Biomaterials 31 (2010) 1787–1797

Contents lists available at ScienceDirect

Biomaterials
journal homepage: www.elsevier.com/locate/biomaterials

The use of RGDGWK-lipopeptide to selectively deliver genes to mouse tumor

vasculature and its complexation with p53 to inhibit tumor growth
Sanjoy Samanta a, Ramakrishna Sistla b, Arabinda Chaudhuri a, *
a
Division of Lipid Science and Technology, Indian Institute of Chemical Technology, Hyderabad 500607, India
b
Pharmacology Division, Indian Institute of Chemical Technology, Hyderabad 500607, India

a r t i c l e i n f o a b s t r a c t

Article history: In vivo selection of phage display libraries have been exploited successfully in the past to isolate various
Received 8 August 2009 high affinity conformationally strained cyclic peptide ligands (CX5–7C, peptides flanked by a cysteine
Accepted 9 October 2009 residue on each side) for integrin receptors capable of selectively homing to tumor vasculatures.
Available online 3 November 2009
Previously, such phase display library studies have shown that integrin a5b1 binds with high affinity to
cyclic peptides containing CRGDGWC motif. Herein we show that a lipopeptide with just the RGDGW
Keywords:
motif (without the flanking cysteine groups) covalently attached to the lysine residue of a mono-
Tumor vasculature targeting
lysinylated cationic amphiphile (RGDGWK-lipopeptide 1) delivers genes to cultured cells preferably via
Integrin targeting
Tumor growth inhibition a5b1 integrins. Importantly, remarkable tumor growth inhibition was observed when the electrostatic
complex of the RGDGWK-lipopeptide 1 and the anti-cancer p53 gene was intravenously administered in
C57BL/6 J mice bearing the aggressive B16F10 tumor. Immunohistochemical staining of mice tumor
cryosections with vasculature markers combined with monitoring expression of the green fluorescence
protein in the same tumor cryosections revealed that the RGDGWK-lipopeptide 1 targets genes to tumor
vasculatures. The colocalization of the TUNEL (terminal deoxyuridine triphosphate nick-end labeling,
a widely used marker of apoptosis) and VE-cadherin (markers of tumor endothelial cells) positive cells in
tumor cryosections support the notion that the remarkable tumor growth inhibition property of the
RGDGWK-lipopeptide 1:p53 complex is initiated through apoptosis of the tumor endothelial cells.
Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction drugs/genes via integrin receptors. For instance, RGD-modified

Integrins, the important class of ab-heterodimeric trans-
membrane glycoprotein receptors and mediators of both cell–
matrix and cell-cell interactions [1], are unique molecular markers
of the tumor endothelial cells. Because of their crucial roles in
mediating tumor angiogenesis, integrins receptors are gaining
increasing importance as drug targets in anti-angiogenic cancer
therapy [2]. The short amino acid sequence arginine-glycine-
aspartic acid (RGD), the most evolutionary conserved recognition
motif for many natural integrin binding extracellular matrix protein
ligands [3], has widely been exploited in the past in targeting
Abbreviations: EDCI, 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide hydro-
chloride; HOBt, 1-Hydroxybenztriazole; BOC, Tert-butyloxy carbonyl; Cbz, Benzyl-
oxycarbonyl; Chol, Cholesterol; DCM, Dichloromethane; DMEM, Dulbecco’s Modifid
Eagles Medium; DMF, N,N-dimethylformamide; FBS, Fetal bovine serum; PBS,
Phosphate buffered saline; GFP, Green fluorescence protein; TUNEL, Terminal
deoxynucleotidyl transferase dUTP nick end labeling; FACS, Fluorescence activated
cell sorter.
* Corresponding author. Tel.: þ91 40 27193201; fax: þ91 40 27193370.
E-mail address: arabinda@iict.res.in (A. Chaudhuri).
liposomes [4–6], lipid based nanoparticles with covalently grafted
RGD motif [7,8], cRGD-functionalized polymeric micelles [9] etc.
have been used previously for selective delivery of drugs/genes to
tumor vasculatures via integrin receptors in anti-angiogenic cancer
therapy.
The interactions between the short RGD recognition motifs of
the extracellular matrix proteins and the integrin receptors are
usually of low affinity and many integrin receptors interact with
secondary sites (also known as synergistic sites) of their native
ligands [10–12]. To this end, in vivo selection of phage display
libraries have been exploited successfully to isolate various high
affinity conformationally strained cyclic peptides (CX5–7C, peptides
flanked by a cysteine residue on each side) for integrins capable of
selectively homing to tumor vasculatures [13–17]. Studies on phage
display libraries have demonstrated that the specificity and affinity
of integrin binding is remarkably dependent on the nature of the
amino acid residues lying C-terminal to RGD. For example, integrin
a5b1 has been shown to bind with high affinity to peptides
containing CRGDGWC motif [16,17]. As to the molecular level
explanation for such high binding affinities of integrin a5b1 to
CRGDGWC peptides, Humphries and coworkers showed that

0142-9612/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biomaterials.2009.10.027
1788 S. Samanta et al. / Biomaterials 31 (2010) 1787–1797

(CH2)4 NHCbz
H3C(H2C)15 NH2 (CH2)4 NHCbz H
N i) H3C(H2C)15 N
H3C(H2C)15 + N NHBoc
HOOC NHBoc H3C(H2C)15
O
I
N-2-aminoethyl-N,N-di-n- α ε
N Boc-N -z-L-Lysine 2-(N,N-di-n-hexadecylamino)-ethylamino-Lys(z)-NHBoc
hexadecylamine

O O
- H H
H + Cl N N
MeOOC C NH3 MeO NHBoc HO NHBoc
ii ) iii )
O O

HN HN HN

L-Tryptophan-methyl ester III

II
hydrochloride

Cbz HN

O
H H
iv ) H3C(H2C)15 N N
I III N N NHBoc
H3C(H2C)15 H
v) O O

HN

IV

2-(N,N-di-n-hexadecylamino)-ethylamino-Lys(z)-Trp-G ly -NHBoc

O O
O H
NHBoc N
NHBoc O NHBo c ix)
HO vi) O vii)
O x)
viii ) BnOOC
BnOOC BnOOC

N- t- butyloxycarbonyl-L - V VI
Aspartic acid - -benzylester
O O O
O H
H N
N NHCbz NHCbz
O N xi) HO N
H H
O O
BnOOC BnOOC
NCbz NCbz
HN HN
NHCbz VIII NHCbz
VII
HO- Asp( OBzl )- Gly- Arg( di -z )- NHz
Fig. 1. Scheme used for syntheses of RGDGWK-lipopeptide 1 and structure of the control RGDLFK-lipopeptide 2.
 S. Samanta et al. / Biomaterials 31 (2010) 1787–1797 1789

CbzHN

Cl O O O
H H H
xii) N NHCbz
R N N
IV VIII N N N N
xiii) R H H H
H O O O
BnOOC
HN NCbz
HN
R = -(CH 2 )15 CH3 NHCbz
IX
Cl NH3

O O O
H H H Cl
Cl N N N (CH2 )15 CH3
H3N
xiv) N N N N
H H H (CH2 )15 CH3
IX O O O H
COOH
HN NH
NH 2 Cl
H2N

RGDGWK-Lipopeptide 1

Reagents and conditions: i) EDCI, HOBt, DIPEA, dry DCM, 12 h, rt; ii) N -t-
butyloxycarbonyl -Glycine, EDCI, HOBt, DIPEA, dry DCM, 12 h, rt; iii)
THF/MeOH/H2O (3:1:1), LiOH, 2 h, 0 ºC; iv) 3:1 Dry DCM:TFA (v/v); v) EDCI, HOBt,
DIPEA, dry DCM, 12 h, rt; vi ) Allyl alcohol, DCC , DMAP, dry DCM, 12 h, rt.; vii) 3:1
Dry DCM:TFA (v/v); viii ) N -t-butyloxycarbonyl-Glycine, EDCI, HOBt, DIPEA, dry
DCM, 12 h, rt; ix) 3:1 Dry DCM:TFA (v/v); x ) N α, N δ, N ω-tri-benzoyloxycarbonyl-L-
Arginine, EDCI, HOBt, DIPEA, dry DCM, 12 h, rt; xi) Dry THF, (Ph 3P)4Pd, NMM, N 2
atm., 3 h, rt; xii) Dry DCM/TFA/thioanisole (1:4:4, v/v), 2 h, N2 atm., 0 ºC; xiii ) Dry
DCM, EDCI, HOBt, pH 4-6, 4 h, rt.; xiv) Dry THF, 6N HCl, Pd(OH)2, H2, 8 h, rt.

Structure of control RGDLFK-lipopeptide 2. NH3 Cl

O O
H O Cl
Cl H 3N H H
N N
N N N (CH2 )15CH3
H N N
H H
O O (CH2 )15 CH3
COO H O H
HN
NH2 Cl
H2N

RGDLFK- Lipopeptide 2
Fig. 1. (continued).
1790 S. Samanta et al. / Biomaterials 31 (2010) 1787–1797
mutation of Trp157 of a5 subunit to alanine resulted into loss of high
affinity binding of the a5b1 integrin to RGDGW [18]. These findings
are consistent with the supposition that Trp157 of a5 subunit
participates in a favorable hydrophobic interaction with the tryp-
tophan of the RGDGW resulting into its high affinity binding with
a5b1 integrin. We have recently demonstrated that a lipopeptide
containing a non-cyclic conformationally unstrained simple RGDK
tetrapeptide sequence in its polar head-group region can selec-
tively target genes to proangiogenic a5b1 integrin receptors and
can target genes to mouse tumor vasculatures [19]. Taking such
a5b1 integrin receptor selective tumor vasculature targeting effi-
ciencies of the simple RGDK-lipopeptide into account, we became
interested in probing the integrin receptor selectivity and tumor
vasculature targeting property of lipopeptide having just the
‘‘RGDGW’’ motif i.e. without the terminal cysteine residues of the
‘‘CRGDGWC’’ ligand (obtained from phage display studies
mentioned above) in its head-group region.
Herein, we show that RGDGWK-lipopeptide 1 (Fig. 1) with
covalently grafted RGDGW motif to the lysine residue of a monoly
sinylated cationic amphiphile delivers genes to cultured cells
preferably via a5b1 integrins and that RGDGWK-lipopeptide 1
selectively targets genes to mouse tumor vasculatures. We also
demonstrate that significant tumor growth inhibition can be
accomplished through intravenous administration of the electro-
static complex of the RGDGWK-lipopeptide 1 and the anti-cancer
p53 gene in C57BL/6J mice bearing the aggressive B16F10 tumor.
Furthermore, the immunohistochemical findings described below
are consistent with the supposition that the remarkable tumor
growth inhibition property of the RGDGWK-lipopeptide 1:p53
complex is mediated through apoptosis of the tumor endothelial
cells. Taken together, the presently described RGDGWK-lipopeptide
1 holds therapeutic potential for future use in anti-angiogenic
cancer therapy.
2. Materials and methods
2.1. General procedures and reagents
p-CMV-SPORT-b-gal and p-a5GFP plasmids were obtained as generous gifts
from Dr. Nalam Madhusudhana Rao and Dr. Gopal Pande respectively, Centre for
Cellular and Molecular Biology, Hyderabad, India. Mouse anti-human a5b1 mono-
clonal antibody, mouse anti-human avb3 monoclonal antibody, mouse anti-human
avb5 monoclonal antibody and the blood vessel staining kit were purchased from
Chemicon, USA. FITC-conjugated rat anti-mouse secondary antibody, o-nitrophenyl-
b-D-galactopyranoside, cholesterol, cell culture media and fetal bovine serum were
purchased from Sigma, St. Louis, USA. NP-40, antibiotics and agarose were
purchased from Hi-media, India. TUNEL assay kit was purchased from Promega
Corp. USA. VE-Cadherin mAb and Texas Red-conjugated anti-mouse secondary
antibody were procured from Santacruz Biotech, USA. A549 (Human Pulmonary
Carcinoma cell) cell line was procured from ATCC (USA). B16F10 (murine melanoma
cells) was procured from the National Centre for Cell Sciences, NCCS, Pune, India.
Cells were grown at 37  C in Dulbecco’s modified Eagle’s medium (DMEM) with 10%
FBS in a humidified atomosphere containing 5% CO2. 6–8 weeks old female C57BL/6J
mice (each weighing 20–22 g) were purchased from National Institute of Nutrition,
Hyderabad, India. All the in vivo experiments were performed in accordance with
the Institutional Bio-Safety and Ethical Committee Guidelines using an approved
animal protocol. Mass spectral data were acquired by using a commercial LCQ ion
trap mass spectrometer (ThermoFinnigan, SanJose, CA, USA) equipped with an ESI
source or micromass Quatro LC triple quadrapole mass spectrometer for ESI analysis.
Data were acquired by liquid secondary ion mass spectrometry (LSIMS) using meta-
nitrobenzyl alcohol as the matrix. LSIMS analysis was performed in the scan range
100–1000 amu at the rate of 3 scans/s. 1H NMR spectra were recorded on Varian FT
200 MHz and 400 MHz Spectrometer. NaBoc-N3-Z-L-Lysine, N-t-butyloxycarbonyl-L-
Asparticacid-b-benzylester, L-Tryptophan, Thioanisole were purchased from Fluka
(Switzerland). EDCI and Tetrakis-(triphenylphosphin)-palladium(0) were purchased
from Sigma (USA) and Aldrich (USA), respectively. Na, Nd, Nu-tri-benzyloxycarbonyl-
L-Arginine was purchased from Senn Chemicals (Switzerland). Column chroma-
tography was performed with silica gel (Acme Synthetic Chemicals, India,
60–120 mesh) and acidic alumina (s. d. Fine chemicals, India).
2.2. Syntheses
The synthetic route used for preparing the RGDGWK-lipopeptide 1 is shown in
Fig. 1. The starting mixed primary-tertiary amine namely, N-2-aminoethyl-N,N-di-n-
hexadecylamine was prepared as described previously [20]. Structures of all the
synthetic intermediates shown below in Fig. 1 were confirmed by 1H NMR and
LSIMS. Severe line broadening, particularly in the range d 3–5 ppm, was observed in
the 1H NMR spectra of all the final lipopeptides presumably due to presence of
multiple exchangeable protons in their head-group regions. Thus, the final lipid was
characterized by the molecular ion peaks in their ESIMS. The purity of the target
RGDGWK-lipopeptide 1 was confirmed by reversed phase analytical HPLC analysis.
Details of synthetic steps for preparing the RGDGWK-lipopeptide 1, 1H NMR & mass
spectral data for RGDGWK-lipopeptide 1 and those for all its synthetic intermediates
shown in Fig. 1 are provided below.
Step i). Solid HOBt (1.2 g, 8.9 mmol) and EDCI (1.7 g, 8.9 mmol) were added
sequentially to an ice cold and stirred solution of NaBOC-N3-Z-L-Lysine (3.4 g,
8.9 mmol) in dry DCM (15 mL). After half an hour, N-2-aminoethyl-N,N-di-n-hexa-
decylamine (3 g, 5.9 mmol) dissolved in dry DCM (10 mL) was added to the reaction
mixture. Di-isopropylethyl amine (DIPEA) was added dropwise to the stirred reac-
tion mixture until it became alkaline to litmus. The resulting solution was left stirred
at room temperature for 12 h. The reaction mixture was then transferred in chlo-
roform (100 mL) and washed sequentially with ice-cooled 1 N HCl (2  100 mL),
saturated sodium bicarbonate (2  100 mL) and brine (1 100 mL). The organic
layer was dried over anhydrous sodium sulfate, filtered and the solvent from the
filtrate removed by rotary evaporation. The residue upon column chromatographic
purification with 60–120 mesh silica gel using 10% acetone-hexane (v/v) as eluent
afforded intermediate I (3.1 g, 60% yield, Rf ¼ 0.5, 30:70 acetone:hexane, v/v).
1
H NMR (200 MHz, CDCl3):d/ppm ¼ 0.9 [t, J ¼ 6.6 Hz, 6H, CH3–(CH2)15–]; 1.2–1.9
[bs, 56H, –(CH2)14–; s, 9H, CO–O–C(CH3)3; m, 4H, LysCgH2 þ LysCdH2; m, 2H,
b
LysC H2]; 2.3–2.6 [t, J ¼ 7.3 and 5.9 Hz, 4H, –N(–CH2–CH2–)2; t, J ¼ 5.8 and 5.1 Hz, 2H,
–N–CH2–CH2– NH–CO]; 3.1–3.4 [m, 2H, LysuCH2; m, 2H, –N–CH2–CH2– NH–CO–];
4.0 [m, 1H, LysCaH]; 4.9–5.2 [m, 1H, NH–CO–O–CH2–C6H5; 2H, COO–CH2–C6H5;
m,1H, LysCaH–NH–CO–]; 6.5 [m, 1H, –CH2–CH2–NH–CO–]; 7.2–7.4 [m, 5H, COO–
CH2–C6H5]. LSIMS: m/z ¼ 872 [M þ 1]þ for C53H98N4O5.
Step ii). L-tryptophan methyl ester hydrochloride (3.4 g, 15.5 mmol) was dis-
solved in dry DMF (10 mL) and the solution was added to an ice cold reaction
mixture (which has been under stirring condition for ½ h) containing solid HOBt
(3.6 g, 23.3 mmol), EDCI (4.5 g, 23.3 mmol) and N-t-butyloxycarbonyl-Glycine (4 g,
23.3 mmol) in dry DCM (30 mL). Di-isopropylethyl amine (DIPEA) was added
dropwise to the stirred reaction mixture until it became alkaline to litmus. The
resulting solution was left stirred at room temperature for 12 h. The reaction
mixture was then transferred in chloroform (100 mL) and washed sequentially with
ice-cooled 1 N HCl (2  100 mL), saturated sodium bicarbonate (2  100 mL) and
brine (1 100 mL). The organic layer was dried over anhydrous sodium sulfate,
filtered and the solvent from the filtrate removed by rotary evaporation. The residue
upon column chromatographic purification with 60–120 mesh silica gel using 1%
methanol-chloroform (v/v) as eluent afforded intermediate II (3.5 g, 60% yield,
Rf ¼ 0.5, 5:95 methanol:chloroform, v/v).
1
H NMR (200 MHz, CDCl3):d/ppm ¼ 1.2 [s, 9H, CO–O–C(CH3)3]; 3.1 [m, 2H,
TrpCbH2]; 3.6 [s, 3H, –OCH3; d, 2H, –CO–CH2–NH–]; 4.8 [m, 1H, TrpCaH]; 5.4 [t, 1H,
–CH2–NH–CO–]; 6.8–7.4 [m, 1H, –CH–NH–CO–; 5H, –CH2–C8H5NH]; 8.8 [s, 1H, Ind-
NH]. LSIMS: m/z ¼ 375 [M þ 1]þ for C19H25N3O5.
Step iii). To a stirred solution of intermediate II (2 g, 1.46 mmol) in THF/methanol/
water (3:1:1, v/v), lithium hydroxide (0.8 g, 4.4 mmol) was added at 0  C and the
reaction mixture was left under stirring for 2 h at 0  C. Excess lithium hydroxide was
filtered off and the reaction mixture was neutralized with aqueous potassium
hydrogen sulphate to pH 3–4. The product was extracted with chloroform and dried
over anhydrous sodium sulphate. Removal of the solvent by rotary Evaporation
afforded 1.8 g of intermediate III (98% yield, Rf ¼ 0.2, 5:95 methanol:chloroform, v/v).
1
H NMR (200 MHz, CDCl3):d/ppm ¼ 1.2 [s, 9H, CO–O–C(CH3)3]; 3.1 [m, 2H,
TrpCbH2]; 3.6 [d, J ¼ 5 Hz, 2H, –NH–CH2–CO–]; 4.8 [m, 1H,TRpCaH]; 5.4 [t, 1H, –CH2–
NH–CO–]; 6.8–7.4 [m, 1H, –CH–NH–CO–; 5H, –CH2–C8H5NH]; 8.8 [s, 1H, Ind-NH].
LSIMS: m/z ¼ 361 [M þ 1]þ for C18H23N3O5.
Steps iv) & v). The intermediate I (2.7 g, 3.1 mmol) obtained above in step i) was
dissolved in dry DCM (15 mL) and TFA (5 mL) was added at 0  C. The resulting
solution was left stirred at 0  C for 3 h to ensure complete deprotection. Excess TFA
was removed by nitrogen flushing. The resulting compound was dissolved in
chloroform (100 mL) and washed sequentially with aqueous saturated NaHCO3
(3  100 mL) and brine (1 100 mL). The organic layer was dried over anhydrous
sodium sulfate and filtered. The filtrate upon removal of solvent by rotary evapo-
ration afforded 2.21 g deprotected amine intermediate. The amine intermediate
(2.21 g, 2.87 mmol) was dissolved in dry DCM (15 mL) and the solution was added to
an ice cold reaction mixture (which has been under stirring condition for ½ h)
containing solid HOBt (0.66 g, 4.3 mmol), EDCI (0.83 g, 4.3 mmol) and intermediate
III (1.6 g, 4.3 mmol) in dry DCM/DMF (5:1, 10 mL). Di-isopropylethyl amine (DIPEA)
was added dropwise to the stirred reaction mixture until it became alkaline to
litmus. The resulting solution was left stirred at room temperature for 12 h. The
reaction mixture was then transferred in chloroform (50 mL) and washed sequen-
tially with ice-cooled 1 N HCl (2  100 mL), saturated sodium bicarbonate
 S. Samanta et al. / Biomaterials 31 (2010) 1787–1797
(2  100 mL) and brine (1 100 mL). The organic layer was dried over anhydrous
sodium sulfate, filtered and the solvent from the filtrate removed by rotary evapo-
ration. The residue upon column chromatographic purification with 60–20 mesh
silica gel using 1.5% methanol/chloroform (v/v) as eluent afforded 2.8 g of inter-
mediate IV (87% yield, Rf ¼ 0.6, 10:90 methanol:chloroform, v/v).
1
H NMR (200 MHz, CDCl3):d/ppm ¼ 0.9 [t, J ¼ 6.6 Hz, 6H, CH3–(CH2)15–]; 1.2–1.9
[bs, 56H, –(CH2)14–; s, 9H, CO–O–C(CH3)3; m, 4H, LysCgH2 þ LysCdH2; m, 2H,
LysCbH2]; 2.3–2.6 [t, J ¼ 7.2 and 5.8 Hz, 4H, –N(–CH2–CH2–)2; t, J ¼ 5.1 and 5.3 Hz, 2H,
–N–CH2–CH2– NH–CO]; 3–3.3 [m, 2H, LysuCH2; 2H, –N–CH2–CH2–NH–CO–; m, 2H,
TrpCbH2]; 3.75 [d, 2H, GlyCaH2]; 4.25 [m, 1H, LysCaH]; 4.75 [m, 1H, TRpCaH]; 5.1 [s,
2H, COO–CH2–C6H5]; 5.4 [m,1H, LysCaH–NH–CO–]; 6.8–7.6 [m, 1H, LysuCH2–NH–;
m, 5H, –CH2–C8H5NH; m, 5H, COO–CH2–C6H5; m, 1H, –CH2–CH2–NH–CO–]; 9.6 [s,
1H, Ind-NH]. LSIMS: m/z ¼ 1116 [M þ 1]þ for C53H98N4O5.
Step vi). A mixture of DCC (3.2 g, 15.5 mmol), DMAP (0.095 g, 0.7 mmol) and N-t-
butyloxycarbonyl-L-Aspartic acid-b-benzylester (5 g, 15.5 mmol) was stirred for half
an hour in dry DCM (50 mL). The allyl alcohol (5.4 g, 93 mmol) was added to the
mixture and the reaction mixture was left stirred at room temperature for 12 h. The
solvent was removed by rotary evaporation and resuspended in ethylacetate (20 mL).
DCU was filtered off and organic layer was washed with 1 N HCl (2  100 mL), satu-
rated sodium bicarbonate (2  100 mL) and brine (1 100 mL). The organic layer was
dried over anhydrous sodium sulfate, filtered and the solvent from the filtrate
removed by rotary evaporation. The residue upon column chromatographic purifi-
cation with 60–120 mesh silica gel using 0.5% methanol/chloroform (v/v) as eluent
afforded intermediate V (4.7 g, 65% yield, Rf ¼ 0.6, 20:80 acetone:hexane, v/v).
1
H NMR (200 MHz, CDCl3):d/ppm ¼ 1.4 [s, 9H, CO–O–C(CH3)3]; 2.8–3.1 [m, 2H,
AspCbH2]; 4.6 [m, 1H, AspCaH; m, 2H, –CH2CH]CH2]; 5.1 [s, 2H, –CH2C6H5]; 5.2–5.4
[m, 2H, –CH2CH]CH2]; 5.8 [m, 1H, –CH2CH]CH2]; 7.2–7.4 [m, 1H, –NH–COO–; m,
5H, –CH2C6H5]. LSIMS: m/z ¼ 363 [M þ 1]þ for C19H25NO6.
Steps vii) & viii). The intermediate V (3.6 g, 10 mmol) obtained above in step vi)
was dissolved in dry DCM (15 mL) and TFA (5 mL) was added at 0  C. The resulting
solution was left stirred at 0  C for 3 h to ensure complete deprotection. Excess TFA was
removed by nitrogen flushing. The amine intermediate (2.5 g, 9.5 mmol) was dis-
solved in dry DCM (15 mL) and the solution was added to an ice cold reaction mixture
(which has been under stirring condition for ½ h) containing solid HOBt (2.2 g,
14.3 mmol), EDCI (2.7 g, 14.3 mmol) and N-t-butyloxycarbonyl-Glycine (2.5 g,
14.3 mmol) in dry DCM (20 mL). Di-isopropylethyl amine (DIPEA) was added drop-
wise to the stirred reaction mixture until it became alkaline to litmus. The resulting
solution was left stirred at room temperature for 12 h. The reaction mixture was then
transferred in chloroform (100 mL) and washed sequentially with 1 N HCl
(2  100 mL), saturated sodium bicarbonate (2  80 mL) and brine (1  50 mL). The
organic layer was dried over anhydrous sodium sulfate, filtered and the solvent from
the filtrate removed by rotary evaporation. The residue upon column chromatographic
purification with 60–120 mesh silica gel using 1% methanol/chloroform (v/v) as eluent
afforded intermediate VI (2.7 g, 67% yield, Rf ¼ 0.5, 5:95 methanol:chloroform, v/v).
1
H NMR (200 MHz, CDCl3):d/ppm ¼ 1.4 [s, 9H, CO–O–C(CH3)3]; 2.8–3.1 [m, 2H,
AspCbH2]; 3.8 [d, 2H, GlyCaH2–]; 4.6 [m, 2H, –CH2CH]CH2]; 4.9 [m, 1H, AspCaH]; 5.1
[s, 2H, –CH2C6H5]; 5.2–5.4 [m, 2H, –CH2CH]CH2]; 5.8 [m, 1H, –CH2CH]CH2]; 7.0 [d,
1H, –NHCOCH2–]; 7.2–7.4 [m, 1H, –NH–COO–; m, 5H, –CH2C6H5]. LSIMS: m/z ¼ 420
[M þ 1]þ for C21H28N2O7.
Steps ix) & x). The intermediate VI (2.6 g, 6.2 mmol) obtained above in step viii)
was dissolved in dry DCM (12 mL) and TFA (4 mL) was added at 0  C. The resulting
solution was left stirred at 0  C for 3 h to ensure complete deprotection. Excess TFA
was removed by nitrogen flushing. The amine intermediate (1.9 g, 5.9 mmol) was
then dissolved in dry DCM (15 mL) and the solution was added to an ice cold
reaction mixture (which has been under stirring condition for ½ h) containing solid
HOBt (0.72 g, 4.7 mmol), EDCI (0.9 g, 4.7 mmol) and Na, Nd, Nu-tri-benzyloxy
carbonyl-L-Arginine (2.75 g, 4.7 mmol) dry DCM (20 mL). Di-isopropylethyl amine
(DIPEA) was added dropwise to the stirred reaction mixture until it became alkaline
to litmus. The resulting solution was left stirred at room temperature for 12 h. The
reaction mixture was then transferred in chloroform (100 mL) and washed
sequentially with 1 N HCl (2  80 mL), saturated sodium bicarbonate (2  80 mL)
and brine (1  50 mL). The organic layer was dried over anhydrous sodium sulfate,
filtered and the solvent from the filtrate removed by rotary evaporation. The residue
upon column chromatographic purification with 60–120 mesh silica gel using 1%
methanol/chloroform (v/v) as eluent afforded intermediate VII (3.7 g, 71% yield,
Rf ¼ 0.6, 5:95 methanol:chloroform, v/v).
1
H NMR (200 MHz, CDCl3):d/ppm ¼ 1.9 [m, 4H, ArgCbH2 þ ArgCgH2]; 2.9 [m, 2H,
Asp CbH2]; 3.4 [m, 2H, ArgCdH2]; 3.8 [d, 2H, GlyCaH2–]; 4.4 [m, 1H, ArgCaH]; 4.6 [m,
2H, –CH2CH]CH2]; 4.9 [m, 1H, AspCaH]; 5.0–5.4 [m, 8H, –CH2C6H5; m, 2H,
–CH2CH]CH2]; 5.8 [m, 1H, –CH2CH]CH2]; 6.4 [m, 1H, ArgCaH–NH–]; 7.0 [d, 1H,
–NHCOCH2–]; 7.2–7.4 [m, 1H, –NH–COO–; m, 20H, –CH2C6H5]; 9.4 [s, 1H, –CH]NH].
LSIMS: m/z ¼ 879 [M þ 1]þ for C46H50N6O12.
Step xi). Tetrakis (triphenylphosphine) palladium (0.26 g, 0.23 mmol) was added
to a stirred solution of intermediate VII (2 g, 2.27 mmol) in dry THF (15 mL) and then
N-methylmorpholine (0.69 g, 6.8 mmol) was added dropwise at 0  C. The reaction
mixture was left under stirring for 3 h under nitrogen atmosphere. The reaction
mixture was diluted with chloroform (50 mL) and washed sequencially with 1 N HCl
(2  100 mL) and brine (200 mL). The organic layer was dried over anhydrous
sodium sulfate, filtered and the solvent from the filtrate removed by rotary
1791
evaporation. The residue upon column chromatographic purification with 60–
120 mesh silica gel using 1% methanol/chloroform (v/v) as eluent afforded inter-
mediate VIII (1 g, 52% yield, Rf ¼ 0.2, 5:95 methanol:chloroform, v/v).
1
H NMR (200 MHz, CDCl3):d/ppm ¼ 1.9 [m, 4H, ArgCbH2 þ ArgCgH2]; 2.9 [m, 2H,
Asp CbH2]; 3.6–4.1 [m, 2H, ArgCdH2; d, 2H, GlyCaH2–]; 4.4 [m, 1H, ArgCaH]; 4.9 [m,
1H, AspCaH]; 5.0–5.2 [m, 8H, –CH2C6H5]; 6.4 [m, 1H, ArgCaH–NH–]; 7.0 [d, 1H,
–NHCOCH2–]; 7.2–7.6 [m, 1H, –NH–COO–; m, 20H, –CH2C6H5]; 9.4 [s, 1H, –CH]NH].
LSIMS: m/z ¼ 840 [M þ 1]þ for C43H46N6O12.
Steps xii) & xiii). The intermediate IV (2.0 g, 1.8 mmol) obtained above in step v)
was dissolved in dry DCM/thioanisole (1:4, 10 mL) and TFA (8 mL) was added at 0  C
under nitrogen atmosphere. The resulting solution was left stirred at 0  C for 2 h to
ensure complete deprotection. Excess TFA was removed by nitrogen flushing and
thioanisole was removed by repeated washing with hexane. The amine intermediate
salt (1.5 g, 1.47 mmol) was then dissolved in dry DCM (15 mL) and the solution was
added to an ice cold reaction mixture (which has been under stirring condition for
½ h) containing solid HOBt (0.18 g, 1.19 mmol), EDCI (0.227 g, 1.19 mmol) and
intermediate VIII (0.99 g, 1.19 mmol) in dry DCM (10 mL). Di-isopropylethyl amine
(DIPEA) was added dropwise to the stirred reaction mixture until pH 4–6. The
resulting solution was left stirred at room temperature for 4 h. The reaction mixture
was then transferred in chloroform (50 mL) and washed sequentially with 1 N HCl
(2  80 mL) and brine (1  50 mL). The organic layer was dried over anhydrous
sodium sulfate, filtered and the solvent from the filtrate removed by rotary evapo-
ration. The residue upon column chromatographic purification with acidic alumina
using 2% methanol/chloroform (v/v) as eluent afforded intermediate IX (1 g, 37%
yield, Rf ¼ 0.7, 10:90 methanol:chloroform, v/v)).
1
H NMR (300 MHz, CDCl3):d/ppm ¼ 0.9 [t, J ¼ 6.2 Hz, 6H, CH3–(CH2)15–]; 1.2–1.9
[m, 56H, –(CH2)14–; 4H, LysCgH2 þ LysCdH2; ArgCbH2 þ ArgCgH2; m, 2H, LysCbH2];
2.4–3.4 [m, 2H, AspCbH2; m, 4H, –N(–CH2–CH2–)2; 2H, –N–CH2–CH2– NH–CO; m, 2H,
–N–CH2–CH2–NH–CO; m, 2H, LysuCH2; d, 2H, TrpCbH2]; 3.6–4.4 [d, 2H, GlyCaH2; m,
2H, ArgCdH2; m, 1H, Arg CaH; m, 1H, LysCaH ]; 4.5–5.2 [m, 1H, AspCaH; m, 1H,
TrpCaH; m, 10H, COO–CH2–C6H5; m, 1H, LysuCH2-NH–CO–O–CH2–C6H5]; 5.4–7.6 [m,
25 H, COO–CH2–C6H5; 5H, –CH2–C8H5NH(ind); ]; 7.8–9.4 [s, 1H, –CH]NH; s, 1H, Ind-
NH; m, 1H]. ESIMS: m/z ¼ 1836 [M þ 1]þ for C104H148N13O16.
Step xiv). The intermediate IX (0.1 g, 0.054 mmol) obtained above in step xiii)
was dissolved in HPLC grade THF (10 mL) and 3 drops of 6 (N) HCl was added. In the
resulting solution catalytic amount of Pd(OH)2 was added and the reaction mixture
was kept stirred under hydrogen atmosphere for 8 h to ensure complete hydroge-
nation. The reaction mixture was filtered and the filtrate was dried in rotary evap-
orator. The solid residue was dissolved in 2–3 drops of methanol and the pure
RGDGWK-lipopeptide 1 was precipitated with addition of excess acetone (20 mL).
The precipitate upon filtration and vacuum drying afforded the target RGDGWK-
lipopeptide 1 as pure white solid (0.04 g, 61% yield).
1
H NMR (400 MHz, CDCl3 þ CD3OD):d/ppm ¼ 0.9 [t, J ¼ 6.2 Hz, 6H, CH3–(CH2)15–
]; 1.2–1.9 [m, 56H, –(CH2)14–; m, 4H, LysCgH2 þ LysCdH2; m, 4H, ArgCbH2 þ ArgCgH2;
m, 2H, LysCbH2]; 2.8–4.0 [m, 2H, AspCbH2; m, 4H, –N(–CH2–CH2–)2; 2H, –N–CH2–
CH2– NH–CO; m, 2H, –N–CH2–CH2–NH–CO; m, 2H, LysuCH2; d, 2H, TrpCbH2; d, 4H,
GlyCaH2; m, 2H, ArgCdH2; m, 1H, Arg CaH; m, 1H, LysCaH ]; 4.4–4.8 [m, 1H, AspCaH;
m, 1H, TrpCaH; m, 1H, LysuCH2–NH–CO–O–CH2–C6H5]; 7.0–7.8 [m, 5H, –CH2–
C8H5NH(ind.)]. ESIMS: m/z ¼ 1210 [M þ 1]þfor C65H118N13O8.
The control RGDLFK-lipopeptide 2 was synthesized using the same synthetic
scheme as used for preparing RGDGWK-lipopeptide 1 using the appropriately pro-
tected leucine and phenylalanine and its structure was confirmed using the same 1H
NMR and mass spectral technique as used for RGDGWK-lipopeptide 1. As in case of
RGDGWK-lipopeptide 1, the purity of the target control RGDLFK-lipopeptide 2 was
confirmed by analytical HPLC analysis.
1
H NMR for RGDLFK-lipopeptide 2 (400 MHz, CDCl3 þ CD3OD):d/ppm ¼ 0.6–
0.9 [m, 6H, –CH(CH3)2; 6H, CH3–(CH2)15]; 1.2–2.0 [m, 1H, –CH(CH3)2; m, 56H,
–(CH2)14–; 4H, LysCgH2 þ LysCdH2; m, 2H, Lys CbH2; m, 2H, –CH2CH(CH3)2; m,
4H, ArgCbH2 þ ArgCgH2]; 2.8–4.8 [m, 2H, AspCbH2; m, 4H, –N(–CH2–CH2–)2; 2H,
–N–CH2–CH2– NH–CO; m, 2H, –N–CH2–CH2–NH–CO; m, 2H, LysuCH2; m, 2H,
PheCaH –CH2Ph; d, 2H, GlyCaH2; m, 2H, ArgCdH2; m, 1H, ArgCaH; m, 1H, LysCaH;
m, 1H, PheCaH; m, 1H, LeuCaH; m, 1H, AspCaH]; 7.1–7.4 [m, 5H, –CH2–C6H5].
ESIMS: m/z ¼ 1226 [M þ 1]þfor C67H124N12O8.
2.3. Preparation of liposomes
The cationic lipid and cholesterol in 2:1 mole ratio was dissolved in a mixture of
chloroform and methanol (3:1, v/v) in a glass vial. The solvent was removed with
a thin flow of moisture free nitrogen gas and the dried lipid film was then kept under
high vacuum for 8 h. Sterile deionized water was added to the vacuum dried lipid
film and the mixture was allowed to swell overnight. The vial was then vortexed for
2–3 min at room temperature to produce multilamellar vesicles (MLVs). MLVs were
then sonicated in an ice bath until clarity using a Branson 450 sonifier at 100% duty
cycle and 25 W output power to produce small unilamellar vesicles (SUVs).
2.4. FACS protocol
Approximately 4  106 A549 cells were fixed for 1 h in 2% formaldehyde solu-
tion, centrifuged and supernatant was discarded. Cells were suspended in 4 mL of
1792 S. Samanta et al. / Biomaterials 31 (2010) 1787–1797

Fig. 2. Gene transfection by RGDGWK-lipopeptide 1 is mediated more selectively via a5b1 integrin receptor among a5b1, avb3, avb5 integrin receptors. A. Evaluation of the a5b1, avb3
and avb5 integrins in A549 cells by FACS. In each of the three experiments, about one million cells were fixed and labeled with monoclonal antibodies against a5b1, avb3 and avb5
integrins and then examined by flow cytometry. The green bordered areas on the left side in each FACS profile represents the signal of the control isotype antibody and the traces on
the right (blue bordered) represent the signal with anti-a5b1 (left), anti-avb3 (middle) and anti-avb5 (right) integrin antibodies; B. Gene transfer efficiencies of the RGDGWK-
lipopeptide 1 in A549 cells (measured using the lipofection method described in reference 24) in presence of monoclonal antibodies against a5b1, avb3 and avb5 integrins (*P < 0.005
when compared to the control transfection experiments in absence of any added antibody). C. Gene transfer efficiencies of the control RGDLFK-lipopeptide 2 in A549 cells in
presence of monoclonal antibodies against a5b1, avb3 and avb5 integrins (*P < 0.002 when compared to the control transfection experiments in absence of any added antibody).

wash buffer (PBS containing 2% FBS and 0.1% sodium azide). After 30 min of incu-
bation at 37  C, it was divided equally to four Eppendorff tubes. The tubes were
centrifuged and the supernatant was discarded. 100 mL of three different commer-
cially available monoclonal antibodies (at a 1:100 dilution in wash buffer) was added
to three of the four eppendorff tubes and in the fourth one, 100 mL only wash buffer
was added as a control. The cell suspensions were then incubated at room
temperature for 1.5 h, centrifuged, the supernatant discarded and the cell pellets
were washed with wash buffer (2  1 mL). 100 mL of FITC-conjugated secondary
antibody (at 1:200 dilution) was added to all the four eppendroff tubes and incu-
bated for 1 h. Centrifugation and washing were repeated for 2–3 times. Finally, the
cell pellets were suspended in 1 mL wash buffer and the flow cytometry histograms
were recorded in a FACS-canto II instrument (Becton Dickinson).
2.5. In vitro transfection
The efficiencies of the RGDGWK-lipopeptide 1 as well that of the control
RGDLFK-lipopeptide 2 in transfecting A549 cells were evaluated following a previ-
ously described transfection protocol [9]. Briefly, cells were seeded at a density of
12,000 cells per well in a 96-well plate 18–24 h before the transfection. The elec-
trostatic complexes of the lipopeptides and p-CMV-SPORT-b-Gal plasmid (con-
taining 0.3 mg of DNA and appropriate amount of cationic liposomes required to
maintain lipid:DNA charge ratio of 9:1) were added to plain DMEM medium, the
total volume was made up to 100 mL and the solution was kept under mild shaking at
room temperature for 20 min. The complexes were then added to the cells. After 3 h
of incubation, 100 mL of DMEM with 20% FBS was added to the cells. The medium was
changed to 10% complete medium after 24 h and the reporter gene activity was
estimated after 48 h. The cells were washed twice with PBS (100 mL each) and lysed
in 50 mL lysis buffer [0.25 M Tris-HCl pH 8.0, 0.5% NP40]. Care was taken to ensure
complete lysis. The b-galactosidase activity per well was estimated by adding 50 mL
of 2X-substrate solution [1.33 mg/mL of ONPG, 0.2 M sodium phosphate (pH 7.3) and
2 mM magnesium chloride] to the lysate in a 96-well plate. Absorption at 405 nm
was converted to b-galactosidase units using a calibration curve constructed with
pure commercial b-galactosidase enzyme. The values of b-galactosidase units in
triplicate experiments assayed on the same day varied by less than 20%. The
transfection experiment was carried in duplicate and the transfection efficiency
values shown in Fig. 2 are the average of triplicate experiments performed on the
same day. Each transfection experiment was repeated two times and the day to day
variation in average tansfection efficiency was found to be within 2-fold.
2.6. Integrin receptor selectivity studies
Cells were seeded at a density of 12,000 cells per well in a 96-well plate 18–24 h
before transfection. Cells were pre-incubated with 50 mL of three different mono-
clonal antibodies (at a dilution of 1:25 in 10% complete medium) for 45 min at room
temperature. After 45 min, medium containing the antibody from the cells were
taken out. Then a fresh amount of 50 mL of three different monoclonal antibodies (at
a dilution of 1:25 in 10% complete medium) were added to the cell wells. 50 mL of
lipopeptide:DNA complexes (containing 9:1 þ/ charge ratio) was added to each
well. After 1 h incubation at 37  C medium was discarded and 200 mL of 10% complete
medium was added to each well. The reporter gene expression was monitored after
48 h. Each transfection experiment was conducted in duplicate on the same day.
2.7. Immunohistochemical studies
Immunohistochemical staining of tumor endothelial cells were performed on
the 23rd day after tumor inoculation. On the 22nd day after tumor inoculation, mice
were intravenously injected with the complex of RGDGWK-lipopeptide 1 and a5-
GFP/or p53 plasmid DNA. Tumors were excised and cryosectioned. Ten-micrometer
frozen sections were first treated with TUNEL assay kit for marking apoptotic cells
and then immunostained stained with vWF- and VE-Cadherin mAbs for marking
tumor endothelial cells.
2.8. Tumor growth inhibition studies
A suspension of B16F10 melanoma cells in Hank’s buffer salt solution (HBSS) was
prepared (1.5  106 cells/mL). Each of the 6–8 weeks old female C57BL/6 J mice
(weighing 20–22 g) was injected subcutaneously with 0.1 mL of the cell suspension
into the left flanks on day 0. Mice were then randomly sorted into five groups and
 S. Samanta et al. / Biomaterials 31 (2010) 1787–1797
Fig. 3. RGDGWK-lipopeptide 1: p53 complex, upon intravenous administration in C57
BL/6 J mice, shows significant tumor inhibition properties. A. 6–8 weeks old female C57
BL/6 J mice (each weighing 20-22 g) with aggressive B16F10 tumors (produced by
subcutaneous injections of 1.5  105 B16F10 cells in 100 mL Hank’s buffer salt solution
(HBSS) into the left flanks on day 0) were randomly sorted into five groups and each
group (n ¼ 5) was administered intravenously with: RGDGWK-lipopeptide 1:p53 lipo
plex in 5% aqueous glucose containing 50 mg DNA with lipopeptide:DNA charge ratio of
6:1 per injection (filled diamonds); RGDLFK-lipopeptide 2:p53 lipoplex in 5% aqueous
glucose containing 50 mg DNA with lipopeptide:DNA charge ratio of 6:1 per injection
(filled squares); liposomes of RGDGWK-lipopeptide 1 alone (stars) and 5% aqueous
glucose alone (filled triangles) on day 5, 7, 9, 11 and 13. Tumor volumes (V ¼ 1/2.ab2
where, a ¼ maximum length of the tumor and b ¼ minimum length of the tumor
measured perpendicular to each other) were measured with a slide calipers for up to
23 days. Results represent the means þ/ SD (for n ¼ 5 tumors) (*P < 0.005 compared
to RGDLFK/p53); B. Representative samples of B16F10 tumors excised on day 23. I.
tumors treated with RGDGWK-liposome/p53 complexes; II. tumors treated with
RGDLFK-liposome/p53 complexes; III. tumor treated with vehicle only; and IV. tumors
treated with RGDGWK-liposome only.
each group (n ¼ 5) was administered intravenously with different lipo
peptide:p53gene complexes (containing 50 mg DNA with lipopeptide:DNA charge
ratio of 6:1 per injection) or control liposomes (without p53) or naked p53 on day 5,
7, 9, 11 and 13 after the inoculation of tumor cells. Tumor volumes (V ¼ 1/2  ab2
where, a ¼ maximum length of the tumor and b ¼ minimum length of the tumor
measured perpendicular to each other) were measured with a slide calipers for up to
23 days. The mice were euthanized and the tumors were collected at the end of each
in vivo experiment. All the in vivo experiments were performed in accordance with
the Institutional Bio-Safety and Ethical Committee Guidelines using an approved
animal protocol.
2.9. Statistical analysis
Error bars represent mean values  SEM. The statistical significance of the
experiments was determined by two-tailed Student’s test. *P < 0.05 were considered
statistically significant.
1793
3. Results
3.1. a5b1 Integrin selectivity of RGDGWK-lipopeptide 1
Toward evaluating whether the RGDGWK-lipopeptide 1 delivers
genes to cells selectively/preferably via any specific integrin
receptor, we first tested the efficiencies of the RGDGWK-lipo-
peptide 1 in delivering the p-CMV-SPORT-b-Gal reporter gene into
both untreated A549 cells (express significant amounts of all the
three avb3, avb5 and a5b1 integrin receptors as demonstrated by
the FACS profile shown in Fig. 2A) as well as in A549 cells pretreated
with mAbs against avb3, avb5 and a5b1 integrins. The gene
transfer efficiency of the RGDGWK-lipopeptide 1 was affected most
(by more than 50%) when the cells were pre-incubated with anti-
a5b1 integrin mAb (Fig. 2B). The transfection efficiency of the
RGDGWK-lipopeptide 1 was affected by lesser degree in cells pre-
treated with monoclonal antibodies against the integrins avb3 and
avb5 (Fig. 2B). Thus, the findings summarized in Fig. 2B are
consistent with the proposition that RGDGWK-lipopeptide 1
transfects cells more selectively via a5b1 integrin receptors.
3.2. Role of the amino acids lying C-terminal to RGD
Studies on phage display libraries previously showed that the
specificity and the affinity of integrin binding for RGD peptides are
remarkably dependent on the nature of the amino acid residues
lying C-terminal to RGD [16,17]. Based on this prior findings, it is not
unlikely that the more a5b1 integrin selective transfection prop-
erties of the presently described RGDGWK-lipopeptide 1 (Fig. 2B) is
critically governed by the amino acids lying C-terminal to the RGD.
Toward confirming this, we designed and synthesized a represen-
tative control RGDLFK-lipopeptide 2 in which the ‘‘GW’’ dipeptide
of the RGDGWK-lipopeptide 1 was replaced with ‘‘LF’’ dipeptide
and evaluated its transfection properties in both untreated A549
cells and in A549 cells treated with mAbs against avb3, avb5 and
a5b1 integrins. Unlike RGDGWK-lipopeptide 1, the control RGDLFK-
lipopeptide 2 was found to be significantly less efficient in trans-
fecting A549 cells (Fig. 2C). In sharp contrast to the more a5b1
integrin selective transfection properties of the RGDGWK-lipo
peptide 1 (Fig. 2B), the transfection property of the control RGDLFK-
lipopeptide 2 remained essentially unaffected in both untreated
A549 cells and in A549 cells pretreated with mAbs against avb3,
avb5 and a5b1 integrins (Fig. 2C). Thus, consistent with the findings
in phase display studies [16,17], the GW dipeptide sequence lying
C-terminal to RGD in RGDGWK-lipopeptide 1 are likely to play
a dominant role in imparting more a5b1 integrin selective trans-
fection properties to the presently described RGDGWK-lipopeptide
1. The observed insensitivity of the control RGDLFK-lipopeptide 2 to
any of the a5b1, avb3 and avb5 integrins (Fig. 2C) is also consistent
with the previously reported finding that RGD peptides lacking the
tryptophan residue are weak inhibitors of the binding of mAb 16
(the anti-human a5 monoclonal antibody) to a5b1 integrin
compared to their structural analog with tryptophan [18].
3.3. Tumor growth inhibition studies
Toward evaluating the therapeutic potential of the RGDGWK-
lipopeptide 1, next we studied the tumor growth inhibition prop-
erties of the RGDGWK-lipopeptide 1:CMV-p53 plasmid complex by
intravenously administering the complex in C57BL/6J mice bearing
aggressive murine B16F10 melanoma tumor. Remarkable inhibition
of tumor growth was achieved only when the anti-cancer
pCMV-p53 gene was administered in complexation with RGDGWK-
lipopeptide 1 (Fig. 3A). Toward probing the role of the RGDGWK-
lipopeptide 1 alone, if any, behind the observed tumor growth
1794 S. Samanta et al. / Biomaterials 31 (2010) 1787–1797

Fig. 4. RGDGWK-lipopeptide 1 selectively targets genes to tumor vasculature. A & C. Immunohistochemical staining of two representative tumor sections. On day 23rd after tumor
inoculation, mice were intravenously injected with the complex of RGDGWK-lipopeptide 1 and a5-GFP plasmid DNA. After 24 h, tumors were excised, sectioned and the repre-
sentative tumor sections were immunostained with both anti-vWF and anti-VE-cadherin mAbs (tumor endothelial cell markers). B & D. RGDGWK-lipopeptide 1 targets genes to
tumor neovasculature. The stained tumor slides shown in parts A & C when observed in the same positions under fluorescent microscope (at 10x magnification) using a green filter
revealed GFP expression in the tumor vasculatures. Bar ¼ 2.0 mm.

inhibition, we intravenously administered only the liposomes of
the RGDGWK-lipopeptide 1 in mice (without being in complexation
with any anti-cancer gene). As depicted in Fig. 3A & 3B, essentially
no tumor growth inhibition was observed upon intravenous
administration of the liposome of the RGDGWK-lipopeptide 1
alone. This finding convincingly demonstrate that the observed
tumor growth inhibition property of the RGDGWK-lipopeptide
1:p53 (Fig. 3A and B) is unlikely to originate from any inherent anti-
cancer properties of the RGDGWK-lipopeptide 1 itself. Similarly, no
tumor growth inhibition was observed when naked p53 plasmid
was intravenously administered alone (data not shown) presum-
ably due to degradation by serum nucleases while in circulation
under naked condition. Stated differently, the remarkable tumor
growth inhibition property of the RGDGWK-lipopeptide 1:p53
(Fig. 3A & 3B) is consistent with the supposition that RGDGWK-
lipopeptide 1 stabilize the anti-cancer p53 gene by forming
a compact electrostatic complex while under circulation. Impor-
tantly, the degree of tumor growth inhibition was significantly less
when mice were intravenously administered with the p-CMV-p53
gene in complexation with the control RGDLFK-lipopeptide 2
 S. Samanta et al. / Biomaterials 31 (2010) 1787–1797 1795

Fig. 5. Tumor growth inhibition property of the RGDGWK-lipopeptide 1:p53 complex is mediated through apoptosis of tumor endothelial cells. Two female C57 BL/6 J mice were
subcutaneously implanted with B16F10 melanoma cells (1.5  105 cells) and the tumors were allowed to grow to w3000 mm3. Two mice were then given single intravenous
injection of RGDGWK-lipopeptide 1:p53 complex. One mouse was sacrificed 24 h post injection and the tumor was excised, cryosectioned, fixed and the fixed frozen sections were
treated with TUNEL assay kit for marking apoptotic cells (2nd panels from the left, green). Subsequently, the same tumor cryosections were immunounostained with mAb against
VE-cadherin for marking tumor endothelial cells (3rd panels from left, red). The 4th panels (from the left) show superimposed images (yellow). The stained tumor slides were
observed in the same positions under fluorescent microscope (10X magnification) using green and red filters. The 1st panels from the left in both images (24 & 72 h) show the tissue
architecture when observed in bright field. Bar ¼ 2.0 mm.

(Fig. 3A–B). Mice intravenously administered with vehicle alone
(5% aqueous glucose solution) developed large tumor on day 23
(Fig. 3A) and were sacrificed at that point.
3.4. Tumor vasculature targetability of RGDGWK-lipopeptide 1
With a view to probe whether the RGDGWK-lipopeptide 1 is
capable of targeting genes to tumor vasculatures under systemic
settings, aggressive B16F10 murine melanoma tumors were
produced in 6–8 weeks old female C57BL/6J mice (each weighing
20–22 g). On day 22 after tumor cell inoculation, when large
tumors were grown, mice were intravenously administered with
the electrostatic complexes of RGDGWK-lipopeptide 1 and p-a5GFP
(plasmid DNA encoding green fluorescence protein). After 24 h,
tumors were excised, cryosectioned and ten-micrometer frozen
sections of tumors were immunostained with both anti-vWF and
anti-VE-cadherin monoclonal antibodies (markers of tumor endo-
thelial cells). Fixed tumor cryosections immunostained with anti-
vWF antibodies revealed presence of the tumor vasculatures when
observed in bright fields (Fig. 4A & C). The same tumor cryosections
when observed under fluorescent microscope revealed GFP
expression in only those areas stained with blood vessel markers
(Fig. 4A–D). Thus, the findings summarized in Fig. 4A–D, taken
together, convincingly demonstrated the ability of the RGDGWK-
lipopeptide 1 in delivering genes selectively to tumor vasculatures.
3.5. TUNEL assay & immunohistochemical studies
Since p53 gene is a well known apoptosis inducer and since the
presently described RGDGWK-lipopeptide 1 is efficient in targeting
1796 S. Samanta et al. / Biomaterials 31 (2010) 1787–1797
genes to tumor vasculatures (Fig. 4), we envisaged that the tumor
growth inhibition properties of the RGDGWK-lipopeptide 1:p53
complex might be induced by apoptois of the tumor endothelial cells.
With a view to address this mechanistic issue, we implanted (s.c.)
B16F10 murine melanoma cells in female C57BL/6J mice and the
tumors were allowed to grow and become vascularized for 15 days.
At this point, RGDGWK-lipopeptide 1:p53 complexes were intrave-
nously administered in two mice. One mouse was sacrificed 24 h
post injection and the tumor was excised, cryosectioned, fixed and
the fixed frozen sections were treated with TUNEL assay kit for
marking apoptotic cells. Subsequently, the same tumor crysections
were immunounostained with vascular endothelial (VE)-cadherin-
specific antibodies to identify tumor vasculatures. The TUNEL-posi-
tive cells (i.e. cells undergoing apoptosis) were found to be co-lo-
calized with tumor endothelial cells across the entire cryosections
when they were prepared 24 h post injection (Fig. 5). The same
processes when repeated using cryosections prepared 72 h post
injection from the second mouse showed presence of some TUNEL-
positive cells not co-localized with tumor endothelial cells (Fig. 5).
4. Discussion
Angiogenesis, the sprouting of new blood vessels from pre-
existing vessels, is one of the distinguished factors contributing to
tumor malignancy [21]. Prevention of new blood vessel formation
around tumor tissues is a promising anti-angiogenic therapeutic
approach. One modality of this approach is based on selective tar-
geting of anti-cancer drugs/genes to tumor vasculatures via integ-
rin receptors over expressed on the surface of the tumor
endothelial cells [22]. An overwhelming majority of studies has
exploited either linear or cyclic forms of RGD-based peptides for
targeting avb3 integrin receptors over expressed on tumor endo-
thelial cells [7,23–25]. An elegant strategy for identifying high
affinity integrin ligands is based on phage display studies. Selecting
for phage that specifically home to tumor vasculatures upon
injection into mice has provided a panel of peptides capable of
homing to tumor vasculatures via binding to angiogenesis-related
molecules such as, integrins, expressed on the surface of the tumor
endothelial cells [13,26,27]. Ever since the reports on the distinct
proangiogenic roles of the a5b1 integrins [28–30] were disclosed,
efforts toward developing efficient a5b1 integrin receptor ligands
and antagonists are gaining increasing importance [31–35]. To this
end, cyclic peptides such as, ACRGDGWCG & ACDCRGDCFCG, were
synthesized based on phage sequences that bound with high
affinity to a5b1 integrin [17]. These latter findings emphasized the
importance of introducing conformational constrains in the design
of efficient ligands for a5b1 integrin.
As to the mechanistic insights into the high binding affinities of
integrin a5b1 to CRGDGWC peptides, Humphries and coworkers
showed that mutation of Trp157 of a5 subunit to alanine resulted
into loss of high affinity binding of the a5b1 integrin to CRGDGWC
[18]. Presumably, the Trp157 of a5 subunit participates in a favorable
hydrophobic interaction with the tryptophan of the CRGDGWC
resulting into its high affinity binding with a5b1 integrin. Very
recently, we have demonstrated that a lipopeptide containing
a non-cyclic conformationally unstrained simple RGDK tetrapep-
tide sequence in its polar head-group region can selectively target
genes to proangiogenic a5b1 integrin receptors and can target
genes to mouse tumor vasculatures [19]. These findings prompted
us to study the integrin receptor selectivity and tumor vasculature
targeting property of lipopeptide having just the ‘‘RGDGW’’ motif
(i.e. without the terminal cysteine residues of the a5b1 integrin
binding ‘‘CRGDGWC’’ peptide identified from phage display studies
mentioned above) in its head-group region. With this question in
mind, we synthesized RGDGWK-lipopeptide 1 (Scheme I, Fig. 1). An
important synthetic issue needs to be mentioned at this point of
discussion. We synthesized the RGDGWK-lipopeptide 1 in conju-
gate fashion, i.e. by coupling the protected RGD tripeptide unit (VIII,
Scheme I, Fig. 1) with the protected GWK-lipopeptide unit IV
(Scheme I, Fig. 1) and not by directly coupling the protected
RGDGWK-OH to the amino group of the N-2-aminoethyl-N,N-di-n-
hexadecylamine. In the latter route, while preparing the RGDGWK-
OH via selective allyl deprotection of the benzyloxycarbonyl pro-
tected RGDGWK-O-Allyl with Pd(PPh3)4/N-methylmorpholine
(NMM), an undesired Aspartimide product containing a cyclic
imide bond between the aspartic acid side chain carbonyl and
glycine NH residues was obtained (the formation of which was
presumably promoted by NMM used as the allyl cation scavenger
during selective allyl deprotection, data not shown).
Interestingly, the in vitro gene transfer efficiencies of RGDGWK-
lipopeptide 1 in A549 cells (expresses all the three widely exploited
a5b1, avb3 and avb5 integrins, Fig. 2A) was affected most when the
cells were pre-incubated with mAb against a5b1 integrins (Fig. 2B).
This finding is consistent with the supposition that the RGDGW motif
(without the presence of the flanking cysteines residues in
CRGDGWC) too exhibits stronger affinity for a5b1 integrin receptors
compared to its affinity for the avb3 and avb5 integrins. It is not clear
at this stage of investigation whether the decreased transfection
efficiency of the RGDGWK-lipopeptide 1 in A549 cells pre-incubated
with mAb against a5b1 integrins (Fig. 2B) originates from lower DNA
uptake compared to those in cells pretreated with mAbs against avb3
and avb5 integrins. Future DNA uptake studies aimed at measuring
cell associated DNA quantities upon incubation of antibody treated
cells with each lipoplexes are likely to throw more insights to this end.
From the therapeutic application point of view, next we studied
the tumor growth inhibition properties of the RGDGWK-lipopeptide
1:p53 complex. Importantly, the complex when intravenously
administered in C57 BL/6J mice bearing the aggressive B16F10
melanoma tumors showed remarkable tumor growth inhibition
(Fig. 3). The finding that intravenous administration of the control
RGDLFK-lipopeptide 2:p53 complex showed significantly less
tumor growth inhibition demonstrated the importance of the role of
the GW dipeptide sequence lying C-terminal to the RGD sequence of
the RGDGWK-lipopeptide 1 under systemic settings too. Colocali-
zation of cells expressing green fluorescence and tumor endothelial
cells markers (Fig. 4) provided compelling evidence that the
RGDGWK-lipopeptide 1 targets genes to tumor vasculatures.
Colocalization of the VE-cadherin positive cells (markers of tumor
endothelial cells) and TUNEL-positive cells (marker of cells under-
going apoptosis) across the entire tumor cryosections prepared 24 h
post injection of the RGDGWK-lipopeptide 1: p53 complexes
together with the presence of few TUNEL-positive apoptotic cells
with no co-localized tumor vasculatures (absence of VE-cadherin
positive cells) in tumor cryosections prepared at 72 h (Fig. 5)
supports the notion that apoptosis of tumor cells presumably begins
after initiation of apoptosis of tumor endothelial cells. The degree of
tumor growth inhibition with RGDGWK-lipoplexes was found to be
higher than that with the control RGDLFK-lipoplexes (Fig. 3) sug-
gesting that the selectivity for tumor vasculature is better with the
former. To this end, it is worth mentioning here that although the
immunohistochemical findings summarized in Fig. 4 provided
evidence for selective tumor vasculature targeting ability of the
RGDGWK-lipopeptide 1, they do not provide insights into why
RGDGWK-lipopeptide 1 is more tumor vasculature selective than
the control RGDLFK-lipopeptide 2.
5. Conclusions
We have demonstrated that a lipopeptide with just the RGDGW
motif (without the flanking cysteine groups present in many
 S. Samanta et al. / Biomaterials 31 (2010) 1787–1797
ligands isolated from phage display library) covalently attached to
the lysine residue of a monolysinylated cationic amphiphile
(RGDGWK-lipopeptide 1) can deliver genes to cultured cells pref-
erably via a5b1 integrins. Importantly, remarkable tumor growth
inhibition was observed when the electrostatic complex of the
RGDGWK-lipopeptide 1 and the anti-cancer p53 gene was intra-
venously administered in C57BL/6J mice bearing the aggressive
B16F10 tumor. Immunohistochemical staining of mice tumor cry-
osections with vasculature markers combined with monitoring
expression of the green fluorescence protein in the same tumor
cryosections revealed that the RGDGWK-lipopeptide 1 is capable of
targeting genes to tumor vasculatures. The colocalization of the
TUNEL (terminal deoxyuridine triphosphate nick-end labeling,
a widely used marker of apoptosis) and VE-cadherin (markers of
tumor endothelial cells) positive cells in tumor cryosections
support the notion that the remarkable tumor growth inhibition
property of the RGDGWK-lipopeptide 1:p53 complex is initiated
through apoptosis of the tumor endothelial cells.
Acknowledgements
AC gratefully acknowledges the financial support received from
the Council of Scientific and Industrial Research, Government of
India, New Delhi (NWP 0036) for this work. We thank Prof. Robert.
J. Debs for providing us the p-CMV-p53 plasmids as a gift and Mr.
Dipankar Pramanik for his assistance in tumor growth inhibition
experiments. SS thanks the Council of Scientific and Industrial
Research, Government of India, New Delhi for providing doctoral
research fellowships.
Appendix
Figures with essential color discrimination. Figs. 2–5 in this
article are difficult to interpret in black and white. The full color
images can be found in the on-line version, at doi:10.1016/
j.biomaterials.2009.10.027.
Appendix. Supplementary data
Supplementary data associated with this article can be found in
the on-line version, at doi:10.1016/j.biomaterials.2009.10.027.
References
[1] Hynes RO. Integrins: bidirectional, allosteric signaling machines. Cell
2002;110:673–87.
[2] Stupack DJ, Cheresh DA. Integrins and angiogenesis. Curr Top Dev Biol
2004;64:207–38.
[3] Ruoslahti E. RGD and other recognition sequences for integrins. Annu Rev Cell
Dev Biol 1996;12:697–715.
[4] Xiong X, Huang Y, Lu WL, Zhang X, Zhang H, Nagai T, et al. Intracellular delivery
of doxorubicin with RGD-modified sterically stabilized liposomes for an
improved antitumor efficacy: in vitro and in vivo. J Pharm Sci 2005;94:1782–93.
[5] Koppitz M, Huenges M, Gratias R, Kessler H. Synthesis of unnatural lipohilic N-
(9H-Fluoren-9-ylmethoxy)carbonyl-substituted a-amino acids and their
incorporation into cyclic RGD-peptides: a structure-activity study. Helv Chim
Acta 1997;80:1280–300.
[6] Hu B, Finsinger D, Peter K, Guttenberg Z, Bärmann M, Kessler H, et al. Inter-
vesicle cross-linking with integrin alpha IIb beta 3 and cyclic-RGD-lipopeptide.
A model of cell-adhesion processes. Biochemistry 2000;39:12284–94.
[7] Hood JD, Bednarski M, Frausto R, Guccione S, Reisfeld RA, Xiang R, et al. Tumor
regression by targeted gene delivery to the neovasculature. Science
2002;296:2404–7.
[8] Murphy EA, Majeti BK, Barnes LA, Makale M, Weis SM, Lutu-Fuga K, et al.
Nanoparticle-mediated drug delivery to tumor vasculature suppresses
metastasis. Proc Natl Acad Sci U S A 2008;105:9343–8.
1797
[9] Nasongkla N, Shuai X, Ai H, Weinberg BD, Pink J, Boothman DA, et al. cRGD-
functionalized polymer micelles for targeted doxorubicin delivery. Angew
Chem Int Ed Engl 2004;43:6323–7.
[10] Aota S, Nomizu M, Yamada KM. The short amino acid sequence Pro-His- Ser-
Arg-Asn in human fibronectin enhances cell-adhesive function. J Biol Chem
1994;269:24756–61.
[11] Bowditch RD, Hariharan M, Tommina EF, Smith JW, Yamada KM, Getzoff ED,
et al. Identification of a novel integrin binding site in fibronectin. Differential
utilization by beta 3 integrins. J Biol Chem 1994;269:10856–63.
[12] Tan K, Casasnovas JM, Liu J, Briskin MJ, Springer TA, Wang J. The structure of
immunoglobulin superfamily domains 1 and 2 of MAdCAM-1 reveals novel
features important for integrin recognition. Structure 1998;6:793–801.
[13] Arap W, Pasqualini R, Ruoslahti E. Cancer treatment by targeted drug delivery
to tumor vasculature in a mouse model. Science 1998;279:377–80.
[14] Kraft S, Diefenbach B, Mehta R, Jonczyk A, Luckenbach GA, Goodman S.
Definition of an unexpected ligand recognition motif for alphav beta6 integrin.
J Biol Chem 1999;274:1979–85.
[15] Koivunen E, Gay D, Ruoslahti E. Selection of peptides binding to the alpha
5 beta 1 integrin from phage display library. J Biol Chem 1993;268:
20205–10.
[16] Koivunen E, Wang B, Ruoslahti E. Isolation of a highly specific ligand for the
alpha 5 beta 1 integrin from a phage display library. J Cell Biol 1994;124:
373–80.
[17] Koivunen E, Wang B, Ruoslahti E. Phage libraries displaying cyclic peptides
with different ring sizes: ligand specificities of the RGD-directed integrins. Bio/
Technology 1995;13:265–70.
[18] Humphries JD, Askari JA, Zhang X-P, Takada Y, Humphries MJ, Mould AP.
Molecular basis of ligand recognition by integrin alpha5beta 1. II. Specificity of
arg-gly-Asp binding is determined by Trp157 OF THE alpha subunit. J Biol
Chem 2000;275:20337–45.
[19] Pramanik D, Majeti BK, Mondal G, Karmali PP, Ramakrishna S, Ovula GR, et al.
Lipopeptide with a RGDK tetrapeptide sequence can selectively target genes to
proangiogenic a5b1 integrin receptor and mouse tumor vasculature. J Med
Chem 2008;51:7298–302.
[20] Karmali PP, Valluripalli VK, Chaudhuri A. Design, syntheses and in vitro gene
delivery efficacies of novel mono-, di- and trilysinated cationic lipids:
a structure-activity investigation. J Med Chem 2004;47:2123–32.
[21] Carmeliet P. Angiogenesis in life, disease and medicine. Nature 2005;438:
932–6.
[22] Temming K, Schiffelers RM, Molema G, Kok RJ. RGD based strategies for
selective delivery of therapeutics and imaging agents to the tumor vascula-
ture. Drug Resist Updat 2005;8:381–402.
[23] Pastorino F, Brignole C, Marimpietri D, Cilli M, Gambini C, Ribatti D, et al.
Vascular damage and anti-angiogenic effects of tumor vessel-targeted lipo-
somal chemotherapy. Cancer Res 2003;63(21):7400–9.
[24] Schiffelers RM, Koning GA, ten Hagen TL, Fens MH, Schraa AJ, Janssen AP, et al.
Anti-tumor efficacy of tumor vasculature-targeted liposomal doxorubicin.
J Control Release 2003;91(1–2):115–22.
[25] Murphy EA, Majeti BK, Barnes LA, Makale M, Weis SM, Lutu-Fuga K, et al.
Nanoparticle-mediated drug delivery to tumor vasculature suppresses
metastasis. Proc Natl Acad Sci U S A 2008;105(27):9343–8.
[26] Pasqualini R, Ruoslahti E. Organ targeting in vivo using phage display peptide
libraries. Nature 1996;380(6572):364–6.
[27] Ruoslahti E, Rajotte D. An address system in the vasculature of normal tissues
and tumors. Annu Rev Immunol 2000;18:813–27.
[28] Kim S, Bell K, Mousa SA, Varner JA. Regulation of angiogenesis in vivo by
ligation of integrin alpha5beta1 with the central cell-binding domain of
fibronectin. Am J Pathol 2000;156:1345–62.
[29] George EL, Georges-Labouesse EN, Patel-King RS, Rayburn H, Hynes RO.
Defects in mesoderm, neural tube and vascular development in mouse
embryos lacking fibronectin. Development 1993;119:1079–91.
[30] Yang JT, Rayburn H, Hynes RO. Embryonic mesodermal defects in alpha 5
integrin-deficient mice. Development 1993;119:1093–105.
[31] Heckmann D, Meyer A, Marinelli L, Zahn G, Stragies R, Kessler H. Probing
integrin selectivity: rational design of highly active and selective ligands for
the a5b1 and avb3 intrgrin receptor. Angew Chem Int Ed Engl 2007;46:
3571–4.
[32] Stragies R, Osterkamp F, Zischinsky G, Vossmeyer D, Kalkhof H, Reimer U, et al.
Design and synthesis of a new class of selective integrin alpha5beta1 antag-
onists. J Med Chem 2007;50:3786–94.
[33] Meyer A, Auernheimer J, Modlinger A, Kessler H. Targeting RGD recognizing
integrins: drug development, biomaterial research, tumor imaging and tar-
geting. Curr Pharm Des 2006;12:2723–47.
[34] Smallheer JM, Weigelt CA, Woerner FJ, Wells JS, Daneker WF, Mousa SA, et al.
Synthesis and biological evaluation of nonpeptide integrin antagonists con-
taining spirocyclic scaffolds. Bioorg Med Chem Lett 2004;14:383–7.
[35] Zimmermann D, Guthöhrlein EW, Malesević M, Sewald K, Wobbe L,
Heggemann C, et al. Integrin a5b1 ligands: biological evaluation and confor-
mational analysis. Chem Bio Chem 2005;6:272–6.