Asymmetric syntheses of pectenotoxins-4 and -8, Part I: Synthesis of the C1-C19 subunit

Article abstract of DOI:10.1002/1521-3773(20021202)41:23<4569::AID-ANIE4569>3.0.CO;2-V

The total syntheses of the marine macrolides pectenotoxins-4 and -8 have been realized. The pectenotoxins are a new class of marine macrolides that exhibit a high level of biological activity. Key fragment couplings include a Felkin-selective Grignard add

Full text of DOI:10.1002/1521-3773(20021202)41:23<4569::AID-ANIE4569>3.0.CO;2-V

COMMUNICATIONS
[1] D. C. Neckers, J. Chem. Educ. 1975, 52, 695.
[2] E. Atherton, M. J. Gait, R. C. Sheppard, B. J. Williams, Bioorg. Chem.
1979, 8, 351.
[3] G. Barany, R. B. Merrifield, Peptides 1980, 2, 1.
[4] T. T. Nguyen, U. Asseline, Biochimie 1985, 67, 673.
[5] K. N. Ganesh, Curr. Sci. 1985, 54, 103.
[6] R. B. Merrifield, Science 1986, 232, 341.
[7] G. Barany, N. Kneib-Cordonier, D. G. Mullen, Int. J. Pept. Protein Res.
1987, 30, 705.
[8] A. R. Brown, P. H. H. Hermkens, H. C. J. Ottenheijm, D. C. Rees,
Synlett 1998, 8, 817.
[9] B. Clapham, T. S. Reger, K. D. Janda, Tetrahedron 2001, 57, 4637.
[10] A. M. Harned, P. R. Hanson, Org. Lett. 2002, 4, 1007.
[11] S. Booth, P. H. H. Hermkens, H. C. J. Ottenheijm, D. C. Rees,
Tetrahedron 1998, 54, 15385.
5’-amino and 3’-acetaldehyde groups remain at the end of the
reaction.
22]
As in the polymerase chain reaction (PCR),[20 S(dAp)8
templates the ligation of multiple monomers in a single
reaction cycle. Unlike PCR, where both strands of the DNA
duplex are amplified to give exponential growth, the DNA-
templated polymerization reactions employ only a single
strand as the template and growth is presumably linear with
each reaction cycle. In addition, the requirement for primers
compatible with the double-strand-binding polymerase is
avoided, and short sequences of DNA are amplified effi-
ciently.
Finally, the reaction does not synthesize native DNA, but a
backbone analogue. Therefore, solid-supported oligomeric
DNA can be used to catalyze the rapid synthesis of polymers
containing different backbones simply by changing the
structure of the reactant, in this case (T)1 and (TN)2, Quite
unlike other solid-supported syntheses, S(dAp)n can be used
repeatedly both to catalyze the polymerization as well as
purify the product, which greatly reduces the time and effort
for the synthesis of modified DNA-analogues. The extension
of this chemistry to mixed-sequence templates should enable
the rapid amplification of DNA sequence information into
specific backbone-modified analogues. Moreover, this general
strategy for solid-phase synthesis can now be extended more
broadly, through other molecular recognition elements, to
accomplish chain-length-specific polymerizations.
[12] B. Doerner, R. Steinauer, P. White, Chimia 1999, 53, 11.
[13] C. Saluzzo, R. ter Halle, F. Touchard, F. Fache, E. Schulz, M. Lemaire,
J. Organomet. Chem. 2000, 603, 30.
[14] A. Kirschning, H. Monenschein, R. Wittenberg, Angew. Chem. 2001,
113, 670; Angew. Chem. Int. Ed. 2001, 40, 650.
[15] S. J. Shuttleworth, S. M. Allin, P. K. Sharma, Synthesis 1997, 11, 1217.
[16] K. C. Hultzsch, J. A. Jernelius, A. H. Hoveyda, R. R. Schrock, Angew.
Chem. 2002, 114, 609; Angew. Chem. Int. Ed. 2002, 41, 589.
[17] J. S. Kingsbury, S. B. Garber, J. M. Giftos, B. L. Gray, M. M. Okamoto,
R. A. Farrer, J. T. Fourkas, A. H. Hoveyda, Angew. Chem. 2001, 113,
4381; Angew. Chem. Int. Ed. 2001, 40, 4251.
[18] X. Li, Z.-Y. J. Zhan, R. Knipe, D. G. Lynn, J. Am. Chem. Soc. 2002,
124, 746.
[19] Z.-Y. J. Zhan, D. G. Lynn, J. Am. Chem. Soc. 1997, 119, 12420.
[20] R. K. Saiki, S. Scharf, F. Faloona, K. B. Mullis, G. T. Horn, H. A.
Erlich, N. Arnheim, Science 1985, 230, 1350.
[21] H. A. Erlich, D. Gelfand, J. J. Sninsky, Science 1991, 252, 1643.
[22] T. J. White, Trends Biotechnol. 1996, 14, 478.
[23] A. Luther, R. Brandsch, G. von Kiedrowski, Nature 1998, 396, 245.
Experimental Section
DNA Synthesis: All native DNA oligomers were prepared by the Emory
University Microchemical Facility on a PE-Biosystems 394 DNA Synthe-
sizer. The DNA S(dAp)8 template was synthesized on OAS-PS (Glen
Research, Batch No. G008062, Cat. No. 26-4001) solid supports by standard
cyanoethyl phosphoramidite chemistry. The linker of OAS-PS is stable to
the last step of ammonium hydroxide deprotection treatment in the
automated synthesis. To confirm purity, the DNA oligomers were removed
from the resin and analyzed by Rainin HPXL RP-HPLC: Phenomenex
Prodigy 5 analytical ODS(2) C18; Rainin Dynamax UV detector at
260 nm, and eluted with MeOH in H2O (0 100% in 50 min), and the
structure confirmed by MALDI-MS: C80H97N40O38P7 calcd m/z: 2443.69
[MþHþ]; found 2444.87.
Polymerization: The substrates, 8 mm for (TN)2 and 16 mm for (T)1, were
mixed with S(dAp)8 at the indicated stoichiometry, heated to 758C for 2 3
min, and cooled to 48C for 3 h.[23] NaBH3CN (20 equiv) was added at room
temperature and the mixture stirred vigorously in a vortex mixer for 72 h,
with the addition of more NaBH3CN (10 equiv) after 48 h to ensure complete
reaction. The mixture was subjected to centrifugation in the Ultrafree-MC
(Millipore Corp.) centrifugal filter tube at room temperature to remove the
reagents, and the beads resuspended in MeOH, heated to 758C, and subjected
to centrifugation whilst hot, with 3î1 mL MeOH. The MeOH washes were
combined, evaporated to dryness, and analyzed by reverse-phase HPLC
(same system as above), and eluted with MeOH in H2O: 0 5% from 0 8 min,
5 20% from 8 9 min, 20 35% from 9 24 min, 35 60% from 24 25 min, 60
100% from 25 50 min. The product assignments were justified by co-
injection with standard samples and by mass spectrometry. MALDI-TOF
(2’,4’,6’-trihydroxyacetophenone monohydrate (THAP)/citrate): (TN)2-[NH-
(TN)2]3-CHO, C96H128N24O29 calcd m/z: 2080.9279 [Mþ]; found 2082.85
[Mþ2H]þ; (TN)2-[NH-(TN)2]3-CH2OH, C96H130N24O29 calcd m/z: 2082.9436
[Mþ]; found 2083.84 [MþH]þ; (T)1-[NH-(T)1]7, C96H136N24O25, calcd m/z:
2025.0109 [Mþ]; found 2026.12 [Mþ H]þ.
Asymmetric Syntheses of Pectenotoxins-4 and
-8, Part I: Synthesis of the C1 C19 Subunit**
David A. Evans,* Hemaka A. Rajapakse, and
Dirk Stenkamp
The first members of the pectenotoxin family of marine
natural products were isolated off the northeastern coast of
Japan in 1985. Subsequently, ten members of this group have
been identified.[1] The structural diversity within the pecte-
notoxins stems from variations in the oxidation state at C43,
as well as the differing configurations of the AB spiroketal
portion of the structures. Pectenotoxin-2 (C43 ¼ Me) is
cytotoxic towards human lung, colon, and breast cancer cell
lines with LC50 values in the nanomolar range.[1c] Pectenotox-
[*] Prof. D. A. Evans, H. A. Rajapakse, D. Stenkamp
Department of Chemistry & Chemical Biology
Harvard University
Cambridge, MA 02138 (USA)
Fax : (þ 1)617-495-1460
E-mail: evans@chemistry.harvard.edu
[**] Financial support has been provided by the National Institutes of
Health (GM33327-16). A BASF-Forschungsstipendum Postdoctoral
Fellowship to D.S. and a Howard Hughes Medical Institute Predoc-
toral Fellowship to H.A.R. are gratefully acknowledged. The NIH
BRS Shared Instrumentation Grant Program 1-S10RR04870 and the
NSF (CHE 88-140019) are acknowledged for providing NMR
facilities.
For resin recovery, the beads were washed 3 î 1 mL with deionized H2O
and resuspended in H2O overnight under vortex mixing before being used
for the next reaction cycle.
Received: June 5, 2002 [Z19473]
Angew. Chem. Int. Ed. 2002, 41, No. 23
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4123-4569 $ 20.00+.50/0
4569

COMMUNICATIONS
ins-1, -4, and -8 (C43 ¼ CH2OH) are differentiated by their
AB spiroketal subunits that are interconvertable by acid
equilibration (Figure 1, Eq. (1)).[1b] While no total synthesis of
intermediate I, a synthesis subgoal that might be assembled
from fragments II and III through a metalloenamine-epoxide
alkylation.[4] The synthesis plan calls for masking the ketones
at C14 and C36 as protected alcohols until late in the
synthesis. Protection of the carboxylate terminus at C1 as its
derived N-phenylamide would enable the preservation of the
C1 carboxyl carbon atom in its correct oxidation state
throughout the synthesis.[5] In the retrosynthesis of the ABC
tricyclic fragment II, convenient disconnections could be
made about the C16 C17 and C11 C12 bonds (Scheme 1).
The stereocenters at C11 and C16 could be constructed by
chelate-controlled reduction of the corresponding ketone, or
by Felkin-controlled addition of an appropriate nucleophile to
the respective aldehydes.
The integration of catalytic enantioselective processes into
our synthesis projects has been an ongoing objective. In this
regard, the enantioselective Sn2þ-catalyzed aldol reaction
[Eq. (2)] provides the foundation for the synthesis of the C8
C11 acetate and C36 C39 propionate fragments.[6,7] The
asymmetric Michael addition reaction reported from these
laboratories provides a useful alternate route for the con-
struction of the C36 C40 propionate fragment [Eq. (3)].[8]
42
O
H
Me
C14
10
H
Pectenotoxin-4 (1)
11
B
O
H
O
O
Me
7
O
43
OH
H
OH
O
CH2OH
H
O
1
47
G
H
D
H
36
A
Me
O
H
E
33
O
O
20
O
H
40
30
HO
H
H
10
B
Me
Me
Me
O
OH
O
45
C14
7
11
A
Me
H
O
O
Pectenotoxin-8 (2)
H
H
O
43
H
H
3
OH
O
CH2OH
O
1
G
47
H
D
H
36
O
Me
Me
33
E
O
O
20
O
H
40
HO
30
H
H
Me
Me
Me
45
AB Spiroketals
HO
H
R
H
H
10
11
10
10
B
R
11
OH
B
H+
H+
O
R
O
11
7
O
(1)
B
O
3
7
OH
A
O
H
O
7
H
A
H
R'
3
3
H
A
R'
Me Me
R'
Pectenotoxin-8 (57%)
OTMS
O
O
Pectenotoxin-1 (29%)
Pectenotoxin-4 (14%)
O
OH
N
N
PhS
Figure 1. The pectenotoxin family of natural products, and their inter-
conversion by acid equilibration [Eq. (1)].
Sn
R'
R'
OEt
R
(2)
O
TfO OTf
PhS
+
OEt
R
O
H
R = H (3), R' = Bn 86% yield, 98% ee
R = Me (4), R' = Ph, 97% yield, 95% ee, 94:6 d.r.
any pectenotoxin has been reported, there have been efforts
directed toward the synthesis of these architecturally complex
natural products.[2] In this and the following communication,[3]
we describe our efforts culminating in the synthesis of
pectenotoxin-4 (1).
Pectenotoxin-4 was chosen as our initial target. Although
this spiroketal is the most stable isomer in acyclic precursors it
becomes the least favored when constrained in the macrolide
framework.[1b,d] Our principal disconnections are illustrated in
Scheme 1. Application of the macrolactonization transform
and ring opening of the D-ring bicyclic ketal reveals
O
2+
Me Me
O
O
–
2 SbF6
N
N
N
RO
O
OR
Cu
tBu
tBu
OTMS
+
O
N
N
(3)
O
Me
O
N
Me
O
R = Bn (5) 87% yield, 98% ee, 98:2 d.r.
R = TBS (6), 75% yield, 95% ee, 98:2 d.r.
O
O
14
O
Pectenotoxin-4 (1)
H
Me
OP
Me
14
C
H
H
B
H
Me
7
O
O
O
OP
C
O
B
Macrolactonization
O
7
O
H
Me
1
OH
H
A
O
O
OH
O
CH2OH
NHPh
H
H
O
1
D
H
OP
H
A
36
Me
O
H
E
O
OH
33
O
20
19
II
O
O
G
19
HO 40
Epoxide
H
H
Me
Me
alkylation
Me
Me
17
42
H
OP
O
7
O
O
OP
B
C19–C20
X
11
+
+
1
19
12
A
16
X
OP
O
Br
Me
NHPh
H
14
SnBu3
H
O
OP
Me
OP
H
7
O
O
C
B
OH
OH
1
A
O
O
NHPh
H
+
CH2OH
OP
OP
I
19
NMe2
33
33
PO
N
PO
20
Me
H
Epoxide
alkylation
Me
H
F
F
O
36
O
36
Me
40
O
H
O
40
O
H
E
20
E
H
H
PO
Me
OP
Me
Me
PO
Me
OP
Me
Me
III
Scheme 1. Retrosynthetic analysis of pectenotoxin-4.
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
4570
0044-8249/02/4123-4570 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 23

COMMUNICATIONS
Synthesis of the AB spiroketal began with the addition of
the boron enolate derived from 7 to 5-hexenal, which afforded
the desired syn-aldol adduct in 87% yield (d.r. > 95:5;
Scheme 2).[9] Transamidation to N-phenylamide 8 (AlMe3,
aniline, CH2Cl2, 98%), olefin ozonolysis, followed by in situ
ether formation afforded the A-ring lactol methyl ether 9a. In
preparation for a Wittig-based fragment coupling, 9a was
OPMBOTBS
Me
H
16
7
PhHN
O
1
B
12
X
c
X
11
A
O
94%
OBn
O
H
14a, X = Br
14b, X = Li
O
d.r. >95:5
11, X = OEt
12, X = N(Me)OMe
17, X = H
e,f
86%
a,b
O
69%
OTBS
Me
H
OPMB OTBS
16
PhHN
O
1
16
B
11
H
A
O
O
H
13
OBn
O
O
15
O
OBn
Me
O
OH
a,b
PhHN
O
X
c
7
1
d.r. = 73:27 d 80%
O
N
1
PhHN
17, X = H
85%
A
Me
96%
Me
7
O
H
OPMBOTBS
Me
1
7
PhHN
O
g
Bn
B
16
11
8
9a, X = OMe
9b, X = PPh3+BF4
A
68%
d
O
–
O
H
d.r. = 86:14
O
OH
O
OTES
(see [Eq. (2)])
PhS
OH
OBn
16a
e,f
11
OEt
11 OEt
Scheme 3. Synthesis of C1 C16 Fragment 16a. a) [Ti(OiPr)Cl3], 2-bro-
moallyltrimethylsilane, CH2Cl2, À788C; 87%; b) PMB-trichloroacetimi-
date, TfOH, Et2O, À208C!08C; 79%; c) 14a, tBuLi, Et2O/pentane,
À788C then 12; 94%; d) Zn(BH4)2, Et2O/CH2Cl2, À408C; 80%; e) LiBH4,
THF, À108C; 93%; f) SO3¥Py, Et3N, DMSO/CH2Cl2 (1:1); 91%; g) 14a,
tBuLi, Et2O, À788C, then MgBr2, then aldehyde 17, Et2O, À788C; 68%.
See reference [7] for abbreviations.
g,h; 80%
d.r. >95:5
H
96%
3
10
O
O
Me
Me
H
H
Me
N
PhHN
O
1
PhHN
O
B
1
B
i
11
OEt
11
7
7
A
OMe
O
A
O
91%
O
H
O
H
O
O
12
11
Scheme 2. Synthesis of the AB spiroketal. a) 1. nBu2BOTf, Et3N, CH2Cl2,
À788C; 2. 5-hexenal, CH2Cl2, À788C; 87%; b) AlMe3, aniline; CH2Cl2;
98%; c) 1. O3, MeOH, then DMS, À788C; 2. CSA; 96%; d) PPh3¥HBF4,
CH3CN, 608C; e) TESOTf, lut., CH2Cl2, À788C; 99%; f) Pd/C, Et3SiH,
acetone; 97%; g) 1. 2 equiv nBuLi, THF, À788C; 2. aldehyde 10; h) CSA,
MeOH/CH2Cl2 (1:1); 80% (three steps); i) MeONHMe¥HCl,
iPrMgCl,THF, À788C!À208C; 91%. See reference [7] for abbreviations.
LiBHBu3 afforded better selectivity[15] (d.r. 81:19, 58% yield),
but resulted in competitive 1,4-reduction (38%).
Aviable alternative to the construction of 16a was to access
the stereocenter at C11 directly by a Felkin-controlled
addition to the corresponding spiroketal aldehyde 17.[16] The
desired aldehyde 17 was accessed in two steps:[17] full
reduction of the ethyl ester (LiBH4, THF, À108C, 94%)
followed by Parikh Doering oxidation (SO3¥Py, Et3N, CH2Cl2/
DMSO, À108C, 91%).[18] The Grignard reagent derived from
vinyl bromide 14 added to aldehyde 17 to afford 16a with
good selectivity (86:14) and yield (68%).[19]
The completion of the fragment synthesis began with an
investigation into electrophile-induced heterocyclizations to
form ring-C [Eq. (4)]. Unfortunately, all attempts to induce
cyclization of either 16a or 16b with either halogenating
agents or with mercuric salts afforded the undesired ring-C
C12 diastereomer 19.
transformed into the anomeric phosphonium salt 9b
(PPh3¥HBF4, CH3CN, 608C).[10] The aldehyde partner 10 was
readily accessed by protection of aldol adduct 3, followed by
chemoselective half reduction of the S-phenyl thioester under
conditions (Pd/C, Et3SiH, acetone, 97%) reported by Fu-
kuyama et al.[11]
Wittig coupling of the ylide derived from 9b and aldehyde
10 afforded the corresponding enol ether, which was imme-
diately subjected to camphorsulfonic acid (CSA) in MeOH to
provide the desired AB spiroketal 11 in 80% overall yield
(d.r. > 95:5). In preparation for the next fragment
coupling, the ethyl ester group at C11 of 11 was
transformed to the Weinreb amide 12 by using a
modified literature procedure.[12]
With the spiroketal in hand, the C11 C12 bond
construction was examined. The desired C12 C16
syn-diol was readily accessed from the allylation
OPMB
X
Me
H
H
OPMBOTBS
H
O
O
R
O
B
B
C
16
(4)
11
11
1
12
A
A
O
O
O
16
H
NHPh
H
OH
OTBS
OH
OR
19, X = I, HgOAc; d.r. > 10:1
16a, R = Bn; 16b, R = H
of the differentially protected glyceraldehyde
derivative 13 (d.r. > 95:5; Scheme 3).[13] PMB
protection of the resulting C14 alcohol required the use of
PMB-trichloroacetimidate.[14] Attempts to employ basic con-
ditions resulted in the formation of alkyne by-products from
the elimination of hydrogen bromide. Lithium halogen
exchange of vinyl bromide 14a (tBuLi, Et2O/pentane,
À788C), followed by addition of Weinreb amide 12 afforded
the desired C1 C16 carbon chain. Reduction of the resulting
carbonyl group at C11 in 15 afforded moderate chelate-
controlled diastereoselectivity with Zn(BH4)2 (d.r. 73:27, 80%
yield). The use of superior chelating reagents such as
The successful solution to the ring-C construction and the
synthesis of the ABC epoxide fragment began with the
selective deprotection of the benzyl ether group at C15 under
reducing conditions[20] (LiDBB, THF, À788C, 81%)
(Scheme 4). This step was followed by C11 hydroxy-directed
epoxidation[21] ([VO(acac)2], TBHP, PhH, 99%, d.r. > 95:5)
to afford a single epoxide diastereomer 18, which was cyclized
to 20a under mildly acidic conditions (PPTS, CH2Cl2, 87%).
Acylation of the primary alcohol with phenylthiono chlor-
oformate (DMAP, CH2Cl2, À108C, 75%), followed by
Angew. Chem. Int. Ed. 2002, 41, No. 23
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4123-4571 $ 20.00+.50/0
4571

COMMUNICATIONS
Me
OPMB
H
Me
42
42
Me
H
OPMB OTBS
H
OPMB OTBS
H
O
O
O
O
O
R
11
B
c
a,b
16
O
O
B
16
11
B
C
1
11
1
1
A
O
A
O
87%
70%
A
O
O
16
NHPh
H
NHPh
H
NHPh
H
OH
OH
OH
OBn
OH
OTBS
16a
18 (d.r. >95:5)
20a, R = CH2OH
20b, R = CH3
d,e
OPMB
H
54%
H
OPMB
Me
Me
77%
f,g
OPMB
H
H
H
Me
Me
Me
H
O
O
O
O
B
C
OTES
B
C
OH
11
Me
j,k,l
80%
OH
h,i
78%
OBn
1
11
O
O
1
B
A
O
O
C
O
A
O
O
16
16
11
1
NHPh
H
NHPh
H
A
O
OTBODPS
16
OTBODPS
NHPh
H
OTBODPS OH
O
23 (d.r. = 95:5)
22 (d.r. >95:5)
21
TBODPS = Si(OtBu)Ph2
19
19
Scheme 4. Completion of C1 C19 ABC fragment. a) LiDBB, THF, À788C; 81%; b) [VO(acac)2], TBHP, PhH; 99%; c) PPTS, CH2Cl2; 87%; d) PhOC(S)Cl,
DMAP, CH2Cl2, À108C; 75%; e) Bu3SnH, AIBN, xylenes, 1608C; 72%; f) TBODPSCl, DMAP, Et3N, DMF; 90%; g) CSA, CH2Cl2/MeOH (1:1), À208C;
85%; h) SO3¥Py, Et3N, CH2Cl2/DMSO (1:1), À108C; i) 2-(benzyloxymethyl) allyltributylstannane, BF3¥OEt2, CH2Cl2, À788C; 78% (two steps); j) TESCl, Im,
CH2Cl2; 92%; k) LiDBB, THF, À788C; 87%; l) Ti(OtBu)4, (l)-(þ)-DET, TBHP, 3 ä MS, CH2Cl2, À308C; 99%. See reference [7] for abbreviations.
Chem. Soc. 1999, 121, 7540 7552; b) D. A. Evans, D. M. Fitch, T. E.
Smith, V. J. Cee, J. Am. Chem. Soc. 2000, 122, 10033 10046.
[6] D. A. Evans, D. W. C. MacMillan, K. R. Campos, J. Am. Chem. Soc.
1997, 119, 10859 10860.
[7] Abbreviations: OTf ¼ trifluoromethanesulfonyl; Bn ¼ benzyl, TBS ¼
deoxygenation under standard Barton conditions completed
the synthesis of the C-ring tetrahydrofuran 20b.[22,23]
The hydroxy group at C11 was protected as a tert-
butoxydiphenylsilyl (TBODPS) ether.[24] Selective deprotec-
tion of the primary TBS ether (CSA, CH2Cl2/MeOH),
oxidation of the resulting alcohol 21 (SO3¥Py, Et3N, CH2Cl2/
DMSO, À108C)[18] to the unstable a,b-bisalkoxy aldehyde,
followed by Felkin controlled allylation with 2-(benzyloxy-
methyl)allyltributylstannane[25] (BF3¥OEt2, CH2Cl2, À788C,
78% over two steps, d.r. 95:5) provided homoallyl alcohol 22,
the complete carbon chain of the C1 C19 fragment. Protec-
tion of the hydroxy group at C16 (TESCl, Im, CH2Cl2, 92%),
benzyl deprotection[20] (LiDBB, THF, À788C, 87%), and a
stoichiometric Sharpless asymmetric epoxidation[26] ([Ti(Ot-
Bu)4], (l)-(þ)-DET, TBHP, 3 ä MS, CH2Cl2, À308C, 99%,
d.r. 95:5) afforded the desired ABC epoxide 23 with excellent
selectivity.
tert-butyldimethylsilyl; d.r. ¼ diastereomeric ratio; TES ¼ triethylsil-
yl; lut. ¼ 2,6-lutidine; PMB ¼ 4-methoxybenzyl; DMSO ¼ dimethyl
sulfoxide; LiDBB ¼ lithium di-tert-butyl biphenylide; DMAP ¼ 4-
(N,N-dimethylamino)pyridine; AIBN ¼ 2,2’-azobisisobutyronitrile;
TBODPS ¼ tert-butoxydiphenylsilyl; Im ¼ imidazole; PPTS ¼ pyridi-
nium para-toluenesulfonate; DMF ¼ dimethylformamide; CSA ¼
camphorsulfonic acid; TBHP ¼ tert-butyl hydroperoxide; DET¼ di-
ethyl tartrate; MS ¼ molecular sieves.
[8] D. A. Evans, K. A. Scheidt, J. N. Johnston, M. C. Willis, J. Am. Chem.
Soc. 2001, 123, 4480 4491.
[9] D. A. Evans, J. Bartroli, T. L. Shih, J. Am. Chem. Soc. 1981, 103, 2127
2129.
[10] a) J. Godoy, S. V. Ley, B. Lygo, J. Chem. Soc. Chem. Commun. 1984,
1381 1382; b) D. Culshaw, P. Grice, S. V. Ley, G. A. Strange,
Tetrahedron Lett. 1985, 26, 5837 5840.
[11] T. Fukuyama, S.-L. Lin, L. Li, J. Am. Chem. Soc. 1990, 112, 7050
7051; See also: D. A. Evans, B. W. Trotter, P. J. Coleman, B. Cote, L.
Carlos Dias, H. A. Rajapakse, A. N. Tyler, Tetrahedron 1999, 55,
8671 8726.
[12] The typical conditions of adding iPrMgCl to a cooled solution of ester
and MeNHOMe¥HCl resulted in the formation of significant amounts
of isopropyl ketone. Preforming the magnesium amide of the Weinreb
salt, thus ensuring that all the Grignard reagent was consumed prior to
ester addition, was critical for the success of this reaction. For the
original procedure, see: J. M. Williams, R. B. Jobson, N. Yasuda, G.
Marchesini, U.-H. Dolling, E. J. Grabowski, Tetrahedron Lett. 1995,
36, 5461 5464.
The preceding discussion describes the stereoselective
synthesis of the C1 C19 ABC epoxide fragment of the
pectenotoxin skeleton. The following communication de-
scribes the syntheses of the C20 C30 E-ring and C31 C40 FG
fragments, and the fragment assemblage to pectenotoxin-4.
Received: September 18, 2002 [Z50190]
[13] M. T. Reetz, K. Kesseler, J. Org. Chem. 1985, 50, 5434 5436.
[14] a) H. P. Wessel, T. Iversen, D. R. Bundle, J. Chem. Soc. Perkin Trans. 1
1985, 2247 2250; b) N. Nakajima, K. Horita, R. Abe, O. Yonemitsu,
Tetrahedron Lett. 1988, 29, 4139 4142.
[15] A.-M. Faucher, C. Brochu, S. R. Landry, I. Duchesne, S. Hantos, A.
Roy, A. Myles, C. Legault, Tetrahedron Lett. 1998, 39, 8425
8428.
[16] N. T. Ahn, O. Einstein, Nouv. J. Chim. 1977, 1, 61 70.
[17] Attempts to access aldehyde 17 directly by half reduction with
electrophilic reagents such as diisobutylaluminum hydride resulted in
N-phenylamide decomposition. This was the only reagent incompat-
ibility observed throughout the synthesis.
[1] a) T. Yasumoto, M. Murata, Y. Oshima, M. Sano, G. K. Matsumoto, J.
Clardy, Tetrahedron 1985, 41, 1019 1025; b) K. Sasaki, J. L. C.
Wright, T. Yasumoto, J. Org. Chem. 1998, 63, 2475 2480; c) J. H.
Jung, C. J. Sim, C.-O. Lee, J. Nat. Prod. 1995, 58, 1722 1726; d) M.
Daiguji, M. Satake, K. J. James, A. Bishop, L. Mackenzie, H. Naoki, T.
Yasumoto, Chem. Lett. 1998, 653 654.
[2] a) S. Amano, K. Fujiwara, A. Murai, Synlett 1997, 1300 1302; b) D.
Awakura, K. Fujiwara, A. Murai, Synlett 2000, 1733 1736; c) G. C.
Micalizio, W. R. Roush, Org. Lett. 2001, 3, 1949 1952; d) L. A.
Paquette, X. Peng, D. Bondar, Org. Lett. 2002, 4, 937 940.
[3] D. A. Evans, H. A. Rajapakse, A. Chiu, D. Stenkamp, Angew. Chem.
2002, 114, 4755; Angew. Chem. Int. Ed. 2002, 41, 4573.
[18] J. R. Parikh, W. von E. Doering, J. Am. Chem. Soc. 1967, 89, 5505
5507.
[19] The Grignard reagent was prepared by adding a Et2O/PhH solution of
MgBr2 to vinyllithium 14b. See: M. Nakatsuka, J. A. Ragan, T.
Sammakia, D. B. Smith, D. E. Uehling, S. L. Schreiber, J. Am. Chem.
Soc. 1990, 112, 5583 5601.
[4] For similar bond constructions in a complex setting see: a) D. A.
Evans, S. L. Bender, J. Morris, J. Am. Chem. Soc. 1988, 110, 2506
2526; b) D. A. Evans, R. P. Polniaszek, K. M. DeVries, D. E. Guinn,
D. J. Mathre, J. Am. Chem. Soc. 1991, 113, 7613 7630.
[5] For the use of N-phenylamides as carboxyl surrogates see: a) D. A.
Evans, P. H. Carter, E. M. Carreira, J. A. Prunet, M. Lautens, J. Am.
4572
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4123-4572 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 23

COMMUNICATIONS
[20] a) P. K. Freeman, L. L. Hutchinson, J. Org. Chem. 1980, 45, 1924
1930; b) R. E. Ireland, M. G. Smith, J. Am. Chem. Soc. 1988, 110, 854
860.
[21] B. E. Rossiter, T. R. Verhoeven, K. B. Sharpless, Tetrahedron Lett.
1979, 20, 4733 4736.
epoxide alkylation (transform T1) precedes diene formation
(transform T2).
The plan for the construction of the F-ring tetrahydrofuran
IV was to involve a C37 hydroxy-directed epoxidation of
olefin VI with a subsequent ring closure by the C32 hydroxy
moiety (transform T3). Finally, the stereoselective formation
of the E-ring tetrahydrofuran V from its acyclic precursor VII
was based on an iodoetherification precedent provided by
Bartlett and Rychnovsky (transform T4).[3]
The synthesis of the ring-E synthon V began with the
known aldol adduct adduct 2 (Scheme 1).[4] Reduction of 2
(LiBH4, THF, 08C), and selective protection of the primary
alcohol (TBSCl, Im, CH2Cl2, 100% over two steps) afforded
allylic alcohol 3.[5] Acylation of 3 with the PMB-protected
lactic acid 4[6] (DCC, DMAP, CH2Cl2, 52%), followed by
carbonyl olefination of 5a with Tebbe reagent[7] afforded the
1,5-diene 5b. Claisen rearrangement of 5b in refluxing
toluene gave the desired rearrangement product 6 in 82%
yield for the two steps. Chelate-controlled reduction of the
resulting ketone (Zn(BH4)2, Et2O, À788C, 86%, d.r. 86:14)
provided the precursor for the key iodoetherification reac-
tion. In spite of the modest selectivity that was observed for
the formation of the desired tetrahydrofuran 7 (NIS, CH3CN,
À408C, 89%, d.r. 72:28), this outcome proved sufficient to
pursue the planned route.
[22] D. H. Barton, W. B. Motherwell, A. Stange, Synthesis 1981, 743 745.
[23] The stereochemistry of the newly formed C12 stereocenter of
tetrahydrofuran 20b was assigned by using 2D NMR spectroscopy.
[24] J. W. Gillard, R. Fortin, H. E. Morton, C. Yoakim, C. A. Quesnelle, S.
Daignault, Y. Guidan, J. Org. Chem. 1988, 53, 2602 2608. This
unusual protecting group offered the necessary acid stability during
the synthesis, as well as the required fluoride ion lability for the final
global deprotection with tris(dimethyamino)sulfur(trimethylsilyl) di-
fluoride (See: K. A. Scheidt, H. Chen, B. C. Follows, S. R. Chemler,
D. S. Coffey, W. R. Roush, J. Org. Chem. 1998, 63, 6436 6437). Under
these conditions, the time required for C11-OTBS deprotection was
prohibitively long. The use of a more base-labile TES ether did not
afford the required acid stability during the synthesis.
[25] S. V. Ley, L. R. Cox, J. Chem. Soc. Perkin Trans. 1 1997, 3315 3324.
[26] L. D.-L Lu, R. A. Johnson, M. G. Finn, K. B. Sharpless, J. Org. Chem.
1984, 49, 728 731. Sharpless has noted (See: Y. Gao, R. M. Hanson,
J. M. Klunder, S. Y. Ko, H. Masamune, K. B. Sharpless, J. Am. Chem.
Soc. 1987, 109, 5765 5780, and references therein) that 1,1-disub-
stitued olefins are problematic substrates for epoxidation under
stoichiometric [Ti(OiPr)4] conditions due to epoxide opening by
isopropyl alcohol. The use of [Ti(OtBu)4] conveniently avoided any
such side reactions.
Successive radical dehalogenation of 7 (Bu3SnH, AIBN,
toluene, 100%) and deprotection of the primary TBS ether
(TBAF, THF, 95%) afforded alcohol 8. Oxidation with Dess
Martin reagent[8] (py, CH2Cl2, 99%), Wittig homologation
(EtOC(O)CC(CH3)PPh3, THF, 658C; 100% E:Z > 95:5), and
ester reduction (LiAlH4, Et2O, 0 8C, 92%) completed the
carbon assembly of the E-ring fragment. Benzyl protection
(NaH, BnBr, TBAI, THF/DMF, 94%) followed by PMB
deprotection (DDQ, CH2Cl2/pH 7 buffer, 95%) gave alcohol
10. Oxidation to the methyl ketone[8] (Dess Martin period-
inane, py, CH2Cl2, 93%), and hydrazone formation (TMSCl,
CH2Cl2/Me2NNH2, 100%) completed the synthesis of hydra-
zone 11.
As summarized in Figure 1, the first stage of the synthesis of
the ring-F fragment IV will be simplified to the construction
of the C31 C35 phosphonium salt, the C36 C40 aldehyde,
and their union through a Wittig coupling to afford the Z-
olefin VI.
The synthesis of the C31 C35 phosphonium salt began with
the known triol derivative 12 (Scheme 2).[9] Protection of the
hydroxy group at C33 of 12 as a PMB ether (PMBBr, NaH,
THF/DMF, 95%) followed by reductive ozonolysis (O3,
EtOH, then DMS, then NaBH4, 95%) afforded alcohol 13.
Transformation of 13 to the corresponding iodide (I2, Im,
Ph3P, CH 2Cl2, 08C, 89%) proceeded smoothly, but careful
control of the temperature was required to access phospho-
nium salt 14 (Ph3P, CH3CN, 558C, 89%).[10]
Asymmetric Syntheses of Pectenotoxins-4 and -
8, Part II: Synthesis of the C20 C30 and C31
C40 Subunits and Fragment Assembly**
David A. Evans,* Hemaka A. Rajapakse,
Anna Chiu, and Dirk Stenkamp
In the preceding communication,[1] the proposed synthesis
plan identified the two principal pectenotoxin-4 subunits II
and III (Figure 1). It was our intention to couple these
fragments through the alkylation of the metalloenamine
derived from hydrazone III, readily available from the
coupling of advanced intermediates IV and V (transform
T2), by epoxide II. However, this investigation revealed that
the above bond construction was not feasible due to the
decomposition of metalloeneamine III under the reaction
conditions.[2] Accordingly, the objective in the present com-
munication is the synthesis of the subunits IV and V, and the
completion of the syntheses of pectenotoxin-4 (1) and
pectenotoxin-8 by a revised fragment coupling strategy, where
[*] Prof. D. A. Evans, H. A. Rajapakse, A. Chiu, D. Stenkamp
Department of Chemistry & Chemical Biology
Harvard University
Cambridge, MA 02138 (USA)
Fax : (þ 1)617-495-1460
The synthesis of the aldehyde partner 17 began with
protection of the hydroxy group at C37 of aldol adduct 15[11]
as a base-sensitive triphenylsilyl ether (TPSCl, Im, DMAP,
DMF, 08C, 98%; Scheme 2). Half reduction of the S-phenyl
thioester[12] (Pd/C, Et3SiH, acetone, 95%), and olefination
under modified Lombardo conditions[13] ([Cp2ZrCl2], Zn dust,
CH2I2, THF, 08C, 84%) afforded olefin 16. Rhodium-
E-mail: evans@chemistry.harvard.edu
[**] Financial support has been provided by the National Institutes of
Health (GM33327-16). A BASF-Forschungsstipendum Postdoctoral
Fellowship to D.S. and a Howard Hughes Medical Institute Predoc-
toral Fellowship to H.A.R. are gratefully acknowledged. The NIH
BRS Shared Instrumentation Grant Program 1-S10RR04870 and the
NSF (CHE 88-140019) are acknowledged for providing NMR
facilities.
Angew. Chem. Int. Ed. 2002, 41, No. 23
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4123-4573 $ 20.00+.50/0
4573