A practical entry to the crambescidin family of guanidine alkaloids. Enantioselective total syntheses of ptilomycalin A, crambescidin 657 and its methyl ester (neofolitispates 2), and crambescidin 800

Article abstract of DOI:10.1021/ja000234i

Among the most structurally remarkable guanidine natural products are the crambescidin/ptilomycalin A family of marine alkaloids. The evolution of a practical strategy for preparing pharmacologically significant crambescidin/ptilomycalin A alkaloids that lack oxidation at C13 is described. The first total syntheses of crambescidin 800 (2), crambescidin 657 (6), and neofolitispate 2 (7) are reported in full detail. The central strategic step in these convergent total syntheses is tethered Biginelli condensation of β-keto ester 24 with ureido aminal 61 to combine all carbons of the guanidine nucleus and set the pivotal C10-C13 stereorelationship. The total synthesis of crambescidin 800 (2) was accomplished in 3% overall yield from commercially available 3-butyn-1-ol by way of 16 isolated and purified intermediates. Full details of our earlier total synthesis of ptilomycalin A (1) are also presented. The total syntheses described in this disclosure confirm the stereochemical assignments of 1, 2, 6, and 7 and rigorously establish that the absolute configuration of the hydroxyspermidine side chain of crambescidin 800 (2) is S.

Full text of DOI:10.1021/ja000234i

J. Am. Chem. Soc. 2000, 122, 4893-4903
4893
A Practical Entry to the Crambescidin Family of Guanidine Alkaloids.
Enantioselective Total Syntheses of Ptilomycalin A, Crambescidin
657 and Its Methyl Ester (Neofolitispates 2), and Crambescidin 800
D. Scott Coffey,† Andrew I. McDonald, Larry E. Overman,* Michael H. Rabinowitz,‡ and
Paul A. Renhowe§
Contribution from the Department of Chemistry, 516 Rowland Hall, UniVersity of California,
IrVine, California 92697-2025
ReceiVed January 21, 2000
Abstract: Among the most structurally remarkable guanidine natural products are the crambescidin/ptilomycalin
A family of marine alkaloids. The evolution of a practical strategy for preparing pharmacologically significant
crambescidin/ptilomycalin A alkaloids that lack oxidation at C13 is described. The first total syntheses of
crambescidin 800 (2), crambescidin 657 (6), and neofolitispate 2 (7) are reported in full detail. The central
strategic step in these convergent total syntheses is tethered Biginelli condensation of â-keto ester 24 with
ureido aminal 61 to combine all carbons of the guanidine nucleus and set the pivotal C10-C13 stereorelationship.
The total synthesis of crambescidin 800 (2) was accomplished in 3% overall yield from commercially available
3-butyn-1-ol by way of 16 isolated and purified intermediates. Full details of our earlier total synthesis of
ptilomycalin A (1) are also presented. The total syntheses described in this disclosure confirm the stereochemical
assignments of 1, 2, 6, and 7 and rigorously establish that the absolute configuration of the hydroxyspermidine
side chain of crambescidin 800 (2) is S.
Introduction
were isolated from Crambe crambe, a bright red conspicuous
species of sponge found at shallow depths along the rocky coast
of the Mediterranean. Ptilomycalin A (1), several of the
crambescidins,5b and neofolitispates 2 (7, crambescidin 657
methyl ester)7 have been reported from extracts of other warm-
water sponges, whereas 1, crambescidin 800 (2), cereromycalin
(5), and fromiamycalin (8) have been isolated also from
starfishes collected off New Caledonia.8 C. crambe is likewise
the source of 13,14,15-isocrambescidin 800 (9), which has a
different stereochemistry of the pentacyclic guanidine unit.4b,5a
With the exception of 5, 8, and 9, members of this guanidine
alkaloid family differ from each other either in the length of
the ω-hydroxycarboxylic acid unit, the presence or absence of
a hydroxyl group at C13 of the guanidine moiety, or in the nature
of the side chain terminus (carboxylic acid, ester, or polyamine
amide). Structural assignments for the pentacyclic guanidine
fragment derive almost exclusively from spectroscopic and mass
spectrometric data. Extensive NMR studies demonstrated that
the relative stereochemistry of the pentacyclic guanidine cores
of ptilomycalin A (1) and crambescidins 800 (2), 816 (3), and
A striking variety of structurally novel guanidines have been
isolated from marine organisms.1 Diverse biological activities
are associated with many of these alkaloids, likely reflecting
the multiple ways that a guanidinium cation can participate in
noncovalent interactions. Among the most remarkable marine
guanidine natural products are the family of alkaloids depicted
in Figure 1 that have a rigid pentacyclic guanidine carboxylic
acid core linked to an ω-hydroxycarboxylic acid, ester, or
polyamine amide. The first member of this group to be isolated
was ptilomycalin A (1), which was originally obtained by
Kusumi, Kashman, and co-workers from a red sea sponge,
Hemimycale sp., and a sponge found in the Caribbean.2,3 The
groups of Rinehart,4 Braekman,5 and patent applications from
Rinehart and Pharma Mar6 describe a large series of cognate
alkaloids, the crambescidins (e.g., 2-4 and 6). These latter
alkaloids, of which crambescidin 816 (3) is the most abundant,
† Current address: Lilly Research Laboratories, Lilly Corporate Center,
Indianapolis, IN 46285.
‡ Current address: Glaxo, Inc. Research Institute, 5 Moore Dr., Research
Triangle Park, NC 27709.
(4) (a) Jares-Erijman, E. A.; Sakai, R. Rinehart, K. L. J. Org. Chem.
1991, 56, 5712-5715. (b) Jares-Erijman, E. A.; Ingrum, A. L.; Carney, J.
R.; Rinehart, K. L.; Sakai, R. J. Org. Chem. 1993, 58, 4805-4808. (c)
Rinehart, K. L. Abstracts of Papers, 213th National Meeting of the American
Chemical Society, San Franciso, CA, April 13-17, 1997; American
Chemical Society: Washington, DC, 1997; ORG 348.
§ Current address: Chiron Corporation, 4560 Horton St., Emeryville, CA
94608-2916.
(1) For reviews, see: (a) Berlinck, R. G. S. Nat. Prod. Rep. 1999, 16,
339-365. (b) Berlinck, R. G. S. Nat. Prod. Rep. 1996, 13, 377-409. (c)
Berlinck, R. G. S. Prog. Chem. Org. Nat. Prod. 1995, 66, 119-295. (d)
Faulkner, D. J. Nat. Prod. Rep. 1999, 16, 155-198, and earlier reviews in
this series.
(2) (a) Kashman, Y.; Hirsh, S.; McConnell, O. J.; Ohtani, I.; Kusumi,
T.; Kakisawa, H. J. Am. Chem. Soc. 1989, 111, 8925-8926. (b) Ohtani, I.;
Kusumi, T.; Kakisawa, H.; Kashman, Y.; Hirsh, S. J. Am. Chem. Soc. 1992,
114, 8472-8479. (c) Ohtani, I.; Kusumi, T.; Kakisawa, H. Tetrahedron
Lett. 1992, 33, 2525-2528.
(5) (a) Berlinck, R. G. S.; Braekman, J. C.; Daloze, D.; Bruno, I.; Riccio,
R.; Ferri, S.; Spampinato, S.; Speroni, E. J. Nat. Prod. 1993, 56, 1007-
1015. (b) Tavares, R.; Daloze, D.; Braekman, J. C.; Hajdu, E.; Muricy, G.;
Van Soest, R. W. M. Biochem. Syst. Ecol. 1994, 22, 645-646.
(6) (a) Shi, J.-G.; Sun, F.; Rinehart, K. L. WO Patent 3,756,734, 1998.
(b) Rinehart, K. L.; Jares-Erijman, E. A. U.S. Patent 5,756,734, 1998.
(7) Venkateswarlu, Y.; Reddy, M. V. R.; Rao, P. R.; Rao, J. V. Indian
J. Chem., Sect. B 1999, 38, 254-256.
(3) The Caribbean sponge was originally identified as Ptilocaulis aff.
Spiculifer. A reexamination of the voucher specimen has led to this sponge
being characterized as belonging to the genus Batzella Topsent, 1891, which
is closely related to the genera Crambe and Monanchora.5
(8) Palagiano, E.; De Marino, S.; Minale, L.; Riccio, R.; Zollo, F.; Iorizzi,
M.; Carre, J. B.; Debitus, C.; Lucarain, L.; Provost, J. Tetrahedron 1995,
51, 3675-3682.
10.1021/ja000234i CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/05/2000

4894 J. Am. Chem. Soc., Vol. 122, No. 20, 2000
Coffey et al.
ATP at its binding site.10 Ptilomycalin A (1) and crambescidins
800 (2), 816 (3), 844 (4), and 657 (6) are reported to exhibit
low- to mid-nanomolar activities in vitro against L1210 murine
leukemia cells,4c,6 and similar levels of cytotoxicity against
several human cancer cell lines including lung carcinoma A-549,
colon carcinoma HT-29, and melanoma MEL-28.2,5,6,11 Synthetic
ptilomycalin A (NSC 700559) exhibits low- to mid-nanomolar
activity against many human tumors in the National Cancer
Institute (NCI) in vitro panel, with particular selectivity realized
against nonsmall cell lung cancer (HOP-62) and melanoma
(M14).12 Crambescidin 816 is also reported to be a powerful in
vitro Ca2+-channel blocker (IC50 ) 0.15 nM; nifedipine ) 1.2
µM in this assay).5a
Synthesis Plan. A molecular mechanics model of the methyl
ester of the crambescidin/ptilomycalin A pentacyclic guanidine
core is shown in Figure 2. The central triazaacenaphthalene ring
system of these alkaloids is nearly planar with the seven- and
six-membered cyclic ethers being oriented on one face. Because
the two C-O bonds are axial, we surmised that the C8 and
C15 spirocenters would assemble with the required stereochem-
istry if the central triazaacenaphthalene unit had the proper cis
stereochemistry. At the outset of our endeavors, we envisaged
that setting the cis stereorelationship of the angular hydrogens
at C10 and C13 and relating the chirality of this unit to the C3
and C19 stereogenic centers of the oxepene and hydropyran
rings would be critical elements in evolving a stereocontrolled
strategy for preparing the crambescidin/ptilomycalin A class of
guanidine alkaloids.
As illustrated in Scheme 1, disconnection of the C8 aminal
and retrosynthetic cleavage of the C15-O bond of 13 leads to
1-oxohexahydropyrrolo[1,2-c]pyrimidine carboxylic ester (X )
O) or 1-iminohexahydropyrrolo[1,2-c]pyrimidine carboxylic
ester (X ) NH2) intermediates 14. The 5-alkoxycarbonyl-
1,2,3,4-tetrahydropyrimidine part structure of 14 (X ) O)
suggested to us that this bicyclic intermediate might be prepared
by a modification of the three-component Biginelli condensation
in which the urea and aldehyde reactants would be linked as
Figure 1. Pentacyclic marine guanidine alkaloids.
657 (6) are identical, whereas 13,14,15-isocrambescidin 800 (9)
is epimeric at C13, C14, and C15.2,4-6 The absolute configu-
ration of the guanidine moieties of 3 and 9 was established by
oxidative degradation of the oxepene rings of these alkaloids
to yield (S)-2-hydroxybutanoic acid,4a,b whereas the absolute
configuration of the hydroxyspermidine unit of 3 was assigned
1
using Mosher’s method.5a,9 Because H NMR and 13C NMR
chemical shifts of the hydroxyspermidine fragment of 3 are
nearly identical to those of 2 and 9, it has been assumed that
the stereochemistry of the hydroxyspermidine is the same for
all crambescidins containing this unit.
Little chemistry of these natural products has been described.
Ptilomycalin A (1) was reported to give intractable mixtures
upon attempted hydrolysis under both acidic and basic condi-
tions or upon metal hydride reduction. Methanolysis of di-p-
bromobenzoate derivative 10 was reported to yield side-chain
fragment 11 and tetracyclic vinylogous urethane 12 (eq 1).2b In
no instance has degradation of these alkaloids provided an intact
pentacyclic guanidine carboxylic acid.
Ptilomycalin A and the crambescidin alkaloids exhibit a
variety of pharmacological activities. Antiviral activity against
Herpes simplex virus, type 1, antifungal activity against Candida
albicans, and anti-HIV activities have been described.2b,4a
Ptilomycalin A is reported to be the first nonnucleotide analogue
that inhibits Na+, K+- and Ca2+-ATPases by interaction with
(10) Ohizumi, Y.; Sasaki, S.; Kusumi, T.; Ohtani, I. I. Eur. J. Pharmacol.
1996, 310, 95-98.
(11) Greater activity against L1210 (5-10 ×) has been reported for
chlorospermidine analogues of crambescidin 800 and 816.4c,6a
(12) National Cancer Institute Developmental Therapeutics Program
testing to L.E.O., dated March 12, 1998.
(9) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512-519.

Crambescidin Family of Guanidine Alkaloids
J. Am. Chem. Soc., Vol. 122, No. 20, 2000 4895
Scheme 1
Figure 2. Molecular mechanics model of the pentacyclic core of the
crambescidin/ptilomycalin A alkaloids.
depicted in 15.13 This analysis had the appeal of high conver-
gence, because the left-hand three rings of 13 would derive from
acyclic fragment 15, while the right two rings and the ester side
chain would be incorporated as the simple â-keto ester unit 16.
In 1993, we reported initial model studies that established
the viability of “tethered-Biginelli” condensations (eq 2) and
verified that the cis orientation of the angular methine hydrogens
could be preferentially realized when this dehydrative condensa-
tion was promoted under Knoevenagel conditions.14 A more
thorough study of stereoselection in tethered Biginelli condensa-
tions was recently published and revealed that either bicyclic
stereoisomer can be formed preferentially using the proper acyl
substituent and reaction conditions.15
guanidine (or a guanidine precursor) to stitch together an acyclic
dihydroxy dienedione. The ranking accomplishment of these
studies is Snider’s early construction of the pentacyclic nucleus
of ptilomycalin A from 19 (R1 ) CO2Me, R2 ) Et, R3 ) Me).18b
The appeal of this approach to the ptilomycalin A/crambescidin
alkaloids is compromised in part by the inability of the remote
secondary alcohol stereocenters C3 and C19 to control in any
way the orientation of the cis C10 and C13 angular hydrogens.
The biomimetic approach has proven particularly powerful for
the preparation of simple pentacyclic congeners from the direct
reaction of guanidinium salts with achiral dienediones (dihydro
analogues of 19 with R1 ) R2 ) R3 ) H).19
Before describing our investigations that led to the first total
syntheses of members of the crambescidin/ptilomycalin A
alkaloid group,16 the accomplishments registered by the groups
of Snider and Murphy in assembling the basic guanidinium units
of these alkaloids from acyclic precursors must be mentioned.17-19
The presumed biomimetic strategy illustrated in eq 3 has the
appeal that pentacycles such as 20 would be assembled using
Results and Discussion
Enantioselective Total Synthesis of Ptilomycalin A (1). At
the time our investigations began, ptilomycalin A was the sole
member of this alkaloid group to have been reported. In light
of the difficulty experienced during degradation studies in
removing the ester side-chain of 1,2 we chose to incorporate
the 16-hydroxyhexadecanoic acid fragment from the outset. The
synthesis of â-ketoester 24, which incorporates this unit, is
summarized in Scheme 2. Alkylation of the dianion of methyl
acetoacetate (21)20 with enantiopure (R)-siloxy iodide 22
provided 23 in 73% yield. Iodide 22 is available in high yield
from methyl (R)-2-hydroxybutanoate.21a Selective transesteri-
fication of 23 with allyl 16-hydroxyhexadecanoate using
4-(dimethylamino)pyridine (DMAP) as catalyst gave 24 in 64%
overall yield from 21.22
(13) (a) Biginelli, P. Gazz. Chim. Ital. 1893, 23, 360. (b) For a recent
review, see: Kappe, C. O. Tetrahedron 1993, 49, 6937-6963.
(14) Overman, L. E.; Rabinowitz, M. H. J. Org. Chem. 1993, 58, 3235-
3237.
(15) McDonald, A. I.; Overman, L. E. J. Org. Chem. 1999, 64, 1520-
1528.
(16) Overman, L. E.; Rabinowitz, M. H.; Renhowe, P. A. J. Am. Chem.
Soc. 1995, 117, 2657-2658.
(17) For brief reviews of synthetic work in this area, see refs 1a and 1b.
(18) (a) Snider, B. B.; Shi, Z. Tetrahedron Lett. 1993, 34, 2099-2102.
(b) Snider, B. B.; Shi, Z. J. Am. Chem. Soc. 1994, 116, 549-557.
(19) (a) Murphy, P. J.; Williams, H. L.; Hursthouse, M. B.; Abdul Malik,
K. M. J. Chem. Soc., Chem. Commun. 1994, 119-120. (b) Murphy, P. J.;
Williams, H. L. J. Chem. Soc., Chem. Commun. 1994, 819-820. (c)
Murphy, P. J.; Williams, H. L.; Hibbs, D. E.; Hursthouse, M. B.; Abdul
Malik, K. M. Tetrahedron 1996, 52, 8315-8332. See also, Nagasawa, K.;
Georgieva, A.; Nakata, T. Tetrahedron 2000, 56, 187-192.
(20) Huckin, S. N.; Weiler, L. J. Am. Chem. Soc. 1974, 96, 1082-1087.
(21) (a) Kitamura, M.; Tokunaga, M.; Ohkuma, T.; Noyori, R. Org.
Synth., Coll. Vol. 9 1998, 589-595. (b) Taber, D. F.; Silverberg, L. J.
Tetrahedron Lett. 1991, 32, 4227-4230.

4896 J. Am. Chem. Soc., Vol. 122, No. 20, 2000
Coffey et al.
Scheme 2
Scheme 4
Scheme 3
Under the conditions developed during our original model
study,14 Biginelli condensation of crude 28 and â-keto ester 24
took place in low yield. A number of reaction parameters were
then surveyed and it was soon found that the efficiency of the
Biginelli reaction improved considerably in polar solvents. Best
results were achieved by heating a mixture of crude 28, 1.5
equiv of â-keto ester 24, 1 equiv of morpholinium acetate, a
catalytic amount of acetic acid, and excess Na2SO4 at 70 °C in
EtOH. Purification of the resulting product on silica gel provided
cis adduct 29 in 61% yield and trans adduct 30 in 8% yield.
Stereochemical assignments for these hexahydropyrrolo[1,2-c]-
pyrimidines followed from the similarity of their angular
methine hydrogen signals (29: 4.25 and 4.11 ppm; 30: 4.44
and 4.09 ppm) with those of 18 and its trans epimer, the latter
of which had earlier been analyzed by single-crystal X-ray
diffraction analysis.14 The condensation reported in Scheme 3
was carried out on a large scale (17 g of 24 and 4 g of 28) and
the yield obtained in this experiment appears to be reliable.
However, high efficiency under these conditions was not
reproducible, presumably due to the indeterminate nature of 28.
In a recent detailed examination of stereoselection in related
Biginelli condensations, a reliable procedure for generating the
electrophilic reaction component and for carrying out the
Biginelli condensation was developed; these conditions
consistently provide cis adducts such as 29 in yields of 60-
65%.15
Because the tethered Biginelli condensation had just been
developed,14 we elected in this first generation effort to arrive
at the central condensation reaction as early as possible in the
synthetic sequence. For this reason, the electrophilic component
of the Biginelli condensation was simplified from that depicted
in Scheme 1 by deletion of the C1-C7 fragment. The precursor
of this less elaborate intermediate, unsaturated urea 27, was
prepared in three steps from enantiopure methyl (R)-3-hydroxy-
7-methyloct-6-enoate (25)21 as summarized in Scheme 3.
Mitsunobu displacement of alcohol 25 with hydrazoic acid
followed by reduction of the crude â-azido ester with LiAlH4
gave S amino alcohol 26 in 72% yield and in >98% enantio-
meric excess (ee).23 Use of other nitrogen nucleophiles such as
phthalimide in the Mitsunobu reaction led to significant amounts
of the corresponding R,â-unsaturated ester. Reaction of 26 with
potassium cyanate and HCl under standard conditions pro-
vided unsaturated urea 27 in 82% yield after recrystallization.
Ozonolysis of 27 in MeOH at -78 °C followed by reduction
of the intermediate hydroperoxide with Me2S and concentration
furnished a viscous yellow oil. Further concentration of this
product at 0.1 Torr for 5 days at 50 °C to remove residual
Me2SO led to a nearly colorless amorphous powder. This crude
intermediate is more complex than formulation 28 implies.
Multiple signals were observed for many carbon atoms in the
Although 29 could be converted in one step to spirotricyclic
intermediate 31 by exposure to a slight excess of p-toluene-
sulfonic acid monohydrate (p-TsOH‚H2O), the reaction was
more reproducible on a large scale if the tert-butyldimethylsilyl
(TBDMS) group was first discharged with pyridinium p-
toluenesulfonate (PPTS) in MeOH and the resulting alcohol was
then cyclized at room temperature in CHCl3 with a catalytic
amount of p-TsOH‚H2O (Scheme 4). This sequence provided a
single tricyclic product 31 in nearly quantitative yield. The 11.5
Hz diaxial-coupling constant of the C14 methine hydrogen of
31 signaled that this intermediate was epimeric to ptilomycalin
A at C14.24
1
13C NMR spectra and the H NMR spectrum was broad. No
That diastereoselection would be high in forming the spiro-
hydropyran had been established in our earliest model study14
and can be rationalized as outlined in Scheme 5. Axial
protonation of the vinylogous carbamate 34 would generate
N-acyliminium cation 35. Spirocyclization of this intermediate
from the â-face to generate 36 would be favored to maximize
a staggered conformation with respect to the forming bond.25,26
aldehyde signal was apparent, and mass spectral data indicated
the presence of higher molecular weight materials derived from
dehydrative oligomerization. Because all attempts to enhance
the purity of 28 by chromatography were unsuccessful, we
proceeded with this crude intermediate.
(22) Taber, D. F.; Amedio, J. C., Jr.; Patel, Y. K. J. Org. Chem. 1985,
50, 3618-3619.
(23) Enantiomeric excess was determined by evaluation of the 19F NMR
spectra of the corresponding (R)- and (S)-Mosher amides,9 see Supporting
Information.
(24) The crambescidin numbering system is employed in the discussion
of all synthetic intermediates. Proper IUPAC designations of intermediates
can be found in the Experimental Section or Supporting Information.

Crambescidin Family of Guanidine Alkaloids
J. Am. Chem. Soc., Vol. 122, No. 20, 2000 4897
Scheme 5
Scheme 6
Although epimerization of 31 to the axial ester might have
been possible at this point, we chose to defer this adjustment
to the final stage of the synthesis, hoping to benefit from a
presumed (incorrect as it turned out) thermodynamic preference
for this group to be axial in the natural product. To prepare for
the addition of the remaining carbons of the guanidine core, 31
was oxidized with the Swern reagent27 to provide 32 and the
urea functional group was protected and activated for subsequent
guanidine formation by O-methylation (Scheme 4). This delicate
methylation had to be performed under carefully prescribed
conditions, and pseudourea product 33 had to be purified rapidly
on Et3N-treated silica gel, or else significant epimerization at
C10 was observed. Reversible â-elimination of the pseudourea
group is undoubtedly responsible for the erosion of C10
stereochemistry.
At this juncture, the remaining seven carbons of the penta-
cyclic guanidine nucleus needed to be appended, an elaboration
that proved to be extremely challenging. Bromide 38 was
available in high yield from alcohol 37 (86% ee) by reaction
with CBr4 and 1,2-bis(diphenylphosphino)ethane (dppe) (Scheme
6).28 In early scouting studies, we were unsuccessful in
efficiently coupling lithium or cerium reagents derived from
bromide 38 with a benzyl ester congener of aldehyde 33.
Epimerization of 33 at C10 in the presence of Lewis acidic
reagents emerged as a critical complicating issue. We eventually
found that the Grignard reagent derived from 38 could be joined
to 33 in acceptable yield at -78 °C in tetrahydrofuran (THF).
Quenching this reaction at low temperature with morpholinium
acetate and immediate filtration of the reaction mixture to
remove magnesium salts provided the adduct as a mixture of
alcohol epimers. Direct oxidation of this intermediate under
Swern conditions27 then provided 39 in 58% yield from 33.
Approximately 5% of a diastereomer arising from the minor
enantiomer of alcohol 37 was removed at this point. This
sequence was extremely delicate and yields were markedly
eroded if magnesium salts resulting from the Grignard step were
not removed.
Cleavage of the silyl protecting group of 39 with tetra-
butylammonium fluoride (TBAF) furnished alcohol 40, which
was exposed to ammonia and ammonium acetate under condi-
tions similar to those originally reported by Snider.18b After
purification of the crude product on silica gel using an eluent
containing formic acid, 41 was isolated in 51% as its formate
salt (1H NMR: δ 8.23; 13C NMR: δ 165.8). Formation of the
second spiroaminal took place exclusively by axial C-O bond
formation, because only a single pentacyclic guanidine was
detected.
A model of the tetracyclic cation 46, which is the likely direct
precursor of pentacyclic guanidine 41, is shown in Figure 3;
the C1-C7 side chain was replaced with a methyl group to
generate this model. It is apparent from examining the model
that torsional interactions would be minimized by axial addition
of the oxygen nucleophile to the electron-deficient carbon.25
The total synthesis of (-)-ptilomycalin A was readily
completed from 41. The allyl ester of this intermediate was
cleaved using palladium(0) catalysis29 and the resulting acid was
(25) See, inter alia: (a) Cherest, M.; Felkin, H. Tetrahedron Lett. 1968,
2199-2204, 2205-2208. (b) Lucero, M. J.; Houk, K. N. J. Org. Chem.
1998, 63, 6973-6977.
(26) Torsional steering as a potential control element in spirocyclization
is developed more fully later in the text in the discussion of forming the
spirooxepene.
(27) Mancuso, A. J.; Huang, S.-L.; Swern, D. J. Org. Chem. 1978, 43,
2480-2482.
(28) The alcohol precursor 37 of bromide 38 used in these experiments
was prepared in 86% ee by asymmetric reduction of an ynone precursor.16
A better synthesis of alcohol 37, which provides material of high
enantiopurity, is shown in Scheme 7.
(29) Deziel, R. Tetrahedron Lett. 1987, 28, 4371-4372.

4898 J. Am. Chem. Soc., Vol. 122, No. 20, 2000
Coffey et al.
Scheme 7
Figure 3. Model showing the expected preference for axial addition
in forming the oxepene ring.
coupled with di-tert-butoxycarbonyl (BOC)-protected spermi-
dine 4230 to generate amide 43. The ester was then epimerized
by heating in MeOH in the presence of excess Et3N. Epimer-
ization under these conditions, however, favored the â epimer
to the extent of 2-3:1. Consequently, three recycles were
required to obtain axial ester 44 in 50% yield. The equatorial
C14 methine hydrogen of 44 showed a diagnostic doublet (J )
initially protected to give p-methoxybenzyl (PMB) ether 48
(Scheme 7).32 Deprotonation of 48 with n-butyllithium at -40
°C, trapping the acetylide intermediate with anhydrous dimeth-
ylformamide (DMF), and quenching the intermediate R-amino-
alkoxide into aqueous phosphate buffer33provided ynal 49 in
90% overall yield. The C3 stereocenter was introduced by the
method of Weber and Seebach34 by reaction of 49 with Et2Zn
in the presence of 20 mol % of (4R,5R)-2,2-dimethyl-R,R,R′,R′-
tetra(naphth-2-yl)-1,3-dioxolan-4,5-dimethanol [(-)-TADDOL]
and Ti(Oi-Pr)4 to give 50 in 94% yield. This excellent
asymmetric transformation could be performed reliably on scales
as large as 50 g to provide 50 in 99% ee.9 A triisopropylsilyl
(TIPS) group was next introduced to protect the C3 hydroxyl
group, because this ether was expected to survive the conditions
of the Biginelli condensation. Semi-hydrogenation of 51 using
Lindlar’s catalyst yielded cis-alkene 52 and removal of the
p-methoxybenzyl group from this intermediate with 2,3-dichloro-
5,6-dicyano-1,4-benzoquinone DDQ provided alcohol 53. By
purifying only intermediates 49 and 50, the synthesis of 53 could
be realized in 80% overall yield from commercially available
3-butyn-1-ol. Conversion of alcohol 53 to iodide 54 was
accomplished in standard fashion in high yield.35
1
4.8 Hz) at δ 2.93 in the H NMR spectrum. Finally, cleavage
of the BOC protecting groups of 44 with HCO2H, followed by
concentration and washing with aqueous NaOH-NaCl provided
(-)-ptilomycalin A trihydrochloride (1) in good yield. Synthetic
1 showed 1H and 13C NMR spectra consistent with those
reported for (-)-ptilomycalin A2a,b and was indistinguishable
from an authentic sample by thin-layer chromatography (TLC)
comparisons on three adsorbents. Synthetic 1 was converted to
1
ditrifluoroacetate derivative 45, which also exhibited H and
13C NMR spectra indistinguishable from those reported.2b The
specific rotation of synthetic 45, [R]23 -15.9 (c 0.8, CHCl3),
D
was identical within experimental precision to the rotation,
[R]23D -15.8 (c 0.7, CHCl3), reported for this well-characterized
derivative of the natural product.2b
Second-Generation Synthesis Plan. Motivated by the thera-
peutic potential of the crambescidin/ptilomycalin A class of
guanidine alkaloids and the opportunity to use the tools of
organic synthesis to probe the molecular origins of antitumor
activity in this series, a second-generation synthesis approach
to these targets was developed, which we hoped would provide
practical synthetic access to these structures. A weakness of
the strategy we employed to prepare ptilomycalin A was the
series of delicate transformations that had to be carried out on
advanced intermediates to introduce C1-C7. This elaboration
would be avoided, and the overall convergency of the synthe-
sis would be enhanced, if all carbons of the pentacyclic core
could be joined in the pivotal Biginelli condensation. This
optimally convergent approach is outlined in Scheme 1 and was
initially implemented to achieve the inaugural total syntheses
of crambescidin 800 (2), crambescidin 657 (6), and its methyl
ester neofolitispates 2 (7).
The remaining carbons of the C1-C13 fragment came from
methyl (R)-3-hydroxy-7-methyloct-6-enoate (25) (Scheme 8).
This intermediate, which is readily available in 97% ee,21 was
transformed to amide 55 in 88% yield by reaction with N,O-
dimethylhydroxylamine hydrochloride,36 followed by protection
of the secondary alcohol as a triethylsilyl (TES) ether. The
lithium reagent derived from iodide 54 was then coupled with
55 at -78 °C to generate dienone 56 in 60-70% yield. Masking
(31) A preliminary report of the synthesis of 59 has appeared, see:
Coffey, D. S.; McDonald, A. I.; Overman, L. E.; Stappenbeck, F. J. Am.
Chem. Soc. 1999, 121, 6944-6945.
Total Synthesis of Crambescidins 800 (2) and 657 (6) and
Neofolitispates 2 (7). Developing a rapid enantioselective
construction of the common C1-C13 fragment (15 of Scheme
1) was critical to evolving an efficient route for the total
synthesis of the ptilomycalin A/crambescidin alkaloids.31 The
synthesis of this unit began with 3-butyn-1-ol (47), which was
(32) (a) Takaku, H.; Ueda, S.; Ito, T. Tetrahedron Lett. 1983, 24, 5363-
5366. (b) Nakajima, N.; Horita, K.; Abe, R.; Yonemitsu, O. Tetrahedron
Lett. 1988, 29, 4139-4142.
(33) Journet, M.; Cai, D.; DiMichele, L. M.; Larsen, R. D. Tetrahedron
Lett. 1998, 39, 6427-6428.
(34) Weber, B.; Seebach, D. Tetrahedron 1994, 50, 7473-7484.
(35) Singh, A. K.; Bakshi, R. K.; Corey, E. J. J. Am. Chem. Soc. 1987,
109, 6187-6189.
(30) Cohen, G. M.; Cullis, P. M.; Hartley, J. A.; Mather, A.; Symons,
M. C. R.; Wheelhouse, R. T. J. Chem. Soc., Chem. Commun. 1992, 298-
300.
(36) Garigipati, R. S.; Tschaen, D. M.; Weinreb, S. M. J. Am. Chem.
Soc. 1985, 107, 7790-7792.

Crambescidin Family of Guanidine Alkaloids
J. Am. Chem. Soc., Vol. 122, No. 20, 2000 4899
Scheme 8
Scheme 9
the C8 carbonyl of 54 as a ketal became necessary to prevent
dehydration of the â-hydroxy ketone under the Mitsunobu
conditions employed in a subsequent step to install the C10
amino group. However, ketalization of 56 was quite sluggish.
We eventually discovered that when the â-hydroxy group of
this intermediate was not protected, ketalization was much
easier. Ultimately, conditions were developed that cleaved the
TES group of 56, did not promote dehydration of the intermedi-
ate â-hydroxy ketone, and effected ketalization. Thus, reaction
of 56 with ortho ester 5837 and 1,3-propanediol in the presence
of Amberlyst-15 at room temperature provided hydroxy ketal
57 in 80% yield. Mitsunobu displacement of the secondary
alcohol of 57 with azide followed by reduction to the amine
delivered 59 in 77% yield from 57. The sequence summarized
in Schemes 7 and 8 is readily scaled and provided amine 59 on
multigram scales in 11 steps and ∼30% overall yield from
commercially available 3-butyn-1-ol.
Scheme 10
The critical tethered Biginelli condensation was next realized
with little difficulty. Condensation of amine 59 with trimethyl-
silyl isocyanate provided urea 60 in high yield (Scheme 9).
Selective dihydroxylation of the trisubstituted double bond of
60,38 followed by cleavage of the vicinal diol with Pb(OAc)4
in toluene and addition of morpholinium acetate yielded
intermediate 61, which was used without purification in the
critical Biginelli condensation step. Using conditions that we
had previously optimized with simpler urea aminal intermedi-
ates,15 crude 61 was condensed with 2.8 equiv of â-ketoester
24 at 60 °C in 2,2,2-trifluoroethanol to give a 6-7:1 mixture
of cis and trans hexahydropyrrolopyrimidines 62 and 63 in 61%
overall yield from 60. Although these stereoisomers could be
separated by HPLC, they were difficult to resolve on a
preparative scale. Therefore, this mixture was carried on to a
tricyclic intermediate where isomer separation was straightfor-
ward. Stereochemical assignments for the hexahydropyrrolo-
[1,2-c]pyrimidines 62 and 63 again followed from the similarity
of 1H NMR signals for their H13 methine hydrogens (62: 4.22
ppm; and 63: 4.44 ppm) with those of 18 and its trans epimer.14
The remaining three rings of the pentacyclic guanidine
nucleus were generated as follows. First, the silyl protecting
groups of 62 were removed with TBAF to provide the
corresponding urea diol (Scheme 10). Brief exposure of this
intermediate to 1 equiv of p-TsOH‚H2O induced formation of
the spirohydropyran and discharge of the ketal to generate 64
in 71% yield for the two steps. As in our earlier synthesis of
ptilomycalin A, cyclization to form the spirocyclic hydropyran
took place with untarnished stereoselectivity.
(37) (a) Roush, W. R.; Gillis, H. R. J. Org. Chem. 1980, 45, 4283-
4287. (b) Baganz, H.; Domaschke, L. Chem. Ber. 1958, 91, 650-653.
(38) Sharpless, K. B.; Williams, D. R. Tetrahedron Lett. 1975, 3045-
3046.
Initial survey experiments showed that it would not be easy
to selectively activate the urea functional group of 64. Conse-

4900 J. Am. Chem. Soc., Vol. 122, No. 20, 2000
Coffey et al.
quently, the secondary alcohol of 64 was protected as its
chloroacetate derivative, this protecting group being chosen
because it would be removed upon exposure to ammonia. At
this stage, the minor trans isomer (∼12%) resulting from the
Biginelli condensation was removed by silica gel chromatog-
raphy and isomerically pure 65 was obtained in 86% yield.
Exposure of 65 to excess MeOTf in the presence of 2,6-di-tert-
butylpyridine delivered methyl pseudourea 66, a labile inter-
mediate that was directly cyclized to the pentacyclic guanidine.
It was critical that 66 not be exposed to silica gel, because
epimerization at C10 and some decomposition resulted under
typical chromatographic purification conditions. After consider-
able experimentation, it was found that saturating an allyl alcohol
solution of crude 66 (buffered with 2 equiv of NH4Cl) with
anhydrous ammonia at room temperature, followed by heating
the resulting solution in a resealable tube at 60 °C for 1 day
provided a 1.5:1 mixture of pentacyclic guanidines 67 and 68
in 81% yield. That 1.5:1 represented the equilibrium ratio of
the C14 epimers under these conditions was readily established
by resubmission of 67 and 68 to the reaction conditions.
Separation of these isomers on silica gel and two recycles of
67 (NH3-NH4Cl, allyl alcohol, 60 °C) provided 68 (1H NMR
of H14: J ) 4.8 Hz) in 52% overall yield from urea 65.
Scheme 11
The sequence summarized in Scheme 10 for assembling
pentacyclic guanidine 68 represents a significant improvement
over that employed in our first-generation synthesis of ptilo-
mycalin A. Not only is the overall yield higher, but material
throughput is increased because the pentacyclic guanidine
nucleus is assembled under conditions that equilibrate the C14
ester epimers. Moreover, the guanidine products are obtained
directly with chloride counterions, obviating the need for
detrimental aqueous washes. The use of allyl alcohol as solvent
allowed higher concentrations of ammonia to be employed
without complicating transesterification of the allyl hexa-
decanoate side chain.
purification of the crude product provided the trihydrochloride
1
salt of crambescidin 800 (2) in 75% yield. H and 13C NMR
data for synthetic 2 were in accord with those reported for
natural 2,43 and synthetic 2 was indistinguishable from a natural
specimen by HPLC comparisons using three eluents.4a,5a Syn-
The total syntheses of crambescidin 657 (6), neofolitispates
2 (7), and crambescidin 800 (2) were culminated as summarized
in Scheme 11. Initially the allyl protecting group of 68 was
cleaved with Pd(PPh3)4 and morpholine29 and the product was
washed with dilute HCl and purified on silica gel to provide
acid 69 in 94% yield (Scheme 11). This acid was quantitatively
converted to the carboxylate inner salt 6 by washing with dilute
NaOH. The 13C NMR spectrum of this latter product in CDCl3
was identical (mean difference < (0.1 ppm) to that reported
for natural crambescidin 657.39 The specific rotation of synthetic
6 was [R]23 -13.6 (c 0.45, MeOH), whereas [R]23 -12.1 (c
1
thetic 2 was also converted to the triacetate derivative 72. H
and 13C NMR data for synthetic 72 were in perfect accord with
those reported for naturally derived 72.4a,5a The specific rotation
of synthetic 72 was [R]25D -37 (c 0.20, CHCl3), whereas [R]25
D
-43 (c 0.15, CHCl3) was reported for this derivative of the
natural isolate.5a
No criteria had suggested that synthetic 2 might differ from
natural crambescidin 800. Nonetheless, because the configura-
tion of crambescidin 800 at C43 had not been established, we
thought it prudent to vouchsafe that we could distinguish a C43
epimer. If the guanidine and hydroxy spermidine domains did
not interact appreciably, differences between these two permuta-
tions might be difficult to discern. To this end, we prepared the
C43 epimer of crambescidin 800 from 69 and the R enantiomer
of 70.40,44 19F NMR studies demonstrated that the Mosher
derivative 73 prepared from synthetic 2 was indistinguishable
from this derivative prepared from a 150 µg sample of natural
2, yet different from the identical derivative of the C43 epimer
of synthetic 2.
D
D
0.34, MeOH) is reported for the natural isolate.6a Reaction of
synthetic 69 with diazomethane provided methyl ester 7,6a [R]23
D
-17 (c 0.2, CHCl3), whose 13C NMR spectrum agreed in all
respects (mean difference (0.13 ppm) with that reported for
neofolitispates 2, [R]23 -18 (c 1, CHCl3), obtained from the
D
sponge Neofolitispa dianchora collected at the Andaman Islands,
India.7
Coupling of carboxylic acid 69 with (S)-7-hydroxyspermidine
7040 using benzotriazol-1-yloxytris(dimethylamino)phosphonium
hexafluorophosphate (BOP reagent)41 provided the correspond-
ing amide 71 in 82% yield. Removal of the BOC groups with
3 M HCl in ethyl acetate42 followed by reversed-phase HPLC
Conclusion. The first total syntheses of crambescidin 800
(2), crambescidin 657 (6), and neofolitispates 2 (7) were
accomplished in enantioselective fashion using a convergent
strategy. The central strategic step is tethered Biginelli conden-
(39) Natural crambescidin 657 is reported as an inner salt.6a
(40) Coffey, D. S.; McDonald, A. I.; Overman, L. E. J. Org. Chem. 1999,
64, 8741-8742.
(43) Table 2 of ref 4a lists MeOD as the solvent for NMR spectra of
natural crambescidin 800. This designation is apparenly in error; the solvent
should be CDCl3.
(44) A more detailed discussion of the challenge in distinguishing C43
epimers in the crambescidin series is found in the accompanying paper:
Coffey, D. S.; Overman, L. E.; Stappenbeck, F., following paper in this
issue.
(41) Castro, B.; Dormoy, J. R.; Evin, G.; Selve, C. Tetrahedron Lett.
1975, 1219-1222.
(42) Stahl, G. L.; Walter, R.; Smith, C. W. J. Org. Chem. 1978, 43,
2285-2286.

Crambescidin Family of Guanidine Alkaloids
J. Am. Chem. Soc., Vol. 122, No. 20, 2000 4901
(6S,11Z,13S)-6-Amino-8-(1′,3′-dioxan-2′-yl)-2-methyl-13-triiso-
propylsiloxypentadeca-2,11-diene (59). Triphenylphosphine (2.89 g,
11.0 mmol) and hydrazoic acid (5.8 mL, 12 mmol, 2.1 M in toluene)46
were added to a 0 °C solution of alcohol 57 (2.65 g, 5.49 mmol) and
THF (55 mL). Diethyl azodicarboxylate (DEAD, 2.6 mL, 16 mmol)
was then added dropwise over a period of 15 min. The resulting solution
was maintained at 0 °C for 1.5 h, then approximately half of the solvent
was removed in vacuo. The resulting solution was diluted with hexanes
(30 mL) and filtered through a plug of silica gel using 97:3 hexanes-
Et2O. The eluent was concentrated, and the crude product was purified
by flash chromatography (97:3 hexanes-Et2O) to give 2.45 g (88%)
of the corresponding azide as a clear oil: [R]25D +9.5, [R]25577 +10.3,
[R]25546 +12.1, [R]25435 +24.1, [R]25405 +31.2 (c 1.6, CHCl3); 1H NMR
(500 MHz, CDCl3) δ 5.41-5.29 (m, 2 H), 5.10 (broad t, J ) 7.1 Hz,
1 H), 4.47 (app q, J ) 7.4 Hz, 1 H), 3.96-3.86 (m, 4 H), 3.71-3.66
(m, 1 H), 2.12-2.07 (m, 3 H), 2.00-1.72 (m, 6 H), 1.70 (s, 3 H), 1.64
(s, 3 H), 1.63-1.42 (m, 5 H), 1.05 (s, 21 H), 0.87 (t, J ) 7.4 Hz, 3 H);
13C NMR (125 MHz, CDCl3) δ 134.6, 132.4, 127.7, 123.3, 99.1, 69.9,
59.6, 59.6, 58.3, 42.2, 36.1, 32.2, 31.7, 25.7, 25.1, 24.7, 22.2, 18.1,
18.1, 17.6, 12.4, 9.4 ppm; IR (film) 2101 cm-1; HRMS (FAB) m/z
506.3776 (M - H, 506.3781 calcd for C28H52N3O3Si). Anal. Calcd for
C28H53N3O3Si: C, 66.22; H, 10.52. Found: C, 66.27; H, 10.50.
sation of â-keto ester 24 with ureido aminal 61 to combine all
carbons of the guanidine nucleus and set the pivotal C10-C13
stereorelationship. The total synthesis of crambescidin 800 (2)
was accomplished in 3% overall yield from commercially
available 3-butyn-1-ol by way of 16 isolated and purified
intermediates. Full details of our earlier total synthesis of
ptilomycalin A (1) are also provided. These total syntheses
confirm the stereochemical assignments of 1, 2, 6, and 7 and
rigorously establish that the absolute configuration of the
hydroxyspermidine side chain of crambescidin 800 (2) is S. The
fully convergent synthesis route detailed herein constitutes a
viable way to prepare the major group of crambescidin alkaloids
that lack oxidation at C13 on multigram scales.
Experimental Section45
(6R,11Z,13S)-2-Methyl-6-triethylsilyloxy-13-triisopropylsiloxy-
pentadeca-2,11-dien-8-one (56). A pentane solution of t-BuLi (23.5
mL, 40 mmol, 1.7 M) was added dropwise to a -78 °C solution of
iodide 54 (6.67 g, 16.8 mmol) and Et2O-hexanes (1:1, 100 mL). The
solution was maintained at -78 °C for 30 min and a solution of amide
55 (6.10 g, 18.5 mmol) and Et2O-hexanes (1:1, 40 mL) was added
dropwise. The resulting solution was maintained at -78 °C for 30 min,
allowed to warm to 0 °C, and maintained at 0 °C for 2 h. The solution
was then added to saturated aqueous NH4Cl (150 mL), the phases were
separated, and the aqueous phase was extracted with Et2O (2 × 150
mL). The combined organic extracts were dried (MgSO4), filtered, and
concentrated. Purification of the residue by flash chromatography (98:2
hexanes-Et2O) gave 5.93 g (65%) of 56 as a clear oil. The product
was ca. 95% pure and was used without further purification in the next
step.
A solution of this azide (2.45 g, 4.82 mmol) and Et2O (18 mL) was
added dropwise to a 0 °C solution of LiAlH4 (12 mL, 12 mmol, 1.0 M
in Et2O) and Et2O (18 mL). The ice bath was removed, and the solution
was allowed to warm to room temperature. After 1 h, the reaction was
quenched by sequential addition of H2O (600 µL), NaOH (600 µL, 3
N), and H2O (1.8 mL). The resulting mixture was stirred for 1 h, MgSO4
was added, the mixture was filtered through Celite, and the eluent was
concentrated to afford a brown oil. Purification of this oil by flash
chromatography (10:1:0.1 CHCl3-MeOH-concentrated NH4OH) gave
2.05 g (88%) of amine 59 as a light yellow oil: [R]25D +21.2, [R]25
577
+22.7, [R]25 +26.1, [R]25 +47.2, [R]25 +58.1 (c 1.6, CHCl3);
546
435
405
A small sample was further purified by flash chromatography (98:2
1H NMR (500 MHz, CDCl3) δ 5.39-5.29 (m, 2 H), 5.11 (br t, J ) 7.1
Hz, 1 H), 4.46 (app q, J ) 7.4 Hz, 1 H), 3.95-3.84 (m, 4 H), 3.15-
3.11 (m, 1 H), 2.10-1.96 (m, 4 H), 1.83-1.69 (m, 4 H), 1.68 (s, 3 H),
1.63-1.31 (m, 6 H), 1.61 (s, 3 H), 1.05 (s, 21 H), 0.86 (t, J ) 7.5 Hz,
3 H); 13C NMR (125 MHz, CDCl3) δ 134.3, 131.4, 127.9, 124.1, 100.4,
69.8, 59.4, 59.2, 46.7, 43.1, 38.8, 32.7, 31.6, 25.6, 25.3, 24.6, 22.1,
18.0, 18.0, 17.6, 12.3, 9.3 ppm; IR (film) 3387, 3310 cm-1; HRMS
(FAB) m/z 482.4011 (M + H, 482.4029 calcd for C28H56NO3Si). Anal.
Calcd for C28H55NO3Si: C, 69.80; H, 11.51. Found: C, 69.85; H, 11.56.
hexanes-Et2O) to obtain an analytical specimen: [R]25D +4.1, [R]25
577
1
+4.8, [R]25 +4.9, [R]25 +11.0, [R]25 +14.3 (c 1.6, CHCl3); H
546
435
405
NMR (400 MHz, CDCl3) δ 5.41-5.36 (m, 1 H), 5.29-5.24 (m, 1 H),
5.08 (tt, J ) 7.1, 1.3 Hz, 1 H), 4.45 (app q, J ) 6.7 Hz, 1 H), 4.18
(quintet, J ) 6.0 Hz, 1 H), 2.60 (A of ABX, JAB ) 15.3, JAX ) 7.2
Hz, 1 H), 2.48-2.43 (m, 3 H), 2.30-2.24 (m, 2 H), 2.05-1.93 (m, 2
H), 1.68 (s, 3 H), 1.64-1.40 (m, 4 H), 1.59 (s, 3 H), 1.04 (s, 21 H),
0.94 (t, J ) 7.9 Hz, 9 H), 0.85 (t, J ) 7.5 Hz, 3 H), 0.58 (q, J ) 7.9
Hz, 6 H); 13C NMR (100 MHz, CDCl3) δ 208.9, 135.1, 131.8, 126.7,
123.8, 69.8, 68.7, 50.2, 44.1, 37.9, 31.6, 25.7, 23.8, 21.9, 18.1, 18.0,
17.6, 12.3, 9.3, 6.9, 4.9 ppm; IR (film) 1717 cm-1; high-resolution
mass spectroscopy (HRMS) fast atom bombardment (FAB) m/z
537.4141 (M - H, 537.4159 calcd for C31H61O3Si2).
(6S,11Z,13S)-8-(1′,3′-Dioxan-2′-yl)-2-methyl-13-triisopropylsiloxy-
6-ureidopentadeca-2,11-diene (60). Trimethylsilyl isocyanate (0.55
mL, 4.1 mmol) was added to a room temperature solution of 59 (1.61
g, 3.35 mmol), CH2Cl2 (6.8 mL), and i-PrOH (0.31 mL). After 15 h,
additional i-PrOH (3 mL) was added and the solution was maintained
for 1 h and then concentrated. The resulting oil was purified on silica
gel (100% EtOAc) to provide 1.57 g (89%) of 60 as a colorless oil:
[R]25 +7.0, [R]25 +12.0, [R]25 +17.3, [R]25 +20.7, [R]25
(6R,11Z,13S)-8-(1′,3′-Dioxan-2′-yl)-6-hydroxy-2-methyl-13-triiso-
propylsiloxypentadeca-2,11-diene (57). A mixture of crude ketone
56 (3.74 g, 6.94 mmol), ortho ester 5837b (4.10 g, 34.7 mmol), 1,3-
propanediol (12.6 mL, 174 mmol), Amberlyst-15 resin (278 mg), and
MeCN (70 mL) was maintained at room temperature for 7 h. The
mixture was then filtered through Celite and the filtrate was partitioned
between Et2O (150 mL) and H2O (50 mL). The organic phase was
washed with H2O (2 × 50 mL), dried (MgSO4), filtered, and the filtrate
was concentrated. Purification of the residue by flash chromatography
D
577
546
435
405
1
+25.4 (c 1.0, CHCl3); H NMR (400 MHz, CDCl3) δ 5.24-5.36 (m,
2H), 5.03-5.15 (m, 4H), 4.41 (dd, J ) 13.2, 7.1 Hz, 1H), 3.80-3.91
(m, 4H), 3.64 (m, 1H), 1.71-2.03 (m, 8H), 1.63 (s, 3H), 1.55 (s, 3H),
1.36-1.63 (m, 6H), 1.00 (s, 21H), 0.82 (t, J ) 7.4 Hz, 3H); 13C NMR
(100 MHz, CDCl3) 159.3, 134.4, 131.8, 127.5, 123.7, 99.9, 69.7, 59.4,
59.2, 46.7, 36.9, 31.5, 31.1, 25.6, 25.0, 24.5, 22.1, 17.9, 17.8, 17.5,
12.2, 9.2 ppm; IR (film) 3354, 1660, 1600 cm-1. Anal. Calcd for
C29H56N2O4Si: C, 66.36; H, 10.75; N, 5.34. Found: C, 66.31; H, 10.70;
N 5.41.
(85:15 hexanes-Et2O) gave 2.68 g (80%) of ketal 57 as clear oil: [R]25
D
+13.3, [R]25 +14.2, [R]25 +16.8, [R]25 +30.1, [R]25 +37.4
577
546
435
405
1
(c 1.6, CHCl3); H NMR (500 MHz, CDCl3) δ 5.42-5.29 (m, 2 H),
5.14 (broad t, J ) 7.1 Hz, 1 H), 4.45 (app q, J ) 7.5 Hz, 1 H), 4.11-
4.08 (m, 1 H), 4.02-3.85 (m, 4 H), 3.80 (s, 1 H), 2.16-1.96 (m, 6 H),
1.84-1.76 (m, 2 H), 1.68 (s, 3 H), 1.65-1.36 (m, 6 H), 1.61 (s, 3 H),
1.05 (s, 21 H), 0.86 (t, J ) 7.4 Hz, 3 H); 13C NMR (125 MHz, CDCl3)
δ 134.6, 131.5, 127.5, 124.3, 101.1, 69.9, 67.0, 59.5, 59.5, 43.7, 37.5,
31.7, 31.3, 25.7, 25.2, 24.1, 22.3, 18.1, 18.0, 17.6, 12.4, 9.3 ppm; IR
(film) 3532 cm-1; HRMS (FAB) m/z 505.3683 (M + Na, 505.3691
calcd for C28H54O4SiNa). Anal. Calcd for C28H54O4Si: C, 69.65; H,
11.27. Found: C, 69.40; H, 11.28.
(4aR,7S)-4-[15-(Allyloxycarbonyl)pentadecyloxycarbonyl]-1,2,
4a,5,6,7-hexahydro-3-[(4S)-tert-butyldimethylsiloxypentyl)]-7-[(7S,5Z)-
2-(1′,3′-dioxan-2′-yl)-7-triisopropylsiloxynon-5-enyl)]-1-oxo-pyrrolo-
[1,2-c]pyrimidine (62). Osmium tetroxide (0.75 mL, 0.1 M in t-BuOH)
was added to a solution of 60 (524 mg, 1.00 mmol), N-methylmor-
pholine-N-oxide (NMO, 406 mg, 3.46 mmol), and 10:1 THF-H2O (25
mL). After 1.5 h, Florisil (3 g), NaHSO3 (3 g), and EtOAc (50 mL)
were added and the reaction mixture was stirred vigorously. After 30
min, this mixture was filtered, and the filtrate was concentrated to
(45) General experimental details have been described: Metais, E.;
Overman, L. E.; Rodriguez, M. I.; Stearns, B. A. J. Org. Chem. 1997, 62,
9210-9216. Additional details are presented in Supporting Information.
(46) Hydrazoic acid was generated by the procedure reported in: Org.
React. 1946, 3, 327.

4902 J. Am. Chem. Soc., Vol. 122, No. 20, 2000
Coffey et al.
provide the corresponding diol, a colorless oil which was used without
further purification.
Calcd for C41H68N2O8: C, 68.68; H, 9.56; N, 3.91. Found: C, 68.71;
H, 9.51; N 3.84.
(3R,4R,4aR,6′R,7S)-4-[15-(Allyloxycarbonyl)pentadecyloxy-
carbonyl]-1,2,4a,5,6,7-hexahydro-1-oxo-7-[(7S,5Z)-7-chloroacetoxy-
2-oxo-5-nonenyl]-pyrrolo[1,2-c]pyrimidine-3-spiro-6′-(2′-methyl)-
3′,4′,5′,6′-tetrahydro-2H-pyran (65). Chloroacetyl chloride (0.34 mL,
0.46 mmol) was added dropwise to a 0 °C solution of 64 (0.63 g, 0.88
mmol, containing ∼12% of the C4a S epimer), pyridine (1.4 mL, 18
mmol), and CH2Cl2 (50 mL). The solution was allowed to warm to
room temperature, and after 1 h, was quenched by adding Et2O (200
mL). This solution was washed with 1 N NaOH (25 mL), CuSO4 (2 ×
25 mL), and brine (25 mL), dried (MgSO4), filtered, and the filtrate
was concentrated. The resulting residue was purified on silica gel (2:1
hexanes-EtOAc; 1:1 hexanes-EtOAc; 1:2 hexanes-EtOAc) to yield
600 mg (86%) of isomerically pure 65 as a colorless oil, and ∼85 mg
A solution of this crude diol, Pb(OAc)4 (532 mg, 1.20 mmol), and
toluene (60 mL) was maintained at room temperature for 30 min. The
reaction mixture was filtered through a plug of Celite, morpholinium
acetate (300 mg, 2.0 mmol) was added, and the solution was
concentrated to provide the crude aminal 61 as a slightly yellow oil.
A solution of this crude aminal, 24 (1.95 g, 3.36 mmol), and 2,2,2-
trifluoroethanol (1.0 mL) was maintained at 60 °C for 2 days. The
reaction was then quenched by adding Et2O (20 mL) and 50% aqueous
NH4Cl (5 mL). The layers were separated, the organic layer was dried
(MgSO4), concentrated, and the resulting oil was purified on silica gel
(10:1 hexanes-EtOAc; 7:1 hexanes-EtOAc; 3:1 hexanes-EtOAc) to
provide 1.5 g of 24 and 638 mg (61%) of a ∼6.5:1 mixture of 62 and
63, respectively, which was carried forward without separation.
For characterization purposes, a 50 mg sample of this mixture was
separated by HPLC (7:1 hexanes-EtOAc; Altima 5 µm silica) to give
(∼12%) of the C4a S isomer that was derived from 61. 65: [R]25
D
+42.7, [R]25 +47.0, [R]25 +52.6, [R]25 +96.1, [R]25 +120,
577
546
435
405
1
a pure sample of 62: [R]25D -4.5, [R]25577 -4.9, [R]25546 -5.7, [R]25
(c 1.00, CHCl3); H NMR (500 MHz, CDCl3) δ 6.34 (s, 1H), 5.87-
5.94 (m, 1H), 5.48-5.56 (m, 2H), 5.27-5.32 (m, 2H), 5.22 (d, J )
10.4 Hz, 1H), 4.56 (d, J ) 5.7 Hz, 2H), 4.31-4.33 (m, 1H), 4.09-
4.19 (m, 2H), 4.03 (s, 2H), 4.00-4.06 (m, 1H), 3.77-3.81 (m, 1H),
3.34 (d, J ) 16.6 Hz, 1H), 2.40-2.48 (m, 3H), 2.25-2.38 (m, 5H),
2.05-2.17 (m, 3H), 1.69-1.74 (m, 4H), 1.55-1.62 (m, 7H), 1.42-
1.50 (m, 1H), 1.24-1.31 (m, 22H), 1.06-1.15 (m, 1H), 1.05 (d, J )
6.0 Hz, 3H), 0.89 (t, J ) 7.5 Hz, 3H); 13C NMR (125 MHz, CDCl3)
207.9, 173.4, 168.8, 166.6, 153.4, 133.0, 132.3, 127.8, 118.0, 82.1,
73.7, 66.1, 64.9, 64.8, 54.9, 53.9, 53,1, 46.3, 42.3, 41.1, 34.2, 32.2,
32.0, 29.5, 29.49, 29,4, 29.3, 29.2, 29.1, 29.07, 29.0, 28.6, 27.4, 25.9,
21.8, 21.6, 18.5, 9.3 ppm;47 IR (film) 3296, 2928, 2855, 1732, 1652
cm-1. Anal. Calcd for C43H69N2O9Cl: C, 65.09; H, 8.77; N, 3.53.
Found: C, 65.16; H, 8.79; N 3.57.
435
-15.5, [R]25405 -22.7 (c 0.75, CHCl3); 1H NMR (500 MHz, CDCl3) δ
6.72 (s, 1H), 5.87-5.95 (m, 1H), 5.21-5.37 (m, 4H), 4.56 (d, J ) 5.7
Hz, 2H), 4.51 (dd, J ) 12.7, 7.1 Hz, 1H), 4.22 (dd, J ) 11.0, 4.6 Hz,
1H), 4.06-4.13 (m, 3H), 3.97-3.98 (m, 1H), 3.76-3.88 (m, 4H), 2.47-
2.58 (m, 3H), 2.39 (d, J ) 13.6 Hz, 1H), 2.32 (t, J ) 7.5 Hz, 2H),
2.26-2.32 (m, 1H), 2.15 (dd, J ) 13.0, 6.0 Hz, 1H), 1.99-2.03 (m,
1H), 1.50-1.90 (m, 13H), 1.41-1.48 (m, 3H), 1.11-1.40 (m, 23H),
1.10 (d, J ) 6.1 Hz, 3H), 0.91-1.07 (m, 21H), 0.82-0.91 (m, 12H),
0.03 (s, 3H), 0.02 (s, 3H); 13C NMR (125 MHz, CDCl3) 173.5, 166.0,
151.9, 151.2, 134.3, 132.2, 128.2, 118.0, 102.1, 99.2, 69.9, 68.3, 64.9,
64.2, 59.3, 57.7, 52.7, 39.0, 37.4, 34.5, 34.2, 31.8, 31.3, 30.4, 29.6,
29.57, 29.5, 29.4, 29.3, 29.2, 29.1, 29.0, 26.1, 25.9, 25.3, 24.9, 24.4,
23.6, 21.8, 18.1, 12.3, 9.3, -4.4, -4.7 ppm;47 IR (film) 3211, 1741,
1682, 1627 cm-1. Anal. Calcd for C59H108N2O9Si2: C, 67.77; H, 10.41;
N, 2.68. Found: C, 67.68; H, 10.27; N 2.65.
(3R,4R,4aR,6′R,7S)-4-[15-(Allyloxycarbonyl)pentadecyloxy-
carbonyl]-1,2,4a,5,6,7-hexahydro-1-oxo-7-[(7S,5Z)-7-hydroxy-2-oxo-
5-nonenyl]-pyrrolo[1,2-c]pyrimidine-3-spiro-6′-(2′-methyl)-3′,4′,5′,6′-
tetrahydro-2H-pyran (64). A solution of the 62/63 mixture (1.30 g,
1.24 mmol), TBAF (6.2 mL, 1.0 M solution in Et2O), and DMF (31
mL) was maintained at room temperature for 5 h, the solution was
diluted with Et2O (150 mL), and washed with H2O (50 mL) and brine
(2 × 50 mL). The organic layer was dried (MgSO4), filtered, and the
filtrate was concentrated to give a residue that was used without further
purification.
Pentacycles 67 and 68. A solution of 65 (327 mg, 0.412 mmol),
MeOTf (1.29 mL, 8.21 mmol), 2,6-di-tert-butylpyridine (0.46 mL, 2.1
mmol), and CH2Cl2 (20 mL) was maintained at room temperature for
8 h. The solution was then poured into Et2O (100 mL) and washed
with 1 N NaOH (2 × 10 mL) and brine (10 mL). The organic layer
was dried (MgSO4), filtered, and the filtrate concentrated. The resulting
crude pseudourea 66 was used without further purification.
Ammonia was bubbled through a solution of this sample of 66,
NH4Cl (50 mg, 0.93 mmol), and allyl alcohol (5 mL) at room
temperature until the solution was saturated (∼20 min). The reaction
vessel was sealed and heated to 60 ° C for 28 h. The reaction was then
cooled to room temperature, concentrated, and the resulting oil was
purified by silica gel medium-pressure liquid chromatography (MPLC)
(100:0.6 CHCl3-i-PrOH) to provide 147 mg of 67 and 98 mg of 68.
Resubjecting 67 to these reaction conditions (2×), followed by
chromatographic separation provided an additional 60 mg of 68 (52%
combined yield of 68).
A solution of this crude diol, TsOH‚H2O (236 mg, 1.24 mmol), and
CHCl3 (180 mL) was maintained at 60 °C for 15 min. The reaction
was quenched by adding saturated aqueous NaHCO3 (20 mL), the layers
were separated, and the organic layer was washed with brine (20 mL).
The organic layer was dried (MgSO4), concentrated, and the resulting
oil was purified on silica gel (1:3 hexanes-EtOAc; 100% EtOAc) to
provide 630 mg (71%) of 64 as a ∼6.5:1 mixture of epimers. This
mixture of stereoisomers was not separated, but directly used in the
next step.
67: [R]25D +12.2, [R]25577 +13.1, [R]25546 +14.1 (c 2.0, CHCl3); 1H
NMR (500 MHz, CDCl3) δ 8.68 (s, 1H), 8.56 (s, 1H), 5.88-5.95 (m,
1H), 5.64-5.67 (m, 1H), 5.48 (d, J ) 10.9 Hz, 1H), 5.33 (dd, J )
17.2, 1.5 Hz, 1H), 5.25 (dd, J ) 10.4, 1.2 Hz, 1H), 4.57 (d, J ) 5.7
Hz, 2H), 4.48 (d, J ) 10.3 Hz, 1H), 4.32-4.38 (m, 1H), 4.10-4.24
(m, 3H), 3.78-3.81 (m, 1H), 2.56-2.61 (m, 2H), 2.45 (d, J ) 11.6
Hz, 1H), 2.32 (t, J ) 7.6 Hz, 2H), 2.26-2.36 (m, 3H), 2.15-2.18 (m,
2H), 2.00 (dt, J ) 13.8, 4.7 Hz, 1H), 1.87 (dd, J ) 14.6, 5.4 Hz, 1H),
1.61-1.78 (m, 10H), 1.53-1.58 (m, 1H), 1.42-1.49 (m, 1H), 1.23-
1.35 (m, 22H), 1.05-1.15 (m, 1H), 1.05 (d, J ) 6.1 Hz, 3H), 0.81 (t,
J ) 7.2 Hz, 3H); 13C NMR (125 MHz, CDCl3) 173.5, 167.6, 147.3,
133.4, 132.3, 129.7, 118.0, 83.9, 81.7, 70.9, 67.6, 65.7, 64.9, 53.4, 53.3,
36.8, 36.2, 34.2, 31.8, 30.6, 29.7, 29.6, 29.5, 29.46, 29.4, 29.2, 29.1,
29.08, 29.0, 28.5, 25.9, 24.9, 23.6, 21.3, 17.9, 10.1 ppm;47 IR (film)
3268, 1732, 1660, 1608 cm-1; HRMS (FAB) m/z 698.5096 (M - Cl,
698.5108 calcd for C41H68N3O6).
A pure sample of 64 was obtained by HPLC (7:1 hexanes-EtOAc;
Altima 5 µm silica): [R]25 +42.2, [R]25 +42.7, [R]25 +49.8,
D
577
546
[R]25 +91.0, [R]25 +114 (c 0.6, CHCl3); 1H NMR (500 MHz,
435
405
CDCl3) δ 5.87-5.95 (m, 1H), 5.56 (s, 1H), 5.34-5.43 (m, 2H), 5.31
(dd, J ) 17.2, 1.5 Hz, 1H), 5.22 (dd, J ) 10.6, 1.3 Hz, 1H), 4.57 (dd,
J ) 4.3, 1.3 Hz, 2H), 4.38 (dd, J ) 14.5, 6.8 Hz, 1H), 4.29-4.31 (m,
1H), 4.08-4.18 (m, 2H), 4.02 (dt, J ) 11.1, 4.8 Hz, 1H), 3.77-3.80
(m, 1H), 3.37 (d, J ) 16.8 Hz, 1H), 2.52-2.60 (m, 2H), 2.43-2.50
(m, 1H), 2.32 (t, J ) 7.5 Hz, 2H), 2.22-2.27 (m, 2H), 2.04-2.20 (m,
4H), 1.69-1.76 (m, 4H), 1.56-1.65 (m, 7H), 1.42-1.48 (m, 3H), 1.24-
1.28 (m, 21H), 1.06-1.09 (m, 1H), 1.05 (d, J ) 6.0 Hz, 3H), 0.89 (t,
J ) 7.5 Hz, 3H); 13C NMR (125 MHz, CDCl3) 209.0, 173.5, 168.9,
153.0, 134.1, 132.3, 129.8, 118.1, 82.2, 68.4, 66.2, 65.1, 64.9, 55.0,
54.0, 53.2, 46.2, 42.7, 34.3, 32.2, 32.1, 30.3, 30.0, 29.6, 29.57, 29.5,
29.4, 29.3, 29.2, 29.1, 28.7, 26.0, 24.9, 22.0, 21.7, 18.8 ppm;47 IR (film)
3450, 3231, 3081, 2927, 2855, 1732, 1715, 1659, 1651 cm-1. Anal.
68: [R]25D -9.6, [R]25577 -10.5, [R]25546 -9.5, [R]25435 -16.5, [R]25
405
1
-17.2 (c 0.75, CHCl3); H NMR (500 MHz, CDCl3) δ 8.54 (s, 1H),
8.43 (s, 1H), 5.88-5.95 (m, 1H), 5.64-5.67 (m, 1H), 5.48 (d, J )
10.9 Hz, 1H), 5.31 (dd, J ) 17.2, 1.5 Hz, 1H), 5.22 (dd, J ) 10.4, 1.2
Hz, 1H), 4.57 (dd, J ) 5.7, 1.2 Hz, 2H), 4.48 (d, J ) 9.7 Hz, 1H),
4.29-4.33 (m, 1H), 4.08 (t, J ) 6.8 Hz, 2H), 3.99-4.05 (m, 1H), 3.84-
3.87 (m, 1H), 2.93 (d, J ) 4.8 Hz, 1H), 2.55-2.63 (m, 2H), 2.32 (t, J
(47) 13C NMR signals of many of the methylene carbons of the
hexadecanoate side chain overlap.

Crambescidin Family of Guanidine Alkaloids
J. Am. Chem. Soc., Vol. 122, No. 20, 2000 4903
) 7.6 Hz, 2H), 2.26-2.36 (m, 2H), 2.13-2.24 (m, 3H), 1.98 (dd, J )
14.7, 5.3 Hz, 1H), 1.78-1.84 (m, 1H), 1.51-1.76 (m, 10H), 1.38-
1.48 (m, 2H), 1.21-1.30 (m, 22H), 1.07-1.20 (m, 1H), 1.05 (d, J )
6.1 Hz, 3H), 0.81 (t, J ) 7.2 Hz, 3H); 13C NMR (125 MHz, CDCl3)
173.5, 168.3, 148.4, 133.5, 132.3, 129.7, 118.0, 84.0, 80.8, 70.9, 67.2,
65.5, 64.9, 54.1, 52.2, 49.9, 36.7, 36.1, 34.4, 31.8, 31.7, 30.5, 29.6,
29.56, 29.5, 29.45, 29.4, 29.2, 29.1, 29.06, 28.4, 26.7, 25.8, 24.9, 23.6,
21.5, 17.7, 10.0 ppm;47 IR (film) 3263, 1732, 1652, 1614, cm-1; HRMS
(FAB) m/z 698.5106 (M - Cl, 698.5108 calcd for C41H68N3O6).
Carboxylic Acid Hydrochloride 69 and Crambescidin 657 (6):
A solution of 68 (27 mg, 37 µmol), Pd(PPh3)4 (21 mg, 18 µmol),
morpholine (13 µL, 0.15 mmol), and MeCN (1.0 mL) was maintained
at room temperature for 5 h. The solution was diluted with Et2O (30
mL), and washed with 0.1 N HCl (2 × 5 mL) and brine (5 mL). The
organic layer was dried (MgSO4), filtered, and the filtrate was
concentrated. The resulting residue was purified on silica gel (100:1
CHCl3-MeOH; 33:1 CHCl3-MeOH) to yield 24 mg (94%) of 69 as
a colorless oil: [R]25D 15.2 (c 0.5, MeOH); 1H NMR (500 MHz, CDCl3)
δ 5.63-5.66 (m, 1H), 5.46-5.49 (m, 1H), 4.48 (br d, J ) 10.2 Hz,
1H), 4.27-4.31 (m, 1H), 4.04-4.12 (m, 2H), 3.96-4.03 (m, 1H), 3.85-
3.88 (m, 1H), 2.92 (d, J ) 4.9 Hz, 1H), 2.62 (t, J ) 13.8 Hz, 1H),
2.55 (dd, J ) 12.7, 4.7 Hz, 1H), 2.12-2.32 (m, 7H), 1.86 (dd, J )
14.8, 5.3 Hz, 1H), 1.77-1.81 (m, 1H), 1.60-1.73 (m, 9H), 1.51-1.59
(m, 1H), 1.37-1.45 (m, 2H), 1.20-1.30 (m, 22H), 1.16-1.20 (m, 1H),
1.04 (d, J ) 6.1 Hz, 3H), 0.80 (t, J ) 7.2 Hz, 3H), the NH and OH
signals were not observed; 13C NMR (125 MHz, CDCl3) 179.1, 168.4,
148.7, 133.6, 129.8, 83.9, 80.8, 70.8, 67.0, 65.4, 54.0, 52.0, 50.0, 36.7,
36.0, 31.9, 31.8, 30.5, 29.4, 29.3, 29.2, 29.1, 29.0, 28.4, 26.7, 25.8,
25.5, 23.7, 21.5, 17.8, 10.0 ppm;47 IR (film) 3261, 3138, 2919, 2849,
1728, 1658 cm-1; HRMS (FAB) m/z 658.4791 (M - Cl, 658.4795
calcd for C38H64N3O6).
2.26-2.46 (m, 6H), 2.10-2.20 (m, 1H), 2.00 (dd, J ) 13.9, 5.8 Hz,
1H), 1.79-1.85 (m, 3H), 1.50-1.77 (m, 11H), 1.36-1.47 (m, 20H),
1.22-1.35 (m, 25H), 1.09 (d, J ) 6.1 Hz, 3H), 0.85 (t, J ) 6.1 Hz,
3H); 13C NMR (125 MHz, d4-MeOD) 176.6/176.2, 170.2, 158.5, 150.2,
134.3, 131.3, 85.1, 82.2, 80.0, 79.96, 72.3, 69.1, 68.4, 68.37, 66.5, 55.6,
55.0, 54.2, 53.5, 50.7, 45.1, 38.9, 38.7, 38.3, 38.1, 37.9, 36.2, 34.3,
34.1, 33.0, 32.6, 31.5, 30.8, 30.7, 30.6, 30.5, 30.3, 30.2, 29.6, 28.8,
28.7, 27.6, 27.0, 26.7, 26.6, 24.4, 21.8, 19.5, 10.8 ppm;47 IR (film)
3356, 1732, 1706, 1657, 1613 cm-1; HRMS (FAB) m/z 1001.7 (M -
Cl, 1001.7 calcd for C55H97N6O10).
Crambescidin 800 Trihydrochloride (2). A solution of 71 (13 mg,
13 µmol) and 1.3 mL of a 3.0 M solution of HCl in EtOAc was
maintained at room temperature for 20 min and then concentrated.
Purification of the residue by reversed-phase HPLC (4:1 MeOH-0.1
M NaCl, Altima C18, 5 µm column) gave 11 mg (75%) of crambescidin
800 (2) as its trihydrochloride salt (a light yellow oil): [R]25 -4.4,
D
[R]25577 -5.0, [R]25546 -4.0, [R]25435 -6.3, [R]25405 -6.2, (c 0.7, CHCl3).
Spectroscopic and mass spectrometric data for this sample were
consistent with data published for natural 2.4a,5a,43
Peracetylcrambescidin 800 Chloride (72). A solution of crambes-
cidin 800 (2) (5.0 mg, 5.5 µg), Ac2O (0.5 mL), and pyridine (1 mL)
was maintained at room temperature for 23 h and then concentrated
(0.9 mm, 23 °C).5a The residue was diluted with CHCl3 (20 mL) and
washed with 0.1 M HCl (5 mL), and brine (5 mL). The organic layer
was dried (MgSO4), filtered, and the filtrate was concentrated. The
resulting residue was purified on silica gel (20:1 CHCl3-MeOH; 10:1
CHCl3-MeOH) to yield 2 mg (35%) of peracetylcrambescidin 800
(72) as a colorless wax: [R]25 -37 (c 0.2, CHCl3). NMR and mass
D
spectrometric data for this sample were consistent with published data.5a
Acknowledgment. This research was supported by a grant
from NIH NHLBIS (HL-25854), through an NIH Postdoctoral
Fellowship to D.S.C. (CA-75616), and a graduate fellowship
to A.I.M. from Pfizer. NMR and mass spectra were determined
at UCI using instruments acquired with the assistance of NSF
and NIH shared instrumentation grants. We thank Professor
Kenneth Rinehart for providing a sample of natural 2, Professor
Kashman for providing a sample of natural ptilomycalin A, and
Professor K. N. Houk for stimulating discussions concerning
torsional steering in the spirocyclization steps.
Carboxylic acid 69 was quantitatively converted to the carboxylate
inner salt by washing a CHCl3 (5 mL) solution of the acid (5 mg) with
1 N NaOH (1 mL) and brine (1 mL). The organic layer was dried
(Na2SO4) and then concentrated to provide 6 as a colorless oil: [R]25
D
-13.6 (c 0.45, MeOH). Spectroscopic and mass spectrometric data for
this sample were consistent with data published for natural 6.6a
41,45-di-(tert-Butoxycarbonyl)crambescidin 800 (71). Benzotri-
azol-1-yloxytris(dimethylamino)phosphonium hexafluorophosphate (22
mg, 50 µmol) was added to a room temperature solution of carboxylic
acid 69 (23 mg, 33 µmol), amine 70 (18 mg, 50 µmol),40 Et3N (0.15
mL, 1.1 mmol), and CH2Cl2 (5 mL). After 4 h, the reaction was diluted
with Et2O (20 mL), and washed with saturated aqueous NH4Cl (5 mL)
and brine (5 mL). The organic layer was dried (MgSO4), filtered, and
the filtrate was concentrated. The resulting residue was purified on silica
gel (50:1 CHCl3-MeOH) to yield 28 mg (82%) of 71 as a colorless
oil: [R]25 -3.0, [R]25 -2.2, [R]25 -2.8, [R]25 -3.5, [R]25
Supporting Information Available: Characterization data
and detailed experimental procedures for the preparation of
ptilomycalin A (1), neofolitispates 2 (7), 22-33, 38-41, 43-
45 and 48-55. 1H and 13C NMR spectra for 1, 2, 6, 7, 45, and
72; 19F NMR spectra for synthetic 73 and natural 73. This
material is available free of charge via the Internet at
http://pubs.acs.org.
D
577
1
546
435
405
-3.6, (c 0.75, CHCl3); H NMR (500 MHz, d4-MeOD) δ 5.70-5.73
(m, 1H), 5.47-5.52 (m, 1H), 4.40 (br d, J ) 10.3 Hz, 1H), 4.33-4.37
(m, 1H), 4.10-4.16 (m, 2H), 4.02-4.09 (m, 1H), 3.75-3.85 (m, 2H),
3.34-3.59 (m, 2H), 3.23-3.29 (m, 2H), 3.12-3.20 (m, 2H), 3.07 (d,
J ) 4.8 Hz, 1H), 2.94-3.06 (m, 2H), 2.64 (dd, J ) 13.0, 4.7 Hz, 1H),
JA000234I