Ready routes to key myo-inositol component of GPIs employing microbial arene oxidation or Ferrier reaction

Article abstract of DOI:10.1039/a800282g

Microbial arene oxidation or Ferrier reaction of enol acetates provides versatile complementary routes that greatly facilitate preparation of inositol synthon(s) for GPI assembly.

Full text of DOI:10.1039/a800282g

View Article Online / Journal Homepage / Table of Contents for this issue
Ready routes to key myo-inositol component of GPIs employing
microbial arene oxidation or Ferrier reaction
Zhaozhong J. Jia, Lars Olsson and Bert Fraser-Reid*
Natural Products and Glycotechnology Research Institute, Inc., (NPG)
4118 Swarthmore Road, Durham, North Carolina, 27707, USA
Microbial arene oxidation or Ferrier reaction of enol ace-
tates provides versatile complementary routes that greatly
facilitate preparation of inositol synthon(s) for GPI
assembly.
OH
Cl
HO
O
OH
OH
i
O
3
It was apparent that glycosylphosphatidylinositols (GPIs) had a
special place in glycobiology1 even before the first assignment
of covalent structure to a member of this relatively new class of
biomolecules by Ferguson and co-workers.2,3 Although over
ii
71%
OBn
OBn
BnO
BnO
100 proteins/enzymes have now been identified4 that are
anchored to cell surfaces by GPIs, answers to the question
‘What do GPIs do?’ continue to be evasive.5 Despite having a
central ‘conserved core’,6 the glyco domains of GPIs show
impressive variations, as may be judged by comparing struc-
tures from Trypanosoma,3 rat brain Thy-17 and Leishmania.8
The inositol moiety is more uniform; however, in our recently
completed synthesis of the rat brain Thy-1 anchor,9 the prepar-
ation of the inositol glycosyl acceptor 110 caused the most angst
of the entire project.11 We have therefore been examining other
routes that make the universal component 1 more accessible.
We report herein on two of them.
OBn
OBn
OBn
OBn
O
S
iii
HO
O
83%
O
OH
O
5
1
4
v
iv
70–80%
OBn
BnO
OBn
OBn
vi (43%); vii (35%)
HO
AcO
6
Our original route12 employed a plan where the regioisomers
2a and 2b, prepared from myo-inositol, were independently
processed to give optically active 1 (Scheme 1). As an alterna-
tive, we were attracted to tetrol 3 whose ready preparation
from chlorobenzene via microbial oxidation has been elegantly
explored by Hudlicky et al.13 The axial OH of the routinely
prepared derivative 4 had to be inverted, but attempts to facili-
tate this by selectively protecting the equatorial OH were not
encouraging.
Scheme 2 Reagents and conditions: i, ref. 13; ii, BnBr, NaH, DMF,
Bu4NI, then AcOH, H2O, 70 ЊC, 1 h; iii, SOCl2, Et3N, 0 ЊC, 20 min, then
RuCl3, NaIO4, 0 ЊC, 1 h; iv, CsOAc, DMSO, 50 ЊC, 10–20 min, then
THF, H3Oϩ, 1 h; v, NaOAll, AllOH; vi, NaOMe, MeOH; vii, AllBr
(1 equiv.), NaH, DMF, Bu4NI, 0 ЊC
We now had to differentiate between the three hydroxy
groups of 12. Bearing in mind David’s precedent with galactose
derivatives,18 triol 12 was subjected to stannylene-mediated
derivatization, whereby the allyl group was selectively installed
at the desired C1 site of 13 in 75% yield based on recovered 12,
along with small amounts of diallylated products.
Benzylation of the axial C2 OH was now necessary, and in
keeping with standard protocol, 13 was treated with 1 equiv. of
p-methoxybenzyl chloride (PMBCl), followed by excess BnBr
to give compound 14 along with a mono-PMB-containing
derivative. That the latter was 15 became evident when oxid-
ative cleavage yielded 16 instead of the expected, known isomer
1.10
OBn
BnO
O
OBn
OBn
O
O
HO
O
R'O
RO
O
1
2a R = PMB, R′ = H
b R = H, R′ = PMB
Scheme 1
The reactivity of the hydroxy groups of 13 was therefore the
reverse of the usual axial versus equatorial expectation. Indeed,
treatment of 13 with 1 equiv. of BnBr gave 1 directly in 66%
unoptimized yield.
On the other hand, the Sharpless-inspired14 cyclic sulfate 5,
obtained in excellent yield from 4 (Scheme 2), reacted with cae-
sium acetate to give 6. Unfortunately a comparable reaction,
which would have led directly to 1, could not be effected with an
allyloxy nucleophile. However, deacetylation of 6 and treatment
of the resulting diol with 1 equiv. of allyl bromide did afford 1,
albeit accompanied by ca. 12% of the regioisomeric allyl ether.
The second route to 1 was based on Bender’s discovery15 that
enol esters undergo the Ferrier reaction16 with excellent stereo-
control. Prestwich had already tested the procedure with the
tri-O-p-methoxybenzyl analog of 8.17 Accordingly, methyl
α--glucopyranoside 7 was converted into 8 by four standard
operations in 80% overall yield, and transformation to enol
acetate 9 was readily achieved.The Ferrier reaction afforded 10,
isolated in 63% yield, and chelation controlled reduction gave
11 and thence triol 12 (Scheme 3).
Experimental
1-O-Allyl-3,4,5-tri-O-benzyl-D-myo-inositol 13
A mixture of triol 12 (994 mg, 2.21 mmol) and bis(tributyltin)
oxide (660 mg, 2.65 mmol) in benzene (25 ml) was refluxed for
20 h, with a Dean–Stark trap. Tetrabutylammonium iodide
(815 mg, 2.21 mmol) and allyl bromide (956 µl, 11.05 mmol)
were added. The mixture was stirred at 70 ЊC under Ar for 2 h.
The solvent was evaporated in vacuo. Flash chromatography of
the residue gave 13 (387 mg, 0.790 mmol) as a solid in 75% yield
[based on recovered triol 12 (520 mg, 1.16 mmol)]. Rf (1:1 light
1
petroleum–EtOAc) 0.54; H NMR (400 MHz; CDCl3: δ 7.37–
J. Chem. Soc., Perkin Trans. 1, 1998
631

View Article Online
OAc
O
HO
HO
HO
HO
BnO
BnO
O
O
O
O
BnO
BnO
ii
i
iii
BnO
BnO
83%
91%
BnO
HO
BnO
BnO
OMe
OMe
OMe
OBn
OMe
9
7
8
iv
63%
6
4 O
HO
6
1
5
OH
OH
BnO
BnO
v
BnO
BnO
OAc
1
OAc
2
BnO
BnO5
5
vi
OBn
OBn
HO
HO
HO
6
2
4
2
3
79%
3
OBn
4
BnO
AcO
BnO
65%
1
3
OH
OH
12
10
11
11
(sugar numbering)
(myo-inositol numbering)
vii
75%
OBn
OBn
OPMB
BnO
OH
BnO
O
BnO
viii
OBn
OBn
OBn
BnO
BnO
+
HO
OBn
OBn
O
O
15 (51%)
14 (20%)
13
ix
x
66%
56%
OBn
OH
BnO
BnO
OBn
OBn
OBn
OBn
HO
BnO
O
O
1
16
Scheme 3 Reagents and conditions: i, ref. 16; ii, DCC, DMSO, TFA, Py, iii, Ac2O, K2CO3, MeCN; iv, Hg(OAc)2, acetone, H2O; v, NaBH(OAc)3,
AcOH, MeCN; vi, NaOMe, MeOH; vii, Bu2SnO, PhH, reflux, then AllBr, Bu4NI, 70 ЊC; viii, PMBCl (1 equiv.), NaH, Bu4NI, DMF, 0 ЊC, 2 h, then
BnBr (3 equiv.), room temp.; ix, BnBr (1 equiv.), NaH, Bu4NI, DMF, Ϫ5 to 0 ЊC, 1–2 h; x, CAN, CH2Cl2, MeCN, H2O
2 M. A. J. Ferguson, S. W. Homans, R. A. Dwek and T. W.
Rademacher, Science, 1988, 239, 753.
3 S. W. Homans, C. J. Edge, M. A. J. Ferguson, R. A. Dwek and T. W.
Rademacher, Biochemistry, 1989, 28, 2881.
4 P. Gerold, V. Eckert and R. T. Schwarz, Trends Glycosci.
Glycotechnol., 1996, 8, 265.
5 M. G. Low, in Abstracts: International Meeting on Interactions of
GPI-anchors with Biological Membranes, Spluegen, Switzerland,
September 14–17, 1997.
7.22 (m, 15H), 5.97–5.87 (m, 1H, allyl), 5.28 (ddd, J 17.2, 3.2,
1.6, 1H, allyl), 5.19 (ddd, J 10.4, 2.8, 1.2, 1H, allyl), 4.92–4.81
(m, 4H, Bn ×2), 4.72 (s, 2H, Bn), 4.22 (dd, J 2.6, 2.8, 1H), 4.21–
4.08 (m, 2H, allyl), 4.02 (dd, J 9.4, 9.2, 1H), 3.96 (dd, J 9.6, 9.6,
1H), 3.42 (dd, J 9.6, 2.8, 1H), 3.35 (dd, J 9.4, 9.2, 1H), 3.15 (dd,
J 9.4, 2.8, 1H), 2.63 (s, 1H, OH), 2.55 (s, 1H, OH); 13C NMR:
δ 138.6 (s, Bn ×2), 137.8 (s, Bn), 134.4 (d, allyl), 128.4–127.5
(d, Bn ×3), 117.7 (t, allyl), 82.8, 80.8, 79.9, 78.8, 72.1 and 66.9
(d, C1, C2, C3, C4, C5 and C6), 75.8, 75.3, 72.6 and 71.2 (t, Bn
×3 and allyl).
6 M. McConville and M. A. J. Ferguson, Biochem. J., 1993, 294, 305.
7 S. W. Homans, M. A. J. Ferguson, R. Anand and A. F. Williams,
Nature, 1988, 333, 269.
8 P. Schneider, M. A. J. Fergusson, M. J. McConville, A. Mehlert and
S. W. Homans, J. Biol. Chem., 1990, 265, 16 955.
9 A. S. Campbell and B. Fraser-Reid, J. Am. Chem. Soc., 1995, 117,
10 387.
10 C. J. J. Elie, R. Verduyn, C. E. Dreef, D. M. Bounts, G. A. van der
Marel and J. H. van Boom, Tetrahedron, 1990, 46, 8243.
11 R. Madsen, U. E. Udodong, C. Roberts, D. R. Mootoo,
P. Konradsson and B. Fraser-Reid, J. Am. Chem. Soc., 1995, 117,
1554.
12 A. J. Bridges, Chemtracts: Org. Chem., 1996, 9, 215.
13 T. Hudlicky, M. Mandel, J. Rouden, R. S. Lee, B. Bachmann,
T. Dudding, K. J. Yost and J. S. Merola, J. Chem. Soc., Perkin
Trans. 1, 1994, 1553.
14 B. M. Kim and K. B. Sharpless, Tetrahedron Lett., 1989, 30, 655.
15 S. L. Bender and R. J. Budhu, J. Am. Chem. Soc., 1991, 113, 9883.
16 R. J. Ferrier and S. Middleton, Chem. Rev., 1993, 93, 2779.
17 G. D. Prestwich, Acc. Chem. Res., 1996, 29, 503.
18 S. David, A. Thieffry and A. Veyrieres, J. Chem. Soc., Perkin Trans.
1, 1981, 1796.
1-O-Allyl-2,3,4,5-tetra-O-benzyl-D-myo-inositol 1
A mixture of diol 13 (600 mg, 1.22 mmol), NaH (60% in min-
eral oil, 195 mg, 4.88 mmol) and tetrabutylammonium iodide
(450 mg, 1.22 mmol) in anhydrous DMF (40 ml) was chilled in
an ice bath and covered by Ar. To the stirred mixture was added
benzyl bromide (160 µl, 1.34 mmol). After stirring for 2 h at
0 ЊC, the reaction was quenched with methanol. The mixture
was evaporated in vacuo. Flash chromatography of the residue
gave 1 (470 mg, 0.81 mmol) as a crystalline material in 66%
1
yield. Rf (4:1 light petroleum–EtOAc) 0.36; H and 13C NMR
(CDCl3) data were identical to those reported in ref. 10.
Acknowledgements
This work was supported by grants from the NIH (GM 40171)
and INSMED Pharmaceuticals, Inc., Richmond, Virginia.
Paper 8/00282G
Received 9th January 1998
Accepted 9th January 1998
References
1 M. G. Low and A. R. Saltiel, Science, 1988, 239, 268.
632
J. Chem. Soc., Perkin Trans. 1, 1998