(±)-1,2:5,6-Di-O-isopropylidene-myo-inositol and (±)-6-O-benzoyl-1,2:4,5-di-O-isopropylidene-myo-inositol: A practical preparation of key intermediates for myo-inositol phosphates

Article abstract of DOI:10.1016/S0008-6215(01)00286-5

A simple and practical synthetic procedure for the versatile intermediates, (±)-1,2:5,6-di-O-isopropylidene-myo-inositol and (±)-6-O-benzoyl-1,2:4,5-di-O-isopropylidene-myo-inositol, is described.

Full text of DOI:10.1016/S0008-6215(01)00286-5

Carbohydrate Research 337 (2002) 75–78
www.elsevier.com/locate/carres
Note
(9)-1,2:5,6-Di-O-isopropylidene-myo-inositol and
(9)-6-O-benzoyl-1,2:4,5-di-O-isopropylidene-myo-inositol: a
practical preparation of key intermediates for myo-inositol
phosphates
Sonya M. Khersonsky, Young-Tae Chang*
Department of Chemistry, New York Uni6ersity, 100 Washington Square East, New York, NY 10003, USA
Received 16 August 2001; accepted 27 September 2001
Abstract
A simple and practical synthetic procedure for the versatile intermediates, (9)-1,2:5,6-di-O-isopropylidene-myo-inositol and
(9)-6-O-benzoyl-1,2:4,5-di-O-isopropylidene-myo-inositol, is described. © 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Cyclitol; Inositol phosphate intermediate; Debenzoylation
other minor isomers. Separation of compounds 5 and 6
requires column chromatography, which may not be
appropriate for large-scale synthesis.3 In situ benzoyla-
tion of the mixture, a procedure developed by Gigg et
al., greatly facilitates the separation of 3,6-dibenzoate
isomer 3 due to its highly crystalline nature (26%
yield).4 However, unlike the case of isomer 3, further
purification of 3,4-dibenzoate 2 from the filtrate is
neither efficient (13% yield from inositol) nor straight-
forward.5 One of the problems in purification is the
presence of a significant amount of (9)-3,4:5,6-tetra-
O-benzoyl-1,2-O-isopropylidene-myo-inositol (4) (28%)
in the filtrate, which competitively crystallizes along
with the desirable compound 2.5 An alternative ap-
proach was introduced for an easy purification of 2 and
3, but the procedure involves a two-step reaction from
myo-inositol, using (9)-1,2-O-isopropylidene-myo-
inositol as an intermediate.6
Several myo-inositol phosphates are widely recog-
nized for their biological roles as second messengers in
intracellular signaling.1 The discovery of novel inositol
phosphates and their functions in living systems is still
ongoing.2 The preparation of properly protected inosi-
tol intermediates is a key step in the synthesis of
inositol phosphates as well as other related products
such as conduritols and pseudosugars. In this paper, we
report a simple and practical synthesis of (9)-1,2:5,6-
di-O-isopropylidene-myo-inositol and (9)-6-O-ben-
zoyl-1,2:4,5-di-O-isopropylidene-myo-inositol,
which
are versatile intermediates (Scheme 1).
Acetal derivatives are one of the most popular classes
of inositol phosphate intermediates; (9)-1,2:4,5-di-O-
isopropylidene-myo-inositol (6) and (9)-1,2:5,6-di-O-
isopropylidene-myo-inositol (5) are two examples of
these derivatives. When myo-inositol is treated with
acetone (or its ketal derivatives) in the presence of an
acid catalyst, a mixture of compounds 5, 6, and (9)-
1,2-O-isopropylidene-myo-inositol is formed along with
To simplify the purification of compound 2, we at-
tempted to modify the original procedure of Gigg et al.4
by reducing the formation of tetrabenzoate 4. A mix-
ture of myo-inositol, 2,2-dimethoxypropane, p-toluene-
sulfonic acid in DMF was stirred at 120 °C for 3 h. The
reaction mixture was further benzoylated by addition of
* Corresponding author. Tel.: +1-212-9988400; fax: +1-
212-2607905.
E-mail address: yt.chang@nyu.edu (Y.-T. Chang).
0008-6215/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(01)00286-5

76
S.M. Khersonsky, Y.-T. Chang / Carbohydrate Research 337 (2002) 75–78
benzoyl chloride and pyridine at 0 °C. After the filtra-
tion of compound 3, HPLC–MS (High-Performance
Liquid Chromatography-Mass Spectrometry) was used
to monitor the product ratio of compound 2 versus
tetrabenzoate 4. An increased amount of 2,2-
dimethoxypropane (up to 16 equiv) in the first step did
not significantly reduce the amount of tetrabenzoate 4
(nearly equimolar amount with 2). The only result was
an increased amount of brown oil. Similarly, the addi-
tion of molecular sieves to the reaction mixture did not
decrease the amount of tetraol.
Finally, we undertook a second approach—to re-
move any existing (9)-1,2-O-isopropylidene-myo-inos-
itol from the reaction mixture. After 3 h of stirring at
120 °C, the reaction mixture was concentrated by vac-
uum distillation and filtered through a plug of silica gel.
The plug was washed with ethyl acetate to elute a
mixture of diols leaving most of the (9)-1,2-O-iso-
propylidene-myo-inositol behind. After benzoylation of
the diols, compound 3 was isolated by filtration in 26%
yield. HPLC–MS examination of the filtrate indicated
less than a 4% of tetrabenzoate 4 contamination of
compound 2. The filtrate was stirred with an excess
amount of water to give a crude solid material. The
solid was then easily recrystallized from acetone to give
pure compound 2 in 22% yield.
step in place of neutralization–extraction. Base-cata-
lyzed transacylation of the benzoyl groups in com-
pound 2 or 3 was performed using NaOMe in MeOH.
The crude reaction mixture, after dilution with
dichloromethane, was filtered through a plug of silica
and evaporated to dryness. Washing the solid with
ethyl acetate–hexane solvent mixture to remove the
byproduct, methyl benzoate, afforded pure diol prod-
ucts 5 or 6 in 91 and 94% yields, respectively.
Finally, we developed regioselective hydrolysis reac-
tion conditions of 3 to synthesize (9)-6-O-benzoyl-
1,2:4,5-di-O-isopropylidene-myo-inositol (7). Treatment
of 3 with sodium hydroxide in a dichloromethane–
methanol solvent mixture generated the major compo-
nent 7, along with its isomer (9)-3-benzoyl-
1,2:4,5-di-O-isopropylidene-myo-inositol (8), and diol 6
as minor components. GC–MS was used to monitor
relative amounts of 3, 7, 8, and 6 (tR 26.59, 22.03,
21.55, and 15.88 min, respectively). The optimum con-
ditions to maximize the amount of compound 7 were
established after a series of screening of solvents, cata-
lysts, reaction times, and temperatures. The reaction
mixture was purified by column chromatography to
give pure compound 7 in 49% yield. It is noteworthy
that the selective benzoylation of diol 6 gives the 3-ben-
zoate 8 as a major product.7 Thus, this selective deben-
zoylation is a complementary method to synthesize
novel isomer 7, which otherwise has to be prepared by
multiple protection–deprotection procedures.
A previously reported procedure for the hydrolysis of
2 and 3 with NaOH in MeOH requires a mild neutral-
ization of the reaction mixture with solid carbon diox-
ide to suppress acetal migration or hydrolysis, followed
by an extraction.4,5 This neutralization procedure is
usually time consuming due to vigorous gas generation
and hypercooling. In order to obtain a high recovery of
the product, the extraction step requires a large excess
of organic solvent. To simplify this procedure, we used
sodium methoxide, combined with a silica gel filtration
1. Experimental
General methods.—All reagents were commercially
available from Aldrich or Acros and used without
Scheme 1. Reagents: (a) 2,2-dimethoxypropane, p-TSA, DMF; (b) BzCl, pyridine; (c) NaOCH3, MeOH; (d) 0.4 equiv NaOH,
MeOH, CH2Cl2, H2O.

S.M. Khersonsky, Y.-T. Chang / Carbohydrate Research 337 (2002) 75–78
77
further purification. NMR spectra were recorded on a
Gemini 300 MHz spectrometer. GC–MS data were
obtained on a Hewlett–Packard HP5971 GC–MS spec-
trometer (HP–5MS ultra low bleed 5%-diphenyl–95%-
dimethylsiloxane copolymer column, oven temperature
range 80–310 °C). HPLC–MS data were obtained on a
Hewlett–Packard HP1100 spectrometer equipped with
a C18 column (20×4 mm, 3 mm) using 5–95% MeCN
(over water) gradient. The identities of the compounds
were confirmed by electrospray-ionization mass spectra
(ESIMS), and the relative quantities were measured by
UV absorption at 230 nm. Silica filtrations were per-
off-white solid was thoroughly washed with 1:5
EtOAc–hexanes solvent mixture to remove methyl ben-
zoate. The white solid 5 was obtained in 91% yield (5.03
g). 1H NMR data were in agreement with those re-
ported in the literature.5 ESIMS: [M+H]+ Calcd,
261.1; Found, 261.1.
Preparation of (9)-1,2:4,5-di-O-isopropylidene-myo-
inositol (6).—Compound 3 (10 g, 21.4 mmol) was
suspended in 300 mL of MeOH and treated with
NaOMe (1.38 g, 25.6 mmol). The reaction mixture was
allowed to stir at reflux until all starting material had
reacted (solution turned clear). After an additional 30
min of stirring under reflux, the reaction mixture was
cooled to rt and diluted with CH2Cl2 (1500 mL). The
solution was then filtered through a plug of silica (4 cm
in height, 7 cm in diameter, 50 g). The silica was
washed with 150 mL of EtOAc, and the solvents were
evaporated in vacuo. The off-white solid was thor-
oughly washed with 1:5 EtOAc–hexanes solvent mix-
ture to remove methyl benzoate. The white solid 6 was
,
formed on Sorbent Technologies, 60 A silica gel (63–
200 mesh). TLC was done on SAI F254 precoated silica
gel plates (250-mm layer thickness).
Preparation of (9)-3,6-di-O-benzoyl-1,2:4,5-di-O-
isopropylidene-myo-inositol (3).—A mixture of myo-
inositol (25 g, 139 mmol), 2,2-dimethoxypropane (75
mL, 600 mmol), and p-toluenesulfonic acid monohy-
drate (0.5 g, 2.6 mmol) in DMF (100 mL) was heated
to 120 °C for 3 h until no more solid remained. The
reaction mixture was concentrated by vacuum distilla-
tion (50 °C at 20 barr), and Et3N (0.36 mL) was added
to the residual oil. The oil was diluted with 100 mL of
EtOAc and vacuum evaporated together with silica (40
g) to complete dryness. The resulting powder was
loaded on top of a plug of silica (5.5 cm in height, 8.5
cm in diameter, 120 g) and washed thoroughly with
EtOAc (700 mL). After the solvent was removed in
vacuo, pyridine (70 mL) and PhCOCl (50 mL, 0.42
mol) were added to the residual oil at 0 °C. The reac-
tion mixture was then allowed to stir at rt for 2 h. The
precipitate was collected (the initial filtrate was set aside
for further use) and washed with pyridine, water,
Me2CO, and Et2O to give 3 as a white solid in 26%
1
obtained in 94% yield (5.2 g). H NMR data were in
agreement with those reported in the literature.5 ES-
IMS: [M+H]+ Calcd, 261.1; Found, 261.1.
Preparation of (9)-6-O-benzoyl-1,2:4,5-di-O-iso-
propylidene-myo-inositol (7).—Compound 3 (5 g, 10.7
mmol) was completely dissolved in CH2Cl2 (350 mL)
and MeOH (200 mL). NaOH (0.17 g, 4.27 mmol) in
water (50 mL) was diluted with MeOH (100 mL) and
added to the reaction mixture over 20 min. The reac-
tion mixture was stirred overnight at rt, and the sol-
vents were removed in vacuo. The solid was dissolved
in EtOAc and vacuum evaporated together with silica
(10 g) to complete dryness. The resulting powder was
loaded on top of a plug of silica (4 cm in height, 7 cm
in diameter, 50 g) and washed with 5:1 hexanes–EtOAc
to elute methyl benzoate and starting material 3. The
silica gel was further washed with 1:1:1 solvent mixture
of hexanes–EtOAc–CH2Cl2 to elute 7, that, after evap-
oration of solvents, was obtained in 49% yield (1.9 g).
1
yield (16.9 g). H NMR data were in agreement with
those reported in the literature.5 ESIMS: [M+H]+
Calcd, 469.2; Found, 469.0.
Preparation of (9)-3,4-di-O-benzoyl-1,2:5,6-di-O-
isopropylidene-myo-inositol (2).—The initial pyridine
filtrate from the preparation of 3 was diluted with an
excess of water (1 L) and stirred at rt for 48 h. The
precipitate was collected, thoroughly washed with wa-
ter, and recrystallized from acetone to give 2 as a white
solid in 22% yield (14.2 g). 1H NMR data were in
agreement with those reported in the literature.5 ES-
IMS: [M+H]+ Calcd, 469.2; Found, 469.0.
Preparation of (9)-1,2:5,6-di-O-isopropylidene-myo-
inositol (5).—Compound 2 (10 g, 21.4 mmol) was
dissolved in MeOH (150 mL) and treated with NaOMe
(1.38 g, 25.6 mmol). The reaction mixture was allowed
to stir at reflux for 1.5 h, cooled down to rt, and diluted
with CH2Cl2 (750 mL). The solution was then filtered
through a plug of silica (4 cm in height, 7 cm in
diameter, 50 g). The silica was washed with EtOAc (150
mL), and the solvents were removed in vacuo. The
1
No trace of 8 was detected by GC–MS analysis. H
NMR (CDCl3): l 1.31, 1.38, 1.43, 1.59 (4s, each 3 H, 2
CMe2), 2.47 (d, 1 H, J 8.5 Hz, 3-OH), 3.51 (dd, 1 H,
J5,4 9.0 Hz, H-5), 3.98 (dd, 1 H, J4,3 10.0, H-4), 4.08
(ddd, 1 H, J3,2 4.5 Hz, H-3), 4.29 (dd, 1 H, J1,6 6.7 Hz,
H-1), 4.48 (dd, 1 H, J2,1 4.8 Hz, H-2), 5.47 (dd, 1 H, J6,5
11.0 Hz, H-6), 6.37 (dd, 2 H, Ph), 7.51 (dd, 1 H, Ph),
8.03 (d, 2 H, Ph); ESIMS: [M+H]+ Calcd, 365.2;
Found, 365.0; Anal. Calcd for C19H24O7: C, 62.63; H,
6.64. Found: C, 62.48; H, 6.73.
Acknowledgements
This project was supported by a New York Univer-
sity Research Challenge Fund Grant.

78
S.M. Khersonsky, Y.-T. Chang / Carbohydrate Research 337 (2002) 75–78
3. de la Pradilla, R. F.; Jaramillo, C.; Jimenez-Barbero, J.;
References
Martin-Lomas, M.; Penades, S.; Zapata, A. Carbohydr.
Res. 1990, 207, 249–257.
4. Gigg, J.; Gigg, R.; Payne, S.; Conant, R. Carbohydr. Res.
1985, 142, 132–134.
1. Potter, B. V. L.; Lampe, D. Angew. Chem., Int. Ed. Engl.
1995, 34, 1933–1972.
2. (a) Shears, S. B. Bioessays 2000, 22, 786–789;
5. Chung, S. K.; Ryu, Y. Carbohydr. Res. 1994, 258, 145–
167.
(b) Chi, T. H.; Crabtree, G. R. Science 2000, 287, 1937–
1939;
6. Gigg, J.; Gigg, R.; Payne, S.; Conant, R. J. Chem. Soc.,
Perkin Trans. 1 1987, 2411–2414.
(c) Fukuda, M.; Mikoshiba, A. K. Bioassays 1997, 19,
593–603;
7. Estevez, V. A.; Prestwich, G. D. Tetrahedron Lett. 1991,
32, 1623–1626.
(d) Shamsuddin, A. M. Anticancer Res. 1999, 19, 3733–
3736.