Preparation of 3,6-di-O-benzylallosamizoline from natural allosamidin

Article abstract of DOI:10.1016/S0957-4166(00)00156-7

Preparation of 3,6-di-O-benzylallosamizoline, the aglycon required for the total synthesis of allosamidin and its related compounds, from natural allosamizoline obtained from natural allosamidin is described. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0957-4166(00)00156-7

Tetrahedron: Asymmetry 11 (2000) 2083±2091
Preparation of 3,6-di-O-benzylallosamizoline from
natural allosamidin
Takanori Wakabayashi, Hironao Saito and Masao Shiozaki*
Exploratory Chemistry Research Laboratories, Sankyo Co. Ltd, Hiromachi 1-2-58, Shinagawa-ku, Tokyo 140-8710, Japan
Received 23 February 2000; accepted 17 April 2000
Abstract
Preparation of 3,6-di-O-benzylallosamizoline, the aglycon required for the total synthesis of allosamidin
and its related compounds, from natural allosamizoline obtained from natural allosamidin is described.
# 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction
Allosamidin, an inhibitor of enzymes that hydrolyze chitin to a monomer, was isolated from
the mycelia of Streptomyces sp. 1713 and characterized as 1 by Suzuki et al.1 as shown in Fig. 1.
Chitin is a major structural constituent of the exo-skeleton of arthropods and fungal cell walls. It
has been thought that this chitinase inhibitor 1 and its related analogues may control the insect's
growth by larval ecdysis and may function as a fungicide. It has recently become important to
discover the potential of allosamidin analogues as insecticides and/or fungicides.
Figure 1. Allosamidin 1
Synthetic approaches to allosamidin 1 and its analogues starting from many types of readily
available substances have been investigated and accomplished by many chemists.2 However, it
has been dicult to obtain a large amount of natural allosamidin 1 and allosamizoline 2 (the
aglycon part of 1) from the fermentation broth of Streptomyces sp. 1713. Recently, Suzuki and
Isogai reported in detail the production of natural allosamidin 1 from the fermentation broth of
Streptomyces sp. SANK62785, and the isolation of 2 from the broth by aqueous HCl hydrolysis
* Corresponding author. E-mail: shioza@shina.sankyo.co.jp
0957-4166/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(00)00156-7

2084
T. Wakabayashi et al. / Tetrahedron: Asymmetry 11 (2000) 2083±2091
in good yield.3 Therefore, it is now possible to obtain a fairly large amount of 2 from this fermentation
broth. Consequently, we investigated the availability of 2 as a useful starting material for the
synthesis of many types of allosamidin-related compounds. Here, we describe the conversion of 2
to an important intermediate, 3,6-di-O-benzylallosamizoline 5, which has been used for the total
synthesis of allosamidin and its related compounds.2d,i,l
2. Results and discussion
The starting material, natural allosamizoline (unpuri®ed-2), obtained from the fermentation
broth of Streptomyces sp. SANK62785 by the reported method,3 is contaminated with an un-
identi®ed impurity. It was puri®ed by acetylation, and successive deacetylation of the resulting
triacetate 3 gave pure allosamizoline (puri®ed-2) in two steps in 61% yield (Scheme 1). We used
both unpuri®ed-2 and puri®ed-2 in this investigation.
Scheme 1. Reagents and conditions: (a) Ac2O±pyridine, DMAP, rt, 15 h, 69%; (b) cat. 28% NaOMe in MeOH, rt, 1 h 87%
Firstly, we attempted the benzylation of unpuri®ed-2 with 1.5 equivalents of powdered KOH in
dimethylformamide (DMF) and crown ether (18-crown-6), and then excess benzyl chloride as
shown in Scheme 2. The major product was 6-O-benzylallosamizoline 4, in which the primary
alcohol of unpuri®ed-2 was benzylated, accompanying many other minor benzylated products,
that is, 3- and 4-O-monobenzylated, 3,6-, 4,6- and 3,4-di-O-benzylated, and 3,4,6-tri-O-benzylated
compounds. Separation of these compounds by chromatography was dicult. Conversion of 4 to
the title compound 5 in 35% yield has already been reported in the literature.2s Even though this
route is very short, the results indicate that it is not practical in terms of separation and total
yield.
Scheme 2. Reagents and conditions: (a) 1.5 equiv. KOH, 18-crown-6, DMF, rt, 10 min, then 20 equiv. BnCl, rt, 16 h,
24%; (b) for a reported method, see Ref. 2s, 35%
Secondly, as shown in Scheme 3, the amino-oxazoline ring of unpuri®ed-2 was cleaved by treat-
ment with hydrazine according to the reported method,2m and acetylated with pyridine±acetic

T. Wakabayashi et al. / Tetrahedron: Asymmetry 11 (2000) 2083±2091
2085
anhydride to obtain pure 6 in 46% yield. Hydrolysis of 6 with aqueous hydrochloric acid gave
aminocyclitol hydrochloride 7. Treatment of 7 with 1,10-thiocarbonyldiimidazole in THF±H2O
gave cyclic thiocarbamate 8. Benzylation of 8 using 1 equivalent of benzyl bromide and sodium
hydride at ^60±24ꢀC in DMF gave S-benzylated oxazoline 9 in 72% yield. The benzylation of 8
using 3 equivalents of benzyl bromide and sodium hydride at room temperature in DMF was
attempted. Although the 6-O-benzylated compound was obtained as a major product in this
reaction, many other benzylated compounds were also obtained. Therefore, we needed another
route for regioselective protection of 3- and 6-hydroxy groups.
Scheme 3. Reagents and conditions: (a) (1) NH2NH2, EtOH, 80ꢀC, 6 h, for a reported method, see Ref. 2m; (2) Ac2O±
pyridine, rt, 15 h, two steps, 46%; (b) 2 M HCl, 80ꢀC, 1 h, quantitatively; (c) Im2CˆS, THF±H2O, 0ꢀC, 1 h, then rt, 1
h, 60%; (d) BnBr (1 equiv.), NaH, DMF, ^60ꢀC±rt, 72%
Thirdly, as shown in Scheme 4, protection of 4- and 6-hydroxy groups of 8 with 1,3-dichloro-
1,1,3,3-tetraisopropyldisiloxane (TIPDSCl2) and 4-(dimethylamino)pyridine (DMAP) yielded
cyclic 4,6-di-O-siloxane 10. Conversion of thiocarbamate 10 to dimethylamino-oxazoline 11 with
Scheme 4. Reagents and conditions: (a) TIPDSCl2, DMAP, pyridine, 0ꢀC, 30 min, then rt, 1 h, 84%; (b) 2 M Me2NH in
MeOH, re¯ux, 30 h, 65%; (c) BnBr, Ag2O, toluene, re¯ux, 87%; (d) Bu4NF, THF, rt, 20 min, 93%; (e) Bu2SnO, toluene,
re¯ux, 4 h, then BnBr, CsF, DMF, rt, 16 h, 52% (recovered 13, 32%)

2086
T. Wakabayashi et al. / Tetrahedron: Asymmetry 11 (2000) 2083±2091
dimethylamine was accomplished according to the reported method.2n Benzylation of the 3-hydroxy
group of 11 with benzyl bromide and silver(I) oxide a€orded 12. Desilylation of 12 with tetra-
butylammonium ¯uoride gave 13. Treatment of 13 with dibutyltin oxide, and then benzyl bromide
and cesium ¯uoride a€orded the title compound 5 (Scheme 4). Although this route needs fairly
long steps, the dimethylamino-oxazoline moiety of 2 can also be changed to other amino-oxazoline
groups for the synthesis of the allosamidin analogues.2m
Finally, as shown in Scheme 5, direct conversion of both unpuri®ed-2 and puri®ed-2 to com-
pound 11 was attempted. Treatment of both unpuri®ed-2 and puri®ed-2 with TIPDSCl2 and
DMAP in pyridine provided cyclic 4,6-di-O-siloxane 11 in 38 and 80% yield, respectively. Judging
from the short steps (four steps) to 5 from 2 and the total yield (16% from unpuri®ed-2 and 34%
from puri®ed-2), this route should be practical for the synthesis of title compound 5 from natural
allosamizoline.
Scheme 5. Reagents and conditions: (a) TIPDSCl2, DMAP, pyridine, rt, 3 h, 38% from unpuri®ed-2, and 80% from
puri®ed-2, respectively
3. Conclusion
The puri®ed allosamizoline obtained by the acid hydrolysis of natural allosamidin was converted
to 3,6-di-O-benzylallosamizoline 5 in four steps in 34% overall yield.
4. Experimental
4.1. General
Melting points were determined on a Yanagimoto micro melting point apparatus and were
uncorrected. Optical rotations were obtained by the use of a JASCO P-1030 polarimeter. 1H and
13C NMRs were recorded with a JEOL-GSX 400 spectrometer using TMS as the internal standard.
IR absorption spectra were determined with an IR A-2 spectrophotometer, and mass spectra
were obtained with a JMS-700 mass spectrometer. Elemental analyses were performed by the
Institute of Science and Technology, Inc. Separation of the compounds by column chromatog-
raphy was done with silica gel 60 (230±400 mesh ASTM, Merck) under a slightly elevated
pressure (1.2±1.5 atm) for easy elution, and the quantity of silica gel used was 50±100 times the
weight charged on the column. THF was distilled under the existence of radical anion generated
by benzophenone and sodium. Dichloromethane was dried by being passed through an ICN
Alumina B-Super I. DMF and pyridine were dried by storage over 4 A molecular sieves.

T. Wakabayashi et al. / Tetrahedron: Asymmetry 11 (2000) 2083±2091
2087
4.2. 3,4,6-Tri-O-acetylallosamizoline 34
To a suspension of allosamizoline (unpuri®ed-2, 217 mg, 1 mmol, obtained from natural allosamidin
according to the reported method)3 in pyridine (5 ml) and Ac2O (1 ml) was added DMAP (10 mg,
0.082 mmol). This mixture was stirred for 15 h at room temperature, quenched with water, and
extracted with EtOAc. The organic layer was washed with aqueous sat. NaHCO3 and then brine,
dried over Na2SO4, ®ltered, and evaporated in vacuo to give a residue, which was chromato-
graphed on a silica gel column. Elution with CHCl3:MeOH (98:2, then 97:3) gave 3 (239 mg,
1
69%) as an amorphous residue. IR ꢀmax (®lm) 1745, 1660 cm^1; 270 MHz H NMR (CDCl3) ꢁ
2.02 (3H, s), 2.05 (3H, s), 2.10 (3H, m), 2.56 (1H, m), 2.91 (6H, s), 4.16 (1H, dd, J=6.3, 11.5 Hz),
4.21 (1H, dd, J=5.4, 11.5 Hz), 4.37 (1H, dd, J=3.8, 8.6 Hz), 4.76 (1H, dd, J=5.1, 8.6 Hz), 5.04
(1H, dd, J=5.8, 7.8 Hz), 5.14 (1H, dd, J=3.8, 5.8 Hz). FAB MS (positive) m/z 343 [M+H]+. High
resolution FAB MS (positive) calcd for C15H23N2O7: 343.1505; found: 343.1523.
4.3. Allosamizoline (puri®ed-2)
To a solution of 3 (242 mg, 1 mmol) in MeOH (20 ml) was added 0.1 ml of 28% NaOMe
solution in MeOH. The solution was stirred at room temperature for 1 h and concentrated in
vacuo to give a residue, which was chromatographed on a silica gel column. Elution with
MeOH:CHCl3:sat. 28% aqueous NH3 (25:75:1) gave puri®ed-2 (132 mg, 87%) as a white powder.
‰ꢂŠ2D4=+6.3 (c=0.8, H2O), and its HCl salt; ‰ꢂŠD24=^23.3 (c=0.8, H2O) (lit. HCl salt: ‰ꢂŠ2D2=^22.2
(c ˆ 0:5, H2O),3 [ꢂ]D=^22.9 (c=1.0, H2O),2b and ‰ꢂŠ2D3=^21.7 (c=0.9, H2O)2h); 400 MHz
1H NMR (CD3OD) ꢁ 2.01 (1H, m, C5-H), 2.89 (6H, s, NMe2), 3.61±3.70 (3H, m), 3.78 (1H, dd,
J=4.2, 11.2 Hz, C6-H), 3.98 (1H, dd, J=4.8, 8.8 Hz, C2-H), 4.80 (1H, dd, J=5.5, 8.8 Hz, C1-H).
FAB MS (positive) m/z 217 [M+H]+. Anal. calcd for C9H16N2O4: C, 49.99; H, 7.46; N, 12.96.
Found: C, 49.68; H, 7.24; N, 13.24.
4.4. 6-O-Benzylallosamizoline 4
A solution of allosamizoline (unpuri®ed-2, 30 mg, 0.139 mmol, obtained from natural allosamidin
according to the reported method3), powdered KOH (11.7 mg, 0.209 mmol, 1.50 equiv.) and 18-
crown-6 (1.50 mg, 0.04 equiv.) in DMF (0.90 ml) was stirred for 10 min at 24ꢀC. To this solution
was added benzyl chloride (0.32 ml, 2.78 mmol), and the mixture was stirred for 17 h at room
temperature. The reaction mixture was concentrated in vacuo to remove DMF, diluted with
EtOAc, washed with brine, dried over MgSO4, ®ltered, and concentrated in vacuo to give a residue,
which was chromatographed on a silica gel (3 g) column. Elution with EtOAc:MeOH (1:1) gave a
mixture of the product and a small amount of silica gel. This mixture was dissolved in EtOAc to
remove the powdered silica gel. Filtration of this suspension and concentration gave 4 (9.8 mg,
ꢀ
ꢀ
1
24%) as a solid; mp 146±147 C (recrystallized from EtOAc) (lit. mp 143±144 C2s); 270 MHz H
NMR (CDCl3) ꢁ 2.23 (1H, m, C5-H), 2.88 (6H, s, NMe2), 3.69 (1H, dd, J=5.2, 9.5 Hz, C6-H),
3.72 (1H, dd, J=5.2, 9.5 Hz, C6-H), 3.81 (1H, t, J=7.1 Hz, C4-H), 3.87 (1H, dd, J=4.8, 7.1 Hz,
C3-H), 4.08 (1H, dd, J=4.8, 9.0 Hz, C2-H), 4.56, 4.58 (2H, AB-q, J=12.4 Hz, CH2Ph), 4.69
(1H, dd, J=5.8, 9.0 Hz, C1-H), 7.27±7.37 (5H, m, C6H5). EI MS m/z 306 (M+), 289, 245, 215,
190, 185, 179, 167, 158, 139. High resolution EI MS m/z, calcd for C16H22N2O4: 306.1579; found:
306.1576.

2088
T. Wakabayashi et al. / Tetrahedron: Asymmetry 11 (2000) 2083±2091
4.5. [1R-(1ꢂ,2ꢃ,3ꢂ,4ꢃ,5ꢃ)]-5-Acetamido-3-(acetoxymethyl)-1,2,4-cyclopentanetriol 1,2,4-triacetate 6
To a suspension of allosamizoline (unpuri®ed-2, 205 mg, 0.95 mmol, obtained from natural
allosamidin according to the reported method3) in EtOH (99.5%, 5 ml) was added hydrazine
monohydrate (2.5 ml). After the solution was warmed at 80ꢀC for 6 h, the reaction mixture was
concentrated in vacuo, and pyridine (20 ml) and Ac2O (10 ml) were added. The mixture was
stirred for 15 h at room temperature, and diluted with EtOAc. The solution was washed with sat.
NaHCO3 aq. and then brine, dried over MgSO4, ®ltered and concentrated in vacuo to give a
residue, which was chromatographed on a silica gel column. Elution with benzene:EtOAc (5:1)
gave 6 (164 mg, 46%) as an oil; 270 MHz 1H NMR (CDCl3) ꢁ 1.97 (3H, s), 2.06 (3H, s), 2.09 (3H,
s), 2.13 (6H, s), 2.23±2.32 (1H, m), 4.23 (1H, dd, J=5.3, 11.2 Hz), 4.38 (1H, dd, J=4.6, 11.2 Hz),
4.54 (1H, ddd, J=5.9, 8.6, 11.2 Hz), 4.95 (1H, d, J=6.6 Hz), 5.10 (1H, dd, J=5.3, 7.3 Hz), 5.45
(1H, dd, J=7.3, 11.2 Hz), 5.90 (1H, d, J=8.6 Hz). EI MS m/z 373 (M+).
4.6. [1R-(1ꢂ,2ꢃ,3ꢂ,4ꢃ,5ꢃ)]-5-Amino-3-(hydroxymethyl)-1,2,4-cyclopentanetriol 7
A suspension of 6 (127 mg, 0.340 mmol) in aqueous 2 M HCl (22 ml) was stirred for 1 h at
80ꢀC. The mixture became a solution, which was concentrated and dried in vacuo to give an
amine hydrochloride 7 (68 mg, quantitatively). This compound was employed for the next reaction
1
without puri®cation: 270 MHz H NMR (D2O, internal standard=DSS) ꢁ 1.96 (1H, m), 3.34
(1H, dd, J=6.9, 9.7 Hz), 3.63±3.73 (2H, m), 3.78 (1H, dd, J=4.5, 11.6 Hz), 4.10±4.16 (2H, m).
4.7. [3aR-(3aꢂ,4ꢂ,5ꢃ,6ꢂ,6aꢂ)]-4,5-Dihydroxy-hexahydro-6-(hydroxymethyl)-2H-cyclopentoxazol-
2-thione 8
The above obtained hydrochloride 7 (204 mg, 1.020 mmol) was dissolved in H2O (1 ml) and
THF (4 ml). Then 1,10-thiocarbonyldiimidazole (276 mg, 1.546 mmol) was added to this solution
at ice-cooling temperature. After stirring for 1 h at 0ꢀC, the mixture was stirred further for 1 h at
24ꢀC, and concentrated in vacuo to give a mixture, which was chromatographed on a silica gel
column. Elution with EtOAc:MeOH:AcOH (92:8:1) gave triol 8 (125 mg, 60%). IR ꢀmax (KBr)
1
3465, 3240, 1557, 1532, 1158, 1053, 1003 cm^1. 400 MHz H NMR (D2O) ꢁ 2.13 (1H, m), 3.51
(1H, dd, J=7.1, 11.4 Hz), 3.57 (1H, dd, J=7.1, 8.8 Hz), 3.68 (1H, dd, J=4.5, 11.4 Hz), 3.83 (1H,
dd, J=4.9, 7.1 Hz), 4.00 (1H, dd, J=4.9, 9.5 Hz), 5.02 (1H, dd, J=5.4, 9.5 Hz); 100 MHz 13C
NMR (D2O) ꢁ 51.8, 59.9, 64.5, 74.9, 81.9, 86.3, 188.2 (C=S). FAB MS (positive) m/z 206
[M+H]+. High resolution FAB MS (positive), calcd for C7H12NO4S: 206.0487; found: 206.0490.
4.8. [3aR-(3aꢂ,4ꢂ,5ꢃ,6ꢂ,6aꢂ)]-2-(Benzylthio)-3a,5,6,6a-tetrahydro-6-(hydroxymethyl)-4H-cyclo-
pentoxazole-4,5-diol 9
To a suspension of powdered NaH (2.4 mg, 0.1 mmol) in DMF (1 ml) was gradually added a
solution of 8 (11.0 mg, 0.054 mmol) in DMF (0.5 ml) at ^64ꢀC. After stirring for 10 min, a solution
of benzyl bromide (10.9 mg, 0.064 mmol) in DMF (0.5 ml) was added, and the temperature was
slowly elevated to 24ꢀC. The reaction mixture was quenched with AcOH, and evaporated in
vacuo at 50ꢀC (bath temperature) to give a residue, which was chromatographed on a silica gel
column. Elution with MeOH:CH2Cl2 (9:91) gave 9 (11.4 mg, 72%). IR ꢀmax (CHCl3) 1592 cm^1;
1
400 MHz H NMR (CDCl3) ꢁ 2.01 (1H, m), 3.74±3.88 (4H, m), 4.16 (1H, m), 4.19 (2H, s), 4.83

T. Wakabayashi et al. / Tetrahedron: Asymmetry 11 (2000) 2083±2091
2089
(1H, dd, J=6.3, 9.2 Hz), 7.21±7.31 (5H, m). 100 MHz 13C NMR (CDCl3) ꢁ 36.0 (SCBn), 51.6,
59.7, 74.5, 74.6, 82.3, 83.4, 127.8, 128.7 (2ÂC), 129.0 (2ÂC), 135.9, 166.5 (NˆC±S). FAB MS
(positive) m/z 296 [M+H]+. High resolution FAB MS (positive), calcd for C14H18NO4S: 296.0957;
found: 296.0948.
4.9. [3aR-(3aꢂ,4ꢂ,5ꢃ,6ꢂ,6aꢂ)]-4,5-Dihydroxy-6-hydroxymethyl-hexahydro-2H-cyclopentoxazol-
2-thione 5-O-6-(O-methyl)-1,1,3,3-tetraisopropyldisiloxane 10
To a solution of 8 (6.8 mg, 0.033 mmol) and DMAP (1.4 mg, 0.012 mmol) in pyridine (0.5 ml)
was added TIPDSCl2 (0.05 ml, 0.165 mmol) at 0ꢀC under nitrogen with stirring. After 30 min,
this mixture was stirred for 1 h at 24ꢀC, and concentrated in vacuo to give a residue, which was
chromatographed on a silica gel column. Elution with EtOAc:hexane (1:3) gave 10 (12.5 mg,
24
1
84%) as a powder. ‰ꢂŠD =^51.4 (c=1.5, CHCl3); 400 MHz H NMR (CDCl3) ꢁ 1.00±1.06 (28H,
m), 2.10 (1H, m), 2.83 (1H, bs, OH), 3.82 (1H, dd, J=8.8, 11.4 Hz), 3.93±4.12 (4H, m), 5.11 (1H,
dd, J=4.8, 9.5 Hz), 7.92 (1H, broad, NH); 100 MHz 13C NMR (CDCl3) ꢁ 12.6, 12.8, 13.2, 13.6,
17.1 (2ÂC), 17.2, 17.3 (2ÂC), 17.4 (3ÂC), 50.9, 56.9, 62.3, 71.8, 81.6, 81.9, 189.0. FAB MS
(positive) m/z 448 [M+H]+. High resolution FAB MS (positive), calcd for C19H38NO5SSi2:
448.2009; found: 448.2000.
4.10. 4,6-Di-O-1,1,3,3-tetraisopropyldisiloxane-allosamizoline 11
A solution of 10 (8.9 mg, 0.020 mmol) in 2 M Me2NH in MeOH (2 ml) was re¯uxed for 30 h
under nitrogen. This solution was concentrated in vacuo to give a residue, which was chromato-
graphed on a silica gel column. Elution with MeOH:CHCl3:28% aqueous NH3 (5:95:0, and then
5:95:1) gave 11 (5.9 mg, 65%) as a powder. ‰ꢂŠ2D4=^25.0 (c=0.7, CHCl3); IR ꢀmax (KBr) 3172,
2946, 2868, 1647, 1466, 1410, 1035 cm^1; 400 MHz 1H NMR (CDCl3) ꢁ 1.00±1.10 (28H, m), 1.80
(1H, m), 2.89 (6H, s), 3.79 (1H, dd, J=8.8, 11.4 Hz), 3.90±4.11 (4H, m), 4.73 (1H, t, J=8.8 Hz);
100 MHz 13C NMR (CDCl3) ꢁ 12.7, 13.0, 13.3, 13.5, 17.1 (2ÂC), 17.2 (2ÂC), 17.3 (2ÂC), 17.5
(2ÂC), 37.7 (2ÂC), 51.5, 57.9, 71.8, 72.5, 78.5, 83.1, 161.8. FAB MS (positive) m/z 459 [M+H]+.
High resolution FAB MS (positive), calcd for C21H43N2O5Si2: 459.2711; found: 459.2713.
4.11. Compound 11 from unpuri®ed-2
(a) To a solution of unpuri®ed-2 (21.7 mg, 0.100 mmol) and 4-dimethylaminopyridine (2 mg) in
pyridine (1 ml) was added 1,3-dichloro-1,1,3,3-tetraisopropyldisiloxane (0.10 ml, 0.313 mmol) at
0ꢀC under nitrogen with stirring. After stirring for 30 min at 0ꢀC and for 3 h at 24ꢀC, the reaction
mixture was concentrated in vacuo to give a mixture, which was chromatographed on a silica gel
column. Elution with MeOH:CHCl3:28% aqueous NH3 (5:95:0, and then 5:95:1) gave 11 (17.5
mg, 38%) as a powder, which was identical with that obtained from 10 in all respects. (b) The
same treatment of puri®ed-2 as described in (a) above gave 11 in 80% yield.
4.12. 3-O-Benzyl-4,6-di-O-1,1,3,3-tetraisopropyldisiloxane-allosamizoline 12
A mixture of 11 (7.8 mg, 0.017 mmol), benzyl bromide (29 mg, 0.170 mmol) and Ag2O (19.8 mg,
0.085 mmol) in toluene (2 ml) was re¯uxed for 4 h, and the reaction mixture was ®ltered through
Celite. The ®ltrate was concentrated in vacuo to give a residue, which was chromatographed on a

2090
T. Wakabayashi et al. / Tetrahedron: Asymmetry 11 (2000) 2083±2091
silica gel column. Elution with CHCl3:MeOH (49:1) gave 12 (8.1 mg, 87%) as a paste. IR ꢀmax
(®lm) 2946, 2929, 2896, 2868, 1656 cm^1; 400 MHz 1H NMR (CDCl3) ꢁ 0.99±1.11 (28H, m), 1.84
(1H, m), 2.90 (6H, s), 3.74 (1H, dd, J=4.8, 8.4 Hz), 3.91±4.00 (3H, m), 4.11 (1H, dd, J=4.8, 9.2
Hz), 4.71 (1H, t, J=8.8 Hz), 4.73, 4.89 (2H, AB-q, J=12.1 Hz), 7.23±7.34 (3H, m), 7.41±7.43
(2H, m); 100 MHz 13C NMR (CDCl3) ꢁ 12.6, 13.0, 13.3, 13.5, 17.1 (3ÂC), 17.2 (2ÂC), 17.4
(2ÂC), 17.5, 37.7 (2ÂC), 51.5, 57.9, 71.2, 71.6, 71.8 (OCH2Ph), 78.4, 90.7, 127.1, 127.6 (2ÂC),
128.0 (2xC), 138.9 (penyl-C1), 161.7. FAB MS (positive) m/z 549 [M+H]+. High resolution FAB
MS (positive), calcd for C28H49N2O5Si2: 549.3180; found: 549.3172.
4.13. 3-O-Benzylallosamizoline 13
To a solution of 12 (27.1 mg, 0.049 mmol) in THF (1 ml) was added 0.25 ml of 1.0 M Bu4NF
solution in THF at 24ꢀC with stirring. After 20 min, the mixture was concentrated in vacuo to give
a residue, which was chromatographed on a silica gel column. Elution with CHCl3:MeOH:28%
aqueous NH3 (95:5:1, then 92:8:1) gave 13 (14.0 mg, 93%) as an amorphous residue. ‰ꢂŠ2D4=^9.4
(c=1.0, CHCl3). IR ꢀmax (CHCl3) 3610, 3435, 2933, 1700, 1649 cm^1; 400 MHz 1H NMR (CDCl3)
ꢁ 2.21 (1H, m), 2.90 (6H, s), 3.76±3.92 (4H, m), 4.27 (1H, dd, J=3.7, 8.6 Hz), 4.65, 4.81 (2H, AB-q,
J=11.6 Hz), 4.72 (1H, dd, J=5.2, 8.6 Hz), 7.28±7.38 (5H, m); 100 MHz 13C NMR (CDCl3) ꢁ
37.7 (2ÂC), 53.2 (C5), 62.5 (C6), 72.0, 73.0 (OCH2Ph), 77.0, 82.7, 91.3, 127.7, 127.9 (2ÂC), 128.4
(2ÂC), 138.0 (pentyl-C1), 162.0. FAB MS (positive) m/z 307 [M+H]+. High resolution FAB MS
(positive), calcd for C16H23N2O4: 307.1658; found: 307.1678.
4.14. 3,6-Di-O-benzylallosamizoline 5
A solution of 13 (8.3 mg, 0.027 mmol) and Bu2SnˆO (10.1 mg, 0.041 mmol) in toluene (5 ml)
was re¯uxed for 4 h, and concentrated in vacuo to give a residue, which was dissolved in DMF (1 ml).
To this solution were added CsF (8.2 mg, 0.054 mmol) and BnBr (23.1 mg, 0.135 mmol), and this
mixture was stirred for 16 h at 24ꢀC, and ®ltered. The ®ltrate was concentrated in vacuo, and
chromatographed on a silica gel column. Elution with CHCl3:MeOH:28% aqueous NH3
(96:4:0.5, then 92:8:1) gave the recovered 13 (2.5 mg, 30%) and 5 (5.6 mg, 52%) as an amorphous
residue. ‰ꢂŠ2D4=^17.7 (c=0.2, CH2Cl2); [lit. [ꢂ]D=^19.6 (c=1.2, CH2Cl2)].2s IR nmax (CHCl3)
1
3608, 3090±2800, 1649 cm^1; 400 MHz H NMR (CDCl3) ꢁ 2.28 (1H, m), 2.91 (6H, s), 3.63 (1H,
dd, J=6.5, 9.3 Hz), 3.71 (1H, dd, J=5.9, 9.3 Hz), 3.78 (1H, dd, J=4.3, 6.5 Hz), 3.95 (1H, dd,
J=6.6, 8.6 Hz), 4.25 (1H, dd, J=4.3, 8.7 Hz), 4.55 (2H, s), 4.67, 4.81 (2H, AB-q, J=11.7 Hz),
4.68 (1H, dd, J=7.1, 7.4 Hz), 7.25±7.39 (10H, m); 100 MHz 13C NMR (CDCl3) ꢁ 37.7 (2ÂC),
51.2 (C5), 69.5, 71.9, 72.6, 73.3, 76.1, 82.3, 91.4, 127.5 (2ÂC), 127.6 (2ÂC), 127.7 (2ÂC), 127.9
(2ÂC), 128.3, 128.4, 138.0, 138.4, 161.9. FAB MS (positive) m/z 397 [M+H]+. High resolution
FAB MS (positive), calcd for C23H29N2O4: 397.2127; found: 397.2137.
References
1. (a) Sakuda, S.; Isogai, A.; Matsumoto, S.; Suzuki, A.; Koseki, K. Tetrahedron Lett. 1986, 27, 2475±2478. (b)
Sakuda, S.; Isogai, A.; Matsumoto, S.; Suzuki, A. J. Antibiotics 1987, 40, 296±300. (c) Koga, D.; Isogai, A.;
Sakuda, S.; Matsumoto, S.; Suzuki, A.; Kimura, S.; Ide, A. Agric. Biol. Chem. 1987, 51, 471±476. (d) Sakuda, S.;
Isogai, A.; Makita, T.; Matsumoto, S.; Koseki, K.; Kodama, H.; Suzuki, A. Agric. Biol. Chem. 1987, 51, 3251±
3259. (e) Sakuda, S.; Isogai, A.; Matsumoto, S.; Suzuki, A.; Koseki, K.; Kodama, H.; Yamada, Y. Agric. Biol.

T. Wakabayashi et al. / Tetrahedron: Asymmetry 11 (2000) 2083±2091
2091
Chem. 1988, 52, 1615±1617. (f) cf. Nishimoto, Y.; Sakuda, S.; Takayama, S.; Yamada, Y. J. Antibiotics 1991, 44,
716±722.
2. (a) Trost, B. M.; van Vranken, D. L. J. Am. Chem. Soc. 1990, 112, 1261±1263. (b) Takahashi, S.; Terayama, H.;
Kuzuhara, H. Tetrahedron Lett. 1991, 32, 5123±5126. (c) Nakata, M.; Akazawa, S.; Kitamura, S.; Tatsuta, K.
Tetrahedron Lett. 1991, 32, 5363±5366. (d) Grith, D. A.; Danishefsky, S. J. J. Am. Chem. Soc. 1991, 113, 5863±
5864. (e) Maloisel, J.-L.; Vasella, A.; Trost, B. M.; van Vranken, D. L. J. Chem. Soc., Chem. Commun. 1991, 1099±
1101. (f) Maloisel, J.-L.; Vasella, A. Helv. Chim. Acta 1992, 75, 1491±1514. (g) Maloisel, J.-L.; Vasella, A.; Trost,
B. M.; van Vranken, D. L. Helv. Chim. Acta 1992, 75, 1515±1526. (h) Simpkins, N. S.; Stokes, S.; Whittle, A. J.
J. Chem. Soc., Perkin Trans. 1 1992, 2471±2477. (i) Takahashi, S.; Terayama, H.; Kuzuhara, H. Tetrahedron Lett.
1992, 33, 7565±7568. (j) Trost, B. M.; van Vranken, D. L. J. Am. Chem. Soc. 1993, 115, 444±458. (k) Kitahara, T.;
Suzuki, N.; Koseki, K.; Mori, K. Biosci. Biotech. Biochem. 1993, 57, 1906±1909. (l) Terayama, H.; Kuzuhara, H.;
Takahashi, S.; Sakuda, S.; Yamada, Y. Biosci. Biotech. Biochem. 1993, 57, 2067±2069. (m) Takahashi, S.;
Terayama, H.; Kuzuhara, H. Tetrahedron Lett. 1994, 35, 4149±4152. (n) Goering, B. K.; Ganem, B. Tetrahedron
Lett. 1994, 35, 6997±7000. (o) Blattner, R.; Furneaux, R. H.; Kemmitt, T.; Tyler, P. C.; Ferrier, R. J.; Tiden, A.-K.
J. Chem. Soc., Perkin Trans. 1 1994, 3411±3421. (p) Takahashi, S.; Terayama, H.; Kuzuhara, H.; Sakuda, S.;
Yamada, Y. Biosci. Biotech. Biochem. 1994, 58, 2301±2302. (q) Blattner, R.; Furneaux, R. H.; Lynch, G. P.
Carbohydr. Res. 1996, 294, 29±40. (r) Shrader, W. D.; Imperiali, B. Tetrahedron Lett. 1996, 37, 599±602. (s)
Grith, D. A.; Danishefsky, S. J. J. Am. Chem. Soc. 1996, 118, 9526±9538. (t) Berecibar, A.; Grandjean, C.;
Siriwardena, A. Chem. Rev. 1999, 99, 779±844.
3. Suzuki, A.; Isogai, A. Japan Kokai Tokkyo Koho, JP 07002807, 1995, 75±82 (in Japanese).
4. The numbering of allosamizoline in the nomenclature is di€erent from that of 2H-cyclopentoxazole or 4H-
cyclopentoxazole.