Erufosine (ErPC3) Cationic Prodrugs as Dual Gene Delivery Reagents for Combined Antitumor Therapy

Article abstract of DOI:10.1002/chem.201903976

Sixteen cationic prodrugs of the antitumor alkylphospholipid (APL) erufosine were rationally synthesized to provide original gene delivery reagents with improved cytotoxicity profile. The DNA complexation properties of these cationic lipids were determined and associated transfection rates were measured. Furthermore, the self-assembly properties of the pro-erufosine compounds were investigated and their critical aggregation concentration was determined. Their hydrolytic stability under pH conditions mimicking the extracellular environment and the late endosome milieu was measured. Hemolytic activity and cytotoxicity of the compounds were investigated. The results obtained in various cell lines demonstrate that the prodrugs of erufosine display antineoplastic activity similar to that of the parent antitumor drug but are not associated with hemolytic toxicity, which is a dose-limiting side effect of APLs and a major obstacle to their use in anticancer therapeutic regimen. Furthermore, by using lipoplexes prepared from a prodrug of erufosine and a plasmid DNA encoding a pro-apoptotic protein (TRAIL), evidence was provided for selective cytotoxicity towards tumor cells while nontumor cells were resistant. This study demonstrates that the combination approach involving well tolerated erufosine cationic prodrugs and cancer gene therapy holds significant promise in tumor therapy.

Full text of DOI:10.1002/chem.201903976

A Journal of
Accepted Article
Title: Erufosine (ErPC3) Cationic Prodrugs as Dual Gene Delivery
Reagents for Combined Antitumor Therapy
Authors: Boris Gaillard, Cendrine Seguin, Jean-Serge Remy,
Françoise Pons, and Luc Lebeau
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201903976
Link to VoR: http://dx.doi.org/10.1002/chem.201903976
Supported by

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
Erufosine (ErPC3) Cationic Prodrugs as Dual Gene Delivery
Reagents for Combined Antitumor Therapy
Boris Gaillard, Cendrine Seguin, Jean-Serge Remy, Françoise Pons, and Luc Lebeau *[a]
Abstract:
Sixteen
cationic
prodrugs
of
the
antitumor
the action of acyltransferase into PCs, or by lysophospholipases
into glycerophosphocholine, efforts have been devoted to the
synthesis of metabolically stable analogs to act as modulators of
signaling pathways and alter biochemical routes (Figure 1).
Edelfosine (1-O-octadecyl-2-O-methyl-rac-glycero-3-phosphocholine),
the first molecule of this drug class was described in the late
60s.[3] Though it was proven effective in in vitro and in vivo
antiproliferative assays, clinical use was limited mainly due to
metabolic instability, high hemolytic activity, and gastrointestinal
toxicity.[4] Ilmofosine (1-hexadecylthio-2-methoxymethyl-rac-
glycero-3-phosphocholine), a thio-ether analog was proposed but,
although it revealed as an effective inhibitor of cell proliferation,
the chemical modification did not alter significantly its metabolic
stability or cytotoxic effect.[5] In the late 80s, miltefosine (HePC,
hexadecylphosphocholine), a new analog lacking the glycerol
moiety, was discovered and brought APLs to a clinical important
level.[6] Similarly to the earlier APLs, miltefosine presented
appreciable in vitro antiproliferative activity but was found highly
hemolytic and provoked marked gastrointestinal toxicity.
Consequently, the clinical use of miltefosine has been limited
mainly to the topical treatment of skin metastases in patients with
breast cancer. To improve the therapeutic potency of APLs,
replacement of the choline moiety in miltefosine by a dimethyl
piperidinium group resulted in perifosine (octadecyl-1-
(dimethylpiperidinio-4-yl)-phosphate).[7] Stability and half-life of
the compound were significantly increased, preventing rapid
metabolic degradation.[8] However, hemolytic activity still was high
and toxicity profile after oral administration was similar to that of
miltefosine with no successful outcome in the treatment of various
alkylphospholipid (APL) erufosine were rationally synthesized to
provide original gene delivery reagents with improved cytotoxicity
profile. The DNA complexation properties of these cationic lipids were
determined and associated transfection rates were measured.
Besides, the self-assembly properties of the pro-erufosine
compounds were investigated and their critical aggregation
concentration was determined. Their hydrolytic stability under pH
conditions mimicking the extracellular environment and the late
endosome milieu was measured. Hemolytic activity and cytotoxicity of
the compounds were investigated. The results obtained in various cell
lines demonstrate that the prodrugs of erufosine display antineoplastic
activity similar to that of the parent antitumor drug but are not
associated with hemolytic toxicity, a dose-limiting side effect of APLs
and a major obstacle to their use in anticancer therapeutic regimen.
Furthermore, using lipoplexes prepared from a prodrug of erufosine
and a plasmid DNA encoding a pro-apoptotic protein (TRAIL),
evidence was provided for selective cytotoxicity towards tumor cells
while non-tumor cells were resistant. This study demonstrates that the
combination approach involving well tolerated erufosine cationic
prodrugs and cancer gene therapy holds significant promise in tumor
therapy.
Introduction
Alkylphospholipids (APLs) constitute a new class of antitumor
drugs which do not interact directly with DNA but target the cell
membrane where they accumulate and interfere with lipid
metabolism and signaling pathways, thereby affecting the growth,
cell cycle progression, and survival of tumor cells.[1] APLs
originally derived from lysophosphatidylcholines (lysoPCs),[2]
endogenous cell membrane components that result from
hydrolysis of phosphatidylcholines (PCs) by phospholipases.
Since lysoPCs are not stable and become inactivated, either by
O
O
P
LysoPC
N
O
O
O
O
O
O
O
H
H
OH
O
O
P
Edelfosine
N
O
O
P
Ilmofosine
Miltefosine
N
S
O
O
O
O
[a]
Dr. B. Gaillard, C. Seguin, Dr. J.-S. Remy, Prof. Dr. F. Pons, Dr. L.
Lebeau
O
P
N
N
O
O
O
O
O
Laboratoire de Conception et Application de Molécules Bioactives
UMR 7199 CNRS - Université de Strasbourg
Faculté de Pharmacie, 74 route du Rhin - BP 60024
67401 Illkirch (France)
O
P
Perifosine
Erufosine
O
E-mail: llebeau@unistra.fr
O
P
Supporting information for this article is given via a link at the end of
the document.
O
O
N
O
Figure 1. Structure of 18:0 lysoPC and APLs.
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
cancers.[9] As a member of the fourth generation of APLs,
erufosine (erucylphosphohomocholine) resulted from the
simultaneous introduction of an w-9-cis-unsaturated alkyl chain
with 22 carbon atoms and a homocholine moiety on the
phosphate group of the APL. Extension of the hydrophobic chain
decreased aqueous solubility of the molecule that likely self-
assembles into lamellar structures instead of micelles, as was
nucleic acid phosphates then becomes thermodynamically
disfavored, as compared to intramolecular zwitterionic interaction.
This process thus triggers and facilitates nucleic acid
decondensation, which is required for nucleic acid processing by
the cell translation machinery. Besides, from the sorting
endosomes, most amphiphilic molecules enter the endocytic
recycling pathway.[18] Lipid recycling at the plasma membrane is
claimed for a closely related structural analog.[10] As a consequence, a highly dynamic and effective process,[19] and antiproliferative
the hemolytic potential was significantly reduced, which enabled
intravenous (iv) administration that was not feasible with previous
APLs.[11]
APL produced in situ from the nucleic acid carrier is prone to
trafficking at the plasma membrane where it can operate intrinsic
apoptotic activity.[20] Starting from erufosine, one of the most
promising APLs to date, original lipid constructs were thus
imagined for simultaneously implementing gene therapy and
chemotherapy. A series of 16 biolabile cationic lipids have been
designed which can regenerate erufosine in situ under a chemical
or enzyme stimulus. The properties of these erufosine prodrugs
as gene delivery reagents have been investigated using a
luciferase reporter gene assay, and their intrinsic cytotoxicity has
been determined. Finally, using a plasmid DNA encoding the
tumor necrosis factor-related apoptosis-inducing ligand (TRAIL),
the in vitro synergistic antiproliferative effect of TRAIL and
erufosine prodrugs has been examined.
Due to their particular mechanism of action targeting the cell
membrane, APLs may be useful in combination with other
chemotherapy agents and radiation therapy. Several clinical
studies have reported significant benefits of combining APLs with
diverse antineoplastic agents (cytarabine, gemcitabine,
capecitabine, idarubicin, doxorubicin, etoposide, docetaxel,
staurosporine, sorafenib, lenalidomide-dexamethasone, bortezomib-
dexamethasone, temsirolimus…).[12] Besides, some of the
pathways by which APLs lead to cytotoxicity have been shown to
influence radiosensitivity or radioresistance of tumor cells, which
resulted in successful clinical outcome of the combination of APLs
with radiotherapy.[13] On the other hand, there are only very few
reports on the combination of APLs with cancer gene therapy. In
2003, Zeisig et al. described gene transfer with liposomal
formulations incorporating APLs as helper lipids,[14] assuming that
APLs may promote the transmembrane transport of the plasmid-
lipid complex (lipoplex) into the target cell due to their "detergent-
like properties". Introduction of 2 % tetradecylphosphocholine
(TPC), a lower carbon homologue of miltefosine, in the lipoplexes
improved transfer efficiency of the LacZ gene into HCT15 and
HCT116 human colon carcinoma cells up to 43 %. Similarly,
Settelen et al. reported the intratumoral co-administration of
naked DNA with miltefosine.[15] Though this non condensing-
plasmid formulation failed to promote transgene expression in
vitro, it showed a ten-fold increased reporter gene expression with
increasing miltefosine concentration in Renca tumor following
intratumor injection, as compared to plasmid alone. More recently,
perifosine was used in combination with siRNA lipoplexes
silencing c-FLIP, an inhibitory protein that blocks the extrinsic
pathway of apoptosis by preventing caspase-8 activation by death
receptors.[16]
In the course of our research towards developing new nucleic
acid carriers, we were attracted by APLs as they can be
conveniently transformed into cationic lipids, i.e., molecules that
may be potentially powerful nucleic acid condensing reagents.
Therefore, the transformation of an APL into a phosphotriester
derivative generates a net positive charge on the molecule. This
derivative can thus establish electrostatic interactions with
negatively charged nucleic acid and promote its condensation into
discrete particles as required for cell uptake and intracellular
delivery through endocytosis.[17] Furthermore, when the
transformation of the APL compound is reversible and sensitive,
e.g., to a pH or enzyme stimulus as met along the endo-lysosome
pathway, in situ regeneration of the zwitterionic APL may be
expected. Intermolecular electrostatic interaction between the
ammonium headgroup of the newly unmasked APL molecule and
Results and Discussion
Design of erufosine prodrugs as gene carriers
Erufosine is currently the most promising APL, essentially due to
its structure with a 22-carbon long chain and an w-9-cis-double
bond that imposes formation of lamellar instead of micellar
structures in aqueous solution. This results in a significant
decrease in hemolytic activity as compared to other APLs and
allows intravenous administration at clinically relevant high
doses.[21] Transformation of erufosine into a cationic species with
nucleic acid complexation ability can be straightforwardly
achieved through esterification into phosphotriester compounds.
However, non-selective hydrolysis of such compounds into
phosphodiesters would produce a mixture of erufosine together
with two other phosphodiesters resulting from the alternate
leaving of the homocholine or heneicosenol phosphate
substituent. Consequently, to favor formation of the parent APL
upon hydrolysis of these phosphotriesters, it is valuable to
introduce a (bio)labile substituent at the phosphate that may be
preferentially removed under a chemical or enzyme stimulus.
Based on previous reports on the development of DOPC-based
phosphoacetals as labile cationic lipids for gene delivery,[22] we
aimed to extend this methodology to produce potent prodrugs of
erufosine (Figure 2). Consequently, erufosine was derivatized to
incorporate various n-carbon alkyl chains connected to the
phosphate group, directly or through a mixed acetal moiety.
Compounds were developed both in the ester (compounds noted
EEn) and carbonate (compounds noted ECn) mixed phosphoacetal
series to modulate their hydrolytic stability which presumably at
least partly determines their potency as gene delivery reagents
and the rate of erufosine release in situ. Pursuing the same idea,
a methyl substituent was introduced on the acetal bridge
(compounds noted EE'n and EC'n) as it was recently demonstrated
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
O
P
R-X, 8 eq.
CHCl3, Δ, 24 h
extracellular and late endosome environment, i.e., pH 7.4 and 4.5,
respectively. This was achieved through 31P-NMR measurements.
As phosphotriesters and phosphodiesters display 31P chemical
shifts differing by ca. 5–6 ppm, it is possible to precisely monitor
the quantity of both species in a sample and thus determine the
hydrolysis rate of a phosphotriester in solution. The pro-APLs
were formulated into liposomes by using an injection technique,[23]
at pH 7.4 and 4.5, and incubated at 25 °C. Periodical acquisition
of the 31P-NMR spectra using a 4-sec pulse cycle (enabling for
quantitative determination of the phosphorylated species through
the integration of their resonance signal) was carried out and the
rate of hydrolysis was determined (Table 1). Whatever the
compound investigated, only one single 31P resonance did appear
during the course of the experiments. This signal did correspond
to that of erufosine and no other phosphodiester could be
detected. This confirmed that hydrolysis selectively occurred at
the phosphoacetal center. Consistently, phosphotriester E12
lacking acetal moiety revealed fully stable both under neutral and
acidic conditions, even after an incubation period of 31 days. All
the other erufosine derivatives were sensitive to hydrolysis. As a
general trend, hydrolysis was quicker at pH 7.4 than at pH 4.5.
This indicates that mixed acetals of carboxylic or carbonic and
phosphoric esters display a reactivity that is different from that of
dialkyl acetals which are highly labile species in acidic media, and
is consistent with results reported for other phosphoacetals.[22e, f]
Erufosine
O
O
N
X
O
R
Entry
Name
E12
Yield (%)
R
X
(
1
OTf
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
97
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
2
3
EE12
EE'12
EC12
EC'12
EE18:1
EE'18:1
EC18:1
EC'18:1
EE2
41
50
31
41
35
47
58
49
6
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
4
O
O
5
6
7
8
O
O
9
10
11
12
13
14
15
16
EE6
30
69
74
32
74
55
EE8
EE10
EE14
EE16
EE18
Figure 2. Synthetic route to and structure of the 16 erufosine prodrugs
investigated in this work.
Table 1. Hydrolytic stability of the pro-erufosine compounds. Compounds
formulated into liposomes were incubated at 25 °C and hydrolysis at pH 7.4
and 4.5 was monitored by quantitative 31P-NMR measurements. H120 refers to
the extent of hydrolysis (%) after incubation over 120 h. When possible, time
required for 50 % hydrolysis (t1/2) was calculated from the theoretical curve
fitting with the experimental data.
that it can significantly improve the hydrolysis rate of phosphoacetal
compounds.[22f] All these compounds were producedaccording to
the same protocol, reacting erufosine with a large excess of
electrophilic reagent (8 eq.) in boiling anhydrous chloroform for
24 h. The reaction yields were highly variable (6 to 97 %) and did
depend on the reactivity of the electrophilic reagents and on the
stability of the reaction products, especially for compound E2. Due
to the presence of two stereogenic centers, compounds
H120 (%)
t1/2 (h)
Entry Compound
pH 7.4
pH 4.5
pH 7.4
pH 4.5
displaying a methyl-substituted acetal bridge (EE'12, EC'12, EE'18:1
,
[a]
1
2
E12
0
0
2
–
–
–
–
and EC'18:1) were produced as a mixture of four diastereomers.
When possible, the couples of enantiomers were separated for
analytic purpose, but evaluations (vide infra) were performed on
the original mixture of the four isomers.
EE12
EE'12
EC12
EC'12
EE18:1
EE'18:1
EC18:1
EC'18:1
EE2
14
56
8
3
53
1
96
–
100
–
4
5
6
5
–
–
Hydrolytic stability of the erufosine prodrugs
6
23
56
10
4
18
54
0
–
–
7
96
–
104
–
Aqueous formulation of the erufosine prodrugs results in the
formation of nanosized lipid aggregates that most likely are
internalized by cells through the endocytic route. To express the
expected intrinsic antitumor activity (vide infra), these compounds
need to be transformed back into erufosine which is the bioactive
species. Hydrolysis of the erufosine prodrugs may occur under a
chemical stimulus, e.g., acidification of the environment as
observed in the endosome during its maturation process, or may
be under the control of hydrolytic enzymes that are abundantly
introduced into the endosome compartment upon fusion with
lysosomes. To get an insight into regeneration of erufosine from
its prodrugs, we determined the hydrolytic stability of the pro-
erufosine compounds under pH conditions mimicking the
8
9
6
–
–
10
11
12
13
14
15
16
–
–
–
–
EE6
18
16
16
14
16
14
8
–
–
EE8
1
–
–
EE10
EE14
EE16
EE18
2
–
–
2
–
–
2
–
–
3
–
–
[a] No hydrolysis was observed after an incubation period of 31 days.
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
Table 2. Particle size and zeta potential (z) of lipoplexes obtained from pro-
erufosine compounds and pCMV-Gluc, as measured by dynamic light scattering.
Lipoplexes were prepared at 25 °C, in 5 % glucose, at a N/P ratio of 3 and with
Based on these former reports, introduction of a methyl substituent
on the acetal bridge was expected to improve the hydrolysis rate
of the erufosine prodrugs. This however only held true in the ester
series and EE'12 and EE'18:1 indeed hydrolyzed faster than EE12 and
EE18:1, respectively. In the carbonate series, compounds were
significantly more resistant to hydrolysis and introduction of a
methyl substituent on the acetal bridge had only little effect, if any,
on the rate of the hydrolytic reaction (EC'12 vs. EC12, and EC'18:1 vs.
EC18:1). With regards to the influence of the length of the pendant
arm tethered to erufosine through the acetal linker on the
hydrolysis rate, no significant difference was observed, except for
EE6 that revealed ca. four times more labile than higher homologs
under acidic conditions. For technical reasons, we could not
measure the hydrolytic stability of the lower homolog EE2. Though
these results allowed a comparison of the hydrolytic stability of
the erufosine prodrugs, it must be kept in mind that it only
projected an incomplete and truncated image of the in cellulo
stability of these compounds, as in situ enzymatic hydrolysis of
the pro-erufosine compounds most probably occurs much faster
than "purely" pH-controlled chemical hydrolysis, as was reported
with other phosphoacetals.[22c]
1
molar equivalent of DOPE. Data are the mean of three independent
measurements (± SD).
Entry
Compound
Particle size (nm)
z (mV)
1
2
E12
194 ± 28
141 ± 5
+ 42 ± 1
+ 48 ± 2
+ 39 ± 2
+ 40 ± 1
+ 54 ± 2
– 35 ± 4
+ 45 ± 2
+ 47 ± 3
+ 44 ± 1
– 31 ± 2
+ 47 ± 6
+ 36 ± 1
+ 41 ± 3
+ 41 ± 1
+ 42 ± 3
+ 43 ± 1
EE12
EE'12
EC12
EC'12
EE18:1
EE'18:1
EC18:1
EC'18:1
EE2
3
149 ± 15
117 ± 15
92 ± 10
4
5
6
263 ± 9
7
597 ± 19
611 ± 12
608 ± 44
133 ± 10
144 ± 31
110 ± 4
8
9
10
11
12
13
14
15
16
EE6
EE8
EE10
EE14
EE16
EE18
122 ± 11
121 ± 29
138 ± 29
120 ± 19
Lipoplex formation and characterization
The ability of the erufosine prodrugs to interact electrostatically
with nucleic acids and form complexes was studied by
conventional electrophoretic DNA retardation assay, in which a
full retard of plasmid DNA (pDNA) is observed at a lipid/pDNA
phosphate ratio (N/P) corresponding to electroneutrality of the
complexes. The erufosine prodrugs generally led to a full pDNA
complexation at N/P > 1.2-2.0 as illustrated for compound EE12 in
Figure 3. In the case of EE2 however, electroneutrality was not
reached even at N/P 4, presumably because this compound was
too labile (vide supra) and was partly hydrolyzed during gel
migration. The size (hydrodynamic diameter) and charge (zeta
potential, z) of the lipoplexes prepared at a N/P ratio of 3 with one
molar equivalent of DOPE were investigated by DLS (Table 2).
Most lipoplexes displayed a size in the range of 100-150 nm.
Noteworthy, introduction of an unsaturated chain on the
phosphate group of erufosine, regardless of the linker, translated
into the formation of significantly larger lipoplexes (250-600 nm,
entry 6-9). As could be expected considering electrophoretic
behavior, nearly all lipoplexes were highly positively charged at
N/P 3, z values ranging from +36 to +54 mV. Again, EE2-based
lipoplexes distinguished as they were negatively charged, which was
consistent with significant hydrolysis of the compound in aqueous
media. More intriguing was the negative charge measured for
lipoplexes prepared from EE18:1 (–35 mV ± 4 mV), revealing
significant differences in the organization of the lipid molecules
and pDNA within the transfection particles (affecting a more precise
matching of the ammonium and phosphate species[24]), as
compared to those obtained with other lipids, especially EE'18:1
,
EC18:1 and EC'18:1
.
DNA transfection
The efficacy of the pro-erufosine compounds to deliver a plasmid
DNA intracellularly was investigated in the A549 cell line using the
reporter gene pCMV-Gluc encoding the Gaussia princeps
luciferase, under the control of a CMV promoter. As this luciferase
is excreted by cells, transgene expression was directly assessed
by standard bioluminescence measurements on cell culture
supernatant. Various structural features of the pro-erufosine
compounds were examined, in order to identify structure-activity
relationships, namely the nature of the biolabile linker (acetal of
esters and acetal of carbonates), presence of a substituent on the
acetal bridge, and length of the carbon chain introduced on the
erufosine phosphate group.
0
0.8 1.0 1.2 1.5
2
3
4
Nature of the biolabile linker
Considering a fixed-length substituent (C12) at the phosphate
Figure 3. pDNA binding ability of pro-erufosine EE12 at increasing N/P ratio (0
to 4). Lipoplexes were prepared by mixing 0.8 µg of plasmid DNA (pCMV-Gluc)
and increasing amounts of pro-erufosine in 10 mM Hepes buffer. After an
incubation period of 20 min, samples (25 µL) were analyzed by 1 % agarose gel
electrophoresis using Tris-acetate-EDTA buffer. DNA was visualized after
ethidium bromide staining.
group of erufosine, the five pro-erufosine compounds E12, EE12
,
EE'12, EC12, and EC'12 were compared in order to investigate the
influence of the nature of the biolabile linker on transfection
efficiency (Figure 4). No significant transgene expression was
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
12
1,E+12
10
11
1,E+11
10
10
1,E+10
10
9
1,E+09
10
8
1,E+08
10
7
1,E+07
10
DOPE (eq.)
0
1
2
3 0 1 2 3 0 1 2 3
N/P
1
1
3
3
5
5
C
C
L
L
E12
E12
12
1,E+12
12
1,E+12
10
10
11
1,E+11
11
1,E+11
10
10
10
1,E+10
10
1,E+10
10
10
9
9
1,E+09
1,E+09
10
10
8
8
1,E+08
1,E+08
10
10
7
7
1,E+1007
1,E+07
10
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
DOPE (eq.)
DOPE (eq.)
1
1
3
3
5
5
1
3
5
N/P
N/P
1
3
5
C
C
L
L
EE'12
EE’12
C
L
EE12
EE12
C
L
12
1,E+12
10
12
1,E+12
10
11
11
1,E+11
1,E+11
10
10
10
10
1,E+10
1,E+10
10
10
9
9
1,E+09
1,E+09
10
10
8
8
1,E+08
10
1,E+1008
7
7
1,E+07
10
1,E+1007
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
0
1
2
3
DOPE (eq.)
DOPE (eq.)
1
1
3
3
5
5
1
1
3
3
5
5
N/P
N/P
C
C
L
L
EC12
EC12
C
C
L
L
EC'12
EC’12
Figure 4. Transfection profile of pro-erufosine compounds as a function of the structure of the biolabile linker installed on the phosphate group. Experiments were
carried out on A549 cells, varying the amount of DOPE in the lipid formulations and the charge ratio N/P (DNA: 0.4 µg/well), in the presence of 10 % serum. Control
(C) refers to basal bioluminescence measured in untreated cells. Lipofectamine®-based lipoplexes (L) were prepared under experimental conditions that were
optimized according to the supplier's instructions. Data shown are representative of a triplicate determination (mean ± SD).
generally observed in the absence of DOPE, a fusogenic helper
lipid routinely used in transfection experiments.[25] Formulations
incorporating up to three molar equivalents of DOPE were
systematically evaluated, and one molar equivalent of the fusogenic
lipid with respect to the pro-erufosine compound most often led to
optimal transfection efficiency. With regards to the chemical
nature of the linker installed between the phosphate group of
erufosine and the C12 carbon chain (i.e., docecyl or dodecanoyl
substituent), acetal ester (EE12) revealed more efficient to deliver
the pCMV-Gluc transgene to cells and was almost matching the
results obtained with Lipofectamine® 2000, a gold standard gene
delivery reagent. Transfection rate was increased by a ca. twenty-
fold factor, as compared to that of E12 lacking specific biolabile
resulted in a two- to three-fold decrease in transfection efficiency.
Besides, introduction of a methyl substituent on the acetal bridge
also translated into a loss of transgene expression. The effect was
more pronounced in the ester series (8.5 1010 vs.
1.1 1010 RLU/mg protein for EE12 and EE'12, resp.) than in the
carbonate series (3.3 1010 vs. 1.5 1010 RLU/mg protein for EC12
and EC'12, resp.). Thus, it was deduced that acceleration of pro-
erufosine hydrolysis was not in favor of a higher transfection rate.
This might indicate either that some early decondensation of
plasmid in the extracellular milieu decreased cell uptake or that
intracellular plasmid decondensation was not achieved in a timely
manner, possibly resulting in enhanced DNA degradation by
nucleases before reaching the nucleus for processing.
linker. Replacement of the ester by a carbonate (i.e., EC12
)
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
Considering the compound series displaying an unsaturated
C18:1 pendant arm (i.e., EE18:1, EE'18:1, EC18:1, and EC'18:1), results
were slightly different (Figure S1). Whereas optimum transfection
rate required the use of one molar equivalent of DOPE or more
with EE'18:1, EC18:1, and EC'18:1, this lipid "helper" was deleterious
with EE18:1, decreasing transfection rate by ca. an order of
magnitude. With respect to the nature of the biolabile linker,
introduction of a methyl substituent on the acetal bridge in the
ester series slightly improved transfection efficiency (EE18:1 vs.
EE'18:1), whereas opposite effect was observed in the carbonate
series (EC18:1 vs. EC'18:1). Together with the results obtained in the
C12 series, this highlighted that transfection rate did depend both
on the structure of the biolabile linker and on that of the pendant
alkyl chain. One hypothesis is that conformation and orientation
of the polar head of the molecules facing the aqueous milieu
depend on intrinsic structural parameters of the lipid (structure of
the pendant hydrophobic arm, nature of the linker, steric
hindrance at the acetal bridge…), with direct implications on
lipase activity with cascading effects (modification of the
hydrolysis rate of the compounds, of pDNA decondensation rate
under the threat of nucleases, and ultimately of transfection
efficiency).
Lipoplexes of large size are frequently reported to be more
effective to promote in vitro transgene expression than smaller
ones due to sedimentation and facilitated size-dependent cellular
uptake via a switch from a clathrin-dependent to a caveolae-
mediated cell entry pathway.[26] That was not the case in this study
as larger lipoplexes (i.e. from EE18:1, EE'18:1, EC18:1, and EC'18:1, with
size ranging from 263 to 608 nm) revealed less efficient than
smaller ones (i.e. from EE12 EE'12, EC12, and EC'12, with size ranging
from 92 to 149 nm). Similar results for cationic lipids such as
SAINTs[27] or COPA[28] have been reported previously. In our
study, that probably revealed that due the structure of the pro-
erufosine they were made of, and although internalized by cells to
12
1,E+12
10
11
1,E+11
10
10
1,E+10
10
9
1,E+09
10
8
1,E+08
10
7
1,E+07
10
C
C
E2 E6 E8 E10 E12 E14 E16 E18 1E8:01
E2 E6 E8 E10 E12 E14 E16 E18 E18:1
Figure 5. Transfection profile of pro-erufosine EEn compounds as a function of
the length of the alkyl chain tethered to the phosphate group. Experiments were
carried out on A549 cells, in the presence of 10 % serum. Lipoplexes were
formulated with DOPE (1 eq.) at N/P 3 (DNA: 0.4 µg/well). Control (C) refers to
basal bioluminescence measured in untreated cells. Data shown are
representative of a triplicate determination (mean ± SD).
DNA dose and N/P ratio
As an important parameter governing transgene expression, we
then investigated the influence of the pDNA dose deposited onto
the cultured cells. A549 cells were treated with increasing
amounts of pCMV-Gluc (from 0.1 to 0.4 µg/well) formulated into
lipoplexes (EE12/DOPE 1/1) with gradual increase of the N/P ratio
(1 to 4). As a general trend and as might be expected, the higher
the DNA dose, the higher the transfection rate (Figure 6).
However, when increasing the pDNA dose, optimum transfection
efficiency was attained at lower N/P ratio. Indeed, at the lower
pDNA dose (0.1 µg/well), transfection rate did not yet reach a
maximum with N/P increasing up to 4. At 0.2 µg DNA/well,
maximum transfection efficiency was attained at N/P ratio of 3-4,
whereas it was observed at N/P of 2.75-3 at 0.3 µg/well and at
N/P of 2.5-2.75 at 0.4 µg/well. This shift of charge ratio for optimal
transfection may be consistently explained by intrinsic cytotoxicity
of the DNA carrier (vide infra).
a
lesser extent, the smaller lipoplexes were more timely
processed in the intracellular compartment.
Serum compatibility
Influence of the carbon chain length
A serious limitation of the cationic lipid-mediated gene delivery
is that transfection efficiency is drastically decreased in the
presence of serum proteins. This likely results from the
formation of the so-called "biocorona" that anionic proteins and
other serum components (fatty acids, heparin…) form around
the cationic transfection particles, provoking their destabilization
and/or stifling their cell uptake and endosome escape. Although
some formulations remain efficient at the level of serum content
typically used in cell culture (≤ 10 %), the effect of higher serum
levels has been scarcely investigated in prior studies.[29] To gain
an insight into the effect of serum on the intracellular delivery of
pDNA by erufosine prodrugs, EE12/DOPE/pCMV-Gluc lipoplexes
(1 equiv. DOPE, N/P 3) were incubated with cells in the
presence of an increasing concentration of serum, from 10 to
75 % (Figure 7). Considering the lower DNA dose (0.2 µg/well),
transfection rate was decreased by ca. two orders of magnitude
as serum content increased from 10 to 25 %, and in 50 % serum
(that roughly corresponds to concentration in blood), no
transgene expression was detected. When a higher DNA dose
(0.4 µg/well) was deposited onto the cells,
Selecting the acetal ester backbone for further structural
optimization, systematic variation of the length of the alkyl chain
tethered to the biolabile acetal moiety, from 2 to 18 carbon atoms,
was realized for investigating its influence on gene delivery
properties (Figure 5). Transfection rate clearly depended on alkyl
chain length, high transgene expression being obtained provided
the alkyl chain contained more than 2 and less than 18 carbon
atoms. Results distributed in a quasi-bell-shaped curve, with
maximum efficiency attained with the dodecanoyl chain (EE12).
Compound with the shorter chain (EE2) failed to mediate high
transgene expression, maybe in relation to the lack of stability of
the compound already aforementioned. At the opposite, degraded
transfection rate was observed for compounds with alkyl chains
beyond C16 (EE18), which was yet exacerbated by the presence of
an unsaturation in the chain (EE18:1). Similar results were obtained
in the 16HBE cell line (Figure S2). In this case, degradation of the
transfection rate with EE2 and EE18:1 was even more marked, which
confirmed the deleterious effect of the introduction of a short or
unsaturated chain on the carrier backbone.
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
12
1,E+12
10
11
1,E+11
10
10
1,E+10
10
9
1,E+09
10
108
1,E+08
107
N/P
1,E+07
.
1
1
1.25
1.25
1.5
1.5
1.75
1.75
2
2
2.25
2.25
2.5
2.5
2.75
2.75
3
3
4
C
C
i
i
i
i
i
i
i
i
i
i 4
EE12/DOPE (1/1)
Figure 6. Rate of transfection mediated by pro-erufosine EE12 at various charge ratios (N/P varying from 1 to 4), with increasing dose of pCMV-Gluc (0.1, 0.2, 0.3,
and 0.4 µg/well, from light blue to dark blue, resp.). Experiments were carried out on A549 cells, in the presence of serum (10 %). Lipoplexes were formulated with
DOPE (1 eq.) and control (C) refers to basal bioluminescence measured in untreated cells. Data shown are representative of a triplicate determination (mean ± SD).
10
1,E+10
compartments which makes them key players in various
membrane signaling pathways involved in carcinogenesis.[1c] In
the absence of serum, the lytic concentration for different APLs
has been shown to reflect their critical micellar concentration
(CMC) which are in the low µM range.[30] We thus determined the
CMC for erufosine and for the pro-erufosine compounds using the
fluorescent probe technique that has been extensively applied to
study surfactant micellization. Noteworthy, due to tethering of an
"additional" hydrophobic moiety to the erufosine scaffold, pro-
10
***
**
9
8
7
1,E+09
10
*
***
1,E+08
10
1,E+07
10
6
1,E+06
10
5
1,E+05
10
10
25
50
75
erufosine compounds display
a lower hydrophilic-lipophilic
Serum (%)
balance (HLB) than their parent compound and most likely
preferentially form lamellar structures in aqueous environment.
Consequently, the term critical aggregation concentration (CAC)
is suggested in place of CMC.[31] In this study, pyrene was
selected as the hydrophobic fluorescent dye. Fluorescence of
pyrene is sensitive to the polarity of the solubilizing medium and
the dye exhibits different fluorescence behaviors in hydrophilic
and lipophilic environments. Measurements were performed in
water and data obtained for erufosine and its prodrugs are
collected in Table 3. The CAC value measured for erufosine was
Figure 7. Transfection efficiency as
a function of serum content for
EE12/DOPE/pCMV-Gluc. Lipoplexes were prepared with DOPE (1 equiv.) at N/P
3. Experiments were carried out on A549 cells (6.000 cells/well) and cell
supernatant (containing 10 % serum) was replaced with culture medium
complemented with 10, 25, 50, and 75 % serum just before adding the
lipoplexes (0.2 and 0.4 µg of plasmid per well; light blue and dark blue bars,
resp.). Control (grey bars) refers to basal bioluminescence measured in
untreated cells. Data shown are representative of a triplicate determination
(mean ± SD). Statistically significant differences when compared to untreated
cells are indicated: *** p < 0.001, ** p < 0.01, or * p < 0.05.
Table 3. Critical aggregation concentration of erufosine and its prodrugs as
measured with pyrene by the fluorescent probe technique.
deleterious effect of serum was counteracted and significant
luciferase production still was measured in 50 % serum.
Entry
Compound CAC (µM)
Entry
Compound CAC (µM)
Intrinsic antitumor activity
E
0.7 ± 0.3
1
2
3
4
5
6
7
8
9
1.6 ± 0.5
10
11
E12
Pro-erufosine compounds have been designed so they may be
hydrolyzed in situ under a pH or enzymatic stimulus into erufosine,
EE2
7.7 ± 2.1
EE12
0.6 ± 0.2
1.1 ± 0.7
1.1 ± 0.2
0.9 ± 0.2
0.9 ± 0.6
1.0 ± 0.6
EE6
a
compound with intrinsic antitumor activity. In order to
1.1 ± 0.7
12
13
14
15
16
EE'12
EC12
characterize the antineoplastic effect of the pro-erufosine
compounds, we investigated their physical properties, hemolytic
activity, intrinsic cytotoxicity, and combined antitumor activity
when used as a carrier for intracellular delivery of a plasmid DNA
encoding a proapoptotic protein.
EE8
1.0 ± 0.3
EE10
0.8 ± 0.5
EC'12
EE18:1
EE'18:1
EE12
0.6 ± 0.2
EE14
1.2 ± 0.3
Self-assembly properties of the erufosine prodrugs
EE16
Alkylphospholipids are membrane-active compounds that can
disrupt phospholipid bilayers. They can thus interfere with
biological membranes delimitating the cellular and intracellular
1.3 ± 0.4
17
18
EC18:1
EC'18:1
0.8 ± 0.4
0.6 ± 0.2
EE18
1.4 ± 0.7
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
1.56 ± 0.47 µM and no earlier data was found in the literature.
Interestingly, CMC values below 10 nM have been suggested for
erucylphosphocholine (ErPC), an erufosine homolog compound.[32] It
cannot be excluded that replacement of the ethylene group with a
propylene one between the phosphate and ammonium moieties
can have an effect on the solvation state of the zwitterionic head
group of the lipid and, consequently, on the CMC of the compound.
However, if such an effect has been already reported, e.g., for
hexadecylphosphocholine (miltefosine) and hexadecylphospho-
homocholine (with CMC values of 13 and 20 µM, resp.),[33] it was
of lower magnitude. As expected, transformation of erufosine into
pro-erufosine compounds through the replacement of the phosphate
negative charge by a hydrophobic substituent translated into a
decrease in the CAC values. This was not the case however for
EE2. The reason for this is unclear. It might be proposed that the
introduction of an acetoxymethyl substituent (a short polyoxygenated
group that can establish hydrogen bonds with water molecules)
on the phosphate group of erufosine did not much change the
global hydrophobicity of the molecule while better complexation of
the trimethylammonium moiety by water molecules (as compared
to zwitterion) did enhance hydrophilicity of the polar head group,
thus improving aqueous solubility of the compound. CAC slightly
decreased with alkyl chain elongation up to C12, then a reverse
effect was observed up to C18 (Figure S3). This however must be
considered with caution, as standard deviations in CAC
measurements were significant. For the same reason, no trend
could be detected with respect to the structure of the biolabile
spacer (i.e., ester vs. carbonate and non-substituted vs.
substituted acetal), or to the introduction of an unsaturation in the
Table 4. Hemolytic activity of erufosine and pro-erufosine compounds upon
incubation with sheep RBC for 1 and 24 h at 37 °C.
Entry
Compound
1 h
24 h
HA200 (%)
HC50 (µM) HA200 (%)
HC50 (µM)
1
2
3
4
5
6
7
8
9
130
–
E
84.9 ± 4.6
25.0 ± 0.3
–
–
EE6
140
85.2 ± 6.2
E12
0.5 ± 0.4
< 0.5
–
–
–
–
–
12.0 ± 1.9
18.6 ± 1.5
8.7 ± 1.1
9.6 ± 2.3
7.3 ± 0.8
11.3 ± 4.0
1.2 ± 0.7
–
EE12
EE'12
EC12
EC'12
–
< 0.5
–
< 0.5
–
0.7 ± 0.2
–
EE18
< 0.5
–
–
–
EE18:1
1.0 ± 0.3
–
[a] Concentration of compound inducing 50 % hemolysis. [b] Percentage of
hemolysis provoked by the compound at the higher concentration tested
(200 µmol).
effect of most of the pro-erufosine compounds was below a
measurable amount (< 1 %) except for EE6 (HA200: 25.0 ± 0.3). In
order to better characterize the pro-erufosine compounds, the
hemolytic effect of two pro-erufosine compounds was measured
over extended periods of time (Figure 8 and Figure S5-S6). As
might be expected, hemolysis increased with time and reached
significant values after 18-24 h. Consequently, a 24-h incubation
time was chosen in the following evaluations so as to discriminate
against the pro-APLs (Figure S7). Even under these tough
incubation conditions, the pro-erufosine compounds revealed
only slightly harmful to RBC, with HA200 values below 20 % (Table
4). Again, EE6 stood as an exception displaying a HA200 value of
85.2 ± 6.2 %, whereas HC50 was around 140 µM. At the opposite,
no general trend was observed with regards to the structure of the
biolabile linker, and lower hydrolytic stability did not systematically
translate into higher hemolytic effect.
C18 chain. Differences were small and getting
a deeper
understanding of the structure-property relationships would
require a more precise determination of the CAC values.
Hemolytic activity
Hemolytic activity is associated with most of the antitumor APLs
described to date, so these compounds are not compatible with
intravenous administration.[30a] When APLs are used in liposomal
formulations (e.g., co-formulated with cholesterol and 1,2-
dipalmitoyl-sn-glycero-phosphoglycerol), lysis of red blood cells
can be partly prevented.[34] This however is accompanied by a
significant reduction in the antitumor activity of the compound, as
previously shown in vitro.[35] In serum-free conditions, the lytic
concentration for various APLs and related compounds generally
corresponds well to their CMC.[30] Decreasing the CMC of APLs
through their reversible transformation into more hydrophobic
derivatives should thus result in increasing their biocompatibility.
To investigate this point, the hemolytic activity of erufosine and
pro-erufosine compounds was determined using an erythrocyte
leakage assay. Sheep red blood cells (RBC) were treated with
increasing amounts of aqueous dispersion of the compounds at
37 °C and hemolysis (hemoglobin leakage) was monitored
spectrophotometrically. First, we investigated the hemolytic effect
of erufosine over an incubation period of 1 h. As expected, damage
to the RBC increased with the APL concentration. The concentration
provoking 50 % hemolysis (HC50) was around 130 µM and
84.9 ± 4.6 % hemolysis was achieved at 200 µM (HA200) (Table 4
and Figure S4). Under the same incubation conditions, hemolytic
25
20
15
10
5
0
0
5
10
15
20
25
Time (h)
Figure 8. Time course of hemolytic activity of pro-erufosine EE12 (◼) and E12
(▲). Sheep RBC were incubated at 37 °C with 200 µM of the compounds.
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
Considering these data, it was concluded that the
transformation of erufosine into prodrugs drastically reduces the
hemolytic activity, which makes the pro-erufosine compounds
highly compatible with intravenous administration.
Table 5. Antitumor activity (IC50) of erufosine and pro-erufosine compounds
in various cell lines exposed for 24 h at 37 °C, as measured by MTT dye
reduction assay.
IC50 (µM)
Entry
Compound
Cytotoxic effect
A549
H292
16HBE
The intrinsic cytotoxicity of APLs is not strictly related to their
ability to permeabilize membranes and lyse cells. It is widely
assumed that, at pharmacologically relevant concentration, APLs
1
2
E
57 ± 8
46 ± 8
19 ± 3
EE2
171 ± 44
174 ± 39
208 ± 66
insert into the cell membrane, provoking
a biophysical
disturbance through altered cholesterol homeostasis, and
perturbing various signal transduction pathways.[1c] In order to
estimate the antiproliferative potency of the pro-erufosine
compounds, a dose-response study was conducted in three
human pulmonary cell lines: A549 (alveolar carcinoma epithelial
cells), H292 (carcinoma mucoepidermoid cells), and 16HBE
(bronchial epithelial cells). Thirteen pro-erufosine compounds
were tested in parallel with parent erufosine. Cells were exposed
for 24 h to concentration of lipids between 500 nM and 1 mM, and
cell survival was determined using the MTT assay.[36] For the
three cell lines, a concentration dependent decrease in cell
viability was seen in response to erufosine and pro-erufosine
compounds (Figure S8-S10). Interpolated from the experimental
concentration–effect curves, the IC50 values (concentration that
inhibited the cell growth by 50 %) reflecting the cytotoxicity of the
compounds are presented in Table 5. The three cell lines showed
varying sensitivities to the compounds. The IC50 values ranged
from 57 to 199 µM for A549 cells, from 43 to 174 µM for H292
cells, and from 19 to 257 µM for non-cancerous 16HBE cells.
Cytotoxicity of the erufosine prodrugs did not parallel their
hydrolytic stability as determined by 31P-NMR measurements in
an "enzyme-free" model aqueous medium (vide supra). This likely
originates from the in situ enzymatic hydrolysis of the pro-APLs
compounds in the complemented culture medium which
competes with pure pH-controlled chemical hydrolysis. Depending
on the pro-APL, its structure and the nature of the substituent
installed on the phosphate group, the compound may or not be a
substrate for some specific lipases and, thus, may or may not be
metabolized into parent cytotoxic erufosine. However, all the pro-
erufosine compounds displayed marked cytotoxicity that roughly
compared to that of erufosine (i.e., with IC50 in the low µM range),
revealing that they were productively metabolized into erufosine,
though not at the same rate. It is noteworthy that EE2 that
distinguished by lower chemical stability as compared to the other
pro-erufosine compounds (vide supra) was among the less
cytotoxic prodrugs in the series, with IC50 values that were three
to ten times higher than that of erufosine, depending on the cell
line. Considering the prodrugs in the ester series (EEn) and except
for EE2 (vide supra), a general trend was that the IC50 value steadily
increased with the length of the pendant arm tethered to erufosine
(i.e., with n), up to C16. A longer chain (i.e., C18) tended to
decrease the IC50 value, and unsaturation in the C18 chain further
enhanced this fall. These effects were marked in the A549 cell
line (Figure S11), and the same trends were identified in the H292
and 16HBE cell lines, although all the compounds in these series
have not been systematically assayed. Finally, with regard to the
aggregation properties of the compounds, no relationship could
EE6
60 ± 8
88 ± 10
136 ± 40
81 ± 7
43 ± 7
–
58 ± 14
3
EE8
–
4
EE10
E12
–
–
5
124 ± 11
125 ± 26
73 ± 15
167 ± 56
161 ± 25
–
225 ± 31
144 ± 66
63 ± 7
252 ± 112
257 ± 81
–
6
EE12
EE’12
EC12
EC’12
EE14
EE16
151 ± 42
192 ± 44
161 ± 21
179 ± 63
193 ± 59
199 ± 30
7
8
9
10
11
12
13
14
–
–
EE18
170 ± 72
97 ± 35
141 ± 49
61 ± 20
165 ± 63
–
EE18:1
be established between CAC and IC50 values, thus confirming that
the cytotoxic effect of erufosine and pro-erufosine compounds
was not achieved through their membrane disrupting properties.
Combined antitumor activity of TRAIL and erufosine prodrugs
The use of combination therapy for cancer treatment is well
established.[37] Indeed, single-agent therapy aims to suspend one
signaling pathway and cancer cells can thrive through the initial
oncogenic route and activate another parallel signaling
pathway.[38] Multi-agent therapy can simultaneously modulate
several signaling pathways in diseased cells, maximizing the
therapeutic effect and, possibly, overcoming resistance
mechanisms. Whereas chemotherapy drugs are normally
associated with severe side-effects, administration of
a
combination of agents hitting various targets in different pathways
and displaying different toxicity profiles can improve the
therapeutic index, either in the form of better efficacy, or in the
form of comparable efficacy and reduced toxicity. In recent years,
combination anticancer therapy between nucleic acids and small-
molecule drugs has been recommended for cancer treatment
because of such advantages and because it can serve to
overcome genetic heterogeneity and existence of complicated
signaling pathways.[39] Besides, compared with coordinate
treatment with two drugs in separated carriers, the co-delivery of
two agents in one single carrier may allow resolving the
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
complexity of variation in pharmacokinetics and ensures the co-
localization of the two agents to maximize the additive or
synergistic effect, thereby reducing the amount of each drug and
promoting the efficacy of the combination therapy.[40] Our
hypothesis herein was that the combination of pro-APLs and
pDNA encoding a pro-apoptotic protein would diminish tumor cell
survival and show at least an additive antineoplastic effect.
Therefore, in order to investigate antitumor effect of the pro-
erufosine compounds as both gene delivery reagents and
prodrugs of antiproliferative erufosine, a plasmid DNA encoding
the tumor necrosis factor-related apoptosis-inducing ligand
(TRAIL, also called Apo2L or TNFSF10) was selected. As a
member of the tumor necrosis factor (TNF) superfamily, TRAIL
can induce apoptosis in malignant tumors while sparing normal
cells.[41] In humans, TRAIL activates extrinsic apoptotic pathway
after binding TRAIL death receptors (DRs), TRAIL-R1 (DR4,
TNFRSF10A) and TRAIL-R2 (DR5, TNFRSF10B), and TRAIL
decoy receptors that do not possess functional death domain,
TRAIL-R3 (DcR1, TNFRSF10C) and TRAIL-R4 (DcR2, TNFRSF10D).
Like most TNF superfamily members, TRAIL is a homotrimeric,
type II transmembrane protein. It can thus simultaneously recruit
several receptors. If the recruited receptors are all death receptors
(i.e., DR4 or DR5), a death-inducing signaling complex (DISC) is
formed that can strongly activate various caspases inducing
apoptosis. At the opposite, recruitment of inhibitor receptors (i.e.,
DcR1 or DcR2) leads to a DISC that cannot or only weakly
activate caspases and therefore impairs TRAIL-induced cell
death.[42] Nevertheless, the situation may be further complicated
by the cell-type dependency of the apoptotic response, itself
influenced by the effect on ligand binding mode of factors such as
the level of TRAIL oligomerization or glycosylation.[43]
was observed in A549 cells, since cell viability still was higher than
70 % at the SPK dose of 200 ng/mL. Once again these data are
consistent with those in the literature (IC50 > 1500 ng/mL).[45]
Taken together, these results suggest that induction of apoptosis
in the three cell models using a plasmid DNA encoding TRAIL
should be feasible, though A549 cells that expressed the higher
level of TRAIL-R1 appeared unexpectedly less sensitive to SPK
than H292 and 16HBE cells.
TRAIL-R1
TRAIL-R2
TRAIL-R3
TRAIL-R4
A549
H292
16HBE
Isotype control
TRAIL-R monoclonal antibodies (R1, R2, R3, R4)
Figure 9. Expression of the TRAIL receptors in the three cell lines as monitored
by flow cytometry. The fluorescence histograms were obtained using PE-
conjugated antibodies.
Characterization of TRAIL receptor expression and activity in the
different cell lines
Combined cytotoxic activity
Pro-erufosine EE12 was selected as the best pro-erufosine
candidate for the following experiments. The three cell lines were
treated for 24 h with lipoplexes prepared with either pCMV-Gluc
or pUNO1-hTRAIL, a plasmid DNA encoding TRAIL. The transfection
particles were prepared with one equivalent of DOPE at a charge
ratio of 3, corresponding to the previously established optimized
conditions (vide supra). The dose of pDNA applied to the cells
was varied from 0.1 to 0.4 µg DNA per well and incubation was
carried out for 24 h. For comparison, Lipofectamine® was assayed
in parallel.
Luciferase expression was determined in the three cell lines,
and A549 and H292 cells displayed very similar expression
profiles (Figure S13). On the other hand, 16HBE cells revealed a
little bit less responsive, since EE12-mediated transgene expression
was approximately four times weaker in this cell line, while
remaining at the higher rate with Lipofectamine®. In any case, the
transfection rate in the three cell lines was highly significant,
suggesting that replacing pCMV-Gluc with pUNO1-hTRAIL
should give rise to significant expression of TRAIL.
Characterization of the expression and activity of the TRAIL
receptors in the different cell lines was a prerequisite to validate
the cell models suitable for evaluation of combined antitumor
activity of TRAIL and erufosine prodrugs. Expression of the TRAIL
receptors was assessed by flow cytometry (Figure 9). Death
receptors TRAIL-R1 and TRAIL-R2 were expressed in the three
cell lines in similar amounts, except for TRAIL-R1 which
expression was significantly higher in A549 cells. TRAIL-R3 was
expressed only in H292 and 16HBE cells while TRAIL-R4 was not
expressed in any cell line. The sensitivity of the three cell lines to
TRAIL was then checked using SuperKillerTRAIL® (SPK), a
hexameric TRAIL recombinant protein providing significantly
enhanced apoptosis.[44] Cells were incubated with SPK for 24 h
and cell viability was subsequently measured (Figure S12). SPK
induced apoptosis of H292 and 16HBE cells in a dose-dependent
manner, with EC50 of 0.49 ± 0.18 and 0.68 ± 0.39 nM,
respectively (i.e., 12.7 ± 4.7 and 17.6 ± 10.2 ng/mL). Noteworthy,
the EC50 value for SPK in H292 cancer cells was significantly
lower than that reported for TRAIL (25 ± 5 ng/mL)[45] which is
consistent with the hexameric structure of SPK. Sensitivity of the
non-tumor 16HBE cells to TRAIL was unexpected and revealed
that recombinant SPK may not reflect properly all the properties
of monomeric TRAIL. By contrast, only low induction of apoptosis
Cell viability was monitored in the three cell lines treated with
pCMV-Gluc and pUNO1-hTRAIL (Figure 10). Control experiments
with Lipofectamine® revealed no decrease in cell viability using
pUNO1-hTRAIL as compared to pCMV-Gluc. Although this could
indicate that no functional TRAIL was produced, it is worth noting
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
cationic charge in DNA delivery experiments. We also showed
that these gene carriers are further degraded into parent
antitumor erufosine and thus express intrinsic antitumor
properties but are not associated with any hemolytic effect, as is
the case for erufosine. Finally, using a plasmid DNA encoding the
pro-apoptotic protein TRAIL, we evidenced selective toxicity
towards tumor cells while non-tumor cells were resistant.
Together, it appears that the combination approach involving well
tolerated erufosine cationic prodrugs and cancer gene therapy
holds significant promise and, e.g., might be useful in the
management of some drug-resistant tumors.
120
100
80
60
40
20
0
*
* *
* *
* *
*
L
0.1 0.2 0.3 0.4
L
0.1 0.2 0.3 0.4
L
0.1 0.2 0.3 0.4
A549
H292
16HBE
16HBE
A549
H292
Figure 10. Antiproliferative effect of lipoplexes prepared from EE12 or
Lipofectamine® 2000 (L), and pCMV-Gluc (light blue and white bars, resp.) or
pUNO1-hTRAIL (dark blue and black bars, resp.) on three cell lines. Pro-
erufosine-based lipoplexes were formulated with DOPE (1 eq.) at N/P 3, with a
pDNA dose varying from 0.1 to 0.4 µg/well. Lipofectamine-based lipoplexes
(0.1 µg plasmid/well) were prepared according to the optimized experimental
conditions provided by the supplier. Cell viability was measured by MTT assay.
Data shown are representative of a triplicate determination (mean ± SD).
Statistically significant differences are indicated: *** p < 0.001, ** p < 0.01, or
* p < 0.05.
Experimental Section
Materials
Unless otherwise stated, all chemical reagents were purchased from Alfa
Aesar (Bischeim, France) and used without purification. When required,
solvents were dried just before use as described elsewhere.[[48]] Thin layer
chromatography (TLC) was performed on precoated plates (0.25 mm
Silica Gel 60, F254, Merck, Darmstadt, Germany). Products were purified
by flash chromatography over silica gel (Silica Gel 60, 40-63 µm, Merck,
Darmstadt, Germany). NMR spectra were recorded on Bruker 400 MHz
Avance III instrument. 1H-, 13C-, and 31P-NMR chemical shifts δ are
reported in ppm relative to their standard reference (1H: CHCl3 at 7.27 ppm,
CD2HOD at 3.31 ppm; 13C: CDCl3 at 77.0 ppm, CD3OD at 49.0 ppm; 31P:
(MeO)2P(O)Me at 38.78 ppm). IR spectra were recorded on a FT-IR
Nicolet 380 spectrometer in the ATR mode and absorption values ν are in
wave numbers (cm-1). Mass Spectra (MS) were recorded on an Agilent
Technologies 6520 Accurate Mass QToF instrument, using electrospray
ionization (ESI) mode. Mass data are reported in mass units (m/z). 1,2-
Dioleoyl-sn-glycero-3- phosphoethanolamine (DOPE) was from Avanti
Polar Lipids (Alabaster, USA). Lipofectamine® 2000 was obtained from
Invitrogen (Cergy Pontoise, France). Plasmid pCMV-Gluc (5.7 kbp) and
coelenterazine substrate for monitoring Gaussia luciferase activity were
from Nanolight Technology (Pinetop, Az, USA). Plasmid pUNO1-hTRAIL
(4026 bp) was from InvivoGen (Toulouse, France) and SuperKillerTRAIL®
(SPK, 26 kDa) was from Enzo Life Sciences (Villeurbanne, France). PE-
conjugated anti-TRAIL antibodies used in flow cytometry analysis were
from Diaclone (Besançon, France). 3-(4,5-Dimethylthiazol-2-yl)-2,5-
diphenyl tetrazolium bromide (MTT) was from Sigma-Aldrich (Saint-
Quentin Fallavier, France). Culture media DMEM, DMEM/F12, RPMI, and
supplements were from GIBCO-BRL (Cergy-Pontoise, France). Fetal calf
serum (FCS) was from Perbio (Brebières, France). A549 cells (human lung
carcinoma; CCL-185), and NCI-H292 (H292) cells (human lung
mucoepidermoid carcinoma; CRL-1848) were obtained from ATCC-LGC
(Molsheim, France). The 16HBE14o- (16HBE) cells were a generous gift
from Dr D. Gruenert (California Pacific Medical Center Research Institute,
San Francisco, CA, USA). Defibrinated sheep blood was from Eurobio
(Les Ulis, France).
that intrinsic cytotoxicity of this transfection reagent was basically
very high both in A549 and H292 cells (with less than 40-50 % cell
survival), and thus could mask some effect of TRAIL. In the non-
tumor 16HBE cell line, no such intrinsic cytotoxicity of
Lipofectamine® could be invoked (~20 %), but data were
consistent with the resistance of non-tumor cells to TRAIL.[46] With
regards to the pro-erufosine pDNA delivery reagent EE12
,
statistically significant enhancement of cytotoxicity was observed
with pUNO1-hTRAIL, as compared to pCMV-Gluc, in A549 and
H292 cells, but not in 16HBE cells. This effect was clearly visible
from the intermediate plasmid dose of 0.2 ng/mL and indicated
that the transgene product was indeed functional and did express
intrinsic antineoplastic effect in the two tumor cell lines.
Resistance of the non-tumor 16HBE cells to TRAIL again was
consistent with results in the literature. Finally, the combination of
pro-erufosine compound EE12 and pUNO1-hTRAIL induced death
of A549 and H292 tumor cells selectively (cell death > 40-50 %)
while preserving normal 16HBE cells (cell death < 20-25 %).
Extensive work now is required on various testing platforms to
determine whether there is synergism between the pro-erufosine
compound and pUNO1-hTRAIL or simply additive anticancer
effect, and identify optimal dose combination as certain dose
ratios of combined drugs can be synergistic, while other ratios
may be antagonistic.[47]
Conclusion
Abbreviations: PE, petroleum ether; s, singlet; d, doublet; t, triplet; q,
quadruplet; b, broad.
In the present study, we designed and optimized cationic
prodrugs of erufosine, and investigated their properties as both
gene delivery reagents and antineoplastic drugs. We
demonstrated that these pro-erufosine compounds can efficiently
deliver plasmid DNA to mammal cells and that those incorporating
a pH- or enzyme-specific cleavage site were invariably more
potent –at least by an order of magnitude– than an analog
compound lacking such specific cleavage site. These results thus
demonstrate the superiority of biolabile lipids bearing transient
Synthesis
Erufosine (E). Triethylamine (1.30 mL, 9.27 mmol) was added dropwise to
phosphorus oxychloride (288 µL, 3.09 mmol) in anhydrous CH2Cl2 (10 mL)
at 0 °C under inert atmosphere. Then erucyl alcohol[[49]] (1.00 g,
3.09 mmol) in CH2Cl2 (15 mL) was added dropwise over a 90-min period
and the reaction mixture was allowed to warm to rt for 30 min, time after
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
which all the alcohol had reacted. Temperature was brought back to 0 °C
and 1,3-propanediol (223 µL, 3.09 mmol) in CH2Cl2 (15 mL) was added
dropwise within 60 min. The reaction mixture was stirred at rt for 16 h
before quenching by addition of 10 % HCl (10 mL). It was extracted twice
with CH2Cl2, and the organic layer was washed with brine, dried over
MgSO4, filtered, and reduced under vacuum. The crude residue was
purified by flash chromatography (Et2O) to yield 2-erucyloxy-1,3,2-
dioxaphosphinane 2-oxide (0.82 g, 60 %). Rf : 0.30 (Et2O). 1H-NMR (400
MHz, CDCl3) δ 0.88 (t, J = 6.6 Hz, 3H); 1.26 (m, 30H); 1.61 and 1.77 (2
m, 1H); 1.72 (m, 2H); 2.00 (m, 4H); 2.26 (dtt, J1 = 15.3 Hz, J2 = 10.3 Hz,
J3 = 5.3 Hz, 1H); 4.09 (td, J1 = J2 = 6.9 Hz, 2H); 4.32-4.45 (m, 4H); 5.34 (m,
2H). 13C-NMR (100.7 MHz, CDCl3) δ 14.3; 22.9; 25.7; 26.4; 27.4 (2C);
29.4; 29.5 (3C); 29.7 (2C); 29.8 (4C); 30.0 (2C); 30.5; 32.1; 67.8;
68.6 (2C); 130.1 (2C). 31P-NMR (162 MHz, CDCl3) δ – 7.08. IR n 457; 594;
721; 771; 828; 883; 934; 966; 1029; 1141; 1237; 1297; 1464; 2852; 2921.
The previous compound (200 mg, 0.45 mmol) in CHCl3/CH3CN/i-
PrOH 3:5:5 (8 mL) was treated with 45 % (w/w) aqueous trimethylamine
(3 mL, 20.2 mmol) at 70 °C. After 24 h, all starting material was consumed
(checked by TLC). Volatile was removed under vacuum, the residue was
extracted twice with CHCl3/MeOH 9/1 and the organic layer was washed
with brine, reduced under vacuum, diluted with CHCl3, dried over MgSO4,
filtered and evaporated. The crude residue was purified by flash
chromatography (CH2Cl2/MeOH/H2O 75:22:3 to 45:45:10) to yield
erufosine (123 mg, 75 %). Rf : 0.28 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR
(400 MHz, CD3OD/CDCl3 1:1) δ 0.86 (t, J = 6.2 Hz, 3H); 1.25 (m, 30H);
1.60 (tt, J1 = J2 = 6.8 Hz, 2H); 2.00 (m, 4H); 2.04 (m, 2H); 3.11 (s, 9H);
3.47 (m, 2H); 3.82 (td, J1 = J2 = 6.9 Hz, 2H); 3.90 (td, J1 = J2 = 6.1 Hz, 2H);
5.31 (m, 2H). 13C-NMR (100.7 MHz, CD3OD/CDCl3 1:1) δ 14.3; 23.2; 25.1;
26.4; 27.7 (2C); 29.8 (2C); 29.9; 30.0; 30.1 (2C); 30.2 (4C); 30.3 (2C);
31.4; 32.4; 53.4 (3C); 61.9; 64.9; 66.2; 130.3 (2C). 31P-NMR (162 MHz,
CD3OD/CDCl3 1:1) δ + 0.66. IR n 528; 553; 721; 840; 945; 1059; 1092;
1223; 1466; 1480; 1640; 2850; 2919; 3271. HR-MS (ESI+) m/z [M+H]+
calcd for C28H59NO4P+ 504.4176, found 504.4193.
1.65 (tt, J1 = J2 = 7.2 Hz, 2H); 2.36 (t, J = 7.5 Hz, 2H); 5.68 (s, 2H). 13C-
NMR (100.7 MHz, CDCl3) δ 14.1; 22.5; 24.4; 31.3; 34.2; 68.8; 172.0. IR
n 718; 773; 988; 1039; 1094; 1137; 1223; 1245; 1708; 1761; 2931; 2957.
Chloromethyl octanoate. Yield: 82 %. Rf : 0.75 (PE/Et2O 90:10). 1H-NMR
(400 MHz, CDCl3)
δ 0.88 (t, J = 6.6 Hz, 3H); 1.30 (m, 8H);
1.65 (tt, J1 = J2 = 7.2 Hz, 2H); 2.38 (t, J = 7.5 Hz, 2H); 5.68 (s, 2H). 13C-
NMR (100.7 MHz, CDCl3) δ 14.2; 22.7; 24.7; 30.0 (2C); 31.7; 34.1; 68.7;
171.9. IR n 462; 716; 1038; 1102; 1136; 1260; 1441; 1763; 2856; 2926.
Chloromethyl decanoate. Yield: 80 %. Rf : 0.75 (PE/Et2O 90:10). 1H-NMR
(400 MHz, CDCl3)
δ 0.88 (t, J = 6.8 Hz, 3H); 1.26 (m, 12H);
1.65 (tt, J1 = J2 = 7.3 Hz, 2H); 2.38 (t, J = 7.5 Hz, 2H); 5.70 (s, 2H). 13C-
NMR (100.7 MHz, CDCl3) δ 14.3; 22.9; 24.8; 29.2; 29.4 (2C); 29.6; 32.1;
34.2; 68.8; 172.0. IR n 463; 718; 758; 1034; 1105; 1135; 1260; 1338; 1441;
1465; 1764; 2854; 2924.
Chloromethyl tetradecanoate. Yield: 68 %. Rf : 0.75 (PE/Et2O 90:10). 1H-
NMR (400 MHz, CDCl3) δ 0.88 (t, J = 6.4 Hz, 3H); 1.26 (m, 20H);
1.65 (tt, J1 = J2 = 7.2 Hz, 2H); 2.38 (t, J = 7.5 Hz, 2H); 5.70 (s, 2H). 13C-
NMR (100.7 MHz, CDCl3) δ 14.3; 22.8; 24.7; 29.1; 29.3; 29.5; 29.6; 29.7;
29.8 (3C); 32.1; 34.2; 68.7; 172.0. IR n 463; 719; 758; 1036; 1108; 1135;
1217; 1340; 1441; 1465; 1765; 2853; 2922.
Chloromethyl hexadecanoate. Yield: 71 %. Rf : 0.75 (PE/Et2O 90:10). 1H-
NMR (400 MHz, CDCl3) δ 0.88 (t, J = 6.8 Hz, 3H); 1.26 (m, 24H);
1.65 (tt, J1 = J2 = 7.2 Hz, 2H); 2.38 (t, J = 7.6 Hz, 2H); 5.70 (s, 2H). 13C-
NMR (100.7 MHz, CDCl3) δ 14.3; 22.8; 24.7; 29.1; 29.3; 29.5; 29.6; 29.7;
29.8 (5C); 32.1; 34.2; 68.7; 171.9. IR n 437; 686; 718; 772; 994; 1017;
1035; 1098; 1147; 1192; 1216; 1260; 1383; 1438; 1461; 1472; 1755; 2848;
2915; 2964.
Chloromethyl octadecanoate. Yield: 76 %. Rf : 0.6 (PE/Et2O 90:10). 1H-
NMR (400 MHz, CDCl3) δ 0.90 (t, J = 6.7 Hz, 3H); 1.32 (m, 28H);
1.65 (tt, J1 = J2 = 7.0 Hz, 2H); 2.36 (t, J = 7.5 Hz, 2H); 5.68 (s, 2H). 13C-
NMR (100.7 MHz, CDCl3) δ 14.3; 22.8; 24.7; 29.1; 29.3; 29.5; 29.6; 29.7;
29.8 (7C); 32.1; 34.2; 68.7; 171.9. IR n 686; 718; 793; 1017; 1035; 1045;
1099; 1147; 1191; 1259; 1383; 1438; 1462; 1472; 1755; 2848; 2916; 2959.
Preparation of chloromethyl esters and 1-chloroethyl esters
Typical procedure: Chloromethyl dodecanoate. Dodecanoyl chloride
(3.5 mL, 14.7 mmol) was added dropwise over 10 min to a mixture of
anhydrous zinc chloride (45 mg, 0.29 mmol) and paraformaldehyde
(663 mg, 22.1 mmol) at 0 °C under argon. After the addition was complete,
the reaction mixture was stirred for 1 h at 0 °C, and allowed to warm to
room temperature overnight. The crude mixture was directly purified by
flash chromatography (PE/Et2O 100:0 to 95:5) to provide chloromethyl
dodecanoate (2.61 g, 71 %) as a colorless oil. Rf : 0.6 (PE/Et2O 95:5). 1H-
NMR (400 MHz, CDCl3) δ 0.88 (t, J = 6.8 Hz, 3H); 1.26 (m, 16H);
1.65 (tt, J1 = J2 = 7.2 Hz, 2H); 2.38 (t, J = 7.5 Hz, 2H); 5.70 (s, 2H). 13C-
NMR (100.7 MHz, CDCl3) δ 14.3; 22.8; 24.7; 29.1; 29.3; 29.5; 29.6;
29.7 (2C); 32.1; 34.2; 68.7; 171.8. IR n 462; 718; 771; 1034; 1106; 1135;
1338; 1441; 1465; 1765; 2853; 2922.
Chloromethyl oleate. Yield: 60 %. Rf : 0.5 (PE/Et2O 95:5). 1H-NMR (400
MHz, CDCl3) δ 0.88 (t, J = 6.7 Hz, 3H); 1.29 (m, 20H); 1.65 (tt, J1 = J2 =
7.1 Hz, 2H); 2.01 (m, 4H); 2.38 (t, J = 7.5 Hz, 2H); 5.34 (m, 2H); 5.70 (s,
2H). 13C-NMR (100.7 MHz, CDCl3) δ 14.3; 22.8; 24.7; 27.3 (2C); 29.1; 29.2
(2C); 29.5 (2C); 29.7; 29.8; 29.9; 32.1; 34.1; 68.7; 129.9; 130.2; 171.9. IR
n 462; 719; 771; 1036; 1093; 1114; 1219; 1260; 1338; 1440; 1463; 1766;
2853; 2922; 3003.
1-Chloroethyl oleate. Paraformaldehyde was replaced by acetaldehyde.
Yield: 27 %. Rf : 0.5 (PE/Et2O 95:5). 1H-NMR (400 MHz, CDCl3) δ 0.88 (t,
J = 6.7 Hz, 3H); 1.28 (m, 20H); 1.64 (tt, J1 = J2 = 7.1 Hz, 2H);
1.78 (d, J = 5;8 Hz, 3H); 2.01 (m, 4H); 2.34 (t, J = 7.5 Hz, 2H);
5.34 (m, 2H); 6.55 (q, J = 5.8 Hz, 1H). 13C-NMR (100.7 MHz, CDCl3) δ
14.3; 22.8; 24.7; 25.4; 27.4 (2C); 29.1; 29.2 (2C); 29.5 (2C); 29.7; 29.8;
29.9 ; 32.1; 34.3; 80.7; 129.9; 130.2; 171.6. IR n 666; 721; 772; 934; 1019;
1083; 1137; 1220; 1272; 1378; 1463; 1761; 2853; 2922; 3004.
1-Chloroethyl dodecanoate. Paraformaldehyde was replaced by
acetaldehyde. Yield: 42 %. Rf : 0.6 (PE/Et2O 95:5). 1H-NMR (400 MHz,
CDCl3) δ 0.88 (t, J = 6.7 Hz, 3H); 1.26 (m, 16H); 1.64 (tt, J1 = J2 = 7.3 Hz,
2H); 1.78 (d, J = 5.8 Hz, 3H); 2.34 (t, J = 7.5 Hz, 2H); 6.55 (q, J = 5.8 Hz,
1H). 13C-NMR (100.7 MHz, CDCl3) δ 14.3; 22.9; 24.8; 25.4; 29.2; 29.4;
29.5; 29.6; 29.7; 29.8; 32.1; 34.4; 80.8; 171.8. IR n 664; 721; 771; 934;
1019; 1084; 1110; 1142; 1219; 1272; 1378; 1465; 1760; 2853; 2923.
Preparation of chloromethyl carbonates and 1-chloroethyl carbonates
1
Chloromethyl acetate. Yield: 90 %. H-NMR (400 MHz, CDCl3) δ 2.14 (s,
3H); 5.68 (s, 2H). 13C-NMR (100.7 MHz, CDCl3) δ 20.8; 68.7; 172.0. IR
n 461; 598; 713; 818; 982; 1015; 1041; 1194; 1371; 1712; 1762.
Typical procedure: Chloromethyl dodecyl carbonate. To a solution of
dodecanol (3.02 g, 16.2 mmol) and anhydrous Et3N (4.5 mL, 32.0 mmol)
in Et2O (35 mL) at 0 °C under argon, chloromethylchloroformate (2.2 mL,
24.7 mmol) in anhydrous Et2O (10 mL) was added dropwise over 15 min.
The resulting mixture was warmed to rt and stirred for 2 h. Precipitate was
Chloromethyl hexanoate. Yield: 77 %. Rf : 0.75 (PE/Et2O 75:25). 1H-NMR
(400 MHz, CDCl3)
δ 0.90 (t, J = 6.6 Hz, 3H); 1.32 (m, 4H);
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
discarded by filtration and the filtrate was reduced under vacuum. The
crude residue was purified by flash column chromatography over silica gel
(PE/Et2O 100:0 to 95:5) to yield chloromethyl dodecyl carbonate (2.60 g,
58 %) as a clear oil. TLC Rf 0.50 (PE/Et2O 95:5). 1H-NMR (400 MHz,
CDCl3) δ 0.88 (t, J = 6.7 Hz, 3H); 1.26 (m, 18H); 1.70 (tt, J1 = J2 = 6.8 Hz,
2H); 4.22 (t, J = 6.7 Hz, 2H); 5.73 (s, 2H). 13C-NMR (100.7 MHz, CDCl3) δ
14.3; 22.9; 25.8; 28.7; 29.3; 29.5; 29.6; 29.7; 29.8 (2C); 32.1; 69.6; 72.3;
153.6. IR n 718; 787; 961; 1113; 1239; 1343; 1443; 1765; 2853; 2922.
δ 0.86 (t, J = 6.8 Hz, 6H); 1.25 (m, 46H); 1.63 (tt, J1 = J2 = 6.8 Hz, 2H);
1.68 (tt, J1 = J2 = 7.4 Hz, 2H); 1.99 (m, 4H); 2.19 (m, 2H); 2.40 (t,
J = 7.5 Hz, 2H); 3.16 (s, 9H); 3.49 (m, 2H); 4.07 (td, J1 = J2 = 6.9 Hz, 2H);
4.17 (td, J1 = J2 = 6.9 Hz, 2H); 5.31 (m, 2H); 5.62 (ABX syst., JAB = 5.2 Hz,
JAX = 13.4 Hz, JBX = 11.4 Hz, 2H). 13C-NMR δ (100.7 MHz, CD3OD/CDCl3
1:1) δ 14.5 (2C); 23.3 (2C); 24.8; 25.3; 26.1; 27.8 (2C); 29.7; 29.8; 30.0
(5C); 30.2; 30.3 (6C); 30.4 (2C); 30.5 (2C); 30.9; 32.6 (2C); 34.6; 53.7 (3C);
64.3; 65.5; 69.9; 83.5; 130.5; 130.6; 173.2. 31P-NMR (162 MHz,
CD3OD/CDCl3 1:1) δ – 3.38. IR n 491; 721; 757; 858; 968; 1029; 1116;
1157; 1261; 1466; 1760; 2850; 2919; 2956; 3389. HR-MS (ESI+) m/z [M-
Cl]+ calcd for C41H83NO6P+ 716.5953, found 716.5954.
1-Chloroethyl dodecyl carbonate. Chloromethyl chloroformate was
replaced by 1-chloroethyl chloroformate. Yield: 58 %. Rf : 0.6 (PE/Et2O
95:5). 1H-NMR (400 MHz, CDCl3) δ 0.88 (t, J = 6.8 Hz, 3H); 1.26 (m, 18H);
1.68 (tt, J1 = J2 = 6.8 Hz, 2H); 1.83 (d, J = 5.8 Hz, 3H); 4.20 (t, J = 6.7 Hz,
2H); 6.43 (q, J = 5.8 Hz, 1H). 13C-NMR (100.7 MHz, CDCl3) δ 14.3; 22.9;
25.4; 25.8; 28.7; 29.3; 29.5; 29.6; 29.7; 29.8 (2C); 32.1; 69.3; 84.7; 153.1.
IR n 662; 785; 892; 1008; 1067; 1111; 1243; 1348; 1385; 1466; 1762;
2853; 2923.
3-(((1-(Dodecanoyloxy)ethoxy)(erucyloxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EE'12). Compound EE'12 (152 mg,
50 %) was obtained as a mixture of four diastereomers from erufosine
(201 mg, 0.40 mmol) and 1-chloroethyl dodecanoate (841 mg, 3.20 mmol)
according to the general procedure. Rf : 0.4 and 0.5 (CH2Cl2/MeOH/H2O
75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3 1:1) δ 0.86 (t, J = 6.8 Hz, 6H);
1.25 (m, 46H); 1.56 (d, J = 5.2 Hz, 3H); 1.63 (tt, J1 = J2 = 6.8 Hz, 2H); 1.68
(tt, J1 = J2 = 7.4 Hz, 2H); 1.99 (m, 4H); 2.19 (m, 2H); 2.36 (t, J = 7.4 Hz,
2H); 3.17 (s, 9H); 3.52 (m, 2H); 4.07 (td, J1 = J2 = 6.9 Hz, 2H); 4.17 (m,
2H); 5.30 (m, 2H); 6.44 (qd, J1 = J2 = 5.2 Hz, 1H). 13C-NMR (100.7 MHz,
CD3OD/CDCl3 1:1) δ 14.4 (2C); 21.7; 23.3 (2C); 24.8; 25.3; 26.1; 27.8
(2C); 29.7; 29.8; 30.0 (5C); 30.2 (4C); 30.3 (5C); 30.4 (2C); 30.9; 32.6 (2C);
34.7; 53.7 (3C); 64.3; 65.2; 69.6; 92.0; 130.5 (2C); 173.7. 31P-NMR δ (162
MHz, CD3OD/CDCl3 1:1) δ – 5.15 and – 5.85. IR n 515; 721; 850; 969;
1041; 1090; 1167;1269; 1379; 1466; 1753; 2850; 2919; 3388. HR-MS
(ESI+) m/z [M-Cl]+ calcd for C42H85NO6P+ 730.6109, found 730.6120.
Chloromethyl oleyl carbonate. Yield: 80 %. Rf : 0.6 (PE/Et2O 95:5). 1H-
NMR (400 MHz, CDCl3) δ 0.88 (t, J = 6.7 Hz, 3H); 1.28 (m, 22H);
1.69 (tt, J1 = J2 = 7.0 Hz, 2H); 2.01 (m, 4H); 4.22 (t, J = 6.8 Hz, 2H); 5.34
(m, 2H); 5.72 (s, 2H). 13C-NMR (100.7 MHz, CDCl3) δ 14.3; 22.8; 25.7;
27.4 (2C); 28.6; 29.3 (2C); 29.5 (2C); 29.7 (2C); 29.8; 29.9; 31.9; 69.5;
72.3; 129.9; 130.2; 153.6. IR n 719; 759; 787; 965; 1114; 1243; 1343;
1443; 1766; 2853; 2922; 3004.
1-Chloroethyl oleyl carbonate. Chloromethyl chloroformate was replaced by
1
1-chloroethyl chloroformate. Yield: 85 %. Rf : 0.6 (PE/Et2O 95:5). H-NMR
(400 MHz, CDCl3)
δ 0.88 (t, J = 6.7 Hz, 3H); 1.29 (m, 22H);
3-((((((Dodecyloxy)carbonyl)oxy)methoxy)(erucyloxy)phosphoryl)oxy)-
N,N,N-trimethylpropan-1-aminium chloride (EC12). Compound EC12 (97 mg,
31 %) was obtained from erufosine (200 mg, 0.40 mmol) and chloromethyl
dodecyl carbonate (884 mg, 3.17 mmol) according to the general
procedure. Rf : 0.45 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz,
CD3OD/CDCl3 1:1) δ 0.88 (t, J = 6.6 Hz, 6H); 1.27 (m, 48H); 1.68 (m,
4H); 2.00 (m, 4H); 2.21 (m, 2H); 3.17 (s, 9H); 3.50 (m, 2H);
4.10 (td, J1 = J2 = 6.8 Hz, 2H); 4.17 (t, J = 6.4 Hz, 2H); 4.25 (m, 2H); 5.32
(m, 2H); 5.65 (m, 2H). 13C-NMR δ (100.7 MHz, CD3OD/CDCl3 1:1) δ 14.6
(2C); 23.6 (2C); 25.0; 26.3; 26.6; 28.0 (2C); 29.5; 30.1; 30.2 (2C); 30.3
(2C); 30.4; 30.5 (4C); 30.6 (6C); 30.7 (2C); 31.1; 32.8 (2C); 53.8 (3C);
64.5; 65.9; 70.2 (2C); 86.8; 130.7; 130.7; 156.0. 31P-NMR (162 MHz,
CD3OD/CDCl3 1:1) δ – 3.06. IR n 490; 721; 790; 860; 946; 1024; 1115;
1257; 1466; 1761; 2852; 2920; 3399. HR-MS (ESI+) m/z [M-Cl]+ calcd for
C42H85NO7P+ 746.6058, found 746.6069.
1.69 (tt, J1 = J2 = 7.0 Hz, 2H); 1.83 (d, J = 5.8 Hz, 3H); 2.01 (m, 4H); 4.20
(t, J = 6.8 Hz, 2H); 5.34 (m, 2H); 6.43 (q, J = 5.8 Hz, 1H). 13C-NMR (100.7
MHz, CDCl3) δ 14.3; 22.8; 25.4; 25.8; 27.4 (2C); 28.7; 29.3 (2C); 29.5 (2C);
29.7; 29.9 (3C); 31.9; 69.3; 84.6; 129.9; 130.2; 153.1. IR n 663; 722; 758;
894; 1008; 1111; 1245; 1348; 1384; 1464; 1763; 2853; 2922; 3004.
Preparation of erufosine prodrugs
General procedure. Electrophilic reagent (4.0 mmol) and erufosine
(0.5 mmol) were reacted in refluxing anhydrous CHCl3 (12 mL) for 24 h
with stirring under inert atmosphere. The solvent was removed under
vacuum and the crude residue was purified by flash chromatography
(CH2Cl2/MeOH 10:0 to 7:3) to yield the corresponding pro-erufosine.
3-(((Dodecyloxy)(erucyloxy)phosphoryl)oxy)-N,N,N-trimethylpropan-1-
aminium triflate (E12). Compound E12 (139 mg, 97 %) was obtained from
erufosine (102 mg, 0.16 mmol) and dodecyl triflate[[50]] (170 mg,
0.53 mmol) according to the general procedure except the reaction was
conducted at rt. Rf : 0.7 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz,
CDCl3) δ 0.88 (t, J = 6.6 Hz, 6H); 1.26 (m, 48H); 1.67 (tt, J1 = J2 = 6.8 Hz,
4H); 2.02 (m, 4H); 2.20 (m, 2H); 3.24 (s, 9H); 3.55 (m, 2H);
4.04 (td, J1 = J2 = 6.8 Hz, 4H); 4.14 (td, J1 = J2 = 6.8 Hz, 2H); 5.35 (m, 2H).
13C-NMR (100.7 MHz, CDCl3) δ 14.4 (2C); 22.8 (2C); 24.5; 25.6 (2C);
27.4 (2C); 29.3 (2C); 29.5 (4C); 29.8 (5C); 29.9 (5C); 30.0 (2C); 30.4 (2C);
32.1 (2C); 53.6 (3C); 63.8; 64.1; 68.5; 68.6; 130.0; 130.1. 31P-NMR δ (162
MHz, CDCl3) δ – 1.43. IR n 516; 572; 638; 772; 866; 1030; 1158; 1225;
1257; 1465; 2852; 2921; 3500. HR-MS (ESI+) m/z [M-Cl]+ calcd for
C40H83NO4P+ 672.6054, found 672.6039.
3-(((1-((((Dodecyloxy)carbonyl)oxy)ethoxy))(erucyloxy)phosphoryl)oxy)-
N,N,N-trimethylpropan-1-aminium chloride (EC'12). Compound EC'12
(128 mg, 41 %) was obtained as a mixture of four diastereomers from
erufosine (199 mg, 0.39 mmol) and 1-chloroethyl dodecyl carbonate
(923 mg, 3.15 mmol) according to the general procedure. Rf : 0.45
(CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3 1:1) δ
0.86 (t, J = 6.8 Hz, 6H); 1.25 (m, 48H); 1.59 (d, J = 5.2 Hz, 3H);
1.67 (m, 4H); 2.00 (m, 4H); 2.19 (m, 2H); 3.17 (s, 9H); 3.50 (m, 2H); 4.07
(m, 2H); 4.16 (m, 2H); 4.20 (m, 2H); 5.30 (m, 2H); 6.34 (m, 1H). 13C-NMR
δ (100.7 MHz, CD3OD/CDCl3 1:1) δ 14.6 (2C); 21.5; 23.6 (2C); 25.0; 26.3;
26.6; 28.0 (2C); 29.5; 29.9 (3C); 30.0 (3C); 30.2 (2C); 30.3 (2C); 30.4 (6C);
30.5 (2C); 30.8; 32.6 (2C); 53.7 (3C); 64.3; 65.3; 69.8 (2C); 95.5; 130.5
(2C); 156.8. 31P-NMR (162 MHz, CD3OD/CDCl3 1:1) δ – 5.16 and – 5.31.
IR n 515; 720; 789; 889; 969; 1030; 1259; 1393; 1465; 1757; 2852; 2921;
3387. HR-MS (ESI+) m/z [M-Cl]+ calcd for C43H87NO7P+ 760.6215, found
760.6221.
3-((((Dodecanoyloxy)methoxy)(erucyloxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EE12). Compound EE12 (124 mg,
41 %) was obtained from erufosine (201 mg, 0.40 mmol) and chloromethyl
dodecanoate (791 mg, 3.17 mmol) according to the general procedure. Rf :
0.5 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3 1:1)
3-(((Erucyloxy)((oleoyloxy)methoxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EE18:1). Compound EE18:1 (115 mg,
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
1
35 %) was obtained from erufosine (200 mg, 0.40 mmol) and chloromethyl
oleate (1.05 g, 3.17 mmol) according to the general procedure. Rf : 0.65
(CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CDCl3) δ 0.88 (t,
J = 7.0 Hz, 6H); 1.27 (m, 50H); 1.63 (tt, J1 = J2 = 6.8 Hz, 2H); 1.68 (tt,
J1 = J2 = 7.4 Hz, 2H); 2.00 (m, 8H); 2.28 (m, 2H); 2.38 (t, J = 7.5 Hz, 2H);
3.48 (s, 9H); 3.81 (m, 2H); 4.07 (td, J1 = J2 = 6.9 Hz; 2H); 4.23 (m, 2H);
5.34 (m, 4H); 5.62 (ABX syst., JAB = 4.8 Hz, JAX = 12.7 Hz, JBX = 10.9 Hz,
2H). 13C-NMR δ (100.7 MHz, CDCl3) δ 14.3 (2C); 22.9 (2C); 24.5 (C); 24.7
(C); 25.6; 27.4 (4C); 29.2; 29.3; 29.4; 29.5 (7C); 29.7 (5C); 29.8 (3C); 29.9
(2C); 30.4; 32.1 (2C); 34.2; 53.8 (3C); 63.7; 64.7; 69.1; 82.8; 129.8; 130.0;
130.1; 130.3; 172.4. 31P-NMR (162 MHz, CDCl3) δ – 3.42. IR n 495; 721;
796; 844; 970; 1055; 1146; 1241; 1465; 1647; 1750; 2852; 2921; 3387.
HR-MS (ESI+) m/z [M-Cl]+ calcd for C47H93NO6P+ 798.6735, found
798.6739.
2H); 4.02-4.23 (m, 6H); 5.35 (m, 4H); 6.35 (qd, J1 = J2 = 5.2 Hz, 1H). H-
NMR (400 MHz, CDCl3) E2 δ 0.88 (t, J = 6.6 Hz, 6H); 1.26 (m, 52H); 1.56
(d, J = 5.2 Hz, 3H); 1.67 (tt, J1 = J2 = 6.8 Hz, 4H); 2.00 (m, 8H); 2.28 (m,
2H); 3.49 (s, 9H); 3.81 (m, 2H); 4.02-4.23 (m, 6H); 5.33 (m, 4H); 6.41 (qd,
J1 = J2 = 5.2 Hz, 1H). 13C-NMR δ (100.7 MHz, CDCl3) E1 δ 14.3 (2C); 21.6;
22.8 (2C); 24.5; 25.5; 25.8; 27.4 (4C); 28.7; 29.3; 29.4 (2C); 29.5 (6C);
29.6 (3C); 29.7 (2C); 29.8 (3C); 29.9 (3C); 30.1; 31.9 (2C); 53.7 (3C); 63.7;
64.3; 68.9; 69.2; 94.7; 129.9; 130.0; 130.1; 130.2; 153.5. 13C-NMR δ
(100.7 MHz, CDCl3) E2 δ 14.3 (2C); 21.6; 22.8 (2C); 24.5; 25.5; 25.8; 27.4
(4C); 28.8; 29.4 (3C); 29.5 (6C); 29.6 (3C); 29.7 (5C); 29.9 (3C); 30.1; 31.9
(2C); 53.7 (3C); 63.7; 64.3; 69.1 (2C); 94.6; 129.9; 130.0; 130.1; 130.2;
153.3. 31P-NMR (162 MHz, CDCl3) δ – 5.00 (E1); – 5.71 (E2). IR n 512;
721; 787; 887; 969; 1032; 1259; 1392; 1465; 1757; 2852; 2921; 3379. HR-
MS (ESI+) m/z [M-Cl]+ calcd for C49H97NO7P+ 842.6997, found 842.7005
(E1) and 842.7001 (E2).
3-(((Erucyloxy)(1-(oleoyloxy)ethoxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EE'18:1). Compound EE'18:1 (158 mg,
47 %) was obtained as two separated couples of enantiomers from
erufosine (200 mg, 0.40 mmol) and 1-chloroethyl oleate (995 mg,
2.88 mmol) according to the general procedure. Rf : 0.7 (E1) and 0.65 (E2)
(CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3 1:1) E1
δ 0.86 (t, J = 7.0 Hz, 6H); 1.25 (m, 50H); 1.55 (d, J = 5.3 Hz, 3H); 1.61 (tt,
J1 = J2 = 6.8 Hz, 2H); 1.68 (tt, J1 = J2 = 7.4 Hz, 2H); 2.00 (m, 8H); 2.18 (m,
2H); 2.36 (t, J = 7.4 Hz, 2H); 3.18 (s; 9H); 3.52 (m, 2H); 4.04 (td,
J1 = J2 = 6.9 Hz, 2H); 4.16 (m, 2H); 5.31 (m, 4H); 6.44 (dq, J1 = J2 = 5.3 Hz,
1H). 1H-NMR (400 MHz, CD3OD/CDCl3 1:1) E2 δ 0.86 (t, J = 7.0 Hz, 6H);
1.25 (m, 50H); 1.55 (d, J = 5.3 Hz, 3H); 1.61 (tt, J1 = J2 = 6.8 Hz, 2H); 1.68
(tt, J1 = J2 = 7.4 Hz, 2H); 2.00 (m, 8H); 2.18 (m, 2H); 2.36 (t, J = 7.4 Hz,
2H); 3.18 (s, 9H); 3.52 (m, 2H); 4.04 (td, J1 = J2 = 6.9 Hz, 2H); 4.16 (m,
2H); 5.31 (m, 4H); 6.45 (dq, J1 = J2 = 5.3 Hz, 1H). 13C-NMR δ (100.7 MHz,
CD3OD/CDCl3 1:1) E1 δ 14.5 (2C); 21.7; 23.3 (2C); 24.8; 25.3; 26.0; 27.8
(4C); 29.7; 29.8 (3C); 30.0 (3C); 30.2 (4C); 30.3 (5C); 30.4 (4C); 30.9; 32.6
(2C); 34.7; 53.7 (3C); 64.3; 65.2; 69.6; 92.0; 130.3; 130.5; 130.6; 130.7;
173.0. 13C-NMR δ (100.7 MHz, CD3OD/CDCl3 1:1) E2 δ 14.5 (2C); 21.7;
23.3 (2C); 24.8; 25.1; 26.0; 27.8 (4C); 29.7; 29.8 (3C); 30.0 (3C); 30.2 (4C);
30.3 (5C); 30.4 (4C); 30.9; 32.6 (2C); 34.5; 53.7 (3C); 64.3; 65.2; 69.6;
92.0; 130.3; 130.5; 130.6; 130.7; 172.8. 31P-NMR (162 MHz,
CD3OD/CDCl3 1:1) δ – 5.16 (E1); – 5.68 (E2). IR n 516; 721; 849; 968;
1030; 1082; 1160; 1270; 1465; 1754; 2852; 2921; 3377. HR-MS (ESI+)
m/z [M-Cl]+ calcd for C48H95NO6P+ 812.6892, found 812.6883.
3-((((Acetyloxy)methoxy)(erucyloxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EE2). Compound EE2 (8 mg, 6 %) was
obtained from erufosine (120 mg, 0.24 mmol) and chloromethyl acetate
(215 mg, 1.91 mmol) according to the general procedure. Rf : 0.15
(CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3 1:1) δ
0.85 (t, J = 6.8 Hz, 3H); 1.25 (m, 30H); 1.68 (tt, J1 = J2 = 7.2 Hz, 2H); 2.01
(m, 4H); 2.14 (s, 3H); 2.19 (m, 2H); 3.17 (s, 9H); 3.49 (m, 2H); 4.07 (td,
J1 = J2 = 6.9 Hz, 2H); 4.18 (m, 2H); 5.31 (m, 2H); 5.60 (ABX syst.,
JAB = 5.6 Hz, JAX = 13.4 Hz, JBX = 12.2 Hz, 2H). 13C-NMR δ (100.7 MHz,
CD3OD/CDCl3 1:1) δ 14.5; 20.9; 23.2; 24.8; 26.0; 27.8 (2C); 29.9; 30.0
(2C); 30.2 (3C); 30.3 (4C); 30.4 (2C); 30.9; 32.6; 53.7 (3C); 64.3; 65.5;
69.8; 83.4; 130.5 (2C); 171.8. 31P-NMR (162 MHz, CD3OD/CDCl3 1:1) δ
– 3.43. IR n 671; 975; 1033; 1120; 1463; 1767; 2852; 2922; 3366. HR-MS
(ESI+) m/z [M-Cl]+ calcd for C31H63NO6P+ 576.4388, found 576.4390.
3-(((Erucyloxy)((hexanoyloxy)methoxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EE6). Compound EE6 (47 mg, 30 %)
was obtained from erufosine (120 mg, 0.24 mmol) and chloromethyl
hexanoate (315 mg, 1.89 mmol) according to the general procedure. Rf :
0.3 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3 1:1)
δ 0.85 (t, J = 6.8 Hz, 3H); 0.88 (t, J = 6.6 Hz, 3H); 1.25 (m, 34H); 1.63 (tt,
J1 = J2 = 6.8 Hz, 2H); 1.68 (tt, J1 = J2 = 7.2 Hz, 2H); 2.01 (m, 4H); 2.19 (m,
2H); 2.41 (t, J = 7.5 Hz, 2H); 3.17 (s, 9H); 3.49 (m, 2H); 4.09 (td, J1 = J2 =
6.9 Hz, 2H); 4.20 (td, J1 = J2 = 6.9 Hz, 2H); 5.30 (m, 2H); 5.63 (ABX syst.,
JAB = 5.2 Hz, JAX = 13.5 Hz, JBX = 11.7 Hz, 2H). 13C-NMR δ (100.7 MHz,
CD3OD/CDCl3 1:1) δ 14.2; 14.4; 22.8; 23.2; 24.6; 24.7; 25.9; 27.7 (2C);
29.7; 29.8 (3C); 30.0 (3C); 30.1; 30.2 (2C); 30.3 (2C); 30.8; 31.7; 32.4;
34.3; 53.7 (3C); 64.1; 65.2; 69.7; 83.2; 130.4 (2C); 173.0. 31P-NMR (162
MHz, CD3OD/CDCl3 1:1) δ – 3.41. IR n 493; 755; 857; 970; 1028; 1157;
1269; 1465; 1760; 2852; 2922; 3372. HR-MS (ESI+) m/z [M-Cl]+ calcd for
C35H71NO6P+ 632.5014, found 632.5009.
3-(((Erucyloxy)((((oleyloxy)carbonyl)oxy)methoxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EC18:1). Compound EC18:1 (200 mg,
58 %) was obtained from erufosine (200 mg, 0.40 mmol) and chloromethyl
oleyl carbonate (1.16 g, 3.18 mmol) according to the general procedure.
Rf : 0.65 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CDCl3) δ 0.88
(t, J = 6.6 Hz, 6H); 1.26 (m, 52H); 1.68 (tt, J1 = J2 = 6.8 Hz, 4H); 2.00 (m,
8H); 2.29 (m, 2H); 3.49 (s, 9H); 3.78 (m, 2H); 4.07 (td, J1 = J2 = 6.8 Hz,
2H); 4.18 (t, J = 6.8 Hz, 2H); 4.21 (m, 2H); 5.34 (m, 4H); 5.63 (m, 2H). 13C-
NMR δ (100.7 MHz, CDCl3) δ 14.3 (2C); 22.9 (2C); 24.6; 25.4; 25.6; 27.4
(4C); 28.7; 29.4 (3C); 29.5 (4C); 29.6; 29.8 (4C); 29.9 (4C); 30.0 (4C); 30.3;
32.1 (2C); 53.7 (3C); 63.7; 64.7; 69.2; 69.5; 85.8; 129.9; 130.0; 130.1;
130.2; 153.1. 31P-NMR (162 MHz, CDCl3) δ – 3.03. IR n 502; 721; 772;
858; 947; 1027; 1155; 1256; 1464; 1761; 2852; 2921; 3390. HR-MS (ESI+)
m/z [M-Cl]+ calcd for C48H95NO7P+ 828.6841, found 828.6837.
3-(((Erucyloxy)((octanoyloxy)methoxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EE8). Compound EE8 (192 mg, 69 %)
was obtained from erufosine (202 mg, 0.40 mmol) and chloromethyl
octanoate (620 mg, 3.22 mmol) according to the general procedure. Rf :
0.35 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3 1:1)
δ 0.86 (t, J = 6.8 Hz, 6H); 1.26 (m, 38H); 1.63 (tt, J1 = J2 = 6.8 Hz, 2H);
1.68 (tt, J1 = J2 = 7.4 Hz, 2H); 2.01 (m, 4H); 2.19 (m, 2H); 2.41 (t,
J = 7.5 Hz, 2H); 3.17 (s, 9H); 3.49 (m, 2H); 4.09 (td, J1 = J2 = 6.9 Hz, 2H);
4.20 (td, J1 = J2 = 6.9 Hz, 2H); 5.32 (m, 2H); 5.63 (ABX syst.; JAB = 5.2 Hz,
JAX = 13.8 Hz, JBX = 11.1 Hz, 2H). 13C-NMR δ (100.7 MHz, CD3OD/CDCl3
1:1) δ 14.4; 14.5; 23.2; 23.3; 24.7; 25.2; 26.1; 27.8 (2C); 29.7 (2C); 29.9;
30.0 (3C); 30.2 (3C); 30.3 (3C); 30.4 (2C); 30.9; 32.4; 32.6; 34.5; 53.7
(3C); 64.3; 65.5; 69.9; 83.4; 130.5; 130.6; 173.2. 31P-NMR (162 MHz,
CD3OD/CDCl3 1:1) δ – 3.50. IR n 506; 756; 858; 970; 1028; 1152; 1269;
1466; 1760; 2852; 2922; 3384. HR-MS (ESI+) m/z [M-Cl]+ calcd for
C37H75NO6P+ 660.5327, found 660.5317.
3-(((Erucyloxy)(1-((((oleyloxy)carbonyl)oxy)ethoxy))phosphoryl)oxy)-
N,N,N-trimethylpropan-1-aminium chloride (EC'18:1). Compound EC'18:1
(169 mg, 49 %) was obtained as two separated couples of enantiomers
from erufosine (199 mg, 0.40 mmol) and 1-chloroethyl oleyl carbonate
(1.18 mg, 3.14 mmol) according to the general procedure. Rf : 0.7 (E1) and
0.6 (E2) (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CDCl3) E1 δ
0.88 (t, J = 6.6 Hz, 6H); 1.26 (m, 52H); 1.58 (d, J = 5.2 Hz, 3H); 1.68 (tt,
J1 = J2 = 6.8 Hz, 4H); 2.00 (m, 8H); 2.28 (m, 2H); 3.49 (s, 9H); 3.81 (m,
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
3-((((Decanoyloxy)methoxy)(erucyloxy)phosphoryl)oxy)-N,N,N-
CD3OD/CDCl3 1:1) δ – 3.29. IR n 508; 721; 771; 859; 967; 1032; 1161;
1267; 1467; 1761; 2849; 2917; 3377. HR-MS (ESI+) m/z [M-Cl]+ calcd for
C47H95NO6P+ 800.6892, found 800.6891.
trimethylpropan-1-aminium chloride (EE10). Compound EE10 (220 mg,
74 %) was obtained from erufosine (206 mg, 0.41 mmol) and chloromethyl
decanoate (716 mg, 3.27 mmol) according to the general procedure. Rf :
0.45 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3 1:1)
δ 0.86 (t, J = 6.8 Hz, 6H); 1.25 (m, 42H); 1.63 (tt, J1 = J2 = 6.8 Hz, 2H);
1.68 (tt, J1 = J2 = 7.4 Hz, 2H); 2.01 (m, 4H); 2.19 (m, 2H); 2.41 (t,
J = 7.5 Hz, 2H); 3.17 (s, 9H); 3.49 (m, 2H); 4.09 (td, J1 = J2 = 6.9 Hz, 2H);
4.20 (td, J1 = J2 = 6.9 Hz, 2H); 5.32 (m, 2H); 5.63 (ABX syst., JAB = 5.2 Hz,
JAX = 13.8 Hz, JBX = 11.1 Hz, 2H). 13C-NMR δ (100.7 MHz, CD3OD/CDCl3
1:1) δ 14.6 (2C); 23.6 (2C); 25.1; 25.5; 26.4; 28.1 (2C); 30.0; 30.2 (3C);
30.3 (3C); 30.4 (2C); 30.5 (2C); 30.6 (3C); 30.7 (2C); 31.2; 32.9 (2C); 34.8;
53.9 (3C); 64.5; 66.0; 70.1; 83.8; 130.8 (2C); 173.6. 31P-NMR (162 MHz,
CD3OD/CDCl3 1:1) δ – 3.17. IR n 512; 752; 787; 858; 970; 1027; 1149;
1273; 1466; 1761; 2852; 2922; 3373. HR-MS (ESI+) m/z [M-Cl]+ calcd for
C39H79NO6P+ 688.5640, found 688.5632.
Methods
Nucleic acid retardation assay. Freshly prepared lipoplexes (vide infra) at
the desired N/P ratio (where N is the concentration of the lipid ammonium
group and
P that of nucleic acid phosphate) were analyzed by
electrophoresis through a 1 % agarose gel. The gel was run in a 40 mM
Tris-acetate-EDTA buffer, pH 8.0 and nucleic acid was further stained
using an ethidium bromide solution (0.5 µg/mL).
Dynamic light scattering (DLS) measurements. The average particle size
and zeta potential of lipoplexes were measured using a Zetasizer nanoZS
apparatus (Malvern Instruments, Paris, France). All measurements were
performed on freshly prepared lipoplexes (vide infra) at 25 °C and in
triplicate. Data were analyzed using the multimodal number distribution
software supplied with the instrument and expressed as mean (± SD).
3-(((Erucyloxy)((tetradecanoyloxy)methoxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EE14). Compound EE14 (100 mg,
32 %) was obtained from erufosine (200 mg, 0.40 mmol) and chloromethyl
tetradecanoate (890 mg, 3.21 mmol) according to the general procedure.
Rf : 0.5 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3
1:1) δ 0.86 (t, J = 6.8 Hz, 6H); 1.25 (m, 50H); 1.63 (tt, J1 = J2 = 6.8 Hz, 2H);
1.68 (tt, J1 = J2 = 7.4 Hz, 2H); 2.01 (m, 4H); 2.19 (m, 2H); 2.41 (t,
J = 7.5 Hz, 2H); 3.17 (s, 9H); 3.49 (m, 2H); 4.09 (td, J1 = J2 = 6.9 Hz, 2H);
4.20 (td, J1 = J2 = 6.9 Hz, 2H); 5.32 (m, 2H); 5.63 (ABX syst., JAB = 5.2 Hz,
JAX = 13.8 Hz, JBX = 11.1 Hz, 2H). 13C-NMR δ (100.7 MHz, CD3OD/CDCl3
1:1) δ 14.5 (2C); 23.4 (2C); 24.9; 25.3; 26.2; 27.9 (2C); 29.8; 29.9; 30.0
(2C); 30.1 (2C); 30.3 (4C); 30.6 (8C); 30.5 (2C); 31.0; 32.7 (2C); 34.6; 53.7
(3C); 64.3; 65.6; 70.0; 83.5; 130.6 (2C); 173.3. 31P-NMR (162 MHz,
CD3OD/CDCl3 1:1) δ – 3.29. IR n 515; 721; 772; 859; 966; 1030; 1160;
1269; 1467; 1760; 2849; 2918; 2956; 3395. HR-MS (ESI+) m/z [M-Cl]+
calcd for C43H87NO6P+ 744.6266, found 744.6259.
Erufosine prodrug hydrolytic stability determination. Hydrolysis rate of pro-
APLs was measured by 31P-NMR spectroscopy. Compounds were
formulated as liposomes using a solvent injection technique.[23] Briefly, the
pro-APLs (10 µmol) were dissolved in i-PrOH (200 µL) and then injected
with a syringe at a flow rate of approximately 600 µL min-1 and a stirring
speed of 400 rpm into the appropriate aqueous buffer medium (300 µL,
either HEPES 10 mM pH 7.4, or AcOK/AcOH 10 mM pH 4.5) containing
10 mM Triton X-100. The resulting preparations were complemented with
buffer (400 µL) and D2O (100 µL) and introduced into 5-mm NMR tubes,
and 31P-NMR spectra were recorded periodically at 20 °C with a 4-s pulse
cycle for quantitative measurements.
Pro-APLs self-assembly studies. Self-assembly properties of erufosine
and pro-APLs were determined using a fluorescent probe technique.[[51]]
A
3-(((Erucyloxy)((hexadecanoyloxy)methoxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EE16). Compound EE16 (237 mg,
74 %) was obtained from erufosine (200 mg, 0.40 mmol) and chloromethyl
hexadecanoate (980 mg, 3.21 mmol) according to the general procedure.
Rf : 0.45 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3
1:1) δ 0.86 (t, J = 6.8 Hz, 6H); 1.24 (m, 54H); 1.63 (tt, J1 = J2 = 6.8 Hz, 2H);
1.68 (tt, J1 = J2 = 7.4 Hz, 2H); 2.01 (m, 4H); 2.19 (m, 2H); 2.41 (t,
J = 7.5 Hz, 2H); 3.17 (s, 9H); 3.49 (m, 2H); 4.09 (td, J1 = J2 = 6.9 Hz, 2H);
4.20 (td, J1 = J2 = 6.9 Hz, 2H); 5.32 (m, 2H); 5.63 (ABX syst., JAB = 5.2 Hz,
JAX = 13.8 Hz, JBX = 11.1 Hz, 2H). 13C-NMR δ (100.7 MHz, CD3OD/CDCl3
1:1) δ 14.5 (2C); 23.3 (2C); 24.8; 25.2; 26.1; 27.8 (2C); 29.7; 29.8; 30.0
(6C); 30.2 (4C); 30.3 (8C); 30.4 (2C); 30.9; 32.6 (2C); 34.5; 53.7 (3C);
64.3; 65.4; 69.9; 83.4; 130.5 (2C); 173.2. 31P-NMR (162 MHz,
CD3OD/CDCl3 1:1) δ – 3.34. IR n 506; 720; 758; 859; 969; 1021; 1159;
1272; 1467; 1760; 2850; 2918; 3379. HR-MS (ESI+) m/z [M-Cl]+ calcd for
C45H91NO6P+ 772.6579, found 772.6569.
lipid film obtained by evaporation in a SpeedVac apparatus for 1 h of the
adequate volume of an ethanolic solution (2 mM) of the selected
compound was dispersed in H2O (2 mM) and diluted in a concentration
range from 2 mM to 2 nM. Pyrene (0.1 mM in DMSO, 10 µL) was added to
the samples under vigorous stirring, and the mixtures were incubated in
the dark for 30 min at 25 °C. Fluorescence spectra were then recorded at
350 to 450 nm with excitation at 330 nm, with
a Fluoromax-4
spectrophotometer (Horiba Jobin Yvon). Emission spectral ratio at 375 and
387 nm (I375/I387) was expressed as the logarithm of concentration, and the
concentration above which the molecules spontaneously associate to form
micelles or aggregates was graphically determined as the first break point
in the fluorescence behavior of pyrene.
Hemolytic activity determination. Red blood cell (RBC) leakage assay was
carried out according to a procedure described elsewhere.[[52]] RBC were
prepared from defibrinated sheep blood. Briefly, blood sample (5 mL) was
centrifuged at 500 x g for 5 min. After removal of the yellow upper layer,
an equal volume of PBS pH 7.4 was added to the sample, and the mixture
was homogenized. This treatment (centrifugation, removal of upper layer,
addition of PBS and homogenization) was repeated three times. The last
homogenate was diluted to 1/25 in PBS and the resulting suspension of
RBCs (140 µL) was added to lipid samples (vide supra, 10 µL) prepared
at increasing concentration (from 45 µM to 3 mM) in 96-well plates. After
incubation for 1 or 24 h at 37 °C and 5 % CO2, plates were centrifuged at
250 x g for 10 min to pellet intact RBC and absorbance of the supernatants
(100 µL) was measured at 450 nm. Hemolysis induced by the compounds
was expressed as percent, calculated from negative and 100% controls,
and plotted as a function of lipid concentration. Negative control was
obtained on suspensions of RBC in PBS alone. Complete hemolysis
3-(((Erucyloxy)((octadecanoyloxy)methoxy)phosphoryl)oxy)-N,N,N-
trimethylpropan-1-aminium chloride (EE18). Compound EE18 (91 mg, 55 %)
was obtained from erufosine (100 mg, 0.20 mmol) and chloromethyl
octadecanoate (530 mg, 1.59 mmol) according to the general procedure.
Rf : 0.5 (CH2Cl2/MeOH/H2O 75:22:3). 1H-NMR (400 MHz, CD3OD/CDCl3
1:1) δ 0.88 (t, J = 6.8 Hz, 6H); 1.24 (m, 58H); 1.62 (tt, J1 = J2 = 6.8 Hz, 2H);
1.68 (tt, J1 = J2 = 7.4 Hz, 2H); 2.00 (m, 4H); 2.28 (m, 2H); 2.38 (t, J = 7.5
Hz, 2H); 3.47 (s, 9H); 3.80 (m, 2H); 4.05 (td, J1 = J2 = 6.9 Hz, 2H); 4.23 (m,
2H); 5.34 (m, 2H); 5.63 (ABX syst., JAB = 4.8 Hz, JAX = 12.9 Hz,
JBX = 11.1 Hz, 2H). 13C-NMR δ (100.7 MHz, CD3OD/CDCl3 1:1) δ 14.3
(2C); 22.8 (2C); 24.5; 24.7; 25.5; 27.4 (2C); 29.2; 29.3; 29.4; 29.5 (4C);
29.7 (3C); 29.8 (6C); 29.9 (6C); 30.0 (2C); 30.4; 32.0 (2C); 34.1; 53.8 (3C);
63.7; 64.6; 69.1; 82.7; 130.0 (2C); 172.4. 31P-NMR (162 MHz,
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
(100 % control) was obtained by adding Triton X-100® (10 µM, 10 µL) to
suspensions of RBC in PBS alone. Lipid concentrations inducing 50 %
hemolysis (HC50) and the percentage of hemolysis provoked by 200 µM of
the lipids (HA200) were graphically determined from the curve fitted to the
data.
Analysis of cell TRAIL receptor expression and sensitivity. To assess
TRAIL expression, A549, H292, and 16HBE cells (106 cells) were washed
with PBS containing 2 % FCS and incubated for 4 h at 4 °C with a PE-
labeled monoclonal antibody targeting TRAIL-R1, R2, R3, or R4 according
to concentration recommended by the manufacturer. After two washes
with PBS-2 % FCS, TRAIL receptor expression was monitored by flow
cytometry (Guava EasyCyteTM, Merck Milipore, Darmstadt, Germany)
and data were analyzed with InCyte Software (Merck Milipore).
Cell culture. All cell lines were grown in culture flasks (Becton-Dickinson)
at 37 °C in a 5 % CO2 humidified chamber. Human lung carcinoma (A549)
and bronchial (16HBE) epithelial cells were grown in DMEM/F12 medium
containing FBS (10 %), penicillin (100 units/mL), streptomycin (100 µg/mL),
and Hepes (5 mM). Lung mucoepidermoid carcinoma cells (H292) were
grown in RPMI 1640 supplemented with FBS (10 %), sodium pyruvate
(1 mM), L-glutamine (2 mM), penicillin (100 units/mL), streptomycin (100
µg/mL), and Hepes (10 mM). At confluence, cells were released from
flasks with trypsin (0.5 % in PBS), centrifuged (4 °C, 5 min, 120 g) counted
and transferred into 96-well plates (Becton-Dickinson) in 100 µL culture
For determination of cell sensitivity to TRAIL, A549, H292, and 16HBE
cells were washed with PBS containing 2 % FCS and incubated for 4 h in
96-well plates (30,000 cells, 100 µL) before addition of increasing amounts
(in 10 µL) of SPK. After 24 h incubation at 37 °C, cell viability was
determined by MTT assay as described above.
Statistical analysis. All data are presented as the mean ± standard
deviation (SD). Statistical significance between treatments was assessed
by one-way analysis of variance (ANOVA) followed by the Dunnett multiple
comparison test, using the Kaleidagraph 4.5 software (Synergy Software,
Reading, PA, USA). Data were considered as statistically significant for p
value less than 0.05 (*** p < 0.001, ** p < 0.01, * p < 0.05).
medium (A549:
6,000 cells/well; H292: 9,000 cells/well; 16HBE:
12,000 cells/well) for transfection experiments, cytotoxicity assays, or IC50
determinations. Plates were maintained at 37 °C in a 5 % CO2 humidified
chamber for 24 h before experiment.
Preparation of lipoplexes. Typically, the appropriate volume of a freshly
prepared solution of pro-erufosine compound (2 mM in EtOH) was
deposited at the bottom of a 500 µL polyethylene tube and dried under
vacuum for 1 h. Then, pCMV-Gluc DNA (40 µL at the required
concentration in 5 % glucose, DNA concentration refers to phosphate
content) was added to the resulting lipid film. After stirring by vortex for 20
s, the preparation was allowed to stand at rt for 30 min before use.
Acknowledgements
The authors are grateful for financial support to BG from Labex
Medalis, Région Alsace, and Alsace contre le Cancer.
Transfection experiments. Cells were seeded into 96-well plates (Becton-
Dickinson) at the required density (vide supra) in 100 µL of serum
containing culture medium. Twenty-four hours later, freshly prepared DNA
lipoplexes (10 µL, i.e. 0.4 µg DNA) were added to the wells and cells were
then let to grow in the incubator without further handling for 24 h. Negative
control was obtained by adding glucose 5 % (10 µL) to the wells instead of
lipoplexes. Then, Gaussia luciferase production was measured by
monitoring light production on an aliquot of culture supernatant (20 µL of a
1/100th dilution of supernatant prepared in non-supplemented culture
medium) for 1 sec upon addition of the coelenterazine substrate (50 µL,
1.5 µM) using a luminometer (Berthold Centro LB960 XS, Thoiry, France).
Protein content of the culture supernatant was determined using a
bicinchoninic acid (BCA) assay and luciferase activity was normalized to
protein content (RLU/mg protein). Value for each sample is the mean of a
triplicate determination (± SD).
Keywords: erufosine • prodrug • gene therapy • chemotherapy •
hemolytic toxicity
[1] a) J. de Almeida Pachioni, J. G. Magalhães, E. J. C. Lima, L. de Moura Bueno,
J. F. Barbosa, M. M. de Sá and C. O. Rangel-Yagui, J. Pharm. Pharm. Sci. 2013,
16, 742-759; b) P.-A. Jaffrès, C. Gajate, A. M. Bouchet, H. Couthon-Gourvès,
A. Chantôme, M. Potier-Cartereau, P. Besson, P. Bougnoux, F. Mollinedo and
C. Vandier, Pharmacol. Ther. 2016, 165, 114-131; c) P. Ríos-Marco, C. Marco,
X. Gálvez, J. M. Jiménez-López and M. P. Carrasco, Biochim. Biophys. Acta-
Biomembr. 2017, 1859, 1657-1667.
[2] A. Kostadinova, T. Topouzova-Hristova, A. Momchilova, R. Tzoneva and M.
R. Berger in Advances in Protein Chemistry and Structural Biology, Vol. 101 (Ed.
R. Donev), 2015, pp. 27-66.
[3] H. Eibl, D. Arnold, H. U. Weltzien and O. Westphal, Liebigs Ann. Chem. 1967,
709, 226-230.
[4] C. Gajate and F. Mollinedo, Curr. Drug Metab. 2002, 3, 491-525.
[5] a) R. Andreesen, M. Modolell, H. U. Weltzien, H. Eibl, H. H. Common, G. W.
Löhr and P. G. Munder, Cancer Res. 1978, 38, 3894-3899; b) P. G. Munder, M.
Modolell, R. Andreesen, H. U. Weltzien and O. Westphal, Springer Semin.
Immunopathol. 1979, 2, 187-203.
[6] D. Berkovic, Gen. Pharmacol. 1998, 31, 511-517.
[7] P. Hilgard, T. Klenner, J. Stekar, G. Nössner, B. Kutscher and J. Engel, Eur.
J. Cancer 1997, 33, 442-446.
[8] S. R. Vink, J. H. M. Schellens, W. J. van Blitterswijk and M. Verheij, Invest.
New Drugs 2005, 23, 279-286.
[9] P. G. Richardson, C. Eng, J. Kolesar, T. Hideshima and K. C. Anderson,
Expert Opin. Drug Metab. Toxicol. 2012, 8, 623-633.
[10] J. Kötting, M. R. Berger, C. Unger and H. Eibl, Cancer Chemother.
Cytotoxicity assay. Mitochondrial activity measurements (MTT assay) was
used to assess cytotoxicity of the compounds. Lipid samples were
prepared from a lipid film (3.6 µmol of compound evaporated from an
ethanolic solution) hydrated with aqueous glucose 5 % and vigorously
vortexed for 20 s before serial dilution in aqueous glucose 5 %. Part of the
culture medium in the plates prepared the day before (25 µL) was removed
and replaced by lipid samples (25 µL). Negative control was obtained by
adding glucose 5 % to the cells instead of lipoplexes. After a 24-h
incubation period at 37 °C, culture supernatant was removed, cells were
carefully washed with PBS and 0.5 mg/mL MTT (100 µL) in complete
culture medium was added. After a 1-h incubation period at 37 °C, MTT
solution was removed and DMSO (100 µL) was added to lyse cells and
dissolve reduced MTT. Intensity of MTT reduction was then evaluated by
measuring absorbance at 492 nm. Viability of cells treated with lipids or
lipoplexes was expressed as the percentage of the absorbance measured
in untreated cells. Value for each sample is the mean of triplicate
determinations (±SD). IC50 values were determined within the 95 %
confidence interval using nonlinear regression analysis with the GraphPad
Prism software (GraphPad Software, San Diego, CA, USA).
Pharmacol. 1992, 30, 105-112.
[11] R. G. Bagley, L. Kurtzberg, C. Rouleau, M. Yao and B. A. Teicher, Cancer
Chemother. Pharmacol. 2011, 68, 1537-1546.
[12] a) J. C. Bendell, J. Nemunaitis, S. J. Vukelja, C. Hagenstad, L. T. Campos,
R. C. Hermann, P. Sportelli, L. Gardner and D. A. Richards, J. Clin. Oncol. 2011,
29, 4394-4400; b) S. Fu, B. T. Hennessy, C. S. Ng, Z. Ju, K. R. Coombes, J. K.
Wolf, A. K. Sood, C. F. Levenback, R. L. Coleman, J. J. Kavanagh, D. M.
Gershenson, M. Markman, K. Dice, A. Howard, J. Li, Y. Li, K. Stemke-Hale, M.
Dyer, E. Atkinson, E. Jackson, V. Kundra, R. Kurzrock, R. C. Bast and G. B.
Mills, Gynecol. Oncol. 2012, 126, 47-53; c) I. Gojo, A. Perl, S. Luger, M. R. Baer,
K. J. Norsworthy, K. S. Bauer, M. Tidwell, S. Fleckinger, M. Carroll and E. A.
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
Sausville, Invest. New Drugs 2013, 31, 1217-1227; d) A. Guidetti, C. Carlo-
Stella, S. L. Locatelli, W. Malorni, R. Mortarini, S. Viviani, D. Russo, A.
Marchianò, R. Sorasio, A. Dodero, L. Farina, L. Giordano, M. Di Nicola, A.
Anichini, P. Corradini and A. M. Gianni, Clin. Cancer Res. 2014, 20, 5641-5651;
e) P. G. Richardson, J. Wolf, A. Jakubowiak, J. Zonder, S. Lonial, D. Irwin, J.
Densmore, A. Krishnan, N. Raje, M. Bar, T. Martin, R. Schlossman, I. M.
Ghobrial, N. Munshi, J. Laubach, J. Allerton, T. Hideshima, K. Colson, E.
Poradosu, L. Gardner, P. Sportelli and K. C. Anderson, J. Clin. Oncol. 2011, 29,
4243-4249.
[13] a) C. Belka, V. Jendrossek, M. Pruschy, S. Vink, M. Verheij and W. Budach,
Int. J. Radiat. Oncol. Biol. Phys. 2004, 58, 542-554; b) S. R. Vink, W. J. van
Blitterswijk, J. H. M. Schellens and M. Verheij, Cancer Treat. Rev. 2007, 33,
191-202.
[31] M. C. Parlato, J. P. Jee, M. Teshite and S. Mecozzi, J. Org. Chem. 2011,
76, 6584-6591.
[32] a) H. Eibl and P. Kaufmann-Kolle, J. Liposome Res. 1995, 5, 131-148; b)
F. Schuettauf, K. H. Eibl, S. Thaler, K. Shinoda, R. Rejdak, C. A. May, G.
Blatsios and U. Welge-Lussen, Curr. Eye Res. 2005, 30, 813-820.
[33] M. Lukac, M. Mrva, M. Garajova, G. Mojzisova, L. Varinska, J. Mojzis, M.
Sabol, J. Kubincova, H. Haragova, F. Ondriska and F. Devinsky, Eur. J. Med.
Chem. 2013, 66, 46-55.
[34] a) I. Fichtner, R. Zeisig, H. Naundorf, S. Jungmann, D. Arndt, G. Asongwe,
J. A. Double and M. C. Bibby, Breast Cancer Res. Treat. 1994, 32, 269-279; b)
P. Kaufmann-Kolle, J. Drevs, M. R. Berger, J. Kotting, N. Marschner, C. Unger
and H. Eibl, Cancer Chemother. Pharmacol. 1994, 34, 393-398.
[35] R. Zeisig, S. Jungmann, D. Arndt, A. Schutt and E. Nissen, Anti-Cancer
Drugs 1993, 4, 57-64.
[14] R. Zeisig, A. Ress, I. Fichtner and W. Walther, Cancer Gene Ther. 2003,
10, 302-311.
[36] T. Mosmann, J. Immunol. Methods 1983, 65, 55-63.
[37] J. M. W. Gee, A. Howell, W. J. Gullick, C. C. Benz, R. L. Sutherland, R. J.
Santen, L. A. Martin, F. Ciardiello, W. R. Miller, M. Dowsett, P. Barrett-Lee, J. F.
R. Robertson, S. R. Johnston, H. E. Jones, A. E. Wakeling, R. Duncan and R.
I. Nicholson, Endocr.-Relat. Cancer 2005, 12, S1-S7.
[15] N. Settelen, O. Roch, D. Bock, R. Rooke, S. Braun and O. Meyer, J. Control.
Release 2004, 94, 237-244.
[16] H. A. Elrod, Y.-D. Lin, P. Yue, X. Wang, S. Lonial, F. R. Khuri and S.-Y. Sun,
Mol. Cancer Ther. 2007, 6, 2029-2038.
[17] Z. U. Rehman, I. S. Zuhorn and D. Hoekstra, J. Control. Release 2013, 166,
46-56.
[38] H. J. Lee, G. L. Zhuang, Y. Cao, P. Du, H. J. Kim and J. Settleman, Cancer
Cell 2014, 26, 207-221.
[18] S. Mukherjee, T. T. Soe and F. R. Maxfield, J. Cell Biol. 1999, 144, 1271-1284.
[19] R. L. Juliano, Nucl. Acid Ther. 2018, 28, 3.
[39] W. Huang, L. Q. Chen, L. Kang, M. J. Jin, P. Sun, X. Xin, Z. G. Gao and Y.
H. Bae, Adv. Drug Deliv. Rev. 2017, 115, 82-97.
[20] B. Gaillard, J.-S. Remy, F. Pons and L. Lebeau, submitted.
[21] M. C. Georgieva, S. M. Konstantinov, M. Topashka-Ancheva and M. R.
Berger, Cancer Lett. 2002, 182, 163-174.
[40] V. Tsouris, M. K. Joo, S. H. Kim, I. C. Kwon and Y. Y. Won, Biotechnol. Adv.
2014, 32, 1037-1050.
[41] Y. S. Shi, X. Pang, J. Q. Wang and G. Liu, Adv. Healthc. Mater. 2018, 7,
15.
[42] D. Merino, N. Lalaoui, A. Morizot, P. Schneider, E. Solary and O. Micheau,
Mol. Cell. Biol. 2006, 26, 7046-7055.
[43] S. Shirley, A. Morizot and O. Micheau, Recent Patents Anti-Canc. Drug
Discov. 2011, 6, 311-323.
[44] N. Chekkat, C. M. Lombardo, C. Seguin, M.-C. Lechner, F. Dufour, Y.
Nominé, M. De Giorgi, B. Frisch, O. Micheau, G. Guichard, D. Altschuh and S.
Fournel, Oncotarget 2018, 9, 15566.
[45] K. Azijli, I. A. M. van Roosmalen, J. Smit, S. Pillai, M. Fukushima, S. de
Jong, G. J. Peters, I. V. Bijnsdorp and F. A. E. Kruyt, Cancer Chemother.
Pharmacol. 2014, 73, 1273-1283.
[22] a) P. Pierrat, A. Casset, P. Didier, D. Kereselidze, M. Lux, F. Pons and L.
Lebeau, ChemBioChem 2016, 17, 1771-1783; b) P. Pierrat, A. Casset, D.
Kereselidze, M. Lux, F. Pons and L. Lebeau, Macromol. Biosci. 2016, 16, 984-
989; c) P. Pierrat, G. Creusat, G. Laverny, F. Pons, G. Zuber and L. Lebeau,
Chem. Eur. J. 2012, 18, 3835-3839; d) P. Pierrat, D. Kereselidze, M. Lux, L.
Lebeau and F. Pons, Int. J. Pharm. 2016, 511, 205-218; e) P. Pierrat, D.
Kereselidze, P. Wehrung, G. Zuber, F. Pons and L. Lebeau, Pharm. Res. 2013,
30, 1362-1379; f) P. Pierrat, G. Laverny, G. Creusat, P. Wehrung, J.-M. Strub,
A. VanDorsselaer, F. Pons, G. Zuber and L. Lebeau, Chem. Eur. J. 2013, 19,
2344-2355.
[23] P. Gentine, A. Bubel, C. Crucifix, L. Bourel-Bonnet and B. Frisch, J.
Liposome Res. 2012, 22, 18-30.
[46] P. M. Nair, H. Flores, A. Gogineni, S. Marsters, D. A. Lawrence, R. F. Kelley,
H. Ngu, M. Sagolla, L. Komuves, R. Bourgon, J. Settleman and A. Ashkenazi,
Proc. Natl. Acad. Sci. U. S. A.a 2015, 112, 5679-5684.
[47] L. D. Mayer and A. S. Janoff, Mol. Interv. 2007, 7, 216-223.
[48] L. Lebeau, S. Olland, P. Oudet and C. Mioskowski, Chem. Phys. Lipids
1992, 62, 93-103.
[24] N. Dan and D. Danino, Adv. Colloid Interface Sci. 2014, 205, 230-239.
[25] I. S. Zuhorn, U. Bakowsky, E. Polushkin, W. H. Visser, M. C. A. Stuart, J.
Engberts and D. Hoekstra, Mol. Ther. 2005, 11, 801-810.
[26] J. Rejman, V. Oberle, I. S. Zuhorn and D. Hoekstra, Biochem. J.l 2004, 377,
159-169.
[27] S. Audouy, G. Molema, L. de Leij and D. Hoekstra, J. Gene Med. 2000, 2,
465-476.
[49] M. Kim, A. Pitchaiah, N.-J. Park, I. T. Hwang and K.-I. Lee, Lett. Org. Chem.
2012, 9, 628-631.
[28] S. E. Han, H. Kang, G. Y. Shim, M. S. Suh, S. J. Kim, J. S. Kim and Y. K.
Oh, Int. J. Pharm. 2008, 353, 260-269.
[50] J. A. Heyes, D. Niculescu-Duvaz, R. G. Cooper and C. J. Springer, J. Med.
Chem. 2002, 45, 99-114.
[29] a) C. L. Chan, K. K. Ewert, R. N. Majzoub, Y. K. Hwu, K. S. Liang, C. Leal
and C. R. Safinya, J. Gene Med. 2014, 16, 84-96; b) O. Zelphati, L. S. Uyechi,
L. G. Barron and F. C. Szoka, Biochim. Biophys. Acta 1998, 1390, 119-133; c)
L. Li, H. M. Song, K. Luo, B. He, Y. Nie, Y. Yang, Y. Wu and Z. W. Gu, Int. J.
Pharm. 2011, 408, 183-190.
[51] a) E. D. Goddard, N. J. Turro, P. L. Kuo and K. P. Ananthapadmanabhan,
Langmuir 1985, 1, 352-355; b) L. Piñeiro, M. Novo and W. Al-Soufi, Adv. Colloid
Interface Sci. 2015, 215, 1-12.
[52] B. C. Evans, C. E. Nelson, S. S. Yu, K. R. Beavers, A. J. Kim, H. Li, H. M.
Nelson, T. D. Giorgio and C. L. Duvall, J. Vis. Exp. 2013, 73, e50166.
[30] a) J. Kötting, N. W. Marschner, W. Neumuller, C. Unger and H. Eibl, Prog.
Exp. Tumor Res. 1992, 34, 131-142; b) E. A. M. Fleer, D. Berkovic, C. Unger
and H. Eibl, Prog. Exp. Tumor Res. 1992, 34, 33-46.
This article is protected by copyright. All rights reserved.

10.1002/chem.201903976
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
Text for Table of Contents
Author(s), Corresponding Author(s)*
Page No. – Page No.
Title
((Insert TOC Graphic here: max.
width: 5.5 cm; max. height: 5.0 cm))
Layout 2:
FULL PAPER
O
P
Boris Gaillard, Cendrine Seguin, Jean-
Serge Remy, Françoise Pons, Luc
Lebeau*
O
O
N
X
- Gene delivery
O
R2
O
- in situ erufosine release
- Hemocompatibility
Erufosine
R1
pUNO1-hTRAIL
O
Page No. – Page No.
"Pro-erufosine"
Selective antitumor effect
Erufosine (ErPC3) Cationic Prodrugs
as Dual Gene Delivery Reagents for
Combined Antitumor Therapy
Dual erufosine prodrugs: Biolabile cationic derivatives of the antiproliferative
alkylphospholipid erufosine have been synthesized and show combined antitumor
activity with a plasmid encoding the tumor necrosis factor-related apoptosis-
inducing ligand (TRAIL), with selectivity for tumor vs. non-tumor cells.
This article is protected by copyright. All rights reserved.