Streamlined synthesis of phosphatidylinositol (PI), PI3P, PI3, 5P 2, and deoxygenated analogues as potential biological probes

Article abstract of DOI:10.1021/jo060702s

Highly direct total syntheses of phosphatidylinositol (PI), phosphatidylinositol-3-phosphate (PI3P), phosphatidylinositol-3,5-bisphosphate (PI3,5P₂), and a range of deoxygenated versions are reported. Each synthesis is carried out to deliver th

Full text of DOI:10.1021/jo060702s

Streamlined Synthesis of Phosphatidylinositol (PI), PI3P, PI3,5P2,
and Deoxygenated Analogues as Potential Biological Probes
Yingju Xu, Bianca R. Sculimbrene, and Scott J. Miller*
Department of Chemistry, Boston College, Chestnut Hill, Massachusetts 02467
scott.miller.1@bc.edu
ReceiVed April 2, 2006
Highly direct total syntheses of phosphatidylinositol (PI), phosphatidylinositol-3-phosphate (PI3P),
phosphatidylinositol-3,5-bisphosphate (PI3,5P2), and a range of deoxygenated versions are reported. Each
synthesis is carried out to deliver the target in optically pure form. The key step for each synthesis is a
catalytic asymmetric phosphorylation reaction that affects desymmetrization of an appropriate myo-inositol
precursor. Elaboration to each target compound is then carried out employing a diversity-oriented strategy
from the common precursors. In addition to three natural products, several additional streamlined total
syntheses of deoxygenated PI analogues are reported. These syntheses set the stage for high-precision
biological investigations of polar headgroup/biological target interactions of these membrane-associated
signaling molecules.
Introduction
Phosphatidylinositol (PI) and its various phosphorylated
derivatives are now known to be central players in signal
transduction and are ubiquitous in biochemistry (Figure 1).1
Atomic resolution-level investigations of their structure-activity
relationships would be greatly facilitated by streamlined syn-
theses in each enantiomeric series.2 Furthermore, access to
analogues where individual hydroxyl groups might be modified
or deleted would create new opportunities for chemical biologi-
cal studies.3 Numerous synthetic approaches to these molecules
have appeared.4-7 Many strategies have capitalized on the chiral
pool for both building blocks as well as chiral auxiliaries.
(1) Irvine, R. F.; Schell, M. J. Nature ReV. Mol. Cell. Biol. 2001, 2,
327-338.
(2) Prestwich, G. D. Chem. Biol. 2004, 11, 619-637.
(3) For examples of syntheses of deoxy-PI-type compounds with saturated
side chains, see: (a) Wang, D. S.; Chen, C. S. Bioorg. Med. Chem. 2001,
9, 3165-3172. (b) Hu, Y.; Meuillet, E. J.; Qiao, L.; Berggren, M. M.; Powis,
G.; Kozikowski, A. P. Tetrahedron Lett. 2000, 41, 7415-7418. (c)
Andresen, T. L.; Skytte, D. M.; Madsen, R. Org. Biomol. Chem. 2004, 2,
2951-2957.
(4) For syntheses of PIP-compounds with unsaturated side chains, see:
(a) Kubiak, R. J.; Bruziak, K. S. J. Org. Chem. 2003, 68, 960-968. (b)
Gaffney, P. R. J.; Reese, C. B. J. Chem. Soc., Perkin Trans. 1 2001, 192-
205. (c) Watanabe, Y.; Nakatomi, M. Tetrahedron 1999, 55, 9743-9754.
FIGURE 1. Phosphatidylinositol and representative derivative phos-
phates.
Biomimetic approaches have also inspired elegant approaches
to many members of this class, as well as PI-based probes of
PI-dependent biological events.8
We recently described an alternative synthetic approach based
on asymmetric catalysis. In particular, we reported a new method
10.1021/jo060702s CCC: $33.50 © 2006 American Chemical Society
Published on Web 05/20/2006
J. Org. Chem. 2006, 71, 4919-4928
4919

Xu et al.
SCHEME 1
for catalytic asymmetric phosphorylation of myo-inositol deriva-
tives that led to desymmetrization of starting material.9 We then
developed synthetic methodology such that the product of
asymmetric phosphorylation could be elaborated to certain
inositol phosphates and deoxygenated derivatives.10 These
compounds were utilized in chemical biological studies that
probed the stereospecificity of several inositol monophos-
phatase enzymes. We also used the asymmetric phosphorylation
as a springboard for the synthesis of each enantiomer of
phosphatidylinositol-3-phosphate (PI3P), in each optically pure
form.11 We now report new total syntheses in this family of
complex phospholipids, culminating in total syntheses of
phosphatidylinositol (PI), phosphatidylinositol-3,5-bisphosphate
(PI3,5P2), and several key deoxygenated analogues. A key
element of the present study is efficient Mitsunobu-type
methodology for uniting the lipid portions of these molecules
with appropriate polar headgroups. As in our previous studies,
both the naturally occurring enantiomers of these phosphoi-
nositides and the corresponding enantiomeric series are acces-
sible in enantiopure form, with equivalent efficiency, based on
the choice of catalyst used in the desymmetrization of a key
common intermediate. The streamlined synthesis of these
compounds sets the foundation for new chemical biological
studies of these membrane-associated signaling molecules and
their stereospecific interactions with their various biological
targets.
Results and Discussion
(5) For representative syntheses of PI3P-compounds with saturated side
chains, see: (a) Morisaki, N.; Morita, K.; Nishikawa, A.; Nakatsu, N.; Fukui,
Y.; Hashimoto, Y.; Shirai, R. Tetrahedron 2000, 56, 2603-2614. (b) Falck,
J. R.; Krishna, U. M.; Capdevila, J. H. Bioorg. Med. Chem. Lett. 2000, 10,
1711-1713. (c) Painter, G. F.; Grove, S. J. A.; Gilbert, I. H.; Holmes, A.
B.; Raithby, P. R.; Hill, M. L.; Hawkins, P. T.; Stephens, L. R. J. Chem.
Soc., Perkins Trans. 1 1999, 923-936. (d) Chen, J.; Feng, L.; Prestwich,
G. D. J. Org. Chem. 1998, 63, 6511-6522. (e) Wang, D. S.; Chen, C. S.
J. Org. Chem. 1996, 61, 5905-5910. (f) Bruzik, K. S.; Kubiak, R. J.
Tetrahedron Lett. 1995, 36, 2415-2418.
(6) For representative syntheses of PI3,5P2-compounds with saturated
side chains, see: (a) Falck, J. R.; Krishna, U. M.; Katipally, K. R.; Capderila,
J. H.; Ulug, E. T. Tetrahedron Lett. 2000, 41, 4271-4275. (b) Han, F.;
Hayashi, M.; Watanabe, Y. Chem. Lett. 2003, 32, 724-725. (c) Han, F.;
Hayashi, M.; Watanabe, Y. Eur. J. Org. Chem. 2004, 558-566. (d)
Nishikawa, A.; Saito, S.; Hashimoto, Y.; Koga, K.; Shiri, R. Tetrahedron
Lett. 2001, 42, 9195-9198.
(7) For representative syntheses of PI-compounds, see: Watanabe, Y.;
Kiyosawa, Y.; Hyodo, S.; Hayashi, M. Tetrahedron Lett. 2005, 46, 281-
284.
(8) Xu, Y.; Lee, S. A.; Kutateladze, T. G.; Sprissa, D.; Shisheva, A.;
Prestwich, G. D. J. Am. Chem. Soc. 2006, 128, 885-897.
(9) (a) Sculimbrene, B. R.; Miller, S. J. J. Am. Chem. Soc. 2001, 123,
10125-10126. (b) Sculimbrene, B. R.; Morgan, A. J.; Miller, S. J. J. Am.
Chem. Soc. 2002, 124, 11653-11656. (c) Sculimbrene, B. R.; Morgan, A.
J.; Miller, S. J. Chem. Commun. 2003, 1781-1785.
(10) Morgan, A. J.; Wang, Y. K.; Roberts, M. F.; Miller, S. J. J. Am.
Chem. Soc. 2004, 126, 15370-15371.
(11) Sculimbrene, B. R.; Xu, Y.; Miller, S. J. J. Am. Chem. Soc. 2004,
126, 13182-13183.
The major hurdles associated with short, efficient synthesis
of the PI-signaling molecules are the management of the six
stereogenic hydroxyl groups of myo-inositol and the typically
long step-counts associated with protecting group manipula-
tions.12 To address this challenge generally, we began with the
development of catalytic, enantioselective phosphorylation
catalysts. Specifically, we were able to find complementary
pentapeptide catalysts 1 and 2 that enabled enantiospecific
syntheses of both inositol-1-phosphate (I-1P) and the enantio-
meric inositol-3-phosphate, I-3P (Scheme 1). Each peptide
affected desymmetrization of meso-myo-inositol precursor 3
such that phosphates 4 and ent-4 could be isolated with total
enantiopurity. Each could then be subjected to one-step dis-
solving metal reductive cleavage of the protecting groups such
that highly efficient syntheses of I-1P and I-3P were in hand.
To extend these studies to syntheses of naturally occurring
PIs, we needed to develop a synthetic scheme that was
compatible with the presence of the sensitive arachidonate side
chain that is typical of the C2-ester of the glycerol moiety in
the naturally occurring PI targets. As a result, the original
desymmetrizations that we developed with benzyl groups on
(12) Gani, D. Nature 2001, 414, 703-705.
4920 J. Org. Chem., Vol. 71, No. 13, 2006

Streamlined PI Syntheses
SCHEME 2
SCHEME 3 a
a Key: (a) 7a, Hunig’s base, THF, -78 °C; then 30% H2O2/H2O, 31%; (b) TBSCl, imidazole, DMF, 89%; (c) NaH, BnOH, THF, 84%; (d) HF‚pyridine,
THF, 86%; (e) dicyanoimidazole, toluene/CH2Cl2 (1:1), 7b; then 30% H2O2/H2O; (f) H2, Pd(OH)2/C, t-BuOH/H2O, NaHCO3, 85%.
the 2-, 4-, and 6-positions of the inositol headgroups were not
likely to be compatible with these more involved syntheses. That
is, neither hydrogenolytic nor Birch-type cleavage of protecting
groups would be compatible with arachidonoyl esters. As a
result, we followed the precedent of Bruzik,13 targeting a final,
global deprotection that would be based on trimethylsilyl
bromide dealkylation. The strategy requires that the desymme-
trization of the myo-inositol headgroup be carried out on inositol-
(13) Kubiak, R. J.; Bruzik, K. S. J. Org. Chem. 2003, 68, 960-968.
J. Org. Chem, Vol. 71, No. 13, 2006 4921

Xu et al.
SCHEME 4 a
a Key: (a) TBSCl, imidazole, DMF, 89%; (b) NaH, BnOH, THF 99%; (c) HF‚pyridine, THF, 77%; (d) 4,5-dicyanoimidazole, toluene/CH2Cl2 (1:1), 14;
then 30% H2O2/H2O, 39%; (e) TMSBr (20 equiv)/PhCH3, 70 °C; then NH4OH, 61%.
derived triol 5 (Scheme 2), employing p-methoxybenzyl groups
instead of benzyl protection. Indeed, the peptide-based phos-
phorylation catalysts are equivalently effective in the face of
this minor change to the inositol structure, and each peptide
enables the desymmetrization to be carried out with analogous
efficiency.
PI3P and ent-PI3P Syntheses. In our initial studies to
demonstrate the applicability of these approaches to the natural
products and probes of relevance to phosphatidylinositol-
dependent signaling, we targeted molecules with both the
naturally occurring unsaturated side chains (e.g., arachidonoyl)
and side chains that are saturated. We also wished to demon-
strate syntheses in each enantiomeric series since the power of
asymmetric catalysis translates into access to either optically
pure series.14 Our studies of application of catalytic asymmetric
phosphorylation to phosphoinositide commenced with synthesis
of PI3P, the product of phosphoinositide-3-kinase (PI3K), an
enzyme known to be important in the biochemistry of cell cycle
progression.15
For PI3P with saturated side chains, we were able to convert
either compound 4 or ent-4 to the corresponding target following
a route that was compatible with 2,4,6-tribenzyl protection on
the inositol. It was necessary, however, to convert the phenyl
phosphate ester groups to benzyl phosphate ester groups
following a high-yielding three-step sequence (Scheme 3).
Compound 4 is, in principle, two steps from the target PI3P
with saturated side chains. For example, we showed that we
could convert 4 to compound 9a or 9b following standard
chlorophosphite (7a) or phosphoramidite (7b) coupling meth-
ods.16 However, we were not able to identify conditions that
could efficiently remove the benzyl ether and phenyl phosphate
ester protecting groups in one step, without competitive
degradation of the material. Instead, we adopted a procedure
wherein 4 was temporarily converted to a 3,5-disilylated
analogue (TBSCl, imidazole; 89%). Sequential phosphate phen-
yl-to-benzyl transesterification/desilylation then afforded com-
pound 8 in 72% yield (two steps). Diol 8 could then be subjected
to a site-selective P(III)-coupling reaction to deliver protected
PI3P analogue 10.17 Global hydrogenolysis then delivered ent-
PI3P with the saturated side chains in 85% yield, after ion
exchange. This synthetic plan, despite the phosphate ester
exchange sequence, may represent the most efficient access to
PI3P extant.
It is possible that an even further degree of streamlining may
be achieved if the protecting group swap could be avoided.
In this spirit, we have attempted to develop a catalytic
asymmetric desymmetrization of 3 or 5 using dibenzyl phos-
phochloridate, instead of our standard diphenyl phospho-
chloridate employed in Schemes 1 and 2 (eq 1). Such a
transformation would allow conversion of 3 to 8 in a single
step, rather than four steps. Unfortunately, when the cata-
lytic P(V) chemistry is attempted under the peptide-catalyzed
conditions, dibenzyl phosphochloridate reacts as a carbon-based
SN2-type electrophile as shown in eq 2.18 We are continuing
to study alternatives to the ester swap through different
strategies.
4922 J. Org. Chem., Vol. 71, No. 13, 2006

Streamlined PI Syntheses
The desymmetrization of 5 also enables synthesis of PI3P
with the naturally occurring side chains with unprecedented
efficiency. Compound 6 may be converted to 13 through an
analogous phosphate ester exchange (Scheme 4). Then, com-
pound 13 may be subjected to a site-selective phosphoramidite
coupling with reagent 14 to give 15a in 39% yield. While the
isolated yield is modest, it is a reflection of the difficulty in
removing the minor product derived from phosphoramidite
coupling at the C5-position of the inositol (15b). Since the next
step is the final deprotection, and since final purification of
phosphoinositides is challenging, it is essential that the penul-
timate precursor be exceptionally pure prior to deprotection. In
fact, treatment of 15 with TMSBr delivers ent-PI3P in 61%
isolated yield, after ion exchange, in extremely clean fashion.
PI3,5P2 Synthesis. We next turned our attention to the PI3,-
5P2 targets, once again with an eye to enantiospecific syntheses
of targets with either saturated or unsaturated side chains. PI3,-
5P2 is one of the most recently identified phosphoinositidyl
bisphosphates, and it is implicated in a myriad of membrane-
associated biochemical events.19 As a result of our PI3P
syntheses, we were able to extend these routes to streamlined
syntheses of compounds in the PI3,5P2 series. As shown in
Scheme 5, ent-8 was converted to ent-10 as before (via ent-
7b). Then, the remaining hydroxyl group at C5 of the myo-
inositol ring was subjected to coupling with dibenzyl diisopro-
pylphosphoramidite using standard conditions to give 16 in 86%
yield. Global hydrogenolysis delivers PI3,5P2-DiC8 in the
naturally occurring enantiomeric series in 97% isolated yield.
For the PI3,5P2 analogues with naturally occurring unsaturated
side chains, we also followed the route developed for PI3P. As
shown in Scheme 6, ent-13 could be converted to the C1-
functionalized material ent-15 in analogous yield to those
observed in the PI3P studies. The second phosphorylation
sequence delivered protected PI3,5P2, 17, in 88% isolated yield.
In this case, the complete deprotection was achieved employing
TMSI under carefully optimized conditions. PI3,5P2 was thus
obtained with excellent purity and in 41% isolated yield.
PI Synthesis. We next turned our attention to the synthesis
of phosphatidylinositol in the optically pure form. Although this
compound would seem the simplest of the PI-type natural
products that we set out to synthesize, it is also the one that
requires the development of a different method than those
employed for PI3P or PI3,5P2. Although we considered devel-
oping a catalytic asymmetric delivery of a glycerol-function-
alized P(V) reagent, our experiences with dibenzyl phospho-
chloridate diverted our attention to an approach that would
utilize compounds such as 8 (Scheme 7), wherein the desym-
metrization had already been performed. As such, we examined
phosphotriester synthesis under Mitsunobu-type conditions.20
Accordingly, monohydrolysis of 8 is achieved under basic
SCHEME 5 a
a Key: (a) 4,5-dicyanoimidazole, PhCH3/CH2Cl2 (1:1), ent-7b; then 30% H2O2/H2O, 40%; (b) 4,5-dicyanoimidazole CH2Cl2; dibenzyl diisopropylphos-
phoramidite; then 30% H2O2/H2O, 86%;(c) Pd(OH)2/C, H2, t-BuOH/H2O (5:1), Chelex 100 sodium form; 97%
J. Org. Chem, Vol. 71, No. 13, 2006 4923

Xu et al.
SCHEME 6 a
a Key: (a) 4,5-dicyanoimidazole, toluene/CH2Cl2 (1:1), ent-14; then 30% H2O2/H2O, 42%; (b) 4,5-dicyanoimidazole PhCH3/CH2Cl2(1:1); dibenzyl
diisopropylphosphoramidite; then 30% H2O2/H2O, 88%; (c) TMSI, CDCl3, 41%.
SCHEME 7a
sizing optically pure analogues of the PI-derived compounds.
For example, we have previously found deoxygenated inositol
phosphate compounds to be valuable in studying the stereo-
specificity of IP interactions with the various enzymes that
operate upon them.10 We have therefore prepared analogous
compounds in the PI series based on the combination of catalytic
asymmetric phosphorylation and Mitsunobu-based phosphotri-
ester synthesis. Such compounds, if available, could assist in
elucidation of the key interactions between the various PI-
compounds and their often membrane-associated protein tar-
gets.22
For 3-deoxygenated variants, we have previously demon-
strated that precursor 4 could be monofunctionalized with
phenylthionochloroformate at the 3-position (Scheme 8, 70%).
The remaining free C5-hydroxyl group is then readily converted
to TBS ether 21 (90%). Radical deoxygenation may then be
induced under Barton-type conditions. Phosphate ester exchange
and desilylation then afford compound 22. Mono-deprotection
of dibenzyl phosphate 22, followed by Mitsunobu coupling to
19 results in compound 23 (89% over two steps.) Notably,
compound 23 could be monophosphorylated in analogy to the
other syntheses presented above, if a 5-phosphorylated analogue
is desired. Nevertheless, hydrogenolysis then yields 3-deoxy-
PI with, in this case, saturated C8-fatty acid side chains in 97%
yield.
a Key: (a) LiBr, acetone, reflux; then DOWEX 50 × 2-200; (b) DEAD,
Ph3P, 19, THF 40% over two steps; (c) Pd(OH)2/C, H2, t-BuOH/H2O (5:
1), Chelex 100, sodium form; 98%.
conditions to deliver phosphoric acid 18.21 Coupling to 19 could
then be achieved employing DEAD/PPh3 conditions, delivering
20 in 40% yield from 8 (two steps). Hydrogenolysis of the
protecting groups then allowed isolation of optically pure PI-
DiC8 in 98% yield.
Synthesis of Deoxy-PI Analogues. We next wished to
demonstrate the utility of the Mitsunobu chemistry for synthe-
(14) The sequences described below have indeed been executed equiva-
lently in each enantiomeric series for a number of cases. Details of these
syntheses, including characterization of enantiomers, are included in the
Supporting Information where appropriate. For concise presentation, we
describe in the text the preparation of only one enantiomer of each target.
(15) Vazquez, F.; Sellers, W. R. Biochim. Biophys. Acta 2000, 1470,
M21-M35.
The 3,5-dideoxy variant of PI may also be achieved by a
related route, shown in Scheme 9. In this case, exhaustive
thiocarbonylation is carried out to give bis(thionocarbonate) 24.
Radical deoxygenation followed by transesterification then
(16) Martin, S. F.; Josey, J. A.; Wong, Y. L.; Dean, D. W. J. Org. Chem.
1994, 59, 4805-4820.
(20) (a) Mitsunobu, O. Synthesis 1981, 1-28. (b) Hughes, D. L. Org.
React. 1992, 42, 335-656. (c) Hughes, D. L. Org. Prep. Proc. Int. 1996,
28, 127-164. (d) Mitsunobu, O.; Kato, K.; Kimura, J. J. Am. Chem. Soc.
1969, 91, 6510-6511.
(17) Reddy, K. K.; Saady, M.; Falck, J. R. J. Org. Chem. 1995, 60,
3385-3390.
(18) This reactivity manifold is precedented. See: Atherton, F. R.;
Howard, H. T.; Todd, A. R. J. Chem. Soc. 1948, 1106-1111.
(19) Dove, S. K.; Piper, R. C.; McEwen, R. K.; Yu, J. W.; King, M. C.;
Hughes, D. C.; Thuring, J.; Holmes, A. B.; Cooke, F. T.; Michell, R. H.;
Parker, P. J.; Lemmon, M. A. EMBO J. 2004, 23, 1922-1933.
(21) Mahmoodi, N. O. Phosphorus, Sulfur Silicon 2002, 177, 2887-
2893.
(22) Hurley, J. H.; Grobler, J. A. Curr. Opin. Struct. Biol. 1997, 7, 557-
565.
4924 J. Org. Chem., Vol. 71, No. 13, 2006

Streamlined PI Syntheses
SCHEME 8 a
a Key: (a) phenyl chlorothionoformate, DMAP, pyridine, CH2Cl2; 70%; (b) TBSOTf, 2,6-lutidine, CH2Cl2; 90% (c) AIBN, Bu3SnH, toluene; 70%; (d)
BnOH, NaH, THF; 91%; (e) HF pyridine, THF; 75%; (f) LiBr, acetone, reflux; then DOWEX 50 × 2-200; (g) DEAD, Ph3P, 19, THF; 89% over two steps;
(h) Pd(OH)2/C, H2, t-BuOH/H2O (5:1), Chelex 100, sodium form; 97%.
SCHEME 9 a
a Key: (a) phenyl chlorothionoformate, DMAP, pyridine, CH2Cl2; 70%; (b) AIBN, Bu3SnH, toluene; 60% (c) BnOH, NaH, THF; 67%; (d) LiBr, acetone,
reflux; then DOWEX 50 × 2-200; (e) DEAD, Ph3P, 19, THF 55% over two steps; (f) Pd(OH)2/C, H2, tBuOH/H2O (5:1), Chelex 100, sodium form; 93%.
delivers 25, which may be carried through the hydrolysis-
Mitsunobu sequence to afford 26. As in the previous case, global
hydrogenolysis delivers the final product, 3,5-dideoxy-PI, with
the C8 fatty acid side chains installed in this case. Notably, the
Mitsunobu sequences should be entirely compatible with either
enantiomeric series and a wide variety of side chain functional
types, including unsaturated versions, or variants that are
functionalized with biologically interesting tethers as well.
the high precision study of biological processes wherein the
title compounds play significant roles.
Experimental Section
Protected PI3,5P2-DiC8 (16). To a stirred solution of ent-10
(0.023 g, 0.019 mmol) in CH2Cl2 (4.0 mL) was added dibenzyl
diisopropylphosphoramidite (0.063 mL, 0.19 mmol) followed by
4,5-dicyanoimidazole (0.027 g, 0.23 mmol). The reaction was stirred
at room temperature for 14 h and then cooled to 0 °C. 30% H2O2/
H2O (2 mL) was added, and the reaction was stirred at 0 °C for
another 1 h. The reaction was then quenched with saturated Na2SO3
solution (15 mL) and the mixture extracted with CH2Cl2 (3 × 50
mL). The organic layers were combined, dried over sodium sulfate,
and then concentrated under reduced pressure to afford an oil. The
crude product was purified using silica gel flash chromatography
eluting with a gradient of 0-50% ethyl acetate/hexanes to afford
pure product as a clear oil (0.024 g, 86% yield): 1H NMR (CDCl3,
400 MHz) δ 7.36-7.12 (m, 36H), 7.00 (m, 4H), 5.06 (m, 1H),
5.00-4.66 (m, 16H), 4.53 (m, 1H), 4.46 (m, 1H), 4.37 (m, 1H),
4.26 (m, 1H), 4.13-3.93 (m, 4H), 3.86 (m, 1H), 3.78 (m, 1H),
2.14 (m, 4H), 1.54 (m, 4H), 1.25 (m, 16H), 0.88 (m 6H); 13C NMR
(CDCl3, 100 MHz) δ 173.3, 173.0, 172.9, 140.0, 138.5, 138.1,
138.0, 135.8, 135.7, 135.6, 129.0, 128.9, 128.8, 128.6, 128.5, 128.4,
128.3, 128.1, 128.0, 127.9, 127.7, 127.6, 127.5, 80.2, 78.3, 78.1,
76.1, 76.0, 75.0, 74.9, 70.2, 70.0, 69.9, 69.8, 69.7, 69.5, 69.4, 65.9,
61.7, 34.3, 34.2, 31.9, 31.2, 29.3, 29.2, 29.1, 25.0, 24.9, 22.8, 14.3;
31P NMR (CDCl3, 162 MHz) δ -0.73, -0.79, -0.84, -0.88; IR
(film, cm-1) 2927, 2855, 1742, 1497, 1455, 1271, 1158, 1013, 880,
Conclusions
In summary, we have completed a range of total syntheses
of phosphatidylinositol derivatives. In either the PI3P, PI3,5P2,
or PI series, the sequences we have carried out are highly
efficient and streamlined in comparison to prior approaches.
The key step is the catalytic asymmetric phosphorylation
reaction that effectively desymmetrizes the myo-inositol ring,
such that either enantiomeric series may be synthesized. Efficient
sequences that rely on a common approach to each target family
allow unified reaction conditions to be applied throughout the
series. For the PI compounds, a Mitsunobu coupling reaction
was used that enables a wide variety of glycerol surrogates to
be used. The schemes also enable the syntheses of a range of
compounds wherein hydroxyls and/or phosphates at strategic
positions in these PI-based targets may be deleted. These
syntheses now set the stage for application of these probes to
J. Org. Chem, Vol. 71, No. 13, 2006 4925

Xu et al.
735, 696; TLC Rf 0.24 (50% ethyl acetate/hexanes); exact mass
calcd for [C81H97O19P3 + H]+ requires m/z 1467.5915, found
1467.5924 (ESI+); [R]D ) +2.0 (4.0, CHCl3).
10 h. It was then cooled to rt and concentrated. The residue was
purified by silica gel chromatography eluting with EtOAc/hexanes
(1:1) to CH2Cl2/MeOH (1:1) to give the crude product as a lithium
salt. The salt was dissolved in MeOH, and H+ ion-exchange resin
was added. The mixture was then filtered through cotton. The filtrate
was concentrated to give the crude phosphoric diester product. The
crude residue was then dissolved in 5 mL of THF, and diacyl-
glycerol 19 (0.076 g, 0.12 mmol) and triphenylphosphine (0.032
g, 0.12 mmol) were added followed by DEAD (19 µL, 0.12 mmol).
The reaction mixture was stirred at rt under N2 for 12 h. The mixture
was then concentrated under reduced pressure and purified by silica
gel chromatography eluting with 10-50% EtOAc/hexanes to yield
32 mg of 20 (yield ) 40% over two steps): 1H NMR (CDCl3, 300
MHz) δ 7.36-7.20 (m, 20H), 5.10 (m, 1H), 5.03 (m, 2H), 4.88-
4.71 (m, 6H), 4.31-4.09 (m, 4H), 4.07-3.93 (m, 2H), 3.88 (m,
1H), 3.67 (t, J ) 9.6 Hz, 1H), 3.54 (m, 2H), 2.42 (m, 1H), 2.21
(m, 5H), 1.55 (m, 4H), 1.23 (m, 16H), 0.88 (m, 6H); 13C NMR
(CDCl3, 100 MHz) δ 173.1, 172.7, 138.5, 138.3, 128.8, 128.7,
128.6, 128.5, 128.4, 128.1, 128.0, 127.9, 127.8, 127.7, 127.6, 81.4,
81.3, 80.2, 80.1, 79.1, 79.0, 78.9, 75.7, 75.6, 75.5, 75.4, 75.2, 75.0,
74.9, 72.0, 70.0, 69.9, 69.8, 69.7, 69.5, 69.4, 65.9, 65.8, 65.7, 65.6,
61.8, 61.7, 34.4, 34.2, 31.9, 30.0, 29.4, 29.3, 29.2, 25.1, 22.9, 14.4;
31P NMR (CDCl3, 121 MHz): δ -0.4, -0.5; IR (film, cm-1): 3550,
3461, 3064, 3031, 2926, 2856, 2360, 1743, 1497, 1455, 1259, 1109,
1020, 895, 737, 697; TLC: Rf ) 0.54 (1:1EtOAc/Hexanes); exact
mass calcd for [C53H71O13P + H]+ requires m/z 947.4711, found
947.4708 (ESI+); [R]D ) +2.4 (1.0, CHCl3).
PI-DiC8. To a stirred solution of 20 (0.028 g, 0.030 mmol) in
t-BuOH/H2O (5:1, 3 mL) was added sodium ion-exchange resin
(Chelex 100 sodium form, 50-100 dry mesh; washed by H2O)
followed by Pd(OH)2/C (washed by H2O). The reaction was then
stirred at 1 atm of H2 for 24 h. The reaction was filtered through
Celite, and the filtrate was concentrated under reduced pressure
and lyophilized to afford a white solid (0.018 g, 98% yield): 1H
NMR (D2O, 300 MHz) δ 5.13 (m, 1H), 4.27 (m, 1H), 4.06 (m,
2H), 3.90 (m, 2H), 3.77 (t, J ) 7.8 Hz, 1H), 3.58 (t, J ) 9.6 Hz,
1H), 3.45 (t, J ) 9.6 Hz, 1H), 3.38 (m, 1H), 3.17 (t, J ) 9.3 Hz,
1H), 2.26-2.05 (m, 4H), 1.43 (m, 4H), 1.11 (m, 16H), 0.69 (m,
6H); 31P NMR (D2O, 121 MHz) δ 0.4; exact mass calcd for [C25H46-
NaO13P + H]+ requires m/z 609.2652, found 609.2669 (ESI+);
[R]D ) +2.8 (1.0, H2O at pH ) 9).
2,4,6-Tri-O-benzyl-5-O-tert-butyldimethylsilyl-D-myo-inositol-
1-diphenylphosphate-3-O-thiocarbonic Acid Phenyl Ester (21).
To a stirred solution of 2,4,6-tri-O-benzyl-D-myo-inositol-1-diphenyl
phosphate-3-O-thiocarbonic acid phenyl ester (0.572 g, 0.698 mmol)
in dichloromethane (20 mL) was added 2,6-lutidine (326 µL, 2.79
mmol). TBSOTf (481 µL, 2.10 mmol) was then added and the
reaction mixture stirred for 3 h at which time saturated NH4Cl
solution (25 mL) was added and the mixture was extracted with
dichloromethane (3 × 75 mL). The organic layers were combined,
dried over sodium sulfate, and then concentrated under reduced
pressure to afford a yellow, oily solid. The crude product was
purified using silica gel flash chromatography eluting with a
gradient of 0-20% ethyl acetate/hexanes to afford pure product as
a clear oil (580 mg, 90% yield): 1H NMR (CDCl3, 400 MHz) δ
7.43-7.13 (m, 28H), 6.89 (m, 2H), 5.41 (dd, J ) 2.4, 10.0 Hz,
1H), 4.92-4.69 (m, 8H), 4.14 (t, J ) 10.0 Hz, 1H), 4.04 (t, J )
9.2 Hz, 1H), 3.74 (t, J ) 9.2 Hz, 1H), 0.97 (s, 9H), 0.06 (s, 3H),
-0.01 (s, 3H); 13C NMR (CDCl3, 126 MHz) δ 194.5, 153.5, 150.7,
138.7, 138.5, 138.3, 130.0, 129.9, 129.7, 128.6, 128.4, 128.2, 128.1,
128.0, 127.6, 127.5, 127.4, 127.2, 126.8, 125.6, 125.5, 122.0, 120.3,
120.1, 83.6, 80.0, 79.7, 79.6, 79.5, 76.0, 75.7, 75.6, 74.8, 26.2, 18.2,
0.2, -3.8, -3.9; 31P NMR (CDCl3, 162 MHz) δ -11.6; IR (film,
cm-1) 2927, 2855, 2360, 1590, 1490, 1288, 1217, 1190, 1039, 1025,
960, 837, 777, 689; TLC Rf 0.50 (20% ethyl acetate/hexanes); exact
mass calcd for [C52H57O10PSSi + H]+ requires m/z 933.3258, found
933.3232 (ESI+); [R]D ) +22 (1.0, CHCl3).
PI3,5P2-DiC8. To a stirred solution of 16 (0.023 g, 0.016 mmol)
in t-BuOH/H2O (5:1, 3 mL) was added sodium ion-exchange resin
(Chelex 100 sodium form, 50-100 dry mesh, washed by H2O)
followed by Pd(OH)2/C (40 mg, washed by H2O). The reaction
was then stirred at 1 atm of H2 for 32 h. The reaction mixture was
filtered through Celite, and the filtrate was concentrated under
reduced pressure and lyophilized to afford a white solid (0.013 g,
97% yield): 1H NMR (D2O, 300 MHz) δ 5.16 (m, 1H), 4.30 (m,
2H), 4.15 (m, 1H), 3.90-3.88 (m, 4H), 3.74 (m, 3H), 2.29 (t, J )
7.2 Hz, 2H), 2.25 (t, J ) 7.8 Hz, 2H),1.45 (m, 4H), 1.14 (m, 16H),
0.72 (m, 6H); 31P NMR (D2O, 121 MHz) δ 3.8, 2.3, 0.2; exact
mass calcd for [C25H49O19P3 - H]- requires m/z 745.2003, found
745.2016 (ESI-); [R]D ) +5.2 (1.0, H2O at pH ) 9).
Protected PI3,5P2 (17). To a stirred solution of ent-15 (0.0500
g, 0.0310 mmol) in CH2Cl2/toluene (1.5 mL/ 1.5 mL) was added
dibenzyl diisopropylphosphoramidite (0.216 g, 0.626 mmol) fol-
lowed by 4,5-dicyanoimidazole (0.0920 g, 0.782 mmol). The
reaction mixture was stirred at room temperature for 12 h and then
cooled to 0 °C. 30% H2O2/H2O (3 mL) was added, and the reaction
was stirred at 0 °C for another 1 h. The reaction was then quenched
with saturated Na2SO3 solution (20 mL) and the mixture extracted
with CHCl3 (3 × 75 mL). The organic layers were combined, dried
over sodium sulfate, and then concentrated under reduced pressure
to afford an oil. The crude product was purified using silica gel
flash chromatography eluting with a gradient of 10-50% ethyl
acetate/hexanes to afford pure product as a clear oil (0.050 g, 88%
yield): 1H NMR (CDCl3, 400 MHz) δ 7.32-7.18 (m, 25H), 7.06
(m, 5H), 6.84 (m, 3H), 6.76-6.66 (m, 4H), 5.38-5.30 (m, 8H),
5.10 (m, 1H), 5.00-4.85 (m, 8H), 4.76 (t, J ) 10.0 Hz, 4H), 4.70-
4.63 (m, 4H), 4.50 (m, 1H), 4.36 (m, 2H), 4.26 (m, 2H), 4.12 (m,
1H), 4.01 (t, J ) 9.6 Hz, 3H), 4.00-3.83 (m, 1H), 3.82 (s, 3H),
3.73-3.70 (s, 6H), 2.80 (m, 6H), 2.21 (m, 4H), 2.04 (m, 4H), 1.60
(m, 4H), 1.29 (m, 34H), 0.88 (m 6H); 13C NMR (CDCl3, 100 MHz)
δ 159.2, 159.0, 130.7, 130.3, 129.5, 129.4, 129.2, 129.1, 129.0,
128.8, 128.7, 128.6, 128.5, 128.3, 128.1, 128.0, 127.9, 127.7, 113.9,
113.7, 74.6, 69.7, 60.7, 55.6, 55.5, 34.3, 33.9, 32.3, 31.9, 30.1, 30.0,
29.9, 29.8, 29.7, 29.6, 27.6, 26.9, 26.1, 26.0, 25.2, 25.1, 23.1, 23.0,
21.5, 14.7, 14.6, 14.5; 31P NMR (CDCl3, 162 MHz) δ -0.48, -0.51,
-0.58, -0.61, -0.64, -0.70; IR (film, cm-1) 2924, 2853, 1743,
1613, 1514, 1456, 1379, 1249, 1015, 881, 823, 737, 697; TLC Rf
0.69 (50% ethyl acetate/hexanes); exact mass calcd for [C106H139O22P3
+ Na]+ requires m/z 1879.8869, found 1879.8850 (ESI+); [R]D )
+0.6 (6.0, CHCl3).
PI3,5P2. Compound 17 (0.012 g, 0.0064 mmol) was dissolved
in 1.0 mL of CDCl3, and iodotrimethylsilane (0.028 mL, 0.19 mmol)
was added. The reaction mixture was stirred at room temperature
for 45 min and then cooled to 0 °C and concentrated via vacuum
transfer. The residue was dissolved in 1.5 mL of toluene and then
azeotroped at 0 °C via vacuum transfer. This procedure was repeated
twice. The residue was then dissolved in 1.0 mL of methanol and
stirred at 0 °C for 1 h. This solution was then concentrated under
reduced pressure. The residue was dissolved in a minimal amount
of CH3OH/CH2Cl2 (1:1) and purified by silica gel chromatography
eluting with CHCl3/CH3OH/2.2 M NH4OH (9:7:2) to yield 3 mg
(41% yield) of PI3,5P2 as the ammonium salt: 1H NMR (CD3-
OD/CDCl3/D2O (4:3:1), 400 MHz) δ 5.39 (m, 8H), 5.30 (m, 1H),
4.23 (m, 3H), 4.09 (m, 5H), 3.93 (m, 3H), 2.84 (m, 5H), 2.40 (t, J
) 6.8 Hz, 2H), 2.33 (t, J ) 6.8 Hz, 2H), 2.10 (m, 4H), 1.73 (m,
2H), 1.61 (m, 2H), 1.30 (m, 34H), 0.92 (m, 6H); 31P NMR (CD3-
OD/CDCl3/D2O (4:3:1), 121 MHz) δ 3.0, 1.9, 0.7; TLC Rf 0.10
(9:7:2 CHCl3/CH3OH/2.2 M NH4OH); exact mass calcd for
[C47H85O19P3 + H]+ requires m/z 1047.4976, found 1047.5001
(ESI+).
Protected PI-DiC8 (20). To a stirred solution of 8 (0.058 g,
0.082 mmol) in 5 mL of acetone (reagent grade) was added LiBr
(0.011 g, 0.12 mmol), and the reaction mixture was refluxed for
2,4,6-Tri-O-benzyl-5-O-tert-butyldimethylsilyl-D-myo-inositol-
3-deoxy-1-diphenyl Phosphate. To a stirred solution of 21 (0.180
4926 J. Org. Chem., Vol. 71, No. 13, 2006

Streamlined PI Syntheses
g, 0.193 mmol) in toluene (5 mL) was added tributyltin hydride
(105 µL, 0.386 mmol) followed by AIBN (0.010 g, 0.058 mmol).
The reaction mixture was stirred at reflux for 4 h. The mixture
was cooled to room temperature and concentrated under reduced
pressure to afford an oil. The crude product was then purified using
silica gel flash chromatography eluting with a gradient of 0-20%
ethyl acetate/hexanes to afford pure product as a clear oil (106 mg,
70% yield): 1H NMR (CDCl3, 500 MHz) δ 7.38-7.09 (m, 25H),
4.82 (m, 2H), 4.62 (m, 2H), 4.49-4.34 (m, 3H), 4.14 (m, 1H),
3.90 (t, J ) 9.0 Hz, 1H), 3.64 (m, 2H), 2.16 (dt, J ) 4.0, 14.0 Hz,
1H), 1.30 (m, 1H), 0.92 (s, 9H), 0.08 (s, 3H), -0.01 (s, 3H); 13C
NMR (CDCl3, 100 MHz) δ 150.9, 150.8, 138.9, 138.8, 138.6, 130.0,
129.9, 129.8, 129.6, 128.5, 128.3, 128.1, 128.0, 127.8, 127.5, 127.1,
125.5, 125.3, 120.4, 120.3, 120.2, 120.1, 82.4, 82.3, 80.7, 80.6,
77.0, 75.5, 74.8, 72.9, 72.2, 30.4, 26.3, 18.4, -3.8, -3.9; 31P NMR
(CDCl3, 162 MHz) δ -11.4; IR (film, cm-1) 2927, 2855, 1590,
1490, 1288, 1191, 1088, 1070, 1024, 955, 835, 777; TLC Rf 0.29
(20% ethyl acetate/hexanes); exact mass calcd for [C45H53O8PSi
+ H]+ requires m/z 781.3326, found 781.3340 (ESI+); [R]D ) -24
(1.0, CHCl3).
2,4,6-Tri-O-benzyl-5-O-tert-butyldimethylsilyl-D-myo-inositol-
3-deoxy-1-dibenzyl Phosphate. To a stirred solution of 2,4,6-tri-
O-benzyl-5-O-tert-butyldimethylsilyl-D-myo-inositol-3-deoxy-1-
diphenyl phosphate (0.116 g, 0.148 mmol) in THF (5 mL) was
added benzyl alcohol (34.0 µL, 0.327 mmol) followed by sodium
hydride (0.0150 g, 0.594 mmol). The reaction mixture was stirred
and monitored by thin-layer chromatography. Starting material was
consumed after 3 h, and the reaction was quenched with 0.5 M
citric acid until pH ) 7. The reaction mixture was then concentrated
under reduced pressure and extracted with diethyl ether (3 × 75
mL) in 25 mL of brine. The organic layers were combined, dried
over sodium sulfate, and then concentrated under reduced pressure
to afford an oil. The crude product was purified using silica gel
flash chromatography eluting with a gradient of 0-20% ethyl
acetate/hexanes to afford pure product as a clear oil (110 mg, 91%
yield): 1H NMR (CDCl3, 300 MHz) δ 7.37-7.07 (m, 25H), 4.87-
(m, 3H), 4.77 (m, 3H), 4.59-4.31 (m, 4H), 4.27 (m, 1H), 4.02 (m,
1H), 3.77 (t, J ) 9.0 Hz, 1H), 3.63-3.49 (m, 2H), 2.05 (dt, J )
4.2, 14.1 Hz, 1H), 1.16 (m, 1H), 1.10 (s, 9H), 0.03 (s, 3H), -0.04
(s, 3H); 13C NMR (CDCl3, 100 MHz) δ 139.0, 138.7, 138.6, 128.7,
128.6, 128.5, 128.4, 128.3, 128.2, 128.1, 128.0, 127.9, 127.8, 127.3,
127.2, 81.3, 81.2, 81.0, 75.5, 74.8, 72.9, 72.5, 69.5, 69.4, 69.3, 30.7,
26.5; 31P NMR (CDCl3, 121 MHz) δ -0.5; IR (film, cm-1) 2927,
2889, 2855, 1497, 1455, 1259, 1144, 1088, 1071, 1015, 1001, 835,
735, 696; TLC Rf 0.12 (25% ethyl acetate/hexanes); exact mass
calcd for [C47H57O8PSi + H]+ requires m/z 809.3639, found
809.3624 (ESI+); [R]D ) -21 (1.0, CHCl3).
2,4,6-Tri-O-benzyl-D-myo-inositol-3-deoxy-1-dibenzyl Phos-
phate (22). To a stirred solution of 2,4,6-tri-O-benzyl-5-O-tert-
butyldimethylsilyl-D-myo-inositol-3-deoxy-1-dibenyl phosphate (0.060
g, 0.074 mmol) in THF (5 mL) (Teflon container) was added HF/
pyridine (0.5 mL). After 48 h, the reaction was neutralized with
saturated sodium bicarbonate solution (50 mL) until basic and
extracted with chloroform (3 × 75 mL). The organic layers were
combined and concentrated under reduced pressure to afford a
viscous oil. The crude product was purified using silica gel flash
chromatography eluting with a gradient of 20-50% ethyl acetate/
hexanes to afford pure product as a clear oil (38 mg, 75% yield):
1H NMR (CDCl3, 400 MHz) δ 7.35-7.19 (m, 25H), 5.01-4.91
(m, 4H), 4.80 (m, 2H), 4.60-4.45 (m, 4H), 4.30 (m, 1H), 4.08 (m,
1H), 3.87 (t, J ) 9.2 Hz, 1H), 3.68 (m, 1H), 3.57 (t, J ) 9.2 Hz,
1H), 2.61 (m, 1H), 2.18 (dt, J ) 4.0, 14.0 Hz, 1H), 1.24 (m, 1H);
13C NMR (CDCl3, 100 MHz) δ 138.6, 138.4, 135.9, 128.6, 128.5,
128.4, 128.0, 127.9, 127.8, 127.7, 127.6, 127.5, 80.7, 80.6, 80.3,
80.2, 76.9, 76.2, 75.4, 75.1, 72.7, 72.5, 69.6, 69.5, 69.4, 69.3, 30.6;
31P NMR (CDCl3, 121 MHz) δ -0.5; IR (film, cm-1) 3411, 3063,
3031, 2929, 2874, 1497, 1455, 1267, 1213, 1069, 1017, 911, 735,
697; TLC Rf 0.33 (50% ethyl acetate/hexanes); exact mass calcd
for [C41H43O8P + H]+ requires m/z 695.2774, found 695.2759
(ESI+); [R]D ) -15 (1.0, CHCl3).
Protected 3-Deoxy-PI-DiC8 (23). To a stirred solution of 22
(0.060 g, 0.086 mmol) in reagent grade acetone (5 mL) was added
lithium bromide (0.011 g, 0.12 mmol). The reaction mixture was
stirred at reflux for 12 h. The mixture was then cooled to room
temperature and concentrated under reduced pressure to afford a
white solid. The crude product was then purified using silica gel
flash chromatography eluting with CHCl3/MeOH/2.2 M NH4OH
(9:7:1) to afford the product as a lithium salt. The lithium salt was
then dissolved in a minimum amount of methanol and loaded onto
a proton ion-exchange resin column. The acidic fractions were
combined and concentrated under reduced pressure (no heat) to
afford a clear oil (acid form). The oil was then dissolved in THF
(5 mL). Triphenylphosphine (0.034 g, 0.13 mmol) was added
followed by diacylglycerol 19 (0.081 g, 0.13 mmol). DEAD (21
µL, 0.13 mmol) was added, and the reaction was stirred for 12 h
and then concentrated under reduced pressure to afford an orange
oil. The crude product was purified using silica gel flash chroma-
tography eluting with a gradient of 5-50% ethyl acetate/hexanes
to afford pure product as a clear oil (0.073 g, 89% yield, two
steps): 1H NMR (CDCl3, 300 MHz) δ 7.39-7.22 (m, 20H), 5.11
(m, 1H), 4.97 (m, 2H), 4.81 (m, 2H), 4.65-4.46 (m, 4H), 4.27-
3.92 (m, 6H), 3.86 (t, J ) 9.6 Hz, 1H), 3.69 (m, 1H), 3.60 (q, J )
7.5 Hz, 1H), 2.60 (m, 1H), 2.20 (m, 5H), 1.55 (m, 5H), 1.26 (m,
16H), 0.87 (m, 6H); 13C NMR (CDCl3, 100 MHz) δ 173.1, 173.0,
172.7, 138.6, 138.4, 138.3, 128.7, 128.6, 128.5, 128.4, 128.0, 127.9,
127.8, 127.7, 127.6, 80.8, 80.7, 80.6, 80.2, 80.1, 76.2, 75.5, 75.4,
75.2, 75.1, 72.7, 72.6, 72.5, 72.4, 69.8, 69.7, 69.6, 69.5, 65.9, 65.8,
65.6, 65.5, 61.9, 61.8, 34.4, 34.3, 31.9, 30.5, 30.4, 29.4, 29.3, 29.2,
25.2, 22.9, 14.1; 31P NMR (CDCl3, 121 MHz) δ -0.6, -0.7; IR
(film, cm-1) 3447, 3031, 2928, 2857, 1742, 1497, 1455, 1360, 1265,
1155, 1102, 1069, 1023, 736, 697; TLC Rf 0.48 (50% ethyl acetate/
hexanes); exact mass calcd for [C53H71O12P + H]+ requires m/z
931.4761, found 931.4771 (ESI+); [R]D ) -8.6 (1.0, CHCl3).
3-Deoxy-PI-DiC8. To a stirred solution of 23 (0.026 g, 0.028
mmol) in t-BuOH/H2O (5:1, 3 mL) was added sodium ion-exchange
resin (Chelex 100 sodium form, 50-100 dry mesh; washed by H2O)
followed by Pd(OH)2/C (washed by H2O). The reaction mixture
was then stirred at 1 atm of H2 for 24 h. The mixture was filtered
through Celite, and the filtrate was concentrated under reduced
pressure and lyophilized to afford a white solid (0.016 g, 97%
yield): 1H NMR (D2O, 300 MHz) δ 5.17 (m, 1H), 4.28 (m, 1H),
4.11 (m, 2H), 3.93 (m, 2H), 3.80 (t, J ) 7.8 Hz, 1H), 3.62 (m,
1H), 3.57 (t, J ) 9.6 Hz, 1H), 3.18 (t, J ) 9.3 Hz, 1H), 2.24 (m,
4H), 1.97 (dt, J ) 3.9, 13.5 Hz, 1H), 1.44 (m, 5H), 1.14 (m, 16H),
0.73 (m, 6H); 31P NMR (D2O, 121 MHz) δ 0.7; exact mass calcd
for [C25H47O12P + H]+ requires m/z 593.2703, found 593.2687
(ESI+); [R]D ) +3.2 (1.0, H2O at pH ) 9).
2,4,6-Tri-O-benzyl-D-myo-inositol-3,5-dideoxy-1-dibenzyl Phos-
phate (25). To a stirred solution of 2,4,6-tri-O-benzyl-D-myo-
inositol-3,5-dideoxy-1-diphenyl phosphate (0.172 g, 0.264 mmol)
in THF (10 mL) was added benzyl alcohol (68.0 µL, 0.661 mmol)
followed by sodium hydride (0.0200 g, 0.793 mmol). The reaction
mixture was stirred and monitored by thin-layer chromatography.
Starting material was consumed after 3 h, and the reaction was
quenched with 0.5 M citric acid until pH ) 7. The reaction was
then concentrated under reduced pressure and extracted with diethyl
ether (3 × 75 mL) in 20 mL of brine. The organic layers were
combined, dried over sodium sulfate, and then concentrated under
reduced pressure to afford an oil. The crude product was purified
using silica gel flash chromatography eluting with a gradient of
0-25% ethyl acetate/hexanes to afford pure product as a clear oil
(0.120 g, 67% yield): 1H NMR (CDCl3, 400 MHz) δ 7.35-7.18
(m, 25H), 4.99 (m, 4H), 4.65-4.43 (m, 6H), 4.39 (dt, J ) 2.8, 8.4
Hz, 1H), 4.12 (m, 1H), 3.91 (m, 1H), 3.77 (m, 1H), 2.45 (m, 1H),
2.24 (m, 1H), 1.53-1.40 (m, 2H); 13C NMR (CDCl3, 100 MHz) δ
138.6, 128.6, 128.5, 128.4, 128.3, 127.9, 127.8, 127.7, 127.6, 127.5,
75.4, 75.0, 74.9, 72.6, 72.2, 71.5, 70.9, 69.4, 69.3, 69.2; 31P NMR
J. Org. Chem, Vol. 71, No. 13, 2006 4927

Xu et al.
(CDCl3, 162 MHz) δ -0.5; IR (film, cm-1) 3062, 3031, 2942, 2866,
1497, 1454, 1359, 1278, 1212, 1071, 1016, 736, 696; TLC Rf 0.22
(30% ethyl acetate/hexanes); exact mass calcd for [C41H43O7P +
H]+ requires m/z 679.2825, found 679.2797 (ESI+); [R]D ) -16
(4.8, CHCl3).
29.4, 29.3, 29.2, 25.2, 25.2, 23.0, 14.4; 31P NMR (CDCl3, 162 MHz)
δ -0.6, -0.7; IR (film, cm-1) 3063, 3031, 2928, 2857, 1742, 1497,
1455, 1360, 1282, 1158, 1097, 1070, 1023, 737, 697; TLC Rf 0.64
(50% ethyl acetate/hexanes); exact mass calcd for [C53H71O11P +
H]+ requires m/z 915.4812, found 915.4796 (ESI+); [R]D ) -10
(1.0, CHCl3).
Protected 3,5-Dideoxy-PI-DiC8 (26). To a stirred solution of
25 (0.095 g, 0.14 mmol) in reagent grade acetone (8 mL) was added
lithium bromide (0.016 g, 0.18 mmol). The reaction mixture was
stirred at reflux for 12 h. The mixture was then cooled to room
temperature and concentrated under reduced pressure to afford a
white solid. The crude product was then purified using silica gel
flash chromatography eluting with CHCl3/MeOH/2.2 M NH4OH
(9:7:1) to afford the product as a lithium salt. The lithium salt was
then dissolved in a minimum amount of methanol and loaded onto
a proton ion-exchange resin column. The acidic fractions were
combined and concentrated under reduced pressure (no heat) to
afford a clear oil (acid form). The oil was then dissolved THF (5
mL). Triphenylphosphine (0.040 g, 0.15 mmol) was added followed
by diacylglycerol 19 (0.096 g, 0.15 mmol). DEAD (24 µL, 0.15
mmol) was added, and the reaction was stirred for 12 h and then
concentrated under reduced pressure to afford an orange oil. The
crude product was purified using silica gel flash chromatography
eluting with a gradient of 5-50% ethyl acetate/hexanes to afford
pure product as a clear oil (0.070 g, 55% yield, two steps): 1H
NMR (CDCl3, 400 MHz) δ 7.28 (m, 20H), 5.10 (m, 1H), 4.97 (m,
2H), 4.65-4.43 (m, 6H), 4.36 (m, 1H), 4.19-3.98 (m, 5H), 3.90
(m, 1H), 3.76 (m, 1H), 2.46 (m, 1H), 2.20 (m, 5H), 1.55 (m, 4H),
1.47 (m, 2H), 1.23 (m, 16H), 0.88 (m, 6H); 13C NMR (CDCl3, 100
MHz) δ 173.1, 172.7, 138.5, 138.4, 136.0, 135.9, 128.6, 128.5,
128.4, 128.3, 127.8, 127.7, 127.6, 81.7, 74.9, 72.6, 72.1, 72.0, 71.4,
71.3, 70.9, 69.6, 69.5, 69.4, 65.7, 65.5, 62.0, 34.4, 34.3, 34.2, 32.0,
3,5-Dideoxy-PI-DiC8. To a stirred solution of 26 (0.063 g, 0.069
mmol) in t-BuOH/H2O (5:1, 3 mL) was added sodium ion-exchange
resin (Chelex 100 sodium form, 50-100 dry mesh; washed by H2O)
followed by Pd(OH)2/C (washed by H2O). The reaction was then
stirred at 1 atm of H2 for 24 h. The reaction was filtered through
Celite, and the filtrate was concentrated under reduced pressure
and lyophilized to afford a white solid (0.037 g, 93% yield): 1H
NMR (D2O, 300 MHz) δ 5.11 (m, 1H), 4.25 (m, 1H), 4.07 (m,
2H), 3.88 (m, 3H); 3.74 (m, 2H), 2.22 (t, J ) 6.6 Hz, 2H), 2.16 (t,
J ) 7.5 Hz, 2H), 2.06 (m, 1H), 1.94 (m, 1H), 1.42 (m, 5H), 1.26
(m, 1H), 1.10 (m, 16H), 0.69 (m, 6H); 31P NMR (D2O, 121 MHz)
δ 0.8; exact mass calcd for [C25H47O11P + H]+ requires m/z
577.2754, found 577.2738 (ESI+); [R]D ) -4.0 (1.0, H2O at pH
) 9).
Acknowledgment. We thank the National Institutes of
General Medical Sciences of the National Institutes of Health
for support (GM-068649).
Supporting Information Available: Experimental procedures
and product characterization for additional intermediates discussed
in the manuscript. This material is available free of charge via the
Internet at http://pubs.acs.org.
JO060702S
4928 J. Org. Chem., Vol. 71, No. 13, 2006