A new highly diastereoselective synthesis of epi-inositol from D- galactose

Article abstract of DOI:10.1016/S0040-4039(00)00360-9

The inosose derivative 3 was obtained with high stereoselectivity by intramolecular aldol condensation of the aldohexos-5-ulose derivative 2, and it was selectively reduced and debenzylated to give epi-inositol in high yield. The stereochemistry and the preferred conformations of compounds 3-7 were determined through 1D and 2D NMR experiments. (C) 2000 Published by Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00360-9

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 3253–3256
A new highly diastereoselective synthesis of epi-inositol from
D-galactose1
Venerando Pistarà,b Pier Luigi Barili,a Giorgio Catelani,a,∗ Antonino Corsaro,b,∗
Felicia D’Andrea a and Salvatore Fisichella b
aDipartimento di Chimica Bioorganica e Biofarmacia, Università di Pisa, Via Bonanno 33, I-56126, Pisa, Italy
bDipartimento di Scienze Chimiche, Università degli Studi di Catania, Viale A. Doria 6, I-95125, Catania, Italy
Received 27 January 2000; accepted 28 February 2000
Abstract
The inosose derivative 3 was obtained with high stereoselectivity by intramolecular aldol condensation of the
aldohexos-5-ulose derivative 2, and it was selectively reduced and debenzylated to give epi-inositol in high yield.
The stereochemistry and the preferred conformations of compounds 3–7 were determined through 1D and 2D
NMR experiments. © 2000 Published by Elsevier Science Ltd. All rights reserved.
Keywords: intramolecular aldol condensations; carbohydrates; inositols; reduction.
Inositols (hexahydroxycyclohexanes) are an interesting group of carbocyclic sugar analogues which
continue to attract a great deal of attention from synthetic chemists2 because of their various biological
activities.3 Of the nine possible stereoisomers (seven achiral forms and one enantiomeric couple) myo-
inositol is present in all living organisms and readily available; D- and L-chiro-inositols are prepared from
the naturally occurring methyl ethers (+)-pinitol and (−)-quebrachitol, respectively.4 The remaining six
inositols are unnatural and many synthetic routes to them have been developed.5
Having recently developed a general approach to at least one enantiomer of the four diastereoisomeric
aldohexos-5-uloses,6 we have started a project for their transformation into inositols through an intra-
molecular aldol cyclization followed by reduction of the intermediate inosose. Although several synthetic
approaches to inositols from carbohydrates are based on an aldol-type cyclization step,2 for instance in
the case of the mercury(II)-promoted rearrangement of hex-5-enopyranosides, the so called Ferrier(II)
reaction,2,7 few examples of a direct aldol condensation of a 1,5-dicarbonyl hexose are reported, the
most relevant being Kiely’s biomimetic synthesis of myo-inositol from D-xylo-hexos-5-ulose,8 leading,
without isolation and characterization of the intermediate inosose, to a mixture of inositols. This paper
reports the results obtained starting from 2,6-di-O-benzyl-L-arabino-aldohexos-5-ulose (2) which leads
to a highly stereoselective route (Scheme 1) to epi-inositol (6), an unnatural and expensive isomer,
∗
Corresponding authors. E-mail: giocate@farm.unipi.it (G. Catelani), acorsaro@dipchi.unict.it (A. Corsaro)
0040-4039/00/$ - see front matter © 2000 Published by Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00360-9
tetl 16640

3254
which was previously prepared5f in very limited yield starting from myo-inositol by inversion of the
configuration of one –CHOH group through an oxidation with nitric acid, followed by reduction with
sodium amalgam or by catalytic (Pt) hydrogenation.
Scheme 1.
Compound 2 (1.18 mmol), easily obtained from the commercially available methyl β-D-
galactopyranoside 1 through a previously described6a synthetic route, was treated with DBU (0.16
mmol) in toluene (20 ml) at room temperature; a smooth reaction took place leading, after 6 h, to
the complete disappearance of 2 and the formation of a major product (TLC), accompanied only by
unidentified products with Rf=0. After neutralization with Amberlyst® 15, flash chromatography of the
residue gave 2L-2,4-di-O-benzyl-(2,3,5,6/4)-pentahydroxycyclohexanone 39 in 60% yield. The catalytic
debenzylation (10% Pd/C in EtOH) of 3 produced quantitatively the previously unreported meso inosose
4.10 The stereochemistry and the preferred conformer of these two inososes was determined by 1D
and 2D (NOE, COSY and HETCOR) NMR experiments; in particular, inosose 4 having a symmetry
plane through C-1 and C-4, shows in its 1H NMR spectrum only three signals,10 likewise its 13C NMR
spectrum shows signals at 68.97, 74.13 and 79.20 ppm together with the signal for the carbonyl group
(208.84 ppm). Furthermore, the J values (3.4 and 4 Hz) of H-2(6), H-3(5) and H-4 establish that the
preferred conformer has three axial hydroxyl groups. The inosose 3 had the same conformational
situation as evidenced by the analogous values of the vicinal protonic coupling constants and by the
presence of long-range couplings between H-2 and H-6 (J=1.3 Hz) and H-3 and H-5 (J=2.3 Hz).
The cis-axial/equatorial disposition of the oxygenated substituents of the two new chiral centres of 3,
point to a kinetically controlled aldol condensation, that could be ascribed to the preferred conformation
of the intermediate chiral enolate and of the following ring closure transition state. When the aldol
condensation was carried out with 0.005 M NaOH in aqueous methanol the formation of 3 was not
stereospecific, a second inosose being formed as minor product (ratio of about 2:1, overall yield of
70%), thus indicating that the solvent and/or the base promoter play a specific role in the stereochemical
outcome of the reaction. The structure of the second inosose is under investigation.
Inosose 3 was then selectively reduced both by catalytic hydrogenation (Ni–Raney/H2 40 psi, EtOH,
rt 12 h; 92% yield) and with hydride (NaBH4, MeOH, −78°C 1 h; 90% yield) to give, exclusively, the
epi-inositol derivative 511 deriving from the attack of hydrogen or hydride on the less hindered side of
the ketone. The structure of 5 was unequivocally proven by the NMR data of its 1,2,3,5-tetraacetate 712
where the H-6 proton appears as a double doublet (4.46 ppm) with a large Jax/ax (9.61 Hz) coupling with
the adjacent H-1 and H-5 protons; the H-5 proton (5.17 ppm) shows a small Jax/eq because it couples with

3255
the H-4 proton (3.30 Hz) as in the case of the H-3 proton (5.04 ppm), which has a small Jax/eq coupling
with the H-2 (3.23 Hz) and H-4 (3.10 Hz) protons.
Finally, epi-inositol 613 was obtained in nearly quantitative yields by catalytic debenzylation of inositol
derivative 5, or by reduction of inosose 4. The structure of epi-inositol 6 was confirmed by a comparison
of its analytical and spectroscopic data with those reported in the literature.14
In conclusion, our synthetic strategy offers a simple method to synthesize epi-inositol in a highly
stereoselective way and with a good chemical yield, by means of a process which leads to a single
stereoisomer, although it involves the formation of three new chiral centres. We are now engaged in
a program aimed at extending this synthetic scheme to other diastereoisomeric aldohexos-5-uloses,
with various kinds of functionalities, with the scope of establishing a general route to the biologically
interesting cyclitols, starting from readily available D-galactose derivatives.
Acknowledgements
This work was performed in the frame of the national relevance project ‘Chimica dei Composti
Organici di Interesse Biologico’ financed by MURST, Italy (PRIN-97), and the project ‘Metodologie di
Sintesi a basso impatto ambientale’ of Consorzio Interuniversitario Nazionale La Chimica per l’Ambiente
(INCA) granted to a loan of MURST, Italy, in accordance with law no. 488/92.
References
1. Part 12 of the series ‘Rare and complex saccharides from D-galactose and other milk-derived carbohydrates’. For Part 11
see Ref. 15.
2. Hudlicky, T.; Cebulak, M. Cyclitols and Their Derivatives. A Handbook of Physical, Spectral and Synthetic Data; VCH
Weinheim: New York, Cambridge, 1993; Ferrier, R. J.; Midleton, S. Chem. Rev. 1993, 93, 2779–2831; Daljko, P. I.; Sinaÿ,
P. Angew. Chem., Int. Ed. 1999, 38, 773–777.
3. Berridge, M. J. Nature 1989, 341, 197; Billington, D. C. The Inositol Phosphates–Chemical Synthesis and Biological
Significance; VCH Weinheim, 1993.
4. Kiddle, J. J. Chem. Rev. 1995, 95, 2189–2202; Tegge, W.; Ballou, C. E. Proc. Natl. Acad. Sci. USA. 1989, 86, 94–98.
5. (a) For allo, see: Mandel, M.; Hudlicky, T. J. Chem. Soc., Perkin Trans. 1 1993, 741–743; (b) for muco, see: Brammer Jr.,
L. E.; Hudlicky, T. Tetrahedron: Asymmetry 1998, 9, 2011–2014; (c) for neo, see: Riley, A. M.; Jenkins, D. J.; Potter, B. V.
L. Carbohydr. Res. 1998, 314, 277–281; (d) for scyllo, see: Husson, C.; Odier, L.; Vottéro, Ph. J. A. Carbohydr. Res. 1998,
307, 163–165; (e) for cis, see: Angyal, S. J.; Odier, L.; Tate, M. E. Carbohydr. Res. 1995, 266, 143–146; (f) for epi, see:
Posternak, T. Helv. Chim. Acta 1946, 29, 1991–1998.
6. (a) For L-arabino, see: Barili, P. L.; Berti, G.; Catelani, G.; D’Andrea, F. Gazz. Chim. Ital. 1992, 122, 135–142; (b) for D-
xylo, see: Barili, P. L.; Berti, G.; Catelani, G.; D’Andrea, F.; De Rensis, F. Tetrahedron 1997, 53, 8665–8674; (c) for L-lyxo,
see: Barili, P. L.; Berti, G.; Catelani, G.; D’Andrea, F.; De Rensis, F.; Goracci, G. J. Carbohydr. Chem. 1998, 17, 1167–1180;
(d) for L-ribo, see: Barili, P. L.; Bergonzi, M. C.; Berti, G.; Catelani, G.; D’Andrea, F.; De Rensis, F. J. Carbohydr. Chem.
1999, 18, 1037–1049.
7. (a) Ferrier, R. J. J. Chem. Soc., Perkin Trans. 1 1979, 1455–1458; (b) for a recent example, see: Takahashi, H; Kittaka, H;
Ikegami, S. Tetrahedron Lett. 1998, 39, 9703–9706.
8. Kiely, D. E.; Fletcher Jr., H. G. J. Org. Chem. 1969, 34, 1386–1390.
9. 2L-2,4-Di-O-benzyl-(2,3,5,6/4)-pentahydroxycyclohexanone 3: yield 60%, m.p. 110–112°C, white crystals from ethyl
25
acetate (found: C, 67.22; H, 6.44. C20H22O6 requires: C, 67.03; H, 6.29%); [α]D −69.23 (c 0.26, CHCl3); νmax (KBr)
3473, 3452, 3380, 1724, 1625 cm−1; δH (CD3CN, 500 MHz) 4.06 (dd, 1H, J3,4=3.1, J4,5=3.2 Hz, H-4), 4.46 (ddd, 1H,
J5,6=4.3, J3,5=2.3 Hz, H-5), 4.53 (1H, dd, J2,3=3.4, J2,6=1.3 Hz, H-2), 4.55 (ddd, 1H, H-3), 4.58 (dd, 1H, H-6), 4.56 and 4.83
(AB system, 2H, JAB=12 Hz, benzylic CH2), 4.68 and 4.77 (AB system, 2H, JAB=11.5 Hz, benzylic CH2), 7.37–7,46 (m,
10H, phenyl H); δC (CD3CN, 125 MHz) 72.58 (CH2), 73.25 (CH2), 75.96 (C-6), 76.75 (C-3), 76.85 (C-4), 77.46 (C-5),
81.63 (C-2), 126.65–139.07 (phenyl C), 207.52 (C_O).

3256
10. 2,3,5,6/4-Pentahydroxycyclohexanone 4: yield 96%, m.p. 118–120°C, white crystals from ethanol (found: C, 40.47; H,
5.71. C6H10O6 requires: C, 40.43; H, 5.66%); νmax (KBr) 3351, 2924, 1728, 1629 cm−1; δH (DMSO-d6/D2O 200 MHz)
3.96 (dd, 1H, spl. ›3.4 and 4.0 Hz), 4.09 (dd, 2H, spl. ›3.4 and 4.0 Hz), 4.43 (d, 2H, spl. ›4.0 Hz); δC (DMSO-d6 50 MHz)
68.97, 74.13 (2×), 79.20 (2×) and 208.84 (C_O).
11. 1D-2,6-Di-O-benzyl-epi-inositol 5: yield 92%, m.p. 178–180°C, white crystals from ethyl acetate (found: C, 66.73; H, 6.66.
25
C20H24O6 requires: C, 66.65; H, 6.71%); [α]D −271.85 (c 0.94, CHCl3); νmax (KBr) 3390, 2922, 1642, 1451 cm−1
.
12. 1D-1,3,4,5-Tetraacetyl-2,6-di-O-benzyl-epi-inositol 7: yield 98%, syrup (found: C, 63.70; H, 6.18. C28H32O10 requires: C,
63.63; H, 6.10%); [α]D −6.23 (c 0.96, CHCl3); νmax (NaCl) 3478, 2924, 2890, 1739 cm−1; δH (C6D6, 200 MHz) 1.83
25
(s, 3H, CH3), 1.70 (s, 6H, CH3), 1.67 (s, 3H, CH3), 4.16 (ddd, 1H, J1,2=3.23, J2,3=3.1, J2,4=1.2 Hz, H-2), 4.46 (dd, 1H,
J1,6=9.6, J5,6=9.6 Hz, H-6), 4.54 (s, 2H, CH2), 4.61 (s, 2H, CH2), 5.04 (dd, 1H, J3,4=3.3, Hz, H-3), 5.17 (dd, 1H, H-1) 5.33
(dd, 1H, J4,5=3.5 Hz, H-5), 5.92 (ddd, 1H, H-4), 7.05–7.40 (m, 10H, phenyl H); δC (C6D6, 50 MHz) 20.62 (CH3), 20.51
(CH3), 20.34 (CH3 ×2), 68.97 (C-3), 69.52 (C-4), 70.82 (C-5), 73.02 (C-1), 74.89 (CH2), 75.12 (CH2), 75.72 (C-6), 77.08
(C-2), 139.16–139.28 (phenyl C), 169.13, 169.24, 169.57, 170.10 (C_O).
13. epi-Inositol 6: yield 95%, m.p. 280–282°C, white crystals from ethanol (lit.5f 285°C); νmax (KBr) 3410, 3220, 2950, 1638
cm−1 13C NMR data identical to those previously reported.14
.
14. Salazar-Pereda, V.; Martinez-Martinez, F. J.; Contreras, R.; Flores-Parra, A. J. Carbohydr. Chem. 1997, 16, 1479–1507.
15. Barili, P. L.; Catelani, G.; D’Andrea, F.; Puccioni, L. J. Carbohydr. Chem. 2000, 19, 79–91.