Optimized synthesis of phosphatidylserine

Article abstract of DOI:10.1007/s00726-009-0447-0

The synthesis of phosphatidyl serine containing saturated fatty acids was thoroughly studied and optimized in order to establish a protocol amenable to large-scale synthesis. The key step was a one-pot multicomponent reaction involving an O-benzyl phosphorodiamidite, protected serine and diacylglycerol, followed by in situ oxidation of the resulting phosphite. In order to replace expensive and poorly stable tetrazole, a screening of substitutes was carried out and imidazolium chloride was selected as the best suited one. Springer-Verlag 2010.

Full text of DOI:10.1007/s00726-009-0447-0

Amino Acids (2010) 39:367–373
DOI 10.1007/s00726-009-0447-0
ORIGINAL ARTICLE
Optimized synthesis of phosphatidylserine
•
Giuseppe Guanti L. Banfi A. Basso L. Bondanza
•
•
•
•
G. Guglieri K. Powles R. Riva
•
Received: 29 July 2008 / Accepted: 9 December 2009 / Published online: 9 February 2010
Ó Springer-Verlag 2010
Abstract The synthesis of phosphatidyl serine containing
saturated fatty acids was thoroughly studied and optimized
in order to establish a protocol amenable to large-scale
synthesis. The key step was a one-pot multicomponent
reaction involving an O-benzyl phosphorodiamidite, pro-
tected serine and diacylglycerol, followed by in situ oxi-
dation of the resulting phosphite. In order to replace
expensive and poorly stable tetrazole, a screening of sub-
stitutes was carried out and imidazolium chloride was
selected as the best suited one.
Thus, treatment with PS was demonstrated to attenuate the
age-related loss of neuronal functions and thus to improve
cognitive functions (Pepeu et al. 1996). Phosphatidyl serine
is also implicated in the blood coagulation process. For this
last effect, it has been shown that both the PS headgroup
per se and unsaturation of the 1,2 fatty acid components are
important (Smirnov et al. 1999).
The main source of PS has been, to date, extraction from
calf brain: a 92% pure phosphatidyl serine can be isolated
from brain ‘‘cephalin’’ by means of solvent fractionation.
This method has currently become problematic due to
increased consumer concern over bovine spongiform
encephalopathy (BSE) (mad cow disease). Although a
partial solution to this problem may be the extraction from
eggs or plant sources (soybean), a fully synthetic procedure
may be advantageous not only from an economical point of
view, but also because of the possibility to obtain com-
pounds containing a single fatty acid type, thus allowing a
better assessment of the differences in biological activity
related to the different fatty acid composition. It should be
noted that the fatty acid content varies greatly among the
various natural sources of PS, and that it is thought to have
a great importance for biological activity (Blokland et al.
1999). Particularly interesting are, from this point of view,
the completely saturated phosphatidylserines, containing
palmitic or stearic acid, because PS extracted from natural
sources is typically characterized by a high percentage of
monounsaturated (soybean) or polyunsaturated (calf brain)
acids.
Keywords Phospholipids ꢀ Multicomponent reactions ꢀ
Phosphorodiamidites
Introduction
Phosphatidyl serine (PS) 1 is one of the most important
natural phospholipids present in the cellular membrane. It
occurs quite widely in nature but usually in concentrations
less than 10% of the cell phospholipid pool. The highest
amount is present in brain tissue, particularly in myelin
(white matter). PS plays an important role in functioning of
the membranes in neurons. Apart from maintenance of the
internal cell’s environment, it has been proved to be
implicated in cell-to-cell communication, cell growth reg-
ulation, secretory vesicle release, and signal transduction.
Although synthetic methods for the synthesis of phos-
pholipids in general and PS in particular were already
known (Baer and Maurukas 1955; Eibl 1984; Hermetter
et al. 1982; Lindh and Stawinski 1989; Martin et al. 1994),
all of them present some practical problems that make
them not well suited for a commercial production. So we
G. Guanti (&) ꢀ L. Banfi ꢀ A. Basso ꢀ L. Bondanza ꢀ
G. Guglieri ꢀ K. Powles ꢀ R. Riva
Dipartimento di Chimica e Chimica Industriale,
`
Universita di Genova, via Dodecaneso 31,
16146 Genoa, Italy
e-mail: guanti@chimica.unige.it
123

368
G. Guanti et al.
wanted to re-evaluate PS synthesis, in order to develop a
robust procedure for a potential large-scale production.
and the base. The best results were obtained at 0°C, in the
presence of 2,6-lutidine, by sequential addition of 2, 3, and
finally excess BnOH. The overall yield of 9 was however
only 25%; moreover, the isolation of 9 from the various
phosphates arising from inclusion of 2 equal alcohols
required a complex chromatography. These and other
preliminary results indicated that a synthesis employing
POCl3, benzyl phosphorodichloridate (Misiura et al. 1983),
or other reagents at the phosphoric oxidation state (Burgos
et al. 1987; van Boeckel et al. 1981, 1985; Xia and Hui
1997a, b; Yamauchi et al. 1986) was not suited for our
purposes.
Results and discussion
The most obvious methodology would be a multicompo-
nent reaction of protected glycerol 2, protected serine 3 and
a suitable phosphorus reagent, to give phosphate 4, which
contains a further protecting group R (Scheme 1). Removal
of the acetal, followed by acylation would then furnish
triprotected PS 5. Finally, deblocking of the masking
groups would give PS 1.
An alternative is represented by the use of derivatives of
phosphorous acid, followed by in situ oxidation of the
resulting phosphite. For example, Martin (Martin and Josey
1988; Martin et al. 1994) has reported a synthesis of pro-
tected PS by using phenyl dichlorophosphite. This method
was not optimal for our purposes for two main reasons: it
involved a phenyl R protection (whereas we preferred a
benzyl group, easier to deprotect), and it required very low
temperatures (–78°C) for the coupling step in order to
achieve selectivity. Although the first problem may be
overcome by using benzyl dichlorophosphite (Bannwarth
and Trzeciak 1987; Claesen et al. 1985; Ogilvie et al. 1980)
instead of phenyl dichlorophosphite, the former was
reported to be rather unstable and even to explode during
distillation (Miller et al. 1971), and thus it was not advis-
able to employ it in a large-scale production.
The crucial step is undoubtedly the multicomponent
reaction leading to the phosphate 4, which contains three
different alcoholic residues, but also the choice of the
protecting groups is important. In previous works, the
phosphate protecting group R was often Ph (Martin and
Josey 1988; Martin et al. 1994). However this option
requires the use of platinum or palladium black for the final
deprotection. Since the use of the less expensive Pd–C was
preferable, we decided to use three protections that may be
cleaved simultaneously by hydrogenation with this cata-
lyst. Therefore we selected Bn and Cbz for the serine and
Bn as R protecting group for the phosphate. While usually
N-Cbz-L-serine benzyl ester 3 had been prepared intro-
ducing first the benzyl ester and then the urethane (Itaya
et al. 1993), we found the benzylation of N-Cbz-L-serine
with benzyl bromide more suitable for large-scale synthesis
(Baer and Maurukas 1955). For this step we employed
a phase-transfer methodology (see the ‘‘Experimental’’
section).
Other coupling reactions that lead to phosphites employ
O-alkyl chlorophosphoroamidites (Bruzik et al. 1986;
Inami et al. 1990; Prestwich et al. 1991) or O-alkyl phos-
phorodiamidites (Bannwarth and Trzeciak 1987; Chen
et al. 1998; Mayer et al. 1999; Reddy et al. 1995; Dreef
et al. 1988) instead of dichlorophosphites as reagents.
These methodologies were originally developed in the
realm of oligonucleotide synthesis, but later they have
proved to be useful also for phospholipid synthesis. Among
the two, the second method seemed to us particularly
appealing, since it involves mild conditions and the same
type of catalyst for the two subsequent steps (typically
tetrazole), therefore allowing in principle an easier one-pot
procedure. The selectivity is ensured by the strong differ-
ence in the rate of substitution of the two amino groups by
alcohol molecules.
In order to couple 2, 3 and benzyl alcohol to give
phosphate 4, we explored various alternative multicom-
ponent reactions. A particularly straightforward possibility
is the sequential addition of these three alcohols to POCl3
in the presence of a base. We tried to optimize this protocol
by changing the order of addition of the three components
O
O
OR
O
O
P
O
OH
O
OBn
O
O
2
O
4
CbzHN
In order to exploit this strategy, we prepared the
O-benzyl phosphorodiamidite 6 following the literature
procedures (Scheme 2). There are two alternative reported
methods for preparing 6: (a) by reaction of diisopropyl-
amine with benzyl dichlorophosphite (Bannwarth and
Trzeciak 1987); (b) by reaction of PCl3 with diisopropyl-
amine to give chloro bis(diisopropylamino)phosphorami-
dite (King and Sundaram 1984) followed by coupling with
benzyl alcohol. This second method can be carried out
HO
OBn
NHCbz
3
O
R4
R4
O
OR
P
O
O
deprotection
O
PS
1
O
OBn
O
O
5
CbzHN
Scheme 1 General plan for the synthesis of phosphatidylserine
123

Optimized synthesis of PS
369
stepwise (Dreef et al. 1988; Dreef-Tromp et al. 1992) or
one-pot (Wakamiya et al. 1996). In our hands all methods
worked well, but we eventually selected the second one
because of the already cited potential problems associated
with benzyl dichlorophosphite. In particular, we preferred
the one-pot method developed by Wakamiya and
coworkers (Wakamiya et al. 1996). Crude 6 obtained in
this way is not pure, being contaminated by substantial
amounts of dibenzyl N,N-diisopropylphosphoramidate and
benzyl N,N-diisopropylphosphoramidic acid, easily detec-
ted by GC–MS. Although we tried to optimize the reaction,
we were not able to completely suppress these side prod-
ucts, whose presence was deleterious for the yield of the
ensuing multicomponent reaction. However, an extractive
workup with n-hexane/acetonitrile (see ‘‘Experimental’’)
allowed us to obtain 98% pure 6 (GC) in 56% yield.
Although some 6 was lost in the acetonitrile washing, we
preferred to have a higher purity, paying the fee for a lower
yield.
Starting from 6 we first studied the stepwise process to
9, reacting it with one of the two alcohols 2 (protected
glycerol) and 3 (protected serine), in the presence of tet-
razole and isolating, through silica gel chromatography, the
intermediate phosphoramidites 7 and 8 in 84 and 91%
yield, respectively. Since phosphites are sensitive to acids,
we always added 0.5–1% Et3N in the eluent. It should be
noted that the synthesis of 7 by this method was already
described by Prestwich et al. (Chen et al. 1998). A second
acid-catalyzed substitution, followed by in situ oxidation
with mCPBA (Chen et al. 1998), was then used for con-
verting both 7 (by reaction with protected serine 3) and 8
(by reaction with protected glycerol 3) into 9. Once again
the overall yields were fully satisfactory (83 and 88%).
However, tetrazole was not well suited for our purposes:
it is a rather expensive compound, and furthermore, it is not
very stable. We therefore carefully studied the first step
(conversion of 6 into 7 or 8), trying to replace tetrazole
with a cheaper and more stable catalyst, reasoning that any
acid with a similar pKa could work similarly well.
Unfortunately this was not true. We tried several alterna-
tive weak acids experiencing yields remarkably lower than
those achieved with tetrazole itself (see Scheme 2). It is
clear that in this reaction tetrazole does not act merely as a
specific acid catalyst, but is involved in a more complex
catalytic process. After a thorough screening we could
obtain acceptable results with pyridinium p-toluenesulfo-
nate (PPTS). However, while the synthesis of 8 worked
nearly as well as with tetrazole, the related preparation of 7
was less satisfactory. Moreover, while transformation of 7
into 9 proceeded with 72% yield, conversion of 8 into 9
afforded an unsatisfactory 51% yield.
iPr
iPr
N
P
a
iPr
Bn
Bn OH
O
N
O
6
iPr
c
OBn
b
O
O
f,g
CbzHN
iPr
iPr
O
N
O
e,g
d,g
N
P
P
iPr
iPr
OBn
OBn
8
7
O
OBn
O
O
P
O
OBn
O
O
9
CbzHN
h
These results indicate that the most critical step appears
to be the introduction of the protected glycerol. We
therefore carried out a second screening of acid catalysts
for the conversion of 8 to 9. The best results were obtained
with imidazolium chloride (78%). It is interesting to note
that imidazolium chloride, as tetrazole, has an additional
pyridine-type nitrogen. Although the overall yield with
imidazolium chloride (78%) was still slightly lower than
that achieved with tetrazole (88%), the low cost and high
stability of imidazolium chloride makes it the catalyst of
choice for this reaction.
O
OBn
P
HO
O
O
O
OBn
O
HO
10
CbzHN
i
C15H31
O
O
O
C15H31
OBn
P
O
O
OBn
O
O
Moreover, imidazolium chloride was also found to be
efficient for the transformation of 6 into 8, affording a yield
similar to that achieved with PPTS and tetrazole. We then
repeated the whole process with this catalyst, this time
without isolation of the intermediates, obtaining compound
9 in 49% (unoptimized) yield after chromatography.
The next step involved deprotection of the isopropyli-
dene group followed by acylation with a fatty acid.
11
CbzHN
Scheme 2 Reagens and conditions: a PCl3, (iPr)2NH, n-hexane; then
add BnOH/Et3N, hexane/CH3CN extraction, 56%. b 2 ? tetrazole,
84% or 2 ? PPTS, 67%. c 3 ? tetrazole, 91% or 3 ? PPTS, 87% or
3 ? imidazolium chloride, 88%. d 2 ? tetrazole or 2 ? PPTS, or
2 ? imidazolium chloride. e 3 ? tetrazole or 3 ? PPTS. f 3 ? im-
idazolium chloride, then 2, 49%. g MCPBA, 0°C. h BF3, MeOH.
i Palmitic acid, DCC, DMAP, CH2Cl2, 37%
123

370
G. Guanti et al.
Throughout this study we always used palmitic acid. one-pot procedure eventually produced 11 in 78% yield
Unfortunately the deblocking step turned out to be rather from 12. Most importantly the crude product could be
troublesome, maybe because of the partial lability of ben- conveniently purified by crystallization, notwithstanding
zyl phosphates. We tried a lot of alternative acid condi- the obvious presence of two diastereoisomers due to the
tions, but the yield was invariably low. The best result that chiral phosphorus atom.
we could achieve was a 37% yield for the two steps from 9
to 11 [deblocking followed by acylation of crude 10 with palmitoyl PS 1a in nearly quantitative yield. This com-
fatty acid and DCC/DMAP (Chen et al. 1996)]. pound is poorly soluble in most solvents. Therefore, in
Finally, hydrogenation over Pd–C afforded pure di-
This disappointing result prompted us to explore an order to separate it from the catalyst we had to carry out the
alternative route, inserting the two fatty acids before the filtration using hot CHCl3/MeOH 9:1.
assembly of the phosphate (Scheme 3). In the meantime,
we had developed an efficient chemoenzymatic synthesis
of 1,2-diacyl-sn-glycerols (DAG) in high e.e., that is Conclusions
amenable to up-scaling, since it does not need chromato-
graphic purifications (Guanti et al. 2004). So we had an In conclusion, we have reported a highly convergent and
easy access to 1,2-dipalmitoyl-sn-glycerol 12 as input for efficient synthesis of PS 1a in 74% yield from DAG 12
our multicomponent phosphate assembly (Scheme 3). through a two-step procedure involving only two crystal-
Taking advantage of the experience gained in the synthesis lizations for the purification of intermediates (no chroma-
of 9, we performed the coupling of 6, 3 and 12 in a one-pot tography). This procedure was used in our laboratory for
fashion, using imidazolium chloride as catalyst. This the synthesis of multigram quantities of PS.
reaction was carefully optimized in order to improve the
The main improvements compared to the previously
yield and to obtain a crude product amenable to crystalli- reported syntheses of PS are: (a) the use of the phospho-
zation. During this optimization we realized that, for var- rodiamidite method, that allows a multicomponent, one-pot
ious reasons, MCPBA was not the best oxidant: it is not procedure at 0°C–r.t.; (b) the use of cheap imidazolium
particularly cheap, nor safe, it requires low temperatures, chloride as catalyst; (c) the use of benzyl protection for the
and the resulting side product (meta-chlorobenzoic acid) phosphate, that allows final hydrogenation under milder
complicates purification of the product. We investigated conditions than the phenyl group; (d) a more convergent
´
several other oxidants (Hebert and Just 1990; Martin route, since the fatty acids are already bound to one of the
and Josey 1988) and found that tert-butyl hydroperoxide building blocks.
(Kozikowski et al. 1994; Martin and Wagman 1996) was
Coupled with the already reported (Guanti et al. 2004)
best suited for this task. After all these optimizations, the efficient enantioselective preparation of 12 via a chemo-
enzymatic route, this process should be very likely ame-
nable to scale-up. Moreover, the introduction of other
saturated fatty acids (e.g., stearic acid) is expected to be
feasible by the same procedure.
C15H31
O
iPr
iPr
O
OBn
OH
O
N
P
C15H31
iPr
Bn
O
+
OH
+
N
O
CbzHN
O
3
6
iPr
12
a
C15H31
O
Experimental
O
O
O
C15H31
OBn
NMR spectra were taken in CDCl3 at 200 MHz (1H), and
50 MHz (13C), using TMS as internal standard for 1H
NMR and the central peak of CDCl3 (at 77.02 ppm) for 13
O
P
O
OBn
O
O
C
11
CbzHN
NMR. Chemical shifts are reported in ppm (d scale),
coupling constants are reported in hertz. Peak assignment
in 13C spectra was made with the aid of DEPT experiments.
GC–MS was carried out on a HP-5971A instrument, using
an HP-1 column (12 m long, 0.2 mm wide), electron
impact at 70 eV, and a mass temperature of about 170°C.
Only m/z [ 33 were detected. All analyses were performed
with a constant He flow of 0.9 ml/min with initial tem-
perature of 100°C, initial time 2 min, rate 20°C/min, final
temperature 260°C, final time 4 min, injector temperature
250°C, detector temperature 280°C. Rt are in min. Melting
b
C15H31
O
O
O
C15H31
O
P
O
O
O
OH
O
O
1a
H3N
Scheme 3 Reagents and conditions: a 6 ? 3, imidazolium chloride,
0°C, then 12, imidazolium chloride, 0°C, then tBuOOH, 0°C, 78%.
b H2, Pd–C, AcOH, 95%
123

Optimized synthesis of PS
371
points were measured on a Bu¨chi 535 apparatus and are
uncorrected. TLC analyses were carried out on silica gel
plates and developed at UV or with ‘‘molybdic’’ reagent
(21 g (NH4)MoO4ꢀ4H2O, 1 g Ce(SO4)2, 469 ml H2O,
31 ml H2SO4), with Dittmer–Lester reagent (Dittmer and
Lester 1964), or with ninhydrin. Rf was measured after an
elution of 7–9 cm. Chromatographies were carried out on
220- to 400-mesh silica gel using the ‘‘flash’’ methodology.
Petroleum ether (40–60°C) is abbreviated as PE. Dry sol-
vents were purchased from Fluka. All reactions employing
dry solvents were carried out under a nitrogen atmosphere.
reunited organic phases were washed with brine, well dried
with Na2SO4 and evaporated to dryness to give an oil. It
was taken up in n-hexane (40 ml) and stirred until a well
solid separates. It was filtered on a Buchner filter and
washed with n-hexane. Finally, the solid was crystallized
from Et2O/PE to give pure 3 (11.03 g, 75%) as a white
solid, m.p. 81.9–82.6°C. Rf = 0.34 (PE/Et2O 3:7). The
other analytical data were identical with those reported
(Itaya et al. 1993).
O-Benzyl O-[(S)-2-(benzyloxycarbonyl)-2-
(benzyloxycarbonylamino)ethyl] O-(1,2-O-dipalmitoyl-
sn-glyceryl) phosphate 11
O-Benzyl N,N,N0,N0-tetraisopropyl phosphorodiamidite
6 (Dreef-Tromp et al. 1992; Wakamiya et al. 1996)
A solution of 3 (2.195 g, 6.66 mmol) in dry CH2Cl2
(30 ml) was treated with a solution of diamidite 6 (2.505 g,
7.40 mmol) in dry CH2Cl2 (15 ml). The solution was
cooled to 0°C and treated with imidazolium chloride
(200 mg, 1.91 mmol). After 30 min at 0°C, the solution
was stirred at r.t. for 21 h. Then dipalmitoylglycerol
(4.128 g, 7.25 mmol) was added. Then imidazolium chlo-
ride (1.957 g, 18.73 mmol) was added in three equal por-
tions, spaced out by 45 min each. After 150 min from the
last addition the mixture was cooled again to 0°C, and
treated with a 3 M solution of tert-butylhydroperoxide in
toluene (3.5 ml, 10.36 mmol). After 2 h at 0°C, the reac-
tion was quenched with 10% NaHSO3 (20 ml) and 5%
NaHCO3 (20 ml). When the mixture was negative at
iodine–starch paper, the phases were separated and the
aqueous one extracted with CH2Cl2 three times. The united
organic extracts were washed with brine, dried (Na2SO4)
and evaporated to dryness to give a yellow foam (8.434 g).
It was crystallized from methanol to give a white solid
(5.431 g, 78%), corresponding to 11 as a diastereoisomeric
mixture due to the chiral phosphorus (m.p., 49.8–54.6°C,
Rf = 0.5 (PE/ACOEt 7:3). Found: C, 69.1; H, 9.2; N, 1.25.
Diisopropylamine (49 ml, 346 mmol) and n-hexane
(140 ml) were placed in a two-necked flask equipped with
a dropping funnel and kept under nitrogen atmosphere. The
flask was cooled to 0°C and treated dropwise (during
15 min) through the dropping funnel, with a solution of
PCl3 (7.87 g, 57.0 mmol) in n-hexane (10 ml). After
washing the funnel with few milliliters of n-hexane, the
temperature was allowed to rise to r.t. After 3 h the milky
suspension was refluxed for 19 h. After cooling again to
0°C, Et3N (24.0 ml, 172.4 mmol) was added followed by
slow addition (15 min) of a solution of Et3N (8.0 ml,
57.4 mmol) and BnOH (6.0 ml, 57.9 mmol) in Et2O
(20 ml). After stirring for 25 min at 0°C and 35 min at r.t.,
the mixture was filtered through a sintered funnel and
evaporated to dryness. The resulting oil was taken up in
n-hexane (100 ml) and washed three times with CH3CN
(80 ? 60 ? 60 ml). The hexane phase was checked at
GC–MS (purity = 98%) and evaporated to dryness to give
6 (10.92 g, 56%) as a slightly yellow liquid. Rf = 0.38 (PE/
AcOEt 1:1). GC–MS: Rt = 7.85. m/z 338 (M?, 0.6), 247
(2.2), 238 (10.3), 121 (2.6), 91 (100), 90 (3.9), 79 (6.1), 70
1
1
(3.4), 65 (3.3), 58 (5.8), 43 (4.9), 42 (5.4), 41 (4.0). H
C60H92NO12P requires: C, 68.61; H, 8.83; N, 1.33%. H
NMR: 7.42–7.20 [5 H, m, aromatics]; 4.65 [2 H, d, CH2Ph,
J = 7.4]; 3.58 [4 H, d of septuplets, CH(CH3)2, Jd = 10.8,
Js = 6.7]; 1.18 [24 H, dd, CH3, J = 2.3, 6.7].
NMR: 7.33 [15 H, s, aromatics]; 5.89 and 5.85 [1 H, 2d,
NH of the two diast., J = 8.4]; 5.25–5.05 [2 H, m, P–O–
CH2Ph]; 5.11 [2 H, s, CH2Ph]; 5.01 and 4.96 [2 H, 2 s,
CH2Ph of two diast.]; 4.65–4.55 [1 H, m]; 4.50–4.35 [1 H,
m]; 4.35–4.18 [2 H, m]; 4.12–3.95 [4 H, m]; 2.27 [4 H, t,
CH2C = O, J = 7.5]; 1.88–1.73 [4 H, m]; 1.25 [48 H, m];
0.88 (mc) [6 H, m, terminal CH3].
N-(Benzyloxycarbonyl)-L-serine benzyl ester
N-(benzyloxycarbonyl)-L-serine (10.63 g, 44.4 mmol) was
suspended in CH3CN (50 ml), and treated with 4 M
aqueous NaOH (11.175 ml, 44.7 mmol). The resulting
solution was treated with n-Bu4NHSO4 (1.52 g,
4.48 mmol) and BnBr (6.40 ml, 53.9 mmol) and refluxed
for 4 h. After cooling to r.t. (a white solid separates), 5%
aqueous NaHCO3 (55 ml) was added and the suspension
stirred for 20 min. After removal of the solid through fil-
tration, the two phases were separated and the aqueous one
extracted three times with AcOEt (40 ml each). The
1,2-O-Dipalmitoyl-sn-glycero-3-phosphoserine 1
Compound 11 (1.939 g, 1.835 mmol) is dissolved in AcOH
(40 ml) and hydrogenated over 10% Pd–C (90 mg) for
48 h at r.t. The mixture was filter through a warm Celite
cake prepared in a sintered funnel with pores = 10–20 lm.
The use of a low pore size sintered funnel was necessary in
order to avoid passage of the catalyst. The cake was
123

372
G. Guanti et al.
Inami K, Teshima T, Emura J, Shiba T (1990) Synthesis of
lysophosphatidylserine with 19:4 acyl group, as
thoroughly washed with hot CHCl3–MeOH 9:1 (160 ml
overall). The filtrate was evaporated to dryness, removing
completely AcOH azeotropically with the aid of n-heptane.
The resulting white solid was transferred into a sintered
funnel and washed with acetic acid, acetone, water, acetone
and finally with few Et2O. After drying, 1.29 g of 1 was
obtained (95%). Rf = 0.43 (CHCl3/CH3OH/25% NH3 50/
25/6, developed with ninhydrin). M.p.: 159–160°C (dec.).
Found: C, 61.1; H, 9.4; N, 1.7. C38H74NO10P requires: C,
62.01; H, 10.13; N, 1.90%. The other spectroscopic data
a novel
sodium-potassium ATPase inhibitor, in relation to DLIS-2, an
endogenous digoxin-like substance. Tetrahedron Lett 31:4033–
4036
Itaya T, Iida T, Shimizu S, Mizutani A, Morisue M et al (1993) Wittig
reaction with N-protected 3-(triphenylphosphonio)alaninates:
synthesis of optically active (E)-(2-arylvinyl)glycine derivatives.
Chem Pharm Bull 41:252–261
King RB, Sundaram PM (1984) Bis(dialkylamino)phosphines. J Org
Chem 49:1784–1789
Kozikowski AP, Powis G, Fauq AH, Tu¨ckmantel W, Gallegos A
(1994) Synthesis and biological activity of the D-3-deoxy-3-
fluoro and D-3-chloro-3-deoxy analogs of phosphatidylinositol.
J Org Chem 59:963–971
1
were coincident with those reported (Eibl 1984). At H or
13C NMR in CDCl3–MeOH the signals were rather broad.
Lindh I, Stawinski J (1989) A general method for the synthesis of
glycerophospholipids and their analogs via H-phosphonate
intermediates. J Org Chem 54:1338–1342
Acknowledgments We gratefully thank Euticals for financing this
project, and Prof. Carlo Scolastico for many helpful advices.
Martin SF, Josey JA (1988) A general protocol for the preparation of
phospholipids via phosphite coupling. Tetrahedron Lett
29:3631–3634
Martin SF, Wagman AS (1996) Synthesis and kinetic evaluation of
inhibitors of the phosphatidylinositol-specific phospholipase c
from Bacillus cereus. J Org Chem 61:8016–8023
Martin SF, Josey JA, Wong Y-L, Dean DW (1994) General method
for the synthesis of phospholipid derivatives of 1, 2-o-diacyl-sn-
glycerols. J Org Chem 59:4805–4820
Mayer TG, Weingart R, Mu¨nstermann F, Kawada T, Kurzchalia T
et al (1999) Synthesis of labeled glycosyl phosphatidyl inositol
(gpi) anchors. Eur J Org Chem, pp 2563–2571
Miller PS, Fang KN, Kondo NS, Ts’O POP (1971) Conformation and
interaction of dinucleoside mono- and diphosphates, V. Synthe-
ses and properties of adenine and thymine nucleoside alkyl
phosphotriesters, the neutral analogs of dinucleoside monophos-
phates. J Am Chem Soc 93:6657–6665
Misiura K, Okruszek A, Pankiewicz K, Stec WJ, Czownicki Z et al
(1983) Stereospecific synthesis of chiral metabolites of ifosfamide
and their determination in the urine. J Med Chem 26:674–679
Ogilvie KK, Theriault NY, Seifert J-M, Pon RT, Nemer MJ (1980)
The chemical synthesis of oligoribonucleotides. IX. A compar-
ison of protecting groups in the dichloridite procedure. Can
J Chem 58:2686–2693
References
Baer E, Maurukas J (1955) Phosphatidyl serine. J Biol Chem 212:25–
38
Bannwarth W, Trzeciak A (1987) A simple and effective chemical
phosphorylation procedure for biomolecules. Helv Chim Acta
70:175–186
Blokland A, Honig W, Brouns F, Jolles J (1999) Cognition-enhancing
properties of subchronic phosphatidylserine (PS) treatment in
middle-aged rats: comparison of bovine cortex PS with egg PS
and soybean PS. Nutrition 15:778–783
Bruzik KS, Salamonczyk G, Stec WJ (1986) A general method for the
synthesis of glycerophospholipids. J Org Chem 51:2368–2370
Burgos CE, Ayer DE, Johnson RA (1987) A new, asymmetric
synthesis of lipids and phospholipids. J Org Chem 52:4973–4977
Chen J, Profit AA, Prestwich GD (1996) Synthesis of photoactivat-
able 1,2-o-diacyl-sn-glycerol derivatives of 1-l-phosphatidyl-D-
myo-inositol 4,5-bisphosphate (ptdinsp2) and 3,4,5-trisphos-
phate (ptdinsp3). J Org Chem 61:6305–6312
Chen J, Feng L, Prestwich GD (1998) Asymmetric total synthesis of
phosphatidylinositol 3-phosphate and 4-phosphate derivatives.
J Org Chem 63:6511–6522
Claesen CAA, Segers RPAM, Tesser GI (1985) Ar(alk)ylsulfonyl
ethyl groups as phosphorus protecting functions. Recl Trav Chim
Pays-Bas 104:119–122
Pepeu G, Marconcini Pepeu I, Amaducci L (1996) A review of
phosphatidylserine pharmacological and clinical effects. Is
phosphatidylserine a drug for the ageing brain? Pharm Res
33:73–80
Dittmer JC, Lester RL (1964) A simple, specific spray for the
detection of phospholipids on thin-layer chromatograms. J Lipid
Res 5:126–127
Prestwich GD, Marecek JF, Mourey RJ, Theibert AB, Ferris CD et al
(1991) Tethered ip3. Synthesis and biochemical applications of
the 1-o-(3-aminopropyl) ester of inositol (1,4,5)-trisphosphate.
J Am Chem Soc 113:1822–1825
Reddy KK, Saady M, Falck JR, Withed G (1995) Intracellular
mediators: synthesis of l-alpha-phosphatidyl-d-myo-inositol
3,4,5-trisphosphate and glyceryl ether analogs. J Org Chem
60:3385–3390
Smirnov MD, Ford DA, Esmon CT, Esmon NL (1999) The effect of
membrane composition on the hemostatic balance. Biochemistry
38:3591–3598
van Boeckel CAA, Oltvoort JJ, van Boom JH (1981) Phosphotriester
approach to the synthesis of a-glucosylated mono- and diphos-
phatidyl glycerols: bacterial cell-wall components. Tetrahedron
37:3751–3761
Dreef CE, Elie CJJ, Hoogerhout P, van der Marel GA, van Boom JH
(1988) Synthesis of 1-o-(1, 2-di-o-palmitoyl-sn-glycero-3-phos-
pho)-D-myo-inositol 4,5-bisphosphate: an analogue of naturally
occurring (ptd)ins(4, 5)p2. Tetrahedron Lett 29:6513–6516
Dreef-Tromp CM, Lefeber AWM, van der Marel GA, van Boom JH
(1992) Synthesis and phosphorylating properties of hydroxya-
mino acid phosphoramidites. Synthesis, pp 1269–1272
Eibl H (1984) Phospholipids as functional constituents of biomem-
branes. Angew Chem Int Ed Engl 23:257–271
Guanti G, Basso A, Banfi L, Bevilacqua E, Bondanza L et al (2004)
Efficient chemoenzymatic enantioselective synthesis of diacyl-
glycerols (dag). Tetrahedron Asymmetry 15:2889–2892
´
Hebert N, Just G (1990) Synthesis of phospholipids using an inverse
phosphite triester approach. J Chem Soc Chem Commun, pp
1497–1498
Hermetter A, Paltauf F, Hauser H (1982) Synthesis of diacyl and
alkylacyl glycerophosphoserines. Chem Phys Lipids 30:35–45
van Boeckel CAA, Visser GM, van Boom JH (1985) Synthesis of
phosphatidyl-b-glucosyl glycerol containing a dioleoyl diglyc-
eride moiety: application of the tetraisopropyldisiloxane-1,3-diyl
(tips) protecting group in sugar chemistry. Part IV. Tetrahedron
41:4557–4565
123

Optimized synthesis of PS
373
Wakamiya T, Nishida T, Togashi R, Saruta K, Yasuoka J et al (1996)
Preparations of N-alpha-fmoc-o-[(benzyloxy)hydroxyphosphi-
nyl] beta-hydroxy alpha-amino acid derivatives. Bull Chem
Soc Jpn 69:465–468
Xia J, Hui Y-Z (1997a) The efficient synthesis of mixed diacyl
phospholipids with polyunsaturated fatty acid in sn-2 position of
glycerol. Tetrahedron Asymmetry 8:3019–3021
Xia J, Hui Y-Z (1997b) The stereospecific synthesis of mixed-acid
phospholipids with polyunsaturated fatty acid from D-mannitol.
Tetrahedron Asymmetry 8:451–458
Yamauchi K, Une F, Tabata S, Kinoshita M (1986) Synthesis of
glycerophospholipids containing aminoalkylphosphonic acids.
J Chem Soc Perkin Trans 1:765–770
123