Dispiroketals in synthesis. Part 23. A new route to (+)-D-conduritol B from myo-inositol

Article abstract of DOI:10.1039/a700422b

Using (2S,2′S)-2,2′-diphenyl-6,6′-bi(3,4-dihydro-2H-pyran) to effect a simultaneous protection-resolution of a myo-inositol derivative, a new synthesis of (+)-D-conduritol B has been achieved.

Full text of DOI:10.1039/a700422b

View Online / Journal Homepage / Table of Contents for this issue
Dispiroketals in synthesis. Part 23.1 A new route to (ϩ)-D-conduritol
B from myo-inositol
Jean E. Innes, Paul J. Edwards and Steven V. Ley*
Department of Chemistry, Lensfield Road, Cambridge CB2 1EW, UK
U sing
(2S,2ЈS)-2,2Ј-diphenyl-6,6Ј-bi(3,4-dihydro-2H -
TBDPS protection was chosen to enhance the solubility of
the resulting diol in chloroform, the solvent used for the sub-
sequent dispiroketalisation. Derivatives with different protec-
tion had given poor results due to the insolubility of the sub-
strate; diol 5, however, was completely soluble in chloroform
and reacted with (2S,2ЈS)-2,2Ј-diphenyl-6,6Ј-bi(3,4-dihydro-
2H-pyran) 2 to give the corresponding dispiroketal 6 (Scheme 2).
pyran) to effect a simultaneous protection–resolution of a
myo-inositol derivative, a new synthesis of (؉)-D -conduritol
B has been achieved.
Conduritols and their derivatives have attracted considerable
attention from synthetic chemists owing to their diverse bio-
logical activity.2–6 Conduritol B is the least accessible of the
chiral conduritols because it is the only one which does not
possess a cis-vicinal diol pair, thus making it less amenable to
the two most common synthetic strategies, the action of Pseudo-
monas putida on aromatic precursors7 and the asymmetric
dihydroxylation reaction of Sharpless.8 Five different asym-
metric approaches to conduritol B have been reported.9 We
describe here a new route to (ϩ)--conduritol B 1 using
OH
O
O
RO
O
BnO
OH
i
iv
5
Ph
Ph
OR
O
O
OBn
O
O
O
O
O
HO
Ph
Ph
Ph
O
O
Ph
HO
OH
6 R = TBDPS
7 R = H
9
ii
iii
OH
8 R = Bn
v, vi
1
2
(2S,2ЈS)-2,2Ј-diphenyl-6,6Ј-bi(3,4-dihydro-2H-pyran) 210 to
effect the simultaneous resolution and protection of a myo-
inositol-derived diol, a strategy which has the potential to offer
equal access to both enantiomers.
BnO
RO
RO
vii
Ph
O
O
OBn
OR
O
We have demonstrated previously the use of dispiroketals in
synthesis, especially in the carbohydrate area and in the gener-
ation of enantiomerically-enriched compounds from racemic
or meso diols.10 Useful results continue to be produced as new
generations of substituted bi(dihydropyran)s are developed and
becomecommerciallyavailable.†Furthermore, related chemistry
using other 1,2-diacetals, which we introduced some years ago,
is beginning to impact upon the area.11
Racemic 1,2:4,5-di-O-cyclohexylidene-myo-inositol 3, which
is available in one step from myo-inositol,12 is the starting point
for the current synthesis. Protection of 3 with tert-butyl-
diphenylsilyl (TBDPS) groups to give 4 followed by cleavage of
the trans-cyclohexylidene acetal results in diol 5 (Scheme 1).
O
OR
1 R = H
11 R = OAc
Ph
viii
10
Scheme 2 Reagents and conditions: i, 2 (0.51 equiv.), PPh3ؒHBr (cat.),
CHCl3, room temp., 96 h, 35% (70% based on theoretical maximum for
a resolution); ii, Bu4NF (2 equiv.), THF, 60 ЊC, 10 h, quant.; iii, NaH
(2.5 equiv.), DMF, 0 ЊC, 1 h, then Bu4NI (cat.), BnBr (2.4 equiv.), room
temp., 36 h, 61%; iv, 1-sulfanylpropan-2-ol (3 equiv.), BF3ؒOEt2 (1
equiv.), CHCl3, 45 ЊC, 40 min, quant. (ref. 14); v, thiocarbonyldiimid-
azole (2 equiv.), DMAP (cat.), PhMe, reflux, 16 h, 97%; vi, P(OMe)3,
reflux, 5 h, quant.; vii, Li (ca. 10 equiv.), NH3(1), Et2O, Ϫ78 ЊC, 2 h,
quant.; vii, Ac2O, pyridine, room temp. 16 h, quant.
This reaction constitutes a simultaneous resolution and pro-
tection of the racemic diol 5. The complete diastereoselectivity
of the reaction (the product 6 is a single diastereoisomer by 1H
and 13C NMR spectroscopy) is a consequence of a chirality
match, only one diastereoisomer of 6 allowing an all-chair
structure with all substituents equatorial and full anomeric
stabilisation at the newly-formed spiro centres (Fig. 1).10
O
O
O
RO
O
TBDPSO
HO
ii
OR
OTBDPS
O
O
OH
The resolution–protection proceeds in a good 70% yield
(based on the theoretical maximum for a resolution). The other
enantiomer of diol 5 does not react with 2 because its chirality
is mis-matched; it is readily removed by chromatography, an
advantage over conventional resolutions which frequently
involve a tedious separation of diastereoisomers. As the bulky
TBDPS groups prevented further transformations later in the
synthesis they were removed to afford diol 7, which was di-
benzylated to provide the less sterically encumbered derivative
8. It should be noted that it was not possible to have the benzyl
5
3 R = H
4 R = TBDPS
i
Scheme 1 Reagents and conditions: i, TBDPSCl (10 equiv.), pyridine
(excess), reflux, 16 h, 72%; ii, ethylene glycol (1 equiv.), toluene-p-
sulfonic acid (cat.), CHCl3, 50 ЊC, 1.5 h, 76%
† Unsubstituted diene 6,6Ј-bi(3,4-dihydro-2H-pyran) is available from
the Aldrich Chemical Company.
J. Chem. Soc., Perkin Trans. 1, 1997
795

View Online
CDCl3) [1.07 (9 H, s) and 1.08 (9 H, s), 2 × But], [1.12–1.41
Ph
(14 H, m), 1.43–1.57 (2 H, m), 1.61–1.72 (2 H, m), 1.74–1.83
(2 H, m) and 1.89–2.05 (2 H, m), 11 × CH2], 3.35 (1 H, t, J 9.9,
5-H), 3.56 (1 H, t, J 4.1, 2-H), 3.66–3.73 (2 H, m, two out of
1-H or 3-H and 4-H or 6-H), 3.89 (1 H, t, J 9.9, one of 4-H
or 6-H), 3.98 (1 H, dd, J 9.9, 4.1, one of 1-H or 3-H), [4.71–4.76
(1 H, m) and 4.86–4.92 (1 H, m), 2 × dispoke CHPh], [7.15–
7.45 (22 H, m), 7.57–7.60 (2 H, m), 7.73–7.77 (2 H, m) and
7.82–7.89 (4 H, m), 30 × dispoke and silyl Ar-H]; δC(100 MHz;
CDCl3; asterisk indicates overlapping signals as judged by
relative intensities) 17.4 and 18.7 [2 × SiC(CH3)3], 19.3, 19.7,
23.7, 23.9 and 25.1 (5 × CH2), 26.9 and 27.2 [2 × SiC(CH3)3],
27.8, 28.1, 31.6, 33.4, 35.1 and 37.3 (6 × CH2), 67.5, 68.9, 70.1,
70.2, 70.5, 75.4, 75.9 and 79.5 (1-C, 2-C, 3-C, 4-C, 5-C, 6-C
and 2 × dispoke CHPh), 96.9* (dispoke acetal C), 109.4
(cyclohexylidene acetal C), 125.5, 126.0, 126.8, 126.9, 127.0,
127.1, 127.3, 127.5, 128.1*, 128.2, 128.9, 129.4 and 129.7 (dis-
poke ortho-, meta- and para-C and silyl meta- and para-C),
131.9*, 133.8 and 134.2 (silyl ipso-C), 135.8, 136.2* and 137.3
(silyl ortho-C), 143.4 and 144.1 (dispoke ipso-C); m/z (ϩFAB)
1078 (M ϩ Na)ϩ, 840 (M ϩ Na Ϫ TBDPS)ϩ, 661, 603, 563, 485,
407, 319, 253 [HRMS (ϩFAB) on (M ϩ Na)ϩ: found,
1077.5167. C66H78O8Si2Na requires 1077.5133]; and (with
20% EtO–light petroleum) unreacted starting material (376
mg, 33%), spectroscopically identical to that isolated
previously.15
O
O
TBDPSO
OTBDPS
O
O
O
O
Ph
6
Fig. 1
protecting groups in place from the beginning since they did not
have sufficient solubilising effect to allow the dispiroketalisation
to proceed. Cleavage of the remaining cyclohexylidene group 13
in 8 gave diol 9 which was converted, according to the Corey–
Winter procedure,14 to alkene 10. Finally, treatment with lith-
ium in liquid ammonia achieved global deprotection to afford
the target molecule 1 in quantitative yield (purification on
reversed phase silica gel). Conversion of 1 to tetra-O-acetyl-
conduritol B 11 gave a less polar material which was easier to
handle. While the melting points and the proton and carbon
NMR spectra of 11 were in accordance with the literature,9a,b
the optical rotation {[α]D25 ϩ70.0 (c 0.85, CHCl3)} was far too low
compared to literature values {for the (Ϫ)-isomer, [α]D20 Ϫ176.8 (c
1.18, CHCl3)9a and [α]D20 Ϫ172.4 (c 1.2, CHCl3)9b}. A second
purification by sublimation did not alter the value of the
rotation. For this reason a sample of racemic conduritol B‡ was
peracetylated and chiral GC was found to give clear separation
of the enantiomers. When our synthetic material was compared
with the racemic material using this technique only one enan-
tiomer was detectable. A double-injection experiment proved
that our synthetic material consisted of a single enantiomer of
conduritol B.
Acknowledgements
We thank Glaxo-Wellcome for a Research Studentship (to
J. E. I.), Dr Duncan Holmes (Glaxo-Wellcome) for useful
discussions and BP and Ciba for research endowments (to
S. V. L.).
We estimate the enantiomeric excess to be >99% by this
method. The discrepancy between our value for the optical
rotation and the literature is attributed either to a strongly
rotating impurity present in very small amount, to concen-
tration effects or to incorrect determination of the literature
values. No chiral GC studies were described in these cases.9a,b
References
1 Part 22: S. V. Ley, S. Mio and B. Meseguer, Synlett, 1996, 791.
2 D. C. Billington, F. Perron-Sierra, I. Picard, S. Beaubras, J. Duhalt,
J. Espinal and S. Challal, Bioorg. Med. Chem. Lett., 1994, 4, 2307.
3 G. Legler, Methods Enzymol., 1977, 46, 368.
4 G. Legler and M. Herrchen, FEBS Lett., 1981, 135, 139.
5 G. Legler and R. Bollhagen, Carbohydr. Res., 1992, 233, 113.
6 Z.-X. Guo, A. H. Haines, S. M. Pyke, S. G. Pyke and R. J. K. Taylor,
Carbohydr. Res., 1994, 264, 147.
Experimental
7 See, for example; (a) S. V. Ley, M. Parra, A. J. Redgrave and
F. Sternfeld, Tetrahedron, 1990, 46, 4995; (b) T. Hudlicky, H. Luna,
H. F. Olivio, C. Andersen, T. Nugent and J. D. Price, J. Chem. Soc.,
Perkin Trans. 1, 1991, 2907.
8 See, for example, R. Angelaud and Y. Landais, J. Org. Chem., 1996,
61, 5020.
9 (a) H. Paulsen, W. Röben and F. R. Heiker, Chem. Ber., 1981, 114,
3342; (b) C. Le Drain, J.-P. Vionnet and P. Vogel, Helv. Chim. Acta,
1990, 73, 161; (c) T. Akiyama, H. Shima and S. Ozaki, Tetrahedron
Lett., 1991, 32, 5593; (d) L. Yu, R. Cabrera, J. Ramirez,
V. A. Malinovskii, K. Brew and P. G. Wang, Tetrahedron Lett., 1995,
36, 2897; (e) Two almost identical approaches; N. Chida,
M. Ohtsuka, K. Nakazawa and S. Owaga, J. Chem. Soc., Chem.
Commun., 1989, 436; K.-I. Sato, M. Bokura and M. Taniguchi, Bull.
Chem. Soc. Jpn., 1994, 67, 1633.
10 For a review of the synthesis and chemistry of dispiroketals, see
S. V. Ley, R. Downham, P. J. Edwards, J. E. Innes and M. Woods,
Contemp. Org. Synth., 1995, 365.
11 N. L. Douglas, S. V. Ley, H. M. I. Osborn, D. R. Owen, H. W. M.
Priepke and S. L. Warriner, Synlett, 1996, 793 and references cited
therein.
12 C. Jiang and D. C. Baker, J. Carbohydr. Chem., 1986, 5, 615.
13 K. S. Bruzik and M.-D. Tsai, J. Am. Chem. Soc., 1992, 114, 6361.
14 E. J. Corey and R. A. E. Winter, J. Am. Chem. Soc., 1963, 85, 2677.
15 J. E. Innes, Ph.D. Thesis, University of Cambridge, 1996.
Preparation of 1-L-(2ЈS,2ЉS,6ЈS,6ЉS)-1,4-di-O-(tert-butyl-
diphenylsilyl)-2,3-O-cyclohexylidene-5,6-O-[6Ј,6Љ-diphenyl-3Ј,3Љ,
4Ј,4Љ,5Ј,5Љ,6Ј,6Љ-octahydro-2Ј,2Љ-bi(2H-pyran-2Ј,2Љ-diyl)]-myo-
inositol 6
Diol 5 (1.14 g, 1.54 mmol) and diene 2 (225 mg, 0.708 mmol,
0.51 eq.) were dried by azeotroping with anhydrous toluene.
The crystalline residue was dissolved in chloroform (freshly dis-
tilled from calcium hydride), triphenylphosphonium bromide (a
few crystals, catalytic) was introduced and the solution was
stirred at room temp. over 72 h, with addition of a further
portion of triphenylphosphonium bromide after each 24 h
period. Silica gel (ca. 10 g) was then added, the solvent was
removed in vacuo and the resulting powder was dry-loaded onto
a packed silica gel column. Flash column chromatography with
gradient elution (compounds listed in order of elution with
their eluent) gave (with 2% Et2O–light petroleum) the dispiroke-
talised myo-inositol 6 (570 mg, 70% based on theoretical maxi-
mum 50% yield of one diastereoisomer) as a white foam; [α]D24
Ϫ13.2 (c 1.03, CHCl3) (Found: C, 75.03; H, 7.44. C66H78O8Si2
requires C, 75.10; H, 7.45%); νmax(CHCl3)/cmϪ1 3153, 3028,
2933, 2856, 1463, 1379, 1162, 1135, 1104, 1064; δH (500 MHz;
‡ We are indebted to Professor R. J. K. Taylor, University of York, Dr
A. H. Haines, University of East Anglia and Dr H. A. J. Carless, Uni-
versity College, London for supplying us with samples of racemic con-
duritol B for the chiral GC studies.
Paper 7/00422B
Received 17th January 1997
Accepted 27th January 1997
796
J. Chem. Soc., Perkin Trans. 1, 1997