A novel carbocyclic ring closure of hex-5-enopyranosides and pent-4- enofuranosides

Article abstract of DOI:10.1055/s-1999-2792

Nitrile oxide cycloadditions to hex-5-enopyranosides and pent-4- enofuranosides give spiroisoxazolines, which upon reductive opening of the heterocyclic ring undergo spontaneous intramolecular aldol-like condensation to give six- and five-membered densely

Full text of DOI:10.1055/s-1999-2792

LETTER
1289
A Novel Carbocyclic Ring Closure of Hex-5-enopyranosides and Pent-4-
enofuranosides
John K. Gallos,* Theocharis V. Koftis, Alexandros E. Koumbis, Vassilios I. Moutsos
Department of Chemistry, Aristotle University of Thessaloniki, Thessaloniki 540 06, Greece
Fax +31-997679, E-mail: igallos@chem.auth.gr
Received 16 April 1999
Dedicated with appreciation to Professor A. Varvoglis on the occasion of his 60th birthday
Abstract: Nitrile oxide cycloadditions to hex-5-enopyranosides
and pent-4-enofuranosides give spiroisoxazolines, which upon re-
ductive opening of the heterocyclic ring undergo spontaneous in-
tramolecular aldol-like condensation to give six- and five-
membered densely functionalised carbocycles.
Key words: carbohydrates, carbocycles, enaminones, hydrogena-
tion, nitrile oxides
Carbohydrates have been increasingly used as a chiral
pool in natural product synthesis.1,2 Polyoxygenated cy-
clopentane and cyclohexane rings, - ubiquitously occur-
ring in biologically interesting compounds (cyclitols,
aminocyclitols, modified nucleosides, prostaglandins,
etc.) - are most often prepared by carbocyclic ring closure
of sugar templates; a range of such methods have been de-
veloped in the last twenty years,2 including carbanion cy-
clisations,3 cycloaddition reactions,2,4 free radical
cyclisations,1,5 cyclisations via organometallic intermedi-
ates,6 ring closing olefin metathesis7 and other reactions.8
Among them, the Ferrier cyclisation reaction of hex-5-
enopyranosides affording six-membered carbocycles in
the presence of mercury(II) salts9 is perhaps the most pop-
ular one. However, this method suffers from its inability
to be used for the preparation of cyclopentanoids, not to
mention the use of the toxic mercury salts. In an extension
of our recent work on the synthesis of carbocycles srarting
from sugars,10 we report herein a novel conversion of both
hex-5-enopyranosides and pent-4-enofuranosides to six-
and five-membered carbocycles, respectively, in two
steps. Our strategy involves nitrile oxide cycloaddition to
hex-5-enopyranosides and pent-4-enofuranosides, fol-
lowed by reductive cleavage of the resulting isoxazoline
ring.11 Subsequent intramolecular aldol-like condensation
gives the carbocyclic products.
Scheme 1 Reagents and Conditions: i. (a) 2,6-Cl2C6H3CNO (1.1
equiv.), CH2Cl2, reflux, 4 h, or (b) p-MeC6H4CH=NOH, NCS, pyridi-
ne, Et3N, CHCl3, reflux, 90 min, or (c) EtNO2, PhNCO, Et3N, benze-
ne, 20 ºC, 5 days. ii. Ranay Ni, MgSO4, H2 (1 atm), MeOH/CH3CO2H,
(6:1), 20 ºC, 90 min.
was obtained in the same solvent mixture with 20 equiva-
lents of boric acid.
With these results in hands, we extended our study by us-
ing in situ prepared nitrile oxides by oxidation of
aldoximes16 or dehydration of primary nitroparaffins17
and also by performing the cycloadditions with glucose
template 4. The ribose adducts 2a, b, c were formed as
single diastereoisomers, whereas the glucose analogs 5a,
b, c were found to be a ca. 9:1 mixture of diastereoisomers
- the major one being depicted in Scheme 1.14,15 It is inter-
esting that both isomers gave the same products in the
next step.
We initially started by adding the stable 2,6-dichloroben-
zonitrile oxide to the ribose derivative 1.12 The resulting
spiro-isoxazoline 2a (Scheme 1) was further subjected to
N-O bond cleavage by catalytic Raney nickel13 hydroge-
nation in methanol/acetic acid 6:1 to give products 3a1
and 3a2 in high combined yields, which were easily sepa-
rated by column chromatography.14,15 Conditions proved
to be crucial for the reaction outcome. Poorer yields of 3a1
and 3a2 were obtained in methanol-water 9:1 with 2-10
equivalents boric acid additive,11 while no cyclic product
Reductive N-O bond cleavage under the conditions estab-
lished in the case of 2a14 led again to carbocyclic products
in generally good yields, but the MeO or OH group at po-
sition C-1 (sugar numbering) was missing and in the case
Synlett 1999, No. 8, 1289–1291 ISSN 0936-5214 © Thieme Stuttgart · New York

1290
J. K. Gallos et al.
LETTER
of R=Me, a 1,3-diketone was formed together with the In conclusion, a novel method for preparing carbocycles
enaminone. All new compounds (2, 3, 5, 6)14 were isolat- from carbohydrates has been developed, which comple-
ed by column chromatography (silica gel, hexane/ethyl ments the Ferrier method as it can be applied to the syn-
acetate) and characterised by their spectral and analytical thesis of 5-membered rings. Further studies on the scope
data. Despite their dense functionalisation, all compounds and limitations of our method as well as its application in
prepared were found to be stable enough at room temper- the synthesis of target molecules are in progress.
ature, while they can be stored in the refrigerator for sev-
eral months without appreciable decomposition.
Acknowledgement
The absolute configuration of the newly formed stereo-
centers in cycloadducts 2 and 5 was deduced from NOE
experiments: the mutual signal enhancement observed be-
A. E. Koumbis.
tween one of the methylene protons (5-H for 2 and 6-H for
5) and one methyl of the acetonide group in 2 or the MeO
We are grateful to Dr. E. Coutouli-Argyropoulou for the NOE ex-
periments and the “Leonidas Zervas” Foundation for a fellowship to
References and Notes
group in the major isomer of 5, strongly support the as-
signed structures.
(1) Fraser-Reid, B. Acc. Chem. Res. 1996, 29, 57.
(2) (a) Ferrier, R. J.; Middleton, S. Chem. Rev. 1993, 93, 2779.
(b) Berecibar, A.; Grandjean, C.; Siriwardena, A. Chem. Rev.
1999, 99, 779.
(3) (a) Chretien, F.; Khaldi, M.; Chapleur, Y. Tetrahedron Lett.
1997, 38, 5977. (b) Cere, V.; Peri, F.; Pollicino, S.
Tetrahedron Lett. 1997, 38, 7797.
(4) Gothelf, K. V.; Jorgensen, K. A. Chem. Rev. 1998, 98, 863.
(5) (a) Martinez-Grau, A.; Marco-Contelles, J. Chem. Soc. Rev.
1998, 27, 155. (b) Marco-Contelles, J.; Alhambra, C.;
Martinez-Grau, A. Synlett 1998, 693.
(6) Selected recently published papers: (a) Ito, H.; Motoki, Y.;
Taguchi, T.; Hanzawa, Y. J. Am. Chem. Soc. 1993, 115, 8835.
(b) Mukai, C.; Uchiyama, M.; Sakamoto, S.; Hanaoka, M.
Tetrahedron Lett. 1995, 36, 5761. (c) Iimori, T.; Takahashi,
H.; Ikegami, S. Tetrahedron Lett. 1996, 37, 649. (d) Das, S.
K.; Mallet, J.-M.; Sinay, P. Angew. Chem. Int. Ed. Engl. 1997,
36, 493. (e) Sollogoub, M.; Mallet, J.-M.; Sinay, P.
Tetrahedron Lett. 1998, 39, 3471.
(7) (a) Ovaa, H.; Codee, J. D. C.; Lastdrager, B.; Overleeft, H. S.;
van der Marel, G. A., van Boom, J. H. Tetrahedron Lett. 1998,
39, 3025. (b) Sellier, O.; Van de Weghe, P.; Le Nouen, D.;
Strehler, C.; Eustache, J. Tetrahedron Lett. 1999, 40, 853.
(8) (a) Gallos, J. K.; Koftis, T. V.; Koumbis, A. E. J. Chem. Soc.,
Perkin Trans. 1, 1994, 611. (b) Ohira, S.; Sawamoto, T.;
Yamato, M. Tetrahedron Lett. 1995, 36, 1537. (c) Sudha, A.
V. R. L.; Nagarajan, M. Chem. Commun., 1998, 925.
(9) Ferrier, R. J. J. Chem. Soc., Perkin Trans. 1 1979, 1455.
(10) (a) Gallos, J. K.; Koumbis, A. E.; Apostolakis, N.E. J. Chem.
Soc., Perkin Trans. 1 1997, 2457. (b) Gallos, J. K.; Koftis,
T.V.; Karamitrou, V. I.; Koumbis, A. E. J. Chem. Soc., Perkin
Trans. 1 1997, 2461.
With the exception of 6a, the proton shifts (in CDCl3 at
20 oC) of the amino group in all enaminones appeared at
d ~5.5 and ~10.0 as two broad single signals, which re-
veals an intramolecular hydrogen bond demonstrating,
thus, the geometry of the double bond.18 These signals
o
were missing in compound 6a (CDCl3, 20 C), but ap-
peared at d 8.0 and 10.05 in DMSO-d6 at 20 oC and coa-
lesced at 50 oC in the same solvent. The configuration of
the C-1 asymmetric centre (ribose numbering) in both 3a1
and 3a2 was unequivocally assigned by NOE experiments,
which have shown no significant signal enhancement be-
tween 1-H and 2-H, whereas NOE was observed between
1-H and one methyl group of the acetonide. Also, the val-
ue J1,2=0 is characteristic for the trans arrangement of the
substituents at C-1 and C-2 positions.8a The diketones 3c2
and 6c2 exist exclusively as enols, the enolic proton ap-
pearing at d ~15.5.
Scheme 2 could account for the hydrogenation-ring clo-
sure reaction. The N-O bond scission triggered a sequence
of reactions with intermediate formation of enaminone I
(as hemiacetal or aldehyde), which easily underwent in-
tramolecular aldol-like condensations to II with subse-
quent hydrogenation of the double bond formed to give
the final products III. It is apparent that the reaction con-
ditions, the nature of substituents and the size of the ring
expected to be formed could affect the reaction pathway.
Thus in the case of hydrogenation of adduct 2a (Scheme
1), the condensation stops after the first step and it is not
followed by elimination, possibly due to steric reasons.
When R=Me, the enaminone I is not stable enough and is
partially hydrolysed to a 1,3-diketone, which finally af-
fords the diketones 3c2 and 6c2.
(11) Curran, D. P. Adv. Cycloaddition 1988, 1, 129.
(12) Hill, J. M.; Hutchinson, E. J.; Le Grand, D. M.; Roberts, S. M.,
Thorpe, A. J.; Turner, N. J. J. Chem. Soc., Perkin Trans. 1
1994, 1483.
(13) Suspension in water, purchased from Fluka.
Typical experimental procedure for the synthesis of 2a: A
solution of 1 (186 mg, 1.0 mmol) and 2,6-dichlorobenzonitrile
oxide (207 mg, 1.1 mmol) in CH2Cl2 (5 mL) was refluxed for
4 h. The mixture was then concentrated by a rotary evaporator
and chromatographed (silica gel, hexane/ethyl acetate 30:1) to
25
give 246 mg (66%) of 2a, mp 62-64 ºC (from hexane), [a]D
+77.2 (c 0.17, MeOH). Typical experimental procedure for
the synthesis of 5b: A solution of p-tolualdehyde oxime (73
mg, 0.54 mmol), N-chlorosuccinimide (80 mg, 0.6 mmol) and
pyridine (1 drop) in CHCl3 (5 mL) was refluxed for 20 min.
The reaction mixture was then cooled to room temperature, a
solution of 4 (201 mg, 0.45 mmol) and Et3N (0.07 mL, 0.5
mmol) in CHCl3 (5 mL) was added dropwise and the mixture
was refluxed overnight. Evaporation of the solvent followed
Scheme 2
Synlett 1999, No. 8, 1289–1291 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
A Novel Carbocyclic Ring Closure of Enopyrano(furano)sides
1291
by column chromatography (silica gel, hexane/ethyl acetate
CDCl3) 3.30 (1H, d, J 18.4), 3.50 (3H, s), 3.69 (1H, dd, J 5.3
and 4.3), 3.88 (2H, m), 3.93 (1H, d, J 18.4), 4.65 (1H, d, J
12.0), 4.67 (1H, d, J 4.3), 4.85 (5H, m) and 7.35 (18H, m); dC
(75 MHz, CDCl3) 42.80, 57.08, 73.90, 75.42, 75.89, 78.67,
78.81, 80.30, 100.04, 111.58, 127.65, 127.70, 129.94, 128.06,
128.08, 128.18 (overlapping peaks), 128.34, 128.41, 128.47,
128.52, 131.16, 134.99, 137.84, 137.89, 138.45 and 156.07.
Compound 6a: dH (300 MHz, CDCl3) 2.18 (1H, dd, J 14.6 and
5.0), 2.22 (1H, dd, J 14.6 and 8.4), 3.68 (1H, ddd, J 8.4, 7.3
and 5.0), 3.83 (1H, dd as a t, J 7.3), 4.03 (1H, d, J 7.3), 4.48
(1H, d, J 12.0), 4.53 (1H, d, J 12.0), 4.73 (1H, d, J 11.0), 4.76
(1H, d, J 11.0), 4.78 (1H, d, J 11.5), 5.50 (1H, d, J 11.5) and
7.3 (18H, m); dC (75 MHz, CDCl3) 28.24, 71.65, 73.77, 74.21,
78.52, 84.12, 99.00, 127.41, 127.49, 127.57, 127.92, 128.07,
128.20, 128.26, 128.49, 130.82, 133.37, 134.05, 138.47,
138.52, 138.55, 155.32 and 195.99.
10:1) gave 256 mg (44%) of the two diastereoisomers in 9:1
ratio (by 1H-NMR). The major diastereoisomer 5b was
crystallised from hexane-ether, mp 178-180 ºC, [a]D25 -47.8
(c = 0.5, CHCl3). General experimental procedure for
reductive cleavage of 2 and 5: To a solution of 2 or 5 (0.2
mmol) in MeOH (6 mL) and CH3CO2H (1 mL), MgSO4 (200
mg) and Raney Ni (catalytic) were added and the mixture was
stirred under 1 atm H2 for 90 min.The solids were filtered off,
the solvent evaporated and the residue was chromatographed
on silica gel using hexane/ethyl acetate as the eluent to give
the corresponding products 3 and 6.
All new compounds gave spectral and analytical data
consistent with their assigned structures. Selected NMR data
(J values are given in Hz): Compound 2a: dH (300 MHz,
CDCl3) 1.36 (3H, s), 1.45 (3H, s), 3.20 (1H, d, J 18.7), 3.44
(3H, s), 3.78 (1H, d, J 18.7), 4.78 (1H, d, J 5.5), 4.87 (1H, d, J
5.5), 4.95 (1H, s) and 7.36 (3H, m); dC (75 MHz, CDCl3)
24.84, 26.17, 43.53, 54.67, 83.16, 84.25, 107.99, 112.91,
118.85, 127.94, 128.23, 131.03, 134.92 and 154.61.
(16) Boa, A. N.; Booth, S. E.; Dawkins, D. A.; Jenkins, P. R.;
Fawcett, J.; Russell, D. R. . J. Chem. Soc., Perkin Trans. 1
1993, 1277.
(17) (a) Mukaiyama, T.; Hoshino, T. J. Am. Chem. Soc. 1960, 82,
5339. (b) Hassner, A.; Dehaen, W. J. Org. Chem. 1990, 55,
5505.
Compound 3a1: dH (300 MHz, CDCl3) 1.36 (3H, s), 1.38 (3H,
s), 2.95 (3H, s), 3.90 (1H, s), 4.43 (1H, d, J 5.9), 4.71 (1H, d,
J 5.9), 5.76 (1H, s br, NH), 7.4 (3H, m) and 9.90 (1H, s br,
NH); dC (75 MHz, CDCl3) 25.81, 27.43, 55.90, 79.21, 81.27,
82.87, 103.19, 111.88, 127.96, 128.03, 130.97, 132.40,
132.70, 134.43, 159.33 and 198.50 (restricted rotation of the
aromatic ring). Compound 5a (major isomer): dH (300 MHz,
(18) Zhuo, J.-C. Magn. Res. Chem. 1998, 36, 565.
Article Identifier:
1437-2096,E;1999,0,08,1289,1291,ftx,en;G11899ST.pdf
Synlett 1999, No. 8, 1289–1291 ISSN 0936-5214 © Thieme Stuttgart · New York