Studies on the synthesis of (-)-gymnodimine. Subunit synthesis and coupling

Article abstract of DOI:10.1021/jo062396o

Two principal subunits of the marine algal toxin (-)-gymnodimine were synthesized. A trisubstituted tetrahydrofuran representing C10-C18 of the toxin was prepared via a highly stereoselective iodine-mediated cyclization of an acyclic alkene bearing a bis-

Full text of DOI:10.1021/jo062396o

Studies on the Synthesis of (-)-Gymnodimine. Subunit Synthesis
and Coupling
James D. White,* Laura Quaranta, and Guoqiang Wang
Department of Chemistry, Oregon State UniVersity, CorVallis, Oregon 97331-4003
james.white@oregonstate.edu
ReceiVed NoVember 21, 2006
Two principal subunits of the marine algal toxin (-)-gymnodimine were synthesized. A trisubstituted
tetrahydrofuran representing C10-C18 of the toxin was prepared via a highly stereoselective iodine-
mediated cyclization of an acyclic alkene bearing a bis-2,6-dichlorobenzyl (DCB) ether. The formation
of a cis-2,5-disubstituted tetrahydrofuran in this process conforms to a stereodirecting effect by the DCB
group proposed by Bartlett and Rychnovsky. A cyclohexene subunit corresponding to the C1-C8, C19-
C24 portion of gymnodimine was synthesized via Diels-Alder cycloaddition of a 1,2,3-trisubstituted
diene to a symmetrical dienophile obtained from Meldrum’s acid. Differentiation of carbonyl groups in
the cycloadduct was made by an intramolecular reaction with a neighboring alcohol to form a γ-lactone.
Linkage of the two subunits at C18-C19 was accomplished by using a B-alkyl Suzuki coupling in which
a borane prepared from the pendent alkenyl chain of the cyclohexene domain was reacted with the (E)-
iodoalkene attached at C16 of the tetrahydrofuran sector. Subsequent transformations positioned functional
groups in the coupled product for a future macrocyclization event that would close the 15-membered
ring of gymnodimine.
Introduction
or proliferation of the organism under favorable conditions, a
phenomenon currently categorized under the broad term Harmful
Algal Blooms (HABs) but previously known as “red tides”.
Although red tides were described as long as 3000 years ago,
their occurrence and geographical distribution have increased
markedly in recent times, with the result that incidents of human
poisoning from marine algal sources have also increased.2
Chemical studies of bioactive compounds produced by HABs
have led to the identification of several highly potent neuro-
toxins, including the brevetoxins,3 prymnesins,4 and gymnocin-
A.5 Blooms of the dinoflagellate Gymnodinium sp. are frequently
associated with production of these toxins and it was during
Marine algal toxins are notorious not only for their undesir-
able effects on the human gastric system but also for the
devastation they can cause in the natural environment.1 In the
United States, illnesses associated with seafood consumption
account for approximately 20% of all foodborne disease
outbreaks, and half of those are caused by naturally occurring
algal toxins. Around the world, more than 60 000 incidents of
sickness due to ingestion of marine algal toxins are reported
each year with 1.5% of those cases resulting in fatality. The
major source of marine algal toxins is unicellular algae
(phytoplankton), which form the base of the marine food chain.
Enormous blooms of these algae are created by aggregation and/
(3) Lin, Y.-Y.; Rish, M.; Ray, S. M.; Van Engen, D.; Clardy, J.; Golik,
J.; James, J. C.; Nakanishi, K. J. Am. Chem. Soc. 1981, 103, 6773.
(4) (a) Igarashi, T.; Satake, M.; Yasumoto, T. J. Am. Chem. Soc. 1996,
118, 479. (b) Igarashi, T.; Satake, M.; Yasumoto, T. J. Am. Chem. Soc.
1999, 121, 8499.
(1) Van Dolah, F. M. EnViron. Health Perspect. 2000, 108, 133.
(2) Luckas, B.; Dahlmann, J.; Erler, K.; Gerdts, G.; Wasmund, N.;
Hummert, C.; Hansen, P. D. EnViron. Toxicol. 2005, 20, 1.
10.1021/jo062396o CCC: $37.00 © 2007 American Chemical Society
Published on Web 01/31/2007
J. Org. Chem. 2007, 72, 1717-1728
1717

White et al.
SCHEME 1. Retrosynthetic Analysis of Gymnodimine
one such bloom in 1994 in Foveaux Strait off the South Island
of New Zealand that gymnodimine is believed to have made
its appearance. A year later, oysters collected from the same
area by Yasumoto yielded sufficient gymnodimine for its planar
structure to be determined.6 Subsequently, the relative and
absolute configuration of the toxin was established as 1 by X-ray
crystallographic analysis of the p-bromobenzamide derivative
3 of the reduction product gymnodamine (2).7 More recently,
several analogues of 1, gymnodimines B (4) and C (5) as well
as deoxygymnodimine B (6), have been isolated from Gymno-
dinium sp.,8 and in 2002 gymnodimine itself was found in clams
harvested in Tunisia.9
rotoxic spirolides12 and pinnatoxins13 which also contain a
spiroimine unit show much reduced activity when the imine
function is reduced or hydrolyzed to an amino ketone, suggesting
that there is a common pharmacophore bearing the spiroimine
signature in marine toxins.
Progress toward the total synthesis of gymnodimine has been
disclosed by Murai14 and Romo,15 and two reports from this
laboratory have outlined independent approaches to the tetrahy-
drofuran subunit16 of 1 and to the spiroimine portion.17 This
paper presents further details of our routes to these two domains
of gymnodimine and describes their coupling to give a structure
that we foresee leading to 1. Our strategy is outlined in Scheme
1 and envisions construction of the central macrocyclic portion
of gymnodimine from two sectors that are connected first at
C18-C19 (bond A) followed by ring closure through linkage
of C10 and C11 at bond B. Elaboration of the spiroimine and
butenolide subunits of 1 would take place after the macrocycle
7 is completed. The two segments that merge into 7 are defined
as the pentasubstituted cyclohexene 8 and the trisubstituted
tetrahydrofuran 9, the means for their connection being open
to variation. This blueprint mandates the preparation of 8 and
9 in enantiopure form and thus imposes certain restrictions upon
the routes chosen for their syntheses.
Gymnodimine has been shown to be a potent neurotoxin and
is believed to be responsible for several incidents of neurotoxic
shellfish poisoning that have occurred in New Zealand.10 In a
mouse bioassay, 1 showed acute toxicity by injection (LD50 96
µg/kg) and oral administration (LD50 755 µg/kg); however,
toxicity was diminished when 1 was ingested with food.11 The
toxicity of gymnodamine (2) is negligible (MLD > 4040 µg/
kg) compared to that of 1, indicating that the imine moiety of
gymnodimine is necessary for activity. Interestingly, the neu-
(12) (a) Hu, T.; Curtis, J. M.; Oshima, Y.; Quilliam, M. A.; Walter,
J. A.; Waston-Wright, W. M.; Wright, J. L. C. Chem. Commun. 1995, 2159.
(b) Hu, T.; Curtis, J. M.; Walter, J. A.; Wright, J. L. C. Tetrahedron Lett.
1996, 37, 7671. (c) Hu, T.; Burton, I. W.; Cembella, A. D.; Curtis, J. M.;
Quilliam, M. A.; Walter, J. A.; Wright, J. L. C. J. Nat. Prod. 2001, 64,
308. (d) Cembella, A. D.; Lewis, N. I.; Quilliam, M. A. Nat. Toxins 1999,
7, 197.
(13) (a) Uemura, D.; Chou, T.; Hanio, T.; Nagatsu, A.; Fukuzawa, S.;
Zheng, S.; Chen, H. J. Am. Chem. Soc. 1995, 117, 1155. (b) Chou, T.;
Kamo, O.; Uemura, D. Tetrahedron Lett. 1996, 37, 4023. (c) Chou, T.;
Haino, T.; Kuramoto, M.; Uemura, D. Tetrahedron Lett. 1996, 37, 4027.
(14) (a) Ishihara, J.; Miyakawa, J.; Tsujimoto, T.; Murai, A. Synlett 1997,
1417. (b) Tsujimoto, T.; Ishihara, J.; Horie, M.; Murai, A. Synlett 2002,
399.
(5) Satake, M.; Shoji, M.; Oshima, Y.; Naoki, H.; Fujita, T.; Yasumoto,
T. Tetrahedron Lett. 2002, 43, 5829.
(6) Seki, T.; Satake, M.; Mackenzie, L.; Kaspar, H.; Yasumoto, T.
Tetrahedron Lett. 1995, 36, 7093.
(7) Stewart, M.; Blunt, J. W.; Munro, M. H. G.; Robinson, W. T.;
Hannah, D. J. Tetrahedron Lett. 1997, 38, 4889.
(8) (a) Miles, C. O.; Wilkins, A. L.; Stirling, D. J.; Mackenzie, A. L.
J. Agric. Food Chem. 2000, 48, 1373. (b) Miles, C. O.; Wilkins, A. L.;
Stirling, D. J.; Mackenzie, A. L. J. Agric. Food Chem. 2003, 51, 4838.
(9) Bire, R.; Krys, S.; Fremy, J.-M.; Dragacci, S.; Stirling, D.; Kharrat,
R. J. Nat. Toxins 2002, 11, 269.
(15) (a) Yang, J.; Cohn, S.; Romo, D. Org. Lett. 2000, 2, 763. (b) Ahn,
Y.; Cardenas, G. I.; Yang, J.; Romo, D. Org. Lett. 2001, 3, 751.
(16) White, J. D.; Wang, G.-Q.; Quaranta, L. Org. Lett. 2003, 5, 4109.
(17) White, J. D.; Wang, G.-Q.; Quaranta, L. Org. Lett. 2003, 5, 4983.
(10) Stirling, D. J. N. Z. J. Mar. Freshwater Res. 2001, 35, 851.
(11) Munday, R.; Towers, N. R.; Mackenzie, L.; Beuzenberg, V.;
Holland, P. T.; Miles, C. O. Toxicon 2004, 44, 173.
1718 J. Org. Chem., Vol. 72, No. 5, 2007

Synthesis of (-)-Gymnodimine
SCHEME 2. Synthetic Plan for the Tetrahydrofuran
Subunit
SCHEME 4. The Bartlett-Rychnovsky
Cis-2,5-Disubstituted Tetrahydrofuran Synthesis
SCHEME 3. First Generation Synthesis of the
Tetrahydrofuran Precursor
genation20 completely unproductive. Furthermore, although the
mesylate of 13 could be prepared, this unstable substance
underwent elimination to a conjugated diene before its displace-
ment by hydride could be effected. This, again, reflects a high
degree of steric compression around C3 of 13. In the hope that
inversion at this center would lead to a more compliant substrate
for erasure of the C3 hydroxyl group, 13 was oxidized to ketone
14, which gave predominantly (9:1) the (3R) alcohol 15 upon
reduction with L-Selectride. This circuitous tactic was to no
avail, however, since 15 was as unreactive as 13 toward reagents
designed to activate the free alcohol for its removal.
The difficulty associated with deletion of the superfluous
hydroxyl group in 13 and 15 persuaded us to postpone removal
of this function to a later stage, and our advance toward 9 was
therefore continued from 13 (Scheme 4). After protection of
this alcohol as its p-methoxybenzyl (PMB) ether 16, the
acetonide was cleaved in acidic methanol to give diol 17 as the
putative substrate for cyclization to a tetrahydrofuran. Precedent
suggested there would be strong preference for an electrophile
initiated cyclization of 17 to yield a tetrahydrofuran rather than
a tetrahydropyran, but precedent also indicated that a preponder-
ance of the cis-2,5-disubstituted tetrahydrofuran would prevail
only under a narrow set of conditions. One of those protocols,
discovered by Bartlett and Rychnovsky,21 entails iodoetherifi-
cation of a 2,6-dichlorobenzyl 4-pentenyl ether. Pursuing this
line of inquiry, diol 17 was exposed to 2,6-dichlorobenzyl
bromide, tetra-n-butylammonium iodide, and sodium hydride
to afford bis ether 18. Treatment of 18 with iodine in acetonitrile
at low temperature then gave tetrahydrofuran 19 as the sole
detectable stereoisomer. The configuration of 19 was initially
established by NOE measurements which defined a cis orienta-
tion of the C2 hydrogen and C3 methyl group as well as a cis
relationship of the C3 proton with the iodomethyl substituent
(Figure 1). Subsequently, X-ray crystallographic analysis of 19
fully confirmed this assignment.
Results and Discussion
Synthesis of the Tetrahydrofuran Subunit. Alkynyl-
substituted tetrahydrofuran 10 (Scheme 2) seemed a logical
precursor to 9 since reliable methodology exists for transforming
a terminal alkyne into the trisubstituted (E)-iodoalkene moiety
present in 9. Establishing the 2,3,5-trisubstitution pattern of 10
was more problematic, however. A substituted furan as the point
of departure toward 10 seemed untenable, and preference was
therefore given to a pathway from an acyclic precursor in the
hope that the cis-2,5 orientation of alkynyl and alkoxymethyl
groups in 10 could be generated stereoselectively. From this
perspective, (2S,4R)-4-methyl-5-hexen-1,2-diol (11) appeared
to be an attractive entry to 10, especially since 11 can be
envisioned as the asymmetric crotylation product of (R)-
glyceraldehyde.
Treatment of the acetonide (12)18 of (R)-glyceraldehyde with
trans-2-butene and (-)-diisopinylcampheylmethoxyborane un-
der conditions described by Brown19 afforded homoallylic
alcohol 13 in good yield and with excellent stereoselectivity
(Scheme 3). However, subsequent removal of the unwanted
hydroxyl substituent from C3 of 13 proved impossible, steric
encumbrance around this site making attempts at derivatization
of the alcohol for purposes such as Barton-McCombie deoxy-
The highly stereoselective formation of 19 from 18 is to be
contrasted with an alternative pathway (Scheme 5) in which
(18) Schmid, C. R.; Bryant, J. D. J. Org. Chem. 1991, 56, 4056.
(19) Brown, H. C.; Bhat, K. S. J. Am. Chem. Soc. 1986, 108, 293.
(20) Barton, D. H. R.; McCombie, S. W. J. Chem. Soc., Perkin, Trans.
1 1975, 1574.
(21) Rychnovsky, S. D.; Bartlett, P. A. J. Am. Chem. Soc. 1981, 103,
3963. A mechanism accounting for this effect is presented by the authors
and its application to 18 has been discussed in ref 16.
J. Org. Chem, Vol. 72, No. 5, 2007 1719

White et al.
SCHEME 6. Deoxygenation and Elaboration of a
Tetrahydrofuran Subunit for Gymnodimine
FIGURE 1. Nuclear Overhauser enhancements (NOE) measured from
NMR spectra of 19 and 21.
SCHEME 5. Synthesis of a Trans-2,5-Substituted
Tetrahydrofuran
diol 17 was converted to pivalate 20 before intramolecular
iodoetherification. In that case, tetrahydrofuran 21 in which the
p-methoxybenzyl ether had been cleaved was formed exclu-
sively, its trans configuration at C2 and C5 again being assigned
on the basis of NOE measurements (Figure 1). The divergent
results from iodoetherification of 18 and 20 lend credence to
the Bartlett-Rychnovsky mechanism for cyclization of acyclic
alkenols to tetrahydrofurans which postulates that increasing
steric bulk attached to the oxygen incorporated into the
tetrahydrofuran favors cis-2,5-disubstitution.21
With a route to 19 of correct absolute configuration for
incorporation into the tetrahydrofuran substructure of gym-
nodimine now firmly established, two important tasks remained.
These were deletion of the oxygen function at C4 and removal
of the dichlorobenzyl group from its site in the C5 side chain.
The latter was required in order for this substituent to be
extended and functionalized in a manner that would facilitate
closure of the gymnodimine macrocycle at C9-C10 (linkage
B in 7). According to this scenario, functional group modifica-
tion at C2 of 19 en route to iodoalkene 9 would be delayed
until those two processes were completed. However, it was
found necessary to exchange the iodo substituent of 19 for an
alcohol before proceeding with our planned sequence and this
was accomplished by displacement of iodide from 19 with
cesium trifluoroacetate followed by treatment with diethylamine
(Scheme 6).21 Alcohol 22 was converted to its TBS ether 23
and the p-methoxybenzyl ether was cleaved with DDQ to afford
24. Reductive removal of the C4 hydroxyl group from this
substrate proved to be straightforward. Alcohol 24 was converted
to its imidazoyl thioate and the latter was reduced with tri-n-
butylstannane in the presence of a radical initiator.20 The
resulting desoxy product 25 now required cleavage of the
dichlorobenzyl ether, a transformation that was readily ac-
complished with lithium-ammonia at low temperature. Primary
alcohol 26 was oxidized under Swern conditions to aldehyde
27, which was reacted with the anion of phosphonate 28 to yield
trans-R,â-unsaturated ester 29. Catalytic hydrogenation of 29
yielded 30 from which TBS protection was removed to
give 31.
Installation of an (E)-iodoalkene chain at C2 of 31 employed
the stannylcupration-iodination protocol developed by Lip-
shutz,22 an operation that first required conversion of 31 to
alkyne 32. This was achieved by Swern oxidation of 31 to the
corresponding aldehyde and reaction of the latter with the anion
of Gilbert-Seyferth reagent 33.23 Before functionalizing the
alkyne, ester 32 was reduced to an alcohol that was protected
as TIPS ether 34. Alkyne 34 was then smoothly transformed to
iodoalkene 35, the species programmed schematically as 9 for
coupling with cyclohexene subunit 8.
Second Generation Synthesis of 9. The lengthy sequence
from glyceraldehyde acetonide 12 to tetrahydrofuran 35 por-
trayed in Schemes 3, 4, and 6 suffers from a strategic flaw
arising from the presence of an extraneous oxygen substituent
that must be erased before reaching 35. This defect prompted
consideration of a new, shorter route to 35 that avoided this
(22) Lipshutz, B. H.; Ellsworth, E. L.; Dimock, S. H.; Reuterm, D. C.
Tetrahedron Lett. 1989, 30, 2065.
(23) (a) Gibert, J. C.; Weerasooriya, U. J. Org. Chem. 1979, 44, 4997.
(b) Seyferth, D.; Marmor, R. S. Tetrahedron Lett. 1970, 11, 2493.
1720 J. Org. Chem., Vol. 72, No. 5, 2007

Synthesis of (-)-Gymnodimine
SCHEME 7. Second Generation Approach to
Tetrahydrofuran Domain of Gymnodimine
SCHEME 8. Implementation of Second Generation
Tetrahydrofuran Synthesis
impediment. Specifically, it appeared that closure of the
tetrahydrofuran from the opposite direction to that shown in
Scheme 4, i.e., from right-to-left instead of left-to-right, could
circumvent the deoxygenation problem. The new approach is
outlined in Scheme 7 and reverses the tactical moves shown in
Scheme 2 by generating tetrahydrofuran 36, absent the unwanted
C4 oxygen but with correct absolute configuration, from acyclic
precursor 37.
Pursuing this plan led to aldehyde 38 as the point of departure,
and asymmetric crotylation of this substance under Brown’s
conditions19 furnished homoallylic allylic alcohol 39 at an
acceptable level of stereochemical purity (Scheme 8). Cleavage
of silyl protection from 39 gave diol 40, which, following
precedent established with 17, was converted to its bis(2,6-
dichlorobenzyl) ether. Hydroboration-oxidation of the vinyl
group produced alcohol 41, and Dess-Martin oxidation of this
substance to an aldehyde followed by Wittig olefination gave
cyclization substrate 42. As with 18, iodoetherification of 42
resulted in highly efficient formation of a tetrahydrofuran, in
this case 43 with cis substituents at C2 and C5. A barely
SCHEME 9. Retrosynthetic Analysis of the Cyclohexene
Subunit
1
detectable (by H NMR) amount of the trans-2,5 isomer of 43
was present in the reaction product.
Convergence of 43 with material obtained by the earlier route
from glyceraldehyde acetonide was accomplished by displace-
ment of the iodo substituent with the anion of diethyl malonate,
decarboxylation of diester 44 to monoester 45, and final cleavage
of the dichlorobenzyl ether to furnish 31. This substance was
identical with the hydroxy ester prepared previously. The final
operation in the second generation sequence, cleavage of DCB
ether 45, was incompatible with the lithium-ammonium reduc-
tion used previously to cleave the dichlorobenzyl ether from
25 since competing reduction of the ester function of 45
intervened. Fortunately, an efficient cleavage of DCB ether 45
was realized with boron trichloride to yield 31. The ten-step
sequence from 38 to 31 shortens the previous pathway from 12
by five steps and delivers 31 in an overall yield of 15%.
Synthesis of the Cyclohexene Subunit. The second major
subunit required in our synthesis plan for gymnodimine was
the pentasubstituted cyclohexene 8. This fragment incorporates
three stereocenters, including the extracyclic stereogenic carbon
that will become part of the butenolide appendage of 1, a
quaternary carbon bearing functionality from which the spiroimine
can be fabricated, and a functional moiety that accommodates
bond formation with the tetrahydrofuran sector 9. After con-
sidering several routes to 8, our plan settled on a Diels-Alder
approach that would combine diene 46, already endowed with
a stereogenic center corresponding to C4 of 1, with dienophile
47 (Scheme 9). A pathway to 46 was foreseen from (S)-
glyceraldehyde derivative 48 in combination with an alkenyl-
metal reagent 49.
The preparation of diene 46 commenced with saponification
of commercially available ester 50, conversion of the resulting
J. Org. Chem, Vol. 72, No. 5, 2007 1721

White et al.
SCHEME 10. Synthesis of a 1,3-Diene for Diels-Alder
Cycloaddition
SCHEME 11. Preparation of a Dienophile from Meldrum’s
Acid
SCHEME 12. Diels-Alder Cycloaddition Leading to a
Cyclohexene Subunit for 1
carboxylic acid to a mixed anhydride with isobutyl chlorofor-
mate, and then treatment of the anhydride with N,O-dimethyl-
hydroxylamine hydrochloride in the presence of N-methylmor-
pholine (Scheme 10). Amide 5124 was produced in good yield
by this sequence. The partner for conjunction with 51 was
prepared from ethyl 2-butynoate (52) by reaction with hexa-n-
butyldistannane in the presence of n-butyllithium and copper-
(I) cyanide.25 The resulting ester 53 was reduced to alcohol 54,
which was protected as its p-methoxybenzyl ether 55. Trans-
metallation of (E)-stannane 55 with n-butyllithium followed by
addition of 51 afforded R,â-unsaturated ketone 56,26 and Wittig
olefination of this ketone then furnished diene 57.
neighboring functionality at C7. There was also the possibility
that the single (but remote) stereogenic center in 57 would steer
a dienophile toward the face of the diene that would produce
the desired (S) configuration at C7. In the event, this overly
optimistic outcome did not materialize and the reaction of 57
with 60 in acetic acid produced cycloadducts 61 and 62 in nearly
equal amounts (Scheme 12). Fortunately, 61 and 62 were easily
separable by chromatography and both were crystalline. It was
therefore possible to ascertain by X-ray crystallographic analysis
the precise stereostructure of each cycloadduct.
The choice of a dienophilic partner for Diels-Alder reaction
with 57 was made with due consideration for the fact that
cycloaddition would be required to generate a quaternary
stereogenic carbon (C22) adjacent to a hindered stereogenic
center at C7. In trial runs, unsymmetrical dienophiles reacted
sluggishly with diene 57 and showed regio- or stereoselectivity
in the cycloaddition that was too low for a feasible advance on
8. We therefore chose the highly reactive dienophile 58,27
prepared from Meldrum’s acid (59) via zwitterion 60 (Scheme
11), as the partner for 57, recognizing that this would not lead
to stereogenicity at the new quaternary carbon in the cycload-
duct, but that a distinction could probably be made later between
the pair of carboxyl substituents through their interaction with
A previous study with 62 showed that this cycloadduct could
be converted in ten steps to spiroimine 63 having correct
configuration at the spiro carbon for gymnodimine.17 Unfortu-
nately, this left us with the problem of inverting configuration
at C7 for which we saw no obvious solution that would avoid
a highly likely cyclohexene double bond migration toward C6-
C7. We therefore decided to adopt cycloadduct 61 as our
platform for departure toward substructure 8. The p-methoxy-
benzyl ether of 61 was cleaved with 2,3-dichloro-5,6-dicy-
anobenzoquinone in a reaction that led directly to lactone
carboxylic acid 64 and the acid was converted to its more
tractable methyl ester 65 with trimethylsilyldiazomethane.
Spontaneous formation of cis-fused lactone 64 upon deben-
zylation of 61 conveniently differentiated the two carbonyl
functions of 61, and efforts were therefore focused upon
reduction of 64 to exploit this felicitous result. Exposure of 65
to sodium borohydride terminated in exhaustive reduction of
the lactone ester to the undesired triol 66, but controlled
(24) Lee, C. E.; Kick, E. K.; Ellman, J. A. J. Am. Chem. Soc. 1998,
120, 9735.
(25) Piers, E.; Wong, T.; Ellis, K. A. Can. J. Chem. 1992, 70, 2058.
(26) Gilbson, C. L.; Handa, S. Tetrahedron: Asymmetry 1996, 1281.
(27) Zia-Ebrahimi, M.; Huffman, G. W. Synthesis 1996, 215.
1722 J. Org. Chem., Vol. 72, No. 5, 2007

Synthesis of (-)-Gymnodimine
SCHEME 13. Elaboration of a Cis-Fused γ-Lactone from
Diels-Alder Cycloaddition
SCHEME 14. Preparation of a Trans-Fused γ-Lactone for
Advancement of the Cyclohexene Subunit of 1
treatment of 65 with diisobutylaluminum hydride led to the more
promising hemiacetal 67 (Scheme 13). It was assumed that 67
would exist in an equilibrium mixture with its open hydroxy
aldehyde isomer and that the primary alcohol could be trapped
as a silyl ether, but to our disappointment, treatment of 67 with
triisopropylsilyl chloride gave only silyl acetal 68. On the other
hand, Horner-Wadsworth-Emmons reaction of 67 with the
anion of diethyl cyanomethylphosphonate (69)28 led transiently
to R,â-unsaturated nitrile 70, which closed spontaneously to 71
upon neutralization. Although 71 could, in principle, revert to
a ring-opened isomer, e.g., 70, via â-elimination, no means could
be found for accomplishing this transformation. Consequently,
further progress from 71 was effectively blocked.
With the preparation of lactone 76, we were in position to
move forward using chemistry already developed in Scheme
13 but without the unhappy prospect of encountering a ring-
locked species such as 71. The first task with 76 was selective
reduction of the lactone moiety without perturbing the nitrile,
a process that was accomplished with diisobutylaluminum
hydride in ether at low temperature (Scheme 15). The reduction
product was immediately revealed as a mixture of hemiacetal
stereoisomers 77 and its opened form 78. The mobile equilib-
rium interconnecting 77 and 78 permitted clean silylation of
the latter to give primary ether 79. Although this substance
contains an exposed nitrile alongside a sterically hindered
aldehyde, reaction of 79 with vinyllithium at low temperature
was completely selective, affording a 4:1 mixture of stereoiso-
meric allylic alcohols in which the Felkin-Anh isomer 80
predominated. Swern oxidation of the mixture of alcohols
furnished vinyl ketone 81, our putative coupling partner for
tetrahydrofuran subunit 9.
Fragment Coupling. Our initial plan for connecting the
tetrahydrofuran domain to the cyclohexene subunit of gym-
nodimine projected conjugate addition of a vinylcopper species
derived from 35 to enone 81. To preserve our limited quantities
of these materials, model coupling experiments were conducted
on structurally modified variants of 35 and 81. First, a truncated
version of tetrahydrofuran 35 was prepared from 25 by cleaving
the TBS ether and oxidizing the resultant primary alcohol to
aldehyde 82 (Scheme 16). This aldehyde was reacted with
Gilbert-Seyferth reagent 3323 in the presence of a base to yield
alkyne 83. Stannylcupration of 83 and subsequent reaction with
methyl iodide furnished stannane 84, which upon treatment with
iodine gave (E)-iodoalkene 85.
We now realized that spontaneous formation of cis-fused
lactone acid 64 upon debenzylation of 61 had ensnared us in a
cul-de-sac and that a different route from 61 was needed if we
were to gain access to 8. The possibility of using 61 to reach a
trans-fused γ-lactone seemed attractive, since increased strain
in that system should offer better prospect for opening the five-
membered ring and incorporating the functional substituents
needed for 8. In this context, a return to 64 provided hope,
reduction of this substance with sodium borohydride affording
diol 72 rather than triol 66 (Scheme 14). Conversion of
carboxylic acid 72 to a mixed anhydride with methyl chloro-
formate followed by treatment with triethylamine resulted in
trans-fused lactone 73 and the latter underwent Swern oxidation
to furnish crystalline aldehyde 74. X-ray crystallographic
analysis of 74 confirmed that this substance indeed possessed
a trans ring fusion. The angular aldehyde of 74 could now be
employed in a Horner-Wadsworth-Emmons condensation with
phosphonate 6928 to give (E)-R,â-unsaturated nitrile 75 in which
the disubstituted alkene was selectively hydrogenated under
carefully controlled conditions to yield nitrile lactone 76.
A coupling partner for 84 or 85 was conveniently at hand in
the form of enone 86, the C7 epimer of 81, which had been
prepared previously in the course of advancing Diels-Alder
(28) Deschamps, B.; Lefebvre, G.; Seyden-Penne, J. Tetrahedron 1972,
28, 4209.
J. Org. Chem, Vol. 72, No. 5, 2007 1723

White et al.
SCHEME 15. Synthesis of a Vinyl Ketone for Coupling
with a Tetrahydrofuran
SCHEME 17. Conjugate Addition of Alkenylstannane 84 to
Vinyl Ketone 86
Failure of our conjugate addition strategy for uniting two
major subunits of 1 caused us to revise our approach in favor
of one that did not hinge upon transmetalation of lithioalkene
87. A B-alkyl Suzuki-Miyaura coupling30 came to mind as a
possible method for linking iodoalkene 35 with a cyclohexene
fragment such as 80, but to implement this plan it was first
necessary to protect the secondary alcohol of 80. This was
accomplished by acetylation, and acetate 90 was reacted with
9-borabicyclo[3.3.1]nonane followed by degassed water to
quench excess 9-BBN (Scheme 18). Alkylborane 91 was not
isolated from this reaction but was carried forward immediately
to the coupling step with iodoalkene 35. This reaction, which
was conducted in the presence of cesium carbonate, triphenyl-
arsine, and a catalytic quantity of palladium bis(diphenylphos-
phino)ferrocene dichloride, produced 92 in 62% yield. The use
of triphenylarsine as an additive in this process was clearly
beneficial since the yield of 92 dropped to 36% when this
substance was omitted.
SCHEME 16. Synthesis of a Tetrahydrofuran for Coupling
with 81
Further progress from 92 toward a precursor from which the
macrocycle of gymnodimine could be fashioned now required
manipulation of the C7 alkyl substituent that had lain dormant
since its incorporation as silyl ether 79. An end-game for 1 was
conceived in which this substituent would be replaced by an
alkyne, the alkyne would then be transformed to an (E)-
iodoalkene, and the latter would be employed in an intramo-
lecular Nozaki-Hiyama-Kishi reaction31 with an aldehyde at
the distal terminus of the tetrahydrofuran side chain. Attempts
to realize this plan began with selective cleavage of the TES
ether from 92, which was followed by oxidation of the liberated
primary alcohol 93 to aldehyde 94 (Scheme 19). Previous
experience with an aldehyde related to 94 had informed us that
adduct 62 to spiroimine 63.17 However, when 86 was treated
with the organocopper reagent obtained by transmetalation of
stannane 84 with n-butyllithium and then a further transmeta-
lation of lithio species 87 with thienylcopper reagent 88,29 the
yield of coupled product 89 was a disappointing 25% (Scheme
17). This result was attributed, on the basis of subsequent
experiments, to inefficient formation of an organocopper species
from 87. However, replacement of 84 by iodoalkene 85 as a
coupling partner for 86 resulted in no improvement to the yield
(22%) of 89.
(30) Chemler, S. R.; Trauner, D.; Danishefsky, S. J. Angew. Chem., Int.
Ed. 2001, 40, 4544.
(31) (a) Jin, H.; Uenishi, J.-I.; Christ, W. J.; Kishi. Y. J. Am. Chem.
Soc. 1986, 108, 5644. (b) Takai, K.; Kimura, K.; Kuroda, T.; Hiyama, T.;
Nozaki, H. Tetrahedron Lett. 1983, 24, 5282.
(29) Lipshutz, B. H.; Koerner, M.; Parker, D. A. Tetrahedron Lett. 1987,
28, 945.
1724 J. Org. Chem., Vol. 72, No. 5, 2007

Synthesis of (-)-Gymnodimine
SCHEME 18. B-Alkyl Suzuki Coupling of Iodoalkene 35
with Acetate 90
SCHEME 19. Modification of C7 Functionality of Coupled
Product 92
the cyclohexene double bond readily migrates into conjugation
and therefore 94 was advanced without delay to alkyne 95 with
diazophosphonate 96.32 The C9 methyl substituent of gym-
nodimine could now be imported into 95 if methylation of the
terminal alkyne could be achieved without impacting either the
nitrile or acetate functions, but it was clear that deprotonation
of the alkyne under strongly basic conditions could not be used
for this purpose. Fortunately, a method due to Suzuki33 involving
methylation of an alkynylstannane appeared to be well suited
to our needs. Following Suzuki’s protocol, reaction of 95 with
neat tri-n-butylmethoxystannane34 at elevated temperature led
cleanly to 97 and this afforded methyl-substituted alkyne 98
upon treatment with methyl iodide in the presence of potassium
fluoride and a catalytic quantity of bis(tri-tert-butylphosphine)-
palladium. Subsequent attempts to move 98 toward a substrate
suitable for closing the macrocyclic core of 1 met difficulties
that stem from the relatively hindered environment of the alkyne.
For example, attempted silylcupration-iodination,35 hydrozir-
conation-iodination,36 and palladium-catalyzed hydrostanny-
lation-iodination37 of 98 all returned unreacted starting material.
It is clear from these failures that alkyne 98 is not a viable entity
for completing our route to 1 and that a different strategy must
be found for creating the macrocyclic portion of gymnodimine.
This will be the subject of future efforts directed toward the
total synthesis of this intriguing structure.
In summary, a convergent approach to an advanced inter-
mediate containing the C3-C32 sector of gymnodimine has
been developed. In complementing a previous study that led to
the spiroimine nucleus of 1, the present work lays groundwork
upon which the final stages of a completed synthesis of
gymnodimine can be based.
(32) (a) Ohira, S. Synth. Commun. 1989, 19, 561. (b) Muller, S.; Liepold,
B.; Roth, G. J.; Bestmann, H. J. Synlett 1996, 521.
(33) Hosoya, T.; Wakao, M.; Kondo, Y.; Doi, H.; Suzuki, M. Org.
Biomol. Chem. 2004, 2, 24.
(34) Logue, M. W.; Teng, K. J. Org. Chem. 1982, 47, 2549.
(35) Fleming, I.; Newton, T. W.; Roessler, F. J. Chem. Soc., Perkin Trans.
I 1981, 1, 2527.
(36) (a) Panek, J. S.; Hu, T. J. Org. Chem. 1997, 62, 4912. (b) Hart,
D. W.; Blackburn, T. F.; Schwartz, J. J. Am. Chem. Soc. 1975, 97, 680.
(37) Semmelhack, M. F.; Hooley, R. J. Tetrhedron Lett. 2003, 44, 5737.
Experimental Section
(2S,4R,5R)-2-Iodomethyl-5-(2,6-dichlorobenzyloxymethyl)-4-
methyltetrahydrofuran (43). A solution of alkene 42 (1.60 g, 3.59
J. Org. Chem, Vol. 72, No. 5, 2007 1725

White et al.
mmol) in dry CH3CN (300 mL) was cooled to -20 °C and a
solution of iodine (1.4 g, 5.38 mmol) in dry CH3CN (50 mL) was
added slowly via syringe. The dark-brown mixture was stirred for
50.3, 55.7, 67.5, 68.9, 73.5, 74.3, 105.7, 109.4, 114.1, 128.5, 129.2,
129.7, 129.9, 159.7, 165.5, 170.7; HRMS (mixture) (CI) m/z
474.2249 ([M]+; calcd for C26H34O8 474.2254).
3 h at -20 °C and then diluted with Et2O and saturated Na2S2O3
(50 mL). The phases were separated and the aqueous phase was
extracted with Et2O (2 × 200 mL). The combined organic layer
was washed with brine, dried over Na2SO4, filtered, and concen-
trated. Flash chromatography (hexane-EtOAc, 10:1) of the residue
gave 1.24 g (84%) of 43: [R]2D3 -24.8 (c 1.25, CHCl3); IR (neat)
2957, 2927, 2872, 1579, 1564, 1437, 1197, 1101, 994, 778, 767
(2R,3R,5S)-5-(2,6-Dichlorobenzyloxymethyl)-2-ethynyl-3-me-
thyltetrahydrofuran (83). To a solution of potassium tert-butoxide
(1M in i-PrOH, 4.3 mL, 4.3 mmol) in THF (10 mL) at -78 °C
was added a solution of diethyl (diazomethyl)-phosphonate 33 (0.77
g, 4.3 mmol) in THF (2 mL). The mixture was stirred for 5 min,
and aldehyde 82 (0.66 g, 2.2 mmol) was added slowly. The mixture
was allowed to slowly warm to room temperature and the reaction
was quenched by addition of water (5 mL). The solvent was
removed under vacuum and the residue was partitioned between
EtOAc and water. The phases were separated, the aqueous phase
was extracted with EtOAc (2×), and the combined organic layer
was washed with brine, dried over Na2SO4, and concentrated under
vacuum. Flash chromatography (hexane-EtOAc, 15:1) of the
residue afforded 0.51 g (79%) of 83: [R]2D3 -20.7 (c 1.04,
CHCl3); IR (neat) 3302, 2961, 2931, 2874, 1735, 1563, 1437, 1197,
1102, 779, 767 cm-1; 1H NMR (300 MHz, CDCl3) δ 1.13 (d, J )
6.7 Hz, 3H), 1.70 (m, 1H), 2.09 (m, 1H), 2.38 (m, 1H), 2.48 (d,
J ) 1.9 Hz, 1H), 3.55 (dd, J ) 5.8, 9.9 Hz, 1H), 3.70 (dd, J ) 6.0,
10.0 Hz, 1H), 4.07 (dd, J ) 2.2, 7.5 Hz, 1H), 4.25 (m, 1H), 4.85
(d, J ) 10.7 Hz, 1H), 4.90 (d, J ) 10.7 Hz, 1H), 7.21 (dd, J ) 7.7,
8.4 Hz, 1H), 7.35 (d, J ) 8.0 Hz, 2H); 13C NMR (100 Mz, CDCl3)
δ 17.1, 36.8, 41.1, 66.3, 68.2, 73.5, 75.5, 83.5, 128.8, 130.3, 135.7,
137.3; HRMS (CI) m/z 298.0524 ([M]+; calcd for C15H16O2Cl2
298.0527).
1
cm-1; H NMR (300 MHz, CDCl3) δ 1.05 (d, J ) 7.7 Hz, 3H),
1.77 (ddd, J ) 7.7, 12.7, 15.4 Hz, 1H), 2.02 (ddd, J ) 5.1, 8.1,
12.8 Hz, 1H), 2.19 (m, 1H), 3.17 (dd, J ) 8.0, 10.0 Hz, 1H), 3.28
(dd, 4.8, 10.0 Hz, 1H), 3.64 (m, 2H), 3.71 (m, 1H), 4.16 (m, 1H),
4.78 (d, J ) 11.8 Hz, 1H), 4.82 (d, J ) 11.8 Hz, 1H), 7.22 (dd,
J ) 7.6, 8.4 Hz, 1H), 7.34 (d, J ) 8.0 Hz, 2H); 13C NMR (100
Mz, CDCl3) δ 11.3, 17.9, 35.3, 40.1, 68.2, 72.6, 78.4, 86.5, 128.8,
130.3, 133.9, 137.3; HRMS (FAB) m/z 414.9728 ([M + H]+; calcd
for C14H18O2Cl2I 414.9729).
9-(2′,2′-Dimethyl[1′,3′]dioxolan-4′-yl)-7-(4′′-methoxybenzyl-
oxymethyl)-3,3,8-trimethyl-2,4-dioxaspiro[5.5]undec-8-en-1,5-di-
ones (61 and 62). To a solution of 60 (172 mg, 0.54 mmol) in
anhydrous EtOH (8 mL) was added glacial acetic acid (37 µL, 0.65
mmol) followed by a solution of diene 57 (153 mg, 0.65 mmol) in
EtOH (4 mL). The mixture was stirred for 48 h at room temperature,
the solvent was removed under vacuum, and the residue was diluted
with cold water and EtOAc. The separated organic layer was washed
with brine, dried over Na2SO4, filtered, and concentrated under
vacuum. Flash chromatography (hexane-EtOAc, 1:1) of the residue
afforded 217 mg (85%) of a mixture (1.2:1) of 61 and 62. The
mixture was separated by spinning rotor chromatography (CH2-
Cl2-EtOAc, 100:5) to provide pure 61 and 62.
(2R,3R,5S)-5-(2,6-Dichlorobenzyloxymethyl)-2-[(E)-2-iodo-1-
methylvinyl]-3-methyltetrahydrofuran (85). To flame-dried CuCN
(33 mg, 0.37 mmol) was added THF (1.5 mL) under argon, and
the mixture was cooled to -78 °C. n-BuLi (2.58 M, 0.29 mL, 0.74
mmol) was added slowly, and the resulting mixture was stirred for
10 min at -65 °C. The suspension gradually became a homoge-
neous, colorless solution that was re-cooled to -78 °C. To this
solution was added n-Bu3SnH (0.13 mL, 0.74 mmol), and the
resulting mixture was stirred for 10 min, during which time the
solution became yellow. A solution of alkyne 83 (32 mg, 0.11
mmol) in THF (0.5 mL) was slowly added to the reaction mixture.
After the resulting mixture was stirred for 30 min at -78 °C, MeI
(50 µL) and DMPU (80 µL) were added. This mixture was allowed
to warm slowly to room temperature. The reaction was quenched
by addition of a solution of saturated NH4Cl/NH3‚H2O (9/1), and
the mixture was diluted with EtOAc. The phases were separated
and the aqueous phase was extracted with EtOAc (2×). The organic
layer was washed with brine, dried over Na2SO4, and concentrated
under vacuum to give vinyl stannane 84.
(4′R,7S)-61: [R]2D3 +8.6 (c 1.05, CHCl3); IR (mixture) (neat)
2986, 2936, 2873, 1770, 1736, 1613, 1514, 1456, 1379, 1304, 1278,
1249, 1209, 1157, 1055, 1033, 860, 822, 735 cm-1; 1H NMR (300
MHz, CDCl3) δ 1.35 (s, 3H), 1.38 (s, 3H), 1.41 (s, 3H), 1.61 (s,
3H), 1.69 (s, 3H), 1.97 (m, 1H), 2.08 - 2.32 (m, 3H), 3.44 (dd,
J ) 8.0, 11.2 Hz, 1H), 3.45 (t, J ) 8.0 Hz, 1H), 3.58 (m, 1H), 3.77
(s, 3H), 3.76 (m, 1H), 4.05 (dd, J ) 7.0, 8.0 Hz, 1H), 4.34 (s, 2H),
5.01 (t, J ) 7.0 Hz, 1H), 6.83 (m, 2H), 7.18 (m, 2H); 13C NMR
(100 Mz, CDCl3) δ 16.7, 21.1, 25.7, 26.8, 28.9, 29.8, 31.8, 45.7,
50.2, 55.6, 67.8, 68.5, 73.4, 75.2, 105.7, 109.5, 114.1, 127.0, 129.6,
129.8, 129.9, 159.6, 165.5, 170.7; HRMS (mixture) (CI) m/z
474.2249 ([M]+; calcd for C26H34O8 474.2254).
(4′R,7R)-62: [R]2D3 -10.4 (c 0.5, CHCl3); IR (mixture) (neat)
2986, 2936, 2873, 1770, 1736, 1613, 1514, 1456, 1379, 1304, 1278,
1249, 1209, 1157, 1055, 1033, 860, 822, 735 cm-1; 1H NMR (300
MHz, CDCl3) δ 1.36 (s, 3H), 1.39 (s, 3H), 1.43 (s, 3H), 1.67 (s,
3H), 1.71 (s, 3H), 1.98-2.27 (m, 3H), 2.39-2.53 (m, 1H), 3.37
(m, 1H), 3.65 (t, J ) 8.0 Hz, 1H), 3.79 (s, 3H), 3.70-3.82 (m,
2H), 4.00 (t, J ) 8.0 Hz, 1H), 4.37 (m, 2H), 5.04 (t, J ) 7.0 Hz,
1H), 6.85 (d, J ) 7.9 Hz, 2H), 7.18 (d, J ) 7.8 Hz, 2H); 13C NMR
(100 Mz, CDCl3) δ 16.4, 20.1, 26.0, 26.7, 28.9, 29.7, 31.8, 45.7,
To a solution of stannane 84 in THF (1 mL) at 0 °C was added
I2 (28 mg, 0.11 mmol), and the mixture was stirred for 30 min at
room temperature. The mixture was diluted with water and extracted
with Et2O. The combined organic layers were washed with brine,
dried over Na2SO4, and concentrated under vacuum. Flash chro-
matography (hexane-EtOAc, 40:1) of the residue yielded 40 mg
(86%) of 85: [R]2D3 -13.8 (c 0.16, CHCl3); IR (neat) 2957, 2928,
1563, 1437, 1197, 1102, 1044, 777, 768 cm-1; 1H NMR (300 MHz,
CDCl3) δ 1.01 (d, J ) 6.4 Hz, 3H), 1.66 (m, 1H), 1.82 (s, 3H),
1726 J. Org. Chem., Vol. 72, No. 5, 2007

Synthesis of (-)-Gymnodimine
2.05 (m, 2H), 3.60 (m, 2H), 3.90 (d, J ) 8.7 Hz, 1H), 4.25 (m,
1H), 4.81 (d, J ) 10.8 Hz, 1H), 4.87 (d, J ) 10.8 Hz, 1H), 6.29 (s,
1H), 7.21 (m, 1H), 7.36 (d, J ) 8.1 Hz, 2H); 13C NMR (100 Mz,
CDCl3) δ 17.4, 20.1, 36.9, 37.2, 68.0, 73.7, 77.8, 79.1, 91.1, 128.8,
130.3, 133.8, 137.3, 147.6; HRMS (CI) m/z 439.9811 ([M - H]+;
calcd for C16H19Cl2IO2 439.9807).
1.91-2.07 (m, 3H), 2.14 (s, 3H), 2.29 (dd, J ) 7.2, 16.1 Hz, 1H),
2.58 (m, 2H), 3.58 (dd, J ) 2.1, 11.4 Hz, 1H), 3.62-3.68 (m, 2H),
4.06 (dd, J ) 6.8, 8.1 Hz, 1H), 5.01 (t, J ) 7.5 Hz, 1H), 5.15 (d,
J ) 17.0 Hz, 1H), 5.26 (d, J ) 6.6 Hz, 1H), 5.31 (d, J ) 10.5 Hz,
1H), 5.80 (ddd, J ) 6.6, 10.5, 17.1 Hz, 1H); 13C NMR (100 Mz,
CDCl3) δ 4.7, 7.3, 13.5, 19.3, 19.9, 21.6, 24.2, 26.0, 26.7, 31.2,
41.0, 50.2, 62.3, 67.1, 75.0, 76.3, 109.4, 119.6, 121.2, 128.7, 130.0,
133.2, 170.3; HRMS (ES) m/z ([M]+ 491.3058; calcd for C27H45-
NO5Si 491.3067).
3-((1R,2R)-1-((E)-5-((2R,3R,5S)-Tetrahydro-5-(2,6-dichloroben-
zyloxylmethyl)-3-methyfuran-2-yl)hex-4-enoyl)-2-(triethylsila-
nyloxymethyl)-3-methyl-4-((R)-2,2-dimethyl-1,3-dixolan-4-yl)-
cyclohex-3-enyl)propanenitrile (89). To a solution of vinylstannane
84 (10.5 mg, 17.4 µmol) in THF (0.1 mL) at -78 °C was added
n-BuLi (7 µL of a 2.5 M solution in hexane, 17.5 µmol). The
mixture was stirred for 1 h and a solution of 2-thienyl(cyano)copper
lithium (88, 70 µL of a 0.25 M solution in THF, 17.4 µmol) was
added. The mixture was stirred for 10 min at -78 °C and a solution
of 86 (6.5 mg, 14.5 µmol) in THF (0.1 mL) was added. After 2 h
at -78 °C, the reaction was quenched by addition of aqueous
saturated NaHCO3. The mixture was diluted with EtOAc, the phases
were separated, and the aqueous layer was extracted with EtOAc
(2×). The combined organic extract was washed with saturated
NaHCO3 and brine, dried over Na2SO4, and concentrated under
vacuum. Flash chromatography (hexane-EtOAc, 4:1) of the residue
afforded 2.8 mg (25%) of 89: IR (neat) 2954, 2930, 2875, 2248,
1700, 1582, 1564, 1456, 1437, 1378, 1260, 1210, 1156, 1101, 1061,
(4E)-1-((1R,2S)-1-(2-Cyanoethyl)-2-(triethylsilanyloxymethyl)-
3-methyl-4-((R)-2,2-dimethyl-1,3-dioxolan-4-yl)cyclohex-3-enyl)-
5-((2R,3R,5R)-tetrahydro-5-(3-triisopropylsilanyloxypropyl)-3-
methyfuran-2-yl)hex-4-enyl Acetate (92). To alkene 90 (27 mg,
55 µmol) was added 9-borabicyclo[3.3.1]nonane (0.44 mL of a 0.5
M solution in THF, 220 µmol) and the mixture was stirred at room
temperature for 3 h. Degassed water (16 µL) was carefully added
and the mixture was stirred at room temperature for 1 h. In a
separate flask, a solution of Cs2CO3 (35 mg, 108 µmol), PdCl2-
(dppf) (8.3 mg, 11 µmol), AsPh3 (3.3 mg, 11 µmol), and iodide 35
(31 mg, 66 µmol) in DMF (0.2 mL) was prepared and the THF
solution of borane 91 was added. The mixture was stirred at room
temperature for 10 h and then was diluted with EtOAc. The mixture
was washed with water and brine, dried over Na2SO4, and
concentrated. Flash chromatography (hexane-EtOAc, 4:1) of the
residue gave 28.1 mg (62%) of 92: [R]2D3 -14.7 (c 0.30, CHCl3);
IR (neat) 2954, 2918, 2867, 2849, 2245, 1737, 1590, 1461, 1375,
1236, 1159, 1059, 1016, 857, 810, 752 cm-1; 1H NMR (300 MHz,
CDCl3) δ 0.59 (q, J ) 7.6 Hz, 6H), 0.94 (d, J ) 6.4 Hz, 3H), 0.96
(t, J ) 7.7 Hz, 9H), 0.91-1.00 (m, 3H), 1.05 (s, 18H), 1.38 (s,
3H), 1.40 (s, 3H), 1.54 (s, 3H), 1.75 (s, 3H), 2.08 (s, 3H), 1.50-
2.25 (m, 18H), 2.56 (m, 2H), 3.55-3.80 (m, 6H), 3.97 (m, 1H),
4.05 (t, J ) 7.7 Hz, 1H), 4.98 (d, J ) 10.2 Hz, 1H), 5.00 (t, J )
7.7 Hz, 1H), 5.36 (t, J ) 8.6 Hz, 1H); 13C NMR (75 MHz, CDCl3)
δ 4.7, 7.3, 11.6, 12.4, 13.7, 14.5, 14.5, 17.6, 18.4, 19.4, 21.5, 22.7,
23.1, 26.1, 26.7, 29.9, 32.0, 33.2, 36.4, 40.3, 41.2, 50.8, 63.8, 67.0,
75.2, 75.3, 78.1, 92.6, 109.3, 121.2, 126.6, 128.5, 130.0, 135.8,
171.0; HRMS (ES) m/z ([M + Na]+ 854.5772; calcd for C47H85-
NO7NaSi2 854.5762).
1
1017, 863, 813, 767, 746 cm-1; H NMR (400 MHz, CDCl3) δ
0.54 (q, J ) 7.9 Hz, 6H), 0.92 (t, J ) 7.9 Hz, 9H), 1.28 (s, 3H),
1.31 (m, 2H), 1.43 (s, 3H), 1.48 (s, 3H), 1.64, (s, 3H), 1.66 (m,
1H), 1.77 (m 1H), 1.79 (s, 3H), 1.95-2.4 (m, 10H), 2.6 (m, 1H),
3.44-3.55 (m, 4H), 3.62 (dd, J ) 5.2, 9.7 Hz, 1H), 3.68 (d, J )
8.0 Hz, 1H), 4.05 (t, J ) 7.6 Hz, 1H), 4.20 (m, 1H), 4.81 (d, J )
10.6 Hz, 1H), 4.85 (d, J ) 10.6, 1H), 5.04 (t, J ) 7.2 Hz, 1H),
5.39 (t, J ) 6.5 Hz, 1H), 7.21 (t, J ) 8.0 Hz, 1H), 7.34 (d, J ) 8.0
Hz, 2H); HRMS (FAB) m/z 760.3576 ([M - H]+; calcd for
C41H60O6N35Cl2Si 760.3567).
1-((1R,2S)-1-(2-Cyanoethyl)-2-(triethylsilanyloxymethyl)-3-
methyl-4-((R)-2,2-dimethyl-1,3-dioxolan-4-yl)cyclohex-3-enyl)-
allyl Acetate (90). To a solution of 80 (18.5 mg, 0.044 mmol) in
CH2Cl2 (0.3 mL) at 0 °C was added Et3N (0.1 mL, 0.71 mmol),
DMAP (2.5 mg), and Ac2O (2 mL, 0.21 mmol). The mixture was
stirred for 2 h at room temperature and the reaction was quenched
by addition of a saturated solution of NH4Cl. The mixture was
diluted with CH2Cl2, the two phases were separated, and the aqueous
layer was extracted with CH2Cl2 (2×). The combined organic
extract was washed with water and brine, dried over Na2SO4, and
concentrated. Flash chromatography (hexane-EtOAc, 4:1) of the
residue furnished 18.6 mg (93%) of 90: [R]2D3 +15.7 (c 0.81,
CHCl3); IR (neat) 2954, 2924, 2875, 2248, 1724, 1456, 1370, 1233,
(4E)-1-((1R,2S)-1-(2-Cyanoethyl)-2-(hydroxymethyl)-3-meth-
yl-4-((R)-2,2-dimethyl-1,3-dioxolan-4-yl)cyclohex-3-enyl)-5-
((2R,3R,5R)-tetrahydro-5-(3-triisopropylsilanyloxypropyl)-3-
methyfuran-2-yl)hex-4-enyl Acetate (93). A solution of 92 (28.1
mg, 33 µmol) in a mixture of H2O/HOAc/THF (3/5/11, 1.5 mL) at
0 °C was stirred for 7 h. The reaction was quenched by addition of
saturated NaHCO3 and the mixture was diluted with EtOAc. The
layers were separated and the aqueous layer was extracted with
EtOAc (2×). The combined organic extract was washed with brine,
1
1157, 1056, 1081, 970, 861, 808, 748 cm-1; H NMR (400 MHz,
CDCl3) δ 0.65 (q, J ) 7.9 Hz, 6H), 0.99 (t, J ) 8.0 Hz, 9H), 1.41
(s, 3H), 1.44 (s, 3H), 1.59 (m, 1H), 1.78 (s, 3H), 1.79 (m, 2H),
J. Org. Chem, Vol. 72, No. 5, 2007 1727

White et al.
dried over Na2SO4, and concentrated. Flash chromatography
(hexane-EtOAc, 2:1) of the residue gave 20.1 mg (83%) of 93:
[R]2D3 -21.8 (c 0.28, CHCl3); IR (neat) 3393, 2928, 2865, 2248,
1734, 1451, 1434, 1383, 1236, 1158, 1097, 1056, 1032, 882, 858,
MHz, CDCl3) δ 0.97 (d, J ) 6.7 Hz, 3H), 1.06-1.10 (m, 3H),
1.09 (s, 18H), 1.42 (s, 3H), 1.45 (s, 3H), 1.60 (s, 3H), 1.55-1.71
(m, 7H), 1.75-1.86 (m, 2H), 1.85 (s, 3H), 1.89-2.16 (m, 7H),
2.11 (s, 3H), 2.21 (d, J ) 2.5 Hz, 1H), 2.25 (m, 1H), 2.50 (m, 1H),
2.62 (m, 1H), 2.81 (s, 1H), 2.62 (d, J ) 5.8 Hz, 1H), 3.64 (dd, J
) 2.3, 8.1 Hz, 1H), 3.73 (t, J ) 5.7 Hz, 2H), 4.06 (dd, J ) 6.7, 8.1
Hz, 1H), 4.39 (m, 1H), 5.96-5.05 (m, 2H), 5.35 (t, J ) 7.1 Hz,
1H); 13C NMR (100 Mz, CDCl3) δ 11.7, 12.4, 13.3, 17.6, 18.1,
18.5, 20.2, 21.4, 24.8, 25.6, 26.0, 26.7, 29.9, 30.3, 30.5, 33.2, 36.5,
40.2, 41.0, 41.1, 63.8, 67.4, 73.2, 75.0, 75.1, 78.1, 82.8, 92.6, 109.6,
120.7, 126.4, 127.6, 129.4, 136.1, 170.9; HRMS (ES) m/z ([M +
H]+ 712.4984; calcd for C42H70NO6Si 712.4972).
1
794 cm-1; H NMR (300 MHz, CDCl3) δ 0.97 (d, J ) 6.6 Hz,
3H), 1.07 (m, 3H), 1.09 (s, 18H), 1.42 (s, 3H), 1.45 (s, 3H), 1.52-
1.78 (m, 12H), 1.58 (s, 3H), 1.80 (s, 3H), 1.85-2.10 (m, 6H), 2.12
(s, 3H), 2.26 (m, 1H), 2.57 (m, 2H), 3.63 (d, J ) 8.1 Hz, 1H),
3.70-3.82 (m, 5H), 3.98 (m, 1H), 4.09 (dd, J ) 6.7, 8.2 Hz, 1H),
4.98 (d, J ) 8.7 Hz, 1H), 5.05 (t, J ) 7.3 Hz, 1H), 5.34 (t, J ) 6.7
Hz, 1H); 13C NMR (75 MHz, CDCl3) δ 11.3, 12.0, 13.1, 17.2, 18.1,
18.6, 19.5, 21.1, 24.2, 24.5, 25.7, 26.3, 29.5, 30.3, 30.5, 32.8, 36.1,
39.9, 40.6, 49.7, 61.0, 63.4, 66.7, 74.8, 77.7, 92.2, 109.0, 120.7,
126.1, 127.5, 131.2, 135.6, 170.6; HRMS (ES) m/z ([M + Na]+
740.4905; calcd for C41H71NO7NaSi 740.4898).
(4E)-1-((1R,2S)-1-(2-Cyanoethyl)-3-methyl-4-((R)-2,2-dimethyl-
1,3-dioxolan-4-yl)-2-(prop-1-ynyl)cyclohex-3-enyl)-5-((2R,3R,5R)-
tetrahydro-5-(3-triisopropylsilanyloxypropyl)-3-methyfuran-2-
yl)hex-4-enyl Acetate (98). A solution of 95 (16.8 mg, 0.024 mmol)
in Bu3SnOMe (0.05 mL) was stirred for 36 h at 85 °C. After the
solution had cooled to room temperature, flash chromatography
(hexane-EtOAc, 4:1) of the mixture gave 16.0 mg (68%) of
stannane 97.
(4E)-1-((1R,2S)-1-(2-Cyanoethyl)-2-ethynyl-3-methyl-4-((R)-
2,2-dimethyl-1,3-dioxolan-4-yl)cyclohex-3-enyl)-5-((2R,3R,5R)-
tetrahydro-5-(3-triisopropylsilanyloxypropyl)-3-methyfuran-2-
yl)hex-4-enyl Acetate (95). To a solution of alcohol 93 (22.4 mg,
31 µmol) in CH2Cl2 (1.0 mL) at 0 °C was added Dess-Martin
periodinane (27 mg, 62 µmol). After removal of the cooling bath,
the mixture was stirred for 1 h and then diluted with Et2O. Most of
the solvent was removed under vacuum, and the residue was taken
up in Et2O. A mixture of 10% Na2S2O3/saturated NaHCO3 (1/1)
was added and the mixture was stirred at room temperature for 30
min. The two phases were separated and the organic layer was
washed with water and brine. The aqueous washings were back-
extracted with Et2O, and the organic extract was washed with water
and brine. The combined Et2O extract was dried over Na2SO4, and
concentrated to give 20.5 mg (92%) of aldehyde 94: 1H NMR (300
MHz, CDCl3) δ 0.93 (d, J ) 6.6 Hz, 3H), 1.04 (m, 3H), 1.05 (s,
18H), 1.39 (s, 3H), 1.44 (s, 3H), 1.50-1.82 (m, 10 H), 1.54 (s,
3H), 1.61 (s, 3H), 1.83-2.09 (m, 6H), 2.10 (s, 3H), 2.36-2.62
(m, 3H), 2.66 (d, J ) 4.3 Hz, 1H), 3.61 (d, J ) 8.1 Hz, 1H), 3.63-
3.72 (m, 3H), 3.95 (m, 1H), 4.08 (dd, J ) 6.7, 8.2 Hz, 1H), 4.94
(dd, J ) 3.4, 8.8 Hz, 1H), 5.05 (t, J ) 7.1 Hz, 1H), 5.30 (t, J ) 6.7
Hz, 1H), 9.47 (d, J ) 4.9 Hz, 1H).
To a solution of (1-diazo-2-oxopropyl)phosphonic acid dimethyl
ester (96, 23 mg, 120 µmol) in THF (0.5 mL) at -78 °C was added
a solution of sodium methoxide (40 µL of a 0.26 M solution in
THF, 10.4 µmol), followed by slow addition of a solution of
aldehyde 94 (20.5 mg, 29 µmol) in THF (0.3 mL). After the cooling
bath was removed, the mixture was stirred at room temperature
for 30 min. A saturated NH4Cl solution was added and the mixture
was stirred vigorously for 20 min and then diluted with EtOAc.
The phases were separated and the aqueous layer was extracted
with EtOAc (2×). The combined organic extract was washed with
brine, dried over Na2SO4, and concentrated. Flash chromatography
(hexane-EtOAc, 4:1) of the residue afforded 16.8 mg (83%) of
95: [R]2D3 +14.1 (c 0.17, CHCl3); IR (neat) 3308, 2931, 2865,
2352, 2246, 1733, 1699, 1652, 1471, 1452, 1436, 1419, 1394, 1372,
1233, 1153, 1095, 1057, 1032, 881, 854, 787 cm-1; 1H NMR (400
A mixture of Pd(t-Bu3P)2 (0.1 mg) and KF (0.5 mg) in 0.1 mL
of DMF was stirred for 5 min at room temperature, after which
MeI (4.5 µL of a 1 M solution in DMF, 4.5 µmol) and a solution
of stannane 97 (3.0 mg, 3.0 µmol) in DMF (0.1 mL) were added.
The mixture was stirred for 5 min at 60 °C, then was cooled to
room temperature and diluted with EtOAc. The solution was washed
with water and brine, dried over Na2SO4, and concentrated. Flash
chromatography (hexane-EtOAc, 4:1) of the residue furnished 1.4
mg (64%) of 98: [R]2D3 +22.7 (c 0.41, CHCl3); IR (neat) 2951,
2853, 2242, 1740, 1462, 13777, 1234, 1110, 1060, 1039 cm-1; 1H
NMR (400 MHz, CDCl3) δ 0.98 (d, J ) 6.6 Hz, 3H), 1.08 (m,
3H), 1.09 (s, 18H), 1.42 (s, 3H), 1.46 (s, 3H), 1.60 (s, 3H), 1.56-
1.72 (m, 11H), 1.82 (s, 3H), 1.83 (s, 3H), 1.87-2.07 (m, 6H), 2.11
(s, 3H), 2.23 (m, 1H), 2.52 (m, 1H), 2.63 (m, 1H), 2.71 (s, 1H),
3.61-3.67 (m, 2H), 3.74 (t, J ) 5.7 Hz, 2H), 3.99 (m, 1H), 4.04
(d, J ) 6.7, 8.0 Hz, 1H), 4.98 (dd, J ) 2.2, 9.7 Hz, 1H), 5.03 (t,
J ) 7.4 Hz, 1H), 5.35 (t, J ) 7.0 Hz, 1H); 13C NMR (100 Mz,
CDCl3) δ 3.7, 11.3, 12.0, 12.9, 17.2, 18.1, 18.2, 19.8, 21.1, 24.5,
25.1, 25.7, 26.3, 29.5, 30.2, 32.8, 36.0, 39.8, 40.8, 41.0, 63.4, 67.0,
74.7, 74.8, 78.0, 92.2, 109.0, 120.6, 126.2, 127.9, 128.5, 135.6,
170.5; HRMS (ES) m/z ([M + H]+ 726.5110; calcd for C43H72-
NO6Si 726.5129).
Acknowledgment. LQ is grateful to the Swiss National
Science Foundation for a Postdoctoral Fellowship. Financial
support for this work was provided by the National Institute of
General Medical Sciences through grant GM58889.
Supporting Information Available: Experimental procedures
and characterization data for compounds not included in the paper,
and 1H and 13C NMR spectra of new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
JO062396O
1728 J. Org. Chem., Vol. 72, No. 5, 2007