An intramolecular nitrone-olefin dipolar cycloaddition-based approach to total synthesis of the cylindricine and lepadiformine marine alkaloids

Article abstract of DOI:10.1021/jo990266s

A synthetic route to the cylindricine skeleton as well as to the reported structure of the marine alkaloid lepadiformine has been achieved using an intramolecular nitrone/1,3-diene dipolar cycloaddition as the key step. The synthesis began with sequential alkylations of acetone oxime to afford key intermediate oxime 30, which contains all of the carbons necessary to form the tricyclic skeleton of the alkaloids. Nitrone 40, available from oxime 30 by standard transformations, underwent an intramolecular 1,3- dipolar cycloaddition to provide isoxazolidine 43. Related 1,3-dipolar cycloadditions were also explored on two additional nitrone-olefin substrates 41 and 42, which were prepared in a manner similar to that of 40. The tricyclic alkaloid core 52 was formed stereoselectively by a tandem oxidation-Michael addition of amino alcohol 49 derived from isoxazolidine 43. Cleavage of the O-phenyl ether of 52 provided 2-epi-cylindricine C (53). Several unsuccessful attempts were made to convert 52 to cylindricine C by epimerization at C2. Tricyclic ketone 52 was deoxygenated to give amine 59, whose structure and relative stereochemistry were confirmed by single- crystal X-ray analysis of its picrate salt. Removal of the O-phenyl protecting group from 59 provided tricyclic amino alcohol 60 having the putative structure of lepadiformine, but whose NMR data did not correspond to those of the natural product.

Full text of DOI:10.1021/jo990266s

J . Org. Chem. 1999, 64, 4865-4873
4865
An In tr a m olecu la r Nitr on e-Olefin Dip ola r Cycloa d d ition -Ba sed
Ap p r oa ch to Tota l Syn th esis of th e Cylin d r icin e a n d
Lep a d ifor m in e Ma r in e Alk a loid s
Kim M. Werner, J esus M. de los Santos, and Steven M. Weinreb*
Department of Chemistry, The Pennsylvania State University, University Park, Pennsylvania 16802
Maoyu Shang†
Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, Indiana 46556
Received February 12, 1999
A synthetic route to the cylindricine skeleton as well as to the reported structure of the marine
alkaloid lepadiformine has been achieved using an intramolecular nitrone/1,3-diene dipolar
cycloaddition as the key step. The synthesis began with sequential alkylations of acetone oxime to
afford key intermediate oxime 30, which contains all of the carbons necessary to form the tricyclic
skeleton of the alkaloids. Nitrone 40, available from oxime 30 by standard transformations,
underwent an intramolecular 1,3-dipolar cycloaddition to provide isoxazolidine 43. Related 1,3-
dipolar cycloadditions were also explored on two additional nitrone-olefin substrates 41 and 42,
which were prepared in a manner similar to that of 40. The tricyclic alkaloid core 52 was formed
stereoselectively by a tandem oxidation-Michael addition of amino alcohol 49 derived from
isoxazolidine 43. Cleavage of the O-phenyl ether of 52 provided 2-epi-cylindricine C (53). Several
unsuccessful attempts were made to convert 52 to cylindricine C by epimerization at C2. Tricyclic
ketone 52 was deoxygenated to give amine 59, whose structure and relative stereochemistry were
confirmed by single-crystal X-ray analysis of its picrate salt. Removal of the O-phenyl protecting
group from 59 provided tricyclic amino alcohol 60 having the putative structure of lepadiformine,
but whose NMR data did not correspond to those of the natural product.
In the early 1990s, Blackman and co-workers described
the isolation of a series of structurally related alkaloids
from the marine ascidian Clavelina cylindrica, collected
at sites off the coast of Tasmania.1 The most abundant
of these alkaloids are cylindricines A (1) and B (7), whose
structures were firmly established by X-ray crystal-
lography of the corresponding picrates. Interestingly,
possess a butyl chain at C2 rather than the hexyl group
(e.g. cylindricines H (8) and I (9)). Similarly, compounds
in the cylindricine B (pyridoquinoline) series exist which
also have a butyl appendage (e.g. cylindricine J (10)). It
might be noted that the X-ray crystal structure and NMR
data, as well as molecular mechanics calculations,1c
indicate that these molecules prefer to exist in the
conformations shown.
In 1994, Biard et al. reported the isolation of lepadi-
formine from the tunicate Clavelina lepadiformis Muller
collected in the Mediterranean near Tunisia.2 It was
found that this material has moderate in vitro cytotoxic
activity against various tumor cell lines. On the basis of
spectroscopic data, the structure of lepadiformine was
formulated as the zwitterionic pyrroloquinoline 11, re-
lated to the cylindricine A series. Thus, lepadiformine is
epimeric at C2 and lacks the C4 oxygen of the cylindri-
cines. Also, NMR NOESY data suggested that lepadi-
formine has conformation 11a , rather than the flip form
11b.2 However, we have performed molecular mechanics
these two compounds exist as a 3:2 equilibrium mixture,
presumably interconverting through the aziridinium
intermediate 6.1a Subsequent investigations led to the
isolation of some minor compounds having the cylindri-
cine A pyrroloquinoline framework and stereochemistry,
but differing in the functionality at C14. Examples of
these alkaloids include cylindricines C (2), D (3), E (4),
and F (5). In addition, a few alkaloids were found which
† Author to be contacted about the X-ray crystal structure determi-
nation.
(1) (a) Blackman, A. J .; Li, C.; Hockless, D. C. R.; Skelton, B. W.;
White, A. H. Tetrahedron 1993, 49, 8645. (b) Li, C.; Blackman, A. J .
Aust. J . Chem. 1994, 47, 1355. (c) Li, C.; Blackman, A. J . Aust. J . Chem.
1995, 48, 955.
(2) Biard, J . F.; Guyot, S.; Roussakis, C.; Verbist, J . F.; Vercauteren,
J .; Weber, J . F.; Boukef, K. Tetrahedron Lett. 1994, 35, 2691.
10.1021/jo990266s CCC: $18.00 © 1999 American Chemical Society
Published on Web 05/29/1999

4866 J . Org. Chem., Vol. 64, No. 13, 1999
Werner et al.
Sch em e 1
Sch em e 2
such as 17 via N-alkylation to a cyclic nitrone 16.8 We
anticipated that a subsequent intramolecular cyclization
of nitrone-olefin 16 would produce spirocyclic isoxazol-
idine 15 having the requisite cylindricine/lepadiformine
stereochemistry at C5, 10, and 13 (cylindricine number-
ing1) (vide infra).9,10 Cleavage of the N-O bond in 15
should provide amino alcohol 14, which would then be
processed to the various alkaloids, perhaps via an enone
like 13, depending on the nature of R′.
calculations which indicate that conformer 11b is ∼2.6
kcal/mol more stable than 11a .3 More recently, a group
at SmithKline Beecham reported the structure of the
alkaloid fasicularin (12), isolated from the ascidian
Neptheis fasicularis.4 Compound 12 is closely related
structurally to the cylindricine B group of alkaloids but
is epimeric at C10 and lacks the C4 oxygenation. Fasicu-
larin has activity against a DNA repair deficient strain
of yeast and is also cytotoxic.
Our approach to a nitrone precursor such as 17 began
with acetone oxime, which was converted to its dianion
and alkylated with commercially available epoxide 18 to
afford hydroxy oxime 19 (Scheme 2).11 Oxime 19, which
is presumably initially Z,11,12 was found to be a mixture
of E/Z isomers after an aqueous workup, and upon
further purification by chromatography on silica gel
equilibrated completely to the E isomer. The E oxime 19
was then again metalated and alkylated with 5-bromo-
1-pentene to afford oxime 20. The oxime functionality in
compound 20 was first O-silylated, and the secondary
hydroxyl group was converted to the corresponding
chloride 21 via Mitsunobu methodology.13 Unfortunately,
all attempts to cyclize 21 to a nitrone led to the 1,2-
oxazine 22 as the only characterizable product. Although
the expectation was that chloro oxime 21 would exist as
an equilibrating E/Z mixture,14 these results show that
cyclization onto oxygen leading to the oxazine 22 is
apparently more favorable than closure onto nitrogen,
which would provide the desired nitrone. In a similar
vein, it has previously been observed by Tiecco and co-
workers14 that electrophile-promoted closure of alkenyl
oximes to 1,2-oxazines and/or five-membered nitrones is
not dependent upon starting oxime geometry. However,
in the cases they examined, the cyclic nitrone was usually
the preferred product.
In view of the unique structures of these marine
metabolites, along with their interesting biological activ-
ity, we have recently been involved in a program leading
to a stereoselective total synthesis of the pyrroloquinoline
subgroup of the alkaloids.5 While this work was in
progress, Snider and Liu described the first total syn-
theses of cylindricines A, D, and E.6 Moreover, Pearson
and co-workers have recently described some interesting
synthetic studies aimed at lepadiformine involving an
azaallyl anion cycloaddition strategy.7
Our original retrosynthetic approach to these alkaloids
is briefly outlined in Scheme 1. The intent was to utilize
the methodology of Grigg to convert an oxime substrate
(8) (a) Grigg, R.; Hadjisoteriou, M.; Kennewell, P.; Markandu, J .;
Thornton-Pett, M. J . Chem. Soc., Chem. Commun. 1992, 1388. (b)
Grigg, R.; Hadjisoteriou, M.; Kennewell, P.; Markandu, J . J . Chem.
Soc., Chem. Commun. 1992, 1537.
(9) For a brief review of intramolecular nitrone-olefin cycloaddi-
tions, see: Wade, P. A. Intramolecular 1,3-Dipolar Cycloadditions. In
Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Eds.;
Pergamon: Oxford, 1991; Vol. 4, p 1111.
(10) For some intramolecular nitrone-olefin spirocyclizations, see:
Tufariello, J . J .; Trybulski, E. J . J . Org. Chem. 1974, 39, 3378.
Gossinger, E.; Imhof, R.; Wehrli, H. Helv. Chim. Acta 1975, 58, 96.
(11) (a) J ung, M. E.; Blair, P. A.; Lowe, J . A. Tetrahedron Lett. 1976,
1439. (b) Kofron, W. G.; Yeh, M.-K. J . Org. Chem. 1976, 41, 439.
(12) For NMR determination of oxime geometry, see: Bunnell, C.
A.; Fuchs, P. L. J . Org. Chem. 1977, 42, 2614.
(3) Although at first glance it would seem that 11b, having an axial
hexyl group, should be unfavorable, there is a destabilizing 1,3-diaxial
interaction between C9 and C14 in 11a which leads to the former
conformation being preferred. We thank Professor R. L. Funk for
assistance in performing these calculations with PC Model.
(4) Patil, A. D.; Freyer, A. J .; Reichwein, R.; Carte, B.; Killmer, L.
B.; Faucette, L.; J ohnson, R. K.; Faulkner, D. J . Tetrahedron Lett. 1997,
38, 363.
(5) For a preliminary account of portions of this work, see: Werner,
K. M.; De los Santos, J . M.; Weinreb, S. M.; Shang, M. J . Org. Chem.
1999, 64, 686.
(6) Snider, B. B.; Liu, T. J . Org. Chem. 1997, 62, 5630.
(7) (a) Pearson, W. H.; Barta, N. S.; Kampf, J . W. Tetrahedron Lett.
1997, 38, 3369. (b) Pearson, W. H.; Ren, Y. J . Org. Chem. 1999, 64,
688.
(13) Rollin, P. Synth. Commun. 1986, 16, 611.
(14) Tiecco, M.; Testaferri, L.; Tingoli, M.; Bagnoli, L.; Marini, F. J .
Chem. Soc., Perkin Trans. 1 1993, 1989.

Cylindricine and Lepadiformine Marine Alkaloid Synthesis
J . Org. Chem., Vol. 64, No. 13, 1999 4867
Sch em e 3
Sch em e 5
Sch em e 4
In view of these disappointing results, we decided to
approach synthesis of the cyclic nitrone system 16 via a
more reliable strategy involving closure of a keto hy-
droxylamine. At the same time, we have investigated
some systems which differ in the R′ moiety shown in the
structures in Scheme 1 (vide infra). It appeared to us that
the most convergent and efficient route would utilize a
system containing all of the carbon atoms needed for the
alkaloids (cf. 13), and therefore a high priority was to
investigate the cycloaddition of a nitrone like 16 contain-
ing a 1,3-diene as the dipolarophile.15
The required diene subunit was first prepared by the
sequence shown in Scheme 3. Unsaturated aldehyde 23
was converted to dienol 24, which underwent a J ohnson
ortho ester Claisen rearrangement to afford a mixture
of regioisomeric diene esters.16 This mixture was reduced
with lithium aluminum hydride to afford a chromato-
graphically separable 1:1.3 mixture of diene alcohols 25
and 26. Although the isolated yield of the desired E,E-
dienol 26 was only 19% for the two steps from intermedi-
ate 24, the simplicity of the procedure allowed for
preparation of reasonable amounts of this material. This
alcohol could then be converted to iodide 27 in high yield.
A more efficient alternative procedure was subsequently
developed for preparation of the necessary dienol (Scheme
4). Thus, Pd-catalyzed coupling of the known hydroxy
vinylboronic acid 2917 with the E-vinyl iodide 2818 using
the Suzuki-Miyaura methodology19,20 afforded 26 in good
yield.
To continue the synthesis, oxime alcohol 19 was
converted to the trianion and alkylated with iodo diene
27 to yield 30 (Scheme 5). Cleavage of the oxime
functionality of 30 to the corresponding ketone 31 with
21
TiCl3 and silylation of the alcohol afforded 32. The
carbonyl group in 32 was ketalized, giving 33, which was
desilylated to alcohol 34. Swern oxidation of 34 to ketone
35, followed by oxime formation provided 36. Finally,
reduction of oxime 36 with sodium cyanoborohydride22
afforded hydroxylamine 37. The yields in this sequence
of reactions were uniformly high.
We were pleased to find that treatment of hydroxyl-
amino ketal 37 with aqueous HCl led to the desired cyclic
nitrone diene 40 which could be isolated in 92% yield
(Scheme 6). Initial cyclization experiments involved
heating 40 in o-dichlorobenzene at 170 °C, which led to
formation of the desired isoxazolidine cycloadduct 43, but
in low yield. However, it was eventually found that
thermolysis of 40 in DMSO overnight at 195 °C produced
adduct 43 in 63% yield. We were also pleased to find that
43 was formed as a single stereoisomer, whose structure
was eventually confirmed by X-ray analysis of a subse-
quent intermediate (vide infra).
(15) An example of an intramolecular nitrone-conjugated diene
cycloaddition has been described: Holmes, A. B.; Hughes, A. B.; Smith,
A. L.; Williams, S. F. J . Chem. Soc., Perkin Trans. 1 1992, 1089.
Intermolecular 1,3-dipolar cycloaddition reactions of nitrones with
conjugated dienes leading to indolizidines and quinolizidines have been
reported: (a) Tufariello, J . J . Acc. Chem. Res. 1979, 12, 396. (b)
Tufariello, J . J .; Dyszlewski, A. D. J . Chem. Soc., Chem. Commun.
1987, 1138 and references cited. Conjugated dipolarophiles also tend
to show higher reactivity than nonconjugated ones: Sustmann, R.
Tetrahedron Lett. 1971, 2717.
(16) Hudlicky, T.; Koszyk, F. J .; Kutchan, T. M.; Sheth, J . P. J . Org.
Chem. 1980, 45, 5020.
(17) Roush, W. R.; Moriarty, K. J .; Brown, B. B. Tetrahedron Lett.
1990, 31, 6509.
We have also briefly investigated cyclizations of some
other related alkenyl nitrones which seemed to have
some potential as useful intermediates for synthesis of
the cylindricines. Thus, using the same methodology
outlined in Scheme 5, we prepared hydroxylamine olefins
38 and 39 (for experimental details, see Supporting
Information). Compounds 38 and 39 could be converted
(18) Tius, M. A.; Busch-Petersen, J .; Yamashita, M. Tetrahedron
Lett. 1998, 39, 4219.
(19) Miyaura, N.; Suginome, H.; Suzuki, A. Tetrahedron 1983, 39,
3271.
(20) We are grateful to Ann Bullion for synthesis of diene alcohol
26 via the Suzuki-Miyaura coupling route.
(21) Timms, G. H.; Wildsmith, E. Tetrahedron Lett. 1971, 195.
(22) Borch, R. F.; Bernstein, M. D.; Durst, H. D. J . Am. Chem. Soc.
1971, 93, 2897.

4868 J . Org. Chem., Vol. 64, No. 13, 1999
Werner et al.
Sch em e 6
Sch em e 7
cleanly to cyclic nitrones 41 and 42, respectively, by mild
acid hydrolysis. Interestingly, cyclization of the terminal
alkene nitrone 41 occurred in xylene at 150 °C to give a
1:2 mixture of spirocyclic isoxazolidine 44 and the
unwanted bridged isomer 46 in 69% total yield.23a In the
case of the E-olefin nitrone 42, cyclization was extremely
sluggish. At best, we were only able to isolate 15% (38%
based on recovered nitrone) of a single cycloadduct which
has tentatively been assigned spirocyclic structure 45.
It would appear that these cyclizations can occur via
the two conformations shown in 40-42 and 47 (the
bridging chain must be boatlike for stereoelectronic
reasons). In both situations, the olefin resides exclusively
on the face of the nitrone opposite to the bulky phen-
oxymethyl group. When R is large, conformer 47 is
probably disfavored relative to 40-42 for steric reasons,
therefore leading to the spirocyclic series of adducts. The
alternative regioisomeric isoxazolidine system 48 was not
detected in any case, presumably for stereoelectronic
reasons.23b
With the desired cycloadduct 43 now in hand, we
turned to developing methodology for converting it to one
or more of the marine alkaloids of interest. Cleavage of
the N-O bond of this isoxazolidine could be effected in
high yield with Zn/HOAc to give the amino alcohol 49
(Scheme 7). Oxidation of 49 with the Dess-Martin
reagent in the presence of tert-butyl alcohol24 led directly
to tricycle 52, presumably via enone 50. It might be noted
that in the absence of tert-butyl alcohol, the Dess-Martin
oxidation went poorly.24 Compound 52 is a single stereo-
isomer whose configuration and conformation have been
assigned as shown on the basis of 2D NMR experiments
(NOESY, HMQC, and 13C Inadequate), along with the
results of subsequent transformations (vide infra). The
formation of the C2 axial epimer shown is not surprising,
since for stereoelectronic reasons conjugate addition of
the amino group to the enone in 50 must occur through
a transition state initially leading to a boat ketone 51,
which ring flips to the observed product 52. As a further
confirmation of structure, ketone 52 was subsequently
Sch em e 8
protected as its ethylene ketal (35% unoptimized), and
the phenyl protecting group was removed by Birch
reduction, followed by acid hydrolysis (26% unoptimized),
to afford 2-epi-cylindricine C (53).25 The proton NMR data
for this material were similar to, but clearly different
from, those of authentic cylindricine C (2).26
We have also made some attempts to epimerize tricycle
52 at C2 to produce the cylindricine C stereochemistry.
It was our hope that 52 could be converted via a retro-
Mannich process into iminium ion 54, which might then
equilibrate to isomer 55 (Scheme 8). Mannich reclosure
of 55 would afford O-phenylcylindricine C (56). However,
all attempts to promote this interconversion using vari-
ous acids failed, and in most cases only starting material
was recovered.
Another approach to adjusting the C2 stereochemistry
is outlined in Scheme 9. It was found that activated MnO2
oxidation of amino ketone 52 afforded the vinylogous
amide 57 (33% unoptimized yield). Unfortunately, reduc-
tion of 57 with sodium cyanoborohydride/HCl gave back
ketone 52. Similarly, reduction of 57 using Li/NH3 rather
surprisingly also gave only the original ketone 52.
At this point we turned to removing the carbonyl group
of intermediate ketone 52 to access the proposed lepadi-
(25) Aryl protection for alcohols has not been widely used. Cf.
Marshall, J . A.; Partridge, J . J . J . Am. Chem. Soc. 1968, 90, 1090. In
this case a p-chlorophenyl ether protecting group was used for oxygen.
(26) We are grateful to Professors Barry Snider and A. J . Blackman
for copies of the NMR spectra of several of the cylindricines including
cylindricine C.
(23) (a) For formation of both fused and bridged isoxazolidines in a
related system, see: LeBel, N. A.; Lajiness, T. A. Tetrahedron Lett.
1966, 2173. (b) Cf. LeBel, N. A.; Post, M. E.; Whang, J . J . J . Am. Chem.
Soc. 1964, 86, 3759.
(24) Dess, D. B.; Martin, J . C. J . Org. Chem. 1983, 48, 4155.

Cylindricine and Lepadiformine Marine Alkaloid Synthesis
J . Org. Chem., Vol. 64, No. 13, 1999 4869
Sch em e 9
Sch em e 10
F igu r e 1. ORTEP plot of the picrate of tricycle 59 (anion
omitted for clarity).
appears, therefore, that the structure of lepadiformine
is not as originally formulated and requires revision. It
might also be added that Pearson and Ren7b,32 have
recently synthesized the three other possible diastereo-
mers of 60 at C2 and C13 and have found that none of
these stereoisomers correspond to lepadiformine. Another
possibility which has not yet been tested is that lepadi-
formine is epimeric at C10 and thus is in the fasicularin
(12) series.
In summary, we have described a stereoselective and
convergent route to the pyrroloquinoline framework of
the cylindricines/lepadiformine via an intramolecular
nitrone dipolar cycloaddition involving a conjugated diene
as the dipolarophile. In these studies, we have synthe-
sized the putative structure of lepadiformine and have
shown that the constitution of this alkaloid requires
revision. Future work will require the development of
alternative methodology to set the proper C2 stereochem-
istry found in the cylindricine series of marine alkaloids.
formine structure 11. All attempts to deoxygenate 52 by
a Wolff-Kishner reduction,27 or by reduction of the
tosylhydrazone,28 failed. In addition, we were unable to
convert ketone 52 to the corresponding thioketal. Alter-
natively, Clemmensen reduction29 of 52 did provide some
of the deoxygenated tricyclic amine 59 (6%), but the
primary product of this reaction was the trisubstituted
olefin 58 (51%) (Scheme 10). Although alkenes have
previously been observed as minor byproducts in Clem-
mensen reductions of ketones, the fact that compound
58 was the major product is quite surprising.29,30 It was
possible, however, to stereoselectively hydrogenate olefin
58 to produce the desired intermediate 59. To firmly
establish the structure and stereochemistry of 59, a
single-crystal X-ray analysis was performed on its picrate
salt. An ORTEP plot of this structure is depicted in
Figure 1. Interestingly, tricycle 59 has the conformation
as is shown in 11b, which was predicted by molecular
modeling for lepadiformine (vide supra).3
Finally, the phenyl protecting group in 59 was removed
by a dissolving metal reduction, followed by acid hydroly-
sis, providing tricyclic amino alcohol 60. Although 60 has
the structure proposed for lepadiformine (cf. 11), direct
comparison of the proton and carbon NMR spectra of this
material with those of the natural alkaloid kindly sup-
plied by Professor Biard31 clearly showed that the two
compounds were in fact different. Moreover, there is no
indication that our synthetic tricycle 60 exists in the
unusual zwitterionic form proposed for lepadiformine. It
Exp er im en ta l Section
5-Hyd r oxy-6-p h en oxyh exa n -2-on e Oxim e (19). To a 0
°C solution of acetone oxime (2.74 g, 37.5 mmol) in THF (125
mL) was added nBuLi (30 mL, 2.5 M solution in hexanes). The
mixture was stirred at this temperature for 30 min, and a
solution of 1,2-epoxy-3-phenoxypropane (3.38 mL, 25.0 mmol)
in THF (25 mL) was then added. After warming to room
temperature, the solution was stirred for 2 h, diluted with
water, and extracted with CH2Cl2. The combined organic layers
were washed with water, dried over MgSO4, and concentrated.
The crude residue was crystallized from Et2O/hexanes (1:4)
to yield 4.79 g (86%) of oxime 19 as a white solid. An analytical
sample was obtained by recrystallization from CH2Cl2/hex-
1
anes: mp 108-110 °C; H NMR (300 MHz, CDCl3) δ 7.56 (br
s, 1H), 7.29 (t, J ) 7.4 Hz, 2H), 6.97 (t, J ) 7.3 Hz, 1H), 6.93
(d, J ) 7.8 Hz, 2H), 4.05-3.90 (m, 3H), 3.30 (br s, 1H), 2.44 (t,
J ) 7.2 Hz, 2H), 1.93 (s, 3H), 1.86-1.83 (m, 2H); 13C NMR (75
MHz, CDCl3) δ 158.9, 158.6, 129.5, 121.0, 114.5, 71.8, 69.2,
32.1, 29.5, 13.8; IR (CH2Cl2) 3244 (br), 2927, 1601, 1497, 1250
cm-1; EIMS m/z (relative intensity) 223 (M+, 4); HRMS calcd
for C12H17NO3 223.1208, found 223.1201.
(4E)-Un d eca -1,4-d ien -3-ol (24). To a solution of vinyl-
magnesium bromide (428 mL, 1 M in THF) in THF (500 mL)
at 0 °C was added (E)-2-nonenal (59.0 mL, 357 mmol). The
reaction was quenched after 15 min by the addition of a
saturated aqueous NH4Cl solution. The mixture was diluted
with water and extracted with EtOAc. The combined organic
extracts were washed with brine, dried over MgSO4, and
concentrated. The product was distilled (85 °C, 12 mmHg) to
(27) Yasuda, S.; Hanaoka, M.; Arata, Y. Chem. Pharm. Bull. 1980,
28, 831.
(28) Natsume, M.; Ogawa, M. Heterocycles 1981, 15, 237.
(29) Martin, E. L. Org. React. 1942, 1, 155.
(30) There seems to be little reason to further optimize the deoxy-
genation of ketone 52 since our synthetic compound 60 does not
correspond to natural lepadiformine.
(31) We thank Professor J . F. Biard for the proton and carbon NMR
spectra, as well as a sample of lepadiformine.
(32) We appreciate frequent helpful discussions and exchanges of
information with Professor William H. Pearson.

4870 J . Org. Chem., Vol. 64, No. 13, 1999
Werner et al.
afford 51.6 g (86%) of dienol 24 as a clear colorless oil: 1H
NMR (300 MHz, CDCl3) δ 5.91 (ddd, J ) 17.2, 10.4, 5.7 Hz,
1H), 5.71 (ddt, J ) 15.2, 6.6, 0.7 Hz, 1H), 5.49 (ddt, J ) 15.4,
6.6, 1.3 Hz, 1H), 5.25 (dt, J ) 17.3, 1.4 Hz, 1H), 5.12 (dt, J )
10.4, 1.3 Hz, 1H), 4.58 (t, J ) 5.9 Hz, 1H), 2.04 (q, J ) 6.8 Hz,
2H), 1.59 (t, J ) 5.9 Hz, 1H), 1.43-1.26 (m, 8H), 0.88 (t, J )
6.7 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ 139.9, 132.8, 130.9,
114.5, 73.7, 32.1, 31.6, 29.0, 28.8, 22.5, 14.0; IR (neat) 3342,
1H), 6.94 (d, J ) 8.3 Hz, 2H), 6.03 (m, 2H), 5.60 (m, 2H), 4.50
(br s, 1H), 4.03 (m, 1H), 3.94 (d, J ) 5.5 Hz, 2H), 2.45 (t, J )
6.9 Hz, 2H), 2.40 (m, 2H), 2.10 (quin, J ) 7.7 Hz, 4H), 1.88
(m, 2H), 1.59-1.30 (m, 12H), 0.92 (t, J ) 6.6 Hz, 3H); 13C NMR
(75 MHz, CDCl3) δ 161.5, 158.5, 132.6, 131.5, 130.7, 130.1,
129.4, 120.9, 114.5, 71.6, 69.2, 32.5, 32.2, 31.7, 30.4, 29.6, 29.5,
29.3, 28.8, 27.8, 25.1, 22.5, 14.0; IR (CHCl3) 3583, 3260 (br),
1599, 1588, 1241, 990 cm-1; EIMS m/z (relative intensity) 401
(M+, 3), 384 (M+ - OH, 13), 294 (12), 276 (11), 236 (11), 94
(PhOH, 100); HRMS calcd for C25H39NO3 401.2930, found
401.2936.
985, 961, 920 cm-1
.
(4E,6E)-Tr id eca -4,6-d ien -1-ol (26). A solution of alcohol
24 (5.00 g, 29.8 mmol), triethyl orthoacetate (109 mL, 595
mmol), propionic acid (400 µL, 5.36 mmol), and Hg(OAc)2 (191
mg, 0.600 mmol) was heated at reflux. After 16 h, the reaction
mixture was diluted with ether (200 mL) and washed with
aqueous 3 N HCl. The organic layer was dried over MgSO4
and concentrated to a yellow liquid. The crude residue was
purified by flash chromatography (2% Et2O/hexanes) to afford
2.8 g of a mixture of isomeric dienols which was used without
further purification. The mixture was dissolved in THF (5 mL)
and added to a suspension of 95% LiAlH4 (470 mg, 11.8 mmol)
in THF (40 mL) at 0 °C. After 15 min, the mixture was warmed
to room temperature and quenched by the addition of Na2SO4‚
10H2O. After stirring for 30 min, the mixture was filtered, and
the filtrate was concentrated. The crude product was purified
by flash chromatography (10-20% Et2O/hexanes gradient) to
afford 836 mg (19% over 2 steps) of alcohol 26 as a clear
colorless oil: 1H NMR (300 MHz, CDCl3) δ 6.02 (m, 2H), 5.58
(m, 2H), 3.66 (t, J ) 6.5 Hz, 2H), 2.15 (q, J ) 7.2 Hz, 2H),
2.05 (q, J ) 7.0 Hz, 2H), 1.66 (quin, J ) 6.8 Hz, 2H), 1.38-
1.26 (m, 9H), 0.88 (t, J ) 6.7 Hz, 3H); 13C NMR (75 MHz,
CDCl3) δ 132.9, 131.0, 130.9, 130.0, 62.2, 32.5, 32.2, 31.7, 29.3,
28.8 (2C), 22.5, 14.0; IR (neat) 3342 (br), 985 cm.-1
Hyd r oxy Keton e 31. To a solution of oxime 30 (12.1 g, 7.46
mmol) in p-dioxane (150 mL) were added ammonium acetate
(30.3 g, 392 mmol) and 50% aqueous acetic acid (15 mL). The
mixture was stirred, and 20% aqueous titanium trichloride
(56.0 g, 72.5 mmol) was added slowly at room temperature.
The mixture was stirred for 20 min, diluted with Et2O, and
washed with aqueous 1 N HCl, a saturated aqueous NaHCO3
solution, and brine. The organic extract was dried over MgSO4
and concentrated. The residue was purified by flash chroma-
tography (20% EtOAc/hexanes) to afford 10.0 g (86%) of
1
hydroxy ketone 31 as a white solid: mp 57-60 °C; H NMR
(300 MHz, CDCl3) δ 7.30 (t, J ) 8.7 Hz, 2H), 6.97 (t, J ) 7.3
Hz, 1H), 6.90 (d, J ) 7.8 Hz, 2H), 5.99 (m, 2H), 5.55 (m, 2H),
4.00-3.93 (m, 3H), 2.66 (t, J ) 7.0 Hz, 2H), 2.60 (d, J ) 3.8
Hz, 1H), 2.45 (t, J ) 7.4 Hz, 2H), 2.05 (quin, J ) 7.1 Hz, 4H),
1.86 (m, 2H), 1.56 (quin, J ) 7.7 Hz, 2H), 1.33 (m, 10H), 0.88
(t, J ) 6.7 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ 211.3, 158.4,
132.8, 131.4, 130.8, 130.1, 129.5, 121.1, 114.5, 71.8, 69.5, 42.7,
38.6, 32.6, 32.3, 31.7, 29.3, 28.9 (2C), 26.8, 23.4, 22.6; IR
(CH2Cl2) 3588, 3452 (br), 1706, 1599, 1587, 1229, 990 cm-1
;
ESIMS m/z (relative intensity) 409 ([M + Na]+, 46), 369 ([M
+ H - H2O]+, 100); HRMS calcd for C25H38O3Na 409.2719,
found 409.2734. Anal. Calcd for C25H38O3: C, 77.68; H, 9.91.
Found: C, 77.78; H, 9.82.
(4E,6E)-Tr id eca -4,6-d ien -1-ol (26). To boronic acid17 29
(910 mg, 7.00 mmol) were added benzene (20 mL), (E)-1-
iodooctene18 (1.83 g, 7.70 mmol), tetrakis(triphenylphosphine)-
palladium (404 mg, 0.350 mmol), and NaOEt (7.0 mL of a 2
M solution in ethanol), and the mixture was heated at reflux.
After 2 h, the mixture was cooled to room temperature, and
15% aqueous NaOH (2 mL) was added followed by the
dropwise addition of a 30% aqueous H2O2 solution (2 mL). The
reaction mixture was diluted with brine and extracted with
Et2O. The organic extracts were combined, dried over MgSO4,
and concentrated. The crude residue was purified by flash
chromatography (10-15% EtOAc/hexanes gradient) to afford
593 mg (64%) of dienol 26 as a clear colorless oil.
Siloxy Keton e 32. To a solution of hydroxy ketone 31 (10.0
g, 25.9 mmol) and imidazole (3.53 g, 51.8 mmol) in CH2Cl2 (130
mL) was added TBDPSCl (8.10 mL, 31.1 mmol). The mixture
was stirred for 16 h, diluted with CH2Cl2, and washed with
brine. The organic layer was dried over MgSO4 and concen-
trated. The residue was purified by flash chromatography (3%
EtOAc/hexanes) to afford 14.8 g (92%) of siloxy ketone 32 as
a clear colorless oil: 1H NMR (300 MHz, CDCl3) δ 7.69 (m,
4H), 7.44-7.30 (m, 6H), 7.19 (m, 2H), 6.89 (m, 1H), 6.65 (m,
2H), 5.99 (m, 2H), 5.53 (m, 2H), 4.09 (quin, J ) 5.5 Hz, 1H),
3.81 (AB of ABX, J AB ) 9.6 Hz, J AX ) 5.5 Hz, J BX ) 5.8 Hz,
∆υAB ) 23.7 Hz, 2H), 2.46 (m, 2H), 2.29 (t, J ) 7.3 Hz, 2H),
2.02 (m, 4H), 1.88 (m, 2H), 1.50 (m, 2H), 1.28 (m, 10H), 1.07
(s, 9H), 0.88 (t, J ) 6.7 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ
210.5, 158.4, 135.8, 133.8, 133.7, 132.7, 131.4, 130.8, 130.1,
129.7, 129.6, 129.2, 127.6, 127.5, 120.6, 114.2, 70.6 (2C), 42.5,
37.8, 32.6, 32.3, 31.7, 29.3, 28.9 (2C), 28.1, 27.0, 23.3, 22.6,
19.4, 14.1; IR (neat) 1714, 1598, 1587, 1243, 988 cm-1. Anal.
Calcd for C41H56O3Si: C, 78.79; H, 9.03. Found: C, 78.67; H,
9.08.
(4E,6E)-1-Iod otr id eca -4,6-d ien e (27). To a solution of
triphenylphosphine (918 mg, 3.50 mmol) in CH2Cl2 (9 mL) at
0 °C were added imidazole (5.5 mg, 7.0 mmol) and iodine (890
mg, 3.50 mmol). After 30 min, a solution of alcohol 26 (343
mg, 1.75 mmol) in CH2Cl2 (2 mL) was added. The mixture was
warmed to room temperature and stirred for 40 min. Brine
was added, and the mixture was extracted with CH2Cl2. The
combined organic layers were dried over MgSO4 and concen-
trated. The crude residue was purified by flash chromatogra-
phy (hexanes) to afford 510 mg (95%) of iodide 27 as a pale
yellow oil: 1H NMR (300 MHz, CDCl3) δ 6.02 (m, 2H), 5.60
(ddd, J ) 14.1, 6.9, 6.9 Hz, 1H), 5.48 (ddd, J ) 13.9, 6.9, 6.9
Hz, 1H), 3.18 (t, J ) 6.9 Hz, 2H), 2.17 (q, J ) 7.1 Hz, 2H),
2.05 (q, J ) 7.0 Hz, 2H), 1.90 (quin, J ) 7.0 Hz, 2H), 1.28 (m,
8H), 0.88 (t, J ) 6.5 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ
133.4, 131.9, 129.8, 129.2, 33.1, 32.9, 32.6, 31.7, 29.2, 28.9, 22.6,
F or m a tion of Keta l 33. To a solution of siloxy ketone 32
(14.8 g, 23.7 mmol) and pTsOH (1.35 g, 7.11 mmol) in benzene
(120 mL) was added ethylene glycol (13 mL, 237 mmol). The
flask was fitted with a Dean-Stark trap, and the mixture was
heated at reflux for 16 h. The solution was cooled to room
temperature, and an aqueous saturated NaHCO3 solution was
added. The mixture was extracted with Et2O, dried over
MgSO4, and concentrated. The residue was purified by flash
chromatography (3% EtOAc/hexanes) to afford 15.4 g (97%)
of ketal 33 as a clear colorless oil: 1H NMR (300 MHz, CDCl3)
δ 7.73 (m, 4H), 7.47-7.34 (m, 6H), 7.20 (m, 2H), 6.90 (t, J )
7.3 Hz, 1H), 6.67 (d, J ) 7.8 Hz, 2H), 6.01 (m, 2H), 5.57 (m,
2H), 4.09 (quin, J ) 5.2 Hz, 1H), 3.88 (m, 6H), 2.06 (m, 4H),
14.1, 6.4; IR (neat) 985 cm-1
.
P r ep a r a tion of Oxim e 30. To a solution of oxime 19 (4.07
g, 18.2 mmol) in THF (60 mL) at 0 °C was added nBuLi (23.0
mL, 2.5 M solution in hexanes). The mixture was warmed to
room temperature and stirred for 45 min. After the solution
was cooled to 0 °C, a solution of iodide 27 (6.13 g, 20.0 mmol)
in THF (5 mL) was added. The mixture was warmed to room
temperature and stirred for 16 h. The mixture was diluted with
Et2O and washed with brine. The organic layer was dried over
MgSO4 and concentrated. The crude product was purified by
flash chromatography (25-40% EtOAc/hexanes gradient) to
yield 5.55 g (76%) of oxime isomers 30 as a white solid. Data
1.74-1.27 (m, 18H), 1.09 (s, 9H), 0.90 (t, J ) 6.7 Hz, 3H); 13
C
NMR (75 MHz, CDCl3) δ 158.6, 136.0, 135.9, 134.1, 133.9,
132.5, 132.0, 130.5, 130.2, 129.6, 129.5, 129.2, 127.6, 127.4,
120.4, 114.3, 111.6, 71.3, 70.8, 64.8, 36.9, 32.6 (2C), 31.8, 31.7,
29.7, 29.4, 28.9, 28.5, 27.0, 23.4, 22.6, 19.4, 14.1; IR (neat) 1602,
1245, 987 cm-1. Anal. Calcd for C43H60O4Si: C, 77.20; H, 9.04.
Found: C, 77.10; H, 9.15.
1
for major isomer: mp 94-95 °C; H NMR (300 MHz, CDCl3)
δ 9.52 (br s, 1H), 7.30 (t, J ) 7.9 Hz, 2H), 6.98 (t, J ) 7.3 Hz,

Cylindricine and Lepadiformine Marine Alkaloid Synthesis
J . Org. Chem., Vol. 64, No. 13, 1999 4871
Alcoh ol 34. To a solution of ketal 33 (15.4 g, 23.1 mmol) in
THF (65 mL) was added tetrabutylammonium fluoride (35 mL,
1 M solution in THF). The solution was stirred for 16 h, at
which time the mixture was diluted with Et2O and washed
with brine. The organic layer was dried over MgSO4 and
concentrated. The residue was purified by flash chromatog-
raphy (15-20% EtOAc/hexanes gradient) to afford 9.90 g (99%)
of alcohol 34 as a clear colorless oil: 1H NMR (300 MHz,
CDCl3) δ 7.29 (t, J ) 8.2 Hz, 2H), 6.96 (t, J ) 7.3 Hz, 1H),
6.91 (d, J ) 7.8 Hz, 2H), 5.99 (m, 2H), 5.55 (m, 2H), 4.02-
3.83 (m, 7H), 2.71 (br d, J ) 3.0 Hz, 1H), 2.05 (quin, J ) 6.9
Hz, 4H), 1.94-1.60 (m, 6H), 1.40-1.21 (m, 12H), 0.88 (t, J )
6.7 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ 158.5, 132.3, 131.7,
130.4, 130.1, 129.3, 120.8, 114.4, 111.4, 71.9, 69.9, 64.7, 36.8,
32.7, 32.4, 32.3, 31.6, 29.5, 29.2, 28.7, 27.3, 23.3, 22.4, 13.9;
IR (neat) 3448 (br), 1595, 1588, 1243, 979 cm-1. Anal. Calcd
for C27H42O4: C, 75.31; H, 9.83. Found: C, 75.13; H, 9.79.
Keton e 35. To a solution of oxalyl chloride (13.8 mL, 2 M
solution in CH2Cl2) in CH2Cl2 (80 mL) at -60 °C was added
DMSO (3.92 mL, 55.2 mmol) dropwise. The mixture was
stirred for 15 min, and a solution of alcohol 34 (9.90 g, 23.0
mmol) in CH2Cl2 (20 mL) was then added. After the mixture
was stirred for 15 min, TEA (16.0 mL, 115 mmol) was added
and the mixture was warmed to room temperature. The
mixture was diluted with CH2Cl2, washed with water and
brine, dried over MgSO4, and concentrated. The residue was
purified by flash chromatography (7% EtOAc/hexanes) to
afford 9.15 g (93%) of ketone 35 as a clear colorless oil: 1H
NMR (300 MHz, CDCl3) δ 7.30 (t, J ) 8.6 Hz, 2H), 6.99 (t, J
) 7.3 Hz, 1H), 6.89 (d, J ) 8.6 Hz, 2H), 5.99 (m, 2H), 5.55 (m,
2H), 4.57 (s, 2H), 3.90 (s, 4H), 2.62 (t, J ) 7.3 Hz, 2H), 2.02
(m, 6H), 1.59 (m, 2H), 1.28 (m, 12H), 0.88 (t, J ) 6.6 Hz, 3H);
13C NMR (75 MHz, CDCl3) δ 207.3, 157.8, 132.6, 131.8, 130.6,
130.2, 129.6, 121.6, 114.5, 110.9, 87.1, 72.6, 64.9, 37.1, 33.6,
32.6, 32.4, 31.7, 30.4, 29.6, 29.3, 28.8, 23.4, 22.6, 14.0; IR (neat)
1721, 1600, 1244, 988 cm-1; ESIMS m/z (relative intensity)
429 ([M + H]+, 36), 367 (100), 349 (11); HRMS calcd for
EIMS m/z (relative intensity) 443 (M+ - H2, 3), 265 (34), 202
(27), 94 (PhOH, 100); HRMS calcd for C27H43NO4 445.3192,
found 445.3175. Anal. Calcd for C27H43NO4: C, 72.77; H, 9.73;
N, 3.14. Found: C, 72.69; H, 9.85; N, 3.17.
F or m a tion of Nitr on e Dien e 40. A mixture of hydroxyl-
amine 37 (8.45 g, 19.0 mmol) and aqueous 3 N HCl (60 mL)
in THF (120 mL) was stirred at room temperature for 3.5 h.
The solution was diluted with EtOAc and then washed with
water, 5% aqueous NaOH solution, and brine. The organic
layer was dried over MgSO4 and concentrated. The residue
was purified by flash chromatography (10% MeOH/EtOAc) to
afford 6.73 g (92%) of nitrone diene 40 as a clear colorless oil:
1H NMR (300 MHz, CDCl3) δ 7.26 (t, J ) 8.0 Hz, 2H), 6.95 (t,
J ) 7.4 Hz, 1H), 6.90 (d, J ) 7.9 Hz, 2H), 5.98 (m, 2H), 5.53
(m, 2H), 4.67 (dd, J ) 10.1, 3.5 Hz, 1H), 4.31 (m, 1H), 4.14
(dd, J ) 10.1, 2.8 Hz, 1H), 2.82 (dt, J ) 15.9, 8.5 Hz, 1H), 2.61
(m, 2H), 2.44 (dt, J ) 15.2, 7.1 Hz, 1H), 2.29 (q, J ) 7.5 Hz,
2H), 2.06 (quin, J ) 7.5 Hz, 4H), 1.58-1.26 (m, 12H), 0.88 (t,
J ) 6.6 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ 158.3, 148.7,
132.6, 131.2, 130.7, 130.0, 129.3, 121.0, 114.5, 72.0, 66.3, 32.4,
32.0, 31.6, 30.1, 29.2, 29.0, 28.7, 26.4, 24.4, 22.4, 19.6, 13.9;
IR (neat) 1602, 1590, 1237, 985 cm-1; EIMS m/z (relative
intensity) 383 (M+, 6), 367 (8), 276 (58), 202 (33), 94 (PhOH,
100); HRMS calcd for C25H37NO2 383.2824, found 383.2833.
5-(H ex-5′-en yl)-2-p h en oxym et h yl-2,3-d ih yd r o4H -p yr -
r ole 1-Oxid e (41). A mixture of ketal 38 (0.68 g, 2.03 mmol)
and 3 N aqueous HCl (2 mL) in THF (20 mL) was stirred at
room temperature overnight. The mixture was diluted with
water, extracted with EtOAc, dried over MgSO4, and concen-
trated. The residue was purified by flash chromatography
eluting with 4:1 EtOAc/MeOH to yield 0.48 g (87%) of nitrone
41 as a pale yellow oil: 1H NMR (300 MHz, CDCl3) δ 1.35-
1.55 (m, 4H), 1.99-2.06 (m, 2H), 2.18-2.26 (m, 2H), 2.36-
2.79 (m, 4H), 4.07 (dd, J ) 2.8, 10.1 Hz, 1H), 4.24 (br s, 1H),
4.62 (dd, J ) 3.4, 10.1 Hz, 1H), 4.87-4.98 (m, 2H), 5.65-5.78
(m, 1H), 6.84-7.27 (m, 5H); 13C NMR (75 MHz, CDCl3) δ 19.5,
24.2, 26.3, 28.4, 30.0, 33.1, 66.2, 71.9, 114.4, 114.5, 120.9, 129.2,
138.1, 148.8, 158.2; IR (neat) 3385 (br), 2927 (br), 1600, 1497,
1245 cm-1; EIMS m/z (relative intensity) 273 (M+, 4); CIMS
C
27H41O4 429.3005, found 429.3023.
Oxim e 36. To a solution of ketone 35 (9.15 g, 21.4 mmol)
and pyridine (2.60 mL, 32.1 mmol) in ethanol (80 mL) was
added hydroxylamine hydrochloride (1.78 g, 25.6 mmol). After
stirring for 30 min, the mixture was diluted with CH2Cl2,
washed with brine, dried over MgSO4, and concentrated. The
residue was purified by flash chromatography (15% EtOAc/
hexanes) to afford 9.60 g (100%) of oxime 36 as a pale yellow
oil, which was a 1:1 mixture of geometric isomers: 1H NMR
(300 MHz, CDCl3) δ 7.80 (s, 1H), 7.69 (s, 1H), 7.30 (m, 4H),
6.95 (m, 6H), 5.98 (m, 4H), 5.56 (m, 4H), 4.94 (s, 2H), 4.59 (s,
2H), 3.93 (s, 4H), 3.91 (s, 4H), 2.55 (m, 2H), 2.44 (m, 2H), 2.04
(q, J ) 6.8 Hz, 8H), 1.90 (m, 4H), 1.62 (m, 4H), 1.37-1.21 (m,
24H), 0.88 (t, J ) 6.8 Hz, 6H); 13C NMR (75 MHz, CDCl3) δ
159.2, 158.3, 158.2, 157.9, 132.5, 131.9, 130.5, 130.3, 129.6,
129.5, 121.2, 114.7, 114.3, 111.2, 68.2, 64.9 (2C), 62.4, 36.9 (2C),
33.2, 32.6, 32.5 (2C), 31.8, 31.7, 29.7, 29.4, 28.9, 25.4, 23.4,
m/z (relative intensity) 274 (MH+, 100); HRMS calcd for C17H23
-
NO2 273.1729, found 273.1726.
Cycloa d d ition of Nitr on e 41. A solution of nitrone 41
(0.23 g, 0.84 mmol) in m-xylene (10 mL) was heated at 150 °C
for 48 h. The solvent was removed, and the residue was
purified by flash chromatography eluting first with 1:20
EtOAc/hexanes to afford 108 mg (47%) of the less polar
compound 46 as a pale yellow oil: 1H NMR (300 MHz, CDCl3)
δ 1.42-1.65 (m, 4H), 1.69-1.83 (m, 6H), 2.03-2.10 (m, 2H),
2.22-2.26 (m, 2H), 3.37-3.46 (m, 1H), 3.84 (t, J ) 8.9 Hz, 1H),
4.30 (dd, J ) 4.0, 8.9 Hz, 1H), 4.72 (br d, J ) 8.9 Hz, 1H),
6.90-7.30 (m, 5H); 13C NMR (75 MHz, CDCl3) δ 23.3, 23.5,
25.6, 33.2, 33.5, 40.0, 41.2, 66.5, 71.0, 72.8, 79.7, 114.3, 120.4,
129.2, 158.8; IR (neat) 2918, 1599, 1496, 1245 cm-1; EIMS m/z
(relative intensity) 273 (M+, 8); CIMS m/z (relative intensity)
274 (MH+, 99); HRMS calcd for C17H23NO2 273.1729, found
273.1728. Further elution with 1:10 EtOAc/hexanes furnished
the more polar product 44, 51 mg (22%) as a pale yellow oil:
1H NMR (300 MHz, CDCl3) δ 1.23-2.31 (m, 12H), 2.43-2.51
(m, 1H), 3.46-3.56 (m, 1H), 3.80-3.88 (m, 2H), 4.20 (t, J )
7.9 Hz, 1H), 4.26 (dd, J ) 4.3, 9.0 Hz, 1H), 6.90-7.31 (m, 5H);
13C NMR (75 MHz, CDCl3) δ 21.1, 22.9, 23.1, 26.5, 32.8, 34.1,
45.0, 65.3, 71.0, 71.5, 71.6, 114.5, 120.6, 129.3, 158.9; IR (neat)
2930, 1600, 1496, 1246 cm-1; EIMS m/z (relative intensity) 273
(M+, 10); HRMS calcd for C17H23NO2 273.1729, found 273.1728.
5-(Hep t-5′-en yl)-2-p h en oxym eth yl-2,3-d ih yd r o-4H-p yr -
r ole 1-Oxid e (42). A mixture of ketal 39 (1.02 g, 2.93 mmol)
and 3 N aqueous HCl (3 mL) in THF (30 mL) was stirred at
room temperature for 4.5 h. The mixture was diluted with
water, extracted with EtOAc, dried over MgSO4, and concen-
trated. The residue was purified by flash chromatography
eluting with 4:1 EtOAc/MeOH to yield 0.65 g (78%) of nitrone
42 as a pale yellow oil: 1H NMR (300 MHz, CDCl3) δ 1.34-
1.44 (m, 2H), 1.46-1.60 (m, 2H), 1.62 (d, J ) 4.6 Hz, 3H),
1.96-2.02 (m, 2H), 2.24-2.32 (m, 2H), 2.41-2.85 (m, 4H), 4.13
(dd, J ) 2.8, 10.1 Hz, 1H), 4.29 (br s, 1H), 4.67 (dd, J ) 3.4,
22.6, 20.5, 14.1; IR (neat) 3354 (br), 1595, 1590, 1237, 990 cm-1
.
Anal. Calcd for C27H40O4: C, 75.66; H, 9.41. Found: C, 75.50;
H, 9.58.
Hyd r oxyla m in e 37. To a solution of oxime 36 (2.00 g, 4.50
mmol) in methanol (15 mL) at 0 °C was added a trace of
bromocresol green followed by sodium cyanoborohydride (426
mg, 6.80 mmol). The solution was kept at pH ) 4 by addition
of a 2 M solution of HCl in methanol. The mixture was stirred
for 2 h, and 15% aqueous KOH was then added until the
solution became basic. The product was extracted with CH2Cl2,
and the combined organic extracts were dried over MgSO4 and
concentrated. The residue was purified by flash chromatog-
raphy (35% EtOAc/hexanes) to afford 1.90 g (95%) of hydroxyl-
amine 37 as a clear colorless oil: 1H NMR (300 MHz, CDCl3)
δ 7.29 (m, 2H), 6.95 (t, J ) 7.3 Hz, 1H), 6.92 (d, J ) 7.8 Hz,
2H), 6.00 (m, 2H), 5.58 (m, 2H), 4.01 (d, J ) 5.3 Hz, 2H), 3.94
(s, 4H), 3.18 (quin, J ) 5.7 Hz, 1H), 2.05 (m, 4H), 1.82-1.59
(m, 6H), 1.34 (m, 12H), 0.89 (t, J ) 6.7 Hz, 3H); 13C NMR (75
MHz, CDCl3) δ 158.6, 132.4, 131.8, 130.4, 130.2, 129.3, 120.8,
114.5, 111.4, 66.4, 64.8, 60.8, 37.0, 33.2, 32.5, 32.4, 31.6, 29.6,
29.3, 28.8, 23.4, 22.5, 14.0; IR (neat) 3264, 1599, 1587 cm-1
;

4872 J . Org. Chem., Vol. 64, No. 13, 1999
Werner et al.
10.1 Hz, 1H), 5.33-5.45 (m, 2H), 6.87-7.29 (m, 5H); 13C NMR
(75 MHz, CDCl3) δ 17.8, 19.7, 24.5, 26.6, 29.3, 30.2, 32.1, 66.5,
72.1, 114.6, 121.1, 125.1, 129.4, 130.8, 149.3, 158.4; IR (neat)
2928, 1599, 1497, 1454, 1244 cm-1; EIMS m/z (relative
intensity) 287 (M+, 3).
129.2, 120.4, 114.3, 72.0, 68.0, 63.6, 58.7, 50.7, 43.0, 40.4, 36.9,
36.2, 31.7, 29.2, 26.0, 25.9, 24.3, 23.0, 22.4, 21.5, 13.9; IR (neat)
1707, 1602, 1584, 1243 cm-1; CIMS m/z (relative intensity) 384
(MH+, 11), 290 (MH+ - PhOH - H2O, 7), 276 (9), 69 (100);
HRMS calcd for C25H37NO2 383.2824, found 383.2815. Anal.
Calcd for C25H37NO2: C, 78.28; H, 9.72; N, 3.65. Found: C,
78.26; H, 9.77; N, 3.61.
Cycloa d d ition of Nitr on e 42. A solution of nitrone 42
(0.097 g, 0.338 mmol) in toluene (2 mL) was heated in a sealed
tube at 120 °C for 72 h. The solvent was removed, and the
residue was purified by flash chromatography eluting with
1:10 EtOAc/hexanes to afford 14 mg (15%) (38% based on
recovered nitrone) of cycloadduct 45 as a pale yellow oil: 1H
NMR (300 MHz, CDCl3) δ 1.18-1.30 (m, 2H), 1.33 (d, J ) 6.0
Hz, 3H), 1.51-1.84 (m, 8H), 1.95-2.02 (m, 2H), 2.24-2.33 (m,
1H), 3.49-3.57 (m, 1H), 3.81 (t, J ) 9.0 Hz, 1H), 4.22 (dd, J )
6.0, 10.8 Hz, 1H), 4.30 (dd, J ) 4.0, 9.0 Hz, 1H), 6.90-7.31
(m, 5H); 13C NMR (75 MHz, CDCl3) δ 20.5, 21.0, 22.4, 23.6,
26.9, 32.2, 34.9, 52.1, 66.6, 70.6, 72.4, 79.5, 114.4, 120.5, 129.4,
158.9; IR (neat) 2927, 1600, 1497, 1247 cm-1; EIMS m/z
(relative intensity) 287 (M+, 4.6); HRMS calcd for C18H25NO2
287.1885, found 287.1870.
Isoxa zolid in e 43. A solution of nitrone diene 40 (485 mg,
1.27 mmol) in DMSO (7.0 mL) was heated at 195 °C in a sealed
tube for 16 h. The solvent was removed, and the residue was
purified by flash chromatography (10% EtOAc/hexanes) to
afford 305 mg (63%) of isoxazolidine 43 as a yellow oil: 1H
NMR (500 MHz, CDCl3) δ 7.27 (m, 2H), 6.92 (m, 3H), 5.66 (dt,
J ) 15.2, 6.7 Hz, 1H), 5.37 (dd, J ) 15.2, 8.3 Hz, 1H), 4.43
(dd, J ) 10.8, 8.4 Hz, 1H), 4.32 (dd, J ) 9.0, 3.9 Hz, 1H), 3.81
(t, J ) 9.0 Hz, 1H), 3.51 (m, 1H), 2.30 (m, 1H), 2.02 (m, 4H),
1.80-1.53 (m, 8H), 1.39-1.16 (m, 10H), 0.87 (t, J ) 6.9 Hz,
3H); 13C NMR (75 MHz, CDCl3) δ 158.8, 135.8, 130.2, 129.3,
120.5, 114.4, 85.7, 72.3, 70.7, 66.2, 50.4, 34.6, 32.2, 31.7, 31.6,
28.9, 28.7, 26.4, 23.6, 22.5, 21.6, 20.9, 14.0; IR (neat) 1602,
1584, 1243, 950 cm-1; EIMS m/z (relative intensity) 383 (M+,
39), 366 (17), 276 (100), 218 (23); HRMS calcd for C25H37NO2
383.2824, found 383.2822.
Am in o Alcoh ol 49. To a solution of isoxazolidine 43 (200
mg, 0.522 mmol) in AcOH:H2O (1:1, 3 mL) was added Zn dust
(680 mg, 10.4 mmol), and the mixture was heated at 45 °C for
3 h. The solution was cooled, basified with 15% aqueous NaOH,
and extracted with EtOAc. The combined organic layers were
dried over MgSO4 and concentrated. The residue was purified
by flash chromatography (2.5-5% MeOH/CH2Cl2) to afford 183
mg (91%) of amino alcohol 49 as a yellow oil: 1H NMR (300
MHz, CDCl3) δ 7.28 (m, 2H), 6.96 (t, J ) 7.3 Hz, 1H), 6.89 (d,
J ) 8.7 Hz, 2H), 5.67 (dt, J ) 15.1, 6.6 Hz, 1H), 5.39 (dd, J )
15.2, 8.1 Hz, 1H), 4.45 (t, J ) 8.8 Hz, 1H), 4.00 (AB of ABX,
J AB ) 9.5 Hz, J AX ) 7.3 Hz, J BX ) 7.7 Hz, ∆υAB ) 33.4 Hz,
2H), 3.49 (m, 1H), 2.10-1.68 (m, 11H), 1.55-1.20 (m, 16H),
0.87 (t, J ) 6.5 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ 158.6,
132.3, 131.7, 129.4, 120.9, 114.4, 73.4, 68.2, 66.3, 56.7, 46.0,
37.7, 34.2, 32.1, 31.6, 29.0, 28.7, 28.2, 25.4, 24.1, 22.5, 19.7,
14.0; IR (neat) 3025 (br), 3025, 1596, 1237, 967 cm-1; CIMS
m/z (relative intensity) 386 (MH+, 6), 368 (MH+ - H2O, 21),
274 (MH+ - PhOH - H2O, 7), 260 (6), 69 (100); HRMS calcd
for C25H39NO2 385.2981, found 385.2997. Anal. Calcd for
Clem m en sen Red u ction of Keton e 52. Concentrated HCl
(1.5 mL) was added to a solution of tricyclic ketone 52 (82.0
mg, 0.214 mmol) in toluene (750 µL). Zn(Hg) (100 mg) was
added, and the mixture was heated to 90 °C and stirred for
16 h. Additional Zn(Hg) (100 mg) and concentrated HCl (1 mL)
was added, and after 7 h the reaction mixture was cooled to
room temperature, basified with a 15% aqueous NaOH solu-
tion, and extracted with EtOAc. The combined organic layers
were dried over MgSO4 and concentrated. The crude residue
was purified by flash chromatography (5-10% EtOAc/hexanes
gradient) to afford 40 mg (51%) of olefin 58 and 5 mg of tricyclic
indolizidine 59 (6%). Data for olefin 58: 1H NMR (300 MHz,
CDCl3) δ 7.27 (m, 2H), 6.93 (m, 3H), 5.35 (br d, J ) 6.5 Hz,
1H), 4.00 (dd, J ) 9.1, 4.1 Hz, 1H), 3.76 (t, J ) 8.4 Hz, 1H),
3.25 (m, 1H), 2.90 (ddd, J ) 6.0, 6.0, 6.0 Hz, 1H), 2.33 (m,
1H), 2.18-1.95 (m, 4H), 1.75-1.10 (m, 19H), 0.88 (t, J ) 6.6
Hz, 3H); 13C NMR (75 MHz, CDCl3) δ 159.1, 143.0, 129.3,
120.4, 114.7, 114.5, 73.0, 63.6, 62.9, 54.8, 40.8, 35.6, 34.6, 34.5,
31.9, 29.7, 27.9, 27.3, 27.0, 25.1, 24.5, 22.7, 14.1; IR (neat) 1602,
1590, 1243 cm-1; CIMS m/z (relative intensity) 386 (MH+, 80),
274 (MH+ - PhOH, 93), 260 (100); HRMS calcd for C25H38NO
368.2953, found 368.2929.
Hyd r ogen a tion of Olefin 58. To a solution of olefin 58
(32.6 mg, 0.089 mmol) in ethanol (1.5 mL) under argon was
added 10% Pd/C (100 mg). The flask was evacuated and filled
with H2. After 6 h, the reaction mixture was filtered through
Celite, and the filtrate was concentrated. The residue was
purified by flash chromatography (2.5-10% EtOAc/hexanes
gradient) to afford 23 mg (70%) of indolizidine 59 as a clear
colorless oil: 1H NMR (300 MHz, CDCl3) δ 7.27 (m, 2H), 6.91
(m, 3H), 3.82 (dd, J ) 8.9 Hz, 1H), 3.60 (t, J ) 8.8 Hz, 1H),
3.31 (m, 1H), 2.79 (m, 1H), 2.10-1.18 (m, 27H), 0.88 (t, J )
6.6 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ 159.2, 129.4, 120.4,
114.5, 72.6, 63.6, 63.2, 56.2, 36.5, 36.4, 36.0, 35.8, 32.0, 30.5,
29.6, 27.9, 26.2, 25.8, 25.3, 22.7, 22.0, 21.0, 14.1; IR (neat) 1600,
1586, 1244 cm-1; CIMS m/z (relative intensity) 370 (MH+, 24),
336 (7), 276 (MH+ - PhOH, 58), 262 (100); HRMS calcd for
C
25H40NO 370.3110, found 370.3122. The picrate salt of amine
59 was recrystallized from CH2Cl2/hexanes (1/1) by slow
evaporation for X-ray analysis (see Figure 1).
2-epi-Cylin d r icin e C (53). To a solution of ketone 52 (15
mg, 0.039 mmol) in benzene (6 mL) was added ethylene glycol
(0.190 mL, 3.39 mmol), followed by a catalytic amount of PPTS.
The reaction flask was fitted with a Dean-Stark trap, and
the mixture was heated at reflux and stirred for 48 h. The
reaction mixture was cooled to room temperature and concen-
trated. The mixture was diluted with EtOAc and washed with
a 15% NaOH solution. The organic layer was washed with
brine, dried over MgSO4, and concentrated. The crude reaction
mixture was purified by preparative TLC (5% EtOAc/hexanes,
4 elutions) to afford 5.9 mg (35%) of the ketal as a pale yellow
oil: 1H NMR (200 MHz, CDCl3) δ 7.30 (m, 2H), 6.92 (m, 3H),
4.07-3.87 (m, 4H), 3.80 (dd, J ) 9.0, 4.5 Hz, 1H), 3.65 (t, J )
9.0 Hz, 1H), 3.32 (m, 1H), 2.93 (m, 1H), 2.30-1.20 (m, 25H),
0.88 (m, 3H).
Anhydrous ammonia (5 mL) was added to a solution of the
ketal (0.014 mmol) in 0.800 mL of EtOH at -50 °C. Lithium
wire (100 mg) was added in portions over 2 h, and the reaction
mixture was diluted with aqueous NH4Cl solution. The
mixture was slowly warmed to room temperature, allowing
the ammonia to evaporate. The crude product was partitioned
between EtOAc and water, and the organic layer was washed
with brine, dried over MgSO4, and concentrated. The crude
reaction mixture was purified by flash chromatography (5%
EtOAc/hexanes) to afford 3.9 mg of a mixture of the desired
reduction product and starting material (2.5:1 ratio).
The mixture was dissolved in THF (1 mL), 3 N aqueous HCl
(0.250 mL) was added, and the reaction was stirred for 16 h.
C
25H39NO2: C, 77.87; H, 10.19; N, 3.63. Found: C, 77.71; H,
10.20; N, 3.59.
P r ep a r a tion of Tr icyclic Keton e 52. To a suspension of
Dess-Martin periodinane (325 mg, 0.768 mmol) in CH2Cl2 (1
mL) was added tert-butyl alcohol (98 µL, 1.02 mmol). After
stirring for 15 min, the suspension was added slowly to a
solution of alcohol 49 (197 mg, 0.512 mmol) in CH2Cl2 (2.5 mL).
After 40 min, the reaction mixture was quenched by the
addition of a 15% aqueous NaOH solution. The mixture was
diluted with water and extracted with CH2Cl2. The organic
extracts were combined, dried over MgSO4, and concentrated.
The crude residue was purified by flash chromatography (5%
EtOAc/hexanes) to afford 139 mg (71%) of tricyclic ketone 52
as a clear colorless oil: 1H NMR (300 MHz, CDCl3) δ 7.30 (m,
2H), 6.93 (m, 3H), 3.92 (dd, J ) 9.0, 4.5 Hz, 1H), 3.70 (t, J )
8.9 Hz, 1H), 3.47 (m, 1H), 3.31 (m, 1H), 2.64 (dd, J ) 15.5, 5.2
Hz, 1H), 2.51 (br s, 1H), 2.29 (m, 2H), 2.21 (dd, J ) 15.5, 6.9
Hz, 1H), 2.10 (m, 2H), 1.85 (m, 2H), 1.70-1.20 (m, 16H), 0.87
(t, J ) 6.7 Hz, 3H); 13C NMR (126 MHz, CDCl3) δ 211.7, 158.8,

Cylindricine and Lepadiformine Marine Alkaloid Synthesis
J . Org. Chem., Vol. 64, No. 13, 1999 4873
The reaction mixture was basified with a 15% NaOH solution
and extracted with EtOAc. The combined organic layers were
dried over MgSO4 and concentrated. The residue was purified
by flash chromatography (10-50% EtOAc/hexanes gradient)
to afford 1 mg (24%) of 2-epi-cylindricine C (53) as a clear
colorless oil: 1H NMR (300 MHz, CDCl3) δ 3.57 (dd, J ) 10.6,
4.0 Hz, 1H), 3.40-3.28 (m, 2H), 3.23 (t, J ) 5.9 Hz, 1H), 2.78
(m, 1H), 2.67 (dd, J ) 15.7, 5.9 Hz, 1H), 2.54 (m, 1H), 2.26 (m,
1H), 2.17 (dd, J ) 15.4, 6.0 Hz, 1H), 2.09-2.02 (m, 2H), 1.84-
1.20 (m, 19H), 0.88 (t, J ) 7.2 Hz, 3H); CIMS m/z (relative
intensity) 308 (MH+, 56), 290 (8), 276 (20), 222 (9).
trated. The crude product was purified by flash chromatogra-
phy (20% EtOAc/hexanes and then 0.5% NH4OH/9.5% MeOH/
90% CH2Cl2) to afford 18 mg (71%) of amino alcohol 60 as a
pale yellow oil: IR (neat) 3436 (br) cm-1; 1H NMR (300 MHz,
CDCl3) δ 3.48 (dd, J ) 10.3, 3.2 Hz, 1H), 3.27 (d, J ) 10.3 Hz,
1H), 3.13 (m, 1H), 2.65 (m, 1H), 2.12-1.17 (m, 27H), 0.88 (t, J
) 6.6 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ 64.1, 63.5, 63.1,
54.2, 37.1, 35.8, 31.9, 30.2, 29.5, 28.1, 26.1, 25.1, 22.7, 21.7,
21.0, 14.1; ESIMS m/z (relative intensity) 294 ([M + H]+, 100),
278 (5); HRMS calcd for C19H36NO 294.2797, found 294.2785.
Tr icyclic Am in o Alcoh ol 60. Anhydrous ammonia (6 mL)
was added to a solution of indolizidine 59 in 2 mL of THF/
EtOH (1/4) at -50 °C. Lithium wire (200 mg) was added in
portions over 2 h, and after stirring for an additional 2 h, the
reaction was quenched with an aqueous NH4Cl solution. The
mixture was slowly warmed to room temperature, allowing
the ammonia to evaporate. The crude product was partitioned
between ether and water, and the organic layer was washed
with brine, dried over MgSO4, and concentrated. As shown in
Ack n ow led gm en t. We are grateful to the National
Science Foundation (CHE-97-32038) for financial sup-
port of this research. Dr. Kiyohiro Samizu provided
helpful suggestions during the initial stages of the
project, and Dr. Alan Benesi assisted with 2D NMR
experiments. J .M.D. acknowledges the Consejeria de
Educacion del Gobierno Vasco (Spain) for a postdoctoral
fellowship.
1
the H NMR spectrum, the crude product obtained was a 2:3
mixture of starting material:reduction product and therefore
the mixture was resubmitted to the above procedure. The
resulting residue was dissolved in THF (1 mL), and 2 N HCl
in methanol (0.100 mL) was added. After stirring for 6 h at
room temperature, the mixture was basified with a 15%
aqueous NaOH solution and extracted with ether. The com-
bined organic extracts were dried over MgSO4 and concen-
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails for preparation of nitrone precursors 38 and 39, proton
and carbon NMR spectra of new compounds, and the NOESY
spectrum of tricyclic ketone 52. This material is available free
of charge via the Internet at http://pubs.acs.org.
J O990266S