Reinvestigation of the iodocyclization of 4,5,7-tri-O-benzyl-3-(N-benzylacetamido)-1,2,3-trideoxy-D-gluco-hept-1- enitol: Unexpected formation of a 1,3-imino-heptitol derivative

Article abstract of DOI:10.1016/S0008-6215(01)00314-7

The NIS-mediated iodocyclization of 4,5,7-tri-O-benzyl-3-(N-benzylacetamido)-1,2,3-trideoxy-D-gluco-hept-1-enitol gave unexpectedly a 1,3-imino-heptitol derivative, namely 2-O-acetyl-N-benzyl-4,5,7-tri-O-benzyl-1,3-dideoxy-1,3-imino-D-glycero-D- ido-heptitol. This compound is a new example of a rare class of azetidine imino alditol derivatives which have interesting properties such as glycosidase inhibitors. The physical and spectral data for this imino heptitol were essentially identical to those reported for 2,6-anhydro-4,5,7-tri-O-benzyl-3-(N-benzylacetamido)-3-deoxy-D-glycero-D- ido-heptitol, a derivative of C-(2-acetamido-2-deoxy-α-D-glucopyranosyl)methanol obtained from the same precursor [Lay, L.; Nicotra, F.; Panza, L.; Verani, A. Gazz. Chim. Ital. 1992, 122, 345-348]; these findings cast doubts on the structure reported for the latter product.

Full text of DOI:10.1016/S0008-6215(01)00314-7

Carbohydrate Research 337 (2002) 273–277
www.elsevier.com/locate/carres
Note
Reinvestigation of the iodocyclization of
4,5,7-tri-O-benzyl-3-(N-benzylacetamido)-1,2,3-trideoxy-
-gluco-hept-1-enitol: unexpected formation of a 1,3-imino-heptitol
derivative
D
Adewale Eniade, Olivier R. Martin*
Department of Chemistry, State Uni6ersity of New York, PO Box 6016, Binghamton, NY 13902-6016, USA
Received 18 September 2001; accepted 28 November 2001
Abstract
The NIS-mediated iodocyclization of 4,5,7-tri-O-benzyl-3-(N-benzylacetamido)-1,2,3-trideoxy-
pectedly a 1,3-imino-heptitol derivative, namely 2-O-acetyl-N-benzyl-4,5,7-tri-O-benzyl-1,3-dideoxy-1,3-imino-
D
-gluco-hept-1-enitol gave unex-
-glycero- -ido-
D
D
heptitol. This compound is a new example of a rare class of azetidine imino alditol derivatives which have interesting properties
such as glycosidase inhibitors. The physical and spectral data for this imino heptitol were essentially identical to those reported
for 2,6-anhydro-4,5,7-tri-O-benzyl-3-(N-benzylacetamido)-3-deoxy-
D-glycero-D-ido-heptitol, a derivative of C-(2-acetamido-2-de-
oxy-a- -glucopyranosyl)methanol obtained from the same precursor [Lay, L.; Nicotra, F.; Panza, L.; Verani, A. Gazz. Chim. Ital.
D
1992, 122, 345–348]; these findings cast doubts on the structure reported for the latter product. © 2002 Elsevier Science Ltd. All
rights reserved.
Keywords: Azetidines, polyhydroxylated; 1,3-Imino heptitols; C-Glycosides
In the course of our investigations on glycoside mim-
ics,1 we became interested in the synthesis of non-trans-
ferable analogs of UDP-GlcNAc for the study of the
mechanism of biosynthesis of chitin.2 In particular, we
designed as target compound the homo analog of
UDP-GlcNAc (A, Scheme 1), a potential inhibitor of
chitin synthase and of related glycosyl transferases.
Although C-glycosyl derivatives of GlcNAc are notori-
ously difficult to prepare,3 we considered that the C-
sponding sugar nucleotide analog. We now report that
the NIS-mediated iodocyclization of 4,5,7-tri-O-benzyl-
3-(N-benzylacetamido)-1,2,3-trideoxy-D-gluco-hept-1-
enitol (4) did not give B, but an unexpected 1,3-
dideoxy-1,3-imino heptitol derivative 5; this compound
has properties nearly identical to those reported for B,
which suggests that the product reported as B4 may not
have the correct structure. A synthesis of the homo-
analog of UDP-GlcNAc A has recently been achieved
by Thiem and Scha¨fer5 by way of a different route.
(2-acetamido-2-deoxy-a-D -glucopyranosyl)methanol
derivative
B reported in 1992 by Nicotra and
coworkers4 would constitute a convenient precursor of
the desired homoglycosyl phosphate and the corre-
* Corresponding author. Present address: ICOA, Universite´
d’Orle´ans, BP 6759, F-45067 Orle´ans, France; tel.: +33-23-
8494581; fax: +33-23-8417281.
E-mail address: olivier.martin@univ-orleans.fr (O.R. Mar-
tin).
Scheme 1.
0008-6215/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(01)00314-7

274
A. Eniade, O.R. Martin / Carbohydrate Research 337 (2002) 273–277
Scheme 2.
Heptenitol derivative 4 was prepared from tri-O-ben-
zyl-
-arabinofuranose (1) using Nicotra’s approach4
under slightly modified conditions (Scheme 2). In par-
ticular, we confirmed the high degree of stereoselectivity
of the addition of vinylmagnesium bromide to the
with both H-1A and H-1B as well as a weaker one with
D
H-3 (see Fig. 1). Finally, the cis-stereochemistry of the
azetidine ring system was determined on the basis of a
series of unambiguous NOE correlations (Fig. 1) and
3
by comparison of the JH,H coupling constants with
those reported for related cis-configured azetidines (1,3-
furanosylamine 2: the
D
-gluco epimer 3 could be ob-
dideoxy-1,3-imino-hexitol derivatives7).
tained exclusively by reversing the mode of addition
and using a smaller excess of reagent. The procedure
was completed by the selective N-acetylation of 3, thus
affording the interesting chain-extended derivative of
GlcNAc 4. The NMR spectra of 4 were complicated by
the presence of signals of the two amide conformers.
Compound 4 was then reacted with NIS (instead of
iodine4), with the goal of promoting iodocyclization,6
and the reaction mixture was treated, after disappear-
ance of the starting material, with aqueous NaHCO3
for several hours. This sequence of reactions afforded a
major product, isolated in 41% yield, whose physical
and spectral properties were essentially identical to
those reported for B4 (see for example Table 1 for a
comparison of 13C NMR chemical shifts). The detailed
study of the spectral parameters of the isolated product,
as well as its chemical behavior, led us however to
conclude that this product was the isomeric 2-O-acetyl-
Fig. 1. Long-range H,H-COSY correlations and NOE’s in 5.
Table 1
13C NMR spectral data for 5–7 (90 MHz) a
Position
5
Lit. data4
6 b
7 b
N-benzyl-4,5,7-tri-O-benzyl-1,3-dideoxy-1,3-imino-
glycero- -ido-heptitol (5).
D-
C-1
C-2
C-3
C-4
C-5
C-6
C-7
58.0 (−)
67.2 c
58.7
67.8
68.3
79.2
77.8
70.6
71.8
74.0
74.0
75.6
63.3
21.6
58.2 (−) 59.3 (−)
D
67.4
67.3
79.1
76.8
73.7
65.3
69.5
80.4
76.8
71.0
Several features indicated that 5 did not contain an
acetamido function any more, but an acetate ester: the
CꢀO absorption band appearing at 1735 cm−1 and the
ready cleavage of the acetyl group under Zemple´n
conditions to give compound 7. Furthermore, the
acetylation of 5, to give the diacetate 6, promoted a
67.6 c
78.6 d
77.3 d
70.0
71.3 (−)
68.6 (−) 71.0 (−)
73.1 (−) 73.4 (−)
73.8 (−) 73.6 (−)
75.4 (−) 74.1 (−)
62.7 (−) 62.6 (−)
20.9
21.2
170.2
170.4
OꢁCH2Ph 73.4 (−)
73.4 (−)
1
large downfield shift of the signal of H-6 (see H NMR
74.9 (−)
data in Table 2), thus providing strong evidence that
the 6-OH group of precursor 4 did not participate in
the iodocyclization reaction. The chemical shift of C-1
(l 58.0 ppm in 5) suggested a bond to nitrogen rather
than to oxygen. Chemical shift changes on (de)-
acetylation also indicated that the acetate group in 5
was at C-2. The four-membered ring constitution of the
new product was firmly established on the basis of
long-range H,H-COSY correlations: one of the N-ben-
zylic protons was found to exhibit strong correlations
NꢁCH2Ph 62.6 (−)
CH3CO
20.9
CH3CO
170.5
171.3
a Chemical shifts in ppm (lTMS 0). Solvent CDCl3.
b Assignments confirmed by HETCOR experiments.
c These assignments can be interchanged.
d These assignments can be interchanged.

A. Eniade, O.R. Martin / Carbohydrate Research 337 (2002) 273–277
275
Table 2
in marked contrast with that of hexose-derived hepteni-
tols which constitute convenient precursors of various
pyranoid C-glycosyl compounds under the same condi-
tions.8 The process reported herein provides an unusual
access to 1,3-imino-alditols, a rare class of azetidine-
containing imino sugars. The related, five-carbon 1,3-
1H NMR spectral data for 5–7 (360 MHz) a
b
Position
5
6 c,d
7 c
H-1A (pro-R)
(J1A,1B, J1A,2)
H-1B (pro-S)
3.04, dd
(9.9, 6.1)
3.26 d
(ꢀ1)
3.02 dd
(9.9, 6.1)
3.23 d
(1.4)
2.99 dd
(9.0, 6.0)
3.24 d
(B1)
ꢀ4.23 m
(5.7)
3.56 dd
(8.7)
ꢀ4.23 m
(3.5)
3.83 dd
(8.0)
3.98 q
(?)
3.71 ꢀd
(?)
3.71 ꢀd
3.45 d
(13.2)
4.13 d
4.57 d
(11.9)
4.65 d
4.57 AB
(10.8)
4.57 AB
4.75 AB
(11.1)
dideoxy-1,3-imino-L-xylitol was shown recently to be a
potent and very selective glucoamylase inhibitor.9 Fi-
nally, our study suggests that the product obtained by
Nicotra from 4 under similar conditions is actually 5
(J1B,2
)
H-2
4.79 td
(6.5)
4.62 td
(6.5)
(J2,3
H-3
(J3,4
H-4
(J4,5
H-5
(J5,6
H-6
)
)
)
)
3.75 dd
(9.4)
3.67 dd
(9.7)
and not
a
C-(2-acetamido-2-deoxy-a-D-glucopyran-
osyl)methanol derivative as reported.4
4.35 dd
(2.3)
4.20 dd
(2.5)
3.47 dd
(7.6)
3.64 dd
(5.5)
1. Experimental
4.01 m
(5.2, 3.3)
3.63 dd
(9.8)
5.22 td
(5.9, 2.4)
3.78 dd
(11.2)
General methods.—Solvents were evaporated under
reduced pressure and below 40 °C. Optical rotations
were measured with an automatic polarimeter for solu-
(J6,7A, J6,7B
H-7A
)
(J7A,7B
)
1
tions in a 0.1 dm cell at 2293 °C. H and 13C NMR
H-7B
3.70 dd
3.41 d
(13.3)
3.91 dd
3.37 d
(13.3)
NꢁCHA
(JA,B
spectra were recorded at 360 and 90 MHz, respectively,
on a Bruker AM-360 instrument using chloroform-d as
the solvent unless otherwise indicated and TMS as the
internal reference (lTMS 0.00 ppm). For 13C NMR
spectra in CDCl3, the signal of the solvent (l 77.00) was
used as the reference. Chemical shifts (in ppm) and
coupling constants (in Hz) were obtained from first-or-
der analysis of the spectra. A (−) sign is used to
indicate negative 13C NMR-signals in the DEPT mode
(qy 135°). Assignments were generally made on the
basis of 2D H,H-COSY and HETCOR correlation
spectra. IR spectra were recorded using a Perkin–
Elmer 1600 FT-IR instrument. Analytical TLC was
performed on glass plates precoated with Silica Gel 60
F-254 as the adsorbent (layer thickness: 0.25 mm). The
developed plates were air-dried, exposed to UV light
for inspection, sprayed with a solution of ammonium
phosphomolybdate, and heated to 120–140 °C. Flash
chromatography was performed using Silica Gel 60
(230–400 mesh). Mass spectral analyses were carried
out at Midwest Center for Mass Spectrometry, Univer-
sity of Nebraska, Lincoln, NE 68588. Elemental analy-
ses were performed by Atlantic Microlab, Inc.,
Norcross, GA 30091.
)
NꢁCHB
4.17 d
4.51 d
(11.7)
4.16 d
4.46 AB
(12.0)
OꢁCHA1
(JA,B
)
OꢁCHB1
OꢁCHA2
4.59 d
4.53 AB
ꢀ4.50 AB
4.51 d
(11.8)
(JA,B
)
OꢁCHB2
OꢁCHA3
4.53 AB
4.86 d
(11.1)
4.68 d
4.70 d
(10.7)
(JA,B
)
OꢁCHB3
5.00 d
5.00 d
4.75 AB
a Chemical shifts in ppm (lTMS 0) and coupling constants in
Hz. Solvent: CDCl3.
b l CH3CO: 2.00 (s).
c Assignments confirmed by H,H-COSY.
d l CH3CO: 2.04, 2.05 (2s).
What is the origin of this unexpected product? Al-
though no detailed mechanistic investigations have
been undertaken, the mechanism described in Scheme 3
provides a working hypothesis. The NIS-mediated
iodocyclization might involve, as the internal nucle-
ophile, the carbonyl oxygen of the neighboring amide
function, to give iodinated intermediate II. Upon hy-
drolytic treatment, the resulting orthoamide III could
evolve either by regeneration of the acetamido group
with formation of a 1,2-epoxide or by cleavage of the
CꢁN bond to form the ester function at O-2 and the
azetidine ring by an internal SN2 process. The factors
responsible for the unexpected evolution of III remain
to be elucidated.
N-Benzyl-2,3,5-tri-O-benzyl-
(2).—Prepared from 2,3,5-tri-O-benzyl-
D
-arabinofuranosylamine
-arabinofura-
D
nose (1) as described by Nicotra.4 The product was
recrystallized from EtOH (yield: 87%): mp 75–77 °C
(lit.4 72–74 °C). 13C NMR (CDCl3, mixture of
anomers, main signals): l 49.1, 49.9 (NꢁCH2Ph), 70.7,
71.0, 71.6, 71.7, 71.9, 72.1 (OCH2Ph), 73.3 (C-5), 79.6,
80.5, 82.6, 83.0, 83.5, 86.5 (C-2–4), 90.4 and 93.5 (C-1).
In conclusion, the NIS-mediated reaction of the un-
4,5,7-Tri-O-benzyl-3-benzylamino-1,2,3-trideoxy-D-
saturated heptenitol 4, a derivative of 2-acetamido-2-
gluco-hept-1-enitol (3).—The procedure4 was modified
as follows: to compound 2 (2.5 g, 4.9 mmol) in THF (5
deoxy-D-glucose, did not give the expected
six-membered C-glycosyl compound. This behavior is

276
A. Eniade, O.R. Martin / Carbohydrate Research 337 (2002) 273–277
Scheme 3.
mixture was stirred vigorously overnight. Ethyl acetate
(20 mL) was then added and the aqueous phase was
separated. The organic phase was washed successively
with aq Na2S2O3, water and brine, dried (MgSO4) and
concentrated. The residue was purified by flash chro-
matography (3:1 hexane–EtOAc) which afforded com-
pound 5 as a syrup (150 mg, 41%). [h]D −21.3° (c 1,
CHCl3) {reported4 [h]D for B: −20.8° (c 1, CHCl3)};
IR: wmax 3455 (OH), 1735 cm−1 (CꢀO); NMR: see
Tables 1 and 2. FAB-MS: m/z 596 (100%, [M+H]+).
2,6-Di-O-acetyl-N-benzyl-4,5,7-tri-O-benzyl-1,3-di-
mL) was added, at rt and under nitrogen, a 1 M
solution of vinyl magnesium bromide in THF (25 mL,
5 equiv). The mixture was stirred overnight. The solu-
tion was then carefully added to aq NH4Cl (50 mL).
THF was partially evaporated under reduced pressure.
Ethyl acetate (50 mL) was then added, the aqueous
phase was separated, washed with EtOAc (25 mL), and
the combined organic phases were dried and concen-
trated, thus affording essentially pure 3 (2.7 g, quant);
[h]D +12° (c 1, CHCl3) {lit.4 [h]D +12.3° (c 1,
CHCl3)}. The D-manno epimer could not be detected in
the reaction mixture.
deoxy-1,3-imino-D-glycero-D
-ido-heptitol† (6).—Com-
pound 5 was acetylated in essentially quantitative yield
using DMAP in Ac2O. An analytical sample was ob-
tained by flash chromatography (4:1 hexane–EtOAc):
[h]D −29.9° (c 1, CHCl3); IR: wmax 1736 cm−1 (CꢀO);
NMR: see Tables 1 and 2. FAB-MS: m/z 638 (100%,
[M+H]+). Anal. Calcd for C39H43NO7: C, 73.45; H,
6.80; N, 2.20. Found: C, 73.40; H, 7.11; N, 2.07.
4,5,7-Tri-O-benzyl-3-(N-benzylacetamido)-1,2,3-tri-
deoxy- -gluco-hept-1-enitol (4).—Prepared from 3 in
D
94% yield by the procedure described.4 [h]D +14.6° (c
1, CHCl3) {lit.4 [h]D +14° (c 1, CHCl3)}; IR: wmax
(KBr) 3405 (OH), 1625 cm−1 (CꢀO); 13C NMR
(CDCl3, 25°C; mixture of two amide rotamers): l 22.7,
22.9 (NCOCH3), 46.3, 54.3 (NCH2Ph), 61.9 (C-3), 70.1,
70.6 (C-6), 71.1, 71.3 (C-7), 73.4, 73.5, 73.6, 74.8, 74.9
(OCH2Ph), 78.3, 78.7, 79.7, (C-4,5), 119.8, 121.0 (C-2),
126.6–128.5 (Ar-CH), 134.0–134.9 (Ar-C), 171.4, 172.0
(NCOCH3).
N-Benzyl-4,5,7-tri-O-benzyl-1,3-dideoxy-1,3-imino-
D-
glycero-
D
-ido-heptitol† (7).—Compound 5 (260 mg,
0.44 mmol) was deacetylated in MeOH (20 mL) con-
taining a catalytic amount of MeONa. The solvent was
evaporated, the residue was taken into EtOAc (20 mL),
the solution was washed with water, then with brine,
dried (MgSO4) and concentrated, thus affording homo-
geneous 7 (240 mg, 99%): [h]D −22.9° (c 1, CHCl3);
IR: wmax 3433 cm−1 (OH); NMR: see Tables 1 and 2.
FAB-MS: m/z 554 (30%, [M+H]+). Anal. Calcd for
C35H39NO5: C, 75.92; H, 7.10; N, 2.53. Found: C,
75.33; H, 7.15; N, 2.49.
2-O-Acetyl-N-benzyl-4,5,7-tri-O-benzyl-1,3-dideoxy-
1,3-imino-D-glycero-D
-ido-heptitol† (5).—To a solution
of 4 (350 mg, 0.6 mmol) in anhyd THF (10 mL) was
added, under N2 at 0 °C, N-iodosuccinimide (290 mg,
1.3 mmol). The reaction mixture was stirred at rt until
complete disappearance of 4 (ꢀ2 h). Saturated aq
Na2S2O3 was then added dropwise until the brown
color had disappeared. The phases were then separated
and the organic layer was transferred into a reaction
flask. Saturated aq NaHCO3 (4 mL) was added and the
Acknowledgements
† Strictly, these compounds should be named as derivatives
of a 5,7-dideoxy-5,7-imino-D-glycero-L-gulo-heptitol.
This work was supported, in part, by a grant from
the National Institutes of Health (grant DK 35766).

A. Eniade, O.R. Martin / Carbohydrate Research 337 (2002) 273–277
5. Scha¨fer, A.; Thiem, J. J. Org. Chem. 2000, 65, 24–29.
277
References
6. Harding, K. E.; Tiner, T. H In Comprehensi6e Organic
Synthesis; Trost, B. M., Ed.; Pergamon: Oxford, 1991;
Vol. 4, pp. 363–421.
7. Michaud, T.; Chanet-Rey, J.; Chou, S.; Gelas, J. Carbo-
hydr. Res. 1997, 303, 123–127.
8. (a) Lancelin, J.-M.; Pougny, J.-R.; Sinay¨, P. Carbohydr.
Res. 1985, 136, 369–375;
1. (a) Martin, O. R. In Carbohydrate Mimics: Concepts and
Methods; Chapleur, Y., Ed.; Wiley–VCH: Weinheim,
1997; pp. 259–282;
(b) Spak, S. J.; Martin, O. R. Tetrahedron 2000, 56,
217–224.
2. Saxena, I. M; Brown, R. M., Jr.; Fevre, M.; Geremia, R.
A.; Henrissat, B. J. Bacteriol. 1995, 177, 1419–1424 and
references cited therein.
(b) Boschetti, A.; Nicotra, F.; Panza, L.; Russo, G. J. Org.
Chem. 1988, 53, 4181–4185;
(c) Freeman, F.; Robarge, K. D. J. Org. Chem. 1989, 54,
346–359.
3. Xie, J. Recent Res. De6. Org. Chem. 1999, 3, 505–523.
4. Lay, L.; Nicotra, F.; Panza, L.; Verani, A. Gazz. Chim.
Ital. 1992, 122, 345–348.
9. Kra¨mer, B.; Franz, T.; Picasso, S.; Pruschek, P.; Ja¨ger, V.
Synlett 1997, 295–297.