The reaction of (4R,5R)- and (4S,5S)-4,5-epoxy-2(E)-hexenoates and secondary amines

Article abstract of DOI:10.1248/cpb.49.849

A reaction of methyl (4R,5R)-4,5-epoxy-2(E)-hexenoate 1 with N-benzylmethylamine gave a diastereomerically pure methyl (4R,5R)-4,5-epoxy-(3S)-N-benzylmethylamino hexanoate 6 and methyl (4S,5R)-4-N-benzylmethylamino-5-hydroxy-2(E)-hexenoate 7. The former was chemoenzymatically converted to (-)-osmundalactone 11, which is an aglycone of osmundalin. On the other hand, the directly conjugated addition of dimethylamine to methyl (4S,5S)-4,5-epoxy-2(E)-hexenoate 1 followed by treatment with MeOH at 40°C exclusively provided methyl (4R,5S)-4-dimethylamino-5-hydroxy-2(E)-hexenoate 16, which was converted into L-(-)-forosamine 18.

Full text of DOI:10.1248/cpb.49.849

July 2001
Chem. Pharm. Bull. 49(7) 849—853 (2001)
849
The Reaction of (4R,5R)- and (4S,5S)-4,5-Epoxy-2(E)-Hexenoates and
Secondary Amines
Chikako SAOTOME, Machiko ONO, and Hiroyuki AKITA*
School of Pharmaceutical Sciences, Toho University, 2–2–1 Miyama, Funabashi, Chiba 274–8510, Japan.
Received January 31, 2001; accepted March 26, 2001
A reaction of methyl (4R,5R)-4,5-epoxy-2(E)-hexenoate 1 with N-benzylmethylamine gave a diastereomeri-
cally pure methyl (4R,5R)-4,5-epoxy-(3S)-N-benzylmethylamino hexanoate 6 and methyl (4S,5R)-4-N-benzyl-
methylamino-5-hydroxy-2(E)-hexenoate 7. The former was chemoenzymatically converted to (؊)-osmundalac-
tone 11, which is an aglycone of osmundalin. On the other hand, the directly conjugated addition of dimethyl-
amine to methyl (4S,5S)-4,5-epoxy-2(E)-hexenoate 1 followed by treatment with MeOH at 40 °C exclusively pro-
vided methyl (4R,5S)-4-dimethylamino-5-hydroxy-2(E)-hexenoate 16, which was converted into L-(؊)-
forosamine 18.
Key words Michael addition; secondary amine; osmundalactone; forosamine
In the preceding paper, we reported the syntheses of each
optically pure stereoisomer of 4,5-epoxy-2(E)-hexenoates
(4R,5R)-1 and (4S,5S)-1 based on a chemoenzymatic method
from an achiral precursor, methyl sorbate.1) As a part of the
useful application of (4S,5S)-1 to the syntheses of amino
sugars or related compounds, formal total syntheses of D-
acosamine 2 and D-ristosamine 3 from (4S,5S)-1 were
achieved.2) The reaction of (4S,5S)-1 with 4 eq of benzy-
lamine gave the 1,4-addition products (3R,4S,5S)-4 (53%)
and (3S,4S,5S)-5 (32%). The intramolecular nucleophilic at-
tack by an ester carbonyl group upon the epoxy ring of the
substrates, (3R,4S,5S)-4 and (3S,4S,5S)-5 resulted in the for-
mal total syntheses of D-acosamine 2 and D-ristosamine 3, re-
spectively. In the course of our continuing interest in the re-
action of 4,5-epoxy-2(E)-hexenoates (4S,5S)-1 with amine,
the reaction of (4S,5S)-1 with a secondary amine aroused our
interest.
The reaction of (Ϯ)-1 with 2 eq of N-methylbenzylamine
at 40 °C for 4 d gave a diastereomerically pure (Ϯ)-6 (68%)
and (Ϯ)-7 (22%), while this reaction in the presence of
MeOH provided (Ϯ)-6 (30%) and (Ϯ)-7 (66%). In case of
the latter, a solvent effect appeared and the product ratio of
(Ϯ)-6 and (Ϯ)-7 was reversed. In order to determine the
structure of (Ϯ)-6, two possible authentic samples were pre-
pared. The reaction of (Ϯ)-3,4-syn 4 and (Ϯ)-3,4-anti 5 with
an excess of methyl iodide at 0 °C provided the N-methylated
amines (Ϯ)-3,4-syn 6 (48%) and (Ϯ)-3,4-anti 8 (33%), re-
spectively. Physical data of the present (Ϯ)-6 were identical
with those of authentic (Ϯ)-3,4-syn 6. For the purpose of de-
termining the structure of (Ϯ)-7, compound (Ϯ)-7 was con-
verted into the acetate (Ϯ)-9 (88%) and the acetal (Ϯ)-10
(47%) by applying the Evans method.3) A chemical shift due
to the C5-H of (Ϯ)-9 appeared in the lower field (d 5.21, dq,
Jϭ6, 9 Hz) in comparison with that (d 4.11, quintet, Jϭ6 Hz)
of (Ϯ)-7, thus the hydroxyl group was located at the C5-posi-
tion. The anti-stereochemistry of (Ϯ)-7 was confirmed by the
following experimental fact. Nuclear Overhauser effect
(NOE) experiments of (Ϯ)-10 were shown in Fig. 1, and the
coupling constants of the C3-axial and C4-axial protons, and
the C4-axial and C5-axial protons were 10 Hz and 10 Hz, re-
spectively, clearly indicating that the starting (Ϯ)-7 pos-
sessed 4,5-anti-configurations. Then, the formation of di-
astereomerically pure (Ϯ)-6 and 4,5-anti 7 was explained by
the following experiments.
When a solution of (Ϯ)-6 in MeOH was allowed to stand
at 40 °C for a long time (4 d), the rearranged (Ϯ)-7 was ob-
tained in 54% yield along with the starting (Ϯ)-6 (30%). The
same reaction was carried out for a short time (2 d), after
which (Ϯ)-7 (43%) and the intermediary (Ϯ)-1 (16%) were
obtained along with the starting (Ϯ)-6 (21%). A solution of
(Ϯ)-7 in MeOH was exposed at 40 °C for 4 d, no change of
(Ϯ)-7 was observed. On the other hand, exposure of the syn-
Chart 1
To whom correspondence should be addressed. e-mail: akita@phar.toho-u.ac.jp
© 2001 Pharmaceutical Society of Japan

850
Vol. 49, No. 7
Chart 2
Chart 3
thetic (Ϯ)-8 in MeOH at 40 °C for 4 d gave (Ϯ)-7 exclusively (Ϯ)-8 could be converted into (Ϯ)-1, and the partial conver-
in 83% yield. These experimental results could be explained sion of (Ϯ)-6 into (Ϯ)-1 probably occurs. Finally, the reac-
by the thermodynamic stability of (Ϯ)-6 and (Ϯ)-8. Based on tion of (Ϯ)-1 with N-methylbenzylamine gave 3,4-syn 6 cor-
inspection of the stability of (Ϯ)-6 and (Ϯ)-8 using Dreiding responding to the kinetically controlled product and 4,5-anti
stereomodels, (Ϯ)-8 was presumably more unstable than 7 corresponding to the thermodynamically controlled prod-
(Ϯ)-6. Because, in case of (Ϯ)-8, steric repulsion between uct. The difference in the product ratio of (Ϯ)-6 and (Ϯ)-7
the C5-methyl group and C3-substituent appeared to be larger between the absence of MeOH and the presence of MeOH is
than that of (Ϯ)-6. Consequently, the rate of conversion of explained by the solvent effect. Methanol presumably accel-
the unstable (Ϯ)-8 into (Ϯ)-1 is presumably faster than that erates the rate of the retro-Michael process to afford (Ϯ)-1,
of the conversion of (Ϯ)-6 into (Ϯ)-1. According to the then the thermodynamically favored (Ϯ)-7 could be obtained
above mentioned conversion experiments, the intermediary from (Ϯ)-1 as a major product.
product resulting from exposure of (Ϯ)-6 in MeOH for 2 d
This reaction was effectively applied to the synthesis of os-
and (Ϯ)-8 in MeOH for 4 d are the retro-Michael reaction mundalactone 11, which is an aglycone of osmundalin iso-
product, 4,5-epoxy-2(E)-hexenoate ((Ϯ)-1). From these ex- lated from Osmunda japonica THUNBERG (Akaboshi zenmai),
perimental results, the epoxy ester ((Ϯ)-1) is generated from which acts as a feeding inhibitor for the larvae of butterfly
(Ϯ)-6 or/and (Ϯ)-8, and the liberated N-methylbenzylamine Eurema hecabe mandrarina.4a,b) The reaction of (ϩ)-
again attacks at the C4-position of (Ϯ)-1 in the manner of (4R,5R)-1 with N-methylbenzylamine gave (ϩ)-(3S,4R,5R)-6
anti-stereochemistry to afford the (4,5)-anti 7 corresponding (65%, [a]D ϩ17.3° (cϭ0.98, CHCl3)) and (ϩ)-(4S,5R)-7
to the thermodynamically controlled product. On the whole, (25%, [a]D ϩ52.2° (cϭ1.01, CHCl3)). Treatment of (ϩ)-6
the reaction of (Ϯ)-1 with N-methylbenzylamine is explained with trifluoromethanesulfonic acid (CF3SO3H) exclusively
as follows: At first, a competitive attack by the nucleophile at provided d-lactone ((ϩ)-(3S,4R,5S)-12) (72% [a]D ϩ53.2°
the C3- and C4-positions of (Ϯ)-1 presumably occurs to af- (cϭ0.76, CHCl3)), which was then subjected to acetylation5)
ford (Ϯ)-6, (Ϯ)-8 and the 4,5-anti 7. Then, more unstable with Ac2O–AcOH (2 : 1) to afford an acetate (Ϫ)-(4R,5S)-13

July 2001
851
Chart 4
Chart 5
(82%, [a]D Ϫ169.8° (cϭ0.73, CHCl3)) with the elimination Hydrogenation of (Ϫ)-(4R,5S)-16, followed by treatment
of N-methylbenzylamine. In this case, the liberated N- with 80% AcOH, gave the d-lactone ((4R,5S)-17) which was
methylbenzylamine was acetylated in 98% yield and did not reduced with diisobutylaluminum hydride (DIBAH) to pro-
work as a nucleophilic reagent against the 2H-pyran-2-one vide an amino sugar L-(Ϫ)-(4R,5S)-18 (20% from (Ϫ)-16,
moiety of the generated substrate (Ϫ)-13. The physical data [a]D Ϫ85.7° (cϭ0.78, MeOH)) and a diol (ϩ)-(4R,5S)-19
([a]D and NMR) of (Ϫ)-13 were identical with those ([a]D (24% from (Ϫ)-16, [a]D ϩ10.7° (cϭ0.93, MeOH)). The
Ϫ172° (cϭ2.8, CHCl3) and NMR) of the reported (Ϫ)-13.4a) physical data of the synthesized L-(Ϫ)-18 were consistent
Finally, the acetate ((Ϫ)-13) was exposed to enzymatic hy- with those ([a]D ϩ86.1° (cϭ0.9, MeOH) and NMR) of D-
drolysis using the lipase “Amano P” from Pseudomonas sp. forosamine 18.6)
and converted to the hydroxy d-lactone ((Ϫ)-(4R,5S)-11)
In conclusion, in the reaction of 4,5-epoxy-2(E)-hexenoate
(92%, [a]D Ϫ69.0° (cϭ0.46, H2O)) whose physical data 1 with a secondary amine, the 1,4-conjugated addition of
([a]D and NMR) were identical with those ([a]D Ϫ70.6° secondary amine to the a,b-unsaturated ester moiety may
(cϭ2.0, H2O)) of the natural (Ϫ)-osmundalactone 11.4a)
occur to provide a diastereomeric mixture of (4,5)-epoxy-3-
Then, the reaction of (Ϫ)-(4S,5S)-1 with dimethylamine N-substituted amino esters. From this mixture, the product
was carried out. The reaction of (Ϫ)-(4S,5S)-1 with dimethyl- distribution between a 4,5-anti-2(E)-hexenoate such as 7 or
amine hydrochloride (2 eq)–triethylamine (2 eq) in MeOH 16 and an enantiomerically pure 4,5-epoxy-3-N-substituted
gave a 1.7 : 1 diastereomeric mixture of (3R,4S,5S)-14 and amino ester such as 6 was found to depend upon the reaction
(3S,4S,5S)-15 in 81% yield. This mixture was again exposed condition and the nature of the secondary amine used.
to MeOH at 40 °C for 2 d to provide the rearranged (Ϫ)-
Experimental
(4R,5S)-16 (61%, [a]D Ϫ115.9° (cϭ1.02, CHCl3)) and
All melting points were measured on a Yanaco MP-3S micro melting
(4S,5S)-1 (13%) along with the starting mixture (25% recov-
ery) of 14 and 15. This phenomenon was also understood by
the same explanation as the preparation of (Ϯ)-7 from (Ϯ)-1.
point apparatus and are uncorrected. 1H- and 13C-NMR spectra were
recorded on JEOL AL 400 spectrometer in CDCl3. Carbon substitution de-
grees were established by Distortionless Enhancement by Polarization

852
Vol. 49, No. 7
Transfer (DEPT) pulse sequence. The fast atom bombardment mass spectra was washed with saturated brine, dried over MgSO4 and evaporated. The
(FAB-MS) was obtained with a JEOL JMS-DX 303 spectrometer. IR spectra residue was chromatographed on silica gel (10 g, n-hexane : AcOEtϭ20 : 1)
were recorded on a JASCO FT/IR-300 spectrometer. Optical rotations were to give (Ϯ)-10 (0.145 g, 47%) as a homogeneous oil. (Ϯ)-10: IR (neat):
measured with a JASCO DIP-370 digital polarimeter. All evaporations were 1739 cmϪ1; NMR: d 1.47 (3H, d, Jϭ6 Hz), 2.32 (3H, s), 2.42 (1H, t, Jϭ10
performed under reduced pressure. For column chromatography, silica gel Hz), 2.57 (1H, dd, Jϭ8, 16 Hz), 3.09 (1H, dd, Jϭ4, 16 Hz), 3.69 (3H, s),
(Kieselgel 60) was employed.
3.79, 3.85 (each 1H, d, Jϭ14 Hz), 4.10 (1H, dq, Jϭ6, 10 Hz), 4.37 (1H, ddd,
Methyl 3b-Benzylamino-4b,5b-epoxyhexanoate (؎)-4 and Methyl 3a- Jϭ4, 8, 10 Hz), 5.54 (1H, s), 7.23—7.48 (10H, m). Anal. Found: C, 71.56;
Benzylamino-4b,5b-epoxyhexanoate (؎)-5 A mixture of (Ϯ)-1 (3.11 g, H, 7.77; N, 3.69. Calcd for C22H27NO4: C, 71.52; H, 7.37; N, 3.79%. FAB-
21.9 mmol) and BnNH2 (9.39 g, 88 mmol) was allowed to stand for 2 d at MS m/z: 370 (Mϩϩ1).
40 °C. The reaction mixture was directly chromatographed on silica gel (200
Treatment of (؎)-6 with MeOH i) A solution of (Ϯ)-6 (0.1 g, 0.4
g, n-hexane : AcOEtϭ2 : 1) to afford (Ϯ)-4 (2.95 g, 53%) as a colorless oil mmol) in MeOH (2 ml) was allowed to stand for 4 d at 40 °C. The reaction
and (Ϯ)-5 (1.91 g, 32%) as a colorless oil in eluate order. (Ϯ)-4: the spectral mixture was worked up in the same way as for the preparation of (Ϯ)-6 and
data (IR and NMR) of (Ϯ)-4 were identical with those of the reported (Ϯ)-7 from (Ϯ)-1 to afford (Ϯ)-6 (0.03 g, 30%) and (Ϯ)-7 (0.054 g, 54%).
(3R,4S,5S)-4: FAB-MS m/z: 250 (Mϩϩ1). (Ϯ)-5: the spectral data (IR and ii) A solution of (Ϯ)-6 (0.1 g, 0.4 mmol) in MeOH (2 ml) was allowed to
NMR) of (Ϯ)-5 were identical with those of the reported (3S,4S,5S)-5: FAB- stand for 2 d at 40 °C. The reaction mixture was worked up in the same way
MS m/z: 250 (Mϩϩ1).
Methyl 3b-N-Benzylmethylamino-4b,5b-epoxyhexanoate (؎)-6 and (0.044 g, 43%).
Methyl (4a,5b)-4-N-Benzylmethylamino-5-hydroxy-2(E)-hexenoate (؎)- Treatment of (؎)-8 with MeOH A solution of (Ϯ)-8 (0.14 g, 0.5
i) A mixture of (Ϯ)-1 (1.00 g, 7 mmol) and N-benzylmethylamine (1.71 mmol) in MeOH (2 ml) was allowed to stand for 4 d at 40 °C. The reaction
g, 14 mmol) was allowed to stand for 4 d at 40 °C. The reaction mixture was mixture was worked up in the same way as for the preparation of (Ϯ)-7 from
directly chromatographed on silica gel (60 g) to afford (Ϯ)-1 (0.08 g, 8%), (Ϯ)-1 to afford (Ϯ)-7 (0.116 g, 83%).
as for i) to afford (Ϯ)-1 (0.009 g, 16%), (Ϯ)-6 (0.021 g, 21%) and (Ϯ)-7
7
(Ϯ)-6 (1.259 g, 68%) as a colorless oil from n-hexane : AcOEtϭ9 : 1 eluate
Methyl (3S)-N-Benzylmethylamino-(4R,5R)-epoxyhexanoate 6 and
and (Ϯ)-7 (0.37 g, 22%) as a colorless oil from n-hexane : AcOEtϭ4 : 1 elu- Methyl (4S,5R)-4-N-Benzylmethylamino-5-hydroxy-2(E)-hexenoate
7
ate. (Ϯ)-6: IR (neat): 1736 cmϪ1; NMR: d 1.31 (3H, d, Jϭ5 Hz), 2.31 (3H, A mixture of (4R,5R)-1 (1.00 g, 7 mmol) and N-benzylmethylamine (1.71 g,
s), 2.42 (1H, dd, Jϭ7, 14 Hz), 2.61 (1H, dd, Jϭ7, 14 Hz), 2.82—2.88 (2H, 14 mmol) was allowed to stand for 4 d at 40 °C. The reaction mixture was di-
m), 3.08 (1H, dt, Jϭ7, 7 Hz), 3.69 (3H, s), 3.71, 3.75 (each 1H, d, Jϭ12 Hz), rectly chromatographed on silica gel (60 g) to afford (3S,4R,5R)-6 (1.203 g,
7.21—7.36 (5H, m). FAB-MS m/z: 264 (Mϩϩ1). (Ϯ)-7: IR (neat): 3439, 65%) as a colorless oil from n-hexane : AcOEtϭ9 : 1 eluate and (4S,5R)-7
1723 cmϪ1; NMR: d 1.20 (3H, d, Jϭ6 Hz), 2.22 (3H, s), 2.88 (1H, dd, Jϭ6, (0.463 g, 25%) as a colorless oil from n-hexane : AcOEtϭ4 : 1 eluate.
10 Hz), 3.49, 3.67 (each 1H, d, Jϭ13 Hz), 3.77 (3H, s), 4.11 (1H, dq, Jϭ6, 6 (3S,4R,5R)-6: Spectral data (IR, NMR) of (3S,4R,5R)-6 were identical with
Hz), 5.98 (1H, d, Jϭ16 Hz), 7.01 (1H, dd, Jϭ10, 16 Hz), 7.23—7.34 (5H, those of the reported (Ϯ)-6. [a]D25 ϩ17.3° (cϭ0.98, CHCl3); Anal. Found: C,
m). FAB-MS m/z: 264 (Mϩϩ1). ii) To a solution of (Ϯ)-1 (1.42 g, 10 mmol) 68.47; H, 8.52; N, 5.07. Calcd for C15H21NO3: C, 68.41; H, 8.04; N, 5.32%.
in MeOH (5 ml) was added N-benzylmethylamine (2.69 g, 22 mmol), and the (4S,5R)-7: Spectral data (IR and NMR) of (4S,5R)-7 were identical with
whole mixture was allowed to stand for 2 d at 40 °C. The reaction mixture those of the reported (Ϯ)-7. [a]D24 ϩ52.2° (cϭ1.01, CHCl3); HR-MS (FAB-
was evaporated and the residue was directly chromatographed on silica gel MS); m/z: 264.1551 (Mϩϩ1), Calcd for C15H22NO3: 264.1600.
(60 g, n-hexane : AcOEtϭ9 : 1) to afford (Ϯ)-6 (0.789 g, 30%) as a colorless
oil and (Ϯ)-7 (1.735 g, 66%) as a colorless oil in eluate order.
Treatment of (3S,4R,5R)-6 with CF3SO3H To a solution of (3S,4R,5R)-
6 (0.42 g, 1.6 mmol) in CH2Cl2 (40 ml) was added trifluoromethanesulfonic
Methylation of (؎)-4 and (؎)-5 i) To a solution of (Ϯ)-4 (0.14 g, 0.6 acid (0.4 ml) at Ϫ20 °C, and the whole mixture was stirred for 2 h at the
mmol) in CH2Cl2 (1 ml) was added methyl iodide (1 ml), and the whole mix- same temperature. The reaction mixture was diluted with cooled 7% aque-
ture was allowed to stand for 12 h at 0 °C. The reaction mixture was diluted ous NaHCO3 and extracted with CH2Cl2. The organic layer was washed with
with 7% aqueous NaHCO3 and extracted with CH2Cl2. The organic layer saturated brine, dried over MgSO4 and evaporated. The residue was chro-
was washed with saturated brine, dried over MgSO4 and evaporated. The
matographed on silica gel (10 g, n-hexane : AcOEtϭ3 : 1) to give (3S,4R,5S)-
residue was chromatographed on silica gel (10 g) to give (Ϯ)-6 (0.07 g, 12 (0.286 g, 72%) as a homogeneous oil. (3S,4R,5S)-12: [a]D24 ϩ53.2°
48%) from n-hexane : AcOEtϭ9 : 1 eluate and (Ϯ)-4 (0.03 g, 23%) from n- (cϭ0.76, CHCl3); IR (neat): 3376, 1713 cmϪ1; NMR: d 1.48 (3H, d, Jϭ5
hexane : AcOEtϭ1 : 1 eluate. The spectral data (IR and NMR) of (Ϯ)-6 were Hz), 2.25 (3H, s), 2.60 (1H, dd, Jϭ11, 18 Hz), 2.77 (1H, dd, Jϭ6, 18 Hz),
identical with those of the previous (Ϯ)-6. ii) To a solution of (Ϯ)-5 (0.12 g, 3.01 (1H, dt, Jϭ6, 10 Hz), 3.44 (1H, t, Jϭ10 Hz), 3.50, 3.72 (each 1H, d,
0.5 mmol) in CH2Cl2 (1 ml) was added methyl iodide (1 ml), and the whole Jϭ13 Hz), 4.10 (1H, dq, Jϭ6, 10 Hz), 7.26—7.37 (5H, m). FAB-MS m/z:
mixture was allowed to stand for 12 h at 0 °C. The reaction mixture was di- 250 (Mϩϩ1). HR-MS (FAB-MS); m/z: 250.1454 (Mϩϩ1), Calcd for
luted with 7% aqueous NaHCO3 and extracted with CH2Cl2. The organic C14H20NO3: 250.1443.
layer was washed with saturated brine, dried over MgSO4 and evaporated.
Acetylation of (3S,4R,5S)-12 A solution of (3S,4R,5S)-12 (0.26 g, 1
The residue was chromatographed on silica gel (10 g) to give (Ϯ)-8 (0.04 g, mmol) and Ac2O (4 ml) in AcOH (2 ml) was stirred for 12 h at room temper-
33%) from n-hexane : AcOEtϭ9 : 1 eluate and (Ϯ)-5 (0.05 g, 46%) from n- ature. The reaction mixture was diluted with toluene and evaporated. To the
hexane : AcOEtϭ1 : 1 eluate. (Ϯ)-8: IR (neat): 1736 cmϪ1; NMR: d 1.33
residue was added 7% aqueous NaHCO3, and the whole was extracted with
(3H, d, Jϭ5 Hz), 2.28 (3H, s), 2.49 (1H, dd, Jϭ6, 15 Hz), 2.64 (1H, dd, Jϭ8, ether. The organic layer was washed with saturated brine and dried over
15 Hz), 2.80—2.84 (2H, m), 3.08 (1H, dt, Jϭ6, 8 Hz), 3.64, 3.71 (each 1H,
d, Jϭ14 Hz), 3.71 (3H, s), 7.22—7.33 (5H, m). Anal. Found: C, 68.41; H,
MgSO4. The organic layer was evaporated to give a residue which was chro-
matographed on silica gel (10 g, n-hexane : AcOEtϭ4 : 1) to afford (4R,5S)-
8.04; N, 5.32. Calcd for C15H21NO3: C, 68.41; H, 8.04; N, 5.32%. FAB-MS 13 (0.145 g, 82%) as a homogeneous oil. (4R,5S)-13: [a]D27 Ϫ169.8°
m/z: 264 (Mϩϩ1).
(cϭ0.73, CHCl3); IR (neat): 1739 cmϪ1; NMR: d 1.44 (3H, d, Jϭ6 Hz), 2.14
Acetylation of (؎)-7 A solution of (Ϯ)-7 (1.11 g, 4.2 mmol) and Ac2O (3H, s), 4.60 (1H, dq, Jϭ7, 7 Hz), 5.28 (1H, ddd, Jϭ1, 3, 7 Hz), 6.11 (1H,
(0.53 g, 5.2 mmol) in pyridine (10 ml) was stirred for 12 h at room tempera- dd, Jϭ1, 10 Hz), 6.78 (1H, dd, Jϭ3, 10 Hz). Anal. Found: C, 56.31; H, 6.13.
ture. The reaction mixture was diluted with H2O and extracted with ether. Calcd for C8H10O4: C, 56.47; H, 5.92%. FAB-MS m/z: 171 (Mϩϩ1).
The organic layer was washed with 2 M aqueous HCl, 7% aqueous NaHCO3
(؊)-Osmundalactone 11 A mixture of (4R,5S)-13 (0.09 g, 0.5 mmol)
and saturated brine, and dried over MgSO4. The organic layer was evapo- and lipase Amano P (0.1 g) in 0.1 M phosphate buffer solution (20 ml) was
rated to give a residue which was chromatographed on silica gel (20 g, n- stirred for 2 d at 30 °C. The reaction mixture was extracted with AcOEt and
hexane : AcOEtϭ9 : 1) to afford (Ϯ)-9 (1.13 g, 88%) as a homogeneous oil.
the organic layer was dried over MgSO4. The organic layer was evaporated
(Ϯ)-9: IR (neat): 1737 cmϪ1; NMR: d 1.32 (3H, d, Jϭ6 Hz), 1.97 (3H, s), to give a residue which was chromatographed on silica gel (10 g, n-hexane :
2.24 (3H, s), 3.06 (1H, t, Jϭ9 Hz), 3.48, 3.66 (each 1H, d, Jϭ13 Hz), 3.78 AcOEtϭ1 : 1) to afford (Ϫ)-11 (0.062 g, 92%). Crystallization of (Ϫ)-11
(3H, s), 5.21 (1H, dq, Jϭ6, 9 Hz), 5.91 (1H, d, Jϭ16 Hz), 6.95 (1H, dd, Jϭ9, from benzene gave a colorless crystal (Ϫ)-11. (Ϫ)-11: mp 81—82 °C (lit.
16 Hz), 7.23—7.34 (5H, m). Anal. Found: C, 66.67; H, 7.90; N, 4.49. Calcd mp 82—82.5 °C)4a); [a]D22 Ϫ69.0° (cϭ0.46, H2O) (lit. [a]D22 Ϫ70.6° (cϭ2.0,
for C17H23NO4: C, 66.86; H, 7.59; N, 4.59%. FAB-MS m/z: 306 (Mϩϩ1).
Acetal Formation from (؎)-7 To a solution of (Ϯ)-7 (0.22 g, 0.8 (1H, dt, Jϭ2, 9 Hz), 4.39 (1H, dq, Jϭ6, 9 Hz), 5.97 (1H, dd, Jϭ2, 10 Hz),
mmol) in Tetrahydrofuran (THF) (2 ml) was added benzaldehyde (0.3 g, 2.7 6.88 (1H, dd, Jϭ2, 10 Hz). Anal. Found: C, 56.14; H, 6.37. Calcd for
H2O))4a); IR (neat): 3410, 1710 cmϪ1; NMR: d 1.49 (3H, d, Jϭ6 Hz), 4.25
mmol) and potassium tert-butoxide (0.15 g, 1.2 mmol) at 0 °C, and the whole
C6H8O3: C, 56.25; H, 6.29%. FAB-MS m/z: 129 (Mϩϩ1).
mixture was stirred for 15 min at 0 °C. The reaction mixture was diluted
Methyl (4R,5S)-4-Dimethylamino-5-hydroxy-2(E)-hexenoate 16 To a
with saturated aqueous NH4Cl and extracted with ether. The organic layer solution of (4S,5S)-1 (0.81 g, 5.7 mmol) in MeOH (16 ml) were added di-

July 2001
853
Ϫ85.7° (cϭ0.78, MeOH) (lit. D-(ϩ)-18: [a]D25 ϩ86.1° (cϭ0.9, MeOH))6); IR
(neat): 3388, 2937 cmϪ1; NMR: d 1.20, 1.28 (each 3H, d, Jϭ6.5 Hz), 1.40—
2.05 (8H, m), 2.20—2.35 (2H, m), 2.21, 2.24 (each 6H, s), 3.66—3.72 (1H,
m), 3.56, 4.38 (each 1H, dq, Jϭ6.6, 9 Hz), 4.71 (1H, dd, Jϭ1.2, 9 Hz), 5.19
(1H, t, Jϭ2 Hz). Anal. Found: C, 60.72; H, 10.90; N, 8.57. Calcd for
C8H17NO2: C, 60.34; H, 10.76; N, 8.80%. (4R,5S)-19: [a]D24 ϩ10.7°
(cϭ0.93, MeOH); IR (KBr): 3357, 2936 cmϪ1; NMR: d 1.19 (3H, d, Jϭ6
Hz), 1.53-1.86 (4H, m), 2.28—2.32 (1H, m), 2.38 (6H, s), 3.51—3.57,
3.64—3.69 (each 1H, m), 4.10 (1H, dq, Jϭ2, 6 Hz). Anal. Found: C, 59.96;
H, 12.01; N, 8.79. Calcd for C8H19NO2: C, 59.59; H, 11.88; N, 8.69%.
methylamine hydrochloride (0.93 g, 11.4 mmol) and triethylamine (1.15 g,
11.4 mmol), and the whole mixture was allowed to stand for 1 d at 40 °C.
The reaction mixture was evaporated to give a residue which was chro-
matographed on silica gel (20 g, CHCl3 : MeOHϭ50 : 1) to afford a mixture
(0.846 g, 81%) of (3R,4S,5S)-14 and (3S,4S,5S)-15 as an oil. This mixture
in MeOH (16 ml) again stood for 2 d at 50 °C. The reaction mixture was
evaporated to give a residue which was again chromatographed on silica gel
(30 g) to afford (4S,5S)-1 (0.084 g, 13%) from CHCl3 eluate and (4R,5S)-16
(0.516 g, 61%) as a colorless oil from CHCl3 : MeOHϭ50 : 1 eluate. NMR
data of (4S,5S)-1 were identical with those of (Ϯ)-1. (4R,5S)-16: [a]D20
Ϫ115.9° (cϭ1.02, CHCl3); IR (neat): 3430, 1724 cmϪ1; NMR: d 1.09 (3H,
d, Jϭ6 Hz), 2.28 (6H, s), 2.55 (1H, dd, Jϭ4, 10 Hz), 3.76 (3H, s), 4.08 (1H,
dq, Jϭ4, 6 Hz), 5.93 (1H, d, Jϭ16 Hz), 6.91 (1H, dd, Jϭ10, 16 Hz). Anal.
Found: C, 57.34; H, 9.18; N, 7.52. Calcd for C9H17NO3: C, 57.74; H, 9.15;
N, 7.48%. FAB-MS m/z: 188 (Mϩϩ1).
Conversion of (4R,5S)-16 into L-(؊)-Forosamine 18 i) A solution of
(4R,5S)-16 (0.47 g, 2.5 mmol) in MeOH (10 ml) was hydrogenated at an or-
dinary temperature and pressure over 20% Pd(OH)2–C (0.08 g). After hydro-
gen absorption had ceased, the catalyst was filtered off and the filtrate was
evaporated to give a residue. A solution of the crude residue in 80% aqueous
AcOH (10 ml) was refluxed for 2.5 h and the reaction mixture was diluted
with benzene. The organic layer was evaporated to give a residue which was
chromatographed on silica gel (15 g, CHCl3 : MeOHϭ4 : 1) to afford a d-lac-
tone (4R,5S)-17 (0.32 g) as an oil. ii) To a solution of (4R,5S)-17 (0.31 g) in
THF (8 ml) was added a 1.5 M solution of DIBAH (5.9 ml, 8.8 mmol) in
toluene at Ϫ30 °C, and the whole mixture was stirred for 1 h at the same
temperature. Aqueous MeOH (MeOH : H2Oϭ3 : 1, 8 ml) was added to the
reaction mixture at Ϫ30 °C and the whole mixture was filtered off with the
aid of Celite. The precipitate was washed with acetone. The washing was
combined with the filtrate and the combined organic layer was evaporated to
give a residue, which was chromatographed on silica gel (10 g) to afford L-
(Ϫ)-forosamine 18 (0.08 g, 20% yield from (4R,5S)-16) as an oil from
CHCl3 : MeOHϭ9 : 1 eluate and (4R,5S)-19 (0.1 g, 24% yield from (4R,5S)-
16) as an oil from CHCl3 : MeOHϭ4 : 1 eluate. L-(Ϫ)-forosamine 18: [a]D24
References and Notes
1) Ono M., Saotome C., Akita H., Tetrahedron:Asymmetry, 7, 2595—
2602 (1996).
2) Saotome C., Ono M., Akita H., Tetrahedron:Asymmetry, 11, 4137—
4151 (2000).
3) Evans D. A., Gauchet-Prunet J. A., J. Org. Chem., 58, 2446—2453
(1993).
4) Isolation and bioactivity: a) Hollenbeak K. H., Kuehne M. E., Tetrahe-
dron, 30, 2307—2316 (1974); b) Numata A., Hokimoto K., Takemura
T., Katsuno T., Yamamoto K., Chem. Pharm. Bull., 32, 2815—2820
(1984); Synthesis of (Ϯ)-11: c) Torssell K., Tyagi M. P., Acta Chem.
Scand., B, 31, 297—301 (1977); d) Yamagiwa S., Sato H., Hoshi N.,
Kosugi H., Uda H., J. Chem. Soc., Perkin Trans. I, 1979, 570—583.
Synthesis of (Ϫ)-11: e) Murayama T., Sugiyama T., Yamashita K.,
Agric. Biol. Chem., 50, 2347—2351 (1986); f ) Noshita T., Sugiyama
T., Kitazumi Y., Oritani T., Tetrahedron Lett., 35, 8259—8262 (1994).
5) As a preliminary experiment, the acetylation of (Ϯ)-12 with Ac2O–
pyridine gave the corresponding acetate (95%) which was treated with
Ac2O/AcOH (2 : 1) to afford (Ϯ)-13 (87%) and N-methylbenzylamine
acetate (68%).
6) Dyong I., Knollmann R., Jersch N., Angew. Chem., Int. Ed. Engl., 15,
302—302 (1976).