Synthesis of mono- and dideoxygenated α,α-trehalose analogs

Article abstract of DOI:10.1016/j.carres.2007.05.009

In this work, we describe the synthesis and NMR characterization of four mono- and four dideoxygenated analogs of α,α-d-trehalose. The symmetrical (2,2′-, 3,3′-, 4,4′- and 6,6′-) dideoxy analogs were obtained via selective protection and subsequent radica

Full text of DOI:10.1016/j.carres.2007.05.009

Carbohydrate Research 342 (2007) 2014–2030
Synthesis of mono- and dideoxygenated a,a-trehalose analogs
Fiona L. Lin,a Herman van Halbeeka and Carolyn R. Bertozzia,b,c,d,
*
aDepartment of Chemistry, University of California, Berkeley, CA 94720, United States
bDepartment of Molecular and Cell Biology, University of California, Berkeley, CA 94720, United States
cHoward Hughes Medical Institute, University of California, Berkeley, CA 94720, United States
dThe Molecular Foundry, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, United States
Received 8 April 2007; received in revised form 2 May 2007; accepted 7 May 2007
Available online 18 May 2007
Abstract—In this work, we describe the synthesis and NMR characterization of four mono- and four dideoxygenated analogs of
a,a-D-trehalose. The symmetrical (2,20-, 3,30-, 4,40- and 6,60-) dideoxy analogs were obtained via selective protection and subsequent
radical deoxygenation of the desired hydroxyl group set. The unsymmetrical (20-, 30-, 40- and 60-) monodeoxy analogs were synthe-
sized by desymmetrization of a,a-trehalose and subsequent deoxygenation under radical conditions. Complete assignment of all 1H
and 13C resonances in the spectra of these deoxytrehaloses was achieved through the extensive use of 2D {1H,1H} and {1H,13C}
correlation NMR experiments. The synthesis of these trehalose analogs sets the stage for future biochemical and NMR-based stud-
ies to probe the substrate interactions of trehalose with the recently identified mycobacterial sulfotransferase Stf0.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Keywords: Deoxytrehalose; Deoxygenation; Mycobacterial sulfotransferase; NMR spectroscopy
1. Introduction
well as oxygen radicals.3 While produced non-constitu-
tively in many organisms in response to changing
environment, interestingly, trehalose biosynthesis in
mycobacteria is constitutive and subjected to constant
turnover,4,5 suggesting a utility for trehalose in these bac-
teria beyond stress response. Mycobacterial trehalose is
found in both the cytoplasm and the cell wall,6 where it
is incorporated into a variety of glycolipids including tre-
halose 6,60-dimycolate (cord factor), sulfolipids (acylated
trehalose-2-sulfate) and lipooligosaccharides.7 Among
the known trehalose-containing metabolites, sulfolipid-
1 (SL-1) (2, Fig. 1) is the most abundant sulfur-contain-
ing lipid in Mycobacterium tuberculosis (accounting for
up to 0.7% of cellular dry weight).8–13 The high abun-
dance of SL-1 suggests that it plays crucial roles in the
survival and/or virulence of M. tuberculosis. Further-
more, in vitro cell-based assays have demonstrated a
strong correlation between the presence of SL-1 and a
range of host immunological responses.14–17
Trehalose [a-D-Glc-(1M1)-a-D-Glc] (1, Fig. 1) is a sym-
metrical, non-reducing disaccharide found in many
organisms from plants and insects to microbes.1 In many
biological systems, trehalose is associated with fortifica-
tion of cells against various types of physiological stres-
ses including osmotic shock,1 heat and cold shock,2 as
First discovered by Dubos and Middlebrook,8 and
later characterized by Goren and co-workers,10–13 SL-1
consists of a trehalose core that is acylated with three
distinct lipids at the 2-, 3-, 6- and 60-positions and is
Figure 1. Structures of a,a-trehalose (1) and sulfolipid-1 (2).
*
Corresponding author. Tel.: +1 510 6431682; fax: +1 510 6432628;
e-mail: crb@berkeley.edu
0008-6215/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2007.05.009

F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
2015
sulfated at the 20-position. Although SL-1 was discov-
ered and characterized several decades ago, the molecu-
lar details of SL-1 biosynthesis have not yet been fully
elucidated. Our laboratory has recently identified a
novel mycobacterial sulfotransferase, Stf0, which initi-
ates SL-1 biosynthesis by the transfer of a sulfuryl group
from 30-adenosyl-50-phosphosulfate (PAPS) to the 20-
hydroxyl group of trehalose (Scheme 1).18 The lack of
SL-1 precursors in Stf0 knockout mutants provided evi-
dence that this is the obligate first step in SL-1 biosyn-
thesis. Given that trehalose is not biosynthesized in
humans, disruption of SL-1 biosynthesis presents a
potential target for anti-tubercular drug development.
Towards this goal, a detailed understanding of Stf0
substrate recognition is critical. Herein we present the
synthesis of eight deoxygenated trehalose analogs
(Fig. 2), which set the stage for future biochemical stud-
ies of Stf0 substrate recognition.
The chemical manipulation of trehalose, as a C2-sym-
metrical, non-reducing disaccharide, presents two major
challenges, namely: (1) construction of the a,a-1,1 glyco-
sidic linkage, and (2) differentiation between the chemi-
cally similar secondary hydroxyl groups. Synthetic
trehalose analogs that have been reported to date pri-
marily contain modifications of the C4 and C6 hydroxyl
groups. Synthetic manipulation of these positions is
facilitated by the use of the benzylidene protecting
group, and the unique reactivity of the primary C6
hydroxyl residue can furthermore be selectively
targeted.19–28 Reports on the modifications of the C2
and C3 hydroxyl groups are rare, except for a series of
papers published by Hough, Richardson and co-workers
on the use of trehalose diepoxides in the synthesis of the
2,20- and 3,30-dideoxytrehalose analogs.29–31 These
compounds, however, do not possess the native stereo-
chemical configuration of trehalose, but rather have an
inverted stereochemistry at the carbon adjacent to the
deoxygenated site. The work reported herein represents
the first systematic approach to synthesize a complete
panel of mono- and symmetrical dideoxygenated treha-
lose analogs (Fig. 2).
Figure 2. Synthetic targets: symmetrical dideoxygenated trehalose
analogs 3–6 and unsymmetrical monodeoxygenated trehalose analogs
7–10.
fully protected and appropriately deoxygenated glucosyl
donors and acceptors would allow for a convergent ap-
proach, but requires the construction of the synthetically
challenging a,a-1,1 linkage and also entails lengthy
preparation of the individual deoxy-glucosyl building
blocks.32 An alternative approach involves desymmetri-
zation of commercially available a,a-trehalose, requiring
judicious choice of orthogonal protecting groups in
order to isolate the desired set of hydroxyl groups.
Removal of the hydroxyl functionality can then be
conveniently achieved by using Robins’ radical deoxy-
genation conditions.33,34 Herein we employ the latter
approach to efficiently construct the proposed panel of
deoxygenated trehalose analogs.
2.1. Synthesis of 6,60-dideoxy (3) and 60-monodeoxy (7)
analogs
The syntheses of 6,60-dideoxytrehalose (3) and 6-deoxy-
trehalose (7) have been previously reported. The reported
synthesis of 7 relies on the selective mono-bromination
of a,a-trehalose;28 unfortunately, in our hands, we only
obtained the dibrominated product when using this
procedure. While the 6,60-dideoxy analog 3 could be
prepared from the dibrominated compound, or via the
selective sulfonylation of the 6- and 60-hydroxyl
groups of a,a-trehalose dihydrate, followed by conver-
sion to the iodide and subsequent reductive dehalogen-
ation,26 we chose an alternative synthetic route that
2. Results and discussion
We considered two synthetic strategies for preparation
of the target deoxytrehalose analogs. Dimerization of
Scheme 1. Stf0-catalyzed sulfation of trehalose in SL-1 biosynthesis.

2016
F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
enables preparation of both the mono- and also
the dideoxygenated products from a common inter-
mediate.
2.2. Synthesis of 4,40-dideoxy (4) and 40-monodeoxy (8)
analogs
The synthesis of compound 3 began with selective silyl-
ation of the primary alcohols of trehalose (Scheme 2).
Global protection of the remaining hydroxyl groups as
benzoyl esters afforded 11, and removal of the two silyl
ethers using TBAF provided 6,60-diol 12. Compound 12
was then activated as thionocarbonate 13 and subjected
to radical deoxygenation conditions. The reaction yields
of the radical cleavage of the primary thionocarbonate
were highly variable, with multiple side products
observed as shown by TLC analysis. Removal of all
benzoyl protecting groups provided 6,60-dideoxytreha-
lose (3) (Scheme 2).
The synthesis of 60-deoxytrehalose (7) also utilized
diol 12. Using 1 equiv of benzoyl chloride, diol 12 was
mono-acylated (Scheme 3). The low overall yield from
12 to 14 was predicated from overbenzoylation of 12.
Nevertheless, the remaining free 60-hydroxyl group was
removed as described for 13 via thionocarbonate 14 to
provide 6-deoxytrehalose (7) in moderate yield.
The known dibenzylidene acetal of trehalose35 15 served
as a common intermediate in the syntheses of the 4,40-;
4-; 2,20- and 3,30-deoxytrehalose analogs. For the sym-
metrical 4,40-dideoxy analog 4, we employed regioselec-
tive reductive cleavage36 of benzylidene acetal 16 using
TfOH/NaCNBH3 to expose the 4- and 40-hydroxyl
groups (Scheme 4). The reaction was very sluggish,
and complete conversion to 17a could not be achieved
even with the addition of large excesses (>15 equiv) of
reagents. The use of HCl instead of TfOH offered no
improvement. By TLC analysis, both the mono- and
dibenzylidene reduction products could be observed,
albeit with very similar Rf values. Since the monoreduc-
tion product 17b could be used to synthesize the 40-
monodeoxy analog 8, we isolated both compounds.
Although we were unable to fully separate 17a and
17b even after extensive column chromatography, sub-
jecting the mixture to thionocarbonylation afforded
18a and 18b that are separable by silica gel chromato-
Scheme 2. Reagents: (a) (i) TBSCl, Py (ii) BzCl, Py (92% over two steps); (b) TBAF, 1:9 AcOH–THF (72%). (c) PTC-Cl, DMAP, MeCN;
(d) (i) Bu3SnH, AIBN (cat.), PhMe; (ii) NaOMe, MeOH (32% over three steps).
Scheme 3. Reagents: (a) (i) BzCl (1.0 equiv), Py; (ii) PTC-Cl, DMAP, MeCN (10% over two steps); (b) (i) Bu3SnH, AIBN (cat.), PhMe; (ii) NaOMe,
MeOH (60% over two steps).

F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
2017
Scheme 4. Reagents: (a) NaH, BnBr, DMF (90%); (b) NaCNBH3, TfOH, THF; (c) pentafluorophenyl thionochloroformate, DMAP, MeCN (10%
for 18a, 10% for 18b over two steps); (d) (i) Bu3SnH, AIBN (cat.), PhMe; (ii) H2, Pd(OH)2/C, MeOH (36% over two steps).
graphy (Scheme 4). Deoxygenation of 18a and deprotec-
tion provided 4,40-dideoxytrehalose 4 in 36% yield over
two steps.
Although monobenzylidene acetal 17b could be
obtained from the incomplete cleavage of dibenzylidene
16 (Scheme 4), it was difficult to achieve consistent yields
for the conversion of 16 to 17b. Therefore, an alternative
route was devised for the targeted synthesis of 4-mono-
deoxytrehalose 8 (Scheme 5). Subjecting dibenzylidene
16 (Scheme 4) to a brief acidic, rather than reductive,
cleavage removed only one of the two benzylidene ace-
tals,19 furnishing 19 as the major product. Silylation of
the primary alcohol of 19 provided 20 in 80% yield.
Interestingly, thionocarbonylation of the 40-hydroxyl
of 20 did not proceed to completion even after multiple
additions of phenyl thionochloroformate (PTC-Cl) and
DMAP, presumably due to the bulky silyl group at
the nearby 60-position. Nonetheless, cleavage of the
semi-purified thionocarbonate under radical deoxygen-
ation conditions afforded the 40-deoxygenated product
21 in 58% yield over two steps. Simultaneous removal
of benzyl and silyl ethers by acidic hydrogenolysis pro-
vided the final monodeoxy analog 8.
2.3. Synthesis of 3,30-dideoxy (5) and 2,20-dideoxy (6)
analogs
For the synthesis of the 3,30-dideoxy analog 5, we first
attempted dimerization of 2-acetoxyglucal via the
in situ generation of glucosyl iodide as previously
reported.37 However, the starting material and product
have similar Rf values, rendering the key dimerization
step difficult to monitor by TLC and the purification
of the product problematic. As an alternative, we
attempted selective benzoylation of the 2- and 20-hydroxyl
Scheme 5. Reagents: (a) (i) BnBr, NaH, DMF; (ii) 2:8 TFA–MeOH,
2 h (40% over two steps); (b) TBSCl, Et3N, CH2Cl2 (80%); (c) (i) PTC-
Cl, DMAP, MeCN; (ii) Bu3SnH, AIBN (cat.), PhMe (58% over two
steps); (d) H2, Pd(OH)2/C, 1:9 AcOH–MeOH (52%).

2018
F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
groups of dibenzylidene acetal 15 using the mild acylat-
ing reagent, benzoyl imidazole, which has been reported
to give the desired 3,30-diol 22 (Scheme 6) as the major
product in 75% yield.38 In our hands, we obtained a
4:1 mixture of 22 and the 2,30-dibenzoate regioisomer,
which were inseparable by silica gel chromatography.
However, treating 15 with 2 equiv of another mild
benzoylating reagent, benzoyl cyanide,39 provided pure
22 with virtually no other contaminating isomers
(45% isolated yield plus starting material) (Scheme 6).
Subsequent conversion of diol 22 to the respective
thionocarbonate, followed by radical deoxygenation
and deprotection, yielded desired analog 5.
Synthesis of the 2,20-dideoxy analog from intermedi-
ate 15 utilizes the regioselective reductive etherification
method as described by Wang et al. (Scheme 7),40 and
excellent regioselectivity was observed for this reaction.
However, in our hands, 24 was obtained in 38% yield for
this three-step procedure, but we observed a significant
amount of unreacted TMS–ether starting material dur-
ing the reductive etherification reaction. Further addi-
tion of reagents led to nonspecific etherification of
both the C2- and C3-hydroxyl groups. Nonetheless, diol
24 was obtained and subjected to thionocarbonylation
and radical deoxygenation to yield compound 25. Glo-
bal deprotection gave the 2,20-dideoxytrehalose analog
6.
group in compound 26 was achieved previously using
DCC–DMAP–palmitic acid in the synthesis of treha-
lose-2-palmitate.41 Consistent with this previous finding,
benzoylation using benzoic acid–DCC–DMAP provided
the desired 20-O-benzoate ester 27 in 80% isolated yield
(Scheme 8). As a side note, benzoylation using benzoyl
chloride (instead of benzoic acid) and DCC–DMAP
produced a 1:1 mixture of the 20- and 30-monobenzoyl
derivatives. The poor regioselectivity observed presum-
ably was due to benzoyl chloride being a more reactive
acylating reagent than DCC-activated benzoic acid.
With the 20-hydroxyl protected, conversion of the 30-
hydroxyl to the corresponding thionocarbonate, fol-
lowed by radical deoxygenation and deprotection,
yielded the 30-deoxy analog 9 (Scheme 8).
To access the 20-monodeoxytrehalose analog 10, we
first attempted direct, selective thionocarbonylation of
the 20-hydroxyl of 26 (Scheme 8). Unfortunately, no
selectivity between the 20- and 30-hydroxyl groups was
observed, presumably due to the high reactivity of the
thionochloroformate. We then focused on a suitable
protecting group for the 30-hydroxyl of compound 27,
which would have to be installed under relatively neutral
conditions due to the base-sensitive 20-O-benzoyl group
and the highly acid-sensitive trans-fused 2,3-O-cyclo-
hexylidene. A few candidates were surveyed, including
the introduction of a benzyl group at the 30-hydroxyl
using benzyl trichloroacetimidate and
a catalytic
amount of trifluoromethanesulfonic acid (TfOH) or
benzyl bromide–Ag2O. The first set of conditions led
to degradation of the starting material, even when using
a catalytic amount (5 mol %) of acid. The Ag2O-cata-
lyzed benzylation was very sluggish, with little product
formation (by TLC) even after heating at 60 ꢁC for three
days. Eventually, we successfully installed a methoxy-
2.4. Synthesis of 3-monodeoxy (9) and 2-monodeoxy (10)
analogs
The synthesis of the two remaining monodeoxy analogs
9 and 10 utilizes the known tricyclohexylidene acetal of
trehalose 26 as a common intermediate (Scheme 8).41,42
The regioselective palmitoylation of the 2-hydroxyl
Scheme 6. Reagents: (a) BzCN, Et3N, MeCN (45%); (b) (i) PTC-Cl, DMAP, MeCN; (ii) Bu3SnH, AIBN (cat.), PhMe (67% over two steps); (c) (i)
NaOMe, MeOH; (ii) TFA–MeOH (14% over two steps).
Scheme 7. Reagents: (a) (i) TMS–Cl, Et3N, MeCN; (ii) Et3SiH, PhCHO, TMSOTf (cat.), CH2Cl2; (iii) 1.0 M TBAF in THF (38% over three steps);
(b) (i) PTC-Cl, DMAP, MeCN; (ii) Bu3SnH, AIBN (cat.), PhMe (51% over two steps); (c) H2, Pd(OH)2/C, AcOH–MeOH (91%).

F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
2019
Scheme 8. Reagents: (a) Cyclohexanone dimethyl ketal, CSA (cat.), DMF (36%); (b) BzOH, DCC, DMAP, CH2Cl2 (80%); (c) (i) PTC-Cl, DMAP,
MeCN; (ii) Bu3SnH, AIBN (cat.), PhMe (55% over two steps); (c) (i) NaOMe, MeOH; (ii) 10% aq HCl–THF (60% over two steps).
Scheme 9. Reagents: (a) MOM-Cl, TBAI, DIPEA (90%); (b) (i) NaOMe, MeOH; (ii) PTC-Cl, DMAP, MeCN (75% over two steps); (c) (i) Bu3SnH,
AIBN (cat.), PhMe; (ii) 10% aq HCl–THF (40% over two steps).
methyl (MOM) group on the 30-hydroxyl to afford com-
pound 29 (Scheme 9). Following the standard procedure
for installing MOM groups using MOM-Cl (in large
including {1H,1H} COSY, {1H,1H} TOCSY, {1H,13C}
HSQC and {1H,13C} HMBC was employed to accom-
plish this task. The strategies utilized for assignment
followed those outlined in the carbohydrate NMR
literature,44,45 with a few modifications tailored to the
spectral complexity of the compounds at hand.
excess) in Hunig’s base,43 the reaction was extremely
¨
slow even when heated at 60 ꢁC. The addition of tetra-
butylammonium iodide (TBAI), however, greatly accel-
erated the reaction rate. Greater than 90% conversion of
27 to 29 could be obtained in 30 min using stoichiome-
tric amounts of TBAI at 60 ꢁC. Removal of the 20-ben-
The complete assignment of the 1H and 13C signals of
20-monodeoxytrehalose 10 is discussed here as an exam-
1
ple. Figure 3a shows the 1D H NMR spectrum of 10.
´
zoyl group of 29 under Zemplen conditions exposed the
Taking into consideration the characteristic patterns
1
20-hydroxyl group, which was subsequently converted to
the corresponding thionocarbonate (30). Radical deoxy-
genation of 30, followed by acidic cleavage of the
cyclohexylidene and the MOM groups, yielded 20-mono-
deoxytrehalose 10. The moderate yields of the last two
steps reflect a partial loss of the fully deprotected, highly
polar disaccharide product during purification.
for H signals from a-D-glucopyranosyl rings, most of
1
the non-overlapping H signals (including the H1–H2,
H4–H5, and H10–H20ax–H20eq–H30–H40 connectivity
trails) were assigned by 2D {1H,1H} COSY and
TOCSY. (2D NMR spectra are available in the Supple-
mentary data). The 13C signals corresponding to the
aforementioned hydrogens then were assigned via 2D
one-bond {1H,13C} correlation (HSQC). However, the
region of the 1H NMR spectrum between 3.6 and
3.8 ppm is very crowded. Unambiguous assignment of
all signals based solely on analysis of their splitting pat-
terns was not possible in this area, due to higher order
effects even at 800 MHz.
2.5. Complete assignment of 1H and 13C NMR resonances
for the mono- and dideoxytrehaloses
1
The assignment of all the H and 13C resonances for
each of the disaccharides proved to be nontrivial, due
1
Sorting of H signals using 13C chemical shifts (by
1
to severe overlap of signals in the H NMR spectra. A
HSQC) helped tremendously, due to the larger chemical
combination of two-dimensional (2D) NMR techniques

2020
F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
1
Figure 3. (a) 1D H NMR spectrum of 20-deoxytrehalose (10) in D2O recorded at 500 MHz. (b) 2D {1H,13C} HMBC spectrum of 10. Cross-peaks
resulting from long-range coupling between H40 and C60, and between H4 and C6, allowed the assignment of these two carbon resonances. (c) 2D
{1H,13C} HSQC spectrum of 10 showing the assignment of H6a0/H6b0 and H6a/H6b via their correlations to the C60 and C6 signals, respectively. (d)
2D {1H,13C} HMBC spectrum of 10, with a small sweep width (10 ppm) in the 13C dimension. Cross-peaks resulting from H10–C50, H1–C3, H40–C50
and H2–C3 long-range couplings allowed the definitive assignment of C50 and C3, the signals of which are just 0.02 ppm apart.
shift dispersion of the 13C signals. Six overlapping proton
resonances corresponding to H3, H50, H6a, H6b, H6a0
and H6b0 occur within this 0.2 ppm window. The two
methylenic C6 carbons were readily identified via a 13C
DEPT-135 experiment. Assignment of these two C6 car-
bons was achieved by analyzing the 2D {1H,13C} HMBC
spectrum of 10 (Fig. 3b). With the H4 and H40 signals
lowing the assignment of C3 and C50, the exact chemical
shifts of the H3 and H50 resonances from the HSQC
spectrum were obtained.
Tables 1–3 summarize the 1H and 13C chemical shifts,
as well as the JHH coupling constants, for trehalose and
the mono- and dideoxygenated analogs synthesized in
1
this study. The H and 13C chemical shifts at the sites
3
assigned (see above), the JCH cross-peak between H40
of deoxygenation in the deoxytrehaloses are shifted up-
field relative to those in unmodified trehalose, consistent
with the removal of the deshielding effects exerted by the
electronegative oxygen atom (Tables 1 and 2). The car-
bons adjacent to the deoxygenation sites are also influ-
enced by deoxygenation, as revealed by a slight upfield
shift of ꢀ5 ppm (Table 2). Furthermore, in the mono-
deoxygenated trehaloses, the 1H and 13C chemical shifts
in the unmodified glucose rings are almost identical to
those observed for trehalose, whereas the values for
the deoxygenated rings are remarkably similar to those
of the corresponding dideoxygenated analog. This
observation indicates that the effects imparted by deoxy-
genation are localized in the modified glucose ring
(Tables 1 and 2). Severe spectral overlap hampered the
and C60, and the cross-peak between H4 and C6 allowed
the unambiguous assignment of C6 and C60. Conse-
1
quently, the multiplets at 3.82–3.88 ppm in the H spec-
trum were attributed to two H6 protons (H6a and H6a0)
based on cross-peaks between these signals and C6 or
C60, respectively, in the HSQC spectrum (Fig. 3c). Protons
H6b and H6b0 were assigned in an analogous fashion, by
virtue of their correlations to C6 and C60 (Fig. 3c).
As for the remaining two signals, not only were H3
1
and H50 found to overlap in the H spectrum, but their
corresponding 13C resonances also were very close to
each other, differing by only 0.02 ppm. Definitive assign-
ment of C3 and C50 was based on the C3–H1 and C50–
H40 cross-peaks in the HMBC spectrum (Fig. 3d). Fol-

F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
2021
Table 1. 1H chemical shifts (in ppm) of a,a-trehalose (1) and deoxytrehalose analogs 3–10
H-1
H-2
H-3
H-4
H-5
H-6a H-6b H-10 H-20
H-30
H-40
H-50 H-6a0 H-6b0
a,a-Trehalose (1)
6-Dideoxy (3)
4-Dideoxy (4)
5.18 3.64 3.84 3.43 3.81 3.85
3.75
—
—
—
5.09
5.20
—
3.64
3.55
—
3.77
4.08
—
3.18
2.00 (eq) 4.08
1.46 (ax)
—
3.86
—
1.26
3.66
—
—
—
—
—
—
—
—
—
—
—
—
—
—
3.58
3.71
3.76
3-Dideoxy (5)
2-Dideoxy (6)
—
—
—
—
—
—
—
—
—
—
—
—
—
—
5.13
5.25
3.88
2.20 (eq) 3.65
1.90 (ax)
3.67
3.65
3.88
3.84
3.84
2.16 (eq) 3.97
1.73 (ax)
3.65
3.56
3.38
6-Monodeoxy (7)
4-Monodeoxy (8)
5.15 3.63 3.83 3.43 3.80 3.84
5.18 3.63 3.84 3.43 3.83 3.86
3.75
3.75
5.12
5.21
3.78
4.09
3.18
2.01 (eq) 4.09
1.47 (ax)
1.26
3.67
—
3.58
3-Monodeoxy (9)
5.22 3.64 3.70 3.45 3.78 3.84
3.74
3.74
5.08
5.29
3.86
2.19 (eq) 3.65
1.88 (ax)
3.69
3.83
3.86
3.70
3.77
2-Monodeoxy (10) 5.14 3.60 3.76 3.42 3.67 3.84
2.19 (eq) 4.04
1.76 (ax)
3.38
3.76
Prime denotes deoxygenated glucose ring.
Table 2. 13C chemical shifts (in ppm) of a,a-trehalose (1) and deoxytrehalose analogs 3–10
C-1
C-2
C-3
C-4
C-5
C-6
C-10
C-20
C-30
C-40
C-50
C-60
a,a-Trehalose (1)
6-Dideoxy (3)
4-Dideoxy (4)
3-Dideoxy (5)
2-Dideoxy (6)
6-Monodeoxy (7)
4-Monodeoxy (8)
3-Monodeoxy (9)
2-Monodeoxy (10)
93.84
—
—
—
—
93.96
92.90
93.76
93.89
71.66
—
—
—
—
71.70
70.92
71.72
71.67
73.14
—
—
—
—
73.18
72.38
73.16
73.33
70.32
—
—
—
—
70.34
69.53
70.29
70.35
72.76
—
—
—
—
72.79
71.90
72.65
72.87
61.16
—
—
—
—
61.15
60.35
61.15
61.14
—
—
—
—
—
—
93.91
94.33
92.67
93.13
93.83
93.64
92.67
92.97
71.95
73.52
66.95
36.97
71.94
72.64
66.78
37.02
72.94
69.77
34.50
68.74
72.93
69.04
34.42
68.60
75.86
34.82
64.91
71.61
75.85
34.01
64.83
71.67
68.84
67.37
73.40
73.39
68.84
66.49
73.34
73.36
17.21
64.28
61.25
61.30
17.20
63.44
61.25
61.30
Prime denotes deoxygenated glucose ring.
Table 3. JHH coupling constants (in Hz) of a,a-trehalose (1) and deoxytrehalose analogs 3–10
Compound
1H Coupling constants, J (Hz)
a,a-Trehalose (1)
6,60-Dideoxy (3)
4,40-Dideoxy (4)
3,30-Dideoxy (5)
2,20-Dideoxy (6)
3.8 (J1,2), 9.8 (J2,3), 9.8 (J3,4), 9.8 (J4,5), 2.0 (J5,6a), 5.1 (J5,6b), 11.9 (J6a,6b
)
0
0
0
0
0
0
0
0
0
0
3.8 ðJ1 ;2 Þ, 9.6 ðJ2 ;3 Þ, 9.6 ðJ3 ;4 Þ, 9.6 ðJ4 ;5 Þ, 6.2 ðJ5 ;6
Þ
a
a
0
0
0
0
0
0
0
0
0
0
0
0
3.8 ðJ1 ;2 Þ, 9.8 ðJ2 ;3 Þ, 12.7 ðJ3 ;4ax0 Þ, 12.7 ðJ4eq ;4ax Þ, 12.7 ðJ4ax0;5 Þ, 2.1 ðJ3 ;4eq0 Þ or ðJ4eq0;5 Þ, 5.2 ðJ3 ;4eq0 Þ or ðJ4eq0;5 Þ
b
0
0
0
3:4ðJ1 ;2 Þ, 12.0 ðJ6a0;6b
Þ
0
0
0
0
0
0
0
0
0
0
0
0
0
3:7ðJ1 ;2ax0 Þ, 1.0 ðJ
Þ, 11.8 ðJ2ax0;3 Þ, 5.1 ðJ2eq0;3 Þ, 13.4 ðJ2ax ;2eq Þ, 9.6 ðJ3 ;4 Þ, 9.6 ðJ4 ;5 Þ, 2.3 ðJ5 ;6a0 Þ, 5.5 ðJ5 ;6b Þ,
1 ;2eq0
0
12.2 ðJ6a0;6b
Þ
60-Monodeoxy (7)
40-Monodeoxy (8)
3.8 (J1,2), 9.8 (J2,3), 9.8 (J3,4), 9.8 (J4,5), 3.8 ðJ1 ;2 Þ, 9.8 ðJ2 ;3 Þ, 9.8 ðJ3 ;4 Þ, 9.8 ðJ4 ;5 Þ, 6.2 ðJ5 ;6
Þ
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3.7 (J1,2), 9.6 (J2,3), 9.6 (J3,4), 9.6 (J4,5), 5.3 (J5,6b), 12.1 (J6a,6b), 3.8 ðJ1 ;2 Þ, 9.8 ðJ2 ;3 Þ, 12.7 ðJ3 ;4ax0 Þ, 12.7 ðJ4ax ;4eq Þ,
c
c
0
0
0
0
0
0
0
0
0
12.7 ðJ4ax0;5 Þ, 2.0 ðJ3 ;4eq0 Þ or ðJ4eq0;5 Þ, 5.5 ðJ3 ;4eq0 Þ or ðJ4eq0;5 Þ, 3.1 ðJ5 ;6a0 Þ, 6.4 ðJ5 ;6b Þ, 12.1 ðJ6a0;6b
Þ
30-Monodeoxy (9)
2-Monodeoxy (10)
3.8 (J1,2), 9.7 (J2,3), 9.7 (J3,4), 3.6 ðJ1 ;2 Þ, 4.5 ðJ2 ;3eq0 Þ, 11.2 ðJ
Þ, 11.2 ðJ3eq ;3ax Þ, 4.5 ðJ3eq0;4 Þ, 11.2 ðJ3ax0;4 Þ
0
0
0
0
0
0
0
0
2 ;3ax0
0
0
0
0
3.8 (J1,2), 9.7 (J2,3), 9.7 (J3,4), 9.7 (J4,5), 2.2 (J5,6a), 5.5 (J5,6b), 3.8 ðJ
Þ, 1.0 ðJ
Þ, 5.2 ðJ2eq0;3 Þ, 11.8 ðJ2ax0;3 Þ,
1 ;2ax0
1 ;2eq0
0
0
0
0
0
5.1 ðJ2eq0;3 Þ, 13.3 ðJ2eq ;2ax Þ, 9.6 ðJ3 ;4 Þ
Prime denotes deoxygenated glucose ring.
Some coupling constants cannot be measured due to overlap of signals in the spectrum.
a H-30 and H-50 have the same chemical shift values; therefore, the coupling between H-4eq0 and these two protons cannot be distinguished.
b Due to spectral overlap and the complex appearance of H-3ax0 and H-3eq0, only two coupling constants can be measured for this compound. This
observation is consistent with previously reported spectral data in Ref. 37.
c H-30 and H-50 have the same chemical shift values; therefore, the coupling between H-4eq0 and these two protons cannot be distinguished.
measurement of all 1H–1H coupling constants; however,
the coupling constants measured for the deoxytrehaloses
are very similar to those measured for trehalose (Table
3). Presumably, the removal of one or two hydroxyl
groups does not affect the predominant solution confor-
mations of the disaccharide to a large extent.

2022
F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
3. Conclusions
broadband, inverse) probe head with a z-gradient coil.
1D H NMR spectra of trehalose and its deoxygenated
1
This study represents the first systematic synthesis of a
panel of mono- and dideoxygenated a,a-trehalose ana-
logs, four of which (6, 8, 9 and 10) have not been previ-
ously reported. NMR experiments of these compounds
have furthermore allowed for the full assignment of all
1H and 13C resonances for each compound. The
synthetic trehalose analogs will be used in future
experiments to characterize the substrate binding
interactions in the active site of mycobacterial sulfo-
transferase Stf0.
analogs were also recorded at 800 MHz and ambient
temperature, using a Bruker Avance 800 spectrometer
equipped with a 5-mm TCI cryoprobe. 1D 13C and
DEPT experiments were performed at ambient temper-
ature on a Bruker DRX-500 spectrometer, equipped
with a 5-mm broadband-observe probe head. Data were
processed off-line using Bruker’s TopSpin for MS Win-
dows 2000/XP software (version 1.3) (Bruker BioSpin,
1
Billerica, MA). H chemical shifts are reported relative
to 2,2-dimethyl-silapentane-5-sulfonic acid (DSS) (but
were actually measured by reference to internal acetone
at d 2.225 ppm in D2O) with an accuracy of 0.003 ppm.
13C chemical shifts are referenced to the methyl signal of
acetone in D2O (d CH3 30.89 ppm, relative to DSS).
2D {1H,1H} gs-COSY data sets were collected in
magnitude mode;46–48 2D {1H,1H} TOCSY49,50 data sets
were collected in phase-sensitive mode using the TPPI
method.51 The TOCSY pulse program contained a 30–
120 ms MLEV-17 spin-lock pulse train.52 2D {1H,13C}
HSQC53,54 and HMBC55,56 experiments were performed
with pulsed field gradients for coherence selection, using
echo–antiecho phase-sensitive detection. The 2D data
were typically processed with shifted squared sine bell
functions applied in both the t2 and the t1 dimensions,
along with zero-filling in the t1 dimension.
4. Experimental
4.1. General materials and methods
All chemical reagents were of analytical grade, obtained
from commercial suppliers and used without further
purification. Flash chromatography was performed
using silica gel. Analytical TLC was performed on silica
gel plates, and visualized by charring with ethanolic
H2SO4 or ceric ammonium molybdate (CAM) stain,
and/or by absorbance of UV light at 254 nm. All air-
and moisture-sensitive reactions were performed under
an atmosphere of N2. CH2Cl2, MeCN and toluene were
dried by distilling from calcium hydride. THF was dis-
tilled from sodium metal/benzophenone, dry DMF
was obtained from commercial suppliers and MeOH
was dried over Mgꢁ. All organic extracts were dried over
Na2SO4, filtered and concentrated under reduced pres-
sure using a rotary evaporator.
4.3. Syntheses
The known compounds 15,28 16,24 2440 and 2642 were
prepared according to literature procedures.
Low and high-resolution fast-atom bombardment
(FAB) and low-resolution electrospray ionization (ESI)
mass spectra were obtained at the UC Berkeley Mass
Spectrometry Laboratory. High-resolution electrospray
ionization mass spectra were obtained at the Scripps
Mass Spectrometry Laboratory. Elemental analyses
were obtained at the UC Berkeley Microanalytical
Laboratory.
4.3.1. 6,60-Di-O-tert-butyldimethylsilyl-2,3,4,20,30,40-hexa-
O-benzoyl-a,a-D-trehalose (11). To a solution of anhy-
drous a,a0-trehalose (5.0 g, 15 mmol) in 60 mL of dry
pyridine was added tert-butyldimethylchlorosilane
(4.8 g, 32 mmol) at 0 ꢁC (ice-water bath). A catalytic
amount of DMAP (0.5 g, 5 mmol) was added, and the
resulting pale-yellow reaction mixture was stirred over-
night, after which TLC analysis indicated complete con-
sumption of starting material. The reaction mixture was
cooled to 0 ꢁC, and benzoyl chloride (20.3 mL,
175 mmol) was added. The reaction mixture was
warmed to rt and stirred overnight, after which the reac-
tion was complete as indicated by TLC analysis. The
reaction mixture was diluted with CH2Cl2, and the pre-
cipitate removed by filtration. The filtrate was washed
with H2O and brine. The crude product, a yellow oil,
was purified by flash column chromatography (3:7
EtOAc–hexanes) to yield 16 g (92%) of 11 as white
amorphous solids (Rf 0.45, 2:8 EtOAc–hexanes). 1H
NMR (500 MHz, CDCl3): d ꢂ0.15 (d, 12H, J = 7.5),
0.79 (s, 18H), 3.10–3.17 (m, 4H), 4.00 (d, 2H,
J = 10.0), 5.40 (dd, 2H, J = 10.0, 4.0), 5.64–5.68 (m,
4H), 6.20 (app. t, 2H, J = 10.0), 7.32–7.56 (m, 18H),
4.2. NMR spectroscopy
Prior to NMR studies, hydroxyl hydrogens in the
unprotected disaccharides were exchanged with deute-
rium by repeated dissolution in 99.9% D2O with inter-
mittent lyophilization. Each of the deuterium-
exchanged saccharides (20–40 lmol) was dissolved in
0.6 mL of D2O (99.96% D; Cambridge Isotope Labs,
Andover, MA) and subjected to 1H and 13C NMR anal-
ysis. The majority of 1D and 2D NMR experiments,
unless otherwise noted, were performed at 27 ꢁC on a
Bruker Avance 500 spectrometer, interfaced with a
Hewlett–Packard workstation running XWinNMR 3.5
under Linux 7.1, using a 5-mm TBI (triple-channel,

F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
2023
7.80 (d, 4H, J = 7.5), 7.94 (d, 4H, J = 7.0), 8.14 (d, 4H,
J = 7.5); 13C NMR (125 MHz, CDCl3): d ꢂ5.46, ꢂ5.47,
18.5, 26.0, 60.9, 68.2, 70.9, 71.2, 71.8, 93.1, 128.49,
128.52, 128.9, 129.1, 129.5, 129.6, 129.98, 130.02,
130.2, 133.2, 133.4, 133.7, 164.8, 165.5, 165.9.
LRFABMS (m/z): [M+Li]+ calcd for C66H74O17Si2Li,
1201.5; found, 1201.4. Anal. Calcd for C66H74O17Si2:
C, 66.31; H, 6.24. Found: C, 66.06; H, 6.12.
165.5, 165.8, 194.8. HRFABMS (m/z): [M+Li]+ calcd
for C68H54O19S2Li, 1245.2861; found, 1245.2875.
4.3.4. 6,60-Dideoxy-a,a-D-trehalose (3). Compound 13
(0.60 g, 0.48 mmol) and azobisisobutylonitrile (AIBN)
(0.032 g, 0.19 mmol) were dissolved in 10 mL of dry tol-
uene. Tributyltin hydride (Bu3SnH) (1.4 mL, 4.8 mmol)
was added in one portion. The solution was heated at
reflux for 30 min, after which TLC analysis indicated
complete consumption of the starting material. The
reaction mixture was cooled to rt, and PhMe was
removed under reduced pressure. The resulting residue
was redissolved in 40 mL of MeCN and washed with
pentane to remove excess Bu3SnH and other byprod-
ucts. Following removal of MeCN, the crude product
was resuspended in 20 mL of dry MeOH and 0.18 mL
of a 1.0 M solution of NaOMe in MeOH was added. After
stirring at rt for 12 h, TLC analysis indicated the presence
of a highly polar compound (Rf 0.52, 7:2:1 EtOAc–
MeOH–H2O). The reaction mixture was neutralized by
addition of Dowex 50X-W8 resin (H+ form). After re-
moval of the resin by filtration, the filtrate was concen-
trated to a clear oil as crude product. The crude product
was purified by flash chromatography (7:2:1 EtOAc–
MeOH–H2O) to yield a white residue. Residual water
was removed by lyophilization to yield 20 mg (36%) of
3 as a white, fluffy solid. NMR data are summarized
in Tables 1–3. HRFABMS (m/z): [M+Na]+ calcd for
C12H22O9Na, 333.1162; found, 333.1164.
4.3.2. 2,3,4,20,30,40-Hexa-O-benzoyl-a,a-D-trehalose (12).
Compound 11 (3.42 g, 2.86 mmol) was dissolved in dry
THF (50 mL) with 10% (v/v) AcOH under N2. A
1.0 M solution of TBAF in THF (12 mL, 12 mmol)
was added, and the reaction mixture was stirred and
monitored by TLC. After ꢀ24 h, the reaction appeared
to be complete, and the solvent was removed. The crude
product was purified by flash chromatography (1:1
EtOAc–hexanes) to yield 2.0 g of 12 (72%) as a white
1
solid (Rf 0.3, 1:1 EtOAc–hexanes). H NMR (500 MHz,
CDCl3): d 2.43 (app. t, 2H, J = 7.0), 2.83–2.87 (m,
2H), 3.04–3.08 (m, 2H), 3.83 (app. d, 2H, J = 10.0),
5.32 (dd, 2H, J = 10.5, 4.0), 5.42 (app. t, 2H, J =
10.0), 5.64 (d, 2H, J = 4.0), 6.24 (app. t, 2H, J = 10.0),
7.18–7.39 (m, 16H), 7.50–7.53 (m, 2H), 7.83–7.86 (m,
8H), 7.98 (d, 4H, J = 7.5); 13C NMR (125 MHz,
CDCl3): d 60.5, 68.9, 70.1, 70.66, 71.70, 93.4, 128.6,
128.7, 128.8, 129.3, 129.9, 130.0, 130.3, 133.5, 133.9,
134.0, 165.5, 165.8, 166.4. LRFABMS (m/z): [M+H]+
calcd for C64H46O17, 966.3; found, 967.6. Anal. Calcd
for C64H46O17: C, 67.08; H, 4.80. Found: C, 66.87; H,
4.94.
4.3.5. 2,3,4,6,20,30,40-Hepta-O-benzoyl-60-O-phenoxythio-
carbonyl-a,a-D-trehalose (14). Diol 12 (1.59 g, 1.64
mmol), benzoyl chloride (0.21 mL, 1.8 mmol) and
DMAP (ꢀ10 mg) were dissolved in 10 mL of dry pyri-
dine. The reaction mixture was stirred at rt and moni-
tored by TLC. After ꢀ12 h, two new, less polar spots
in ꢀ1:1 ratio were observed, corresponding to the
mono- and dibenzoylated products, as well as some
unreacted starting material. The reaction mixture was
concentrated and subjected to flash chromatography
(2:8 EtOAc–hexanes). The partially purified mono-
benzoylated product (0.198 g, 0.185 mmol) was treated
with PTC-Cl (30 lL, 0.22 mmol) and DMAP (0.045 g,
0.37 mmol) in 10 mL of dry MeCN. After stirring over-
night, the reaction mixture was concentrated, resus-
pended in EtOAc (50 mL), and the organic layer was
washed with H2O and brine. The crude product was
purified by flash chromatography (3:7 EtOAc–hexanes)
to yield 140 mg (63%) of 14 as a white foam (Rf 0.69,
4.3.3. 6,60-Di-O-phenoxythiocarbonyl-2,3,4,20,30,40-hexa-
O-benzoyl-a,a-D-trehalose (13). Diol 12 (0.84 g, 0.87
mmol) and DMAP (0.43 g, 3.5 mmol) were dissolved
in 40 mL of dry MeCN. Phenyl thionochloroformate
(PTC-Cl) (260 lL, 1.9 mmol) was added to this solution,
causing the solution to become bright yellow. The reac-
tion mixture was stirred at rt overnight, after which the
bright yellow color had faded and TLC analysis indi-
cated complete consumption of starting material. The
solvent was removed, and the resulting pale-yellow resi-
due was subjected to flash chromatography (3:7 EtOAc–
hexanes) to yield 0.70 g (65%) of 13 as a white solid (Rf
0.26, 2:8 EtOAc–hexanes). 1H NMR (500 MHz,
CDCl3): d 3.98 (dd, 2H, J = 12.0, 2.5), 4.07 (dd, 2H,
J = 12.0, 4.0), 4.28–4.30 (m, 2H), 5.47 (dd, 2H, J =
10.5, 4.0), 5.65 (app. t, 2H, J = 10.0), 5.75 (d, 2H,
J = 4.0), 6.28 (app. t, 2H, J = 10.0), 7.10 (app. d, 4H,
J = 8.0), 7.26–7.47 (m, 22H), 7.56 (app. t, 2H, J =
7.5), 7.82 (app. d, 4H, J = 7.5), 7.94 (app. d, 4H,
J = 7.5), 8.13 (app. d, 4H, J = 7.5); 13C NMR
(125 MHz, CDCl3): d 68.4, 68.5, 70.3, 70.4, 71.5, 93.4,
122.2, 126.8, 128.6, 128.7, 128.9, 129.1, 129.3, 129.7,
130.0, 130.2, 130.2, 133.5, 133.8, 134.1, 153.7, 165.2,
1
3:7 EtOAc–hexanes). H NMR (400 MHz, CDCl3): d
3.90 (dd, 1H, J = 12.4, 4.6), 4.02–4.10 (m, 2H), 4.11
(dd, 1H, J = 12.4, 4.0), 4.32–4.39 (m, 2H), 5.51–5.59
(m, 2H), 5.68 (app. t, 1H, J = 9.8), 5.71 (app. t, 1H,
J = 9.8), 5.76 (d, 1H, J = 3.8), 5.79 (d, 1H, J = 3.8),
6.31 (app. dt, 2H, J = 9.8, 2.0), 7.08–7.14 (m, 2H),
7.25–7.51 (m, 21H), 7.53–7.61 (m, 3H), 7.85–7.90 (m,

2024
F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
4H), 8.07–8.11 (m, 6H), 8.12–8.17 (m, 4H); 13C
DEPT135 NMR (100 MHz, CDCl3): d 62.1 (C6, CH2),
68.4, 68.6, 68.8, 68.9, 70.4 (C60, CH2), 70.5, 70.6, 71.4,
71.6, 93.0, 122.2, 126.8, 128.7, 129.0, 129.1, 129.8,
130.0, 130.16, 130.19, 130.21, 133.4, 133.5, 133.8,
134.1, 134.2. HRFABMS (m/z): [M+Li]+ calcd for
C68H54O19SLi, 1213.3140; found, 1213.3141.
oil) was partially purified by flash chromatography
(2:8 EtOAc–hexanes) to yield 4 g of a 1:1 mixture of
17a (Rf 0.56, 3:7 EtOAc–hexanes) and monobenzylidene
17b (Rf 0.62, 3:7 EtOAc–hexanes). The mixture was
taken on to the following step without further purifica-
tion. A small amount (200 mg) of pure 17 was isolated
1
and fully characterized. H NMR (500 MHz, CDCl3):
d 2.58 (d, 2H, J = 3.0), 3.45–3.58 (m, 4H), 3.60 (dd,
2H, J = 9.5, 3.8), 3.70 (m, 2H), 3.92 (app. t, 2H,
J = 9.3), 4.10–4.15 (m, 2H), 4.47 (d, 2H, J = 12.0),
4.54, (d, 2H, J = 12.0), 4.66 (d, 2H, J = 12.0), 4.73 (d,
2H, J = 12.0), 4.83 (d, 2H, J = 11.3), 5.03 (d, 2H,
J = 11.3), 5.28 (d, 2H, J = 3.8), 7.22–7.44 (m, 30H).
13C DEPT135 NMR (100 MHz, CDCl3): d 69.1 (C6,
CH2), 69.3, 70.8 (Bn, CH2), 72.6 (Bn, CH2), 73.7 (Bn,
CH2), 75.4, 79.1, 81.1, 94.1, 126.3, 127.7, 127.80,
127.83, 127.9, 128.5, 128.5, 128.7. HRFABMS (m/z):
[M+Li]+ calcd for C54H58O11Li, 889.4139; found,
889.4141.
4.3.6. 6-Deoxy-a,a-D-trehalose (7). Compound 14
(120 mg, 0.099 mmol) and AIBN (3.0 mg, 0.02 mmol)
were dissolved in 10 mL of dry toluene. Bu3SnH
(0.14 mL, 0.50 mmol) was added in one portion. The
reaction mixture was heated at reflux for 30 min, after
which TLC analysis indicated complete consumption
of starting material. The reaction mixture was cooled
to rt, and toluene was removed under reduced pressure.
The resulting residue was redissolved in 40 mL of
MeCN and washed with pentane (3 · 20 mL) to remove
excess Bu3SnH and other byproducts. The crude prod-
uct was partially purified by silica gel flash chromato-
graphy (2:8 EtOAc–hexanes) to yield 90 mg (87%) of a
white foam. This product (73 mg) was resuspended in
5 mL of anhyd MeOH, and 30 lL of a 1.0 M solution
of NaOMe in MeOH was added. After stirring at rt
for 12 h, TLC analysis indicated the presence of a highly
polar compound (Rf 0.52 in 7:2:1 EtOAc–MeOH–H2O).
The reaction mixture was neutralized by addition of
Dowex 50X-W8 resin (H+ form). After removal of the
resin by filtration, the filtrate was concentrated to a clear
oil as crude product. The crude product was purified by
flash chromatography (7:2:1 EtOAc–MeOH–H2O) to
yield a white residue. Residual water was removed by
lyophilization to yield 16 mg (69%) of 7 as a white, fluffy
solid. NMR data are summarized in Tables 1–3.
HRESIMS (m/z): [M+Na]+ calcd for C12H22O10Na,
349.1105; found, 349.1105.
4.3.8. 4,40-Di-O-pentafluorophenoxy thiocarbonyl-2,3,6,
20,30,60-hexa-O-benzyl-a,a-D-trehalose (18a). The 1:1
mixture of 17a and monobenzylidenated byproduct
17b was dissolved in 50 mL of dry MeCN, to which
DMAP (2.22 g, 18.2 mmol) and pentafluorophenyl
thionochloroformate (2.55 mL, 15.9 mmol) were added.
The resulting orange solution was stirred at rt for
12 h, by which time a white precipitate had formed.
TLC analysis indicated complete conversion of starting
materials to two new compounds. The reaction mixture
was filtered, and the filtrate was adsorbed onto silica gel
and purified by flash chromatography (1:9 to 2:8
EtOAc–hexanes) to yield 1.20 g (9%, two steps) of 18a
as an off-white solid (Rf 0.53, 2:8 EtOAc–hexanes), as
well as 1.0 g (11%, two steps) of the monobenzylide-
nated product 18b (Rf 0.62, 3:7 EtOAc–hexanes). Note
that both compounds decompose slowly upon heating
at 45 ꢁC on a rotary evaporator. 1H NMR of 18a
(500 MHz, CDCl3): 3.24 (dd, 2H, J = 11.1, 4.0), 3.39
(dd, 2H, J = 11.1, 2.0), 3.71 (dd, 2H, J = 9.7, 3.4),
4.24 (app. t, 2H, J = 9.3), 4.35–4.40 (m, 4H), 4.46 (d,
2H, J = 11.9), 4.64, (d, 2H, J = 11.9), 4.75 (d, 2H,
J = 11.8), 4.86 (d, 2H, J = 11.1), 4.90 (d, 2H, J =
11.1), 5.23, (d, 2H, J = 3.4), 5.68 (app. t, 2H, J = 9.7),
7.21–7.38 (m, 30H). 13C DEPT135 NMR (125 MHz,
CDCl3): d 67.9 (C6, CH2), 69.1, 73.6 (C6, Bn), 74.1
(C6, Bn), 75.8 (C6, Bn), 78.7, 79.1, 82.9, 94.0, 127.7,
127.9, 128.0, 128.1, 128.2, 128.49, 128.52, 128.6.
HRFABMS (m/z): [M+Li]+ calcd for C68H56O13F10-
S2Li, 1341.3163; found, 1341.3176. Anal. Calcd for
C68H56F10O13S2: C, 61.17; H, 4.23; S, 4.80. Found: C,
61.03; H, 4.34; S, 4.62.
4.3.7. 2,3,6,20,30,60-Hexa-O-benzyl-a,a-D-trehalose (17a).
Compound 16 (8.5 g, 9.7 mmol) was dissolved in 80 mL
˚
of dry THF, to which 3 A molecular sieves (1 g) and
NaCNBH3 (1.40 g, 22.3 mmol) were added. The result-
ing pale-yellow suspension was stirred at rt for 10 min.
A solution of 1.65 mL (22.3 mmol) of trifluoromethane-
sulfonic acid (TfOH) in 5 mL of THF was added drop-
wise to the reaction mixture, resulting in significant gas
evolution. Addition of TfOH was discontinued when no
more bubbling was observed. The reaction progressed
very slowly as monitored by TLC, even after being stir-
red for several hours. Additional NaCNBH3 and TfOH
(2 equiv each) were added every 2 h, but the reaction
had not progressed any further after addition of an extra
4 equiv of reagents. At this point, the reaction mixture
was filtered through Celite to remove the molecular
sieves and salt, concentrated and redissolved in CH2Cl2.
The organic layer was washed with H2O and brine.
After removal of solvent, the crude product (yellow
4.3.9. 4,40-Dideoxy-a,a-D-trehalose (4). Compound 18a
(1.2 g, 0.90 mmol) was dissolved in 40 mL of dry tolu-
ene, to which AIBN (0.059 g, 0.36 mmol) was added,

F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
2025
followed by Bu3SnH (2.42 mL, 8.99 mmol). The result-
ing solution was heated at reflux for 1 h, at which point
TLC analysis indicated complete consumption of the
starting material. The reaction mixture was cooled to
rt, concentrated and redissolved in 100 mL of MeCN.
The MeCN solution was washed with pentane to
remove Sn-containing byproducts, and the combined
pentane layers were back-extracted with 50 mL of
MeCN. The combined MeCN layers were adsorbed
onto silica gel and subjected to flash chromatography
(2:8 EtOAc–hexanes) to yield 0.66 g of a clear oil. This
oil was dissolved in 10 mL of dry MeOH and subjected
to hydrogenolysis using Pd(OH)2/C. (Caution! Extreme
fire hazard.) The reaction was monitored by TLC, and
hydrogen was bubbled into the solution until all starting
material was consumed. The catalyst was removed by
filtering through Celite, and the filtrate was adsorbed
onto silica gel. Purification by flash chromatography in
7:2:1 EtOAc–MeOH–H2O yielded a white amorphous
solid, from which residual water was removed by lyoph-
ilization to yield 102 mg (36%, two steps) of 4 (Rf 0.30,
7:2:1 EtOAc–MeOH–H2O). NMR data are summarized
in Tables 1–3. HRESIMS (m/z): [M+Na]+ calcd for
C12H22O9Na, 333.1156; found, 333.1155.
7.56 (app. dd, 2H, J = 8.0, 1.4). 13C DEPT135 NMR
(125 MHz, CDCl3): d 62.3 (CH2), 63.2, 69.2 (CH2),
70.5, 71.6, 73.2 (CH2), 73.8 (CH2), 75.48 (CH2), 75.50,
78.8, 78.9, 79.4, 81.2, 82.5, 94.3, 95.0, 101.4, 126.3,
127.8, 127.89, 127.93, 127.98, 128.02, 128.11, 128.14,
128.4, 128.5, 128.6, 128.73, 128.75, 129.1. HRFABMS
(m/z): [M+Li]+ calcd for C47H50O11Li, 797.3513; found,
797.3508.
4.3.11.
2,3,20,30-Tetra-O-benzyl-4,6-benzylidene-60-O-
tert-butyldimethylsilyl-a,a-D-trehalose (20). Compound
19 (950 mg, 1.20 mmol) was dissolved in dry CH2Cl2
(20 mL), and DMAP (14.7 mg, 0.120 mmol) and Et3N
(0.7 mL, 5 mmol) were added to the solution, followed
by TBDMS-Cl (54.3 mg, 3.60 mmol). The resulting
pale-yellow solution was stirred at rt for 48 h, and
TLC analysis indicated conversion of the starting mate-
rial to a less polar product. The reaction mixture was
adsorbed onto silica gel and purified by flash chroma-
tography in 1:9 EtOAc–hexanes to yield 870 mg (80%)
1
of 20 as a clear oil (Rf 0.51, 2:8 EtOAc–hexanes). H
NMR (500 MHz, CDCl3): d 0.079 (s, 3H), 0.93 (s,
9H), 2.57 (br s, 1H), 3.60 (dd, 1H, J = 9.2, 3.4), 3.64–
3.73 (m, 5H), 3.77 (dd, 1H, J = 11.2, 2.8), 3.96 (app. t,
1H, J = 9.2), 4.06 (app. dt, 1H, J = 9.7, 3.6), 4.15 (dd,
1H, J = 9.7, 4.8), 4.18 (app. t, 1H, J = 9.2), 4.28–4.35
(m, 1H), 4.77–4.91 (m, 6H), 5.01 (d, 1H, J = 11.2),
5.05 (d, 1H, J = 11.2), 5.23 (d, 2H, J = 3.5), 5.60 (s,
1H), 7.30–7.46 (m, 23H), 7.54–7.57 (m, 2H). 13C
DEPT135 NMR (125 MHz, CDCl3): d ꢂ5.2, ꢂ5.1,
26.2, 63.1, 63.3 (CH2), 69.2 (CH2), 71.2, 71.6, 73.2
(CH2), 73.6 (CH2), 75.5 (CH2), 75.6 (CH2), 78.8, 79.1,
79.4, 81.3, 82.6, 94.3, 94.9, 101.4, 126.4, 127.81,
127.86, 127.89, 127.95, 127.96, 127.97, 128.2, 128.3,
128.4, 128.5, 128.6, 128.7, 128.8, 129.1. HRFABMS
(m/z): [M+Li]+ calcd for C53H64O11SiLi, 911.4378;
found, 911.4354.
4.3.10. 2,3,20,30-Tetra-O-benzyl-4,6-O-benzylidene-a,a-
D-trehalose (19). Compound 15 (2.0 g, 3.9 mmol) was
suspended in 40 mL of dry DMF and cooled to 0 ꢁC
in an ice/water bath. NaH (60 wt %, 0.93 g, 32 mmol)
was added, and the resulting mixture was stirred for
10 min. Benzyl bromide (2.2 mL, 18 mmol) was added
dropwise over 5 min, and the reaction mixture was stir-
red at rt for 3 h. At this point the reaction appeared to
be complete by TLC analysis, and water was added to
quench excess NaH. This reaction mixture was diluted
with EtOAc and washed with H2O and brine. The or-
ganic layer was dried over Na2SO4 and concentrated
to yield a pale-yellow residue. This crude product (Rf
0.73, 2:8 EtOAc–hexanes) was redissolved in 20 mL of
9:1 MeOH–CH2Cl2, and trifluoroacetic acid (TFA)
(2 mL) was added. After 3 h, TLC anaylsis indicated
the formation of two new products along with unreacted
starting material. The acid was neutralized with solid
NaHCO3. Water was added, and the resulting mixture
was extracted with EtOAc, which was then washed with
water and brine. The organic layer was adsorbed onto
silica gel, after which flash chromatography with 1:1
EtOAc–hexanes yielded 1.22 g (40%, two steps) of 19
as a white foam (Rf 0.31, 2:98 MeOH–CH2Cl2). 1H
NMR (500 MHz, CDCl3): d 1.94 (br s, 1H), 2.66 (br s,
1H), 3.58 (dd, 1H, J = 9.8, 3.5), 3.61–3.73 (m, 6H),
3.93 (app. t, 1H, J = 9.2), 4.05 (m, 1H), 4.14–4.20 (m,
2H), 4.27–4.32 (m, 1H), 4.73–4.76 (m, 3H), 4.83 (app.
t, 1H, J = 12.0), 4.90 (d, 1H, J = 11.2), 5.01 (d, 1H,
J = 11.2), 5.05 (d, 2H, J = 11.3), 5.18 (d, 1H, J = 3.7),
5.20 (d, 1H, J = 3.4), 5.60 (s, 1), 7.28–7.48 (m, 23H),
4.3.12. 2,3,20,30-Tetra-O-benzyl-4,6-benzylidene-40-deoxy-
60-O-tert-butyldimethylsilyl-a,a-D-trehalose (21). Com-
pound 20 (1.33 g, 1.47 mmol) was dissolved in dry
MeCN (40 mL), and DMAP (0.717 g, 5.87 mmol) and
PTC-Cl (0.40 mL, 2.9 mmol) were added to this solu-
tion. The reaction mixture was stirred at rt overnight,
after which TLC analysis indicated the formation of a
new, slightly less polar product as well as unreacted
starting material. Complete consumption of the starting
material could not be achieved even after adding more
PTC-Cl and DMAP. At this point, the reaction mixture
was concentrated and adsorbed onto silica gel and par-
tially purified by flash chromatography (1:9 EtOAc–hex-
anes). This semi-purified product was dissolved in dry
toluene (60 mL), and AIBN (50 mg, 0.30 mmol) and
Bu3SnH (1.9 mL, 6.0 mmol) were added. The resulting
solution was heated at reflux for 15 min. After cooling
to rt, the crude reaction mixture was adsorbed onto

2026
F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
silica gel and purified by flash chromatography (1:9
EtOAc–hexanes) to yield 743 mg (58%, two steps) of
21 as an amorphous white solid (Rf 0.56, 2:8 EtOAc–
hexanes). 1H NMR (500 MHz, CDCl3): d 0.034 (s,
3H), 0.041 (s, 3H), 0.89 (s, 9H), 1.51 (dd, 1H, J = 13,
12.2), 2.05–2.11 (m, 1H), 3.51–3.53 (m, 3H), 3.61–3.69
(m, 3H), 4.03–4.09 (m, 1H), 4.10–4.20 (m, 3H), 4.27–
4.33 (m, 1H), 4.68–4.78 (m, 5H), 4.84 (d, 2H,
J = 11.1), 4.95 (d, 1H, J = 11.2), 5.21 (d, 1H, J = 2.4),
5.22, (d, 1H, J = 3.2), 5.56 (s, 1H), 7.24–7.43 (m,
23H), 7.51 (dd, 2H, J = 8.2, 2.1). 13C DEPT135 NMR
(125 MHz, CDCl3): d ꢂ5.1, ꢂ3.3, 26.2, 33.5 (CH2),
62.9, 65.9 (CH2), 69.0, 69.3 (CH2), 72.4, 73.56 (CH2),
73.59 (CH2), 75.5 (CH2), 75.7, 78.8, 79.4, 80.5, 82.7,
94.2, 94.9, 101.5, 126.4, 127.8, 127.9, 128.0, 128.1,
128.2, 128.4, 128.5, 128.6, 128.7, 129.1. HRFABMS
(m/z): [M+Li]+ calcd for C53H64O10SiLi, 895.4429;
found, 895.4411.
102.1, 126.7, 128.4, 129.1, 129.6, 130.1, 133.9. Lit.:57
1H NMR (400 MHz): d 2.66 (br s, 2H), 3.49 (t, 2H,
J = 10.0), 3.56 (t, 2H, J = 9.6), 3.60 (dd, 2H, J = 4.8,
10.0), 3.86 (dt, 2H, J = 4.8, 10.0), 4.39 (t, 2H, J = 9.6),
5.11 (dd, 2H, J = 4.0, 9.6), 5.40 (s, 2H), 5.48 (d, 2H,
J = 4.0), 7.13–7.64 (m, 16H), 8.16 (d, 4H, J = 7.2).
4.3.15. 2,20-Di-O-benzoyl-4,6:40,60-di-O-benzylidene-3,30-
dideoxy-a,a-D-trehalose (23). Compound 22 (0.300 g,
0.413 mmol) and DMAP (0.202 g, 1.65 mmol) were dis-
solved in dry MeCN (20 mL), and PTC-Cl (0.12 mL,
0.91 mmol) was added. The resulting bright-yellow solu-
tion was stirred at rt overnight, after which all starting
material had been consumed as shown by TLC analysis;
two major products had formed (Rf 0.46 and 0.58, 3:7
EtOAc–hexanes), presumably the mono- and diacylated
species. The reaction mixture was adsorbed onto silica
gel, after which purification by flash chromatography
with 2:8 EtOAc–hexanes afforded 0.33 g (80%) of the
Rf 0.58 spot as a white powder. This thionocarbonate
(0.276 g, 0.277 mmol) was dissolved in 50 mL of dry tol-
uene, AIBN (9 mg, 0.005 mmol) and Bu3SnH (0.75 mL,
2.8 mmol) were added, and the solution was heated at
reflux for 30 min. TLC analysis indicated complete con-
version of the starting material to a new, less polar prod-
uct. The reaction mixture was adsorbed onto silica gel,
after which purification by flash chromatography (1:9
to 2:8 EtOAc–hexanes) yielded 0.161 g (84%) of 23 as
4.3.13. 4-Deoxy-a,a-D-trehalose (8). Compound 21
(360 mg, 0.41 mmol) was dissolved in 20 mL of 9:1
MeOH–EtOAc with 10% AcOH (v/v). The solution
was purged under N2 for 10 min, and subjected to
hydrogenolysis over Pd(OH)2/C. (Caution! Extreme fire
hazard.) After 3.5 h, most of the starting material had
been converted to two new, more polar spots (Rf 0.76
and 0.28, 7:2:1 EtOAc–MeOH–H2O). An additional
0.5 mL of AcOH was added to the reaction mixture,
and the reaction vessel was heated at 50 ꢁC overnight.
The reaction appeared to be complete, and the reaction
mixture was filtered through Celite to remove the Pd
catalyst. The filtrate was concentrated and purified by
flash chromatography (7:2:1 EtOAc–MeOH–H2O) to
yield a clear oil, from which residual water was removed
by lyophilization to yield 55 mg (45%) of 8 as a fluffy,
white solid. NMR data are summarized in Tables 1–3.
HRFABMS (m/z): [M+Na]+ calcd for C12H22O10Na,
349.1111; found, 349.1119.
1
a white powder. H NMR (500 MHz, CDCl3): d 2.38
(app. q, 2H, J = 11.7), 2.41–2.46 (m, 2H), 3.56 (app. t,
2H, J = 10.4), 3.70–3.77 (m, 2H), 3.75 (dd, 2H,
J = 10.6, 4.8), 3.97 (app. td, 2H, J = 10.0, 4.8), 5.23–
5.27 (m, 2H), 5.42 (d, 2H, J = 3.4), 5.47 (s, 2H), 7.34–
7.43 (m, 14H), 7.54 (app. tt, 2H, J = 7.5, 1.3), 8.15
(app. dd, 4H, J = 8.5, 1.3). 13C DEPT135 NMR
(125 MHz, CDCl3): d 29.8 (CH2), 65.1, 69.0 (CH2),
69.2, 76.2, 92.7, 101.9, 126.6, 128.4, 129.0, 129.3,
130.0, 133.8. HRFABMS (m/z): [M+Li]+ calcd for
C40H38O11Li, 701.2574; found, 701.2565.
4.3.14. 2,20-Di-O-benzoyl-4,6:40,60-di-O-benzylidene-a,a-
D-trehalose (22). To a solution of 5.0 g (9.6 mmol) of
15 in 100 mL of dry MeCN was added 2.3 mL
(19 mmol) of benzoyl cyanide followed by 0.5 mL
(3 mmol) of Et3N. After stirring for 2 h, the yellow reac-
tion mixture was adsorbed onto silica gel and subjected
to flash chromatography using 5:95 to 10:90 EtOAc–tol-
uene to provide 3.15 g (45%) of the desired dibenzoate
22 (Rf 0.28, 15:85 EtOAc–toluene), essentially free of
other contaminating isomers. 1H NMR (500 MHz,
CDCl3) d 2.54 (d, 2H, J = 3.2), 3.49 (app. t, 2H,
J = 10.0), 3.56 (app. t, 2H, J = 9.6), 3.60 (dd, 2H,
J = 4.8, 10.0), 3.86 (app. dt, 2H, J = 4.8, 10.0), 4.39
(app. t, 2H, J = 9.6), 5.11 (dd, 2H, J = 4.0, 9.6), 5.40
(s, 2H), 5.48 (d, 2H, J = 4.0), 7.13–7.64 (m, 16H), 8.16
(app. d, 4H, J = 7.2). 13C DEPT135 NMR (125 MHz,
CDCl3) d 63.2, 68.4 (CH2), 69.0, 73.7, 81.2, 93.9,
4.3.16. 3,30-Dideoxy-a,a-D-trehalose (5). Compound 23
(0.160 g, 0.230 mmol) was dissolved in dry MeOH
(10 mL)–THF (2 mL), and a solution of 0.5 M NaOMe
in MeOH (5 mL) was added. The reaction mixture was
stirred at rt for 2 h, at which point TLC analysis indi-
cated complete consumption of the starting material
(Rf of product 0.31, 3:7 EtOAc–hexanes). The solvent
was removed to yield a residue, which was redissolved
in EtOAc and washed with H2O and brine. The organic
layer was concentrated and redissolved in 1:1 MeOH–
TFA. After being stirred at rt for 10 min, complete con-
sumption of the starting material was shown by TLC,
and the reaction mixture was adsorbed onto silica gel.
Purification by flash chromatography (7:2:1 EtOAc–
MeOH–H2O) yielded 13 mg (19%, two steps) of 5 as a
white solid following lyophilization (Rf = 0.56, 7:2:1

F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
2027
EtOAc–MeOH–H2O). NMR data are summarized in
Tables 1–3. HRFABMS (m/z): [M+Na]+ calcd for
C12H22O9Na, 333.1162; found, 333.1170. Lit.:37 1H
NMR (300 MHz, MeOH-d4): d 5.08 (d, 1H, H-1), 3.8–
3.5 (unresolved m, 5H, H-2, H-4, H-5 and two H-6),
2.09 (m, 1H, H-3b), 1.94 (m, 1H, H-3a, J1,2 = 3.4).
carbodiimide (DCC) (0.269 g, 1.26 mmol) and DMAP
(0.141 g, 1.16 mmol) were dissolved in 25 mL of dry
CH2Cl2. The solution was stirred at rt overnight, after
which time a white precipitate was observed and TLC
analysis indicated almost complete conversion of the
starting material to a new product. The reaction mixture
was concentrated, re-suspended in ice-cold acetone and
filtered to remove insoluble urea byproduct. The filtrate
was concentrated to yield an off-white residue, which
was purified by flash chromatography (1:9 EtOAc–hex-
anes) to yield 0.58 g (80%) of 27 as a white solid (Rf 0.36,
4.3.17. 2,20-Dideoxy-3,30-di-O-benzyl-4,6:40,60-di-O-benz-
ylidene-a,a-D-trehalose (25). Compound 24 (0.650 g,
0.931 mmol) was dissolved in dry MeCN (30 mL), and
DMAP (0.546 g, 4.47 mmol) and PTC-Cl (300 lL,
2.23 mmol) were added. The resulting bright-yellow
solution was stirred at rt overnight, after which a white
precipitate was observed and TLC analysis indicated
complete conversion of the starting material to a new
product. The precipitate was removed by filtration,
and the filtrate was absorbed onto silica gel. Flash chro-
matography with 2:8 EtOAc–hexanes yielded 510 mg
(56%) of a white foam. This product (0.40 g, 0.41 mmol)
was dissolved in 60 mL of dry toluene, and AIBN
(33 mg, 0.20 mmol) and Bu3SnH (1.0 mL, 3.7 mmol)
were added. The resulting clear solution was heated at
reflux for 30 min, after which TLC analysis indicated
complete consumption of the starting material. The
reaction mixture was cooled to rt, absorbed onto silica
gel and subjected to flash chromatography (2:8
EtOAc–hexanes) to yield 125 mg (91%) of 25 as a white
amorphous solid (Rf 0.32, 2:8 EtOAc–hexanes), contam-
inated with a small amount of Sn-containing byprod-
ucts. 1H NMR (500 MHz, CDCl3): d 1.84–1.90 (m,
2H), 2.23 (app. dd, 2H, J = 14.0, 4.9), 3.71–3.83 (m,
6H), 4.00–4.10 (m, 2H), 4.22–4.28 (m, 2H), 4.71 (d,
2H, J = 11.9), 4.86 (d, 2H, J = 11.9), 5.20 (d, 2H, J =
3.5), 5.65 (s, 2H), 7.26–7.43 (m, 16H), 7.52–7.55 (m,
4H). 13C DEPT135 NMR (125 MHz, CDCl3): d 36.5
(CH2), 64.3, 69.4 (CH2), 73.1 (CH2), 73.4 (CH2), 84.1,
93.7, 101.8, 126.5, 128.09, 128.13, 128.7, 128.9, 129.4.
HRFABMS (m/z): [M+Li]+ calcd for C40H42O9Li,
673.2989; found, 673.3002.
1
2:8 EtOAc–hexanes). H NMR (400 MHz, CD3CN): d
1.45–2.06 (m, 30H), 3.12 (dd, 1H, J = 6.5, 6.0), 3.40–
3.43 (m, 1H), 3.51–3.58 (m, 3H), 3.71 (app. t, 1H, J =
12.0), 3.74–3.86 (m, 2H), 3.89 (app. t, 1H, J = 11.5),
3.90–3.97 (m, 1H), 4.03 (app. t, 1H, J = 11.5), 4.04–
4.09 (m, 1H), 4.93 (dd, 1H, J = 12.0, 4.5), 5.35 (d, 1H,
J = 4.0), 5.37 (d, 1H, J = 4.5), 7.52 (app. t, 2H, J =
9.5), 7.64 (app. t, 1H, J = 9.5), 8.12 (app. d, 2H,
J = 9.5); 13C DEPT135 NMR (125 MHz, CD3CN): d
23.2 (CH2), 24.5 (CH2), 26.0 (CH2), 26.2 (CH2), 28.5
(CH2), 36.5 (CH2), 36.9 (CH2), 38.4 (CH2), 38.8
(CH2), 61.7 (CH2), 62.1 (CH2), 65.0, 67.1, 69.5, 73.2,
73.6, 74.4, 75.0, 76.7, 94.4, 94.8, 129.7, 130.5, 134.4.
HRFABMS (m/z): [M+Li]+ calcd for C37H50O12Li,
693.3; found, 693.5. Anal. Calcd for C37H50O12: C,
64.71; H, 7.34. Found: C, 64.44; H, 7.52.
4.3.20. 2-O-Benzoyl-4,6:20,30:40,60-tri-O-cyclohexylidene-
3-deoxy-a,a-D-trehalose (28). Compound 27 (1.00 g,
1.46 mmol) and DMAP (0.356 g, 2.91 mmol) were dis-
solved in 50 mL of dry MeCN, and PTC-Cl (240 lL,
1.75 mmol) was added to this solution in one portion.
The resulting bright-yellow solution was stirred at rt
for 12 h, after which time a precipitate had formed.
TLC analysis indicated complete consumption of the
starting material. The reaction mixture was filtered
through Celite, and the filtrate was concentrated and
partially purified by flash chromatography (1:9 to 2:8
EtOAc–hexanes) to yield the thionocarbonate as a white
foam (0.71 g, 59%). This foam (0.21 g, 0.26 mmol) was
dissolved in 20 mL of dry toluene, and AIBN (8 mg,
0.05 mmol) and Bu3SnH (0.34 mL, 1.3 mmol) were
added. The resulting solution was heated at reflux for
30 min, at which time complete conversion of the thi-
onocarbonate (Rf 0.41, 2:8 EtOAc–hexanes) to a new,
slightly less polar spot (Rf 0.51, 2:8 EtOAc–hexanes)
was observed. The reaction mixture was concentrated
to yield a pale-yellow residue that was dissolved in
30 mL of MeCN and washed with pentane to remove
the excess organotin byproducts. The crude product ob-
tained was then purified by flash chromatography (1:9 to
2:8 EtOAc–hexanes) to yield 150 mg (88%) of 28 as a
4.3.18. 2,20-Dideoxy-a,a-D-trehalose (6). Compound 25
(215 mg, 0.320 mmol) was dissolved in 9:1 MeOH–
AcOH (10 mL total volume) and subjected to hydrogen-
olysis using Pd(OH)2/C. (Caution! Extreme fire hazard.)
The reaction was monitored by TLC until all starting
material was consumed. The Pd catalyst was removed
by filtering through Celite, and the filtrate was adsorbed
onto silica gel and purified by flash chromatography
(7:2:1 EtOAc–MeOH–H2O) to yield 91 mg (91%) of 6
as a white, fluffy solid following lyophilization. NMR
data are summarized in Tables 1–3. HRFABMS (m/z):
calcd for C12H22O9Na, 333.1162; found, 333.1167.
1
4.3.19. 2-O-Benzoyl-4,6;20,30;40,60-tri-O-cyclohexylidene-
a,a-D-trehalose (27). Compound 26 (0.612 g, 1.05
mmol), benzoic acid (0.128 g, 1.05 mmol), dicyclohexyl
white solid. H NMR (500 MHz, CD3CN): d 1.2–2.3
(m, 32H), 3.26 (dd, 1H, J = 10.2, 4.9), 3.45–3.91 (m,
1H), 3.54 (dd, 1H, J = 9.4, 3.0), 3.61 (app. t, 1H,

2028
F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
10.5), 3.72 (d, 2H, J = 7.2), 3.81–3.90 (m, 2H), 3.91
(app. t, 1H, J = 9.3), 4.05 (app. t, 1H, J = 9.4), 5.05–
5.12 (m, 1H), 5.30 (d, 1H, J = 3.4), 5.41 (d, 1H,
J = 3.0), 7.49 (app. t, 2H, J = 8.0), 7.61 (app. td, 1H,
J = 7.4, 1.2), 8.04 (dd, 2H, J = 8.0, 1.2). 13C DEPT135
NMR (125 MHz, CD3CN): d 22.3 (CH2), 22.5 (CH2),
22.6 (CH2), 23.6 (CH2), 25.1 (CH2), 25.3 (CH2),
27.6 (CH2), 27.7 (CH2), 29.8 (CH2), 35.6 (CH2), 36.0
(CH2), 37.5 (CH2), 37.8 (CH2), 60.8 (CH2), 61.5
(CH2), 65.8, 66.1, 67.1, 69.2, 72.4, 73.5, 75.9, 92.2,
93.6, 128.7, 129.4, 133.4. HRFABMS (m/z): [M+Li]+
calcd for C37H50O11Li, 677.3513; found, 677.3527.
J = 7.5), 8.13 (d, 2H, J = 8.0); 13C DEPT135 NMR
(125 MHz, CDCl3): d 22.36 (CH2), 22.44 (CH2), 22.6
(CH2), 23.5 (CH2), 23.8 (CH2), 25.5 (CH2), 36.0
(CH2), 36.2 (CH2), 37.9 (CH2), 55.6, 60.9 (CH2), 61.6
(CH2), 63.8, 66.2, 72.5, 72.7, 73.2, 73.4, 76.0, 76.7,
94.5, 94.9, 96.9 (CH2), 128.7, 129.6, 133.4. LRFABMS
(m/z): [M+Li]+ calcd for C39H54O13Li, 737; found,
737. Anal. Calcd for C39H54O13: C, 64.09; H, 7.45.
Found: C, 64.30; H, 7.62.
4.3.23. 4,6:20,30:40,60-Tri-O-cyclohexylidene-3-O-meth-
oxymethyl-2-O-phenoxythionocarbonyl-a,a-D-trehalose
(30). Compound 29 (0.175 g, 0.240 mmol) was dis-
solved in 5 mL of dry MeOH, and 0.5 mL of a solution
of 0.5 M NaOMe in MeOH was added. Once the deben-
zoylation reaction was complete as shown by TLC, the
reaction mixture was concentrated in vacuo. The result-
ing residue was resuspended in EtOAc and washed with
NH4Cl solution and brine. Upon removal of the solvent,
the residue was redissolved in 20 mL of dry MeCN, and
DMAP (0.058 g, 0.48 mmol) and PTC-Cl (35 lL,
0.26 mmol) were added. The resulting bright-yellow
solution was stirred at rt overnight, after which TLC
analysis indicated complete consumption of the starting
material. The reaction mixture was concentrated and
purified by flash chromatography (2:8 EtOAc–hexanes)
to yield 0.137 g (75%) of 30 as a white solid (Rf 0.55,
4.3.21. 3-Deoxy-a,a-D-trehalose (9). Compound 28
(120 mg, 0.179 mmol) was dissolved in 5 mL of anhyd
MeOH, and 3.6 mL of a 0.5 M NaOMe solution in
MeOH was added. The reaction mixture was stirred at
rt for 5 h, after which TLC analysis indicated almost
complete conversion of the starting material to the
debenzoylated product. The reaction mixture was con-
centrated and redissolved in 20 mL of 1:1 10% aq HCl–
THF and stirred at rt for 48 h. The residual acid was neu-
tralized by the addition of Dowex 1 · 8 200–400 mesh re-
sin (Clꢂ form, exchanged to OHꢂ form). The resin was
removed by filtration, and the filtrate was concentrated
and purified by flash chromatography (7:2:1 EtOAc–
MeOH–H2O) to yield 35 mg (60%) of 9 as a white solid
following lyophilization. NMR data are summarized
in Tables 1–3. HRFABMS (m/z): [M+Na]+ calcd for
C12H22O10Na, 349.1111; found, 349.1109.
1
2:8 EtOAc–hexanes). H NMR (500 MHz, CD3CN): d
1.28–1.80 (m, 28H), 2.03–2.19 (m, 2H), 3.48 (s, 3H),
3.53 (dd, 1H, J = 9.5, 3.0), 3.55–3.90 (m, 6H), 3.95–
4.01 (m, 2H), 4.14 (app. t, 1H, J = 9.5), 4.20 (app. t,
1H, J = 9.5), 4.76 (d, 1H, J = 6.5), 4.96 (d, 1H, J =
6.5), 5.25 (dd, 1H, J = 9.5, 4.0), 5.39 (d, 1H, J = 3.0),
5.69 (d, 1H, J = 4.0), 7.22 (d, 2H, J = 7.5), 7.27 (t,
1H, J = 7.5), 7.40 (t, 2H, J = 7.5); 13C DEPT135
NMR (125 MHz, CDCl3): d 15.2 (CH2), 22.4 (CH2),
22.5 (CH2), 22.7 (CH2), 22.8 (CH2), 23.5 (CH2), 23.8
(CH2), 24.98 (CH2), 25.52 (CH2), 27.7 (CH2), 27.8
(CH2), 35.9 (CH2), 36.2 (CH2), 37.7 (CH2), 37.8
(CH2), 55.7, 61.4 (CH2), 61.5 (CH2), 63.8, 66.3, 72.9,
73.0, 73.2, 76.0, 80.7, 91.8, 93.6, 97.0 (CH2), 121.9,
126.6, 129.5. HRFABMS (m/z): [M+Li]+ calcd for
C39H54O13SLi: 769.3445; found, 769.3444.
4.3.22. 2-O-Benzoyl-4,6:20,30:40,60-tri-O-cyclohexylidene-
3-O-methoxymethyl-a,a-D-trehalose (29). Compound
27 (0.617 g, 0.898 mmol) and tetrabutylammonium
iodide (TBAI) (0.66 g, 1.8 mmol) were dissolved in 6 mL
of freshly distilled Hunig’s base (ethyldiisopropylamine).
¨
MOM-Cl (1.0 mL, 13 mmol) was added to the above
solution. The reaction mixture was heated in a 55 ꢁC
oil bath for 2 h, after which an orange sticky tar was
observed and TLC analysis indicated almost complete
conversion of the starting material to a less polar com-
pound (Rf 0.48, 2:8 EtOAc–hexanes). The reaction mix-
ture was concentrated under high vacuum, and the
resulting orange residue was redissolved in EtOAc.
The organic layer was washed with water and brine
and adsorbed onto silica gel. Purification by flash chro-
matography (1:9 to 2:8 EtOAc–hexanes) yielded 0.59 g
(90%) of 29 as a white solid. 1H NMR (500 MHz,
CD3CN): d 1.28–1.66 (m, 27H), 1.85–1.95 (m, 1H),
1.99–2.07 (m, 1H), 2.17–2.25 (m, 1H), 3.09–3.14 (dd,
1H, J = 10.5, 5.5), 3.26 (s, 3H), 3.49 (dd, 1H, J = 9.5,
3.0), 3.50–3.53 (m, 2H), 3.74 (app. t, 1H, J = 9.5), 3.76
(app. t, 1H, J = 9.5), 3.82–3.89 (m, 2H), 4.10–4.18 (m,
3H), 4.71 (d, 1H, J = 6.5), 4.94 (d, 1H, J = 6.6), 5.13
(dd, 1H, J = 9.5, 4.0), 5.34 (d, 1H, J = 3.0), 5.39 (d,
1H, J = 4.0), 7.48 (t, 2H, J = 8.0), 7.59 (t, 1H,
4.3.24. 2-Deoxy-a,a-D-trehalose (10). Compound 30
(0.23 g, 0.30 mmol) was dissolved in 20 mL of dry tolu-
ene. AIBN (10 mg, 0.006 mmol) and Bu3SnH (0.40 mL,
1.5 mmol) were added to this solution, and it was heated
at reflux for 30 min. At this point, a new product (Rf
0.40, 2:8 EtOAc–hexanes) was evident by TLC. The
reaction mixture was concentrated in vacuo to yield a
clear oil. This oil was redissolved in MeCN and washed
with pentane to remove organotin byproducts. The
MeCN layer was concentrated and dissolved in 1:1
10% aq HCl–THF. After stirring at rt overnight, depro-
tection appeared to be complete by TLC. The residual

F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
2029
14. Goren, M. B.; D’Arcy Hart, P.; Young, M. R.; Arm-
strong, J. A. Proc. Natl. Acad. Sci. U.S.A. 1976, 73, 2510–
2514.
15. Zhang, L.; English, D.; Andersen, B. R. J. Immunol. 1991,
146, 2730–2736.
acid was neutralized by addition of Dowex 1 · 8 200–
400 mesh resin (Clꢂ form, exchanged to OHꢂ form).
The resin was removed by filtration. The filtrate was
concentrated and purified by flash chromatography
(7:2:1 EtOAc–MeOH–H2O) to yield 40 mg (40%, two
steps) of 10 as a white solid following lyophilization
(Rf 0.32, 7:2:1 EtOAc–MeOH–H2O). NMR data are
summarized in Tables 1–3. HRFABMS (m/z):
[M+H]+ calcd calcd for C12H23O10, 327.1291; found,
327.1289.
16. Pabst, M. J.; Gross, J. M.; Brozna, J. P.; Goren, M. B. J.
Immunol. 1988, 140, 634–640.
17. Gilleron, M.; Stenger, S.; Mazorra, Z.; Wittke, F.;
Mariotti, S.; Bohmer, G.; Prandi, J.; Mori, L.; Puzo, G.;
De Libero, G. J. Exp. Med. 2004, 199, 649–659.
18. Mougous, J. D.; Petzold, C. J.; Senaratne, R. H.; Lee, D.
H.; Akey, D. L.; Lin, F. L.; Munchel, S. E.; Pratt, M. R.;
Riley, L. W.; Leary, J. A.; Berger, J. M.; Bertozzi, C. R.
Nat. Struct. Mol. Biol. 2004, 11, 721–729.
19. Hadfield, A. F.; Hough, L.; Richardson, A. C. Carbohydr.
Res. 1978, 63, 51–60.
20. Hadfield, A. F.; Hough, L.; Richardson, A. C. Carbohydr.
Res. 1980, 80, 123–130.
Acknowledgements
We thank Dr. Jeffrey Pelton for technical assistance with
NMR experiments performed at the 800 MHz spectro-
meter in Lawrence Berkeley National Laboratory; and
Drs. Clifton Leigh, Carlos Valdez and Jennifer Czlapin-
ski and Tanya Leavy for helpful discussions. This work
was supported by a grant to CRB from the National
Institutes of Health (AI51622).
21. Rose, J. D.; Maddry, J. A.; Comber, R. N.; Suling, W. J.;
Wilson, L. N.; Reynolds, R. C. Carbohydr. Res. 2002, 337,
105–120.
22. Azuma, I.; Sakurai, T.; Ishida, H.; Kitajima, T.; Yama-
zaki, M. Carbohydr. Res. 1991, 212, 47–53.
23. Wang, J.; Elchert, B.; Hui, Y.; Takemoto, J. Y.; Bensaci,
M.; Wennergren, J.; Chang, H.; Rai, R.; Chang, C.-W. T.
Bioorg. Med. Chem. 2004, 12, 6397–6413.
24. Hui, Y.; Chang, C.-W. T. Org. Lett. 2002, 4, 2245–2248.
25. Birch, G. G. J. Chem. Soc. C 1966, 1072–1074.
26. Birch, G.; Richardson, A. C. Carbohydr. Res. 1968, 8,
411–415.
27. Birch, G. G.; Lee, C.-K.; Richardson, A. C. Carbohydr.
Res. 1974, 36, 97–109.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.carres.
2007.05.009.
´
28. Hanessian, S.; Lavallee, P. Carbohydr. Res. 1973, 28, 303–
311.
29. Hough, L.; Munroe, P. A.; Richardson, A. C. J. Chem.
Soc. C 1971, 1090–1094.
30. Hough, L.; Richardson, A. C.; Tarelli, E. J. Chem. Soc. C
1971, 1732–1738.
31. Hough, L.; Richardson, A. C.; Tarelli, E. J. Chem. Soc. C
1971, 2122–2127.
References
32. Pratt, M. R.; Leigh, C. D.; Bertozzi, C. R. Org. Lett. 2003,
5, 3185–3188.
33. Robins, M. J.; Wilson, J. S. J. Am. Chem. Soc. 1981, 103,
932–933.
34. Robins, M. J.; Wilson, J. S.; Hansske, F. J. Am. Chem.
Soc. 1983, 105, 4059–4065.
1. Wolf, A.; Kra¨mer, R.; Morbach, S. Mol. Microbiol. 2003,
49, 1119–1134.
2. Singer, M. A.; Lindquist, S. TIBTECH 1998, 16, 460–468.
3. Benaroudj, N.; Lee, D. H.; Goldberg, A. L. J. Biol. Chem.
2001, 276, 24261–24267.
35. Hanessian, S.; Plessas, N. R. J. Org. Chem. 1969, 34,
1035–1044.
4. Elbein, A. D.; Pan, Y. T.; Pastuszak, I.; Carroll, D.
Glycobiology 2003, 13, 17R–27R.
36. Garegg, P. J.; Hultberg, H.; Wallin, S. Carbohydr. Res
1982, 108, 97–101.
37. Lichtenthaler, F. W.; Werner, B. Carbohydr. Res. 1999,
319, 47–54.
5. De Smet, K. A. L.; Weston, A.; Brown, I. N.; Young, D.
B.; Robertson, B. D. Microbiology 2000, 146, 199–208.
6. Elbein, A. D.; Mitchell, M. J. Bacteriol. 1973, 113, 863–
873.
38. Birch, G. G.; Lee, C. K.; Richardson, A. C.; Ali, Y.
Carbohydr. Res. 1976, 49, 153–161.
39. Holy, A.; Soucek, M. Tetrahedron Lett. 1971, 185–188.
40. Wang, C.-C.; Lee, J.-C.; Luo, S.-Y.; Fan, H.-F.; Pai,
C.-L.; Yang, W.-C.; Lu, L.-D.; Hung, S.-C. Angew.
Chem., Int. Ed. 2002, 41, 2360–2362.
41. Wallace, P. A.; Minnikin, D. E. Carbohydr. Res. 1994,
263, 43–59.
42. Wallace, P. A.; Minnikin, D. E. J. Chem. Soc., Chem.
Commun. 1993, 1292–1293.
43. Stork, G.; Takahashi, T. J. Am. Chem. Soc. 1977, 99,
1275–1276.
44. Van Halbeek, H. In Encyclopedia of Nuclear Magnetic
Resonance; Grant, D. M., Harris, R. K., Eds.; Wiley: New
York, 1996; pp 1107–1137.
7. Besra, G. S.; Chatterjee, D. Lipids and Carbohydrates of
Mycobacterium tuberculosis. In Tuberculosis: Pathogene-
sis, Protection and Control; Bloom, B. R., Ed.; American
Society for Microbiology: Washington, DC, 1994; pp 285–
306.
8. Dubos, R. J.; Middlebrook, G. Am. Rev. Tuberc. 1948, 58,
698–699.
9. Middlebrook, G.; Coleman, C. M.; Schaefer, W. B. Proc.
Natl. Acad. Sci. U.S.A. 1959, 45, 1801–1804.
10. Goren, M. B.; Brokl, O.; Roller, P.; Fales, H. M.; Das, B.
C. Biochemistry 1976, 15, 2728–2735.
11. Goren, M. B.; Brokl, O.; Das, B. C.; Lederer, E.
Biochemistry 1971, 10, 72–81.
12. Goren, M. B. Biochim. Biophys. Acta 1970, 210, 127–138.
13. Goren, M. B. Biochim. Biophys. Acta 1970, 210, 116–126.

2030
F. L. Lin et al. / Carbohydrate Research 342 (2007) 2014–2030
45. Duus, J. Ø.; Gotfredsen, C. H.; Bock, K. Chem. Rev. 2000,
100, 4589–4614.
52. Bodenhausen, G.; Ruben, D. J. Chem. Phys. Lett. 1980,
69, 185–189.
46. Piantini, U.; Sørensen, O. W.; Ernst, R. R. J. Am. Chem.
Soc. 1982, 104, 6800–6801.
53. Palmer, A. G., III; Cavanagh, J.; Wright, P. E.; Rance, M.
J. Magn. Reson. 1991, 93, 151–170.
47. Hurd, R. E. J. Magn. Reson. 1990, 87, 422–428.
48. Von Kienlin, M.; Moonen, C. T. W.; Van der Toorn, A.;
Van Zijl, P. C. M. J. Magn. Reson. 1991, 93, 423–429.
49. Braunschweiler, R.; Ernst, R. R. J. Magn. Reson. 1983, 53,
521–528.
54. Kay, L. E.; Keifer, P.; Saarinen, T. J. Am. Chem. Soc.
1992, 114, 10663–10665.
55. Bax, A.; Summers, M. F. J. Am. Chem. Soc. 1986, 108,
2093–2094.
56. Cicero, D. O.; Barbato, G.; Bazzo, R. J. Magn. Reson.
2001, 148, 209–213.
50. Bax, A.; Davis, D. G. J. Magn. Reson. 1985, 65, 355–360.
51. Marion, D.; Wuthrich, K. Biochem. Biophys. Res. Com-
57. Baer, H. H.; Bell, A. J. Carbohydr. Res. 1979, 75, 175–
184.
¨
mun. 1983, 113, 967–974.