Lipophilic pyrylium salts in the synthesis of efficient pyridinium-based cationic lipids, gemini surfactants, and lipophilic oligomers for gene delivery

Article abstract of DOI:10.1021/jm0601755

Several new classes of pyridinium cationic lipids were synthesized and tested as gene delivery agents. They were obtained through a procedure that generates simultaneously the heterocyclic ring and the positively charged nitrogen atom, using lipophilic pyrylium salts as key intermediates that react with primary amines, yielding pyridinium salts. The choice of the appropriately substituted primary amine, diamine or polyamine, allows the design of the shape of the final lipids, gemini surfactants, or lipophilic polycations. We report also a comprehensive structure-activity relationship study that identified the most efficient structural variables at the levels of the hydrophobic anchor, linker, and counterion for these classes of pyridinium cationic lipids. This study was also aimed at finding the best liposomal formulation for the new transfection agents.

Full text of DOI:10.1021/jm0601755

3
872
J. Med. Chem. 2006, 49, 3872-3887
Lipophilic Pyrylium Salts in the Synthesis of Efficient Pyridinium-Based Cationic Lipids,
Gemini Surfactants, and Lipophilic Oligomers for Gene Delivery
†
†
‡
‡
‡
‡
Marc Antoniu Ilies, William A. Seitz, Betty H. Johnson, Edward L. Ezell, Aaron L. Miller, E. Brad Thompson, and
,
†
Alexandru T. Balaban*
Texas A & M UniVersity at GalVeston, MARS, 5007 AVenue U, GalVeston, Texas 77551, and Department of Biochemistry and Molecular
Biology, UniVersity of Texas Medical Branch, 301 UniVersity BouleVard, GalVeston, Texas 77555
ReceiVed February 15, 2006
Several new classes of pyridinium cationic lipids were synthesized and tested as gene delivery agents. They
were obtained through a procedure that generates simultaneously the heterocyclic ring and the positively
charged nitrogen atom, using lipophilic pyrylium salts as key intermediates that react with primary amines,
yielding pyridinium salts. The choice of the appropriately substituted primary amine, diamine or polyamine,
allows the design of the shape of the final lipids, gemini surfactants, or lipophilic polycations. We report
also a comprehensive structure-activity relationship study that identified the most efficient structural variables
at the levels of the hydrophobic anchor, linker, and counterion for these classes of pyridinium cationic
lipids. This study was also aimed at finding the best liposomal formulation for the new transfection agents.
6
7
1. Introduction
categories, namely, cationic lipids and cationic polymers. Both
1
classes have much lower immunogenicity and cytotoxicity than
viral vectors, allow the use of plasmids of practically unlimited
size, and can be manufactured and stored in bulk quantities
under GMP-compliant norms.
Cationic lipids are amphiphilic molecules containing a polar
Gene therapy is a revolutionary form of therapy that promises
to treat diseases at the cellular level using the ultimate drug:
the DNA. Its success relies on finding efficient vectors for the
transfer and expression (transfection) of the genetic material to
8
2
the desired location in the living organism. Thus, when the
(cationic) head linked via a spacer to a hydrophobic tail. When
cellular machinery is impaired, a copy of a gene correcting a
deficient (abnormal) gene can be transfected into cells, tissues,
or organs affected by specific hereditary or acquired diseases.
In cancer therapy, on the other hand, one hopes to deliver a
gene that selectively kills or inhibits the malignant cells.
Permanent incorporation of such a gene into the genome often
is not desired. In contrast to conventional therapies, the limiting
factor is not the quantity of the DNA but the efficiency of its
overall delivery. Side effects are often limiting in conventional
therapy. A new paradigm will have to be devised, replacing
a certain concentration is reached, they self-assemble via
cooperative hydrophobic binding forming cationic liposomes.
In this form they can interact electrostatically with the negatively
charged DNA molecules, forming cationic lipid-DNA com-
9
plexes (lipoplexes) in which the genetic material is highly
compacted and protected from the action of endogenous
nucleases. The characteristics of the lipoplexes (lipid nature and
composition of the parent liposomes, properties of the plasmid,
and the size, shape, and ú potential of the supramolecular
complexes) are essential for achieving high levels of transfec-
a
the classical ADME/Tox paradigm.
1
0
tion.
Viral vectors are still the most efficient transfection systems
available, being structurally evolved to circumvent effectively
the various delivery barriers encountered by a foreign plasmid
3
With the focus on the nature of the cationic lipid, it must be
stressed that after 2 decades of synthetic work that generated
several commercial cationic lipid transfection systems such as
DOTAP (1), DMRIE (2), and DOSPA (3) (Chart 1), there is
still room for improvement of the transfection efficiency while
4
before reaching the nucleus of the target cell. Although current
research is attempting to improve viral vectors,5a,b their use is
still limited by serious side effects such as immunogenicity,
4
,6,11
mutagenicity, and sometimes fatal toxicity.5c Other important
disadvantages are the limited size of the plasmid that can be
minimizing the cytotoxicity.
In recent years, it has been
shown by a few different groups that progress can be achieved
1
2-14
via the heterocyclic cationic lipids (4-7, Chart 2).
Among
inserted into the virion and the difficulties associated with good
manufacturing practice (GMP) production.5
these heterocyclic cationic lipids, the pyridinium cationic lipids
reported first by Engberts’ group and then by ourselves
1
3
14
An alternative approach to efficient gene delivery vectors is
to develop synthetic transfer agents, which fall into two main
reached or surpassed the transfection efficiency of commercial
cationic lipid formulations, both in vitro and in vivo, while
maintaining a low cytotoxicity.
*
To whom correspondence should be addressed. Phone: 409-741-4313.
Fax: 409-740-4787. E-mail: balabana@tamug.edu.
7
Similar to cationic lipids, cationic polymers, when mixed
†
Texas A & M University at Galveston.
University of Texas Medical Branch.
9
‡
with DNA, form compact polyplexes. Although it is somehow
a
Abbreviations: ADME/Tox, absorption, distribution, metabolism, and
easier to control polyplex properties than those of lipoplexes,
owing to the templating features of the polymeric backbone,
cationic polymers usually display higher cytotoxic and im-
munogenic profiles than cationic lipids. Recently, another class
emerged to fill the gap between cationic lipids and cationic
polymer systems, borrowing advantages from both classes. This
excretion/toxicity; GMP, good manufacturing practice; DOTAP, N-(2,3-
dioleoyloxypropyl)-N,N,N-trimethylammonium chloride; DMRIE, N-(2,3-
dimiristyloxypropyl)-N-(hydroxyethyl)-N,N-dimethylammonium bromide;
DOSPA, N-(2,3-dioleyloxypropyl)-N-[2-(sperminecarboxamido)ethyl]-N,N-
dimethylammonium trifluoroacetate; DOPE, dioleoylphosphatidylethano-
lamine; Chol or C, cholesterol; DSC, differential scanning calorimetry; PBS,
phosphate buffer isotonic saline at pH 7.4; BCA, bicinchoninic acid; bovine
BSA, serum albumin; RLU, relative luminescence units; Lipofect, Lipo-
fectamine; Boc, tert-butyloxycarbonyl.
15
new class contains cationic gemini surfactants and oligomeric
1
6
cationic surfactants.
1
0.1021/jm0601755 CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/01/2006

Pyrylium Salts for Synthesis
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3873
Chart 1
various designs while maintaining a truncated cone shape for
the cationic amphiphile. It also provides a simple and direct
synthetic route to gemini and oligomeric surfactants. We report
now a detailed structure-activity relationship study aimed at
identifying the most efficient structural variables for these new
types of pyridinium cationic vectors. This study also sought the
best formulation for these novel transfection agents.
2
. Results and Discussion
One of the most important variables in the structure of
cationic lipids is the length of the hydrophobic chain.6
,14,20
This
parameter influences directly the shape of the individual
amphiphilic molecules and acts with the polar head to influence
the supramolecular associations that can be generated through
Chart 2
2
1
the self-assembling process. Therefore, we synthesized a
library of lipophilic pyridinium salts in order to evaluate the
influence of the chain length on the physicochemical and
biological properties of these new transfection vectors. The
synthesis (Scheme 1) is based on the original synthetic strategy
14a,b
reported previously,
which involves the reaction of pyrylium
salts with primary amines to generate the cationic head in a
2
2
single, high-yield step. Eight lipophilic pyridinium salts with
linear alkyl side chains ranging from 10 to 17 carbon atoms
and with hexafluorophosphate as counterion were obtained via
1
9
a modified literature procedure. Briefly, mesityl oxide was
2
2
monoacylated with fatty acid chlorides 12 in the presence of
aluminum chloride to yield the pyrylium chloroaluminates 13,
which were converted into the more stable hexafluorophosphates
1
4 by hexafluorophosphoric acid. The pyrylium hexafluoro-
phosphates were subsequently reacted with primary linear
alkylamines ranging from 10 to 18 carbon atoms, thus generating
the library of pyridinium cationic lipids 15-28 (Scheme 1 and
Table 1).
Chart 3
Physicochemical data of the new compounds show that
pyrylium salts 14 display alternating melting points; compounds
with an odd number of carbon atoms in the side chain exhibit
lower melting points than their congeners with an even number
of carbon atoms in the side chain. This behavior, similar to that
of fatty acids, is due to the differences in the crystal packing of
the amphiphilic molecules. A similar behavior is observed for
the critical temperatures of the pyridinium cationic lipids; those
compounds with two identical fatty chains (each with an even
number of carbon atoms) display slightly higher Tc values than
their congeners with an odd and an even number of carbon
atoms in the two long side chains (compare 18 with 20 and
compare 25 with 26). Otherwise, the critical temperatures are
generally increasing while elongating the alkyl chains.
Gemini surfactants are synthetic transfection vectors compris-
ing two simple surfactant moieties linked together at the level
of the polar head. Introduced in the 1970s, they experienced a
new outburst in the early 1990s because of their unusual
properties.17 They possess a higher charge per mass ratio than
cationic lipids and also have superior surfactant properties. They
can self-assemble at very low concentrations, forming a variety
of supramolecular assemblies (micelles, bilayers, inverted
micelles). Because of these special properties, they have been
To assess how the chain length influences the transfection
activity, the lipids were conditioned using a previously optimized
14b,c
protocol for these compounds.
Thus, the cationic lipids were
15b
recently used as vectors for gene therapy. Similar properties
mixed with cholesterol at a 1:1 molar ratio, dissolved in
chloroform/methanol, and dried under vacuum to generate a lipid
film. The dried lipid film was hydrated with sterile phosphate
buffer isotonic saline at pH 7.4 (PBS), vortex-mixed, and
sonicated to yield cationic liposomes. The resulting liposomal
solution was allowed to react with a DNA solution to form the
final cationic lipid-DNA complexes (lipoplexes), which were
assessed for transfection efficiency. We used a pGL3 plasmid
were reported for oligomeric surfactants.18
Our previous structure-activity relationship study14b of three
series of pyridinium cationic lipids (8-10, Chart 3) revealed
that the most efficient representative was the N-(1,3-dimyris-
toyloxypropane-2-yl)collidinium derivative 9. This amphiphilic
compound, because of its truncated cone shape, induces a higher
radius of curvature of the lipid bilayer, thus generating smaller,
more transfection-efficient liposomes. The same features were
found to be beneficial by Engbert’s group.13
(Promega, Madison, WI), encoding a firefly luciferase gene
under the control of the constitutively active SV40 promoter,
as the reporter for the transfection efficiency. An optimal1
electrostatic charge ratio for cationic lipid/DNA of 2:1 was used
in all preparations. Cholesterol (Chol) was used as colipid in
the liposomal formulations because it had proved to be superior
4b,c
These findings prompted us to investigate new pyridinium
cationic lipids with the alkyl chains directly attached on the
aromatic ring, using the reaction between a lipophilic pyrylium
19
14,19
salt and a fatty amine. This original, flexible strategy
allows

3
874 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Ilies et al.
Scheme 1
Table 1. Lipophilic Pyrylium Salts 14 and Dialkylpyridinium Cationic
Lipids 15-28 with Their Critical Temperatures (Tc)
compd
4a
4b
4c
4d
4e
4f
4g
4h
R′
R
mp/Tc (°C) compd
R′
R
mp/Tc (°C)
1
1
1
1
1
1
1
1
C10
C11
C12
C13
C14
C15
C16
C17
67.6
62.8
78.7
73.3
85.2
79.8
88.1
81.8
46.9
56.7
52.2
18
19
20
21
22
23
24
25
26
27
28
C12 C11
C10 C12
C12 C12
C10 C13
C12 C13
C14 C14
C14 C15
C16 C15
C16 C16
C16 C17
C18 C17
54.8
57.6
58.6
58.4
67.2
68.5
74.7
74.1
76.2
82.2
82.1
1
1
1
5
6
7
C10 C10
C10 C12
C10 C11
Figure 2. Optimization of helper lipid and its molar ratio relative to
the cationic lipid 23 in liposome preparation. NCI-H23 cell extracts
were assayed 24 h later for luminescence (RLU/µg of protein).
diacylglycerol cationic lipids, more efficient transfection could
be achieved with cationic lipids having hydrophobic tails of
unequal lengths than with the symmetrical congeners.24
To validate the conditions optimized by the experiment of
Figure 1, several formulations based on 23 were tested, using
cholesterol or DOPE in various molar ratios as helper lipids.
Lipofectamine was used as a reference at 2:1 electrostatic charge
ratio to DNA. The results confirmed that the best helper lipid
for these compounds is cholesterol, coformulated with the
cationic lipid at a 1:1 molar ratio (Figure 2).
Pyridinium Cationic Lipids Coupled with Gemini Sur-
factants. An important parameter in gene delivery is the amount
of compacting cationic species added to the DNA in the delivery
complex in order to reduce the dimensions of DNA and mask
its negative charge. One wants to keep the amount of cationic
entity as low as possible in order to minimize its immunoge-
nicity. As mentioned before, a higher (positive) charge per mass
ratio than in cationic lipids can be obtained by using cationic
Figure 1. Transfection data for cationic lipids 15-28 compared with
DOTAP as reference. All lipids were conditioned with cholesterol as
a helper lipid, at 1:1 molar ratio. An optimized electrostatic charge
ratio of cationic lipid/DNA of 2:1 was used. NCI-H23 cell extracts
were assayed 24 h later for luminescence (RLU/µg of protein).
to dioleoylphosphatidylethanolamine (DOPE) for this type of
pyridinium cationic lipid.14b,c All liposomal preparations were
sonicated for 30 min at 65-68 °C (twice for 15 min, with a
15
gemini surfactants. Moreover, these amphiphilic compounds
have the ability to help the endosomal release of the DNA
complexes by destabilizing the endosomal membrane via
1
5-min pause between) for the new compounds, and at 37 °C
2
5
for the chosen reference cationic formulation DOTAP/Chol (1:1
molar ratio),23 similar to our previous studies.14b,c Under these
conditions liposomes with sizes ranging from 87 to 191 nm were
obtained as determined by dynamic light scattering, with a ú
potential of around 30 mV. The resulting lipoplexes had a ú
potential of around -15 mV. We used the lung cancer cell line
NCI-H23 as the main test system for in vitro experiments, since
our previous transfection studies1 on several cultured cell lines
showed it to be the most sensitive toward pyridinium cationic
lipids. The results are summarized in Figure 1.
associations with the gemini anionic lipids. This class of
15b
compounds was reported to be efficient in gene delivery;
therefore, we decided to investigate the efficiency of the
pyridinium polar head, in a gemini surfactant design, for in vitro
gene delivery.
Two libraries of pyridinium gemini surfactants were obtained
using the same synthetic strategy (Schemes 2-4). The key
intermediate, tetradecyldimethylpyrylium hexafluorophosphate
14e (whose side chain length was proved to be optimal), was
condensed with different primary diamines, generating the linked
polar heads in a single, high-yield step.
The linker’s structure has an essential role in the size and
shape and hence in the supramolecular self-assembling proper-
ties of this class of amphiphilic molecules. If hydrophobic, it
will tend to pack between the two hydrophobic chains, generat-
ing a tapered molecular shape. If hydrophilic, it will be inclined
to stay at the hydrophobic/water interface, thus changing the
4b,c
The maximum transfection efficiency was obtained when two
myristyl chains were attached to the pyridinium ring (23, Figure
1
). The level of transfection was twice that of 24 (which has
only one more carbon atom in one of the side chains) or the
standard transfection system DOTAP/Chol.
We note than in contrast to our results with these new
compounds other authors have reported that in the case of

Pyrylium Salts for Synthesis
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3875
Scheme 2
two to eight carbon atoms in order to finely tune the properties
of the new species and to assess the influence of the linker (and
hence of the shape of the molecule) on the transfection
efficiency. The seven representatives thus obtained and their
physicochemical properties are presented in Table 2.
The second library (Schemes 3 and 4) was designed to test
the influence of a hydrophilic linker, containing secondary amino
groups, on the transfection efficiency of pyridinium gemini
surfactants. Since some of these secondary amino moieties are
positively charged at the physiologic pH, they will also
contribute to the binding and compacting of the DNA and are
expected to generate a different transfection profile for com-
pounds 41 and 47.
Table 2. Gemini Cationic Lipids 29-35, 41, and 47 and Their Critical
Temperatures (Tc)
compd
spacer
Tc (°C)
2
3
3
3
3
3
3
4
4
4
4
9
0
1
2
3
4
5
1a
1b
1c
7
(CH2)2
(CH2)3
(CH2)4
(CH2)5
(CH2)6
(CH2)7
(CH2)8
178.2
186.6
185.6
137.8
189.9
83.6
68.8
65.1
45.4
67.2
a
The synthetic approach used in this case involved polyamines
36a-c and 42, with various numbers of methylene units between
the amino moieties, as starting materials, which were selectively
protected with tert-butyloxycarbonyl (Boc) groups on the
secondary amino moieties using a well-established literature
(CH2)2NH(CH2)2
(CH2)3NH(CH2)3
(CH2)3NH(CH2)4
(CH2)3NH(CH2)4NH(CH2)3
2
6
procedure. Thus, compounds 36a-c and 42 were condensed
with ethyl trifluoroacetate when the primary amino moieties
were selectively protected with trifluoroacetyl groups. Treatment
of trifluoroacetamides 37a-c and 43 with di-tert-butyl dicar-
bonate in the presence of triethylamine afforded compounds
38a-c and 44, respectively, in which the secondary amino
groups were Boc-protected. A selective hydrolysis in basic
medium generated the diamines 39a-c and 45, which were
subsequently condensed with the key intermediate 14e to
generate the pyridinium gemini compounds 40a-c and 46. The
a
Amorphous (glass transition at 9.6 °C).
solid angle between the two lipophilic chains and the shape of
the amphiphilic molecule.
The first library comprised gemini surfactants 29-35 with
nonpolar linkers, generated from n-alkyldiamines and the
pyrylium salt 14e (Scheme 2). We varied the spacer length from
Scheme 3
Scheme 4

3
876 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Ilies et al.
Figure 4. Transfection data for Boc-gemini cationic lipids 40 and 46,
compared with DOTAP as reference. All lipids were conditioned with
cholesterol as helper lipid, at 1:1 molar ratio. An electrostatic charge
ratio cationic lipid/DNA of 2:1 was used. NCI-H23 cell extracts were
assayed 24 h later for luminescence (RLU/µg of protein).
Figure 3. Transfection data for gemini cationic lipids 29-35, 41, and
7 compared with DOTAP as reference. All lipids were conditioned
with cholesterol as a helper lipid, at 1:1 molar ratio. An electrostatic
charge ratio cationic lipid/DNA of 2:1 was used. NCI-H23 cell extracts
were assayed 24 h later for luminescence (RLU/µg of protein).
4
possibility is the inactivation of the lipoplexes in contact with
2
7
serum, similar to the case of Lipofectamine. Other groups
removal of Boc protecting groups from these precursors via HCl/
28
reported similar findings.
MeOH yielded the final polycationic amphiphiles 41a-c and
In the case of gemini lipids 29-35 another interesting
behavior was observed (Figure 3). On elongation of the linker
from two carbon atoms (compound 29, which is somehow
similar in shape to compound 23) to three carbon atoms
4
7. Their physicochemical characteristics are shown in Table
2.
The use of this synthetic strategy was made necessary because
22
of the interference of the secondary amino groups in a direct
reaction between polyamines 36 or 42 and pyrylium salt 14e,
which translates into a complex mixture of products, which are
very difficult to separate.
(compound 30), there was a steep decrease of transfection.
Further elongation by four or five carbon atoms gave little
improvement, but thereafter the transfection efficiency increased
monotonically with the linker length. Even so, the transfection
efficiency of gemini amphiphile 35 was lower than of its
congener 29, which transfected similarly to the standard
transfection system DOTAP/Chol 1:1 (Figure 3, “DOTAP”).
Changing the helper lipid and its molar ratio in the formulation,
as well as the lipid/DNA charge ratio, did not improve the
transfection efficiency (data not shown). The transfection results
The data in Table 2 show that the critical temperatures of
the cationic amphiphiles depend markedly on the structure of
the linker. As a general trend, gemini cationic lipids with alkyl
linkers display higher transition temperatures than their conge-
ners bearing polar secondary amino moieties in the spacer.
Within the gemini series 29-35 it can be observed that when
the spacer is elongated from two to eight carbon atoms, two
opposite effects can be noticed. Thus, when the spacing is
increased from two to three carbon atoms in the linker, the
critical temperature increases because of a better separation of
the positive charges on the pyridinium rings. It remains elevated
with further elongation of the spacer up to six methylene units,
after which it decreases abruptly, possibly because of a change
in the packing of hydrophobic elements.
(
with the exception of 29) are similar to the ones reported by
28b
MacDonald and co-workers for quaternary ammonium-based
gemini surfactants.
Since the polycationic lipids 41 and 47 displayed much lower
transfection efficiency than the related gemini congeners 29-
3
5, we tested the transfection efficiency of their precursors 40
and 46, hypothesizing that they would give higher transfection
efficiency. These compounds have the secondary amino moieties
Boc-protected, behaving therefore as normal gemini cationic
surfactants. We found that their critical temperatures are higher
than their congeners having n-alkyl linkers with the same
dimensions (Table 3).
In the case of hydrophilic linkers, substitution of a methylene
unit by an amino group decreased the critical temperature
(compare 32 with 41a and compare 34 with 41b). Interestingly,
when the linker is sufficiently long, the differences between its
characteristics are blurred (compare 35 with 41c), the thermal
stability of the compounds being dictated predominantly by other
structural elements. Compound 47, with two secondary amino
moieties in the linker, is a viscous liquid at room temperature.
Therefore, it would be expected that the gemini surfactants with
hydrophilic spacers 41 and 47 would display higher transfection
efficiencies than their hydrophobic congeners 29-35, since the
fluidity of the complexes with DNA would be much higher
under physiological conditions.10 However, after the lipids were
tested as a 1:1 molar ratio mixture with cholesterol, the reverse
was observed (Figure 3). Polycationic compounds 41 and 47
displayed much lower transfection efficiencies than their
hydrophobic congeners 29-35. This fact may be explained by
considering that a transfection event implies both an effective
compaction of the DNA via electrostatic association of the lipid
with the DNA (favored in the case of polycationic surfactants)
and an efficient release of the plasmid from the complex after
passing the cell’s membrane, which is slowed by the presence
of multiple positive charges on the lipid’s polar head. Another
Table 3. Gemini Cationic Lipids 40 and 46 and Their Critical
Temperatures (Tc)
compd
spacer
Tc (°C)
4
4
40c
0a
0b
(CH2)2NBoc (CH2)2
(CH2)3NBoc (CH2)3
(CH2)3NBoc (CH2)4
(CH2)3NBoc(CH2)4NBoc(CH2)3
139.9
134.4
143.5
103.3
46
Compounds 40 and 46 were conditioned (1:1 molar) with
cholesterol, assayed under the same conditions as those used
for the gemini series 29-35. The results of transfections (Figure
4) confirmed the working hypothesis. The transfection efficiency
was found to depend on the structure of the linker; the best
results were obtained with 40a and 46 (Figure 4).
Interestingly, 40a was at least 5 times more active than its
alkyl congener 32. The Boc derivative 46 was the most active,
being about 3 times more efficient than the commercial
formulation DOTAP/Chol 1:1. These facts indicated that ef-

Pyrylium Salts for Synthesis
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3877
Scheme 5
Table 4. Multitail Cationic Lipids 48-50 and Their Critical
Temperatures (Tc)
compd
n
Tc (°C)
4
4
4
5
8a
8b
9
12
14
125.8
124.9
69.4
0
121.2
ficient pyridinium gemini cationic lipids can be obtained using
hydrophilic linkers decorated with lipophilic moieties.
These findings prompted us to test the effect of substituting
the Boc groups in 46 with more lipophilic moieties, introduced
either by acylation of 47 (with myristoyl or palmitoyl chloride)
or by alkylation of the same intermediate with palmityl bromide.
Compounds 48a, 48b, and 49 were thus obtained in good yields
Figure 5. Transfection data for multitail cationic lipids 48-50
compared with Lipofectamine as reference. Cholesterol (C) was
coformulated at 1:1 molar ratio with the test lipid. DOPE (D) was
coformulated at 1:1 or 1:2 molar ratio. An electrostatic charge ratio
cationic lipid/DNA of 2:1 was used. NCI-H23 cell extracts were assayed
(Scheme 5). All these compounds were expected to form tapered
dendritic units, either flat or conical. Another conical lipophilic
dendritic amphiphile, the tris(2,4-dimethyl-6-myristylpyridinium-
2
4 h later for luminescence (RLU/µg of protein).
1
-ethylenyl)amine 50, was synthesized by condensation of the
myristylpyrylium salt 14e with tris(aminoethyl)amine (Scheme
).
The physicochemical properties of these dendritic molecules
superior to cholesterol as a colipid for all these multitail
compounds. When the amount of DOPE was doubled, the
transfection efficiency decreased markedly to about one-third
of the initial value. The transfection efficiencies of the lipids
alone are very small. By contrast, the tris(myristylpyridinium)
5
are presented in Table 4. As expected, the two bispyridinium-
bisamides 48a and 48b have critical temperatures slightly higher
than that of their Boc congener 46, a fact that can be related to
the association of the four hydrophobic tails. In contrast, the
tetraalkyl derivative 49, lacking the two amido groups, displays
a much lower transition temperature. The trispyridinium 50 has
a relatively high critical temperature probably because of the
presence of the third heterocyclic ring, which confers a better
packing.
5
0 displayed a more homogeneous transfection profile, being
significantly active even without helper lipids. Taking into
consideration all these facts and knowing that the Lipofectamine
formulation consists of the polycationic lipid DOSPA (Chart
1
), coformulated with DOPE, we hypothesize that the mecha-
nism of action of pyridinium polycationic lipids is similar to
the one of DOSPA and therefore Lipofectamine.
The transfection efficiency of these dendritic molecules was
tested on the same NCI-H23 cell line, using Lipofectamine as
reference and a constant lipid/DNA electrostatic charge ratio
of 2:1. The lipids were either tested alone or coformulated with
cholesterol or DOPE as colipids at molar ratios of 1:1 or 1:2
A comparison between the transfection results for polycationic
amphiphiles 49 and 50 and their congeners 41 and 47 reveals
the important contribution of the hydrophobic elements in the
design of polycationic surfactants. Moreover, an analysis of the
most efficient transfection systems obtained in the gemini
surfactant series (compounds 29 and 48b) shows that they
actually share common structural elements (Figure 6). Thus,
48b (as well as 48a and 49) can be considered to be dimers of
a cationic lipid having a charged pyridinium group (red circle
in Figure 6) and a second polar group (blue circle) at the water/
lipid interface. The two polar elements are spaced by a short
(Figure 5).
Data from Figure 5 reveal that 48a, 48b, and 49 have the
same transfection profile, being most active when coformulated
with DOPE in 1:1 molar ratio. In this form, the lipid 49 was
more efficient than the commercial transfection system Lipo-
fectamine. At the 1:1 molar ratio, DOPE was found to be

3
878 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Ilies et al.
Figure 6. Energy-minimized structures of gemini lipids 29 and 48b. It can be considered that both structures share a common structural motif
shown at right), consisting of a pyridinium polar head (indicated by a large blue circle) and a second polar moiety (indicated by a red circle) placed
(
at the water/hydrophobic interface, spaced by a short linker (two or three carbon atoms). For segment A in the right-hand structural motif, we chose
ester and amide polar groups (see text).
Scheme 6
linker (with two or three carbon atoms). In 29, the second
pyridinium ring can be considered to act as the secondary polar
group.
N-ethylmercaptopyridinium salt 55 with n-alkylthiols. In contrast
with compounds 52 and 54, which possess a hydrophilic element
in the second chain that induces a tapered shape in aqueous
media, disulfides 56 were expected to self-assemble in a way
similar to lipophilic pyridinium salts 15-28. We used n-
alkylthiols with chain lengths ranging from 10 to 14 carbon
atoms, thus generating lipophilic pyridinium salts structurally
similar to the most efficient congeners 15-28. Additionally,
the same key intermediate 55 was oxidized alone with iodine
in ethanol to yield the “dimeric” disulfidic gemini pyridinium
salt 57.
Therefore, we decided to investigate different constructs that
match the basic design depicted in Figure 6. We used ester and
amide polar groups (segment A in Figure 6) because they are
biodegradable and can reduce the cytotoxicity of the new
cationic lipids. The synthesis is outlined in Scheme 6.
Thus, the starting myristylpyrylium salt 14e was condensed
either with ethanolamine or with ethylenediamine to yield the
hydroxy- or aminoethylpyridinium salts 51 and 53, which were
subsequently acylated with fatty acid chlorides to generate the
final cationic lipids 52 and 54. We varied the second chain
length from 12 to 16 carbon atoms in order to find the optimum
size for the best transfection efficiency in these new classes of
cationic lipids.
Pyridinium amides 54 were obtained as mixtures of coun-
terions; thus, they had to be subsequently transformed into
chlorides by anion exchange over a Dowex resin in chloride
1
4b
form. We had demonstrated in our previous study that there
are essentially no differences in the transfection efficiency and
cytotoxicity between hexafluorophosphate and chloride anions,
the chloride anion being more biocompatible. The characteristics
of these new series of cationic lipids are outlined in Table 5.
We tested the three new series of cationic lipids 52, 54, and
Also, to test the validity of the new design, a lipophilic
disulfide group was introduced instead of the polar ester and
amidic groups, via condensation of 14e with 2-aminoethanethiol
followed by the oxidative cross-coupling with I2/EtOH of the

Pyrylium Salts for Synthesis
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3879
Table 5. Cationic Lipids of Type 52, 54, 56, and 57 with Their Critical
Temperatures (Tc)
compd
Tc (°C)
counterion
compd
Tc (°C)
counterion
Cl-
-
52a
52b
52c
54a
54b
54c
56a
61.3
68.5
77.8
41.9
49.9
46.8
76.0
PF6
PF6
PF6
Cl
Cl
Cl
56a
56b
56b
56c
56c
57
77.4
81.0
75.1
81.8
86.9
-
-
PF6
-
-
Cl
-
-
PF6
Cl
PF6
Cl
-
-
-
-
107.0
139.3
-
-
PF6
57
Figure 9. Transfection efficiency for compounds 53 and 54 compared
with standard transfection reagents. In all preparations an electrostatic
charge ratio of cationic lipid/DNA of 2:1 was used. NCI-H23 cell
extracts were assayed 24 h later for luminescence (RLU/µg of protein).
Figure 7. Transfection efficiency for compounds 51 and 52 compared
with standard transfection reagents. In all preparations an electrostatic
charge ratio of cationic lipid/DNA of 2:1 was used. NCI-H23 cell
extracts were assayed 24 h later for luminescence (RLU/µg of protein).
Figure 10. Cytotoxicity data for 54 compared with standard trans-
fection reagents. In all preparations an electrostatic charge ratio of
cationic lipid/DNA of 2:1 was used. NCI-H23 cell extracts were assayed
2
4 h later for luminescence (RLU/µg of protein).
exploit the high intracellular levels of glutathione,3
0b-e
revealing
once again the importance of structure-activity correlation
studies in the design of novel cationic lipid-based transfection
systems.
In contrast to the disulfide series, the esters 52 and the amides
54 showed a good transfection profile, the esters being generally
more active and less cytotoxic than amides, despite having a
-
higher Tc (mostly due to the PF6 anion) and a less biocom-
patible counterion (Table 5, Figures 7 and 9). In both cases the
transfection efficiency peaked at the myristoyl derivatives,
similar to 23 and 29, a fact that proves that the structural
assumptions were correct. Thus, 52b and 54b, when conditioned
with DOPE as colipid at a 1:1 molar ratio, were 1.5-6 times
more efficient than commercial transfection systems, including
the very efficient DMRIE-C formulation. In particular, the
transfection efficiency of 52b was associated with a very low
(practically negligible) cytotoxicity, which is particularly im-
portant for further in vivo applications. The amide series 54,
although also very transfection efficient, showed a higher
cytotoxic side effect, probably due in part to the possible in
vivo degradation product 53, which may be much more toxic
than its hydroxyl congener 51 (Figures 8 and 10).
Good results were also obtained when cholesterol was used
as colipid (at 1:1 molar ratio to cationic lipid), especially in the
ester series 52, where its effect seems to be correlated with the
second chain’s length. The members of the amide series 54
showed a homogeneous transfection profile when conditioned
with cholesterol, and were rather insensitive to the modifications
of the secondary chain length. These differences between the
two hydrophilic groups might be explained considering that the
ester group is a hydrogen-bond acceptor and therefore sensitive
to the hydrogen-bond-donating properties of the cholesteric OH
group, whereas the amide group is a hydrogen-bond donor and
Figure 8. Cytotoxicity for 52 compared with standard transfection
reagents. In all preparations an electrostatic charge ratio of cationic
lipid/DNA of 2:1 was used. NCI-H23 cell extracts were assayed 24 h
later for luminescence (RLU/µg of protein).
2
0
56 against several commercial formulations (DOTAP/Chol 1/1,
DMRIE/Chol 1/1, Lipofectamine) on the NCI-H23 cell line at
a constant electrostatic charge ratio of cationic lipid/DNA of
2
:1. The new lipids were tested alone or coformulated with
cholesterol or DOPE in a 1:1 molar ratio, since this ratio was
the most efficient in previous cases. A WST-1 cytotoxicity
assay29 was performed in parallel (Figures 7-10). Lower
absorbance in this latter assay corresponds to higher toxicity.
The disulfides 56 and 57, together with the precursor 55 were
also tested in the same conditions. However, they were found
to be devoid of any significant transfection efficiency (efficiency
less than 3000 RLU/µg of protein), showing moreover signifi-
cant cytotoxic effects (data not shown). The transfection
efficiency did not improve when the counterion was changed
-
-
from PF6 to Cl , using the DOWEX anion exchanger resin.
These findings are similar to the conclusions of a previous
3
0a
study but at variance with the results of other groups that
used a strategically placed disulfide group to enhance the
transfection efficiency in various redox-sensitive designs that

3
880 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Ilies et al.
Figure 11. Transfection data for the most efficient cationic lipids in MCF-7 (blue bars), MDA-231â (red bars), DU-145 (black bars), and SWB-95
green bars) tumor cell lines against different commercial formulations. In all preparations an electrostatic charge ratio of cationic lipid/DNA of 2:1
(
was used.
thus rather inert toward these interactions. The liposomes
generated from lipids 52 and 54 had sizes between 74 and 157
nm, with a ú potential ranging from + 23 to + 44 mV; their
corresponding lipoplexes had a negative ú potential, around -15
mV, for the lipids alone or conditioned with cholesterol, and
around -50 mV, when conditioned with DOPE. Mention must
be made that these new lipids are able to transfect NCI-H23
cells with low cytotoxicity even without a helper lipid.
Since it is known that the transfection efficiency of synthetic
vectors varies with the cell line, we tested the most efficient
cationic amphiphilic formulations on four additional tumor cell
lines. We chose two breast cancer cell lines, MCF-7 and MDA-
case with the ester 52b, conditioned with DOPE at 1:1 molar
ratio, which was 6 times more efficient than DMRIE-C, largely
surpassing the other two commercial formulations tested.
3
. Conclusions
An extensive structure-activity relationship study was carried
out to identify the structural parameters required for the best
transfection efficiency in different classes of gene transfer
agents. The optimization of liposomal formulations and the
transfection efficiency of the corresponding lipoplexes on several
tumor cell lines were discussed. Incorporation of groups that
may be removed enzymatically (such as ester and amide) in
the structure of the lead compounds 52b and 54c reduced the
cytotoxicity and enhanced the transfection efficiency by over-
coming the intracellular delivery barriers.
231â, a prostate tumor line (DU-145), and a glioma cell line
(SWB-95), all representative of important targets in cancer
therapy. The transfection conditions were kept constant for all
preparations.
By making use of the high reactivity of pyrylium salts toward
primary amines, we present an alternative to traditional ap-
proaches for constructing cationic lipids that are efficient as
gene delivery agents. The approach is versatile, allowing access
to cationic lipids, gemini surfactants, and oligomeric surfactants.
The simplicity of the procedure makes it useful in supramo-
lecular chemistry, when hydrophobicity-mediated self-assem-
bling elements are required in the design of novel materials with
various applications.
The results (Figure 11) showed that optimized formulations
based on the new cationic lipids synthesized in this study
reached or surpassed the efficiency of commercial transfection
systems DOTAP/Chol, DMRIE-C, and Lipofectamine. For the
MCF-7 cell line, the best results were obtained with the cationic
lipids 52b (conditioned with DOPE at 1:1 molar ratio) and 54c
(conditioned with DOPE or cholesterol, at the same 1:1 molar
ratio), which surpassed the most efficient commercial system
DMRIE-C. Interestingly, 52b/Chol 1/1, which was very efficient
on the NCI-H23 cell line, showed a lower activity on MCF-7
cells.
4. Experimental Protocols
c
Chemistry. General. The phase transition temperatures (T ) for
The gemini surfactant 29 was the most efficient system on
the MDA-231â cell line, being 2 times more efficient than
DMRIE-C and largely surpassing DOTAP and Lipofectamine.
Ester 52b and amide 54c also gave good transfection results.
The latter compound also proved to be efficient on the DU-
cationic lipids were determined by differential scanning calorimetry,
using a TA-Instruments Q100 DSC and a heating rate of 5 °C/
min. The same instrument was used for recording the melting points
for the pyrylium salts and other intermediates. A Mel-Temp II
(Laboratory Devices) melting point device was also used for this
purpose. NMR spectra were recorded at ∼300 K with a Varian
145 cell line, with transfection levels similar to those of DMRIE-
Unity Plus spectrometer equipped with a 5 mm indirect detection
C.
1
13
probe, operating at 400 MHz for H NMR and at 100 MHz for
C
The most spectacular results were obtained on the brain
NMR. Chemical shifts are reported as δ values, using tetrameth-
ylsilane (TMS) as internal standard for proton spectra and the
solvent resonance for carbon spectra. Assignments were made on
carcinoma SWB-95 cell line, known to be very difficult to
transfect. Formulations based on 29, 52b, and 54c were 2-6
times more efficient than the most effective commercial
formulation, DMRIE-C. The best results were obtained in this
1
1
the basis of signal intensity, selective decoupling, COSY ( H- H),
TOCSY, APT, HMQC, and HMBC sequences. Elemental analyses

Pyrylium Salts for Synthesis
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3881
were performed by combustion, using a Perkin-Elmer 2400 series
II CHNS analyzer.
Fatty acids, triethylamine, acetic anhydride, acetic acid, acyl
chlorides, and other solvents were from Acros (Fisher Scientific,
Pittsburgh, PA). Thin-layer chromatography (TLC) was performed
on silicagel 60-F254 plates (Merck, Whitehouse Station, NJ), eluted
4,6-Dimethyl-1-dodecyl-2-decylpyridinium hexafluorophos-
+
-
phate 16: T
c
) 56.7 °C; yield 71%. Anal. (C29
H
54N PF
6
) C, H,
N.
4,6-Dimethyl-1-decyl-2-undecylpyridinium hexafluorophos-
+
-
phate 17: T
c
) 52.2 °C; yield 69%. Anal. (C28
H
52N PF
6
) C, H,
N.
with MeOH/CHCl
General Procedure for the Synthesis of 4,6-Dimethyl-2-
alkylpyrylium Hexafluorophosphates 14 (Adapted from ref
3
20:80 (v/v).
4,6-Dimethyl-1-dodecyl-2-undecylpyridinium hexafluorophos-
+
-
phate 18: T
c
) 54.8 °C; yield 60%. Anal. (C30
H
56N PF
6
) C, H,
N.
4
,6-Dimethyl-1-decyl-2-dodecylpyridinium hexafluorophos-
19a). The fatty acid (0.1 mol) was refluxed with 45 mL of thionyl
+
-
phate 19: T
c
) 57.6 °C; yield 72%. Anal. (C29
H
54N PF
6
) C, H,
chloride for 90 min. Excess thionyl chloride was removed by
rotevaporation under reduced pressure; 20 mL of hexane was added
and then evaporated again under reduced pressure to remove the
traces of any remaining thionyl chloride. Anhydrous aluminum
chloride (13.5 g, 0.1 mol) was added to the crude fatty acid chloride
under stirring and external cooling (ice bath). The mixture became
homogeneous in about 10 min. Then 12 mL (0.15 mmol) of mesityl
oxide was added dropwise, and the reaction mixture was stirred at
room temperature for 12 h (fuming occurs because of HCl
generation). The resulting dark viscous mass was hydrolyzed by
carefully adding 50 mL of ice-cold 5% hydrochloric acid and 50
mL of diethyl ether, with external cooling (ice bath). The ethereal
layer was separated and extracted a second time with 20 mL of
diluted hydrochloric acid, then discarded. The combined aqueous
solution of pyrylium tetrachloroaluminate was extracted with 20
mL of diethyl ether to remove ketonic byproducts, separated, and
treated with 15 mL of hexafluorophosphoric acid when the crude
pyrylium hexafluorophosphate separated as a brown oil. After
extraction with chloroform, separation, drying on anhydrous sodium
sulfate, and evaporation of the solvent, the crude compound was
recrystallized twice from ethanol to afford the pure pyrylium
hexafluorophosphate.
N.
4,6-Dimethyl-1,2-didodecylpyridinium hexafluorophosphate
+
-
20: T
c
6
) 58.6 °C; yield 63%. Anal. (C31H58N PF ) C, H, N.
4,6-Dimethyl-1-decyl-2-tridecylpyridinium hexafluorophos-
+
-
phate 21: T
c
) 58.4 °C; yield 59%. Anal. (C30
H
56N PF
6
) C, H,
N.
4,6-Dimethyl-1-dodecyl-2-tridecylpyridinium hexafluorophos-
+
-
phate 22: T ) 67.2 °C; yield 62%. Anal. (C H N PF ) C, H,
c
32 60
6
N.
4,6-Dimethyl-1,2-ditetradecylpyridinium hexafluorophos-
+
-
phate 23: T ) 68.5 °C; yield 70%. Anal. (C H N PF ) C, H,
c
35 66
6
N.
4
,6-Dimethyl-1-tetradecyl-2-pentadecylpyridinium hexafluo-
+
-
c
6
rophosphate 24: T ) 74.7 °C; yield 61%. Anal. (C36H68N PF )
C, H, N.
4
,6-Dimethyl-1-hexadecyl-2-pentadecylpyridinium hexafluo-
+
-
rophosphate 25: T
C, H, N.
c
6
) 74.1 °C; yield 63%. Anal. (C38H72N PF )
4
,6-Dimethyl-1,2-dihexadecylpyridinium hexafluorophos-
+
-
phate 26: T
c
6
) 76.2 °C; yield 59%. Anal. (C39H74N PF ) C, H,
N.
4,6-Dimethyl-2-decylpyrylium hexafluorophosphate 14a: mp
4
,6-Dimethyl-1-hexadecyl-2-heptadecylpyridinium hexafluo-
+
-
67.6 °C; yield 27%. Anal. (C17
H
29O PF
6
) C, H.
+
-
rophosphate 27: T
C, H, N.
c
6
) 82.2 °C; yield 71%. Anal. (C40H76N PF )
4,6-Dimethyl-2-undecylpyrylium hexafluorophosphate 14b:
+
-
mp 62.8 °C; yield 21%. Anal. (C18
H31O PF
6
) C, H.
4,6-Dimethyl-1-octadecyl-2-heptadecylpyridinium hexafluo-
+
-
4
,6-Dimethyl-2-dodecylpyrylium hexafluorophosphate 14c:
c
6
rophosphate 28: T ) 82.1 °C; yield 65%. Anal. (C42H80N PF )
+
-
mp 78.7 °C; yield 31%. Anal. (C19
H33O PF
6
) C, H.
C, H, N.
General Procedure for the Synthesis of Pyridinium Gemini
Surfactants 29-35. 4,6-Dimethyl-2-tetradecylpyrylium hexafluo-
rophosphate (14e, 1 g, 2.22 mmol) was dissolved in 10 mL of
chloroform and treated with 0.75 mmol of diamine. Triethylamine
4
,6-Dimethyl-2-tridecylpyrylium hexafluorophosphate 14d:
+
-
mp 73.3 °C; yield 19%. Anal. (C20
H
35O PF
6
) C, H.
4,6-Dimethyl-2-tetradecylpyrylium hexafluorophosphate 14e:
+
-
mp 85.2 °C; yield 32%. Anal. (C21
H
37O PF
6
) C, H.
(0.2 mL) was then added to the brown-orange solution, which was
4,6-Dimethyl-2-pentadecylpyrylium hexafluorophosphate 14f:
subsequently refluxed for 5 min, treated with 0.5 mL of acetic acid,
and refluxed for another hour. The homogeneous mixture was
treated with 0.2 mL of concentrated aqueous ammonia, refluxed
for 5 min, cooled, and extracted with 10 mL of water to dissolve
the precipitated inorganic salts. The combined aqueous extracts were
back-extracted once with 10 mL of chloroform and discarded. The
chloroform extracts were combined, dried on sodium sulfate, and
evaporated to dryness to yield the crude product. Recrystallization
from ethanol, followed by flash chromatography on silica gel 60
+
-
mp 79.8 °C; yield 25%. Anal. (C22
H
39O PF
6
) C, H.
4,6-Dimethyl-2-hexadecylpyrylium hexafluorophosphate 14g:
+
-
mp 88.1 °C; yield 26%. Anal. (C23
H
41O PF
6
) C, H.
4,6-Dimethyl-2-heptadecylpyrylium hexafluorophosphate 14h:
+
-
6
mp 81.8 °C; yield 20%. Anal. (C24H43O PF ) C, H.
General Procedure for the Synthesis of 4,6-Dimethyl-1,2-
dialkylpyridinium Hexafluorophosphates 15-28. 4,6-Dimethyl-
-alkylpyrylium hexafluorophosphate 14 (3.5 mmol) was dissolved
in 15 mL of chloroform and treated with a solution of fatty
alkylamine (3.5 mmol) in 3 mL of chloroform. Triethylamine (0.5
mL) was then added to the brown-yellow solution, which was
subsequently refluxed for 5 min, treated with 1 mL of acetic acid,
and refluxed for another hour. The homogeneous mixture was
treated with 0.2 mL of concentrated aqueous ammonia to convert
any unreacted pyrylium salt into the corresponding highly soluble
pyridine, refluxed for 5 min, cooled, and extracted two times with
2
3
(MeOH/CHCl gradients 5/95 to 20/80 v/v as mobile phase)
afforded the pure compound, which was crystallized from methanol
or ethanol to yield the final gemini surfactant.
1,2-Bis(4,6-dimethyl-2-tetradecylpyridinium-1-yl)ethane di-
hexafluorophosphate 29: T ) 178.2 °C; yield 57%. Anal.
c
+
-
(C44
H
78
N
2
2PF
1,3-Bis(4,6-dimethyl-2-tetradecylpyridinium-1-yl)propane di-
hexafluorophosphate 30: T ) 186.6 °C; yield 53%. Anal.
6
) C, H, N.
c
+
-
1
0 mL of water to dissolve the precipitated inorganic salts. The
80 2 6
(C45H N 2PF ) C, H, N.
combined aqueous extracts were back-extracted once with 10 mL
of chloroform and discarded. The chloroform extracts were
combined, dried on sodium sulfate, and evaporated to dryness to
yield the crude product. Flash chromatography on silica gel 60,
1,4-Bis(4,6-dimethyl-2-tetradecylpyridinium-1-yl)butane di-
hexafluorophosphate 31: T ) 185.6 °C; yield 57%. Anal.
c
+
-
(C H N 2PF6 ) C, H, N.
4
6
82
2
1,5-Bis(4,6-dimethyl-2-tetradecylpyridinium-1-yl)pentane di-
using a 20/80 MeOH/CHCl
3
(v/v) mobile phase, afforded the pure
hexafluorophosphate 32: T ) 137.8 °C; yield 62%. Anal.
c
+
-
compound, which was crystallized from ethanol to yield the final
cationic lipid.
47 84 2
(C H N 2PF6 ) C, H, N.
1,6-Bis(4,6-dimethyl-2-tetradecylpyridinium-1-yl)hexane di-
4
,6-Dimethyl-1,2-didecylpyridinium hexafluorophosphate 15:
hexafluorophosphate 33: T ) 189.9 °C; yield 52%. Anal.
c
+
-
+
-
T
c
) 46.9 °C; yield 65%. Anal. (C27
H
50N PF
6
) C, H, N.
86 2 6
(C48H N 2PF ) C, H, N.

3
882 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Ilies et al.
1,7-Bis(4,6-dimethyl-2-tetradecylpyridinium-1-yl)heptane di-
aminoalkylcarbamate 39 or 45 (1 mmol) in CHCl
3
(5 mL) was
hexafluorophosphate 34: T ) 83.6 °C; yield 61%. Anal.
c
treated with 1.25 g of 4,6-dimethyl-2-tetradecylpyrylium hexafluo-
rophosphate 14e (2.5 mmol) dissolved in 5 mL of CHCl , under
+
-
(C
49
H
88
N
2
2PF
,8-Bis(4,6-dimethyl-2-tetradecylpyridinium-1-yl)octane di-
hexafluorophosphate 35: T ) 68.8 °C; yield 64%. Anal.
6
) C, H, N.
3
1
stirring, when a yellow-brown solution was formed. The homoge-
neous reaction mixture was refluxed for 1 h, cooled, and treated
with 0.3 mL of concentrated aqueous ammonia. After refluxing
for another 5 min, the mixture was allowed to reach room
temperature and was diluted with chloroform to a final volume of
25 mL. The solution was washed with water (10 mL), dried on
c
+
-
(C H
50 90
N
2
2PF
6
) C, H, N.
General Procedure for the Synthesis of Pyridinium Gemini
Surfactants 41 and 47. A. General Procedure for the Synthesis
of Trifluoroacetyl-Protected Polyamines 37 and 43 (after Ref
2
6). In a round-bottom flask 15 mmol of polyamine 36 or 42 were
Na
followed by flash chromatography on silica gel 60 (MeOH/CHCl
gradients 5/95 to 25/75 v/v as mobile phase) afforded the pure
compound, which was crystallized from ethanol to yield the final
gemini pyridinium carbamate.
2 4
SO , and evaporated to dryness. Recrystallization from ethanol,
dissolved in 10-20 mL of MeCN and treated dropwise, under
stirring, with ethyl trifluoroacetate (35 + 15k mmol, where k is
the number of secondary amino moieties) and water (18k mmol).
The reaction mixture was subsequently refluxed for 3-4 h (TLC
3
control, MeOH/CHCl
3
20/80 v/v) and then cooled, and most of
tert-Butyl-N,N-bis(2-(4,6-dimethyl-2-tetradecylpyridinium-1-
the solvent was removed under vacuum. Crystallization of the
desired product occurred upon cooling and was completed by
c
yl)ethylenyl) carbamate dihexafluorophosphate 40a: T ) 139.9
+
-
°C; yield 84%. Anal. (C51
tert-Butyl-N,N-bis(3-(4,6-dimethyl-2-tetradecylpyridinium-1-
yl)propylenyl) carbamate dihexafluorophosphate 40b: T
91 3 2 6
H N O 2PF ) C, H, N.
gradually adding CH
stirring. The solid was filtered, washed with cold CH
under vacuum to afford the pure trifluoroacetate salts.
2
Cl
2
(30-40 mL) to the mother liquor, under
2
Cl , and dried
2
c
)
+
-
95 3 2 6
134.4 °C; yield 86%. Anal. (C53H N O 2PF ) C, H, N.
1
7
N ,N -Bis(trifluoroacetyl)diethylenetriamine trifluoroacetate
tert-Butyl-N-(3-(4,6-dimethyl-2-tetradecylpyridinium-1-yl)-
propylenyl)-N-(4-(4,6-dimethyl-2-tetradecylpyridinium-1-yl)bu-
2
6a
3
3
9
7a: yield 93%; mp 89-90 °C; lit. mp 89-90 °C.
1 9
N ,N -Bis(trifluoroacetyl)dipropylenetriamine trifluoroacetate
7b: yield 97%; mp 120-121 °C; see ref 26i for the free base.
c
tylenyl) carbamate dihexafluorophosphate 40c: T ) 143.5 °C;
+
-
97 3 2 6
yield 79%. Anal. (C54H N O 2PF ) C, H, N.
1
8
N ,N -Bis(trifluoroacetyl)spermidine trifluoroacetate 37c: yield
N,N′-Di-tert-butyloxycarbonyl-N,N′-bis(3-(4,6-dimethyl-2-tet-
2%; mp 145-146 °C; lit.2 mp 144-146 °C; lit. mp 146-147
6f
26a
radecylpyridinium-1-yl)propyl)butanediamine dihexafluoro-
+
°
C.
phosphate 46: T
PF ) C, H, N.
6
c
) 103.3 °C; yield 81%. Anal. (C62H N O
112 4 4
2-
1
14
-
N ,N -Bis(trifluoroacetyl)spermine ditrifluoroacetate 43: yield
9
4%; mp 198-199 °C; lit.2 mp 195-198 °C; lit. mp 199-200
6a
26g
E. General Procedure for the Synthesis of Pyridinium
Polycationic Amphiphiles 41 and 47. The pyridinium gemini
carbamate 40 or 46 (0.5 mmol) was dissolved in methanol (20 mL)
and was treated dropwise, under stirring, with a solution obtained
by diluting 1 mL of 37% aqueous HCl with 4 mL of MeOH. The
mixture was stirred at room temperature until TLC showed complete
deprotection of the amino group (0.5-4 h). The solvent was
evaporated in a vacuum, and the residue was taken into 25 mL of
°
C.
B. General Procedure for the Boc Protection of the Secondary
Amino Moiety of Trifluoroacetamides 37 and 43 (after Ref 26).
The trifluoroacetate salt 37 or 43 from the previous step A (12
mmol) was dissolved/suspended in triethylamine (25 + 12k mmol,
where k is the number of secondary amino moieties) and treated
dropwise under argon, at 0 °C (ice bath), with a solution of (Boc)
12.6k mmol) in THF (3-8 mL). The cooling bath was removed,
and the homogeneous yellowish solution was stirred at room
temperature for 3-6 h (TLC control, MeOH/CHCl 20/80 v/v).
2
O
(
CH
NaPF
mL of CH
dried on Na
on silica gel 60 (NH
60 v/v as mobile phase) afforded the pure compound, which was
crystallized from Et O/EtOH mixtures to yield the final gemini
2
Cl
. The aqueous layer was separated and reextracted with 20
Cl , then discarded. The combined CH Cl extracts were
SO and evaporated to dryness. Flash chromatography
OH/MeOH/CH Cl gradients 5/15/80 to 5/35/
2
and extracted with 10% aqueous NaOH saturated with
6
3
2
2
2
2
Then the reaction was quenched with water (150-250 mL) and
the desired product was extracted with ethyl acetate (3 × 150 mL).
The combined organic extracts were washed with water (70 mL)
2
4
4
2
2
and dried on Na
2 4
SO . Evaporation of the solvent yielded the crude
2
compound as a yellow oil, which was crystallized from ether/hexane
mixtures to afford pure white crystals of Boc-protected trifluoro-
acetamides.
pyridinium amphiphile.
N,N-Bis(2-(4,6-dimethyl-2-tetradecylpyridinium-1-yl)ethylenyl)-
amine dihexafluorophosphate 41a: T ) 65.1 °C; yield 49%. Anal.
c
+
-
tert-Butyl-N,N-bis(2-trifluoroacetamidoethyl)carbamate 38a:
yield 97%; mp 122-123 °C; lit.26h mp 113-115 °C;
tert-Butyl-N,N-bis(3-trifluoroacetamidopropyl)carbamate 38b:
(C46
H
83
N
3
2PF
N,N-Bis(3-(4,6-dimethyl-2-tetradecylpyridinium-1-yl)propy-
lenyl)amine dihexafluorophosphate 41b: T ) 45.4 °C; yield 52%.
6
) C, H, N.
c
26d,26i
+
-
yield 90%; mp 67-68 °C.
87 3 6
Anal. (C48H N 2PF ) C, H, N.
tert-Butyl-N-(3-trifluoroacetamidopropyl)-N-(4-trifluoroaceta-
N-(3-(4,6-Dimethyl-2-tetradecylpyridinium-1-yl)propylen-
yl)-N-(4-(4,6-dimethyl-2-tetradecylpyridinium-1-yl)butylen-
26f
midobutyl)carbamate 38c: yield 98%; mp 74-75 °C; lit. mp
4 °C; lit.2 mp 74-75 °C.
6a
7
c
yl)amine dihexafluorophosphate 41c: T ) 67.2 °C; yield 47%.
+
-
N,N′-Di-tert-butyloxycarbonyl-N,N′-bis(3-trifluoroacetami-
89 3 6
Anal. (C49H N 2PF ) C, H, N.
2
6c
dopropyl)butanediamine 44: yield 94%; mp 107-108 °C; lit.
N,N′-Bis(3-(4,6-dimethyl-2-tetradecylpyridinium-1-yl)propyl)-
butanediamine dihexafluorophosphate 47: Tglass ) 9.6 °C; yield
mp 93.5-95 °C; lit.26g mp 104-105 °C.
+
-
C. General Procedure for the Selective Deprotection of the
Primary Amino Moiety of Trifluoroacetamidoalkyl Carbamates
96 4 6
45%. Anal. (C52H N 2PF ) C, H, N.
General Procedure for the Synthesis of Dendritic Pyridinium
Amphiphiles 48. The pyridinium gemini amphiphile N,N′-bis(3-
(4,6-dimethyl-2-tetradecylpyridinium-1-yl)propyl)butanediamine di-
hexafluorophosphate 47 (0.32 g, 0.3 mmol) was dissolved in MeCN
3
9 and 45 (after Ref 26). The carbamate 38 or 44 from step B (10
mmol) was dissolved in methanol (120 mL), cooled to 5-10 °C
ice bath), and treated dropwise with a methanolic 0.2 M NaOH
(
solution (130 mL). When addition was complete, the cooling bath
was removed and the mixture was stirred at 22 °C overnight. The
methanol was removed in a vacuum, and the oily solution was
3
(10 mL) together with 0.1 g (1 mmol) of NEt . The solution was
cooled to 0 °C (ice bath) and was treated dropwise, under stirring,
with a solution obtained by diluting 0.9 mmol of myristoyl or
palmitoyl chloride with 2 mL of MeCN. The reaction mixture was
allowed to reach room temperature, and the stirring was continued
until TLC showed complete acylation (2 h). The solvent was
extracted with a mixture of MeOH/CHCl
The combined organic extracts were dried over Na
concentrated to dryness to afford the pure bis-aminoalkylcarbamate
3
(1/9 v/v, 3 × 50 mL).
2
SO and
4
3
9 or 45 (TLC control) as a pale-yellow oil (yields 93-97%), which
rotevaporated. The residue was taken into 25 mL of CH
extracted with 10 mL of 10% aqueous NaOH saturated with NaPF
and the organic layer was dried on Na SO . Evaporation of the
solvent afforded the crude product, which was crystallized from
2 2
Cl and
was used directly into the next step.
D. General Procedure for the Synthesis of Lipophilic Pyri-
dinium Gemini Carbamates 40 and 46. A solution of bis-
6
,
2
4

Pyrylium Salts for Synthesis
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3883
EtOH (5 mL). Flash chromatography on silica gel 60 (MeOH/CHCl
gradient 5/95 to 20/80 v/v as mobile phase) afforded the pure
3
gradients 10/90 to 30/70 v/v as mobile phase) afforded the pure
compound, which was crystallized from Et O/EtOH to yield the
2
compound, which was crystallized from Et
2
O/EtOH mixtures to
final hydroxyethylpyridinium salt in 86% yield.
yield the final dendritic pyridinium amphiphile.
N,N′-Bis(3-(4,6-dimethyl-2-tetradecylpyridinium-1-yl)propy-
lenyl)butane-N,N′-bismyristoylamide dihexafluorophosphate 48a
4,6-Dimethyl-2-tetradecyl-1-(2-hydroxyethyl)pyridiniumhexaflu-
orophosphate 51: mp 55.6 °C. Anal. (C H NO PF ) C, H, N.
23 42 6
B. Acylation of Hydroxyethylpyridinium Hexafluorophos-
phate 51 with Fatty Acid Chlorides. The hydroxyethylpyridinium
hexafluorophosphate 51 (0.4 g, 0.8 mmol) was dissolved in 10 mL
of MeCN with 0.84 mmol of n-alkyl acid chloride. Triethylamine
+
-
+
-
(
n ) 12): T
c
148 4 6
) 125.8 °C; yield 55%. Anal. (C80H N 2PF ) C,
H, N.
N,N′-Bis(3-(4,6-dimethyl-2-tetradecylpyridinium-1-yl)propy-
lenyl)butane-N,N′-bispamitoylamide dihexafluorophosphate 48b
(91 mg, 0.9 mmol) was added to the reaction mixture as a solution
+
-
(
n ) 14): T
c
156 4 6
) 124.9 °C; yield 62%. Anal. (C84H N 2PF ) C,
in 3 mL of acetonitrile, while stirring under external cooling (ice
bath). The cooling bath was then removed, and the resultant
suspension was subsequently refluxed until TLC showed complete
acylation (6-10 h). The solvent was removed in a vacuum, and
H, N.
General Procedure for the Synthesis of Dendritic Pyridinium
Amphiphile 49. The pyridinium gemini amphiphile N,N′-bis(3-
(
4,6-dimethyl-2-tetradecylpyridinium-1-yl)propyl)butanediamine di-
hexafluorophosphate 47 (0.32 g, 0.3 mmol) was dissolved in MeCN
10 mL) with 0.3 g (1 mmol) of 1-bromohexadecane. The mixture
the residue was partitioned between CH
aqueous NaPF (25 mL). The aqueous layer was back-extracted
with 25 mL of CH Cl and discarded. The methylene chloride
2 2
Cl (25 mL) and saturated
6
(
2
2
was refluxed under inert atmosphere until TLC showed complete
alkylation (10 h). The solvent was removed by rotevaporation. The
extracts were combined, dried on sodium sulfate, and evaporated
to dryness to yield the crude product. Flash chromatography on
silica gel 60 (MeOH/CHCl gradients 5/95 to 25/75 v/v as mobile
3
phase) afforded the pure compound, which was crystallized from
EtOH to yield the final pyridinium cationic lipid.
residue was taken into 25 mL of CH
0% aqueous NaOH saturated with NaPF
Na CO . Evaporation of the solvent afforded the crude product,
which was crystallized from EtOH (5 mL). Flash chromatography
on silica gel 60 (NH OH/MeOH/CH Cl gradients 5/15/80 to 5/35/
0 v/v/v as mobile phase) afforded the pure compound, which was
2
Cl
2
, extracted with 10 mL of
1
6
, and dried on anhydrous
2
3
4
,6-Dimethyl-2-tetradecyl-1-(2-dodecanoyloxyethyl)pyridin-
4
2
2
ium hexafluorophosphate 52a: T ) 61.3 °C; yield 87%. Anal.
c
6
+
-
(
C
35
H
64NO
2
PF
,6-Dimethyl-2-tetradecyl-1-(2-tetradecanoyloxyethyl)pyri-
dinium hexafluorophosphate 52b: T ) 68.5 °C; yield 90%. Anal.
PF
4,6-Dimethyl-2-tetradecyl-1-(2-hexadecanoyloxyethyl)pyri-
dinium hexafluorophosphate 52c: T ) 77.8 °C; yield 91%. Anal.
6
) C, H, N.
2
crystallized from Et O/EtOH mixtures to yield the final dendritic
pyridinium amphiphile.
4
c
N,N′-Bis(3-(4,6-dimethyl-2-tetradecylpyridinium-1-yl)propy-
lenyl)butane-N,N′-bispalmitylamine dihexafluorophosphate 49:
+
-
(C
37
H68NO
2
6
) C, H, N.
+
-
T
c
) 69.4 °C; yield 65%. Anal. (C84
160 4 6
H N 2PF ) C, H, N.
c
General Procedure for the Synthesis of Dendritic Tris-
+
-
(C
39
H72NO
2
PF
6
) C, H, N.
Pyridinium Amphiphile 50. 4,6-Dimethyl-2-tetradecylpyrylium
hexafluorophosphate (14e, 0.9 g, 2 mmol) was dissolved in 10 mL
General Procedure for the Synthesis of Pyridinium Amide
Cationic Lipids 54. A. Synthesis of Aminoethylpyridinium
Hexafluorophosphate 53. 4,6-Dimethyl-2-tetradecylpyrylium
hexafluorophosphate (14e, 2.25 g, 5 mmol) was dissolved in 50
of CH
2
Cl
2
and treated with a solution obtained by dissolving 73
Cl . The
mg (0.5 mmol) of tris(2-aminoethyl)amine in 2 mL of CH
2
2
resulting homogeneous brown-orange solution was subsequently
refluxed for 5 min, treated with 0.5 mL of acetic acid, and refluxed
for another hour. The reaction mixture was cooled to 22 °C, treated
with 0.2 mL of concentrated aqueous ammonia, and refluxed for
mL of CH
by dissolving 0.3 g (5 mmol) of ethylenediamine in 20 mL of CH
Cl . Triethylamine (0.1 mL) was added to the reaction mixture,
2 2
Cl and treated under stirring with a solution obtained
2
-
2
and the resultant homogeneous orange-brownish solution was
subsequently refluxed for 5 min, treated with 1 mL of acetic acid,
and refluxed for another hour. The reaction mixture was cooled to
room temperature, treated with 1 mL of concentrated aqueous
ammonia, and heated to reflux for another 5 min. After cooling, it
was extracted with water (2 × 25 mL) in order to dissolve the
precipitated inorganic salts. The combined aqueous extracts were
back-extracted once with 20 mL of CH Cl and discarded. The
2 2
methylene chloride extracts were combined, dried on sodium sulfate,
and evaporated to dryness to yield the crude product. Flash
2 2
another 5 min. After cooling, it was diluted with 20 mL of CH Cl
and extracted with 10 mL of water to dissolve the precipitated
inorganic salts. The combined aqueous extracts were back-extracted
once with 10 mL of CH Cl and discarded. The methylene chloride
2 2
extracts were combined, dried on sodium sulfate, and evaporated
to dryness to yield the crude product, which was crystallized from
ethanol. Flash chromatography on silica gel 60 (MeOH/CHCl
gradients 5/95 to 30/70 v/v as mobile phase) afforded the pure
compound, which was crystallized from Et O/EtOH (1/2 v/v
3
2
mixture) to yield the final crystalline amphiphilic trispyridinium
salt.
Tris(2-(4,6-dimethyl-2-tetradecylpyridinium-1-yl)ethyl)am-
chromatography on silica gel 60 (NH
/20/75 to 5/40/55 v/v/v as mobile phase) afforded the pure
compound, which was crystallized from Et O/EtOH to yield the
4 2 2
OH/MeOH/CH Cl gradients
5
ine trishexafluorophosphate 50: T ) 121.2 °C; yield 57%. Anal.
c
2
+
-
final aminoethylpyridinium salt in 53% yield.
(C H N
69 123 4
3PF
6
) C, H, N.
General Procedure for the Synthesis of Pyridinium-Ester
4,6-Dimethyl-2-tetradecyl-1-(2-aminoethyl)pyridinium hexaflu-
+
-
Cationic Lipids 52. A. Synthesis of Hydroxyethylpyridinium
Hexafluorophosphate 51. 4,6-Dimethyl-2-tetradecylpyrylium
hexafluorophosphate (14e, 1.98 g, 4.4 mmol) was dissolved in 20
43 2 6
orophosphate 53: mp 61.4 °C. Anal. (C23H N PF ) C, H, N.
B. Acylation of Aminoethylpyridinium Hexafluorophosphate
53 with Fatty Acid Chlorides. The aminoethylpyridinium hexaflu-
orophosphate 53 (0.2 g, 0.4 mmol) was dissolved in 10 mL of
MeCN with 50 mg (0.5 mmol) of triethylamine and treated
dropwise, while stirring under external cooling (ice bath), with 0.44
mmol of n-alkyl acid chloride dissolved in 3 mL of MeCN. The
cooling bath was removed, and the resulting suspension was
refluxed until TLC showed complete acylation (3-6 h). The solvent
was removed in vacuo, and the residue was partitioned between
mL of CH
by dissolving 0.24 g (4 mmol) of 2-aminoethanol in 15 mL of CH
Cl . Triethylamine (0.1 mL) was added to the reaction mixture,
2 2
Cl and treated, while stirring, with a solution obtained
2
-
2
and the resultant homogeneous brown-orange solution was subse-
quently refluxed for 5 min, treated with 1 mL of acetic acid, and
refluxed for another hour. The reaction mixture was cooled to room
temperature, treated with 1 mL of concentrated aqueous ammonia,
and refluxed for another 5 min. After cooling, it was diluted with
CH
2
Cl
2
(25 mL) and saturated aqueous NaPF
6
(25 mL). The
Cl and
6
5 mL of CH
to dissolve the precipitated inorganic salts. The combined aqueous
extracts were back-extracted once with 20 mL of CH Cl and
2
Cl
2
and extracted with water (2 × 25 mL) in order
aqueous layer was back-extracted with 25 mL of CH
2
2
discarded. The methylene chloride extracts were combined, dried
on sodium sulfate, and evaporated to dryness to yield the crude
2
2
discarded. The methylene chloride extracts were combined, dried
on sodium sulfate, and evaporated to dryness to yield the crude
product. Flash chromatography on silica gel 60 (MeOH/CHCl
3
gradients 20/80 to 30/70 v/v as mobile phase) afforded the pure
compound, which was crystallized from ethanol to yield the final
product. Flash chromatography on silica gel 60 (MeOH/CHCl
3

3
884 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Ilies et al.
pyridinium cationic lipid in 75-85% yield. Although the NMR
4,6-Dimethyl-2-tetradecyl-1-(3,4-dithiaoctadecyl)pyridini-
analysis of the resultant lipids showed pure compounds, elemental
c
um hexafluorophosphate 56c: T ) 81.8 °C; yield 65%. Anal.
-
-
+
-
analysis showed a mixture of counterions (Cl and PF
6
), which
(C37
H
70NS
2
PF
6
) C, H, N. Chloride: T
c
) 86.9 °C; yield 78%
+
-
was confirmed by DSC. Therefore, the compounds were dissolved
(from hexafluorophosphate). Anal. (C37
H
2
70NS Cl ) C, H, N.
in a minimum amount of 96% EtOH and passed through a DOWEX
C. Oxidative Coupling of Mercaptoethylpyridinium Hexaflu-
orophosphate 55. A solution of mercaptoethylpyridinium hexafluo-
rophosphate 55 (0.25 g, 0.5 mmol) in 15 mL of EtOH/CH Cl , 2/1,
was cooled to 0 °C (ice bath) and treated dropwise, under nitrogen
atmosphere and with vigorous stirring, with a solution of iodine
(0.076 g, 0.3 mmol) in 3 mL of EtOH until the yellow-brown color
of free iodine was persistent. The solvent was removed in a vacuum,
-
1X8-400 anion exchange column (Cl form), packed in the same
solvent. Elution with 96% EtOH afforded the pure compounds as
chlorides, which were concentrated and crystallized from EtOH.
2
2
4
,6-Dimethyl-2-tetradecyl-1-(2-dodecanoylamidoethyl)pyri-
dinium chloride 54a: T ) 41.9 °C; yield 51%. Anal. (C35H N -
c
65 2
+
-
O Cl ) C, H, N.
,6-Dimethyl-2-tetradecyl-1-(2-tetradecanoylamidoethyl)py-
ridinium chloride 54b: T ) 49.9 °C; yield 55%. Anal.
and the residue was partitioned between CH
saturated aqueous NaHCO (25 mL). The aqueous layer was back-
extracted with 25 mL of CH Cl and discarded. The methylene
chloride extracts were combined and treated with 20% aqueous
NaHSO (25 mL), dried on Na SO , and evaporated to dryness to
yield the crude product. Flash chromatography on silica gel 60
MeOH/CHCl gradients 3/97 to 25/75 v/v as mobile phase)
afforded the pure compound, which was crystallized from Et O/
EtOH to yield the final disulfidic pyridinium cationic lipid. The
2 2
Cl (25 mL) and
4
3
c
+
-
2
2
(
C
37
H
69
N
2
O Cl ) C, H, N.
,6-Dimethyl-2-tetradecyl-1-(2-hexadecanoylamidoethyl)py-
ridinium chloride 54c: T ) 46.8 °C; yield 48%. Anal.
4
3
2
4
c
+
-
(
C
39
H
73
N
2
O Cl ) C, H, N.
(
3
General Procedure for the Synthesis of Pyridinium Disulfide
2
Cationic Lipids 56. A. Synthesis of Mercaptoethylpyridinium
Hexafluorophosphate 55. Aminothioethanol hydrochloride (0.34
g, 3 mmol) was added, while stirring, to a solution of 0.2 g of
EtONa in 10 mL of absolute ethanol, followed by 1.42 g of 4,6-
dimethyl-2-tetradecylpyrylium hexafluorophosphate (14e, 3.15 mmol)
dissolved in ethanol (10 mL). The resulting homogeneous orange-
brownish solution was refluxed for 10 min, treated with 0.4 mL of
acetic acid, and refluxed for another hour. The reaction mixture
was cooled to room temperature, and 0.25 mL of concentrated
aqueous ammonia was added. After the mixture was refluxed for
another 5 min, the solvent was removed by rotevaporation, and
-
-
PF
as for the preparation of amides 54, the pure chloride being
crystallized from Et O/EtOH.
,6-Bis(4,6-dimethyl-2-tetradecylpyridinium-1-yl)-3,4-dithia-
hexane dihexafluorophosphate hexafluorophosphate 57: T
6
counterion can be exchanged to Cl using the same procedure
2
1
c
)
+
-
1
07.0 °C; yield 71%. Anal. (C46
H
82
2
N S
2
6
PF ) C, H, N. Chloride:
T
c
) 139.3 °C; yield 79% (from hexafluorophosphate). Anal.
+
-
(C
46
H
82
N
2
S
2
Cl ) C, H, N.
Preparation and Evaluation of Liposomes. Dioleoylphosphati-
dylethanolamine (DOPE), cholesterol (Chol), and N-(2,3-dioleoy-
loxypropyl)-N,N,N-trimethylammonium chloride (DOTAP) were
purchased from Sigma (St. Louis, MO). The purity of the lipids
was checked by TLC and also by differential scanning calorimetry
using a TA Instruments Q100 DSC instrument (New Castle, DE).
For the preparation of liposomes a protocol optimized for this
type of compound1 was used. The cationic amphiphile (with or
without helper lipid) was dissolved in chloroform/methanol (2/1,
v/v) to reach a final concentration of 1 mM for the charged species.
From this stock solution, a volume of 50 µL was transferred to a
the residue was partitioned between CH
aqueous NaPF (25 mL). The aqueous layer was back-extracted
with 25 mL of CH Cl and discarded. The methylene chloride
extracts were combined, dried on sodium sulfate, and evaporated
to dryness to yield the crude product. Flash chromatography on
2 2
Cl (25 mL) and saturated
6
2
2
silica gel 60 (MeOH/CHCl
phase) afforded the pure compound, which was crystallized from
Et O/EtOH to yield the final mercaptoethylpyridinium salt in 66%
yield.
3
gradients 3/97 to 20/80 v/v as mobile
4b,c
2
4
,6-Dimethyl-2-tetradecyl-1-(2-mercaptoethyl)pyridinium hexa-
1
mL microcentrifuge tube (USA Scientific, catalog no. 1415-2500),
+
-
6
fluorophosphate 55: mp 71.9 °C. Anal. (C23H42NS PF ) C, H,
and the solvent was evaporated in a SpeedVac evaporator under
vacuum for 1 h. Traces of organic solvent were removed by drying
the lipid film in a vacuum desiccator for 3 h over Drierite. The
dried film was hydrated overnight after adding 500 µL of sterile
phosphate buffer isotonic saline at pH 7.4 (PBS) (GibcoBRL/
Invitrogen, Carlsbad, CA), purging the tube with sterile nitrogen,
and vortex-mixing it for 1 min. Liposomes were generated the next
day by sonicating the tube in a water bath sonicator (Branson, model
N.
B. Oxidative Coupling of Mercaptoethylpyridinium Hexaflu-
orophosphate 55 with n-Alkyl Thiols. The mercaptoethylpyri-
dinium hexafluorophosphate 55 (0.25 g, 0.5 mmol) was suspended
2 2
in 20 mL of EtOH with 3 mmol of n-alkylthiol, and CH Cl was
added until all the solids were dissolved (4-8 mL). The solution
was cooled to 0 °C (ice bath), and a solution of iodine (0.46 g, 1.8
mmol) in 10 mL of EtOH was added dropwise, under nitrogen
atmosphere and with vigorous stirring, until the yellow-brown color
of free iodine was persistent. The solvent was removed in vacuo,
1
210) for 30 min (two reprises of 15 min sonication, with 15 min
pause between) at 65-68°C for formulations based on lipids 15-
5, 40, 41, 46-50, 52, 54, 56, and 57 in a mixture with cholesterol
3
and the residue was partitioned between CH
saturated aqueous NaHCO (25 mL). The aqueous layer was back-
extracted with 25 mL of CH Cl and discarded. The methylene
chloride extracts were combined and treated with 20% aqueous
NaHSO (25 mL), dried on Na SO , and evaporated to dryness to
2 2
Cl (25 mL) and
or DOPE and at 37 °C for DOTAP/cholesterol. After sonication
the preparations were allowed to reach 22 °C and assessed on the
specified cells within 30 min from the preparation. Lipofectamine
and DMRIE liposomes were prepared according to the manufac-
turer’s specifications.
The size of the lipoplexes was determined in parallel by dynamic
light scattering, using a DynaPro particle sizer (Protein Solutions,
Piscataway, NJ) and employing the Dynamics data collection and
analysis software. In each experiment the liposomal preparation
3
2
2
3
2
4
yield the crude product. Trituration with hexane followed by flash
chromatography on silica gel 60 (petroleum ether/acetone gradients
3
0/70 to 0/100 v/v as mobile phase) afforded the pure compound,
which was crystallized from Et O/EtOH to yield the final disulfidic
pyridinium cationic lipid. The PF
2
-
6
counterion can be exchanged
to Cl using the same procedure as for the preparation of amides
4, the pure chlorides being crystallized from Et O/EtOH.
,6-Dimethyl-2-tetradecyl-1-(3,4-dithiatetradecyl)pyridini-
um hexafluorophosphate 56a: T ) 76.0 °C; yield 58%. Anal.
(20 µL) was diluted with PBS to a final volume of 200 µL. From
-
this solution an aliquot (15 µL) was introduced in a quartz cuvette
precleaned with distilled water and methanol and dried to eliminate
any contaminants. The cuvette was sealed with Parafilm, and the
mean average diameter of the vesicles was determined at 25 °C as
the result of at least three reliable readings.
5
2
4
c
+
-
(
(
C
33
H
62NS
2
PF
6
) C, H, N. Chloride: T
c
) 77.4 °C; yield 82%
+
-
from hexafluorophosphate). Anal. (C33
H62NS
2
Cl ) C, H, N.
The ú potentials of the liposomes and lipoplexes were determined
using a Zetasizer 2000 instrument (Malvern Instruments, U.K.)
using the Zeta2000 software. The liposomal/lipoplex formulation
(200 µL) was diluted with PBS to a final volume of 2 mL and
injected into the measuring cuvette of the instrument. Calibration
4
,6-Dimethyl-2-tetradecyl-1-(3,4-dithiahexadecyl)pyridini-
um hexafluorophosphate 56b: T ) 81.0 °C; yield 63%. Anal.
c
+
-
(
(
C
35
H
66NS
2
PF
6
) C, H, N. Chloride: T
c
) 75.1 °C; yield 76%
+
-
from hexafluorophosphate). Anal. (C35
2
H66NS Cl ) C, H, N.

Pyrylium Salts for Synthesis
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3885
was performed using the DTS5050 carboxylated polystyrene latex.
The final value for the ú potential of the particles was the average
of at least three separate readings.
Preparation of Lipoplexes. The plasmid used in this study was
pGL3 (Promega, Madison, WI), encoding a firefly luciferase gene
under control of the constitutively active SV40 promoter. The
plasmid was amplified in Escherichia coli and purified using a
Genopure Plasmid MaxiKit (Roche, Indianapolis, IN).
after transfection, 20 µL of WST-1 tetrazolium dye solution (Roche,
Mannheim, Germany) was added to each well (still containing the
serum and the liposomal preparation). A blank was prepared by
mixing 100 µL of Optimem, 100 µL of serum, and 20 µL of
tetrazolium dye solution, and the plate was incubated at 37 °C in
2
the CO incubator. After 3 h the colorimetric measurement was
performed at 450 nm (with a reference wavelength of 595 nm that
was subtracted) by means of a Vmax kinetic microplate reader
(Molecular Devices, Sunnyvale, CA). The value corresponding to
the blank was deducted from the value corresponding to each well.
For generation of the the lipoplexes, 2.8 µL of plasmid DNA
solution (0.5 µg/µL) was diluted with 100 µL of Optimem (Gibco/
Invitrogen) to make plasmid DNA stock solutions. A liposome
solution was prepared separately by diluting 40 µL of the initial
liposomal stock solution (0.1 mM in cationic lipid) with Optimem
in order to reach a final volume of 50 µL. To this liposome solution,
Acknowledgment. Financial support from the Welch Foun-
dation and NCI Grant CA 041407 (to E.B.T.) is acknowledged.
The authors thank Dr. Wayne Bolen and Dr. Massoud Motamedi
for their help with the sizing and ú potential measurement
experiments. We acknowledge Dr. Nancy Smith Templeton and
Dr. Malcolm Brenner for interesting discussions and suggestions
and Dr. David Gorenstein for his support, and the Sealy Center
for Structural Biology for NMR Support.
50 µL of the DNA stock solution was added, and (after mixing)
the tube was incubated for 30 min at 22 °C. The content was then
diluted with 500 µL of Optimem (final volume of the lipoplex stock
solution was 600 µL).
Cell Transfection. The lipoplexes were tested for their ability
to transfect five cancer cell lines: a lung carcinoma (NCI-H23,
found to be the most sensitive to pyridinium-based transfection
formulations), two breast carcinomas (MCF-7 and MDA-231â), a
prostate carcinoma (DU-145), and a brain glioma (SWB-95). The
cells were maintained in 10% fetal bovine serum (FBS) enriched
Supporting Information Available: Analytical data and results
from elemental analysis. This material is available free of charge
via the Internet at http://pubs.acs.org.
References
2
medium at 37 °C in a humidified atmosphere of 95% air/5% CO .
The following media were used: RPMI 1640 (CellGro, Houston,
TX) for NCI-H23, Iscove’s modified Dulbecco’s medium (Gibco/
Invitrogen) for MCF-7, Dulbecco’s modified Eagle medium (Gibco/
Invitrogen) for MDA-231â, minimum essential medium (Gibco/
Invitrogen) for DU-145, and minimum essential medium Eagle R
modification (Sigma) for the SWB-95 cell line. Twenty-four hours
prior to transfection, the cells were transferred to 96-well microtiter
plates (Costar, catalog no. 3596) at a density of 20 000 cells/well.
Each well received 100 µL of appropriate medium, and the plate
was incubated in the same conditions as above. All experiments
were done in triplicate or quadruplicate. The error bars in figures
represent one standard deviation from the average value.
Immediately before transfection the medium was removed, and
the cells from each well were briefly washed with 200 µL of sterile
PBS. After removal of the PBS solution, each well received 100
µL of lipoplex stock solution, and the plate was returned to the
incubator for 1 h. An additional 100 µL of medium (20% FBS)
was added to each well, and the plate was incubated for a further
(1) (a) Templeton, N. S., Lasic, D. D., Eds. Gene Therapy. Therapeutical
Mechanisms and Strategies; Marcel Dekker: New York, 2000. (b)
Lemoine, N. R.; Cooper, D. N. Gene Therapy. In Human Molecular
Genetics; Cooper, D. N., Humphries, S. E., Strachan, T., Eds.; BIOS
Scientific Publishers: Oxford, U.K., 1996. (c) Smith, K. R. Gene
therapy: theoretical and bioethical concepts. Arch. Med. Res. 2003,
3
4, 247-268. (d) Smith, K. R. Gene transfer in higher animals:
theoretical considerations and key concepts. J. Biotechnol. 2002, 99,
1-22. (e) Friedmann, T. Gene therapy: delivering the medicines of
the 21st century. Nat. Med. 1996, 2, 144-147.
(
2) (a) Greco, O.; Scott, S. D.; Marples, B.; Dachs, G. U. Cancer gene
therapy: “delivery, delivery, delivery”. Front. Biosci. 2002, 7,
d1516-d1524. (b) Templeton, N. S. Developments in liposomal gene
delivery systems. Expert Opin. Biol. Ther. 2001, 1, 567-570.
(3) (a) Kay, M. A.; Glorioso, J. C.; Naldini, L. Viral vectors for gene
therapy: the art of turning infectious agents into vehicles of
therapeutics. Nat. Med. 2001, 7, 33-40. (b) Walther, W.; Stein, U.
Viral vectors for gene transfer: a review of their use in the treatment
of human diseases. Drugs 2000, 60, 249-271. (c) Strayer, D. S. Viral
gene delivery. Expert Opin. InVest. Drugs 1999, 8, 2159-2172. (d)
Robbins, P. D.; Ghivizzani, S. C. Viral vectors for gene therapy.
Pharmacol. Ther. 1998, 80, 35-47.
2
4 h, after which the luciferase content and protein content were
assessed.
Luciferase and Protein Content Assay. Twenty-four hours after
(
4) (a) Tseng, W.-C.; Huang, L. Liposome-based gene therapy. Pharm.
Sci. Technol. Today 1998, 1, 206-213. (b) Nishikawa, M.; Huang,
L. Nonviral vectors in the new millennium: delivery barriers in gene
transfer. Hum. Gene Ther. 2001, 12, 861-870. (c) Pedroso de Lima,
M. C.; Simoes, S.; Pires, P.; Faneca, H.; Duzgunes, N. Cationic lipid-
DNA complexes in gene delivery: from biophysics to biological
applications. AdV. Drug DeliVery ReV. 2001, 47, 277-294. (d) Read,
M. L.; Logan, A.; Seymour, L. W. Barriers to gene delivery using
synthetic vectors AdV. Genet. 2005, 53, 19-46.
transfection, the medium was aspirated and the wells were washed
briefly with 200 µL of PBS. After removal of PBS, the cells were
lysed by adding 100 µL of 1X reporter lysis buffer (Promega) to
each well and incubating the plate at 37 °C for 10 min. The cell
lysate was collected and used for luciferase and protein assays.
For the luciferase assay, 20 µL of cell lysate was transferred to
a test tube and assessed directly by means of a Monolight 2010
luminometer (Analytical Luminescence Laboratory, San Diego, CA)
using a luciferase assay kit from Promega.
(
5) (a) Petry, H.; Goldmann, C.; Ast, O.; Luke, W. The use of virus-like
particles for gene transfer. Curr. Opin. Mol. Ther. 2003, 5, 524-
528. (b) Mountain, A. Gene therapy: the first decade. Trends
Biotechnol. 2000, 18, 119-128. (c) Rochlitz, C. F. Gene therapy of
cancer. Swiss Med. Wkly. 2001, 131, 4-9.
The protein content was quantified using a bicinchoninic acid
(
6) (a) Miller, A. D. Cationic liposomes for gene therapy. Angew. Chem.,
Int. Ed. 1998, 37, 1768-1785. (b) Ilies, M. A.; Seitz, W. A.; Balaban,
A. T. Cationic lipids in gene delivery: principles, vector design and
therapeutical applications. Curr. Pharm. Des. 2002, 8, 2441-2473.
(c) Ilies, M. A.; Balaban, A. T. Recent developments in cationic lipid-
mediated gene delivery and gene therapy. Expert Opin. Ther. Pat.
(BCA) assay (Pierce, Rockford, IL). The BCA assay was prepared
as specified in its manufacturer’s instructions. An amount of 40
µL of cell lysate was treated with 1 mL of BCA reagent in an
acryl cuvette, and the solution was incubated for 1 h at 37 °C. The
light absorption of the solution was then read at 562 nm by means
of a Beckman DU-600 UV-vis spectrometer (Palo Alto, CA), and
the protein content was estimated by comparison to bovine serum
albumin standards. The luciferase activity was normalized by the
protein content and expressed as relative luminescence units/µg of
protein (RLU/µg protein).
2001, 11, 1729-1752. (d) Niculescu-Duvaz, D.; Heyes, J.; Springer,
C. J. Structure-activity relationship in cationic lipid mediated gene
transfection. Curr. Med. Chem. 2003, 10, 1233-1261. (e) Kostarelos,
K.; Miller, A. D. What Role can chemistry play in cationic liposome-
based gene therapy research today? AdV. Genet. 2005, 53, 71-118.
(7) (a) Kabanov, A. V.; Lemieux, P.; Vinogradov, S.; Alakhov, V.
Pluronic block copolymers: novel functional molecules for gene
therapy. AdV. Drug DeliVery ReV. 2002, 54, 223-233. (b) Davis,
M. E. Non-viral gene delivery systems. Curr. Opin. Biotechnol. 2002,
Viability Assay. To quantify the relative cytotoxicity of the
29
nonviral cationic vectors, a WST-1 standard viability method was
performed along with the luciferase and BCA assays. Five wells
of cells for each liposomal preparation were set up in parallel and
transfected identically as in the pGL3 assay. Twenty-four hours
1
3, 128-131. (c) Hwang, S. J.; Davis, M. E. Cationic polymers for
gene delivery: designs for overcoming barriers to systemic admin-
istration. Curr. Opin. Mol. Ther. 2001, 3, 183-191. (d) De Smedt,

3
886 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13
Ilies et al.
S. C.; Demeester, J.; Hennink, W. E. Cationic polymer based gene
delivery systems. Pharm. Res. 2000, 17, 113-126. (e) Kabanov, A.;
Zhu, J.; Alakhov, V. Pluronic block copolymers for gene delivery.
AdV. Genet. 2005, 53, 231-261.
47, 3744-3754. (c) Ilies, M. A.; Johnson, B. H.; Makori, F.; Miller,
A.; Seitz, W. A.; Thompson, E. B.; Balaban, A. T. Pyridinium
cationic lipids in gene delivery: an in vitro and in vivo comparison
of transfection efficiency versus a tetraalkylammonium congener.
Arch. Biochem. Biophys. 2005, 435, 217-226.
(
8) (a) Li, S.; Huang, L. Nonviral gene therapy: promises and challenges.
Gene Ther. 2000, 7, 31-34. (b) Felgner, P. L. Prospects for synthetic
self-assembling systems in gene delivery. J. Gene Med. 1999, 1, 290-
(
15) (a) Menger, F. M.; Keiper, J. S. Gemini surfactants. Angew. Chem.,
Int. Ed. 2000, 39, 1906-1920. (b) Kirby, A. J.; Camilleri, P.;
Engberts, J. B.; Feiters, M. C.; Nolte, R. J.; Soderman, O.; Bergsma,
M.; Bell, P. C.; Fielden, M. L.; Garcia Rodriguez, C. L.; Guedat, P.;
Kremer, A.; McGregor, C.; Perrin, C.; Ronsin, G.; van Eijk, M. C.
Gemini surfactants: new synthetic vectors for gene transfection.
Angew. Chem., Int. Ed. 2003, 42, 1448-1457.
292. (c) Kostarelos, K.; Miller, A. D. Synthetic, self-assembly ABCD
nanoparticles; a structural paradigm for viable synthetic non-viral
vectors. Chem. Soc. ReV. 2005, 34, 970-994. (d) Martin, B.; Sainlos,
M.; Aissaoui, A.; Oudrhiri, N.; Hauchecorne, M.; Vigneron, J. P.;
Lehn, J. M.; Lehn, P. The design of cationic lipids for gene delivery.
Curr. Pharm. Des. 2005, 11, 375-394. (e) Christine, C.; Conwell,
C. C.; Huang, L. recent advances in non-viral gene delivery AdV.
Genet. 2005, 53, 3-18.
(
16) Zana, R. Dimeric and oligomeric surfactants. Behavior at interfaces
and in aqueous solution: a review. AdV. Colloid Interface Sci. 2002,
97, 205-253.
(
9) Felgner, P. L.; Barenholz, Y.; Behr, J.-P.; Cheng, S. H.; Cullis, P.;
Huang, L.; Jessee, J. A.; Seymour, L.; Szoka, F. C., Jr.; Thierry, A.
R. Nomenclature for synthetic gene delivery systems. Hum. Gene
Ther. 1997, 8, 511-512.
(
17) (a) Menger, F. M.; Littau, C. A. Gemini surfactants: a new class of
self-assembling molecules. J. Am. Chem. Soc. 1993, 115, 10083-
10090. (b) Menger, F. M.; Littau, C. A. Gemini-surfactants: synthesis
and properties. J. Am. Chem. Soc. 1991, 113, 1451-1452.
18) (a) Danino, D.; Talmon, Y.; Levy, H.; Beinert, G.; Zana, R. Branched
threadlike micelles in an aqueous solution of a trimeric surfactant.
Science 1995, 269, 1420-1421. (b) Zana, R.; Levy, H.; Papoutsi,
D.; Beinert, G. Micellization of two triquaternary ammonium
surfactants in aqueous solution. Langmuir 1995, 11, 3694-3698. (c)
Esumi, K.; Goino, M.; Koide, Y. Adsorption and adsolubilization
by monomeric, dimeric, or trimeric quaternary ammonium surfactant
at silica/water interface. J. Colloid Interface Sci. 1996, 183, 539-
545. (d) Maiti, P. K.; Lansac, Y.; Glaser, M. A.; Clark, N. A. Self-
assembly in surfactant oligomers: a coarse-grained description
through molecular dynamics simulations. Langmuir 2002, 18, 1908-
(
10) (a) Chesnoy, S.; Huang, L. Structure and function of lipid-DNA
(
complexes for gene delivery. Annu. ReV. Biophys. Biomol. Struct.
2000, 29, 27-47. (b) Ewert, K.; Ahmad, A.; Evans, H. M.; Safinya,
C. R. Cationic lipid-DNA complexes for non-viral gene therapy:
relating supramolecular structures to cellular pathways. Expert Opin.
Biol. Ther. 2005, 5, 33-53. (c) Ewert, K.; Slack, N. L.; Ahmad, A.;
Evans, H. M.; Lin, A. J.; Samuel, C. E.; Safinya, C. R. Cationic
lipid-DNA complexes for gene therapy: understanding the relation-
ship between complex structure and gene delivery pathways at the
molecular level. Curr. Med. Chem. 2004, 11, 133-149. (d) Simberg,
D.; Danino, D.; Talmon, Y.; Minsky, A.; Ferrari, M. E.; Wheeler,
C. J.; Barenholz, Y. Phase behavior, DNA ordering, and size
instability of cationic lipoplexes. Relevance to optimal transfection
activity. J. Biol. Chem. 2001, 276, 47453-47459. (e) Zuidam, N. J.;
Barenholz, Y. Characterization of DNA-lipid complexes commonly
used for gene delivery. Int. J. Pharm. 1999, 183, 43-46. (f) Xu, Y.
H.; Hui, S. W.; Frederik, P.; Szoka, F. C. Physicochemical
characterization and purification of cationic lipoplexes. Biophys. J.
1918. (e) Thalody, B.; Warr, G. G. The selective binding of anions
to gemini and trimeric surfactants at air/solution interfaces. Aust. J.
Chem. 2004, 57, 193-196.
(
19) (a) Bogatian, M.; Deleanu, C.; Mihai, G.; Balaban, T. S. Pyrylium
salts with long alkyl substituents. II. 2,4-Dimethyl-6-undecylpyrylium
perchlorate and derived pyridinium salts. Z. Naturforsch. 1992, 47b,
1999, 77, 341-353. (g) Perrie, Y.; Gregoriadis, G. Liposome-
1011-1015. (b) Bogatian, M.; Vinatoru, M.; Bogatian, G.; Mihai,
entrapped plasmid DNA: characterization studies. Biochim. Biophys.
Acta 2000, 1475, 125-132.
G.; Balaban, T. S. Artificial membranes by the self-assembly of
pyridinium salts derived from pyrylium salts with long alkyl
substituents. ReV. Roum. Chim. 2004, 49, 819-826.
(
11) Templeton, N. S. Cationic liposomes as in vivo delivery vehicles.
Curr. Med. Chem. 2003, 10, 1279-1287.
(
20) (a) Felgner, J. H.; Kumar, R.; Sridhar, C. N.; Wheeler, C. J.; Tsai,
Y. J.; Border, R.; Ramsey, P.; Martin, M.; Felgner, P. L. Enhanced
gene delivery and mechanism studies with a novel series of cationic
lipid formulations. J. Biol. Chem. 1994, 269, 2550-2561. (b) Ren,
T.; Liu, D. Synthesis of diether linked cationic lipids for gene
delivery. Bioorg. Med. Chem. Lett. 1999, 9, 1247-1250. (c) Massing,
U.; Kley, J. T.; Gurtesch, L.; Fankhaenel, S. A simple approach to
DOTAP and its analogs bearing different fatty acids. Chem. Phys.
Lipids 2000, 105, 189-191. (d) Floch, V.; Legros, N.; Loisel, S.;
Guillaume, C.; Guilbot, J.; Benvegnu, T.; Ferrieres, V.; Plusquellec,
D.; Ferec, C. New biocompatible cationic amphiphiles derivative from
glycine betaine: a novel family of efficient nonviral gene transfer
agents. Biochem. Biophys. Res. Commun. 1998, 251, 360-365. (e)
Banerjee, R.; Das, P. K.; Srilakshmi, G. V.; Chaudhuri, A.; Rao, N.
M. Novel series of non-glycerol-based cationic transfection lipids
for use in liposomal gene delivery. J. Med. Chem. 1999, 42, 4292-
(
12) (a) Solodin, I.; Brown, C. S.; Bruno, M. S.; Chow, C. Y.; Jang, E.
H.; Debs, R. J.; Heath, T. D. A novel series of amphiphilic
imidazolinium compounds for in vitro and in vivo gene delivery.
Biochemistry 1995, 34, 13537-13544. (b) Heyes, J. A.; Niculescu-
Duvaz, D.; Cooper, R. G.; Springer, C. J. Synthesis of novel cationic
lipids: effect of structural modification on the efficiency of gene
transfer. J. Med. Chem. 2002, 45, 99-114. (c) Majeti, B. K.; Karmali,
P. P.; Reddy, B. S.; Chaudhuri, A. In vitro gene transfer efficacies
of N,N-dialkylpyrrolidinium chlorides: a structure-activity investiga-
tion. J. Med. Chem. 2005, 48, 3784-3795. (d) Singh, R. S.;
Goncalves, C.; Sandrin, P.; Pichon, C.; Midoux, P.; Chaudhuri, A.
On the gene delivery efficacies of pH-sensitive cationic lipids via
endosomal protonation: a chemical biology investigation. Chem. Biol.
2004, 11, 713-723. (e) Kumar, V. V.; Pichon, C.; Refregiers, M.;
Guerin, B.; Midoux, P.; Chaudhuri, A. Single histidine residue in
head-group region is sufficient to impart remarkable gene transfection
properties to cationic lipids: evidence for histidine-mediated mem-
brane fusion at acidic pH. Gene Ther. 2003, 10, 1206-1215.
13) (a) van der Woude, I.; Wagenaar, A.; Meekel, A. A.; ter Beest, M.
B.; Ruiters, M. H.; Engberts, J. B.; Hoekstra, D. Novel pyridinium
surfactants for efficient, nontoxic in vitro gene delivery. Proc. Natl.
Acad. Sci. U.S.A. 1997, 94, 1160-1165. (b) Meekel, A. A. P.;
Wagenaar, A.; Smisterova, J.; Kroeze, J. E.; Haadsma, P.; Bosgraaf,
B.; Stuart, M. C. A.; Brisson, A.; Ruiters, M. H. J.; Hoekstra, D.;
Engberts, J. B. F. N. Synthesis of pyridinium amphiphiles used for
transfection and some characteristics of amphiphile/DNA complex
formation. Eur. J. Org. Chem. 2000, 665-673. (c) Roosjen, A.;
Smisterova, J.; Driessen, C.; Anders, J. T.; Wagenaar, A.; Hoekstra,
D.; Hulst, R.; Engberts, J. B. F. N. Synthesis and characteristics of
biodegradable pyridinium amphiphiles used for in vitro DNA
delivery. Eur. J. Org. Chem. 2002, 1271-1277.
4299.
(
21) Israelachvili, J. Intermolecular and Surface Forces, 2nd ed.; Aca-
demic Press: New York, 1991.
22) (a) Balaban, T. S.; Balaban, A. T. Pyrylium Salts. Science of Synthesis.
Houben-Weyl Methods of Molecular Transformations; G. Thieme
Verlag: Stuttgart, Germany, 2003; pp 11-200. (b) Schroth, W.;
Balaban, A. T. Pyrylium Salts. Methoden der Organischen Chemie
(
(
(Houben-Weil) (Methods of Organic Chemistry (Houben-Weil)); G.
Thieme Verlag: Stuttgart, Germany, 1992; pp 755-963. (c) Balaban,
A. T.; Dinculescu, A.; Dorofeenko, G. N.; Fischer, G. W.; Koblik,
A. V.; Mezheritskii, V. V.; Schroth, W. Pyrylium Salts: Syntheses,
Reactions and Physical Properties. In AdVances in Heterocyclic
Chemistry; Academic Press: New York, 1982; pp 8-360.
23) (a) Leventis, R.; Silvius, J. R. Interactions of mammalian cells with
lipid dispersions containing novel metabolizable cationic amphiphiles.
Biochim. Biophys. Acta 1990, 1023, 124-132. (b) Templeton, N.
S.; Lasic, D. D.; Frederik, P. M.; Strey, H. H.; Roberts, D. D.;
Pavlakis, G. N. Improved DNA: liposome complexes for increased
systemic delivery and gene expression. Nat. Biotechnol. 1997, 15,
647-652.
(
(
14) (a) Ilies, M. A.; Seitz, W. A.; Caproiu, M. T.; Wentz, M.; Garfield,
R. E.; Balaban, A. T. Pyridinium-Based Cationic Lipids as Gene-
Transfer Agents. Eur. J. Org. Chem. 2003, 2645-2655. (b) Ilies,
M. A.; Seitz, W. A.; Ghiviriga, I.; Johnson, B. H.; Miller, A.;
Thompson, E. B.; Balaban, A. T. Pyridinium cationic lipids in gene
delivery: a structure-activity correlation study. J. Med. Chem. 2004,

Pyrylium Salts for Synthesis
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 13 3887
(
24) Balasubramaniam, R. P.; Bennett, M. J.; Aberle, A. M.; Malone, J.
G.; Nantz, M. H.; Malone, R. W. Structural and functional analysis
of cationic transfection lipids: the hydrophobic domain. Gene Ther.
Milroy, D. A.; Wood, P. J.; Pouton, C. W.; Threadgill, M. D.
Synthesis and characterisation of polyamine-poly(ethylene glycol)
constructs for DNA binding and gene delivery. Bioorg. Med. Chem.
2000, 8, 1779-1797.
1996, 3, 163-172.
(
25) Matsuo, H.; Chevallier, J.; Mayran, N.; Le Blanc, I.; Ferguson, C.;
Faure, J.; Blanc, N. S.; Matile, S.; Dubochet, J.; Sadoul, R.; Parton,
R. G.; Vilbois, F.; Gruenberg, J. Role of LBPA and Alix in
multivesicular liposome formation and endosome organization.
Science 2004, 303, 531-534.
(27) Vitiello, L.; Bockhold, K.; Joshi, P. B.; Worton, R. G. Transfection
of cultured myoblasts in high serum concentration with DODAC:
DOPE liposomes. Gene Ther. 1998, 5, 1306-1313.
(
28) (a) Karmali, P. P.; Kumar, V. V.; Chaudhuri, A. Design, syntheses
and in vitro gene delivery efficacies of novel mono-, di- and
trilysinated cationic lipids: a structure-activity investigation. J. Med.
Chem. 2004, 47, 2123-2132. (b) Rosenzweig, H. S.; Rakhmanova,
V. A.; MacDonald, R. C. Diquaternary ammonium compounds as
transfection agents. Bioconjugate Chem. 2001, 12, 258-263.
29) Berridge, M. V.; Tan, A. S.; McCoy, K. D.; Wang, R. The
biochemical and cellular basis of cell proliferation assays that use
tetrazolium salts. Biochemica 1996, 14-19.
30) (a) Kumar, V. V.; Chaudhuri, A. On the disulfide-linker strategy for
designing efficacious cationic transfection lipids: an unexpected
transfection profile. FEBS Lett. 2004, 571, 205-211. (b) Tang, F.;
Hughes, J. A. Introduction of a disulfide bond into a cationic lipid
enhances transgene expression of plasmid DNA. Biochem. Biophys.
Res. Commun. 1998, 242, 141-145. (c) Tang, F.; Wang, W.; Hughes,
J. A. Cationic liposomes containing disulfide bonds in delivery of
plasmid DNA. J. Lipid Res. 1999, 9, 331-347. (d) Byk, G.; Wetzer,
B.; Frederic, M.; Dubertret, C.; Pitard, B.; Jaslin, G.; Scherman, D.
Reduction-sensitive lipopolyamines as a novel nonviral gene delivery
system for modulated release of DNA with improved transgene
expression. J. Med. Chem. 2000, 43, 4377-4387. (e) Dauty, E.;
Remy, J. S.; Blessing, T.; Behr, J. P. Dimerizable cationic detergents
with a low cmc condense plasmid DNA into nanometric particles
and transfect cells in culture. J. Am. Chem. Soc. 2001, 123,
(
26) (a) O’Sullivan, M. C.; Dalrymple, D. M. A one-step procedure for
the selective trifluoroacetylation of primary amino groups of
polyamines. Tetrahedron Lett. 1995, 36, 3451-3452. (b) Xu, D.;
Prasad, K.; Repic, O.; Blacklock, T. J. Ethyl trifluoroacetate: a
powerful reagent for differentiating amino groups. Tetrahedron Lett.
(
(
1
995, 36, 7357-7360. (c) O’Sullivan, M. C.; Zhou, Q.; Li, Z.;
Durham, T. B.; Rattendi, D.; Lane, S.; Bacchi, C. J. Polyamine
derivatives as inhibitors of trypanothione reductase and assessment
of their trypanocidal activities. Bioorg. Med. Chem. 1997, 5, 2145-
2155. (d) Carrington, S.; Blagbrough, I. S.; Fairlamb, A. H.
Macrocyclic polyamine lactam synthesis by diphenyl ether closure
of 23-, 24- and 28-membered rings. Chem. Commun. 1998, 21, 2335-
2
336. (e) Blagbrough, I. S.; Geall, A. J. Practical synthesis of
unsymmetrical polyamine amides. Tetrahedron Lett. 1998, 39, 439-
4
42. (f) Rauter, H.; Di Domenico, R.; Menta, E.; Oliva, A.; Qu, Y.;
Farrell, N. Selective platination of biologically relevant polyamines.
Linear coordinating spermidine and spermine as amplifying linkers
in dinuclear platinum complexes. Inorg. Chem. 1997, 36, 3919-
3927. (g) Stromgaard, K.; Bjornsdottir, I.; Andersen, K.; Brierley,
M. J.; Rizoli, S.; Eldursi, N.; Mellor, I. R.; Usherwood, P. N.; Hansen,
S. H.; Krogsgaard-Larsen, P.; Jaroszewski, J. W. Solid phase synthesis
and biological evaluation of enantiomerically pure wasp toxin
analogues PhTX-343 and PhTX-12. Chirality 2000, 12, 93-102. (h)
Koscova, S.; Budesinski, M.; Hodacova, J. A facile synthesis of
selectively protected linear oligoamines. Collect. Czech. Chem.
Commun. 2003, 68, 744-750. (i) Garrett, S. W.; Davies, O. R.;
9227-9234.
JM0601755