Enhanced gene expression in epithelial cells transfected with amino acid-substituted gemini nanoparticles

Article abstract of DOI:10.1016/j.ejpb.2010.04.007

Gemini surfactants are versatile gene delivery agents because of their ability to bind and compact DNA and their low cellular toxicity. Through modification of the alkyl tail length and the chemical nature of the spacer, new compounds can be generated wit

Full text of DOI:10.1016/j.ejpb.2010.04.007

European Journal of Pharmaceutics and Biopharmaceutics 75 (2010) 311–320
Contents lists available at ScienceDirect
European Journal of Pharmaceutics and Biopharmaceutics
journal homepage: www.elsevier.com/locate/ejpb
Research paper
Enhanced gene expression in epithelial cells transfected with
amino acid-substituted gemini nanoparticles
Peng Yang a, Jagbir Singh b, Shawn Wettig c, Marianna Foldvari c, Ronald E. Verrall a, Ildiko Badea b,
*
a Department of Chemistry, University of Saskatchewan, Saskatoon, Canada
b Drug Design and Discovery Research Group, College of Pharmacy and Nutrition, University of Saskatchewan, Saskatoon, Canada
c School of Pharmacy, University of Waterloo, Waterloo, Canada
a r t i c l e i n f o
a b s t r a c t
Article history:
Gemini surfactants are versatile gene delivery agents because of their ability to bind and compact DNA
and their low cellular toxicity. Through modification of the alkyl tail length and the chemical nature of
the spacer, new compounds can be generated with the potential to improve the efficiency of gene deliv-
ery. Amino acid (glycine and lysine) and dipeptide (glycyl-lysine and lysyl-lysine) substituted spacers of
gemini surfactants were synthesized, and their efficiency of gene delivery was assessed in epithelial cells
for topical cutaneous and mucosal applications.
Three different epithelial cell lines, COS-7, PAM212 and Sf 1Ep cells, were transfected with plasmid
DNA encoding for interferon gamma and green fluorescent protein complexed with the amino acid-
substituted gemini compounds in the presence of 1,2 dioleyl-sn-glycero-phosphatidyl-ethanolamine as
a helper lipid. Gene expression was quantified by ELISA. Size, zeta potential and circular dichroism mea-
surements were used to characterize the plasmid–gemini (PG) and plasmid–gemini surfactant–helper
lipid (PGL) complexes.
Received 30 October 2009
Accepted in revised form 16 April 2010
Available online 22 April 2010
Keywords:
Gene therapy
In vitro transfection
Non-viral
Epithelial cells
Nanoparticles
Delivery efficiency
Gene expression was found to increase up to 72 h and then declined by the 7th day. In general, the gly-
cine-substituted surfactant showed consistently high gene expression in all three cell lines. Results of
physicochemical and spectroscopic studies of the complexes indicate that substitution of the gemini
spacer does not interfere with compaction of the DNA.
The superior performance of these spacer-substituted gemini surfactants might be attributed to their
better biocompatibility compared to the surfactants possessing unsubstituted spacers.
Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction
shown to be versatile vectors for non-viral gene delivery [5–8].
This stems from the fact that there are many possibilities, through
Topical gene delivery is an important approach in the quest to
find treatment or cure for a multitude of genetic and acquired dis-
orders, such as scleroderma, epidermolysis bullosa, cystic fibrosis
and others. To correct the imbalances created by a missing or
incorrect gene, efficient non-invasive gene delivery systems should
chemical modification of the alkyl tails and the spacer group of
these compounds, to modulate their physicochemical behaviour
in a way that can enhance the efficiency of gene transfer. For
example, modification of the length, degree of unsaturation and
substitution of different functional groups in the spacer and alkyl
tails can provide opportunities to tailor their chemical structure
for specific needs [9,10].
Cationic gemini surfactants are capable of binding DNA [11]
and have several advantages compared to classic monovalent sur-
factants: lower cellular toxicity [7], lower critical micelle concen-
tration (cmc), generally one or two orders of magnitude higher
efficiency in reducing surface tension and greater tendency to
self-assemble [12]. The higher order assembly of the gemini sur-
factants depends on the shape of the molecule (i.e., nature of the
tail and spacer) and ranges from micellar to inverted micellar or
bilayer structures. In the presence of the DNA, these supramolec-
ular particles become more complex, resulting in polymorphic
phase behaviour with the appearance of inverted micellar and
be employed. Interferon gamma (IFN-c) is a secreted cytokine with
immunomodulatory, antiviral, antifibrotic and antitumour activity.
Based on these properties, its therapeutic potential in conditions
such as scleroderma [1], atopic dermatitis [2] and post-operatory
fibrosis [3] have been discussed, and delivery of IFN-
c gene into
epidermal skin and mucosal epithelial tissue can be considered
for these applications. Gemini surfactants, N,N-bis(dimethylalkyl)-
a,x-alkanediammonium halide derivatives (Fig. 1) [4], have been
* Corresponding author. College of Pharmacy and Nutrition, University of
Saskatchewan, 110 Science Place, Thorvaldson Building, Room G22, Saskatoon,
Saskatchewan, Canada S7N 5C9. Tel.: +1 306 966 6349.
E-mail address: ildiko.badea@usask.ca (I. Badea).
0939-6411/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ejpb.2010.04.007

312
P. Yang et al. / European Journal of Pharmaceutics and Biopharmaceutics 75 (2010) 311–320
protect the genetic material against intracellular degradation,
R
N
have the ability to undergo polymorphic phase changes and un-
dergo facile release of the plasmid once inside the cell to effect
protein expression. Gemini surfactants with C3–C12 alkyl spacers
are able to compact the DNA into 100–200 nm particles with
positive surface charge [5], which enables interaction with the
cell surface and endocytosis. The transfection efficiency of these
agents depends on the spacer length, the 3-carbon spacer (12-
3-12) being the most efficient. CD spectra of the plasmid–gemini
surfactant complexes have previously been discussed in terms of
the formation of highly compact Wꢀ DNA [5]. Addition of a helper
lipid, 1,2 dioleyl-sn-glycero-phosphatidyl-ethanolamine (DOPE),
induces polymorphic phase behaviour [13], which correlates with
increased transfection efficiency [5]. Cellular toxicity of the
transfection complexes was significantly lower compared to the
commercial Lipofectamine Plus [5]. Substitution of the alkyl
spacer with pH-sensitive imino groups increases the transfection
efficiency of the gemini surfactants [8]. The 1,9-bis(dodecyl)-
1,1,9,9-tetramethyl-5-imino-1,9-nonanediammonium dibromide
surfactant (12-7NH-12) transfected cells at a ninefold higher level
compared to the alkyl spacer derivatives. This improvement was
attributed to its pH sensitivity and ability to form multiple
phases, which enable membrane fusion and release of the DNA
in the cells [8].
N
N
R = Glycine
Lysine
(G)
(K)
Glycine-lysine (GK)
Lysine-lysine (KK)
Counter ion: Cl
Fig. 1. General structure of gemini surfactants. The two hydrophobic alkyl chains
(tail) with charged head groups (ion) are connected with a carbon spacer [4].
cubic structures [13]. To be efficient transfection agents, they
have to fulfil several criteria: ability to bind DNA and create com-
plexes of a certain size and morphology, the vector–plasmid com-
plexes must retain an overall positive surface charge so as to
facilitate interaction with the negative cell surface and enhance
cellular uptake (in the absence of specific targeting), induce
changes in the DNA structure favourable for efficient delivery,
We report here the results of a study focusing on the modifica-
tion of this derivative by chemically coupling amino acid moieties
to the imino spacer so as to further increase and prolong gene
expression. It has been shown that small molecular weight pep-
NH2 HCl
O
H
N
N
N
Cl
Cl
N
N
N
1
c
a
3
NHBoc
O
N
N
N
12- 7N(Gly)-12
2
ClHH2N
d
NH2HCl
O
NHBoc
HN
f
BocHN
O
O
g
NH2
N
N
N
e
NH
O
O
Cl
Cl
N
N
N
N
N
N
6
4
5
12- 7N(Lys-Gly)-12
Fig. 2. Synthetic scheme #1: (a) Boc-glycine, EEDQ, CHCl3; (b) 1-bromododecane, MeCN; (c) (i) TFA, CHCl3; (ii) Clꢀ ion exchange; (d) (i) TFA, CHCl3; (ii) NaOH, EtOH; (e) Boc-
Lys(Boc)-OH,EEDQ,CHCl3; (f) 1-bromododecane, MeCN; (g) (i) TFA, CHCl3; (ii) Clꢀ ion exchange.

P. Yang et al. / European Journal of Pharmaceutics and Biopharmaceutics 75 (2010) 311–320
NH2HCl
313
O
ClHH2N
N
N
N
Cl
Cl
NHBoc
O
BocHN
8
N
N
N
12- 7N(Lys)-12
h
7
H
N
N
N
NH2 HCl
1
ClH H2N
O
NH
NHBoc
O
ClH H2N
BocHN
l
N
N
N
NHCbz
O
m
n
O
Cl
Cl
NH
o
BocHN
O
BocHN
N
N
N
N
N
N
9
11
12-7 N(Lys-Lys)-12
10
Fig. 3. Synthetic scheme #2: (h) Boc-Lys(Boc)-OH, EEDQ, CHCl3; (f) 1-bromododecane, MeCN; (g) (i) TFA, CHCl3; (ii) Clꢀ ion exchange; (k) Cbz-Lys(Boc)-OH, EEDQ,CHCl3; (l)
HCOONH4, 10% Pd/C; (m) Boc-Lys(Boc)-OH, EEDQ,CHCl3; (n) 1-bromododecane, MeCN; (o) (i) TFA, CHCl3; (ii) Clꢀ ion exchange.
tides such as arginine-rich protamine fragments are able to bind
and condense DNA into 100-nm particles. Transfection of cells re-
sulted in prolonged gene expression for up to 12 days [14]. We
hypothesized that an additive effect could be achieved by coupling
the highly efficient imino-substituted spacer of a gemini surfactant
with amino acid/dipeptide residues.
bath. The specific conductance was measured after each addition
of an aqueous solution of concentrated surfactant into water under
a nitrogen atmosphere, allowing for the solution to equilibrate.
2.3. Preparation of the gene delivery systems
The pGTꢁIFN-GFP plasmid was used for the transfection experi-
ments. The murine interferon gamma (IFN-c) and green fluores-
2. Materials and methods
cent protein (GFP) genes were inserted into the pGT backbone as
a bicistronic system [5]. Transfection agents of plasmid/gemini sur-
factant (PG) complexes and plasmid/gemini surfactant/helper lipid
(PGL) particles were prepared as previously described [5]. The plas-
mid:gemini surfactant charge ratio was 1:10. DOPE was used as a
helper lipid.
2.1. Synthesis of amino acid/dipeptide-substituted gemini surfactants
Boc-lys-OH, Boc-lys(Cbz)-OH and 3,30-iminobis (N,N-dimethyl-
propylamine) were purchased from Sigma–Aldrich; Boc-gly-OH
and EEDQ were purchased from Alfa Aesar. All solvents used were
of the highest grade and purchased from Sigma–Aldrich. These
chemicals were used without further purification. 1H NMR spec-
tra were recorded in either CDCl3 or CD3OD using a 500 MHz Bru-
ker spectrometer, and mass spectra were obtained using a QSTAR
XL MS/MS System.
2.4. Transfection of COS-7, PAM212 and Sf 1Ep cells using amino acid/
dipeptide-substituted gemini surfactants and DOPE
COS-7 African green monkey kidney fibroblasts (ATCC,
CRL-1651), PAM212 murine keratinocytes (kindly provided by Dr.
S. Yuspa, NCI) and Sf 1Ep cottontail rabbit epithelial cells (ATCC,
CCL-68) were seeded on Falcon 96-well tissue culture plates (BD,
Mississauga, ON) at a density of 2 ꢂ 105 cells/well. Three plates
were seeded with each cell line at the same passage number after
initiation of the cultures. Transfection experiments were carried
out using 100 ng plasmid DNA/well. Supernatants were collected
every 24 h and replaced with fresh cell culture medium. The results
are the average of three plates of triplicate wells.
2.2. Specific conductance – determination of the critical micelle
concentrations
Specific conductivities were measured as previously described
using a Wayne-Kerr precision component analyzer (Model 6425)
operating at 1.5 kHz, with a Tacussel electrode having a cell con-
stant of 1.11 cmꢀ1 [15]. Experimental temperatures were main-
tained at 25 0.05 °C using a Haake (Model F3) circulating water

314
P. Yang et al. / European Journal of Pharmaceutics and Biopharmaceutics 75 (2010) 311–320
2.4.1. ELISA
Step b: Synthesis of 1,9-bis(dodecyl)-1,1,9,9-tetramethyl-5-
ELISA was used to detect IFN-
c
secreted into the supernatant
(N0-Boc-glycyl)amino-1,9-nonanediammonium dibromide.
and was performed according to the BD Pharmingen protocol. Pro-
tein concentration was calculated from a standard curve using re-
combinant IFN-c.
One gram (2.9 mmol) of (2) and 3.0 g (12 mmol) of 1-bromodo-
decane were dissolved in 20 mL of dry MeCN, and the reaction
mixture was refluxed at 60 °C for 3 days under a N2 atmosphere.
The solvent was removed under vacuum, and a large excess of
diethyl ether was added to the residue and left overnight to make
sure that the fine, suspended solid would settle. The organic phase
was removed under vacuum. This process was repeated four times
to ensure that all of the 1-bromododecane had been removed. The
white solid was collected and dried under vacuum to obtain 1.85 g
of product (yield, 76%). 1H NMR(CDCl3) d/ppm, 5.59 (brs, 1H),
4.11–3.99 (m, 2H), 3.81–3.59 (m, 6H), 2.55–3.41 (m, 4H), 3.39–
3.25(s, 14H), 2.34 (brs, 2H), 2.18 (brs, 2H), 1.69 (brs, 4H), 1.38 (s,
9H), 1.34–1.14 (m, 36H), 0.83 (brs, 6H); ESI-MS m/z: 341.3349
[M]2+/2.
2.5. Size and zeta potential measurements
Size and zeta potential measurements of PG and PGL complexes,
prepared at 1:10 plasmid:gemini surfactant charge ratio, the same
as for the transfection experiments, were carried out using a
Zetasizer Nano ZS instrument (Malvern Instruments, Worcester-
shire, UK). The results reported are the average of triplicate mea-
surements on a % volume basis. Size measurements of gemini
compounds at two concentrations above the cmc were also carried
out on % volume basis.
2.6. Circular dichroism
Step c: Synthesis of 12-7N(Gly)-12 (3).
PG and PGL complexes were prepared at the same charge and
(i) De-protection of (3) was carried out by dissolving 1.0 g in
30 mL of CHCl3 and stirring at 0 °C for 30 min followed by
addition of 10 mL of TFA, drop-wise. The solution was
removed from the ice bath, and the reaction was contin-
ued at room temperature for 24 h. The solvent was
removed under vacuum.
molar ratio as for transfection experiments, using 20
lg/mL plas-
mid DNA. Spectra were recorded using
P
ꢃ180 instrument (Applied
Photo Physics, UK), with 2-nm slit under a N2 atmosphere.
2.7. Statistics
(ii) The bromide salt of (3) was converted to chloride using
Amberlite IRA-400(Cl) ion exchange resin. The chloride
salt was then dissolved in 50 mL of methanol, and 3 g
of charcoal was added. The mixture was refluxed for
30 min, cooled to room temperature and stirred over-
night. The charcoal was removed by filtration, and the
solution was reduced to a small volume, and the product
was salted-out with acetone. This was repeated several
times to yield 0.2 g of compound (3), 1,9-bis(dodecyl)-
1,1,9,9-tetramethyl-5-glycylamino-1,9-nonanediammo-
nium dichloride, hydrochloride salt (yield, 24%). 1H NMR
(CD3OD) d/ppm, 4.11 (brs, 2H), 3.75–3.23 (m, 14H), 3.24–
3.06 (m, 12H), 2.31–2.02 (m, 4H), 1.83 (brs, 4H), 1.51–
1.22 (m, 36H), 0.92 (brs, 6H); ESI-MS m/z: 291.3090
[M]2+/2.
Results are expressed as the mean of n P 3 standard devia-
tion. One way analysis of variance (ANOVA) and Bonferroni post
hoc tests were used for statistical analyses (SPSS version 16.0). Sig-
nificant differences were considered at p < 0.05 level.
3. Results
3.1. Synthesis of amino acid/dipeptide-substituted gemini surfactants
3.1.1. General synthetic procedure
The coupling reactions between 3,30-iminobis (N,N-dimethyl-
propylamine) and protected amino acids were carried out in the
presence of EEDQ, and the intermediates were purified by separa-
tion on a silica column [16]. Subsequently, quaternization was car-
ried out as previously described [17] followed by de-protection of
the amino acids to give the target compounds.
Step d: Synthesis of N-2-amino-N,N-bis[3-(dimethylamino)-
propyl]-acetamide (4).
(i) De-protection of (2) followed the procedure outlined in
step c (i) to give a white solid. 1H NMR (CD3OD) d/ppm,
4.05 (brs, 2H), 3.49–3.33 (m, 4H), 3.20–3.07 (m, 4H),
2.87 (s, 6H), 2.85 (brs, 8H), 2.09–1.91 (m, 4H); ESI-MS
m/z: 245.2341 [M+H]+.
3.1.2. Synthesis of 12-7N(Gly)-12 (3) and 12-7N(Lys-Gly)-12 (6)
(Fig. 2)
Step a: Synthesis of N-2-(tert-butyloxycarbonyl) amino-N,N-
bis[3-(dimethylamino)propyl]-acetamide (2).
(ii) The white solid was dissolved in 10 mL of EtOH, 0.2 g of
KOH pellets was added, and the mixture stirred at 0 °C
for 1 h. Afterwards, the EtOH was removed under vac-
uum, and the residue was transferred to a separatory
funnel containing a mixture of CHCl3/H2O, 150 mL/
20 mL. The organic phase was collected, dried over
Na2SO4, and the solvent was removed under vacuum to
obtain 0.3 g of oil-like (4) (yield, 45%).
Three grams (17.1 mmol) of Boc-glycine and 6.3 g (25.7 mmol)
of N-ethoxycarbonyl-2-ethoxy-1,2-dihydroquinoline (EEDQ) were
dissolved in 30 mL of CHCl3 and stirred at room temperature for
10 min. Then, 3.85 g (20.6 mmol) of 3,30-iminobis (N,N-dimethyl-
propylamine) (1) dissolved in 10 mL of CHCl3 was added, gradually,
to the solution, and the reactants were stirred for an additional
10 min upon completion of the addition of amine. Finally, the solu-
tion was refluxed for 2 days under a N2 atmosphere.
Step e: Synthesis of N,N-bis[3-(dimethylamino)propyl]-[N0-
a
e
(N -N -di-Boc-L-lysyl)]-glycinamide (5).
Upon completion of the reaction, monitored by TLC, the solvent
was removed under vacuum, and the product was separated by
column chromatography (230–400 mesh silica gel) eluting with
CHCl3/MeOH (8:2, v/v) to obtain 4.75 g of colourless, oil-like prod-
uct (2) (yield 81%). 1H NMR (CDCl3) d/ppm: 5.55 (brs, 1H), 3.99–
3.96 (m, 2H), 3.40–3.22 (m, 4H), 2.35–2.16 (m, 16H), 1.76–1.67
(m, 4H), 1.43 (s, 9H); ESI-MS m/z: 345.2968 [M+].
The coupling of intermediate (4) and Boc-lys (Boc)-OH followed
the procedure outlined in step a to give 0.51 g of (5) (yield, 74%). 1H
NMR (CDCl3) d/ppm, 7.09 (s, 1H), 5.39 (s, 1H), 4.94 (s, 1H), 4.20–
3.89 (m, 3H), 3.36–3.12 (m, 4H), 2.97 (brs, 2H), 2.24–2.12 (m,
4H), 2.11–2.05 (brs, 12H), 1.80–1.21 (m, 28H); ESI-MS m/z:
573.4360 [M+H]+.

P. Yang et al. / European Journal of Pharmaceutics and Biopharmaceutics 75 (2010) 311–320
315
Step f: Synthesis of 1,9-bis(dodecyl)-1,1,9,9-tetramethyl-5-
a
e
[N0-(N -N -di-Boc-
dibromide.
L-lysyl)glycyl]amino-1,9-nonanediammonium
The synthesis and purification of this intermediate followed the
procedure outlined in step b (yield 88%).
1H NMR (CDCl3) d/ppm, 7.87 (s, 1H), 5.42 (s, 1H), 4.92 (s, 1H),
4.24–4.06 (m, 3H), 3.85–3.51 (m, 8H), 3.50–3.20 (m, 16H), 3.02
(brs, 2H), 2.44–2.09 (m, 4H), 1.85–1.57 (m, 6H), 1.41–1.13 (m,
58H), 0.83–0.81 (t, 6H, J = 5.0 Hz).
Step g: Synthesis of 12-7N(Lys-Gly)-12 (6).
The synthesis and purification of 12-7N(Lys-Gly)-12 (6) fol-
lowed the procedure outlined in step c (yield 30%). 1H NMR (CDCl3)
d
/ppm, 4.26 (brs, 2H), 4.07 (brs, 1H), 3.75–2.88 (m, 31H), 2.33–1.18
(m, 50H), 0.92 (brs, 6H); ESI-MS m/z: 355.3559 [M]2+/2.
3.1.3. Synthesis of 12-7N(Lys)-12 (8) and 12-7N(Lys-Lys)-12 (11)
(Fig. 3)
a
e
Step h: Synthesis of N,N-bis[3-(dimethylamino)propyl]-N -N -di-
Boc- -lysinamide (7).
L
The synthesis and purification of (7) followed the procedure
outlined in step a (yield 83%).1H NMR (CDCl3) d/ppm, 5.38–5.31
(m, 1H), 5.15 (brs, 1H), 4.35 (brs, 1H), 4.0 (brs, 1H), 3.36–3.05 (m,
2H), 2.99–2.79 (m, 3H), 2.08–2.01 (m, 4H), 1.98 (brs, 12H), 1.62–
1.29 (m, 7H), 1.23–1.17 (brs, 21H); ESI-MS m/z: 516.4095 [M+1]+.
a
Step i: Synthesis of 1,9-bis(dodecyl)-1,1,9,9-tetramethyl-5-(N -
e
N -di-Boc-
L-lysyl)amino-1,9-nonanediammonium dibromide.
The synthesis and purification of this intermediate followed the
procedure outlined in step b (yield 69%). 1H NMR (CDCl3) d/ppm,
5.33 (brs, 1H), 4.96 (brs, 1H), 4.42 (brs, 1H), 3.95–3.26 (m, 24H),
3.08 (brs, 2H), 2.49–2.17 (m, 4H), 1.90–1.18 (m, 64H), 0.87–0.85
(t, 6H, J = 5.0 Hz); ESI-MS m/z: 426.8980 [M]2+/2.
Fig. 4. Specific conductivities of the amino acid-modified gemini surfactants: (A)
12-7NH-12ꢁ2Cl h, 12-NK-12 s, 12-NG-12
D
, 12-7NKG-12
r
; (B) 12-NKK-12. Lines
Step j: Synthesis of 12-7N(Lys)-12 (8).
represent fit to Eq. (1). (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)
The synthesis and purification of 12-7N(Lys)-12 (8) (1,9-
bis(dodecyl)-1,1,9,9-tetramethyl-5-(L-lysyl)amino-1,9-nonanedi-
ammonium dichloride, dihydrochloride salt) was carried out
according to step c (yield 35%). 1H NMR (CDCl3) d/ppm, 4.53 (brs,
1H), 3.71–3.36 (m, 16H), 3.27–3.10 (m, 12H), 3.00 (brs, 2H),
2.32–1.51 (m, 14H), 1.51–1.25 (m, 36H), 0.92 (brs, 6H); ESI-MS
m/z: 326.8460 [M]2+/2.
vacuum, followed by transfer of the residue to a separatory funnel
containing a 2-phase system of CHCl3 and 1 wt.% Na2CO3 in water.
The organic phase was collected and washed, twice, with water
and dried over Na2SO4. The CHCl3 was removed under vacuum to
obtain 0.31 g product (isolated yield, 42%). 1H NMR (CDCl3)
d/ppm, 3.93 (brs, 1H), 3.63–3.43 (m, 2H), 3.42–3.23 (m, 3H), 3.06
(brs, 2H), 2.59 (brs, 2H), 2.48 (brs, 8H), 2.37 (s, 6H), 1.95–1.33
(m, 19H); ESI-MS m/z: 416.3598 [M+H]+.
e
Step k: Synthesis of N,N-bis[3-(dimethylamino)propyl]-N -Boc-
a
N -Cbz-
L
-lysinamide (9).
The synthesis and purification of this intermediate followed the
procedure outlined in step a (yield, 76%). 1H NMR (CDCl3) d/ppm,
7.41–7.30 (brs, 5H), 5.68 (brs, 1H), 5.08 (brs, 2H), 4.82 (brs, 1H),
4.65 (brs, 1H), 3.63–2.99 (m, 6H), 2.36–2.15 (m, 16H), 1.85–1.34
(m, 19H); ESI-MS m/z: 550.3954 [M+H]+.
a
Step m: Synthesis of N,N-bis[3-(dimethylamino)propyl]-N -
a0 e0
e
(N -N -di-Boc-L-lysyl)-N -Boc-L-lysinamide (10).
The synthesis and purification of (10) followed the procedure
outlined in step a (yield 83%).1H NMR (CDCl3) d/ppm, 6.89–6.78
(m, 1H), 5.22 (brs, 1H), 4.98–4.85 (m, 3H), 4.06 (brs, 1H), 3.60–
3.50 (m, 1H), 3.47–3.38 (m, 1H), 3.33–3.24 (m, 1H), 3.18–3.03
(m, 5H), 2.31–2.22 (m, 4H), 2.20 (brs, 12H), 1.84–1.40 (m, 39H),
1.38–1.29 (m, 4H); ESI-MS m/z: 744.5591 [M+H]+.
e
Step l: Synthesis of N,N-bis[3-(dimethylamino)propyl]-N -Boc-
L-lysinamide.
One gram (1.8 mmol) of (9) and 0.4 g (6.3 mmol) of HCOONH4
were dissolved in 10 mL of anhydrous MeOH and stirred for
10 min at room temperature under a N2 atmosphere. Then, 0.3 g
of 10% Pd/C was added, gradually, over 30 min, and the reaction
was continued for a further 24 h at room temperature. The catalyst
was removed by filtration, and the MeOH was removed under
Step n: Synthesis of 1,9-bis(dodecyl)-1,1,9,9-tetramethyl-5-
a
a0 e0
e
[N -(N -N -di-Boc-
L-lysyl)-N -Boc-L-lysyl)]amino-1,9-nonane-
diammonium dibromide.

316
P. Yang et al. / European Journal of Pharmaceutics and Biopharmaceutics 75 (2010) 311–320
Table 1
Critical micelle concentration fitting parameters.
Surfactant
A1
(l
S cmꢀ1 mMꢀ1
)
A2
(l
S cmꢀ1 mMꢀ1
)
d (mM)
cmc (mM)
aa
12-7NG-12
12-7NK-12
12-7NGK-12
12-7NKK-12
12-7NH-12ꢁ2Clꢀ
12-7NH-12ꢁ2Brꢀb
313
320
340
432
290 30
196
2
3
2
7
103.2 0.2
87.8 0.7
93.3 0.5
0.97 0.01
1.16 0.03
1.32 0.02
1.8 0.1
2.4 0.3
0.54 0.5
2.94 0.03
3.81 0.03
3.72 0.03
7.6 0.2
2.1 0.6
1.24 0.06
0.330 0.002
0.274 0.003
0.274 0.002
0.454 0.008
0.19 0.02
196
55
1
4
5
7
49
0.25 0.05
a
Equal to A2/A1.
Data from [8].
b
parameters, including the cmc and degree of micelle ionization
) for the amino acid-substituted gemini surfactants, are reported
in Table 1. Critical micelle concentrations were determined from a
fit of the specific conductivity data according to the method of
Carpena et al. [18], using the relation:
The synthesis and purification of this intermediate followed the
procedure outlined in step b (yield 85%). 1H NMR (CDCl3) d/ppm,
7.45 (brs, 1H), 5.63 (brs, 1H), 5.04 (brs, 1H), 4.86 (brs, 1H), 4.70
(brs, 1H), 4.01 (brs, 1H), 3.87–3.36 (m, 12H), 3.31 (brs, 12H), 3.08
(brs, 4H), 2.61–2.11 (m, 4H), 1.90–1.20 (m, 79H), 0.88–0.86 (t,
6H, J = 5.0 Hz); ESI-MS m/z: 540.9698 [M]2+/2.
(a
ꢀ
ꢁ
1 þ eðc-cmcÞ=d
j
¼
j0 þ A1c þ dðA2 ꢀ A1Þ ln
ð1Þ
1 ꢀ eꢀcmc=d
Step o: Synthesis of 12-7N(Lys-Lys)-12 (11).
where A1 and A2 are the pre-cmc and post-cmc slopes of the
j
vs.
The synthesis and purification of 12-7N(Lys-Lys)-12 (11) (1,9-
concentration curve and d is the width of the cmc transition region.
Initial estimates of A1 and A2 were obtained from the pre-cmc and
post-cmc slopes, respectively. Initial values of the cmc were ob-
tained from the approximate breakpoints in the specific conduc-
tance vs. concentration curves, and the initial estimate of d was
obtained from the approximate spacing between data points. The
a
bis(dodecyl)-1,1,9,9-tetramethyl-5-[N -(
L
-lysyl)-
L
-lysyl)]amino-
1,9-nonanediammonium dichloride, trihydrochloride salt) was
carried out according to the procedure followed in step c
(yield 38%). 1H NMR (CDCl3) d/ppm, 4.73 (brs, 1H), 4.07 (brs,
1H), 3.93 (brs, 1H), 3.81–3.31 (m, 18H), 3.22–3.11 (m, 12H),
3.07–2.97 (m, 4H), 2.44–1.26 (m, 56H), 0.93–0.91 (t, 6H,
J = 5.0 Hz); ESI-MS m/z: 390.8914 [M]2+/2.
initial specific conductance value,
ing at zero concentration.
j0, was obtained from the read-
3.2. Physiochemical characterization of amino acid/dipeptide-
substituted gemini compounds
3.2.2. Hydrodynamic size of self-aggregates
Estimates of the hydrodynamic diameters (Table 2) of the self-
aggregates of these gemini compounds (in the absence of DNA)
were obtained from DLS measurements made at two concentra-
tions above the cmc. The sizes of all the amino acid-substituted
gemini were found to decrease as the surfactant concentration
above the cmc was increased.
3.2.1. Critical micelle concentration
Specific conductance vs. concentration graphs for the amino
acid-substituted gemini surfactants are shown in Fig. 4. Fitting
Table 2
Aggregate size of the amino acid/dipeptide-substituted gemini surfactants above cmc.
3.3. Transfection of cells with amino acid/dipeptide-substituted gemini
surfactants
Surfactant
12-7NG-12
Concentration (mM)
Diameter (nm)
6
3.67
3.22
3.3.1. Transfection of COS 7 cells
10
The screening of the new derivatives was carried out using the
generally easy to transfect COS7 cell line. As controls, the 12-7NH-
12 parent gemini surfactant and Lipofectamine Plus were used. All
transfections were carried out in the presence of helper lipid,
DOPE. Gene expression increased significantly (p < 0.001) in all
PGL complexes and Lipofectamine Plus from 24 to 72 h (Fig. 5).
Some gene expression was observed after 1 week, but it was
12-7NK-12
12-7NGK-12
12-7NKK-12
6
10
5.68
3.94
6
10
5.85
3.54
12
20
4.21
3.73
p<0.001
24h
48h
72h
1w
p<0.001
6000
5000
4000
3000
2000
1000
0
12-7NG-12
12-7NK-12
12-7NGK-12 12-7NKK-12 12-7NH-12
Lipof
Plasmid
Fig. 5. Gene expression in COS7 cells transfected with PGL complexes. Triplicate wells of each sample were loaded on three different plates. The results are expressed as mean
of the plates (n = 3). Bars represent standard deviation.

P. Yang et al. / European Journal of Pharmaceutics and Biopharmaceutics 75 (2010) 311–320
317
PAM212
A
24h
48h
72h
5000
4500
4000
3500
3000
2500
2000
1500
1000
500
p<0.001
0
12-7NG-12 12-7NK-12 12-7NGK-12 12-7NKK-12 12-7NH-12 Lipofectamine
Sf 1 Ep
B
p<0.001
400
24h
48h
72h
350
300
250
200
150
100
50
0
12-7NG-12 12-7NK-12 12-7NGK-12 12-7NKK-12 12-7NH-12 Lipofectamine
Fig. 6. Gene expression in PAM212 (A) and Sf 1Ep (B) cells transfected with PGL complexes. Triplicate wells of each sample were loaded on three different plates. The results
are expressed as mean of the plates (n = 3). Bars represent standard deviation.
Table 3
significantly higher in the supernatants collected from cells trans-
fected with the 12-7NGK-12 PGL, 4.1 1.4 ng IFN-
c
/2 ꢂ 105 cells,
Size and zeta potential measurements. Values are average of four measure-
ments STD. Polydispersity (PDI) index is indicated for the size measurements as
average of four measurements STD in brackets.
compared to Lipofectamine Plus (p = 0.001). The other two amino
acid-substituted gemini surfactants, 12-7NK-12 and 12-7NKK-12,
performed similarly to the controls. Minimal gene expression
was observed in the cells transfected with the plasmid and DOPE,
in the absence of gemini surfactant.
PGL
Size (nm) (PDI)
Zeta potential (mV)
P/12-7NG-12/DOPE
P/12-7NK-12/DOPE
P/12-7NGK-12/DOPE
P/12-7NKK-12/DOPE
P/12-7NH-12/DOPE
P/DOPE
116.3 1.3 (0.252 0.008)
112.7 1.7 (0.265 0.006)
117.3 0.6 (0.232 0.012)
105.3 0.4 (0.245 0.008)
133 2.8 (0.279 0.006)
98.2 0.3 (0.224 0.008)
32.6 1.8
28.6 3.6
34.5 5.6
31.3 4.0
24.9 3.4
ꢀ32.7 0.6
3.3.2. Transfection of PAM212 and Sf 1Ep cells
The amino acid-substituted gemini surfactants are designed
for topical (cutaneous and mucosal) gene delivery. Two different
cell lines were transfected, the previously used PAM 212 kerati-
nocytes and the Sf 1Ep epithelial cells (Fig. 6). Similar to COS7
cells, gene expression in the PAM212 cell line increased signifi-
cantly in time (p < 0.001). However, there was little or no protein
detected after 1 week (results not shown). In PAM212 cells, the
highest gene expression was observed with the 12-7NK-12 at
drastically diminished compared to the previous time point. At
72 h, the highest level of gene expression was measured in the cells
transfected with the 12-7NG-12 PGL, 4.8 0.6 ng IFN-
cells, significantly higher compared to the 12-7NH-12 PGL,
1.9 0.4 ng IFN-
c
/2 ꢂ 105
c
c
/2 ꢂ 105 cells, (p < 0.001) and Lipofectamine Plus,
2.7 1.2 ng IFN-
/2 ꢂ 105 cells, (p < 0.001). IFN-
c
levels were also
72 h, 631 218 pg IFN-
c
/2 ꢂ 105 cells. There was no significant

318
P. Yang et al. / European Journal of Pharmaceutics and Biopharmaceutics 75 (2010) 311–320
B
A
Plasmid
Plasmid
12-7NG-12
12-7NK-12
12-7NGK-12
12-7NKK-12
12-7NH-12
12-7NG-12
12-7NK-12
12-7NGK-12
12-7NKK-12
12-7NH-12
240
260
280
300
320
240
260
280
300
320
Wavelength [nm]
Wavelength [nm]
Fig. 7. CD spectra of the PG complexes (A) and the PGL particles (B). Values are average of three measurements.
difference among the substituted derivatives (12-7NG-12, 12-
7NK-12, 12-7NGK-12 and 12-7NKK-12), but all exhibited signifi-
cantly higher gene expression compared to the parent compound,
The observed break points are somewhat broader than that ob-
served for the 12-7NH-12 2Brꢀ surfactant but more in line with
that shown for the gemini 12-7NH-12 2Clꢀ (Fig. 4). Note that these
amino acid/peptide gemini surfactants have, in addition to the tra-
ditional two quaternary ammonium charged sites with Clꢀ counte-
rions, 1–3 primary amine sites that are in the form of HCl salts. The
increase in the magnitude of the specific conductance for these
compounds correlates well with the increase in number of the lat-
ter sites in the gemini surfactant. These spacer substituents are
hydrophilic and make the spacer more hydrated than in the case
of the 12-7NH-12 parent cation, and this is consistent with the
order of the cmc values (Table 1); 12-7NKK-12 > 12-7NGK-
12 > 12-7NK-12 > 12-7NG-12 > 12-7NH-12 and the broad break
in the specific conductance lines of Fig. 4. Alternatively, the in-
creased hydrophilicity of the head group will also likely result in
a change in the micelle dynamics and may also be partially respon-
sible for this observation.
The shape of the conductance profile, specifically the lack of a
sharp break for 12-7NH-12, also suggests that there may be a dif-
ference in the dynamics of self-aggregation associated with the
chloride when compared to the bromide counterion. Such an
observation is not inconsistent with previous results for gemini
surfactants where different counterions were found to produce dif-
ferent aggregation behaviour [19]; however, this feature requires
additional study.
12-7NH-12 of 168 18 pg IFN-
c
/2 ꢂ 105 cells (p < 0.001). As ob-
served in earlier studies [5], the in vitro transfection efficiency
of Lipofectamine in PAM 212 cells is higher compared to the gem-
ini surfactants. Overall, gene expression in PAM212 was 1–6-fold
higher compared to Sf 1Ep cells. In the Sf 1Ep cell line, the trend
was similar to the COS7 cells, the 12-7NG-12 and 12-7NGK-12
PGLs showing higher gene expression at 72 h compared to the
12-7NK-12 and 12-7NKK-12 PLGs at 72 h. IFN-
12-7NG-12 PGL-transfected cells (329 46 pg IFN-
was significantly higher compared to all other delivery systems
(p < 0.001). The transfection efficiency of the Lipofectamine was
very low in the Sf 1Ep cells.
c
c
levels in the
/2 ꢂ 105 cells)
3.4. Physicochemical characterization of the amino acid/dipeptide-
substituted gemini surfactant based transfection complexes
3.4.1. Size and zeta potential measurements
PGL complexes were prepared in a similar manner to those used
for the transfection experiments. Particle sizes for all complexes
were in the 110–140 nm range (Table 3). The plasmid/DOPE com-
plexes showed slightly smaller particle size, slightly less than
100 nm. The zeta potential of ꢀ32.7 0.6 mV of the plasmid/DOPE
complexes shifted to +20–40 mV.
The aggregate sizes (Table 2) determined from DLS measure-
ments show that they increase with the size of the amino acid/
dipeptide substitution on the imino spacer of the gemini and their
magnitude is approximately consistent with the expected overall
length of a fully extended C12 alkyl tail, quaternary head group
and extended amino acid. A decrease in the hydrodynamic diame-
ters of the self-aggregates of these gemini compounds was found
to occur as the post-micelle concentration was increased. Esti-
3.4.2. Circular dichroism
Circular dichroism was used to evaluate compaction of the DNA
by the substituted gemini surfactants. Plasmid–gemini surfactant
without DOPE and PGL complexes were prepared in a similar man-
ner to those for the transfection experiments. The plasmid DNA
alone showed two positive peaks at 295 and 255 nm, while all
PG complexes had a blue shift of the 295 nm peak and depression
of the 255-nm peak (Fig. 7A). The 12-7NG-12 PGL shows a flatten-
ing of the 295-nm peak, and the other compounds, including the
parent 12-7NH-12 surfactant, show a slight depression (Fig. 7B).
The pattern of the plasmid–DOPE mixture showed a positive peak
at 275 nm.
mates of micelle counterion dissociation
a (Table 1) measured near
the cmc show that Clꢀ is less tightly bound in the gemini surfactant
with amino acid-substituted spacer than in the case of the parent
compound, and this can lead to a higher surface charge density
in the former aggregates. As the surfactant concentration is in-
creased, growth of the aggregates must occur in a manner such
as to minimize the repulsive forces in the head group region of
the aggregates. This may be achieved by appropriate changes in
the shape of the aggregates to overcome this stress. A change to
a bilayer-like structure would be enhanced by favourable intermo-
lecular hydrogen bonding between the keto-oxygen and the pri-
mary amine hydrogens of adjacent amino acids. This could lead
to compaction of the aggregates as the amino acids become less ex-
tended from the aggregate surface. Thus, one might expect a trans-
formation from a polydisperse to a less disperse distribution of
4. Discussion
4.1. Characterization of the amino acid/dipeptide-substituted gemini
surfactants
The specific conductivity plots are characteristic of gemini sur-
factants in general, with no evidence of pre-micelle aggregation.

P. Yang et al. / European Journal of Pharmaceutics and Biopharmaceutics 75 (2010) 311–320
319
aggregates. High polydispersity index (PDI) values (ca. 1) were
found for all amino acid/dipeptide spacer-substituted gemini at
the lower concentration, i.e., nearer to the cmc. However, at the
higher concentration, the PDI values decreased to 0.8 < PDI < 1.0
in several cases. The rather broad break points in the specific con-
ductance graphs (Fig. 4) are also qualitatively consistent with the
higher polydispersity of the aggregates as they form at the cmc;
the d values from Eq. (1) (Table 1) indicate the width of the transi-
tion between the pre-micellar and post-micellar slopes and are
indicative of this broadness.
the particles is approximately +30 mV for all compounds, which
indicates that the system is colloidally stable [27]. There was no
difference between the zeta potential of the unsubstituted 12-
7NH-12 and amino acid/dipeptide-substituted gemini surfactants,
indicating that substitution did not shield the surface charge on
the particles. The CD spectra obtained are consistent with those
previously reported for gemini surfactants [8,13,28]. Previously,
such changes in the native DNA structure were interpreted in
terms of the surfactants inducing formation of Wꢀ DNA [29],
which usually required high salt concentrations. The changes ob-
served for the gemini surfactants are more consistent with a B-
DNA conformation in which both the helicity and base stacking
are perturbed as a result of compaction of the DNA molecule by
the surfactant [30]. The similarity between the profiles obtained
for the amino acid-substituted vs. unsubstituted gemini surfac-
tants (Fig. 7) indicates that the substitution on the spaced does
not interfere with the ability of the gemini surfactant to compact
DNA.
4.2. In vitro transfection efficiency of the amino acid/dipeptide-
substituted gemini compounds
We have studied the in vitro transfection efficiency of novel
amino acid-substituted gemini surfactants in three different epi-
thelial cell lines. The epithelial cell lines were used primarily to
evaluate the performance of these compounds with potential
application in scleroderma, atopic dermatitis and wound healing.
Gemini surfactants self-assemble into higher order structures, and
their aggregation depends on the length and chemical nature of
the spacer [20] that also determines the fit between the positively
charged nitrogens of the head group and the phosphate backbone
of the DNA [21]. Previously, the transfection efficiency of the alkyl
spacer gemini surfactants was found to vary inversely with their
head group area determined from Gibbs interfacial energy [5].
Imino substitution in the spacer introduced a better fit with the
DNA and conferred pH sensitivity to the gemini surfactants [8],
which could reduce cytotoxicity and facilitate endosomal escape
of the DNA. Polylysine and polyallylamine showed relatively high
cytotoxicity due to their polycationic nature and the presence of
primary amine groups. Substitution of the primary amines with
imidazoyl derivatives not only reduced cytotoxicity, but also im-
proved endosomal escape of the genetic material [22]. The amino
acid-substituted gemini surfactants show higher gene expression
compared to the unsubstituted parent compound 12-7NH-12 in
all three cell lines (Figs. 5 and 6) investigated in this work. Cat-
ionic agents non-specifically bind the cell surface through ionic
interactions with membrane-associated proteoglycans [23]. Dif-
ferences in the gene expression among the three cell lines might
be due to the differences in the distribution of proteoglycans on
the cell surface. Transgene expression depends not only on com-
paction by the cationic agent and delivery of the genetic material
into the cells but also on endosomal escape and release of the
DNA from the complexes. It appears that grafting of amino acids
or peptides on the gemini surfactant molecule can improve the
delivery and release of the DNA in the cytoplasm. It should be
noted that a weak binding in the PG complex does not protect
and compact sufficiently, whereas strong binding will not allow
for the release of the DNA. While gemini surfactants bind the
DNA via electrostatic interactions, the nature of binding by the
amino acid and peptide-substituted gemini surfactants might be
‘softened’ by van der Waals and hydrogen bonding forces [24],
thus increasing the flexibility and plasticity of the supramolecular
structures [25]. Increased hydrophilicity of the amino acid-substi-
tuted gemini surfactants compared to the parent compound
might translate into better biocompatibility. The increased bio-
compatibility of these peptidomimetic particles could lead to
lower cytotoxicity. The gene expression pattern showed similari-
ties with higher gene expression in the cells transfected with gly-
cine-substituted and glycyl-lysine substituted gemini surfactants.
Although it was expected that the substituted gemini surfactants
would form larger complexes due to a larger hydrophilic head,
there was no significant difference compared to the parent com-
pound (Table 3). Complex sizes of 100–150 nm are optimal for
endocytotic internalization by the cells [26]. Zeta potential of
5. Conclusion
Amino acid and peptide-substituted imino spacer of gemini sur-
factants transfect epithelial cells at higher efficiency compared to
unsubstituted imino spacers. The substitution does not increase
the size of the particles or reduce zeta potential, nor weaken
DNA compaction. The higher transfection efficiency is attributed
to the higher biocompatibility and flexibility of the amino acid/
peptide-substituted gemini surfactants and demonstrates the fea-
sibility of using amino acid-substituted gemini surfactants as gene
carriers for the treatment of diseases affecting epithelial tissue.
Acknowledgements
The authors thank the Saskatchewan Structural Sciences Centre
(SSSC) for the use of NMR, ESI-MS and CD instruments and Mr.
Jason Maley for analysis and discussions. Financial assistance
provided to R.E.V. by the Natural Science and Engineering Research
Council of Canada (NSERC) is gratefully acknowledged.
References
[1] B. Freundlich, S.A. Jimenez, V.D. Steen, T.A. Medsger Jr., M. Szkolnicki, H.S. Jaffe,
Treatment of systemic sclerosis with recombinant interferon-gamma. A phase
i/ii clinical trial, Arthritis Rheum. 35 (1992) 1134–1142.
[2] L.M. Buys, Treatment options for atopic dermatitis, Am. Fam. Physician 75
(2007) 523–528.
[3] H. Emmez, O. Kardes, F. Dogulu, G. Kurt, L. Memis, M.K. Baykaner, Role of
antifibrotic cytokine interferon-gamma in the prevention of postlaminectomy
peridural fibrosis in rats, Neurosurgery 62 (2008) 1351–1357 (discussion 7–8).
[4] F.M. Menger, J.S. Keiper, Gemini surfactants, Angew. Chem. Int. Ed. Engl. 39
(2000) 1906–1920.
[5] I. Badea, R. Verrall, M. Baca-Estrada, et al., In vivo cutaneous interferon-gamma
gene delivery using novel dicationic (gemini) surfactant-plasmid complexes, J.
Gene Med. 7 (2005) 1200–1214.
[6] M.L. Fielden, C. Perrin, A. Kremer, et al., Sugar-based tertiary amino gemini
surfactants with a vesicle-to-micelle transition in the endosomal ph range
mediate efficient transfection in vitro, Eur. J. Biochem. 268 (2001) 1269–1279.
[7] H.S. Rosenzweig, V.A. Rakhmanova, R.C. MacDonald, Diquaternary ammonium
compounds as transfection agents, Bioconjugate Chem. 12 (2001) 258–263.
[8] S.D. Wettig, I. Badea, M. Donkuru, R.E. Verrall, M. Foldvari, Structural and
transfection properties of amine-substituted gemini surfactant-based
nanoparticles, J. Gene Med. 9 (2007) 649–658.
[9] E. Fisicaro, C. Compari, E. Duce, G. Donofrio, B. Rozycka-Roszak, E. Wozniak,
Biologically active bisquaternary ammonium chlorides: physico-chemical
properties of long chain amphiphiles and their evaluation as non-viral
vectors for gene delivery, Biochim. Biophys. Acta 1722 (2005) 224–233.
[10] R. Zana, Dimeric (gemini) surfactants: effect of the spacer group on the
association behavior in aqueous solution, J. Colloid Interf. Sci. 248 (2002) 203–
220.
[11] L. Karlsson, M. van Eijk, O. Soderman, Compaction of DNA by gemini
surfactants: effects of surfactant architecture, J. Colloid Interf. Sci. 252
(2002) 290–296.

320
P. Yang et al. / European Journal of Pharmaceutics and Biopharmaceutics 75 (2010) 311–320
[12] C. Wang, X. Li, S.D. Wettig, I. Badea, M. Foldvari, R.E. Verrall, Investigation of
complexes formed by interaction of cationic gemini surfactants with
deoxyribonucleic acid, Phys. Chem. Chem. Phys. 9 (2007) 1616–1628.
[13] M. Foldvari, I. Badea, S. Wettig, R. Verrall, M. Bagonluri, Structural
characterization of novel gemini non-viral DNA delivery systems for
cutaneous gene therapy, J. Exp. Nanosci. 1 (2006) 165–176.
[14] R. Kharidia, K.A. Friedman, J.F. Liang, Improved gene expression using low
molecular weight peptides produced from protamine sulfate, Biochemistry
(Mosc.) 73 (2008) 1162–1168.
[21] X. Chen, J. Wang, N. Shen, et al., Gemini surfactant/DNA complex monolayers
at the air–water interface: effect of surfactant structure on the assembly,
stability, and topography of monolayers, Langmuir 18 (2002) 6222–6228.
[22] A. Pathak, A. Aggarwal, R.K. Kurupati, et al., Engineered polyallylamine
nanoparticles for efficientinvitro transfection, Pharm. Res. 24 (2007) 1427–1440.
[23] K.A. Mislick, J.D. Baldeschwieler, Evidence for the role of proteoglycans in
cation-mediated gene transfer, Proc. Natl. Acad. Sci. USA 93 (1996) 12349–
12354.
[24] E. Jabbari, Targeted delivery with peptidomimetic conjugated self-assembled
nanoparticles, Pharm. Res. 26 (2009) 612–630.
[15] S.D. Wettig, R.E. Verrall, Thermodynamic studies of aqueous m–s–m gemini
surfactant systems, J. Colloid Interf. Sci. 235 (2001) 310–316.
[25] G. Colombo, P. Soto, E. Gazit, Peptide self-assembly at the nanoscale:
a
[16] T.Y. Zakharian, A. Seryshev, B. Sitharaman, B.E. Gilbert, V. Knight, L.J. Wilson, A
fullerene-paclitaxel chemotherapeutic: synthesis, characterization, and study
of biological activity in tissue culture, J. Am. Chem. Soc. 127 (2005) 12508–
12509.
[17] X. Li, S.D. Wettig, C. Wang, M. Foldvari, R.E. Verrall, Synthesis and solution
properties of gemini surfactants containing oleyl chains, Phys. Chem. Chem.
Phys. 7 (2005) 3172–3178.
challenging target for computational and experimental biotechnology, Trends
Biotechnol. 25 (2007) 211–218.
[26] M.R. Almofti, H. Harashima, Y. Shinohara, A. Almofti, Y. Baba, H. Kiwada,
Cationic liposome-mediated gene delivery: biophysical study and mechanism
of internalization, Arch. Biochem. Biophys. 410 (2003) 246–253.
[27] W. Mehnert, K. Mader, Solid lipid nanoparticles: production, characterization
and applications, Adv. Drug Deliv. Rev. 47 (2001) 165–196.
[18] P. Carpena, J. Aguiar, P. Bernaola-Galvan, C.C. Ruiz, Problems associated with
the treatment of conductivity–concentration data in surfactant solutions:
simulations and experiments, Langmuir 18 (2002) 6054–6058.
[28] I. Badea, S. Wettig, R. Verrall, M. Foldvari, Topical non-invasive gene delivery
using gemini nanoparticles in interferon-gamma-deficient mice, Eur. J. Pharm.
Biopharm. 65 (2007) 414–422.
[19] N. Jiang, P.X. Li, Y.L. Wang, J.B. Wang, H.K. Yan, R.K. Thomas, Aggregation
behavior of hexadecyltrimethylammonium surfactants with various
counterions in aqueous solution, J. Colloid Interf. Sci. 286 (2005) 755–760.
[20] E. Alami, G. Beinert, P. Marie, R. Zana, Alkanediyl-alpha, omega-
bis(dimethylalkylammonium bromide) surfactants. 3. Behavior at the air–
water-interface, Langmuir 9 (1993) 1465–1467.
[29] D. Simberg, D. Danino, Y. Talmon, et al., Phase behavior, DNA ordering, and size
instability of cationic lipoplexes. Relevance to optimal transfection activity, J.
Biol. Chem. 276 (2001) 47453–47459.
[30] C.S. Braun, G.S. Jas, S. Choosakoonkriang, G.S. Koe, J.G. Smith, C.R. Middaugh,
The structure of DNA within cationic lipid/DNA complexes, Biophys. J. 84
(2003) 1114–1123.