"Endo" and "Exo" Bicyclo[4.2.0]-octadiene Isomers from the Electrocyclization of Fully Substituted Tetraene Models for SNF 4435C and D. Control of Stereochemistry by Choice of a Functionalized Substituent

Article abstract of DOI:10.1021/ol036048+

(Matrix presented) A tandem electrocyclic closure, perceived as the key step in a biomimetic approach to SNF 4435C and D, was tested with 1,1,8-trisubstituted tetraene substrates. The ratio of endo:exo products could be controlled by the choice of the R

Full text of DOI:10.1021/ol036048+

ORGANIC
LETTERS
2004
Vol. 6, No. 2
161-164
“Endo” and “Exo”
Bicyclo[4.2.0]-octadiene Isomers from
the Electrocyclization of Fully
Substituted Tetraene Models for SNF
4435C and D. Control of
Stereochemistry by Choice of a
Functionalized Substituent
Kathlyn A. Parker* and Yeon-Hee Lim
Department of Chemistry, State UniVersity of New York at Stony Brook,
Stony Brook, New York, 11794-3400
kparker@notes.cc.sunysb.edu
Received October 18, 2003
ABSTRACT
A tandem electrocyclic closure, perceived as the key step in a biomimetic approach to SNF 4435C and D, was tested with 1,1,8-trisubstituted
Z
tetraene substrates. The ratio of endo:exo products could be controlled by the choice of the R substituent at C-1. On the basis of these
results, a short stereoselective route to an advanced SNF 4435 intermediate was devised.
SNF4435C and D are natural products reported in 20011 as
isolates from Streptomyces spectabilis, a soil organism from
Okinawa. Their reported biological activities are significant,
and their novel structures present interesting questions of
biosynthetic origins and biomimetic synthetic approaches.
multidrug resistance (MDR) in vitro in several cell lines at
nanomolar concentrations. Furthermore, SNF 4435C showed
positive effects in vincristine-treated mice that bore VCR-
resistant P388 leukemia.3 On the basis of these results,
Kurosawa and co-workers suggested that these drugs act in
the MDR test by modulating cellular P-glycoprotein via
direct binding.
The SNF compounds originally attracted attention in a
screen for immunosuppressants, exhibiting activity at sub-
micromolar concentrations. Subsequently, they were shown
to act by suppression of B-cell proliferation induced by
lipopolysaccaharide (LPS) versus T-cell proliferation induced
by Con A.2 This mechanism of action is distinct from that
of FK-506. Additional interest in the SNF compounds arose
from the discovery that they demonstrated reversal of
Structure elucidation showed SNF 4435C and D to be
substituted bicyclo[4.2.0]octadienes, spiro-fused to a pyrone-
(1) Kurosawa, K.; Takahashi, K.; Tsuda, E. J. Antibiot. 2001, 54, 541.
(2) Kurosawa, K.; Takahashi, K.; Fujise, N.; Yamashita, Y.; Washida,
N.; Tsuda, E. J. Antibiot. 2002, 55, 71.
(3) Kurosawa, K.; Takahashi, K.; Tsuda, E.; Tomida, A.; Tsuruo, T. Jpn.
J. Cancer Res. 2001, 92, 1235.
10.1021/ol036048+ CCC: $27.50 © 2004 American Chemical Society
Published on Web 12/17/2003

substituted tetrahydrofuran substituent.4 Relative stereochem-
istry was established on the basis of NOE experiments. The
two compounds are stereoisomers in which four of the five
chiral centers have the same stereochemistry relative to each
other and fifth chiral center is epimeric (C-6).
As these two isomers may be derived from a single
biosynthetic precursor (see below), it seems likely that they
are diastereomeric at all centers except C-6. On the basis of
this supposition, Beaudry and Trauner proposed that the
structures of SNF 4435C and D should be drawn as 1a and
1b, respectively, isomeric at carbons 8, 10, 15, and 16 (Figure
1).5 We have adopted this convention in this paper. Ac-
isomer (3a and 3b w 4 and/or 5). Indeed, a similar 8π, 6π
closure has been postulated as the key ring-forming process
in the biosynthesis of the endiandric acids,8 an insight that
led to the biomimetic synthesis of these polyacetates.9 An
analysis similar to ours was made by Trauner, who was the
first to describe the key features of the biosynthetic scheme
in the literature.5 Both Trauner and, more recently, Baldwin10
have now reported model studies for the double-electro-
cyclization sequence with 1,8-disubstituted tetraenes.
Scheme 1. Suggested Biosynthesis of SNF 4435C and D
(relative stereochemistry only)
Figure 1. Structures of SNF 4435C and SNF 4435D. Absolute
stereochemistry as predicted in this communication.
ceptance of this premise leads to a further postulate. Because
SNF 4435C and D are congeners of spectinabilin 2,6 a
structural isomer for which the stereocenter at C-6 has been
assigned as R,7 it is reasonable to believe that structures of
SNF 4435C and D have the absolute stereochemistry shown
with the 6R-configuration.
Our own approach has been to explore the electrocycliza-
tion requirements of 1,1,8-trisubstituted tetraenes, substrates
with substitution patterns more pertinent to the apparent
biosynthetic pathway. Although synthetic equivalents of
either tetraene 4 or 5 might be expected to undergo the
symmetry-allowed double cyclization, there are significant
differences between the closure of these substrates and those
that have been studied previously.11 The literature does not
describe the 8π electrocyclization of a 1,1,8-trisubstituted
tetraene. Examination of the putative helical transition state12
for either of the proposed 8π cyclizations places the C-1
substituents squarely above the distal double bond, introduc-
ing what may be substantial steric interactions (Figure 3).
On the other hand, because the vinyl methyl groups at
positions 11, 13, and 15 should favor the reactive conforma-
tion (syn, syn, syn) of the substrate, the desired closure
Figure 2. Spectinabilin, (E,E,E,Z)-tetraene isomer.
Seeking a biomimetic synthetic approach, we considered
the possible biosynthetic origin of the SNF compounds. This
led us to the premise that the 6,4-ring system is the product
of a 6π electrocyclization within a cyclooctatriene (i.e., 1a
and 1b w 3a and 3b), that the cyclooctatriene is, itself, an
electrocyclic ring-closure product and that the linear precur-
sor is the polypropionate-derived tetraene or its geometric
(8) (a) Bandaranayake, W. M.; Banfield, J. E.; Black, D. St. C.; Falon,
G. D.; Gatehouse, B. M. Aust. J. Chem. 1981, 34, 1655. (b) Bandaranayake,
W. M.; Banfield, J. E.; Black, D. St. C. J. Chem. Soc., Chem. Commun.
1980, 19, 902.
(9) Nicolaou, K. C.; Petasis, N. A.; Zipkin, R. E. J. Am. Chem. Soc.
1982, 104, 5560.
(4) Takahashi, K.; Tsuda, E.; Kurosawa, K. J. Antibiot. 2001, 54, 548.
(5) Beaudry, C. M.; Trauner, D. Org. Lett. 2002, 4, 2221.
(6) Kakinuma, K.; Hanson, C. A.; Rinehart, K. L., Jr. Tetrahedron 1976,
32, 217.
(7) Ishibashi, Y.; Nishiyama, S.; Shizuri, Y.; Yamamura, S. Tetrahedron
Lett. 1992, 33, 521.
(10) Moses, J. E.; Baldwin, J. E.; Marquez, R.; Adlington, R. M.; Cowley,
A. R. Org. Lett. 2002, 4, 3731.
162
Org. Lett., Vol. 6, No. 2, 2004

Table 1. Coupling-Tandem Cyclization of Fully Substituted
Tetraene Models for SMF 4435C and Da
Figure 3. Helical transition state for thermal 8π electrocyclization
of tetraene 5 or its equivalent.
should be favored, relative to that of internally unsubstituted
tetraenes, by an entropic effect.
Thus, we addressed the preparation of substrates that
would be informative models for the cyclization that might
take place in S. spectabilis and for cyclizations that we might
exploit in total synthesis. We suspected that the electro-
cyclization of substrates with the geometry of the proposed
5 would be more facile than that of its isomer 4,8 and
therefore we focused on the preparation and cyclization
behavior of (Z,Z,Z,E)-1,1,3,5,7,8-substituted 1,3,5,7-tetraenes.
As we suspected that Z-E isomerization of conjugated
polyene intermediates might be facile, we planned a con-
vergent, late-stage coupling to provide the presumably
sensitive substrate. We believed that it would be prudent to
construct the tetraene directly from two diene precursors
rather than from a triene, as trienes themselves are subject
to electrocyclization reactions (noted above and observed by
Beaudry and Trauner5 in their studies).
entry diene
RZ
RE
prodb yield endo exo
1
2
3
4
5
6
8a
8b
8c
8d
8e
8f
CH2OTBS Me
11a
CO2Me 11b
31%c
62%e 100
11cd 38%e 100
50
50
0
0
H
CO2Et
Me
CO2Et
CN
H
CO2Et
Me
Me
11d
11e
11f
56%e
54%e
59%e
40
10
90
60
90
10
a Reaction conditions: (a) Sn2Me6, Pd(PPh3)4, PhH, reflux; (b)
Pd(CH3CN)2Cl2, DMF, rt. b Identified by NOE experiment. c Isolated yield;
endo and exo products not separated. d RE and RZ interchanged in the
product. e Sum of isolated yield of endo and exo products.
As a model for key conversions in both the biosynthesis
and a biomimetic total synthesis, we chose the electro-
cyclization of the tetraene 9a, viewed as the Stille coupling
product of dienes 65 and 8a.13 In fact, palladium-catalyzed
coupling was accompanied by double electrocylic closure,
providing a mixture of the anticipated endo product and the
unprecedented exo product (represented as endo- and exo-
11a). The results of this experiment are described in Table
1,14 entry 1.15
reported for the 1,8-disubstituted tetraene ester 9b in the
literature5 (and confirmed in our own lab, Table 1, entry 2),
we next moved to elucidate the requirements for specific
closure to endo-bicyclo[4.2.0]systems.
The experiment described in entry 3 leads to product 11c
with the same stereochemistry as that obtained in entry 2,
indicating isomerization in one of the steps of the cascade.
Tetraene 9d, an alkylated analogue of the endo-selective 9b,
gave a nearly 1:1 mixture of isomers, and tetraene 9e gave
a 1:9 mixture of the desired endo-11e and the undesired exo-
11e.
In light of the lack of stereospecificity in the double closure
of our 1,1,8-trisubstituted tetraene ether 9a and the specificity
In search of guidance for the design of cyclization
substrates that would favor endo products, we examined the
conformations available to the cyclooctatriene intermediates
10. Although only qualitative, the picture that emerged
allowed us to develop a hypothesis for further work.
In the allowed pericyclic closure that leads to endo-11b,
the cyclooctatriene adopts a conformation that we designate
“endo” (Figure 4) in which the nitrophenyl substituent is
tucked under the ring (nitrophenyl endo). The alternative,
allowed closure would proceed through the conformation
labeled “exo”, but this is disfavored because of a steric
interaction between the nitrophenyl substituent and the
adjacent methyl group.
(11) Although 8π, 6π closures of 1,8-disubstituted 1,3,5,7-(E,Z,Z,E)-
tetraenes proceed at low temperature (e.g., 2,4,6,8-(E,Z,Z,E)-decatetraene
closes at -10 °C over 30 h and the subsequent electrocylization occurs at
20 °C over 8 h), closure of (Z,Z,Z,E)-tetraenes requires higher temperatures
(9 °C for the first step and 40 °C for the second step) and the closure of
Z,Z,Z,Z-substrates is even slower, affording the [4.2.0] product directly at
65 °C. See: (a) Huisgen, R.; Dahmen, A.; Huber, H. J. Am. Chem. Soc.
1967, 89, 7130. (b) Huisgen, R.; Dahmen, A.; Huber, H. Tetrahedron Lett.
1969, 1461. (c) For a related example and a literature survey, see: Hayashi,
R.; Fernandez, S.; Okamura, W. H. Org. Lett. 2002, 4, 851.
(12) Thomas, B. E., IV; Evanseck, J. D.; Houk, K. N. J. Am. Chem.
Soc. 1993, 115, 4165.
(13) Stannane compounds 8 were prepared from the corresponding
iododienes 7; see Supporting Information.
(14) For each of the entries, the linear tetraene 9 underwent the double
closure within the time frame of the coupling experiment (i.e., we never
saw linear tetraene or cyclooctatriene in any of these experiments). Nuclear
Overhauser experiments confirmed the stereochemical assignments of all
products in Table 1.
For other substrates in the Table, the situation is more
complex. The endo transition state is somewhat disfavored
by substrates in which RZ is a methyl group (compare entries
(15) These isomers were not separated. Nevertheless, a NOE experiment
on the mixture allowed assignment.
Org. Lett., Vol. 6, No. 2, 2004
163

Scheme 2a
Figure 4. Endo and exo conformations for the cyclization of 10.
2 and 4), presumably because of steric repulsion between
this methyl group and the adjacent aryl substituent. The endo
transition state becomes more disfavored in substrates in
which RZ is an ester group (entries 3 and 5); the plane of
the ester and the plane of the aryl group must be parallel, a
situation that leads to transannular interactions between the
ester group and the cyclooctatriene ring.
Considering mechanisms for minimizing the steric repul-
sion between RZ and the aromatic group and also between
RZ and the cyclooctatriene ring, we imagined that RZ might
be the long but slender nitrile. Indeed, entry 6 shows that
tetraene 9f closed to give endo-11f as the almost exclusive
product.
a Reagent and conditions: (a) TBDMSCl, Et3N, DMAP, DMF,
40 °C (92%); (b) MsCl, Et3N, DMAP, CH2Cl2, 0 °C (96%); (c)
diethyl(cyanomethyl)phosphonate, NaH, DME, 0 °C (62%); (d) (Z)-
3-iodo-2-methylpropenal, KHMDS, toluene, -78 °C (89%); (e)
Sn2Me6, Pd(PPh3)4, PhH, reflux (84%); (f) 6, Pd(CH3CN)2Cl2,
DMF, rt (53%).
and stannylation provided the reagent for the coupling
protocol.
Thus, a dramatic reversal of product distribution results
from the replacement of an ester group (entry 5) with a nitrile
group (entry 6). We attribute the lowering of the transition
state energy for the endo product to the smaller steric
requirement for the RZ substituent, which must be tucked
into the endo cavity as the reaction proceeds.
Viewing the functional group pattern in the target struc-
tures 1a and 1b and noting the requirements for endo-
selective tandem closure, we concluded that tetraene 9g or
a similar cyanotetraene might prove to be a key intermediate
in a total synthesis. We have now demonstrated the acces-
sibility of the appropriate precursors and the good behavior
of the 1,3,6-triene system under the Stille coupling condi-
tions.
The three-step, one-pot transformation of vinyl stannane
8g to bicyclo 11g endo provided a 53% yield of crystalline
material after chromatography. A small amount (4%) of the
exo isomer was recovered from a slow-moving fraction.
With impressive stereoselectivity in these most recent
cyclizations established and with appropriately functionalized
substrates readily available, we are continuing our pursuit
of a biomimetic synthesis of SNF 4435C and D.
Acknowledgment. This work was supported by the
National Institutes of Health and by SUNY at Stony Brook.
We are grateful to Professor Nancy Goroff for assistance
with molecular modeling studies. We also thank Dr. Tun-Li
Shen (Brown University) for mass spectroscopic data.
The preparation of the vinyl tin reagent 8g, required for
this approach, proved to be quite straightforward. The
commercially available trans-2-butene-1,4-diol (12) was
subjected to a three-step sequence that afforded the Horner-
Wadsworth-Emmons reagent 13. Condensation with (Z)-
3-iodo-2-methyl-propenal gave the 1,3,6-(Z,Z,E)-triene 7g,
Supporting Information Available: Experimental pro-
cedures and spectroscopic and analytical data for all new
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
OL036048+
164
Org. Lett., Vol. 6, No. 2, 2004