Intramolecular 1,3-dipolar cycloaddition of unsaturated nitrones derived from methyl α-d-glucopyranoside

Article abstract of DOI:10.1016/j.tet.2005.04.066

The intramolecular 1,3-dipolar cycloaddition of unsaturated nitrones derived from methyl α-d-glucopyranoside with 2-furaldehyde has been studied. This cycloaddition was found to afford three 9-oxa-1-azabicyclo[4.2.1] nonane diastereomers in a 3:1:1 ratio

Full text of DOI:10.1016/j.tet.2005.04.066

Tetrahedron 61 (2005) 6816–6823
Intramolecular 1,3-dipolar cycloaddition of unsaturated nitrones
derived from methyl a-D-glucopyranoside
a
Petra Padar, Miklos Hornyak, Peter Forgo, Zoltan Kele, Gabor Paragi, Nicola M. Howarth
a
b
a
c
d
´ ´
´
´
´
´
´
and Lajos Kovacs
´
a,
*
´
a
´ ´
Department of Medicinal Chemistry, University of Szeged, Dom ter 8, H-6720 Szeged, Hungary
b
´ ´
Department of Organic Chemistry, University of Szeged, Dom ter 8, H-6720 Szeged, Hungary
c
´
´
Protein Chemistry Research Group, Hungarian Academy of Sciences, University of Szeged, Dom ter 8, H-6720 Szeged, Hungary
dChemistry, School of Engineering and Physical Sciences, William H. Perkin Building, Heriot-Watt University, Riccarton,
Edinburgh EH14 4AS, UK
Received 23 December 2004; revised 8 April 2005; accepted 29 April 2005
Abstract—The intramolecular 1,3-dipolar cycloaddition of unsaturated nitrones derived from methyl a-D-glucopyranoside with
2-furaldehyde has been studied. This cycloaddition was found to afford three 9-oxa-1-azabicyclo[4.2.1]nonane diastereomers in a 3:1:1
ratio [with the principal isomer possessing a (3S,4R,5S,6S,8S) configuration, determined by NMR spectroscopy]. The effects of different
Lewis acid catalysts (MgCl2, ZnCl2 and BF3$OEt2) on yields and diastereomeric ratios have been examined in detail. The best result (90%
yield) was achieved when MgCl2 was present (in toluene, 120 8C bath temperature, 12 h). The stereoselectivity of the 1,3-dipolar
cycloaddition was not significantly altered under the conditions investigated.
q 2005 Elsevier Ltd. All rights reserved.
1. Introduction
We envisage that it will be possible to control the
stereochemical outcome of the intramolecular 1,3-dipolar
cycloaddition by virtue of steric constraint so that the actual
number of isoxazolidine isomers produced would be
reduced compared to the theoretical. The use of different
Lewis acid catalysts is anticipated to improve both the
stereoselectivity and reactivity of the nitrone. The mechan-
ism of such 1,3-dipolar cycloadditions has been extensively
studied by several authors.3 Nitrones are nucleophiles which
Peptide nucleic acids (PNAs) are nucleic acid mimics
bearing a pseudopeptide backbone (Fig. 1).1,2 They possess
very favourable hybridisation properties with nucleic acid
targets, display high chemical and biological stabilities and
have the potential to be used as both antisense and antigene
therapeutic agents. Unfortunately, PNA has low lipid
penetration and, consequently, poor cellular uptake. In an
attempt to overcome these undesirable attributes, we have
designed a conformationally restricted oligonucleotide
analogue whose backbone should be positively charged
under physiological conditions. These new chiral nucleo-
side analogues are termed azetidine nucleic acids (ANAs,
Fig. 1). The pivotal step in the synthesis of the ANA
monomers, needed for construction of the oligomers, is a
diastereoselective intramolecular 1,3-dipolar cycloaddition
involving unsaturated nitrones derived from carbohydrate
precursors. Subsequent transformations on the correspond-
ing isoxazolidines obtained should then afford the desired
azetidine derivatives (Fig. 2).
Keywords: 1,3-Dipolar cycloaddition; Isoxazolidines; Bicyclic 1,2-oxaze-
panes; 9-Oxa-1-azabicyclo[4.2.1]nonanes; Asymmetric synthesis.
*
Corresponding author. Tel.: C36 62 545145; fax: C36 62 545141;
e-mail: kovacs@ovrisc.mdche.u-szeged.hu
Figure 1. The structure of PNA and ANA oligomers (B, nucleobase). Only
one diastereoisomer is shown for each ANA structure.
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.04.066

´ ´
P. Padar et al. / Tetrahedron 61 (2005) 6816–6823
6817
Figure 2. Retrosynthetic analysis of ANA monomers.
co-ordinate strongly to Lewis acids to form nitrone/Lewis
acid complexes. These complexes are generated easily and
they serve to assist in 1,3-dipolar cycloadditions through
stabilization of the corresponding transition state and
decreasing the energy gap between the LUMO and
HOMO of one of the substrates.4,5 Thus, a series of
catalysts (metals and their complexes) has been developed
for use in either normal or inverse electron demand 1,3-
dipolar cycloadditions, for example, Mg2C, Ti2C, Zn2C
,
Ni2C, Pd2C, Yb3C, B(III), Al3C, Cu2C.3 The 1,3-dipolar
cycloaddition of nitrones bearing heteroaryl rings with
electron-deficient alkenes has been investigated in detail by
Merino et al.6–8
Our interest in the development of an efficient route for the
synthesis of chiral isoxazolidines and, ultimately, azetidines
led us to therefore consider using a similar strategy, utilizing
unsaturated nitrones derived from carbohydrates, for their
preparation. The starting material employed for the work
reported herein was methyl a-D-glucopyranoside (1)
(Fig. 3). In this case, the 1,3-dipolar cycloaddition
investigated was a model reaction in order that suitable
reaction procedures for performing such cycloadditions
could be identified. It is envisaged that future employment
of appropriate carbohydrate derivatives from the D-manno
and D-galacto series in place of 1 will permit formation of
isoxazolidines in which the hydroxyl groups in the
carbohydrate portion are differentiated. The preparation of
such compounds is an integral part of our synthetic
approach to the desired ANA monomers.
Figure 3. a: (2/3, 9/10): Zn, sonication (1.6 g scale: 90%, 5 g scale:
30%) or Zn and Co(II)phthalocyanine (5 g scale: 70%), b: NH2OH$HCl,
NaHCO3 (70%), c: NaBH3CN, HCl/dioxane, d: 2-furaldehyde, toluene (cC
d: 30–40%), 4 A MS, 50 8C, 18 h, e: toluene, 120 8C, 4 A MS, Lewis acid
catalyst, 7: 9-oxa-1-azabicyclo-[4.2.1]nonane, 8: 8-oxa-1-azabicyclo-
[4.2.1]nonane skeletons.
˚
˚
reactions (up to 1.6 g of 2 or 9 afforded 3 or 10 in ca. 90%
yields). Thus, several attempts were made to increase the
scale of this reaction by employing activated zinc instead.
This was prepared according to established methods,13 for
example, zinc–copper alloy14 or the reduction of anhydrous
zinc chloride with various alkali metals in the presence of
naphthalene.15,16 Heating solutions of 2 or 9 in ethanol at
reflux in the presence of activated zinc produced by either
method, afforded the same result, with respect to scale and
yield (1 g scale ca. 70%, 5 g scale ca. 30% yield). When
more than 5 g of the starting 6-deoxy-6-iodo derivative (2 or
9, respectively) was used and the activated zinc was
prepared in situ from zinc chloride and lithium, the reaction
also failed to go to completion. In this case, though, the
remaining lithium in the reaction mixture caused decompo-
sition of the unsaturated aldehyde and, also, prevented
addition of water to the reaction mixture, which is necessary
to dissolve zinc salts from the surface of zinc. Thus, all these
procedures gave optimum product yields up to a maximum
1 g scale. Upon conducting further investigations into this
Boord reaction, we found that, for large scale reactions,
reasonable yields of the products 3 and 10 could be obtained
2. Results and discussion
The starting material, methyl a-D-glucopyranoside 1, was
successfully converted into the benzoylated (2) and
benzylated (9) 6-deoxy-6-iodo derivatives according to the
methods described by Garegg9 and Vasella,10 respectively
(Fig. 3). The subsequent Boord reaction on the halo
derivatives [2/3, 9/10 (Fig. 3)] was accomplished
upon treatment with zinc followed by sonication.11
However, it was discovered that acceptable yields of the
products 3 and 10 were only obtained for small scale

´ ´
P. Padar et al. / Tetrahedron 61 (2005) 6816–6823
6818
when zinc and cobalt(II) phthalocyanine was utilised
(Kleban et al.12 employed zinc and vitamin B12 for the
same purpose) rather than zinc and sonication (5 g scale, ca.
70% yield).
Having prepared unsaturated aldehydes 3 and 10, the next
step in our synthetic pathway involved treatment with
hydroxylamine at room temperature17 to give oximes 4 and
11, respectively (E/Z isomers in 1:1 ratio) (Fig. 3).
Subsequently, 4 or 1118 were reduced with sodium cyano-
borohydride and HCl/1,4-dioxane at the appropriate pH,
depending on the protecting groups present, to afford 5 or
12, respectively (Fig. 3). These hydroxylamines were used
in the next step without further purification in order to avoid
their decomposition. The mixture of HCl/1,4-dioxane had to
be added slowly due to the acid sensitive nature of the
benzoyl protecting groups and because further reduction of
the hydroxylamine could easily occur at low pH which
would result in formation of the amine instead. It was
envisaged that this amine by-product would hinder the
subsequent condensation step as it could react with
2-furaldehyde to give a Schiff’s base, drastically reducing
the yield of the cycloaddition reaction. Therefore, in an
attempt to overcome this limitation, we have investigated
performing the reduction of 4 and 11 in phosphate buffer
solutions at various pHs, ranging from 4 to 8. Unfortunately,
to date, all attempts have proved unsuccessful and so our
original approach for preparing hydroxylamines 5 and 12
has been retained for the present work. Finally, crude
hydroxylamines 5 and 12 were condensed with 2-furalde-
hyde to furnish the desired nitrones, 6 and 13, required for
investigation of the 1,3-dipolar intramolecular cyclo-
addition reaction (Fig. 3). These were afforded in overall
yields of 30–40% for the two steps, after purification.
Figure 4. The structures of 9-oxa-1-azabicyclo[4.2.1]nonanes 7a–7d.
whereas the equivalent protons in isomers 7c and 7d assume
a cis arrangement. The NOESY spectrum of 7b (Fig. 5)
indicates the spatial proximity of protons H-2a, H-4, and H-
8 and that of H-7b, H-4, H-8. Protons H-7b and H-2a reside
on the bottom face of the structure relative to the oxazepane
ring. In addition, it appears that proton H-7b is located far
from its neighbours, H-2a, H-4 and H-8, as no coupling
between H-7b and H-2a was detected (Table 4).
Table 1. Selected 1H NMR chemical shifts of compounds 7b–7d
With nitrones 6 and 13 to hand, it was now possible to
investigate the intramolecular 1,3-dipolar cycloaddition.
This was simply accomplished by heating a solution of the
Compound/
atom
7b
7c
7d
H-2a (dd)a
H-2b (dd)
H-3
H-4 (dd)
H-5
H-6
H-7a (ddd)
H-7b (ddd)
H-8 (dd)
H-30 (d)
H-40(dd)
3.22
4.22
5.85 (m)
6.00
5.68 (dd)
5.02 (m)
2.75
3.18
4.63
3.60
2.89
5.96 (ddd)
6.21
5.51 (d)
4.82 (dd)
2.83
3.14
4.74
3.02
3.53
5.89 (ddd)
5.98
5.69 (dd)
4.96 (ddd)
2.67
2.86
4.71
˚
appropriate nitrone in toluene at reflux in the presence of 4 A
molecular sieves (Fig. 3). Unfortunately, for the benzyl
protected nitrone 13, this reaction proved to be sluggish
(toluene, reflux, 1 week) and very low yielding (!10%);
therefore it was abandoned. For the benzoyl nitrone 6, this
reaction was found to be more successful and gave
isoxazolidine 7 (Fig. 3) as a mixture of diastereoisomers
in 17–90% yield as expected. These isomers were
subsequently separated by column chromatography and
characterised by NMR spectroscopy as the 9-oxa-1-
azabicyclo[4.2.1]nonane diastereoisomers 7b–7d (Fig. 4).
6.28
6.32
6.60
6.44
6.51
6.42
a Multiplicities in parenthesis.
Table 2. Selected 13C NMR chemical shifts of compounds 7b–7d
We were unable to isolate the fourth diastereisomer from the
9-oxa series (7a) (Fig. 4) as its yield was negligible. We
assume that the other alternative product from this
cycloaddition, 8-oxa-1-azabicyclo[4.2.1]nonane 8 (Fig. 3),
did not form because of steric hindrance between the furyl
side chain and the oxazepane ring.
Compound/
atom
7b
7c
7d
C-2
C-3
C-4
C-5
C-6
C-7
C-8
C-20
C-30
C-40
C-50
58.7
68.8
73.1
73.1
77.7
34.2
65.4
154.3
106.2
110.3
142.3
54.7
68.1
74.7
79.2
81.8
38.1
60.9
147.1
111.2
110.8
143.6
52.2
66.8
74.6
73.2
77.6
31.2
64.1
148.1
111.1
110.7
143.4
Tables 1 and 2 show selected proton and carbon chemical
shifts recorded in the 1H and 13C NMR spectra of
compounds 7b–7d. Upon assignment of the individual
1
resonances by means of H, 13C, HSQC and HMBC NMR
measurements, it was shown that protons H-6 and H-8 adopt
a relative trans arrangement in the principal isomer 7b

´ ´
P. Padar et al. / Tetrahedron 61 (2005) 6816–6823
6819
NOESY spectrum recorded. By taking into account the
coupling constants for all the isoxazolidine and 1,2-
oxazepane ring protons in all three isomers, we have been
able to determine the configuration of each of the newly
formed chiral centres [7b: (6S,8S); 7c: (6R,8S); and 7d:
(6S,8R) (Fig. 4)]. The stereochemistry of the remaining
chiral centres have been deduced from D-glucose and the
conformation of the 1,2-oxazepane ring.
The effects of different Lewis acid catalysts (BF3$OEt2, ZnCl2,
˚
MgCl2), solvents, absence or presence of 4 A molecular
sieves and reflux time on yields and diastereomeric ratios
for this intramolecular 1,3-dipolar cycloaddition with
nitrone 6 (Fig. 3) have been examined in detail (Table 5).
The diastereomeric ratios of 7b–7d obtained from each
reaction were initially determined by the combined use of
TLC and RP-HPLC. However, as this method proved
cumbersome and inaccurate, alternatives were sought. The
1
fortunate finding that, in the H NMR spectra, the peaks
assigned to protons H-30 and H-40 of the furyl ring were
located in unique positions for each of the three isomers,
that is, 7b–7d, led us to investigate using 1H NMR
spectroscopy instead for these measurements. This afforded
ratios which were in good agreement with those obtained
previously from RP-HPLC experiments and so, due to its
convenience and accuracy, it became the method of choice.
Figure 5. NOESY spectrum of compound 7b. Crucial correlations between
underlined protons are shown in boxes.
Table 3. Coupling constants for compounds 7b–7d (Hz)
Compd/coupling
constant
7b
7c
7d
From Table 5, it can be seen that when the intramolecular
1,3-dipolar cycloaddtion was performed in toluene, the
reflux time (24 or 48 h) had no effect on yield or
diastereomeric ratio. However, since nitrone 6 decomposed
quickly at temperatures above 100 8C, even under an argon
atmosphere, it was not advantageous to heat the reaction in
toluene at reflux for more than 24 h. 1,4-Dioxane was found
to be an unsuitable solvent for this reaction; the mixture of
isomers 7b–7d was afforded in only 17% yield.
J2a,2b
J2a,3
J2b,3
J3,4
J4,5
J5,6
J6,7a
J6,7b
J7a,7b
J7a,8
J7b,8
13.8
7.9
4.9
9.2
7.9
6.0
8.8
3.8
13.1
3.6
8.5
3.2
1.9
15.0
10.9
4.0
10.9
7.6
—
6.1
10.0
13.5
11.0
9.5
12.5
10.3
4.0
10.1
8.7
5.8
8.7
6.2
13.2
7.2
11.2
3.1
We have established that the main diastereoisomer obtained
from these cycloaddition reactions (except when 1,4-
dioxane and ZnCl2 was used) was 7b, bearing the (6S,8S)
configuration at the newly formed chiral centres (Table 5).
The maximum yield for this intramolecular 1,3-dipolar
cycloaddition was achieved when MgCl2 was added [90%,
toluene, 120 8C (bath temperature), 12 h, Table 5].
Unexpectedly, in the presence of the harder Lewis acid
catalyst, BF3$OEt2, most of the starting nitrone 6 decom-
posed after only a few hours to give an undesired product
which contained one less benzoyl group (as determined
from MS data). As a result of this finding, we propose that
this also attributed to the reduced yield observed for the
reaction performed in the presence of ZnCl2, although here
decomposition of the nitrone was slower.
0
J3 ,4
0
3.2
2.0
0
J4 ,5
0
1.9
Table 4 reveals that compounds 7b and 7d adopt similar
structures. Naturally, though, the position of protons H-8
and H-30 in the furyl ring are reversed for isomer 7d
compared to isomer 7b [7b: (8S), 7d: (8R)]. This was
confirmed by the coupling constants measured between
protons H-6, H-7 and H-8 (Table 3). The only real difference
between their structures is that proton H-7b is located closer
to protons H-2a and H-4 in isomer 7d (evaluated from the
dihedral angles). Thus, for isomer 7d, a cross-correlation
peak was visible in the NOESY spectrum (H-7b/H-2a/H-4).
In the case of isomer 7c, protons H-30/H-3/H-7a and H-5/
H-3/H-7a on the top face of the oxazepane ring are found to
be in close proximity to each other, according to the
In conclusion, we have ascertained that the optimum
conditions for performing this 1,3-dipolar cycloaddition
Table 4. NOESY data of compounds 7b–7d
Compd Connected protons (upside positions)a
Connected protons (downside positions)a
Connected protons (peripheral positions)a
H-6/H-7a; H-7a/H-30
H-7b/H-8; H-7b/H-6
H-7a/H-6; H-7a/H-8
7b
7c
7d
H-3/H-2b; H-5/H-6
H-7b/H-8/H-4; H-8/H-2a/H-4
H-2a/H-4
H-7b/H-2a/H-4/H-30
H-30/H-3/H-7a H-3/H-5/H-7a H-3/H-2b
H-3/H-2b; H-5/H-6
a Relative to the 9-oxa-1-azabicyclo[4.2.1]nonane skeleton as shown in Figure 6.

´ ´
P. Padar et al. / Tetrahedron 61 (2005) 6816–6823
6820
Table 5. The effect of Lewis acids and solvents on yields and diastereomeric ratios
Solvent, mol
sieves
Catalyst
Time (h)
Temperature (8C)a
Yield (%)
Diastereomeric
ratios (HPLC)b
Diastereomeric
ratios (NMR)b,c
˚
Toluene, 4 A
—
—
—
—
24
48
24
24
24
12
24
120
120
80
100
120
120
120
70
75
32
17
50
90
2.8:1:1
2.8:1:1.2
2.1:1:1
1:1.1:1.2
4:1:1e
2:1:1
3:1:1
3:1:1
2:1:1
Toluene
Benzene, 4 A
˚
d
˚
1,4-Dioxane, 4 A
—
˚
Toluene, 4 A
ZnCl2
MgCl2
BF3$OEt2
0.5:1:1
2:1:1
—
˚
Toluene, 4 A
f
˚
Toluene, 4 A
—
—
a Bath temperature.
b Ratio of isomers 7b:7c:7d isolated from the reaction mixture.
c Determined from the integral of protons H-30 and H-40.
d Not determined due to the presence of impurities.
e Diastereomer 7b could not be separated from an impurity.
f
The starting material decomposed and a by-product was formed (see text).
reaction with nitrone 6 (Fig. 3) involve using toluene as the
solvent and MgCl2 as the Lewis acid catalyst. It appears that
if the Lewis acid catalyst added is hard, a side reaction
involving elimination of a benzoyl protecting group from
the starting nitrone becomes significant and this may be
accompanied by decomposition and conversion of the furyl
group, too. The stereoselectivity of the cycloaddition was
found not to alter much under the conditions investigated
here.
calculations were conducted on a representative for each of
the different energy clusters. These utilised the Hartree–
Fock method with 3-21 Gaussian basis set and applied the
Gaussian03 code.20 Although this is one of the simplest
methods, it was reasoned that this was sufficient for our
purposes. The conformations of the oxazepane rings for
isomers 7b–7d derived from the computational studies were
found to be in very good agreement with the structures
obtained previously from NMR studies. In Figure 6, the
optimized geometry of isomer 7b is presented. The
spatial proximity of protons H-7b, H-8, H-4 and H-2a on
the bottom face can be clearly seen. In addition to the
theoretical calculations providing information about the
geometry of isomers 7a–7d, we had hoped that they could
also be used to predict the diastereomeric ratio afforded by
the cycloaddition reaction (7a:7b:7c:7dz0:3:1:1) based on
the total energy of each of the isomers.
3. Theoretical investigations
In order to analyse the geometry of all four diastereoisomers
of the isozaxolidine derivative (i.e., 7a–7d (Fig. 4))
produced from the intramolecular 1,3-dipolar cycloaddition
with nitrone 6 (Fig. 3), we have conducted a systematic
computational investigation. This involved performing
molecular dynamics simulations followed by high-level ab
initio calculations. For the molecular dynamics studies, the
‘simulated annealing’ protocol described in the SYBYL
program package19 was employed to obtain the required
starting geometries for isomers 7a–7d for the subsequent
higher level investigations. The Merck’s force field
parameter set (MMFF94) was applied with its own charge
distribution. The molecules were equilibrated for 2000 fs at
1200 K and then cooled to 50 K exponentially over
10,000 fs. In this way, 1000 conformations were provided
for each isomer. Next a semi-empirical optimization using
the PM3 method was performed and the results afforded
were grouped according to their energies. Finally, ab initio
However, it appeared that, at this level of theory, significant
differences could not be observed with the total energies for
diastereoisomers 7a–7d being approximately the same. In
Table 6, the HF/3-21G total energies of the minimized
geometries for each of the different isozaxolidine isomers
are presented.
Table 6. Calculated total energies of compounds 7a–7da
Compound
Total energy (hartree)
7a
7b
7c
7d
K1870.388634
K1870.389014
K1870.389014
K1870.389272
a Ab initio (HF/3-21G method).
4. Conclusion
We have successfully synthesized a variety of isoxazolidine
derivatives of chiral moieties employing a Lewis acid-
catalyzed 1,3-dipolar cycloaddition. In our preliminary
studies, the exclusive formation of 9-oxa-1-azabicyclo-
[4.2.1]nonane diastereomers 7b–7d (Fig. 4) from nitrone 6
(Fig. 3) has been observed. The maximum yield for the 1,3-
dipolar cycloaddition reaction was achieved with MgCl2 as
the Lewis acid catalyst [toluene, 120 8C (bath temperature),
12 h]. The ratio of the diastereoisomers of 7b–7d was 3:1:1
and this did not alter significantly under the different
Figure 6. The lowest-energy conformation of compound 7b calculated by
HF/3-21 method. Proximal hydrogen atoms of 9-oxa-1-azabicyclo[4.2.1]-
nonane skeleton, supporting the configuration and conformation of the
above compound (NMR evidence), are shown with the H prefix. Carbon
atoms of the above skeleton and those of the furan skeleton (primed
numbers) are labelled without the C prefix for clarity.

´ ´
P. Padar et al. / Tetrahedron 61 (2005) 6816–6823
6821
reaction conditions investigated here. The theoretical total
energies of the minimized geometries for 7a–7d obtained
computationally failed to rationalise the experimentally
determined diastereomeric ratios of these cycloadducts
produced from the intramolecular 1,3-dipolar cycloaddition.
Further studies are in progress to obtain chiral 1,2,4-
substituted azetidines from the cycloadducts afforded and
new cycloaddition reactions are envisaged.
hydroxylamine hydrochloride (3.41 g, 41.19 mmol,
4.5 equiv) and sodium hydrogen carbonate (3.41 g,
49.80 mmol, 4.6 equiv). After 2 h stirring at room tempera-
ture, the ethanol was removed under reduced pressure, and
the residue was dissolved in dichloromethane (100 mL).
The solution was washed with water (3!50 mL), dried
(MgSO4) and evaporated in vacuo. Further purification was
accomplished by column chromatography [10%–50% (v/v)
EtOAc in hexane] to give oxime 4 as a light yellow oil
(5.16 g, 70%, 1:1 mixture of E and Z isomers). Rf: 0.60; 0.53
(E/Z isomer), (1:1, hexane/EtOAc); IR (film, nmax/cmK1):
3426m, 3069w, 2984w, 1726s, 1601m, 1450m, 1315m,
1260s, 1246s, 1177m, 1105s, 1094s, 1069s, 1026m, 942w,
710s; dH (500 MHz, CH3OD): 5.36 (m, 2H, H-6a, H-6b),
6.02 (m, 4H, H-2, H-3, H-4, H-5), 6.60 (m, 1H, H-1, E) 6.80
(d, 1H, J1,2Z5.4 Hz, H-1, Z), 7.34–7.56 (m, 9H, arom. H),
7.96–8.04 (m, 6H, arom. H), 8.63 (s, 1H, OH). dc (125 MHz,
CH3OD): 68.1 (C-2, Z), 71.8 (C-2, E), 73.8 (C-3, Z), 74.5
(C-3, E), 74.6 (C-4, E), 74.8 (C-4, Z), 120.4 (C-6, E), 121.0
(C-6, Z), 129.6–129.7 (6!arom. CH), 130.7 (3!arom. Cq),
130.8 (6!arom. CH), 132.7 (C-5, Z), 134.6 (C-5, E), 134.7
(3!arom. CH), 145.9 (C-1, E) 147.0 (C-1, Z), 166.6
(C]O), 166.7 (C]O), 167.1 (C]O); LRMS (ESI): m/z
474 (57%, [MCH]C), 491 (100%, [MCNH4]C), 496 (75%,
[MCNa]C); HRMS (FAB, glycerol): Calcd for C27H24NO7
[MCH]Cm/z 474.15473, found m/z 474.15418.
5. Experimental
5.1. General procedures
The following abbreviations are employed: ACN (aceto-
nitrile); ANA (azetidine nucleic acid(s)); anh. (anhydrous);
Bn (benzyl); Bz (benzoyl); CDCl3 (deuterochloroform);
CH2Cl2 (dichloromethane); ESI (electrospray ionization);
EtOAc (ethyl acetate); FAB (fast atom bombardment);
HRMS (high resolution mass spectrometry); LRMS (low
resolution mass spectrometry); MeOD (deuteromethanol);
MeOH (methanol); PNA [peptide nucleic acid(s)]; rt (room
temperature); THF (tetrahydrofuran).
Chemicals were purchased from Aldrich, Fluka, Merck or
Reanal (Budapest, Hungary). 2-Furaldehyde and BF3$OEt2
were freshly distilled prior to use. Anhydrous solvents and
anhydrous Lewis acid catalysts were prepared as
described.21 Organic solutions were dried using anhydrous
MgSO4 and evaporated in Bu¨chi rotary evaporators. TLC:
Kieselgel 60 F254 (Merck), solvent systems: CH2Cl2/MeOH,
hexane/EtOAc, visualization: UV light, H2SO4/ethanol.
Mp: Electrothermal IA 8103 apparatus. IR spectra: Bio-
5.1.2. (2S,3S,4R)-2,3,4-Tris(benzoyloxy)-N-(furan-2-
ylmethylene)hex-5-en-1-amine oxide (6). To a stirred
solution of oxime 4 (0.50 g, 1.06 mmol, 1 equiv) in dioxane
(5 mL) was added sodium cyanoborohydride (0.22 g,
3.20 mmol, 3 equiv) in small portions, while the solution
was carefully treated with HCl/dioxane (1.7 M, w1 mL), to
maintain the pH between 3 and 5. After completion of
reaction (TLC), the solution was evaporated in vacuo, the
residue was dissolved in EtOAc (50 mL), the organic layer
was washed with aqueous sodium carbonate (40 mL), water
(40 mL) and brine (40 mL), dried (MgSO4) and evaporated
in vacuo. The resulting hydroxylamine 5 was used without
any further purification.
Rad FTS-60A (KBr pellets unless otherwise stated, nmax
/
cmK1, s, strong; m, medium; w, weak). NMR: Bruker
Avance DRX 400 and 500 spectrometers (1H: 400.13,
500.13 MHz; 13C: 125.76 MHz, respectively), MeOD,
CDCl3 solutions, d (ppm), J (Hz). Spectral patterns: s,
singlet; d, doublet; dd, doublet of doublet; t, triplet; m,
multiplet; br, broad; deut, deuterable. The superscripts *, #
denote interchangeable assignments. For the 2D experi-
ments (HSQC, HMBC, NOESY) the standard Bruker
software packages (INV4GSSW, INV4GSLRNDSW)
were applied. LRMS: Finnigan MAT TSQ 7000, ESI
technique. HRMS: VG ZAB SEQ high resolution mass
spectrometer using FAB ion source. Samples were dis-
solved in glycerol, the resolution of the instrument was
10,000. TLC/MS; TLC/HPLC: the analyte solution has been
applied onto a 5 cm wide silica gel TLC plate as a band to
obtain sufficient material. After developing in a solvent
system the appropriate band was removed, the silica gel was
suspended in MeOH (100 mL for MS, 1000 mL for HPLC),
sonicated, centrifuged and the supernatant was used for MS
analysis and HPLC (for HPLC 10 mL was injected). HPLC:
SHIMADZU UV/VIS detector: SPD-10A VP, pump: LC-
10AC VP, column: LiChrospher 6.1, RP select B (5 mm).
Eluent system: gradient: 70–90% ACN within 25 min, flow
rate: 1 mL/min.
Compound 5 was dissolved in toluene (20 mL) and treated
with freshly distilled 2-furaldehyde (176 mL, 2.12 mmol) in
˚
the presence of 4 A molecular sieves. After stirring at 50 8C
for 18 h, the solution was filtered, evaporated under reduced
pressure and co-evaporated with ACN (3!50 mL). The
crude residue was purified by column chromatography (10–
30% (v/v) EtOAc in hexane) to give nitrone isomers 6 as a
pale yellow foam (0.23 g, 40%). Rf: 0.17; 0.23 (E/Z isomer,
7:3, hexane/EtOAc); IR (KBr, nmax/cmK1): 3429w, 3065w,
2980w, 1728s, 1612m, 1601w, 1450w, 1315m, 1261s,
1240s, 1179w, 1107s, 1096s, 1069m, 1026w, 948w, 863m,
708s; dH (500 MHz, CDCl3): 4.31 (d, 1H, J1,2Z5.5 Hz,
H-1), 5.35 (d, 1H, J5,6bZ10.1 Hz H-6b, Z), 5.45 (d, 1H,
H-6a, J5,6aZ16.6 Hz, E), 5.95–6.07 (m, 4H, H-2, H-3, H-04,
0
0
0
0
H-5), 6.16 (dd, 1H, J3 ,4 Z1.7 Hz, J4 ,5 Z3.4 Hz, H-4 ),
6.52 (d, 1H, J3 ,4 Z1.7 Hz, H-30), 7.26 (s, 1H,
0
0
KOCN]CH), 7.34–7.42 (m, 9H, arom. H), 7.79 (d, 1H,
0
0
0
J4 ,5 Z3.4 Hz, H-5 ), 7.97–8.04 (m, 6H, arom. H). dC
(125 MHz, CDCl3): 65.3 (C-1), 69.5 (C-2), 73.0 (C-4*),
73.5 (C-3*), 112.4 (C-30), 116.3 (C-40), 120.7 (C-6), 127.3
(CH]N), 128.4 (3!arom. CH), 128.5 (3!arom. CH),
128.6 (arom. Cq), 129.7 (3!arom. CH), 129.9 (3!arom.
5.1.1. (2S,3S,4R)-1-(Hydroxyimino)hex-5-ene-2,3,4-triyl
tribenzoate (4). The unsaturated aldehyde 311 (5.00 g,
10.92 mmol, 1 equiv) was dissolved in ethanol (100 mL)
and distilled water (30 mL). The solution was treated with

´ ´
P. Padar et al. / Tetrahedron 61 (2005) 6816–6823
6822
CH), 131.50(C-5), 133.2 (arom. CH), 133.5 (2!arom. CH),
144.0 (C-5 ), 146.3 (C-20), 165.1 (C]O), 165.3 (C]O),
165.7 (C]O); LRMS (ESI): m/z 554 (100%, [MCH]C),
576 (20%, [MCNa]C); HRMS (FAB, glycerol): Calcd for
C32H28NO8 [MCH]Cm/z 554.18094, found m/z 554.18324.
J7a,8Z11.0 Hz, J7b,8Z9.5 Hz, H-8), 4.82 (dd, 1H, J6,7b
10.0 Hz, J6,7aZ6.1 Hz, H-6), 5.51 (d, 1H, J4,5Z7.6 Hz,
H-5), 5.96 (ddd, 1H, J3,4Z10.9 Hz, J2a,3Z10.9 Hz, J2b,3
Z
Z
4.0 Hz, H-3), 6.21 (dd, 1H, J3,4Z10.9 Hz, J4,5Z7.6 Hz,
0
0
0
0
0
H-4), 6.44 (dd, 1H, J3 ,4 Z3.2 Hz, J4 ,5 Z2 Hz, H-4 ), 6.60
0
0
0
(d, 1H, J3 ,4 Z3.2 Hz, H-3 ), 7.19–7.54 (m, 9H, arom. H,
1H, H-50), 7.75–8.00 (m, 6H, arom. H); dC (125.76 MHz,
CDCl3): 38.1 (C-7), 54.7 (C-2), 60.9 (C-8), 68.1 (C-3), 79.2
(C-5), 74.7 (C-4), 81.8 (C-6), 110.8 (C-40) 111.2 (C-30),
127.6 (3!arom. CH), 127.9 (3!arom. CH), 129.0 (3!
arom. CH, 3!arom. Cq), 132.5 (3!arom. CH), 132.9 (3!
arom. CH), 143.6 (C-50), 147.1 (C-20), 164.3 (Cq), 165.1
(Cq), 165.7 (Cq); LRMS (ESI): m/z 554 (100%, [MCH]C),
576 (43%, [MCNa]C); HRMS (FAB, glycerol): Calcd for
C32H28NO8 [MCH]Cm/z 554.18094, found m/z 554.18268.
5.1.3. (3S,4R,5S,6S,8S)-8-(Furan-2-yl)-9-oxa-1-aza-
bicyclo[4.2.1]nonane-3,4,5-triyl tribenzoate (7b); (3S,
4R,5S,6R,8S)-8-(furan-2-yl)-9-oxa-1-azabicyclo-[4.2.1]-
nonane-3,4,5-triyl tribenzoate (7c) and (3S,4R,5S,
6S,8R)-8-(furan-2-yl)-9-oxa-1-azabicyclo[4.2.1]-nonane-
3,4,5-triyl tribenzoate (7d). General procedure. The pure
nitrone 6 (0.40 g, 0.72 mmol) was dissolved in dry toluene
˚
(20 mL) in the presence of 4 A molecular sieves and the
solution was heated at 80–100–120 8C (bath temperature)
for 12–48 h under an argon atmosphere. The reaction
mixture was cooled, filtered, evaporated in vacuo and co-
evaporated with ACN (3!50 mL). Chromatographic puri-
fication [0–10% (v/v) EtOAc in hexane], yielded three
isomers (7b,7c,7d) as an oil or foam, in different isomer
ratios depending on the reaction conditions, in 17–90%
yield (Table 6). The ratios of the three diastereisomers were
determined by TLC, RP-HPLC and 1H NMR (Table 6). The
reactions were carried out under different conditions, for
example, solvents, temperature, reflux time and with or
without Lewis acid catalysts (see Table 5). The catalysts,
anh. ZnCl2, anh. MgCl2 and freshly distilled BF3$OEt2,
were used in 0.1–0.2 mol equiv with respect to nitrone. Rf:
0.32 (7:3, hexane/EtOAc as a single spot; after 5!
development in an 8:2, hexane/EtOAc eluent it could be
separated into three isomers).
Compound 7d: IR (KBr, nmax/cmK1): 3067w, 2928w,
1730s, 1601w, 1450w, 1315w, 1281s, 1260s, 1179w,
1109m, 1096m, 1070m, 1026w, 710m; dH (500 MHz,
CDCl3): 2.67 (ddd, 1H, J7a,7bZ13.2 Hz, J6,7aZ8.7 Hz,
J7a,8Z7.2 Hz, H-7a), 2.86 (ddd, 1H, J7a,7bZ13.2 Hz, J7b,8
11.2 Hz, J6,7bZ6.2 Hz, H-7b), 3.02 (dd, 1H, J2a,2bZ12.5 Hz,
J2a,3Z10.3 Hz, H-2a), 3.53 (dd, 1H, J2a,2bZ12.5 Hz, J2b,3
4.0 Hz, H-2b), 4.71 (dd, 1H, J7a,8Z7.2 Hz, J7b,8Z11.2 Hz,
H-8), 4.96 (ddd, 1H, J6,7aZ8.7 Hz, J6,7bZ6.2 Hz, J5,6
Z
Z
Z
5.8 Hz, H-6), 5.89 (ddd, 1H, J2a,3Z10.3 Hz, J3,4Z10.1 Hz,
J2b,3Z4.0 Hz, H-3), 5.69 (dd, 1H, J4,5Z8.7 Hz, J5,6Z5.8 Hz,
H-5), 5.98 (dd, 1H, J3,4Z10.1 Hz, J4,5Z8.7 Hz, H-4), 6.42
0
0
0
0
0
(dd, 1H, J3 ,4 Z3.1 Hz, J4 ,5 Z1.9 Hz, H-4 ), 6.51 (d, 1H,
0
0
0
J3 ,4 Z3.1 Hz, H-3 ), 7.20–7.25 (m, 3H, arom. H), 7.39–7.41
(m, 3H, arom. H), 7.49–7.52 (m, 4H, arom. H, H-50), 7.78–
7.82 (m, 3H, arom. H), 7.97–7.99 (m, 3H, arom. H); dC
(125 MHz, CDCl3): 31.2 (C-7), 52.2 (C-2), 64.1 (C-8), 66.8
(C-3), 73.2 (C-5), 74.6 (C-4), 77.6 (C-6), 111.1 (C-30), 110.7
(C-40), 128.1, 128.2 (3!arom. CH), 128.5 (3!arom. Cq,)
129.6 (3!arom. CH) 129.8 (3!arom. CH), 133.0 (3!arom.
CH), 133.4 (3!arom. CH), 143.4 (C-50), 148.1 (C-20), 165.0
(Cq), 165.4 (Cq), 165.9 (Cq); LRMS (ESI): m/z 554 (100%,
[MCH]C), 576 (56%, [MCNa]C); HRMS (FAB, glycerol):
Calcd for C32H28NO8 [MCH]Cm/z 554.18094, found m/z
554.18379.
Compound 7b: IR (KBr, nmax/cmK1): 3429w, 3065w,
2961w, 1726s, 1601w, 1450w, 1315m, 1281s, 1261s,
1179w, 1107s, 1096s, 1069m, 1026w, 708s; dH (500 MHz,
CDCl3): 2.75 (ddd, 1H, J7a,7bZ13.1 Hz, J6,7aZ8.8# Hz,
J7a,8Z3.6* Hz, H-7a), 3.18 (ddd, 1H, J7a,7bZ13.1 Hz,
J7b,8Z8.5# Hz, J6,7bZ3.8* Hz, H-7b), 3.22 (dd, 1H,
J2a,2bZ13.8 Hz, J2a,3Z7.9 Hz, H-2a), 4.22 (dd, 1H,
J2a,2bZ13.8 Hz, J2b,3Z4.9 Hz, H-2b), 4.63 (dd, 1H,
J7b,8Z8.5# Hz, J7a,8Z3.6* Hz, H-8), 5.02 (m, 1H, H-6),
5.68 (dd, 1H, J4,5Z7.9 Hz, J5,6Z6.0 Hz, H-5), 5.85 (m, 1H,
NOESY data: Table 4.
H-3), 6.00 (dd, 1H, J3,4Z9.2 Hz J4,5Z7.9 Hz, H-4), 6.28 (d,
0
1H, J3 ,4 Z3.2 Hz, H-3 ), 6.32 (dd, 1H, J3 ,40Z3.2 Hz,
0
0
0
0
0
0
0
J4 ,5 Z1.9 Hz, H-4 ), 7.24–7.50 (m, 10H, arom. H, H-5 ),
7.86–8.00 (m, 6H, arom. H); dC (125 MHz, CDCl3): 34.2
(C-7), 58.7 (C-2), 65.4 (C-8), 68.8 (C-3), 73.1 (C-4, C-5),
77.7 (C-6), 106.2 (C-30), 110.3 (C-40), 128.3 (3!arom.
CH), 128.4 (3!arom. C), 129.6 (3!arom. Cq, 3!arom.
CH), 129.7 (3!arom. CH), 133.2 (3!arom. CH), 142.3
(C-50), 154.3 (C-20), 165.0 (Cq), 165.1 (Cq), 165.6 (Cq);
LRMS (ESI): m/z 554 (100%, [MCH]C), 576 (50%,
[MCNa]C); HRMS (FAB, glycerol): Calcd for C32H28NO8
[MCH]Cm/z 554.18094, found m/z 554.18213.
Acknowledgements
´
We would like to thank Mr. Balazs Leitgeb (Biological
Research Center of the Hungarian Academy of Sciences,
Szeged, Hungary) and Dr. Ferenc Bogar (Protein Chemistry
Research Group of the Hungarian Academy of Sciences,
Szeged, Hungary) for their help with the molecular
dynamics calculations, and for the useful scripts, Mrs.
´
¨
Gyorgyi Ferenc for her assistance with the LRMS
measurements, Dr. Pal T. Szabo (Chemical Research
Center, Mass Spectrometry Department, Budapest,
´
Hungary) for the HRMS measurements, Dr. Otto Berkesi
(Department of Physical Chemistry, University of Szeged,
Compound 7c: IR (KBr, nmax/cmK1): 3433w, 3063w,
2924w, 1726s, 1601w, 1450m, 1315m, 1281s, 1271s,
1179w, 1107s, 1069m, 1026m, 710s; dH (500 MHz,
CDCl3): 2.83 (ddd, 1H, J7a,7bZ13.5 Hz, J7a,8Z11.0 Hz,
J6,7aZ6.1 Hz, H-7a), 2.89 (dd, 1H, J2a,2bZ15.0 Hz,
J2b,3Z4.0 Hz, H-2b), 3.14 (ddd, 1H, J7a,7bZ13.5 Hz,
J6,7bZ10.0 Hz, J7b,8Z9.5 Hz, H-7b), 3.60 (dd, 1H,
J2a,2bZ15.0 Hz, J2a,3Z10.9 Hz, H-2a), 4.74 (dd, 1H,
´
´
´
Hungary) for IR spectra and Dr. Tamas Martinek (Depart-
ment of Pharmaceutical Chemistry, University of Szeged,
Hungary) for some NMR measurements. This research has
been supported by the Wellcome Trust (Grant:

´ ´
P. Padar et al. / Tetrahedron 61 (2005) 6816–6823
6823
14. Smith, R. D.; Simmons, H. E. In Organic Synthesis, Coll. Vol.
1973 pp 855–858.
063879/Z/01/Z) and EU FP6 (Grant: Targeting Replication
and Integration of HIV (TRIoH, LSHB-CT-2003-503480)).
15. Rieke, R. D.; Li, P. T. J.; Burns, T. P.; Uhm, S. T. J. Org.
Chem. 1981, 46, 4323–4324.
16. Rieke, R. D. Aldrichimica Acta 2000, 33, 52–60.
17. Vasella, A. Helv. Chim. Acta 1977, 60, 426–446.
18. Gallos, J. K.; Koumbis, A. E.; Xiraphaki, V. P.; Dellios, C. C.;
Coutouli-Argyropoulou, E. Tetrahedron 1999, 55,
15167–15180.
References and notes
1. Egholm, M.; Buchardt, O.; Christensen, L.; Behrens, C.;
Freier, S. M.; Driver, D. A.; Berg, R. H.; Kim, S. K.; Norden,
B.; Nielsen, P. E. Nature 1993, 365, 566–568.
2. Peptide Nucleic Acids: Protocols and Applications; Nielsen,
P. E., Egholm, M., Eds.; Horizon Scientific: Wymondham,
UK, 1999.
19. Tripos Inc. Sybyl 6.9. 1699 South Hanley Rd, St. Louis,
Missouri 63144, USA. 2002.
20. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.;
Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar,
S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada,
M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida,
M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.;
Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala,
P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg,
J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain,
3. Gothelf, K. V. In Cycloaddition Reactions in Organic
Synthesis; Kobayashi, S., Jorgensen, K. A., Eds.; Wiley-
VCH: Weinheim, 2002; pp 211–247.
4. Gothelf, K. V.; Jorgensen, K. A. Acta Chem. Scand. 1996, 50,
652–660.
5. Tanaka, J.; Kanemasa, S. Tetrahedron 2001, 57, 899–905.
6. Merino, P.; Anoro, S.; Merchan, F. L.; Tejero, T. Molecules
2000, 5, 132–152.
7. Merino, P.; Anoro, S.; Cerrada, E.; Laguna, M.; Moreno, A.;
Tejero, T. Molecules 2001, 6, 208–220.
¨
M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghava-
8. Merino, P.; Tejero, T.; Laguna, M.; Moreno, A.; Cerrada, E.;
Lopez, J. A. Org. Biomol. Chem. 2003, 1, 2336–2342.
9. Garegg, P. J.; Samuelsson, B. J. Chem. Soc., Perkin Trans. 1
1980, 2866–2869.
chari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.;
´
Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.,
Gaussian 03, Revision A.1. Gaussian, Inc., Pittsburgh, PA.
2003.
10. Bernet, B.; Vasella, A. Helv. Chim. Acta 1979, 62, 1990–2016.
11. Skaanderup, P. R.; Hyldtoft, L.; Madsen, R. Monatsh. Chem.
2002, 133, 467–472.
12. Kleban, M.; Kautz, U.; Greul, J.; Hilgers, P.; Kugler, R.; Dong,
¨
H. Q.; Jager, V. Synthesis 2000, 1027–1033.
13. Erdik, E. Tetrahedron 1987, 43, 2203–2212.
21. Armarego, W. L. F.; Chai, C. L. L. Purification of Laboratory
Chemicals, 5th ed.; Elsevier: Amsterdam, 2003.