00 (m, 3H), 3.94 (s, 2H), 3.90 (s, 2H), 3.59-3.46 (m, 10H),
3.09-2.66 (m, 9H), 2.40-2.33 (m, 1H), 1.90-1.86 (m, 2H), 1.75-1.38 (m, 8H), 1.24
(d, 3H, J ) 6.9), 0.89 (t, 3H, J ) 7.3).
(DMSO-d6 + D2O) δ 9.19 (s, 1H), 8.50-8.40 (m, 2H), 8.19 (d, 1H, J ) 8.4), 7.80 (t, 1H,
J ) 8.2), 7.74 (t, 1H, J ) 8.2), 7.45 (d, 2H, J ) 8.5), 7.39 (s, 1H), 7.19-6.95 (m, 11H),
5.48-5.30 (m, 3H), 5.19 (s, 1H), 4.89 (s, 2H), 4.69 (s, 1H), 4.60-4.24 (m, 8H), 4.20 (s,
2H), 4.13 (m, 1H), 4.02-3.98 (m, 2H), 3.89-3.75 (m, 2H), 3.52-3.37 (m, 10 H),
3.30-3.22 (m, 2H), 3.10-2.62 (m, 9H), 2.40-2.30 (m, 3H), 1.91-1.81 (m, 2H),
1.76-1.38 (m, 8H), 1.34 (d, 3H, J ) 6.9), 0.86 (t, 3H, J ) 7.3).
(DMSO-d6 + D2O) δ 9.13 (s, 1H), 8.47-8.35 (m, 2H), 8.20-8.09 (m, 1H), 7.90-7.81
(m, 1H), 7.79-7.69 (m, 1H), 7.45-7.30 (m, 3H), 7.19-6.90 (m, 11H), 5.49-5.20 (m,
4H), 4.86 (s, 2H), 4.60-3.90 (m, 15H), 3.65-3.50 (m, 4H), 3.49-3.33 (m, 20H),
3.32-3.10 (m, 4H), 3.08-2.60 (m, 10H), 2.45-2.20 (m, 6H), 1.92-1.37 (m, 14H), 1.33
(d, 3H, J ) 5.1), 0.85 (t, 3H, J ) 5.7).
(DMSO-d6) δ 9.94 (s, 1H), 9.28 (s, 1H), 9.04 (s, 1H), 8.58 (d, 1H, J ) 9.1), 8.52 (s, 1H),
8.30-8.16 (m, 3H), 8.15-7.98 (m, 2H), 7.97-7.72 (m, 5H), 7.62-7.48 (m, 3H), 7.37 (s,
1H), 7.25 (d, 2H, J ) 8.5), 7.20-6.90 (m, 7H), 6.82 (d, 2H, J ) 7.8), 6.56 (d, 2H, J )
7.6), 6.41 (s, 1H), 5.98-5.83 (m, 2H), 5.48-5.20 (m, 8H), 4.95 (s, 2H), 4.65-4.53 (m,
3H), 4.52-4.32 (m, 4H), 4.31-4.12 (m, 4H), 4.11-3.96 (m, 4H), 3.84 (t, 2H, J ) 5.1),
3.65-3.35 (m, 24H), 3.20-2.80 (m, 16H), 2.45-2.33 (m, 4H), 1.98-1.83 (m, 2H),
1.82-1.40 (m, 12H), 1.35 (d, 3H, J ) 7), 0.89 (t, 3H, J ) 7.2).
(D2O) δ 8.73 (s, 1H), 8.52 (s, 1H), 7.83 (s, 1H), 7.75 (s, 1H), 7.62 (s, 1H), 7.39 (s, 1H),
7.19 (d, 2H, J ) 7.9), 7.05 (d, 2H, J ) 7.9), 7.00-6.80 6H), 6.63 (d, 2H, J ) 8.3),
5.65-5.40 (m, 4H), 5.05-4.85 (m, 2H), 4.70-4.62 (m, 2H), 4.60-4.25 (m, 14H),
4.10-4.00 (m, 2H), 3.95-3.35 (m, 34H), 3.30-3.10 (m, 10H), 3.00-2.50 (m, 16H),
2.15-2.05 (m, 2H), 2.00-1.45 (m, 17H), 1.15-1.05 (m, 3H).
(DMSO-d6 + D2O) δ 9.30 (s, 1H), 8.57 (d, 1H, J ) 8.2), 8.40 (s, 1H), 8.22 (d, 1H, J )
8.2), 7.91 (t, 1H, J ) 7.1), 7.71 (t, 1H, J ) 7.3), 7.52 (d, 2H, J ) 8.5), 7.38 (s, 1H),
7.23 (d, 2H, J ) 8.5), 7.18-7.04 (m, 8H), 6.81 (d, 4H, J ) 8), 6.56 (d, 4H, J ) 8.2),
5.42 (s, 2H), 5.38 (s, 2H), 5.22 (s, 1H), 5.10 (s, 2H), 4.93 (s, 1H), 4.74 (s, 1H),
4.38-4.34 (m, 9H), 4.26-4.14 (m, 7H), 4.06-4.03 (m, 2H), 3.90 (s, 4H), 3.91-3.83 (m,
2H), 3.62 (t, 4H, J ) 6.1), 3.55-3.36 (m, 10H), 3.29-2.79 (m, 20H), 2.39 (t, 4H, J )
6.1), 2.20-2.08 (m, 2H), 1.89-1.82 (m, 3H), 1.75-1.33 (m, 13H), 1.23 (d, 3H, J )
6.9), 0.90-0.84 (m, 3H).
(DMSO-d6 + D2O) δ 9.05 (s, 1H), 8.31 (d, 1H, J ) 8.2), 8.10 (d, 1H, J ) 8.2), 7.80 (t,
1H, J ) 7.2), 7.77 (s, 1H), 7.73-7.61 (m, 3H), 7.40 (s, 1H), 7.43-7.29 (m, 2H),
7.06-6.99 (m, 10H), 6.80 (d, 4H, J ) 8.2), 6.53 (d, 4H, J ) 8.2), 5.49 (s, 2H),
5.43-5.11 (m, 4H), 4.79 (s, 1H), 4.57 (s, 1H), 4.43-4.11 (m, 24H), 3.70 (s, 6H),
3.60-3.30 (m, 78H), 3.17 (s, 3H), 3.05-2.70 (m, 20H), 2.40-2.22 (m, 10H), 2.20-1.23
(m, 16H), 1.19 (d, 3H, J ) 7), 0.82 (t, 3H, J ) 7).
C4a
C4b
11.2, 15.4 (30%)
10.8, 15.2 (29%)
2106.0
2120.89
C5
C6
11.4, 16.2 (28%)
9.2, 12.6 (29%)
2480.0
2988.78
C7
10.0, 12.5 (29%)
3721.0
a The conjugates show two peaks corresponding to E/Z isomers of the original cytotoxic molecule. b Acetonitrile % in the mobile phase is indicated
in brackets.

1964 Bioconjugate Chem., Vol. 21, No. 11, 2010
Dal Pozzo et al.
Table 3. Physico-Chemical Characterizations of Fragments F1-F6, Containing Cyclopeptide Residues
entry
HPLCa, rt, min
Maldi mass [M + H]+
1H NMR
F1
9.14 (20%)
870.13
(DMSO-d6 + D2O) δ 7.14-7.09 (m, 5H), 7.02-6.97 (m, 4H), 4.57-4.52 (m, 2H),
4.46-4.44 (m, 2H, 4.26 (s, 2H), 4.00 (d, 1H, J ) 15), 3.95-3.90 (m, 2H), 3.57-3.51
(m, 8H), 3.26 (d, 1H, J ) 15), 3.04 (t, 2H, J ) 6.7), 2.87-2.64 (m, 5H), 2.38-2.33 (m,
1H), 1.75-1.72 (m, 1H), 1.51-1.47 (m, 1H), 1.34-1.32 (m, 2H).
(DMSO-d6 + D2O) δ 7.13-6.98 (m, 9H), 4.57-4.52 (m, 1H), 4.47-4.40 (m, 1H),
4.31-4.28 (m, 1H), 4.27 (s, 2H), 4.03-3.98 (m, 2H), 3.62-3.48 (m, 12H), 3.32 (t, 2H,
J ) 5), 3.30 (d,1H, J ) 9), 3.04 (t, 2H, J ) 6), 2.93-2.60 (m, 5H), 2.54-2.35 (m, 3H),
1.85-1.67 (m, 1H), 1.65-1.45 (m, 1H), 1.43-1.35 (m, 2H).
(D2O) δ 7.29-7.11 (m, 9H), 4.65 (t, 1H, J ) 8.2), 4.40 (s, 2H), 4.35-4.17 (m, 4H),
3.76-3.56 (m, 24H), 3.49 (t, 2H, J ) 4.7), 3.39-3.36 (m, 2H), 3.18-3.06 (m, 4H),
2.97-2.83 (m, 4H), 2.75-2.65 (m, 1H), 2.58 (t, 3H, J ) 7.6), 1.85-1.34 (m, 8H).
(D2O) δ 7.43 (d, 2H, J ) 8.2), 7.30 (d, 2H, J ) 8.2), 7.18 (d, 2H, J ) 8.8), 6.94 (d, 2H, J
) 8.8), 4.93 (t, 1H, J ) 6.5), 4.59 (s, 2H), 4.53-4.36 (m, 4H), 4.01-3.74 (m, 26H),
3.67 (t, 2H, J ) 4.7), 3.59-3.52 (m, 2H), 3.35 (t, 2H, J ) 7.0), 3.27 (t, 2H, J ) 7.0),
3.14-3.04 (m, 5H), 2.95-2.87 (m, 1H), 2.79-2.72 (m, 4H), 2.03-1.58 (m, 8H).
(D2O), δ 7.25 (d, 2H, J ) 8.2), 7.12 (d, 2H, J ) 8.2), 7.00 (d, 2H, J ) 8.8), 6.76 (d, 2H,
J ) 8.8), 4.60 (t, 1H, J ) 6.5), 4.42 (s, 2H), 4.31-4.19 (m, 6H), 3.83-3.57 (m, 40H),
3.52-3.40 (m, 5H), 3.36 (s, 4H), 3.21-3.07 (m, 6H), 2.96- 2.67 (m, 2H), 2.62-2.51 (m,
6H), 1.92-1.30 (m, 12H).
F2
8.3 (30%)
881.0
F3a
F3b
10.8 (25%)
8.9 (22%)
1241.0
1256.96
F4
11.6 (21%)
1617.31
F5
F6
10.9 (22%)
8.7 (26%)
1935.22
2679.79
(DMSO-d6 + D2O) δ 7.78 (s, 1H), 7.76 (s, 1H), 7.08 (d, 4H, J ) 7.7), 7.01 (d, 4H, J )
8.2), 6.78 (d, 4H, J ) 8.3), 6.54 (d, 4H, J ) 8.2), 4.52-4.44 (m, 2H), 4.42-4.00 (m,
16H), 3.73 (t, 4H, J ) 5.3), 3.60-3.54 (m, 7H), 3.50-3.37 (m, 67H), 3.31 (s, 1H), 3.26
(s, 1H), 3.20 (s, 3H), 3.02 (t, 5H, J ) 6.5), 2.90-2.52 (m, 11H), 2.36 (t, 8H, J ) 6.5),
2.19-2.10 (m, 2H), 2.0-1.90 (m, 1H), 1.80-1.68 (m, 4H), 1.55-1.42 (m, 2H),
1.35-1.28 (m, 4H)
a Acetonitrile % in the mobile phase is indicated in brackets.
Table 4. Physico-Chemical Characterization of Fragments F7-F11, Containing the Cytotoxic Drug NMT
entry
HPLCa, rt, minb
mass
1H NMR
F7
18.0, 25.0 (35%)
965.0c
(DMSO-d6) δ 9.96 (s, 1H), 9.28 (s, 1H), 8.56 (d, 1H, J ) 8.5), 8.28 (s, 1H), 8.21 (d, 1H,
J ) 8.5), 8.05-8.00 (m, 1H), 7.90 (t, 1H, J ) 7.4 Hz), 7.75 (t, 1H, J ) 7.4), 7.52 (d,
1H, J ) 8.4), 7.36 (s, 2H), 7.30-7.19 (m, 9H), 6.41 (s, 1H), 5.86-5.80 (m, 1H),
5.42-5.36 (m, 2H), 5.26-5.11 (m, 2H), 4.95 (s, 2H), 4.43-4.37 (m, 6H), 3.48 (t, 2H, J
) 5.7), 3.00-2.61 (m, 4H), 1.90-1.82 (m, 2H), 1.74-1.50 (m, 2H), 1.49-1.35 (m, 2H),
1.30-1.18 (m, 2H), 0.82 (t, 3H, J ) 7.2).
(DMSO-d6) δ 10.10 (s, 1H), 9.31 (s, 1H), 8.59 (d, 1H, J ) 8.6), 8.22 (d, 1H, J ) 8.5),
8.04 (s, 3H), 7.91 (t, 1H, J ) 7.9), 7.76 (t, 1H, J ) 8.1), 7.54 (d, 2H, J ) 8.2), 7.36 (s,
1H), 7.25 (d, 2H, J ) 8.6), 6.51 (s, 1H), 6.02 (t, 1H, J ) 6.3), 5.45 (s, 2H), 5.42 (s, 2H),
5.36 (s, 2H), 4.95 (s, 2H), 4.50-4.47 (m, 1H), 4.37 (t, 2H, J ) 5.6), 3.95-3.85 (m, 1H),
3.49-3.42 (m, 2H), 3.10-2.92 (m, 2H), 1.91-1.82 (m, 2H), 1.75-1.38 (m, 4H), 1.34
(d, 3H, J ) 6.9), 0.88 (t, 3H, J ) 7.2).
F8
12.0, 16.9 (26%)
812.32d
F9
F10
11.4, 16.2 (31%)
11.6, 16.3 (32%)
885.8c
1053.4d
(DMSO-d6) δ 9.89 (s, 1H), 9.28 (s, 1H), 8.58 (d, 1H, J ) 8.2), 8.22 (d, 1H, J ) 8.2), 8.11
(d, 1H, J ) 7.8), 8.03-7.99 (m, 1H), 7.90 (t, 1H, J ) 7.9), 7.76 (t, 1H, J ) 7.9), 7.64
(d, 1H, J ) 7.2), 7.55 (d, 2H, J ) 8.4), 7.37 (s, 1H), 7.24 (d, 2H, J ) 8.4), 6.41 (s, 1H),
5.91 (t, 1H, J ) 5.7), 5.42 (s, 2H), 5.37 (s, 2H), 5.32 (s, 2H), 4.95 (s, 2H), 4.43-4.36
(m, 3H), 3.92 (s, 2H), 3.89 (s, 4H), 3.88-3.86 (m, 1H), 3.60-3.58 (m, 8H), 3.48-3.44
(m, 2H), 3.01-2.95 (m, 3H), 1.92-1.86 (m, 2H), 1.70-1.38 (m, 4H), 1.25 (d, 3H, J )
6.9), 0.89 (t, 3H, J ) 7.3).
(DMSO-d6) δ 9.81 (s, 1H), 9.28 (s, 1H), 8.51 (d, 1H, J ) 9.0), 8.20 (d, 1H, J ) 8.1), 7.89
(t, 1H, J ) 7.1), 7.86-7.75 (m, 2H), 7.49 (d, 2H, J ) 8.1), 7.39 (s, 1H), 7.21 (d, 2H, J
) 8.4), 6.40 (s, 1H), 5.90-5.88 (m, 1H), 5.42 (s, 2H), 5.37 (s, 2H), 5.31 (s, 2H), 4.95 (s,
2H), 4.39-4.27 (m, 4H), 3.60-3.49 (m, 14H), 3.37 (t, 2H, J ) 4.8), 3.01-2.95 (m, 2H),
2.40-2.38 (m, 2H), 1.90-1.86 (m, 2H), 1.71-1.38 (m, 4H), 1.20 (d, 3H, J ) 7.0), 0.89
(t, 3H, J ) 7.2).
F11
9.0, 12.0 (36%)
1041.4d
a Acetonitrile % in the mobile phase is indicated in brackets. b These fragments show two peaks corresponding to E/Z isomers of the original
c
d
cytotoxic molecule. Maldi [M + Na]+. Esi [M + H]+.
other for the length and disposition of the hydrophilic glycol
chains, whereas the spacers of conjugates C6 and C7 contain
two RGD ligands connected to the drug through multiple glycol
chains in a branched conformation. In Table 5, the conjugate
affinity to Rvꢀ3 integrin and their tumor cell adhesion is reported
in parallel with the ligand cyclopeptides and the drug alone.
C1-C3, bearing a short glycol between ligand and drug, showed
very poor solubility and lost affinity to the receptor if compared
with the ligand alone, most probably due to steric hindrance
induced by the drug. However, when the glycol spacers were
elongated in C4-C5, affinity and solubility improved, but the
stability decreased, as it is possible to observe in Table 6. This
is consistent with the results reported in Table 7: as stability
decreased, cytotoxicity increased, obviously due to some
amounts of the free drug in the culture media. Only the dimeric
conjugates C6 and C7 exhibited good receptor affinity and
cytotoxicity, together with sufficient stability and optimal
solubility.
Cell adhesion of the conjugates resulted generally increased
with respect to RGD peptides alone, and this may be attributed
to cooperative adhesion of the whole molecule.
DISCUSSION
In the present work, RGD peptides as targeting moieties
were conjugated to the cytotoxic agent Namitecan through

RGD Peptide-Camptothecin Conjugates
Bioconjugate Chem., Vol. 21, No. 11, 2010 1965
Table 5. Binding to Isolated rvꢀ3 Integrina of Conjugates C1-C7
and Their Effect on Adhesion to PC3 (Prostatic Carcinoma) and
A2780 (Ovarian Carcinoma) Cell Lines, in the Presence of
Vitronectin, after 3 h Treatment
release CPT, probably because of the steric crowding of the
bulky drug), even though some of them have been used with
success in the literature, mainly for immunoconjugates.
Finally, the Ala-Cit dipeptide was found to have a signifi-
cantly longer life in murine blood, together with sufficient
cleavability by tumor cell associated proteases, as demon-
strated by the cytotoxic activity of the corresponding
conjugates; thus, Ala-Cit was adopted in all subsequent
conjugate spacers. However, the stability was greatly influ-
enced by the rest of the molecule, and this is observable with
conjugates C3-C5, where elongation of the hydrophilic
glycol chains led to a decrease of the half-life in the murine
blood (see Table 6). To improve the solubility, we chose to
substitute the commonly used large poly(ethylene glycol)s
(PEGs) with a number of small glycol chains. In fact, PEGs,
which are unique water-solubilizing agents currently em-
ployed for bioactive systems, are commercially available as
polydisperse oligomers, and would yield undesirable hetero-
geneous products; moreover, because of their conformational
flexibility, they tend to create bulky loops that can interfere
with the ligand binding to the receptor. On the contrary, short
glycols can be distributed on strategical positions of the
molecular construct with respect to both the targeting and
cleavable peptides. At first, we introduced in the spacers short
glycol fragments connected in series through citrulline
residues, which, beyond increasing hydrophilicity, should
impart rigidity to the chain by their amide bonds, without
shielding the ligand from recognition. Nevertheless, this
strategy proved unsuccessful: in fact, as solubility improved,
stability decreased, and cytotoxicity increased, due to the
presence of significant amounts of the parent drug. This
behavior indicates that the cleavable peptide is more exposed
to the blood proteases. A deshielding effect is also observable
in the conjugate C3, where the rigid triazole linkage orients
the spacer chain away from the cleavable dipeptide moiety,
where it is more accessible. On the other hand, solubility
issues are a potentially negative aspect that could be a serious
obstacle for preclinical development, and introduction of
hydrophilic moieties is unavoidable. In an attempt to increase
pegylation, while maintaining plasma stability and affinity
to the receptor, we designed dimeric conjugates containing
two RGD linked to a dendron unit bearing different glycol
chains, to obtain a multifunctional platform where different
pendant domains can be strategically allocated, without
disturbing each other. Moreover, this approach exploits the
multivalency effect on the ligand affinity (16, 18, 34). In
this way, we could find conjugates C6 and C7, endowed with
the desired characteristics (increased avidity for the receptor
together with appreciable stability and solubility), which
demonstrated superior in vitro efficacy.
binding, IC50 nM
cell adhesion, IC50 µM
entry
Rvꢀ3
PC3
A2780
P1
P2
C1
C2
1.70 ( 0.10
1.08 ( 0.12
9.71 ( 0.06
30.40 ( 2.25
11.00 ( 0.81
2.99 ( 0.26
4.36 ( 0.22
9.38 ( 0.17
3.00 ( 0.10
6.20 ( 0.08
no binding
23.20 ( 5.02
25.20 ( 4.50
1.80 ( 0.40
2.10 ( 0.20
0.56 ( 0.07
1.60 ( 0.80
0.50 ( 0.09
1.40 ( 0.20
0.39 ( 0.09
1.00 ( 0.05
3.80 ( 0.30
4.50 ( 0.20
2.70 ( 0.30
0.90 ( 0.05
0.67 ( 0.02
1.70 ( 0.20
1.70 ( 0.10
5.40 ( 0.30
0.45 ( 0.02
1.00 ( 0.00
C3
C4a
C4b
C5
C6
C7
NMT
no adhesion
a Nonspecific binding was determined in the presence of 1 µM
Echistatin.
Table 6. Stability of Conjugates C1-C7 in the Whole Blood of CD1
Mice, Expressed as Percentage of Residual Amounts Found at
Different Hours vs Time Zero
entry
1 h
1.8
2 h
3 h
C1
C2
C3
C4a
C4b
C5
C6
C7
n.d.
100
34
8
n.d.
100
20
0.9
1.5
2.5
67
100
56
34
35
38
76
67
9
11
73
61
51
Table 7. Cytotoxicity of Conjugates C1-C7, after 3 h Treatment,
IC50 µM
entry
PC3
A2780
C1
C2
C3
C4a
C4b
C5
C6
C7
0.22 ( 0.003
4.60 ( 0.80
1.00 ( 0.10
2.50 ( 0.70
0.84 ( 0.20
0.42 ( 0,05
1.00 ( 0.02
1.00 ( 0.01
0.36 ( 0.05
>3000
0.0084 ( 0.0006
0.095 ( 0.02
0.03 ( 0.003
0.009 ( 0.0007
0.013 ( 0.0006
0.02 ( 0.003
0.008 ( 0.0005
0.008 ( 0.0001
0.0049 ( 0.0003
>3000
NMT
P1
P2
>3000
>3000
different spacers, consisting of linear or branched glycol
residues and a peptide suitable for releasing the drug by
enzymatic hydrolysis inside the tumor cell. The products were
evaluated for their in vitro tumor cell-targeting efficacy,
stability in murine blood, solubility, and cytotoxicity, with
the purpose of studying their structure-activity relationship.
Since the spacers between carrier and drug play the main
role in the fate of small targeted drug conjugates, we
systematically modified these molecular bridges in a continu-
ing effort to reach a fine balance between sufficient systemic
stability and cleavability inside the tumor cells, while
preserving high receptor affinity and solubility of the whole
conjugate. In the first conjugate C1, a spacer containing the
Phe-Lys dipeptide was introduced, because this sequence had
been proposed by Dubowchick et al. (26, 27) for rapid
lysosomal hydrolysis and high human plasma stability.
However, we assisted at the total cleavage of C1 within one
hour in the murine blood. In a search for more stable peptides,
we evaluated other peptide sequences (such as D-Ala-Phe-
Lys, Val-Cit, Aib-Cit, Tic-Cit), but with negative results (Dal
Pozzo, A., Ni, M., and Bucci, F. Data not published. We
also confirmed with our representative examples that com-
pounds lacking the PABC self-immolative group do not
CONCLUSION
Our study highlights the crucial role of the spacer between
targeting device and drug in tumor-targeted conjugates.
Unlike immunoconjugates (where the bulky nature of the
targeting molecule is less sensitive to the influence of the
drug and spacer counterparts, and can shield and stabilize
the connecting bond from premature hydrolysis), engineering
conjugates with low-molecular-weight ligands is a most
challenging problem, its success depending on a plateau of
different factors. In fact, the small targeting molecule must
retain tumor binding after substitution with molecules, that
may be of similar or larger size, and the addition of new
domains results in the modification of existing domains
affecting their properties. Accordingly, careful selection of
the enzymatically cleavable peptide and the length and
disposition of the hydrophilic chains shoud be necessary for

1966 Bioconjugate Chem., Vol. 21, No. 11, 2010
Dal Pozzo et al.
(13) Arap, W., Pasqualini, R., and Ruoslathi, E. (1998) Cancer
treatment by targeted drug delivery to tumor vasculature in a
mouse model. Science 279, 377–380.
(14) Haubner, R., and Wester, H. J. (2004) Radiolabelled traces
for imaging of tumor angiogenesis and evaluation of antiangio-
genic therapies. Curr. Pharm. Des. 10, 1439–455.
(15) 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.
(16) Chen, X., Plasencia, C., Hou, Y., and Neamati, N. (2005)
Synthesis and biological evaluation of dimeric RGD peptide-
placlitaxel conjugate as a model for integrin-targeted drug
delivery. J. Med. Chem. 48, 1098–1106.
(17) Meyer, A., Auernheimer, J., Modlinger, A., and Kessler, H.
(2006) Targeting RGD recognizing integrins: drug development,
biomaterial research, tumor imaging and targeting. Curr. Pharm.
Des 12, 2723–2747.
(18) Ryppa, C., Mann-Steinberg, H., Fichtner, I., Weber, H., Satchi-
Fainaro, R., Biniossek, M. L., and Kratz, F. (2008) In vitro and
in vivo evaluation of doxorubicin conjugates with the divalent
peptide E-[c(RGDfK)2)] that targets integrin Rvꢀ3. Bioconjugate
Chem. 19, 1414–422.
(19) McNerny, A. M., Kukowska-Latallo, J. F., Mullen, D. G.,
Wallace, J. M., Desai, A. M., Shukla, R., Huang, B., Holl,
M. M. B., and Baker, J. R. (2009) RGD dendron bodies: synthetic
avidity agents with defined and potentially interchangeable
effector sites that can substitute for antibodies. Bioconjugate
Chem. 20, 1853–1859.
the development of each single small targeting molecule for
therapeutic application, since the strategy is not generalizable.
We tried to help bring order to this sometimes unruly subject,
and optimized the properties of our conjugates after a
systematic in vitro study. Our results demonstrate the
potential of tumor-targeted delivery of conjugates based on
the specific recognition of the integrin RGD ligands and
selective release of the drug by tumor-associated enzymes.
However, we are aware of possible in vivo problems, such
as the following: (1) Partial distribution of free drug generated
through extracellular proteases cannot be ruled out. (2) The
hydrophilic compounds synthesized may undergo rapid renal
clearance, and larger PEGs would be needed to enhance the
retention time in the circulation and allow proper tumor
accumulation. Nevertheless, an improved therapeutic index
could be expected by tumor specific delivery of at least a
part of the targeted drug, resulting in lower and therefore
less toxic systemic doses necessary to obtain the antitumor
effect. Further evaluation of conjugates C6 and C7 in
preclinical animal models for tumor growth inhibition and
acute toxicity studies are currently in progress.
ACKNOWLEDGMENT
We are grateful to Sigma-Tau, Roma, Italy, for financial
assistance. We thank Gianni Dottori for technical assistance.
LITERATURE CITED
(20) Pisano, C., De Cesare, M., Beretta, G. L., Zuco, V., Pratesi,
G., Penco, S., Vesci, L., Fodera`, R., Ferrara, F. F., Guglielmi,
M. B., Carminati, P., Dalla valle, S., Morini, G., Merlini, L.,
Orlandi, A., and Zumino, F. (2008) Preclinical profile of
antitumor activity of a novel Hydrophilic camptothecin, ST1968.
Mol. Cancer Ther. 7, 2051–2059.
(21) Conover, C. D., Greenwald, R. B., Pendri, A., Gilbert, K. W.,
and Shum, K. L. (1998) Camptothecin delivery systems.
Enhanced efficacy and tumor accumulation of camptothecin
following its conjugation to poly(ethylene glycol) via a glycine
linker. Cancer Chemother. Pharmacol 42, 407–414.
(22) Caiolfa, V. R., Zamai, M., Fiorino, A., Frigerio, E., Pellizzoni,
C., d’Argi, R., Ghigleri, A., Castelli, M. G., Farao, M., Pesenti,
E., Gigli, M., Angelucci, F., and Suarato, A. (2000) Polymer-
bound camptothecin: initial biodistribution and antitumor activity
studies. J. Controlled Release 65, 105–119.
(23) Lerchen, H. G., Baumgarten, J., von dem Bruch, K., Lehmann,
T. E., Sperzel, M., Kempka, G., and Fiebig, H. H. (2001) Design
and optimization of 20-O-linked camptothecin glycoconjugates
as anticancer agents. J. Med. Chem. 44, 4186–4195.
(24) Dal Pozzo, A., Ni, M. H., Esposito, E., Dallavalle, S., Musso,
L., Bargiotti, A., Pisano, C., Vesci, L., Bucci, F., Castorina, M.,
Fodera`, R., Giannini, G., and Penco, S. (2010) Novel tumor-
targeted RGD peptide-camptothecin conjugates: synthesis and
biological evaluation. Biorg. Med. Chem. 18, 64–72.
(25) Duncan, R., Cable, H. C., Lloyd, J. B., Rejmanova, P., and
Kopecek, J. (1983) Polymers containing enzymatically degrad-
able bond, 7. Design of oligopeptides side chains in poly N-2-
hydroxypropyl)methacrylamide copolymers to promote degrada-
tion by lysosomal enzymes. Macromol. Chem. 184, 1997–2008.
(1) Doronina, S. O., Toki, B. E., Torgov, M. Y., Mendelsohn, B. A.,
Cerveny, C. G., Chace, D. F., DeBlank, R. I., Gearing, R. P.,
Bovee, T. D., Siegall, C. B., Francisco, J. A., Wahl, A. F., Meyer,
D. I., and Senter, P. D. (2003) Development of potent monoclonal
antibody auristatin conjugates for cancer therapy. Nat. Biotechnol.
21, 778–784.
(2) Jaracz, S., Chen, J., Kuznetsova, L. V., and Ojima, I. (2005)
Recent advances in tumor- targeting drug conjugates. Biorg. Med.
Chem. 13, 5043–5054.
(3) Temming, K., Meyer, D. L., Zabinski, R., Dijkers, E. C. F.,
Poelstra, K., Molema, G., and Kok, R. J. (2006) Evaluation of
RGD-targeted albumin carriers for specific delivery of auristatin
E to tumor blood vessels. Bioconjugate Chem 17, 1385–1394.
(4) Chari, R. V. J. (2008) Targeted cancer therapy: conferring
specificity to cytotoxic drugs. Acc. Chem. Res. 41, 98–107.
(5) Ducry, L., and Stump, B. (2010) Antibody-drug conjugates:
linking cytotoxic payloads to monoclonal antibodies. Bioconju-
gate Chem. 21, 5–13.
(6) Hamman, P. R., Hinman, L. M., Hollander, I., Beyer, C. F.,
Lindh, D., Holcomb, R., Hallett, W., Tsou, H. R., Upeslacis, J.,
Shochat, D., Mountain, A., Flowers, D. A., and Bernstein, I.
(2002) Gentuzumab ozogamicin: a potent and selective anti-
CD33 antibody-calicheamicin conjugate for treatment of acute
myeloid leukemia. Bioconjugate Chem 13, 47–58.
(7) Okarvi, S. M. (2004) Peptide-based radiopharmaceutical; future
tools for diagnostic imaging of cancer and other diseases. Med.
Res. ReV. 24, 357–397.
(8) Zaccaro, L., del Gatto, A., Pedone, C., and Saviano, M. (2009)
Peptides for tumor therapy and diagnosis: current status and
future directions. Curr. Med. Chem. 16, 780–795.
(26) Dubowchik, G. M., and Firestone, R. A. (1998) Cathepsin
B-sensitive dipeptide prodrugs. 1. A model study of structural
requirements for efficient release of Doxorubicin. Biorg. Med-
.Chem. Lett. 8, 3341–3346.
(27) Dubowchik, G. M., Firestone, R. A., Padilla, L., Willner, D.,
Hofstead, S. J., Mosure, K., Knipe, J. O., Lash, S. J., and Trail,
P. A. (2002) Cathepsin B-labile dipeptide linkers for lysosomal
release of doxorubicin from internalizing immunoconjugates:
model studies of enzymatic drug release and antigen-specific in
vitro antitumor activity. Bioconjugate Chem. 13, 858–869.
(9) Cristofanilli, M., Charnsangavej, C., and Hortobagyi, G. N.
(2002) Angiogenesis modulation in cancer research: novel
clinical approaches. Nat. ReV. Drug DiscoVery 1, 415–426.
(10) Jain, R. K. (2003) Molecular regulation of vessel maturation.
Nat. Med. 9, 685–693.
(11) Ruoslathi, E., and Pierschbacher, M. D. (1987) New perspec-
tive in cell adhesion RGD and integrins. Science 238, 491–497.
(12) Pasqualini, R., Koivunen, E., and Ruoslathi, E. (1997) R
v
integrins as receptors for tumor targeting by circulating ligands.
Nat. Biotechnol. 15, 542–546.

RGD Peptide-Camptothecin Conjugates
Bioconjugate Chem., Vol. 21, No. 11, 2010 1967
(28) Shiose, Y., Kuga, H., Ohki, H., Ikeda, M., Yamashita, F., and
Hashida, M. (2009) Systematic research of peptide spacers
controlling drug release from macromolecular prodrugs system,
carboxymethyldextran-polyalcohol-peptide drug conjugates. Bio-
conjugate Chem. 20, 60–70.
(31) Carl, P. L., Chakravarty, P. K., and Katzenellenbogen, J. A.
(1980) A novel connector linkage applicable in prodrug design.
J. Med. Chem. 24, 479–480.
(32) Wang, P., Layfield, R., Landon, M., Mayer, R. J., and Ramage,
R. (1998) Transfer active ester condensation: A novel technique
for peptide segment coupling. Tetrahedron Lett. 85, 8711–8714.
(33) Kolb, H. C., and Sharpless, K. B. (2003) The growing impact
of click chemistry on drug discovery. Drug DiscoVery Today. 8,
1128–1137.
(29) Haubner, R., Finsinger, D., and Kessler, H. (1997) Stere-
oisomeric peptide libraries and peptidomimetics for designing
selective inhibitors of the Rvꢀ3 integrin for a new cancer
therapy. Angew. Chem. 36, 1374–1389.
(34) Kitov, P. I., and Bundle, D. R. (2003) On the nature of
multivalency effect: a thermodynamic model. J. Am. Chem. Soc.
125, 16271–16284.
(30) Dal Pozzo, A., Penco, S., Giannini, G., Tinti, O., Pisano,
C., Vesci and L., Ni, M. H. Cyclopeptide derivatives
with anti-integrin activity, EP 1 751 176 B1, October 3,
2010.
BC100097R