Oligosaccharides as inhibitors of mycobacterial arabinosyltransferases. Di- and trisaccharides containing C-3 modified arabinofuranosyl residues

Article abstract of DOI:10.1016/j.bmc.2004.11.003

The assembly of the arabinan portions of cell wall polysaccharides in mycobacteria involves a family of arabinosyltransferases (AraT's) that promote the polymerization of decaprenolphosphoarabinose. Mycobacterial viability depends upon the ability of the organism to synthesize an intact arabinan and thus compounds that inhibit these AraT's are both useful biochemical tools as well as potential lead compounds for new anti-tuberculosis agents. We describe here the preparation of oligosaccharide fragments of mycobacterial arabinan that contain arabinofuranosyl residues modified at C-3 by the replacement of the hydroxyl group with an amino, azido or methoxy functionality. Subsequent testing of these oligosaccharides as inhibitors of mycobacterial AraT's revealed that all inhibited the enzymes, but to varying degrees. In further studies, each compound was shown to have only low activity as an inhibitor of mycobacterial growth.

Full text of DOI:10.1016/j.bmc.2004.11.003

Bioorganic & Medicinal Chemistry 13 (2005) 1369–1379
Oligosaccharides as inhibitors of mycobacterial
arabinosyltransferases. Di- and trisaccharides containing
C-3 modified arabinofuranosyl residues
Oana M. Cociorva,a, Sudagar S. Gurcha,b, Gurdyal S. Besrab and Todd L. Lowarya,*
aDepartment of Chemistry, The Ohio State University, 100 West 18th Avenue, Columbus, OH 43210, USA
bSchool of Biosciences, The University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
Received 6 October 2004; revised 29 October 2004; accepted 4 November 2004
Available online 23 November 2004
Abstract—The assembly of the arabinan portions of cell wall polysaccharides in mycobacteria involves a family of arabinosyltrans-
ferases (AraTÕs) that promote the polymerization of decaprenolphosphoarabinose. Mycobacterial viability depends upon the ability
of the organism to synthesize an intact arabinan and thus compounds that inhibit these AraTÕs are both useful biochemical tools as
well as potential lead compounds for new anti-tuberculosis agents. We describe here the preparation of oligosaccharide fragments of
mycobacterial arabinan that contain arabinofuranosyl residues modified at C-3 by the replacement of the hydroxyl group with an
amino, azido or methoxy functionality. Subsequent testing of these oligosaccharides as inhibitors of mycobacterial AraTÕs revealed
that all inhibited the enzymes, but to varying degrees. In further studies, each compound was shown to have only low activity as an
inhibitor of mycobacterial growth.
Ó 2004 Elsevier Ltd. All rights reserved.
1. Introduction
fying compounds that inhibit enzymes involved in cell
wall biosynthesis.9–11
The reemergence of mycobacterial diseases such as
tuberculosis and AIDS-associated M. avium as human
health threats has spurred increasing interest in the
development of new antibiotics for treating these infec-
tions.1,2 The urgency of this task has been heightened
by the emergence of drug resistant mycobacterial
strains3,4 and difficulties in treating these diseases in
AIDS-patients.5 Among the arsenal of front-line anti-
mycobacterial agents used for the treatment of these dis-
eases are drugs, for example, isoniazid and ethambutol,
that prevent the assembly of the mycobacterial cell wall,
a structure that is critical for the viability of these organ-
isms.6–8 Therefore, significant effort in the quest for new
anti-mycobacterial agents has been expended on identi-
In earlier papers, we described syntheses of arabinofu-
ranosyl-containing oligosaccharides that were designed
to be substrates12–14 or inhibitors15,16 of mycobacterial
arabinosyltransferases (AraTÕs), enzymes that are
responsible for assembling the arabinan portions of
two key cell-wall polysaccharides, lipoarabinomannan
(LAM) and arabinogalactan (AG). The arabinan of
AG and LAM consists of a linear chain of a-(1!5)-
linked D-arabinofuranose (Araf) residues, with periodic
a-(1!3)-linked branch points to which additional a-
(1!5)-linked arabinan chains are connected.7,8 At the
nonreducing ends of each arabinan chain in AG is at-
tached a hexasaccharide motif, 1, (Chart 1) while in
LAM, the terminal motifs are either 1, or tetrasaccha-
ride 2.7,8,17 Thus, three types of arabinofuranosyl link-
ages, a-(1!5), a-(1!3), and b-(1!2), are present in
these polysaccharides. The glycosyltransferases respon-
sible for the assembly of these glycans catalyze the con-
densation of decaprenolphosphoarabinose (DPA, 3)
with an acceptor species (e.g., 4) ultimately yielding
the polysaccharide (Fig. 1).18,19 Mycobacterial AraTÕs
have been validated as targets for drug action through
the demonstration that ethambutol, an anti-tuberculosis
Keywords: Arabinosyltransferases; Mycobacteria; Inhibitors; Oligo-
saccharides; Tuberculosis.
*
Corresponding author at present address: Alberta Ingenuity Centre
for Carbohydrate Science and Department of Chemistry, The
University of Alberta, Gunning-Lemieux Chemistry Centre, Edmon-
ton, Canada, AB T6G 2G2. Tel.: +1 7804921861; fax: +1
7804927705; e-mail: tlowary@ualberta.ca
Both authors contributed equally to this work.
0968-0896/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.11.003

1370
O. M. Cociorva et al. / Bioorg. Med. Chem. 13 (2005) 1369–1379
HO
HO
O
HO
HO
HO
O
O
O
HO
HO
O
HO
O
O
OH
HO
O
O
OH
HO
O
OH
O
HO
O
O
OH
O
O
Arabinan
O
HO
O
HO
HO
HO
O
HO
O
Arabinan
O
2
HO
1
HO
Chart 1.
O
P
HO
O
HO
O
O
O
8
HO
OH
O
HO
3, (Decaprenolphosphoarabinose, DPA)
Arabinan
AraT
OH
O
O
HO
+
OH
HO
OR
O
HO
OR
5
HO
4
Figure 1. Prototypical arabinosyltransferase-catalyzed reaction.
drug, inhibits at least one of these enzymes, likely the a-
(1!3) AraT.18,20–22 Thus, other compounds that pre-
vent arabinan assembly will not only be useful biochem-
ical tools, but also potential lead compounds for new
anti-tuberculosis agents.
inhibitors of mycobacterial AraTÕs. In addition, these
compounds have been screened for their ability to inhi-
bit mycobacterial growth. It is well-established24–30 that
oligosaccharide analogs in which one or more hydroxyl
groups are replaced by other functionalities frequently
inhibit the enzymes that normally bind their unmodified
parent structures. Thus, we prepared and evaluated
derivatives of 6–8 in which the hydroxyl group at
C-30 and/or C-300 has been substituted with methoxy,
azido, or amino groups (oligosaccharides 9–17,
Chart 2).
It has previously been demonstrated that di- and
triarabinofuranoside fragments of AG and LAM (e.g.,
6–8) are substrates for mycobacterial AraTÕs.18,23
We describe here the synthesis of analogs of these
oligosaccharides and their subsequent evaluation as
OH
HO
O
OH
OH
HO
HO
X
O
O
OH
O
X
O
O(CH2)7CH3
O
OH
X
O
O
X
O(CH2)7CH3
O
O
O(CH2)7CH3
HO
HO
OH
O
HO
OH
6, X = OH
7, X = OH
8, X = OH
9, X = OCH3
12, X = N3
10, X = OCH3
13, X = N3
11, X = OCH3
14, X = N3
15, X = NH2
16, X = NH2
17, X = NH2
Chart 2.

O. M. Cociorva et al. / Bioorg. Med. Chem. 13(2005) 1369–1379
1371
2. Results and discussion
2.1. Synthesis of 9–17
The 3-O-methylated thioglycoside 21 was obtained
(Scheme 1) in four steps from 23. First, reaction with so-
dium methoxide in methanol at reflux afforded diol 24 in
57% yield. The remainder of the mass balance in this
reaction was unreacted 23. Although allowing the reac-
tion to continue for a longer period of time resulted in
total conversion of the starting material, prolonging
the reaction also led to additional decomposition prod-
ucts. We therefore found it more efficient to stop the
reaction prior to the complete consumption of 23. Diol
24 could be readily separated from epoxide 23, and the
latter could then be re-subjected to the opening reaction.
With 24 in hand, it was then acetylated (giving 25) and
converted to triacetate 26 upon acetolysis with acetic
anhydride and sulfuric acid. This two-step sequence
gave 26 in 75% overall yield as an inseparable 20:1 a/b
mixture of anomers. Reaction of 26 with p-thiocresol
and boron-trifluoride etherate yielded 21 (80%) again
as an inseparable mixture of anomers in which the a-gly-
coside predominated (20:1 a/b). An analogous series of
transformations was used to convert 23 into the 3-azido
thioglycoside 22. Opening of the epoxide ring with so-
dium azide afforded 27 and acetylation of this diol gave
28 in 63% overall yield from 23. Acetolysis of the methyl
glycoside 28 provided peracetate 29, which, without fur-
ther purification, was reacted with p-thiocresol and bor-
on-trifluoride etherate yielding 22 in 83% yield, as a 5:1
a/b mixture of anomers that could be separated by
chromatography.
The preparation of 9–17 required first the synthesis of
the monosaccharide building blocks 18–22 (Chart 3).
Acceptors 18–20 are known and were prepared as de-
scribed earlier.15 The general strategy used to access 21
and 22 involved the introduction of the appropriate
functionality via the nucleophilic opening of methyl
2,3-anhydro-a-D-xylofuranoside (23, Scheme 1).31 The
addition of nucleophiles to 2,3-anhydrofuranosides pos-
sessing the a-D-xylo stereochemistry proceeds in a highly
regioselective manner, with attack at C-3 being pre-
ferred, presumably due to a combination of steric and
electronic factors.32 As outlined below, this approach
was successful and in all cases the regioselectivity of
the epoxide ring opening reactions could be established
1
by NMR spectroscopy. In the H NMR spectra of the
products, the anomeric hydrogen appeared as a singlet
or doublet with 3JH1,H2 < 2Hz; in the 13C NMR spectra,
the anomeric carbon resonated between 105 and
110ppm. These data clearly demonstrate that the
products formed have the D-arabino stereochemistry.33
Had the epoxide opened with the opposite regio-
selectivity, providing a product with the D-xylo stereo-
3
chemistry, the JH1,H2 value would be 4–5Hz and the
anomeric carbons would have appeared between 100
and 105ppm.
OBz
HO
OBz
OBz
HO
BzO
O
O
O
O(CH2)7CH3
BzO
O(CH2)7CH3
O(CH2)7CH3
HO
HO
18
19
20
9−17
OAc
OAc
AcO
O
AcO
O
STol
STol
CH3O
N3
22
21
Chart 3.
OR
OAc
RO
AcO
O
a
c
O
d
OAc
21
OCH3
CH3O
CH3O
HO
24, R = H
25, R = Ac
26
O
O
b
OCH3
OR
O
23
OAc
RO
AcO
c
O
d
e
OAc
22
OCH3
N3
N3
29
27, R = H
28, R = Ac
f
Scheme 1. Reagents and conditions: (a) NaOCH3, CH3OH, reflux, 57%. (b) Ac2O, pyridine, 0°C ! rt, 83%. (c) Ac2O, H2SO4, 0°C, 91% (for 25),
99% (for 28). (d) p-Thiocresol, BF3–OEt2, CH2Cl2, 0°C, 80% (for 26), 83% (for 29). (e) NaN3, NH4Cl, EtOH, reflux. (f) Ac2O, pyridine, rt, 63% (from
23).

1372
O. M. Cociorva et al. / Bioorg. Med. Chem. 13 (2005) 1369–1379
OAc
AcO
O
OBz
O
O
X
O(CH2)7CH3
BzO
a
b
c
c
9
30, X = OCH3
31, X = N3
d
18
15
12
OBz
BzO
O
X
O(CH2)7CH3
O
O
AcO
OAc
c
c
a
b
10
13
32, X = OCH3
33, X = N3
d
20
16
OAc
AcO
O
OBz
O
X
O
X
O(CH2)7CH3
O
O
AcO
OAc
c
c
a
b
11
14
34, X = OCH3
35, X = N3
d
19
17
Scheme 2. Reagents and conditions: (a) 21, N-iodosuccinimide, AgOSO2CF3, CH2Cl2, 0°C, 71% (for 18), 76% (for 20), 78% (for 19). (b) 22, N-
iodosuccinimide, AgOSO2CF3, CH2Cl2, 0°C, 82% (for 18), 90% (for 20), 78% (for 19). (c) NaOCH3, CH3OH, rt, 92% (for 30), 96% (for 31), 70% (for
32), 95% (for 33), 97% (for 34), 84% (for 35). (d) Ph3P, H2O, THF, rt, 32% (for 12), 73% (for 13), 80% (for 14).
Having successfully synthesized thioglycosides 21 and
22, these donors were then coupled with acceptors 18–
20 to provide the required oligosaccharides (Scheme
2). The glycosylations were all carried out using N-iodo-
succinimide and silver triflate as the promoters and the
reactions afforded the protected oligosaccharides 30–35
in 71–90% yield. The stereochemistry of the glycosyla-
tions was established by NMR spectroscopy. In all
cases, the 3JH1,H2 of the newly introduced arabinofuran-
osyl residue was 0–2Hz, indicative of the a-stereochem-
istry of this residue. Similarly, the chemical shifts of the
anomeric carbons were between 105 and 110ppm, which
is expected for the a-arabinofuranosides.33 Had the b-
by chromatography and the re-isolated starting material
was treated again with triphenylphosphine and water.
2.2. Screening of compounds as substrates and inhibitors
of mycobacterial arabinosyltransferases
With the oligosaccharides in hand, they were first evalu-
ated in the mycobacterial AraT assay to determine their
activity as substrates. The available assay uses a mem-
brane fraction from mycobacteria as the enzyme source.
Previous studies have demonstrated that this membrane
fraction possesses both a-(1!5) and b-(1!2) AraT
activities, but appears to lack the enzyme responsible
for the installation of the a-(1!3)-linked Araf residues
in mycobacterial arabinan.18 As expected,18,23 the parent
oligosaccharides 6–8,12 showed activity as substrates for
these enzymes. More interesting, however, was that
none of the modified analogs were substrates when
screened at a concentration of 3.6mM.
3
arabinofuranosides been produced, the JH1, H2 values
would have been larger (4–5Hz) and the chemical shifts
of the anomeric carbons at higher field (100–105ppm).33
The protected derivatives were then deacylated with
sodium methoxide to give 9–14 in 70–97% yields. The
aminosugar-containing oligosaccharides were obtained
upon reduction of 12–14 with triphenylphosphine in
wet THF to give the products 15, 16, and 17 in 32%,
73%, and 80% yields, respectively. The reduction of 12
was very slow and the reaction appeared to stop after
stirring for 24h. Therefore, 12 and 15 were separated
These oligosaccharides were therefore next tested as
inhibitors of the enzymes. For a given compound, the
corresponding parent oligosaccharide 6, 7, or 8 was
used as the substrate at 0.4mM, while each potential

O. M. Cociorva et al. / Bioorg. Med. Chem. 13(2005) 1369–1379
1373
Table 1. Activity of oligosaccharide analogs against mycobacterial
AraTÕsa and M. tuberculosis strain H37Rvb
which can be quantitated as a percent inhibition of
growth (Table 1).
Compound
% Inhibition of
AraT activity
% Inhibition of
mycobacterial growth
From the data presented in Table 1 it is clear that
although all of the oligosaccharides screened do inhibit
mycobacterial growth, they do so only to a minimal de-
gree. The most potent compound is the 30,300-dimethoxy
trisaccharide, 11, but only a 20% inhibition of growth is
observed.35 In addition, there is no apparent correlation
between the potency of a given compound as an inhibi-
tor of mycobacterial AraTÕs and its activity against bac-
terial growth. For example, the most potent AraT
inhibitor, disaccharide 13, does not inhibit bacterial
growth to any more appreciable level than does disac-
charide 12, which is a substantially weaker inhibitor of
the enzymes.
9
34
59
37
36
75
2
8
14
20
13
17
3
10
11
12
13
15
16
17
4
49
2
10
5
12
a All compounds were screened at a concentration 3.6mM as outlined
in Section 3.27.
b All compounds were screened at a concentration of 6.25lg/mL using
the Alamar Blue assay.34
The lack of activity of these compounds against the bac-
teria could be attributed to the relatively hydrophilic
character of the oligosaccharides, which would prevent
them from entering the organism by diffusion through
the highly lipophilic outer layer of the mycobacterial cell
wall. Similarly, these compounds would be expected to
be too large to enter the organism through cell wall
porins, which serve as the entry point for some anti-
tuberculosis drugs.36 The results here parallel those
obtained with other oligosaccharide fragments of myco-
bacterial cell wall polysaccharides. To date, in general
only low levels of activity have been observed and efforts
to increase the activity of these compounds by making
them more lipophilic through the addition of hydropho-
bic groups, for example, as in 36 and 37 (Chart 4), have,
for the most part, resulted in only marginal improve-
ments in activity.37–39 A notable exception is disaccha-
ride 38, which has been shown to have an MIC only
2–4 fold higher than ethambutol.40 However, it remains
to be demonstrated that the potency 38 is a consequence
its ability to inhibit mycobacterial AraTÕs. The detergent
nature of this compound, or the ones whose synthesis is
described here, could result in an inhibition of bacterial
growth or enzymatic activity.
inhibitor was added at a concentration of 3.6mM. All of
the modified oligosaccharides are inhibitors of the
AraTÕs, but to varying degrees (Table 1). The most
potent inhibitor was azido disaccharide 13 (75% inhibi-
tion), while the weakest was the amino disaccharide 15
(23% inhibition). Because all of the compounds tested
were AraT inhibitors, it is difficult to draw conclusions
about the origin of the inhibition. The analysis of the
data is further complicated by the presence of two AraT
activities, a-(1!5) and b-(1!2), in the incubation mix-
ture. We note, however, that the analogs of the a-(1!3)-
linked disaccharide 7, compounds 10, 13, and 16, were
the most potent inhibitors. Although the significance
of this observation is unclear, these results suggest that,
in future analog synthesis to identify inhibitors of these
enzymes, the focus should be on using this disaccharide
scaffold, rather than the a-(1!5)-linked disaccharide 6
or trisaccharide 8.
2.3. Screening of compounds as inhibitors of M. tuber-
culosis H37Rv growth
The oligosaccharides that were tested as inhibitors of
mycobacterial AraTÕs were next screened for their ability
to prevent the growth of M. tuberculosis strain H37Rv.
To assess inhibitory potency, each compound was evalu-
ated at a concentration of 6.25lg/mL in the fluores-
cence-based Alamar Blue microplate assay,34 which
makes use of a dye that changes from blue (and nonfluo-
rescent) to pink (and fluorescent) in the presence of the
growing bacteria. When a compound that inhibits bacte-
rial growth is added, a decrease in fluorescence results,
2.4. Conclusions
In summary, we describe here the synthesis of a panel of
oligosaccharide fragments of mycobacterial arabinan
and their subsequent testing as inhibitors of mycobacte-
rial AraTÕs and mycobacterial growth. All the com-
pounds do, to some degree, inhibit the AraTÕs, with
those based upon the a-(1!3)-linked disaccharide 7
OBz
O
OH
BzO
HO
O
O
OH
N
O
OCH3
O
OBn
BzO
O
HO
O
O
OH
HO
O
O(CH2)7CH3
O(CH2)7CH3
CH3O
BnO
O(CH2)7CH3
HO
36
37
38
Chart 4.

1374
O. M. Cociorva et al. / Bioorg. Med. Chem. 13 (2005) 1369–1379
being the most potent. Subsequent testing of these com-
pounds as inhibitors of mycobacterial growth revealed
only poor levels of activity.
up: Rf 0.57 (CH2Cl2/CH3OH, 6:1); [a]D +116.8 (c 1.4,
CH3OH); H NMR (400MHz, D2O, dH) 5.13 (s, 1H),
1
4.92(s, 1H), 4.15 (s, 2H), 4.03–3.96 (m, 3H), 3.80–3.76
(m, 2H), 3.73–3.62 (m, 3H), 3.59 (d, 1H, J = 5.8Hz),
3.45–3.39 (m, 1H), 3.38 (s, 3H), 1.55–1.54 (m, 2H),
1.27–1.25 (m, 10H), 0.88–0.82 (m, 3H); 13C NMR
(100MHz, D2O, dC) 107.9 (2), 87.3, 83.5, 82.9, 82.2,
80.6, 79.3, 68.0, 61.9, 61.0, 58.8, 32.2, 29.8, 29.8, 29.6,
26.3, 22.9, 14.2; HR-ESI-MS Calcd for [C19H36O9]Na+
431.2257, found 431.2248.
3. Experimental
3.1. General methods
Reactions were carried out in oven-dried glassware. Sol-
vents were distilled from appropriate drying agents be-
fore use. Unless stated otherwise, all reactions were
carried out at room temperature under a positive pres-
sure of argon and were monitored by TLC on silica
gel 60 F254. Spots were detected under UV light or by
charring with 10% H2SO4, in EtOH. Unless otherwise
indicated, all column chromatography was performed
on silica gel 60 (40–60lM). Iatrobeads refers to a
beaded silica gel 6RS–8060, which is manufactured by
Iatron Laboratories (Tokyo). The ratio between silica
gel and crude product ranged from 100 to 50:1(w/w).
Optical rotations were measured at 22 2°C and are
3.4. Octyl 3,5-di-O-(3-O-methyl-a-D-arabinofuranosyl)-
a-D-arabinofuranoside (11)
Trisaccharide 34 (90mg, 0.1mmol) was debenzoylated
in dry CH3OH (10mL), as described for the preparation
of 9. The product was purified by chromatography (hex-
anes/EtOAc, 1:4) to give 11 (54mg, 97%) as a syrup: Rf
0.30 (hexanes/EtOAc, 1:4); [a]D +114.4 (c 1.2, CH3OH);
1H NMR (400MHz, D2O, dH) 5.19 (s, 1H), 5.11 (s, 1H),
5.00 (s, 1H), 4.26 (s, 1H), 4.23–4.17 (m, 4H), 4.12–4.09
(m, 1H), 4.06–4.05 (m, 1H), 3.96–3.94 (m, 1H), 3.87–
3.84 (m, 2H), 3.78–3.68 (m, 6H), 3.52–3.51 (m, 1H),
3.47 (s, 3H), 3.46 (s, 3H), 1.63 (m, 2H), 1.34 (m, 10H),
0.93–0.91 (m, 3H); 13C NMR (100MHz, D2O, dC)
108.2, 108.0, 108.0, 87.4, 87.3, 83.6, 83.5, 82.7, 81.7,
80.4, 79.4, 78.8, 68.1, 66.1, 62.1, 61.9, 58.0, 57.9, 32.2,
29.7, 29.6, 26.3, 22.9, 21.0, 14.2; HR-ESI-MS Calcd
for [C25H46O13]Na+ 577.2836, found 577.2847.
1
in units of degreesmL/gdm. H NMR spectra were re-
corded at 250, 400, or 500MHz, and chemical shifts
are referenced to either TMS (0.0, CDCl3) or HOD
(4.78, D2O and CD3OD). 13C NMR spectra were re-
corded at 63, 100, or 125MHz, and 13C chemical shifts
were referenced to internal CDCl3 (77.23, CDCl3), exter-
nal dioxane (67.40, D2O) or CD3OD (48.9, CD3OD).
Electrospray mass spectra were recorded on samples
suspended in mixtures of THF with CH3OH and added
NaCl. Oligosaccharides 6–8 were prepared as previously
described.12
3.5. Octyl 5-O-(3-azido-3-deoxy-a-D-arabinofuranosyl)-
a-D-arabinofuranoside (12)
Disaccharide 31 (50mg, 0.08mmol) was debenzoylated
in dry CH3OH (10mL), as described for the preparation
of 9. The product was purified by chromatography
(CH2Cl2/CH3OH, 10:1) to give 12 (32mg, 96%) as a syr-
up: Rf 0.21 (CHCl3/CH3OH, 10:1); [a]D +98.5 (c 0.7,
CHCl3); 1H NMR (400MHz, CD3OD, dH) 5.06 (s,
1H), 4.99 (s, 1H), 4.16–4.15 (m, 2H), 4.10–4.08
(m, 1H), 4.01 (s, 1H), 3.99 (d, 1H, J = 1.9Hz), 3.95
(dd, 1H, J = 2.5, 11.1Hz), 3.92–3.88 (m, 2H), 3.74–
3.68 (m, 3H), 3.46–3.41 (m, 1H), 1.57–1.55 (m, 2H),
1.28–1.27 (m, 10H), 0.89–0.86 (m, 3H); 13C NMR
(100MHz, CD3OD, dC) 108.6, 108.0, 85.4, 83.1, 79.8,
79.6, 78.2, 68.1, 67.4, 66.5, 61.9, 32.0, 29.7, 29.5, 29.4,
26.3, 22.8, 14.3; HR-ESI-MS Calcd for [C18H33O8N3]-
Na+ 442.2165, found 442.2169.
3.2. Octyl 5-O-(3-O-methyl-a-D-arabinofuranosyl)-a-D-
arabinofuranoside (9)
To a solution of 30 (75mg, 0.12mmol) in dry CH3OH
(10mL), was added dropwise NaOCH3 (0.1M in
CH3OH) until the pH of the reaction mixture was 11.
After stirring overnight, the reaction mixture was
neutralized with few drops of acetic acid and then
concentrated. The resulting residue was purified by
chromatography (hexanes/EtOAc, 1:2) to give
9
(45mg, 92%) as a syrup: Rf 0.51 (hexanes/EtOAc, 1:2);
[a]D +102.7 (c 0.9, CH3OH); 1H NMR (400MHz,
D2O, dH) 5.12(s, 1H), 4.98 (s, 1H), 4.27 (s, 1H), 4.16–
4.12(m, H2 ), 4.09–4.08 (m, 1H), 4.04–4.01 (m, 1H),
3.93–3.87 (m, 2H), 3.80–3.70 (m, 4H), 3.53–3.49 (m,
1H), 3.48 (s, 3H), 1.65 (m, 2H), 1.36–1.34 (m, 10H),
0.93–0.91 (m, 3H); 13C NMR (100MHz, D2O, dC)
108.4, 107.9, 87.4, 83. 7, 82. 2, 81.8, 78.7, 77.3, 68.6,
66.7, 62.1, 58.0, 33.2, 29.8 (2), 29.7, 26.3, 22.9, 14.2;
HR-ESI-MS Calcd for [C19H36O9]Na+ 431.2257, found
431.2271.
3.6. Octyl 3-O-(3-azido-3-deoxy-a-D-arabinofuranosyl)-
a-D-arabinofuranoside (13)
Disaccharide 33 (110mg, 0.15mmol) was debenzoylated
in dry CH3OH (15mL), as described for the preparation
of 9. The product was purified by chromatography
(CH2Cl2/CH3OH, 10:1) to give 13 (60mg, 95%) as a syr-
up: Rf 0.54 (CH2Cl2/CH3OH, 6:1); [a]D +98.7 (c 0.8,
CHCl3); 1H NMR (400MHz, CD3OD, dH) 5.21 (s,
1H), 4.95 (s, 1H), 4.23 (s, 2H) 4.14–4.12 (m, 1H),
4.06–4.04 (m, 1H), 4.01–3.99 (m, 1H), 3.89 (d, 2H,
J = 12.2Hz), 3.77 (dd, 2H, J = 3.2, 6.1Hz), 3.69–3.65
(m, 2H), 3.45–3.41 (m, 1H), 1.61–1.57 (m, 2H), 1.29–
1.28 (m, 10H), 0.91–0.87 (m, 3H); 13C NMR
3.3. Octyl 3-O-(3-O-methyl-a-D-arabinofuranosyl)-a-D-
arabinofuranoside (10)
Disaccharide 32 (100mg, 0.14mmol) was debenzoylated
in dry CH3 OH (10mL), as described for the preparation
of 9. The product was purified by chromatography
(CH2Cl2/CH3OH, 6:1) to give 10 (40mg, 70%) as a syr-

O. M. Cociorva et al. / Bioorg. Med. Chem. 13(2005) 1369–1379
1375
(100MHz, CD3OD, dC) 107.80, 107.67, 82.36, 82.37,
82.14, 80.56, 80.40, 67.96, 65.97, 61.61, 60.87, 32.05,
29.69, 29.60, 29.48, 26.25, 22.86, 14.28; HR-ESI-MS
Calcd for [C18H33O8N3]Na+ 442.2165, found 442.2149.
ESI-MS Calcd for [C18H35O8N]Na+ 416.2260, found
416.2259.
3.10. Octyl 3,5-di-O-(3-amino-3-deoxy-a-D-arabinofuran-
osyl)-a-D-arabinofuranoside (17)
3.7. Octyl 3,5-di-O-(3-azido-3-deoxy-a-D-arabinofuranos-
yl)-a-D-arabinofuranoside (14)
A solution of 14 (28mg, 0.05mmol) and Ph3P (26mg,
0.1mmol) in THF (5mL) containing a few drops of
water was stirred for 24h. The solution was then concen-
trated and the product was purified by chromatography
on Iatrobeads (CHCl3/CH3OH, 2:1 with 2% Et3N) to
give 17 (21mg, 80%): Rf 0.11 (CHCl3/CH3OH, 4:1);
[a]D +109.7 (c 0.7, CH3OH); 1H NMR (400MHz,
D2O, dH) 5.10 (s, 1H), 5.04 (s, 1H), 4.99 (s, 1H), 4.19–
4.16 (m, 1H), 4.15–4.14 (m, 1H), 4.01–3.98 (m, 1H),
3.97–3.96 (m, 3H), 3.89–3.86 (m, 2H), 3.79–3.74 (m,
3H), 3.70–3.62(m, 3H), 3.55–3.49 (m, 1H), 3.05 (dd,
1H, J = 3.0, 5.9Hz), 3.01 (dd, 1H, J = 3.1, 6.1Hz),
1.56–1.52 (m, 2H), 1.26–1.17 (m, 10H), 0.82–0.79 (m,
3H); 13C NMR (100MHz, D2O, dC) 107.9, 107.8,
107.7, 85.7, 85.2, 83.0, 82.1, 81.8, 81.4, 79.4, 68.4, 66.6,
62.0, 62.0, 58.6, 58.6, 31.5, 29.0, 28.8, 28.6, 25.6, 22.4,
13.8; HR-ESI-MS Calcd for [C23H44O11N2]Na+
547.2843, found 547.2856.
Trisaccharide 35 (50mg, 0.06mmol) was debenzoylated
in dry CH3OH (5mL), as described for the preparation
of 9. The product was purified by chromatography on
Iatrobeads (CHCl3/CH3OH, 10:1) to give 14 (29mg,
84%) as a colorless oil: Rf 0.46 (CHCl3/CH3OH, 6:1);
[a]D +78.9 (c 0.8, CH3OH); 1H NMR (400MHz,
CD3OD, dH) 5.10 (s, 1H), 5.03 (s, 1H), 4.88 (s, 1H),
4.18–4.17 (m, 2H), 4.12–4.11 (m, 1H), 4.07–4.05 (m,
4H), 3.96–3.95 (m, 1H), 3.83–3.79 (m, 4H), 3.67–3.59
(m, 4H), 3.37–3.35 (m, 1H), 1.53–1.51 (m, 2H), 1.22–
1.20 (m, 10H), 0.83–0.79 (m, 3H); 13C NMR
(100MHz, CD3OD, dC) 109.5, 109.0, 108.9, 84.0, 83.7,
81.9, 81.6, 81.4, 78.7, 69.3, 67.8, 67.5 (2), 63.1 (2),
33.3, 31.2, 31.0, 30.9, 30.8, 27.5, 24.1, 15.6; HR-ESI-
MS Calcd for [C23H40O11N6]Na+ 599.2652, found
599.2653.
3.8. Octyl 5-O-(3-amino-3-deoxy-a-D-arabinofuranosyl)-
a-D-arabinofuranoside (15)
3.11. p-Toluyl 2,5-di-O-acetyl-3-O-methyl-1-thio-a-D-
arabinofuranoside (21)
A solution of 12 (41mg, 0.09mmol) and Ph3P (40mg,
0.15mmol) in THF (5mL) containing a few drops of
water was stirred for 24h. The solution was then concen-
trated and the product was purified by chromatography
on Iatrobeads (CHCl3/CH3OH, 2:1 with 2% Et3N) to
give 15 (11mg, 32%): Rf 0.14 (CHCl3/CH3OH, 4:1);
To a solution of 26 (1.1g, 3.8mmol) in dry CH2Cl2
(25mL) at 0°C was added p-thiocresol (564mg,
4.5mmol). After stirring for 10min at 0°C, BF3–OEt2
(0.72mL, 5.7mmol) was added dropwise and then the
reaction mixture was warmed slowly to room tempera-
ture. After 90min, Et3N (0.2mL) was added and the
reaction mixture was concentrated. The residue was
purified by chromatography (hexanes/EtOAc, 3:1) to
provide 21 (1.07g, 80%) as a syrup (a/b, 20:1): Rf 0.7
(hexanes/EtOAc, 2:1); 1H NMR (400MHz, CDCl3,
dH) 7.41 (d, 2H, J = 8.1Hz), 7.12(d, 2H, J = 8.1Hz),
5.47 (s, 1H), 5.26 (dd, 1H, J = 1.5, 1.5Hz), 4.48–4.44
(m, 1H), 4.36–4.24 (m, 2H), 3.67 (d, 1H, J = 5.1Hz),
3.46 (s, 3H), 2.33 (s, 3H), 2.11 (s, 6H); 13C NMR
(100MHz, CDCl3, dC) 170.9, 169.8, 138.0, 132.7,
130.4, 123.0, 91.8, 86.3, 81.3, 80.7, 63.6, 58.6, 21.3,
21.1, 21.0; HR-ESI-MS Calcd for [C17H22O6S]Na+
377.1035, found 377.1021.
1
[a]D +76.8 (c 0.9, CH3OH); H NMR (400MHz, D2O,
dH) 5.24 (s, 1H), 5.13 (s, 1H), 4.16–4.12 (m, 3H), 4.04–
4.03 (m, 2H), 3.87–3.82 (m, 2H), 3.77–3.74 (m, 3H),
3.59–3.54 (m, 1H), 3.22–3.20 (m, 1H), 1.64–1.62 (m,
2H), 1.31 (m, 10H), 0.90–0.88 (m, 3H); 13C NMR
(100MHz, D2O, dC) 107.8, 106.4, 87.9, 84.9, 83.0,
81.0, 75.3, 68.7, 61.8, 60.0, 58.3, 31.8, 29.3, 29.2, 29.1,
25.9, 22.6, 14.0; HR-ESI-MS Calcd for [C18H35O8N]-
Na+ 416.2260, found 416.2265.
3.9. Octyl 3-O-(3-amino-3-deoxy-a-D-arabinofuranosyl)-
a-D-arabinofuranoside (16)
A solution of 13 (30mg, 0.07mmol) and Ph3P (21mg,
0.08mmol) in THF (5mL) containing a few drops of
water was stirred for 24h. The solution was then concen-
trated and the product was purified by chromatography
on Iatrobeads (CHCl3/CH3OH, 2:1 with 2% Et3N) to
give 16 (20mg, 73%): Rf 0.48 (CHCl3/CH3OH, 2:1);
[a]D +79.6 (c 1.0, CH3OH); 1H NMR (400MHz,
CD3OD, dH) 5.07 (s, 1H), 4.84 (s, 1H), 4.04 (dd, 1H,
J = 1.3, 2.7Hz), 4.01–3.97 (m, 1H), 3.93 (dd, 1H,
J = 2.7, 5.9Hz), 3.88 (dd, 1H, J = 1.2, 2.9Hz), 3.86–
3.85 (m, 1H), 3.75 (dd, 1H, J = 3.4, 11.8Hz), 3.70–3.61
(m, 4H), 3.41–3.37 (m, 1H), 2.98 (dd, 1H, J = 2.9,
5.2Hz), 1.58–1.51 (m, 2H), 1.35–1.12 (m, 10H), 0.89–
0.86 (m, 3H); 13C NMR (100MHz, CD3OD, dC)
109.7, 109.6, 87.7, 84.5, 83.6, 83.3, 82.2, 69.9, 63.7,
63.1, 60.8, 33.2, 30.8, 30.6, 30.6, 27.4, 23.9, 14.6; HR-
3.12. p-Toluyl 2,5-di-O-acetyl-3-azido-3-deoxy-1-thio-D-
arabinofuranoside (22)
Glycosyl acetate 29 (350mg, 1.16mmol) was converted
into 22 with p-thiocresol (173mg, 1.39mmol) and
BF3–OEt2 (0.21mL, 1.7mmol) in dry CH2Cl2 (15mL)
as described for the preparation of 21. The product
was purified by chromatography (hexanes/EtOAc, 6:1)
to provide 22 (351mg, 83%) as two separable anomers
in a 5:1 a/b ratio. a-Anomer: Rf 0.66 (hexanes/EtOAc,
1
2:1); [a]D +120.0 (c 1.0, CHCl3); H NMR (400MHz,
CDCl3, dH) 7.38 (d, 2H, J = 8.1Hz), 7.12(d, H2 ,
J = 8.1Hz), 5.45 (d, 1H, J = 2.3Hz), 5.17 (dd, 1H,
J = 2.3, 2.7Hz), 4.32–4.27 (m, 3H), 3.87 (dd, 1H, J =
3.4, 6.9Hz), 2.32 (s, 3H), 2.11 (s, 3H), 2.10 (s, 3H);
13C NMR (100MHz, CDCl3, dC) 170.4, 169.7, 138.3,

1376
O. M. Cociorva et al. / Bioorg. Med. Chem. 13 (2005) 1369–1379
134.0, 133.8, 129.9, 129.1, 90.4, 82.3, 79.0, 66.8, 62.7,
21.2, 20.7 (2); HR-ESI-MS Calcd for [C16H19O5N3-
S]Na+ 388.0943, found 388.0940. b-Anomer: Rf 0.72
(hexanes/EtOAc, 2:1); [a]D ꢀ89.7 (c 0.7, CHCl3); 1H
NMR (400MHz, CDCl3, dH) 7.37 (d, 2H, J = 8.1Hz),
7.16 (d, 2H, J = 8.1Hz), 5.63 (d, 1H, J = 5.7Hz), 5.22
(dd, 1H, J = 5.7, 6.6Hz), 4.41 (dd, 1H, J = 5.9,
11.6Hz), 4.31 (dd, 1H, J = 5.2, 11.6Hz), 4.17 (dd, 1H,
J = 6.6, 6.7Hz), 4.01–3.96 (m, 1H), 2.33 (s, 3H), 2.20
(s, 3H), 2.13 (s, 3H); 13C NMR (100MHz, CDCl3, dC)
170.7, 170.1, 138.4, 133.1, 130.1, 129.4, 89.3, 78.9,
78.0, 65.6, 63.9, 21.3, 21.0, 20.9; HR-ESI-MS Calcd
for [C16H19O5N3S]Na+ 388.0943, found 388.0931.
(s, 3H), 2.12 (s, 3H), 2.11 (s, 3H), 2.10 (s, 3H); 13C
NMR (100MHz, CDCl3, dC) 170.8, 169.7, 169.7,
100.0, 85.9, 82.7, 80.4, 63.8, 58.7, 21.3, 21.0 (2); HR-
ESI-MS Calcd for [C12H18O8]Na+ 313.0899, found
313.0884.
3.16. Methyl 3-azido-3-deoxy-a-D-arabinofuranoside (27)
A solution of 23 (1.0g, 6.85mmol), NaN3 (884mg,
13.7mmol) and NH4Cl (800mg, 15.1mmol) in ethanol
(20mL) and water (8mL) was heated at reflux for 24h.
The reaction mixture was cooled and concentrated to
give crude 27 (1.3g) as a yellow-white solid that was
used crude in the next reaction: Rf 0.21 (toluene/EtOAc,
2:1). The NMR data for this compound were identical
to those reported previously.42
3.13. Methyl 3-O-methyl-a-D-arabinofuranoside (24)
To a solution of 23 (200mg, 1.37mmol) in dry CH3OH
(10mL), was added a 1M solution of NaOCH3 in
CH3OH (10mL, 10mmol). The reaction mixture was
heated at reflux for 24h before being cooled and neutral-
ized with acetic acid. The solution was concentrated and
purified by chromatography (hexanes/EtOAc, 1:2) to
give 24 (140mg, 57%) as a colorless oil: Rf 0.4 (hexa-
nes/EtOAc, 1:2). The NMR data for this compound
were identical to those reported previously.41
3.17. Methyl 2,5-di-O-acetyl-3-azido-3-deoxy-a-D-arabi-
nofuranoside (28)
To a solution of 27 (1.3g crude, 6.85mmol) in dry pyri-
dine (25mL) was added dropwise Ac2O (6.4mL,
68mmol). The reaction mixture was stirred at rt over-
night, then diluted with ether and washed successively
with 0.1M HCl, water, and brine. After drying
(Na2SO4), the solution was filtered and concentrated.
The residue was purified by chromatography (hexanes/
EtOAc, 6:1) to give 28 (1.17g, 63% from 23) as a syrup:
Rf 0.51 (hexanes/EtOAc, 2:1); [a]D +84.5 (c 0.9, CHCl3);
1H NMR (400MHz, CDCl3, dH) 4.99 (d, 1H,
J = 2.0Hz), 4.93 (s, 1H), 4.32–4.21 (m, 2H), 4.11–4.07
(m, 1H), 3.70 (dd, 1H, J = 2.0, 6.3Hz), 3.37 (s, 3H),
2.10 (s, 6H); 13C NMR (100MHz, CDCl3, dC) 170.7,
168.0, 106.6, 82.4, 79.5, 66.6, 63.3, 55.1, 20.9; HR-ESI-
MS Calcd for [C10H15O6N3]Na+ 296.0858, found
296.0845.
3.14. Methyl 2,5-di-O-acetyl-3-O-methyl-a-D-arabinofur-
anoside (25)
To a solution of 24 (900mg, 5.05mmol) in dry pyridine
(10mL) at 0°C was added dropwise Ac2O (4.7mL,
50mmol). The reaction mixture was stirred at rt over-
night, then diluted with CH2Cl2 and washed successively
with 0.1M HCl, water, and brine. After drying
(Na2SO4), the solution was filtered and concentrated
and the residue purified by chromatography (hexanes/
EtOAc, 2:1) to give 25 (1.1g, 83%) as a syrup: Rf 0.7
(hexanes/EtOAc, 1:2); [a]D +101.3 (c 0.8, CHCl3); 1H
NMR (400MHz, CDCl3, dH) 5.05 (d, 1H, J = 1.2Hz),
4.92 (s, 1H), 4.36–4.32 (m, 1H), 4.24–4.19 (m, 2H),
3.58 (m, 1H), 3.44 (s, 3H), 3.41 (m, 3H), 2.11 (s, 3H),
2.10 (s, 3H); 13C NMR (100MHz, CDCl3, dC) 169.9,
107.4, 85.4, 81.0, 80.6, 64.1, 58.7, 55.3, 21.1, 21.0; HR-
ESI-MS Calcd for [C11H18O7]Na+ 285.0950, found
285.0933.
3.18. 1,2,5-Tri-O-acetyl-3-azido-3-deoxy-D-arabinofura-
nose (29)
A solution of 28 (340mg, 1.24mmol) in Ac2O (10mL)
was cooled to 0°C for as a mixture of H2SO4/Ac2O
(0.3:2, 0.4mL) was added slowly. The reaction mixture
was stirred at 0°C for 1h before a saturated aqueous
solution of NaHCO3 was added until no more evolution
of gas was observed. The mixture was diluted with
CH2Cl2 (10mL) and the organic layer was separated
and washed successively with a saturated solution of
NaHCO3, water, and brine. After drying (Na2SO4),
the solution was filtered and concentrated to give 29
(370mg, 99%) as a syrup that was used crude in the
preparation of 22: Rf 0.60 (hexanes/EtOAc, 2:1).
3.15. 1,2,5-Tri-O-acetyl-3-O-methyl-a-D-arabinofuranose
(26)
A solution of 25 (1.09g, 4.2mmol) in Ac2O (20mL) was
cooled to 0°C and a H2SO4/Ac2O solution (0.3:2, 1mL)
was added. The reaction was stirred at 0°C for 60min
then cooled and a saturated aqueous solution of NaH-
CO3 (20mL) was added. After stirring for 10min,
CH2Cl2 (40mL) was added and the organic layer was
washed successively with a saturated aqueous solution
of NaHCO3, water, and brine. After drying (Na2SO4),
the organic phase was filtered and concentrated and
the residue was purified by chromatography (hexanes/
EtOAc, 2:1) to give 26 (1.1g, 91%) as a syrup (a/b,
20:1): Rf 0.33 (hexanes/EtOAc, 1:1); 1H NMR
(400MHz, CDCl3, dH) 6.17 (s, 1H), 5.17 (d, 1H,
J = 1.4Hz), 4.35–4.25 (m, 3H), 4.22–4.17 (m, 1H), 3.45
3.19. Octyl 5-O-(2,3-di-O-acetyl-3-O-methyl-a-D-arabino-
furanosyl)-2-O-benzoyl-a-D-arabinofuranoside (30)
A solution of alcohol 1815 (150mg, 0.42mmol), thiogly-
coside 21 (171mg, 0.48mmol) and powdered molecular
˚
sieves (4A, 250mg) in dry CH2Cl2 (15mL) was cooled
to 0°C and then N-iodosuccinimide (135mg, 0.6mmol)
and silver triflate (20mg, 0.08mmol) were added. After
stirring for 5min at 0°C, Et3N (0.1mL) was added
and the reaction mixture was then diluted with CH2Cl2

O. M. Cociorva et al. / Bioorg. Med. Chem. 13(2005) 1369–1379
1377
(10mL) and filtered through Celite. The filtrate was
washed successively with a saturated aqueous solution
of Na2S2O3, water, and brine. After drying (Na2SO4),
the organic phase was filtered and concentrated and
the residue was purified by chromatography (hexanes/
EtOAc, 6:1 ! 2:1) to give 30 (178mg, 71%) as a syrup:
Rf 0.35 (hexanes/EtOAc, 2:1); [a]D +89.0 (c 0.9, CHCl3);
1H NMR (400MHz, CDCl3, dH) 8.05–8.01 (m, 2H),
7.61–7.56 (m, 1H), 7.47–7.44 (m, 2H), 5.20 (s, 1H),
5.12(s, 1H), 5.11 (m, 1H), 5.08 (d, 1H, J = 0.9Hz),
4.32–4.16 (m, 4H), 3.95 (dd, 1H, J = 4.2, 11.1Hz),
3.77–3.73 (m, 2H), 3.54 (d, 1H, J = 4.8Hz), 3.50–3.46
(m, 1H), 3.30–3.28 (m, 4H), 2.09 (s, 3H), 2.08 (s, 3H),
1.63–1.56 (m, 2H), 1.27 (m, 10H), 0.89–0.86 (m, 3H);
13C NMR (100MHz, CDCl3, dC) 170.9, 169.7, 166.8,
133.7, 130.0, 129.5, 128.7, 106.3, 105.5, 86.3, 85.9,
82.3, 81.1, 80.4, 77.3, 68.1, 66.4, 64.0, 58.4, 32.0, 29.7,
29.5, 29.4, 26.3, 22.8, 21.0 (2), 14.3; HR-ESI-MS Calcd
for [C30H44O12]Na+ 619.2731, found 619.2715.
2H), 3.43 (s, 3H), 2.11 (s, 3H), 2.02 (s, 3H), 1.66–1.62
(m, 2H), 1.29 (m, 10H), 0.88–0.86 (m, 3H); 13C NMR
(100MHz, CDCl3, dC) 170.8, 169.6, 166.3, 165.6,
133.5, 133.1, 130.0, 129.8, 129.5, 128.6, 128.4, 105.9,
105.6, 86.4, 82.9, 81.0, 80.8, 80.7, 80.6, 67.7, 63.8, 63.3,
58.5, 32.0, 29.5, 29.5, 29.3 26.2, 22.8, 21.1, 20.8, 14.2;
HR-ESI-MS Calcd for [C37H48O13]Na+ 723.2992, found
723.3003.
3.22. Octyl 3-O-(2,5-di-O-acetyl-3-azido-3-deoxy-a-D-
arabinofuranosyl)-2,5-O-benzoyl-a-D-arabinofuranoside
(33)
Alcohol 2015 (120mg, 0.25mmol) was glycosylated with
thioglycoside 22 (112mg, 0.30mmol) in dry CH2Cl2
(15mL) with N-iodosuccinimide (67mg, 0.30mmol), sil-
ver triflate (12mg, 0.04mmol), and powdered molecular
˚
sieves (4A, 250mg) as described for the preparation of
30. Purification of the product by chromatography (hex-
anes/EtOAc, 4:1) gave 33 (160mg, 90%) as a syrup: Rf
0.61 (hexanes/EtOAc, 2:1); [a]D +123.7 (c 0.9, CHCl3);
1H NMR (400MHz, CDCl3, dH) 8.00 (d, 4 H,
J = 7.4Hz), 7.60–7.57 (m, 1H), 7.51–7.47 (m, 1H),
7.43–7.39 (m, 2H), 7.28–7.24 (m, 2H), 5.51 (s, 1H),
5.29 (s, 1H), 5.25 (s, 1H), 5.16 (d, 1H, J = 2.2Hz), 4.66
(dd, 1H, J = 3.0, 12.1Hz), 4.53 (dd, 1H, J = 4.3,
12.1Hz), 4.43–4.40 (m, 1H), 4.33 (d, 1H, J = 5.4Hz),
4.25–4.21 (m, 1H), 4.18–4.15 (m, 1H), 4.11–4.07 (m,
1H), 3.79–3.73 (m, 1H), 3.69 (dd, 1H, J = 2.6, 6.5Hz),
3.54–3.48 (m, 1H), 2.13 (s, 3H), 2.03 (s, 3H), 1.64–1.62
(m, 2H), 1.31–1.28 (m, 10H), 0.90–0.86 (m, 3H); 13C
NMR (100MHz, CDCl3, dC) 170.5, 169.8, 166.3,
165.6, 133.6 (2), 129.9, 129.8 (2), 129.4, 128.6, 128.4,
105.8, 104.9, 82.7, 82.5, 81.1, 80.6, 80.0, 67.7 (2), 63.2,
63.1, 32.0, 29.5, 29.4, 29.3, 26.2, 22.7, 20.8, 20.7, 14.2;
HR-ESI-MS Calcd for [C36H45O12N3]Na+ 734.2901,
found 734.2915.
3.20. Octyl 5-O-(2,3-di-O-acetyl-3-azido-3-deoxy-a-D-
arabinofuranosyl)-2-O-benzoyl-a-D-arabinofuranoside
(31)
Alcohol 1815 (35mg, 0.09mmol) was glycosylated with
thioglycoside 22 (52mg, 0.14mmol) in dry CH2Cl2
(10mL) with N-iodosuccinimide (60mg, 0.26mmol), sil-
ver triflate (5mg, 0.02mmol), and powdered molecular
˚
sieves (4A, 100mg) as described for the preparation of
30. Purification of the product by chromatography (hex-
anes/EtOAc, 3:1) gave 31 (45mg, 82%) as a syrup: Rf
0.64 (hexanes/EtOAc, 1:1); [a]D +99.7 (c 0.9, CHCl3);
1H NMR (400MHz, CDCl3, dH) 8.04–8.02(m, H2 ),
7.61–7.57 (m, 1H), 7.48–7.44 (m, 2H), 5.23 (s, 1H),
5.15 (s, 1H), 5.07 (d, 1H, J = 2.7Hz), 5.05 (d, 1H,
J = 2.2Hz), 4.32–4.12 (m, 4H), 3.91 (dd, 1H, J = 4.7,
11.2Hz), 3.77–3.74 (m, 2H), 3.68 (dd, 1H, J = 2.1,
6.4Hz), 3.49–3.45 (m, 1H), 3.32(d, 1H, J = 5.4Hz),
2.1 (s, 6H), 1.63 (m, 2H), 1.34–1.27 (m, 10H), 0.89–
0.85 (m, 3H); 13C NMR (100MHz, CDCl3, dC) 170.7,
169.9, 167.0, 133.8, 130.0, 129.3, 128.7, 105.7, 105.4,
86.1, 82.5, 82.3, 79.7, 77.0, 68.2, 66.7, 66.4, 63.2, 32.0,
29.6, 29.5, 29.4, 26.2, 22.8, 20.9, 14.3; HR-ESI-MS
Calcd for [C29H41O11N3]Na+ 630.2639, found 630.2652.
3.23. Octyl 3,5-di-O-(2,5-di-O-acetyl-3-O-methyl-a-D-
arabinofuranosyl)-2-O-benzoyl-a-D-arabinofuranoside
(34)
Alcohol 1915 (150mg, 0.41mmol) glycosylated with thio-
glycoside 21 (387mg, 1.09mmol) in dry CH2Cl2 (15mL)
with N-iodosuccinimide (236mg, 1.06mmol) silver tri-
flate (55mg, 0.21mmol), and powdered molecular sieves
3.21. Octyl 3-O-(2,5-di-O-acetyl-3-O-methyl-a-D-arabino-
furanosyl)-2,5-di-O-benzoyl-a-D-arabinofuranoside (32)
˚
(4A, 500mg) as described for the preparation of 30.
Purification of the product by chromatography (hexa-
nes/EtOAc, 6:1 ! 2:1) gave 34 (270mg, 78%) as a syrup:
Rf 0.27 (hexanes/EtOAc, 2:1); [a]D +104.2( c 1.2,
CHCl3); 1H NMR (400MHz, CDCl3, dH) 8.07–8.04
(m, 2H), 7.59–7.55 (m, 1 H), 7.46–7.43 (m, 2H), 5.34
(s, 1H), 5.27 (d, 1H, J = 1.4Hz), 5.16 (d, 1H,
J = 1.3Hz), 5.15 (s, 1H), 5.13 (s, 1H), 5.08 (d, 1H,
J = 0.9Hz), 4.33–4.15 (m, 8 H), 3.94 (dd, 1H, J = 3.9,
11.3Hz), 3.76–3.69 (m, 2H), 3.56 (d, 1H, J = 5.2Hz),
3.50–3.44 (m, 2H), 3.42 (s, 3H), 3.22 (s, 3H), 2.10 (s,
3H), 2.09 (s, 3H), 2.08 (s, 3H), 2.07 (s, 3H), 1.66–1.59
(m, 2H), 1.29–1.27 (m, 10H), 0.89–0.86 (m, 3H); 13C
NMR (100MHz, CDCl3, dC) 170.9 (2), 169.7, 169.6,
165.8, 133.5, 130.0, 129.8, 128.6, 106.1, 105.9, 105.6,
86.5, 86.5, 83.0, 81.4, 81.0, 81.0, 80.7, 80.5 (2), 67.7,
65.1, 64.0, 63.7, 58.5, 58.3, 32.0, 29.6, 29.4, 26.2, 22.8,
Alcohol 2015 (171mg, 0.36mmol) was glycosylated with
thioglycoside 21 (155mg, 0.43mmol) in dry CH2Cl2
(15mL) with N-iodosuccinimide (122mg, 0.54mmol),
silver triflate (19mg, 0.07mmol), and powdered molecu-
˚
lar sieves (4A, 250mg) as described for the preparation
of 30. Purification of the product by chromatography
(hexanes/EtOAc, 5:1 ! 2:1) gave 32 (190mg, 76%) as
a syrup: Rf 0.47 (hexanes/EtOAc, 2:1); [a]D +89.7 (c
1.0, CHCl3); 1H NMR (400MHz, CDCl3, dH) 8.01–
7.99 (m 4H), 7.58–7.56 (m, 1H), 7.49–7.47 (m, 1H),
7.43–7.39 (m, 2H), 7.28–7.24 (m, 2H), 5.41 (s, 1H),
5.31 (s, 1H), 5.22 (s, 1H), 5.19 (s, 1H), 4.68 (dd, 1H,
J = 2.8, 12.1Hz), 4.53 (dd, 1H, J = 4.4, 12.1Hz), 4.44–
4.41 (m, 1H), 4.38–4.36 (m, 1H), 4.27–4.20 (m, 1H),
4.18–4.13 (m, 2H), 3.77–3.73 (m, 1H), 3.56–3.48 (m,

1378
O. M. Cociorva et al. / Bioorg. Med. Chem. 13 (2005) 1369–1379
21.1 (2), 20.9, 14.3; HR-ESI-MS Calcd for
[C40H58O18]Na+ 849.3521, found 849.3506.
tor activity at this concentration) and [14C]-DPA (3,
40,000cpm, 18mM, 20lL [stored in chloroform/metha-
nol, 2:1]), were dried under a stream of argon in a micro-
centrifuge tube (1.5mL) and placed in a vacuum
desiccator for 15min to remove any residual solvent to-
gether with the appropriate acceptor compounds (6, 7,
or 8) to determine any potential inhibitory properties
of the modified compounds. The dried constituents of
the assay were then resuspended in 8mL of a 1% aque-
ous solution of Igepal CA-630 (Sigma). The remaining
constituents of the arabinosyltransferase assay contain-
ing 50mM MOPS (adjusted to pH8.0 with KOH),
5mM b-mercaptoethanol, 10mM MgCl2, 1mM ATP,
and membranes (1mg) were added to a final reaction
volume of 80lL. The reaction mixtures were then incu-
bated at 37°C for 1h. A CHCl3/CH3OH (1:1, 533lL)
solution was then added to the incubation tubes and
the entire contents centrifuged at 18,000g. The superna-
tant was recovered and dried under a stream of argon
and re-suspended in C2H5OH/H2O (1:1, 1mL) and
loaded onto a pre-equilibrated [C2H5OH/H2O (1:1)]
1mL Whatmann strong anion exchange (SAX) car-
tridge, which was washed with 3mL of ethanol. The elu-
ate was dried and the resulting products partitioned
between the two phases arising from a mixture of n-but-
anol (3mL) and H2O (3mL). The resulting organic
phase was recovered following centrifugation at 3500g
and the aqueous phase was again extracted twice with
3mL of water saturated n-butanol, the pooled extracts
were back-washed twice with water saturated with n-
butanol (3mL). The water saturated n-butanol fraction
was dried and re-suspended in 200lL of n-butanol.
The total cpm of radiolabeled material extractable into
the n-butanol phase was measured by scintillation
counting using 10% of the labeled material and 10mL
of EcoScintA (National Diagnostics, Atlanta, GA,
USA). The incorporation of [14C]Araf was determined
by subtracting counts present in control assays (incuba-
tion of the reaction components in the absence of the
compounds). The remainder of the labeled material
was subjected to thin-layer chromatography in CHCl3/
CH3OH/NH4OH/H2O (65:25:0.5:3.6) on aluminum
backed Silica Gel 60 F254 plates (E. Merck, Darmstadt,
Germany). Autoradiograms were obtained by exposing
TLCs to X-ray film (Kodak X-Omat) for 3 days to
determine the extent of product formation. Inhibition
of product formation using compounds 6, 7, and 8 at
0.4mM as acceptors was determined by exposure of
phosphorimaging screens (Kodak) for 24h. The de-
tected radiation was quantified using a Molecular Ima-
ger FX (BIO RAD) and Quantity One software.
Inhibition was then calculated by expressing the radio-
active product of each acceptor as a percentage of the
total radiation in all acceptor/inhibitor derived
products.
3.24. Octyl 3,5-di-O-(2,5-di-O-acetyl-3-azido-3-deoxy-a-
D-arabinofuranosyl)-2-O-benzoyl-a-D-arabinofuranoside
(35)
Alcohol 1915 (193mg, 0.53mmol) was glycosylated with
thioglycoside 22 (481mg, 1.32mmol) in dry CH2Cl2
(10mL) with N-iodosuccinimide (297mg, 1.32mmol) sil-
ver triflate (85mg, 0.33mmol), and powdered molecular
˚
sieves (4A, 500mg) as described for the preparation of
30. Purification of the product by chromatography (hex-
anes/EtOAc, 6:1 ! 2:1) to give 35 (350mg, 78%) as a
syrup: Rf 0.71 (hexanes/EtOAc, 1:1); [a]D +98.7 (c 0.8,
CHCl3); 1H NMR (400MHz, CDCl3, dH) 8.06–8.04
(m, 2H), 7.61–7.57 (m, 1H), 7.48–7.44 (m, 2H), 5.45 (s,
1H), 5.25 (d, 1H, J = 1.2Hz), 5.16 (d, 2H, J = 2.3Hz),
5.13 (d, 1H, J = 2.3Hz), 5.03 (d, 1H, J = 1.9Hz), 4.31–
4.20 (m, 6H), 4.13–4.07 (m, 2H), 3.91 (dd, 1H, J = 3.7,
11.1Hz), 3.73–3.68 (m, 4H), 3.49–3.43 (m, 1H), 2.13
(s, 3H), 2.10 (s, 3H), 2.09 (s, 3H), 2.08 (s, 3H), 1.65–
1.58 (m, 2H), 1.43–1.26 (m, 10H), 0.93–0.86 (m, 3H);
13C NMR (100MHz, CDCl3, dC) 170.6 (2), 169.9,
169.8, 165.8, 133.7, 130.0, 129.6, 128.7, 105.9, 105.6,
105.0, 82.8, 82.5 (2), 81.1, 80.8, 80.0 (2), 67.7, 66.9,
65.5, 63.2, 32.0, 29.6, 29.5, 29.4, 26.2, 22.8, 20.9 (3),
20.9, 14.3; HR-ESI-MS Calcd for [C38H52O16N6]Na+
871.3337, found 871.3348.
3.25. Bacterial strains and growth conditions
M. smegmatis mc2 155 was a generous gift from W. R.
Jacobs, Albert Einstein College of Medicine, Bronx,
New York. Liquid cultures of M. smegmatis were grown
at 37°C in Luria Bertoni (LB) broth medium (Difco)
supplemented with 0.05% Tween 80, biomass harvested,
washed with phosphate buffered saline (PBS), and
stored at ꢀ20°C until further use.
3.26. Preparation of membrane fractions
M. smegmatis cells (10g wet weight) were washed and
re-suspended at 4°C in 30mL of buffer A, which con-
tained 50mM MOPS (adjusted to pH8.0 with KOH),
5mM b-mercaptoethanol, and 10mM MgCl2, and sub-
jected to probe sonication (Soniprep 150, MSE Sanyo
Gallenkamp, Crawley, Sussex, UK; 1cm probe) for a to-
tal time of 10min in 60s pulses and 90s cooling intervals
between pulses. The sonicate was centrifuged at 27,000g
for 20min at 4°C. Membrane fractions were obtained by
centrifugation of the 27,000g supernatant at 100,000g
for 1h at 4°C. The supernatant was carefully removed
and the membranes gently resuspended in buffer A at
a protein concentration of 20mg/mL. Protein concentra-
tions were determined using the BCA Protein Assay
Reagent kit (Pierce Europe, Oud-Beijerland, The
Netherlands).
3.28. Measurement of anti-tuberculosis activity of the
oligosaccharides
3.27. Arabinosyltransferase assay
Measurement of the anti-tuberculosis activity of the tar-
get compounds was carried out as previously reported
using the fluorescence-based Alamar Blue microplate as-
say.34 All compounds were tested against M. tuberculosis
Compounds 9, 10, 11, 12, 13, 15, 16, and 17 at a concen-
tration of 3.6mM (previously shown to have no accep-

O. M. Cociorva et al. / Bioorg. Med. Chem. 13(2005) 1369–1379
1379
´
20. Mikusova, K.; Slayden, R. A.; Besra, G. S.; Brennan, P. J.
Antimicrob. Agents Chemother. 1995, 39, 2484–2489.
strain H37Rv (ATCC 27294) at a concentration 6.25lg/
mL. The results of these assays, expressed as a percent
inhibition of growth of the bacteria, are provided in
Table 1.
´
21. Deng, L.; Mikusova, K.; Robuck, K. G.; Scherman, M.;
Brennan, P. J.; McNeil, M. R. Antimicrob. Agents
Chemother. 1995, 39, 694–701.
22. Khoo, K.-H.; Douglas, E.; Azadi, P.; Inamine, J. M.;
Besra, G. S.; Brennan, P. J.; Chatterjee, D. J. Biol. Chem.
1996, 271, 28682–28690.
Acknowledgements
23. Ayers, J. D.; Lowary, T. L.; Morehouse, C. B.; Besra, G.
S. Bioorg. Med. Chem. Lett. 1998, 8, 437–442.
24. Spohr, U.; Hindsgaul, O.; Lemieux, R. U. Can. J. Chem.
1985, 63, 2644–2652.
25. Mukherjee, A.; Palcic, M. M.; Hindsgaul, O. Carbohydr.
Res. 2000, 326, 1–21.
26. Hindsgaul, O.; Kaur, K. J.; Srivastava, G.; Blaszczyk-
Thurin, M.; Crawley, S.; Heerze, L. D.; Palcic, M. M. J.
Biol. Chem. 1991, 266, 17858–17862.
27. Glaudemans, C. P. J.; Kovac, P.; Nashed, E. M. Methods
Enzymol. 1994, 247, 305–322.
The National Institutes of Health (AI44045–01) sup-
ported this work. GSB, a Lister-Jenner Research
Fellow, acknowledges support from The Medical
Research Council, The Wellcome Trust and the Lister
Institute for Preventive Medicine. Antimycobacterial
data were provided by the Tuberculosis Antimicrobial
Acquisition and Coordinating Facility (TAACF)
through a research and development contract with the
U.S. National Institute of Allergy and Infectious
Diseases.
28. Bundle, D. R.; Eichler, E.; Gidney, M. A. J.; Meldal, M.;
Ragauskas, A.; Sigurskjold, B. W.; Sinnott, B.; Watson,
D. C.; Yaguchi, M.; Young, N. M. Biochemistry 1994, 33,
5172–5182.
References and notes
29. Lowary, T. L.; Hindsgaul, O. Carbohydr. Res. 1994, 251,
33–67.
1. Kumari, S.; Ram, V. J. Drugs Today 2004, 40, 487–500.
2 . OÕBrien, R. J.; Nunn, P. P. Am. J. Respir. Crit. Care Med.
2001, 163, 1055–1058.
3. Wade, M. M.; Zhang, Y. Front. Biosci. 2004, 9, 975–
994.
´
30. Laferte, S.; Chan, N. W. C.; Sujino, K.; Lowary, T. L.;
Palcic, M. M. Eur. J. Biochem. 2000, 267, 4840–4849.
31. Martin, M. G.; Ganem, B.; Rasmussen, J. R. Carbohydr.
Res. 1983, 123, 332–334.
4. Nachega, J. B.; Chaisson, R. E. Clin. Infect. Dis. 2003, 36,
S24–S30.
32. Williams, N. R. Adv. Carbohydr. Chem. Biochem. 1970, 25,
109–179.
5. De Jong, B. C.; Israelski, D. M.; Corbett, E. L.; Small, P.
M. Annu. Rev. Med. 2004, 55, 283–301.
6. Blanchard, J. S. Annu. Rev. Biochem. 1996, 65, 215–239.
7. Brennan, P. J. Tuberculosis 2003, 83, 91–97.
8. Lowary, T. L. Mycobacterial Cell Wall Components. In
Glycoscience: Chemistry and Chemical Biology; Fraser-
Reid, B., Tatsuta, K., Thiem, J., Eds.; Springer: Berlin,
2001; pp 2005–2080.
9. Young, D. B.; Duncan, K. Annu. Rev. Microbiol. 1995, 49,
641–673.
10. Lowary, T. L. Mini-Rev. Med. Chem. 2003, 3, 689–702.
11. Crick, D. C.; Mahapatra, S.; Brennan, P. J. Glycobiology
2001, 11, 107R–118R.
12. Han, J.; Gadikota, R. R.; McCarren, P. R.; Lowary, T. L.
Carbohydr. Res. 2003, 338, 581–588.
13. Yin, H.; DÕSouza, F. W.; Lowary, T. L. J. Org. Chem.
2002, 67, 892–903.
14. DÕSouza, F. W.; Ayers, J. D.; McCarren, P. R.; Lowary,
T. L. J. Am. Chem. Soc. 2000, 122, 1251–1260.
15. Cociorva, O. M.; Lowary, T. L. Tetrahedron 2004, 60,
1481–1489.
16. Cociorva, O. M.; Lowary, T. L. Carbohydr. Res. 2004,
339, 853–865.
33. Mizutani, K.; Kasai, R.; Nakamura, M.; Tanaka, O.;
Matsuura, H. Carbohydr. Res. 1989, 185, 27–38.
34. Collins, L. A.; Franzblau, S. G. Antimicrob. Agents
Chemother. 1997, 41, 1004–1009.
35. As a point of reference, in this assay, the MIC of a
compound can be estimated to be the concentration at
which >90% inhibition of bacterial growth is observed
(Ref. 34). For ethambutol, an accurate MIC has been
determined to be 1.64lg/mL, which would translate into
>90% inhibition at this concentration. Therefore, the
oligosaccharides tested here are sufficiently poorer inhibi-
tors of bacterial growth than ethambutol.
36. Lambert, P. A. J. Appl. Microbiol. 2002, 92, 46S–54S.
37. Pathak, A. K.; Pathak, V.; Suling, W. J.; Gurcha, S. S.;
Morehouse, C. B.; Besra, G. S.; Maddry, J. A.; Reynolds,
R. C. Bioorg. Med. Chem. 2002, 10, 923–928.
38. Pathak, A. K.; Pathak, V.; Seitz, L.; Maddry, J. A.;
Gurcha, S. S.; Besra, G. S.; Suling, W. J.; Reynolds, R. C.
Bioorg. Med. Chem. 2001, 9, 3129–3143.
39. Pathak, A. K.; Pathak, V.; Maddry, J. A.; Suling, W. J.;
Gurcha, S. S.; Besra, G. S.; Reynolds, R. C. Bioorg. Med.
Chem. 2001, 9, 3145–3151.
40. Pathak, A. K.; Pathak, V.; Kulshrestha, M.; Kinnaird, D.;
Suling, W. J.; Gurcha, S. S.; Besra, G. S.; Reynolds, R. C.
Tetrahedron 2003, 59, 10239–10248.
17. Nigou, J.; Gilleron, M.; Puzo, G. Biochemie 2003, 85, 153–
166.
18. Lee, R. E.; Brennan, P. J.; Besra, G. S. Glycobiology 1997,
7, 1121–1128.
19. Lee, R. E.; Brennan, P. J.; Besra, G. S. J. Am. Chem. Soc.
1995, 117, 11829–11832.
41. Houseknecht, J. B.; McCarren, P. R.; Lowary, T. L.;
Hadad, C. M. J. Am. Chem. Soc. 2001, 123, 8811–8824.
42. Unger, F. M.; Christian, R.; Waldstaetten, P. Carbohydr.
Res. 1979, 69, 71–77.