Synthetic explorations towards 3-deoxy-3-fluoro derivatives of D-perosamine

Article abstract of DOI:10.1016/S0008-6215(01)00188-4

Based on a literature precedent, preparation of methyl 4-azido-3,4,6-trideoxy-3-fluoro-α-D-mannopyranoside (18) was attempted via fluorination of methyl 4-azido-2-O-benzyl-4,6-dideoxy-α-D-altropyranoside with diethylaminosulfur trifluoride (DAST). Contrary to expectations, the reaction took place with retention of configuration at the site of the fluorination yielding methyl 4-azido-2-O-benzyl-3,4,6-trideoxy-3-fluoro-α-D-altropyranoside. Treatment with DAST of methyl 4-azido-2-O-benzyl-4,6-dideoxy-α-D-allopyranoside (8), or its 2-(p-methoxybenzyl) analog 9 resulted in fluorination with inversion of configuration at position 3, to give the corresponding 3-deoxy-3-fluoro glucopyranosides 10 and 11, respectively. Accordingly, compound 18 was prepared from 11, by de-p-methoxybenzylation at O-2, followed by inversion of configuration at C-2 in the resulting methyl 4-azido-3,4,6-trideoxy-3-fluoro-α-D-glucopyranoside. The 2-O-methyl analog of 18 (19) was prepared by methylation of 18. Compounds 18 and 19 were converted, conventionally, into the 3-fluoro analogs of the terminal determinants of the O-PS of Vibrio cholerae O:1, serotype Inaba and Ogawa, respectively.

Full text of DOI:10.1016/S0008-6215(01)00188-4

www.elsevier.com/locate/carres
Carbohydrate Research 334 (2001) 195–205
Synthetic explorations towards 3-deoxy-3-fluoro
derivatives of
D-perosamine
Emmanuel Poirot,a Alex H.C. Chang,a,b,† Derek Horton,b Pavol Kova´cˇa,*
aNational Institutes of Health, NIDDK, LMC, Building 8, Room B1A24, Bethesda, MD 20892-0815, USA
bDepartment of Chemistry, American Uni6ersity, Washington, DC 20016, USA
Received 6 April 2001; accepted 29 June 2001
Dedicated to Professor Joachim Thiem on the occasion of his 60th birthday.
Abstract
Based on a literature precedent, preparation of methyl 4-azido-3,4,6-trideoxy-3-fluoro-a-
was attempted via fluorination of methyl 4-azido-2-O-benzyl-4,6-dideoxy-a- -altropyranoside with diethylaminosul-
fur trifluoride (DAST). Contrary to expectations, the reaction took place with retention of configuration at the site
of the fluorination yielding methyl 4-azido-2-O-benzyl-3,4,6-trideoxy-3-fluoro-a- -altropyranoside. Treatment with
DAST of methyl 4-azido-2-O-benzyl-4,6-dideoxy-a- -allopyranoside (8), or its 2-(p-methoxybenzyl) analog 9 re-
D
-mannopyranoside (18)
D
D
D
sulted in fluorination with inversion of configuration at position 3, to give the corresponding 3-deoxy-3-fluoro
glucopyranosides 10 and 11, respectively. Accordingly, compound 18 was prepared from 11, by de-p-methoxybenzy-
lation at O-2, followed by inversion of configuration at C-2 in the resulting methyl 4-azido-3,4,6-trideoxy-3-fluoro-a-
D
-glucopyranoside. The 2-O-methyl analog of 18 (19) was prepared by methylation of 18. Compounds 18 and 19
were converted, conventionally, into the 3-fluoro analogs of the terminal determinants of the O-PS of Vibrio cholerae
O:1, serotype Inaba and Ogawa, respectively. © 2001 Elsevier Science Ltd. All rights reserved.
Keywords: 3-Deoxy-3-fluoro-
Nucleophilic displacement; Vibrio cholerae O:1
D
-mannose; Methyl 4-azido-3-fluoro-3,4,6-trideoxy-a-D-mannopyranoside; Fluorination with DAST;
components of synthetic vaccines. We have
studied the interaction of carbohydrate anti-
gens and antibodies in many antigen–anti-
body systems.1–6 Using fragments of
polysaccharide antigens and analogs thereof
that have been specifically deoxygenated or
fluorinated, we have been able to obtain a
great deal of detailed information on binding
at the molecular level. As part of our current
work towards a synthetic vaccine for cholera,
we have synthesized a number of specifically
deoxygenated and fluorinated analogs of the
terminal determinants of the O-specific
polysaccharide (O-PS) of Vibrio cholerae O:1
(compare Ref. 7 and papers cited therein).
1. Introduction
During the past few years, we have been
interested in developing conjugate vaccines for
infectious diseases using synthetic fragments
of lipopolysaccharides as their antigenic com-
ponents. It is generally believed that anti-
body–antigen binding characteristics of such
fragments can reveal their potential utility as
* Corresponding author. Tel.: +1-301-4963569; fax: +1-
301-4020589.
E-mail address: kpn@helix.nih.gov (P. Kova´cˇ).
† Present address: Ferro Pfanstiehl Laboratories, Inc., 1219
Glen Rock Avenue, Waukegan, IL 60085 USA.
0008-6215/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(01)00188-4

196
E. Poirot et al. / Carbohydrate Research 334 (2001) 195–205
Herein we describe an approach towards sub-
stances related to 3-deoxy-3-fluoro-
perosamine (4-amino-4,6-dideoxy- -mannose)
and their conversion into the terminal deter-
minants of the O-PS of V. cholerae O:1 fluori-
nated at position 3.
sponding
D
-altro compound 13. Subsequent
D
-
SN2 process of a suitable leaving group at C-3
mediated with fluoride ion was expected to
give the desired methyl 4-azido-2-O-benzyl-
D
3,4,6-trideoxy-3-fluoro-a- -mannopyranoside.
D
The route starting from 4 seemed feasible, as
the theoretical analysis13 of stereochemical
factors suggested that the displacement at C-3
in a- and b-altropyranosides should occur
readily. More importantly, we surmised, the
claim of a successful, analogous displacement
with a fluoride ion, effecting the altro“
manno conversion reported by Sharma et al.14
rendered an important literature precedent.
2. Results and discussion
Syntheses of 2-, 2%- and 4%-deoxyfluoro
analogs of the terminal monosaccharide deter-
minants of the O-PS of Vibrio cholerae O:1,
serotype Inaba (1) and Ogawa (2) have been
described previously.8 The objective of this
work was to prepare the 3-fluoro analogs,
compounds 23 and 21, respectively. The syn-
thesis of the key precursor of these substances,
methyl 4-azido-4,6-dideoxy-a- -mannopyran-
D
oside (3), is well established.9–11 Therefore, we
focused initially on preparation of the 3-fluoro
derivatives 18 and 19 from the known12 2-O-
benzyl derivative of 3, azide 4. The synthesis
was to involve first inversion of the configura-
tion at position 3 in 4, to arrive at the corre-
Following the foregoing strategy (Scheme
1), the inversion of configuration at C-3 in the
known 2-benzyl ether (4) was attempted via
Scheme 1.

E. Poirot et al. / Carbohydrate Research 334 (2001) 195–205
197
Table 1
1H NMR chemical shifts a and peak multiplicities b in CDCl3 for compounds 6–11, 13, 14, and 16–27
Compound Chemical shifts (l)
H-1
H-2
4.11dd
H-3
H-4
3.58dd
H-5
H-6
OCH3-1 OCH3-2 OCH3-Ph
6 c
7
4.97d
4.92d
3.87m
3.86m
4.05m
4.04m
4.05m
3.40m
1.40d
1.40d
1.34d
1.31d
1.20d
1.05d
1.06d
1.38d
1.38d
1.34d
0.92d
1.36
1.35d
1.29d
1.25d
1.32d
1.23d
1.28d
1.36d
1.27d
1.23d
3.40s
3.39s
3.44s
3.42s
2.89s
2.92s
2.94s
3.40s
3.39s
3.41s
2.76s
3.33s
3.33s
3.32s
3.37s
3.35s
3.39s
3.44s
3.94s
3.28s
3.41s
4.08dd
3.42t
3.39t
3.04t
3.36dd
3.57dd
2.72dd
2.72dd
2.24dd
2.82ddd
2.84ddd
3.24dd
3.25ddd
3.24dt
3.81s
3.30s
8 d
4.72dd
4.68bd f,g
4.50bd l
4.46bt
4.47bt
4.72s
4.39m
4.36m
4.30m
4.94dt
4.98dt
4.13m
4.85dt
4.54dt
4.57dt
4.67ddd
4.71ddd 3.66–3.40m
4.48ddd 3.03q
4.76ddd 4.29–4.20m
4.57ddd
4.81ddd
5.78t
9 e
9 h,i
10 h,j
11 h,k
13 l
14 m
16
3.46–3.38m
3.62dd
3.46–3.38m
4.00m
4.66d
3.74ddd
3.77m n
4.55ddd
4.13ddd
3.66–3.40m
3.59m
4.15m
4.76bt
4.57bt
4.71dd
4.72dd
4.71bd
4.76dd
4.71bd
4.80dd
4.59d
3,57m
17 g
18
19
2.48ddd
3.14m
3.53m
3.62dt o
3.66–3.40m
3.45m
3.82–3.67m
3.53m
3.53s
3.48
3.53
20
21 p
22
3.82–3.67m
4.06m
3.07q
4.15q
2.96dd
3.39dd
4.18m
23 q
24 r
25 s
26 t
27 u
4.22ddd
3.46t
3.66dd
5.30ddd
ꢀ3.7m
3.90
4.31m
4.14m
3.72m
4.58s
4.71dd
4.73dd
5.21t
4.82ddd
4.78ddd
ꢀ4.11m
ꢀ3.7m
3.50s
a Nuclei of the
L-tetronic acid residue are referred to with primes.
b d, doublet; t, triplet; q, quartet; m, multiplet, b, broad.
2
c lPhCH 4.78, J 12.4 Hz.
2
2
d lPhCH 4.68, J 12.4 Hz.
2
e lPhCH , 4.61, J 12.0 Hz; lPhOCH 3.80; lOH 3.5, J J3,OH 8.2 Hz, disappears on deuteration.
2
2
3
f Partially overlapped with the PhCH2 signal.
g Broadening of the signal due to long-range interaction with H-3.
h Measured in C6D6.
i lPhCH , 4.31, J 11.5 Hz; lPhOCH 3.30; lOH 3.66, J 7.8 Hz, disappears on deuteration.
2
2
2
3
j lPhCH 4.52, J 12.2 Hz.
2
2
k lPhCH 4.54, J 11.9 Hz.
2
l lPhCH 2 4.61(s).
2
m lPhCH 4.64, J 11.9 Hz.
2
n On de2uteration, it appears as ddd, J1,2 4.0, J2,3 8.8, JF,2 13.1 Hz.
o Partially overlapped with the H-5 signal.
p Measured in ꢀ3:1 CDCl3–CD3OD; lH-2% 4.27 (dd, partially overlapped, J 3.9, 8.5), lH-3%a,b 2.05, 1.84 (2 m), lH-4%a,b 3.79 (dd,
partially overlapped, J 5.1, 6.3 Hz).
q Measured in D2O; lH-2% 4.30 (dd, J 4.0, 8.6), lH-3%a,b 2.03, 1.84 (2 m), lH-4%a,b 3.72 (dd, J 6.2, 7.9 Hz).
r lPhCH 4.64, J 12.7 Hz, lCOCH 2.19.
2
2
3
s lCOCH 2.10, lPhCH 4.68, J 12.0 Hz.
2
3
2
t lNH, 6.03, J4,NH 9.2 Hz, lH-2% 5,24 (dd, J 5.1, 7.3), lH-3%a,b 2.2 (m), lH-4%a,b 4.15(m), lCOCH 2.18, 2.17, 2.15.
3
u lNH, 6.08, J4,NH 9.7 Hz, lH-2% 5.2 (dd, J 4.9, 7.3), lH-3%a,b 2.1 (m), lH-4%a,b 4.15 (m), lCOCH 2.15, 2.03.
3
Swern15,16 oxidation (4“6), followed by re-
duction with NaBH4 of the ketone formed.
The conformational change, as a result of the
conversion 4“6, evidenced from NMR spec-
tra of 6 (Tables 1 and 2), prevented ready
identification of any configurational change
that may have occurred during the oxidation.
[To obtain a definitive proof of the stereo-
chemistry at C-2 in 6 (and 7) would be beyond
the scope of this work, and these structures
are shown (Scheme 1), arbitrarily, as having
the manno configuration at that position.]
However, the 1H NMR spectrum of the
product of the following reduction showed

198
E. Poirot et al. / Carbohydrate Research 334 (2001) 195–205
clearly that inversions at both C-3 and C-2
had occurred during the oxidation/reduction
steps, yielding the allo compound 8. The J1,2
coupling constant in the desired product hav-
ing the altro configuration would be expected
to remain essentially the same as in the start-
ing manno compound 4 (compare, J1,2 1.6
Hz9), and the large diaxial J3,4 in 4 (compare,
J3,4 9.6 Hz9) was expected to become small,
typical for an axial–equatorial system. The 1H
NMR spectrum of 8 (Tables 1 and 2) showed
the expected change of J3,4, but it also showed
J1,2 3.1 Hz, a value typical for a-glucopyran-
osides. Notwithstanding that the desired com-
pound 13 was not obtained in this way, we
proceeded with fluorination of the allo com-
pound 8 with diethylaminosulfur trifluoride
(DAST). NMR spectra of the major product
isolated showed that the fluorination took
place with inversion of configuration at C-3
giving the gluco compound 10. This further
confirmed the inversion of configuration at
C-2 that had occurred during the oxidation/
reduction steps (4“6“8, compare, Tables 1
and 2, showing the insignificant change of J1,2
value as a result of the conversion). A plausi-
ble explanation for such inversion lies in the
basic conditions during the foregoing transfor-
mations. Thus, we contemplated, it was im-
portant that the manno“altro conversion be
carried out under nonbasic conditions, so that
the undesired inversion of configuration at
C-2 would not take place. This was accom-
plished by nucleophilic displacement of a 3-O-
trifluoromethanesulfonyl (triflyloxy) group in
12 with nitrite ion, as described by Albert et
al.,17 which yielded directly the epi-hydroxy
compound 13. However, when the foregoing
Table 2
1H NMR coupling constants a,b for compounds 6–11, 13, 14, and 16–27
Compound
Coupling constants (Hz)
J1,2 J2,3
4.5
J3,4
J4,5
J5,6
J1,F
J2,F
J3,F
J4,F
6 c
7 d
8
9.9
9.8
10.2
10.4
10.1
9.9
6.2
6.1
6.1
6.2
6.4
6.0
6.0
6.1
6.3
6.0
6.2
6.1
6.1
6.7
6.2
6.2
6.1
6.3
6.3
6.3
6.1
4.2
3.1
3.3
2.5
9
3.5
3.5
ꢀ3.4
ꢀ3.6
1.2
1.2
ꢀ3.9
3.7
ꢀ3.5
ꢀ3.4
9.3
9.0
3.6
3.7
9.2
9.1
3.4
3.4
3.4
3.4
3.2
3.3
3.2
3.6
3.6
3.0
2.7
2.6
9.3
9.0
3.5
2.5
9.2
9.1
9.5
9 e
10
11
13
14
16
17
18
19
20
21
22
23
24
25
26
27
f
g
g
53.4
53.9
13.4
13.0
f
9.7
10.3
10.0
9.5
10.0
ꢀ10
g
7.7
48.4
52.8
55.1
48.9
48.8
48.9
48.6
48.9
48.0
30.1
12.2
19.1
g
ꢀ3.9
ꢀ3.7
5.0
18.0
1.8
6.4
11.5
g
g
1.6
9.9
9.9
4.8
f
f
g
g
g
10.2
10.4
9.9
10.4
3.2
3.2
10.7
10.6
9.7
9.7
g
g
1.8
4.9
f
f
10.0
10.2
10.2
10.0
10.2
g
1.8
4.2
1.5
1.7
2.0
6.6
9.6
g
g
g
4.8
ꢀ4.5
5.7
48.1
49.0
g
g
a For conditions of measurements, see Table 1.
b Nuclei of the
L-tetronic acid residue are referred to with primes.
c J2,4 1.2 Hz (long range).
d J2,4 1.2 Hz (long range).
e Measured in C6D6.
f Not determined due to broad signals.
g Not determined due to overlapping signals.

E. Poirot et al. / Carbohydrate Research 334 (2001) 195–205
199
Scheme 2.
compound was treated with DAST, the fluori-
nation occurred with retention of configura-
tion at C-3, yielding methyl 4-azido-2-O-ben-
While other routes to compounds 21 and 23
could be envisioned (Scheme 2), we concluded
from the observations just described that the
most straightforward route to the mannopy-
ranoside 18 would result from taking advan-
tage of the serendipitous inversion of
configuration at C-2 (4“6“8). The fact that
fluorination at C-3 in 8 positions the fluorine
atom at that carbon atom with the desired
stereochemistry (“10, Scheme 1) opens a rel-
atively simple way to the manno derivative 18.
It only requires inversion of configuration at
C-2 in the 3-fluoro gluco compound 10. This
idea was pursued with the p-methoxybenzyl
derivative 11 (Scheme 1), which was obtained
from the known21 2-ether 5 through a similar
sequence of transformations as described for
10 (5“7“9“11). The use of the p-methoxy-
benzyl group as a temporary protecting group
at O-2 was dictated by the presence of the
azido group at C-4, as it allowed high-yielding
regeneration of HO-2 under nonreductive
conditions.
zyl-3,4,6-trideoxy-3-fluoro-a- -altropyran-
D
oside (14). This was revealed unequivocally by
1H NMR spectroscopy (Tables 1 and 2). The
spectra did not show the expected, structurally
significant changes of either J2,3 or J3,4 in the
product of fluorination, as compared with
these values in the starting altropyranoside 13,
thereby providing evidence that the fluoro
compound 14 had been formed. A retention of
configuration during an analogous fluorina-
tion has been previously observed18. Consider-
ing the foregoing results, together with the
lack of evidence Sharma et al.14 provide for
the structure of the product of their fluorina-
tion (compare, Ref. 14, compound 16 therein),
it is not inconceivable that that fluorination
had also occurred with retention of configura-
tion. No products of fluorination could be
isolated from treatments of triflate 15, pre-
pared conventionally from the altro alcohol
13, with either tetrabutylammonium fluoride
(TBAF) or tris(dimethylamino)sulfonium di-
fluorotrimethylsilicate (TASF).19,20
Different results have been reported on the
efficiency of SN2 displacements at C-2 in a-
glucopyranosides. When treated with the ben-

200
E. Poirot et al. / Carbohydrate Research 334 (2001) 195–205
zoate anion, a 2-O-mesyl derivative gave only
a poor yield of the corresponding manno com-
pound.22 Using the better leaving, O-triflyloxy
group, Haradahira et al.23 reported a 30%
yield in the displacement with fluoride ion. On
the other hand, yields reported more recently
of the displacement of the 2-O-triflyloxy
group with fluoride and azide ions, were 60
and 90%, respectively.18 Treatment of the cor-
responding 2-hydroxy-a-glucopyranoside with
DAST led mainly to a rearrangement,18 an
observation that also others have made.24–26
Thus, it seemed interesting to attempt the
a-gluco“a-manno conversion (16“18) by
inversion of configuration at C-2 via a nucle-
ophilic SN2 displacement reaction in triflate
17. Accordingly, the p-methoxybenzyl group
in 11 was removed by treatment with 2,3-
dichloro - 5,6 - dicyano - 1,4 - benzoquinone
(DDQ), and the triflate 17, obtained from 16
conventionally, was treated17 with NaNO2 in a
polar aprotic solvent. That the product of this
reaction was the desired manno compound 18
the corresponding peracetates 27 and 26, re-
spectively. These were deacetylated to give the
respective desired products whose NMR spec-
tra were consistent with anticipated structures.
3. Experimental
General methods.—Unless stated otherwise,
optical rotations were measured at ambient
temperature for solutions in CHCl3 (c 1) with
a Perkin–Elmer automatic polarimeter, model
241 MC. All reactions were monitored by
thin-layer chromatography (TLC). The detec-
tion was effected by charring with 5% H2SO4
in EtOH and, when applicable, by UV light.
Preparative chromatography was performed
by gradient elution from columns of Silica Gel
60 (particle size 0.04–0.063 mm, E. Merck)
using, at the onset of development, a solvent
mixture slightly less polar than that used for
TLC. NMR spectra (25 °C) were obtained at
1
300 MHz for H and 75 MHz for 13C with a
Varian Mercury spectrometer. 1H NMR
chemical shifts reported in Table 1 are refer-
enced to that of Me4Si (0 ppm) and the HDO
solvent peak (4.78 ppm) for measurements in
organic solvents and D2O, respectively. 13C
NMR chemical shifts reported in Table 3 are
referenced to that of CDCl3 (77.0 ppm) and
the center MeOH peak (49.0 ppm). Assign-
ments of signals were made by first-order
analysis of spectra and, when feasible, they
were supported by APT and/or DEPT experi-
ments, homonuclear decoupling and/or homo-
and heteronuclear two-dimensional correla-
tion spectroscopy, using commercial software
supplied with the spectrometer. Chemical ion-
ization-mass spectra (CIMS) were measured
with a Finnigan spectrometer using ammonia
as the reagent gas. Palladium-on-charcoal cat-
alyst (5%, ESCAT 103) was a product of
Engelhard Industries. (S)-(−)-a-Hydroxy-g-
1
(60%) was evident from the H NMR spec-
trum showing J1,2 1.8 Hz (a change of J1,2
from 3.9 Hz as a result of the conversion
16“18), as expected for the newly formed
diequatorial arrangement of H-1 and H-2.
Also, the spectrum reflected a change of the
diaxial arrangement of H-2 and H-3 in 16 to
the axial–equatorial arrangement of the same
protons in 18, by showing J2,3 3.4 Hz (cf., J2,3
9.2 Hz in 16, Table 2). Conversion of 18 into
21 and 23 required now only well established
chemistry. Methylation of 18 gave the fully
protected compound 19, which was converted
into amine 20 by catalytic hydrogenation.
Amine 22 was similarly obtained. N-Acylation
of each of amines 20 and 22 with the commer-
cially available 3-deoxy- -glycero-tetronolac-
L
tone then gave the target compounds 21 and
23, which were most conveniently isolated as
butyrolactone (3-deoxy- -glycero-tetronolac-
L
tone) was purchased from Aldrich Chemical
Company, and used as supplied. HATU (O-7-
azobenzotriazol - 1 - yl) - 1,1,3,3 - tetramethyl-
uronium hexafluorophosphate) was purchased
from PerSeptive Biosystem, Inc. Unless stated
otherwise, solutions in organic solvents were
dried with anhyd Na2SO4, and concentrated at
40 °C/2 kPa.

E. Poirot et al. / Carbohydrate Research 334 (2001) 195–205
201
Methyl 4-azido-2-O-benzyl-4,6-dideoxy-h-
-arabino(ribo?)-hexopyranosid-3-ulose (6).—
for 30 min. Triethylamine (13 mL, 93 mmol)
was added at −60 °C, and the mixture was
allowed to warm to rt. TLC (4:1 hexane–
EtOAc) then showed disappearance of the
starting material and formation of a major,
faster moving product. The mixture was parti-
tioned between CH2Cl2 and water, the organic
layer was dried and concentrated. Chromatog-
raphy (10:1 hexane–EtOAc) gave pure 6 (3.78
g, 94.5%): mp 105–106 °C (from EtOAc); [h]D
−9.9° (c 0.7); CIMS: m/z 309 ([M+18]+).
Anal. Calcd for C14H17N3O4: C, 57.72; H,
D
A solution of oxalyl chloride (1.63 mL, 18.75
mmol) in dry CH2Cl2 (33 mL) was cooled to
−60 °C. A solution of Me2SO (2.66 mL, 37.5
mmol) in CH2Cl2 (8.5 mL) was added drop-
wise, while the inside temperature was kept
slightly below −60 °C. When the addition
was complete, the mixture was stirred at the
same temperature for 30 min. A solution of 4
(4 g, 18.75 mmol) in CH2Cl2 (20 mL) was
added at −60 °C, and the mixture was stirred
Table 3
13C NMR chemical shifts a,b for compounds 6–11, 14, and 16–27
Compound
Chemical shifts
C-1
C-2
C-3
C-4
C-5
C-6
PhCH2
OCH3-1
OCH3-2
PhOCH3
6
7
8
100.85
100.89
98.94
99.00
99.77
97.48
99.05
98.62
99.20
97.03
100.60
99.01
99.02
99.19
93.76
101.01
97.81
99.51
98.54
99.32
79.99
79.57
73.58
73.16
74.69
76.55
77.58
74.29
70.95
82.16
68.42
77.52
77.44
77.36
68.12
67.43
72.32
73.86
68.89
77.48
198.31
198.63
68.76
68.63
69.23
93.07
94.56
87.89
94.28
89.74
91.65
86.66
93.59
89.87
93.76
89.66
66.68
69.36
86.64
89.63
69.17
69.13
63.86
63.81
64.13
65.91
67.39
60.34
66.12
65.95
63.18
63.55
53.36
51.58
53.16
51.30
62.15
59.83
52.05
52.36
68.63
68.57
61.34
61.26
61.81
64.37
65.67
61.65
65.63
65.34
66.29
66.44
68.86
67.04
68.67
66.82
61.45
62.67
66.88
67.06
18.11
19.06
17.94
17.9
72.54
72.15
70.41
70.04
70.16
71.90
73.01
72.98
55.57
55.49
55.92
55.82
55.14
54.06
55.43
55.48
55.56
55.20
9
55.21
55.80
9 c
18.32
16.96
18.35
18.15
17.97
17.26
17.97
17.96
17.37
17.38
17.51
16.61
17.88
18.31
17.38
17.54
10 d
11 b,e
14
55.12
16 f
17 g
18 h
19 i
20 j
21 k
22 l
23 m
24 n
25 o
26 p
27 q
55.26
55.04
54.85
54.47
54.89
54.77
54.98
55.82
55.38
55.01
55.00
59.60
59.27
59.61
70.74
72.71
59.83
a For conditions of measurements, see Table 1.
b Nuclei of the
L-tetronic acid residue are referred to with primes.
c Measured in C6D6.
d J1,F 9.9, J2,F 15.5, J3,F 183.7, J4,F 15.5, J5,F 5.7 Hz.
e J1,F 11.8, J2,F 18.0, J3,F 183.5, J4,F 15.0, J5,F 6.2 Hz.
f J1,F 10.0, J2,F 17.3, J3,F 184.1, J4,F 16.2, J5,F 6.2 Hz.
g J1,F 7.8, J2,F 16.6, J3,F 189.7, J4,F 15.2, J5,F 6.6 Hz
h J1,F 7.8, J2,F 15.1, J3,F 182.3, J4,F 15.8, J5,F 6.8 Hz.
i J1,F 8.5, J2,F 14.5, J3,F 187.5, J4,F 15.5, J5,F 6.5 Hz.
j J1,F 9.7, J2,F 14.6, J3,F 182.2, J4,F 17.5, J5,F 7.8, J6,F 2.0, JF,OMe-2 2.0 Hz.
k J1,F 8.7, J2,F 14.6, J3,F 187.9, J4,F 17.5, J5,F 6.2, JF,OMe-2 3.1 Hz.
l J1,F 9.6, J2,F 17.4, J3,F 179.2, J4,F 17.1, J5,F 8.3, J6,F 2.0 Hz.
m J1,F 8.5, J2,F 16.6, J3,F 184.3, J4,F 17.7, J5,F 7.0 Hz.
n lCOCH 21.05 ppm.
3
o lCOCH 20.98 ppm.
3
p lC-2% 71.12, lC-3% 30.63, lC-4% 59.76 ppm; lCOCH 20.88, 20.75, 20.68 ppm.
3
q lC-2% 71.26, lC-3% 30.66, lC-4% 59.83 ppm; lCOCH (2 C) 20.79 ppm.
3

202
E. Poirot et al. / Carbohydrate Research 334 (2001) 195–205
Methyl 4-azido-2-O-benzyl-3,4,6-trideoxy-
3-fluoro-h- -glucopyranoside (10).—DAST
5.88; N 14.42. Found: C, 57.64; H, 5.97; N,
14.39.
D
(0.26 mL, 2.13 mmol) was added at −20 °C
to a solution of 8 (420 mg, 1.42 mmol) in
toluene (4 mL), and the mixture was stirred at
rt for 30 min. TLC (4:1 hexane–EtOAc) then
showed that all starting material was con-
sumed and that one major product was
formed. Methanol (1 mL) was added with
stirring at −20 °C, followed by aq NaHCO3
and, when effervescence ceased, the mixture
was partitioned between CH2Cl2 and water.
The organic phase was concentrated and chro-
matography (9:1“4:1 hexane–EtOAc) gave
10 (210 mg, 50%): [h]D +99.6° (c 0.5); CIMS:
m/z 313 ([M+18]+). Anal. Calcd for
C14H18FN3O3: C, 56.94; H, 6.14; N 14.23.
Found: C, 56.77; H, 6.05; N, 14.14.
Methyl 4-azido-4,6-dideoxy-2-O-p-meth-
oxybenzyl - h -
D
- arabino(ribo?) - hexopyran-
osid-3-ulose (7).—Treatment of 5 (21.1 g,
65.26 mmol) as described for the preparation
of 6 gave 7 (19 g 91%): mp 95–96 °C (from
EtOH): [h]D −31.5° (c 1.1); CIMS: m/z 339
([M+1]+). Anal. Calcd for C15H19N3O5: C,
56.07; H, 5.96; N, 13.08. Found: C, 55.85, H,
5.96, N, 13.06.
Methyl 4-azido-2-O-benzyl-4,6-dideoxy-h-
-allopyranoside (8).—Sodium borohydride
D
(277 mg, 7.32 mmol) was added at 0 °C to a
solution of 6 (3.05 g, 10.5 mmol) in 7:2
EtOH–H2O (90 mL), the cooling was re-
moved, and the mixture was stirred for 1 h.
TLC (10:1 toluene–acetone) then showed that
no starting material remained. Acetone (2 mL)
was added to decompose excess reagent, the
mixture was concentrated, and the residue was
partitioned between water and CH2Cl2. The
organic phase was dried, concentrated and the
residue was chromatographed to give first a
small amount of 3 resulting from the non-
stereospecific reduction.
Eluted later was pure, amorphous 8 (2.33 g,
75%): [h]D +1.2° (c 0.4; CIMS: m/z 311
([M+18]+). Anal. Calcd for C14H19N3O4: C,
57.33; H, 6.53; N 14.33. Found: C, 57.45; H,
6.45; N, 14.36.
A small amount of an intermediate, mixed
fraction was also obtained.
Methyl 4-azido-3-fluoro-2-O-p-methoxyben-
zyl-3,4,6-trideoxy-h- -glucopyranoside (11).
D
—Fluorination of alcohol 9 (9.8 g, 30.31
mmol) with DAST in toluene, as described for
the preparation of 10 gave, after chromatogra-
phy, the fluoro compound 11 (4.85 g, 49%):
mp 62–63.5 °C (from ether–hexane): [h]D
+
35° (c 1.1); CIMS: m/z 343 ([M+1]+). Anal.
Calcd for C15H20FN3O4: C, 55.38; H, 6.20; N,
12.92. Found: C, 55.23, H, 6.22, N, 12.90.
Methyl 4-azido-2-O-benzyl-4,6-dideoxy-h-
D
-altropyranoside (13).—Trifluoromethanesul-
fonic anhydride (0.2 mL, 1.2 mmol) was
added at −5 °C to a solution of alcohol 412
(293 mg, 1 mmol) in CH2Cl2 (5 mL) contain-
ing pyridine (0.25 mL, 2.5 mmol), and the
mixture was stirred at rt until TLC (4:1 hex-
ane–EtOAc) showed that the reaction was
complete (ꢀ1 h). The mixture was succes-
sively washed with 1 M HCl, aq NaHCO3 and
water, the organic phase was dried and con-
centrated at B35 °C, to give methyl 4-azido-
2-O-benzyl-3-O-trifluoromethylsulfonyl-4,6-
Compound 8 was further characterized
as methyl 3-O-acetyl-4-azido-2-O-benzyl-4,6-
dideoxy-a- -allopyranoside (24), which was
D
prepared by conventional acetylation: mp 61–
62 °C (from ether–hexane): [h]D −41.9° (c
0.7); CIMS: m/z 353 ([M+18]+). Anal. Calcd
for C16H21N3O5: C, 57.30; H, 6.31; N 12.53.
Found: C, 57.41; H, 6.26; N, 12.53.
Methyl 4-azido-4,6-dideoxy-2-O-p-meth-
dideoxy-a- -mannopyranoside (12, CIMS:
D
oxybenzyl-h-
D
-allopyranoside (9).—Reduction
m/z 425 ([M+18]+)), which was sufficiently
pure for the next step.
of ketone 7 (10.07 g, 31.34 mmol), as de-
scribed for the preparation of 8 gave, after
chromatography (7:3 hexane–EtOAc), the
allo compound 9 (9.80 g, 97%) as an oil: [h]D
−4.2° (c 1.1); CIMS: m/z 341 ([M+1]+).
Anal. Calcd for C15H21N3O5: C, 55.72; H,
6.55; N, 13.00. Found: C, 55.54, H, 6.59, N,
12.79.
Sodium nitrite (0.150 g, 0.17 mmol) was
added to a solution of the foregoing triflate in
DMF (3.5 mL), and the mixture was stirred at
50 °C for 2 h, when TLC (4:1 hexane–EtOAc)
showed that all starting material was con-
sumed. The mixture was concentrated, the
residue was triturated with acetone, the solids

E. Poirot et al. / Carbohydrate Research 334 (2001) 195–205
203
Methyl 4-azido-3-fluoro-3,4,6-trideoxy-h- -
D
were removed by filtration, the filtrate was
concentrated, and the residue was chro-
matographed to give the amorphous methyl
glucopyranoside (16).—Water (ꢀ0.2 mL) fol-
lowed by DDQ (1.736 g, 7.6 mmol) was added
to a solution of the methoxybenzylated com-
pound 11 (1.77 g, 5.46 mmol) in CH2Cl2
(ꢀ50 mL). The solution, which became dark,
was stirred at rt overnight, when TLC (4:1
hexane–EtOAc) showed that the reaction was
complete. After dilution with CH2Cl2, the
mixture was washed with brine, dried, and
concentrated. Chromatography (9:1“6:1 hex-
ane–EtOAc) gave 16 in essentially theoretical
yield: mp 60–61 °C (from ether–pentane):
[h]D +24.5° (c 0.8). CIMS: m/z 223 ([M+1]+).
Anal. Calcd for C7H12FN3O3: C, 40.97; H,
5.89; N, 20.48. Found: C, 41.09, H, 5.76, N,
20.42.
4-azido-2-O-benzyl-4,6-dideoxy-a- -altropy-
D
ranoside (13, 150 mg, 51%): [h]D +70° (c 1.5);
CIMS: m/z 313 ([M+18]+). Anal. Calcd for
C14H19N3O4: C, 57.33; H, 6.53; N 14.33.
Found: C, 57.49; H, 6.63; N, 14.03.
For further characterization by NMR spec-
troscopy, a portion of 13 was acetylated with
Ac2O–pyridine, to give methyl 3-O-acetyl-4-
azido-2-O-benzyl-4,6-dideoxy-a- -altropyran-
D
oside (25): CIMS: m/z 335 ([M+18]+).
Attempted fluorinations of 13.—(a) DAST
(0.12 mL, 1.02 mmol) was added at −10 °C
to a solution of the altroside 13 (100 mg, 0.34
mmol) in dry toluene (1 mL), and the mixture
was stirred at 50 °C for 30 min, when TLC
(4:1 hexane–EtOAc) showed that all starting
material was consumed. The mixture was pro-
cessed as described above for similar fluorina-
tions and chromatographed to give methyl
4-azido-2-O-benzyl-3-fluoro-3,4,6-trideoxy-a-
Methyl 4-azido-3-fluoro-3,4,6-trideoxy-h- -
D
mannopyranoside (18).—Pyridine (1.1 mL,
13.6 mmol) followed by triflic anhydride (1.2
mL, 7.1 mmol) was added with stirring at 0 °C
to a solution of the fluoro compound 16 (1.12
g, 5.46 mmol) in CH2Cl2 (ꢀ50 mL). After 30
min, TLC (4:1 hexane–EtOAc) showed that
the reaction was complete. The solution was
diluted with CH2Cl2, washed successively with
aq HCl (M), NaHCO3 (saturated, twice), wa-
ter, dried, and concentrated. Chromatography
(95:5 hexane–EtOAc) gave triflate 17 (1.63 g,
88.5%): CIMS: m/z 355 ([M+1]+), which was
used immediately for the next step.
D
-altropyranoside (14, 55 mg, 55%), as shown
by NMR spectroscopy; CIMS: m/z 313 ([M+
18]+).
(b) Compound 13 (293 mg, 1 mmol) was
converted into the corresponding 3-tri-
fluoromethanesulfonate (15), as described for
the preparation of 12, and the product was
divided into two equal portions.
To a solution of 15 (ꢀ0.5 mmol) in CH2Cl2
(8 mL) was added, at −30 °C, TASF (0.826
g, 3 mmol), and the mixture was stirred at rt
for 20 min, when TLC (4:1 hexane–EtOAc)
showed that all starting material was con-
sumed. After concentration, the residue was
partitioned between water and CH2Cl2, and
the organic phase was dried and concentrated
to give material containing essentially one
charring component. NMR spectra of the ma-
terial showed that fluorination did not occur,
and the material was not investigated further.
(c) To a solution of 15 (ꢀ0.5 mmol) in dry
THF (5 mL) was added 1 M solution of
Bu4NF in THF (1 mL), and the mixture was
stirred overnight at rt, when TLC (4:1 hex-
ane–EtOAc) showed that the reaction was
almost complete. NMR spectroscopy indi-
cated that the same product as in (b), devoid
of fluorine, was formed.
Sodium nitrite (666 mg, 9.66 mmol) was
added to a solution of the foregoing com-
pound 17 (1.63 g, 4.83 mmol) in DMF (ꢀ20
mL), and the mixture was stirred at 85 °C for
4 h. Heating was discontinued, and after 12 h
TLC (4:1 hexane–EtOAc) showed that all
starting material had been consumed. The
mixture was concentrated, and the residue trit-
urated with CH2Cl2. After filtration and con-
centration of the filtrate through a Celite pad,
chromatography afforded the fluorinated a-
mannopyranoside 18, 590 mg (60%) as a col-
orless oil: [h]D+124° (c 1.3); CIMS: m/z 223
([M+1]+). Anal. Calcd for C7H12FN3O3: C,
40.97; H, 5.89; N, 20.48. Found: C, 40.94, H,
5.85, N, 20.39.
Methyl 2-O-acetyl-4-(2,4-di-O-acetyl-3-de-
oxy-
deoxy-h-
of azide 18 (263 mg, 1.28 mmol) and 5%
L
-glycero-tetronamido)-3-fluoro-3,4,6-tri-
D
-mannopyranoside (26).—A mixture

204
E. Poirot et al. / Carbohydrate Research 334 (2001) 195–205
([M+1]+). Anal. Calcd for C8H14FN3O3: C,
43.83; H, 6.44; N, 19.17. Found: C, 44.07, H,
6.44, N, 18.89.
Pd–C (160 mg) in EtOH (ꢀ4 mL) was stirred
under hydrogen for 3 h, when TLC (4:1 hex-
ane–EtOAc and neat EtOAc) showed com-
plete disappearance of the starting material
and formation of one product. After filtration
through a Celite pad, the filtrate was concen-
trated, and chromatography gave methyl 4-
Methyl 2-O-acetyl-4-(2,4-di-O-acetyl-3-de-
oxy -
L
- glycero - tetronamido) - 3 - fluoro - 2 - O-
methyl - 3,4,6 - trideoxy - h -
D
- mannopyranoside
(27).—Hydrogenation of the foregoing azide
19 (311 mg), as described for the preparation
of 22, gave methyl 4-amino-3-fluoro-2-O-
amino-3-fluoro-3,4,6-trideoxy-a- -mannopyran
D
oside (22, 213 mg, 93%) as a colorless oil:
CIMS: m/z 180 ([M+1]+), 197 ([M+18]+).
A mixture of the foregoing amine (200 mg,
methyl-3,4,6-trideoxy-a- -mannopyranoside
D
(20) (175 mg, 61%): CIMS: m/z 194 ([M+1]+),
211 ([M+18]+).
1.116 mmol) and 3-deoxy- -glycero-tetrono-
L
A mixture of the foregoing amine 20 (175
lactone [(S)-(−)-a-hydroxy-l-butyrolactone]
(228 mg, 2.23 mmol) was stirred at 95 °C for
24 h. The mixture was cooled to rt and chro-
matographed (twice, 1st column, 6:1 CH2Cl2–
methanol; 2nd column, 10:1 EtOAc–MeOH)
to give amide 23 (277 mg, 88%) as a yellow
oil. It was dissolved in pyridine (4 mL), and
the mixture was treated with Ac2O (2 mL)
overnight at rt. The mixture was concentrated,
and the residue was chromatographed (1:1
hexane–EtOAc) to give 26 (293 mg, 73%): mp
101–102 °C (from EtOAc); [h]D −9.3° (c 1.2):
CIMS: m/z 408 ([M+1]+), 425 ([M+18]+).
Anal. Calcd for C17H26FNO9: C, 50.12; H,
6.43; N, 3.44. Found: C, 49.89, H, 6.58, N,
3.34.
mg, 0.905 mmol), and 3-deoxy- -glycero-
L
tetronolactone [(S)-(−)-a-hydroxy-l-butyro-
lactone, 185 mg, (1.811 mmol)] was stirred at
95 °C for 24 h. After cooling to rt the crude
product was treated with a mixture of 3:1
pyridine–Ac2O (4 mL), and the product iso-
lated as described for the preparation of 26.
After conventional processing and chromatog-
raphy (2:1 hexane–EtOAc), the crystalline
acetate 27 (212 mg, 62%) showed mp 107–
109 °C (from EtOH–Et2O) and [h]D +13.0°
(c 0.7); CIMS: m/z 397 ([M+1]+). Anal.
Calcd for C16H26FNO8: C, 50.65; H, 6.91; N,
3.69. Found: 50.10, H, 7.16, N, 3.44. We
could not obtain closer analytical figures than
those just shown, which fit a tetartohydrate
(0.25 water) of 27.
Methyl 4-(3-deoxy-
L
-glycero-tetronamido)-
-mannopyranoside
3-fluoro-3,4,6-trideoxy-h-
D
Methyl 4-(3-deoxy-
L
-glycero-tetronamido)-
(23).—Deacetylation (Zemple´n), of acetate 26
gave compound 23 in virtually theoretical
yield: mp 126–127.5 °C (from EtOH–EtOAc);
[h]D +29° (c 0.8, MeOH); CIMS: m/z 282
([M+1]+). Anal. Calcd for C11H20FNO6: C,
46.97; H, 7.17; N, 4.98. Found: C, 46.87, H,
7.26, N, 4.92.
3 - fluoro - 2 - O - methyl - 3,4,6 - trideoxy - h -
D
-
mannopyranoside (21).—Deacetylation (Zem-
ple´n) of 27 (147 mg) gave pure 21 (96 mg,
84%): mp 173–175 °C (from acetone): [h]D
+27.3° (c 0.7); CIMS: m/z 296 ([M+1]+),
313
([M+18]+).
Anal.
Calcd
for
C12H22FNO6: C, 48.81; H, 7.51; N, 4.74.
Found: C, 48.85, H, 7.69, N, 4.69.
Methyl 4-azido-3-fluoro-2-O-methyl-3,4,6-
trideoxy-h- -mannopyranoside (19).—To a
D
solution of the foregoing mannoside 18 (323
mg, 1.57 mmol) in DMF (ꢀ3 mL) was added
NaH (82 mg, 2.04 mmol), followed by MeI
(0.13 mL, 2.04 mmol), and the mixture was
stirred at rt for 30 min. TLC (4:1 hexane–
EtOAc) then showed that the methylation was
complete. A few drops of MeOH were added
to decompose the excess of reagents, and the
mixture was concentrated. Chromatography
(9:1 hexane–EtOAc) of the material in the
residue gave 19 (311 mg, 90%) as a colorless
syrup: [h]D +115.0° (c 1.5); CIMS: m/z 237
References
1. Nashed, E. M.; Perdomo, G. R.; Padlan, E. A.; Kova´cˇ,
P.; Kabat, E. A.; Glaudemans, C. P. J. J. Biol. Chem.
1990, 265, 20699–20707.
2. Glaudemans, C. P. J.; Kova´cˇ, P.; Nashed, E. M.; Padlan,
E. A.; Arepalli, S. R. ACS Symp. Ser. 1994, 580, 157–
183.
3. Glaudemans, C. P. J.; Kova´cˇ, P.; Nashed, E. M. Methods
Enzymol. 1994, 247B, 305–322.
4. Glaudemans, C. P. J.; Kova´cˇ, P.; Rao, A. S. Carbohydr.
Res. 1989, 190, 267–277.

E. Poirot et al. / Carbohydrate Research 334 (2001) 195–205
205
5. Glaudemans, C. P. J.; Kova´cˇ, P.; Rasmussen, K. Bio-
chemistry 1984, 23, 6732–6736.
15. Mancuso, A. J.; Huang, S. L.; Swern, D. J. Org. Chem.
1978, 43, 2480–2482.
6. Pavliak, V.; Pozsgay, V.; Kova´cˇ, P.; Karpas, A.; Chu, C.;
Schneerson, R.; Robbins, J.; Glaudemans, C. P. J. J.
Biol. Chem. 1993, 268, 25797–25802.
7. Wang, J.; Zhang, J.; Miller, C. E.; Villeneuve, S.; Ogawa,
Y.; Lei, P.-s.; Lafaye, P.; Nato, F.; Karpas, A.; Bystricky´,
S.; Szu, S. C.; Robbins, J. B.; Kova´cˇ, P.; Fournier, J.-M.;
Glaudemans, C. P. J. J. Biol. Chem. 1998, 273, 2777–
2783.
8. Chang, A. H. C.; Horton, D.; Kova´cˇ, P. Tetrahedron:
Asymmetry 2000, 11, 595–606.
9. Eis, M. J.; Ganem, B. Carbohydr. Res. 1988, 176, 316–
323.
10. Bundle, D. R.; Gerken, M.; Peters, T. Carbohydr. Res.
1988, 174, 239–251.
11. Gotoh, M.; Kova´cˇ, P. J. Carbohydr. Chem. 1994, 13,
1193–1213.
12. Peters, T.; Bundle, D. R. Can. J. Chem. 1989, 67, 497–
502.
13. Hough, L.; Richardson, A. C. In Nucleophilic Displace-
ment Reactions, Part F; Coffey, S., Ed.; New York:
Elsevier, 1967; Vol. I, pp. 403–407.
16. Omura, K.; Swern, D. Tetrahedron 1978, 34, 1651–1660.
17. Albert, R.; Dax, K.; Link, R. W.; Stu¨tz, A. E. Carbohydr.
Res. 1983, 118, C5–C6.
18. Faghih, R.; Escribano, C. F.; Castillon, S.; Garcia, J.;
Lukacs, G.; Olesker, A.; Thang, T. T. J. Org. Chem.
1986, 51, 4558–4564.
19. Szarek, W. A.; Hay, G. W.; Doboszewski, B. J. Chem.
Soc., Chem. Commun. 1985, 663–664.
20. Doboszewski, B.; Hay, G. W.; Szarek, W. A. Can. J.
Chem. 1987, 65, 412–419.
21. Lei, P.-s.; Ogawa, Y.; Flippen-Anderson, J. L.; Kova´cˇ, P.
Carbohydr. Res. 1995, 275, 117–129.
22. Miljkovic, M.; Gligorijevic, M.; Glisin, D. J. Org. Chem.
1974, 39, 3223–3226.
23. Haradahira, T.; Maeda, M.; Omae, H.; Yano, Y.; Ko-
jima, M. Chem. Pharm. Bull. 1984, 32, 4758–4766.
24. Hasegawa, A.; Goto, M.; Kiso, M. J. Carbohydr. Chem.
1985, 4, 627–637.
25. Nicolaou, K. C.; Ludduwahetty, T.; Randall, J. L.; Chu-
cholowski, A. J. Am. Chem. Soc. 1986, 108, 2466–2467.
26. Kova´cˇ, P.; Yeh, H. J. C.; Jung, G. L.; Glaudemans, C. P.
J. J. Carbohydr. Chem. 1986, 5, 497–512.
14. Sharma, M.; Bernacki, R. J.; Hillman, M. J.; Korytnyk,
W. Carbohydr. Res. 1993, 240, 85–93.
.