GENE DELIVERY

Journal of Controlled Release 143 (2010) 215–221

Contents lists available at ScienceDirect

Journal of Controlled Release

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j c o n r e l

Tumor regression after systemic administration of a novel tumor-targeted gene

delivery system carrying a therapeutic plasmid DNA
Swati Koppu a, Yew Jinn Oh a, RuAngelie Edrada-Ebel a, David R. Blatchford a, Laurence Tetley b,
Rothwelle J. Tate a, Christine Dufès a,⁎
a
Strathclyde Institute of Pharmacy and Biomedical Sciences, University of Strathclyde, 27 Taylor Street, Glasgow G4 0NR, United Kingdom
b
Faculty of Biomedical and Life Sciences, University of Glasgow, Glasgow G12 8QQ, United Kingdom

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

Article history: The possibility of using genes as medicines to treat cancer is limited by the lack of safe and efficacious
Received 27 August 2009 delivery systems able to deliver therapeutic genes selectively to tumors by intravenous administration. We
Accepted 15 November 2009 investigate if the conjugation of the polypropylenimine dendrimer to transferrin, whose receptors are
Available online 26 November 2009
overexpressed on numerous cancers, could result in a selective gene delivery to tumors after intravenous
administration, leading to an increased therapeutic efficacy. The objectives of this study are to evaluate the
Keywords:
Cancer therapy
targeting and therapeutic efficacies of a novel transferrin-bearing polypropylenimine dendrimer.
Transferrin The intravenous administration of transferrin-bearing polypropylenimine polyplex resulted in gene
Tumor targeting expression mainly in the tumors. Consequently, the intravenous administration of the delivery system
Gene delivery complexed to a therapeutic DNA led to a rapid and sustained tumor regression over one month, with long-
Polypropylenimine dendrimer term survival of 100% of the animals (90% complete response, 10% partial response).The treatment was well
tolerated by the animals, with no apparent signs of toxicity. Transferrin-bearing polypropylenimine may thus
be a promising gene delivery system for cancer therapy.
© 2009 Elsevier B.V. All rights reserved.

1. Introduction has previously been used successfully as a tumor-targeting ligand for

many drug and gene delivery systems [4–9].
The possibility of using genes as medicines to treat cancer is The objectives of this study are therefore 1) to prepare and
currently limited by the lack of safe and efficacious delivery systems characterize a novel transferrin-bearing generation 3 polypropyleni-
able to deliver therapeutic genes selectively to tumors by intravenous mine dendrimer (DAB-Tf) and 2) to evaluate in vitro and in vivo the
administration, without secondary effects to healthy tissues [1]. therapeutic and targeting efficacies of this therapeutic system.
With the long-term aim of developing an efficacious cancer-targeted
gene medicine, we have recently demonstrated that systemically 2. Materials and methods
administered generation 3-diaminobutyric polypropylenimine dendri-
mer (DAB) combined with Tumor Necrosis Factor (TNFα) expression 2.1. Cell lines and reagents
plasmid driven by a tumor-specific promoter, leads to tumor regression
in a number of murine models, with excellent long-term response [2]. Quanti-iT™ PicoGreen® dsDNA reagent and tissue culture media
Although the expression of the therapeutic genes was tumor-specific, were obtained from Invitrogen (Paisley, UK). Vectashield® mounting
this gene delivery system was widely distributed in the body, which medium with 4′,6-diamidino-2-phenylindole (DAPI) was obtained
may have reduced the therapeutic potential of this system. from Vector Laboratories (Peterborough, UK). Passive lysis buffer was
In order to further improve the tumor delivery capability of this purchased from Promega (Southampton, UK). Label IT® Cy3 Nucleic
system, we hypothesize that the conjugation of DAB dendrimer to the Acid Labeling kit was from Cambridge Biosciences (Cambridge, UK).
iron-carrier transferrin, whose receptors are overexpressed on N-[1-(2,3-Dioleoyloxy) propyl]-N,N,N-trimethylammonium methyl-
numerous cancer cell lines [3], could result in a selective receptor- sulfate (DOTAP) liposomal transfection reagent was purchased from
mediated gene delivery to tumors after intravenous administration Roche (Burgess Hill, UK). Polypropylenimine dendrimer generation
and therefore lead to an increased therapeutic efficacy. Transferrin 3 (PPI G3; DAB-Am16), transferrin and all other chemicals were pur-
chased from Sigma Aldrich (Poole, UK). A431 epidermoid carcinoma
and T98G glioblastoma were obtained from the European Collection
of Cell Cultures (Salisbury, UK). The expression plasmids encoding
⁎ Corresponding author. Tel.: + 44 141 548 3796; fax: + 44 141 552 2562. Tumor necrosis factor (TNF)α (pORF9-mTNFα) and β-galactosidase
E-mail address: [email protected] (C. Dufès). (pCMVsport β-galactosidase) were obtained respectively from

0168-3659/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jconrel.2009.11.015
GENE DELIVERY
216 S. Koppu et al. / Journal of Controlled Release 143 (2010) 215–221

InvivoGen (San Diego, CA) and Invitrogen (Paisley, UK). They were mer:DNA weight ratio 5:1) served as positive controls. DNA
purified using an Endotoxin-free Giga Plasmid Kit (Qiagen, Hilden, concentration (1 µg/well) was kept constant for all the formulations
Germany). tested. After 72 h incubation, cells were lysed with 1× passive lysis
buffer (PLB) (50 µL/well) for 20 min. The cell lysates were subse-
2.2. Synthesis and characterization of transferrin-bearing quently analyzed for β-galactosidase expression [11]. Briefly, 50 µL of
polypropylenimine dendrimer the assay buffer (2 mM magnesium chloride, 100 mM mercaptoetha-
nol, 1.33 mg/mL ο-nitrophenol-β-galactopyranoside, 200 mM sodium
2.2.1. Conjugation of transferrin to DAB dendrimer phosphate buffer, pH 7.3) was added in each well containing
Polypropylenimine dendrimer generation 3 (DAB) was conjugated the lysates. After 2 h incubation at 37 °C, the absorbance of the sam-
to transferrin (Tf) by using dimethylsuberimidate (DMSI) as a cross- ples was read at 405 nm with a plate reader (Thermo Lab Systems,
linking agent in a similar manner to that reported for transferrin- Multiscan Ascent, UK).
bearing vesicles [10]. Polypropylenimine dendrimer DAB (24 mg) was
added to transferrin (6 mg) and DMSI (12 mg) in triethanolamine HCl 2.3.3. Cellular uptake
buffer (pH 7.4, 2 mL). The coupling reaction was allowed to take place Imaging of the cellular uptake of the DNA carried by DAB-Tf was
at room temperature for 2 h whilst stirring. The final product was carried out using confocal microscopy. Labeling of plasmid DNA with
purified by size exclusion chromatography using a Sephadex G75 the fluorescent probe Cy3 was performed using a Label IT® Cy3
column and freeze-dried. DAB-Tf was then dissolved in D2O at a Nucleic Acid Labeling kit, as described by the manufacturer. A431 and
concentration of 5 mg/mL. The grafting of the transferrin to DAB was T98G cells were grown on microscope slides (0.6 × 106 cells/90-mm
assessed by 1H NMR spectroscopy, using a Jeol Oxford NMR AS 400 Petri dish) at 37 °C for 24 h. The cells were then incubated for further
spectrometer. 24 h with Cy3-labeled DNA (6 µg/dish) complexed to DAB-Tf and DAB
at the polymer:DNA weight ratios giving their highest transfection
2.2.2. Characterization of polyplex formation by PicoGreen® assay efficacy (DAB-Tf:DNA: 10:1; DAB:DNA: 5:1 [11]). Control slides were
The degree of DNA accessibility following complexation with DAB-Tf treated with naked DNA or remained untreated. The slides were then
was assessed by PicoGreen® assay, performed according to the washed with PBS and fixed in methanol for 30 min. Upon staining of
supplier's protocol. The fluorescence of PicoGreen® significantly the nuclei with DAPI, the cells were examined using a Leica TCS SP5
increases on intercalation with double stranded DNA. The electrostatic confocal microscope. DAPI was excited with the 405 nm laser line
interaction between the anionic DNA and cationic group of the polymer (bandwidth: 415–491 nm), whereas Cy3 was excited with the
on formation of the Tf-bearing DAB–DNA polyplex condenses the DNA 453 nm laser line (bandwidth: 550–620 nm).
and reduces the number of PicoGreen® binding sites, ultimately
reducing the fluorescence intensity for the PicoGreen® solution. 2.3.4. In vitro anti-proliferative activity
PicoGreen® reagent was diluted 200-fold in Tris–EDTA (TE) buffer Anti-proliferative activity of transferrin-bearing DAB complexed
(10 mM Tris, 1 mM EDTA, pH 7.5) on the day of the experiment. One mL with plasmid DNA encoding TNFα was assessed in A431 and T98G
of complex polymer–DNA at various polymer:DNA weight ratios cancer cell lines. Cells (2 × 103 cells per well in 96-well plates seeded
(20:1, 10:1, 5:1, 2:1, 1:1, 0.5:1 and 0:1) was added to 1 mL PicoGreen® 72 h prior treatment) were incubated for 72 h with the DNA formu-
solution. The DNA concentration in the cuvette (10 µg/mL) was kept lations at final concentrations of 10− 3 to 10 000 µg/mL. Anti-
constant throughout the experiment. The fluorescence intensity of the proliferative activity was evaluated by measurement of the growth
complexes in the presence of PicoGreen® was analyzed at various inhibitory concentration for 50% of the cell population (IC50) in a
time points with a Varian Cary Eclipse Fluorescence spectrophotometer standard MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium
(λexc: 480 nm, λem: 520 nm). Results were represented as percentage of bromide] assay. Absorbance was measured at 570 nm using a plate
DNA condensation (=100 − % relative fluorescence to DNA control) reader. Dose–response curves were fitted to percentage absorbance
(n = 4). values to obtain IC50 values (three independent experiments with n = 5
for each concentration level).
2.2.3. Polyplex size and zeta potential measurement
Size and zeta potential of Tf-bearing DAB complexed to DNA were 2.4. In vivo study
measured by photon correlation spectroscopy, using a Malvern
Zetasizer Nano series Nano-ZS (Malvern Instruments, Malvern, UK). 2.4.1. Animals
Female immunodeficient BALB/c mice were housed in groups of
2.3. In vitro biological characterization five at 19 °C to 23 °C with a 12-hour light–dark cycle. They were fed a
conventional diet (Rat and Mouse Standard Expanded, B&K Universal,
2.3.1. Cell culture Grimston, United Kingdom) with mains water ad libitum. Experimen-
A431 and T98G cell lines overexpressing Tf receptors were grown tal work was carried out in accordance with UK Home Office
as monolayers in Dulbecco's Modified Eagle Medium (DMEM) regulations and approved by the local ethics committee.
supplemented with 10% (v/v) fetal bovine serum, 1% (v/v) L-
glutamine and 0.5% (v/v) penicillin–streptomycin. Cells were cultured 2.4.2. Biodistribution of gene expression
at 37 °C in a 5% carbon dioxide atmosphere. Groups of mice (n = 5, initial mean weight 18 g) bearing sub-
cutaneously implanted A431 tumors overexpressing transferrin
2.3.2. In vitro transfection receptors, were treated intravenously with a single injection of Tf-
Transfection efficacy of the DNA carried by the transferrin-bearing bearing and control DAB dendrimers carrying β-galactosidase
dendrimer was assessed with a plasmid coding for β-galactosidase expression plasmid (50 µg of DNA). Mice were sacrificed 24 h after
(pCMV βgal), using a β-galactosidase transfection assay. A431 and injection. Their organs were removed, immediately frozen in liquid
T98G cells were seeded in quintuplicate at a density of 2000 cells/well nitrogen and analyzed for their β-galactosidase levels as previously
in 96-well plates. After 72 h incubation, the cells were treated with described [12].
DAB-Tf complexed to plasmid DNA encoding β-galactosidase, at the
polymer:DNA weight ratios used for the DNA condensation experi- 2.4.3. In vivo tumoricidal activity
ment. Naked DNA served as a negative control, formulations of Mice (n = 5, initial mean weight 18 g) bearing well-established,
DOTAP–DNA (polymer:DNA weight ratio 5:1) and DAB–DNA (poly- vascularized subcutaneous A431 tumors, were treated by intravenous
 GENE DELIVERY
S. Koppu et al. / Journal of Controlled Release 143 (2010) 215–221 217

injection with the Tf-bearing DAB dendrimer carrying TNFα expres-

sion plasmid, the non-targeted dendrimer carrying the same
therapeutic plasmid, the targeted dendrimer either alone or carrying
a non-therapeutic β-galactosidase expression plasmid and naked
DNA. Gene therapy treatment was administered by intravenous tail
vein injection (50 µg of DNA) once daily for ten days. Animals were
weighed daily and tumor volume was determined by caliper
measurements (volume = d3 × π/6). Results were expressed as rela-
tive tumor volume (rel. Voltx = Voltx / Volt0) and responses classified
analogous to Response Evaluation Criteria in Solid Tumors (RECIST,
[13]). Progressive disease is defined as an increase in relative tumor
volume higher than 1.2-fold, stable disease as a relative volume
between 0.7 and 1.2 of starting volume, partial response as measure-
able tumor with a reduction of more than 30% (0–0.7) and complete
response as the absence of any tumor.

2.5. Statistical analysis

Results were expressed as means ± standard error of the mean

(S.E.M). Statistical significance was determined by one-way analysis
of variance (ANOVA) followed by the Bonferroni multiple compar-
ison post-test (GraphPad Prism software). Differences were consid-
ered as significant when P b 0.05.

3. Results and discussion

3.1. Conjugation of transferrin to DAB dendrimer

The synthesis of DAB-Tf was confirmed by 1H NMR spectrum, as

follows (Fig. 1).
1
H NMR (D2O): δ DAB (H2N-CH2-CH2) = 2.65; DAB (N-CH2-CH2) =
2.45; DAB (CH2-CH2-CH2) = 1.64; DAB-Tf (CH2-CH2-NH-CO) = 3.70;
DAB-Tf (H2N-CH2-CH2) = 2.80; DAB-Tf (N-CH2-CH2) = 2.55; DAB-
Tf (CH2-CH2-CH2) = 1.64; DAB-Tf = 2.90–2.40 (m); 2.25 (t); 1.20–
Fig. 1. 1H NMR spectra of generation 3 polypropylenimine dendrimer conjugated to
1.90 (m).
transferrin (DAB-Tf) (A) (B: 1H NMR spectra of the unmodified dendrimer).
The NMR spectrum of DAB had a characteristic broad triplet peak
for the CH2 adjacent to peripheral primary amino group at δ 2.65,
which was slightly shifted to 2.80 ppm in the NMR spectrum of DAB- PicoGreen® assay. Transferrin-bearing DAB was able to condense
Tf. However, in the 1H NMR spectrum of DAB-Tf, an additional broad more than 70% of the DNA, independently of the weight ratio tested
triplet at δ 3.70 was observed which is compatible to the methylene (Fig. 2A). DNA condensation occurred almost instantaneously and was
protons adjacent to an amide unit. This indicated that some of the found to be stable over at least 24 h. It increased with increasing
peripheral amino groups had reacted with the carboxyl group of the weight ratios, reaching a plateau near 85% at a dendrimer:DNA weight
DMSI to form an amide linkage with transferrin. Multiple small peaks ratio of 0.5:1 and an almost complete condensation at a dendrimer:
between δ 1.30 and 3.40 corresponded to protons for transferrin. DNA weight ratio of 20:1. These results demonstrated that complexes
These results demonstrated that DAB has been successfully conjugat- can be formed through electrostatic interactions between DNA and
ed to transferrin by cross-linking with DMSI. DAB-Tf, although an excess of dendrimer is required to ensure
1
H NMR spectra showed that the dendrimer is 50% bound to efficient DNA condensation. The conjugation of transferrin to DAB did
transferrin as signified by the ratio of the integrals of resonances at δ not destabilize DNA condensation, which is a prerequisite for the
3.70 and 2.80 for methylene units (e and a) attached to the amide- transport of this macromolecule. The formation of nanoparticles of
linked transferrin and free amine, respectively. transferrin-bearing DAB complexed to DNA was also demonstrated by
This synthesis is a simple one-step procedure unlike the synthetic transmission electron microscopy (Fig. 1 in Supplementary data). A
methods previously reported, which generally involve dicyclohexylcar- gel retardation assay confirmed the complete and partial DNA
bodiimide (DCC). DCC is widely used for amide and ester formation. condensation obtained respectively for polymer:DNA weight ratios
However, its by-product N, N-dicyclohexyl urea can only be removed by of 10:1 and 1:1, as demonstrated by the DNA condensation
lengthy processes such as filtration followed by extensive dialysis. Using experiment (Fig. 2 in Supplementary data).
dimethylsuberimidate as a crosslinker avoids this problem.
3.3. Polyplex size and zeta potential measurement
3.2. Characterization of polyplex formation by PicoGreen® assay
Transferrin-bearing DAB polyplex displayed a z-average mean
The polypropylenimine dendrimer DAB can efficiently form diameter of 287 nm (polydispersity index: 0.393). The conjugation of
complexes with plasmid DNA through electrostatic interactions transferrin to the periphery of DAB led to a slight increase of the
between its protonated primary amines and the negatively charged polyplex size compared to the unmodified DAB polyplex, which had an
phosphodiester groups of the DNA. However, the conjugation of average size of 196 nm (polydispersity index: 0.683). As the cut-off size
transferrin may affect the dendrimer ability to complex DNA. For that for extravasation has been found to be 400 nm for most tumors [14,15],
reason, we examined the influence of the conjugation of transferrin on this delivery system has the required properties to access the tumor
the dendrimer ability to form complexes with DNA, by using a cells. Zeta-potential experiments demonstrated that transferrin-bearing
GENE DELIVERY
218 S. Koppu et al. / Journal of Controlled Release 143 (2010) 215–221

level after treatment with DAB-Tf at various dendrimer:DNA ratios was

obtained at a DAB-Tf:DNA weight ratio of 10:1 on T98G (Fig. 2B) and
A431 cells (Fig. 2C). The conjugation of Tf to DAB led to an improved
transfection compared to the unmodified DAB and DOTAP on both the
tested cell lines. On T98G cells, gene expression following treatment
with DAB-Tf (12.27.10− 3 ± 0.27.10− 3 U/mL) was 1.3 times higher than
that of unmodified DAB polyplex (9.05.10− 3 ± 0.25.10− 3 U/mL) and 2.8
times higher than that of DOTAP–DNA (4.38.10− 3 ± 0.21.10− 3 U/mL).
On A431 cells, gene expression following treatment with DAB-Tf
polyplex (11.07.10−3 ± 0.37.10− 3 U/mL) was 2.2 times higher than
that of unmodified DAB polyplex (5.07.10− 3 ± 0.21.10− 3 U/mL) and 3
times higher than that of DOTAP–DNA (3.67.10− 3 ± 0.16.10− 3 U/mL).
No gene expression was observed after treatment with control DNA, as
expected.
The improved β-gal expression induced by Tf-bearing DAB com-
pared to unmodified DAB most likely resulted from the transferrin-
specific uptake by the cancer cells overexpressing Tf receptors. The
higher zeta potential of the DAB polyplex would have otherwise
resulted in a higher transfection efficacy of the unmodified dendrimer,
due to the strong correlation between cellular uptake and positive
charge density of polyplexes [16]. This improved β-gal expression was
in accordance with that observed with other Tf-bearing gene delivery
systems [17,18].

Fig. 2. In vitro quantitative characterization of DAB-Tf polyplex: A) DNA condensation of

DAB-Tf polyplex using PicoGreen® reagent at various durations and polymer:DNA weight
ratios: 20:1 (■), 10:1 (●), 5:1 (▲), 2:1 (▼), 1:1 (♦), 0.5:1 (◄) and DNA only (►). Results are
expressed as mean± SEM (n = 4). B, C) Transfection efficacy of DAB-Tf polyplex at various
polymer:DNA weight ratios relative to DOTAP–DNA and native DAB–DNA in T98G (B) and
A431 cells (C). DOTAP–DNA and DAB–DNA were dosed at their optimal carrier:DNA ratio
of 5:1 (n = 15). ⁎: P b 0.05 vs the highest transfection treatment.

and control DAB polyplexes were bearing a positive surface charge. The
zeta potential of DAB-Tf polyplex was 1.03 mV, smaller than that of the
unmodified DAB polyplex (6.42 mV). This decrease is most likely due to
the presence of the negatively charged amino acids of transferrin and
would lead to a minimization of the non-specific interactions of the
polyplex with the negatively charged cellular membranes [16].

3.4. In vitro transfection Fig. 3. Confocal microscopy imaging of the cellular uptake of Cy3-labeled DNA (6 µg/
dish) either complexed with DAB-Tf (A, B), DAB (C, D) or free in solution (E, F) after
incubation for 24 h with T98G (left) and A431 cells (right). Blue: nuclei stained with
Transfection efficacy was determined by quantifying the expression DAPI (excitation: 405 nm laser line, bandwidth: 415–491 nm), green: Cy3-labeled DNA
of β-galactosidase encoded on the plasmid. The highest transfection (excitation: 453 nm laser line, bandwidth: 550–620 nm) (Magnification: ×40).
 GENE DELIVERY
S. Koppu et al. / Journal of Controlled Release 143 (2010) 215–221 219

3.5. Cellular uptake Similar improvements have been obtained by Kircheis et al. [19]
when using Tf-bearing polyethylenimine (PEI). The authors demon-
The uptake of Cy3-labeled DNA by A431 and T98G cells was strated that Tf-bearing PEI polyplexes led to a specific expression in
qualitatively confirmed using confocal microscopy (Fig. 3). Co- the tumor, whereas expression in the lungs and the other organs was
localization of DNA in the nuclei was clearly visible in both cell lines dramatically reduced compared to the non-targeted PEI.
treated with Tf-bearing DAB polyplex, as well as in A431 cells after Transferrin receptors are expressed in a range of cancer cells, but
treatment with DAB polyplex. It was however much weaker after also on rapidly growing normal cells. The combination of active
treatment with DAB polyplex in T98G cells. Cy3-labeled DNA was also targeting, based on the use of ligands, and passive targeting, based on
disseminated in the cytoplasm after treatment with targeted and non- the accumulation of particulate delivery systems due to the enhanced
targeted formulations in both cell lines. By contrast, cells treated with permeability and retention effect [20], provides a tumor-selective
free DNA did not show any Cy3-derived fluorescence. The grafting of targeting strategy.
Tf to DAB therefore seems to improve DNA uptake by T98G cells, but
has little influence on A431 cell line, for which the nuclear uptake of
DNA seems to be similar after treatment with both targeted and non- 3.8. In vivo tumoricidal activity
targeted dendrimers.
Tf-targeted DAB dendrimer complexed with TNFα plasmid induced
3.6. In vitro anti-proliferative activity a tumor regression within 24h compared to untreated tumors (Fig. 5).

The conjugation of transferrin to DAB led to an increase of the in vitro

anti-proliferative activity compared to the unmodified polyplex (IC50 of
7.09 ± 0.88 µg/mL and 9.47± 1.15 µg/mL respectively for the targeted
polyplex and for the non-targeted polyplex in A431 cells, 21.72 ±
2.68 µg/mL and 29.84 ± 2.79 µg/mL respectively for the targeted
polyplex and for the non-targeted polyplex in T98G cells). DNA alone
did not exert any cytotoxicity to the cells, thus emphasizing the need of a
gene delivery system for the delivery of therapeutic DNA to cancer cells.
These results may be attributed to the improved transfection efficacy
when treated with Tf-bearing DAB polyplex. Although the improved
transfection levels after treatment with Tf-bearing DAB polyplex at a
polymer:DNA ratio of 10:1 were very close for the two cell lines, the
anti-proliferative activity was more pronounced on A431 than on T98G
cells, probably due to different sensitivities of the cell lines toward
TNFα-mediated apoptosis.

3.7. Biodistribution of gene expression

The intravenous administration of control DAB polyplex resulted in

gene expression mainly in the liver (42.4 ± 8.7 mU β-galactosidase per
organ) followed by the tumor to a lesser extent (25.8 ± 5.6 mU β-
galactosidase per organ) (Fig. 4). By contrast, the grafting of transferrin
to DAB led to gene expression mostly in the tumor (38.1 ±1.3 mU β-
galactosidase per tumor, significantly higher than that obtained after
treatment with control DAB polyplex), with a β-galactosidase amount
4-fold higher than in the liver (8.9± 1.1 mU β-galactosidase per organ).
The β-galactosidase amounts in the other organs were dramatically
reduced (less than 4 mU β-galactosidase per organ), thus confirming
the targeting efficacy of transferrin to the tumors.

Fig. 5. A) Tumor growth studies in a well-established mouse A431 xenograft model after
intravenous administration of transferrin-bearing DAB polyplex carrying plasmid DNA
encoding TNFα (50 µg/injection) (green), non-targeted DAB polyplex (orange), uncom-
plexed DAB-Tf (dark blue), DAB-Tf polyplex carrying a non-therapeutic DNA encoding β-
galactosidase (pale blue), naked DNA (red) and untreated tumors (black). Treatment was
administered by intravenous tail vein injection once daily for ten days. Results were
expressed as relative tumor volume (rel. Voltx = Voltx / Volt0) (n = 5). B) Magnification of
Fig. 4. Biodistribution of gene expression after a single intravenous administration of Tf- the above graph. C) Tumor histology at Day 6 after treatment with transferrin-bearing DAB
bearing DAB polyplex (light grey) and non-targeted DAB polyplex (dark grey) (50 µg polyplex carrying plasmid DNA encoding TNFα (left), non-targeted DAB polyplex (partial
DNA administered). Results were expressed as milliunits β-galactosidase per organ tumor: whole tumor size: 3.5 mm× 4.5 mm) (middle) and untreated tumors (partial
(n = 5). *: P b 0.05: Tf-bearing polyplex vs unmodified polyplex for each organ. tumor: whole tumor size: 9 mm× 10 mm) (right) (H & E staining) (bar: 1 mm).
GENE DELIVERY
220 S. Koppu et al. / Journal of Controlled Release 143 (2010) 215–221

This treatment led to a marked tumor regression over the following non-therapeutic DNA encoding β-galactosidase (pale blue) led to tumor
month and long-term survival of 100% of the animals. On the last day of growth. On the last day of the experiment, 100% of the tumors treated
the experiment, 90% of the tumors treated with Tf-bearing DAB polyplex with Tf-bearing DAB only or carrying β-galactosidase expression
had completely disappeared, while the remaining 10% showed a partial plasmid, free DNA or left untreated were progressive. Histological
response (Fig. 6B). Treatment with non-targeted DAB complex led to sections of tumors taken on Day 6 showed widespread necrosis, a
tumor regression as well, but with a less successful outcome (40% known mechanism for the antitumor effects of TNFα, after treatment
complete response, 60% partial response). Although the injection of free with Tf-bearing and non-targeted DAB polyplex and a pronounced
DAB dendrimer led to a marked growth retardation on the same cell line reduction of the size of the tumor after treatment with the targeted
due to the intrinsic antitumor activity of the dendrimer [2], the grafting polyplex (Fig. 5C).
of transferrin to this dendrimer seems to inhibit this effect for these As a result, the extended survival of the mice treated with targeted
experimental settings. The administration of DAB-Tf alone or carrying a DAB polyplex and unmodified DAB polyplex was 34 days compared to
untreated mice (Fig. 6C). Interestingly, treatment with DAB-Tf
uncomplexed with DNA led to an extended survival of the mice of
16 days compared to untreated animals. The treatments were well
tolerated by the mice and no apparent signs of toxicity or significant
weight loss were observed (Fig. 6A). These therapeutic effects,
together with the lack of apparent toxicity, potentially make the
transferrin-bearing polypropylenimine dendrimer an interesting gene
delivery system as part of a gene medicine.
In conclusion, novel transferrin-bearing DAB polypropylenimine
dendrimer led to an improved tumor specificity of gene expression and
sustained tumor regression after intravenous administration without
visible secondary effect to the mice. Transferrin-bearing DAB dendrimer
is therefore a highly promising delivery system for cancer therapy.

Acknowledgments

This work was supported by a Tenovus Scotland grant to Christine

Dufès.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in

the online version, at doi:10.1016/j.jconrel.2009.11.015.

References
[1] D. Luo, W.M. Saltzman, Synthetic DNA delivery systems, Nat. Biotechnol. 18
(2000) 33–37.
[2] C. Dufès, W.N. Keith, A. Bisland, I. Proutski, I.F. Uchegbu, A.G. Schätzlein, Synthetic
anticancer gene medicine exploits intrinsic antitumor activity of cationic vector to
cure established tumors, Cancer Res. 65 (2005) 8079–8084.
[3] A. Calzolari, I. Oliviero, S. Deaglio, G. Mariani, M. Biffoni, N.M. Sposi, F. Malavasi, C.
Peschle, U. Testa, Transferrin receptor 2 is frequently expressed in human cancer
cell lines, Blood Cells Mol. Dis. 39 (2007) 82–91.
[4] R. Kircheis, L. Wightman, A. Schreiber, B. Robitza, V. Rossler, M. Kursa, E. Wagner,
Polyethylenimine/DNA complexes shielded by transferrin target gene expression
to tumors after systemic administration, Gene Ther. 8 (2001) 28–40.
[5] R. Kircheis, E. Ostermann, M. Wolschek, C. Lichtenberger, C. Magin-Lachmann, L.
Wightman, M. Kursa, E. Wagner, Tumor-targeted gene delivery of tumor necrosis
factor-α induces tumor necrosis and tumor regression without systemic toxicity,
Cancer Gene Ther. 9 (2002) 673–680.
[6] C. Dufès, J.M. Muller, W. Couet, J.C. Olivier, I.F. Uchegbu, A.G. Schätzlein, Anticancer
drug delivery with transferrin targeted polymeric chitosan vesicles, Pharm. Res. 21
(2004) 101–107.
[7] M. Nakase, M. Inui, K. Okumura, T. Kamei, S. Nakamura, T. Tagawa, p53 gene
therapy of human osteosarcoma using a transferrin-modified cationic liposome,
Mol. Cancer Ther. 4 (2005) 625–631.
[8] T.R. Daniels, T. Delgado, G. Helguera, M.L. Penichet, The transferrin receptor part II:
targeted delivery of therapeutic agents into cancer cells, Clin. Immunol. 121
(2006) 159–176.
[9] M. Hong, S. Zhu, Y. Jiang, G. Tang, Y. Pei, Efficient tumor targeting of hydroxycamp-
tothecin loaded PEGylated niosomes modified with transferrin, J. Control. Release
133 (2009) 96–102.
[10] C. Dufès, A.G. Schätzlein, L. Tetley, A.I. Gray, D.G. Watson, J.C. Olivier, W. Couet, I.F.
Uchegbu, Niosomes and polymeric chitosan based vesicles bearing transferrin and
glucose ligands for drug targeting, Pharm. Res. 17 (2000) 1250–1258.
[11] B.H. Zinselmeyer, S.P. Mackay, A.G. Schätzlein, I.F. Uchegbu, The lower-generation
Fig. 6. In vivo studies (continued). A) Variations of the animal body weight throughout
polypropylenimine dendrimers are effective gene-transfer agents, Pharm. Res. 19
the treatment regime (n = 5) (Color coding as in Fig. 5). B) Overall tumor response to (2002) 960–967.
treatments on the last day of the experiment, stratified according to change in tumor [12] B.H. Zinselmeyer, N. Beggbie, I.F. Uchegbu, A.G. Schätzlein, Quantification of beta-
volume: progressive disease (increase N1.2-fold, red), stable disease (0.7–1.2, orange), galactosidase activity after non-viral transfection in vivo, J. Control. Release 91
partial response (0–0.7, yellow) and complete response (0, green) over the duration of (2003) 201–208.
the experiment (40 days). C) Time to disease progression. Animals were removed from [13] P. Therasse, S.G. Arbuck, E.A. Eisenhauer, J. Wanders, R.S. Kaplan, L. Rubinstein, J.
the study once their tumor reached 12 mm diameter (Color coding as in Fig. 5). Verweij, M. Van Glabbeke, A.T. Van Oosterom, M.C. Christian, S.G. Gwyther, New
 GENE DELIVERY
S. Koppu et al. / Journal of Controlled Release 143 (2010) 215–221 221

guidelines to evaluate the response to treatment in solid tumors, European [17] C. Tros, N. de Ilarduya, Dϋzgϋneş, Efficient gene transfer by transferrin lipoplexes
Organization for Research and Treatment of Cancer, National Cancer Institute of in the presence of serum, Biochim. Biophys. Acta 1463 (2000) 333–342.
the United States, National Cancer Institute of Canada, J. Natl. Cancer Inst. 92 [18] M. Kursa, G.F. Walker, V. Roessler, M. Ogris, W. Roedl, R. Kircheis, E. Wagner, Novel
(2000) 205–216. shielded transferrin-polyethylene glycol-polyethylenimine/DNA complexes for
[14] F. Yuan, M. Dellian, D. Fukumura, M. Leunig, D.A. Berk, V.P. Torchilin, R.K. Jain, systemic tumor-targeted gene transfer, Bioconjug. Chem. 14 (2003) 222–231.
Vascular permeability in a human tumor xenograft: molecular size dependence [19] R. Kircheis, L. Wightman, E. Ostermann, E. Wagner, Tumor-targeted gene delivery:
and cutoff size, Cancer Res. 55 (1995) 3752–3756. an attractive strategy to use highly active effector molecules in cancer treatment,
[15] T. Dutta, N.K. Jain, N.A.J. McMillan, H.S. Parekh, Dendrimer nanocarriers as Gene Ther. 9 (2002) 731–735.
versatile vectors in gene delivery, Nanomedicine, NBM (2009), doi:10.1016/j. [20] H. Maeda, The tumor blood vessel as an ideal target for macromolecular anticancer
nano.2009.05.005. agents, J. Control. Release 19 (1992) 315–324.
[16] R.I. Mahato, L.C. Smith, A. Rolland, Pharmaceutical perspectives of nonviral gene
therapy, Adv. Genet. 41 (1999) 95–156.