The chemical synthesis of phosphorothioate and phosphorodithioate analogues of lysophosphatidic acid (LPA) and cyclic phosphatidic acid (CPA)

Article abstract of DOI:10.1039/b9nj00704k

The chemical synthesis of new sulfur analogues of lysophospholipids has been described, including phosphorothioate/phosphorodithioate derivatives of lysophosphatidic acids (LPA) and phosphorothioate/phosphorodithioate derivatives of cyclic phosphatidic acids (cPA). For the preparation of LPA and cPA derivatives both oxathiaphospholane and dithiaphospholane approaches have been employed. Each lysophospholipid analogue has been synthesized as a series of five compounds, bearing five different fatty acid residues, both saturated (12:0, 14:0, 16:0, 18:0) and unsaturated (18:1), in the form of ammonium salts. The phosphorodithioate analogues of LPA were obtained as triethylammonium salts, however these were not stable and decomposed when transformed into the ammonium salt by ion exchange in aqueous methanol solution. The new sulfur analogues of LPA and cPA may share interesting biological properties of their parent compounds, and previously synthesized derivatives may behave as regulators of many metabolic processes and hopefully show new biological activity.

Full text of DOI:10.1039/b9nj00704k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
The chemical synthesis of phosphorothioate and phosphorodithioate
analogues of lysophosphatidic acid (LPA) and cyclic phosphatidic
acid (CPA)w
Przemys$aw Rytczak, Maria Kozio$kiewicz and Andrzej Okruszek*
Received (in Montpellier, France) 24th November 2009, Accepted 16th February 2010
First published as an Advance Article on the web 26th March 2010
DOI: 10.1039/b9nj00704k
The chemical synthesis of new sulfur analogues of lysophospholipids has been described,
including phosphorothioate/phosphorodithioate derivatives of lysophosphatidic acids (LPA) and
phosphorothioate/phosphorodithioate derivatives of cyclic phosphatidic acids (cPA). For the
preparation of LPA and cPA derivatives both oxathiaphospholane and dithiaphospholane
approaches have been employed. Each lysophospholipid analogue has been synthesized as a series
of five compounds, bearing five different fatty acid residues, both saturated (12 : 0, 14 : 0, 16 : 0,
18 : 0) and unsaturated (18 : 1), in the form of ammonium salts. The phosphorodithioate analogues
of LPA were obtained as triethylammonium salts, however these were not stable and decomposed
when transformed into the ammonium salt by ion exchange in aqueous methanol solution. The
new sulfur analogues of LPA and cPA may share interesting biological properties of their parent
compounds, and previously synthesized derivatives may behave as regulators of many metabolic
processes and hopefully show new biological activity.
human blood and its components, the highest abundance was
Introduction
found for the linoleoyl-(18 : 2)-derivative among the LPA
compounds4b and for the oleoyl-(18 : 1)-derivative in the cPA
Both LPA (1-acyl-sn-glycerol-3-phosphate) (1) and cPA
series.4d
(1-acyl-sn-glycerol-2,3-cyclic phosphate) (2) belong to the
family of lysophospholipids,1,2 containing one acyl group
LPA is produced extracellularly from lysophosphatidyl-
(fatty acid residue) and a phosphate moiety (either acyclic or
choline (LPC) through hydrolytic action of lysophospholipase
cyclic) as depicted in Fig. 1.
D (lysoPLD), identical to autotaxin (ATX) or nucleotide
pyrophosphatase/phosphodiesterase (NPP2),4c and degraded
Although LPA and cPA are minor lipid species present in
human blood and other biological media, they are involved in
by lipid phosphate phosphatases (LPPs) to monoacyloglycerol
(MAG).5 Lysophosphatidylcholine is also a substrate for
numerous intracellular processes and are important inter-
mediates in lipid biosynthesis.3 From the structural point of
biosynthesis of cPA in blood.2 This lipid is formed by intra-
view, LPA and cPA bear a polar phosphate head group and a
molecular transphosphatidylation of LPC catalyzed by
single hydrophobic chain. In fact, both species are present in
lysoPLD-like enzymes, although the mechanism of this
various biological fluids as a mixture of several compounds
enzyme, which directs the degradation of LPC to LPA or
cPA, is still obscure.6
containing different fatty acid residues, either saturated
(14 : 0, 16: 0, 18 : 0) or unsaturated (16 : 1, 18 : 1, 18: 2, 20 : 4).4
In spite of its simple structure, LPA is regarded as an
Following the mass spectrometry measurements performed on
important signalling molecule with a wide range of biological
activities. For example, LPA elicits cell proliferation and
survival, platelet aggregation, smooth-muscle contraction,
neurite retraction and cytokine production.5 Elevated levels
of LPA promote cell migration and invasion and this
compound in general plays an important role in cancer
progression7 and angiogenesis.8 It has also been found that
in experimental animals LPA promotes wound healing.9
Fig. 1 Structures of natural LPA (1) and cPA (2). Substituents R–C(O)–
Although physiological concentration of cPA in human
blood is significantly (ca. 10 times) lower than that of LPA,4d
its biological activity is also remarkable.2 Despite structural
similarity, some of its biological properties are different or
even opposite to those of LPA.2 For instance cPA, in contrast
to LPA, inhibits cell proliferation10 and tumor cell invasion,11
thus showing anti-metastatic properties.
refer to various fatty acid residues present in lysophospholipids.
Institute of Technical Biochemistry, Faculty of Biotechnology and
´
Food Sciences, Technical University of Łodz, Stefanowskiego 4/10,
´
´
´
90-924 Łodz, Poland. E-mail: andrzej.okruszek@p.lodz.pl;
Fax: +48 42 636 66 18; Tel: +48 42 631 34 45
w This article is part of a themed issue on Biophosphates, and is
´
dedicated to Professor Wojciech J. Stec, Łodz´ , Poland, on the occasion
of his 70th birthday.
Both LPA and cPA mediate their cellular responses through
the activation of specific lysophosphatidic acid receptors
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1008 | New J. Chem., 2010, 34, 1008–1017

View Article Online
coupled to G proteins.12 Thus, such receptors are being
considered as potential drug targets. Taking into account
that either agonists or antagonists of lysophosphatidic acid
receptors may have therapeutic potential, the synthesis and
structure–activity studies of numerous analogues of LPA and
cPA were undertaken. These derivatives of lysophosphatidic
acids are often described as ‘‘metabolically stabilized analogues’’
to stress the fact that labile structural elements of naturally
occurring compounds are chemically stabilized.13 Some of
these compounds also act as inhibitors of enzymes involved
in lysophosphatidic acid metabolism.13
Fig. 2 Structures of the LPA and cPA sulfur analogues (phosphoro-
thioates and phosphorodithioates) 3–6.
The most frequently encountered modifications involve
introducing fluorine atoms in different parts of LPA/cPA
molecules, including their phosphonate and/or thiophosphonate
analogues.13,14 Typical modifications of LPA include the
addition of an alkylated oxygen atom in position 2 of glycerol
in order to prevent acyl migration.15 cPA analogues are often
synthesized as so-called ‘‘carba derivatives’’, containing
substituted or unsubstituted phosphonate/thiophosphonate
moieties.16 In the LPA series, 1-oleoyl-2-O-methyl-sn-glycero-
phosphorothioate (OMPT) was synthesized, as a metaboli-
cally stabilized analogue containing both an alkyl-protected
2-oxygen and a phosphorothioate function. This compound
was prepared both in racemic and enantiomeric forms.17 Racemic
OMPT was found to be a potent agonist for the LPA3 G-protein-
coupled receptor.17a Further studies revealed, that (2S)-OMPT is
5- to 20-fold more active than (2R)-OMPT.17b
The structures of synthesized sulfur analogues of LPA (3a–e, 4a–e)
and cPA (5a–e, 6a–e) are shown in Fig. 2.
Needless to say, the pleiotropic biological/therapeutic
activity of LPA and cPA and their hitherto known analogues
kindled hopes that the phosphorothioate/phosphorodithioate
derivatives 3–6 may have interesting biological properties as
well. It should also be added, that for preliminary biological
studies all phosphorothioate/phosphorodithioate compounds
3–6 were synthesized in a racemic form.
The synthesis of 1-acylglycerols
The starting point for the synthesis of the sulfur analogues of
both LPA and cPA were 1-acylglycerols, which were prepared
according to the procedure described by Lok et al.21 Thus,
racemic glycidol (7) was incubated for 5 h at ca. 85 1C with
equimolar amounts of corresponding fatty acids in the
presence of a catalytic quantity of tributylamine (Scheme 1).
Crude 1-acylglycerols (8a–e) were purified by crystallization
from diethyl ether–diisopropyl ether (1 : 1, v/v) to give
products in 42–68% isolated yield. The purity of 8 was
checked by TLC, and their identity confirmed by comparison
It should also be added that other thiolated acyl glycerol
systems were recently described.18 The new lipid analogues
show interesting PLA2 activation properties18a–c and can be
employed in constructing novel gene delivery vehicles.18d
Results and discussion
1
of melting points with literature data and by H NMR. The
The interesting biological properties of LPA and cPA as well
as their known analogues prompted us to undertake studies on
the synthesis of new modifications of lysophosphatidic acids.
We focused our attention on the derivatives of LPA and cPA,
in which either one or two nonbridging phosphate oxygen
atoms were substituted by sulfur to give phosphorothioate
or phosphorodithioate analogues, respectively. From the
previous studies performed on the synthesis of phosphorothioate
and phosphorodithioate derivatives of nucleotides and
oligonucleotides,19 we expected that these lysophospholipid
sulfur analogues should have similar physicochemical properties
as natural lysophosphatidic acids, but should be more resistant
towards hydrolytic enzymes.20 In order to prevent possible acyl
migration in LPA analogues, the oxygen atom in position 2 of
glycerol was methylated. To ensure a representative range of
lysophosphatidic acid analogues, each congener was prepared
as a series of five different compounds, bearing the residues of
the following fatty acids: (a) lauric (12 : 0), (b) myristic (14 : 0),
(c) palmitic (16 : 0), (d) stearic (18 : 0), (e) oleic (18 : 1). Thus,
sulfur LPA analogues were prepared as phosphorothioates
(3a–e) and phosphorodithioates (4a–e). Similarly, sulfur cPA
analogues were also obtained as phosphorothioates (5a–e) and
phosphorodithioates (6a–e). All aforementioned compounds
were previously unknown, except the phosphorothioate
oleoyl analogue of LPA (3e), which is known as OMPT.17
details are given in the experimental section.
The synthesis of sulfur analogues of LPA
The first step in the functionalization of 1-acylglycerols (8a–e)
was the selective protection of the primary hydroxyl function
in the presence of a secondary one. For this purpose a
tert-butyldimethylsilyl group (TBDMS) was chosen, which
has been successfully employed for selective protection of
primary hydroxyls in the synthesis of prostaglandins,22
nucleosides23 and modified lipids.24 The TBDMS group was
introduced in 42–78% yield by reacting each of 8a–e with
tert-butyldimethylsilyl chloride and triethylamine in the
presence of catalytic amounts of 4-dimethylaminopyridine
(DMAP) (see Scheme 2).
The resulting 3-O-silyl ethers 9a–e were further methylated
at the 2-hydroxyl group with trimethylsilyldiazomethane
Scheme 1 The synthesis of the 1-monoacylglycerols (8a–e). The fatty
acid residues R–C(O)– correspond to those shown in Fig. 2.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1008–1017 | 1009

View Article Online
Scheme 2 The synthesis of phosphorothioate (3a–e) and phosphorodithioate (4a–e) analogues of LPA. The fatty acid residues R–C(O)–
correspond to those shown in Fig. 2.
(TMSCHN2) in the presence of 40% aqueous fluoroboric acid
(HBF4).25 Fortunately, the acidic procedure did not affect the
TBDMS group present in the molecule,24 and fully protected
lipids 10a–e were isolated in 48–92% yield. The TBDMS
group was removed by a standard method, employing
tetra(n-butyl)ammonium fluoride (TBAF) in THF solution22–24
to give 3-OH compounds 11a–e in 64–92% yield. For the
introduction of sulfur-containing phosphate groups, oxathia-
phospholane and dithiaphospholane approaches were used,
previously developed by us for the synthesis of modified
nucleotides.19g,k
were isolated as ammonium salts in 40–48% yield, after
+
ion–exchange on a Amberlysts-NH4 resin in 90% aqueous
methanol solution. It should be added that ammonium salts
were found to be more stable than independently prepared
sodium salts of 3a–e (not shown). Physicochemical data of the
synthesized racemic 1-acyloxy-2-O-methyl-sn-glycerol-3-O-
phosphorothioates (ammonium salts) 3a–e are listed in
Table 1. The compounds 3a–e were found to be soluble in
alcohols (methanol, ethanol) and in water. They were isolated
in the form of white solids, and were stable when stored at ꢁ20 1C.
The ammonium phosphorothioates (3a–e) and all inter-
mediate compounds (9a–e; 10a–e; 11a–e; 12a–e, X = O;
13a–e, X = O) were isolated by silica gel flash chromato-
graphy (FC). Their purity and identity was confirmed by TLC,
1H/31P NMR and MALDI TOF MS. The details are given in
the experimental section.
LPA phosphorothioates (3a–e)
Phosphorothioate derivatives of lysophospholipids (3a–e)
were synthesized according to the oxathiaphospholane
approach, and successfully employed by us for the preparation
of nucleoside phosphorothioates.19k Thus, the 3-OH lipids
11a–e were reacted with 2-N,N-diisopropylamino-1,3,2-
oxathiaphospholane in the presence of S-ethylthiotetrazole,
and then with elemental sulfur, to give 3-O-(2-thio-1,3,2-
oxathiaphospholane) lipid derivatives 12a–e (X = O) in
42–65.5% yield (two signals for the stereoisomers were
observed in the 31P NMR spectrum at d ca. 105 ppm). Such
a chemical shift was described earlier as a characteristic
of 2-alkoxy-2-thio-1,3,2-oxathiaphospholane derivatives.19a
Compounds 12a–e (X = O) were treated with an excess of
3-hydroxypropionitrile in the presence of 1,8-diazabicyclo-
[5.4.0]undec-7-ene (DBU) to result, after elimination of
episulfide, in the formation of 3-O-(O-2-cyanoethyl)-
LPA phosphorodithioates (4a–e)
For the synthesis of the phosphorodithioate derivatives of
LPA (4a–e) the dithiaphospholane method was employed,
which was previously introduced by us for the preparation
of nucleoside phosphorodithioates.19g The starting materials
were again 3-OH lipids (11a–e), which were reacted with
2-N,N-diisopropylamino-1,3,2-dithiaphospholane
in
the
presence of S-ethylthiotetrazole, and then with elemental
sulfur, to give 3-(2-thio-1,3,2-dithiaphospholane) lipid deriva-
tives 12a–e (X = S) in 45–51% yield, with a 31P NMR d
ca. 123 ppm. Such a chemical shift value was earlier described
as characteristic for this class of compounds.19b,g The dithia-
phospholane ring was cleaved by the action of an excess of
3-hydroxypropionitrile in the presence of DBU to give, after
elimination of episulfide, 3-O-(O-2-cyanoethyl)-phosphoro-
dithioates 13a–e (X = S). The observed 31P NMR chemical
shifts for 13a–e (ca. 115 ppm) were characteristic for phos-
phorodithioate O,O-diesters,19g however the products isolated
by column chromatography were found to contain ca. 30% of
3-hydroxypropionitrile (1H NMR examination). This impurity
phosphorothioates 13a–e (X
= O) in 61–88% yield
(two signals were observed for the stereoisomers in the 31P
NMR spectrum at d ca. 58 ppm, characteristic of phosphoro-
thioate O,O-diesters). The O-2-cyanoethyl groups were
removed according to the procedure described by Sekine et al.26
by the action of DBU on 13a–e (X = O), after its silylation
with N,O-bis(trimethylsilyl)acetamide (BSA). The final
1-acyloxy-2-O-methyl-sn-glycerol-3-O-phosphorothioates 3a–e
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1010 | New J. Chem., 2010, 34, 1008–1017

View Article Online
Table 1 Physicochemical characteristics of 1-acyloxy-2-O-methyl-sn-glycerol-3-phosphorothioates (ammonium salts) 3a–e
Molecular weight [Da]
Calculateda
Measuredb
Sulfur modified lysophospholipid
d
31P NMR (CD3OD) [ppm]
Lauroyloxy-thio-LPA 3a
Myristoyloxy-thio-LPA 3b
Palmitoyloxy-thio-LPA 3c
Stearoyloxy-thio-LPA 3d
Oleoyloxy-thio-LPA 3e
54.40
53.25
54.65
54.65
53.04
384.46
412.51
440.56
468.62
466.60
383.2
410.8
438.9
467.0
465.3
a
b
Acidic form. MALDI TOF MS (m/z, M+ ꢁ H ions).
was removed by prolonged (48 h) maintaining of samples
under high vacuum and finally 13a–e (X = S) were obtained
in 53–61% yield. The removal of the O-2-cyanothyl group
from 13a–e (X = S) was performed by the BSA/DBU
procedure.26 Analysis of the 31P NMR spectra of the crude
products of reaction showed that their major components had
chemical shifts close to those expected for 4a–e. For the crude
products 31P NMR chemical shifts of ca. 101 ppm (CDCl3)
were observed, which were similar to those observed for the
nucleoside monoester phosphorodithioates (88.6–90.6 ppm,
Et3NH+ salts, D2O solution).19g,27 However, attempts to
isolate the product by silica gel column chromatography with
‘‘normal’’ CHCl3–MeOH elution resulted in its complete
decomposition. Purification of the product could be
performed when 1% of triethylamine was added to the eluting
system. The increased stability of nucleoside phosphorodithioates
at basic conditions, with the strong tendency to decomposition
at lower or even neutral pH, was observed earlier by us and
other authors.19g,27 After silica gel column chromatography
with CHCl3–MeOH elution in the presence of 1% of Et3N,
single products were isolated which were identified by 31P
NMR (d ca. 91 ppm, CDCl3), 1H NMR and MALDI TOF MS
as desired phosphorodithioates 4a–e. Unfortunately, the
decomposition of 4a–e occurred when freshly prepared triethyl-
Scheme 3 The synthesis of phosphorothioate (5a–e) and phosphoro-
dithioate (6a–e) analogues of cPA. The fatty acid residues R–C(O)–
correspond to those shown in Fig. 2.
protected adenosine 50-O-(2-thio-1,3,2-dithiaphospholane)
with potassium tert-butoxide in DMF solution. The general
procedure for our synthesis of sulfur analogues of cPA is
outlined in Scheme 3.
The synthesis of cPA phosphorothioates (5a–e)
Phosphorothioate derivatives of cyclophospholipids (5a–e)
were synthesized by reacting each of 8a–e with 2-N,N-
diisopropylamino-1,3,2-oxathiaphospholane in the presence
of S-ethylthiotetrazole, and then with elemental sulfur, in
CH2Cl2 solution. The crude product was further reacted
without isolation with an excess of DBU and was found to
be a major component, the 31P NMR spectrum (CDCl3) shows
two closely located peaks in the 71.5–74.5 ppm range, in
ca. 1 : 1 ratio. Such 31P NMR chemical shift values were close
to those described for nucleoside cyclic 20,30-O,O-phosphoro-
thioates (74.7–76.6 ppm in D2O),29 thus suggesting, for 5a–e, a
five-membered phosphorothioate structure. It can be assumed,
that the reaction of 8 with the oxathiaphospholane derivative
occurred predominantly at the more reactive primary oxygen
atom, and the addition of strongly basic DBU induced the
nucleophilic attack of the other oxygen atom leading to
the formation of the five-membered ring phosphorothioate.
The presence of the two NMR signals for each product can be
attributed to the presence of two centres of chirality, one at the
C2–carbon and the other at the phosphorus. Each product
5a–e is then a mixture of four stereoisomers constituting two
racemates. Compounds having isomeric relationship showed
different 31P NMR chemical shifts and slightly different
chromatographic mobility. We were able to separate them
by silica gel column chromatography into faster-moving
(FAST) and slower-moving (SLOW) isomers. The purified
and separated isomers of each of 5a–e were isolated as
ammonium salts were passed through the Amberlysts-NH4
+
resin in 90% aqueous methanol solution in order to obtain
ammonium salts suitable for biological testing. The decom-
position products could not be identified either by NMR or by
MS. These observations suggest that the phosphorodithioate
derivatives of LPA (4a–e) can be obtained by using the
procedure outlined in Scheme 2 and the products can be
purified by silica gel chromatography with basic protection.
However, the compounds decompose during ion exchange and
therefore can not be used for biological testing.
The synthesis of sulfur analogues of cPA
The racemic 1-acylglycerols (8a–e), described above as
precursors of modified LPA, were also employed by us for
the chemical synthesis of sulfur analogues of cPA. On the basis
of previous experiences we decided to employ oxathia-
phospholane- and/or dithiaphospholane derivatives of 8a–e
as convenient intermediates for the transfer of sulfur atoms
into lysophospholipid molecules. In fact, the only example of
using this type of compounds for the synthesis of cyclic
organophosphorus derivatives was described by Baraniak
and Stec,28 who obtained a six-membered ring adenosine cyclic
(30–50)-phosphorodithioate by the reaction of an appropriately
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1008–1017 | 1011

View Article Online
Table 2 Physicochemical characteristics of 1-acyloxy-sn-glycerol-2,3-cyclic phosphorothioates (ammonium salts) 5a–e
Molecular weight [Da]
Calculateda
Measuredb
Sulfur modified cyclophospholipid
Yield [%]
Separated isomers
d
31P NMR (CD3OD) [ppm]
Lauroyloxy-thio-cPA 5a
66
—
68
—
70
—
58
—
59
—
FAST
SLOW
FAST
SLOW
FAST
SLOW
FAST
SLOW
FAST
SLOW
72.85
72.67
72.71
72.50
73.65
73.40
72.83
72.64
74.86
73.95
352.42
—
380.47
—
408.52
—
436.58
—
351.6
351.4
379.3
380.0
407.7
407.0
435.4
435.6
433.2
433.2
Myristoyloxy-thio-cPA 5b
Palmitoyloxy-thio-cPA 5c
Stearoyloxy-thio-cPA 5d
Oleoyloxy-thio-cPA 5e
434.56
—
a
b
Acidic form. MALDI TOF MS (m/z, M+ ꢁ H ions).
ammonium salts after ion–exchange on Amberlysts-NH4
resin, which were found to be more stable than the inde-
pendently prepared sodium salts (not shown). Their purity and
identity was confirmed by TLC, 1H/31P NMR and MALDI
TOF MS. The yields and physicochemical characteristics of
the synthesized racemic 1-acyloxy-sn-glycerol-2,3-cyclic phos-
phorothioates (ammonium salts) 5a–e are listed in Table 2.
The details are given in the experimental section. The
compounds 5a–e were found to be soluble in alcohols (methanol,
ethanol) and in water. They were isolated in the form of white
solids, and were stable when stored at ꢁ20 1C.
than the independently prepared sodium salts (not shown).
Their purity and identity was confirmed by TLC, 1H/31P
NMR and MALDI TOF MS. The yields and physicochemical
data of prepared racemic 1-acyloxy-sn-glycerol-2,3-cyclic
phosphorodithioates (ammonium salts) 6a–e are listed in
Table 3.
+
The details are given in the experimental section. The
compounds 6a–e were found to be soluble in alcohols (methanol,
ethanol) and in water. They were isolated in the form of white
solids, and were stable when stored at ꢁ20 1C.
Conclusions
The synthesis of cPA phosphorodithioates (6a–e)
Chemically modified lysophospholipids and cyclophospho-
lipids may share interesting biological properties of their
parent natural compounds, and hitherto synthesized deriva-
tives may be regulators of many metabolic processes and can
therefore be regarded as potential therapeutics. In this paper
the chemical synthesis of compounds of this series was
described, including phosphorothioate/phosphorodithioate
analogues of LPA and phosphorothioate/phosphorodithioate
analogues of cPA. Oxathiaphospholane and dithiaphospholane
derivatives were employed as reactive intermediates. Each
analogue was synthesized as a series of five compounds,
bearing five different fatty acid residues both saturated
(12 : 0, 14 : 0, 16 : 0, 18 : 0) and unsaturated (18 : 1). For
preliminary biological studies, all compounds were prepared
as racemates. Unfortunately, phosphorodithioate analogues
of LPA were not stable and could be isolated only as triethyl-
ammonium salts by column chromatography under basic
conditions.
The synthesis of phosphorodithioate analogues of cPA (6a–e)
was performed in an analogous manner to the corresponding
phosphorothioate derivatives (Scheme 3). Thus, 1-acylglycerols
(8a–e) were reacted with 2-N,N-diisopropylamino-1,3,2-
dithiaphospholane in the presence of S-ethylthiotetrazole,
and then with elemental sulfur. Further reaction of the crude
product with DBU resulted in formation of a mixture,
showing a signal in the 31P NMR spectrum (in CDCl3) at
ca. 130 ppm as the major component. Similar values of 31P
NMR chemical shifts (136.9–138.2 ppm in D2O) were reported
for nucleoside cyclic 20,30-O,O-phosphorodithioates,30 suggesting
the five-membered phosphorodithioate structure for 6a–e. The
reaction of 8 with the dithiaphospholane derivative most
probably occurred at the more reactive primary oxygen atom.
Further reaction with strongly basic DBU induced cyclization
of the intermediate 3-O-(2-thio-1,3,2-dithiaphospholane)
compound. The chromatographically purified products 6a–e
were isolated as ammonium salts after ion-exchange on
Due to the presence of the additional centre of chirality
at phosphorus, phosphorothioate analogues of cPA were
Amberlysts-NH4 resin, which appeared to be more stable
+
Table 3 Physicochemical characteristics of the 1-acyloxy-sn-glycerol-2,3-cyclic phosphorodithioates (ammonium salts) 6a–e
Molecular weight [Da]
Calculateda
Measured)b
Sulfur modified cyclophospholipid
Yield [%]
d
31P NMR (CD3OD) [ppm]
Lauroyloxy-dithio-cPA 6a
Myristoyloxy-dithio-cPA 6b
Palmitoyloxy-dithio-cPA 6c
Stearoyloxy-dithio-cPA 6d
Oleoyloxy-dithio-cPA 6e
45
40
47
41
42
133.93
133.89
132.46
132.50
132.40
368.49
396.54
424.59
452.65
450.63
366.9
395.0
423.5
451.0
449.0
a
b
Acidic form. MALDI TOF MS (m/z, M+ ꢁ H ions).
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1012 | New J. Chem., 2010, 34, 1008–1017

View Article Online
obtained in the form of two stereoisomers, which were
separated by silica gel column chromatography.
1-Palmitoylglycerol (8c). Yield 68% (white crystalline solid).
Mp 76–77 1C (lit.,21 mp 76 1C). 1H NMR (CDCl3) d 0.88
(t, 3H, J 6.8 Hz), 1.26 (br s, 24H), 1.63 (m, 2H), 2.06 (br s, 1H),
2.38 (t, 2H, J 7.5 Hz), 2.60 (br s, 1H), 3.65 (m, 2H), 3.94
(m, 1H), 4.18 (m, 2H).
All synthesized compounds will be tested under in vitro
conditions as potential factors influencing wound healing. The
biological testing of synthesized sulfur analogues of LPA and
cPA would also involve determination of their stability
towards lysophospholipid phosphatases at cell culture
conditions. In preliminary experiments we were able to show
that both phosphate (unmodified) and phosphorothioate
oleoyl cPA were stable after 48 h incubation at 37 1C at
100 mM concentration in thermally inactivated bovine fetal
serum or in human serum (TLC control).
1-Stearoylglycerol (8d). Yield 62% (white crystalline solid).
1
Mp 80–81 1C (lit.,21 mp 80–81 1C). H NMR (CDCl3) d 0.88
(t, 3H, J 6.8 Hz), 1.25 (br s, 28H), 1.63 (m, 2H), 2.10 (br s, 1H),
2.35 (t, 2H, J 7.3 Hz), 2.50 (br s, 1H), 3.65 (m, 2H), 3.93
(m, 1H), 4.18 (m, 2H).
1-Oleoylglycerol (8e). Yield 42% (white crystalline solid).
Mp 35 1C (lit.,21 mp 34–35 1C). 1H NMR (CDCl3) d 0.88
(t, 3H, J 6.8 Hz), 1.30 (m, 20H), 1.60 (br s, 1H), 1.63 (m, 2H),
2.02 (m, 4H), 2.18 (br s, 1H), 2.35 (t, 2H, J 7.8 Hz), 2.49
(br s, 1H), 3.71 (m, 2H), 3.93 (m, 1H), 4.18 (m, 2H), 5.35
(m, 2H).
Experimental
General procedures
The purity of all products and intermediates was controlled by
TLC, performed on Kieselgel 60F254 alufoil plates (Merck)
with iodine or phosphoromolybdic acid detection. Flash
chromatography (FC) was run on silica gel 230–400 mesh
(Merck) and followed by TLC. 1H and 31P NMR spectra were
recorded on a Bruker DPX 250 spectrometer (250.13 MHz for 1H)
and referenced to Me4Si (internal) or 85% H3PO4 (external
standard). 31P NMR spectra were obtained with standard
proton decoupling. MALDI TOF mass spectra were recorded
on a Voyager-Elite spectrometer in negative (M+ ꢁ H) mode
(dihydroxybenzoic acid matrix, nitrogen laser at 337 nm with
310 mJ energy, spectra acquired at ca. 10% above the threshold
level). The experimental procedures will be described in detail
for palmitoyl derivatives. The syntheses with other fatty acid
residues were performed in an analogous way.
The synthesis of sulfur analogues of LPA
The synthesis of sulfur analogues of LPA (phosphorothioates
and phosphorodithioates) was performed starting from racemic
1-acylglycerols (8a–e). Here, the syntheses starting from the
palmitoyl derivative (8c) will be described in detail. The
analogues containing other fatty acid residues were prepared
in exactly the same way and identified by NMR spectroscopy
and MALDI TOF MS.
The synthesis of LPA phosphorothioates (3a–e)
1-O-Palmitoyl-3-O-tert-butyldimethylsilyl-sn-glycerol (9c).
A solution of 1.0 g (6.6 mmol) of TBDMS-Cl in anhydrous
methylene chloride (2.5 cm3) was added dropwise at room
temperature into a solution containing 2.1 g (6.3 mmol) of 8c,
0.67 g (6.6 mmol) of Et3N and 38 mg (0.31 mmol) of DMAP.
Stirring at room temperature was continued for 17 h, and the
resulting solution was diluted with chloroform (50 cm3) and
washed with water (50 cm3). The organic layer was dried with
anhydrous Na2SO4 and evaporated under reduced pressure.
The residue was purified by flash column chromatography
(FC) on silica gel, using chloroform with gradually increasing
content of methanol (from 0 to 2.5%, v/v) as eluent, yielding
Materials
Unless otherwise noted, reagents, solvents and other materials
were obtained from commercial sources (Aldrich, Fluka). 2-N,N-
Diisopropylamino-1,3,2-oxathiaphospholane19a and 2-N,N-
diisopropylamino-1,3,2-dithiaphospholane19b were synthesized
as described by us earlier.
The synthesis of racemic 1-acylglycerols (8a–e) (general procedure)21
1
9c (1.19 g, 42%) as a white solid. H NMR (CDCl3) d 0.066
Into a mixture of corresponding fatty acid (52 mmol) and
tributylamine (0.2 g) was added (dropwise over 20 min, with
stirring at 85 1C) 4 g (54 mmol) of freshly distilled racemic
glycidol. The mixture was maintained at 85 1C for 5 h, cooled,
and the crude product crystallized and recrystallized from a
diisopropyl ether–diethyl ether mixture (ca. 1 : 1, v/v).
(s, 6H), 0.88 (t, 3H, J 6.8 Hz), 0.89 (s, 9H), 1.22 (br s, 24H),
1.59 (m, 2H), 2.30 (t, 2H, J 7.6 Hz), 2.43 (br s, 1H), 3.61
(m, 2H), 3.84 (m, 1H), 4.10 (m, 2H).
1-O-Palmitoyl-2-O-methyl-3-O-tert-butyldimethylsilyl-sn-
glycerol (10c). TMSCHN2 (2.7 cm3 of 2 M hexane solution)
was added at 0 1C to a vigorously stirred mixture of 9c (1.19 g,
2.7 mmol) in dichloromethane (20 cm3) and 40% aqueous
HBF4 (0.4 cm3). Stirring at 0 1C was continued and after
30 min three portions of 2 M TMSCHN2 solution (1.35 cm3
each) were added at 20 min intervals. After addition of last
portion of TMSCHN2, stirring was continued at 0 1C for
30 min and at room temperature for 30 min. Then, 2 cm3 of
10% NaHCO3 solution was added with stirring and the layers
were separated. The organic layer was dried over anhydrous
Na2SO4 and evaporated. The residue was purified by silica gel
FC using chloroform with gradually increasing content of
methanol (from 0 to 2.5%, v/v) as eluent. The methylated
1-Lauroylglycerol (8a). Yield 61% (white crystalline solid).
Mp 62–63 1C (lit.,31 mp 62 1C). 1H NMR (CDCl3) d 0.88
(t, 3H, J 6.8 Hz), 1.26 (br s, 16H), 1.63 (m, 2H), 2.10 (br s, 1H),
2.35 (t, 2H, J 7.6 Hz), 2.52 (br s, 1H), 3.65 (m, 2H), 3.93
(m, 1H), 4.20 (m, 2H).
1-Myristoylglycerol (8b). Yield 62% (white crystalline solid).
Mp 68–69 1C (lit.,31 mp 69 1C). 1H NMR (CDCl3) d 0.88
(t, 3H, J 6.8 Hz), 1.26 (br s, 20H), 1.63 (m, 2H), 2.12 (br s, 1H),
2.35 (t, 2H, J 7.8 Hz), 2.56 (br s, 1H), 3.65 (m, 2H), 3.94
(m, 1H), 4.18 (m, 2H).
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1008–1017 | 1013

View Article Online
product 10c (0.78 g, 60.5%) was isolated as a colourless oil. 1H
NMR (CDCl3) d 0.068 (s, 6H), 0.88 (t, 3H, J 6.8 Hz), 0.90
(s, 9H), 1.26 (br s, 24H), 1.63 (m, 2H), 2.33 (t, 2H, J 7.8 Hz),
3.42 (m, 1H), 3.43 (s, 3H), 3.69 (m, 2H), 4.19 (m, 2H).
J 6.6 Hz), 1.20 (br s, 24H), 1.53 (m, 2H), 2.26 (t, 2H, J 7.4 Hz),
3.37 (s, 3H), 3.58 (m, 1H), 3.92 (m, 2H), 4.14 (m, 2H), 4.80
(br s, 8H); 31P NMR (CD3OD) d 54.65; MALDI TOF MS
(M+ ꢁ H) m/z 438.9 (calculated MW 440.56 Da for the
protonated form).
1-O-Palmitoyl-2-O-methyl-sn-glycerol (11c). TBAF (2.27 g,
9.6 mmol) was added to a solution of 10c (0.78 g, 1.63 mmol)
in anhydrous THF (25 cm3) and the mixture was stirred at
room temperature for 2 h. The solvent was evaporated and the
residue was purified by silica gel FC with chloroform–methanol
gradient (from 1 to 2% MeOH, v/v) as eluent. The product 11c
(0.35 g, 64%) was isolated as a white solid. 1H NMR (CDCl3)
d 0.85 (t, 3H, J 6.9 Hz), 1.21 (br s, 24H), 1.57 (m, 2H), 2.28
(t, 2H, J 7,7 Hz), 2.87 (br s, 1H), 3.43 (s, 3H), 3.45 (m, 1H),
3.60 (m, 2H), 4.19 (m, 2H).
1-O-Lauroyl-2-O-methyl-sn-glycerol-3-O-phosphorothioate (3a).
Prepared exactly as for 3c. White solid (ammonium salt).
1H NMR (CD3OD) d 0.88 (t, 3H, J 6.9 Hz), 1.25 (br s,
16H), 1.60 (m, 2H), 2.33 (t, 2H, J 7.8 Hz), 3.46 (s, 3H), 3.68
(m, 1H), 3.98 (m, 2H), 4.22 (m, 2H), 5.01 (br s, 8H); 31P NMR
(CD3OD) d 54.40; MALDI TOF MS (M+ ꢁ H) m/z 438.9
(calculated MW 440.56 Da for the protonated form).
1-O-Myristoyl-2-O-methyl-sn-glycerol-3-O-phosphorothioate
(3b). Prepared exactly as 3c. White solid (ammonium salt).
1H NMR (CD3OD) d 0.90 (t, 3H, J 6.8 Hz), 1.30 (br s, 20H),
1.61 (m, 2H), 2.35 (t, 2H, J 7.7 Hz), 3.46 (s, 3H), 3.68 (m, 1H),
4.00 (m, 2H), 4.23 (m, 2H), 4.89 (br s, 8H); 31P NMR
(CD3OD) d 53.25; MALDI TOF MS (M+ ꢁ H) m/z 410.8
(calculated MW 412.51 Da for the protonated form).
1-O-Palmitoyl-2-O-methyl-sn-glycerol-3-O-(2-thio-1,3,2-oxa-
thiaphospholane) (12c, X = O). 2-N,N-Diisopropylamino-
1,3,2-oxathiaphospholane19a (0.36 g, 1.04 mmol) was added
dropwise, with stirring at room temperature, to a solution of
11c (0.35 g, 1.04 mmol) and S-ethylthiotetrazole (ChemGenes)
(0.135 g, 1.04 mmol) in anhydrous CH2Cl2 (8 cm3). After
stirring for 2 h at room temperature, 100 mg of dry elemental
sulfur was added and stirring was continued overnight. The
reaction mixture was filtered, the solvent was evaporated and
the residue was purified by silica gel FC with a chloroform–
methanol gradient (from 1 to 2% MeOH, v/v) as eluent. The
product 12c (X = O) (0.29 g, 59%) was isolated as a pale-
1-O-Stearoyl-2-O-methyl-sn-glycerol-3-O-phosphorothioate (3d).
Prepared exactly as 3c. White solid (ammonium salt).
1H NMR (CD3OD) d 0.81 (t, 3H, J 6.8 Hz), 1.20 (br s,
28H), 1.52 (m, 2H), 2.26 (t, 2H, J 7.8 Hz), 3.37 (s, 3H), 3.60
(m, 1H), 3.92 (m, 2H), 4.13 (m, 2H), 4.80 (br s, 8H); 31P NMR
(CD3OD) d 54.65; MALDI TOF MS (M+ ꢁ H) m/z 467.0
(calculated MW 468.62 Da for the protonated form).
1
yellow oil. H NMR (CDCl3) d 0.86 (t, 3H, J 6.8 Hz), 1.23
1-O-Oleoyl-2-O-methyl-sn-glycerol-3-O-phosphorothioate (3e).
Prepared exactly as 3c. White solid (ammonium salt).
1H NMR (CD3OD) d 0.81 (t, 3H, J 6.8 Hz), 1.30 (m, 20H),
1.61 (m, 2H), 2.02 (m, 4H), 2.35 (t, 2H, J 7.7 Hz), 3.45 (s, 3H),
3.68 (m, 1H), 3.99 (m, 2H), 4.21 (m, 2H), 4.98 (br s, 8H), 5.33
(m, 2H); 31P NMR (CD3OD) d 53.04; MALDI TOF MS
(M+ ꢁ H) m/z 465.3 (calculated MW 466.60 Da for the
protonated form). Lit.17b for (2R)-OMPT: 31P NMR
(CD3OD) d 52.76; MALDI TOF MS (M++Na), m/z 489.
(br s, 24H), 1.60 (m, 2H), 2.23 (t, 2H, J 7.6 Hz), 3.446
(s, 1.5H), 3.449 (s, 1.5H), 3.51 (m, 2H), 3.63 (m, 1H),
4.09–4.28 (m, 4H), 4.34–4.59 (m, 2H); 31P NMR (CDCl3)
d 105.31 (s), 105.42 (s), the peaks were in 1 : 1 ratio.
1-O-Palmitoyl-2-O-methyl-sn-glycerol-3-O-(O-2-cyanoethyl)-
phosphorothioate (13c, X = O). DBU (88.5 mg, 0.58 mmol)
was added dropwise, at room temperature, into a stirred
solution of 12c (X = O) (0.28 g, 0.58 mmol) and 3-hydroxy-
propionitrile (0.21 g, 2.9 mmol) in anhydrous dichloromethane
(8 cm3). The solution was stirred for 1 h, evaporated, and the
residue purified by silica gel FC with a chloroform–methanol
gradient (from 8 to 33% MeOH, v/v) as eluent to give 13c,
The synthesis of LPA phosphorodithioates (4a–e)
1-O-Palmitoyl-2-O-methyl-sn-glycerol-3-O-(2-thio-1,3,2-dithia-
phospholane) (12c, X = S). 2-N,N-Diisopropylamino-1,3,2-
dithiaphospholane19b (315 mg, 0.91 mmol) was added dropwise,
with stirring at room temperature, to a solution of 11c
(320 mg, 0.91 mmol) and S-ethylthiotetrazole (ChemGenes)
(120 mg, 0.91 mmol) in anhydrous CH2Cl2 (7 cm3). After
stirring for 2 h at room temperature, 100 mg of dry elemental
sulfur was added and stirring was continued overnight. The
reaction mixture was filtered, the solvent was evaporated and
the residue was purified by silica gel FC with a chloroform–
methanol gradient (from 1 to 2% MeOH, v/v) as eluent. The
product 12c (X = S) (205 mg, 45%) was isolated as a
pale-yellow oil. 1H NMR (CDCl3) d 0.85 (t, 3H J 6.7 Hz),
1.23 (br s, 24H), 1.61 (m, 2H), 2.32 (t, 2H, J 7.8 Hz), 3.46
(s, 3H), 3.59 (m, 2H), 3.53–3.73 (m, 3H), 4.06–4.30 (m, 4H);
31P NMR (CDCl3) d 123.50; MALDI TOF MS (positive
mode) m/z 521.9 (M++Na) (calculated MW 498.74).
1
X = O (0.216 g, 75%) as a colourless oil. H NMR (CDCl3)
d 0.81 (t, 3H, J 6.9 Hz), 1.19 (br s, 24H), 1.55 (m, 2H), 2.26
(t, 2H, J 7.7 Hz), 2.71 (t, 2H, J 6.0 Hz), 3.37 (s, 3H), 3.40
(m, 2H), 3.58 (m, 1H), 3.83 (m, 2H), 4.11 (m, 2H); 31P NMR
(CDCl3) d 58.01 (s), 58.08 (s), the peaks were in 1 : 1 ratio.
1-O-Palmitoyl-2-O-methyl-sn-glycerol-3-O-phosphorothioate
(3c). Into a solution of 13c, X = O (0.216 g, 0.44 mmol) in
anhydrous dichloromethane (10 cm3) was added BSA (1.42 g,
6.9 mmol) followed by DBU (0.26 g, 1.7 mmol). After 2 h at
room temperature the solution was evaporated and the residue
was purified by silica gel FC with a chloroform–methanol
gradient (from 10 to 100% MeOH, v/v) as eluent. Fractions
containing phosphorothioate monoester (31P NMR (CDCl3)
d 44.61) were combined, evaporated and subjected to ion–
exchange on Amberlysts-NH4+ resin in 90% aqueous methanol
solution, yielding 3c in the form of ammonium salt (0.12 g,
63%) as a white solid. 1H NMR (CD3OD) d 0.81 (t, 3H,
1-O-Palmitoyl-2-O-methyl-sn-glycerol-3-O-(O-2-cyanoethyl)-
phosphorodithioate (13c, X = S). DBU (62.6 mg, 0.41 mmol)
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1014 | New J. Chem., 2010, 34, 1008–1017

View Article Online
was added dropwise, at room temperature, into a stirred
solution of 12c (X = S) (205 mg, 0.41 mmol) and 3-hydroxy-
propionitrile (146 mg, 2.0 mmol) in anhydrous dichloro-
methane (5 cm3). The solution was stirred for 2 h, evaporated,
and the residue purified by silica gel FC with chloroform–
methanol gradient (from 8 to 33% MeOH, v/v) as eluent to
give the product as a colourless oil. Preliminary 1H NMR
investigation (CDCl3) revealed the presence of ca. 30% of
3-hydroxypropionitrile in the sample (two triplets, d 2.51 and
3.72, J 6.2 Hz). The sample was kept for 48 h at high vacuum
(ca. 0.001 mmHg) at room temperature in order to remove
impurity. This treatment gave 0.16 g (59%) of 13c (X = S) in
containing other fatty acid residues were prepared in exactly
the same way.
1-O-Palmitoyl-sn-glycerol-2,3-cyclic phosphorothioate (5c).
2-N,N-Diisopropylamino-1,3,2-oxathiaphospholane19a (0.31 g,
1.5 mmol) was added dropwise, with stirring at room
temperature, to a solution of 11c (0.50 g, 1.5 mmol) and
S-ethylthiotetrazole (ChemGenes) (0.195 g, 1.5 mmol) in
anhydrous CH2Cl2 (10 cm3). After stirring for 2 h at room
temperature, 100 mg of dry elemental sulfur was added and
stirring was continued overnight. Then, DBU (0.52 g, 3.2 mmol)
was added with stirring and the reaction mixture was kept for
2 h at room temperature. The resulting solution was filtered
and evaporated. The 31P NMR spectrum of the residue
(in CDCl3) showed the presence of two major peaks at
d 71.51 and 74.42 ppm. The crude product was carefully
chromatographed on silica gel with a chloroform–methanol
gradient (from 3 to 17% MeOH, v/v) as eluent. This procedure
allowed for complete separation of the stereoisomers of 5c,
having different chromatographic mobility and therefore
named FAST and SLOW. The stereoisomers were separately
subjected to ion–exchange on Amberlysts-NH4+ resin in 90%
aqueous methanol solution giving FAST 5c and SLOW 5c in
the form of ammonium salts as white solids in a total yield of
0.41 g (70%). 1H NMR (CD3OD) FAST d 0.82 (t, 3H,
J 6.7 Hz), 1.26 (br s, 24H), 1.61 (m, 2H), 2.36 (t, 2H,
J 7.7 Hz), 3.97 (m, 1H), 4.15–4.35 (m, 3H), 4.57 (m, 1H);
31P NMR (CD3OD) FAST d 72.89, SLOW d 72.65; MALDI
TOF MS (M+ ꢁ H) FAST m/z 407.7, SLOW m/z 407.0
(calculated MW 408.52 Da for the protonated form).
1
the form of a colourless oil. H NMR (CDCl3) d 0.82 (t, 3H
J 6.8 Hz), 1.19 (br s, 24H), 1.54 (m, 2H), 2.28 (t, 2H, J 7.7 Hz),
3.41 (s, 3H), 3.42–3.67 (m, 5H), 3.57–4.26 (m, 4H); 31P NMR
(CDCl3) d 114.99; MALDI TOF MS (M+ ꢁ H) m/z 508.0
(calculated MW 509.70 Da for the protonated form).
1-O-Palmitoyl-2-O-methyl-sn-glycerol-3-O-phosphorodithioate
(4c). (A) Into a solution of 13c (X = S) (160 mg, 0.24 mmol) in
anhydrous dichloromethane (7 cm3) was added BSA (770 mg,
3.74 mmol) followed by DBU (670 mg, 4.64 mmol). After 31 h
at room temperature the solution was evaporated and the
residue was examined by 31P NMR (in CDCl3) showing
the presence of signal at 101.58 ppm as the major component.
The crude product was chromatographed on silica gel column
with a chloroform–methanol gradient (from 8 to 33% MeOH, v/v)
as eluent. Unfortunately, the 31P NMR analysis of the
products eluted from the column (CDCl3) showed a mixture
of compounds with chemical shifts in 41–88 ppm range as a
result of decomposition of a product 4c.
1-O-Lauroyl-sn-glycerol-2,3-cyclic phosphorothioate (5a).
Prepared exactly as for 5c in the form of two separated
stereoisomers in 66% total yield. White solids (ammonium
salts). 1H NMR (CD3OD) FAST d 0.89 (t, 3H, J 6.9 Hz), 1.28
(br s, 16H), 1.62 (m, 2H), 2.37 (t, 2H, J 7.8 Hz), 3.98 (m, 1H),
4.15–4.35 (m, 3H), 4.58 (m, 1H); 31P NMR (CD3OD) FAST
d 72.85, SLOW d 72.67; MALDI TOF MS (M+ ꢁ H) FAST
m/z 351.6, SLOW m/z 351.4 (calculated MW 352.42 Da for the
protonated form).
(B) The reaction of 13c (X = S) with DBU/BSA was
repeated exactly as in (A). The crude product was chromato-
graphed as above, with 1% addition of triethylamine to the
eluting solvent. The fraction with a mobility similar to that of
corresponding sulfur analogue was collected and examined by
31P NMR, showing the single product with a chemical shift
91.14 ppm (CDCl3), that was identified as triethylammonium
1
salt of 4c (pale-yellow oil, 92 mg, 58%). H NMR (CDCl3)
d 0.79 (t, 3H, J 6.7 Hz), 1.16 (br s, 24H), 1.26 (t, 18H,
J 7.3 Hz), 1.52 (m, 2H), 2.23 (t, 2H, J 7.6 Hz), 3.07 (q, 12H,
J 7.3 Hz), 3.36 (s, 3H), 3.60 (m, 1H), 3.91–4.07 (m, 3H),
4.19–4.27 (m, 1H); 31P NMR (CDCl3) d 91.15; MALDI TOF
MS (M+ ꢁ H) m/z 455.0 (calculated MW 456.64 Da for the
protonated form).
1-O-Myristoyl-sn-glycerol-2,3-cyclic phosphorothioate (5b).
Prepared exactly as for 5c in the form of two separated
stereoisomers in 59% total yield. White solids (ammonium
salts). 1H NMR (CD3OD) FAST d 0.80 (t, 3H, J 6.8 Hz), 1.20
(br s, 20H), 1.52 (m, 2H), 2.26 (t, 2H, J 7.9 Hz), 3.88 (m, 1H),
4.05–4.24 (m, 3H), 4.49 (m, 1H); 31P NMR (CD3OD) FAST
d 72.71, SLOW d 72.50; MALDI TOF MS (M+ ꢁ H) FAST
m/z 379.3, SLOW m/z 380.0 (calculated MW 380.47 Da for the
protonated form).
(C) The triethylammonium salt of 4c (60 mg) was dissolved
in 90% aqueous methanol (20 cm3) and passed through a
+
column filled with Amberlysts-NH4 resin (20 cm3). The
column was washed with 90% aqueous methanol (60 cm3)
and the solvent was evaporated. Unfortunately, 31P NMR
examination of the residue (in CD3OD) revealed the presence
of several decomposition products in the 68–76 ppm range.
The attempts to identify the products of decomposition of 4c
were unsuccessful.
1-O-Stearoyl-sn-glycerol-2,3-cyclic phosphorothioate (5d).
Prepared exactly as for 5c in the form of two separated
stereoisomers in 58% total yield. White solids (ammonium
salts). 1H NMR (CD3OD) FAST d 0.87 (t, 3H, J 6.7 Hz), 1.24
(br s, 20H), 1.60 (m, 2H), 2.34 (t, 2H, J 7.5 Hz), 3.98 (m, 1H),
4.16–4.36 (m, 3H), 4.56 (m, 1H); 31P NMR (CD3OD) FAST
d 72.83, SLOW d 72.64; MALDI TOF MS (M+ ꢁ H) FAST
m/z 435.4, SLOW m/z 435.6 (calculated MW 436.58 Da for the
protonated form).
The synthesis of cPA phosphorothioates (5a–e)
In this paragraph the synthesis starting from palmitoyl
derivative (8c) will be described in detail. The analogues
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1008–1017 | 1015

View Article Online
1-O-Oleoyl-sn-glycerol-2,3-cyclic phosphorothioate (5e).
Prepared exactly as for 5c in the form of two separated
stereoisomers in 59% total yield. White solids (ammonium
salts). 1H NMR (CD3OD) FAST d 0.83 (t, 3H, J 6.7 Hz), 1.24
(m, 20H), 1.55 (m, 2H), 1.97 (m, 4H), 2.31 (t, 2H, J 7.8 Hz),
4.04 (m, 1H), 4.20-4.39 (m, 3H), 4.59 (m, 1H); 31P NMR
(CD3OD) FAST d 74.86, SLOW d 73.95; MALDI TOF MS
(M+ ꢁ H) FAST m/z 433.2, SLOW m/z 433.2 (calculated MW
434.56 Da for the protonated form).
(m, 1H), 4.13–4.39 (m, 3H), 4.56 (m, 1H); 31P NMR (CD3OD)
d 132.50; MALDI TOF MS (M+ ꢁ H) m/z 451.0 (calculated
MW 452.65 Da for the protonated form).
1-O-Oleoyl-sn-glycerol-2,3-cyclic phosphorodithioate (6e).
Prepared exactly as for 6c in 42% yield as a white solid
(ammonium salt). 1H NMR (CD3OD)
d 0.88 (t, 3H,
J 6.8 Hz), 1.30 (m, 20H), 1.56 (m, 2H), 2.01 (m, 4H), 2.36
(t, 2H, J 8.0 Hz), 4.01 (m, 1H), 4.16–4.38 (m, 3H), 4.59
(m, 1H), 5.32 (m, 2H); 31P NMR (CD3OD) d 132.40; MALDI
TOF MS (M+ ꢁ H) m/z 449.0 (calculated MW 450.63 Da for
the protonated form).
The synthesis of cPA phosphorodithioates (6a–e)
In this paragraph the synthesis starting from palmitoyl
derivative (8c) will be described in detail. The analogues
containing other fatty acid residues were prepared in exactly
the same way.
Acknowledgements
This work was supported by a grant (PBZ-MNiSW-07/I/2007)
from the Polish Ministry of Science and Higher Education.
1-O-Palmitoyl-sn-glycerol-2,3-cyclic phosphorodithioate (6c).
2-N,N-Diisopropylamino-1,3,2-dithiaphospholane19b (330 mg,
1.5 mmol) was added dropwise, with stirring at room
temperature, to a solution of 11c (500 mg, 1.5 mmol) and
S-ethylthiotetrazole (ChemGenes) (195 mg, 1.5 mmol) in
anhydrous CH2Cl2 (10 cm3). After stirring for 2 h at room
temperature, 100 mg of dry elemental sulfur was added and
stirring was continued overnight. Then, DBU (0.52 g,
3.2 mmol) was added with stirring and the reaction mixture
was kept for 3 h at room temperature. The resulting solution
was filtered and evaporated. The 31P NMR spectrum of the
residue (in CDCl3) showed the presence of a major peak at
d 130.2 ppm. The crude product was chromatographed by
silica gel FC with a chloroform–methanol gradient (from 3 to
20% MeOH, v/v) as eluent. Fractions containing cyclic
phosphorodithioate were combined, evaporated and subjected
References
1 A. L. Parrill, Biochim. Biophys. Acta, 2008, 1781, 540–546.
2 Y. Fujiwara, Biochim. Biophys. Acta, 2008, 1781, 519–524.
3 E. E. Kooijman, K. M. Carter, E. G. van Laar, V. Chupin,
K. N. J. Burger and B. de Kruijff, Biochemistry, 2005, 44,
17007–17015.
4 (a) T. Kobayashi, R. Tanaka-Ishii, H. Ikezawa and K. Murakami-
Murofushi, Life Sci., 1999, 65, 2185–2191; (b) D. L. Baker,
D. M. Desiderio, D. D. Miller, B. Tolley and G. J. Tigyi, Anal.
Biochem., 2001, 292, 287–295; (c) J. Aoki, A. Inoue and
S. Okudaira, Biochim. Biophys. Acta, 2008, 1781, 513–518;
(d) L. Shan, S. Li, K. Jaffe and L. Davis, J. Chromatogr., B: Anal.
Technol. Biomed. Life Sci., 2008, 862, 161–167.
5 L. A. van Meeteren and W. H. Moolenaar, Prog. Lipid Res., 2007,
46, 145–160.
6 S. Tsuda, S. Okudaira, K. Moriya-Ito, C. Shimamoto, M. Tanaka,
J. Aoki, H. Arai, K. Murakami-Murofushi and T. Kobayashi,
J. Biol. Chem., 2006, 281, 26081–26088.
to ion-exchange on Amberlysts-NH4 resin in 90% aqueous
+
methanol solution, yielding 6c in the form of ammonium salt
(300 mg, 47%) as white solid. 1H NMR (CD3OD) d 0.88
(t, 3H, J 6.9 Hz), 1.27 (br s, 24H), 1.62 (m, 2H), 2.37 (t, 2H,
J 7.7 Hz), 4.01 (m, 1H), 4.16–4.37 (m, 3H), 4.60 (m, 1H);
31P NMR (CD3OD) d 132.46; MALDI TOF MS (M+ ꢁ H) m/z
423.5 (calculated MW 424.59 Da for the protonated form).
7 G. B. Mills and W. H. Moolenaar, Nat. Rev. Cancer, 2003, 3,
582–589.
8 C. M. Rivera-Lopez, A. L. Tucker and K. R. Lynch, Angiogenesis,
2008, 11, 301–310.
9 A. Sturm and A. U. Dignass, Biochim. Biophys. Acta, 2002, 1582,
282–288.
10 D. J. Fischer, K. Liliom, Z. Guo, N. Nusser, T. Virag,
K. Murakami-Murofushi, S. Kobayashi, J. R. Erickson, G. Sun,
D. D. Miller and G. Tigyi, Mol. Pharmacol., 1998, 54, 979–988.
11 R. Ishihara, M. Tatsuta, H. Iishi, M. Baba, N. Uedo,
K. Higashino, M. Mukai, S. Ishiguro, S. Kobayashi and
K. Murakami-Murofushi, Int. J. Cancer, 2004, 110, 188–193.
12 (a) D. J. Fischer, K. Murakami-Murofushi and G. Tigyi, Ann. N.
Y. Acad. Sci., 2000, 905, 287–289; (b) K. Noguchi, S. Ishii and
T. Shimizu, J. Biol. Chem., 2003, 278, 25600–25606.
1-O-Lauroyl-sn-glycerol-2,3-cyclic phosphorodithioate (6a).
Prepared exactly as for 6c in 45% yield as a white solid
1
(ammonium salt). H NMR (CD3OD) d 0.89 (t, 3H, J 6.8 Hz),
1.26 (br s, 16H), 1.61 (m, 2H), 2.37 (t, 2H, J 7.6 Hz), 4.01
(m, 1H), 4.16-4.38 (m, 3H), 4.59 (m, 1H); 31P NMR (CD3OD)
d 133.93; MALDI TOF MS (M+ ꢁ H) m/z 366.9 (calculated
MW 368.49 Da for the protonated form).
13 G. D. Prestwich, Y. Xu, L. Qian and G. Jiang, Biochem. Soc.
Trans., 2005, 33, 1357–1361.
14 Y. Xu, J. Aoki, K. Shimizu, M. Umezu-Goto, K. Hama,
Y. Takanezawa, S. Yu, G. B. Mills, H. Arai, L. Qian and
G. D. Prestwich, J. Med. Chem., 2005, 48, 3319–3327.
15 L. Qian, Y. Xu, H. Arai, J. Aoki, T. M. McIntyre and
G. D. Prestwich, Org. Lett., 2003, 5, 4685–4688.
16 (a) D. L. Baker, Y. Fujiwara, K. R. Pigg, R. Tsukahara,
S. Kobayashi, H. Murofushi, A. Uchiyama, K. Murakami-
Murofushi, E. Koh, R. W. Bandle, H. S. Byun, R. Bittman,
D. Fan, M. Murph, G. B. Mills and G. Tigyi, J. Biol. Chem.,
2006, 281, 22786–22793; (b) A. Uchiyama, M. Mukai, Y. Fujiwara,
S. Kobayashi, N. Kawai, H. Murofushi, M. Inoue, S. Enoki,
Y. Tanaka, T. Niki, T. Kobayashi, G. Tigyi and K. Murakami-
Murofushi, Biochim. Biophys. Acta, 2007, 1771, 103–112;
(c) G. D. Prestwich, J. Gajewiak, H. Zhang, X. Xu, G. Yang
and M. Serban, Biophys. Acta, 2008, 1781, 588–594.
1-O-Myristoyl-sn-glycerol-2,3-cyclic phosphorodithioate (6b).
Prepared exactly as for 6c in 40% yield as a white solid
1
(ammonium salt). H NMR (CD3OD) d 0.88 (t, 3H, J 6.9 Hz),
1.31 (br s, 20H), 1.61 (m, 2H), 2.36 (t, 2H, J 7.5 Hz), 4.02
(m, 1H), 4.16–4.39 (m, 3H), 4.61 (m, 1H); 31P NMR (CD3OD)
d 133.88; MALDI TOF MS (M+ ꢁ H) m/z 395.0 (calculated
MW 396.54 Da for the protonated form).
1-O-Stearoyl-sn-glycerol-2,3-cyclic phosphorodithioate (6d).
Prepared exactly as for 6c in 41% yield as a white solid
1
(ammonium salt). H NMR (CD3OD) d 0.86 (t, 3H, J 7.1 Hz),
1.24 (br s, 28H), 1.59 (m, 2H), 2.37 (t, 2H, J 7.3 Hz), 3.98
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1016 | New J. Chem., 2010, 34, 1008–1017

View Article Online
17 (a) Y. Hasegawa, J. R. Erickson, G. J. Goddard, S. Yu, S. Liu,
K. W. Cheng, J. Aoki, R. Jarosz, A. D. Schrier, K. R. Lynch,
G. B. Mills and X. Fang, J. Biol. Chem., 2003, 278, 11962–11969;
(b) L. Qian, Y. Xu, Y. Hasegawa, J. Aoki, G. B. Mills and
G. D. Prestwich, J. Med. Chem., 2003, 46, 5575–5578.
18 (a) S. Ghosh, K. R. K. Easwaran and S. Bhattacharya, Tetra-
hedron Lett., 1996, 37, 5769–5772; (b) S. Bhattacharya, S. Ghosh
and K. R. K. Easwaran, J. Org. Chem., 1998, 63, 9232–9242;
(c) A. Bajaj, P. Kondaiach and S. Bhattacharya, J. Med. Chem.,
2008, 51, 2533–2540; (d) S. Bhattacharya and A. Bajaj, Chem.
Commun., 2009, 4632–4656.
(j) B. Karwowski, A. Okruszek, J. Wengel and W. J. Stec, Bioorg.
Med. Chem. Lett., 2001, 11, 1001–1003; (k) M. Olesiak,
D. Krajewska, E. Wasilewska, D. Korczynski, J. Baraniak,
´
A. Okruszek and W. J. Stec, Synlett, 2002, 967–971; (l) P. Guga,
A. Okruszek and W. J. Stec, Top. Curr. Chem., ed. J. P. Majoral,
Springer Verlag, Berlin-Heidelberg, 2002, 220, pp. 169–200.
20 M. Wojcik, M. Cieslak, W. J. Stec, J. W. Goding and
´ ´
M. Kozio"kiewicz, Oligonucleotides, 2007, 17, 134–145.
21 C. M. Lok, A. P. J. Mank and J. P. Ward, Chem. Phys. Lipids,
1985, 36, 329–334.
22 E. J. Corey and A. Venkateswarlu, J. Am. Chem. Soc., 1972, 94,
6190–6191.
19 (a) W. J. Stec, A. Grajkowski, M. Kozio"kiewicz and B. Uznanski,
´
Nucleic Acids Res., 1991, 19, 5883–5888; (b) A. Okruszek,
M. Olesiak and J. Balzarini, J. Med. Chem., 1994, 37,
3850–3854; (c) A. Okruszek, A. Sierzcha"a, K. L. Fearon and
W. J. Stec, J. Org. Chem., 1995, 60, 6998–7005; (d) W. J. Stec,
´
A. Grajkowski, A. Kobylanska, M. Kozio"kiewicz, K. Misiura,
A. Okruszek, A. Wilk, P. Guga and M. Boczkowska, J. Am. Chem.
Soc., 1995, 117, 12019–12029; (e) J. B"aszczyk, M. W. Wieczorek,
23 K. K. Ogilvie, Can. J. Chem., 1973, 51, 3799–3807.
24 Y. Xu, L. Qian and G. D. Prestwich, Org. Lett., 2003, 5,
2267–2270.
25 T. Aoyama and T. Shiori, Tetrahedron Lett., 1990, 31, 5507–5508.
26 M. Sekine, H. Tsuruoka, S. Iimura and T. Wada, Nat. Prod. Lett.,
1994, 5, 41–46.
27 P. H. Seeberger, E. Yau and M. H. Caruthers, J. Am. Chem. Soc.,
1995, 117, 1472–1478.
A. Okruszek, A. Sierzcha"a, A. Kobylanska and W. J. Stec,
´
J. Chem. Crystallogr., 1996, 26, 33–42; (f) A. Sierzcha"a,
A. Okruszek and W. J. Stec, J. Org. Chem., 1996, 61, 6713–6716;
(g) A. Okruszek, M. Olesiak, D. Krajewska and W. J. Stec, J. Org.
28 J. Baraniak and W. J. Stec, Reviews on Heteroatom Chemistry,
ed. S. Oae, MYU, Tokyo, 1993, vol. 8, pp. 143–164.
29 J. Jankowska, M. Wenska, M. Sobkowski, J. Stawinski and
´
Chem., 1997, 62, 2269–2272; (h) A. Kobylanska, A. Okruszek and
´
A. Kraszewski, Tetrahedron Lett., 2000, 41, 2227–2229.
30 M. Wenska, J. Jankowska, M. Popenda, J. Stawin
A. Kraszewski, Tetrahedron Lett., 2001, 42, 8055–8058.
31 H. Shimada, V. E. Tyler and J. L. McLaughlin, J. Nat. Prod., 1997,
60, 417–418.
W. J. Stec, Nucleosides, Nucleotides Nucleic Acids, 1998, 17,
1977–1982; (i) W. J. Stec, B. Karwowski, M. Boczkowska,
P. Guga, M. Kozio"kiewicz, M. Sochacki, M. W. Wieczorek and
J. B"aszczyk, J. Am. Chem. Soc., 1998, 120, 7156–7167;
´
ski and
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1008–1017 | 1017