P. Borrachero et al. / Tetrahedron Letters 45 (2004) 4835–4839
4837
O
HO
HO
9
OH
COOMe
4
O
(R = H)
ii
Bn
N
R
N
2
6
i
H
O
H
O
HN
H
O
iii
HN
HO
H
O
O
O
O
O
1'
O
O
OH
OH
1'
1
O
O
O
Me2C
Me2C
Me2C
O
O
O
OH
10 (98%)
CMe2
9 (96%)
CMe2
CMe2
6a
7 (R = Bn, 18%)
+ 8 (R = H, 62%)
Scheme 3. Reagents and conditions: (i) Mo(CO)6, MeCN/H2O, reflux; (ii) LiAlH4; (iii) 80% TFA.
(1,2:3,4-di-O-isopropylidene-a-
D
-galacto-pentopyranos-
5-yl)pyrrolidine (9) in 96% yield, after column chroma-
resolution mass spectrometry corroborated the loss of
both isopropylidene protecting groups.**
20
D
tography. Compound 9: ½aꢀ ꢁ 30 (c 1.0, CHCl3).– The
HRCIMS data were in agreement with its structure, also
corroborated by the absences of any carbonyl IR band
and 13C signal (NMR spectrum).
Some conformational features for 6a and its derivatives
(7–10) can be deduced from the spectral data. Thus, the
C(50)H/C(azol ring)H coupling constant takes the values
9.9, 3.4, 2.3, 6.5, and 10.0 Hz, respectively, for 6a, 7, 8, 9,
and 10; among them, first and last high values are
indicative of a preferential anti relationship in solution
between these protons (compounds 6a and 10), while for
compounds 7 and 8, the low values of J suggest that
these protons have a preferential gauche relationship in
solution, and the medium J value for 9 may indicate no
conformational preference in solution.
Compounds 7 and 8 should keep the (6R,8R) configu-
ration, coming from that of isoxazolidine C(3) and C(5)
atoms, respectively. No epimerization at C(3) of 6a is to
be expected, so that 7 and 8 must have the (6R) con-
figuration. However, the C(5) of 6a might have under-
gone epimerization. For compound 7, the C(6)H/C(8)H
and C(8)H/C(6)H contacts (1D NOESY experiments)
and the absence of HO/C(6)H and C(6)H/OH contacts
are in agreement with a 6,8-cis relationship, and there-
fore corroborate the expected (8R) configuration. The
overlapping between some signals in the 1H NMR
spectrum of 8 did not allow us to perform similar 1D
NOESY experiments, but it was possible for its reduc-
tion product 9, for which the C(3a)H/C(2)H, C(3a)H/
C(4)H, C(5b)H/C(2)H, and C(5b)H/C(4)H contacts
observed as well as the absence of both C(3b)H/C(2)H
and C(3b)H/C(4)H contacts indicate the 2,4-cis rela-
tionship again, so that the (2R,4R) configuration is
assigned for 9 and (6R,8R) for 8.
In view of the foregoing, fairly good results, we planned
to extend the 1,3-cycloaddtion reaction with the same
dipolarophile 5 to the nitrone 4, which has a primary
hydroxy group at the terminal carbon atom, and thus
should allow us to save a reduction step of the synthetic
route to polyhydroxy-perhydroazazulenes. A first assay
of its reaction with 5 at 50 °C led to a mixture of
products. Two diastereomeric 3-(penta-O-benzyl-
pentitol-1-yl)isoxazolidines (11a
D
-gluco-
and 11bàà) were
**
Selected spectral data for 10: IR (KBr) mmax 3396 (OH and NH) and
1094 cmꢁ1 (C–OH); 1H NMR (500 MHz, CD3OD) d 4.77 (d, 1H,
Deprotection of 9 with aqueous 80% trifluoroacetic acid
almost quantitatively yielded (2R,4R)-4-hydroxy-2-
J1 ;2 ¼ 6:0, H-10), 4.43 (dddd, 1H, J3a;4 ¼ 5:5, J3b;4 ¼ 4:5, J4;5a ¼ 7:5,
0
0
J4;5b ¼ 7:0, H-4), 3.77 (dd, 1H, J2;5 ¼ 10:0, H-50), 3.18 (dd, 1H,
0
(a-D-galacto-pentopyranos-5-yl)pyrrolidine (10, 98%),
after cation-exchange chromatography; it showed
J5a;5b ¼ 9:5, J4;5a ¼ 7:5, H-5a), 3.06 (dd, 1H, J4;5b ¼ 7:0, H-5b), 3.00
(ddd, 1H, J2;3a ¼ 8:5, J2;3b ¼ 2:5, H-2), 2.35 (ddd, 1H, J3a;3b ¼ 14:0, H-
3a), and 1.75 (ddd, 1H, H-3b); 13C NMR (75.4MHz, CD3OD) d 90.2
(C-10), 71.4 (C-4), 68.4 (C-50), 59.6 (C-2), 58.1 (C-5), and 39.2
(C-3). HRFABMS: m=z 258.0953 (calcd for C9H17NO6+Na: 258.0954).
Selected spectral data for 11a: IR (film) mmax 1751 cmꢁ1 (ester C@O);
1H NMR (500 MHz, CDCl3) d (locant numerals for the sugar
moiety are maintained, but primed, such as for the starting sugar
derivative) 4.38 (dd, 1H, J4a;5 ¼ 8:7, J4b;5 ¼ 7:6, H-5), 3.93 (dd, 1H,
20
D
in agreement with its hemiacetal structure, and high-
mutarotation: ½aꢀ ꢁ 7 to ꢁ 43:6 (24 h; c 0.3, MeOH),
J4 ;5 ꢂ J3 ;4 ꢂ 5:0, H-40), 3.73 (dd, 1H, J2 ;3 ꢂ J3 ;4 ꢂ 5:0, H-30), 3.72
0
0
0
0
0
0
0
0
(s, 3H, COOMe), 3.68 (overlapped signal, 1H, H-20), 3.39 (ddd, 1H,
–
0
Selected spectral data for 9: IR (KBr) mmax 3183 (OH and NH), and
1067 cmꢁ1 (C–OH); 1H NMR (300 MHz, CDCl3) d (locant numerals
for the sugar moiety are maintained, but primed, such as for the
J3;4a ¼ 3:8, J3;4b ¼ 7:7, J2 ;3 ꢂ 0, H-3), 2.78 (ddd, 1H, J4a;4b ¼ 12:7, H-
4a), and 2.46 (ddd, 1H, H-4b); 13C NMR (75.4 MHz, CDCl3) d 172.9
(COOMe), 78.5 (C-20), 76.7 (C-5), 67.2 (C-3), 52.1 (COOMe), and 33.5
(C-4). HRCIMS: m=z 822.3989 (calcd for C52H55NO8+H: 822.4006).
Selected spectral data for 11b: IR (film) mmax 1753 cmꢁ1 (ester C@O);
1H NMR (300 MHz, CDCl3) d 4.62 (overlapped signal, 1H, H-5),
starting sugar derivative) 5.51 (d, 1H, J1 ;2 ¼ 5:0, H-10), 4.27 (m,
0
0
àà
overlapped signal, H-4), 3.70 (dd, 1H, J2;5 ¼ 6:5, H-50), 3.34 (ddd,
0
1H, J2;3a ¼ 9:2, J2;3b ¼ 5:4, H-2), 2.95 (ddd, 1H, J5a;5b ¼ 11:3,
4
3.99 (dd, 1H, J3 ;4 ¼ 4:9, J4 ;5 ¼ 5:0, H-40), 3.89 (dd, 1H, J5 ;6 a ¼ 3:7,
0
0
0
0
0
0
J4;5a ¼ 1:8, J3b;5a ¼ 1:8, H-5a), 2.82 (dd, 1H, J4;5b ¼ 4:1, H-5b),
J6 a;6 b ¼ 9:8, H-60a), 3.82 (dd, 1H, J2 ;3 ¼ 1:4, H-30), 3.77 (dd, 1H,
0
0
0
0
ꢃ2.5 (br s, 2H, NH and OH), 2.14 (ddd, 1H, J3a;3b ¼ 14:6, J3a;4 ¼ 5:8,
H-3a), 1.83 (dddd, J3b;4 ¼ 5:2, H-3b); NOE contacts (1D NOESY):
H-3a, H-2, H-4; H-2, H-3a, H-5b, H-50, H-40; H-5b, H-2, H-4; 13C
NMR (75.4 MHz, CDCl3) d 96.3 (C-10), 71.9 (C-4), 70.2 (C-50), 56.9
(C-2), 55.2 (C-5), 37.7 (C-3). HRCIMS: m=z 316.1757 (calcd for
J2 ;3 ¼ 3:7, H-20), 3.73 (s, 3H, COOMe), 3.12 (ddd, 1H, J3;4a ¼ 5:0,
0
J3;4b ¼ 8:6, H-3), 2.85 (ddd, 1H, J4a;4b ¼ 13:0, J4a;5 ¼ 3:7, H-4a), and
2.61 (ddd, 1H, J4b;5 ¼ 9:9, H-4b); 13C NMR (75.4 MHz, CDCl3) d
77.7 (C-20), 74.8 (C-5), 67.7 (C-3), 51.8 (COOMe), and 33.6 (C-4).
HRCIMS: m=z 822.3995 (calcd for C52H55NO8+H: 822.4006).
C
15H25NO6+H: 316.1760).