Efficacious redox-responsive gene delivery in serum by ferrocenylated monomeric and dimeric cationic cholesterols

Article abstract of DOI:10.1039/c4ob02513j

Herein, we present the design and synthesis of new redox-active monomeric and dimeric (gemini) cationic lipids based on ferrocenylated cholesterol derivatives for gene delivery. The cationic cholesterols are shown to be transfection efficient after being formulated with the neutral helper lipid DOPE in the presence of serum (FBS). The redox activity of the resulting co-liposomes and their lipoplexes could be regulated using the alkanyl ferrocene moiety attached to the ammonium head groups of the cationic cholesterols. Atomic force microscopy (AFM), dynamic light scattering (DLS) and zeta potential measurements were performed to characterize the co-liposomal aggregates and their complexes with pDNA. The transfection efficiency of lipoplexes could be tuned by changing the oxidation state of the ferrocene moiety. The gene transfection capability was assayed in terms of green fluorescence protein (GFP) expression using pEGFP-C3 plasmid DNA in three cell lines of different origins, namely Caco-2, HEK293T and HeLa, in the presence of serum. The vesicles possessing ferrocene in the reduced state induced an efficient transfection, even better than a commercial reagent Lipofectamine 2000 (Lipo 2000) as evidenced by flow cytometry and fluorescence microscopy. All the co-liposomes containing the oxidized ferrocene displayed diminished levels of gene expression. Gene transfection events from the oxidized co-liposomes were further potentiated by introducing ascorbic acid (AA) as a reducing agent during lipoplex incubation with cells, leading to the resumption of transfection activity. Assessment of transfection capability of both reduced and oxidized co-liposomes was also undertaken following cellular internalization of labelled pDNA using confocal microscopy and flow cytometry. Overall, we demonstrate here controlled gene transfection activities using redox-driven, transfection efficient cationic monomeric and dimeric cholesterol lipids. Such systems could be used in gene delivery applications where transfection needs to be performed spatially or temporally. This journal is

Full text of DOI:10.1039/c4ob02513j

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Efficacious redox-responsive gene delivery in
serum by ferrocenylated monomeric and dimeric
cationic cholesterols†
Cite this: Org. Biomol. Chem., 2015,
3, 4310
1
a
a
b
Gururaja Vulugundam,‡ Krishan Kumar,‡ Paturu Kondaiah and
Santanu Bhattacharya*
a
Herein, we present the design and synthesis of new redox-active monomeric and dimeric (gemini) cat-
ionic lipids based on ferrocenylated cholesterol derivatives for gene delivery. The cationic cholesterols are
shown to be transfection efficient after being formulated with the neutral helper lipid DOPE in the pres-
ence of serum (FBS). The redox activity of the resulting co-liposomes and their lipoplexes could be regu-
lated using the alkanyl ferrocene moiety attached to the ammonium head groups of the cationic
cholesterols. Atomic force microscopy (AFM), dynamic light scattering (DLS) and zeta potential measure-
ments were performed to characterize the co-liposomal aggregates and their complexes with pDNA. The
transfection efficiency of lipoplexes could be tuned by changing the oxidation state of the ferrocene
moiety. The gene transfection capability was assayed in terms of green fluorescence protein (GFP)
expression using pEGFP-C3 plasmid DNA in three cell lines of different origins, namely Caco-2, HEK293T
and HeLa, in the presence of serum. The vesicles possessing ferrocene in the reduced state induced an
efficient transfection, even better than a commercial reagent Lipofectamine 2000 (Lipo 2000) as evi-
denced by flow cytometry and fluorescence microscopy. All the co-liposomes containing the oxidized
ferrocene displayed diminished levels of gene expression. Gene transfection events from the oxidized co-
liposomes were further potentiated by introducing ascorbic acid (AA) as a reducing agent during lipoplex
incubation with cells, leading to the resumption of transfection activity. Assessment of transfection capa-
bility of both reduced and oxidized co-liposomes was also undertaken following cellular internalization of
labelled pDNA using confocal microscopy and flow cytometry. Overall, we demonstrate here controlled
gene transfection activities using redox-driven, transfection efficient cationic monomeric and dimeric
cholesterol lipids. Such systems could be used in gene delivery applications where transfection needs to
be performed spatially or temporally.
Received 30th November 2014,
Accepted 19th February 2015
DOI: 10.1039/c4ob02513j
www.rsc.org/obc
reversible physical or chemical changes in response to external
Introduction
4
–6
7
5
or internal stimuli like pH, heat, light, and the presence of
8
9,10
Recently, there has been a surge in research activity on the reducing agents and enzymes.
In the case of biological
development of gene and drug delivery systems that are also cells, membrane fusion and disassembly in response to an
1
–3
responsive to external stimuli.
These systems undergo imposed membrane potential is a common phenomenon.
Thus, stimuli-responsive systems find applications in the con-
trolled and selective release of the enveloped cargo (DNA/drug)
at the target site whenever it is required.
Ferrocene (Fc) is one of the most commonly used redox-
switchable moieties because of its chemical stability over a
wide range of solution conditions and favorable working redox
a
Department of Organic Chemistry, Indian Institute of Science, Bangalore, 560012,
India. E-mail: sb@orgchem.iisc.ernet.in; Fax: +91-80-23600529;
Tel: +91-80-22932664
b
Department of Molecular Reproduction Development and Genetics, Indian Institute
of Science, Bangalore, 560012, India. E-mail: paturu@mrdg.iisc.ernet.in
†
1
1,12
Electronic supplementary information (ESI) available: Synthetic procedures potentials.
The transformation of a hydrophobic ferrocene
Fc(II)] moiety to a hydrophilic ferrocenium [Fc(III)] moiety has
been utilized by many groups to control the aggregation pro-
and characterization details, DLS and zeta measurements of co-liposomes in
HEPES buffer, UV-Vis spectra, XRD spectra of co-liposomes, schematic illus-
tration of lipid packing, optimization of DOPE content in liposomal formu-
lations and transfection efficacies in 50% serum, and FACS analysis of labelled
pDNA internalization. See DOI: 10.1039/c4ob02513j
[
1
3
perties of a number of synthetic molecular systems. Gokel
and coworkers reported for the first time vesicle formation
from a steroidal derivative of ferrocene. These acted as pro-
‡
These authors contributed equally to this work.
4310 | Org. Biomol. Chem., 2015, 13, 4310–4320
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
amphiphiles, which formed vesicles upon oxidation to the the spacer joining the polar head groups in the case of gemini
1
4,15
ferrocenium moiety as a polar head group.
The vesicles lipid series. The lipoplexes formed from these redox-active
collapsed when Fc(III) was reduced to Fc(II). Abe’s group has lipids formulated with a helper lipid, 1,2-dioleoyl-sn-glycero-3-
reported spontaneous vesicle formation by a mixture of a cat- phosphatidyl ethanolamine (DOPE), showed an excellent trans-
ionic ferrocenyl surfactant (FTMA) and an anionic surfactant fection ability, significantly better than a commercial formu-
(
SDBS) which were reversibly converted into micelles upon oxi- lation, Lipofectamine 2000 (Lipo 2000) in the presence of serum.
1
6
dation of FTMA and reformed into vesicles on reduction.
Importantly, the transfection activity could be controlled via a
These authors also demonstrated vesicle formation by a ferro- change in the redox state of the ferrocene moiety. The oxidation
cene terminated dialkyl chain surfactant, bis-(11-ferrocenyl- of ferrocene in liposomal formulations led to a decrease in the
(undecyl)) dimethylammonium bromide (BFDMA), which dis- transfection capability of the resulting lipoplexes. The transfec-
integrated into smaller aggregates and micelles after oxi- tion activity could be resumed after exposing the oxidized lipo-
1
7,18
dation.
Abbott et al. studied the interaction of BFDMA plexes to ascorbic acid (AA) in the cell culture medium. All the
with DNA and successfully demonstrated spatial and temporal co-liposomes (reduced and oxidized) and their transfection opti-
1
9–27
control over lipoplex mediated DNA delivery.
These mized lipoplexes were characterized using zeta, DLS and AFM
authors showed that the oxidation state of Fc which was measurements. Such transfection efficient systems with distinct
present in the hydrophobic domain played a key role in med- activity under the influence of the redox phenomenon could be
1
9,20
iating high or low levels of gene transfection.
shown that these transfection ‘inactive’ lipoplexes can be in a spatially or temporally controlled manner.
turned on” by chemically activating them both before and
It was also very useful where cellular gene expression needs to be achieved
“
2
2
after the introduction of lipoplexes into the cells. Apart from
this, a recent study with covalently attached ferrocene to the Experimental details
phospholipid head group permitted an electrochemical
Materials
2
8
modulation of the liposomal membrane permeability. Also,
recently ferrocene functionalized branched PEIs have been
All reagents and solvents used in the present study were of the
highest grade commercially available and all solvents used
were purified, dried, or freshly distilled, as required. Starting
materials, such as ferrocene, 6-bromohexanoic acid, 11-bromo-
undecanoic acid, anhydrous aluminium chloride and sodium
borohydride, were purchased from Sigma-Aldrich. 1,2-Dioleoyl-
sn-glycero-3-phosphatidyl ethanolamine (DOPE) was pur-
chased from Avanti Polar Lipids. Column chromatography was
performed using a 60–120 mesh silica gel. NMR spectra were
2
9
developed which show good transfection activity.
Cholesterol being an important constituent of animal
cell membranes, it is implicated in various biological pro-
30–32
cesses.
As part of our continuing efforts to develop efficient
cytofectins with improved gene delivery efficacies, we earlier
reported several cholesterol based monomeric or gemini cat-
ionic lipids and lipopolymers where we studied their aggregate
33–41
formation and DNA delivery efficiency.
However, to date, no
1
recorded using a Bruker spectrometer (400 MHz for H NMR
ferrocene based cholesterol derivative has been developed that
can also induce gene transfection. Herein, we report the gene
delivery efficacies of ferrocene containing cationic monomeric
and dimeric lipids based on the cholesterol backbone (Fig. 1).
The redox-active ferrocene moiety is located in the polar head
group region in the monomeric lipid series, while it is a part of
1
3
and 100 MHz for C NMR). The chemical shifts (δ) are
reported in ppm downfield from the internal standard, TMS,
1
13
for H-NMR and C-NMR. Mass spectra were recorded on a
MicroMass ESI-TOF spectrometer. All the compounds were
1
13
adjudged pure by H NMR, C NMR, mass spectroscopy
ESI-MS) and elemental analysis.
(
Synthesis
All the intermediates (1–8) and the final lipids (CHM-C6F,
CHM-C11F, CHD-C6F and CHD-C11F) were prepared as
described in the ESI.†
Preparation of reduced and oxidized co-liposomes
Mixed liposomes of each cationic lipid with DOPE in the
desired molar ratio were prepared by dissolving in chloroform
in a wheaton glass vial. The organic solvent was evaporated by
passing N gas to form a thin film along the walls of the tube
2
which was dried completely under vacuum for 6 h. Each lipid
was then hydrated in the presence of 1 mM aqueous Li SO
2 4
solution (pH 6) which was filtered through 0.1 μm cut-off
Millex titration cartridges to remove any extraneous particles.
The electrolyte was required for the electrolytic oxidation of
the ferrocenylated liposomes. The final concentration of each
Fig. 1 Molecular structures of redox-active ferrocene containing cat-
ionic monomeric (CHM-C6F and CHM-C11F) and gemini (CHD-C6F and
CHD-C11F) cholesteryl lipids.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 4310–4320 | 4311

View Article Online
Paper
Organic & Biomolecular Chemistry
cationic lipid was maintained at 1 mM which was left for modified Eagle’s medium) containing 10% fetal bovine serum
hydration for 12 h at 4 °C and repeatedly freeze–thawed (ice- (FBS). The cell culture medium also contained pen strep
cold water to 70 °C) with intermittent vortexing to ensure (100 units ml− penicillin and 100 µg ml streptomycin) as an
1
−1
optimal hydration. Sonication of these suspensions for 15 min antibiotic. Cells were maintained in an incubator (37 °C) at a
in a bath sonicator at 70 °C afforded liposomes which were relative humidity of >95% with a CO level of 5%. All the cell
2
used for all the studies. Oxidized co-liposomes were prepared lines were passaged using 0.5% trypsin–EDTA at proper con-
by electrochemical oxidation of the 1 mM suspension of fluency levels.
2
9
reduced cationic lipid/DOPE as described earlier.
The
Transfection biology (EGFP expression)
cationic lipid suspensions were heated up to 70 °C and electro-
chemical oxidation was accomplished using a CHI 660A We used pEGFP-C3 plasmid DNA to check the gene trans-
electrochemical analyzer (CH Instruments, TX, USA) with a fection levels mediated by redox responsive cationic monomeric
conventional three electrode assembly. A platinum mesh and gemini lipids. In a typical gene transfection experiment,
2
(
4.0 cm ) was used as the working and auxiliary electrodes and first 24-well cell culture plates were seeded at a density of
a saturated calomel electrode was used as the reference elec- 40 000 cells per well and allowed to grow for 24 h. Lipoplexes
trode. A constant potential of 600 mV was applied between the were prepared using 0.8 µg of pEGFP-C3 plasmid DNA and
working electrode and the reference electrode along with con- different co-liposomal formulations at various molar ratios
tinuous purging of N gas. The progress of oxidation was fol- (cationic lipid/pDNA). Cells were treated with lipoplexes for 6 h
2
lowed by monitoring current passed at the working electrode in 10% serum (FBS) containing cell culture medium. The oxi-
and by UV/visible spectrophotometry. The UV-Vis spectra of dized lipoplexes were incubated in cell culture medium with
the redox-active lipid suspensions before and after oxidation and without ascorbic acid (10 molar equivalents of cationic
are presented in Fig. S1, ESI.† The absorption peak corres- lipid). After lipoplex treatment, cells were incubated for
ponding to ferrocene [Fc(II)] (430 nm) got suppressed after another 42 h. At the end, cells were washed twice properly with
oxidation and a new peak at 645 nm characteristic of the ferroce- DPBS and detached by trypsinization using 0.5% trypsin–
nium moiety [Fc(III)] appeared. The oxidized co-liposomes EDTA. Cells were pooled in 5% FBS containing DPBS and
appeared to be quite stable where the characteristic peak of fer- taken for analysis using a FACS Calibur flow cytometer
rocenium was evident even after 96 h of oxidation (Fig. S1, ESI†). (Becton-Dickinson). Flow cytometry data analysis was per-
formed using the WinMDI 2.9 program. Gene transfection
Dynamic light scattering (DLS) and zeta potential
measurements
capabilities of different liposomal formulations were expressed
in terms of % GFP-positive cells and geometric mean
The average particle sizes and the zeta potential of the co-lipo- fluorescence intensity (MFI). The visual analysis of GFP was
somes were determined using a Zetasizer nano (Malvern, undertaken with the help of fluorescence microscopy
Worcestershire, UK) instrument which employed an incident (Olympus IX-81).
laser beam of 633 nm wavelength. The mean diameters
reported were obtained from a Gaussian analysis of the inten-
In vitro Cytotoxicity
sity-weighted particle size distributions. The lipid concen- An MTT [3-(4,5-dimethylthiazole-2-yl)-2,5-diphenyltetrazolium
trations for size and zeta potential measurements were bromide] assay was used to evaluate the viabilities of cells
maintained at 0.1 mM. Lipoplexes of the reduced or oxidized treated with different lipoplexes under gene transfection experi-
cationic lipid formulations at various lipid/pDNA molar ratios ment conditions. In a typical experiment, cells were treated
were prepared by mixing pDNA and a given amount of the cat- with lipoplexes after 24 h post plating of cells. Lipoplexes were
ionic lipid formulation such that the final concentration of removed after 6 h and fresh 10% serum containing medium
pDNA is 4 μg mL− . The DLS and zeta potential measurements (200 µl) was added to wells. After 38 h, MTT (20 µL, 5 mg ml
1
−1
were performed after an incubation of 30 min.
in DPBS) was added to each well for 4 h. The assay was termi-
nated by removing the entire medium from wells, followed by
addition of DMSO (200 µl) and reading the absorbance at
Atomic force microscopy
AFM images were recorded on an atomic force microscope 570 nm on a microplate reader. Lipoplex treatment in each
AFM; JPK NanoWizard) by tapping mode in air. The aqueous experiment was performed in triplicates and the results
(
suspensions were prepared as described above and deposited expressed here are based on at least three such independent
onto freshly cleaved mica. Immobilization was achieved using experiments.
electrostatic forces between the opposite charges of mica and
polymeric suspensions. The sample was carefully dried over-
Confocal microscopy
night under vacuum before each measurement. Lipoplexes Intracellular delivery of fluorescein labelled pDNA was followed
were prepared as described above for the DLS measurements.
using a confocal microscope. For microscopic analysis cells
were first plated on glass cover slips in 12-well cell culture
plates. Transfection experiments were performed using 1 µg
Cell culture
Three different cell lines, namely Caco-2, HEK293T and HeLa, of labelled pDNA as discussed above. The molar ratios of
were cultured in a cell culture medium, DMEM (Dulbecco’s maximum transfection efficiency for different formulations
4312 | Org. Biomol. Chem., 2015, 13, 4310–4320
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
were taken into account for transfection experiments. The cells
were washed twice properly with DPBS 48 h post transfection
and fixed using 4% paraformaldehyde solution for 10 min.
Again, after a proper wash with DPBS, cell nuclei staining were
performed using DAPI (300 nM). Cells were again washed to
remove excess dye and the cover slips were mounted on glass
slides over an antifade reagent. Thus-prepared samples were
viewed under a confocal laser scanning microscope (LSM
meta, Zeiss).
Results and discussion
Earlier work from this laboratory has shown a significant
improvement in the transfection efficiency of cationic chole-
sterol based cytofectins when the cationic head group is linked
Scheme 2 Reaction conditions. (i) ω-Bromoalkanoyl chloride, anhyd.
AlCl
3
, CH
2
Cl
2
, RT, 12 h, 45–55%; (ii) NaBH
4 3
/AlCl , THF, RT, 4 h, 75–80%;
(
iii) Chol-OCH
2
CH NMe , MeOH–EtOAc (1 : 1), screw-top pressure tube,
2
2
to the steroid backbone via an ether type linkage as compared 70 °C, 40–50%.
4
2,43
to its ester or urethane counterparts.
Subsequent studies
showed that cholesterol-based gemini lipids based on ether
linkage with varying length of the spacer between the
ammonium head groups formed stable formulations, which
by acylation of ferrocene using an appropriate ω-bromoalka-
noyl chloride in the presence of anhydrous AlCl in dry CHCl
This was followed by the complete reduction of the ketone
group by NaBH /AlCl in dry THF to produce 1-(ω-bromoalkyl)
ferrocenes (3, 4) and 1,1′-bis(ω-bromoalkyl) ferrocenes (7, 8)
respectively in 85–90% and 75–80% yield. Monomeric
lipids (CHM-C6F and CHM-C11F) and gemini lipids (CHD-C6F
and CHD-C11F) were obtained by refluxing the key
precursor cholest-5-en-3β-oxyethan-N,N-dimethyl amine (Chol
3
5,36
3
.
3
turned out to be also superior transfecting agents.
In the
present work, we describe the effect of introducing a redox-
active ferrocene moiety on the gene delivery capability of two
series of redox-active cholesterol based cationic lipids. For the
monomeric lipids (CHM-C6F and CHM-C11F), the redox-
responsive ferrocene moiety was attached on to the cationic
head group via either six-carbon or eleven-carbon spacers. In
the case of the gemini lipids (CHD-C6F and CHD-C11F), the
two cationic cholesterol units were connected via six- and
eleven-carbon spacers on either side of the cyclopentadienyl
rings of ferrocene. Here, the ferrocene moiety becomes a part
of the polymethylene spacer which joins the two cationic head
groups. The synthesis of these cationic lipids as outlined in
Schemes 1 and 2 was accomplished as briefly described below.
4
3
OCH
and 7 or 8 (0.07 mmol) respectively in dry MeOH–EtOAc (4 mL,
: 1, v/v) over a period of 4–6 days in a screw-top pressure tube,
2 2 2
CH NMe ) (0.2 mmol) with either 3 or 4 (0.22 mmol)
1
until TLC indicated the completion of each reaction. After
that, the reaction mixture was cooled and the solvent was evap-
orated to furnish a crude solid, which was purified by column
chromatography over neutral alumina using CHCl
MeOH. The isolated yields of the pure lipids ranged from 40 to
0%. The purities of these lipids were ascertained by TLC; the
ranged from 0.2 to 0.3 in 20 : 1 and 10 : 1 CHCl –MeOH for
the monomeric and gemini lipids, respectively. Each new
3 3
to CHCl –
Chemistry
5
Briefly, ω-bromo aliphatic-1-monoacylferrocenes (1, 2) and
ω-bromo aliphatic-1,1′-diacylferrocenes (5, 6) were synthesized
R
f
3
1
13
lipid was fully characterized by H-NMR, C-NMR, mass
spectrometry and elemental analysis (ESI†).
Physicochemical characterizations of co-liposomes
The co-liposomes (DOPE/cationic lipid) considered for
physicochemical characterizations were those which showed
maximum gene transfection capabilities (DOPE–CHM-C6F,
2.5 : 1; DOPE–CHM-C11F, 1.5 : 1; DOPE–CHD-C6F, 4 : 1; and
DOPE–CHD-C11F, 3 : 1) as discussed in the Gene transfection
biology section.
Atomic force microscopy (AFM)
We examined the shape and size distribution of the co-liposo-
mal formulations containing each of the cationic ferrocenyl
lipids along with DOPE at the optimized ratio (from gene
transfection) by atomic force microscopy. Although the lipid
Scheme 1 Reaction conditions. (i) ω-Bromoalkanoyl chloride, anhyd.
AlCl
3
, CH
2
Cl
2
, RT, 12 h, 55–60%; (ii) NaBH
CH NMe , MeOH–EtOAc (1 : 1), screw-top pressure tube,
0 °C, 40–50%.
4 3
/AlCl , THF, RT, 4 h, 85–90%;
(
iii) Chol-OCH
2
2
2
7
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 4310–4320 | 4313

View Article Online
Paper
Organic & Biomolecular Chemistry
Dynamic light scattering (DLS) and zeta potential
measurements
The average hydrodynamic diameters and zeta potential of
each redox lipid/DOPE mixed aggregate (transfection opti-
mized) in their unoxidized and oxidized states are shown in
Table 1. When the ferrocene moiety is present in the native
unoxidized state, monomeric lipid–DOPE (1 : 2.5, CHM–C6F
and 1 : 1.5, CHM–C11F) suspensions formed aggregates of size
1
1
30–150 nm while the gemini lipid–DOPE (1 : 4, CHD–C6F and
: 3, CHD–C11F) suspensions formed slightly larger aggregates
of size 170–200 nm. In the case of the oxidized lipid/DOPE sus-
pensions, the size of the resulting aggregates was found to be
larger than that of their counterparts prior to oxidation. This
could be due to the formation of ferrocenium species upon
oxidation, which probably introduced a greater inter-lipidic
Fig. 2 Representative atomic force micrographs of the co-liposomes of repulsion into their vesicular aggregates. It may be worth
cationic ferrocenyl cholesterol and DOPE at the transfection optimized
ratio. (A) CHM-C6F (Red.), (B) CHM-C6F (Ox.), (C) CHD-C6F (Red.) and
noting that the sizes of various ferrocenyl lipid/DOPE co-aggre-
gates as obtained from the DLS studies were comparable to
those obtained from the AFM measurements.
(
D) CHD-C6F (Ox.).
The mixed lipid suspensions were further characterized by
suspensions from both the reduced and oxidized lipids zeta potential measurements which allowed the determination
formed nearly spherical vesicles, they differed in their size. of the effective charge of the exposed surface of the lipid co-
As shown in Fig. 2, we observed that the vesicles dilated upon aggregates in aqueous medium. As we moved from the mono-
oxidation. The co-liposomal formulations containing meric to the gemini series, we observed an increase in the zeta
CHM-C6F in its reduced state formed vesicles of size potential of the liposomes from +30 mV to +43 mV. This may
7
0–100 nm which upon oxidation increased to 120–150 nm. A be due to the presence of two cationic ammonium head
similar trend was observed in the case of the dimeric redox groups present in the gemini lipids. All the oxidized lipid/
lipid, CHD-C6F, which increased in size from 120–150 nm to DOPE formulations displayed much higher zeta potential
1
60–200 nm after oxidation of ferrocene in co-liposomes. It values than their reduced counterparts. This could be ascribed
may be noted that the net charges of these redox-active cat- to an increase in the global charge density of all the redox-
ionic lipids change upon oxidation. In the case of monomeric active lipids upon oxidation of the ferrocene moiety.
lipids, complete oxidation of the ferrocene moiety of CHM-C6F
In addition, we performed the DLS and zeta potential
and CHM-C11F to ferrocenium leads to a change in overall measurements of all co-liposomes prepared in 10 mM HEPES
molecular charge from +1 to +2. On the other hand, for gemini buffer (pH = 7.4) and compared them with the liposomes pre-
lipids CHD-C6F and CHD-C11F, the net charge changes from pared in Li SO as shown in Table S1, ESI.† We did not
2
4
+
2 (reduced) to +3 (oxidized) upon complete oxidation. Thus it observe any significant change in the sizes of the co-liposomes
is likely that owing to the increased electrostatic repulsions prepared in Li SO and HEPES buffer (pH = 7.4). However, we
2
4
among the positively charged ferrocenium moieties along with found a slight decrease in zeta potential values for co-lipo-
that of the cationic ammonium head groups, we observed somes prepared in HEPES buffer compared with those pre-
4
4
larger liposomes.
pared in Li SO .
2 4
a
Table 1 Average hydrodynamic diameters and zeta potential of the redox-active cationic cholesterol/DOPE aggregates
Avg. hydrodynamic diameter (nm)
PDI
Zeta potential (mV)
Red.
Mixed liposomes
Red.
Ox.
Red.
Ox.
Ox.
DOPE/CHM-C6F
DOPE/CHM-11F
DOPE/CHD-C6F
DOPE/CHD-C11F
130 ± 5
180 ± 25
200 ± 20
225 ± 30
260 ± 30
0.14
0.11
0.09
0.16
0.22
0.25
0.24
0.28
31 ± 0.5
30 ± 1
43 ± 1.7
41 ± 2
38 ± 2
39 ± 2.5
51 ± 2
53 ± 3
145 ± 10
170 ± 10
195 ± 20
a
2 4
Hydrodynamic diameters and zeta potential of co-liposomes of each redox-active cationic cholesterol/DOPE prepared in Li SO at the
transfection optimized molar ratios. DOPE–CHM-C6F (2.5 : 1), DOPE–CHM-C11F (1.5 : 1), DOPE–CHD-C6F (4 : 1) and DOPE–CHD-C11F (3 : 1).
The measurements were performed at a final cationic lipid concentration of 0.1 mM. Each value is the mean ± S.D of five independent
measurements. PDI: polydispersity index.
4314 | Org. Biomol. Chem., 2015, 13, 4310–4320
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
X-ray diffraction (XRD) analysis
efficiency at DOPE ratios 4 : 1 and 3 : 1 respectively (Fig. S4,
ESI†).
In order to provide a plausible illustration of the lipid mole-
cular arrangement in the bilayer, we performed X-ray diffrac-
tion (XRD) experiments for self-supported cast films of the co-
liposomes (Fig. S2, ESI†). The XRD analysis revealed a series of
higher order reflections in their diffraction pattern that are
characteristic of lamellar phases. The bilayer widths calculated
from the first order diffraction peak using Bragg’s equation
All the DOPE optimized formulations of monomeric and
gemini cholesterols as discussed above were used in sub-
sequent gene transfection studies. We used different cationic
cholesterol/pDNA ratios (molar ratios) to elicit maximum
transfection efficacies. The ranges of molar ratios used for the
monomeric and gemini cholesterols were 0.5 : 1 to 2 : 1 and
0.25 : 1 to 1 : 1 respectively. We analyzed GFP expression for
(
nλ = 2d sin θ) for the co-liposomes of CHM-C6F and CHD-C6F
various controls, i.e., cells alone, cells treated with only co-lipo-
somes (at various concentrations used in different molar
ratios) and cells treated with pDNA alone (0.8 µg) which did
not reveal any noticeable activity (Fig. S5, ESI†). Transfection
data obtained from the flow cytometry for lipoplexes of various
co-liposomes revealed that the cationic lipids CHM-C6F and
CHD-C6F with C6 alkyl chain exhibited an optimal GFP
expression in Caco-2 cells that was significantly better than
that of a commercial formulation, Lipofectamine 2000 (Lipo
were found to be 37.8 Å and 45.1 Å respectively. Based on the
energy optimized structure, the theoretical bilayer width of
each lipid, which is two times the distance from the ‘head’
nitrogen atom to the ‘tail’ terminal carbon, is known to be
3
4
4
2 Å. The lower value of theoretical bilayer widths for mono-
meric lipid when compared with experimentally obtained
values suggests that the co-liposomes of the monomeric lipids
could adopt an interdigitated bilayer arrangement with the
ferrocene alkyl chain being looped into the bilayer owing to its
hydrophobic nature. Similarly, the experimentally obtained
bilayer widths for gemini lipids were approximately similar to
the theoretical value, which suggested that the co-liposomes of
gemini lipids could assume a regular bilayer like structure
where the hydrophobic spacer consisting of 1,1′-dialkyl ferro-
cene could also loop into the bilayer to avoid its unfavourable
contact with water (Fig. S3, ESI†).
2
000). The transfection activity of the cationic lipid with a C11
alkyl chain, CHM-C11F, also appeared to be better than Lipo
000, and CHD-C11F remained comparable to Lipo 2000
Fig. 3). Among cationic monomeric lipids, CHM-C6F appeared
2
(
to be a relatively better gene transfectant than CHM-C11F. In
the case of gemini lipids, also the lipids with a C6 alkyl chain
showed better GFP expression than those with a C11 alkyl
chain.
Transfection experiments in HEK293T cells also revealed
the same order of transfection capability of cationic chole-
Gene transfection biology
To assess the gene transfection capability of ferrocenylated cat- sterols as that observed in Caco-2 cells. Cationic cholesterol
ionic monomeric and dimeric cholesterols, we conducted with C6 alkyl chain (CHM-C6F and CHD-C6F) showed relatively
transfection experiments in human epithelial colorectal better gene expression in comparison with Lipo 2000 (Fig. 4).
adenocarcinoma (Caco-2) cells, human embryonic kidney
(
(
HEK)293T cells and human epithelial cervical carcinoma
HeLa) cells. Transfection experiments were carried out using
0
.8 µg of pEGFP-C3 plasmid DNA (pDNA) in 10% serum (FBS)
containing cell culture medium (DMEM). Flow assisted cell
sorting (FACS) was undertaken to quantify the transfection
levels both in terms of the number of cells showing the GFP
expression in the transfected population (%GFP positive cells)
and the level of GFP expression (mean fluorescence intensity
(MFI)). In order to prepare the co-liposomal formulations,
each of the cationic lipids was admixed with the neutral helper
phospholipid DOPE at different molar ratios (DOPE–lipid;
0
.5 : 1 to 3 : 1 for the monocationic lipids and 1 : 1 to 6 : 1 for
the dimeric lipids) which afforded stable co-liposomes.
Gene transfection experiments were first performed in
Caco-2 cells in the presence of 10% serum (FBS) to optimize
the best proportion of DOPE in liposomal formulations
leading to maximum transfection efficacies. Formulations with
varying DOPE/cationic cholesterols were used to transfect
Fig. 3 Optimization of maximum transfection activity of redox-active
cationic monomeric (A and C) and gemini cholesterols (B and D) in
pDNA at a fixed molar ratio (cationic lipid/pDNA) of 1 : 1 and Caco-2 cells. The EGFP expression profile is represented as (A and B) %
GFP positive cells and (C and D) the mean fluorescence intensity (MFI).
0
.5 : 1 for monomeric and gemini cholesterols respectively.
Transfection studies were performed using pDNA (pEGFP-C3, 0.8 µg) in
0% FBS containing cell culture medium and GFP expression was ana-
lyzed 48 h post transfection. Transfection activities were analyzed
Monocationic cholesterols CHM-C6F and CHM-C11F showed
maximum transfection efficiency at relatively lower DOPE
ratios of 2.5 : 1 and 1.5 : 1 respectively than the gemini lipids
1
for statistical significance in comparison with Lipo 2000 (*P < 0.05,
CHD-C6F and CHD-C11F that showed maximum transfection **P < 0.01 and ***P < 0.001, two-tailed Student’s t-test).
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 4310–4320 | 4315

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 5 Effect of oxidation of ferrocene on the transfection activity of
optimized lipoplexes (Red.) of monomeric and gemini cholesterols in
Caco-2, HEK293T and HeLa cell lines. Transfection efficacies are rep-
resented as % GFP positive cells (A) and the mean fluorescence intensity
(B). Transfections were performed using pDNA (pEGFP-C3, 0.8 µg) in
10% FBS containing cell culture medium and GFP expression was ana-
lyzed 48 h post transfection. Transfection activities of oxidized co-lipo-
somes were analyzed for statistical significance in comparison with
reduced co-liposomes (*P < 0.05, **P < 0.01 and ***P < 0.001, two-
tailed Student’s t-test).
which had a ferrocene moiety in the unoxidized neutral state,
we proceeded to check the transfection efficacies of these
Fig. 4 Optimization of transfection activity of the redox-active cationic
monomeric (A, C, E and G) and gemini cholesterols (B, D, F and H) in
HEK293T and HeLa cells. GFP expression profile is represented as (A, B, lipids after oxidation of the ferrocene. Each of the transfection
C and D) % GFP positive cells and (E, F, G and H) the mean fluorescence
intensity (MFI). Transfection studies were performed using pDNA
experiments involved freshly prepared oxidized co-liposomes
for lipoplex preparation. The lipoplexes derived from oxidized
(
pEGFP-C3, 0.8 µg) in 10% FBS containing cell culture medium and GFP
co-liposomes led to significantly reduced levels of gene trans-
fection at all the molar ratios studied in each cell line. The oxi-
dation state of ferrocene was found to have a profound effect
on gene transfection efficacies of both cationic monomeric
and gemini cholesterols wherein GFP expression, especially in
terms of mean fluorescence intensity (MFI), dropped down
drastically. It should be noted that transfections with higher
and lower MFI values reveal high and low GFP expressions
expression was analyzed 48 h post transfection. Transfection activities
were analyzed for statistical significance in comparison with Lipo 2000
(
*P < 0.05, **P < 0.01 and ***P < 0.001, two-tailed Student’s t-test).
However, the cationic lipid with a C11 alkyl chain showed a
relatively lower expression of GFP.
Transfection studies were also performed in HeLa cells.
Interestingly both the sets of cationic cholesterols with C6 and
C11 alkyl chains showed good expression of GFP which was
also higher than the commercial formulation (Fig. 4). Cationic
monomeric cholesterols showed a relatively better transfection
capability than their gemini (dimeric) counterparts. Among
the gemini lipids, the one with a C6 alkyl chain was slightly
better than the one with a C11 alkyl chain. GFP expression was
also analyzed using fluorescence microscopy where the level of
GFP expression was in good agreement with the data obtained
from flow cytometry in all the cell lines examined (Fig. 6).
45
respectively. This in turn reported that co-liposomes of
the cationic cholesterols lost their transfection activity after
oxidation of ferrocene (Fig. 5).
The differences in GFP expression in all the three cell lines
were also analyzed using fluorescence microscopy where cells
treated with lipoplexes containing cationic cholesterols with
oxidized ferrocene showed a significantly lower level of GFP
expression when compared to the lipoplexes of reduced co-
liposomes. Representative images of GFP expression for trans-
fections mediated by reduced and oxidized CHM-C6F co-lipo-
somes are shown below as a comparison. Transfection using
Lipo 2000 following the manufacturer’s protocol has also been
shown as a positive control (Fig. 6).
Oxidation state of ferrocene, a key determinant of the
transfection efficiency
After the optimization of the maximum transfection activity
We also looked at the gene expression levels at higher
mediated by cationic monomeric and gemini cholesterols serum concentration where
a lipoplex incubation with
4316 | Org. Biomol. Chem., 2015, 13, 4310–4320
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 7 (A) Representative flow cytometry histograms for the cellular
internalization of labelled pDNA after 6 h lipoplex treatment in HeLa
cells in the presence of 10% serum containing medium. (B) GFP
expression (MFI) profile derived from the treatment of lipoplexes of
CHM-C6F (Red. and Ox.) and AA treated oxidized lipoplexes. Fluo-
rescence microscopic images of HeLa cells depicting GFP expression in
transfections mediated by AA treated CHM-C6F (Ox.) lipoplexes (D) in
comparison with CHM-C6F (Ox.) lipoplexes (C).
Fig. 6 Representative fluorescence microscopic images for the com-
parison of the GFP expression obtained from transfections mediated by
CHM-C6F co-liposomes (Red. and Ox.) in all the three cell lines (Caco-2,
HEK293T and HeLa) using Lipo 2000 as a positive control. Images were
taken 48 h post transfection. Scale bar = 100 µm.
cells was performed in 50% serum concentration. We chose
CHM-C6F, which showed significant level of gene expression
in all the three cell lines in the presence of 10% serum concen-
tration. Higher serum concentration of course reduced the
transfection efficiency of reduced co-liposomes slightly while
transfections mediated by oxidized ferrocene based lipoplexes
were still inactive (Fig. S6, ESI†). GFP expression analysis using
fluorescence microscopy substantiated the flow cytometry data
where oxidized co-liposome transfected cells showed signifi-
cantly lower intensity of GFP expression (Fig. S7, ESI†).
This observation was in accordance with previously reported
assist the delivery of DNA molecules to a specific cell
population.
EGFP expression that originated from the lipoplexes derived
from the co-liposomal formulations containing the reduced
and oxidized ferrocene was also followed at the level of intra-
cellular delivery of labelled pDNA using flow cytometry and
confocal microscopy experiments. HeLa cells were transfected
with fluorescein labelled pDNA-lipoplexes derived from the
reduced or oxidized CHM-C6F and CHD-C6F co-liposomes for
an incubation period of 6 h and analyzed under a flow
cytometer. It showed that a greater pDNA fluorescence was
observed in the cells transfected with lipopolexes of reduced
co-liposomes than that of oxidized ones (Fig. 7A). A representa-
tive quantification of pDNA fluorescence positivity in
CHM-C6F co-liposome (Red. and Ox.) transfected HeLa cells is
also shown in Fig. S8 (ESI†) wherein the reduced and oxidized
co-liposome transfected cells showed ∼80% and ∼30% pDNA
fluorescence positivity respectively. To corroborate the obser-
vation we undertook confocal microscopy experiments. Inter-
estingly, higher pDNA fluorescence was observed in the HeLa
cells transfected with reduced CHM-C6F co-liposomes and, as
expected, a lower number of cells was found to exhibit pDNA
fluorescence after transfection with oxidized CHM-C6F co-lipo-
somes as shown in Fig. 8A and B. Lipo 2000 served as a posi-
tive control in the experiment and the observed pDNA
fluorescence appeared almost similar to reduced CHM-C6F as
shown in Fig. 8C.
1
9,20
literature on other ferrocene based carriers.
Once the transfection levels were optimized for both
reduced and oxidized ferrocenylated cholesterol containing
lipoplexes, we performed the study by asking whether the
transfection activity of oxidized lipoplexes could be resumed
by following the reduction of lipoplexes by ascorbic acid (AA)
2
1
as reported elsewhere. Interestingly, transfections performed
by CHM-C6F (Ox.) in AA containing cell culture medium could
lead to gene expression levels comparable to those of
CHM-C6F (Red.), and the phenomenon was quite reproducible
in all the three cell lines studied (Fig. 7B). This was also con-
firmed by visual analysis performed under a fluorescence
microscope where ostensibly significant difference in GFP
expression was observed for CHM-C6F (Ox.) treatment and
CHM-C6F (Ox.)/AA treatment (Fig. 7C and D). This could be
attributed to the extracellular exposure of the oxidized lipo-
plexes to AA leading to activity resumption after reduction.
Thus, the liposomal formulations shown here could be
harnessed under the influence of redox response with tren-
chant control extracellularly. Such a phenomenon could be
Characterization of the lipoplexes
quite useful for gene delivery applications where the gene Physical properties like the aggregate size and zeta potential
expressions are likely to be considered spatially or temporally. are crucial parameters which influence the transport of DNA
4
6,47
This controlled activation of inactive lipoplexes could also across the cell membranes.
We performed dynamic light
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 4310–4320 | 4317

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 8 Confocal microscopy images of HeLa cells after 6 h of lipoplex
treatment in the presence of serum. Lipoplex treatments consisted of
CHM-C6F (Red.) (A), CHM-C6F (Ox.) (B) and Lipo 2000 as a positive
control (C). Panels A, B and C represent (left to right) pDNA fluor-
escence, DAPI stained nuclei and overlay of the previous two along with
bright field images. Lipoplexes were prepared using 1 µg of fluorescein
labelled pDNA at molar ratios of maximum transfection and incubated
for 30 min before adding to the cells. Scale bar = 20 µm.
scattering measurements to characterize the lipoplexes pre-
pared from pDNA and the co-liposomes of redox-active lipids.
Fig. 9A shows the hydrodynamic size of lipoplexes obtained
from both the reduced and oxidized forms of the cationic
cholesterol/DOPE formulations at various cationic cholesterol/
pDNA molar ratios. At lower ratios, CHM-C6F co-liposomes
formed lipoplexes of ∼250 nm size, which increased to 500 nm
at a molar ratio of 1 and then decreased to a much smaller
Fig. 9 Hydrodynamic diameters (A) and zeta potential (B) of lipoplexes
size of ∼180 nm. On the other hand, the gemini lipid formed using solutions of pDNA and co-liposomes of redox-active lipids
at various lipid/pDNA molar ratios. Atomic force microscopy images of
optimized lipoplexes of (C) CHM-C6F (Red.), (D) CHM-C6F (Ox.), (E)
CHD-C6F (Red.) and (F) CHD-C6F (Ox.).
CHM-C6D co-liposomes in the reduced state formed lipoplexes
of relatively larger size (∼350 nm) at lower ratios and formed
smaller sized lipoplexes (∼200 nm) from a molar ratio of 1 to
5
. The lipoplexes formulated using oxidized CHM-C6F and
CHD-C6F co-liposomes were considerably larger than the lipo-
plexes formed using their reduced counterparts at all the
We also measured the zeta potential and size of the repre-
molar ratios investigated. This might be another reason for the sentative oxidized lipoplexes (lipid/pDNA; molar ratio, 2) of
low transfection capability of the oxidized lipid formulations. CHM-C6F treated with ascorbic acid (AA). It revealed an
Fig. 9B illustrates the zeta potential of the lipoplexes evident increase in zeta potential values of oxidized lipoplexes
formed using reduced and oxidized lipids and pDNA at after AA treatment (from −3.5 ± 3 mV to 16 ± 5 mV) which was
various molar ratios. Lipoplexes formed using the reduced comparable to that of lipoplexes of reduced co-liposomes (19 ±
forms of both CHM-C6F and CHD-C6F co-liposomes showed a 6 mV) (Fig. S9B, ESI†). It was also observed that there was a sig-
transition from negative to positive zeta potential with increas- nificant decrease in the size of oxidized lipoplexes after AA
ing molar ratios with the change in sign occurring at molar treatment (from ∼900 nm to ∼200 nm) (Fig. S9A, ESI†). This
ratios of 1.5 and 1 respectively. The oxidized forms of the lipid could have possibly led to the resumption in transfection
formulations under consideration also showed negative zeta activity of oxidized lipoplexes upon treatment with AA.
potential at lower ratios but showed much less positive values
even at higher molar ratios.
The morphology of redox-active cationic lipids/pDNA was
further investigated by atomic force microscopy, and all the
4318 | Org. Biomol. Chem., 2015, 13, 4310–4320
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
lipoplexes were prepared at the molar ratio that showed give relatively lower cell viabilities. This trend was consistent
optimum transfection efficiency. Lipoplexes formed from the among all the cell lines except that the monomeric lipid with
cationic cholesterols in their reduced state were observed as C11 alkyl chain, CHM-C11F, appeared to be slightly more toxic
near-spherical shaped aggregates with an average size of than CHM-C6F only in HeLa cells. All the formulations showed
1
00–200 nm (Fig. 9C and E). However, the lipoplexes formed better cell viabilities when compared to the commercial formu-
from their oxidized counterparts exhibited larger particle sizes lation, Lipo 2000, in all three cell lines. We observed that
around 500–900 nm (Fig. 9D and F). This indicates that the cationic gemini cholesterols appeared relatively less cytotoxic
particle size of the lipoplexes could be greatly influenced by in comparison with the cationic monomeric cholesterols. Cell
the oxidation state of the redox-active cationic cholesterols. viability experiments revealed that transfection with lipoplexes
The plausible reason for the relatively larger size of the containing lipids with oxidized ferrocene led to improved cell
lipoplexes derived from oxidized co-liposomes could be their viabilities than the lipids with unoxidized ferrocene. A repre-
inability to effectively condense the pDNA into compact aggre- sentative quantitation of cell viabilities is shown in Fig. 10 for
gates. This also resulted in lower zeta potential values due to all efficient formulations in the three cell lines 48 h post
the presence of loosely bound pDNA on the outer surfaces of transfection.
these lipoplexes of oxidized co-liposomes. This might have led
to the poor transfection efficiency of oxidized co-liposomes.
It may also be noticed that the lipoplex sizes observed by AFM Conclusions
are somewhat smaller than those observed from the DLS
measurements. This variation in AFM and DLS results could
be due to shrinkage of the particles induced during the partial
We demonstrate here efficient pDNA delivery to three cell lines
of different origins mediated by co-liposomes of cationic
cholesterol lipids (monomeric and gemini) under the influ-
drying process for the AFM sample preparation, whereas DLS
4
8
ence of redox control. The monomeric cholesterols possess
ferrocene at the head group region separated by alkanediyl
spacers of different lengths (six and eleven methylene units).
In the gemini cholesterols, the ferrocene unit is flanked by
alkanediyl chains on both its cyclopentadienyl rings, and
forms part of the spacer segment which connects the cationic
head groups of the monomeric units. The oxidation state of
the ferrocene moiety in the cationic cholesterols (that is,
whether they are present in the reduced or oxidized state) is a
critical determinant to mediate cell transfection. Co-liposomal
formulations possessing ferrocenylated cholesterol in its
natural unoxidized state are shown to produce high GFP levels
even better than those of Lipofectamine 2000 in the presence
of serum. On the other hand, oxidized ferrocenylated chole-
sterols exhibited significantly diminished levels of GFP
expression. The transfection activities could easily be resumed
after exposure of oxidized lipoplexes to ascorbic acid (AA). GFP
expression observed under fluorescence microscopy was in
good agreement with the flow cytometry (FACS) results. Trans-
fection experiments conducted using fluorescein labelled
pDNA showed higher lipoplex internalization for reduced co-
liposomes than the oxidized ones. Lipoplexes derived from
these liposomal formulations did not impart any acute toxicity
in the range of optimized transfection concentration levels.
All the co-liposomal formulations and their lipoplexes were
thoroughly characterized by DLS, zeta potential and AFM
measurements. Overall, the transfection efficient redox-active
cholesterols presented here could be useful for on-demand/
controlled and effective delivery of pDNA to cancer cells under
the influence of redox sensitive behavior of the ferrocene
moiety.
data provide the sizes of ‘intact’ hydrated particles. Interest-
ingly when ferrocenes are located in the polar head group
region, in the cationic cholesterols under the current investi-
gation, we observed higher zeta potential and smaller size for
the lipoplexes formed from the cationic cholesterols in their
reduced state. This might have led to better internalization by
the cells which promoted high levels of transgene expression
by the lipids with the ferrocene in its reduced state. This
result is similar to an earlier report where the ferrocenes were
2
7
at the end of the hydrophobic tails.
Cell viability assay
All transfection efficient formulations were screened for their
toxicity using the MTT assay towards all the cell lines used in
the study under actual transfection experiment conditions. As
a whole there was no obvious toxicity observed from the lipo-
plexes of all the cationic cholesterols except the monomeric
lipid with C6 alkyl chain, CHM-C6F, which appeared to
Fig. 10 Cytotoxicity profile of all the transfection optimized lipoplexes
of reduced and oxidized co-liposomes of monomeric and gemini
cholesterols in Caco-2, HEK293T and HeLa cells. Cell viability experi-
ments (MTT assay) were performed under actual transfection conditions
using Lipo 2000 as a positive control. Experiments were conducted in Notes and references
triplicates and results presented here are based on at least three inde-
pendent experiments.
1 X. Guo and F. C. Szoka, Acc. Chem. Res., 2003, 36, 335.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem., 2015, 13, 4310–4320 | 4319

View Article Online
Paper
Organic & Biomolecular Chemistry
2
3
4
5
6
7
8
9
P. Brown, C. P. Butts and J. Eastoe, Soft Matter, 2013, 9, 26 B. S. Aytar, J. P. E. Muller, S. Golan, Y. Kondo, Y. Talmon,
2
365.
M. S. Shim and Y. J. Kwon, Adv. Drug Delivery Rev., 2012,
4, 1046.
N. L. Abbott and D. M. Lynn, J. Colloid Interface Sci., 2012,
387, 56.
27 J. P. E. Muller, B. S. Aytar, Y. Kondo, D. M. Lynn and
N. L. Abbott, Soft Matter, 2012, 8, 6608.
28 C. L. Debby, A. A. Arnold and J. Mauzeroll, J. Am. Chem.
Soc., 2010, 132, 15120.
29 K. Kumar, G. Vulugundam, P. Kondaiah and
S. Bhattacharya, J. Mater. Chem. B, 2015, 3, 2318.
30 M. S. Brow and J. L. Goldstein, Cell, 1997, 89, 331.
6
V. Budker, V. Gurevich, J. E. Hagstrom, F. Bortzov and
J. A. Wolff, Nat. Biotechnol., 1996, 14, 760.
O. V. Gerasimov, J. A. Boomer, M. M. Qualls and
D. H. Thompson, Adv. Drug Delivery Rev., 1999, 38, 317.
D. C. Drummond, M. Zignani and J. C. Leroux, Prog. Lipid
Res., 2000, 39, 409.
M. B. Yatvin, J. N. Weinstein, W. H. Dennis and 31 M. T. Bengoechea-Alonso and J. Ericsson, Curr. Opin. Cell
R. Blumenthal, Science, 1978, 202, 1290. Biol., 2007, 19, 215.
Z. H. Huang, W. J. Li, J. A. MacKay and F. C. Szoka, Mol. 32 K. Simons and D. Toomre, Nat. Rev. Mol. Cell Biol., 2000,
Ther., 2005, 11, 409.
1, 31.
P. Meers, Adv. Drug Delivery Rev., 2001, 53, 265.
33 A. Bajaj, P. Kondaiah and S. Bhattacharya, Bioconjugate
1
0 C. Wang, Q. Chen, Z. Wang and X. Zhang, Angew. Chem.,
Int. Ed., 2010, 49, 8612.
Chem., 2008, 19, 1640.
34 S. Bhattacharya and Y. K. Ghosh, Langmuir, 2001, 17,
1
1
1 K. Hoshino, K. Suga and T. Saji, Chem. Lett., 1986, 979.
2 T. Saji, K. Hoshino and S. Aoyagui, J. Am. Chem. Soc., 1985, 35 A. Bajaj, P. Kondiah and S. Bhattacharya, J. Med. Chem.,
2067.
1
07, 6865.
2007, 50, 2432.
1
1
3 R. L. McCarley, Annu. Rev. Anal. Chem., 2012, 5, 391.
4 J. C. Medina, I. Gay, Z. Chen, L. Echegoyen and
G. W. Gokel, J. Am. Chem. Soc., 1991, 113, 366.
36 A. Bajaj, P. Kondaiah and S. Bhattacharya, Bioconjugate
Chem., 2007, 18, 1537.
37 S. Bhattacharya and J. Biswas, Langmuir, 2010, 26,
4642.
38 A. Bajaj, P. Kondaiah and S. Bhattacharya, J. Med. Chem.,
2008, 51, 2533.
39 J. Biswas, S. K. Mishra, P. Kondaiah and S. Bhattacharya,
Org. Biomol. Chem., 2011, 9, 4600.
40 S. K. Misra, S. Naz, P. Kondaiah and S. Bhattacharya, Bio-
materials, 2014, 35, 1334.
41 S. K. Misra, J. Biswas, P. Kondaiah and S. Bhattacharya,
PLoS One, 2013, 8, e68305.
42 Y. K. Ghosh, S. S. Visweswariah and S. Bhattacharya, FEBS
Lett., 2000, 473, 341.
43 Y. K. Ghosh, S. S. Visweswariah and S. Bhattacharya, Bio-
conjugate Chem., 2002, 13, 378.
1
1
1
1
1
2
2
5 K. Wang, S. Munoz, L. Zhang, R. Castro, A. E. Kaifer and
G. W. Gokel, J. Am. Chem. Soc., 1996, 118, 6707.
6 H. Sakai, H. Imamura, Y. Kakizawa, Y. Kondo, N. Yoshino
and M. Abe, Colloids Surf., A:, 2004, 232, 221.
7 Y. Kakizawa, H. Sakai, K. Nishiyama and M. Abe, Langmuir,
1
996, 12, 921.
8 Y. Kakizawa, H. Sakai, A. Yamaguchi, Y. Kondo, N. Yoshino
and M. Abe, Langmuir, 2001, 17, 8044.
9 N. L. Abbott, C. M. Jewell, M. E. Hays, Y. Kondo and
D. M. Lynn, J. Am. Chem. Soc., 2005, 127, 11576.
0 C. M. Jewell, M. E. Hays, Y. Kondo, N. L. Abbott and
D. M. Lynn, J. Controlled Release, 2006, 112, 129.
1 B. S. Aytar, J. P. E. Muller, S. Golan, S. Hata, H. Takahashi,
Y. Kondo, Y. Talmon, N. L. Abbott and D. M. Lynn, 44 Z. Wang, H. Mohwald and C. Gao, Langmuir, 2011, 27,
J. Controlled Release, 2012, 157, 249. 1286.
2 S. Golan, B. S. Aytar, J. P. E. Muller, Y. Kondo, D. M. Lynn, 45 P. Chabaud, M. Camplo, D. Payet, G. Serin, L. Molreau,
2
2
2
N. L. Abbott and Y. Talmon, Langmuir, 2011, 27, 6615.
3 M. E. Hays, C. M. Jewell, Y. Kondo, D. M. Lynn and
N. L. Abbott, Biophys. J., 2007, 93, 4414.
P. Barthelemy and M. W. Grinstaff, Bioconjugate Chem.,
2006, 17, 466.
46 B. Ma, S. Zhang, H. Jiang, B. Zhao and H. Lv, J. Controlled
Release, 2007, 123, 184.
4 C. L. Pizzey, C. M. Jewell, M. E. Hays, D. M. Lynn,
N. L. Abbott, Y. Kondo, S. Golan and Y. Talmon, J. Phys. 47 Z. U. Rehman, I. S. Zuhorn and D. Hoekstra, J. Controlled
Chem. B, 2008, 112, 5849. Release, 2013, 166, 46.
5 M. E. Hays, C. M. Jewell, D. M. Lynn and N. L. Abbott, 48 C. M. Hoo, N. Starostin, P. West and M. L. Mecartney,
2
Langmuir, 2007, 23, 5609.
J. Nanopart. Res., 2008, 10, 89.
4320 | Org. Biomol. Chem., 2015, 13, 4310–4320
This journal is © The Royal Society of Chemistry 2015