Synthesis of aza-C-galacto disaccharides from C1-substituted galactals

Article abstract of DOI:10.1039/b101598m

C1-substituted galactals are transformed via the double reductive amination on their 1,5-diketone derivatives, to aza-β-C-galacto-disaccharides.

Full text of DOI:10.1039/b101598m

Synthesis of aza-C-galacto disaccharides from C1-substituted galactals
Xuhong Cheng, Govindaraj Kumaran and David R. Mootoo*
Department of Chemistry, Hunter College/CUNY, 695 Park Avenue, New York, New York 10021, USA.
E-mail: dmootoo@shiva.hunter.cuny.edu
Received (in Corvallis, OR, USA) 15th February 2001, Accepted 13th March 2001
First published as an Advance Article on the web 17th April 2001
C1-substituted galactals are transformed via the double
reductive amination on their 1,5-diketone derivatives, to
aza-b-C-galacto-disaccharides.
Interest in polyhydroxylated azaheterocycles as biochemical
tools and therapeutic agents has been extensively documented.1
Attention has focused on their potent activity as glycosidase
inhibitors.1,2 Guidelines for the design of glycone specific
glycosidase inhibitors are well known, but models for aglycone
specific inhibitors,2 or for mimetics associated with other
carbohydrate mechanisms are not as well developed.3 Struc-
tures with a high degree of structural complexity, for example as
found in disaccharide analogues, are of interest as probes of
recognition specificity.2a,b Of these, C-disaccharides (e.g. 1, 2)
have the added benefit of stability towards chemical and
enzymic hydrolysis. We and others have shown that C1-
substituted glycals 3 are attractive precursors to C-disaccharides
Scheme 1
1.4–6 Herein, we illustrate the versatility of such C1 substituted
glycals, by the synthesis of the novel, biologically interesting
addition of monosaccharide lithio acetylides to pyranolactones.
aza-b-C-galactosides 4–6. N-linked lipid iminocyclitols related
to 4 have recently shown potential as inhibitors of gp-
120/galactosylceramide binding.3c Azasugar 5 is an analog of
the ubiquitous lactose subunit and 6 is a potential mimetic of a
recently discovered selectin antagonist.7
While applicable to 1 ? 6 type disaccharide systems, this type
of organometallic coupling strategy is not generally applicable
to methylene linked structures (e.g. 1 ? 2, 1 ? 3 and 1 ? 4
linked disaccharide analogues). By comparison, the diketones
in our approach are obtained through dihydroxylation of C1-
substituted glycals (e.g. 10). Such glycals may be obtained
through our oxocarbenium ion cyclization methodology, start-
ing from the ester derived from the 1-thio-1,2-isopropylidene
acetal (TIA)-alcohol 8 and the acid partner 9,4 or via other
protocols.6 The esterification reaction used for the ‘glycone’–
‘aglycone’ coupling is experimentally straightforward and
allows access to C1-pyrano substituted glycals (and hence C-
disaccharide derivatives) with intersaccharide linkers of any
length.
The methodology was first applied to the galactal 11
(Scheme 2). Dihydroxylation of 11 with osmium tetroxide–
NMNO proceeded with complete a-selectivity to provide lactol
12 in 80% yield. Selective acetylation of 12 and PCC oxidation
of the monoacetate 13 led to the diketone 14 in 78% overall
yield from 12. Treatment of 14 with 1.5 equiv. of NaCNBH3,
1.2 equiv. of ammonium formate in anhydrous methanol in the
presence of freshly activated 4 Å molecular sieves afforded the
We envisaged a route to aza-C-disaccharides which is based
on the stereoselective double reductive amination on a complex
diketone 7 (Scheme 1).8 Our optimism was guided by several
examples of highly stereoselective double reductive aminations
in polyhydroxylated dicarbonyl substrates.9 An attractive aspect
of this approach is the introduction of the amine in a single
operation at a relatively late stage in the synthesis, thereby
reducing protecting group inefficiencies. In addition the
synthetic precursors are C1-substituted glycals which may be
assembled via convergent syntheses, and therefore appropriate
for complex disaccharide structures. Prompted by a recent
report of a similar double reductive amination strategy to aza-b-
(1 ? 6)-C-disaccharides,10 we describe a preliminary account
of our results. In the former case, the diketone substrates were
prepared from ‘disaccharide’ ketal precursors, obtained via the
Scheme 2 (i) OsO4, NMNO, acetone, 80%; (ii) Ac2O, DMAP, EtOAc, 95%;
(iii) PCC, CH2Cl2, NaOAc; Celite, 4 Å MS, 82%; (iv) NaCNBH3,
NH4HCO2, 4 Å MS, anhyd. MeOH, 72%; (v) NaOMe, MeOH; (vi) Bu4NF,
THF; (vii) HCl, MeOH, 58% over three steps.
DOI: 10.1039/b101598m
Chem. Commun., 2001, 811–812
This journal is © The Royal Society of Chemistry 2001
811

herein is expected to provide access to a wide variety of aza-b-
C-glycosides. These directions are currently being explored.
We thank the National Institutes of Health (NIH), General
Medical Sciences (GM 57865) for their support of this research.
‘Research Centers in Minority Institutions’ award RR-03037
from the National Center for Research Resources of the NIH,
which supports the infrastructure (and instrumentation) of the
Chemistry Department at Hunter, is also acknowledged.
Notes and references
1 For recent reviews: (a) N. Asano, R. J. Nash, R. J. Molyneux and
G. W. J. Fleet, Tetrahedron: Asymmetry, 2000, 11, 1645; (b) P. Sears
and C. H. Wong, Angew. Chem., Int. Ed. Eng., 1999, 38, 2300.
2 (a) T. D. Heightman and A. T. Vasella, Angew. Chem., Int. Ed., 1999,
38, 750; (b) L. A. G. M. van den Broek, in Carbohydrates in Drug
Design, ed. Z. J. Witczak and K. A. Nieforth, Marcel Dekker, Inc., New
York, 1997, pp. 471–493; (c) G. Legler, in Carbohydrate Mimics:
Concepts and Methods, ed Y. Chapleur, Wiley-VCH, New York, 1998,
pp. 461-490; (d) B. Ganem, Acc. Chem. Res., 1996, 29, 340.
3 (a) C.-H. Wong, R. L Halcomb, Y. Ichikawa and T. Kajimoto, Angew.
Chem., Int. Ed. Engl., 1995, 34, 521; (b) T. D. Butters, L. A. G. M. van
den Broek, G. W. J. Fleet, T. M. Krulle, M. R. Wormald, R. A. Dwek
and F. M. Platt, Tetrahedron: Asymmetry, 2000, 11, 113; (c) K. T.
Weber, D. Hammache, J. Fantini and B. Ganem, Biorg. Med. Chem.
Lett., 2000, 10, 1011.
Scheme 3 (i) OsO4, NMNO, acetone, 17a (81%), or K3FeCN6, DBU,
acetone, 17b (83%); (ii) BnBr, AgOTf, CH2Cl2, 18a (87%), 18b (62%); (iii)
NaBH4, EtOH; (iv) Swern’s Ox. 19a (60%), 19b (64%), over two steps; (v)
NaCNBH3, NH4HCO2, 4 Å MS, anhyd. MeOH, 20a (72%), 20b (68%); (vi)
Bu4NF, THF; (vii) HCl, MeOH; (viii) H2, Pd/C, EtOH, HCOOH, 5 (67%),
6 (60%) over three steps.
4 N. Khan, X. Cheng and D. R. Mootoo, J. Am. Chem. Soc., 1999, 121,
4918.
5 T. Eisele, M. Ishida, G. Hummel and R. Schmidt, Liebigs Ann., 1995,
2113.
6 M. H. D. Postema, D. Calimente, L. Liu and T. L. Behrmann, J. Org.
Chem., 2000, 65, 6061.
7 K. Hiruma, T. Kajimoto, G. Weitz-Schmidt, I. Ollmann and C.-H.
Wong, J. Am. Chem. Soc., 1996, 118, 9265.
8 For other approaches to aza-C-disaccharides: (a) B. A. Johns, Y. T. Pan,
A. D. Elbein and C. R. Johnson, J. Am. Chem. Soc., 1997, 119, 4856; (b)
C. Marquis, S. Picasso and P. Vogel, Synthesis, 1999, 1441; (c) O. R.
Martin and O. M. Saavedra, J. Org. Chem., 1996, 61, 6987; O. R.
Martin, in Carbohydrate Mimics: Concepts and Methods, ed Y.
Chapleur, Wiley-VCH, New York, 1998, pp. 259–282.
9 For other stereoselective double reductive aminations on dicarbonyl
substrates: (a) K. Abe, T. Okumura, T. Tsugoshi and N. Nakamura,
Synthesis, 1984, 597; (b) W. Zou and W. A. Szarek, Carbohydr. Res.,
1993, 242, 311; (c) E. W. Baxter and A. B. Reitz, J. Org. Chem., 1994,
59, 3175; (d) H. Zhao and D. R. Mootoo, J. Org. Chem., 1996, 61,
6762.
10 M. A. Leeuwenburgh, S. Picasso, H. S. Ovekleeft, G. A. van der Marel,
P. Vogel and J. H. van Boom, Eur. J. Org. Chem., 1999, 1185.
11 Anhydrous conditions were required for optimal yields in the reductive
amination reactions. See ref. 8c for the same observation.
12 The coupling constants are similar to those expected for 3,4-O-
isopropylidene-b-O-galactosides. For example, see: P. L. Barili, G.
Catelani, F. D’Andrea and E. Mastrorilli, J. Carbohydr. Chem., 1997,
16, 1001.
aza-b-C-galactoside 15 as a single stereoisomer in 72% yield.11
The stereochemistry of 15 was assigned on the basis of J values
(J1,2 = 9.9, J2,3 = 7.7, J3,4 = 5.1, J4,5 = 2.2 Hz) and
observation of 1% NOE effects between H1, and H3 and H5
respectively.12 Removal of protecting groups in 15 under
standard conditions provided the desired aza-b-C-galactoside 4
in 58% overall yield from 15.13
Dihydroxylation of galactal 16a (Scheme 3) under the
aforementioned conditions produced a single ketal 17a in 80%
yield. These conditions led to lower yields for the reaction of the
gluco linked galactal 16b. However use of potassium ferricya-
nide instead of NMNO as co-oxidant in the presence of DBU led
to high yield of the desired hydroxy ketal 17b.14 Although the
secondary alcohols in 17a,b could be selectively converted to
their acetate derivatives, this protecting group proved to be
incompatible with subsequent steps. Compounds 17a,b were
therefore converted to the respective benzyl ethers 18a,b.
Unlike the case for ketal 12, direct oxidation of 18a,b to the
requisite diketones was not successful. Diketones 19a,b were
eventually obtained via a two step reduction–oxidation proce-
dure on 18a,b. Treatment of 19a,b under reductive amination
conditions provided the C-aza derivatives 20a,b in 72 and 64%
yields respectively. The stereochemistry in 20a,b was assigned
from NOESY experiments in a similar fashion as described for
15. Standard deprotection procedures on 20a and 20b provided
the desired b-aza-C-disaccharides 5 and 6 respectively.15
The high stereoselectivity of these double reductive amina-
tions is consistent with the model developed by Stevens for the
hydride reduction of six membered cyclic iminium ions.16 Thus
preferred a-face reduction on imium ions like 21 and 22 would
13 For 4: clear oil; 1H NMR (500 MHz, CD3OD) d 0.90 (t, J = 7.0 Hz, 3H),
1.30 (m, 28H), 1.50 (m, 2H), 1.62 (m, 1H), 1.94 (m, 1H), 2.78 (m, 1H),
3.15 (t, J = 6.5 Hz, 1H), 3.42 (dd, J = 3.0, 8.5 Hz, 1H), 3.62 (t, J = 10.0
Hz, 1H), 3.78 (m, 2H), 4.00 (br s, 1H). 13C NMR (75 MHz, CD3OD) d
14.6, 23.9, 26.8, 30.6, 30.7, 30.9, 31.1, 32.0, 32.2, 60.9, 61.2, 61.5, 68.9,
71.7, 75.6. FAB HRMS calcd for C23H48NO4 (M + H) 402.3583, found
402.3584.
14 M. Minamoto, K. Yamamoto and J. Tsuji, J. Org. Chem., 1990, 55,
766.
15 For 5: white powder; 1H NMR (500 MHz, D2O) d 2.05 (m, 1H), 2.32 (br
d, J = 15.5 Hz, 1H), 3.29 (br t, J = 9.5 Hz, 1H), 3.45 (t, J = 6.5 Hz,
1H), 3.63 (dd, J = 3.0, 9.5 Hz, 1H), 3.68–3.90 (m, 9H), 4.15 (br s, 1H),
4.20 (m, 1H). 13C NMR (90 MHz, D2O) d 28.7, 59.0, 59.7, 60.1, 66.8,
68.0, 69.5, 70.4, 70.5, 73.2, 74.9, 76.0. FAB HRMS calcd for
C13H26NO9 (M + H) 340.1608, found 340.1608. For 6: 1H NMR (500
MHz, D2O) d 1.29 (m, 1H), 1.67 (m, 1H), 2.09 (br d, J = 16.0 Hz, 1H),
2.44 (t, J = 10.0 Hz, 1H), 2.83 (t, J = 6.5 Hz, 1H), 3.36 (t, J = 10.0 Hz,
1H), 3.41 (s, 3H), 3.50–3.70 (m, 6H), 3.83 (m, 2H), 4.01 (d, J = 3.0 Hz,
1H), 4.83 (d, J = 3.5 Hz, 1H). 13C NMR (90 MHz, D2O) d 31.6, 42.6,
55.1, 58.2, 59.3, 61.6, 61.9, 69.5, 71.9, 72.1, 72.3, 72.6, 75.0, 99.6.
ESMS: 354.2 (M + H).
lead to the observed b-aza-C-galacto motif. The observation of
the same sense of stereochemical bias in gluco and manno type
diketones, supports this stereochemical model.8c,10 Therefore,
given the availability of C1-substituted glucals and galactals,
and the amenability of such compounds to conversion to
different C2 substituted ketals, the methodology described
16 R. V. Stevens, Acc. Chem. Res., 1984, 17, 289.
812
Chem. Commun., 2001, 811–812