Synthesis and Evaluation of Antitumor Alkylphospholipid Prodrugs

Article abstract of DOI:10.1007/s11095-020-02830-y

Purpose: Hemolysis is a serious side effect of antitumor alkylphospholipids (APLs) that limits dose levels and is a constraint in their use in therapeutic regimen. Nine prodrugs of promising APLs (miltefosine, perifosine, and erufosine) were synthesized so as to decrease their membrane activity and improve their toxicity profile while preserving their antineoplastic potency. Methods: The synthesis of the pro-APLs was straightforwardly achieved in one step starting from the parent APLs. The critical aggregation concentration of the prodrugs, their hydrolytic stability under various pH conditions, their blood compatibility and cytotoxicity in three different cell lines were determined and compared to those of the parent antitumor lipids. Results: The APL prodrugs display antitumor activity which is similar to that of the parent alkylphospholipids but without associated hemolytic toxicity. Conclusion: The pro-APL compounds may be considered as intravenously injectable derivatives of APLs. They could thus address one of the major issues met in cancer therapies involving antitumor lipids and restricting their utilization to oral and topical administration because of limited maximum tolerated dose.

Full text of DOI:10.1007/s11095-020-02830-y

Pharm Res
(2020) 37:106
https://doi.org/10.1007/s11095-020-02830-y
RESEARCH PAPER
Synthesis and Evaluation of Antitumor Alkylphospholipid
Prodrugs
1
1
1
1
Boris Gaillard & Jean-Serge Remy & Françoise Pons & Luc Lebeau
Received: 10 January 2020 /Accepted: 21 April 2020
#
Springer Science+Business Media, LLC, part of Springer Nature 2020
ABSTRACT
INTRODUCTION
Purpose Hemolysis is a serious side effect of antitumor alkyl-
phospholipids (APLs) that limits dose levels and is a constraint
in their use in therapeutic regimen. Nine prodrugs of promis-
ing APLs (miltefosine, perifosine, and erufosine) were synthe-
sized so as to decrease their membrane activity and improve
their toxicity profile while preserving their antineoplastic
potency.
Methods The synthesis of the pro-APLs was straightforward-
ly achieved in one step starting from the parent APLs. The
critical aggregation concentration of the prodrugs, their hy-
drolytic stability under various pH conditions, their blood
compatibility and cytotoxicity in three different cell lines were
determined and compared to those of the parent antitumor
lipids.
Lysophosphatidylcholines (lysoPCs) are endogenous cell
membrane components produced from phosphatidylcholines
by phospholipases. They act as modulators of several signaling
pathways and biochemical routes. Since these compounds are
not stable and are substrates for acyltransferases and lysophos-
pholipases, efforts have been made to develop alkylphospho-
lipids (APLs) as metabolically stable synthetic analogs of
lysoPCs for translational health research and clinical trials.
Some of these analogs revealed effective immune modulating
properties but also demonstrated selective antineoplastic ac-
tivities (1). This novel class of antitumor agents thus served as
the basis for the synthesis of a wider diversity of structural
variants that have been reviewed elsewhere (2). Among them,
miltefosine, perifosine, and erufosine proved especially prom-
ising and have been tested in phase I and phase II clinical trials
for their antitumor activity (Fig. 1).
Results The APL prodrugs display antitumor activity which is
similar to that of the parent alkylphospholipids but without
associated hemolytic toxicity.
Conclusion The pro-APL compounds may be considered as
intravenously injectable derivatives of APLs. They could thus
address one of the major issues met in cancer therapies involving
antitumor lipids and restricting their utilization to oral and top-
ical administration because of limited maximum tolerated dose.
Miltefosine has been described for the first time by Eibl and
Unger in 1987 (3). It has been evaluated in vitro (4) and in vivo
(5), and its clinical development for oral treatment of patients
with solid tumors was prevented by dose-limiting gastrointes-
tinal toxicity (6). The intravenous (iv) route being excluded
because of high hemolytic activity (7), miltefosine is currently
used for topical administration. Perifosine results from phar-
macomodulation at the headgroup and was developed in or-
der to get improved biological stability so as to expand the
therapeutic range of the compound (8). Replacement of the
choline substituent by the 4-hydroxy-N,N-dimethylpiperidi-
nium moiety and elongation of the hydrophobic chain to 18
carbon atoms were assumed to lead to a less toxic compound,
while preserving the antineoplastic potency. These expecta-
tions have been fulfilled to a large extent and phase I and
phase II clinical trials were conducted (9). However, whereas
perifosine had extended half-life in plasma (137 h vs. 96 h for
miltefosine) (10), it was associated with high hemolytic activity
(11). Oral administration caused toxicity that was similar to
.
.
KEY WORDS alkylphospholipid erufosine hemolytic
.
.
.
toxicity miltefosine perifosine prodrug
Electronic supplementary material The online version of this article
(https://doi.org/10.1007/s11095-020-02830-y) contains supplementary
material, which is available to authorized users.
*
Luc Lebeau
llebeau@unistra.fr
1
Laboratoire de Conception et Application de Molécules Bioactives, UMR
7
199 CNRS - Université de Strasbourg, Faculté de Pharmacie, 74 route
du Rhin – BP 60024, 67401 Illkirch, France

1
06 Page 2 of 12
Pharm Res
(2020) 37:106
Fig. 1 Structure of the three APLs
considered in this work.
that of miltefosine and the treatment of various cancers
proved unsatisfactory (12–14). Consequently, perifosine can-
not be used as monotherapy but provides interesting results in
combination with radiotherapy (15). Erufosine is a member of
the third generation of APLs. It combines an ω-9-cis-unsatu-
rated alkyl chain with 22 carbon atoms and a homophospho-
choline moiety. Extension of the hydrophobic alkyl chain, as
compared to the previous APLs, results in a decrease in aque-
ous solubility of the compound which organizes into lamellar
structures instead of micelles (16). The critical micellar con-
centration (CMC) for such compounds was thus evaluated in
and examined their intrinsic physicochemical properties.
This has been realized through determination of their hy-
drolytic stability as a function of pH, and analysis of their
surface active and aggregation properties. The antitumor
activity of the pro-APLs in various normal and tumor cell
lines was characterized by IC50 determination, and blood
compatibility was investigated. Whereas compounds
retained full antitumor activity of the parent APLs, they
did not reveal harmful to erythrocytes and did not provoke
hemolysis, the dose-limiting side-effect associated with
APLs.
−7
−8
the range of 10 –10 M which is 2 to 3 orders of magnitude
below that for miltefosine (10− M) (17), suggesting that intra-
venous administration can be considered. Favorable pharma-
cokinetic and pharmacodynamic properties such as potent
cytotoxicity, low gastrointestinal toxicity and reduced myelo-
toxic and hemolytic effects were found (18,19). Erufosine has
been involved in multiple cell-based and preclinical studies
5
MATERIALS AND METHODS
Materials
General information on materials can be found as
Supplementary Information. The preparation of erufo-
sine and Ea-c prodrugs has been described elsewhere
(26).
(20) and phase I clinical trials are expected.
APLs offer an advantage over conventional DNA-
interacting chemotherapeutic agents as they have a unique
mode of action on cell membranes. However, the clinical
development of these compounds is seriously restricted by
their intrinsic surface active properties that translate into
unfavorable pharmacokinetics and pharmacodynamics.
This prompted us to investigate prodrugs of APLs to ad-
dress this issue. Searching carefully in the literature, we
identified only very few derivatives of APLs that could be
potential prodrugs (Fig. 2). The first one (PDΦPC) was
developed in McDonald’s group in 2006 and was formally
the result of the esterification of miltefosine with a
branched diacylglyceride (21). Two other compounds de-
scribed by Øpstad et al. consist in conjugates between mil-
tefosine and polyenoic chromophores, retinoic and carote-
noic acids, through a glycol linker (M-Car and M-Ret,
resp.) (22). These 3 compounds carry a net positive charge
and were designed as DNA transfer reagents. Whether
they can regenerate antitumor miltefosine when exposed
to degradation enzymes in a biological environment was
not investigated by the authors. Starting from our previous
work on DOPC-based biolabile gene transfer reagents
Abbreviations: PE, Petroleum Ether; s, singlet; d, doublet;
t, triplet; q, quadruplet; m, multiplet; br, broad
Miltefosine (M)
Triethylamine (822 μL, 5.90 mmol) was slowly added to fresh-
ly distilled POCl (500 μL, 5.36 mmol) in dry THF (20 mL) at
3
0°C under argon. Then 1-hexadecanol (1.30 g, 5.36 mmol) in
THF (10 mL) was added over 30 min and the resulting mix-
ture was warmed to rt. When all the alcohol had reacted
(checked by TLC), the mixture was cooled to 0°C and a sec-
ond portion of triethylamine (3.3 mL, 21.4 mmol) was added,
followed by 2-bromoethanol (380 μL, 5.36 mmol) in THF
(10 mL). The reaction mixture was stirred overnight at rt.,
decomposed with HCl 10% (10 mL) and heated at 40°C for
2 h. THF was removed under vacuum and the aqueous resi-
due was extracted with CH Cl . The organic layer was dried
2
2
over Na SO , filtered, reduced under vacuum, and the crude
2
4
residue was purified by flash chromatography (CH Cl /
2
2
MeOH 10:0 to 8:2) to yield intermediate 2-bromoethyl hex-
(
23–25), we have designed prodrugs of APLs (pro-APLs)
adecyl hydrogenophosphate (1.01 g, 44%). R : 0.43 (CH Cl /
f
2
2

Pharm Res
(2020) 37:106
Page 3 of 12 106
Fig. 2 Structure of the three
miltefosine derivatives described in
the literature and developed for
gene delivery purposes (21,22).
1
MeOH/H O 75:22:3). H-NMR (400 MHz, MeOD/CDCl
Perifosine (P)
2
3
1
:1) δ 0.85 (t, J = 6.7 Hz, 3H); 1.24 (m, 26H); 1.60 (tt, J =
1
J = 6.4 Hz, 2H); 3.51 (t, J = 6.4 Hz; 2H); 3.86 (td, J = J =
Phosphorus oxychloride (864 μL, 9.2 mmol) was added
2
1
2
1
3
6
.5 Hz, 2H); 4.10 (t, J = J = 6.5 Hz, 2H). C-NMR
to 1-octadecanol (2.50 g, 9.2 mmol) in anhydrous Et O
1
2
2
(
3
100.7 MHz, MeOD/CDCl 1:1) δ 14.3; 23.3; 26.4; 30.0;
0.3 (br); 31.3; 32.6; 66.0; 66.7. P-NMR (162 MHz;
(100 mL) at 0°C, followed by dropwise addition of trie-
thylamine (1.29 mL, 9.2 mmol). The reaction mixture
was stirred at rt. for 16 h, filtered and the filtrate was
reduced under vacuum. The dry residue was dissolved
in anhydrous CHCl3 (100 mL) and N,N-dimethyl-4-
hydroxypiperidinium tosylate (2.78 g, 9.24 mmol), trie-
thylamine (3.00 mL, 21.5 mmol), and 4-DMAP (65 mg,
3
3
1
MeOD/CDCl 1:1) δ – 0.56. IR (ATR) ν 495; 596; 695;
7
3
33; 772; 879; 1007; 1218; 1270; 1456; 2852; 2921.
The previous compound (1.0 g, 2.33 mmol) in
CHCl /CH CN/i-PrOH 3:5:5 (13 mL) was treated with
4
3
3
5% (w/w) aqueous NMe (5 mL, 34.3 mmol) at 70°C.
3
When all starting material was consumed (checked by
TLC), volatile was removed under vacuum and the res-
0.53 mmol) in CHCl (100 mL) were introduced in the
reaction flask and allowed to react for 2 d at rt. The
reaction mixture was then reduced under vacuum, sus-
3
idue was extracted with CHCl /MeOH 1:3. The organ-
3
ic layer was washed with sat. NaCl, reduced under vac-
uum, diluted with CHCl , dried over MgSO , filtered
pended in THF (100 mL) and refluxed with H O
2
(2 mL) for 6 h. The crude mixture was reduced under
vacuum and the residue was directly purified by flash
chromatography (CHCl /MeOH/NH OH 28% 10:6:1)
3
4
and evaporated. The crude residue was purified by flash
chromatography (CH Cl /MeOH/H O 75:22:3 to
2
2
2
3
4
4
5:45:10) to yield miltefosine (0.77 g, 81%). R : 0.1
to yield perifosine (0.70 g, 16%). R : 0.2 (CH Cl /
f 2 2
f
1
1
(
CH Cl /MeOH/H O 75:22:3). H-NMR (400 MHz,
MeOH/H O 75:22:3). H-NMR (400 MHz,
2
2
2
2
MeOD/CDCl3 1:1) δ 0.85 (t, J = 6.2 Hz, 3H); 1.24
MeOD/CDCl3 1:1) δ 0.85 (t, J = 6.5 Hz, 3H); 1.24
(
m, 26H); 1.61 (tt, J = J = 6.4 Hz, 2H); 3.31 (s, 9H);
(m, 30H); 1.59 (m, 2H); 2.10 (m, 4H); 3.09 (s, 3H);
1
2
3
2
1
5
.62 (m, 2H); 3.85 (td, J = J = 6.5 Hz, 2H); 4.23 (m,
H). C-NMR (100.7 MHz, MeOD/CDCl3 1/1) δ
3.16 (s, 3H); 3.25–3.53 (m, 4H); 3.82 (m, 2H); 4.40
1
2
1
3
13
(m, 1H). C-NMR (100.7 MHz, MeOD/CDCl 1:1) δ
3
4.3; 23.2; 26.4; 30.1 (2C); 30.3 (8C); 31.3; 32.5;
14.3; 23.2; 26.4; 27.2 (2C); 29.9; 30.0; 30.2 (10C); 31.5;
3
1
31
4.5; 59.7; 66.7; 67.0.
P-NMR (162 MHz;
32.5; 55.4 (2C); 59.2 (2C); 65.0; 66.6.
P-NMR
MeOD/CDCl3 1:1) δ – 0.32. IR (ATR) ν 497; 716;
(162 MHz; MeOD/CDCl3 1:1) δ – 0.40. IR ν 479;
7
2
46; 853; 927; 957; 1050; 1126; 1246; 1471; 1633;
720; 846; 919; 994; 1062; 1124; 1156; 1225; 1468;
+
+
849; 2914; 3372. HR-MS (ESI+) m/z [M + H] calcd
2848; 2916. HR-MS (ESI+) m/z [M + Na] calcd for
+
+
for C H NO P 408.3237, found 408.3239.
C H NNaO P 484.3526, found 484.3528.
2
1
47
4
25 52
4

1
06 Page 4 of 12
Pharm Res
(2020) 37:106
General Procedure for the Preparation of pro-APLs
general procedure. R : 0.5 (E1) and 0.4 (E2) (CH Cl /
f 2 2
1
MeOH/H O 75:22:3). H-NMR (400 MHz,
2
Pro-APLs were prepared according to a previously reported
procedure (26). Briefly, electrophilic reagent and APL were
MeOD/CDCl3 1:1) E1 δ 0.85 (t, J = 6.1 Hz, 6H);
1.24 (m, 42H); 1.58 (d, J = 5.3 Hz, 3H); 1.61 (tt, J =
1
stirred in refluxing dry CHCl for 24 h under argon. Volatile
J = 6.8 Hz, 2H); 1.69 (tt, J = J = 6.6 Hz, 2H); 2.37 (t,
3
2
1
2
was distilled off and the residue was purified by flash chroma-
tography (CH Cl /MeOH 10:0 to 7:3) to yield the
corresponding pro-APL.
J = 7.2 Hz, 2H); 3.24 (s, 9H); 3.76 (m, 2H); 4.09 (td,
J = J = 6.7 Hz, 2H); 4.50 (m, 2H); 6.42 (qd, J = J =
2
2
1
2
1
2
1
5.3 Hz, 1H). H-NMR (400 MHz, MeOD/CDCl 1:1)
3
E2 δ 0.86 (t, J = 6.2 Hz, 6H); 1.24 (m, 42H); 1.56 (d,
J = 5.3 Hz, 3H); 1.61 (tt, J = J = 6.8 Hz, 2H); 1.68 (tt,
2
-(((Dodecyloxy)(hexadecyloxy)phosphoryl)oxy)
1
2
-
N,N,N-trimethylethan-1-aminium triflate (Ma)
J = J = 6.6 Hz, 2H); 2.35 (t, J = 7.3 Hz, 2H); 3.24 (s,
1 2
9
H); 3.77 (m, 2H); 4.09 (td, J = J = 6.3 Hz, 2H); 4.48
1 2
1
3
Compound M (100 mg, 84%) was obtained from miltefosine
(m, 2H); 6.48 (qd, J = J = 5.3 Hz, 2H). C-NMR
1 2
a
(
0
80 mg, 0.20 mmol) and dodecyl triflate (27) (162 mg,
.53 mmol) according to the general procedure except the
reaction was conducted at rt. R : 0.7 (CH Cl /MeOH/H O
(100.7 MHz, MeOD/CDCl3 1:1) E1 δ 14.5 (2C);
21.7; 23.4 (2C); 25.3; 26.1; 29.6; 29.7; 29.8; 29.9;
30.0 (2C); 30.1 (2C); 30.2 (4C); 30.3 (4C); 30.7; 32.5
f
2
2
2
1
13
7
6
9
5:22:3). H-NMR (400 MHz, CDCl ) δ 0.88 (t, J = 6.6 Hz,
(2C); 34.5; 54.6 (3C); 62.3; 66.5; 70.0; 92.4; 173.3. C-
3
H); 1.26 (m, 44H); 1.68 (tt, J = J = 6.8 Hz, 4H); 3.31 (s,
NMR (100.7 MHz, MeOD/CDCl 1:1) E2 δ 14.5 (2C);
1
2
3
1
3
H); 3.83 (m, 2H); 4.08 (m, 4H); 4.48 (m, 2H). C-NMR
21.6; 23.4 (2C); 25.3; 26.1; 29.6; 29.7; 29.8; 29.9; 30.0
(2C); 30.1 (2C); 30.2 (4C); 30.3 (4C); 30.7; 32.5 (2C);
(100.7 MHz, CDCl ) δ 14.3 (2C); 22.9 (2C); 25.6 (2C); 29.4
3
3
1
(2C); 29.6 (2C); 29.7 (2C); 29.8 (4C); 29.9 (6C); 30.5 (2C); 32.1
34.5; 54.6 (3C); 62.0; 66.3; 70.2; 92.3; 172.8. P-NMR
3
1
31
(2C); 54.6 (3C); 61.2; 65.9; 69.2 (2C). P-NMR (162 MHz,
(162 MHz, MeOD/CDCl 1:1) E1 δ – 5.96. P-NMR
3
CDCl ) δ – 2.14. IR ν 517; 574; 638; 1030; 1161; 1226; 1253;
(162 MHz, MeOD/CDCl 1:1) E2 δ – 5.57. IR ν 509;
3
3
+
1
467; 2852; 2921; 3500. HR-MS (ESI+) m/z [M-Cl] calcd
721; 934; 976; 1050; 1082; 1165; 1238; 1267; 1467;
1755; 2849; 2916; 2956; 3389. HR-MS (ESI+) m/z
+
for C H NO P 576.5115, found 576.5108.
3
3
71
4
+
+
[
M-Cl] calcd for C H NO P 634.5170, found
35 73 6
2
-((((Dodecanoyloxy)methoxy)(hexadecyloxy)phosphoryl)oxy)
634.5167. Note: IR and HR-MS were measured on
the mixture of the four diastereomers.
-
N,N,N-trimethylethan-1-aminium chloride (Mb)
Compound M (136 mg, 42%) was obtained from miltefosine
4-(((Dodecyloxy)(octadecyloxy)phosphoryl)oxy)
b
(
(
0
200 mg, 0.49 mmol) and chloromethyl dodecanoate (24)
-1,1-dimethylpiperidin-1-ium chloride (Pa)
979 mg, 3.94 mmol) according to the general procedure. R :
f
1
.5 (CH Cl /MeOH/H O 75:22:3). H-NMR (400 MHz,
Compound P (55 mg, 48%) was obtained from perifosine
2
2
2
a
MeOD/CDCl 1:1) δ 0.85 (t, J = 6.3 Hz, 6H); 1.24 (m,
(79 mg, 0.17 mmol) and dodecyl triflate (141 mg, 0.44 mmol)
according to the general procedure except the reaction was
conducted at rt. R = 0.68 (CH Cl /MeOH/H O 75:22:3).
3
4
6
2
2H); 1.63 (tt, J = J = 6.8 Hz, 2H); 1.69 (tt, J = J =
1 2 1 2
.6 Hz, 2H); 2.40 (t, J = 7.5 Hz, 2H); 3.23 (s, 9H); 3.76 (m,
H); 4.11 (td, J = J = 6.3 Hz, 2H); 4.50 (m, 2H); 5.64 (ABX
f
2
2
2
1
H-NMR (400 MHz, MeOD/CDCl 1:1) δ 0.86 (t, J =
1
2
3
1
3
syst., JAB = 5.2 Hz, JAX = 13.4 Hz, JBX = 11.4 Hz, 2H). C-
6.7 Hz, 6H); 1.24 (m, 48H); 1.68 (tt, J = J = 7.0 Hz, 4H);
1 2
NMR (100.7 MHz, MeOD/CDCl 1:1) δ 14.5 (2C); 23.3
2.10–2.32 (m, 4H); 3.14 (s, 3H); 3.20 (s, 3H); 3.41–3.55 (m,
4H); 4.06 (m, 4H); 4.62 (m, 1H). C-NMR (100.7 MHz,
3
1
3
(
2C); 25.1; 25.9; 29.7; 29.8; 29.9; 30.0 (2C); 30.1 (4C); 30.2
7C); 30.7; 32.5; 34.6; 54.7 (3C); 62.3; 66.5; 70.2; 84.2; 173.2.
(
CDCl ) δ 14.4 (2C); 23.0 (2C); 25.7 (2C); 26.9 (2C); 29.6
3
3
1
P-NMR (162 MHz, MeOD/CDCl 1:1) δ – 3.80. IR ν 458;
(2C); 29.7 (2C); 29.9 (2C); 30.1 (5C); 30.2 (6C); 30.5; 30.6
3
3
1
4
2
90; 721; 832; 873; 965; 1042; 1159; 1264; 1468; 1760; 2849;
(2C); 32.2 (2C); 47.1 (2C); 59.1 (2C); 69.2; 69.3 (2C). P-
+
917; 2956; 3382. HR-MS (ESI+) m/z [M-Cl] calcd for
NMR (162 MHz; MeOD/CDCl 1:1) δ – 2.07. IR ν 516;
3
+
C H NO P 620.5014, found 620.5004.
573; 638; 721; 764; 923; 1029; 1161; 1245; 1467; 2851;
3
4
71
6
+
2
920; 3252. HR-MS (ESI+) m/z [M-Cl] calcd for
+
2
-(((1-(Dodecanoyloxy)ethoxy)(hexadecyloxy)phosphoryl)oxy)
C H NO P 630.5585, found 630.5581.
37 77 4
-
N,N,N-trimethylethan-1-aminium chloride (Mc)
4-((((Dodecanoyloxy)methoxy)(octadecyloxy)phosphoryl)oxy)
Compound Mc (169 mg, 51%) was obtained as two
separated couples of enantiomers (E1 and E2) from mil-
tefosine (201 mg, 0.49 mmol) and 1-chloroethyl dodec-
anoate (24) (1.04 g, 3.95 mmol) according to the
-1,1-dimethylpiperidin-1-ium chloride (Pb)
Compound P (51 mg, 22%) was obtained from perifosine
b
(150 mg, 0.33 mmol) and chloromethyl dodecanoate

Pharm Res
(2020) 37:106
Page 5 of 12 106
(
0
675 mg, 2.70 mmol) according to the general procedure. R :
.6 (CH Cl /MeOH/H O 75:22:3). H-NMR (400 MHz,
2 2 2
RESULTS AND DISCUSSION
f
1
MeOD/CDCl 1:1) δ 0.86 (t, J = 6.7 Hz, 6H); 1.24 (m,
Design and Synthesis of the pro-APLs
3
4
7
3
6H); 1.63 (tt, J = J = 6.9 Hz, 2H); 1.68 (tt, J = J =
.2 Hz, 2H); 2.10–2.32 (m, 4H); 2.39 (t, J = 7.6 Hz, 2H);
1 2 1 2
Phosphotriesters as those presented at Fig. 2 may be consid-
ered as potential APL prodrugs. Nevertheless, three different
phosphodiesters will result from their non-selective hydrolysis,
of which the expected APL. The two other phosphodiesters
will be produced through the alternate leaving of the 1-
dodecanol or choline substituent. In order to avoid these un-
wanted side reactions, the parent APL should be derivatized
through the grafting of a labile substituent. Under a suitable
chemical or enzyme stimulus, selective removal of the latter
may be obtained. Thus, considering our previous work on
membrane phospholipid-based phosphoacetals as labile cat-
ionic lipids for gene delivery (24,25), we explored the possibil-
ity to develop our methodology for producing potent prodrugs
of APLs (Fig. 3). To this end, a dodecyl chain was tethered to
the phosphate group of APLs, directly or through an acetal
moiety. It has been previously demonstrated that a methyl
substituent on the acetal bridge improves hydrolytic suscepti-
bility of the phosphotriester (25). Therefore, this modification
was introduced in the pro-APL series. As reference com-
pounds and starting materials for the preparation of pro-
APLs, miltefosine, perifosine and erufosine were first prepared
(Sch. 1). Miltefosine was synthesized in two steps according to
a protocol described by North et al. (28) with modifications
allowing improvement of the reaction yield from 5.1 to
35%. Perifosine was obtained using a one-pot procedure start-
ing from 1-octadecanol, phosphorus oxychloride and N,N-di-
methyl-4-hydroxypiperidinium tosylate. Erufosine was pre-
pared in two steps from erucyl alcohol, 1,3-propanediol, and
phosphorus oxychloride (26). The intermediate 1,3-dioxa-2-
phosphinane was reacted with trimethylamine to afford the
target compound in 75% yield. The nine pro-APL com-
pounds were then prepared through the direct O-alkylation
of APLs by adequate electrophilic reagents, namely 1-dodecyl
triflate, chloromethyl dodecanoate, and 1-chloroethyl dodec-
anoate. Reaction yields were variable, depending on the nu-
cleophilicity of the parent APL, perifosine offering lower re-
activity presumably due to higher stability of the zwitterion.
M , P , and E were obtained as two couples of enantiomers
.17 (s, 3H); 3.22 (s, 3H); 3.45–3.59 (m, 4H); 4.08 (td, J =
1
J = 6.8 Hz, 2H); 4.70 (m, 1H); 5.62 (ABX syst., J = 5.2 Hz,
2
AB
1
3
JAX = 13.8 Hz, JBX = 11.1 Hz, 2H). C-NMR (100,7 MHz;
MeOD/CDCl 1:1) δ 14.5 (2C); 23.4 (2C); 25.3; 26.1; 27.1
3
(
2C); 29.8; 29.9; 30.0; 30.1 (2C); 30.2; 30.3 (5C); 30.4 (7C);
3
1
0.9; 32.6 (2C); 34.6; 54.7 (2C); 58.9 (2C); 69.9; 70.0; 83.5;
3
1
73.1. P-NMR (162 MHz, MeOD/CDCl 1:1) δ – 4.22.
3
IR ν 531; 666; 720; 760; 802; 923; 966; 1020; 1148; 1264;
1
)
468; 1644; 1764; 2849; 2916; 2955; 3378. HR-MS (ESI+
+
+
m/z [M-Cl] calcd for C H NO P 674.5483, found
3
8
77
6
6
74.5477.
4
-(((1-(Dodecanoyloxy)ethoxy)(octadecyloxy)phosphoryl)oxy)
-
1,1-dimethylpiperidin-1-ium chloride (Pc)
Compound P (96 mg, 20%) was obtained as two sep-
c
arated couples of enantiomers (E1 and E2) from perifo-
sine (300 mg, 0.65 mmol) and 1-chloroethyl dodeca-
noate (1.33 g, 5.06 mmol) according to the general pro-
cedure. R : 0.3 (E1) and 0.4 (E2) (CH Cl /MeOH/H O
f
2
2
2
1
7
5:22:3). H-NMR (400 MHz, MeOD/CDCl 1:1) D1
3
δ 0.86 (t, J = 6.7 Hz, 6H); 1.24 (m, 46H); 1.56 (d, J =
.2 Hz, 3H); 1.61 (tt, J = J = 6.9 Hz, 2H); 1.68 (tt,
5
1
2
J = J = 7.2 Hz, 2H); 2.08–2.30 (m, 4H); 2.36 (t, J =
1
2
7
4
.6 Hz, 2H); 3.17 (s, 3H); 3.24 (s, 3H); 3.45–3.59 (m,
H); 4.06 (td, J = J = 6.8 Hz, 2H); 4.70 (m, 1H); 6.46
1
2
1
(
qd, J = J = 5.2 Hz, 1H). H-NMR (400 MHz,
1 2
MeOD/CDCl3 1:1) E2 δ 0.86 (t, J = 6.7 Hz, 6H);
.24 (m, 46H); 1.54 (d, J = 5.2 Hz, 3H); 1.61 (tt, J =
1
1
J = 6.9 Hz, 2H); 1.68 (tt, J = J = 7.2 Hz, 2H); 2.08–
2
1
2
2
3
6
1
1
2
3
9
1
2
.30 (m, 4H); 2.35 (t, J = 7.6 Hz, 2H); 3.18 (s, 3H);
.24 (s, 3H); 3.45–3.59 (m, 4H); 4.08 (td, J = J =
1
2
.8 Hz, 2H); 4.70 (m, 1H); 6.50 (qd, J = J = 5.2 Hz,
1
2
1
3
H). C-NMR (100.7 MHz, MeOD/CDCl 1:1) E1 δ
3
4.5 (2C); 21.7; 23.4 (2C); 25.3; 26.1; 26.8; 27.1;
9.7; 29.8; 30.0; 30.1 (2C); 30.2; 30.3 (5C); 30.4 (7C);
0.8; 32.6 (2C); 34.7; 55.2 (2C); 58.6 (2C); 69.5; 69.7;
c
c
c
1
3
2.1; 173.1. C-NMR (100.7 MHz, MeOD/CDCl3
:1) E2 δ 14.5 (2C); 21.7; 23.4 (2C); 25.3; 26.1; 27.0;
7.1; 29.7; 29.8; 30.0; 30.1 (2C); 30.2; 30.3 (5C); 30.4
due to the two stereogenic centers in the molecule. In some
cases, the isomers could be separated but evaluations (vide infra)
were performed on the original mixture of diastereomers.
(
7C); 30.8; 32.6 (2C); 34.7; 54.8 (2C); 58.8 (2C); 69.7;
3
1
6
1
1
1
9.9; 92.3; 172.8. P-NMR (162 MHz, MeOD/CDCl3
Self-Assembly Properties of the pro-APLs
3
1
:1) E1 δ – 6.1. P-NMR (162 MHz, MeOD/CDCl3
:1) E2 δ – 6.0. IR ν 553; 638; 772; 752; 974; 1012;
090; 1163; 1263; 1466; 1745; 2851; 2920; 3380. HR-
There is broad consensus that APLs may act as detergent
molecules and display membrane-active properties. At high
concentration, these compounds impair the orderly structure
of the plasma membrane. Micellar clusters are produced,
leading to the formation of pores and holes in the membranes
+
+
MS (ESI+) m/z [M-Cl] calcd for C H NO P
3
9
79
6
6
88.5640, found 688,5627. Note: IR and HR-MS were
measured on the mixture of the four diastereomers.

1
06 Page 6 of 12
Pharm Res
(2020) 37:106
Fig. 3 Structure of the pro-APL
compounds described in this work.
of tumor cells (29). In order to investigate the behavior of
APLs and pro-APLs in aqueous environment, the fluorescence
probe technique was selected. This technique has been widely
used to investigate micellization of surfactants and determine
their CMC. In the case of amphiphilic compounds with a low
hydrophilic-lipophilic balance (HLB) as it is most likely the
case with the pro-APLs described herein, aggregation may
preferentially proceed through the formation of lamellar
structures. In this case, the concentration above which self-
assembly occurs should be referred to the critical aggregation
concentration (CAC) instead of CMC. In our study, we select-
ed pyrene as the hydrophobic fluorescent dye. Pyrene fluores-
cence vary with the polarity of the solubilizing medium and
various fluorescence patterns are observed where the dye en-
vironment is hydrophilic or lipophilic. Measurements were
performed in water and data obtained for the APLs and
pro-APLs are collected in Table I. The CMC values obtained
for miltefosine, perifosine and erufosine were 11.2, 4.4, and
1.4 μM, respectively. The CMC determined for miltefosine
was consistent with that previously determined by the Du
Noüy ring method (12 μM) (30) whereas the Langmuir
trough method yielded a somehow lower value (2.5–
3.0 μM) (31). For perifosine, the CMC was found to be
4.4 μM which is similar to the value determined by the
Langmuir trough method (2.5 μM) (11). No CMC data for
erufosinecould befoundintheliterature. Asexpected, trans-
formation of APLs into pro-APLs significantly decreased the
CAC of the compounds. In the miltefosine series, the effect
was appreciably greater with a ten-fold decrease in the CAC
value when compared to the parent compound, whereas a
two- to three-fold decrease was observed in the perifosine
and erufosine series. This is explained by the longest alkyl
chain residue of these two APLs and consequently, their
strongest initial lipophilic nature.
Scheme 1 Synthesis of the parent APLs, miltefosine, perifosine and erufosine, and subsequent pro-APLs.

Pharm Res
(2020) 37:106
Page 7 of 12 106
Tab le I APLs and Pro-APLs Critical Aggregation Concentration as
Measured with Pyrene by the Fluorescent Probe Technique
phosphotriesters and phosphodiesters differ by 5–6 ppm.
This variation allows accurate monitoring of the hydrolysis
reaction. The pro-APLs were processed into liposomes at neu-
tral and acidic pH, and incubated at 25°C. Periodical acqui-
M
M
a
M
b
M
c
P
P
a
P
b
P
c
E
E
a
E
b
E
c
CAC (μM) 11.2 1.1 4.3 1.2 4.4 1.2 2.9 2.3 1.4 0.5 0.5 0.5
31
sition of the P-NMR spectra under conditions enabling
quantitative measurements (4-s pulse cycle) were carried out
to determine the rate of hydrolysis (Fig. 4, Table II). As might
be anticipated, phosphotriesters without acetal linker (M , P ,
Hydrolytic Properties of the pro-APLs
a
a
and E ) retained their full structural integrity under neutral
a
Aqueous formulation of pro-APLs results in the formation of
nanosized lipid aggregates (liposomes) that most likely are in-
ternalized by cells through the endocytic route. To express
their intrinsic antitumor activity, pro-APLs need to be trans-
formed into APLs which are the active drugs. Such a transfor-
mation may be triggered by a chemical stimulus, e.g., a pH
decrease, such as the one observed in the endosomal compart-
ment during the maturation process, or can be mediated by
hydrolytic enzymes that are abundant in lysosomes. We thus
investigated the resistance of the compounds to hydrolysis un-
der neutral and acidic conditions, mimicking the extracellular
and acidic conditions, even over an extended period of time (>
31d). Though to various extents, all the other compounds
revealed sensitive to hydrolysis. Overall, hydrolysis of the
compounds was accelerated at pH 7.4 as compared to
pH 4.5. This indicates that mixed acetals of phosphoric
and carboxylic esters do not display the same reactivity as
dialkyl acetals which are fairly stable in neutral or basic
media and highly labile under acidic conditions. These
results are consistent with those previously obtained with
other phosphoacetals (24,25) and point out that hydrolysis
of such compounds cannot be analyzed just considering
the intrinsic reactivity of the acetal moiety but must be
interpreted with regard to the respective sensitivity of the
3
1
and late endosome environment, respectively, using P-
3
1
NMR spectroscopy. P chemical shifts displayed by
Fig. 4 Hydrolytic stability of the
pro-APLs. Compounds formulated
into liposomes were incubated at
2
5°C and hydrolysis was monitored
31
by quantitative P-NMR
measurements, at pH 7.4 and 4.5
(
(a): miltefosine prodrugs; (b):
perifosine prodrugs; (c): erufosine
prodrugs).

1
06 Page 8 of 12
Pharm Res
(2020) 37:106
Table II Susceptibility of the pro-APLs to hydrolysis. Pro-APLs were incu-
and hemolysis was monitored spectrophotometrically. As
expected, hemolysis increased in a concentration-dependent
manner (Fig. 5). However, whereas APLs provoked massive
hemolysis within 1 h (Fig. 5a), pro-APLs revealed much less
deleterious to RBC. At the highest concentration tested
(200 μM), M , M , and M only provoked 21%, 16%, and
bated at pH 7.4 and 4.5, at 25°C. Extent of the hydrolysis reaction was
31
determined by quantitative P-NMR spectroscopy. H120 refers to the extent
of hydrolysis (%) after a 120-h incubation period. When possible, time re-
quired for 50% hydrolysis (t1/2) was calculated from the theoretical curve
fitting with the experimental data
*
a
*
*
a
a
b
c
M
M
b
M
c
P
a
P
b
P
c
E
E
b
E
c
73% hemolysis, respectively (Fig. 5b). By comparison, milte-
H
120 (%) pH 7.4
0
20 89
0
0
–
–
16 60
0
0
–
–
17
2
56
53
96
fosine provoked complete disruption of RBC at the concen-
pH 4.5
0
–
–
3
–
–
91
18
17
2
–
–
60
84
90
tration of ca. 100 μM. The higher hemolytic toxicity of M , as
c
t1/2 (h)
pH 7.4
pH 4.5
–
compared to M and M , was consistent with the lower hy-
a
b
–
104
drolytic stability of this compound (vide supra), hemolysis being
caused by miltefosine that was produced in situ. All the other
pro-APLs investigated herein did not significantly alter RBC
*
No hydrolysis was observed after an incubation period of 31 d
(
hemolysis < 2%) over a 1-h incubation period. Thus, in order
ester moieties (carboxylic and phosphoric) towards hydro-
lysis in aqueous solutions, depending on the reaction
mechanism involved (specific base catalyzed, water cata-
lyzed, or specific acid catalyzed). Fully consistent with pri-
or examples in the literature, introduction of a methyl
substituent on the acetal bridge also invariably provoked
a drastic increase in the hydrolysis rate of the pro-APLs,
both under basic and acidic conditions. In the miltefosine
series, whereas only 3–20% hydrolysis were measured af-
ter 5 days incubation of the compound lacking acetal sub-
to characterize and discriminate the pro-APL compounds,
incubation time was extended and hemolysis was measured
at 24 h (Fig. 6). Concentration of the compounds inducing
5
0% hemolysis (hemolytic concentration for 50% effect,
HC ) and percentage of hemolysis induced at the concentra-
5
0
tion of 200 μM (hemolytic activity at 200 μM, HA200) are
displayed in Table III. Among the three pro-APLs series, mil-
tefosine prodrugs displayed the higher residual hemolytic tox-
icity, whereas pro-erufosine compounds had only minimal
effect on RBC, and perifosine prodrugs exhibited intermedi-
ate properties. Besides, as might be anticipated, the more la-
bile the compounds, the higher the hemolysis, and M , P , and
Ec were invariably associated with the higher residual hemo-
lytic toxicity in each prodrug series. Nevertheless, the data
clearly show that pro-APLs display drastically reduced hemo-
lytic effect as compared to the parent APLs. Hemolysis pro-
voked by perifosine and miltefosine prodrugs after 24 h was
much lower in most cases than that induced by erufosine itself
after only 1 h. Besides, erufosine prodrugs were practically
devoided of any hemolytic activity (HA200 < 25%, after
24 h). As erufosine is the only APL currently considered for
intravenous administration (37), it can be concluded that all
the pro-APLs described herein could be safely qualified for
this administration route.
stitution (M ), methyl substitution (M ) led to 50% hydro-
b
c
lysis within 17–18 h, regardless of the pH. This roughly
corresponded to an acceleration of the hydrolysis rate by
a factor of 5 at pH 7.4, and by a factor of 120 at pH 4.5.
The same trend was observed in the perifosine and eru-
fosine series.
c
c
Hemolytic Activity of the pro-APLs
Most APLs described to date display hemolytic activity that is
incompatible with administration through the iv route (7,13).
Though hemolysis can be partly prevented if the APLs are
processed in liposomal form (i.e., co-formulation with other
lipids such as cholesterol and 1,2-dipalmitoyl-sn-glycero-phos-
phoglycerol) (32,33), a significant reduction in therapeutic ac-
tivity is achieved in this case, as previously shown in vitro (34).
The presence of micelles is important for the cell lysis mecha-
nism, which involves micellar dissolution and partition into
the cell membrane, phase transition between lamellae and
micelles and decrease in the size of micelles (35). In the ab-
sence of serum, it has been shown that the lytic concentrations
for different APLs and related compounds correspond rather
well to their CMC which are in the low μM range (7,36).
Thus, lowering the CMC of an APL through the formation
of a more hydrophobic prodrug, a pro-APL, might result in
improving the biocompatibility of the compound. In order to
examine this hypothesis, the hemolytic activity of APLs and
pro-APLs was measured. Sheep red blood cells (RBC) were
incubated at 37°C with increasing doses of the compounds
a
Concentration of pro-APL inducing 50% hemolysis.
Percentage of hemolysis provoked by the compound at
b
200 μmol
Cytotoxic Activity of the pro-APLs
In the presence of serum, direct membrane permeabilization
and cell lysis is not a major cytotoxic mechanism of action of
APLs at pharmacologically relevant concentration (38,39). It
is widely appreciated that, at clinical dose, APLs insert into the
cell membrane, producing a biophysical disturbance by alter-
ing cholesterol homeostasis, and finally disturbing various

Pharm Res
(2020) 37:106
Page 9 of 12 106
Fig. 5 Hemolytic activity of APLs
a) and miltefosine prodrugs (b)
(
incubated with sheep RBC at 37°C
for 1 h.
signal transduction pathways (40). To evaluate the antiproli-
ferative effect of the APL prodrugs, a concentration-response
study was performed in tumor and non-tumor pulmonary
cells: A549 (human epithelial lung carcinoma), NCI-H292
Fig. 6 Hemolytic properties of the
pro-APLs ((a): miltefosine prodrugs;
(b): perifosine prodrugs; (c): erufo-
sine prodrugs) incubated with
sheep red blood cells at 37°C for
2
4 h.

1
06 Page 10 of 12
Pharm Res
(2020) 37:106
Table III Hemolytic Activity of pro-APLs upon Incubation with Sheep RBC
for 24 h at 37°C
Table IV Cytotoxic Effect (IC50) of APLs and Pro-APLs in Three Cell Lines
Exposed for 24 h at 37°C, as Measured by the MTT Assay
M
a
M
b
M
c
P
a
P
b
P
c
E
a
E
b
E
c
Compound
IC50 (μM)
a
HC50 (μM) 84 89 69 > 200 141 84 > 200 > 200 > 200
A549
NCI-H292
16HBE
73 ± 8
b
HA200 (%) 100 100 100
36 21 80
12
19
25
M
77 ± 15
46 ± 11
39 ± 16
139 ± 32
127 ± 39
56 ± 7
70 ± 9
56 ± 7
M
a
M
b
M
c
P
107 ± 14
69 ± 26
49 ± 6
52 ± 28
59 ± 10
77 ± 3
(
mucoepidermoid pulmonary carcinoma), and 16HBE (bron-
chial epithelial cells). The nine pro-APLs were assayed and
compared with parent miltefosine, perifosine, and erufosine.
Cells were exposed to concentration of lipids between 500 nM
and 1 mM, and cell survival was determined after a 24-
h incubation period by the MTT dye reduction assay. A de-
crease in cell viability was observed in response to increasing
concentration of APLs and pro-APLs in the three cell lines
116 ± 33
72 ± 7
P
a
63 ± 5
P
P
E
b
139 ± 4
148 ± 38
57 ± 8
168 ± 7
116 ± 25
46 ± 8
231 ± 70
60 ± 5
c
19 ± 3
E
a
E
b
E
c
81 ± 7
124 ± 11
125 ± 26
73 ± 15
225 ± 31
144 ± 66
63 ± 7
151 ± 42
192 ± 44
(
Fig. 7). However, the cell lines showed different sensitivity to
micellar APLs and lamellar pro-APLs, possibly indicating var-
ious mechanisms for the binding or uptake. The observation
that no complete suppression of cell viability was attained,
even at the highest concentration of 1 mM, is related to the
known partial underestimation of antiproliferative agents by
the MTT assay (41). The IC50 values (concentration that
inhibited the cell growth by 50%) reflecting the cytotoxicity
of the compounds interpolated from the experimental
concentration–effect curves are presented in Table IV. The
IC50 values were in the range 39–192 μM for A549 cells, 46–
1
68 μM for NCI-H292 cells, and 19–231 μM for non-
cancerous 16HBE cells. The cytotoxicity of pro-APLs did
not correlate with their previously measured hydrolytic stabil-
ity. This likely originated from in situ enzyme-mediated hydro-
lysis of the pro-APL compounds, a process competing with
pure pH-controlled chemical hydrolysis. Depending on the
pro-APL (i.e., the structure of the parent APL and that of
Fig. 7 Survival of A549, NCI-H292, and 16HBE cells treated with increasing concentrations of APLs and pro-APLs.

Pharm Res
(2020) 37:106
Page 11 of 12 106
the substituent grafted to the phosphoryl moiety), the com-
pound may be metabolized into the the parent cytotoxic
APL at various rates by lipases in the milieu. Anyhow, every
pro-APL tested herein displayed a cytotoxicity that compared
to that of the parent APL, revealing that it was productively
metabolized into the cells. Nevertheless, depending on the cell
line investigated, some pro-APLs revealed in some cases more
active than the corresponding APL (see e.g. miltefosine and
perifosine prodrugs in A549 and NCI-H292 cells). One expla-
nation would be a higher internalization rate through the
endocytic uptake routes and/or a more efficient addressing
of the pro-APL or its hydrolysis product to the plasma mem-
brane. In the erufosine series, such an effect was not observed
and none of the prodrugs offered enhanced cytotoxicity when
compared to the APL.
option, not only for cancer, but also for the treatment of other
diseases (e.g., visceral and cutaneous leishmaniasis).
ACKNOWLEDGMENTS AND DISCLOSURES. The authors are
grateful for financial support to BG from Labex Medalis,
Région Alsace, and Alsace contre le cancer.
COMPLIANCE WITH ETHICAL STANDARDS
Conflict of Interest The authors have no competing interests to
declare.
REFERENCES
1
.
Modolell M, Andreesen R, Pahlke W, Brugger U, Munder PG.
Disturbance of phospholipid metabolism during the selective de-
struction of tumor cells induced by alkyl lysophospholipids.
Cancer Res. 1979;39(11):4681–6.
CONCLUSION
2
3
.
.
Kostadinova A, Topouzova-Hristova T, Momchilova A, Tzoneva
R, Berger MR. Antitumor lipids - structure, functions, and medical
applications. Adv Protein Chem Struct Biol. 2015;101:27–66.
Breiser A, Kim DJ, Fleer EAM, Damenz W, Drube A, Berger M,
et al. Distribution and metabolism of hexadecylphosphocholine in
mice. Lipids. 1987;22:925–6.
In the current study, we investigated biolabile prodrugs of
APLs for the first time. The prodrugs were obtained through
direct O-alkylation of miltefosine, perifosine, and erufosine
with various electrophilic reagents (dodecyl triflate, chloro-
methyl dodecanoate, and 1-chloroethyl dodecanoate).
Aggregation properties of the resulting pro-APLs were exam-
ined and their critical aggregation concentration was deter-
mined. Transformation of APLs into pro-APLs resulted in a
significant decrease in water solubility of the compounds and,
consequently, in their CAC. Besides, the hydrolytic stability of
the pro-APLs was determined under neutral and acidic con-
ditions mimicking the extracellular environment and the late
endosome milieu, respectively. Pro-APLs incorporating a
methyl substituted phosphoacetal bridge revealed exacerbat-
ed sensitivity to hydrolysis, both under neutral and acidic con-
ditions. Furthermore, the decrease in CAC for the pro-APLs,
as compared to the parent APLs, did translate into a drastic
reduction of the hemolytic toxicity of the compounds. All the
pro-APLs were less harmful to red blood cells than erufosine,
the only APL that has been considered for intravenous admin-
istration to date. In order to determine whether pro-APLs
inherited the antineoplastic activity of the parent APLs, three
different cell lines were exposed to increasing concentrations
of the compounds and IC50 were determined. Cytotoxicity of
the pro-APLs was found in the low micromolar range, similar
to that of APLs and thus revealing that the APL prodrugs were
productively metabolized by cells into the active parent com-
pounds. In conclusion, our results indicate that pro-APLs are
intravenously injectable derivatives of APLs with similar anti-
proliferative efficacy. Their use could thus address one of the
major issues met in cancer therapies involving antitumor lipids
and restricting their utilization to oral and topical administra-
tion, because of limited maximum tolerated dose. Therefore
pro-APLs may be suggested as an interesting therapeutic
4. Yanapirut P, Berger MR, Reinhardt M, Schmähl D. In vitro inves-
tigations on the antineoplastic effect of hexadecylphosphocholine.
Arzneimittelforschung. 1991;41:652–5.
5
6
7
.
.
.
Eibl H, Hilgard P, Stekar J, Voegeli R, Harleman JH.
Experimental therapeutic studies with miltefosine in rats and mice.
Prog Exp Tumor Res. 1992;34:116–30.
Verweij J, Planting AST, Stoter G, Gandia D, Armand JP. Phase II
study of oral miltefosine in patients with squamous cell head and
neck cancer. Eur J Cancer. 1993;29:778–9.
Kötting J, Marschner NW, Neumuller W, Unger C, Eibl H.
Hexadecylphosphocholine and octadecyl-methyl-glycero-3-phos-
phocholine: a comparison of hemolytic activity, serum binding
and tissue distribution. Prog Exp Tumor Res. 1992;34:131–42.
Hilgard P, Klenner T, Stekar J, Nössner G, Kutscher B, Engel J.
D-21266, a new heterocyclic alkylphospholipid with antitumour
activity. Eur J Cancer. 1997;33:442–6.
8
9
.
.
Gills JJ, Dennis PA. Perifosine: update on a novel Akt inhibitor.
Curr Oncol Rep. 2009;11:102–10.
1
0. Vink SR, Schellens JHM, van Blitterswijk WJ, Verheij M. Tumor
and normal tissue pharmacokinetics of perifosine, an oral anti-
cancer alkylphospholipid. Investig New Drugs. 2005;23:279–86.
11. Mravljak J, Reiner Z, Pečar S. Synthesis and biological evaluation
of spin-labeled alkylphospholipid analogs. J Med Chem. 2005;48:
6
393–9.
1
2. Hideshima T, Catley L, Yasui H, Ishitsuka K, Raje N, Mitsiades C,
et al. Perifosine, an oral bioactive novel alkylphospholipid, inhibits
Akt and induces in vitro and in vivo cytotoxicity in human multiple
myeloma cells. Blood. 2006;107:4053–62.
3. Ernst DS, Eisenhauer E, Wainman N, Davis M, Lohmann R, Baetz
T, et al. Phase II study of perifosine in previously untreated patients
with metastatic melanoma. Investig New Drugs. 2005;23:569–75.
14. Leighl NB, Dent S, Clemons M, Vandenberg TA, Tozer R, Warr
DG, et al. A phase II study of perifosine in advanced or metastatic
breast cancer. Breast Cancer Res Treat. 2008;108:87–92.
5. Cirstea D, Hideshima T, Rodig S, Santo L, Pozzi S, Vallet S, et al.
Dual inhibition of Akt/mammalian target of rapamycin pathway
by nanoparticle albumin-bound-rapamycin and perifosine induces
1
1

1
06 Page 12 of 12
Pharm Res
(2020) 37:106
antitumor activity in multiple myeloma. Mol Cancer Ther. 2010;9:
lysophosphatidylcholine analogs. Bioorg Med Chem. 2009;17:
3433–42.
9
63–75.
1
6. Kötting J, Berger MR, Unger C, Eibl H. Alkylphosphocholines:
influence of structural variation on biodistribution at antineoplasti-
cally active concentrations. Cancer Chemother Pharmacol.
29. Noseda A, White JG, Godwin PL, Jerome WG, Modest EJ.
Membrane damage in leukemic cells induced by ether and ester
lipids - an electron microscopic study. Exp Mol Pathol. 1989;50:
69–83.
1
992;30:105–12.
1
7. Eibl H, Kaufmann-Kolle P. Medical application of synthetic phos-
pholipids as liposomes and drugs. J Liposome Res. 1995;5:131–48.
8. Yosifov DY, Todorov PT, Zaharieva MM, Georgiev KD, Pilicheva
BA, Konstantinov SM, Berger MR. Erucylphospho-N,N,N-trime-
thylpropylammonium (erufosine) is a potential antimyeloma drug
devoid of myelotoxicity. Cancer Chemother Pharmacol 2011;67:
30. Yaseen M, Wang Y, Su TJ, Lu JR. Surface adsorption of zwitter-
ionic surfactants: n-alkyl phosphocholines characterised by surface
tensiometry and neutron reflection. J Colloid Interface Sci.
2005;288:361–70.
31. Rakotomanga M, Loiseau PM, Saint-Pierre-Chazalet M.
Hexadecylphosphocholine interaction with lipid monolayers.
Biochim Biophys Acta-Biomembr. 1661;2004:212–8.
32. Fichtner I, Zeisig R, Naundorf H, Jungmann S, Arndt D, Asongwe
G, et al. Antineoplastic activity of allkylphosphocholines (APC) in
human breast carcinomas in vivo and in vitro - use of liposomes.
Breast Cancer Res Treat. 1994;32:269–79.
33. Kaufmann-Kolle P, Drevs J, Berger MR, Kotting J, Marschner N,
Unger C, et al. Pharmacokinetic behavior and antineoplastic activ-
ity of liposomal hexadecylphosphocholine. Cancer Chemother
Pharmacol. 1994;34:393–8.
34. Zeisig R, Jungmann S, Arndt D, Schutt A, Nissen E. Antineoplastic
activity in vitro of free and liposomal alkylphosphocholines. Anti-
Cancer Drugs. 1993;4:57–64.
1
1
3–25.
1
2
2
9. Bagley RG, Kurtzberg L, Rouleau C, Yao M, Teicher BA.
Erufosine, an alkylphosphocholine, with differential toxicity to hu-
man cancer cells and bone marrow cells. Cancer Chemother
Pharmacol. 2011;68:1537–46.
0. Ríos-Marco P, Marco C, Gálvez X, Jiménez-López JM, Carrasco
MP. Alkylphospholipids: an update on molecular mechanisms and
clinical relevance. Biochim Biophys Acta-Biomembr. 1859;2017:
1
657–67.
1. Wang L, Koynova R, Parikh H, MacDonald RC. Transfection
activity of binary mixtures of cationic O-substituted phosphatidyl-
choline derivatives: the hydrophobic core strongly modulates phys-
ical properties and DNA delivery efficacy. Biophys J. 2006;91:
35. Heerklotz H. Interactions of surfactants with lipid membranes. Q
Rev Biophys. 2008;41:205–64.
3
692–706.
2
2. Opstad CL, Zeeshan M, Zaidi A, Sliwka H-R, Partali V, Nicholson
DG, et al. Novel cationic polyene glycol phospholipids as DNA
transfer reagents-lack of a structure-activity relationship due to un-
controlled self-assembling processes. Chem Phys Lipids. 2014;183:
36. Fleer EAM, Berkovic D, Unger C, Eibl H. Cellular uptake and
metabolic-fate of hexadecylphosphocholine. Prog Exp Tumor
Res. 1992;34:33–46.
37. Berger M, Sobottka S, Konstantinov SM, Eibl H.
Erucylphosphocholine is the prototype of i.v. injectable alkylphos-
phocholines. Drugs today. 1998;34:73–81.
38. Lohmeyer M, Workman P. Growth arrest vs direct cytotoxicity and
the importance of molecular structure for the in vitro antitumor
activity of ether lipids. Brit J Cancer. 1995;72:277–86.
39. Henke G, Lindner LH, Vogeser M, Eibl H-J, Woerner J, Mueller
AC, et al. Pharmacokinetics and biodistribution of Erufosine in nude
mice - implications for combination with radiotherapy. Radiat
Oncol. 2009;4:46.
40. Ríos-Marco P, Marco C, Cueto FJ, Carrasco MP, Jimenez-Lopez
JM. Pleiotropic effects of antitumour alkylphospholipids on choles-
terol transport and metabolism. Exp Cell Res. 2016;340:81–90.
41. Sobottka SB, Berger MR. Assessment of antineoplastic agents by
MTT assay - partial underestimation of antiproliferative properties.
Cancer Chemother Pharmacol. 1992;30:385–93.
1
17–36.
2
2
2
2
2
2
3. Pierrat P, Creusat G, Laverny G, Pons F, Zuber G, Lebeau L. A
cationic phospholipid-detergent conjugate as a new efficient carrier
for siRNA delivery. Chem-Eur J. 2012;18:3835–9.
4. Pierrat P, Kereselidze D, Wehrung P, Zuber G, Pons F, Lebeau L.
Bioresponsive deciduous-charge amphiphiles for liposomal delivery
of DNA and siRNA. Pharm Res. 2013;30:1362–79.
5. Pierrat P, Laverny G, Creusat G, Wehrung P, Strub J-M,
VanDorsselaer A, et al. Phospholipid-detergent conjugates as novel
tools for siRNA delivery. Chem-Eur J. 2013;19:2344–55.
6. Gaillard B, Remy J-S, Pons F, Lebeau L. Cationic erufosine
(
ErPC3) prodrugs as gene delivery reagents for antitumor com-
bined therapy. Chem-Eur J. 2019;25:15662–79.
7. Heyes JA, Niculescu-Duvaz D, Cooper RG, Springer CJ. Synthesis
of novel cationic lipids: effect of structural modification on the effi-
ciency of gene transfer. J Med Chem. 2002;45:99–114.
8. North EJ, Osborne DA, Bridson PK, Baker DL, Parrill AL.
Autotaxin structure–activity relationships revealed through
Publisher’s Note Springer Nature remains neutral with regard to jurisdic-
tional claims in published maps and institutional affiliations.