On the synthesis of the 2,6-dideoxysugar l-digitoxose

Article abstract of DOI:10.1016/j.carres.2007.09.012

As deoxysugars are integral components of many natural products, the development of efficient chemical and enzymatic routes to prepare these compounds is of particular interest. Herein, we report a comparison of several synthetic methodologies used to prepare protected derivatives of the 2,6-dideoxysugar l-digitoxose. A novel, stereoselective synthetic route to efficiently access methyl 4-O-tert-butyldimethylsilyl-2,6-dideoxy-3-O-trimethylsilyl-α-l-ribo-hexopyranoside in 35% yield over nine facile steps is described.

Full text of DOI:10.1016/j.carres.2007.09.012

Available online at www.sciencedirect.com
Carbohydrate Research 342 (2007) 2695–2704
On the synthesis of the 2,6-dideoxysugar L-digitoxose
Shannon C. Timmonsa and David L. Jakemanb,*
aDepartment of Chemistry, Dalhousie University, Halifax, Nova Scotia, Canada B3H 4J3
bCollege of Pharmacy, Dalhousie University, 5968 College Street, Halifax, Nova Scotia, Canada B3H 3J5
Received 20 June 2007; received in revised form 24 September 2007; accepted 28 September 2007
Available online 5 October 2007
Abstract—As deoxysugars are integral components of many natural products, the development of efficient chemical and enzymatic
routes to prepare these compounds is of particular interest. Herein, we report a comparison of several synthetic methodologies used
to prepare protected derivatives of the 2,6-dideoxysugar L-digitoxose. A novel, stereoselective synthetic route to efficiently access
methyl 4-O-tert-butyldimethylsilyl-2,6-dideoxy-3-O-trimethylsilyl-a-L-ribo-hexopyranoside in 35% yield over nine facile steps is
described.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Keywords: L-Digitoxose; 2,6-Dideoxysugars; Glycosylation; Natural products
1. Introduction
The 2,6-dideoxysugar L-digitoxose (2) is of particular
interest in our laboratory because it is the single carbo-
hydrate component of the structurally unique angu-
cycline jadomycin B (1) (Fig. 1).9,10 L-Digitoxose (2) is
also a component of numerous other bacterial second-
ary metabolites including polyketide-derived antitumor
antibiotics11,12 and antifungal antibiotics.13 Interest-
ingly, D-digitoxose (3) is also a well-known constituent
of cardiac14 and plant-derived15,16 steroidal glycosides.
With respect to L-digitoxose, a number of chemical
syntheses based on significantly different strategies have
been described.17–22 Several of these routes rely on
chiral non-carbohydrate precursors and stereoselective
Natural products continue to serve as reliable scaffolds
for drug discovery and development.1 Interestingly,
many therapeutically relevant natural products are
glycosylated and it is widely acknowledged that carbo-
hydrate functionalities are often critical in conferring
bioactivity,2–4 with the most common biologically rele-
vant sugar residues being 2,6-dideoxyhexoses and
2,3(4),6-trideoxysugars.5 The functional role of deoxy-
sugar residues varies greatly between biomolecules and
remains ambiguous in many cases, due in part to the di-
verse expanse of chemical space accessed by glycosides.6
Thus, a clear understanding of the biological implica-
tions of natural product glycosylation has significantly
lagged behind that of proteins and nucleic acids7 and
the potential of carbohydrate-based therapeutic agents
remains largely under-explored.8 As this research area
is expanding, there is a renewed interest in the efficient
chemical and enzymatic synthesis of deoxysugars of
medicinal value, as well as the total and semi-synthesis
of carbohydrate-containing natural products.
HO
OH
OH
O
O
HO
HO
H
O
2
N
O
O
O
OH
O
O
HO
OH
OH
1
3
*
Figure 1. Structures of jadomycin
D-digitoxose (3).
B (1), L-digitoxose (2), and
Corresponding author. Tel.: +1 902 494 7159; fax: +1 902 494
1396; e-mail: david.jakeman@dal.ca
0008-6215/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2007.09.012

2696
S. C. Timmons, D. L. Jakeman / Carbohydrate Research 342 (2007) 2695–2704
reactions,17–19 which often necessitate numerous dia-
stereomeric chromatographic separations and result in
low overall yields.18,19 Other syntheses of L-digitoxose
originate from monosaccharide precursors,20–22 but
typically involve a large number of synthetic steps and
result in similarly low overall yields.20,21 Of the synthetic
routes previously described, the methodology of Brima-
combe and co-workers was first investigated to prepare
L-digitoxose due to its reasonable number of synthetic
steps (six) and its commercially available monosaccha-
ride origin (L-rhamnose).22
with n-BuLi at <ꢀ30 ꢁC in THF.25 Interestingly, 4-
hydroxyl and 4-O-benzyl derivatives produced only
complex mixtures of products under the aforementioned
reaction conditions.25
In light of these results, a-L-rhamnopyranosides 4a,
4b, and 4c were prepared as previously reported22,25 to
test the efficacy of the Klemer–Rodemeyer elimination
reaction in synthesizing 2-deoxy-3-ketosugars 5a, 5b,
and 5c, respectively, under various reaction conditions.
Initial experiments were conducted using 4-O-MEM-
protected derivative 4a under similar reaction conditions
to those described by Brimacombe and co-workers22
and, in our experience, the highest isolated yield of 2-
deoxy-3-ketosugar 5a achieved was 22% (Table 1, entry
1). The use of smaller C-4 protecting groups, as in the
case of 4-O-MOM- and 4-O-methyl-protected deriva-
tives 4b and 4c, with 6 equiv of n-BuLi resulted in simi-
larly low yields of 5b and 5c, 16% and 19%, respectively,
after silica gel chromatography (Table 1, entries 2–3).
Using 4c, many attempts were made to improve the
yield of the desired 2-deoxy-3-ketosugar by varying the
number of equivalents of n-BuLi (frequently titrated
with 2,5-dimethoxybenzyl alcohol as an indicator26),
the internal temperature of the reaction mixture, and
the reaction time (Table 1, entries 4–7) with little
success.
2. Results and discussion
2.1. Synthesis of L-digitoxose: method of Brimacombe
The key step in the synthetic strategy of Brimacombe
and co-workers for preparing L-digitoxose involves the
reaction of fully protected a-L-rhamnopyranoside 4a
or 4b with n-butyllithium (n-BuLi) at <ꢀ30 ꢁC in THF
(Fig. 2).22 This elimination reaction, first described by
Klemer and Rodemeyer in 1974,23 is based on the
preferential abstraction of the quasi-axial hydrogen
atom attached to C-3 of the carbohydrate ring.24 In
the hands of Brimacombe and co-workers, reaction of
4a and 4b with 6 equiv of n-BuLi successfully produced
desired 2-deoxy-3-ketosugars 5a and 5b in 41% and 38%
yields, respectively, after careful silica gel chromato-
graphy.22 In addition to these rhamnopyranosides,
4-O-methyl-protected a-L-rhamnopyranoside 4c was
also previously used by Horton and co-workers to pre-
pare 2-deoxy-3-ketosugar 5c in 40% yield upon reaction
These disappointingly low yields prompted a careful
1
analysis of the H NMR spectra of crude reaction mix-
tures. Surprisingly, this analysis revealed that only one
diastereomer of 4a, 4b, and 4c had primarily reacted
after reaction times of 1–3 h under various reaction con-
ditions, which were equal to or longer than the 30 min
reaction time reported by Horton25 with 4c, as well as
Li Bu
RO
OCH3
OCH3
H
OCH3
H
n
-BuLi, THF
O
O
O
O
RO
RO
O
O
O
5a: R = MEM (41%)
5b: R = MOM (38%)
5c: R = CH3 (40%)
H
Ph
4a: R = MEM
4b: R = MOM
4c: R = CH3
Figure 2. Mechanism of the Klemer–Rodemeyer elimination reaction with results reported by Brimacombe22 and Horton.25
Table 1. Selected results from varying conditions of Klemer–Rodemeyer elimination reaction with 4a, 4b, and 4c
Entry
4-O-R group (compound)
Equiv of n-BuLi
Temperature (ꢁC)
Time (h)
Isolated yield (%) (product)
1
2
3
4
5
6
7
MEM (4a)
MOM (4b)
CH3 (4c)
CH3 (4c)
CH3 (4c)
CH3 (4c)
CH3 (4c)
6
6
6
3
3
6
3
ꢀ40 to ꢀ30
ꢀ40 to ꢀ30
ꢀ40 to ꢀ30
ꢀ40 to ꢀ30
ꢀ75 to ꢀ65
ꢀ75 to ꢀ65
ꢀ10 to 0
1
1
1
22 (5a)
16 (5b)
19 (5c)
23 (5c)
5 (5c)
No desired product
No desired product
3
20
20
1

S. C. Timmons, D. L. Jakeman / Carbohydrate Research 342 (2007) 2695–2704
2697
the 1 h reaction time reported by Brimacombe22 with 4a
and 4b. Through one-dimensional nuclear Overhauser
effect (NOE) NMR experiments, using 4c as a model,
it was determined that the exo diastereomer reacted
significantly faster than the endo diastereomer (Fig. 3).
Given the 3:2 exo:endo diastereomeric ratio of 4a, 4b,
and 4c, these NMR results were useful in rationalizing
the low yields described in Table 1.
atom and H-3 must be greater in the exo diastereomers
as compared to the endo diastereomers. These confor-
mational observations may provide insight into the
differential reactivity of exo and endo diastereomers of
a-L-rhamnopyranosides 4a, 4b, and 4c, as the synthetic
utility of this reaction has been linked to the conforma-
tional rigidity of the monosaccharide ring.24
In addition to establishing the lower reactivity of endo
diastereomers, careful analysis of both 1H NMR spectra
of crude reaction mixtures and chromatographically
separated products facilitated the identification of sev-
eral Klemer–Rodemeyer elimination reaction byprod-
ucts (Fig. 4). The first byproduct, 3-deoxy-2-ketosugar
6, was observed in only trace quantities in the 1H
NMR spectra of several crude reaction mixtures and is
attributed to proton abstraction from C-2 of the carbo-
hydrate ring.24 Secondly, a significant quantity of alkene
7 was frequently identified and chromatographically iso-
lated from crude reaction mixtures. Byproduct 7 is
attributed to abstraction of the benzylic proton, H-2 in
the 1,3-dioxolane ring, in lieu of H-3 in the carbohydrate
ring.24 In addition to these two expected byproducts, a
third compound (8), also frequently observed in signifi-
cant quantity, was chromatographically separated and
characterized from numerous Klemer–Rodemeyer reac-
tion mixtures. Byproduct 8 appears to result from two
additional reactions of n-BuLi with desired 2-deoxy-3-
ketosugars 5a, 5b, and 5c. One reaction presumably
involves an SN2 attack of n-BuLi at the anomeric center,
displacing the methoxy substituent, while the second
reaction seemingly involves the abstraction of H-4,
resulting in ring opening and the formation of 8.
In an attempt to address the diastereoselectivity issues
associated with benzylidene acetal-protected a-L-
rhamnopyranosides 4a, 4b, and 4c, an isopropylidene
acetal was installed to protect the C-2 and C-3 hydroxyls
of methyl-a-L-rhamnopyranoside. Although largely
unprecedented in a Klemer–Rodemeyer reaction, an iso-
propylidene acetal-protected a-L-rhamnopyranoside
was desirable for two reasons: (i) C-2 of the 1,3-dioxo-
lane ring would no longer be chiral, eliminating diastereo-
selectivity issues; and, (ii) there would be no proton
attached to C-2 of the 1,3-dioxolane ring, eliminating
the possibility of a byproduct resulting from the abstrac-
tion of the benzylic proton. To facilitate this proposed
Klemer–Rodemeyer reaction, isopropylidene-protected
a-L-rhamnopyranoside 11 was prepared from 925
Of particular interest with respect to these observa-
´
tions are the conformational studies of Nanasi and
co-workers on methyl 2,3-O-benzylidene-a-L-rhamno-
pyranoside, as well as 4-O-acetyl and 4-O-benzyl deriv-
atives.27 Their NMR spectroscopy-based investigations
revealed that significant conformational differences exist
in the pyranoside and 1,3-dioxolane rings of exo and
endo diastereomers of the three aforementioned a-L-
rhamnopyranosides. With respect to pyranoside rings,
3JH-2,H-3 values were consistently larger for endo diaste-
reomers as compared to exo diastereomers, suggesting
that the corresponding dihedral angles are smaller in
the case of endo diastereomers. Moreover, the coupling
constant between the acetal carbon atom and quasi-
axial proton H-3, which is also dependent on dihedral
angle,28 was consistently larger for exo diastereomers
of these a-L-rhamnopyranosides. This indicates that
the C–O–C–H dihedral angle between the acetal carbon
OCH
OCH
H
3
3
O
O
H CO
3
H CO
3
O
H
O
H
O
O
Ph
Ph
H
4c exo
4c endo
Figure 3. exo and endo Diastereomers of 4c (arrows indicate strong 1D
NOE interactions).
O
OCH
O
O
OCH
HO
3
3
RO
RO
RO
O
6
7
8
R = MEM, MOM, or CH
3
Figure 4. Identified Klemer–Rodemeyer elimination reaction
byproducts.
OCH3
OCH3
OCH3
OCH3
NaH, CH3I
DMF
n-BuLi
THF
2,2-dimethoxypropane
p-TsOH
O
O
O
O
O
HO
HO
H3CO
H3CO
(92%)
(66%)
OH
O
O
OH
O
O
12
9
10
11
Scheme 1. Proposed synthesis of 2-deoxy-3-ketosugar 12.

2698
S. C. Timmons, D. L. Jakeman / Carbohydrate Research 342 (2007) 2695–2704
according to the isopropylidene installation method of
resulted in degradation, as no desired product was
isolated after 23 h at ꢀ40 to 20 ꢁC with 4 equiv of base
(Table 2, entry 3). Similarly poor results were obtained
using tert-butyllithium (t-BuLi), which produced a yield
of only 5% after 3 h at ꢀ50 to ꢀ25 ꢁC with 2 equiv of
base (Table 2, entry 4). Interestingly, the use of s-BuLi
as an alternative base provided an essential mixture of
strong basicity and steric hindrance, limiting byproducts
resulting from nucleophilic attack of the base. Specifi-
cally, reaction of 4a with 3 equiv of s-BuLi at ꢀ50 to
ꢀ35 ꢁC for 4 h resulted in a significantly improved
71% yield of 2-deoxy-3-ketosugar 5a after silica gel chro-
matography (Table 2, entry 5). Consistent with afore-
mentioned stereochemical observations, NMR analysis
of crude reaction mixtures revealed that the exo dia-
stereomer of 4a reacted fastest when s-BuLi was used
as the base.
Liptak and co-workers29 and the methylation conditions
´
of Brimacombe30 (Scheme 1). The Klemer–Rodemeyer
elimination reaction was first attempted with 11 at
ꢀ50 to ꢀ40 ꢁC using 5 equiv of n-BuLi, but after 2 h a
significant quantity of starting material remained and
no 2-deoxy-3-ketosugar 12 was formed as determined
1
by analysis of the H NMR spectrum of the crude reac-
tion mixture. Warmer reaction temperatures (ꢀ10 to
0 ꢁC) and lower equivalents of n-BuLi (2.5) facilitated
the reaction of 11, but only complex mixtures of prod-
ucts were obtained and no 2-deoxy-3-ketosugar 12 was
observed in the 1H NMR spectra of crude reaction
mixtures.
Interestingly, the conformation of methyl 4-O-acetyl-
2,3-O-isopropylidene-a-L-rhamnopyranoside has also
´
been studied by Nanasi and co-workers using NMR
spectroscopy.27 In terms of the pyranose ring, the con-
formation of this compound was found to be similar
to the exo diastereomers of benzylidene acetal-protected
a-L-rhamnopyranosides, however; with respect to the
1,3-dioxolane ring, this isopropylidene acetal-protected
a-L-rhamnopyranoside possessed a conformation simi-
lar to the endo diastereomers of benzylidene acetal-pro-
tected a-L-rhamnopyranosides.
The lack of success obtained using 11, as well as the
nucleophilic role of n-BuLi in byproduct formation,
motivated us to explore the use of different bases.
Although the majority of Klemer–Rodemeyer elimina-
tion reactions published to date report the use of n-
BuLi,22–25,31,32 one procedure has been described using
sec-butyllithium (s-BuLi) as a base.33 Thus, 4-O-methyl-
and 4-O-MEM-protected a-L-rhamnopyranosides 4c
and 4a were used in Klemer–Rodemeyer reactions with
various bases in attempts to improve the yields of
2-deoxy-3-ketosugars 5c and 5a, respectively (Table 2).
Reaction of 4c with lithium diisopropylamide (LDA)
remained sluggish at successively higher temperatures,
producing only a 6% yield of 2-deoxy-3-ketosugar 5c
after 4 h at ꢀ30 to 0 ꢁC with 2 equiv of base (Table 2,
entries 1–2). Longer reaction times using 4c and LDA
With a viable synthetic route to 2-deoxy-3-ketosugar
5a in hand, a sodium borohydride reduction was
attempted to prepare 13 (Scheme 2). This reaction pro-
ceeded stereoselectively, as reported by Brimacombe
and co-workers,22 resulting in a C3–OH axial:equatorial
ratio of 10:1 and an 80% yield of 13 after silica gel
chromatography. This high level of stereoselectivity pre-
sumably results from sodium borohydride attack via an
equatorial trajectory, avoiding an energetically unfavor-
able 1,3-diaxial interaction with the axial methoxy
group located at the anomeric center.34,35
We next chose to deprotect 2,6-dideoxysugar 13 to
facilitate the conversion of 2 into a glycosyl donor suit-
able for sugar nucleotide synthesis with the aim of using
this substrate to confirm the in vitro activity of the gly-
cosyltransferase involved in jadomycin B biosynthesis.
Deprotection of 13 using 29% aq acetic acid under
gentle reflux for 2 h, as reported by Brimacombe and
co-workers,22 was unsuccessful. Interestingly, NMR
and MS analysis of the reaction product revealed that,
although the anomeric methoxy group had been
removed under the aforementioned conditions, the 4-
O-MEM ether functionality was still intact (Scheme 2).
Higher concentrations of aq acetic acid and longer
Table 2. Selected Klemer–Rodemeyer elimination reactions results with different bases
Entry
4-O-R group (compound)
Base (equiv)
Temperature (ꢁC)
Time (h)
Isolated yield (%) (product)
1
2
3
4
5
CH3 (4c)
CH3 (4c)
CH3 (4c)
CH3 (4c)
MEM (4a)
LDA (2)
LDA (2)
LDA (4)
t-BuLi (2)
s-BuLi (3)
ꢀ50 to ꢀ25
ꢀ30 to 0
3
4
23
3
4 (5c)
6 (5c)
No desired product
5 (5c)
71 (5a)
ꢀ40 to 20
ꢀ50 to ꢀ25
ꢀ50 to ꢀ35
4
OCH
OCH
3
3
NaBH
MeOH
4
OH
OH
O
OH
O
O
29% aq AcOH
O
MEMO
MEMO
HO
(80%)
5a
13
2
Scheme 2. Sodium borohydride reduction of 5a and deprotection of 13.

S. C. Timmons, D. L. Jakeman / Carbohydrate Research 342 (2007) 2695–2704
2699
reaction times unfortunately resulted in the complete
degradation of this 2,6-dideoxysugar derivative.
In summary, the importance of the base in Klemer–
Rodemeyer elimination reactions involving conforma-
tionally flexible monosaccharides has been demonstrated.
The regioselectivity of this elimination reaction is also
dependent on the type and steric bulk of the C-4 protect-
ing group,36 complicating the selection of another suit-
able C-4 protecting group. Thus, efforts were re-directed
toward the design and execution of an alternative
synthetic route to access L-digitoxose, or a protected
derivative thereof, which could be used directly in the
coupling of an L-digitoxosyl halide with a phosphate
or nucleoside 50-diphosphate, as previously demon-
strated in the synthesis of other 2,6-dideoxysugar
nucleotides.37,38
co-workers.41 The original procedure involved the use
of 0.008 equiv of potassium carbonate and a reaction
time of 24 h at room temperature, however, by increas-
ing the equivalents of potassium carbonate 5-fold to
0.04, the reaction time was decreased to only 3 h at
room temperature. Following a simple filtration through
a silica gel plug, crystalline 6-deoxy-L-glucal (16) was
obtained in quantitative yield.
Regioselective oxidation of 16 was accomplished
using pyridinium dichromate as previously reported by
Czernecki and co-workers.42 In our experience, precipi-
tation of the resulting Cr(III) species using toluene
proved tedious and largely ineffective. Alternatively,
the reaction mixture was passed successively through
two short silica gel plugs, which effectively bound the
green Cr(III) species, facilitating an efficient purification
of enulose 17 and affording crystals of 17 in 81% yield.
In attempts to streamline the proposed synthetic
strategy, the sodium borohydride reduction of enulose
17 was investigated (Scheme 4). Using conditions previ-
ously described for the preparation of 13 from 2-deoxy-
3-ketosugar 5a,22 reduction with sodium borohydride
provided only an inefficient route back to 6-deoxy-L-
glucal (16). Interestingly, Luche reduction conditions,43
which include a lanthanide salt such as CeCl3Æ7H2O in
addition to sodium borohydride, sometimes result in
stereoselectivities opposite to reductions involving
sodium borohydride alone.44 The high regio- and
stereoselectivity associated with Luche-type reductions
was first noted by Danishefsky45 and is postulated to
result from the coordination of CeCl3, a Lewis acid, with
the carbonyl group, essentially blocking one face of the
molecule.46 In the case of enulose 17, Luche-type reduc-
tion did result in a difference in stereoselectivity, but the
2.2. Synthesis of L-digitoxose: a second approach
Given the previous success of the stereoselective sodium
borohydride reduction of 5a, a new synthetic plan was
devised to synthesize L-digitoxose, which included the
preparation of a 2-deoxy-3-ketosugar via a different
route. This new strategy also originated from L-rham-
nose and involved the use of a glycal (Scheme 3).39
The first step in this synthetic route involved the
preparation of 3,4-di-O-acetyl-6-deoxy-L-glucal (15)
from L-rhamnose (14) using a one-pot, three-step proce-
dure originally developed by Koreeda and co-workers.40
Following a facile purification using a silica gel plug, 15
was isolated in 85% yield over three steps in only 8 h.
Deacetylation of 15 using potassium carbonate in
methanol was accomplished using a procedure similar
to the one previously described by van Heerden and
1. Ac2O, HBr/AcOH
2. HBr/AcOH
K2CO3
OH
pyridinium dichromate
AcOH, EtOAc
3. CuSO4, Zn0, NaOAc, H2O, AcOH
MeOH
O
HO
HO
16
AcO
O
HO
O
O
HO
(81%)
(85% over 3 steps)
(quant)
OH
14
AcO
15
OH
O
17
TBDMS-Cl, imidazole
DMF
(80%)
NaBH4
MeOH
OCH3
NaOMe
TMS-Cl
pyridine
OCH3
OCH3
OTMS
in MeOH
(86%)
OH
O
TBDMSO
O
O
O
O
TBDMSO
TBDMSO
TBDMSO
(78%)
(95%)
O
18
21
(35% over 9 steps)
20
19
10:1 α:β
Scheme 3. Novel synthesis of silyl-protected L-digitoxose derivative 21.
NaBH4
MeOH
NaBH4, CeCl3·7H2O
1:1 MeOH:THF
OH
O
HO
HO
HO
HO
HO
HO
O
O
O
+
O
17
16
(73%)
16
(50%)
22
(10%)
Scheme 4. Reduction of enulose 17 under various reaction conditions.

2700
S. C. Timmons, D. L. Jakeman / Carbohydrate Research 342 (2007) 2695–2704
5:1 ratio of 6-deoxy-L-glucal (16):L-digitoxal (22) was
very unfavorable.
the bioactive 2,6-dideoxysugar L-daunosamine31 and
related sugars.32 Although the deprotection of the 4-O-
MEM ether was problematic, this work illustrates the
potential use of the Klemer–Rodemeyer elimination
reaction with s-BuLi and conformationally flexible
monosaccharides.
A novel synthetic route was also developed to provide
efficient access to silyl-protected L-digitoxose derivative
21 in 35% overall yield over nine steps. The benefits of
this route over other methods previously described to
prepare this 2,6-dideoxysugar include the ease and speed
of synthetic steps as well as the high degree of regio- and
stereoselectivity achieved in key transformations. This
synthetic strategy could potentially be extended for use
in the synthesis of other 2,6-dideoxy- and 2-deoxysugars
where a change in stereochemistry at C-3, as compared
to the starting glycal, is desired.
To facilitate the efficient reduction of the 3-keto-
functionality of 17 to an axial hydroxyl group, access
to a 2-deoxy-3-ketosugar analogous to those previously
described (5a, 5b, and 5c) was required. This transfor-
mation was accomplished in two steps. The remaining
free hydroxyl group of enulose 17 was first protected
with a tert-butyldimethylsilyl (TBDMS),47 which affor-
ded TBDMS-protected enulose 18 in 80% yield. A
TBDMS protecting group was chosen to protect enulose
17 because of its well-documented, non-acidic ease of
removal
using
tetra-n-butylammonium
fluoride
(Bu4NF),47,48 as well as its stability to reaction condi-
tions in forthcoming synthetic steps.
To prepare 2-deoxy-3-ketosugar 19 from 18, a simple
procedure involving sodium methoxide in methanol was
employed, which afforded 19 in 86% yield as a 10:1 a:b
mixture of diastereomers.35 This diastereomeric mixture
was used directly in the next synthetic step without any
further purification.
4. Experimental
Reduction of 19 with sodium borohydride22 provided
straightforward, stereoselective access to a 2,6-dideoxy-
sugar containing an axially configured C-3 hydroxyl
group (20). Interestingly, diastereomeric mixture 19
(10:1 a:b) was reduced to a crude 2,6-dideoxysugar
mixture with a C3–OH axial:equatorial ratio of 10:1,
demonstrating the importance of the axial anomeric
methoxy substituent in conferring diastereoselectivity.
Following silica gel chromatography, 20 was isolated
in 78% yield.
Lastly, the remaining C-3 hydroxyl group was pro-
tected with a trimethylsilyl (TMS) group,49 which affor-
ded fully protected L-digitoxose derivative 21 in 95%
yield. The 3-O-TMS functionality was chosen as a pro-
tecting group at this last step in the synthetic strategy
because it could be easily removed under the same con-
ditions as the 4-O-TBDMS group with Bu4NF after 21
is coupled with a nucleoside 50-diphosphate in due
course.
4.1. General methods
With the exclusion of certain solvents, chemicals were
purchased commercially and used without further puri-
fication unless otherwise specified. Anhydrous CH2Cl2
and THF were dried and purified via filtration through
alumina using an Innovative Technology solvent
purification system. All other solvents were reagent
grade unless otherwise noted. Analytical thin-layer chro-
matography was performed on glass-backed TLC plates
pre-coated with silica gel (SiliCycleTM, 250 lm) and com-
pounds were detected by UV absorbance (254 nm) and/
or by spraying with a KMnO4 visualization solution (3 g
KMnO4, 20 g K2CO3, 5 mL of 5% w/v aq NaOH,
300 mL H2O). Automated normal-phase silica gel
chromatography was performed using a Biotage SP1TM
flash chromatography system. Normal-phase silica
gel benchtop chromatography was performed using
SiliCycleTM 230–400 mesh ultra pure silica. Melting
points were measured in open capillary tubes using a
Gallenkamp melting point apparatus and are reported
uncorrected. Nuclear magnetic resonance experiments
were conducted using a Bruker AVANCE 500 MHz
spectrometer. 1D NOE NMR experiments were per-
formed using a mixing time of 750 ms. Chemical shifts
are reported in parts per million (ppm) relative to a
tetramethylsilane internal standard at 0.00 ppm for sam-
ples in CDCl3, while spectra recorded in D2O were ref-
erenced to the solvent peak at 4.79 ppm and externally
to 2,2-dimethyl-2-silapentane-5-sulfonate (DSS). High-
resolution EI-MS measurements were obtained using a
DuPont CEC 21-110B double-focusing magnetic sector
instrument and high-resolution ESI-MS measurements
were obtained using a Bruker Daltonics micrOTOF
instrument.
3. Conclusions
In conclusion, the yields of Klemer–Rodemeyer elimina-
tion reactions can potentially be improved through the
use of s-BuLi as a base, although careful monitoring
of reaction temperature and time is critical and regio-
selectivity is often determined only on a case by case
basis. Higher predictability can be established for
conformationally rigid carbohydrates (e.g., monosac-
charides containing two benzylidene acetals) as com-
pared to carbohydrates with more conformational
flexibility (e.g., monosaccharides containing a single
benzylidene acetal).24 The utility of this reaction with
conformationally rigid monosaccharides has previously
been demonstrated in the large-scale (20 g) synthesis of

S. C. Timmons, D. L. Jakeman / Carbohydrate Research 342 (2007) 2695–2704
2701
4.2. General procedure for the synthesis of 2-deoxy-3-
ketosugars 5a, 5b, and 5c
colorless syrup: Rf = 0.38 (35:65 EtOAc:hexanes); 1H
NMR (CDCl3): d 6.03 (br d, 1H, J2,3 = 10.0 Hz, H-3),
5.74 (ddd, 1H, J2,4 = 0.5 Hz, J1,2 = 2.5 Hz, H-2), 4.86
(d, 1H, JH,H = 7.0 Hz, OCH2O), 4.81 (m, 1H, H-1),
4.77 (d, 1H, OCH2O), 3.88–3.54 (m, 6H, H-4, H-5,
OCH2CH2O), 3.42 (s, 3H, OCH3), 3.39 (s, 3H, OCH3),
1.30 (d, 3H, J5,6 = 5.5 Hz, H3-6); 13C NMR (CDCl3):
d 131.7 (C-3), 126.6 (C-2), 95.5 (C-1), 95.1 (OCH2O),
75.5, 71.8, 67.3, 66.0 (C-4, C-5, OCH2CH2O), 59.2
(OCH3), 55.7 (OCH3), 18.3 (C-6); HRMS (ESI+) m/z:
calcd for C11H20O5Na, 255.1203; found, 255.1181.
A diastereomeric mixture of a-L-rhamnopyranoside
acetals (4a, 4b, or 4c; exo:endo = 3:2) was dissolved in
anhydrous THF under a nitrogen atmosphere in a
two-necked round-bottomed flask fitted with a ther-
mometer. The reaction mixture was cooled in a dry
ice-acetone bath to the desired reaction temperature
and the desired base (n-BuLi or s-BuLi) was added
dropwise. Upon reaction completion, as determined by
TLC, the reaction mixture was allowed to warm to
ꢀ10 ꢁC before pouring into an ice-water mixture con-
taining 10% w/v NH4Cl. The resulting aqueous solution
was extracted with equal volumes of CH2Cl2 (thrice) and
the combined organic extracts were then extracted with
an equal volume of H2O. The organic layer was dried
over Na2SO4, filtered, and concentrated under reduced
pressure to afford the crude product as a yellow syrup.
See Tables 1 and 2 for specific reaction conditions and
corresponding yields.
4.2.5. (6R)-6-Hydroxy-3-methoxy-dec-2-en-4-one (8).
This compound was isolated as a byproduct from sev-
eral Klemer–Rodemeyer elimination reactions in which
4c was used as a starting material to prepare 5c. Similar
1
byproducts were also observed in the crude H NMR
spectra of analogous reactions aimed at preparing
2-deoxy-3-ketosugars 5a and 5b. Compound 8 was a
colorless syrup: Rf = 0.27 (20:80 EtOAc:hexanes); 1H
NMR (CDCl3): d 6.29 (q, 1H, J1,2 = 7.0 Hz, H-2), 4.07
(m, 1H, H-6), 3.66 (s, 3H, OCH3), 3.09 (br s, 1H,
OH), 2.80 (dd, 1H, J5a,6 = 2.5 Hz, J5a,5b = 17.5 Hz,
H-5a), 2.66 (dd, 1H, J5b,6 = 9.0 Hz, H-5b), 1.84 (d,
3H, H3-1), 1.57–1.22 (m, 6H, CH2CH2CH2), 0.91 (t,
3H, JH,H = 5.0 Hz, CH2CH3); 13C NMR (CDCl3): d
198.5 (C@O), 154.9 (C-3), 124.8 (C-2), 67.9 (C-6), 60.0
(OCH3), 44.4 (C-5), 36.3 (CH2), 27.7 (CH2), 22.6
(CH2), 14.0 (CH2CH3), 11.4 (C-1); HRMS (EI+) m/z:
calcd for C11H20O3, 200.1412; found, 200.1408.
4.2.1.
Methyl
2,6-dideoxy-4-O-(2-methoxyethoxy)-
methyl-a-L-erythro-hexopyranosid-3-ulose (5a). Com-
pound 5a was prepared as described above starting
from 0.354 g (1.00 mmol) of 4a. Purification via bench-
top silica gel chromatography (isocratic 35:65 EtOAc:
hexanes) afforded 5a as a colorless syrup: Rf = 0.21
(35:65 EtOAc:hexanes); 1H NMR spectral data matched
that reported.22
4.2.2. Methyl 2,6-dideoxy-4-O-methoxymethyl-a-L-ery-
thro-hexopyranosid-3-ulose (5b). Compound 5b was
prepared as described above starting from 0.266 g
(1.00 mmol) of 4b. Purification via benchtop silica gel
chromatography (isocratic 22:78 EtOAc:hexanes) affor-
4.3. Methyl 2,3-O-isopropylidene-a-L-rhamnopyranoside
(10)
This compound was prepared using a procedure
29
´
described by Liptak and co-workers. Starting from
ded 5b as
a
colorless syrup: Rf = 0.45, (35:65
6.39 g (35.9 mmol) of 9, compound 10 was isolated in
92% yield (7.20 g, 33.0 mmol) as a pale yellow syrup:
Rf = 0.43 (6:3:1 CH2Cl2:EtOAc:EtOH); 1H NMR
(CDCl3): d 4.86 (br s, 1H, H-1), 4.14–4.00 (m, 2H,
H-2, H-3), 3.58 (m, 1H, H-5), 3.36 (s, 3H, OCH3),
3.31 (m, 1H, H-4), 1.51 (s, 3H, C(CH3)2), 1.34 (s, 3H,
C(CH3)2), 1.28 (d, 3H, J5,6 = 6.6 Hz, H3-6).
1
EtOAc:hexanes); H NMR spectral data matched that
reported.22
4.2.3. Methyl 2,6-dideoxy-4-O-methyl-a-L-erythro-hexo-
pyranosid-3-ulose (5c). Compound 5c was prepared as
described above starting from 0.253 g (1.00 mmol) of
4c. Purification via benchtop silica gel chromatography
(isocratic 10:90 EtOAc:hexanes) afforded 5c as a color-
4.4. Methyl 2,3-O-isopropylidene-4-O-methyl-a-L-
rhamnopyranoside (11)
1
less syrup: Rf = 0.17 (10:90 EtOAc:hexanes); H NMR
spectral data matched that reported.25
This compound was prepared using a procedure de-
scribed by Brimacombe.30 Starting from 2.75 g
(12.6 mmol) of 10, compound 11 was isolated in 66%
yield (1.93 g, 8.32 mmol) following benchtop silica gel
chromatography (isocratic 10:90 EtOAc:hexanes) as a
colorless syrup: Rf = 0.78 (35:65 EtOAc:hexanes); 1H
NMR (CDCl3): d 4.83 (br s, 1H, H-1), 4.14–4.08 (m,
2H, H-2, H-3), 3.57 (m, 1H, H-5), 3.53 (s, 3H, OCH3),
3.36 (s, 3H, OCH3), 2.96 (dd, 1H, J3,4 = 9.6 Hz,
4.2.4. Methyl 2,3,6-trideoxy-4-O-(2-methoxyethoxy)-
methyl-a-L-erythro-hex-2-enopyranoside (7). This com-
pound was isolated as a byproduct from several
Klemer–Rodemeyer elimination reactions in which 4a
was used as a starting material to prepare 5a. Similar
byproducts were also observed in the crude H NMR
spectra of analogous reactions aimed at preparing
2-deoxy-3-ketosugars 5b and 5c. Compound 7 was a
1

2702
S. C. Timmons, D. L. Jakeman / Carbohydrate Research 342 (2007) 2695–2704
J4,5 = 6.3 Hz, H-4), 1.54 (s, 3H, C(CH3)2), 1.35 (s, 3H,
C(CH3)2), 1.28 (d, 3H, J5,6 = 6.3 Hz, H3-6); HRMS
(ESI+) m/z: calcd for C11H20O5Na, 255.1203; found,
255.1185.
the reaction mixture was filtered through two short silica
gel plugs in scintillated funnels (washing with HPLC
grade EtOAc), to remove the Cr(III) species instead of
precipitating with toluene as previously reported.42
Starting from 2.74 g (21.1 mmol) of 16, compound 17
was isolated in 81% yield (2.19 g, 17.1 mmol) as white
crystals: mp 87–88 ꢁC, lit.42 92–93 ꢁC, lit.50 86 ꢁC;
Rf = 0.77 (100% EtOAc); 13C NMR (125 MHz, CDCl3):
d 194.2 (C@O), 164.7 (C-1), 103.5 (C-2), 80.0 (C-5), 72.8
4.5. Methyl 2,6-dideoxy-4-O-(2-methoxyethoxy)methyl-
a-L-ribo-hexopyranoside (13)
This compound was prepared using a procedure
described by Brimacombe and co-workers.22 Starting
from 0.560 g (2.26 mmol) of 5a, compound 13 was iso-
lated in 80% yield (0.453 g, 1.81 mmol) following bench-
top silica gel chromatography (isocratic 75:25 EtOAc:
hexanes) as a pale yellow syrup: Rf = 0.22 (75:25
1
(C-4), 18.0 (C-6). H NMR spectral data matched that
reported.50
4.9. 1,5-Anhydro-4-O-tert-butyldimethylsilyl-2,6-
dideoxy-L-erythro-hex-1-en-3-ulose (18)
1
EtOAc:hexanes); H NMR spectral data matched that
reported.22
1,5-Anhydro-2,6-dideoxy-L-erythro-hex-1-en-3-ulose (17)
(1.54 g, 12.0 mmol) was dissolved in anhydrous DMF
(15 mL) in a round-bottomed flask under a nitrogen
atmosphere. Imidazole (2.05 g, 30.1 mmol) and tert-
butyldimethylsilyl chloride (2.17 g, 14.4 mmol) were
added and the reaction mixture was stirred for 19 h at
rt, after which time TLC analysis revealed the complete
consumption of 17. The reaction mixture was diluted
with EtOAc (50 mL) and extracted with H2O
(3 · 50 mL). After drying the organic extracts over
Na2SO4 and concentration under reduced pressure,
TBDMS-protected derivative 18 was obtained in 80%
4.6. Di-O-acetyl-6-deoxy-L-glucal (15)
This compound was prepared using a procedure
described by Koreeda and co-workers40 except that the
bromination reaction was complete as determined by
TLC after 1.5 h and was not stirred overnight as previ-
ously reported.40 Starting from 5.00 g (27.4 mmol) of
14, compound 15 was isolated in 85% yield (4.99 g,
23.3 mmol) over three steps following purification using
a short silica gel plug in a scintillated funnel (isocratic
10:90 EtOAc:hexanes) as a colorless syrup: Rf = 0.60
yield (2.33 g, 9.61 mmol) as
a
colorless syrup:
1
(25:75 EtOAc:hexanes); H and 13C NMR spectral data
Rf = 0.38 (10:90 EtOAc:hexanes); 1H NMR (CDCl3):
d 7.29 (d, 1H, J1,2 = 6.0 Hz, H-1), 5.35 (d, 1H, H-2),
4.32 (dq, 1H, J4,5 = 11.5 Hz, J5,6 = 6.0 Hz, H-5), 4.02
(d, 1H, H-4), 1.52 (d, 3H, H3-6), 0.95 (s, 9H,
C(CH3)3), 0.24 (s, 3H, Si(CH3)2), 0.12 (s, 3H, Si(CH3)2);
13C NMR (CDCl3): d 193.7 (C@O), 162.3 (C-1), 105.0
(C-2), 80.2 (C-5), 75.3 (C-4), 26.0 (C(CH3)3), 25.83
(C(CH3)3), 25.80 (C(CH3)3), 18.7 (C(CH3)3), 18.2 (C-
6), ꢀ3.9 (Si(CH3)2), ꢀ5.4 (Si(CH3)2); HRMS (ESI+)
m/z: calcd for C12H22O3SiNa, 265.1230; found,
265.1220.
matched that reported.41
4.7. 6-Deoxy-L-glucal (16)
This compound was prepared using a procedure de-
scribed by van Heerden and co-workers41 except that
the number of equivalents of potassium carbonate was
increased 5-fold to significantly reduce the reaction time
to 3 h. Vigorous stirring was also necessary. Starting
from 2.84 g (13.3 mmol) of 15, compound 16 was iso-
lated in quantitative yield (1.73 g, 13.3 mmol) following
filtration through a short silica gel plug in a scintillated
funnel (isocratic HPLC grade MeOH), which afforded
white crystals: mp 73–74 ꢁC, lit.42 72–73 ꢁC; Rf = 0.36
(100% EtOAc); 1H NMR (CDCl3): d 6.31 (dd, 1H,
J1,2 = 6.0 Hz, J1,3 = 1.5 Hz, H-1), 4.69 (dd, 1H,
J2,3 = 2.0 Hz, H-2), 4.21 (ddd, 1H, J3,4 = 7.5 Hz, H-3),
3.84 (dq, 1H, J4,5 = 10.0 Hz, J5,6 = 6.0 Hz, H-5), 3.40
(dd, 1H, H-4), 1.38 (d, 3H, H3-6); 13C NMR (CDCl3):
d 144.8 (C-1), 102.7 (C-2), 75.2 (C-4), 74.5 (C-5), 70.2
(C-3), 17.1 (C-6).
4.10. Methyl 4-O-tert-butyldimethylsilyl-2,6-dideoxy-
a/b-L-erythro-hexopyranosid-3-ulose (19)
1,5-Anhydro-4-O-tert-butyldimethylsilyl-2,6-dideoxy-L-
erythro-hex-1-en-3-ulose (18) (85 mg, 0.35 mmol) was
dissolved in 0.01 M NaOMe in MeOH (7.1 mL) and
the resulting reaction mixture was stirred for 1 h at rt,
after which time TLC analysis revealed the complete
consumption of 18. The reaction mixture was neutral-
ized to pH 7 using Amberliteꢂ IR-120 (H+) ion
exchange resin (free acid form), filtered, and concen-
trated under reduced pressure to afford 19 in 86% yield
(81 mg, 0.30 mmol) in an a:b ratio of 10:1 as a pale
yellow syrup: Rf = 0.35, 0.27 (10:90 EtOAc:hexanes);
1H NMR (CDCl3): d a diastereomer: 5.06 (br d, 1H,
J1,2ax = 4.5 Hz, H-1), 3.96 (dq, 1H, J4,5 = 9.5 Hz,
4.8. 1,5-Anhydro-2,6-dideoxy-L-erythro-hex-1-en-3-ulose
(17)
This compound was prepared using a procedure
described by Czernecki and co-workers42 except that

S. C. Timmons, D. L. Jakeman / Carbohydrate Research 342 (2007) 2695–2704
2703
J5,6 = 6.5 Hz, H-5), 3.88 (d, 1H, H-4), 3.36 (s, 3H,
OCH3), 2.75 (dd, 1H, J2ax,2eq = 14.0 Hz, H-2ax), 2.61
(br d, 1H, H-2eq), 1.44 (d, 3H, H3-6), 0.95 (s, 9H,
C(CH3)3), 0.20 (s, 3H, Si(CH3)2), 0.06 (s, 3H, Si(CH3)2),
(320 lL, 2.49 mmol) was added dropwise and the reac-
tion mixture was stirred for 1 h at 0–5 ꢁC, after which
time TLC analysis revealed the complete consumption
of 20. The reaction mixture was diluted with hexanes
(50 mL) and washed with H2O (5 · 20 mL). The organic
layer was dried over Na2SO4 and concentrated under
reduced pressure to afford 21 in 95% yield (549 mg,
1.57 mmol) as a colorless syrup: Rf = 0.50 (10:90
EtOAc:hexanes); 1H NMR (CDCl3): d 4.62 (dd, 1H,
J1,2ax = 4.0 Hz, J1,2eq = 3.5 Hz, H-1), 4.08 (dq, 1H,
J4,5 = 10.0 Hz, J5,6 = 6.5 Hz, H-5), 3.94 (m, 1H, H-3),
3.32 (dd, 1H, J3,4 = 3.0 Hz, H-4), 3.31 (s, 3H, OCH3),
1.97 (ddd, 1H, J2ax,2eq = 14.0 Hz, J2eq,3 = 5.0 Hz,
H-2eq), 1.78 (dt, 1H, J2ax,3 = 4.0 Hz, H-2ax), 1.19 (d,
3H, H3-6), 0.90 (s, 9H, C(CH3)3), 0.12 (s, 9H, Si(CH3)3),
0.07 (s, 3H, Si(CH3)2), 0.06 (s, 3H, Si(CH3)2); 13C NMR
(CDCl3): d 97.9 (C-1), 74.9 (C-4), 68.1 (C-3), 65.9 (C-5),
55.1 (OCH3), 36.5 (C-2), 26.08 (C(CH3)3), 26.07
(C(CH3)3), 26.06 (C(CH3)3), 18.33 (C(CH3)3), 18.31
(C(CH3)3), 18.27 (C-6), 0.66 (Si(CH3)3), 0.65 (Si(CH3)3),
0.64 (Si(CH3)3), ꢀ3.8 (Si(CH3)2), ꢀ4.6 (Si(CH3)2);
HRMS (EI+) m/z: calcd for C16H36O4Si2, 348.2152;
found, 348.2151.
b
diastereomer: 4.59 (dd, 1H, J1,2a = 9.5 Hz,
J1,2e = 2.5 Hz, H-1); 13C NMR (CDCl3): d a diastereo-
mer: 203.4 (C@O), 99.6 (C-1), 80.5 (C-4), 70.7 (C-5),
54.8 (OCH3), 46.5 (C-2), 25.94 (C(CH3)3), 25.93
(C(CH3)3), 25.91 (C(CH3)3), 19.2 (C-6), 18.7
(C(CH3)3), ꢀ4.1 (Si(CH3)2), ꢀ5.4 (Si(CH3)2); HRMS
(ESI+) m/z: calcd for C13H26O4SiNa, 297.1493; found,
297.1488.
4.11. Methyl 4-O-tert-butyldimethylsilyl-2,6-dideoxy-a-
L-ribo-hexopyranoside (20)
Methyl 4-O-tert-butyldimethylsilyl-2,6-dideoxy-a/b-L-
erythro-hexopyranosid-3-ulose (19) (1.78 g, 6.49 mmol)
was dissolved in HPLC grade MeOH (72 mL) in a
round-bottomed flask. NaBH4 (2.33 g, 61.7 mmol) was
added in small portions and the reaction mixture was
stirred for 1 h at rt, after which time TLC analysis
revealed the complete consumption of 19. The reaction
mixture was concentrated under reduced pressure,
re-dissolved in CH2Cl2 (50 mL), and washed with H2O
(25 mL). The organic layer was dried over MgSO4 and
concentrated under reduced pressure to yield crude 20
as a pale yellow syrup, which was purified via automated
silica gel chromatography using the following gradient:
5:95 A:B (10 CV), linear gradient to 10:90 A:B (5 CV),
10:90 A:B (2 CV) where A = EtOAc and B = hexanes.
Following chromatography diastereomerically pure 20
was obtained in 78% yield (1.40 g, 5.06 mmol) as a
colorless syrup: Rf = 0.16 (10:90 EtOAc:hexanes); 1H
NMR (CDCl3): d 4.70 (br d, 1H, J1,2ax = 4.0 Hz, H-1),
3.96 (dq, 1H, J4,5 = 9.5 Hz, J5,6 = 6.0 Hz, H-5), 3.91
(m, 1H, H-3), 3.36 (s, 3H, OCH3), 3.32 (dd, 1H,
J3,4 = 3.0 Hz, H-4), 3.05 (d, 1H, J3,OH = 5.5 Hz, OH),
2.14 (dd, 1H, J2ax,2eq = 15.0 Hz, J2eq,3 = 3.0 Hz,
H-2eq), 1.88 (dt, 1H, J2ax,3 = 3.5 Hz, H-2ax), 1.24 (d,
3H, H3-6), 0.92 (s, 9H, C(CH3)3), 0.11 (s, 3H, Si(CH3)2),
0.10 (s, 3H, Si(CH3)2); 13C NMR (CDCl3): d 98.2 (C-1),
74.8 (C-4), 67.8 (C-3), 63.2 (C-5), 55.3 (OCH3), 35.4
(C-2), 25.94 (C(CH3)3), 25.93 (C(CH3)3), 25.92
(C(CH3)3), 18.3 (C-6), 18.2 (C(CH3)3), ꢀ4.1 (Si(CH3)2),
ꢀ4.5 (Si(CH3)2); HRMS (ESI+) m/z: calcd for
C13H28O4SiNa, 299.1649; found, 299.1638.
Acknowledgments
This work was supported by grants from the Canadian
Institutes of Health Research and the Nova Scotia
Health Research Foundation. We thank Professor T.
Bruce Grindley (Department of Chemistry, Dalhousie
University) for helpful discussions.
Supplementary data
Supplementary data (NMR spectra of compounds 5a–c,
7–8, 10–11, 13, and 15–21) associated with this article
can be found, in the online version, at doi:10.1016/
j.carres.2007.09.012.
References
1. Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2007, 70, 461–
477.
´
´
2. Krˇen, V.; Martınkova, L. Curr. Med. Chem. 2001, 8,
1303–1328.
3. Thorson, J. S.; Hosted, T. J., Jr.; Jiang, J.; Biggins, J. B.;
Ahlert, J. Curr. Org. Chem. 2001, 5, 139–167.
4. Weymouth-Wilson, A. C. Nat. Prod. Rep. 1997, 14, 99–
110.
4.12. Methyl 4-O-tert-butyldimethylsilyl-2,6-dideoxy-3-
O-trimethylsilyl-a-L-ribo-hexopyranoside (21)
5. Rupprath, C.; Schumacher, T.; Elling, L. Curr. Med.
Chem. 2005, 12, 1637–1675.
6. Le, G. T.; Abbenante, G.; Becker, B.; Grathwohl, M.;
Halliday, J.; Tometzki, G.; Zuegg, J.; Meutermans, W.
Drug Discovery Today 2003, 8, 701–709.
Methyl
4-O-tert-butyldimethylsilyl-2,6-dideoxy-a-L-
ribo-hexopyranoside (20) (460 mg, 1.66 mmol) was
dissolved in anhydrous pyridine (8 mL) in a round-
bottomed flask under a nitrogen atmosphere and cooled
to 0–5 ꢁC in an ice-water bath. Trimethylsilyl chloride
7. Seeberger, P. H. Nature 2005, 437, 1239.

2704
S. C. Timmons, D. L. Jakeman / Carbohydrate Research 342 (2007) 2695–2704
´
´ ´
8. Lowary, T. L. Carbohydr. Res. 2006, 341, 1207.
29. Liptak, A.; Imre, J.; Nanasi, P. Carbohydr. Res. 1981, 92,
9. Jakeman, D. L.; Borissow, C. N.; Graham, C. L.;
Timmons, S. C.; Reid, T. R.; Syvitski, R. T. Chem.
Commun. 2006, 3738–3740.
154–156.
30. Brimacombe, J. S. Methods Carbohydr. Chem. 1972, 6,
376–378.
10. Doull, J. L.; Ayer, S. W.; Singh, A. K.; Thibault, P. J.
Antibiot. 1993, 46, 869–871.
31. Horton, D.; Weckerle, W. Carbohydr. Res. 1975, 44, 227–
240.
11. Mallams, A. K.; Puar, M. S.; Rossman, R. R. J. Am.
Chem. Soc. 1981, 103, 3938–3940.
32. Cheung, T.-M.; Horton, D.; Sorenson, R. J.; Weckerle, W.
Carbohydr. Res. 1978, 63, 77–89.
12. Tomita, F.; Tamaoki, T.; Shirahata, K.; Kasai, M.;
Morimoto, M.; Ohkubo, S.; Mineura, K.; Ishii, S. J.
Antibiot. 1980, 33, 668–670.
13. Zielinski, J.; Jereczek, E.; Sowinski, P.; Falkowski, L.;
Rudowski, A.; Borowski, E. J. Antibiot. 1979, 32, 565–
568.
33. Klemer, A.; Balkau, D. J. Chem. Res. 1978, 303, 3823–
3830.
34. Hodgson, R.; Nelson, A. Org. Biomol. Chem. 2004, 2,
373–386.
35. Ko¨pper, S.; Thiem, J. J. Carbohydr. Chem. 1987, 6, 57–85.
36. Morin, C. Chem. Lett. 1986, 1055–1056.
37. Minami, A.; Eguchi, T. J. Am. Chem. Soc. 2007, 129,
5102–5107.
14. Reichstein, T.; Weiss, E. Adv. Carbohydr. Chem. 1962, 17,
65–120.
15. Huan, V. D.; Ohtani, K.; Kasai, R.; Yamasaki, K.; Tuu,
N. V. Chem. Pharm. Bull. 2001, 49, 453–460.
16. Abe, F.; Mori, Y.; Okabe, H.; Yamauchi, T. Chem.
Pharm. Bull. 1994, 42, 1777–1783.
38. Oberthur, M.; Leimkuhler, C.; Kahne, D. Org. Lett. 2004,
¨
6, 2873–2876.
`
39. Veyrieres, A. In Carbohydrates in Chemistry and Biology;
Ernst, B., Hart, G. W., Sinay, P., Eds.; Wiley-VCH Verlag
¨
17. Andreana, P. R.; McLellan, J. S.; Chen, Y.; Wang, P. G.
Org. Lett. 2002, 4, 3875–3878.
GmbH: Weinheim, DE, 2000; Vol. 1, pp 367–405.
40. Shull, B. K.; Wu, Z.; Koreeda, M. J. Carbohydr. Chem.
1996, 15, 955–964.
41. Dixon, J. T.; van Heerden, F. R.; Holzapfel, C. W.
Tetrahedron: Asymmetry 2005, 16, 393–401.
42. Czernecki, S.; Vijayakumaran, K.; Ville, G. J. Org. Chem.
1986, 51, 5472–5475.
18. Braun, M.; Moritz, J. Synlett 1991, 750–752.
19. Fronza, G.; Fuganti, C.; Grasselli, P.; Pedrocchi-Fantoni,
G.; Zirotti, C. Tetrahedron Lett. 1982, 4143–4146.
20. Toshima, K.; Yoshida, T.; Mukaiyama, S.; Tatsuta, K.
Carbohydr. Res. 1991, 222, 173–188.
21. Bock, K.; Lundt, I.; Pedersen, C. Acta Chem. Scand. 1984,
B38, 555–561.
22. Brimacombe, J. S.; Hanna, R.; Saeed, M. S.; Tucker, L. C.
N. J. Chem. Soc., Perkin Trans. 1 1982, 2583–2587.
23. Klemer, A.; Rodemeyer, G. Chem. Ber. 1974, 107, 2612–
2614.
43. Luche, J.-L. J. Am. Chem. Soc. 1978, 100, 2226–2227.
44. Cousins, R. P. C.; Pritchard, R. G.; Raynor, C. M.; Smith,
M.; Stoodley, R. J. Tetrahedron Lett. 2002, 489–492.
45. Larson, E. R.; Danishefsky, S. J. Am. Chem. Soc. 1983,
105, 6715–6716.
46. Bartoli, G.; Bartolacci, M.; Giuliani, A.; Marcantoni, E.;
Massaccesi, M. Eur. J. Org. Chem. 2005, 14, 2867–
2879.
24. Horton, D.; Weckerle, W. Carbohydr. Res. 1988, 174, 305–
312.
25. Clode, D. M.; Horton, D.; Weckerle, W. Carbohydr. Res.
1976, 49, 305–314.
47. Corey, E. J.; Venkateswarlu, A. J. Am. Chem. Soc. 1972,
94, 6190–6191.
26. Winkle, M. R.; Lansinger, J. M.; Ronald, R. C. J. Chem.
Soc., Chem. Commun. 1980, 87–88.
48. Corey, E. J.; Snider, B. B. J. Am. Chem. Soc. 1972, 94,
2549–2550.
´
´
27. Harangi, J.; Liptak, A.; Szurmai, Z.; Nanasi, P. Carbo-
49. Uchiyama, T.; Hindsgaul, O. J. Carbohydr. Chem. 1998,
17, 1181–1190.
hydr. Res. 1985, 136, 241–248.
28. Lemieux, R. U.; Nagabhushan, T. I.; Paul, B. Can. J.
Chem. 1972, 50, 773–776.
50. Paulsen, H.; Bunsch, H. Chem. Ber. 1978, 111, 3484–
¨
3496.