Synthesis of amino-sugars using the directed dihydroxylation reaction

Article abstract of DOI:10.1039/a904991f

The synthesis of protected forms of two amino-sugars, talosamine 1 and allosamine 2, is described; the syn, syn stereochemistry at C-2, C-3 and C-4 was controlled by the use of a hydrogen-bonding directed dihydroxylation reaction using stoichiometric and catalytic OsO₄; proof of the relative stereochemistry of the allosamine series was obtained through an X-ray crystal structure of a protected derivative.

Full text of DOI:10.1039/a904991f

Synthesis of amino-sugars using the directed dihydroxylation reaction
Timothy J. Donohoe,* Kevin Blades and Madeleine Helliwell†
Department of Chemistry, The University of Manchester, Oxford Road, Manchester, UK, M13 9PL.
E-mail: t.j.donohoe@man.ac.uk
Received (in Cambridge, UK) 22nd June 1999, Accepted 29th July 1999
The synthesis of protected forms of two amino-sugars,
talosamine 1 and allosamine 2, is described; the syn, syn
stereochemistry at C-2, C-3 and C-4 was controlled by the
use of a hydrogen-bonding directed dihydroxylation reac-
tion using stoichiometric and catalytic OsO4; proof of the
relative stereochemistry of the allosamine series was ob-
tained through an X-ray crystal structure of a protected
derivative.
We then turned our attention to allosamine 2 and proceeded
to prepare allylic amide 9 by a process analogous to that
described earlier (Scheme 2).5‡ The modified conditions
reported by Isobe and co-workers were crucial to success in the
Overman rearrangement.6 The oxidation of 9 (again the major
anomer was isolated and is shown for clarity) proved to be most
interesting. In this case the standard Upjohn reaction was
moderately syn selective for the allosamine isomer (step vi).
Unfortunately the addition of TMEDA to OsO4 only marginally
improved the syn selectivity of the oxidation, (step vii).
Occasionally, our attempts to use TMEDA/OsO4 for directing
the dihydroxylation reaction have been frustrated by either
electrostatic or steric repulsion between the substrate and the
oxidant.7 In such cases we have found that addition of a
monodentate amine (quinuclidine is best) to OsO4 can give
better syn stereoselectivity via a hydrogen bonding process; this
was indeed the case (step viii), and the selectivity attainable was
now high enough to be synthetically useful. The reagent formed
during this reaction is not as effective a hydrogen bond acceptor
as OsO4/TMEDA but can give higher levels of syn stereo-
selectivity, presumably as a consequence of its reduced steric
bulk. One of the advantages of using a monodentate amine
promoter in the dihydroxylation reaction is that the process can
be made catalytic in osmium by using the amine as its
2-Amino-sugars are the key constituents of a variety of
naturally occurring polysaccharides. We are interested in the
synthesis of two of the rarer amino-sugars, talosamine and
allosamine: N-acetyltalosamine has been isolated from both
bovine and ovine cartilage,1 whilst N-acetylallosamine is the
repeat unit in the sugar portion of allosamidin2 (which is an
interesting chitinase inhibitor).
We have recently described that combination of OsO4 with an
amine (usually TMEDA) produces a reagent which dihydrox-
ylates allylic amides: efficient hydrogen bonding between the
oxidant and the allylic amide means that the stereoselectivity
observed is contra-steric and therefore opposite to that obtained
under standard oxidation conditions.3 Examination of talos-
amine and allosamine reveals that they both have syn,syn
stereochemistry at C2–C4 and so are ideally configured for
synthesis via our chemistry.
Our synthesis of N-acetyl talosamine began with tri-O-acetyl-
d-glucal, which was converted into the allylic trichloro-
acetamide 5 using standard chemistry (Scheme 1).4 Compounds
3, 4 and 5 are formed as mixtures of anomers (9+1) and the
major one was separated and is shown for clarity. The
dihydroxylation of 5 was then examined under a variety of
conditions to investigate ways of forming the syn,syn isomer
which corresponds to the talosamine series. As can be observed
from Scheme 1, oxidation under standard Upjohn conditons
(step viii) gave predominantly the syn,anti isomer (this
corresponds to the altrose series of amino sugars) as expected.
However, use of OsO4 and TMEDA at low temperature was
successful in completely reversing the facial bias of this
molecule and provides the syn,syn isomer with complete
1
stereoselectivity (as measured by H NMR spectroscopy, step
ix) and excellent yield. The relative stereochemistry of the
adducts shown in Scheme 1 was proven by correlation to a
compound of known configuration (vide infra). Note that the
products from the stoichiometric and catalytic oxidations are
slightly different; the acidic work-up used with stoichiometric
OsO4 removes the TBDMS group at C-6; the resulting triol was
then peracetylated with Ac2O in order to expedite isolation.
Scheme 1 Reagents and conditions: i, BnOH (1.1 equiv.), BF3·OEt2 (0.1
equiv.), CH2Cl2, 220 °C room temp., 1 h, 98%; ii, K2CO3 (0.5 equiv.),
MeOH–H2O (4+1), room temp., 1 h, 93%; iii, TBDMSCl (1.1 equiv.), Et3N
(2.2 equiv.), CH2Cl2, room temp., 24 h, 74%; iv, PPh3 (4.0 equiv.), PhCO2H
(4.0 equiv.), DEAD (4.0 equiv.), THF, 0 °C, 2 h, 88%; v, NaOMe (0.5
equiv.), MeOH, room temp., 86%; vi, DBU (1.2 equiv.) CCl3CN (1.3
equiv.), CH2Cl2, 220 °C–room temp., 97%; vii, xylene, cat. K2CO3, D, 18
h, 93%; viii, NMO (1.5 equiv.), cat. OsO4, acetone–H2O (4:1), room temp.,
77%; ix, TMEDA (1.05 equiv.), OsO4 (1.05 equiv.), CH2Cl2, 278 °C–room
temp., then MeOH–HCl, then Ac2O, 89%.
† Author to whom correspondence regarding the crystallography should be
addressed.
Chem. Commun., 1999, 1733–1734
1733

Scheme 3 Reagents and conditions: i, NaOH, EtOH, 10 min; ii, Ac2O, cat.
DMAP, pyridine, 24 h, 82%; iii, NaOH, EtOH, 10 min; iv Ac2O, cat.
DMAP, pyridine, 24 h, 86%.
introduced a novel oxidising system, using quinudidine N-
oxide, which is capable of dihydroxylating certain allylic
amides with hydrogen-bonding control∑ and which uses cata-
lytic OsO4.
We wish to thank the EPSRC (K. B.) and Zeneca Pharmaceu-
ticals (Strategic Research Fund) for financial support.
Notes and references
Scheme 2 Reagents and conditions: i, MeOH (1.1 equiv.), BF3·OEt2 (0.5
equiv.), toluene, 220 °C–room temp., 1 h, 89%; ii, K2CO3 (0.5 equiv.),
MeOH–H2O (4+1), room temp., 1 h, 92%; iii, TBDMSCl (1.1 equiv.), Et3N
(2.2 equiv.), CH2Cl2, room temp., 24 h, 91%; iv, DBU (1.2 equiv.), CCl3CN
(1.3 equiv.), CH2Cl2, 220 °C–room temp., 98%; v, Ph2O, cat. K2CO3,
195 °C, 3 h, 64%; vi, NMO (1.5 equiv.), cat. OsO4, acetone–H2O (4+1),
room temp., 75%; vii, TMEDA (1.05 equiv.), OsO4 (1.05 equiv.), CH2Cl2,
278 °C–room temp., then MeOH–HCl, then Ac2O, 92%; viii, quinuclidine
(1.1 equiv.), OsO4 (1.05 equiv.), CH2Cl2, 278 °C–room temp., then
MeOH–HCl, then Ac2O, 90%; ix, quinuclidine N-oxide (2.0 equiv.), cat.
OsO4, CH2Cl2, room temp., 80%; x, Me3NO·2H2O (1.5 equiv.), cat. OsO4,
CH2Cl2, room temp., 87%.
‡ A similar strategy has been adopted previously for the synthesis of
allosamine. However, the reported route is troubled by both a low yielding
Overman rearrangement (30%) and oxidation with RuO4 (50%).
§ Crystal data for C18H20NO7Cl3 12: M = 468.70, orthorhombic, a =
12.862(7), b = 27.763(6), c = 6.007(4) Å3, T = 296 K, space group P2,2,2,
(no. 19), Z = 4, m(Cu-Ka) = 4.225 mm21, 2179 independent reflections
which were used in all calculations. The final wR(F2) was 0.1554 (all data).
R(F) was 0.0499 using 1703 reflections with I > 2s(I). The structure was
solved using direct methods and developed using difference Fourier
techniques, then refined by full matrix least-squares on F2. CCDC
182/1357.
¶ Selected data for 13: [a]D +69 (c 5.33, EtOH); dH(300 MHz; CDCl3) 5.83
(d, 1H, J 9.3), 5.51 (dd, 1H, J 3.0, 3.5), 4.97 (dd, 1H, J 3.0, 10.2), 4.68 (d,
1H, J 4.4), 4.48 (ddd, 1H, J 3.5, 4.4, 9.3), 4.30-4.15 (m, 3H), 3.42 (s, 3H),
2.16 (s, 3H), 2.09 (s, 3H), 1.99 (s, 3H) 1.97 (s, 3H); dC (75 MHz; CDCl3)
170.6, 170.4, 169.2, 169.0, 97.6, 68.3, 66.0, 63.0, 62.1, 55.7, 47.4, 23.0,
20.9, 20.6, 20.4. For 14: [a]D +72 (c 0.80, CHCl3); dH(300 MHz; CDCl3)
6.39 (d, 1H, J 9.6), 5.40 (m, 1H), 5.34 (dd, 1H, J 3.5, 4.5), 4.91 (d, 1H, J 1.1),
4.68 (d, 1H, J 11.6), 4.54 (d, 1H, J 11.6), 4.48 (dddd, 1H, J 9.6, 4.7, 1.1, 1.1),
4.27 (ddd, 1H, J 7.1, 6.2, 1.3), 4.12 (ABX, 2H, JAB 11.1, JAX 6.2 JBX 7.1),
2.18 (s, 3H), 2.07 (s, 3H), 2.03 (s, 3H) 1.99 (s, 3H); dC(75 MHz; CDCl3)
170.3, 169.4, 169.4, 168.9, 136.2, 128.5, 128.1, 128.0, 99.1, 69.8, 67.7,
66.5, 64.7, 61.7, 48.5, 23.4, 20.6, 20.5.
corresponding N-oxide (with sufficient water added to allow the
catalytic cycle to turn over). So, the use of quinuclidine N-oxide
(2 equiv.)8 and catalytic OsO4 (5 mol%) with approximately 4
equiv. of water, also gave an excellent level of syn stereo-
selectivity (step ix). This novel oxidising system gave better
stereoselectivity than that observed by using trimethylamine N-
oxide as a reoxidant (step x).9
The relative stereochemistry of the allosamine series of
compounds was proven by an X-ray crystal structure on the
benzylidine protected derivative 12 (Fig. 1).§
∑ Oxidation of 5 with quinuclidine N-oxide under the conditions described
in Scheme 2 gave a 1+1 mixture of stereoisomers. However, other allylic
amides do give syn selective results using this oxidant; details will be
reported in due course.
1 M. J. Crumpton, Nature, 1957, 180, 605; R. Heyworth and P. G. Walker,
Proc. Int. Congr. Biochem. 4th, 1958, 1, 7.
2 See, D. J. Kassab and B. Ganem, J. Am. Chem. Soc., 1999, 64, 1782 and
references therein.
3 T. J. Donohoe, K. Blades, P. R. Moore, J. J. G. Winter, M. Helliwell and
G. Stemp, J. Org. Chem., 1999, 64, 2980; T. J. Donohoe, P. R. Moore,
M. J. Waring and N. J. Newcombe, Tetrahedron Lett., 1997, 38, 5027;
for early studies in this area see: H. C. Kolb, M. S. VanNieuwenhze and
K. B. Sharpless, Chem. Rev., 1994, 94, 2483.
4 K. Takeda, E. Kaji, Y. Konda, N. Sato, H. Nakamura, N. Miya, A.
Morizane, Y. Yanagisawa, A. Akiyama, S. Zen and Y. Harigaya,
Tetrahedron Lett., 1992, 33, 7145.
5 T. Sugai, H. Okazaki, A. Kuboki and H. Ohta, Bull. Chem. Soc. Jpn.,
1997, 70, 2535.
6 T. Nishikawa, M. Asai, N. Ohyabu and M. Isobe, J. Org. Chem., 1998,
63, 188.
Fig. 1 X-Ray structure of 12.
As amino-sugars frequently occur in nature as their N-acetyl
derivatives, we removed the trichloroacetyl group and added an
acetyl functionality. This was accomplished readily in a two
stage, one pot process (Scheme 3).¶ Compound 14 prepared by
this route is known in the literature10 and displayed character-
istics identical to those previously reported (1H NMR, [a]D);
this correlation proves the stereochemistry of the talosamine
series described in Scheme 1.
7 T. J. Donohoe, K. Blades, M. Helliwell, M. J. Waring, and N. J.
Newcombe, Tetrahedron Lett., 1998, 39, 8755.
To conclude, we have prepared protected forms of two
naturally occurring and rare amino-sugars, talosamine and
allosamine: both compounds are available in high yield and
with excellent levels of stereoselectivity. These derivatives are
ideally configured for further elaboration into polysaccharides
and work is continuing in this direction. Moreover, we have
8 I. A. O’Neil, J. Y. Q. Lai and D. Wynn, Chem. Commun., 1999, 59.
9 G. Poli, Tetrahedron Lett., 1989, 30, 7385; T. J. Donohoe, R. Garg and
P. R. Moore, Tetrahedron Lett., 1996, 37, 3407; T. J. Donohoe, P. R.
Moore and R. L. Beddoes, J. Chem. Soc., Perkin Trans. 1, 1997, 43.
10 A. Banaszek and W. Karpiesiuk, Carbohydr. Res., 1994, 251, 233.
Communication 9/04991F
1734
Chem. Commun., 1999, 1733–1734