Design, synthesis, and evaluation of original carriers for targeting vascular endothelial growth factor receptor interactions

Article abstract of DOI:10.1007/s11095-005-5265-9

Purpose. Angiogenesis is a key event in tumor growth and metastasis, chronic inflammatory disease, and cardiovascular disease. It is controlled by positive and negative regulators, which include vascular endothelial growth factor (VEGF) as the most active

Full text of DOI:10.1007/s11095-005-5265-9

#
Pharmaceutical Research, Vol. 22, No. 8, August 2005 ( 2005)
DOI: 10.1007/s11095-005-5265-9
Short Communication
Design, Synthesis, and Evaluation of Original Carriers for Targeting Vascular
Endothelial Growth Factor Receptor Interactions
Mario Gonc¸alves,1,2 Karine Estieu-Gionnet,2,3 Thomas Berthelot,2 Georges Laı¨n,2 Mireille Bayle,2
Xavier Canron,3 Natacha Betz,1 Andreas Bikfalvi,3 and Ge´rard De´le´ris2,4
Received November 19, 2004; accepted March 17, 2005
Purpose. Angiogenesis is a key event in tumor growth and metastasis, chronic inflammatory disease, and
cardiovascular disease. It is controlled by positive and negative regulators, which include vascular
endothelial growth factor (VEGF) as the most active of these. VEGF/VEGF receptors are important
targets not only for therapy but also for imaging. Based on the structural study of VEGF, we developed a
novel cyclopeptide (cyclo-VEGI) that exhibits powerful antitumor properties. We herein report the design
of novel molecules derived from cyclo-VEGI as potential targeting agents in cancer and other
angiogenesis-related diseases.
Methods. We performed selective chemical modification of the most active VEGF-derived cyclopeptide
(cyclo-VEGI). Original hydrophilic linkers were synthesized and coupled to cyclo-VEGI. These
reactions provide nanocarriers for delivery. The inhibitory effect of the different compounds on VEGF
binding was evaluated in competition assays with 125I-VEGF. A fluorescent cyclo-VEGI peptide was
synthezised to assess direct binding and internalization of cyclo-VEGI.
Results. Chemical modifications of cyclo-VEGI do not diminish the biological activity of cyclo-VEGI as
measured in competition assays; in fact, it is even increased. Moreover there is a strong cellular accumu-
lation of the fluorescent-labeled cyclo-VEGI. Conjugates synthesized in this study may be useful leads to
design delivery systems for targeting approaches in cancer and other angiogenesis-related diseases.
Conclusion. The modified cyclo-VEGIs may have a wide range of applications and represent a useful
tool to develop delivery/carrier systems for therapeutic targeting or imaging.
KEY WORDS: angiogenesis; chemotherapy; cyclopeptide modification; imaging; receptors targeting;
solid-phase peptide synthesis; vascular endothelial growth factor.
1 CEA Saclay, DSM/DRECAM/LSI/LPI, 91191, Gif-sur-Yvette INTRODUCTION
Cedex, France.
2 CNRS UMR5084-Chimie Bio-Organic Chemistry Group, Univer-
Angiogenesis is a process by which new capillaries
site´ Victor Segalen Bordeaux 2, 146 rue Le´o Saignat, F-33076
sprout from preexisting blood vessels (1,2). It is involved
Bordeaux Cedex, France.
both in physiological and pathological phenomena. Physiolog-
3 INSERM E0113-Molecular Angiogenesis Laboratory, Universite´
ical angiogenesis occurs in the female reproductive system
Bordeaux 1, 33405 Talence, France.
(3) such as during wound healing (4,5). In contrast, patho-
4 To whom correspondence should be addressed. (e-mail: gerard.
logical neoangiogenesis occurs in a number of diseases such
as cancer, mainly solid tumors and metastasis, or rheumatoid
arthritis (6Y8). Several growth factors have been proved to
deleris@u-bordeaux2.fr)
ABBREVIATIONS: Abs301, absorbance at 301 nm; AcOH, acetic
acid; All, allyl; Boc, tert-butoxycarbonyl; CHO, Chinese hamster
ovary; cyclo-VEGI, cyclo-vascular endothelial growth inhibitor; Da, regulate these phenomena. Among these, vascular endothe-
dalton; DMF, dimethylformamide; Flt-1, fms-like tyrosine kinase-1;
Fmoc, 9-fluorenylmethoxycarbonyl; HBTU, 2-(1H-benzotriazol-
1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate; HOBt,
hydroxybenzotriazol; KDR, kinase domain receptor; MALDI,
matrix-assisted laser desorption/ionization; MsCl, mesyl chloride;
NMM, N-methylmorpholine; NMR, nuclear magnetic resonance;
Pbf, 2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl; PhtNK,
potassium phtalimidate; PVDF, poly(vinylidene difluoride); SPPS,
solid-phase peptidic synthesis; TEG, triethyleneglycol; THF,
tetrahydrofuran; Tis, triisopropylsilane; TsCl, tosyl chloride; Trt,
trityl; VEGF, vascular endothelial growth factor.
lial growth factor (VEGF) is a key regulator of both phy-
siological and pathological angiogenesis (9Y13). Inhibition of
VEGF activity using various inhibitors results in suppression
of tumor growth in vivo (14Y17). VEGF receptors are also
overexpressed in ischemic tissues (18). VEGF exists as dif-
ferent spliced forms, the most important being VEGF165. Its
signaling activity is mainly mediated by two tyrosine kinase
receptors, VEGFR-1 (Flt-1) and VEGFR-2 (KDR). VEGFY
KDR interaction has been demonstrated as the signaling
pathway leading to tumoral neoangiogenesis (19,20). Domi-
#
1411
0724-8741/05/0800-1411/0
2005 Springer Science + Business Media, Inc.

1412
Gonc¸alves et al.
nant regions for VEGF binding to KDR have been investi- (18Y36-C). The beam diameter at the sample location was 6
gated by structural studies and directed mutagenesis (21Y23). mm. In all experiments, a 2.0 cmj1 resolution was used and
They highlight that Arg82, Lys84, and His86 are the key res- acquisitions were performed using 32 scans in transmittance.
idues for binding of VEGF to KDR; they are clustered within
The N-9-fluorenylmethoxycarbonyl (Fmoc)-protected
region 79Y93 which forms a b-hairpin (21). Based on these natural amino acids and 2-chlorotrityl chloride resin were
structural studies, we designed a series of linear and cyclic purchased from Advanced Chemtech. All trifunctional amino
peptides aimed to inhibit the VEGFYKDR binding. As acids were suitably protected. The a-carboxyl group of glu-
described in previous reports (24Y26) several cyclopeptides tamic acid was protected as allyl (All) group. The (-amino
exhibited a good affinity for KDR receptor. All cyclopeptides group of lysine was protected as the (tert-butoxy)carbonyl
which presented good affinities also inhibited endothelial cell (Boc) group. The histidine imidazole ring and the amide
proliferation and in vivo angiogenesis inhibition on CAM group of glutamine were protected with the triphenylmethyl
assays (differentiated chicken chorioallantoic membrane). (trityl, Trt) group (27,28) and the guanidino function of argi-
One of most potent among these original derivatives is a nine as 2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl
17-mer cyclopeptide we named cyclo-VEGI (for vascular (Pbf) group (29,30). A 0.45 M solution of 2-(1H-benzotria-
endothelial growth inhibitor). It comprises VEGF residues zol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate
79Y93, and among them basic residues Arg82, Lys84, and His (HBTU) in N-hydroxybenzotriazole (HOBt) was purchased
86 which have been characterized as the essential residues in from Applied Biosystems. Diisopropylethylamine (DIEA),
VEGFY KDR binding and subsequent cell signaling. This magnesium sulfate, ninhydrine, phosphorus pentoxide, potas-
novel active compound significantly decreases the growth of sium hydroxide, potassium phthalimidate (PhtNK), sodium
established intracranial human glioma in nude and syngenic azide, sodium hydrogencarbonate (NaHCO3), thioanisole, tri-
mice and improves survival without side effects (24,26). fluoroacetic acid (TFA), triisopropylsilane (Tis), and tri-
Preliminary results also indicate a highly synergistic effect phenylphosphine (PPh3) were purchased from Aldrich.
of cyclo-VEGI with alpha radioimmunotherapy.
Diethyl ether, N-methylmorpholine (NMM), piperidine, so-
Targeting molecules to neoangiogenic areas is of out- dium hydroxide, sodium diethyldithiocarbamate, and triethy-
standing importance. Anticancer drugs may be delivered to lamine were purchased from Avocado. Dimethylformamide
neoangiogenic sites as well as radioisotopes for therapy or (DMF) and hydrazine hydrate were purchased from Acros.
imaging. In addition to its role in tumors, imaging is im- Acetic acid (AcOH), benzotriazol-1-yloxytris(pyrrolidino)-
portant in cerebral infarct because penumbra areas over- phosphonium hexafluorophosphate (PyBOP), hydrochloride
express VEGF receptors that may be specifically targeted. acid, mesyl chloride (MsCl), tetrakis(triphenylphophine)pal-
This would be of benefit since functional magnetic resonance ladium [Pd(PPh3)4], and tosyl chloride (TsCl) were pur-
imaging (MRI) is inadequate owing to blood flow alterations. chased from Lancaster. Ammonia solution and phenol were
These data encouraged us to perform a series of chem- purchased from Prolabo (France). Acetonitrile (CH3CN),
ical modifications of cyclo-VEGI to further improve these dichloromethane, ethanol, ethyl acetate, n-hexane, and me-
substrates for therapy or imaging. The design of scaffolds re- thanol were purchased from J.T. Baker.
quires a specific strategy for synthesis to protect critical pep-
tide side groups responsible for the biological activity of
Binding Assays
cyclo-VEGI. We herein report our results along with a pre-
liminary in vitro evaluation of variously decorated cyclo-
VEGI-derived peptides.
VEGF165 was labeled with 125I-Na using Iodogen (Pierce,
Rockford, IL, USA) as coupling agent according to the
manufacturer’s instructions. CHO-KDR cells or bovine
aortic endothelial cells (BAE) were seeded at 2.5 Â 105 den-
MATERIALS AND METHODS
sity in gelatin-coated six-well plates and cultured in complete
medium for 2 days. Cells were washed twice with ice-cold
General Information
phosphate-buffered saline (PBS) and incubated with 10 ng/ml
of 125I-VEGF and peptides at indicated concentrations in
Mass spectra were run using a matrix-assisted laser de-
binding medium [Dulbecco’s modified Eagle’s medium
sorption ionization-time of flight (MALDI-TOF) Reflex III
(DMEM); 20 nM 4-(2-hydroxyethyl)-1-piperazineethanesul-
Bruker apparatus for MALDI spectra and a Micromas
fonic acid (HEPES), pH 7.4; 0.15% gelatin] on a shaker at
VGAutospec-Q for high-resolution mass spectra (HRMS).
4-C. After 2 h, cells were washed three times with PBS and
solubilized by the addition of 2% Triton, 10% glycerol, and
For the nuclear magnetic resonance (NMR) spectra a Bruker
Avance 300 was used. Chemicals shifts are reported in parts
per million relative to tetramethylsilane as an internal
standard (in NMR description, s = singlet, d = doublet, t =
1 mg/ml of bovine serum albumin prior to g-counting. Each
condition was tested in duplicate and repeated at least two
times. Data are expressed as percentage of total specific
triplet, q = quartet, m = multiplet, br = broad peak). Ultra-
radioactivity.
violet (UV) measurments were taken on a Genesys 5
Spectronic Instruments spectrophotometer using a quartz cell
with 1-cm path length. For column chromatography 63Y200 Targeting of Coumarin-Labeled Cyclo-VEGI Peptides
mesh silica gel 60 (VWR International) was used as the to Endothelial Cells in Vitro
stationary phase. Infrared (IR) spectra were recorded on a
Bruker Tensor 27 (Germany, Ettlingen) spectrometer
BAEs were grown on Labtek chamber slides (Nunc,
equipped with an HTS-XT autosampler, a Globar (MIR) Naperville, USA) for 2 days. Pre´confluent BAEs were
source (7 V), a KBr beam splitter, and a DTGS/B detector incubated for different time intervals with coumarin-cyclo-

Original Carriers for Targeting VEGF Receptor Interactions
1413
VEGI peptides or with uncoupled coumarin as a control. with 40 ml of 1 M HOBt in NMP, 40 ml of NMP, and 40 ml
After specified time intervals, cells were washed three times of CH2Cl2, and was dried overnight under high vacuum.
with PBS and fixed with 1% glutaraldehyde. Cellular loc- Peptidyl resin was weighed (1.11 g, 0.20 mmol) and mixed
alization was analyzed with a fluorescence confocal micro- with a solution of PyBOP (312.18 mg, 0.60 mmol), HOBt
scope (Zeiss, LSM510 Meta). Experiments were repeated two (81.07 mg, 0.60 mmol), and DIEA (207.63 ml, 1.2 mmol) in 20
times with identical results.
ml of NMP. The mixture was swelled at room temperature
until Kaiser’s test was negative. The peptidyl resin was
washed with 50 ml of each NMP, CH2Cl2, and MeOH and
was dried under high vacuum. Final cleavage of cyclo-VEGI
from the resin without loss of any side-chain protecting group
was performed with a solution of 1% TFA in CH2Cl2 (10 ml/
g of resin). Cyclopeptidyl resin was weighed (1.02 g), mixed
with dilute TFA, and shaken for 2 min. Then the solution was
filtered and the filtrate was collected in a flask containing a
solution of 10% pyridine in MeOH (2 ml/10 ml of 1% TFA).
It was repeated 10 times and cyclopeptidyl resin was washed
with 3 Â 30 ml of CH2Cl2, 3 Â 30 ml of MeOH, 2 Â 30 ml of
CH2Cl2 and 3 Â 30 ml of MeOH. Then filtrate was
evaporated under reduced pressure to 5% of the volume.
Thereafter, 40 ml of cold water were added to the residue to
aid precipitation of the product which was subsequently
isolated by filtration through a sintered glass funnel. Product
was washed three times with fresh water and dried in a
dessicator under vacuum over KOH, and later over P2O5.
The product was analyzed by reverse-phase HPLC under
the following conditions: eluant A, 0.05% TFA in water;
eluant B, 0.05% TFA in CH3CNYH2O (70/30); gradient,
50Y100% over 30 min; flow rate, 1 ml/min; detector 214
nm. Then 580 mg (65%) of protected cyclo-VEGI were
obtained as a white solid without further purification. A
portion of the product was treated with 0.75 g of phenol in a
TISYthioanisoleYH2OYTFA solution (1:2:2:40) for 3 h at
room temperature. The product was precipitated from cold
diethyl ether and filtered. Then MALDI mass spectrometry
analysis gave the expected molecular mass of deprotected
cyclo-VEGI (theoretical value: 1998.36 Da; experimental
value: 1998.04 Da).
Anchoring of N-(9-Fluorenylmethoxycarbonyl)-Glutamic
Acid Allyl Ester to the 2-Chlorotrityl Chloride Resin
This was accomplished by a general procedure recom-
mended by Novabiochem. Fmoc-Glu-OAll (638.2 mg, 1.56
mmol) was dissolved in 10 ml of CH2Cl2 with DIEA (1.08 ml,
6.29 mmol). The solution was added to the 2-chlorotrityl
chloride resin and was swollen for 2 h at room temperature.
At the end of this time, the resin was washed with 3 Â 10 ml
of CH2Cl2YMeOHYDIEA (17:2:1), followed by 3 Â 10 ml of
CH2Cl2, 2 Â 10 ml of DMF, and 2 Â 10 ml of CH2Cl2. Then
the resin was dried in vacuo over KOH. The substitution
level was determined spectrophotometrically by Fmoc cleav-
age (31). Fmoc-Glu(resin)-OAll (5.05 mg) was introduced
into a test tube and a solution of 20% piperidine in DMF was
added (0.5 ml). 20% piperidine in DMF (0.5 mL) was also
added to an empty test tube to serve as a blank. Over the
next 15 min, the test tube with the resin was swirled two or
three times to make sure all the resin had come in contact
with the piperidine solution. DMF was added to both tubes
to reach a 50-ml volume. The blank was used to zero the UV
spectrophotometer at 301 nm. The absorbance of the solution
is 0.346. The substitution level was calculated from the for-
mula: Abs301 Â Vol (ml)/[7800 Â m (g)] and was determined
to be 0.44 mmol/g.
Peptide Synthesis
Cyclo-VEGI (D-FPQIMRIKPHQGQHIGE) was syn-
thesized by Fmoc/t-Bu batch solid-phase synthesis on an
Applied Biosystems 433A automated peptide synthesizer. 1-Allyloxy-8-Hydroxy-3,6-Dioxaoctane ( 2 )
Preloaded Fmoc-Glu(2-ClTrt resin)-OAll was used for the
linear chain assembly. Subsequent Fmoc amino acids were
Sodium hydride (3 g, 0.08 mol) was added portionwise to
coupled using a fourfold excess of amino acids activated as a solution of triethyleneglycol (20 ml, 0.15 mol) in dry THF
HOBt ester by means of a 0.45 M HBTUYHOBt solution. (50 ml). The solution was stirred at room temperature for 15
Removal of the allyl protecting group was performed be- min. Then allyl bromide (4.3 ml, 0.05 mol) was added and
fore N-terminal Fmoc deprotection. The peptidyl resin was stirring was maintained at room temperature for 1 h. Then
weighed (1.59 g, 0.27 mmol) and dried at 40-C under high solvent was evaporated under reduced pressure and the
vacuum. The reaction vessel was flushed with a stream of residue was dissolved in dichloromethane. The solution was
argon. Pd(PPh3)4 (965.83 mg, 0.83 mmol) was dissolved in a filtrated and washed with brine (40 ml). The aqueous phase
solution of CHCl3YAcOHYNMM (37:2:1) (15 ml/g of resin) was extracted with dichloromethane (2 Â 100 ml). The
by bubbling a stream of argon through the solution. There- organic phase was dried over MgSO4 and solvent were
after this solution was added to the peptidyl resin under evaporated under reduced pressure. The resulting yellowish
argon and was swelled manually every 15 min for 2 h at room oil was purified by chromatography (n-hexaneYethyl acetate
temperature. Then the peptidyl resin was filtered and washed 70:30 and then 50:50). Thereafter evaporation of the solvent
consecutively with 50 ml of 0.5% DIEAYNMP [N-methyl allowed to obtain 6.93 g (73%) of compound 2.
pyrrolidone] (% v/v), 50 ml of 0.5% sodium diethyldithio-
carbamateYNMP (% w/w), 50 ml of CH2Cl2, 50 ml of 1 M 2H, CH2CH=CH2, JH-H=5.61 Hz); 5.94Y5.84 (m, 1H,
HOBt in NMP, 50 ml of NMP, and 50 ml of CH2Cl2. The CH=CH2); 5.285.14 (m, 2H, CH=CH2).
1H NMR (CDCl3) d ppm: 3.53Y3.76 (m, 12H); 3.99 (d,
3
resin was dried under high vacuum over KOH overnight.
13C NMR (CDCl3) d ppm: 61.6 (CH2OH); 70.2Y70.5
Fmoc removal was achieved with a solution of 20% (CH2O); 72.2 (CH2CH=CH2); 72.5 (CH2CH2OH); 117.1
piperidine in NMP. The peptidyl resin was newly washed (CH=CH2); 134.6 (CH=CH2).

1414
Gonc¸alves et al.
3
1-[(p-Toluenesulphonyl)oxy]-9-Allyl-3,6,9-Trioxaoctane ( 3 )
1H NMR (CDCl3) d ppm: 2.82 (t, 2H, CH2NH2, JH-H
=
3
5.3 Hz); 3.43 (t, 2H CH2CH2NH2, JH-H = 5.3 Hz); 3.54Y3.68
A solution of compound 2 (2.14 g, 11.3 mmol) in distilled (m, 10H, YCH2YCH2YOY); 4.01 (d, 2H, CH2CH=CH2, 3JH-H
=
triethylamine (30 ml) was cooled to 0-C. Then tosyl chloride 5.61 Hz); 5.14Y5.28 (m, 2H, CH=CH2); 5.84Y5.94 (m, 1H,
was added (3.22 g, 16.9 mmol) to the medium and was stirred CH=CH2).
at room temperature for 24 h. Then the reaction mixture was
13C NMR (CDCl3) d ppm: 40.2 (CH2NH2); 68.5
poured into ice and ethyl acetate was added (100 ml). The ( C H 2 C H 2 N H 2 ) , 6 8 . 7 Y7 1 . 2 ( C H 2 YC H 2 YO Y a n d
aqueous phase was extracted with ethyl acetate (2 Â 100 ml). CH2CH=CH2); 116.2 (CH=CH2); 133.9 (CH=CH2).
The combined organic layers were washed with 2% AcOH
and H2O, and then dried over MgSO4. The solvent was re-
moved under vacuum to give a yellowish oil that was purified
by chromatography (n-hexane/ethyl acetate, 75:25, 50:50, and
100% ethyl acetate). Then 3.1 g of compound 3 (80%) were
obtained as a colorless oil.
IR (cmj1): 2908; 2869; 1674; 1116; 997; 929.
1-[(p-Toluenesulphonyl)oxy]-11-Hydroxy-3,6,9-
Trioxaundecane ( 7 )
1H NMR (CDCl3) d ppm: 2.44 (s, 3H, YCH3); 3.66Y3.51
To a solution of tetraethyleneglycol (165 ml; 950 mmol)
in CH2Cl2 (425 ml) under nitrogen atmosphere was added
triethylamine (20 ml, 142.5 mmol). The solution was cooled
at 0-C and then tosyl chloride was added (18.1 g; 95 mmol).
Then the solution was heated at room temperature and was
stirred for 20 h. Thereafter the solvent was evaporated under
reduced pressure and the resulting residue was dissolved in
dichloromethane (200 ml). This organic phase was washed
with H2O (3 Â 300 ml) and the combined aqueous layers
were extracted with dichloromethane (3 Â 200 ml). The
combined organic layers were washed with 2% AcOH and
H2O, and then dried over MgSO4. After filtration the solvent
was evaporated under reduced pressure to yield a colorless
oil which was purified by silica gel chromatography (n-hex-
ane/ethyl acetate, 70:30 to 50:50) to give 28.18 g (85%) of
compound 7.
(m, 10H, YCH2YCH2YOY); 4.01 (d, 2H, CH2CH=CH2, 3JH-H
=
3
5.61 Hz); 4.14 (t, 2H, YCH2YOTs, JH-H = 4.7 Hz); 5.14Y5.28
(m, 2H, CH=CH2); 5.84Y5.94 (m, 1H, CH=CH2); 7.33 (d, 2H,
3
3
YCHArY, JH-H = 8.1 Hz); 7.77 (d, 2H, YCHArY, JH-H = 8.1
Hz)
13C NMR (CDCl3) d ppm: 21.2 (YCH3); 68.2Y68.9
(YCH2YCH2YOY); 70.1Y70.2 (CH2CH=CH2, CH2CH2-OTs);
71.7 (CH2YOTs); 116.6 (CH=CH2); 127.5Y129.5 (YCHArY);
132.5 (YCArYSO2Y); 134.3 (CH=CH2); 144.5 (YCArYCH3).
1-Phthalimido-9-Allyl-3,6,9-Trioxaoctane ( 4 )
To a solution of compound 3 (3.1 g, 9 mmol) in distilled
DMF (100 ml) was added potassium phthalimidate (1.83 g,
9.9 mmol). The solution was heated at 120-C for 12 h and
then the solution was cooled to room temperature. The sol-
vent was evaporated under reduced pressure and the result-
ing residue was dissolved in CH2Cl2 (100 ml). H20 (100 ml)
was added. The aqueous phase was extracted three times
with CH2Cl2 (3 Â 100 mL). The combined organic layers
were washed with H2O (3 Â 100 ml) and then dried over
MgSO4. Filtration and evaporation of the solvent made it
possible to obtain 2.73 g of compound 4 (94%) without
further purification.
1H NMR (CDCl3) d ppm: 2.43 (s, 3H, YCH3); 2.68 (s, 1H,
YOH); 3.58Y3.68 (m, 14H, YCH2YCH2YOY); 4.14 (t, 2H,
3
3
YCH2YOTs, JH-H = 4.7 Hz); 7.33 (d, 2H, YCHArY, JH-H
=
3
8.1 Hz); 7.77 (d, 2H, YCHArY, JH-H = 8.1 Hz)
13C NMR (CDCl3) d ppm: 23.0 (YCH3); 64.9
(YCH2YOH); 68.7Y71.9 (YCH2YCH2YOY); 73.8; 128.2Y132.3
(YCHArY); 134.1 (YCArYSO2Y); 146.2 (YCArYCH3).
IR (cmj1): 3422; 2915; 2876; 1189.
1H NMR (CDCl3) d ppm: 3.34Y3.47 (m, 10H,
3
1-Azido-11-Hydroxy-3,6,9-Trioxaundecane ( 8 )
YCH2YCH2YOY); 3.52 (t, 2H, CH2YNPht, JH-H = 5.76 Hz);
3
3.67 (t, 2H, CH2CH2YNPht, JH-H = 5.85 Hz); 3.77 (d, 2H,
CH2CH=CH2, 3JH-H = 5.61 Hz); 4.91Y5.09 (m, 1H, CH=CH2);
5.57Y5.76Y (m, 2H, CH=CH2); 7.48Y7.55 (m, 2H, CHAr);
7.57Y7.63 (m, 2H, CHAr).
To a solution of compound 7 (14 g, 40.18 mmol) in ab-
solute ethanol (200 ml) was added sodium azide (5.22 g,
80.36 mmol). The solution was stirred at 85-C for 7 h.
Thereafter the solution was cooled to room temperature. To
the resulting residue were added CH2Cl2 (200 ml) and H20
(200 ml). The organic phase was washed with H2O (3 Â 200
ml) and the combined aqueous layers were extracted with
CH2Cl2 (3 Â 200 ml). The solvent was dried over MgSO4,
filtrated, and evaporated under vacuum. The residue was
then purified by silica gel chromatography (n-hexane/ethyl
acetate, 50:50 to 25:75) to give 6.71 g (76%) of compound
8 as a brown oil.
13C NMR (CDCl3) d ppm: 37.1 (CH2YNPht); 67.8
(CH2CH2YNPht); 69.3Y72.0 (YCH2YCH2YOY); 116.9
(CH=CH2); 123.1 (CHAr); 132.0 (CAr); 133.8 (CHAr);
134.6 (CH=CH2); 168.1 (C=O).
IR (cmj1): 2906; 2867; 1774; 1713; 1615; 1107; 1006; 928.
1-Amino-9-Allyl-3,6,9-Trioxaoctane ( 5 )
To a solution of compound 4 (2.73 g, 8.55 mmol) in
1H NMR (CDCl3) d ppm: 2.89 (s, 1H, YOH); 3.31 (br s,
absolute ethanol (100 ml) was added hydrazine hydrate (1.66 2H, YCH2YN3); 3.52 (br s, 2H, YCH2YCH2YN3); 3.59 (m, 12H,
ml, 34.23 mmol). The solution was heated at 80-C for 24 h. YCH2YCH2YOY).
The solvent was evaporated under reduced pressure and the
13C NMR (CDCl3) d ppm: 50.6 (YCH2YN3); 61.7 (YCH2
resulting residue was dissolved in CH2Cl2 (50 ml). After YOH); 70.0Y70.6 (YCH2YCH2YOY and YCH2YCH2YN3); 72.4
filtration the solvent was evaporated and then 1.53 g (95%) (YCH2YCH2YOH).
of compound 5 were obtained as a colorless oil.
IR (cmj1): 3429; 2914; 2875; 2113; 1125.

Original Carriers for Targeting VEGF Receptor Interactions
1415
1-Azido-11-[(Methanesulfonyl)oxy]-3,6,9-
Trioxaundecane ( 9 )
1H NMR (CDCl3, 300 MHz) d: 2.26 (br s, 2H,
YCH2YCH=CH2); 3.31 (br s, 2H, YCH2YN3); 3.45 (br s, 2H,
YCH2YCH2YN3); 3.54Y3.59 (m, 14H, YCH2YCH2YOY et
To a solution of compound 8 (17.39 g; 79.32 mmol) in YCH2YCH2YCH=CH2); 4.93Y5.03 (m, 2H, YCH=CH2);
CH2Cl2 (500 ml) was added triethylamine (30 ml, 214.16 5.70Y5.76 (m, 1H, YCH=CH2).
mmol). The solution was cooled to 0-C and then mesyl chlo-
13C NMR (CDCl3, 75 MHz) d: 33.9 (YCH2YCH=CH2);
ride (12.3 ml; 158.65 mmol) was added. The reaction mixture 50.5 (YCH2YN3); 69.8Y70.4 (YCH2YCH2YOY and YCH2Y
was heated at room temperature, stirred for 24 h, and then a CH2YCH=CH2); 116.1 (YCH=CH2); 136.8 (YCH=CH2).
solution of saturated NaHCO3 (200 ml) was added. The
organic layer was washed with H2O (3 Â 100 ml), dried over
MgSO4, and filtrated. The solvent is evaporated under
reduced pressure to give a dark oil that was purified by silica
gel chromatography (n-hexaneYethyl acetate, 10:90). Then
21.2 g (88%) of compound 9 were obtained as a brown oil.
1H NMR (CDCl3, 300 MHz) d: 2.96 (s, 3H, YCH3); 3.28
(br s, 2H, YCH2YN3); 3.51Y3.54 (m, 10H, YCH2YCH2YOY);
3.65 (br s, 2H, YCH2YCH2YOMs); 4.25 (br s, 2H, YCH2Y
OMs).
IR (cmj1): 2911; 2872; 2110; 1125.
1-Amino-10-Azido- 3,6,9-Trioxaundecane ( 12 )
To a solution of compound 10 (3.51 g, 10.09 mmol) in
absolute ethanol (100 ml) was added hydrazine hydrate (1.95
ml, 40.36 mmol) under an inert atmosphere. The reaction
mixture was heated at 80-C for 24 h. The solvent was evap-
orated under reduced pressure and the resulting residue was
diluted with water (50 ml) and concentrated HCl (8 ml) and
heated at reflux for 1 h. The resulting suspension was cooled
to 0-C and the solids were filtered. The aqueous filtrate was
neutralize with 1 M NaOH and then concentrated under
reduced pressure. The residue was diluted with dichloro-
methane (50 ml) washed with 4 M NaOH and dried over
MgSO4. Thereafter the solvent was evaporated to give 2.03 g
13NMR (CDCl3, 75 MHz) d: 37.5 (YCH3); 50.0
(YCH2YN3); 68.3Y69.9 (YCH2YCH2YOY and YCH2YCH2YN3).
IR (cmj1): 2912; 2873; 2110; 1678; 1134.
1-Azido-10-Phthalimido-3,6,9-Trioxaundecane ( 10 )
To a solution of compound 9 (7.88 g, 26.5 mmol) in dry of compound 12 (92%) as a yellowish oil.
DMF (150 ml) was added potassium phthalimidate (5.4 g,
1H NMR (CDCl3, 300 MHz) d: 2.71 (t, 2H, CH2NH2);
29.15 mmol). The solution was heated at 120-C for 12 h, and 3.25 (t, 2H, CH2N3); 3.37 (t, 2H, CH2CH2NH2); 3.48Y3.72 (m,
then the solvent was evaporated under reduced pressure. 10H, CH2CH2O and CH2CH2N3).
CH2Cl2 (100 ml) and H20 (100 ml) were added to the re-
13C NMR (CDCl3, 75 MHz) d: 41.2 (CH2NH2), 50.3
sulting residue. The aqueous layer was extracted with CH2Cl2 (CH2N3), 69.9Y70.3 (YCH2YCH2YOY and CH2CH2N3), 72.8
(3 Â 100 ml) and the combined organic layers were washed (CH2CH2NH2).
with H20 (3 Â 100 ml). The organic phase was dried over
IR (cmj1): 3360; 2909; 2870; 1774; 2110; 1124.
HRMS (FAB+): calculated for C8H19N4O3 (MH)+
MgSO4, filtrated and then the solvent was evaporated under
reduced pressure. The residue was purified by silica gel 219.1457, found 219.1462.
chromatography (n-hexane/ethyl acetate, 80:20) to give 6.47 g
of compound 10 (70%) as a brown oil.
1-Amino-12-[but-3-enyl]-3,6,9,12-Tetraoxaundecane ( 13 )
1H NMR (CDCl3) d ppm: 3.12 (t, 2H, CH2NPht, 3JH-H
5.25 Hz); 3.34Y3.47 (m, 10H, YCH2YCH2YOY); 3.55 (t, 2H,
CH2YN3, 3JH-H = 5.76 Hz); 3.69 (t, 2H, CH2CH2YNPht, 3JH-H
5.85 Hz); 7.48Y7.55 (m, 2H, CHAr); 7.57Y7.63 (m, 2H, CHAr).
13C NMR (CDCl3) d ppm: 36.7 (CH2NPht), 50.1 (CH2N3),
67.3 (CH2CH2NPht and CH2CH2N3), 69.5Y70.1 (CH2O), 123.5
(CHAr), 131.5 (CAr), 134.8 (CHAr), 167.6 (C=O).
=
To a solution of compound 11 (1.02 g; 3.73 mmol) in dry
THF (50 ml) was added triphenylphosphine (1.27 g; 4.7
mmol) under inert atmosphere. Stirring was maintained at
room temperature overnight. The solvent was evaporated
until 10% of the initial content and a solution of 1:1
etherYpetroleum ether was added to the solution which
cooled to 0-C. The solution was filtered to remove precipi-
tated triphenylphosphine oxide. The filtrate was then evap-
orated. Purification of the crude product was achieved by
=
1-Azido-12-[but-3-enyl]-3,6,9,12-Tetraoxaundecane ( 11 )
To a solution of but-3-en-1-ol (1.1 ml; 13.28 mmol) in dry silica gel chromatography (CH2Cl2YMeOH, 50:50). Thereaf-
THF (10 ml), cooled at 0-C, was added sodium hydride (400 ter the solvent was evaporated to give 1.07 g (78%) of
mg; 13.26 mmol) under an inert atmosphere. A solution of compound 13.
compound 9 (1.58 g, 5.31 mmol) in dry THF (5 ml) was
1H NMR (CDCl3, 300 MHz) d: 2.23 (br s, 2H,
added. The reaction mixture was heated at room temperature YCH2YCH=CH2); 2.75 (br s, 2H, YCH2YNH2); 3.39 (br s,
and was stirred overnight. Then a solution of 32% of aqueous 2H, YCH2YCH2YNH2); 3.51Y3.54 (m, 14H, YCH2YCH2YOY
ammonia was added until the reaction mixture was clear. and YCH2YCH2YCH=CH2); 4.89Y4.99 (m, 2H, YCH=CH2);
Thereafter the solvent was evaporated under reduced pres- 5.66Y5.72 (m, 1H, YCH=CH2).
sure. The resulting residue was dissolved in dichloromethane
13C NMR (CDCl3, 75 MHz) d: 33.9 (YCH2YCH=CH2);
(50 ml) and washed with H2O (3 Â 50 ml). The aqueous 41.5 (YCH2YNH2); 70.1Y70.4 (YCH2YCH2YO and YCH2Y
layers were extracted with dichloromethane (3 Â 50 ml). CH2YCH=CH2); 73.2 (YCH2YCH2YNH2); 116.2 (YCH=CH2);
After drying over MgSO4 the solvent was filtrated and eva- 136.7 (YCH=CH2).
porated under vacuum. The residue was then purified by
silica gel chromatography (n-hexaneYethyl acetate, 40:60) to
give 1.07 g (74%) of compound 11.
IR (cmj1): 2907; 2868; 1117.
HRMS (FAB+): calculated for C12H26NO4 (MH)+
248.1861 found 248.1869.

1416
Gonc¸alves et al.
Compound 14
Compound 16
To a solution of cyclo-VEGI (300 mg; 84.2 mmol) in
To a solution of cyclo-VEGI (220.4 mg; 61.8 mmol) in
NMP (0.1 ml/mmol) were added PyBOP (98.3 mg; 189 mmol) NMP (0.1 ml/mmol) were added PyBOP (133.6 mg; 256.7
and HOBt (25.5 mg; 189 mmol). After complete dissolution, mmol) and HOBt (34.7 mg; 256.7 mmol). After complete
DIEA (65.4 ml; 378 mmol) and the reaction mixture was stirred dissolution, DIEA (90 ml; 520.1 mmol) was added and the
over 5 min at room temperature. Then compound 12 (21.2 mg; reaction mixture was stirred over 5 min at room temperature.
81.9 mmol) was added and the mixture was stirred over 24 Then compound 13 (27.4 mg; 110.9 mmol) was added and the
h at room temperature. The solution was concentrated and mixture was stirred over 24 h at room temperature. The
the product was precipitated with cold water, filtered, and solution was concentrated and the product precipitated with
dissolved in a solution of CH3CNYH2O (70/30) with 0.1% cold water, filtered, and dissolved in a solution of CH3CNY
TFA (eluant B) and few drops of THF, and then loaded onto H2O (70:30) with 0.1% TFA (eluant B) and few drops of
a preparative Hibar Purosphere C18 column. The elution was THF. Then it was loaded onto a preparative Hibar Puro-
achieved using the following conditions: eluant A, 0.1% sphere C18 column. The elution was achieved under the
TFA in water; eluant B, 0.1% TFA in CH3CN/H2O (70:30); following conditions: eluant A, 0.1% TFA in water; eluant B,
gradient, 0Y100% B over 35 min; flow rate, 4 ml/min; 0.1% TFA in CH3CN/H2O; gradient 0Y100% B over 25 min;
detector, 214 nm. Then 220.9 mg (95%) of compound 14 flow rate, 4 ml/min; detector, 214 nm. The purified product
were obtained. The purified product was analyzed by HPLC was analyzed by HPLC with a flow rate of 1 ml/min and using
with a flow rate of 1 ml/min and using the same other the same other conditions as described earlier. The MALDI
conditions as described earlier. The MALDI mass spectrom- mass spectrometry analysis of the purified product was un-
etry analysis of the purified product was unsuccessful, successful, certainly because of the presence of the protecting
certainly because of the presence of the protecting groups groups which render the ionization of compound 16 difficult.
which render the ionization of compound 14 difficult. A A portion of the product was treated with 0.75 g of phenol in
portion of the product was treated with 0.75 g of phenol in a a TISYthioanisoleYH2OYTFA solution (1:2:2:40) for 3 h at
TISYthioanisoleYH2OYTFA solution (1:2:2:40) for 3 h at room temperature. The product was precipitated from cold
room temperature. The product was precipitated from cold diethyl ether and filtered. Then MALDI mass spectrometry
diethyl ether and filtered. Then MALDI mass spectrometry analysis revealed that the product was a mixture of cyclo-
analysis gave the expected result (theoretical value: 2198.60 VEGI and compound 16 (theoretical values: 2227.67 Da and
Da; experimental value: 2198.28 Da).
1998.36; experimental values: 2227.35 Da and 1998.04 Da)
which were unseparable under the protected form.
Compound 17
Compound 15
Compound 14 (240.7 mg; 63 mmol), triphenylphosphine
(18.2 mg; 69 mmol), and water (1.7 ml; 94 mmol) were mixed
with 5 ml of THF. After the solution was stirred for 48 h, it
was concentrated and the product was precipitated with cold
water and filtered. The product was dissolved in a solution of
CH3CNYH2O (70:30) with 0.05% TFA (eluant B) and a few
drops of THF, and then loaded onto a preparative Hibar
Purosphere column C18. The elution was achieved using the
following conditions: eluant A, 0.05% TFA in water; eluant
B, 0.05% TFA in CH3CNYH2O (70/30); gradient, 0Y100%
over 35 min; flow rate, 4 ml/min; detector, 214 nm. Then
127.1 mg (54%) of compound 14 were obtained. The purified
product was analyzed by HPLC with a flow rate of 1 ml/min
and under the same conditions as described earlier. The
MALDI mass spectrometry analysis of the purified product
was unsuccessful certainly owing to the presence of the
protecting groups which render the ionization of compound
17 difficult. A portion of the product was treated with 0.75 g
of phenol in a TISYthioanisoleYH2OYTFA solution (1:2:2:40)
for 3 h at room temperature. The product was precipitated
from cold diethyl ether and filtered. Then MALDI mass
spectrometry analysis gave the expected result (theoretical
value: 2172.60 Da; experimental value: 2172.85 Da).
To a solution of cyclo-VEGI (733.5 mg; 0.2 mmol) in
NMP (0.1 ml/mmol) were added PyBOP (325.1 mg; 0.62
mmol) and HOBt (84.4 mg; 0.62 mmol). After complete
dissolution, DIEA (166.8 ml; 1.23 mmol) and the reaction
mixture was stirred over 5 min at room temperature. Then
compound 5 (50 mg; 0.26 mmol) was added and the mix-
ture was stirred over 24 h at room temperature. The
solution was concentrated and the product was precipitat-
ed with cold water, filtered, and dissolved in a solution of
CH3CN/H2O (70:30) with 0.1% TFA (eluant B) and few
drops of THF, and then loaded onto a preparative Hibar
Purosphere C18 column. The elution was achieved using the
following conditions: eluant A, 0.1% TFA in water; eluant
B, 0.1% TFA in CH3CNYH2O (70:30); gradient, 50Y100% B
over 40 min; flow rate, 4 ml/min; detector, 214 nm. Then
186 mg (94%) of compound 15 were obtained. The purified
product was analyzed by HPLC with a flow rate of 1 ml/min
and using the same other conditions as described earlier.
The MALDI mass spectrometry analysis of the purified
product was unsuccessful certainly because of the presence
of the protecting groups which render the ionization of
compound 15 difficult. A portion of the product was treated
with 0.75 g of phenol in a TISYthioanisoleYH2OYTFA
solution (1:2:2:40) for 3 h at room temperature. The product
was precipitated from cold diethyl ether and filtered. Then Compound 18
MALDI mass spectrometry analysis gave the expected result
(theoretical value: 2169.59 Da; experimental value: 2169.32
Da).
Fully experimental procedure and biological properties
will be developed in a future paper. Analytical HPLC on

Original Carriers for Targeting VEGF Receptor Interactions
1417
Scheme 1. Synthesis of a heterobifunctional linker with a terminal olefin.
Hibar purosphere column C18 show one compound with a chain anchoring of Glu to this resin will allow its protection
retention time of 11.2 min. Elution was achieved using the during SPPS and its selective deprotection during the final
following conditions: eluant A, 0.05% TFA in water; eluant peptidic-resin cleavage performed under mild acidic condi-
B, 0.05% TFA in CH3CNYH2O (70:30); gradient: 10% at 0 tions. We prepared this noncommercially available preloaded
min, 20% at 5 min, 100% at 7 min and over a period of 30 resin by formation of an ester bond between the g-COOH
min; flow rate, 1 ml/min; detector, 410 nm. Then ES mass group of Fmoc-Glu-OAll and the 2-chlorotrityl chloride
spectrometry analysis gave m/e (m+3) 806.8 Da.
resin, without any coupling reagent in the presence of
diisopropylethylamine. Substitution level was determined by
UV quantification at 301 nm (31) and suitable protecting
groups were used for all trifunctional amino acids.
RESULTS AND DISCUSSION
Chemical modifications of cyclo-VEGI must not affect
After SPPS completion, removal of allyl (All) protection
residues Arg82, Lys84, and His86 because they play a central was carried out according to the method of Kates and co-
role in the cyclo-VEGIYKDR interaction. The side chain workers (32) utilizing Pd(PPh3)4YCHCl3YAcOHYNMM. This
group of Glu93, which is the only to be acidic, was selected to reaction was performed under neutral conditions allowing
introduce new molecular decoration into cyclo-VEGI. More- the selective deprotection of the a-COOH group with the
over, Glu93 is at the opposite site of key residues Arg82, linear peptide still anchored to the resin, and thereafter on-
Lys84, and His86 in cyclo-VEGI. Our first goal was to resin cyclization. The cyclic peptide was then cleaved from
propose a synthetic route allowing to obtain a protected the solid support using repeated mild acidic conditions to
cyclopeptide with one deprotected side chain group, by solid- prevent any side-chain deprotection by extended exposure of
phase peptidic synthesis (SPPS), to introduce selectively the protected cyclopeptide to residual TFA. The final product
molecular decoration. To our knowledge this strategic route was purified by reverse-phase HPLC. It was impossible to
had never been described previously. For this purpose, we obtain any analytical result by MALDI mass spectrometry. It
performed the total synthesis of cyclo-VEGI on the extreme- was hypothesized that the presence of the protecting groups
ly acid-labile 2-chlorotrityl chloride resin. The linear precur- and the cyclic structure prevent ionization of the protected
sor peptide of cyclo-VEGI was assembled by standard Fmoc cyclopeptide, and thus MALDI mass spectrometry analysis.
chemistry on a Fmoc-Glu(resin)-OAll carrier. Indeed, side- Nevertheless, a portion of protected cyclo-VEGI was taken
Scheme 2. Synthesis of a common synthon for terminal amine/olefin linkers.

1418
Gonc¸alves et al.
Scheme 3. Synthesis of terminal amine/olefin linker from compound 9.
and deprotected by a mixture of TFA in the presence of suit- grafted polyvinylidene fluoride (PVDF) films (33). The
able scavengers. The free cyclopeptide was purified by reverse- introduction of a linker molecule with an amino function
phase HPLC and showed the expected molecular mass (theo- would enable the grafting of cyclo-VEGI onto polymeric
retical value: 1998.36 Da, experimental value: 1998.04 Da). nanocarriers and therefore allow the design of drug delivery/
Based on the initial loading, overall yield of the supported- carrier systems. Moreover, chemical modifications of cyclo-
solid synthesis of protected cyclo-VEGI was calculated at 65%. VEGI should be possible via linker molecules by amide bond
Our second goal was to determine possible molecular formation with the g-COOH group of Glu93. These linker
decorations of cyclo-VEGI for its use in angiogenesis molecules should also act as spacer arms to minimize steric
imaging or in nanocarrier preparation. We have reported hindrance between cyclo-VEGI and the solid support.
elsewhere the immobilization of linear peptides through their We next designed and synthesized three linkers. Oligo-
terminal amino function, onto poly(acrylic acid) radiation- ethylene glycol derivatives are ideal for this purpose because
Scheme 4. Functionalization of cylo-VEGI.

Original Carriers for Targeting VEGF Receptor Interactions
1419
Fig. 1. Inhibition of 125I-VEGFYKDR binding to Chinese hamster ovary (CHO) cells
expressing KDR by modified cVEGI peptides. Cells were incubated in duplicates with
increasing concentration of peptides in the presence of 10 ng/ml of 125I-VEGF. Binding was
analyzed after 2 h of incubation at 4-C as indicated in materials and methods. Results are
expressed as percentages of specific binding in comparison to control.
they are water soluble and available in a variety of lengths. We alizations such as radiolabeling as we might be afraid of a too
first synthesized a heterobifunctional linker that possesses a high lability of the allyloxy group. This could have a nefast
terminal amino group for its anchoring to cyclo-VEGI, and effect on the functionalization step. The reduction of the azido
a terminal allyl group allowing further modifications (i.e, group of compound 11 was performed with triphenylphos-
radiolabeling). This linker was prepared from commercially phine in THF to give the original linker 13 which was ob-
available triethylene glycol 1 (TEG) (Scheme 1).
tained with an overall yield of 33%. Thereafter cyclo-VEGI
The critical step is the selective mono-allylation of TEG. was functionalized with linkers 5, 12, and 13 (Scheme 4) using
It was successfully performed by using an excess of TEG with a mixture of PyBOP and HOBt with DIEA in THF to give
sodium azide and allyl bromide in dry THF to give compounds 14, 15, and 16.
compound 2 in good yield. Tosylation of compound 2 gave
Compound 17 was obtained from compound 14 by
compound 3 which was subjected to a Gabriel’s amination to reduction of the azido group under mild conditions possible
give the original linker molecule 5 with an overall yied of because of the presence of triphenylphosphine in THF.
65%. Thereafter we synthesized two further linkers from a Compounds 14Y16 were purified by reverse-phase HPLC.
common synthon taking advantage of the procedure de- The presence of all the side-chain protecting groups in cyclo-
scribed by Bertozzi and Bednarski (34) (Scheme 2) yielding a VEGI made the ionization of compounds 14Y16 impossible in
synthon allowing to obtain a linker with a terminal olefin or MALDI mass spectrometry analysis. Therefore, in a fraction
amino group under protected form (azido or phthalimido form). of each compound, all the protecting groups were completely
The critical step is the selective monotosylation of removed to allow MALDI mass spectrometry analysis which
tetraethylene glycol 6. It was performed in triethylamine confirmed the integrity of compounds 14, 15, and 17. MALDI
with tosyl chloride and a large excess of tetraethylene glycol analysis of unprotected compound 16 also showed the
to give compound 7 in good yield. Compound 7 was reacted presence of non-modified cyclo-VEGI. Purification by re-
with sodium azide in DMF or absolute ethanol to provide verse-phase HPLC of protected or unprotected compound 16
compound 8. Mesylation of compound 8 yielded synthon 9 and cyclo-VEGI was unsuccessful owing to similar retention
from which two linker molecules with a terminal olefin or a times.
terminal amino group could be obtained (Scheme 3).
We then investigated the ability of compounds 15 and 17
Compound 9 subjected to a Gabriel’s amination yielded to inhibit the binding of 125I-VEGF to KDR in comparison to
linker 12 with an overall yield of 49%. The azido group nonmodified cyclo-VEGI. This was achieved by investigating
would give the terminal amino group by reduction after
decoration of cyclo-VEGI with linker 12. On the other side,
when compound 9 was subjected to the nucleophilic substi-
tution of its mesylate group by an alcohol with a terminal
olefin it gave compound 11. We chose to replace the allyl
group by a but-1-enyl group to determine what kind of
terminal olefin would be the most convenient for function-
Scheme 5. Structure of compound 18.

1420
Gonc¸alves et al.
Fig. 2. (A, B) Targeting of endothelial cells by coumarin-labeled cyclo-VEGI in vitro. Endothelial cells
were incubated with 1 mM of coumarin-labeled cyclo-VEGI (peptide 18; A) or coumarin only (B) for
2 h. After 2 h, cells were washed three times and fixed with 1% glutaraldehyde. Fluorescence pictures
were taken at 405 nm wavelength.
the capacity of these cyclopeptides to interfere with binding newly synthezised heterobifunctional linkers, two of them
of 125I-labeled VEGF165 to Chinese hamster ovary (CHO) being original. Functionalized cyclo-VEGI conserved a
cells expressing KDR (Fig. 1).
strong inhibitory potency for VEGF-binding to KDR,
As shown in Fig. 1, functionalization of cyclo-VEGI compound 17 for instance being more efficient than non-
does not abrogate its ability to compete with 125I-VEGF modified cyclo-VEGI. Furthermore, functionalized cyclo-
binding. Modified cyclo-VEGIs interfere even more strongly VEGI is targeted to endothelial cells in vitro where it is
in 125I-VEGF binding than nonmodified cyclo-VEGI. (IC50 accumulated after binding in the endosomal compartment.
values of 1, 0.94, 0.16 mM for compound 14, 15, and 17, An important point to be noticed at is that for all previously
respectively). This may be explained by an increase in prepared cyclopeptides binding affinities were well correlat-
basicity of the compounds through blocking the acidic ed with in vitro cellular activity as well as in vivo activity. The
function of the Glu residue by our linkers since basic full biological evaluation is in progress.
residues play a key role in the VEGFYKDR interaction.
These compounds could be useful not only as inhibitors
Compound 17 showed even higher inhibitory activity when or targeting agents in tumors or other angiogenesis-related
compared to the other modified peptides. This may be diseases but also for molecular imaging Chemical conjugation
explained by the presence of another basic residue coming to polymers and nanoparticules as well as extensive biolog-
from the linker. We also tested the effect of coumarin- ical evaluation of the different systems is in progress.
labeled cyclo-VEGI peptide on VEGF binding (compound
18). Cyclo-VEGI was labeled with 7-diethylaminocoumarin-
ACKNOWLEDGMENTS
3-carboxylic acid as previously described (35,36) (compound
18; Scheme 5). Inhibitory activity was similar to that of
compounds 14 and 15 (IC50 = 1 mM for 18).
This work was carried out in the frame of a collaboration
between Comissariat a` l’Energie Atomique (CEA), Bio-
Organic Chemistry Group (LRC DSM-98-15, CNRS
UMR5084), and INSERM E0113. The authors gratefully
acknowledge CEA and Conseil Re´gional d’Aquitaine for
providing financial support through the Ph.D. grant of M. G.
This work was supported by grants from the Ligue Nationale
contre le Cancer (Comite´ de la Gironde, G. D.; Equipela-
bellise´e, A. B.), the Conseil Re´gional d’Aquitaine (G. D., A.
B.), and the European Commission (FP6,LSCH-CT-2003-
503233, STROMA, to A. B.; BThe European Commission is
not liable for any use that may be made of the information
contained^).
Furthermore, we investigated the ability of compound 18
to target endothelial cells in vitro (Fig. 2). Coumarin-labeled
cyclo-VEGI (compound 1 mM) was incubated for 2 h with
BAE cells (Fig. 2). Fluorescence labeling was seen in cells
that were exposed to coumarinYcyclo-VEGI but not in cells
exposed to coumarin only. There was a time-dependent
accumulation of the fluorescent peptide. Fluorescence was
already visible at 30 min and increased significantly up to 2 h.
At 2 h, fluorescence was mainly intracellular and probably
localized in the endosomal compartment.
CONCLUSION
REFERENCES
As an innovative approach toward the preparation of
original molecular nanocarriers, we herein described a new
and convenient strategy for the decoration of cyclo-VEGI.
These derivatives were obtained by an original synthetic
route. Conjugation of cyclopeptide was achieved with three
1. G. Bergers and L. E. Benjamin. Tumorigenesis and the
angiogenic switch. Nat. Rev. Cancer 3:401Y410 (2003).
2. R. Kalluri. Basement membranes: structure, assembly and role
in tumour angiogenesis. Nat. Rev. Cancer 3:422Y433 (2003).

Original Carriers for Targeting VEGF Receptor Interactions
1421
3. L. P. Reynolds, A. T. Grazul-Bilska, and D. A. Redmer.
Angiogenesis in the female reproductive organs: pathological
implications. Int. J. Exp. Pathol. 83:151 (2002).
factor: crystal structure and functional mapping of the kinase
domain receptor binding site. Proc. Natl. Acad. Sci. USA
94:7192Y7197 (1997).
4. S. M. Hyder and G. M. Stancel. Regulation of angiogenic growth 22. C. Wiesman, G. Fuh, H. W. Christinger, C. Eigenbrot, J. A.
˚
Wells, and A. M. de Vos. Crystal structure at 1.7 A resolution
of VEGF in complex with domain 2 of the Flt-1 receptor.
Cell 91:695Y704 (1997).
factors in the female reproductive tract by estrogens and
progestins. Mol. Endocrinol. 13:806Y822 (1999).
5. J. Li, Y.-P. Zang, and R. Kirsner. Angiogenesis in wound repair:
angiogenic growth factors and the extracellular matrix. Micros. 23. B. A. Keyt, H. V. Nguyen, L. T. Berleau, C. M. Duarte, J. Parke,
Res. Tech. 60:107Y114 (2003).
6. J. Folkman. Angiogenesis in cancer, vascular, rheumatoid and
H. Chen, and N. Ferrara. Identification of vascular endothelial
growth factor determinants for binding KDR and FLT-1
receptors. J. Biol. Chem. 271:5638Y5646 (1996).
other disease. Nat. Med. 1:27Y31 (1995).
7. L. Liotta, P. S. Steeg, and W. G. Stetler Stevenson. Cancer 24. L. Bello, V. Lucini, F. Costa, M. Pluderi, C. Giussani, F. Acerbi,
metastasis and angiogenesis: an imbalance of positive and
negative regulation. Cell 64:327Y336 (1991).
8. E. Bodolay, A. E. Koch, J. Kim, and Z. Szegedi. Angiogenesis
and chemokines in rheumatoid arthritis and other systemic
inflammatory rheumatic diseases. J. Cell. Mol. Med. 6:357Y376
(2002).
G. Carrabba, M. Pannacci, D. Caronzolo, S. Grosso, S.
Shinkaruk, F. Colleoni, X. Canron, G. Tomei, G. Deleris, and
A. Bikfalvi. Combinatorial administration of molecules that
simultaneously inhibit angiogenesis and invasion leads to
increased therapeutic efficacy in mouse models of malignant
glioma. Clin. Cancer Res. 10:4527Y4537 (2004).
9. N. Ferrara. Role of vascular endothelial growth factor in the 25. A. Bikfalvi. Recent developments in the inhibition of angiogen-
esis: examples from studies on platelet factor-4 and the VEGF/
regulation of angiogenesis. Kidney Int. 56:794Y814 (1999).
10. H. F. Dvorak, T. M. Sioussat, L. F. Brown, B. Berse, J. A. Nagy,
VEGFR system. Biochem. Pharmacol. 68:1017Y1021 (2004).
A. Sotrel, E. J. Manseau, L. Van de Water, and D. R. Senger. 26. L. Zilberberg, S. Shinkaruk, O. Lequin, B. Rousseau, M.
Distribution of vascular permeability factor (vascular endothe-
lial growth factor) in tumors: concentration in tumor blood
vessels. J. Exp. Med. 174:1275Y1278 (1991).
Hagedorn, F. Costa, D. Caronzolo, M. Balke, X. Canron, O.
Convert, G. Lain, K. Gionnet, M. Goncalves, M. Bayle, L. Bello,
G. Chassaing, G. Deleris, and A. Bikfalvi. Structure and
inhibitory effects on angiogenesis and tumor development of a
new vascular endothelial growth factor. J. Biol. Chem.
278:35564Y35573 (2003).
11. M. Relf, S. Lejeune, P. A. Scott, S. Fox, K. Smith, R. Leek, A.
Moghaddam, R. Whitehouse, R. Bicknell, and A. L. Harris.
Expression of the angiogenic factors vascular endothelial cell
growth factor, acidic and basic fibroblast growth factor, tumor 27. M. Mergler, R. Tanner, J. Costeli, and P. Grogg. Peptide syn-
growth factor beta-1, platelet-derived endothelial cell growth
factor, placenta growth factor, and pleiotrophin in human
primary breast cancer and its relation to angiogenesis. Cancer
Res. 57:963Y969 (1997).
thesis by a combination of solid-phase an solution methods I: a
new very acid-labile anchor group for the solid-phase synthesis
of fully protected fragments. Tertrahedr. Lett. 29:4005Y4008
(1988).
12. N. Ferrara, K. Houck, L. Jakeman, and D. W. Leung. Molecular
and biological properties of the vascular endothelial growth
factor family of proteins. Endocr. Rev. 13:18Y32 (1992).
13. A. Bikfalvi and R. Bicknell. Recent advances in angiogenesis
and vascular targeting. Trends Pharmacol. Sci. 23:576Y582
(2002).
28. M. Mergler, R. Nyfeler, R. Tanner, J. Costelli, and P. Grogg.
Peptide synthesis by a combination of solid-phase and solution
methods II: synthesis of fully protected peptide fragments on
2-methoxy-4-alcoxy-benzyl alcohol resin. Tetrahedr. Lett.
29:4009Y4012 (1988).
29. L. A. Carpino. The 2,2,4,6,7-pentamethyldihydrobenzofuran-5-
sulfonyl group (Pbf) as arginine side chain protectant. Tetrahedr.
Lett. 34:7829 (1993).
14. K. J. Kim, B. Li, J. Winer, M. Armanini, N. Gillett, H. S.
Phillips, and N. Ferrara. Inhibition of vascular endothelial
growth factor-induced angiogenesis suppresses tumour growth 30. C. G. Fields. Minimization of tryptophan alkylation following 9-
in vivo. Nature 362:841Y844 (1993).
fluorenylmethoxycarbonyl solid-phase peptide synthesis. Tetra-
hedr. Lett. 34:6661 (1993).
15. B. Millauer, L. K. Shawver, K. H. Plate, W. Risau, and A.
Ullrich. Glioblastoma growth inhibited in vivo by a dominant-
negative Flk-1 mutant. Nature 367:576Y579 (1994).
16. B. Millauer, M. P. Longhi, K. H. Plate, L. K. Shawver, W. Risau,
A. Ullrich, and L. M. Strawn. Dominant-negative inhibition of
Flk-1 suppresses the growth of many tumor types in vivo. Cancer
Res. 56:1615Y1620 (1996).
31. J. Meienhofer, M. Waki, E. P. Heimer, T. J. Lambros, R. C.
Makofske, and C. D. Chang. Solid phase synthesis without
repetitive acidolysis. Preparation of leucyl-alanyl-glycyl-valine
using 9-fluorenylmethyloxycarbonylamino acids. Int. J. Peptide
Protein Res. 13:35Y42 (1979).
32. S.A. Kates, In Peptides, Chemistry, Structure, Biology, Proc 13th
American Peptide Symposium, ESCOM Leiden, 1994.
17. P. Borgstrom, K. J. Hillan, P. Sriramarao, and N. Ferrara.
Complete inhibition of angiogenesis and growth of microtumors 33. M.-C. Clochard, N. Betz, M. Goncalves, C. Bittencourt, J.-J.
by anti- vascular endothelial growth factor neutralizing antibody:
novel concepts of angiostatic therapy from intravital video-
microscopy. Cancer Res. 56:4032Y4039 (1996).
Pireaux, K. Gionnet, G. De´le´ris, and A. Le Moe¨l. Peptide
immobilization onto radiation grafted PVDF-g-poly(acrylic
acid) films. In press (2004).
18. N. Ferrara and K. Alitalo. Clinical applications of angiogenic
growth factors and their inhibitors. Nat. Med. 5:1359Y1364 (1999).
19. F. Shalaby, J. Rossant, T. P. Yamaguchi, M. Gertsenstein, X. F.
34. C. R. Bertozzi and M. D. Berdnarski. The synthesis of
heterobifunctional linkers for the conjugation of ligands to
molecular probes. J. Org. Chem. 56:4326Y4329 (1991).
Wu, M. L. Breitman, and A. G. Schuh. Failure of blood-island 35. T. Berthelot, J. C. Talbot, G. Laı¨n, G. Deleris, and L. Latxague.
formation and vasculogenesis in Flk-1-deficient mice. Nature
376:62Y66 (1995).
Synthesis of N" -(7-diethylaminocoumarin-3-carboxyl)- and N"-
(7-methoxycoumarin-3-carboxyl)-L-Fmoc lysine as tools for pro-
tease cleavage detection by fluorescence. J. Peptide Sci. 11:153Y160
(2005).
20. G.-H. Fong, J. Rossant, M. Gertsenstein, and M. L. Breitman.
Role of the Flt-1 receptor tyrosine kinase in regulating the
assembly of vascular endothelium. Nature 376:66Y74 (1997).
21. Y. A. Muller, B. Li, H. W. Christinger, J. A. Wells, B. C.
Cunningham, and A. M. de Vos. Vascular endothelial growth
36. T. Berthelot, G. Laı¨n, L. Latxague, and G. Deleris. Synthesis of
novel fluorogenic L-Fmoc lysine derivatives as potential tools
for imaging cells. J. Fluoresc. 14:671Y675 (2004).