Synthesis of acridine-nuclear localization signal (NLS) conjugates and evaluation of their impact on lipoplex and polyplex-based transfection

Article abstract of DOI:10.1016/j.ejmech.2005.07.015

We report on the synthesis of various acridine(Acr)-spacer-nuclear localization signal (NLS) peptide conjugates and explore whether their use as NLS-labeling agent of plasmidic DNA could improve gene nuclear import and expression into cells when mediated by synthetic DNA complexes. As the conditions of successful use of the NLS properties to enhance gene transfer are not clear, and with the aim of detecting and defining the requirements of NLS-enhanced transfection, we investigated gene delivery and expression into various cell lines with various DNA complexes (lipoplexes or polyplexes) that were formulated for various N/P ratios from various preformed Acr-spacer-NLS/DNA complexes (1:1, 5:1 and 10:1 molar ratio). For the in vitro transfection assays, the lipoplexes and polyplexes were formulated from the preformed Acr-spacer-NLS/DNA complexes and dioctadecylamidoglycylspermine (DOGS)/dioleylphosphatidylethanolamine (DOPE) 1:1:mol and branched polyethyleneimine (PEI) 25:kDa, respectively, which are very efficient in vitro gene transfer systems. We show by fluorescence experiments that part of the acridine-NLS-conjugates remains intercalated within the plasmid for most of the N/P lipoplexes and polyplexes investigated. We show that, as several other studies performed with NLS-conjugates that are not covalently linked to DNA, the expression of the transgene is in most cases not improved upon complexation of plasmidic DNA with NLS-intercalating conjugates prior to its formulation as lipoplexes or polyplexes.

Full text of DOI:10.1016/j.ejmech.2005.07.015

European Journal of Medicinal Chemistry 40 (2005) 1295–1306
www.elsevier.com/locate/ejmech
Original article
Synthesis of acridine-nuclear localization signal (NLS) conjugates
and evaluation of their impact on lipoplex and polyplex-based transfection
Caroline Boulanger, Christophe Di Giorgio, Pierre Vierling *
Laboratoire de Chimie Bioorganique, UMR 6001 CNRS, Faculté des Sciences, Université de Nice Sophia-Antipolis, 06108 Nice cedex 2, France
Received 6 April 2005; received in revised form 6 July 2005; accepted 18 July 2005
Available online 11 October 2005
Abstract
We report on the synthesis of various acridine(Acr)-spacer-nuclear localization signal (NLS) peptide conjugates and explore whether their
use as NLS-labeling agent of plasmidic DNA could improve gene nuclear import and expression into cells when mediated by synthetic DNA
complexes. As the conditions of successful use of the NLS properties to enhance gene transfer are not clear, and with the aim of detecting and
defining the requirements of NLS-enhanced transfection, we investigated gene delivery and expression into various cell lines with various
DNA complexes (lipoplexes or polyplexes) that were formulated for various N/P ratios from various preformed Acr-spacer-NLS/DNA com-
plexes (1:1, 5:1 and 10:1 molar ratio). For the in vitro transfection assays, the lipoplexes and polyplexes were formulated from the preformed
Acr-spacer-NLS/DNA complexes and dioctadecylamidoglycylspermine (DOGS)/dioleylphosphatidylethanolamine (DOPE) 1:1 mol and
branched polyethyleneimine (PEI) 25 kDa, respectively, which are very efficient in vitro gene transfer systems. We show by fluorescence
experiments that part of the acridine-NLS-conjugates remains intercalated within the plasmid for most of the N/P lipoplexes and polyplexes
investigated. We show that, as several other studies performed with NLS-conjugates that are not covalently linked to DNA, the expression of
the transgene is in most cases not improved upon complexation of plasmidic DNA with NLS-intercalating conjugates prior to its formulation
as lipoplexes or polyplexes.
© 2005 Elsevier SAS. All rights reserved.
Keywords: Gene transfer; DNA complex; Non-viral vector; Synthetic vector
1. Introduction
mal cells, gene suicide, inhibition of tumor vasculature
growth). Gene transfer systems based on lipoplexes or poly-
plexes have gained wide acceptance over the last decade as
gene transfer vectors. At present, however, their usefulness
and applications as therapeutical devices are limited by tran-
sient and low levels of gene expression observed in vivo. One
of the limiting steps responsible for a low gene expression
with these non-viral vectors resides in an inefficient intracel-
lular trafficking of DNA from the cytoplasm to the nucleus
[1,2]. This is particularly of concern with postmitotic and qui-
escent cells for which the nuclear membrane breakdown does
not occur periodically. Nevertheless, they present several
advantages as compared with adenoviral or retroviral vectors
including low-cost and large-scale production, and safety.
To overcome the cytoplasmic degradation of the gene via
an effective transport into the nucleus and to improve the effi-
ciency of gene expression, the use of nuclear localization sig-
nal (NLS) peptides for non-viral gene transfer has been widely
investigated ([3–17] and references therein). Except during
The development of gene transfer vectors allowing gene
expression into cells constitute not only a most attractive thera-
peutical approach (gene therapy) and a major breakthrough
in the biomedical field for the treatment of various inherited
and acquired diseases but also a powerful tool to study gene
and protein function and regulation. For example, the deliv-
ery of a therapeutical gene into cancer cells is indeed a prom-
ising technology for the development of various anticancer
strategies (replacement of a deficient tumor suppressor gene
to re-establish the balance between growth and apoptosis, inhi-
bition of a dominant oncogene, stimulation and modification
of immune effector cells to recognize and reject cancer cells,
amplification of tumor cell immunogenicity, chemo-/radio-
sensitization of tumor cells, chemo-/radio-protection of nor-
* Corresponding author. Tel.: +33 4 92 07 61 43; fax: +33 4 92 07 61 51.
E-mail address: vierling@unice.fr (P. Vierling).
0223-5234/$ - see front matter © 2005 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2005.07.015

1296
C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
mitosis, macromolecules such as proteins or nucleic acids can-
not enter the nucleus through the nuclear pore. The intra-
nuclear transfer of cellular and viral proteins, DNA, and/or
RNA occurs by means of an energy-dependent mechanism
which involves peptidic NLS sequences that bind to struc-
tures called nuclear pore complex (NPC) via transport recep-
tors such as importins a/b [3].
duplex by a combination of hydrophobic, electrostatic, hydro-
gen bonding, and dipolar forces [18]. The framework of the
Acr-spacer-NLS-conjugates consists of an Acr moiety bear-
ing a spacer of variable length in an attempt to separate the
NLS peptide from DNA and to overcome the problem of the
electrostatic interactions between the cationic NLS and
anionic DNA [8,9]. NLS attachment to this Acr-spacer was
performed via a chemically stable thioether bond between
the maleimide-labeled spacer unit and a cysteine residue at
the carboxy terminal of the peptide. As the conditions of suc-
cessful use of the NLS properties to enhance gene transfer
are not clear from literature data [3–5] and with the aim of
detecting and defining the requirements of NLS-enhanced
transfection, we investigated gene delivery and expression into
various cell lines (NIH-3T3, A549) with various DNA com-
plexes (lipoplexes or polyplexes) that were formulated for
various N/P ratios (0.8, 1.25, 2.5, 5 and 10) from various pre-
formed Acr-spacer-NLS/DNA complexes (1:1, 5:1 and
10:1 molar ratio). For the in vitro transfection assays, the
lipoplexes and polyplexes were formulated from the pre-
formed Acr-spacer-NLS/DNA complexes and dioctadecyla-
midoglycylspermine (DOGS)/dioleylphosphatidyletha-
nolamine (DOPE) (1:1 mol) and branched polyethyleneimine
(PEI) 25 kDa which are very efficient in vitro gene transfer
systems ([19–22] and references therein). Moreover, various
amounts of DNA (0.1 and 0.5 µg per well) were tested.
To date, literature indicates that the NLS approach has
potential for improving DNA nuclear delivery and expres-
sion with non-viral vectors. Most of the studies involved the
well-characterized simian virus (SV)40 T large antigen NLS-
sequence 126PKKKRKV132. The various approaches explored
differ mainly in the method used for the attachment (covalent
[4,8,6,10] or non-covalent [5,9,11–17]) of the NLS pep-
tide(s) to DNA and in the use of linear DNA [4,5,9] or circu-
lar (plasmidic) DNA [6,8–17]. While many of these ap-
proaches have met with limited success, a significant (10–
1000-fold) enhanced gene expression was obtained following
ligation of a NLS-oligonucleotide conjugate to one or both
ends of a linear DNA [4]. There is further a controversy con-
cerning the optimum number of NLS per DNA to enhance its
nuclear delivery [4–6,9]. Moreover, the neutralization of the
positive charges of the lysine and arginine residues of many
NLS peptides (which are critical for their interactions with
transport receptors and importins) by the negatively charged
DNA phosphates is likely expected to hinder cargo recogni-
tion and binding to NPC [8]. Very recently, a highly cationic
NLS peptide based on the HTLV sequence was shown to be
more effective for condensing plasmid DNA into discrete par-
ticles and for enhancing gene expression levels compared to
polylysine controls [9].
2. Chemistry
The synthetic approach to the targeted Acr-spacer-NLS-
conjugates (Scheme 1) is based on the elaboration of the key
acridine-spacer-maleimido synthons 7a–d onto which the
NLS peptide (via its cysteine-thiol function) was added. The
various synthons 7a–d were elaborated in three steps starting
from the dichloro-acridine derivative 4 and the mono-
protected a,x-diamino spacers 3a–c (which were obtained
using conventional procedures) and commercially available
3d, respectively. The first step consisted of nucleophilic dis-
placement of one aromatic chlorine of 4 by the amine of the
spacer units 3a–d which was performed in phenol and in the
presence of N-methyl morpholine (yields ranging from 70%
to 80%). The next step was the Boc-deprotection of the ter-
minal spacer amine function in 5a–d, which was achieved,
quantitatively in excess TFA. The target maleimide function
in 7a–d was then introduced by acylation of the resulting
amine in derivatives 6a–d with the heterobifunctional reagent
N-succinimidyl-4-(maleimidomethyl)cyclohexane carboxy-
late (SMCC) with yields ranging from 80% to 90%).
As a part of our contribution into this field, we explored
whether the use of a non-covalent attachment of a SV40 NLS
peptide to plasmidic DNA via intercalating conjugates (see
structures in Fig. 1) could improve gene nuclear import and
expression mediated by lipoplexes and polyplexes. As DNA
intercalating agent, we selected an acridine-derived moiety.
Acridine (Acr) is a heteroaromatic polycyclic molecule, which
inserts tightly, but reversibly, between two base pairs in a DNA
Condensation of the maleimides 7a–d with the NLS
peptide-thiol leading to the formation of a thioether bond was
best performed by adding a DMF solution of 7a–d to a solu-
tion of the NLS peptide in a 20 mM phosphate buffer at pH 7.
Prior to use, the commercial NLS peptide was stirred over-
night in a phosphate buffer 20 mM at pH 5–6 with tris(2-
carboxyethyl)phosphine hydrochloride (TCEP) [23] in order
Fig. 1
. Chemical structure and code name of the acridine-spacer-NLS-
conjugates used in this study.

C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
1297
Scheme 1. Synthetic route to the acridine-spacer-NLS-conjugates. i) H2N–(CH2)yNHBoc (y = 2 or 5)/EDC/HOBt/NEt3/DMF/25 °C (75–85% yield); ii)
Pd(PPh3)4/CH2Cl2/HNEt2 (35–70% yield); iii) NMM/phenol/100 °C (70–80% yield); iv) TFA/CH2Cl2 (quantitative); v) SMCC/DMF/NEt3 (80–90% yield);
vi) HCl 0.1 N in DMF, then NLS–(SH) (0.125 equiv.) in phosphate buffer pH 7 (NLS–(SH) = CGYGPKKKRKVGG).
to reduce the traces of (NLS-S)2 disulfide. The condensation
reaction was followed by analytical HPLC, and the Acr-
spacer-NLS-conjugates were purified by semi-preparative
HPLC. Their structure was confirmed by MALDI-TOF mass
spectrometry and that of their keyAcr-spacer-maleimido start-
ing materials 7a–d was attested by ESI-MS, and by 1H and
13C NMR.
3.1. Lipoplex and polyplex formation and characterization
The lipoplexes and polyplexes (also used for the in vitro
transfection assays described below) were formulated with
the various Acr-spacer-NLS-conjugates and pTG11236 plas-
mid (pCMV-SV40-luciferase-SV40pA; 5739 bps), and for 1,
5 and 10 molecules of NLS-conjugates per plasmid. They
were prepared in 5% glucose by preforming the acridine-
NLS/DNA complexes which were then mixed with a liposo-
mal dispersion of DOGS/DOPE 1:1 mol or a solution of
branched PEI 25 kDa, respectively, and for N/P ratios of 5,
2.5, 1.25 and 0.8 for the lipoplexes, and 10, 5, 1.25 and 0.8 for
the polyplexes (N = number of DOGS or PEI amine equiva-
lents; P = number of DNA phosphate equivalents). The
Acr-spacer-NLS: DNA complexes, and the lipoplexes and
polyplexes formulated therefrom were analyzed by gel elec-
trophoresis for attesting DNA complexation (results not
shown). The lipoplexes and polyplexes were further investi-
gated by dynamic light scattering for particle size measure-
ments. For comparison, various control lipoplexes and poly-
plexes were also formulated and analyzed. These controls
consisted of lipoplexes and polyplexes that were formulated
from (i) DNA alone, (ii) dichloro-Acr (4)/DNA, (iii) Acr-
spacer (7a–d)/DNA, or (iv) NLS peptide/DNA. Moreover,
phenanthroline (Phen) and ethidium bromide (ETB) as DNA
control intercalating agents were also tested for comparison.
3. Pharmacology
The observation that direct microinjection of DNA into
the cell nucleus led to protein expression in over 50% of
microinjected cells, whereas its injection into the cytoplasm
led to protein expression in less than 0.01% of cells, indi-
cates that inefficient DNA traffic from the cytoplasm to the
nucleus is a major barrier restricting transgene expression
[24].
The objectives of this study were to determine whether
acridine-spacer-NLS (SV40) peptide conjugates used as NLS-
labeling agent of DNA (via non-covalent intercalation of the
acridine moiety into DNA) can improve gene nuclear import
and expression of non-viral gene transfer systems (e.g.
lipoplexes and polyplexes). Aiming at this goal, we explored
the impact of the acridine-spacer-NLS-conjugates (Fig. 1) on
the luciferase expression level in cells transfected with
lipoplexes and polyplexes that were formulated from DNA
(= luciferase reporter gene)/acridine-spacer-NLS complexes
in comparison with non-labeled lipoplexes and polyplexes.
3.2. In vitro transfection
To explore the ability of the Acr-spacer-NLS-conjugates
to improve gene nuclear import and expression, we assayed

1298
C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
the transfection potential of all the DOGS/DOPE-based
lipoplexes and PEI-based polyplexes that were formulated
from Acr-spacer-NLS/DNA complexes described in the pre-
cedent section. To unambiguously highlight the impact of the
Acr-spacer-NLS-conjugates on transfection, all the control
formulations also described in the precedent section were
tested as well. Moreover, these DNA complexes were assayed
for low DNA amounts (i.e. 0.1 and 0.5 µg per well) and on
two different cell lines (i.e. NIH-3T3 and lung epithelial
A549 cells, from human pulmonary carcinoma) which pos-
sess the a-importins that bind to the NLS SV40 peptide
sequence used. These assays were performed in the presence
of 10% fetal calf serum for 48 h. The transfection efficiency
of the lipoplexes and polyplexes was expressed in femto-
gram (fg) of luciferase per mg of proteins. Cells treated with
naked DNA under equivalent conditions showed expression
levels of about 102–3 fg of luciferase per mg of proteins. The
cell viability of the lipoplexes and polyplexes was also
checked by determining the total protein amount per well of
the transfected cells relative to that measured for untreated
cells (for which the total protein amount per well is in a
30–60 µg per well range).
or PEI, fluorescently-labeled lipoplexes and polyplexes were
detected but only for a N/P ratio of 0.8 and 1.25. For higher
N/P ratios, a fluorescence trace was no more detected indi-
cating that partial to full ETB decomplexation from DNA has
occurred.
Analyses of the Acr-spacer-NLS-labeled and control
lipoplexes and polyplexes by gel electrophoresis with ETB
spreading (data not shown) indicated that fully complexed
DNA (not accessible to ETB intercalation) was detected for
N/P 2.5, 5 and 10, while “free” plasmid and partially com-
plexed plasmid (accessible to ETB intercalation) was detected
for N/P 1.25 and 0.8. Moreover, no electrophoretic migration
differences between the Acr-spacer-NLS-labeled lipoplexes
and polyplexes and their respective controls could be detected.
These results show that complexation of DNA with such low
amounts of Acr-spacer-NLS-conjugates, NLS, Acr, Phen or
ETB per plasmid prior to complexation with DOGS/DOPE
or PEI had no significant impact on DNA complexation nor
on its electrophoretic migration.
The gel electrophoresis assays performed in the presence
of excess sodium dodecylsulfate (SDS, data not shown)
showed that the polyplexes displayed a greater stability than
the lipoplexes with respect to SDS-induced dissociation.
Indeed, for N/P > 1.25, the plasmid remained still partially
complexed to PEI while fully-decomplexed plasmid was
detected for any of the DOGS/DOPE lipoplexes investigated
here.
4. Results and discussion
4.1. Lipoplex and polyplex formation and characterization
To ascertain the presence of the intercalating conjugates
within the plasmid in the lipoplexes and polyplexes, and more
particularly for high N/P ratios, we used the fluorescence prop-
erties of the Acr-spacer-NLS-conjugates. These conjugates
are fluorescent with a maximum of emission at 480–485 nm
whether in the absence or presence of the DOGS/DOPE lipo-
somes (Fig. 2A curve a), or of PEI (Fig. 2B curve a) but are
no more fluorescent when intercalated into DNA (Fig. 2A,
curve b). We found that upon adding the DOGS/DOPE lipo-
The gel electrophoresis assays performed on the Acr-
spacer-NLS/DNA (1:1, 5:1 and 10:1 mol) adducts and with-
out ETB spreading showed the absence of a fluorescence trace,
indicating that formation of these adducts cannot be ascer-
tained using this technique (data not shown). However, when
using ETB as intercalating agent, the ETB/DNA adducts were
detectable but only for a ETB/DNA ratio of 10:1. When these
ETB/DNA 10:1 adducts were complexed with DOGS/DOPE
Fig. 2. (A): Fluorescence emission spectrum of (i) Acr-C13-NLS alone or in the presence of DOGS/DOPE liposomes (curve a), (ii) Acr-C13-NLS/DNA adduct
or DNA alone (curve b), (iii) DOGS/DOPE/DNA N/P 5 lipoplexes (curve c(–)), (iv) DOGS/DOPE/[Acr-C13-NLS/DNA] N/P 5 lipoplexes, i.e. the lipids were
added to the Acr-C13-NLS/DNA adduct (curve c(+)), and (iv) Acr-C13-NLS/[DOGS/DOPE/DNA] N/P 5 lipoplexes immediately after addition of Acr-C13-
NLS to the lipoplexes (curve d, t = 0 min) and after 20 min of incubation (curve e, t = 20 min). (B): Fluorescence emission spectrum of (i) Acr-C13-NLS alone
or in the presence of PEI (curve a), (ii) PEI/DNA N/P 1.25, 5 and 10 lipoplexes (curves b(–), c(–) and d(–), respectively), (iv) PEI/[Acr-C13-NLS/DNA] N/P
1.25, 5 and 10 polyplexes, i.e. PEI was added to the Acr-C13-NLS/DNA adduct (curves b(+), c(+) and d(+), respectively). The Acr-C13-NLS/DNA molar ratio
is of 10:1. DNA is plasmid pTG11236. The concentration of Acr-C13-NLS is of 0.26 µM, that of DNA is of 15 µM phosphate. The emission spectra were
recorded after excitation at 409 nm. For more details, see materials and methods section.

C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
1299
somes to the preformed Acr-spacer-NLS/DNA 10:1 com-
plexes, the intercalating agent remained still complexed to
DNA for a DOGS/DNA N/P ratio of 5. Indeed, the fluores-
cence emission curves recorded for the DOGS/DOPE/[Acr-
spacer-NLS/DNA] and DOGS/DOPE/DNA lipoplexes were
very close (Fig. 2A, curve c(+) and c(–), respectively). The
full ejection of Acr-spacer-NLS from DNA upon complex-
ation by DOGS/DOPE would have resulted in a more sub-
stantial fluorescence emission increase at 480 nm with respect
to that recorded for the DOGS/DOPE/DNA lipoplexes. More-
over, we found that the addition of the Acr-spacer-NLS-
conjugates to the preformed N/P 5 DOGS/DOPE/DNA
lipoplexes (at a NLS/DNA molar ratio of 10:1) resulted in a
decrease of the fluorescence signal with time (Fig. 2A, curve
d for t = 0 vs. curve e for t = 20 min). This indicates that the
intercalating agent penetrates progressively into lipid-
complexed DNA which is still accessible at a N/P ratio of 5.
By contrast, partial to almost full ejection of Acr-spacer-
NLS from DNA was observed upon addition of PEI to the
preformed Acr-spacer-NLS/DNA adducts. As illustrated in
Fig. 2B, the progressive increase of the PEI/DNA N/P ratio
from 1.25 (curve b) to 5 (curve c) then to 10 (curve d), resulted
in a progressive increase of the fluorescence emission at
~485 nm [Fmax(PEI/[Acr-spacer-NLS/DNA])–Fmax(PEI/
DNA) for N/P 1.25 (arrow 1) < N/P 5 (arrow 2) < N/P 10
(arrow 3)], indicating the progressive release of Acr-spacer-
NLS from DNA. This release was almost complete for N/P
10 (arrow 3 ≤ Fmax(Acr-spacer-NLS) at 485 nm (curve a).
Altogether, these fluorescence experiments indicate that
part of the Acr-spacer-NLS-conjugates remains intercalated
within the plasmid for any of the N/P lipoplexes and poly-
plexes investigated except for the N/P 10 polyplexes.
Fig. 3
.
Means and 95.0% Tukey HSD intervals of the percentage of cell via-
bility for the DOGS/DOPE/[Acr-spacer-NLS/DNA] lipoplexes or PEI/[Acr-
spacer-NLS/DNA] polyplexes in A549 and NIH-3T3 cells, irrespectively of
(i) the spacer nature (C5, C7, C13 and C16), (ii) the Acr-spacer-NLS/DNA
molar ratio (1:1, 5:1 and 10:1), and (iii) the N/P ratio (0.8, 1.25, 2.5 and 5 for
the lipoplexes; 0.8, 1.25, 5 and 10 for the polyplexes) with respect to
control/DOGS/DOPE or control/PEI formulations (control = DNA,
Acr/DNA, Acr-spacer/DNA, Phen/DNA, ETB/DNA, NLS/DNA) and for a
DNA dose of 0.1 or 0.5 µg per well. For more details concerning the statis-
tical analysis, see materials and methods section.
mulations pointed out that theAcr-spacer-NLS-conjugates do
not improve luciferase expression neither in NIH-3T3 cells
(Fig. 4A, B) nor in A549 cells (Fig. 4C). Indeed, the detailed
analyses which took into account (i) the spacer length of the
Acr-spacer-NLS-conjugates, (ii) the N/P ratio (from 0.8 to 5)
of the formulation, (iii) theAcr-spacer-NLS/DNA molar ratio
(1, 5 or 10), and (iv) the DNA dose (0.5 or 0.1 µg per well),
showed that the Acr-spacer-NLS-labeled lipoplexes display
transfection efficiencies comparable to those of the control
and DOGS/DOPE reference lipoplexes (Figs. 4A, C for a
DNA dose of 0.5 µg per well; comparable tendencies were
found for a dose DNA dose of 0.1 µg per well, data not
shown). The presence of these conjugates was even found to
decrease luciferase expression in the NIH-3T3 cells when their
transfection is mediated with the labeled PEI-polyplexes
(Fig. 4B). Moreover, neither the length of the spacer between
the acridine intercalating and NLS units (Fig. 4A) nor the
Acr-spacer-NLS/DNA ratio, even for low N/P ratios of the
lipoplexes or polyplexes (see all cartoons throughout Fig. 4A–
C), were found to have a significant impact on transfection.
In terms of effect on lipoplex or polyplex mean size of the
Acr-spacer-NLS-conjugates, NLS peptide,Acr, Phen or ETB,
these additives had no significant impact. Indeed, lipoplexes
and polyplexes with mean particle sizes in the 70–130 nm
range for N/P ratios of 2.5, 5 and 10, and in the 100–200 nm
range for N/P 0.8 and 1.25 were observed whether they were
formulated or not with 1, 5 or 10 molecules of these additives
per plasmid.
4.2. In vitro transfection
The various statistical analyses of the cell viability and
transfection data obtained with the NIH and A549 cells are
illustrated in Figs. 3 and 4, respectively. It is noticeable that
the use of Acr-spacer-NLS-conjugates (as well as NLS, Acr,
Phen or ETB) for the formulation of the lipoplexes and poly-
plexes had no detectable cytotoxic effects on the NIH-
3T3 and A549 cell growth (Fig. 3) at the highest concen-
trations tested. Indeed, the 10:1 Acr-spacer-NLS/DNA,
Acr/DNA, Acr-spacer/DNA, Phen/DNA, ETB/DNA, or
NLS/DNA lipoplexes and polyplexes displayed a cell viabil-
ity that is identical to that of their respective reference
DOGS/DOPE lipoplexes and PEI polyplexes.
5. Conclusion
Importantly, the statistical analyses of the luciferase expres-
sion levels obtained with the different labeled and control for-
It appears, in line with several other studies performed with
NLS-conjugates not covalently linked to DNA [5,9,11–17],

1300
C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
Fig. 4
.
Means and 95.0% Tukey HSD intervals of the logarithmic transformation of transfection levels (Log10(fg luciferase per mg protein)) for the
DOGS/DOPE/[Acr-spacer-NLS/DNA] lipoplexes or PEI/[Acr-C16-NLS/DNA] polyplexes in NIH-3T3 (A and B, respectively) and A549 (C) cells, with res-
pect to (i) the spacer length, i.e. C5, C7, C13 and C16 (A), (ii) the Acr-spacer-NLS/DNA molar ratio, i.e. 1:1, 5:1 and 10:1 (A–C), and (iii) the N/P ratio (0.8,
1.25, 2.5 and 5 for the lipoplexes (A, C); 0.8, 1.25, 5 and 10 for the polyplexes (B)) as compared to control/DOGS/DOPE or control/PEI formulations
(control = DNA, Acr/DNA, Acr-spacer/DNA, Phen/DNA, ETB/DNA, NLS/DNA). The DNA dose was 0.5 µg per well. For more details concerning the statis-
tical analyses, see materials and methods section.
that the expression of the transgene is not improved upon com-
plexation of plasmidic DNA with NLS-intercalating conju-
gates prior to its formulation as lipoplexes or polyplexes.
These unsuccessful results might be due to (i) a too low con-
centration of NLS-labeled DNA (whether naked or partially
complexed as lipoplexes and polyplexes) that has been
released into the cytoplasm, (ii) dissociation of the NLS-
conjugates from DNA inside the cytoplasm, (iii) too strong
electrostatic interactions between the NLS signal and DNA,
thus hampering the recognition of the NLS by the nuclear
transport and import proteins, and/or (iii) to the globular plas-
mid used which owing to its size and topology does not per-
meate efficiently into the nucleus via the NPC. This later draw-
back could be circumvented by the use of linear double-
stranded DNAs which have shown enhanced gene expression
in the presence of NLS-PNA conjugates [9], or following liga-
tion of an oligonucleotide-NLS conjugate [4] or incorpora-
tion of a modified nucleotide for chemical linkage of the NLS
peptide [5].
6. Experimental protocols
Most of the reactions were performed in anhydrous sol-
vents under dry and oxygen-free nitrogen. Anhydrous sol-
vents were prepared by standard methods. The purification
by column chromatography were carried out using silica gel
60 (70–230 mesh) and chloroform (CHCl3), dichloromethane
(CH2Cl2), acetone, methanol (MeOH), N,N-dimethyl-
formamide (DMF) or mixtures thereof as indicated. Unless
noted otherwise, the ratios describing the composition of sol-
vent mixtures represent relative volume. Advancing of the
reaction was followed by thin layer chromatography (TLC)
on silica plates F254 or by HPLC. The following developing
systems were used: UV light, KMnO4, H2SO4/EtOH, Dragen-
dorff reagent, ninhydrin reagent. DOGS was synthesized in
our laboratory, as described elsewhere [19]. DOPE was pur-
chased from Sigma. The SV40 nuclear transport signal pep-
tide analogue CGYGPKKKRKVGG was from Bachem. 1H
and 13C spectra were recorded at 200 and 50.3 MHz, respec-

C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
1301
3
tively, on a Bruker AC-200. Chemical shifts were measured
relative to CHCl3 (d 7.26 ppm) or CH3OD (d 3.31 ppm) for
1H, relative to CDCl3 (d 77.1 ppm) for 13C and expressed
indirectly in relation to TMS. The atom numbering used for
the description of the acridine derivative 1H and 13C spectra
is shown in Scheme 1. The following abbreviations are used
to describe the signal multiplicities: bs (broad signal), s (sin-
glet), d (doublet), t (triplet), q (quadruplet), and m (multip-
let). Chemical shifts are expressed in parts per million (ppm)
and listed as follows: shift in ppm (multiplicity, coupling, inte-
gration and attribution). Electrospray ionization mass spec-
trometry (ESI-MS) in positive mode was performed on a
Finnigan MAT LCQ equipped with an atmospheric pressure
ionization (API) source. Mass spectrometry of the final mol-
ecules were performed on a MALDI-TOF applied Biosys-
temsVoyager DE-PRO. HPLC analyses were performed using
either a Merck L-6200 pump system equipped with
L-4000 UV detector (set at 280 nm), or on a Waters 600 pump
equipped with a Waters 996 photodiode array detector (PDA,
UV detector from 195 to 360 nm) and a column (250 × 4 mm)
packed with Lichrospher 100 RP-18 (5 µm). A gradient with
water (0.1% TFA) as solvent A and acetonitrile (0.1% TFA)
as solvent B was used with a flow = 1 ml/min.
2.55 (t, J = 5.9 Hz, 2H, CH2CO2H); 3.32–3.51 (m, 2H,
CH2N); 4.42–4.67 (m, 2H, CH2O); 5.10–5.36 (m, 2H,
H2C=CH); 5.38–5.57 (m, 1H, NH); 5.76–6.02 (m, 1H,
H2C=CH); 9.77–10.25 (bs, 1H, CO2H). 13C NMR (CDCl3):
d 34.3 (CH2CO2H); 36.4 (CH2N); 65.8 (CH2O); 117.9
(H2C=CH); 132.7 (H2C=CH); 156.5 (C(O)NH); 177.2
(CO2H).
6.1.1.3. Synthesis of [11-(5-N-(tert-butoxycarbonyl)amino-
pentylcarbamoyl)-undecyl]-carbamic acid allyl ester 2a. To
a solution of 375 mg (1.25 mmol) of 1a in 15 ml DMF and
0.52 ml (3.76 mmol) of triethylamine stirred at 0 °C, 203 mg
(1.50 mmol) of 1-hydroxybenzotriazole (HOBt), 0.31 ml
(1.50 mmol) of N-Boc-1,5-diaminopentane, and 288 mg
(1.50 mmol) of N-ethyl-N′-(3-dimethylaminopropyl)-
carbodiimide (EDC) were added. After 30 min at 0 °C, the
solution was stirred at room temperature during 1 day. The
solvents were evaporated and the residue obtained was dis-
solved in CHCl3. The organic phase was washed succes-
sively with a 5% citric acid solution, a 5% Na2CO3solution,
and water, then dried over Na2SO4 and evaporated, leading
to 2a (521 mg; 1.08 mmol, 86%) as a white powder. TLC
(CHCl3/MeOH: 97:3: v/v; ninhydrin, KMnO4) Rf = 0.45.
1H NMR (CDCl3): d 1.11–1.69 (m, 33H, (CH3)3,
(CH2)3CH2NHBoc, N–CH2–(CH2)9); 2.11 (t, 3J = 7.5 Hz, 2H,
CH2C(O)); 2.98–3.28 (m, 6H, NCH2); 4.45–4.59 (m, 2H,
CH2O); 4.59–4.75 and 4.75–4.97 (m, 2H, HN carbamic);
5.10–5.35 (m, 2H, H2C=CH); 5.67–6.01 (m, 2H,
H2C=CH and NH amide). 13C NMR (CDCl3): d 24.0
(CH2(CH2)2NHBoc); 25.9 (NH(CH2)9CH2); 26.8
(AllocNH(CH2)2CH2); 28.5 (C(CH3)3); 29.3, 29.4,
29.5, 29.5, 29.8, 30.0 (NHCH2CH2CH2(CH2)6 and
CH2CH2CH2CH2NHBoc); 36.9 (NH(CH2)10CH2); 39.3, 40.3
(CH2(CH2)3CH2NHBoc); 41.1 (AllocNHCH2); 65.4 (CH2O);
79.1 (C(CH3)3); 117.5 (H2C=CH); 133.1 (H2C=CH); 156.2,
156.4 (C(O)Boc and C(O)Alloc); 173.3 ((CH2C(O)).
6.1. Chemistry
6.1.1. Synthesis of the BocNH-spacer-NH2 spacers 3a–c
6.1.1.1. Synthesis of 12-N-(allyloxycarbonyl)aminodode-
canoic acid 1a. A solution of 0.59 ml (5.6 mmol) of allyl-
chloroformate in 5 ml dioxane was added at 0 °C to a solu-
tion of 12-aminododecanoic acid (1.0 g, 4.6 mmol) in 20 ml
THF and 20 ml NaOH (1 M). The resulting mixture was
stirred during 1 h at 0 °C then at room temperature for 2 h.
The solvents were evaporated in a Rotavap evaporation sys-
tem. The residue was dissolved in a 5% citric acid solution.
The resulting aqueous phase was extracted with CHCl3. The
organic phase was then washed with water, dried over Na2SO4
and evaporated, leading to 751 mg (2.51 mmol, 54%) of 1a
as a white powder. TLC (CHCl3/MeOH: 9:1: v/v; ninhydrin,
KMnO4, H2SO4) Rf = 0.62. 1H NMR (CDCl3): d 1.12–1.38
(m, 14H, (CH2)7(CH2)2CO2H); 1.38–1.71 (m, 4H, NCH2CH2
6.1.1.4. Synthesis of [11-(2-N-(tert-butoxycarbonyl)amino-
ethylcarbamoyl)-undecyl]-carbamic acid allyl ester 2b. The
procedure described for the synthesis of 2a when applied to
1a (375 mg, 1.25 mmol) and N-Boc-1,2-diaminoethane
(0.24 ml, 1.50 mmol) afforded, after purification by liquid–
liquid extraction (CHCl3–H2O), 415 mg (0.94 mmol, 75%)
of 2b as a white powder. TLC (CHCl3/MeOH: 97:3: v/v; nin-
hydrin, KMnO4) Rf = 0.48. 1H NMR (CDCl3): d 1.04–1.70
3
and CH2CH2CO2H); 2.31 (t, J = 7.4 Hz, 2H, CH2CO2H);
3.04–3.23 (m, 2H, NCH2); 4.47–4.65 (m, 2H, CH2O); 4.72–
4.94 (m, 1H, NH); 5.11–5.38 (m, 2H, H2C=CH); 5.78–6.10
(m, 1H, H2C=CH); 10.80–11.48 (bs, 1H, CO2H). 13C NMR
(CDCl3): d 24.8 (CH2CH2CO2H); 26.8 (N(CH2)2CH2); 29.1,
29.3, 29.5, 30.0 (NCH2CH2CH2(CH2)6); 34.2 (CH2CO2H);
41.2 (NCH2); 65.5 (CH2O); 117.7 (H2C=CH); 133.2
(H2C=CH); 156.4 C(O)NH); 179.6 (CO2H).
3
(m, 27H, NHCH2(CH2)9 and (CH3)3); 2.13 (t, J = 7.6 Hz,
2H, CH2C(O)); 3.04–3.39 (m, 6H, C(O)NHCH2); 4.46–4.62
(m, 2H, CH2O); 4.71–4.94 (m, 1H, HN carbamic); 5.03–5.35
(m, 3H, H2C=CH and NH carbamic); 5.76–6.01 (m,
1H,H2C=CH); 6.31–6.48 (m, 1H, NH amide). 13C NMR
(CDCl3): d 25.8 (NH(CH2)9CH2); 26.8 (NH(CH2)2CH2); 28.5
(C(CH3)3); 29.3, 29.3, 29.5, 29.5 (NH(CH2)3(CH2)6); 30.0
(AllocNHCH2CH2); 36.8 (NH(CH2)10CH2); 40.4, 40.7
(HN(CH2)2NH); 41.1 (AllocNHCH2); 65.5 (CH2O); 79.6
(C(CH3)3); 117.6 (H2C=CH); 133.1 (H2C=CH); 156.4, 157.0
(C(O)Boc and C(O)Alloc); 174.1 (CH2C(O)).
6.1.1.2. Synthesis of 3-N-(allyloxycarbonyl)amino-propio-
nic acid 1b. The procedure described for the synthesis of 1a
when applied to b-alanine (500 mg, 5.6 mmol) led to 865 mg
(5.0 mmol, 52%) of 1b as a colorless oil. TLC (CHCl3/MeOH:
9:1: v/v; ninhydrin, KMnO4) Rf = 0.54. 1H NMR (CDCl3): d

1302
C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
6.1.1.5. Synthesis of [5-(3-N-(allyloxycarbonyl)aminopro-
pionylamino)-pentyl]-carbamic acid tert-butyl ester 2c. The
procedure described for the synthesis of 2a when applied to
400 mg (2.31 mmol) of 1b and 0.58 ml (2.77 mmol) of N-Boc-
1,5-diaminopentane, afforded 719 mg (2.01 mmol, 87%) of
2c as a white powder. TLC (CHCl3/MeOH: 9:1: v/v; ninhy-
drin, KMnO4) Rf = 0.46 . 1H NMR (CDCl3): d 1.18–1.56 (m,
CH2C(O)); 2.45 (t, 3J = 6.9 Hz, 2H, H2NCH2); 2.89–3.13 (m,
4H, HN(CH2)2NH). 13C NMR (CDCl3/CD3OD): d 25.6
(H2N(CH2)9CH2); 26.6 (H2N(CH2)2CH2); 28.0 (C(CH3)3);
29.0, 29.1, 29.2, 29.2, 29.3 (H2N(CH2)3(CH2)6); 32.5
(H2NCH2CH2); 36.2 (H2N(CH2)10CH2); 39.5, 39.7
(HN(CH2)2NH); 41.3 (H2NCH2); 79.3 (C(CH3)3); 157.1
(C(O)Boc); 175.0 (C(O)NH).
3
15H, (CH3)3 and (CH2)3CH2NH); 2.36 (t, J = 6.0 Hz, 2H,
CH2C(O)); 2.96–3.12 (m, 2H, AllocNHCH2); 3.12–3.27 (m,
2H, CH2NHBoc); 3.34–3.49 (m, 2H, CH2C(O)NHCH2),
4.44–4.54 (m, 2H, H2CO); 4.69–4.83 (m, 1H, NH carbamic);
5.09–5.32 (m, 2H, H2C=CH); 5.58-5.73 (m, 1H, NH car-
bamic); 5.73–5.96 (m, 1H, H2C=CH); 6.16–6.35 (m, 1H, NH
amide). 13C NMR (CDCl3): d 28.5 (C(CH3)3); 29.1, 29.6
((CH2)2CH2C(O)); 36.0 (AllocNHCH2CH2); 37.2
(CH2C(O)NH); 39.2 (C(O)NHCH2); 40.3 (AllocNHCH2);
65.5 (CH2NHBoc); 79.2 (C(CH3)3); 117.5 (H2C=CH); 132.9
(H2C=CH); 156.3, 156.5 (OC(O)NH); 171.4 (C(O)NH).
6.1.1.8. Synthesis of [5-(3-amino-propionylamino)-pentyl]-
carbamic acid tert-butyl ester 3c. The Alloc-deprotection
procedure when applied to 719 mg (2.01 mmol) of 2c
afforded, after chromatography over a silica gel column (90 g,
CHCl3/MeOH: from 95:5 to 80:20: v/v) and steric exclusion
chromatography over Sephadex LH-20 gel column (7 g,
MeOH), 198 mg (0.72 mmol, 36%) of 3c as a white powder.
TLC (CHCl3/MeOH: 95:5: v/v; ninhydrin, KMnO4)
Rf = 0.63. 1H NMR (CDCl3): d 1.00–1.59 (m, 17H,
3
(CH2)3CH2NH, C(CH3)3, NH2); 2.14 (t, J = 5.9 Hz, 2H,
CH2C(O)); 2.64–3.15 (m, 6H, NCH2); 5.00–5.14 (m, 1H, NH
carbamic); 7.33–7.49 (m, 1H, NH amide). 13C NMR (CDCl3):
d 23.8 (HN(CH2)2CH2); 28.2, 29.0 (HN–CH2–CH2–CH2–
CH2); 29.4 (C(CH3)3); 38.2 (H2NCH2); 38.5 (H2NCH2CH2);
38.8 (C(O)NHCH2); 40.1 (CH2NHBoc); 78.6 (C(CH3)3);
156.0 (C(O)Boc); 172.3 (C(O)NH).
6.1.1.6. Synthesis of [5-(12-N-(aminododecanoylamino)-
pentyl]-carbamic acid tert-butyl ester, 3a (Alloc-deprotection
procedure). To a solution of 521 mg (1.08 mmol) of 2a in
20 ml CH2Cl2 and 1.7 ml (16.2 mmol) of diethylamine,
125 mg (0.11 mmol) of tetrakis(triphenylphosphine)palladium
(Pd(PPh3)4) were added. The mixture was stirred during 2 h
at room temperature. After evaporation of the solvents, the
residue was dissolved in CHCl3 and washed with water. The
organic phase was dried over Na2SO4 and evaporated. Chro-
matography of the residue over a silica gel column (40 g,
CHCl3/MeOH: from 95:5 to 80:20: v/v) led to an orange pow-
der. Steric exclusion chromatography over Sephadex LH-
20 gel column (7 g, MeOH) led to 172 mg (0.43 mmol, 40%)
of 3a as a white powder. TLC (CHCl3/MeOH: 9:1: v/v; nin-
6.1.2. Synthesis of the acridine-maleimide conjugates 7a–d
6.1.2.1. Synthesis of 4-(2,5-dioxo-2,5-dihydro-pyrrol-1-
ylmethyl)-cyclohexanecarboxylic acid {5-[12-(6-chloro-2-
methoxy-acridin-9-ylamino)-dodecanoylamino]-pentyl}-
amide (7a). Synthesis of 5a (General condensation
procedure A). 3a (172 mg, 0.43 mmol) was added to a solu-
tion of 2-methoxy-6,9-dichloroacridine 4 (60 mg, 0.21 mmol)
in phenol (2.5 g, 26.6 mmol) and 4-methylmorpholine (NMM)
(0.07 ml, 0.65 mmol) heated at 100 °C. The mixture was
stirred at 100 °C during 2 h, then phenol was distilled under
reduce pressure. Chromatography over a silica gel column
(15 g, CH2Cl2/acetone: from 100:0 to 70:30: v/v), of the brown
oily residue led to 113 mg (0.18 mmol, 82%) of 5a as a dark
yellow oil. TLC (CH2Cl2/MeOH: 9:1: v/v; ninhydrin, UV)
Rf = 0.49. 1H NMR (CDCl3): d 1.03–1.80 (m, 33H, (CH3)3,
(CH2)3CH2NHBoc,AcrNHCH2(CH2)9); 2.09 (t, 3J = 7.6 Hz,
2H, CH2C(O)); 2.96–3.10 (m, 2H, CH2NHBoc); 3.10–3.25
1
hydrin, KMnO4) Rf = 0.46. H NMR (CDCl3/CD3OD,
50:50): d 0.90–1.65 (m, 33H, (CH2)3CH2NHBoc,
3
H2NCH2(CH2)9); 1.90 (t, J = 7.5 Hz, 2H, CH2C(O)); 2.40
(t, 3J = 7.1 Hz, 2H, H2NCH2); 2.78 (t,3J = 6.7 Hz, 2H, C(O)N-
3
HCH2); 2.92 (t, J = 6.8 Hz, 2H, CH2NHBoc). 13C NMR
(CDCl3/CD3OD, 50:50): d 23.7 (CH2(CH2)2NHBoc); 25.7
(H2N(CH2)9CH2); 26.6 (H2N(CH2)2CH2); 28.0 (C(CH3)3);
28.6, 29.0, 29.1, 29.1, 29.2, 29.3 (H2N(CH2)3(CH2)6 and
CH2CH2CH2CH2NHBoc); 32.3 (H2NCH2CH2); 36.1
(H2N(CH2)10CH2); 38.9, 39.9 (CH2(CH2)3CH2NHBoc); 41.2
(H2NCH2); 78.9 (C(CH3)3); 156.7 (C(O)Boc); 174.6
(C(O)NH).
3
(m, 2H, C(O)NHCH2); 3.63 (t, J = 7.1 Hz, 2H, AcrN-
HCH2); 3.84–3.94 (bs, 3H, OCH3); 4.63–4.82 (m, 1H,
NHBoc); 5.87–6.03 (m, 1H, NH amide); 7.14–7.39 (m and
3
3
6.1.1.7. Synthesis of [2-(12-amino-dodecanoylamino)-ethyl]-
carbamic acid tert-butyl ester 3b. The Alloc-deprotection
procedure when applied to 415 mg (0.94 mmol) of 2b
afforded, after chromatography over a silica gel column (30 g,
CHCl3/MeOH: from 95:5 to 80:20: v/v) and steric exclusion
chromatography over Sephadex LH-20 gel column (7 g,
MeOH), 235 mg (0.66 mmol, 70%) of 3b as a white powder.
TLC (CHCl3/MeOH: 9:1: v/v; ninhydrin, KMnO4) Rf = 0.50.
1H NMR (CDCl3/CD3OD): d 0.97–1.53 (m, 27H,
1dd, J = 9.4 Hz, J = 2.5 Hz, 3H, AcrH1, AcrH3, AcrH7);
7.84–8.01 (m, 3H,AcrH4,AcrH5,AcrH8). 13C NMR (CDCl3):
d 24.0 (CH2(CH2)2NHBoc); 25.8 (AcrNH(CH2)9CH2); 26.9
(AcrNH(CH2)2CH2); 28.5 (C(CH3)3); 29.3, 29.4, 29.4, 29.9
(AcrNH(CH2)3(CH2)6, CH2CH2CH2CH2NHBoc); 31.7
(AcrNHCH2CH2); 36.8 (AcrNH(CH2)10–CH2); 39.3, 40.3
(CH2(CH2)3CH2NHBoc); 50.6 (AcrNHCH2); 55.5 (OCH3);
79.1 (C(CH3)3); 99.4 (AcrC1); 115.5, 117.7 (AcrCa, AcrCd);
124.2, 124.4, 124.5, 127.8, 131.0 (AcrC3, AcrC4, AcrC5,
AcrC7, AcrC8); 134.8, 146.4, 148.2 (AcrC6, AcrCb, AcrCc);
3
H2NCH2(CH2)9 and (CH3)3); 1.96 (t, J = 7.5 Hz, 2H,

C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
1303
150.1, 155.9 (AcrC2, AcrC9); 156.2 (C(O)Boc); 173.3
(C(O)NH).
6.1.2.2. Synthesis of 4-(2,5-dioxo-2,5-dihydro-pyrrol-1-
ylmethyl)-cyclohexanecarboxylic acid {2-[12-(6-chloro-2-
methoxy-acridin-9-ylamino)-dodecanoylamino]-ethyl}-
amide 7b. Synthesis of 5b. The above described condensation
procedure A when applied to 3b (235 mg, 0.66 mmol) and 4
(91 mg, 0.33 mmol) afforded, after chromatography over a
silica gel column (20 g, CH2Cl2/acetone: from 100:0 to 70:30:
v/v), 150 mg (0.25 mmol, 76%) of 5b as a dark yellow pow-
der. TLC (CH2Cl2/MeOH: 9:1: v/v; ninhydrin, UV) Rf = 0.50.
1H NMR (CDCl3): d 1.00–1.83 (m, 27H, AcrNHCH2(CH2)9
and (CH3)3); 2.10 (t, 3J = 7.6 Hz, 2H, CH2C(O)); 3.12–3.40
Synthesis of 6a (Boc-deprotection procedure). The Boc-
deprotection was quantitatively achieved by dissolving
113 mg of 5a (0.18 mmol) in CH2Cl2 with a large excess of
TFA during 1 h at room temperature. The excess TFA was
removed by co-evaporation with cyclohexan leading to 6a
(162 mg, 0.18 mmol). 1H NMR (CD3OD): d 1.09–1.80 (m,
+
22H, (CH2)3CH2NH3 , AcrNHCH2CH2(CH2)8); 1.83–2.05
(m, 2H,AcrNHCH2CH2); 2.16 (t, 3J = 7.5 Hz, 2H, CH2C(O));
+
2.92 (t, 3J = 7.5 Hz, 2H, CH2NH3 ); 3.18 (t, 3J = 6.8 Hz, 2H,
C(O)NHCH2); 3.89–3.99 (bs, 3H, OCH3); 4.05 (t, 3J = 7.5 Hz,
(m, 4H, HN(CH2)2NH); 3.63 (t, J = 7.1 Hz, 2H, AcrN-
3
3
3
2H, AcrNHCH2); 7.39 (dd, J = 9.3 Hz, J = 2.0 Hz, 1H,
HCH2); 3.81–3.97 (bs, 3H, OCH3); signals for AcrH as for
5a. 13C NMR (CDCl3): d 25.7 (AcrNH(CH2)9CH2); 26.9
(AcrNH(CH2)2CH2); 28.4 (C(CH3)3); 29.3, 29.4, 29.4
(AcrNH(CH2)3(CH2)6); 31.7 (AcrNHCH2CH2); 36.7
3
3
AcrH1); 7.38 and 7.52 (2dd, J = 9.3 Hz, J = 2.0 Hz, 2H,
AcrH3, AcrH7); 7.60–7.73 (m, 3H, AcrH5, AcrH4, AcrH1);
8.32 (d, 3J = 9.5 Hz, 1H, AcrH8). 13C NMR (CD3OD): d 24.7
(CH2(CH2)2NH3+); 27.1 (AcrNH(CH2)9CH2); 27.8
(AcrNH(CH2)2CH2); 28.1, 29.9, 30.2, 30.3, 30.4, 30.5
(AcrNH(CH2)3(CH2)6 and CH2CH2CH2CH2NHBoc); 30.7
(AcrNHCH2CH2); 37.1 (AcrNH(CH2)10CH2); 39.9, 40.5
(AcrNH(CH2)10CH2);
40.4
(CH2NHBoc);
40.7
(CH2CH2NHBoc); 50.6 (AcrNHCH2); 55.6 (OCH3); 79.6
(C(CH3)3); signals for AcrC and C(O) as for 5a.
Synthesis of 6b. The Boc-deprotection procedure when
applied to 150 mg of 5b (0.25 mmol) led quantitatively to 6b
(213 mg; 0.25 mmol): 1H NMR (CD3OD): d 1.04–1.71 (m,
16H, AcrNHCH2CH2(CH2)8); 1.80–2.04 (m, 2H,
+
(CH2(CH2)3CH2NH3 ); 50.4 (AcrNHCH2); 56.6 (OCH3);
103.8 (AcrC1); 111.2, 115.3 (AcrCa, AcrCd); 118.5, 121.4,
125.0, 128.6, 129.1 (AcrC3, AcrC4, AcrC5H, AcrC7, AcrC8);
135.7, 141.3, 141.7 (AcrC6, AcrCb, AcrCc); 157.6, 158.1
(AcrC2, AcrC9); 162.4 (q, CF3CO2H); 173.3 (C(O)–NH).
Synthesis of 7a (general condensation procedure B).
To a solution of 6a (185 mg, 0.21 mmol) and 0.12 ml NEt3
(0.84 mmol) in 10 ml DMF, 77 mg (0.23 mmol) of 4-(N-
maleimidomethyl)cyclohexane-carboxylic acid N-hydro-
xysuccinimide ester (SMCC) were added. The resulting mix-
ture was stirred overnight at room temperature, then the
solvents were evaporated under reduced pressure. Chroma-
tography of the residue over a silica gel column (15 g,
CH2Cl2/acetone: from 100:0 to 70:30: v/v) led to 130 mg
(0.17 mmol, 81%) of 7a. TLC (CH2Cl2/MeOH: 9:1: v/v; nin-
3
AcrNHCH2CH2); 2.21 (t, J = 7.6 Hz, 2H, CH2C(O)); 3.05
3
+
3
(t, J = 5.9 Hz, 2H, CH2NH3 ); 3.45 (t, J = 6.0 Hz, 2H,
C(O)NHCH2); 3.82–4.09 (m, 5H, OCH3 and AcrNHCH2);
signals for AcrH as for 6a. 13C NMR (CD3OD): d 26.7
(AcrNH(CH2)9CH2); 27.8 (AcrNH(CH2)2CH2); 30.2, 30.3,
30.4, 30.5 (AcrNH(CH2)3(CH2)6); 30.7 (AcrNHCH2CH2);
36.9 (AcrNH(CH2)10CH2); 38.1 (CH2NH3+); 40.7
+
(CH2CH2NH3 ); 50.3 (AcrNHCH2); 56.6 (OCH3); signals for
AcrC and C(O) as for 6a.
Synthesis of 7b. The condensation procedure B when
applied to 6b (135 mg, 0.27 mmol) led to 163 mg (0.23 mmol,
84%) of 7b. TLC (CH2Cl2/MeOH: 9:1: v/v; ninhydrin, UV)
Rf = 0.44. 1H NMR (CDCl3): d 0.70–2.27 (m, 30H,
AcrNHCH2(CH2)10 and cyclohexyl); 3.15–3.51 (m, 6H,
HN(CH2)2NH and CHCH2N); 3.66–4.00 (m, 4H, AcrN-
HCH2 and OCH3); 6.48–6.82 (m, 3H, CH=CH and NH
amide); signals for AcrH and H(maleimido) as for 7a. 13C
1
hydrin, UV) Rf = 0.48. H NMR (CDCl3): d 0.71–2.32 (m,
36H, (CH2)3CH2NHSMCC, AcrNHCH2(CH2)10, and cyclo-
hexyl); 2.99–3.42 (m, 6H, C(O)NHCH2 and CHCH2N); 3.64–
3.90 (m, 5H, AcrNHCH2 and OCH3); 6.03–6.34 (m, 2H, NH
amide); 6.62–6.73 (bs, 2H, CH=CH); 6.92–7.12 (m, 2H,
AcrH7, AcrH3); 7.34–7.52 (m, 2H, AcrH1, AcrH8); 7.56–
NMR (CDCl3):
d 25.8 (AcrNH(CH2)9CH2); 26.8
7.68 (m, 1H, AcrH5); 7.94 (d, 3J = 9.5 Hz, 1H, AcrH4). 13
C
(AcrNH(CH2)2CH2); 28.8, 29.1, 29.2, 29.2, 29.3, 29.8, 30.2
(AcrNHCH2CH2CH2(CH2)6 and CH2(cyclohexyl)); 36.4
(CHCH2N); 36.6 (AcrNH(CH2)10-CH2); 39.9, 40.2
(HN(CH2)2NH); 43.7 (CH2NC(O)); 45.0 (C(O)CH); 49.2
(AcrNHCH2); 55.8 (OCH3); signals for AcrC, C(O) and
C(maleimido) as for 7a. MS (ESI+): m/z = 718.7 in agree-
ment with the calculated mass for [M]+ = C40H52ClN5O5.
NMR (CDCl3): d 23.8 (CH2(CH2)2NH–C(O)); 25.9
(AcrNH(CH2)9CH2); 26.9 (AcrNH(CH2)2CH2); 28.9, 29.0,
29.1, 29.3, 29.4, 29.8, 30.1 (AcrNHCH2CH2CH2(CH2)6,
CH2CH2CH2CH2NHC(O) and CH2(cyclohexyl)); 36.4
(CHCH2N); 36.8 (AcrNH(CH2)10CH2); 39.0, 39.1
(CH2(CH2)3CH2NH); 43.7 (CH2NC(O)); 45.1 (C(O)CH);
49.0 (AcrNHCH2); 55.8 (OCH3); 102.5 (AcrC1); 109.8, 113.9
(AcrCd, AcrCa); 118.2, 120.0, 120.5, 123.6, 126.7,
127.1 (AcrC3, AcrC4, AcrC5, AcrC6, AcrC7, AcrC8);
134.1 (CH=CH); 134.5, 140.0 (AcrCb, AcrCc); 154.9, 156.0
(AcrC9, AcrC2); 171.1 (C(O)–CH=CH–C(O)); 173.8
(C(O)NH(CH2)5); 176.1 (C(O)NHCH). MS (ESI+):
m/z = 760.7 in agreement with the calculated mass for
[M]+ = C43H58ClN5O5.
6.1.2.3. Synthesis of 4-(2,5-dioxo-2,5-dihydro-pyrrol-1-
ylmethyl)-cyclohexanecarboxylic acid {5-[3-(6-chloro-2-
methoxy-acridin-9-ylamino)-propionylamino]-pentyl}-
amide 7c. Synthesis of 5c. The above described condensation
procedure A when applied to 3c (198 mg, 0.66 mmol) and 4
(101 mg, 0.36 mmol) afforded, after chromatography over a
silica gel column (30 g, CH2Cl2/acetone: from 100:0 to 70:30:

1304
C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
v/v), 148 mg (0.29 mmol, 79%) of 5c as a dark yellow pow-
der. TLC (CH2Cl2/MeOH: 9:1: v/v; ninhydrin, UV) Rf = 0.62.
1H NMR (CDCl3): d 1.17–1.59 (m, 15H, (CH2)3CH2NHBoc,
C(CH3)3); 2.47 (t, 3J = 5.6 Hz, 2H, CH2C(O)); 2.98–3.15 (m,
2H, CH2NHBoc); 3.18–3.33 (m, 2H, CH2–C(O)NHCH2);
3.84–4.02 (m, 5H, OCH3, AcrNHCH2); 4.56–4.73 and 6.23–
6.39 (m, 2H, AcrNH, NHBoc); 7.17–7.28 (m, 2H, AcrH1 and
AcrH3); 7.30–7.39 (m, 2H, AcrH7, NH amide); 7.88 (d,
3J = 10.2 Hz, 1H, AcrH8); 7.96 (d, 3J = 1.8 Hz, 1H, AcrH4);
8.04 (d, 3J = 9.2 Hz, 1H, AcrH5). 13C NMR (CDCl3): d 24.3
(CH2(CH2)2NHBoc); 28.8 (C(CH3)3); 29.3, 30.2
(CH2CH2CH2CH2CH2); 36.1 (CH2C(O)); 39.8, 40.2
(CH2(CH2)3CH2NHBoc); 47.0 (AcrNHCH2); 56.1 (OCH3);
79.6 (C(CH3)3); signals for AcrC and C(O) as for 6a.
4.90–5.16 (m, 2H, NH); signals forAcrH as for 5a. 13C NMR
(CDCl3): d 24.1 (NH–(CH2)2–CH2); 28.5 (CH3(Boc)); 29.9,
31.2 (NHCH2CH2CH2CH2); 40.3 (CH2NHBoc); 50.4 (AcrN-
HCH2); 55.6 (OCH3); 79.2 (C(Boc)); 156.2 (C(O)Boc); sig-
nals for AcrC as for 5a.
Synthesis of 6d. The Boc-deprotection procedure applied
to 170 mg of 5d (0.38 mmol) led to 6d (263 mg, 0.38 mmol).
1H NMR (CD3OD): d 1.43–1.67 (m, 2H, NH(CH2)2CH2);
1.67–1.87 (m, 2H, AcrNHCH2CH2); 1.87–2.13 (m, 2H,
+
CH2CH2NH3 ); 2.84 (t, 3J = 7.3 Hz, 2H, AcrNHCH2); 3.86
3
+
(s, 3H, OCH3); 4.03 (t, J = 7.1 Hz, 2H, CH2NH3 ); 7.20–
7.63 (m, 5H, AcrH1, AcrH3, AcrH5, AcrH7 and AcrH8); 8.19
(d, J = 9.3 Hz, 1H, AcrH4). 13C NMR (CD3OD): d 24.8
3
(NH(CH2)2CH2); 28.2 and 30.2 (NHCH2CH2CH2CH2); 40.5
+
Synthesis of 6c. The Boc-deprotection procedure applied
to 148 mg of 5c (0.29 mmol) led quantitatively to 6c (218 mg,
0.29 mmol). 1H NMR (CDCl3/CD3OD): d 1.01–1.55 (m, 8H,
CH2C(O) and (CH2)3CH2NH3+3); 2.56–2.79 (m, 4H,
CH2(CH2)3CH2NH3+); 3.01 (t, J = 6.3 Hz, 2H, AcrN-
(CH2NH3 ); 49.9 (AcrNHCH2); 56.6 (OCH3); signals for
AcrC as for 6a.
Synthesis of 7d. The condensation procedure B when
applied to 6d (204 mg, 0.27 mmol) led to 133 mg (0.21 mmol,
88%) of 7d. TLC (CH2Cl2/MeOH: 9:1: v/v; ninhydrin, UV)
Rf = 0.52. 1H NMR (CDCl3): d 0.73–2.07 (m, 16H,
NHCH2(CH2)3, CH2(cyclohexyl) and CHCH2); 3.00–3.39 (m,
4H, CH2NHC(O) and CH2N(C(O))2); 3.49–3.73 (m, 2H,
AcrNHCH2); 3.87 (s, 3H, OCH3); 5.76–5.94 (m, 1H, NH);
6.63 (s, 2H, CH=CH); 7.04–7.40 (m, 4H, AcrH1, AcrH3,
AcrH7 and NH); 7.76–8.03 (m, 3H,AcrH4,AcrH5 andAcrH8).
13C NMR (CDCl3): d 24.2 (NH(CH2)2CH2); 28.9 (C(O-
)CHCH2); 29.5 (CH2CH2NHC(O)); 29.8 (NCH2CHCH2);
31.1 (AcrNHCH2CH2); 36.3 (CHCH2N); 38.9 (C(O)CH);
43.7 (CH2NHC(O)); 45.2 (CH2N(C(O))2); 50.4 (AcrN-
HCH2); 55.6 (OCH3); signals for AcrC and C(maleimido) as
for 7a. MS (ESI+): m/z = 563.5 in agreement with the calcu-
lated mass for [M]+ = C31H35ClN4O4.
HCH2); 3.82 (s, 3H, OCH3); signals for AcrH as for 6a. 13
C
+
NMR (CDCl3/CD3OD): d 23.2 (CH2(CH2)2NH3 ); 26.5, 28.0
+
+
(CH2CH2CH2CH2NH3 ); 34.4 (CH2C(O)); 38.7 (CH2NH3 );
39.1 (C(O)NHCH2); 46.2 (AcrNHCH2); 55.7 (OCH3); sig-
nals for AcrC and C(O) as for 6a.
Synthesis of 7c. The condensation procedure B when
applied to 6c (204 mg, 0.27 mmol) led to 133 mg (0.21 mmol,
78%) of 7c. TLC (CH2Cl2/MeOH: 9:1: v/v; ninhydrin, UV)
Rf = 0.47. RMN 1H (CDCl3/CD3OD): d 0.71–2.03 (m, 16H,
3
(CH2)3CH2NH and cyclohexyl); 2.67 (t, J = 5.9 Hz, 2H,
CH2C(O)); 2.88–3.08 (m, 4H, C(O)NHCH2); 3.25 (d,
2J = 7.1 Hz, 2H, CH2N(C(O))2); 3.84 (s, 3H, OCH3); 4.18 (t,
3J = 5.9 Hz, 2H, AcrNHCH2); 6.61 (s, 2H, CH=CH);
6.72–6.95 (m, 1H, NH amide); 7.17–7.25 (m, 1H, AcrH3);
7.30–7.40 (m, 1H, AcrH7); 7.49–7.68 (m, 3H, AcrH1, AcrH5
3
and AcrH8); 8.15 (d, J = 9.3 Hz, 1H, AcrH4). 13C NMR
6.1.3. Synthesis of the acridine-spacer-NLS-conjugates
Acr-spacer-NLS
(CDCl3/CD3OD): d 23.8 (CH2(CH2)2NHCOCH); 28.4, 28.6,
29.6, 34.6 (CH2(cyclohexyl), CH2CH2NH); 36.2 (CHCH2N);
38.7, 39.1 (C(O)NHCH2); 43.5 (CH2N(C(O))2); 44.8
(C(O)CH); 46.3 (AcrNHCH2); 55.7 (OCH3); signals forAcrC,
C(O) and C(maleimido) as for 7a. MS (ESI+): m/z = 634.6
in agreement with the calculated mass for [M]+ =
C34H40ClN5O5.
To a solution of the NLS SV40 peptide (585 µg, 0.21 µmol;
M = 1377 g/mol as checked by MALDI-TOF MS) in 64.1 µl
phosphate buffer (pH 7, 20 mM), 3 µl of HCl 0.1 N and 4 µl
of a 0.53 mM solution of TCEP were added. The mixture
was stirred overnight at room temperature. Then, 22.5 µl of
DMF and 6.7 µl of a solution containing 1.70 µmol of 7a–d
in DMF were added. The reaction was monitored by HPLC.
After completion of the reaction (about 15 min at room tem-
perature), the solvents were evaporated. The resulting Acr-
spacer-NLS compounds were purified by reverse-phase semi-
preparative HPLC (A/B 90:10 to 0:100 over 35 min).
6.1.2.4. Synthesis of 4-(2,5-dioxo-2,5-dihydro-pyrrol-1-
ylmethyl)-cyclohexanecarboxylic acid [5-(6-chloro-2-
methoxy-acridin-9-ylamino)-pentyl]-amide 7d. Synthesis of
5d. The condensation procedure A described above when
applied to 0.22 ml (1.08 mmol) of N-Boc-1,5-diaminopentane
and 150 mg (0.54 mmol) of 4 afforded, after chromatography
over a silica gel column (30 g, CH2Cl2/acetone: from 100:0 to
70:30: v/v), 170 mg (0.38 mmol, 71%) of 5d as a dark yellow
powder. TLC (CH2Cl2/MeOH: 9:1: v/v; ninhydrin, UV)
Rf = 0.46. 1H NMR (CDCl3): d 1.14–1.57 (m, 13H,
AcrNH(CH2)2(CH2)2, C(CH3)3); 1.57–1.83 (m, 2H,
AcrNHCH2CH2); 2.90–3.20 (m, 2H, CH2NHBoc); 3.44–
3.70 (m, 2H,AcrNHCH2); 3.86 (s, 3H, OCH3); 4.63–4.90 and
6.1.3.1. Acr-C16-NLS. tR = 30.43 min, MS (MALDI-TOF):
m/z = 2138.3 in agreement with the calculated mass for
[M]+ = C103H162ClN25O20S.
6.1.3.2. Acr-C13-NLS. tR = 28.58 min, MS (MALDI-TOF):
m/z = 2096.2 in agreement with the calculated mass for
[M]+ = C100H156ClN25O20S.

C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
1305
6.1.3.3. Acr-C7-NLS. tR = 22.44 min, MS (MALDI-TOF):
m/z = 2012.2 in agreement with the calculated mass for
[M]+ = C94H144ClN25O20S.
ing the lipoplexes and polyplexes with sodium dodecyl sul-
fate, following the procedures described elsewhere [19].
6.5. Fluorescence assays
6.1.3.4. Acr-C5-NLS. tR = 24.55 min, MS (MALDI-TOF):
m/z = 1941.1 in agreement with the calculated mass for
[M]+ = C91H139Cl18N24O19S.
The labeled lipoplexes and polyplexes were prepared as
described above from a liposomal DOGS/DOPE dispersion
or PEI solution in 5% glucose, respectively, which was added
to preformed Acr-spacer-NLS/DNA complexes in 5% glu-
cose to reach a final volume of 1.5 ml, a Acr-spacer-NLS
concentration of 0.26 µM, a DNA concentration of 15 µM
phosphate. These preparations were then immediately placed
in thermoregulated cuvettes at 25.0 0.5 °C. Then, the emis-
sion fluorescence spectrum was recorded using a Perkin–
Elmer spectrofluorometer LS 50B after excitation at 409 nm.
Alternatively, the Acr-spacer-NLS-conjugates in 5% glucose
were added to DOGS/DOPE liposomes, PEI, preformed
lipoplexes or polyplexes in the thermoregulated fluorescence
cuvettes. The quantities of compounds used were calculated
according to the final DNA concentration, the N/P ratio (2.5,
5 and 10), and theAcr-spacer-NLS/DNA molar ratio (1, 5 and
10).
6.2. Preparation of the DNA complexes
The plasmid pTG11236 (pCMV-SV40-luciferase-
SV40pA) used for the preparation of the DNA complexes
and for transfection assays was produced by Transgène (Stras-
bourg, France). pTG11236 is a plasmid of 5739 bp. The quan-
tities of compounds used were calculated according to the
desired DNA concentration of 0.1 mg/ml in the lipoplex or
polyplex dispersion, the N/P ratio, the molar weight and the
number of amino groups in the cationic lipid (DOGS) or cat-
ionic polymer (PEI). The N/P ratio of 5, for example, corre-
sponds to the molar amount of DOGS or PEI necessary to
have a ratio of five amino group nitrogens (for 1 mol of DOGS
or PEI) per one phosphate in the DNA (330 Da mean MW for
one nucleotide unit), as described elsewhere [19–22].
First of all, DOGS/DOPE 1:1 mol liposomal solutions in a
5% glucose were prepared at a DOGS concentration of
1 mg/ml. Thus, for the preparation of N/P 5 DOGS/DOPE/
(Acr-NLS/DNA) complexes, 100 µl of DOGS solution
(10 mg/ml in EtOH) and 58.91 µl of DOPE solutions
(10 mg/ml in CHCl3) were transferred to a borosilicate glass
tube (16 × 100 mm). The solvent was evaporated in a Rotavap
evaporation system (45 °C, 30 pm, 0.2 bar, 40 tr/min). One
milliliter of a 5% glucose solution was added to the film
obtained. The preparation was vortexed for 12 h, then soni-
cated for 5 min to yield a liposomal preparation. The lipoplex
dispersions were then prepared by mixing 1.96 µl of
1.29 mg/ml DNA solution with (i) 10 µl of the Acr-NLS con-
jugate (the concentration of this solution depends on the
desired DNA/Acr-NLS mol:mol ratio (1:1, 1:5 or 1:10), then
(ii) with the liposomal solution (11.96 µl for N/P 5) and glu-
cose 5% to reach the final volume of 25 µl and a DNA con-
centration of 0.1 mg/ml. This preparation was vortexed for
10 s and used within 1 h for the particle size measurements
and the in vitro transfection experiments.As for PEI, the poly-
mer solution was prepared at a concentration of 0.431 mg/ml
in glucose 5%. The polyplexes were prepared according to
the same procedure than for the preparation of the lipoplexes.
6.6. In vitro transfection of NIH-3T3 and A549 cells
NIH-3T3 (mouse fibroblasts) and A549 cells (epithelial
cells derived from human pulmonary carcinoma) were grown
24 h before the in vitro transfection experiments in Dulbeco-
modified Eagle culture medium (DMEM, Gibco-BRL) con-
taining 10% foetal calf serum (FCS, Sigma), in 96-well plates
(2 × 104 cells per well), in a wet (37 °C) and 5% CO2/95% air
atmosphere. Five microliters of the lipoplex or polyplex prepa-
ration were diluted to 100 µl in DMEM. The cell culture
medium was removed and replaced with this 100 µl lipoplex
(or polyplex solution which corresponds to a DNA concen-
tration of 0.5 µg and a lipid concentration of 18.94 µM for
DOGS (N/P 5) and 0.26 µM for PEI (N/P 5) in each well.
After 4 and 24 h, 50 and 100 µl DMEM supplemented with
30% and 10% FCS, respectively, were added. Forty-eight
hours after transfection, the culture medium was discarded
and the cells were washed twice with 100 µl phosphate-
buffered saline (PBS) and then lysed with 50 µl lysis buffer
(Promega, Charbonnières, France). The lysates were frozen
at –40 °C awaiting analysis of luciferase activity. This mea-
surement was done for 10 s on 10 µl lysis mixture in a LB96P
luminometer (Berthold, Evry, France) in dynamic mode, using
the ‘luciferase’ determination system (Promega) in 96 well
plates. The total protein concentration per well was deter-
mined by the BCA test (Pierce, Montluçon, France). For cells
grown in the absence of lipoplexes or polyplexes, a well con-
tained approximately 30–50 µg protein. Luciferase activity
was calculated as femtograms (fg) of luciferase per milli-
gram (mg) of protein. The percentage of cell viability of the
lipoplexes or polyplexes was calculated as the ratio of the
total amount of protein per well of the transfected cells rela-
tive to that measured for untreated cells ×100%.
6.3. Measurement of the size of the DNA complexes
The sample was diluted with glucose 5% in the measure-
ment tube and homogenized and the average sizes were mea-
sured by photon correlation spectroscopy using a Coulter
N4Plus particle size analyzer, as described elsewhere [25].
The formulations and analyses were reproduced twice.
6.4. Agarose gel electrophoresis
Each sample was analyzed and plasmid integrity in each
sample was confirmed by electrophoresis after decomplex-

1306
C. Boulanger et al. / European Journal of Medicinal Chemistry 40 (2005) 1295–1306
[9] K.H. Bremner, L.W. Seymour, A. Logan, M.L. Read, Factors influ-
6.7. Statistical analysis
encing the ability of nuclear localization sequence peptides to enhance
nonviral gene delivery, Bioconjugate Chem. 15 (2004) 152–161.
[10] C. Neves, V. Escriou, G. Byk, D. Scherman, P. Wils, Intracellular fate
and nuclear targeting of plasmid DNA, Cell Biol. Toxicol. 15 (1999)
193–202.
[11] L.J. Branden, A.J. Mohamed, C.I. Smith, A peptide nucleic acid-
nuclear localization signal fusion that mediates nuclear transport of
DNA, Nat. Biotechnol. 17 (1999) 784–787.
Statistical tests were performed with STATGRAPHICS
Plus5.0 software. Analysis of variance (ANOVA) was run on
the logarithmic transformation of transfection levels
(Log10(fg luciferase per mg protein)) and on the cell viabil-
ity to fit normal distributions of the data.
Five factors, i.e. nature of the Acr-spacer-NLS-conjugates
(length of the spacer) or of the intercalating agent (Acr, Phen,
BET), Acr-spacer-NLS/DNA ratios (1, 5, and 10), N/P ratios
(0.8, 1.25, 2.5, 5 and 10), and amount of DNA (0.1 and 0.5 µg
per well) were analyzed as source of the variation of logarith-
mic transformation of the transfection levels and of cell viabil-
ity percentages using a multiple comparison procedure. The
Tukey’s honestly significant difference (HSD) method was
used to discriminate among the means of cell viability per-
centages and the logarithmic transformation of luciferase
expression levels.
[12] B. Schwartz, M.A. Ivanov, B. Pitard, V. Escriou, R. Rangara, G. Byk,
P. Wils, J. Crouzet, D. Scherman, Synthetic DNA-compacting pep-
tides derived from human sequence enhance cationic lipid-mediated
gene transfer in vitro and in vivo, Gene Ther. 6 (1999) 282–292.
[13] C.K. Chan, D.A. Jans, Enhancement of polylysine-mediated transfer-
rinfection by nuclear localization sequences: polylysine does not
function as a nuclear localization sequence, Hum. Gene Ther. 10
(1999) 1695–1702.
[14] C.K. Chan, D.A. Jans, Enhancement of MSH receptor- and GAL4-
mediated gene transfer by switching the nuclear import pathway,
Gene Ther. 8 (2001) 166–171.
[15] P. Collas, P. Alestrom, Nuclear localization signals: a driving force for
nuclear transport of plasmid DNA in zebrafish, Biochem. Cell Biol. 75
(1997) 633–640.
[16] A. Subramanian, P. Ranganathan, S.L. Diamond, Nuclear targeting
peptide scaffolds for lipofection of nondividing mammalian cells,
Nat. Biotechnol. 17 (1999) 873–877.
Acknowledgements
[17] F. Zhang, P. Andreassen, P. Fender, E. Geissler, J.F. Hernandez,
J. Chroboczek, A transfecting peptide derived from adenovirus fiber
protein, Gene Ther. 6 (1999) 171–181.
We wish to thank TRANSGENE, SA (Strasbourg, France)
for the generous gift of pcTG11236.
[18] J. Feigon, W.A. Denny, W. Leupin, D.R. Kearns, Interactions of
antitumor drugs with natural DNA: 1H NMR study of binding mode
and kinetics, J. Med. Chem. 27 (1984) 450–465.
References
[19] J. Gaucheron, C. Santaella, P. Vierling, Highly fluorinated liposper-
mines for gene transfer: synthesis and evaluation of their in vitro
transfection efficiency, Bioconjugate Chem. 12 (2001) 114–128.
[20] J. Gaucheron, C. Santaella, P.Vierling, Improved in vitro gene transfer
mediated by fluorinated lipoplexes in the presence of a bile salt
surfactant, J. Gene Med. 3 (2001) 338–344.
[21] O. Boussif, J. Gaucheron, C. Boulanger, C. Santaella, H.V. Kolbe,
P. Vierling, Enhanced in vitro and in vivo cationic lipid-mediated gene
delivery with a fluorinated glycerophosphoethanolamine helper lipid,
J. Gene Med. 3 (2001) 109–114.
[22] J. Gaucheron, C. Boulanger, C. Santaella, N. Sbirrazzuoli, O. Boussif,
P. Vierling, In vitro cationic lipid-mediated gene delivery with fluori-
nated glycerophospho-ethanolamine helper lipids, Bioconjugate
Chem. 12 (2001) 949–963.
[23] J.A. Burns, J.C. Butler, J. Moran, G.M. Whitesides, Selective reduc-
tion of disulfides by Tris(2-carboxyethyl)phosphine, J. Org. Chem. 56
(1991) 2648–2650.
[24] H. Pollard, J.S. Remy, G. Loussouarn, S. Demolombe, J.P. Behr,
D. Escande, Polyethylenimine but not cationic lipids promotes trans-
gene delivery to the nucleus in mammalian cells, J. Biol. Chem. 273
(1998) 7507–7511.
[25] G. Verderone, N. Van Craynest, O. Boussif, C. Santaella, R. Bischoff,
H.V. Kolbe, P. Vierling, Lipopolycationic telomers for gene transfer:
synthesis and evaluation of their in vitro transfection efficiency, J.
Med. Chem. 43 (2000) 1367–1379.
[1] J. Zabner, A.J. Fasbender, T. Moninger, K.A. Poellinger, M.J. Welsh,
Cellular and molecular barriers to gene transfer by a cationic lipid, J.
Biol. Chem. 270 (1995) 18997–19007.
[2] S. Brunner, T. Sauer, S. Carotta, M. Cotten, M. Saltik, E. Wagner, Cell
cycle dependence of gene transfer by lipoplex, polyplex and recombi-
nant adenovirus, Gene Ther. 7 (2000) 401–407.
[3] M. Ohno, M. Fornerod, I.W. Mattaj, Nucleocytoplasmic transport: the
last 200 nanometers, Cell 92 (1998) 327–336.
[4] M.A. Zanta, P. Belguise-Valladier, J.P. Behr, Gene delivery: a single
nuclear localization signal peptide is sufficient to carry DNA to the
cell nucleus, Proc. Natl. Acad. Sci. USA 96 (1999) 91–96.
[5] J.J. Ludtke, G. Zhang, M.G. Sebestyen, J.A. Wolff, A nuclear local-
ization signal can enhance both the nuclear transport and expression
of 1 kb DNA, J. Cell Sci. 112 (1999) 2033–2041.
[6] C. Ciolina, G. Byk, F. Blanche, V. Thuillier, D. Scherman, P. Wils,
Coupling of nuclear localization signals to plasmid DNA and specific
interaction of the conjugates with importin alpha, Bioconjugate
Chem. 10 (1999) 49–55.
[7] R. Cartier, R. Reszka, Utilization of synthetic peptides containing
nuclear localization signals for nonviral gene transfer systems, Gene
Ther. 9 (2002) 157–167.
[8] T. Nagasaki, T. Myohoji, T. Tachibana, S. Futaki, S. Tamagaki, Can
nuclear localization signals enhance nuclear localization of plasmid
DNA? Bioconjugate Chem. 14 (2003) 282–286.