発癌促進物質,アプリシトキシンの形式全合成

九州大学学術情報リポジトリ

Kyushu University Institutional Repository

発癌促進物質,アプリシトキシンの形式全合成

岡村, 浩昭

https://doi.org/10.11501/3099951

出版情報：Kyushu University, 1994, 博士（理学）, 論文博士

A FORMAL TOTAL SYNTHESIS OF APLY SIATOXIN, A POTENT CANCER

PROMOTER

November, 1994 Hiroaki Okamura

Chapter 1.

Chapter 2.

Chapter 3.

3-1 3-2

Chapter 4.

Contents

Introduction

Retrosynthetic Analysis

Synthesis of Aplysiatoxin Synthesis of Fragments Assembly of Fragments

Experimental

Acknowledgements

References and Notes

1

8

13 17

31

68

69

Chapter 1

Introduction

Even in the Ron1an age, toxin of sea hares which are gastropod mollusks without shell, was well known to the people I) but their chemical characterization has not been elucidated until recently. In 197 4, Kato and Sheuer isolated the strong toxins (LD 100 0.3 mg/Kg, ip mouse) which narned aplysiatoxin (1a) and debr01noaplysiatoxin (lb ), from the digestive gland of sea hare, Stylocheilus longicauda, and clarified their plain structures.2a) In 1978, Moore et al. obtained debromoaplysiatoxin (1b), oscillatoxin A (1c), and their brominated derivatives (1d-f) as causative agents of a severe contact dermatitis that sometirnes affects swimmers in Hawaii, from the mixture of marine algae, Oscillatoria nigroviridis and Scizothrix calcicola.2d) In 1984, they also isolated 1 a and 1 b frorn blue-green alga, Lyngbya majuscula, and elucidated their stereochemistries including absolute configurations based on I H NMR analyses, NOE experiments, CD spectra, and X ray crystallographic study of dibromoaplysiatoxin (1g) which was obtained by brornination of 1b.2e) Since S. longicauda feeds primarily on the alga, L.

majuscula, it has been presumed at present that the genuine producer of this series of compounds is not sea hare but marine algae.

x2

OH

R XI X2 x3

la CH3 Br H H aplysiatoxin

1 b CH3 H H H debromoaplysiatoxin

1c H H H H oscillatoxin A

1d H Br H H bromooscillatoxin

1e H Br Br H dibromooscillatoxin

1f CH3 Br Br H bromoaplysiatoxin

1g CH3 Br Br Br dibromoaplysiatoxin

Along with the strong cytotoxicity and inflammatory activity, potent tumor promoting activity was reported for this series of compounds by Fujiki et al. in 1982_3) At that time, two classes of tumor protnoters, phorbol ester derivatives and teleocidin derivatives, had been known.

12- 0-tetradecanoylphorbole-13-acetate (TPA)

dihydroteleocidin B

Since the potency of tumor pr01notion by the compound in aplysiatoxin family is comparable to that of those known tumor promoters, la and its related compounds have been classified into the third class today. Although these three classes of tumor promoters have quite different structures, it has been reported that they all cotnbine to the same receptor, protein kinese C. However, this interesting but riddling structure-activity relationship remained unclarified to date.

In addition to the biological interests, aplysiatoxin family provides attractive targets for total synthesis due to their characteristic structures, which consist of ten stereogenic centers including four consecutive asytnmetric centers (C9-C 12, hereafter the number of the position is according to aplysiatoxin numbering), rnacro diolide structure, spiro acetal, and hemiacetal hydroxy group that is extremely sensitive to acid and base.2c,2e) Since Kato and Scheuer's report, several groups attempted the synthesis of this class of compounds4) and, in 1987, Kishi's group accomplished the first and only total synthesis of aplysiatoxin (la) and debromoaplysiatoxin (lb) in opticaly active forms.5) Later, Ireland4a) and Yamamura4b,4c) reported the synthesis of dehydroxyaplysiatoxin which lacks the labile C3 hemiacetal hydroxy group. Still no other total synthesis of this series of compounds has been reported and it remains as a challenging target.

To achieve an efficient synthesis of the compounds in a ply iatoxin family, Inethodologies for enantioselective construction of consecutive asymmetric centers (polypropionate segment) and for macrocyclization under very mild conditions are required.

In the last decade, many useful methodologies for the stereoselective construction of polypropionate segment have been reported by several groups as shown in Scheme 1. These methodologies have used epoxidation and subsequent regioselective methylation,6) stereoselective catalytic hydrogenation,7) hydroboration,8) aldol reaction of �-silyl aldehyde,9) hetero Diels-Alder reaction, I 0) and therrnodynamically stereocontrolled spiroacetalization and subsequent stereoselective reduction, II) respectively, as key steps.

B n_� '

^{n A.}

cHO Bn � OH

1) Asymm.

epoxidation

2) LiCuMe2

Bn � ^OH

OH

1) Asymm.

n

1 l l

H �

epoxidation

H� ';') Y: J

HO OH

"-.'.·.·Ko 0

0 OBn

Ox 0

2) LiCuMe2

0 ^{H Ox} 0

Kishi's method

�0 �

) 0�.0 ^OBn �0� ) ^� 0

,,···

- ^{... OBn}

.... o

^s

ⁿ 0

Stork's method Scheme 1. Continued

EtOO

1) thexylborane 2) B2H6

3) [0]

�

OR OR

TrO

TrO

�

^OTr

OH OH OH

COOEt

- -

TrO OH OH OH

OTr OH OH OH OH OH

Still's method

SiMe3

$)l!

^Bn

^�

_MgBr ^Me3Si

_�

^{OH OH OBn}

1) ^TBHP, V5+

2) Base

Sato's method

Scheme 1. Continued

OTr

OMe

Me3Si

�

0

QDMPM

OR

Y 9

?R \

0

�

: H :.

Y 9

?R \

HO

�

: H :

Danishefsky's method

�

Thermodynamic ^control

1&-

0 ^� 0

I I

0 0

H

�

^{- HQ}

HO

Albizati's method Scheme 1

Recently a different but equally efficient methodology based on stereoselective [2,3]Wittig rearrangement has been reported for the stereoselective synthesis of polypropionate segment from our group_l2) This tnethodology has been successfully applied to the synthesis of Ireland alcohol, 12d, 12f) key intermediate for the synthesis of tirandamycin 13) (Scheme 2)

�

^OBn [2,3]Wittig

rearrangement ^OH I odol acton iza tion

�

⁰

BnO : : OH

I ⁼

OH OTHP

�

^OPiv

HO . .

X

^OTHP

WrY

OMPM 0

� .

BnO�C02Pri

OTHP

0 ⁼

�

^OPiv

BnO :

�

^{0 OTHP}

�

^{. . CHO}

0

�

^-

Scheme 2

Ireland alcohol 0H

tirandamycin

By taking the advantage of this procedure and titanium mediated asymmetric epoxidation 14) (hereafter referred to as A.E.) that is the most appropriate method for the introduction of isolated asymmetric centers, the author achieved the synthesis of aplysiatoxin by way of Kishi's aldehyde (2), which has been converted into aplysiatoxin (la) in six steps.5) This paper describes in full

detail the results obtained in this study _I 5)

0 H

�:uy

OBOM OBOM

2: Kishi's aldehyde

Chapter 2

Retrosynthetic Analysis

Aplysiatoxin (la) and debromoaplysiatoxin (lb) have the co1nn1on structure except for the aromatic moiety, Cn-carbon in which is not brominated in lb _{but in} la. Since it has been already known that lb can be convetied into la by treatment with aqueous bromine solution at pH 6,2e) the synthesis of lb is a prin1ary goal of this research.

In the decision of the synthetic plan of the aplysiatoxins, the special attention rnust be paid to the time of the construction of herniacetal moiety, because it is extremely sensitive to either acid or base conditions2c, 2e) and the manipulation of most protecting groups is considered to be very difficult in the presence of the hemiacetal moiety. Therefore, the formation of the hen1iacetal moiety was planned to be carried out at the last stage of the synthesis. Another problem in this synthesis is the introduction of complex stereochemistries. Compound lb has ten stereogenic centers, but two of them, C3 and C7 acetal carbons, are expected to be introduced with the desired chirality upon the lactonization of the con·esponding seco-acid. Four of the remaining eight asymn1etric centers are consecutive ( C9-C 12). These asymrnettic centers seemed to be constructed with stereoselective titanium-mediated [2,3]Wittig rearrangen1ent as a key step. Based on these analyses, the synthetic strategy was elaborated as described in a retrosynthetic manner in Scheme 3.

Dissociation of C 1 ester linkage in lb gave seco-acid A as an immediate precursor of lb.

Further cleavage of the bond C2-C3 provided acid B and acetic acid, which might be rec01nbined by Claisen condensation reaction. Dissociation of another C27 ester linkage envisioned two

fragments C and D as plausible intermediates. Further disconnection of fragn1ent C or of its opened form C' between bonds C7-C3 and C12-C13 generated three fragments E, F, and G, which were considered to be recombined by nucleophilic opening of two different epoxides in F with Grignard reagent (fragment E) and with lithiodithiane derivative generated by base treatment of G, respectively. In the actual synthesis, the terminal epoxide in F was masked as a protected diol until the coupling with G. The carboxylic acid in G was also masked as a protected diol until an appropriate stage in order to avoid the epimerization at C4. Stereogenic carbons in fragments D, E, and G seemed to be introduced by using ti tani urn mediated A.E. 14) and the contiguous stereogenic carbons in fragment F was considered to be derived stereospecifically from £­

benzyloxylated a-hydroxy acid (I) which could be readily prepared by the use of titanium-

mediated [2,3]Wittig rearrangement6d) of optically active substance. Although the configuration of a-carbon in I was opposite to that of C9 in 1, a-hydroxy ester was expected to be converted into a terminal epoxide with the inversion of its a-configuration. According to this synthetic plan, the synthesis of fragments D, E, F, and G were star1ed.

C02H

1 17

OH

Py

^o

^c=:=>

^{''o OH}

- ₀ Z7 :rl 0

�

OH OP1 OP2

1b A

rY

2

+ CH3C02H

9 ·,,0 OH

OP1

0�

OP2

rY

B

D

0

(1+� ^{¢ �}

Scheme 3. Continued

OP2 OH

OMe 3

MgBr ₊

�

⁺

c

^{s OP5}

\.

)2

^�

\

^·

s

f

^/'-.,

^�

^{- /OP6}

OP,

�

� 7 ^{(\ T}

^{3 '-.._/}

E F

QH

BnO

� 12

^{� 9 C02Pr-i}

Scheme 3

G

Chapter 3

Synthesis of Aplysiatoxin

3-1. Synthesis

^of

Fragments

Synthesis of fragment D started with epoxy alcohol

4

which was readily prepared frotn crotyl alcohol

(3)

according to the literature procedurel4b)

(Scheme 4).

30 28

� OH

^a

3

OR

� 0

6----{

e(6R=H 0

7

R

⁼

BOM

�s

OR S�

k (10 R

⁼

H

11

R

⁼

MPM

0

�OH

b, c

4 90% ee

OBOM

�OH

g,

h, i

l,m --

OH

8

OMPM H2oc_ _.J.

^{'lo /}

� �

0 0

�O)lNHPh

5 100% ee

OBOM

�0

9

OBOM

1 2

d

a) Ti(OiPr)4, (-)-DET, TBHP, MS 4A b) Et3N, PhNCO c) recrystalization from AcOEt (74°/o for 3 steps) d) dil. HCI04 (15°/o) e) ipr2NEt, BOMCI (70°/o) f) K2C03, MeOH (93°/o)

g)

(CH3)3CCOCI, Et3N then MsCI h) DIBAL i) KOH, MeOH (55°/o for 3 steps)

j)

1,3-dithiane, n-Buli (99°/o)

k)

NaH, MPMCI (82°/o) I) Mel, CaC03 m) NaCI02, 2-methyl- 2-butene (65°/o for 2 steps)

Scheme 4

Treatment of

4

with phenyl isocyanate gave crystalline carbamate 5. Since the optical purity of the starting

4

was 90o/o ee, the carbamate was recrystallized to afford an optically pure compound. Compound 5 thus obtained was treated with perchloric acid and the resulting alcohol 6 was protected as a benzyloxyn1ethyl (BOM) ether 7. _Compound 7 was subjected to alcoholysis

to give diol 8 and then converted into epoxide 9 _{by the sequence:} i) protection of the resulting primary hydroxyl group as a pivalate, ii) mesylation of the remaining secondary hydroxyl group, iii) reductive cleavage of pivaloyl group, and iv) alkaline treatment of the hydroxy mesylate. One

was transformed into the desired carboxylic acid 12 corresponding to fragment D in three step : i) protection of hydroxyl group as a p-methoxybenzyl

(MPM)

ether 1 1,16) ii) hydroly is of dithioacetal, and iii) oxidation of the resulting aldehyde into carboxylic acid with sodium chlorite. ^l7)

For the synthesis of fragment E, ester 13 was employed as a starting material which was readily prepared from m-hydroxycinnamic acid by a conventional manner

(Scheme

_5).

17 15 13

a

YOH

Y C02Et

0

OBOM

1 3

OH

d, e, f

OBOM OH

1 6

+ 16:16' = 17:1

[VLOH] OBOM

_{1 6,}

OBOM

1 4

OMe

OBOM

1 7

b �OH

OBOM

1 5

OMe

96% ee

Br

g

OBOM

1 8

c

MgBr

a) DIBAL (92°/o)

b)

Ti(O;Pr)4, (-)-DIPT, TBHP, MS

4A

_(90°/o)

c)

Red-al, then Nai04 (91 °/o) d) TsCI, Et3N, DMAP e) Mel, NaH f) NaBr (63°/o for 3 steps) g) Mg

Scheme

₅

Ester 13 was converted into allylic alcohol 14 in good yield by diisobutylaluminum hydride

(DIBAL)

reduction.

A.E.

of 14 proceeded smoothly with enantioselectivity of 96% ee to give

epoxy alcohol15, although structurally similar p-methoxycinnamyl alcohol was a poor substrate for this titanium-mediated A.E. I 8) Reduction of 15 with sodium bis( methoxyethoxy)aluminum hydride (Red-al) I 9) gave a mixture of 1,3- ( 16) and 1 ,2-diol ( 16') in a ratio of 17: 1.20) To remove the undesired I ,2-diol, the mixture was treated with Nai04 and sub

j

ected to silica gel chromatography to give the desired 16 in a pure form. Transformation of 16 into bromide 17

was carried out in three steps: i) tosylation of primary hydroxyl group, ii) Inethylation of

secondary hydroxyl group, and iii) substitution of the tosylate with bromide anion. At the econd step in this sequence, the addition of methyl iodide to a solution of the hydroxy n1esylate ptior to the addition of sodium hydride is indispensable to prevent the forrnation of the undesired oxetane.

The resulting bromide 17 was converted in a usual manner into Grignard reagent

18

to be used for the coupling with fragment F.

Synthesis of fragment F started from racemic (3£)-1-benzyloxy-3-buten-2-ol

(dl-19)

as shown in Scheme 6.

� 11

Y!OBn

a

OH

(d/)-19

QH

�OBn OH

(R)-19

71% _ee

e, f

b, c � 11

�

OBn

8

20

<t--fo

d ¹¹ � .

BnO�C02Pri

� ^g,h,i

BnO : : OH

21

OTHP

0 ::

BnO �OH .

23

j,

k

I :.

22

OTHP

�OTBDMS

HO .

₇

24

a) Ti(OiPr)4, (-)-DIPT, TBHP

(36 °/o)

b) bromoacetic acid, NaH c) iPrl, Na2C03

(76°/o

for

2

steps) d) LOA, Cp2 TiCI2

(72°/o)

e) KOH f) l2

(62°/o

for

2

steps)

g)

DHP, PPTS

h)

K2C03, MeOH

i)

^LAH

(75°/o

for

3

steps) j) TBDMSCI, lmH

k)

H2, Pd/C

(84°/o

for 2 steps)

Scheme 6

Racemic 19 was first subjected to kinetic resolution by using a Ti(QiPr4), (-)-DJPT, and TBHP system5b that is generally very effective for the resolution of racemic secondary allylic alcohols. Contrary to the previous expectation, however, kinetic resolution of dl-19 was not so effective. At the stage of 64o/o conversion of the starting material, the remained

(R)-19

showed

only 71 o/o ee. Although (R)-19 was not optically pure, it was directly used for the next reaction because it was expected that the stereoisomer originating in the unde ired enantiomer (S)-19 could be removed at the latter stage of the synthesis. Compound (R)-19 was converted into ester 20 according to the reported procedure,21) and titanium-n1ediated [2,3]Wittig rearrangen1ent of ²⁰ afforded £-benzyloxylated hydroxy ester 21 with quantitative chirality transfer together with high syn, E-selectivity.22) After alkaline hydrolysis, 21 was subjected to iodolactonization.23) The resulting lactone 22 was converted into 2,3-syn-3,4-anti-epoxy alcohol 23 in three steps; i) protection of hydroxyl group as a tetrahydropyranyl (THP) ether,24) ii) rnethanolysis of lactone along with epoxide formation, and iii) lithium aluminum hydride (LAH) reduction of the resulting methyl ester. Hydroxy protection as a t-butyldimethylsilyl (TBDMS) ether followed by hydrogenolysis afforded epoxy alcohol 24 which was a synthetic equivalent of fragment F.

Fragment G was derived from easily available aldehyde 2525) (Scheme 7). Compound 25 was first converted into THP ether 26 in a conventional manner. Conversion of 26 into 27 was achieved by oxidative cleavage of olefin and subsequent Wittig-Honer olefination.6) DIBAL reduction of 27 afforded allylic alcohol 28 which was transformed into epoxy alcohol 29 of 9So/o ee. After hydroxy protection as a trityl ether 30, compound 29 was exposed to methylmagnesium bromide in the presence of a catalytic amount of Cui giving a rnixture of 31 and its regioisomer in a ratio of 2.7: 1.26) Acid treatment of the mixture followed by chromatographic separation afforded the desired triol 32 which was converted into carbonate 33 by treatment with carbonyldiimidazole. Transformation of 33 to dithioacetal 35 corresponding to fragment G was carried out by the sequence; i) oxidation of primary alcohol to aldehyde,27) ii) dithioacetalization of the resulting aldehyde giving dithioacetal 34, iii) alcoholysis of carbonate, and iv) reprotection of the resulting diol ^{as an} acetonide.

OHC

�

^a,

^{b c,}

^d

�

^{C02Et ___ e __ �}

7 3

25

THPO

�

^{OH ---}

THPO

28

THPO

2 6 27

�

^{OR h}

THPO

g

(

_{29 R = H}

30 R = Tr

�

^OTr

[+ ^{)( r}

^/'--OT

J

THPO OH THP

r � lr -

H

�

_OR

OR'

3 1 3 1 '

31:31' ⁼ 2.7:1

k.l

cfClq

^R

R'

m,n

(

_{34 R =}

_�·

_{= o}

35 R, R ⁼ Me

j (

_{32 R = H, R' = H}

33 R = R' =CO

a) NaBH4

b)

^{DHP, TsOH}

(56°/o

for 2 steps)

c)

Os04, Nai04 d) (;Pr0)2P(O)CH2C02Et, NaH

(87°/o

for 2 steps) e) DIBAL

(95°/o)

f) Ti(O;Pr)4, (+)-DIPT, TBHP

(86°/o)

g) TrCI, Et3N, DMAP

(85°/o)

h) MeMgBr, Cui i) CSA, MeOH

(56°/o

for 2 steps) j) C01m2, DMAP

then dil. HCI

(89°/o)

k) Swern oxdn. I) 1 ,3-propanedithiol, BF3•0Et2

(72°/o

for 2 steps) m) K2C03, MeOH n) 2,2-dimethoxypropane, PPTS

(90°/o

for 2 steps)

Scheme 7

3-2. Assembly of Fragments

With fragments D, E, F, and G in hand, the stage was set for the construction of 1. It was started with the coupling of fragments E and F as desctibed in Scheme 8. Treatment of epoxy alcohol 24 with Grignard reagent 18 in the presence of a catalytic atnount of Cui provided l ,3-diol 36 and a small amount (ca.

15o/o)

of the undesired stereoisomer due to insufficient optical purity of 24 (vide

supra).

Since the separation of 36 and its diastereomer was difficult at this stage, the

mixture was used for the next step without separation. A mixture of 36 and its diastereOiner was

subjected to mesylation followed by metal hydride reduction in order to establish the structure of the Cg-C21 fragtnent. The desired mesylation proceeded smoothly but the treatment of the resulting mesylate with LiBHEt3 (super hydride) which has been reported to be the most efficient reagent for substitution of mesyloxy group with hydride,28) gave the oxetane derivative 38 predominantly.

OMe

MgBr

OBOM ^a

+

OTHP

0 =

�OTBDMS

HO .

12 : ₈

24 71%ee

MeO HO QTHP

OBOM

₃₇

OBOM

₃₈

MeO HO QTHP

OTBDMS

b, c

OBOM

36

OH d, e, f,

g

OBOM

₃₉

a) Cui (71 °/o) b) MsCI, Et3N c) LAH (62°/o for 2 steps) d) Ac20, DMAP e) PPTS, MeOH f) MsCI, Et3N, DMAP g) KOH, MeOH (24°/o for 4 steps)

Scheme 8

This unexpected result was considered to be attributable to that super hydride abstracted proton from the primary hydroxy group and that the resulting alkoxy group attacked the mesylate

to give oxetane ring. However, it was expected that, if lithium alun1inum hydride (LAH) was used instead of super hydride, the aluminum hydride which was caught by the alkoxide would attack the mesylate in a intramolecular fashion to give the desired 1 ,3-diol (Scheme 9).

/"'). _)

^__

} - u•

Mso

\ fast slow\

H-BEt3Li

oxetane formation

Intramolecular hydride delivery

Scheme 9

/y.

^{.. ,,OH}

Based on this analysis, we examined the reduction of the mesylate with LAH. As expected, the reduction of the mesylate proceeded smoothly but the cleavage of the TBDMS ether also occurred simultaneously to give I ,4-diol 37. Therefore, 37 was reprotected as a diacetate and converted into 39 _{by the sequence;} i) acid hydrolysis of THP ether,24) ii) mesylation of the resulting hydroxyl group, and iii) alkaline hydrolysis of acetates accompanying epoxide ring formation. The undesired minor diastereomer produced at the coupling of 18 _{and 24, could be} removed at this stage by repeated silica gel colutnn chron1atography. Although c01npound 39 _was obtained with sufficient purity and set the stage for the coupling with 35 corresponding to fragtnent G, difficulty in separation of diastereomeric by-product and the tedious operation of protecting

groups caused by the unexpected cleavage of TBDMS ether, prompted us to explore a more efficient approach to 39.

In order to avoid these difficulties, we examined the approach to 39, using A.E. of E-allylic alcohols as a tool of introducing chirality at C9-C 12, because A.E. of E-allylic alcohols has been well established to proceed with high enantioselectivity20) (Scheme 10).

11 � OH BnO� ₈

40

a

- �'-

OH

BnO./��

b,c

41

0

+

Et02C

�

^{O g}

0

+

�

⁰

43

�

OMe ^MgBr OBOM

1 8

MeO

HO :

44

HO

0

8

OH OBOM

m,n

OBOM

46

OR

0

( 49

^{R = MPM}

39 R ⁼ H

j, k

d,e,f

42

h ^HO

~

₁₂ ^: ₈

45

MeO OR

0

OBOM

1

(47

^{R = H}

48

^{R = MPM}

a) Ti(OiPr)4, (+)-OIPT, TBHP

(79°/o) b)

Me 3AI

c)

2,2-dimethoxypropane, CSA

(58°/o

for 2 steps) d) H2, Pd/C e) Swern oxdn. f) (iPr0)2P(O)CH2C02Et, 1BuOK

(44°/o

for 3 steps)

g)

OIBAL

(96 °/o)

h) Ti(OiPr)4, (+)-OIPT, TBHP

(82°/o)

i) Cui

(76°/o) j)

TsCI, Et3N

k)

LAH

(90°/o

for 2 steps) I) MPMCI, NaH

(74°/o)

m) PPTS, MeOH

(48°/o)

n) TsCI, 1BuOK

(77°/o)

o) 000, H20

(95o/o)

Scheme 10

Thus, the synthesis started with epoxy alcohol 41

(95o/a

ee) which was readily prepared by A.E. of allylic alcohol 40.29) Epoxy alcohol 41 was treated with ttimethylaluminum, according to Oshima's method)O) Although this procedure provided a mixture of 1,2-and l ,3-diols in a ratio

of 5:1, the undesired l ,3-diol was readily separated with silica gel column chrotnatography after their conversion into the corresponding acetonides. Further conver ion of acetonide

42

into two carbon elongated £-allylic alcohol 44 wa carried out in four tep ; i) hydrogenoly is of benzyl

ether, ii) Swern oxidation of the resulting alcohol to aldehyde,27) iii) Wittig-Horner olefination,6) wherein an inseparable mixture of 43 and its epimer3l) was produced in a ratio of 13: I, and iv) DIBAL reduction, after which 44 and its epimer were separated by column chromatography. A.

E. of 44 proceeded with diastereoselectivity of 86o/o de,20) and the resulting mixture of 45 and its diastereomer was used for the next reaction without separation. Coupling of 18 and the above mixture of epoxy alcohols proceeded regioselectively to give the corresponding diastereomeric mixture of 1 ,3-diols which were separated chromatographically to give 46 as a single isomer.

The following transforn1ation of hydroxy methyl group in 46 to the conesponding methyl group was carried out in the same manner as described for the preparation of 37. Conversion of the resulting 47 into the terminal epoxide 49 was effected straightforwardly by the sequence: i) protection of hydroxyl group as a MPM ether 48, 16) ii) acid hydrolysis of acetonide, and iii) treatment of the diol with p-toluene ulfonyl chloride in the presence of excess potassium t­

butoxide. The resulting protected epoxide

49

was converted into desired

39

by 2,3-dichloro-5,6- dicyano- l ,4-benzoquinone

(DDQ)

oxidation.l6)

Having established an efficient route to

39,

we next examined its conversion into fragment C as shown in

scheme 11.

The coupling of

39

and lithiodithiane which was derived from

35

and butyl lithium, proceeded smoothly to afford

50

in good yield. Deacetalization followed by dethioacetalization32) also proceeded smoothly to give the desired spiro acetal

52

spontaneously.

However, attempts to oxidatively cleave C2-C3 bond by using RuQ433) in order to give lactone

53

turn out a failure, because the other oxygen functionalities in

52

were not stable to the reaction conditions. Therefore, the author examined the construction of spirolactone

53

by way of one carbon contracted aldehyde

5

⁴

(scheme 12).

MeO

OBOM 3 ₉

b

OH

OBOM

MeO

OBOM 52

s

�

o

C

s

�

35

;

^---

8

a

OBOM 50

c

51

OH

Oxidaive cleavage

�

^\

OBOM

a) n-Buli, TMEDA (92%} b) PPTS, MeOH (66%} c) NCS (77%}

Scheme 11

o

-\--

0

53

The one carbon contracted aldehyde 54 could be obtained in good yield by treatment of terminal diol 51 with lead tetraacetate34) that is compatible with thioacetal labile to oxidation, at low temperature. Deprotection of the thioacetal group in 54 was effected by the treatment with N­

chlorosuccinin1ide (NCS)32) to give spiroacetal 55 as a single isomer. The orientation of the resulting hemiacetal hydroxyl group was determined as equatorial, based on 1 H NMR analysis of the anomeric proton (J = 9.5 Hz) of its diacetate 58. This configurational assignment was also supported from the observation that the acetyl group on C3 hydroxy group readily migrated to C9- hydroxy group (vide infra). This means that the C3-equatorial hydroxy group locates in the very vicinity of C9-hydroxy group. Spiroacetal 55 thus obtained was subjected to oxidation. Although

the oxidation of hemiacetal to lactone in general proceeds smoothly even in the presence of secondary aliphatic alcohol, all the attempts to selectively oxidized the lactol55 were un uccessfull.

Oxidation of 55 with pyridinium dichromate (PDC), AgC03-Celite, Swern, Pt-mediated molecular oxygen, or ozone, gave the unexpected 9-oxo con1pound 56 instead of the desired 53.

Therefore, the protection of C9-hydroxy group was examined prior to oxidation of the hemiacetal.

MeO

a

51

OBOM 54

oxidation

OBOM

55

a) Pb(0Ac)4, AcOK (99%) b) ^NCS (52%)

Scheme 12

0

H b

53

a b

''oAc 55

''oAc

58

57 59

Scheme 13

It was expected that diacety lation of 55 followed by selective reductive cleavage of C3- acetoxy group would afford the desired monoacetate 59, since C3-acetoxy group seemed sterically less hindered than C9-acetoxy group. In accord with this expectation, the acetylation of 55 firstly gave monoacetate 57 that was further acetylated to give diacetate 58. This supported that C9- hydroxy group was sterically more hindered than C3-hydroxy group (Scheme 13). _However, reductive cleavage of C3-acetoxy group did not give the desired 59 but the unnecessary 57 due to the presence of proximal C9-hydroxy group.

Thus, the author planned to synthesize lactone 53 by way of aldehyde 61, the C9-hydroxy group of which was protected as a t-butyldimethylsilyl (TBDMS) ether that is reluctant to migrate (Schen1e 14). After deprotection of the acetonide followed by oxidative cleavage of the resulting diol, 60 was protected as TBDMS ether 61. Treatment of 62 with DDQ followed by Swern oxidation, however, did not give the desired lactone but only 9-oxo compound 63 with the unexpected Inigration of TBDMS group.

MeO

s

�

OMPM 0

C

s

�J

_Meo

35

;---;

-:?"

8 a

OBOM 49 OBOM 60

0

b,c,d

.. ,OTBDMS e

OBOM 6 1

0 0

·.,,OTBDMS

f, g ^MeO

OBOM

6 2 OBOM

6 3

a) n-Buli, TMEDA (59%) b) PPTS, MeOH (91 %) c) Pb(Ac0)4, AcOK (95%) d) TBDMSCI, lmH (91 %) e) NCS (76%) f) DDQ, H20 g) Swern oxdn.

Scheme 14

These results indicated that C9-hydroxy group should be protected with a group that has no migratory ability and that remains intact during the removal of C 11-protective group. For this purpose, we chose MPM and 2,4-dinlethoxybenzyl (DMPM) ethers 16) as protective groups of C9-and C 11-hydroxy groups, respectively.

MeO OR

OBOM

(

^{39 R = H}

a ^{64 R} = DMPM

OBOM

g

OBOM

69

b

s

I

c

1 V I

^{3 _ �}

L'Y �b-1

^OBOM

3s

1 ---

OBOM 6 7

h

''OM PM

OBOM

(

^{65 R = H}

c ^{66 R} = MPM

0

68

''OM PM

70

a) DMPMCI, NaH (79%) b) n-Buli, TMEDA (99%) c) MPMCI, NaH (67%) d) PPTS, MeOH (98%) e) Pb(Ac0)4 (94%) f) NCS (86%) g) DDQ, H20 (71%) h) PDC (82%)

Schen1e 15

0

Along this line, cornpound 39 was protected as DMPM ether 64 (Scheme 15). Coupling of 35 and 64 was carried out in the same manner as previously described in the preparation of 50, to give 65 which was further protected as MPM ether 66. Selective deprotection of DMPM group in 68 was conducted with DDQ under Yonemitsu's conditions, 16) to afford spiroacetal 69. PDC oxidation of hemiacetal 69 finally afforded the desired spiro lactone 70.

The remaining problems for the present synthesis were the introduction of the C l-C2 and C22-C26 units and lactonization. At first, the introduction of C22-C26 unit was examined, since the resulting ester bond was expected to be sterically tnore crowded and, therefore, to be les reactive against nucleophiles than a 6-membered lactone, the nucleophilic opening of which was the following step for the introduction of the C l-C2 unit.

a

70

OBOM 71

b

_OBOM

0 OMPM HO

JVy

12 ^OBOM

a)

DDQ,

H20 (90%) b)

TCBC, Et3N, DMAP

(48%)

Scheme 16

o:cy

72

+

0

MPMO

�

OBOM

73

OBOM

Treatment of

70

with

DDQ

gave hydroxy lactone

71.

The condensation of

71

and

12

was effected by using Yamaguchi procedure35) which was proven to be the most efficient method for esterification and lactonization, to give the desired ester

72

along with a sn1all amount of unsaturated p-methoxybenzyl ester

73 (Scheme 16).

The undesired

73

was provably formed by the esterification of p-methoxybenzyl alcohol which was generated by P-elimination of MPM ether at the stage of mixed anhydride. Fortunately, no contamination of

74

was detected, probably due to the low reactivity of cx,p-unsaturated mixed anhydride

(Scheme 17).

At this stage, all the asymmetric centers required for the synthesis of 1 was settled.

0 0 OMPM

CI3C6H

�

^O

�

OBOM

olJ:;

_OBOM

72

0 0

)

^slow

)lJl�/--

CI3C6H3 ₀

... ^{"-� y}

)

_fast ^OBOM

MPMO

0

�

_OBOM

74 Scheme 17

0

MPMO

�

OBOM

With 72 in hand, the author examined the introduction of Cl-C2 unit. However, treatn1ent of 72 with Li enolate of ethyl acetate under various conditions afforded only a complex mixture (Scheme 18).

MeO CH=C(OLi)OEt

�

OBOM

72: R =

~

_OBOM ^OBOM

Scheme 18

Since all the attempts of two carbon elongation at the lactone moiety ended in a failure, the author decided to examine another approach to 1 starting from 60.

OBOM ₆₀

OMPM HOO�

12

OBOM

a

MeO MPMO

3

CHO

OBOM

.klY,,O OMPM

₂₇

0

75 d

(76

^{R,R' =}

n

2 R ⁼ R' ⁼ 0

a) TCBC, Et3N, then DMAP (85%) c) PPTS, MeOH d) Pb(Ac0)4, KOAc (77% for 2 ^steps) d) NCS, AgN03 (70%)

Scheme 19

OBOM

Condensation of the alcohol 60 and carboxylic acid 12 was again effected by using Yamaguchi method to give ester 75 which contained all the asymmetric centers in 1 except for two acetal carbons (Scheme 19). After acid hydrolysis of the terminal acetonide in 75, treatment of the resulting diol with lead tetraacetate gave aldehyde 76 which was an intermediate in Kishi's synthesis of 1. For the further structure confirmation, 76 was converted into another Kishi's

intermediate 2. Both 2 and 76 showed identical 1 H NMR spectra in every respect with those of the corresponding authentic samples.

Since 2 has been reported to be convertible to 1 in six steps,5) this accomplishment constitutes a formal total synthesis of optically active aplysiatoxin in a highly convergent and en anti oselecti ve manner.

In this synthesis, the complex architecture of 1 including many asymmetric centers have been constructed in a highly cnantioselective manner by using [2,3]Wittig rearrangement and titanium-mediated asymmetric epoxidation as key steps, demonstrating the usefulness of these methodologies for the synthesis of various types of natural products.

Chapter 4

Experimental

Experimental

NMR spectra were recorded at 400 MHz on a JEOL GX -400 or at 90 MHz on a JEOL

FX-90Q instrument in CDC}], unless otherwise noted. All signals were expressed as ppm down field from tetramethylsilane used as an internal standard (o-value in CDCl3). IR spectra were obtained with a JASCO IR-700 instrument. Optical rotation was measured with a JASCO DIP-360 automatic digital polarimeter. Column chromatography was conducted on Silica Gel 60, 70-230 mesh ASTM, available from E. Merck. Preparative thin layer chromatography was

performed on 0.5 mm x 20 em x 20 em E. Merck silica gel plate (60 F-2.54). Solvents were dried and distilled shortly before use. Reactions were carried out under an atmosphere of nitrogen if necessary.

( 2R, 3R) -1-(N -Phenylcarbamoyloxy)-2,3-epoxybutane (5).

To a suspension of MS 4A (20 g) in dichloromethane (700 ml) was added (-)-diisopropyl tartrate ( 4.30 g, 18.4 mmol). Titanium tetraisopropoxide ( 4.19 ml, 14.1 mmol) and t-butyl hydropcroxide (48 ml, 3.53 mol dm-3 in toluene, 170 mmol) were then added to the mixture at -20 °C. After stirring for 30 min, (E)-2-butenol (12.0 ml, 141 mmol) was added at the same temperature. After another 1 h, the mixture was left in refrigerator ( -20 °C) for 36 h. To the solution was added dimethylsulfide (5.8 ml, 79 mmol) and the reaction temperature was gradually raised to room temperature. To this solution were added triethylamine (23.4 ml, 168 mmol) and phenyl isocyanate (20.1 ml, 185 mmol), and the mixture was further stirred for 24 h. The mixture was treated with aqueous acetone (205 ml, acetone-HzO ⁼ 40: 1) and stirred for 12 h. The resulting precipitate was filtrated off and the filtrate was diluted with ethyl acetate. The organic layer was washed with water, dried over MgS04, filtrated through a short silica gel column, and concentrated to give 5 (21.4 g, 103 mmol, 73 %);

[aJ56

+46.8° (c 1.29, MeOH).

IR (KBr): 3280, 1730, 1597, 1547, 1499, 1439, 1307, 1222, 1052, 899, 860, 746cm-I. lH NMR (400 MHz): 7.39�7.33 (m, 2H), 7.32�7.29 (m, 2H), 7.07 (t, J ₌ 7.33 Hz, 1H), 6.90--6.75 (br s, 1H), 4.50 (dd, J ₌ 2.93, 12.21 Hz, 1H), 3.99, (dd, J ₌ 6.34, 12.21 Hz, lH), 3.01"'2.96, (m, 2H), 1.35, (d, J ₌ 5.40 Hz, 3H). Calcd. for C11H13NOJ: c,' 63.76; H, 6.32;

N, 6.76%. Found: C, 63.83; H, 6.19; N, 6.86o/o. Three recrystallizations of the product from hexane-ethyl acetate, gave optically pure carbamate 5 _as a crystalline [2.4 g,

[aJ56

+48.1°, (c 0.964, MeOH)].

( 2S, 3R)-1,2-Carbonyldioxybutan-3-ol (6).

Aqueous HCJ04 (166 ml, 5 %) was added to a solution of carbamate 5 (23.5 g, 114 mmol) in acetonitrile ( 150 ml) and the mixture was stirred at room temperature. After 24 h, saturated aqueous NaHCO] ( 100 ml) was added and bulk of acetonitrile was removed under reduced pressure. The residue was diluted with ether, washed with water, dried over MgS04, and concentrated. Silica gel column chromatography of the residue (hexane-ethyl acetate = 6:4) gave carbonate 6 (2.27 g, 17.2 mmol, 15 %) as an oil;

[a]b6

-5.4° (c 5.0, CHC]J). IR (neat):

3450, 2976, 1789, 1393, 1185, 1075, 772cm-l. lH NMR (90 MHz): 4.65--4.30 (m, 3H), 4.05 (dq, J = 6.56, 3.50 Hz, 1H), 2.38-2.04 (br s, 1H), 1.14 (d, J = 6.56 Hz, 3H). Calcd.

for CsHs04: C, 45.46; H, 6.10o/o. Found: C, 45.43; H, 6.12%.

( 2S, 3R)-3-Benzyloxymethoxy-1 ,2-carbonyldioxybutane (7).

To a solution of carbonate 6 (2.11 g, 16.0 mmol) and diisopropylethylamine (6.0 ml, 34 mmol) in dichloromethane (20 ml) was added benzyl chloromethyl ether (3.1 ml, 20 mmol) at room temperature. After stirring for 48 h, methanol (3 ml) was added to the solution and the mixture was stirred for another 12 h. The solution was concentrated under reduced pressure.

The resulting slurry was diluted with ether, washed with water, dried over MgS04, and concentrated. Silica gel column chromatography of the residue (hexane-ethyl acetate= 8:2-7:3) gave BOM ether 7 (2.82 g, 11.2 mmol, 70 o/o) as an oil;

[a]b6

+ 7.4° (c 1.2, CHCl3). IR (neat):

2928, 1796, 1451, 1375, 1171, 1022, 742, 697cm-l. 1H NMR (400 MHz): 7.38-7.29 (m, 5H), 4.83 (ABq, J = 6.84 Hz, 2H), 4.63 (ABq, J = 11.72 Hz, 2H), 4.59-4.56 (m, 1H), 4.47 (s, 1H), 4.45 (d, J = 1.96 Hz, IH), 4.09 (dq, J = 6.35, 3.90 Hz, IH), 1.22 (d, J = 6.35 Hz, 3H). Calcd. for Ct3Ht60s: C, 61.90; H, 6.39%. Found: C, 61.87; H, 6.42%.

( 2S, 3R)-3-Benzyloxymethoxybutane-1 ,2-diol (8).

KzC03 (1.0 g, 7.2 tnmol) was added to a solution of BOM ether 7 (1.41 g, 5.59 mmol) in methanol (30 ml) at room temperature. After stirring for 10 h, bulk of methanol was removed under reduced pressure. The residue was diluted with ether, washed with \Vater, dried over MgS04, and concentrated. Silica gel chromatography of the residue (hexane-ethyl acetate ⁼ 6:4-3:7) gave diol 8 (1.18 g, 5.21 mmol, 93 o/o) as an oil�

[a]f>6

-30.8° (c 3.10, CHCl3). IR (neat): 3404,2884, 1641, 1378, 1167, 1037, 740, 697cm-1. lH NMR (90 MHz): 7.54---7.26 (m, 5H), 4.80 (s, 2H), 4.61 (s, 2H), 4.05--3.49 (m, 4H), 2.99,.·-2.74 (br s, 1H), 2.54---2.23 (br s, 1H), 1.20 (d, J ₌ 5.78 Hz, 3H); Anal. Calcd. for C12H1804: C, 63.70; H, 7.91 %. Found:

C, 63.43; H, 8.02%.

( 2R, 3R)-3-Benzyloxymethoxy-1 ,2-epoxybutane (9).

To a solution of diol 8 (1.14 g 5.03 mmol), 4-dimethylaminopyridine (100 mg, 0.82 mmol), and triethylamine ( 1.6 ml, 12 mmol) in dichloromethane (20 ml) were added pivaloyl chloride (630 �tl, 5.12 mtnol) at room temperature. After stirring for 10 h, methanesulfonyl chloride (430 !J.l, 5.6 mmol) was added and the mixture was stirred for another 1 h. The solution was concentrated under reduced pressure, diluted with ether, and washed with water.

The organic layer was separated, dried over MgS04, filtrated through a pad of silica gel, and concentrated to give the corresponding mesylate (1.29 g, 3.32 mmol, 66 %) as an oil; lH NMR (90 MHz): 7.36-7.16 (br s, 5H), 4.8&---4.68 (m, 1H), 4.73 (s, 2H), 4.56 (s, 2H), 4.42,..,3.92 (m, 3H), 2.98 (s, 3H), 1.21 (d, J ₌ 6.56 Hz, 3H), 1.15 (s, 9H).

The mesylate (360 mg, 0.927 mmol) was dissolved in dichloromethane (10 ml) and cooled to -78 °C. To the solution was added diisobutylaluminum hydride ( 1.9 ml, 1.0 mol dm-3 in hexane, 1.9 mmol). After stirring for 1 h, the mixture was quenched with methanol (0.5 ml) at the temperature. After additional 5 min, the solution was gradually warmed to room temperature. At this point, the solution became a white gel. To the gel, saturated aqueous potassium sodium tartrate ( 10 ml) was added and the whole mixture was left with stirring until it became a clear solution. The resulting solution was extracted with ether, dried with MgS04, filtrated through a pad of silica gel, and concentrated to give 13-mesyloxy alcohol (271 mg, 0.890 mmol, 96 o/o) as an oil; lH NMR (400 MHz): 7.38--7.30 (m, 5H), 4.'81 (s, 2H), 4.69

(ddd, J = 5.85, 3.90, 3.90 Hz, 1H), 4.64 (ABq, J = 11.72 Hz, 2H), 4.07 (dq, J = 3.90, 6.53 Hz, 1H), 3.88�3.84 (m, 2H), 3.10 (s, 3H), 2.58-2.45 (br s, 1H), 1.27 (d, J = 6.35 Hz, 3H).

The P-mesyloxy alcohol (707 mg, 2.32 mmol) was added to the mixture of methanol (10 ml) and aqueous KOH ( 4.6 ml, 1.0 mol dm-3, 4.6 mmol), and stirred at room temperature.

After 1 h, bulk of methanol was removed under diminished pressure. The resulting solution was extracted with ether, dried over MgS04, concentrated, and chromatographed on silica gel (hexane-ethyl acetate= 8:2) lo give cpoxidc 9 (450 mg, 2.16 mmol, 93 o/o) as an oil�

[aJ56

+24.0° (c 5.38, CHC]J). IR (neat): 3026, 2882, 1449, 1376, 1039, 738, 697cm-1. lH NMR (90 MHz): 7.41,..,7.26 (n1, 5H), 4.88 (ABq, J = 6.90 Hz, 2H), 4.65 (s, 2H), 3.58 (dq, J = 6.56, 6.56 Hz, 1H), 3.02 (ddd, J = 2.62, 5.30, 6.56 Hz, 1H), 2.78 (br t, J = 5.03 Hz, 1H), 2.56 (dd, J = 2.62, 5.03 Hz, 1H), 1.27 (d, J = 6.56 Hz, 3H). Calcd. for C12H160J: C, 69.21 � H, 7.74%. Found: C, 69.04� H, 7.67%.

( 3R, 4R)-4-Benzyloxymethoxy-3-hydroxy-1,1-propylenedithiopentane (1 0).

Butyllithium (3.5 ml, 1.6 mol dm-3 in hexane, 5.6 mmol) was added to a solution of 1,3- dithiane (678 mg, 5.64 mmol) in THF (20 ml) at 0 °C. After stirring for 1 h, a solution of epoxide 9 (783 mg, 3.76 mmol) in THF (10 ml) was added to the mixture at the temperature.

After stirring for another 10 h, the mixture was quenched with aqueous H3P04 ( 10 ml, 5 %) and allowed to warm to room temperature. The solution was extracted with ether, dried over MgS04, and concentrated. Silica gel column chromatography of the residue (hexane-ethyl acetate= 8:2�:4) gave thioacetall 0 (1.23 g, 3.74 mmol, 99 %) as an oil�

[aJ56

+3.6° (c 0.73, CHCl3). IR (neat): 3460, 2892, 1378, 1275, 1101, 1038, 739, 699cm-1. lH NMR (400 MHz): 7.38,..,7.26 (m, 5H), 4.84 (ABq, J = 7.33 Hz, 2H), 4.64 (ABq, J = 11.72 Hz, 2H),4.33 (dd, J = 4.88, 9.76 Hz, 1H), 3.83,..,3.77 (m, 1H), 3.64 (dq, J = 6.35, 6.35 Hz, 1H), 2.97,..,2.82 (m, 4H), 2.63 (d, 1= 4.88 Hz, 1H), 2.16,..,2.09 (m, 1H), 1.97,..,1.83 (m, 3H), 1.22 (d, J = 6.35 Hz, 3H). Calcd. for C16H240JS2: C, 58.50� H, 7.38%. Found: C, 58.54� H, 7.35o/o.

... •""""-.. .... �.-��.-.-., .,..,. .. -•• _,._ r'"''"" ���,.,�· ... .,..,.,..., .. � • , ' , • •• • .•' ... ' •• ,.

( 3R, 4R)-4-Ben zy ^I ox y met h ox y-3- (p- met h ox y benz y ^I ox y) -1, 1-prop y ^I en e d i­

thiopentane (11).

To a suspension of thioacetal 10 (1.23 g, 3.74 mmol) and NaH (180 mg, 60 o/o in mineral oil, 4.5 mmol) in THF-DMF (40 ml, 3: 1) was added p-methoxybenzyl chloride (560 �tl, 4.1 mmol) at room temperature. After stirring for 24 h, the mixture was quenched with aqueous H3P04 (20 ml, 5%), extracted with ether, dried over MgS04, and concentrated. Silica gel

chromatography of the residue (hexane-ethyl acetate= 9: 1,-7:3) gave MPM ether 11 ^{( 1.38 g,} 3.08 mmol, 82 %) as an oil;

[a]E/

^{+18.3 o} _(c ^0.731, CHCb). IR (neat): 2890, 1608, 1510, 1453, 1246, 1038, 821, 739, 699cn1-l. lH NMR (90 MHz): 7.48-7.26 (m, 5H), 7.28 (d, 1 = 13.61 Hz, 2H), 6.82 (d, 1 = 13.61 Hz, 2H), 4.75 (s, 2H), 4.56 (s, 2H), 4.21-3.75 (m, 3H), 3.70 (s, 3H), 3.04-2.68 (m, 4H), 2.24-1.80 (m, 4H), 1.18 (d, 1 ⁼ 9.72 Hz, 3H). Calcd. for C24-fi3204S2: C, 64.25; H, 7.19%. Found: C, 64.16; H, 7.11o/a.

(3R,4R)-4-Benzyloxymethoxy-3-{p-methoxybenzyloxy)pentanoic acid (12).

Methyl iodide (3.50 �1, 5.6 mmol) was added to a suspension of MPM ether 11 ^{( 491 mg,}

1.10 mmol) and calcium carbonate (1.1 g, 11 mmol) in aqueous acetonitrile (10 ml, acetonitrile­

H20 = 4: 1). After stirring for 10 h at room temperature, the mixture was extracted with ether, dried over MgS04, and concentrated. Silica gel chromatography of the residue (hexane-ethyl acetate= 8:2) gave the corresponding aldehyde (256 mg, 0.714 mmol, 65 %) as an oil; lH NMR ( 400 MHz): 9.74, (t, 1 = 1.46 Hz, 1H), 7.38-7.21 (m, 5H), 7.22 ( d, 1 = 8.30 Hz, 2H), 6.85 (d, 1 = 8.30 Hz, 2H), 4.78 (ABq, 1 = 7.32 Hz, 2H), 4.59 (s, 2H), 4.52 (ABq, 1 = 11.23 Hz, 2H), 4.04--3.97 (m, 2H), 3.79 (s, 3H), 2.68-2.62 (m, 2H), 1.19 (d, 1= 5.86 Hz, 3H).

To a mixture of the aldehyde (66.0 mg, 0.184 mmol), t-butanol (1 ml), saturated aqueous NaH2P04 (1 ml), and 2-methylbutene (100 �1, 0.95 mmol) were added NaCl02 (20 mg, 0.22 mmol) at 0 °C. After 5 min, aqueous H3P04 (5 ml, 5 %) was added and the solution was extracted with ether, dried over MgS04, and concentrated. Silica gel chromatography of the residue (CHCl3-methanol = 200:8) gave carboxylic acid 12 ^(56.5 mg, 0.151 mmol, 82 %) as ^an oil; IR (neat): 2932, 1709, 1609, 1511, 1247, 1037, 822, 740, 699cm-1. lH NMR (90 MHz):

7.46-7.18, (m, 5H), 7.18 (d, 1 = 8.86 Hz, 2H), 6.78 (d, 1 = 8.86 Hz, 2H), 4:73 (s, 2H), 4.53

(s, 2H), 4.47 (s, 2H), 4.02�--3.70 (m, 2H), 3.70 (s, 3H), 2.65--2.45 (m, 2H), 1.10 (d, J = 5.82 Hz, 3H).

m-Benzyloxymethoxycinnamyl alcohol (14).

Diisobutylaluminum hydride (70 ml, 1.0 mol dm-3 in hexane, 70 mmol) was added to a solution of ester 13 (10.3 g, 33.0 mmol) in dichloromethane ( 160 ml) at -78 ^oc and the mixture was stirred for 1 h at the temperature. Methanol (10 ml) was added to this solution and the whole mixture was stirred for another 5 min. The solution was gradually warmed to room temperature. At this point, the solution became a white gel. To the gel was added saturated

aqueous potassium sodium tartrate (200 ml) and the whole mixture was left until it became a clear solution. The solution was extracted with ether, dried over MgS04, and concentrated to

give allylic alcohol 14 (8.17 g, 30.2 mmol, 92 %) as an oil; IR (neat): 3374,3026, 2896, 1577, 1087, 1018, 773,741, 694cm-l. lH NMR (400 MHz): 7.37-7.23, (m, 6H), 7.11 (t, J ₌ 1.95 Hz, 1H), 7.05 (br d, J ₌ 7.82 Hz, 1H), 6.98 (dd, J = 1.95, 8.32 Hz, 1H), 6.59 (br d, J ₌ 16.12 Hz, 1H), 6.36 _(ddd, ^J = 5.37, 5.86, 16.12 Hz, 1H), 5.30 _(s, 2H), 4.73 _(s, 2H), 4.32 (d, J ₌ 5.86 Hz, 1H), 4.31 _(d, ^J ₌ 5.37 Hz, 1H), 1.50'"" 1.44 _{(br s,} 1H). HREIMS m/z calcd.

for C17HisG.3: 270.12549, _found 270.12573 (M+).

(2R, 3R)-3-(m-Benzyloxymethoxyphenyl)-2,3-epoxypropan-1-ol (15).

To a suspension of (-)-diisopropyl tartrate ( 1.4 ml, 6.6 mmol) and powdered MS 4A ( 1.7 g) in dichloromethane (120 ml) were added titanium tetraisopropoxide ( 1.7 ml, 5.7 mmol) and t-butyl hydroperoxide (31 _ml, 3.7 mol dm-3 in toluene, 110 _{mmol) at} -20 °C. After stirring for 30 min, a solution of ally lie alcohol 14 ( 15.7 _g, 58.1 mmol) in dichloromethane (20 _{ml) was} added at the temperature. After another 1 h, the mixture was left in refrigerator ( -20 °C) _for 10 h. The mixture was quenched with pre-cooled ( -20 °C) aqueous acetone (60 ml, acetone-HzO = 5: 1), and the reaction temperature was gradually raised to room temperature. After stirring for 3 h, the resulting precipitate was filtered off and the filtrate was concentrated. Silica gel column chromatography of the residue (hexane-ethyl acetate= 6:4) gave epoxy alcohol 15 ( 14.8 _g, 51.7 mmol, 89 %) as an oil;

[aJl}

+26.8° (c 0.821, CHCI)). IR _(neat): 3440, 2

9oo

, 1586, 1489,

1233, 1158, 1088, 789, 742, 6⁹⁸cm-l^. lH NMR (400 MHz): 7.37-7.24 (m, 6^H), 7.05-6.97 (m, 2H), 6.95 (d, J = 7.33 Hz, 1H) 5.29 (ABq, J = 7.33 Hz, 2H), 4.72 (s, 2H), 4.04 (ddd, J

= 2.44, 4.88, 12.70 Hz, 1H), 3.91 (d, J = 1.95 Hz, 1H), 3.79 (ddd, J = 3.90, 7.80, 12.70 Hz, 1H), 3.20 (ddd, J = 1.95, 2.44, 3.90 Hz, 1H), 1.78 (dd, J = 4.88, 7.82 Hz, 1H). Calcd.

for ^{Ct 7Ht s}04^: C, 71.31; H, 6.34%. Found: C, 71.24; H, 6.30%.

(S) -3- (m-Benzyloxymethoxyphenyl)propane-1,3-diol (16).

Red-al ( ¹⁰ n1l, ^3.6 mol dm-3 in toluene, ³⁶ mmol) was added to a stirring solution of epoxy alcohol15 (6.57 g, 22.9 mmol) in THF (50 ml) at 0 °C and stirred for 10 min. Then the mixture was left in refrigerator ^{(0 °C)} for ¹⁰ h. The mixture was quenched with aqueous NaOH ( ¹⁰ ml, ¹⁵ %),allowed to warm to room temperature, and poured into water. The mixture was extracted with ether, dried with MgS04, and concentrated. The residue was dissolved in aqueous THF (50 ml, THF-HzO = 1: 1) and to this solution was added Nai04 ( ^1.0 g, ^4.7 mmol) at room temperature. After vigorous stirring for ³ h, the mixture was extracted with ether, dried over MgS04, and concentrated. Silica gel column chromatography of the residue (hexane-ethyl acetate= 1:1-3:7) gave 1,3-diol 16 ^(6.00 g, 20.8 mn1ol, ^{91 %)} as an oil;

[aJb9

^-29.1° (c 1.33, CHCl3). IR (neat): 3402, 2938, 1585, 1239, 1158, 1018, 788, 742, 699cm-1. ^lH NMR (400 M^Hz)^{: 7.36-7.25} (m, 6H), ^7.10 (br s, 1H), 7.03--7.00 (m, ^{2H) 5.30} (s, 2H), 4.96-4.93 (m,

1 H), 4.73 (s, 2H), 3.86 (br t, J = 5.62 Hz, 2H), 2.90--2.70 (br s, lH), 2.35-2.20 (br s, ^{1 H),} 2.06-1.90 (m, ^2H). Calcd. for ^{C1 7Hzo}04^: C, 70.81; H, 6^.99^%. Found: C, 70.86; H, 6.77%.

(S) -3- (m-Benzyloxymethoxyphenyl)-1-bromo-3-methoxypropane (1 7).

p-Toluenesulfonyl chloride ^(5.1 g, ²⁷ mmol) was added to a solution of diol 16 ^(7.33 g, 25.4 mmol), 4-dimethylaminopyridine ⁽¹ 00 mg, ^0.82 mmol), and triethylamine ( ^4.3 ml, ³¹ mmol) in dichloromethane (250 ml) at room temperature. After ⁴ h, bulk of dichloromethane

was removed under diminished pressure, and the residue was diluted with ether, washed with water, dried over MgS04, filtrated through a pad of silica gel, and concentrated to give tosylate (10.5 g, ^23.7 mmol, ^{93 %)} as an oil.

To a solution of the above tosylate (6.34 g, 14.3 mmol) and methyl iodide (2.5 ml, 40 mmol) in DMF-THF (120 ml, 3: 1) was added sodium hydride (900 mg, 60 % in mineral oil, 23 mmol) at room temperature. After stirring for 5 h, the mixture was quenched with aqueous H3P04 ( 10 ml, 5 % ). The mixture was poured into water and extracted with ether. The organic layer was dried over MgS04, concentrated, and diluted with DMF (50 ml). To the solution was added NaBr (11 g, 110 mmol) and the mixture was stirred at room temperature for 2 d, then poured into water (300 ml), and extracted with hexane-ethyl acetate (8:2). The organic layer was dried over MgS04, concentrated, and chromatographed on silica gel (hexane-ethyl acetate=

9:1,...,8:2) to give bromide 17 ^(3.63 g, 9.94 mmol, 70 %) as an oil�

[aJ5°

^-46.1° (c ^3.63, CHCl3). IR (neat): 2896, 1586, 1482, 1449, 1241, 1156, 1091, 1021, 788, 737, 697cm-l.

lH NMR (400 ^{MHz): 7.3�7.26} (m, 6H), 7.04-7.02 (m, 2H), 6.96 (d, 1 = 7.31 Hz, 1H), 5.31 (s, 2H), 4.75 (s, 2H), 4.31 (dd, J = 4.39, 7.81 Hz, 1H), 3.55 (ddd, 1= 5.86, 8.30, 9.76

Hz, 1H), 3.37 (ddd, J = 5.86, 5.86, 9.76 Hz, lH), 3.25 (s, 3H), 2.32,...,2.25 (m, lH), 2.13,...,2.05 (m, 1H). Calcd. for C1gH210JBr: C, 59.19� H, 5.79%. Found: C, 59.15� H, 5.79o/a.

Isopropyl [ ( 1R, 2E)-1-benzyloxymethyl-2-butenyloxy ]acetate (2 0).

To a stirred mixture of the allylic alcohol 1 ^{9 ( 1.05} g, 5.46 mmol) and sodi urn hydride (0.670 g, 60 % in mineral oil, 17 mmol) in THF (9 ml) was added dropwise a solution of bromoacetic acid (0.823 g, 5.92 mmol) in THF (9 ml). The mixture was refluxed for 12 h, cooled to room temperature, poured into water, and extracted with ether. The aqueous layer was adjusted to pH 1 and extracted with dichloromethane. The organic layers were combined, dried, and concentrated. The residue was added to a solution of Na2C03 (0.360 g, 3.40 mmol) and water (6 drops) in hexamethylphosphoric triamide (6 ml) and stirred for 5 min. After isopropyl iodide (0.810 ml, 8.11 mmol) was added, the mixture was further stirred for 12 h. The reaction mixture was then poured to water, extracted with hexane, dried, and concentrated. Column chromatography of the residue (hexane-ethyl acetate = 5: 1) gave the ester 2 0 ^{( 1.20} g, 4.10 mmol, 75 o/o) as an oil� lH NMR (90 MHz): 7.03 (s, 5H), 6.0(}-4.86 (m, 2H), 4.58 (s, 2H), 4.05 (s, 2H), 4.02 (m, lH), 3.6�3.34 (m, 2H), 1.71 (d, 1 = 5.4 Hz, 3H)·, 1.�1.69 (m,

lH), 1.23 _{(d, 1} = 6.3 Hz., 6H). Calcd for C17H240Lr C, 69.84; _H, 8.27o/o. Found: C, 69.64;

H, 8.32%.

Isopropyl (2S, 3R, 4E)-6-Benzyloxy-2-hydroxy-3-methyl-4-hexenoat e (21).

A solution of ester 2 0 (2.14 _g, 7.32 mmol) in THF (5 ml) was added dropwise to a solution of LDA (9.45 _n1l, 0.815 mol dm-3 in THF-hexane = 1:1) _at -100 °C. After 1h, a solution of Cp2TiCb (2.37 _g, 9.52 mmol) in THF ( 100 ml) was added to the mixture at the same temperature. After another 15 min, the reaction temperature was gradually raised to -20 oc and the mixture was kept standing in refrigerator ( -20 _{°C) for} 19 h. The mixture was quenched with a saturated aqueous solution of KF (3.6 ml) and allowed to warm to room temperature.

The mixture was then filtered through a pad of Celite and concentrated in vacuo. Column chromatography of the residue (hexane-ethyl acetate= 5: 1) gave hexenoate 21 ( 1.54 _g, 5.27 mmol, 72 o/iJ) as an oil; lH NMR (400 MHz): 7.38--7.26 _(m, 5H), 5.77 _(dd, J ₌ 15.6, 6.8 Hz, 1H), 5.70 _(dt, J _{= 1}5_.6, 5.4 Hz, 1H), _{5.11 (m,} 1H), 4.12 (br s, 1H), 4.01 (d, J ₌ 5.4 _Hz, 2H), 2.80 (m, 1H) , 2.67 _(m, lH) , 1.29 _(d, J ₌ 5.4 Hz, 3H) , 1.28 (d, J ₌ 5.9 Hz, 3H), 1.01 (d, J ₌ 6.8 Hz, 3H). Calcd for C17H2404: C, 69.84; _H, 8.27%. _{Found: C,} 69.67; H, 8.25%.

( 2S, 3S, 4S, SR) -6-Benzyloxy-2-hydroxy-5-iodo-3-methyl hexan -4-olide (2 2).

Aqueous (X)tassium hydroxide (4.4 _ml, 1.0 mol dm-3, 4.4 mmol) was added at room temperature to a solution of hexenoate 21 (0.430 _g, 1.47 mmol) in methanol (14 _{ml). After} 1 day, a bulk of methanol was removed under diminished pressure. The residual solution was diluted with water, adjusted to pH 4 by using aqueous H3P04 (5 %), and extracted with dichloromethane. The organic layer was concentrated under vacuum and diluted with acetonitrile (20 _ml). To this solution was added I2 ( 1.12 g, 4.41 mmol) and the mixture was stirred at 0 _{oc for} 18 _h. The mixture was decolorized with aqueous Na2S203, extracted with ether, dried over Na2S04, and concentrated. Column chromatography of the residue (hexane­

ethyl acetate= 7:3) gave the icx:lolactone 22 (0.345 _g, 0.917 _mmol, 62 o/o) as an oil; 1H NMR (90 MHz): 7.32 _(s, 5H), 4.54 _(s, 2H), 4.51-4.35 _(m, lH), 4.32-3.95 _(m, 2H), 3.94-3.70