to the best of our knowledge. Hence, we first considered the
class of C-glycosyl glycines wherein a single carbon-carbon
bond is holding the two molecular fragments.6 We envisaged
a direct entry to these compounds with either R- or
S-configuration by stereospecific R-amination of glycosyl
acetaldehydes via an organocatalyzed asymmetric carbon-
nitrogen bond-forming reaction.7 The ready access to R
(axial) C-glycopyranosyl acetaldehydes by double bond
oxidative cleavage of allyl C-glycopyranosides and the
proline-catalyzed epimerization of R to ꢀ (equatorial) ano-
mers developed in our laboratory8 furnished the required
substrates for our synthetic program. The execution of this
plan starting from ꢀ-C-glucopyranosyl ethanal 1a to give
the corresponding R-configured C-glucosyl glycine 5a under
the effect of the simplest organocatalyst such as L-proline is
presented in Scheme 1. Guided by previous studies on direct
reduced in situ by sodium borohydride and ethanol to give
the nonepimerizable alcohol 3a which was isolated in very
good yield and high diastereomeric purity (de >95%).10
Hydrogenation of 3a over Raney Nickel selectively removed
the Cbz protective groups and cleaved the N-N bond to
give the free amine.11 Treatment of crude amine with Boc2O
furnished the N-Boc protected amino alcohol 4a (Scheme
1). This was readily transformed into the target R-amino ester
5a by oxidation with Jones reagent and esterification with
diazomethane. The configuration of the carbon stereocenter
of the glycinyl group of 5a was assigned7 as R by the
Dale-Mosher NMR method.12 To this end, the H NMR
1
analysis of the pair of Mosher amides derived from 5a
allowed us to calculate ∆δRS values (see Supporting Infor-
mation) as previously performed in our13a,b and other13c
laboratories for other C-glycosyl amino acids.
To firmly establish the key role of L-proline on the
stereochemical outcome of the aldehyde 1a and DBAD
asymmetric coupling, the reaction was carried out in
MeCN at 0 °C in the presence of an achiral base and a
Brønsted acid. To this aim, the couples pyrrolidine/acetic
acid and pyrrolidine/trifluoroacetic acid were employed.
In both cases, the reaction afforded a mixture of R-hy-
drazino aldehyde diastereoisomers as shown by the
isolation of alcohols 3a and epimer 6a in a 1:1 ratio (Table
1). Hence, an internal asymmetric induction by the chiral
glycoside moiety was reasonably excluded while the
presence of proline appeared to be crucial. Accordingly,
a reversal of diastereoselectivity was observed by per-
forming the reaction of 1a with DBAD in the presence of
D-proline as the catalyst (Table 1). This reaction led to
the S configured R-hydrazino alcohol 6a as a single
product in 73% isolated yield. This alcohol was processed
as described for 3a in Scheme 1 and transformed into the
corresponding ꢀ-C-glucosyl amino ester 7a (Table 1).
The organocatalytic enamine-enamine tandem sequence
which involved the proline-catalyzed anomerization of R-C-
glucosyl acetaldehyde 1b to the ꢀ-anomer 1a and the
subsequent R-amination of the latter was also investigated
(see Scheme S1 in Supporting Information). The combination
Scheme 1. L-Proline-Catalyzed R-Amination of Model
Perbenzylated ꢀ-C-Glucosyl Acetaldehyde 1a
electrophilic R-amination of simple R-enolizable ketones and
aldehydes by using enamine catalysis,9 the sugar aldehyde
1a was allowed to react in acetonitrile at 0 °C with a typical
electrophilic nitrogen source such as dibenzyl azodicarboxy-
late (DBAD) in the presence of 30 mol % of L-proline as
the catalyst. To our great delight, this coupling process turned
out to be highly effective because it afforded exclusively
the R-hydrazino aldehyde 2a as judged by TLC and NMR
analyses of the reaction mixture. This product, however, was
(10) The diastereomeric purity of hydrazino-alcohols 3 was evaluated
by 1H NMR analysis (DMSO-d6, 120-160 °C) of crude reaction mixtures
and by comparison with spectra of authentic samples of the corresponding
R-epimers. These were isolated from epimeric mixtures of 3 which were
obtained by treatment of crude aldehydes 2 with imidazole (CH2Cl2, rt,
72 h) and then with NaBH4. The diastereomeric purity of 3 was also
confirmed by 1H NMR analysis of crude N-Boc amino alcohols 4. The same
analysis was extended to hydrazino-alcohols 9a, 11a, 14a, and 16a
(Table 2).
(11) Crucial for the effective execution of this synthetic pathway was
the cleavage of the hydrazino group by the use of high-quality Raney Nickel
catalyst. Apparently deteriorated batches of this catalyst resulted in partial
or total debenzylation of the sugar moiety and the formation of hardly
processable intermediates.
(7) Preliminary data have been succinctly reported in ref 1. It has to be
noted, however, that because of a misprinting the representative ꢀ-C-glucosyl
glycine reported therein was erroneously depicted as having the S config-
uration.
(12) (a) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512–
519. (b) Seco, J. M.; Quin˜oa´, E.; Riguera, R. Chem. ReV. 2004, 104, 17–
117.
(8) Massi, A.; Nuzzi, A.; Dondoni, A. J. Org. Chem. 2007, 72, 10279–
10282.
(13) (a) Dondoni, A.; Massi, A.; Sabbatini, S. Chem.-Eur. J. 2005, 11,
7110–7125. (b) Dondoni, A.; Massi, A.; Minghini, E. Synlett 2006, 539–
542. (c) Ro¨hrig, C. H.; Takhi, M.; Schmidt, R. R. Synlett 2001, 1170–
1172.
(9) A selection: (a) List, B. J. Am. Chem. Soc. 2002, 124, 5656–5667.
(b) Bøgevig, A.; Juhl, K.; Kumaragurubaran, N.; Zhuang, W.; Jørgensen,
K. A. Angew. Chem., Int. Ed. 2002, 41, 1790–1793. (c) Liu, T.-Y.; Cui,
H.-L.; Zhang, Y.; Jiang, K.; Du, W.; He, Z.-Q.; Chen, Y.-C. Org. Lett.
2007, 9, 3671–3674. (d) Mukherjee, S.; Yang, J. W.; Hoffmann, S.; List,
B. Chem. ReV. 2007, 107, 5471–5569.
(14) Attempts to perform both proline-catalyzed cycles in either MeOH
or MeCN produced poorer results as confirmed by the lower yields of
isolated 3a. For a detailed scheme showing all the species involved, see
Scheme S1 in the Supporting Information.
4486
Org. Lett., Vol. 10, No. 20, 2008